Peripheral Neuropathy: 2-Volume Set

The Curtis Center 170 S Independence Mall W 300E Philadelphia, Pennsylvania 19106

PERIPHERAL NEUROPATHY Copyright © 2005, Elsevier Inc. All rights reserved.

ISBN 0–7216–9491–8 (2 Volume Set) (Volume 1) 9997637887 (Volume 2) 9997637895

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permissions may be sought directly from Elsevier’s Health Sciences Rights Department in Philadelphia, PA, USA: phone: (⫹1) 215 238 7869, fax: (⫹1) 215 238 2239, e-mail: [email protected]. You may also complete your request online via the Elsevier homepage (http://www.elsevier.com), by selecting ‘Customer Support’ and then ‘Obtaining Permissions.’ Exceptions to copyright are as follows: Artwork for Chapters 13, 25, 27, 28, 37, 43, 47, 99, 100, 108, 116, and 121.

NOTICE Neurology is an ever-changing field. Standard safety precautions must be followed, but as new research and clinical experience broaden our knowledge, changes in treatment and drug therapy may become necessary or appropriate. Readers are advised to check the most current product information provided by the manufacturer of each drug to be administered to verify the recommended dose, the method and duration of administration, and contraindications. It is the responsibility of the licensed prescriber, relying on experience and knowledge of the patient, to determine dosages and the best treatment for each individual patient. Neither the publisher nor the author assumes any liability for any injury and/or damage to persons or property arising from this publication. THE PUBLISHER First edition 1975. Second edition 1984. Third edition 1993.

Library of Congress Cataloging-in-Publication Data Peripheral neuropathy / [edited by] Peter James Dyck, P. K. Thomas.—4th ed. p. ; cm. Includes bibliographical references and index. ISBN 0–7216–9491–8 (set) 1. Nerves, Peripheral–Diseases. I. Dyck, Peter James, II. Thomas, P. K. (Peter Kynaston). [DNLM: 1. Peripheral Nervous System Diseases. 2. Peripheral Nerves—pathology. WL 500 P445 2005] RC409.P46 2005 616.8⬘5607–dc22 2004051387 Publishing Director, Global Medicine: Susan Pioli Developmental Editor: Jennifer Ehlers Editorial Assistant: Joan Ryan Designer: Gene Harris Marketing Manager: Michael Passanante

Working together to grow libraries in developing countries www.elsevier.com | www.bookaid.org | www.sabre.org

Printed in the United States of America Last digit is the print number:

9

8

7

6

5

4

3

2

1

Contributing Authors

AMMAR AL-CHALABI, PH.D., F.R.C.P. Senior Lecturer in Neurology, Institute of Psychiatry, London, United Kingdom; Instructor in Complex Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York; Visiting Scientist in Neurology, Massachusetts General Hospital, Boston, Massachusetts; Consultant Neurologist, King’s College Hospital, London,United Kingdom Sporadic Motor Neuron Degeneration; Inherited Motor Neuron Degeneration

DORIS-EVA BAMIOU, M.D., M.SC.(DISTINCTION) Clinical Research Fellow, Neuro-otology Department, The National Hospital for Neurology and Neurosurgery, London, United Kingdom Diseases of the Eighth Cranial Nerve

ROBERT W. BANKS, PH.D., D.SC. Lecturer, School of Biological and Biomedical Sciences, University of Durham, Durham, United Kingdom The Muscle Spindle

RICHARD J. BAROHN, M.D. Professor and Chairman, Department of Neurology, and Professor of Pathology, University of Kansas Medical Center, Kansas City, Kansas Polyneuropathy Caused by Nutritional and Vitamin Deficiency

TIMOTHY J. BENSTEAD, M.D., F.R.C.P.(C) Professor, Division of Neurology, Dalhousie University; QEII Health Sciences Centre, Halifax, Nova Scotia, Canada Differential Diagnosis of Polyneuropathy

ALAN R. BERGER, M.D. Professor and Associate Chairman, Department of Neurology, University of Florida; Director, Neuroscience Institute, Shands Jacksonville, Jacksonville, Florida Human Toxic Neuropathy Caused by Industrial Agents

C.-H. BERTHOLD, PH.D., M.D. Professor Emeritus, Section of Neuroanatomy, Department of Anatomy and Cell Biology, Göteborg University, Göteborg, Sweden Microscopic Anatomy of the Peripheral Nervous System

ADIL E. BHARUCHA, M.D. Associate Professor of Medicine, Mayo Clinic College of Medicine; Consultant in Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota Autonomic and Somatic Systems to the Anorectum and Pelvic Floor; Management of Gut Dysmotility

ROLFE BIRCH, M.A., M.B., B.CHIR., F.R.C.S.EDIN., F.R.C.S.&P.GLAS., F.R.C.S.ENG., M.CHIR. (Personal Chair) Professor of Neurological Orthopaedic Surgery, University College, London; Visiting Professor, Imperial College, London; Orthopaedic Surgeon, and Head of Department, Peripheral Nerve Injury Unit, Royal National Orthopaedic Hospital, Stanmore, Middlesex, United Kingdom Operating on Peripheral Nerves

HERBERT L. BONKOVSKY, M.D. Professor, and Director, The Liver-Biliary-Pancreatic Center, Department of Medicine, Biochemistry & Molecular Biology, University of Massachusetts, Worcester, Massachusetts Porphyric Neuropathy

AUGUST M. BOOTH, PH.D. Research Assistant Professor Emeritus, Department of Pharmacology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania Autonomic Systems to the Urinary Bladder and Sexual Organs

E. PETER BOSCH, M.D. Professor of Neurology, Mayo Medical School, Rochester, Minnesota; Consultant, Department of Neurology, Mayo Clinic Scottsdale, Scottsdale, Arizona Peripheral Neuropathy Associated with Lymphoma, Leukemia, and Myeloproliferative Disorders

v

vi

Contributing Authors

HUGH BOSTOCK, M.SC., PH.D. Professor of Neurophysiology, Institute of Neurology, University College London, London, United Kingdom Nerve Excitability Measures: Biophysical Basis and Use in the Investigation of Peripheral Nerve Disease

FRANK BRADKE, PH.D. Independent Principal Investigator on Associate Professor Level, Max Planck Institute of Neurobiology, Munich, Germany Guidance of Axons to Targets in Development and in Disease

ROSCOE O. BRADY, M.D. Branch Chief, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland Fabry’s Disease

STEPHEN BRIMIJOIN, PH.D. Professor of Pharmacology, Mayo Clinic College of Medicine, Rochester, Minnesota Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease

DEBORAH BUCK, PH.D. Senior Research Associate, Department of Primary Care, University of Liverpool, Liverpool, United Kingdom Health Outcomes and Quality of Life

RICHARD P. BUNGE, M.D.* Professor of Neurological Surgery, Cell Biology and Anatomy, and Neurology; Scientific Director, The Miami Project to Cure Paralysis; University of Miami School of Medicine, Miami, Florida Gross Anatomy of the Peripheral Nervous System

DAVID BURKE, M.D., D.SC. Dean of Research and Development, College of Health Sciences, University of Sydney, Sydney, New South Wales, Australia Nerve Excitability Measures: Biophysical Basis and Use in the Investigation of Peripheral Nerve Disease

JAMES P. BURKE, PH.D. Assistant Professor of Epidemiology, Mayo Medical School, Mayo Clinic, Rochester, Minnesota

MICHAEL CAMILLERI, M.D., M.PHIL. (LOND.), F.R.C.P.(LOND.), F.R.C.P.(EDIN.), F.A.C.P., F.A.C.G. Atherton and Winifred W. Bean Professor, Professor of Medicine and Physiology, Mayo Medical School; Consultant in Gastroenterology, Mayo Clinic, Rochester, Minnesota Management of Gut Dysmotility

J. AIDAN CARNEY, M.D., PH.D. Professor of Pathology (Emeritus), Mayo Medical School; Consultant in Pathology (Emeritus), Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, Minnesota Multiple Endocrine Neoplasia, Type 2B

COLIN CHALK, M.D., C.M., F.R.C.P.(C.) Associate Professor, Department of Neurology and Neurosurgery, McGill University; Associate Physician, Division of Neurology, Department of Medicine, McGill University Health Centre, Montréal General Hospital site, Montréal, Québec, Canada Diseases of Spinal Roots

PHILLIP F. CHANCE, M.D. Professor of Pediatrics and Neurology, Division Head, Genetics and Development, Research Affiliate, Center on Human Development and Disability, and Coordinator, Research Emphasis Area on Joubert Syndrome, University of Washington, Seattle, Washington Hereditary Motor and Sensory Neuropathies: An Overview of Clinical, Genetic, Electrophysiologic, and Pathologic Features; Hereditary Motor and Sensory Neuropathies Involving Altered Dosage or Mutation of PMP22: The CMT1A Duplication and HNPP Deletion

S. Y. CHIU, PH.D. Professor of Physiology, University of Wisconsin, Madison, Wisconsin Channel Function in Mammalian Axons and Support Cells

MICHAEL P. COLLINS, M.D. Neuromuscular Diseases Consultant, Neurosciences Department, Marshfield Clinic, Marshfield, Wisconsin Neuropathies with Systemic Vasculitis

Epidemiologic Approaches to Peripheral Neuropathy

TED M. BURNS, M.D. Assistant Professor, Department of Neurology, University of Virginia, Charlottesville, Virginia Mechanisms of Acute and Chronic Compression Neuropathy; Peripheral Neuropathies in Infants and Children: Polyneuropathies, Mononeuropathies, Plexopathies, and Radiculopathies

*Deceased.

JOHN H. COOTE, PH.D., D.SC. Professor of Physiology (Emeritus), University of Birmingham, Birmingham, United Kingdom Neural Control of Cardiac Function

JAMES J. CORBETT, M.D. McCarty Professor and Chairman of Neurology, and Professor of Ophthalmology, University of Mississippi School of Medicine, Jackson, Mississippi; Instructor Visiting Lecturer in Ophthalmology, Harvard Medical School, Boston, Massachusetts The Pupil

Contributing Authors

DAVID R. CORNBLATH, M.D. Professor of Neurology, Johns Hopkins University School of Medicine; Neurologist, and Director, Neurology EMG Laboratory, The Johns Hopkins Hospital, Baltimore, Maryland Peripheral Neuropathies in Human Immunodeficiency Virus Infection

T. COWEN, PH.D. Professor of Neurobiology of Aging, Department of Anatomy and Developmental Biology, University College London, Royal Free Campus, London, United Kingdom Aging in the Peripheral Nervous System

PAULA CUDIA, M.D. Research Fellow, Centre for Neuromuscular Disease, The National Hospital for Neurology, London, United Kingdom Peripheral Nerve Diseases Associated with Mitochondrial Respiratory Chain Dysfunction

BASIL T. DARRAS, M.D. Professor of Neurology (Pediatrics), Harvard Medical School; Senior Associate in Neurology, Director, Neuromuscular Program, and Director, Residency Training Program and Outpatient Clinics, Department of Neurology, Children’s Hospital Boston, Boston, Massachusetts Peripheral Neuropathies in Infants and Children: Polyneuropathies, Mononeuropathies, Plexopathies, and Radiculopathies

JENNY L. DAVIES, B.A. Data Analyst II, Peripheral Neuropathy Research Center, Mayo Clinic, Rochester, Minnesota Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores

WILLIAM C. DE GROAT, M.SC., PH.D. Department of Pharmacology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania

vii

MICHAEL DONAGHY, D.PHIL., F.R.C.P. Reader in Clinical Neurology, University of Oxford, Oxford; Consultant Neurologist, Department of Clinical Neurology, The Radcliffe Infirmary, Oxford; Honorary Civilian Consultant in Neurology to the Army, United Kingdom Lumbosacral Plexus Lesions

PETER J. DYCK, M.D. Professor of Neurology and Roy E. and Merle Meyer Professor of Neuroscience, Mayo Clinic College of Medicine; Head of the Peripheral Neuropathy Research Center and Consultant in Neurology, Mayo Clinic, Rochester, Minnesota Pathologic Alterations of Nerves; Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores; Compound Action Potentials of Sural Nerve in Vitro in Peripheral Neuropathy; Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes; Quantitative Sensation Testing; Hereditary Motor and Sensory Neuropathies: An Overview of Clinical, Genetic, Electrophysiologic, and Pathologic Features; HMSN II (CMT2) and Miscellaneous Inherited System Atrophies of Nerve Axon: Clinical–Molecular Genetic Correlates; HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics; Chronic Inflammatory Demyelinating Polyradiculoneuropathy; Neuropathy Associated with the Monoclonal Gammopathies; Nonmalignant Inflammatory Sensory Polyganglionopathy; Microvasculitis; Amyloidosis and Neuropathy

P. JAMES B. DYCK, M.D. Associate Professor of Neurology, Mayo Medical School; Consultant in Neurology, Mayo Clinic, Rochester, Minnesota Pathologic Alterations of Nerves; Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores; Quantitative Sensation Testing; Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties; Microvasculitis

ANDREW G. ENGEL, M.D. McKnight-3M Professor of Neuroscience, Mayo Clinic College of Medicine, Rochester, Minnesota Diseases of the Neuromuscular Junction

Autonomic Systems to the Urinary Bladder and Sexual Organs

ANGELA DISPENZIERI, M.D. Assistant Professor of Medicine, Mayo Graduate School of Medicine; Consultant, Mayo Clinic, Rochester, Minnesota POEMS Syndrome (Osteosclerotic Myeloma)

MARY L. DOMBOVY, M.D., M.H.S.A. Clinical Associate Professor of Neurology and Physical Medicine and Rehabilitation, University of Rochester; Department Chair, Physical Medicine and Rehabilitation, Unity Health System, Rochester, New York Rehabilitation Management of Neuropathies

JANEAN ENGELSTAD, H.T. Histology Technician, Peripheral Neuropathy Research Center, Mayo Clinic, Rochester, Minnesota Pathologic Alterations of Nerves; Microvasculitis

MARK A. FERRANTE, M.D. Clinical Associate Professor of Neurology, Tulane University Medical Center, New Orleans, Louisiana; Director, EMG Laboratory, Bienville Orthopaedic Specialists, Biloxi, Mississippi Upper Limb Neuropathies: Long Thoracic (Nerve to the Serratus Anterior), Suprascapular, Axillary, Musculocutaneous, Radial, Ulnar, and Medial Antebrachial Cutaneous

viii

Contributing Authors

JOHN P. FRAHER, M.B., B.CH., B.A.O., F.R.C.S.ED., PH.D., D.SC., M.R.I.A. Professor of Anatomy, University College Cork, Cork, Ireland Microscopic Anatomy of the Peripheral Nervous System

MASON W. FREEMAN, M.D. Associate Professor of Medicine, Harvard Medical School; Physician, and Chief, Lipid Metabolism Unit, Massachusetts General Hospital, Boston, Massachusetts Tangier Disease and Neuropathy

ROY FREEMAN, M.B., CH.B. Associate Professor of Neurology, Harvard Medical School; Director, Autonomic and Peripheral Nerve Laboratory, Beth Israel Deaconess Medical Center, Boston, Massachusetts Treatment of Cardiovascular Autonomic Failure

THOMAS R. FRITSCHE, M.D., PH.D. Associate Director, The JONES Group/JMI Laboratories, North Liberty, Iowa

RALF GOLD, M.D. Consultant, and Head, Research Group for MS and Neuroimmunology, Julius-Maximilians-University Faculty of Medicine; Professor of Neurology, Department of Neurology, University Hospital Würzburg, Würzburg, Germany Introduction to Immune Reactions in the Peripheral Nervous System; Experimental Autoimmune Neuritis

IAN A. GRANT, M.D., F.R.C.P.(C) Assistant Professor, Division of Neurology, Dalhousie University; QEII Health Sciences Centre, Halifax, Nova Scotia, Canada Differential Diagnosis of Polyneuropathy; Cryptogenic Sensory Polyneuropathy

NORMAN A. GREGSON, PH.D. Reader in Neuroimmunology, Department of Clinical Neurosciences, Guy’s, King’s and St. Thomas’ School of Medicine, London, United Kingdom Peripheral Nerve Antigens

Parasitic Infections of the Peripheral Nervous System

ANNEKE GABREËLS-FESTEN, M.D., PH.D. Assistant Professor, University Nijmegen; Institute of Neurology, University Medical Center Nijmegen, Nijmegen, The Netherlands Autosomal Recessive Hereditary Motor and Sensory Neuropathies

ERNEST D. GARDNER, M.D.* Professor of Neurology, Orthopaedic Surgery, and Anatomy, University of California, Davis, School of Medicine, Davis, California Gross Anatomy of the Peripheral Nervous System

JOHN W. GRIFFIN, M.D. Professor and Director, Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, Maryland The Control of Axonal Caliber; The Guillain-Barré Syndromes

MICHAEL J. GROVES, B.SC.(HONS.), PH.D. Non-Clinical Lecturer in Peripheral Nerve Pathology, Division of Neuropathology, Institute of Neurology, University College London, London, United Kingdom Pathology of Peripheral Neuron Cell Bodies

CATERINA GIANNINI, M.D. Assistant Professor, Department of Pathology, and Consultant, Mayo Clinic, Rochester, Minnesota Tumors and Tumor-like Conditions of Peripheral Nerve

DONALD H. GILDEN, M.D. Louise Baum Professor and Chairman, Department of Neurology, and Professor of Microbiology, University of Colorado Health Sciences Center; Professor and Chairman, Department of Neurology, University of Colorado Hospital, Denver, Colorado Herpesvirus Infection and Peripheral Neuropathy

HANS H. GOEBEL, M.D. Professor of Neuropathology, University of Mainz, Medical Center; Professor of Neuropathology, Johannes Gutenberg University, Mainz, Germany Lysosomal and Peroxisomal Disorders

*Deceased.

THOMAS M. HABERMANN, M.D. Professor of Medicine, Mayo Medical School; Consultant, Division of Hematology and Internal Medicine, Mayo Clinic Rochester, Rochester, Minnesota Peripheral Neuropathy Associated with Lymphoma, Leukemia, and Myeloproliferative Disorders

ANGELIKA F. HAHN, M.D., F.R.C.P.(C) Professor of Neurology, University of Western Ontario, and London Health Sciences Centre, London, Ontario, Canada Chronic Inflammatory Demyelinating Polyradiculoneuropathy

SUSAN HALL, B.SC., PH.D., D.SC. Head, Division of Anatomy, Cell and Human Biology, Guy’s, King’s and St. Thomas’ School of Biomedical Sciences, London, United Kingdom Mechanisms of Repair after Traumatic Injury

Contributing Authors

JOHN R. HALLIWILL, PH.D. Assistant Professor, Exercise and Movement Science, University of Oregon, Eugene, Oregon Sympathetic Nerves and Control of Blood Vessels to Human Limbs

MICHAEL G. HANNA, M.B.CH.B.(HONS.), M.D., F.R.C.P.(UK) Reader in Clinical Neurology, Department of Molecular Neuroscience, Institute of Neurology, University College London; Consultant Neurologist, Centre for Neuromuscular Disease, The National Hospital for Neurology and Neurosurgery, London, United Kingdom Peripheral Nerve Diseases Associated with Mitochondrial Respiratory Chain Dysfunction

A. E. HARDING, M.D., F.R.C.P.* Professor of Clinical Neurology, Institute of Neurology, London; Consultant Neurologist, The National Hospital for Neurology and Neurosurgery, London, United Kingdom Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons

HANS-PETER HARTUNG, M.D. Professor and Chairman, Department of Neurology, Heinrich-Heine Universität Düsseldorf, Düsseldorf, Germany Introduction to Immune Reactions in the Peripheral Nervous System; Experimental Autoimmune Neuritis; Principles of Immunotherapy; Chronic Inflammatory Demyelinating Polyradiculoneuropathy

STEVEN HERSKOVITZ, M.D. Associate Professor of Clinical Neurology, Albert Einstein College of Medicine; Montefiore Medical Center, Bronx, New York Neuropathy Caused by Drugs

AHMET HÖKE, M.D., PH.D. Assistant Professor, Departments of Neurology and Neuroscience, Johns Hopkins University School of Medicine; Staff Neurologist and Pathologist, and Assistant Director of Neuromuscular Histopathology Laboratory, The Johns Hopkins Hospital, Baltimore, Maryland The Control of Axonal Caliber; Peripheral Neuropathies in Human Immunodeficiency Virus Infection

RICHARD A. C. HUGHES, M.D., F.R.C.P., F.MED.SCI. Professor and Head, Department of Clinical Neurosciences, Guy’s King’s and St. Thomas’ School of Medicine, King’s College London, University of London, London, United Kingdom Peripheral Nerve Antigens; Principles of Immunotherapy; Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes; Diseases of the Fifth Cranial Nerve

*Deceased.

ix

CLARE HUXLEY, PH.D. Reader, Imperial College London, London, United Kingdom Transgenic Models of Inherited Neuropathy

ROBERT R. JACOBSON, M.D., PH.D. Former Director, Leprosy Consultant, and Lecturer, National Hansen’s Disease Center, Baton Rouge, Louisiana Leprosy

ANN JACOBY, PH.D. Professor of Medical Sociology, Department of Primary Care, University of Liverpool, Liverpool, United Kingdom Health Outcomes and Quality of Life

KRISTJÁN R. JESSEN, M.SC., PH.D. Professor of Developmental Neurobiology, Department of Anatomy and Developmental Biology, University College London, London, United Kingdom Molecular Signaling in Schwann Cell Development

DAVID M. JOHNSON, B.S. Engineering Technical Specialist, Division of Engineering and Technology Services, Mayo Clinic, Rochester, Minnesota Quantitative Sensation Testing

H. ROYDEN JONES, JR., M.D. Jamie Ortiz-Patino Chair in Neurology, and Chairman Emeritus of the Division of Medical Specialties and Department of Neurology, Lahey Clinic, Burlington; Clinical Professor of Neurology, Harvard Medical School, Boston; Director of the EMG Laboratory, Children’s Hospital Boston, Boston, Massachusetts Peripheral Neuropathies in Infants and Children: Polyneuropathies, Mononeuropathies, Plexopathies, and Radiculopathies

MICHAEL J. JOYNER, M.D. Professor of Anesthesiology, Mayo Medical School; Professor of Anesthesiology, and Vice-Chair, Department of Physiology, Mayo Clinic, Rochester, Minnesota Sympathetic Nerves and Control of Blood Vessels to Human Limbs

BASHAR KATIRJI, M.D. Professor of Neurology, Case Western Reserve University; Chief, Neuromuscular Division, and Director, EMG Laboratory, University Hospitals of Cleveland, Cleveland, Ohio Mononeuropathies of the Lower Limb

KENTON R. KAUFMAN, PH.D. Associate Professor of Bioengineering, Mayo Medical School; Director, Motion Analysis Laboratory, and Consultant, Department of Orthopedic Surgery, Mayo Clinic, Rochester, Minnesota Quantitative Muscle Strength Assessment

x

Contributing Authors

JOHN J. KELLY, M.D. Professor and Chair, Department of Neurology, George Washington University, Washington, D.C. Amyloidosis and Neuropathy

WILLIAM R. KENNEDY, M.S., M.D. Professor of Neurology, Department of Neurology, University of Minnesota, Minneapolis, Minnesota Pathology and Quantitation of Cutaneous Innervation

MATTHEW C. KIERNAN, M.R.G.S.(HONS.), PH.D., F.R.A.C.P. Senior Lecturer in Neurology, University of New South Wales and Prince of Wales Medical Research Institute; Consultant Neurologist, Princes of Wales Hospital, Randwick, Sydney, New South Wales, Australia Nerve Excitability Measures: Biophysical Basis and Use in the Investigation of Peripheral Nerve Disease

BERND C. KIESEIER, M.D. Assistant Professor of Neurology, Heinrich-HeineUniversity, Düsseldorf, Germany Introduction to Immune Reactions in the Peripheral Nervous System; Experimental Autoimmune Neuritis

JUN KIMURA, M.D. Professor, Department of Neurology, University of Iowa Health Care, and University of Iowa, Iowa City, Iowa Nerve Conduction and Needle Electromyography

R. H. M. KING, PH.D., M.SC., F.R.C.PATH. Principal Research Fellow, Department of Clinical Neurosciences, Royal Free and University College Medical School, London, United Kingdom Microscopic Anatomy of the Peripheral Nervous System; Aging in the Peripheral Nervous System

CHRISTOPHER J. KLEIN, M.D. Assistant Professor of Neurology, Mayo Medical School; Consultant in Neurology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores; Quantitative Sensation Testing; Hereditary Motor and Sensory Neuropathies: An Overview of Clinical, Genetic, Electrophysiologic, and Pathologic Features; HMSN II (CMT2) and Miscellaneous Inherited System Atrophies of Nerve Axon: Clinical–Molecular Genetic Correlates; Hereditary Brachial Plexus Neuropathy; HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

KLEOPAS A. KLEOPA, M.D. Senior Consultant Neurologist, The Cyprus Institute of Neurology and Genetics, Nicosia, Cyprus X-linked Charcot-Marie-Tooth Disease

CHRISTOPHER J. KLINGELE, M.D. Assistant Professor of Obstetrics and Gynecology, Mayo Clinic College of Medicine; Consultant in Obstetrics and Gynecology, Mayo Clinic, Rochester, Minnesota Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

DAVID L. KREULEN, PH.D. Professor, Departments of Physiology and Neurology, Michigan State University, East Lansing, Michigan Neurobiology of Autonomic Ganglia

ROBERT A. KYLE, M.D. Consultant, Division of Hematology and Internal Medicine, Mayo Clinic; Professor of Medicine and Laboratory Medicine, Mayo Medical School, Rochester, Minnesota Neuropathy Associated with the Monoclonal Gammopathies; Amyloidosis and Neuropathy; POEMS Syndrome (Osteosclerotic Myeloma)

CATHERINE LACROIX, M.D. Hospital Practitioner – Neuropathy, Centre Hospitalier de Bicetre, Le Kremlin, Bicetre, France Sarcoid Neuropathy

JOHN T. KISSEL, M.D. Professor and Vice-Chair, Department of Neurology, and Director, Division of Neuromuscular Disease, The Ohio State University, Columbus, Ohio Neuropathies with Systemic Vasculitis

TERRENCE D. LAGERLUND, M.D., PH.D. Associate Professor of Neurology, Mayo Clinic College of Medicine; Consultant in Neurology, Mayo Clinic, Rochester, Minnesota Nerve Blood Flow and Microenvironment

CAROLINE M. KLEIN, M.D., PH.D. Assistant Professor of Neurology, Department of Neurology, University of North Carolina School of Medicine, Chapel Hill, North Carolina Diseases of the Seventh Cranial Nerve; The Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

EDWARD H. LAMBERT, M.D., PH.D.* Professor Emeritus, Mayo Clinic, Rochester, Minnesota Compound Action Potentials of Sural Nerve in Vitro in Peripheral Neuropathy

*Deceased.

Contributing Authors

SALLY N. LAWSON, PH.D. Professor, Department of Physiology, School of Medical Sciences, University of Bristol, University Walk, Bristol, Avon, United Kingdom The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

xi

LINDA M. LUXON, B.SC.(HONS.), M.B., F.R.C.P. Professor of Audiological Medicine, University College London (Institute of Child Health), University of London; Consultant Physician in Neuro-otology, The National Hospital for Neurology and Neurosurgery, London, United Kingdom Diseases of the Eighth Cranial Nerve

JACQUELINE A. LEAVITT, M.D. Associate Professor of Ophthalmology, Mayo Medical School; Consultant, Department of Ophthalmology, Rochester Methodist Hospital, St. Mary’s Hospital, Rochester, Minnesota Diseases of the Third, Fourth, and Sixth Cranial Nerves

P. NIGEL LEIGH, PH.D., M.B.B.S., F.R.C.P.(UK) Professor of Clinical Neurology, and Head, Department of Neurology, Institute of Psychiatry; Director, King’s MND Care & Research Centre, King’s College Hospital, London, United Kingdom Sporadic Motor Neuron Degeneration; Inherited Motor Neuron Degeneration

J. G. LLEWELYN, B.SC.(HONS), M.D., F.R.C.P. Consultant Neurologist, Royal Gwent Hospital, Newport and Cardiff University School of Medicine, Cardiff, Wales; Honorary Consultant Neurologist, The National Hospital for Neurology and Neurosurgery, London, United Kingdom Diabetic Neuropathies

GLENN LOPATE, M.D. Assistant Professor of Neurology, Department of Neurology, Washington University; Assistant Professor of Neurology, Barnes-Jewish Hospital, St. Louis, Missouri Polyneuropathies and Antibodies to Nerve Components

PHILLIP A. LOW, M.D., F.R.A.C.P. Professor of Neurology, and Chairman, Division of Clinical Neurophysiology, Mayo Medical School; Consultant in Neurology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota Oxidative Stress and Excitatory Neurotoxins in Neuropathy; Nerve Blood Flow and Microenvironment; Quantitation of Autonomic Impairment; Management of Autonomic Failure

JAMES R. LUPSKI, M.D., PH.D. Cullen Professor of Molecular and Human Genetics, and Professor of Pediatrics, Baylor College of Medicine; Consulting Medical Geneticist and Pediatrician, Texas Children’s Hospital and Ben Taub General Hospital; Consulting Geneticist, The Methodist Hospital, Texas Women’s Hospital, and St. Joseph’s Hospital, Baylor College of Medicine, Houston, Texas Hereditary Motor and Sensory Neuropathies: An Overview of Clinical, Genetic, Electrophysiologic, and Pathologic Features; Hereditary Motor and Sensory Neuropathies Involving Altered Dosage or Mutation of PMP22: The CMT1A Duplication and HNPP Deletion; Hereditary Motor and Sensory Neuropathy Related to Early Growth Response 2 (EGR2) Gene

RUDOLF MARTINI, PH.D. Professor and Head, Developmental Neurobiology, Department of Neurology, School of Medicine, University of Wuerzburg, Wuerzburg, Germany Myelination; Transgenic Models of Nerve Degeneration

CHRISTOPHER J. MATHIAS, D.PHIL., D.SC., F.R.C.P., F.MED.SCI. Professor of Neurovascular Medicine, Division of Neuroscience and Psychological Medicine, and Faculty of Medicine, Imperial College, London; Division of Clinical Neurology, Institute of Neurology, University College London; Consultant Physician and Director, Neurovascular Medicine Unit, St. Mary’s Hospital, London; Autonomic Unit, The National Hospital for Neurology and Neurosurgery, London, United Kingdom Quantitation of Autonomic Impairment; Diseases of the Ninth, Tenth, Eleventh, and Twelfth Cranial Nerves

JUSTIN C. MCARTHUR, M.B.B.S., M.P.H. Professor of Neurology and Epidemiology, Johns Hopkins University; Deputy Director, Department of Neurology, Johns Hopkins Hospital, Baltimore, Maryland Pathology and Quantitation of Cutaneous Innervation

ELIZABETH S. MCDONALD, M.D., PH.D. Radiology Resident, Mayo Clinic College of Medicine, Rochester, Minnesota Neurotrophic Factors in the Peripheral Nervous System

JAMES G. MCLEOD, M.B.B.S., D.PHIL., D.SC., F.R.A.C.P., F.R.C.P. Emeritus Professor, University of Sydney; Honorary Consulting Neurologist, Royal Prince Alfred Hospital, Camperdown, New South Wales, Australia Paraneoplastic Neuropathy

PHILIP G. MCMANIS, M.B.B.S., M.D., F.R.A.C.P.* Associate Professor, University of Sydney; Senior Staff Neurologist and Director of Clinical Neurophysiology, Royal North Shore Hospital, Sydney, New South Wales, Australia Nerve Blood Flow and Microenvironment

L. JOSEPH MELTON III, M.D. Professor of Epidemiology, Mayo Medical School, Mayo Clinic; Mayo Clinic, Rochester, Minnesota Epidemiologic Approaches to Peripheral Neuropathy

*Deceased.

xii

Contributing Authors

ALBEE MESSING, V.M.D., PH.D. Professor of Pathology and Neuroscience, School of Veterinary Medicine and Waisman Center, University of Wisconsin-Madison, Madison, Wisconsin Diphtheritic Polyneuropathy

VIRGINIA V. MICHELS, M.D. Professor of Medical Genetics, Mayo Clinic/Foundation, Rochester, Minnesota Mendelian and Mitochondrial Inheritance, Gene Identification, and Clinical Testing

RHONA MIRSKY, PH.D. Professor of Developmental Neurobiology, Department of Anatomy and Developmental Biology, University College London, London, United Kingdom Molecular Signaling in Schwann Cell Development

JOHN D. POLLARD, M.B.B.S., B.SC.(MED), PH.D., F.R.A.C.P. Bushell Professor of Neurology, University of Sydney; Head of Neurology, Royal Prince Alfred Hospital, Camperdown, New South Wales, Australia Principles of Immunotherapy; Neuropathy in Diseases of the Thyroid and Pituitary Glands

MICHAEL POLYDEFKIS, M.D. Assistant Professor of Neurology, Johns Hopkins University, Baltimore, Maryland Pathology and Quantitation of Cutaneous Innervation

SUDHA POTTUMARTHY, M.B.B.S., F.R.C.P.A. Senior Fellow, Department of Laboratory Medicine, and Acting Director, Antimicrobics Testing Laboratory, University of Washington School of Medicine, Seattle, Washington Parasitic Infections of the Peripheral Nervous System

PETER C. O’BRIEN, PH.D. Professor of Biostatistics, Division of Biostatistics, Mayo Clinic, Rochester, Minnesota Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores; Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes; Quantitative Sensation Testing

GRAHAM M. O’HANLON, PH.D. Research Assistant, Division of Clinical Neurosciences, University of Glasgow, Glasgow, Scotland, United Kingdom

MARY M. REILLY, M.D., F.R.C.P.I., F.R.C.P. Honorary Senior Lecturer, Department of Molecular Neurosciences, Institute of Neurology; Consultant Neurologist, The National Hospital for Neurology and Neurosurgery, London, United Kingdom Hereditary Amyloid Neuropathy

ANDREA ROBERTSON, PH.D. Career Development Fellow, Medical Research Council, London, United Kingdom Transgenic Models of Inherited Neuropathy

Peripheral Nerve Antigens

GILMORE N. O’NEILL, M.B., M.R.C.P.I. Instructor in Neurology and Assistant in Neurology, Massachusetts General Hospital, Boston, Massachusetts Tangier Disease and Neuropathy

GUSTAVO C. ROMAN, M.D., F.A.C.P., F.A.A.N., F.R.S.M.(LOND.) Professor of Medicine/Neurology, Division of Neurology, Department of Medicine, University of Texas Health Science Center at San Antonio; Neurologist, University Hospital and VA Hospital, San Antonio, Texas Tropical Myeloneuropathies

DAVID J. PATERSON, M.SC., M.A., D.PHIL. Professor of Physiology, University of Oxford, Director of Pre-clinical Studies, Fellow of Merton College, Oxford, University of Oxford, Oxford, United Kingdom Neural Control of Cardiac Function

ALAN PESTRONK, M.D. Professor of Neurology and Pathology, Washington University Medical School; Professor, Barnes-Jewish Hospitals, St. Louis, Missouri Polyneuropathies and Antibodies to Nerve Components

DAVID PLEASURE, M.D. Professor, Neurology and Pediatrics, University of Pennsylvania School of Medicine; Director, Joseph Stokes Jr. Research Institute, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania Diphtheritic Polyneuropathy

MICHAEL C. ROWBOTHAM, M.D. Professor of Clinical Neurology and Anesthesia, University of California, San Francisco; Director, UCSF Pain Clinical Research Center; Senior Attending Neurologist, UCSF-Mount Zion Pain Management Center, San Francisco, California Mechanisms and Pharmacologic Management of Neuropathic Pain

MONIQUE M. RYAN, M.B.B.S., M. MED., F.R.A.C.P. Paediatric Neurologist, Institute for Neuromuscular Research, and Senior Lecturer, Discipline of Paediatrics and Child Health, University of Sydney, The Children’s Hospital at Westmead, Westmead, New South Wales, Australia Peripheral Neuropathies in Infants and Children: Polyneuropathies, Mononeuropathies, Plexopathies, and Radiculopathies

Contributing Authors

MARTIN RYDMARK, M.D., PH.D. Associate Professor, Senior Lecturer, Mednet – Medical Informatics & Computer Assisted Education, Institute of Anatomy and Cell Biology, The Sahlgrenska Academy at Göteborg University, Göteborg, Sweden

xiii

MARTIN SCHMELZ, M.D., PH.D. Professor, Department of Anesthesiology and Intensive Care Medicine, Faculty of Clinical Medicine Mannheim, University of Heidelberg, Mannheim, Germany Single-Unit Recordings of Afferent Human Peripheral Nerves by Microneurography

Microscopic Anatomy of the Peripheral Nervous System

THOMAS D. SABIN, M.D. Professor and Vice Chair of Neurology, Tufts University School of Medicine; Acting Chief, Department of Neurology, Tufts-New England Medical Center, Boston, Massachusetts Leprosy

GÉRARD SAID, M.D. Professor of Neurology, University Paris-Sud, Paris; Chairman, Service of Neurology, Hopital de Bicetre, Le Kremlin Bicetre, Val De Marne, France Lyme Disease; Sarcoid Neuropathy

DAVID S. SAPERSTEIN, M.D. Assistant Professor of Neurology and Pathology, and Chief, Neuromuscular Disease Service, University of Kansas Medical Center, Kansas City, Kansas Polyneuropathy Caused by Nutritional and Vitamin Deficiency

FRANCESCO SCARAVILLI, M.D., PH.D., D.SC., F.R.C.PATH. Professor of Neuropathology, Honorary Consultant Neuropathologist, and Head of Division of Neuropathology, Institute of Neurology, The National Hospital for Neurology and Neurosurgery, University College London, London,United Kingdom Pathology of Peripheral Neuron Cell Bodies

JON J. A. SCOTT, B.SC., PH.D. Senior Lecturer in Physiology, Department of Pre-Clinical Sciences, University of Leicester, Leicester, United Kingdom The Golgi Tendon Organ

KAZIM SHEIKH, M.D. Assistant Professor, Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, Maryland The Guillain-Barré Syndromes

JOHN T. SHEPHERD, M.D. Professor Emeritus, Mayo Clinic, Rochester, Minnesota Sympathetic Nerves and Control of Blood Vessels to Human Limbs

MICHAEL E. SHY, M.D. Professor of Neurology, Professor of Molecular Medicine and Genetics, Director of the Inherited Neuropathy Clinic, and Co-Director of the Neuromuscular Program, Wayne State University School of Medicine, Detroit, Michigan Hereditary Motor and Sensory Neuropathies: An Overview of Clinical, Genetic, Electrophysiologic, and Pathologic Features; Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

WOLFGANG SINGER, M.D. Resident, Neurology, Mayo Clinic, Rochester, Minnesota Management of Autonomic Failure

HERBERT H. SCHAUMBURG, M.D. Edwin S. Lowe Professor and Chairman of Neurology, Albert Einstein College of Medicine; Chairman of Neurology, Montefiore Medical Center, Bronx, New York Human Toxic Neuropathy Caused by Industrial Agents; Neuropathy Caused by Drugs

STEVEN S. SCHERER, M.D., PH.D. William N. Kelley Professor of Neurology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania X-linked Charcot-Marie-Tooth Disease

RAPHAEL SCHIFFMANN, M.D. Section Chief, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland Fabry’s Disease

BENN E. SMITH, M.D. Assistant Professor of Neurology, Mayo Clinic College of Medicine, Rochester, Minnesota Nonmalignant Inflammatory Sensory Polyganglionopathy

ERIC J. SORENSON, M.D. Assistant Professor, Mayo Foundation Hospitals, Mayo Medical School; Head of Section, Neuromuscular Diseases, and Director of Motor Neuron Program, Mayo Clinic, Rochester, Minnesota Transgenic Animal Models of Amyotrophic Lateral Sclerosis

JUDITH M. SPIES, M.B.B.S., PH.D., F.R.A.C.P. Senior Lecturer, University of Sydney; Staff Specialist Neurologist, Royal Prince Alfred Hospital, Camperdown, New South Wales, Australia Paraneoplastic Neuropathy

xiv

Contributing Authors

ERIK V. STÅLBERG, M.D., PH.D., F.R.C.P. Professor Emeritus, Department of Neurosciences, Clinical Neurophysiology, Uppsala University; Professor Emeritus, Department of Clinical Neurophysiology, Uppsala University Hospital, Uppsala, Sweden Single-Fiber Electromyography and Other Electrophysiologic Techniques for the Study of the Motor Unit

J. CLARKE STEVENS, M.D. Professor of Neurology, Mayo Medical School, Rochester, Minnesota Median Neuropathy

GUIDO STOLL, M.D. Consultant, Julius-Maximilians-University Faculty of Medicine; Professor of Neurology, Department of Neurology, University Hospital Würzburg, Würzburg, Germany Introduction to Immune Reactions in the Peripheral Nervous System; Experimental Autoimmune Neuritis

GUILLERMO A. SUAREZ, M.D. Associate Professor of Neurology, Mayo Clinic College of Medicine; Consultant in Neurology, Mayo Clinic, Rochester, Minnesota Immune Brachial Plexus Neuropathy; POEMS Syndrome (Osteosclerotic Myeloma)

UELI SUTER, PH.D. Professor, Institute of Cell Biology, Swiss Federal Institute of Technology, ETH Zu¯rich, ETH Ho¯nggerberg, Zu¯rich, Switzerland Myelination

THOMAS R. SWIFT, M.D. Professor Emeritus of Neurology, Medical College of Georgia, Augusta, Georgia Leprosy

BRUCE V. TAYLOR, M.B., M.D., F.R.A.C.P. Clinical Senior Lecturer, University of Tasmania; Head, Department of Neurology, Royal Hobart Hospital, Hobart, Tasmania, Australia Multifocal Motor Neuropathy and Conduction Block

AYALEW TEFFERI, M.D. Professor of Medicine and Hematology, Mayo Medical School and Mayo Clinic, Rochester, Minnesota Peripheral Neuropathy Associated with Lymphoma, Leukemia, and Myeloproliferative Disorders

STEPHEN N. THIBODEAU, PH.D. Professor of Laboratory Medicine, Mayo Clinic/Foundation, Rochester, Minnesota Mendelian and Mitochondrial Inheritance, Gene Identification, and Clinical Testing

P. K. THOMAS, C.B.E., M.D., D.SC., F.R.C.P., F.R.C.(PATH.) Emeritus Professor of Neurology, University College London School of Medicine and The National Hospital for Neurology and Neurosurgery, London, United Kingdom Clinical Patterns of Peripheral Neuropathy; Diseases of the Ninth, Tenth, Eleventh, and Twelfth Cranial Nerves; Autosomal Recessive Hereditary Motor and Sensory Neuropathies; Lysosomal and Peroxisomal Disorders; Diabetic Neuropathies

PHILIP D. THOMPSON, M.B.B.S., PH.D., F.R.A.C.P. Professor of Neurology, University of Adelaide; Head, Department of Neurology, Royal Adelaide Hospital, Adelaide, South Australia Clinical Patterns of Peripheral Neuropathy

ERIK C. THORLAND, PH.D. Fellow, Clinical Molecular Genetics, Mayo Clinic/Foundation, Rochester, Minnesota Mendelian and Mitochondrial Inheritance, Gene Identification, and Clinical Testing

D. R. TOMLINSON, PH.D., D.SC. Professor of Neuropharmacology, Family of Life Sciences, The University of Manchester, Manchester, England Diabetic Neuropathies

ERIK TOREBJÖRK, M.D., PH.D. Professor, Department of Clinical Neurophysiology, University of Uppsala, Uppsala, Sweden Single-Unit Recordings of Afferent Human Peripheral Nerves by Microneurography

KLAUS V. TOYKA, M.D., F.R.C.P. Professor of Neurology, Julius-Maximilians-University Faculty of Medicine; Neurologist-In-Chief, Department of Neurology, University Hospital Würzburg, Würzburg, Germany Introduction to Immune Reactions in the Peripheral Nervous System; Experimental Autoimmune Neuritis

JO Zˇ E V. TRONTELJ, M.D., PH.D. Professor, Institute of Clinical Neurophysiology, University Medical Center, Ljubljana, Slovenia Single Fiber Electromyography and Other Electrophysiologic Techniques for the Study of the Motor Unit

KENNETH L. TYLER, M.D. Reuler-Lewin Family Professor of Neurology and Professor of Medicine, Microbiology and Immunology, University of Colorado Health Sciences Center; Chief, Neurology Service, Denver Veterans Affairs Medical Center, Denver, Colorado Herpesvirus Infection and Peripheral Neuropathy

Contributing Authors

B. ULFHAKE, PH.D. Professor, Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden Aging in the Peripheral Nervous System

PAUL M. VANHOUTTE, M.D., PH.D. Distinguished Visiting Professor, Department of Pharmacology, University of Hong Kong, Hong Kong Sympathetic Nerves and Control of Blood Vessels to Human Limbs

ANNABEL K. WANG, M.D. Assistant Professor, Department of Neurology, Mount Sinai School of Medicine; Assistant Attending Physician, Mount Sinai Hospital, New York, New York The Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

LAURA E. WARNER, PH.D. Research Scientist, University of Washington, Seattle, Washington Hereditary Motor and Sensory Neuropathy Related to Early Growth Response 2 (EGR2) Gene

HENRY DEF. WEBSTER, M.D. Emeritus Scientist, NINDS, National Institutes of Health, Bethesda, Maryland Introduction

ANANDA WEERASURIYA, M.PHIL., PH.D. Professor of Neuroscience and Physiology, Mercer University School of Medicine, Macon, Georgia Blood-Nerve Interface and Endoneurial Homeostasis

GWEN WENDELSCHAFER-CRABB, M.S. Senior Scientist, Department of Neurology, University of Minnesota, Minneapolis, Minnesota Pathology and Quantitation of Cutaneous Innervation

EELCO F. M. WIJDICKS, M.D. Professor of Neurology, and Chair, Division of Critical Care Neurology, Department of Neurology, Mayo Clinic, Rochester, Minnesota Management of Patients with Acute Neuromuscular Disease in the Intensive Care Unit

ASA J. WILBOURN, M.D. Clinical Professor of Neurology, Case Western Reserve University; Director, EMG Laboratory, Department of Neurology, Cleveland Clinic, Cleveland, Ohio Brachial Plexus Lesions; Upper Limb Neuropathies: Long Thoracic (Nerve to the Serratus Anterior), Suprascapular, Axillary, Musculocutaneous, Radial, Ulnar, and Medial Antebrachial Cutaneous; Mononeuropathies of the Lower Limb

xv

HUGH J. WILLISON, M.B.B.S., PH.D., F.R.C.P. Professor of Neurology, Division of Clinical Neurosciences, University of Glasgow; Honorary Consultant Neurologist, Institute of Neurological Sciences, Southern General Hospital, Glasgow, Scotland, United Kingdom Peripheral Nerve Antigens; Multifocal Motor Neuropathy and Conduction Block

ANTHONY J. WINDEBANK, M.D. Professor of Neurology, Mayo Graduate School and Mayo Medical School; Consultant in Neurology, Mayo Clinic and Mayo Foundation; Dean, Mayo Medical School, Mayo Clinic College of Medicine, Rochester, Minnesota Neurotrophic Factors in the Peripheral Nervous System; Hereditary Brachial Plexus Neuropathy; Porphyric Neuropathy; Nonmalignant Inflammatory Sensory Polyganglionopathy; Metal Neuropathy

HARALD WITTE Graduate Student, Max-Planck-Institute of Neurobiology, Munich, Germany Guidance of Axons to Targets in Development and in Disease

JACKIE D. WOOD, M.S., PH.D. Professor of Physiology and Cell Biology and Internal Medicine, Ohio State University College of Medicine and Public Health, Columbus, Ohio Neurobiology of the Enteric Nervous System

BRIAN R. YOUNGE, M.D. Associate Professor of Ophthalmology, Mayo Medical School, Mayo Graduate School; Consultant in Ophthalmology, Mayo Clinic, Rochester, Minnesota Diseases of the Third, Fourth, and Sixth Cranial Nerves

DOUGLAS W. ZOCHODNE, M.D., F.R.C.P.(C.) Professor, University of Calgary; Consultant Neurologist, Calgary Health Region, Foothills Hospital, Calgary, Alberta, Canada Neuropathies Associated with Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness

Preface

More than a decade has passed since the third edition of Peripheral Neuropathy. The editors have found it necessary to create an essentially new textbook in order to encompass the rapid, exciting advances in neurobiology, molecular genetics, chemical and cellular pathology, and treatment. With the exception of chapters on gross anatomy, the compound action potential of the sural nerve in vitro, and the chapter on progressive muscular atrophy, most chapters are completely rewritten to produce the broadest and most up-to-date reviews of the neurobiology of the nerve and its diseases. The editors hope that the readers will find the new edition comprehensive and informative. The only credit the editors take is for their choice of authors, who are authorities in their fields. We are deeply grateful for their hard work and new insights. As in previous editions, we honor a pioneer and friend who has made major contributions to our understanding of nerve biology and disease. In the first edition, Wilhelm Krücke, of Frankfurt am Main, was chosen as he was considered to be the Dean of Neuropathology of peripheral nerve. It was not surprising that in his Introduction he mentioned the contributions of T. Schwann (description of the myelin cell named after him), A. Waller and R. Cajal (key contributors to cellular events of nerve fiber degeneration), L. Ranvier (the node that bears his name), A. Gombault (segmental demyelination), R. Virchow (inflammation), and P. Weis (axonal flow). For the second edition, we honored Fritz Buchthal, who along with E. H. Lambert, pioneered clinical nerve conduction and clinical electromyography. It is of some interest that these new techniques have become so informative that the earlier use of direct nerve and muscle stimulation with galvanic and faradic currents (Duchenne, and later Erb) are hardly mentioned in modern textbooks. In this fourth edition, Kiernan and colleagues (Chapter 5) argue that these earlier approaches (or modifications) might still be used in specific indications. J. Z. Young, the special mentor of one of us (PKT), wrote the Introduction to the third edition. Young

(with P. K. Thomas) had returned to the old technique of teasing peripheral nerve fibers—a methodology that remains useful and is extensively described in Chapter 32. In the present edition, we honor Henry deF. Webster. Harry characterized the ultrastructural features of mammalian nerve both in development and after maturation, and provided early studies of pathologic alteration in disease. His electron micrographs in Peters, Palay, and deF. Webster, “The Fine Structure of the Nervous System: The Neurons and Supporting Cells” (W. B. Saunders, 1976), remains as an example of how it should be done. Quite unfairly to authors whose chapters are not mentioned here, we list some of the special coverage. Sally Lawson in her chapter on dorsal root ganglion (spinal ganglion) neurons, reviews sensory properties, electrophysiology, differences between non-nociceptor and nociceptor somas, immunocytochemical properties, trophic factors, cytokines, receptors, ion channels, membrane properties, genes, and injury reactions—a real tour de force! Jackie Wood makes the case for a third nervous system (in addition to the central and peripheral nervous systems), the enteric system. An insight into the system is crucial for the understanding of diseases affecting the gut. The chapter by A. G. Engel provides a concise review of the diseases of the neuromuscular junction. His electron micrographs of pathologic alterations are unmatched. The chapters on quantitating neuropathic impairment, disability, symptoms, outcomes, and quality of life as it relates to neuropathy are more focused on neuropathy and more comprehensive than discussed elsewhere. Ian Grant’s chapter on differential diagnosis of neuropathy is an important read for physicians who want to improve their diagnostic ability. The sections on inherited motor, motor and sensory, and sensory and autonomic neuropathies have undergone major changes. Comprehensive and detailed reviews are focused first on clinical features (inheritance and natural history pattern, population, level and pathologic type of neuron involvement, electrophysiologic and pathologic characteristics) followed by molecular xvii

xviii

Preface

genetic causes. P. James B. Dyck emphasizes the need to recognize that diabetic neuropathy is not a single disorder. Discussing diabetic radiculoplexus neuropathies, he stresses underlying immune mechanisms and the possible benefit of immune-modulating therapy. The chapter on necrotizing vasculitis of the peripheral nervous system by Michael Collins and John Kissel is a definitive and beautifully illustrated contribution. In the pathology chapter, we include the new approaches used in identifying not only the focal or multifocal abnormality of proximal nerves, but also the underlying cause. Many chapters deal with an increasing emphasis on diagnosis and management of diseases of the autonomic nervous system. The strong support of P. A. Low, MD is acknowledged.

The editors are grateful to the editors at Elsevier (Jennifer Ehlers) and at Bermedica Production, Ltd. (Berta Steiner). PJD is grateful to Mary Lou Hunziker for her enormous help. I (PJD) hereby also acknowledges the heroic help of Nok (Sam) Ponsford, M.D., who stepped in to help when one of us, her husband (PKT), developed a stroke, and other health complications, during the editing of this book. Colleagues from The Royal Free Hospital also helped. NOK and PK—thank you. One of us (PJD) is grateful for the forbearance of Isabelle, Ernest Carl, Fred Howard, P. James B., and M. Katharine E. and to grandchildren (Sophie, Jacob, Chloe, and Abbie), and to Marian and Scott. PETER J. DYCK P. K. THOMAS

1 Introduction HENRY DEF. WEBSTER

Early Electron Microscopic Studies of Peripheral Nerve Segmental Demyelination and Remyelination Wallerian Degeneration and Regeneration

Development of Peripheral Nerve Fibers Early Axon–Schwann Cell Interactions Early Schwann Cell–Axon Relationships in Nerves

EARLY ELECTRON MICROSCOPIC STUDIES OF PERIPHERAL NERVE About 50 years ago, the first electron microscopic observations of peripheral nerve were reported. The limited resolution of the light microscope had left some important questions unanswered and had only partially answered others. What were the relationships of axons, their myelin sheaths and Schwann cells? Did myelin have a layered structure and were there discontinuities in the myelin sheath at Schmidt-Lanterman clefts? How was myelin formed and how did myelinated and unmyelinated fibers develop? What structural changes did trauma or disease produce in peripheral nerves? Finally, what cellular changes occurred during nerve regeneration and how did they differ from those seen during development? The electron microscope permitted investigators to visualize relationships of myelin, cell membranes, organelles, fibrils and other cellular inclusions at substantially higher resolution. This was an important advantage and led to a substantial increase in the use of this instrument. Increased use of the electron microscope led investigators to describe the fine structure of the nervous system,27 to quantitate nerve biopsy findings,14,24,36 and to create better methods for preserving peripheral nervous tissue.9,45,47,50,52 Subsequent advances in methodology included immunocytochemical techniques for localizing myelin constituents16,25,38 and an in situ hybridization method to study the distribution of myelin protein messenger RNAs (mRNAs) in sections prepared for electron microscopy.48

Formation and Growth of Myelin Sheaths Distribution of the Protein Constituents of Myelin Concluding Comment

Early studies using the electron microscope showed that axons in the peripheral nervous system were ensheathed by Schwann cells and that myelin was located in the cytoplasm of Schwann cells.13,26,30 Myelin segments terminated at nodes of Ranvier in a series of loops32,41 and had a compact layered structure except in Schmidt-Lanterman clefts.31 There, myelin lamellae were separated by helically arranged strips of cytoplasm. The layered pattern of compact myelin was evident after fixation and was thought to be consistent with molecular models of myelin derived from observations utilizing polarized light and x-ray diffraction.11,23,28,33 Early in development, Schwann cells surrounded bundles of axons and segregated larger ones for subsequent myelination.26 Then myelin formation occurred by spiral growth of the mesaxon, an extension of the Schwann cell surface membrane.13,30 Further myelin membrane growth produced a compact spiral of myelin membrane with inner and outer mesaxons connecting it to the Schwann cell’s periaxonal and external surface membranes. The above findings were important and proved to be useful for investigators who began studying peripheral nervous system pathology with the electron microscope.

SEGMENTAL DEMYELINATION AND REMYELINATION Segmental demyelination, a sequence of changes characterized by myelin damage and relative sparing of axons, is a cardinal feature of human diphtheritic neuritis12 and its 3

4

Introduction

model, experimental diphtheritic neuritis in guinea pigs.43 Because of the importance of segmental demyelination in human demyelinating diseases, we selected this pathologic process for our first electron microscopic study.52 Early observations showed that the preservation of large myelin sheaths had to be improved so that lesions could be distinguished from preparative artifacts. Then early focal lesions were identified paranodally and elsewhere along the length of myelin segments without evidence of inflammation.52 Later breakdown of myelin into ovoids occurred within Schwann cells; nearby macrophages also contained myelin remnants. Severely demyelinated nerves also contained some Schwann cells that were actively regenerating myelin sheaths. Interestingly, some of these sheaths had major variations in contour along their lengths52 that resembled those seen paranodally in normal adult fibers.51 Since published models of myelination did not explain how these myelin irregularities were formed, their development and distribution were investigated in a later study of myelin formation.45 In acute idiopathic neuritis (Guillain-Barré syndrome) and its experimental model, experimental allergic neuritis (EAN42), segmental demyelination is preceded by a perivascular invasion of lymphocytes and mononuclear cells. In a serial section study of early EAN lesions, Åström found that lymphocytes migrated through endothelial cells of rat sciatic nerve venules, a process he considered to be a form of emperipolesis.2,3 Demyelination then began within the perivenular inflammatory infiltrates and involved contact of mononuclear cells with myelin and subsequent stripping of myelin remnants from axon surfaces by mononuclear cells.18

WALLERIAN DEGENERATION AND REGENERATION Nerve transection or crush produces distal degeneration of axons and their myelin sheaths, a process called wallerian degeneration. When the early stages of this process were examined, there were transient accumulations of mitochondria in paranodal regions of large myelinated fibers.44 They occurred also in some unmyelinated axons. Comparisons of proximal and distal paranodal regions showed that most accumulations were in distal paranodal axoplasm. Migration of mitochondria to paranodal regions and their proliferation were considered as possible explanations of these transient accumulations. A later study by others demonstrated that both anterograde and retrograde transport of axonal mitochondria and other organelles occurred during wallerian degeneration.40 After nerve transection, axons and their myelin sheaths regenerate. This process begins in the distal end of the proximal stump, the site we selected for our studies of nerve regeneration. When axons form growth cones and begin regenerating, Schwann cells divide rapidly. They

also synthesize laminin, a major constituent of their basement membranes. Laminin stimulates cell division17 and has an important role in the extension of neurites during nerve development and regeneration.4 After transecting sciatic nerves, we prepared supernatants of proximal and distal stumps and compared their effects on Schwann cell proliferation and laminin production in vitro.53 Lysates of cultures treated with 24-hour proximal supernatants contained significantly higher levels of laminin than those prepared from other supernatant intervals, from distal segments, or from control nerves. Twenty-four–hour and 2-day proximal supernatants also increased proliferation of cultured Schwann cells significantly. These findings suggested that proximal stump axons were possible sources of substances that were responsible for these effects.53 Since the addition of vasoactive intestinal peptide (VIP) to cultured Schwann cells also increased laminin synthesis,54 Zhang and collaborators then tested the effects of locally administered VIP on regenerative responses of transected nerve fibers. They found that injections of VIP into transection sites twice daily for 2 weeks accelerated ensheathment and myelination of regenerating axons.55

DEVELOPMENT OF PERIPHERAL NERVE FIBERS As noted above, the contour variations we described in normal adult51 and regenerating52 myelin sheaths were not explained by contemporary models of myelination. Our interest in their formation and distribution led to a series of later studies of Schwann cell and nerve fiber development that are described and illustrated in an earlier edition of this text.46

Early Axon–Schwann Cell Interactions In the developing peripheral nervous system, outgrowth of neurites is soon followed by the appearance of Schwann cells that originate from the neural crest.21 Speidel described their early association with axons in the transparent tail fins of tadpoles, along with important parameters of myelin formation.34 Billings-Gagliardi extended these findings by correlating in vivo observations of early interactions of Schwann cells and axons in Xenopus tadpoles with their appearance in electron micrographs.5,6 Schwann cells, which were seen moving freely between axons, did not have a basal lamina and did not resemble either fibroblasts or macrophages. These Schwann cells were ovoid in shape and had several long processes that ended in blunt expansions.6 When first interacting with axons, Schwann cells moved “inchworm” style along axonal surfaces. Movement was sporadic with alternating periods of rapid (5 ␮m/min) and no movement.5 Later, when Schwann cells spread along axons and ensheathed them,

5

Introduction

they became more spindle shaped and acquired a basal lamina.6 Subsequent tissue culture studies clearly demonstrated that axon–Schwann cell contact was required for basal lamina formation.7,8 The in vivo and electron microscopic observations of tadpole nerve fibers by BillingsGagliardi5,6 also showed that relationships between ensheathing Schwann cells and axons were more complex than those described by Speidel.34 Instead, they resembled those found in developing mouse and rat nerves.

By the end of prophase, the Schwann cell was spindle shaped and remained so through metaphase and much of anaphase. The axis of mitosis was parallel to the long axis of the cell. Radial processes that surrounded axons reappeared in telophase and extended along processes of neighboring interphase Schwann cells or the basal lamina that enclosed the family. These shape changes that occur during mitosis were thought to have a role in increasing the number of axons contacted by Schwann cells early in nerve development.22

Early Schwann Cell–Axon Relationships in Nerves

Formation and Growth of Myelin Sheaths

In developing mammalian nerves, the relationships between Schwann cells and axons change rapidly. During this complex sequence of events, axons are sorted into a population of fibers that becomes myelinated and another population that remains unmyelinated. Some geometric and quantitative aspects of these changing relationships were examined in skip serial sections of a fiber population found at the margin of the rat sciatic nerve’s posterior tibial fascicle.22,49 At birth, none of this marginal bundle’s axons was myelinated and almost all of the bundle’s transverse area was occupied by “Schwann cell families,” a term used to describe all of the axons and Schwann cell processes found within a common basal lamina. Big axon bundles were located in the center of each family, and larger axons were found more commonly at the edge of a bundle, segregated in a separate furrow, or in a 1⬊1 relationship with a Schwann cell located on the family’s outer surface. This concentric arrangement, which persisted during axon bundle subdivision and the onset of myelination, suggested that radial sorting of axons destined to be myelinated occurred in sheaths formed by longitudinal columns of Schwann cell families. The sorting sequence included initial contact with a Schwann cell process, segregation in a separate furrow of a family sheath, Schwann cell division, and establishment of a 1⬊1 relationship with one of the daughter Schwann cells, which then became isolated from the family sheath before myelination began.49 Counts during the week after birth also showed that Schwann cell families did the sorting in this population of fibers and permitted us to estimate rates for axon bundle subdivision, segregation of larger axons in separate furrows, and establishment of 1⬊1 relationships. Because approximately half of the Schwann cells that surrounded axon bundles also enveloped larger axons in a separate furrow, this process of segregation was thought to be an essential intermediate step in the establishment of the 1:1 relationship that preceded myelination.49 In order to understand the geometry and dynamics of this sorting process, we also examined the relationships of dividing Schwann cells.22 In newborn rat sciatic nerves, virtually all of the mitotic Schwann cells were located in the family sheaths described above. As mitosis began, the radial extent of the processes that surrounded axons decreased.

The electron microscopic observations and hypothesis of Geren clearly established the basic morphologic parameters of peripheral myelination.13 The mesaxon, an extension of the Schwann cell surface membrane, grows and forms a spiral sheet around the axon. Further growth and apposition of the spiral layers occur as the myelin sheath matures. Her findings, which were confirmed by many others, did not explain two important features of myelin sheath development. First, the contour of the myelin spiral is not uniform along the internode; complex variations occur.51 Second, after the compact sheath is formed, its internal circumference increases to accommodate the growing axon. Because little was known about the geometry and dimensions of the myelin spiral as it formed and grew in a Schwann cell, these parameters were studied in skip serial sections of the marginal bundle of the rat sciatic nerve’s posterior tibial fascicle.45 All of this bundle’s fibers were unmyelinated at birth, so the onset and duration of myelination were easily established. At appropriate intervals, approximate dimensions of the bundle and its largest fibers were measured and calculated at the same relative level in littermates’ nerves. These data showed that the myelin membrane’s area and transverse length increased exponentially with time; the growth rate increased rapidly during the formation of the first four to six spiral layers and remained relatively constant during the subsequent rapid enlargement of the compact sheath.45 Later more precise measurements in the developing sixth cranial nerve also showed that the myelin membrane grows exponentially during myelin formation.15 Along myelin internodes composed of two to six spiral turns, there were many variations in the number of lamellae and their contour. Near the mesaxon’s origin, longitudinal strips of cytoplasm separated the myelin layers. Thicker sheaths were larger in circumference, more circular in transverse sections and more uniform at different levels. Separation of lamellae by cytoplasm became discontinuous and generally occurred at Schmidt-Lanterman clefts. Variations in sheath contour similar to those described earlier51 became less frequent and usually were located in paranodal regions.45 These findings showed that myelin

6

Introduction

remodeling and redistribution of Schwann cell cytoplasm both occur during peripheral nerve myelination and that they are important features of this dynamic process.45 How does a Schwann cell in developing nerve rapidly increase the internal circumference, the number of compact layers, and the length of its myelin segment? There is general agreement that this is a difficult unsolved problem. Several mechanisms have been suggested and are being investigated.

and diseased peripheral nervous system has expanded rapidly. The electron microscope is still widely used, especially in correlative studies. Three recent examples concern the myelinated fiber’s molecular architecture1 and the role of an integrin receptor in myelination.10,29 These reports and the contents of this edition suggest a bright future for research that will continue to increase our understanding of the etiology, pathogenesis and treatment of peripheral nerve diseases.

Distribution of the Protein Constituents of Myelin

REFERENCES

In peripheral myelin, the major protein constituent is a glycoprotein called P0. To study its localization in Schwann cells, Trapp and his collaborators immunostained semithin Epon sections with P0 antiserum and traced the distribution of cytoplasmic staining on electron micrographs of the same Schwann cells in adjacent thin sections.38 They showed that very thin myelin sheaths in newborn trigeminal nerve were detected more easily in semithin sections stained with P0 antiserum than with other cellular stains. In developing and adult myelin-forming Schwann cells, P0 staining also was present in cytoplasmic regions occupied by Golgi profiles.38 Later P0 gene expression was studied by using light and electron microscopic in situ hybridization techniques to explore the distribution of P0 mRNA during the postnatal development of the trigeminal ganglion.19,20,48 Of particular interest was the transient detection of P0 mRNA in developing Schwann cells that surround sensory neurons, even though these cells never form myelin.19 There are two basic proteins in nerve myelin called P1 and P2 with molecular weights of 18,500 and 13,500 respectively. P1 was localized in dense line regions of compact myelin25 and, like P0, was readily detected in myelin sheaths of newborn rats.16,39 P2 is present in smaller amounts and was not detected until day 4 in a quantitative densitometric study of developing rat sixth nerves.16 However, since it was detected in 85% of myelin sheaths at day 20, the authors concluded that the sensitivity of the immunostaining method was limited and that all myelinforming Schwann cells probably expressed P2.16 Myelin-associated glycoprotein, a member of the immunoglobulin super family, is present in developing peripheral myelin.35,46 In mature sheaths, it is found in Schmidt-Lanterman clefts, paranodal areas, and periaxonal regions of myelin and Schwann cells.37

CONCLUDING COMMENT During the 30 odd years since the first edition of this reference was planned, the literature devoted to the structure, function, biochemistry and genetics of the normal

1. Arroyo, E. J., and Scherer, S. S.: On the molecular architecture of myelinated fibers. Histochem. Cell Biol. 113:1, 2000. 2. Åström, K. E.: Migration of lymphocytes through the endothelium of venules in experimental allergic neuritis. Experientia 24:589, 1968. 3. Åström, K. E., Webster, H. deF., and Arnason, B. E.: The initial lesion in experimental allergic neuritis: a phase and electron microscopic study. J. Exp. Med. 128:469, 2002. 4. Baron-Van Evercooren, A., Kleinman, H. K., Ohno, S., et al.: Nerve growth factor, laminin, and fibronectin promote neurite growth in human fetal sensory ganglia cultures. J. Neurosci. Res. 8:179, 1982. 5. Billings-Gagliardi, S.: Mode of locomotion of Schwann cells migrating in vivo. Am. J. Anat. 150:73, 1977. 6. Billings-Gagliardi, S., Webster, H. deF., and O’Connell, M. F.: In vivo and electron microscopic observations on Schwann cells in developing tadpole nerve fibers. Am. J. Anat. 141:375, 1974. 7. Bunge, M. B., Williams, A. K., and Wood, P. M.: NeuronSchwann cell interaction in basal lamina formation. Dev. Biol. 92:449, 1982. 8. Bunge, M. B., Williams, A. K., and Wood, P. M., et al.: Comparison of nerve cell and nerve cell plus Schwann cell cultures, with particular emphasis on basal lamina and collagen formation. J. Cell Biol. 84:184, 1980. 9. Descarries, L., and Schröder, J. M.: Fixation du tissu nerveux par perfusion a grand débit. J. Microscopie 7:281, 1968. 10. Feltri, M. L., Porta, D. G., Previtali, S. G., et al.: Conditional disruption of beta 1 integrin in Schwann cells impedes interactions with Schwann cells. J. Cell Biol. 156:199, 2002. 11. Fernandez-Moran, H., and Finean, J. B.: Electron microscope and low-angle x-ray diffraction studies of the nerve myelin sheath. J. Biophys. Biochem. Cytol. 3:725, 1957. 12. Fisher, C. M., and Adams, R. D.: Diphtheritic polyneuritis—a pathological study. J. Neuropathol. Exp. Neurol. 15:243, 1956. 13. Geren, B. B.: The formation from the Schwann cell surface of myelin in the peripheral nerves of chick embryos. Exp. Cell Res. 7:558, 1954. 14. Gutrecht, J. A., and Dyck, P. J.: Quantitative teased-fiber and histologic studies of human sural nerve during postnatal development. J. Comp. Neurol. 138:117, 1970. 15. Hahn, A. F., Chang, Y., and Webster, H. deF.: Development of myelinated nerve fibers in the sixth cranial nerve of the rat: a quantitative electron microscope study. J. Comp. Neurol. 260:491, 1987.

Introduction 16. Hahn, A. F., Whitaker, J. N., Kachar, B., and Webster, H. deF.: P2, P1, and P0 myelin protein expression in developing rat sixth nerve: a quantitative immunocytochemical study. J. Comp. Neurol. 260:501, 1987. 17. Kleinman, H. K., Cannon, F. B., Laurie, G. W., et al.: Biological activities of laminin. J. Cell. Biochem. 27:317, 1985. 18. Lampert, P. W.: Mechanism of demyelination in experimental allergic neuritis. Lab. Invest. 20:127, 1969. 19. Lamperth, L., Manuelidis, L., and Webster, H. deF.: Non myelin-forming perineuronal Schwann cells in rat trigeminal ganglia express P0 myelin glycoprotein mRNA during postnatal development. Mol. Brain Res. 5:177, 1989. 20. Lamperth, L., Manuelidis, L., and Webster, H. deF.: P0 glycoprotein mRNA distribution in myelin-forming Schwann cells of the developing rat trigeminal ganglion. J. Neurocytol. 19:756, 1990. 21. Le Douarin, N.: The Neural Crest, 2nd ed. Cambridge, UK, Cambridge University Press, 1999. 22. Martin, J. R., and Webster, H. deF.: Mitotic Schwann cells in developing nerve: their changes in shape, fine structure and axon relationships. Dev. Biol. 32:417, 1973. 23. Napolitano, L. M., and Scallen, T. J.: Observations on the fine structure of peripheral nerve myelin. Anat. Rec. 163:1, 1969. 24. Ochoa, J., and Mair, W. G. P.: The normal sural nerve in man. I. Ultrastructure and numbers of fibres and cells. Acta Neuropathol. (Berl.) 13:197, 1969. 25. Omlin, F. X., Webster, H. deF., Palkovits, C. G., and Cohen, S. R.: Immunocytochemical localization of basic protein in major dense line regions of central and peripheral myelin. J. Cell Biol. 95:242, 1982. 26. Peters, A., and Muir, A. R.: The relationship between axons and Schwann cells during development of peripheral nerves in the rat. Q. J. Exp. Physiol. 44:117, 1959. 27. Peters, A., Palay, S. L., and Webster, H. deF.: The Fine Structure of the Nervous System: The Cells and Their Processes, New York, Harper & Row, 1970. 28. Peterson, R. G., and Pease, D. C.: Myelin embedded in polymerized gluteraldehyde-urea. J. Ultrastruct. Res. 41:115, 1972. 29. Podratz, J. L., Rodriguez, E., and Windebank, A. J.: Role of the extracellular matrix in myelination of peripheral nerve. Glia 35:35, 2001. 30. Robertson, J. D.: The ultrastructure of adult vertebrate peripheral myelinated fibers in relation to myelinogenesis. J. Biophys. Biochem. Cytol. 1:271, 1955. 31. Robertson, J. D.: The ultrastructure of Schmidt-Lanterman clefts and related shearing defects of the myelin sheath. J. Biophys. Biochem. Cytol. 4:39, 1958. 32. Robertson, J. D.: Preliminary observations on the ultrastructure of nodes of Ranvier. Z. Zellforsch. 50:553, 1959. 33. Robertson, J. D.: Origin of the unit membrane concept. Protoplasma 63:218, 1967. 34. Speidel, C. C.: In vivo studies of myelinated nerve fibers. Int. Rev. Cytol. 16:173, 1964. 35. Sternberger, N. H., Quarles, R. H., Itoyama, Y., and Webster, H. deF.: Myelin-associated glycoprotein demonstrated immunocytochemically in myelin and myelinforming cells of developing rats. Proc. Natl. Acad. Sci. U. S. A. 76:1510, 1979.

7

36. Thomas, P. K.: The quantitation of nerve biopsy findings. J. Neurol. Sci. 11:285, 1970. 37. Trapp, B. D., Andrews, S. B., Wong, A., et al.: Colocalization of the myelin-associated glycoprotein and the microfilament components, F-actin and spectrin, in Schwann cells of myelinated nerve fibres. J. Neurocytol. 18:47, 1989. 38. Trapp, B. D., Itoyama, Y., Sternberger, N. H., et al.: Immunocytochemical localization of P0 protein in Golgi complex membranes and myelin of developing rat Schwann cells. J. Cell Biol. 90:1, 1981. 39. Trapp, B. D., Mclntyre, L. J., Quarles, R. H., et al.: Immunocytochemical localization of rat peripheral nervous system myelin proteins: P2 protein is not a component of all peripheral nervous system myelin sheaths. Proc. Natl. Acad. Sci. U. S. A. 76:3552, 1979. 40. Tsukita, S., and Ishikawa, H.: The movement of membranous organelles in axons. J. Cell Biol. 84:513, 1980. 41. Uzman, B. G., and Nogueira-Graf, G.: Electron microscope studies of the formation of nodes of Ranvier in mouse sciatic nerves. J. Biophys. Biochem. Cytol. 3:589, 1957. 42. Waksman, B. H., and Adams, R. D.: A comparative study of experimental allergic neuritis in the rabbit, guinea pig and mouse. J. Neuropathol. Exp. Neurol. 15:293, 1956. 43. Waksman, B. H., Adams, R. D., and Mansmann, H. C.: Experimental study of diphtheritic polyneuritis in the rabbit and guinea pig. J. Exp. Med. 105:591, 1957. 44. Webster, H. deF.: Transient, focal accumulation of axonal mitochondria during the early stages of wallerian degeneration. J. Cell Biol. 12:361, 1962. 45. Webster, H. deF.: The geometry of peripheral myelin sheaths during their formation and growth in rat sciatic nerves. J. Cell Biol. 48:348, 1971. 46. Webster, H. deF.: Development of peripheral nerve fibers. In Dyck, P. J., Thomas, P. K., Griffen, J. W., et al.: (eds.): Peripheral Neuropathy, 3rd ed, Philadelphia, W. B. Saunders, p. 243–266, 1993. 47. Webster, H. deF., and Collins, G. H.: Comparison of osmium tetroxide and glutaraldehyde perfusion fixation for the electron microscopic study of the normal rat peripheral nervous system. J. Neuropathol. Exp. Neurol. 23:109, 1964. 48. Webster, H. deF., Lamperth, L., Favilla, J. T., et al.: Use of a biotinylated probe and in situ hybridization for light and electron microscopic localization of P0 mRNA in myelin-forming Schwann cells. Histochemistry. 86:441, 1987. 49. Webster, H. deF., Martin, J. R., and O’Connell, M. F.: The relationships between interphase Schwann cells and axons before myelination: a quantitative electron microscopic study. Dev. Biol. 32:401, 1973. 50. Webster, H. deF., Schröder, J. M., Asbury, A. K., and Adams, R. D.: The role of Schwann cells in the formation of “onion bulbs” found in chronic neuropathies. J. Neuropathol. Exp. Neurol. 26:276, 1967. 51. Webster, H. deF., and Spiro, D.: Phase and electron microscopic studies of experimental demyelination. I. Variations in myelin sheath contour in normal guinea pig sciatic nerve. J. Neuropathol. Exp. Neurol. 19:42, 1960. 52. Webster, H. deF., Spiro, D., Waksman, B. H., and Adams, R. D.: Phase and electron microscopic studies of experimental

8

Introduction

demyelination. II. Scwann cell changes in guinea pig sciatic nerves during experimental diphtheritic neuritis. J. Neuropathol. Exp. Neurol. 20:5, 1961. 53. Zhang, Q.-L., Lin, P.-X., Chang, Y., and Webster, H. deF.: Effects of nerve segment supernatants on cultured Schwann cell proliferation and laminin production. J. Neurosci. Res. 37:612, 1994.

54. Zhang, Q. L., Lin, P. X., Shi, D., et al.: Vasoactive intestinal peptide: mediator of laminin synthesis in cultured Schwann cells. J. Neurosci. Res. 43:496, 1996. 55. Zhang, Q.-L., Liu, J., Lin, P.-X., and Webster, H. deF.: Local administration of vasoactive intestinal peptide after nerve transection accelerates early myelination and growth of regenerating axons. J. Peripher. Nerv. Syst. 7:118, 2002.

2 Gross Anatomy of the Peripheral Nervous System ERNEST D. GARDNER AND RICHARD P. BUNGE

General Features Spinal and Peripheral Nerves Autonomic Nervous System

Gross Anatomy Head, Neck, and Upper Limb Thorax

By common definition, the peripheral nervous system includes the cranial nerves, the spinal nerves with their roots and rami, the peripheral nerves, and the peripheral components of the autonomic nervous system. This chapter presents the principles of organization and distribution of the peripheral nervous system, and outlines its major gross anatomic features; the cranial nerves and their autonomic components are considered in other chapters. Books and atlases that provide more detailed information, most of which have valuable bibliographies, are included in the references.18,26,28,53,56,57,69,72 Illustrations are limited to those that present concepts of the arrangement and distribution of spinal and peripheral nerves. The terminology used follows the Nomina Anatomica,49 translated into English where appropriate.

GENERAL FEATURES Spinal and Peripheral Nerves The dorsal and ventral roots are attached to the spinal cord by a series of filaments. After traversing the subarachnoid space, each root enters a dural pouch, or sac, and then a dural sheath, the sac being subdivided by a septum. Immediately peripheral to the spinal ganglion of the dorsal root, corresponding dorsal and ventral roots join to form a spinal nerve. The dural sheaths become confluent at the ganglion, and then merge with the epineurium of the spinal nerve. Each spinal nerve quickly

Abdomen, Pelvis, and Lower Limb Back

divides into a dorsal and a ventral (primary) ramus. The dorsal rami supply the back; the ventral rami supply the limbs and ventrolateral part of the body wall. In the cervical and lumbosacral regions, the ventral rami intermingle and form plexuses from which the major peripheral nerves emerge. Cranial nerves are also included in the term peripheral nerve. Although the funicular organization and connective tissue elements of nerves are described in other chapters, some general aspects are worth noting here. The sheaths convey the intrinsic blood and lymphatic vessels as well as the nervi nervorum, which supply the connective tissue and vessels with sensory and autonomic fibers. The sheaths also impart strength, especially against tensile stresses (see Chapter 3). Spinal roots have thinner and less well-defined sheaths and are therefore more fragile. The funicular arrangement of nerves is a matter of considerable importance, especially in connection with injury, surgical repair, and regeneration. Sunderland has described these arrangements in detail, especially with respect to the variation in number and size of funiculi, their intraneural course, and their redistribution through the formation of funicular plexuses.64 Cranial nerves differ significantly from spinal nerves, especially in their mode of embryologic development and their relation to the special senses, and because some cranial nerves supply branchial arch structures. They are attached to the brain at irregular rather than regular intervals; they are not formed of dorsal and ventral roots; some have more than one ganglion, whereas others have none; 11

12

Structure of the Peripheral Nervous System

and the optic nerve is a fiber tract of the central nervous system and not a peripheral nerve. Distribution of Spinal and Peripheral Nerves When the ventral ramus of a spinal nerve enters a plexus and joins other such rami, its component funiculi ultimately enter several of the peripheral nerves emerging from the plexus. Thus, as a general principle, each ramus entering a plexus contributes to several peripheral nerves, and each such peripheral nerve contains fibers derived from several ventral rami. Thus each spinal nerve has a pattern of ultimate distribution referred to as segmental or dermatomal, in contrast to that characteristic of peripheral nerves20,24,64 (Fig. 2–1). The term segmental refers to the fact that the longitudinal extent of spinal cord to which a right and left pair of spinal roots is attached constitutes a segment of the spinal cord. The term dermatome refers to the area of skin supplied by the sensory fibers of a single dorsal root through the dorsal and ventral rami of its spinal nerve. It is evident, then, that because most dorsal rami have cutaneous branches, the area of skin supplied by just a ventral ramus is usually not a complete dermatome. The mixture of nerve fibers in plexuses is such that it is difficult, if not impossible, to trace their course by dissection, and dermatomal distribution has been determined by physiologic experimentation and by studying disorders and

FIGURE 2–1 Schematic diagram of spinal and peripheral nerve distribution. Only sensory fibers to the skin are represented. Two nerve fibers of spinal nerve A are shown entering a plexus. One of the fibers joins peripheral nerve X, and the other joins peripheral nerve Y. Two fibers of spinal nerve B also join the two peripheral nerves. Thus the areas supplied by the two spinal nerves are different from the areas supplied by the two peripheral nerves, as shown in the subdivided rectangle. (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

surgical sections of spinal roots and nerves. The results of such studies have yielded complex maps, chiefly because of variation, overlap, and differences in method. Variation results from intrasegmental rootlet anastomoses adjacent to the cervical and lumbosacral portions of the spinal cord50 and from individual differences in plexus formation and peripheral nerve distribution. Overlap is such that section of a single root does not produce complete anesthesia in the area supplied by that root; at the most, some degree of hypalgesia may result. Overlap is greater for touch than for pain—hence the more common occurrence of hypalgesia rather than hypesthesia after section of a single root. Problems of interpretation also result from differences in method. As pointed out by Pearson and co-workers, the fundamental problem is one of segmentation and involves the question: Do certain dermatomes extend as a series of bands from the median plane of the back into the limbs, and are these bands of dermatomes arranged in an uninterrupted sequence?51 A full discussion of this problem and its embryologic basis is beyond the scope of this chapter. Nevertheless, it is important to note that the maps published by Keegan and Garrett indicate that all dermatomes, from C2 to S1, form an uninterrupted, simply arranged series extending from the median plane of the back.35 Their data were based largely on the detection of hypalgesia following compression of a single nerve root by a herniated nucleus pulposus. The degree of compression was usually not verified. Valuable as Keegan and Garrett’s maps might be as diagnostic aids in such disorders, it must be emphasized that their methods yield zones of hypalgesia that are not complete dermatomes. Moreover, as Pearson and colleagues51 pointed out, the discussion presented by Keegan and Garrett35 contains significant conceptual defects, and some of their data are at serious odds with anatomic arrangements. In contrast, Foerster published data on dermatomes that, while admittedly incomplete, derived from studies based on sound physiologic methods as follows13,14: 1. Method of residual sensitivity—a single root was left intact, as contiguous roots above and below were divided. This gave the total distribution of the intact dorsal root. 2. Constructive method—a series of contiguous roots was divided; consequently, the superior border of the resulting anesthesia represented the inferior border of the dermatome corresponding to the next higher intact root, and the inferior border of the anesthetic area represented the superior border of the next lower intact root. 3. Stimulation of dorsal roots—electrical stimulation yielded vasodilatation in areas of skin that were smaller than the dermatomes determined by the residual method, but that were similar in shape and

Gross Anatomy of the Peripheral Nervous System

general location. These areas resembled the zones reported by Head and Campbell in their classic study of the distribution of herpes zoster.25 The maps shown in Figures 2–2 and 2–3 are based on Foerster’s study of a large number of human patients in which he used combinations of the methods just outlined; they have been shown to be useful clinically.12–14 It must, however, be emphasized that the maps depict presumably normal dermatomes, not areas of sensory deficit. This

FIGURE 2–2 Front view of dermatomal distribution, based on Foerster’s data.13,14 The right and left halves show the total distribution of alternating dermatomes, thereby illustrating the degree of overlap. In some regions, however, overlap is very great and more than two dermatomes may be involved. Hence, additional figures (insets) are necessary to show the total distribution of certain dermatomes. (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

13

statement is qualified because spinal cord sensitivity may be altered by root sections in such a way as to affect the size of the dermatome being tested by peripheral stimulation. It is of interest that there is little correspondence between dermatomes and underlying muscles. Many tables have been published that list the segmental supply of muscles or of the segments controlling movements at joints. Schemes have been proposed that attempt to bring some logical proximodistal order into the segmental

14

Structure of the Peripheral Nervous System

enlargements supply the more proximal limb muscles (note, however, that segments C5 to C8 are involved in shoulder movements), and that the more caudal segments supply the more distal muscles (T1 is the chief motor supply of the intrinsic muscles of the hand). A muscle usually receives fibers from each of the spinal nerves that enter the peripheral nerve supplying it (although one spinal nerve may be its chief supply), and section of a single spinal nerve weakens but does not usually paralyze a muscle. Finally, published diagrams that illustrate the segmental sensory supply of bones and joints are based chiefly on dissections and clinical observations. Their clinical value is doubtful, especially in view of the diffuse and often referred nature of pain from these deep structures. The distribution of peripheral nerves in the limbs is fundamentally different from that of spinal nerves. These enable one to contrast cutaneous nerve supply (Figs. 2–4 and 2–5) with the dermatomes (Figs. 2–2 and 2–3). In the trunk, however, peripheral and spinal nerves are usually identical in their cutaneous distribution. Peripheral nerves to the limbs also overlap, but to a lesser extent than do spinal nerves. Thus, if a peripheral nerve is cut, the muscles supplied by the nerve are greatly weakened or completely paralyzed, autonomic dysfunction occurs, and sensation is lost in the central part of the area of distribution of the nerve and diminished at the edges of its area of supply. The last is due to overlap from adjacent peripheral nerves; the overlap is less than in the case of spinal nerves and is often less for touch than for pain. Peripheral nerves vary in their course and distribution, and adjacent nerves may communicate with each other.18,26 Such communications sometimes account for unexpected residual sensation or movement after section of a nerve. Whatever the pattern of distribution, major peripheral nerves usually have five general kinds of branches:

FIGURE 2–3 Back view of dermatomes, based on Foerster’s data13,14 as explained in legend for FIGURE 2–2. (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

1. Muscular—motor, sensory, and autonomic fibers; the sensory fibers are from muscle fibers and associated connective tissue and tendons and sometimes joints 2. Cutaneous or mucosal—each has sensory and autonomic fibers; the cutaneous branches often include fibers from subjacent joints, ligaments, and tendons, especially in the case of digital nerves 3. Articular—arising where the nerve crosses a joint and containing sensory and autonomic fibers 4. Vascular—sensory and autonomic to adjacent blood vessels 5. Terminal—one, several, or all of the foregoing

supply of limb muscles.44 These tables and schemes, however, tend to be limited in value because of variation, overlap, and incomplete information about segmental motor origin and distribution. The general arrangement is that the more rostral segments of the cervical and lumbosacral

Also of importance, in addition to peripheral distribution, are the order and site of origin of individual branches, including distances to the muscles they supply, as well as the order in which structures are innervated, and the nature and extent of variations.63,66–68 Moreover, in several instances peripheral nerves pass through osteofibrous

Gross Anatomy of the Peripheral Nervous System

15

FIGURE 2–4 Approximate areas of cutaneous nerve distribution illustrated in the right upper limb. Neither variation nor overlap is shown. (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

canals, where they may be subject to compression. The chief sites of compression and related literature are summarized by Smorto and Basmajian.61

Autonomic Nervous System Certain general principles of organization are presented here to clarify the relationships of spinal and peripheral nerves with the autonomic nervous system. By classic anatomic definition, the autonomic nervous system, sometimes called the visceral or vegetative nervous system, is that system of motor nerve fibers that supplies cardiac muscle, smooth muscle, and glands and consists, in its simplest form, of a pathway of two succeeding nerve cells. The first cell is located in the brain or spinal cord, the second in a ganglion outside the brain and spinal cord. Anatomically and functionally, however, the autonomic nervous system is much more complex than the simplistic definition just given would indicate, to a degree far beyond the scope of this chapter. The autonomic pathway or outflow begins with certain nerve cells in the brainstem and spinal cord. The axons of these cells, termed preganglionic fibers, leave the brainstem and spinal cord over certain cranial nerves and ventral roots and synapse in peripheral autonomic ganglia (including the suprarenal medullae and certain chromaffin cells). The axons of the ganglion cells are termed postganglionic fibers and are distributed to cardiac muscle, smooth muscle, and certain gland cells. The locations, arrangements, connections, and patterns of distribution of

the preganglionic and postganglionic autonomic fibers are grouped, on the basis of an anatomic classification of subdivisions of the pathways, into sympathetic, parasympathetic, and enteric systems or divisions. Most viscera are supplied by both sympathetic and parasympathetic divisions; the enteric system is limited to the wall of the bowel. Works cited in the references, several of which contain valuable bibliographies, provide additional information about the autonomic nervous system.9,28,38,40,46,54 Sympathetic (Thoracolumbar) System This part of the autonomic nervous system comprises the preganglionic fibers that issue from the thoracic and upper lumbar levels of the spinal cord. These fibers travel in ventral roots and spinal nerves to reach peripheral sympathetic ganglia, where they synapse with ganglion cells. The locations of these ganglia are related to the embryonic migration of cells from the neural tube and neural crest, which form the ganglia of the sympathetic trunk and prevertebral plexuses. Sympathetic Trunk and Ganglia. Most preganglionic fibers leave the spinal nerves or ventral rami and reach the adjacent sympathetic trunk and ganglia by way of rami communicantes (Fig. 2–6). The sympathetic trunks are long nerve strands, one on each side of the vertebral column, extending from the base of the skull to the coccyx. Each usually contains 21 to 25 ganglia of varying sizes, but broader ranges have been recorded.

16

Structure of the Peripheral Nervous System

FIGURE 2–5 Approximate areas of cutaneous nerve distribution to the right lower limb. Neither variation nor overlap is shown. (From Gardner, E., Gray, D. J., and O’Rahilly, R.:Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

Many of the preganglionic fibers that enter the sympathetic trunks synapse in the ganglia of the trunks and in accessory ganglia; those that do not synapse continue through and reach ganglia of the prevertebral plexuses. Of the postganglionic fibers arising in the trunk ganglia, some go directly to adjacent viscera and blood vessels; the others return to spinal nerves and dorsal and ventral rami by way of rami communicantes. The fibers of both the ventral and dorsal6 rami eventually supply secretory fibers to sweat glands, motor fibers to the smooth muscle of the hair follicles (arrectores pilorum), and motor fibers to the smooth muscle of the blood vessels of the limbs and walls of the trunk. However, some of the

postganglionic fibers to the back and to the proximal parts of the limbs reach these parts by accompanying blood vessels. Each trunk ganglion has one to four rami communicantes, which connect it with the corresponding nerve and often with the nerve above or below. The ramus or rami containing the most postganglionic fibers tend to connect with the corresponding nerve (each spinal nerve or one of its rami receives such fibers). In the thorax, the rami containing the most preganglionic fibers are more oblique in direction (coming from the spinal nerve above or below), and they are more lateral in position, that is, farther from the spinal cord.55

Gross Anatomy of the Peripheral Nervous System

17

viscera or blood vessels. The preganglionic fibers that enter these ganglia are those that traverse the sympathetic trunks without synapsing. They reach the prevertebral ganglia by way of branches termed splanchnic nerves (the term splanchnic is also applied to certain visceral branches in the pelvis). These nerves may contain ganglia along their course. The postganglionic fibers from these various ganglia go directly to adjacent viscera and blood vessels. These ganglia, their preganglionic and postganglionic fibers, and the thoracic and abdominal branches of the vagus nerves form what are termed the prevertebral plexuses. Many sensory fibers from viscera (for pain as well as reflexes) traverse these plexuses and reach the spinal cord by way of splanchnic nerves, rami communicantes, and dorsal roots, or the brainstem by way of the vagus nerves. Some of the migrating embryonic cells form the chromaffin system, in particular the medullae of the adrenal glands. Correspondingly, some of the preganglionic fibers in the splanchnic nerves end in relation to these cells. Accessory Ganglia. Some of the migrating embryonic cells stop along spinal nerves, ventral rami, and rami communicantes, especially in the cervical, lower thoracic, and upper lumbar levels. Here they form scattered sympathetic cells, which often are collected into definite accessory (intermediate) ganglia.60,76 The postganglionic fibers from these cells continue in their associated nerves; hence sympathectomies in these regions may not be completely effective.48

FIGURE 2–6 The sympathetic trunks. Preganglionic rami communicantes are shown as interrupted lines, postganglionic rami as solid black. Based on studies by Pick and Sheehan.55 (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

Prevertebral Plexuses and Ganglia and Chromaffin System. Many embryonic ganglion cells migrate to a position in front of the vertebral column, where they form ganglionic masses that are named according to adjacent

Parasympathetic (Craniosacral) System This part of the autonomic system comprises the preganglionic fibers that issue from the brainstem by way of the 3rd, 7th, 9th, 10th, and 11th cranial nerves, and from the sacral cord by way of its second and third or third and fourth ventral roots. The ganglion cells are usually in or near the organ to be innervated; hence, the postganglionic fibers are short. Moreover, none seems to go to the blood vessels, smooth muscle, or glands of the limbs or body walls. The cranial nerves are described elsewhere, but it is worth noting that the parasympathetic ganglia of the third, seventh, and ninth cranial nerves form what are termed the cephalic or cranial parasympathetic ganglia (ciliary, pterygopalatine, otic, and submandibular). Their postganglionic fibers supply the eye, lacrimal and salivary glands, and mucous and serous glands of the oral and nasal cavities. The preganglionic fibers of the 11th nerve are distributed with the vagus nerves; the ganglion cells are near or in the walls of the viscera of the neck, thorax, and abdomen. As mentioned earlier, vagal branches contribute to the prevertebral plexuses.

18

Structure of the Peripheral Nervous System

The sacral preganglionic parasympathetic fibers leave the sacral plexus as pelvic splanchnic nerves, enter the inferior hypogastric plexus, and reach ganglion cells in the walls of pelvic organs. The Enteric System In his extensive treatise on the autonomic nervous system, Langley classified the enteric plexuses of the gut as a third autonomic division.43 Langley recognized that the enteric nervous system receives sympathetic and parasympathetic inputs, but he considered this extensive ganglionic network relatively autonomous in its control of gut function. More recent studies have established the remarkable extent and complexity of the enteric nervous system. Furness and Costa have estimated that there may be 108 ganglion cells in the gut wall of the human,15 and Gershon has reviewed evidence that the many types of neurons present employ at least three known neurotransmitters plus as many as nine putative transmitters or modulators.19 Physiologic aspects of visceral functional control are considered in Chapter 13. The enteric nervous system extends the length of the gastrointestinal tract from esophagus to rectum. It is composed of two major plexuses of ganglion cells, as well as interconnecting fibers. The more internally located neurons (Meissner’s plexus) are located in the submucosa of the gut wall; the more externally located neurons (Auerbach’s or the myenteric plexus) lie between the external longitudinal and internal circular smooth muscle layers of the muscularis externa. Neurons occur in these regions as small aggregates; these aggregates are interconnected by fascicles of unmyelinated nerve fibers, and extensive connections are also found between the submucosal neurons and the myenteric components. Certain histologic aspects of these plexuses are unusual. The small ganglia are entirely nonvascular, and the cytoarchitecture is more like that of the central nervous system than of peripheral sensory or autonomic ganglia.16,19 The similarity to the central nervous system derives from the lack of space and extracellular matrix material between the individual cells of the ganglia and the diversity of neuronal and synaptic types present. Details of the histology and cytology of this unique portion of the autonomic nervous system are included in Chapter 12.

GROSS ANATOMY The gross anatomy of the peripheral nervous system of each region of the body and the distribution and branching of nerves are outlined here, with emphasis on certain regional or topographic features. Additional special details are to be found in references dealing with peripheral nerve injuries.24,58,61,64

Head, Neck, and Upper Limb Cervical Plexus The ventral rami of the upper four cervical nerves unite to form the cervical plexus; those of the lower four, together with the greater part of that of the first thoracic, join to form the brachial plexus. (The distribution of the dorsal rami of the cervical nerves is described later.) Each cervical ramus receives one or more rami communicantes (containing postganglionic fibers) from a cervical sympathetic ganglion, and often from the vertebral nerve and plexus as well. The cervical plexus is arranged as an irregular series of loops located in front of the levator scapulae and scalenus medius muscles, under cover of the sternocleidomastoid muscle and internal jugular vein. The branches of the loops are superficial (to the skin of the back of the head, the neck, and the shoulder) and deep (to certain neck muscles and the diaphragm). Superficial Branches. These cutaneous branches emerge near the middle of the posterior border of the sternocleidomastoid. They include the lesser occipital nerve, which ascends behind the ear and supplies some of the skin on the side of the head and on the cranial surface of the ear; the great auricular nerve, which ascends obliquely across the sternocleidomastoid muscle to supply the skin over the parotid gland, over the mastoid process, and on both surfaces of the ear; the transverse cervical nerves, which supply the skin on the side and front of the neck; and the supraclavicular nerves, derived from a common trunk that divides into anterior, middle, and posterior supraclavicular nerves, which supply the skin over the shoulder and the front of the thorax. Deep Branches. These are the ansa cervicalis, the phrenic nerve, and the muscular branches to the sternocleidomastoid, trapezius, levator scapulae, scalene, and prevertebral muscles. There are also small communicating branches to the 10th, 11th, and 12th cranial nerves. The ansa cervicalis (ansa hypoglossi) is a loop formed by fibers of the first three (or the second and third) cervical nerves. It presents a superior root (the so-called descending branch of the hypoglossal nerve), which connects it with the hypoglossal nerve, but which consists of fibers from the second or first cervical nerve, and an inferior root (nervus descendens cervicalis), which connects it with branches from the second and third cervical nerves. The ansa cervicalis supplies the infrahyoid muscles. The thyrohyoid, however, receives its cervical fibers by way of the hypoglossal nerve. The phrenic nerve, which supplies the diaphragm, arises chiefly from the fourth cervical nerve, but commonly has a root from the fifth as well, and sometimes

Gross Anatomy of the Peripheral Nervous System

from the third. Rarely, it may arise entirely from the accessory phrenic nerve. It descends in front of the scalenus anterior muscle, enters the thorax, and descends between the pericardium and mediastinal pleura. The right phrenic nerve passes in front of the root of the right lung, pierces the diaphragm near the opening for the inferior vena cava (or traverses the opening for that vessel), and distributes most of its motor fibers from below. The left phrenic nerve passes in front of the root of the left lung and pierces the diaphragm immediately to the left of the pericardium. Each nerve carries motor fibers to the diaphragm and sensory and autonomic fibers for the diaphragm, pleura, and peritoneum. Referred pain from the area of supply of a phrenic nerve is commonly felt in the skin over the trapezius (C4 and C5). Pain is sometimes referred to the region of the ear; this is probably related to a contribution from the third cervical nerve. The accessory phrenic nerve is present in about one third of instances. It usually arises from the fifth cervical nerve through the nerve to the subclavius, but may arise from the cervical plexus or from a cardiac branch of a cervical sympathetic ganglion. The accessory nerve runs a variable course before joining the phrenic nerve in the thorax. If such a nerve is present, section of the phrenic nerve in the neck will not paralyze the corresponding half of the diaphragm completely. Brachial Plexus The brachial plexus, which is situated partly in the neck and partly in the axilla, is formed by the ventral rami of the lower four cervical nerves and the greater part of the ventral ramus of the first thoracic nerve. These rami lie first between the scalenus anterior and the scalenus medius, and then in the posterior triangle of the neck. Here the plexus is situated above the clavicle, posterior and lateral to the sternocleidomastoid. It lies above and behind the third part of the subclavian artery and is crossed by the inferior belly of the omohyoid. In this situation, the plexus may be injected with a local anesthetic. The general features of the plexus are discussed in greater detail elsewhere (see Chapter 55).11,22,23,36 The plexus descends behind the concavity of the medial two thirds of the clavicle, and accompanies the axillary artery under cover of the pectoralis major. It is enclosed with the axillary vessels in the axillary sheath, and its cords are arranged around the second part of the axillary artery behind the pectoralis minor. The terminal branches of the plexus arise at the inferolateral border of the pectoralis minor. The brachial plexus can be marked on the surface by a line from the posterior margin of the sternocleidomastoid at the level of the cricoid cartilage to the midpoint of the clavicle. The plexus can be palpated both above and below the omohyoid, in the angle between the clavical and the sternocleidomastoid.

19

The common, though not invariable, arrangement of branches is as follows. The ventral rami of the fifth and sixth cervical nerves unite to form the upper trunk, that of the seventh remains single as the middle trunk, and the eighth cervical and first thoracic rami form the lower trunk. Each trunk then divides into an anterior and a posterior division (for the front and back of the limb, respectively). The anterior divisions of the upper and middle trunks unite to form the lateral cord, the anterior division of the lower trunk forms the medial cord, and the three posterior divisions form the posterior cord. The terminal branches arise from the three cords. The brachial plexus is thus composed successively of (1) ventral rami and trunks that lie in the neck in relation to the subclavian artery (the lowest trunk lies on the first rib behind the subclavian artery), (2) divisions that lie behind the clavicle, and (3) cords and branches that lie in the axilla in relation to the axillary artery. Variations. The brachial plexus frequently receives contributions from the fourth cervical or the second thoracic nerve also, or from both. When the contribution from the fourth cervical is large and that from the first thoracic small, the plexus is described as being prefixed in relation to the vertebral column. When the contributions from the first and second thoracic nerves are large, the plexus is termed postfixed; this latter situation is prominent when the first rib is rudimentary.3,10 It is, however, uncommon to find prefixation or postfixation in the sense of a complete shift in which a full nerve is gained at one end while one is completely lost at the other end. Other variations (in gross form, component arrangements, and branching) are common. Branches of the Ventral Rami. These are the dorsal scapular and long thoracic nerves, and twigs to the scalene and longus colli muscles. The dorsal scapular nerve (chiefly C5) sometimes supplies the levator scapulae but is distributed chiefly to the rhomboids. The long thoracic nerve arises by three roots from C5 to C7, descends behind the brachial plexus, and enters the external surface of the serratus anterior. Branches of the Trunks. These are the nerves to the subclavius and the suprascapular nerve and occasionally medial and lateral pectoral nerves also. The nerve to the subclavius (C5) descends behind the clavicle, in front of the brachial plexus to supply the subclavius muscle and the sternoclavicular joints. It often sends fibers to the phrenic nerve by a communicating branch, the accessory phrenic nerve described earlier. The suprascapular nerve (C5, C6) passes through the scapular notch, supplies the acromioclavicular and shoulder joints and the supraspinatus muscle, and then passes through the spinoglenoid notch to end in the infraspinatus.

20

Structure of the Peripheral Nervous System

Branches of the Cords. These include a number of cutaneous and muscular branches and the important terminal branches, namely the median, ulnar, radial, musculocutaneous, and axillary nerves. Several lateral pectoral nerves (C5 to C7) arise from the lateral cord (or upper and middle trunks) and supply both pectoral muscles as well as the acromioclavicular and shoulder joints. Medial pectoral nerves (C8, T1) from the medial cord (or lower trunk) supply both pectoral muscles. Also arising from the medial cord is the medial antebrachial cutaneous nerve (C8, T1), which descends with the brachial artery to the lower part of the arm, where it becomes cutaneous. It divides into anterior and ulnar branches, which supply the skin of the medial half of the forearm. The medial brachial cutaneous nerve (chiefly T1), also a branch of the medial cord, is a small nerve that supplies the medial and posterior aspects of the arm and communicates with the intercostobrachial nerve. From the posterior cord, there arise the upper subscapular nerve (or nerves) (C5), for the subscapularis; the thoracodorsal nerve (C7, C8), for the latissimus dorsi; and the lower subscapular nerve (or nerves) (C5, C6), for the subscapularis and teres major. The median nerve (C5, C6 to C8, T1) arises from the medial and lateral cords by medial and lateral roots, which unite in a variable fashion (Fig. 2–7). Descending as a part of the neurovascular bundle, it enters the cubital fossa, where it lies under cover of the bicipital aponeurosis and supplies the elbow joint. It then passes between the two heads of the pronator teres and descends on the deep surface of the flexor digitorum superficialis. It is indicated on the surface by a line down the middle of the forearm to the midpoint between the styloid processes. Just above the flexor retinaculum, it is quite superficial in the interval between the flexor carpi radialis and palmaris longus tendons, and completely so if the latter muscle is absent. The median nerve enters the hand by passing through the carpal canal, behind the flexor retinaculum. It then spreads out in an enlargement and divides into its terminal branches under cover of the palmar aponeurosis and the superficial palmar arch. Usually it divides first into lateral and medial portions or divisions. The median nerve has no muscular branches in the upper arm (i.e., above the elbow). In the forearm, it supplies all the muscles of the front of the forearm except the flexor carpiulnaris and the medial half of the flexor digitorum profundus. In the cubital fossa, a bundle of muscular branches is given to the pronator teres, flexor carpi radialis, palmaris longus, and flexor digitorum superficialis. The anterior interosseous nerve also arises in the cubital fossa. It descends on the front of the interosseous membrane, supplies the flexor pollicis longus and flexor digitorum profundus (lateral part), passes behind and supplies the pronator quadratus, and ends in twigs to the wrist and intercarpal joints. In the lower part of the

FIGURE 2–7 Schematic representation of the sequences of muscular and cutaneous branches of the musculocutaneous and median nerves. The black dot in each instance represents a muscle or a muscle group. The sequences of muscular branches are the more common ones. Based on studies by Sunderland and Ray.68 (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

forearm, the median nerve gives off an inconstant palmar branch for the supply of a small area of the skin of the palm. The median and ulnar nerves may communicate in the forearm, and the anterior interosseous nerve may communicate with the ulnar nerve. In the hand, the lateral division gives off an important muscular branch (recurrent branch) for the abductor pollicis, flexor pollicis brevis, and opponens pollicis, which anastomoses with the deep branch of the ulnar nerve.21 The lateral division then divides in such a way as to furnish three palmar digital nerves for both sides of the thumb and the lateral aspect of the index finger and the first lumbrical. The medial division divides in such a way as to furnish four palmar digital nerves for the adjacent sides of the index and middle fingers, middle and ring fingers, and second lumbrical. All digital nerves, near their terminations, send branches dorsally to the backs of the distal parts of the fingers, and all digital nerves, in both hand and foot

Gross Anatomy of the Peripheral Nervous System

supply ligaments, joints, and tendons of the digits, as well as skin. Rarely the median nerve supplies the first dorsal interosseous or the adductor pollicis. As a general rule, the median nerve tends to supply the thenar muscles, and the ulnar nerve the remainder. However, the dividing line between the two distributions is variable, and either nerve may invade the territory of the other. The ulnar nerve (C7, C8, T1) arises from the medial cord, but usually has a lateral root also from the lateral cord, which carries the C7 fibers (Fig. 2–8). It descends with the neurovascular bundle, pierces the medial intermuscular septum, and descends behind the medial epicondyle, between the two heads of the flexor carpi ulnaris. Here it supplies the elbow joint. It then descends on the flexor digitorum profundus to the middle of the forearm and then along the lateral side of the ulnar artery. Both enter the hand by passing in front of the flexor retinaculum, a slip of which covers them, between the pisiform and the hook of the hamate. The nerve then divides into its

FIGURE 2–8 Schematic representation of the sequences of muscular and cutaneous branches of the ulnar nerve. The sequences of muscular branches are the more common ones. Based on studies by Sunderland and Hughes.67 (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

21

superficial and deep terminal branches. The course of the nerve in the forearm may be indicated on the surface by a line from the front of the medial epicondyle to the lateral margin of the pisiform. There are no muscular branches in the upper arm (i.e., above the elbow). In the upper forearm, muscular branches are given to the flexor carpi ulnaris and the medial part of the flexor digitorum profundus. In the middle of the forearm, the large cutaneous dorsal branch arises and then descends to the back of the hand. In the lower part of the forearm, a variable palmar branch is given to the skin on the medial side of the palm. In the hand, the dorsal branch gives twigs to the skin of the back of the hand and then divides into three dorsal digital nerves that supply the dorsal aspects of the medial side of the little finger and the adjacent sides of the ring and middle fingers (the most lateral of these digital nerves communicates with the adjoining digital branch of the superficial radial nerve). The dorsal digital nerves (including those from the radial nerve) do not reach the tips of the fingers. They extend as far as the nail in the first and fifth fingers, but only to the middle or proximal phalanges in the second, third, and fourth fingers.5 The innervation is completed distally in all fingers by palmar digital nerves. It should be emphasized that there is considerable variation in the number of digits supplied by the median, ulnar, and radial nerves32,34 and that an increase in the number of digits supplied by one nerve is accompanied by a decrease in the number of digits supplied by an adjacent nerve.62 Also, the area of skin supplied on the dorsum of the hand may be invaded by one of the antebrachial cutaneous nerves.62 In the palm of the hand, the superficial branch of the ulnar nerve supplies the palmaris brevis and then divides in such a way as to provide palmar digital nerves for the medial side of the little finger and the adjacent sides of the little and ring fingers. The deep branch, after supplying the hypothenar muscles, extends laterally with the deep palmar arch. In its course it supplies all the interossei, the third and fourth lumbricals, and the adductor pollicis and ends in the flexor pollicis brevis (deep head). The musculocutaneous nerve (C5 to C7) arises from the lateral cord, pierces the coracobrachialis, and descends between the biceps and brachialis, supplying all three muscles and the elbow joint (see Fig. 2–7). On reaching the lateral side of the arm, it continues as the lateral antebrachial cutaneous nerve. This nerve divides into anterior and posterior branches, which supply the skin on the lateral part of the forearm as far as the thumb. Either branch may reach the dorsum of the hand. The radial nerve (C5, C6 to C8, T1), a continuation of the posterior cord, is the largest branch of the brachial plexus (Fig. 2–9). It begins at the medial margin of the biceps, opposite the posterior axillary fold, and descends

22

Structure of the Peripheral Nervous System

FIGURE 2–9 Schematic representation of the sequence of muscular branches of the axillary and radial nerves. The sequences of muscular branches are the more common ones. Based on studies by Sunderland.63 (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

behind the axillary and brachial arteries. It then runs obliquely across the back of the arm, winding around the humerus under cover of the lateral head of the triceps. Lying first on the medial head, and then in the groove on the humerus, it pierces the lateral intermuscular septum at the superior point of trisection of a line between the deltoid insertion and the lateral epicondyle, and comes forward into the cubital fossa. Here it is deeply placed, between the brachialis and brachioradialis. At about the level of the lateral epicondyle, it divides into superficial and deep branches. In thin persons, the radial nerve may be palpated as it winds around the humerus, especially about 1 to 2 cm below the deltoid insertion, and also in the interval between the brachialis and brachioradialis. In the arm, the posterior brachial cutaneous nerve arises from the radial nerve in the axilla. It supplies the skin on the back of the arm. Several muscular branches are given to the three heads of the triceps; that to the medial head accompanies the ulnar nerve and is known as the ulnar

collateral nerve. A branch is also given to the anconeus, and this branch also supplies the elbow joint. The lower lateral brachial cutaneous nerve arises in the arm (sometimes in common with the next branch) and supplies the lateral surface of the lower part of the arm. The posterior antebrachial cutaneous nerve arises in the groove, pierces the lateral head of the triceps, and supplies the skin of the back of the forearm as far as the wrist. Muscular branches of the radial nerve supply the brachioradialis, extensor carpi radialis longus and brevis, and brachialis (probably sensory), and small twigs are given to the elbow joint. The superficial branch is the continuation of the radial nerve. It descends under cover of the brachioradialis, lateral to the radial artery. It then turns dorsally and becomes subcutaneous. It supplies the lateral side of the dorsum of the hand and then divides into a number of dorsal digital nerves for the thumb and index and middle fingers that cross the anatomic snuffbox. The superficial branch may be palpable before entering the snuffbox, and its terminal digital branches can be felt as they cross the tendon of the extensor pollicis longus. A distribution to all five digits has been recorded.45 The deep branch of the radial nerve winds laterally around the radius between the superficial and deep layers of the supinator, where it is often in contact with a bare area of the radius.8 On the back of the forearm, it lies between the superficial and deep extensors. After supplying the supinator and superficial group (the extensor digitorum, extensor digitiminimi, extensor carpi ulnaris, and often extensor carpiradialis brevis), it continues as the posterior interosseous nerve (the entire deep branch is sometimes given this name). This nerve descends on the interosseous membrane and ends on the back of the carpus, where it supplies the wrist and intercarpal joints. During its course, it supplies the abductor pollicis longus, the extensor pollicis brevis and longus, and the extensor indicis. The axillary nerve (C5, C6) is a branch of the posterior cord (see Fig. 2–9). It lies behind the axillary artery, lateral to the radial nerve. At the lower border of the subscapularis, it turns posteriorly, near the joint capsule, and passes through the quadrangular space, at a level indicated by a horizontal plane through the middle of the deltoid. As it does so, it supplies the shoulder joint. It passes medial to the surgical neck of the humerus and divides into anterior and posterior branches. The anterior branch supplies the deltoid. The posterior branch supplies the deltoid and teres minor and then turns around the lower border of the deltoid to supply the skin on the back of the arm as the upper lateral brachial cutaneous nerve. Autonomic Nerve Supply The parasympathetic supply to the head and neck is by way of the cranial nerves mentioned previously and is described with the individual nerves.

23

Gross Anatomy of the Peripheral Nervous System

Sympathetic Trunk and Ganglia. The preganglionic fibers for the head and neck arise chiefly from the first and second thoracic segments of the spinal cord (the range is C8 to T4) (see Fig. 2–6). The fibers reach the thoracic part of the sympathetic trunk, in which they ascend to the cervical part, where they synapse and from which postganglionic fibers are distributed to the blood vessels, smooth muscle, and glands of the head and neck. The preganglionic fibers for the upper limb arise from approximately the 2nd to the 9th or 10th thoracic segments of the spinal cord. They enter the sympathetic trunk and ascend to synapse in the stellate and adjacent ganglia. Postganglionic fibers reach the upper limb by way of fibers that accompany the subclavian and axillary arteries and their branches. The cervical part of the sympathetic trunk consists of three or four ganglia connected by an intervening cord or cords. Postganglionic fibers leave by rami communicantes (one or more to each cervical ventral ramus) by branches that accompany adjacent blood vessels and by branches that go directly to certain cranial nerves and to viscera of the neck and thorax. The superior cervical ganglion lies behind the internal carotid artery, in front of the longus capitis muscle, and extends from the first to the second or third cervical vertebra. The ganglion gives rami communicantes to the upper cervical ventral rami and the last four cranial nerves, twigs to the carotid body and sinus and to the pharyngeal plexus, and cervical cardiac nerves to the heart. Several branches form a plexus along the external carotid artery,17 with some of the fibers reaching the salivary glands. One or more large branches form an internal carotid nerve,39 which ascends with the internal carotid artery to supply the eye, orbit, and intracranial structures by forming first an internal carotid plexus and then subsidiary plexuses along the branches of the artery and by giving twigs to various nerves (tympanic, greater petrosal, and third, fourth, fifth, and sixth cranial) and to the ciliary ganglion. The middle cervical ganglion is quite variable and is often fused with either the superior or the vertebral ganglion. It usually lies just above the arch formed by the inferior thyroid artery (along which twigs form a plexus), at the level of the sixth cervical vertebra. Rami are given to cervical nerves, usually C4 to C6, as well as to the heart. The vertebral ganglion usually lies in front of the vertebral artery, just below the arch of the inferior thyroid artery, and about at the level of the seventh cervical vertebra. As the sympathetic trunk extends downward from the vertebral to the stellate (or inferior cervical) ganglion, it forms two or more cords that pass on each side of the vertebral artery. Rami from the ganglion reach some of the lower cervical nerves and thereby enter the branchial plexus. Twigs are given to the vertebral plexus (discussed later), and a cord termed the ansasubclavia loops in front of and below the first part of the subclavian

artery to the stellate (or inferior cervical) ganglion. Branches of the ansa contribute to a plexus along the subclavian artery. The cervicothoracic (stellate) ganglion has two parts: the inferior cervical and the first thoracic (occasionally a second and third also). These ganglia may be completely fused.29 The ganglionic mass lies usually at the level of the seventh cervical and first thoracic vertebrae, in front of the eight cervical and first thoracic nerves and the seventh cervical transverse process and neck of the first rib, and behind the vertebral artery. The stellate ganglion receives preganglionic fibers from the first, or first and second, thoracic nerves. Its postganglionic fibers supply chiefly the upper limb by way of rami to the lower cervical and upper thoracic nerves and branches to the subclavian and vertebral arteries. Branches of the vertebral and stellate ganglia form a vertebral plexus, which accompanies the vertebral artery into the posterior cranial fossa. During its course, postganglionic fibers are given to the lower cervical nerves.65 Some fibers ascend separately, behind the artery, as a distinct vertebral nerve, which extends to the level of the axis or atlas. Postganglionic fibers in it are given to the cervical nerves and spinal meninges.

Thorax Thoracic Nerves Each of the 12 thoracic nerves gives off a meningeal branch and then, after emerging from an intervertebral foramen, divides into a dorsal and a ventral ramus. The meningeal branches and dorsal rami are described with the back (see later). Each ventral ramus is connected to the sympathetic trunk by a variable number of rami communicantes, and each runs a separate course forward, supplying the skin, muscles, and serous membranes of the thoracic and abdominal walls. The ventral rami of the first 11 nerves are called intercostal nerves; that of the 12th is the subcostal nerve. Typical Intercostal Nerves. The fourth, fifth, and sixth intercostal nerves are typical intercostal nerves and supply only the thoracic wall (Fig. 2–10). Each passes below the neck of the numerically corresponding rib and enters the costal groove below the posterior intercostal vessels. At the anterior end of the intercostal spaces, the nerves turn forward through the overlying muscles and, as anterior cutaneous branches, are distributed to the skin of the front of the thorax. Here they give off medial mammary branches. At the angle of the rib, each nerve supplies the external intercostal muscle and gives off a collateral and a lateral cutaneous branch. The collateral branch passes forward in the intercostal space and ends anteriorly as a lower anterior cutaneous nerve. The lateral cutaneous branch pierces the overlying muscles and divides into

24

Structure of the Peripheral Nervous System

FIGURE 2–10 Diagrammatic representation of the nerves of the thoracic wall. The thickness of the intercostal muscles is exaggerated. (From Gardner, F., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

anterior and posterior branches that supply the skin of the thorax. Some of the anterior branches form lateral mammary branches. The intercostal nerves supply the intercostal, subcostal, serratus posterior superior, and transversus thoracis muscles. Special Intercostal Nerves. The first, second, and third intercostal nerves are special in that they supply the arm as well as the thorax. The first thoracic nerve is the largest of the thoracic spinal nerves. It divides into a larger upper and a smaller lower part. The upper joins the brachial plexus, and the lower becomes the first intercostal nerve. Its distribution is like that of a typical intercostal nerve,3 except that its lateral cutaneous branch supplies the skin of the axilla and may communicate with the intercostobrachial nerve. The second intercostal nerve, which often contributes to the brachial plexus, also has a distribution like that of a typical intercostal nerve, except that its lateral cutaneous branch passes into the arm as the intercostobrachial nerve. This nerve pierces the overlying muscles and supplies the skin on the back and medial side of the arm as far as the elbow. It usually anastomoses with the posterior and medial brachial cutaneous nerves and supplies the axillary arch when that muscle is present. The third intercostal nerve similarly has a distribution like that of a typical intercostal nerve, but its lateral cutaneous branch often gives a twig to the medial side of the arm. The 7th to 11th intercostal nerves are also special in that they supply the abdominal wall as well as the thorax (Fig. 2–11). Known as thoracoabdominal nerves, they course

FIGURE 2–11 The cutaneous distribution of the thoracoabdominal nerves. (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

forward and downward to the anterior ends of the intercostal spaces. They continue between the transversus and internal oblique muscles and then between the rectus abdominis and the posterior wall of its sheath. Here, each nerve divides into two branches. The larger one forms a plexiform arrangement that gives twigs to the rectus and from which an anterior cutaneous branch pierces the rectus to supply the overlying skin. The smaller branch also supplies the rectus and may pierce the rectus and become cutaneous. During its course, each thoracoabdominal nerve gives off a lateral cutaneous branch that pierces the external oblique muscle and divides into anterior and posterior branches. These supply the back, side, and front, of the abdominal wall. The thoracoabdominal nerves also supply the intercostal, subcostal, serratus posterior inferior, transversus abdominis, external and internal oblique, and rectus abdominis muscles, and give sensory twigs to adjacent diaphragm, pleura, and peritoneum. The ventral ramus of the 12th thoracic nerve is special in that it is subcostal rather than intercostal in position and is known as the subcostal nerve. It enters the abdomen, courses downward and laterally behind the kidney, pierces the transversus abdominis, and passes between this muscle

Gross Anatomy of the Peripheral Nervous System

and the internal oblique. It then enters the sheath of the rectus, turns forward, and becomes cutaneous between the umbilicus and the public symphysis. Its lateral cutaneous branch supplies the skin of the gluteal region and upper thigh as far down as the greater trochanter. Muscular branches are given to the transversus, oblique, and rectus muscles. Sensory twigs are given to the adjacent peritoneum. Autonomic Nerve Supply The parasympathetic supply of the thoracic viscera is furnished by the vagus nerves (see Chapter 52). Sympathetic Trunk and Ganglia. The sympathetic trunks enter the thorax from the neck, descend in front of the heads of the ribs and the posterior intercostal vessels and accompanying nerves, and enter the abdomen by piercing the crura of the diaphragm or by passing behind the medial arcuate ligaments (see Fig. 2–6). In the thorax, each trunk usually has 11 or 12 (occasionally 10 or 13) separate ganglia of varying size. As described earlier, the first thoracic ganglion is often fused with the inferior cervical to form the cervicothoracic or stellate ganglion. The trunk itself may be very slender between two adjacent ganglia; sometimes it is double. Each ganglion has one to four rami communicantes whose distinguishing characteristics were described under General Features. Preganglionic fibers for the thoracic and abdominal walls and back arise from all levels of the thoracic spinal cord and reach the sympathetic trunk by way of rami communicantes. The postganglionic fibers return to the spinal nerves by way of rami communicantes and are distributed by way of dorsal and ventral rami and the meningeal branches of the spinal nerves. The postganglionic fibers for thoracic and abdominal viscera are distributed by visceral branches. These include cardiac branches of the ansa subclavia, stellate ganglion, and upper four or five thoracic ganglia and pulmonary branches of the upper four or five thoracic ganglia. Scattered filaments of the trunk are also given to the aortic and esophageal plexuses. The major visceral branches, however, are the three splanchnic nerves. The greater splanchnic nerve is formed by three or four large roots and an inconstant number of smaller ones, all of which arise in the variable fashion from the trunk and ganglia; the upper and lower limits are usually the 5th and 10th. The nerve pierces the diaphragm or passes through its aortic opening and ends in the celiac ganglion and plexuses. It and the other splanchnic nerves contain sensory as well as preganglionic fibers, and relatively large splanchnic ganglion is usually present along the nerve near the diaphragm. The lesser splanchnic nerve, which may be absent, usually arises from the lower thoracic ganglia by one to the three rootlets. Descending slightly lateral to the greater

25

splanchnic nerve, it pierces the diaphragm and joins the aorticorenal ganglion and celiac plexus. The lowest splanchnic nerve, which also may be absent, usually arises from the lowest thoracic ganglion. Descending medial to the sympathetic trunk, it enters the abdomen and joins the aorticorenal ganglion and adjacent plexus. More detailed information about the sympathetic trunks, ganglia, and branches is available in various other sources.30,46,55 (See also references cited under General Features.) Prevertebral Plexuses. These are formed by the branches of the vagus nerves and sympathetic trunks that supply the thoracic viscera and blood vessels. The cardiac plexus may be considered as comprising inter-connected aortic arch, right and left coronary, and right and left atrial plexuses, all lying in adventitia or under epicardium.47 The sympathetic components reach the plexuses by way of cervical, cervicothoracic, and thoracic cardiac nerves (the preganglionic fibers arise from the upper four to five or six segments of the thoracic spinal cord). A variable number of cardiac ganglia occur along the cervicothoracic nerves. The pulmonary plexuses, which are usually described as anterior and posterior (with respect to the root of the lung), are essentially continuous with the cardiac plexus. Lying mostly posterior to the root of the lung and formed chiefly by the vagus nerves, the plexuses receive branches directly from the upper four to five or six segments of the thoracic spinal cord. The esophageal plexus is formed in a variable fashion by the vagus nerves after they leave the pulmonary plexuses. Lying in the fibrous wall of the esophagus, the plexus receives filaments from the sympathetic trunks and greater splanchnic nerves; the preganglionic fibers arise mainly from the lower segments of the thoracic spinal cord. The thoracic aortic plexus receives filaments from the vagus nerves and sympathetic trunks (the preganglionic fibers arise from the upper thoracic segments) that ramify in its adventitia and form a delicate plexus that continues for a short distance along the branches of the aorta. The plexus is also continuous through the aortic opening of the diaphragm with the abdominal aortic and celiac plexuses.

Abdomen, Pelvis, and Lower Limb Lumbar Plexus The dorsal rami of the lumbar nerves, which provide part of the nerve supply of the back, are described later. The ventral rami enter the psoas major muscle, where they combine in a variable fashion to form the lumber plexus (a division into anterior and posterior divisions that then combine, as occurs in the trunks of the brachial plexus, has been described but is difficult to demonstrate).

26

Structure of the Peripheral Nervous System

Within the muscle they are connected to the lumbar sympathetic trunk by rami communicates. Strictly speaking, the second to fourth nerves are usually (in about three fourths of instances) described as forming the lumbar plexus proper. However, the lower part of the fourth lumbar nerve and all of the fifth enter the sacral plexus (the combined trunk is known as the lumbosacral trunk or furcal nerve), and the two plexuses are commonly known as the lumbosacral plexus; the fourth lumbar nerve is then the one ventral ramus common to both plexuses. Moreover, the branches of the first lumbar nerve also are usually described with the lumbar plexus. The general features of the lumbar plexus are discussed further elsewhere (see Chapter 56).1,27,71 Variations. As in the case of the brachial plexus, prefixation and postfixation of the lumbrosacral plexus in the sense of complete shifts upward or downward are uncommon. Nevertheless, the plexus is often spoken of as prefixed when the upper level is the 11th or 12th thoracic nerve and postfixed when the lower border is the fifth sacral or first coccygeal nerve. The total range may therefore be from the 11th thoracic to the first coccygeal. The rami supplying the limbs, exclusive of the cutaneous branches of T12 and L1, can range from the first lumbar to the third sacral. Moreover, minor variations in pattern are common, and the plexuses are often asymmetrical with respect to right and left. Branches. Direct branches (L1 to L4) are given to the quadratus lumborum, psoas major, and psoas minor muscles. The first lumbar nerve has variable connections with the subcostal nerve and with the second lumbar and gives twigs to adjacent muscles. It resembles an intercostal nerve in giving off a collateral branch, the ilioninguinal nerve, and then continuing as the iliohypogastric nerve, which has lateral and anterior cutaneous branches.7 The terminal branches of the lumbar plexus are the lateral femoral cutaneous and the femoral nerves, which emerge laterally from the psoas major, the genitofemoral nerve, which emerges from its front, and the obturator and sometimes the accessory obturator nerves, which emerge medially. The iliohypogastric nerve emerges from the lateral side of the psoas major, runs behind the kidney, pierces the transversus muscle above the iliac crest, and divides into lateral and anterior cutaneous branches. The lateral branch supplies the skin over the side of the buttock, and the anterior branch the skin above the pubis (see Fig. 2–11). Muscular branches, if any, are probably sensory. A motor branch is occasionally given to the pyramidalis. The ilioinguinal nerve runs a similar course to the iliac crest. Here it pierces the transversus and internal oblique, accompanies the spermatic cord through the

inguinal canal, emerges from the superficial ring, and gives cutaneous branches to the thigh and anterior scrotal or anterior labial branches. The lateral femoral cutaneous nerve (L2, or L2 and L3, or L1 and L2) runs obliquely across the iliacus toward the anterior superior iliac spine and enters the thigh by passing behind the inguinal ligament. It divides into anterior and posterior branches that supply the skin of the anterior and lateral aspects of the thigh. The femoral nerve (chiefly L4, plus L2 and L3) is the largest branch of the lumbar plexus (Fig. 2–12). It descends between the psoas and iliacus and enters the thigh behind the middle of the inguinal ligament in the muscular compartment lateral to the femoral vessels. Entering the femoral triangle, it breaks up into a number of terminal branches. In the iliac fossa it supplies the iliacus and the femoral artery. The nerve to the pectineus arises here (or in the femoral triangle) and passes behind the femoral sheath to supply the pectineus and the hip joint. The terminal branches of the femoral nerve are sometimes classified into an anterior division (anterior cutaneous and a branch to the sartorius) and a posterior division (muscular and saphenous). The muscular branch of the anterior division goes directly to the sartorius. The anterior cutaneous branches of the same division are subdivided into the intermediate and medial cutaneous nerves. The intermediate nerve, which is usually double, gives branches to the sartorius and supplies the skin on the front of the thigh; distally it contributes to the patellar plexus. The medial nerves supply the skin on the medial side of the thigh and contribute to the subsartorial and patellar plexuses. The muscular branches of the posterior division supply the rectus femoris (a twig from this is given to the hip joint), the vastus medialis (a branch continues to the knee joint), the vastus intermedius (branches continue to the articularis genus and the knee joint), and the vastus lateralis (a branch continues to the knee joint). The saphenous nerve can be regarded as the termination of the femoral nerve. It descends with the femoral vessels through the femoral triangle and subsartorial canal, crosses the femoral artery from lateral to medial, and then becomes cutaneous. It descends in the leg with the large saphenous vein and supplies the skin on the medial side of the leg and foot. It gives a branch to the knee joint and contributes to the subsartorial and patellar plexuses. The saphenous nerve may be joined by filaments constituting an accessory femoral nerve. This is a common variant that arises from the lumbar plexus, runs a separate course into the thigh, and usually ends by joining one of the cutaneous branches of the femoral nerve. The subsartorial plexus, which lies in the adductor canal deep to the sartorius, is formed by branches from the medial femoral cutaneous, saphenous, and obturator nerves. The patellar plexus, which lies in front of the knee,

Gross Anatomy of the Peripheral Nervous System

27

FIGURE 2–12 The sequences of branches of the sciatic, tibial, and common peroneal nerves, based on studies by Sunderland and Hughes. The sequences of branches of the femoral and obturator nerves are based on studies by Pitres and Testut. (From Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975, with permission.)

is formed by branches of the lateral, intermediate, and medial femoral cutaneous and saphenous nerves. The genitofemoral nerve (L2, or L1 and L2, occasionally L3) descends on the front of the psoas major and divides into genital and femoral branches. The genital branch enters the inguinal canal through the deep ring, supplies the cremaster, and continues to supply the scrotum or labium majus and the adjacent part of the thigh. The femoral branch enters the femoral sheath, lateral to the artery, and turns forward to supply the skin of the femoral triangle.

The obturator nerve (L3, L4, sometimes L2 or L5 also) emerges from the medial side of the psoas major at the inlet of the pelvis (see Fig. 2–12). It runs downward and forward on the lateral wall of the pelvis to the obturator groove, where it divides into anterior and posterior branches. These pass through the obturator foramen into the thigh, where they are separated by the adductor brevis. Branches to the hip joint arise from the trunk and from one or both branches. The anterior branch lies in front of the adductor brevis, behind the pectineus and adductor longus. It descends along the latter muscle;

28

Structure of the Peripheral Nervous System

supplies it and the gracilis, adductor brevis, and sometimes pectineus also; and ends as a filament to the femoral artery and subsartorial plexus; twigs from it supply the overlying skin and occasionally the knee joint. The posterior branch pierces the obturator externus and descends behind the adductor brevis, in front of the adductor magnus. It then pierces the adductor magnus and accompanies the popliteal artery to the back of the knee joint. It supplies the obturator externus, adductor magnus, and sometimes the adductor brevis. The accessory obturator nerve (L3, L4), which is persent in nearly 10% of cases, descends medial to the psoas and enters the thigh deep to the pectineus.75 It supplies the hip joint and pectineus. Sacral Plexus The dorsal rami of the sacral nerves are described with the back. The ventral rami of the first four sacral nerves emerge from the sacral canal though the pelvic sacral foramina. The fourth then divides into upper and lower divisions; the upper division and the first three ventral rami combine with the lumbosacral trunk to form the sacral plexus. Anterior and posterior divisions of the first three rami have been described, but, as in the case of the lumbar plexus, they are difficult to demonstrate. Each ramus contributing to the sacral plexus is connected to a single ganglion of the sacral sympathetic trunk by one or more rami communicates. The sacral plexus lies in front of the piriformis muscle, separated from the internal iliac vessels and ureter anteriorly by the parietal pelvic fascia. From the complex and variable union of the ventral rami, 12 named branches arise. Five supply pelvic strucures; the remainder are distributed to the buttock and lower limb. Branches to Pelvic Structures. The chief branch is the pudendal nerve (S2, S3, S4), which supplies most of the perineum. It passes through the greater sciatic notch below the piriformis and, after crossing the back of the ischial spine, enters the perineum through the lesser sciatic notch, accompanied by the internal pudendal artery. Running forward in the pudendal canal in the ischiorectal fossa, it gives off the inferior rectal nerve and then divides into the perineal nerve and the dorsal nerve and then divides into the perineal nerve and the dorsal nerve of the penis or clitoris. The inferior rectal nerve (which may arise separately from S3 and S4 of the sacral plexus) pierces the medial wall of the pudendal canal and divides into branches that traverse the ischiorectal fossa. They supply the skin around the anus, the sphincter ani externus, and the anal mucosa as far upward as the pectinate line. The perineal nerve divides into superficial and deep branches. The superficial branch gives off two posterior scrotal or labial nerves (medial and lateral), which pierce the superficial and deep perineal fasciae and

run forward to supply the scrotum or labium majus. The deep branch of the perineal nerve gives twigs to the sphincter ani externus and levator ani and then enters the superficial perineal space to supply the bulbospongiosus, ischiocavernosus, superficial transversus perinei, and bulb of the penis. The dorsal nerve of the penis or clitoris, the other terminal branch of the pudendal nerve, pierces the posterior edge of the urogenital diaphragm. As it runs forward, it supplies the deep transversus perinei and sphincter urethrae muscles. It then pierces the inferior fascia of the diaphragm, gives a branch to the corpus cavernosum penis or clitoris, and then runs forward on the dorsum of the penis or clitoris to supply the skin, prepuce, and glans. In addition to the pudendal nerve, the sacral plexus gives the following additional branches to pelvic structures: the nerve to the piriformis (S1, S2), the nerve to the levator ani and coccugeus (S3, S4), the nerve to the sphincter ani externus (perineal branch of S4), and the pelvid splanchnic nerves (S2, S3 and S4, S5), which contain parasympathetic preganglionic as well as sensory fibers and which enter the inferior hypogastric plexus. Branches to the Buttock and Lower Limb. These are seven in number. The superior gluteal nerve (L4, L5, S1) passes backward through the greater sciatic notch, above the piriformis. An upper branch supplies the gluteus medius, and a lower branch supplies the gluteus minimus, tensor fasciae latae, and hip joint. The inferior gluteal nerve (L5, S1, S2) passes through the greater sciatic foramen below the piriformis and supplies the gluteus maximus. The nerve to the obturator internus (L5, S1, S2) leaves the pelvis below the piriformis, supplies the superior gemellus, and then passes through the lesser sciatic foramen to the obturator internus. The nerve to the quadratus femoris (L4, L5, S1) leaves the pelvis below the piriformis in front of the sciatic nerve. It supplies the inferior gemellus, the quadratus femoris, and the hip joint. The posterior femoral cutaneous nerve (S1 to S3) leaves the pelvis below the piriformis. It descends in company with the sciatic nerve, becomes superficial near the popliteal fossa, and accompanies the small saphenous vein to the middle of the calf. Its branches are inferior clunial nerves (gluteal branches) to the skin of the buttock, perineal branches to the skin of the genitalia, and femoral and sural branches to the skin on the back of the thigh and calf. The perforating cutaneous (inferior medial clunial) nerve (S2, S3) pierces the sacrotuberous ligament and supplies the skin over the lower part of the buttock. The sciatic nerve (L4, L5, S1 to S3), the largest nerve in the body, consists of peroneal and tibial parts, which are usually bound together, leaving the pelvis through the greater sciatic foramen, below the piriformis (see Fig. 2–12).2 Sometimes they leave separately, the peroneal portion piercing the piriformis and the tibial portion passing

Gross Anatomy of the Peripheral Nervous System

below it. The exit of the sciatic nerve from the pelvis is indicated by the superior point of trisection of a line from the posterior superior iliac spine to the ischial tuberosity. Its downward course is indicated by a line down the middle of the back of the thigh (from the midpoint of a line between the greater trochanter and ischial tuberosity). The sciatic nerve descends under cover of the gluteus maximus, between the greater trochanter and ischial tuberosity. In the thigh it lies anteriorly on the adductor magnus and is accompanied by the posterior femoral cutaneous nerve and the companion artery. In the lower third of the thigh the nerve separates into its two components, the tibial and common peroneal nerves (the separation may occur at any level in the gluteal region or thigh). Its branches arise mostly on the medial side and supply the semitendinosus, semimembranosus, long head of the biceps, adductor magnus (all by the tibial nerve), and short head of the biceps (by the common peroneal nerve). The tibial (medial popliteal) nerve (L4 to S3) descends separately through the popliteal fossa (see Fig. 2–12). It then lies on the popliteus muscle, under cover of the gastrocnemius, and at the lower border of the popliteus passes deep to the fibrous arch of the soleus to reach the back of the leg. Here it descends first on the tibialis posterior and flexor digitorum longus and then on the tibia. Then, becoming more superficial and crossing the posterior tibial artery posteriorly to gain its lateral side, it ends by dividing into medial and lateral plantar nerves under cover of the flexor retinaculum. In the thigh, muscular branches arise as listed with the sciatic nerve. In the popliteal fossa, branches are given to the knee joint and muscular branches to the gastrocnemius, soleus, plantaris, popliteus, and tibialis posterior. A branch of the nerve to the popliteus, the interosseous nerve of the leg, descends on the interosseous membrane. The medial sural cutaneous nerve joins the peroneal communicating branch of the common peroneal nerve to form the sural nerve. The sural nerve descends in company with the small saphenous vein, gives lateral calcaneal branches to the skin of the back of the leg and lateral aspect of the foot and heel, gives twigs to the ankle joints, and continues forward along the lateral side of the little toe as the lateral dorsal cutaneous nerve. In the leg the tibial nerve gives muscular branches to the soleus, tibialis posterior, flexor hallucis longus, and flexor digitorum longus. Medial calcaneal branches supply the skin of the heel and sole, and a twig is given to the ankle joint. The course of the tibial nerve in the leg is indicated on the surface by a line from about the level of the tibial tuberosity downward to the midpoint between the medial malleolus and the heel. The medial plantar nerve, the larger of the two terminal branches of the tibial nerve, at first lies deep to the abductor hallucis. It then runs forward in the sole between the abductor and the flexor digitorum brevis. It supplies these muscles and the skin on the medial side of the sole.

29

Its terminal branches are four plantar digital nerves for muscles (flexor hallucis and first lumbrical) and for the medial side of the big toe and the adjacent sides of the first and second, second and third, and third and fourth toes. The nerves extend on to the dorsum and supply the nail beds and tips of the toes. The lateral plantar nerve runs forward and laterally between the quadratus plantae and the flexor digitorum brevis and divides into superficial and deep branches. During its course it supplies the quadratus plantae and abductor digiti minimi and the skin of the lateral side of the sole. The superficial branch supplies the flexor digiti minimi brevis and the lateral side of the sole and little toe and, by plantar digital nerves, the adjacent sides of the fourth and fifth toes. The deep branch turns medially; supplies the interossei, the second, third, and fourth lumbricals, and the adductor hallucis; and gives off articular twigs. The common peroneal (lateral popliteal) nerve (L4 to S2) descends through the popliteal fossa, following the medial edge of the biceps closely (see Fig. 2–12). It crosses the lateral head of the biceps, gains the back of the head of the fibula, and winds around the neck of that bone (where it is often palpable, and where it is susceptible to injury) under cover of the peroneus longus. Here it divides into its terminal branches, the superficial and deep peroneal nerves. While a part of the sciatic nerve, it supplies the short head of the biceps and sometimes the knee joint also. In the popliteal fossa, it supplies the knee joint and gives rise to a branch that divides into the lateral sural cutaneous nerve (for the skin on the lateral side of the leg) and the peroneal communicating branch (which joins the medial sural cutaneous nerve to form the sural nerve). At the neck of the fibula, it gives off a small recurrent branch that supplies the knee and tibiofibular joints and the tibialis anterior. The common peroneal nerve sometimes supplies the peroneus longus or extensor digitorum longus or both. The superficial peroneal (musculocutaneous) nerve, one of the two terminal branches of the common peroneal nerve, descends in front of the fibula, between the peronei and the extensor digitorum longus. Muscular branches supply the peroneus longus and peroneus brevis; the branch to the latter is often prolonged to the extensor digitorum brevis and adjacent joints and is termed the accessory deep peroneal nerve.42,73 In the lower part of the leg, the superficial peroneal nerve divides into medial and intermediate dorsal cutaneous nerves. These pass in front of the extensor retinacula, the medial branch supplying the skin and joints of the medial side of the big toe and (by dorsal digital nerves) the adjacent sides of the second and third toes; the intermediate nerve (by dorsal digital nerves) supplies the adjacent sides of the third and fourth and the fourth and fifth toes. As in the hand, the territories of distribution of the cutaneous nerves of the foot show considerable variation in size and overlap and reciprocal changes in size.33

30

Structure of the Peripheral Nervous System

The deep peroneal nerve continues the winding course of the common peroneal nerve around the neck of the fibula, then pierces the anterior intermuscular septum and extensor digitorum longus, and descends on the interosseous membrane. It meets the anterior tibial artery, and both pass deep to the extensor retinacula. Branches are given to the tibialis anterior, extensor hallucis longus, extensor digitorum longus, peroneus tertius, and ankle joint. In the foot, where it lies about midway between the malleoli, the nerve divides into its terminal branches, medial and lateral. The medial branch gives dorsal digital nerves for the adjacent sides of the first and second toes, and the lateral branch supplies the extensor digitorum brevis and adjacent joints. It may also send twigs (probably afferent) to the first three dorsal interossei. Coccygeal Plexus The ventral ramus of the fifth sacral nerve enters the pelvis between the sacrum and coccyx; that of the coccygeal nerve passes forward below the rudimentary transverse process of the first piece of the coccyx. The coccygeal (or sacrococcygeal) plexus is formed by these two ventral rami, together with the lower division of the ventral ramus of the fourth sacral nerve.59 The plexus supplies the coccyx, the sacrococcygeal joint, and the skin over the coccyx. Autonomic Nerve Supply The parasympathetic supply of the abdominal viscera is furnished chiefly by the vagus nerves. The descending and sigmoid portions of the colon and the pelvic viscera are supplied by the sacral parasympathetics. Details of the relationship between the vagal nerve fibers, the sympathetic nerve fibers, and the nerve cell bodies and fibers of the enteric nerve plexuses are given in Chapter 12. Sympathetic Trunk and Ganglia. The two trunks enter the abdomen by piercing the diaphragm or by passing behind the medial arcuate ligaments (see Fig. 2–6). In descending on the vertebral column, adjacent to the psoas major muscles, the right trunk lies behind the inferior vena cava, the left one beside the aorta. The trunks continue into the pelvis, where they lie on the pelvic surface of the sacrum, medial to the upper three pelvic sacral foramina and usually in front of the fourth. They end by uniting in front of the coccyx to form an enlargement, the ganglion impar. In the lumbar region, the two trunks are seldom symmetrical and the ganglia are irregular in size, number, and position. There are usually four or five ganglia (three or four, according to Webber70), but there may be from two to six, and occasionally a trunk is an elongated ganglionic mass. The variations are such as to make identification of the proper level of a specific ganglion very difficult. Nor can counting from the highest lumbar ganglion be depended upon. When the first lumbar ganglion is

present, it lies between the crus of the diaphragm and the vertebral column, is difficult to reach, and is often overlooked. Rami communicantes are a better means of identification. In the pelvis, the number of ganglia is variable, but there are usually three or four. Each lumbar ganglion has two or more rami communicantes to two or more spinal nerves. The lowest ramus at upper levels contains the most preganglionic fibers and is the clue to the identification of a ganglion. For example, the second lumbar ganglion is connected to the first and second lumbar nerves; hence the lowest ramus (from the second nerve) leads to the second ganglion. Nevertheless, the identification of ganglia during surgical procedures remains an uncertain and difficult task. The second lumbar nerve is usually the lowest one to contain preganglionic sympathetic fibers. Each sacral ganglion tends to be connected by rami communicantes with only one spinal nerve. The postganglionic fibers in the lumbar and sacral rami communicantes are for the supply of the lower part of the abdominal wall and back; the anal canal, perineum, and external genitalia; and the lower limb by way of the lumbosacral and coccygeal plexuses. The preganglionic fibers for the lower limbs arise from the lower thoracic and upper lumbar levels of the spinal cord and descend in the trunks to the lumbar and sacral ganglia, where they synapse. The preganglionic fibers for abdominal viscera descend mostly in the thoracic splanchnic nerves. Some, however, descend in the sympathetic trunks and leave by way of visceral branches. These consist of four or more lumbar splanchnic nerves of variable size, which, depending upon level, join the celiac or intermesenteric and adjacent plexuses or the superior hypogastric plexus.39 These nerves also contain preganglionic and sensory fibers, and some of their postganglionic fibers reach pelvic viscera by way of the hypogastric plexuses, and the iliac fossa and upper thigh by way of the aortic plexus. The preganglionic fibers for pelvic viscera arise in the upper lumbar levels of the spinal cord. They descend and synapse in lumbar and sacral ganglia. The postganglionic fibers reach pelvic viscera by way of rami communicantes or lumbar splanchnic nerves (just described) and by way of sacral splanchnic nerves. These last are a variable number of fine visceral branches of the sacral sympathetic trunks that join the inferior hypogastric plexus. More detailed information about the sympathetic trunks and ganglia is available in the references cited under General Features and elsewhere.41,55 Prevertebral Plexuses. In the abdomen, these are best considered as a single great plexus formed by the splanchnic nerves, branches of both vagus nerves, and masses of ganglion cells, with various parts named

31

Gross Anatomy of the Peripheral Nervous System

according to the arteries with which they are associated. The plexus lies in front of the upper part of the abdominal aorta. Its chief ganglia are the irregularly shaped celiac ganglia at the origin of the celiac trunk, each lying on the corresponding crus of the diaphragm. The inferolateral extensions of the ganglia are termed the aorticorenal ganglia. The superior mesenteric ganglion (or ganglia) is usually fused with the celiac ganglia. Smaller ganglia, for example, phrenic and renal, are often found along the smaller subsidiary plexuses. The celiac and superior mesenteric plexuses lie on the front and sides of the aorta at the origins of the celiac trunk and the superior mesenteric and renal arteries. They contain the celiac, aorticorenal, and superior mesenteric ganglia and many smaller unnamed masses. Branches of the plexuses extend along arteries and are named accordingly—for example, hepatic, gastric, phrenic, suprarenal, renal, ureteric, and testicular or ovarian. Downward continuations of the prevertebral plexus form the aortic and inferior mesenteric plexuses. The aortic plexus consists of a variable number of interconnected strands that, as they descend along the aorta, receive branches from the lumbar splanchnic nerves. The part of the plexus between the origins of the superior and inferior mesenteric arteries is also known as the intermesenteric plexus, and below the bifurcation of the aorta the aortic plexus becomes the superior hypogastric plexus. Some filaments from the aortic plexus accompany the lumbar arteries and provide a pathway for postganglionic fibers to the abdominal wall and back. Other filaments, together with branches of the lumbar splanchnic nerves, form a plexus along the common and external iliac arteries. This plexus is reinforced by a branch of the femoral nerve and continues into the thigh along the femoral artery. The inferior mesenteric plexus is a continuation of the aortic plexus along the inferior mesenteric plexus. It contains one or more inferior mesenteric ganglia and it forms the superior rectal plexus, which supplies the rectum with sympathetic and sensory fibers. In the pelvis, in front of the fifth lumbar vertebra, the aortic plexus becomes the superior hypogastric plexus (or presacral nerve). This then divides in front of the sacrum into two elongated narrow networks termed the right and left hypogastric nerves. Each of these nerves descends on the side of the rectum (or rectum and vagina) and is joined by the pelvic splanchnic nerves to form the right and left inferior hypogastric (or pelvic) plexuses. Each plexus contains small, scattered pelvic ganglia, and each is joined by the sacral splanchnic nerves from the sympathetic trunk. Many branches of the inferior hypogastric plexus supply the rectum, with some of them forming a middle rectal plexus. Large portions of the inferior hypogastric plexuses form either the prostatic plexus (which continues forward as the cavernous nerves of the penis and from

which is derived the vesical plexus) or the uterovaginal plexus (fibers from the lowermost part of this plexus continue as the cavernous nerves of the clitoris). The inferior hypogastric plexuses carry sensory, postganglionic sympathetic, and preganglionic parasympathetic fibers. Some of the preganglionic parasympathetic fibers supply the descending and sigmoid colonic segments, usually by a single ascending branch of the hypogastric nerve.74 Some of the sensory fibers are pain fibers that reach the spinal cord by way of the superior hypogastric plexuses and lumbar splanchnic nerves as well as by way of pelvic splanchnic nerves. The other sensory fibers, concerned with reflexes and visceral sensations, reach the spinal cord by way of the pelvic splanchnic nerves.

Back The nerve supply of the back is provided by the meningeal branches and dorsal rami of spinal nerves. Each spinal nerve gives off a meningeal branch (or sinuvertebral nerve), which is often connected with adjacent sympathetic ganglia and which reenters the vertebral canal and supplies dura mater, posterior longitudinal ligament, periosteum, and epidural and intraosseous blood vessels.52 The meningeal branches of the upper three cervical nerves give off branches that ascend through the foramen magnum and supply the dura mater of the floor of the posterior cranial fossa.37 The dorsal rami arise after the spinal nerves emerge from the intervertebral foramina (sacral nerves divide within the sacrum), and the dorsal rami pass backward to supply the muscles, bones, joints, and skin of the back. Most divide into medial and lateral branches, which descend as they run dorsally, each supplying muscles and each anastomosing with nerves above and below to form a plexus in the muscles. In the upper half of the trunk the medial branches supply skin, whereas in the lower half the lateral branches become cutaneous. However, the level of shift is variable (see Thoracic Dorsal Rami below). Cervical Dorsal Rami The connecting loops between the dorsal rami of the first three or four cervical nerves form what is termed the posterior cervical plexus,4 which gives branches to adjacent muscles. The dorsal ramus of the first cervical nerve, which usually lacks cutaneous fibers, is known as the suboccipital nerve. It emerges above the posterior arch of the atlas, below the vertebral artery, and supplies the semispinalis capitis and the muscles of the suboccipital triangle. The dorsal ramus of the second cervical nerve supplies the obliquus capitis inferior and then divides into medial and lateral branches. The medial branch, termed the greater occipital nerve, pierces the semispinalis capitis and trapezius and then accompanies the occipital artery and supplies the skin of the scalp as far forward as the vertex.

32

Structure of the Peripheral Nervous System

The medial branch of the dorsal ramus of the third cervical nerve continues as the third occipital nerve, piercing the trapezius and supplying the skin on the back of the head. The dorsal rami of C6, C7, and C8 usually have no cutaneous branches4,31,51; hence the C5 dermatome is adjacent to T1 and, with overlap, C4 meets T2. This junction marks what might be called the dorsal axial line (from about the spine of C7 to near the deltoid insertion). Thoracic Dorsal Rami Each ramus passes backward, supplying the deeply placed muscles, and divides into a medial and a lateral cutaneous branch, which are separated by slips of the longissimus thoracis muscle. The medial branches pass backward and downward, supplying the erector spinae and its divisions. The medial branches of the upper thoracic nerves (T1 to T3) also become cutaneous. The lateral branches supply the levatores costarum, the longissimus thoracis, and the iliocostalis thoracis. They have a long downward course,31 the lower ones (T9 to T12) piercing the latissimus dorsi and supplying the skin of the back as far down as the gluteal region. In what might be termed a transitional zone, both medial and lateral branches of T4 to T8 give cutaneous twigs. Lumbar, Sacral, and Coccygeal Dorsal Rami The lateral branches of the upper lumbar rami give rise to the superior clunial nerves, which supply the skin of the buttock. The lateral branches of the lower lumbar rami, together with those of the sacral dorsal rami, form the dorsal sacral plexus. The dorsal rami of the first four sacral nerves pass backward through the dorsal sacral foramina, whereas those of the fifth sacral and the coccygeal nerves emerge through the sacral hiatus. The medial branches of the first dorsal rami supply the erector spinae. The lateral branches, together with those of the lower lumbar and a contribution from the fifth sacral, form the dorsal sacral plexus immediately behind the sacrum and coccyx.27 Loops of this plexus give off two or three middle clunial nerves (though not invariably), which pierce the overlying gluteus maximus and supply the skin of the buttock. The dorsal rami of the fifth sacral and coccygeal nerves lack medial and lateral branches; they communicate (often forming a single nerve) and supply adjacent ligaments and overlying skin.

REFERENCES 1. Bardeen, C. R., and Elting, A. W.: A statistical study of the variations in the formation and position of the lumbo-sacral plexus in man. Anat. Anz. 19:124, 209, 1901. 2. Beaton, L. E., and Anson, B. J.: The relation of the sciatic nerve and of its subdivisions to the piriformis muscle. Anat. Rec. 70:1, 1937.

3. Cave, A. J. E.: The distribution of the first intercostal nerve and its relation to the first rib. J. Anat. 63:367, 1929. 4. Cave, A. J. E.: The innervation and morphology of the cervical intertransverse muscles. J. Anat. 71:497, 1937. 5. Dankmeijer, J., and Waltman, J. M.: Sur I’innervation de la face dorsale des doigts humains. Acta Anat. 10:377, 1950. 6. Dass, R.: Sympathetic components of the dorsal primary divisions of human spinal nerves. Anat. Rec. 113:493, 1952. 7. Davies, F.: A note on the first lumbar nerve (anterior ramus). J. Anat. 70:177, 1935. 8. Davies, F., and Laird, M.: The supinator muscle and the deep radial (posterior interosseous) nerve. Anat. Rec. 101:243, 1948. 9. Delmas, J., and Laux, G.: Système Nerveux Sympathique. Paris, Masson et Cie, 1952. 10. Dow, D. R.: The anatomy of rudimentary first thoracic ribs with special reference to the arrangement of the brachial plexus. J. Anat. 59:166, 1925. 11. Fenart, R.: La morphogénése du plexus brachial, ses rapports avec la formation du cou et du membre supérieur. Acta Anat. 32:322, 1958. 12. Fender, F. A.: Foerster’s scheme of the dermatomes. Arch. Neurol. Psychiatry 41:688, 1939. 13. Foerster, O.: The dermatomes in man. Brain 56:1, 1933. 14. Foerster, O.: Symptomatologie der Erkankungen des Rücken-marks and seiner Wurzeln. In Bumke, I., and Foerster, O. (eds.): Handbuch der Neurologie, Vol. 5. Berlin, Springer, 1936. 15. Furness, J. B., and Costa, M.: Types of nerves in the enteric nervous system. Neuroscience 5:1, 1980. 16. Gabella, G.: Structure of the Autonomic Nervous System. London, Chapman and Hall, 1976. 17. Gardner, E.: Surgical anatomy of the external carotid plexus. Arch. Surg. 46:238, 1943. 18. Gardner, E., Gray, D. J., and O’Rahilly, R.: Anatomy, 4th ed. Philadelphia, W. B. Saunders, 1975. 19. Gershon, M. D.: The enteric nervous system. Annu. Rev. Neurosci. 4:227, 1981. 20. Hansen, K., and Schliack, H.: Segmentale Innervation. Stuttgart, Thieme, 1962. 21. Harness, D., and Sekeles, E.: The double anastomotic innervation of thenar muscles. J. Anat. 109:461, 1971. 22. Harris, W.: The true form of the brachial plexus, and its motor distribution. J. Anat. 38:399, 1904. 23. Harris, W.: The Morphology of the Brachial Plexus. London, Humphrey Milford, 1939. 24. Haymaker, W., and Woodhall, B.: Peripheral Nerve Injuries, 2nd ed. Philadelphia, W. B. Saunders, 1953. 25. Head, H., and Campbell, A. W.: The pathology of herpes zoster and its bearing on sensory localisation. Brain 23:353, 1900. 26. Hollinshead, H.: Anatomy for Surgeons, 2nd ed. New York, Hoeber Medical Division, Harper & Row, 1968, 1969, 1971. 27. Horwitz, M. T.: The anatomy of (A), the lumbosacral nerve plexus—its relation to variations of vertebral segmentation, and (B), the posterior sacral nerve plexus. Anat. Rec. 74:91, 1939. 28. Hovelacque, A.: Anatomie des Nerfs Craniens et Rachidiens et du Système Grande Sympathique Chez L’homme. Paris, Gaston Doin et Cie, 1927.

Gross Anatomy of the Peripheral Nervous System 29. Jamieson, R. W., Smith, D. B., and Anson, B. J.: The cervical sympathetic ganglia. Bull. Northwest. Univ. Med. Sch. 26:219, 1952. 30. Jit, I., and Mukerjee, R. N.: Observations on the anatomy of the human thoracic sympathetic chain and its ganglia; with an anatomical assessment of operations for hypertension. J. Anat. Soc. India 9:55, 1960. 31. Johnston, H. M.: The cutaneous branches of the posterior primary divisions of the spinal nerves, and their distribution in the skin. J. Anat. 43:80, 1908. 32. Jones, F. W.: The Principles of Anatomy as Seen in the Hand, 2nd ed. London, Baillière, Tindall, & Cox, 1942. 33. Jones, F. W.: Structure and Function as Seen in the Foot, 2nd ed. London, Baillière, Tindall, & Cox, 1942. 34. Kaplan, E. B.: Functional and Surgical Anatomy of the Hand, 2nd ed. Philadelphia, J. B. Lippincott, 1965. 35. Keegan, J. J., and Garrett, F. D.: The segmental distribution of the cutaneous nerves in the limbs of man. Anat. Rec. 102:409, 1948. 36. Kerr, A. T.: The brachial plexus of nerves in man, the variations in its formation and branches. Am. J. Anat. 23:285, 1918. 37. Kimmel, D. L.: Innervation of spinal dura mater and dura mater of the posterior cranial fossa. Neurology (Minneap.) 11:800, 1961. 38. Kuntz, A.: The Autonomic Nervous System, 4th ed. Philadelphia, Lea & Febiger, 1953. 39. Kuntz, A.: Components of splanchnic and intermesenteric nerves. J. Comp. Neurol. 105:251, 1956. 40. Kuntz, A., Hoffman, H. H., and Napolitano, L. M.: Cephalic sympathetic nerves. Arch. Surg. 75:108, 1957. 41. Labbok, A.: Anatomische Untersuchungen and Typen des Kreuzabschnittes der Trunci sympathici. Anat. Anz. 85:14, 1937. 42. Lambert, E. H.: The accessory deep peroneal nerve. Neurology (Minneap.) 19:1169, 1969. 43. Langley, J. N.: The Autonomic Nervous System. Part 1. Cambridge, UK, Heffer, 1926. 44. Last, R. J.: Innervation of the limbs. J. Bone Joint Surg. Br. 31:452, 1949. 45. Learmonth, J. R.: A variation in the distribution of the radial branch of the musculo-spinal nerve. J. Anat. 53:371, 1919. 46. Mitchell, G. A. G.: Anatomy of the Autonomic Nervous System. Edinburgh, E. & S. Livingstone, 1953. 47. Mizeres, N. J.: The cardiac plexus in man. Am. J. Anat. 112:141, 1963. 48. Monro, P. A. G.: Sympathectomy. London, Oxford University Press, 1959. 49. Nomina Anatomica, 4th ed. Amsterdam, Excerpta Medica Foundation, 1977. 50. Pallie, W.: The intersegmental anastomoses of posterior spinal rootlets and their significance. J. Neurosurg. 16:188, 1959. 51. Pearson, A. A, Sauter, R. W., and Buckley, T. F.: Further observations on the cutaneous branches of the dorsal primary rami of the spinal nerves. Am. J. Anat. 118:891, 1966. 52. Pedersen, H. E., Blunck, C. F. J., and Gardner, E.: The anatomy of lumbosacral posterior rami and meningeal

53.

54. 55. 56. 57. 58. 59. 60. 61. 62. 63.

64. 65.

66.

67.

68.

69. 70. 71. 72. 73. 74. 75. 76.

33

branches of spinal nerves (sinu-vertebral nerves). J. Bone Joint Surg. Am. 38:377, 1956. Pernkopf, E.: Atlas of Topographical and Applied Human Anatomy (ed., H. Ferner; transl., H. Monsen). Philadelphia, W. B. Saunders, 1963. Pick, J.: The Autonomic Nervous System. Philadelphia, J. B. Lippincott, 1970. Pick, J., and Sheehan, D.: Sympathetic rami in man. J. Anat. 80:12, 1946. Pitres, A., and Testut, L.: Les Nerfs en Schémas. Paris, Gaston Doin et Cie, 1952. Schadé, J. P.: The Peripheral Nervous System. Amsterdam, Elsevier, 1966. Seddon, H.: Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingstone, 1972. Sicard, A., and Bruézière, J.: Le plexus sacro-coccygien. Arch. Anat. Histol. Embryol. 33:43, 1950. Skoog, T.: Ganglia in the communicating rami of the cervical sympathetic trunk. Lancet 2:457, 1947. Smorto, M. P., and Basmajian, J. V.: Clinical Electroneurography. Baltimore, Williams & Wilkins, 1972. Stopford, J. S. B.: The variation in distribution of the cutaneous nerves of the hand and digits. J. Anat. 53:14, 1918. Sunderland, S.: Metrical and non-metrical features of the muscular branches of the radial nerve. J. Comp. Neurol. 85:93, 1946. Sunderland, S.: Nerves and Nerve Injuries, 2nd ed. New York, Churchill Livingstone, 1978. Sunderland, S., and Bedbrook, G. M.: The relative sympathetic contribution to individual roots of the brachial plexus in man. Brain 72:297, 1949. Sunderland, S., and Hughes, E. S. R.: Metrical and nonmetrical features of the muscular branches of the sciatic nerve and its medial and lateral popliteal divisions. J. Comp. Neurol. 85:205, 1946. Sunderland, S., and Hughes, E. S. R.: Metrical and nonmetrical features of the muscular branches of the ulnar nerve. J. Comp. Neurol. 85:113, 1946. Sunderland, S., and Ray, L. J.: Metrical and non-metrical features of the muscular branches of the median nerve. J. Comp. Neurol. 85:191, 1946. von Lanz, T., and Wachsmuth, W.: Praktische Anatomie. (Some parts in 2nd edition.) Berlin, J. Springer, 1953–1972. Webber, R. H.: A contribution on the sympathetic nerves in the lumbar region. Anat. Rec. 130:581, 1958. Webber, R. H.: Some variations in the lumbar plexus of nerves in man. Acta Anat. 44:336, 1961. Williams, P. L, and Warwick, R. (eds.): Gray’s Anatomy, 36th ed. Philadelphia, W. B. Saunders, 1980. Winckler, G.: Le nerf péronier accessoire profond. Arch. Anat. Histol. Embryol. 18:181, 1934. Woodburne, R. T.: The sacral parasympathetic innervation of the colon. Anat. Rec. 124:67, 1956. Woodburne, R. T.: The accessory obturator nerve and the innervation of the pectineus muscle. Anat. Rec. 136:367, 1960. Wrete, M.: Die intermediären vegetativen Ganglien der Lum-balregion beim Menschen. Z. Mikrosk. Anat. Forsch. 53:122, 1943.

3 Microscopic Anatomy of the Peripheral Nervous System C.-H. BERTHOLD, JOHN P. FRAHER, R. H. M. KING, AND MARTIN RYDMARK

Nerve Trunks and Spinal Roots Connective Tissues Endoneurium Perineurium Epineurium

Node of Ranvier The Constricted Axon Segment The Node–Paranodal Apparatus The Schwann Cell The Myelin Sheath

Myelinated Nerve Fibers

Unmyelinated Nerve Fibers

General Organization of Myelinated Axons Periaxonal Space Axolemma Axoplasm The Paranode-Node-Paranode Region Internodal End Region

Historical Background Development General Organization Relationship of Unmyelinated Axons with Schwann Cells Spatial Relationships between Axons, Schwann Cells, and Collagen

Nerve Trunks and Spinal Roots R. H. M. King The nerve fibers comprising the peripheral nerves are collected together in bundles or fascicles. A nerve fiber is defined as an axon plus its associated Schwann cells. Each fascicle is delineated by a perineurial sheath composed of flattened squamous cells connected together by tight junctions. The individual fascicles are embedded in a collagenous matrix or epineurium. The connective tissues of peripheral nerves were described by Ranvier332 and Key and Retzius.216 The latter suggested that the nerve was divided into perifascicular connective tissues, lamellated sheath, and intrafascicular tissue, which they named epineurium, perineurium, and endoneurium, respectively (Fig. 3–1). This remains the most satisfactory and convenient subdivision and is in universal use today. Unlike epineurium and endoneurium, however, the term perineurium does not just

Morphometry Regeneration

The CNS-PNS Transitional Zone General Form Fiber Transition Tissue Composition Historical Background Development Comparative Anatomy Blood Supply Mechanical Features

refer to the collagenous connective tissue but also includes the cellular component. The perineurium is part of the membranous covering that ensheaths the whole of the nervous system (Fig. 3–2). Having said that, many authors use the term endoneurium to refer to the contents of the fascicle within the perineurium, including Schwann cells and axons, and not just to the connective tissue (e.g., Dyck et al.,99 Midroni et al.270). The nerve fascicles may branch so that any particular nerve fiber may change from one fascicle to another and may change association with fibers of other types. Electrophysiologic studies of sensory median nerve fascicles have shown that there is intrafascicular segregation by modality of both myelinated and unmyelinated fibers. These studies suggested that some Schwann cells contain only afferent C fibers while others contain efferent sympathetic fibers.177 This may result from alterations of Schwann cell proteins to reflect the type of axon with which they are associated. This segregation may improve the efficiency of regeneration because axons could make useful contact with similar end organs even if they were not identical to the 35

36

Structure of the Peripheral Nervous System

original one. Earlier studies, however, showed that some axons in a Remak fiber (unit of Schwann cell and unmyelinated axons) contain dense-cored vesicles while others do not. This observation implies that adrenergic axons can share the same Schwann cell as nonadrenergic ones.389 Axons conduct electrical impulses at rates varying between 1 and 100 m/s. Saltatory conduction in myelinated fibers, whereby the nerve impulse jumps from one node of Ranvier (see Node of Ranvier later) to the next, results in the faster transmission speeds, whereas continuous transmission of the nerve impulse, found in unmyelinated axons, occurs at the lowest velocities. In addition to propagation of electrical impulses, axons also have chemical conduction mechanisms. There are fast axonal anterograde and retrograde transport (200 to 400 mm/day) systems that depend on the tubulin and intermediate filament cytoskeleton. In addition, there are much slower transport systems whereby the components of this network are transported at rates of about 1 to 2 mm/day. These are discussed in more detail later (see Axoplasm). FIGURE 3–1 Transverse section of normal human sural nerve branch showing epineurium (ep), perineurium (arrowhead), and endoneurium (en). There are fat globules (arrows) and blood vessels (asterisk) in the epineurium. Lymph vessels are not identifiable at this magnification. Fleming-fixed, Kultschitsky-stained paraffin section. Bar: 100 ␮m. See Color Plate

CONNECTIVE TISSUES Endoneurium Very approximately 50% of the space enclosed by the perineurium is occupied by axons and their Schwann cells. The endoneurial fluid and fibrous collagen takes up another

FIGURE 3–2 Portion of transverse section through part of the brachial plexus of rabbit, showing disposition of epineurium (Ep), perineurium (P), and endoneurium (En). The structure labeled endoneurium is an intrafascicular partition or septum (see text for explanation). (From Key, A., and Retzius, G.: Studien in der Anatomie des Nervensystems und des Bindegewebes. Stockholm, Samson & Wallin, 1876.)

Microscopic Anatomy of the Peripheral Nervous System

20% to 30% of the space in peripheral nerves; there is very little collagen in spinal roots. Fibrous collagen provides mechanical strength to the fascicle and supports the nerve fibers, both myelinated and unmyelinated, and their associated Schwann cells. Spinal roots are in a protected environment, so mechanical strength is less important than in peripheral nerves, which are subject to stresses by skeletal movements. The fascicle also contains fibroblasts, occasional macrophages and mast cells, and small blood vessels. Approximately 10% of the nuclei within the perineurium belong to fibroblasts,297 and another 2% to 9% to endogenous macrophages,170 and there are very small numbers of mast cells; the remainder are Schwann cell nuclei. Lymphocytes positive for CD4 or CD8 antigen are not usually found in the endoneurium of normal nerves, although they may occasionally be present in the epineurium. Fibroblast counts show that they make up a relatively greater proportion of cell nuclei in nerves containing a higher percentage of large myelinated fibers compared to those in which most fibers are small or unmyelinated, possibly because there are more Schwann cell nuclei per unit area in the latter situation. There are no lymph vessels in the endoneurium. The small blood vessels are mostly capillaries but may be slightly larger, with a variable number of pericytes. Arterioles are occasionally present.16 Nerve biopsies for diagnostic purposes are most commonly taken from the sural nerve because this is relatively easily accessible at the ankle, where it lies between the lateral malleolus and the tendo calcaneus. The sural nerve is a sensory branch of the sciatic nerve innervating the dorsum of the foot. Having been the subject of most attention, this nerve is the best characterized. The numbers of myelinated and unmyelinated fibers have been counted by several authors and are often used as an aid to diagnosis. The use of total myelinated fiber counts is quite rare because this requires a biopsy of the whole sural nerve (6 to 14 fascicles) rather than just a small number of fascicles. More commonly only a proportion of fascicles are removed at biopsy in order to minimize any resultant sensory deficit. Total fiber counts are also very time consuming because there may be as many as 10,000 in the sural nerve at the level of the lateral malleolus.17 The more frequently used myelinated fiber density is, however, susceptible to sampling artifact, and the results are dependent on shrinkage resulting from fixation and on other processing artifacts. The density found therefore varies slightly among different authors but is between 7000 and 9000/mm2 for myelinated fibers and 30,000 and 40,000/mm2 for unmyelinated fibers in adults.207 Myelinated fiber densities may not be helpful in interpretation, however, especially in situations in which there are amyloid deposits or Renaut bodies (see later) or excessive numbers of Schwann cells, as in hereditary motor and sensory neuropathy type 1a. Unmyelinated axons are particularly difficult to quantify because of identification problems, patchy distribution within the fascicle, and their susceptibility to fixation artifacts.

37

At birth the density of nerve fibers is approximately 26,000/mm2, and this reduces as the fibers enlarge and the quantity of fibrous collagen increases.207 The total number of fibers reaches a peak at about age 5 years, but the density continues to fall as a result of increases in myelin thickness and in the quantity of fibrous collagen. The process of maturation is quite slow; although the axon reaches its final diameter around the age of 5 years, the myelin sheath does not reach its maximum thickness until 15 years of age.368 The proportion of myelinated fibers of different calibers also changes with maturation and is unimodal in early childhood, becoming bimodal around the age of 5 years.207 Endoneurial collagen is mainly composed of collagens I and III forming long fibrils with a characteristic crossbanding every 67 nm and a mean diameter of 51 nm.375 This fibrillar material occupies much of the spaces between the nerve fibers and other components of the nerve trunk. In addition, there are other less conspicuous collagenous sheaths of the nerve fibers themselves. In large fibers two sheaths may be identified. The inner one, immediately adjacent to the Schwann cell plasma membrane, is the sheath of Plenk and Laidlaw. This is a very fine sheath of collagen fibers with a circular and oblique orientation that is continuous across the nodes of Ranvier but inflected to follow the Schwann cell basal lamina.227,319,320 By light microscopy this is indistinguishable from the Schwann cell basal lamina. Adjacent to this is another sheath of longitudinal collagen fibrils described by Key and Retzius that is not inflected at the nodes.216 The electron microscopic appearances of these structures were described by Thomas in very early electron microscopic studies of peripheral nerves.419 They have also been studied by scanning electron microscopy, which confirmed the transmission electron microscopic findings.142,439 In small and medium-sized nerve fibers, it is not possible to distinguish two separate sheaths. Perineurial collagen has a similar composition and also has a diameter of 51 nm. This forms wavy bundles that are oriented predominantly along the length of the fascicle, although some may be oriented circumferentially or obliquely. Epineurial collagen, in contrast, forms fibrils of a noticeably larger diameter (80 to 100 nm) and may have a different chemical composition, with a higher proportion of collagen I.375 True elastic tissue is not found in the endoneurium, only the oxytalan component. This is seen by electron microscopy as bundles of nonbranching fibrils with a diameter of 10 to 12.5 nm and an axial periodicity of 17.5 nm and is probably the scaffold on which amorphous elastin is deposited to make elastic fibers.157,264 The amorphous component, elaunin, is not usually seen in the endoneurium or perineurium but may be encountered in the epineurium. Fibrous long-spacing collagen (FLSC) may be occasionally encountered in the perineurium (Fig. 3–3B) or endoneurium, sometimes merging with Schwann cell basal laminae in the latter situation. In addition to nerve trunk perineurium, it may also be present in between the layers of perineurial cells that form the capsules of Meissner and pacinian corpuscles.

38

Structure of the Peripheral Nervous System

FIGURE 3–3 A, Electron micrograph showing transverse section of perineurium of normal sural nerve. Where adjacent perineurial cells overlap, they are connected by tight junctions (arrow) and there are localized regions of densely packed fibrils (asterisk). There are pinocytotic vesicles on both surfaces of the cells (arrowheads). The basal laminal layer (BL) is thick and does not possess a lamina lucida. Collagen fibrils (C) are unstained. Bar: 0.5 ␮m. B, There are bundles of fibrous long-spacing collagen (FLS) lying adjacent to the basal lamina of some cells (asterisk). The bundles of normal collagen are mostly cut in cross section, but some run obliquely or radially (arrowheads). Regions of cell overlap are indicated by an arrow. Bar: 0.5 ␮m.

FLSC forms the Luse bodies described first in schwannomas253 but has also been reported in many other situations, particularly neurofibromas. Luse bodies are well-delineated, usually fusiform structures containing a proteoglycan, probably dermatan sulfate proteoglycan,87 and have a periodicity of 100 to 150 nm. Although some immunochemical studies suggest that they are a version of collagen VI,59,60 others have failed to confirm this.87,359 These conflicting identifications may imply that FLSC can be formed from the breakdown or turnover of either collagen I, III, or VI87 or a reaction to an abnormal extracellular environment. The incidence increases with age, and it is not found in the perineurium of children. FLSC has not been reported in animal nerves. The basal laminal sheaths of Schwann cells have two regions, a dense region (the lamina densa) and a clear space

(the lamina lucida) separating it from the cell membrane. These result from the arrangement of laminin, collagen IV glycosaminoglycans, and other components. Perineurial cell basal lamina often lacks the lamina lucida and also may be considerably thicker than Schwann cell or endoneurial cell basal laminae (see Fig. 3–3A). Renaut bodies may occasionally be encountered in the endoneurium, especially at sites of compression. These are fusiform bodies mainly consisting of fibrous material, predominantly oxytalan with some fibrous collagen and fine cell processes resembling fibroblasts. These are often epithelial membrane antigen positive, suggesting that they are derived from perineurial cells.455 Renaut bodies are typically 200 to 300 ␮m in diameter and rather longer and lie along the long axis of the fascicle. They do not contain

Microscopic Anatomy of the Peripheral Nervous System

axons or Schwann cells. They most commonly lie adjacent to the inner surface of the perineurium but may be deeper in the endoneurium (Fig. 3–4). In the past they have been sometimes misinterpreted as infarcts or other abnormal structures, but they are now accepted as components of normal nerves.

Perineurium The perineurium is composed of modified fibroblasts that are connected together by tight (zonula occludens) and gap (zonula adherens) junctions to form concentric cylindrical sheaths separated by fibrous collagen (see Fig. 3–3). In early development, the nerve fibers are separated into bundles by a loose arrangement of fibroblasts. The mature arrangement with development of tight junctions and a basal laminal coating does not appear until after the commencement of myelination. In the rat, myelination commences at birth but the diffusion barrier properties of the perineurium are not detectable until 30 days after birth.35,380 This may render the immature nerve susceptible to toxins and infectious agents that would be excluded by the perineurial diffusion barrier in the adult.223 It is unknown at what age these barrier properties develop in humans.

FIGURE 3–4 Renaut bodies (asterisks) are most commonly found adjacent to the perineurium but may also be deeper in the endoneurium (human sural nerve at the ankle). Resin section stained with thionin and acridine orange.379 Bar: 100 ␮m. See Color Plate

39

Overlapping perineurial cells do not possess basal laminae between the adjacent membranes. They are linked in these regions by tight junctions; this is the mechanical arrangement underlying their diffusion barrier properties. The multiple interconnecting network formed by tight junctions is well shown by freeze-fracture investigations15,336 (Fig. 3–5). These studies showed that these junctions are far more widespread than suggested by electron microscopy of ultrathin sections. The pinocytotic vesicles (caveolae) seen by electron microscopy take up substances to be transported across the endoneurium. Gap junctions are much more frequently seen in developing nerve and are rare in mature nerve.15 These junctions allow small molecules to pass directly between perineurial cells. There are high levels of various enzymes in perineurial cells, such as phosphorylating enzymes, ATPase, and creatine kinase,372 consistent with the perineurium’s function as an active diffusion barrier. The endoneurial fluid is hypertonic compared with interstitial fluid of general connective tissues.225,280 Presumably the blood-nerve barrier created by blood vessel epithelial cells has a role in regulating this together with the barrier created by the perineurium. If the perineurium is damaged, the contents bulge outward into the epineurium and the larger myelinated fibers are demyelinated in the underlying region.399,421 During recovery after perineurial injury, fibroblasts first encircle the damaged fibers, and these then form small perineurial septa giving rise to minifascicles rather like those seen in neuromas.411 The size of the perineurial window is important in determining the degree of injury. Traumatic nerve injury in which sharp objects penetrate the perineurium may result in persistent neuropathy caused by a persisting defect in the perineurium.411 Perineurial cells also possess cytoplasmic actin-containing structures similar to the filamentous aggregates seen in smooth muscle cells. By analogy with the contractile sheath of seminiferous tubules,77 it has been suggested that these may give the whole structure contractile properties,348 but this has not been confirmed.425 This could explain the ability of the nerve fascicle to maintain its shape and adapt to changes in the volume of the perineurium. Peristaltic-like movements could also move endoneurial fluid along the fascicle. Although perineurial cells were originally thought to derive from the neural crest,373 regeneration experiments and developmental studies suggested that they were derived from fibroblasts.422 This has been confirmed by the use of a retroviral marker.62 Transgenic mice deficient in the signaling protein desert hedgehog have been used to show that its production by Schwann cells converts loosely arranged fibroblasts into mature perineurium.312 Once this has occurred, they can be maintained in vitro and retain their typical characteristics despite repeat passaging and lack of contact with any other cell type.315

40

Structure of the Peripheral Nervous System

FIGURE 3–5 Zonula occludens between overlapping borders of perineurial cells from rabbit sciatic nerve showing multiple interconnecting ridges with intervening areas containing caveolae (p). (A ⫻19,000; B ⫻63,000.) (Courtesy of Dr. G. Gabriel.)

As already discussed, the perineurium is a metabolically active structure that maintains a constant microenvironment in the endoneurial space and forms a barrier to diffusion between the epineurium and endoneurium. This barrier is only breached at the distal termination and where blood vessels traverse the perineurium. Inflammatory cells can use transperineurial vessels as a route into the endoneurium. Although these vessels are accompanied by a sleeve of perineurial cells, this sleeve terminates within the endoneurium and the cells do not fit tightly against the vascular endoneurial cells. Despite this, the combination of collagen fibrils and perineurial cells forms a sufficiently tight seal to maintain a pressure differential between the epineurium and endoneurium. The perineurium may also perform a filtering function by selectively taking up and transporting some substances by pinocytosis305 (see Fig. 3–3A). In axonal degeneration, the perineurial cells take up lipid droplets from myelin breakdown.81 There may also be lipid-laden macrophages between the cellular laminae, implying a movement of these cells from endoneurium to epineurium. The perineurium connects to the pia-arachnoid at the proximal end of the nerve trunk. In sensory nerves the perineurium extends over the dorsal root ganglion, forming the inner layers of its capsule (Fig. 3–6).177 There is an extensive meshwork of tight junctions in this location similar to that seen in the nerve trunk.337 The termination

at the distal end depends on the type of fiber. In encapsulated sensory endings the perineurium appears continuous with the outer capsule layers.371 This is confirmed by immunohistochemical studies showing that the outer layers and capsule of the pacinian corpuscle (Fig. 3–7) are derived from perineurial cells, whereas the inner laminae immunostain for S100 and Leu-7 antigen and are therefore derived from Schwann cells.442 Pacinian corpuscles have frequently been noted in the epineurium and even occasionally in the endoneurium.175 Timofeew’s corpuscles are smaller encapsulated end organs only found in late fetal and early postnatal life. These have been reported in association with autonomic nerves and ganglia in the pelvis.13 In contrast to the situation in which the perineurium is continuous with a capsule surrounding the ending, and hence the axons are always enclosed, in free sensory endings the perineurium terminates before the ending, leaving the axon exposed. Little work has been done on the occurrence of nerve endings in the nerve trunks themselves, but free endings have been identified in the epineurium, perineurium, and endoneurium.201 Investigating the distal regions of the dental pulp shows that the perineurium progressively loses basal lamina and tight junctions, commencing with the outer aspects of the sheath, until the nerve fibers are only surrounded by a loose collection of fibroblasts.295 Similarly, at the

Microscopic Anatomy of the Peripheral Nervous System

41

FIGURE 3–6 Diagram illustrating relationships of the peripheral nerve sheaths to the meningeal ensheathment of the spinal cord. The epineurium (EP) is in continuity with the dura mater (DM). The endoneurium (EN) persists from the peripheral nerves through the spinal roots to their junction with the spinal cord. At the subarachnoid angle (SA), the greater portion of the perineurium (P) passes between the dura and the arachnoid (A), but a few layers appear to continue over the roots as the inner layer of the root sheath (RS). The arachnoid is reflected over the roots at the subarachnoid angle and becomes continuous with the outer layers of the root sheath. At the junction with the spinal cord, the outer layers become continuous with the pia mater (PM). (From Haller, F. R., and Low, F. N.: The fine structure of the peripheral nerve root sheath in the subarachnoid space in the rat and other laboratory animals. Am. J. Anat. 131:1, 1971, with permission.)

motor end plate, the perineurium terminates before the formation of the end plate proper and forms a bellshaped covering that does not make direct contact with the nerve or the muscle.353 A gap of about 1.5 ␮m separates the perineurium from the muscle basal lamina, also rendering the nerve fiber susceptible to external influences at this point, as is also the case with free sensory endings.

Epineurium The quantity of connective tissues surrounding the nerve fascicles varies considerably depending on the location of the nerve. There is only a loose connection with the

surrounding connective tissues, so that the nerve trunks are relatively mobile. This presumably reduces the likelihood of damage by entrapment during movement of joints. In general there is a more extensive epineurium at joints and in trunks with greater numbers of fascicles than in smaller nerves with few fascicles. The epineurium can be separated into two layers, the epifascicular epineurium and the interfascicular epineurium. Mice deficient in desert hedgehog not only lack a true perineurium but have very sparse epineurial collagen.312 This suggests that the development of well-defined nerve fascicles is required for the formation of an epineurium. There is no epineurium in spinal roots, where the root sheath is a looser, less compact layer.

42

Structure of the Peripheral Nervous System

protect the nerve fascicles from compression damage.412 The loss of these fat deposits in generalized wasting may be a factor in the development of pressure palsies in wasted, bedridden patients.

Myelinated Nerve Fibers C.-H. Berthold, R. H. M. King, and M. Rydmark

FIGURE 3–7 Transverse section through a pacinian corpuscle within the epineurium of a fascicle in human sural nerve. A large myelinated fiber is just visible at the center of the concentric layers of cell processes (arrow). Resin section stained with thionin and acridine orange. Bar: 100 ␮m. See Color Plate

Nerve trunks may be distinguished from other longitudinal structures by naked eye observation as a result of the white spiral bands seen on their surface. These are termed the spiral bands of Fontana.115 Stretching or damaging the nerve changes or destroys these bands and results in a homogeneous gray appearance. Fontana deduced that these bands resulted from the zigzag course of the nerve fibers within the nerve trunk.115 This allows stretching of the nerve by up to 20% without damaging the nerve fibers.76 More recent work, however, showed that the wavy pattern actually originates in the arrangement of the collagen fibrils in the inner layer of the epineurium, which lie in wavy bundles.403 These waves may be necessary to allow the nerve to straighten without stretching the nerve fibers, because collagen fibrils are not extensible. A longitudinal anastomotic network of arterioles and venules provides the blood supply to the nerve fascicles. Epineurial capillaries may be fenestrated, unlike those in the endoneurium. In addition, there is a lymphatic capillary network in the epineurium that accompanies the arteries and is connected to the regional lymph nodes but without connections to the endoneurium.413 There may also be considerable quantities of adipose tissue in the outer layer of the epineurium, especially in the sciatic nerve in the buttocks and thigh. This distribution has led to the suggestion that this fat may act as a buffer to

The microscopic and molecular anatomy of peripheral myelinated nerve fibers has been reviewed several times during the past decade.18,20,34,156,228,229,322,360–362,425,452 The following description is based mainly on the microscopic organization as found in peripheral myelinated mammalian (rat, guinea pig, rabbit, cat, human, and other) nerve fibers more than 4 to 5 ␮m in diameter. Particular emphasis is given to the node of Ranvier and its close internodal neighborhood. The peripheral myelinated nerve fiber (Fig. 3–8) consists of a single continuous neuronal process, the axon, and a set of Schwann cells arranged serially along the outside of the axon. Each Schwann cell enwraps the axon segment with which it is associated with an electrically insulating, multilayered cell membrane specialization, the myelin sheath. The meeting points of consecutive Schwann cells appear as short (approximately 1 ␮m long) constricted and myelin-free fiber segments, the nodes of Ranvier. The intervening 300- to 2000-␮m-long fiber segments, each corresponding to the extension of one Schwann cell with a length proportional to fiber size, are the internodes (see Reynolds and Heath339). An internode can be separated into three parts: a main internodal region that constitutes about 95% of the internodal length, holds the Schwann cell nucleus at its midpoint, and displays the biconical myelin sheath clefts referred to as Schmidt-Lanterman incisures; and a proximal and a distal end region, each corresponding to 2% to 3% of the internode (“proximal” and “distal” refer to the cell body). The transition from the main internodal region to the end regions is gradual and indicated by an increasing amount of mitochondrion-rich Schwann cell cytoplasm outside a more or less crenated myelin sheath (Fig. 3–9A to E). As a rule, the end regions are dilated, the distal one slightly more than the proximal one, forming the paranodal bulbs of classic light and electron microscopy.250,460–462 In transverse section, a myelinated nerve fiber shows two main concentric zones: an outer Schwann cell zone and an inner axonal zone. The two zones are separated by the periaxonal space (see Fig. 3–10C later). Internodally the Schwann cell zone can be divided into four concentric regions: (1) an outermost basal lamina that sharply

Microscopic Anatomy of the Peripheral Nervous System

43

FIGURE 3–8 Schematic drawing to show the major components of a large myelinated nerve fiber and the terminology used. ASN ⫽ axon–Schwann cell network; B ⫽ mitochondrion bag; CON ⫽ constricted axon segment; IE Reg. ⫽ internodal end region; IN ⫽ internode; JPS ⫽ juxtaparanodal segment; NoR ⫽ node of Ranvier; PS ⫽ paranodal segment; PNP ⫽ paranode-node-paranode region; SCN ⫽ Schwann cell nuclear region, which is situated more or less at the midpoint of an internode. According to this view, an internode consists of a proximal and a distal end region (IE Reg.) and between these the main internodal region. An internodal end region is separated in a juxtaparanodal and a paranodal segment (IE Reg. ⫽ JPS ⫹ PS). A node of Ranvier and its two bordering internodal end regions make up a PNP region.

demarcates the fiber from the endoneurial space and runs uninterrupted all the way between the central nervous system (CNS) and the fiber terminations, (2) the outer (abaxonal) cytoplasmic compartment, (3) the myelin sheath, and (4) the inner (adaxonal) cytoplasmic compartment. At a node of Ranvier, the Schwann cell zone consists of the basal lamina and nodal constituents derived from the outer cytoplasmic compartment. The transverse contour of the myelin sheath varies according to transverse level as a result of the distribution of the Schwann cell cytoplasm outside the sheath. Thus the contour of the myelin sheath is more or less indented by the Schwann cell perikaryon, equipped with shallow notches (see Fig. 3–9A) along the stretch between the perikaryon and the internodal end regions and crenated in the latter (see Figs. 3–9C to E). The axonal contour adapts to that of the myelin sheath except at the nodes of Ranvier, where it is approximately circular (see Fig. 3–9D to F). A close-to-circular myelin sheath/axonal contour is also noted at sites where a Schmidt-Lanterman incisure enters the myelin from the outer Schwann cell compartment (see Fig. 3–9B). Three morphometric variables are currently used in order to define a myelinated nerve fiber, assuming a circular transverse fiber section: (1) d ⫽ the diameter of

the axon measured in the main internodal region, avoiding the Schwann cell perikaryon and Schmidt-Lanterman incisures; (2) nl ⫽ number of myelin sheath lamellae; and (3) il ⫽ internodal length. From these three a number of secondary variables can be calculated, including fiber diameter (D ⫽ d ⫹ 2 ⫻ nl ⫻ lamellar thickness), myelin sheath transverse area, and internodal myelin volume.32,287 The size of mammalian peripheral myelinated axons, given as a d value, is in the range 1 to 20 ␮m; few exceed 15 ␮m. A rough estimate of il is obtained from il ⫽ d ⫻ 100. Local variations in fiber variables such as axon diameter, myelin sheath thickness, internodal distance, branching, and size of branches all influence axon function and impart integrative capacities to the axon.448 The microscopic organization of internodal main regions is relatively simple compared with that of a node of Ranvier and its two bordering internodal end regions. These parts of myelinated nerve fibers—the paranode-node-paranode (PNP) regions—constitute, structurally as well as functionally, the most spectacular parts of a myelinated nerve fiber. The PNP regions are responsible for the generation and propagation of the action potential.342 They interfere with axoplasmic transport,21,29,327 and they are reactive centers in

44

Structure of the Peripheral Nervous System

A

B

C

D

FIGURE 3–9 Electron micrographs of transversely sectioned alpha motor nerve fibers; L7 ventral spinal root, adult cat. A, Main internodal region. The sectioning plane runs outside the Schwann cell perikaryon and in between Schmidt-Lanterman incisures. Strands of the outer Schwann cell compartment, containing a few mitochondria, are at the arrowheads. At this level the transverse contour of the myelin sheath is slightly undulated. Ax ⫽ axon. Bar: 2 ␮m. B, Main internodal region. The sectioning plane runs through a level where a Schmidt-Lanterman incisure reaches the outer Schwann cell compartment (arrowheads). At such levels the myelin contour is approximately circular and surrounded by a substantial layer of the outer Schwann cell compartment; in this way there is a thin cytoplasmic girdle that cross-connects the longitudinal strands of cytoplasm (arrowheads in A) that run between the perikaryon and the node of Ranvier. Note the more or less angulated contours of adjacent fibers without Schmidt-Lanterman incisures. Ax ⫽ axon. Bar: 2 ␮m. C, Juxtaparanodal segment about 10 ␮m from the node. The arrow indicates one of six mitochondrion bags. Ax ⫽ axon. Bar: 2 ␮m. D, Transition level between the juxtaparanodal and paranodal segments, about 5 ␮m from the node of Ranvier. The innermost myelin lamellae have formed the “first” turns of the terminal cytoplasmic spiral and attach to a constricted and rather circular axonal contour. Asterisks indicate three of the six extensive axon–Schwann cell networks situated inside the myelin crests. Ax ⫽ axon. Bar: 2 ␮m. Figure continued on opposite page

45

Microscopic Anatomy of the Peripheral Nervous System

F

E

FIGURE 3–9 Continued E, Paranodal segment about 3 ␮m from the node of Ranvier. Arrow indicates a mitochondrion bag. Ax ⫽ axon. Bar: 2 ␮m. F, Node of Ranvier. The nodal axon (Ax) is surrounded by a nodal gap (Ng) filled with radiating Schwann cell microvilli. Note the tuftlike arrangements of the microvilli at the lower left and the upper right part of the gap. PN ⫽ perinodal space. Bar: 2 ␮m.

a large number of neuropathies (see later),50,51,187,205,277,381 as well as during the early phase of wallerian degeneration,453 and in collateral sprouting.166,199,200,263,461 The structural organization of the PNP regions differs according to fiber size. Fibers of a diameter less than 4 to 5 ␮m have relatively simple PNP regions that are referred to as “nodes of Ranvier type I” (see Fig. 3–11A later). Larger fibers, the ones treated here, have “nodes of Ranvier type II” (see Fig. 3–11B later).318,328 In the present context, “paranode” has been used (1) in a broad sense, according to traditional concepts229,332,454,463 that can be traced back to Ranvier himself, referring to the whole “internodal end region” of Figure 3–8 and used here only in the notion of the “PNP region”; and (2) in the more usual contemporary, restricted sense, referring to those segments of a myelinated nerve fiber that include the attachment of the myelin sheath to the axon.

GENERAL ORGANIZATION OF MYELINATED AXONS The axon consists of a relatively firm gelatinous cord of neuronal cytoplasm, the axoplasm,174,249 enclosed by the axolemma (i.e., by the cell membrane of the axon). The axon is separated from the adaxonal Schwann cell membrane by the periaxonal space.

Periaxonal Space According to the preparative procedure used, the periaxonal space has been reported to measure from about 30 nm down to just a few nanometers (Fig. 3–10C). At some sites the two facing cell membranes appear to fuse and form five-layered membrane complexes, reminiscent of tight junctions (Fig. 3–10C). Freeze-fracture studies, however, have not shown tight junctions between the axolemma and the adaxonal Schwann cell membrane. In contrast, an extensive tight junctional complex has been demonstrated in the mouth of the inner mesaxon (see later).365 The Schwann cell membrane that faces the periaxonal space throughout an internode shows a strong myelin-associated glycoprotein (MAG) positivity.324 The periaxonal space is probably sealed off from the mesaxonal space of the compact myelin, being open only at the nodes, where it apparently communicates more or less freely with the endoneurial space via the nodal gaps as indicated by the accessibility of different tracer substances.176 In large fibers, however, not even lanthanum enters the periaxonal space from the nodes.427

Axolemma The axolemma appears as a conventional three-layered unit membrane about 8 nm thick when examined in standard electron microscopic preparations. Freeze-fracturing demonstrates that the outside of the inner leaflet of the

FIGURE 3–10 A, Electron micrograph of an alpha motor nerve fiber sectioned transversely about 15 ␮m from the node of Ranvier; L7 ventral spinal root, adult cat. Microtubular (MT) and neurofilament (NF) domains are well separated. Arrows point at axoplasmic reticulum profiles; arrowheads at vesiculotubular membranous profiles. m ⫽ mitochondrion. Bar: 0.2 ␮m. B, Freeze-etched stereo pair electron micrographs of saponin-treated frog axon showing 11-nm neurofilaments linked by a ladder-like arrangement of 4- to 6-nm cross-bridges 20 to 30 nm in length. Bar: 0.1 ␮m. (From Hirokawa, N.: Cross-linker system between neurofilaments, microtubules, and membranous organelles in frog axons revealed by the quick-freeze, deep-etching method. J. Cell Biol. 94:129, 1982, with permission.) C, Electron micrograph of an alpha motor nerve fiber sectioned transversely through the main internodal region; L7 ventral spinal root, adult cat. The specimen was fixed in glutaraldehyde ⫹ tannic acid followed by OsO4 ⫹ potassium ferrocyanide. With this method, organelles appear coated by a diffuse layer of electron-dense material. The lipid (intermediate) layer of the cytomembranes appears like a “white” (clear) line, approximately 3 nm in width. There is a thick axoplasmic cortex (arrowheads) inside the axolemma (white arrow). Neurofilaments (NF) are interconnected by diffuse electron-dense bridges. One of several microtubules is encircled. Asterisks indicate axoplasmic reticulum profiles associated with the axoplasmic cortex. Periaxonal space is marked with one dot (•) and the inner adaxonal-cytoplasmic Schwann cell compartment by two dots (••). The axolemma and the Schwann cell membrane seem to fuse at several sites (black arrows). My ⫽ myelin sheath; m ⫽ mitochondria. Bar: 0.1 ␮m.

Microscopic Anatomy of the Peripheral Nervous System

axolemma, the so-called P face, is comparatively rich in intramembranous particles (IMPs) (see Fig. 3–12C later). The inside of the outer leaflet, the so called E face, displays, except at nodes of Ranvier (see later), relatively few IMPs.145,272,405 Many E-face IMPs are about 10 nm in size and may represent voltage-sensitive Na⫹ channels.224,344,404,449 The axolemma conveys signals between the neuron and its Schwann cells that control the proliferative and myelinproducing functions of the Schwann cells and partly regulate axon size.46,49,84,85,171,172,194,239,355,429 The additional presence of receptor and transducer proteins involved in axon–Schwann cell communication has been postulated by several authors.108,109,209,244 The axolemma is stabilized by the immediate subjacent part of the axoplasm, which here forms the axoplasmic cortex (see Fig. 3–10C), an outer condensed part of the cytoskeletal microtrabecular matrix containing ankyrin, fodrin, actin, and A-60.112,151,184,189,192,194,241,335,366,438

Axoplasm The axoplasm consists of a cytosol and formed elements suspended in it. The formed elements are (1) the axoplasmic organelles (i.e., mitochondria, the axoplasmic [endoplasmic] reticulum, dense lamellar bodies, multivesicular bodies, vesiculotubular profiles, and membranous cisterns); (2) the cytoskeleton; and (3) axoplasmic inclusions. Organelles and inclusions show their highest occurrence and their most elaborate arrangements in the PNP regions of large myelinated axons. Ribosomes and Golgi apparatuses are absent, though stray ribosome-like granules have been reported (see Alvarez3). Organelles Numerical data referring to the frequency of various axoplasmic organelles in the main internodal region are meager.20,475 Mitochondria. Axonal mitochondria are 0.1 to 0.3 ␮m in diameter and 0.5 to more than 10 ␮m in length. Usually they lack electron-dense granules and have flattened or tubelike, often longitudinally oriented, cristae. The concentration of axonal mitochondria decreases with increasing axon diameter from about 1/␮m2 transverse area of axoplasm in the thinnest myelinated axons to about 0.1/␮m2 in the thickest ones.20 A close relationship between mitochondria and the axoplasmic reticulum (AR) has been demonstrated ultrastructurally, and formation of axonal mitochondria in relation with the AR has been suggested.396 Mitochondria are transported both anterogradely and retrogradely in the axon. Transportation is saltatory: Short periods of movement at a velocity similar to that of other organelles are interrupted by long periods of rest.116,117 The outer membrane of the mitochondrion is considered multivalent in terms of sites capable of interacting with the force-generating transport mechanism.167,256

47

The Axoplasmic Reticulum. The AR is to be distinguished from the smooth membranous vesicular, tubular, and vesiculotubular profiles that occur scattered in small amounts all along the axon and accumulate in large numbers proximal to a transportation block.168,246,431 The AR forms a system of delicate varicose membranous tubules about 20 to 60 nm in diameter and flattened sacs, up to 100 nm in width, that branch and interconnect frequently as they extend lengthwise throughout the axon from the hillock into the terminals.246,330,331 The rough endoplasmic reticulum of the soma connects via a “transitional endoplasmic reticulum” to the AR in the proximal part of the axon hillock.246 A connection between the AR and the Golgi apparatus has also been demonstrated.325,402 The AR can be subdivided into a minor, outer, subaxolemmal and a major, inner, central compartment. The former is situated just inside the axolemma, where, embedded in the axoplasmic cortex, it forms a single layer of tubes and sacs that interconnects with the central AR.21,331 In view of abundant branching and a minimum tubular caliber of less than 20 nm, the proposition that the AR extends uninterrupted throughout the axon has been hard to contradict. The results of cold-block experiments indicate that the AR itself is not rapidly transported either anterogradely or retrogradely.104 Most likely, the AR is a vector in slow anterograde transport and is not involved in retrograde transport.246 It acts as a reservoir for Ca2⫹ ions.97 Dense Lamellar Bodies and Multivesicular Bodies. These bodies vary in shape and size from short tubes 0.1 ⫻ 0.2 to 0.5 ␮m to rounded entities about 0.5 ␮m in diameter (Fig. 3–11B; see also Figs. 3–15A and 3–15D later). Dense lamellar bodies (DLBs) and multivesicular bodies (MVBs) are rare outside the PNP regions.28 The ultrastructural appearance of the DLBs and the MVBs and their occasional content of acid hydrolases indicate that they belong to a heterogeneous group of organelles that includes secondary lysosomes and residual bodies.148–150 They are considered to be transported retrogradely.104,432 Vesiculotubular Profiles. These elements are 0.05 to 0.1 ␮m in diameter and up to 0.5 ␮m (the tubes) in length (see Fig. 3–11B; see also Fig. 3–15A and C later). Some seem empty, whereas others contain granular electron-dense material. As a rule, the tubes appear varicose or even beaded. Tubes and vesicles lie close together in the PNP regions, forming rows or strands together with AR profiles in association with bundles of microtubules (MTs). 106,431 These organelles are transported anterogradely in the axon. 104,432 They probably derive from the Golgi apparatus of the soma and act as fast-transport vectors for newly synthesized protein and lipid.106,179,208,246,406,428

48

Structure of the Peripheral Nervous System

B B PN

Ax

C

B B

Ax

Microscopic Anatomy of the Peripheral Nervous System

Membranous Cisterns. All membrane-demarcated and empty-looking entities more than 0.15 ␮m in size are here classified as cisterns. This is a heterogeneous group that includes vacuoles, of which some are probably large endosomes and some preparatory artifacts. The Cytoskeleton This, the most conspicuous of the axoplasmic components,64 consists of the MTs, the neurofilaments (NFs), and the microtrabecular matrix, including the microfilaments (Fig. 3–12B; see also Fig. 3–10).101,105,191,435 It defines the size, shape, stability, and growth pattern of the axon. It contains the machinery necessary for bidirectional axoplasmic transport, that is, the interaction with the anterograde transport “motor” (kinesin) and the retrograde “motor” (dynein).167,374 Axoplasmic transport and its molecular correlates differ in several aspects from that in the cell body and the dendrites.61,197 There are also differences in the composition of the axonal cytoskeleton with respect to type of peripheral nerve, distance from cell body, and age.291,417,447 Microtubules extend longitudinally in the axon and form the tracks along which the fast bidirectional axoplasmic transport of membranous organelles takes place.456 They are approximately 25 nm thick, are unbranched, and vary from a few to more than 1000 ␮m in length.45,65,433 They lie solitary in the axoplasm or form small bundles.243 The number of MTs varies inversely with axon size from 30 to 40/␮m2 cross-sectioned axoplasm in small axons to 10 to 15/␮m2 in large axons.20,107,195 The MT packing varies along the course of an axon354 as well as between afferent and efferent ones.309,354,475–477 MTs are fragile structures, and about 50% of the MT mass is particularly labile and disassembles easily.11,12 They are, however, well preserved after glutaraldehyde–tannic acid fixation.

49

Neurofilaments run longitudinally in the axon, often with a slight spiraling course. They are approximately 10 nm thick, of undefined length, and unbranched. There are about 150 to 200 NFs/␮m2 cross-sectioned axoplasm, their density normally being relatively unaffected by the size of the axon. NFs are considered to be the “major determinant” of axon size.193–195,311,435,444 Three kinds of NF monomers polymerize together to form a neurofilament: NF-L (68 kDa), NF-M (150 kDa), and NF-H (200 kDa), with all three existing in many different isoforms depending on the degree of phosphorylation.257,289 The less phosphorylated, the higher is the probability that the NF units exist as mobile mono- or oligomers and the more labile are polymers of such units. NF density and degree of phosphorylation increase with distance from the cell body and with age. The microtrabecular matrix is only faintly indicated in electron micrographs of axons submitted to conventional preparative protocols. It then appears as thin strings of a wispy material that radiate from and interconnect the NFs and some MTs (see Fig. 3–10A). Other preparatory procedures, including quick freeze in combination with deepetching and rotatory shading (see Fig. 3–10B), examination of unfixed tissue in the high-voltage electron microscope, and the use of tannic acid as a mordant in combination with glutaraldehyde fixation (see Figs. 3–10C and 3–12B), have all added new features to an otherwise rather emptylooking axoplasm.21,105,261,434 The most spectacular of these is the plethora of microtrabecules (cross-linkers) that radiate from the shafts of the NFs and the MTs, linking them into a dense lattice that extends throughout the axoplasm (see Fig. 3–10).189,192,266,366 The axoplasmic organelles are suspended in the NF-MT lattice and cross-linked to it by microtrabecules. Three main types of microtrabecules, all 4 to 6 nm thick and 20 to 150 nm long, have been

FIGURE 3–11 A, Electron micrograph of longitudinal section from a human sural nerve showing a small myelinated fiber with a type I node of Ranvier. The paranodal segments (Para) are smoothly tapered. Few nodal microvilli are distinguishable. There is an extensive collection of mitochondria (asterisk) together with glycogen deposits (G). The paranodal myelin lamellae are connected together by a strip of dense desmosome-like structures (arrow). Bar: 1 ␮m. B, Electron micrograph of longitudinal section through a highly segregated CON segment and its proximal (P) and distal (D) internodal end regions; alpha motor nerve fiber, L7 ventral spinal root, adult cat. Dense lamellar bodies (DLBs), multivesicular bodies (MVBs), filled vesiculotubular profiles, and some mitochondria occupy the clear axoplasm distal to the nodal midlevel. Large numbers of empty-looking vesiculotubular profiles and some mitochondria are present in the finely granular axoplasm proximal to the nodal midlevel. Note the bulging contour of the proximal paranodal segment (Para). ASN ⫽ axon–Schwann cell network. Bar: 2 ␮m. C, Electron micrograph of longitudinal section through a PNP region; alpha motor nerve fiber, L7 ventral spinal root, adult cat (sectioning plane is outside the nodal gap). PN ⫽ perinodal space between the nodal ends of the two adjacent paranodes. The myelin sheaths are crenated and equipped with longitudinal crests and intervening furrows. The latter are filled with mitochondrion-rich Schwann cell cytoplasm forming the mitochondrion bags (B). Note electron-dense glycogen granules dispersed among the mitochondria. Ax ⫽ axoplasm inside a myelin crest. Bar: 2 ␮m. D, Electron micrograph showing detail from a longitudinally sectioned mitochondrion bag of an internodal end region; alpha motor nerve fiber, L7 ventral spinal root, adult cat. The Schwann cell cytoplasm contains masses of densely packed mitochondria (Mi) and clusters of glycogen particles (G). My ⫽ myelin sheath; E ⫽ endoneurial space. Bar: 0.5 ␮m.

B

C

FIGURE 3–12 A, Electron micrograph of longitudinal section through a nodal gap (NG) containing densely packed microvilli emanating from the Schwann cell nodal collars (SC); alpha motor nerve fiber, L7 ventral spinal root, adult cat. The nodal collars separate the nodal gap from the perinodal space (P). Asterisks in the myelin sheath indicate the tips of “ear-of-barley”–like arrangements of the terminal cytoplasmic pockets. The nodal axolemma extends between the two bars (note thick electron-dense coating inside the axolemma in this GMA embedding), while the node gap proper extends between the two arrowheads. A nodal recess extends between a bar and an arrowhead. The nodal gap walls (W) overlie each recess and delineate the node gap from the myelin sheath. Ax ⫽ axon. Bar: 0.5 ␮m. (From Berthold, C. H., and Rydmark, M.: Electron microscopic serial section analysis of nodes of Ranvier in lumbosacral spinal roots of the cat: ultrastructural organization of nodal compartments in fibres of different sizes. J. Neurocytol. 12:475, 1983, with permission.) B, Electron micrograph of transversely sectioned nodal axon segment; alpha motor nerve fiber, L7 ventral spinal root, adult cat. Two arrowheads indicate the axolemma. The axoplasmic cortex is particularly well developed (white bar) and measures 60 to 100 nm. Matrix substance coats all formed elements and connects some of them directly with the axoplasmic cortex. Arrows mark out crosscut AR profiles. MTs are encircled. Asterisks are in organelles of the vesiculotubular type. Ng ⫽ node gap. Glutaraldehyde and tannic acid followed by osmium tetroxide and potassium ferrocyanide. Bar: 0.1 ␮m. Legend continued on opposite page

Microscopic Anatomy of the Peripheral Nervous System

identified in frog myelinated axons.189,192,232,241,243 After tannic acid treatment, all membranous organelles, the NFs, the MTs, and the inside of the axolemma appear coated by a 3- to 30-nm thick, uneven, at times coarsely granular layer of material from which the microtrabecules seem to emanate (see Fig. 3–10C). The matrix layer coating the inside of the axolemma (see Figs. 3–10C and 3–12B) is referred to as the axoplasmic cortex.105 Previous observations have suggested that the cytoskeleton is a rather immobile element in the axoplasm. The idea formerly accepted—that the whole NF-MT lattice slides slowly down the axon235,236—was seriously questioned,300,302,303 as it had been already from some earlier experimental studies.397 The slow component-a (SCa) of axoplasmic transport that contains tubulin and the NF appeared to be gradually incorporated in a “stationary” polymer lattice and renewed it by local disassembly and reassembly.301 More recent observations have shown that MTs and NFs are, indeed, actively transported along axons by fast motors, although in an infrequent and asynchronous way interrupted by long pauses (see Baas,10 Miller et al.,271 and Wang and Brown446). A similar situation may also apply to members of the other slowly transported component (SCb). In addition to proteins heading for the terminals, such as synapsin and clathrin-uncoating protein complexes,38 this component contains proteins that belong to the microtrabecular matrix, such as actin (during transport probably complexed with an actin depolymerizing factor47), fodrin (brain spectrin), myosin, tau,221 tropomyosin, and calmodulin.39,112,236,237,243 The axoplasm can, according to the distribution of its cytoskeletal components, be separated into three principal domains: (1) a subaxolemmal domain, (2) domains characterized by NFs, and (3) domains characterized by MTs and diffuse granular material.366 AR is present in all three domains. The subaxolemmal domain, mainly studied in squid giant axons, where it is several micrometers thick,44 should in vertebrate axons correspond to the axoplasmic cortex and the outer aspect of the subjacent axoplasmic zone. The domain interconnects the axolemma and the cytoskeleton and contains ankyrin, fodrin (brain spectrin), actin, and A-60.184,189,192,241,268,366,434 It includes the outer compartment of the AR, which covers 10% to 20% of the inner aspect of the axolemma.21 Fast anterograde axoplasmic transport seems to have a predilection for the subaxolemmal domain.331,393 The NF domain constitutes the shape- and

51

size-supporting framework of the axon. It is, as seen in transverse sections, pierced and separated in interconnecting subdomains by the “channels,” “tunnels,” or “streets” formed by the MT domains (see Fig. 3–10A), along which rapid anterograde and retrograde transport of membranous organelles takes place. Granular Material The presence of granular material (in the present context this refers to electron-dense granules as seen after conventional preparative procedures) is sporadic and is usually restricted to the PNP regions. There are two main types of granules, coarse ones of high electron density and minute ones that aggregate in areas with a pepper-like appearance.32 Granules with high electron density measure from about 25 nm to greater than 100 nm. Of these, the smaller are classified as glycogen granules.20,318,327,474 They usually form clusters and become more common with increasing age.31,424 Tiny granules that form territories of a pepper-like texture seem to be present only in the PNP regions, where they, as a rule, occupy either the proximal or the distal half (see Fig. 3–11B). They are less than 10 nm in diameter and lie packed closely together, obscuring MTs and NFs. Most likely, the accumulations of these tiny granules give rise to the diffuse staining seen on light microscopy in the PNP regions of some axons.28,273,328 These may represent specific proteins, but the precise identity of the granular material is unknown.

THE PARANODE-NODE-PARANODE REGION A PNP region consists of a central node of Ranvier and two bordering internodal end regions. The latter can, with reference to the node, be separated in two parts: (1) a short (irrespective of fiber size) 3- to 5-␮m long paranodal segment situated adjacent to the node; and (2) more internodally, a juxtaparanodal segment the length of which is proportional to fiber size. The juxtaparanodal segment (see Fig. 3–9C) is defined by the changing shape and size of the Schwann cell, the myelin sheath, and the axon as compared with their appearance in the main internodal region. The paranodal segment is defined by the termination of the myelin sheath and its attachment to the axon (see Fig. 3–9E). The transition between the juxtaparanodal and the paranodal segment is, in fibers more than 4 to 5 ␮m

FIGURE 3–12 Continued C, Electron micrograph of freeze-fracture preparation of rodent sciatic nerve exposing the P face of the axolemma of the node (N) and paranode region of a myelinated fiber. There is a dense population of intramembraneous particles on the nodal axolemma. The paranodal region is characterized by the presence of alternating ridges (arrows) and troughs in the axolemma related to the attachment of the helically arranged terminal myelin loops (TL). The parallel transverse bands associated with their attachment are visible on the axolemma. Direction of platinum shadowing is indicated by the arrowhead in the lower left corner. Bar: 0.2 ␮m. (Courtesy of N. G. Beamish.).

52

Structure of the Peripheral Nervous System

in diameter, indicated by a reduction of the axon transverse area by 75% or more350 and by the axon adopting an almost circular transverse contour (see Fig. 3–9D). The axon keeps a circular transverse outline as it passes through the proximal and distal paranodal myelin sheath cuffs and the intervening nodal gap. In this way there is at the middle of a PNP region a 7- to 10-␮m-long constricted, fairly cylindrical axon segment: the CON segment. The CON segments and their radial Schwann cell surroundings form together the functionally and structurally most complex and also the most vulnerable parts of a myelinated nerve fiber. The molecular machinery

necessary to generate trains of action potentials is localized here (for review and references, see Waxman and Ritchie452). Here also are the conspicuous zones of myelin sheath attachment to the axon. The CON segments furthermore have to maintain axoplasmic flow through a tube whose transverse area is reduced by 75% to 90% as compared with that of the main internodal region, a reduction nevertheless shown to be advantageous for conduction velocity.178,262 The CON segment is considered separately later, after descriptions of the internodal end region and the node of Ranvier. Morphometric data are presented in Table 3–1.

Table 3–1. Paranode and Node of a Large Cat Alpha Axon: A Morphometric Description General fiber characteristics Diameter (D) ⫽ 17.5 ␮m Axon diameter (d) ⫽ 12.5 ␮m Number of myelin lamellae ⫽ 140 Myelin sheath thickness ⫽ 2.5 ␮m Internodal length ⫽ 1800 ␮m Internodal End Region Juxtaparanodal segment (values refer to 1 region) Length Axon mean cross-sectional area Axon maximum circumference Axon mean circumference Axon membrane area Myelin sheath maximum cross-sectional area Myelin sheath maximum circumference Schwann cell adaxonal membrane area Schwann cell adaxonal cytoplasmic maximum cross-sectional area Schwann cell adaxonal cytoplasmic volume Schwann cell endoneurial membrane area Schwann cell outer cytoplasmic compartment maximum cross-sectional area Schwann cell outer cytoplasmic compartment volume Schwann cell mitochondria, maximum number noted in a cross section Schwann cell mitochondria, calculated total number (assumed mitochondrion size: 0.15 ⫻ 0.5 ␮m) Paranodal segment (values refer to 1 region) Length Axon diameter Axon cross-sectional area Axon circumference Axon membrane area Axon volume Terminal cytoplasmic spiral, number of turns Terminal cytoplasmic spiral, number of turns attached to axolemma Terminal cytoplasmic spiral, length Terminal cytoplasmic spiral, diameter of cord Terminal cytoplasmic spiral, membrane area Terminal cytoplasmic spiral, volume Schwann cell outer cytoplasmic compartment cross-sectional area Schwann cell outer cytoplasmic compartment volume Schwann cell endoneurial membrane area

75 ␮m 102 ␮m2 63 ␮m 49 ␮m 3680 ␮m2 188 ␮m2 88 ␮m 3680 ␮m2 0.25 ␮m2 15 ␮m3 4330 ␮m2 28 ␮m2 1214 ␮m3 450 10.000 4 ␮m 4.7 ␮m 17.7 ␮m2 14.9 ␮m 60 ␮m2 71 ␮m3 140 26 2530 ␮m 0.1 ␮m 795 ␮m2 20 ␮m3 8 ␮m2 32 ␮m3 200 ␮m2

4%* 83% 160% 125% 5% 160% 160% 5% 160% 5% 5% 230% 6% ⬎2000% — 0.2% 38% 14% 38% 0.08% 0.03% — — — — — — 65% 0.2% 0.3%

Table continued on opposite page

53

Microscopic Anatomy of the Peripheral Nervous System

Table 3–1. Paranode and Node of a Large Cat Alpha Axon: A Morphometric Description—Continued Node of Ranvier Axon compartment Axon length Axon diameter Axon cross-sectional area Axon circumference Axon membrane area Axon volume Schwann cell compartment Extracellular interconnection area between endoneurial space and node gap Node gap width Node gap height Node gap extracellular volume Node gap Schwann cell microvilli number mean length mean diameter total membrane area total volume Node gap walls, membrane area Node gap recesses, ceiling area Paranode-Node-Paranode axon region Mitochondria Multivesicular bodies Dense lamellar bodies Vesiculotubular membraneous profiles

1 ␮m 5 ␮m 20 ␮m2 16 ␮m 24 ␮m2 19 ␮m3

0.006% 40% 16% 40% 0.03% 0.009%

0.2 ␮m2 0.3 ␮m 1 ␮m 0.7 ␮m3

— — — —

800 1 ␮m 84 nm 210 ␮m2 4.4 ␮m3 43 ␮m2 11 ␮m2

— — — — — —

— — — —

2.6% 100% 92.6% 97.0%

*Percentage values represent parts of the corresponding internodal total. Data are taken from Rydmark and Berthold,351 Rydmark et al.,352 and Berthold et al.28 and are compensated for preparative dimensional changes.27

Internodal End Region The juxtaparanodal segment is characterized by an irregular myelin sheath contour and a comparatively voluminous outer Schwann cell cytoplasmic compartment. In many mammalian species, including humans, this cytoplasm is concentrated in longitudinal cords that extend with an increasing volume from the main internodal region to the node of Ranvier (see Figs. 3–9C and D). The cords, one or two in small fibers and up to seven in the largest ones, “indent” the underlying myelin and axon. This gives rise to conspicuous myelin sheath irregularities, often like longitudinal crests, that in transverse sections produce the typical appearance of a deeply crenated myelin sheath that surrounds a correspondingly fluted axon.20,231,327,463 In fibers more than 5 to 6 ␮m in diameter, the cords of Schwann cell cytoplasm become extraordinarily rich in mitochondria as the node is approached and form the “mitochondrion bags” of the internodal end region (see Figs. 3–9C and 3–11C and D; see also Table 3–1).352 The number of mitochondria reaches a maximum value at a level about 10 ␮m from the node. More nodally, it decreases rapidly and becomes close to nil a few micrometers from the node. Mitochondrion bags are rich in glycogen, many contain lipid droplets, and

some hold Marchi-positive myelinoid bodies and crystalline lamellar bodies (Reich granules; see Figs. 3–17 and 3–18 later). The Juxtaparanodal Segment The juxtaparanodal axon segment is a cast of the surrounding myelin sheath and characterized by an irregular, usually fluted, shape. The slender longitudinal axon ridges (or bulges) cease at the transition between the juxtaparanodal and the paranodal segment, that is, at the level where the myelin sheath turns to the axon in order to terminate along the paranodal part of the CON segment (see Fig. 3–9D). In many juxtaparanodes, particularly in those of large alpha motor axons in old animals,424 the more nodal parts of the axon ridges break up into thin, winding, irregular processes that are embedded in the adaxonal Schwann cell compartment (Fig. 3–13A; see also Fig. 3–9A). This complex of interwoven axonal and Schwann cell profiles is the axon–Schwann cell network (ASN).20,151,398 Both kinds of profiles contain DLBs and MVBs, some of which are acid phosphatase positive and should be classified as secondary lysosomes and residual bodies (see Fig. 3–13A, inset).149,155

54

Structure of the Peripheral Nervous System

Ax A

FIGURE 3–13 A, Electron micrograph of transverse section close to the transition between the juxtaparanodal and the paranodal segments; alpha motor nerve fiber, L7 ventral spinal root, adult cat. The axon crests inside the myelin crests have fragmented and form, together with the inner cytoplasmic Schwann cell compartment, typical axon–Schwann cell networks (asterisks). Arrows point at myelin crests covered by a substantial layer of Schwann cell cytoplasm; arrowheads, at mitochondrion bags. Bar: 2 ␮m. Inset, Detail from axon–Schwann cell network after incubation for demonstration of acid phosphatase. One of the larger axon profiles contains a multivesicular body filled with the reaction product (arrow). Bar: 0.5 ␮m. (Inset micrograph courtesy of Dr. K. Gatzinsky.) B, Electron micrograph of longitudinal section through a paranodal segment that illustrates the attachment of the myelin sheath on the paranodal axon (Ax); alpha motor nerve fiber, L7 ventral spinal root, adult cat. Asterisks identify two “ear-of-barley” arrangements in the terminal cytoplasmic spiral of the Schwann cell. Some of the crosscut cords in the spiral are indicated with asterisks. G ⫽ nodal gap; W ⫽ nodal gap wall. Bar: 0.2 ␮m.

The ASN has the ability to trap, sequester, and probably disintegrate effete axonal constituents and may reflect the role of the PNP region as a site where local lysosomal degradation of transported materials can take place.148,149,153,398 The network appears strikingly hypertrophic in a number of neuropathies36,398 and just the opposite in CNTF knockout mice, in which it develops early during the postnatal period and then withers away.152 The major changes in the organization of the axon, as seen when tracing along the

juxtaparanodal segment toward the node, are the gradual reduction of the transverse axon area, the increasing delicacy and final disintegration of the axon ridges, and the displacement of MTs from the periphery to the core of the axon. As a result, the MT concentration increases in proportion to the reduction of the transverse area of the axon.20,433 At its nodal end the juxtaparanodal axolemma contains voltage-gated (Kv) potassium channels of the delayed-rectifier type (␣ subunits Kv1.1 and Kv1.2 and the

Microscopic Anatomy of the Peripheral Nervous System

cytoplasmic ␤ subunit Kv␤2) in complex with contactinassociated protein-2 (Caspr-2), so far of unknown significance in the mature nerve.8,56,269,321,333,346 The Paranodal Segment This contains the termination—the nodal end—of the outer, abaxonal cytoplasmic Schwann cell compartment. The abaxonal Schwann cell plasma membrane contains Kv1.5 potassium channels.269 The abaxonal Schwann cell cytoplasm coats the inturning myelin sheath, faces the perinodal space, and extends into the nodal gap, where it forms the Schwann cell brush border. Here the Schwann cell compartment is virtually devoid of mitochondria but rich in moderately electron-dense round or tubelike “juxtanodal Schwann cell bodies” 50 to 200 nm in size and of unknown significance (see Fig. 3–11C).32 Their localization to a cytoplasmic domain between numerous mitochondria and a brush border, as well as their general ultrastructural appearance, recall the “dense apical tubules” that appear in proximal tubule cells after oil infusion.78 The paranodal segment is defined by the termination of the myelin sheath. Viewed in a median longitudinal section, each terminating myelin lamella splits into two leaflets that enclose a drop-shaped portion of Schwann cell cytoplasm: the terminal cytoplasmic pocket (see Figs. 3–11B, 3–12A, and 3–13B). The whole set of pockets thus displayed in a longitudinal section is the sagittal representation of a single, very thin and continuous cord of Schwann cell cytoplasm that encircles the paranodal axon cylinder in a complex helical manner. Each turn of the spiral corresponds to the termination of one myelin lamella. The total length of this terminal cytoplasmic spiral (TCS) is about 2500 ␮m in a large cat alpha motor fiber and provides a membranous mantle area of about 800 ␮m2 (see Table 3–1). The TCS represents the nodal margin (the lateral belt) of the hypothetically unrolled myelin sheath (see Fig. 3–16 later). It connects the inner, adaxonal Schwann cell compartment with the outer, abaxonal, one. In this respect the TCS is equivalent to the cytoplasmic spiral (Golgi-Rezzonico spiral) associated with an incisure of Schmidt-Lanterman, although it is less than half the length of the latter. The inherent stability of the TCS, as well as that of the Schmidt-Lanterman cytoplasmic spiral, depends on E-cadherin and MAG.111,430 The turn of the TCS that belongs to the innermost myelin lamella attaches to the paranodal axon segment farthest from the node. Subsequent lamellae then terminate in consecutive order up to the node. Only some 10% to 20% of the lamellae of a thick fiber attach directly to the axolemma. The remaining 80% to 90% coil on top of one another in sets of 10 to 25 turns. In a longitudinal section this gives the picture of “ear-of-barley”–like aggregations of terminal cytoplasmic pockets (see Figs. 3–12A and 3–13B) that penetrate the surrounding compact myelin for a distance of 0.5 to 1.5 ␮m. The specific staining properties of the TCS

55

explain the light microscopic appearance of the so-called spinous cuffs or spiny bracelets of Nageotte.190,470 The TCS is claimed to be rich in Na⫹ and Ca2⫹.102 The turns of the TCS that join the axolemma indent it and give it a serrated outline. They attach to it by gap junction–like membrane complexes that, when viewed in freeze-fracture preparations, form the so-called transverse bands188,248,345,459 and complex patterns of IMPs in both P and E faces. The various types of membrane connections noted between the paranodal axon segment and the TCS are as a whole referred to as the glia-axon-junction (GAJ) complex,204 the largest of the mammalian cell adhesion complexes.56,314 The GAJ complex includes a periaxonal space, 3 to 5 nm in width, that, at least in its most nodal aspect, opens into a node gap (see below). The turns of the TCS are joined together by a system of more or less continuous tight junctions.365 The GAJ complex and the TCS react strongly with antibodies against connexin 32 and GM1,275,400 both substances known to take part in cell adhesion and intercellular communication. In addition, the GAJ complex is endowed with Caspr-1 (paranodin) and contactin on its axonal side and neurofascin-155 on its Schwann cell side.42,98,267,430 The demonstration of Na⫹-K⫹-Cl⫺ cotransporter protein in the TCS, in nodal compartments of the Schwann cell, and in association with the Schmidt-Lanterman cytoplasmic spiral suggests involvement in K⫹ redistribution during and after impulse activity.4 Disordered and/or disrupted GAJ complexes correlate strongly with axonal dysfunction in various neuropathies.57,158,206,367,418,420 Those turns of the TCS that belong to the outermost two to five myelin lamellae do not attach to the axolemma but are separated from it by a space 20 to 50 nm in height. They form, in this way, the “ceiling” of a nodal gap recess. The most nodal turn of the TCS (i.e., the cytoplasmic pocket of the outermost myelin lamella) is high and voluminous and constitutes the nodal gap wall. The subaxolemmal part of the AR is particularly well developed in the paranodal axon segment. Whether it is involved in the impulse-generating mechanism is not known, although the GAJ complex and its axonal and Schwann cell environments apparently release Ca2⫹ during impulse activity.102,242,441,468,469

Node of Ranvier The nodes of Ranvier are the only sites along a myelinated nerve fiber that support the fast de- and repolarization process necessary for generation of action potentials.449 This ability is mainly a result of the high concentration of voltage-sensitive Na⫹ channels in the nodal axolemma. The ultrastructural study of nodes of Ranvier is hampered by their vulnerability to manipulation and chemical influences.176 Serious artifacts develop when the preparatory protocol includes dehydration, as it usually does.27,438,468,469 In particular, the nodal gaps and the paranodal myelin

56

Structure of the Peripheral Nervous System

become distorted as the two myelin sheaths that meet at the node shrink, are torn apart, and split in relation to their TCSs. As a result, the nodal gap appears as a welldefined and relatively large and clear opening in the myelin sheath, a picture not at all consistent with the in vivo or pre-dehydration appearance, where nodal gaps are hard to define.27,154 The thicker the myelin sheath, the more serious are the preparatory artifacts. These difficulties probably explain why there are few systematic studies dealing with quantitative aspects of nodal ultrastructure in large fibers.20,21,274,318,327,351,438 The nodes of Ranvier can, just like other parts of a myelinated fiber, be described as consisting of two concentrically arranged zones: an outer Schwann cell compartment (here without myelin) and an inner axonal compartment. Cell adhesion molecules (CAMs) both of the neuron-glia and of the neuronal type coexist specifically in the node of Ranvier, where they appear in the Schwann cell compartment as well as in the axon. This localization of both CAMs in the adult myelinated nerve fiber emphasizes the importance of nodal structural integrity.220,340 The endoneurial territory outside a node of Ranvier and at the meeting point of two adjacent internodes is denoted the “perinodal space.” The Schwann Cell Compartment This consists of three regions separated from the perinodal space by the basal lamina: (1) the outer demarcation, (2) the nodal gap, and (3) the nodal gap walls (see Figs. 3–9F and 3–12A). The outer demarcation is defined by more or less overlapping extensions of the outer cytoplasmic compartments of the two adjoining internodes. The extensions are referred to as nodal collars. The nodal collars send a conventional brush border of radially arranged microvilli into the nodal gap. The brush border, as a rule, is particularly well developed in those sectors of the outer demarcation where the nodal collars are direct continuations of a mitochondrion bag. In some sectors of the outer demarcation, the nodal collars of the paranodal segments overlap and join, forming five-layered membrane complexes reminiscent of tight junctions. At some points the collars are joined by clumps of lamellar material.33 In other sectors of the same demarcation region, the nodal collars are minimal and lie 0.1 to 0.2 ␮m apart. Here the nodal gap is separated from the perinodal space only by the Schwann cell basal lamina. Solitary tufts of Schwann cell microvilli are common outside the nodal gap facing the perinodal space. Similar tufts are occasionally observed elsewhere on the outer cytoplasmic Schwann cell compartment (see Pannese et al.310). The nodal gap, when viewed in a median longitudinal section, is shaped like an isosceles trapezoid (see Fig. 3–12A), the longer and shorter sides of which correspond to the nodal axolemma and the plane of the outer demarcation, respectively. The slanting sides of the trapezoid are formed

by the nodal gap walls.33,351 The nodal gap contains the Schwann cell brush border and the gap matrix substance. The matrix substance, containing glycosaminoglycans, occupies the extracellular space of the gap. Its role as a cation exchanger and/or buffer has been much discussed but remains elusive.230,233 Its strong and capricious affinity for heavy cations such as Ag2⫹, Cu2⫹, Fe2⫹, and Pb2⫹ is noteworthy in connection with electron histochemistry based on the precipitation of metal salts.5,183,222,229,233,234,451,471–473 The metallophilia of the nodal gap substance lies behind the appearance of both the so-called nodal cementing disc and the shorter bar of Ranvier’s cross.186 The latter develops when the CON segment together with the nodal gap display metallophilia, such as after incubation in a solution of AgNO3. The nodal gap contains NG2, a chondroitin sulfate proteoglycan, presumably delivered into the gap from the outside by “perineural fibroblasts.”255 Speculation has it that NG2 might restrain an otherwise sprouting-happy nodal axon. The microvilli of the brush border end less than 5 nm from the axolemma. In a large fiber (D ⫽ ~15 ␮m), the height of the nodal gap varies in different sectors of the node from 0.1 ␮m to as much as 2 ␮m. There are about 800 to 1000 microvilli, each 70 to 80 nm in diameter and on the average 1 ␮m long, in a nodal gap (see Table 3–1). The seven or eight microfilaments (F-actin) in a microvillus and the lack of a terminal web in the nodal collars suggest structural kinship with the brush border of the proximal tubule cell of the kidney (see Kenny and Booth215). The actin filaments connect to the cell membrane of a microvillus with three ERM proteins: ezrin, radixin, and moesin.265,363 Freeze-fractured microvilli show P and E faces in which 45% of the intramembranous particles are about 10 nm, a size considered to correspond to that of voltage-sensitive Na⫹ channels.224,344,450 The nodal Schwann cell microvilli are furnished with inwardrectifying K⫹ channels of the IRK family269 and possibly are able to perform “K⫹-buffering” (see Orkand et al.307). The nodal gap forms low recesses that extend for a distance of 0.1 to 0.4 ␮m both proximally and distally between the axolemma and the most nodal turns of the TCS of the paranodal segment. Calculations based on morphometric data show that, in a large cat alpha fiber, the Schwann cell faces the nodal gap with a cell membrane area of about 250 ␮m2. Approximately 75%, 20%, and 5% of this area are provided by the brush border, the nodal gap walls, and the ceiling of the gap recess, respectively (see Table 3–1). The use of watersoluble embedding media after glutaraldehyde–osmium tetroxide fixation or the use of glutaraldehyde–tannic acid fixation followed by fixation in osmium tetroxide and potassium ferricyanide produces more or less closed nodal gaps with a maximal close packing of the microvilli (see Fig. 3–12A).27,32,33 The Schwann cell plasma membrane of the villi and related to the nodal gap (and to the paranodal axon segment) may contain, in addition to acetylcholine receptors, Na⫹ and fast and slow K⫹ channels.72,74,449,452 It has been suggested that some of the Na⫹ channels integrated

Microscopic Anatomy of the Peripheral Nervous System

57

in the nodal axolemma may have been synthesized in the Schwann cell and delivered to the nodal axolemma via the node gap microvilli169 (however, see Brophy56). The nodal gap walls are formed by the comparatively high and voluminous cytoplasmic pocket of the outermost myelin lamella of the two meeting internodes (see Figs. 3–12A and 3–13B). They are in direct communication with the paranodal outer Schwann cell compartment and contain juxtanodal Schwann cell bodies and occasional mitochondria. The Axonal Compartment This consists of the nodal axon segment, which corresponds to the midregion of the CON segment (see Figs. 3–9F and 3–11B). The segment is about 1 to 1.5 ␮m long regardless of fiber size.351 It extends between the most nodally and closely attached turn of the TCS of each of the two adjoining paranodes. The segment is usually barrel shaped with a diameter a few percent larger than that of the adjacent paranodal segments. The size of the mantle area of the nodal axon segment increases linearly with increasing axon size. In cat spinal roots, the nodal membrane area varies from about 4 ␮m2 in the smallest fibers to about 30 ␮m2 in the largest ones.351 The nodal axolemma (review by Salzer357) is characterized by a high content of P-face particles (about 1500/␮m2)345 (see Fig. 3–12) and an inside coating of electron-dense material. The inside coating, or the nodal axoplasmic cortex, is about 30 nm thick after conventional electron microscopic preparation (see Fig. 3–9F) and up to 100 nm after application of unconventional preparative methods (see Figs. 3–12A and B). The undercoating contains actin,478 ␤IV spectrin,19 and the 480- and 270-kDa isoforms of ankyrin G, which probably bind directly to the voltagegated Na⫹ channels459 and anchor them to the cytoskeleton by way of neurofascin, NrCAM, and L1.18,80,204,254,356 After staining with ferric ions and ferrocyanide, the nodal axolemma develops a thick, electron-dense precipitate, a reaction taken to indicate the presence of Na⫹ ion channels and their anchoring support.325,451 The thickness of the nodal axoplasmic cortex as shown in Figure 3–12B corresponds well with the assumed size of the ankyrin-neurofascin complex.204 In very small fibers (D ⫽ 1 to 2 ␮m), in which the nodal gap contains just a few microvilli, the inside coating is patchy and restricted only to sites related to the tip of a microvillus. In some nodes the axolemma projects spine- or crestlike outgrowths.33,438 The crests run across the longitudinal fiber axis. Both kinds of projections contain AR profiles. Coated pits are common in the axolemma at the bases of spines and crests and in relation to the nodal recesses. The presence of coated invaginations also in the nodal collars may suggest an interplay between axon and Schwann cell mediated via the nodal gap.33,116,238 The Schwann cell:axon membrane ratio of the node of Ranvier is plotted against axon diameter in Figure 3–14, which illustrates that the Schwann cell:axon membrane

FIGURE 3–14 Ratio between nodal Schwann cell membrane area and nodal axon membrane area in the cat and plotted against axon diameter as estimated in internodal main regions. Open circles ⫽ L7 ventral root axons; filled circles ⫽ L7 dorsal root axons. (From Rydmark, M., and Berthold, C. H.: Electron microscopic serial section analysis of nodes of Ranvier in lumbar spinal roots of the cat: a morphometric study of nodal compartments in fibres of different sizes. J. Neurocytol. 12:537, 1983, with permission.)

ratio reaches its highest values of 10 to 15 in the d interval of 5 to 10 ␮m. Provided that a high ratio reflects the ability of the Schwann cell to control the ionic milieu outside the axolemma in a way favorable for impulse conduction, nerve fibers with the highest conduction stamina would appear in the 5- to 10-␮m interval. This is the size of large gamma axons, small to medium-sized alpha axons, and type II afferents, which are all known to fire tonically (i.e., have high conduction stamina). This and the suggested accumulation of K⫹ channels in the paranodal axolemma recall the role of the Schwann cells as potassium clearing units outside squid giant axons.1,58

The Constricted Axon Segment The length of the CON segment (Fig. 3–15A; see also Fig. 3–8) is about 6 to 10 ␮m and increases little with fiber size.351 The axolemma of the CON segment is formed by a median nodal field, flanked symmetrically by a proximal and a distal paranodal field. Voltage-clamp experiments on normal mammalian myelinated fibers have shown that fast-depolarizing Na⫹ currents are generated at the nodes during impulse conduction, whereas fast-repolarizing K⫹ currents, well known to be present at active amphibian nodes, are lacking.48,49,73,218 A fast K⫹ current appears, however, at the active mammalian node if the myelin is loosened

58

Structure of the Peripheral Nervous System

P

I N

D

E

2 μm

Microscopic Anatomy of the Peripheral Nervous System

from the paranodal axon segments.73 These observations, in combination with results from freeze-fracturing and immunocytochemistry, have given the picture of a CON segment in which the nodal axolemma is rich in Na⫹ channels, slow tetraethylammonium-blockable K⫹ channels, and Na⫹,K⫹-ATPase5,6,36,103,370,378,410,466 but devoid of fast 4-aminopyridine–blockable K⫹ channels.14,37,75,96,219,341,343 The paranodal axolemma, in contrast, seems to be rich in fast K⫹ channels but poor in Na⫹ channels and slow K⫹ channels, and may contain chloride channels.407 In addition to its numerous large P-face particles and thick axoplasmic cortex, the nodal field of the axolemma of the CON segment is characterized by ␣-bungarotoxin binding sites140 and numerous filipinsterol complexes.40 The axolemma of the paranodal fields and the cytoplasm of the adjoining TCS contain Ca2⫹-activated ATPase.258 As a whole, the axolemma of the CON segment is 5'-nucleotidase positive, becomes impregnated when treated with bismuth iodide, and binds ruthenium red and lectins.88–92 The magnitude of the Schwann cell membrane area provided by the TCS that faces the extracellular space outside the CON segment (the periaxonal and the nodal gap spaces) is about 10 times that of the axolemma of the CON segment. Elsewhere along the fiber, this Schwann cell:axon membrane ratio is 1:1.33 The Schwann cell membrane domains that directly face the axolemma of the CON segment (i.e., some or all turns of the TCS and the ends of nodal microvilli) appear to be anchored to the axonal cytoskeleton by filamentous transmembrane linkers.204

59

The appearance of the axoplasm in the CON segments and their immediate proximal and distal juxtaparanodal extensions is highly variable with regard to both the amount and the distribution of membranous organelles and granular material.20,21,28,29,328 The various patterns run the gamut from an appearance that, except for the high concentration of MTs, barely deviates from that noted in the main internodal region, to that of an axoplasm crammed with organelles and granular material (see Figs. 3–11B and 3–15). As a rule, DLBs and MVBs (i.e., retrogradely transported organelles) occupy the axoplasm in the distal half of the CON segment and the adjacent part of the juxtaparanodal axon. Organelles transported anterogradely (vesiculotubular profiles) occupy the proximal half of the CON segment and adjacent parts of the juxtaparanodal axon. They are also numerous in the nodal segment, and some appear in the distal paranode. Such a proximodistal segregation of retrogradely and anterogradely transported elements is easily understood in terms of a local reduction of the axonal stream at the node. This is similar to the pattern noted at experimentally induced blocks of axoplasmic transport.79,104,390,392,432 The obvious variations noted in the segregation pattern, the presence of lysosomes, and the low occurrence of DLBs and MVBs outside the PNP regions (see Table 3–1)26,29,149,150 are, however, difficult to understand as passive rheologic phenomena. One possible explanation could be that these variations indicate differences in the functional/metabolic state of a neuron in general and of a PNP region in particular. The various patterns may reflect different operative

FIGURE 3–15 Electron micrographs of longitudinally sectioned nodes of Ranvier; alpha motor nerve fibers, L7 ventral spinal roots (A–D) and medial gastrocnemius nerve (E), adult cat. A, Two consecutive digitized electron micrographs are aligned and fused with an arithmetic minimum filter. The strandlike arrangements of amassed vesiculotubular organelles in the axoplasm of the proximal internodal end region (thin arrows) are particularly distinct. CON ⫽ constricted part of the nodal region; M ⫽ myelin sheath; N ⫽ nodal region. Thick arrow points towards the cell body. Bar: 2 ␮m. (From Pascher, R., Berthold, C. H., and Rydmark, M.: Computer-assisted simulation of high-voltage electron microscopy using serial images recorded by conventional transmission electron microscopy. J. Electron Microsc. [Tokyo] 51:113, 2002, with permission.) B, Detail from a proximal paranodal segment. Arrows point to axoplasmic reticulum profiles. VT ⫽ vesiculotubular profiles; D ⫽ distal; P ⫽ proximal. Bar: 1 ␮m. C, Detail from a proximal paranodal segment. The axoplasm is dominated by longitudinal strands of vesicotubular profiles (asterisks). D ⫽ distal; P ⫽ proximal. Bar: 1 ␮m. D, Detail from the transition between the distal paranodal and juxtaparanodal segments. The axoplasm is rich in microtubular profiles (MT) with associated vesicles and mitochondria (M). Asterisk indicates a strand of vesicotubular profiles. D ⫽ distal; P ⫽ proximal. Bar: 1 ␮m. E, Stereo view of a node of Ranvier from a HRP-transporting medial gastrocnemius nerve. The stereo pair is generated from a digitized image series of eight consecutive sections by fusing with an arithmetic minimum filter. The axoplasmic HRP activity appears as black rounded bodies of different sizes. The axoplasm proximal to the constricted segment contains very few HRP-positive bodies. The stereo view illustrates the continuous strings of HRP-positive bodies that extend from the distal internodal end region (D), where they are comparatively large, through the constricted axon segment at the node (N), where their size decreases and they become incorporated in the vesiculotubular strands. From here, small HRP-positive bodies continue to the proximal level (P) of the constricted axon segment, where they seem to aggregate into a thin, transverse, disclike configuration (white vertical bar). Uncontrasted sample. The viewing angle is 7 degrees. Bar: 2 ␮m. (From Pascher, R., Berthold, C. H., and Rydmark, M.: Computer-assisted simulation of high-voltage electron microscopy using serial images recorded by conventional transmission electron microscopy. J. Electron Microsc. [Tokyo] 51:113, 2002, with permission.)

60

Structure of the Peripheral Nervous System

steps in a local interaction between, on the one hand, the axoplasm of a CON segment and its immediate juxtaparanodal extensions, including the ASN, and, on the other hand, various transported elements. There are observations that indicate that the axoplasm of the CON segment may interact actively both with material transported retrogradely from the terminal region, such as intramuscularly injected horseradish peroxidase (HRP) (see Fig. 3–15E),26,29,149 and with materials transported anterogradely from the soma, such as newly synthesized proteins.7,208,323,478 Our studies of alpha motor neuron PNP regions that partake in retrograde transport of intramuscularly injected HRP have shown organelle patterns similar to those of unexposed animals but with a specific temporal order of appearance after the injection.29,313 This segregation and ordered trapping of HRP-positive organelles more or less ceases as the alpha axon enters the CNS (see Fabricius et al.110). Evidently CON segments of peripheral myelinated nerve fibers are sites where elements of the retrograde endocytotic stream are forced into close, probably interactive, contact with elements that belong to the anterograde exocytotic (lysosomal) stream.198,203,440 Extreme segregation patterns appear in the PNP regions of the distal stump in the first days after a crush injury; the axoplasm distal to nodes of Ranvier accumulates large numbers of mitochondrion-like organelles and DLBs.241,453 The MTs of highly segregated CON segments form coherent bundles of up to 20 members. 21,282 The bundles usually project a few micrometers proximally and distally into the juxtaparanodal segments before they become separated into smaller bundles or individual MTs. The various organelles are organized outside and along the MT bundles, to which they seem to cling in a grapevine fashion (see Fig. 3–15A–D). In more relatively empty-looking and less segregated CON segments, the bundling of MTs is less conspicuous and the MTs are more dispersed. No doubt the distribution of MTs in the axonal cross section will, in view of their fundamental role as the rails and promotors of axoplasmic transport, influence the transport capacity of the CON segment with regard to both the amount and the size of transported bodies.30,328 A hypothetical mechanism that was able to influence the distribution of MTs and operated in the PNP region would give the CON segment the combined properties of a sieve and a throttle valve. The claims that the Schwann cells may influence the axon, and in particular the NF and the MT of the PNP regions, are indeed noteworthy in connection with the axoplasmic transport through CON segments.82–84,288 The possibility exists that a highly segregated CON segment could be such a hindrance to the progression of anterogradely and retrogradely transported organelles that turnaround phenomena391 are triggered in the orifices of the CON segment. As a result, the organelles would begin to shuttle back and forth in the respective internode.29,30

The Node–Paranodal Apparatus Landon and Williams, who gave the first comprehensive ultrastructural description of the PNP region in mammalian nerve fibers more than just a few micrometers thick,231,463 used the term paranodal apparatus, referring to the Schwann cell mitochondrion bags, the nodal gap microvilli, and the nodal gap substance. This emphasized the significance of these parts of the Schwann cell in nodal function, which at that time assumed that a fast K⫹ current repolarized the nodal axolemma during the action potential but did not account for the interference of the CON segment with axoplasmic transport. In view of current knowledge and ideas, it seems reasonable to extend the concept of a paranodal apparatus to include the TCS of the Schwann cell, the whole CON segment, and the immediately adjoining parts of the proximal and distal juxtaparanodal segments and their associated ASNs. With this conceptual broadening, the term node–paranodal apparatus could be more informative. Many interesting speculations have been presented with regard to the functional significance of the node– paranodal apparatus and the ways its various components may cooperate102,279,426,441,459 (see also Landon228,231). Hard facts have, however, been remarkably meager, which is surprising in view of the many fundamental functions expressed by the PNP region and its central role in a host of peripheral neuropathies. This is, however, no longer so. During the last two decades an almost overwhelming amount of new molecular and genetic data focusing on the PNP region have been presented.322,414

THE SCHWANN CELL Schwann cells are the satellite cells of peripheral nerves. The majority originate in the neural crest,181 although others appear later as they migrate along the neural tube. The suggestion that Schwann cells in motor nerves arise in the ventral portion of the neural tube226 was supported by experiments using xenoplastic grafts334 and also by autoradiographic studies.457 The contributions of the neural tube and neural crest have been discussed by several authors; other possibilities seem less likely.9,71,458 In the past they have been thought to be merely supporting cells for the production of the myelin sheath where appropriate, but recently their effects on the axon have become a matter of great interest. During embryonic development of early Schwann cells, two growth factors are particularly important: endothelin and ␤-neuroregulin-1. They seem, however, to be required for survival and some aspects of migration rather than establishment of the peripheral nervous system (PNS) glial cell lineage, and it is still unclear what initial signal is required. The transcription factor Sox10 may be important in early development. These

Microscopic Anatomy of the Peripheral Nervous System

factors have been reviewed by Jessen and Mirsky.210 The development of transgenic mouse models to investigate the role of individual proteins in development of Schwann cells and formation of the myelin sheath, axon, cytoplasm, and basal lamina have emphasized the vital role the Schwann cell plays in nerve function and that, without a healthy, fully functional Schwann cell, the axon will degenerate. As already mentioned, Schwann cells multiply dramatically during development. Once they have established a one-to-one relationship with an axon and myelination has started, no further mitosis occurs unless a pathologic process occurs. If contact with a myelin sheath is lost, the Schwann cell is free to divide again. Although there superficially appear to be two types of Schwann cells, those associated with myelinated and with unmyelinated axons, numerous studies have shown that this is due to axonal signals and that they can change from one to another on the receipt of different signals. Recent studies have defined the proteins expressed on the surface of Schwann cells. All Schwann cells from embryonic day 16 in rat nerve express S100 and O4. Myelinating Schwann cells express Krox20, periaxin, the peripheral myelin proteins P zero (P0), peripheral myelin protein-22 (PMP22), P1, myelin basic protein, P2, MAG, connexin 32, proteolipid protein, and galactosylceramide.

FIGURE 3–16 Diagrammatic representation of an unrolled Schwann cell. The white areas indicate the presence of cytoplasm. The lightly stippled areas at the outer and inner portions of the cell represent semicompacted myelin and the heavily stippled areas, compact myelin. Cytoplasmic channels form a continuous network in both compact and semicompact myelin domains. Cytoplasm is present along the whole margin of the cell, the outer and inner “belts” constituting the marginal rim of the abaxonal and adaxonal Schwann cell cytoplasm, respectively. The lateral belt is the terminal cytoplasmic rim at the node of Ranvier, the outer portion bearing the nodal microvilli. One complete and one incomplete Schmidt-Lanterman incisure have been included together with two longitudinal incisures. (From Mugnaini, E., Osen, K. K., Schnapp, B., et al.: Distribution of Schwann cell cytoplasm and plasmalemmal vesicles [caveolae] in peripheral myelin sheaths: an electron microscope study with thin sections and freeze-fracturing. J. Neurocytol. 6:647, 1977, with permission.)

61

Schwann cells associated with unmyelinated axons express a different set of proteins: p75, neural CAM, L1, glial fibrillary acidic protein, growth-associated protein-43, A5E3, and Ran-2. (For a detailed discussion of the signals that determine Schwann cell identity, see Jessen and Mirsky.210) Abnormalities in the genes for some of these proteins have been related to some human genetic diseases; for example, abnormalities in PMP22 and P0 are related to various subtypes of Charcot-Marie-Tooth (CMT) disease type 1 and mutations in connexin 32 are related to X-linked CMT disease (see Chapter 76). Myelin is formed by the compaction of layers of the Schwann cell surface membrane. A diagram of the myelinating cell unrolled shows a trapezoid sheet (Fig. 3–16). The adaxonal Schwann cells form the innermost portion of cytoplasm adjacent to the axon while the abaxonal region is that outside the myelin sheath. There are cytoplasmic channels (incisures of Schmidt-Lanterman) connecting the abaxonal and adaxonal regions, thus permitting the movement of metabolites from one to the other. The nucleus is midway between nodes on the outer surface of the myelin sheath. Both surfaces of the cell are connected to the myelin sheath by extensions of the cell membrane, the inner and outer mesaxons (Fig. 3–17). These are occluded by tight junctions. There is very little cytoplasm and few organelles except in the nuclear and nodal regions.

62

Structure of the Peripheral Nervous System

responsiveness to antigen, cytokines, and T cells suggest that this may be significant in the pathologic development of inflammatory demyelinating neuropathies245 and also possibly in hereditary demyelinating neuropathy.162

THE MYELIN SHEATH Compact myelin is formed by the spiral wrapping of Schwann cell plasma membranes. This forms an alternating structure of light and dark lines. The light band is also divided by a faint line known as the less-dense line. The dark line (the major dense line) is formed by the apposition of the intracytoplasmic surfaces of the Schwann cell and the less-dense line from the juxtaposition of the extracellular surfaces (Fig. 3–19). X-ray crystallography showed that the periodicity of myelin in fresh, unfixed nerve is 18 nm but fixed material shows a considerable reduction to 12 to 15 nm depending on the processing involved. Compaction may also be affected by osmotic effects and x-irradiation and particularly by abnormal serum proteins. Changes in compaction are most commonly seen in the neuropathy related to immunoglobulin M paraproteinemia. The result is an increase in periodicity to 2.5 times normal because of the failure of the extracellular spaces to connect tightly together. In this situation the less-dense line is seen as two components each the correct distance from the major dense line but widely separated from each other (Fig. 3–20). For a discussion of the detailed structure of the myelin sheath, the reader is referred to Chapter 19. FIGURE 3–17 A, An electron micrograph of a transverse section through a small myelinated nerve fiber from a normal adult human sural nerve. Ax ⫽ axon; My ⫽ myelin; R ⫽ Reich granule; Sn ⫽ Schwann cell nucleus. Bar: 0.5 ␮m. B, Detail from A showing inner and outer mesaxons (arrows). Bar: 0.2 ␮m.

Unmyelinated Nerve Fibers R. H. M. King HISTORICAL BACKGROUND

Whereas there may be lipofuscin granules in the Schwann cells of Remak fibers, these are not found where the Schwann cells are associated with myelinated fibers. However, the age-related acid phosphatase–positive granules of myelinating Schwann cells are the ␲ granules of Reich. These are lamellar structures about 1 ␮m in length, with metachromatic properties resulting from the presence of sulfatide or phosphatide (Fig. 3–18), and may have a lysosomal function. They are not found in normal Remak fibers but may persist in the cytoplasm after myelinated fiber degeneration even when the Schwann cell has become associated with unmyelinated axons. Schwann cells can express major histocompatibility complex class II molecules and hence can act as antigenprocessing cells. In vitro observations on Schwann cell

Very fine fibrous structures were described by Remak in 1838 when he teased autonomic nerve bundles apart in water. He distinguished myelinated nerve fibers (tubuli primitivi) from other delicate fibers (fibrae organicae). Although he identified axons within the tubuli primitivi, there is nothing to suggest that he realized that the fibrae organicae also contained axons.338 It is a convenient shorthand to refer to the bundles of Schwann cells containing unmyelinated axons as Remak fibers. In 1839, Schwann identified the nucleated corpuscles on the fibers as cells.369 It was not for another 60 years, however, that Tuckett, using an osmic acid technique, identified both sheath cells and cores (i.e., Schwann cells and axons). He also described the changes produced in these fibers by degeneration.436

Microscopic Anatomy of the Peripheral Nervous System

63

FIGURE 3–18 Electron micrograph of Reich granules from control human sural nerve. Bar: 0.5 ␮m. Inset illustrates the lamellar structure at higher magnification.

FIGURE 3–19 Electron micrograph showing detail of the myelin sheath structure in a control human sural nerve. Arrows indicate the major dense lines and arrowheads the intermediate lines. Bar: 0.5 ␮m.

FIGURE 3–20 Electron micrograph showing detail of an abnormal myelin sheath. Part of the sheath has normal compact myelin (CM) where major dense lines (white arrows) are separated by a less-dense intermediate line that appears double in places. In the adjacent region of the myelin, the intermediate line has failed to compact and the two components are separated by a wide gap (double-headed arrow). Bar: 0.05 ␮m.

64

Structure of the Peripheral Nervous System

Unmyelinated axons vary in diameter from 0.2 to 3 ␮m, and so are difficult to see by light microscopy, especially in paraffin sections whose thickness is greater than the axon diameter. It was only with the advent of electron microscopy that it was confirmed that there was no myelin at all on these small axons in Remak fibers. Phase-contrast microscopy had previously been interpreted to show a very thin myelin sheath, but electron microscopy revealed that some axons did not possess any myelin at all.185 Electron microscopy later confirmed that the axons were enfolded along their length by a series of separate Schwann cells; earlier workers such as Nageotte had thought that this was a syncytium as opposed to separate cells.281 He described the fibers as forming interlacing trabeculae that divided and reunited, with each containing several axons (see Relationship of Unmyelinated Axons with Schwann Cells later). There is very little overlap in size between myelinated and unmyelinated fibers.297 Unmyelinated axons range in size from 0.1 to 3 ␮m. It has been shown that the presence of a myelin sheath does not increase conduction speed for very small fibers.308 In addition, myelin is complex and easily damaged, so an unmyelinated fiber should be more robust and less sensitive to metabolic insults.

DEVELOPMENT Schwann cells develop from the neural crest ectoderm and already show a basal lamina by 9 weeks’ gestation in humans. There are no reports of naked axons being identified before Schwann cells can be seen, as may be the case in regeneration, although before Schwann cells have a basal lamina, they are difficult to distinguish from axons or fibroblast processes. The axons lie in large bundles each enfolded by a single Schwann cell process (Fig. 3–21). In a study of a 14-week fetal nerve, these bundles were seen to contain up to 90 small axons most of which are in contact only with each other, with only the outer ones contacting the surrounding Schwann cell process.146 The apparently greater density of Schwann cell nuclei in transverse sections of fetal compared with adult nerve is probably partly due to the cells being shorter (rounder) than in adult nerves. The Schwann cells divide in response to axonal signals and send cytoplasmic processes out to encircle individual axons on the edge of the bundle to form promyelin fibers with a 1:1 ratio of Schwann cell to axon. This is a necessary precursor to myelination. Myelination commences at different times in different parts of the PNS but seems to start at about 18 weeks’ gestation in the sural nerve in humans.296 By 21 weeks the myelinated fiber density is approximately 5000/mm2, with on average nine lamellae per fiber.376 It has been estimated that Schwann cells only need to divide three or four times during development to reduce the axon:Schwann cell ratio from

FIGURE 3–21 Electron micrograph of nerve from 24-week-old fetus showing a bundle of numerous small unmyelinated axons (asterisks) embedded in one Schwann cell. Schwann cell processes (arrows) only contact the outermost axons in the bundle. Two single, larger unmyelinated axons (A) have formed a one-to-one relationship with a single Schwann cell process (promyelin fiber). Bar: 1 ␮m.

1:100 to 1:12 to the 1:6 seen at birth.146 Further development after birth reduces this still further until eventually only one or two fibers remain small and unmyelinated. In general this progression from small fetal axons to larger mature ones depends on the axon diameter; as they become larger, they become separated and then myelinated. Further axonal growth occurs after myelination.465 The axonal signal leading to growth of some axons and not others is still unclear. Growth factors are clearly involved; high doses of glial cell line–derived neurotrophic factor (GDNF) will change the phenotype of an axon from unmyelinated to myelinated.196 GDNF is also associated with Schwann cell proliferation and increase in size of unmyelinated axons together with a reduction in the number of axons per Remak fiber. The mechanism could be a reaction of the Schwann cells to GDNF or a reaction of a subpopulation of GNDF-dependent, c-Ret–expressing unmyelinated axons to GDNF. In vitro studies show that both nerve growth factor and GDNF will induce myelination.196,278 The axons that are destined to remain unmyelinated become rearranged to lie in closer contact with a Schwann cell process but separated from each other,

Microscopic Anatomy of the Peripheral Nervous System

forming a Remak fiber. The density of unmyelinated axons falls from 235,000/mm2 at birth to 60,000 to 80,000/mm2 at 3 to 10 years,306 and further to 22,000 to 34,000/mm2 by 15 to 20 years.297 Slightly different counts have been obtained by other workers; Jacobs and Love found a density of 30,000 to 40,000/mm2 by 10 years of age.207 The differences may be due to the difficulty of distinguishing small unmyelinated axons from small, circular Schwann cell processes. There is also considerable variation in density in different parts of the fascicle, rendering sampling procedures liable to produce errors. Unmyelinated axons differ considerably from myelinated ones in reaching their final maximum diameter as early as 23 weeks’ gestation.297,376 The number of axons in one Remak fiber varies between species, being up to 10 in rodents and more in cats but usually only 1 or 2 in adult human nerves (Fig. 3–22).214 In neonates there are numerous axons per Schwann cell unit. By about 6 months of age the numbers have reduced to two to six in each Schwann cell, with each axon having its own mesaxon; further maturation occurs later.207 In vitro studies have shown that the presence of fibroblasts is necessary for the maturation of Remak fibers, and the

FIGURE 3–22 Electron micrograph of normal human sural nerve showing numerous unmyelinated axons (A) embedded in denser Schwann cell processes. Two Schwann cell nuclei are sectioned (SN). There are occasional collagen pockets formed by Schwann cells surrounding bundles of collagen fibrils as if they were axons (arrows). Only rare axons share a Schwann cell process with another axon. Bar: 1 ␮m.

65

addition of individual basal laminal components separately causes extensive multiplication of Schwann cell processes.292 Human nerves are not fully mature until after the end of the first decade of life. The first detailed ultrastructural study of unmyelinated fibers in the human sural nerve was by Ochoa and Mair in 1969.297 They found no evidence of the syncytium postulated by Nageotte281 and still supported by Gasser in 1952, despite his electron microscope studies.147

GENERAL ORGANIZATION The components of the unmyelinated axon are the same as in axons with a myelin sheath. The proportions of NFs and MTs, however, differ somewhat. There tend to be fewer NFs and the axons are less densely populated. Conversely, the surrounding Schwann cell processes may have a greater density of intermediate filaments than the density of NFs in the axon. Although there is often a mesaxon as in myelinated fibers, this is not necessarily the case, and unmyelinated axons may merely lie in grooves in the Schwann cell (see Fig. 3–22). They are always covered by Schwann cell basal lamina even when not embedded in cytoplasm. Where adjacent Schwann cell processes touch, there are often dense patches forming structures suggesting junctional complexes.100 The axolemma in well-fixed material is often distinguishable from Schwann cell plasma membrane by its greater electron density and slight irregularity. The proportion of myelinated to unmyelinated fibers varies in different nerves. In the sural nerve in humans, the most commonly studied peripheral nerve, it is approximately 1:4.297 The spatial arrangement within the fascicle is not random but rather patchy. Remak fibers are also often associated with small myelinated axons. Although it has been suggested that this is developmental, it seems possible that these may be sympathetic bundles. The arrangement of unmyelinated axons in a normal Remak fiber is easily distinguishable from a bundle of unmyelinated sprouts originating from a damaged myelinated fiber during regeneration after axonal damage. In the latter situation the original basal laminal ensheathment of the degenerated parent myelinated fiber may be still present, but even if it is not, the group of fibers forms a round collection of axons and Schwann cell processes that still conforms to the original circular profile. The cytoplasm of regenerating Schwann cell processes is also usually less electron dense than that of stable Schwann cells. These collections of Schwann cell processes within a parent basal lamina are often referred to as bands of Büngner after the original description.63 Axonal sprouting is also suggested by the presence of more than three unmyelinated axons in one Schwann cell process.

66

Structure of the Peripheral Nervous System

Relationship of Unmyelinated Axons with Schwann Cells Because unmyelinated axons are not insulated from each other, there is the possibility that an impulse in one could affect a neighboring axon. Laborious studies showing that axons constantly switch between Schwann cells mean that unmyelinated axons are only in close apposition for a very short distance before they move off separately into different Schwann cells. This constant switching between cells means that “cross talk” between axons is very unlikely. Reconstruction of rat nerves from serial sections shows that the Schwann cells are between 200 and 500 ␮m in length and consist of several thin processes (Fig. 3–23).

Spatial Relationships between Axons, Schwann Cells, and Collagen Bundles of collagen fibrils may be enveloped by Schwann cells as if they were axons.297 These are known as “collagen pockets.” It has been shown that they are quite short in longitudinal extent, so they probably do not represent replacement of axons by collagen. However, they are definitely more common in pathologic situations. They might be “hooks” that attach Schwann cells to collagen bundles to provide additional mechanical stability. Although they were specifically looked for, they were not identified in a study of human fetal nerve,146 suggesting that they are a later development. When unmyelinated axons degenerate, the surrounding Schwann cell processes collapse to form a stack of flattened Schwann cell processes (Fig. 3–24). These are readily distinguishable from the rounded profiles that constitute the bands of Büngner formed by myelinated fiber degeneration.423 Sometimes, however, when there is extensive Schwann cell reduplication (e.g., in the hereditary motor and sensory neuropathies), the excess Schwann cell processes unassociated with axons may form similar flat sheets usually near or in classic onion bulbs. The presence of Reich granules in Schwann cells associated with unmyelinated fibers identifies the cells as having originally been associated with a myelinated axon. These axons may therefore be small unmyelinating sprouts arising from a regenerating myelinated fiber, but it is also possible that Schwann cells left without an axon may become associated with an unmyelinated axon rather than the myelinated one that was present originally.217

MORPHOMETRY In well-fixed material, unmyelinated axons may be measured and counted by electron microscopy. At least 10% of the fascicular area should be photographed to give

FIGURE 3–23 Three-dimensional reconstruction of interrelating Schwann cell units from a normal rat cervical sympathetic trunk based upon electron micrographs of serial transverse sections (D to U) cut at 5-␮m intervals. Four of the electron micrographs are reproduced. The arrangement of axons within Schwann cell units varies from level to level because of separation and rejoining of adjacent units. Schwann cell units are present at levels D, O, and U. (Courtesy of A. J. Aguayo and G. M. Bray.)

a good sample, but because of the uneven distribution within the fascicle, this should ideally be chosen by scattering measuring fields across the fascicle. However, in practice grid bars will obscure part of the fascicle, so a compromise is to montage the whole of several grid squares and measure all the axons within them. The average density

Microscopic Anatomy of the Peripheral Nervous System

67

subtle differences in cytoplasmic contents, with axons having more MTs than NFs and a clearer cytoplasmic background. Intermediate filaments in Schwann cell cytoplasm are often grouped, and this rarely happens in axons. With ideal fixation, the axolemma is denser than the Schwann cell plasma membrane. However, when fixation is less than ideal and in pathologic conditions, it may be impossible to distinguish axons from Schwann cell process with sufficient confidence to make morphometry worthwhile.

REGENERATION

FIGURE 3–24 Electron micrograph of flat sheets of Schwann cell processes (SC) formed after unmyelinated axon loss. The basal laminal ensheathment is indicated by an arrow. Bar: 0.2 ␮m.

found by Ochoa and Mair was about 29,000/mm2,298 but intrafascicular variability was estimated as being between 17,000 and 49,000/mm2.297 The size distribution is unimodal until the fifth decade of life,207 and in normal material the smallest fibers are about 0.5 ␮m. Axons with a smaller diameter than this become more common later in life and probably represent unmyelinated fiber regeneration. The highest density in this study was similar to that found by Ochoa and Mair298 in the younger subjects, being 36,000/mm2 at the age of 44 years, and the lowest was 17,300/mm2 in a 67-year-old man. Fibers without myelin and larger than 3 ␮m are almost certainly demyelinated fibers rather than unmyelinated ones. These are easily distinguished by the differences in the relationship of unmyelinated and demyelinated axons with their Schwann cells. As already mentioned, it may also be difficult to distinguish axons from circular Schwann cell profiles. The encircling of Schwann cell processes by other Schwann cell process was described by Ochoa and Vial.299 Axons usually have a lower electron density, but this difference may be obscured in very thin and/or poorly stained sections. If axons are completely encircled by Schwann cell cytoplasm, they will possess a mesaxon, but not all unmyelinated axons are completely indented into the Schwann cell (see Fig. 3–22). In addition, there are very

Schwann cells originally associated with unmyelinated axons but left without an axon after its degeneration still have the ability of adopting regenerative sprouts from myelinated fibers and can construct a myelin sheath on receiving the appropriate axonal signals. The converse can also apply; in regeneration, Schwann cells containing Reich granules, which are only normally found in association with myelinated fibers, may become associated with unmyelinated axons. Schwann cells that have changed from one type of fiber to another will then also change the repertoire of proteins that they express so that they take on the phenotype of the new type of Schwann cell.

The CNS-PNS Transitional Zone J. P. Fraher GENERAL FORM Almost all cranial and spinal nerves are formed from roots. Most roots are in turn formed from a number of rootlets that are attached to an area of the CNS surface (Fig. 3–25). The area through which motor rootlets emerge is the exit zone. The corresponding area for sensory rootlets is the attachment or entry zone. That for the spinal sensory rootlets is the dorsal root entry zone. Each rootlet possesses an individual CNS-PNS interface, at which the two tissue territories meet. It marks the junction of the myelination territories of oligodendrocytes and Schwann cells. The interface generally lies near the plane of the CNS surface and comprises the external surface of the glia limitans, the limiting membrane of the CNS, with its associated basal lamina. The glia limitans is specialized at the transitional zone (TZ), where it is thicker than elsewhere. Its astrocyte processes typically form a barrier across the rootlet that is pierced only by axons. The barrier is irregular, so that CNS and PNS tissues interdigitate over a length of a rootlet, the TZ (see Fig. 3–25).

68

Structure of the Peripheral Nervous System

FIGURE 3–25 A, Diagrammatic transverse section of spinal cord with ventral and dorsal roots, showing the attachment zones (AZ) of their rootlets. B and C, Enlargements of areas outlined in A showing the extents of both dorsal and ventral rootlet transitional zones (TZ) and a central tissue projection (CTP) in the dorsal rootlet TZ. D, Schematic diagram of ventral rootlet TZ indicating an individual myelinated fiber (arrowhead) and a bundle of unmyelinated axons (asterisk) traversing the thickened glia limitans (arrow) in individual tunnels.

The extent of overlap of CNS and PNS tissues at the TZ is often considerable, because the CNS tissue extends outward into the majority of rootlets as a central tissue projection (CTP) (Fig. 3–26). This most frequently takes the form of a tapering cone, embedded in a surrounding PNS tissue compartment (Fig. 3–27D). Its surface is a thick weave of astrocyte processes, while its core resembles white matter, though astrocyte and oligodendrocyte perikarya are absent from smaller examples.127 TZ morphology varies widely among individual nerves (see Fig. 3–27) and also among the constituent rootlets of a given nerve. In some cases the CTP is absent. The interface is then flush with the CNS surface (Fig. 3–27A) or even slightly below it, and the fiber bundle is generally surrounded by a tapering astrocytic collar at and below the plane of the CNS surface (see Fig. 3–25). The thickening of the astrocytic tissue of the glia limitans at

the TZ is often marked, sometimes to the extent that rootlets emerge through an extensive astrocytic pad raised above the plane of the surrounding CNS surface (Fig. 3–27B). Where the root runs close over the CNS surface, the glia limitans may also be thickened, forming a pad underlying the nerve.137 Not uncommonly, the segment of the rootlet nearest the cord consists entirely of CNS tissue (Fig. 3–27E). This arrangement reaches its most extreme expression in the cochlear nerve, which possesses a particularly long CTP. Consequently, in the rat the entire nerve consists of CNS tissue, and PNS tissue is present only in its branches (Fig. 3–27F). Many longer CTPs, such that of the facial nerve, branch distally, with individual branches extending into each division of the nerve (Fig. 3–27G). Unusually, in some rat dorsolateral vagal rootlets an invagination of PNS tissue extends centrally below the brainstem surface349 (Fig. 3–27H). These peripheral tissue insertions have irregular surfaces. As a result, some myelinated axons alternate between the two tissue compartments as they run through the TZ. These variations in form are not systematically related to the fiber composition or rootlet function.136,137 While all TZs tend to be of the same form in any spinal nerve root, the rootlets of many cranial nerves contain a variety of types of TZ, the pattern of which is constant.124,127,136,137,349 CTP size may be associated with fiber bundle size.52,135,388 All larger roots seem to contain one, and the bigger the root, the longer the CTP. Only the smallest rootlets (usually motor) consistently lack one. With regard to spinal roots, the CTP becomes longer in a caudal direction. As a rule, this is more extensive in sensory than in motor rootlets at corresponding levels.135,285,388,415,416 The CTP surface is generally irregular (see Fig. 3–27). Added to this, astrocytic processes project distally from the CTP into the rootlet endoneurium as a glial fringe, which greatly increases the surface area of the CNS-PNS interface.22,24,70 The fringe is much more prominent in the cat than in the rat. The cat possesses dorsal roots that are considerably larger than those of the rat. Its TZ is correspondingly more complex.22,70 On cross section it shows four concentric zones: outer (PNS), glial fringe, mantle, and core (Fig. 3–28). The rat spinal rootlet CTPs lack the well-defined mantle zone found on the cat CTP surface, and that possesses a regular layer of astrocyte cell bodies.

FIBER TRANSITION Myelinated axons traverse the CNS-PNS interface individually, being fully segregated from one another by the abundant astrocytic processes of the TZ as they do so (Fig. 3–29). At the level of the interface, each possesses a transitional node (Fig. 3–30A, B, D, and E).114,141,377 The myelin sheath

Microscopic Anatomy of the Peripheral Nervous System

69

FIGURE 3–26 Light micrograph of longitudinally sectioned rootlets; toluidine blue–stained cryostat sections (⫻40). A ⫽ dorsal rootlet; B ⫽ ventral rootlet; SC ⫽ spinal cord.

FIGURE 3–27 Diagrammatic sections longitudinal to rootlets and at right angles to the CNS surface, showing common types of TZ (A, B, and D–H) and glial islands (C) (see text).

distal to the node is formed by the transitional Schwann cell, and that central to it by the transitional oligodendrocyte (see Fig. 3–30D). The territories affected by demyelinating and other diseases involving these two cell types are correspondingly limited at the TZ. The central end of the Schwann cell commonly lies in a groove or invagination, walled by astrocytic tissue and lined by basal lamina22,70,122,128,131,135–137,260,276,285,347,349,401 (see Fig. 3–30A, D, and F). Invaginations are generally closely packed, so that the TZ astrocyte tissue resembles a honeycomb54,247 (see Fig. 3–30C). The transitional node is essentially a hybrid between central and peripheral types, but with additional unique features.129 For example, an astrocytic flange projects into the node gap, where it separates the oligodendrocyte and Schwann cell paranodes.129,131 The same is true of the Schwann cell and glial basal laminae; these also project into the node gap, where they become continuous with one another and form a further CNS-PNS barrier (see Fig. 3–30E). Because each myelinated fiber possesses a transitional node22–24,53,122,129,130,141,377 over the short TZ segment, node density is several times higher there than in the root or in the cord.118 The possibility of any cross talk taking place between nodes is decreased by their being offset. This is true both for those of ventral motoneurons and those of

70

Structure of the Peripheral Nervous System

FIGURE 3–28 Schematic representation of the transitional region (TR) at the cat nerve root–spinal cord junction. Different central-peripheral borderline arrangements. A, Concave borderline (dashed line) and inverted TR. B, Flat borderline situated at the level of the rootlet (r)–spinal cord junction. C and D, Convex dome-shaped borderline; the CNS expansion into the rootlet is moderate in C and extensive in D. Stippling denotes CNS tissue. The glial fringe is not shown. E, Pointed borderline. The extent of the transitional region (TR) is indicated. The cross-sectional appearance at four different TR levels (A, B, O, and C) and the distribution of the different TR zones are shown in the lower part of the illustration. Light area, endoneurial zone; shaded or cross-hatched areas, glial fringe; light stippling, mantle zone; heavy stippling, core zone. F, Root–spinal cord junction. The root (R) splits into rootlets (r), each with its own TR and attaching separately to the spinal cord (SC). G, Arrangement noted in several cranial nerve roots. The peripheral nervous system component of the root separates into a bundle of closely packed minirootlets, each equipped with a TR. The minirootlets reunite centrally. BS ⫽ brainstem. (From Berthold, C. H., Carlstedt T., and Corneliuson O.: Anatomy of the nerve root at the central-peripheral transition region. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. [eds.]: Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 156, 1984, with permission.)

sensory dorsal rootlet fibers on the CTP surface. In addition, the latter are displaced distally into the rootlet, away from the CNS fibers of the dorsolateral cord. Also, the abundant concentric astrocytic processes that wall the invagination and surround them may insulate them from one another (see Figs. 3–29 and 3–30). Central to the TZ, many of the myelin sheaths become closely apposed to one another. Here also, in rat ventral motoneurons, internodes of one and the same fiber have a smaller axon caliber and thinner myelin sheaths in the CNS than in the PNS (Figs. 3–31 and 3–32).121,123 (The opposite is true of cat S1 dorsal root axons.25) The basal laminae at the TZ form a single compos-

ite continuous surface. That which overlies the glia limitans is projected distally as an array of tubes, each of which surrounds either a single myelinated axon or an unmyelinated axon bundle. Basal lamina thus forms an extensive barrier between the endoneurium and the axonal and glial elements at the TZ. Unmyelinated mammalian axons also traverse the TZ in tunnels. They generally do so as bundles rather than singly (see Fig. 3–25). However, some of the largest examples traverse the TZ glia limitans in individual tunnels, like myelinated axons.121,124 Central to the TZ, unmyelinated axons tend to be less segregated than is

Microscopic Anatomy of the Peripheral Nervous System

71

FIGURE 3–29 Transverse sections through the same fiber bundle at the ventral rootlet (A), TZ (B), and intraspinal (C) levels. B, The TZ is in the upper half of the illustration, transitional fibers being surrounded by pale astrocytic tissue. C, The fibers are sectioned mostly transversely, while those of the surrounding cord are sectioned longitudinally. Bars: 10 ␮m. (Based on Bristol, D. C., et al.: Spacing of central-peripheral transitional nodes of rat motoneurones. Acta Anat. [Basel] 148:206, 1993, with permission of S. Karger AG.)

the case peripherally.349 This is because, over their CNS courses, they lack the segregating processes provided by the Schwann cells peripherally. The mode of ensheathment of some small axons differs in the two milieux. Many that are myelinated peripherally are unmyelinated in the CNS. A few nerves lack a clearly defined TZ glial barrier. These are the olfactory93–95 and vomeronasal (J. P. Fraher, unpublished observations, 1981). In both, the tiny axons traverse the TZ as very large bundles. These are segregated only to a very limited degree, with just a few processes penetrating among them from the sleeve of olfactory ensheathing cells around the periphery of the bundle. These ensheathing cells accompany them through the TZ and into the olfactory bulb.86,126,329

Not uncommonly, isolated clumps of glial tissue are found lying freely in the rootlets, separated from the CNS interface (see Fig. 3–27C). These comprise glial islands.124,415,416 They may consist only of astrocytic tissue, but some also contain oligodendrocytes. They are generally traversed by axons. As these axons enter and leave them, their transitions between the endoneurium and the islands resemble those at the TZ proper.

TISSUE COMPOSITION In addition to axons, myelinating cells, and astrocytes, the TZ contains associated tissues in which they are embedded. These consist peripherally of endoneurium

72

Structure of the Peripheral Nervous System

Microscopic Anatomy of the Peripheral Nervous System

FIGURE 3–30 Myelinated fibers at the TZ. A, Diagram of longitudinally sectioned myelinated axon traversing the TZ. The central end of the Schwann cell is inserted into an invagination bounded by astrocyte processes (A). The CNS-PNS transitional node lies at the bottom of this. Here the basal laminae (BL) of Schwann cell and astrocyte are continuous with one another where they project into the node gap. (From Fraher, J. P.: The transitional zone and CNS regeneration. J. Anat. 196:137, 2000, with permission of Blackwells.) B, Longitudinally sectioned abducent nerve fiber showing transitional node (arrowhead) deep to the plane of the CNS surface (arrow). Thickened glia limitans is denoted by asterisk. (From Fraher, J. P., Smiddy, P. F., and O’Sullivan, V. R.: The central-peripheral transitional regions of cranial nerves: trochlear and abducent nerves. J. Anat. 161:115, 1988, with permission of Blackwells.) C, Scanning electron micrograph of experimentally avulsed cervical ventral rootlet showing the honeycomb-like arrangement of astrocytic tissue at the TZ. The spaces of the honeycomb are the empty invaginations left after avulsion of the Schwann cells. (From Bristol, D. C., and Fraher, J. P.: Experimental traction injuries of ventral spinal nerve roots: a scanning electron microscopic study. Neuropathol. Appl. Neurobiol. 15:549, 1989, with permission of Blackewells.) D, Longitudinally sectioned transitional myelinated fiber showing Schwann cell (top) and oligodendrocytic (bottom) paranodal regions. The invagination surrounding the former is bounded externally by astrocytic processes (asterisk) and is lined by basal lamina (arrows). (From Fraher, J. P., Smiddy, P. F., and O’Sullivan, V. R.: The central-peripheral transitional regions of cranial nerves: trochlear and abducent nerves. J. Anat. 161:115, 1988, with permission of Blackwells.) E, Longitudinal section through a transitional node, showing an astrocytic process (asterisk) extending into the node gap from an adjacent perikaryon. (From Fraher, J.: Node distribution and packing density in the rat CNS-PNS transitional zone. Microsc. Res. Tech. 34:507, 1996, with permission of Wiley-Liss.) F, Electron micrograph of a transversely sectioned transitional Schwann cell just distal to the paranode lying in an invagination bounded by TZ astrocyte processes and lined by basal lamina (arrows). G, Electron micrograph showing the thick, multilayered glia limitans at the TZ. (F and G from Fraher, J. P.: The CNS-PNS transitional zone of the rat: morphometric studies at cranial and spinal levels. Prog. Neurobiol. 38:261, 1992, with permission.) Bars: B and C, 10 ␮m; D and E, 1 ␮m; F and G, 0.5 ␮m.

FIGURE 3–31 Graphs showing myelin sheath thickness plotted against distance along entire peripheral (left) and central (right) internodes belonging to the same typical rat ventral motoneuron axons at 6 days (A) and at maturity (B). The midlevel of the transitional node is arrowed. Note the variation of myelin sheath thickness along the immature internodes. (Based on Fraher, J. P.: Quantitative studies on the maturation of central and peripheral parts of individual ventral motoneuron axons. II. Internodal length. J. Anat. 127:1, 1978.)

73

74

Structure of the Peripheral Nervous System

Microscopic Anatomy of the Peripheral Nervous System

and rootlet sheath,212,413 and centrally of astrocyte processes, which are continuous around the perimeter of the TZ with the glia limitans.129,130,259,260,349 TZ astrocytes have distinctive morphologic features. At least some of them produce a variety of types of processes serving a number of different roles. These include contributing to the TZ glia limitans and to the concentric sheaths walling the invagination around the central ends of the transitional Schwann cells, as well as extending into the TZ node gap (see earlier), where they sometimes contact the nodal axon. Those processes forming the bulk of the CTP and its surface are mainly fine, resembling those of the glia limitans generally41,43,182,260,316,386,409,437,445,464 (see Fig. 3–30G). It is clear, therefore, that individual astrocytes at the TZ give rise to a variety of processes appropriate to their location, and are not specialized as subclasses each of which produces only a specific type of process.

HISTORICAL BACKGROUND Studies on TZ structure (reviewed by Berthold et al.25) date back at least to the mid-19th century. The fundamental light microscopic picture of the TZ was described by Skinner388 and Tarlov.415,416 CTPs were noted from early on,290,443 and Skinner demonstrated the greater size of these in sensory than in motor roots.388 He also identified both astrocytes and oligodendrocytes within them. Tarlov415,416 identified glial islands, and showed that CTPs increase in length rostrocaudally along the neuraxis. Foncin,114 in the first ultrastructural TZ study, showed that the apparent myelin deficiency in the TZ, the Aufhellungszone,202,240,290,443 was a fixation artifact caused

75

by pronounced lamellar separation in the oligodendrocytic paranodes.

DEVELOPMENT The features of TZ development have been studied extensively in rat cervical ventral rootlets.121,124,125,293,294 Axons of developing ventral motoneurons converge on specific exit points in the presumptive glia limitans, possibly influenced by chemoattraction.173,251 They emerge from the neural tube as bundles (the future rootlets) and grow onward across the prospective leptomeninges, where they become associated with cells derived from the neural crest, some of which develop into Schwann cells. At the cord surface these axon bundles are eventually segregated by astrocytic processes, which grow into them from their margins and form a barrier across each (see Fig. 3–32). The segregating processes arise from perikarya located in the glia limitans around the perimeter of the bundle (see Fig. 3–32G). Some of these cells may have migrated outward from the ventricular zone along motoneuron axon bundles.124,125,293,294 Even before barrier formation, clear differences in cellular density (low centrally, high peripherally) commonly mark the presumptive interface. 127,130,135 Segregation begins around embryonic day 13, but is not completed until shortly after birth. Over this period the rootlet axons are extensively segregated in the PNS, a short distance beyond the plane of the cord surface. This is brought about by processes arising from cell clusters that appear on the rootlet surface shortly after axon emergence from the cord (see Fig. 3–32A and C to E).122,124,130,133,134 These form a fine interlocking matrix that segregates the axons and forms a barrier across the rootlet distal

FIGURE 3–32 A and B, Diagrams summarizing early rat ventral rootlet TZ development. A, An axon bundle emerges from the cord (bottom) to form a rootlet that is associated with a cell cluster. Arrows show plane of section of cluster in C. B, The cell cluster disperses in late fetal life, as the bundle begins to be segregated by processes growing in from perikarya surrounding it. Arrows show the plane of section of G. C–E, Electron micrographs of cell clusters. C and D, Transverse sections of a prominent cluster surrounding an axon bundle (C), and the matrix of processes that arises from it and subdivides the bundle (D). E, Longitudinal section of the bundle, showing ensheathing processes extending centrally from cluster cells below the plane of the cord surface, where they are apposed to astrocyte processes of the glia limitans. F–H, Electron micrographs of transversely sectioned axon bundles traversing the TZ, showing near-absence of glial segregating processes at embryonic day 18 (F), axon segregation by astrocyte processes in the immediate postnatal period (G), and subsequent myelination of a fully segregated axon lying in a developing invagination (H). Bars: C, 5 ␮m; D and E, 1 ␮m; F and H, 0.5 ␮m; G, 2 ␮m. (C, D, G, and H from Fraher, J. P.: Axon-glial relationships in early CNS-PNS transitional zone development: an ultrastructural study. J. Neurocytol. 26:41, 1997 with permission of Chapman & Hall.)

76

Structure of the Peripheral Nervous System

to the cord surface (see Fig. 3–32A, C, and D). This precedes formation of the TZ glial barrier and regresses as the former becomes more complete.125 The clusters themselves disperse just before birth. Their component cells serve as a source of Schwann cells.133,134 However, some of these fail to make axonal contact and lie dormant in the rootlet, at least for some time following birth.211 Subsequent TZ development varies between nerves, reflecting the variety of forms of the mature TZ.124 During development, the CNS-PNS interface oscillates and continually changes its form and position as the two tissue classes establish their mutually exclusive territories. (The formation of glial islands is likely to result from glial tissue becoming separated from the CNS during this process and thereafter coming to lie in the endoneurium.124,415,416) Because the astrocytic barrier is at first flush with the adjacent glia limitans, the interface is initially formed in the plane of the surrounding CNS surface. It retains this position, however, in only a minority of locations.124 It most often becomes displaced distally, as CNS tissue grows into the rootlet to form a CTP. Growth of the CTP continues even after transitional node formation. It therefore entails relative elongation of central fiber segments, because the transitional node remains at the interface. Concurrently, the node comes to lie progressively deeper in the invagination as a result of relative distal overgrowth of astrocytic processes around the central end of the transitional Schwann cell. Axons destined for myelination are the first to be fully segregated (see Fig. 3–32B and F to H). Each of these becomes associated with a premyelin Schwann cell and an oligodendrocyte, at some distance peripheral and central, respectively, to the glia limitans. They are at first separated by a bare axon segment. They subsequently extend along the axon toward one another until they reach the interface, where they become closely associated and where the transitional node develops between them.122,129 This entails formation of the Schwann cell and oligodendrocytic paranodes, which come about much as at typical PNS and CNS nodes, respectively. Astrocytic processes and associated basal laminae grow into the nodal gap to provide the basis of a barrier between the two territories. The myelination process is broadly typical for each cell type, though it is specialized in some details.121,143,358 For instance, there is a striking delay in myelination close to the root-cord junction, as compared with the intraspinal and more peripheral parts of the axons (see Fig. 3–31). Thus, during the first postnatal week, there are stretches lacking myelin along axons that are myelinated elsewhere.124 Also, the length of rootlet lying immediately distal to the TZ, the proximal rootlet segment, lags behind the rest of the root in its

development and even in the adult retains some immature features, such as the presence of unusually short internodes.124,130,133,134 Many of these transitional internodes are particularly short, and their myelin shows unusual features.70,134 In the cat, the Schwann cell internuclear distance along unmyelinated axons averages about 12 ␮m in the PNS compartment of the TZ, which should be compared with the 100- to 300-␮m distance found in peripheral nerves generally.68,69 Presumptively unmyelinated axons undergo variable degrees of segregation within the TZ. Some larger ones become segregated completely. Bundles of smaller axons remain much less so. The underlying factors influencing TZ development have been increasingly the subject of recent investigations, many in nonmammalian vertebrates.55,139,163–165,286,395 These studies have shown that the TZ loci of avian cranial nerve rootlet attachments may be determined in the presumptive glia limitans of the early neural tube.286 Cells derived from the neural crest form clusters termed boundary caps (BCs) associated with them. Similar BCs occur at rat presumptive dorsal rootlet TZs.164 They influence sensory axon ingrowth: Axon bundles are arrested near the plane of the developing cord. They may also prevent abnormal cell migration across the CNS-PNS interface, namely, outgrowth of ventral motoneurons into the rootlets.55,139,163–165,395 The cell clusters described earlier on rat ventral rootlets resemble BCs122,124,130,136,137 but are located some distance superficial to the cord surface. Also, they appear well after axon bundle outgrowth and so, unlike BCs, seem not to influence this. The TZ barrier retains its integrity from the outset. In principle, Schwann cells could invade the CNS without penetrating basal lamina. This is because their sheaths of basal lamina are continuous with that covering the glia limitans (see Fig. 3–30). However, this occurs rarely; they are only occasionally found in the undamaged CNS.2,326 Potential sources of these include the isolated cells derived from the cell clusters211 and the Schwann cells of autonomic nerves related to CNS blood vessels. Experimental studies suggest that the TZ astrocytic processes constitute the main barrier that prevents Schwann cell invasion of the CNS. This occurs following irradiation damage to the glia limitans.138,159–161,382–385 The barrier function is also exemplified in the Jimpy mouse, in which, despite severe central myelin deficiency, Schwann cells do not invade the CNS to remedy the defect.276 Neither do Schwann cell precursors invade the CNS before formation of the astrocytic TZ barrier. The cluster cell barrier may play a part in this.124,126 Thus between them the two barriers could prevent Schwann cell invasion during development and subsequently. Resistance could be aided by the expression of the nerve cell adhesion molecule HNK-1 by central axon

Microscopic Anatomy of the Peripheral Nervous System

segments.293 Their consequent close adherence could prevent axons from being pried apart by centrally directed Schwann cell processes.

COMPARATIVE ANATOMY The TZs that have been studied most extensively are those of the cat S1 dorsal root22–24,67,68 and of rat cranial and spinal nerves, in which TZ development has also been investigated in considerable detail.119,124,126 These and other investigations show that, in its general form, TZ structure is similar in a variety of mammals, including humans,285,387,388 rats, mice, cats, monkeys, and cattle.260,276,283,284,349,364,394,401 Although the range of nonmammalian vertebrates studied (including birds, amphibians, cartilaginous fishes, and cyclostomes) is necessarily incomplete, it seems likely that they all typically possess a glial barrier at the TZ. Data from some lower vertebrate species (Squalus acanthias, Tinca vulgaris, and Rana esculenta) indicate that, in these, motor and sensory rootlets do not differ in this respect.467 It is not yet known if, like the mammals, they lack a clearly defined glial barrier at the olfactory and vomeronasal nerve TZs. In the cyclostome Petromyzon (the sea lamprey), the large ventral motoneuron axons traverse the distinctive glia limitans in individual tunnels.120 In this respect they follow the general vertebrate pattern, though they are unmyelinated. They differ from this, however, in being associated at the point of emergence from the cord with an extensive labyrinth of paraxonal spaces. This is bounded by glial processes, Schwann cell processes, and basal lamina, and bears some comparison with the mammalian node gap in that it involves exposing a large surface area to the extracellular spaces, a feature that could play a part in electrolyte control (Fig. 3–33A). In those animals studied that are classified below the cyclostomes, a TZ barrier seems to be absent. For example, in the cephalochordate Amphioxus, neurites traverse the dorsal nerve attachments to the cord.119 While there is no glial barrier here, glial density is increased near the junction (Fig. 3–33B). Amphioxus has no ventral root. Instead, the motoneuron axons remain in the cord and make specialized contacts resembling motor end plates with extensions of the segmental and notochordal musculature, which are directly apposed to the cord surface. In these locations the glia limitans is absent113,144 (Fig. 3–33C and D). Invertebrate nervous systems vary considerably in form. Nevertheless, at least in those insects, copepods, and snails studied at the ultrastructural level, clearly defined glial interfaces seem not to occur where nerve fibers pass through junctions between nerve cords, interconnecting plexuses, and ganglia (J. P. Fraher, unpublished observations).

77

BLOOD SUPPLY The rat ventral rootlet TZ has a very rich blood supply, perhaps to meet the metabolic requirements of the high concentration of nodes there.213 The vessels are mainly capillaries or postcapillary venules.66,180,252,304,317,408,419 They form a densely knit network in the labyrinth of spaces within the pia mater and superficial to the glia limitans. Here they are closely packed around the rootlets, rather than within them. The latter are very small and their endoneurial spaces rarely if ever contain capillaries. Many vessels are continued centrally into the cord along the ventral motoneuron axon bundles, which they commonly follow across the ventral white column as far as the ventral horn gray matter.122,130,212 This arrangement differs from that at the cat S1 dorsal rootlet TZ. Here, the blood vessels do not accompany the axons between PNS and CNS. Instead they deviate from the endoneurial space and join vessels on the cord surface.22,70 Dorsal roots may as a consequence be more susceptible to ischemia than ventral roots.

MECHANICAL FEATURES In the subarachnoid space, each ventral rootlet is covered by a cellular and connective tissue rootlet sheath.212 The outer layers of this are continuous with the pia mater. This arrangement strengthens the rootlet and the junctional region generally. Central to this, the rootlets lie in the labyrinthine intrapial space. Here the rootlet-cord attachments are strengthened by the intricate interdigitation and adhesion that take place between sheath cells and TZ astrocyte processes.132 Longitudinally running rootlet endoneurial collagen fiber bundles may also strengthen the TZ by their becoming attached to the basal lamina. This also helps to anchor the rootlet to the CNS. The high node packing density at the TZ is likely to detract from its mechanical strength, because the node, being bare, is likely to be a weak point on the fiber.54 However, the transitional Schwann cells help to compensate for this by resisting traction through the collagenous connections between their basal laminae and that of the astrocytes bounding the invagination surrounding the central end of the Schwann cell. The invagination could also act as a suction cup, resisting distraction of the paranode. These mechanical features may protect the TZs against stress-related damage during relative movements between the neuraxis and its bony casing, once any slack has been taken up. The effectiveness of these mechanisms is shown by the fact that experimental rootlet traction tends to result in rupture at the level of the rootlet, rather than at the TZ.

FIGURE 3–33 A, Electron micrograph of a longitudinal section of part of a lamprey TZ, showing Schwann cell processes, astrocytic processes, and basal lamina (arrowheads) bounding a complex labyrinth of spaces (asterisks) in relation to the axon (top right) and the glia limitans (bottom). Schwann cell and astrocyte processes are not separated by basal lamina in the vicinity of the axon (bottom right). (From Fraher, J., and Cheong, E.: Glial-Schwann cell specialisations at the central-peripheral nervous system transition of a cyclostome: an ultrastructural study. Acta Anat. (Basel) 154:300, 1995, with permission of S Karger AG, Basel.) B and C, Transverse sections of Amphioxus spinal cord. B, The dorsal nerve extends to the left. At its attachment there is no glial barrier. Glial perikarya are most densely packed in the region of the tapering transitional segment (asterisk). Their density decreases progressively in a central-peripheral direction. (From Fraher, J. P.: The transitional zone and CNS regeneration. J. Anat. 196:137, 2000, with permission of Blackwells, Oxford, UK.) C, In the ventrolateral part of the cord, its surface is closely related to muscle elements (arrows): laterally to processes of myotomal muscles and ventrally to notochordal muscle-like processes. The glia limitans is deficient at both locations and the notochordal sheath is deficient at the second. D, Electron micrograph of the apposed cord surface and a muscle element, showing contacts resembling complex, wide neuromuscular junctions between the two. Here, terminals of propriospinal axons within the cord (arrows) are separated by a synaptic cleft from a muscle element (asterisk). Bars: A and D, 0.5 ␮m; B and C, 25 ␮m.

Microscopic Anatomy of the Peripheral Nervous System

REFERENCES 1. Abbott, N. J., Lieberman, E. M., Pichon, Y., et al.: Periaxonal K⫹ regulation in the small squid Alloteuthis: studies on isolated and in situ axons. Biophys. J. 53:275, 1988. 2. Adelman, L. S., and Aronson, S. M.: Intramedullary nerve fiber and Schwann cell proliferation within the spinal cord (schwannosis). Neurology 22:726, 1972. 3. Alvarez, J.: The autonomous axon: a model based on local synthesis of proteins. Biol. Res. 34:103, 2001. 4. Alvarez-Leefmans, F. J., Leon-Olea, M., Mendoza-Sotelo, J., et al.: Immunolocalization of the Na(⫹)-K(⫹)-2Cl(⫺) cotransporter in peripheral nervous tissue of vertebrates. Neuroscience 104:569, 2001. 5. Ariyasu, R. G., and Ellisman, M. H.: The distribution of (Na⫹ ⫹ K⫹)ATPase is continuous along the axolemma of unensheathed axons from spinal roots of ‘dystrophic’ mice. J. Neurocytol. 16:239, 1987. 6. Ariyasu, R. G., Nichol, J. A., and Ellisman, M. H.: Localization of sodium/potassium adenosine triphosphatase in multiple cell types of the murine nervous system with antibodies raised against the enzyme from kidney. J. Neurosci. 5:2581, 1985. 7. Armstrong, R., Toews, A. D., and Morell, P.: Axonal transport through nodes of Ranvier. Brain Res. 412:196, 1987. 8. Arroyo, E. J., Xu, Y. T., Zhou, L., et al.: Myelinating Schwann cells determine the internodal localization of Kv1.1, Kv1.2, Kvbeta2, and Caspr. J. Neurocytol. 28:333, 1999. 9. Asbury A. K.: The biology of Schwann cells. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 201, 1975. 10. Baas, P. W.: Microtubule transport in the axon. Int. Rev. Cytol. 212:41, 2002. 11. Baas, P. W., and Black, M. M.: Individual microtubules in the axon consist of domains that differ in both composition and stability. J. Cell Biol. 111:495, 1990. 12. Baas, P. W., Slaughter, T., Brown, A., and Black, M. M.: Microtubule dynamics in axons and dendrites. J Neurosci. Res. 30:134, 1991. 13. Bacsich, P.: On the presence of Timofeew’s sensory corpuscles in the autonomic plexuses of the human prostate and seminal vesicles. J. Anat. 104:182, 1969. 14. Baker, M., Bostock, H., Grafe, P., and Martius, P.: Function and distribution of three types of rectifying channel in rat spinal root myelinated axons. J. Physiol. (Lond.) 383:45, 1987. 15. Beamish, N. G., Stolinski, C., Thomas, P. K., and King, R. H. M.: Freeze-fracture observations on normal and abnormal human perineurial tight junctions: alterations in diabetic polyneuropathy. Acta Neuropathol. (Berl.) 81:269, 1991. 16. Beggs, J., Johnson, P. C., Olafsen, A., et al.: Transperineurial arterioles in human sural nerve. J. Neuropathol. Exp. Neurol. 50:704, 1991. 17. Behse, F.: Morphometric studies on the human sural nerve. Acta Neurol. Scand. Suppl. 132:1, 1990. 18. Bennett, V., Lambert, S., Davis, J. Q., and Zhang, X.: Molecular architecture of the specialized axonal membrane at the node of Ranvier. Soc. Gen. Physiol. Ser. 52:107, 1997.

79

19. Berghs, S., Aggujaro, D., Dirkx, R. Jr., et al.: BetaIV spectrin, a new spectrin localized at axon initial segments and nodes of Ranvier in the central and peripheral nervous system. J. Cell Biol. 151:985, 2000. 20. Berthold, C. H.: Morphology of normal peripheral axons. In Waxman, S. G. (ed.): Physiology and Pathobiology of Axons. New York, Raven Press, p. 3, 1978. 21. Berthold, C. H.: Some aspects of the ultrastructural organization of peripheral myelinated axons in the cat. In Weiss, D. G. (ed.): Axoplasmic Transport. Berlin, Springer-Verlag, p. 40, 1982. 22. Berthold, C. H., and Carlstedt, T.: Observations on the morphology at the transition between the peripheral and the central nervous system in the cat. II. General organization of the transitional region in S1 dorsal rootlets. Acta Physiol. Scand. Suppl. 446:23, 1977. 23. Berthold, C. H., and Carlstedt, T.: Observations on the morphology at the transition between the peripheral and the central nervous system in the cat. III. Myelinated fibres in S1 dorsal rootlets. Acta Physiol. Scand. Suppl. 446:43, 1977. 24. Berthold, C. H., and Carlstedt, T.: Observations on the morphology at the transition between the peripheral and the central nervous system in the cat. V. A light microscopical and histochemical study of S1 dorsal rootlets in developing kittens. Acta Physiol. Scand. Suppl. 446:73, 1977. 25. Berthold, C. H., Carlstedt, T., and Corneliuson, O.: Anatomy of the nerve root at the central-peripheral transition region. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 156, 1984. 26. Berthold, C. H., Corneliuson, O., and Mellstrom, A.: Peroxidase activity at nodes of Ranvier in lumbosacral ventral spinal roots and in the PNS-CNS transitional region after intramuscular administration of horseradish peroxidase. J. Neurocytol. 15:253, 1986. 27. Berthold, C. H., Corneliuson, O., and Rydmark, M.: Changes in shape and size of cat spinal root myelinated nerve fibers during fixation and Vestopal-w embedding for electron microscopy. J. Ultrastruct. Res. 80:23, 1982. 28. Berthold, C. H., Fabricius, C., Rydmark, M., and Andersen, B.: Axoplasmic organelles at nodes of Ranvier. I. Occurrence and distribution in large myelinated spinal root axons of the adult cat. J. Neurocytol. 22:925, 1993. 29. Berthold, C. H., and Mellstrom, A.: Peroxidase activity at consecutive nodes of Ranvier in the nerve to the medial gastrocnemius muscle after intramuscular administration of horseradish peroxidase. Neuroscience 19:1349, 1986. 30. Berthold, C. H., Nilsson, I., and Rydmark, M.: Axon diameter and myelin sheath thickness in nerve fibres of the ventral spinal root of the seventh lumbar nerve of the adult and developing cat. J. Anat. 136:483, 1983. 31. Berthold, C. H., Nordborg, C., Hildebrand, C., et al.: Sural nerve biopsies from workers with a history of chronic exposure to organic solvents and from normal control cases: morphometric and ultrastructural studies. Acta Neuropathol. (Berl.) 62:73, 1983. 32. Berthold, C. H., and Rydmark, M.: Anatomy of the paranode-node-paranode region in the cat. Experientia 39:964, 1983.

80

Structure of the Peripheral Nervous System

33. Berthold, C. H., and Rydmark, M.: Electron microscopic serial section analysis of nodes of Ranvier in lumbosacral spinal roots of the cat: ultrastructural organization of nodal compartments in fibres of different sizes. J. Neurocytol. 12:475, 1983. 34. Berthold, C. H., and Rydmark, M.: Morphology of normal peripheral axons. In Waxman, S. G., Kocsis, J. D., and Stys, P. K. (eds.): The Axon: Structure, Function and Pathophysiology. Oxford, UK, Oxford University Press, p. 13, 1995. 35. Bischoff, A., and Thomas, P. K.: Microscopic anatomy of myelinated nerve fibers. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 104, 1975. 36. Black, J. A., Friedman, B., Waxman, S. G., et al.: Immunoultrastructural localization of sodium channels at nodes of Ranvier and perinodal astrocytes in rat optic nerve. Proc. R. Soc. Lond. B Biol. Sci. 238:39, 1989. 37. Black, J. A., Kocsis, J. D., and Waxman, S. G.: Ion channel organization of the myelinated fiber. Trends Neurosci. 13:48, 1990. 38. Black, M. M., Chestnut, M. H., Pleasure, I. T., and Keen, J. H.: Stable clathrin: uncoating protein (hsc70) complexes in intact neurons and their axonal transport. J. Neurosci. 11:1163, 1991. 39. Black, M. M., and Lasek, R. J.: Slow components of axonal transport: two cytoskeletal networks. J. Cell Biol. 86:616, 1980. 40. Blanchard, C. E., Mackenzie, M. L., Sikri, K., and Allt, G.: Filipin-sterol complexes at nodes of Ranvier. J. Neurocytol. 14:1053, 1985. 41. Bondareff, W., and McLone, D. G.: The external glial limiting membrane in Macaca: ultrastructure of a laminated glioepithelium. Am. J. Anat. 136:277, 1973. 42. Boyle, M. E., Berglund, E. O., Murai, K. K., et al.: Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30:385, 2001. 43. Braak, E.: On the fine structure of the external glial layer in the isocortex of man. Cell Tissue Res. 157:367, 1975. 44. Brady, S. T., Lasek, R. J., and Allen, R. D.: Video microscopy of fast axonal transport in extruded axoplasm: a new model for study of molecular mechanisms. Cell Motil. 5:81, 1985. 45. Bray, D., and Bunge, M. B.: Serial analysis of microtubules in cultured rat sensory axons. J. Neurocytol. 10:589, 1981. 46. Bray, G. M., Rasminsky, M., and Aguayo, A. J.: Interactions between axons and their sheath cells. Annu. Rev. Neurosci. 4:127, 1981. 47. Bray, J. J., Fernyhough, P., Bamburg, J. R., and Bray, D.: Actin depolymerizing factor is a component of slow axonal transport. J. Neurochem. 58:2081, 1992. 48. Brismar, T.: Potential clamp experiments on myelinated nerve fibres from alloxan diabetic rats. Acta Physiol. Scand. 105:384, 1979. 49. Brismar, T.: Potential clamp analysis of membrane currents in rat myelinated nerve fibres. J. Physiol. (Lond.) 298:171, 1980. 50. Brismar, T., Hildebrand, C., and Berglund, S.: Nodes of Ranvier above a neuroma in the rat sciatic nerve: voltage

51.

52.

53.

54.

55.

56. 57.

58.

59. 60.

61. 62.

63.

64.

65. 66.

67.

68.

clamp analysis and electron microscopy. Brain Res. 378:347, 1986. Brismar, T., Hildebrand, C., and Tegner, R.: Nodes of Ranvier in acrylamide neuropathy: voltage clamp and electron microscopic analysis of rat sciatic nerve fibres at proximal levels. Brain Res. 423:135, 1987. Bristol, D. C.: Aspects of the ventral rootlet transitional zone in the rat: a morphometric and morphological analysis of the L, ventral rootlet transitional zone and the distribution of transitional nodes of Ranvier within it. Ph.D. Thesis, Dublin, National University of Ireland, 1989. Bristol, D. C., and Fraher, J. P.: A morphometric study of the CNS-PNS transitional zone in rat lumbar ventral roots. J. Anat. 152:236, 1987. Bristol, D. C., and Fraher, J. P.: Experimental traction injuries of ventral spinal nerve roots: a scanning electron microscopic study. Neuropathol. Appl. Neurobiol. 15:549, 1989. Britsch, S., Goerich, D. E., Riethmacher, D., et al.: The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev. 15:66, 2001. Brophy, P. J.: Axoglial junctions: separate the channels or scramble the message. Curr. Biol. 11:R555, 2001. Brown, A. A., Xu, T., Arroyo, E. J., et al.: Molecular organization of the nodal region is not altered in spontaneously diabetic BB-Wistar rats. J. Neurosci. Res. 65:139, 2001. Brown, E. R., and Abbott, N. J.: Ultrastructure and permeability of the Schwann cell layer surrounding the giant axon of the squid. J. Neurocytol. 22:283, 1993. Bruns, R. R.: Beaded filaments and long-spacing fibrils: relation to type VI collagen. J. Ultrastruct. Res. 89:136, 1984. Bruns, R. R., Press, W., Engvall, E., et al.: Type VI collagen in extracellular, 100-nm periodic filaments and fibrils: identification by immunoelectron microscopy. J. Cell Biol. 103:393, 1986. Bunge, M. B.: The axonal cytoskeleton: its role in generating and maintaining cell forms. Trends Neurosci. 9:477, 1986. Bunge, M. B., Wood, P. M., Tynan, L. B., et al.: Perineurium originates from fibroblasts: demonstration in vitro with a retroviral marker. Science 243:229, 1989. Bungner, O. v.: Versuche zum Studium der histologischen Degeneration und Regeneration verletzer Nerven, ihre Anordnung und ihre Ergebnisse. Zieglers Beitr. 10:321, 1891. Burgoyne, R. D.: Cytoskeleton is a major neuronal organelle. In Burgoyne, R. D. (ed.): The Neuronal Cytoskeleton. New York, Wiley-Liss, p. 1, 1991. Burton, P. R.: Microtubules of frog olfactory axons: their length and number/axon. Brain Res. 409:71, 1987. Caley, D. W., and Maxwell, D. S.: Development of the blood vessels and extracellular spaces during postnatal maturation of rat cerebral cortex. J. Comp. Neurol. 138:31, 1970. Carlstedt, T.: Observations on the morphology at the transition between the peripheral and the central nervous system in the cat. I. A preparative procedure useful for electron microscopy of the lumbosacral dorsal. Acta Physiol. Scand. Suppl. 446:5, 1977. Carlstedt, T.: Observations on the morphology at the transition between the peripheral and the central nervous system in the cat. IV. Unmyelinated fibres in S1 dorsal rootlets. Acta Physiol. Scand. Suppl. 446:61, 1977.

Microscopic Anatomy of the Peripheral Nervous System 69. Carlstedt, T.: Internodal length of nerve fibres in dorsal roots of cat spinal cord. Neurosci. Lett. 19:251, 1980. 70. Carlstedt, T.: An electron-microscopical study of the developing transitional region in feline S1 dorsal rootlets. J. Neurol. Sci. 50:357, 1981. 71. Causey, G.: The Cell of Schwann. Baltimore, Williams & Wilkins, 1960. 72. Chiu, S. Y.: Changes in excitable membrane properties in Schwann cells of adult rabbit sciatic nerves following nerve transection. J. Physiol. (Lond.) 396:173, 1988. 73. Chiu, S. Y., and Ritchie, J. M.: Evidence for the presence of potassium channels in the paranodal region of acutely demyelinated mammalian single nerve fibres. J. Physiol. (Lond.) 313:415, 1981. 74. Chiu, S. Y., Schrager, P., and Ritchie, J. M.: Neuronal-type Na⫹ and K⫹ channels in rabbit cultured Schwann cells. Nature 311:156, 1984. 75. Chiu, S. Y., and Schwarz, W.: Sodium and potassium currents in acutely demyelinated internodes of rabbit sciatic nerves. J. Physiol. (Lond.) 391:631, 1987. 76. Clarke, E., and Bearn, J. G.: The spiral nerve bands of Fontana. Brain 95:1, 1972. 77. Clermont, Y.: Contractile elements in the limiting membranes of the seminiferous tubules. Exp. Cell Res. 15:438, 1958. 78. Cui, S. Y., Christensen, E. I., and Nielsen, S.: Membrane traffic after inhibition of endocytosis in renal proximal tubules. J. Struct. Biol. 107:201, 1991. 79. Dahlstrom, A. B., Czernik, A. J., and Li, J. Y.: Organelles in fast axonal transport: what molecules do they carry in anterograde vs retrograde directions, as observed in mammalian systems? Mol. Neurobiol. 6:157, 1992. 80. Davis, J. Q., Lambert, S., and Bennett, V.: Molecular composition of the node of Ranvier: identification of ankyrin-binding cell adhesion molecules neurofascin (mucin ⫹ /third FNIII domain–) and NrCAM at nodal axon segments. J. Cell Biol. 135:1355, 1996. 81. de la Motte, D. J., Hall, S. M., and Allt, G.: A study of the perineurium in peripheral nerve pathology. Acta Neuropathol. 33:257, 1975. 82. De Waegh, S., and Brady, S. T.: Altered slow axonal transport and regeneration in a myelin-deficient mutant mouse: the trembler as an in vivo model for Schwann cell-axon interactions. J. Neurosci. 10:1855, 1990. 83. De Waegh, S. M., and Brady, S. T.: Local control of axonal properties by Schwann cells: neurofilaments and axonal transport in homologous and heterologous nerve grafts. J. Neurosci. Res. 30:201, 1991. 84. De Waegh, S. M., Lee, V. M., and Brady, S. T.: Local modulation of neurofilament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68:451, 1992. 85. DeCoster, M. A., and DeVries, G. H.: Evidence that the axolemmal mitogen for cultured Schwann cells is a positively charged, heparan sulfate proteoglycan-bound, heparin-displaceable molecule. J. Neurosci. Res. 22:283, 1989. 86. Devon, R., and Doucette, R.: Olfactory ensheathing cells do not require l-ascorbic acid in vitro to assemble a basal lamina or to myelinate dorsal root ganglion neurites. Brain Res. 688:223, 1995.

81

87. Dingemans, K. P., and Teeling, P.: Long-spacing collagen and proteoglycans in pathologic tissues. Ultrastruct. Pathol. 18:539, 1994. 88. Dolapchieva, S., Ichev, K., and Ovtscharoff, W.: Bismuth iodide impregnation of the peripheral nerve fibres. Z. Mikrosk. Anat. Forsch. 99:804, 1985. 89. Dolapchieva, S., Ichev, K., and Ovtscharoff, W.: Lectin binding-sites in axon-myelin-Schwann cell complex. Acta Histochem. Cytochem. 19:253, 1986. 90. Dolapchieva, S., Ichev, K., and Ovtscharoff, W.: Ultrastructural cytochemical localization of 5⬘-nucleotidase activity in axonmyelin-Schwann cell complex. Acta Histochem. 83:125, 1988. 91. Dolapchieva, S., Ichev, K., and Ovtscharoff, W.: Cytochemical localization of ATPase in axon-myelin-Schwann cell complex type. Z. Mikrosk. Anat. Forsch. 103:151, 1989. 92. Dolapchieva, S., Ovtscharoff, W., and Ichev, K.: Localizations of ruthenium red positive material in rabbit peripheral nerves. Acta Histochem. 78:19, 1986. 93. Doucette, R.: Glial influences on axonal growth in the primary olfactory system. Glia 3:433, 1990. 94. Doucette, R.: PNS-CNS transitional zone of the first cranial nerve. J. Comp. Neurol. 312:451, 1991. 95. Doucette, R.: Glial cells in the nerve fiber layer of the main olfactory bulb of embryonic and adult mammals. Microsc. Res. Tech. 24:113, 1993. 96. Dubois, J. M.: Capsaicin blocks one class of K⫹ channels in the frog node of Ranvier. Brain Res. 245:372, 1982. 97. Duce, I. R., and Keen, P.: Can neuronal smooth endoplasmic reticulum function as a calcium reservoir? Neuroscience 3:837, 1978. 98. Dupree, J. L., Girault, J. A., and Popko, B.: Axo-glial interactions regulate the localization of axonal paranodal proteins. J. Cell Biol. 147:1145, 1999. 99. Dyck, P. J., Giannini, C., and Lais, A.: Pathologic alterations in nerves. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 514, 1993. 100. Eames, R. A., and Gamble, H. J.: Schwann cell relationships in normal human cutaneous nerves. J. Anat. 106:417, 1970. 101. Ellisman, M. H.: Beyond neurofilaments and microtubules. Neurosci. Res. Program Bull. 19:43, 1981. 102. Ellisman, M. H., Friedman, P. L., and Hamilton, W. J.: The localization of sodium and calcium to Schwann cell paranodal loops at nodes of Ranvier and of calcium to compact myelin. J. Neurocytol. 9:185, 1980. 103. Ellisman, M. H., and Levinson, S. R.: Immunocytochemical localization of sodium channel distributions in the excitable membranes of Electrophorus electricus. Proc. Natl. Acad. Sci. U. S. A. 79:6707, 1982. 104. Ellisman, M. H., and Lindsey, J. D.: The axoplasmic reticulum within myelinated axons is not transported rapidly. J. Neurocytol. 12:393, 1983. 105. Ellisman, M. H., and Porter, K. R.: Microtrabecular structure of the axoplasmic matrix: visualization of cross-linking structures and their distribution. J. Cell Biol. 87:464, 1980. 106. Ellisman, M. H., Wiley, C. A., Lindsey, J. D., et al.: Structure and function of the cytoskeleton and endomembrane systems at the node of Ranvier. In Zagoren, J. C., and Fedoroff, S. (eds.): The Node of Ranvier. New York, Academic Press, p. 153, 1984.

82

Structure of the Peripheral Nervous System

107. Espejo, F., and Alvarez, J.: Microtubules and calibers in normal and regenerating axons of the sural nerve of the rat. J. Comp. Neurol. 250:65, 1986. 108. Evans, P. D., Reale, V., Merzon, R. M., and Villegas, J.: The effect of a glutamate uptake inhibitor on axon-Schwann cell signalling in the squid giant nerve fibre. J. Exp. Biol. 173:251, 1992. 109. Evans, P. D., Reale, V., Merzon, R. M., and Villegas, J.: N-methyl-D-aspartate (NMDA) and non-NMDA (metabotropic) type glutamate receptors modulate the membrane potential of the Schwann cell of the squid giant nerve fibre. J. Exp. Biol. 173:229, 1992. 110. Fabricius, C., Berthold, C. H., and Rydmark, M.: Axoplasmic organelles at nodes of Ranvier. II. Occurrence and distribution in large myelinated spinal cord axons of the adult cat. J. Neurocytol. 22:941, 1993. 111. Fannon, A. M., Sherman, D. L., Ilyina-Gragerova, G., et al.: Novel E-cadherin-mediated adhesion in peripheral nerve: Schwann cell architecture is stabilized by autotypic adherens junctions. J. Cell Biol. 129:189, 1995. 112. Fath, K. R., and Lasek, R. J.: Two classes of actin microfilaments are associated with the inner cytoskeleton of axons. J. Cell Biol. 107:613, 1988. 113. Flood, P. R.: A peculiar mode of muscular innervation in Amphioxus: light and electron microscopic studies of the so-called ventral roots. J. Comp. Neurol. 126:181, 1966. 114. Foncin, J. F.: Structure fine de la zone de passage radiculomedullaire. Rev. Neurol. (Paris) 105:509, 1961. 115. Fontana, F.: Traité sur le Vénin de la Vipere sur les Poisons Americains. Florence, 1781. 116. Forman, D. S.: Axonal transport of mitochondria. In Smith, R. S., and Bisby, M. A. (eds.): Axonal Transport. New York, Alan R Liss, p. 155, 1987. 117. Forman, D. S., Lynch, K. J., and Smith, R. S.: Organelle dynamics in lobster axons: anterograde, retrograde and stationary mitochondria. Brain Res. 412:96, 1987. 118. Fraher, J.: Node distribution and packing density in the rat CNS-PNS transitional zone. Microsc. Res. Tech. 34:507, 1996. 119. Fraher, J.: Axons and glial interfaces: ultrastructural studies. J. Anat. 200:415, 2002. 120. Fraher, J., and Cheong, E.: Glial-Schwann cell specialisations at the central-peripheral nervous system transition of a cyclostome: an ultrastructural study. Acta Anat. (Basel) 154:300, 1995. 121. Fraher, J. P.: The growth and myelination of central and peripheral segments of ventral motoneurone axons: a quantitative ultrastructural study. Brain Res. 105:193, 1976. 122. Fraher, J. P.: The maturation of the ventral root-spinal cord transitional zone: an ultrastructural study. J. Neurol. Sci. 36:427, 1978. 123. Fraher, J. P.: Quantitative studies on the maturation of central and peripheral parts of individual ventral motoneuron axons. I. Myelin sheath and axon calibre. J. Anat. 126:509, 1978. 124. Fraher, J. P.: The CNS-PNS transitional zone of the rat: morphometric studies at cranial and spinal levels. Prog. Neurobiol. 38:261, 1992. 125. Fraher, J. P.: Axon-glial relationships in early CNS-PNS transitional zone development: an ultrastructural study. J. Neurocytol. 26:41, 1997.

126. Fraher, J. P.: The transitional zone and CNS regeneration. J. Anat. 196:137, 2000. 127. Fraher, J. P., and Delanty, F. J.: The development of the central-peripheral transitional zone of the rat cochlear nerve: a light microscopic study. J. Anat. 155:109, 1987. 128. Fraher, J. P., and Kaar, G. F.: The maturation of the ventral root-spinal cord transitional zone. Part 2. A quantitative ultrastructural study of the dynamics of its early development. J. Neurol. Sci. 53:63, 1982. 129. Fraher, J. P., and Kaar, G. F.: The transitional node of Ranvier at the junction of the central and peripheral nervous systems: an ultrastructural study of its development and mature form. J. Anat. 139:215, 1984. 130. Fraher, J. P., and Kaar, G. F.: The lumbar ventral root-spinal cord transitional zone in the rat: a morphological study during development and at maturity. J. Anat. 145:109, 1986. 131. Fraher, J. P., Kaar, G. F., Bristol, D. C., and Rossiter, J. P.: Development of ventral spinal motoneurone fibres: a correlative study of the growth and maturation of central and peripheral segments of large and small fibre classes. Prog. Neurobiol. 31:199, 1988. 132. Fraher, J. P., and O’Leary, D.: Morphological specialisations of rat cranial nerve transitional zones. J. Anat. 184:119, 1994. 133. Fraher, J. P., and Rossiter, J. P.: Cell clusters on fetal rat ventral roots: prenatal development. J. Anat. 136:111, 1983. 134. Fraher, J. P., and Rossiter, J. P.: Cell clusters on rat ventral roots: postnatal development. J. Anat. 137:555, 1983. 135. Fraher, J. P., and Sheehan, M. M.: The CNS-PNS transitional zone of rat cervical dorsal roots during development and at maturity: a morphological and morphometric study. J. Anat. 152:189, 1987. 136. Fraher, J. P., Smiddy, P. F., and O’Sullivan, V. R.: The centralperipheral transitional regions of cranial nerves: oculomotor nerve. J. Anat. 161:103, 1988. 137. Fraher, J. P., Smiddy, P. F., and O’Sullivan, V. R.: The centralperipheral transitional regions of cranial nerves: trochlear and abducent nerves. J. Anat. 161:115, 1988. 138. Franklin, R. J., and Blakemore, W. F.: Requirements for Schwann cell migration within CNS environments: a viewpoint. Int. J. Dev. Neurosci. 11:641, 1993. 139. Franz, T.: Defective ensheathment of motoric nerves in the Splotch mutant mouse. Acta Anat. (Basel) 138:246, 1990. 140. Freedman, S. D., and Lentz, T. L.: Binding of horseradish peroxidase-alpha-bungarotoxin to axonal membranes at the node of Ranvier. J. Comp. Neurol. 193:179, 1980. 141. Friede, R. L.: A Histochemical Atlas of Tissue Oxidation in the Brain Stem of the Cat. Basel, S. Karger, p. 18, 1961. 142. Friede, R. L., and Bischhausen, R.: The organization of endoneural collagen in peripheral nerves as revealed with the scanning electron microscope. J. Neurol. Sci. 38:83, 1978. 143. Friede, R. L., and Samorajski, T.: Myelin formation in the sciatic nerve of the rat—a quantitative electron microscopic, histochemical and radioautographic study. J. Neuropathol. Exp. Neurol. 27:546, 1968. 144. Fritzsch, B., and Northcutt, R. G.: Cranial and spinal nerve organization in Amphioxus and lampreys: evidence for an ancestral craniate pattern. Acta Anat. (Basel) 148:96, 1993.

Microscopic Anatomy of the Peripheral Nervous System 145. Gabriel, G., Thomas, P. K., King, R. H. M., et al.: Freezefracture observations on human peripheral nerve. J. Anat. 146:153, 1986. 146. Gamble, H. J., and Breathnach, A. S.: An electron-microscope study of human foetal peripheral nerve. J. Anat. 99:573, 1965. 147. Gasser, H. S.: Discussion of the hypothesis of saltatory conduction. Cold Spring Harb. Symp. Quant. Biol. 17:32, 1952. 148. Gatzinsky, K. P.: Node-paranode regions as local degradative centres in alpha-motor axons. Microsc. Res. Tech. 34:492, 1996. 149. Gatzinsky, K. P., and Berthold, C. H.: Lysosomal activity at nodes of Ranvier during retrograde axonal transport of horseradish peroxidase in alpha-motor neurons of the cat. J. Neurocytol. 19:989, 1990. 150. Gatzinsky, K. P., Berthold, C. H., and Corneliuson, O.: Acid phosphatase activity at nodes of Ranvier in alpha-motor and dorsal root ganglion neurons of the cat. J. Neurocytol. 17:531, 1988. 151. Gatzinsky, K. P., Berthold, C. H., and Rydmark, M.: Axon-Schwann cell networks are regular components of nodal regions in normal large nerve fibres of cat spinal nerves. Neurosci. Lett. 124:264, 1991. 152. Gatzinsky, K. P., Holtmann, B. B., Daraie, B., et al.: Early onset of degenerative changes at nodes of Ranvier in alpha-motor axons of CNTF null (-/-) mutant mice. Glia 42:340, 2003. 153. Gatzinsky, K. P., Persson, G. H., and Berthold, C. H.: Removal of retrogradely transported material from rat lumbosacral alpha-motor axons by paranodal axon-Schwann cell networks. Glia 20:115, 1997. 154. Gerhardt, E.: Are Ranvier’s nodes artifacts by preparation? Acta Anat. 141:132, 1991. 155. Ghabriel, M. N., and Allt, G.: Incisures of SchmidtLanterman. Prog. Neurobiol. 17:25, 1981. 156. Ghabriel, M. N., and Allt, G.: The node of Ranvier. Prog. Anat. 2:137, 1982. 157. Ghadially, F. N.: Ultrastructural Pathology of the Cell and Matrix. London, Butterworths, p. 1252, 1988. 158. Giannini, C., and Dyck, P. J.: Axoglial dysjunction: a critical appraisal of definition, techniques, and previous results. Microsc. Res. Tech. 34:436, 1996. 159. Gilmore, S. A., and Duncan, D.: On the presence of peripheral-like nervous and connective tissue within irradiated spinal cord. Anat. Rec. 160:675, 1968. 160. Gilmore, S. A., and Sims, T. J.: The role of Schwann cells in the repair of glial cell deficits in the spinal cord. In Das, G. D., and Wallace, R. B. (eds.): Neural Transplantation and Regeneration. New York, Springer-Verlag, p. 245, 1986. 161. Gilmore, S. A., and Sims, T. J.: Glial-glial and glial-neuronal interfaces in radiation-induced, glia-depleted spinal cord. J. Anat. 190:5, 1997. 162. Ginsberg, L., Malik, O., Kenton, A. R., et al.: Co-existence of hereditary and inflammatory neuropathy. Brain 127:1, 2003. 163. Golding, J., Shewan, D., and Cohen, J.: Maturation of the mammalian dorsal root entry zone—from entry to no entry. Trends Neurosci. 20:303, 1997. 164. Golding, J. P., and Cohen, J.: Border controls at the mammalian spinal cord: late-surviving neural crest boundary

165.

166.

167.

168. 169. 170.

171.

172.

173.

174.

175.

176.

177.

178.

179.

180.

181.

182.

83

cap cells at dorsal root entry sites may regulate sensory afferent ingrowth and entry zone morphogenesis. Mol. Cell. Neurosci. 9:381, 1997. Golding, J. P., Shewan, D., Berry, M., and Cohen, J.: An in vitro model of the rat dorsal root entry zone reveals developmental changes in the extent of sensory axon growth into the spinal cord. Mol. Cell. Neurosci. 7:191, 1996. Gorio, A.: Sprouting and regeneration of peripheral nerve. In Zagoren, J. C., and Fedoroff, S. (eds.): The Node of Ranvier. New York, Academic Press, p. 353, 1984. Grafstein, B.: Axoplasmic transport: function and mechanisms. In Waxman, S. G., Kocsis, J. D., and Stys, P. K. (eds.): The Axon: Structure, Function and Pathophysiology. Oxford, UK, Oxford University Press, p. 185, 1995. Grafstein, B., and Forman, D. S.: Intracellular transport in neurons. Physiol. Rev. 60:1167, 1980. Gray, P. T. A., and Ritchie, J. M.: Ion channels in Schwann and glial-cells. Trends Neurosci. 8:411, 1985. Griffin, J. W., George, R., and Ho, T.: Macrophage systems in peripheral nerves: a review. J. Neuropathol. Exp. Neurol. 52:553, 1993. Gupta, S. K., Poduslo, J. F., Dunn, R., et al.: Myelin-associated glycoprotein gene expression in the presence and absence of Schwann cell-axonal contact. Dev. Neurosci. 12:22, 1990. Gupta, S. K., Poduslo, J. F., and Mezei, C.: Temporal changes in PO and MBP gene expression after crush-injury of the adult peripheral nerve. Brain Res. 464:133, 1988. Guthrie, S., and Lumsden, A.: Motor neuron pathfinding following rhombomere reversals in the chick embryo hindbrain. Development 114:663, 1992. Haak, R. A., Kleinhans, F. W., and Ochs, S.: The viscosity of mammalian nerve axoplasm measured by electron spin resonance. J. Physiol. (Lond.) 263:115, 1976. Hall, S., Hughes, R., and Atkinson, P.: Lamellated sensory corpuscles within the endoneurium. J. Neurol. Neurosurg. Psychiatry 54:744, 1991. Hall, S. M., and Williams, P. L.: The distribution of electrondense tracers in peripheral nerve fibres. J. Cell Sci. 8:541, 1971. Hallin, R. G., Ekedahl, R., and Frank, O.: Segregation by modality of myelinated and unmyelinated fibers in human sensory nerve fascicles. Muscle Nerve 14:157, 1991. Halter, J. A., and Clark, J. W. Jr.: The influence of nodal constriction on conduction velocity in myelinated nerve fibers. Neuroreport 4:89, 1993. Hammerschlag, R., and Stone, G. C.: Further studies on the initiation of fast axonal transport. In Smith, R. S., and Bisby, M. A. (eds.): Axonal Transport. New York, Alan R Liss, p. 37, 1987. Hannah, R. S., and Nathaniel, E. J.: The postnatal development of blood vessels in the substantia gelatinosa of rat cervical cord—an ultrastructural study. Anat. Rec. 178:691, 1974. Harrison, R. G.: Neuroblast versus sheath cell in the development of peripheral nerves. J. Comp. Neurol. 37:123, 1924. Haug, H.: The membrane limitans gliae superficialis of cat’s visual cortex [in German]. Z. Zellforsch. Mikrosk. Anat. 115:79, 1971.

84

Structure of the Peripheral Nervous System

183. Herbst, F.: Untersuchungen ueber Metallsalzreaktionen an den Ranvierschen Schnuerringen. Acta Histochem. 22:223, 1965. 184. Heriot, K., Gambetti, P., and Lasek, R. J.: Proteins transported in slow components a and b of axonal transport are distributed differently in the transverse plane of the axon. J. Cell Biol. 100:1167, 1985. 185. Hess, A., and Lansing, A. I.: The fine structure of peripheral nerve fibres. Anat. Rec. 117:175, 1953. 186. Hess, A., and Young, J. Z.: The nodes of Ranvier. Proc. R. Soc. Lond. B 140:301, 1952. 187. Hirano, A.: Nodes of Ranvier in pathological conditions. In Zagoren, J. C., and Fedoroff, S. (eds.): The Node of Ranvier. New York, Academic Press, p. 213, 1984. 188. Hirano, A., and Dembitzer, H. M.: Further studies on the transverse bands. J. Neurocytol. 11:861, 1982. 189. Hirokawa, N.: Cross-linker system between neurofilaments, microtubules, and membranous organelles in frog axons revealed by the quick-freeze, deep-etching method. J. Cell Biol. 94:129, 1982. 190. Hirokawa, N.: 270K microtubule-associated protein crossreacting with anti-MAP2 IgG in the crayfish peripheral nerve axon. J. Cell Biol. 103:33, 1986. 191. Hirokawa, N.: Molecular architecture and dynamics of the neuronal cytoskeleton. In Burgoyne, R. D. (ed.): The Neuronal Cytoskeleton. New York, Wiley-Liss, p. 5, 1991. 192. Hirokawa, N., Bloom, G. S., and Vallee, R. B.: Cytoskeletal architecture and immunocytochemical localization of microtubule-associated proteins in regions of axons associated with rapid axonal transport: the beta,beta⬘iminodipropionitrile-intoxicated axon as a model system. J. Cell Biol. 101:227, 1985. 193. Hoffman, P. N., Cleveland, D. W., Griffin, J. W., et al.: Neurofilament gene expression: a major determinant of axonal caliber. Proc. Natl. Acad. Sci. U. S. A. 84:3472, 1987. 194. Hoffman, P. N., Griffin, J. W., and Price, D. L.: Control of axonal caliber by neurofilament transport. J. Cell Biol. 99:705, 1984. 195. Hoffman, P. N., Thompson, G. W., Griffin, J. W., and Price, D. L.: Changes in neurofilament transport coincide temporally with alterations in the caliber of axons in regenerating motor fibers. J. Cell Biol. 101:1332, 1985. 196. Hoke, A., Ho, T., Crawford, T. O., et al.: Glial cell linederived neurotrophic factor alters axon Schwann cell units and promotes myelination in unmyelinated nerve fibers. J. Neurosci. 23:561, 2003. 197. Hollenbeck, P. J.: The transport and assembly of the axonal cytoskeleton. J. Cell Biol. 108:223, 1989. 198. Holtzman, E.: Membrane trafficking in neurons. Curr. Opin. Neurobiol. 2:607, 1992. 199. Hopkins, W. G., and Brown, M. C.: The distribution of nodal sprouts in a paralysed or partly denervated mouse muscle. Neuroscience 7:37, 1982. 200. Hopkins, W. G., Brown, M. C., and Keynes, R. J.: Nerve growth from nodes of Ranvier in inactive muscle. Brain Res. 222:125, 1981. 201. Hromada, J.: On the nerve supply of the connective tissue of some peripheral nervous system components. Acta Anat. 55:343, 1963.

202. Hulles, E.: Beiträge zur Kenntnis der sensiblen Wurzeln der Medulla oblongata beim Menschen. Arb. Neurol. Inst. Wien Univ. 7:392, 1900. 203. Huttner, W. B., and Dotti, C. G.: Exocytotic and endocytotic membrane traffic in neurons. Curr. Opin. Neurobiol. 1:388, 1991. 204. Ichimura, T., and Ellisman, M. H.: Three-dimensional fine structure of cytoskeletal-membrane interactions at nodes of Ranvier. J. Neurocytol. 20:667, 1991. 205. Jacobs, J. M.: Toxic effects on the node of Ranvier. In Zagoren, J. C., and Fedoroff, S. (eds.): The Node of Ranvier. New York, Academic Press, p. 245, 1984. 206. Jacobs, J. M.: Morphological changes at paranodes in IgM paraproteinaemic neuropathy. Microsc. Res. Tech. 34:544, 1996. 207. Jacobs, J. M., and Love, S.: Qualitative and quantitative morphology of human sural nerve at different ages. Brain 108:897, 1985. 208. Janetzko, A., Zimmermann, H., and Volknandt, W.: Intraneuronal distribution of a synaptic vesicle membrane protein: antibody binding sites at axonal membrane compartments and trans-Golgi network and accumulation at nodes of Ranvier. Neuroscience 32:65, 1989. 209. Jessen, K. R., and Mirsky, R.: Schwann cells and their precursors emerge as major regulators of nerve development. Trends Neurosci. 22:402, 1999. 210. Jessen, K. R., and Mirsky, R.: Signals that determine Schwann cell identity. J. Anat. 200:367, 2002. 211. Kaar, G. F. Development of the transitional region between central and peripheral nervous systems: a morphological and quantitative study of the L5 ventral root. Ph.D. Thesis, Dublin, National University of Ireland, 1984. 212. Kaar, G. F., and Fraher, J. P.: The sheaths surrounding the attachments of rat lumbar ventral roots to the spinal cord: a light and electron microscopical study. J. Anat. 148:137, 1986. 213. Kaar, G. F., and Fraher, J. P.: The vascularisation of the central-peripheral transitional zone of rat lumbar ventral rootlets: a morphological and morphometric study. J. Anat. 150:145, 1987. 214. Kanda, T., Tsukagoshi, H., Oda, M., et al.: Morphological changes in unmyelinated nerve fibres in the sural nerve with age. Brain 114:585, 1991. 215. Kenny, A. J., and Booth, A. G.: Microvilli: their ultrastructure, enzymology and molecular organization. Essays Biochem. 14:1, 1978. 216. Key, A., and Retzius, G.: Studien in der Anatomie des Nervensystems und des Bindegewebes. Stockholm, Samson & Wallin, 1876. 217. King, R. H. M., and Thomas, P. K.: Electron microscope observations on aberrant regeneration of unmyelinated axons in the vagus nerve of the rabbit. Acta Neuropathol. (Berl.) 18:150, 1971. 218. Kocsis, J. D.: Functional organization of potassium channels in normal and pathological mammalian axons. In Zagoren, J. C., and Fedoroff, S. (eds.): The Node of Ranvier. New York, Academic Press, p. 183, 1984. 219. Kocsis, J. D., and Waxman, S. G.: Ionic channel organization of normal and regenerating mammalian axons. Prog. Brain Res. 71:89, 1987.

Microscopic Anatomy of the Peripheral Nervous System 220. Kordeli, E., Davis, J., Trapp, B., Bennett, V.: An isoform of ankyrin is localized at nodes of Ranvier in myelinated axons of central and peripheral nerves. J. Cell Biol. 110:1341, 1990. 221. Kosik, K. S.: The molecular and cellular biology of tau. Brain Pathol. 3:39, 1993. 222. Krammer, E. B., and Lischka, M. F.: Schwermetallaffine Strukturen des peripheren Nerven. I. Potentieller Stürfaktor beim cytochemischen AChENachweis. Histochemie 36:269, 1973. 223. Kristensson, K., and Olsson, Y.: The perineurium as a diffusion barrier to protein tracers: differences between mature and immature animals. Acta Neuropathol. (Berl.) 17:127, 1971. 224. Kristol, C., Sandri, C., and Akert, K.: Intramembranous particles at the nodes of Ranvier of the cat spinal cord: a morphometric study. Brain Res. 142:391, 1978. 225. Krnjevic, K.: The distribution of Na and K in cat nerves. J. Physiol. (Lond.) 128:473, 1955. 226. Kuntz, A.: Experimental studies on the histogenesis of the sympathetic nervous system. J. Comp. Neurol. 34:1, 1922. 227. Laidlaw, G. F.: Silver staining of the endoneurial fibers of the cerebrospinal nerves. Am. J. Pathol. 6:435, 1930. 228. Landon, D. N.: Structure of normal peripheral myelinated nerve fibres. In Waxman, S. G., and Ritchie, J. M. (eds.): Demyelinating Disease: Basic and Clinical Electrophysiology. New York, Raven Press, p. 25, 1981. 229. Landon, D. N., and Hall, S.: The myelinated nerve fibre. In Landon, D. N. (ed.): The Peripheral Nerve. London, Chapman and Hall, p. 1, 1976. 230. Landon, D. N., and Langley, O. K.: The local chemical environment of nodes of Ranvier: a study of cation binding. J. Anat. 108:419, 1971. 231. Landon, D. N., and Williams, P. L.: Ultrastructure of the node of Ranvier. Nature 199:575, 1963. 232. Langford, G. M., Allen, R. D., and Weiss, D. G.: Substructure of sidearms on squid axoplasmic vesicles and microtubules visualized by negative contrast electron microscopy. Cell Motil. Cytoskeleton 7:20, 1987. 233. Langley, O. K.: A comparison of the binding of Alcian blue and inorganic cations to polyanions in peripheral nerve. Histochem. J. 3:251, 1971. 234. Langley, O. K., and Landon, D. N.: Copper binding at nodes of Ranvier: a new electron histochemical technique for the demonstration of polyanions. J. Histochem. Cytochem. 17:66, 1969. 235. Lasek, R. J.: Axonal transport: a dynamic view of neuronal structures. Trends Neurosci. 3:87, 1980. 236. Lasek, R. J.: Translocation of the neuronal cytoskeleton and axonal locomotion. Philos. Trans. R. Soc. Lond. B Biol. Sci. 299:313, 1982. 237. Lasek, R. J., Garner, J. A., and Brady, S. T.: Axonal transport of the cytoplasmic matrix. J. Cell Biol. 99:212s, 1984. 238. Le Beau, J. M., Powell, H. C., and Ellisman, M. H.: Node of Ranvier formation along fibres regenerating through silicone tube implants: a freeze-fracture and thin-section electron microscopic study. J. Neurocytol. 16:347, 1987. 239. Lemke, G., and Chao, M.: Axons regulate Schwann cell expression of the major myelin and NGF receptor genes. Development 102:499, 1988.

85

240. Levi, E.: Studien zur normalen und pathologischen Anatomie der hinteren Rückenmarkswurzeln. Arb. Neurol. Inst. Wien Univ. 13:62, 1906. 241. Levine, J., and Willard, M.: Fodrin: axonally transported polypeptides associated with the internal periphery of many cells. J. Cell Biol. 90:631, 1981. 242. Lev-Ram, V., and Ellisman, M. H.: Axonal activation-induced calcium transients in myelinating Schwann cells, sources, and mechanisms. J. Neurosci. 15:2628, 1995. 243. Lewis, S. A., Ivanov, I. E., Lee, G. H., Cowan, N. J: Organization of microtubules in dendrites and axons is determined by a short hydrophobic zipper in microtubuleassociated proteins MAP2 and tau. Nature 342:498, 1989. 244. Lieberman, E. M., Abbott, N. J., and Hassan, S.: Evidence that glutamate mediates axon-to-Schwann cell signaling in the squid. Glia 2:94, 1989. 245. Lilje, O.: The processing and presentation of endogenous and exogenous antigen by Schwann cells in vitro. Cell. Mol. Life Sci. 59:2191, 2002. 246. Lindsey, J. D., and Ellisman, M. H.: The neuronal endomembrane system. III. The origins of the axoplasmic reticulum and discrete axonal cisternae at the axon hillock. J. Neurosci. 5:3135, 1985. 247. Livesey, F. J., and Fraher, J. P.: Experimental traction injuries of cervical spinal nerve roots: a scanning EM study of rupture patterns in fresh tissue. Neuropathol. Appl. Neurobiol. 18:376, 1992. 248. Livingston, R. B., Pfenniger, K., Moor, H., Akert, K.: Specialized paranodal and interparanodal glial-axonal junctions in the peripheral and central nervous system: a freeze-etching study. Brain Res. 58:1, 1973. 249. LoPachin, R. M., Castiglia, C. M., Lehning, E., and Saubermann, A. J.: Effects of acrylamide on subcellular distribution of elements in rat sciatic nerve myelinated axons and Schwann cells. Brain Res. 608:238, 1993. 250. Lubinska, L., and Lukaszewska, I.: Shape of myelinated nerve fibres and proximodistal flow of axoplasm. Acta Biol. Exp. 17:115, 1956. 251. Lumsden, A., and Keynes, R.: Segmental patterns of neuronal development in the chick hindbrain. Nature 337:424, 1989. 252. Lundborg, G., and Brånemark, P.-I.: Microvascular structure and function of peripheral nerves. Adv. Microcirc. 1:66, 1968. 253. Luse, S. A.: Electron microscopic study of brain tumours. Neurology 10:881, 1960. 254. Lustig, M., Zanazzi, G., Sakurai, T., et al.: Nr-CAM and neurofascin interactions regulate ankyrin G and sodium channel clustering at the node of Ranvier. Curr. Biol. 11:1864, 2001. 255. Martin, S., Levine, A. K., Chen, Z. J., et al.: Deposition of the NG2 proteoglycan at nodes of Ranvier in the peripheral nervous system. J. Neurosci. 21:8119, 2001. 256. Martz, D., Lasek, R. J., Brady, S. T., Allen R. D.: Mitochondrial motility in axons: membranous organelles may interact with the force generating system through multiple surface binding sites. Cell. Motil. 4:89, 1984. 257. Mata, M., Kupina, N., and Fink, D. J.: Phosphorylationdependent neurofilament epitopes are reduced at the node of Ranvier. J. Neurocytol. 21:199, 1992. 258. Mata, M., Staple, J., and Fink, D. J.: Cytochemical localization of Ca2⫹-ATPase activity in peripheral nerve. Brain Res. 445:47, 1988.

86

Structure of the Peripheral Nervous System

259. Maxwell, D. S.: Fine structure of the normal trigeminal ganglion in the cat and monkey. J. Neurosurg. 26:127, 1967. 260. Maxwell, D. S., Kruger, L., and Pineda, A.: The trigeminal nerve root with special reference to the central-peripheral transition zone: an electron microscopic study in the macaque. Anat. Rec. 164:113, 1969. 261. McDonald, K.: Osmium ferricyanide fixation improves microfilament preservation and membrane visualization in a variety of animal cell types. J. Ultrastruct. Res. 86:107, 1984. 262. McIntyre, C. C., Richardson, A. G., and Grill, W. M.: Modeling the excitability of mammalian nerve fibers: influence of afterpotentials on the recovery cycle. J. Neurophysiol. 87:995, 2002. 263. McQuarrie, I. G.: Effect of conditioning lesion on axonal sprout formation at nodes of Ranvier. J. Comp. Neurol. 231:239, 1985. 264. Mecham, R. P., and Heuser, J. E.: The elastic fiber. In Day, E. D. (ed.): Cell Biology of the Extracellular Matrix. New York, Plenum Press, p. 79, 1991. 265. Melendez-Vasquez, C. V., Rios, J. C., Zanazzi, G., et al.: Nodes of Ranvier form in association with ezrin-radixin-moesin (ERM)-positive Schwann cell processes. Proc. Natl. Acad. Sci. U. S. A. 98:1235, 2001. 266. Meller, K.: Early structural changes in the axoplasmic cytoskeleton after axotomy studied by cryofixation. Cell Tissue Res. 250:663, 1987. 267. Menegoz, M., Gaspar, P., Le Bert, M., et al.: Paranodin, a glycoprotein of neuronal paranodal membranes. Neuron 19:319, 1997. 268. Metuzals, J., and Tasaki, I.: Subaxolemmal filamentous network in the giant nerve fiber of the squid (Loligo pealei L.) and its possible role in excitability. J. Cell Biol. 78:597, 1978. 269. Mi, H., Deerinck, T. J., Ellisman, M. H., Schwarz, T. L.: Differential distribution of closely related potassium channels in rat Schwann cells. J. Neurosci. 15:3761, 1995. 270. Midroni, G., Bilbao, J. M., and Cohen, S. M.: Normal anatomy of peripheral (sural) nerve. In Biopsy Diagnosis of Peripheral Neuropathy. Boston, Butterworth-Heinemann, p. 13, 1995. 271. Miller, C. C., Ackerley, S., Brownlees, J., et al.: Axonal transport of neurofilaments in normal and disease states. Cell. Mol. Life Sci. 59:323, 2002. 272. Miller, R. G., and da Silva, P. P.: Particle rosettes in the periaxonal Schwann cell membrane and particle clusters in the axolemma of rat sciatic nerve. Brain Res. 130:135, 1977. 273. Mohammed, U. H. M., and Landon, D. N.: Axoplasmic asymmetry at the node of Ranvier. J. Anat. 137:820, 1983. 274. Mohammed, U. H. M., Landon, D. N., and Love, S.: Ultrastructural morphometry of frog nodes of Ranvier. J. Anat. 137:822, 1983. 275. Molander, M., Berthold, C. H., Persson, H., et al.: Monosialoganglioside (GM1) immunofluorescence in rat spinal roots studied with a monoclonal antibody. J. Neurocytol. 26:101, 1997. 276. Moll, C., and Meier, C.: The central-peripheral transition zone of cervical spinal nerve roots in Jimpy mutant and normal mice: light- and electron-microscopic study. Acta Neuropathol. (Berl.) 60:241, 1983.

277. Moore, G. R., Boegman, R. J., Robertson, D. M., and Raine, C. S.: Acute stages of batrachotoxin-induced neuropathy: a morphologic study of a sodium-channel toxin. J. Neurocytol. 15:573, 1986. 278. Moya, F., Bunge, M. B., and Bunge, R. P.: Schwann cells proliferate but fail to differentiate in defined medium. Proc. Natl. Acad. Sci. U. S. A. 77:6902, 1980. 279. Muller-Mohnssen, H., Tippe, A., Hillenkamp, F., and Unsold, E.: Is the rise of the action potential of the Ranvier node controlled by a paranodal organ? Naturwissenschaften 61:369, 1974. 280. Myers, R. R., Heckman, H. M., and Powell, H. C.: Endoneurial fluid is hypertonic: results of microanalysis and its significance in neuropathy. J. Neuropathol. Exp. Neurol. 42:217, 1983. 281. Nageotte, J.: L’Organisation de la Matière. Paris, Felix Alcan, 1922. 282. Nakazawa, E., and Ishikawa, H.: Occurrence of fasciculated microtubules at nodes of Ranvier in rat spinal roots. J. Neurocytol. 24:399, 1995. 283. Nathaniel, E. J., and Nathaniel, D. R.: Electron microscopic observations on the dorsal root-spinal cord junction. J. Cell Biol. 19:52a, 1963. 284. Nathaniel, E. J. H., and Pease, D. C.: Collagen and basement membrane formation by Schwann cells during nerve regeneration. J. Ultrastruct. Res. 9:550, 1963. 285. Nemecek, S., Parizek, J., Spacek, J., and Nemeckova, J.: Histological, histochemical and ultrastructural appearance of the transitional zone of the cranial and spinal nerve roots. Folia Morphol. (Warsz.) 17:171, 1969. 286. Niederlander, C., and Lumsden, A.: Late emigrating neural crest cells migrate specifically to the exit points of cranial branchiomotor nerves. Development 122:2367, 1996. 287. Nilsson, I., and Berthold, C. H.: Axon classes and internodal growth in the ventral spinal root L7 of adult and developing cats. J. Anat. 156:71, 1988. 288. Nixon, R. A.: The regulation of neurofilament protein dynamics by phosphorylation: clues to neurofibrillary pathobiology. Brain Pathol. 3:29, 1993. 289. Nixon, R. A., and Sihag, R. K.: Neurofilament phosphorylation: a new look at regulation and function. Trends Neurosci. 14:501, 1991. 290. Obersteiner, H., and Redlich, E.: Uber Wesen und Pathogenese der Tabischen Hinterstrangsdegeneration. Arb. Neurol. Inst. Wien Univ. 1:158, 1894. 291. Oblinger, M. M., Brady, S. T., McQuarrie, I. G., and Lasek, R. J.: Cytotypic differences in the protein composition of the axonally transported cytoskeleton in mammalian neurons. J. Neurosci. 7:453, 1987. 292. Obremski, V. J., Wood, P. M., and Bunge, M. B.: Fibroblasts promote Schwann cell basal lamina deposition and elongation in the absence of neurons in culture. Dev. Biol. 160:119, 1993. 293. O’Brien, D., Dockery, P., McDermott, K., and Fraher, J. P.: The ventral motoneurone axon bundle in the CNS—a cordone system? J. Neurocytol. 27:247, 1998. 294. O’Brien, D., Dockery, P., McDermott, K., and Fraher, J. P.: Early development of rat ventral root transitional zone: an immunohistochemical and morphometric study. J. Neurocytol. 30:11, 2001.

Microscopic Anatomy of the Peripheral Nervous System 295. Obst, T.: Über das Endgebeit des Perineuriums an den Zahnnerven der Ratte. Z. Zellforsch. Mikrosk. Anat. 114:515, 1971. 296. Ochoa, J.: The sural nerve of the human foetus: electron microscope observations and counts of axons. J. Anat. 108:231, 1971. 297. Ochoa, J., and Mair, W. G.: The normal sural nerve in man. I. Ultrastructure and numbers of fibres and cells. Acta Neuropathol. (Berl.) 13:197, 1969. 298. Ochoa, J., and Mair, W. G.: The normal sural nerve in man. II. Changes in the axons and Schwann cells due to ageing. Acta Neuropathol. (Berl.) 13:217, 1969. 299. Ochoa, J., and Vial, J. D.: Behaviour of peripheral nerve structures in chronic neuropathies, with special reference to the Schwann cell. J. Anat. 102:95, 1967. 300. Ochs, S., Jersild, R. A. Jr., and Li, J. M.: Slow transport of freely movable cytoskeletal components shown by beading partition of nerve fibers in the cat. Neuroscience 33:421, 1989. 301. Okabe, S., and Hirokawa, N.: Microtubule dynamics in nerve cells: analysis using microinjection of biotinylated tubulin into PC12 cells. J. Cell Biol. 107:651, 1988. 302. Okabe, S., and Hirokawa, N.: Turnover of fluorescently labelled tubulin and actin in the axon. Nature 343:479, 1990. 303. Okabe, S., and Hirokawa, N.: Do photobleached fluorescent microtubules move? Re-evaluation of fluorescence laser photobleaching both in vitro and in growing Xenopus axon. J. Cell Biol. 120:1177, 1993. 304. Olsson, Y.: Vascular permeability in the peripheral nervous system. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 190–200, 1975. 305. Olsson, Y., and Reese, T. S.: Permeability of vasa nervorum and perineurium in mouse sciatic nerve studied by fluorescence and electron microscopy. J. Neuropathol. Exp. Neurol. 30:105, 1971. 306. Origuchi, Y.: Quantitative histological study in the sural nerves of children. Brain. Dev. 3:395, 1981. 307. Orkand, R. K., Nicholls, J. G., and Kuffler, S. W.: Effect of nerve impulses on the membrane potential of glial cells in the central nervous system of Amphibia. J. Neurophysiol. 29:788, 1966. 308. Paintal, A. S.: A comparison of the nerve impulses of mammalian non-medullated nerve fibres with those of the smallest diameter medullated fibres. J. Physiol. (Lond.) 193:523, 1967. 309. Pannese, E., Procacci, P., Ledda, M., et al.: A quantitative study of microtubules in motor and sensory axons. Acta Anat. (Basel) 118:193, 1984. 310. Pannese, E., Procacci, P., Ledda, M., et al.: Internodal microvilli of Schwann cells of myelinated fibres in lizard spinal roots project onto unmyelinated axons. J. Neurocytol. 18:295, 1989. 311. Parhad, I. M., Clark, A. W., and Griffin, J. W.: The effect of impairment of slow transport on axonal caliber. In Smith, R. S., and Bisby, M. A. (eds.): Axonal Transport. New York, Alan R Liss, p. 263, 1987. 312. Parmantier, E., Lynn, B., Lawson, D., et al.: Schwann cell-derived desert hedgehog controls the development of peripheral nerve sheaths. Neuron 23:713, 1999.

87

313. Pascher, R., Berthold, C. H., and Rydmark, M.: Computerassisted simulation of high-voltage electron microscopy using serial images recorded by conventional transmission electron microscopy. J. Electron Microsc. (Tokyo) 51:113, 2002. 314. Pedraza, L., Huang, J. K., and Colman, D. R.: Organizing principles of the axoglial apparatus. Neuron 30:335, 2001. 315. Peltonen, J., Jaakkola, S., Virtanen, I., and Pelliniemi, L.: Perineurial cells in culture: an immunocytochemical and electron microscopic study. Lab. Invest. 57:480, 1987. 316. Peters, A., Palay, S. L., Webster, H. deF.: The Fine Structure of the Nervous System: The Neurons and Supporting Cells. Philadelphia, W. B. Saunders, p. 233, 1976. 317. Phelps, C. H.: The development of glio-vascular relationships in the rat spinal cord: an electron microscopic study. Z. Zellforsch. Mikrosk. Anat. 128:555, 1972. 318. Phillips, D. D., Hibbs, R. G., Ellison, J. P., and Shapiro, H.: An electron microscopic study of central and peripheral nodes of Ranvier. J. Anat. 111:229, 1972. 319. Plenk, H.: Über argyrophile Fasern (Gitterfasern) und ihre Bildungszellen. Ergeb. Anat. Entwicklungsgesch. 27:302, 1927. 320. Plenk, H.: Der Schwannsche Scheide der markhaltigen Nervenfasern. Z. Mikrosk. Anat. Forsch. 36:191, 1934. 321. Poliak, S., Gollan, L., Salomon, D., et al.: Localization of Caspr2 in myelinated nerves depends on axon-glia interactions and the generation of barriers along the axon. J. Neurosci. 21:7568, 2001. 322. Poliak, S., and Peles, E.: The local differentiation of myelinated axons at nodes of Ranvier. Nat. Rev. Neurosci. 4:968, 2003. 323. Price, D. L., and Griffin, J. W.: Neurons and ensheathing cells as targets of disease processes. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicity. Baltimore, Williams & Wilkins, p. 1, 1980. 324. Quarles, R. H., Hammer, J. H., and Trapp, B. D.: The immunoglobulin gene superfamily and myelination. In Dynamic Interactions of Myelin Proteins. New York, Alan R Liss, p. 49, 1990. 325. Quatacker, J.: The axonal reticulum in the neurons of the superior cervical ganglion of the rat as a direct extension of the Golgi apparatus. Histochem. J. 13:109, 1981. 326. Raine, C. S.: On the occurrence of Schwann cells within the normal central nervous system. J. Neurocytol. 5:371, 1976. 327. Raine, C. S.: Differences between the nodes of Ranvier of large and small diameter fibres in the P.N.S. J. Neurocytol. 11:935, 1982. 328. Raine, C. S., Finch, H., and Masone, A.: Axoplasmic asymmetry at the node of Ranvier. J. Neurocytol. 12:533, 1983. 329. Raisman, G.: Specialized neuroglial arrangement may explain the capacity of vomeronasal axons to reinnervate central neurons. Neuroscience 14:237, 1985. 330. Rambourg, A., Clermont, Y., and Beaudet, A.: Ultrastructural features of six types of neurons in rat dorsal root ganglia. J. Neurocytol. 12:47, 1983. 331. Rambourg, A., and Droz, B.: Smooth endoplasmic reticulum and axonal transport. J. Neurochem. 35:16, 1980. 332. Ranvier, M. L.: Leçons sur l’Histologie du Système Nerveux. Paris, Librairie F. Savy, 1878.

88

Structure of the Peripheral Nervous System

333. Rasband, M. N., Trimmer, J. S., Schwarz, T. L., et al.: Potassium channel distribution, clustering, and function in remyelinating rat axons. J. Neurosci. 18:36, 1998. 334. Raven, C. P.: Experiments on the origin of sheath cells and sympathetic neuroblasts in Amphibia. J. Comp. Neurol. 67:221, 1937. 335. Rayner, D. A., and Baines, A. J.: A novel component of the axonal cortical cytoskeleton, A60, defined by a monoclonal antibody. J. Cell Sci. 94(Pt. 3):489, 1989. 336. Reale, E., Luciano, L., and Spitznas, M.: Freeze-fracture faces of the perineurial sheath of the rabbit sciatic nerve. J. Neurocytol. 4:261, 1975. 337. Reale, E., Luciano, L., and Spitznas, M.: Freeze-fracture aspects of the perineurium of spinal ganglia. J. Neurocytol. 5:385, 1976. 338. Remak, R.: Observationes Anatomicae et Microscopicae de Systematis Nervosi Structura. Berlin, Sumptibus et Formis Reimerianis, 1838. 339. Reynolds, R. J., and Heath, J. W.: Patterns of morphological variation within myelin internodes of normal peripheral nerve: quantitative analysis by confocal microscopy. J. Anat. 187:369, 1995. 340. Rieger, F., Daniloff, J. K., Pincon-Raymond, M., et al.: Neuronal cell adhesion molecules and cytotactin are colocalized at the node of Ranvier. J. Cell Biol. 103:379, 1986. 341. Ritchie, J. M., and Chiu, S. Y.: Distribution of sodium and potassium channels in mammalian myelinated nerve. In Waxman, S. G., and Ritchie, J. M. (eds.): Demyelinating Disease: Basic and Clinical Electrophysiology. New York, Raven Press, p. 329, 1981. 342. Rogart, R.: Structural and functional relationships of ion conduction in the myelinated and demyelinated axon. In Zagoren, J. C., and Fedoroff, S. (eds.): The Node of Ranvier. New York, Academic Press, p. 109, 1984. 343. Roper, J., and Schwarz, J. R.: Heterogeneous distribution of fast and slow potassium channels in myelinated rat nerve fibres. J. Physiol. (Lond.) 416:93, 1989. 344. Rosenbluth, J.: Intramembranous particle distribution at the node of Ranvier and adjacent axolemma in myelinated axons of the frog brain. J. Neurocytol. 5:731, 1976. 345. Rosenbluth, J.: Membrane specializations at the nodes of Ranvier and paranodal and juxtaparanodal regions of myelinated central and peripheral nerve fibers. In Zagoren, J. C., and Fedoroff, S. (eds.): The Node of Ranvier. New York, Academic Press, p. 31, 1984. 346. Rosenbluth, J.: Role of glial cells in the differentiation and function of myelinated axons. Int. J. Dev. Neurosci. 6:3, 1988. 347. Ross, M. D., and Burkel, W.: Electron microscopic observations of the nucleus, glial dome, and meninges of the rat acoustic nerve. Am. J. Anat. 130:73, 1971. 348. Ross, M. H., and Reith, E. J.: Perineurium: evidence for contractile elements. Science 165:604, 1965. 349. Rossiter, J. P., and Fraher, J. P.: Intermingling of central and peripheral nervous tissues in rat dorsolateral vagal rootlet transitional zones. J. Neurocytol. 19:385, 1990. 350. Rydmark, M.: Nodal axon diameter correlates linearly with internodal axon diameter in spinal roots of the cat. Neurosci. Lett. 24:247, 1981. 351. Rydmark, M., and Berthold, C. H.: Electron microscopic serial section analysis of nodes of Ranvier in lumbar spinal

352.

353.

354.

355.

356.

357. 358.

359.

360.

361. 362.

363.

364.

365.

366. 367.

368.

369.

roots of the cat: a morphometric study of nodal compartments in fibres of different sizes. J. Neurocytol. 12:537, 1983. Rydmark, M., Berthold, C. H., and Gatzinsky, K. P.: Paranodal Schwann cell mitochondria in spinal roots of the cat: an ultrastructural morphometric analysis. J. Neurocytol. 27:99, 1998. Saito, A., and Zachs, S. I.: Ultrastructure of Schwann and perineurial sheaths at the mouse neuromuscular junction. Anat. Rec. 164:379, 1969. Saitua, F., and Alvarez, J.: Microtubular packing varies along the course of motor and sensory axons: possible regulation of microtubules by environmental cues. Neurosci. Lett. 104:249, 1989. Salzer J. L.: Mechanisms of adhesion between axons and glial cells. In Waxman, S. G., Kocsis, J. D., and Stys, P. K. (eds.): The Axon: Structure, Function and Pathophysiology. Oxford, UK, Oxford University Press, p. 164, 1995. Salzer, J. L.: Clustering sodium channels at the node of Ranvier: close encounters of the axon-glia kind. Neuron 18:843, 1997. Salzer, J. L.: Nodes of Ranvier come of age. Trends Neurosci. 25:2, 2002. Samorajski, T., and Friede, R. L.: A quantitative electron microscopic study of myelination in the pyramidal tract of rat. J. Comp. Neurol. 134:323, 1968. Sawada, H., Furthmayr, H., Konomi, H., and Nagai, Y.: Immunoelectronmicroscopic localization of extracellular matrix components produced by bovine corneal endothelial cells in vitro. Exp. Cell Res. 171:94, 1987. Scherer, S. S.: Molecular specializations at nodes and paranodes in peripheral nerve. Microsc. Res. Tech. 34:452, 1996. Scherer, S. S.: Nodes, paranodes, and incisures: from form to function. Ann. N. Y. Acad. Sci. 883:131, 1999. Scherer, S. S., and Arroyo, E. J.: Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Syst. 7:1, 2002. Scherer, S. S., Xu, T., Crino, P., et al.: Ezrin, radixin, and moesin are components of Schwann cell microvilli. J. Neurosci. Res. 65:150, 2001. Schlaepfer, W. W., Freeman, L. A., and Eng, L. F.: Studies of human and bovine spinal nerve roots and the outgrowth of CNS tissues into the nerve root entry zone. Brain Res. 177:219, 1979. Schnapp, B., and Mugnaini, E.: Membrane architecture of myelinated fibers as seen by freeze-fracture. In Waxman, S. G. (eds.): Physiology and Pathobiology of Axons. New York, Raven Press, p. 83, 1978. Schnapp, B. J., and Reese, T. S.: Cytoplasmic structure in rapid-frozen axons. J. Cell Biol. 94:667, 1982. Schroder, J. M.: Developmental and pathological changes at the node and paranode in human sural nerves. Microsc. Res. Tech. 34:422, 1996. Schroder, J. M., Bohl, J., and Brodda, K.: Changes of the ratio between myelin thickness and axon diameter in the human developing sural nerve. Acta Neuropathol. (Berl.) 43:169, 1978. Schwann, T.: Mikroskopische Untersuchungen über die Übereinstimmung in der Struktur und dem Wachstum der Tiere und Pflanzen. Berlin, Sander, 1839.

Microscopic Anatomy of the Peripheral Nervous System 370. Schwartz, M., Ernst, S. A., Siegel, G. J., Agranoff, B. W.: Immunocytochemical localization of (Na⫹, K⫹)-ATPase in the goldfish optic nerve. J. Neurochem. 36:107, 1981. 371. Shantha, T. R., and Bourne, G. H.: The perineurial epithelium—a new concept. In The Structure and Function of Nervous Tissue. New York, Academic Press, p. 379, 1968. 372. Shanthaveerappa, T. R., and Bourne, G. H.: ‘The perineurial epithelium,’ a metabolically active, continuous, protoplasmic cell barrier surrounding peripheral nerve fasciculi. J. Anat. 96:527, 1962. 373. Shanthaveerappa, T. R., and Bourne, G. H.: Perineurial epithelium: a new concept of its role in the integrity of the peripheral nervous system. Science 153:1464, 1966. 374. Sheetz, M. P., and Martenson, C. H.: Axonal transport: beyond kinesin and cytoplasmic dynein. Curr. Opin. Neurobiol. 1:393, 1991. 375. Shellswell, G. B., Restall, D. J., Duance, V. C., and Bailey, A. J.: Identification and differential distribution of collagen types in the central and peripheral nervous systems. FEBS Lett. 106:305, 1979. 376. Shield, L. K., King, R. H. M., and Thomas, P. K.: A morphometric study of human fetal sural nerve. Acta Neuropathol. (Berl.) 70:60, 1986. 377. Shimazono, J.: Über das Verhalten der Zentralen und der peripheren Nervensubstanz bei verschiedenen Vergiftungen und Ernährungsstörungen. Arch. Psychiatr. Nervenkr. 53:872, 1914. 378. Shinogami, H.: Cytochemical-localization of K⫹-Nppase in the heminode of Ranvier in the rat optic-nerve. Acta Histochem. Cytochem. 22:207, 1989. 379. Sievers, J.: Basic two-dye stains for epoxy-embedded 0.3-1␮ sections. Stain Technol. 46:195, 1971. 380. Sima, A., and Sourander, P.: The effect of perinatal undernutrition on perineurial diffusion barrier to exogenous protein: an experimental study on rat sciatic nerve. Acta Neuropathol. (Berl.) 24:263, 1973. 381. Sima, A. A., and Brismar, T.: Reversible diabetic nerve dysfunction: structural correlates to electrophysiological abnormalities. Ann. Neurol. 18:21, 1985. 382. Sims, T. J., and Gilmore, S. A.: Interactions between intraspinal Schwann cells and the cellular constituents normally occurring in the spinal cord: an ultrastructural study in the irradiated rat. Brain Res. 276:17, 1983. 383. Sims, T. J., and Gilmore, S. A.: Interactions between Schwann cells and CNS axons following a delay in the normal formation of central myelin. Exp. Brain Res. 75:513, 1989. 384. Sims, T. J., and Gilmore, S. A.: Regeneration of dorsal root axons into experimentally altered glial environments in the rat spinal cord. Exp. Brain Res. 99:25, 1994. 385. Sims, T. J., and Gilmore, S. A.: Regrowth of dorsal root axons into a radiation-induced glial-deficient environment in the spinal cord. Brain Res. 634:113, 1994. 386. Sims, T. J., Gilmore, S. A., Waxman, S. G., and Klinge, E.: Dorsal-ventral differences in the glia limitans of the spinal cord: an ultrastructural study in developing normal and irradiated rats. J. Neuropathol. Exp. Neurol. 44:415, 1985.

89

387. Sindou, M., Quoex, C., and Baleydier, C.: Fiber organization at the posterior spinal cord-rootlet junction in man. J. Comp. Neurol. 153:15, 1974. 388. Skinner, H. A.: Some histologic features of cranial nerves. Arch. Neurol. Psychiatry 25:356, 1931. 389. Small, J. R., Scadding, J. W., and Landon, D. N.: Ultrastructural localization of sympathetic axons in experimental rat sciatic nerve neuromas. J. Neurocytol. 25:573, 1996. 390. Smith, R. S.: The short term accumulation of axonally transported organelles in the region of localized lesions of single myelinated axons. J. Neurocytol. 9:39, 1980. 391. Smith, R. S.: Control of the direction of rapid axonal transport in the vertebrates. In Smith, R. S., and Bisby, M. A. (eds.): Axonal Transport. New York, Alan R Liss Inc, p. 139, 1987. 392. Smith, R. S.: Real-time imaging of axonally transported subresolution organelles in vertebrate myelinated axons. J. Neurosci. Methods 26:203, 1989. 393. Smith, R. S., and Forman, D. S.: Organelle dynamics in lobster axons: anterograde and retrograde particulate organelles. Brain Res. 446:26, 1988. 394. Snyder, R.: The organization of the dorsal root entry zone in cats and monkeys. J. Comp. Neurol. 174:47, 1977. 395. Sonnenberg-Riethmacher, E., Miehe, M., Stolt, C. C., et al.: Development and degeneration of dorsal root ganglia in the absence of the HMG-domain transcription factor Sox10. Mech. Dev. 109:253, 2001. 396. Spacek, J., and Lieberman, A. R.: Relationships between mitochondrial outer membranes and agranular reticulum in nervous tissue: ultrastructural observations and a new interpretation. J. Cell Sci. 46:129, 1980. 397. Spencer, P. S.: Reappraisal of the model for “bulk axoplasmic flow.” Nat. New Biol. 240:283, 1972. 398. Spencer, P. S., and Thomas, P. K.: Ultrastructural studies of the dying-back process. II. The sequestration and removal by Schwann cells and oligodendrocytes of organelles from normal and diseased axons. J. Neurocytol. 3:763, 1974. 399. Spencer, P. S., Weinberg, H. J., Raine, C. S., and Prineas, J. W.: The perineurial window—a new model of focal demyelination and remyelination. Brain Res. 96:323, 1975. 400. Spray, D. C., and Dermietzel, R.: X-linked dominant CharcotMarie-Tooth disease and other potential gap-junction diseases of the nervous system. J. Comp. Neurol. 18:256, 1995. 401. Steer, J. M.: Some observations on the fine structure of rat dorsal spinal nerve roots. J. Anat. 109:467, 1971. 402. Stelzner, D. J.: The relationship between synaptic vesicles, Golgi apparatus, and smooth endoplasmic reticulum: a developmental study using the zinc iodide-osmium technique. Z. Zellforsch. Mikrosk. Anat. 120:332, 1971. 403. Stolinski, C.: Structure and composition of the outer connective tissue sheaths of peripheral nerve. J. Anat. 186:123, 1995. 404. Stolinski, C., and Breathnach, A. S.: Freeze-fracture replication of mammalian peripheral nerve—a review. J. Neurol. Sci. 57:1, 1982. 405. Stolinski, C., Breathnach, A. S., Martin, B., et al.: Associated particle aggregates in juxtaparanodal axolemma and adaxonal Schwann cell membrane of rat peripheral nerve. J. Neurocytol. 10:679, 1981.

90

Structure of the Peripheral Nervous System

406. Stone, G. C., and Hammerschlag, R.: Molecular mechanisms involved in sorting of fast-transported proteins. In Smith, R. S., and Bisby, M. A. (eds.): Axonal Transport. New York, Alan R Liss, p. 15, 1987. 407. Strupp, M., and Grafe, P.: A chloride channel in rat and human axons. Neurosci. Lett. 133:237, 1991. 408. Sturrock, R. R.: A quantitative and morphological study of vascularisation of the developing mouse spinal cord. J. Anat. 132:203, 1981. 409. Suarez, I., and Fernandez, B.: Structure and ultrastructure of the external glial layer in the hypothalamus of the hamster. J. Hirnforsch. 24:99, 1983. 410. Suarez, I., and Raff, M. C.: Subpial and perivascular astrocytes associated with nodes of Ranvier in the rat optic-nerve. J. Neurocytol. 18:577, 1989. 411. Sugimoto, Y., Takayama, S., Horiuchi, Y., Toyama, Y.: An experimental study on the perineurial window. J. Peripher. Nerv. Syst. 7:104, 2002. 412. Sunderland, S.: The connective tissues of peripheral nerves. Brain 88:841, 1965. 413. Sunderland, S.: Nerve and Nerve Injuries. Edinburgh, Churchill Livingstone, 1978. 414. Suter, U., and Scherer, S.: Disease mechanisms in inherited neuropathies. Nat. Rev. Neurosci. 4:714, 2003. 415. Tarlov, I. M.: Structure of the nerve root. I. Nature of the junction between the central and the peripheral nervous system. Arch. Neurol. Psychiatry 37:555, 1937. 416. Tarlov, I. M.: Structure of the nerve root. II. Differentiation of sensory from motor roots: observations on identification of function in roots of mixed cranial nerves. Arch. Neurol. Psychiatry 37:1338, 1937. 417. Tashiro, T., and Komiya, Y.: Maturation and aging of the axonal cytoskeleton: biochemical analysis of transported tubulin. J. Neurosci. Res. 30:192, 1991. 418. Thomas, F. P.: Antibodies to GM1 and Gal(beta 1–3)GalNAc at the nodes of Ranvier in human and experimental autoimmune neuropathy. Microsc. Res. Tech. 34:536, 1996. 419. Thomas, P. K.: The connective tissue of peripheral nerve: an electron microscope study. J. Anat. 97:35, 1963. 420. Thomas, P. K., Beamish, N. G., Small, J. R., et al.: Paranodal structure in diabetic sensory polyneuropathy. Acta Neuropathol. (Berl.) 92:614, 1996. 421. Thomas, P. K., and Bhagat, S.: The effect of extraction of the intrafascicular contents of peripheral nerve trunks on perineurial structure. Acta Neuropathol. (Berl.) 43:135, 1978. 422. Thomas, P. K., and Jones, D. G.: The cellular response to nerve injury. II. Regeneration of the perineurium after nerve section. J. Anat. 101:45, 1967. 423. Thomas, P. K., and King, R. H. M.: The degeneration of unmyelinated axons following nerve section: an ultrastructural study. J. Neurocytol. 3:497, 1974. 424. Thomas, P. K., King, R. H. M., and Sharma, A. K.: Changes with age in the peripheral nerves of the rat. Acta Neuropathol. (Berl.) 52:1, 1980. 425. Thomas, P. K., Ochoa, J. L., Berthold, C. H., et al.: Microscopic anatomy of the peripheral nervous system. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 28, 1993. 426. Tippe, A., and Muller-Mohnssen, H.: Further experimental evidence for the synapse hypothesis of Na⫹-current activation

427.

428.

429.

430.

431.

432.

433.

434.

435.

436.

437.

438.

439.

440. 441.

442.

443.

and inactivation at the Ranvier node. Naturwissenschaften 62:490, 1975. Todd, B. A., Inman, C., Sedgwick, E. M., and Abbott, N. J.: Ionic permeability of the frog sciatic nerve perineurium: parallel studies of potassium and lanthanum penetration using electrophysiological and electron microscopic techniques. J. Neurocytol. 29:551, 2000. Toews, A. D., Armstrong, R., Holshek, J., et al.: Unloading and transfer of axonally transported lipids. In Smith, R. S., and Bisby, M. A. (eds.): Axonal Transport. New York, Alan R Liss, p. 327, 1987. Trapp, B. D., Hauer, P., and Lemke, G.: Axonal regulation of myelin protein mRNA levels in actively myelinating Schwann cells. J. Neurosci. 8:3515, 1988. Trapp, B. D., and Kidd, G. J.: Axo-glial septate junctions: the maestro of nodal formation and myelination? J. Cell Biol. 150:F97, 2000. Tsukita, S., and Ishikawa, H.: Three-dimensional distribution of smooth endoplasmic reticulum in myelinated axons. J. Electron Microsc. (Tokyo) 25:141, 1976. Tsukita, S., and Ishikawa, H.: The movement of membranous organelles in axons: electron microscopic identification of anterogradely and retrogradely transported organelles. J. Cell Biol. 84:513, 1980. Tsukita, S., and Ishikawa, H.: The cytoskeleton in myelinated axons—serial section study. Biomed. Res. (Tokyo) 2:424, 1981. Tsukita, S., Tsukita, S., Kobayashi, T., and Matsumoto, G.: Subaxolemmal cytoskeleton in squid giant axon. II. Morphological identification of microtubule- and microfilamentassociated domains of axolemma. J. Cell Biol. 102:1710, 1986. Tsukita, S., Usukura, J., Tsukita, S., and Ishikawa, H.: The cytoskeleton in myelinated axons: a freeze-etch replica study. Neuroscience 7:2135, 1982. Tuckett, I. L.: On the structure and degeneration of nonmyelinated nerve fibres. J. Physiol. (Lond.) 19:267, 1895. Uehara, M., and Ueshima, T.: Scanning electron microscopy of the superficial glial limiting membrane in the cat brain and spinal cord. Nippon Juigaku Zasshi 50:115, 1988. Uhrik, B., and Stampfli, R.: Ultrastructural observations on nodes of Ranvier from isolated single frog peripheral nerve fibres. Brain Res. 215:93, 1981. Ushiki, T., and Ide, C.: Three-dimensional architecture of the endoneurium with special reference to the collagen fibril arrangement in relation to nerve fibers. Arch. Histol. Jpn. 49:553, 1986. van Deurs, B., Petersen, O. W., Olsnes, S., and Sandvig, K.: The ways of endocytosis. Int. Rev. Cytol. 117:131, 1989. Vasilescu, V., and Filip, D. A.: The Ranvier node as a chemo-electric pulsatory unit: a study of its structurefunctions relations. Physiologie 16:83, 1979. Vega, J. A., Del Valle, M. M., Haro, J. J., et al.: The inner-core, outer-core and capsule cells of the human pacinian corpuscles: an immunohistochemical study. Eur. J. Morphol. 32:11, 1994. Virchow, R.: Rückenmark und Gehirn. In Die Cellularpathologie in ihrer Begrundung auf Physiologische und Pathologische Gewebelehre. Berlin, Hirschwald, p. 238, 1858.

Microscopic Anatomy of the Peripheral Nervous System 444. Voyvodic, J. T.: Target size regulates calibre and myelination of sympathetic axons. Nature 342:430, 1989. 445. Wagner, H. J., Barthel, J., and Pilgrim, C.: Permeability of the external glial limiting membrane of rat parietal cortex. Anat. Embryol. (Berl.) 166:427, 1983. 446. Wang, L., and Brown, A.: Rapid movement of microtubules in axons. Curr. Biol. 12:1496, 2002. 447. Watson, D.: Regional variation in the abundance of axonal cytoskeletal proteins. J. Neurosci. Res. 30:226, 1991. 448. Waxman, S. G.: Regional differentiation of the axon: a review with special reference to the concept of the multiplex neuron. Brain Res. 47:269, 1972. 449. Waxman, S. G.: Voltage-gated ion channels in axons: localization, function, and development. In Waxman, S. G., Kocsis, J. D., and Stys, P. K. (eds.): The Axon: Structure, Function and Pathophysiology. Oxford, UK, Oxford University Press, p. 218, 1995. 450. Waxman, S. G., and Black, J. A.: Macromolecular structure of the Schwann cell membrane: perinodal microvilli. J. Neurol. Sci. 77:23, 1987. 451. Waxman, S. G., and Quick, D. C.: Intra-axonal ferric ion-ferrocyanide staining of nodes of Ranvier and initial segments in central myelinated fibers. Brain Res. 144:1, 1978. 452. Waxman, S. G., and Ritchie, J. M.: Molecular dissection of the myelinated axon. Ann. Neurol. 33:121, 1993. 453. Webster, H. deF.: Transient, focal accumulation of axonal mitochondria during the early stages of wallerian degeneration. J. Cell Biol. 12:361, 1962. 454. Webster, H. deF., and Spiro, D.: Phase and electron microscopic studies of experimental demyelination. I. Variations in myelin sheath contour in normal guinea pig sciatic nerve. J. Neuropathol. Exp. Neurol. 19:42, 1960. 455. Weis, J., Alexianu, M. E., Heide, G., and Schröder, J. M.: Renaut bodies contain elastic fibre components. J. Neuropathol. Exp. Neurol. 52:444, 1993. 456. Weiss, D. G., Seitz-Tutter, D., Langford, G. M., et al.: The native microtubule as the engine for bidirectional organelle movements. In Smith, R. S., and Bisby, M. A. (eds.): Axonal Transport. New York, Alan R Liss, p. 91, 1987. 457. Weston, J. A.: A radioautographic analysis of the migration and localisation of trunk neural crest cells in the chick. Dev. Biol. 6:279, 1963. 458. Weston, J. A.: The migration and differentiation of neural crest cells. In Abercrombie, M., Brachet, J., and King, T. (eds.): Advances in Morphogenesis. New York, Academic Press, p. 127, 1970. 459. Wiley, C. A., and Ellisman, M. H.: Rows of dimericparticles within the axolemma and juxtaposed particles within glia, incorporated into a new model for the paranodal glial-axonal junction at the node of Ranvier. J. Cell Biol. 84:261, 1980. 460. Williams, P. L., and Hall, S. M.: In vivo observations on mature myelinated nerve fibres of the mouse. J. Anat. 107:31, 1970. 461. Williams, P. L., and Hall, S. M.: Prolonged in vivo observations of normal peripheral nerve fibres and their acute

462. 463.

464.

465.

466.

467.

468.

469.

470.

471.

472.

473.

474.

475.

476.

477.

478.

91

reactions to crush and deliberate trauma. J. Anat. 108:397, 1971. Williams, P. L., and Kashef, R.: Asymmetry of the node of Ranvier. J. Cell Sci. 3:341, 1968. Williams, P. L., and Landon, D. N.: Paranodal apparatus of peripheral nerve fibres of mammals. Nature 198:670, 1963. Williams, V.: Intercellular relationships in the external glial limiting membrane of the neocortex of the cat and rat. Am. J. Anat. 144:421, 1975. Windebank, A. J., Wood, P., Bunge, R. P., and Dyck, P. J.: Myelination determines the caliber of dorsal root ganglion neurons in culture. J. Neurosci. 5:1563, 1985. Wood, J. G., Jean, D. H., Whitaker, J. N., et al.: Immunocytochemical localization of the sodium, potassium activated ATPase in knifefish brain. J. Neurocytol. 6:571, 1977. Wulfhekel, U.: The passage from central to peripheral structure of cranial nerves in the lower vertebrates [in German]. Acta Anat. (Basel) 74:183, 1969. Wurtz, C. C., and Ellisman, M. H.: Alterations in the ultrastructure of peripheral nodes of Ranvier associated with repetitive action potential propagation. J. Neurosci. 6:3133, 1986. Wurtz, C. C., and Ellisman, M. H.: Activity associated ultrastructural changes in peripheral nodes of Ranvier are independent of fixation. Exp. Neurol. 101:87, 1988. Yamamoto, K., Merry, A. C., and Sima, A. A.: An orderly development of paranodal axoglial junctions and bracelets of Nageotte in the rat sural nerve. Brain Res. Dev. Brain Res. 96:36, 1996. Zagoren, J. C.: Cation binding at the node of Ranvier. In Zagoren, J. C., and Fedoroff, S. (eds.): The Node of Ranvier. New York, Academic Press, p. 69, 1984. Zagoren, J. C., and Arezzo, J. C.: Cation binding at the node of Ranvier: II. Redistribution of binding sites during electrical stimulation. Brain Res. 242:27, 1982. Zagoren, J. C., Raine, C. S., and Suzuki, K.: Cation binding at the node of Ranvier: I. Localization of binding sites during development. Brain Res. 242:19, 1982. Zelena, J.: Arrays of glycogen granules in the axoplasm of peripheral nerves at pre-ovoid stages of wallerian degeneration. Acta Neuropathol. (Berl.) 50:227, 1980. Zenker, W., and Hohberg, E.: A-alpha-nerve-fiber: number of neurotubules in the stem fibre and in the terminal branches. J. Neurocytol. 2:143, 1973. Zenker, W., Mayr, R., and Gruber, H.: Axoplasmic organelles: quantitative differences between ventral and dorsal root fibres of the rat. Experientia 29:77, 1973. Zenker, W., Mayr, R., and Gruber, H.: Neurotubules: different densities in peripheral motor and sensory nerve fibres. Experientia 31:318, 1975. Zimmermann, H., and Vogt, M.: Membrane proteins of synaptic vesicles and cytoskeletal specializations at the node of Ranvier in electric ray and rat. Cell Tissue Res. 258:617, 1989.

4 Channel Function in Mammalian Axons and Support Cells S. Y. CHIU

Na⫹ Channels Na⫹ Channels on Axons Molecular Identity of Nav Channels on Axons Mechanisms for Nav Channel Segregation

Na⫹ Channels on Schwann Cells K⫹ Channels Molecular Identity of Axonal Kv Channels: A Tripartite of Kv1.1/Kv1.2/Kv␤2

Voltage-gated ion channels are porelike, membranelodged proteins that mediate rapid ionic flux (106 ions/s) across cell membranes. Voltage-gated channels are important for axons because they allow action potentials to be generated; the initial rapid influx of sodium ions through sodium channels causes the rapid upstroke of an action potential followed by a delayed efflux of potassium ions through potassium channels to repolarize the membrane. This cycle of depolarization and hyperpolarization also generates axial current along the axoplasm, which excites neighboring membrane patches and propels action potentials along nerve fibers at a high speed. This model of excitation, referred to as the Hodgkin-Huxley model of nerve excitation,52 has, since its inception in 1952, formed the basis for understanding fast electrical signaling in all nerve models, in both health and disease. In the early 1980s, Neher and Sakmann ushered in the phase of molecular study of ion channels with the invention of the patchclamp technique, which allows the electrical events associated with the opening and closing of a single ion channel protein to be observed. Cloning of ion channels then followed and dominated the field. As of the writing of this chapter, 50 potassium channel genes and 10 sodium channel genes have been identified in mammals; many of them will undoubtedly be given proper addresses in the human genome once the Human Genome Project is completed. Molecular biology also allows ion channel crystals to be grown and their three-dimensional structure solved. In a

Mechanisms for Clustering Kv Channels Function of Kv Channels K⫹ Channels on Schwann Cells

brief span of 3 years (1998 to 2001), MacKinnon and colleagues42,51,84,149 revolutionized the biophysical study of ion channels by solving the pore structure of potassium channels by x-ray crystallography at 2.0- to 3.2-Å resolution, producing a vivid molecular contour of the selectivity filter through which K⫹ ions are seen navigating, captured in a stunning series of frozen molecular snapshots. Basic studies of ion channels in axons and ensheathing cells have gained considerably from genetic cloning, because many ion channel genes are expressed in peripheral nerves, allowing ion channels in these nerves to be studied in molecular terms unthinkable 9 years ago when this chapter was last written. On the clinical front, neuropathy increasingly is linked to mutations of known ion channel genes, offering the potential for designing treatment based on precise molecular targets. This chapter summarizes the current status of ion channel research in axons and ensheathing cells of the peripheral nerves with a primary focus on three major issues: the molecular identity of ion channels, the mechanisms responsible for localization of ion channels, and the normal and pathophysiologic functions of ion channels. When necessary, I have tried to give a brief account of the background to the rapidly evolving molecular research that, for the past 5 years, has shaped progress and defined future questions. Readers interested in a fuller description of the pre–molecular biologic background should consult my earlier chapter written for the third edition of Peripheral Neuropathy.29 95

96

Function of the Peripheral Nervous System

Naⴙ CHANNELS

A

intact

paranodal demyelination

ⴙ

Na Channels on Axons The peripheral nervous system has two types of axon: the large, fast-conducting myelinated axons and the small, slower conducting nonmyelinated axons. Voltage-gated Na⫹ channels (Nav) are sodium-selective pores whose opening and closing is gated by voltage. Nav channels have different distribution patterns on these two types of axon. In the nonmyelinated axons, Nav channels are uniformly distributed; in contrast, Nav channels in the myelinated axon are clustered at the node of Ranvier. Nodal clustering of ion channels was first hypothesized by Rosenbluth,108 who, based on freeze-fracture studies of myelinated axons, suggested that Nav channels are localized to the nodal membrane and trapped there by the paranodal junction. The first experimental evidence for Nav channel segregation came from studies of binding of 3H-saxitoxin to sodium channels of myelinated nerves. Noting that there is no difference in saxitoxin binding to intact and homogenized sciatic nerves, Ritchie and Rogart106 concluded that Nav channels are concentrated at the nodal region and absent in the internodal membrane masked by the myelin. This finding was later confirmed in voltage-clamp studies of single mammalian myelinated fibers in which acute paranodal demyelination was found to add little to the nodal sodium currents (Fig. 4–1A).31 Interestingly, by directly dissolving the myelin of a single internode, voltage clamp studies do show a low density of Nav channel in the internodes of adult myelinated nerves.37 Measurement of gating current (the tiny capacitative current associated with the activation of a single Nav channel) suggests that a single mammalian node of Ranvier has approximately 82,000 channels,24 which, with an assumed nodal area of approximately 60 ␮m2, translates to a very high channel density of approximately 1400/␮m2.

Molecular Identity of Nav Channels on Axons Do axonal Nav channels form a single homogeneous pool? Do different types of axons express different types of Nav channels? Earlier voltage-clamp studies revealed that nodal sodium currents in myelinated fibers inactivate in a double exponential time course, a finding consistent either with a single Nav type with second-order inactivation kinetics23 or two types of Nav channels with different inactivation kinetics.10 Furthermore, Nav channels exhibit different tetrodotoxin sensitivity. Hence, classic studies have pointed to diversity in axonal Nav expression. This issue is now solved in spectacular molecular detail by the cloning of Nav channel genes. At the writing of this chapter, 10 Nav channel genes (Nav1.1 through Nav1.9) have been cloned in mammals, and many of these are expressed

10

10

nA

nA 0

0

-10

-10

0 nA -5

0 nA -5

0

B

C

msec

0

4

msec

4

nodal Nav1.6

node

Juxtaparanodal Kv1.1

FIGURE 4–1 Complementary distribution of Na⫹ and K⫹ channels in mammalian PNS myelinated axons. A, Discovery of complementary distribution of Na⫹ and K⫹ channels in 1980. A single mammalian myelinated fiber was voltage-clamped before (left) and 15 minutes after (right) acute paranodal demyelination. Demyelination reveals a large outward K⫹ current (upward deflection) without affecting the inward Na⫹ current (downward deflection), suggesting that Na⫹ channels are localized to the node and K⫹ channels are under the myelin sheath. (From Chiu, S. Y., and Ritchie, J. M.: Potassium channels in nodal and internodal axonal membrane of mammalian myelinated fibres. Nature 284:170, 1980, with permission.) B and C, Recent immunohistochemical confirmation of complementary distribution of Na⫹ and K⫹ channels. B, Staining of Nav1.6 at the node (red). The paranodal adhesion molecule contactin is stained green. (Courtesy of Rock Levinson.) C, Staining of Kv1.1 in the juxtaparanode (green). Note the lack of Kv1.1 staining at the nodal gap. Paranodal contactin is stained red. (Courtesy of Steve Scherer and Arroyo Edgardo.)

in neurons.47,48 These different Nav channel genes encode sodium channel proteins with similar structural motifs but with different kinetic properties as a result of variation in other regions of their primary amino acid sequence.47

Channel Function in Mammalian Axons and Support Cells

Expression of different Nav isoforms in different axon types would greatly enrich electrophysiologic signaling in the peripheral nervous system (PNS). When investigators perform systematic histochemical analysis of Nav channel expression on axons with isoform-specific antibodies, a fascinating picture of Nav channel expression emerges. Small C-type nonmyelinated axons of neurons express Nav1.9 channels.71 This channel generates a persistent sodium current, potentially rendering C fibers sensitive to subthreshold stimulations. Nav1.2 is expressed predominantly in nonmyelinated axons,49,137 whereas Nav1.6 is expressed predominantly at nodes of Ranvier (red in Fig. 4–1B).21 The story took an intriguing turn when it was found that the Nav subtype switched from Nav1.2 to Nav1.6 when axons become myelinated, a switch first identified in central nervous system (CNS) axons14 but that since has been found applicable to PNS axons as well.99 At present, the functional consequence of this developmental switch in Nav channel subtype is unclear. One idea is that this developmental switch reflects a change in the anchoring proteins needed to cluster Nav channels during myelinogenesis; apparently changing the lock changes the key that fits. Mutations of Nav channels are neuropathologic.62,119 For example, mutation in Nav1.2 causes seizures in mice in transgenic studies, and it has been suggested that humans with mutation of this gene may be candidates for seizure disorders.62 A mutation of the Nav1.6 channel in mice (med) results in a truncated, nonfunctional gene product,20,64 causing mice to die 21 to 28 days after birth. Interestingly, even though the med mice lack immunostaining of Nav1.6 channels in the Ranvier nodes in sciatic nerves,21 the conduction velocity is only mildly affected.43,104 Furthermore, [3H]-tetrodotoxin binding is significantly increased in med sciatic nerves,104 suggesting an upregulation of a yet-to-be identified Nav isoform that compensates for the loss of Nav1.6 channels. The interesting point here is that deleting Nav1.6 also causes paranodal dysmyelination and nodal widening. This finding, together with the switch from Nav1.2 to Nav1.6 channels during myelination, suggests that inappropriate expression of nodal Nav channel isoforms is incompatible with normal nodal structure.

Mechanisms for Nav Channel Segregation Clustering of Nav channels at the Ranvier node is critical for generating high-speed nerve conduction in adult myelinated fibers. How does this clustering occur during development, and how are the clusters maintained after they are formed? For obvious reasons, both in terms of normal physiology and in terms of the neuropathologic impact in multiple sclerosis, the molecular basis for Nav channel clustering has become one of the most intensely researched areas in the field. As will be evident in the

97

discussion that follows, at least four different mechanisms contribute to Nav channel clustering at the Ranvier nodes: the physical barrier imposed by the paranodal junction, direct anchoring of Nav channels to extracellular and intracellular molecules, linkage to auxiliary ␤ subunits of Nav channels, and diffusible factors. In an elegant study combining Nav staining and electrophysiology in rat sciatic nerves, Vabnick and co-workers127 detected Nav clustering as early as postnatal day 1. As Nav clusters started to mature during myelinogenesis, there was a dramatic increase in conduction velocity up to 10 m/s. Clearly, insulation of leakage current across the internodal axon by myelin ensheathment contributes to the increase in conduction velocity. Nevertheless, the concomitant clustering of nodal Nav channels at high density is absolutely required on pure electrophysiologic grounds to allow the tiny nodal membranes to generate an intense focal Na influx to propel saltatory conduction. By following Nav staining before and after myelination formation in the rat sciatic nerve, Vabnick and co-workers127 found that the early Nav channel clusters appear to be pushed ahead by the elongating edge of Schwann cells as they stretch and occupy the future internodal domain on the growing axons. This picturesque finding suggests that Nav clustering requires physical contact between axons and Schwann cells. However, diffusible factors might also contribute to channel clustering. The first serious evidence for this theory comes from in vitro studies of CNS nerves. Kaplan and colleagues61 found that mammalian retinal ganglion cells, when cultured in the absence of direct physical contact with oligodendrocytes (the myelinating cells in the CNS), could form Nav clusters on axons if oligodendrocyteconditioned medium is present. These two types of clustering mechanism might cofunction and are not necessarily mutually exclusive; nevertheless, the debate on the relative role of diffusible factors and physical contact in Nav channel clustering will likely go on with renewed vigor as these factors are identified in the future and subjected to genetic manipulations. Intracellular cytoskeletal elements and extracellular adhesion molecules also contribute to Nav channel clustering. Nav channels have extensive cytoplasmic domains containing structural motifs that might interact with intracellular anchoring proteins. One key intracellular anchoring protein has been identified as ankyrin G, which interacts with purified Nav proteins in biochemical assays120 and colocalizes with Nav channels at the Ranvier nodes in immunohistochemical studies. A cerebellar-specific gene knockout of ankyrin G greatly reduces Nav channels at the initial segments of granule cells.9,145 Another cytoskeletal protein, spectrin isoform ␤IV, co-localizes with Nav channel clusters.11 Extracellular cell adhesion molecules (CAMs) found on the nodal surface with potential for Nav channel clustering activity include neurofascin and Nr-CAM.4 Extracellular and intracellular anchoring proteins apparently interact.

Function of the Peripheral Nervous System

Na⫹ Channels on Schwann Cells Although the main function of Schwann cells is myelin formation and axon insulation, these cells surprisingly express voltage-gated Na⫹ and K⫹ channels, the hallmark of

excitable cells. This discovery was first made in patch-clamp recordings from Schwann cells in culture (Fig. 4–2A).36 Subsequent patch-clamp studies on acutely isolated Schwann cells, with axons still attached to them, confirmed that Nav channels are normally expressed in Schwann cells in vivo (Fig. 4–2B).25 Interestingly, Nav channel activity is downregulated at the Schwann cell soma during myelin elaboration,25 suggesting that the expression of Nav channels on Schwann cells is related to myelinogenesis.30 It has been suggested that during myelin formation, Nav channels on the Schwann cells are redistributed from the soma to distal processes.28 During wallerian degeneration, dying

A

Cultured Schwann cell

400

pA

For example, ankyrin interacts with the cytoplasmic domains of neurofascin, Nr-CAM, and Nav channels,4,120 and ankyrin G knockout mice exhibit altered neurofascin expression in the initial segments of granule neurons in the cerebellum.145 An intriguing finding is that there is a shift in the ankyrin isoform expression (from ankyrin B to ankyrin G) that parallels a shift in the Nav isoform expression (Nav1.2 to Nav1.6) during myelinogenesis.69 The emerging molecular picture appears to be one in which each Nav channel subtype is tailor-made to fit only a specific subset of anchoring proteins, and that maintaining a stable Nav channel cluster in a mature nerve requires a dynamic deployment of specific anchoring proteins. Once a mature Nav channel cluster is formed, anchoring to axonal cytoskeletal elements appears to be more important than the Schwann cells in maintaining the Nav channel clusters. This conclusion comes from a recent study in which contactin, a paranodal protein involved in tightening the paranodal seal, was genetically deleted from mice. Interestingly, despite loosening of the paranodal seal, which causes a dispersion of the juxtaparanodal potassium channels (see later in this chapter), the nodal Nav channel clusters remain intact, suggesting that the paranodal seal is not very important in preserving Nav channel clusters after they are formed. Besides physical constraint by the paranodal junction and anchorage to an assortment of extracellular and intracellular proteins, auxiliary ␤ subunits (Nav␤1, ␤2, ␤3) of Nav channels also have been suggested to contribute to Nav channel clustering at the Ranvier nodes. Nav␤ subunits are transmembrane proteins with an extracellular immunoglobulinlike domain and a cytoplasmic domain.54–56 The cytoplasmic domain has structural motifs that have been hypothesized to interact with ankyrin, whereas the extracellular domain has been suggested to interact with like molecules from Schwann cells in a homophilic fashion75 or with extracellular adhesion molecules such as tenascins and neurofascin.101 Neurofascin and Nav␤1 are co-localized to Ranvier nodes in rat sciatic nerves,101 and in vitro studies using co-cultures of Schwann cells and dorsal root ganglion neurons show that perturbing the function of Nr-CAM, a CAM expressed at the axon surface, blocks Nav channel and ankyrin accumulation at the nodes of Ranvier.73 Researchers now envision a Nav channel raft using the membrane-spanning Nav␤ auxiliary subunit as an intermediary to coordinate extracellular and intracellular interactions with other anchoring elements.73,75 Gene knockouts of Nav␤ will undoubtedly be forthcoming to shed light on the role of ␤ auxiliary subunits on Nav channel clustering.

200 0 -200

5 msec

Acutely isolated Schwann cell

B 200

ionic current (pA)

98

0

Na+ current -300 5 msec

FIGURE 4–2 Na⫹ and K⫹ channels in mammalian Schwann cells. A, Wholecell ionic currents recorded from a single cultured Schwann cell showing Na⫹ currents (downward) and K⫹ currents (upward). (From Chiu, S. Y., Schrager, P., and Ritchie, J. M.: Neuronal-type Na⫹ and K⫹ channels in rabbit cultured Schwann cells. Nature 311:156, 1984, with permission.) B, Na⫹ currents recorded from a single, acutely isolated mammalian Schwann cell. The picture shows the glass patch-pipette used to record the currents from the cell soma. (From Chiu, S. Y.: Sodium currents in axonassociated Schwann cells from adult rabbits. J. Physiol. [Lond.] 386:181, 1987, with permission.)

Channel Function in Mammalian Axons and Support Cells

axons apparently provide a signal to Schwann cells to reexpress Nav channels at the soma.26 Intriguingly, messenger RNA (mRNA) for Nav␤1 appears to be present in Schwann cells of normal myelinated fibers.87 It is unclear where on the myelinating Schwann cell the Nav␤1 protein is expressed, an issue that could be of particular interest given the hypothesis that homophilic interactions between the extracellular domain of Nav␤1 from axons and Schwann cells might play a role in Nav channel clustering at the Ranvier nodes.75 What are the functions of Nav channels on Schwann cells? Their low density appears to preclude Schwann cells from generating action potentials.25,36,113 One possible function is that Nav channels in Schwann cells act to recruit the auxiliary Nav␤1 subunits to the Schwann cell surface where, as discussed previously, they could provide homophilic interactions with axonal Nav␤1 to stabilize Nav channel clusters on axons. The viability of this hypothesis clearly awaits definitive immunohistochemical staining of Nav␤1 on mature Schwann cells to see if it co-localizes and apposes to Nav␤1 expressed on the axons. In a different type of functional speculation, Chiu and colleagues proposed that Schwann cells synthesize Nav channels for axons.36 With the identification of Nav channel subtypes through gene cloning, this hypothesis can be examined by looking for a common set of Nav isoforms expressed in Schwann cells and axons. Teleologically, a transfer of Nav channels from Schwann cells to axons would relieve the biosynthetic load of the neurons.105,118 Finally, Sontheimer and colleagues have advanced the theory that the sodium influx through glial cell Nav channels can be used to fuel potassium uptake through the glial Na⫹,K⫹-ATPase, thereby contributing to potassium buffering.118 If Schwann cell Nav channels are activated during axonal activity, Sontheimer et al.’s model allows an activity-dependent augmentation of potassium uptake through the Schwann cell Na⫹,K⫹-ATPase that would be matched to nerve activity.

K⫹ CHANNELS In contrast to Nav channels, whose primary role is to mediate the rapid upstroke of an action potential, voltagegated potassium channels (Kv) play an important role in the modulation of excitability such as frequency coding. In their pioneering work on excitability of the nonmyelinated squid giant axon, Hodgkin and Huxley52 concluded that Kv channels play a major role in terminating the action potential by providing a delayed potassium efflux to repolarize the membrane. The function of Kv channels in myelinated axons was addressed only when a new voltageclamp technique was invented for studying single myelinated fibers.41,86 The first studies with these new voltage-clamp techniques showed that Kv channels are

99

present in amphibian nodes of Ranvier and play a central role in action potential repolarization (Fig. 4–3). Surprisingly, subsequent voltage-clamp studies on single mammalian myelinated axons from peripheral nerves, including human nerves, showed that the nodal membrane of these fibers lacks Kv channels.17,34,53,110 Action potential repolarization in the mammalian PNS, in contrast to the amphibian PNS myelinated nerves, relies only on Nav channel inactivation and passive leakage current (Fig. 4–3).34 Are Kv channels absent from mammalian myelinated nerves, or are they present in extranodal regions? The first clue for the latter possibility came from a study111 showing that Kv channels are present on chronically demyelinated axons. Chiu and Ritchie31 then showed in acute paranodal demyelination studies that Kv channels are present normally under the myelin sheath (see Fig. 4–1A). By early 1980, the conclusion was established that Kv channels are segregated in a fashion opposite to that of Nav channels, in that Kv channels are absent in the node but present under the myelin (see Fig. 4–1A). This was followed by a period of extensive electrophysiologic characterization of mammalian internodal potassium channels. Baker and colleagues6 defined the presence, as well as the probable function, of three types of rectifying channel (one sensitive to 4-aminopyridine [4-AP], one to tetraethylammonium [TEA], and the third to cesium) under intact myelin in rat spinal root myelinated axons. A major breakthrough in electrophysiologic analysis of axonal channels occurred when highly skilled investigators demonstrated that they can acutely peel away the myelin and subject the naked but tiny axon to patch-clamp recordings.58 These studies identified up to eight different types of axonal potassium channels, including Kv channels, leakage K⫹ channels that show weak voltage dependence, and K⫹ channels sensitive to intracellular metabolic chemistry. Readers interested in the electrophysiologic classification of axonal K⫹ channels should consult several excellent reviews on this subject.130,133 Our main focus here is on the molecular study of axonal Kv channels, which has seen a quantum leap in new knowledge fueled by the pace-setting application of molecular biology and transgenic techniques.

Molecular Identity of Axonal Kv Channels: A Tripartite of Kv1.1/Kv1.2/Kv␤2 Cloning of K⫹ channel genes from mammalian brain has identified three large families of K⫹ channels classified structurally according to the number of transmembrane domains (six, four, or two) in each ␣ subunit.57 Of the three families, the one with six transmembrane regions constitutes the largest family and is the primary focus of this chapter. Within this family, at least 16 subfamilies have been identified based on similarity in the primary amino acid sequence. Most, but not all, of the K⫹ channels in these 16 subfamilies

100

Function of the Peripheral Nervous System

FIGURE 4–3 Macroscopic ionic currents (top) and action potentials (bottom) in nodes of Ranvier. Top, Na⫹ currents are displayed as downward deflections and K⫹ currents as upward deflections. Note that K⫹ current is present in the amphibian (right) but absent in the mammal (left). Bottom, Computer reconstruction of the mammalian action potential based only on Na⫹ and leakage currents: empirical data (left) and computer reconstruction (right). Repolarization is accounted for by Na⫹ channel inactivation and passive leakage current. Unlike the amphibian myelinated nodes, voltage-gated K⫹ channels are not required. (From Chiu, S. Y., Ritchie, J. M., Rogart, R. B., and Stagg, D.: A quantitative description of membrane currents in rabbit myelinated nerve. J. Physiol. [Lond.] 292:149, 1979, with permission.)

are voltage gated (Kv), and each subfamily consists of numerous members. For example, the Kv1 subfamily (also called Shaker subfamily) has up to nine members, Kv1.1 through Kv1.9. Each functional K⫹ channel is composed of four ␣ subunits, and only members of the same subfamily can co-assemble.70 This tetramultimeric nature of K⫹ channels provides great functional diversity through mixing and matching different members of the same subfamily to form potassium channels with unique properties. Auxiliary ␤ subunits (Kv␤s) further diversify Kv channel function.96,125 Three Kv␤ subfamilies have been identified (Kv␤1, Kv␤2, and Kv␤3) and various functions of the Kv␤ subunits have been proposed. Some of these cytoplasmic Kv␤ subunits can induce Kv␣ channel inactivation, leading to the hypothesis that Kv␤ subunit association modulates frequency coding in excitable cells.96 Kv␤ subunits have also been

suggested to be chaperones to facilitate surface expression of Kv␣ subunits.112 Intriguingly, Kv␤ belongs to the aldoketoreductase (AKR) family,40,79 which has led to the hypothesis that the association of Kv␤ with Kv␣ subunits allows excitability of cells to be modulated by intracellular oxidation-reduction chemistry. As for Nav channels, axonal Kv channels in the mammalian peripheral nerves do not form a homogeneous pool. Using isoform-specific antibodies in immunohistochemistry, the most conspicuous landmark is a tripartite Kv1.1/Kv1.2/Kv␤2 clustered at the juxtaparanodes of mature myelinated axons (see Fig. 4–1C),98,100,103,132 a finding confirming the conclusion of earlier electrophysiologic studies regarding Kv channel localization under the myelin.31 It should be noted that this immunohistochemical survey might not be exhaustive; given the diverse population

101

Channel Function in Mammalian Axons and Support Cells

of K channels characterized in patch-clamp studies, other molecular species might well be discovered in the future. Nevertheless, the discovery of the Kv1.1/Kv1.2/Kv␤2 juxtaparanodal tripartite in mammalian peripheral axons has directed research along two major lines of questions. First, what maintains the Kv1.1/Kv1.2/Kv␤2 tripartite in the juxtaparanode of a mature myelinated axon? Second, what is the function of the juxtaparanodal Kv tripartite?

Mechanisms for Clustering Kv Channels During development of the sciatic nerves, Kv1.1/Kv1.2/ Kv␤2 clusters first appear at the nodal membrane, then quickly migrate toward the paranode before stabilizing in the juxtaparanode in a mature myelinated axon,128 an observation that accounts for the gradual diminution of sensitivity of sciatic nerves to Kv channel blockers (TEA and 4-AP) during development.63 What maintains the Kv channel clusters at the juxtaparanode in a mature nerve? As in the case of Nav channels, a physical barrier imposed by the paranodal junction and anchorage to an assortment of proteins have been shown to play a role. The role of Schwann cells in organization of axonal Kv channels has been investigated using various PNS dysmyelinating mutants, including the Trembler mice. Axonal Kv1.1 channels are desegregated in the Trembler dysmyelinated nerves,131 suggesting that the Schwann cell–axon junction

at the paranode is important for maintaining channel clustering. Recently, the hypothesis regarding the paranodal junction as a key factor in positioning the Kv channel clusters has been definitively confirmed by molecular biology studies in which the nuts and bolts of the protein assembly at the junction have been identified and selectively deleted in knockout mice.13,16,91 At least three assembly proteins have been identified that pull the paranodal Schwann cell loops toward the axon at the paranode to form a tight junction. They are Caspr1 and Contactin (which form a functional complex) from the axonal membrane,16,44,80 which interlock with neurofascin (NF)-155 protruding from the Schwann cell loops122 (Fig. 4–4C).16 After the contactin gene is genetically inactivated,16 Caspr expression is abolished from the paranodes. The NF-155 from the Schwann cell fails to interlock to the axons, which causes the Schwann cell loops to recede slightly from the axon, resulting in loosening of the paranodal junction and slowing of the nerve conduction (Fig. 4–4A and B). Kv1.1 is found to migrate from the juxtaparanode to the paranode and stops next to the nodal Nav channels, whose cluster is not affected by contactin deletion (see Fig. 4–4C). Anchoring proteins also have been suggested to contribute to Kv1 channel clustering. Recently, Kv1 channels in the juxtaparanode were found to co-localize and coimmunoprecipate with Caspr2, another member of the neurexin superfamily to which Caspr1 belongs.95 The large

C

A

wild type

rec

stim

NF-155 Caspr +-

distance

- + Na+ - channels

B

juxtaparanode

paranode

node

juxtaparanode

paranode

node

Contactin

mutant + /+ CNTN

1 msec

CNTN

-/-

FIGURE 4–4 Neuropathology of genetic ablation of the paranodal adhesion protein contactin. A, Experimental configuration for measurement of conduction of compound action potential in sciatic nerves of mice. B, Slowing of nerve conduction in the contactin knockout (bottom) compared with the wild type (top). C, Model (top) showing the three assembly proteins (NF-155 from Schwann cells and Caspr and Contactin from the axon) in a normal paranode. Deletion of contactin (bottom) results in lifting of the Schwann cell paranodal loops and migration of Kv channels from the juxtaparanode to the paranode. (From Boyle, M. E., Berglund, E. O., Murai, K. K., et al.: Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30:385, 2001, with permission.)

102

Function of the Peripheral Nervous System

extracellular domain of Caspr2 might interact with binding partners from Schwann cells, helping to regulate Kv1 channel localization to not only the juxtaparanodes but along the inner mesaxon as well.5 Binding to various intracellular proteins, as in the case of Nav channels, might also determine Kv1 localization. For example, there is a PDZbinding domain in the cytoplasmic C-terminus of the Kv1 channel that could act as an anchoring site for other cytoplasmic PDZ-containing elements such as PSD-95.95 As discussed in the case of Nav channel clustering, cytoplasmic ankyrin and spectrin ␤IV appear to anchor Kv channels. The role of spectrin ␤IV in channel organization has been clarified by the autosomal recessive mutation quivering, a spontaneous mutation in mice known since 1953 to produce ataxia, hind limb paralysis, and deafness. Recently, the quivering mutation has been traced to a lossof-function mutation in spectrin ␤IV, which interestingly correlates with an alteration of Kv1 channel localization in myelinated axons of quivering mice.90 Does Kv␤2 promote trafficking and appropriate localization of the Kv1.1/Kv1.2/Kv␤2 tripartite in myelinated axons? As for Nav channels, auxiliary subunits have also been proposed to organize axonal localization of Kv channels. Kv␤2 binds to K⫹ channel ␣ subunits and is a member of the AKR superfamily. However, the Kv␤ subunits, unlike the Nav␤ subunits, are strictly cytoplasmic proteins with no transmembrane domains, a feature that precludes extracellular homophilic interactions with apposing Schwann cell Kv␤s as a mechanism for clustering Kv␣ subunits on axons. Proposed functions for Kv␤s include a chaperone-like role in Kv1 ␣-subunit biogenesis and catalytic activity as an AKR oxidoreductase. In cultured mammalian cells, the absence of Kv␤2 results in unglycosylated forms of Kv1.1 and Kv1.2 that are not efficiently transported to the membrane surface, leading to the hypothesis112 that Kv␤2 facilitates the glycosylation of Kv1␣ subunits in the endoplasmic reticulum (ER), thereby promoting the trafficking of Kv␣/␤ complexes from the ER to the surface cell membrane. Furthermore, in cultured hypocampal neurons, Kv1.2 was appropriately targeted to axons only when cotransfected with Kv␤2.22 Whether Kv␤2 is important for Kv1.1/Kv1.2 clustering in vivo has recently been tested by the generation of a Kv␤2 knockout mouse.78 These mutant mice have a shortened lifespan and a cold swim–induced myotonia indicating hyperexcitability. Surprisingly, Kv1.1/Kv1.2 clusters form normally in the juxtaparanodes of mature PNS myelinated axons, suggesting normal biogenesis and trafficking of Kv1␣s in the absence of Kv␤2.78 The differences between the in vivo and in vitro studies might be reconciled if, for example, other Kv␤ subunits (such as Kv␤1) were upregulated in the Kv␤2-null mice to compensate for the loss of Kv␤2. However, PNS myelinated axons normally do not express Kv␤1, and Western blot analysis of whole-brain lysates revealed no significant change in the Kv␤1 protein

expression in the Kv␤2-null mice.78 Alternatively, anchorage to cytoskeletal elements such as spectrin ␤IV might compensate for the loss of Kv␤2. Finally, other yet-to-be identified Kv1␣ subunits in axons could facilitate efficient surface expression of heteromultimeric Kv1 channels at the juxtaparanodes in the absence of Kv␤2.77 It is likely that multiple binding partners coordinate the juxtaparanodal clustering of Kv1 channels with a high degree of redundancy, and simultaneous genetic inactivation of multiple binding partners might be needed to shed light on the role of protein-protein interactions in Kv1 channel localization. Figure 4–5 summarizes the evolution of our knowledge on the organization of ion channels in myelinated axons from the early 1960s (top) to 2002 (bottom).

Function of Kv Channels Physiologic Functions Having reviewed the mechanisms for Kv1 channel clustering, we next turn to the physiologic and pathophysiologic functions of Kv1 channels in PNS myelinated axons. During development, the transient presence of Kv1 channels on the nodal membrane has been suggested to prevent abnormal excitation in developing myelinating nerves.128 However, what might the functions of Kv1 channels be after they have been fully sequestrated in the juxtaparanode, which is thought to be sealed off from the nodal membrane by the paranodal junction108? In the early 1980s, it was proposed that a mature paranodal junction has significant finite leakage across it, allowing electrotonic interactions between the nodal and the internodal axonal membrane.8,33 As a consequence, it has been proposed that nodal excitation can evoke a delayed, secondary depolarization of the juxtaparanodal membrane, which when amplified by a trace amount of internodal Nav channels,37 can lead to reentrant excitation of the node.32 The juxtaparanodal Kv1 channels have been suggested to dampen this putative reentrant excitation, preserving faithful electrical transmission along the axons. This hypothesis has been difficult to test because of the difficulty in experimentally manipulating the juxtaparanodal Kv1 channels, which are normally wrapped by the myelin. However, a rare human neurologic disorder, episodic ataxia type 1 (EA-1), provides clues to the function of Kv1 channels in myelinated axons. Episodic ataxia is characterized by stress-induced hyperexcitability; EA-1 patients appear normal but intermittently display a protracted phase of ataxia and tremors brought on by stress, including startle and exercise.19 The myokymia component in the clinical manifestations suggests PNS hyperexcitability. Recently, EA-1 has been traced to various missense and nonsense mutations of the human Kv1.1 gene, which result in reduced levels of functional channels, abnormal trafficking, and intracellular aggregation of mutated Kv1.1 proteins.1,19,76,143 Genetic ablation of Kv1.1 in mice was therefore used to assess Kv1.1 functions in vivo.114 Kv1.1-null mice display

103

Channel Function in Mammalian Axons and Support Cells

Node of Ranvier

Initial segment Axon hillock

Dendrite

Na+ Myelin

+

+ -

Axon

+ Myelin

+

-

Na+

microvilli

internode

basal lamina

incisure

juxtapara para

Kv1.1 Kv1.2 Caspr Kvβ2 Caspr2

node

Nav1.6 ankyrinG

PNS CNS

Perinodal astrocyte Compact Myelin

Adaxonal Membrane Microvilli or Perinodal Astrocyte

15 5 Co

nt a cti n

NF

sp r Ca

Kvβ2

Tenascin-R β2

Kv1.1 Kv1.2

NF186

NF

18 6

Paranodal Myelin Loop

Tenascin-C

Caspr-2

K+ Channels

Nr-CAM

FIGURE 4–5 Summary of the evolution of our knowledge on ion channel localization in the mammalian myelinated nerves from the 1950s to 2002. Top, Early view (the 1950s) of myelinated nerves. Action potentials are generated in the initial segment of the neuron and propagate in a saltatory fashion from node to node. The myelin sheath at the internode prevents capacitative and resistive loss, allowing highspeed nerve conduction. (From Peles, E., and Salzer, J. L.: Molecular domains of myelinated axons. Curr. Opin. Neurobiol. 10:558, 2000, with permission.) Bottom, Current view (2002) of myelinated nerves. Molecular biology and gene cloning have identified the ion channel species, clarified their localization, and identified many anchoring proteins involved in organizing the ion channels. (Courtesy of Steve Scherer and Arroyo Edgardo.)

Band 4.1B PDZ?

Spectrin βIVS1 AnkyrinG Na+Channel

104

Function of the Peripheral Nervous System

epilepsy and ataxia resulting from deleting Kv1.1 expression in the CNS.114,144 What is the physiologic impact of deleting Kv1.1 on PNS myelinated axons? First, there is a slight increase in the refractory period of nerve transmission in the adult sciatic nerves,114 suggesting that juxtaparanodal Kv1.1 contributes to electrogenesis. Second, the Kv1.1-null mice display cold swim–induced myokymia, which hints at a stress-inducible hyperexcitability of the neuromuscular junction. Immunohistochemical, electrophysiologic, and theoretical analysis of the neuromuscular transmission in the mutants revealed a temperaturedependent repetitive discharge from a short myelinated segment just preceding the nerve terminal.147,148 The hypothesis is that a preshortening of the internodes prior to the nerve terminal, while needed to facilitate safe invasion of the nerve terminal by action potentials,97,102,134 also increases the electrotonic coupling between the nodal and internodal axonal membrane, which is inherently destabilizing. This juxtaparanodal Kv1.1 is theorized to stabilize the preterminal, myelinated segment of the axon from reentrant excitation.147,148 This proposed action of juxtaparanodal Kv1.1 could be of general applicability to branch points. Computer simulations have shown that an inappropriate shortening of internodes prior to a branch point, while providing safe conduction across the branch point, also creates inherent instability in the pre–branch point segments.146 Juxtaparanodal Kv1.1 might act to prevent reentrant excitation in myelinated segments in the immediate vicinity of branch points, preserving fidelity of electrical signaling in axonal trees. It should be pointed out that, because Kv1.2 channels still remain in the juxtaparanodes in the Kv1.1-null mutants, a double knockout of Kv1.1 and Kv1.2 might be needed to assess fully the physiologic role of juxtaparanodal Kv1 channels. Pathophysiologic Functions Kv1 channels in myelinated axons are highly relevant in demyelinating diseases. Once exposed by demyelination, the internodal Kv1 channels act like a brake to retard conduction across the lesion. Although the scarcity of internodal Nav channels coupled with impedance mismatch134 contributes to conduction failures in demyelination, neutralizing the braking action of Kv channels by pharmacologic intervention has been explored as a means to restore conduction in demyelination. Even a slight migration of Kv1 channels from the juxtaparanode to the paranode, as occurs in the contactin knockout mice (see above), profoundly affects nerve conduction in myelinated fibers.16 Normally, Nav and Kv1 channels are physically delineated by a 2- to 4-␮m paranodal seal. Studies of the contactin knockout mice show that inappropriate encroachment of Kv1 channels on Nav channels consequent on a slight loosening of the paranodal pockets may be incompatible with normal nerve function.16 Schauf and Davis109 first pointed out on electrophysiologic grounds that blockage of Kv1

channels with 4-AP should increase the safety factor of conduction in demyelinated axons. This idea was first confirmed in animal models of demyelination,15,123 then followed by clinical trials121 showing that oral ingestion of 4-AP transiently improves neurologic functions in patients with multiple sclerosis over those given placebo. The use of K⫹ channel blockers in clinical trials in patients with multiple sclerosis has been reviewed.12,83,121 Other reviews on the role of Kv channels in electrogenesis of myelinated axons can be found elsewhere.39,135 The identification of the juxtaparanodal tripartite Kv1.1/Kv1.2/Kv␤2 should provide a more specific target for therapeutic design in pharmacologic intervention strategies in multiple sclerosis.

K⫹ Channels on Schwann Cells Kv channels were first described in mammalian Schwann cells in culture conditions using patch-clamp techniques (see Fig. 4–2A).36 Subsequent patch-clamp studies from acutely isolated Schwann cells demonstrate that Kv channel expression is not a tissue culture artifact.25 Electrophysiologic studies have distinguished various types of Kv channels in Schwann cells according to kinetics and pharmacology.2,7,65,129 Furthermore, inward rectifier K⫹ channels (Kir) have also been described in mammalian Schwann cells.66,67,138,139 Inward rectifiers, unlike Kv, pass inward currents more readily than outward currents, as a result of a cytoplasmic blocking particle (Mg2⫹ or polyamines such as spermine and spermidine) that gets swept into the pore when the cell is depolarized, thereby obstructing the efflux of K⫹ ions. With the presence of Kv and Kir channels firmly established in Schwann cells, research has focused on three broad issues: the relationship between Schwann cell K⫹ channel expression and myelinogenesis, the molecular identity of Schwann cell K⫹ channels, and the function of Schwann cell K⫹ channels. Expression of Schwann Cell K⫹ Channels Linked to Myelinogenesis Patch-clamp studies from Schwann cells acutely isolated from sciatic nerves at different stages of myelination have shown that Kv and Kir channel activity is downregulated at the Schwann cell soma as myelin is elaborated (Fig. 4–6, left).139 Interestingly, in an adult myelinated fiber, both Kv and Kir channel activity can be detected at Schwann cell processes near the node of Ranvier,138 leading to the suggestion of a developmental redistribution of both channels to the Schwann cell microvilli during myelin formation. This pattern of Schwann cell channel expression during normal myelination is reversed when axons are transected in mature myelinated nerve fibers26; as the axons and myelin sheath degenerate during wallerian degeneration, Nav and Kv channel activity reappears on the soma of Schwann cells (Fig. 4–6, right).26 It is likely that a combination of trophic signals both from axons and the myelin

Channel Function in Mammalian Axons and Support Cells

105

FIGURE 4–6 Relation between K⫹ channel activity and proliferation in mammalian Schwann cells. Left, Reduction in Schwann cell proliferation during normal myelinogenesis is paralleled by a reduction in K⫹ channel activity at the Schwann cell soma. (From Wilson, G. F., and Chiu, S. Y.: Potassium channel regulation in Schwann cells during early developmental myelinogenesis. J. Neurosci. 10:1615, 1990, with permission.) Right, Increase in Schwann cell proliferation during wallerian degeneration is paralleled by an increase in K⫹ channel activity at the Schwann cell soma. (From Chiu, S. Y.: Changes in excitable membrane properties in Schwann cells of adult rabbit sciatic nerves following nerve transection. J. Physiol. [Lond.] 396:173, 1988, with permission.)

sheath act in combination to program channel expression in the Schwann cells. Molecular Identity and Localization of Schwann Cell Kv and Kir Channels The first clue to the molecular identity of Kv channels in Schwann cells came from mRNA analysis of sciatic nerves. Because axons are commonly thought to lack transcripts, mRNA extracted from sciatic nerves probably has a predominant Schwann cell origin. Kv1.1 and Kv1.2 mRNAs have been detected in mammalian sciatic nerves.35 Sobko and colleagues116 carried out a more exhaustive analysis of Kv channel identity in Schwann cells, at the protein level, in Schwann cells cultured from neonatal mouse sciatic nerves. Under these conditions, Schwann cells express four members of the Kv1 subfamily (Kv1.1, Kv1.2, Kv1.4, and Kv1.5), one member of the Kv2 subfamily (Kv2.1), and two members of the Kv3 subfamily (Kv3.1 and Kv3.2). These K⫹ channels are also detected in Western blots of sciatic nerves; however, an axonal contribution to the signal in the Western blots cannot be ruled out.116 Electron microscopic immunohistochemistry shows that Kv1.1 and Kv1.5 are differentially localized in a mature myelinating Schwann cell; Kv1.5 is localized to the Schwann cell membrane near the nodes of Ranvier and in bands that run along the outer surface of the myelin, whereas Kv1.1 is present predominantly in the perinuclear regions and in intracellular compartments.81 Perinuclear location of Kv1.1 proteins in Schwann cells has also been

observed in remyelinating sciatic nerves following lysolecithin-induced demyelination.98 The identity of Kir channels has also been clarified in Schwann cells with the identification of Kir2.1 and Kir2.3 on the Schwann cell microvilli facing the nodal surface.81 Thus, as in axons, K⫹ channel localization in a fully differentiated Schwann cell is highly polarized. What are the molecular mechanisms responsible for the differential localization of K⫹ channels in myelinating Schwann cells? Of particular interest is Kv1.1. This channel in neurons is transported away from the cell soma (dorsal root ganglion) and inserted on the axon at the juxtaparanode, which can be contrasted with the situation in a mature myelinating Schwann cell, in which Kv1.1 is trapped in intracellular compartments and apparently fails to be transported to the cell surface. It is possible that the normal absence of Kv␤2 in Schwann cells98 fails to provide a chaperone for surface expression of Kv1.1 in Schwann cells. However, Kv1.5 shows efficient surface expression in Schwann cells81 in the absence of Kv␤2. It is known from transfection studies using cultured mammalian cell lines that different Kv␣ subunits have different surface expression efficiency77: the surface expression efficiency is good for Kv1.2 homotetramers but poor for Kv1.1 homotetramers, the latter being trapped in the ER. Interestingly, in neonatal sciatic nerves, most of the Kv1.5 subunits are involved in heteromulteric association with Kv1.2,116 a finding that might explain why Kv1.5 shows efficient surface expression on Schwann cells.81 It appears that Kv1.1

106

Function of the Peripheral Nervous System

subunits might exist in Schwann cells as homotetramers, which are inherently poorly transported and trapped in the intracellular compartments.81 Here, the comparison with Kv1.1 trafficking in axons is interesting. As discussed previously, Kv1.1 trafficking to the juxtaparanodes is not affected by genetic deletion of Kv␤2 proteins.78 Perhaps the efficient surface expression of Kv1.1 in the Kv␤2-null axons is the result of its association with Kv1.2, an association that might be lacking in Schwann cells. Functions of K⫹ Channels in Schwann Cells Three hypotheses concerning K⫹ channel functions in Schwann cells have been proposed. The first is that K⫹ channels are synthesized in Schwann cells and transferred to axons to relieve synthetic burden on the neurons. The second is that K⫹ channels in Schwann cells contribute to potassium buffering. The third is that K⫹ channels in Schwann cells modulate Schwann cell proliferation. Schwann Cell–to-Axon Transfer of Kv Channels. With regard to the first hypothesis of a Schwann cell–toaxon transfer of K⫹ channels, a minimum requirement for the hypothesis has been satisfied by the demonstration that Kv1.1/Kv1.2, two members of the axonal juxtaparanodal tripartite (Kv1.1/Kv1.2/Kv␤2), are expressed in Schwann cells.35,116 However, the present difficulty with this hypothesis, at least for Kv1.1, is that this protein appears to be retained in the intracellular compartments of Schwann cells. In contrast, the vesicular form of Kv1.1 proteins in Schwann cells might reflect proteins that are destined for transfer to axons; this transfer, if it exists, could be below our current detection limit. Kv1.2, however, should be expressed efficiently on the Schwann cell surface, but immunohistochemical localization of Kv1.2 in myelinating Schwann cells has not been examined. In squid giant axons, a bilateral transfer of large proteins between Schwann cells and axons has been demonstrated.50,126 Our laboratory, in collaboration with Albee Messing and Larry Wrabetz, is currently exploring transgenic techniques45 to test the transfer hypothesis by selectively expressing Kv1.1 in Schwann cells of the Kv1.1 global knockout mouse to see if juxtaparanodal expression of Kv1.1 can be restored. Potassium Buffering. The level of endoneurial fluid potassium is known to have a profound effect on the excitability of peripheral nerves; depending on the extracellular potassium level, excitability of nerves can be either enhanced or depressed. Normal nerve activity, as well as damaged tissues, could be sources of extracellular potassium. Abnormal elevation of endoneurial potassium could be responsible for positive symptoms such as paresthesias and pain in peripheral neuropathies caused by spontaneous firing of axons,124 and an elevation of endoneurial potassium concentration has been demonstrated in demyelinated and regenerating nerves.72 Glial

cells have traditionally been suggested to play a role in potassium buffering.68,88 Transport systems such as Na⫹,K⫹-ATPase on glial cells can transport potassium ions into glial cells and lower the extracellular potassium level. In addition, the various Kv and Kir channels described on Schwann cells could make an important contribution to potassium buffering. In the peripheral nerves, activitydependent release of potassium is spatially homogeneous in neonates before myelin formation and becomes highly localized to the nodal region once myelin is formed. The expression of Kv and Kir channels in Schwann cells appears to be functionally matched to the developmental change in potassium release. In neonates, activitydependent release of axonal K⫹ is large because of the participation of the entire axonal surface in electrogenesis; correspondingly, Schwann cell expression of Kir and Kv channels is large in neonates. As myelination proceeds, Schwann cell expression of both K⫹ channels is reduced, consistent with a marked reduction in K⫹ released per unit axonal length as the myelin sheath is deposited on the axons. However, theoretical calculations have shown that the intense and highly focal ionic flux at the nodal regions, coupled with a restricted extracellular volume at the nodalparanodal-juxtaparanodal compartments, could produce significant local accumulation of K⫹ ions.28 The Schwann cell microvilli facing the nodal membrane, with a large surface-to-volume ratio, are an ideal system for absorbing K⫹ ions released during activity (note that, even though Kv channels are absent at the node, the leakage channels that help repolarize an action potential are likely potassium channels that mediate K⫹ efflux during the repolarization phase). Once inside the Schwann cells, the K⫹ ions could flow down the concentration gradient and exit elsewhere, away from the node. The Kir2.1 and Kir2.3 channels on the microvilli have been hypothesized to mediate the nodal absorption, whereas the Kv1.5 channels located on the outer myelin surface have been suggested to divert the absorbed K⫹ ions and release them away from the node.81,138 Testing this model with global genetic ablation of Kir2.1 is hampered by lethality in mutant mice at 5 to 10 hours after birth because of cleft palate.142 Functional Linkage between Ion Channels and Proliferation. Insights into channel function in Schwann cells can be gained by examining channel expression at various stages of the life cycle of a Schwann cell. Myelinogenesis and proliferation, two of the best studied features of Schwann cell biology, undergo characteristic changes during Schwann cell–axon interactions. During the first postnatal week, there is a wave of Schwann cell proliferation as cells contact growing axons.18,136,141 This initial proliferation soon subsides and gives way to myelinogenesis as cells are committed to a 1:1 relation with axons destined to become myelinated. As described in previous sections, patch-clamp analyses have shown that

Channel Function in Mammalian Axons and Support Cells

the Kv and Kir channels cease to be active on the Schwann cell soma as the cells leave the proliferation phase and enter the myelinogenic phase. The parallel reduction in K⫹ channel activity and Schwann cell proliferation suggests a functional linkage between the two (see Fig. 4–6, left). This functional linkage was explored in a study in which quiescent, adult Schwann cells were driven into proliferation by nerve transection.26 Here again, the K⫹ channel activity and proliferation increase in parallel (see Fig. 4–6, right). Blocking Kv channels with quinine, TEA, and 4-AP also, in most cases, reversibly blocks Schwann cell proliferation.38 Finally, application of Schwann cell mitogens (glial growth factor and myelin and axolemmal fragments) to cultured Schwann cells enhances Kv channel activity.140 Taken together, these studies suggest that potassium channel activity is an important early signal that modulates the transition of Schwann cells from a proliferative to a myelinogenic phase. The mechanism underlying this signal transduction remains unclear. One possibility is that Kv channel activity alters the resting potential of the Schwann cell,38 which produces ion fluxes that alter the intracellular concentration of second messengers such as sodium and calcium, which are thought to be important for mitosis.117 Since the discovery that K⫹ channels are linked to Schwann cell proliferation, research has focused on identifying the molecular species of K⫹ channels responsible for modulating Schwann cell proliferation. As discussed in previous sections, molecular biology has identified at least seven Kv channels and two Kir channels in Schwann cells.35,81,82,116 Which of these K⫹ channel subtypes are key modulators of Schwann cell proliferation? One approach is to examine the antiproliferative effects of K⫹ channel blockers specific for certain K⫹ channel subtypes. Besides the broad-spectrum K⫹ channel blockers such as TEA and 4-AP, various venom toxins (e.g., the dendrotoxin family) have been found to block certain K⫹ channel subtypes specifically. Based on the lack of an antiproliferative effect of dendrotoxin-I116 and dendrotoxin-␣,89 Kv1.1 and Kv1.2 have been excluded as possible candidates. Of the remaining Kv channels, Kv2.1 and Kv1.5 are strong contenders as important modulators of Schwann cell proliferation. The evidence comes from both CNS and PNS studies. In the CNS, astrocyte proliferation is also linked functionally to Kv channel activity. Using an antisense oligodeoxynucleotide technique, MacFarlane and Sontheimer74 found that Kv1.5 is the major Kv channel expressed in cultured astrocytes. Furthermore, inhibition of Kv1.5 protein expression by antisense oligodeoxynucleotides significantly inhibits astrocyte proliferation.74 The next interesting question that MacFarlane and Sontheimer74 addressed is whether the reduction in proliferation accompanying normal astrocyte differentiation is due to a reduction in protein expression level for Kv1.5 or a reduction in the activity of an existing pool of Kv1.5

107

channels. The correct picture that is emerging appears to be the latter one. How might the activity of existing Kv1.5 channels be modulated without a change in the protein expression level? MacFarlane and Sontheimer74 identified the key to be channel phosphorylation; phosphorylation of the Kv1.5 channel increases channel activity and increases proliferation, whereas dephosphorylation of the channel decreases its activity and decreases proliferation. What, then, controls phosphorylation of Kv channels in glial cells? Immunoprecipitation shows that Kv1.5 is associated with the Src family of protein tyrosine kinases, and this association remains unaltered during astrocyte proliferation.74 Similarly in Schwann cells, Kv1.5 and Kv2.1 have been shown to be physically associated with a Src family tyrosine kinase, which is probably Fyn.115 Indeed, in patchclamp studies of cultured Schwann cells, pharmacologic inhibition of tyrosine kinases inhibits Kv channel activity, and artificial activation of tyrosine kinases increases Kv channel activity.93,115 Taken together, the emerging picture is that certain Kv channels (such as Kv1.5/Kv2.1) and certain Src tyrosine kinases (such as Fyn) coexist physically as a modulatory complex in Schwann cells, that these Kv channels are constitutively phosphorylated by Src tyrosine kinase, and that an understanding of the endogenous factors that regulate Src tyrosine activity might be the key to understand how Kv channel activity modulates Schwann cell proliferation during development. It is important to emphasize that this model for linking K⫹ channels to Schwann cell proliferation rests mostly on tissue culture and explant studies. The central question that remains unanswered is whether Kv channels modulate Schwann cell proliferation in the more complex setting of a whole animal. Nevertheless, two intriguing in vivo studies using global gene knockouts suggest that the in vitro model might be applicable in vivo. First, Peretz and co-workers92 examined whether a particular tyrosine phosphatase, protein tyrosine phosphatase ␧ (PTP␧), might be an in vivo factor regulating phosphorylation of the Kv1.5/Kv2.1/Src complex in Schwann cells. In general, tyrosine phosphatases counter the action of Src tyrosine protein kinases, causing dephosphorylation of Kv1.5 and reducing its channel activity. Consistent with this, global genetic ablation of PTP␧ in mice leads to hyperphosphorylation of Kv1.5 and Kv2.1 proteins and enhancement of Kv channel activity in Schwann cells.92 More importantly, this increased Kv1.5/Kv2.1 channel activity presumably prolongs the proliferative phase of Schwann cells, thereby delaying their entrance into the myelinogenic phase and causing a transient hypomyelination in the PTP␧-null sciatic nerves shortly after birth.92 Interestingly, the PTP␧null mice exhibit normal myelination as adults, suggesting that other PTP species might be present to downregulate Kv channel activity in the adult, a prerequisite for myelination.92 Also, patch-clamp recordings reveal a large Kv current in Schwann cells of dysmyelinating mutant Trembler

108

Function of the Peripheral Nervous System

mice,27 suggesting that pathologic hyperphosphorylation of Kv channels in adult Schwann cells might be incompatible with myelination. Whether Kv1.5 is the Schwann cell channel species expressed in the Trembler Schwann cell is unclear. Clinically, pathologic upregulation of Kv channel activity in Schwann cells might cause schwannomas, and blockage of Kv channels has been suggested as a possible therapeutic intervention to slow the growth of this type of tumor.60,107 Second, genetic ablation of K⫹ channels in mice has in some cases produced hypomyelination. A striking example can be found in genetic ablation of a Kir channel (Kir4.1), which causes severe hypomyelination in the CNS axons.85 Kir4.1 is expressed in oligodendrocytes, the myelinating glia of CNS axons, and deleting Kir4.1 channels from oligodendrocytes results in a depolarized resting membrane potential and immature morphology in these cells.85 Global genetic ablation in mice of Kir2.1, the Kir species found in Schwann cells, results in lethality, which has prevented any study of possible perturbation on myelination.142 Interestingly, mutation in Kir2.1 in humans causes Andersen’s syndrome (AS), a rare disorder characterized by periodic paralysis and cardiac arrhythmias.3,94 Whether patients with AS have dysmyelination in the PNS remains unexplored. Kir2.1 channels maintain a stable resting potential that is important for cell function,59 and it remains possible that mutation of this channel causes a deviation of the resting potential of Schwann cells to a pathologic level that is incompatible with normal myelination. The limitation of the global knockout is the lack of tissue specificity, and in some cases lethality, a limitation that can be ameliorated by the use of the Cre-Lox recombinase system to limit gene inactivation to Schwann cells. This technique has recently been successfully applied to selectively ablate non–ion channel genes in Schwann cells,46 and selective inactivation of ion channel genes in Schwann cells using this technique will probably be the next step for shedding light on in vivo functions of K⫹ channels in Schwann cells.

3.

4. 5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

ACKNOWLEDGMENTS

16.

Work in the author’s laboratory was supported by grants from the National Institutes of Health, the National Multiple Sclerosis Society, and a Pew Scholar Award in the Biomedical Sciences.

17.

REFERENCES

19.

1. Aelman, J. P., Bond, C. T., Pessia, M., and Maylie, J.: Episodic ataxia results from voltage-dependent potassium channels with altered functions. Neuron 15:1449, 1995. 2. Amedee, T., Ellie, E., Dupouy, B., and Vincent, J. D.: Voltagedependent calcium and potassium channels in Schwann cells

20.

18.

cultured from dorsal root ganglia of the mouse. J. Physiol. (Lond.) 441:35, 1991. Andelfinger, G., Tapper, A. R., Welch, R. C., et al.: KCNJ2 mutation results in Andersen syndrome with sex-specific cardiac and skeletal muscle phenotypes. Am. J. Hum. Genet. 71:663, 2002. Arroyo, E. J., and Scherer, S. S.: On the molecular architecture of myelinated fibers. Histochem. Cell Biol. 113:1, 2000. Arroyo, E. J., Xu, Y. T., Zhou, L., et al.: Myelinating Schwann cells determine the internodal localization of Kv1.1, Kv1.2, Kv␤2, and Caspr. J. Neurocytol. 28:333, 1999. Baker, M., Bostock, H., Grafe, P., and Martius, P.: Function and distribution of three types of rectifying channel in rat spinal root myelinated axons. J. Physiol. (Lond.) 383:45, 1987. Baker, M., Howe, J. R., and Ritchie, J. M.: Two types of 4-aminopyridine-sensitive potassium current in rabbit Schwann cells. J. Physiol. (Lond.) 464:321, 1993. Barrett, E. F., and Barrett, J. N.: Intracellular recording from vertebrate myelinated axons: mechanism of the depolarizing afterpotential. J. Physiol. (Lond.) 323:117, 1982. Bennett, V., and Lambert, S.: Physiological roles of axonal ankyrins in survival of premyelinated axons and localization of voltage-gated sodium channels. J. Neurocytol. 28:303, 1999. Benoit, E., Corbier, A., and Dubois, J. M.: Evidence for two transient sodium currents in the frog node of Ranvier. J. Physiol. (Lond.) 361:339, 1985. Berghs, S., Aggujaro, D., Dirkx, R., et al.: Beta IV spectrin, a new spectrin localized at axon initial segments and nodes of Ranvier in the central and peripheral nervous system. J. Cell Biol. 151:985, 2000. Bever, C. T. Jr., Young, D., Anderson, P. A., et al.: The effects of 4-aminopyridine in multiple sclerosis patients: results of a randomized, placebo-controlled, double-blind, concentration-controlled, crossover trial. Neurology 44:1054, 1994. Bhat, M. A., Rios, J. C., Lu, Y., et al.: Axon-glia interactions and the domain organization of myelinated axons requires Neurexin IV/Caspr/Paranodin. Neuron 30:369, 2001. Boiko, T., Rasband, M. N., Levinson, S. R., et al.: Compact myelin dictates the differential targeting of two sodium channel isoforms in the same axon. Neuron 30:91, 2001. Bostock, H., Sears, T. A., and Sherratt, R. M.: The effects of 4-aminopyridine and tetraethylammonium ions on normal and demyelinated mammalian nerve fibres. J. Physiol. (Lond.) 313:301, 1981. Boyle, M. E., Berglund, E. O., Murai, K. K., et al.: Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30:385, 2001. Brismar, T.: Potential clamp analysis of membrane currents in rat myelinated nerve fibres. J. Physiol. (Lond.) 298:171, 1980. Brown, M. J., and Asbury, A. K.: Schwann cell proliferation in the postnatal mouse: timing and topography. Exp. Neurol. 74:170, 1981. Browne, D. L., Gancher, S. T., Nutt, J. G., et al.: Episodic ataxia/myokymia syndrome is associated with point mutations in the human potassium channel gene, KCNA1 [see comments]. Nat. Genet. 8:136, 1994. Burgess, D. L., Kohrman, D. C., Galt, J., et al.: Mutation of a new sodium channel gene, Scn8a, in the mouse mutant “motor endplate disease.” Nat. Genet. 10:461, 1995.

Channel Function in Mammalian Axons and Support Cells 21. Caldwell, J. H., Schaller, K. L., Lasher, R. S., et al.: Sodium channel Na(v)1.6 is localized at nodes of Ranvier, dendrites, and synapses. Proc. Natl. Acad. Sci. U. S. A. 97:5616, 2000. 22. Campomanes, C. R., Carroll, K. I., Manganas, L. N., et al.: Kv beta subunit oxidoreductase activity and Kv1 potassium channel trafficking. J. Biol. Chem. 277:8298, 2002. 23. Chiu, S. Y.: Inactivation of sodium channels: second order kinetics in myelinated nerve. J. Physiol. (Lond.) 273:573, 1977. 24. Chiu, S. Y.: Asymmetry currents in the mammalian myelinated nerve. J. Physiol. (Lond.) 309:499, 1980. 25. Chiu, S. Y.: Sodium currents in axon-associated Schwann cells from adult rabbits. J. Physiol. (Lond.) 386:181, 1987. 26. Chiu, S. Y.: Changes in excitable membrane properties in Schwann cells of adult rabbit sciatic nerves following nerve transection. J. Physiol. (Lond.) 396:173, 1988. 27. Chiu, S. Y.: Patch-clamp studies on myelin-deficient Schwann cells from mutant trembler mice [abstract]. Soc. Neurosci. Abstr.14:48.2, 1988. 28. Chiu, S. Y.: Functions and distribution of voltage-gated sodium and potassium channels in mammalian Schwann cells. Glia 4:541, 1991. 29. Chiu, S. Y.: Channel function in mammalian axons and support cells. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 94, 1993. 30. Chiu, S. Y.: Differential expression of sodium channels in acutely isolated myelinating and non-myelinating Schwann cells of rabbits. J. Physiol. (Lond.) 470:485, 1993. 31. Chiu, S. Y., and Ritchie, J. M.: Potassium channels in nodal and internodal axonal membrane of mammalian myelinated fibres. Nature 284:170, 1980. 32. Chiu, S. Y., and Ritchie, J. M.: Evidence for the presence of potassium channels in the paranodal region of acutely demyelinated mammalian single nerve fibres. J. Physiol. (Lond.) 313:415, 1981. 33. Chiu, S. Y., and Ritchie, J. M.: On the physiological role of internodal potassium channels and the security of conduction in myelinated nerve fibres. Proc. R. Soc. Lond. B Biol. Sci. 220:415, 1984. 34. Chiu, S. Y., Ritchie, J. M., Rogart, R. B., and Stagg, D.: A quantitative description of membrane currents in rabbit myelinated nerve. J. Physiol. (Lond.) 292:149, 1979. 35. Chiu, S. Y., Scherer, S. S., Blonski, M., et al.: Axons regulate the expression of Shaker-like potassium channel genes in Schwann cells in peripheral nerve. Glia 12:1, 1994. 36. Chiu, S. Y., Schrager, P., and Ritchie, J. M.: Neuronal-type Na⫹ and K⫹ channels in rabbit cultured Schwann cells. Nature 311:156, 1984. 37. Chiu, S. Y., and Schwarz, W.: Sodium and potassium currents in acutely demyelinated internodes of rabbit sciatic nerves. J. Physiol. (Lond.) 391:631, 1987. 38. Chiu, S. Y., and Wilson, G. F.: The role of potassium channels in Schwann cell proliferation in Wallerian degeneration of explant rabbit sciatic nerves. J. Physiol. (Lond.) 408:199, 1989. 39. Chiu, S. Y., Zhou, L., Zhang, C. L., and Messing, A.: Analysis of potassium channel functions in mammalian axons by gene knockouts. J. Neurocytol. 28:349, 1999. 40. Chouinard, S. W., Wilson, G. F., Schlimgen, A. K., and Ganetzky, B.: A potassium channel beta subunit related to

41.

42.

43.

44.

45.

46.

47. 48.

49.

50.

51. 52.

53.

54. 55. 56.

57. 58.

59.

109

the aldo-keto reductase superfamily is encoded by the Drosophila hyperkinetic locus. Proc. Natl. Acad. Sci. U. S. A. 92:6763, 1995. Dodge, F. A., and Frankenhaeuser, B.: Membrane currents in isolated frog nerve fibre under voltage clamp conditions. J. Physiol. (Lond.) 143:76, 1958. Doyle, D. A., Morais, C. J., Pfuetzner, R. A., et al.: The structure of the potassium channel: molecular basis of K⫹ conduction and selectivity. Science 280:69, 1998. Duchen, L. W., and Stefani, E.: Electrophysiological studies of neuromuscular transmission in hereditary “motor end-plate disease” of the mouse. J. Physiol. (Lond.) 212:535, 1971. Einheber, S., Zanazzi, G., Ching, W., et al.: The axonal membrane protein Caspr, a homologue of neurexin IV, is a component of the septate-like paranodal junctions that assemble during myelination. J. Cell Biol. 139:1495, 1997. Feltri, M. L., D’Antonio, M., Quattrini, A., et al.: A novel P0 glycoprotein transgene activates expression of lacZ in myelinforming Schwann cells. Eur. J. Neurosci. 11:1577, 1999. Feltri, M. L., Graus, P. D., Previtali, S. C., et al.: Conditional disruption of beta 1 integrin in Schwann cells impedes interactions with axons. J. Cell Biol. 156:199, 2002. Goldin, A. L.: Diversity of mammalian voltage-gated sodium channels. Ann. N. Y. Acad. Sci. 868:38, 1999. Goldin, A. L., Barchi, R. L., Caldwell, J. H., et al.: Nomenclature of voltage-gated sodium channels. Neuron 28:365, 2000. Gong, B., Rhodes, K. J., Bekele-Arcuri, Z., and Trimmer, J. S.: Type I and type II Na(⫹) channel alpha-subunit polypeptides exhibit distinct spatial and temporal patterning, and association with auxiliary subunits in rat brain. J. Comp. Neurol. 412:342, 1999. Grossfeld, R. M., Klinge, M. A., Lieberman, E. M., and Stewart, L. C.: Axon-glia transfer of a protein and a carbohydrate. Glia 1:292, 1988. Hille, B., Armstrong, C. M., and MacKinnon, R.: Ion channels: from idea to reality. Nat. Med. 5:1105, 1999. Hodgkin, A. L., and Huxley, A. F.: A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. (Lond.) 117:500, 1952. Horackova, M., Nonner, W., and Stampfli, R.: Action potentials and voltage-clamp currents of single rat Ranvier nodes [abstract]. Proc. Int. Union Physiol. 7:198, 1968. Isom, L. L., and Catterall, W. A.: Na⫹ channel subunits and Ig domains. Nature 383:307, 1996. Isom, L. L., De Jongh, K. S., and Catterall, W. A.: Auxiliary subunits of voltage-gated ion channels. Neuron 12:1183, 1994. Isom, L. L., Ragsdale, D. S., De Jongh, K. S., et al.: Structure and function of the beta 2 subunit of brain sodium channels, a transmembrane glycoprotein with a CAM motif. Cell 83:433, 1995. Jan, L. Y., and Jan, Y. N.: Cloned potassium channels from eukaryotes and prokaryotes. Annu. Rev. Neurosci. 20:91, 1997. Jonas, P., Brau, M. E., Hermsteiner, M., and Vogel, W.: Single-channel recording in myelinated nerve fibers reveals one type of Na channel but different K channels. Proc. Natl. Acad. Sci. U. S. A. 86:7238, 1989. Jongsma, H. J., and Wilders, R.: Channelopathies: Kir2.1 mutations jeopardize many cell functions. Curr. Biol. 11:R747, 2001.

110

Function of the Peripheral Nervous System

60. Kamleiter, M., Hanemann, C. O., Kluwe, L., et al.: Voltagedependent membrane currents of cultured human neurofibromatosis type 2 Schwann cells. Glia 24:313, 1998. 61. Kaplan, M. R., Meyer-Franke, A., Lamber, S., et al.: Induction of sodium channel clustering by oligodendrocytes. Nature 386:724, 1997. 62. Kearney, J. A., Plummer, N. W., Smith, M. R., et al.: A gainof-function mutation in the sodium channel gene Scn2a results in seizures and behavioral abnormalities. Neuroscience 102:307, 1901. 63. Kocsis, J. D., Ruiz, J. A., and Waxman, S. G.: Maturation of mammalian myelinated fibers: changes in action-potential characteristics following 4-aminopyridine application. J. Neurophysiol. 50:449, 1983. 64. Kohrman, D. C., Smith, M. R., Goldin, A. L., et al.: A missense mutation in the sodium channel Scn8a is responsible for cerebellar ataxia in the mouse mutant jolting. J. Neurosci. 16:5993, 1996. 65. Konishi, T.: Voltage-dependent potassium channels in mouse Schwann cells. J. Physiol. (Lond.) 411:115, 1989. 66. Konishi, T.: cAMP-mediated expression of inwardly rectifying potassium channels in cultured mouse Schwann cells. Brain Res. 594:197, 1992. 67. Konishi, T.: Activity-dependent regulation of inwardly rectifying potassium currents in non-myelinating Schwann cells in mice. J. Physiol. (Lond.) 474:193, 1994. 68. Kuffler, S. W., Nicholls, J. G., and Orkand, R. K.: Physiological properties of glial cells in the central nervous system of amphibia. J. Neurophysiol. 29:768, 1966. 69. Lambert, S., Davis, J. O., and Bennett, V.: Morphogenesis of the node of Ranvier: co-clusters of ankyrin and ankyrinbinding integral proteins define early developmental intermediates. J. Neurosci. 17:7025, 1997. 70. Li, M., Jan, Y. N., and Jan, L. Y.: Specification of subunit assembly by the hydrophilic amino-terminal domain of the Shaker potassium channel. Science 257:1225, 1992. 71. Liu, C. J., Dib-Hajj, S. D., Black, J. A., et al.: Direct interaction with contactin targets voltage-gated sodium channel Na(v)1.9/NaN to the cell membrane. J. Biol. Chem. 276:46553, 2001. 72. Low, P. A.: Endoneurial potassium is increased and enhances spontaneous activity in regenerating mammalian nerve fibers-implications for neuropathic positive symptoms. Muscle Nerve 8:27, 1985. 73. Lustig, M., Zanazzi, G., Sakurai, T., et al.: Nr-CAM and neurofascin interactions regulate ankyrin G and sodium channel clustering at the node of Ranvier. Curr. Biol. 11:1864, 2001. 74. MacFarlane, S. N., and Sontheimer, H.: Modulation of Kv1.5 currents by Src tyrosine phosphorylation: potential role in the differentiation of astrocytes. J. Neurosci. 20:5245, 2000. 75. Malhotra, J. D., Kazen-Gillespie, K., Hortsch, M., and Isom, L. L.: Sodium channel beta subunits mediate homophilic cell adhesion and recruit ankyrin to points of cell-cell contact. J. Biol. Chem. 275:11383, 2000. 76. Manganas, L. N., Akhtar, S., Antonucci, D. E., et al.: Episodic ataxia type-1 mutations in the Kv1.1 potassium channel display distinct folding and intracellular trafficking properties. J. Biol. Chem. 276:49427, 2001. 77. Manganas, L. N., and Trimmer, J. S.: Subunit composition determines Kv1 potassium channel surface expression. J. Biol. Chem. 275:29685, 2000.

78. McCormack, K., Connor, J. X., Zhou, L., et al.: Genetic analysis of the mammalian K⫹ channel beta-subunit Kvbeta 2 (Kcnab2). J. Biol. Chem. 277:13219, 2002. 79. McCormack, T., and McCormack, K.: Shaker K⫹ channel ␤ subunits belong to an NAD(P)H-dependent oxidoreductase superfamily. Cell 79:1133, 1994. 80. Menegoz, M., Gaspar, P., Le, B. M., et al.: Paranodin, a glycoprotein of neuronal paranodal membranes. Neuron 19:319, 1997. 81. Mi, H., Deerinck, T. J., Ellisman, M. H., and Schwarz, T. L.: Differential distribution of closely related potassium channels in rat Schwann cells. J. Neurosci. 15:3761, 1995. 82. Mi, H. Y., Deerinck, T. J., Jones, M., et al.: Inwardly rectifying K⫹ channels that may participate in K⫹ buffering are localized in microvilli of Schwann cells. J. Neurosci. 16:2421, 1996. 83. Miller, A.: Current and investigational therapies used to alter the course of disease in multiple sclerosis. South. Med. J. 90:367, 1997. 84. Morais-Cabral, J. H., Zhou, Y., and MacKinnon, R.: Energetic optimization of ion conduction rate by the K⫹ selectivity filter. Nature 414:37, 2001. 85. Neusch, C., Rozengurt, N., Jacobs, R. E., et al.: Kir4.1 potassium channel subunit is crucial for oligodendrocyte development and in vivo myelination. J. Neurosci. 21:5429, 2001. 86. Nonner, W.: A new voltage clamp method for Ranvier nodes. Pflugers Arch. 309:176, 1969. 87. Oh, Y., and Waxman, S. G.: The beta 1 subunit mRNA of the rat brain Na⫹ channel is expressed in glial cells. Proc. Natl. Acad. Sci. U. S. A. 91:9985, 1994. 88. Orkand, R. K.: Glial electrophysiology and transport. Ann. N. Y. Acad. Sci. 633:245, 1991. 89. Pappas, C. A., and Ritchie, J. M.: Effect of specific ion channel blockers on cultured Schwann cell proliferation. Glia 22:113, 1998. 90. Parkinson, N. J., Olsson, C. L., Hallows, J. L., et al.: Mutant beta-spectrin 4 causes auditory and motor neuropathies in quivering mice. Nat. Genet. 29:61, 2001. 91. Pedraza, L., Huang, J. K., and Colman, D. R.: Organizing principles of the axoglial apparatus. Neuron 30:335, 2001. 92. Peretz, A., Gil-Henn, H., Sobko, A., et al.: Hypomyelination and increased activity of voltage-gated K⫹ channels in mice lacking protein tyrosine phosphatase ␧. EMBO J. 19:4036, 2000. 93. Peretz, A., Sobko, A., and Attali, B.: Tyrosine kinases modulate K⫹ channel gating in mouse Schwann cells. J. Physiol. (Lond.) 519(pt. 2):373, 1999. 94. Plaster, N. M., Tawil, R., Tristani-Firouzi, M., et al.: Mutations in Kir2.1 cause the developmental and episodic electrical phenotypes of Andersen’s syndrome. Cell 105:511, 2001. 95. Poliak, S., Gollan, L., Martinez, R., et al.: Caspr2, a new member of the neurexin superfamily, is localized at the juxtaparanodes of myelinated axons and associates with K⫹ channels. Neuron 24:1037, 1999. 96. Pongs, O., Leicher, T., Berger, M., et al.: Functional and molecular aspects of voltage-gated K⫹ channel beta subunits. Ann. N. Y. Acad. Sci. 868:344, 1999. 97. Quick, D. C., Kennedy, W. R., and Donaldson, L.: Dimensions of myelinated nerve fibers near the motor and sensory terminals in cat tenuissimus muscles. Neuroscience 4:1089, 1979.

Channel Function in Mammalian Axons and Support Cells 98. Rasband, M., Trimmer, J. S., Schwarz, T. L., et al.: Potassium channel distribution, clustering, and function in remyelinating rat axons. J. Neurosci. 18:36, 1998. 99. Rasband, M. N., and Trimmer, J. S.: Developmental clustering of ion channels at and near the node of Ranvier [review]. Dev. Biol. 236:5, 2001. 100. Rasband, M. N., and Trimmer, J. S.: Subunit composition and novel localization of K⫹ channels in spinal cord. J. Comp. Neurol. 429:166, 2001. 101. Ratcliffe, C. F., Westenbroek, R. E., Curtis, R., and Catterall, W. A.: Sodium channel beta1 and beta3 subunits associate with neurofascin through their extracellular immunoglobulin-like domain. J. Cell Biol. 154:427, 2001. 102. Revenko, S. V., Timin, E. N., and Khodorov, B. I.: Nerve impulse conduction from the myelinated portion of the axon to the nonmyelinated terminal [in Russian]. Biofizika 18:1074, 1973. 103. Rhodes, K. J., Strassle, B. W., Monaghan, M. M., et al.: Association and colocalization of the Kv␤1 and Kv␤2 ␤-subunits with Kv1 ␣-subunits in mammalian K⫹ channel complexes. J. Neurosci. 17:8246, 1997. 104. Rieger, F., Pincon-Raymond, M., Lombet, A., et al.: Paranodal dysmyelination and increase in tetrodotoxin binding sites in the sciatic nerve of the motor end-plate disease (med/med) mouse during postnatal development. Dev. Biol. 101:401, 1984. 105. Ritchie, J. M.: Sodium-channel turnover in rabbit cultured Schwann cells. Proc. R. Soc. Lond. B Biol. Sci. 233:423, 1988. 106. Ritchie, J. M., and Rogart, R. B.: Density of sodium channels in mammalian myelinated nerve fibers and nature of the axonal membrane under the myelin sheath. Proc. Natl. Acad. Sci. U. S. A. 74:211, 1977. 107. Rosenbaum, C., Kamleiter, M., Grafe, P., et al.: Enhanced proliferation and potassium conductance of Schwann cells isolated from NF2 schwannomas can be reduced by quinidine. Neurobiol. Dis. 7:483, 2000. 108. Rosenbluth, J.: Intramembranous particle distribution at the node of Ranvier and adjacent axolemma in myelinated axons of the frog brain. J. Neurocytol. 5:731, 1976. 109. Schauf, C. L., and Davis, F. A.: Impulse conduction in multiple sclerosis: a theoretical basis for modification by temperature and pharmacological agents. J. Neurol. Neurosurg. Psychiatry 37:152, 1974. 110. Schwarz, J. R., Reid, G., and Bostock, H.: Action potentials and membrane currents in the human node of Ranvier. Pflugers Arch. 430:283, 1995. 111. Sherratt, R. M., Bostock, H., and Sears, T. A.: Effects of 4-aminopyridine on normal and demyelinated mammalian nerve fibres. Nature 283:570, 1980. 112. Shi, G., Nakahira, K., Hammond, S., et al.: ␤ Subunits promote K⫹ channel surface expression through effects early in biosynthesis. Neuron 16:843, 1996. 113. Shrager, P., Chiu, S. Y., and Ritchie, J. M.: Voltage-dependent sodium and potassium channels in mammalian cultured Schwann cells. Proc. Natl. Acad. Sci. U. S. A. 82:948, 1985. 114. Smart, S. L., Lopantsev, V., Zhang, C. L., et al.: Deletion of the Kv1.1 potassium channel causes epilepsy in mice. Neuron 20:809, 1998. 115. Sobko, A., Peretz, A., and Attali, B.: Constitutive activation of delayed-rectifier potassium channels by a src family tyrosine kinase in Schwann cells. EMBO J. 17:4723, 1998.

111

116. Sobko, A., Peretz, A., Shirihai, O., et al.: Heteromultimeric delayed-rectifier K⫹ channels in Schwann cells: developmental expression and role in cell proliferation. J. Neurosci. 18:10398, 1998. 117. Soltoff, S. P., and Cantley, L. C.: Mitogens and ion fluxes. Annu. Rev. Physiol. 50:207, 1988. 118. Sontheimer, H., Black, J. A., and Waxman, S. G.: Voltagegated Na⫹ channels in glia: properties and possible functions. Trends Neurosci. 19:325, 1996. 119. Spampanato, J., Escayg, A., Meisler, M. H., and Goldin, A. L.: Functional effects of two voltage-gated sodium channel mutations that cause generalized epilepsy with febrile seizures plus type 2. J. Neurosci. 21:7481, 2001. 120. Srinivasan, Y., Elmer, L., Davis, J., et al.: Ankyrin and spectrin associate with voltage-dependent sodium channels in brain. Nature 333:177, 1988. 121. Stefoski, D., Davis, F. A., Faut, M., and Schauf, C. L.: 4-Aminopyridine improves clinical signs in multiple sclerosis. Ann. Neurol. 21:71, 1987. 122. Tait, S., Gunn-Moore, F., Collinson, J. M., et al.: An oligodendrocyte cell adhesion molecule at the site of assembly of the paranodal axo-glial junction. J. Cell Biol. 150:657, 2000. 123. Targ, E. F., and Kocsis, J. D.: 4-Aminopyridine leads to restoration of conduction in demyelinated rat sciatic nerve. Brain Res. 328:358, 1985. 124. Torebjork, H. E., Ochoa, J. L., and McCann, F. V.: Paresthesiae: abnormal impulse generation in sensory nerve fibres in man. Acta Physiol. Scand. 105:518, 1979. 125. Trimmer, J. S.: Regulation of ion channel expression by cytoplasmic subunits. Curr. Opin. Neurobiol. 8:370, 1998. 126. Tytell, M., and Lasek, R. J.: Glial polypeptides transferred into the squid giant axon. Brain Res. 324:223, 1984. 127. Vabnick, I., Novakovic, S. D., Levinson, S. R., et al.: The clustering of axonal sodium channels during development of the peripheral nervous system. J. Neurosci. 16:4914, 1996. 128. Vabnick, I., Trimmer, J. S., Schwarz, T. L., et al.: Dynamic potassium channel distributions during axonal development prevent aberrant firing patterns. J. Neurosci. 19:747, 1999. 129. Verkhratsky, A., Hoppe, D., and Kettenmann, H.: K⫹ channel properties in cultured mouse Schwann cells: dependence on extracellular K⫹. J. Neurosci. Res. 28:210, 1991. 130. Vogel, W., and Schwarz, J. R.: Voltage-clamp studies in axons: macroscopic and single-channel currents. In Waxman, S. G., Kocsis, J. D., and Stys, P. K. (eds.): The Axon: Structure, Function and Pathophysiology. New York, Oxford University Press, p. 257, 1995. 131. Wang, H., Allen, M. L., Grigg, J. J., et al.: Hypomyelination alters K⫹ channel expression in mouse mutants shiverer and Trembler. Neuron 15:1337, 1995. 132. Wang, H., Kunkel, D. D., Martin, T. M., et al.: Heteromultimeric K⫹ channels in terminal and juxtaparanodal regions of neurons. Nature 365:75, 1993. 133. Waxman, S. G.: Voltage-gated ion channels in axon: localization, function, and development. In Waxman, S. G., Kocsis, J. D., and Stys, P. K. (eds.): The Axon: Structure, Function and Pathophysiology. New York, Oxford University Press, p. 218, 1995. 134. Waxman, S. G., and Brill, M. H.: Conduction through demyelinated plaques in multiple sclerosis: computer simulations of facilitation by short internodes. J. Neurol. Neurosurg. Psychiatry 41:408, 1978.

112

Function of the Peripheral Nervous System

135. Waxman, S. G., and Ritchie, J. M.: Molecular dissection of the myelinated axon. Ann. Neurol. 33:121, 1993. 136. Webster, H. D., and Favilla, J. T.: Development of peripheral nerve fibers. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. P. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 329, 1984. 137. Westenbroek, R.E., Noebels, J.L., and Catterall, W.A.: Elevated expression of type II Na⫹ channels in hypomyelinated axons of shiverer mouse brain. J. Neurosci. 12:2259, 1992. 138. Wilson, G. F., and Chiu, S. Y.: Ion channels in axon and Schwann cell membranes at paranodes of mammalian myelinated fibers studied with patch clamp. J. Neurosci. 10:3263, 1990. 139. Wilson, G. F., and Chiu, S. Y.: Potassium channel regulation in Schwann cells during early developmental myelinogenesis. J. Neurosci. 10:1615, 1990. 140. Wilson, G. F., and Chiu, S. Y.: Mitogenic factors regulate ion channels in Schwann cells cultured from newborn rat sciatic nerve. J. Physiol. (Lond.) 470:501, 1993. 141. Wood, P. M., and Bunge, R. P.: Evidence that sensory axons are mitogenic for Schwann cells. Nature 256:662, 1975. 142. Zaritsky, J. J., Eckman, D. M., Wellman, G. C., et al.: Targeted disruption of Kir2.1 and Kir2.2 genes reveals the essential role of the inwardly rectifying K(⫹) current in K(⫹)-mediated vasodilation [see comments]. Circ. Res. 87:160, 2000.

143. Zerr, P., Adelman, J. P., and Maylie, J.: Episodic ataxia mutations in Kv1.1 alter potassium channel function by dominant negative effects or haploinsufficiency. J. Neurosci. 18:2842, 1998. 144. Zhang, C. L., Messing, A., and Chiu, S. Y.: Specific alteration of spontaneous GABAergic inhibition in cerebellar Purkinje cells in mice lacking the potassium channel Kv1.1. J. Neurosci. 19:2852, 1999. 145. Zhou, D., Lambert, S., Malen, P. L., et al.: AnkyrinG is required for clustering of voltage-gated Na channels at axon initial segments and for normal action potential firing. J. Cell Biol. 143:1295, 1998. 146. Zhou, L., and Chiu, S. Y.: Computer model for action potential propagation through branch point in myelinated nerves. J. Neurophysiol. 85:197, 2001. 147. Zhou, L., Messing, A., and Chiu, S. Y.: Determinants of excitability at transition zones in Kv1.1-deficient myelinated nerves. J. Neurosci. 19:5768, 1999. 148. Zhou, L., Zhang, C. L., Messing, A., and Chiu, S. Y.: Temperature-sensitive neuromuscular transmission in Kv1.1 null mice: role of potassium channels under the myelin sheath in young nerves. J. Neurosci. 18:7200, 1998. 149. Zhou, Y., Morais-Cabral, J. H., Kaufman, A., and MacKinnon, R.: Chemistry of ion coordination and hydration revealed by a K⫹ channel-Fab complex at 2.0 Å resolution. Nature 414:43, 2001.

5 Nerve Excitability Measures: Biophysical Basis and Use in the Investigation of Peripheral Nerve Disease MATTHEW C. KIERNAN, DAVID BURKE, AND HUGH BOSTOCK

Early Excitability Studies Nerve Excitability and Nerve Conduction Studies Compared Currently Used Measures of Nerve Excitability and Their Biophysical Basis Strength-Duration Behavior Recovery Cycles: Refractoriness, Superexcitability, and Late Subexcitability Threshold Electrotonus and Current-Threshold Relationship

Automated Recording of Multiple Nerve Excitability Properties Nerve Excitability Properties and Membrane Potential Effects of Ischemia on Axonal Excitability Modality and Regional Differences between Different Axons Changes in Axonal Membrane Potential in Peripheral Neuropathies

Nowadays, electrical stimulation of peripheral nerves for the purpose of clinical diagnosis nearly always employs supramaximal stimuli, to measure the velocity and amplitude of compound sensory or motor action potentials (see Chapter 35). This chapter is concerned with the alternative approach of using submaximal stimuli to determine the ease with which axons can be excited. Such studies of electrical excitability may seem novel, but they actually have a much longer history than velocity measurements, albeit a history tarnished with confusion caused by erroneous theories. More recent excitability studies are based on (and have contributed to) a much better understanding of axonal biophysics, and it is argued that they have the potential to make important contributions to our understanding of the pathophysiology of peripheral neuropathy.

EARLY EXCITABILITY STUDIES Fifty years ago, Hodgkin and Huxley’s revolutionary theory of nerve excitability36 and conduction was paralleled by an

Multifocal Motor Neuropathy and Axonal Hyperpolarization Uremic Neuropathy and Axonal Depolarization Impulse Conduction in Demyelinating Neuropathies Potential Role for Excitability Studies in the Neuropathy Clinic

independent revolution in the clinical electrophysiology of peripheral nerve, as measurements of excitability gave way to measurements of velocity. A 1952 textbook84 lists five methods of electrodiagnosis in neurology: electromyography, electroencephalography, psychogalvanic reflex, galvanic vertigo, and “the assessment of tissue excitability and what is termed ‘accommodation’.” Conduction velocity measurements are not mentioned (although an early clinical study35 had been published in 1948). The then “classical method” of excitability testing, which had been “the mainstay of electro-diagnosis for over a century,” involved testing nerves and muscles with “faradic” current from an induction coil, and “galvanic” current from a battery, to give short and long pulses, respectively.84 With the improvements of electronics during the Second World War, the simple galvanic-faradic test gave way to accurate determination of strength-duration (or intensitytime) curves using electronically generated rectangular pulses.87 For durations beyond the temps utile, threshold was a constant minimum (rheobase), while chronaxie was defined as the duration of a threshold current that was twice the rheobase. Strength-duration curves determined at the motor 113

114

Function of the Peripheral Nervous System

agreement as to standards has so far been reached.”84 It never was. The most important pre-1952 studies of accommodation in human nerves were probably Kugelberg’s thesis of 194453 and his demonstration of a dramatic increase in accommodation during ischemia, and decrease following release of ischemia, and the loss of accommodation associated with tetany or latent tetany.53–55 These early excitability studies helped to classify neuromuscular disorders as myopathic, neuropathic, or myelopathic, depending on whether they primarily affected muscle, nerve fiber, or nerve cell. During the 1950s, pioneering measurements of motor nerve conduction by Lambert and colleagues at Mayo Clinic, and of sensory nerve action potentials by Dawson, Scott, Sears, and Gilliatt at Queen Square, London (reviewed by Fowler30), set the scene for a rapid development of nerve conduction studies, and by the mid-1960s velocity measurements were being used to divide neuropathies into axonal and demyelinating types. Within 20 years, excitability studies were almost entirely displaced by conduction studies, which have held sway ever since.

point of a healthy muscle differ drastically from the responses of a denervated muscle, and were used in the clinical evaluation of the degree of muscle innervation (Fig. 5–1). The fact that normal motor point strength-duration curves resemble those obtained by stimulating the nerve trunk is simply due to the lower threshold of myelinated axon than muscle at the motor point. Unfortunately, however, the observation was misconstrued by Lapique, in his 1926 treatise on chronaxie.59 He formulated an electrical theory of neuromuscular transmission in which successful transmission required isochronism, or matching chronaxie between nerve and muscle. Lucas had already shown that nerve and muscle are not normally isochronic as early as 1906,67 and the principle of isochronism was soon soundly criticized.29,90 Nevertheless, Lapique’s fallacy lent measurements of chronaxie a spurious significance for many years, and isochronism persisted in medical texts as late as 1945.83 The other type of excitability measurement described in 1952 was accommodation, the tendency of a subthreshold current to slowly reduce nerve excitability. Studies of accommodation were also misguided by faulty theory. Hill’s theory of accommodation, published in 1936,33 was based in part on experiments on depolarized frog nerves, which led him to assume that accommodation to a steady current was always complete (so that the threshold to a superimposed test pulse should return to its original value) if the current was maintained long enough. According to this theory, accommodation was described by a single time constant, which could be deduced equally well from excitability tests made with current ramps, condenser discharges, or sinusoidal currents. For human nerves with normal resting potentials, however, accommodation is never more than partial (e.g., see Fig. 5–5 below) and Hill’s equations do not work well.4 It is therefore not surprising that the 1952 text stated: “Several methods of measuring and expressing the degree of accommodation present in a tissue have been suggested, but for technical reasons no

NERVE EXCITABILITY AND NERVE CONDUCTION STUDIES COMPARED Over the last 30 years there have been few significant changes to the neurophysiologic investigation of patients with suspected neuromuscular disorders. Motor and sensory nerve conduction studies, in combination with electromyography, have remained the methods of choice for the clinician investigating peripheral nerve function. Routine nerve conduction studies can document the presence of a neuropathy with or without resting conduction block, but they may provide little further insight into disease pathophysiology. Measurements of action potential amplitude and latency provide information only on the

100 16

we

De

80

ee

w

60 Volts

24

we e

40 N or m a

20

at

rv

20

ne

ek s

ks

ks

ed

Chronaxie v o lu

nta r y

move m

e nt l contra la teral muscle

Rheobase

4 1 0.2 0.2 0.08

Temps utile 0 0 0.01 0.03

0.1

0.3 1 3 Time (msec)

10

30

100 300

FIGURE 5–1 The normal strength-duration curve from a motor point is shown by the dashed line, with the other curves from the denervated muscle of the other hand. Determinations at different times during the reinnervation show the return of the strength-duration curve toward normal. For stimulus durations greater than the temps utile (arrow), the threshold is constant (i.e., the rheobase). Chronaxie is the stimulus duration at which threshold is twice the rheobase. (Adapted from Ochs, S.: Elements of Neurophysiology. New York, Wiley, 1965. Original data from Ritchie, A. R.: The electrical diagnosis of peripheral nerve injury. Brain 67:314, 1944.)

115

Nerve Excitability Measures

CURRENTLY USED MEASURES OF NERVE EXCITABILITY AND THEIR BIOPHYSICAL BASIS

number of conducting fibers and the conduction velocity of the fastest. However, conduction velocity is a nonspecific indicator of pathophysiology: It may be decreased by cooling, membrane depolarization or hyperpolarization, sodium channel blockade, axonal thinning, demyelination, or remyelination with short internodes. Measurements of axonal excitability are known to be sensitive to and can provide an indirect measure of resting membrane potential. Therefore, these measures provide complementary information to conventional nerve conduction studies. The currently preferred technique for assessing nerve excitability relies on the stimulus-based method of threshold tracking.18 With threshold tracking, changes in the intensity of the stimulus current required to generate a test potential of fixed amplitude are measured on line by a computer that in turn adjusts stimulus intensity to keep the amplitude of the subsequent test potential constant. “Threshold” in this context indicates the stimulus current required to produce a nerve or muscle action potential of target size (e.g., 40% of maximum). Threshold can be measured on line (“tracked”) during different maneuvers that alter nerve excitability. For example, when axons become hyperpolarized, the test potential will be smaller, and the computer will gradually increase stimulus intensity until the test potential has returned to its target size. The target is usually around 40% to 50% of maximum (because this is on the fast rising phase of the S-shaped stimulusresponse curve for the compound potential). “Threshold” is inversely related to axonal excitability, and can therefore provide an indirect measure of membrane potential.

The recent revival of nerve excitability studies by the authors and others has stemmed from basic research on the roles of different ion channels in myelinated axon.5 The principal ion channel types required to account for the electrical behavior of large myelinated axons in human peripheral nerves are indicated schematically in Figure 5–2. The contribution of the internodal axon compartment to action potential repolarization, afterpotentials, and slow excitability changes was first demonstrated by the microelectrode studies on amphibian axons of Ellen and John Barrett.7 A quantitative electrical model, showing how the nodal and internodal ion channels could interact to generate slow changes in excitability, was subsequently proposed on the basis of excitability measurements on human axons. 15 That model incorporated all the channel types shown in Figure 5–2 except for the persistent Na channels, which later excitability studies on human axons showed to be important in accounting for their surprisingly long strength-duration time constants, and for some important differences between motor and sensory axons.20 For further information on the complex electrical structure of myelinated axons, the reader is referred to recent reviews6,18,23,88 and papers describing equivalent electrical circuits and mathematical models.15,18,20,93,95 A detailed review of ion channel function is found in Chapter 4.

Na+

Nat Nap Ks

Kf

ATP

Ks

Ih

Lk

Kf

Ca2+ Na+

K+

Na+

K+

Na+

FIGURE 5–2 Schematic diagram of myelinated axon, showing principal components affecting electrical excitability. Nodal Na channels contribute both transient (Na t) and persistent (Na p) currents on depolarization: Na t is responsible for the regenerative upstroke of the action potential, and Na p affects subthreshold excitability. The nodal slow K conductance (Ks) is activated too slowly to contribute to action potential repolarization, but produces accommodation to subthreshold depolarizing currents and spike-frequency adaptation. Fast K channels (Kf) are concentrated in the juxtaparanodal region under the myelin, limit internodal depolarization during the spike, and so prevent nodal reexcitation by the depolarizing afterpotential. Hyperpolarization-activated cation current (Ih) limits post-tetanic hyperpolarization by the electrogenic sodium pump (Na,K-ATPase). There are also voltage-independent or leak conductances (Lk). The passive Na,Ca2 ion exchanger normally contributes to Ca2 homeostasis, but can be reversed with pathologic consequences during anoxia/ischemia. Ks, Kf, Lk, Na p, and Na,K-ATPase contribute to the resting potential.

116

Function of the Peripheral Nervous System

Here we focus on excitability properties, and introduce the electrical components in Figure 5–2 as they are reflected in excitability measurements. In addition to the measures featured in early excitability studies, namely strength-duration behavior and accommodation, another useful technique involves estimation of the “excitability cycle” or recovery cycle of excitability changes following an impulse. The phenomena of refractoriness and supernormality were described before the First World War.1 Clinical studies were initiated in the 1960s,32 and a trickle of subsequent studies have bucked the trend to concentrate exclusively on nerve conduction measurements.52 Of the many possible ways of measuring accommodation, the use of subthreshold rectangular current pulses to measure “threshold electrotonus”13 has been the most useful because the responses are normally closely related to electrotonus (the subthreshold voltage changes), and hence to the activity of voltage-dependent ion channels.12,13

Strength-Duration Behavior If a nerve membrane behaved simply as a passive electrical circuit before the threshold for excitation was reached, then the strength-duration measurements should result in

exponential membrane potential trajectories, and the strength-duration relationship should be described by Lapique’s59 equation: I  /[1  exp(t/)] where I is the threshold current for a pulse of duration t,  is the rheobase, and the strength-duration time constant  (defined as the ratio of threshold charge to rheobase for very brief stimuli76) is equal to chronaxie  loge2. However, the nerve membrane does not behave in this way, because sodium channels are activated before threshold is reached, giving rise to a local response. Instead the strength-duration relationship is better described by the earlier equation of Weiss103: Q  (t  ) where Q is the threshold charge (i.e., Q  It), and   chronaxie. This simple empirical relationship cannot be justified from first principles, but works remarkably well both for model axons following Hodgkin-Huxley type equations and for real rat and human axons11,21,71 (Fig. 5–3). The linearity of the relationship is such that estimates of the strength-duration time constant based on just two

50

A

Current (mA)

40 Sensory Motor

30 20 10 0 0.0

0.2

0.4

0.6

0.8

1.0

Stimulus duration (ms)

4

B

Sensory Motor

y = 0.76372 + 1.4122x r =0.997 y = 0.90940 + 2.1814x r =0.999

1

417 μs

2 540 μs

Charge (μC)

3

0 –1.0 –0.8 –0.6 –0.4 –0.2 0.0 0.2 0.4 Stimulus duration (ms)

0.6

0.8

1.0

FIGURE 5–3 A, Strength-duration curves for sensory and motor potentials of 30% of maximum recorded in the same experiment from the median nerve at the wrist. Arrows indicate chronaxies determined from the straight-line fits of the data in B to Weiss’s formula, using plots of threshold charge against stimulus duration. The time constant is given by the (negative) intercept on the duration axis, and the rheobase current is given by the slope of the regression line. (From Mogyoros, I., Kiernan, M. C., and Burke, D.: Strength-duration properties of human peripheral nerve. Brain 119:439, 1996, with permission from Oxford University Press.)

117

Nerve Excitability Measures

stimulus durations (echoing the earlier galvanic-faradic test) have been found to be just as reliable as estimates based on fitting a line to multiple points.71 Nevertheless, it should be remembered that Weiss’s equation is only approximate, and breaks down beyond the temps utile (see Fig. 5–1) when threshold is constant. It also overestimates threshold for currents of very short duration.66 The modeling of strength-duration behavior11 showed that  is linearly related to the passive nodal time constant, and therefore increases with the increased nodal capacitance in demyelination. However, because of its dependence on the subthreshold local response,  is also strongly dependent on sodium channels activated by subthreshold depolarization. Later modeling and excitability measurements20 provided evidence that differences in a persistent sodium conductance, and not differences in passive nodal properties, were responsible for the puzzling observation71,78 that sensory axons in human peripheral nerves have longer strength-duration time constants than the motor axons. The modeling accurately predicted an increase in , resulting from increased subthreshold activation of sodium channels, when nerves are depolarized.20

absolute refractoriness, axons are relatively refractory for up to 4.5 ms, during which greater current is required to generate an action potential. Refractoriness is dependent on fiber size and temperature: the smaller or cooler the fiber, the longer its duration. The refractory period is followed by a superexcitable (or supernormal) phase, characterized by a reduction in threshold occurring over a 10- to 15-ms interval. Finally, there is a phase of late subexcitability (or late subnormality) ending around 100 ms. These changes in threshold are associated with measurable changes in latency. Latency is initially increased during the refractory period, decreased during superexcitability, and increased during the late phase of subexcitability.8,96 The refractory period results from inactivation of transient Na channels.31 The refractory period is prolonged by cooling, as a result of slowed channel kinetics. It increases with membrane depolarization as a result of inactivation of Na channels and decreases with membrane hyperpolarization because approximately 30% of Na channels are inactivated at resting membrane potential. Consequently, the refractory period can be used as an indicator of membrane potential. The superexcitable period is due to a depolarizing afterpotential that results from the capacitative charging of the internode by the action potential, with subsequent discharge occurring through or under the myelin sheath.7 The superexcitable period changes reciprocally with membrane potential through effects on paranodal K channels, which open with membrane depolarization and close with membrane hyperpolarization. As a result, the superexcitable period reflects the status of the internodal membrane and can also be used as an indicator of membrane potential.

Recovery Cycles: Refractoriness, Superexcitability, and Late Subexcitability Axonal excitability oscillates in a stereotypical fashion following conduction of a single impulse, a process referred to as the recovery cycle8,49 (Fig. 5–4). Following conduction of a single action potential, the axon is initially absolutely refractory, so that it cannot produce further action potentials regardless of stimulus strength. Following this phase of

A

B 100

100

Threshold increase (%)

Threshold increase (%)

29 C

Refractoriness

0

32 C

Superexcitability

0 Late subexcitability 35 C

0

100 Inter-stimulus interval (ms)

10 100 Inter-stimulus interval (ms)

FIGURE 5–4 A, Recovery cycles for motor axons in the median nerve, averaged over 10 subjects at three different skin temperatures. B, Same data replotted with logarithmic time axis to show early part of recovery more clearly, and approximately uniform horizontal displacement of curves with temperature. (Data from Kiernan, M. C., Cikurel, K. and Bostock, H.: Effects of temperature on the excitability properties of human motor axons. Brain 124:816, 2001.)

118

Function of the Peripheral Nervous System

The final late subexcitable period reflects a corresponding membrane hyperpolarization and results from activation of a nodal slow K conductance.5,8 The driving force for K currents is the difference between the potassium equilibrium potential and the resting potential (ER  EK), so that subexcitability is sensitive to changes in the extracellular potassium concentration as well as the membrane potential.

Threshold Electrotonus and Current-Threshold Relationship Changes in potential of the nodal membrane spread to the internode, but slowly because of the resistance of the myelin sheath and consequently the slow charging of the internodal capacitance. This results in slow activation/deactivation of voltage-dependent channels on the internodal membrane. Although Na channel density is insufficient for the internodal membrane to generate an action potential, the changes in resistance of the internodal membrane and in the current stored on it will affect the behavior of the node.6 Approximately 99.9% of the axonal membrane is internodal so that, despite the lower channels densities, there are many more channels on the internode than on the node. The only technique that provides some insight into internodal conductances in human subjects in

A

vivo is the technique of threshold electrotonus. This term describes the changes in threshold produced by longlasting direct current (DC) pulses, which under most circumstances parallel the underlying electrotonic change in membrane potential. In response to depolarizing current pulses, there is an initial fast phase that is almost instantaneous and is proportional to the applied current (the “F” phase) (Fig. 5–5). This is followed by further depolarization that develops slowly over some tens of milliseconds as the current spreads to and depolarizes the internodal membrane (the “S1” phase). The threshold decrease (i.e., the extent of depolarization) reaches a peak approximately 20 ms after the onset of the current pulse, depending on its strength, and then threshold starts to return slowly toward the control level. This lessening of the degree of depolarization is due to activation of a hyperpolarizing conductance with slow kinetics. The use of K channel blocking agents indicates that the slow accommodative process is due to activation of slow K channels, which are located on both the node and the internode.5,13,18 When the DC pulse is terminated, threshold increases rapidly and there is then a slow overshoot before it gradually recovers to control level. The slow overshoot is due to the slow deactivation of a hyperpolarizing conductance, and is further evidence that the current pulse activated a slow K conductance.

B 100

100

F

0

F S1 –100

0 –100 Sensory

–200

S3 Motor

–300 Current (% threshold)

Threshold reduction (%) Current (% threshold)

Threshold reduction (%)

S2

S1

50 0 –50 0

100 Delay (ms)

200

0 –100 0

100 200 300 400 500 Delay (msec)

FIGURE 5–5 Normal threshold electrotonus waveforms and components. A, Averaged motor nerve responses to 100-ms polarizing currents ( 40% of threshold) from 38 subjects, showing fast (F) and slow (S1, S2) components (see text) and small intersubject variability (thick lines, mean; thin lines, mean  SD). Stimuli applied to ulnar nerve at wrist, and compound muscle action potential recorded from hypothenar muscle. (From Bostock, H., Cikurel, K., and Burke, D.: Threshold tracking techniques in the study of human peripheral nerve. Muscle Nerve 21:137, 1998, with permission.) B, Comparison between mean motor and sensory responses to 300-ms current pulses ( 40%, 80% of threshold) from eight subjects, showing additional component of threshold electrotonus (S3) resulting from inward rectification activated by hyperpolarization. (From Bostock, H., Burke, D., and Hales, J. P.: Differences in behaviour of sensory and motor axons following release of ischaemia. Brain 117:225, 1994, with permission of Oxford University Press.)

119

Nerve Excitability Measures

With long-lasting hyperpolarizing DC pulses, there is again a fast threshold change that increases threshold proportionally to the applied current, analogous to the comparable phase with depolarizing currents (the “F” phase). Then threshold continues to increase as the hyperpolarization spreads to the internode. This “S1” phase starts as a mirror image of the S1 phase with depolarizing current but soon diverges, because hyperpolarization closes K channels (slow nodal and fast and slow internodal K channels), and this increases the amplitude and time constant of S1. At approximately 150 ms after the start of DC hyperpolarization, S1 reaches a maximum and threshold begins to decrease toward the control level. This accommodative phase (“S3”) (see Fig. 5–5B) produces “inward rectification” and is due to activation of the hyperpolarization-activated current (IH).79 Although IH has slow activation and deactivation kinetics, it is activated and affects threshold earlier than 100 ms (where S1 peaks): Without IH, the hyperpolarizing threshold increase would be even greater. On termination of the hyperpolarizing DC pulse, threshold rapidly decreases and then undergoes a slow, depolarizing undershoot as IH is slowly deactivated and the slow K conductance is reactivated. Just as the activation of different voltage-dependent channels in a cell can be revealed by plotting a currentvoltage (I/V) relationship, it can be convenient to plot the threshold changes at the ends of a series of long current pulses of different amplitude as a current-threshold relationship (e.g., see Fig. 5–6C below). This is conventionally done with threshold increase (hyperpolarization) to the left, threshold decrease (depolarization) to the right, depolarizing current to the top, and hyperpolarizing current to the bottom, to correspond to a conventional I/V plot. Outward rectification resulting from K channel activation causes a steepening of the curve in the top right quadrant, and the steepening of the relationship in the bottom left quadrant indicates inward rectification resulting from activation of IH.

AUTOMATED RECORDING OF MULTIPLE NERVE EXCITABILITY PROPERTIES A considerable spur to the reintroduction of nerve excitability testing into clinical neurophysiology has been the development of a computer program for automatic tracking of threshold changes, and a protocol that sequentially measures a number of different excitability parameters in a motor nerve within a recording period of about 10 minutes44 (Fig. 5–6). The program first measures stimulus-response curves with two different stimulus widths (Fig. 5–6A and B), from which the strength-duration time constant can be estimated for fibers recruited at different fractions of the compound muscle action potential (CMAP) (Fig. 5–6D). A target response of 40% maximal CMAP amplitude is then set, and threshold tracking is implemented, using the slope of the stimulus-response curve to anticipate changes in threshold efficiently. Threshold electrotonus is recorded for

100-ms polarizing currents  40% of the control threshold (Fig. 5–6E), the current-threshold relationship for 200-ms pulses of currents from 50% to 100% of threshold (Fig. 5–6C), and the recovery cycle for interstimulus intervals from 2 to 200 ms (Fig. 5–6F). Many of the excitability recordings referenced in this chapter have been obtained with this convenient protocol. A similar program has been described for measuring excitability properties of sensory axons.48 Unfortunately there is currently no stimulator available commercially for making these automatic measurements, and this has limited the number of centers performing nerve excitability studies.

NERVE EXCITABILITY MEASURES AND MEMBRANE POTENTIAL The extent of the dependence of each of the excitability parameters on membrane potential was recently explored in studies that measured the effect of DC polarizing currents and ischemia on nerve excitability.43 Depolarizing the axonal membrane by 1 mA resulted in significant changes in the relatively refractory period (increased by 37%), supernormality (decreased by 82%), and threshold electrotonus (“fanning in” by 18% in the depolarizing phase and 71% in the hyperpolarizing phase). A smaller change was documented in strength-duration time constant (increased by 5%). These findings illustrate the sensitivity of axonal excitability measures to changes in membrane potential and establish quantitative relationships that have proved useful in the interpretation of the axonal abnormalities recorded in diseases affecting peripheral nerves, such as uremia (see below).

Effects of Ischemia on Axonal Excitability Different disease processes produce focal compression or ischemia and, from a clinical perspective, the changes in excitability associated with ischemia and its release, which indirectly alter resting membrane potential, are more relevant than the “pure” changes in potential induced by polarizing currents. Ischemia is a common cause of peripheral nerve dysfunction, and the excitability changes demonstrated in such studies may well be similar to those responsible for the generation of ectopic activity leading to the symptoms of paresthesias, fasciculation, and myokymia in patients suffering from peripheral nerve disease.15,17,70,72 Experimental models of ischemia have evolved from initial studies on the effects of inflation of a blood pressure cuff around the arm, with the subsequent development of postischemic paresthesias in the digits and palm.61 Ischemia paralyzes energy-dependent processes, particularly the Na,K pump (Na,K-ATPase), and this results in accumulation of Na ions inside the axon and K ions outside the axon and thereby depolarization of sensory and

120

Function of the Peripheral Nervous System

CMAP (% max)

B CMAP (mV)

A 10 1

.1 1

100

0

10 Stimulus (mA)

0

C

100 200 Stimulus (% mean threshold)

D CMAP (% max)

Current (% threshold)

100 0

–100

0 –500 0 Threshold reduction (%)

0

1 Strength-duration time constant (ms)

E

F 100 Threshold change (%)

Threshold reduction (%)

100

0

–100

0

–200 0

100 Conditioning-test delay (ms)

10

200

100

Inter-stimulus interval (ms)

motor axons. There is first an increase and then a reduction in nerve excitability.14,17,43,62,66,72 With release of ischemia, axons rapidly hyperpolarize as a result of a rebound increase in function of the pump. A series of recordings made before (control), during (at 5 and 15 minutes), and after ischemia are illustrated for a single subject in Figure 5–7, using threshold tracking to record multiple measures of axonal excitability. It can be seen that ischemia produces marked and distinctive changes in all measures of axonal excitability—changes consistent with axonal depolarization. The stimulusresponse curve, a nonspecific measure of excitability, shifts to the left as the stimulus current required to evoke the target response becomes less (Fig. 5–7A). The slope of the current-threshold relationship reflects the rectifying properties of the axon, both nodal and internodal, and it becomes steeper as potassium conductances are activated (except with strong hyperpolarizing currents, when the slope is reduced because of a reduction in inward rectification) (Fig. 5–7B). There is a fanning in of threshold

FIGURE 5–6 Multiple nerve excitability measures, recorded with a standard 10-minute protocol. Solid lines and filled circles, recording from a single subject; thin broken lines, means and 95% confidence limits for 29 normal subjects. A, Stimulus-response curves for stimulus durations of 1 ms (thick line) and 0.2 ms (thin line). Filled circle indicates 50% maximum response to 1-ms stimulus, and broken ellipse indicates 95% confidence limits for this point. B, Normalized stimulus-response curves. C, Current-threshold relationship. D, Strengthduration time constants, estimated from data in A for axons recruited at different fractions of the maximum CMAP. E, Threshold electrotonus. F, Recovery cycle. (From Kiernan, M. C., Burke, D., Andersen, K., and Bostock, H.: Multiple measures of axonal excitability: a new approach in clinical testing. Muscle Nerve 23:399, 2000, with permission.)

electrotonus curves as a result of increased resting activation of K channels and short-circuiting of the internodal axon (Fig. 5–7C). The short-circuiting of the internodal axon also reduces superexcitability (Fig. 5–7D), and this, together with increased sodium channel inactivation, prolongs the relative refractory period. Taken together, these changes are those expected from axonal depolarization.43 In some sensory axons, the increased activation of persistent sodium conductances temporarily outweighs the effects of transient sodium channel inactivation and increased potassium conductance, resulting in spontaneous activity and mild, short-lasting paresthesias. In contrast, release of ischemia reverses the changes in excitability described above, such that the axon becomes hyperpolarized14,15,17,43,63,72: the stimulus-response curve shifts to the right (Fig. 5–7A), the current-threshold relationship becomes less steep (Fig. 5–7B), threshold electrotonus curves “fan out” (Fig. 5–7C), and the relative refractory period shortens and gives way to enormous superexcitability (Fig. 5–7D). Postischemic axonal

121

Nerve Excitability Measures

A

B I1

C PI

Current (% threshold)

CMAP (mV)

10

1

.1

100

0

I2 I1 C PI

–100

–200 0

100 200 Conditioning-test delay (ms)

I1 I2

–500 0 Threshold reduction (%)

D Threshold change (%)

C Threshold reduction (%)

C

10 Stimulus (mA)

hyperpolarization is induced by rebound overactivity of the Na,K pump, attempting to normalize the ionic gradients that have developed during the period of ischemia when the pump was paralyzed. The pump is electrogenic and generates a net outward, hyperpolarizing current, because it exports three Na ions for each two K ions imported. While paresthesias experienced during the period of ischemia are typically mild, it is with the release of ischemia, when axons become hyperpolarized, that paresthesias paradoxically become most intense.72 In motor axons, the postischemic hyperexcitability has been attributed to the combination of increased Na,K pump activity and the high level of extracellular K ions accumulated during the paralysis of the pump. The pump is able to make the axonal membrane hyperpolarized relative to the equilibrium potential for K. This reversal of the normal electrochemical gradient for K means that K currents become inward and depolarizing, and, like normal Na currents, they therefore become regenerative because they cause more K channels to open. This results in spontaneous depolarizations toward EK, which produce bursts of action potentials. Following release of ischemia of sufficiently long duration to produce fasciculation, a bimodal distribution of thresholds has been demonstrated for motor axons, with a small population of axons existing in a depolarized state while the majority are hyperpolarized as a result of overactivity of the Na,K pump.15 Small changes in pump current then lead to rapid transition of axons from a depolarized to a hyperpolarized condition, generating spontaneous discharges.

PI –100

1

FIGURE 5–7 The effects of ischemia on nerve excitability properties in a single subject (C  control; I1  5 minutes of ischemia; I2  15 minutes of ischemia; PI  5 minutes postischemia). A, Absolute stimulus-response relationship. B, Current-threshold relationship. C, Threshold electrotonus. D, Recovery cycle. (From Kiernan, M. C., and Bostock, H.: Effects of membrane polarization and ischaemia on the excitability properties of human motor axons. Brain 123:2542, 2000, with permission from Oxford University Press.)

0

100

I2 0

I1 C PI 10 100 Inter-stimulus interval (ms)

Similar studies in sensory axons have failed to document such clear evidence for a bimodal distribution of thresholds in the postischemic state, as was demonstrated for motor axons.17,63 It appears that the high-threshold state is too short lived in sensory axons, because of the greater inward rectification17 (see below).

MODALITY AND REGIONAL DIFFERENCES BETWEEN DIFFERENT AXONS Axons function to transmit an impulse train securely from one end to the other and, presumably, this is done with minimal expenditure of energy. However, the pattern of activity varies widely for axons of different modality. They are subjected to impulse patterns that have very different peak and sustained discharge rates, regularity of discharge, and train length. It might be expected that different axons require slightly different strategies to achieve secure, energy-efficient conduction. In human subjects, there are a number of differences between sensory and motor axons at the same site in the same nerve and between sensory axons innervating the upper and lower limbs. For example, sensory axons in the ulnar nerve17 and in the median nerve63 undergo greater accommodation to hyperpolarizing stimuli than motor axons in the same nerves, presumably reflecting a greater expression of the hyperpolarization-activated cation conductance, IH. Sensory axons in the median nerve have a longer strength-duration time constant71,78 and shorter rheobase than motor axons in the same nerve,71 differences

122

Function of the Peripheral Nervous System

that probably relate to a greater expression of persistent Na conductance (INaP) on sensory axons.20 Sensory axons undergo less refractoriness, supernormality, and late subnormality as they recover excitability following a single discharge.49 Finally, the available evidence suggests that resting membrane potential is similar for median sensory and motor axons,24 but it is possible that median sensory axons are more dependent on the electrogenic Na,K pump to maintain resting membrane potential,63 something that might be necessary to offset their greater persistent Na channel activity.20 Either way, median sensory axons appear to undergo greater ischemic depolarization and a more complex but probably greater postischemic hyperpolarization than median motor axons.63 There are also regional differences between cutaneous afferents in the median and sural nerves that cannot be explained by the slightly smaller size of the sural afferents. For example, median afferents accommodate better to both depolarizing and hyperpolarizing stimuli than do sural afferents, differences that probably reflect greater expression of nodal and internodal slow K conductances and of the internodally located IH on median afferents.65,66 In addition, sural afferents undergo less ischemic depolarization and postischemic hyperpolarization than median afferents, but the precise mechanisms underlying these differences remain uncertain.66 In addition to these differences, there is considerable evidence that the excitability of an axon population changes along its length. For example, the changes in excitability are greater in more proximal segments, and fasciculation occurs more readily following release of ischemia when the site of the ischemia is proximal.14,53 The biophysical basis for this length-dependent difference in axonal excitability has not been the subject of exhaustive study, and the few preliminary studies have not clarified the issue.56,70 The modality-dependent and regional differences in biophysical properties (but not any length-dependent differences) could well represent appropriate differences for the different physiologic roles played by those axons (i.e., an adaptation designed for optimal function). However, this may not be the case when excitability is disturbed and the ability to conduct impulses is jeopardized. The greater expression of two essentially depolarizing conductances (INaP and IH) on sensory than on motor axons would render them more excitable when the nerve is stressed (e.g., during and after ischemia), and it is entirely logical that sensory axons would become ectopically active more readily under such conditions. Conversely, when impulse conduction is jeopardized, hyperpolarizing stresses would be more likely to precipitate conduction failure in motor axons,24 and it is perhaps not surprising that inflammatory demyelinating neuropathies commonly present with predominantly motor deficits even when sensory axons are demonstrably involved. The regional differences in those conductances responsible for accommodation to depolarizing and

hyperpolarizing stimuli imply that median sensory axons are better placed to react to depolarizing and hyperpolarizing stresses than are sural sensory axons. It is conceivable that this could be a factor in the explanation for why the clinical manifestations of many acquired polyneuropathies are more prominent in the distal lower limb.

CHANGES IN AXONAL MEMBRANE POTENTIAL IN PERIPHERAL NEUROPATHIES To date a number of clinical studies have been completed using the new protocol developed to allow rapid measurement of multiple parameters of nerve excitability.44 Studies in patients with uremic neuropathy and with multifocal motor neuropathy (MMN) are described to illustrate the effects that changes in membrane potential can have on axonal function and how these changes can be documented by studies of axonal excitability. This section concludes with a consideration of how physiologic changes in membrane potential can produce conduction block in chronically demyelinated axons.

Multifocal Motor Neuropathy and Axonal Hyperpolarization MMN is a disease process characterized by motor nerve involvement with sparing of sensory function (other than tendon jerks), producing a syndrome in affected patients of slowly progressive muscle atrophy, weakness, and fasciculation.80,82,89,99 A critical diagnostic feature in MMN is the demonstration of conduction block in multiple peripheral nerves on electrophysiologic investigation.98 Conduction block involves only motor axons, with sensory conduction spared across the lesion.42,81 The presence of “positive” symptoms and signs such as cramp, myokymia, and fasciculation in the context of predominantly negative features such as depressed tendon jerks, muscle atrophy, and weakness remains unexplained in these patients. Neither has the mechanism of the conduction block itself been elucidated. Limited information is available from pathologic studies, some of which provide evidence in favor of demyelination.42,81 Axonal properties remote from foci of conduction block are normal,25 findings that support the contention that MMN is not a generalized disease with a focal presentation. However, excitability studies have also been undertaken in MMN patients just distal to the site of conduction block in affected nerves, with strikingly abnormal findings.45 The most significant alterations in excitability properties in the MMN patients were (1) a reduction in minimum slope of the current-threshold (I/V) relationship (indicating a reduction in input conductance); (2) a “fanning out” of threshold

123

Nerve Excitability Measures

despite the changes consistent with hyperpolarization in the other excitability parameters. This discrepancy between ischemia and depolarization in Figure 5–9D presumably reflects the dependence of late subexcitability on (ER  EK), rather than just on ER (see above), and that endoneurial K increases during ischemia. Similarly, the small change in late excitability on release of ischemia probably indicates that the endoneurial fluid becomes hypokalemic as a result of the pump activity. These comparisons suggest that the nerve distal to the site of conduction block in these patients with MMN was hyperpolarized, in a way similar to what occurs in postischemic nerve. For a critical test of whether hyperpolarization was responsible for the abnormal membrane properties, a depolarizing current of 0.5 mA was applied to the nerves of the MMN patients, to test whether depolarization could reverse the abnormalities. This was indeed the case (see Fig. 5–9): All the excitability abnormalities were reduced by depolarization. It is notable that, in these MMN patients, the hyperpolarization seems to represent a stable steady state. Excitability recordings taken 10 weeks apart showed almost identical excitability abnormalities. Persistent overactivity of the Na,K pump would require a persistent intra-axonal source of Na ions to drive the pump. These Na ions could not be continuously entering the axon at the recording site, or there would be a net membrane depolarization.

electrotonus; and (3) an increase in superexcitability recorded during the recovery cycle. The changes in threshold electrotonus and recovery cycle are illustrated for six patients in Figure 5–8A and B. All these excitability parameters, highly abnormal in the MMN patients, depend on the resting conductance of the paranodal and internodal axon membrane. One way that the resting conductance of axonal membrane can be reduced is by membrane hyperpolarization, and the excitability properties in Figure 5–8A and B are very similar to the changes previously seen during hyperpolarization after release from ischemia (see Fig. 5–7C and D). To explore this possibility in more detail, more excitability parameters of the patient nerves were compared to those from polarized and ischemic nerves in healthy subjects43 (Fig. 5–9). For simplicity, the data points plotted are restricted to the mean values for both control groups, and for the effects of 1-mA depolarizing and hyperpolarizing currents, and for recordings started 5 minutes after applying a pressure cuff and 5 minutes after its release. The patient axons behaved like postischemic axons.43 In Figure 5–9D, superexcitability is plotted against late subexcitability. Whereas both hyperpolarization and recovery from ischemia increase superexcitability, late subexcitability is decreased by hyperpolarizing currents but remains unchanged when the nerve is postischemic, and unchanged or even increased in MMN patients,

A

B Threshold change (%)

100

0

–100

0

–200 0

100 Conditioning-test delay (ms)

200

C

10 100 Inter-stimulus interval (ms)

D 100 Threshold change (%)

100 Threshold reduction (%)

FIGURE 5–8 Pathologic changes in nerve excitability measures related to membrane potential. A and B, Threshold electrotonus and recovery cycle recorded distal to the site of conduction block in 6 patients with multifocal motor neuropathy (circles and error bars, means  SEM) compared with 29 normal controls (thin solid and broken lines, means  SEM). (Data from Kiernan, M. C., Guglielmi, J.-M., Kaji, R., et al.: Evidence for axonal membrane hyperpolarization in multifocal motor neuropathy with conduction block. Brain 125:664, 2002.) C and D, Threshold electrotonus and recovery cycle recorded prior to dialysis in 9 patients with chronic renal failure (circles and error bars, means  SEM) compared with 29 normal controls (thin solid and broken lines, means  SEM). (Data from Kiernan, M. C., Walters, J. L., Andersen, K. V., et al.: Nerve excitability changes in chronic renal failure indicate membrane depolarization due to hyperkalaemia. Brain 125:1366, 2002.)

Threshold reduction (%)

10

0

–100

0

–200 0

100 Conditioning-test delay (ms)

200

10 Inter-stimulus interval (ms)

100

124

Function of the Peripheral Nervous System

A

B .5

10

D

RD R

H

I

MD

M

R D

RRP (ms)

Minimum I/V slope

I

RD

PI H

0 0

1 Resting I/V slope

PI

MD M

30

35 Temperature ( C)

C

D 30 H M PI MD D

RD 50

R

Subexcitability (%)

TEd (10–20ms)

100

D 20 M PI

I RD

10

MD

H

R

I 0

–200

–100 TEh (90–100ms)

0

–50 Superexcitability (%)

Longitudinal diffusion or transport along the axon could provide an intra-axonal source of Na ions. Given that the lesion in these patients is focal, Na,K pump activity could be blocked by, for example, edema, or by antibodies directed against the protein components of the pump. In either case, interference with Na,K pump function could produce a depolarizing block with intracellular accumulation of Na at the site of the lesion. Disruption of the blood-nerve barrier42 might increase the potassium concentration in the endoneurial fluid, further aggravating any depolarization block. If Na influx continued, a steady state would be achieved only by Na ions moving intracellularly along the axon to a site where the pump was still working, and this would result in overactivity and membrane hyperpolarization. Therefore, depolarization at the site of the lesion would coexist with chronic hyperpolarization on one or both sides of this site. Such a lesion, with adjacent ischemic and postischemic lengths of nerve, would be likely to generate ectopic activity, a hallmark of MMN, with patients experiencing fasciculation or myokymia in the presence of conduction block. Other clinical symptoms and studies provide further support for this hypothesis. Cold paralysis, described in monomelic amyotrophy,34,47 also occurs in patients with MMN.41 The electrogenic Na,K pump is temperature sensitive, with slower kinetics at lower temperatures. Cooling would exacerbate the effect of the already compromised Na,K pump function, increasing the depolarizing conduc-

FIGURE 5–9 Changes in multiple excitability parameters related to changes in membrane potential. Changes with polarization (H  1-mA hyperpolarizing current; D  1-mA depolarizing current) and ischemia (I  5 minutes of ischemia; PI  5 minutes postischemia) compared with mean data from three MMN patients, including the normalizing effect of depolarization (M  MMN control; MD  MMN with 0.5-mA depolarizing current) and with data from six patients with chronic renal failure (R  before dialysis; RD  after dialysis). A, Minimum versus resting current-threshold slope. B, Relative refractory period versus skin temperature. C, Early depolarizing versus late hyperpolarizing threshold electrotonus. D, Superexcitability versus late subexcitability. Dotted lines and ellipses indicate 95% confidence limits for normal control values.

tion block. This contrasts with the expected effect of cooling in alleviating conduction block caused by demyelination.86,94 Similarly digitalis, a known blocker of Na,K pump activity, can paradoxically exacerbate the fanning out changes seen in threshold electrotonus in MMN.41 Unfortunately, attempts to track excitability changes more proximally toward the site of the focal lesion in patients with MMN have failed because the nerve becomes too inexcitable.105

Uremic Neuropathy and Axonal Depolarization Uremic neuropathy is a distal, symmetrical, mixed sensorimotor, predominantly axonal polyneuropathy affecting legs more than arms.85 The pathologic findings are similar to those in other toxic neuropathies, but the mechanism of nerve damage is unknown. Uremic neuropathy can be minimized by early dialysis68 or reversed by renal transplantation.9,75,77 Because uremic neuropathy improves with dialysis, it is attributed to the accumulation of dialyzable metabolites, but the nature of these uremic toxins remains obscure.10 Substances in the 300 to 5000 range of molecular weights have been implicated on the basis of the supposed superiority of peritoneal dialysis in preventing neuropathy,91 and this “middle molecule” theory3 is still the most popular, although all attempts to identify the putative middle-molecular-weight neurotoxins have been

Nerve Excitability Measures

inconclusive.10,69,85 It has been proposed that uremic neurotoxins cause nerve damage by inhibiting Na,K-ATPase, leading to membrane depolarization,74 because maintenance of a normal membrane potential and ionic gradients is essential for axonal survival.97 In a group of uremic patients treated by hemodialysis, the most striking abnormalities in excitability parameters were revealed by the recordings of threshold electrotonus and recovery cycle (see Fig. 5–8C and D), both properties that are strongly sensitive to membrane potential.51 Axonal excitability was highly abnormal prior to dialysis, with increased refractory period, increased accommodation in threshold electrotonus, and reduced superexcitability. These changes indicate axonal depolarization, and are the opposite of those described above in MMN.45 Hemodialysis produced rapid and significant normalization of these excitability parameters (see Fig. 5–9). Compared with nerves depolarized by applied currents or by ischemia, the excitability changes in chronic renal failure (CRF) were more like those during ischemia, but differences in subexcitability suggested that the increase in endoneurial potassium was more important in CRF than in ischemia (see Fig. 5–9D). The excitability indices of membrane potential in the predialysis recordings were strongly correlated with serum potassium, but not with other electrolytes or serum markers of renal dysfunction. Moreover, CRF patients with highly abnormal serum levels of urea and creatinine, but with normal potassium, had normal nerve excitability properties. These data all pointed to hyperkalemia as the primary cause of membrane depolarization in these subjects. In summary, axons in many patients with CRF appear to be chronically depolarized, and it is likely that this membrane depolarization is due to hyperkalemia.51 It is a reasonable conjecture that this hyperkalemic membrane depolarization contributes to the etiology of uremic neuropathy, but this has not yet been investigated.

Impulse Conduction in Demyelinating Neuropathies Acute demyelination lowers the safety margin for impulse conduction, such that axons can become sensitive to shifts in membrane potential, even when those shifts occur through normal physiologic mechanisms.19 In critically conducting axons, impulse conduction can be impaired by the effects of heating and activity and probably by any mechanism that produces a significant shift in membrane potential, whether depolarizing or hyperpolarizing. Raising temperature by 0.5 °C can be sufficient to precipitate conduction failure in a critically conducting axon,86 and warming commonly accentuates the deficit in multiple sclerosis, even to the point that warming is sometimes used as a provocative test. Conduction failure occurs because warming speeds up channel gating, affecting both activation

125

and inactivation, and this decreases the time integral of the Na current at the node of Ranvier. In a critically conducting axon, the duration of the driving current at the blocking node can reach 1.0 ms,38 and conduction block can be precipitated or relieved by maneuvers that manipulate the time course of the driving current, such as changing temperature or administration of agents that interfere with Na channel inactivation. Activity hyperpolarizes axons. Different mechanisms are responsible for hyperpolarization by short-and long-lasting trains of impulses. With brief, high-frequency trains ( 10 to 20 impulses at 100 to 200 Hz), there is cumulative activation of the nodal slow K conductance.5,8,100 The resulting hyperpolarization can increase threshold by approximately 40%, more than sufficient to jeopardize conduction in impaired axons, but the return to control excitability occurs over 100 to 150 ms,64 such that this mechanism might disrupt the discharge pattern and limit the discharge rate, but it could not produce conduction failure by itself. When an axon conducts long impulse trains, particularly at high frequency, there is an accumulation of Na ions within the axon, and this activates the electrogenic Na,K pump, as occurs on release of ischemia, to restore ionic balance. This activity-dependent hyperpolarization has been demonstrated in human sensory2,50 and motor8,16 axons and, importantly, it can be produced by natural activity.57,101 The extent and duration of the hyperpolarization depend on discharge rate and train length, and can result in an increase in threshold of approximately 40% that takes many tens of minutes to decay to control excitability. Voluntary contractions lasting as little as 15 seconds can increase the threshold of motor axons by 10% to 15%, and this increase decays over some 5 to 10 minutes.101 Such changes are likely to be clinically relevant: A conservative estimate of the safety margin for impulse conduction in a series of patients with chronic inflammatory demyelinating polyneuropathy (CIDP) suggested that significant conduction failure would occur if the axons hyperpolarized by approximately 14%.26 For the same impulse load, the extent of the activitydependent hyperpolarization seems to be greater for motor than for sensory axons101 (M. C. Kiernan, C. S.-Y. Lin, and D. Burke, unpublished findings). An important factor in this difference is probably the difference in IH (see above). In patients with inflammatory demyelinating polyneuropathies, the normal activity-dependent hyperpolarization can precipitate conduction failure at sites of impaired function. This was demonstrated first in MMN40 and subsequently in CIDP.26 There is nothing special about activity: Any process that produces sufficient hyperpolarization will produce clinically significant conduction block at pathologic sites in these disorders, provided that a sufficient number of axons are critically conducting. The release of ischemia results in a postischemic hyperpolarization, and this too can precipitate conduction failure.28 Paradoxically, in CIDP conduction failure may also occur during ischemia, during

126

Function of the Peripheral Nervous System

the depolarizing shift in membrane potential.28 This probably occurs because depolarization inactivates transient Na channels, thus decreasing the availability of functioning channels in axons that are critically dependent on the size of the Na current. It is also possible that ischemia produces an ischemic metabolite that blocks Na channels,62 a mechanism that would further limit the number of Na channels available for the action current. The important message is that critically conducting axons are delicately poised. Conduction may block if membrane potential is too far from threshold (i.e., the axon is hyperpolarized) or if the Na current becomes inadequate (because of heating or because of a limitation on the number of functioning Na channels). Significant changes in activity or significant shifts in membrane potential, whether depolarizing or hyperpolarizing, may be sufficient to produce a transient worsening of symptoms. Clinical fluctuations are well documented in multiple sclerosis; they are as likely in demyelinating polyneuropathies if a sufficient number of axons can only just maintain conduction.

POTENTIAL ROLE FOR EXCITABILITY STUDIES IN THE NEUROPATHY CLINIC Automated nerve excitability testing is still in its infancy, and it is not yet clear whether it will earn a place in the clinic, back alongside nerve conduction studies, which displaced the earlier excitability tests in the 1960s. So far, although many different abnormalities in nerve excitability measures have been found in a range of diseases affecting peripheral nerve, none has been sufficiently sensitive or specific to warrant use as a routine diagnostic tool. Excitability testing is, however, undoubtedly capable of yielding unique information about the biophysical state of nerve fibers, and this information can reasonably be expected to lead to improvements in patient treatment. Although we have concentrated above on the new information it provides about resting membrane potentials, excitability testing can also give warning of a wide variety of other changes in axonal membrane properties. Indeed, measurements of an excitability parameter that is only weakly potential dependent, the strength-duration time constant, have led to the first example of a new treatment initiated as a result of excitability testing. Kanai et al.39 recorded increased strength-duration time constants in Machado-Joseph disease, and this led the authors to introduce treatment with mexiletine to reduce persistent sodium currents. This treatment only partially normalized the strength-duration time constants, but dramatically relieved disabling muscle cramps in eight patients. It is this type of use, as a research tool for comparing groups of patients with defined pathology, in whom excitability testing can provide interesting new insights into pathophysiology, that it has proved most useful so far.

In addition to the changes in membrane potential described above, excitability changes documented in disease include resistance to ischemic depolarization in diabetics102 and a reduction in hyperpolarization-activated current (IH) associated with diabetic neuropathy,37 which in rats is prevented by blocking the polyol pathway with aldose reductase inhibitor.104 Abnormal nerve excitability in amyotrophic lateral sclerosis may involve both reduced K channel and increased persistent Na currents, both leading to membrane instability, but there is a high variability among patients.18,22,37,73 In contrast to nerve conduction studies, demyelination often fails to result in detectable excitability abnormalities,18 either because the demyelination is too distal (see below) or because the density of affected nodes is too low, and affected nodes will have high thresholds and not contribute to the 40% maximal responses that are usually tracked. Thus a study of CIDP patients found high thresholds and prominent changes in stimulus-response curves, but no consistent changes in threshold electrotonus.27 However, it has recently been found that the diffuse demyelination in Charcot-Marie-Tooth disease type 1A does produce characteristic and consistent abnormalities in threshold electrotonus (H. Nodera, S. Kuwabara, and R. Kaji, personal communication). A major limitation of current nerve excitability testing methods is that they are restricted to sites where the peripheral nerve passes close to the skin, such as the ulnar and median nerves at the elbow and wrist. However, in many diseases affecting nerve excitability, it is the most distal portions of the nerve where membrane abnormalities are most pronounced, as inferred from the predominantly distal origin of most ectopic discharges.60 Thus a study of motor axons in neuromyotonia46 failed to find evidence for a membrane abnormality responsible for the hyperexcitability, and a study of Guillain-Barré patients58 found normal membrane properties at the wrist, despite prolonged distal latencies and an increased refractory period of transmission in patients with the axonal form of the disease. Excitability testing is therefore most likely to find clinical application in conditions in which axons are affected rather uniformly along their length, such as in giving early warning of neuropathy in patients being treated with neurotoxic drugs.92

REFERENCES 1. Adrian, E. D., and Lucas, K.: On the summation of propagated disturbances in nerve and muscle. J. Physiol. (Lond.) 44:68, 1912. 2. Applegate, C., and Burke, D.: Changes in excitability of human cutaneous afferents following prolonged high frequency stimulation. Brain 112:147, 1989. 3. Babb, A. L., Ahmad, S., Bergström, J., and Scribner, B. H.: The middle molecule hypothesis in perspective. Am. J. Kidney Dis. 1:46, 1981.

Nerve Excitability Measures 4. Baker, M., and Bostock, H.: Depolarization changes the mechanism of accommodation in rat and human motor axons. J. Physiol. (Lond.) 411:545, 1989. 5. Baker, M., Bostock, H., Grafe, P., and Martius, P.: Function and distribution of 3 types of rectifying channel in rat spinal root myelinated axons. J. Physiol. (Lond.) 383:45, 1987. 6. Baker, M. D.: Axonal flip-flops and oscillators. Trends Neurosci. 25:514, 2000. 7. Barrett, E. F., and Barrett, J. N.: Intracellular recording from vertebrate myelinated axons: mechanism of the depolarizing afterpotential. J. Physiol. (Lond.) 323:117, 1982. 8. Bergmans, J.: The physiology of single human nerve fibres. Louvain, Vander, 1970. 9. Bolton, C. F., Baltzan, M. A., and Baltzan, R. G.: Effects of renal transplantation on uremic neuropathy. N. Engl. J. Med. 284:1170, 1971. 10. Bolton, C. F., and Young, G. B.: Neurological Complications of Renal Disease. Boston, Butterworths, 1990. 11. Bostock, H.: The strength-duration relationship for excitation of myelinated nerve: computed dependence on membrane parameters. J. Physiol. (Lond.) 341:59, 1983. 12. Bostock, H.: Mechanisms of accommodation and adaptation in axons. In Waxman, S. G., Stys, P. K., and Kocsis, J. D. (eds.): The Axon. Oxford, UK, Oxford University Press, p. 311, 1995. 13. Bostock, H., and Baker, M.: Evidence for two types of potassium channel in human motor axons in vivo. Brain Res. 462:354, 1988. 14. Bostock, H., Baker, M., Grafe, P., and Reid, G.: Changes in excitability and accommodation of human motor axons following brief periods of ischaemia. J. Physiol. (Lond.) 441:513, 1991. 15. Bostock, H., Baker, M., and Reid, G.: Changes in excitability of human motor axons underlying post-ischaemic fasciculations: evidence for two stable states. J. Physiol. (Lond.) 441:537, 1991. 16. Bostock, H., and Bergmans, J.: Post-tetanic excitability changes and ectopic discharges in a human motor axon. Brain 117:913, 1994. 17. Bostock, H., Burke, D., and Hales, J. P.: Differences in behaviour of sensory and motor axons following release of ischaemia. Brain 117:225, 1994. 18. Bostock, H., Cikurel, K., and Burke, D.: Threshold tracking techniques in the study of human peripheral nerve. Muscle Nerve 21:137, 1998. 19. Bostock, H., and Grafe, P.: Activity-dependent excitability changes in normal and demyelinated rat spinal root axons. J. Physiol. (Lond.) 365:239, 1985. 20. Bostock, H., and Rothwell, J. C.: Latent addition in motor and sensory fibres of human peripheral nerve. J. Physiol. (Lond.) 498:227, 1997. 21. Bostock, H., Sears, T. A., and Sherratt, R. M.: The spatial distribution of excitability and membrane current in normal and demyelinated mammalian nerve fibres. J. Physiol. (Lond.) 341:41, 1983. 22. Bostock, H., Sharief, M. K., Reid, G., and Murray, N. M. F.: Axonal ion channel dysfunction in amyotrophic lateral sclerosis. Brain 118:217, 1995. 23. Burke, D., Kiernan, M. C., and Bostock, H.: Excitability of human axons. Clin. Neurophysiol. 112:1575, 2001.

127

24. Burke, D., Kiernan, M. C., Mogyoros, I., and Bostock, H.: Susceptibility to conduction block: differences in biophysical properties of cutaneous afferents and motor axons. In Kimura, J., and Kaji, R. (eds.): Physiology of ALS and Related Diseases. Amsterdam, Elsevier Science, p. 43, 1997. 25. Cappelen-Smith, C., Kuwabara, S., Lin, C. S.-Y., and Burke, D.: Abnormalities of axonal excitability are not generalized in early multifocal motor neuropathy. Muscle Nerve 26:769, 2002. 26. Cappelen-Smith, C., Kuwabara, S., Lin, C. S.-Y., et al.: Activity-dependent hyperpolarization and conduction block in chronic inflammatory demyelinating polyneuropathy. Ann. Neurol. 48:826, 2000. 27. Cappelen-Smith, C., Kuwabara, S., Lin, C. S.-Y., et al.: Membrane properties in chronic inflammatory demyelinating polyneuropathy. Brain 124:2439, 2001. 28. Cappelen-Smith, C., Lin, C. S., Kuwabara, S., and Burke, D.: Conduction block during and after ischaemia in chronic inflammatory demyelinating polyneuropathy. Brain 125:1850, 2002. 29. Davis, H., and Forbes, A.: Chronaxie. Physiol. Rev. 16:407, 1936. 30. Fowler, C.: Early history of nerve conduction studies and electromyography. In Osselton, J. W. (ed.): Clinical Neurophysiology. Oxford, UK, Butterworth-Heinemann, p. 45, 1995. 31. Frankenhaeuser, B., and Huxley, A. F.: The action potential in myelinated nerve fibres of Xenopus laevis as computed on the basis of voltage clamp data. J. Physiol. (Lond.) 171:302, 1964. 32. Gilliatt, R. W., and Willison, R. G.: The refractory and supernormal periods of the human median nerve. J. Neurol. Neurosurg. Psychiatry 26:136, 1963. 33. Hill, A. V.: Excitation and accommodation in nerve. Proc. R. Soc. B 119:305, 1936. 34. Hirayama, K., Tsubaki, T., Toyokura, Y., and Okinaka, S.: Juvenile muscular atrophy of unilateral upper extremity. Neurology 13:373, 1963. 35. Hodes, R., Larrabee, M.G., and German, W.: The human electromyogram in response to nerve stimulation and the conduction velocity of motor axons. Arch. Neurol. Psychiatry 60:240, 1948. 36. Hodgkin, A.L., and Huxley, A.F.: A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. (Lond.) 117:500, 1952. 37. Horn, S., Quasthoff, S., Grafe, P., et al.: Abnormal axonal inward rectification in diabetic neuropathy. Muscle Nerve 19:1268, 1996. 38. Inglis, J. T., Leeper, J. B., Wilson, L. R., et al.: The development of conduction block in single human axons following a focal nerve injury. J. Physiol. (Lond.) 513:127, 1998. 39. Kanai, K., Kuwabara, S., Arai, K., et al.: Muscle cramp in Machado-Joseph disease: altered motor axonal excitability properties and mexiletine treatment. Brain 126(Pt. 4):965, 2003. 40. Kaji, R., Bostock, H., Kohara, N., et al.: Activity-dependent conduction block in multifocal motor neuropathy. Brain 123:1602, 2000.

128

Function of the Peripheral Nervous System

41. Kaji, R., and Kojima, Y.: Pathophysiology and clinical variants of multifocal motor neuropathy. In Kimura, J., and Kaji, R. (eds.): Physiology of ALS and Related Diseases. Amsterdam, Elsevier, p. 85, 1997. 42. Kaji, R., Oka, N., Tsuji, T., et al.: Pathological findings at the site of conduction block in multifocal motor neuropathy. Ann. Neurol. 33:152, 1993. 43. Kiernan, M. C., and Bostock, H.: Effects of membrane polarization and ischaemia on the excitability properties of human motor axons. Brain 123:2542, 2000. 44. Kiernan, M. C., Burke, D., Andersen, K., and Bostock, H.: Multiple measures of axonal excitability: a new approach in clinical testing. Muscle Nerve 23:399, 2000. 45. Kiernan, M. C., Guglielmi, J.-M., Kaji, R., et al.: Evidence for axonal membrane hyperpolarization in multifocal motor neuropathy with conduction block. Brain 125:664, 2002. 46. Kiernan, M. C., Hart, I. K., and Bostock, H.: Excitability properties of motor axons in patients with spontaneous motor activity. J. Neurol. Neurosurg. Psychiatry 70:56, 2001. 47. Kiernan, M. C., Lethlean, A. K., and Blum, P. W.: Monomelic amyotrophy: juvenile muscular atrophy of the upper limb. J. Clin. Neurosci. 6:353, 1999. 48. Kiernan, M. C., Lin, S. Y., Andersen, K. V., et al.: Clinical evaluation of excitability measures in sensory nerve. Muscle Nerve 24:883, 2001. 49. Kiernan, M. C., Mogyoros, I., and Burke, D.: Differences in the recovery of excitability in sensory and motor axons of human median nerve. Brain 119:1099, 1996. 50. Kiernan, M. C., Mogyoros, I., Hales, J. P., et al.: Excitability changes in human cutaneous afferents induced by prolonged repetitive axonal activity. J. Physiol. (Lond.) 500:255, 1997. 51. Kiernan, M. C., Walters, J. L., Andersen, K. V., et al.: Nerve excitability changes in chronic renal failure indicate membrane depolarization due to hyperkalaemia. Brain 125:1366, 2002. 52. Kimura, J.: Refractory period measurements in the clinical domain. In Waxman, S. G., and Ritchie, R. M. (eds.): Demyelinating Disease: Basic and Clinical Electrophysiology. New York, Raven Press, p. 239, 1981. 53. Kugelberg, E.: Accommodation in human nerves and its significance for the symptoms in circulatory disturbances and tetany. Acta Physiol. Scand. 8(Suppl. 24):1, 1944. 54. Kugelberg, E.: Neurologic mechanism for certain phenomena in tetany. Arch. Neurol. Psychiatry 56:507, 1946. 55. Kugelberg, E.: Activation of human nerves by hyperventilation and hypocalcemia: neurologic mechanism of symptoms of irritation in tetany. Arch. Neurol. Psychiatry 60:153, 1948. 56. Kuwabara, S., Cappelen-Smith, C., Lin, C. S.-Y., et al.: Differences in accommodative properties of median and peroneal motor axons. J. Neurol. Neurosurg. Psychiatry 70:372, 2001. 57. Kuwabara, S., Lin, S.-Y., Mogyoros, I., et al.: Voluntary contraction impairs the refractory period of transmission in healthy human axons. J. Physiol. 531:265, 2001. 58. Kuwabara, S., Ogawara, K., Sung, J. Y., et al.: Differences in membrane properties of axonal and demyelinating GuillainBarre syndromes. Ann. Neurol. 52:180, 2002.

59. Lapique, L.: l’Excitabilité en Fonction du Temps: La Chronaxie, Sa Signification et Sa Mesure. Paris, Presses Univ. de Paris, 1926. 60. Layzer, R. B.: The origin of muscle fasciculations and cramps. Muscle Nerve 17:1243, 1994. 61. Lewis, T., Pickering, G. W., and Rothschild, P.: Centripetal paralysis arising out of arrested blood flow to the limb, including notes on a form of tingling. Heart 16:1, 1931. 62. Lin, C. S.-Y., Grosskreutz, J., and Burke, D.: Sodium channel function and the excitability of human cutaneous afferents during ischaemia. J. Physiol. (Lond.) 538:435–446, 2002. 63. Lin, C. S.-Y., Kuwabara, S., Cappelen-Smith, C., and Burke, D.: Responses of human sensory and motor axons to the release of ischaemia and to hyperpolarizing currents. J. Physiol. (Lond.) 541:1025, 2002. 64. Lin, C. S.-Y., Mogyoros, I., and Burke, D.: Recovery of excitability of cutaneous afferents in the median and sural nerves following activity. Muscle Nerve 23:763, 2000. 65. Lin, C. S.-Y., Mogyoros, I., Kuwabara, S., et al.: Accommodation to depolarizing and hyperpolarizing current in cutaneous afferents of the human median and sural nerves. J. Physiol. (Lond.) 529:483, 2000. 66. Lin, C. S.-Y., Mogyoros, I., Kuwabara, S., et al.: Differences in responses of cutaneous afferents in the human median and sural nerves to ischemia. Muscle Nerve 24:1503, 2001. 67. Lucas, K.: The excitable substance of amphibian muscle. J. Physiol. (Lond.) 36:113, 1906. 68. Manis, Y., and Friedman, E. A.: Dialytic therapy for irreversible uremia. N. Engl. J. Med. 301:1260, 1979. 69. Merrill, J. P.: The search for ‘factor X’. Clin. Nephrol. 11:56, 1979. 70. Mogyoros, I., Bostock, H., and Burke, D.: Mechanisms of paresthesias arising from healthy axons. Muscle Nerve 23:310, 2000. 71. Mogyoros, I., Kiernan, M. C., and Burke, D.: Strengthduration properties of human peripheral nerve. Brain 119:439, 1996. 72. Mogyoros, I., Kiernan, M. C., Burke, D., and Bostock, H.: Excitability changes in human sensory and motor axons during hyperventilation and ischaemia. Brain 120:317, 1997. 73. Mogyoros, I., Kiernan, M. C., Burke, D., and Bostock, H.: Strength-duration properties of sensory and motor axons in amyotrophic lateral sclerosis. Brain 121:851, 1998. 74. Nielsen, V. K.: The peripheral nerve function in chronic renal failure. V. Sensory and motor conduction velocity. Acta Med. Scand. 194:445, 1973. 75. Nielsen, V. K.: The peripheral nerve function in chronic renal failure. VIII. Recovery after renal transplantation: clinical aspects. Acta Med. Scand. 195:163, 1974. 76. Noble, D., and Stein, R. B.: The threshold conditions for initiation of action potentials by excitable cells. J. Physiol. (Lond.) 187:129, 1966. 77. Oh, S. J., Clemens, R. S. Jr., Lee, Y. W., and Diethelm, A. G.: Rapid improvement in nerve conduction velocity following renal transplantation. Ann. Neurol. 4:369,1978. 78. Panizza, M., Nilsson, J., Roth, B. J., et al.: The time constants of motor and sensory peripheral nerve fibers measured with the method of latent addition. Electroencephalogr. Clin. Neurophysiol. 93:147, 1994.

Nerve Excitability Measures 79. Pape, H. C.: Queer current and pacemaker: the hyperpolarization-activated cation current in neurons. Annu. Rev. Physiol. 58:299, 1980. 80. Parry, G. J.: Pure motor neuropathy with multifocal conduction block masquerading as motor neuron disease. Muscle Nerve 11:103, 1988. 81. Parry, G. J.: Multifocal motor neuropathy: pathology and treatment. In Kimura, J., and Kaji, R. (eds.): Physiology of ALS and Related Diseases. Amsterdam, Elsevier, p. 73, 1997. 82. Pestronk, A., Cornblath, D. R., Ilyas, A. A., et al.: A treatable multifocal motor neuropathy with antibodies to GM1 ganglioside. Ann. Neurol. 24:73, 1988. 83. Purves-Stewart, J.: The Diagnosis of Nervous Diseases, 9th ed. London, Edward Arnold, 1945. 84. Purves-Stewart, J., and Worster-Drought, C.: The Diagnosis of Nervous Diseases, 10th ed. London, Edward Arnold, 1952. 85. Raskin, R.: Neurological complications of renal failure. In Aminoff, M. J. (ed.): Neurology and General Medicine. New York, Churchill Livingstone, 2001. 86. Rasminsky, M.: The effects of temperature on conduction in demyelinated single nerve fibres. Arch. Neurol. 28:287, 1973. 87. Ritchie, A. R.: The electrical diagnosis of peripheral nerve injury. Brain 67:314, 1944. 88. Ritchie, J. M.: Physiology of axons. In Waxman, S. G., Stys, P. K., and Kocsis, J. D. (eds.): The Axon. Oxford, UK, Oxford University Press, p. 68, 1995. 89. Roth, G., Rohr, J., Magistris, M. R., and Ochsner, F.: Motor neuropathy with proximal multifocal persistent conduction block, fasciculations and myokymia: evolution to tetraplegia. Eur. Neurol. 25:416, 1986. 90. Rushton, W. A. H.: The time factor in electrical excitation. Biol. Rev. 10:1, 1935. 91. Scribner, B. H.: Discussion. Trans. Am. Soc. Artif. Intern. Organs 11:29, 1965. 92. Schilling, T., Heinrich, B., Kau, R., et al.: Paclitaxel administered over 3h followed by cisplatin in patients with advanced head and neck squamous cell carcinoma: a clinical phase I study. Oncology 54:89, 1997. 93. Schwarz, J. R., Reid, G., and Bostock, H.: Action potentials and membrane currents in the human node of Ranvier. Pflügers Arch. 430:283, 1995.

129

94. Sears, T. A., and Bostock, H.: Conduction failure in demyelination: is it inevitable? Adv. Neurology 31:357, 1981. 95. Stephanova, D. I., and Bostock, H.: A distributed-parameter model of the myelinated human motor nerve fibre: temporal and spatial distributions of electrotonic potentials and ionic currents. Biol. Cybernet. 74:543, 1996. 96. Stys, P. K., and Ashby, P.: An automated technique for measuring the recovery cycle of human nerves. Muscle Nerve 17:969, 1994. 97. Stys, P. K., Ransom, B. R., Black, J. A., and Waxman, S. G.: Anoxic/ischemic injury in axons. In Waxman, S. G., Kocsis, J. D., and Stys, P. K. (eds.): The Axon. Oxford, UK, Oxford University Press, p. 462, 1995. 98. Sumner, A.: Consensus criteria for the diagnosis of partial conduction block and multifocal motor neuropathy. In Kimura, J. and Kaji, R. (eds.) Physiology of ALS and Related Diseases. Amsterdam, Elsevier, p. 221, 1997. 99. Taylor, B. V., Wright, R. A., Harper, C. M., and Dyck, P. J.: Natural history of 46 patients with multifocal motor neuropathy with conduction block. Muscle Nerve 23:900, 2000. 100. Taylor, J. L., Burke, D., and Heywood, J.: Physiological evidence for a slow K conductance in human cutaneous afferents. J. Physiol. (Lond.) 453:575, 1992. 101. Vagg, R., Mogyoros, I., Kiernan, M. C., and Burke, D.: Activity-dependent hyperpolarization of human motor axons produced by natural activity. J. Physiol. (Lond.) 507:919, 1998. 102. Weigl, P., Bostock, H., Franz, P., et al.: Threshold tracking provides a rapid indication of ischaemic resistance in motor axons of diabetic subjects. Electroencephalogr. Clin. Neurophysiol. 73:369, 1989. 103. Weiss, G.: Sur la possibilité de rendre comparable entre eux les appareils servant à l’excitation électrique. Arch. Ital. Biol. 35:413, 1901. 104. Yang, Q., Kaji, R., Takagi, T., et al.: Abnormal axonal inward rectifier in streptozotocin-induced experimental diabetic neuropathy. Brain 124:1149, 2001. 105. Yokota, T., Saito, Y., Yuki, N., and Tanaka, H.: Persistent increased threshold of electrical stimulation selective to motor nerve in multifocal motor neuropathy. Muscle Nerve 19:823, 1996.

6 The Muscle Spindle ROBERT W. BANKS

Muscle Spindles of the Cat Soleus Abundance of Muscle Spindles and Afferent Axons

Development of Muscle Spindles Intrafusal Muscle Fibers

One of the primary functions of the peripheral nervous system is the execution of motor commands issued by the brain and spinal cord to the skeletal muscles. The complexity of this task is such that the central nervous system requires a constant stream of information about the kinematic and dynamic states of its muscles, enabling them to be continuously modified so as to complete a command successfully in a wide variety of circumstances. There are two principal kinds of encapsulated sense organ in skeletal muscles that are responsible for providing this information: the tendon organ, which senses the force of contraction (see Chapter 7); and the muscle spindle, which responds to muscle length. The afferent axons that supply their sensory endings are among the largest (and therefore most rapidly conducting) in the body. Muscle spindles and tendon organs together account for virtually all group I axons (conventionally differentiated as Ia and Ib), and muscle spindles alone account for a substantial proportion of group II axons. Ruffini’s morphologic categories of primary and secondary endings99 have long been identified as the intrafusal terminals of Ia and II afferents, respectively. Furthermore, unlike the passive tendon organ, the responses of the muscle spindle’s sensory endings are modifiable by motor input supplied by an exclusive system of small (gamma) motoneurons, as well as collateral branches of alpha motoneurons. Among the various sense organs, therefore, the eye and the ear are the only ones to exceed the muscle spindle in structural and functional complexity. Thus, although at one level it is possible to state quite simply that the function of the muscle spindle is to act as a length sensor, and this will be true wherever one occurs, at the higher level of integrated motor control systems,

Sensory Innervation Motor Innervation

such as the cyclic movement of the legs in walking or running, there is no statement that is both simple and general to describe the muscle spindle’s function. What is clear, however, is that the information required of the muscle spindle is put to various uses according to the nature of the particular motor task at issue. For readers who wish to explore further the higher level function of the muscle spindle in motor control, the recent reviews by Hultborn61 and McCrea84 provide an entry to the extensive literature. This chapter concerns the structure and function of the mammalian muscle spindle, primarily considered at the level of the individual length sensor, and focusing on the neural components. However, it also examines quantitative aspects of sensory provision in different muscles, specifically the relative abundance of muscle spindles and the richness of their afferent nerve supply. A significant proportion of the myelinated nerve fibers in a peripheral motor nerve, including the largest of these, are involved directly in motor control (as distinct from motor output) and thus constitute an expensive necessity. Consideration of adaptive processes in evolution leads us to suppose that the sensory complement of a particular muscle will tend towards an optimum in which there is a balance between the cost of its provision and the benefit gained in terms of motor control. If suitable, unbiased, measures can be found, it should be possible to obtain comparative data useful in modeling different motor control strategies. Inevitably, most of our information about mammalian muscle spindles has been derived from a very few species, including rat, mouse, and human, but in particular the cat. The muscle spindles of these species share a basic 131

132

Function of the Peripheral Nervous System

organization that may be summarized as follows: (1) with very few exceptions, each muscle spindle consists of three types of specialized muscle fiber forming a small bundle; (2) these intrafusal fibers are fully cross-striated for most of their lengths, except for the segments contacted by the sensory endings, where groups of euchromatic nuclei occur (in the equator, or equatorial region); (3) there is a variable number of sensory endings, at least one of which is always distinguished as a primary ending; (4) the striated portions of the intrafusal fibers on either side of the sensory endings (the poles, or polar regions) are innervated by a very variable motor supply; and (5) most of the innervated part of the muscle spindle is enclosed by a multilayered capsule that is expanded in the region of the sensory endings, so as to enclose the sensory endings in a gel-filled periaxial space whose contents are further individually wrapped by fibroblast-like inner capsule cells. Free afferents and a nonsomatic motor (autonomic) innervation sometimes occur in muscle spindles, but their sporadic appearance shows that they are not essential for spindle function. Neither they, nor the capsule, is considered further in the present chapter; additional information on these topics may be found in Banks and Barker.9 Some information is available on the structure of muscle spindles from a wide variety of mammalian species, often in the older literature that has been thoroughly reviewed by Barker19 and classified by Eldred et al.46,47 These descriptions are consistent with the idea that the basic organization given above is common to all mammals, though the number of types of intrafusal fiber (initially in rat, rabbit, cat, and monkey) was not finally established until the 1970s.14,89

MUSCLE SPINDLES OF THE CAT SOLEUS With this essential background in place, we may now proceed to take a particular muscle as a paradigm and give a fuller description of its complement of muscle spindles. We may take the soleus of the cat, one of very few muscles known in sufficient detail, for this purpose. The soleus is the deepest head of the ankle extensor, or triceps surae; its accessibility, simple structure, and homogeneous composition make it especially suitable for physiologic study. In the adult cat, the soleus weighs just over 4 g and is almost 85 mm long.100 Apart from the specialized intrafusal fibers, the muscle is made up entirely of slow-twitch (type S) fibers.34 These constitute the great bulk of the muscle, numbering 22,000 to 30,00037; nevertheless, it is convenient to adopt the usual fusicentric epithet of extrafusal to describe them. The extrafusal fibers have a mean length of about 42 mm,100 giving the muscle an index of architecture of 0.49 in a unipennate arrangement. There are, on average, 56 muscle spindles in the cat soleus36 (Fig. 6–1A). Individual spindles vary consider-

ably in size and complexity while usually conforming to the basic organization summarized above. This is illustrated by a representative sample of six muscle spindles shown in Figure 6–1B to F, all of which are derived from a single soleus muscle. Two of them (Fig. 6–1B) are linked end to end in tandem fashion, though only one of the intrafusal fibers (the bag2; see below) is continuous through both spindles. Their combined length overall is 10.2 mm. The remaining separate spindles (Fig. 6–1C to F) range in length from 5.3 to 8.5 mm, with a mean of 7.5 mm. The locations of the sensory endings are marked alongside the expanded portions of the capsules: three spindles (Fig. 6–1C, F, and the smaller of the tandem pair in B) each contain just a single primary (P) ending; two each contain a primary and one secondary (S1) ending (Fig. 6–1D and the larger of the tandem pair in B); and one (Fig. 6–1E) contains a primary and two secondary endings. In this case both secondary endings lie on the same side of the primary and are thus designated S1 and S2, but in general two or more secondary endings may be located on either or both sides of the primary, apparently at random. The three types of intrafusal fiber are designated as bag1, bag2 and chain.14,89 Their properties and distinguishing features are described below; for the moment we note that, in the sample shown in Figure 6–1, the intrafusal bundle of each muscle spindle consists of one bag1, one bag2, and several chain fibers, except for the smaller of the tandem pair (Fig. 6–1B), which lacks a bag1 fiber. We may thus conveniently distinguish b1b2c spindles from b2c spindles, the proportions of the two kinds differing according to muscle type.18 The lengths of the muscle spindles given above usually correspond to those of one or other bag fiber, typically the bag2. The encapsulated portion of the muscle spindle is often considerably shorter, and both bag fibers usually extend beyond the ends of the capsule. In the four separate muscle spindles of Figure 6–1, the mean length of the capsules is 3.9 mm; one spindle (Fig. 6–1F) has a particularly long extracapsular pole, whereas another (Fig. 6–1C) has one very short pole that inserts directly into tendon (note the tendon organ alongside) without any extracapsular portion whatever. Despite the similarity in the complements of intrafusal fibers of most soleus muscle spindles, the pattern of motor innervation varies widely. Two examples from the sample of Figure 6–1 are shown in greater detail in Figure 6–2 to illustrate the range from simple to complex, without in either case marking the extremes of variability. The muscle spindle shown in Figures 6–1F, 6–2B, and 6–2C is the simpler of the two. Its entire somatic innervation is enclosed within the capsule, as is commonly the case. It is supplied by a single small nerve, which contains four separate motor axons in addition to the Ia afferent. In a remarkably symmetrical arrangement, each pole receives two motor axons, one ending on the bag1 fiber in a single

The Muscle Spindle

133

FIGURE 6–1 Muscle spindles in the soleus muscle of the cat. A, Projection plan of a soleus muscle of a kitten, reconstructed from serial longitudinal sections. Lines extending between the origin (o) and insertion (i) indicate the unipennate fascicular structure. Muscle spindles (dots show the locations of their equatorial regions) are predominantly distributed close to the intramuscular branches of the nerve (n). (Modified from Chin, N. K., Cope, M., and Pang, M.: Number and distribution of spindle capsules in seven hindlimb muscles of the cat. In Barker, D. [ed.]: Symposium on Muscle Receptors. Hong Kong, Hong Kong University Press, p. 241, 1962.) B through F, Examples of whole muscle spindles, stained with silver nitrate and teased from a single soleus muscle of an adult cat, showing the locations of their primary (P) and secondary (S1, S2) sensory endings. Note that one spindle is closely associated with a tendon organ (t o). Curvature of the spindles is artifactual, resulting from the teasing process.

large end plate (Fig. 6–2Bii and Biv), and the other, after branching repeatedly, ending on the bag2 and the chain fibers (Fig. 6–2Bi and Biii). Further enlargement (Fig. 6–2C) of the axon shown in Figure 6–2Biii allows its mostly myelinated preterminal branches (blue) to be differentiated from the terminal branches (red) that constitute the separate endings. The endings of this one axon are themselves of very variable form, ranging from large, well-

defined end-plate–like structures in more polar locations, such as those on the bag2, to structures close to the sensory ending of such simplicity that it is impossible to be sure whether they are in neuroeffective contact with the associated chain fibers. In the more complex muscle spindle (Figs. 6–1E and 6–2A), the innervation of the encapsulated part enters the spindle via five small nerves. Progressing from pole to

134

Function of the Peripheral Nervous System

FIGURE 6–2 A and B, The complete innervation of two soleus muscle spindles, as revealed by silver staining and teasing. (The complete spindles are shown in Figure 6–1E and 6–1F, respectively.) In each case (Ai to Ax and Bi to Biv), the intrafusal branching and motor endings of every motor axon entering the spindles are shown at larger scale (calibration refers to these details). Six axons (Aiii, Avii, Aviii, Ax, Bii, and Biv) supply the bag1 fibers exclusively, two supply chain fibers alone (Aiv and Avi), and two (Av and Aix) supply a bag2 fiber alone. Bag2 and chain fibers are innervated together by three axons (Aii, Bi, and Biii). The small endings shown in Ai are located on a bag1 fiber and an unusually long chain fiber, but whether they were innervated by a single axon or by two separate axons is unknown. C, Further enlargement of the intrafusal branches (blue) and motor endings (red) of the axon Biii, which provides the entire motor innervation of the bag2 and chain fibers in one pole of spindle B.

pole, and beginning with that which contains the two secondary endings, we see the following axonal distributions in each small nerve: (1) two unbranched motor axons supplying separate endings to the bag1 fiber (Fig. 6–2Ax) and bag2 (Fig. 6–2Aix) fibers; (2) a group II afferent supplying the S2 sensory ending together with three motor axons, two of which are unbranched and provide separate endings to the bag1 fiber (Fig. 6–2Avii and Aviii), whereas the third branches to some three separate endings, all on the bag2 fiber (Fig. 6–2Av); (3) a group II afferent supplying the S1 sensory ending together with two motor axons, both ending only on chain fibers, one after more extensive branching (Fig. 6–2Aiv) than the other (Fig. 6–2Avi); (4) the group Ia afferent alone entering the equatorial region; and finally (5) two motor axons, one unbranched to a single large ending on the bag1 fiber (Fig. 6–2Aiii) and the other

much branched and supplying several endings of varying form to the bag2 and chain fibers (Fig. 6–2Aii). Thus, of the 13 motor axons that innervate the encapsulated portions of the two spindles, 6 supply the bag1 fibers exclusively, 2 supply a bag2 and 2 some chain fibers separately (all in the same pole), and each of the remaining 3 supply both bag2 and a group of chain fibers. The more complex muscle spindle has, in addition, a small group of endings in the extracapsular portion of one pole (Fig. 6–2Ai). They resemble, in form and size, the end plates of extrafusal fibers and are likely to be derived by collateral branching of one or two axons with an otherwise skeletomotor distribution. In this case they are located on the bag1 fiber and an unusually long and wide chain fiber, but the supplying axons are broken off close to the terminals.

The Muscle Spindle

The number of chain fibers present in each intrafusal bundle is not readily discernible in teased, whole-muscle spindles, but according to the data in Table 4 of Boyd,28 the mean number of chain fibers in soleus muscle spindles is 4.4 and the mean total number of intrafusal fibers is 6.4. In the complete muscle there are thus about 360 (56  6.4) intrafusal fibers in total. In round figures, the nerve to the soleus contains 450 myelinated axons of which perhaps 180 are sensory and 270 are motor.31,88 Most of the afferents can be accounted for by the 56 muscle spindles and 45 tendon organs,103 with the spindles receiving some 70 group II afferents (estimated from data in Banks and Stacey18); the balance would be made up of free-ending and paciniform afferents.19 Boyd and Davey31 estimated that some 115 of the motor axons are gamma efferents and therefore purely fusimotor in distribution. Thus the 360 intrafusal fibers are innervated exclusively by about 125 afferent and 115 efferent axons, or 53% of the myelinated fibers, whereas the (say) 25,000 extrafusal fibers are innervated by 155 alpha motoneurons, representing only 34% of all myelinated fibers in the nerve. Moreover, an unknown proportion (in this muscle, but see below) of the alpha efferents would in fact have a skeletofusimotor distribution, further increasing the numerical dominance of the control over the output systems in the innervation of the soleus.

ABUNDANCE OF MUSCLE SPINDLES AND AFFERENT AXONS Now that we have taken the soleus of the cat as our prime example, we must inquire how typical is its complement of muscle spindles, and the richness of their innervation, in comparison with other mammalian muscles. Of course, different exemplars of a particular muscle exhibit variability in the number of muscle spindles that they contain; nevertheless, it is clear that each muscle has a characteristic complement.36 The coefficient of variation for the number of muscle spindles in soleus of the cat is 13% (56  7 [mean  standard deviation]36). It so happens that the mean coefficient of variation was also 13% for a population of one forelimb and nine hind limb muscles of the cat, in which the mean number of muscle spindles ranged from 25 (interosseus manus V, 0.21 g) to 114 (semitendinosus, 6.41 g).36 Perhaps not surprisingly, therefore, it appears that larger muscles may possess absolutely more muscle spindles than smaller ones. But are muscle spindles relatively more abundant in larger or smaller muscles, if, indeed, there are any differences? For many authors, the answer is simply provided by dividing the number of muscle spindles by the adult weight (or mass) of the muscle to give a parameter known, following Voss,106 as spindle frequency or spindle density. By this measure, smaller muscles tend to have higher spindle densities than larger ones, so, for

135

example, the spindle densities of the human lumbricalis manus III and the gastrocnemius (both medial and lateral heads together) are 12.2 and 0.4, respectively.18 This seemingly large disparity is often taken to reflect the different functional roles of the lumbricalis in fine control of hand movements and of the gastrocnemius in gross movement of the whole body in locomotion, on the intuitive assumption that the former requires more length feedback control than the latter. Nevertheless, it is also clear that there is a nonlinear relationship between the number of muscle spindles in a particular muscle and the mass of the muscle. This prevents a straightforward comparison of the spindle densities of muscles of different sizes, even those belonging to a single species.18 That there may be some more fundamental relationship involved is indicated by comparison of the human data for the lumbricalis manus III and gastrocnemius with the corresponding data for the cat. Here the spindle densities are 173 and 6.5, respectively, although the cat is not noted for manual dexterity. But notice that both these spindle densities are about 15 times those of the human muscles. Encouraged by such observations, Banks and Stacey,18 using an allometric approach, found a linear regression relationship between the logarithms of spindle number and muscle mass. The sample consisted mainly of hind limb muscles and included most of the data then available from the rat and cat, together with an arbitrarily chosen sample of human muscles. The slope of the regression line was 0.32, a value geometrically significant as essentially the same as that to be expected if the number of muscle spindles were scaling isometrically with a parameter of length (i.e., proportional to the cube root of muscle mass, assuming constant physical density). A similar regression slope has been found for the soleus muscle alone, with data from the mouse, rat, cat, and human,8 indicating that the isometric relationship may be a common feature of homologous muscle series. Of course a regression line is a statistical abstraction, and individual muscles deviate more or less from it, but since the deviation is characteristic of a given muscle, it is presumably of functional significance, and its value can be taken as a measure of the relative abundance of spindles in that particular muscle. At first it appeared as though this musclespecific variability might show a regression against size for the different muscles of a single species similar to that for homologous muscles of different species,18 but it is now clear that it does not.8 Voss107 has tabulated an almost complete set of data for the human musculature, allometric analysis of which8 again revealed a linear regression between the logarithms of numbers of muscle spindles and muscle masses, but with a slope of 0.5. Comparison of a group of homologous feline and human muscles further indicated that a similar relationship may occur in the cat and, by extension, other mammals. Isometric scaling therefore applies only to comparisons between species; within a species, the scaling effect of muscle size appears to be

136

Function of the Peripheral Nervous System

a square-root function. Where does this leave the relative abundance of muscle spindles in human lumbricalis manus III and gastrocnemius? If we express the relative abundance as the observed number of spindles divided by the number expected from the regression relationship, that of lumbricalis manus III is 1.01 and of gastrocnemius is 0.41, so the lumbricalis emerges as almost precisely average, whereas the gastrocnemius is clearly rather poorly supplied with spindles for a muscle of its size. The muscle with the greatest relative abundance is the longissimus capitis (8.94) and that with the least is the digastric (0.13, ignoring the only two muscles recorded by Voss107 as possessing no spindles). The soleus, incidentally, has a relative abundance of 0.99. The sensory complement of the individual spindles of a single muscle varies widely, as is clear even from the small sample of Figure 6–1. In the cat soleus it has been found to range from a single primary up to a primary and four secondaries, all supplied by separate afferent axons.12 However, the overall population of spindles of a particular muscle has a characteristic pattern of sensory innervation, as shown by a specific distribution of the frequency of occurrence of the various individual complements,18 For the b1b2c spindles from a sample of 11, mostly hind limb, muscles of the cat, the distributions were adequately described by binomial statistics, the parameters of which (n, p) varied independently among the heteronymous muscles.18 Thus n varied from 3 (superficial and deep lumbrical muscles, and interosseus) to 22 (extensor digitorum longus), whereas p varied from approximately 0.07 (extensor digitorum longus) to 0.58 (complexus), for the theoretical distributions that best fitted the observed distributions. The values for soleus were n  10 and p ⬵ 0.14, corresponding to (and in part derived from) the observed mean number of 1.36 afferents per b1b2c spindle18 that were additional to the single Ia that is both necessary and sufficient for the spindle’s initial differentiation. In cat spindles these additional afferents usually form secondary endings, whereas in the rat masseter many of them contribute to multiply innervated primary endings.11 The proportion of muscle spindles of the b2c type seems to be another factor contributing to the characteristic sensory complement of different muscles. In Banks and Stacey’s18 sample this ranged from 0% in interosseus to 34% in complexus, the proportion in soleus being 11%. Almost all of these have a single sensory ending that is perhaps best recognized as a b2c primary, though its axon is intermediate in caliber between Ia and II afferents12 and its response to stretch cannot be modulated by dynamic fusimotor input (see below). A single secondary ending was additionally present in only 10% of the total number of b2c spindles in the sample of Banks et al.,12 and none of these spindles had more than two afferents. There is very little comparative information available for the sensory complements of the muscles of other species.

However, Banks and Stacey have collected some unpublished data on 98 spindles from the soleus, peroneus longus, and deep lumbrical of the rat, all of the b1b2c type. In the soleus sample, the number of afferents ranged from 1 to 4, the mean number of “additional” afferents being 1.08, and the best fitting binomial parameters were n  2 and p  0.54.

DEVELOPMENT OF MUSCLE SPINDLES In this short review there is space only for a consideration of the events that occur during normal development. (For an extensive review of abnormal development following nerve or muscle injury, as well as regeneration and reinnervation in the adult, see Zelená.109) The following account is based mainly on the rat, which has been the focus of most structural and immunohistochemical studies. In recent years the genetic basis of muscle spindle development in the mouse has become tractable through the application of knockout and transgenic technology. Gestation in the rat lasts for 21 to 23 days. At about embryonic day 15 (E15), the muscle primordia in the hind limb consist of undifferentiated primary myotubes that are immunoreactive for a tonic myosin heavy chain (MHC).75 In the absence of direct evidence to the contrary, it is simplest to assume that these myotubes are able to differentiate either as intrafusal or extrafusal muscle fibers and that they will differentiate as intrafusal fibers if, and only if, they are contacted by afferent axons (for a contradictory view, see Soukup et al.102). The arrival of the innervation, both sensory (the presumptive Ia afferents) and motor (the presumptive alpha motoneurons), is virtually coincident with the first clear differentiation of the primary myotubes at E17, when the tonic MHC is upregulated in some of the myotubes (the presumptive bag2 fibers) and downregulated in the rest (presumptive type I extrafusal fibers),75,76 although both types express similar levels of neonatal and slow-twitch MHC at this time.90 The secondary myotubes that develop in association with the presumptive bag2 myotube give rise successively to the bag1 fiber and the chain fibers, in a pattern that seems to be a modification of the extrafusal version.85,86 The presumptive bag1 fiber expresses only neonatal MHC when it first appears at E19; around 2 days later the neonatal MHC is replaced by slow-twitch and tonic MHCs.90,102 When the future chain fibers first appear, from E21 to postnatal day 4 (P4), they also express neonatal MHC, although there is disagreement as to whether it is expressed alone90,102 or together with slow- and fast-twitch isoforms.74 It may be that the myoblasts that give rise to these secondary myotubes are themselves multipotent; equally, there may be separate lineages, since the presumptive bag1 and chain myotubes appear together rather than sequentially in human spindles.91 However, it should be noted that the

137

The Muscle Spindle

secondary, unlike the primary, myotubes are assembled in the presence of the afferent innervation. If it is indeed true that at least the primary myotubes are undifferentiated until the arrival of the innervation, a relatively simple model of random instructional contact by the first afferents to enter the muscle primordium may then be postulated, thus obviating any necessity for the afferents to locate specific myotubes. The afferents themselves can then form the pathways to guide the remaining intrafusalspecific innervation (group II afferents and gamma efferents) to their appropriate targets, and such a mechanism can adequately account for many features of the pattern of sensory and motor innervation seen in adult spindles.7 Even the sequential development of recognizable bag1 and chain intrafusal muscle fibers requires the continued presence of the sensory, but not the motor, innervation, although the motor innervation does affect the differential expression of MHC isoforms along the length of the individual fibers.101,109 As yet little is known about the molecular aspects of the interaction between the afferents and the myotubes that they contact. The earliest to have been identified so far is probably the onset in the mouse of the differential expression of Egr-3, a member of the early growth response family of zinc-finger transcription factors, in presumptive intrafusal, but not extrafusal, muscle fibers.105 This occurs at about the equivalent stage of development in the mouse (E15.5) as the upregulation of tonic myosin at E17 in presumptive bag2 fibers of the rat, as described above. When Egr-3 expression is absent, spindles begin to form but the characteristic slow (tonic-like) myosin is not induced; their sensory and motor innervation is then withdrawn and the incipient spindles degenerate.104 A little later than the onset of Egr-3 expression (at E19 in the rat), the bag2 fibers begin to produce NT-3, one of the neurotrophin family of trophic factors, as judged by in situ hybridization of the messenger RNA (mRNA).39 The bag1 fibers follow suit at E21, and the bag fibers, exclusively in muscle, continue to express NT-3 through to maturity, the mRNA occurring especially in the small amounts of sarcoplasm between the nuclei of the nuclear bags. Tyrosine receptor kinase C (TrkC) is the receptor for NT-3 and it is expressed in a subfraction of dorsal root ganglion (DRG) cells, including the large muscle afferents. Absence of NT-3 or its receptor as a result of targeted disruption of the appropriate gene in the mouse results in various deficiencies, including a lack of muscle spindles and tendon organs, at least in the hind limb.52 For muscle afferents of the trigeminal mesencephalic nucleus, the situation seems more complex since 38% of masseter spindles survive NT-3 knockout and all of them survive in the absence of TrkC.53,81 This is a nice story of mutual dependency of the kind that we should expect to see operating in an integrated structure induced by instructional contact: DRG cells expressing TrkC are dependent on a supply of NT-3 for their survival, and they themselves

induce NT-3 expression in their instructed targets, an expression that continues to be nerve dependent in the adult.40 Lack of spindles (and tendon organs) in the NT-3 knockout mouse seems to be a consequence of greatly increased apoptosis from about E10.5.73 Since this is, of course, somewhat earlier than neuromuscular contact, the DRG cells are presumably sustained at first by NT-3 from a source other than the muscle itself.

INTRAFUSAL MUSCLE FIBERS Space permits only a brief consideration of the intrafusal muscle fibers; the following account is based largely on the recent reviews by Zelená,109 Soukup et al.,102 and Banks and Barker,9 where further details may be found. Typically, the chain fibers are considerably shorter than either the bag1 or bag2 fiber, and often only the bag fibers extend well beyond the ends of the capsule. Nevertheless, in almost all cases, each intrafusal muscle fiber extends from pole to pole within the spindle and thus comprises two contractile polar regions separated by a noncontractile equatorial region bearing sensory terminals. This is perhaps the most fundamental way in which intrafusal and extrafusal muscle fibers differ and therefore deserves emphasizing at the outset. Normally both contractile poles receive motor nerve endings, though in only a minority of cases are they supplied by the same axon (see Banks7 for a review). The intimate nature of the contact between the sensory terminals and the underlying muscle fibers leaves little room to doubt that length changes of the muscle in which the spindle lies are transmitted in more or less modified form to the sensory endings by the intrafusal fibers, and their passive and active mechanical properties contribute significantly to shaping the sensory responses. This is not to deny the existence of additional passive components, both parallel and series, that contribute to the sensory response, nor, indeed, additional active components, including voltage- and Ca2-dependent properties of the sensory endings themselves. When the three types of intrafusal fiber were first clearly recognized,14 the criteria for differentiation included morphology and ultrastructure, but especially enzyme histochemical profiling. Although attempts were made to correlate the profiles of intrafusal and extrafusal muscle fiber types, it soon became clear that the intrafusal fibers possessed unique combinations of properties; thus, for example, the fastest contracting fibers (the chains), while exhibiting an appropriate level of alkaline-stable ATPase activity, were also the most oxidative. Moreover, these properties showed marked regional variation with respect to the location of the primary sensory ending in particular. The introduction of immunohistochemical staining techniques further emphasized the uniqueness of the mammalian intrafusal fibers in demonstrating the

138

Function of the Peripheral Nervous System

occurrence and distribution of various isoforms of the MHC: as expected, chain fibers characteristically express a fast-twitch isoform, whereas bag2 fibers are characterized by a slow-twitch form. However, for most of their length the bag1 fibers express a slow, tonic-like myosin that shares some antigenic similarities with vertebrate tonic extrafusal fibers.80 In mammals this type of extrafusal fiber is found only in a few specialized locations, such as the “orbital rim G fibers” of extraocular muscles, where, most interestingly, these fibers may play the role of nuclear-bag fibers in muscle spindles (see Barker19 for a review). Both the bag1 and bag2 fibers of skeletal muscles show regional variation in myosin expression, the bag1 having a slow-twitch myosin like that of the bag2 in its extracapsular polar regions, whereas the tonic myosin also occurs in the juxtaequatorial region of the bag2 (i.e., its contractile portion within the periaxial space).102 Indeed, the expression of tonic myosin in the presumptive bag2 fiber is one of the earliest signs of differentiation of intrafusal from extrafusal fibers (see above). Intracellular recording20 has shown that, in response to electrical stimulation of a fusimotor axon at 1 Hz, neuromuscular transmission evokes action potentials in chain fibers, and rarely in bag2. Only junctional potentials have been recorded from bag1 fibers, and they are more common in the bag2 (seven of eight recordings) than are action potentials. Direct observation of living spindles in situ25 or in isolation32 has shown that the two polar regions of each intrafusal muscle fiber act as independent contractile units, so any excitation, whether junctional or action potentials, presumably fails to cross the equatorial region. Contraction of the chain fibers is rapid and distributed throughout the pole, whereas that of the bag fibers is focal and slower, especially in the bag1.25 Against expectation, the sites of these foci of contraction did not usually correspond to a motor ending, nor did they correspond to the locations from which junctional potentials had been recorded in similar preparations.10 It should be noted, however, that, to elicit these contractions, repetitive stimulation at rates of up to 110 Hz were used. The greatest discrepancies occurred in the bag1 fibers, whose focal contractions could be as much as 2.5 mm away from the nearest motor ending. The contraction sites were usually located extracapsularly where the bag1 fiber exhibits some twitch-like properties in sarcomere structure and (at least in the rat) myosin isoform content.

SENSORY INNERVATION We have seen that, most probably, the Ia afferent induces the differentiation of the intrafusal muscle fibers. Its peripheral terminals occupy, and indeed define, the equatorial region, a length of about 350 m in b1b2c muscle spindles of the cat hind limb. There is such a close correlation between

the disposition of these sensory terminals and the grouping of the underlying myonuclei into nuclear bags and nuclear chains12 that it seems likely to reflect an important mutual functional dependency. The equatorial myonuclei are frequently described in even the most elementary account of muscle spindles, but our familiarity with them should not distract us from their highly unusual nature. Where else can such an accumulation of euchromatic nuclei (about 80 per bag2 fiber, on average12) be found in such a small space? It seems unlikely that the local demands for mRNA are so high as to require these numbers. Could they, perhaps, be essentially passive elements (volumetrically incompressible springs?) that happen to possess mechanical properties suited to the overall mechanosensory transduction process of the equatorial region? From measurements of overall tension and the distribution of strain in isolated spindles both before and after removal of a portion of the capsule, Poppele et al.95 were able to show that the stiffness of the equatorial sensory region, unlike the polar regions, is not length dependent. At the “rest length” (the shortest length they used, with a just-measurable tension), Poppele et al.95 found the stiffness of the equatorial region to be eight times greater than that of the polar regions. With the spindle extended by about 15%, the equatorial region was only twice as stiff as the polar regions, as a result of the nonlinear (apparently exponential) increase in polar-region stiffness. The relative stiffness of the equatorial region is, at first sight, perhaps somewhat surprising, because the intrafusal muscle fibers, as the most conspicuous longitudinal element, exhibit a local minimum in diameter here.12 Poppele et al.95 suggested that the equatorial stiffness might be due mainly to the sensory terminals themselves; alternatively (or additionally), the fine elastic fibers that attach to the surfaces of the bag fibers over about 300 m on either side of the primary ending3 could be the main passive component involved. It is unclear whether these elastic fibers are continuous with the more robust ones that are a prominent feature surrounding the polar regions of the bag2 fiber, and whose branches are extensively dispersed among the inner and outer capsules,38,54 but if so they are also well placed to absorb tension during active or passive length changes in either pole of a bag fiber, and to provide the necessary restoring force on reversal of the change in length. The sensory terminals that occupy the equatorial region of all the intrafusal fibers constitute the primary ending of Ruffini,99 which he described as annulo-spiral in form. He noted that the primary ending of the cat was usually supplied by a single, very large nerve fiber, but that occasionally it was supplied by two fibers, which remained separate as far as they could be traced (see above). For simplicity, we consider further only those endings derived from single fibers. Reconstruction of three such primary endings from serial sections,4,12 one of which is illustrated in Figure 6–3, has shown that, whereas the overall form of the endings is remarkably constant, the contributions made to that form

The Muscle Spindle

139

FIGURE 6–3 The primary ending of a cat tenuissimus muscle spindle. Isometric reconstruction from serial, 1-m thick, toluidine-blue stained transverse sections. The sensory terminals are shaded gray. Equatorial nuclei are shown as ellipsoids containing a nucleolus (black dot), or as the nucleolus alone. Myelinated preterminal branches are unshaded. A, The entire ending. Note that the Ia afferent has already divided into branches that supply the bag1 terminals and the bag2/chain terminals separately (b1br, b2cbr). B, Removal of most of the preterminal branches to show the terminals on the bag1 (below) and bag2 (above) fibers. Note the dense packing of terminals on the bag1 fiber especially. C, Removal of the bag fibers to show the chain terminals. (Modified from Banks, R. W., Barker, D., and Stacey, M. J.: Form and distribution of sensory terminals in cat hindlimb muscle spindles. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 299:329, 1982.)

by individual terminals are very variable (Fig. 6–4). One may infer from this that the determination of terminal form is due to local interaction between the terminal and the underlying intrafusal muscle fiber, and is not purely an intrinsic property of the terminal itself. The sensory terminals are derived from a system of preterminal branches; the following description of the pattern and distribution of the myelinated preterminal branches of primary endings is based on data relating to the cat tenuissimus muscle, since no other muscle has provided so much information. The data are taken from papers by Banks et al.12,16 and Banks.4 If each ending is considered to begin at the first branching node, then the 34 endings included a total of 275 nodes, of which 101 (37%) branched dichotomously, 18 (7%) trichotomously, and 1 (0.4%) pentachotomously. The number of nodes in individual endings ranged from 3 to 14, and the proportion of branched nodes from 22% to 100%. There were a total of 177 ultimate preterminal branches in a range of 3 to 9 per ending. Nineteen of these were first, 83 second, 72 third, and 3 were fourth order. With few exceptions, each ultimate branch supplied only one type of intrafusal muscle fiber. Of the exceptions, most (seven) supplied bag2 and chain fibers together; in the only remaining case, an ultimate myelinated branch supplied a bag1 together

with a single chain fiber.15 Each of the 35 bag1 fibers received on average more ultimate preterminal branches (1.9  0.28 per fiber [mean  standard error]) than either the 36 bag2 fibers (1.5  0.18 per fiber) or the approximately 140 chain fibers (0.55  0.09 per fiber; data based on the sample of Banks et al.16), although those supplying bag1 fibers tended to be of lower order (modal value 2) than those supplying bag2 (modal values 2 and 3) or chain fibers (modal value 3). No two endings showed the same pattern of branching, but one feature did recur repeatedly: In 27 of the 34 endings, the first branching node was dichotomous and one branch exclusively supplied all the terminals on the (single) bag1 fiber, whereas the other, perforce, supplied the terminals of the bag2 and chain fibers. Moreover, in each of the three endings where the first node branched tri- or pentachotomously, one first-order branch again supplied the bag1 fiber exclusively. Unmyelinated branches form the final distributional stage of the preterminal axons. They arise almost exclusively from the heminodes of the ultimate myelinated segments, are about 1 m in diameter and 10 to 20 m in length, and are sometimes branched (Fig. 6–5). Taken together with evidence from their ultrastructure and histochemistry,98 their dimensions relative to the much larger sensory terminals strongly indicate that they are the sites

140

Function of the Peripheral Nervous System

FIGURE 6–4 Schematic diagrams to illustrate the variability in the preterminal branching pattern and terminal distribution in cat tenuissimus primary endings, based on reconstructions from serial transverse (A and B) or longitudinal (C) 1-m thick sections. Information on the distribution of chain terminals was obtained only from the longitudinal sections. Myelinated internodes are shown as elongated rectangles, and unmyelinated segments of axons are shown as simple lines. Differently shaded regions of the bag1 (b1), bag2 (b2), and chain (c) fibers represent separate terminals. Primary A is the same as that in Figure 6–3; the terminals of primary C are shown in more detail in Figure 6–6. (Modified from Banks, R. W.: Observations on the primary sensory ending of tenuissimus muscle spindles in the cat. Cell Tissue Res. 246:309, 1986. © Springer-Verlag.)

where encoding of the receptor potentials occurs. The transition from preterminal to terminal branch is marked by a sudden increase in diameter, the expanded terminals invariably being in contact with intrafusal muscle fibers. The following description of the sensory terminals is again derived mostly from the cat tenuissimus muscle and the papers of Banks et al.12 and Banks.4 Individual terminals may be simple or branched, with rami of very variable lengths. They wind around the muscle fibers, appearing to occupy the available space in a competitive manner, and so making their own, very variable, contribution to the overall form of the primary ending (see Fig. 6–4). The bag1 fiber has the most, and the most highly branched, terminals; individually, chain fibers have the fewest and the least branched (Fig. 6–6). Consequently, of the total neuromuscular contact area in a primary ending, about 35% is due to terminals on the bag1 fiber, 25% to terminals on the bag2 fiber, and 40% to those on the chain

fibers collectively. Again, the density of terminals is greatest on the bag1 fiber (with about 55% coverage of the equatorial surface of the fiber), intermediate on the bag2 (with about 45% coverage), and least on the chains (with about 42% coverage), so that the overall longitudinal extent of the terminals on each fiber is about the same among the different types. In thin (1-m) sections through the long axes of intrafusal muscle fibers, profiles of the sensory terminals are mostly lentiform in outline (Fig. 6–7), especially toward the middle of the ending, where the pitch of any spirals is small. Both the free (outer) boundary of the terminal profile and the neuromuscular (inner) boundary thus approximate to segments of circles, whose common chord is in the plane of the surface of the muscle fiber and is of fairly constant width, at about 6 m. The shape of these profiles suggests a mechanical model of tension transmission from the intrafusal fibers, whether at constant length or undergoing active or passive length change, in which the terminals are

141

The Muscle Spindle

mpt

upt

to c 1

t

to c 2 t

FIGURE 6–5 Detail of the primary ending of Figure 6–3 showing part of the terminals on the bag1 fiber. The final segment of a myelinated preterminal branch (mpt) gives rise, at the heminode, to a much narrower unmyelinated preterminal segment (upt) that immediately branches to supply two separate, greatly expanded, terminals (t) in contact with the muscle fiber. (Modified from Banks, R. W., Barker, D., and Stacey, M. J.: Form and distribution of sensory terminals in cat hindlimb muscle spindles. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 299:329, 1982.)

squeezed between inner (cellular) and outer (extracellular) tension-bearing elements. It may be supposed that the minimum-energy condition (zero tension) would correspond to a circular terminal profile, that increasing tension results in progressively greater lentiform distortion of the profile, and that consequently the sensory-terminal membrane is progressively stretched as the surface area of a volume element of the terminal membrane is increased. Moreover, differential tension distribution between the inner and outer surfaces of the sensory terminals would result in asymmetrical development of the lentiform profile, as shown by the prominently bulging terminals on the bag1 fibers in Figure 6–7, or, at the opposite extreme of form, the deeply indented ones of the chain fibers. A candidate for the extracellular tension element required by this model is the continuous basal lamina that covers the intrafusal muscle fibers and the outer (free) surfaces of the sensory terminals. There is, however, no intervening basal lamina at the neuromuscular boundary, so the terminals appear to have insinuated themselves between the cellular and extracellular components of the sarcolemma. The lack of dystrophin at this boundary, when it is present elsewhere at the intrafusal muscle fiber membrane,87 is perhaps a secondary consequence of the separation of the sarcolemmal components, but nonetheless interesting in attesting to highly localized variations in mechanical properties. In the absence of any evidence to the contrary, it is simplest to assume that the mechanical properties of the basal lamina and the sensory terminals are constant throughout

1 2 3

FIGURE 6–6 Semischematic arrangement of the terminals of primary C in Figure 6–4. Individual terminals (darker shades) are shown above and below their corresponding locations in the overall arrangement (paler shades) on the bag1 (red), bag2 (green), and chain (blue) fibers. Note that there are only two separate terminals on the three chain fibers, with chain 2 receiving its terminals directly from those of the other two fibers (arrows indicate these “sensory-cross terminals”1). Examples of the longitudinal sections used in the reconstruction are shown in Figure 6–7, spindle ii. (Modified from Banks, R. W.: Observations on the primary sensory ending of tenuissimus muscle spindles in the cat. Cell Tissue Res. 246:309, 1986. © Springer-Verlag.)

the equatorial region. If so, the range of differential form of the terminals (bulging on the bag1, more or less symmetrical on the bag2, and deeply indented on the chains) must reflect the mechanical properties of the different types of intrafusal muscle fiber. Whatever the precise mechanism of terminal distortion and its differential properties, we may suppose that longitudinal stretch of the equatorial region leads to a continuously varying receptor potential, generated by the sensory terminals. Hunt and Ottoson63 were able to record such a potential from single afferent axons (Ia or II) in isolated muscle spindles of the cat, when the spiking activity of the afferents had been blocked with tetrodotoxin. Hunt and

142

Function of the Peripheral Nervous System

Spindle Bag1 fibres

Sarcomere length um

(i)

2.2

(ii)

2.5

iii

2.6

Bag2 fibres (i)

2.35

(ii)

2.5

(iii)

2.5 Chain fibres

(i) 1.95 (ii) (iii)

2.05 20 um 2.25

FIGURE 6–7 Tracings of 1-m thick, longitudinal sections, stained with toluidine blue, through the bag1, bag2, and a sample chain fiber of three cat tenuissimus muscle spindles (i to iii). Mean sarcomere lengths show that the spindles were fixed under different (unknown) amounts of static tension increasing in order, i to iii. Analysis4 of the shape of the terminal profiles (colored areas) indicated that they are progressively deformed by the increasing longitudinal tension. (Modified from Banks, R. W.: Observations on the primary sensory ending of tenuissimus muscle spindles in the cat. Cell Tissue Res. 246:309, 1986. © Springer-Verlag.)

Wilkinson64 went on to show that the fundamental frequencies of both the receptor potential and the overall tension of the muscle spindle responded similarly to sinusoidal stretch. The depolarizing current is carried by Na, but a Ca2 current, which can be blocked by D600, is also discernible when Na is removed from the bathing medium.65 Presumably closure of the mechanosensory channels leads to repolarization of sensory-terminal membranes, but at least two types of K current (one tetraethylammonium sensitive, the other not) also seem to

contribute; in primary endings, the K currents can produce hyperpolarization on release of stretch, though the precise location of these currents within the primary ending remains uncertain.65 Kruse and Poppele,71 adopting a systems analytical approach, provided evidence that a Ca2activated K channel contributes to the mid-frequency (0.4 to 4 Hz) dynamics of the primary ending. Although, under normal conditions, Ca2 makes an insignificant contribution to the receptor potential, it nevertheless seems to play a key role, as yet undefined, in the proper functioning of the primary ending. Thus blockage of Ca2 channels by diltiazem or CoCl2 rapidly results in the abolition of spiking activity.71 The importance of Ca2 is underlined by the variety and abundance of Ca2-binding proteins that occur in the ending, including calbindin D-28k, calretinin (Fig. 6–8A), neurocalcin, and, at least in frog spindles, frequenin.44,48,56,66,108 All are members of the EF-hand superfamily, representing two families of proteins with either six or four EF-hands, thought to act respectively as Ca2 buffers (calbindin D-28k, calretinin), or Ca2-activated switches (neurocalcin, frequenin).33,67 Details of the distribution of the various proteins differ: Apart from the primary sensory ending, calbindin D-28k also occurs in the intrafusal muscle fibers of rat spindles, whereas calretinin occurs in the chain fibers of cat spindles and in cat tendon organs, but is absent from secondary endings and from rat tendon organs.44,48 Maintenance of the ionic composition of such large sensory terminals as those of the primary and secondary endings of muscle spindles may be expected to be metabolically demanding. It is therefore not surprising to find that mitochondria are abundant in them, often being concentrated in their central cores. What is perhaps more surprising is the abundant population of vesicles that is also present. Various types of vesicle occur commonly in many kinds of mechanoreceptor terminal, but until recently they appear to have been dismissed as having no obvious or demonstrable function related to mechanosensory transduction.2 Especially prominent among them are large numbers of small, clear vesicles resembling those of chemical synapses1 (Fig. 6–9). The apparent similarity of these “synaptic-like vesicles” to the classic presynaptic vesicles of motor neuromuscular junctions, for example, was enhanced when De Camilli et al.43 showed that synapsin 1 and synaptophysin occur in the sensory terminals of muscle spindles (see Fig. 6–8B) and tendon organs as well as in motor endings. Use of the fluorescent styryl dye FM1-43 to study the dynamics of vesicle release in the (motor) neuromuscular junction26 led directly to the observation that it also stained the sensory endings of muscle spindles (see Fig. 6–8C), but stimulation of the nerve is necessary to induce uptake of dye into the motor endings, whereas the sensory endings accumulate it spontaneously. It is supposed that the

The Muscle Spindle

143

FIGURE 6–8 A, The distribution of calretinin in the equatorial region of a spindle from the abductor digiti quinti medius muscle of the cat, demonstrated using the peroxidase-antiperoxidase technique (25 m-thick frozen section, calibration 50 m). The primary ending (preterminal and terminal branches) is the most intensely stained, but the chain fibers themselves are also positively stained. B, The distribution of synaptophysin, a protein characteristic of synaptic vesicles, in a semispinalis capitis muscle spindle of the rat, using indirect immunofluorescence and confocal microscopy (25-m-thick frozen section, calibration 50 m). Immunostaining is essentially restricted to the sensory terminals, in this case of a primary ending. Those on the bag2 fiber (above) and a chain fiber (below) are visible in this confocal plane. C, FM1-43 labeling of the sensory terminals in a primary ending of a rat lumbrical muscle spindle (calibration 20 m). The living muscle was maintained in vitro during uptake of the dye, which occurs spontaneously. The muscle was then fixed prior to confocal microscopy. The dye is thought to be incorporated into the membranes of “synaptic-like vesicles”13 within the terminals. (The preparations illustrated were made by Amal El-Tarhouni [A], Ian Maclean and Christine Richardson [B], and Guy Bewick [C].)

FIGURE 6–9 A, Electron micrograph of a transverse section through the equatorial region of a bag fiber in a muscle spindle of cat flexor digitorum longus. Note the location of the sensory terminal (t) between basal lamina (bl) and underlying muscle fiber. B, Detail of the box in A showing small, clear (“synaptic-like”) vesicles (v) in the terminal. An “” figure (arrow) indicates that the vesicles may fuse with the terminal membrane. (The preparations illustrated were made by Mohammed Adal.)

dye becomes incorporated into the vesicular membranes, and if so, it appears that the vesicles in the sensory endings are fused with the terminal membrane and recycled constitutively, though there do not seem to be any specialized release sites. We have now provided evidence that this process of recycling can be modulated in an activitydependent manner; that glutamate is relatively enriched in the sensory terminals; and that exogenously applied glutamate has an excitatory action on muscle spindle afferents.13,17,27 These results have revealed a hitherto unknown, possibly even unsuspected, level of complexity of the mechanosensory properties of muscle spindle afferents, one that will almost certainly prove to have a more general applicability among mechanoreceptors. Much

more work will be needed before its functional significance can be fully understood. We have already seen that encoding of the continuously varying potentials generated by terminal deformation (receptor potentials) is likely to occur preferentially at the heminodes of the ultimate myelinated preterminal branches. But in each primary ending there are several such potential encoding sites, and theoretical considerations indicate that the overall output of the ending (in the form of impulses successfully invading the parent afferent) will be dominated by whichever encoding site momentarily has the highest firing rate.45 This is due to resetting of the encoding sites having lower firing rates at that moment by antidromic invasion of impulses from the site with the

144

Function of the Peripheral Nervous System

highest rate. In extreme circumstances it can lead to complete occlusion of the output of a less active site by that of a more active one, a phenomenon first described from muscle spindle studies, and explained in this way, by Crowe and Matthews.41 Banks et al.16 systematically studied the interactions of encoding sites activated by pairs of dynamic and static fusimotor axons (see following section for a description of these terms). Use of the tenuissimus muscle enabled the physiologic and histologic properties of individual primary endings to be correlated. Banks et al.16 defined a coefficient of interaction (Ci), such that Ci  1 for linear summation of the separate inputs Ci  0 for complete occlusion of one input by the other Although Ci was not constant, standardization of conditions allowed a characteristic mean value to be calculated for each primary ending. Whenever the static and dynamically activated sites were separated by myelinated preterminal branches containing at least 4 complete nodes of Ranvier (12 of 16 primaries), their interactions were dominated by occlusion (Ci  0.35). With fewer than four nodes separating the encoding sites, the overall output usually showed greater summation, and the greatest amount of summation (Ci  0.69) was shown by a primary ending whose activated encoding sites were separated by branches containing just a single complete node. While recognizing that several factors could be involved in the summation, Banks et al.16 argued that electrotonic spread of receptor potentials between the encoding sites was probably the most important (at least under the conditions of their experiments), since others such as mechanical coupling of the intrafusal fibers would not be expected to correlate with primary-ending structure. Having considered the mechanosensory transduction and encoding processes, we are now in a position to understand at least some of the complexities in the final output of a sensory ending (especially of a primary) as it responds to the stimulation of changing muscle length. Quantitative study of the relationship between spindle afferent output and muscle length, including its modification during fusimotor stimulation (see following section), was greatly facilitated by the single-unit technique introduced by Kuffler et al.78 This method has been used, more or less modified, in virtually all subsequent investigations of the response properties of the sensory endings. In the present chapter, it is impossible to treat the many important contributions to this work individually. They have been reviewed by some of the principal authors of this work, including Kennedy et al.,69 Matthews,83 Hulliger,59 Boyd,29 Hunt,62 and Prochazka.97 We conclude this section with a brief summary of the passive responses of sensory endings, before, in the final section, seeing how these are modified by fusimotor stimulation. Linear systems analysis of continuous length changes, both sinusoidal94 and random,93 has provided quantitative descriptions of the responses of primary and second-

ary endings that are applicable within their linear ranges. For example, if the amplitude of stretch does not exceed about 0.1% of muscle length, the response of a primary ending may be fully described by the transfer function: Ks(s+0.44)(s+11.3)(s+44) H(s)=

(s+0.44)(s+0.8) where s  j is the Laplace transform variable and K is a parametric gain constant. A comparison of responses to random and sinusoidal stretches outside the linear range revealed at least two types of nonlinearity93: a static one affecting both primary and secondary endings in which sensitivity (impulses/s/mm) decreases with stretch amplitude for stretches up to about 1% of muscle length; and a dynamic one affecting only primary endings, which, for example, produces a fractional power-law relationship between Ia responses and velocity of stretch during ramp stretches.57,58 There is fairly general agreement that the transition from linearity to (static) nonlinearity corresponds with the limit at which attached cross-bridges in the intrafusal sarcomeres are no longer able to absorb their proportion of the applied stretch. It is quite possible that the dynamic nonlinearity is also due to the mechanical properties of the intrafusal fibers, in this case mainly the bag1. We have seen that this fiber is characterized by a tonic myosin, and that it does not twitch, but there is also evidence that stretch activation may contribute to its dynamic behavior.96 The interaction of separate encoding sites located in the first-order branches of the Ia afferent, and thus associated with the bag1 and bag2  chain fibers, respectively, results in a further nonlinearity that we have already considered (see above). Attempting to identify the source of the nonlinearities in the sensory responses is important for understanding the internal working of the spindle, but this knowledge is much less important in trying to model the behavior of the muscle spindle as part of a general motor control system. For this purpose, it may be sufficient to have a nonlinear transfer function available that accurately predicts the output of the spindle under all stimulation conditions. Attempts to derive such a function using white noise analysis to estimate Wiener kernels have made some progress, but face some important technical difficulties (see Kröller70).

MOTOR INNERVATION Examples of primary endings responding to various combinations of muscle stretch and fusimotor stimulation are shown in Figure 6–10, using unpublished results of Banks and Hulliger. They have been chosen to illustrate several important aspects of the functional organization of fusimotor innervation. The stretches, whether trapezoidal or sinusoidal, are of large amplitude and consequently outside the linear range of the primary.

145

The Muscle Spindle

FIGURE 6–10 Examples of the responses of primary endings to a variety of conditions of muscle stretch and fusimotor stimulation in the peroneus digiti quinti of cat. A and B, Responses of the primary endings (instantaneous frequency plots, upper traces) of two spindles in the same muscle to a series of five trapezoidal stretches (lower traces) in the presence (broad line, middle traces) or absence (narrow line, middle traces) of fusimotor stimulation. Pseudorandom intervals of fusimotor stimulation were used with a mean rate of 100 Hz and 20% coefficient of variation. In each horizontal pair (i to iii, A and B) are shown the effects of stimulating the same fusimotor axon with the same pseudorandom pattern on the responses of the two primary endings: i, a dynamic gamma axon; ii, a static gamma axon that activates the bag2 fiber only in each spindle; and iii, a static gamma axon that activates both bag2 and chain fibers in each spindle. Note the strength of the fusimotor action in Aiii; in the first active stretch cycle of Biii, cross-correlation analysis showed a clear tendency to 1:1 driving (Ia fires an impulse for every motor stimulus), indicating a predominantly chain action. C, The same endings as Ai to Aiii except that the muscle was stretched sinusoidally at 1 Hz and 0.5 mm amplitude. Phase advance for passive cycles was 75 degrees. Phase advance for active cycles was bag1 active (Ci), 64 degrees; bag2 active (Cii), 54 degrees; and bag2 and chains active (Ciii), 47 degrees. D, The same endings as Ai to Aiii except that the muscle was held at constant mean length (trace not shown) while the fusimotor axons were stimulated with three cycles of a rate that varied from 0 to 150 to 0 Hz. Both primary response (upper trace) and stimulation rate (lower trace) are shown as instantaneous frequency plots. Note the weak effect of bag1 activation at constant length (dynamic axon, Di), the high gain (and hysteresis) of bag2 activation at relatively low rates of stimulation (static axon, Dii), and a tendency to drive when the chain fibers are strongly activated (static axon, Diii), but recall that the bag2 fiber is also active here. Calibrations: A to C upper trace ordinate 0 to 300 Hz, lower trace ordinate 5 to 5 mm, and abscissa 0 to 40 s; D ordinates (both traces) 0 to 200 Hz and abscissa 0 to 40 s. (The data were obtained together with Manuel Hulliger while on research leave in his laboratory in Calgary.)

(a)

(b)

(i)

(ii)

(iii)

(c)

(i)

(ii)

(iii)

(d)

146

Function of the Peripheral Nervous System

Figures 6–10A and B show the responses of two primary endings (afferent conduction velocities of 76 and 73 m/s, respectively) in the same peroneus digiti quinti muscle of a cat to a series of trapezoidal stretches. Each primary ending could be activated by several fusimotor axons, three of them in common, as shown here. All three are gamma axons, which are exclusively fusimotor in distribution, as shown by Leksell.79 Fusimotor effects are classified into two principal categories: dynamic, in which the sensitivity of the primary to the rate of stretch is increased; and static, in which the mean firing rate of the primary is increased but the dynamic sensitivity is virtually unaltered or reduced.82 Moreover, single fusimotor axons may be described as dynamic or static in that they induce similar effects in each of the muscle spindles that they innervate.24,42 In the examples shown in Figure 6–10A and B, the axons were stimulated at a mean rate of 100 Hz during the periods of the broad horizontal bars. Axon i (conduction velocity 26 m/s) was dynamic, whereas axons ii (conduction velocity 26 m/s) and iii (conduction velocity 23 m/s) were both static. Figure 6–10C shows the same arrangement and combination of primary ending and fusimotor axons as Figure 6–10A except that the muscle is now stretched sinusoidally at 1 Hz. Despite the nonlinearities (especially that of the encoder when the primary output falls silent), it is possible to fit pure sines with minimum error to the responses expressed as probability density functions. When averaged, the passive-cycle sines have the following properties: peak-to-peak modulation, 57 impulses/s, and phase advance, 75 degrees. The corresponding values when the fusimotor axons are active are, for a mean stimulation rate of 30 Hz (not shown), dynamic gamma (i), 68 impulses/s and 70 degrees; static gamma (ii), 50 impulses/s and 55 degrees; and static gamma (iii), 49 impulses/s and 52 degrees. For a mean stimulation rate of 100 Hz (illustrated), the values are dynamic gamma (i), 84 impulses/s and 64 degrees; static gamma (ii), 56 impulses/s and 54 degrees; and static gamma (iii), 50 impulses/s and 47 degrees. These values are quite typical of those obtained from the soleus muscle spindles in the systematic study by Hulliger et al.60 Note that in each case the peak-to-peak modulation is increased by dynamic action with respect to the passive response, whereas it is unaltered or reduced by static action, but both dynamic and static activation reduce the phase advance. Finally, Figure 6–10D shows the effects of steadily increasing, and then decreasing, the rate of stimulation of the fusimotor axons on the primary response while the muscle is held at constant length. This is the same primary ending and set of motor axons as in Figure 6–10A and C. Note that the primary ending has a marked tendency to fire an impulse for every stimulus pulse (“driving”) when static gamma (iii) is activated, indicating that a fast intrafusal mechanism is responsible.

Elucidation of these and other phenomena of fusimotor action has been the work of several decades and has involved some of the most elegantly conceived and executed experiments in muscle spindle physiology. They deserve to be included here, but space considerations forbid it once again; however, details may be found in the review by Banks.7 Accordingly, we summarize only the main conclusions. The functional classification of gamma axons into dynamic and static types corresponds to their segregated distribution, examples of which we encountered in Figure 6–2. There is no distinction between the conduction velocities of the two types, both ranging from 15 to 55 m/s in the cat hind limb. Dynamic axons are distributed to bag1 fibers, whereas static axons supply bag2 and chain fibers. In the cat these distributions are usually exclusive, but in the rat and monkey several instances of co-innervation of the bag1 with either or both of the other types of intrafusal fiber have been reported.72,77 Even in these species, however, the overall distribution of the axons concerned may be predominantly of one kind or the other, so that the distinction between dynamic and static axons still holds. In any particular muscle spindle, the static axons may be distributed to the bag2 and chain fibers together or separately. The variability in the distribution pattern of static gamma axons accounts for a similar variability in the details of their actions,49 such as those illustrated in Figure 6–10, where static axon ii activates the bag2 fiber of both spindles (A) and (B), and static axon iii activates, in different proportions in the two spindles, the bag2 together with the chain fibers. The possibility that there might be two types of static gamma axon30 was refuted by Banks,5 who showed that there was, nevertheless, a statistical tendency for the faster conducting axons to be more widely distributed (i.e., to more spindles) than the more slowly conducting, and that a driving action, when present, tended to be due to the slowest axons supplying the spindle concerned. Apart from these correlations, the complete distribution of individual static axons could be accounted for by chance. Banks6 went on to demonstrate by histologic analysis that the occurrence of a segregated supply of the bag2 and chain fibers in any particular spindle depended on the total number of static axons entering the spindle, and that this was closely correlated with the number of afferent axons present, which was itself apparently determined at random (see Abundance of Muscle Spindles and Afferent Axons above). Banks’5 conclusions regarding the differential distribution of static gamma axons in relation to their conduction velocities were based on his observations on the cat tenuissimus; qualitatively similar results have since been obtained from two other hind limb muscles, the peroneus digiti quinti (peroneus tertius) and peroneus longus.35,51 The results from these, perhaps more typical, muscles were necessarily obtained using only physiologic information. The criteria for identifying the intrafusal

147

The Muscle Spindle

distribution of individual static gamma axons on this basis have recently been critically examined by Petit et al.92 The complete distribution of any single static gamma axon usually includes both bag2 and chain fibers.21 When we are so used to the concept of homogeneous motor units, this may seem strange. It is true that the differential distribution in relation to conduction velocity could provide the central nervous system with a measure of separate control, but Banks6 has pointed out that, when sufficient static gamma axons supply a single spindle to make segregation possible, it is the inputs to the two poles rather than the bag2 and chain fibers that are likely to be segregated. The probable functional significance of the co-innervation of bag2 and chain fibers by single static gamma axons has been investigated by Emonet-Dénand et al.,50 who reported that coactivation of bag2 and chain fibers enables primary endings to signal changes of length continuously over a larger range of stretch velocities and more independently of the average muscle length than is possible during activation of either type of intrafusal fiber alone. In addition to the purely fusimotor (gamma) innervation, muscle spindles may receive some of their motor input as collateral branches of the axons of alpha motoneurons (see Banks7 for a review). The complete distribution of these axons is therefore often described as skeletofusimotor, and the axons are referred to as beta axons. This usage can be quite convenient, but should not be allowed to imply that they constitute a separate category of motor axon. On the contrary, all the evidence indicates that they are identical to the corresponding alpha axons, apart from the fact that they happen to have collateral input to muscle spindles.7 The number of such collateral branches received by any particular spindle is determined at random, but unlike the gamma axons, there is no correlation with the number of afferents present.6 It is likely that this reflects the necessity for gamma axons to locate their intrafusal targets, which they could most easily do by following pathways established by the foregoing afferents during development. The first physiologic observations of beta innervation in mammals were made by Bessou et al.23 The axons were relatively slowly conducting and had dynamic actions, subsequently shown to be due to bag1 activation. Barker et al.22 argued that the proportion of intrafusal motor endings that they identified as derived from beta axons was too high for them all to have been supplied by collaterals of slow-conducting axons. They suggested that the deficit was made up by fast beta axons, a prediction that was confirmed by glycogen depletion studies on peroneus digiti quinti (p. tertius) by Harker et al.55 These beta axons were static in action, selectively activating long chain fibers such as that shown in Figure 6–2Ai. Once regarded as an accident of development, or as an example of the retention of an evolutionarily primitive feature, skeletofusimotor axons are now seen as an integral feature of the motor control

system in mammals. Based on the physiologic identification of beta axons with conduction velocities of 55 m/s or more (any slower beta axons would have been excluded), Jami et al.68 have estimated that at least 50% of the spindles in peroneus digiti quinti (p. tertius) receive a beta innervation, and that one third of the supply is dynamic and two thirds static. They also found that the majority (11/17, or 69%) of alpha motoneurons that conducted at 75 m/s or less were in fact skeletofusimotor in distribution, whereas only a minority (23/82, or 28%) of those that conducted at more than 80 m/s were.

ACKNOWLEDGMENTS I would like to thank Tomasˇ Soukup for helpful comments. The preparations illustrated in Figures 6–8 and 6–9 were made by Amal El-Tarhouni (8A), Ian Maclean and Christine Richardson (8B), Guy Bewick (8C), and Mohammed Adal (9). I am most grateful to all of them for their help with these images. The data shown in Figure 6–10 were obtained together with Manuel Hulliger while on research leave in his laboratory in Calgary. I am completely indebted to Manuel for these results. I am most grateful to Hong Kong University Press, the Royal Society, and Springer-Verlag for permission to reproduce and modify Figures 6–1B and 6–3 through 6–7.

REFERENCES 1. Adal, M. N.: The fine structure of the sensory region of cat muscle spindles. J. Ultrastruct. Res. 26:332, 1969. 2. Akoev, G. N., Alekseev, N. P., and Krylov, B. V.: Mechanoreceptors. Berlin, Springer-Verlag, 1988. 3. Banks, R. W.: On the attachment of elastic fibres in cat tenuissimus muscle spindles. J. Physiol. (Lond.) 348:16P, 1984. 4. Banks, R. W.: Observations on the primary sensory ending of tenuissimus muscle spindles in the cat. Cell Tissue Res. 246:309, 1986. 5. Banks, R. W.: The distribution of static -axons in the tenuissimus muscle of the cat. J. Physiol. (Lond.) 442:489, 1991. 6. Banks, R. W.: Intrafusal motor innervation: a quantitative histological analysis of tenuissimus muscle spindles in the cat. J. Anat. 185:151, 1994. 7. Banks, R. W.: The motor innervation of mammalian muscle spindles. Prog. Neurobiol. 43:323, 1994. 8. Banks, R. W.: On the number of spindles in mammalian muscles. J. Physiol. (Lond.) 511:69P, 1998. 9. Banks, R. W., and Barker, D.: The muscle spindle. In Engel, A. G., and Franzini-Armstrong, C. (eds.): Myology, 3rd ed. New York, McGraw-Hill (in press). 10. Banks, R. W., Barker, D., Bessou, P., et al.: Histological analysis of cat muscle spindles following direct observation of the effects of stimulating dynamic and static motor axons. J. Physiol. (Lond.) 283:605, 1978.

148

Function of the Peripheral Nervous System

11. Banks, R. W., Barker, D., Saed, H. H., and Stacey, M. J.: Innervation of muscle spindles in rat deep masseter. J. Physiol. (Lond.) 406:161P, 1988. 12. Banks, R. W., Barker, D., and Stacey, M. J.: Form and distribution of sensory terminals in cat hindlimb muscle spindles. Philos. Trans. R. Soc. Lond. (Biol.) 299:329, 1982. 13. Banks, R. W., Bewick, G. S., Reid, B., and Richardson, C.: Evidence for activity-dependent modulation of sensoryterminal excitability in spindles by glutamate release from synaptic-like vesicles. In Gandevia, S. G., Proske, U., and Stuart, D. G. (eds.): Sensori-motor Control of Movement and Posture. New York, Plenum, p. 13, 2002. 14. Banks, R. W., Harker, D. W., and Stacey, M. J.: A study of mammalian intrafusal muscle fibres using a combined histochemical and ultrastructural technique. J. Anat. 123:783, 1977. 15. Banks, R. W., Hulliger, M., and Scheepstra, K. A.: Correlated histological and physiological observations on a case of common sensory output and motor input of the bag1 fibre and a chain fibre in a cat tenuissimus spindle. J. Anat. 193:373, 1998. 16. Banks, R. W., Hulliger, M., Scheepstra, K. A., and Otten, E.: Pacemaker activity in a sensory ending with multiple encoding sites: the cat muscle spindle primary ending. J. Physiol. (Lond.) 498:177, 1997. 17. Banks, R. W., Richardson, C., and Bewick, G. S.: Immunocytochemical demonstration of glutamate in the sensory terminals of rat muscle spindles. J. Physiol. (Lond.) 528:62P, 2000. 18. Banks, R. W., and Stacey, M.: Quantitative studies on mammalian muscle spindles and their sensory innervation. In Hník, P., Soukup, T., Vejsada, R., and Zelená, J. (eds.): Mechanoreceptors: Development, Structure and Function. New York, Plenum Press, p. 263, 1988. 19. Barker, D.: The morphology of muscle receptors. In Hunt, C. C. (ed): Handbook of Sensory Physiology, Vol. 3, Pt. 2: Muscle Receptors. Berlin, Springer-Verlag, p. 1, 1974. 20. Barker, D., Bessou, P., Jankowska, E., et al.: Identification of intrafusal muscle fibres activated by single fusimotor axons and injected with fluorescent dye in cat tenuissimus spindles. J. Physiol. (Lond.) 275:149, 1978. 21. Barker, D., Emonet-Dénand, F., Laporte, Y., et al.: Morphological identification and intrafusal distribution of the endings of static fusimotor axons in the cat. J. Physiol. (Lond.) 230:405, 1973. 22. Barker, D., Stacey, M. J., and Adal, M. N.: Fusimotor innervation in the cat. Philos. Trans. R. Soc. Lond. (Biol.) 258:315, 1970. 23. Bessou, P., Emonet-Dénand, F., and Laporte, Y.: Occurrence of intrafusal muscle fibres innervated by branches of slow motor fibres in the cat. Nature 198:594, 1963. 24. Bessou, P., Laporte, Y., and Pagès, B.: Similitude des effets (statiques ou dynamiques) exercés par des fibres fusimotrices uniques sur les terminaisons primaires de plusieurs fuseaux chez le chat. J. Physiol. (Paris) 58:31, 1966. 25. Bessou, P., and Pagès, B.: Cinematographic analysis of contractile events produced in intrafusal muscle fibres by stimulation of static and dynamic fusimotor axons. J. Physiol. (Lond.) 252:397, 1975.

26. Betz, W. J., Mao, F., and Bewick, G. S.: Activity-dependent fluorescent staining and destaining of living vertebrate motor nerve terminals. J. Neurosci. 12:363, 1992. 27. Bewick, G. S., Reid, B., and Banks, R. W.: Investigating the role of small clear vesicles in vertebrate mechanosensory endings using rat muscle spindles. J. Physiol. (Lond.) 528: 62P, 2000. 28. Boyd, I. A.: The structure and innervation of the nuclear bag muscle fibre system and the nuclear chain muscle fibre system in mammalian muscle spindles. Philos. Trans. R. Soc. Lond. (Biol.) 245:81, 1962. 29. Boyd, I. A.: Intrafusal muscle fibres in the cat and their motor control. In Barnes, W. J. P., and Gladden, M. H. (eds.): Feedback and Motor Control in Invertebrates and Vertebrates. London, Croom Helm, p. 123, 1985. 30. Boyd, I. A.: Two types of static -axon in cat muscle spindles. Q. J. Exp. Physiol. 71:307, 1986. 31. Boyd, I. A., and Davey, M. R.: Composition of Peripheral Nerves. Edinburgh, E. & S. Livingstone, 1968. 32. Boyd, I. A., and Ward, J.: Motor control of nuclear bag and nuclear chain intrafusal fibres in isolated living muscle spindles from the cat. J. Physiol. (Lond.) 244:83, 1975. 33. Burgoyne, R. D., and Weiss, J. L.: The neuronal calcium sensor family of Ca2-binding proteins. Biochem. J. 353:1, 2001. 34. Burke, R. E., Levine, D. N., Salcman, M., and Tsairis, P.: Motor units in cat soleus muscle: physiological, histochemical and morphological characteristics. J. Physiol. (Lond.) 238:503, 1974. 35. Celichowski, J., Emonet-Dénand, F., Laporte, Y., and Petit, J.: Distribution of static axons in cat peroneus tertius spindles determined by exclusively physiological criteria. J. Neurophysiol. 71:722, 1994. 36. Chin, N. K., Cope, M., and Pang, M.: Number and distribution of spindle capsules in seven hindlimb muscles of the cat. In Barker, D. (ed.): Symposium on Muscle Receptors. Hong Kong, Hong Kong University Press, p. 241, 1962. 37. Clark, D. A.: Muscle counts of motor units: a study in innervation ratios. Am. J. Physiol. 96:296, 1931. 38. Cooper, S., and Gladden, M. H.: Elastic fibres and reticulin of mammalian muscle spindles and their functional significance. Q. J. Exp. Physiol. 59:367, 1974. 39. Copray, J. C. V. M., and Brouwer, N.: Selective expression of neurotrophin-3 messenger RNA in muscle spindles of the rat. Neuroscience 63:1125, 1994. 40. Copray, J. C. V. M., and Brouwer, N.: Neurotrophin-3 mRNA expression in rat intrafusal muscle fibres after denervation and reinnervation. Neurosci. Lett. 236:41, 1997. 41. Crowe, A., and Matthews, P. B. C.: The effects of stimulation of static and dynamic fusimotor fibres on the response of stretching of the primary endings of muscle spindles. J. Physiol. (Lond.) 174:109, 1964. 42. Crowe, A., and Matthews, P. B. C.: Further studies of static and dynamic fusimotor fibres. J. Physiol. (Lond.) 174:132, 1964. 43. De Camilli, P., Vitadello, M., Canevini, M. P., et al.: The synaptic vesicle proteins synapsin I and synaptophysin (protein p38) are concentrated both in efferent and afferent nerve endings of the skeletal muscles. J. Neurosci. 8:1625, 1988. 44. Duc, C., Barakat-Walter, I., and Droz, B.: Innervation of putative rapidly adapting mechanoreceptors by calbindin- and

The Muscle Spindle

45. 46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

calretinin-immunoreactive primary sensory neurons in the rat. Eur. J. Neurosci. 6:264, 1994. Eagles, J. P., and Purple, R. L.: Afferent fibers with multiple encoding sites. Brain Res. 77:187, 1974. Eldred, E., Yellin, H., DeSantis, M., and Smith, C. M.: Supplement to bibliography on muscle receptors: their morphology, pathology and physiology. Exp. Neurol. 55:1, 1977. Eldred, E., Yellin, H., Gadbois, L., and Sweeney, S.: Bibliography on muscle receptors: their morphology, pathology and physiology. Exp. Neurol. Suppl 3:1, 1967. El-Tarhouni, A., and Banks, R. W.: The distribution of calretinin in muscle receptors of the cat. J. Physiol. (Lond.) 487:77P, 1995. Emonet-Denand, F., Laporte, Y., Matthews, P. B. C., and Petit, J.: On the subdivision of static and dynamic fusimotor actions on the primary ending of the cat muscle spindle. J. Physiol. (Lond.) 268:827, 1977. Emonet-Dénand, F., Laporte, Y., and Petit, J.: Functional consequences of bag2 and chain fiber coactivation by static -axons in cat spindles. J. Neurophysiol. 77:1425, 1997. Emonet-Dénand, F., Laporte, Y., and Petit, J.: Comparison of static fusimotor innervation in cat peroneus tertius and longus muscles. J. Neurophysiol. 80:249, 1998. Ernfors, P., Lee, K. F., Kucera, J., and Jaenisch, R.: Lack of neurotrophin-3 leads to deficiencies in the peripheral nervous system and loss of limb proprioceptive afferents. Cell 77:503, 1994. Fan, G. P., Copray, S., Huang, E., et al.: Formation of a full complement of cranial proprioceptors requires multiple neurotrophins. Dev. Dyn. 218:359, 2000. Gladden, M. H.: Structural features relative to the function of intrafusal muscle fibres in the cat. In Homma, S. (ed.): Understanding the Stretch Reflex. Amsterdam, Elsevier, p. 51, 1976. Harker, D. W., Jami, L., Laporte, Y., and Petit, J.: Fast conducting skeletofusimotor axons supplying intrafusal chain fibers in the cat peroneus tertius muscle. J. Neurophysiol. 40:791, 1977. Hietanen-Peltola, M., Pelto-Huikko, M., Rechardt, L., et al.: Calbindin-D-28k-immunoreactivity in rat muscle spindle: a light and electron microscopic study. Brain Res. 579:327, 1992. Holm, W., Padeken, D., and Schäffer, S. S.: Characteristic curves of the dynamic response of primary muscle spindle endings with and without gamma stimulation. Pflugers Arch. 391:163, 1981. Houk, J. C., Rymer, W. Z., and Crago, P. E.: Dependence of dynamic response of spindle receptors on muscle length and velocity. J. Neurophysiol. 46:143, 1981. Hulliger, M.: The mammalian muscle spindle and its central control. Rev. Physiol. Biochem. Pharmacol. 101:1, 1984. Hulliger, M., Matthews, P. B. C., and Noth, J.: Static and dynamic fusimotor action on the response of Ia fibres to low frequency sinusoidal stretching of widely ranging amplitude. J. Physiol. (Lond.) 267:811, 1977. Hultborn, H.: State-dependent modulation of sensory feedback. J. Physiol. (Lond.) 533:5, 2001.

149

62. Hunt, C. C.: Mammalian muscle spindle: peripheral mechanisms. Physiol. Rev. 70:643, 1990. 63. Hunt, C. C., and Ottoson, D.: Impulse activity and receptor potential of primary and secondary endings of isolated mammalian muscle spindles. J. Physiol. (Lond.) 252:259, 1975. 64. Hunt, C. C., and Wilkinson, R. S.: An analysis of receptor potential and tension of isolated cat muscle spindles in response to sinusoidal stretch. J. Physiol. (Lond.) 302:241, 1980. 65. Hunt, C. C., Wilkinson, R. S., and Fukami, Y.: Ionic basis of the receptor potential in primary endings of mammalian muscle spindles. J. Gen. Physiol. 71:683, 1978. 66. Iino, S., Kobayashi, S., and Hidaka, H.: Neurocalcinimmunopositive nerve terminals in the muscle spindle, Golgi tendon organ and motor endplate. Brain Res. 808:294, 1998. 67. Ikura, M.: Calcium binding and conformational response in EF-hand proteins. TIBS 21:14, 1996. 68. Jami, L., Murthy, K. S. K., and Petit, J.: A quantitative study of skeleto-fusimotor innervation in the cat peroneus tertius muscle. J. Physiol. (Lond.) 325:125, 1982. 69. Kennedy, W. R., Poppele, R. E., and Quick, D. C.: Mammalian muscle spindles. In Sumner, A. J. (ed.): The Physiology of Peripheral Nerve Disease. Philadelphia, W. B. Saunders, p. 74, 1980. 70. Kröller, J.: Reverse correlation analysis of the stretch response of primary muscle spindle afferent fibers. Biol. Cybern. 69:447, 1993. 71. Kruse, M. N., and Poppele, R. E.: Components of the dynamic response of mammalian muscle spindles that originate in the sensory terminals. Exp. Brain Res. 86:359, 1991. 72. Kucera, J.: Characteristics of motor innervation of muscle spindles in the monkey. Am. J. Anat. 173:113, 1985. 73. Kucera, J., Fan, G. P., Jaenisch, R., et al.: Dependence of developing group Ia afferents on neurotrophin-3. J. Comp. Neurol. 363:307, 1995. 74. Kucera, J., and Walro, J. M.: Origin of intrafusal muscle fibers in the rat. Histochemistry 93:567, 1990. 75. Kucera, J., and Walro, J. M.: Origin of intrafusal fibers from a subset of primary myotubes in the rat. Anat. Embryol. 192:149, 1995. 76. Kucera, J., Walro, J. M., and Reichler, J.: Role of nerve and muscle factors in the development of rat muscle spindles. Am. J. Anat. 186:144, 1989. 77. Kucera, J., Walro, J. M., and Reichler, J.: Neural organization of spindles in three hindlimb muscles of the rat. Am. J. Anat. 190:74, 1991. 78. Kuffler, S. W., Hunt, C. C., and Quilliam, J. P.: Function of medullated small nerve fibres in mammalian ventral roots: efferent muscle spindle innervation. J. Neurophysiol. 14:29, 1951. 79. Leksell, L.: The action potential and excitatory effects of the small ventral root fibres to skeletal muscle. Acta Physiol. Scand. Suppl. 31:1, 1945. 80. Liu, J.-X., Eriksson, P.-O., Thornell, L.-E., and PedrosaDomellöf, F.: Myosin heavy chain composition of muscle spindles in human biceps brachii. J. Histochem. Cytochem. 50:171, 2002. 81. Matsuo, S., Ichikawa, H., Silos-Santiago, I., et al.: Proprioceptive afferents survive in the masseter muscle of trkC knockout mice. Neuroscience 95:209, 2000.

150

Function of the Peripheral Nervous System

82. Matthews, P. B. C.: The differentiation of two types of fusimotor fibre by their effects on the dynamic response of muscle spindle primary endings. Q. J. Exp. Physiol. 47:324, 1962. 83. Matthews, P. B. C.: Evolving views on the internal operation and functional role of the muscle spindle. J. Physiol. (Lond.) 320:1, 1981. 84. McCrea, D. A.: Spinal circuitry of sensorimotor control of locomotion. J. Physiol. (Lond.) 533:41, 2001. 85. Milburn, A.: The early development of muscle spindles in the rat. J. Cell Sci. 12:175, 1973. 86. Milburn, A.: Stages in the development of the cat muscle spindle. J. Embryol. Exp. Morphol. 82:177, 1984. 87. Nahirney, P. C., and Ovalle, W. K.: Distribution of dystrophin and neurofilament protein in muscle spindles of normal and mdx-dystrophic mice: an immunocytological study. Anat. Rec. 235:501, 1993. 88. Olson, C. B., and Swett, C. P.: A functional and histochemical characterization of motor units in a heterogeneous muscle (flexor digitorum longus) of the cat. J. Comp. Neurol. 128:475, 1966. 89. Ovalle, W. K., and Smith, R. S.: Histochemical identification of three types of intrafusal muscle fibers in the cat and monkey based on the myosin ATPase reaction. Can. J. Physiol. Pharmacol. 50:195, 1972. 90. Pedrosa, F., and Thornell, L.-E.: Expression of myosin heavy chain isoforms in developing rat muscle spindles. Histochemistry 94:231, 1990. 91. Pedrosa-Domellöf, F., and Thornell, L.-E.: Expression of myosin heavy chain isoforms in developing human muscle spindles. J. Histochem. Cytochem. 42:77, 1994. 92. Petit, J., Banks, R. W., and Laporte, Y.: Testing the classification of static axons using different patterns of random stimulation. J. Neurophysiol. 81:2823, 1999. 93. Poppele, R. E.: An analysis of muscle spindle behavior using randomly applied stretches. Neuroscience 6:1157, 1981. 94. Poppele, R. E., and Bowman, R. J.: Quantitative description of linear behavior of mammalian muscle spindles. J. Neurophysiol. 23:59, 1970. 95. Poppele, R. E., Kennedy, W. R., and Quick, D. C.: A determination of static mechanical properties of intrafusal muscle in isolated cat muscle spindles. Neuroscience 4:401, 1979. 96. Poppele, R. E., and Quick, D. C.: Stretch-induced contraction of intrafusal muscle in cat muscle spindle. J. Neurosci. 1:1069, 1981.

97. Prochazka, A.: Proprioceptive feedback and movement regulation. In Rowell, L., and Sheperd, J. T. (eds.): Handbook of Physiology. Sect. 12. Exercise: Regulation and Integration of Multiple Systems. New York, American Physiological Society, p. 89, 1996. 98. Quick, D. C., Kennedy, W. R., and Poppele, R. E.: Anatomical evidence for multiple sources of action potentials in the afferent fibres of muscle spindles. Neuroscience 5:109, 1980. 99. Ruffini, A.: On the minute anatomy of the neuromuscular spindles of the cat, and on their physiological significance. J. Physiol. (Lond.) 23:190, 1898. 100. Sacks, R. D., and Roy, R. R.: Architecture of the hind limb muscles of cats: functional significance. J. Morphol. 173:185, 1982. 101. Soukup, T., Pedrosa, F., and Thornell, L.-E.: Influence of neonatal motor denervation on expression of myosin heavy chain isoforms in rat muscle spindles. Histochemistry 94:245, 1990. 102. Soukup, T., Pedrosa-Domellöf, F., and Thornell, L.-E.: Expression of myosin heavy chain isoforms and myogenesis of intrafusal fibres in rat muscle spindles. Microsc. Res. Tech. 30:390, 1995. 103. Swett, J. E., and Eldred, E.: Distribution and numbers of stretch receptors in medial gastrocnemius and soleus muscles of the cat. Anat. Rec. 137:453, 1960. 104. Tourtellotte, W. G., Keller-Peck, C., Milbrandt, J., and Kucera, J.: The transcription factor Egr3 modulates sensory axon-myotube interactions during muscle spindle morphogenesis. Dev. Biol. 232:388, 2001. 105. Tourtellotte, W. G., and Milbrandt, J.: Sensory ataxia and muscle spindle agenesis in mice lacking the transcription factor Egr-3. Nat. Genet. 20:87, 1998. 106. Voss, H.: Untersuchungen über Zahl, Anordnung und Länge der Muskelspindeln in den Lumbricalmuskeln des Menschen und einiger Tiere. Z. Micros. Anat. Forsch. 42:509, 1937. 107. Voss, H.: Tabelle der absoluten und relativen Muskelspindelzahlen der menschlichen Skelettmuskulatur. Anat. Anz. 129:562, 1971. 108. Werle, M. J., Roder, J., and Jeromin, A.: Expression of frequenin at the frog (Rana) neuromuscular junction, muscle spindle and nerve. Neurosci. Lett. 284:33, 2000. 109. Zelená, J.: Nerves and Mechanoreceptors. London, Chapman & Hall, 1994.

7 The Golgi Tendon Organ JON J. A. SCOTT

Tendon Organ Structure Distribution Physiologic Responses Responses to Motor Unit Contractions

Static Sensitivity Dynamic Sensitivity Recordings during Natural Movements Transduction

The control of movement is highly dependent on sensory feedback from the limbs, which arises from a variety of sensory receptors in the skeletal muscles, joints, and skin. Skeletal muscles are supplied with two encapsulated mechanoreceptors that have a significant role in providing movement-related feedback, the Golgi tendon organ and the muscle spindle, which monitor the tension and length of the muscle, respectively. The tendon organ, innervated by a single Ib afferent axon, is comparatively simple in structure and has certainly attracted less attention than the muscle spindle. As a consequence, there are aspects of its physiology that are still poorly understood. This chapter reviews the structure, distribution, and physiology of the tendon organ in normal muscles and following reinnervation. Some aspects of the role of tendon organ feedback in the spinal and supraspinal control of movement are also addressed, particularly in relation to the evolving picture regarding the reflex actions mediated via the Ib inhibitory interneuron and the significant changes in reflex action occurring during locomotion.

TENDON ORGAN STRUCTURE Tendon organs are located at the junction between skeletal muscle fibers and their associated tendon. The organs are enclosed in a connective tissue capsule and range from 0.2 to 1.5 mm in length, depending on the species and the muscle, and are approximately 0.1 mm in diameter. The majority of tendon organs have a fusiform shape, though bifid and occasional trifid forms occur. The body of the

Recovery from Nerve Lesions Reflex Actions Central Projections and Proprioception

organ comprises a bundle of collagen strands that are continuous, at their proximal end, with a group of muscle fibers and, at their distal end, merge with the collagen of the tendon itself (Fig. 7–1). Thus the inserting muscle fibers are in series with the collagen strands and their contractions will generate tension within those strands. The strands of collagen arise individually from muscle fibers but do not remain distinct; rather, they are tightly intertwined and branch and fuse together, creating a network of longitudinal interconnections.8,61,67 Overall, there is a net fusion of strands from the muscle to the tendinous end such that the number of strands merging with the tendon is fewer than that arising from the muscle fibers.61 As a consequence, individual muscle fibers do not have unique mechanical inputs through the body of the organ. This feature may have consequences for the pattern of neural output. There is considerable variation in the number of muscle fibers inserting into the body of individual tendon organs, with a range of 3 to 50 depending on the muscle and the form of the organ, though the majority of studies indicate ranges of 10 to 25.4,22,61,65,67 Given that the muscle fibers comprising each motor unit are randomly distributed throughout the muscle, each muscle fiber inserting into the tendon organ is likely to be derived from a different motor unit. Physiologic studies and the use of the glycogendepletion technique have confirmed that each motor unit that activates a tendon organ normally contributes one or two fibers only to the group of inserting muscle fibers.22,32,67 In a given muscle, the proportion of muscle fibers inserting into tendon organs is relatively small, thus the majority of fibers insert directly into the tendon and lie in 151

152

Function of the Peripheral Nervous System

A

B FIGURE 7–1 The Golgi tendon organ. A, Light micrograph of a tendon organ from human abductor pollicis brevis muscle. The distal ends of the muscle fibers insert into the collagen strands of the organ, which, in turn, merge into the distal tendon of the muscle. The Ib afferent axon enters the capsule and gives rise to myelinated branches, from which unmyelinated terminals derive and innervate the collagen strands. B, Schematic of a tendon organ at the junction of the gastrocnemius muscle and the Achilles tendon. (From Proske, U.: The Golgi Tendon Organ. 1993.)

parallel with the organs. These fibers, therefore, have the potential to reduce the mechanical tension on an organ and to diminish its response by “unloading.”32,71 In addition to these fibers, some muscle fibers that lie outside the

body of the organ insert into the outer surface of the capsule.8,67,72 Other fibers may insert into collagen strands within the organ that lack innervation.72 The modes of action, if any, of these latter groups of fibers is unclear,

153

The Golgi Tendon Organ

although they may act as if they were in parallel with the organ. The body of the organ is enclosed within a connective tissue capsule between 3 and 12 ␮m thick, comprising, in cross-section, 5 to 20 concentrically arranged capsular cells.51,61 At the point of entry of the afferent axon, the capsular cells are continuous with the perineural sheath of the axon. At the proximal end of the organ, the capsule constricts to form a tight collar around the collagen bundle, which also demarcates the most distal extent of the muscle fibers.61 A similar, distal collar marks the merging of the organ with the main tendon. The lumen of the organ is divided into three types of compartment by “septal cells.”61 These compartments are identified as neuronal compartments, which contain only the myelinated axon close to its entry into the capsule; terminal compartments, which contain collagen bundles along with unmyelinated branches and terminals of the afferent axon; and fibrous compartments, which contain only collagen bundles.51,61,72 Blood vessels enter the capsule along with the afferent axon and course between the layers of the capsule into the lumen of the organ, although they do not appear to be present in the terminal compartments.51 The innervation to the tendon organ is usually via a single, large-diameter, myelinated group I afferent axon. This afferent fiber is commonly known as a Ib afferent axon, to distinguish it from the Ia afferent axon that gives rise to the primary endings of the muscle spindle, although there is almost complete overlap in conduction velocity, except for the fastest conducting fibers. In the cat, Ib afferent axons have conduction velocities of 60 to 110 m/s and range in diameter from 7 to 15 ␮m close to the point of entry into the capsule.4,8 In some instances, Ib afferents branch and innervate two, or very occasionally three, tendon organs.20,65 The Ib afferent axon divides into two or three myelinated branches either close to the point of entry or within the neuronal compartment of the capsule (see Fig. 7–1). These myelinated branches subsequently subdivide and lose their myelin, with further branchings of the unmyelinated, preterminal branches in the terminal compartments. The terminal branches intertwine among the network of collagen strands. Close to their point of termination, the branches may become devoid of Schwann cell processes and are covered only by the basal lamina.50,61 The terminals themselves take the form of spirals and sprays that are richly supplied with mitochondria and form intimate associations with the collagen strands.50,51,61 Innervation of tendon organs by myelinated afferent axons of only 2 to 3 ␮m in diameter has also been observed,68 but their function is not known.

DISTRIBUTION Tendon organs are located at the myotendinous junction of skeletal muscles, more commonly at the insertion rather than origin of the muscle. In pennate muscles, the organs

are commonly found near the point of nerve entry in a central zone of the muscle.59,65 Occasionally, tendon organs may be found entirely within the tendon,4 but the responsiveness of such receptors is as yet unknown. Tendon organs may also be found in close association with other receptors, such as paciniform corpuscles and, in particular, muscle spindles, where they are referred to as dyads.46 Tendon organs have been found to be present in almost all skeletal muscles examined, including limb, neck, masticatory, and respiratory muscles.36 It is interesting to note that some particularly specialized muscles, such as the cat tenuissimus,52 the human orbicularis oris (J. Scott, unpublished data), and sometimes the small hand muscles such as the lumbricals,14 appear to be devoid of tendon organs. Detailed counts of tendon organs in individual muscles are rare and, unfortunately, have not been carried out systematically. In almost all instances, the number of tendon organs is lower than that of spindles and shows considerable variation: In a survey of 10 cat hind limb muscles, the proportion of tendon organs to spindles ranged from 0.33 to 0.94.36 Interestingly, the limited information available from primate and human intrinsic hand muscles indicates relatively low ratios. Thus for the abductor pollicis, in which there are tendon organ counts of seven and eight in the bonnet monkey and human, respectively, as well as for the lumbricals, which likewise have counts between zero to one and five, respectively14 (J. Scott, unpublished data), the ratios of tendon organs to spindles are below 0.3. Because tendon organs are specifically responsive to the tension generated by muscle contraction, a more meaningful way of expressing receptor complement is as a ratio of the number of motor units in the muscle. This has been done for a number of cat hind limb muscles, giving a range of 1.9 to 9.0 motor units per tendon organ.36 Given that there are typically between 10 and 25 muscle fibers inserting into each tendon organ (see above), this implies that, taken together, the population of tendon organs of the muscles should be sufficient to monitor the contractile activity of all the motor units.

PHYSIOLOGIC RESPONSES The tendon organ lies in series with the muscle fibers that insert into it. Therefore, it can be stretched by stretch or contraction of the muscle. However, the organ has a high threshold to the passive tensions generated by wholemuscle stretch and responds with low rates of firing; indeed, a significant proportion of organs have been shown to be insensitive to stretch within the physiologic range.33 By contrast, tendon organs are sensitive to active contractions of the whole muscle. Furthermore, Houk and Henneman32 first demonstrated that a given tendon organ shows a high sensitivity to the tension generated by active contraction of a small number of individual motor units and

154

Function of the Peripheral Nervous System

fails to respond to the contractions of other units. The range of numbers of response-generating motor units (4 to 15) was found to be approximately the same as that of muscle fibers inserting into the organs. Therefore, each tendon organ serves to sample the contractile tension generated by a small population of motor units,5 each of which contributes one or two muscle fibers to the group that inserts into the body of the organ.22,67 The force threshold for the response to the contraction of single motor units may be as low as a few milligrams.6 Measurements of the sensitivities of isolated cat tendon organ–muscle fiber preparations have given values of 335 impulses/s/g.17 The majority of the motor units within a muscle, when stimulated separately, do not activate an individual tendon organ because none of the muscle fibers comprising these units inserts into the organ. Such in-parallel units have the potential to “unload” the tendon organ32 by reducing the relative tension being channeled through the organ. In theory, activation of an in-parallel motor unit could reduce the response of the organ to an in-series unit. The physiologic significance of such unloading effects is unclear.36 Stuart et al.69 showed that the firing rate of a tendon organ responding to the contraction of an in-series motor unit could be reduced when an in-parallel unit was contracting at the same time.7,31,69 There is also evidence of reciprocal unloading effects elicited by two in-series motor units.31 Such effects would imply that the response of the tendon organ was dependent on the combination of motor units

contracting concurrently and not directly on the force being applied to the organ itself. In support of this, it has been shown that the firing rate of a tendon organ to simultaneous contraction of all the in-series motor units is often higher than that during contraction of the muscle as a whole.5 This phenomenon of unloading has been cited as one of the possible factors underlying the apparent nonlinearity of tendon organ responses17 (see below). However, Gregory and co-workers25–27 reported that the reductions in firing rate were only transient during tetanic contractions. They attributed this to the initial lowering of tension by the in-parallel fibers, which was then recovered as the in-series fibers shortened further.

Responses to Motor Unit Contractions Tendon organs are typically silent in the passive muscle except at extreme muscle lengths. In response to tetanic contraction of an in-series motor unit, the organ will respond with a train of action potentials. There is an initial, high-frequency burst during the rising phase of the tension, followed by adaptation to a plateau rate of firing during the steady-state tension (Fig. 7–2A). The tendon organ response may therefore be characterized by a static sensitivity, the sensitivity shown to steady-state forces, and a dynamic sensitivity, the sensitivity shown to changing force levels. Complications inherent in measuring these 80

Static discharge (impulses s–1)

Impulses s–1

60

40

20

10 g

FR FR FF

60

FR S 40

FF FF FF

20

0 0

20

30

40

50

Plateau tension (g)

0.5 s

A

10

B FIGURE 7–2 Responses of a tendon organ afferent to muscle contraction. A, Instantaneous frequency discharges of a Ib afferent in response to isometric contraction of an activating FR motor unit in the peroneus tertius muscle of the cat. The motor unit was stimulated at 60, 80, 100, and 150/s, and the responses have been overlaid. B, Response relationships between the afferent firing frequency of a tendon organ and the plateau tensions developed by the individual motor units that inserted into the organ. (From Petit, J., Davies, P., and Scott, J. J. A.: Static sensitivity of tendon organs to tetanic contraction of in-series motor units in feline peroneus tertius muscle. J. Physiol. [Lond.] 481:177, 1994, with permission of Cambridge University Press.)

155

The Golgi Tendon Organ

sensitivities and the extent of the apparent nonlinearities have given rise to significant debate and also to difficulties in interpreting the role of tendon organs in the feedback control of muscle contraction.36 Motor units in skeletal muscles vary in terms of their muscle fiber complement and their contractile properties. In terms of the contractile properties, they are classified into three groups, FF, FR, and S.10 The FF units are fast contracting and fast fatiguable, whereas the S units are slow contracting and very fatigue resistant, with the FR units being intermediate in terms of both fatigability and contraction speed. These units also differ in terms of the numbers of muscle fibers, with the FF units having the most and the S units the fewest muscle fibers. Analyses of the motor-unit composition of the muscle fibers inserting into tendon organs have revealed a heterogeneity that reflects that of the muscle as a whole.26,27,37,38,58 Thus a detailed analysis of the motor unit input to a tendon organ in the peroneus tertius muscle of the cat showed it was activated by 11 motor units: 5 FFs, 4 FRs, and 2 Ss.55

Static Sensitivity If tendon organs are to provide the central nervous system with meaningful information regarding muscle force, then it might be expected that there should be consistent relationships between the discharge frequency and the contractile force generated by activating motor units. Tetanic contractions, applied under isometric conditions, have frequently been employed to investigate the sensitivity of tendon organs, because the plateau tension provides a constant stimulus that elicits a steady discharge from individual organs. Under such conditions, a number of studies have shown that there was no consistent relationship between force and discharge frequency; thus a small, type S motor unit might elicit a higher discharge than a powerful FF unit.5,26,32,37,38,58 Furthermore, activation of a motor unit that inserts into more than one tendon organ may give rise to different frequencies of discharge from the different organs,38 and combined stimulation of pairs of activating motor units does not appear to elicit a summed response.27 Fukami17 showed that, for isolated tendon organs, there is a linear relationship between the static discharge and the contractile tension generated by a single, inserting muscle fiber. This has also been shown to hold true for the relationship between discharge frequency and the plateau tension generated by individual activating motor units55 (Fig. 7–2B), although there were differences between the sensitivities to different motor units, the sensitivity being greatest for type S and least for FF motor units. When motor units were activated in combination, linear relationships were again recorded for the discharge frequency against tension for each specific combination. The slopes of the relationships were shallower than for the single units and,

with the exception of combinations of type S units, tended to be similar.55 Measurements of the static sensitivities of a group of tendon organs to the tetanic contractions of single and combined motor units showed that the sensitivities were high for the lowest tensions and dropped rapidly to a steady level of approximately 100 impulse/s/N that was independent of the size and number of the activating units.55 The relationship between sensitivity and contractile tension may be attributable to the unloading effect of the in-parallel muscle fibers of the activating motor units in combination with the compliance of the tendon; this unloading effect will be largest for the largest motor units, and therefore, when they are stimulated singly, they may show lower sensitivities than the smaller units. Beyond a certain level of tension, for large units or as motor units are combined, the unloading resulting from the compliance of the tendon will approach a constant level because the stiffness of the tendon will increase as it is stretched until it reaches a high, constant level.57 As a consequence, the sensitivity of the organs will also approach a low, constant level as reported by Petit et al.55 A further consequence of this unloading effect is that, when motor units are added together, for the first few units the order in which that addition takes place will affect the sensitivity of the organ to the applied tension.26 Although tetanic, isometric contractions represent a simple method for investigating the steady-state input to a tendon organ, they do not constitute a representative physiologic condition, particularly because the motor units are typically stimulated at rates well above those occurring naturally. An alternative approach has been to investigate isotonic contractions, with ongoing muscle shortening or lengthening. These have been generated with groups of motor units stimulated asynchronously, so that a smooth tension profile can be elicited while stimulating the individual motor units at physiologic rates. The responses of the tendon organs were again found to be proportional to the active tension, with the same sensitivity as for the isometric contractions, indicating that the tendon organ has the capacity to provide an accurate signal of the contractile force being applied by the contracting muscle under conditions that more closely reflect natural movements.56

Dynamic Sensitivity During the onset of a contraction, the discharge rate of a tendon organ is higher than that for the equivalent steadystate tension and is related to the rate of rise of the tension. This dynamic sensitivity has particularly been investigated under isometric conditions using rates of motor unit stimulation that generate unfused contractions. A common observation under such conditions is that the tendon organ will respond with an action potential for each rising phase

156

Function of the Peripheral Nervous System

in the tension oscillation,30,39 with such 1:1 responses being referred to as “driving.” One consequence of this dynamic sensitivity and of the phenomenon of driving is that, when a motor unit is stimulated at rates below the fusion frequency for the contraction (up to approximately 40 impulses/s for FF units), the rate of tendon organ discharge will reflect the oscillation frequency rather than the underlying absolute tension. For example, increasing the stimulation frequency from 10 to 20 impulses/s will double the frequency of the force oscillations and therefore double the rate of discharge, even though the mean force level changes by a differing amount.39 The effects of 1:1 driving can also be observed when pairs of motor units are activated asynchronously.30 Because these stimulation rates reflect the main physiologic range of motor units firing, there is the implication that the dynamic sensitivity could dominate the feedback to the central nervous system and therefore, particularly at low levels of contractile force, the organs would be signaling force variation rather than the absolute force levels.36 The nature of the dynamic response appears complex. In a passive muscle at physiologic length, the tendon organs are usually silent. During the rising phase of a tetanic contraction, the sensitivity changes, with an initial high peak in discharge frequency that then declines, even though the absolute tension is still rising, until the steadystate levels are achieved. Early models fitted the response with two exponential components, the first with a fast and the second with a slower time constant.31 Likewise, during the unfused contractions, the initial peak at the onset of the first contraction is higher than the maintained driving response.39 Further analysis of the dynamic response has shown that, for each motor unit–tendon organ combination, the initial instantaneous frequency for the first two impulses (assuming that the tendon organ was not firing before the contraction onset) is linearly related to the rate of rise of the tension, the slope of the relationship being steepest for the type S units. This initial dynamic sensitivity is absent if the tendon organ is already discharging before the onset of the contraction, or if the activating motor unit is repetitively stimulated at relatively short intervals. When stretches are applied to isolated tendon organs, the receptor potential shows an initial overshoot with respect to the applied tension and the discharge rate shows a further overshoot that is dependent on the rate of rise of the receptor potential.18 This latter overshoot was attributed to the accommodative properties of the pacemaker site such that the threshold for action potential generation decreases with increasing rate of rise of the receptor potential13,18 and is absent if the tendon organ is already discharging before the stimulus onset. The second component of the dynamic response is exponentially related to the rate of rise of the tension, the sensitivity being independent of motor unit size.13

Recordings during Natural Movements Although the studies described above have enabled development of significant understanding of the patterns of response to a range of stimulation paradigms, the ideal would be to be able to analyze the responses of tendon organs during natural movements. Unfortunately, this is technically very demanding and, compared with recordings from muscle spindles, there is a paucity of recordings from tendon organs during natural movements. The majority of the available data have come from two sources, recordings made from cats using chronic implants and microneurographic recordings from human subjects. The recordings from cats have shown that the Ib afferents discharge in relation to the electromyographic (EMG) activity within the muscles and therefore to the occurrence of contractile activity.2,43 During stepping movements, Ib afferents from ankle extensors showed maximal discharge during the loading periods of the stance phase.2 During smooth increases in contractile force, the tendon organ discharges were commonly marked by stepwise increments in firing rate, which were attributed to the increments in the mechanical input to the tendon organ resulting from motor unit recruitment.2 Similar steps in discharge were also observed during the generation of crossed-extensor reflexes.12 The development of the microneurographic technique has enabled recordings to be made from single nerve fibers in human subjects. Although the numbers of tendon organs from which recordings have been made is relatively small, the findings have confirmed those from animal studies in that the organs respond to muscle contraction, as evidenced by EMG activity, and show a low sensitivity to muscle stretch.1,15 Although the firing rates appear to be relatively low, the organs display a high sensitivity to variations in the contractile activity, as demonstrated by their responsiveness to the pulsatile motor activity that has been recorded during slow finger movements.70 During progressively developing contractions, recordings from human Golgi tendon organs have also shown steps in the firing rate that coincided with increments in the contractile force and may be attributable to motor unit recruitment.15 Interestingly, some but not all of these steps showed an initial dynamic peak even though the Ib afferent was already discharging (see below).

TRANSDUCTION As has already been described, the adequate stimulus for tendon organs is contraction of the inserting muscle fibers that stretches the collagen strands. This stretching deforms the unmyelinated terminals of the Ib afferent axon that are wrapped around the collagen strands, giving rise to a receptor potential resulting from conductance

157

The Golgi Tendon Organ

changes in the deformed terminals. Early models of tendon organ function proposed that, as the tendon organ is stretched, so the axon terminals are compressed between the collagen strands, giving rise to the mechanical distortion and hence the change in membrane potential.8 Although no direct measurements of the membrane properties and ion channel openings have been made, indirect measurements of the receptor potential in isolated tendon organs indicate that stretch itself is the mechanical stimulus.18,71 Under steady-state conditions, the receptor potentials were found to be linearly related to the stretch applied to the organ and the sensitivity was also related to the compliance of the organ, indicating that stretch itself was the activating stimulus.18 During sinusoidal stretching, wherein the dynamic sensitivity becomes apparent, the relationship between the receptor potential and the applied stretch was nonlinear for all but the smallest amplitude stretches, and appeared dependent both on the mechanics of the organ and on the ionic processes within the terminals.18 The Ib afferent axon branches several times before giving rise to the terminals that are associated with the collagen strands. The first- and second-order branches of the axon are commonly myelinated,61 indicating a capacity to conduct action potentials. This raises the question as to where the site of action potential generation is located, because the terminal node of each branch could act as a pacemaker site. At one extreme, the receptor potential could spread to the most proximal branch point, so that a single train of action potentials is generated representing the summation of the receptor potentials produced in the separate branches. Alternatively, each branch could conduct action potential trains that then interact at one or more branch points. Where there are multiple pacemaker sites, an action potential arising in one branch and arriving at the branching node can be conducted up the parent branch but can also be conducted antidromically down the other daughter branch. In this case, the action potential will collide with any action potentials traveling orthodromically in the second daughter branch. As a consequence, the pattern of discharge recorded in the main ascending branch will depend on the relative firing rates in the daughter branches and will range from probabilistic mixing, giving rise to a degree of summation, to complete occlusion of one branch’s activity. Detailed analysis of the discharge patterns of the Ia afferents from muscle spindles has shown evidence of a range of competitive pacemaker interactions resulting in varying levels of occlusion.3 When two motor units are stimulated together, the frequency of action potential discharge in the Ib afferent axon to the combined stimulation is less than the arithmetic sum of the responses to the units stimulated separately. This has been proposed as evidence of pacemaker interactions,27 although it could also be due to unloading effects.17,55

Further evidence for the operation of multiple pacemaker sites derives from the occurrence of cross-adaptation. As described above, the dynamic response shows an initial peak that is dependent on the accommodative properties of the pacemaker site.13,18 Thus, if an activating motor unit is stimulated twice, tetanically, the tendon organ’s response to the second contraction lacks an initial dynamic peak as a result of adaptation of the pacemaker threshold. However, if the organ is subjected to successive contractions of two different activating motor units, the dynamic peak may or may not be present.26,27 If the two motor units stretch collagen strands that receive innervation from different branches of the Ib afferent axon, then each of these branches may have its own pacemaker site, such that adaptation of one site by the first contraction does not influence the sensitivity of the second site, which displays a full dynamic response to the contraction of the second unit. Conversely, if the two collagen strands are innervated by terminals arising from the same main axon’s branch, then the initial dynamic peak will be absent from the response to the second unit. These findings have been successfully replicated with computer models that incorporate multiple pacemaker sites.24

RECOVERY FROM NERVE LESIONS Two main types of nerve injury have been investigated experimentally, nerve crush (axonotmesis) and nerve transection (neurotmesis) with subsequent repair. These differ significantly in their outcomes. Following nerve crush, there is usually a high proportion of axons that regenerate and successfully reinnervate their original target sites. After nerve section, however, many axons fail to regenerate successfully and so a significant proportion of target sites remains uninnervated. Where reinnervation does take place, it is highly unlikely to be effected by the original axon; indeed, the reinnervation may be by an axon that previously innervated a different type of end organ. The tendon organ is relatively resistant to denervation atrophy, and after nerve crush injury there is usually successful restoration of the Ib afferent ending, although it is reduced in extent.35,62 During the early stages of recovery, the Ib afferents frequently fail to maintain discharge throughout the plateau phase of a tetanic contraction, with a reduced static sensitivity, but these gradually recover to normal levels.63 The two phases of the dynamic response appear to be affected differentially; the second phase, which shows an exponential relationship to the rate of rise of the tension (see above) is the same as normal. The initial sensitivity, which is additionally dependent on the accommodative properties of the pacemaker site, is reduced during the early stages but also recovers toward normal.63 Following nerve transection, the pattern of recovery is much poorer, with many tendon organs remaining

158

Function of the Peripheral Nervous System

uninnervated. Where reinnervation does occur, the endings are frequently abnormal in appearance.62 There are three confounding factors that affect the outcome of the reinnervation: first, the effects of the denervationreinnervation process itself on the properties of the afferent axon, as, for example, the effects observed following nerve crush injury; second, the effects of reinnervation by inappropriate axons; and third, the effects consequent on the reorganization of the motor units. Following nerve transection, there is frequently grouping of the muscle fibers comprising individual motor units.41 This can significantly alter the mechanical input to a tendon organ; for example, all the muscle fibers inserting into an organ could be derived from a single motor unit, or a group of muscle fibers comprising a reinnervated motor unit could lie immediately adjacent to an organ and thereby generate significant unloading.62 Recordings from reinnervated tendon organs display a range of abnormalities, including phasic-only responses and on-off responses, in which the afferent axon discharges one or two action potentials at the onset and relaxation of the contraction but not during the maintained phase, and off-only responses. It is noteworthy that tendon organs showing abnormal responses to some motor units could also display a normal pattern of response to the contractions of other units, suggesting that the abnormalities in the mechanical input do play the major part in determining the patterns of response following reinnervation.63,64 What is unclear at present is what impact the significant disruptions in proprioceptive feedback, resulting from axonal mislocation and abnormal patterns of response, have on the control of movement.

REFLEX ACTIONS Investigation of the spinal interactions mediated by Ib afferents has been complicated by the problems associated with trying to separate the effects of Ib actions from those of other afferents, particularly the group Ia muscle spindle afferents.36 Many early studies used the technique of graded electrical stimulation of the muscle nerves of the cat hind limb, in decerebrate preparations, to investigate the reflex actions of the muscle afferents. These studies showed that, when the stimulation intensity was progressively elevated above the threshold for Ia afferents, and therefore above the threshold for eliciting the monosynaptic stretch reflex, there was evidence of a disynaptic inhibitory action on the homonymous and some synergic muscles and of facilitation of antagonists. This disynaptic inhibition represented the classic autogenetic inhibition that was proposed to be mediated by the group Ib afferents of the tendon organ.21,42 More detailed examination showed that the Ib interactions to the spinal cord do not mirror the Ia actions, in the form of an inverse

myotatic reflex, but rather that the distribution of inhibitory actions is more widespread and the facilitation of antagonists more variable. In particular, the inhibition arising from flexor muscles tends to be weaker than that from extensors.29 The picture of a simple autogenetic inhibitory reflex, mediated via an interneuron—the “Ib inhibitory interneuron,” which provides a negative feedback regulation of contractile force—has undergone considerable revision over recent years. Detailed investigations of the inputs to the Ib inhibitory interneuron in the lumbar segments of the cat spinal cord have shown that this interneuron receives a wide range of converging inputs, which include segmental, excitatory inputs from Ia, joint, and cutaneous afferents as well as Ib afferent inputs. There are also descending excitatory inputs from the corticospinal and rubrospinal tracts and inhibitory inputs from the dorsal and noradrenergic reticulospinal systems.28,34,36 Because of the pattern of convergence on the Ib inhibitory interneuron, which includes Ia and Ib inputs, and the target projections, this pathway may be more appropriately termed nonreciprocal group I inhibition.28 In addition to the widespread capacity for modulating the Ib reflex via the converging inputs, studies of walking movements have demonstrated a more dramatic pattern of Ib action. During the stance phase of locomotion in the cat, when the extensor muscles are active, the Ib afferent feedback contributes to a positive feedback loop that reinforces the extensor muscle action during the weight-bearing phase of walking54; the Ib action shows a change in sign from inhibition to excitation. Furthermore, the Ib feedback from the ankle extensors has an inhibitory action on the flexors, preventing the initiation of the swing phase until the extensor muscles are unloaded as weight is transferred to the contralateral foot.53,54 During a lengthening contraction of the extensors in the stance phase, the positive feedback from the Ib loop can act in concert with the negative feedback Ia-mediated stretch reflex to compensate for the loading.66

CENTRAL PROJECTIONS AND PROPRIOCEPTION The afferent feedback from tendon organs is relayed to both the cerebellum and the cerebral cortex, and therefore might be predicted to have input to central processing of both motor control and proprioception. The cerebellum receives tendon organ inputs from the hind limb via the dorsal spinocerebellar tract (DSCT) and the ventral spinocerebellar tract, which enter the cerebellum via the inferior peduncle and terminate as mossy fibers in the anterior lobe and paramedian lobule. Forelimb input is relayed via the cuneocerebellar and rostral spinocerebellar tracts. Climbing fiber input is routed via the inferior olive.16

159

The Golgi Tendon Organ

The best studied of the cerebellar projections relaying tendon organ feedback is that of the DSCT in the cat, originating in Clarke’s column. The DSCT receives inputs from a variety of muscle, joint, and cutaneous afferents, but the different inputs are segregated to different groups of DSCT cells.44 The input to the DSCT neurons takes a variety of forms. In a study of the effects of weak muscle contraction, just over 50% of the DSCT neurons tested were contraction sensitive, indicating a wide distribution of the muscle-afferent inputs. However, the effects were strongly contrasting, being separable into four main types. The first group of DSCT cells showed excitation that persisted throughout the contraction and was attributed to monosynaptic inputs from Ib afferents. The second group showed declining excitation during the contraction, probably as a consequence of presynaptic inhibition. The third and largest group showed declining inhibition that could be attributed to disynaptic inputs via the Ib inhibitory interneuron, and the fourth group displayed mixed effects.73 Clearly, the cerebellum receives a complex variety of information about ongoing muscle contractions, but how this information is processed within the cerebellum in terms of regulation of movement and muscle tone is yet to be elucidated. In addition to the cerebellar projection, the Ib afferent feedback is also relayed to area 3a, in the somatosensory cortex, of the cerebral cortex. This projection in the cat is routed via collaterals of the dorsal spinocerebellar axons that synapse in nucleus Z and are relayed through the thalamus.48,49 The question therefore arises as to whether the Ib feedback contributes to conscious proprioception and generates some form of sense of force or effort or whether these arise from corollary discharges. Stimulation of populations of low-threshold muscle afferents from the hand gives rise to a short-latency cortical potential and illusions of muscle stretch.19 At the singlefiber level such perceptions are very rare, though in a sample of three Ib afferents, microstimulation of one of the afferents did elicit a sensation of movement.45 Roland and Ladegaard-Pedersen60 demonstrated that a “sense of tension” exists that persists during skin anesthesia and during partial muscle paralysis, which disturbs the relationship between motor output and achieved contractile force. Such a sense of tension should therefore originate from the muscle receptors. However, a number of other studies suggest that the preferential source of information is derived centrally; for example, the perception that the weight of an object increases with developing fatigue40,47 and the sense of force is unaffected by changing the initial muscle length for maximal isometric contractions, thereby altering the actual achieved force.11 Eccentric exercise also results in errors of force estimation,9 but these changes are not matched by changes in tendon organ sensitivity,23 again implying that the dominant information contributing to the sense of force is that derived from the centrally generated

motor commands rather than the peripheral feedback from the tendon organs.

ACKNOWLEDGMENTS My thanks are extended to Dr. Revers Donga for his helpful criticisms of the manuscript.

REFERENCES 1. Al-Falahe, N. A., Nagaoka, M., and Vallbo, A. B.: Response profiles of human muscle afferents during active finger movements. Brain 113:325, 1990. 2. Appenteng, K., and Prochazka, A.: Tendon organ firing during active muscle lengthening in awake, normally behaving cats. J. Physiol. (Lond.) 353:81, 1984. 3. Banks, R. W., Hulliger, M., Scheepstra, K. A., and Otten, E.: Pacemaker activity in a sensory ending with multiple encoding sites: the cat muscle spindle primary ending. J. Physiol. (Lond.) 498:177, 1997. 4. Barker, D.: The morphology of muscle receptors. In Hunt, C. C. (ed.): Handbook of Sensory Physiology, Vol. 3. Berlin, Springer, p. 1, 1985. 5. Binder, M. D.: Further evidence that the Golgi tendon organ monitors the activity of a discrete set of motor units within a muscle. Exp. Brain Res. 43:186, 1981. 6. Binder, M. D., Kroin, J. S., Moore, G. P. and Stuart, D. G.: The response of Golgi tendon organs to single motor unit contractions. J. Physiol. (Lond.) 271:337, 1977. 7. Binder, M. D., and Osborn, C. E.: Interactions between motor units and Golgi tendon organs in the tibialis posterior muscle of the cat. J. Physiol. (Lond.) 364:199, 1985. 8. Bridgman, C. F.: The structure of tendon organs in the cat: a proposed mechanism for responding to muscle tension. Anat. Rec. 162:209, 1968. 9. Brockett, C., Warren, N., Gregory, J. E., et al.: A comparison of the effects of eccentric exercise on force and position sense at the human elbow joint. Brain Res. 771:251, 1997. 10. Burke, R. E., Levine, D. W., Tsairis, P., and Zajac, F. E.: Physiological types and histochemical profiles in motor units of the cat gastrocnemius. J. Physiol. (Lond.) 234:723, 1973. 11. Caferelli, E., and Bigland-Ritchie, B.: Sensation of static force in muscles of different length. Exp. Neurol. 65:511, 1979. 12. Crago, P. E., Houk, J. C., and Rymer, W. Z.: Sampling of total muscle force by tendon organs. J. Neurophysiol. 47:1069, 1982. 13. Davies, P., Petit, J., and Scott, J. J. A.: The dynamic response of Golgi tendon organs to tetanic contraction of in-series motor units. Brain Res. 690:82, 1995. 14. Devanandan, M. S., Ghosh, S., and John, K. T.: A quantitative study of muscle spindles and tendon organs in some intrinsic muscles of the hand in the bonnet monkey (Macaca radiata). Anat. Rec. 207:263, 1983. 15. Edin, B., and Vallbo, A. B.: Muscle afferent responses to isometric contractions and relaxations in humans. J. Neurophysiol. 63:1307, 1990.

160

Function of the Peripheral Nervous System

16. Ekerot, C. F., Larson, B., and Oscarsson, O.: Information carried by the spinocerebellar paths. Prog. Brain Res. 50:79, 1979. 17. Fukami, Y.: Responses of isolated Golgi tendon organs of the cat to muscle contraction and electrical stimulation. J. Physiol. (Lond.) 318:429, 1981. 18. Fukami, Y., and Wilkinson, R. S.: Responses of isolated Golgi tendon organs of the cat. J. Physiol. (Lond.) 265:673, 1977. 19. Gandevia, S. C., Burke, D., and McKeon, B.: The projection of muscle afferents from the hand to cerebral cortex in man. Brain 107:1, 1984. 20. Goldfinger, M. D., and Fukami, Y.: Distribution, density and size of muscle receptors in cat tail dorsolateral muscles. J. Anat. 135:371, 1982. 21. Granit, R.: Reflexes of self-regulation of muscle contraction and autogenetic inhibition. J. Neurophysiol. 13:351, 1950. 22. Gregory, J. E.: Relations between identified tendon organs and motor units in the medial gastrocnemius muscle of the cat. Exp. Brain Res. 81:602, 1990. 23. Gregory, J. E., Brockett, C. L., Morgan, D .L., et al.: Effect of eccentric muscle contractions on Golgi tendon organ responses to passive and active tension in the cat. J. Physiol. (Lond.) 538:209, 2002. 24. Gregory, J. E., Morgan, D. L., and Proske, U.: Site of impulse initiation in tendon organs of cat soleus muscle. J. Neurophysiol. 54:1383, 1985. 25. Gregory, J. E., Morgan, D. L., and Proske, U.: The discharge of cat tendon organs during unloading contractions. Exp. Brain Res. 61:222, 1986. 26. Gregory, J. E., and Proske, U.: The responses of Golgi tendon organs to stimulation of different combinations of motor units. J. Physiol. (Lond.) 295: 251, 1979. 27. Gregory, J. E., and Proske, U.: Motor unit contractions initiating impulses in a tendon organ in the cat. J. Physiol. (Lond.) 313:251, 1981. 28. Harrison, P. J., and Jankowska, E.: Sources of input to interneurones mediating group I non-reciprocal inhibition of motoneurones in the cat. J. Physiol. (Lond.) 361:403, 1985. 29. Harrison, P. J., Jankowska, E., and Johannison, T.: Shared reflex pathways of group I afferents in different hindlimb muscles. J. Physiol. (Lond.) 338:113, 1983. 30. Horcholle-Bossavit, G., Jami, L., Petit, J., et al.: Activation of tendon organs by asynchronous contractions of motor units in cat leg muscles. Neurosci. Lett. 103:44, 1989. 31. Horcholle-Bossavit, G., Jami, L., Petit, J., et al.: Unloading of tendon organ discharges by in-series motor units in cat peroneal muscles. J. Physiol. (Lond.) 408:185, 1989. 32. Houk, J. C., and Henneman, E.: Responses of Golgi tendon organs to active contractions of the soleus muscle of the cat. J. Neurophysiol. 30:466, 1967. 33. Houk, J. C., Singer, J. J., and Henneman, E.: Adequate stimulus for tendon organs with observations on mechanics of ankle joint. J. Neurophysiol. 34:1051, 1971. 34. Hultborn, H.: State-dependent modulation of sensory feedback. J. Physiol. (Lond.) 533:5, 2001. 35. Ip, M. C., Vrbova, G., and Westbury, D.: The sensory reinnervation of hind-limb muscles after denervation and deefferentation. Neuroscience 2:423, 1977.

36. Jami, L.: Golgi tendon organs in mammalian skeletal muscle: functional properties and central actions. Physiol. Rev. 72:623, 1992. 37. Jami, L., and Petit, J.: Frequency of tendon organ discharges elicited by the contraction of motor units in cat leg muscles. J. Physiol. (Lond.) 261:633, 1976. 38. Jami, L., and Petit, J.: Heterogeneity of motor units activating single Golgi tendon organs in cat leg muscles. Exp. Brain Res. 24:485, 1976. 39. Jami, L., Petit, J., Proske, U., and Zytnicki, D.: Responses of tendon organs to unfused contractions of single motor units. J. Neurophysiol. 53:32, 1985. 40. Jones, L. A., and Hunter, I. W.: Perceived force in fatiguing isometric contractions. Percept. Psychophys. 33:369, 1983. 41. Karpati, G., and Engel, W. K.: Type grouping in skeletal muscles after experimental reinnervation. Neurology 18: 447, 1968. 42. Laporte, Y., and Lloyd, D. P. C.: Nature and significance of the reflex connections established by large afferent fibers of muscular origin. Am. J. Physiol. 169:609, 1952. 43. Loeb, G. E.: Somatosensory unit input to the spinal cord during normal walking. Can. J. Physiol. Pharmacol. 59:627, 1981. 44. Lundberg, A., and Oscarsson, O.: Functional organisation of the dorsal spinocerebellar tract in the cat. IV. Synaptic connections of afferents from Golgi tendon organs and muscle spindles. Acta Physiol. Scand. 38:53, 1956. 45. Macefield, G., Gandevia, S. C., and Burke, D.: Perceptual responses to microstimulation of single afferents innervating joints, muscles and skin of the human hand. J. Physiol. (Lond.) 429:113, 1990. 46. Marchand, E. R., Bridgman, C. F., Shumpert, E., and Eldred, E.: Association of tendon organs with spindles in muscles of the cat’s leg. Anat. Rec. 169:23, 1971. 47. McCloskey, D. I., Ebeling, P., and Goodwin, G. M.: Estimation of weights and tensions and apparent involvement of a “sense of effort”. Exp. Neurol. 42:220, 1974. 48. Mcintyre, A. K., Proske, U., and Rawson, J. A.: Central projection of afferent information from tendon organs in the cat. J. Physiol. (Lond.) 354:395, 1984. 49. Mcintyre, A. K., Proske, U., and Rawson, J. A.: Pathway to the cerebral cortex for impulses from tendon organs in the cat’s hindlimb. J. Physiol. (Lond.) 369:115, 1985. 50. Merrillees, N. C. R.: Some observations on the fine structure of a Golgi tendon organ of a rat. In Barker, D. (ed.): Symposium on Muscle Receptors. Hong Kong, Hong Kong University Press, p. 199, 1962. 51. Nitatori, T.: The fine structure of human Golgi tendon organs as studied by three-dimensional reconstruction. J. Neurocytol. 17:27, 1988. 52. Palmer, J. M., and Sitwell, D. L.: Analysis of innervation of tenuissimus muscle of the cat. Anat. Rec. 130:434, 1958. 53. Pearson, K. G.: Could enhanced reflex function contribute to improving locomotion after spinal cord repair? J. Physiol. (Lond.) 533:75, 2001. 54. Pearson, K. G., and Collins, D. F.: Reversal of the influence of group Ib afferents from plantaris on activity in medial gastrocnemius muscle during locomotor activity. J. Neurophysiol. 70:1009, 1993.

The Golgi Tendon Organ 55. Petit, J., Davies, P., and Scott, J. J. A.: Static sensitivity of tendon organs to tetanic contraction of in-series motor units in feline peroneus tertius muscle. J. Physiol. (Lond.) 481:177, 1994. 56. Petit, J., Scott, J. J. A., and Reynolds, K. J.: Tendon organ sensitivity to steady-state isotonic contraction of in-series motor units in feline peroneus tertius muscle. J. Physiol. (Lond.) 500:227, 1997. 57. Proske, U., and Morgan, D. L.: Tendon stiffness: methods of measurement and significance for the control of movement. A review. J. Biomech. 20:75, 1987. 58. Reinking, R. M., Stephens, J. A., and Stuart, D. G.: The tendon organs of cat medial gastrocnemius: significance of motor unit type and size for the activation of Ib afferents. J. Physiol. (Lond.) 250:491, 1975. 59. Richmond, F. J. R., and Stuart, D. G.: Distribution of sensory receptors in the flexor carpi radialis muscle of the cat. J. Morphol. 183:1, 1985. 60. Roland, P. E., and Ladegaard-Pedersen, H.: A quantitative analysis of sensations of tension and of kinaesthesia in man. Brain 100:671, 1977. 61. Schoultz, T. W., and Swett, J. E.: The fine structure of the Golgi tendon organ. J. Neurocytol. 1:1, 1972. 62. Scott, J. J. A.: The functional recovery of muscle proprioceptors after peripheral nerve lesions. J. Peripher. Nerv. Syst. 1:19, 1996. 63. Scott, J. J. A., Davies, P., and Petit, J.: The static sensitivity of tendon organs during recovery from nerve injury. Brain Res. 697:225, 1995.

161

64. Scott, J. J. A., Davies, P., and Petit, J.: The dynamic response of feline Golgi tendon organs during recovery from nerve injury. Neurosci. Lett. 207:179, 1996. 65. Scott, J. J. A., and Young, H.: The number and distribution of muscle spindles and tendon organs in the peroneal muscles of the cat. J. Anat. 151:143, 1987. 66. Sinkjaer, T., Andersen, J. B., Ladouceur, M., et al.: Major role for sensory feedback in soleus EMG activity in the stance phase of walking in man. J. Physiol. (Lond.) 523:817, 2000. 67. Spielmann, J. M., and Stauffer, E. K.: Morphological observations of motor units connected in-series to Golgi tendon organs. J. Neurophysiol. 55:147, 1986. 68. Stacey, M. J.: Free endings in skeletal muscle of the cat. J. Anat. 105:231, 1969. 69. Stuart, D. G., Mosher, C. C., Gerlach, R. L., and Reinking, R. M.: Mechanical arrangement and transducing properties of Golgi tendon organs. Exp. Brain Res. 14:274, 1972. 70. Wessberg, J., and Vallbo, A. B.: Coding of pulsatile motor output by human muscle afferents during slow finger movements. J. Physiol. (Lond.) 485:271, 1995. 71. Wilkinson, R. S., and Fukami, Y.: Responses of isolated Golgi tendon organs of cat to sinusoidal stretch. J. Neurophysiol. 49:976, 1983. 72. Zelena, J., and Soukup, T.: The in-series and in-parallel components in rat hindlimb tendon organs. Neuroscience 9:899, 1983. 73. Zytnicki, D., Lafleur, J., Kouchtir, N., and Perrier, J.-F.: Heterogeneity of contraction-induced effects in neurons of the cat dorsal spinocerebellar tract. J. Physiol. (Lond.) 487:761, 1995.

8 The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons SALLY N. LAWSON

Overview Large Light (Neurofilament-Rich) and Small Dark (Neurofilament-Poor) DRG Neurons Soma Size, Conduction Velocity, and Neurofilament Soma Size in Relation to Sensory Properties Afferent Receptive Properties Terms Used to Define Properties of Afferents Functional Classes of Cutaneous Afferent Neurons Cutaneous, Muscle, and Visceral Afferent Receptive Properties Electrical Membrane Properties of DRG Neurons Nociceptors versus LTMs

Non-nociceptive Neurons Changes in Electrophysiology in Chronic Pain Models Chemical Phenotype of DRG Neurons Neurotrophic Factors and Their Receptors Inflammatory Mediators and Tissue Damage Other Receptors Oligosaccharides Expressed by DRG Neurons Transmitters, Neuromodulators, Peptides, and Their Receptors Excitatory Amino Acid Transmitters Neuropeptides Cytochemistry of Cutaneous, Visceral, and Muscle Afferents

OVERVIEW This chapter attempts to relate the morphology, electrophysiology, and cytochemistry of dorsal root ganglion (DRG) neurons to each other and to their sensory receptive properties. The classification of primary afferent neurons in relation to their afferent receptive properties was covered in depth in the previous edition182 and is merely summarized here. Primary afferent neurons are pseudo-unipolar. That is, in their mature form they have only one process leaving the soma. This process is called the initial segment. It branches at its T junction into a peripherally and a centrally projecting process. The initial segment is short in neurons with unmyelinated fibers, and much longer in neurons with large myelinated fibers. The sensory neuron or unit includes the cell body, or soma, plus fibers and terminals. The cell bodies (somata) of primary somatosensory neurons are segmentally arranged in DRGs and in cranial root ganglia. Somatosensory neurons in DRGs may be nocicep-

Membrane Receptors and Sensory Transduction Thermoreceptive Molecules Mechanoreceptive Molecules Chemoreceptive Molecules Acid-Sensing Molecules Ion Currents and Channels Na Currents and Channels K Currents and Channels Hyperpolarization-Activated Currents and Channels Ca2 Currents and Channels Chemical Phenotype of Nociceptive and LTM Neurons Summary

tive or non-nociceptive; may have unmyelinated (C) fibers or myelinated (A) fibers; may respond to mechanical, thermal, and/or chemical stimuli; and may project to skin, muscle, and blood vessels of the trunk and limbs or to visceral organs in the thorax and abdomen. Glial cells in DRGs and cranial root ganglia are of neural crest origin, as are all DRG and jugular ganglion neurons and some neurons in the trigeminal ganglion (Vth cranial nerve) neurons. The rest of the trigeminal ganglion neurons and all neurons in the geniculate (VII), petrosal (IX), and nodose (X) ganglia are of placodal origin.174 This chapter focuses on DRG neurons.

Large Light (Neurofilament-Rich) and Small Dark (Neurofilament-Poor) DRG Neurons Studies of the morphology and histology of DRG neurons from the late 19th century mostly found two main subgroups of neurons (see Lawson164 and Rambourg et al.254), known as LL (large light) or A-type neurons and SD (small dark) or B-type neurons. LL and SD neurons have overlapping 163

164

Function of the Peripheral Nervous System

distributions of cell size (cross-sectional area through the middle of the cell); in mouse and rat the sizes of each are normally distributed,164,171 leading to bimodal distributions of cell size. LL neurons have a large mean size but include neurons of all sizes; SD neurons are small. Bimodal distribu-

A

B

C

tions of DRG neuronal size occur in many species, including mouse, rat, cat,243,248 and human (S. Border and S. N. Lawson, unpublished data) (Fig. 8–1) as well as chick. These populations are not evenly distributed within the length of the DRG. In chick DRGs the larger neurons tend to be more

D

E

F

FIGURE 8–1 Size distributions of dorsal root ganglion (DRG) neuronal profiles that contain nuclei in frozen sections (A–D) and wax sections (E and F). Sections are stained to show neurofilament (NF) immunoreactivity (filled grey histograms show NF-rich neurons and hatched histograms show NF-poor neurons) in all but E, which shows horseradish peroxidase (HRP)–injected cells with no immunoreactivity. Different antibodies against NF were used. Antibodies were RT97 against a highly phosphorylated epitope on the 200-kDa NF subunit (NF200) (A, D, and F), against a phosphorylation-independent epitope (B), and against the 68-kDa subunit (C). A small-diameter population of NF-poor neurons and an NF-rich population with a much larger mean diameter are apparent in all species, but cell sizes depend on species. A, Adult rat DRG, plotted on same scale as cat ganglia (B) and as human (C) ganglia from a human postmortem L5 DRG of a 70-year-old man. D, Same data as in A, plotted for comparison with E and F and showing size ranges (small, medium, and large), which differ slightly as a result of differences in processing in D, E, and F. E and F, Sizes of dye-injected neurons with C-, A-, or A-fiber neurons. After intracellular recording, dye was injected into the soma, enabling cell size to be plotted in relation to conduction velocity. E, Cells injected with HRP, with no immunocytochemistry. F, Fluorescent dye injection followed by immunocytochemistry for NF200. NF-rich neurons have A fibers and NF-poor neurons have C fibers. (B: Data from Perry, M. J., and Lawson, S. N.: Neurofilament in feline primary afferent neurons: a quantitative immunocytochemical study. Brain Res. 607:307, 1993. C: Courtesy of Sandra Border. E: Data replotted from Harper, A. A., and Lawson, S. N.: Electrical properties of rat dorsal root ganglion neurons with different peripheral conduction velocities. J. Physiol. [Lond.] 359:47, 1985. F: Data replotted from Lawson, S. N., and Waddell, P. J.: Soma neurofilament immunoreactivity is related to cell size and fibre conduction velocity in rat primary sensory neurons. J. Physiol. [Lond.] 435:41, 1991.)

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

peripheral (ventrolateral), and the later developing smaller neurons are mesiodorsal.118 Similarly, in rodent (unpublished data) and cat308 DRGs, the earlier developing167,168 LL neurons tend to be concentrated more peripherally. LL DRG neurons have clear Nissl substance (patches of organelles that pick up conventional stains) separated by unstained (with conventional stains) swathes of neurofilament (NF), whereas SD neurons have little NF and more even, darker staining.80,168,281,344 Thus LL neuronal somata are NF rich and SD somata are NF poor.243,245 Immunostaining for NF (e.g., Fig. 8–1A to D; see also Fig. 8–4 later) can distinguish between these populations in several species (e.g., rat, mouse, cat, human) as originally demonstrated in rat.171

Soma Size, Conduction Velocity, and Neurofilament Neuronal cell body size and conduction velocity are loosely related. In adult mice, rats, and cats, DRG neuronal somata with C fibers are small, those with A fibers are small to medium sized, and those with A-fiber neurons mainly medium or large121,175,347 (Fig. 8–1E and F) if these words are defined as follows: small  within size range of SD cell distribution, large  right-hand side of LL cell distribution, and medium  between small and large ranges. With this definition, small neurons have mainly C and fewer A fibers, medium-sized neurons may have C, mainly A, or fewer A fibers, and large neurons A (few) and mainly A fibers. In rat, the NF-poor/SD and NF-rich/LL lumbar DRG neurons have C fibers and A fibers, respectively (see Fig. 8–1F),173 and NF content of the soma is thus frequently used to distinguish between these groups. For deductions to be made about conduction velocity from cell size alone, the size ranges of small, medium, and large neurons need to be established for the neurons under study, because actual sizes vary with species (Fig. 8–1A to C) (also see Gerke and Plenderleith100), postnatal age,164,168 histologic processing, and whether studied in sections, or in vitro, where some neurons may flatten. An additional complication is that cell size (both LL and SD) reduces by about 30% after spinal nerve axotomy, coupled with a loss of DRG neurons. In one study in rat the neuronal loss was 20% after 1 week increasing to 35% by 45 days327; in mouse DRGs 24% and 54% neuronal loss, respectively, was reported 7 and 28 days after midsciatic section.283 Furthermore, the loss after nerve transection appeared to be limited to cutaneous, not muscle, afferents.128 Thus comparisons between axotomized and nonaxotomized neurons should take into account possible changes in cell size; selective cell loss; and altered small, medium, and large cell size boundaries.

Soma Size in Relation to Sensory Properties It is frequently stated that small neurons in DRGs are nociceptive. It is the case that a higher proportion of small neurons than of large neurons is nociceptive. However, some

165

A-fiber nociceptive neurons are large, especially those with A fibers (see later), and some small neurons are C- or A-fiber low-threshold mechanoreceptors (LTMs) (unpublished observations). Thus, although many small neurons are likely to be nociceptive, and many large neurons are likely to be LTMs, there are important exceptions to this.

AFFERENT RECEPTIVE PROPERTIES Although DRG neurons are all afferent neurons, only those that give rise to a conscious sensation are, strictly speaking, sensory. Thus visceral afferent neurons that may not evoke conscious sensation are usually called afferent, not sensory. Primary afferent neurons have classically been described in terms of their modality or adequate stimulus, threshold, conduction velocity, adaptation properties, and sensory receptive properties. These terms are expanded below.

Terms Used to Define Properties of Afferents Modality. The term sensory modality of a primary afferent neuron is used to describe the type of sensation asociated with stimulation of that neuron. Classically, the modalities of human sensation included sight, hearing, taste, smell, and touch. For the somatosensory system as well as touch, it is useful to include temperature (heat, cold), and pain. More precise definition can be achieved by referring to the quality of the evoked sensation, such as “gentle pressure” to further define touch, or “burning pain.” Adequate Stimulus. Adequate stimulus has been defined as the “type of naturally occurring event that most effectively excites” the afferent unit.185 Thus a low-intensity mechanical stimulus is the adequate (most effective) stimulus for a LTM fiber. Threshold. The threshold stimulus is the weakest that can activate the fiber. Low-threshold units respond to lowintensity (usually nondamaging and nonpainful) stimuli. Thus cutaneous LTMs usually give rise to sensations such as touch, vibration or pressure, position, or movement, whereas low-threshold thermoreceptors give rise to sensations of warmth or cooling of the skin. High-intensity stimuli are likely to cause damage to the tissues. Nociceptors and Pain. Stimuli that are damaging or potentially damaging to the tissues are said to be noxious, and primary afferent neurons that respond only, or preferentially, to such stimuli are called nociceptors. To quote Light and Perl, “Nociceptors are defined as primary afferent units that uniquely signal stimuli intense enough to cause damage to the tissue.”185 Nociceptors usually have high thresholds for activation, but these thresholds may be lower where tissues are particularly vulnerable, such as the surface of the cornea. Activation of nociceptors usually

166

Function of the Peripheral Nervous System

results in pain once the information they transmit reaches the higher centers of the brain in a conscious animal. Some fibers may be selectively activated by noxious mechanical, noxious thermal, or noxious chemical stimuli, others by more than one of these types of stimulus. The last group are called polymodal nociceptors.185 In contrast to cutaneous afferents, in which there are clearly separate groups of nociceptive and LTM neurons,185 there is debate about whether afferents projecting to some visceral organs (heart, colon, bladder) may be involved in “organ regulation and non-painful sensations as well as pain,”133 whereas innervation of other organs clearly involves separate lowthreshold and nociceptive afferents (see later). Transduction. Stimuli are transduced into electrical signals at receptive regions of the fiber. The type(s) of natural stimulus (mechanical, thermal, or chemical) that can evoke electrical signals in a given neuron depend on the type(s) of protein receptor molecules (an ion channel or associated with an ion channel) in its receptive membrane. For instance, activation of a Na channel or closure of a K channel may cause a depolarizing receptor potential; if this reaches threshold, one or more action potentials will propagate along the fiber. Transduction usually occurs at the receptive membrane at the peripheral terminals of the neuronal fiber but may occur in receptive non-neuronal cells, such as the Merkel cell, whose possible role in mechanical transduction has been the subject of much debate.117 Possible molecular transduction mechanisms are discussed in a later section. Conduction Velocity. Afferents are divided according to their fiber conduction velocity into C, A, and A fibers (cutaneous afferents) and group IV, III, II, and I fibers (muscle afferents).185 The conduction velocity ranges of these groups of fibers can be determined with compound action potential recordings,185 but need to be determined for each experimental situation. This is because the velocity ranges differ with age, species, temperature, the nerve used, and the proximodistal position along the nerve since some fibers slow toward the periphery.122 For example, the upper limits of 2.5 m/s for C fibers and 30 m/s for A fibers, appropriate for adult cat peripheral nerve,185 are sometimes applied to rat studies, although values determined with compound action potentials for young adult rats are much lower (1 m/s for C fibers and 1 to 6 or 1 to 8 m/s for A fibers).73,85 Adaptation. A stimulus of constant strength applied to the sensory receptive field results in firing of action potentials at a rate that decreases with time. This decrease is known as adaptation, and may be rapid or slow. The terms rapidly adapting (RA) and slowly adapting (SA) were first coined in relation to mechanosensitive primary afferent neurons. RA mechanoreceptors tend to fire only during the initial application or removal of a constant mechanical stimulus. That is, they have a phasic or dynamic pattern of firing and

detect change or movement, and can signal such information to the central nervous system (CNS). The most rapidly adapting mechanoreceptors are those innervating pacinian corpuscles, which signal information about acceleration or rapid vibration.185 Conversely, SA mechanoreceptors continue to fire (with a static or phasic firing pattern) in response to a sustained mechanical stimulus such as stretch or light pressure, and signal information about the amount of pressure or degree of stretch to the CNS. There is some evidence that intrinsic membrane properties contribute to adaptation and that these may be similar at the receptive membrane and in the soma.120

Functional Classes of Cutaneous Afferent Receptive Neurons Non-nociceptive neurons include low-threshold thermoreceptor (cool and warm receptor) neurons, low-threshold chemoreceptive afferents such as taste receptors, and a variety of visceral afferent chemoreceptors (Table 8–1). However, the largest group of non-nociceptive afferent DRG neurons projecting to skin and skeletal muscle are LTMs. Cutaneous LTMs conduct in A-, A-, and C-fiber conduction velocity ranges, and muscle spindle afferent LTMs conduct in group II and group I ranges. Fibers of cutaneous and muscle LTMs241 make up a large proportion of fast-conducting myelinated fibers (A/types I and II). The different types of unit are summarized in Table 8–1, and their properties are covered in greater depth by Light and Perl.185 An overview follows. Low-Threshold Mechanoreceptors (LTMs) Cutaneous A␣␤-Fiber LTM Units. Cutaneous A-fiber LTM units show a variety of adaptation rates and firing patterns. SA units are said to be of two types. The SA type I (SA I) have an irregular discharge and are thought to make contact with Merkel cells in skin (see review by Halata et al.117). The SA II fire more regularly127 and are classically described as contacting Ruffini endings in dermis, although there are few such endings in human glabrous skin.234 RA units in skin are important in the detection of stimuli of changing intensity and movement of mechanical stimuli. They include fibers (glabrous RA units) in glabrous skin that are thought to innervate Meissner’s corpuscles and fibers in hairy skin that innervate guard hair (G hair) follicles (known as G hair units). Afferent units that are best activated by light mechanical stimuli to the skin between hairs or to movement of groups of hairs are known as Field units. G hair and Field units are subdivided according to their sensitivity to rate of movement such that the more slowly adapting G hair and Field units are called type II and the more rapidly adapting units activated only by faster rates of movement are called type I.127 Thus GII hair units respond to sustained deflection of a hair, whereas GI hair units respond best to sudden deflection, or flicking, of a hair. This leads to the following

Phasic/ Slower than A/

Tonic/Slow

Slow movement

Pressure, stretch, position

Phasic/ Rapid and variable

A/ A/

SAI (Merkels cells); irregular firing SAII (? Ruffini-end organs), regular firing, responds to stretch and cooling

C

C mechanoreceptor (C LTM), very slow (not fast) movement over skin, cooling

A/

A/

A

Glabrous RA (Meissners corpuscle)/ movement G Hair follicle/ field unit, variable adaptation to hair deflection or skin indentation

Movement

A/

CV Groups

D hair/extremely sensitive to slow hair movement, also stretch, cooling

Pacinian corpuscles in dermis

Vibration, Phasic/ Very acceleration, rapid tap, jerk

Phasic/ Rapid

Fiber Type/ Responds to

Type of Stimulus

Firing/ Adaptation Rate

Skin (Exteroceptors)

Table 8–1. Classes of Cutaneous Afferent Receptive Neurons*

?

Sustained pressure

? emotional touch

? insect detection

Movement

Movement

Tap, buzz, vibration

Sensation

Pressure on muscle, Contraction sensitive units

Muscle spindle primary and secondary endings, Golgi tendon organs. Ruffini corpuscles in joint capsules & ligaments

?

Primary endings respond with dynamic firing component to rate of change of stretch (muscle spindle intrafusal nuclear bag fibers)

Paciniform endings in ligaments, joint capsules and facia

Fiber Type/ Responds to

III, IV

Ia, Ib, II

Ia

A/

CV Group

Muscle/Tendon/Joint (Proprioceptors)

A, A/ or unknown

A

A

A/

CV Group

Table continued on following page

Organ capsules, walls of GI tract & lung, baroreceptors, renal pelvis

Lung SA stretch receptors

Paciniform corpuscles in mesentery (in cat) & large arteries Lung RA stretch receptors

Fiber Type/ Responds to

Viscera (Enteroceptors)

?

Firing/ Adaptation Rate

Probably Slow

?

Unknown

Polymodal nociceptor

Fibers responsive to heat or cold

C/A

Mostly C also A

C & A

A A/ C

C/A

CV Groups

?

burning pain/ache

Heat/cold

Sharp pricking or dull pain

Warm, cool

Sensation

Unknown

Polymodal nociceptor

?

Joint capsules, ligaments skeletal muscles: compression or stretch.

Thermoreceptor (cool or warm) in muscle Free nerve endings

Fiber Type/ Responds to

IV

III & IV

III & IV

III/IV

IV

CV Group

Muscle/Tendon/Joint (Proprioceptors)

Unknown

Polymodal nociceptor

?

Bladder pressure 60 cm H2O

Free nerve endings

?

e.g., Low pH, glucose, various chemoreceptors, ion concentrations, ischaemia, irritant receptors in lung

Fiber Type/ Responds to

C

C A

C & A

C & A

Vary

CV Group

Viscera (Enteroceptors)

CV  conduction velocity; GI  gastrointestinal tract; HTMs  high-threshold mechanoreceptors; LTMs  low-threshold mechanoreceptors; RA  rapidly adapting; SA  slowly adapting. *See text and references 34, 52, 116, 117, 133, 176, 185, 205, 242, 260, and 273. This table is not intended to be a complete listing but to provide an overview of some of the types of units innervating different types of tissue. The most complete section relates to skin.

Silent nociceptors Normally insensitive become sensitized by, (e.g., inflammatory mediators, heat, skin damage)

Polymodal nociceptors One or more of: noxious Depends on mechanical, noxious stimulus heat, cold or chemical type (e.g., pH, ischemia, inflammatory mediators)

Noxious thermal heat or cold Heat 44˚ C or cold 17˚ C

HTMs mainly A, some A in dermis and in epidermis.

Noxious mechanical

Some are slow

Free nerve endings

Cutaneous, presumed free nerve endings

Taste and olfaction are excluded as DRG neurons are not involved

Fiber Type/ Responds to

NOCICEPTORS

THERMO-RECEPTORS Warm or cool Variable

CHEMORECEPTORS Internal chemical environment

Type of Stimulus

Skin (Exteroceptors)

Table 8–1. Classes of Cutaneous Afferent Receptive Neurons*—Continued

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

groupings: GI hair, GII hair, Field I, and Field II.127 Thus, although Field and G hair units are commonly thought of as rapidly adapting, in reality their rates of adaptation vary. The most rapidly adapting units are those innervating pacinian corpuscles in the dermis112,235 and the mesentery of some species such as cat.224 These units respond best to rapidly changing stimuli such as a tap to the skin, or rapid vibration. Cutaneous A-Fiber LTMs. Most cutaneous A-fiber LTMs projecting to hairy skin are called “D hair” units after the very fine down hairs. D hair units are extremely sensitive to slow movement of hair, but also respond both to stretch and frequently to cooling of the skin.185 Their sensitivity to small movements of fine hairs makes detection of insects in the fur one possible role. Cutaneous C-Fiber LTMs. Cutaneous C-fiber LTMs (C mechanoreceptors) respond to extremely slow (1 mm/s) movement across the skin.185 They have been described in a number of species (rat, cat, rabbit, guinea pig). Recent evidence in the human shows the presence of these units323 and indicates that information from C mechanoreceptors reaches the insula cortex and is associated with a pleasant emotional response (limbic touch).231 Thermoreceptors Cooling and warming receptors have C-fiber conduction velocities in subprimates but may have C or A fibers in primates.185 Cooling receptors fire at higher frequencies as temperature is decreased from about 30° C, and human psychophysical experiments show that sensations of cold/cool result from temperatures below 30° C, ache below 17.5° C, and cold pain below 14° C.65,185 Peripheral warm receptors fire in the range from 30° to 46° C.65,185 Many C-fiber polymodal nociceptors have a thermal threshold of about 43° C, and some A-mechanoheat units have a higher threshold (53° C). Nociceptors C-Fiber Nociceptors. C-fiber nociceptive neurons are the most numerous of the nociceptive fibers and probably have receptive fields in all types of tissues. The majority of those with superficial receptive fields respond to both noxious mechanical and noxious heat stimuli and can thus be classed as C-mechanoheat or C-polymodal nociceptors.185 A smaller proportion of units with superficial receptive fields are C-fiber high-threshold mechanoreceptor (HTM) units, because they respond to noxious mechanical stimuli but not to a single application of heat or cold; a few respond to noxious heat only. A-Fiber Nociceptors. Cutaneous A-fiber nociceptors have A or A fibers and respond to noxious mechanical stimuli or noxious mechanical and noxious heat stimuli (mechanoheat units). Superficial receptive fields are punctate. For a discussion of firing properties and sensitization of these

169

neurons, see Light and Perl.185 A-fiber nociceptors in primate (monkey) have different responses to noxious heat. Type I units (two-thirds A, one-third A, conducting up to 60 m/s) have a higher heat threshold with a later onset, long-lasting response to heat and a lower mechanical threshold, whereas type II units (exclusively A, 30 m/s) have a lower heat threshold with a prompt, early, and short response to heat and a higher mechanical threshold.322 First pain sensation to heat may be served by the type II A fibers, and first pain to noxious mechanical stimuli by type I (A and A fibers).322 Although sometimes overlooked in recent literature, some A-fiber nociceptors conduct in the A range. For a review of the evidence for A nociceptors, see Djouhri et al.73 Briefly, conduction velocities of A-fiber nociceptors range from 5 to 60 m/s in cat33,35 and 5 to 60 m/s in primate (monkey),322 clearly including units with A fibers because in both species the upper end of the A range is less than 30 m/s. The earliest descriptions of properties of cutaneous nociceptive A fibers in cat33,35 and monkey240 included nociceptors with A fibers. Some of these A units were insensitive mechanoreceptors or moderate-pressure receptors, with lower thresholds than many A nociceptors, but because both types fired more enthusiastically to noxious than to nonnoxious pressure, their adequate stimulus is in the noxious range and they are classed as nociceptors. Burgess and Perl33 found in cat that the percentages of nociceptive A fibers that conducted above the A range (30 m/s) were 7% of highthreshold nociceptors, 35% of insensitive mechanoreceptors, and 100% of moderate-pressure receptors. Intracellular recordings in DRG neurons show that about 20% to 65% of A-fiber nociceptive neurons have A fibers in cat,157 rat263 (X. Fang et al., unpublished data), and guinea pig.71 For more details on these percentages, see Djouhri et al.73 Nonetheless, at the time of writing, it is frequently assumed that nociceptors conduct only in the C- and A-fiber range, perhaps because (1) the distribution of conduction velocities of A-fiber nociceptors tends to peak in the upper part of the A range33,35 and (2) the A wave of the compound action potential is dominated by LTM fibers. (For further discussion, see Lawson166 and Djouhri et al.73) Unresponsive DRG Neurons: Possible “Silent” Nociceptors. Inexcitable afferent units are hard to identify with certainty because (1) the appropriate search stimulus may not have been used and (2) in fiber recordings sympathetic C fibers have to be excluded. Nonetheless, apparently inexcitable units have been described in several species projecting not only to joint tissues but also to skin and visceral organs.71,133,273 These units may respond after repeated stimuli or after acute inflammation of the tissues206 to stimuli that did not previously excite them. Such units may include very-high-threshold nociceptive neurons and/or units that are either sensitized or excited by substances released during tissue damage, including inflammatory mediators (see later). The latter may act as detectors of inflammation/tissue damage.133,185

170

Function of the Peripheral Nervous System

ELECTRICAL MEMBRANE PROPERTIES OF DRG NEURONS

Cutaneous, Muscle, and Visceral Afferent Receptive Properties Table 8–1 provides a summary of some of the different types of primary afferent neurons, linking stimulus type, conduction velocity range, adaptation properties, and associated sensations where known. Information about sensations is largely dependent on microneurography in the conscious human. Microneurography employs electrode penetration of a nerve to record from, and stimulate, single afferent fibers, enabling correlation between fiber properties and reported sensations. Unlike cutaneous and muscle afferents, visceral afferents do not all fall into clear nociceptive and non-nociceptive categories. For example, firing of afferent fibers encodes distention pressures from nonpainful through to painful in both bladder and colon, whereas other organs, such as gallbladder and ureter, are innervated only by high-threshold afferents that are mechano- and chemosensitive and are clearly visceral nociceptors; still others (esophagus) are innervated by both high- and low-threshold afferents.133 Only a few examples of visceral afferents are listed in the table. For a detailed list and extensive bibliography of the earlier literature, see the tables in Perl and Burgess.242

A

C HTM

D

Nociceptors versus LTMs There are clear differences in soma membrane properties between nociceptive and LTM neurons. Nociceptive neurons have longer action potential and longer afterhyperpolarization durations in cat, rat, and guinea pig A-fiber neurons71,156,263 (Fig. 8–2), and in rat and guinea pig in C-fiber neurons71 (X. Fang et al., unpublished data). To summarize, in each of the conduction velocity ranges (C, A, and A), nociceptive neurons have significantly longer afterhyperpolarization durations,71,156,263 and larger action

Aδ HTM

B

B

C LTM

E

ns

1

B

I Aα/β LTM G H

H Aα/β LTM MS J

Aα/β LTM T HAIR

ns

ns

G Aα/β HTM

* *

10

C Unresp 20 mV 2 ms

***

ns

Aδ Unresp

*

ns

F C

Aδ LTM

AP durn at base (ms)

A

AHP 80% duration (ms)

A

Although it is hard to record intracellularly from small afferent fibers and terminals, it is possible to make such recordings from the soma. Because the properties of DRG neurons in vitro may be altered by axotomy and removal of the normal cellular and molecular environments, this section is limited (unless stated otherwise) to intracellular recordings from DRG neurons in vivo.

100

10

NOC LTM NOC LTM NOC LTM Aδ Aα/β C ns ns ns ns ns ns

*

* *

* * *

1 NOC LTM NOC LTM NOC LTM Aδ Aα/β C

FIGURE 8–2 A, Examples of action potential configurations from intracellular recordings of guinea pig DRG neurons, with identified sensory properties and conduction velocities (indicated for each trace). HTM  high threshold mechanoreceptor; LTM  low threshold mechanoreceptor; Unresp  unresponsive; G  G hair receptor; MS  muscle spindle afferent; T hair  Tylotrich hair. B, The relationship of the median action potential (AP) duration and afterhyperpolarization (AHP) duration to 80% recovery. NOC  nociceptors; ns  median not significantly different; *  P  0.05; ***  P  .001 Mann Whitney U Test. (Data from Lawson, S. N.: Phenotype and function of somatic primary afferent nociceptive neurons with C-, Adelta or Aalpha/beta-fibres. J. Exp. Physiol. 87:239, 2002.)

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

potential overshoots,75 and in the C- and A-fiber ranges they have significantly longer action potential durations with both longer rise times and fall times71,156,263 (for a summary, see Lawson166 and Fig. 8–2). As well as being related to sensory properties, action potential duration is also inversely related to the conduction velocity of the fiber (see Fig. 8–2B).106,122,166,330 Extracellular recordings of afferent C fibers in rat and pig97 also show a trend of longer action potential durations in nociceptive compared with LTM fibers. These slower membrane kinetics of nociceptive neurons in both soma and fiber, plus their higher activation thresholds, may have an important role in normally limiting the amount of information reaching the CNS via the nociceptive primary afferent pathway.

Non-nociceptive Neurons There are differences in intracellularly recorded membrane properties between different subtypes of LTM neurons. In guinea pig71 and rat (X. Fang et al., unpublished data) DRGs, the action potential duration of muscle spindle Afiber afferents have faster kinetics (shorter durations) than those of cutaneous A-fiber LTM units (see Fig. 8–2A).

Changes in Electrophysiology in Chronic Pain Models The two main types of animal model of chronic pain states are the induction of inflammation in the tissues and damage to the peripheral nerve. During Chronic Peripheral Inflammation Inflammatory mediators cause rapid (minutes to hours) changes in sensitivity of peripheral terminals in inflamed tissues. However, clinically more important is chronic pain that may last for days, weeks, and months. Animal models of chronic inflammatory pain include intradermal injections of complete Freund’s adjuvant (CFA) or carrageenan in one hind limb, which induce both inflammation and enhanced withdrawal from noxious stimuli, both immediately and persisting for days/weeks. Changes in electrophysiologic properties of nociceptive but not LTM DRG neurons in vivo in guinea pig take more than 2 days to be complete after intradermal CFA injection.72 They include increased incidence (threefold) of C- and A-fiber nociceptors that fire spontaneously.72 These changes are accompanied by decreased withdrawal latencies to noxious heat. The slow time course of the changes is compatible with altered protein expression of ion channels or associated molecules. Indeed, Na channel protein overall is upregulated by the first day after CFA injection, with a longer lasting increase in small than large DRG neurons.108 Many inflammation-induced electrophysiologic changes in guinea pig including the increased spontaneous firing, are nerve growth factor (NGF) dependent,72 as is the inflammatory hyperalgesia.338

171

Following Peripheral Nerve Injury In humans, peripheral nerve injury leads to neuropathic pain involving hyperalgesia, tactile and cold allodynia, and spontaneous pain. A number of animal models of neuropathic pain involve peripheral nerve injury. These include chronic constriction injury (CCI)20 and tight ligation or cut of the L5 (or L5 and L6) spinal nerves (Chung model).153 It is not known which neuronal subtypes make the greatest contribution to neuropathic pain: whether it is those with C, A, or A fibers; those originally with nociceptive or LTM properties; those with intact or severed peripheral fibers; those with fibers affected by inflammation at, or distal to, the injury site; and/or those that are spontaneously active or not. Some evidence in human patients indicates a correlation between spontaneous fiber activity and spontaneous pain.110,225,346 In animal models, peripheral nerve injury (axotomy or CCI) leads to a brief burst of firing, and after 2 to 3 days a gradual increase in spontaneous activity.109,188,197 The activity may originate at the lesion site or in the DRG.137,302 Many afferents develop “an ectopic repetitive firing capability” in response to stimuli in the region of the injury,197,331 which may be blocked by Na channel blockers (tetrodotoxin [TTX], lidocaine) and enhanced by veratridine (prolongs open time of Na channels), implicating Na currents.197 One study found no spontaneous activity in damaged cutaneous afferents.209 Axotomy leads in vivo to broader action potentials in A-fiber DRG neurons than usual, coupled with increased spontaneous firing in A- but not C-fiber neurons.154,295 Intact neurons with C- (but not A-) fibers that run alongside axotomized fibers also show spontaneous firing which may prove to be important in generation of spontaneous pain.340 It may therefore be that a combination of spontaneous activity from axotomized A-fiber and adjacent intact C-fiber neurons both contribute to the generation of neuropathic pain.

CHEMICAL PHENOTYPE OF DRG NEURONS There is a vast literature covering the expression of peptides, proteins, enzymes, receptors, signaling proteins, and carbohydrate groups in DRG neurons (for previous reviews, see Lawson165 and Millan211). Some aspects of the chemical phenotype are expressed transiently during development, and/ or show altered expression following damage to the nerves or other tissues, including inflammation. The focus here is on molecules that are selectively localized in clear subgroups (e.g., certain sizes) of DRG neurons, because this may be indicative of a link with selective functions of a neuronal type. (For a detailed review of co-localization as known in 1992, see Lawson.165) Percentages of DRG neurons reported to be positive for a particular marker are dependent on differences in in situ probes, antibodies, immunocytochemical techniques, expression at different ages, rostro-caudal level of DRGs, species or strains within

SMALL

MEDIUM

LARGE

C FIBER / NF POOR Aδ-FIBER NF-RICH Aδ /β-FIBER NF-RICH NOCICEPTORS LOW THRESHOLD MECHANORECEPTORS Peptide Peptide Peptide Peptide REC REC REC REC REC ION CH ION CH ION CH ION CH ION CH / REC Cytokine Cytokine Enzyme CBH / LECTIN Peptide Peptide REC REC ION CH / REC Peptide REC REC REC REC ION CH ION CH / REC ENZYME REC ION CH ION CH ION CH / REC REC REC REC REC REC REC ENZ ENZ ION CH ION CH ION CH ION CH Cytokine CBH CBH ION CH / REC REC ION CH

ET1 Galanin Nociceptin Somatostatin GLUR1,GLUR5,AMPA,Kainate B1, B2, H1 EGF, FGF2 NK1, SSTR2a P2X3 BK, KCa current Ca++ current L-type, N-type Kv1.4 Nav1.9, Naβ3 TRPA1,TPRM8,TRPV1,TRPV3 TGFα, TNFα LIF retrograde transport COX1 IB4 binding / PNA ET1, PACAP, VIP Substance P / NKA GAL2, ETA receptor GFRα3 TRPV4 CGRP GFRα1,GFRα2, RET MOR/DOR/KOR ORL1 TrkA, p75 Nav1.7, Nav1.8 P2X2/3 nNOS TrkB Ca++ current T-Type Kir TRPV2 CB1 cannabinoid receptor GAL1 receptor GLUR2/3 GM1 receptor / Cholera toxin B binding P2Y1 TrkC / NT3 uptake Calbindin Carbonic anhydrase α3 Na+ pump subunit Naβ1.1,Naβ2.1 Kv1.1, 1.2, Kvβ2.1 H current, HCN1, HCN2 IL1β GSAII / LCA / PSA lectin binding SSEA3/4 BNC1/ASIC2 GluR2/3, gp130, CNTF-R, IL6-R, KCNQ2, KCNQ3, KCNQ5

FIGURE 8–3 A visual indication of the approximate reported sizes of neurons labeled by different markers or expressing different currents (details in text). The length of line relates to neuronal relative sizes (small, medium, and large) as used in Figure 8–1, from published size distributions or from terminology used by authors. Thus, for this figure, the word “small” is translated into a line the width of the SD/C-fiber/NF-poor population. This figure does not provide any indication about proportions of neurons expressing these phenotypes. The order used relates to (a) cell size, (b) the type of molecule (see column 1), and (c) within these groups is alphabetical. Abbreviations are found within the text, but those in column 1 are as follows: REC  receptor; ION CH  ion channel; ION CH/REC  ion channel gated chemically, thermally, or mechanically; CBH/lectin  lectin binding to carbohydrate groups or use of antibodies against particular oligosaccharide groups (see text).

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

species, and, importantly, the level of staining required for a neuron to be judged positive, which is rarely explained. These factors together largely explain differences in reported percentages. Although much extremely useful knowledge has been acquired from DRG neurons in tissue culture, cultured neurons may show altered phenotypes compared with their in vivo counterparts as a result of physical damage such as axotomy close to the soma145,274 and loss of in vivo influences such as trophic factors. Percentage values for human DRGs labeled by immunocytochemistry or in situ hybridization may be influenced by long postmortem times, resulting in degradation of molecules under examination. Because of the complexity and amount of information, a number of figures and tables summarize the detail provided in the text, as follows. An indication of the sizes of neurons reported to have different chemical phenotypes is provided in Figure 8–3, and examples of staining for NF, isolectin B4 (IB4), calcitonin gene–related peptide (CGRP), and tyrosine kinase A (TrkA) are shown in Figure 8–4. A cartoon of sources of endogenous substances that may influence DRG neuronal function is given in Figure 8–5A and a summary of altered expression of substances by DRG neurons or non-

173

neuronal cells after axotomy and inflammation is provided in Figure 8–5B. Tables relating chemical phenotype to aspects of afferent neuron function are included in subsequent sections for trophic factor receptors (Table 8–2); properties of neurons projecting to skin, muscle, and viscera (Table 8–3); and putative molecular transduction candidates (Table 8–4).

Neurotrophic Factors and Their Receptors Neurotrophic factors are factors that support growth, repair, or survival of neurons. These may be target derived, that is, produced by the tissues to which the neuron projects, or produced by cells, such as Schwann/satellite cells within the nerve or DRG. These factors tend to bind to specific membrane receptors, and may be internalized and transported to the neuronal soma, where they can affect intracellular events including gene expression. Their direct influence is therefore most powerful in neurons that express appropriate high-affinity receptors, although they may also bind with lower affinity to other receptors or have indirect effects via influence on other cell types. Thus understanding of (1) the effects of these factors and (2) which neurons express the receptors for

FIGURE 8–4 Interference contrast images of 8 m sections of adult rat L5 DRG showing immunostaining using ABC immunocytochemistry (Vector) with a monoclonal antibody RT97 (courtesy of John Wood. Available from Chemicon) raised against the 200-kDa neurofilament subunit (NF200); and with polyclonal antibodies raised against IB4 (Vector Laboratories), CGRP (Peninsula Laboratories) and TrkA (courtesy of L. Reichardt). Note that the NF200 staining pattern appears reciprocal to that of IB4 binding in terms of neuronal size.

174

Function of the Peripheral Nervous System

RELEASE OF SUBSTANCES FROM CELLS OTHER THAN DRG NEURONS Glut SP Somatostatin Noradrenalin Opioids BDNF FGF2 5HT NT4 GDNF CNTF IL1β PGs

ET1 IL1β TGFα EGF

NGF TNFα GDNF CNTF LIF ET1

TNFα ET1

NGF NT3 NT4 GDNF Neurturin FGFs

SKIN

BDNF Glut SP CGRP Somatostatin TNFα ?Nociceptin/orphanin

NT3 GDNF FGF2 ATP Lactate

SKEL MUSC

BDNF Glutamate SP CGRP Somatostatin

RELEASE OF SUBSTANCES FROM NORMAL DRG NEURONS A FIGURE 8–5 Endogenous substance that influence DRG neurons. A, Some of the endogenous chemicals that are known to have a direct influence (usually via specific membrane receptors) or an indirect influence on DRG neurons. Structures represented are spinal cord, a DRG neuron, satellite cells, Schwann cells, leukocytes and mast cells, skin, and skeletal muscle. Top, Chemicals secreted and released by cells other than DRG neurons, which may influence the DRG neuron at the central or peripheral terminals or at the cell body in the DRG. Bottom, Substances released by DRG neurons that may influence the same or other DRG neurons again through the central or peripheral terminals or at the cell body in the DRG. (For details, see text.) See Color Plate Figure continued on opposite page

these factors is essential for a full understanding of their roles. Additional complexity is added by altered expression or transport of these factors and/or of their receptors on the DRG neurons during development and following tissue damage such as inflammation and peripheral nerve injury (see Fig. 8–5B). The resulting plasticity in the properties of DRG neurons is extensive and makes this a very complex field. Trk Receptors for Neurotrophins The neurotrophin family includes NGF, brain-derived neurotrophic factor (BDNF), and neurotrophin-3 (NT-3) and neurotrophin-4/5 (NT-4/5). High-affinity receptors for neurotrophins are members of the tyrosine kinase (Trk) family of neurotrophin receptors and include TrkA, TrkB, and TrkC; p75 is a low-affinity receptor for these neurotrophins. TrkA, TrkB, TrkC, and p75 are all expressed on subpopulations of DRG neurons, and their activation has important effects on sensory neuron function. High-affinity binding occurs between TrkA and NGF, between TrkB and both BDNF and NT-4/5, and between TrkC and NT-3. There is also some cross-reactivity between these factors and

the other receptors12; for example, NT-3 can also interact with TrkA and TrkB on DRG neurons. The selectivity of neurotrophin receptors for high-affinity binding with their ligands is enhanced by p75 (see later section on p75). Neurotrophins and their receptors are essential for the survival of sensory neurons (and other neurons outside the remit of this chapter) and in the development and maintenance of their phenotype in the adult. Studies on transgenic mice have found losses of DRG neurons of 70%, about 30%, and 55% to 80% for NGF, BDNF, and NT-3 knockouts, respectively, indicating that not only do the majority of DRG neurons depend on these neurotrophic factors for survival at some time during development, but that many require more than one.204 About 75% of DRG neurons in adult rats express at least one of the Trk receptors (TrkA, TrkB, and/or TrkC), which are expressed mainly by small-, medium-, and large-sized DRG neurons, respectively.217 Many neurons express more than one Trk and about 4% express all three.140,143 The percentages of adult rat DRG neurons that express TrkA, TrkB, and TrkC are about 40%, 5% to 10%, and 20%, respectively.147

175

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

INFLAMMATION / NGF ↑SP ↑CGRP ↑Nociceptin ↑BDNF ↑B2 ↑GAL2 ↑MOR ↑Nav1.7 ↑Nav1.8

↓↑GAL ↓GAL1 ↓DOR ↓KOR

↑VIP ↑PACAP ↑GAL nNOS ↑EGF-R ↑GRFα3 ↑LIF-R ↑B2 ↑α2δ (Ca++) ↑Nav1.3 ↑Naβ3 ↑KCNQ2,3,5

↓BDNF ↓SP ↓CGRP ↓SOM ↓trkA ↓IB4 ↓P2X3 ↓GAL2 ↓GFRα2 ↓MOR ↓Nav1.9 ↓Nav1.8 ↓IKIR ↓Kv1.4

↑CGRP ↓GAL1

LARGE

SMALL

↑IL1 ↑TNFα

↑IL1 ↑TNFα ↑LIF ↑PGs ↑5HT

↑GLUT ↑NGF ↑NT4 ↑H+ ↑BK ↑ATP ↑Hist ↑K+ ↑ET1 ↑5HT ↑IL1

↑GLUT ↑SP ↑CGRP

SKIN ↑BDNF ↑NPY ↑EGF-R ↑GFRα1 ↑LIFR ↑α2A (adren) ↑TNFα ↑IL6 ↑Ih ↑KCNQ2,3,5

↓CGRP ↓GAL1 ↓fast IA ↓IKIR ↓T Ca2+ ↓Kv1.1, 1.2 ↓Kvβ2.1

↑NGF ↑GDNF ↑NT3 ↑TGFα ↑TNFα ↑p75 ↑TNFR1

↑NGF ↑BDNF ↑NT4 ↑GDNF ↑FGF2 ↑p75 ↑GFRα ↑FGFR1-3 ↑LIF ↑IL1β ↑IL6 ↑TNFα

↑FGF2 ↑FGFR1-3 ↑IL1α,β ↑5HT ↑TNFα,

SKEL MUSC

↑IL6 ↑LIF

↓↑CNTF ? NEURONAL SIZE ↑TNFR1,2 ↑nAChR7

PERIPHERAL AXOTOMY

Color code : PEPTIDES and GLUT (glutamate); RECEPTORS, ION CHANNELS//CURRENTS; TROPHIC FACTORS ; OTHER MEDIATORS RELEASED BY METABOLISM, TISSUE DAMAGE OR INFLAMMATION; OTHER ;

= DRG NEURON

B FIGURE 8–5 Continued Alterations in expression after axotomy or during inflammation. B, Alterations, gleaned from the literature, and detailed in the text, in expression of a variety of molecules in small and large DRG neurons and in non-neuronal cells closely associated with DRG neurons, that may influence DRG neuron function directly or indirectly. Neuronal cells shown are small and large DRG neurons. Non-neuronal cells represented are, from left to right, satellite cells, Schwann cells, leukocytes and mast cells, skin, and skeletal muscle; to the extreme right is a sensory nerve terminal. Top, Changes in expression following inflammation or application of exogenous NGF. Bottom, Changes in expression that occur in these cells following peripheral nerve axotomy. Because most nerve injury or neuropathic states have associated inflammation, it is likely that changes characteristic of both inflammation and axotomy may contribute to the resulting neuropathic pain. This representation is not intended to be complete, but nonetheless serves to illustrate the complexity of the altered environment that might contribute to altered DRG neuron function. The non-neuronal contribution is contained within the box. See Color Plate

Possible relationships between expression of trophic factor receptors and neurons with different sensory receptive properties are shown in Table 8–2, and the background to this table is given in the text below. The expression of these receptors in relation to the peripheral projection sites of the neurons is given in Table 8–3. TrkA. At birth 70% to 80% of rat lumbar DRG neurons express TrkA, and this declines to about 40% of adult rat DRG neurons that express TrkA strongly.15,215 In the adult rat about 20% of the positive neurons are NF rich and thus

are likely173 to have myelinated fibers. The appearance of TrkA-positive neurons can be seen in Figure 8–4. TrkA expression in adult DRG neurons appears to be partially dependent on target-derived NGF.183 NGF binds to TrkA in membranes of the soma and fibers of TrkA-expressing DRG neurons. Part of the signaling mechanism involves endocytosis and transport of the NGF-TrkA complex to the neuronal soma, and can thus chronically influence both the phenotype and function of TrkA-positive DRG neurons,181 whereas local intracellular signaling pathways39 enable acute local responses such as sensitization to heat (see later).

176

Function of the Peripheral Nervous System

Table 8–2. Trophic Factor Receptors and Sensory Receptive Properties Receptor

Ligand

Size (S/M/L)

Expressed by Possible Afferent Type

CV Range

TrkA

NGF  (NT-3)

SML

C  A  A

TrkB

M

TrkC

BDNF NT-4 NT-3

GFR-1 GFR-3 FGF2

GDNF Artemin FGF2

SML SM S

Mainly nociceptor or silent, some cutaneous LTMs (weak), Visceral, Merkel cell neurites*, whisker follicle afferents* LTM SA & D hair?, Ruffini endings*, whisker follicle afferents* D hair development LTM SA, Proprioceptive, Merkel cell neurites* D hair development Must include some C-fiber nociceptors (IB4) ? Mainly nociceptors Possibly muscle nociceptors

SML

A A ? C

This table summarizes information in the text relating to possible expression/dependencies of different trophic factor receptors, and the possible sensory receptive properties of neurons that express them, or show dependencies on the ligand or receptor in knockout mouse studies of ligand or receptor. Apart from TrkA, their expression has not been directly determined in individual identified neurons in vivo. *Results of studies on growth factor dependencies of large-diameter cutaneous afferent fibers projecting to the whisker complex studied using a variety of neurotrophin and neurotrophin receptor knockout mice. These studies show impairments of development of certain types of nerve fibers and cutaneous structures. The interpretation of these studies was that Ruffini endings were dependent on BDNF and TrkB; Merkel cell neurite complexes were dependent on NT-3, TrkC, NGF, TrkA, and p75; and whisker follicle afferents were dependent on NGF, TrkA, and some also on BDNF and Trk.94 Thus there appear to be complex codependencies on different factors.

Axotomy results in a 25% reduction in the percentage of rat DRG neurons that express TrkA after 1 week and 35% after 2 weeks,282 and 2 weeks after inflammation (adjuvantinduced arthritis), an increase in TrkA protein was reported,247 although other studies showed little or no change in TrkA messenger RNA (mRNA). NGF. NGF is a target-derived growth factor that binds with high affinity to TrkA on sensory neurons. Target tissue cells that secrete NGF include epithelial cells (an important source), smooth muscle cells, and fibroblasts and also, especially during development or after nerve injury, Schwann cells.10,123,341 About 70% of DRG neurons are NGF dependent during embryonic development, and some of

these (mainly the IB4-binding small neurons) switch to glial cell line–derived neurotrophic factor (GDNF) dependence in early postnatal life.216 In the adult, target-derived NGF is normally important in maintaining expression of a variety of molecules in TrkA-expressing DRG neurons, many of which are upregulated as a result of increased NGF expression in the tissues during inflammation.268 These include substance P, CGRP,326 possibly TrkA,183 (although reports on this vary), BDNF,207 growth-associated protein-43 (GAP-43),178 some channel subunits (including voltage-gated Na [Nav] channels Nav1.7,107 Nav1.8, and Nav1.986), bradykinin B2 receptors,177 and auxiliary channel proteins such as P11230 (see Fig. 8–5B). Other peptides are downregulated by NGF, and consequently upregulated after axotomy. These include

Table 8–3. Phenotypic Properties of Rat DRG Neurons That Project to Skin, Muscle, and Viscera Neurons† Marker IR/mRNA*

Cutan.

Skel. Muscle

Visc. Pelvic

Visc. Splanchnic‡

References

TrkA IR or mRNA TrkB mRNA TrkC mRNA None of TrkA, B, or C IB4 binding IR Carbonic anhydrase IR CGRP IR Substance P IR Somatostatin IR NF rich 3 Na pump IR

ⴙⴙ/ⴙⴙ   ⴙⴙ ⴙⴙ  ⴙⴙⴙ ⴙ  ⴙⴙ

ⴙ/() ⴙⴙ ⴙⴙⴙ

ⴙⴙⴙⴙ ⴙⴙⴙⴙ () () ⴙ

ⴙⴙⴙ

17,189,201 201 201 201 17,189 244 17,244 244 244 244 77

ⴙ ⴙⴙⴙ ⴙⴙ () ⴙⴙ In fibers

()/

ⴙⴙⴙⴙⴙ/ⴙⴙⴙ ⴙⴙⴙⴙ

ⴙⴙ

*IR  immunoreactivity; messenger RNA (mRNA) shown by in situ hybridization. † Approximate percentages of neurons that were retrogradely labeled via nerves that project to muscle, skin or viscera, that also expressed the marker indicated in the first column. The exception was the 3 Na pump, which is based on fiber staining. ‡ Splanchnic includes retrograde labeling via the splanchnic nerve and from organs innervated by the splanchnic innervation. Note that, of DRG neurons innervating epidermis, 14% were TrkA positive and 70% showed IB4 binding.189 Percentages indicated as follows: ⴙⴙⴙⴙⴙ  100%; ⴙⴙ  50%; ⴙ  20%;   10%; ()  5%;  none.

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

vasoactive intestinal polypeptide (VIP), cholecystokinin (CCK), neuropeptide Y (NPY), and galanin326 (Fig. 8–5B). NGF and Nociception. NGF is important for the development and maintenance of the normal phenotype of nociceptive neurons by controlling expression of many different molecules.204 When NGF levels were lowered over a critical period (postnatal days 4 to 11 in the rat), fewer A nociceptors and more LTM D hair units were reported, perhaps indicating that A neurons may develop into D hair units rather than nociceptors in the absence of NGF.204 Raised NGF is thought to have acute local effects on peripheral terminals of DRG neurons. For example, NGF causes enhanced responses of cultured DRG neurons to capsaicin within 10 minutes, and in vivo it causes heat hyperalgesia that is dependent at 1 hour on mast cells and at 5 hours on neutrophils.19,285 The mast cell involvement is thought to be due to NGF-triggered NGF release presumably via activation of mast cell TrkA receptors.285 Direct analysis of TrkA immunoreactivity in neurons with identified sensory properties in vivo found intense TrkA-LI expression only in nociceptive neurons with C, A, or A fibers; both the percentage of positive neurons and the intensity of the expression were as high in A-fiber as C-fiber nociceptors.86 In addition, a few D hair units were weakly positive.86 NGF and Inflammation. NGF is upregulated in inflamed tissues. For instance, in inflamed skin it is upregulated secondary to increased levels of cytokines such as interleukins and tumor necrosis factor- (TNF-),336 and this increased NGF is “necessary and sufficient” for the production of inflammatory hyperalgesia.204,336 Its transport to DRG neuronal somata is increased, causing altered expression of many molecules important in the function of nociceptive neurons (see earlier), and leading to altered properties of nociceptive neurons. These altered properties probably contribute to hyperalgesia. They include, in adult rats or guinea pigs, lowered threshold to heat but not mechanical stimuli, increased spontaneous firing, and increased fiber firing rate; these changes and the hyperalgesia are prevented if NGF is sequestered in the inflamed tissues.75,159,337 Some of the electrophysiological changes take more than 1 day to develop—enough time for retrograde transport of NGF, altered gene expression (see earlier), and transport of products of gene expression.75 Apart from its effects on primary afferent neurons, elevated NGF has many effects on local inflammatory and immune responses. For instance, it acts as a chemotactic factor for leukocytes (including neutrophils) and is important in wound healing.163 NGF and Axotomy. Axotomy interrupts the supply of peripheral target-derived NGF to the DRG neuronal soma, and some DRG neurons die. After sciatic nerve lesion, NGF is upregulated in non-neuronal cells in the sciatic nerve, triggered, at least partially, by interleukin-1 (IL-1) released from invading macrophages. However, this NGF

177

increase does not fully replace the normal supply from the periphery.123,186 NGF plays a trophic role after nerve injury, including protection against neuron death.314 Regrowth of damaged fibers within the nerve is said not to be dependent on NGF and may even be encouraged by reduced NGF, although increased NGF encourages collateral sprouting of cutaneous afferent fibers within the skin.68,202,314 TrkB. TrkB, the high-affinity receptor for BDNF and NT-4/5,138 is normally expressed in about half the neurons in rat DRGs, in small- to medium-sized neurons, with about 30% having NF-rich somata. TrkB protein was also found in about 30% of human DRG neurons.136 After nerve injury the number of DRG neurons expressing it increases by 15% to 20%.89 BDNF. BDNF is a survival factor for some sensory neurons during development.47 In the adult, BDNF is expressed in several sites, including CNS DRG neurons and nerve bundles in skin.27 It is expressed in 30% to 50% of DRG neurons; positive neurons are small- to medium-sized, mainly TrkA-expressing neurons; it is also expressed in some TrkA-negative/IB4-positive neurons and in some TrkCpositive but in few TrkB-positive neurons.149,207,246 BDNF can be upregulated by NGF in the BDNF-expressing, TrkApositive neurons. It is downregulated by NT-3139 both in TrkC- and non–TrkC-expressing neurons. BDNF is transported both peripherally and centrally in DRG neurons; centrally it is found in dense-cored and clear vesicles in the superficial dorsal horn (laminae I and II), where it is colocalized in many nerve terminals with CGRP and p75.190,207,317 It is released in the dorsal horn in response to bursts of firing in afferent C fibers and has been proposed to act as an excitatory neurotransmitter/neuromodulator, acting on TrkB receptors on spinal dorsal horn neurons.246 As regards its effects on DRG neurons, evidence from transgenic mice implicates it in the maintenance of the sensory properties of SA mechanoreceptive neurons.47 It can also cause acute heat hyperalgesia.285 BDNF and Inflammation. During peripheral inflammation there is an NGF-dependent upregulation of BDNF in TrkA-expressing DRG neurons and in nerve terminals in the dorsal horn; there is evidence to suggest a role for BDNF in inflammation-induced allodynia/hyperalgesia, perhaps by contributing to central sensitization.56,57,195,317 BDNF and Axotomy. Peripheral nerve injury results in complex changes in BDNF expression in affected DRG neurons. Axotomy of peripheral sensory fibers results in long-lasting (several weeks) changes in BDNF expression of two kinds. There is upregulation of BDNF in somata of medium- to large-sized, non–TrkA-expressing DRG neurons, including TrkB- and TrkC-expressing neurons and in their central projections in deep dorsal horn and gracile nucleus. At the same time, a decreased expression of BDNF is seen in small-diameter neurons and in the

178

Function of the Peripheral Nervous System

superficial dorsal horn.55,144,208 Crush,55 CCI,227 and sciatic nerve section139 are complex nerve injuries that result in a mixture of intact and damaged/axotomized fibers in the nerve, or neurons in the DRG. In such injuries increased BDNF expression occurs in intact small- to medium-sized DRG neurons and in superficial layers of the spinal cord, as well as in large DRG neurons. After sciatic nerve section there was a substantial increase in expression in medium to large neurons for at least 3 weeks.139 Possible reasons for the complex pattern after complex nerve injury are as follows: Axotomy results in loss of NT-3 from the periphery (NT-3 normally downregulates BDNF expression in medium to large neurons) and a loss of NGF (NGF normally maintains, i.e., upregulates, BDNF expression in the smaller TrkA-positive neurons). This would result in BDNF downregulation in small neurons and upregulation in large neurons in a purely axotomized population of neurons. After L5 spinal nerve injury, there is increased expression of BDNF in small- to medium-sized neurons in the adjacent “intact” L4 DRG.93,115 These intact DRG neurons, whose fibers run adjacent to axotomized degenerating fibers, may (1) be subject to inflammatory mediators/NGF as a result of degeneration of the axotomized fibers and (2) have more NGF availability at their terminals, both of which may contribute to the BDNF upregulation in smaller DRG neurons. It has been suggested that increased release of BDNF in the dorsal horn after nerve injury might contribute to central sensitization, although the relative roles of the increased expression in axotomized large primary afferent neurons, or the NGF-related increase in intact small- to medium-sized, probably mainly nociceptive neurons is as yet unclear. NT-4. NT-4 is secreted by a variety of cell types, with high levels of expression embryonically, reducing around birth and increasing in some tissues postnatally. In the adult rat and human it is expressed in skin (including hair follicles), mainly in keratinocytes,27,113 and it is upregulated in human inflamed skin.113 Like BDNF, it can cause acute heat hyperalgesia, an effect dependent on functional mast cells.285 NT-4 is also widely expressed in the CNS.179 The loss of about 75% of TrkB-positive DRG neurons in NT-4 knockout mice299 shows NT-4 to be essential for survival of TrkB expressing neurons during development. Studies on knockout mice299 indicate that survival of DRG neurons with properties of D hair receptors is dependent on NT-3 early in postnatal development and on NT-4 later in the mature animal. TrkC. TrkC is the high-affinity receptor for NT-3. It is expressed by a subpopulation of medium to large DRG neurons, mainly those that project to skeletal muscle (see Table 8–3). TrkC knockout mice lack Ia muscle afferent projections to spinal motor neurons, have fewer large myelinated axons in the dorsal root and dorsal columns, and show movement disorders.155 Studies on knockout mice indicate that many DRG neurons depend on TrkC activation

in early embryonic development and that a substantial number of these switch their dependence to TrkA activation between 11.5 and 13.5 embryonic days334; they also provide evidence that survival of D hair LTMs in early postnatal development is dependent on NT-3.299 NT-3. NT-3 is critical for the survival and function of a population of large DRG neurons (in rat, about 20% of all DRG neurons) that express carbonic anhydrase and parvalbumin; these neurons are probably proprioceptors and cutaneous mechanoreceptors (slowly adapting A units or D hair receptors).82,349 NT-3 is necessary for the development, survival, and/or function of a subpopulation of muscle spindles, Golgi tendon organs, Merkel cells, and D hair receptors.82,349 In adult rats NT-3 is detectable in large neurons most (94%) of which express TrkC, suggesting that the NT-3 is derived from target tissues.54,314 In adults NT-3 is expressed in skeletal muscle and in skin, in skin muscles (canniculus parnosus and arrectores pilorum) and hair follicles.27 NT-3 signaling appears to be necessary for the development of the proprioceptive phenotype.226 Furthermore, there is emerging evidence that, although NT-3 does not affect TrkC expression in adult DRG neurons, it can downregulate TrkA expression especially in neurons with high TrkA expression but not TrkC expression, possibly acting via TrkA receptors.111 NT-3 and Axotomy. Application of NT-3 to the axotomized nerve enhances nerve regeneration and reverses the upregulation of NPY in axotomized neurons.296 p75. p75 is a member of the TNF receptor superfamily. It binds to all the neurotrophins (NGF, BDNF, NT-3, and NT4/5) with a similar low affinity (see Chapter 17). It has important roles in both apoptosis (having a death domain similar to that of many of the TNF receptor family) and in cell survival (for review, see Casaccia-Bonnefil and colleagues48), and these roles are modulated by NGF/TrkA signaling. For example, its absence in p75 knockout mice results in a 50% loss of DRG neurons.300 In adult DRG neurons it is expressed in TrkA-, TrkB-, and TrkCexpressing DRG neurons.140,143,339 Relationships (either ligand passing or binding between receptors) between p75 and Trk receptors can influence the function of these receptors. For example, the presence of p75 enhances ligand specificity for the Trk receptors. Thus, although BDNF, NT-3, and NT-4/5 may each bind to TrkB, TrkB activation by NT-3 and NT-4/5 is reduced if p75 is present, making TrkB more selective for BDNF.266 Furthermore, although NGF and NT-3 can both bind to and activate TrkA of p75, TrkA shows a greater affinity for NGF in the presence and the activation of TrkA by NT-3 is reduced.266 Thus the selectivities of TrkB for BDNF and of TrkA for NGF are both strongly enhanced by the presence of p75. There is evidence that the pro-apoptotic effects of p75 are normally suppressed by NGF/TrkA signaling but that after

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

axotomy, with the loss of target-derived NGF, p75 activation does indeed become pro-apoptotic (see Chapter 17). Receptors for the GDNF Family of Trophic Factors Besides the neurotrophins, a number of other neurotrophic factors are thought to influence DRG neurons and may play roles in the regeneration of peripheral nerves. An important family is the GDNF family, which includes GDNF, neurturin, artemin, and persephin; apart from persephin (which is not further discussed) these all support survival of peripheral neurons in vitro. In the adult, GDNF is found in the CNS, ovary, prostate, adrenal gland, and skin; it is also found in skeletal muscle (human, not mouse) and in axons and Schwann cells in peripheral nerve.105,305 Peripheral sources of GDNF for DRG neurons normally include basal keratinocytes and Schwann cells, with upregulation in Schwann cells after nerve injury.7 Sources of artemin are not yet established. Neurturin is expressed in adult mice in certain visceral organs (gastrointestinal tract, bladder, reproductive organs) and skin.105 Many DRG neurons that are NGF dependent in embryonic life switch dependency to GDNF, neurturin, and/or artemin postnatally.13 The GDNF family of growth factors acts through a receptor complex with two components. This includes a common signal-transducing domain receptor tyrosine kinase RET (or c-Ret), and one of four ligand-binding domains (GDNF family receptor  [GFR]-1 through -4). GDNF acts preferentially through GFR-1, neurturin through GFR-2, artemin through GFR-3, and persephin through GFR-4 (see Chapter 17). GFR-1 through -3 receptors are expressed by mammalian DRG neurons. The growth factors are bound and transported centrally in DRG neurons that express the appropriate ligand-binding domains. Overall about 60% of DRG neurons express RET; these neurons are mainly small to medium with some large neurons, and express one or more GFR subunits147 and many show IB4 binding (see later). Reported percentages of adult rat DRG neurons expressing mRNA for RET, GFR-1, GFR-2, and GFR-3 are 60% to 64%, 40% to 50%, 20% to 33%, and 20%, respectively.18,147 RET, GFR-1, and GFR-2 are in both small and large DRG neurons.18 A higher proportion of large neurons express GFR-1 than GFR-2, and 30% of GFR-1– positive neurons are NF rich, and these probably have myelinated fibers.18 About 20% of RET-positive neurons are TrkC positive147 and thus likely to be LTMs. mRNAs for GFR-2 and GFR-3 were shown to be highly co-localized but have low co-localization with GFR-1 mRNA.147 Few GFR-1 (20%) or GFR-2 (3%) positive neurons were reported to express TrkA,18 (see below for comment) but GFR-3 immunoreactivity was described as having substantial colocalization with TrkA.232 These GFR-3–immunoreactive neurons were 20% of mouse and rat lumbar DRG neurons, most of which (90%) were NF poor and positive for RET, TrkA, CGRP, and transient receptor potential protein V1

179

(TRPV1), and about 30% were IB4 positive,232 features generally indicative of slow conduction velocities and nociceptive properties. However, there is a discrepancy in the above literature between the high co-localization between GRF-1 and TrkA and the high co-localization of GFR-2 and GFR3, coupled with the very low co-localization of GRF-2 and TrkA mentioned previously. The reasons for this are not clear. Thus GFR-3 may be expressed in nociceptive neurons, and, unlike other receptor components in this group, is almost exclusively expressed in sensory neurons, and thus is of potential interest as a therapeutic target (see later). Table 8–2 presents a summary of possible sensory properties of neurons that express the different receptor components. Nerve Injury. Nerve injury leads to upregulation of GFR and GDNF in Schwann cells in the nerve distal to the injury; the level in the nerve is related to the extent of infiltration by macrophages and lymphocytes. In the somata of axotomized DRG neurons, upregulation of GFR-1 in larger neurons and GFR-3 in smaller neurons, decreased GFR-2 levels and unchanged RET expression are reported.16 GDNF can reverse the following postaxotomy changes in DRG neurons: reduced IB4 binding; decreased expression of both somatostatin and the P2X purinergic receptor P2X3; increased expression of activation transcription factor 3 (ATF3), galanin, and NPY in large DRG neurons; slowing of conduction velocity and behavioral changes including tactile hypersensitivity and thermal hyperalgesia.216,332 Systemic exogenous artemin can also reverse or prevent a number of changes associated with nerve injury, including behavioral measures of neuropathic pain.95 Comparison of IB4- and TrkA-Labeled Populations. It is frequently stated that IB4 binding and TrkA expression define separate subpopulations of small or nociceptive neurons. Most or all of the population of small neurons that does not express any TrkA, B, or C shows IB4 binding. Thus most small neurons express either TrkA or IB4 or both. It is therefore worth comparing these two populations. IB4 is a lectin from the plant Griffonia simplicifolia that binds to -D-galactose residues in glycoconjugates on the cell surface and Golgi apparatus of small, NF-poor DRG neurons in a variety of species.214 In Figure 8–4 it can be seen that the staining pattern for IB4 appears to be reciprocal to that for the 200-kDa neurofilament subunit (NF200). Percentages of rat lumbar DRG neurons that bind IB4 increase from 9% at birth to 40% at 14 days postnatal,15 as they switch their dependence from NGF in embryonic life to GDNF in postnatal life.216 Reports on adult rat DRGs indicate that most or all (95% to 100%) IB4-positive neurons express RET while 63% of RET-positive neurons express IB4; up to 80% of IB4-positive cells express GFR-1 and about 50% express GFR-2, whereas only about 30% of GFR-3–positive neurons express IB4.18,216,232 The RET/ GFR receptor components show a distribution similar to

180

Function of the Peripheral Nervous System

that of IB4 in the inner part of lamina II in the dorsal horn.216 While all RET-positive neurons show retrograde transport of 125 I-labeled GDNF, only 13% of TrkA-positive neurons show such transport.216 Coexpression with TrkA is low for GFR1– and GFR-2– (although see discrepancy noted earlier) and high for GFR-3–expressing neurons.18,232 Thus overall, although there is some overlap between the RET/GFR1/IB4 and TrkA populations, the former take up and transport GDNF, whereas the TrkA-positive population binds and transports NGF and expresses little GFR-1 or GFR-2 mRNA, and half of this population expresses GFR-3.232 That there is some co-localization between TrkA expression and IB4 binding8,18 was confirmed with intracellular recording and dye injection studies in rat DRGs. These showed (1) that a third of C-fiber neurons were positive for both TrkA and IB4, with a tendency for reciprocal staining intensities for these two markers; (2) that most nociceptors strongly expressed TrkA or IB4 binding sites; (3) that IB4 binding sites were present on C-fiber but not A-fiber nociceptive neurons, whereas TrkA expression was in both C- and Afiber nociceptors86; and (4) that some weak positive labeling for TrkA and IB4 was seen in some D hair units.86 TrkA-, and not IB4-, positive neurons express substance P and CGRP.18 Other differences between these neurons include projection of TrkA-positive neurons to laminae I and IIo and IB4-positive neurons mainly to IIi.214 Compared with IB4-negative small neurons, IB4-positive neurons have longer duration action potentials and a smaller noxious heat-activated current,301 and the TTX-resistant (TTXR) Na channel subunit Nav1.9 is preferentially expressed in IB4-positive cells.87 In summary, A-fiber nociceptors express TrkA but not IB4 binding sites, while most C-fiber nociceptors express one or the other, or both of these. Apart from the NGF and GDNF dependence of TrkA expressing and IB4 binding neurons respectively, the functional differences between these groups of nociceptors are poorly understood. The Fibroblast Growth Factor Family and Their Receptors Almost all tissues express fibroblast growth factors (FGFs). The FGF family can influence growth, development, differentiation, and fiber branching and appears to play important roles in the responses of DRG neurons to injury. These are part of a larger family of at least 22 different FGFs. There are four FGF receptors (FGFR-1 through -4) that are tyrosine kinase receptors. Some of these have several isoforms. mRNAs for at least seven FGFs (2, 5, 7, 9, 10, 13, and 14) and several receptor types are present normally in DRGs, in neurons and/or non-neuronal cells.182 A subpopulation of DRG neurons that is NGF dependent during embryonic life switches to dependence on FGF-2 later in development; this group of neurons expresses the receptor for FGF-2, expresses substance P, and is NF poor, and thus presumably (see earlier) has C fibers.1 Skeletal muscle as well as DRG and spinal cord are sources of FGF-2, and it has been suggested that these

FGF-2–dependent DRG neurons may be muscle C-fiber nociceptive neurons.1 Axotomy. In DRGs following peripheral axotomy, some FGFs are upregulated, as are their receptors (e.g., FGF-1, FGF-2, FGFR-1 through -3), and some are downregulated (e.g., FGF-13).135,182 FGF-2 and FGFR-1 through -3 are upregulated in Schwann cells and macrophages at the injury site, and FGF-2 stimulates nerve regeneration and Schwann cell proliferation after axotomy.114 Neuropoietic Cytokines and Their Receptors These cytokines, expressed by Schwann cells, fibroblasts, and sensory neurons, as well as by macrophages, mast cells, and lymphocytes have local influences on neurons including sensory neurons, and on glial cells. Examples are members of the interleukin-6 (IL-6) family: ciliary neurotrophic factor (CNTF), leukemia inhibitory factor (LIF), IL-6, IL-1, TNF-, and transforming growth factor- and - (TGF- and TGF-). Many of these are upregulated in tissue damage and/or inflammation and may also be known as inflammatory mediators. (See Fig. 8–5A.) The IL-6 Cytokine Family. This family includes the structurally related molecules CNTF, LIF, IL-6, and IL-1. CNTF and LIF both support the survival of sensory neurons after axotomy and in culture.315 They have ligand-specific  receptors (CNTF-R, LIFR, IL-6R, IL-1R). They also have signal-transducing components one of which, gp130, is common to signaling by IL-6, CNTF, and LIF.315 gp130 is expressed in most DRG neurons and its expression is unchanged after axotomy.96 The soluble ligand-specific receptors can be released from cells and act as receptors on other cells.315 CNTF. CNTF is a neurotrophic cytokine expressed by glial cells in the peripheral nervous system and CNS. The CNTF receptor (CNTF receptor ) is expressed in all DRG neurons.200 Following nerve injury, the expression of CNTF in Schwann cells is reduced peripheral to the axotomy but increases as nerve regeneration begins, perhaps because axonal contact with glial cells is required for its expression.314 It enhances neurite outgrowth of DRG neurons and is upregulated by IL-6.287 IL-6. IL-6 is not normally detected in adult DRG neurons but is synthesized in a third of DRG neurons (medium and large sized) and in non-neuronal cells such as Schwann cells after nerve injury.219 After axotomy, endogenous IL-6 is pro-nociceptive and promotes DRG neuron survival; these effects may be in part due to its upregulation of BDNF in medium- to large-sized DRG neurons.219 IL-6 also enhances neurite outgrowth in axotomized DRG neuron fibers, possibly by upregulating CNTF and gp130 expression.287 IL-6 is upregulated in skeletal muscle after denervation.161 IL-6R. Almost all DRG neurons express IL-6R, although IL-6 supports the survival of only about 30% of embryonic

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

DRG neurons; this is thought to be because IL-6 acts by induction of, for example, BDNF expression in DRG neurons.219 IL-6 also supports survival of newborn rat DRG neurons, but only if exogenous IL-6R is also provided.315 IL-1. IL-1 is expressed in most medium to large DRG neurons and some satellite cells.61 It enhances neurite regeneration, causes some afferent fibers in skin to fire spontaneously, can cause substance P release from DRG neurons, and induces cyclooxygenase-2 (COX-2) expression.129 IL-1 is upregulated in inflamed tissues, and an antagonist to IL-1 reduces inflammation-induced hyperalgesia.268 LIF. About 25% of rat DRG neurons retrogradely transport LIF in vivo. These are small neurons, mainly the CGRP/ TrkA-positive population, but also some IB4-binding neurons.318 LIF immunoreactivity was found in Schwann cells and nerve fibers. In damaged peripheral nerve, LIF upregulation is predominantly in Schwann cells. LIF promotes nerve regrowth and normal regeneration of nerve fibers after injury.37 In human nerves following injury, LIF expression is upregulated within hours and can stay high for years where a neuroma is present.79 During an inflammatory response, it is upregulated in neutrophils, macrophages, mast cells, and blood vessel walls.79 After denervation LIF is also upregulated in skeletal muscle cells.161 Although LIF can be pro-inflammatory, it has anti-inflammatory effects following CFA injection to the tissues.303 LIFR. LIFR is normally expressed in the nucleus of most DRG neurons, but in the cytoplasm of only a few; however, following peripheral axotomy, it is expressed for at least 2 weeks in the cytoplasm of a high proportion of small, medium, and large DRG neurons.96 TNF-. TNF- is a pro-inflammatory cytokine, as well as a “potent activator of Schwann cell and other glial cell cytokines,” and plays an important role in the early degenerative changes after nerve injury.286 TNF- is also strongly implicated in both inflammatory and neuropathic pain. It causes nociceptive fibers to fire spontaneously, and intraneural injection of TNF- induces neuropathic pain– related behavior in rats through the activation of its receptor TNF-R1.292 In a neuropathic pain model, it markedly enhances neuropathic pain behavior, including spontaneous foot lifting.270 Normally it is found in small-diameter DRG neurons and is transported peripherally in muscle but not cutaneous afferent DRG neurons.270 Intraneural TNF- is derived from Schwann cells and macrophages, although some may be expressed by DRG neurons themselves.270 Nerve Injury. After nerve injury, TNF- is upregulated in Schwann and satellite cells.228,270 It is transported centrally in DRG neurons via the DRG to the spinal cord, where it is released and taken up by spinal cord neurons (dorsal horn and motoneurons) and glia.286 At this time increased numbers of medium to large DRG neurons were reported to contain TNF- in both ligated L5 (injured) and nonligated adjacent L4 DRG neurons, although to what

181

extent this is derived from glial cells or DRG neurons is not clear.270 The two TNF- receptors TNFR1 (R55) and TNFR2 (R75) are both upregulated in damaged and in adjacent, presumed intact, DRG neurons after spinal nerve injury, peaking within 24 hours.228,271 In addition, TNFR1 is upregulated in satellite cells after nerve injury.228 TGF-. TGF- is a member of the epidermal growth factor (EGF) family, and is important in the control of glial and Schwann cell proliferation and survival of differentiated neurons.294 Within the DRG, TGF- and its receptor (EGF) are expressed mainly in small DRG neurons and in satellite cells surrounding some large- or medium-sized neurons.350 Nerve injury results in upregulation of TGF- in satellite cells, and of its receptor in all DRG neurons from 1 to more than 14 days.350 Its role in the DRG after nerve injury is not fully understood, although it may stimulate proliferation of satellite cells. (See Fig. 8–5B.) TGF-. The TGF- family is a superfamily of cytokines secreted by a variety of immune cells and with a variety of functions at least some of which are anti-inflammatory. During development they are necessary for the survival enhancement offered to DRG neurons by some of the neurotrophins. In the adult, they are involved in Schwann cell proliferation, secretion of extracellular matrix and neurotrophins, and control of expression of cell adhesion molecules. TGF- and TGF receptor 2 are found in satellite cells. TGF-2 is present in about half the DRG neurons and in fibroblasts in the DRG.294 It is expressed in mechanoreceptor-related structures in skin, being present in pacinian corpuscles (TGF-1 and -2) and in Meissner corpuscles (TGF-2), and TGF- receptors (I and II) were found in Merkel cells.294

Inflammatory Mediators and Tissue Damage There is substantial evidence of the importance of inflammatory mediators and the immune system in chronic pain states, with half the clinical cases of human neuropathic pain being associated with infection or inflammation of a peripheral nerve (see review by Watkins and Maier333). Pathogens and other types of tissue damage can activate inflammatory and immune responses. Pro-inflammatory mediators released locally from damaged tissues include bradykinin, ATP, K, H, and 5-hydroxytryptamine (5-HT or serotonin), and activation of the arachidonic acid pathway in these tissues results in production of prostaglandins, prostacyclin, leukotrienes, and thromboxanes that tend to be proinflammatory, acting as chemoattractants to encourage infiltration of immune cells. Such cells, as well as other local cells within peripheral nerves, including macrophages, T lymphocytes, mast cells, fibroblasts, endothelial cells, and Schwann cells in the nerve or satellite cells in the DRG, secrete further mediators such as cytokines and growth

182

Function of the Peripheral Nervous System

factors, including those described earlier. Many of these (e.g., NGF) are also pro-inflammatory.333 For example, peripheral nerve injury involves proliferation and activation of satellite cells; these release pro-inflammatory cytokines such as TNF-,228 as well as neurotrophic factors such as NGF, BDNF, and NT-3, and they also express p75.351 The ensuing Wallerian degeneration involves upregulation of TNF-, IL-1, and IL-1 in Schwann cells and in recruited macrophages in peripheral nerve280; the Schwann cells produce neurotrophic factors (such as NGF, GDNF, NT-4, and p75) (see reviews by Terenghi314 and Watkins and Maier333). Mechanical damage as well as pathogens may result in local tissue damage, including exposure of proteins (such as the P0 and P2 peripheral nerve proteins).333 These are normally hidden from the immune system but, once exposed, can trigger inflammatory and immune responses.333 In addition, mast cells and macrophages, when activated, produce destructive molecules such as digestive enzymes that are not only effective in destroying pathogens, but may also damage nerve tissue,333 thus triggering further involvement of inflammatory and immune processes in a positive feedback loop. Positive feedback loops involving cascades of mediators involved in inflammatory and immune responses may have the effect of developing and sustaining these responses. Primary afferent fibers are intricately involved in these processes. They may be activated or sensitized by a number of pro-inflammatory mediators, leading to greater intrinsic excitability. Sensitization of afferent terminals may be a local acute effect or may result from alteration of the neuronal synthetic activity, leading to altered expression or properties of molecules such as ion channels, receptors and/or neuropeptides, or insertion of novel channels and/or receptors into the membrane of the whole neuron. For example, activation or sensitization of nociceptive fibers by proinflammatory mediators can cause release from both peripheral and central afferent terminals of the pro-inflammatory peptides substance P and CGRP. These further enhance inflammatory and immune responses and may excite afferent terminals peripherally, and enhance transmission in nociceptive pathways centrally. Thus cascades of inflammatory mediators and sensitization of primary afferent neurons can cause positive feedback loops that are likely to be very important for developing and sustaining inflammatory or neuropathic pain.333 Some of the molecules that are up- or downregulated within DRG neurons or that might, via receptors on DRG neurons, influence DRG neurons are illustrated in Figure 8–5B for peripheral inflammation (top) or nerve injury (bottom).

Further Inflammation-related Receptors and Mediators Apart from the neurotrophic factors and cytokines, DRG neurons are influenced by a number of other substances released by damaged tissues, by local cells such as leukocytes,

fibroblasts, epithelial cells, and Schwann or satellite cells or even by the neurons themselves. Many of these substances could be classed as pro-inflammatory, or occasionally as anti-inflammatory, mediators. These are introduced below. Prostaglandins, Prostaglandin Receptors, and COX Enzymes. Prostaglandins are important inflammatory mediators. Apart from widespread pro-inflammatory effects, prostaglandins such as prostaglandin E2 (PGE2) have a number of effects on primary afferent neurons, including sensitization of nociceptors to bradykinin and other mediators as well as to natural stimuli.152 One mechanism contributing to nociceptor sensitization is that PGE2 lowers the activation threshold of the TTXR (tetrodotoxin resistant) inward Na current that is found in nociceptive DRG neurons.102 PGE2 is also implicated in mechanotransduction in the renal pelvis; here stretch causes the release of PGE2, which causes the release of substance P, which in turn activates primary afferent fibers.160 PGE2 binds to prostaglandin E (EP) receptors. Of the four EP receptor isoforms, EP1, EP3, and EP4 have been found in DRG neurons.11 Synthesis of prostaglandins requires COX (cyclooxygenase) enzymes, including the constitutive COX-1 and the inducible COX-2 isoforms. COX-2 is induced in peripheral tissues by cytokines, growth factors, and other inflammatory stimuli,152 resulting in increased prostaglandin release. DRG neurons express COX-1 in many of the SD neuron (C-fiber) population, both CGRP-positive and IB4-positive neurons, but its expression is unchanged by peripheral inflammation.58 COX-2 is not expressed in DRG neurons either normally or during inflammation.58 ATP Receptors. A discussion of purinergic receptors (P2X and P2Y receptors) is presented in the later sections on mechanotransduction and chemotransduction. Histamine1 Receptors. The predominant sensation when histamine is applied to the receptive field of human afferent C fibers is itch.272 Histamine sources include mast cells in rat. The histamine1 (H1) receptors are expressed on a subpopulation of small DRG neurons141,142 with unmyelinated fibers that bind IB4. After nerve injury these IB4 binding neurons downregulate H1, although it is upregulated in other mainly small neurons that express substance P– and CGRP.141 Bradykinin B1 and B2 Receptors. Bradykinin is algesic and pro-inflammatory. It is released as a result of damage to the tissues, and contributes to the inflammatory cascade. It causes pain, sensitization, and hyperalgesia in humans, and in a rat skin-nerve preparation it activates greater than 50% of C-polymodal nociceptors; all bradykinin-sensitive neurons are also capsaicin sensitive.162 Bradykinin acts via B1 and B2 receptors on DRG neurons and sensitizes TRPV1 receptors to heat.304 Both receptors are expressed in some of the small

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

DRG neurons; B2 was strongly upregulated (in the same group of neurons that normally express it) by activation of TrkA receptors by NGF and also to a lesser extent by GDNF,177 and a B2 receptor antagonist (Bradyzide) reduces inflammatory hyperalgesia in animal models.152 Two weeks after CCI, B2 and B1 were upregulated in rat DRGs,180 although B1 expression in DRG neurons was decreased 2 weeks after axotomy (sciatic nerve section).90 Endothelin Receptors. There are three endothelins that bind to two receptors, ETA and ETB, with distinct effects. Endothelin 1 (ET1) can induce pain-related behavior through the ETA receptor.249 ETA receptors, when activated by ET1, activate nociceptive C and A fibers in rats in vivo,101 possibly by causing a lowered activation threshold for TTXR Na channels.352 ET1 is expressed by endothelial cells, macrophages, and Schwann/satellite cells as well as by small- to medium-sized DRG neurons. Skin injury causes its release. ETA receptors are found in small- and some medium-sized DRG neurons, but ETB receptors are mainly found on satellite cells or Schwann cells around unmyelinated fibers.249 Peptides. Several peptides released by DRG neurons can influence the inflammatory and immune systems. Substance P and CGRP are vasoactive, pro-inflammatory and proimmune, and somatostatin is anti-inflammatory and antihyperalgesic. For more detail see the appropriate section later under Neuropeptides. Serotonin (5-Hydroxytryptamine). 5-HT is released following tissue injury from mast cells, can cause pain or hyperalgesia,272,310 and is vasoactive. One of its effects is to influence the properties of the TTXR inward Na current typical of nociceptive neurons, lowering its threshold and making the neurons more excitable.41,102 Of the 15 receptor subtypes known, mRNAs for 6 have been described in DRG neurons: 5-HT1B, 5-HT1D, 5-HT2A, 5-HT2B, 5-HT3B, and 5-HT4 receptors.222 Nitric Oxide. Nitric oxide (NO) is synthesized by the enzyme nitric oxide synthase (NOS). Normally in DRGs endothelial cells express eNOS (the endothelial form), and a small population of small- to medium-sized peptide (substance P/CGRP)–containing DRG neurons express nNOS (the neuronal form), the expression of which is downregulated in these neurons by target-derived NGF.316 Axotomy. After peripheral nerve exotomy, nNOS mRNA and protein are upregulated in small- to medium-sized peptide-containing DRG neurons, perhaps as a result of the loss of target-derived NGF.316 It has been suggested (see review by Thippeswamy and Morris316) that NO release from axotomized neurons in neuroprotective, by (1) upregulation of cyclic GMP in satellite cells, which may initiate

183

expression of neuroprotective molecules and/or inhibit apoptosis, and (2) induction of NGF expression in satellite cells.

Other Receptors Adrenoreceptors. A small number of DRG neurons of all sizes normally express the 2A adrenoreceptor; this number was increased sixfold after peripheral axotomy, with the greatest increase being in medium to large neurons, but was unchanged after inflammation of the tissues.23 In contrast, the relatively low numbers of neurons normally expressing the 2C receptor remained unchanged after axotomy or inflammation.23 This increase in 2A adrenoreceptor expression may be important in sympathetically associated neuropathies. Cholinergic Receptors. Cholinergic agonists can excite peripheral sensory fibers in skin as well as excite isolated DRG neurons; in addition, cholinergic fibers make presynaptic connections with primary afferent fibers in the dorsal horn (for references, see Xiao et al.342) Nicotinic acetylcholine receptor  subunits (nAChR) 2 through 7 have been reported in DRG neurons,98 and nAChR7 is upregulated in DRGs after axotomy.342 Cannabinoid Receptors. Cannabinoids have an antihyperalgesic effect in the periphery, thought to be mediated by their actions on CB1 receptors on DRG neurons and on CB2 receptors on non-neuronal cells involved in generation of inflammation.151 CB1 receptors are expressed mainly by medium to large NF-rich (presumably A-fiber) DRG neurons, but there is also some expression in smaller, possibly C-fiber neurons.30,151 GM1 Receptors. The ganglioside GM1 receptor, which has a high affinity for the cholera toxin B subunit, is normally expressed in nearly all NF-rich LL DRG neurons.265 However, after peripheral axotomy, this ceases to be a selective marker for LL neurons because SD neurons also express it.321

Oligosaccharides Expressed by DRG Neurons Cell surface glycoconjugates decorate the surface of several populations of DRG neurons (for a review, see Lawson165). Other surface markers of large DRG neurons include the antibodies to stage-specific embryonic antigen-3 and -4 (SSEA3 and SSEA4), and the lectins GSAII (Griffonia simplicifolia agglutinin II), PSA (Pisum sativum antigen), and LCA (Lens culinaris antigen), while markers of small cells include many antibodies to glycoconjugates, including lactoseries carbohydrates and lectins IB4 and PNA (peanut agglutinin).78,165,298 (For information about neurons to which the lectin IB4 binds, see earlier.)

184

Function of the Peripheral Nervous System

TRANSMITTERS, NEUROMODULATORS, PEPTIDES, AND THEIR RECEPTORS Some of the possible transmitter or modulator substances expressed by DRG neurons are mentioned briefly here. It is not the remit of this chapter to discuss the postsynaptic receptors for these transmitters. However, brief mention of receptors is made where these are expressed on DRG neurons because this may indicate a function on presynaptic terminals in the dorsal horn and/or on fibers in the periphery or even on the DRG soma. Possible transmitters or modulators known to be expressed by DRG neurons that have so far been identified include the excitatory amino acids glutamate and aspartate, the peptides substance P and CGRP, peptides upregulated after nerve injury (galanin, NPY, and VIP), and BDNF (see earlier under Neurotrophins and Trk Receptors), all of which are released from terminals of DRG neurons and may act as transmitters or neuromodulators. This is not a complete list, nor is it likely that the complete list is yet known. Complex cross-interactions between transmitters and receptors for different transmitters/modulators (e.g., tachykinins on N-methyl-D-aspartate [NMDA] receptors) are not covered here. See summary Figures 8–3 and 8–5.

Excitatory Amino Acid Transmitters Glutamate immunoreactivity is in about 70% of DRG neurons, and is also seen in primary afferent fibers in the dorsal horn, in some cases co-localized with substance P.66 Glutamate and to a lesser extent aspartate immunoreactivity was seen in primary afferent terminals in the superficial dorsal horn, but only glutamate was also in laminae III and IV.324 These excitatory amino acids are thought to act as transmitters from the primary afferent terminals, with the widespread expression of glutamate making it a candidate transmitter for most primary afferent synapses. In addition, DRG neurons express a number of receptors for glutamate (GluR), which they transport both centrally and peripherally along their fibers. Peripherally, glutamate receptors may have a role in sensory transduction at the periphery. Glutamate receptors fall into several classes: the NMDA, amino-5-methyl-4-isoxazole-pro acid (AMPA), and kainate receptors. To date, limited immunocytochemical or in situ hybridization studies have suggested the following. Most DRG neurons express the NMDA receptor subunit 1 (NMDAR1).269a AMPA receptor subunits are also expressed: GluR1 by small DRG neurons and GluR2/3 by both small and large neurons;269a the GluR5 subunit of the kainate receptor is expressed strongly in many small DRG neurons (see review by Ruscheweyh and Sandkuhler267). NMDA receptors are found on central terminals of putative nociceptive neurons; NMDA, AMPA, and kainate receptors are also found on both myelinated and unmyelinated fibers in skin.42 Behavioral responses to glutamate, NMDA, AMPA, or kainate injected into skin or intra-articularly are indicative of allodynia and hyperalgesia, and these behaviors

can be blocked by appropriate antagonists or enhanced by local substance P injection.42 For further information on glutamate receptor function, see reference 211. Nerve Injury. Nerve injury results in increased immunoreactivity for the GluR2/3 AMPA receptors in primary afferents in the dorsal horn.250 Inflammation. Glutamate levels increase in inflamed tissues and in synovial fluid of human arthritic joints. Sources of glutamate may include afferent (possibly nociceptive) fibers, dermal/epidermal cells, macrophages, and Schwann cells.42 Peripherally applied glutamate receptor antagonists can attenuate the behavioral responses to inflammation of the tissues, and 48 hours after inflammation induced by CFA, increased numbers of fibers in the tissues show expression of NMDA, AMPA, or kainate receptors; although a simultaneous decrease in NMDAR1 in DRG neuronal somata is seen, possibly because the receptor is transported out of the somata to the fibers/terminals.42

Neuropeptides The term peptide-containing is commonly used to refer to the population of DRG neurons that normally expresses CGRP/substance P. However, other peptides are expressed in DRG neurons, including somatostatin, VIP, galanin, dynorphin, and enkephalin/leu-enkephalin. Substance P and CGRP. Substance P and neurokinin A are tachykinins that are expressed by DRG neurons.162 About 20% of rat and 15% to 40% of human DRG neurons express substance P, and 40% to 60% of rat or human DRG neurons express CGRP148,165 (S. Border and S. N. Lawson, unpublished data). The most intense expression of both is in small neurons, with clear expression in medium (both peptides) and large (CGRP) neurons. Figure 8–4 illustrates the appearance of rat DRG neurons that express CGRP. Substance P–expressing DRG neurons are small,165 and in rat almost all also express CGRP.291 Most (84%) DRG neurons that express CGRP in rat also express TrkA mRNA, and CGRP expression is upregulated by NGF; in contrast, few express TrkB or TrkC mRNA.8,148 Functions. In guinea pig, substance P was in half the cutaneous nociceptive DRG neurons examined, mainly in those with C fibers, but also in some with A and a few with A fibers.170,172 CGRP-positive neurons were mainly nociceptive with C, A, and A fibers, although a few LTM units with A fibers were also weakly positive.169 Noxious stimuli cause both substance P and CGRP to be released from both peripheral and central terminals of DRG neurons262; both can cause depolarization of dorsal horn neurons, and there is considerable evidence that substance P acts as an important neurotransmitter or neuromodulator in the spinal cord.291 CGRP may act to prolong the central effects of substance P released from primary afferent neurons.291 Peripherally released substance P and CGRP cause vasodilation, substance P causes extravasation of plasma

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

from small blood vessels,126 and both substance P and CGRP are pro-inflammatory and promote wound healing.28 Substance P/CGRP– containing nociceptive afferent fibers are therefore important for integrity of the tissues, in influencing the local blood supply, and in local inflammatory and immune responses. The expression of both peptides is upregulated by NGF; thus in nerve injury both substance P and CGRP are downregulated in DRG neurons, whereas inflammation causes their upregulation.38,125 RECEPTORS. Substance P is thought to act via the neurokinin-1 (NK-1) receptor291; all three neurokinin receptors (NK-1 through -3) are expressed on DRG neurons.29 Substance P injected into rat skin was reported to cause mechanical hyperalgesia and allodynia; this effect may have been via activation of the NK-1 receptors that are seen on 30% of cutaneous unmyelinated fibers.45 CGRP receptor component protein was shown immunocytochemically in small- to medium-sized DRG neurons.192 Inflammation and Nerve Injury. A 50% increase in the proportion of unmyelinated fibers in rat skin that express NK-1 receptors was found 48 hours after CFA-induced inflammation.43 The expression of CGRP receptor component protein was increased by inflammation and decreased by nerve injury.192 Somatostatin. Between 5% and 15% of DRG neurons in rat express somatostatin-LI, with positive neurons being exclusively small, mainly IB4 positive, and NF poor, thus probably with C fibers.146,165 Somatostatin expression is upregulated by GDNF,18 and GDNF can reverse its downregulation after axotomy.18 Unlike substance P and CGRP, somatostatin-expressing neurons do not express TrkA, TrkB, or TrkC mRNA, and thus are unlikely to be regulated by NGF, BDNF, or NT-3/4.148 Possible Functions. Somatostatin is released centrally in the spinal dorsal horn from primary afferents during heat- but not mechano-nociception, and it is also released peripherally from DRG neurons during tissue injury/ inflammation.194 Somatostatin can inhibit firing in both dorsal horn and DRG neurons, and peripherally it attenuates the inflammatory/immune response194 and is thus both antinociceptive and anti-inflammatory (see, e.g., Malcangio et al.194). GDNF acutely promotes activity-induced, but not basal, somatostatin release from sensory nerve terminals in the dorsal horn.194 The afferent receptive properties of somatostatin-expressing neurons are not yet established. Somatostatin Receptors. There are several somatostatin receptor (SSTR) subtypes: SSTR1 through -5, with two forms of SSTR2, a and b. Several are expressed in the dorsal horn, but only SSTR2a has been found in DRG neurons, in a small group of small- to medium-sized DRG neurons that do not express the peptides substance P or CGRP.44,276

185

Galanin. Galanin is detectable in few rat DRG neurons normally, mainly in small substance P/CGRP–expressing neurons, although there is evidence of its synthesis in a higher proportion.187 After nerve injury it is strongly upregulated, and after inflammation it is initially downregulated followed by later upregulation “suggesting transition from an inflammatory to a nerve injury state.”38,187 Galanin Receptors. There are three galanin (GAL) receptors: GAL1, GAL2, and GAL3. GAL1 mRNA is present in half the DRG neurons, in large neurons, whereas GAL2 mRNA is expressed in 80% of DRG neurons, but these are mainly small DRG neurons; in addition, many lamina II neurons express GAL2.187 In DRGs, after axotomy GAL1 is strongly reduced and GAL2 moderately reduced.187 After inflammation, DRGs show decreased GAL1 mRNA and a transient increase in GAL2.187 It has been suggested187 that pronociceptive effects of galanin are due to activation of GAL2 receptors on primary afferent terminals and that high levels of galanin protect against nerve injury–induced pain by acting on postsynaptic GAL1 receptors. VIP and Pituitary Adenyl Cyclase–Activating Protein. VIP and pituitary adenyl cyclase–activating protein (PACAP) are members of the same glucagon/ secretin superfamily and are widely expressed in the CNS. They are normally present in primary afferent fibers in laminae I and II of the dorsal horn and are both expressed at low levels in small- to medium-sized DRG neurons.70 There is evidence that PACAP may promote differentiation of DRG neurons in rat, via the PAC1 receptor.223 Because PACAP and VIP are markedly upregulated (PACAP over 2 days and VIP over 2 weeks) in small DRG neurons after nerve injury, they may play different roles in neuropathic pain development.70 Neuropeptide Y. NPY is not normally expressed in DRG neurons but is upregulated in medium to large DRG neurons after axotomy, in the same neurons that upregulate BDNF after axotomy (see later).144,184 This increase in NPY remains until reinnervation occurs and is partially reversed in large DRG neurons by administration of NT-3, perhaps because NT-3 enhances regeneration, or because NT-3 causes a downregulation of NPY, or both.296 Centrally NPY inhibits activity-evoked substance P release.81 Opioid Receptors. Endogenous enkephalins and opiates affect transmission of nociceptive information by activating postsynaptic K channels in the CNS and inhibiting presynaptic Ca2 channels on DRG neurons.14 They act via opioid receptors on central (CNS) or peripheral terminals (e.g., in skin) of afferent fibers. A combination of in situ hybridization and functional studies indicate the following. The , , and opioid receptors (MORs, DORs, and KORs) are expressed by adult DRG neurons, mainly small neurons, with MORs but not DORs being

186

Function of the Peripheral Nervous System

co-localized with substance P/preprotachykinin A.134,213 Activation of opioid receptors inhibits whole-cell Ca2 currents in small but not large DRG neurons, and activation of DORs on dissociated cultured DRG neurons from young rats reduced N-, L-, P-, and Q- but not R-type Ca2 currents.2 Peripheral activation of MORs can inhibit glutamate-induced activity in (possibly nociceptive) primary afferent fibers in skin.59 Interestingly, MORs are downregulated after axotomy and upregulated after peripheral inflammation, whereas DORs and KORs are downregulated in DRG neurons after inflammation.134 Opioid-like Receptors. Nociceptin, the endogenous ligand for opioid-like receptor 1 (ORL1) receptors, is algesic, inducing thermal hyperalgesia–type behavior in rats. Prepronociceptin in RNA is expressed in a few small DRG neurons that do not themselves express substance P and CGRP but are adjacent to neurons that do express these peptides.210 In contrast, ORL1 is abundantly expressed in small- to medium-sized neurons many of which express substance P and CGRP, suggesting a possible presynaptic action of nociceptin.210 Peripheral inflammation causes rapid (30 minutes, recovering by 6 hours) upregulation of prepronociceptin mRNA in a subpopulation of TRPV1expressing small- to medium-sized DRG neurons.132

CYTOCHEMISTRY OF CUTANEOUS, VISCERAL, AND MUSCLE AFFERENTS Retrograde labeling studies on primary afferent neurons that project to different types of tissue show patterns in the cytochemical properties of afferents projecting to skin, muscle, and viscera. These patterns may reflect differences in afferent receptor type, differences in trophic factor availability in the tissues, or perhaps differences in conduction velocity or modality of neurons projecting to the different types of tissue. Such studies often involve either (1) cutting the nerve projecting selectively to one of these types of tissue and applying dye to the cut end of the nerve, or (2) injecting the dye into the intact tissues. Technical problems accompanying such studies include phenotypic changes of neurons with cut or damaged fibers, damage to the nerve terminals that take up the dye in the tissues, and spread of dye into other tissues (see Fox and Powley92 and Perry and Lawson244). A summary of some of the differences in chemical phenotype of neurons projecting to skin, muscle, and viscera is shown in Table 8–3. Note that the percentage of visceral afferents labeled by a particular marker may differ according to the type of innervation (splanchnic compared with vagal afferents). Overall, splanchnic visceral afferents have the highest expression of TrkA, TrkB, substance P, and CGRP; muscle afferents have the highest expression of TrkC, carbonic anhydrase, and the 3 sodium pump subunit; and cutaneous

afferents have the highest expression of binding for several lectins, including IB4, with particularly high IB4 expression in afferents projecting to epidermis (see Table 8–3). Several markers are selectively associated with a subpopulation of DRG neurons with a generally large size; these may be related to LTMs, including cutaneous or proprioceptive afferents. These include cytochrome oxidase, carbonic anhydrase, parvalbumin, and calbindin D28k (see Carr and Nagy46 and Lawson165). It was suggested that carbonic anhydrase was localized mainly in muscle afferents, and in rat, DRG neurons with carbonic anhydrase activity were those with short-duration action potentials and short afterhyperpolarizations.46,251 More recently, immunocytochemical evidence indicates that the 3 subunit of the sodium pump (Na,K-ATPase) is expressed, perhaps selectively, in afferent and efferent fibers that innervate intrafusal muscle fibers of muscle spindles.77

MEMBRANE RECEPTORS AND SENSORY TRANSDUCTION In recent years a number of molecules have been cloned that are thought to be responsible for transduction of different types of sensory stimuli. From the transduction properties of molecules expressed in subpopulations of sensory neurons, it is possible to make more or less informed deductions about the possible or likely sensory receptive properties of the DRG neurons that express them. These are summarized in Table 8–4 and a brief introduction to them is provided below. Also see summary Figures 8–3 and 8–5A and B. Many studies of the transduction mechanisms have been made in heterologous expression systems, or from patch-clamp studies of dissociated DRG neurons in vitro. Other information is derived from defects in behavioral responses to selected stimuli in transgenic (knockout) mice. Thus the certainty with which predictions of the possible physiologic functions that different receptors might have in sensory neurons in vivo is very variable. This is a fast-moving and changing field, and the table is aimed at providing starter references and illustrating some of the possibilities as seen at the time of writing.

Thermoreceptive Molecules Heat Transduction Several candidate molecules for detection of different thermal stimuli have recently been identified and cloned. These include several members of the TRP protein family. TRPV1 (VR1) was cloned in 1997; it is the capsaicin receptor and responds to noxious heat (42° C).50 It has a thermal threshold similar to that of C-polymodal nociceptors and is expressed in small DRG neurons. Because capsaicin appears to selectively activate C-polymodal nociceptors,150,309 an (unconfirmed) possibility is that TRPV1 may be responsible for noxious heat transduction in these

187

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

Table 8–4. Putative Transduction Molecules Stimulus Type* Molecule

Mechanical

TRPV1 TRPV2 TRPV3 TRPV4 TRPM8 TRPA1/ANKTM1 ASIC/ASIC1 DRASIC/ASIC3 BNC1/ASIC2 Stomatin MEC-2 P2X3 P2Y1

Chemical

Thermal

Size (S/M/L)

? Afferent Type†

References

Low pH, capn

42° C

S

50

52° C 39° C 32° C

SML SML SM NF-rich

C PM or heat nociceptors A-MN Warm A-MN

22° C 17° C

S S

C-cooling C-cold?

(S)ML SML S L

C PM; A-MN, LTMs LTM RA Mech (in C. elegans) Nociceptor LTM

199,238 297 6 6,253 6,252 196 220 220

Noxious Mech, stretch Menthol

Mech? LTM? Mech? Mech? LTM?

DRG Neurons

Low pH Low pH Low pH

Heat ?

ATP ATP

49 6,33,239,290 306,307

A  A fiber; C  C fiber; capn  capsaicin; C PM  C-fiber polymodal nociceptor; L  large; LTM  low-threshold mechanoreeptor; M  medium; Mech  mechanoreceptive; MN  mechanoreceptor; NF  neurofilament; RA  rapidly adapting; S  small; ?  speculative. *Stimulus type refers to the type of stimulus that can activate the molecule, usually in a heterologous expression system. Note that, under “Thermal,” thermal thresholds given are for channels in such expression systems and thresholds may differ in intact DRG neurons, as appears likely for the menthol-sensitive response to cooling (presumably TRPM8) in intact DRG neurons, which has a threshold of about 30° C. An added caveat is that a measured thermal threshold may depend on the previous holding temperature. † Refers to afferent properties of neurons (suggested in the references cited in the last column) that might express these molecules, although in no case has this been directly determined.

neurons and/or in specific heat nociceptors. TRPV2 (VRL-1) is present in medium-sized and large neurons and responds to temperatures greater than 52° C; it was suggested that this may be responsible for the higher thermal threshold of A-fiber mechanoheat nociceptors.49 More recently, two receptors that respond to warm, rather than noxious, heat have been cloned. These are again TRP molecules and are called TRPV3239,290,343 and TRPV4.307 Interestingly these are not only expressed in DRG neurons and fibers but are also reported to be in epidermal keratinocytes, raising the possibility that keratinocytes might contribute in some way to activation of these receptors. TRPV3 shows co-localization with TRPV1 in DRG neurons, with which it can form heteromeric vanilloid channels.290 TRPV4 is expressed in about 10% of lumbar DRG neurons; many (80%) of these are NF rich, and thus presumably myelinated (see earlier), although interestingly most were small.307 Cold Transduction Two cooling/cold receptors have recently been cloned. These are the TRPM8 (CMR1) cold and menthol-sensitive channel and the TRPA1 (ANKTM1) cold-sensitive and menthol-insensitive channel. TRPM8 is present in small DRG neurons,199,238,257 and has a threshold of 22° to 24° C in heterologous expression systems, although in native DRG neurons a channel with properties similar to TRPM8 activates around 30° C (see Reid and colleagues257,259), a threshold consistent with that of low-threshold cooling C

fibers in the human.40 The TRPA1 receptor has a colder threshold for activation (17° C) and is present in a separate population of small DRG neurons.297 DRG neurons that respond to cooling or cold in vivo include specific C-cooling or cold units, C-mechanocold nociceptors, some C-polymodal nociceptors, and several groups of LTMs including C-mechanoreceptors, D hair units, and some A-fiber LTM units (e.g., SA II units). The small sizes of DRG neurons that express TRPM8 or TRPA1 may be indicative of these channels being expressed in some of the above-mentioned C-fiber units, but deduction of the likely sensory receptive properties that such neurons might possess is premature, although these channels seem unlikely to account for cold responses in the larger LTM units. Overall the proportion of neurons activated by cold (see, e.g., Simone and Kajander288) exceeds the proportions expressing either TRPM8 (5% to 10%238) or TRPA1 (4%297). Thus these two channels seem unlikely to account for all cooling or cold transduction in DRG neurons. In addition to the above-noted thermal receptor molecules, inhibition of a background K conductance by cooling was shown to contribute to cold transduction in rat DRG neurons in vitro.258,328 One candidate is TREK1, a two-pore domain background K channel present in DRG neurons (as well as many CNS neurons), that is closed by cooling.193 In addition, the hyperpolarization-activated nonselective cation current Ih328 (see later) may be inhibited by cooling and this could possibly moderate (reduce) neuronal responses to cooling;

188

Function of the Peripheral Nervous System

interestingly this Ih is present in some medium- to largediameter DRG neurons.53

Mechanoreceptive Molecules Several receptors/channels have been implicated in mechanotransduction to date, but this area is less well understood than that of thermal transduction, and transduction mechanisms in different types of neurons are not fully established. There is a group of degenerin/epithelial Na channel (DEG/ENaC) voltage-insensitive sodium channels (mammalian counterparts are the acid-sensing ion channels [ASICs]) some of which (e.g., MEC4, MEC10, and the stomatin-related MEC2) have a role in mechanotransduction in Caenorhabditis elegans. A model of a mechanically gated ion channel that is linked to the cytoskeleton via MEC2 as well as to the extracellular matrix has been proposed in which movement of the membrane causes the channel to open.196 The MEC2 homologue stomatin is widely expressed in DRG neurons, consistent with a possible role in mechanotransduction in many neurons.196 One of the DEG/ENaC family known variously as BNC1, ASIC2, MDEG, or BNaC1 is implicated in mechanotransduction in mammals and is expressed more strongly in large than small DRG neurons: transgenic BNC1 / mice showed reduced sensitivity especially of RA LTM units, consistent with the localization of this protein in lanceolate endings of hair follicle afferents in skin.252 Another of the DEG/ENaC channel group, DRASIC/ASIC3, is thought to be expressed only in primary afferent neurons, and responds to low pH; DRASIC knockout mice show defects in behavioral responses to noxious mechanical stimuli, with decreased responses in mechanical nociceptors and increased responses in LTMs and decreased responses of nociceptors to acid and noxious heat,253 raising the possibility that it may contribute to mechanotransduction mechanisms. Purinergic receptors may also play a role in mechanical transduction. ATP, released through low-level mechanical distortion, can activate heterologously expressed P2Y1 receptors in oocytes, whereas a greater distortion may release enough ATP to activate heterologously expressed P2X3 receptors. P2Y1 receptors are present on large DRG neurons and P2X3 receptors are present on small DRG neurons220; it was therefore suggested that these receptors may contribute to mechanotransduction in LTM neurons (P2Y1) or in nociceptors (P2X3).220 The voltage-gated calcium channel Cav3.2 and the associated T-type calcium current have been suggested to play a role in mechanotransduction in D hair units, because their expression is reduced in mice that are D hair deficient as a result of knocking out the NT-4 gene.284 TRPV4 may have a role in mechanical transduction: there are TRPV4-positive fibers in the skin, and TRPV4

knockout mice show impairment of responses to noxious mechanical stimuli.306 The description of TRPV4 expression in about 10% of DRG neurons, mainly in small, NFrich neurons,307 is consistent with expression in some A-fiber mechanical nociceptors. In addition, BDNF has been implicated in the regulation of mechanosensitivity of low-threshold, SA mechanoreceptors.47

Chemoreceptive Molecules Although it is clear what is meant by chemoreceptor in the context of olfaction and taste, it is much less clear what the limitations of this term should be in the context of the somatosensory system. It is likely that there are fibers that detect, for example, glucose, oxygen, CO2, and other metabolite levels in the tissues, and thus act as homeostatic afferents. In addition, it is known that certain afferent fibers can respond to inflammatory mediators, chemicals released by damaged tissues (such as ATP and bradykinin), and low pH. Information from such afferents could contribute to the detection of inflammatory processes and monitoring of the state of tissue integrity. Many C-fiber nociceptors also respond to capsaicin by activating the capsaicin/noxious heat receptor TRPV1. P2X purinergic receptors respond to ATP,36 and several are expressed in DRG neurons. P2X3 is found only in sensory neurons, where it can form heteromultimers (P2X2/3) with P2X2. Homomeric P2X3 receptors are expressed more highly in smaller diameter DRG neurons, whereas P2X2/3 heteromeric receptors are found more in medium-sized neurons.36 Heterologously expressed P2X3 responds to levels of ATP released by high-intensity (noxious) mechanical distortion, and thus (as mentioned earlier) may play a role in transduction of noxious mechanical stimuli.220 P2Y1 purinergic receptors are expressed in NF-rich large-diameter DRG neurons.220 As mentioned earlier, when expressed in chick oocytes they may respond to ATP released as a result of mild mechanical distortion, and thus may modulate responses to low-intensity mechanical stimulation in large DRG neurons.220

Acid-Sensing Molecules Acid pH in the tissues can be detected by primary afferent neurons resulting in pain. Acid pH may result from lactic acid accumulation in muscle, or from damage to the tissues. Importantly, TRPV1 is activated by low pH.320 The ASICs (also known as DEG/ENaC voltage-insensitive sodium channels, mentioned earlier under Mechanoreceptive Molecules) are activated by extracellular protons. ASIC/ASIC1, BNC1/ASIC2, and DRASIC/ASIC3 are all expressed in DRG neurons—both large and small,6,252,253 but BNC1/ASIC2 is more strongly expressed in large neurons; heteromultimers between these channels give rise to different H-gated channels.21 Other channels

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

may also contribute to transduction of low pH/high proton levels. TREK1 is one such channel. TREK1 is a polymodal background K channel that is closed by stretch, PGE2, cyclic AMP, and intracellular acidosis and is activated by heat.193 For further information on the molecular identities of proton-sensitive channels, see Wood.335

ION CURRENTS AND CHANNELS Na⫹ Currents and Channels Voltage-gated sodium channels are essential for generation and conduction of action potentials. The currents are subdivided into TTX-sensitive (TTXS) and TTX-resistant (TTXR) groups. The  subunits Nav1.1, 1.2, 1.3, 1.6, and 1.7 are TTXS, and Nav1.5, 1.8, and 1.9 are TTXR. Three of the  subunit proteins, Nav1.8 and Nav1.9 (TTXR), and Nav1.7 (TTXS), are expressed more in small than large DRG neurons. Nav1.8 (SNS/PN3). In rats, Nav1.8 is expressed in small- to medium-sized DRG neurons,4 most of which are nociceptive.74 Nav1.8 gives rise to a TTXR Na current that is thought to contribute the majority of the inward Na current in action potentials of small DRG neurons and is thought to be responsible for the TTXR current at nociceptive receptor terminals.5,31 Its properties probably contribute to properties of nociceptive neurons, including its high activation threshold (about 35 mV), its slow kinetics giving rise to long-duration action potentials, its contribution to the large action potential overshoot, and its rapid repriming, which may enable firing even in depolarized fibers.4,5,74,261 Inflammation and Nerve Injury. Several inflammatory mediators can acutely (minutes to hours) decrease the activation threshold, and/or increase the kinetics or magnitude of the TTXR Na current (presumed Nav1.8 related)102 and may contribute to acute hypersensitivity at nerve terminals. In the longer term, Nav1.8 mRNA and TTXR Na current in small DRG neurons and in cutaneous fibers are upregulated during most studies of inflammation.229,312 In some models of neuropathic pain, TTXR current density and Nav1.8 mRNA and protein (see, e.g., Sleeper et al.289) are decreased. Nav1.9 (NaN/SNS2). Nav1.9 also gives rise to a TTXR current. Nav1.9-LI is exclusively in small DRG neurons,69,313 preferentially in those with IB4 binding87 and along C fibers and at nodes of Ranvier of thinly myelinated fibers.88 It gives rise to a persistent, depolarizing TTXR Na current in small DRG neurons as a result of its low activation threshold and ultra-slow inactivation,62 thus it is likely to influence membrane excitability. GDNF, but not NGF, upregulates Nav1.9 mRNA.87 Inflammation and Nerve Injury. Nav1.9 mRNA in DRG neurons is increased after inflammation,313 and after exogenous GDNF administration.87 Axotomy induces a decrease

189

in Nav1.9 mRNA and protein in the DRG87,289,313 that is reversed in IB4-positive neurons by exogenous GDNF.87 Nav1.7 (PN1). Nav1.7 is more highly expressed in small than large DRG neurons76 despite the mRNA being in neurons of all sizes.25,76 It is thought to carry much of the TTXS inward current in action potentials in small neurons.63 Nav1.7 protein is in fibers and terminals of cultured DRG neurons.319 Its slow inactivation, combined with its low activation threshold,63 may be important in generation of receptor potentials and contributing to generation of action potentials. NGF causes a long-lasting (weeks) increase in Nav1.7 protein in DRG neurons in vivo.107 Other ␣ Subunits. mRNAs of several other TTXS  subunits are more abundant in medium and large neurons. These include Nav1.1, 1.2, and 1.6 and Nav2.2 (NaG).25,269 Following axotomy, the appearance of a more rapidly repriming TTXS current thought to contribute to hyperexcitability in axotomized small DRG neurons in vitro24,64 has been ascribed to increased expression of Nav1.3 (also known as III or brain type III), but roles for other TTXS channels have not been ruled out. Afferent/Sensory Receptive Properties of Neurons Expressing Nav1.7, Nav1.8, and Nav1.9. After intracellular recording, identification of sensory properties, dye injection, and immunocytochemistry, Nav1.9 was found to be expressed in nociceptive but not LTM neurons, and Nav1.8 was expressed mainly in nociceptive neurons, although some LTMs were also weakly positive.74 Despite being expressed mainly in small- to medium-sized neurons, Nav1.7 was expressed in both nociceptive and LTM neurons, with a slightly higher mean intensity in nociceptive neurons; its expression was negatively correlated with conduction velocity.76 Naⴙ Channel  Subunits. Na channel  subunits may interact with cytoskeleton or extracellular matrix and play roles in Na channel trafficking within cells and insertion into membrane. When coexpressed with  subunits,  subunits can alter the kinetics, peak current, and/or voltage dependencies of  subunits.131 1 (Na1.1) mRNA is found in medium to large DRG neurons, and a splice variant (1A) is also in DRG neurons.131 2 (Na2.1) mRNA is found in medium to large DRG neurons.25 3 (Na3) mRNA is normally found in small DRG neurons and increases in these neurons in the CCI model of neuropathic pain.279

K⫹ Currents and Channels K channels are central to the control of membrane potential, afterhyperpolarization, and firing frequency, and they influence adaptation. They tend to increase membrane

190

Function of the Peripheral Nervous System

potential stability, and channels that contribute to longduration afterhyperpolarizations may prove to be more highly expressed in nociceptors. Voltage-Gated K⫹ Currents and Kv Channels The two main groups of calcium-insensitive voltage-gated K currents are the depolarization-activated delayed rectifier (IKv) and fast transient (IA) currents. The protein  subunits of the channels that underlie these currents are the Kv subunits. Delayed Rectifier Currents. Delayed rectifier currents (IDR; also called IKv) serve to terminate the action potential rapidly in the soma and inhibit repetitive firing in myelinated axons.9 They are especially prominent in some large cutaneous afferent neurons,84 but are also present in small DRG neurons.103 Information on dendrodotoxin sensitivity is presented in the next section. Fast Transient (A-type) K ⴙ Currents. Fast transient K (IA) currents tend to clamp resting potential at hyperpolarized voltages until they inactivate, thus prolonging the afterhyperpolarization and slowing repetitive discharges.60 IA currents are present in both large cutaneous afferents84 and small DRG neurons103 but are more prominent in slowly conducting afferents.329 IA can be subdivided into rapidly (fast IA) and slowly (slow IA) inactivating types.345 Slow IA is particularly prominent in small DRG neurons with TTXR action potentials (see Yang et al.345), and may therefore contribute to the broad afterhyperpolarizations of nociceptive neurons. In DRG neurons IKv and slow IA are sensitive to dendrotoxin, but fast IA is not.345 Kv1.1 and 1.2 are associated with the delayed rectifier, whereas Kv1.4 is associated with fast IA,255,293 and it has been suggested that Kv1.1/1.2 associated with Kv1.4 (dendrotoxin insensitive) may give rise to the slow IA in DRG neurons.255,345 Kv1.1, 1.2, 1.3, 1.4, 1.5, and 1.6 and Kv2.1 are all expressed in DRG neurons; of these, Kv1.1 and 1.2 are more abundant and more highly expressed in medium- to large-sized neurons, and Kv2.1 is also more highly expressed in medium to large neurons, whereas Kv1.4 is more highly expressed in small neurons.130,255,345 M Currents. M currents are voltage and time dependent, noninactivating, and activated at negative voltages (beginning at approximately 70 mV). Their inhibition by acetylcholine or other agents leads to increased neuronal excitability.32 M currents are associated with KCNQ2/3 and -5 subunits,236,264 members of the 6TM superfamily and are distantly related to Kv channels. M currents as well as KCNQ2, -3, and -5 have been detected in both small and large DRG neurons.236 Block of M current with linopirdine causes increased firing in response to current injection in small DRG neurons,236 indicating that the M current may normally act as a brake to firing in these neurons.

Background K ⴙ Channels. The greater part of the timeindependent resting conductance in a variety of neurons is contributed by background K channels. They are twopore domain, homo- or heterodimeric channels. Some can be inhibited by neurotransmitters212,311 or physical stimuli such as extracellular acidity (the TASK family, which may be involved in acid sensing by nociceptors256); activated by alkalinity (the TALK family), by mechanical stress (TRAAK, the TREK family), or by heat (TREK-1); or be targets for volatile anesthetics (TASK-1, TASK-2, TREK-1, THIK-1).237 Background K currents in DRG neurons play a role in cold transduction (see earlier).258,328 Some members of this family are highly expressed in DRGs in preference to brain (TWIK-2, TASK-1, TASK-2, TREK-1, and TRAAK),203 but at the time of writing nothing is known about expression of these channels in sensory neurons with different modalities, nor whether their expression changes after nerve injury or in inflammation in mammals. Inwardly Rectifying K ⴙ Channels. Inwardly rectifying K channel (KIR) current contributes to the resting membrane potential in a number of cell types, including DRG neurons, where KIR current is mostly in medium-size neurons.278 Little is known of KIR-related channel subunits in DRGs. Ca2ⴙ-Activated K ⴙ Currents. Ca2-activated K currents (IKCa) are of three types, related to BK (slow), IK, and SK channels (big, intermediate, and small conductance channels, respectively), all activated by raised intracellular Ca2; BK is also voltage dependent. Functionally BK is associated with afterhyperpolarizations that develop rapidly and decay in 10 to 100 ms, and SK with slower afterhyperpolarizations that may last for seconds and limit firing frequency (spike frequency adaptation).158,325 Either or both of these could therefore contribute to the long afterhyperpolarization in nociceptors, although this remains to be established. BK and SK currents are found in DRG neurons.3,104,275 BK currents were in two thirds of small DRG neurons. Immunoreactivity for SK1 and IK1 channels was found in many human and rat DRG neurons.26 A related channel, SLACK (sequence like a Ca2-activated K), is expressed in rat DRG neurons.22 SLACK has a conductance similar to that of BK, with which it can interact to generate an intermediate-conductance KCa channel (different from IK1). Its role in DRG neurons is not yet understood. Axotomy Changes in DRG neurons after axotomy suggest decreased K channel expression or activation. Decreases in IKIR and fast IA occur after axotomy in large cutaneous afferent neurons83 and in IKIR in small neurons.345 Reductions in expression of Kv1.1, 1.2, and 1.4 and Kv2.1 have been reported.130,255,345 Reverse transcription– polymerase chain reaction (RTPCR) showed increased KCNQ2, -3, and -5 in both large

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

and small DRG neurons, and enhancement of M currents with retigabine resulted in decreased electrophysiologic and behavioral changes in a model of neuropathic pain.236 Inflammation After acute (2- to 3-hour) inflammation, activation of KCNQ/M currents with retigabine resulted in animals putting increased weight on the inflamed foot.236

Hyperpolarization-Activated Currents and Channels The H current (Ih) is a hyperpolarization-activated, timeand voltage-dependent, nonselective cation current that when activated causes depolarization of the membrane, tending to reduce afterhyperpolarization duration, increase firing frequency, and decrease adaptation.233 An H current is prominent in most or all large DRG neurons but in fewer small neurons.278 The channels that give rise to Ih are made up of HCN protein subunits, four isoforms (HCN1 through -4) of which have been cloned (see Chaplan et al.53). In DRG neurons there is strong expression of HCN1 mRNA in all large- to medium-diameter and most small-diameter DRG neurons, lower expression of HCN2 mRNA in approximately 80% of large and approximately 60% of small neurons, and low or undetectable levels of HCN3 and -4. HCN1 through -3 proteins are concentrated at the membrane especially of large neurons.53 Nerve injury causes increased Ih in large-diameter neurons dissociated in vitro, and ZD 7288 (a specific Ih blocker) blocks ectopic discharge in axotomized A-afferent fibers.53

Ca2ⴙ Currents and Channels The inflection on the falling phase of some of the broader action potentials in DRG neurons is partly due to inward Ca2 current. Ca2 has roles as a second messenger, in transmitter release, and in inhibiting firing by activation of IKCa. Several voltage-gated Ca2 currents have been described in DRG neurons, including L (high voltage activated), T (low voltage activated), and N (intermediate properties)91 (for more on properties of these currents, see review by Catterall51). Their amplitudes differ in neurons of different sizes, with relatively large L-type and N-type and smaller T-type currents in small cells, larger T- but little L- and N-type currents in medium-sized neurons, and little T-type current in large neurons.277 T-type Ca2 channels (Cav3.2) were suggested to be necessary for the normal mechanosensitivity of A-fiber D hair LTMs because expression of this channel was reduced in NT-4 knockout mice that were devoid of D hairs.284 Additionally, L- and N- but not T-type currents cause substance P release from isolated DRG neurons.119 Ca2 currents can be modulated by a variety of agonists. For instance activation of DORs on cultured young rat DRG neurons reduced N-, L-, P-, and Q- but not R-type currents,2 and 5-HT inhibits Ca2

191

currents in small DRG neurons probably via 5-HT1A receptors.67 Limited reports of expression include the following channel subunits demonstrated both immunocytochemically and by in situ hybridization (current type associated with the subunit in parentheses): Cav2.1 (P/Q), Cav2.2 (N), Cav1.2 and Cav1.3 (L), and Cav2.3 (R).51,218,348 NERVE INJURY. The T-type current in medium-sized neurons, as well as total Ca2 currents, were decreased 10 days after CCI of the sciatic nerve.198 In addition, the 2-1 subunit is upregulated (mRNA and protein)191,221 after different types of nerve injury. This may be an important site of action of the analgesic gabapentin.

CHEMICAL PHENOTYPE OF NOCICEPTIVE AND LTM NEURONS Although there is much indirect evidence to connect a number of molecular types to neurons with particular sensory properties, only for a few molecules has this relationship been directly demonstrated, by intracellular recording and determination of sensory receptive properties in skin and subcutaneous tissues followed by intracellular dye injection that enables subsequent immunocytochemistry. To summarize, in rat or guinea pig DRGs, substance P170 and Nav1.985 were found only in nociceptive neurons. CGRP124,169 and Nav1.874 showed intense labeling only in nociceptors, with weak labeling in some cutaneous LTM neurons. Strong TrkA and IB486 immunoreactivity was seen only in nociceptive neurons, and weak labeling of both was seen in a few D hair units.86 C-fiber and A-fiber nociceptors were both labeled for TrkA and IB4. Only C-fiber nociceptors showed IB4 binding,86 although in another study99 3 of 10 A-fiber nociceptors were also labeled by IB4.

SUMMARY It is beginning to be possible to interpret some aspects of the chemical phenotype of primary afferent neurons in terms of their afferent/sensory functions. However, this is a complex field, and the dynamically changing expression of a multitude of factors both by the DRG neurons themselves and by other cells of the tissues in which they reside makes this a complex and fascinating field. A large variety and number of metabolites, cytokines, inflammatory mediators, trophic factors, neuromodulators and transmitters may be released into the close environment of these neurons. That these endogenous chemicals may profoundly influence the properties and expression patterns of these neurons, highlights the advantages of studying them in vivo, both normally and, perhaps more particularly, in their dynamically changing environment following nerve or tissue damage.

192

Function of the Peripheral Nervous System

REFERENCES 1. Acosta, C. G., Fabrega, A. R., Masco, D. H., and Lopez, H. S.: A sensory neuron subpopulation with unique sequential survival dependence on nerve growth factor and basic fibroblast growth factor during development. J. Neurosci. 21:8873, 2001. 2. Acosta, C. G., and Lopez, H. S.: Delta opioid receptor modulation of several voltage-dependent Ca(2) currents in rat sensory neurons. J. Neurosci. 19:8337, 1999. 3. Akins, P. T., and McCleskey, E. W.: Characterization of potassium currents in adult rat sensory neurons and modulation by opioids and cyclic AMP. Neuroscience 56:759, 1993. 4. Akopian, A. N., Sivilotti, L., and Wood, J. N.: A tetrodotoxin-resistant voltage-gated sodium channel expressed by sensory neurons. Nature 379:257, 1996. 5. Akopian A. N., Souslova, V., England, S., et al.: The tetrodotoxin-resistant sodium channel SNS has a specialized function in pain pathways. Nat. Neurosci. 2:541, 1999. 6. Alvarez de la Rosa, D., Zhang, P., Shao, D., et al.: Functional implications of the localization and activity of acid-sensitive channels in rat peripheral nervous system. Proc. Natl. Acad. Sci. U. S. A. 99:2326, 2002. 7. Anand, P.: Neurotrophic factors and their receptors in human sensory neuropathies. Prog. Brain Res. 146:477, 2004. 8. Averill, S., McMahon, S. B., Clary, D. O., et al.: Immunocytochemical localization of trkA receptors in chemically identified subgroups of adult rat sensory neurons. Eur. J. Neurosci. 7:1484, 1995. 9. Baker, M., Bostock, H., Grafe, P., and Martius, P.: Function and distribution of three types of rectifying channel in rat spinal root myelinated axons. J. Physiol. (Lond.) 383:45, 1987. 10. Bandtlow, C. E., Heumann, R., Schwab, M. E., and Thoenen, H.: Cellular localization of nerve growth factor synthesis by in situ hybridization. EMBO J. 6:891, 1987. 11. Bar, K. J., Natura, G., Telleria-Diaz, A., et al.: Changes in the effect of spinal prostaglandin E2 during inflammation: prostaglandin E (EP1-EP4) receptors in spinal nociceptive processing of input from the normal or inflamed knee joint. J. Neurosci. 24:642, 2004. 12. Barbacid, M.: The Trk family of neurotrophin receptors. J. Neurobiol. 25:1386, 1994. 13. Baudet, C., Mikaels, A., Westphal, H., et al.: Positive and negative interactions of GDNF, NTN and ART in developing sensory neuron subpopulations, and their collaboration with neurotrophins. Development 127:4335, 2000. 14. Beedle, A. M., McRory, J. E., Poirot, O., et al.: Agonistindependent modulation of N-type calcium channels by ORL1 receptors. Nat. Neurosci. 7:118, 2004. 15. Bennett, D. L., Averill, S., Clary, D. O., et al.: Postnatal changes in the expression of the trkA high-affinity NGF receptor in primary sensory neurons. Eur. J. Neurosci. 8:2204, 1996. 16. Bennett, D. L., Boucher, T. J., Armanini, M. P., et al.: The glial cell line-derived neurotrophic factor family receptor components are differentially regulated within sensory neurons after nerve injury. J. Neurosci. 20:427, 2000. 17. Bennett, D. L., Dmietrieva, N., Priestley, J. V., et al.: trkA, CGRP and IB4 expression in retrogradely labelled cuta-

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32. 33.

neous and visceral primary sensory neurons in the rat. Neurosci. Lett. 206:33, 1996. Bennett, D. L., Michael, G. J., Ramachandran, N., et al.: A distinct subgroup of small DRG cells express GDNF receptor components and GDNF is protective for these neurons after nerve injury. J. Neurosci. 18:3059, 1998. Bennett, G., al Rashed, S., Hoult, J. R., and Brain, S. D.: Nerve growth factor induced hyperalgesia in the rat hind paw is dependent on circulating neutrophils. Pain 77:315, 1998. Bennett, G. J., and Xie, Y. K.: A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain 33:87, 1988. Benson, C. J., Xie, J., Wemmie, J. A., et al.: Heteromultimers of DEG/ENaC subunits form H-gated channels in mouse sensory neurons. Proc. Natl. Acad. Sci. U. S. A. 99:2338, 2002. Bhattacharjee, A., Gan, L., and Kaczmarek, L. K.: Localization of the Slack potassium channel in the rat central nervous system. J. Comp. Neurol. 454:241, 2002. Birder, L. A., and Perl, E. R.: Expression of alpha2adrenergic receptors in rat primary afferent neurons after peripheral nerve injury or inflammation. J. Physiol.(Lond.) 515(Pt. 2):533, 1999. Black, J. A., Cummins, T. R., Plumpton, C., et al.: Upregulation of a silent sodium channel after peripheral, but not central, nerve injury in DRG neurons. J. Neurophysiol. 82:2776, 1999. Black, J. A., Dib-Hajj, S., McNabola, K., et al.: Spinal sensory neurons express multiple sodium channel alpha-subunit mRNAs. Brain Res. Mol. Brain Res. 43:117, 1996. Boettger, M. K., Till, S., Chen, M. X., et al.: Calcium-activated potassium channel SK1- and IK1-like immunoreactivity in injured human sensory neurons and its regulation by neurotrophic factors. Brain 125:252, 2002. Botchkarev, V. A., Metz, M., Botchkareva, N. V., et al.: Brain-derived neurotrophic factor, neurotrophin-3, and neurotrophin-4 act as “epitheliotrophins” in murine skin. Lab. Invest. 79:557, 1999. Brain, S. D.: Sensory neuropeptides: their role in inflammation and wound healing. Immunopharmacology 37:133, 1997. Brechenmacher, C., Larmet, Y., Feltz, P., and Rodeau, J. L.: Cultured rat sensory neurons express functional tachykinin receptor subtypes 1, 2 and 3. Neurosci. Lett. 241:159, 1998. Bridges, D., Rice, A. S., Egertova, M., et al.: Localisation of cannabinoid receptor 1 in rat dorsal root ganglion using in situ hybridisation and immunohistochemistry. Neuroscience 119:803, 2003. Brock, J. A., McLachlan, E. M., and Belmonte, C.: Tetrodotoxin-resistant impulses in single nociceptor nerve terminals in guinea-pig cornea. J Physiol (Lond) 512:211, 1998. Brown, D. A.: M currents. In Narahashi, T. (ed.): Ion Channels. New York, Plenum Press, p. 55, 1988. Burgess, P. R., and Perl, E. R.: Myelinated afferent fibres responding specifically to noxious stimulation of the skin. J. Physiol. (Lond.) 190:541, 1967.

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons 34. Burgess, P. R., and Perl, E. R.: Cutaneous mechanoreceptors and nociceptors. In Iggo, A. (ed.): Handbook of Sensory Physiology, Vol. 2. Somatosensory System. Berlin, SpringerVerlag, p. 29, 1973. 35. Burgess, P. R., Petit, D., and Warren, R. M.: Receptor types in cat hairy skin supplied by myelinated fibers. J. Neurophysiol. 31:833, 1968. 36. Burnstock, G.: P2X receptors in sensory neurons. Br. J. Anaesth. 84:476, 2000. 37. Cafferty, W. B., Gardiner, N. J., Gavazzi, I., et al.: Leukemia inhibitory factor determines the growth status of injured adult sensory neurons. J. Neurosci. 21:7161, 2001. 38. Calza, L., Pozza, M., Arletti, R., et al.: Long-lasting regulation of galanin, opioid, and other peptides in dorsal root ganglia and spinal cord during experimental polyarthritis. Exp. Neurol. 164:333, 2000. 39. Campenot, R. B., and MacInnis, B. L.: Retrograde transport of neurotrophins: fact and function. J. Neurobiol. 58:217, 2004. 40. Campero, M., Serra, J., Bostock, H., and Ochoa, J. L.: Slowly conducting afferents activated by innocuous low temperature in human skin. J. Physiol. (Lond.) 535:855, 2001. 41. Cardenas, L. M., Cardenas, C. G., and Scroggs, R. S.: 5HT increases excitability of nociceptor-like rat dorsal root ganglion neurons via cAMP-coupled TTX-resistant Na() channels. J. Neurophysiol. 86:241, 2001. 42. Carlton, S. M.: Peripheral excitatory amino acids. Curr. Opin. Pharmacol. 1:52–56, 2001. 43. Carlton, S. M., and Coggeshall, R. E.: Inflammation-induced up-regulation of neurokinin 1 receptors in rat glabrous skin. Neurosci. Lett. 326:29, 2002. 44. Carlton, S. M., Du, J., Davidson, E., et al.: Somatostatin receptors on peripheral primary afferent terminals: inhibition of sensitized nociceptors. Pain 90:233, 2001. 45. Carlton, S. M., Zhou, S., and Coggeshall, R. E.: Localization and activation of substance P receptors in unmyelinated axons of rat glabrous skin. Brain Res. 734:103, 1996. 46. Carr, P. A., and Nagy, J. I.: Emerging relationships between cytochemical properties and sensory modality transmission in primary sensory neurons. Brain Res. Bull. 30:209, 1993. 47. Carroll, P., Lewin, G. R., Koltzenburg, M., et al.: A role for BDNF in mechanosensation. Nat. Neurosci. 1:42, 1998. 48. Casaccia-Bonnefil, P., Gu, C., Khursigara, G., and Chao, M. V.: P75 neurotrophin receptor as a modulator of survival and death decisions. Microsc. Res. Tech. 45:217, 1999. 49. Caterina, M. J., Rosen, T. A., Tominaga, M., et al.: A capsaicin-receptor homologue with a high threshold for noxious heat. Nature 398:436, 1999. 50. Caterina, M. J., Schumacher, M. A., Tominaga, M., et al.: The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389:816, 1997. 51. Catterall, W. A.: Structure and regulation of voltage-gated Ca2 channels. Annu. Rev. Cell Dev. Biol. 16:521, 2000. 52. Cervero, F.: Sensory innervation of the viscera: peripheral basis of visceral pain. Physiol. Rev. 74:95, 1994. 53. Chaplan, S. R., Guo, H. Q., Lee, D. H., et al.: Neuronal hyperpolarization-activated pacemaker channels drive neuropathic pain. J. Neurosci. 23:1169, 2003.

193

54. Chen, C., Zhou, X. F., and Rush, R. A.: Neurotrophin-3 and trkC-immunoreactive neurons in rat dorsal root ganglia correlate by distribution and morphology. Neurochem. Res. 21:809, 1996. 55. Cho, H. J., Kim, J. K., Park, H. C., et al.: Changes in brainderived neurotrophic factor immunoreactivity in rat dorsal root ganglia, spinal cord, and gracile nuclei following cut or crush injuries. Exp. Neurol. 154:224, 1998. 56. Cho, H. J., Kim, J. K., Zhou, X. F., and Rush, R. A.: Increased brain-derived neurotrophic factor immunoreactivity in rat dorsal root ganglia and spinal cord following peripheral inflammation. Brain Res. 764:269, 1997. 57. Cho, H. J., Kim, S. Y., Park, M. J., et al.: Expression of mRNA for brain-derived neurotrophic factor in the dorsal root ganglion following peripheral inflammation. Brain Res. 749:358, 1997. 58. Chopra, B., Giblett, S., Little, J. G., et al.: Cyclooxygenase-1 is a marker for a subpopulation of putative nociceptive neurons in rat dorsal root ganglia. Eur. J. Neurosci. 12:911, 2000. 59. Coggeshall, R. E., Zhou, S., and Carlton, S. M.: Opioid receptors on peripheral sensory axons. Brain Res. 764:126, 1997. 60. Connor, J. A., and Stevens, C. F.: Voltage clamp studies of a transient outward membrane current in gastropod neural somata. J. Physiol. (Lond.) 213:21, 1971. 61. Copray, J. C., Mantingh, I., Brouwer, N., et al.: Expression of interleukin-1 beta in rat dorsal root ganglia. J. Neuroimmunol. 118:203, 2001. 62. Cummins, T. R., Dib-Hajj, S. D., Black, J. A., et al.: A novel persistent tetrodotoxin-resistant sodium current In SNS-null and wild-type small primary sensory neurons. J. Neurosci. 19:RC43, 1999. 63. Cummins, T. R., Howe, J. R., and Waxman, S. G.: Slow closed-state inactivation: a novel mechanism underlying ramp currents in cells expressing the hNE/PN1 sodium channel. J. Neurosci. 18:9607, 1998. 64. Cummins, T. R., and Waxman, S. G.: Downregulation of tetrodotoxin-resistant sodium currents and upregulation of a rapidly repriming tetrodotoxin-sensitive sodium current in small spinal sensory neurons after nerve injury. J. Neurosci. 17:3503, 1997. 65. Davis, K. D., and Pope, G. E.: Noxious cold evokes multiple sensations with distinct time courses. Pain 98:179, 2002. 66. De Biasi, S., and Rustioni, A.: Glutamate and substance P coexist in primary afferent terminals in the superficial laminae of spinal cord. Proc. Natl. Acad. Sci. U. S. A. 85:7820, 1988. 67. Del Mar, L. P., Cardenas, C. G., and Scroggs, R. S.: Serotonin inhibits high-threshold Ca2 channel currents in capsaicin-sensitive acutely isolated adult rat DRG neurons. J. Neurophysiol. 72:2551, 1994. 68. Diamond, J., Foerster, A., Holmes, M., and Coughlin, M.: Sensory nerves in adult rats regenerate and restore sensory function to the skin independently of endogenous NGF. J. Neurosci. 12:1467, 1992. 69. Dib-Hajj, S. D., Tyrrell, L., Black, J. A., and Waxman, S. G.: NaN, a novel voltage-gated Na channel, is expressed preferentially in peripheral sensory neurons and downregulated after axotomy. Proc. Natl. Acad. Sci. U. S. A. 95:8963, 1998.

194

Function of the Peripheral Nervous System

70. Dickinson, T., and Fleetwood-Walker, S. M.: VIP and PACAP: very important in pain? Trends Pharmacol. Sci. 20:324, 1999. 71. Djouhri, L., Bleazard, L., and Lawson, S. N.: Association of somatic action potential shape with sensory receptive properties in guinea pig dorsal root ganglion neurons. J. Physiol. (Lond.) 513:857, 1998. 72. Djouhri, L., Dawbarn, D., Robertson, A., et al.: Time course and nerve growth factor dependence of inflammationinduced alterations in electrophysiological membrane properties in nociceptive primary afferent neurons. J. Neurosci. 21:8722, 2001. 73. Djouhri, L., and Lawson, S. N.: Abeta-fiber nociceptive primary afferent neurons: a review of incidence and properties in relation to other afferent A-fiber neurons in mammals. Brain Res. Brain Res. Rev. 46:131, 2004. 74. Djouhri, L., Fang, X., Okuse, K., et al.: The TTX-resistant sodium channel Nav1.8 (SNS/PN3): expression and correlation with membrane properties in rat nociceptive primary afferent neurons. J. Physiol. (Lond.) 550:739, 2003. 75. Djouhri, L., and Lawson, S. N.: Differences in the size of the somatic action potential overshoot between nociceptive and non-nociceptive dorsal root ganglion neurons in the guinea-pig. Neuroscience 108:479, 2001. 76. Djouhri, L., Newton, R., Levinson, S. R., et al.: Sensory and electrophysiological properties of guinea-pig sensory neurons expressing Na(v)1.7 (PN1) Na() channel alpha subunit protein. J. Physiol. (Lond.) 546:565, 2003. 77. Dobretsov, M., Hastings, S. L., Sims, T. J., et al.: Stretch receptor-associated expression of alpha(3) isoform of the Na(),K()-ATPase in rat peripheral nervous system. Neuroscience 116:1069, 2003. 78. Dodd, J., and Jessell, T. M.: Lactoseries carbohydrates specify subsets of dorsal root ganglion neurons projecting to the superficial dorsal horn of rat spinal cord. J. Neurosci. 12:3278, 1985. 79. Dowsing, B. J., Romeo, R., and Morrison, W. A.: Expression of leukemia inhibitory factor in human nerve following injury. J. Neurotrauma 18:1279, 2001. 80. Duce, I., and Keen, P.: An ultrastructural classification of the neuronal cell bodies of the rat dorsal root ganglion using zinc iodide-osmium impregnation. Cell Tiss. Res. 185:263, 1977. 81. Duggan, A. W., Hope, P. J., and Lang, C. W.: Microinjection of neuropeptide Y into the superficial dorsal horn reduces stimulus-evoked release of immunoreactive substance P in the anaesthetized cat. Neuroscience 44:733, 1991. 82. Ernfors, P., Lee, K. F., Kucera, J., and Jaenisch, R.: Lack of neurotrophin-3 leads to deficiencies in the peripheral nervous system and loss of limb proprioceptive afferents. Cell 77:503, 1994. 83. Everill, B., and Kocsis, J. D.: Reduction in potassium currents in identified cutaneous afferent dorsal root ganglion neurons after axotomy. J. Neurophysiol. 82:700, 1999. 84. Everill, B., Rizzo, M. A., and Kocsis, J. D.: Morphologically identified cutaneous afferent DRG neurons express three different potassium currents in varying proportions. J. Neurophysiol. 79:1814, 1998. 85. Fang, X., Djouhri, L., Black, J. A., et al.: The presence and role of the tetrodotoxin-resistant sodium channel Na(v)1.9 (NaN) in nociceptive primary afferent neurons. J. Neurosci. 22:7425, 2002.

86. Fang, X., Djouhri, L., and Lawson, S. N.: TrkA expression and IB4 binding in functionally identified dorsal root ganglion (DRG) nociceptive neurons in rats in vivo. J. Physiol. (Lond.) 536P:36P, 2001. 87. Fjell, J., Cummins, T. R., Dib-Hajj, S. D., et al.: Differential role of GDNF and NGF in the maintenance of two TTXresistant sodium channels in adult DRG neurons. Brain Res. Mol. Brain Res. 67:267, 1999. 88. Fjell, J., Hjelmstrom, P., Hormuzdiar, W., et al.: Localization of the tetrodotoxin-resistant sodium channel NaN in nociceptors. Neuroreport 11:199, 2000. 89. Foster, E., Robertson, B., and Fried, K.: trkB-like immunoreactivity in rat dorsal root ganglia following sciatic nerve injury. Brain Res. 659:267, 1994. 90. Fox, A., Wotherspoon, G., McNair, K., et al.: Regulation and function of spinal and peripheral neuronal B1 bradykinin receptors in inflammatory mechanical hyperalgesia. Pain 104:683, 2003. 91. Fox, A. P., Nowycky, M. C., and Tsien, R. W.: Kinetic and pharmacological properties distinguishing three types of calcium currents in chick sensory neurons. J. Physiol. (Lond.) 394:149, 1987. 92. Fox, E. A., and Powley, T. L.: False-positive artifacts of tracer strategies distort autonomic connectivity maps. Brain Res. Rev. 14:53, 1989. 93. Fukuoka, T., Kondo, E., Dai, Y., et al.: Brain-derived neurotrophic factor increases in the uninjured dorsal root ganglion neurons in selective spinal nerve ligation model. J. Neurosci. 21:4891, 2001. 94. Fundin, B. T., Silos-Santiago, I., Ernfors, P., et al.: Differential dependency of cutaneous mechanoreceptors on neurotrophins, trk receptors, and P75 LNGFR. Dev. Biol. 190:94, 1997. 95. Gardell, L. R., Wang, R., Ehrenfels, C., et al.: Multiple actions of systemic artemin in experimental neuropathy. Nat. Med. 9:1383, 2003. 96. Gardiner, N. J., Cafferty, W. B., Slack, S. E., and Thompson, S. W.: Expression of gp130 and leukaemia inhibitory factor receptor subunits in adult rat sensory neurons: regulation by nerve injury. J. Neurochem. 83:100, 2002. 97. Gee, M. D., Lynn, B., Basile, S., et al.: The relationship between axonal spike shape and functional modality in cutaneous C-fibres in the pig and rat. Neuroscience 90:509, 1999. 98. Genzen, J. R., Van Cleve, W., and McGehee, D. S.: Dorsal root ganglion neurons express multiple nicotinic acetylcholine receptor subtypes. J. Neurophysiol. 86:1773, 2001. 99. Gerke, M. B., and Plenderleith, M. B.: Binding sites for the plant lectin Bandeiraea simplicifolia I-isolectin B(4) are expressed by nociceptive primary sensory neurons. Brain Res. 911:101, 2001. 100. Gerke, M. B., and Plenderleith, M. B.: Analysis of the distribution of binding sites for the plant lectin Bandeiraea simplicifolia I-isolectin B4 on primary sensory neurons in seven mammalian species. Anat. Rec. 268:105, 2002. 101. Gokin, A. P., Fareed, M. U., Pan, H. L., et al.: Local injection of endothelin-1 produces pain-like behavior and excitation of nociceptors in rats. J. Neurosci. 21:5358, 2001. 102. Gold, M. S., Reichling, D. B., Shuster, M. J., and Levine, J. D.: Hyperalgesic agents increase a tetrodotoxin-resistant

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

103.

104.

105a.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116. 117.

118.

119.

Na current in nociceptors. Proc. Natl. Acad. Sci. U. S. A. 93:1108, 1996. Gold, M. S., Shuster, M. J., and Levine, J. D.: Characterization of six voltage-gated K currents in adult rat sensory neurons. J. Neurophysiol. 75:2629, 1996. Gold, M. S., Shuster, M. J., and Levine, J. D.: Role of a Ca(2)-dependent slow afterhyperpolarization in prostaglandin E2-induced sensitization of cultured rat sensory neurons. Neurosci. Lett. 205:161, 1996. Golden, J. P., DeMaro, J. A., Osborne, P. A., et al.: Expression of neurturin, GDNF, and GDNF family-receptor mRNA in the developing and mature mouse. Exp. Neurol. 158:504, 1999. Görke, K., and Pierau, F. K.: Spike potentials and membrane properties of dorsal root ganglion cells in pigeons. Pflügers Arch. Eur. J. Physiol. 386:21, 1980. Gould, H. J. III, Gould, T. N., England, J. D., et al.: A possible role for nerve growth factor in the augmentation of sodium channels in models of chronic pain. Brain Res. 854:19, 2000. Gould, H. J. III, Gould, T. N., Paul, D., et al.: Development of inflammatory hypersensitivity and augmentation of sodium channels in rat dorsal root ganglia. Brain Res. 824:296, 1999. Govrin-Lippmann, R., and Devor, M.: Ongoing activity in severed nerves: source and variation with time. Brain Res. 159:406, 1978. Gracely, R. H., Lynch, S. A., and Bennett, G. J.: Painful neuropathy: altered central processing maintained dynamically by peripheral input. Pain 51:175, 1992. Gratto, K. A., and Verge, V. M.: Neurotrophin-3 downregulates trkA mRNA, NGF high-affinity binding sites, and associated phenotype in adult DRG neurons. Eur. J. Neurosci. 18:1535, 2003. Gray, J. A. B., and Matthew, P. B. C.: A comparison of the adaptation of the pacinian corpuscle with the accommodation of its own axon. J. Physiol. (Lond.) 114:454, 1951. Grewe, M., Vogelsang, K., Ruzicka, T., et al.: Neurotrophin-4 production by human epidermal keratinocytes: increased expression in atopic dermatitis. J. Invest. Dermatol. 114:1108, 2000. Grothe, C., and Nikkhah, G.: The role of basic fibroblast growth factor in peripheral nerve regeneration. Anat. Embryol. (Berl.) 204:171, 2001. Ha, S. O., Kim, J. K., Hong, H. S., et al.: Expression of brain-derived neurotrophic factor in rat dorsal root ganglia, spinal cord and gracile nuclei in experimental models of neuropathic pain. Neuroscience 107:301, 2001. Halata, Z.: Sensory innervation of the hairy skin: light- and electronmicroscopic study. J. Invest. Dermatol. 101:75S, 1993. Halata, Z., Grim, M., and Bauman, K. I.: Friedrich Sigmund Merkel and his “Merkel cell,” morphology, development, and physiology: review and new results. Anat. Rec. 271A:225, 2003. Hamburger, V., and Levi-Montalcini, R.: Proliferation, differentiation and degeneration in the spinal ganglia of the chick embryo under normal and experimental conditions. J. Exp. Zool. 111:457, 1949. Harding, L. M., Beadle, D. J., and Bermudez, I.: Voltagedependent calcium channel subtypes controlling somatic

120.

121.

122.

123.

124.

125.

126. 127.

128.

129.

130.

131.

132.

133.

134.

135.

136.

195

substance P release in the peripheral nervous system. Prog. Neuropsychopharmacol. Biol. Psychiatry 23:1103, 1999. Harper, A. A.: Similarities between some properties of the soma and sensory receptors of primary afferent neurons. Exp. Physiol. 76:369, 1991. Harper, A. A., and Lawson, S. N.: Conduction velocity is related to morphological cell type in rat dorsal root ganglia. J. Physiol. (Lond.) 359:31, 1985. Harper, A. A., and Lawson, S. N.: Electrical properties of rat dorsal root ganglion neurons with different peripheral conduction velocities. J. Physiol. (Lond.) 359:47, 1985. Heumann, R., Lindholm, D., Bandtlow, C., et al.: Differential regulation of mRNA encoding nerve growth factor and its receptor in rat sciatic nerve during development, degeneration, and regeneration: role of macrophages. Proc. Natl. Acad. Sci. U. S. A. 84:8735, 1987. Hoheisel, U., Mense, S., and Scherotzke, R.: Calcitonin gene-related peptide-immunoreactivity in functionally identified primary afferent neurons in the rat. Anat. Embryol. (Berl.) 189:41, 1994. Hokfelt, T., Zhang, X., and Wiesenfeld-Hallin, Z.: Messenger plasticity in primary sensory neurons following axotomy and its functional implications. Trends Neurosci. 17:22, 1994. Holzer, P.: Neurogenic vasodilatation and plasma leakage in the skin. Gen. Pharmacol. 30:5, 1998. Horch, K. W., Tuckett, R. P., and Burgess, P. R.: A key to the classification of cutaneous mechanoreceptors. J. Invest. Dermatol. 69:75, 1977. Hu, P., and McLachlan, E. M.: Selective reactions of cutaneous and muscle afferent neurons to peripheral nerve transection in rats. J. Neurosci. 23:10559, 2003. Inoue, A., Ikoma, K., Morioka, N., et al.: Interleukin-1beta induces substance P release from primary afferent neurons through the cyclooxygenase-2 system. J. Neurochem. 73:2206, 1999. Ishikawa, K., Tanaka, M., Black, J. A., and Waxman, S. G.: Changes in expression of voltage-gated potassium channels in dorsal root ganglion neurons following axotomy. Muscle Nerve 22:502, 1999. Isom, L. L.: I. Cellular and molecular biology of sodium channel beta-subunits: therapeutic implications for pain? Am. J. Physiol. Gastrointest. Liver Physiol. 278:G349, 2000. Itoh, M., Takasaki, I., Andoh, T., et al.: Induction by carrageenan inflammation of prepronociceptin mRNA in VR1-immunoreactive neurons in rat dorsal root ganglia. Neurosci. Res. 40:227, 2001. Janig, W.: Neurobiology of visceral afferent neurons: neuroanatomy, functions, organ regulations and sensations. Biol. Psychol. 42:29, 1996. Ji, R. R., Zhang, Q., Law, P. Y., et al.: Expression of mu-, delta-, and kappa-opioid receptor-like immunoreactivities in rat dorsal root ganglia after carrageenan-induced inflammation. J. Neurosci. 15:8156, 1995. Ji, R. R., Zhang, Q., Zhang, X., et al.: Prominent expression of bFGF in dorsal root ganglia after axotomy. Eur. J. Neurosci. 7:2458, 1995. Josephson, A., Widenfalk, J., Trifunovski, A., et al.: GDNF and NGF family members and receptors in human fetal and adult spinal cord and dorsal root ganglia. J. Comp. Neurol. 440:204, 2001.

196

Function of the Peripheral Nervous System

137. Kajander, K. C., and Bennett, G. J.: Onset of a painful peripheral neuropathy in rat: a partial and differential deafferentation and spontaneous discharge in A beta and A delta primary afferent neurons. J. Neurophysiol. 68:734, 1992. 138. Kaplan, D. R., and Stephens, R. M.: Neurotrophin signal transduction by the Trk receptor. J. Neurobiol. 25:1404, 1994. 139. Karchewski, L. A., Gratto, K. A., Wetmore, C., and Verge, V. M.: Dynamic patterns of BDNF expression in injured sensory neurons: differential modulation by NGF and NT-3. Eur. J. Neurosci. 16:1449, 2002. 140. Karchewski, L. A., Kim, F. A., Johnston, J., et al: Anatomical evidence supporting the potential for modulation by multiple neurotrophins in the majority of adult lumbar sensory neurons. J. Comp. Neurol. 413:327, 1999. 141. Kashiba, H., Fukui, H., Morikawa, Y., and Senba, E.: Gene expression of histamine H1 receptor in guinea pig primary sensory neurons: a relationship between H1 receptor mRNA-expressing neurons and peptidergic neurons. Brain Res. Mol. Brain Res. 66:24, 1999. 142. Kashiba, H., Fukui, H., and Senba, E.: Histamine H1 receptor mRNA is expressed in capsaicin-insensitive sensory neurons with neuropeptide Y-immunoreactivity in guinea pigs. Brain Res. 901:85, 2001. 143. Kashiba, H., Noguchi, K., Ueda, Y., and Senba, E.: Coexpression of trk family members and low-affinity neurotrophin receptors in rat dorsal root ganglion neurons. Brain Res. Mol. Brain Res. 30:158, 1995. 144. Kashiba, H., and Senba, E.: Up- and down-regulation of BDNF mRNA in distinct subgroups of rat sensory neurons after axotomy. Neuroreport 10:3561, 1999. 145. Kashiba, H., Senba, E., Kawai, Y., et al.: Axonal blockade induces the expression of vasoactive intestinal polypeptide and galanin in rat dorsal root ganglion neurons. Brain Res. 577:19, 1992. 146. Kashiba, H., Uchida, Y., and Senba, E.: Difference in binding by isolectin B4 to trkA and c-ret mRNA-expressing neurons in rat sensory ganglia. Brain Res. Mol. Brain Res. 95:18, 2001. 147. Kashiba, H., Uchida, Y., and Senba, E.: Distribution and colocalization of NGF and GDNF family ligand receptor mRNAs in dorsal root and nodose ganglion neurons of adult rats. Brain Res. Mol. Brain Res. 110:52, 2003. 148. Kashiba, H., Ueda, Y., and Senba, E.: Coexpression of preprotachykinin-A, alpha-calcitonin gene-related peptide, somatostatin, and neurotrophin receptor family messenger RNAs in rat dorsal root ganglion neurons. Neuroscience 70:179, 1996. 149. Kashiba, H., Ueda, Y., and Senba, E.: Systemic capsaicin in the adult rat differentially affects gene expression for neuropeptides and neurotrophin receptors in primary sensory neurons. Neuroscience 76:299, 1997. 150. Kenins, P.: Responses of single nerve fibres to capsaicin applied to the skin. Neurosci. Lett. 29:83, 1982. 151. Khasabova, I. A., Harding-Rose, C., Simone, D. A., and Seybold, V. S.: Differential effects of CB1 and opioid agonists on two populations of adult rat dorsal root ganglion neurons. J. Neurosci. 24:1744, 2004. 152. Kidd, B. L., and Urban, L. A.: Mechanisms of inflammatory pain. Br. J. Anaesth. 87:3, 2001.

153. Kim, S. H., and Chung, J. M.: An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 50:355, 1992. 154. Kim, Y. I., Na, H. S., Kim, S. H., et al.: Cell type-specific changes of the membrane properties of peripherallyaxotomized dorsal root ganglion neurons in a rat model of neuropathic pain. Neuroscience 86:301, 1998. 155. Klein, R., Silos-Santiago, I., Smeyne, R. J., et al.: Disruption of the neurotrophin-3 receptor gene trkC eliminates Ia muscle afferents and results in abnormal movements. Nature 368:249, 1994. 156. Koerber, H. R., Druzinsky, R. E., and Mendell, L. M.: Properties of somata of spinal dorsal root ganglion cells differ according to peripheral receptor innervated. J. Neurophysiol. 60:1584, 1988. 157. Koerber, H. R., and Mendell, L. M.: Functional heterogeneity of dorsal root ganglion cells. In Scott, S. A. (ed.): Sensory Neurons: Diversity, Development and Plasticity. New York, Oxford University Press, p. 77, 1992. 158. Kohler, M., Hirschberg, B., Bond, C. T., et al.: Smallconductance, calcium-activated potassium channels from mammalian brain. Science 273:1709, 1996. 159. Koltzenburg, M., Bennett, D. L., Shelton, D. L., and McMahon, S. B.: Neutralization of endogenous NGF prevents the sensitization of nociceptors supplying inflamed skin. Eur. J. Neurosci. 11:1698, 1999. 160. Kopp, U. C., Cicha, M. Z., Smith, L. A., et al.: Cyclooxygenase-2 involved in stimulation of renal mechanosensitive neurons. Hypertension 35:373, 2000. 161. Kurek, J. B., Austin, L., Cheema S. S., et al.: Up-regulation of leukaemia inhibitory factor and interleukin-6 in transected sciatic nerve and muscle following denervation. Neuromuscul. Disord. 6:105, 1996. 162. Lang, E., Novak, A., Reeh, P. W., and Handwerker, H. O.: Chemosensitivity of fine afferents from rat skin in vitro. J. Neurophysiol. 63:887, 1990. 163. Lawman, M. J., Boyle, M. D., Gee, A. P., and Young, M.: Nerve growth factor accelerates the early cellular events associated with wound healing. Exp. Mol. Pathol. 43:274, 1985. 164. Lawson, S. N.: The postnatal development of large light and small dark neurons in mouse dorsal root ganglia: a statistical analysis of cell numbers and size. J. Neurocytol. 8:275, 1979. 165. Lawson, S. N.: Morphological and biochemical cell types of sensory neurons. In Scott, S. A. (ed.): Sensory Neurons: Diversity, Development and Plasticity. New York, Oxford University Press, p. 27, 1992. 166. Lawson, S. N.: Phenotype and function of somatic primary afferent nociceptive neurons with C-, Adelta or Aalpha/beta-fibres. J. Exp. Physiol. 87:239, 2002. 167. Lawson, S. N., and Biscoe, T. J.: Development of mouse dorsal root ganglia: an autoradiographic and quantitative study. J. Neurocytol. 8:265, 1979. 168. Lawson, S. N., Caddy, K. W. T., and Biscoe, T. J.: Development of rat dorsal root ganglion neurons: studies of cell birthdays and changes in mean cell diameter. Cell Tiss. Res. 153:399, 1974. 169. Lawson, S. N., Crepps, B., and Perl, E. R.: Calcitonin generelated peptide immunoreactivity and afferent receptive properties of dorsal root ganglion neurons in guinea-pigs. J. Physiol. (Lond.) 540:989, 2002.

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons 170. Lawson, S. N., Crepps, B. A., and Perl, E. R.: Relationship of substance P to afferent characteristics of dorsal root ganglion neurons in guinea-pig. J. Physiol. (Lond.) 505(Pt. 1):177, 1997. 171. Lawson, S. N., Harper, A. A., Harper, E. I., et al.: A monoclonal antibody against neurofilament protein specifically labels a subpopulation of rat sensory neurons. J. Comp. Neurol. 228:263, 1984. 172. Lawson, S. N., McCarthy, P. W., and Prabhakar, E.: Electrophysiological properties of neurons with CGRP-like immunoreactivity in rat dorsal root ganglia. J. Comp. Neurol. 365:355, 1996. 173. Lawson, S. N., and Waddell, P. J.: Soma neurofilament immunoreactivity is related to cell size and fibre conduction velocity in rat primary sensory neurons. J. Physiol. (Lond.) 435:41, 1991. 174. Le Douarin, N. M., Kalcheim, C., and Teillet, M. A.: The cellular and molecular basis of early sensory ganglion development. In Scott, S. A. (ed.): Sensory Neurons: Diversity, Development and Plasticity. New York, Oxford University Press, p. 143, 1992. 175. Lee, K. H., Chung, K., Chung, J. M., and Coggeshall, R. E.: Correlation of cell body size, axon size and signal conduction velocity for individually labelled dorsal root ganglion cells in the cat. J. Comp. Neurol. 243:335, 1986. 176. Lee, L. Y., and Pisarri, T. E.: Afferent properties and reflex functions of bronchopulmonary C-fibers. Respir. Physiol. 125:47, 2001. 177. Lee, Y. J., Zachrisson, O., Tonge, D. A., and McNaughton, P. A.: Upregulation of bradykinin B2 receptor expression by neurotrophic factors and nerve injury in mouse sensory neurons. Mol. Cell. Neurosci. 19:186, 2002. 178. Leslie, T. A., Emson, P. C., Dowd, P. M., and Woolf, C. J.: Nerve growth factor contributes to the up-regulation of growth-associated protein 43 and preprotachykinin A messenger RNAs in primary sensory neurons following peripheral inflammation. Neuroscience 67:753, 1995. 179. Lessmann, V., Gottmann, K., and Malcangio, M.: Neurotrophin secretion: current facts and future prospects. Prog. Neurobiol. 69:341, 2003. 180. Levy, D., and Zochodne, D. W.: Increased mRNA expression of the B1 and B2 bradykinin receptors and antinociceptive effects of their antagonists in an animal model of neuropathic pain. Pain 86:265, 2000. 181. Lewin, G. R., and Mendell, L. M.: Nerve growth factor and nociception. Trends Neurosci. 16:353, 1993. 182. Li, G. D., Wo, Y., Zhong, M. F., et al.: Expression of fibroblast growth factors in rat dorsal root ganglion neurons and regulation after peripheral nerve injury. Neuroreport 13:1903, 2002. 183. Li, L., Deng, Y. S., and Zhou, X. F.: Downregulation of TrkA expression in primary sensory neurons after unilateral lumbar spinal nerve transection and some rescuing effects of nerve growth factor infusion. Neurosci. Res. 38:183, 2000. 184. Li, W. P., Xian, C., Rush, R. A., and Zhou, X. F.: Upregulation of brain-derived neurotrophic factor and neuropeptide Y in the dorsal ascending sensory pathway following sciatic nerve injury in rat. Neurosci. Lett. 260:49, 1999. 185. Light, A. R., and Perl, E. R.: Peripheral sensory systems. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.):

186.

187.

188.

189.

190.

191.

192.

193.

194.

195.

196.

197.

198.

199.

200.

201.

202.

203.

197

Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 149, 1993. Lindholm, D., Heumann, R., Meyer, M., and Thoenen, H.: Interleukin-1 regulates synthesis of nerve growth factor in non-neuronal cells of rat sciatic nerve. Nature 330:658, 1987. Liu, H. X., and Hokfelt, T.: The participation of galanin in pain processing at the spinal level. Trends Pharmacol. Sci. 23:468, 2002. Liu, X., Eschenfelder, S., Blenk, K. H., et al.: Spontaneous activity of axotomized afferent neurons after L5 spinal nerve injury in rats. Pain 84:309, 2000. Lu, J., Zhou, X. F., and Rush, R. A.: Small primary sensory neurons innervating epidermis and viscera display differential phenotype in the adult rat. Neurosci. Res. 41:355, 2001. Luo, X. G., Rush, R. A., and Zhou, X. F.: Ultrastructural localization of brain-derived neurotrophic factor in rat primary sensory neurons. Neurosci. Res. 39:377, 2001. Luo, Z. D., Chaplan, S. R., Higuera, E. S., et al.: Upregulation of dorsal root ganglion (alpha)2(delta) calcium channel subunit and its correlation with allodynia in spinal nerve-injured rats. J. Neurosci. 21:1868, 2001. Ma, W., Chabot, J. G., Powell, K. J., et al.: Localization and modulation of calcitonin gene-related peptide-receptor component protein-immunoreactive cells in the rat central and peripheral nervous systems. Neuroscience 120:677, 2003. Maingret, F., Lauritzen, I., Patel, A. J., et al.: TREK-1 is a heat-activated background K() channel. EMBO J. 19:2483, 2000. Malcangio, M., Getting, S. J., Grist, J., et al.: A novel control mechanism based on GDNF modulation of somatostatin release from sensory neurons. FASEB J. 16:730, 2002. Mannion, R. J., Costigan, M., Decosterd, I., et al.: Neurotrophins: peripherally and centrally acting modulators of tactile stimulus-induced inflammatory pain hypersensitivity. Proc. Natl. Acad. Sci. U. S. A. 96:9385, 1999. Mannsfeldt, A. G., Carroll, P., Stucky, C. L., and Lewin, G. R.: Stomatin, a MEC-2 like protein, is expressed by mammalian sensory neurons. Mol. Cell. Neurosci. 13:391, 1999. Matzner, O., and Devor, M.: Hyperexcitability at sites of nerve injury depends on voltage-sensitive Na channels. J. Neurophysiol. 72:349, 1994. McCallum, J. B., Kwok, W. M., Mynlieff, M., et al.: Loss of T-type calcium current in sensory neurons of rats with neuropathic pain. Anesthesiology 98:209, 2003. McKemy, D. D., Neuhausser, W. M., and Julius, D.: Identification of a cold receptor reveals a general role for TRP channels in thermosensation. Nature 416:52, 2002. McLennan, A. J., Vinson, E. N., Marks, L., et al.: Immunohistochemical localization of ciliary neurotrophic factor receptor a expression in the rat nervous system. J. Neurosci. 16:621, 1996. McMahon, S. B., Armanini, M. P., Ling, L. H., and Phillips, H. S.: Expression and coexpression of Trk receptors in subpopulations of adult primary sensory neurons projecting to identified peripheral targets. Neuron 12:1161, 1994. Mearow, K. M.: The effects of NGF and sensory nerve stimulation on collateral sprouting and gene expression in adult sensory neurons. Exp. Neurol. 151:14, 1998. Medhurst, A. D., Rennie, G., Chapman, C. G., et al.: Distribution analysis of human two pore domain potassium

198

204.

205.

206.

207.

208.

209.

210.

211. 212.

213.

214.

215.

216.

217.

218.

Function of the Peripheral Nervous System channels in tissues of the central nervous system and periphery. Brain Res. Mol. Brain Res. 86:101, 2001. Mendell, L. M., Albers, K. M., and Davis, B. M.: Neurotrophins, nociceptors, and pain. Microsc. Res. Tech. 45:252, 1999. Mense, S., and Meyer, H.: Different types of slowly conducting afferent units in cat skeletal muscle and tendon. J. Physiol. (Lond.) 363:403, 1985. Meyer, R. A., Davis, K. D., Cohen, R. H., et al.: Mechanically insensitive afferents (MIAs) in cutaneous nerves of monkey. Brain Res. 561:252, 1991. Michael, G. J., Averill, S., Nitkunan, A., et al.: Nerve growth factor treatment increases brain-derived neurotrophic factor selectively in TrkA-expressing dorsal root ganglion cells and in their central terminations within the spinal cord. J. Neurosci. 17:8476, 1997. Michael, G. J., Averill, S., Shortland, P. J., et al.: Axotomy results in major changes in BDNF expression by dorsal root ganglion cells: BDNF expression in large trkB and trkC cells, in pericellular baskets, and in projections to deep dorsal horn and dorsal column nuclei. Eur. J. Neurosci. 11:3539, 1999. Michaelis, M., Liu, X., and Janig, W.: Axotomized and intact muscle afferents but no skin afferents develop ongoing discharges of dorsal root ganglion origin after peripheral nerve lesion. J. Neurosci. 20:2742, 2000. Mika, J., Li, Y., Weihe, E., and Schafer, M. K.: Relationship of pronociceptin/orphanin FQ and the nociceptin receptor ORL1 with substance P and calcitonin gene-related peptide expression in dorsal root ganglion of the rat. Neurosci. Lett. 348:190, 2003. Millan, M. J.: The induction of pain: an integrative review. Prog. Neurobiol. 57:1, 1999. Millar, J. A., Barratt, L., Southan, A. P., et al.: A functional role for the two-pore domain potassium channel TASK-1 in cerebellar granule neurons. Proc. Natl. Acad. Sci. U. S. A. 97:3614, 2000. Minami, M., Maekawa, K., Yabuuchi, K., and Satoh, M.: Double in situ hybridization study on coexistence of mu-, delta- and kappa-opioid receptor mRNAs with preprotachykinin A mRNA in the rat dorsal root ganglia. Brain Res. Mol. Brain. Res. 30:203, 1995. Molliver, D. C., Radeke, M. J., Feinstein, S. C., and Snider, W. D.: Presence or absence of TrkA protein distinguishes subsets of small sensory neurons with unique cytochemical characteristics and dorsal horn projections. J. Comp. Neurol. 361:404, 1995. Molliver, D. C., and Snider, W. D.: Nerve growth factor receptor TrkA is down-regulated during postnatal development by a subset of dorsal root ganglion neurons. J. Comp. Neurol. 381:428, 1997. Molliver, D. C., Wright, D. E., Leitner, M. L., et al.: IB4-binding DRG neurons switch from NGF to GDNF dependence in early postnatal life. Neuron 19:849, 1997. Mu, X., Silos-Santiago, I., Carroll, S. L., and Snider, W. D.: Neurotrophin receptor genes are expressed in distinct patterns in developing dorsal root ganglia. J. Neurosci. 13:4029, 1993. Murakami, M., Suzuki, T., Nakagawasai, O., et al.: Distribution of various calcium channel alpha(1) subunits in

219.

220.

221.

222.

223.

224.

225.

226.

227.

228.

229.

230.

231.

232.

233.

234.

235.

murine DRG neurons and antinociceptive effect of omegaconotoxin SVIB in mice. Brain Res. 903:231, 2001. Murphy, P. G., Borthwick, L. A., Altares, M., et al.: Reciprocal actions of interleukin-6 and brain-derived neurotrophic factor on rat and mouse primary sensory neurons. Eur. J. Neurosci. 12:1891, 2000. Nakamura, F., and Strittmatter, S. M.: P2Y1 purinergic receptors in sensory neurons: contribution to touch-induced impulse generation. Proc. Natl. Acad. Sci. U. S. A. 93:10465, 1996. Newton, R. A., Bingham, S., Case, P. C., et al.: Dorsal root ganglion neurons show increased expression of the calcium channel alpha2delta-1 subunit following partial sciatic nerve injury. Brain Res. Mol. Brain Res. 95:1, 2001. Nicholson, R., Small, J., Dixon, A. K., et al.: Serotonin receptor mRNA expression in rat dorsal root ganglion neurons. Neurosci. Lett. 337:119, 2003. Nielson, K. M., Chaverra, M., Hapner, S. J., et al.: PACAP promotes sensory neuron differentiation: blockade by neurotrophic factors. Mol. Cell. Neurosci. 25:629, 2004. Nishii, K., Oura, C., and Pallie, W.: Ultrastructure of the mature pacinian corpuscle in the mesentery of the cat. J. Anat. 106:208, 1970. Nystrom, B., and Hagbarth, K. E.: Microelectrode recordings from transected nerves in amputees with phantom limb pain. Neurosci. Lett. 27:211, 1981. Oakley, R. A., Lefcort, F. B., Clary, D. O., et al.: Neurotrophin-3 promotes the differentiation of muscle spindle afferents in the absence of peripheral targets. J. Neurosci. 17:4262, 1997. Obata, K., Yamanaka, H., Fukuoka, T., et al.: Contribution of injured and uninjured dorsal root ganglion neurons to pain behavior and the changes in gene expression following chronic constriction injury of the sciatic nerve in rats. Pain 101:65, 2003. Ohtori, S., Takahashi, K., Moriya, H., and Myers, R. R.: TNF-alpha and TNF-alpha receptor type 1 upregulation in glia and neurons after peripheral nerve injury: studies in murine DRG and spinal cord. Spine 29:1082, 2004. Okuse, K., Chaplan, S. R., McMahon, S. B., et al.: Regulation of expression of the sensory neuron-specific sodium channel SNS in inflammatory and neuropathic pain. Mol. Cell. Neurosci. 10:196, 1997. Okuse, K., Malik-Hall, M., Baker, M. D., et al.: Annexin II light chain regulates sensory neuron-specific sodium channel expression. Nature 417:653, 2002. Olausson, H., Lamarre, Y., Backlund, H., et al.: Unmyelinated tactile afferents signal touch and project to insular cortex. Nat. Neurosci. 5:900, 2002. Orozco, O. E., Walus, L., Sah, D. W., et al.: GFRalpha3 is expressed predominantly in nociceptive sensory neurons. Eur. J. Neurosci. 13:2177, 2001. Pape, H. C.: Queer current and pacemaker: the hyperpolarization-activated cation current in neurons. Annu. Rev. Physiol. 58:299, 1996. Pare, M., Behets, C., and Cornu, O.: Paucity of presumptive Ruffini corpuscles in the index finger pad of humans. J. Comp. Neurol. 456:260, 2003. Pare, M., Smith, A. M., and Rice, F. L.: Distribution and terminal arborizations of cutaneous mechanoreceptors in

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons

236.

237.

238.

239.

240.

241.

242.

243.

244.

245.

246.

247.

248.

249.

250.

251.

252.

253.

the glabrous finger pads of the monkey. J. Comp. Neurol. 445:347, 2002. Passmore, G. M., Selyanko, A. A., Mistry, M., et al.: KCNQ/M currents in sensory neurons: significance for pain therapy. J. Neurosci. 23:7227, 2003. Patel, A. J., and Honore, E.: Properties and modulation of mammalian 2P domain K channels. Trends Neurosci. 24:339, 2001. Peier, A. M., Moqrich, A., Hergarden, A. C., et al.: A TRP channel that senses cold stimuli and menthol. Cell 108:705, 2002. Peier, A. M., Reeve, A. J., Andersson, D. A., et al.: A heatsensitive TRP channel expressed in keratinocytes. Science 296:2046, 2002. Perl, E. R.: Myelinated afferent fibres innervating the primate skin and their response to noxious stimuli. J. Physiol. (Lond.) 197:593, 1968. Perl, E. R.: Function of dorsal root ganglion neurons: an overview. In Scott, S. A. (ed.): Sensory Neurons: Diversity, Development and Plasticity. New York, Oxford University Press, p. 3, 1992. Perl, E. R., and Burgess, P. R.: Classification of afferent dorsal root fibers: mammals. In Altman, P. L., and Dittmer, D. S. (eds.): Biology Data Book. Bethesda, MD, FASEB, p. 1141, 1973. Perry, M. J., and Lawson, S. N.: Neurofilament in feline primary afferent neurons: a quantitative immunocytochemical study. Brain Res. 607:307, 1993. Perry, M. J., and Lawson, S. N.: Differences in expression of oligosaccharides, neuropeptides, carbonic anhydrase and neurofilament in rat primary afferent neurons retrogradely labelled via skin, muscle or visceral nerves. Neuroscience 85:293, 1998. Perry, M. J., Lawson, S. N., and Robertson, J.: Neurofilament immunoreactivity in populations of rat primary afferent neurons: a quantitative study of phosphorylated and nonphosphorylated subunits. J. Neurocytol. 20:746, 1991. Pezet, S., Malcangio, M., and McMahon, S. B.: BDNF: a neuromodulator in nociceptive pathways? Brain Res. Brain Res. Rev. 40:240, 2002. Pezet, S., Onteniente, B., Jullien, J., et al.: Differential regulation of NGF receptors in primary sensory neurons by adjuvant-induced arthritis in the rat. Pain 90:113, 2001. Plenderleith, M. B., Cameron, A. A., Key, B., and Snow, P. J.: Soybean agglutinin binds to a subpopulation of primary afferent neurons in the cat. Neurosci. Lett. 86:257, 1988. Pomonis, J. D., Rogers, S. D., Peters, C. M., et al.: Expression and localization of endothelin receptors: implications for the involvement of peripheral glia in nociception. J. Neurosci. 21:999, 2001. Popratiloff, A., Weinberg, R. J., Rustioni, A.: AMPA receptors at primary afferent synapses in substantia gelatinosa after sciatic nerve section. Eur. J. Neurosci. 10:3220, 1998. Prabhakar, E., and Lawson, S. N.: The electrophysiological properties of rat primary afferent neurons with carbonic anhydrase activity. J. Physiol. (Lond.) 482:609, 1995. Price, M. P., Lewin, G. R., McIlwrath, S. L., et al.: The mammalian sodium channel BNC1 is required for normal touch sensation. Nature 407:1007, 2000. Price, M. P., McIlwrath, S. L., Xie, J., et al.: The DRASIC cation channel contributes to the detection of

254.

255.

256.

257.

258.

259.

260.

261.

262.

263.

264. 265.

266. 267. 268.

269.

269a.

270.

271.

199

cutaneous touch and acid stimuli in mice. Neuron 32:1071, 2001. Rambourg, A., Clermont, Y., and Beaudet, A.: Ultrastructural features of six types of neurons in rat dorsal root ganglia. J. Neurocytol. 12:47, 1983. Rasband, M. N., Park, E. W., Vanderah, T. W., et al.: Distinct potassium channels on pain-sensing neurons. Proc. Natl. Acad. Sci. U. S. A. 98:13373, 2001. Reeh, P. W., and Kress, M.: Molecular physiology of proton transduction in nociceptors. Curr. Opin. Pharmacol. 1:45, 2001. Reid, G., Babes, A., and Pluteanu, F.: A cold- and mentholactivated current in rat dorsal root ganglion neurons: properties and role in cold transduction. J. Physiol. (Lond.) 545:595, 2002. Reid, G., and Flonta, M.: Cold transduction by inhibition of a background potassium conductance in rat primary sensory neurons. Neurosci. Lett. 297:171, 2001. Reid, G., and Flonta, M. L.: Ion channels activated by cold and menthol in cultured rat dorsal root ganglion neurons. Neurosci. Lett. 324:164, 2002. Reinohl, J., Hoheisel, U., Unger, T., and Mense, S.: Adenosine triphosphate as a stimulant for nociceptive and non-nociceptive muscle group IV receptors in the rat. Neurosci. Lett. 338:25, 2003. Renganathan, M., Cummins, T. R., and Waxman, S. G.: Contribution of Na(v)1.8 sodium channels to action potential electrogenesis in DRG neurons. J. Neurophysiol. 86:629, 2001. Richardson, J. D., and Vasko, M. R.: Cellular mechanisms of neurogenic inflammation. J. Pharmacol. Exp. Ther. 302:839, 2002. Ritter, A. M., and Mendell, L. M.: Somal membrane properties of physiologically identified sensory neurons in the rat: effects of nerve growth factor. J. Neurophysiol. 68:2033, 1992. Robbins, J.: KCNQ potassium channels: physiology, pathophysiology, and pharmacology. Pharmacol. Ther. 90:1, 2001. Robertson, B., Perry, M., and Lawson, S. N.: Populations of rat spinal primary afferent neurons with choleragenoid binding compared with those labelled by markers for neurofilament and carbohydrate groups: a quantitative immunocytochemical study. J. Neurocytol. 20:387, 1991. Roux, P. P., and Barker, P. A.: Neurotrophin signalling through the p75 receptor. Prog. Neurobiol. 67:203, 2002. Ruscheweyh, R., and Sandkuhler, J.: Role of kainate receptors in nociception. Brain Res. Brain Res. Rev. 40:215, 2002. Safieh-Garabedian, B., Poole, S., Allchorne, A., et al.: Contribution of interleukin-1 beta to the inflammationinduced increase in nerve growth factor levels and inflammatory hyperalgesia. Br. J. Pharmacol. 115:1265, 1995. Sangameswaran, L., Fish, L. M., Koch, B. D., et al.: A novel tetrodotoxin-sensitive, voltage-gated sodium channel expressed in rat and human dorsal root ganglia. J. Biol. Chem. 272:14805, 1997. Sata, K., Kiyama, H., Park, H. T., et al: AMPA, KA and NMDA receptors are expressed in the rat DRG neurones. Neuroreport 4:1263–1265, 1993. Schafers, M., Geis, C., Brors, D., et al.: Anterograde transport of tumor necrosis factor-alpha in the intact and injured rat sciatic nerve. J. Neurosci. 22:536, 2002. Schafers, M., Forkin, L. S., Geis, C., and Shubayev, V. I.: Spinal nerve ligation induces transient upregulation of

200

272.

273. 274.

275.

276.

277.

278.

279.

280.

281.

282.

283.

284.

285. 286.

287.

Function of the Peripheral Nervous System tumor necrosis factor receptors 1 and 2 in injured and adjacent uninjured dorsal root ganglia in the rat. Neurosci. Lett. 347:179, 2003. Schmelz, M., Schmidt, R., Weidner, C., et al.: Chemical response pattern of different classes of C-nociceptors to pruritogens and algogens. J. Neurophysiol. 89:2441, 2003. Schmidt, R. F.: The articular polymodal nociceptor in health and disease. Prog. Brain Res. 113:53, 1996. Schoenen, J., Delree, P., Leprince, P., and Moonen, G.: Neurotransmitter phenotype plasticity in cultured dissociated adult rat dorsal root ganglia: an immunocytochemical study. J. Neurosci. Res. 22:473, 1989. Scholz, A., Gruss, M., and Vogel, W.: Properties and functions of calcium-activated K channels in small neurons of rat dorsal root ganglion studied in a thin slice preparation. J. Physiol. (Lond.) 513:55, 1998. Schulz, S., Schreff, M., Schmidt, H., et al.: Immunocytochemical localization of somatostatin receptor sst2A in the rat spinal cord and dorsal root ganglia. Eur. J. Neurosci. 10:3700, 1998. Scroggs, R. S., and Fox, A. P.: Multiple Ca2 currents elicited by action potential waveforms in acutely isolated adult rat dorsal root ganglion neurons. J. Neurosci. 12:1789, 1992. Scroggs, R. S., Todorovic, S. M., Anderson, E. G., and Fox, A. P.: Variation in I(H), I(IR), and I(LEAK) between acutely isolated adult rat dorsal root ganglion neurons of different size. J. Neurophysiol. 71:271, 1994. Shah, B. S., Stevens, E. B., Gonzalez, M. I., et al.: Beta3, a novel auxiliary subunit for the voltage-gated sodium channel, is expressed preferentially in sensory neurons and is upregulated in the chronic constriction injury model of neuropathic pain. Eur. J. Neurosci. 12:3985, 2000. Shamash, S., Reichert, F., and Rotshenker, S.: The cytokine network of wallerian degeneration: tumor necrosis factor-, interleukin-1, and interleukin-1. J. Neurosci. 22:3052, 2002. Sharp, G. A., Shaw, G., and Weber, K.: Immunoelectronmicroscopical localisation of the three neurofilament triplet proteins along neurofilaments of cultured dorsal root ganglion neurons. Exp. Cell Res. 137:403, 1982. Shen, H., Chung, J. M., Coggeshall, R. E., and Chung, K.: Changes in trkA expression in the dorsal root ganglion after peripheral nerve injury. Exp. Brain Res. 127:141, 1999. Shi, T. J., Tandrup, T., Bergman, E., et al.: Effect of peripheral nerve injury on dorsal root ganglion neurons in the C57 BL/6J mouse: marked changes both in cell numbers and neuropeptide expression. Neuroscience 105:249, 2001. Shin, J. B., Martinez-Salgado, C., Heppenstall, P. A., and Lewin, G. R.: A T-type calcium channel required for normal function of a mammalian mechanoreceptor. Nat. Neurosci. 6:724, 2003. Shu, X. Q., and Mendell, L. M.: Neurotrophins and hyperalgesia. Proc. Natl. Acad. Sci. U. S. A. 96:7693, 1999. Shubayev, V. I., and Myers, R. R.: Anterograde TNF alpha transport from rat dorsal root ganglion to spinal cord and injured sciatic nerve. Neurosci. Lett. 320:99, 2002. Shuto, T., Horie, H., Hikawa, N., et al.: IL-6 up-regulates CNTF mRNA expression and enhances neurite regeneration. Neuroreport 12:1081, 2001.

288. Simone, D. A., and Kajander, K. C.: Responses of cutaneous A-fiber nociceptors to noxious cold. J. Neurophysiol. 77:2049, 1997. 289. Sleeper, A. A., Cummins, T. R., Dib-Hajj, S. D., et al.: Changes in expression of two tetrodotoxin-resistant sodium channels and their currents in dorsal root ganglion neurons after sciatic nerve injury but not rhizotomy. J. Neurosci. 20:7279, 2000. 290. Smith, G. D., Gunthorpe, M. J., Kelsell, R. E., et al.: TRPV3 is a temperature-sensitive vanilloid receptor-like protein. Nature 418:186, 2002. 291. Snijdelaar, D. G., Dirksen, R., Slappendel, R., and Crul, B. J.: Substance P. Eur. J. Pain 4:121, 2000. 292. Sommer, C., Schmidt, C., and George, A.: Hyperalgesia in experimental neuropathy is dependent on the TNF receptor 1. Exp. Neurol. 151:138, 1998. 293. Song, W. J.: Genes responsible for native depolarizationactivated K currents in neurons. Neurosci. Res. 42:7, 2002. 294. Stark, B., Carlstedt, T., and Risling, M.: Distribution of TGF-beta, the TGF-beta type I receptor and the R-II receptor in peripheral nerves and mechanoreceptors: observations on changes after traumatic injury. Brain Res. 913:47, 2001. 295. Stebbing, M. J., Eschenfelder, S., Habler, H. J., et al.: Changes in the action potential in sensory neurons after peripheral axotomy in vivo. Neuroreport 10:201, 1999. 296. Sterne, G. D., Brown, R. A., Green, C. J., and Terenghi, G.: NT-3 modulates NPY expression in primary sensory neurons following peripheral nerve injury. J. Anat. 193(Pt. 2):273, 1998. 297. Story, G. M., Peier, A. M., Reeve, A. J., et al.: ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell 112:819, 2003. 298. Streit, W. J., Schulte, B. A., Balentine, J. D., and Spicer, S. S.: Histochemical localization of galactose-containing glycoconjugates in sensory neurons and their processes in the central and peripheral nervous system of the rat. J. Histochem. Cytochem. 10:1042, 1985. 299. Stucky, C., Shin, J. B., and Lewin, G. R.: Neurotrophin-4: a survival factor for adult sensory neurons. Curr. Biol. 12:1401, 2002. 300. Stucky, C. L., and Koltzenburg, M.: The low-affinity neurotrophin receptor p75 regulates the function but not the selective survival of specific subpopulations of sensory neurons. J. Neurosci. 17:4398, 1997. 301. Stucky, C. L., and Lewin, G. R.: Isolectin B(4)-positive and -negative nociceptors are functionally distinct. J. Neurosci. 19:6497, 1999. 302. Study, R. E., and Kral, M. G.: Spontaneous action potential activity in isolated dorsal root ganglion neurons from rats with a painful neuropathy. Pain 65:235, 1996. 303. Sugiura, S., Lahav, R., Han, J., et al.: Leukaemia inhibitory factor is required for normal inflammatory responses to injury in the peripheral and central nervous systems in vivo and is chemotactic for macrophages in vitro. Eur. J. Neurosci. 12:457, 2000. 304. Sugiura, T., Tominaga, M., Katsuya, H., and Mizumura, K.: Bradykinin lowers the threshold temperature for heat activation of vanilloid receptor 1. J. Neurophysiol. 88:544, 2002.

The Peripheral Sensory Nervous System: Dorsal Root Ganglion Neurons 305. Suzuki, H., Hase, A., Miyata, Y., et al.: Prominent expression of glial cell line-derived neurotrophic factor in human skeletal muscle. J. Comp. Neurol. 402:303, 1998. 306. Suzuki, M., Mizuno, A., Kodaira, K., and Imai, M.: Impaired pressure sensation in mice lacking TRPV4. J. Biol. Chem. 278:22664, 2003. 307. Suzuki, M., Watanabe, Y., Oyama, Y., et al.: Localization of mechanosensitive channel TRPV4 in mouse skin. Neurosci. Lett. 353:189, 2003. 308. Szarijanni, N., and Rethelyi, M.: Differential distribution of small and large neurons in the sacrococcygeal dorsal root ganglion of the cat. Acta Morphol. Acad. Sci. Hung. 27:25, 1979. 309. Szolcsanyi, J., Anton, F., Reeh, P. W., and Handwerker, H. O.: Selective excitation by capsaicin of mechano-heat sensitive nociceptors in rat skin. Brain Res. 446:262, 1988. 310. Taiwo, Y. O., and Levine, J. D.: Serotonin is a directly-acting hyperalgesic agent in the rat. Neuroscience 48:485, 1992. 311. Talley, E. M., Lei, Q., Sirois, J. E., and Bayliss, D. A.: TASK-1, a two-pore domain K channel, is modulated by multiple neurotransmitters in motoneurons. Neuron 25:399, 2000. 312. Tanaka, M., Cummins, T. R., Ishikawa, K., et al.: SNS Na channel expression increases in dorsal root ganglion neurons in the carrageenan inflammatory pain model. Neuroreport 9:967, 1998. 313. Tate, S., Benn, S., Hick, C., et al.: Two sodium channels contribute to the TTX-R sodium current in primary sensory neurons. Nat. Neurosci. 1:653, 1998. 314. Terenghi, G.: Peripheral nerve regeneration and neurotrophic factors. J. Anat. 194(Pt. 1):1, 1999. 315. Their, M., Marz, P., Otten, U., et al.: Interleukin-6 (IL-6) and its soluble receptor support survival of sensory neurons. J. Neurosci. Res. 55:411, 1999. 316. Thippeswamy, T., and Morris, R.: The roles of nitric oxide in dorsal root ganglion neurons. Ann. N. Y. Acad. Sci. 962:103, 2002. 317. Thompson, S. W., Bennett, D. L., Kerr, B. J., et al.: Brainderived neurotrophic factor is an endogenous modulator of nociceptive responses in the spinal cord. Proc. Natl. Acad. Sci. U. S. A. 96:7714, 1999. 318. Thompson, S. W., Vernallis, A. B., Heath, J. K., and Priestley, J. V.: Leukaemia inhibitory factor is retrogradely transported by a distinct population of adult rat sensory neurons: co-localization with trkA and other neurochemical markers. Eur. J. Neurosci. 9:1244, 1997. 319. Toledo-Aral, J. J., Moss, B. L., He, Z. J., et al.: Identification of PN1, a predominant voltage-dependent sodium channel expressed principally in peripheral neurons. Proc. Natl. Acad. Sci. U. S. A. 94:1527, 1997. 320. Tominaga, M., Caterina, M. J., Malmberg, A. B., et al.: The cloned capsaicin receptor integrates multiple pain-producing stimuli. Neuron 21:531, 1998. 321. Tong, Y. G., Wang, H. F., Ju, G., et al.: Increased uptake and transport of cholera toxin B-subunit in dorsal root ganglion neurons after peripheral axotomy: possible implications for sensory sprouting. J. Comp. Neurol. 404:143, 1999. 322. Treede, R. D., Meyer, R. A., and Campbell, J. N.: Myelinated mechanically insensitive afferents from monkey hairy skin: heat-response properties. J. Neurophysiol. 80:1082, 1998.

201

323. Vallbo, A., Olausson, A., Wessberg, H., and Norrsell, U.: A system of unmyelinated afferents for innocuous mechanoreception in the human skin. Brain Res. 628:301, 1993. 324. Valtschanoff, J. G., Phend, K. D., Bernardi, P. S., et al.: Amino acid immunocytochemistry of primary afferent terminals in the rat dorsal horn. J. Comp. Neurol. 346:237, 1994. 325. Vergara, C., Latorre, R., Marrion, N. V., and Adelman, J. P.: Calcium-activated potassium channels. Curr. Opin. Neurobiol. 8:321, 1998. 326. Verge, V. M., Richardson, P. M., Wiesenfeld-Hallin, Z., and Hokfelt, T.: Differential influence of nerve growth factor on neuropeptide expression in vivo: a novel role in peptide suppression in adult sensory neurons. J. Neurosci. 15:2081, 1995. 327. Vestergaard, S., Tandrup, T., and Jakobsen, J.: Effect of permanent axotomy on number and volume of dorsal root ganglion cell bodies. J. Comp. Neurol. 388:307, 1997. 328. Viana, F., de la Pena, E., and Belmonte, C.: Specificity of cold thermotransduction is determined by differential ionic channel expression. Nat. Neurosci. 5:254, 2002. 329. Villiere, V., and McLachlan, E. M.: Electrophysiological properties of neurons in intact rat dorsal root ganglia classified by conduction velocity and action potential duration. J. Neurophysiol. 76:1924, 1996. 330. Waddell, P. J., and Lawson, S. N.: Electrophysiological properties of subpopulations of rat dorsal root ganglion neurons in vitro. Neuroscience 36:811, 1990. 331. Wall, P. D., and Gutnick, M.: Ongoing activity in peripheral nerves: the physiology and pharmacology of impulses originating from a neuroma. Exp. Neurol. 43:580, 1974. 332. Wang, R., Guo, W., Ossipov, M. H., et al.: Glial cell linederived neurotrophic factor normalizes neurochemical changes in injured dorsal root ganglion neurons and prevents the expression of experimental neuropathic pain. Neuroscience 121:815, 2003. 333. Watkins, L. R., and Maier, S. F.: Beyond neurons: evidence that immune and glial cells contribute to pathological pain states. Physiol. Rev. 82:981, 2002. 334. White, F. A., Silos-Santiago, I., Molliver, D. C., et al.: Synchronous onset of NGF and TrkA survival dependence in developing dorsal root ganglia. J. Neurosci. 16:4662, 1996. 335. Wood, J. N.: Recent advances in understanding molecular mechanisms of primary afferent activation. Gut 53(Suppl. 2):II9, 2004. 336. Woolf, C. J., Allchorne, A., Safieh-Garabedian, B., and Poole, S.: Cytokines, nerve growth factor and inflammatory hyperalgesia: the contribution of tumour necrosis factor alpha. Br. J. Pharmacol. 121:417, 1997. 337. Woolf, C. J., Ma, Q. P., Allchorne, A., and Poole, S.: Peripheral cell types contributing to the hyperalgesic action of nerve growth factor in inflammation. J. Neurosci. 16:2716, 1996. 338. Woolf, C. J., Safieh-Garabedian, B., Ma, Q. P., et al.: Nerve growth factor contributes to the generation of inflammatory sensory hypersensitivity. Neuroscience 62:327, 1994. 339. Wright, D. E., and Snider, W. D.: Neurotrophin receptor mRNA expression defines distinct populations of neurons in rat dorsal root ganglia. J. Comp. Neurol. 351:329, 1995. 340. Wu, G., Ringkamp, M., Hartke, T. V., et al.: Early onset of spontaneous activity in uninjured C-fiber nociceptors after injury to neighboring nerve fibers. J. Neurosci. 21:RC140, 2001.

202

Function of the Peripheral Nervous System

341. Wyatt, S., Shooter, E. M., and Davies, A. M.: Expression of the NGF receptor gene in sensory neurons and their cutaneous targets prior to and during innervation. Neuron 4:421, 1990. 342. Xiao, H. S., Huang, Q. H., Zhang, F. X., et al.: Identification of gene expression profile of dorsal root ganglion in the rat peripheral axotomy model of neuropathic pain. Proc. Natl. Acad. Sci. U. S. A. 99:8360, 2002. 343. Xu, H., Ramsey, I. S., Kotecha, S. A., et al.: TRPV3 is a calcium-permeable temperature-sensitive cation channel. Nature 418:181, 2002. 344. Yamadori, T.: A light and electron microscopic study on the postnatal development of spinal ganglia in rats. Acta Anat. Nippon 45:191, 1970. 345. Yang, E. K., Takimoto, K., Hayashi, Y., et al.: Altered expression of potassium channel subunit mRNA and alpha-dendrotoxin sensitivity of potassium currents in rat dorsal root ganglion neurons after axotomy. Neuroscience 123:867, 2004. 346. Yoon, Y. W., Na, H. S., and Chung, J. M.: Contributions of injured and intact afferents to neuropathic pain in an experimental rat model. Pain 64:27, 1996.

347. Yoshida, S., and Matsuda, Y.: Studies on sensory neurons of the mouse with intracellular recording and dye injection techniques. J. Neurophysiol. 42:1134, 1979. 348. Yusaf, S. P., Goodman, J., Pinnock, R. D., et al.: Expression of voltage-gated calcium channel subunits in rat dorsal root ganglion neurons. Neurosci. Lett. 311:137, 2001. 349. Zhou, X. F., Cameron, D., and Rush, R. A.: Endogenous neurotrophin-3 supports the survival of a subpopulation of sensory neurons in neonatal rat. Neuroscience 86:1155, 1998. 350. Zhou, X. F., Chie, E. T., Deng, Y. S., et al.: Injured primary sensory neurons switch phenotype for brain-derived neurotrophic factor in the rat. Neuroscience 92:841, 1999. 351. Zhou, S.-F., Deng, Y.-S., Chie, E., et al.: Satellite cell-derived nerve growth factor and neurotrophins are involved in noradrenergic sprouting in the dorsal root ganglia following peripheral nerve injury in the rat. Eur. J. Neurosci. 11:1711, 1999. 352. Zhou, Z., Davar, G., and Strichartz, G.: Endothelin-1 (ET-1) selectively enhances the activation gating of slowly inactivating tetrodotoxin-resistant sodium currents in rat sensory neurons: a mechanism for the pain-inducing actions of ET-1. J. Neurosci. 22:6325, 2002.

9 The Pupil JAMES J. CORBETT

Anatomy of Iris Innervation Sympathetic Innervation Parasympathetic Innervation to the Intraocular Muscles Horner’s Syndrome The Size of the Normal Pupil

Simple Anisocoria Pupillary Unrest and Hippus Pupil Cycle Time Iris Damage and Its Effect on the Light, Near, and Drug Responses

The iris can be seen easily, and it supports autonomically innervated constrictor and dilator muscles that, when properly examined, can be used as a rough indicator of damage to the peripheral nerves. The limited utility of such testing is related to the opposing nature of the iris muscles and to the fact that the iris is an extension of the uveal tract. The iris is involved in uveal inflammation that can alter its size, shape, and reactivity to light response, near reaction, and various drugs. The reaction of the iris to light, dark, and near and to adrenergic and cholinergic drugs, as well as supersensitivity reactions and the appearance on slit-lamp examination, have produced considerable literature with conflicting conclusions. This welter of information is best dealt with by separating the responses of the pupil to light, near, dark, and drugs and by examining the effects of classic isolated lesions on each system individually.

ANATOMY OF IRIS INNERVATION Sympathetic Innervation The fibers of the dilator muscle are in the most posterior layer of the iris and are saturated with the melanin that keeps the iris opaque to light (Fig. 9–1). The central cells of origin for sympathetic innervation of the iris dilator are in the posterolateral hypothalamus. The axons traverse the brainstem to spinal cord levels C7-T1, where they synapse in the ciliospinal center of Budge. Axons from the intermediolateral columns of the spinal cord provide the second step in the three-neuron arc. From the spinal cord

Efferent Pupillary Defects Defects of Parasympathetic Innervation The Pupil as a Measure of Autonomic Dysfunction

and beyond, the fibers form part of the peripheral nervous system. The fibers proceed across the apex of the lung, around the ansa subclavia, and pass through the stellate (inferior cervical) ganglion without synapse. They ascend to the superior cervical ganglion (C3-C4 level, angle of the mandible), where they synapse and travel with the internal carotid artery through the foramen lacerum into the cavernous sinus. The sympathetic fibers join the abducens nerve for about 4 mm and then enter the orbit with the ophthalmic division of the trigeminal nerve. They then follow the long posterior ciliary nerves to segmentally innervate the radially arranged fibers of the dilator muscles in the iris (Fig. 9–2).69 The entire three-neuron sympathetic outflow path must be intact to provide tonic output of norepinephrine at the effector cell. A lesion to any one of these three neuron/axon regions stops the tonic output of norepinephrine into the synaptic cleft at the receptors of the dilator muscle and causes a Horner’s syndrome (see below) (Fig. 9–3).

Parasympathetic Innervation to the Intraocular Muscles The iris sphincter and the ciliary muscle are controlled by the Edinger-Westphal nucleus, which is a dorsal-rostral part of the oculomotor nucleus. The third cranial nerve leaves the midbrain in the interpeduncular fossa. Near its epineurium are carried the small-caliber fibers that serve the intraocular muscles. The synapse in this typical two-neuron parasympathetic outflow path is in the ciliary 203

204

Function of the Peripheral Nervous System

FIGURE 9–1 The sphincter muscle (c) is innervated by axons from ciliary ganglion neurons. Slit-lamp examination of the iris suggests that there are between 20 and 30 motor units in the sphincter, but higher estimates (up to 70) have been made. The neuromuscular bundles are arranged in overlapping sectors around the edge of the iris. The dilator muscle (f) is arranged radially; the fibers insert by long extensions called spurs that blend into the sphincter muscle. a ⫽ Anterior bundle; b ⫽ pigment ruff; c ⫽ sphincter muscle; d ⫽ vascular arcades; e ⫽ groups of capillaries, nerves, and melanocytes; f ⫽ dilator muscle; g ⫽ anterior epithelium; h ⫽ Michel’s spur; i ⫽ Fuch’s spur; j ⫽ posterior epithelium. (From Hogan, M. J., Alvarodo, J. A., and Weddell, J. E.: Histology of the Human Eye. Philadelphia, W. B. Saunders, 1971, with permission.)

ganglion, which is located deep in the muscle cone of the orbit about 1 cm behind the globe and just lateral to the optic nerve and medial to the lateral rectus muscle. The postganglionic fibers follow the short ciliary nerves into the sclera to the anterior segment of the globe (see Fig. 9–2). The fibers of the ciliary nerves run between the choroid and the sclera and may be damaged by the retinal photocoagulation used to treat diabetic retinopathy.43,48,53

Horner’s Syndrome Horner’s syndrome is caused by interruption of the ipsilateral sympathetic outflow to the head and neck. It can result from lesions in the brainstem or spinal cord, damage at the lung apex and in the supraclavicular space, or damage to the

carotid plexus along the internal carotid artery all the way to the cavernous sinus. Since the same clinical picture is produced wherever this pathway is interrupted, the combination of miosis and ptosis by itself is not particularly helpful in localizing the lesion.65,73 Recent reports of isolated Horner’s syndrome as the only manifestation of syringomyelia emphasize the need to evaluate this syndrome thoroughly and to be sure that the ptosis and miosis are not due to pseudo-Horner’s syndrome.21,42,63 Clinical Signs Wherever Horner’s syndrome occurs, the lesion also affects neighboring anatomic structures, causing signs and symptoms that help localize the lesion. Unfortunately, efforts to evaluate sweating deficits add little, and the

The Pupil

205

FIGURE 9–2 The optic globe with the sclera peeled back, disclosing the segmental array of short and long ciliary nerves. These penetrate the posterior sclera in proximity to the optic nerve and lie between the choroids and the sclera. In this location, they are vulnerable to photocoagulation burns. At the limbus of the cornea, where they come into contact with the iris, the nerves may be torn loose by blows to the eye. (From Pernkoff, E.: Atlas of Topographical and Applied Human Anatomy. Berlin, Urban & Schwarzenberg, 1963.)

sweat “patterns” tend to be equivocal in their localizing specificity. Studies of sweating patterns in Horner’s syndrome using quinazirine and starch and iodine, with occasional erratic skin temperature results and galvanic skin responses, have been plagued by variability, most likely because of aberrant regeneration.36,49 Ptosis. Müller’s muscle, the sympathetically innervated retractor of the upper lid, is paralyzed in Horner’s syndrome, and this narrows the palpebral fissure. The ptosis in Horner’s syndrome, by itself, never covers the visual axis. With the patient’s eyes wide open and gazing upward, the ptosis almost disappears.65 The lower lid also has some sympathetically innervated retractor fibers. With sympathetic denervation the lower lid rises slightly (“upside-down ptosis”).37 This further adds to the narrowed palpebral fissure and gives the appearance of enophthalmos. There is no true enophthalmos.26,75 Miosis. The miosis of Horner’s syndrome is never intense. Tiny pupils (less than 3 mm in diameter) are due to some

other (or additional) cause, such as miotic drops, loss of supranuclear inhibition (as in coma), or long-standing immobility (as in Adie’s syndrome or Argyll Robertson pupil). Paralysis of the pupillodilator muscle causes only a moderate decrease in pupil size. In bright light, both pupils constrict normally and are nearly equal. Anisocoria in unilateral sympathetic denervation is more evident in dim light because the sphincter relaxes in both eyes; this leaves the normally functioning dilator muscle to dilate the normal pupil and a flaccid dilator in the affected eye, which eventually passively dilates to almost normal size. Dilation Lag. The slowness of the affected pupil to dilate is characteristic of Horner’s syndrome (Fig. 9–4). Since the problem is usually unilateral, this can be of great clinical help in making the diagnosis. Pupil dilation has two components: passive dilation of the iris as a result of inhibition of the sphincter, and active contraction of the sympathetically innervated radial iris muscles. When the lights are turned off, both pupils dilate. However, the

FIGURE 9–3 Parasympathetic and sympathetic pathways to the eye. In the lower part of the figure, the solid line indicates the pathway of the pupillodilator fibers, and the dashed lines show some of the other sympathetic pathways to the orbit and the face. 1, Edinger-Westphal nucleus; 2, oculomotor nerve; 3, branch to inferior oblique muscle; 4, motor root of ciliary ganglion; 5, short ciliary nerves; 6, ciliary muscle and iris sphincter; 7, superior cervical ganglion; 8, internal carotid artery; 9, external carotid artery; 10, sudomotor fibers to the face; 11, carotid plexus; 12, caroticotympanic nerve; 13, tympanic plexus; 14, deep petrosal nerve; 15, lesser superficial petrosal nerve; 16, sympathetic contribution to the vidian nerve; 17, ophthalmic division of trigeminal nerve; 18, nasociliary nerve; 19, long ciliary nerve; 20, ciliary muscle and iris dilator; 21, probable path of sympathetic contribution to retractor muscles of the lids; 22, vasomotor and some sudomotor fibers; 23, ophthalmic artery; 24, lacrimal gland; 25, short ciliary nerves; 26, sympathetic contribution to salivary glands; 27, greater superficial petrosal nerve. (Prepared by H. Stanley Thompson and drawn by Joan Esperson, University of California Medical School, San Francisco, California. From Walsh, F. B., and Hoyt, W. F.: Clinical Neuro-Ophthalmology, Vol. 1. Baltimore, Williams & Wilkins, 1969, with permission.)

The Pupil

207

FIGURE 9–4 Clinical example of “dilation lag” in Horner’s syndrome. Dilation of the pupil occurs when the sphincter stops contracting and the dilator fibers begin to fire. When the iris is sympathetically denervated, the iris dilators do not add speed to opening the pupil, whereas they do in the normal eye. A normal eye dilates almost fully in the dark within 5 seconds. Without sympathetics it takes much longer to accomplish the same phenomenon, hence the name of the phenomenon, dilation lag.

sympathetically denervated pupil dilates more slowly because no active contraction of the dilator muscle occurs; only sphincter inhibition permits the pupil to dilate.41 As the pupil slowly dilates, the anisocoria gradually decreases. Thus there is a lag in dilation when sympathetic denervation is present. This phenomenon can be seen by illuminating both eyes tangentially from below as the room lights are turned off. The anisocoria is much greater at 5 seconds after “lights out” than at 15 seconds. Dilation lag can be documented with flash photos at 5 and then again at 15 seconds; better yet, infrared-sensitive video can be used.76 Facial Anhidrosis. Hemifacial sweating disorders or hemifacial blanching or flushing sometimes draw attention to a sympathetic deficit, but these phenomena are not consistently present in Horner’s syndrome. Again, aberrant regeneration of damaged fibers may contribute to the uncertainty.49,65 Conjunctival and Lid Hyperemia. Conjunctival and lid hyperemia is related to the defect in ipsilateral innervation of all extracranial blood vessels. Loss of vasoconstrictive tone leaves the conjunctiva hyperemic. A fresh Horner’s syndrome is frequently mistaken for conjunctivitis.69 Pupil Pharmacology as Related to Autonomic Denervation. There are many pitfalls in trying to interpret the response of pupils to eyedrops. Individual variability of responsiveness to drugs makes it important to compare the effect of drops in both eyes. Emotional upset or anxiety will dilate the pupils, and drowsiness will constrict them. Both states interfere with evaluation.

Dry eyes, numb eyes, exposed eyes, and eyes that recently have been touched by a wisp of cotton, a topical anesthetic, a tonometer, or a contact lens all are likely to have damaged corneal epithelium.4 Such an eye is likely to soak up more of any drug than an eye with a normal corneal epithelium, which acts as a barrier to absorption. Variation in corneal permeability is especially important when supersensitivity tests are being done. Weak solutions may cause pupillary reactions that would not occur but for highly permeable corneas. In general, the more pigment in the iris, the more slowly the drug takes effect and the longer its action lingers.39,40,46 This is probably due to binding of the drug to iris melanin, from which it is then slowly released. There are wide individual variations in the pupillary responses to topical drugs. The range of responses among blue eyes includes the average response of dark brown eyes,46 so it is probably not just a matter of melanin; there may be genetic factors that influence corneal penetration. Sympathetic Pharmacology Cocaine (5% to 10%) is applied to the conjunctiva as a topical anesthetic, as a mydriatic, and as a test for Horner’s syndrome.11,18,74 Its mydriatic effect is due to accumulation of norepinephrine at the adrenergic receptors (synaptic sites) of the dilator cells. Cocaine causes the transmitter substance to build up at the neuroeffector junction because it prevents the reuptake of the norepinephrine back into the nerve endings. The action of cocaine is analogous to the effect of an anticholinesterase at the cholinergic synapse—it interferes with the mechanism for the prompt disposition of the chemical mediator. Cocaine has no direct action on the effector cell and does not release norepinephrine from the nerve ending. Cocaine

208

Function of the Peripheral Nervous System

does not block the physiologic release of norepinephrine from the nerve endings; it only blocks the norepinephrine reuptake mechanism so that norepinephrine accumulates at the effector junction until the muscle cell of the iris dilator fires continuously. If the nerve action potential is interrupted at any place in the sympathetic pathways, as in Horner’s syndrome, norepinephrine will not accumulate and the pupil will fail to dilate with cocaine drops. The duration of cocaine mydriasis is variable; it may last for more than 24 hours and there is no “rebound miosis.” No more than one drop of 10% cocaine or two drops of the 5% solution should be placed in each eye. More cocaine has a toxic effect on the corneal epithelium, and if too much of it is used, the patient may have an uncomfortable eye after the anesthetic effect wears off. Epinephrine (adrenalin) directly stimulates the receptor sites of the pupil dilator. When epinephrine is applied to the conjunctiva in a weak 1:1000 (0.1%) solution, it does not penetrate into the normal eye in sufficient quantity to have any obvious mydriatic effect. If there has been postganglionic denervation and the receptors have been made supersensitive, this weak solution may dilate the pupil, but the effect is not dependable. Phenylephrine (Neo-Synephrine) is also a directly active adrenergic mydriatic; its action is almost exclusively a direct ␣-adrenergic stimulation of the effector cell. The 1% solution has a moderate mydriatic effect; put in both eyes, it is a better test for adrenergic supersensitivity in postganglionic Horner’s syndrome than is 0.1% epinephrine. Both 5% and 2.5% solutions of phenylephrine are commonly used mydriatics. The pupil recovers in 8 hours and shows a “rebound miosis” lasting for several days. Higher concentrations such as 10% phenylephrine, when absorbed by the nasal mucosa, may produce hypertensive crises, especially in patients with

generalized autonomic neuropathy, if even two drops are placed in each eye.69 Tyramine and especially hydroxyamphetamine (Paredrine) act adrenergically by releasing norepinephrine from the postganglionic nerve endings. This appears to be their only effective action. This effect permits them to be used to separate postganglionic from central and preganglionic Horner’s syndrome.5,70,71 Botulinum toxin blocks the release of acetylcholine, and hemicholinium interferes with synthesis of acetylcholine at both the preganglionic and the postganglionic nerve endings, thus interrupting the parasympathetic pathway in two places. The outflow of sympathetic impulses also is interrupted by systemic doses of these drugs, since the chemical mediator in sympathetic ganglia is also acetylcholine.69

The Size of the Normal Pupil The pupil is not normally quite at the center of the iris but is slightly nasal and inferior. Pupil size varies with age (Fig. 9–5).28 Sympathetic innervation of the dilator muscle appears to be acquired at term. This is based on the observation that the pupils of some premature infants dilate in response to phenylephrine but not to hydroxyamphetamine.16 When a premature baby is awakened, the pupils are small (3.6 mm ⫾ 0.9 mm), according to Isenberg and co-workers,16 and they show no light reaction until a gestational age of 31 weeks is reached. The light reaction gradually improves until it is about 2 mm at term.47 The miosis in newborn babies is due to the fact that they sleep 22 hours out of every 24 and also because the eyeball is not full-grown. By adolescence

FIGURE 9–5 Pupillary size in darkness at various ages. The 1263 subjects were chosen at random from a population survey. Each point represents the average diameter of the two pupils taken together. The ordinate shows horizontal diameter (in millimeters); the abscissa shows age in years. Note the wide scatter among individuals and the obvious age trend. (From Loewenfeld, I. E.: Pupillary changes related to age. In Thompson, H. S. [ed.]: Topics in Neuro-Ophthalmology. Baltimore, Williams & Wilkins, p. 124, 1979, with permission.)

209

The Pupil

the pupils are at their largest. Thereafter they gradually become smaller until age 60, when size levels off.28

associated with hippus, but gradually it has become accepted as a normal phenomenon.60,69

Simple Anisocoria

Pupil Cycle Time

It is common to have a small amount of pupillary inequality. About 20% of the normal population have an easy-to-detect anisocoria of 0.4 mm or more at any given time (based on flash photos taken in dim light). This pupillary inequality is not constant in any given individual; it may increase or decrease or even reverse sides within days or hours. Smaller amounts of asymmetry are even more common. When subjects are examined twice daily for 5 days, 70% will have anisocoria of 0.3 mm or more at least once during this period (Fig. 9–6).24 “Simple” anisocoria decreases in bright light, has not been associated with any disease process, and produces no symptoms. The alternative terms physiologic anisocoria and essential anisocoria are also used. Since the light, dark, and near reactions in both eyes are normal, no further evaluation for a peripheral neuropathic cause for the anisocoria is warranted.

A small beam of light focused on the edge of the pupil induces regular oscillations of the pupil that one can time with a stopwatch. The period of the average complete cycle is called the pupil cycle time.34,68 This technique requires a slit lamp and has been used to look at the iris with sympathetic and parasympathetic denervation.3,31 In both cases the pupil cycle time is slowed. Although pupil cycle time testing may broadly point to an autonomic neuropathy that affects the innervation of the intraocular muscles, it is not useful in distinguishing sympathetic from parasympathetic damage. It is a labor intensive, nonspecific marker of iris muscle dysfunction.

Pupillary Unrest and Hippus The normal iris moves almost continuously, even when there is constant illumination and accommodation. This movement is known as physiologic pupillary unrest or hippus. It is thought to be due to moment-to-moment fluctuations in the activity of the sympathetic and parasympathetic innervation of the iris muscles. Hippus is most obvious in young patients with large pupils and is greatest in moderately bright light. The frequency of the oscillation increases with the intensity of the light. During the 19th century many neurologic diseases were

FIGURE 9–6 The amount of clinically visible anisocoria in a group of 128 normal subjects photographed morning and evening for 5 days. Anisocoria of 0.4 mm is easily discerned and is seen in 19% at any given time; 41% of the subjects had it all the time or some of the time. Thus anisocoria, which is clinically identifiable, is present in one in five patients at any moment, but this is not a fixed population. The mnemonic PERRLA (pupils equal, round, and reactive to light and accommodation) is inaccurate at least 20% of the time. (Data from Lam, B. L., Thompson, H. S., and Corbett, J.: Prevalence of simple anisocoria. Am. J. Ophthalmol. 104:69, 1987.)

Iris Damage and Its Effect on the Light, Near, and Drug Responses Blunt injury to the globe may cause a traumatic iridoplegia. Several factors seem to be at work: (1) the chamber angle may be recessed, tearing the branches of the short ciliary nerves that serve the iris muscle; (2) the sphincter muscle itself may be injured so that it will not constrict in response to pilocarpine; and (3) the short ciliary nerves may be damaged if the choroid is ruptured. These injuries may produce a segmental palsy of the sphincter, a light-near dissociation, and either an undersensitivity or oversensitivity of the sphincter to pilocarpine, depending on the exact nature of the damage. An attack of acute angle-closure glaucoma often produces a pupil of moderate size that fails to react to light or to pilocarpine. This is due to permanent ischemic damage to the iris muscles.

210

Function of the Peripheral Nervous System

EFFERENT PUPILLARY DEFECTS Defects of Parasympathetic Innervation Differential Diagnosis of Sphincter Dysfunction (Fig. 9–7) Dorsal midbrain damage rarely causes isolated pupillary defects. When it does, usually both pupils are affected, and there is often a light-near dissociation. The most common associated ocular motility disorder in this setting is the dorsal midbrain syndrome. Compression of the dorsal midbrain or an intrinsic lesion is the usual cause. Pupil abnormalities caused by ventral midbrain or fascicular lesions are also exceedingly rare in isolation, usually occurring with the other elements of ventral midbrain dysfunction. The pupil fibers are grouped superiorly and medially on the oculomotor nerve as it exits the brainstem, and rarely, small mesencephalic infarcts will affect pupil fibers alone. Tonic Pupil Damage to the ciliary ganglion or short ciliary nerves produces a very characteristic combination of signs consisting of the following: (1) a poor pupillary reaction to light that, under slit-lamp examination, can be seen to be a regional or sector palsy of the iris sphincter (Fig. 9–8);

(2) accommodative paresis, which is frequently incomplete and therefore also “segmental”; and (3) cholinergic supersensitivity of the denervated muscles. The usual recommendation of 0.1% pilocarpine to test for an Adie’s tonic pupil has been recently challenged by a study that recommends 0.0625% pilocarpine. This dose will constrict a tonic pupil but will not constrict a normal pupil. In a study of 11 patients with dilated pupils caused by oculomotor nerve palsy and dilated pupils resulting from Adie’s tonic pupil, no difference was seen between the two groups using 0.1% pilocarpine as the supersensitivity stimulus. Both demonstrated denervation supersensitivity, conclusively showing that this phenomenon clinically occurs in both preganglionic and postganglionic conditions. Denervation supersensitivity cannot be used to separate tonic pupils from those caused by third cranial nerve preganglionic damage because it is seen in both.25 Often the pupillary response to near vision is unusually strong and tonic. Patients who show this constellation of signs are said to have tonic pupils. A tonic pupil can result from any postganglionic, parasympathetic denervation of the intraocular muscles.29 This can, of course, be due to a local infection or injury or a widespread generalized or autonomic peripheral neuropathy, as with diabetes or alcoholism.56,58 If otherwise unexplained, it is called Adie’s syndrome.13,29,66

FIGURE 9–7 Pharmacology of the sympathetic and parasympathetic nerve pathways of the pupil.

The Pupil

211

FIGURE 9–8 A tonic pupil showing segmental constriction of the iris sphincter between 11:00 and 12:30 and 2:30 and 3:30. Partial segmental denervation produces, with hippus, the flickering so-called vermiform movements of the edge of the iris as it is held by illumination. The segments that constrict in response to light are not the same as those that constrict in response to a near stimulus.

Adie’s Syndrome Adie’s syndrome is probably the most common pathologic cause of anisocoria and unilateral poor light reaction. Almost all denervated sphincters in Adie’s syndrome have sectors of residual light reaction when examined under slit lamp.66 The majority of these patients are young women, but Adie’s pupil can be seen in both sexes and at any age. In children younger than 10 years, Adie’s pupil is almost always preceded by a history of chickenpox. The damage in Adie’s tonic pupil occurs in the ciliary ganglion or to the short posterior ciliary nerves. The reaction of the pupil to light is absent in 10% of patients. The reaction of the pupil to a near effort is usually slow and strong and is often tonic. When the patient is asked to refixate at a distance, the tonically constricted pupil redilates very slowly. This is true of any parasympathetically denervated pupil.

FIGURE 9–9 Depiction of unilateral (left eye) anisocoria seen in Adie’s syndrome. Initially, there is a large discrepancy between the two pupils that, over time, becomes less and less. In this example, the right eye then becomes involved at 31/2 years. The anisocoria becomes greater, but gradually both pupils become small, unreactive to light, and tonically reactive to a near stimulus. Involvement of the second eye occurs at a rate of 4% per year. (From Thompson, H. S. [ed.]: Topics in NeuroOphthalmology. Baltimore, Williams & Wilkins, 1979, with permission.)

The basic principles of parasympathetic denervation are the same whatever the cause: 1. Partial denervation produces segmental dysfunction of the iris sphincter depending on which motor units are preserved and which are denervated. 2. Reinnervation is likely to be disproportionately populated by ciliary muscle nerves, thereby producing a pupil that responds better to a near effort than it does to light. 3. While the affected pupil is larger than the normal pupil at first, it gradually becomes smaller until, in all but the brightest of light, it is smaller than that of the unaffected eye. This has been called a “little old Adie’s” (Fig. 9–9). If both eyes are affected, both pupils become small and either weakly or imperceptibly reactive to light. Thus

212

Function of the Peripheral Nervous System

they would seem to be indistinguishable from the light-near dissociation that is associated with syphilis. However a “little old Adie’s” has a tonic near response and the Argyll Robertson pupil does not.17 Drug-Induced and Toxic Mydriasis Atropinic drugs and plants that are rich in the belladonna alkaloids can all cause unilateral or bilateral pupil dilation and paralysis of accommodations. The features of an anticholinergic mydriasis are (1) a 360-degree paralysis of the iris sphincter (no sector loss such as that seen with neuropathic damage), and (2) no constriction in response to 1% pilocarpine eyedrops—a concentration that will vigorously constrict the normal and the denervated iris sphincter.71,72 An iron-containing foreign body in the globe can slowly leach ferric ions into the tissues and poison the intraocular muscles. The result is a dilated and unreactive pupil.35 None of the heavy metals that typically cause peripheral axonal neuropathies has been implicated in pupillary problems. Patients with such intoxication should be tested for sector palsies by careful slit-lamp examination. Light-Near Dissociation Sometimes a striking disparity can be seen between the pupillary response to light and the pupillary response to near vision.67 The light reflex arc may be damaged, but the near reaction pathway may be untouched. This is the case in which there is damage to the afferent limb of the pupillary arc, including the retina, optic nerve, chiasm, an optic tract, brachium of the superior colliculus, or pretectal nucleus. In contrast, the light reflex and the near response may both be damaged in the efferent pathway, and then the near reaction may be restored by aberrant regeneration. Since 97% of the cells in the ciliary ganglion are directed to the ciliary muscle, random regeneration of these cells will be sure to send some fibers to the iris sphincter.69 The pupillary near reaction is frequently forgotten in the course of routine examination. For a near response examination to be accurate, the patient must be able to see what he or she is looking at as a near target, and the study should be done with good room illumination. The patient should be given an accommodative target to look at—an object of interest or one with fine detail on it. Occasionally the response is better if another sensory input is used, such as clicking one’s fingernails. Proprioceptive input, such as using the patient’s own thumbnail, can sometimes elicit a near response when nothing else works. The examiner should watch for the eyes to converge since this reveals how hard the patient is trying. The near response is triggered by blurred or disparaged imagery but has a large volitional component, and verbal encouragement may help. Recognizing a Light-Near Dissociation. It can be difficult to know whether there is a light-near dissociation. When is the near response better than the light response?

When the clinician examines the patient with a pocket light, there are usually three levels of light available: darkness with a light shining tangentially on the pupils from below, room light, and room light with an additional bright light in the eyes. Having the patient look in the distance and shining a bright hand light in the eye three or four times, each time for only 1 or 2 seconds, reveals how small the pupils will become with just a light stimulus. A true light-near dissociation can be present only if the near response (tested in moderate light) is better than the best constriction that bright light can produce (Table 9–1).

The Pupil as a Measure of Autonomic Dysfunction Despite clinical observations and reports of studies of pupil defects in disease (Table 9–2), many problems arise when the pupils are used as an indicator of autonomic dysfunction. The chief difficulty is that the autonomic neuropathy being sought is a generalized process that is likely to affect both pupils. The clearest pupillary signs are the ones that contrast one pupil with the other. Unilateral denervation is easy to identify, but bilateral partial denervation is hard to distinguish from normal. There is hardly a test for Horner’s syndrome, for example, that does not involve comparing one eye with the other. Any test that depends on pupil size or pupillary inequality must contend with the fact that there is a certain amount of simple anisocoria in all of us; we are not, it seems, made with more precision than is necessary. This anisocoria can change from one minute to the next. In addition, one has to be certain that both irises are free of structural damage from prior disease or injury. The fact that pupil size varies with age also must be sidestepped. Both pupils tend to enlarge together with excitement and constrict together with drowsiness, so whenever possible one pupil should be used as a control. Table 9–1. Equipment Useful in Testing and Recording Pupil Function Polaroid CU-5 1:1 camera Filament-free, bright light source (Welch Allyn handle and Finhoff transilluminator) Magnifier with built-in millimeter rule scale to accurately measure pupil photos Accommodative targets Cocaine 4%–10%—ordinarily dispensed in small individual droppers by prescription. Must be disposed of after use. Hydroxyamphetamine 1% (Paredrine)—now available from Pharmics (2350 South Redwood Road, Salt Lake City, UT 84119; Fax: 801-972-4139) Phenylephrine 1% Pilocarpine 0.025% Slit lamp

The Pupil

Table 9–2. Peripheral Neuropathies with Pupil Involvement Acute pandysautonomia1,2,30,62,78,79 Adie’s syndrome13,29,65,66 Alcoholism61,64 Amyloidosis9,51 Botulism10,22,32 Charcot-Marie-Tooth disease20,52 Diabetes mellitus12,15,56–58 Giant cell arteritis6,45 Hereditary sensory neuropathy33 Holmes-Adie syndrome44,54,59 Landry-Guillain-Barré disease77 Miller Fisher variant19,38 Ross syndrome7,14,50 Sarcoidosis23 Shy-Drager syndrome55 Syphilis8,27

Any test using an autonomically acting drug must take into account the influence of the iris pigment, the integrity of the corneal surface and its innervation, the presumed completeness of the denervation, and the amount of reinnervation. Some of the simplest tests for the integrity of innervation of the iris muscles either are not available to the neurologist or are not in common use. For example, the high-magnification view of the iris afforded by a slit lamp provides a clear indication of segmental denervation of the iris sphincter and can also reveal signs of aberrant regeneration. A simple infrared-sensitive video of the pupils gives a valuable view of both pupils in the dark, in the light, and at near. For heavily pigmented irises, a video camera can make the difference between seeing and not seeing the pupillary signs.76 An iris suffering from denervation of its muscles is likely to be smaller in the dark than is appropriate for the patient’s age, and it will probably react relatively weakly to light. The trouble is that both eyes are likely to be affected, and it will be hard to be certain that they are abnormal. If a segment of the iris sphincter is palsied in response to light but still responds to near or to eye movement (a slit-lamp observation), that is evidence of denervation and aberrant reinnervation. If the pupil dilates poorly in the dark and has a clearly stronger reaction to a near stimulus than to light (seen best with infrared video), that indicates abnormality.

REFERENCES 1. Andersen, O., Lindberg, J., Modigh, K., and Reske-Nielsen, E.: Subacute dysautonomia with incomplete recovery. Acta Neurol. Scand. 48:510, 1972.

213

2. Appenzeller, O., and Kornfeld, M.: Acute pandysautonomia: clinical and morphologic study. Arch. Neurol. 29:334, 1973. 3. Blumen, S. C., Feiler-Ofry, V., and Korczyn, A. D.: The pupil cycle time in Horner’s syndrome. J. Clin. Neuro-ophthalmol. 6:232, 1986. 4. Carlson, D. W., and Tychsen, L.: Touching the cornea enhances pharmacologic dilation of the pupil, mainly in the dark iris. Aviat. Space Environ. Med. 60:994, 1989. 5. Cremer, S. A., Thompson, H. S., Digre, K. B., and Kardon, R. H.: Hydroxyamphetamine mydriasis in normal subjects. Am. J. Ophthalmol. 110:66, 1990. 6. Currie, J., and Lessell, S.: Tonic pupil with giant cell arteritis. Br. J. Ophthalmol. 68:135, 1984. 7. Esterly, N. B., Cantolino, S. J., Alte, B. P., and Brusilow, S. W.: Pupillotonia, hyporeflexia, and segmental hypohidrosis: autonomic dysfunction in a child. J. Pediatr. 73:852, 1968. 8. Fletcher, W. A., and Sharpe, J. A.: Tonic pupils in neurosyphilis. Neurology 36:188, 1986. 9. Frewin, D. B., Gilmore, H. R., Ho, J. Q., and Scroop, G. C.: Clinical, physiological and pathological observations in a case of progressive autonomic nervous system degeneration associated with Holmes-Adie syndrome and peripheral neuropathy. Australas. Ann. Med. 17:141, 1968. 10. Friedman, D. I., Fortanasce, V. N., and Sadun, A. A.: Tonic pupils as a result of botulism. Am. J. Ophthalmol. 109:236, 1990. 11. Friedman, J. R., Whiting, D. W., Kosmorsky, G. S., and Burde, R. M.: The cocaine test in normal patients. Am. J. Ophthalmol. 98:808, 1984. 12. Friedman, S. A., Feinberg, R., Podolak, E., and Bedell, R. H.: Pupillary abnormalities in diabetic neuropathy: a preliminary study. Ann. Intern. Med. 67:977, 1967. 13. Harriman, D. G., and Garland, H.: The pathology of Adie’s syndrome. Brain 91:401, 1968. 14. Hedges, T. R., and Gerner, E. W.: Ross’ syndrome (tonic pupil plus). Br. J. Ophthalmol. 59:387, 1975. 15. Hreidarsson, A. B.: Pupil size in insulin-dependent diabetes: relationship to duration, metabolic control, and long-term manifestations. Diabetes 31:442, 1982. 16. Isenberg, S. J., Dang, Y., and Jotterand, V.: The pupils of term and preterm infants. Am. J. Ophthalmol. 108:75, 1989. 17. Kardon, R. H., Corbett, J. J., and Thompson, H. S.: Segmental denervation and reinnervation of the iris sphincter as shown by infrared videographic transillumination. Ophthalmology 105:313, 1998. 18. Kardon, R. H., Denison, C. E., Brown, C. K., and Thompson, H. S.: Critical evaluation of the cocaine test in the diagnosis of Horner’s syndrome. Arch. Ophthalmol. 108:384, 1990. 19. Keane, J. R.: Tonic pupils with acute ophthalmoplegic polyneuritis. Ann. Neurol. 2:393, 1977. 20. Keltner, T. L., Swisher, C. N., Gay, A., and Hepler, R. S.: Myotonic pupils in Charcot-Marie-Tooth disease: successful relief of symptoms with 0.025% pilocarpine. Arch. Ophthalmol. 93:1141, 1975. 21. Kerrison, J. B., Biousse, V., and Newman, N. J.: Isolated Horner’s syndrome and syringomyelia. J. Neurol. Neurosurg. Psychiatry 69:3, 2000. 22. Konig, H., Gassman, H. B., and Jenzer, G.: Ocular involvement in benign botulism B. Am. J. Ophthalmol. 80:430, 1975.

214

Function of the Peripheral Nervous System

23. Kupersmith, M. J., and Aleksic, S. N.: Bilateral pupillary cholinergic supersensitivity in a case of sarcoidosis. Neuro-ophthalmology 4:15, 1984. 24. Lam, B. L., Thompson, H. S., and Corbett, J. J.: The prevalence of simple anisocoria. Am. J. Ophthalmol. 104:69, 1987. 25. Leavitt, J. A., Wayman, L. L., Hodge, D. O., and Brubaker, R. F.: Pupillary response to four concentrations of pilocarpine in normal subjects: application to testing for Adie tonic pupil. Am. J. Ophthalmol. 135:259, 2003. 26. Lepore, F. E.: Enophthalmos and Horner’s syndrome. Arch. Neurol. 40:460, 1983. 27. Loewenfeld, I. E.: The Argyll Robertson pupil, 1869–1969. A critical survey of the literature. Surv. Ophthalmol. 14:199, 1969. 28. Loewenfeld, I. E.: Pupillary changes related to age. In Thompson, H. S. (ed.): Topics in Neuro-Ophthalmology. Baltimore, Williams & Wilkins, p. 124, 1979. 29. Loewenfeld, I. E., and Thompson, H. S.: The tonic pupil: a re-evaluation. Am. J. Ophthalmol. 63:46, 1967. 30. Low, P. A., Dyck, P. J., Lambert, E. H., et al.: Acute panautonomic neuropathy. Ann. Neurol. 13:412, 1983. 31. Manor, R. S., Yassur, Y., Siegal, R., and Ben-Sira, I.: The pupil cycle time test: age variations in normal subjects. Br. J. Ophthalmol. 65:750, 1981. 32. Miller, N. R., and Moses, H.: Ocular involvement in wound botulism. Arch. Ophthalmol. 95:1788, 1977. 33. Miller, R. G., Nielsen, S. L., and Sumner, A. J.: Hereditary sensory neuropathy and tonic pupils. Neurology 26:931, 1976. 34. Miller, S. D., and Thompson, H. S.: Edge-light pupil cycle time. Br. J. Ophthalmol. 62:495, 1978. 35. Monteiro, M. L., Ulrich, R. F., Imes, R. K., et al.: Iron mydriasis. Am. J. Ophthalmol. 97:794, 1984. 36. Morris, J. G. L., Lee, L., and Kim, C. L.: Facial sweating in Horner’s syndrome. Brain 107:751, 1984. 37. Nielsen, P. J.: Upside down ptosis in patients with Horner’s syndrome. Acta Ophthalmol. 61:952, 1983. 38. Okajima, T., Imamura, S., Kawasaki, S., et al.: Fisher’s syndrome: a pharmacological study of the pupils. Ann. Neurol. 2:63, 1977. 39. Patil, P. M., and Jacobowitz, D.: Unequal accumulation of adrenergic drugs by pigmented and nonpigmented iris. Am. J. Ophthalmol. 78:470, 1974. 40. Patil, P. N.: Cocaine-binding by the pigmented and the nonpigmented iris and its relevance to the mydriatic effect. Invest. Ophthalmol. 11:739, 1972. 41. Pilley, S. F., and Thompson, H. S.: Pupillary “dilatation lag” in Horner’s syndrome. Br. J. Ophthalmol. 59:731, 1975. 42. Pomeranz, H.: Isolated Horner syndrome and syrinx of the cervical spinal cord. Am. J. Ophthalmol. 133:702, 2002. 43. Pruett, R. C.: Internal ophthalmoplegia after panretinal therapy. Arch. Ophthalmol. 97:2212, 1979. 44. Purcell, J. J. Jr., Krachmer, J. H., and Thompson, H. S.: Corneal sensation in Adie’s syndrome. Am. J. Ophthalmol. 84:496, 1977. 45. Rabinowich, L., and Mehler, M. F.: Parasympathetic pupillary involvement in biopsy-proven temporal arteritis. Ann. Ophthalmol. 20:400, 1988. 46. Richardson, R. W.: Comparing the mydriatic effect of tropicamide with respect to iris pigmentation. J. Am. Optom. Assoc. 53:885, 1982.

47. Robinson, J., and Fielder, A. R.: Pupillary diameter and reaction to light in preterm neonates. Arch. Dis. Child. 65:35, 1990. 48. Rogell, G. D.: Internal ophthalmoplegia after argon laser panretinal photocoagulation. Arch. Ophthalmol. 97:904, 1979. 49. Rosenberg, M. L.: The friction sweat test as a new method for detecting facial anhidrosis in patients with Horner’s syndrome. Am. J. Ophthalmol. 108:443, 1989. 50. Ross, A. T.: Progressive selective sudomotor denervation: a case with coexisting Adie’s syndrome. Neurology 8:809, 1958. 51. Rubinow, A., and Cohen, A. S.: Scalloped pupils in familial amyloid polyneuropathy. Arthritis Rheum. 29:445, 1986. 52. Salisachs, P., and Lapresle, J.: Argyll-Robertson-like pupils in the neural type of Charcot-Marie-Tooth disease. Eur. Neurol. 16:172, 1977. 53. Schidte, S. N.: Effects on choroidal nerves after panretinal xenon arc and argon laser photocoagulation. Acta Ophthalmol. 62:244, 1984. 54. Selhorst, J. B., Madge, G., and Ghatak, N.: The neuropathology of the Holmes-Adie syndrome. Ann. Neurol. 16:138, 1984. 55. Shy, G. M., and Drager, G. A.: A neurologic syndrome associated with orthostatic hypertension: a clinical pathologic study. Arch. Neurol. 2:511, 1960. 56. Sigsbee, B., Torkelson, R., and Kadis, G.: Parasympathetic denervation of the iris in diabetes mellitus: a clinical study. J. Neurol. Neurosurg. Psychiatry 37:1031, 1974. 57. Smith, S. A., and Smith, S. E.: Evidence for a neuropathic aetiology in the small pupil of diabetes mellitus. Br. J. Ophthalmol. 67:89, 1983. 58. Smith, S. E., Smith, S. A., Brown, P. M., et al.: Pupillary signs in diabetic autonomic neuropathy. Br. Med. J. 2:924, 1978. 59. Spector, R. H., and Bachman, D. L.: Bilateral Adie’s tonic pupil with anhidrosis and hyperthermia. Arch. Neurol. 41:342, 1984. 60. Stark, L., Campbell, F. W., and Atwood, J.: Pupil unrest: an example of noise in a biological servomechanism. Nature 182:857, 1958. 61. Tan, E. T., Lambie, D. G., Johnson, R. H., and Whiteside, E. A.: Parasympathetic denervation of the iris in alcoholics with vagal neuropathy. J. Neurol. Neurosurg. Psychiatry 47:61, 1984. 62. Thomashefsky, A. J., Horwitz, S. J., and Feingold, M. H.: Acute autonomic neuropathy. Neurology 22:251, 1972. 63. Thompson, B. M., Corbett, J. J., Lanning, B. K., and Thompson, H. S.: Pseudo-Horner’s syndrome. Arch. Neurol. 39:108, 1982. 64. Thompson, H. S.: Adie’s syndrome: some new observations. Trans. Am. Ophthalmol. Soc. 75:587, 1977. 65. Thompson, H. S.: Diagnosing Horner’s syndrome. Trans. Am. Acad. Ophthalmol. Otolaryngol. 83:840, 1977. 66. Thompson, H. S.: Segmental palsy of the iris sphincter in Adie’s syndrome. Arch. Ophthalmol. 96:1615, 1978. 67. Thompson, H. S.: Light-near dissociation of the pupil. Ophthalmologica 189:21, 1984. 68. Thompson, H. S.: The pupil cycle time. J. Clin. Neuroophthalmol. 7:38, 1987.

The Pupil 69. Thompson, H. S.: The pupil. In Moses, R. A., and Hart, W. (eds.): Adler’s Physiology of the Eye and Clinical Application. St. Louis, C. V. Mosby, 1992. 70. Thompson, H. S., Digre, K. B., Kardon, R. H., and Cremer, S. A.: Hydroxyamphetamine mydriasis in normal subjects [comment]. Am. J. Ophthalmol. 110:71, 1990. 71. Thompson, H. S., and Mensher, J. H.: Adrenergic mydriasis in Horner’s syndrome: hydroxyamphetamine test for diagnosis of postganglionic defects. Am. J. Ophthalmol. 72:472, 1971. 72. Thompson, H. S., Newsome, D. A., and Loewenfeld, I. E.: The fixed dilated pupil: sudden iridoplegia or mydriatic drops? A simple diagnostic test. Arch. Ophthalmol. 86:21, 1971. 73. Van der Wiel, H. L., and Van Gijn, J.: Horner’s syndrome: criteria for oculosympathetic denervation. J. Neurol. Sci. 56:293, 1982.

215

74. Van der Wiel, H. L., and Van Gijn, J.: Diagnosis of Horner’s syndrome: use and limitations of the hydroxyamphetamine test. J. Neurol. Sci. 73:311, 1986. 75. Van der Wiel, H. L., and Van Gijn, J.: No enophthalmos in Horner’s syndrome. J. Neurol. Neurosurg. Psychiatry 50:498, 1987. 76. Verdick, R. E., and Thompson, H. S.: Infrared videography of the eyes. J. Ophthalmol. Photog. 13:19, 1991. 77. Williams, D., Brust, J. C., Abrams, G., et al.: Landry-GuillainBarre syndrome with abnormal pupils and normal eye movements: a case report. Neurology 29:1033, 1979. 78. Yee, R. D., Trese, M., Zee, D. S., et al.: Ocular manifestations of acute pandysautonomia. Am. J. Ophthalmol. 81:740, 1976. 79. Young, R. R., Asbury, A. K., Adams, R. D., and Corbett, J. L.: Pure pan-dysautonomia with recovery. Trans. Am. Neurol. Assoc. 94:355, 1969.

10 Neural Control of Cardiac Function DAVID J. PATERSON AND JOHN H. COOTE

Organization of Cardiac Vagal and Sympathetic Nervous System Origin of Cardiac Sympathetic Tone Origin of Cardiac Parasympathetic Tone Cardiac Plexus and the “Cardiac Brain” Hypothesis Neurochemical Transmission in the CNS Concluding Remarks on the CNS

Chemical Transmission at the Cardiac Neuroeffector Junction History Postjunctional Signaling Communication between the Cardiac Sympathetic and Parasympathetic Systems Accentuated Antagonism

It is well established that the central and peripheral nervous systems play an important role in regulating the performance of the healthy and diseased heart. This regulation is far reaching, affecting excitability, contraction, blood flow, and cardiac metabolism. Many diseases of the cardiovascular system are also diseases of the autonomic nervous system (dysautonomia), where cardiac dysautonomias may actually influence morbidity and mortality associated with the primary disease itself.17 This chapter is concerned with the neural control of cardiac physiologic function and its relation to disorders of the autonomic nervous system. To appreciate the functional significance of the subject area requires some understanding as to how the cardiac nervous system is organized.

ORGANIZATION OF CARDIAC VAGAL AND SYMPATHETIC NERVOUS SYSTEM The healthy heart is subject to a stream of nerve impulse traffic in both vagus and sympathetic nerves. Resting heart rate (HR) is determined primarily by the magnitude of parasympathetic activity acting directly on the heart or interacting with the sympathetic nerves through several mechanisms in the brain and at the level of the heart. The tonic activity of both systems arises in the central nervous system (CNS) (Fig. 10–1).

Assessing Cardiac Neural Drive: Physiological Outcome Impairment of Cardiac Neural Control: Clinical Outcome

Origin of Cardiac Sympathetic Tone The sympathetic neurons supplying the heart lie in the intermediolateral cell column of gray matter in the upper thoracic segments (T4-T5) of the spinal cord. These fibers are relatively short as they emerge in the white rami communicantes to enter the thoracic sympathetic ganglia, because the ganglia are located close to the spinal cord. They then ascend into the neck via the paravertebral sympathetic chain. The synapse between the pre- and postganglionic cardiac neuron occurs somewhere in the upper thoracic and cervical ganglia. Postganglionic fibers pass gray rami communicantes before forming the cardiac nerves that innervate both pacemaking tissue and the ventricle. The right sympathetic nerves predominantly innervate the sinoatrial node (SAN)137 to increase discharge from the pacemaker (positive chronotrophy). The left sympathetic nerves innervate the atrioventricular node to increase conduction (dromotropism), the excitability of the bundle of His–Purkinje fiber conducting system, and contractility (inotropism) in the atrium and ventricle.83,137 Experimental animal studies suggest that cardiac sympathetic neurons are arranged as a longitudinal column of cells apparently distinct from other columns of sympathetic neurons with different end organ targets.119 They have extensive dendritic fields onto which converge numerous spinal cord and supraspinal afferents. A major input is that from the rostral ventrolateral medulla 217

218

Function of the Peripheral Nervous System

Cerebellum

Baroreceptors IX nerve

ICC

NTS NA

RVL

Brainstem

SAN

FIGURE 10–1 Neural map showing main afferent and efferent pathways that regulate cardiac excitability. The nucleus tractus solitarius (NTS) receives input from baroreceptors that is conveyed by glossopharyngeal (IXth cranial nerve) afferents. It connects with the nucleus ambiguus (NA), which projects via the vagus to the sinoatrial (SA) node of the heart. The connection from NTS to NA might not be direct, as indicated by the dashed projection. Also illustrated are the rostral ventrolateral medulla (RVL) and the intermediolateral cell column (ICC) of the spinal cord. The cerebellum is shown for orientation.

(RVLM), wherein again it is likely that cardiac sympathetic neurons exist as a distinct group.102 It is the input activity in the RVLM projection that most likely determines the level of tonic discharge in the preganglionic neurons.21 The consensus opinion is that this tonic activity is largely dependent on multiple oscillating networks of neurons in the lateral tegmentum of the medulla that fire with a 2- to 6-Hz rhythm that becomes entrained to the arterial baroreceptor input (which is inhibitory and in synchrony with the cardiac cycle).7 This activity is enhanced during inspiration, an effect that is injected at the level of the RVLM.21 Neurochemical transmission between RVLM neuron terminals and spinal sympathetic preganglionic neurons is via the release of glutamate acting on both N-methyl-D-aspartate (NMDA) and non-NMDA receptors.24 Inhibition by the arterial baroreceptor input is mediated by ␥-aminobutyric acid (GABA)151 released by neurons projecting to RVLM neurons from the caudal

ventrolateral medulla, where the neurons receive a direct projection from the nucleus tractus solitarius (NTS).110,139 A further projection from the NTS to GABA interneurons in the spinal cord to inhibit sympathetic preganglionic neurons has also been proposed.85 Numerous other chemical transmitters play a part in modifying the pattern and magnitude of discharge of sympathetic neurons, but little is known about specific influences on cardiac sympathetics. At the level of the RVLM, glutamate, angiotensin II, and vasopressin are excitatory, whereas GABA and nitric oxide (NO) are inhibitory.43 In the spinal cord serotonin, norepinephrine (NE), epinephrine, dopamine, oxytocin, and vasopressin may be released from supraspinal terminals. The monoamines can exert both excitatory and inhibitory actions either indirectly via interneurons or directly via different postsynaptic receptors.18,41 Supramedullary neurons project directly to the spinal cord and release peptides, possibly together with glutamate, which is excitatory. There is also evidence that oxytocin release selectively affects cardiac sympathetic neurons.158 The discharge of sympathetic neurons in the spinal cord is also dependent on NO, which mainly acts to reduce their discharge by causing the release of glycine from spinal interneurons.155,157 A summary of the neuronal pathways regulating cardiac sympathetic preganglionic neurons is illustrated in Figure 10–2.

Origin of Cardiac Parasympathetic Tone Preganglionic cardiac vagal efferents predominantly project from the nucleus ambiguus in humans, although there are also functional projections in many species from the dorsal motor nucleus of the vagus and intermediate zone of the medulla oblongata.57,67 These fibers do not diverge as extensively as those of the sympathetic division, and after exiting the brainstem travel in the Xth cranial nerve, descending in the neck caudally to the common carotid artery. Vagal preganglionic fibers synapse to intracardiac ganglia that are in close proximity to their neuroeffector site. These ganglia in humans are located in the posterior aspect of the atria within the subepicardial layers near the SAN and atrioventricular node. The pattern of vagal innervation is not as evenly distributed compared to the sympathetic innervation.65,71,137 Right-sided innervation is dominant given the location of the SAN as the primary pacemaker for the vagus to decrease excitability and rate (bradycardia). For many years it was thought there was no significant functional vagal innervation beyond the supraventricular structure in humans. This is simply not true.84 There is good evidence for innervation into the septum and subendocardial layers of the human left ventricle. Functionally, vagal stimulation can decrease excitability and contraction of the ventricle. These effects are accentuated in the presence of elevated sympathetic activity and independent of changes in HR brought about by vagal stimulation.86

219

Neural Control of Cardiac Function

NTS

Supramedullary and medullary spinal projecting neurons

Arterial baroreceptor

Lat. tegmental oscillator

Inspiratory N

+

Glutamate

RVLM

+

Cardiac presympathetic neuron

–

+

GABA CVLM

FIGURE 10–2 Schematic summarizing putative neuronal pathways and chemical neurotransmitters determining the tonic activity of cardiac sympathetic nerves. CVLM ⫽ caudal ventrolateral medulla; GABA ⫽ ␥-aminobutyric acid; NTS ⫽ nucleus tractus solitarius; RVLM ⫽ rostral ventrolateral medulla; ⫹, excitation; ⫺, inhibition.

Evidence from a variety of sources indicates that cardiac vagal preganglionic neurons are inherently silent, displaying no spontaneous pacemaker activity. Therefore, their discharge is determined by the activity and chemical transmitters released by neurons that synapse with them.40,104 A summary of the neuronal pathways regulating cardiac parasympathetic preganglionic neurons is illustrated in Figure 10–3. Chief among the excitatory inputs are those related to stimulation of arterial baroreceptors and chemoreceptors, whereas powerful phasic inhibitory inputs arise from central respiratory neurons and airway receptors. As a consequence, in anesthetized animals the firing pattern observed in single cardiac vagal efferent fibers has both a cardiac-related rhythm and a respiratory rhythm.23,76,96,117,136 As stated in the previous section, the effect of these inputs is the opposite to that exerted on cardiac sympathetic neurons. In an elegant series of intracellular experiments in anesthetized animals, the membrane potential of cardiac vagal preganglionic neurons revealed that they exhibit depolarizing potentials correlated with the cardiac cycle, interposed with hyperpolarizations during inspiration. It was shown that these neurons were

+

Monoamines peptides

+

Glutamate

Cardiac preganglionic neuron

–

GABA Heart

receiving inhibitory postsynaptic potentials at a time when the central inspiratory drive was greatest. The cells were most excitable immediately after end inspiration and relatively less excitable as expiration proceeded.40 These studies explained the observation in humans that the cardiac vagal neuron responsiveness to increasing carotid baroreceptor activity (by stretching the arteries with brief neck suction applied externally during controlled-frequency breathing) was greater during expiration than inspiration.32 This effect is referred to as respiratory “gating”95 and is probably the main cause of respiratory sinus arrhythmia.31 However, studies in a variety of vertebrates suggest that there may be two types of cardiac vagal preganglionic neurons: one that displays phasic activity that is cardiorespiratory related and important for coupling cardiac output to ventilation, and another that fires tonically and responds more to states of emergency.141,142 Similar types of cardiac vagal neurons may exist in mammals because, in an exquisite study in anesthetized cats, a population of cardiac vagal neurons was revealed that were only weakly or not at all affected by the respiratory gate.20 As well as arterial baroreceptor, chemoreceptor, and respiratory inputs, a number of other afferent systems can

220

Function of the Peripheral Nervous System

NTS

Lung stretch receptor NO

GABA

– Arterial baroreceptor Arterial chemoreceptor

NO

+

+ + NO

5-HT

NMDA Non NMDA

Inspiratory neuron

+ GABA

GABAA

Cardiac vagal neuron

–

M1

Heart

influence the excitability of cardiac vagal preganglionic neurons when situations demand (e.g., exercise). These are illustrated in the diagram in Figure 10–4.

Cardiac Plexus and the “Cardiac Brain” Hypothesis Although it is widely assumed that neurons in the brainstem and spinal cord are the sole source of cardiac neuronal control, the intrathoracic paravertebral ganglia (containing sympathetic efferent postganglionic neurons) and intrinsic cardiac ganglia (containing parasympathetic efferent postganglionic neurons) are now recognized as being more than efferent relay stations to the heart.5 At the base of the heart lie the cardiac plexuses that are formed by preganglionic and postganglionic sympathetic fibers. There are two divisions within each plexus that are closely interconnected with several small ganglia located within it. The ventral (superficial) portion is formed by the cardiac branch of the left superior cervical ganglion and the lower cervical cardiac branches of the left vagus nerve. The dorsal (deep) aspect is made up of the cardiac branches from the cervical and upper thoracic sympathetic ganglia and the vagus and recurrent laryngeal nerves.106

– Acetyl choline

FIGURE 10–3 Schematic summarizing putative neuronal pathways and chemical neurotransmitters determining the tonic activity of cardiac vagal nerves. GABA ⫽ ␥-aminobutyric acid; 5-HT ⫽ serotonin; M1 ⫽ muscarinic receptor; NMDA, nonNMDA ⫽ two types of glutamate receptor; NO ⫽ nitric oxide; NTS ⫽ nucleus tractus solitarius; ⫹, excitation; ⫺, inhibition.

The idea of a complex nervous system within the heart (“little brain in the heart”) has been recognized for some time, although its precise role is now only being fully appreciated. The principal cells in cardiac ganglia send their axons to pacemaker cells, conducting tissue, and the ventricles. Some ganglia act as interneurons and terminate within the ganglia, whereas others project to other ganglia in the intracardiac plexus. Armour and Ardell’s recent text reviews the role the cardiac neuraxis plays in processing cardiovascular sensory information in order to maintain optimal neural output.5 Whether these local networks of neurons are important during the neural remodeling that takes place with myocardial disease, aging, and neurogenic hypertension has not been firmly established. Nevertheless, the concept of local neural circuits presents an attractive proposition that deserves exploration.

NEUROCHEMICAL TRANSMISSION IN THE CNS In electrophysiologic experiments in rats in which the NTS region was stimulated and postsynaptic responses recorded, cardiac vagal neurons demonstrated excitatory postsynaptic currents mediated by glutamate.98 It seems

221

Neural Control of Cardiac Function

+ INSPIRATORY ACTIVITY

– –

Arterial chemoreceptors

+

– Arterial baroreceptors

+

–

Lung stretch receptors

Superior laryngeal receptors

Post-inspiratory neurons Muscle ergoreceptors

+

FIGURE 10–4 Diagram illustrating the interaction of a variety of peripheral sensory receptors in inspiratory activity and cardiac vagal activity as discussed in the text. ⫹, excitation; ⫺, inhibition.

+

likely that the physiologic role of this pathway is to mediate baroreceptor- and chemoreceptor-induced vagal bradycardia. NO is also likely to be an important mediator of excitation. In anesthetized rats microinjection of NO donors or substrates into the nucleus ambiguus results in a profound vagal bradycardia, and this effect can be prevented by prior microinjection of NO synthesis inhibitors into the same region.38,124 In anesthetized animals NO may be tonically released, because NO synthesis inhibitors microinjected into the nucleus ambiguus on their own lead to a small increase in HR.124 This effect of NO on cardiac vagal efferents is in contrast to its endogenous role in the NTS, the primary relay center for the baroreceptor afferents. Here stimulation of NO release leads to a reduction of a baroreceptor-elicited vagal bradycardia. This effect in the NTS is dependent on NO stimulating GABA release from interneurons that inhibit transmission in the baroreceptor pathway.115 GABA neurons are also located within the nucleus ambiguus, where they make synaptic contact with cardiac vagal preganglionic neurons.28,99 GABA applied in the vicinity of these neurons inhibits their firing. The GABAA receptor antagonist bicuculline prevents this action and also results in an increased cardiac vagal activity, suggesting that there is a tonically active GABA neuron input.66 The functional implication of this input is unclear, but it could mediate the GABAergic inhibitory effects on cardiac vagal neurons shown to occur on stimulation of the NTS.152 A physiologic role of this pathway may be to transmit the cardiac inhibitory action of lung stretch receptors, which have a primary relay in the NTS. A further possibility is that the GABA neurons transmit the inhibition of vagal tone that occurs during physical exercise.

–

–

+

+

+

–

CARDIAC VAGAL PREGANGLIONIC NEURON

Acetylcholine (ACh) has been identified as a putative inhibitory transmitter in the nucleus ambiguus. Its local application reduces the discharge of identified cardiac vagal preganglionic neurons, an action blocked by atropine, indicating a muscarinic receptor is involved.40 Atropine applied alone enhances the discharge of the neurons, particularly during inspiration when they are normally silent, so that there is a complete suppression of respiratory modulation. Such evidence confirms the idea that inspiratory neurons that are cholinergic are directly inhibiting cardiac vagal preganglionic neurons.40 These results also provide an explanation for the observation that very low doses of atropine given in both experimental animals and humans result in augmentation of cardiac vagal activity, probably by an action in the CNS.69,120 The straightforward explanation of the actions of ACh may be more complex than this, because recent in vitro studies in rat brain slices indicate that nicotine, a cholinergic agonist, enhances glutamate neurotransmission to activate cardiac vagal neurons.109 However, until in vivo studies are done, it is too early to discuss the functional significance of these data. Serotonin-containing nerve terminals make synaptic contact with cardiac vagal neurons, and it appears that serotonin excites these neurons via activation of 5-hydroxytryptamine (5-HT)1A receptors and 5-HT7 receptors. Functionally it seems that part of this input is utilized by cardiopulmonary receptors that, when stimulated, elicit a bradycardia as part of the classic BezoldJarisch reflex.152 Microinjections of enkephalins and opiate agonists into the nucleus ambiguus of dogs cause profound bradycardia,77 and intravenous or intracerebroventricular injections of these compounds excite cardiac vagal efferent fibers.61 It appears there is an endogenous opioid system

222

Function of the Peripheral Nervous System

responsible for this enhancing effect on cardiac vagal efferent activity, because recent studies show that low doses of opioid antagonists into the CNS, but not intravenously, block the decrease in HR that accompanies hemorrhagic shock.3 A direct facilitatory action of opioids on cardiac vagal preganglionic neurons was confirmed in a recent study of these neurons recorded in vitro in slices of the rat medulla. This effect of the endorphins is by an inhibitory modulation of voltage-gated calcium currents that secondarily may cause a reduction in calcium-activated potassium currents.63 The effect of endorphins may be further enhanced by a presynaptic action to cause a reduction in GABA release. The important role of GABA in regulating cardiac vagal preganglionic neuron activity is also emphasized by observing the actions of anesthetics.64 A variety of anesthetics cause a clear depression of activity in cardiac vagal efferent fibers, resulting in increased HR.60 Studies on single cardiac vagal neurons recorded in vitro indicate a possible mechanism for this effect. They show that pentobarbital increases the duration of spontaneous inhibitory postsynaptic currents at concentrations that are clinically relevant. This action is via the GABAA receptor because it is prevented if the receptor subunits are modified to make them insensitive.63 The GABAA receptor is also a target for the benzodiazepine drugs, which increase its sensitivity by binding to a specific protein subunit of the receptor.123 As a consequence, they augment the effect of GABA, and it has been shown that in humans intravenous benzodiazepines such as midazolam increase HR by decreasing cardiac vagal tone.37

Concluding Remarks on the CNS This account may give the impression that more details are known about the properties and organization of cardiac vagal preganglionic neurons than cardiac sympathetic neurons. This is because the former are easier to identify as a select group, whereas the characteristics of cardiac sympathetic neurons in general have to be derived from the much larger literature on the sympathetic nervous system. Nonetheless, we can deduce with reasonable certainty— and this has been shown experimentally—that the major afferent regulators of the two populations of cardiac neurons elicit reciprocal actions, although unusually both cardiac vagal and sympathetic activity can increase simultaneously under certain experimental conditions.74 Missing from this account is a description of the role of endogenous brain peptides such as vasopressin and angiotensin. Details of these are outside the scope of this account, but it should not be forgotten that a peptide such as angiotensin can indirectly inhibit cardiac vagal neurons144 and directly excite presympathetic neurons in the RVLM.87 These actions may go awry in disease such as hypertension or heart failure and could explain some of the beneficial

effects of angiotensin-converting enzyme inhibitors that might cross the blood-brain barrier to reduce the production of angiotensin, and hence increase cardiac vagal tone while also decreasing cardiac sympathetic tone.144

CHEMICAL TRANSMISSION AT THE CARDIAC NEUROEFFECTOR JUNCTION Because of its relative ease of access, the cardiac neuroeffector junction became the synapse where the concept of neurotransmission was proved, and in so doing established one of the major physiologic principles of the 20th century, which pharmacologists elegantly exploited for therapeutic purposes. It is worthwhile highlighting briefly the evolution of chemical transmission in a modern text to place recent findings in perspective.

History In the 17th century, Willis first described the vagi or “wandering nerves,” and observed a projection to the aortic arch and speculated that it “may react to changes in the pulse” because bilateral denervation caused “great trembling” in the dog heart. His pupil Richard Lower went on to show that vagal activation affected heartbeat, but it was not until the 1800s that Weber provided conclusive evidence for the right vagus causing cardiac inhibition. This effect could be inhibited by atropine and mimicked by muscarine, strongly suggesting that the vagus releases a chemical neurotransmitter. The Nobel Prize–winning work of Otto Loewi in 1921 proved this concept to be true. Stimulation of the right vagus nerve of an isolated frog heart decreased HR, and the transfer of the perfusate to a second frog heart produced bradycardia in the donor. The chemical substance in the perfusate responsible for this action was named “vagusstoff” and was identified as ACh by Dale in the 1930s. A similar fascinating path of discovery occurred in parallel for the sympathetic nervous system. In 1543 Vesalius distinguished the sympathetic nerves from the vagus, with Willis ascribing some cardiac function to them. However, it was not until Oliver and Schafer, in 1895, had discovered the action of adrenalin and Gaskell, in 1886, divided the nervous system into two distinct antagonistic systems that Langley, in 1898, coined the term autonomic nervous system. T. R. Elliot’s classic experiment in 1905 in Langley’s laboratory in Cambridge suggested that adrenaline or its immediate precursor might be the mediator at the sympathetic neuroeffector junction. This study paved the way for von Euler in 1946 to conclusively prove that norepinephrine (NE) was the sympathetic transmitter, for which he received the Nobel Prize for Medicine or Physiology in 1970.

Neural Control of Cardiac Function

We now know that postganglionic parasympathetic ACh and postganglionic sympathetic NE are packaged into vesicles, and, on depolarization of the nerve terminal, influx of calcium through voltage-gated calcium channels promotes their fusion to the neuronal membrane, causing the release of the transmitter. Almost immediately (2.5 s) after ACh is released it is hydrolyzed by acetylcholinesterase,89,93 whereas NE has a longer active half-life and must be taken up by the sympathetic varicosity by active processes.75 At rest the inhibitory action of the vagus on the SAN dominates over the excitatory action of the sympathetic.82 If vagal tone is removed following the muscarinic antagonist atropine, HR will increase from around 70 beats/min to 130 beats/min in healthy humans. ␤-Adrenergic blockade with propranolol will reduce HR to 58 beats/min78 (Fig. 10–5).

Postjunctional Signaling Acetylcholine diffuses across the synaptic cleft to bind to SAN cell muscarinic M2 receptors, coupled to inhibitory heterotrimeric G proteins. This brings about changes in a range of pacemaking currents that contribute to the spontaneous depolarization of these cells during diastole and the rhythmic generation of action potentials. The upstroke of the SAN action potential is carried by the L-type calcium current (ICaL)73,90 and repolarization by the delayed rectifier potassium current (IK).48,72,126 Deactivation of IK during diastole unmasks a background sodium current (IbNa), which

Vagus nerve Sympathetic varicosity NCa NCa

↑Ca2+ ↑Ca2+ NE +

+ Norepinephrine (NE) β1 M2

ACh

Acetylcholine (ACh)

Cardiac myocyte

HR

HR

Sympathetic stimulation

Vagal stimulation

FIGURE 10–5 Postganglionic sympathovagal signaling on cardiac myocyte. Diagram illustrates voltage activation of neuronal calcium channels, resulting in increased intracellular calcium (Ca2⫹) that facilitates the release of neurotransmitter to either excite (norepinephrine) or inhibit (acetylcholine) cardiac excitability.

223

contributes to diastolic depolarization.45 This is assisted by activation of several other currents, including the hyperpolarization-activated current (If), the sustained inward current (Ist), the T-type calcium current (ICaT), and current generated by the sodium-calcium exchanger (INCE).12,105,156 M2-receptor stimulation is coupled via G␤␥ to an inwardly rectifying potassium current (IKACh), increasing the open probability of the channel proteins GIRK1 and GIRK4.125 This hyperpolarizes SAN cells and increases the time taken to reach action potential threshold. The latency of response to increase R-R intervals is around 200 ms and is a direct result of activation of IKACh.159 These receptors are also coupled to G␣i2, which inhibits the activity of adenylate cyclase. If, one of many currents contributing to the diastolic depolarization, can be directly modulated by nucleotides, such as cyclic GMP (cGMP) and cyclic AMP (cAMP).25 By inhibiting adenylate cyclase and decreasing cAMP levels, acetylcholine reduces If by shifting its activation curve to more negative potentials.26,27 The decrease in cAMP also reduces protein kinase A–dependent phosphorylation of L-type calcium channels. It is not clear how changes in ICaL may alter HR. This may occur via changing intracellular calcium handling and therefore calcium-dependent currents and exchangers involved in diastolic depolarization, by altering action potential threshold, or, more controversially, by contributing to the diastolic depolarization itself.121 The role of these currents in normal pacemaking and its cholinergic modulation in different cells throughout the SAN is an area of ongoing debate, but they nevertheless contribute to the indirect modulation of HR during vagal activation via second messengers (Fig. 10–6).62 Released NE from postganglionic sympathetic varicosities binds to ␤-adrenergic receptors on both pacemaking cells and ventricular myocytes. ␤-Receptor stimulation of these cells is coupled to stimulatory G protein (Gs). When the agonist binds to the receptor, it displaces GDP with GTP and activates the stimulatory alpha subunit (G␣s-GTP) to promote an increase in adenylyl cyclase and the formation of cAMP. This cyclic nucleotide acts as a second messenger to modulate the inward calcium current via protein kinase A.138,154 This in turn facilitates Ca2⫹induced Ca2⫹ release from the sarcoplasmic reticulum (SR) to provide calcium for binding to contractile proteins to increase force of contraction.51 A rise in calcium will also increase HR because the upstroke of the pacemaker potential is calcium dependent.62 cAMP can act directly on If to increase its activation kinetics to more positive potentials, thereby increasing pacemaking. It also acts on phospholamban phosphorylation to increase calcium reuptake by the SR, thereby facilitating relaxation (lusitropic effect).50 Although NE activation of Gs can be directly coupled to ICaL, functional activation takes several (1 to 3) seconds compared to the millisecond activation of the vagus. This is because the sympathetic system

224

Function of the Peripheral Nervous System

ACh β1 Right vagus nerve

Gs

+

AC

Gi

–

M2

Gk

NOS-III Sinoatrial node cell

NO ? – PDE2

Sinoatrial node

PKA

cAMP

sGC cGMP +

PDE3 ICaL

IKACh

If

0 pA -500

+50

Sinoatrial node cell action potentials ACh

0 mV -50 -100

0

300 ms

FIGURE 10–6 Cholinergic modulation of sinoatrial node pacemaking currents. Stimulation of the right vagus nerve leads to the release of acetylcholine (ACh), which binds to muscarinic M2 receptors on sinoatrial node pacemaking cells. This decreases the rate of spontaneous action potential generation by direct G protein (Gk), gating of an inward rectifying potassium current (IKACh), and G protein (Gi) inhibition of adenylyl cyclase (AC) to decrease cAMP- and protein kinase A (PKA)–dependent stimulation of the hyperpolarization-activated current (If) and the L-type calcium current (ICaL). M2 receptor stimulation may also activate endothelial nitric oxide synthase (NOS-III), which increases cGMP levels via NO-dependent stimulation of soluble guanylyl cyclase (sGC). This may inhibit ICaL or If in the presence of high levels of ␤1-adrenergic receptor stimulation via cGMP-dependent stimulation of phosphodiesterase 2 (PDE2) and a decrease in cAMP, although this is controversial. (From Herring, N., Danson, E. J. F., and Paterson, D. J.: Cholinergic control of heart rate by nitric oxide is site specific. News Physiol. Sci. 17:204, 2002, with permission.)

predominantly mobilizes second messenger pathways to activate ion channels, whereas the functional action of the vagus does not rely on the involvement of second messenger phosphorylation pathways to decrease HR.

COMMUNICATION BETWEEN THE CARDIAC SYMPATHETIC AND PARASYMPATHETIC SYSTEMS The close proximity of sympathovagal postganglionic terminal varicosities allows for axoaxonic presynaptic communication. Both NE and ACh can inhibit their own release (autoinhibition) and reciprocally modulate each other’s release. Norepinephrine can bind presynaptically to ␣2 receptors to inhibit its own release by decreasing adenylyl cyclase–cAMP kinase activation of neuronal

calcium channels. Cross talk occurs when ACh binds to presynaptic M2 receptors on sympathetic varicosities to inhibit the release of NE, whereas NE can inhibit ACh release via prejunctional a receptors on vagal terminals.49,81,92,154 Putative cotransmitters have also been identified in both sympathetic and vagal terminals. Sympathetic varicosities contain neuropeptide Y, and vasoactive intestinal polypeptide (VIP) has been identified in vagal terminals.2,153 Because they do not fulfill all the traits of classic neurotransmitters, these peptides probably act as neuromodulators since they are not released in significant quantity unless high-frequency stimulation is performed. Functionally neuropeptide Y can inhibit the release of both NE and ACh94; conversely, VIP can cause a positive chronotropic and inotropic response. Indeed, VIP may be responsible for “vagal-induced tachycardia” when stimulation is stopped.153 ATP can also act on presynaptic

Neural Control of Cardiac Function

purinergic receptors to modulate release of transmitters. Emerging evidence suggests that the gaseous molecule NO is an important messenger because it acts both as a paracrine agent (cotransmission) and as an autocrine agent to modulate neurotransmission. Functionally, NO activates the second messenger cascade involving cGMP signaling to phosphodiesterase and protein kinases. Presynaptically NO inhibits the release of NE130 and decreases sympathetic-induced tachycardias.16 Conversely, it facilitates the release of ACh55 to accentuate vagal-induced bradycardia.54 NO may also have the capacity to diffuse to the postsynaptic cell to inhibit or excite pacemaking. Importantly, its differential action is now recognized as being isoform and site specific, because microdomains of its enzyme nitric oxide synthase (NOS) seem to be critical for encoding function.115

Accentuated Antagonism Both limbs of the cardiac autonomic nervous system act in a synergistic manner, resulting in the phenomenon of accentuated antagonism, whereby the negative chronotropic and inotropic action of the vagus is more pronounced when sympathetic tone is high.79,80 The mechanism underlying this probably involves NO-cGMP, although the exact location of its action is not yet firmly established. Some authors,46,47 although not all,146 support the idea that postjunctional NO is critical because NOS inhibition abolishes inhibition of ICaL by ACh analogues only when ␤-adrenergic tone is high in pacemaking cells. However, recent work suggests that the main functional role of accentuated antagonism occurs presynaptically where NO facilitates the release of ACh during nerve stimulation—effects not mimicked by bath-applied transmitter.115 Disruption of this pathway both anatomically and physiologically may be central to the vagal dysfunction that is seen in hypertension. Conversely, upregulation of peripheral parasympathetic signaling by NO-cGMP is a key component of enhanced vagal responsiveness brought about by physical training.22 Understanding this pathway is clearly important because impaired cardiac parasympathetic activity is a powerful negative prognostic indicator that can predict sudden cardiac death.17

ASSESSING CARDIAC NEURAL DRIVE: PHYSIOLOGIC OUTCOME At a given point in time, HR depends on the net excitatory and inhibitory action of sympathovagal balance. This is controlled by the CNS, which in turn depends on reflex inputs from cardiopulmonary receptors, arterial baroreflexes, and muscle ergoreceptors. Assessing cardiac neural drive noninvasively often relies on methods that can dissect out variabil-

225

ity in both HR and arterial pressure. Rapid changes in rate are predominately due to the vagus, with slower responses being attributed to the sympathetic.84,122 One simple measure is to study the effect of respiration on R-R interval. Inspiration is associated with a decrease and expiration with an increase in R-R interval (respiratory sinus arrhythmia). This response is abolished by atropine but unaffected by propranolol, suggesting involvement of vagal efferents.35,68 However, there is poor reproducibility of responses when respiratory frequency and tidal volume are high.4 Rapid changes in body posture8,35 as well as dynamic36 and isometric39 exercise can also be used to unmask the role of vagal efferents on HR variability. Quantification of HR variability using fast Fourier transform spectral analysis is useful in unmasking autonomic impairment in various pathophysiologic states.53,103,107,111,150 Two main peaks are identified with spectral analysis of cardiovascular waveforms: a high-frequency peak (0.2 to 0.4 Hz) that coincides with breathing frequency and is predominately vagal in origin, and a low-frequency fluctuation (0.03 to 0.15 Hz) that relates to sympathetic activity.53,113,116 Changes in the low-frequency/high-frequency ratio of the R-R interval variability are thought to reflect altered cardiac sympathovagal balance.10,19,52,128 Spectral analysis of cardiovascular waveforms from Holter monitoring studies have shown that patients with heart failure or diabetic neuropathy or following heart transplantation have low HR variability with an impaired high-frequency component (vagal dysfunction) and augmented low-frequency component (enhanced sympathetic activity).9–11,19,34,70,128 In patients with heart failure, augmented sympathetic responses have been correlated with enhanced circulating catecholamines and sympathetic activity (detected by microneurographic recordings).70 This approach has proved popular in the clinical environment and is often used as a prognostic indicator in patients suffering from several cardiac pathologies.11 However, there are pitfalls when using spectral analysis of HR variability during exercise; in particular, evidence indicates that, as the cardiac vagal tone falls with increasing levels of exercise, a greater percentage of the residual power of the high-frequency component may be due to non-neural mechanisms.14 Several methods are available to activate cardiac autonomic reflexes in humans. Infusion of vasoactive drugs that either increase or decrease vascular resistance can activate the vagal or sympathetic limbs of the arterial baroreflex, respectively. The advantage of this technique is that it is relatively easy to use and is minimally invasive, requiring an intravenous infusion or bolus of agents. Arterial pressure can be recorded directly via a catheter, although a Finipres gives a reasonably accurate record of changes in pressure.30,100,135 The main assumption in using vasoactive drugs is that they have no direct action on the heart itself. Phenylephrine or angiotensin has been used to raise blood pressure, although the latter can

226

Function of the Peripheral Nervous System

result in a small increase in HR and also cause the prejunctional release of NE.1 Nitrovasodilators, which release NO, are also not without problems; recent work has shown them to have a direct positive chronotropic effect.56,108 Interpretation of results using nitrovasoactive compounds must therefore be viewed with caution if they are infused over several minutes, because they do not provide an accurate quantitative assessment of cardiac neural drive.15 Furthermore, they only assess the cardiac component and not the vascular component of the reflex. The neck chamber technique overcomes many of the problems of vasoactive drugs. This ingenious approach that was developed by Ernsting and Parry in 1957 allows an airtight chamber applied to the subject’s neck to either increase or decrease carotid sinus transmural pressure, thereby activating or deactivating the high-pressure receptors.33 The downside is that the technique is relatively cumbersome and may not provide a discrete baroreceptor stimulus because transmission of negative pressure is incomplete, with only 65% of the applied pressure being sensed by the carotid arteries.29,98 Both head-up tilt and lower body negative pressure can displace significant blood volume to activate both low- and high-pressure receptors to test neural control of HR. Conversely, the Valsalva maneuver, cold pressure test, and isometric exercise can all be used to activate sympathetic efferent activity. These techniques are discussed in more detail in other chapters.

IMPAIRMENT OF CARDIAC NEURAL CONTROL: CLINICAL OUTCOME Is cardiac sympathovagal innervation clinically important? When the body needs to be in the rest-and-digest phase, the parasympathetic nervous system dominates; conversely, when alertness and activity are required, the sympathetic nervous system prevails. It is generally accepted that high sympathetic activity can facilitate the occurrence of ventricular arrhythmia, whereas vagal activation increases electrical stability and minimizes the incidences of ventricular tachyarrhythmias and fibrillation.118,138 Several pathophysiologic states or trauma can disrupt cardiac sympathovagal communications. Cardiac autonomic disorders can either be localized or generalized. Claude Bernard in 1854 probably described the first case of generalized autonomic failure when he showed that transection of the cervical spinal cord caused a marked fall in blood pressure as a result of loss of sympathetic vasoconstrictor tone, decreasing peripheral vasculature resistance. At the level of the heart, surgical (heart transplantation) or pharmacologic (double autonomic blockade with atropine and propranolol) denervation unmasks the intrinsic rate (approximately 90 beats/min) resulting from the inherent automaticity of the SAN,

which can function independently from the autonomic nervous system. Following heart transplantation or surgical sympathectomy, patients can experience localized autonomic abnormality that may result in life-threatening events, as in chronotropic denervation supersensitivity to adrenergic stimulation. Myocardial ischemia and infarction also create supersensitivity to catecholamines.59,131,147 These enhanced sympathoexcitatory responses lead to greater autonomic heterogeneity, increasing the propensity for arrhythmia and sudden cardiac death.147,160 Various peripheral afferent neuropathies, such as ascending polyneuritis (as in Guillain-Barré syndrome), porphyria, tabes dorsalis, and carotid sinus hypersensitivity, and efferent neuropathies such as diabetic autonomic neuropathy can result in fixed tachycardia at rest and sympathovagal imbalance as detected by power spectral analysis of heart variability.42,88,91,134 In many cases the high-frequency component of the power spectrum is absent, indicating vagal impairment with enhanced indices of sympathetic tone, all decreasing ventricular fibrillation threshold.6,44,97 In diabetics with autonomic neuropathy, there is a high incidence of sudden cardiac death, although cardiovascular reflexes are abnormal in this group before peripheral autonomic neuropathy appears.91 Whether the high mortality and morbidity are a direct consequence of dysautonomia affecting afferent and/or efferent limbs of the reflex arch or a defect in central processing has not been firmly established. Cardiac dysautonomia is also present in other disease states, such as heart failure, hypertension, ischemic heart disease, myocardial infarction, and Shy-Drager syndrome (neurologic manifestations involving striatonigral degeneration and Parkinsonian features, olivopontocerebellar atrophy, or the combination of the two, known as multiple system atrophy).101 Emerging evidence from the new subspeciality of neurocardiology strongly supports a major role for the CNS in the etiology of many cardiac arrhythmias via dysautonomia. Subarachnoid hemorrhage,114,145 cerebral vascular trauma especially to the prefrontal areas, and epilepsy145 may all cause profound alterations in cardiac rhythm. Stimulation of the defense area (hypothalamic region and amygdala) initiates arrhythmia in pigs,13,133 an effect blocked by propranolol.132 Stereotactic surgery in awake patients undergoing implantation of deep brain–stimulating electrodes for Parkinson’s disease has shown that activation of midbrain nuclei causes increases in HR and arterial blood pressure, indicating that complex functional circuits are directly coupled to autonomic outflow from this part of the brain.143 Excessive activation of the sympathetic nervous system can cause marked prolongation of the Q-T interval, as can localized central lesions.112,129,140,147 This facilitates the occurrence of afterdepolarizations caused by abnormal calcium loading in ventricular myocytes and can be amplified by hypokalemia (which further lengthens action potential

Neural Control of Cardiac Function

duration)149 and digitalis. When psychological stress,147 exercise, or even rapid-eye-movement sleep148—all events associated with high sympathetic activity—is experienced by a patient with long Q-T syndrome (disruption of the HERG protein that is responsible for the repolarization current IKr127), there is an increased incidence of potentially lethal ventricular arrhythmia. ␤-Adrenergic blockade minimizes the adverse affects of abnormal sympathetic activation to reduce sympathetic outflow and inhibit the action of catecholamines at the neuroeffector site. Interestingly, vagal stimulation and exercise training have been shown to be as effective as ␤-adrenergic blockade in protecting against the arrhythmogenic effect of myocardial ischemia.58 Indeed, high cardiac vagal tone and responsiveness, both characteristics of the athletic phenotype, are now becoming well-established independent positive prognostic markers against sudden cardiac death.17 Whether a centrally induced asymmetrical increase in cardiac sympathetic activity is involved in the etiology of sudden infant death syndrome and sudden adult death syndrome is not known.

REFERENCES 1. Abboud, F. M., and Chapleau, M. W.: Effects of pulse frequency on single unit activity during sine-wave and natural pulses in dogs. J. Physiol. (Lond.) 401:295, 1988. 2. Allen, J. M., Gjorstrup, P., Bjorkman, J. A., et al.: Studies on cardiac distribution and function of neuropeptide Y. Acta Physiol. Scand. 126:405, 1986. 3. Ang, K. K., McRitchie, R. J., Minson, J. B., et al.: Activation of spinal opioid receptors contributes to hypotension after haemorrhage in conscious rats. Am. J. Physiol. 276: H1552, 1999. 4. Angelone, A., and Coulter, N. A.: Respiratory sinus arrhythmia: a frequency dependent phenomenon. J. Appl. Physiol. 19:479, 1964. 5. Armour, J. A., and Ardell, J. L.: Preface. In Armour, J. A., and Ardell, J. L. (eds.): Basic & Clinical Neurocardiology. Oxford, UK, Oxford University Press, p. v, 2004. 6. Armour, J. A., Hageman, G. R., and Randall, W. C.: Arrhythmias induced by local cardiac nerve stimulation. Am. J. Physiol. 223:1068, 1972. 7. Barman, S. M., Gebber, G. L., and Orer, H. S.: Medullary lateral tegmental field: an important source of basal sympathetic nerve discharge in the cat. Am. J. Physiol. 278:R995, 2000. 8. Bellavere, F., and Ewing, D. J.: Autonomic control of the immediate heart rate response to lying down. Clin Sci. 62:57, 1982. 9. Bigger, J. T. Jr., Fleiss, J. L., Steinman, R. C., et al.: Correlation among time and frequency domain measures of heart rate period variability two weeks after acute myocardial infarction. Am. J. Cardiol. 69:891, 1992. 10. Bigger, J. T. Jr., Fleiss, J. L., Steinman, R. C., et al.: Frequency domain measures of heart period variability and mortality after myocardial infarction. Circulation 85:164, 1992.

227

11. Bigger, J. T. Jr., La Rovere, M. T., Steinman, R. C., et al.: Comparison of baroreflex sensitivity and heart period variability after myocardial infarction. J. Am. Coll. Cardiol. 14:1511, 1989. 12. Brown, H., and Difrancesco, D.: Voltage-clamp investigations of membrane currents underlying pace-maker activity in rabbit sino-atrial node. J. Physiol. (Lond.) 308:331, 1980. 13. Carpeggiani, C., Landisman, C., Montaron, M. F., and Skinner, J. E.: Cryoblockade in the limbic brain (amygdala) prevents or delays ventricular fibrillation after coronary artery occlusion in psychologically stressed pigs. Circ. Res. 70:600, 1992. 14. Casadei, B., Moon, J., Johnston, J., et al.: Is respiratory sinus arrhythmia a good index of cardiac vagal tone in exercise? J. Appl. Physiol. 81:556, 1996. 15. Casadei, B., and Paterson, D. J.: Should we still use nitrovasodilators to test baroreflex sensitivity? J. Hypertens. 18:3, 2000. 16. Choate, J. K., and Paterson, D. J.: Nitric oxide inhibits the positive chronotropic and inotropic responses to sympathetic nerve stimulation in the isolated guinea-pig atria. J. Auton. Nerv. Syst. 75:100, 1999. 17. Cole, C. R., Blackstone, E. H., Pashkow, F. J., et al.: Heartrate recovery immediately after exercise as a predictor of mortality. N. Engl. J. Med. 1341:1351, 1999. 18. Coote, J. H.: The organisation of cardiovascular neurones in the spinal cord. Rev. Physiol. Biochem. Pharmacol. 110:148, 1988. 19. Cripps, T. R., Malik, M., Farrell, T. G., and Camm, A. J.: Prognostic value of reduced heart rate variability after myocardial infarction: clinical evaluation of a new analysis method. Br. Heart J. 65:14, 1991. 20. Daly, M de B., and Kirkman, E.: Differential modulation by pulmonary stretch afferents of some reflex cardioinhibitory responses in the cat. J. Physiol. (Lond.) 417:323, 1989. 21. Dampney, R. A. L.: Functional organisation of central pathways regulating the cardiovascular system. Physiol. Rev. 74:323, 1994. 22. Danson, E. J. F., and Paterson, D. J.: Enhanced neuronal nitric oxide synthase expression is central to cardiac vagal phenotype in exercise-trained mice J. Physiol. (Lond.) 546:225, 2003. 23. Davidson, N. S., Goldner, S., and McCloskey, D. I.: Respiratory modulation of baroreceptor and chemoreceptor reflexes affecting heart-rate and cardiac vagal efferent nerve activity. J. Physiol. (Lond.) 259:523, 1976. 24. Deuchars, S. A., Morrison, S. F., and Gilbey, M. P.: Medullaryevoked EPSPs in neonatal rat sympathetic preganglionic neurones in vitro. J. Physiol. (Lond.) 487:453, 1995. 25. DiFrancesco, D., and Tortora, P.: Direct activation of cardiac pacemaker channels by intracellular cyclic AMP. Nature 351:145, 1991. 26. DiFrancesco, D., and Tromba, C.: Inhibition of the hyperpolarization-activated current (If) induced by acetylcholine in rabbit sino-atrial node myocytes. J. Physiol. (Lond.) 405:477, 1988. 27. DiFrancesco, D., and Tromba, C.: Muscarinic control of the hyperpolarization-activated current (If) in rabbit sino-atrial node myocytes. J. Physiol. (Lond.) 405:493, 1988.

228

Function of the Peripheral Nervous System

28. DiMicco, J. A., Gale, K., Hamilton, B. L., and Gillis, R. A.: GABA receptor control of parasympathetic outflow to heart: characterisation and brainstem location. Science 204:1106, 1979. 29. Eckberg, D. L.: Temporal response patterns of the human sinus node to brief carotid baroreceptor stimuli. J. Physiol. (Lond.) 259:769, 1976. 30. Eckberg, D. L.: Parasympathetic cardiovascular control in human disease: a critical review of methods and results. Am J. Physiol. 239:H581, 1980. 31. Eckberg, D. L.: The human respiratory gate. J. Physiol. (Lond.) 548:339, 2003. 32. Eckberg, D. L., and Orsham, C. R.: Respiratory and baroreceptor reflex interactions in man. J. Clin. Invest. 59:780, 1977. 33. Ernsting, J., and Parry, D. J.: Some observations on the effects of stimulating the stretch receptors in the carotid artery of man. J. Physiol. (Lond.) 137:45P, 1957. 34. Ewing, D. J., Campbell, I. W., Murray, A., et al.: Immediate heart-rate response to standing: simple test for autonomic neuropathy in diabetes. Br. Heart J. 1:145, 1978. 35. Ewing, D. J., and Clarke, B. F.: Diagnosis and management of diabetic autonomic neuropathy. Br. Med. J. 285:916, 1982. 36. Fagraeus, L., and Linnarsson, D.: Autonomic origin of heart rate fluctuations at the onset of muscular exercise. J. Appl. Physiol. 40:679, 1976. 37. Farmer, M. R., Vaile, J. C., Osman, F., et al.: A central ␥-aminobutyric acid mechanism in cardiac vagal control in man revealed by studies with intravenous midazolam. Clin. Sci. 95:241, 1998. 38. Fletcher, J., Moody, W. E., Chowdhary, S. C., and Coote, J. H.: Enhancement of the baroreceptor heart rate reflex by nitric oxide in the nucleus ambiguus of anaesthetised rats [abstract]. J. Physiol. (Lond.) 544:33P, 2002. 39. Friedman, D. B., Jensen, F. B., Mitchell, J. H., and Secher, N. H.: Heart rate and arterial blood pressure at the onset of static exercise in man with complete neural blockade. J. Physiol. (Lond.) 423:543, 1990. 40. Gilbey, M. P., Jordan, D., Richter, D. W., and Spyer, K. M.: Synaptic mechanisms involved in the inspiratory modulation of vagal cardio-inhibitory neurones in the cat. J. Physiol. (Lond.) 356:65, 1984. 41. Gladwell, S. J., and Coote, J. H.: Inhibitory and indirect excitatory effects of dopamine on sympathetic preganglionic neurones in the neonatal rat spinal cord in vitro. Brain Res. 818:397, 1999. 42. Greenland, P., and Griggs, R. C.: Arrhythmic complications in the Guillain-Barré syndrome. Arch. Intern. Med. 140:1053, 1980. 43. Guyenet, P. G., and Stornetta, R. L.: The presympathetic cells of the rostral ventrolateral medulla (RVLM): anatomy, physiology and role in the control of the circulation. In Dun, N. J., Machado, B. H., and Pilowsky, P. M. (eds.): Neural Mechanisms of Cardiovascular Regulation. Boston, Kluver, p. 187, 2004. 44. Hageman, G. R., Goldberg, J. M., Armour, J. A., and Randall, W. C.: Cardiac dysrhythmias induced by autonomic nerve stimulation. Am. J. Cardiol. 32:823, 1973. 45. Hagiwara, N., Irisawa, H., Kasanuki, H., and Hosoda, S.: Background current in sino-atrial cells of the rabbit heart. J. Physiol. (Lond.) 448:53, 1992.

46. Han, X., Shimoni, Y., and Giles, W. R.: An obligatory role for nitric oxide in autonomic control of mammalian heart rate. J. Physiol. (Lond.) 476:309, 1994. 47. Han, X., Shimoni, Y., and Giles, W. R.: A cellular mechanism for nitric oxide-mediated cholinergic control of mammalian heart rate. J. Gen. Physiol. 106:45, 1995. 48. Harris, J. M., and Hutter, O. F.: The action of acetylcholine on the movements of potassium in the sinus venosus of the heart [abstract]. J. Physiol. (Lond.) 133:8, 1956. 49. Hartzell, H. C.: Distribution of muscarinic acetylcholine receptors and presynaptic nerve terminals in amphibian heart. J. Cell. Biol. 86:6, 1980. 50. Hartzell, H. C., and Fischmeister, R.: Direct regulation of cardiac Ca2⫹ channels by G proteins: neither proven nor necessary? Trends Pharmacol. Sci. 13:380, 1992. 51. Hartzell, H. C., Mery, P. F., and Fischmeister, R.: Sympathetic regulation of cardiac calcium current is due exclusively to cAMP-dependent phosphorylation. Nature 351:573, 1991. 52. Hayano, J., Sakakibara, Y., Yamada, M., et al.: Decreased magnitude of heart rate spectral components in coronary artery disease: its relation to angiographic severity. Circulation 81:1217, 1990. 53. Hayano, J., Sakakibara, Y., Yamada, A., et al.: Accuracy of assessment of cardiac vagal tone by heart rate variability in normal subjects. Am. J. Cardiol. 67:199, 1991. 54. Herring, N., Golding S., and Paterson, D. J.: Pre-synaptic NO-cGMP pathway modulates vagal control of heart rate in isolated guinea-pig atria. J. Mol. Cell. Cardiol. 32:1795, 2000. 55. Herring, N., and Paterson D. J.: NO-cGMP pathway facilitates acetylcholine release and bradycardia during vagal nerve stimulation in the guinea pig in-vitro. J. Physiol. (Lond.) 535:507, 2001. 56. Hogan, N. A., Casadei, B., and Paterson, D. J.: Nitric oxide donors can activate heart rate independent of autonomic activation. J. Appl. Physiol. 87:97, 1999. 57. Hopkins, D. A.: The dorsal motor nucleus of the vagus nerve and the nucleus ambiguus: structure and connections. In Hainsworth, R., McWilliam, P. N., and Mary, D. A. S. G. (eds.): Cardiogenic Reflexes. New York, Oxford University Press, p.185, 1987. 58. Hull, S. S., Vanoli, E., Adamson, P. B., et al.: Exercise training confers anticipatory protection from sudden death during acute myocardial ischaemia. Circulation 89:548, 1994. 59. Inoue, H., and Zipes, D. P.: Results of sympathetic denervation in the canine heart: supersensitivity that may be arrhythmogenic. Circulation 75:877, 1987. 60. Inoue, K., and Arndt, J. O.: Efferent vagal discharge and heart rate in response to methohexitone, althesin, ketamine and etomidate in cats. Br. J. Anaesth. 54:1105, 1982. 61. Inoue, K., Nashan, B., and Arndt, J. O.: (D-met2,Pro5) enkephalinamide activates cardioinhibitory efferents in anaesthetised dogs. Eur. J. Pharmacol. 110:233,1985. 62. Irisawa, H., Brown, H. F., and Giles, W.: Cardiac pacemaking in the sinoatrial node. Physiol. Rev. 73:197, 1993. 63. Irnaten, M., Aicher, S. A., Wang, J., et al.: Mu-opioid receptors are located postsynaptically and endorphin-1 inhibits voltage-gated calcium currents in premotor cardiac parasympathetic neurones in the rat nucleus ambiguus. Neuroscience 116:573, 2003.

Neural Control of Cardiac Function 64. Irnaten, M., Walwyn, W. M., Wang, J., et al.: Pentobarbital enhances GABAergic neurotransmission to cardiac parasympathetic neurons, which is prevented by expression of GABAA epsilon sub unit. Anesthesiology 97:717, 2002. 65. James, T. N., and Spence, C. A.: Distribution of cholinesterase within the sinus node and AV node of the human heart. Anat. Rec. 155:151, 1966. 66. Jordan, D., and Spyer, K. M.: Central neural mechanisms mediating respiratory-cardiovascular interactions. In Taylor, E. W. (ed.): Neurobiology of the Cardiorespiratory System. Manchester, UK, Manchester University Press, p. 322, 1987. 67. Kalia, M.: Brain stem localization of vagal preganglionic neurons. J. Auton. Nerv. Syst. 3:451, 1981. 68. Katona, P., and Jih, F.: Respiratory sinus arrhythmia: noninvasive measure of parasympathetic cardiac control. J. Appl. Physiol. 39:801, 1975. 69. Katona, P. G., Lipson, D., and Dauchot, P. J.: Opposing central and peripheral effects of atropine on parasympathetic cardiac control. Am. J. Physiol. 232:H146, 1997. 70. Kienzle, M. G., Ferguson, D. W., Birkett, C. L., et al.: Clinical, hemodynamic and sympathetic neural correlates of heart rate variability in congestive heart failure. Am. J. Cardiol. 69:761, 1992. 71. King, T. S., and Coakley, J. B.: The intrinsic nerve cells of the cardiac atria of mammals and man. J. Anat. 92:353, 1958. 72. Kodama, I., Boyett, M. R., Nikmaram, M. R., et al.: Regional differences in effects of E-4031 within the sinoatrial node. Am. J. Physiol. 276:H793, 1999. 73. Kodama, I., Nikmaram, M. R., Boyett, M. R., et al.: Regional differences in the role of the Ca2⫹ and Na⫹ currents in pacemaker activity in the sinoatrial node. Am. J. Physiol. 272: H2793, 1997. 74. Kollai, M., and Koizumi, K.: Cardiovascular reflexes and interrelationships between sympathetic and parasympathetic activity. J. Auton. Nerv. Syst. 4:135, 1981. 75. Kopin, I. J., Hertting, G., and Gordon, E. K.: Fate of norepinephrine-H3 in the isolated perfused rat heart. J. Pharmacol. Exp. Ther. 138:34, 1962. 76. Kunze, D. L.: Reflex discharge patterns of cardiac vagal efferent fibres. J. Physiol. (Lond.) 222:1, 1972. 77. Laubie, M, Schmitt, H., and Vincent, M.: Vagal bradycardia produced by microinjections of morphine-like drugs into the nucleus ambiguus in anaesthetised dogs. Eur. J. Pharmacol. 59:287, 1979. 78. Leon, D. F., Shaver, J. A., and Leonard, J. J.: Reflex heart rate control in man. Am. Heart J. 80:729, 1970. 79. Levy, M. N.: Sympathetic-parasympathetic interactions in the heart. Circ. Res. 29:437, 1971. 80. Levy, M. N.: Cardiac sympathetic-parasympathetic interactions. Fed. Proc. 43:2598, 1984. 81. Levy, M. N., and Blattberg, B.: Effect of vagal stimulation on the overflow of norepinephrine into the coronary sinus during cardiac sympathetic nerve stimulation in the dog. Circ. Res. 38:81, 1976. 82. Levy, M. N., and Martin, P. J.: Neural control of the heart. In Berne, R. M., and Sperelakis, N. (eds.): Handbook of Physiology, Sect. 2. The Cardiovascular System, Vol. 1. The Heart. Bethesda, MD, American Physiological Society, p. 581, 1979.

229

83. Levy, M. N., Ng, M. L., and Zieske, H.: Functional distribution of the peripheral cardiac sympathetic pathways. Circ. Res. 19:650, 1966. 84. Levy, M. N., and Schwartz, P. J. (eds.): Vagal Control of the Heart: Experimental Basis and Clinical Implications. New York, Futura, 2004. 85. Lewis, D. I., and Coote, J. H.: Chemical mediators of spinal inhibition of rat sympathetic neurones on stimulation in the nucleus tractus solitarii. J. Physiol. (Lond.) 486:48, 1995. 86. Lewis, M. E., Al-Khalidi, A. H., Bonser, R. S., et al.: Vagus nerve stimulation decreases left ventricular contractility in vivo in the human and pig heart. J. Physiol. (Lond.) 534:547, 2001. 87. Li, Y. W., and Guyenet, P. G.: Angiotensin II decreases a resting K⫹ conductance in rat bulbospinal neurons of the C1 area. Circ. Res. 78:274, 1996. 88. Lichtenfield, P.: Autonomic dysfunction in the GuillainBarré syndrome. Am. J. Med. 50:772, 1971. 89. Lindmar, R., Löffelholz, K., and Weide, W.: Interstitial washout and hydrolysis of acetylcholine in the perfused heart. Naunyn Schmiedebergs Arch. Pharmacol. 318:295, 1982. 90. Lipsius, S. L., and Vassalle, M.: Dual excitatory channels in the sinus node. J. Mol. Cell. Cardiol. 10:753, 1978. 91. Lloyd-Mostyn, R. H., and Watkins, P. J.: Defective innervation of heart in diabetic autonomic neuropathy. Br. Med. J. 3:15, 1975. 92. Löffelholz, K., and Muscholl, E.: A muscarinic inhibition of the noradrenaline release evoked by postganglionic sympathetic nerve stimulation. Naunyn Schmiedebergs Arch. Pharmacol. 265:1, 1969. 93. Löffelholz, K., and Pappano, A. J.: The parasympathetic neuroeffector junction of the heart. Pharmacol. Rev. 37:1, 1985. 94. Löffelholz, K., and Weide, W.: Aminopyridines and the release of acetylcholine. Trends Pharmacol. Sci. 3:147, 1982. 95. Lopes, O. U., and Palmer, J. H.: Proposed respiratory ‘gating’ mechanism for cardiac slowing. Nature 264:454, 1976. 96. Lovick, T. A.: Differential control of cardiac and vasomotor activity by neurones in nucleus paragigantocellularis lateralis in the cat. J. Physiol. (Lond.) 389:23, 1987. 97. Malliani, A., Schwartz, P. J., and Zanchetti, A.: Neural mechanisms in life-threatening arrhythmias. Am. Heart J. 100:705, 1980. 98. Mancia, G., Ferrari, A., Gregorini, L., et al.: Circulatory reflexes from carotid and extracarotid baroreceptor areas in man. Circ. Res. 41:309, 1977. 99. Maqbool, A., Batten, T. F. C., and McWilliam, P. N.: Ultrastructural relationship of GABAergic terminals and cardiac vagal preganglionic motoneurones and vagal afferents in cat: combined HRP tracing-immunogold labelling study. Eur. J. Neurosci. 3:501, 1991. 100. Mary, D. A. S. G., and Hainsworth, R.: Methods for the study of cardiovascular reflexes. In Hainsworth, R., and Mark, A. L. (eds.): Cardiovascular Reflex Control in Health and Disease. London, W. B. Saunders, p. 1, 1993. 101. Mathias, C. J.: Disturbances of cardiovascular control in autonomic disorders. In Hainsworth, R., and Mark, A. L. (eds.): Cardiovascular Reflex Control in Health and Disease. London, W. B. Saunders, p. 425, 1993.

230

Function of the Peripheral Nervous System

102. McAllen, R. M., and Dampney, R. A. L.: Vasomotor neurones in the rostral ventrolateral medulla are organised topographically with respect to type of vascular bed but not body region. Neurosci. Lett. 110:91, 1990. 103. McAreavey, D., Neilson, J. M. M., Ewing, D. J., and Russell, D. C.: Cardiac parasympathetic activity during the early hours of acute myocardial infarction. Br. Heart J. 62:165, 1989. 104. Mendelowitz, D.: Firing properties of identified parasympathetic cardiac neurons in nucleus ambiguus. Am J Physiol 271:H2609–H2614, 1996. 105. Mitsuiye, T., Shinagawa, Y., and Noma, A.: Sustained inward current during pacemaker depolarization in mammalian sinoatrial node cells. Circ. Res. 87: 88, 2000. 106. Mizeres, N. J.: The cardiac plexus in man. Am. J. Anat. 112:141, 1963. 107. Murray, A., Ewing, D. J., Campbell, I. W., et al.: RR interval variations in young male diabetics. Br. Heart J. 37:882, 1975. 108. Musialek, P., Lei, M., Brown, H. F., et al.: Nitric oxide can increase heart rate by stimulating the hyperpolarizationactivated inward current If. Circ. Res. 81:60, 1997. 109. Neff, R. A., Humphrey, J., Mihalevich, M., and Mendelowitz, D.: Nicotine enhances presynaptic and postsynaptic glutaminergic neurotransmission to activate cardiac parasympathetic neurons. Circ. Res. 83:1241, 1998. 110. Neff, R. A., Mihalevich, M., and Mendelowitz, D.: Stimulation of NTS activates NMDA and non-NMDA receptors in rat cardiac vagal neurones in the nucleus ambiguus. Brain Res. 792:277, 1998. 111. Oppenheim, A. V., and Schafer, R. W.: Digital Signal Processing. Englewood Cliffs, NJ, Prentice Hall, 1975. 112. Oppenheimer, S. M., Cechetto, D. F., and Hachinski, V. C.: Cerebrogenic cardiac arrhythmias: cerebral electrocardiographic influences and their role in sudden death. Arch. Neurol. 47:513, 1990. 113. Pagani, M., Lombardi, F., Guzzetti, S., et al.: Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ. Res. 59:178, 1986. 114. Parizel, G.: On the mechanism of sudden death with subarachnoid hemorrhage. J. Neurol. 220:71, 1979. 115. Paton, J. F. R., Kasparov, S., and Paterson, D. J.: Site-specific differential modulation of cardiac autonomic control by nitric oxide. Trends Neurosci. 25:626, 2002. 116. Pomeranz, B., Macaulay, R. J. B., Caudill, M. A., et al.: Assessment of autonomic function in humans by heart rate spectral analysis. Am. J. Physiol. 248:H151, 1985. 117. Potter, E. K.: Inspiratory inhibition of vagal responses to baroreceptor and chemoreceptor stimuli in the dog. J. Physiol. (Lond.) 316:177, 1981. 118. Prystowsky, E. N., Jackman, W. M., Rinkenberger, R. L., et al.: Effect of autonomic blockade on ventricular refractoriness and atrioventricular nodal conduction in humans: evidence supporting a direct cholinergic action on ventricular muscle refractoriness. Circ. Res. 49:511, 1981. 119. Pyner, S., and Coote, J. H.: Rostroventrolateral medulla neurons preferentially project to target specified sympathetic preganglionic neurones. Neuroscience 83:617, 1998.

120. Raczkowska, M., Eckberg, D. L., and Ebert, T. J.: Muscarinic cholinergic receptors modulate vagal cardiac responses in man. J. Auton. Nerv. Syst. 7:271, 1983. 121. Rigg, L., Heath, B. M., Cui, Y., and Terrar, D. A.: Localisation and functional significance of ryanodine receptors during beta-adrenoreceptor stimulation in the guineapig sino-atrial node. Cardiovasc. Res. 48: 254, 2000. 122. Rosenblueth, A., and Simeone, F. A.: The interregulations of vagal and accelerator effects on the cardiac rate. Am. J. Physiol. 110:42, 1934. 123. Rudolph, U., Crestani, F., Benke, D., et al.: Benzodiazepine actions mediated by specific ␥-aminobutyric acid A receptor subtypes. Nature 401:796, 1999. 124. Ruggeri, P., Battaglia, A., Ermirio, R., et al.: Role of nitric oxide in the control of the heart rate within the nucleus ambiguus of rats. Neuroreport 11:481, 2000. 125. Sakmann, B., Noma, A., and Trautwein, W.: Acetylcholine activation of single muscarinic K⫹ channels in isolated pacemaker cells of the mammalian heart. Nature 303:250, 1983. 126. Sanguinetti, M. C., Jiang, C., Curran, M. E., and Keating, M. T.: A mechanistic link between an inherited and an acquired cardiac arrhythmia: HERG encodes the Ikr potassium channel. Cell 81:299, 1995. 127. Sanguinetti, M. C., and Keating, M. T.: Role of delayed rectifier potassium channels in cardiac repolarisation and arrhythmias. News Physiol. Sci. 12:152, 1997. 128. Saul, J. P., Arai, Y. H., Berger, R. D., et al.: Assessment of autonomic regulation in chronic congestive heart failure by heart rate spectral analysis. Am. J. Cardiol. 61:1292, 1988. 129. Schwartz, P. J., Periti, M., and Malliani, A.: The long Q-T syndrome. Am. Heart J. 89:378, 1975. 130. Schwarz, P., Diem, R., Dun, N. J., and Forstermann, U.: Endogenous and exogenous nitric oxide inhibits norepinephrine release from rat heart sympathetic nerves. Circ. Res. 77:841, 1995. 131. Shepherd, J. T.: Cardiac mechanoreceptors. In Fozzard, H. A., Haber, E., Jennings, R. B., and Katz, A. M. (eds.): The Heart and Cardiovascular System. New York, Raven Press, p. 1535, 1986. 132. Skinner, J. E.: Neurocardiology shows that the central, not the peripheral, action of propranolol reduces mortality following coronary artery occlusion in the conscious pig. Integr. Physiol. Behav. Sci. 26:85, 1991. 133. Skinner, J. E., and Reed, J. C.: Blockade of frontocorticalbrain stem pathway prevents ventricular fibrillation of ischaemic heart. Am. J. Physiol. 240:H156, 1981. 134. Smith, S. A., and Smith, S. E.: Heart rate variations in the Guillain-Barré syndrome. Br. Med. J. 281:1009, 1980. 135. Smyth, H. S., Sleight, P., and Pickering, G. W.: Reflex regulation of arterial pressure during sleep in man: a quantitative method of assessing baroreflex sensitivity. Circ. Res. 24:109, 1969. 136. Spyer, K. M.: Neural organisation and control of the baroreceptor reflex. Rev. Physiol. Biochem. Pharmacol. 88:23, 1981. 137. Stotler, W. A., and McMahon, R. A.: The innervation and structure of the conductive system of the human heart. J. Comp. Neurol. 87:57, 1947. 138. Stull, J. T., and Mayer, S. E.: Biochemical mechanisms of adrenergic and cholinergic regulation of myocardiac

Neural Control of Cardiac Function

139.

140. 141. 142.

143.

144.

145.

146.

147. 148.

149.

contractility. In Berne, R. M., and Sperelakis, N. (eds.): Handbook of Physiology, Sect. 2. The Cardiovascular System, Vol. 1. The Heart. Bethesda, MD, American Physiological Society, p. 741, 1979. Sun, M. K.: Central neural organisation and control of sympathetic nervous system in mammals. Prog. Neurobiol. 47:465, 1995. Talman, W. T.: Cardiovascular regulation and lesions of the central nervous system. Ann. Neurol. 18:1, 1985. Taylor, E. W.: The evolution of efferent vagal control of the heart in vertebrates. Cardioscience 5:173, 1994. Taylor, E. W., Jordan, D., and Coote, J. H.: Central control of the cardiovascular and respiratory systems and their interactions in vertebrates. Physiol. Rev. 79:855, 1999. Thornton, J. M., Aziz, T., Schlugman, D., and Paterson, D. J.: Electrical stimulation of the midbrain increases heart rate and arterial blood pressure in awake humans. J. Physiol. (Lond.) 593:615, 2002. Vaile, J. C., Chowdhary, S., Osman, F., et al.: Effects of angiotensin II (AT1) receptor blockade on cardiac vagal control in heart failure. Clin. Sci. 101:559, 2001. Van Buren, J. M., and Ajmone-Marsan, C.: A correlation of autonomic and EEG components in temporal lobe epilepsy. Arch. Neurol. 3:683, 1960. Vandecasteele, G., Eschenhagen, T., Scholz, H., et al.: Muscarinic and beta-adrenergic regulation of heart rate, force of contraction and calcium current is preserved in mice lacking endothelial nitric oxide synthase. Nat. Med. 5:331, 1999. Verrier, R. L., and Lown, B.: Behavioural stress and cardiac arrhythmias. Annu. Rev. Physiol. 46:155, 1984. Verrier, R. L., Muller, J. A., and Hobson, J. A.: Sleep, dreams, and sudden cardiac death: the case for sleep as an autonomic stress test for the heart. Cardiovasc. Res. 31:181, 1996 Volders, P. G. A., Vos, M. A., Szabo, B., et al.: Progress in the understanding of cardiac early after depolarisations and torsades de pointes: time to revise current concepts. Cardiovasc. Res. 46:376, 2000.

231

150. Vybiral, T., Bryg, R. J., Maddens, M. E., and Boden, W. E.: Effect of passive tilt on sympathetic and parasympathetic components of heart rate variability in normal subjects. Am. J. Cardiol. 63:1117, 1989. 151. Wang, J., Irnaten, M., and Mendelowitz, D.: Characterisation of spontaneous and evoked GABA synaptic currents in cardiac vagal neurons in rats. Brain Res. 889:78, 2001. 152. Wang, Y., and Ramage, A. G.: The role of central 5-HT1A receptors in the control of B-fibre cardiac and bronchoconstrictor vagal preganglionic neurones in anaesthetised cats. J. Physiol. (Lond.) 536:753, 2001. 153. Wharton, J., Gulbenkian, S., Merighi, A., et al.: Immunohistochemical and ultrastructural localisation of peptide-containing nerves and myocardial cells in the human atrial appendage. Cell Tissue Res. 254:155, 1988. 154. Wikberg, J. E. S., and Lefkowitz, R. J.: Adrenergic receptors in the heart: pre- and postsynaptic mechanisms. In Randall, W. C. (ed.): Nervous Control of Cardiovascular Function. New York, Oxford University Press, p. 95, 1984. 155. Wu, S. Y., and Dun, N. J.: Potentiation of IPSCs by nitric oxide in immature rat sympathetic preganglionic neurones in vitro. J. Physiol. (Lond.) 495:479, 1996. 156. Yanagihara, K., and Irisawa, H.: Inward current activated during hyperpolarization in the rabbit sinoatrial node cell. Pflugers Arch. 385:11, 1980. 157. Yang, Z., Smith, L., and Coote, J. H.: Paraventricular nucleus activation of renal sympathetic neurones is synaptically depressed by nitric oxide and glycine acting at a spinal level. Neuroscience 124:421, 2004. 158. Yashpal, K., Gauthier, S., and Henry, J. L.: Oxytocin administered intrathecally preferentially increases heart rate rather than arterial pressure in the rat. J. Auton. Nerv. Syst. 20:167, 1987. 159. Yatani, A., Hamm, H., Codina, J., et al.: A monoclonal antibody to the subunit of Gk blocks muscarinic activation of atrial K⫹ channels. Science 241:828, 1988. 160. Zipes, D. P., Levy, M. N., Cobb, L. A., et al.: Sudden cardiac death: neural-cardiac interactions. Circulation 76(Suppl. I): I-202, 1987.

12 Neurobiology of the Enteric Nervous System JACKIE D. WOOD

Overview Historical Perspective ENS Conceptual Model Histoanatomy Enteric Neuronal Morphology Enteric Sensory Neurons Visceral Hypersensitivity in Irritable Bowel Syndrome Chemoreception Mechanoreception Interneurons

Enteric Motor Neurons Secretomotor Neurons Excitatory Musculomotor Neurons Inhibitory Musculomotor Neurons Disinhibitory Motor Disorders Chronic Intestinal Pseudo-obstruction Sphincteric Achalasia Neuronal Electrical Behavior Resting Membrane Potentials Action Potentials

OVERVIEW The enteric nervous system (ENS) contains as many neurons as the spinal cord. An estimated 100 million neurons are required for programmatic control of digestive functions (e.g., specialized motility patterns and glandular secretion). The required number of neurons, together with accompanying glia, would greatly expand the volume of the central nervous system (CNS) if placed there. Rather than packing the neural control circuits exclusively within the skull or vertebral column and transmitting control signals over long, unreliable transmission lines to the gut, vertebrate animals have evolved with the neural control networks distributed along the length of the digestive tract in close apposition to the muscles, glands, and blood vessels that must be controlled and whose activity must be integrated for effective whole-organ function. The ENS integrates contraction of the musculature, glandular secretion, and intramural blood flow into organized patterns of digestive behavior. Activity in the ENS may either evoke contractions in the gut musculature or inhibit contractile activity as required for the organization of specific patterns of organ motility. Likewise, the ENS organizes absorption from and secretion into the intestinal lumen and integrates these functions with motility. Integrative microcircuitry in the ENS also controls blood

Synaptic Transmission Fast Excitatory Postsynaptic Potentials Slow Synaptic Excitation Inhibitory Postsynaptic Potentials Presynaptic Facilitation Presynaptic Inhibition Conclusions

flow within the gut’s walls and the distribution of flow to the musculature, mucosa, and lamina propria. Malfunctions of integrative ENS control of the gut’s effector systems are becoming increasingly recognized as underlying factors in gastrointestinal disorders, especially in the functional gastrointestinal disorders.173–175 The musculature, mucosal secretory epithelium, and blood and lymphatic vasculature are the gut’s effector systems. Moment-to-moment behavior of any of the specialized compartments of the gut (e.g., stomach, small and large intestine) reflects neurally integrated activity of the effector systems. The ENS not only controls the activity of each effector, but also coordinates the activity of each to achieve organized behavior of the whole organ. Coordinated activity is achieved through timing of excitatory and/or inhibitory neural inputs to each effector system individually. Coordination of mucosal secretion and muscle contractile activity in a timed sequence during intestinal propulsion of luminal contents is one example of ENS integrative function. Secretion of H2O, electrolytes, and mucus occurs first, followed by propulsive contractions in the musculature. A simultaneous increase in neurally controlled blood flow to the secretory glands supports the enhanced secretory response. The ENS is viewed as a “minibrain” with a library of programs for multiple patterns of small or large intestinal 249

250

Function of the Peripheral Nervous System

behavior. A specific program determines motor behavior in the postprandial state, and another establishes the pattern of intestinal motility that characterizes the fasting state. The specialized motility pattern seen in the upper third of the small intestine during emesis reflects output of another of the programs in the library. During emesis, peristaltic propulsion in the upper third of the small intestine is reversed for rapid movement of the luminal contents toward the stomach. The emetic program can be called up from the library either by commands from the brain or by local sensory detection of threatening substances in the lumen.74

HISTORICAL PERSPECTIVE Modern progress in understanding the neurobiology of the ENS started in the mid-20th century with a transition in conceptual thinking from the traditional view that all control was central, via sympathetic and parasympathetic neural pathways, to a concept that emphasized a local minibrain as the chief determinant of intestinal behavior. The prevailing concept in the first half of the 20th century followed the classic concepts of neural control of other autonomic systems (e.g., heart, blood vasculature, airways, and urinary bladder). Ganglia in the gut wall were presumed to be typical parasympathetic terminal ganglia. Postganglionic neurons with their cell bodies in the terminal ganglia were expected to initiate responses of the musculature or secretory glands through the direct release of acetylcholine at neuromuscular and neuroglandular junctions. All of the excitatory impulses for evoking

Outmoded concept Brain stem Spinal cord

Spinal cord

Preganglionic parasympathetic nerve

Preganglionic sympathetic nerve

muscle contraction and glandular secretion were postulated to be transmitted to the gut along parasympathetic autonomic projections starting in either the brainstem or the sacral region of the spinal cord. The sympathetic division of the autonomic nervous system was the accepted inhibitory component of gut innervation in the early concept. Postganglionic neurons, with their cell bodies in prevertebral sympathetic ganglia, were thought to release norepinephrine to suppress secretion and contraction of the musculature. A brake-accelerator analogy of opposing inhibitory and excitatory input, like that in the heart, was invoked to explain neural control of gut motility. A balance favoring parasympathetic excitation was expected to result in increased activity; balance favoring sympathetic inhibition accounted for suppression of activity. The ganglia in the gut were interpreted as relaydistribution centers for signals transmitted from the CNS (Fig. 12–1). In this model, all of the decision making and automated feedback control functions resided in neural networks in the CNS. No intelligence was attributed to the nervous system in the gut. Sir Johannes Newport Langley, the British autonomic physiologist of Cambridge University’s Trinity College, recognized during the first quarter of the 20th century that the ganglia of the gut did more than robotically relay and distribute information from the CNS. Langley was conceptually unable to reconcile the large disparity between the 2 ⫻ 108 neurons in the gut and the few hundred efferent fibers in the vagus nerves other than to conclude that the nervous system of the gut possessed

Current concept Brain stem Spinal cord

Spinal cord

Parasympathetic nervous system

Sympathetic nervous system

Enteric nervous system (brain-in-the-gut) Interneurons Postganglionic parasympathetic neuron

Postganglionic sympathetic neuron Motor neurons Musculature Glands Blood vessels

Musculature Glands Blood vessels

FIGURE 12–1 Outmoded and current concepts of the functional organization of the autonomic control of the digestive tract.

251

Neurobiology of the Enteric Nervous System

integrative functions independent of the CNS.75 This conclusion was reinforced by his demonstration of organized propulsion of intraluminal boluses in intestinal segments detached from the body in vitro.76 Based on these observations, he proposed that the nervous system in the gut is a distinct third division of the autonomic nervous system and named it the enteric nervous system. Langley’s insightful concept languished for nearly five decades. Now, in the new millennium, three divisions— parasympathetic, sympathetic, and enteric—are accepted by scientists the world over as the construct for the autonomic nervous system.40,41 After Langley, the utmost advance in basic understanding of neuropathy in the digestive tract has been the enlightened realization that the nervous system of the gut is a minibrain with intelligent neural networks. Emergence of the heuristic model of the ENS as a “little brain” has opened the way for basic research that follows many of the approaches used to gain understanding of the neurobiology of the brain and spinal cord. The outcome has been improved insight into the disordered physiology that is manifest in a variety of symptoms, including those of a functional nature. Issues related to functional gastrointestinal disorders today are reminiscent of the historical progression of insight into functional disorders of the nervous system. Functional disorders of the CNS, like those of the digestive tract, were designated as such out of ignorance of the pathophysiologic basis of the disorder. Parkinson’s disease was regarded as an idiopathic disorder in the years before the synaptic microcircuits of the basal ganglia and the importance of the neurotransmitter dopamine were understood for the brain. The history of progress in basic brain research underlies a conviction that expanded understanding of the neural networks in the ENS has potential for uncovering unrealized explanations for gastrointestinal disorders currently described as idiopathic in nature, as well as rational approaches to therapy.

ENS CONCEPTUAL MODEL The heuristic model for the ENS (Fig. 12–2) is the same as for all integrative nervous systems, whether the “simple” nervous systems of invertebrate animals or the vertebrate CNS. Like other integrative nervous systems, the ENS functions with sensory neurons, interneurons, and motor neurons. Sensory neurons have receptor regions specialized for detecting changes in stimulus energy. The gut’s sensory receptors are classified according to the kinds of energy they detect. These include thermo-, chemo-, and mechanoreceptors. The receptor regions transform changes in stimulus energy into signals coded by action

Brain

Effector systems Muscle Sensory neurons

Interneurons

Motor neurons

Enteric nervous system

Secretory Glands Blood vessels

Organ system behavior

FIGURE 12–2 Sensory neurons, interneurons, and motor neurons are synaptically interconnected to form the integrated microcircuits of the enteric nervous system. Like the central nervous system, sensory neurons, interneurons, and motor neurons are connected synaptically for flow of information from sensory neurons to interneuronal integrative networks to motor neurons to effector systems. The enteric nervous system organizes and coordinates the activity of each effector system into meaningful behavior of the whole organ. Bidirectional communication occurs between the central and the enteric nervous system.

potentials that subsequently are transmitted along sensory nerve fibers to other points in the nervous system for processing. Interneurons are interconnected to form networks that process sensory information and control the behavior of motor neurons. Multiple synaptic connections among interneurons organize the neurons into “logic” circuits that decipher action potential codes from sensory neurons and neural signals from elsewhere in the nervous system. Some of the interneuronal circuits are integrative; others are reflex circuits. Integrative circuits are circuits that process input information and reconfigure it into functional output signals that are not direct transforms of the input signal. Plasticity resides in integrative circuits. Reflex circuits, in comparison, are “hardwired,” much like spinal motor reflexes, to generate stereotyped outputs in response to input from specific kinds of sensory receptors. Motor neurons are the final common pathways for transmission of control signals from integrative and reflex circuits to the effector systems. Motor neurons may be stimulated to fire by excitatory synaptic input derived from interneuronal circuitry, or their activity may be suppressed by inhibitory synaptic input from the interneuronal circuitry. In the digestive tract, motor neurons may release excitatory or inhibitory neurotransmitters at their junctions with the muscles and thereby actively excite muscular contraction or actively inhibit contraction.

252

Function of the Peripheral Nervous System

HISTOANATOMY

tation for sustaining the neural networks in a functional state as the ENS is subjected to the mechanical forces and deformations that occur during contraction of the muscle or expansion and stretching of the wall as the lumen fills. Most of the motor neurons that innervate the circular and longitudinal muscle coats reside in the myenteric plexus.12,13 The submucosal plexus is a ganglionated plexus situated throughout the submucosal space between the mucosa and circular muscle coat (Fig. 12–3). It is most prominent as a ganglionated network in the small and large intestine. No ganglionated submucosal plexus exists in the esophagus, and ganglia are sparse in the submucosal space of the stomach. In larger mammals (e.g., pig, human), the submucosal plexus consists of an inner submucosal network (Meissner’s plexus) located at the serosal side of the muscularis mucosae and an outer plexus (Schabadasch’s plexus) adjacent to the luminal side of the circular muscle coat. In human small and large bowel, a third intermediate plexus lies between Meissner’s and Schabadasch’s plexuses.157 Motor innervation of the intestinal crypts and villi is derived from ganglion cells in the submucosal plexus. Some of the neurons in submucosal ganglia send fibers to

Cell bodies of ENS neurons are in ganglia buried inside the walls of the digestive tract. The ganglia are interconnected by interganglionic fiber tracts to form ganglionated plexuses. Interganglionic fiber tracts consist mainly of projections from ganglion cell bodies in one ganglion to connect synaptically with neurons in neighboring ganglia. The ganglionated plexuses form a distributed continuum around the circumference and along the length of each of the specialized organs of the gut. Myenteric and submucosal plexuses are the prominent ganglionated plexuses of the ENS (Fig. 12–3). The myenteric plexus, known also as Auerbach’s plexus, is the major plexus of the ENS. It is located between the longitudinal and circular muscle coats of most all regions of the gastrointestinal tract. It is a two-dimensional array of flat, disclike neurons, ganglia, and interganglionic fiber tracts sandwiched between the longitudinal and circular muscle coats (Fig. 12–3). Unlike other autonomic ganglia, the cell bodies of the ganglion cells are not in grapelike clusters. They lie edge to edge like a single layer of coins placed in a two-dimensional plane. The single-layered, two-dimensional organization is postulated to be an adap-

A

Enteric nervous system Myenteric plexus Circular muscle Submucosal plexus

Longitudinal muscle

Mucosa

80 μm

B

80 μm

Myenteric plexus

C

Submucosal plexus

FIGURE 12–3 Histoanatomic organization of the enteric nervous system. A, Ganglia and interganglionic fiber tracts of the myenteric and submucosal plexuses are notable structures of the enteric nervous system. The myenteric plexus is distributed between the circular and longitudinal muscle coats; the submucous plexus is between the mucosa and circular muscle coat. B, Tyrosine hydroxylase stain of the myenteric plexus as it appears on the underside of the longitudinal muscle coat after microdissection from guinea pig small intestine. C, Tyrosine hydroxylase stain of the submucosal plexus as it appears in a preparation obtained from guinea pig small intestine.

Neurobiology of the Enteric Nervous System

the myenteric plexus and also receive synaptic input from axons projecting from the myenteric plexus.138,139 The interplexus connections link the two networks into a functionally integrated nervous system. Structure, function, and neurochemistry of enteric ganglia differ significantly from other autonomic ganglia. Unlike other autonomic ganglia in which function is mainly as relay-distribution centers for signals transmitted from the CNS, enteric ganglia are interconnected to form a nervous system with mechanisms for integration and processing of information like those found in the brain and spinal cord. Most properties of the ENS mimic those of the CNS. Some examples are 1. Complex integrative functions: Like the brain, the ENS executes complex integrative functions that control single effector systems and coordinates the activity of multiple systems into patterns of overall behavior that achieve homeostasis. Program libraries are present in both the ENS and CNS, but not in other autonomic ganglia. 2. Sensory neurons: Sensory neurons transmit information to the neural circuits of the ENS for processing in the same manner that the CNS receives and processes sensory signals. 3. Interneurons: Interneurons are synaptically interconnected into information processing circuits in both the ENS and CNS, but generally not in other autonomic ganglia. 4. Motor neurons: Enteric motor neurons are the final common pathways for flow of information from interneuronal processing circuits to the effector systems. Their function is analogous to that of alpha and gamma motor neurons from the spinal cord to skeletal muscles. 5. Glial elements: The supporting cells of the ENS are like glial cells in the brain but, unlike the Schwann cells found associated with other autonomic ganglia. Enteric glial elements structurally and chemically resemble astroglia of the brain. 6. Synaptic neuropil: Synaptic neuropils are present in both enteric ganglia and the CNS. Neuropils are tangled meshworks of fine nonmyelinated nerve fibers that are segregated in regions of the ganglia away from the cell bodies that give rise to the fibers. The presence of a neuropil is significant because, in all integrative nervous systems, most of the information processing occurs in microcircuits within a synaptic neuropil. 7. Multiple synaptic mechanisms: As in the brain and spinal cord, the neural circuits of the ENS express multiple forms of neurotransmission, which is involved in the handling of information. 8. Multiple neurotransmitters: Most of the neurotransmitters found in the brain and spinal cord are also expressed in the ENS.

253

9. Absence of connective tissue: Collagen, which forms part of the connective tissue for the structural support of autonomic ganglia, is absent from the interior of enteric ganglia, as it is in the CNS. 10. Reduced extracellular space: Unlike other autonomic ganglia, extracellular space is reduced by close packing of glial support elements. 11. Isolation from blood vessels: No blood vessels enter enteric ganglia of the intestine, and a blood-ganglion barrier analogous to the blood-brain barrier is inserted between the vasculature and the synaptic circuits of the ganglia.

ENTERIC NEURONAL MORPHOLOGY Dogiel types I and II are the principal morphologic classes of neurons in the ENS. The German neuroanatomist A. S. Dogiel described two morphologic types of enteric ganglion cells that are named for him. These are Dogiel types I and II, both of which are found in myenteric and submucosal plexuses. Both types of neurons are distributed in a two-dimensional plane (i.e., the cell bodies are not stacked one on the other). Dogiel type I neurons have cell bodies with multiple short processes and a single long process (Fig. 12–4). These are flat neurons with the processes extending for short distances from the cell body in the circumferential and longitudinal planes of the wall. The short processes are dendrites, which receive synaptic input; the long process is an axon. Axons of Dogiel type I neurons project for relatively long distances through interganglionic fiber tracts and many rows of ganglia. Still, the projections are not long in absolute length; the longest known projections of any enteric ganglion cell are only approximately 2 to 3 cm. Dogiel I neurons projecting in a specific direction are identified by expression of specific neurotransmitters. Aborally projecting Dogiel I axons release nitric oxide and vasoactive intestinal polypeptide (VIP) as neurotransmitters; those projecting in the oral direction release substance P and acetylcholine. Some of the Dogiel type I neurons are motor neurons to the musculature and secretory epithelium; others are interneurons. The cell bodies of ENS neurons with Dogiel type II morphology have smooth surfaces with long and short processes arising in a variety of configurations (Fig. 12–4). The long processes may extend through interganglionic fiber tracts across several rows of ganglia in the circumferential, oral, or aboral direction. Shorter processes may only project within the home ganglion. Almost all Dogiel II neurons in the myenteric plexus project a process to the submucosa/mucosa.36 Terminals of Dogiel II projections into the mucosal regions are fired by mechanical and chemical stimulation of the mucosa (Fig. 12–5).8 Dogiel II

254

Function of the Peripheral Nervous System

FIGURE 12–4 Dogiel morphologic types I and II neurons. A, Dogiel morphologic type I neuron in the submucosal plexus of guinea pig small intestine. Dogiel I neurons have multiple short dendrites at the edges of the cell body and a single axon. B, Dogiel morphologic type II neuron in the myenteric plexus of the guinea pig small intestine. Dogiel II neurons have multiple long processes of variable length that ramify throughout the ganglion and into interganglionic fiber tracts. The neurons in A and B were marked by the injection of biocytin from micropipettes that also served as recording microelectrodes.

neurons are involved at an early stage in the handling of sensory information, as well as interneuronal functions in the microcircuits of the ENS. They are identified by expression of immunoreactivity for the neurotransmitters substance P and acetylcholine and the calcium-binding protein calbindin.

ENTERIC SENSORY NEURONS Sensory neurons belong to one of the three functional categories of enteric neurons illustrated in Figure 12–2. Sensory afferent fibers with neuronal cell bodies in nodose and dorsal root ganglia transmit information from the gastrointestinal tract and gallbladder to the CNS for processing. These afferents also transmit a steady stream of information to the local processing circuits in the ENS and to prevertebral sympathetic ganglia. Mechano-, chemo-, and thermoreceptors are present in the ENS. Mechanoreceptors sense changes in mechanical forces in the mucosa, musculature, serosal surface, and mesentery. They supply both the ENS and the CNS with information on stretch-related tension and muscle length in the wall and on the movement of luminal contents as they brush past the mucosal surface. Mesenteric receptors generate coded information on gross movements of the organ. Chemoreceptors generate information on the concentration of nutrients, osmolarity, and pH in the luminal contents. Recording of nerve impulses in afferent fibers as they exit the gut shows most sensory receptors to be multimodal in that they respond to both mechanical and chemical stimuli.

Thermoreceptors supply the brain with deep-body temperature data used in body temperature regulation and probably account also for the sensations of temperature change in the lumen. Presence in the gastrointestinal tract of pain receptors (nociceptors) equivalent to those connected with C fibers and A␦ fibers elsewhere in the body is likely, but not unequivocally confirmed except for the gallbladder.18 The biophysical principles of sensory function in the gastrointestinal tract are like those found elsewhere in the body. Each sensory neuron has at least one neurite with a generator region that converts a change in stimulus energy into an action potential code. The frequency of action potential discharge is a code for transmission of information on intensity of the stimulus to processing circuits in the ENS and CNS. Information on contractile state of the musculature and distention of the gut wall is coded by mechanoreceptors. Whether the cell bodies of intramuscular and mucosal mechanoreceptors reside in dorsal root ganglia, in enteric ganglia, or in both places is an unresolved question.36,170 Stretch-sensitive mechanoreceptors have pathophysiologic importance because a significant proportion of patients diagnosed with the irritable bowel syndrome (IBS) experience abnormal sensitivity to stretch that is perceived as abdominal pain.77,162

Visceral Hypersensitivity in Irritable Bowel Syndrome The primary causes of abdominal pain of digestive tract origin are distention and exaggerated muscular contraction. Stimuli such as pinching and burning of the mucosa

255

Neurobiology of the Enteric Nervous System

Slow excitatory inputs Neural

Paracrine AH Dogiel II neuron

Myenteric plexus

Myenteric plexus Circular muscle coat Submucosa 5-HT3 Serotonergic receptors

5-HT Mucosa

Mechanosensitive enterochromaffin cell

FIGURE 12–5 Functional neuroanatomy of AH/Dogiel morphologic type II neurons in the neural networks of the ENS. The projectional geometry of the neurites underscores the significance of the cell body as a gating determinant of neurite-to-neurite communication across the cell soma (see Figs. 12–4B and 12–9). Most AH neurons project one or more neurites to the mucosa. The terminals in the mucosa fire in response to acidic pH and in response to serotonin released by mechanical stimulation of mucosal enterochromaffin cells. Like elements in a neural network, AH neurons are stimulated to fire repetitively by slow excitatory inputs. Firing of the neuronal cell body is transformed into slow synaptic output at synapses with neighboring AH neurons in the network (see Fig. 12–4B). The somal gate is closed in the absence of slow synaptic input. In this state, inbound information from the mucosa is not gated to neurites across the cell body. In circumstances in which the somal gate is opened by either slow synaptic or paracrine excitatory input, information from the mucosa is gated in the direction of slow synaptic outputs to AH neurons in the circuit (see Fig. 12–10).

do not evoke pain. Hypersensitivity of the mechanosensors for stretch (distention) and contractile tension are implicated as pain factors in patients diagnosed with IBS. Heightened sensitivity to distention and conscious awareness of gastrointestinal sensations experienced by patients with IBS are generalized phenomena that extend throughout their digestive tract, including the esophagus.120 The locus of neurologic disturbance responsible for the hypersensitivity in patients with IBS is not fully understood. The available evidence suggests disordered neurophysiology in either the peripheral nervous system, the CNS, or both. A peripheral neural hypothesis states that mechanosensitive afferents in the gut wall become hypersensitive to mechanical stimuli and, as a result, transmit erroneous information to the CNS. The alternative hypothesis is for the mechanosensors to be transmitting accurate information, with misinterpretation in the central processing circuits being responsible for the perception of

abnormal sensations from the gut. Results obtained with modern methods of brain imaging suggest abnormal processing of sensory information in the amygdala, anterior cingulate cortex, and prefrontal cortex in patients with IBS.89,118,135 The logical conclusion is that sensory physiology may be disordered at peripheral and/or central levels of neural organization in patients with IBS who experience gastrointestinal hypersensitivity to mechanical stimuli. Hyposensory perception, particularly in the rectosigmoid region, is at the opposite extreme of gastrointestinal sensory abnormality. Sensory suppression, either in the intramural connections for rectoanal stretch reflexes or in the transmission pathway from the rectosigmoid to conscious perception of distention in the CNS, is implicated as an underlying factor in the pathogenesis of chronic constipation and associated symptoms.81 Hyperglycemia represents another of the pathologic disturbances that can lead to sensory neuropathy and suppressed sensations of rectosigmoid fullness that may lead to constipation.19

Chemoreception Sensory transduction of chemical information related to luminal nutrient concentrations in the small bowel involves paracrine signaling from enteroendocrine cells to afferent terminals in the intestinal wall. Mucosal enteroendocrine cells “taste” the luminal contents and release cholecystokinin (CCK) in direct proportion to the concentration of products of protein digestion and lipids. Amounts of CCK released are a transform of the nutrient concentration, which in turn acts to evoke concentrationdependent firing in vagal afferent neurons. Vagal afferents in the stomach and upper small bowel bifurcate for simultaneous transmission of the same information to the CNS and ENS.7 The firing frequency transmitted by the afferents becomes a transform of the stimulus strength for simultaneous decoding in the brain and enteric processing circuitry. The neuronal receptors for CCK belong to the CCK-A subtype.29,136 Vagal sensory mechanisms involving CCK are implicated in the functions involved in the sensations of fullness, satiety, and control of food intake.136 Disordered vagal sensory function is suspect in the pathophysiology of functional dyspepsia.146,151 Firing of impulses in vagal afferent terminals, spinal afferent terminals, and the terminals of the projections of myenteric Dogiel type II neurons in the intestinal mucosa can be initiated by the action of 5-hydroxytryptamine (5-HT) released from mucosal enterochromaffin cells.8,71,151 Distortion of the mucosal surface by mechanical shearing forces or the presence of noxious chemical agents stimulates the enterochromaffin cells to release 5-HT, which then becomes a paracrine signal to serotonergic receptors on spinal afferent and Dogiel type II terminals in the mucosa and elsewhere in the gut wall (see Fig. 12–5). The serotonergic receptors on vagal afferent

256

Function of the Peripheral Nervous System

terminals, on spinal afferent terminals, and on the mucosal terminals of myenteric Dogiel type II neurons belong to the 5-HT3 receptor subtype. Stimulation of the 5-HT3 receptors on vagal afferent terminals is a trigger for nausea and emesis and accounts for the efficacy of 5-HT3 antagonists (e.g., odansetron) as antiemetics. Histologic examination of mucosal biopsies taken from patients with IBS show elevated numbers of enterochromaffin cells and mucosal mast cells, both of which are paracrine sources of serotonin.11,112 Measurements of elevated postprandial concentrations of serotonin in the hepatic portal blood of patients with IBS suggest that elevated release and excessive stimulation of 5-HT3 receptors on intestinal afferent nerve terminals might account for the abdominal pain and discomfort experienced by these patients. Efficacy of 5-HT3 receptor antagonists (e.g., alosetron) in the treatment of abdominal pain in women with the diarrheapredominant form of IBS adds to the findings that suggest a role for disordered release of serotonin and its action to stimulate firing in sensory afferents as underlying pathologic factors in IBS.15

Mechanoreception Low- and high-threshold mechanoreceptors code mechanosensory information from the gastrointestinal and biliary tracts. Low-threshold splanchnic afferents respond to innocuous levels of distention; high-threshold afferents respond to distending volumes (e.g., balloon distention) only at strengths above a set threshold. The latter highthreshold afferents are predominantly unmyelinated C fibers. Low-threshold afferents are myelinated fibers of the A␦ class. Low-threshold mechanoreceptors are believed to be the afferent component of autonomic regulatory reflexes. Whether some degree of activity in lowthreshold pathways reaches the level of conscious perception is uncertain; nevertheless, it is likely that some nonpainful sensations such as abdominal fullness and the presence of gas are derived from the low-threshold– related activity. High-threshold afferents are thought to be the sensory correlates of sharp, localized pain in organs such as the gallbladder, where pain is the only consciously perceived sensation.17 Combined involvement of both lowand high-threshold fibers in nociperception cannot be excluded. Cervero and Janig18 suggested for organs such as the urinary bladder and colon, where distention can evoke sensations ranging from mild fullness to intense pain, that activation of differing proportions of low- and high-threshold mechanoreceptors accounts for the gradation of sensations. Acute visceral pain may emerge from activation of highthreshold nociceptive fibers, whereas chronic forms of visceral pain may be attributed to sensitization of both types of mechanoreceptors by inflammatory conditions or ischemia. Application of acetic acid or other irritants to the large

intestinal mucosa in animals lowers the threshold and sensitizes both the high- and low-threshold distention-sensitive afferents. Involvement of another class of splanchnic afferents termed silent nociceptors is also suspect in chronic abdominal pain. Silent nociceptors are splanchnic afferents that do not discharge impulses in response to the strongest of distending stimuli. Inflammatory mediators sensitize these normally silent receptors. Spontaneous firing and responses to normally innocuous mechanical distention can be recorded in previously unresponsive splanchnic afferents after the bowel is inflamed by experimentally applied irritants or inflammatory substances. Reports that a significant fraction of patients with IBS progress to the development of IBS-like symptoms after an acute bout of infectious enteritis is reminiscent of the basic scientific findings that sensory afferents become sensitized in inflammatory states.25,26 Gwee et al.45 reported that 23% of patients with acute gastroenteritis progressed to IBS-like symptoms within 3 months. Nevertheless, the question of whether the association between acute infectious enteritis and IBS reflects low-level inflammation (e.g., microscopic enteritis) and chronic sensitization of intestinal afferents by inflammatory mediators is not fully resolved. Primary sensory afferent terminals in the intestine express receptors for several different messenger substances, including inflammatory mediators. Receptors for 5-HT, bradykinin, ATP, adenosine, prostaglandins, leukotrienes, and proteases are among those expressed by the afferent terminals.71 Accumulation of any of these substances has potential for increasing the sensitivity of intestinal afferents, especially in disordered conditions such as inflammation and ischemia. Mediators released during mast cell degranulation can sensitize “silent” nociceptors in the large intestine. In animal models, mast cell degranulation results in a reduced threshold for pain responses to balloon distention, and treatment with mast cell–stabilizing drugs prevents this lowering of the pain threshold.92

INTERNEURONS Interneurons represent another of the three functional categories of enteric neurons illustrated in Figure 12–2. Enteric interneurons are synaptically interconnected into integrated circuits that process sensory information and direct the activity of motor neurons to the musculature, secretory epithelium, and blood vasculature. They have individualized properties, including specific morphologic characteristics of the cell somas, directionality of projections, length of axonal projections, electrical and synaptic behavior, and neurotransmitter codes. Multiple synaptic connections contact the cell bodies of interneurons and their axons and dendrites in the ganglionic neuropil. Synaptic connections are made between

Neurobiology of the Enteric Nervous System

axons and cell bodies, between axons and dendrites, and from axon to axon. Whereas the bulk of information processing in integrative nervous systems occurs in the synaptic neuropil, only the cell bodies of enteric neurons are presently accessible for electrophysiologic study of electrical and synaptic behavior. As in the brain, the interneuronal microcircuits account for the higher functions that are required for integrated control and coordination of behavior of effector systems within the complete organ.

ENTERIC MOTOR NEURONS Motor neurons innervate the effector systems. The motor neuron pool of the ENS contains both excitatory and inhibitory neurons. Excitatory motor neurons release neurotransmitters that stimulate muscle contraction and secretion. Inhibitory motor neurons release neurotransmitters that suppress muscular contraction.

Secretomotor Neurons Secretomotor neurons are excitatory motor neurons in the ENS that innervate the intestinal crypts of Lieberkühn (Fig. 12–6). They are uniaxonal neurons with characteristic shapes described as Dogiel type I. When secretomotor neurons fire, they release acetylcholine and/or VIP as neurotransmitters at their junctions with the epithelium of the crypts. Collaterals from secretomotor axons innervate submucosal arterioles (Fig. 12–6). Collateral innervation of the blood vessels links blood flow to secretion by releasing

Secretomotor neuron

dila Va so

(+)

tati

on

Elevated blood flow

Intestinal crypt ¨ of Lieberkuhn

Arteriole

Secretion

FIGURE 12–6 Intestinal crypts of Lieberkühn are innervated by secretomotor neurons. Secretomotor neurons in the submucosal plexus have a single axon that projects to the mucosal epithelium (see Fig. 12–4A). When the neuron fires, neurotransmitters that evoke secretion are released at the junctions with the crypts. Axon collaterals to blood vessels simultaneously dilate the vessels to increase blood flow in support of stimulated secretion.

257

acetylcholine simultaneously at neuroepithelial and neurovascular junctions. Acetylcholine acts at the blood vessels to release nitric oxide from the endothelium, which in turn dilates the vessels and increases blood flow in support of stimulated secretion. Secretomotor neurons have receptors that receive excitatory and inhibitory synaptic input from other neurons in the integrative circuitry of the ENS and from sympathetic postganglionic neurons. They are influenced also by paracrine chemical messages from non-neural cell types in the mucosa and submucosa (e.g., enterochromaffin cells and immune/inflammatory cells). Activation of the excitatory receptors on secretomotor neurons stimulates the neurons to fire and release their transmitters at the junctions with the crypts and regional blood vessels. The overall result of secretomotor neuronal firing is stimulation of the secretion of H2O, electrolytes, and mucus from the crypts into the intestinal lumen. Inhibitory inputs hyperpolarize the membrane potential of secretomotor neurons and thereby decrease the probability of firing. The physiologic effect of inhibiting secretomotor firing is suppression of mucosal secretion. Postganglionic neurons of the sympathetic nervous system are an important source of inhibitory input to the secretomotor neurons. Norepinephrine released from sympathetic axons acts at ␣2a noradrenergic receptors to inhibit firing of the secretomotor neurons. Inhibition of secretomotor firing reduces the release of excitatory neurotransmitters at the junctions with epithelial cells in the crypts. The end result is reduced secretion of water and electrolytes. Suppression of secretion in this manner is part of the mechanism involved in sympathetic nervous “shutdown” of gut function in homeostatic states in which blood is shunted from the splanchnic to the systemic circulation. Knowledge of the neurobiology of submucosal secretomotor neurons is key to understanding the pathophysiology of secretory diarrhea as well as constipation. In general, secretomotor hyperactivity is associated with neurogenic secretory diarrhea; hypoactivity is associated with decreased secretion and a constipated state. Suppression of secretomotor firing by antidiarrheal agents (e.g., opiates, clonidine, and somatostatin analogues) is manifest as harder, drier stools. Stimulation by chemical mediators, such as VIP, serotonin, and histamine, is manifest as more liquid stools. Watery diarrhea may be caused by several different pathophysiologic events. For example, secretomotor neurons may be overly stimulated by excessive serotonin release from mucosal enterochromaffin cells, by histamine release from inflammatory/immune cells in the mucosa/submucosa, or by circulating VIP released from a VIPoma outside the gut. Release of histamine or other inflammatory mediators associated with diarrhea not only stimulates the firing of secretomotor neurons, but also simultaneously acts at presynaptic inhibitory receptors to suppress the release of norepinephrine from the

258

Function of the Peripheral Nervous System

postganglionic sympathetic axons that provide inhibitory input to secretomotor neurons.80 Thus two pathologic factors are involved in the production of neurogenic secretory diarrhea. One is overstimulation of secretomotor neuronal firing and the other is presynaptic suppression of norepinephrine from sympathetic postganglionic neurons that results in removal of the sympathetic braking action on the neurons. Secretomotor neurons have excitatory receptors for several neurotransmitters, including acetylcholine, VIP, substance P, and serotonin (see review by Wood170). One of the serotonergic receptors belongs to the 5-HT3 receptor subtype.33 Efficacy of blockade of 5-HT3 receptors by a 5-HT3 antagonist in the treatment of diarrhea in the diarrhea-predominant population of women with IBS14,15 suggests that overstimulation of secretomotor neurons by serotonin is a significant factor in this form of IBS. Observations that IBS symptoms are commonly exacerbated in the postprandial state20,156 and evidence for elevated postprandial release of serotonin6 raise suspicion that overactive release of serotonin from the enterochromaffin cell population in the mucosa underlies the diarrheal symptoms of IBS. This is reinforced by suggestions of elevated numbers of serotonin-containing enterochromaffin cells and also mast cells in the mucosa of patients with IBS.11,112

Excitatory Musculomotor Neurons Acetylcholine and substance P are the principal neurotransmitters released from excitatory motor neurons to evoke contraction of the muscles.28,47,84 The receptors for acetylcholine on the musculature belong to the muscarinic receptor subtype. Receptors for substance P at neuromuscular junctions are the NK-1 subtype. Axons of excitatory musculomotor neurons to the intestinal circular muscle coat project in the orad direction and excitatory musculomotor axons to the longitudinal muscle coat project in the aborad direction.12,13 Axons of excitatory gastric motor neurons project in the orad direction to both the musculature and mucosa.119,127

Inhibitory Musculomotor Neurons Inhibitory motor neurons release neurotransmitters that suppress contractile activity of the musculature. Early evidence implicated a purine nucleotide, possibly ATP, as the inhibitory transmitter released by enteric inhibitory musculomotor neurons. Consequently, for the gut, the term purinergic neuron temporarily became a synonymous term for inhibitory motor neuron. ATP was joined subsequently by VIP, pituitary adenylate cyclase–activating peptide, and nitric oxide as candidate inhibitory transmitters released by enteric inhibitory musculomotor neurons.43,44,65,97,98 Enteric musculomotor neurons that express VIP and/or

nitric oxide synthase innervate the circular muscle coats of the stomach, intestines, gallbladder, and the various smooth muscle sphincters. Inhibitory musculomotor neurons appear as an evolutionary adaptation for neural control of the specialized self-excitatory properties of the musculature. Electrically conducting junctions connect the smooth muscle fibers one to another to form a functional electrical syncytium that is reminiscent of cardiac muscle. Like cardiac muscle, action potentials and pacemaker potentials spread from muscle fiber to muscle fiber in three dimensions and trigger a contraction as they enter each successive muscle fiber. An interconnected network of non-neuronal pacemaker cells, known as interstitial cells of Cajal (ICCs), extends around the circumference and throughout the longitudinal axis of the small and large intestine.124 The ICC network generates electrical pacemaker potentials (also called electrical slow waves) that trigger action potentials and their associated contractions in the intestinal circular muscle coat. The pacemaker potentials spreading from the ICC networks become an extrinsic stimulus to which the circular muscle responds. The physiologic characteristics of the musculature as a self-excitable electrical syncytium suggest that the pacemaker network should continuously evoke contractions that spread in three dimensions throughout the extent of the syncytium, which in effect is the entire length of the intestine. Nevertheless, in the normal bowel, long stretches of intestine can exhibit no contractile activity and appear as if the musculature were in a paralytic state. Ongoing firing of impulses by enteric inhibitory motor neurons accounts for the occurrence of contractile silence (i.e., physiologic ileus) in the self-excitable electrical syncytium that forms the intestinal circular muscle coat. The circular muscle is able to respond to a pacemaker potential only when the inhibitory musculomotor neurons in a segment of intestine are switched to an inactive state by input from other neurons in the interneuronal control circuits. When the inhibitory motor neurons are firing and releasing their inhibitory neurotransmitters at the neuromuscular junctions, the muscle cannot be excited to contractile threshold by the omnipresent ICC pacemaker activity. Likewise, action potentials and associated contractions can only propagate into regions of the musculature where the inhibitory innervation is “turned off.” Consideration of the physiologic properties of the musculature suggests that inhibitory musculomotor neurons and control of their activity by the integrative microcircuits of the ENS have evolved as the mechanism that determines when the ongoing slow waves initiate a contraction, as well as the distance and direction of propagation once the contraction has been initiated. Inhibitory motor neurons to the circular muscle discharge continuously, with action potentials and contractions in the muscle occurring only when the inhibitory

Neurobiology of the Enteric Nervous System

neurons are “switched off” by input from interneurons in the control circuits. In the various smooth muscle sphincters found along the digestive tract, the inhibitory motor neurons are normally quiescent and are switched to an active state with timing appropriate for coordination of the opening of the sphincter with physiologic events in adjacent regions. When this occurs, the inhibitory neurotransmitter relaxes ongoing muscle contraction in the sphincteric muscle and prevents excitation-contraction in the adjacent muscle from spreading into and closing the sphincter. In nonsphincteric circular muscle, the state of activity of inhibitory motor neurons determines the length of a contracting segment by controlling the distance of spread of action potentials within the three-dimensional electrical geometry of the syncytium. Contraction can occur in segments in which ongoing inhibition has been switched off, while adjacent segments with continuing inhibitory neuronal input cannot contract. The boundaries of the contracted segment reflect the transition zone from inactive to active inhibitory musculomotor neurons. The directional sequence in which the inhibitory motor neurons are switched off establishes the direction of propagation of the contraction. Normally, they are switched off in the aboral direction, resulting in contractile activity that propagates in the aboral direction. In the abnormal conditions associated with emesis, the interneuronal microcircuitry must switch off the inhibitory musculomotor neurons in a reversed sequence to account for small intestinal propulsion that travels toward the stomach.74 Observations of continuous patterned discharge of action potentials by some of the myenteric neurons of dog, cat, guinea pig, and rabbit small intestine were early evidence for the continuous neuronal inhibition of the circular muscle.167 The continuous release of the inhibitory neurotransmitters VIP and nitric oxide coincident with the continuous inhibitory activity has been described for the canine small intestine.82,159 Spontaneously occurring inhibitory junction potentials that reflect the ongoing release of the inhibitory neurotransmitter can sometimes be recorded from the muscle. Nevertheless, in most preparations studied electrophysiologically in vitro, the ongoing inhibitory activity is manifest in the muscle as a steady hyperpolarization of the membrane potential and decreased input resistance. This reduces the probability that electrical pacemaker current from the ICCs will depolarize the muscle to the threshold for action potential discharge and the accompanying contraction. The steady inhibitory hyperpolarization probably reflects smoothing of many inhibitory junction potentials as a result of (1) the release of transmitter from multiple sites during asynchronous discharge of different nerve fibers, (2) long diffusion distances from transmitter release sites to the muscle, and (3) the electrical syncytial properties of the muscle. In general, any treatment or condition that ablates enteric inhibitory musculomotor neurons results in tonic

259

contracture and “achalasia” of the intestinal circular muscle. Several circumstances that involve functional ablation of the intrinsic inhibitory neurons are associated with conversion from a hypoirritable condition of the circular muscle to a hyperirritable state. These are experimental application of local anesthetics164; hypoxic vascular perfusion of an intestinal segment59; surgical ablation130; congenital absence, as in Hirschsprung’s disease165; paraneoplastic syndrome131; and inflammatory neuropathy.73 Most evidence suggests that subpopulations of inhibitory musculomotor neurons are tonically active, and that blockade or ablation releases the circular muscle from the inhibitory influence. The behavior of the muscle in these cases is tonic contracture and disorganized phasic contractile activity reminiscent of fibrillation in cardiac muscle.

DISINHIBITORY MOTOR DISORDERS The neuromuscular physiology of the intestine predicts that spasticity and achalasia will accompany any condition in which ablation of inhibitory musculomotor neurons occurs. Without inhibitory control, the selfexcitable syncytium of nonsphincteric regions will contract continuously and behave as an obstruction. This happens because the muscle is freed to respond to the pacemaker with contractions that propagate without amplitude, distance, or directional control. Contractions spreading in the uncontrolled syncytium collide randomly, resulting in fibrillation-like behavior in the affected intestinal segment. Loss or malfunction of inhibitory musculomotor neurons is the pathophysiologic basis of disinhibitory motor disease. It underlies several forms of chronic intestinal pseudo-obstruction and sphincteric achalasia (Table 12-1). Neuropathic degeneration in the ENS is a progressive disease that, in its earlier stages, can be manifest as symptoms that may be confused with IBS.

Chronic Intestinal Pseudo-obstruction Intestinal pseudo-obstruction is failure of intestinal propulsive motility in the absence of any mechanical form of obstruction. Both myopathic and neuropathic forms of

Table 12–1. Disinhibitory Motor Disease Pseudo-obstruction Chagas’ disease Paraneoplastic syndrome Hirschsprung’s disease Idiopathic Achalasia of the lower esophageal sphincter Biliary dyskinesia (sphincter of Oddi achalasia)

260

Function of the Peripheral Nervous System

chronic intestinal pseudo-obstruction are recognized. Degenerative changes in the musculature underlie the myopathic form, whereas neuropathic forms arise from degenerative changes in the ENS. Failure of propulsive motility in the affected length of bowel reflects loss of the neural microcircuits that program and control the repertoire of motility patterns required for the necessary functions of that region of bowel. Pseudo-obstruction occurs in part because contractile behavior of the circular muscle is hyperactive but disorganized in the denervated regions.140 Hyperactive-disorganized contractile activity, recorded with manometric catheters, is a diagnostic sign of the neuropathic form of chronic small bowel pseudo-obstruction in humans. The hyperactive and disorganized contractile behavior reflects the absence of inhibitory nervous control of the muscles that are self-excitable (autogenic) when released from the braking action imposed by inhibitory motor neurons. Chronic pseudo-obstruction is therefore symptomatic of the advanced stage of a progressive enteric neuropathy. Retrospective review of patients’ records suggest that some IBS-like symptoms can be an expression of early stages of the neuropathy.30,62,137 Degenerative noninflammatory and inflammatory ENS neuropathies are two kinds of disinhibitory motor disorders that culminate in pseudo-obstruction. Noninflammatory neuropathies can be either familial or sporadic.16 The mode of inheritance can be autosomal recessive or dominant. In the former, the neuropathologic findings include a marked reduction in the number of neurons in both myenteric and submucosal plexuses, and the presence of round, eosinophilic intranuclear inclusions in approximately 30% of the residual neurons. Histochemical and ultrastructural evaluations reveal the inclusions to be not viral particles, but rather proteinaceous material forming filaments.92,116 Members of two families have been described with intestinal pseudo-obstruction associated with the autosomal dominant form of ENS neuropathy.88,123 The numbers of enteric neurons were decreased in these patients, with no alterations found in the CNS or parts of the autonomic nervous system outside the gut. Degenerative inflammatory ENS neuropathies are characterized by a dense inflammatory infiltrate confined to enteric ganglia. Paraneoplastic syndrome, Chagas’ disease, and idiopathic degenerative disease are recognizable forms of pseudo-obstruction related to inflammatory neuropathies. Paraneoplastic syndrome is a form of pseudo-obstruction in which commonality of antigens between small cell carcinoma of the lungs and enteric neurons leads to autoimmune attack that results in loss of neurons.131 The majority of patients with symptoms of pseudo-obstruction in combination with small cell lung carcinoma have immunoglobulin G autoantibodies that react with their enteric neurons.78 These antibodies react with a variety of molecules expressed on the nuclear membrane of neurons, including

Hu and Ri proteins, and with cytoplasmic antigens (Yo protein) of Purkinje cells in the cerebellum.2,46,62,122 Immunostaining with sera from paraneoplastic patients shows a characteristic pattern of staining in enteric neurons.2 The detection of antienteric neuronal antibodies is a means to a specific diagnosis; nevertheless, the mechanisms by which the antibodies damage the neurons are unresolved. The association of enteric neuronal loss and symptoms of pseudo-obstruction in Chagas’ disease also reflects autoimmune attack on the neurons with accompanying symptoms that mimic the situation in achalasia and paraneoplastic syndrome. Trypanosoma cruzi, the blood-borne parasite that causes Chagas’ disease, has antigenic epitopes similar to enteric neuronal antigens.163 This antigenic commonality activates the immune system to assault the ENS coincident with its attack on the parasite. Idiopathic inflammatory degenerative ENS neuropathy occurs unrelated to neoplasms, infectious conditions, or other known diseases.30,73,137 De Giorgio et al.30 and Smith et al.137 described two small groups of patients with early complaints of symptoms similar to IBS that progressively worsened and were later diagnosed as idiopathic degenerative inflammatory neuropathy based on full-thickness biopsies taken during exploratory laparotomy that revealed chronic intestinal pseudo-obstruction. Each patient had inflammatory infiltrates localized to the myenteric plexus. Sera from the two cases reported by Smith et al.137 contained antibodies against enteric neurons similar to those found in secondary inflammatory neuropathies (i.e., antiHu), but with different immunolabeling patterns characterized by prominent cytoplasmic rather than nuclear staining. Recognition of the brain-like functions of the ENS leads to the conclusion that early neuropathic changes are expected to be manifest as functional symptoms that worsen with progressive neuronal loss. In diagnostic motility studies (e.g., manometry), degenerative loss of enteric neurons is reflected by hypermotility and spasticity140 because inhibitory motor neurons are included in the missing neuronal population.

Sphincteric Achalasia Sphincteric achalasia is one of the disinhibitory gastrointestinal motor disorders. Smooth muscle sphincters provide functional separation of the various specialized compartments of the digestive tract. The lower esophageal sphincter isolates the esophagus from the acidic environment of the stomach. The pyloric sphincter is a barrier to reflux from the duodenum into the stomach. The sphincter of Oddi guards against reflux from the duodenum into the biliary and pancreatic ducts. The internal anal sphincter contributes to the maintenance of fecal continence. Tonic contracture, which is an inherent myogenic property

Neurobiology of the Enteric Nervous System

of the sphincteric musculature, maintains closure of the orifices that separate the compartments. Enteric inhibitory musculomotor neurons innervate the smooth musculature that forms the sphincters. The function of the inhibitory innervation is to release inhibitory transmitters (e.g., VIP and nitric oxide) that act on the muscle to relax contractile tone and thereby open the sphincter. Inhibitory musculomotor neurons are the motor component of lower esophageal sphincter relaxation during swallowing, relaxation of the sphincter of Oddi to admit bile and pancreatic secretions into the duodenum, and internal anal sphincter relaxation for defecation and passing of flatus. Inhibitory control of sphincteric musculature differs from the tonically active inhibitory outflow to the intestinal musculature. Inhibitory musculomotor neurons to the sphincters are normally silent and are transiently activated with appropriate timing for opening of the sphincter and passage of luminal contents from one compartment to another. The sphincter is opened when the inhibitory musculomotor neurons are stimulated to firing threshold by excitatory synaptic input from interneurons in the ENS microcircuits. Closure of the sphincter returns on termination of the excitatory input and halt in firing of the inhibitory neurons. The requirement for presence of inhibitory musculomotor neurons to open sphincters predicts that ablation of the ENS will lead to sphincteric achalasia. Failure of development of the fetal ENS, including its population of inhibitory musculomotor neurons, accounts for the constricted terminal large intestinal segment and failure of relaxation of the internal anal sphincter in congenital megacolon (i.e., Hirschsprung’s disease).165 Achalasia in the lower esophageal sphincter leads to dilatation of the esophageal body and dysphagia. Achalasia in this group of patients is associated with loss of the inhibitory innervation of the sphincter and the presence of antimyenteric antibodies in the patient’s serum.160 The situation is similar for achalasia in the sphincter of Oddi, which leads to distention in the biliary tree and the characteristic pain localized in the right upper body quadrant. The hormone CCK is used in an effective test of the functional state of the inhibitory innervation of the sphincter of Oddi in patients with biliary pain of undiagnosed origin.55,121 Endoscopic placement of manometric catheters in the sphincter records changes in resting pressure in response to intravenous injection of CCK. Administration of CCK evokes relaxation in a normally innervated sphincter because excitatory receptors for CCK are expressed by the inhibitory musculomotor neurons that innervate the sphincteric smooth muscle. Conversely, CCK evokes a paradoxical contraction and increase in intraluminal pressure in a denervated sphincter. This occurs because the smooth muscle also expresses excitatory receptors for CCK, and contraction of the sphincter results from direct

261

stimulation of the muscle in the absence of input from inhibitory musculomotor neurons.

NEURONAL ELECTRICAL BEHAVIOR Two types of enteric neurons are distinguished by their electrophysiologic behavior in studies in which electrical activity is recorded with intracellular microelectrodes. The two types are identified as AH/type 2 and S/type 1. AH/type 2 and S/type 1 are an arbitrary combination of terms that recognize Nishi and North102 and Hirst et al.48 as the first to publish electrophysiologic descriptions of the two types of neurons. The names combine the alphabetical terms used by Hirst et al. and the numerical designations of Nishi and North. The terms are facilitating because AH/type 2 neurons usually have Dogiel type II multipolar morphology and S/type 1 neurons generally are unipolar like Dogiel type I neurons. AH/type 2 neurons are distinguished electrophysiologically by (1) higher resting membrane potential and lower input resistance than S/type 1 neurons; (2) no spike discharge to depolarizing current injection or discharge of one or two spikes only at the onset of intraneuronal injection of long-duration depolarizing current pulses (Fig. 12–7A) (3) absence of anodal-break excitation at the offset of hyperpolarizing current pulses; (4) prolonged postspike hyperpolarizing potentials; (5) Ca2⫹ contribution to the inward current of the action potential; and (6) tetrodotoxin-resistant action potentials. In addition, exposure to multivalent cationic Ca2⫹ entry blockers such as Mn2⫹, Mg2⫹, or Cd2⫹ depolarizes the neurons, increases input resistance, and augments excitability; and activation of adenylate cyclase and elevation of intracellular cyclic AMP (cAMP) depolarizes the cells, increases input resistance, and augments excitability. Aside from electrophysiologic properties, most AH neurons (⬃80%), but not S neurons, express the calcium-binding protein calbindin.60 AH neurons comprise the largest proportion of neurons in the intestinal myenteric plexus (⬃70%) and the smallest proportion of submucosal plexus neurons (⬍10%). Most of the multipolar AH-Dogiel type II neurons in the intestinal myenteric plexus project one of their long processes out of the ganglion to pass through the circular muscle coat and enter the mucosa (see Fig. 12–5). Because these terminals of AH-type neurons in the mucosa respond to release of serotonin from enterochromaffin cells in the mucosal epithelium, and because mechanical stimulation of the epithelial lining the lumen activates enterochromaffin cells to release serotonin, there has been an inclination toward calling them “intrinsic primary afferent neurons” and referring to them with the acronym IPANs.36,70 The term suggests sensory function. Nevertheless, the behavior of AH neurons overall does not fit the classic neurophysiologic descriptions of primary sensory afferent neurons

262

Function of the Peripheral Nervous System

A

B 20 mV

200 msec

C

20 mV

Resting Membrane Potentials

200 msec Action potential

Fast (⬍50-ms duration) nicotinic excitatory postsynaptic potentials (EPSPs) are evoked experimentally in most all S-type neurons. S-type neurons are more likely to show spontaneously occurring spike discharge than AH-type neurons. Nevertheless, AH neurons discharge spontaneously when activated by excitatory neurotransmitters or paracrine signals (see Slow Synaptic Excitation below).

10 mV

After-hyperpolarization

4 sec

FIGURE 12–7 Enteric neurons are classified as S/type I or AH/type II based on their electrophysiologic behavior. A, AH/type II neurons had low excitability reflected by discharge of only a single action potential during intraneuronal injection of a 200-ms depolarizing current pulse. B, The S/type I neuron had relatively high excitability reflected by repetitive discharge of action potentials during intraneuronal injection of a 200-ms depolarizing current pulse. The bottom traces in A and B show the current pulses and the top traces shows the neuronal responses to the depolarizing current pulses. C, AH/type II neurons are characterized by long-lasting membrane hyperpolarization (i.e., afterhyperpolarization) following the discharge of an action potential.

(i.e., dorsal root and nodose ganglion cells). “Intrinsic primary afferent” is therefore a confusing misnomer for students of neurophysiology. AH neurons are a unique population of neurons in the mammalian nervous system and would best be called “AH/Dogiel type II” or simply “enteric AH” to convey their functional significance in neural control of the bowel and to retain the term that has been used consistently since being applied first by Hirst et al. in 1974.48 S/type 1 neurons are distinguished by (1) lower resting membrane potentials than AH neurons; (2) higher input resistance than AH neurons; (3) elevated excitability reflected by repetitive spike discharge throughout longduration depolarizing current pulses (Fig. 12–7B); (4) increases in frequency of repetitive discharge in direct relation to increases in amplitude of experimentally evoked membrane depolarization; (5) anodal-break excitation (i.e., action potential discharge) at the offset of intraneuronally injected hyperpolarizing current pulses, another reflection of elevated excitability; (6) abolition of somal spikes by tetrodotoxin; (7) insensitivity to stimulation of adenylate cyclase by forskolin and elevation of cAMP; and (8) insensitivity to multivalent cationic Ca2⫹ entry blockers such as Mn2⫹, Mg2⫹, or Cd2⫹.

Potassium conductance is the main determinant of the resting membrane potential in enteric neurons. The resting potential is usually less than the potassium equilibrium potential, which is approximately ⫺90 mV. In AH neurons, a component of the resting potassium conductance, and consequently the resting potential, are dependent on the concentration of free intracellular Ca2⫹.1,42,110,154 A steady influx of Ca2⫹ is responsible for elevated intraneuronal levels that maintain Ca2⫹-activated K⫹ channels in an open state. This is reflected by high K⫹ conductance, lowered input resistance, and hyperpolarized membrane potential in AH neurons in the absence of any exposure to excitatory chemical signals. Neurotransmitters and paracrine neuromodulators act to increase or decrease the Ca2⫹-activated K⫹ conductance, and this is the key mechanism for up and down modulation of excitability and input-output relations in AH neurons. The functional significance of resting potentials less than the K⫹ equilibrium potential is provision of a mechanism whereby the membrane potential can be modulated in the hyperpolarizing or depolarizing direction as determined by whether the messenger substance acts to increase or decrease the K⫹ conductance. In enteric neurons, inhibitory signal substances such as opioid peptides, galanin, and adenosine decrease neuronal excitability by increasing K⫹ conductance and hyperpolarizing the membrane.95,96,115 Excitatory messengers such as substance P, serotonin, and histamine decrease resting potassium conductance, depolarize the membrane, and enhance excitability.66,69

Action Potentials Action potential generation in the cell bodies of AH neurons differs from spike-generating mechanisms in the cell bodies of S neurons and the AH neurons’ own neurites. The ionic mechanism of action potential generation in the cell bodies of AH neurons includes conductance changes for Ca2⫹, Cl⫺, Na⫹, and K⫹. Both Na⫹ and Ca2⫹ carry the inward current of the rising phase of the spike. A characteristic “shoulder,” reminiscent of the plateau on cardiac action potentials, is present at the onset of the falling phase of the spike in AH neurons.153 The “shoulder” reflects activation of voltage-gated Ca2⫹ conductance in N-type, highvoltage–activated Ca2⫹ channels3 as the membrane is

Neurobiology of the Enteric Nervous System

depolarized by inward Na⫹ current. The Na⫹ channels are typical of tetrodotoxin-sensitive channels found in neurons elsewhere.186 Application of tetrodotoxin reduces the rate of rise, the amplitude, and the threshold, but does not abolish the action potentials in AH neurons. The rate of rise of the spike in tetrodotoxin is increased by elevation of external Ca2⫹, and the pure Ca2⫹ spike in this case is abolished by multivalent ions that block Ca2⫹ entry.54,94,103,153,166 Dihydropyridines and other organic calcium entry blockers that suppress Ca2⫹ conductance in smooth and cardiac muscles do not affect the Ca2⫹ component of the spike in AH neurons.176 AH type neurons possess the full compliment of voltage-gated K⫹ currents generally found in neurons elsewhere. These include A-type, delayed rectifier, and inwardly rectifying currents.141,185 The falling phase of the spike is associated with time- and voltage-dependent activation of delayed rectifier K⫹ channels. Long-lasting hyperpolarizing afterpotentials are the hallmark of the action potentials in AH neurons and the basis for the term AH neuron (see Fig. 12–7B). The afterhyperpolarization activates slowly, beginning from 45 to 80 ms after termination of the spike, and lasts for up to 30 s. The amplitude of the afterpotential summates when two or more spikes are fired in close sequence. An increase in membrane conductance reflected by a decrease in the input resistance occurs during the hyperpolarizing afterpotentials. The amplitude of the afterpotential is increased by elevation of extracellular Ca2⫹ and is suppressed by multivalent ions that block Ca2⫹ entry. The amplitude is reduced and the duration of the hyperpolarizing afterpotential is shortened in bathing media with reduced Ca2⫹.103 The polarity of the hyperpolarizing afterpotential is reversed with current clamp of the membrane potentials to values greater than the K⫹ equilibrium potential of approximately ⫺90 mV. These early findings were evidence that an outward current carried by K⫹ ions generated the hyperpolarizing afterpotential. Ca2⫹-activated K⫹ channels carry the outward K⫹ current responsible for the hyperpolarizing afterpotential. Evidence for this comes from voltage-clamp results, which suggest that the slow outward current is coupled temporally to inward Ca2⫹ currents that are activated by depolarizing voltage steps.50 The amplitude of the current responsible for the hyperpolarizing afterpotential is a direct function of the amount of Ca2⫹ that enters the cell during depolarizing clamp steps or as the result of discharge of action potentials. Further evidence for the Ca2⫹ dependence of K⫹ current activation is the suppression of the current and the associated hyperpolarization produced by application of multivalent cations that block Ca2⫹ entry.42,54,103,109 This is consistent with the results obtained with optical imaging of intraneuronal Ca2⫹, which show elevations of intraneuronal free Ca2⫹ during the hyperpolarizing afterpotential.154

263

The hyperpolarizing afterpotentials function to lengthen the refractory period following the spike, and this automatically limits the frequency of spike discharge by the cell body. During slow synaptic excitation (discussed below), the neurotransmitter or paracrine modulator acts to reduce the hyperpolarizing afterpotential, and this allows the somal membrane to fire repetitively at higher frequency. Modulation of the afterpotential by chemical messengers is part of the overall mechanisms by which the excitability and input-output relations of AH neurons are governed.153 Somal action potentials in S-type neurons and spike generation by neurites of both AH- and S-type neurons are generated by classic mechanisms of activation and inactivation of time- and voltage-dependent Na⫹ and K⫹ conductance channels. The spikes are always abolished by sufficient concentrations of tetrodotoxin, and the rate of rise and amplitude are reduced with depleted Na⫹. Unlike AH neurons, mechanisms for generation of repetitive spike discharge are always activated and the cells will often discharge spontaneously, with the spikes preceded by ramplike prepotentials. In contrast to AH neurons, S neurons fire continuously during long-lasting depolarizing current pulses (see Fig. 12–7B), and the frequency of discharge is related directly to the size of the depolarizing pulse. Prominent positive afterpotentials are associated with the spikes and are not followed by Ca2⫹-dependent afterspike hyperpolarizing potentials like those in AH neurons. Unlike AH neurons, depletion of extracellular Ca2⫹ or activation of adenylate cyclase by forskolin does not elevate excitability. S-type electrophysiologic behavior is a characteristic of musculomotor and secretomotor neurons.

SYNAPTIC TRANSMISSION Fundamental mechanisms for chemically mediated synaptic transmission in the ENS are the same as elsewhere in the nervous system. Synaptic transmitters are released by Ca2⫹-triggered exocytosis from stores localized in vesicles at axonal terminals or transaxonal varicosities. Release is triggered by the depolarizing action of action potentials when they arrive at the release site and open voltageactivated Ca2⫹ channels. Once released, enteric neurotransmitters bind to their specific postsynaptic receptors to evoke inotropic or metabotropic synaptic events. When the receptors are directly coupled to the ionic channel, they are classified as inotropic. They are metabotropic receptors when their effects to open or close ionic channels are indirectly mediated by GTP-binding proteins and the induction of cytoplasmic second messengers (e.g., cAMP). Synaptic events in the ENS are basically the same as in the brain and spinal cord. EPSPs, inhibitory postsynaptic potentials (IPSPs), and presynaptic inhibition and

264

Function of the Peripheral Nervous System

facilitation are the principal synaptic events in the ENS. An enteric neuron may express mechanisms for both slow and fast synaptic neurotransmission. Fast synaptic potentials have durations in the millisecond range; slow synaptic potentials last for several seconds or minutes. Fast synaptic potentials are usually EPSPs. The slow synaptic events may be either EPSPs or IPSPs.

Fast Excitatory Postsynaptic Potentials Fast EPSPs were reported for the earliest intracellular studies of myenteric neurons, but were found only in S neurons in the work of Hirst et al.48 and Nishi and North.102 Fast EPSPs are depolarizing responses with durations less than 50 ms (Fig. 12–8A). Fast EPSPs were later reported to occur in AH and S neurons in both myenteric and submucosal plexuses. They appear to be the sole mechanism of transmission between vagal efferents and enteric neurons.125 Most of the fast EPSPs are mediated by acetylcholine acting at nicotinic postsynaptic receptors. The actions of 5-HT at the 5-HT3 serotonergic receptor subtype and purine nucleotides at P2X purinergic receptors behave much like fast EPSPs, and it is possible that some fast EPSPs are serotonergic or purinergic.37,57 Responses A

Action potential Stimulus artififact

mediated by 5-HT3 serotonergic receptors are found on neurons in both plexuses throughout the gastrointestinal tract, including the gastric corpus.129 Results obtained with patch-clamp recording methods show that the nicotinic and 5-HT3 serotonergic receptors are directly coupled to nonspecific cationic channels. Opening of these channels is responsible for the depolarizing event.32 Rapid desensitization is a characteristic of both the nicotinic and 5-HT3 operated channels, both of which are inotropic. Purinergic “fast” depolarizing responses are metabotropic and mediated by second messenger function of phospholipase A and elevation of inositol trisphosphate (IP3).37 Amplitudes of nicotinic fast EPSPs in the intestine become progressively smaller when they are evoked repetitively by focal electrical stimulation applied to the surface of the ganglion or interganglionic fiber tract. Decrease in EPSP amplitude occurs at stimulus frequencies as low as 0.1 Hz, and the rate of decline is a direct function of stimulus frequency. Rundown of this nature does not occur at the synapses in the stomach129,149 or gallbladder.85 The rundown phenomenon reflects presynaptic inhibition of acetylcholine release by additional transmitter substances broadly released by the electrical stimulus or by negative feedback involving autoinhibition of acetylcholine release B Stimulus artififact

IPSP

EPSPs

10 mV 10 mV

0.5 sec

10 msec

C

On Off Stimulus

40 mV 20 sec

FIGURE 12–8 Fast and slow excitatory postsynaptic potentials (EPSPs) and slow inhibitory postsynaptic potentials (IPSPs) are the principal kinds of synaptic events in enteric neurons. A, Two fast EPSPs were evoked by successive stimuli and are shown as superimposed records. Only one of the EPSPs reached threshold for discharge of an action potential. B, Slow IPSP evoked by stimulation of a sympathetic noradrenergic input to the neuron. C, Slow EPSP evoked by repetitive electrical stimulation of the synaptic input to the neuron. Slowly activating depolarization of the membrane potential continues for more than 2 minutes after termination of the stimulus. Repetitive discharge of action potentials reflects enhanced neuronal excitability during the slow EPSP.

Neurobiology of the Enteric Nervous System

(see Presynaptic Inhibition below). Rundown cannot be attributed to postsynaptic changes, because no decrease in the amplitude of the EPSPs occurs during repetitive applications of acetylcholine from microejection pipettes. Fast EPSPs function in the rapid transfer and transformation of neurally coded information between the elements of the enteric microcircuits. They are the bytes of information in the information-processing operations of the logic circuits. One of the fast EPSPs in Figure 12–8A reached threshold for discharge of an action potential, whereas the other EPSP did not reach threshold. Fast EPSPs do not reach threshold when the neuronal membranes are hyperpolarized during slow IPSPs. They are most likely to reach spike threshold when the membranes are depolarized during slow EPSPs. This effect of slow EPSPs and of slow EPSP-like paracrine mediators is an example of neuromodulation whereby the input-output relations of a neuron to one input is modified by a second synaptic or other kind of modulatory input.

Slow Synaptic Excitation Focal electrical stimulation applied experimentally to interganglionic fiber tracts or to the surfaces of ganglia in both myenteric and submucosal plexuses evokes slow EPSPs in both AH and S neurons. Slow EPSPs in AH neurons are accompanied by conversion from hypoexcitability with no action potential discharge to hyperexcitability with high-frequency discharge of action potentials. Neurons with slow EPSPs are found in the small and large intestine and gastric antrum, but not the gastric corpus or gallbladder. Slow EPSPs are restricted to the enteric microcircuits of specialized digestive compartments where propulsive motility is a significant function. Slowly activating membrane depolarization continuing for several seconds to minutes after termination of release of the neurotransmitter from the presynaptic terminal underlies the slow EPSP (Fig. 12–8C). Enhanced excitability reflected by long-lasting trains of action potentials is the hallmark of the event. Enhanced excitability during the slow EPSP is apparent experimentally as repetitive spike discharge during depolarizing current pulses and as anodal-break excitation at the offset of hyperpolarizing current pulses. AH neurons, which fire only a single spike at the beginning of a depolarizing current pulse in the inactivated state, will fire repetitively in response to depolarizing pulses when the slow EPSP is in effect. When activated by slow synaptic inputs, behavior of AH neurons is much like that of S neurons and may be confused as such if the AH neurons happen to be in an activated state as a result of ongoing release of the transmitter or a paracrine mediator (e.g., inflammatory/immune mediators). Postspike hyperpolarization in AH neurons is suppressed during slow EPSPs. Suppression of the afterhyperpolarization is part of the mechanism that permits

265

repetitive spike discharge at increased frequencies during the enhanced state of excitability. Slow EPSPs differ for AH neurons and S neurons. Slow EPSPs in AH neurons are generally associated with an increase in neuronal input resistance that reflects closure of ionic channels and decreased membrane ionic conductance. In S neurons, a large proportion of which are musculomotor or secretomotor neurons, the slow EPSPs are associated with a decrease in neuronal input resistance that reflects opening of ionic channels and increased membrane ionic conductance. Ionic Mechanisms The ionic mechanism for slow EPSPs in AH neurons includes changes in several ionic conductances. The depolarizing phase occurs when Ca2⫹ channels, normally open at rest, are closed.42,93,110 Closure of the Ca2⫹ channels lowers intraneuronal Ca2⫹, which, in turn, leads to closure of Ca2⫹-activated K⫹ channels. Closure of the K⫹ channels accounts for the increased input resistance seen in the neuron during the depolarization phase of the slow EPSP. Ca2⫹ currents during the spike that in the resting state would lead to postspike increase in Ca2⫹-activated K⫹ conductance are suppressed. This accounts for suppression of the afterhyperpolarization (i.e., the AH) during the EPSP. Enhanced excitability, seen as repetitive spike discharge and anodal break excitation during the EPSP, is probably related to suppression of A-type and delayed rectifier K⫹ currents.141 Unlike AH neurons, the ionic mechanism for the depolarization phase of slow EPSPs in S neurons includes elevated conductance for cations. Opening of Na⫹ and Ca2⫹ conductance channels accounts for the depolarization phase and decreased input resistance seen during the EPSP. Elevated Na⫹ conductance is the major contributor to the depolarization phase in the S neurons.56,58 Mediators of Slow EPSPs Several messenger substances found in neurons, endocrine cells, or immune cells mimic slow EPSPs when applied experimentally to enteric neurons. Receptors for more than one of the messenger substances may be present on the same neuron. Table 12–2 contains a current list of these substances together with receptor subtypes. Substance P (a tachykinin), 5-HT, and acetylcholine fulfill criteria for function as a neurotransmitter in the enteric microcircuits. The other substances are implicated mainly by their presence in enteric neurons and/or by mimicry of the slow EPSP when applied experimentally. Immunohistochemical studies find receptors for these substances, as well as combinations of receptors for the other listed substances, to be co-localized on the same enteric neuronal cell body. Histamine and interleukin-1␤ are examples of slow EPSP mimetics of paracrine origin. Histamine and

266

Function of the Peripheral Nervous System

Table 12–2. Slow EPSP Mimetics Acetylcholine (muscarinic/M1) Cholecystokinin (⫺A) Bombesin Calcitonin gene–related peptide Thyrotropin-releasing hormone 5-Hydroxytryptamine (5-HT1P) Norepinephrine Pituitary adenylate cyclase–activating peptide Interleukin-1␤ Histamine (H2 ) Adenosine (A2)

Vasoactive intestinal peptide Cerulein Gastrin-releasing peptide Tachykinins (NK3 /NK1) Corticotropin-releasing hormone ␥-Aminobutyric acid Motilin Glutamate (group I metabotropic) Interleukin-6 Platelet activating factor Bradykinin (B2)

interleukin-1␤ are released from intestinal mast cells to become a neuromodulatory signal that is decoded by the enteric microcircuits. Mast cells have both detector and signal functions in enteric neuroimmune communication.173 They utilize the specificity and memory of the immune system to detect the presence of threatening antigens in the gut, and after doing so release histamine to signal the ENS to reprogram its output for secretory and propulsive motor behavior that effectively eliminates the antigenic threat from the intestinal lumen.173 The evidence for 5-HT is as complete as for any known neurotransmitter, including synthesis, storage, and release from enteric neurons. Nevertheless, the strongest evidence is that agents such as N-acetyl-5-hydroxytryptophyl-5-hydroxytryptophan amide, renzapride, and anti-idiotypic antibodies to 5-HT block both the slow EPSP-like actions of 5-HT and the slow EPSP in the same AH neurons.39 Aside from its role as a putative enteric neurotransmitter, 5-HT also has a paracrine signaling function. A large fraction of the body’s 5-HT is stored in enterochromaffin cells interspersed in the intestinal epithelial lining. Mechanical stimulation (e.g., shearing forces) at the mucosa or noxious stimulation (e.g., stimulant laxatives such as sennosides) releases the 5-HT, which may then reach receptors on AH neuronal projections in the mucosa, other enteric neural elements, and spinal and vagal sensory afferents.9,40 Bornstein et al.10 reviewed five criteria for transmitter function that were fulfilled by substance P: (1) pharmacologic evidence suggests the intramural release of substance P within intestinal segments; (2) the K⫹ conductance decrease produced by substance P and the slow EPSP is the same; (3) chymotrypsin, which digests substance P, reduces both the response to the peptide and the slow EPSP; (4) the widespread occurrence of slow EPSPs implies that terminals of the responsible axons synapse

with most cell bodies in the ganglion, as is the case for the multipolar (AH/Dogiel type II) neurons that contain substance P; and (5) slow EPSPs were evoked in myenteric neurons with ganglia that contained substance P, but no immunocytochemically demonstrable 5-HT–containing fibers. Slow EPSPs can be initiated by substance P released from AH neurons or from collateral projections of spinal afferents inside the intestinal wall.150 Neurokinin-3 receptors mediate the action of substance P.64,126 The slow EPSP-like action of 5-HT is related to the 5-HP1P serotonergic receptor, so named by Mawe and co-workers.87 This receptor subtype is blocked by the drug renzapride, which is a substituted benzamide compound with stimulatory effects on gastric emptying and intestinal transit.40 Slow EPSP mimetic action of acetylcholine is mediated by the M1 muscarinic receptor subtype,21,107 and the action of histamine occurs at the H2 receptor subtype.99,171 Motilin is a slow EPSP mimetic of particular interest because it is implicated as a messenger substance in the initiation of the intestinal migrating motor complex in the interdigestive period.158 Additional interest emerges from findings that macrolide antibiotics (e.g., erythromycin) act at motilin receptors to facilitate gastric emptying in disorders such as diabetic gastroparesis.148 Slow EPSP-like actions of motilin are prominent in AH neurons of the gastric antrum.178 Signal Transduction in AH Neurons Slow activation and the prolonged time course of slow EPSPs are clues to the mechanism of signal transduction. A 30-ms experimental “puff” of a slow EPSP mimetic or electrical stimulation of synaptic inputs for the slow EPSP evokes changes in excitability that last for several minutes.24,176 This happens in experiments in which the exposure is virtually limited to the 30-ms duration of the puff. The slow EPSP mimetics act transiently at discrete sites on the somal membrane, whereas closure of Ca2⫹-activated K⫹ channels in AH neurons and opening of cation channels in S neurons occurs globally. This global activation in the whole cell suggests involvement of an intraneuronal second messenger system that connects localized surface receptors (i.e., metabotropic receptors) to processes in the neuronal interior. Receptor occupancy by the messenger substance in this case stimulates intraneuronal synthesis or release of a second messenger, which then initiates the neurochemical reactions and molecular conformational changes responsible for transformation of the somal membrane from hypo- to hyperexcitability. Several lines of evidence indicate that receptor-mediated activation of adenylate cyclase, elevation of intraneuronal cAMP, and activation of protein kinase A is the signal transduction mechanism for the slow EPSP in AH neurons. Forskolin, a substance that activates adenylate cyclase, proved to be a useful tool for the study of signal transduction

Neurobiology of the Enteric Nervous System

in AH neurons. Application of forskolin elevates cAMP in the ganglia179 and mimics the slow EPSP in AH- but not in S-type neurons.100 Likewise, other treatments that elevate cAMP, such as intraneuronal injection of cAMP, application of membrane-permeant analogues of cAMP, and treatment with phosphodiesterase inhibitors, each produce slow EPSPlike effects.114 Treatment of myenteric ganglia with 5-HT181 or histamine180 stimulates cAMP formation. Stimulation of cAMP by 5-HT is mediated by 5-HT1P receptors and histamine action is mediated by the H2 receptor subtype. All of the evidence supports the hypothesis that cAMP is an intracellular second messenger in the process of signal transduction in AH neurons. Occupancy of receptors for the slow EPSP activates adenylate cyclase, which in turn leads to synthesis of cAMP, phosphorylation of protein kinases and/or membrane channel proteins, and eventually the dramatic changes in neuronal excitability that occur during the slow EPSP in AH neurons. Receptors for several different kinds of mediators are coupled by G proteins to adenylate cyclase to initiate a single postreceptor cascade of events that culminates in the slow EPSP. Experimental use of forskolin improved insight into the behavior of AH neurons that were found to be inexcitable in their resting state.166 In this state, the neurons show hyperpolarized resting membrane potentials that are close to the potassium equilibrium potential (⬃⫺90 mV) and low input resistance indicative of high resting potassium conductance. Injection of depolarizing current does not evoke action potentials in this state. EPSPs that reflect inputs from other neurons in the circuit may be present but do not evoke action potentials in the inexcitable state. Application of forskolin or one of the endogenous slow EPSP mimetics restores excitability to the AH neurons. When this occurs, the neurons may discharge action potentials with hyperpolarizing afterpotentials if the concentration of the agonist is low, or they may discharge repetitively like an S neuron if exposed to higher concentrations. Neurons responding in this way are interpreted as interneuronal circuit elements that are not used continuously by the system. When the digestive state of the gut does not require operation of the section of circuitry that contains the neurons, the circuit is idled by the low excitability state of its component neurons. The circuit is called to activity by overlays of chemical modulators that act to boost the excitability of its neurons. Signal Transduction in S Neurons Early results suggest that signal transduction for the decreased resistance slow EPSPs in S neurons involves a phospholipase C/Ca2⫹-calmodulin second messenger mechanism. Receptors for the slow EPSP on S neurons are G protein coupled to activation of phospholipase C. Phospholipase C catalyzes the formation of IP3, which acts as a second messenger to release Ca2⫹ from intraneuronal

267

membrane stores. Binding of the released Ca2⫹ to calmodulin activates protein kinase C. Activated protein kinase C phosphorylates cation conductance channels that open when phosphorylated to increase conductance for Na⫹ and Ca2⫹ and thereby depolarize the membrane potential. Opening of the cationic conductance channels accounts for the decreased input resistance observed while recording with microelectrodes during the depolarization phase of the EPSP. Termination of the EPSP results from activation of the intraneuronal phosphatase calcineurin, which catalyzes dephosphorylation of the cationic channels.56 Significance of Slow EPSPs in the Integrated System The functional significance of slow EPSPs in AH neurons is twofold. First is an increased probability of impulse discharge in the postsynaptic neuron that is then transformed into excitation or inhibition at either the next-order neuron or an effector, such as the musculature or secretory epithelium. Intestinal peristaltic propulsion, for example, requires sustained discharge by neuronal elements in the microcircuits to account for the sustained neural drive of several seconds’ duration that occurs in the musculature. The behavior of AH neurons during slow EPSPs fits the requirements for a neuronal element whose functional significance in the circuit would be production of either prolonged excitation or inhibition of the intestinal circular or longitudinal muscle layers or a prolonged glandular secretory response. Application of the putative neurotransmitters for the slow EPSP stimulate the release of the neuromuscular or neuroepithelial neurotransmitters acetylcholine, tachykinins, VIP, and nitric oxide from myenteric and submucosal plexuses of intestinal segments in vivo and in vitro.61,161,183 This is consistent with the hypothesis that AH neurons are interneurons that supply coordinated synaptic drive to the excitatory and inhibitory motor innervation of the musculature and excitatory drive to the secretory epithelium. The extensive ramifications of the processes of the AH/Dogiel type II neurons to innervate large numbers of neurons in the same and neighboring ganglia (see Figs. 12–4B and 12–5) and the localization of substance P in these neurons63,170 is consistent with this function. Slow EPSPs underlie a gating mechanism that controls the spread of action potentials between the neurites arising from opposite poles of the cell bodies of multipolar AH neurons. Figure 12–9 illustrates how slow synaptic gating works in AH/Dogiel morphologic type II neurons. Electrophysiologic recording in AH neurons in the low excitability state shows that action potentials propagating toward the cell body in one of its processes do not fire the membrane of the cell soma. If the invading spike should fire the somal membrane, the action potential in the cell body will be followed by the characteristic afterhyperpolarization, which acts to prevent firing of the cell body by any

268

Function of the Peripheral Nervous System

Closed somal gate

Neurite [1]

Open somal gate

+

Action potential propagation [3]

[1]

Positive feed-forward synaptic circuit

[3]

+

Slow excitatory Synaptic input +

Slow EPSP opens somal gate

[2]

[4]

[2]

[4]

FIGURE 12–9 Slow EPSPs and plasticity of membrane excitability in the cell bodies of AH/Dogiel morphology type II neurons underlie a gating mechanism by which the spread of spike information among neurites, which arise from opposite poles of the cell body, is controlled. Left, When the neuron is in the resting state, excitability of the somal membrane is low, and there is low probability that inbound spike information arriving at the soma in neurite 1 will fire the somal membrane. The discharge is limited to one or a few spikes by the postspike hyperpolarizing afterpotentials (see Fig. 12–7C) in case the somal membrane is fired by an inbound spike. With the somal membrane in a hypoexcitable state, the somal gate is totally or partly closed and no or a restricted amount of information is relayed to the neurites at the opposite poles of the cell body. Right, During a slow EPSP, the probability that the somal membrane will be fired by inbound spikes in neurite 1 and that it will fire repetitively at increased frequency is greatly increased because excitability is enhanced, membrane resistance increased, and hyperpolarizing afterpotentials suppressed. The slow EPSP opens the somal gate so that inbound activity is transferred to neurites 2, 3, and 4 and relayed onto synapses with neighboring neurons (see multipolar morphology of AH neurons in Fig. 12–4B).

additional incoming spikes. The probability for the cell body to be fired by inbound spikes is greatly increased during the slow EPSP when excitability is enhanced, membrane resistance is increased, and afterhyperpolarization is suppressed. The membrane of the cell body behaves like a closed gate to the transfer of spike signals between its neurites when it is in the low-excitability state. In this state, spike information, such as inbound information from the mucosa (see Fig. 12–5), is confined to a single neurite (e.g., neurite 1 in Fig. 12–9). The gate is opened and signals are transferred across the somal membrane to other neurites during the slow EPSP. A closed somal gate isolates the initial segments of each neurite, such that spike discharge in one initial segment does not influence another neurite elsewhere around the cell body (see Figs. 12–4B and 12–5). The increase in membrane resistance during the slow EPSP increases the space constant of the somal membrane and facilitates electrotonic spread of the action potential from the neurites into the cell body. This, together with the enhanced excitability, opens the somal gate, resulting in transfer of signals across the cell body to neurites at

Slow excitatory Synaptic input

AH-Dogiel type II neurons

+ Circuit output Motor neurons Effector systems (musculature, glands, vasculature)

FIGURE 12–10 AH/Dogiel morphologic type II neurons are networked into positive feed-forward synaptic circuits. Feed-forward synaptic excitation is a mechanism for rapid initiation of synchronous discharge in the neural elements of the circuit. Output from the feedforward circuit drives populations of motor neurons, which in turn drive the behavior of the musculature, secretory epithelium, and vasculature. Positive feed-forward synaptic networks synchronize firing of large numbers of motor neurons to ensure simultaneous contraction or inhibition of the muscle around the circumference and along an extended length of bowel. Synchronization of firing of secretomotor neurons ensures coordination of secretion around and along an extended length of bowel.

other poles of the neuron. When this happens, all neurites fire in synchrony with the cell body and the action potentials are conducted away to be distributed in regions of the enteric networks lying further along or around the intestine. Somal gating within a pool of AH neurons that are synaptically interconnected in a feed-forward excitatory circuit (Fig. 12–10) is believed to be the mechanism for rapid buildup of action potential discharge in the neuronal pool. This results in simultaneous spike discharge, characteristic of the slow EPSP, in a population of AH neurons positioned around the circumference and along the length of an intestinal segment. Output from the network of AH neurons drives the discharge of the pools of excitatory and inhibitory musculomotor neurons and secretomotor neurons that innervate the intestine’s effector systems. In a segment of bowel, this mechanism ensures that muscular contraction or inhibition occurs simultaneously around the circumference and along the entire length of the segment.155,170 The same can be postulated for secretion in an intestinal segment. As would be the case also for muscular activity, an effective secretory response requires that the response be activated simultaneously around the circumference and along an extended length of intestine. These functional requirements are achieved by feed-forward

Neurobiology of the Enteric Nervous System

buildup of excitation in AH neuronal driver circuits that connect to secretomotor and musculomotor neuronal pools. The projectional geometry of the neurites identifies the significance of slow excitatory input to the cell body of AH neurons as a gating determinant of neurite-to-neurite communication across the cell soma. Most all myenteric AH neurons project one or more neurites to the mucosa (see Fig. 12–5). This multiplies their functional significance because the projections in the mucosa fire in response to acid pH and in response to serotonin released by mechanical stimulation of mucosal enterochromaffin cells.8,36,70 Figures 12–5, 12–9, and 12–10 illustrate the functional complexity contributed by AH neurons at the circuit level of organization. As coupled elements in the control circuits for the musculature and glandular epithelium, they are stimulated by slow synaptic inputs to fire prolonged trains of impulses. This kind of activity in the cell body is transformed into slow synaptic output to neighboring AH neurons in the circuit and results in feed-forward buildup of neuronal firing in the whole circuit. In the absence of slow synaptic input, the somal gate is closed (see Figs. 12–5 and 12–9). In this state, inbound information from the mucosa is not gated to neurites across the cell body, but may become the afferent arm of an axon reflex. In circumstances in which the cell body is activated by slow synaptic or paracrine input, the somal gate is open and the transformed information from the mucosa is gated in the direction of slow synaptic output from both sides of the cell soma, as illustrated in Figure 12–5.

Inhibitory Postsynaptic Potentials Slow IPSPs are hyperpolarizing synaptic potentials found in both myenteric and submucosal ganglion cell somas of the small and large intestine and in myenteric neurons of the gastric antrum.52,108,149,176,184 IPSPs activate slowly and continue for several seconds after termination of the stimulation (see Fig. 12–8B). Slow IPSPs are hyperpolarizing responses associated with decreased input resistance and suppression of excitability. Reduction in the membrane resistance and excitability together with membrane hyperpolarization decrease the probability of action potential discharge that might occur spontaneously in response to excitatory synaptic inputs or to inbound action potentials in a neurite. IPSPs are more readily demonstrated in the submucosal than in the myenteric plexus of in vitro preparations. Earlier studies found slow IPSPs in less than 10% of small intestinal myenteric neurons,67,176 whereas over 80% of submucosal neurons show IPSPs in response to electrical stimulation.108,142 Interaction with slow EPSPs counteracts the inhibitory inputs and is a factor accounting for the low incidence of IPSPs in myenteric plexus studies.

269

Experimental stimulation of interganglionic connectives activates both excitatory and inhibitory inputs, and the excitatory inputs predominate to a variable degree. Selective blockade of slow synaptic inputs by adenosine A1 receptor agonists uncovers and enhances stimulus-evoked slow IPSPs in the myenteric plexus.23 The decrease in input resistance observed during slow synaptic inhibition reflects increased conductance in potassium channels.49,91,134,144 Of the variety of potassium channels found in the membranes of enteric neurons, inwardly rectifying potassium channels are the ones opened by a variety of inhibitory neurotransmitters to account for the increased conductance and hyperpolarization of the membrane potential during the IPSP.49,50,91,144 The Ca2⫹-activated potassium channels that both contribute to the resting membrane potential and account for postspike hyperpolarization are not involved in the generation of the slow IPSPs.106,134,144,145 GTP-binding proteins are suggested to be involved in direct coupling of the receptors to the potassium channels.90,144 Mediators of Slow IPSPs Several putative neurotransmitters and paracrine signal substances evoke slow IPSP-like responses when experimentally applied to enteric neurons. Some of these substances are peptides, others are purinergic compounds, and another is norepinephrine (Table 12–3). Receptors for two or more of these substances may be localized to the cell body of the same neuron. Enkephalins, dynorphin, and morphine all mimic the IPSPs in enteric neurons. Actions of these substances are mediated by opiate receptors and are limited to subpopulations of neurons. Opiate receptors of the ␮ subtype predominate on myenteric neurons, whereas the receptors on submucosal neurons are the ␦-opioid subtype.111 Morphine addiction and effects of withdrawal in enteric neurons is observed as high-frequency discharge of impulses on the addition of naloxone during chronic exposure to morphine.105,106 IPSP-like actions of nociceptin involve an ORL1 receptor that is distinct from typical naloxone-sensitive opioid receptors.79 Table 12–3. Slow IPSP Mimetics Acetylcholine 5-Hydroxytryptamine (5-HT1A) Neurotensin Somatostatin Adenosine (A1) Nociceptin Opioid peptides Norepinephrine (␣2) Cholecystokinin ATP Galanin

270

Function of the Peripheral Nervous System

Norepinephrine binds to ␣2a adrenoceptors to evoke IPSPs. Noradrenergic IPSPs occur primarily in neurons of the intestinal submucosal plexus.108,144 The noradrenergic inputs come from postganglionic fibers of the sympathetic nervous system that are involved in inhibition of secretomotor activity. Galanin is a 29–amino-acid polypeptide that mimics slow IPSPs in almost all of the neurons of the intestinal myenteric and submucosal plexuses.113 Cholecystokinin does this also, but the effect is only seen in a subpopulation of 10% to 15% of myenteric neurons in the small intestine or gastric antrum.101,132,133 The application of adenosine, ATP, and other purinergic analogues simulates the slow IPSP in intestinal myenteric neurons.115 This is seen in nearly all AH neurons and results from the suppression of adenylate cyclase and reduction in intraneuronal levels of cAMP.182 Both the degradative enzyme adenosine deaminase and a selective adenosine A1 receptor antagonist potentiate slow EPSPs in vitro.23 Potentiation of the slow EPSPs reflects reduction of ongoing inhibitory action of endogenously released adenosine and its accumulation in the intestinal wall. Somatostatin is implicated as the endogenous inhibitory transmitter responsible for the intrinsic inhibitory inputs to submucosal ganglion cells. Hirst and Silinsky53 described IPSPs in the submucosal plexus that persisted after sympathectomy and could therefore not be attributed to activation of norepinephrine release from sympathetic postganglionic fibers. This IPSP, unlike the noradrenergic IPSPs, requires repetitive stimulation and occurs with a more slowly developing time course than noradrenergic IPSPs.80 Somatostatincontaining nerve terminals and the nonadrenergic IPSP disappear in parallel after destruction of the myenteric plexus and its projections to the submucosal plexus.177 Functional Significance of Slow IPSPs A reduction in the membrane resistance and excitability of the somal membranes, together with membrane hyperpolarization during the slow IPSPs, decrease the probability of action potential discharge. The probability is reduced that the somal membrane will be fired during electrotonic invasion by spikes in the initial axonal segment or during excitatory synaptic input (see Figs. 12–5 and 12–9). This influence is the inverse of the slow EPSP and acts to close the gate for transfer of spike information across the multipolar soma of AH neurons. Slow synaptic inhibition probably functions to terminate the excitatory state of slow synaptic excitation and reestablish the low excitability state in the ganglion cell soma of AH neurons. This may be a step in the control of sequentially occurring motor events such as the conversion from inhibition to excitation in the circular muscle of an intestinal segment during propagation of propulsive peristalsis. In the intact animal, there may also be inhibitory substances of endocrine or paracrine origin that function in

particular situations to lock the somal membranes in a low excitability state.

Presynaptic Facilitation Presynaptic facilitation refers to enhancement of synaptic transmission that results from actions of chemical mediators at neurotransmitter release sites on enteric axons. This is known to occur at fast excitatory synapses in the myenteric plexus of the small intestine and gastric antrum and at noradrenergic inhibitory synapses in the submucosal plexus. CCK mediates presynaptic facilitation in gallbladder ganglia.86 Presynaptic facilitation is evident as an increase in amplitude of fast EPSPs at nicotinic synapses, where it reflects enhanced release of acetylcholine from axonal release sites. At noradrenergic inhibitory synapses in the submucosal plexus, presynaptic facilitation appears as enhancement of the hyperpolarizing responses to stimulation of sympathetic postganglionic fibers and is associated with elevation of cAMP in the sympathetic nerve terminals.184

Presynaptic Inhibition Presynaptic inhibition refers to mechanisms that suppress release of neurotransmitters from axons. It involves binding of chemical messengers to inhibitory receptors at transmitter release sites on the axon. Presynaptic inhibition is a significant synaptic event within the enteric microcircuits of the gastric corpus and antrum, as well as the small and large intestine and rectum of the guinea pig.51,104,129,149,153,167 Presynaptic inhibitory receptors are found at fast and slow excitatory synapses, at inhibitory synapses, and at neuromuscular junctions. Presynaptic inhibition in many cases involves axoaxonal transmission in which release of a neurotransmitter from one axon acts at receptors on another axon to suppress release of transmitter from the second axon. Presynaptic inhibition also takes the form of autoinhibition and can occur at both neural synapses and neuroeffector junctions. Autoinhibition occurs when the transmitter released from the same enteric neuron accumulates in the vicinity of the release site and activates presynaptic inhibitory receptors to suppress further release. It functions in this way as a negative feedback mechanism that automatically regulates the concentration of neurotransmitter within the synaptic or junctional space. Substances of non-neural origin, released in a paracrine or endocrine manner into the milieu surrounding the synapse, are known to act presynaptically to suppress neurotransmission. Mast cells and enteroendocrine cells are sources of non-neuronal messages to presynaptic receptors in the ENS. Several messenger substances found in neurons, enteroendocrine cells, or immune cells of the gut produce presynaptic inhibition when applied experimentally to enteric synapses (Table 12–4). Norepinephrine acts at

271

Neurobiology of the Enteric Nervous System

Table 12–4. Mimetics for Presynaptic Inhibition Norepinephrine (␣2a) Histamine (H3) Opioid peptides Neuropeptide Y Peptide YY adenosine (A1) Dopamine 5-Hydroxytryptamine (5-HT4?) Acetylcholine (muscarinic) Pancreatic polypeptide Cytokines

presynaptic ␣2a receptors to suppress transmission at both slow and fast excitatory synapses and at excitatory neuroeffector junctions.144 The presynaptic inhibitory actions of norepinephrine are significant because it is the neurotransmitter released at the synaptic interface between postganglionic sympathetic axons and the ENS.172 Electrical stimulation of sympathetic postganglionic fibers in the periarterial mesenteric nerves suppresses fast EPSPs in myenteric ganglion cells without altering the responses to exogenously applied acetylcholine.51,102 Norepinephrine also fulfills the criteria for a presynaptic action by suppressing stimulus-evoked slow EPSPs without effect on the postsynaptic actions of substance P, serotonin, or acetylcholine.177 As expected for a presynaptic action, norepinephrine does not abort a slow EPSP in progress, but blocks ability to evoke subsequent EPSPs.177 Norepinephrine reduces release of 5-HT and substance P from isolated intestinal segments, consistent with presynaptic suppression of slow synaptic excitation.5,68 Electron micrographs show axoaxonal synapses with vesicular profiles characteristic of noradrenergic synapses in the guinea pig myenteric plexus.83 Observations of suppressive effects of norepinephrine on acetylcholine release from intestinal segments in vitro are supporting evidence for a presynaptic action.31,72,117 Histamine, interleukin-1␤, interleukin-6, and plateletactivating factor receive attention because of their role in neuroimmune communication between mast cells and the ENS program circuits. Histamine, like the other listed immune/inflammatory messengers, acts at presynaptic receptors on cholinergic axons to suppress fast EPSPs at nicotinic synapses and at sympathetic nerve terminals to suppress the release of norepinephrine in the ENS microcircuits.80,168,171 This action of histamine is blocked by selective H3 receptor antagonists and mimicked by selective H3 agonists. The evidence indicates that presynaptic action of histamine is at the H3 receptor subtype.80,152 Histamine released during antigenic degranulation of intestinal mast cells acts in like manner.34,35 Application of 5-HT to enteric synapses reduces the amplitude of the fast EPSPs without effects on nicotinic postsynaptic responses to acetylcholine, thereby fulfilling

criteria for serotonergic presynaptic inhibition.104 The receptor for the presynaptic inhibitory action is not unequivocally identified, but fits the profile of the 5-HT1 receptor in the gastric antrum and small intestine38,143,147 or the putative 5-HT4 receptor subtype in the colon.33 The presynaptic action of acetylcholine on fast EPSPs occurs at muscarinic receptors in the myenteric and submucosal plexuses of the small and large intestine107,153 and at myenteric synapses in the gastric antrum.149 The affinity of the presynaptic receptors for selective agonists and antagonists is suggestive of the M2 class of muscarinic receptor.107 The presynaptic muscarinic receptors at fast nicotinic synapses are components of autoinhibitory mechanisms that function in negative feedback regulation of the amount of acetylcholine released by the arrival of the action potential at the release site. Inhibition of acetylcholinesterase by drugs such as eserine results in suppression of stimulus-evoked fast EPSPs, and this effect is reversed by the application of atropine. In this situation, inhibition of the cholinesterase permits accumulation of acetylcholine, which feeds back on the presynaptic muscarinic receptors to suppress further release of acetylcholine. Atropine blocks this presynaptic influence of the esterase and restores the EPSP.153,167 Members of the pancreatic polypeptide family of messenger peptides (e.g., NPY, PYY, PP) act presynaptically to suppress transmission at nicotinic synapses in the microcircuits of the stomach.128 The presynaptic inhibitory action of NPY is significant because it is co-localized with norepinephrine in sympathetic postganglionic axons as well as being found in intrinsic neurons. This suggests that both the CNS and the ENS may use the long-lasting inhibitory action of this mediator to selectively inactivate synapses within the microcircuits. Presynaptic NPY receptors appear to be present on vagal efferent fibers in the gastric corpus and participate in circuit functions by which the ENS microcircuitry of the corpus may ignore its vagal input.128 Presynaptic inhibition by adenosine occurs at the A1 type of P1 purinoreceptor.4,22,23 This occurs at fast nicotinic synapses, slow excitatory synapses, and noradrenergic inhibitory synapses. Suppression of fast EPSPs occurs in the gastric as well as intestinal microcircuits, whereas presynaptic inhibition of IPSPs is seen in the intestinal submucosal plexus.

CONCLUSIONS Many lines of evidence now implicate dysfunction in the ENS as an important factor underlying symptomatology in patient complaints that fit criteria for functional gastrointestinal disorders. This underscores a need for expanded attention to the neural factors that might explain the

272

Function of the Peripheral Nervous System

functional disorders and justifies accelerated expansion of the basic and clinical investigation that defines the new subspecialty of neurogastroenterology. Consideration that the ENS is an independent integrative nervous system, with most of the neurophysiologic complexities found in the CNS, stimulates premonition that functional motor disorders may reflect neuropathies in the “brain-in-the-gut.”174 Lack of understanding of how subtle malfunctions may occur in the synaptic microcircuits of the ENS is undoubtedly the basis for the use of “functional” as the description for functional disorders such as nonulcer dyspepsia and IBS. The current state of understanding of functional gastrointestinal disorders is reminiscent of neurologic disorders such as Parkinsonian tremors, ballisms, and choreas that were classified as “functional” before understanding of neurotransmission in microcircuits of somatic motor centers in the brain. Parkinson’s disease is a classic example of a functional somatic motor disorder that was ultimately explained as malfunction of dopaminergic neurotransmission in localized brain centers. Like the status of understanding of somatic motor control centers in the brain a half-century ago, now as we enter the 21st century, the ENS remains a virtual “black box” that will need to be opened scientifically to understand gastrointestinal disorders more fully.

REFERENCES

9.

10.

11.

12. 13. 14.

15.

16.

17. 18. 19.

1. Akasu, T., and Tokimasa, T.: Potassium currents in submucous neurones of guinea-pig caecum and their synaptic modification. J. Physiol. (Lond.) 416:571, 1989. 2. Altermatt, H., Rodriguez, M., and Scheithauer, B.: Paraneoplastic anti-Purkinje and type 1 antineuronal nuclear autoantibodies bind selectively to central, peripheral autonomic nervous system cells. Lab. Invest. 65:412, 1991. 3. Baidan, L. V., Zholos, A. V., and Wood, J. D.: Modulation of calcium currents by G-proteins and adenosine receptors in myenteric neurones cultured from guinea-pig small intestine. Br. J. Pharmacol. 116:1882, 1995. 4. Barajas-Lopez, C., Surprenant, A., and North, R. A.: Adenosine A1 and A2 receptors mediate presynaptic inhibition and postsynaptic excitation in guinea pig submucosal neurons. J. Pharmacol. Exp. Ther. 258:490, 1991. 5. Bartho, L., Holzer, P., and Lembeck, F.: Sympathetic control of substance P releasing enteric neurones in the guinea pig ileum. Neurosci. Lett. 38:291, 1983. 6. Bearcroft, C. P., Perrett, D., and Farthing, M. J. G.: Postprandial plasma 5-hydroxytryptamine in diarrhoea predominant irritable bowel syndrome: a pilot study. Gut 42:42, 1998. 7. Berthoud, H. R., and Powley, T. L.: Vagal afferent innervation of the rat fundic stomach—morphological characterization of the gastric tension receptor. J. Comp. Neurol. 319:261, 1992. 8. Bertrand, P. P., Kunze, W. A. A., Bornstein, J. C., et al.: Analysis of the responses of myenteric neurons in the small

20.

21.

22.

23.

24.

25.

26. 27.

intestine to chemical stimulation of the mucosa. Am. J. Physiol. 36:G422, 1997. Beubler, E., and Schirgi-Degen, A.: Serotonin antagonists inhibit sennoside-induced fluid secretion and diarrhea. Pharmacology 47(Suppl. 1):64, 1993. Bornstein, J. C., North, R. A., Costa, M., and Furness, J. B.: Excitatory synaptic potentials due to activation of neurons with short projections in the myenteric plexus. Neuroscience 11:723, 1984. Bose, M., Nickols, C., Feakins, R., and Farthing, M. J.: 5-Hyroxytryptamine and enterochromaffin cells in the irritable bowel syndrome. Gastroenterology 118(Suppl. 2):A563, 2000. Brookes, S. J. H.: Classes of enteric nerve cells in the guinea-pig small intestine. Anat. Rec. 262:58, 2001. Brookes, S. J. H.: Retrograde tracing of enteric neuronal pathways. Neurogastroenterol. Motil. 13:1, 2001. Camilleri, M., Chey, W. Y., Mayer, E. A., et al.: A randomized controlled clinical trial of the serotonin type 3 receptor antagonist alosetron in women with diarrhea-predominant irritable bowel syndrome. Arch. Intern. Med. 23:1733, 2001. Camilleri, M., Northcutt, A. R., Kong, S., et al.: Efficacy and safety of alosetron in women with irritable bowel syndrome: a randomized, placebo-controlled trial. Lancet 355:1035, 2000. Camilleri, M., and Phillips, S.: Disorders of small intestinal motility. Gastroenterol. Clin. North Am. 18:405, 1989. Cervero, F.: Afferent activity evoked by natural stimulation of the biliary system in the ferret. Pain 13:137, 1982. Cervero, F., and Janig, W.: Visceral nocireceptor: a new world order? Trends Neurosci. 15:374, 1992. Chey, W. D., Kim, M., Hasler, W. L., and Owyang, C.: Hyperglycemia alters perception of rectal distention and blunts the rectoanal inhibitory reflex in healthy volunteers. Gastroenterology 108:1700, 1995. Chey, W. Y., Jin, H. O., Lee, M. H., et al.: Colonic motility abnormality in patients with irritable bowel syndrome exhibiting abdominal pain and diarrhea. Am. J. Gastroenterol. 96:1499, 2001. Christofi, F. L., Palmer, J. M., and Wood, J. D.: Neuropharmacology of the muscarinic antagonist telenzepine in myenteric ganglia of the guinea-pig small intestine. Eur. J. Pharmacol. 195:333, 1991. Christofi, F. L., Tack, J., and Wood, J. D.: Suppression of nicotinic synaptic transmission by adenosine in myenteric ganglia of the guinea-pig gastric antrum. Eur. J. Pharmacol. 216:17, 1992. Christofi, F. L., and Wood, J. D.: Presynaptic inhibition by adenosine A1 receptors on guinea-pig small intestinal myenteric neurons. Gastroenterology 104:1420, 1993. Clerc, N., Furness, J. B., Kunze, W. A., et al.: Long-term effects of synaptic activation at low frequency on excitability of myenteric AH neurons. Neuroscience 90:279, 2000. Coelho, A. M., Fioramonti, J., and Bueno, L.: Mast cell degranulation induces delayed rectal allodynia in rats: role of histamine and 5-HT. Dig. Dis. Sci. 43:727, 1998. Collins, S. M.: Is the irritable gut an inflamed gut? Scand. J. Gastroenterol. 27(Suppl. 192):102, 1992. Cooke, H. J.: Complexities of nervous control of the intestinal epithelium. Gastroenterology 94:1087, 1988.

Neurobiology of the Enteric Nervous System 28. Daniel, E. E., Parrish, M. B., Watson, E. G., et al.: The tachykinin receptors inducing contractile responses of canine ileum circular muscle. Am. J. Physiol. 31:G161, 1995. 29. Davison, J. S., and Clarke, G. D.: Mechanical properties and sensitivity to CCK of vagal gastric slowly adapting mechanoreceptors. Am. J. Physiol. 255:G55, 1988. 30. De Giorgio, R., Bassotti, G., Stanghellini, V., et al.: Clinical, morpho-functional and immunological features of idiopathic myenteric ganglionitis. Gastroenterology 110:A655, 1996. 31. Del Tacca, M., Soldani, G., Selli, M., and Crema, A.: Action of catecholamines on release of acetylcholine from human taenia coli. Eur. J. Pharmacol. 9:80, 1970. 32. Derkach, V., Surprenant, A., and North, R. A.: 5-HT3 receptors are membrane ion channels. Nature 339:706, 1989. 33. Frieling, T., Cooke, H. J., and Wood, J. D.: Serotonin receptors on submucous neurons in the guinea-pig colon. Am. J. Physiol. 261:G1017, 1991. 34. Frieling, T., Cooke, H. J., and Wood, J. D.: Neuroimmune communication in the submucous plexus of guinea-pig colon after sensitization to milk antigen. Am. J. Physiol. 267:G1087, 1994. 35. Frieling, T., Palmer, J. M., Cooke, H. J., and Wood, J. D.: Neuroimmune communication in the submucous plexus of guinea-pig colon after infection with Trichinella spiralis. Gastroenterology 107:1602, 1994. 36. Furness, J. B., Kunze, W. A. A., Bertrand, P. P., et al.: Intrinsic primary afferent neurons of the intestine. Prog. Neurobiol. 54:1, 1998. 37. Galligan, J. J., LePard, K. J., Schneider, D. A., and Zhou, X. P.: Multiple mechanisms of fast excitatory synaptic transmission in the enteric nervous system. J. Auton. Nerv. Syst. 81:97, 2000. 38. Galligan, J. J., Surprenant, A., Tonini, M., and North, R. A.: Differential localization of 5-HT1 receptors on myenteric and submucosal neurons. Am. J. Physiol. 255:G603, 1988. 39. Gershon, M. D.: Review article: Roles played by 5-hydroxytryptamine in the physiology of the bowel. Aliment. Pharmacol. Ther. 13:15, 1998. 40. Gershon, M. D.: The Second Brain. New York, HarperCollins, 1998. 41. Gershon, M. D., and Erde, S. M.: The nervous system of the gut. Gastroenterology 80:1571, 1981. 42. Grafe, P., Mayer, C. J., and Wood, J. D.: Synaptic modulation of calcium-dependent potassium conductance in myenteric neurons. J. Physiol. (Lond.) 305:235, 1980. 43. Grider, J. R., Cable, M. B., and Said, S. I.: Vasoactive intestinal peptide as a neural mediator of gastric relaxation. Am. J. Physiol. 248:G73, 1985. 44. Grider, J. R., and Makhlouf, G. M.: Colonic peristaltic reflex: identification of vasoactive intestinal peptide as mediator of descending relaxation. Am. J. Physiol. 251:G40, 1986. 45. Gwee, K. A., Leong, Y. L., Graham, C., et al.: The role of psychological and biological factors in post-infective gut function. Gut 44:400, 1999. 46. Heidenreich, F., Schober, R., Brinck, U., and Hartung, H. P.: Multiple paraneoplastic syndromes in a patient with antibodies to neuronal nucleoproteins (anti-Hu). J. Neurol. 242:210, 1995.

273

47. Hellstrom, P. M., Murthy, K. S., Grider, J. R., and Makhlouf, G. M.: Coexistence of three tachykinin receptors coupled to Ca⫹⫹ signaling pathways in intestinal muscle cells. J. Pharmacol. Exp. Ther. 270:236, 1994. 48. Hirst, G. D. S., Holman, M. E., and Spence, I.: Two types of neurones in the myenteric plexus of duodenum in the guinea-pig. J. Physiol. (Lond.) 236:303, 1974. 49. Hirst, G. D. S., Johnson, S. M., and van Helden, D. F.: The calcium current in a myenteric neurone of the guinea pig ileum. J. Physiol. (Lond.) 361:297, 1985. 50. Hirst, G. D. S., Johnson, S. M., and van Helden, D. F.: The slow calcium-dependent current in a myenteric neurone of the guinea pig ileum. J. Physiol. (Lond.) 361:315, 1985. 51. Hirst, G. D. S., and McKirdy, H. C.: Presynaptic inhibition at a mammalian peripheral synapse. Nature (Lond.) 250:430, 1974. 52. Hirst, G. D. S., and McKirdy, H. C.: Synaptic potentials recorded from neurones of the submucous plexus of guinea-pig small intestine. J. Physiol. (Lond.) 249:369, 1975. 53. Hirst, G. D. S., and Silinsky, E. M.: Some effects of 5-hydroxytryptamine, dopamine and noradrenaline on neurones in the submucous plexus of guinea-pig small intestine. J. Physiol. (Lond.) 251:817, 1975. 54. Hirst, G. D. S., and Spence, I.: Calcium action potentials in mammalian peripheral neurones. Nature (Lond.) 243:54, 1973. 55. Hogan, W. J., Greenen, J. E., and Dodds, W. J.: Paradoxical motor response to cholecystokinin-octapeptide (CCK-OP) in patients with suspected sphincter-of-Oddi dysfunction. Gastroenterology 82:A1085, 1982. 56. Hu, H.-Z., Gao, N., Gao, C., et al.: Calmodulin (CaM) and CaM kinase signaling in the enteric nervous system. Neurosci. Abstr. 27:839.11, 2001. 57. Hu, H.-Z., Gao, N., Lin, Z., et al.: P2X7 receptors in the enteric nervous system of guinea-pig small intestine. J. Comp. Neurol. 440:299, 2001. 58. Hu, H.-Z., Gao, N., Ren, J., et al.: A novel metabotropic P2Y1 receptor in submucous plexus of guinea-pig small intestine. Gastroenterology 120:A509, 2001 59. Hukuhara, T., Kotania, S., and Sato, G.: Effect of destruction of intramural ganglion cells on colon motility: possible genesis of congenital megacolon. J. Physiol. (Japan) 11:635, 1961. 60. Iyer, V., Bornstein, J. C., Costa, M., et al.: Electrophysiology of guinea-pig myenteric neurons correlated with immunoreactivity for calcium binding proteins. J. Auton. Nerv. Syst. 22:141, 1988. 61. Javed, N. H., and Cooke, H. J.: Acetylcholine release from colonic submucous neurons in the guinea pig associated with Cl⫺ secretion. Am. J. Physiol. 262:G1331, 1992. 62. Jean, W., Dalmau, J., Ho, A., and Posner, J. B.: Analysis of the IgG subclass distribution and inflammatory infiltrates in patients with anti-hu-associated paraneoplastic encephalomyelitis. Neurology 44:144, 1994. 63. Jenkinson, K. M., Mann, P. T., Southwell, B. R., and Furness, J. B.: Independent endocytosis of the NK(1) and NK(3) tachykinin receptors in neurons of rat myenteric plexus. Neuroscience 100:191, 2000. 64. Jenkinson, K. M., Morgan, J. M., Furness, J. B., and Southwell, B. R.: Neurons bearing NK(3) tachykinin

274

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75. 76.

77.

78.

79.

80.

81.

Function of the Peripheral Nervous System receptors in the guinea-pig ileum revealed by specific binding of fluorescently labelled agonists. Histochem. Cell Biol. 112:233, 1999. Jin, J. G., Murthy, K. S., Grider, J. R., and Makhlouf, G. M.: Stoichiometry of neurally induced VIP release, NO formation, and relaxation in rabbit and rat gastric muscle. Am. J. Physiol. 34:G357, 1996. Johnson, S. M., Katayama, Y., and North, R. A.: Multiple actions of 5-hydroxytryptamine on myenteric neurones of the guinea-pig ileum. J. Physiol. (Lond.) 304:459, 1980. Johnson, S. M., and North, R. A.: Slow synaptic potentials in neurones of the myenteric plexus. J. Physiol. (Lond.) 301:505, 1980. Jonakait, G. M., Tamir, H., Gintzler, A. R., and Gershon, M. D.: Release of [3H] serotonin and its binding protein from enteric neurons. Brain Res. 174:55, 1979. Katayama, Y., North, R. A., and Williams, J. T.: The action of substance P on neurons of the myenteric plexus of the guinea-pig small intestine. Proc. R. Soc. Lond. 206:191, 1979. Kirchgessner, A. L., Tamir, H., and Gershon, M. D.: Identification and stimulation by serotonin of intrinsic sensory neurons of the submucosal plexus of the guinea pig gut—activity-induced expression of Fos immunoreactivity. J. Neurosci. 12:235, 1992. Kirkup, A. J., Brunsden, A. M., and Grundy, D.: Receptors and transmission in the brain-gut axis: I. Receptors on visceral afferents. Am. J. Physiol. 280:G787, 2001. Knoll, J., and Vizi, E. S.: Presynaptic inhibition of acetylcholine release by endogenous and exogenous noradrenaline at high rate of stimulation. Br. J. Pharmacol. 40:400, 1970. Krishnamurthy, S., and Schuffler, M. D.: Pathology of neuromuscular disorders of the small intestine and colon. Gastroenterology 93:610, 1997. Lang, I. M., and Sarna, S. K.: Motor and myoelectric activity associated with vomiting, regurgitation, and nausea. In Wood, J. D. (ed.): Handbook of Physiology: The Gastrointestinal System, Motility and Circulation. New York, Oxford University Press, p. 1179, 1989. Langley, J. N.: The Autonomic Nervous System, Part I. Cambridge, UK, W. Heffer and Sons, 1921. Langley, J. N., and Magnus, R.: Some observations of movements of the intestine before and after degenerative section of the mesenteric nerves. J. Physiol. (Lond.) 33:34, 1905. Lembo, T., Munakata, J., Mertz, H., et al.: Evidence for the hypersensitivity of lumbar splanchnic afferents in irritable bowel syndrome. Gastroenterology 107:1686, 1994. Lennon, V. A., San, D. F., Busk, M. F., et al.: Enteric neuronal autoantibodies in pseudoobstruction with small-cell lung carcinoma. Gastroenterology 100:137, 1991. Liu, S., Hu, H.-Z., Ren, J., et al.: Pre- and postsynaptic inhibition by nociceptin in guinea pig myenteric plexus in vitro. Am. J. Physiol. 281:G237, 2001. Liu, S., Xia, Y., Hu, H.-Z., et al.: Histamine H3 receptormediated suppression of inhibitory synaptic transmission in the guinea-pig submucous plexus. Eur. J. Pharmacol. 397:49, 2000. Loening-Baucke, V.: Sensitivity of the sigmoid colon and rectum in children treated for chronic constipation. J. Pediatr. Gastroenterol. Nutr. 3:454, 1984.

82. Manaka, H., Manaka ,Y., Kostolanska, F., et al.: Release of VIP and substance-P from isolated perfused canine ileum. Am. J. Physiol. 257:G182, 1989. 83. Manber, L., and Gershon, M. D.: A reciprocal adrenergiccholinergic axoaxonic synapse in the mammalian gut. Am. J. Physiol. 5:E738, 1979. 84. Mao, Y. K., Wang, Y. F., and Daniel, E. E.: Characterization of neurokinin type 1 receptor in canine small intestinal muscle. Peptides 17:839, 1996. 85. Mawe, G. M.: Intracellular recording from gall-bladder neurones of the guinea-pig gall bladder. J. Physiol. (Lond.) 429:323, 1990. 86. Mawe, G. M.: The role of cholecystokinin in ganglionic transmission in the guinea-pig gall-bladder. J. Physiol. (Lond.) 439:89, 1991. 87. Mawe, G. M., Branchek, T. A., and Gershon, M. D.: Peripheral neural serotonin receptors: identification and characterization with specific antagonists and agonists. Proc. Natl. Acad. Sci. U. S. A. 83:9799, 1986. 88. Mayer, E., Schuffler, M., Rotter, J., et al.: Familial visceral neuropathy with autosomal dominant transmission. Gastroenterology 91:1528, 1986. 89. Mertz, H., Morgan, V., Tanner, G., et al.: Regional cerebral activation in irritable bowel syndrome and control subjects with painful and nonpainful rectal distention. Gastroenterology 118:842, 2000. 90. Mihara, S., Hirai, K., Katayama, Y., and Nishi, S.: Mechanisms underlying intracellular signal transduction of the slow IPSP in submucous neurones of the guinea-pig caecum. J. Physiol. (Lond.) 436:621, 1991. 91. Mihara, S., and North, R. A.: Opioids increase potassium conductance in submucous neurones of guinea-pig caecum by activating delta receptors. Br. J. Pharmacol. 88:315, 1986. 92. Monoz-Garcia, D., and Ludwin, S.: Adult-onset neuronal intranuclear hyaline inclusion disease. Neurology 36:785, 1986. 93. Morita, K., and Katayama, Y.: Substance P inhibits activation of calcium-dependent potassium conductances in guinea-pig myenteric neurones. J. Physiol. (Lond.) 447:203, 1992. 94. Morita, K., and North, R. A.: Clonidine activates membrane potassium conductance in myenteric neurones. Br. J. Pharmacol. 74:419, 1981. 95. Morita, K., and North, R. A.: Opiate activation of potassium conductance in myenteric neurons: inhibition by calcium ion. Brain Res. 242:145, 1982. 96. Morita, K., North, R. A., and Tokimasa, T.: The calciumactivated potassium conductance in guinea-pig myenteric neurones. J. Physiol. (Lond.) 329:341, 1982. 97. Murthy, K. S., Grider, J. R., Jin, J. G., and Makhlouf, G. M.: Interplay of VIP and nitric oxide in the regulation of neuromuscular activity in the gut. Arch. Int. Pharmacodyn. Ther. 329:27, 1995. 98. Murthy, K. S., and Makhlouf, G. M.: Vasoactive intestinal peptide pituitary adenylate cyclase-activating peptidedependent activation of membrane-bound NO synthase in smooth muscle mediated by pertussis toxin-sensitive G(i1–2). J. Biol. Chem. 269:15977, 1994. 99. Nemeth, P. R., Ort, C. A., and Wood, J. D.: Intracellular study of effects of histamine on electrical behavior of

Neurobiology of the Enteric Nervous System

100.

101.

102.

103.

104.

105.

106. 107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

myenteric neurons in guinea-pig small intestine. J. Physiol. (Lond.) 355:411, 1984. Nemeth, P. R., Palmer, J. M., Wood, J. D., and Zafirov, D. H.: Effects of forskolin on electrical behavior of myenteric neurones in guinea-pig small intestine. J. Physiol. (Lond.) 376:439, 1986. Nemeth, P. R., Zafirov, D. H., and Wood, J. D.: Effects of cholecystokinin, caerulein and pentagastrin on electrical behavior of myenteric neurons. Eur. J. Pharmacol. 116:263, 1985. Nishi, S., and North, R. A.: Intracellular recording from the myenteric plexus of the guinea-pig ileum. J. Physiol. (Lond.) 231:471, 1973. North, R. A.: The calcium-dependent slow after-hyperpolarization in myenteric plexus neurons with tetrodotoxinresistant action potentials. Br. J. Pharmacol. 49:709, 1973. North, R. A., Henderson, G., Katayama, Y., and Johnson, S.M.: Electrophysiological evidence for presynaptic inhibition of acetylcholine release by 5-hydroxytryptamine in the enteric nervous system. Neuroscience 5:581, 1980. North, R. A., and Karras, P. J.: Opiate tolerance and dependence induced in vitro in single myenteric neurones. Nature 272:73, 1978. North, R. A., and Sieglgansberger, W.: Opiate withdrawal signs in single myenteric neurones. Brain Res. 144:208, 1978. North, R. A., Slack, B. E., and Surprenant, A.: Muscarinic M1 and M2 receptors mediate depolarization and presynaptic inhibition in guinea-pig enteric nervous system. J. Physiol. (Lond.) 368:435, 1985. North, R. A., and Surprenant, A.: Inhibitory synaptic potentials resulting from alpha2 adrenoceptor activation in guinea-pig submucous plexus neurones. J. Physiol. (Lond.) 358:17, 1985. North, R. A., and Tokimasa, T.: Depression of calciumdependent potassium conductance of guinea-pig myenteric neurones by muscarinic agonists. J. Physiol. (Lond.) 342:253, 1983. North, R. A., and Tokimasa, T.: Persistent calciumsensitive potassium current and the resting properties of guinea-pig myenteric neurones. J. Physiol. (Lond.) 386:333, 1987. North, R. A., and Tonini, M.: The mechanism of action of narcotic analgesics in the guinea-pig ileum. Br. J. Pharmacol. 61:541, 1977. O’Sullivan, M., Clayton, N., Breslin, N. P., et al.: Increased mast cells in the irritable bowel syndrome. Neurogastroenterol. Motil. 12:449, 2000. Palmer, J. M., Schemann, M., Tamura, K., and Wood, J. D.: Galanin mimics slow synaptic inhibition in myenteric neurons. Eur. J. Pharmacol. 124:379, 1986. Palmer, J. M., Wood, J. D., and Zafirov, D. H.: Elevation of cyclic adenosine monophosphate mimics slow synaptic excitation in myenteric neurones of the guinea-pig. J. Physiol. (Lond.) 376:451, 1986. Palmer, J. M., Wood, J. D., and Zafirov, D. H.: Purinergic inhibition in the small intestinal myenteric plexus of the guinea-pig. J. Physiol. (Lond.) 387:357, 1987. Palo, J., Haltia, M., Carpenter, S., et al.: Neurofilament subunit-related proteins in neuronal intranuclear inclusion. Ann. Neurol. 15:316, 1984.

275

117. Paton, W. D. M., and Vizi, E. S.: The inhibitory action of noradrenaline and adrenaline on acetylcholine output by guinea-pig ileum longitudinal muscle strip. Br. J. Pharmacol. 35:10, 1969. 118. Read, N. W.: Rectal distention: from sensation to feeling. Gastroenterology 118:972, 2000. 119. Reiche, D., and Schemann, M.: Ascending choline acetyltransferase and descending nitric oxide synthase immunoreactive neurones of the myenteric plexus project to the mucosa of the guinea pig gastric corpus. Neurosci. Lett. 241:61, 1998. 120. Richter, J. E., Barish, C. F., and Castell, D. O.: Abnormal sensory perception in patients with esophageal chest pain. Gastroenterology 91:845, 1986. 121. Rolny, P., Arleback, A., Funch-Jensen, P., et al.: Paradoxical response of sphincter of Oddi to intravenous injection of cholecystokinin or ceruletide: manometric findings and results of treatment in biliary dyskinesia. Gut 27:1507, 1986. 122. Rosenblum, M.: Paraneoplastic and autoimmunologic injury of the nervous system: the anti-hu syndrome. Brain Pathol. 3:199, 1993. 123. Roy, A., Bharucha, H., Nevin, N., and Odling-Smee, G. W.: Idiopathic intestinal pseudo-obstruction: a familial visceral neuropathy. Clin. Genet. 18:292, 1980. 124. Sanders, K. M., Ordog, T., Koh, S. D., and Ward, S. M.: A novel pacemaker mechanism drives gastrointestinal rhythmicity. News Physiol. Sci. 15:291, 2000. 125. Schemann, M., and Grundy, D.: Electrophysiological identification of vagally innervated enteric neurons in the guinea pig stomach. Am. J. Physiol. 263:G709, 1992. 126. Schemann, M., and Kayser, H.: Effects of tachykinins on myenteric neurones of the guinea-pig gastric corpus: involvement of NK-3 receptors. Pflugers Arch. 419:566, 1991. 127. Schemann, M., and Schaaf, C.: Differential projection of cholinergic and nitroxidergic neurons in the myenteric plexus of guinea pig stomach. Am. J. Physiol. 32:G186, 1995. 128. Schemann, M., and Tamura, K.: Presynaptic inhibitory effects of the peptides NPY, PYY, and PP on nicotinic EPSPs in guinea-pig gastric myenteric neurones. J. Physiol. (Lond.) 451:79, 1992. 129. Schemann, M., and Wood, J. D.: Synaptic behavior of myenteric neurones in the gastric corpus of the guinea-pig. J. Physiol. (Lond.) 417:519, 1989. 130. Schiller, W., Suriyapa, C., Mutchler, J., and Anderson, M. C.: Surgical alteration of intestinal motility. Am. J. Surg. 125:122, 1973. 131. Schuffler, M., Baird, H., and Fleming, C.: Intestinal pseudo-obstruction as the presenting manifestation of small cell carcinoma of the lung: a paraneoplastic neuropathy of the gastrointestinal tract. Ann. Intern. Med. 98:129, 1983. 132. Schutte, I. W., Akkermans, L. M., and Kroese, A. B.: CCKA and CCKB receptor subtypes both mediate the effects of CCK-8 on myenteric neurons in the guinea-pig ileum. J. Auton. Nerv. Syst. 67:51, 1997. 133. Schutte, I. W., Hollestein, K. B., Akkermans, L. M., and Kroese, A. B.: Evidence for a role of cholecystokinin as

276

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

Function of the Peripheral Nervous System neurotransmitter in the guinea-pig enteric nervous system. Neurosci. Lett. 236:155, 1997. Shen, K.-Z., North, R. A., and Surprenant, A.: Potassium channels opened by noradrenaline and other transmitters in excised membrane patches of guinea-pig submucosal neurones. J. Physiol. (Lond.) 445:581, 1992. Silverman, D. H. S., Munakata, J. A., Ennes, H., et al.: Regional cerebral activity in normal and pathological perception of visceral pain. Gastroenterology 112:64, 1997. Smith, G. P., Jerome, C., and Norgren, R.: Afferent axons in abdominal vagus mediate satiety effect of cholecystokinin in rats. Am. J. Physiol. 249:R638, 1985. Smith, V., Gregson, N., and Foggensteiner, L.: Acquired intestinal aganglionosis and circulating autoantibodies without neoplasia or other neural involvement. Gastroenterology 112:1366, 1997. Song, Z. M., Brookes, S. J. H., and Costa, M.: Identification of myenteric neurons which project to the mucosa of the guinea-pig small intestine. Neurosci Lett. 129:294, 1991. Song, Z. M., Brookes, S. J. H., and Costa, M.: All calbindinimmunoreactive myenteric neurons project to the mucosa of the guinea-pig small intestine. Neurosci. Lett. 180:219, 1994. Stanghellini, V., Camilleri, M., and Malagelada, J.-R.: Chronic idiopathic intestinal pseudo-obstruction: clinical and intestinal manometric findings. Gut 23:824, 1987. Starodub, A. M., and Wood, J. D.: A-type potassium current in myenteric neurons of guinea-pig small intestine. Neuroscience 99:389, 2000. Surprenant, A.: Synaptic transmission in neurons of the submucous plexus. In Nerves and the Gastrointestinal Tract (Falk Symposium 50). Boston, MTP Press, p. 253, 1989. Surprenant, A., and Crist, J.: Electrophysiological characterization of functionally distinct 5-hydroxytryptamine receptors on guinea-pig submucous plexus. Neuroscience 24:283, 1988. Surprenant, A., and North, R. A.: Mechanism of synaptic inhibition by noradrenaline acting at alpha 2-adrenoceptors. Proc. R. Soc. Lond. 234:85, 1988. Surprenant, A., and North, R. A.: Inhibitory receptors and signal transduction in submucosal neurones. In Holle, G. E., and Wood, J. D. (eds.): Advances in the Innervation of the Gastrointestinal Tract. Amsterdam, Elsevier Scientific Press, p. 239, 1992. Tack, J., Caenepeel, P., Fischler, B., et al.: Symptoms associated with hypersensitivity to gastric distention in functional dyspepsia. Gastroenterology 121:526, 2001. Tack, J., Janssens, J., Vantrappen, G., and Wood, J. D.: Actions of 5-hydroxytryptamine on myenteric neurons in the guinea-pig gastric antrum. Am. J. Physiol. 263:G838, 1992. Tack, J., Janssens, J., Vantrappen, G., et al.: Effect of erythromycin on gastric motility in controls and diabetic gastroparesis. Gastroenterology 103:72, 1992. Tack, J. D., and Wood, J. D.: Synaptic behavior in the myenteric plexus of the guinea-pig gastric antrum. J. Physiol. (Lond.) 445:389, 1992.

150. Takaki, M., and Nakayama, S.: Effects of mesenteric nerve stimulation on the electrical activity of myenteric neurons in the guinea pig ileum. Brain Res. 442:351, 1988. 151. Talley, N. J.: Review article: 5-Hydroxytryptamine agonists and antagonists in the modulation of gastrointestinal motility and sensation—clinical implications. Aliment. Pharmacol. Ther. 6:273, 1992. 152. Tamura, K., Palmer, J. M., and Wood, J. D.: Presynaptic inhibition produced by histamine at nicotinic synapses in enteric ganglia. Neuroscience 25:171, 1987. 153. Tamura, K., and Wood, J. D.: Electrical and synaptic properties of myenteric plexus neurones in the terminal large intestine of the guinea-pig. J. Physiol. (Lond.) 415:275, 1989. 154. Tatsumi, H., Hirai, K., and Katayama, Y.: Measurement of the intracellular calcium concentration in guinea-pig myenteric neurons by using fura-2. Brain Res. 451:371, 1988. 155. Thomas, E. A., Bertrand, P. P., and Bornstein, J. C.: Genesis and role of coordinated firing in a feed forward network: a model study of the enteric nervous system. Neuroscience 93:1525, 1999. 156. Thompson, W. G., Longstreth, G. L., Drossman, D. A., et al.: Functional bowel disorders and functional abdominal pain. In Drossman, D. A., Talley, N. J., Thompson, W. G., et al. (eds.): The Functional Gastrointestinal Disorders. McLean, VA, Degnon Associates, p. 351, 2000. 157. Timmermans, J.-P., Adriaensen, D., Cornelissen, W., and Scheuermann, D. W.: Structural organization and neuropeptide distribution in the mammalian enteric nervous system, with special attention to those components involved in mucosal reflexes. Comp Biochem. Physiol. 118:331, 1997. 158. Vantrappen, G., Peeters, T. L., Bloom, S. R., et al.: Motilin and the interdigestive migrating motor complex in man. Am. J. Dig. Dis. 24:497, 1979. 159. Vergara, P., Woskowska, Z., Cipris, S., et al.: Somatostatin excites canine ileum ex vivo: role for nitric oxide? Am. J. Physiol. 269:G12, 1995. 160. Verne, G., Sallustio, J., and Eaker, E.: Anti-myenteric neuronal antibodies in patients with achalasia: a prospective study. Dig. Dis. Sci. 42:307, 1997. 161. Vizi, V. A., and Vizi, E. S.: Direct evidence for acetylcholine releasing effect of serotonin in the Auerbach’s plexus. J. Neural Transm. 42:127, 1978. 162. Whitehead, W. E., Holtkotter, B., Enck, P., et al.: Tolerance for rectosigmoid distention in irritable bowel syndrome. Gastroenterology 98:1187, 1990. 163. Wood, J., Hudson, L., Jessel, T., and Yamamoto, M.: A monoclonal antibody defining antigenic determinants on subpopulations of mammalian neurones and Trypanosoma cruzi parasites. Nature 296:34, 1982. 164. Wood, J. D.: Excitation of intestinal muscle by atropine, tetrodotoxin and Xylocaine. Am. J. Physiol. 222:118, 1972. 165. Wood, J. D.: Electrical activity of the intestine of mice with hereditary megacolon and absence of myenteric ganglion cells. Am. J. Dig. Dis. 18:477, 1973. 166. Wood, J. D.: Neuronal interactions within ganglia of Auerbach’s plexus of the small intestine. In Bülbring, E., and Shuba, M. F. (eds.): Physiology of Smooth Muscle. New York, Raven Press, p. 321, 1976.

Neurobiology of the Enteric Nervous System 167. Wood, J. D.: Electrical and synaptic behavior of enteric neurons. In Wood, J. D. (ed.): Handbook of Physiology: The Gastrointestinal System, Motility and Circulation. New York, Oxford University Press, p. 465, 1989. 168. Wood, J. D.: Histamine signals in enteric neuroimmune interactions. Ann. N. Y. Acad. Sci. 664:275, 1992. 169. Wood, J. D.: Application of classification schemes to the enteric nervous system. J. Auton. Nerv. Syst. 48:179, 1994. 170. Wood, J. D.: Physiology of the enteric nervous system. In Johnson, L. R., Alpers, D. H., Christensen, J., et al. (eds.): Physiology of the Gastrointestinal Tract, 3rd ed. New York, Raven Press, p. 423, 1994. 171. Wood, J. D.: Physiological and pathophysiological paracrine functions of intestinal mast cells in enteric neuro-immune signaling. In Singer, M. F., Zigler, R., and Rohr, G. (eds.): Gastrointestinal Tract and Endocrine System. London, Kluwer Academic Publishers, p. 254, 1995. 172. Wood, J. D.: Neurotransmission at the interface of sympathetic and enteric divisions of the autonomic nervous system. Chin. J. Physiol. 42:201, 1999. 173. Wood, J. D.: Allergies and the brain-in-the-gut. Clin. Perspect. Gastroenterol. 3:343, 2000. 174. Wood, J. D.: Neuropathy in the brain-in-the-gut. Eur. J. Gastroenterol. Hepatol. 12:597, 2000. 175. Wood, J. D.: Neuropathophysiology of IBS. J. Clin. Gastroenterol. 35(Suppl.):11, 2002. 176. Wood, J. D., and Mayer, C. J.: Intracellular study of electrical activity of Auerbach’s plexus in guinea-pig small intestine. Pflugers Arch. 374:265, 1978. 177. Wood, J. D., and Mayer, C. J.: Adrenergic inhibition of serotonin release from neurons in guinea-pig Auerbach’s plexus. J. Neurophysiol. 42:594, 1979.

277

178. Wood, J. D., and Tack, J. F.: Motilin excites myenteric neurons in the gastric antrum of the guinea-pig. Gastroenterology 99:A1236, 1990. 179. Xia, Y., Baidan, L. V., Fertel, R. H., and Wood, J. D.: Determination of levels of cyclic AMP in the myenteric plexus of guinea-pig small intestine. Eur. J. Pharmacol. 206:231, 1991. 180. Xia, Y., Fertel, R., and Wood, J.: Stimulation of formation of adenosine 3⬘,5⬘- phosphate by histamine in myenteric ganglia isolated from guinea-pig small intestine. Eur. J. Pharmacol. 316:81, 1996. 181. Xia, Y., Fertel, R. H., and Wood, J. D.: Stimulation of formation of cAMP by 5-hydroxytryptamine in myenteric ganglia isolated from guinea pig small intestine. Life Sci. 55:685, 1994. 182. Xia, Y., Fertel, R. H., and Wood, J. D.: Suppression of cAMP formation by adenosine in myenteric ganglia of guinea-pig small intestine. Eur. J. Pharmacol. 32:95, 1997. 183. Yau, W. M., and Youther, M. L.: Direct evidence for a release of acetylcholine from myenteric plexus of guinea pig small intestine by substance P. Eur. J. Pharmacol. 81:665, 1982. 184. Zafirov, D. H., Cooke, H. J., and Wood, J. D.: Elevation of cAMP facilitates noradrenergic transmission in submucous neurons of guinea-pig ileum. Am. J. Physiol. 262:G442, 1993. 185. Zholos, A. V., Baidan, L. V., Starodub, A. M., and Wood, J. D.: Potassium channels of myenteric neurons in guinea-pig small intestine. Neuroscience 89:603, 1999. 186. Zholos, A. V., Baidan, L. V., and Wood, J. D.: Sodium conductance in cultured myenteric AH-type neurons from guinea-pig small intestine. Auton. Neurosci. 96:93, 2002.

13 Autonomic and Somatic Systems to the Anorectum and Pelvic Floor ADIL E. BHARUCHA AND CHRISTOPHER J. KLINGELE

Embryology Anatomy Pelvic Floor Rectum and Anal Canal Nerve Supply to the Pelvic Floor Sympathetic Nerves Parasympathetic Nerves Central Nervous System Pathways Regulating the Pelvic Floor Afferent Innervation

Anal Sphincter Tone and Reflexes Internal Anal Sphincter External Anal Sphincter Puborectalis Sacral Reflexes Pharmacology Mechanisms of Continence and Defecation Anorectal Dysfunction in Neurologic Disorders

The anorectum is the caudal end of the gastrointestinal tract, responsible for fecal continence and defecation. In humans, defecation is a viscerosomatic reflex that is often preceded by several attempts to preserve continence. Walter C. Bornemeier captured the intricacies of anorectal function in the following words: There is not a muscle or structure in the body that has a more keenly developed sense of alertness and ability to accommodate itself to varying situations . . . . Yet the sphincter ani can do it! The sphincter can apparently differentiate between solid, fluid and gas. It apparently can tell whether its owner is alone or with someone; whether standing up or sitting down, whether the owner is dressed or undressed. No other muscle in the body is such a protector of the dignity of man, yet so ready to come to his relief. A muscle like this is worth protecting. The anal canal is normally closed by an internal sphincter made of smooth muscle and an external sphincter made of tonically active striated muscle. The internal and external anal sphincters are supplied by autonomic and somatic nerves, respectively. Rectal smooth muscle is primarily regulated by the intrinsic or enteric nervous system while

Obstructed Defecation Fecal Incontinence Management of Fecal Incontinence Bowel Habit Modification Pharmacologic Approaches to Increase Sphincter Pressure Biofeedback Therapy Surgical Approaches

extrinsic (i.e., autonomic) nerves modulate the enteric nervous system. Within the gastrointestinal tract, extrinsic nerves have the most pronounced effects on portions lined by skeletal muscle (i.e., the upper esophageal sphincter and the external anal sphincter). Extrinsic nerves are critical for anorectal functions of storing and evacuating feces; these functions require coordination between visceral components (e.g., rectal contraction, rectal sensation, and internal sphincter relaxation) and somatic components (e.g., external sphincter contraction and relaxation). Consequently, extrinsic neural dysfunction often causes or contributes to defecatory disorders, including obstructed defecation and fecal incontinence.

EMBRYOLOGY The anorectal region in humans derives from four separate embryologic structures: the hindgut, cloaca, proctodeal pit, and anal tubercles.31 The rectum above the pubococcygeal line develops from the primitive hindgut, while the rectum below this line develops from the cloaca. The cloaca is initially a single tube that is subsequently separated by caudad migration of the urorectal septum into urogenital 279

280

Function of the Peripheral Nervous System

ANATOMY

uterus, and rectum. These defects are closed by connective tissue anterior to the urethra, anterior to the rectum (i.e., the perineal body), and posterior to the rectum (i.e., the postanal plate). The levator ani is subdivided into four muscles: pubococcygeus, ileococcygeus, coccygeus, and puborectalis. These muscles are attached peripherally to the pubic body, the ischial spine, and the arcus tendinus, a condensation of the obturator fascia in between these areas (Fig. 13–1). It is unclear whether the puborectalis should be regarded as a component of the levator ani complex or the external anal sphincter. Based on developmental evidence, innervation, and histologic studies, the puborectalis appears distinct from the majority of the levator ani.31 However, the puborectalis and external sphincter complex are innervated by separate nerves originating from S2-S4 (see below), suggesting phylogenetic differences between these two muscles.122

Pelvic Floor

Rectum and Anal Canal

The levator ani, or pelvic diaphragm, is a dome-shaped muscular sheet68 that predominantly contains striated muscle and has midline defects enclosing the bladder,

The rectum is 15 to 20 cm long, and extends from the rectosigmoid junction at the level of third sacral vertebra to the anal orifice (Fig. 13–2). The upper and lower rectum are

and intestinal passages. Mesenchymal elements from the lateral wall of the terminal cloaca fold inward, fusing with the urorectal septum above and the cloacal membrane below. Thereafter the cloacal membrane disintegrates, forming the upper anal canal, which is lined by ectoderm (i.e., columnar epithelium) and endoderm (i.e., squamous epithelium) in the upper and lower portions, respectively. The transition from ectoderm to exoderm may be at, or distal to, the pectinate line. The proctodeal portion of the cloacal membrane disintegrates to form the anal tubercles that join posteriorly and migrate ventrally to encircle a depression, the proctodeal pit. The anal tubercles join the urorectal septum and genital tubercles to form the perineal body, completing the separation between the rectum and urogenital tract.

Symphysis pubis

Deep dorsal vein of clitoris Puborectalis

Urethra Pubococcygeus

Vagina Illiococcygeus Rectum Coccygeus

Piriformis

Sacrum

FIGURE 13–1 Pelvic view of the levator ani demonstrating its four main components: puborectalis, pubococcygeus, iliococcygeus, and coccygeus.

281

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

Pelvi- Longitudinal Circular muscle Obturator rectal muscle layer space space internus

Para-vertebral sympathectic chain

L1 L2

IllioLevator cocygeus ani Pubococygeus

Transverse folds Rectal ampulla Anal columns

Ischiorectal fossa Gluteus maximus Semitendinosus

L3

L4

L5 Deep Superficial Skin Subcutaneous Sphincter Parts of ani internus sphincter ani externus

Anal sinuses

FIGURE 13–2 Diagram of a coronal section of the rectum, anal canal, and adjacent structures. The pelvic barrier includes the anal sphincters and pelvic floor muscles.

separated by a horizontal fold. The upper rectum is derived from the embryologic hindgut, generally contains feces, and can distend toward the peritoneal cavity. The lower part, derived from the cloaca, is confined by a tube of condensed extraperitoneal connective tissue, and is generally empty in normal subjects, except during defecation. In humans, there are fewer enteric ganglia in the rectum compared to the colon and very few ganglia in the anal sphincter.69,168

NERVE SUPPLY TO THE PELVIC FLOOR Sympathetic Nerves The anorectum and pelvic floor are supplied by sympathetic, parasympathetic, and somatic fibers. Sympathetic preganglionic fibers originate from the lowest thoracic ganglion in the paravertebral sympathetic chain. After traveling in the lowest of four (thoracic) splanchnic nerves, these preganglionic fibers join the third lumbar splanchnic nerve and branches from the aortic plexus to form the superior hypogastric plexus (Fig. 13–3). The superior hypogastric plexus is situated in front of the aortic bifurcation at the L4-S1 level. The alternative term for this plexus (i.e., presacral nerve) is misleading because it is seldom condensed, and does not resemble a single nerve. The superior hypogastric plexus provides branches to the uteric and ovarian (or testicular) plexus, and divides into right and left hypogastric nerves that descend, uniting with parasympathetic fibers in the pelvic splanchnic nerves to form corresponding inferior hypogastric plexuses. Branches from each hypogastric nerve supply the testicular or ovarian plexus, the ureteric

Pelvic splanchnic nerves S1

Aortic plexus

Lumbar splanchnic nerves Superior hypogastric plexus

Left hypogastric nerve Right hypogastric nerve

S2 S3 S4 S5

Nerve to levator ani Pudendal nerve Middle rectal plexus Inferior rectal nerve

Inferior hypogastric plexus Uterovaginal plexus Vesical plexus

Labial branches of Perineal nerve perineal nerve

FIGURE 13–3 Sympathetic, parasympathetic, and pudendal nerve supply to the anorectum.

plexus, and the plexus on the internal iliac artery and to the sigmoid colon. Each hypogastric nerve is also joined by the lowest lumbar splanchnic nerve, that from the last lumbar sympathetic ganglion. The inferior hypogastric plexus is located posterior to the urinary bladder, beside the rectum and prostate in men; in women this plexus is also adjacent to the uterus. The inferior hypogastric plexus contains numerous small ganglia with sympathetic fibers conveyed predominantly by the hypogastric nerves, and also by branches from the lumbar splanchnic nerves. The origin and distribution of parasympathetic fibers in the inferior hypogastric plexus are discussed below. The inferior hypogastric plexus gives rise to the middle rectal plexus, vesical plexus, prostatic plexus, and uterovaginal plexus. The nerve supply to the rectum and anal canal is derived from the superior, middle, and inferior rectal plexus. Parasympathetic fibers in the superior and middle rectal plexuses synapse with postganglionic neurons in the myenteric plexus in the rectal wall. The inferior rectal nerve is a branch of the pudendal nerve, conveying motor

282

Function of the Peripheral Nervous System

fibers to the external anal sphincter and sensory input from the lower anal canal.

Parasympathetic Nerves Preganglionic parasympathetic fibers originating from ventral rami of the second, third, and often fourth sacral nerves travel in the pelvic splanchnic nerves, synapsing with neurons in the inferior hypogastric plexus or more distally, in the walls of viscera supplied by the plexus. In addition, ascending fibers from the inferior hypogastric plexus travel via superior hypogastric and aortic plexuses to reach the inferior mesenteric plexus, ultimately innervating the descending and sigmoid colon. After entering the colon, these fibers form the ascending colonic nerves, traveling cephalad in the plane of the myenteric plexus to supply a variable portion of the left colon. Sacral parasympathetic pathways to the colon have excitatory and inhibitory components.58 Excitatory pathways play an important role in colonic propulsive activity, especially during defecation. Consequently, defecation may be affected after surgical section of pelvic nerves in humans.38,145 In other species (e.g., guinea pig), feces transport may be entirely organized by the enteric nervous system; spinal and supraspinal reflexes are also involved in the process.110 Inhibitory pathways allow colonic volume to adapt to its contents, and also mediate descending inhibition, which initiates colonic relaxation ahead of a fecal bolus.

Central Nervous System Pathways Regulating the Pelvic Floor Cortical mapping with transcranial magnetic stimulation suggests that rectal and anal responses are bilaterally represented on the superior motor cortex (Brodmann’s area 4).162 There are subtle differences in the degree of bilateral hemispheric representation between subjects. The external sphincter is innervated by motor neurons in Onuf’s nucleus, located in sacral segments, predominantly in the second sacral segment in cats.94,141 Though they supply striated muscles under voluntary control, these motor neurons are smaller than usual alpha motor neurons and resemble autonomic motor neurons72,140; however, the conduction velocity in pudendal nerve fibers is comparable to that of peripheral nerves.26,135 Unlike other somatic motor neurons in the spinal cord, these neurons are relatively spared in amyotrophic lateral sclerosis, but affected in Shy-Drager syndrome.25,159 Holstege and others have demonstrated that motor neurons in the dorsomedial and ventrolateral parts of Onuf’s nucleus supply the anal and urethral sphincters, respectively. A cellular bridge connects Onuf’s nucleus with the sacral intermediolateral cell column, which innervates the detrusor muscle of the bladder,72 facilitating coordination between the sphincter and detrusor muscles during micturition. Moreover, projections from

the pontine reticular formation project to Onuf’s nucleus and to sacral intermediolateral cell groups, facilitating coordination of processes involved in micturition.72 These dorsolateral pontine reticular formation neurons may also project to the nucleus retroambiguous (NRA) in the caudal medulla oblongata; the NRA is the only area projecting specifically to abdominal muscle motor neurons, and also projects to pelvic floor motor neurons.72 Stimulation of the caudal NRA induced simultaneous contraction of the abdominal and pelvic floor muscles or straining, perhaps explaining why pelvic floor sphincters contract during all maneuvers associated with raised intra-abdominal procedures with the exception of defecation.47 The paraventricular hypothalamic nucleus projects to virtually all autonomic motor neurons in the caudal brainstem and spinal cord, including Onuf’s nucleus and the sacral intermediolateral cell group.70 Diffuse projections from noradrenergic neurons in the locus ceruleus and the ventromedial medullary segmental field to Onuf’s nucleus also exist; their functions are unclear.71 Motor neurons in Onuf’s nucleus lack Renshaw inhibition94 and crossed disynaptic inhibition. In contrast to a joint, the sphincter is not composed of agonist and antagonist muscles, obviating the need for reciprocal inhibition. It is also conceivable that reduced inhibitory input may facilitate maintenance of resting anal sphincter tone.80 Somatic branches originating from Onuf’s nucleus travel in the pudendal nerve, the muscular branches, and the coccygeal plexus. The pudendal nerve branches into inferior rectal, perineal, and posterior scrotal nerves. The inferior rectal nerve supplies the lower part of the external anal sphincter, the anal canal lining, and the skin around the anus. The perineal nerve divides into posterior scrotal (or labial) branches and muscular branches. The posterior scrotal branches innervate the skin, while muscular branches are distributed to the transverse perinei, bulbospongiosus, ischiocavernosus, urethral sphincter, anterior part of the external anal sphincter, and levator ani. Motor fibers of the right and left pudendal nerves have overlapping distributions within the external anal sphincter. Sherrington observed that stimulation of the right pudendal nerve caused circumferential contraction of the external anal sphincter.150 Conversely, tonic external sphincter activity, sphincter inhibition during colonic distention, and the cutaneoanal reflex were not affected by sectioning either pudendal nerve.14

Afferent Innervation Colonic balloon distention in humans evokes a generally ill-defined sensation of gas, perhaps discomfort, and eventually pain at higher intensities. Rectal distention is perceived as a more localized sensation of rectal fullness, interpreted by the patient as a desire for motion or to pass wind. The anal canal responds to distention and to

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

innocuous mucosal proximodistal mechanical shearing stimuli.78 Sensory traffic is conveyed by unmyelinated small C fibers and larger A fibers. Although there are very few rectal mucosal nerve endings, the anal canal is lined by numerous free and organized nerve endings (i.e., Meissner’s corpuscles, Krause’s end bulbs, Golgi-Mazzoni bodies, and genital corpuscles), perhaps explaining why it is exquisitely sensitive to light touch, pain, and temperature. Animal models and clinicopathologic findings in humans suggest that pelvic nerves traveling to the sacral segments are more important for conveying non-noxious and noxious colonic sensations than lumbar colonic (sympathetic) nerves.36,57,58,79,112,113,137,171 There are more afferent neurons supplying the colon in the sacral, compared to lumbar, segments in the cat (7500 vs. 4500 neurons).5,108 However, the number of spinal visceral afferent neurons is relatively small, only 2.5% or less of the total number of spinal afferent neurons supplying skin and deep somatic structures.79 During colonic distention in cats, lumbar and sacral afferents can accurately encode intraluminal pressure ranging from 20 to 100 mm Hg by increasing discharge frequencies.77 Sacral afferents may be better suited for conveying afferent information than lumbar afferents, because they are more likely to lack resting activity and respond to pressure increments with a wider range of discharge frequency.3,15 Janig and Morrison identified three different classes of mechanosensitive visceral afferents in the cat colon. 79 Tonic afferents fired throughout colonic distention and accurately encoded the intensity of distention between 20 and 100 mm Hg. Phasic colonic afferents generally discharged at the onset, and occasionally at the cessation, of a distention stimulus. Tonic afferents were predominantly unmyelinated, slowly conducting C fibers, while most phasic afferents were faster conducting myelinated A fibers. The afferents innervating the anal canal responded to shearing stimuli, but not colonic or anal distention. Two different theories have been proposed to explain visceral pain perception. Proponents of the specificity theory suggested that pain was a distinct sensory modality, mediated by sequential activation of visceral nociceptors and central pain-specific neurons in the spinal dorsal horn.100 However, in the cat colon, Janig and Koltzenburg found no afferent fibers that were selectively activated by noxious stimuli, arguing against the specificity theory. The alternative hypothesis for pain perception—pattern or intensity theory—attributes pain perception to spatial and temporal patterns of impulses generated in nonspecific visceral afferent neurons.78 However, electrophysiologic studies of visceral afferent fibers in other organs, including the colon, have documented high-threshold visceral afferent fibers that only respond to noxious mechanical stimuli.100,147 Subsequently, Cervero and Janig reconciled these opposing concepts in a convergence model wherein input from low- and high-threshold mechanoreceptors

283

converges on spinothalamic and other ascending tract cells.24 Physiologic processes are generally accompanied by low-level activity, mediation of regulatory reflexes, and transmission of nonpainful sensations. High-intensity stimuli increase firing of low-threshold afferents and also activate high-threshold afferents, thereby activating nociceptive pathways and triggering pain.24 More than 90% of all unmyelinated pelvic afferents are silent, being activated by electrical stimulation, but not by even extreme noxious stimuli.79 Silent afferents can respond to chemical stimuli or tissue irritation, however, becoming responsive to even innocuous mechanical stimuli after sensitization.100 These neurophysiologic changes are detectable within minutes after tissue irritation and likely to persist for the duration of irritation. Putative mediators of sensitization include bradykinin, histamine, serotonin, catecholamines, prostanoids, hydrogen and potassium ions, adenosine, substance P, cytokines, and other neuropeptides. Upregulation of silent afferents by inflammation has been implicated to explain visceral hypersensitivity, as is observed in patients with functional gastrointestinal disorders after an acute diarrheal infection.60

ANAL SPHINCTER TONE AND REFLEXES Internal Anal Sphincter The internal anal sphincter is a thickened extension of the circular smooth muscle layer surrounding the colon and is generally considered responsible for maintaining approximately 70% of resting tone, ensuring that the anal canal is closed at rest.51,58 Internal sphincter tone is primarily attributable to smooth muscle activity; the extent to which extrinsic nerves contribute to resting anal sphincter tone is unclear. While resting anal sphincter tone is not affected by spinal transection between C6 and L1,36,49 high spinal anesthesia reduced resting tone, suggesting extrinsic nerves may contribute to resting sphincter tone.51,87 Moreover, resting sphincter tone declined by an average of 30% after bilateral hypogastric nerve sectioning and an average of 63% after lumbar colonic nerve sectioning in dogs,106 returning to baseline 1 month thereafter. Conversely, sympathetic stimulation evoked either internal anal sphincter relaxation,73,93,121 or contraction followed by relaxation.22 Gowers demonstrated that rectal insufflation with air induced anal relaxation; anal pressure returned to normal thereafter.59 The rectoanal inhibitory reflex (RAIR) is mediated by intrinsic nerves and absent in Hirschsprung’s disease. The extrinsic nerves are not essential for the reflex, since it is preserved in patients with cauda equina lesions or after spinal cord transection.36 However, extrinsic nerves may modulate the reflex, since relaxation is more pronounced and prolonged in children with sacral agenesis.107 The RAIR is

284

Function of the Peripheral Nervous System

probably mediated by nitric oxide; morphologic studies reveal an efferent descending nitrergic rectoanal pathway.153 Other nonadrenergic-noncholinergic neurotransmitters (i.e., vasoactive intestinal peptide and ATP) may also participate in the RAIR.12,116

External Anal Sphincter Though resting sphincter tone is predominantly attributed to the internal anal sphincter, studies under general anesthesia or after pudendal nerve block suggest the external anal sphincter generally accounts for approximately 25%, and up to 50%, of resting anal tone.9,33,40,50,144,170 When continence is threatened, the external sphincter contracts to augment anal tone, preserving continence. This “squeeze” response may be voluntary or induced by increased intra-abdominal pressure,47 or by merely moving a finger across the anal canal lining.161 Conversely, the external sphincter relaxes during defecation. The only other striated muscles that display resting activity are the puborectalis, external urethral sphincter, cricopharyngeus, and laryngeal abductors. Resting or tonic activity depends on monosynaptic reflex drive, perhaps explaining why resting anal sphincter tone is reduced, but voluntary contraction of the external sphincter is preserved, in tabes dorsalis.102 However, the presence of muscle spindles in the human external anal sphincter is debatable.16,39 In cats, the tonic reflex is maintained by afferents entering the spinal cord at S2, while the phasic contractile “guarding”

responses to perineal scratch and anal penetration disappear after bilateral section of S3 dorsal roots.13 Thus the afferent axons involved in tonic and phasic reflexes are different and penetrate the spinal cord at different levels. Moreover, afferent discharge in the pudendal nerve is directly related to muscular activity of the external sphincter.13 The fiber distribution also favors tonic activity; type I fibers predominate in the human anal sphincter,142 while that of cats and rabbits predominantly contains type II or fast-twitch muscle fibers.85 External sphincter fibers are circumferentially oriented and very small, separated by profuse connective tissue.142 Although the active lengthtension relationship in the cat external anal sphincter is similar to that in other mammalian skeletal muscles, the muscle generates passive tension even below the optimal length.86 Above optimal length, the passive length-tension relationship is considerably steeper compared to skeletal muscles. Krier and Adams speculated that higher passive stiffness might prevent the external sphincter from being stretched beyond optimal length for force generation during defecation.84

Puborectalis The tonically active puborectalis muscle maintains the resting anorectal angle. Moreover, puborectalis contraction during a sudden rise in abdominal pressure reduces the anorectal angle, preserving continence (Fig. 13–4). While cadaveric studies suggested the puborectalis was supplied

FIGURE 13–4 Magnetic resonance fluoroscopic images of the pelvis at rest (Left), during squeeze (middle), and during simulated defecation (right) in a subject. The asymptomatic rectum was filled with ultrasound gel. At rest, the pelvic floor was well supported; the anorectal angle was relatively obtuse (126 degrees). Pelvic floor contraction during the squeeze maneuver was accompanied by normal upward and anterior motion of the anorectal junction; the angle declined to 95 degrees. During rectal evacuation, the bladder base dropped by 2.5 cm below the pubococcygeal line; a 2.8-cm anterior rectocele emptied completely, and was probably not clinically significant; perineal descent (5 cm) was outside the normal range for evacuation proctography.

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

285

by the pudendal nerve, electrophysiologic stimulation studies in humans suggest this muscle is supplied, strictly ipsilaterally, by branches originating from the sacral plexus above the pelvic floor.122 This spinal reflex response may be mediated by stretch receptors in the levator ani.163 While no or minor incontinence results after surgical division of the internal or external anal sphincter, disruption of the puborectalis inevitably causes significant incontinence, underscoring the importance of this muscle in maintaining continence.104

and pudendal evoked response, can be recorded after electrical stimulation of the dorsal nerves of the glans penis or clitoris.163 Shafik has described several pelvic sphincter responses elicited by stimulating the urethra or anal canal.148,149 The role of these reflexes in normal physiology or as investigative tools is unclear.

Sacral Reflexes

Sympathetic nerves excite while parasympathetic nerves inhibit the sphincters, which is the reverse of their effects on nonsphincteric regions. The internal anal sphincter in humans and monkeys has a dense adrenergic innervation. Stimulation of adrenergic  and  receptors contracted and relaxed human internal sphincter strips, respectively.119 The internal anal sphincter is more sensitive to adrenergic compared to cholinergic agonists; cholinergic agonists either contracted or relaxed internal anal sphincter strips in humans.119 In the vervet monkey, muscarinic receptor stimulation contracted internal sphincter strips, but relaxed them after muscarinic receptors were blocked, probably via nonadrenergic-noncholinergic mechanisms.127 Nicotinic agonists also relaxed internal sphincter strips, probably via nonadrenergic-noncholinergic mechanisms. Table 13–1 summarizes the effects of neurotransmitters and pharmacologic agents that modulate internal anal sphincter tone.

The pelvic floor striated muscles contract reflexly in response to stimulation of perineal skin (i.e., a somatosomatic reflex) or anal mucosa (i.e., a viscerosomatic reflex). The cutaneoanal reflex is elicited by scratching or pricking the perianal skin and involves the pudendal nerves and S4 roots. Electrophysiologic recordings reveal both short latency and long latency responses; the latter corresponds to the visible anal sphincter contraction.160 The short latency response has a latency of 5 to 6 ms, which is too short for a spinal reflex response because the latency from the conus medullaris to the anal sphincter is approximately 7 ms.98 Thus the short latency response is probably attributable to direct stimulation of nerve fibers in or near the sphincter, causing an antidromic volley up to a branch point, followed by orthodromic conduction back to muscle. The longer, stimulus-dependent response is probably polysynaptic, and has an extremely variable latency, limiting its utility as a diagnostic tool.163 In humans, electrical activity of the internal anal sphincter increases during urinary bladder emptying, returning to normal after micturition.138 Vesical distention in cats also increased anal pressure.54 Conversely, the external anal sphincter relaxed during micturition in humans,143 cats,14 and dogs.146 The pudendoanal reflex, also known as the sacral evoked potential, sacral reflex, pudendal sexual reflex,

PHARMACOLOGY

MECHANISMS OF CONTINENCE AND DEFECATION In addition to the pelvic barrier, the rectal curvatures and transverse rectal folds are other anatomic factors impeding fecal evacuation; rectoanal sensation and rectal compliance

Table 13–1. Effects of Neurotransmitters on the Anal Sphincters Neurotransmitter/Pharmacologic Agent

Model

Effects

Adrenergic Phenylephrine

In vitro—human and monkey Healthy subjects and incontinent patients

Acetylcholine52,119 Nicotine52,119 Bethanechol Loperamide Others: glucagon, ketanserin, diltiazem, enkephalin, somatostatin121

In vitro—human and monkey In vitro—human and monkey Healthy human subjects Incontinent patients Healthy human subjects

: contraction; : relaxation q resting pressure in healthy subjects; did not reduce incontinence Either contraction, relaxation, or 4 Relaxation (upper); 4 (lower) p resting pressure q resting pressure p resting pressure

119

q  increased; p  decreased; 4  no change.

286

Function of the Peripheral Nervous System

also help maintain continence. Defecation can often be postponed until convenient because the rectum relaxes or accommodates to retain stool, and the external sphincter and/or puborectalis contract voluntarily; this “squeeze” response is critically dependent on normal rectoanal sensation. Stool is often transferred into the rectum by colonic high-amplitude propagated contractions, which often occur after awakening or meals.8 Denny-Brown and Robertson observed that rectal distention evoked rectal contraction and anal sphincter relaxation, facilitating evacuation.37 Rectal distention by stool is associated with several processes that serve to preserve continence or, if circumstances are appropriate, proceed to defecation. Rectal distention induces reflex relaxation of the internal anal sphincter. The anal sphincter may also relax independently of rectal distention, allowing anal epithelium to periodically “sample” and ascertain whether rectal contents are gas, liquid, or stool.103 Perception of rectal distention prompts voluntary contraction of the external sphincter if defecation is inconvenient, or relaxation if defecation is convenient.156 The voluntary components of defecation include assumption of an appropriate posture, abdominal contraction, and pelvic floor relaxation (see Fig. 13–4). The central nervous system plays a greater role in regulating anorectal sensimotor functions compared to other regions of the gastrointestinal tract. Garry observed that colonic stimulation in cats induced colonic contraction and anal relaxation even after destruction of the lumbosacral cord, concluding that the gut “seems not to have wholly surrendered its independence.”55 However, the elaborate somatic defecation response depends on centers above the lumbosacral cord, and probably craniad to the spinal cord itself.

ANORECTAL DYSFUNCTION IN NEUROLOGIC DISORDERS Common neurologic disorders associated with anorectal manifestations of pelvic floor dysfunction include stroke, autonomic neuropathies, multiple sclerosis, Parkinson’s disease, and some forms of multiple system atrophy. Parkinson’s disease and multiple sclerosis are associated with paradoxical puborectalis contraction, causing obstructed defecation. 30,99 Anorectal sensorimotor disturbances in diabetes mellitus may be symptomatic or asymptomatic.133 Onuf’s nucleus is relatively spared in diseases that affect the somatic motor system (i.e., motor neuron disease),83 but severely affected in diseases that involve visceromotor neurons (e.g., Shy-Drager syndrome,97 mannosidosis, Hurler’s syndrome,158 and Fabry’s disease157).

Obstructed Defecation Obstructed defecation is a disorder of function attributable to either an inadequate “pushing” force or an inability to overcome resistance to expulsion caused by inadequate relaxation of the external anal sphnicter (i.e., anismus) and/or puborectalis sling (i.e., puborectalis dyssynergia).10 In addition to excessive straining during defecation, other features of obstructed defecation include the need for digital removal of stool from the rectum, and a sense of incomplete evacuation after defecation.173 Most patients have “idiopathic” obstructed defecation. Multiple sclerosis and Parkinson’s disease are perhaps the most common neurologic diseases associated with obstructed defecation. The diagnosis is predominantly based on the history and anorectal examination. Anorectal inspection may disclose prominent external hemorrhoids or an anal fissure. Resting anal tone, as gauged by the resistance to inserting a finger in the anal canal, is increased in anismus. On palpation, the puborectalis should contract and generally relax when patients are asked to squeeze and expel the examining finger, respectively. Paradoxical puborectalis contraction during simulated defecation is suggestive, but not necessarily diagnostic, of obstructed defecation. The perineum normally descends by 1 to 3 cm during simulated defecation. Perineal descent is often reduced in anismus. Conversely, perineal descent may be increased in other patients (see Fig. 13–4). One or more tests may be necessary to confirm a clinical suspicion of obstructed defecation. Anal manometry may reveal a high anal resting pressure; though normal ranges are age, gender, and technique dependent, an average resting pressure greater than 100 mm Hg is probably abnormal and suggestive of anismus. Anal pressure may be unchanged, or paradoxically increase, instead of declining during simulated evacuation; the ratio of rectal to anal pressures during expulsion, termed the defecation index, can quantify this process.126 The importance of considering test results in the overall clinical context cannot be overemphasized; paradoxical contraction has been observed in up to 20% of asymptomatic subjects and may not always indicate anismus.166 The balloon expulsion test is more than 80% sensitive and more than 80% specific for identifying pelvic floor dysfunction.123,126 Subjects are asked to expel a rectal balloon, connected over a pulley to a series of weights (Fig. 13–5). Patients without pelvic floor dysfunction can expel the balloon with no or limited external rectal traction. Patients with pelvic floor dysfunction require additional external traction to facilitate expulsion of the rectal balloon. Notwithstanding technical limitations,4 evacuation proctography is often useful for confirming clinically suspected rectocele or pelvic floor dysfunction not confirmed by other diagnostic tests. More recently, rapid magnetic resonance imaging (MRI) sequences have been developed to visualize pelvic floor motion in

287

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

Balloon

Weight

FIGURE 13–5 Rectal balloon expulsion test. Patients are asked to expel a latex balloon filled with 50 mL warm water and connected over a pulley to a series of weights. Graded external traction is provided by adding weights as necessary to facilitate balloon expulsion.

real-time without radiation exposure, as discussed below.46 Finally, since colonic transit is often delayed in obstructed defecation, it is necessary to exclude obstructed defecation before making a primary diagnosis of slow-transit constipation in patients with delayed colonic transit (Fig. 13–6).126

Fecal Incontinence Fecal incontinence is a multifactorial symptom attributable to reduced stool consistency, and/or impairment in one or more factors responsible for maintaining continence (i.e., pelvic floor barrier, rectoanal sensation, and rectal

compliance).11 The most common predisposing factor for incontinence in women is obstetric injury to the pelvic floor and pudendal nerve,154 often compounded by diarrhea. Disorders of the peripheral nervous system causing incontinence include spinal cord lesions, lumbosacral nerve root lesions, and pudendal neuropathy. Lesions of the conus medullaris or cauda equina causing fecal incontinence are also associated with weakness and/or sensory disturbances in the lower extremities and/or paraspinal muscles. Lumbosacral plexopathy is a rare complication of metastatic cancer, pelvic radiotherapy,75 or pregnancy.81 Other neurologic disorders associated with fecal incontinence include dementia,109,111,134 multiple sclerosis,23,66,67 and multiple system atrophy. The evidence for denervation in “idiopathic” fecal incontinence was based on histologic changes, more pronounced in the external sphincter compared to the puborectalis.120 Subsequent studies demonstrated prolonged pudendal nerve latencies in constipated patients with excessive perineal descent,82 and even after 1 minute of acute straining in symptomatic patients.42 It has been suggested that, over the long term, prolonged straining during defecation causes perineal descent, thereby stretching and injuring the pudendal nerves; pudendal nerve injury may lead to denervation atrophy of pelvic floor muscles supplied by these nerves. Perhaps this explains why some patients with pelvic floor dysfunction may progress from constipation-predominant symptoms to excessive perineal descent, ultimately developing fecal incontinence.

2 1 hr

4 hr

24 hr

48 hr

1

3 0 4 5 Geometric Center

Normal Range

1.3

1.6

1.7

0.7-1.7

1.6 - 3.8

3.0 - 4.8

FIGURE 13–6 Scintigraphic assessment of colonic transit in a patient with obstructed defecation. Transit was assessed by a capsule containing 111In. The pH-sensitive methacrylate coating of the capsule dissolved in the alkaline pH of the terminal ileum, releasing the isotope in the cecum at time 0. Scans taken over the next 48 hours monitored isotope progression throughout the colon. At 24 and 48 hours, the isotope was predominantly distributed in the transverse colon, as reflected by a geometric center of 1.7, which corresponds to the mid-transverse colon. The geometric center is the weighted distribution of counts throughout the colon.

288

Function of the Peripheral Nervous System

Clinical Features The clinical assessment is critical for making an accurate diagnosis, establishing rapport with the patient, and formulating a logical strategy for diagnostic testing and treatment. Patients with chronic diarrhea, fecal urgency, constipation, prolapsed hemorrhoids, urinary incontinence, dementia, and diabetes mellitus must be asked whether they have fecal incontinence because they may be embarrassed to disclose the symptom.89 Staining, soiling, seepage, and leakage are terms used to reflect the nature and severity of incontinence. Soiling indicates more leakage than staining of underwear and can be qualified further to indicate severity (i.e., of underwear, outer clothing, or furnishing/bedding). Seepage refers to leakage of small amounts of stool. Severity can also be rated by scales that quantify amount and frequency of stool loss; some scales also incorporate number of pads used, severity of urgency, and the impact of fecal incontinence on coping mechanisms and/or lifestyle-behavioral changes.11 For incontinent patients, the correlation between symptom severity and quality of life may not be perfect. Physicians and patients generally agree that infrequent incontinence for flatus reflects mild, while frequent leakage of liquid and stool Rest Posterion mmHg 100 2.0

reflects severe, incontinence.131 In other areas, physicians and patients may disagree; patients attach greater significance to frequent leakage of gas compared to physicians, while physicians attach greater significance to leakage of solid stool. Consequently, it is vital to ascertain quality of life, reflected not only by items connected with coping, behavior, self-perception, and embarrassment,132 but also practical day-to-day limitations (e.g., the ability to socialize and get out of the house).19 Patients are affected even by the possibility and unpredictability of incontinence episodes;19 conversely, lifestyle or behavioral adjustments may also influence the type and frequency of incontinent episodes. Diagnostic Tests Anal Manometry. Anal manometry provides an overall functional assessment of the anal sphincters. Pressures are measured by withdrawing a catheter with perfused or solid-state transducers by the station pull-through method; resting and squeeze pressures are recorded at serial 1-cm intervals from the rectum to the anal verge (Fig. 13–7). Though anal manometry is technically undemanding and widely available, variations in catheter design, definitions,

Squeeze

Resting P

1

50 mmHg 100

Resting P

Right 2.0

2

50 100

Anterior

Resting P

2.0

3

50 mmHg

Left

Resting Pr

2.0

4

50 mmHg 50 0 mmHg 50

Right pos Resting P 4.0

5

Right Ante Resting P 4.0

6

0 mmHg 50

Left Anter Resting P 4.0

7 0 mmHg

Left Poste Resting P 4.0

8 0

10 second

30 second

FIGURE 13–7 Assessment of squeeze response by manometry. Tracing shows resting and squeeze pressures recorded by circumferentially oriented water-perfused sensors located 2 cm (sensors 1–4) and 4 cm (sensors 5–8) from the anal verge. The squeeze pressure was sustained for approximately 30 seconds in this healthy subject.

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

and methods to calculate average and maximum resting and squeeze pressures between centers should be borne in mind when interpreting pressures. Also, normal values are strongly influenced by technique, age, and gender; both resting and squeeze pressures decline with increasing age, even in asymptomatic subjects.76,101 Therefore, anal pressures should be compared to normal values obtained using the same technique. Average resting and squeeze pressures are generally low in incontinent patients. Other factors causing incontinence (e.g., disordered stool consistency, rectoanal sensation, and/or rectal compliance) should be considered, particularly in incontinent patients with normal sphincter pressures.152 Pudendal Nerve Latencies. Delayed pudendal nerve terminal motor latencies (PNTMLs) are regarded as surrogate markers of pudendal nerve injury, and used to ascertain if anal sphincter weakness is attributable to pudendal nerve injury, sphincter defect, or both.4 The pudendal nerve is stimulated by an electrode on a gloved index finger as the finger courses around the pelvic brim. Measuring the shortest latency between stimulus delivery and recording is critically dependent on close proximity of the examining finger to the nerve. Delayed pudendal nerve latencies have been extensively used to identify pudendal nerve injury caused by parturition,154 and perineal descent related to excessive straining,42 and prior to surgical repair of sphincter defects.96 However, the utility of PNTMLs as a marker for pudendal nerve injury is questionable since pudendal neuropathy is attributed to an axonopathy rather than demyelination. Because PNTML is dependent only on the fastest conducting fibers in the pudendal nerve, nerve latencies may be normal if even a few normally conducting fibers remain. There are several methodologic limitations: test reproducibility is unknown, normative data are limited, and multiple factors (i.e., age, parturition, body mass index) influencing nerve latencies are not generally taken into account. The “normal” ranges for nerve latencies are extremely stringent. Thus latencies that would be regarded as within normal limits for other peripheral nerve injuries are regarded as abnormal for the pudendal nerve. The sensitivity and specificity of the test are uncertain. In one study, approximately 50% of patients with prolonged PNTML had normal anal canal squeeze pressures.169 Finally, in contrast to initial studies, recent studies suggest the test does not predict improvement, or lack thereof, after surgical repair of anal sphincter defects.96 A recent consensus statement concluded that PNTML “cannot be recommended for evaluation of patients with fecal incontinence.”4 Anal Sphincter Electromyography. Endoanal ultrasound (US) has replaced anal sphincter electromyography (EMG) for identifying external anal sphincter defects. Given the limitations of PNTML, concentric needle anal sphincter

289

EMG may be useful for identifying myopathic (small polyphasic motor unit potentials), neurogenic (large polyphasic motor unit potentials), or mixed injury.29,35 The external anal sphincter is examined on each side with one or two needle insertions. The puborectalis muscle is examined by inserting a needle in the midline between the anus and tip of the coccyx, passing the needle through the external anal sphincter and into the deeper puborectalis. Insertional activity at rest and motor unit potential (MUP) amplitude, duration, percent polyphasia, and recruitment following mild to moderate voluntary muscle contraction are assessed. These parameters are measured in a standardized semiquantitative manner that has been demonstrated to have minimal intraobserver and interobserver variability.34,35 Assessment of MUP recruitment is also a reliable measure of muscle weakness in neurogenic lesions. Needle EMG of the puborectalis muscle can be used to distinguish disorders that affect this muscle and the external sphincter muscle selectively or in combination.6 In experienced hands, anal sphincter EMG is not accompanied by severe discomfort. EMG recording by an anal sponge or hard plug electrode correlates well with sphincter pressures, but cannot distinguish among the causes of sphincter weakness specified above. Podnar and colleagues quantified EMG activity in each of four sites in the subcutaneous and/or deeper portion of the external anal sphincter; the taped EMG signal was analyzed off-line.124 Because the level of continuous motor unit activity in the relaxed external sphincter was low, most parameters for the EMG interference pattern were zero. Therefore, the authors counted the number of MUPs, and estimated the lower limits of normal at each of four sites in the subcutaneous and deeper portions of the external anal sphincter. MUP counts were larger in men and in the subcutaneous external sphincter; MUP counts were not affected by age, parity, or fecal continence status.124 Assessment of Corticoanal Pathways. Corticoanal pathways to the external anal sphincter have been investigated by recording motor evoked potentials and pressure profiles from the anal canal in healthy subjects after cortical electrical or magnetic stimulation in research studies62,63; these assessments are not routinely used in clinical practice. Magnetic stimulation is less painful and therefore preferred to electrical stimulation. A more intense stimulus was required to elicit motor responses in the bulbocavernosus muscle and anal sphincter compared to the upper or lower extremities.117 Voluntary facilitation of the pelvic floor muscles reduced the latency for the electrical response in some,117 but not all, studies63; the latency for the pressure response to electrical stimulation was shorter during voluntary facilitation.63 Moreover, contraction times in the external sphincter were similar across a wide range of amplitudes during nonfacilitated,

290

Function of the Peripheral Nervous System

facilitated, and voluntary contractions, indicative of isochronism (i.e., the external sphincter adjusts the rate of rise in pressure during a contraction to attain a wide range of pressure amplitudes within a constant time).63 Lumbosacral and pudendal nerve stimulation facilitated cortical pathways to the external anal sphincter. Magnetic stimulation of the lumbosacral nerve roots or electrical stimulation of the pudendal nerve augmented the amplitude and shortened the latency of the anal sphincteric response to the cortical stimulation.61 Facilitation of sphincteric responses to cortical activation is probably attributable to direct antidromic excitation of sacral motor neurons innervating the external anal sphincter by lumbosacral or pudendal nerve stimulation.92 Moreover, lumbosacral, but not pudendal, nerve stimulation induced a second delayed rise in corticoanal response amplitude, which was maximal when the duration between lumbosacral nerve and cortical stimulation was 100 ms.61 Perhaps this late facilitation was attributable to cortical excitation by activation of ascending afferent pathways, indicative of sensorimotor, peripheral-central neural interactions. The relationship between stimulus intensity and response latency for the anal sphincter differs from that for the tibialis anterior.44 For the tibialis anterior, low-intensity cortical electrical stimulation induced a fast response with a latency of 30 ms, consistent with monosynaptic conduction, while a late response (latency  100 ms) occurred during stimulation at a higher intensity. Controversially, in the external anal sphincter, the late response occurred at a very low stimulus intensity, suggesting activation of highly excitable central motor neurons connected to polysynaptic pathways. The response latency was shorter at higher stimulus intensities;

only one response could be recorded at a time and no inhibition was observed. These observations suggest that, despite different latencies, all responses are transmitted by the same pathway; higher stimulus intensities probably circumvent some interneurons in the pathways. The indications, strengths, and limitations of other tests are detailed below and summarized in Table 13–2. Endoanal Ultrasound. Ultrasound can identify internal sphincter thinning or defects,7,164 and external sphincter scars or defects; these defects are often clinically unrecognized,154 and, in a small proportion of patients, external sphincter defects are amenable to surgical repair (Fig. 13–8).114 Interpretation of external sphincter images is more subjective, operator dependent, and confounded by normal anatomic variations in the external sphincter7,64; there is substantial interobserver variability.41,56 The etiology of internal sphincter thinning is unknown. Since the external sphincter is often asymmetrical in the upper anal canal, particularly in women, it may be difficult to distinguish a normal variant from a sphincter defect.7,74 Moreover, the significance of clinically occult isolated external anal sphincter defects to fecal incontinence is unclear, since these defects have been observed in 1 in 3 women after a vaginal delivery.154 Finally, the external sphincter and perirectal fat are both echogenic and frequently indistinguishable, precluding accurate characterization of external sphincter thickness and identification of external sphincter atrophy. Evacuation Proctography (Defecography). During evacuation proctography, anorectal anatomy and pelvic floor motion are recorded on video with the patient at rest,

Table 13–2. Investigation of Fecal Incontinence Test

Purpose/Common Findings

Limitations

Anorectal manometry

p resting and squeeze pressure reflect internal and external anal sphincter weakness, respectively p or q rectal sensation occur commonly in fecal incontinence and may be remediable by biofeedback therapy Internal and external sphincter defects or scars are most frequently attributable to obstetric or surgical trauma Useful for documenting rectocele, perineal descent, enterocele, and internal rectal intusussception

Technically undemanding, but methods vary between centers

Rectal sensation

Endoanal ultrasound Evacuation proctography (barium defecography) Pelvic MRI

Anal sphincter EMG Pudendal nerve terminal motor latency

• Static imaging visualizes morphology of anal sphincters and entire pelvic floor (i.e., bladder, uterus, and rectum) • Dynamic imaging of pelvic floor motion visualizes prolapse Identify neurogenic, myogenic, and mixed lesions affecting external anal sphincter Delayed nerve latencies reflect pudendal nerve injury

q  increased; p  decreased; EMG  electromyography; MRI  magnetic resonance imaging.

Variations in normal anatomy may be mistaken for an external sphincter defect • Measurements not standardized • Imperfect correlation between symptoms and findings Not widely available

Potentially painful Several methodologic limitations significantly restrict utility of measurements

291

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

FIGURE 13–8 Endoanal ultrasound (US) image of anal sphincters in a patient with fecal incontinence. The internal anal sphincter is hypoechoic. Thick and thin arrows indicate normal internal sphincter and tear, respectively (located approximately between 10 and 5 o’clock) on US images. Large and small arrowheads indicate normal-appearing and partially torn external sphincter (between 10 and 2 o’clock), respectively.

coughing, squeezing, and straining to expel barium from the rectum; the anorectal angle and position of the anorectal junction are tracked during these maneuvers. Prior to dynamic MRI, evacuation proctography was the only modality for identifying excessive perineal descent, internal rectal intussusception, rectoceles, sigmoidoceles or enteroceles; puborectalis dysfunction during squeeze and evacuation can also be characterized.1 However, the methods for conducting and interpreting evacuation proctography are incompletely standardized4; for incontinent patients, a thick barium paste (Anatrast; E-Z-EM, Westbury, NY) is probably preferable to liquid barium. Despite these limitations, evacuation proctography in incontinent patients may be useful, particularly prior to surgery, when there is a high index of suspicion for excessive perineal descent, a significant rectocele (e.g., 2 cm in size, with the need to splint the vagina to facilitate rectal emptying), enterocele, or internal rectal intussusception. These findings may also help educate patients about the nature of the disorder and reinforce the need for treatment (e.g., pelvic floor retraining). Pelvic MRI. With rapid MRI sequences, MRI can visualize both anal sphincter anatomy and global pelvic floor motion in real time, without radiation exposure.46 The anal sphincters are visualized by axial T2-weighted fast spin echo images and corresponding T1-weighted

spin echo images with a disposable endorectal colon coil.46 Disagreement exists over the best technique for the evaluation of the internal sphincter; however, MRI performed the same as or better than US for the assessment of the external sphincter.95,130 In contrast to US, MRI can also identify external sphincter atrophy.18 External sphincter atrophy is a good prognosticator for poor continence after repair of external sphincter defects.17 Dynamic images are acquired as patients squeeze their sphincters or try to expel US gel from the rectum, providing a unique appreciation of global pelvic floor motion—in addition to the anorectum, the bladder and genital organs are also visualized (see Fig. 13–4).46,129 As with any new diagnostic imaging modality, comparisons to age- and gender-matched asymptomatic subjects are necessary to assess the role of static and dynamic MRI as a diagnostic tool in clinical practice. Rectoanal Sensation. Rectal sensation is assessed by progressively distending a latex balloon manually, or a polyethylene balloon by a barostat, while measuring thresholds for first perception, desire to defecate, and severe discomfort (Fig. 13–9).172 Alternatively, perception is recorded by a visual analogue scale during phasic distentions of graded intensity by a barostat.88 Rectal sensation is reduced in diabetes, in multiple sclerosis, and after spinal cord injury23,66,67,133,155; within limits, rectal sensation can be modulated by biofeedback therapy, improving continence.105 Anal sensation is assessed by determining the perception threshold to an electrical

1

Manometric sensor Colonic ballon

2 Barostat 3

Rectal ballon To second barostat

6

5 Volume Pressure

FIGURE 13–9 Barostat-manometer assembly. Highly compliant polyethylene balloons positioned in the colon and/or rectum can be distended to assess motor activity, sensation, and compliance. Motor activity is generally recorded by a balloon inflated to a fixed pressure and opposed to the colonic mucosa; under these circumstances, reduced or increased colonic volume reflects colonic contraction or relaxation, respectively.

292

Function of the Peripheral Nervous System

stimulus, or temperature change in the anal canal. Anal sensitivity to temperature change is reduced in fecal incontinence.139 Rectal Compliance. Compliance is a measure of rectal distensibility, reflecting rectal ability to hold a larger volume of stool at a given pressure, thereby postponing defecation to a convenient time37; the rectum becomes less compliant with aging, perhaps predisposing to incontinence.48 Compliance is measured by assessing rectal pressure-volume relationships with a balloon, either by manually inflating a latex balloon with air or water or, more accurately, by distending a highly compliant polyethylene balloon with a barostat (see Fig. 13–9). Rectal compliance is reduced in ulcerative and radiation proctitis, perhaps contributing to fecal urgency and incontinence.20,32,125

MANAGEMENT OF FECAL INCONTINENCE The management must be tailored to clinical manifestations, and includes treatment of underlying diseases, and other approaches detailed in Table 13–3.

Bowel Habit Modification Modifying irregular bowel habits is the cornerstone to effectively managing incontinence, particularly when pelvic floor weakness is attributable to an irremediable neuromuscular disorder. Several simple approaches are underutilized because of a misperceived lack of benefit. Accurate characterization of bowel habits is necessary to tailor therapy. Diarrhea may be attributable to loss of the sympathetic inhibitory “brake,” or coexistent bacterial overgrowth in patients with an autonomic neuropathy;

Table 13–3. Management of Fecal Incontinence Intervention

Side Effects

Comments

Mechanism of Action

Skin irritation

Disposable products provide skin protection superior to nondisposable products Underpad products were slightly cheaper than body-worn products

Provide skin protection and prevent soiling of linen; polymers conduct moisture away from the skin

Constipation

Titrate dose for each patient; administer before meals and social events

q fecal consistency, purgency; q anal sphincter tone

151

Incontinence pads* Disposable body-worns, reusable body-worns, disposable underpads, and reusable underpads Disposable body-worns are the largest category Antidiarrheal agents* Loperamide (Imodium): up to 16 mg/day in divided doses Diphenoxylate: 5 mg qid Enemas†

Inconvenient; side effects of specific preparations

Biofeedback therapy using anal canal pressure or surface EMG sensors† 115 Rectal balloon for modulating sensation Sphincteroplasty for sphincter defects† 2 Sacral nerve stimulation† Artificial sphincter Gracilis transposition†

Wound infection Recurrent incontinence (delayed) Infection Device erosion, failure, and infection

Rectal evacuation decreases likelihood of fecal incontinence Prerequisites for success include motivation, intact cognition, absence of depression, and some rectal sensation

Improved rectal sensation and coordinated external sphincter contraction;  q anal sphincter tone

Restricted to isolated sphincter defects without denervation

Restore sphincter integrity

Preliminary uncontrolled trials promising Either artificial device or gracilis transposition with/without electrical stimulation

Unclear; q anal sphincter tone; may modulate rectal sensation Restore anal barrier

*Grade A and †grade B therapeutic recommendations are supported by at least one randomized controlled trial, or one high-quality study of nonrandomized cohorts. q  increased; p  reduced;   possible. Data from Bharucha, A. E., and Camilleri, M. H.: GI dysmotility and sphincter dysfunction. In Noseworthy, J. H. (ed.): Neurological Therapeutics: Principles and Practice. London, Martin Dunitz, p. 2255, 2003.

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

evaluation and management are covered in detail in Chapter 122. The sympathetic brake may be restored by clonidine, reducing diarrhea.45 In addition to reducing diarrhea, loperamide, given at an adequate dose (i.e., 2 to 4 mg 30 minutes before meals, up to 16 mg/day), slightly increased internal sphincter tone and reduced incontinence128; the importance of taking loperamide before meals and in adequate doses cannot be overemphasized. The dose must be titrated to reduce diarrhea but avoid constipation. By taking loperamide before social occasions, or meals outside the home, incontinent patients may avoid having an accident outside the home, and gain confidence in their ability to participate in social activities. Diphenoxylate is an alternative option for diarrhea118; the serotonin 5-HT3 antagonist alosetron (Lotronex), currently available under a restricted use program, may benefit patients with severe diarrhea whose symptoms do not respond to other agents. Patients with constipation, fecal impaction, and overflow incontinence may benefit from a regularized evacuation program, incorporating timed evacuation by digital stimulation and/or bisacodyl/glycerol suppositories, fiber supplementation, and selective use of oral laxatives as detailed in recent reviews.90,91

Pharmacologic Approaches to Increase Sphincter Pressure There are no proven approaches. Phenylephrine suppositories increased anal resting pressure by 33% in healthy subjects and incontinent patients.28 However, in a randomized, double-blind, placebo-controlled crossover study, phenylephrine did not significantly improve incontinence scores or resting anal pressure fecal incontinence.21

Biofeedback Therapy Biofeedback therapy is based on the principle of operant conditioning.43 Using a rectal balloon–anal manometry device, patients are taught to contract the external anal sphincter when they perceive balloon distention; perception may be reinforced by visual tracings of balloon volume and anal pressure, and the procedure is repeated with progressively smaller volumes. Several uncontrolled studies suggest continence improves in approximately 70% of patients65; controlled studies are in progress. Sensory assessments (i.e., preserved baseline sensation and improved sensory discrimination after biofeedback therapy) are more likely to be associated with improved continence after biofeedback therapy than are sphincter pressures.105,167

293

Surgical Approaches A colostomy is the last resort for patients with severe incontinence.27 The results of overlapping anterior sphincteroplasty for anal sphincter defects are disappointing96; the procedure may be particularly inadvisable when fecal incontinence is attributable to neuromuscular weakness. The more complicated surgical procedures (e.g., dynamic graciloplasty and artificial anal sphincter) are fraught with significant morbidity, precluding widespread use.27 Sacral nerve stimulation is approved by the Food and Drug Administration for treating urge urinary incontinence in the United States. Sacral nerve stimulation may also augment anal pressures, modulate rectal sensation, and significantly improved continence.53,136,165 Sacral stimulation is conducted as a staged procedure; patients whose symptoms respond to temporary stimulation over approximately 2 weeks proceed to permanent subcutaneous implantation of the device. The procedure for device placement is technically straightforward, and device-related complications are less frequent or significant relative to more invasive artificial sphincter devices discussed above. Further studies of safety and efficacy are awaited.

REFERENCES 1. Agachan, F., Pfeifer, J., and Wexner, S. D.: Defecography and proctography: results of 744 patients. Dis. Colon Rectum 39:899, 1996. 2. Bachoo, P., Brazzelli, M., and Grant, A.: Surgery for faecal incontinence in adults. Cochrane Database Syst. Rev. 4: CD001757, 2000. 3. Bahns, E., Halsband, U., and Janig, W.: Responses of sacral visceral afferents from the lower urinary tract, colon and anus to mechanical stimulation. Pflugers Arch. 410:296, 1987. 4. Barnett, J. L., Hasler, W. L., and Camilleri, M.: American Gastroenterological Association medical position statement on anorectal testing techniques. Gastroenterology 116:732, 1999. 5. Baron, R., Janig, W., and McLachlan, E. M.: The afferent and sympathetic components of the lumbar spinal outflow to the colon and pelvic organs in the cat. III. The colonic nerves, incorporating an analysis of all components of the lumbar prevertebral outflow. J. Comp. Neurol. 238:158, 1985. 6. Bartolo, D. C., Jarratt, J. A., Read, M. G., et al.: The role of partial denervation of the puborectalis in idiopathic faecal incontinence. Br. J. Surg. 70:664, 1983. 7. Bartram, C. I., and Sultan, A. H.: Anal endosonography in faecal incontinence. Gut 37:4, 1995. 8. Bassotti, G., Crowell, M. D., and Whitehead, W. E.: Contractile activity of the human colon: lessons from 24 hour studies. Gut 34:129, 1993. 9. Bennett, R. C., and Duthie, H. L.: The functional importance of the internal sphincter. Br. J. Surg. 51:355, 1964.

294

Function of the Peripheral Nervous System

10. Bharucha, A. E.: Obstructed defecation: don’t strain in vain! [comment]. Am. J. Gastroenterol. 93:1019, 1998. 11. Bharucha, A. E.: Fecal incontinence. Gastroenterology 124:1672, 2003. 12. Biancani, P., Walsh, J., and Behar, J.: Vasoactive intestinal peptide: a neurotransmitter for relaxation of the rabbit internal anal sphincter. Gastroenterology 89:867, 1985. 13. Bishop, B.: Reflex activity of the external anal sphincter of cat. J. Neurophysiol. 22:679, 1959. 14. Bishop, B., Garry, R., Roberts, T., and Todd, J.: Control of the external sphincter of the anus in the cat. J. Physiol. (Lond.) 134:229, 1956. 15. Blumberg, H., Haupt, P., Janig, W., and Kohler, W.: Encoding of visceral noxious stimuli in the discharge patterns of visceral afferent fibres from the colon. Pflugers Arch. 398:33, 1983. 16. Borghi, F., DiMolfetta, L., Garavoglia, M., et al.: Questions about the uncertain presence of muscle spindles in the human external anal sphincter. Panminerva Med. 33:170, 1991. 17. Briel, J. W., Stoker, J., Rociu, E., et al.: External anal sphincter atrophy on endoanal magnetic resonance imaging adversely affects continence after sphincteroplasty. Br. J. Surg. 86:1322, 1999. 18. Briel, J. W., Zimmerman, D. D., Stoker, J., et al.: Relationship between sphincter morphology on endoanal MRI and histopathological aspects of the external anal sphincter. Int. J. Colorectal Dis. 15:87, 2000. 19. Byrne, C. M., Pager, C., Rex, J., et al.: Assessment of quality of life in the treatment of patients with neuropathic fecal incontinence. Dis. Colon Rectum 45:1431, 2002. 20. Camilleri, M., Thompson, W. G., Fleshman, J. W., and Pemberton, J. H.: Clinical management of intractable constipation. Ann. Intern. Med. 121:520, 1994. 21. Carapeti, E. A., Kamm, M. A., and Phillips, R. K.: Randomized controlled trial of topical phenylephrine in the treatment of faecal incontinence. Br. J. Surg. 87:38, 2000. 22. Carlstedt, A., Nordgren, S., Fasth, S., et al.: Sympathetic nervous influence on the internal anal sphincter and rectum in man [comment]. Int. J. Colorectal Dis. 3:90, 1988. 23. Caruana, B. J., Wald, A., Hinds, J. P., and Eidelman, B. H.: Anorectal sensory and motor function in neurogenic fecal incontinence: comparison between multiple sclerosis and diabetes mellitus. Gastroenterology 100:465, 1991. 24. Cervero, F., and Janig, W.: Visceral nociceptors: a new world order? [comment]. Trends Neurosci. 15:374, 1992. 25. Chalmers, D., and Swash, M.: Selective vulnerability of urinary Onuf motoneurons in Shy-Drager syndrome. J. Neurol. 234:259, 1987. 26. Chantraine, A., De Leval, J., and Onkelinx, A.: Motor conduction velocity in the internal pudendal nerves. In Desmedt, J. E. (ed.): New Developments in Electromyography and Clinical Neurophysiology. Basel, Karger, p. 433, 1973. 27. Chapman, A. E., Geerdes, B., Hewett, P., et al.: Systematic review of dynamic graciloplasty in the treatment of faecal incontinence. Br. J. Surg. 89:138, 2002. 28. Cheetham, M. J., Kamm, M. A., and Phillips, R. K.: Topical phenylephrine increases anal canal resting pressure in patients with faecal incontinence. Gut 48:356, 2001. 29. Cheong, D. M., Vaccaro, C. A., Salanga, V. D., et al.: Electrodiagnostic evaluation of fecal incontinence. Muscle Nerve 18:612, 1995.

30. Chia, Y. W., Gill, K. P., Jameson, J. S., et al.: Paradoxical puborectalis contraction is a feature of constipation in patients with multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 60:31, 1996. 31. Cook, T. A., Mortensen, N. J.: Colon, rectum, anus, anal sphincters and the pelvic floor. In Pemberton, J. H., Swash, M., and Henry, M. M. (eds.): The Pelvic Floor: Its Function and Disorders. London, Harcourt Publishers, p. 61, 2002. 32. Corestti, M., Bhoori, S., and Basilisco, G.: Perceptual sensitivity and response bias during rectal distention in patients with irritable bowel syndrome. Neurogastroenterol. Motil. 11:255, 1999. 33. Culver, P. J., and Rattan, S.: Genesis of anal canal pressures in the opossum. Am. J. Physiol. 251(6 Pt. 1):G765, 1986. 34. Daube, J. R.: The description of motor unit potentials in electomyography. Neurology 28:623, 1978. 35. Daube, J. R.: Assessing the motor unit with needle electromyography. In Daube, J. (ed.): Clinical Neurophysiology (Contemporary Neurology Series). Philadelphia, F. A. Davis, p. 257, 1996. 36. Denny-Brown, D., and Robertson, E.: An investigation of the nervous control of defecation. Brain 58:256, 1935. 37. Devroede, G.: Functions of the anorectum: defecation and continence. In Phillips, S., Pemberton, J., and Shorter, R. (eds.): The Large Intestine: Physiology and Disease. New York, Raven Press, p. 115, 1991. 38. Devroede, G., and Lamarche, J.: Functional importance of extrinsic parasympathetic innervation to the distal colon and rectum in man. Gastroenterology 66:273, 1974. 39. Dubrovsky, B., and Filipini, D.: Neurobiological aspects of the pelvic floor muscles involved in defecation. Neurosci. Biobehav. Rev. 14:157, 1990. 40. Duthie, H. L., and Watts, J. M.: Contribution of the external anal sphincter to the pressure zone in the anal canal. Gut 6:64, 1965. 41. Enck, P., Heyer, T., Gantke, B., et al.: How reproducible are measures of the anal sphincter muscle diameter by endoanal ultrasound? Am. J. Gastroenterol. 92:293, 1997. 42. Engel, A. F., and Kamm, M. A.: The acute effect of straining on pelvic floor neurological function. Int. J. Colorectal Dis. 9:8, 1994. 43. Engel, B. T., Nikoomanesh, P., and Schuster, M. M.: Operant conditioning of rectosphincteric responses in the treatment of fecal incontinence. N. Engl. J. Med. 290:646, 1974. 44. Ertekin, C., Hansen, M. V., Larsson, L. E., and Sjodahl, R.: Examination of the descending pathway to the external anal sphincter and pelvic floor muscles by transcranial cortical stimulation. Electroencephalogr. Clin. Neurophysiol. 75:500, 1990. 45. Fedorak, R. N., Field, M., and Chang, E. B.: Treatment of diabetic diarrhea with clonidine. Ann. Intern. Med. 102:197, 1985. 46. Fletcher, J. G., Busse, R. F., Riederer, S. J., et al.: Magnetic resonance imaging of the anatomic and dynamic defects of the pelvic floor in defecatory disorders. Am. J. Gastroenterol. 98:399, 2003. 47. Floyd, W., and Walls, E.: Electromyography of the sphincter ani externus in man. J. Physiol. (Lond.) 122:599, 1953. 48. Fox, J. C., Rath-Harvey, D., Helwig, P. S., et al.: Anal sphincter pressures and rectal compliance decline with aging in

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor

49. 50. 51. 52.

53.

54.

55. 56.

57.

58.

59. 60.

61.

62.

63.

64.

65.

66.

67.

68.

asymptomatic women. Gastroenterology 122(Suppl.): A-69, 2002. Frenckner, B.: Function of the anal sphincters in spinal man. Gut 16:638, 1975. Frenckner, B., and Euler, C. V.: Influence of pudendal block on the function of the anal sphincters. Gut 16:482, 1975. Frenckner, B., and Ihre, T.: Influence of autonomic nerves on the internal and sphincter in man. Gut 17:306, 1976. Friedmann, C. A.: The action of nicotine and catecholamines on the human internal anal sphincter. Am. J. Dig. Dis. 13:428, 1968. Ganio, E., Ratto, C., Masin, A., et al.: Neuromodulation for fecal incontinence: outcome in 16 patients with definitive implant. The initial Italian Sacral Neurostimulation Group (GINS) experience. Dis Colon Rectum 44:965, 2001. Garrett, J. R., Howard, E. R., and Jones, W.: The internal anal sphincter in the cat: a study of nervous mechanisms affecting tone and reflex activity. J. Physiol. (Lond.) 243:153, 1974. Garry, R.: The responses to stimulation of the caudal end of the large bowel in the cat. J. Physiol. (Lond.) 78:208, 1933. Gold, D. M., Halligan, S., Kmiot, W. A., and Bartram, C. I.: Intraobserver and interobserver agreement in anal endosonography. Br. J. Surg. 86:371, 1999. Goligher, J. C., and Hughes, E. S. R.: Sensibility of rectum and colon: its role in the mechanism of anal continence. Lancet 1:543, 1951. Gonella, J., Bouvier, M., and Blanquet, F.: Extrinsic nervous control of motility of small and large intestines and related sphincters. Physiol. Rev. 67:902, 1987. Gowers, W.: The automatic action of the sphincter ani. Proc. R. Soc. Lond. 26:77, 1877. Gwee, K. A., Leong, Y. L., Graham, C., et al.: The role of psychological and biological factors in postinfective gut dysfunction. [comment]. Gut 44:400, 44. Hamdy, S., Enck, P., Aziz, Q., et al.: Spinal and pudendal nerve modulation of human corticoanal motor pathways. Am. J. Physiol. Gastrointest. Liver Physiol. 274(2 Pt. 1): G419, 1998. Herdmann, J., Bielefeldt, K., and Enck, P.: Quantification of motor pathways to the pelvic floor in humans. Am. J. Physiol. Gastrointest. Liver Physiol. 260(5 Pt. 1):G720, 1991. Herdmann, J., Enck, P., Zacchi-Deutschbein, P., and Ostermann, U.: Speed and pressure characteristics of external anal sphincter contractions. Am. J. Physiol. Gastrointest. Liver Physiol. 269(2 Pt. 1):G225, 1995. Heyer, T., Enck, P., Gantke, B., et al.: Anal endosonography: are morphometric measurements of the anal sphincter reproducible? Gastroenterology 108:A613, 1995. Heymen, S., Jones, K. R., Ringel, Y., et al.: Biofeedback treatment of fecal incontinence: a critical review. Dis. Colon Rectum 44:728, Hinds, J. P., Eidelman, B. H., and Wald, A.: Prevalence of bowel dysfunction in multiple sclerosis: a population survey, Gastroenterology 98:1538, 1990. Hinds, J. P., and Wald, A.: Colonic and anorectal dysfunction associated with multiple sclerosis. Am. J. Gastroenterol. 84:587, 1989. Hjartardottir, S., Nilsson, J., Petersen, C., and Lingman, G.: The female pelvic floor: a dome—not a basin. Acta Obstet. Gynecol. Scand. 76:567, 1997.

295

69. Hoffman, S., and Orestano, F.: Histology of the myenteric plexus in relation to rectal biopsy in congenital megacolon. J. Pediatr. Surg. 2:575, 1967. 70. Holstege, G.: Some anatomical observations on the projections from the hypothalamus to brainstem and spinal cord: an HRP and autoradiographic tracing study in the cat. J. Comp. Neurol. 260:98, 1987. 71. Holstege, G., Kuypers, H. G., and Boer, R. C.: Anatomical evidence for direct brain stem projections to the somatic motoneuronal cell groups and autonomic preganglionic cell groups in cat spinal cord. Brain Res. 171:329, 1979. 72. Holstege, G., and Tan, J.: Supraspinal control of motoneurons innervating the striated muscles of the pelvic floor including urethral and anal sphincters in the cat. Brain 110(Pt.5):1323, 1987. 73. Horgan, P. G., O’Connell, P. R., Shinkwin, C. A., and Kirwan, W. O.: Effect of anterior resection on anal sphincter function. Br. J. Surg. 76:783, 1989. 74. Hussain, S. M., Stoker, J., and Lameris, J. S.: Anal sphincter complex: endoanal MR imaging of normal anatomy. Radiology 197:671, 1995. 75. Iglicki, F., Coffin, B., Ille, O., et al.: Fecal incontinence after pelvic radiotherapy: evidences for a lumbosacral plexopathy. Report of a case. Dis. Colon Rectum 39:465, 1996. 76. Jameson, J. S., Chia, Y. W., Kamm, M. A. et al.: Effect of age, sex and parity on anorectal function. Br. J. Surg. 81:1689, 1994. 77. Janig, W., and Koltzenburg, M.: On the function of spinal primary afferent fibres supplying colon and urinary bladder. J. Auton. Nerv. Syst. 30(Suppl.):S89, 1990. 78. Janig, W., and Koltzenburg, M.: Receptive properties of sacral primary afferent neurons supplying the colon. J. Neurophysiol. 65:1067, 1991. 79. Janig, W., and Morrison, J. F. B.: Functional properties of spinal visceral afferents supplying abdominal and pelvic organs, with special emphasis on visceral nociception. Prog. Brain Res. 67:87, 1986. 80. Jankowska, E., Padel, Y., and Zarzecki, P.: Crossed disynaptic inhibition of sacral motoneurones. J. Physiol. (Lond.) 285:425, 1978. 81. Katirji, B., Wilbourn, A. J., Scarberry, S. L., and Preston, D. C.: Intrapartum maternal lumbosacral plexopathy. Muscle Nerve 26:340, 2002. 82. Kiff, E. S., Barnes, P. R., and Swash, M.: Evidence of pudendal neuropathy in patients with perineal descent and chronic straining at stool. Gut 25:1279, 1984. 83. Kihira, T., Yoshida, S., Yoshimasu, F., et al.: Involvement of Onuf’s nucleus in amyotrophic lateral sclerosis. J. Neurol. Sci. 147:81, 1997. 84. Krier, J., and Adams, T.: Autonomic and somatic systems to the anorectum and pelvic floor. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 187, 1993. 85. Krier, J., Adams, T., and Meyer, R. A.: Physiological, morphological, and histochemical properties of cat external anal sphincter. Am. J. Physiol. Gastrointest. Liver Physiol. 255 (6 Pt. 1):G772, 1988. 86. Krier, J., Meyer, R. A., and Percy, W. H.: Length-tension relationship of striated muscle of cat external anal sphincter. Am. J. Physiol. Gastrointest. Liver Physiol. 256 (4 Pt. 1):G773, 1989.

296

Function of the Peripheral Nervous System

87. Langley, J. N., and Anderson, H. K.: On the innervation of the pelvic and adjoining viscera. Part 1. The lower portion of the intestine. J. Physiol. (Lond) 18:67, 1895. 88. Law, N.-M., Bharucha, A. E., Undale, A. S., and Zinsmeister, A. R.: Cholinergic stimulation enhances colonic motor activity, transit and sensation in humans. Am. J. Physiol. 281:G1228, 2001. 89. Leigh, R. J., and Turnberg, L. A.: Faecal incontinence: the unvoiced symptom. Lancet 1:1349, 1982. 90. Locke, G. R. 3rd, Pemberton, J. H., and Phillips, S. F.: AGA technical review on constipation. American Gastroenterological Association. Gastroenterology 119: 1766, 2000. 91. Locke, G. R. 3rd, Pemberton, J. H., and Phillips, S. F.: American Gastroenterological Association Medical Position Statement: guidelines on constipation. Gastroenterology 119:1761, 2000. 92. Loening-Baucke, V., Read, N. W., Yamada, T., and Barker, A. T.: Evaluation of the motor and sensory components of the pudendal nerve. Electroencephalogr. Clin. Neurophysiol. 93:35, 1994. 93. Lubowski, D. Z., Nicholls, R. J., Swash, M., and Jordan, M. J.: Neural control of internal anal sphincter function. Br. J. Surg. 74:668, 94. Mackel, R.: Segmental and descending control of the external urethral and anal sphincters in the cat. J. Physiol. (Lond.) 294: 105, 1979. 95. Malouf, A. J., Halligan, S., Williams, A. B., et al.: Prospective assessment of interobserver agreement for endoanal MRI in fecal incontinence. Abdom. Imaging 26:76, 2001. 96. Malouf, A. J., Norton, C. S., Engel, A. F., et al.: Long-term results of overlapping anterior anal-sphincter repair for obstetric trauma. Lancet 355:260, 2000. 97. Mannen, T., Iwata, M., Toyokura, Y., and Nagashima, K.: The Onuf’s nucleus and the external anal sphincter muscles in amyotrophic lateral sclerosis and Shy-Drager syndrome. Acta Neuropathol. 58:255, 1982. 98. Marsden, C. D., Merton, P. A., and Morton, H. B.: The latency of the anal reflex. J. Neurol. Neurosurg. Psychiatry 45:857, 1982. 99. Mathers, S. E., Kempster, P. A., Law, P. J., et al.: Anal sphincter dysfunction in Parkinson’s disease. Arch. Neurol. 46:1061, 1989. 100. Mayer, E. A., and Gebhart, G. F.: Basic and clinical aspects of visceral hyperalgesia [comment]. Gastroenterology 107:271, 1994. 101. McHugh, S. M., and Diamant, N. E.: Effect of age, gender, and parity on anal canal pressures: contribution of impaired anal sphincter function to fecal incontinence. Dig. Dis. Sci. 32:726, 1987. 102. Melzak, J., and Porter, N. H.: Studies of the reflex activity of the external sphincter ani in spinal man. Paraplegia 1:277, 103. Miller, R., Bartolo, D. C., Cervero, F., and Mortensen, N. J.: Anorectal sampling: a comparison of normal and incontinent patients. Br. J. Surg. 75:44, 1988. 104. Milligan, E. T. C., and Morgan, C. N.: Surgical anatomy of the anal canal with special reference to anorectal fistulae. Lancet 2:1150, 1934.

105. Miner, P. B., Donnelly, T. C., and Read, N. W.: Investigation of mode of action of biofeedback in treatment of fecal incontinence. Dig. Dis. Sci. 35:1291, 1990. 106. Mizutani, M. Neya, T., Ono, K., et al.: Histochemical study of the lumbar colonic nerve supply to the internal anal sphincter and its physiological role in dogs. Brain Res. 598: 45, 1992. 107. Morera, C., and Nurko, S.: Rectal manometry in patients with isolated sacral agenesis. J. Pediatr. Gastroenterol. Nutr. 37:47, 2003. 108. Morgan, C., Nadelhaft, I., and de Groat, W. C.: The distribution of visceral primary afferents from the pelvic nerve to Lissauer’s tract and the spinal gray matter and its relationship to the sacral parasympathetic nucleus. J. Comp. Neurol. 201: 415, 1981. 109. Nakanishi, N., Tatara, K., Naramura, H., et al.: Urinary and fecal incontinence in a community-residing older population in Japan. J. Am. Geriatr. Soc. 45:215, 1997. 110. Nakayama, S., Neya, T., Yamasato, T., et al.: Activity of the spinal defaecation centre in the guinea pig. Ital. J. Gastroenterol. 11:168, 1979. 111. Nelson, R., Furner, S., and Jesudason, V.: Fecal incontinence in Wisconsin nursing homes: prevalence and associations. Dis. Colon Rectum 41:1226, 1998. 112. Ness, T. J., and Gebhart, G. F.: Colorectal distension as a noxious visceral stimulus: physiologic and pharmacologic characterization of pseudaffective reflexes in the rat. Brain Res. 450:153, 1988. 113. Ness, T. J., and Gebhart, G. F.: Visceral pain: a review of experimental studies. Pain 41:167, 1990. 114. Nielsen, M. B., Hauge, C., Pedersen, J. F., and Christiansen, J.: Endosonographic evaluation of patients with anal incontinence: findings and influence on surgical management. AJR Am. J. Roentgenol. 160:771, 1993. 115. Norton, C., Hosker, G., and Brazzelli, M.: Biofeedback and/or sphincter exercises for the treatment of faecal incontinence in adults. Cochrane Database Syst. Rev. 4: CD002111, 2003. 116. Nurko, S., and Rattan, S.: Role of vasoactive intestinal polypeptide in the internal anal sphincter relaxation of the opossum. J. Clin. Invest. 81:1146, 1988. 117. Opsomer, R. J., Caramia, M. D., Zarola, F., et al.: Neurophysiological evaluation of central-peripheral sensory and motor pudendal fibres. Electroencephalogr. Clin. Neurophysiol. 74:260, 1989. 118. Palmer, K. R., Corbett, C. L., and Holdsworth, C. D.: Double-blind cross-over study comparing loperamide, codeine and diphenoxylate in the treatment of chronic diarrhea. Gastroenterology 79:1272, 1980. 119. Parks, A. G., Fishlock, D. J., Cameron, J. D., and May, H.: Preliminary investigation of the pharmacology of the human internal anal sphincter. Gut 10:674, 1969. 120. Parks, A. G., Swash, M., and Urich, H.: Sphincter denervation in anorectal incontinence and rectal prolapse. Gut 18:656, 1977. 121. Penninckx, F., Lestar, B., and Kerremans, R.: The internal anal sphincter: mechanisms of control and its role in maintaining anal continence. Baillieres Clin. Gastroenterol. 6: 193, 1992. 122. Percy, J. P., Neill, M. E., Swash, M., and Parks, A. G.: Electrophysiological study of motor nerve supply of pelvic floor. Lancet 1:16, 1981.

Autonomic and Somatic Systems to the Anorectum and Pelvic Floor 123. Pezim, M. E., Pemberton, J. H., Levin, K. E., et al.: Parameters of anorectal and colonic motility in health and in severe constipation. Dis. Colon Rectum 36:484, 1993. 124. Podnar, S., Mrkaic, M., and Vodusek, D. B.: Standardization of anal sphincter electromyography: quantification of continuous activity during relaxation. Neurourol. Urodyn. 21: 540, 2002. 125. Rao, S. S., Read, N. W., Davison, P. A., et al.: Anorectal sensitivity and responses to rectal distention in patients with ulcerative colitis. Gastroenterology 93:1270, 1987. 126. Rao, S. S., Welcher, K. D., and Leistikow, J. S.: Obstructive defecation: a failure of rectoanal coordination. [comment]. Am. J. Gastroenterol. 93:1042, 1998. 127. Rayner, V.: Observations on the functional internal anal sphincter of the vervet monkey. J. Physiol. (Lond.) 213:27P, 1971. 128. Read, M., Read, N. W., Barber, D. C., and Duthie, H. L.: Effects of loperamide on anal sphincter function in patients complaining of chronic diarrhea with fecal incontinence and urgency. Dig. Dis. Sci. 27:807, 1982. 129. Rentsch, M., Paetzel, C., Lenhart, M., et al.: Dynamic magnetic resonance imaging defecography: a diagnostic alternative in the assessment of pelvic floor disorders in proctology. Dis. Colon Rectum 44:999, 2001. 130. Rociu, E., Stoker, J., Eijkemans, M. J., et al.: Fecal incontinence: endoanal US versus endoanal MR imaging. Radiology 212:453, 1999. 131. Rockwood, T. H., Church, J. M., Fleshman, J. W., et al.: Patient and surgeon ranking of the severity of symptoms associated with fecal incontinence: the fecal incontinence severity index. Dis. Colon Rectum 42:1525, 1999. 132. Rockwood, T. H., Church, J. M., Fleshman, J. W., et al.: Fecal incontinence quality of life scale: quality of life instrument for patients with fecal incontinence. Dis. Colon Rectum 43:9; discussion 16, 2000. 133. Rogers, J., Levy, D. M., Henry, M. M., and Misiewicz, J. J.: Pelvic floor neuropathy: a comparative study of diabetes mellitus and idiopathic faecal incontinence. Gut 29:756, 1988. 134. Romero, Y., Evans, J. M., Fleming, K. C., and Phillips, S. F.: Constipation and fecal incontinence in the elderly population. Mayo Clin. Proc. 71:81, 1996. 135. Roppolo, J. R., Nadelhaft, I., and de Groat, W. C.: The organization of pudendal motoneurons and primary afferent projections in the spinal cord of the rhesus monkey revealed by horseradish peroxidase. J. Comp. Neurol. 234:475, 1985. 136. Rosen, H. R., Urbarz, C., Holzer, B., et al.: Sacral nerve stimulation as a treatment for fecal incontinence. Gastroenterology 121:536, 2001. 137. Ruch, T. C.: Pathophysiology of pain. In Ruch, T. C., and Patton, H. D. (eds.): Physiology and Biophysics: I. The Brain and Neural Function. Philadelphia, W. B. Saunders, p. 272, 1979. 138. Salducci, J., Planche, D., and Naudy, B.: Physiological role of the internal anal sphincter and the external anal sphincter during micturition. In Wienbeck, M. (ed.): Motility of the Digestive Tract. New York, Raven Press, p. 513, 1982. 139. Salvioli, B., Bharucha, A. E., Rath-Harvey, D., et al.: Rectal compliance, capacity and rectoanal sensation in fecal incontinence. Am. J. Gastroenterol. 96:2158, 2001.

297

140. Sato, M., Mizuno, N., and Konishi, A.: Localization of motoneurons innervating perineal muscles: a HRP study in cat. Brain Res. 140: 1978. 141. Schroder, H. D.: Onuf’s nucleus X: a morphological study of a human spinal nucleus. Anat. Embyrol. 162:443, 1981. 142. Schroder, H. D., and Reske-Nielsen, E.: Fiber types in the striated urethral and anal sphincters. Acta Neuropathol. 60: 278, 1983. 143. Schuster, M. M.: Motor action of rectum and anal sphincters in continence and defecation. In Code, C. F. (ed.): Handbook of Physiology: Section 6. Alimentary Canal. Washington, DC, American Physiol Society, p. 2121, 1968. 144. Schweiger, M.: Method for determining individual contributions of voluntary and involuntary anal sphincters to resting tone. Dis. Colon Rectum 22:415, 1979. 145. Scott, H. W. J., and Cantrell, J. R.: Colonmetrographic studies of the effects of section of the parasympathetic nerves of the colon. Bull. Johns Hopkins Hosp. 85:310, 1949. 146. Semba, T., Mishima, H., and Date, T.: Studies on a vesico-anal inhibitory reflex. Jpn. J. Physiol. 6:108, 1956. 147. Sengupta, J. N., Saha, J. K., and Goyal, R. K.: Stimulusresponse function studies of esophageal mechanosensitive nociceptors in sympathetic afferents of opossum. J. Neurophysiol. 64:796, 1990. 148. Shafik, A.: Straining puborectalis reflex: description and significance of a “new” reflex. Anat. Rec. 229:281, 1991. 149. Shafik, A.: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation: mass contraction of the pelvic floor muscles. Int. Urogynecol. J. 9:28, 1998. 150. Sherrington, C. S.: Notes on the arrangement of some motor fibres in the lumbosacral plexus. J. Physiol. (Lond.) 13:672, 1892. 151. Shirran, E., and Brazzelli, M.: Absorbent products for containing urinary and/or faecal incontinence in adults. Cochrane Database Syst. Rev. 4: CD001406, 2003. 152. Siproudhis, L., Bellissant, E., Juguet, F., et al.: Perception of and adaptation to rectal isobaric distension in patients with faecal incontinence. Gut 44:687, 1999. 153. Stebbing, J. F., Brading, A. F., and Mortensen, N. J.: Nitric oxide and the rectoanal inhibitory reflex: retrograde neuronal tracing reveals a descending nitrergic rectoanal pathway in a guinea-pig model. Br. J. Surg. 83:493, 1996. 154. Sultan, A. H., Kamm, M. A., Hudson, C. N., et al.: Anal-sphincter disruption during vaginal delivery [see comments]. N. Engl. J. Med. 329:1905, 1993. 155. Sun, W. M., Read, N. W., and Donnelly, T. C.: Anorectal function in incontinent patients with cerebrospinal disease. Gastroenterology 99:1372, 1990. 156. Sun, W. M., Read, N. W., and Miner, P. B.: Relation between rectal sensation and anal function in normal subjects and patients with faecal incontinence. Gut 31: 1056, 1990. 157. Sung, J. H.: Autonomic neurons affected by lipid storage in the spinal cord in Fabry’s disease: distribution of autonomic neurons in the sacral cord. J. Neuropathol. Exp. Neurol. 38: 87, 1979.

298

Function of the Peripheral Nervous System

158. Sung, J. H. and Mastri, A. R.: Spinal autonomic neurons in Werdnig-Hoffmann disease, mannosidosis, and Hurler’s syndrome: distribution of autonomic neurons in the sacral spinal cord. J. Neuropathol. Exp. Neurol. 39: 441, 1980. 159. Sung, J. H., Mastri, A. R., and Segal, E.: Pathology of Shy-Drager syndrome. J. Neuropathol. Exp. Neurol. 38: 353, 1979. 160. Swash, M.: Early and late components in the human anal reflex. J. Neurol. Neurosurg. Psychiatry 45:767, 1982. 161. Taverner, D., and Smiddy, F. G.: An electromyographic study of the normal function of the external anal sphincter and pelvic diaphragm. Dis. Colon Rectum 2:153, 1959. 162. Turnbull, G. K., Hamdy, S., Aziz, Q., et al.: The cortical topography of human anorectal musculature. Gastroenterology 117:32, 1999. 163. Uher, E. M., and Swash, M.: Sacral reflexes: physiology and clinical application. Dis. Colon Rectum 41:1165, 1998. 164. Vaizey, C. J., Kamm, M. A., and Bartram, C. I.: Primary degeneration of the internal anal sphincter as a cause of passive faecal incontinence. Lancet 349:612, 1997. 165. Vaizey, C. J., Kamm, M. A., Turner, I. C., et al.; Effects of short term sacral nerve stimulation on anal and rectal function in patients with anal incontinence. Gut 44:407, 1999.

166. Voderholzer, W. A., Neuhaus, D. A., Klauser, A. G., et al.: Paradoxical sphincter contraction is rarely indicative of anismus. Gut 41:258, 1997. 167. Wald, A., and Tunuguntla, A. K.: Anorectal sensorimotor dysfunction in fecal incontinence and diabetes mellitus: modification with biofeedback therapy. N. Engl. J. Med. 310:1282, 1984. 168. Weinberg, A. G.: The anorectal myenteric plexus: its relation to hypoganglionosis of the colon. Am. J. Clin. Pathol. 54:637, 1970. 169. Wexner, S. D., Marchetti, F., and Salanga, V. D. et al.: Neurophysiologic assessment of the anal sphincters. Dis. Colon Rectum 34:606, 1991. 170. Wheatley, I. C., Hardy, K. J., and Dent, J.: Anal pressure studies in spinal patients. Gut 18:488, 1977. 171. White, J. C., and Sweet, W. H.: Pain and the Neurosurgeon: A Forty Year Experience. Springfield, IL, Charles C Thomas, 1969. 172. Whitehead, W. E., and Palsson, O. S.: Is rectal pain sensitivity a biological marker for irritable bowel syndrome? Psychological influences on pain perception. Gastroenterology 115:1263, 1998. 173. Whitehead, W. E., and Wald, A.: Functional disorders of the anus and rectum. Gut 45 (Suppl. II): II55, 1999.

14 Autonomic Systems to the Urinary Bladder and Sexual Organs WILLIAM C. DE GROAT AND AUGUST M. BOOTH

Lower Urinary Tract Anatomy Innervation Parasympathetic Pathways Sympathetic Pathways Neural Modulatory Mechanisms Somatic Pathways Afferent Pathways Urothelial-Afferent Interactions Reflex Mechanisms Controlling the Lower Urinary Tract Anatomy of Central Nervous Pathways Controlling the Lower Urinary Tract Pathways in the Spinal Cord

Pathways in the Brain Organization of Urine Storage and Voiding Reflexes Sympathetic Pathways Somatic Pathways to the Urethral Sphincter Voiding Reflexes Neurotransmitters in Micturition Reflex Pathways Excitatory Neurotransmitters Inhibitory Neurotransmitters Neurogenic Dysfunction of the Lower Urinary Tract Spinal Injury

The urogenital system is affected by a variety of neuropathologic conditions and often is the site of the first symptoms produced by diffuse neurologic diseases, such as diabetic neuropathy or multiple sclerosis.12,13,41,62,82,98,111,112,137 This sensitivity to neurogenic disorders is not surprising in view of the complexity and importance of neural mechanisms in the regulation of urogenital function.33,45,48,98,111,131 For example, the activity of the lower urinary tract and certain sex organs is almost totally dependent on central nervous control, whereas many other visceral systems, in particular the gastrointestinal tract and cardiovascular system, continue to function in the absence of extrinsic neural input. In addition, many responses of the urogenital system require the coordination of autonomic and somatic efferent mechanisms at various levels of the lumbosacral spinal cord. These spinal mechanisms are in turn modulated by inputs from higher centers in the brain and in some instances are subject to voluntary control. This chapter reviews the organization of the neural pathways regulating the urogenital system with

Sex Organs Innervation Parasympathetic Pathways Sympathetic Pathways Somatic Pathways Erection Peripheral Mechanisms Central Mechanisms Glandular Secretion Emission-Ejaculation Central Reflex Pathways

the aim of providing the anatomic and physiologic background for understanding urinary and sexual dysfunction resulting from neurologic disorders.

Lower Urinary Tract ANATOMY The storage and periodic elimination of urine is controlled by the activity of two functional units in the lower urinary tract: (1) a reservoir (the bladder) and (2) an outlet (consisting of the bladder neck, urethra, and striated muscles of the pelvic floor). Two regions of the bladder have been distinguished based on embryonic origin. The detrusor, which comprises the major portion of the bladder, is derived from endoderm, whereas the trigone, a small region located posteriorly at the neck of the bladder, is derived from mesoderm and represents the extension of 299

300

Function of the Peripheral Nervous System

the longitudinal muscle layers of the ureters. A more physiologic designation of regions of the bladder has been proposed based on neural innervation and responses to pharmacologic agents.56,98,131 According to this approach, the bladder can be divided circumferentially at the level of the ureteral orifices into cranial and caudal segments termed the body and base, respectively. The trigone occupies the posterior area of the base. The distinguishing characteristics of these two regions are discussed later in this chapter. The luminal surface of the bladder is covered by a multilayered urothelium that functions as a highly efficient barrier to movement of water and ionized substances across the bladder wall.33,56,79,81 The primary urineplasma barrier is attributed to the layer of superficial umbrella cells, which are interconnected by tight junctions and have an unusual asymmetrical unit membrane on their apical surface that consists of an outer leaflet of plaques composed of proteins called uroplakins. Umbrella cells have large numbers of discoid vesicles with asymmetrical membrane structure located just under the apical surface. It is believed that trafficking of these vesicles to the apical surface during stretch allows the umbrella cells to increase their surface area during bladder distention.134 Urothelial cells also have neuronallike properties, including the ability to release neurotransmitters (ATP and nitric oxide [NO]) in response to stretch or chemical stimulation.14,15,18,60,136 Transmitter release may be mediated by exocytosis associated with vesicle trafficking. Transmitters released from the urothelial cells may act in an autocrine/paracrine manner within the urothelium or on subepithelial myofibroblasts, nerves, or blood vessels to influence various functions, including the urothelial barrier, local blood flow, and sensory mechanisms.15,40,58,60,127 According to the traditional view, the bladder smooth muscle is arranged in three layers (inner and outer longitudinal and middle circular), although there is considerable interdigitation between these layers, giving the bladder wall a meshlike appearance.33,56,131 The inner longitudinal layer is continuous with the longitudinal smooth muscle of the urethra, which constitutes the greater portion of the urethral wall. The urethra has a very thin layer of circular muscle but is surrounded by striated muscle that in males is very prominent and extends from the urogenital diaphragm to the neck of the bladder. An anatomic sphincter between the bladder and the urethra has not been identified; however, radiographic studies and measurements of urethral pressure indicate the existence of a physiologic internal sphincter that maintains urinary continence by closure of the bladder neck and proximal urethra.33 The sphincter-like properties of this region have been attributed to the abundance of elastic tissue in the submucosa and the tone of the urethral smooth muscle. Continence is thought to be dependent on

a combination of factors, including urethral wall tension, the caliber of the urethral lumen, and the functional length of the urethra. The striated muscles surrounding the urethra are not essential for urinary continence, but are important in the voluntary termination of urine flow (Fig. 14–1) and in the prevention of stress incontinence.33,129

FIGURE 14–1 Combined cystometrograms and sphincter electromyograms (EMGs) comparing reflex voiding responses in an infant (A) and in a paraplegic patient (C) with a voluntary voiding response in an adult (B). The abscissa in all records represents bladder volume in milliliters and the ordinates represent bladder pressure in centimeters of H2O and electrical activity of the EMG recording. On the left side of each trace, the arrows indicate the start of a slow infusion of fluid into the bladder (bladder filling). Vertical dashed lines indicate the start of sphincter relaxation, which precedes by a few seconds the bladder contraction in A and B. B, Note that a voluntary cessation of voiding (stop) is associated with an initial increase in sphincter EMG followed by a reciprocal relaxation of the bladder. A resumption of voiding is again associated with sphincter relaxation and a delayed increase in bladder pressure. C, Conversely, in the paraplegic patient the reciprocal relationship between bladder and sphincter is abolished. During bladder filling, transient uninhibited bladder contractions occur in association with sphincter activity. Further filling leads to more prolonged and simultaneous contractions of the bladder and sphincter (bladder-sphincter dyssynergia). Loss of the reciprocal relationship between bladder and sphincter in paraplegic patients interferes with bladder emptying.

Autonomic Systems to the Urinary Bladder and Sexual Organs

Under normal conditions, the urinary bladder and outlet exhibit a reciprocal relationship in effecting the storage and elimination of urine (see Fig. 14–1). During storage, the bladder neck and proximal urethra are closed, the sphincter electromyogram (EMG) gradually increases during bladder filling, and maximal intraurethral pressures range from 20 to 50 cm H2O. The detrusor muscle, in contrast, is quiescent, allowing intravesical pressure to remain low (5 to 15 cm H2O) over a wide range of bladder volumes (100 to 400 mL). During voluntary micturition the initial event is a reduction of intraurethral pressure, which reflects a relaxation of the pelvic floor and the paraurethral striated muscles (see Fig. 14–1).129 This coincides with a mechanical shortening of the urethra and the appearance of a cone-shaped opening of the bladder neck. These changes in the urethra are followed in a few seconds by a detrusor contraction and a rise in intravesical pressure that is maintained until the bladder empties. The detrusor contraction produces a further opening of the proximal urethra by shortening of the inner longitudinal muscle bundles, which extend from the bladder into the urethra, and by contractions of outer longitudinal muscle bundles, which loop around the bladder neck. Reflex inhibition of the smooth and striated muscles of the urethra also contributes to the reduction of outlet resistance during micturition. These changes are coordinated by neural pathways at the thoracolumbar and sacral levels of the spinal cord.

INNERVATION The innervation of the lower urinary tract is derived from three sets of peripheral nerves: sacral parasympathetic (pelvic nerves), thoracolumbar sympathetic (hypogastric nerves and sympathetic chain), and sacral somatic nerves (primarily the pudendal nerves) (Fig. 14–2).48,131

Parasympathetic Pathways The sacral parasympathetic outflow, which in humans originates from the S2 to S4 segments of the spinal cord, provides the major excitatory input to the bladder.33,56 Cholinergic preganglionic neurons (PGNs) located in the intermediolateral region of the sacral spinal cord (Fig. 14–3) send axons to cholinergic ganglion cells in the pelvic plexus and in the bladder wall.122 Transmission in bladder ganglia is mediated by a nicotinic cholinergic mechanism, which is sensitive to modulation by various transmitter systems, including muscarinic, adrenergic, purinergic, and peptidergic (Table 14–1).46,50,53,75 The ganglion cells in turn excite the bladder smooth muscle. Histochemical studies of the ganglia and nerves supplying the human lower urinary tract have shown that a large proportion of ganglion cells contain acetylcholinesterase (AChE) as well as vesicular acetyl-

301

choline transporter (VAChT) and therefore are presumably cholinergic. AChE- and VAChT-positive nerves are abundant in all parts of the bladder but are less extensive in the urethra.46,56,98 Neuropeptide Y (NPY) and nitric oxide synthase (NOS) have also been identified in a large percentage (40% to 95%) of intramural ganglia of the human bladder. Several populations of axonal varicosities have been detected in close proximity to intramural ganglion cells, including (1) substance P and calcitonin gene–related peptide (CGRP) positive axons, which are presumably collaterals of extrinsic sensory nerves; (2) tyrosine hydroxylase and NPY axons, which are likely to be sympathetic axons; and (3) vasoactive intestinal polypeptide (VIP), galanin, and NPYcontaining axons, which are presumably preganglionic nerve terminals.56 Parasympathetic neuroeffector transmission in the bladder is mediated by acetylcholine (ACh) acting on postjunctional muscarinic receptors.98 Both M2 and M3 muscarinic receptor subtypes are expressed by bladder smooth muscle; however, examination of subtype-selective muscarinic receptor antagonists and studies of muscarinic receptor knockout mice have revealed that the M3 subtype is the principal receptor involved in excitatory transmission.88,89 In bladders of various animals, stimulation of parasympathetic nerves also produces a noncholinergic contraction that is resistant to atropine and other muscarinic receptor– blocking agents. Activation of M3 receptors triggers intracellular Ca2⫹ release, whereas activation of M2 receptors inhibits adenylate cyclase. The latter may contribute to bladder contractions by suppressing adrenergic inhibitory mechanisms, which are mediated by ␤3-adrenergic receptors and stimulation of adenylate cyclase. ATP (see Table 14–1) has been identified as the excitatory transmitter mediating the noncholinergic contractions.32,106 ATP excites the bladder smooth muscle by acting on P2X purinergic receptors, which are ligand-gated ion channels. Among the seven types of P2X receptors that have been identified in the rat bladder, P2X1 is the major subtype expressed in the rat and also in the human bladder smooth muscle. Although purinergic excitatory transmission is not important in the normal human bladder, it appears to be involved in bladders from patients with pathologic conditions such as chronic urethral outlet obstruction or interstitial cystitis.32,80,98,103,104 Parasympathetic pathways to the urethra induce relaxation during voiding. In various species the relaxation is not affected by muscarinic antagonists and therefore is not mediated by ACh. However inhibitors of NOS block the relaxation in vivo during reflex voiding or block the relaxation of urethral smooth muscle strips induced in vitro by electrical stimulation of intramural nerves, indicating that NO is the inhibitory transmitter involved in relaxation.28,68,98 In some species neurally evoked contractions of the urethra are reduced by muscarinic receptor antagonists or by desensitization of P2X receptors, indicating that

302

Function of the Peripheral Nervous System

FIGURE 14–2 The innervation of the lower urinary tract and male genitalia. (Prepared by Tim Whitney, M.D.)

ACh or ATP is involved in excitatory transmission to urethral smooth muscle.147

Sympathetic Pathways Sympathetic preganglionic pathways that arise from the T11 to L2 spinal segments pass to the sympathetic chain ganglia and then to prevertebral ganglia in the superior hypogastric and pelvic plexuses (see Fig. 14–2), and also

to short adrenergic neurons in the bladder and urethra.48,56 Sympathetic postganglionic nerves that release norepinephrine provide an excitatory input to smooth muscle of the urethra and bladder base, an inhibitory input to smooth muscle in the body of the bladder, and inhibitory and facilitatory input to vesical parasympathetic ganglia.46,56 Histofluorescence microscopy in animals and humans has shown that the smooth muscle of the bladder base is richly innervated by adrenergic terminals, but the

303

Autonomic Systems to the Urinary Bladder and Sexual Organs

FIGURE 14–3 Transneuronal virus tracing of the central pathways controlling the urinary bladder of the rat. Injection of pseudorabies virus into the wall of the urinary bladder leads to retrograde transport of virus (dashed arrows) and sequential infection of postganglionic neurons, preganglionic neurons, and then various central neural circuits synaptically linked to the preganglionic neurons. Normal synaptic connections are indicated by solid arrows. At long survival times virus can be detected with immunocytochemical techniques in neurons at specific sites throughout the spinal cord and brain, extending to the pontine micturition center in the pons (i.e., Barrington’s nucleus) and to the cerebral cortex. Other sites in the brain labeled by virus are (1) the paraventricular nucleus (PVN), medial preoptic area (MPOA), and periventricular nucleus (Peri V.N.) of the hypothalamus; (2) periaqueductal gray (PAG); (3) locus ceruleus (LC) and subceruleus; (4) red nucleus; (5) medullary raphe nuclei; and (6) the noradrenergic cell group designated A5. An L6 spinal cord section shows, on the left side, the distribution of virus-labeled parasympathetic preganglionic neurons (ⵧ) and interneurons (䊉) in the region of the parasympathetic nucleus, the dorsal commissure (DCM), and the superficial laminae of the dorsal horn (DH) 72 hours after injection of the virus into the bladder. The right side shows the entire population of preganglionic neurons (PGN) (ⵧ) labeled by axonal tracing with fluorescent dye (fluorogold) injected into the pelvic ganglia, and the distribution of virus-labeled bladder PGN (䊏). Composite diagram of neurons in 12 spinal sections (42 ␮m).

Table 14–1. Reflexes to the Lower Urinary Tract Afferent Pathway

Efferent Pathway

Central Pathway

Urine Storage Low-level vesical afferent activity (pelvic nerve)

1. External sphincter contraction (somatic nerves) 2. Internal sphincter contraction (sympathetic nerves) 3. Detrusor inhibition (sympathetic nerves) 4. Ganglionic inhibition (sympathetic nerves) 5. Sacral parasympathetic outflow inactive 1. Inhibition of external sphincter activity 2. Inhibition of sympathetic outflow 3. Activation of parasympathetic outflow to the bladder 4. Activation of parasympathetic outflow to the urethra

Spinal reflexes

Micturition High-level vesical afferent activity (pelvic nerve)

Spinobulbospinal reflexes

Spinal reflex

304

Function of the Peripheral Nervous System

bladder body has a considerably weaker adrenergic innervation.98,131 Vesical parasympathetic ganglion cells and the smooth muscle of the proximal urethra also receive an extensive adrenergic innervation.46,48 Radioligand receptor-binding studies showed that ␣-adrenergic receptors are concentrated in the bladder base and proximal urethra, whereas ␤-adrenergic receptors are most prominent in the bladder body.1 These observations are consistent with pharmacologic studies showing that sympathetic nerve stimulation or exogenous catecholamines produce ␤-adrenergic receptor–mediated inhibition of the body and ␣-adrenoceptor–mediated contraction of the base, dome, and urethra. Molecular and contractility studies have shown that ␤3-adrenergic receptors elicit inhibition and ␣1-adrenergic receptors elicit contractions. The ␣1Aadrenergic receptor subtype is most prominent in normal bladders and the ␣1D subtype is upregulated in bladders from patients with outlet obstruction, raising the possibility that ␣1-adrenergic receptor excitatory mechanisms in the bladder might contribute to irritative lower urinary tract symptoms in patients with benign prostatic hyperplasia.55,98

Neural Modulatory Mechanisms Presynaptic modulatory mechanisms and synaptic communication between parasympathetic and sympathetic pathways to the bladder has been demonstrated in bladder ganglia and at postganglionic nerve terminals. In the cat, stimulation of the hypogastric (sympathetic) nerves elicits an initial inhibitory and a delayed facilitatory modulation of cholinergic transmission in parasympathetic bladder ganglia mediated by ␣2- and ␣1-adrenergic receptors, respectively,46,53,75 indicating that sympathetic pathways can influence neural input to the bladder as well as directly affect the bladder smooth muscle (see Table 14–1). Transmission in cat bladder ganglia is also modulated by enkephalins released as cotransmitters along with ACh from preganglionic nerve terminals.49,73 The inhibitory effect of enkephalins can be blocked by the opioid antagonist naloxone and occurs by a presynaptic inhibitory mechanism. Adenosine, presumably derived from ATP released in the bladder ganglia, also exerts an inhibitory action on nicotinic cholinergic transmission in cat bladder ganglia.46 Adrenergic and enkephalinergic inhibitory mechanisms in cat bladder ganglia are frequency dependent, being prominent at low frequencies (0.25 to 0.5 Hz) and markedly reduced at higher frequencies (1 to 10 Hz). This phenomenon is presumably related to the marked temporal facilitation that occurs in cat bladder ganglia at frequencies of preganglionic nerve stimulation above 0.5 Hz.23,54 It has been speculated that cat bladder ganglia function as high-pass filters, eliminating parasympathetic input to the bladder when preganglionic activity is low during urine storage, but significantly amplifying the

input to the bladder when firing is increased during voiding.46 Thus the ganglion acts like a gating circuit. Frequency-dependent adrenergic and enkephalinergic inhibitory mechanisms complement this gating function by effectively inhibiting transmission during urine storage but turning off during voiding to allow complete bladder emptying. In the rat, transmission at synapses in the major pelvic ganglion occurs with a high safety factor, exhibits very little frequency-dependent modulation, and is relatively resistant to the actions of inhibitory transmitters.46 However, the autonomic pathways to the rat urinary bladder are susceptible to modulation at postganglionic sites.46,118–121,132,147 For example, NPY, which is a cotransmitter in cholinergic and adrenergic nerves in the rat, can act prejunctionally to suppress the release of both ACh and norepinephrine in the bladder and urethra.132,147 The NPY inhibition is also frequency dependent, being most prominent at low frequencies of nerve stimulation. In addition, in the rat urinary bladder ACh released from parasympathetic nerves can induce a heterosynaptic facilitation of norepinephrine release from sympathetic nerves by acting on prejunctional M1 muscarinic receptors on the adrenergic terminals. Prejunctional M1 facilitatory and M2/4 inhibitory muscarinic modulatory mechanisms have also been identified on parasympathetic nerve terminals in the rat bladder.119–121 These receptors mediate positive and negative cholinergic feedback mechanisms, respectively, and regulate the release of ACh. Inhibitory M2/4 mechanisms are dominant at low frequencies of nerve activity and therefore could contribute to urine storage, whereas M1 facilitatory mechanisms are dominant at high frequencies of nerve stimulation and could contribute to an enhancement of neurally evoked bladder contractions during micturition to induce complete bladder emptying.

Somatic Pathways The external urethral sphincter (EUS), which is composed of striated muscle, receives a somatic innervation via the pudendal nerve from anterior horn cells in the third and fourth sacral segments (see Fig. 14–2). Branches of the pudendal nerve and other sacral somatic nerves also carry efferent impulses to muscles of the pelvic floor and proprioceptive afferent signals from these muscles, as well as sensory information from the urethra. Analysis of urethral closure mechanisms in the female rat during sneezeinduced stress conditions revealed that the major rise in urethral pressure occurred in the mid-urethra and was mediated by efferent pathways in the pudendal nerve to the EUS as well as pathways in nerves to the iliococcygeus and pubococcygeus muscles, but not by pathways in the sympathetic or parasympathetic nerves.72 Studies of the biomechanical properties of the intact female rat urethra in vitro have confirmed the large contribution of striated

Autonomic Systems to the Urinary Bladder and Sexual Organs

muscle activity and nicotinic receptor mechanisms to the contractions of the mid-urethra.70

Afferent Pathways Afferent axons innervating the urinary tract are present in the three sets of nerves.11,24,42,48,52 The most important afferents for initiating micturition are those passing in the pelvic nerve to the sacral spinal cord. These afferents are small myelinated (A␦) and unmyelinated (C) fibers that convey information from receptors in the bladder wall to second-order neurons in the spinal cord. A␦ bladder afferents in the cat respond in a graded manner to passive distention as well as active contraction of the bladder and exhibit pressure thresholds in the range of 5 to 15 mm Hg, which are similar to those pressures at which humans report the first sensation of bladder filling.33,98,99 These fibers also code for noxious stimuli in the bladder. Conversely, C-fiber bladder afferents in the cat have very high thresholds and commonly do not respond to even high levels of intravesical pressure.65,98 However, activity in some of these afferents is unmasked or enhanced by chemical irritation of the bladder mucosa. These findings indicate that C-fiber afferents in the cat have specialized functions, such as the signaling of inflammatory or noxious events in the lower urinary tract. Nociceptive and mechanoceptive information is also carried in the hypogastric nerves to the thoracolumbar segments of the spinal cord.11 In the rat, A-fiber and C-fiber bladder afferents are not distinguishable on the basis of stimulus modality; thus both types of afferents consist of mechanosensitive and chemosensitive populations.99,108,116,117 C-fiber afferents that respond only to bladder filling have also been identified in the rat bladder and appear to be volume receptors possibly sensitive to stretch of the mucosa. C-fiber afferents are sensitive to the neurotoxins capsaicin and resiniferatoxin as well as to other substances such as tachykinins, NO, ATP, prostaglandins, and neurotrophic factors released in the bladder by afferent nerves, urothelial cells, and inflammatory cells.39,80,98,108,136 These substances can sensitize the afferent nerves and change their response to mechanical stimuli. The properties of lumbosacral dorsal root ganglion cells innervating the bladder, urethra, and EUS in the rat have been studied with patch-clamp recording techniques in combination with axonal tracing methods to identify the different populations of neurons.140,141,144–146 Based on responsiveness to capsaicin, it is estimated that approximately 70% of bladder afferent neurons in the rat are of the C-fiber type. These neurons exhibit high-threshold tetrodotoxin (TTX)–resistant sodium channels and action potentials and phasic firing (one to two spikes) in response to prolonged depolarizing current pulses. Approximately 20% of the bladder C-fiber afferent neurons also are excited by ATP, which induces a depolarization and firing

305

of afferent neurons by activating P2X3 or P2X2/3 receptors. These neurons express isolectin B4 binding, which is commonly used as a marker for ATP-responsive sensory neurons. In contrast, A-fiber afferent neurons are resistant to capsaicin and ATP, exhibiting low-threshold TTX-sensitive sodium channels and action potentials and tonic firing (multiple spikes) in response to depolarizing current pulses. C-fiber bladder afferent neurons also express a slowly decaying A-type K⫹ channel that controls spike threshold and firing frequency.141,144,146 Suppression of this K⫹ channel induces hyperexcitability of bladder afferent nerves. These properties of dorsal root ganglion cells are consistent with the different properties of A-fiber and C-fiber afferent receptors in the bladder. Immunohistochemical studies have shown that a large percentage of bladder afferent neurons contain the peptides CGRP, VIP, pituitary adenyl cyclase–activating peptide (PACAP), tachykinins, galanin, and opioid peptides.43,74,83,98 In the spinal cord, peptidergic nerve terminals have a distribution very similar to the distribution of pelvic nerve afferents labeled with horseradish peroxidase. Nerves containing these peptides are also common in the bladder, in the submucosal and epithelial layers, and around blood vessels.56 Peptidergic bladder afferent neurons in the rat also express tyrosine kinase A, a high-affinity receptor for nerve growth factor (NGF), and receptors for capsaicin (transient receptor potential protein V1 [TRPV1]) and tachykinins (neurokinin [NK]–2 and NK-3 receptors). Capsaicin, a neurotoxin that can release peptides from afferent terminals, produces inflammatory responses, including plasma extravasation and vasodilation, when applied locally to the bladder in experimental animals.83 These findings suggest that the neuropeptides may be important transmitters in the afferent pathways from the lower urinary tract. Tachykinins may also act back on afferent terminals in an auto-feedback manner to modulate the excitability of the terminals.98

Urothelial-Afferent Interactions Recent studies have revealed that the urothelium, which has been traditionally viewed as a passive barrier at the bladder luminal surface,79,81 also has specialized sensory and signaling properties that allow urothelial cells to respond to their chemical and physical environment and to engage in reciprocal chemical communication with neighboring nerves in the bladder wall.14–18,40,60,98,136 These properties include (1) expression of nicotinic, muscarinic, tachykinin, adrenergic, and capsaicin (TRPV1) receptors; (2) responsiveness to transmitters released from sensory nerves; (3) close physical association with afferent nerves; and (4) ability to release chemical mediators such as ATP and NO that can regulate the activity of adjacent nerves and thereby trigger local vascular changes and/or reflex bladder contractions. The role of ATP in urothelial-afferent communication has attracted considerable attention

306

Function of the Peripheral Nervous System

because bladder distention releases ATP from the urothelium and intravesical administration of ATP induces bladder hyperactivity, an effect blocked by administration of P2X purinergic receptor antagonists that suppress the excitatory action of ATP on bladder afferent neurons.98 Mice in which the P2X3 receptor was knocked out exhibited hypoactive bladder activity and inefficient voiding,40 suggesting that activation of P2X3 receptors on bladder afferent nerves by ATP released from the urothelium was essential for normal bladder function. It has also been reported that urothelial cells obtained from patients or cats with a chronic painful bladder condition (interstitial cystitis) release significantly larger amounts of ATP in response to mechanical stretching than do urothelial cells from normal patients.15,128 This raises the possibility that ATP-mediated signaling between the urothelium and afferent nerves is involved in the triggering of painful bladder sensations.

REFLEX MECHANISMS CONTROLLING THE LOWER URINARY TRACT The neural pathways controlling lower urinary tract function are organized as simple on-off switching circuits (Figs. 14–4 and Fig. 14–5; see also Fig. 14–1) that maintain a reciprocal relationship between the urinary bladder and urethral outlet.48,52 The principal reflex components of these switching circuits are listed in Table 14–2 and illustrated in Figure 14–5. Intravesical

FIGURE 14–4 Diagram illustrating the anatomy of the lower urinary tract and the switchlike function of the micturition reflex pathway. During urine storage, a low level of afferent activity activates efferent input to the urethral sphincter. A high level of afferent activity induced by bladder distention activates the switching circuit in the central nervous system (CNS), producing firing in the efferent pathways to the bladder, inhibition of the efferent outflow to the sphincter, and urine elimination.

pressure measurements during bladder filling in both humans and animals reveal low and relatively constant bladder pressures when bladder volume is below the threshold for inducing voiding (see Fig. 14–1). The accommodation of the bladder to increasing volumes of urine is primarily a passive phenomenon dependent upon the intrinsic properties of the vesical smooth muscle and quiescence of the parasympathetic efferent pathway. In addition, in some species urine storage is also facilitated by sympathetic reflexes that mediate an inhibition of bladder activity, closure of the bladder neck, and contraction of the proximal urethra (see Table 14–2 and Fig. 14–5).51,52 During bladder filling the activity of the sphincter EMG also increases (see Fig. 14–1), reflecting an increase in efferent firing in the pudendal nerve and an increase in outlet resistance that contributes to the maintenance of urinary continence. The storage phase of the urinary bladder can be switched to the voiding phase either involuntarily (reflexly) or voluntarily (see Fig. 14–1). The former is readily demonstrated in the human infant (Fig. 14–1A) or in the anesthetized animal when the volume of urine exceeds the micturition threshold. At this point, increased afferent firing from tension receptors in the bladder reverses the pattern of efferent outflow, producing firing in the sacral parasympathetic pathways and inhibition of sympathetic and somatic pathways. The expulsion phase consists of an initial relaxation of the urethral sphincter (Fig. 14–1) followed in a few seconds by a contraction of the bladder, an increase in bladder pressure, and flow of urine. Relaxation of the urethral outlet is mediated by activation of a parasympathetic reflex pathway to the urethra that triggers the release of NO, an inhibitory transmitter, as well as by removal of adrenergic and somatic cholinergic excitatory inputs to the urethra. Secondary reflexes elicited by flow of urine through the urethra facilitate bladder emptying.48 These reflexes require the integrative action of neuronal populations at various levels of the neuraxis (see Fig. 14–5). Certain reflexes, for example, those mediating the excitatory outflow to the sphincters and the sympathetic inhibitory outflow to the bladder, are organized at the spinal level (Fig. 14–5A), whereas the parasympathetic outflow to the detrusor has a more complicated central organization involving spinal and spinobulbospinal pathways (Fig. 14–5B).

ANATOMY OF CENTRAL NERVOUS PATHWAYS CONTROLLING THE LOWER URINARY TRACT The reflex circuitry controlling micturition consists of four basic components: spinal efferent neurons, spinal interneurons, primary afferent neurons, and neurons in

Autonomic Systems to the Urinary Bladder and Sexual Organs

A

307

B

FIGURE 14–5 Diagram showing neural circuits controlling continence and micturition. A, Urine storage reflexes. During the storage of urine, distention of the bladder produces low-level vesical afferent firing, which in turn stimulates (1) the sympathetic outflow to the bladder outlet (base and urethra) and (2) pudendal outflow to the external urethral sphincter. These responses occur by spinal reflex pathways and represent guarding reflexes, which promote continence. Sympathetic firing also inhibits detrusor muscle and modulates transmission in bladder ganglia. A region in the rostral pons (the pontine storage center) increases external urethral sphincter activity. B, Voiding reflexes. During elimination of urine, intense bladder afferent firing activates spinobulbospinal reflex pathways passing through the pontine micturition center, which stimulate the parasympathetic outflow to the bladder and internal sphincter smooth muscle and inhibit the sympathetic and pudendal outflow to the urethral outlet. Ascending afferent input from the spinal cord may pass through relay neurons in the periaqueductal gray (PAG) before reaching the pontine micturition center.

the brain that modulate spinal reflex pathways.44,98 Transneuronal virus tracing, measurements of gene expression, and patch-clamp recording in spinal cord slice preparations have recently provided many new insights into the morphologic and electrophysiologic properties of these reflex components. Neurotropic viruses, such as pseudorabies virus (PRV), have been particularly useful because they can be injected into a target organ (urinary bladder, urethra, urethral sphincter) and then move intra-axonally from the periphery to the central nervous system, where they replicate and then pass retrogradely across synapses to infect secondand third-order neurons in the neural pathways.101,126,135 Because PRV can be transported across many synapses,

it could sequentially infect all of the neurons that connect directly or indirectly to the lower urinary tract (see Fig. 14–3).

Pathways in the Spinal Cord Spinal Cord Anatomy The spinal cord gray matter is divided into three general regions: (1) the dorsal horn, which contains interneurons that process sensory input; (2) the ventral horn, which contains motoneurons; and (3) the intermediate region, which is located between the dorsal and ventral horns and contains interneurons and autonomic PGNs. These regions are further subdivided into layers, or laminae, that are

308

Function of the Peripheral Nervous System

Table 14–2. Receptors for Putative Transmitters in the Lower Urinary Tract* Tissue

Cholinergic

Adrenergic

Other

Bladder body

⫹ (M2) ⫹ (M3)

⫺ (␤3)

Bladder base

⫹ (␣1)

Urothelium

⫹ (M2) ⫹ (M3) ⫹ (M2) ⫹ (M3) ⫹ (N)

Ganglia

⫹ (N) ⫹ (M1)

⫺ (␣2) ⫹ (␣1) ⫹ (␤)

Urethra

⫹ (M)

⫹ (␣1) ⫹ (␣2) ⫺ (␤)

⫹ Purinergic (P2X) ⫺ VIP ⫹ Substance P ⫹ Purinergic (P2X) ⫺ VIP ⫹ TRPV1 ⫹ Purinergic (P2X) ⫹ 5-HT ⫹ Substance P ⫺ Enkephalinergic (␦) ⫹/⫺ Purinergic (P2X, P1) ⫹ Substance P ⫺ Nitric oxide ⫹/⫺ Purinergic (P2X, P1) ⫺ VIP ⫹ NPY

Sphincter striated muscle

⫹ (N)

⫹ (␣) ⫹ (␤)

NPY ⫽ neuropeptide Y; TRPV1 ⫽ transient receptor potential protein V1; VIP ⫽ vasoactive intestinal polypeptide. *Letters in parentheses indicate receptor type: M (muscarinic) and N (nicotinic). ⫹ and ⫺ indicate excitatory and inhibitory, respectively.

numbered starting with the superficial layer of the dorsal horn (lamina I) and extending to the ventral horn (lamina IX) and the commissure connecting the two sides of the spinal cord (lamina X) (Fig. 14–6D). Efferent Pathways Parasympathetic PGNs innervating the lower urinary tract are located in the intermediolateral gray matter (laminae V through VII) in the sacral segments of the spinal cord (see Fig. 14–6),6,47,94,96 whereas sympathetic PGNs are located in both medial (lamina X) and lateral (laminae V through VII) sites in the intermediate gray matter of the rostral lumbar spinal cord. Parasympathetic PGNs send dendrites to discrete regions of the spinal cord, including (1) the lateral and dorsolateral funiculus, (2) lamina I on the lateral edge of the dorsal horn, (3) the dorsal gray commissure (lamina X), and (4) gray matter and lateral funiculus ventral to the autonomic nucleus.44,48 As discussed later, this dendritic structure very likely indicates the origin of important synaptic inputs to these cells. Pudendal motoneurons innervating the EUS in the cat are located in the ventrolateral division of Onuf’s nucleus and send dendritic projections (1) into the lateral funiculus, (2) into lamina X, (3) into the intermediolateral gray matter, and (4) rostrocaudally within the nucleus.44 The dendritic distribution of sphincter motoneurons (i.e., lateral, dorsolateral, and dorsomedial) is similar to that of sacral PGNs, indicating that these two populations of neurons may receive synaptic inputs from the same interneuronal sites and fiber tracts in the spinal cord.

Afferent Projections in the Spinal Cord Afferent pathways from the lower urinary tract project to discrete regions of the dorsal horn that contain the interneurons as well as the soma and/or dendrites of efferent neurons innervating the lower urinary tract. Pelvic nerve afferent pathways from the urinary bladder of the cat and rat project into Lissauer’s tract at the apex of the dorsal horn and then pass rostrocaudally, giving off collaterals that extend laterally and medially through the superficial layer of the dorsal horn (lamina I) into the deeper layers (laminae V through VII and X) at the base of the dorsal horn (see Fig. 14–6A).97,123 The lateral pathway, which is the most prominent projection, terminates in the region of the sacral parasympathetic nucleus and also sends some axons to the dorsal commissure (Fig. 14–6A). Pudendal afferent pathways from the urethra and urethral sphincter exhibit a similar pattern of termination in the sacral spinal cord,48,98 whereas pudendal afferent pathways from cutaneous receptors (e.g., penis) have a prominent projection to deeper layers of the dorsal horn and the dorsal commissure. The overlap of bladder and urethral afferents in the lateral dorsal horn and dorsal commissure indicates these regions are likely to be important sites of viscerosomatic integration and to be involved in coordinating bladder and sphincter activity. Spinal Interneurons As shown in Figure 14–6B and C, interneurons retrogradely labeled by injection of PRV into the urinary bladder of the rat are located in the regions of the

Autonomic Systems to the Urinary Bladder and Sexual Organs

309

FIGURE 14–6 Comparison of the distribution of bladder afferent projections to the L6 spinal cord of the rat (A) with the distribution of c-fos–positive cells in the L6 spinal segment following chemical irritation of the lower urinary tract of the rat (B) and the distribution of interneurons in the L6 spinal cord labeled by transneuronal transport of pseudorabies virus injected into the urinary bladder (C). Afferents are labeled by wheat germ agglutinin–horseradish peroxidase injected into the urinary bladder. C-fos immunoreactivity is present in the nuclei of cells. CC ⫽ central canal; DH ⫽ dorsal horn; SPN ⫽ sacral parasympathetic nucleus. Calibration represents 500 ␮m. D, Drawing shows the laminar organization of the cat spinal cord.

spinal cord receiving afferent input from the bladder.101,126 Interneuronal locations also overlap in many respects with the dendritic distribution of the efferent neurons. A similar distribution of labeled interneurons has been noted following injections of virus into the urethra135 or the external urethral sphincter, indicating a prominent overlap of the interneuronal pathways controlling the various target organs of the lower urinary tract. The spinal neurons involved in processing afferent input from the lower urinary tract have been identified by the expression of the immediate early gene c-fos (see Fig. 14–6B). In the rat, noxious or non-noxious stimulation of the bladder and urethra increases the levels of Fos protein primarily in the dorsal commissure, the superficial dorsal horn, and the area of the sacral parasympathetic nucleus (Fig. 14–6B). Some of these interneurons send long projections to the brain, whereas others make local connections in the spinal cord and participate in segmental spinal reflexes. Patch-clamp recordings from parasympathetic PGNs in the neonatal rat spinal slice preparation have revealed that interneurons located immediately dorsal and medial to

the parasympathetic nucleus make direct monosynaptic connections with the PGNs. 6 Microstimulation of interneurons in both locations elicits glutamatergic, N-methyl-D-aspartic acid (NMDA) and non-NMDA excitatory postsynaptic currents in PGNs. Stimulation of a subpopulation of medial interneurons elicits ␥-aminobutyric acid (GABA)ergic and glycinergic inhibitory postsynaptic currents. Thus local interneurons are likely to play an important role in both excitatory and inhibitory reflex pathways controlling the preganglionic outflow to the lower urinary tract. Glutamatergic excitatory inputs have also been elicited by stimulation of the projections from lamina X and the lateral funiculus.94,95

Pathways in the Brain The neurons in the brain that control the lower urinary tract have been studied with a variety of anatomic tracing techniques in several species. In the rat, transneuronal virus tracing methods have identified many populations of central neurons that are involved in the control of

310

Function of the Peripheral Nervous System

the bladder, urethra, and urethral sphincter, including Barrington’s nucleus (the pontine micturition center [PMC]); the medullary raphe nucleus, which contains serotonergic neurons; the locus ceruleus, which contains noradrenergic neurons; periaqueductal gray (PAG); and the A5 noradrenergic cell group. Several regions in the hypothalamus and the cerebral cortex also exhibited virusinfected cells. Neurons in the cortex were located primarily in the medial frontal cortex. Other anatomic studies in which anterograde tracer substances were injected into brain areas and then identified in terminals in the spinal cord are consistent with the virus tracing data. Tracer injected into the paraventricular nucleus of the hypothalamus labeled terminals in the sacral parasympathetic nucleus as well as the sphincter motor nucleus. In contrast, neurons in the anterior hypothalamus project to the PMC. Neurons in the PMC in turn project primarily to the sacral parasympathetic nucleus and the lateral edge of the dorsal horn and the dorsal commissure, areas containing dendritic projections from the PGNs, sphincter motoneurons, and afferent inputs from the bladder. Finally, projections from neurons in the lateral pons terminate rather selectively in the sphincter motor nucleus. Thus the sites of termination of descending projections from the PMC are optimally located to regulate reflex mechanisms at the spinal level.

ORGANIZATION OF URINE STORAGE AND VOIDING REFLEXES Sympathetic Pathways The integrity of the sympathetic input to the lower urinary tract is not essential for the performance of micturition.48,131 However, physiologic experiments in animals indicate that, during bladder filling, the sympathetic system does provide a tonic inhibitory input to the bladder as well as an excitatory input to the urethra. This sympathetic input is physiologically significant because surgical interruption or pharmacologic blockade of the sympathetic innervation can reduce urethral outflow resistance, reduce bladder capacity, and increase the frequency and amplitude of bladder contractions recorded under constant-volume conditions. Sympathetic reflex activity is elicited by a sacrolumbar intersegmental spinal reflex pathway that is triggered by vesical afferent activity in the pelvic nerves (see Fig. 14–5A).51 The reflex pathway is inhibited when bladder pressure is raised to the threshold for producing micturition. This inhibitory response is abolished by transection of the spinal cord at the lower thoracic level, indicating that it originates at a supraspinal site, possibly the PMC. Therefore, the vesicosympathetic reflex represents a negative feedback mechanism whereby an increase in bladder pressure tends to increase inhibitory input to

vesical ganglia and smooth muscle, thus allowing the bladder to accommodate large volumes (Fig. 14–5A). Increased sympathetic excitatory input to the bladder base and urethra would complement these mechanisms by increasing outflow resistance.

Somatic Pathways to the Urethral Sphincter Motoneurons innervating the striated muscles of the urethral sphincter exhibit a tonic discharge that increases during bladder filling. This activity is mediated in part by low-level afferent input from the bladder (see Fig. 14–5A). During micturition the firing of sphincter motoneurons is inhibited. This inhibition is dependent in part on supraspinal mechanisms, because it is not as prominent in chronic spinal cord–transected animals. Electrical stimulation of the PMC induces sphincter relaxation, suggesting that bulbospinal pathways from the pons may be responsible for maintaining the normal reciprocal relationship between bladder and sphincter.19,33

Voiding Reflexes Spinobulbospinal Micturition Reflex Pathway Micturition is mediated by activation of the sacral parasympathetic efferent pathway to the bladder and the urethra (see Fig. 14–5B) as well as reciprocal inhibition of the somatic pathway to the urethral sphincter (see Table 14–1 and Fig. 14–5B). Studies in animals using brain-lesioning techniques revealed that neurons in the brainstem at the level of the inferior colliculus (i.e., the PMC) have an essential role in the control of the parasympathetic component of micturition.48,85,131 Removal of areas of the brain above the inferior colliculus by intercollicular decerebration usually facilitates micturition by elimination of inhibitory inputs from more rostral centers.138 However, transections at any point below the colliculi abolish micturition. Bilateral lesions in the rostral pons in the region of the locus ceruleus in cats or Barrington’s nucleus in rats also abolish micturition, whereas electrical or chemical stimulation at these sites triggers bladder contractions and micturition.85 These observations led to the concept of a spinobulbospinal micturition reflex pathway that passes through the PMC (Fig. 14–7; see also Fig. 14–5B). The pathway functions as an “on-off” switch (see Fig. 14–4) that is activated by a critical level of afferent activity arising from tension receptors in the bladder and is in turn modulated by inhibitory and excitatory influences from areas of the brain rostral to the pons (e.g., diencephalon and cerebral cortex) (see Fig. 14–7).48,131 In contrast to the reflex control of the bladder, the parasympathetic control of the urethra in the rat appears to be dependent on pathways organized in the spinal cord

Autonomic Systems to the Urinary Bladder and Sexual Organs

311

FIGURE 14–7 Diagram showing the organization of the parasympathetic excitatory reflex pathway to the detrusor muscle. Scheme is based on electrophysiologic studies in cats. In animals with an intact spinal cord, micturition is initiated by a supraspinal reflex pathway passing through a center in the brainstem. The pathway is triggered by myelinated afferents (A␦ fibers), which are connected to the tension receptors in the bladder wall. Injury to the spinal cord above the sacral segments interrupts the connections between the brain and spinal autonomic centers and initially blocks micturition. However, over a period of several weeks following cord injury, a spinal reflex mechanism emerges that is triggered by unmyelinated vesical afferents (C fibers); the A-fiber afferent inputs are ineffective. The C-fiber reflex pathway is usually weak or undetectable in animals with an intact nervous system. Stimulation of the C-fiber bladder afferents by instillation of ice water into the bladder (cold stimulation) activates voiding responses in patients with spinal cord injury. Capsaicin (20 to 30 mg subcutaneously) blocks the C-fiber reflex in chronic spinal cord–injured cats, but does not block micturition reflexes in intact cats. Intravesical capsaicin also suppresses detrusor hyperreflexia and cold-evoked reflexes in patients with neurogenic bladder dysfunction.

that are modulated by input from the brain. NO-mediated relaxation of the urethra that occurs in response to bladder distention is reduced but not eliminated by acute transection of the spinal cord.71 The reflex relaxation of the urethra is also very prominent in chronic spinal cord–transected rats. Electrophysiologic studies in cats and rats have confirmed that the parasympathetic efferent outflow to the urinary bladder is activated by a long-latency supraspinal reflex pathway.36,44,47,52,84 In cats, recordings from sacral parasympathetic PGNs innervating the urinary bladder show that reflex firing occurs with a long latency (65 to 100 ms) following stimulation of myelinated (A␦) vesical afferents in the pelvic nerve. Afferent stimulation also evokes negative field potentials in the rostral pons at latencies of 30 to 40 ms, whereas electrical stimulation in the pons excites sacral PGNs at latencies of 45 to 60 ms. The sum of the latencies for the spinobulbar and bulbospinal components of the reflex pathway approximates the latency for the entire

reflex. In cats it is believed that the ascending afferent pathways from the spinal cord project to a relay station in the PAG, which then connects to the PMC (see Fig. 14–5B).19 Pontine Micturition Center Physiologic and anatomic experiments have provided substantial support for the concept that neuronal circuitry in the PMC functions as a switch in the micturition reflex pathway. The switch seems to regulate bladder capacity and also coordinate the activity of the bladder and EUS. Electrical or chemical stimulation in the PMC of the rat, cat, and dog induces (1) a suppression of urethral sphincter EMG, (2) firing of sacral PGNs, (3) bladder contractions, and (4) release of urine.19,48,84,131 Conversely, microinjections of putative inhibitory transmitters into the PMC of the cat can increase the volume threshold for inducing micturition and in high doses completely block reflex voiding, indicating that synapses in this region are

312

Function of the Peripheral Nervous System

important for regulating the set point for reflex voiding and also are an essential link in the reflex pathway.84 Brain imaging studies using positron emission tomography (PET) or functional magnetic resonance imaging (fMRI) have identified increased neuronal activity in the PMC and PAG during voiding.9,10,21,22 Suprapontine Control of Micturition The organization of suprapontine pathways controlling micturition is less well defined, despite the fact that there is a large body of literature dealing with the responses of the lower urinary tract to lesions or electrical stimulation of the brain. In brief, it appears that the voluntary control of micturition in humans is dependent upon (1) connections between the frontal cortex and the septal and preoptic regions of the hypothalamus and (2) connections between the paracentral lobule and the brainstem and spinal cord.48,131 Lesions to these areas of cortex resulting from tumors, aneurysms, or cerebrovascular disease appear to remove inhibitory control over the anterior hypothalamic area, which normally provides an excitatory input to micturition centers in the brainstem.138 Electrical stimulation of anterior and lateral hypothalamic regions in animals induces bladder contractions and voiding, whereas stimulation of posterior and medial hypothalamic areas inhibits bladder activity. According to results obtained in cats, the inhibitory and excitatory effects of hypothalamic stimulation are believed to be mediated, respectively, by activation of sympathetic inhibitory pathways and activation of parasympathetic excitatory pathways to the bladder. Human PET scan studies have revealed that two cortical areas (the right dorsolateral prefrontal cortex and the anterior cingulate gyrus) were active (i.e., exhibited increased blood flow) during voiding.19,21,22 The hypothalamus, including the preoptic area as well as the pons and the PAG, also showed activity in concert with voluntary micturition. It is noteworthy that the active areas were predominately on the right side of the brain, which is consistent with reports that urge incontinence is correlated with lesions in the right hemisphere. Other PET studies that examined the changes in brain activity during filling of the bladder revealed that increased activity occurred in the PAG, the midline pons, the mid-cingulate gyrus, and bilaterally in the frontal lobes.9,10,90 It was concluded that the results were consistent with the notion that the PAG receives information about bladder fullness and then relays this information to other brain areas involved in the control of bladder storage. A PET study was also conducted in adult female volunteers to identify brain structures involved in voluntary control of pelvic floor muscles.20 The results revealed that the superomedial precentral gyrus, the most medial portion of the motor cortex, is activated during pelvic floor contraction. In addition, the right anterior cingulate gyrus was activated during sustained pelvic floor straining.

NEUROTRANSMITTERS IN MICTURITION REFLEX PATHWAYS Excitatory Neurotransmitters Excitatory transmission in the central pathways to the lower urinary tract may depend on several types of transmitters, including glutamic acid, neuropeptides (substance P, VIP, PACAP), norepinephrine, ACh, and NO.48,98 Experiments in rats have revealed that glutamic acid is an essential transmitter in the ascending, pontine, and descending limbs of the spinobulbospinal micturition reflex pathway and in the reflex pathways controlling the EUS. NMDA and non-NMDA glutamatergic synaptic mechanisms appear to interact synergistically to mediate transmission in these pathways.

Inhibitory Neurotransmitters Damage to central inhibitory mechanisms following disease or injury to the nervous system leads to failure of urine storage and bladder hyperactivity and incontinence.137–139,142,143 Animal studies indicate that transmission in the PMC and spinal cord is regulated by multiple transmitters, including opioid peptides (enkephalins), inhibitory amino acids (GABA, glycine), 5-hydroxytryptamine, ACh, and dopamine.48,98,139,142,143 Experiments in anesthetized animals indicate that GABA and enkephalins exert a tonic inhibitory control in the PMC and regulate bladder capacity. The inhibitory effects are mediated by GABAA and ␮ opioid receptors, respectively. Administration of GABAA or opioid receptor antagonists into the PMC reduces the micturition volume threshold, indicating that the set point for reflex voiding is regulated by inhibitory mechanisms in the brain (see Fig. 14-5).48,55,85 GABA and enkephalins also have inhibitory actions in the spinal cord. Baclofen, a GABAB agonist that mimics the inhibitory effect of GABA, has been used clinically via intrathecal administration in patients with hyperactive bladders to suppress bladder activity and to promote urine storage. Patients with idiopathic Parkinson’s disease often exhibit bladder hyperactivity, suggesting a role of dopamine in the control of bladder function. Animal models for Parkinson’s disease have been developed in monkeys and rats by administering neurotoxins (1-methyl-4-phenyl-1,2,3,6tetrahydropyridine [MPTP] or 6-hydroxydopamine) to destroy dopamine neurons.55,142,143 Following treatment the animals show motor symptoms typical of Parkinson’s disease and also have hyperactive bladders. Pharmacologic studies in MPTP-treated monkeys revealed that the bladder hyperactivity was due to the loss of dopaminergic inhibition mediated by D1 dopaminergic receptors. D2 dopaminergic receptors can mediate a facilitation of micturition. Activation of these receptors also contributes to the bladder hyperactivity after cerebral infarction in the rat.139

Autonomic Systems to the Urinary Bladder and Sexual Organs

NEUROGENIC DYSFUNCTION OF THE LOWER URINARY TRACT Neurogenic disturbances of micturition can be classified into two general categories: failure to store and failure to eliminate urine.137 Problems with storage occur with differing degrees of severity, ranging from reduced bladder capacity and frequency of urination to urgency and incontinence. A common finding is that disorders affecting the brain, particularly suprapontine areas, produce hyperactive or uninhibited bladders. Cerebrovascular accidents, Parkinson’s disease, tumors, and demyelinating diseases are common causes of this problem.13,24,41,61,112 Failure to eliminate urine occurs in various conditions that interrupt the detrusor-to-detrusor excitatory reflex pathway or that interfere with the coordination between detrusor and sphincters.33,137 Areflexic bladders can occur with (1) lower motor neuron lesions, including damage to the pelvic nerve or the sacral spinal cord; (2) lesions of the afferent pathways (e.g., diabetes, tabes dorsalis, pernicious anemia, herniated intervertebral disc); or (3) the acute stage of spinal cord injury (an upper motor neuron lesion).

Spinal Injury Complete spinal cord injury rostral to the lumbosacral level eliminates voluntary and supraspinal control of voiding, leading initially to an areflexic bladder and complete urinary retention followed by a slow development of automatic micturition and bladder hyperactivity mediated by spinal reflex pathways.24,131,137 Following recovery of reflex bladder activity, these patients usually still exhibit urinary retention as a result of a loss of coordination between the bladder and sphincter (detrusor-sphincter dyssynergia). Because sphincter dyssynergia is rarely seen in patients with suprapontine lesions who also exhibit hyperactive bladders, the condition has been attributed to the elimination of bulbospinal inhibitory pathways, which originate in the PMC (see Fig. 14–7). A similar mechanism may account for dyssynergia of the smooth muscle sphincter, which is mediated by sympathetic efferent pathways. Dyssynergia is treated surgically (sphincterotomy) or with drugs (skeletal muscle relaxants, botulinum toxin, adrenergic blocking agents) that depress somatic or sympathetic reflex pathways. The mechanisms contributing to the recovery of bladder function have been studied in rats and cats using electrophysiologic techniques.36,47,50,84,98,140 These studies revealed that the micturition reflex pathways in spinal cord–intact and chronic spinal cord–transected animals are markedly different. In both species, the central delay for the micturition reflex in chronic spinal

313

cord–transected animals is considerably shorter (⬍5 ms in rats; 15 to 40 ms in cats) than in intact animals (60 to 75 ms). In addition, in chronic spinal cord–transected cats the afferent limb of the micturition reflex consists of unmyelinated (C-fiber) afferents, whereas in intact cats it consists of myelinated (A␦) afferents (see Fig. 14–7). This was demonstrated not only with electrophysiologic recording but also by administering capsaicin, a neurotoxin that is known to disrupt the function of C-fiber afferents. In normal cats, capsaicin injected systemically in large doses did not block reflex contractions of the bladder or the A␦-fiber evoked bladder reflex. However, in chronic spinal cord–transected cats (3 to 6 weeks after spinal cord transection), capsaicin completely blocked the rhythmic bladder contractions induced by bladder distention and blocked the C-fiber–evoked reflex firing recorded on bladder postganglionic nerves.36 These data indicate that two distinct central pathways (supraspinal and spinal) utilizing different peripheral afferent limbs (A and C fiber) can mediate detrusor-to-detrusor reflexes in the cat (see Fig. 14–7). The properties of the peripheral C-fiber afferent receptors also appear to be changed in the spinal cord–injured cat. C-fiber bladder afferents in the cat usually do not respond to bladder distention (i.e., silent C fibers).65 However, in chronic spinal cord–transected cats bladder distention initiates automatic micturition by activating C-fiber afferent neurons.36,50 Thus spinal injury must change the properties of C-fiber afferent receptors in the bladder. Experiments in spinal cord–injured rats revealed that C-fiber dorsal ganglion neurons innervating the bladder increase in size,78,140 exhibit increased excitability related in part to a downregulation of A-type K⫹ channels, and exhibit a shift in the expression of Na⫹ channels from a high-threshold TTX-resistant subtype to a TTX-sensitive subtype of Na⫹ channels. Neurotrophic factors appear to contribute to this change because, after spinal cord injury, expression of neurotrophic factors, including NGF, increases in the bladder in concert with bladder hypertrophy.98 The latter most likely occurs as a result of overdistention-elicited detrusor-sphincter dyssynergia and urinary retention. A contribution of NGF to the change in bladder afferent neuron excitability and detrusor hyperreflexia in spinal cord–injured rats is supported by several observations, including (1) administration of exogenous NGF to the bladder or spinal cord can induce bladder hyperactivity and increased excitability of bladder afferent neurons accompanied by downregulation of A-type K⫹ channels, and (2) immunoneutralization of NGF in the spinal cord reduces the bladder hyperreflexia in spinal cord–injured rats and reduces the NGF levels in the lumbosacral dorsal root ganglia.114 The effects of NGF immunoneutralization can be duplicated by systemic or intravesical administration of the C-fiber

314

Function of the Peripheral Nervous System

neurotoxins capsaicin or resiniferatoxin.36,50 Intravesical administration of these toxins in patients with neurogenic bladder dysfunction also reduces bladder hyperactivity and incontinence episodes and increases bladder capacity.55,98 Other reflexes that are unmasked following spinal cord injury also appear to be mediated by C-fiber afferents. For example, it is known that instillation of cold water into the bladder of patients with upper motoneuron lesions induces reflex voiding (the Bors ice water test).24,33,48 This reflex does not occur in normal patients. Recently, it has been shown in the cat that C-fiber bladder afferents are responsible for cold-induced bladder reflexes (see Fig. 14–7). The ice water test is also positive in patients with multiple sclerosis, cerebrovascular disease, Parkinson’s disease, and benign prostatic hypertrophy, as well as in normal infants. These observations suggest that cold-evoked bladder reflexes are mediated by a primitive spinal pathway that is present in the immature nervous system and then is suppressed during postnatal development as supraspinal mechanisms assume the dominant role in controlling micturition. However, when supraspinal controls are eliminated by spinal cord injury or neurologic diseases, such as multiple sclerosis, it appears that the spinal reflexes reemerge. Other reflexes also emerge after spinal cord injury. For example, arterial pressor responses induced by bladder distention (autonomic dysreflexia) occur in spinal cord–injured patients and animals. In normal and spinal cord–injured rats, the increase in blood pressure induced by isometric bladder contractions or distention is suppressed by capsaicin, the C-fiber neurotoxin.39 This suggests that C-fiber afferents mediate bladder vascular reflexes in this species. Similar capsaicin-sensitive afferents may be responsible for autonomic dysreflexia in quadriplegic patients.33

Sex Organs The physiologic changes initiated by erotic stimuli can be divided into four distinct phases (excitement, plateau, orgasm, and resolution) that have been designated collectively the sexual response cycle.87 Although anatomic differences obviously preclude identical responses in male and female during each phase of the cycle, it is clear that similar vascular responses (skin flush, penile and clitoral erection), secretory responses (stimulation of the prostate, bulbourethral gland, and glands of Littre in the male and Bartholin’s and paraurethral glands in the female), and responses of smooth and striated muscles occur in both sexes.4,12,31,45,110,111,115 This section reviews the neural control of the principal autonomic and somatic responses in the male sexual response cycle (erection, secretion, emission, and ejaculation).

INNERVATION Parasympathetic Pathways As described earlier with regard to micturition, three sets of nerves provide an innervation to the urogenital system (see Fig. 14–2). Parasympathetic preganglionic axons from the sacral spinal cord provide the efferent outflow to erectile tissue in the penis and to the seminal vesicles, prostate, and urethral glands (Table 14–3). The sacral pathways have cholinergic ganglionic relay stations in the pelvic plexus and possibly in the effector organs. The postganglionic parasympathetic neurons synthesize and release several transmitters, including NO, ACh, VIP, and ATP. NO is thought to be the major transmitter mediating neurally induced erections,1–4,7,27,29,30,45,59 whereas ACh appears to be important in stimulating

Table 14–3. Male Sexual Reflexes Response

Afferent Source

Efferent Nerves

Central Pathway

Effector Organ

Penile erection Reflexogenic

Pudendal nerve

Sacral parasympathetic

Sacral spinal reflex

Dilation of arterial supply to corpus cavernosum and spongiosum

Auditory, olfactory, visual, imaginative Pudendal nerve

Lumbar sympathetic

Supraspinal origin

Lumbosacral sympathetic and parasympathetic Lumbar sympathetic

Sacral spinal reflex

Psychogenic Glandular secretion Seminal emission

Pudendal nerve

Ejaculation

Pudendal nerve

Somatic efferents in pudendal nerves

Intersegmental spinal reflex (sacrolumbar) Sacral spinal reflex

Seminal vesicles and prostate Contraction of vas deferens, ampulla, prostate and bladder neck closure Rhythmic contractions of bulbo- and ischiocavernous muscles

Autonomic Systems to the Urinary Bladder and Sexual Organs

secretion in glands.25 The functions of VIP and ATP are uncertain.

Sympathetic Pathways The sympathetic innervation of the genital organs consists of PGNs in the thoracolumbar segments of the spinal cord and postganglionic neurons in the paravertebral and prevertebral ganglia (inferior mesenteric and pelvic ganglia), which provide an input to the penile erectile tissue as well as to the smooth muscle of the ductus deferens, seminal vesicles, urethra, and prostate (see Table 14–3).1,3–5,45,77,82,102 Sympathetic postganglionic axons are carried in the pudendal nerve and in nerves arising from the pelvic plexus. Most sympathetic postganglionic neurons are noradrenergic and release norepinephrine, ATP, and neuropeptides such as NPY. These nerves produce constriction of blood vessels and cause a contraction of the vas deferens and urethra– bladder neck.45,57 Some sympathetic nerves arising in the pelvic plexus release NO and presumably induce vasodilation.45,109,111,115

Somatic Pathways The pudendal nerve, arising from the S2 to S4 segments of the spinal cord, provides an efferent excitatory input to the striated muscles (bulbocavernosus and ischiocavernosus) involved in ejaculation.86 The pudendal nerve also contains the principal afferent pathway from the penis (see Table 14–3), and from the clitoris and vagina in the female. Afferent innervation to the uterine cervix and uterus travels in the pelvic and sympathetic nerves.

ERECTION Peripheral Mechanisms Penile erection, which is one of the first responses to occur during sexual arousal, is a vascular phenomenon resulting from a neurally mediated increase in blood flow to the penile erectile tissue (corpora cavernosa and corpus spongiosum).3,27,111 The erectile tissue consists of large venous sinuses that contain very little blood when the penis is flaccid, but distend considerably when blood flow is increased. Dilation in the arterial supply to the cavernous tissue coupled with a relaxation of the sinusoidal smooth muscle in the trabecular tissue (increased compliance) is responsible for erection. During the initiation of erection, corpus blood flow increases 7 to 30 times.111 The neural pathways producing erection arise from both the sacral (parasympathetic) and the thoracolumbar (sympathetic) segments of the spinal cord (see Table 14–3). The postganglionic neurons in these pathways express neuronal

315

NOS (nNOS), and all smooth muscle regions of the penis are richly innervated by nerves containing nNOS.2,27,29,30 The endothelial cells in the penis also express endothelial NOS (eNOS) and can release NO in response to mechanical stimuli (shear stress) associated with changes in blood flow.3,27,29,69 NO directly activates soluble guanylate cyclase in the penile smooth muscle to increase the formation of cyclic GMP (cGMP), which in turn activates cGMPdependent protein kinase I (cGK I).3,66,67,111 Inactivation of cGK I in mice severely reduces reproductive function and markedly reduces the ability of corpus cavernosal tissue to relax in response to neurally, endothelially, or exogenously administered NO.3,67 cGK I is thought to act by multiple mechanisms, including suppression of membrane and sarcoplasmic reticulum Ca2⫹ channels and suppression of inositol trisphosphate (IP3)–induced intracellular Ca2⫹ release. The effects of NO are terminated by the enzymatic breakdown of cGMP by phosphodiesterase. Pharmacologic studies in animals have shown that erections elicited by stimulation of autonomic nerves are reduced by NOS inhibitors and enhanced by phosphodiesterase (PDE) inhibitors.26,27,30 One type of PDE (PDE-5) is highly expressed in penile tissues. Several PDE-5 inhibitors are currently used to treat erectile dysfunction.63 eNOS has also been implicated in neurally mediated erections.3,69 It has been suggested that synthesis of NO in nerves by nNOS initiates erections. However, this response, which is relatively brief, triggers increased blood flow and expansion of the penile vasculature and sinusoidal spaces. The resulting shear force on the endothelium activates a phosphatidylinositol 3-kinase (PI3-kinase) pathway that in turn stimulates the serine/threonine protein kinase Akt, causing direct phosphorylation of eNOS in endothelial cells. After phosphorylation the Ca2⫹ requirement for eNOS activity is reduced, causing increased production of NO, which induces a prolonged penile erection. Intercellular communication via gap junctions is thought to promote the spread of signals throughout the smooth muscle of the penis and amplify the relaxation.3 Exogenous ACh acts on postjunctional muscarinic receptors to elicit a contraction or relaxation of corpus cavernosum preparations. The latter effect is mediated by M3 receptors and release of NO from the endothelium. However, the failure of exogenous ACh to produce erection and the failure of atropine to block erections indicates that cholinergic mechanisms are not essential for neurally mediated increases in penile blood flow.45 However, ACh may act on prejunctional inhibitory receptors on adrenergic nerve terminals to suppress the release of norepinephrine and thereby facilitate NO-mediated erections.111 VIP activates G protein–coupled receptors to stimulate adenylate cyclase and increase cyclic AMP (cAMP), which in turn stimulates cAMP-dependent kinases to suppress contractile mechanisms in vascular smooth muscle and smooth muscle of the trabecular tissue of the penis.1,2,4

316

Function of the Peripheral Nervous System

Although it is clear that exogenous VIP can relax human cavernosal tissue strips in vitro, it has never been shown that VIP released from nerves is responsible for relaxation of penile smooth muscle in vitro or in vivo. The adrenergic innervation of the penis, which provides an excitatory or constrictor input to penile blood vessels, is thought to be involved primarily in detumescence.1,3,4,37,38,45,57,92,93 Electrical stimulation of sympathetic axons in either the hypogastric or pudendal nerves in various animals produces a substantial reduction in penile blood flow.45,110 The effect is blocked by ␣-adrenergic blocking agents. In some animals, a partial erection can be elicited by the administration of ␣-adrenergic blocking agents, suggesting that tonic sympathetic vasoconstrictor mechanisms may have an inhibitory influence on erection.45 Several mechanisms have been implicated in the noradrenergic vasoconstrictor effect, including activation of phospholipase C followed by formation of IP3 and diacylglycerol, which leads to release of intracellular Ca2⫹ as well as sensitization of contractile mechanisms to Ca2⫹.1–3,37,38,93 Ca2⫹ sensitization has been linked to activation of an intracellular signaling pathway involving Rho kinase, which can be stimulated by G protein–coupled receptors such as ␣-adrenergic and endothelin receptors, both of which can elicit contraction of penile smooth muscle. Activation of Rho kinase leads to a change in the phosphorylation state of myosin light chain kinase, resulting in phosphorylation of myosin and subsequent smooth muscle contraction. Administration of a Rho kinase inhibitor triggered a significant increase in intracavernous pressure and an NO-independent erection.38 Similarly, the expression of a dominant negative construct to downregulate Rho kinase enhanced erectile function in rats.37 Other studies have revealed that NO may act to relax penile smooth muscle by suppressing Rho kinase activity.3 These studies have raised the possibility that antagonism of Rho kinase activity may yield new treatments for erectile dysfunction.

Central Mechanisms Penile erection is primarily an involuntary or reflex phenomenon that can be elicited by a variety of stimuli and by at least two distinct central mechanisms (see Table 14–3). For example, psychogenic erections are initiated by supraspinal centers in response to auditory, visual, olfactory, tactile, and imaginative stimuli.4,87,91,111 The efferent limb of the reflex pathway can be in either the thoracolumbar or the sacral autonomic outflow.34,109 Reflexogenic erections, which are initiated by exteroceptive stimulation of the genital regions, are mediated by a sacral spinal reflex mechanism having an afferent limb in the pudendal nerves and an efferent limb in the sacral parasympathetic nerves. Electrophysiologic studies in animals have revealed that the reflex occurs with a short central delay and is not altered by transection of the spinal cord above the lumbar level.124

In patients with lower motor neuron lesions involving the sacral spinal cord, reflexogenic erections are abolished but psychogenic erections may still occur via the sympathetic innervation to the penis.34,45,110 In patients with spinal cord lesions above the level of T12, psychogenic erections are abolished but reflexogenic reactions persist. Under normal conditions it is likely that psychic and reflexogenic stimuli act synergistically in producing erections. It is also known that psychological factors, such as guilt and hostility, or endocrine disturbances that influence libido or supraspinal centers, can interfere with erectile reflexes.12,13,41,62,82,87,112 In addition, any factor that restricts blood flow to erectile tissue will alter erection. Thus atherosclerosis, thrombosis, or a depression of autonomic transmission by either disease or drugs can compromise the vascular responses necessary for tumescence. In adult diabetics, in whom both vascular and neural pathology are common, the incidence of erectile dysfunction is very high.26,62,92

GLANDULAR SECRETION During the second phase of the sexual response cycle (plateau), activity in parasympathetic pathways stimulates mucus secretion from bulbourethral and Littre’s glands and secretion from the seminal vesicles and the prostate gland.25,87,110 Mucus secretion contributes to lubrication of the penis, whereas secretions from the seminal vesicles and prostate provide the bulk of the fluid and chemical factors that contribute to the viability and motility of the spermatozoa. Glandular secretion is thought to be mediated by the parasympathetic system; however, the transmitters have not been identified. Acetylcholine may be involved because cholinomimetic agents stimulate secretion from some glands and AChE-containing nerves have been identified in the prostate and seminal vesicles. However, VIP-containing nerves are also present in these organs. The seminal vesicles and prostate gland are also sensitive to the levels of sex hormones, as evidenced by the stimulation of glandular size and secretion by testosterone and the reverse effects after the administration of estrogens or castration.

EMISSION-EJACULATION The third phase of the sexual act (orgasm), which is accompanied by emission and ejaculation of semen, involves the coordination of autonomic and somatic reflex mechanisms at different levels of the lumbosacral spinal cord. During the first step in this process (emission), reflex activity in the thoracolumbar sympathetic outflow elicits rhythmic contractions of the smooth muscle of the seminal vesicles, prostate, ductus deferens, and ampulla, resulting in the

Autonomic Systems to the Urinary Bladder and Sexual Organs

ejection of sperm and glandular secretions into the urethra and at the same time a closure of the vesical neck to prevent the backflow of semen into the bladder.59,82,102 Pharmacologic studies have shown that these responses are mediated by the adrenergic transmitters norepinephrine and ATP, interacting with ␣-adrenergic receptors and purinergic receptors, respectively.12,76 Thus surgical interruption of sympathetic nerves or the administration of drugs that either block ␣-adrenergic receptors (e.g., tamsulosin), deplete norepinephrine stores, or block norepinephrine release (e.g., guanethidine) blocks emission.5,76 Seminal emission may also be modulated by activity in cholinergic nerves. For example, physostigmine, an anticholinesterase agent, or exogenous ACh depresses the responses of the ductus deferens and other sex organs to electrical stimulation of the sympathetic nerves. The depression, which is accompanied by a reduction in norepinephrine release, is blocked by atropine. These findings suggest that one action of the cholinergic innervation to the sex organs is a muscarinic suppression of excitatory adrenergic transmission. After emission of semen into the proximal urethra, rhythmic contractions of the bulbocavernosus, ischiocavernous, and paraurethral striated muscles result in ejaculation. The afferent and efferent limbs of the ejaculation reflex are contained in the pudendal nerve (see Table 14–3). The sensations accompanying ejaculation constitute the orgasm.102 Orgasm is not necessarily affected by sympathectomy, provided that the pudendal nerves remain intact. Thus neither efferent fibers in the sympathetic nerves nor contractions of smooth muscles of the seminal vesicles and ductus deferens are essential for the occurrence of orgasm.

CENTRAL REFLEX PATHWAYS In the spinal cord, ascending fibers carrying sensory information from the sex organs seem to lie primarily within the anterolateral tracts. Thus bilateral cordotomy usually diminishes or completely abolishes orgasm. Cordotomy also severely compromises ejaculatory and erectile mechanisms, although the latter may recover over a period of time. With more extensive damage of the spinal cord in paraplegics, when the site of injury is located rostral to T12, ejaculation occurs in a relatively small percentage of patients in comparison to reflexogenic erections, which are readily elicited.34,35,64 This observation is consistent with the greater complexity of spinal reflex pathways underlying ejaculation and indicates a considerable dependence of these pathways on supraspinal coordinating mechanisms. Brain imaging studies using fMRI or PET during sexual arousal induced by erotic visual stimuli have identified several brain regions in both males and females that are activated during sexual stimulation.8,100,105,107,125 PET studies in males detected activation in three general regions:

317

limbic/paralimbic areas (anterior cingulate gyrus, orbitofrontal cortex), the striatum (head of the caudate nucleus), and the posterior hypothalamus. Strong signals specifically associated with penile turgidity were observed in the right subinsular region (claustrum, caudate, cingulate gyrus) using fMRI. A PET study also showed increased blood flow in the right prefrontal cortex during orgasm in males, whereas all other cortical areas showed decreases in blood flow.130 It has been suggested that the activation of paralimbic areas is correlated with the emotional and motivational states associated with sexual arousal, whereas the activation of the anterior cingulate and hypothalamic regions is related to the affective, autonomic, and endocrine responses.8 More recent fMRI studies in males have also detected activation in parietal areas known to be involved in attentional processes.100 The human data correlate well with results from animal experiments, which indicate that connections between the limbic system and the hypothalamus (medial preoptic area [MPOA] and paraventricular nucleus) are essential for the stimulation of the descending autonomic projections to the spinal cord that trigger psychogenic penile erections.45,91 Recent studies in rats have identified a population of lumbar spinal neurons located around the central canal in lamina X and medial lamina VII that participate in the ejaculatory reflex.133 These neurons express NK-1 receptors, receive afferent input from the penis, and project to the thalamus. It was initially hypothesized that these neurons might function primarily in transmitting sensory information from the penis to pleasure centers in the brain. However, it was discovered that the neurons were only activated during ejaculation and not during other components of male sexual behavior. When the neurons were destroyed by intrathecal administration of a toxin (saporin-SSP) that selectively destroys neurons expressing NK-1 receptors, ejaculatory responses were completely abolished without altering other components of copulatory behavior (penile erection, mounting, intromission). It was concluded that the neurons may be part of a spinal ejaculation generator. Pharmacologic studies in animals have implicated many neurotransmitter systems in the central control of sexual function.3,7,26,45,91 The dopaminergic system in the MPOA and the oxytocin and NO systems in the paraventricular nucleus exert major excitatory effects on erection, whereas the serotonergic pathways arising in the raphe nuclei seem to be inhibitory. Elsewhere in the brain noradrenergic, GABAergic, and serotonergic pathways are generally inhibitory, whereas dopaminergic, NO, and oxytocin pathways facilitate sexual function. NO appears to tonically modulate the hypothalamic excitatory circuitry because intracerebroventricular injections of NOS inhibitors prevented the penile erectile responses induced by dopamine agonists and oxytocin.113 It is uncertain whether all of these findings in animals are entirely applicable to man.

318

Function of the Peripheral Nervous System

However, recently apomorphine, a nonselective dopamine receptor agonist that stimulates sexual behavior and penile erection in animals, has been marketed to treat penile erectile dysfunction in patients.26

17.

18.

REFERENCES

19.

1. Andersson, K. E.: Pharmacology of lower urinary tract smooth muscle and penile erection tissues. Pharmacol. Rev. 45:253, 1993. 2. Andersson, K. E.: Pharmacology of penile erection. Pharmacol. Rev. 53:417, 2001. 3. Andersson, K. E.: Erectile physiological and pathophysiological pathways involved in erectile dysfunction. J. Urol. 170:S6, 2003. 4. Andersson, K. E., and Wagner, G.: Physiology of penile erection. Physiol. Rev. 75:191, 1995. 5. Andersson, K.-E., and Wyllie, M. G.: Ejaculatory dysfunction: why all ␣-blockers are not equal. BJU Int. 92:876, 2003. 6. Araki, I., and de Groat, W. C.: Synaptic modulation associated with developmental reorganization of visceral reflex pathways. J. Neurosci. 17:8402, 1997. 7. Argiolas, A., and Melis, M.: Neuromodulation of penile erection: an overview of the role of neurotransmitters and neuropeptides. Prog. Neurobiol. 47:235, 1995. 8. Arrow, B. A., Desmond, J. E., Banner, L. L., et al.: Brain activation and sexual arousal in healthy, heterosexual males. Brain 125:1014, 2002. 9. Athwal, B. S., Berkley, K. J., Brennan, A., et al.: Brain activity associated with the urge to void and bladder fill volume in normal men: preliminary data from a PET study. BJU Int. 84:148, 1999. 10. Athwal, B. S., Berkley, K. J., Hussain, I., et al.: Brain responses to changes in bladder volume and urge to void in healthy men. Brain 124:369, 2001. 11. Bahns, E., Ernsberger, U., Janig, W., and Nelke, A.: Functional characteristics of lumbar visceral afferent fibers from the urinary bladder and urethra in the cat. Eur. J. Physiol. 407:510, 1998. 12. Beck, R. O.: Physiology of male sexual function and effect of neurologic disease. In Fowler, C. J. (ed.): Neurology of Bladder, Bowel, and Sexual Dysfunction. Boston, Butterworth-Heinemann, p. 47, 1999. 13. Betts, C. D.: Bladder and sexual function in multiple sclerosis. In Fowler, C. J. (ed.): Neurology of Bladder, Bowel, and Sexual Dysfunction. Boston, Butterworth-Heinemann, p. 289, 1999. 14. Birder, L. A., Apodaca, G., de Groat, W. C., and Kanai, A. J.: Adrenergic and capsaicin evoked nitric oxide release from urothelium and afferent nerves in urinary bladder. Am. J. Physiol. 275:F226, 1998. 15. Birder, L. A., Barrick, S., Roppolo, J. R., et al.: Feline interstitial cystitis results in mechanical hypersensitivity and altered ATP release from bladder urothelium. Am. J. Physiol. 285:F423, 2003. 16. Birder, L. A., Kanai, A. J., de Groat, W. C., et al.: Vanilloid receptor expression suggests a sensory role for urinary blad-

20.

21.

22.

23.

24.

25.

26. 27. 28.

29.

30.

31.

32.

33.

34.

35.

der epithelial cells. Proc. Natl. Acad. Sci. U.S.A. 98:13396, 2001. Birder, L. A., Nakamura, Y., Kiss, S., et al.: Altered urinary bladder function in mice lacking the vanilloid receptor TRPV1. Nature Neurosci. 5:856, 2002. Birder, L. A., Nealon, M., Kiss, S., et al.: ␤-Adrenoceptor agonists stimulate endothelial nitric oxide synthase in rat urinary bladder urothelial cells. J. Neurosci. 22:8063, 2002. Blok, B. F. M.: Brain control of the lower urinary tract. Scand. J. Urol. Nephrol. 36(Suppl. 210):11, 2002. Blok B. F. M., Sturms, L. M., and Holstege, G.: A PET study on cortical and subcortical control of the pelvic floor musculature in women. J. Comp. Neurol. 389:535, 1997. Blok B. F. M., Sturms, L. M., and Holstege, G.: Brain activity during micturition in women. Brain 121:2033, 1998. Blok, B. F. M., Willemsen, A. T. M., and Holstege, G.: A PET study on the brain control of micturition in humans. Brain 120:111, 1997. Booth, A. M., and de Groat, W. C.: A study of facilitation in vesical parasympathetic ganglia of the cat using intracellular recording techniques. Brain Res. 169:388, 1979. Bors, E., and Comarr, A. E.: Neurological Urology: Physiology of Micturition, Its Neurological Disorders and Sequelae. Baltimore, University Park Press, 1971. Bruschini, H., Schmidt, R. A., and Tanagho, E. A.: Neurologic control of prostatic secretion in the dog. Invest. Urol. 15:288, 1978. Burnett, A. L.: Novel pharmacological approaches in the treatment of erectile dysfunction. World J. Urol. 19:57, 2001. Burnett, A. L.: Nitric oxide regulation of penile erection: biology and therapeutic implications. J. Androl. 23:S20, 2002. Burnett, A. L., Calvin, D. C., Chamness, S. L., et al.: Urinary bladder-urethral sphincter dysfunction in mice with targeted disruption of neuronal nitric oxide synthase models idiopathic voiding disorders in humans. Nature Med. 3:571, 1997. Burnett, A. L., Chang, A. G., Crone, J. K., et al.: Noncholinergic penile erection in mice lacking the gene for endothelial nitric oxide synthase. J. Androl. 23:92, 2002. Burnett, A. L., Lowenstein, C. J., Bredt, D. S., et al.: Nitric oxide: a physiologic mediator of penile erection. Science 257:401, 1992. Burnett, A. L., and Truss, M. C.: Mediators of the female sexual response: pharmacotherapeutic implications. World J. Urol. 20:101, 2002. Burnstock, G.: Purinergic signaling in the lower urinary tract. In Abbracchio, M. P., and Williams, M. (eds.): Handbook of Experimental Pharmacology. Berlin, Springer Verlag, p. 423, 2001. Chancellor, M. B., and Yoshimura, N.: Physiology and pharmacology of the bladder and urethra. In Walsh, P. C., Retik, A. B., Vaughn, E. D., and Wein, A. J. (eds.): Campbell’s Urology. Philadelphia, W. B. Saunders, p. 831, 2002. Chapelle, P. A., Durand, J., and Lacert, P.: Penile erection following complete spinal cord injury in man. Br. J. Urol. 52:216, 1980. Chapelle, P. A., Roby-Brami, A. Yakovleff, A., and Bussel, B.: Neurological correlations of ejaculation and testicular size

Autonomic Systems to the Urinary Bladder and Sexual Organs

36.

37.

38.

39.

40.

41. 42. 43. 44.

45.

46.

47.

48.

49.

50.

51.

52.

in men with a complete spinal cord section. J. Neurol. Neurosurg. Psychiatry 51:197, 1988. Cheng, C-L., Liu, J-C., Chang, S-Y., et al.: Effect of capsaicin on the micturition reflex in normal and chronic spinal cats. Am. J. Physiol. 277:R786, 1999. Chitaley, K., Bivalacqua, T. J., Champion, H. C., et al.: Adeno-associated viral gene transfer of dominant negative RhoA enhances erectile function in rats. Biochem. Biophys. Res. Commun. 298:427, 2002. Chitaley, K., Webb, R. C., and Mills, T.: RhoA/Rho-kinase: a novel player in the regulation of penile erection. Int. J. Impot. Res. 13:67, 2001. Chuang, Y. C., Fraser, M. O., Yu, Y. B., et al.: Analysis of the afferent limb of the vesicovascular reflex using the neurotoxins, resiniferatoxin and capsaicin. Am. J. Physiol. Reg. Integ. Comp. Physiol. 281:R1302, 2001. Cockayne, D. A., Hamilton, S. G., Zhu, Q. M., et al.: Urinary bladder hyporeflexia and reduced pain behaviour in P2X3 deficient mice. Nature 407:1011, 2000. DasGupta, R., and Fowler, C. J.: Bladder, bowel, and sexual dysfunction in multiple sclerosis. Drugs 63:153, 2003. de Groat, W. C.: Spinal cord projections and neuropeptides in visceral afferent neurons. Prog. Brain Res. 67:165, 1986. de Groat, W. C.: Neuropeptides in pelvic afferent pathways. Experientia 43:801, 1987. de Groat, W. C., Araki, I., Vizzard, M. A., et al.: Developmental and injury induced plasticity in the micturition reflex pathway. Behav. Brain Res. 92:127, 1998. de Groat, W. C., and Booth, A. M.: Neural control of penile erection. In Maggi, C. A. (ed.): The Autonomic Nervous System, Vol. 3. Nervous Control of the Urogenital System. London, Harwood Academic Publishers, p. 467, 1993. de Groat, W. C., and Booth, A. M.: Synaptic transmission in pelvic ganglia. In Maggi, C. A. (ed.): The Autonomic Nervous System, Vol. 3. Nervous Control of the Urogenital System. London, Harwood Academic Publishers, p. 291, 1993. de Groat, W. C., Booth, A. M., Milne, R. J., and Roppolo, J. R.: Parasympathetic preganglionic neurons in the sacral spinal cord. J. Auton. Nerv. Syst. 5:23, 1982. de Groat, W. C., Booth, A. M., and Yoshimura, N.: Neurophysiology of micturition and its modification in animal models of human disease. In Maggi, C. A. (ed.): The Autonomic Nervous System, Vol. 3. Nervous Control of the Urogenital System. London, Harwood Academic Publishers, p. 227, 1993. de Groat, W. C., and Kawatani, M.: Enkephalinergic inhibition in parasympathetic ganglia of the urinary bladder of the cat. J. Physiol. (Lond.) 413:13, 1989. de Groat, W. C., Kawatani, M., Hisamitsu, T., et al.: Mechanisms underlying the recovery of urinary bladder function following spinal cord injury. J. Auton. Nerv. Syst. 30:S71, 1990. de Groat, W. C., and Lalley, P. M.: Reflex firing in the lumbar sympathetic outflow to activation of vesical afferent fibers. J. Physiol. (Lond.) 226:289, 1972. de Groat, W. C., Nadelhaft, I., Milne, R. J., et al.: Organization of the sacral parasympathetic reflex pathways to the urinary bladder and large intestine. J. Auton. Nerv. Syst. 3:135, 1981.

319

53. de Groat, W. C., and Saum, W. R.: Sympathetic inhibition of the urinary bladder and of pelvic ganglionic transmission in the cat. J. Physiol. (Lond.) 220:297, 1972. 54. de Groat, W. C., and Saum, W. R.: Synaptic transmission in parasympathetic ganglia in the urinary bladder of the cat. J. Physiol. (Lond.) 256:137, 1976. 55. de Groat, W. C., and Yoshimura, N.: Pharmacology of the lower urinary tract. Annu. Rev. Pharmacol. Toxicol. 41:691, 2001. 56. Delancey, J., Gosling, J., Creed, K., et al.: Gross anatomy and cell biology of the lower urinary tract. In Abrams, P., Cardozo, L., Khoury, S., and Wein, A. (eds.): Incontinence. Paris, Health Publications, Ltd., p. 17, 2002. 57. Diederichs, W., Stief, C. G., Lue, T. F., and Tanagho, E. A.: Norepinephrine involvement in penile detumescence. J. Urol. 143:1264, 1990. 58. Drake, M. J., Hedlund, P., Andersson, K.-E., et al.: Morphology, phenotype, and ultrastructure of fibroblastic cells from normal and neuropathic human detrusor: absence of myofibroblast characteristics. J. Urol. 169:1573, 2003. 59. Elbadawi, A., and Goodman, D. C.: Autonomic innervation of accessory male genital glands. In Spring-Mills, E., and Hafez, E. S. E. (eds.): Male Accessory Sex Glands. Amsterdam, Elsevier, p. 101, 1980. 60. Ferguson, D. R., Kennedy, I., and Burton, T. J.: ATP is released from rabbit urinary bladder cells by hydrostatic pressure changes—a possible sensory mechanism? J. Physiol. (Lond.) 505:503, 1997. 61. Fowler, C. J., Jewekes, D., McDonald, W. I., et al.: Intravesical capsaicin for neurogenic bladder dysfunction. Lancet 339:1239, 1992. 62. Goldstein, I., Graziottin, J. R., Heiman, J. R., et al.: Sexual dysfunction in females. In Jardin, A., Wagner, G., Khoury, S., et al. (eds.): Erectile Dysfunction. Paris, Health Publications, Ltd., pp. 507–556, 2000. 63. Goldstein, I., Lue, T. F., Padma-Nathan, H., et al.: Oral sildenafil in the treatment of erectile dysfunction. Sildenafil Study Group. N. Engl. J. Med. 338:1397, 1998. 64. Grossiord, A., Chapelle, P. A., Lacert, P. H., et al.: Le segment medullare lesionnel chez le paraplegique. Rev. Neurol. (Paris) 134:729, 1978. 65. Habler, H. J., Janig, W., and Koltzenberg, M.: Activation of unmyelinated afferent fibers by mechanical stimuli and inflammation of the urinary bladder. J. Physiol. (Lond.) 425:545, 1990. 66. Hashitani, H., Fukuta, H., Dickens, E. J., and Suzuki, H.: Cellular mechanisms of nitric oxide-induced relaxation of corporeal muscle in the guinea pig. J. Physiol. (Lond.) 538:573, 2002. 67. Hedlund, P., Aszodi, A., Pfeifer, A., et al.: Erectile dysfunction in cyclic GMP-dependent kinase I-deficient mice. Proc. Nat. Acad. Sci. U.S.A. 97:2349, 2000. 68. Ho, K. M., Ny, L., McMurray, G., et al.: Co-localization of carbon monoxide and nitric oxide synthesizing enzymes in the human urethral sphincter. J. Urol. 161:1968, 1999. 69. Hurt, J. K., Musicki, B., Palese, M. A., et al.: Akt-dependent phosphorylation of endothelial nitric-oxide synthase mediates penile erection. Proc. Natl. Acad. Sci. U.S.A. 99:1461, 2002.

320

Function of the Peripheral Nervous System

70. Jankowski, R. J., Vorp, D. A., Prantil, R. L., et al.: Development of a system for studying urethral biomechanical function. Am. J. Physiol. 286:F225, 2004. 71. Kakizaki, H., Fraser, M. O., and de Groat, W. C.: Reflex pathways controlling urethral striated and smooth muscle function in the male rat. Am. J. Physiol. 272:R1647, 1997. 72. Kamo, I., Torimoto, K., Chancellor, M. B., et al.: Urethral closure mechanisms under sneeze-induced stress condition in rats: a new animal model for evaluation of stress urinary incontinence. Am. J. Physiol. 285:R356, 2003. 73. Kawatani, M., Shioda, S., Nakai, Y., et al.: Ultrastructural analysis of enkephalinergic terminals in parasympathetic ganglia innervating the urinary bladder of the cat. J. Comp. Neurol. 288:81, 1989. 74. Keast, J. R., and de Groat, W. C.: Segmental distribution and peptide content of primary afferent neurons innervating the urogenital organs and colon of male rats. J. Comp. Neurol. 319:615, 1992. 75. Keast, J., Kawatani, M., and de Groat, W. C.: Sympathetic modulation of cholinergic transmission in cat vesical ganglia is mediated by ␣1 and ␣2 adrenoceptors. Am. J. Physiol. 258:R44, 1990. 76. Kedia, K., and Markland, C.: The effect of pharmacological agents on ejaculation. J. Urol. 114:569, 1975. 77. Kihara, K., and de Groat, W. C.: Sympathetic efferent pathways projecting bilaterally to the vas deferens in the rat. Anat. Rec. 248:291, 1997. 78. Kruse, M. N., Bray, L. A., and de Groat, W. C.: Influence of spinal cord injury on the morphology of bladder afferent and efferent neurons. J. Auton. Nerv. Syst. 54:215, 1995. 79. Lavelle, J. P., Meyers, S., Ruiz, G., et al.: Urothelial pathophysiological changes in feline interstitial cystitis: a human model. Am. J. Physiol. 278: F540, 2000. 80. Lee, H. Y., Bardini, M., and Burnstock, G.: Distribution of P2X3 in the urinary bladder and ureter in the rat. J. Urol. 163:2002, 2000. 81. Lewis, S. A.: Everything you wanted to know about the bladder epithelium but were afraid to ask. Am. J. Physiol. 278:F867, 2000. 82. Lundberg, P. O., Brockett, N. L., Denys, P., et al.: Neurological disorders: erectile and ejaculatory dysfunction. In Jardin, A., Wagner, G., Khoury, S., et al. (eds.): Erectile Dysfunction. Paris, Health Publications, Ltd., p. 591, 2000. 83. Maggi, C. A.: The dual sensory and efferent functions of the capsaicin-sensitive primary sensory neurons in the urinary bladder and urethra. In Maggi, C. A. (ed.): The Autonomic Nervous System, Vol. 3. Nervous Control of the Urogenital System. London, Harwood Academic Publishers, p. 383, 1993. 84. Mallory, B., Steers, W. D., and de Groat, W. C.: Electrophysiological study of micturition reflexes in the rat. Am. J. Physiol. 257:R410, 1989. 85. Mallory, B. S., Roppolo, J. R., and de Groat, W. C.: Pharmacological modulation of the pontine micturition center. Brain Res. 546:310, 1991. 86. Marson, L., and McKenna K.: CNS cell groups involved in the control of the ischiocavernous and bulbospongiosus muscles: a transneural tracing study using pseudorabies virus. J. Comp. Neurol. 374:161, 1996.

87. Masters, W. H., and Johnson, V. E.: Human Sexual Response. Boston, Little Brown & Co., 1966. 88. Matsui, M., Motomura, D., Fujikawa, T., et al.: Mice lacking M2 and M3 muscarinic acetylcholine receptors are devoid of cholinergic smooth muscle contractions but still viable. J. Neurosci. 22:10627, 2002. 89. Matsui, M., Motomura, D., Karasawa, H., et al.: Multiple functional defects in peripheral autonomic organs in mice lacking muscarinic acetylcholine receptor gene for the M3 subtype. Proc. Nat. Acad. Sci. U.S.A. 97:9579, 2000. 90. Matsuura, S., Kakizaki, H., Mitsui, T., et al.: Human brain region response to distension or cold stimulation of the bladder: a positron emission tomography study. J. Urol. 168:2035, 2002. 91. McKenna, K.: Central control of penile erection. Int. J. Impot. Res. 10(Suppl.):S25, 1998. 92. Melman, A., Henry, D. P., Felten, D. F., and O’Connor, B. L.: Effect of diabetes upon penile sympathetic nerves in impotent patients. South. Med. J. 73:307, 1980. 93. Mills, T. M., Chitaley, K., and Lewis, R. W.: Vasoconstrictors in erectile physiology. Int. J. Impot. Res. 5(Suppl. 13):S29, 2001. 94. Miura, A., Kawatani, M., and de Groat, W. C.: Excitatory synaptic currents in lumbosacral parasympathetic preganglionic neurons elicited from the lateral funiculus. J. Neurophysiol. 86:1587, 2001. 95. Miura, A., Kawatani, M., and de Groat, W. C.: Excitatory synaptic currents in lumbosacral parasympathetic preganglionic neurons evoked by stimulation of the dorsal commissure. J. Neurophysiol. 89:382, 2003. 96. Morgan, C., Nadelhaft, I., and de Groat, W. C.: Location of bladder preganglionic neurons within the sacral parasympathetic nucleus of the cat. Neurosci. Lett. 14:189, 1979. 97. Morgan, C., Nadelhaft, I., and de Groat, W. C.: The distribution of visceral primary afferents from the pelvic nerve within Lissauer’s tract and the spinal gray matter and its relationship to the sacral parasympathetic nucleus. J. Comp. Neurol. 201:415, 1981. 98. Morrison, J., Steers, W. D., Brading, A., et al.: Neurophysiology and neuropharmacology. In: Abrams, P., Cardozo, L., Khoury, S., and Wein, A. (eds.): Incontinence. Paris, Health Publications, Ltd., p. 83, 2002. 99. Morrison, J., Wen, J., and Kibble, A.: Activation of pelvic afferent nerves from the rat bladder during filling. Scand. J. Urol. Nephrol. Suppl. 201:73, 1999. 100. Mouras, H., Stoleru, S., Bittoun, J., et al.: Brain processing of visual sexual stimuli in healthy men: a functional magnetic resonance imaging study. Neuroimage 20:855, 2003. 101. Nadelhaft, I., and Vera, P. L.: Central nervous system neurons infected by pseudorabies virus injected into the rat urinary bladder following unilateral transection of the pelvic nerve. J. Comp. Neurol. 359:443, 1995. 102. Newman, H. F., Reiss, H., and Northup, J. C.: Physical basis of emission, ejaculation, and orgasm in the male. Urology 19:341, 1982. 103. O’Reilly, B. A., Kosaka, A. H., Knight, G. F., et al.: P2X receptors and their role in female idiopathic detrusor instability. J. Urol. 167:157, 2002. 104. Palea, S., Artibani, W., Ostardo, E., et al.: Evidence for purinergic neurotransmission in the human urinary bladder affected by interstitial cystitis. J. Urol. 150:2007, 1993.

Autonomic Systems to the Urinary Bladder and Sexual Organs 105. Park, K., Kang, H. K., Seo, J. J., et al.: Blood-oxygenationlevel-dependent functional magnetic resonance imaging for evaluating cerebral regions of female sexual arousal response. Urology 57:1189, 2001. 106. Ralevic, V., and Burnstock, G.: Receptors for purines and pyrimidines. Physiol. Rev. 50:413, 1998. 107. Redoute, J., Stoleru, S., Gregoire, M. C., et al.: Brain processing of visual sexual stimuli in human males. Human Brain Mapping 11:162, 2000. 108. Rong, W., Spyer, K. M., and Burnstock, G.: Activation and sensitization of low and high threshold afferent fibres mediated by P2X receptors in the mouse urinary bladder. J. Physiol. (Lond.) 541:591, 2002. 109. Root, W. S., and Bard, P.: The mediation of feline erection through sympathetic pathways with some remarks on sexual behavior after deafferentation of the genitalia. Am. J. Physiol. 151:80, 1947. 110. Sachs, B. D., and Meisel, R. L.: The physiology of male sexual behavior. In Knobil, E., and Neill, J. (eds.): The Physiology of Reproduction. New York, Raven, p. 1393, 1988. 111. Saenz de Tejada, I., Cadavid, N. G., Heaton, J., et al.: Anatomy, physiology, and pathophysiology of erectile function. In Jardin, A., Wagner, G., Khoury, S., et al. (eds.): Erectile Dysfunction. Paris, Health Publications, Ltd., p. 65, 2000. 112. Sakakibara, R., and Fowler, C. J.: Cerebral control of bladder, bowel, and sexual function and effects of brain disease. In Fowler, C. J. (ed.): Neurology of Bladder, Bowel, and Sexual Dysfunction. Boston, Butterworth-Heinemann, p. 229, 1999. 113. Sato, Y., Zhao, W., and Christ, G. J.: Central modulation of the NO/cGMP pathway affects the MPOA-induced intracavernous pressure response. Am. J. Physiol. Reg. Integ. Comp. Physiol. 281:R269, 2001. 114. Seki, S., Sasaki, K., Fraser, M. O., et al.: Immunoneutralization of nerve growth factor in the lumbosacral spinal cord reduces bladder hyperreflexia in spinal cord injured rats. J. Urol. 168:2269, 2002. 115. Semans, J. H., and Langworthy, O. R.: Observations on the neurophysiology of sexual function in the male cat. J. Urol. 40:836, 1938. 116. Sengupta, J. N., and Gebhart, G. F.: Mechanosensitive properties of pelvic afferent nerve fibers innervating the urinary bladder of the rat. J. Neurophysiol. 72:2420, 1994. 117. Shea, V. K., Cai, R., Crepps, B., et al.: Sensory fibers of the pelvic nerve innervating the rat’s bladder. J. Neurophysiol. 84:1924, 2000. 118. Somogyi, G. T., Tanowitz, M., and de Groat, W. C.: Prejunctional facilitatory ␣1-adrenoceptors in the rat urinary bladder. Br. J. Pharmacol. 114:1710, 1995. 119. Somogyi, G. T., Tanowitz, M., Zernova, G., and de Groat, W. C.: M1 muscarinic receptor facilitation of ACh and noradrenaline release in the rat urinary bladder is mediated by protein kinase C. J. Physiol. (Lond.) 496:245, 1996. 120. Somogyi, G. T., Zernova, G. V., Tanowitz, M., and de Groat, W. C.: Role of L and N type Ca⫹2 channels in muscarinic receptor mediated facilitation of ACh and noradrenaline release in the rat urinary bladder. J. Physiol. (Lond.) 499:645, 1997. 121. Somogyi, G. T., Zernova, G. V., Yoshiyama, M., et al.: Frequency dependence of muscarinic facilitation of trans-

122.

123.

124.

125.

126.

127.

128.

129. 130.

131. 132.

133.

134.

135.

136.

137.

138.

321

mitter release in urinary bladder strips from neurally intact or chronic spinal cord transected rats. Br. J. Pharmacol. 125:241, 1998. Steers, W. D., Ciambotti, J., Erdman, S., and de Groat, W. C.: Morphological plasticity in efferent pathways to the urinary bladder of the rat following urethral obstruction. J. Neurosci. 10:1943, 1990. Steers, W. D., Ciambotti, J., Etzel, B., et al.: Alterations in afferent pathways from the urinary bladder of the rat in response to partial urethral obstruction. J. Comp. Neurol. 310:1, 1991. Steers, W. D., Mallory, B., and de Groat, W. C.: Electrophysiological study of neural activity in penile nerve of the rat. Am. J. Physiol. 254:R989, 1988. Stoleru, S., Gregoire, M. C., Gerard, D., et al.: Neuroanatomical correlates of visually evoked sexual arousal in human males. Arch. Sexual Behav. 28:1, 1999. Sugaya, K., Roppolo, J. R., Yoshimura, N., et al.: The central neural pathways involved in micturition in the neonatal rats as revealed by the injection of pseudorabies virus into the bladder. Neurosci. Lett. 223:197, 1997. Sui, G. P., Wu, C., and Fry, C. H.: Electrical characteristics of suburothelial cells isolated from the human bladder. J. Urol. 171:938, 2004. Sun, Y., Keay, S., de Deyne, P. G., and Chai, T. C.: Augmented stretch activated adenosine triphosphate release from bladder uroepithelial cells in patients with interstitial cystitis. J. Urol. 166:1951, 2001. Tanagho, E. A., and Miller, E. R.: Initiation of voiding. Br. J. Urol. 42:175, 1970. Tiihonen, J., Kuikka, J., Kupila, J., et al.: Increase in cerebral blood flow of right prefrontal cortex in man during orgasm. Neurosci. Lett. 170:241, 1994. Torrens, M., and Morrison, J. F. B.: The Physiology of the Lower Urinary Tract. Berlin, Springer-Verlag, 1986. Tran, L. V., Somogyi, G. T., and de Groat, W. C.: Inhibitory effects of neuropeptide Y on cholinergic and adrenergic transmission in the rat urinary bladder and urethra. Am. J. Physiol. 266:R1411, 1994. Truitt, W. A., and Coolen, L. M.: Identification of a potential ejaculation generator in the spinal cord. Science 297:1566, 2002. Truschel, S. T., Wang, E., Ruiz, G., et al.: Stretch-related exocytosis/endocytosis in bladder umbrella cells. Mol. Biol. Cell 13:830, 2002. Vizzard, M. A., Erickson, V. L., Card, J. P., et al.: Transneuronal labeling of neurons in the adult rat brain and spinal cord after injection of pseudorabies virus into the urethra. J. Comp. Neurol. 355:629, 1995. Vlaskovoska, M., Kasakov, L., Rong, W., et al.: P2X3 knockout mice reveal a major sensory role for urothelially released ATP. J. Neurosci. 21:5670, 2001. Wein, A. J.: Pathophysiology and categorization of voiding dysfunction. In Walsh, P. C., Retik, A. B., Vaughn, E. D., and Wein, A. J. (eds.): Campbell’s Urology. Philadelphia: W. B. Saunders, p. 887, 2002. Yokoyama, O., Yoshimura, M., Namiki, M., and de Groat, W. C.: Role of the forebrain in bladder hyperactivity following cerebral infarction in the rat. Exp. Neurol. 163:469, 2000.

322

Function of the Peripheral Nervous System

139. Yokoyama, O., Yoshiyama, M., Namiki, M., and de Groat, W. C.: Changes in dopaminergic and glutamatergic excitatory mechanisms of the micturition reflex after middle cerebral artery occlusion in conscious rats. Exp. Neurol. 173:129, 2002. 140. Yoshimura, N., and de Groat, W. C.: Plasticity of Na⫹ channels in afferent neurons innervating rat urinary bladder following spinal cord injury. J. Physiol. (Lond.) 503:269, 1997. 141. Yoshimura, N., and de Groat, W. C.: Increased excitability of afferent neurons innervating rat urinary bladder following chronic bladder inflammation. J. Neurosci. 19:4644, 1999. 142. Yoshimura, N., Kuno, S., Chancellor, M. B., et al.: Dopaminergic mechanisms underlying bladder hyperactivity in rats with a unilateral 6-hydroxydopamine (6-OHDA) lesion of the nigrostriatal pathway. Br. J. Pharmacol. 139:1425, 2003. 143. Yoshimura, N., Mizuta, E., Yoshida, O., and Kuno, S.: Therapeutic effects of dopamine D1/D2 receptor

144.

145.

146.

147.

agonists on detrusor hyperreflexia in 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine-lesioned parkinsonian monkeys. J. Pharmacol. Exp. Ther. 286:228, 1998. Yoshimura, N., Seki, S., Erickson, K. A., et al.: Histological and electrical properties of rat dorsal root ganglion neurons innervating the lower urinary tract. J. Neurosci. 23:4355, 2003. Yoshimura, N., Seki, S., Novakovic, S. D., et al.: The role of the tetrodotoxin-resistant sodium channel Nav1.8 (PN3/SNS) in a rat model of visceral pain. J. Neurosci. 21:8690, 2001. Yoshimura, N., White, G., Weight, F. F., and de Groat, W. C.: Different types of Na⫹ and A-type K⫹ currents in dorsal root ganglion neurons innervating the rat urinary bladder. J. Physiol. (Lond.) 494:1, 1996. Zoubek, J. A., Somogyi, G. T., and de Groat, W. C.: A comparison of inhibitory effects of neuropeptide Y on rat urinary bladder, urethra and vas deferens. Am. J. Physiol. 265:R537, 1993.

15 Sympathetic Nerves and Control of Blood Vessels to Human Limbs MICHAEL J. JOYNER, PAUL M. VANHOUTTE, JOHN R. HALLIWILL, AND JOHN T. SHEPHERD

Baroreflexes Direct Measurement of Sympathetic Traffic in Humans via Microneurography Characteristics of Peripheral Sympathetic Traffic in Humans Baroreceptor Control of MSNA MSNA Responses to Orthostatic Challenge MSNA Responses to Chemoreceptor Activation Central Chemoreflexes and Autonomic Function

MSNA and Autonomic Responses to Exercise Sensory Feedback from Active Muscles Baroreflex Resetting during Exercise Interactions between Sympathetic Vasoconstriction and Metabolic Vasodilatation ␣-Receptor Subtypes and Sympathetic Control of Blood Flow in Active Muscles Exercise: Summary

Arterial blood pressure is the key regulated variable in the cardiovascular system, and the autonomic nervous system has a central role in this regulation. In response to a variety of behavioral situations, blood pressure can rise (e.g., exercise) or fall (sleep), but for each situation arterial pressure is usually maintained in a fairly narrow range. Central to the regulation of arterial pressure is the balance that is maintained between local factors such as metabolites from exercising muscle that can cause vasodilatation and the actions of the sympathetic vasoconstrictor nerves.77 In some species there are sympathetic dilator nerves to skeletal muscle, but their existence in humans is now doubtful.35 Additionally, during increases in core temperature, sympathetic vasodilator nerves to the skin evoke marked cutaneous vasodilatation.53 In previous editions of Peripheral Neuropathy, detailed structural, pharmacologic, and biochemical information concerning the autonomic nerves has been presented.77 In this chapter, we focus on how the sympathetic nerves innervating the blood vessels of the human limbs respond to common stimuli.

Autonomic and MSNA Responses to Mental Stress Neurally Mediated Dilatation in Human Skeletal Muscle Aging, Hypertension, and Congestive Heart Failure Autonomic Responses to Body Heating Cutaneous Vascular Responses to Changes in Temperature Summary

BAROREFLEXES For the autonomic nervous system to regulate arterial pressure, the brain must receive information about the level of the arterial pressure. This is accomplished primarily through sensory nerves located in the carotid sinus and aortic arch. Mechanosensitive afferent nerves in the vessel walls at these locations respond to physical deformation in the blood vessels associated with changes in arterial pressure.10,20,45 The carotid receptors are innervated by the glossopharyngeal nerve and the aortic receptors by the vagus nerve (cranial nerves IX and X, respectively). Some of the sensory nerves respond to tonic levels of stretch, while others are more active when the receptors deform with each heart beat.10,20 The baroreceptor afferents are initially integrated at the nucleus of the solitary tract of the brainstem, and the information relayed by the afferents is processed and “sent” to other areas involved in cardiovascular control.20,45 In responses to increases or decreases in arterial pressure, predictable changes in heart rate and efferent 323

324

Function of the Peripheral Nervous System

sympathetic outflow occur. When blood pressure falls, there is an increase in heart rate as a result of vagal withdrawal. When blood pressure increases, there is increased vagal activity and heart rate slows. Likewise, when arterial pressure falls, there is an increase in efferent sympathetic traffic that evokes vasoconstriction to “defend” arterial pressure. By contrast, when pressure increases, sympathetic outflow is suppressed in an effort to reduce the pressure to a normal level. In general, these responses can be described by a sigmoidal curve and the “resting” blood pressure is on the steep linear region of this curve. Over a wide range of pressure, there is a linear relationship (gain) between pressure and either heart rate or sympathetic outflow. However, when blood pressure is extremely low, it is below the “threshold” for normal baroreflex function. By contrast, when pressure is extremely high, the afferent nerves are maximally activated and the system is “saturated.”10,20,45 Figure 15–1 is a schematic of a typical baroreflex response curve.

DIRECT MEASUREMENT OF SYMPATHETIC TRAFFIC IN HUMANS VIA MICRONEUROGRAPHY Microneurography consists of placing small tungsten microelectrodes in peripheral nerves of conscious humans. This technique was initially developed in the 1960s to study the function of motor nerves and skeletal muscle afferents in humans.25,75 During that time, other signals were noted that could be activated by breath holding or the Valsalva maneuver or eliminated by ganglionic blockade, and it was determined that it was also possible to record from multiunit bundles of sympathetic nerves that innervate skeletal muscle or skin.16,17 Although a number of important observations were made using this technique in the 1970s, as the technique became more widespread, a host of findings on almost every aspect of cardiovascular control in humans have been made since the early 1980s. Figure 15–2 depicts arterial blood pressure, breathing, and muscle sympathetic nerve activity (MSNA) from a human volunteer. It is interesting to note that, as blood pressure falls, MSNA increases, and as blood pressure increases, MSNA falls.

Sympathetic nerve activity or heart rate

Threshold

Linear region 150 mm Hg

Rest

100

Saturation 2s

Arterial pressure

FIGURE 15–1 Schematic representation of a baroreflex curve in humans. As blood pressure increases (x axis), there is a progressive fall in both heart rate and efferent sympathetic traffic (y axis). As pressure falls, there are reflex increases in heart rate in efferent sympathetic traffic. These changes are thought to be governed primarily by arterial baroreflexes. The increase in heart rate is due primarily to vagal withdrawal to values of approximately 100 beats/min and thereafter is the result of activation of sympathetic nerves to the heart. The “linear” portion of the stimulus response curve has been described as the gain of the arterial baroreflex, while the flat portion on the upper left is known as the “threshold” and the flat portion in the lower right corner, where further increases in pressure do not evoke further reductions in heart rate of sympathetic traffic, is known as the “saturation point.”

50

FIGURE 15–2 Classic recording of muscle sympathetic nerve activity (MSNA) and mean arterial pressure in a conscious human. In most humans there are continuous subtle changes in arterial pressure associated with either intrinsic cardiac rhythms or breathing. Under these circumstances, as blood pressure increases, baroreflexes are activated and muscle sympathetic nerve activity is reduced. Likewise, when pressure falls, MSNA increases. There is an approximately 1.3-s time lag between changes in arterial pressure and changes in MSNA to account for the time it takes for the baroreceptors to transduce the signal, for the various neural conduction delays, and for central integration of the signal along with the efferent response measured in the lower extremity. Note the large burst associated with aberrant heartbeat in the middle of the record. (From Wallin, B. G., and Fagius, J.: The sympathetic nervous system in man—aspects derived from microelectrode recordings. Trends Neurosci. 9:63, 1986, with permission.)

Sympathetic Nerves and Control of Blood Vessels to Human Limbs

CHARACTERISTICS OF PERIPHERAL SYMPATHETIC TRAFFIC IN HUMANS The most common site for recording peripheral sympathetic traffic in humans is the peroneal nerve. At this site it is possible to record MSNA and skin sympathetic nerve activity (SSNA).17,24 MSNA and SSNA have different characteristics. MSNA is primarily vasoconstrictive under baroreflex control and is “pulse synchronous,” with only one burst seen for each heartbeat. Additionally, MSNA is responsive to chemoreceptor stimuli and rises during hypercapnia or hypoxia.17,75,82,83 Finally, arousal stimuli such as a loud, unexpected noise do not evoke a rise in MSNA.17 In young, healthy, supine humans, the level of resting MSNA typically ranges from 10 to 30 bursts/min. Interestingly, there is little correlation between resting MSNA and baseline values of arterial pressure in normotensive subjects.66,67,83 Skin sympathetic nerve traffic is more complex because it is directed at blood vessels, sweat glands, and piloerector organs. In addition to vasoconstrictor nerves, there is neurally mediated vasodilatation in the skin. In a thermoneutral environment, SSNA is not pulse synchronous, is not under obvious baroreceptor or chemoreceptor control, and is responsive to arousal stimuli.16,75,82 Additionally, the characteristic sharp peaks seen in records of MSNA are less obvious in records of SSNA.24 The differences in MSNA and SSNA seen at rest and during other stimuli also highlight the idea that the sympathetic nervous system does not function in an “all

FIGURE 15–3 Original record of a “modified” Oxford test to assess integrated baroreflex function in conscious humans. Arterial pressure, MSNA, and heart rate (electrocardiogram) are being measured. In this technique, a 100-␮g intravenous bolus dose of nitroprusside is given to lower arterial pressure. One minute later, a 150-␮g bolus of phenylephrine is given intravenously to raise arterial pressure. In response to the fall in pressure, the MSNA bursts increase (middle tracing), and as pressure begins to rise as a result of the phenylephrine bolus, MSNA is suppressed. Likewise, heart rate increases as pressure falls and slows as pressure increases.

or none” unified fashion, and that sympathetic outflow to different organs and regions can differ.

BARORECEPTOR CONTROL OF MSNA Several techniques are available to assess baroreceptor control of MSNA in humans. One of the most useful is the “modified Oxford technique.” In this technique, beat-tobeat measurements of arterial pressure are made and MSNA is recorded during an intravenous bolus infusion of the vasodilator drug sodium nitroprusside followed 1 minute later by a bolus infusion of the vasoconstrictor phenylephrine.19,26,56 This paradigm permits measurement of the heart rate and MSNA responses to a wide range of pressures, and allows for the calculation of baroreflex function curves for both variables. Figure 15–3 is a record of a modified Oxford trial. One limitation with the technique is that the systemic infusion of drugs causes changes in pressure of both the carotid and aortic baroreceptors, and perhaps the cardiopulmonary receptors as well, so determining the relative contribution of each receptor population to the total response is not possible. It also appears that MSNA but not SSNA is under tonic inhibition by arterial baroreflexes. In conscious humans, administration of local anesthesia to the glossopharyngeal and vagus nerves has been used to eliminate afferent traffic from the carotid and aortic receptors.21 Figure 15–4 demonstrates that this evokes large increases in MSNA and arterial pressure with little change in SSNA.

“Modified Oxford” method for baroreflex assessment

Nitroprusside 100 ug bolus iv

Phenylephrine 150 ug bolus iv

60 s

Arterial pressure

Muscle sympathetic nerve activity

ECG

325

326

Function of the Peripheral Nervous System

1 Control

2 After atropine

3 Unilateral block

ECG

MSA SSA 200 mm Hg BP

100 0 5s

4 Bilateral block

5 Bilateral block

6 30 min after bilateral block

ECG

MSA

SSA

200 mm Hg

BP

100 30 s

110 s

0 5s

FIGURE 15–4 Individual record of the effects of bilateral carotid sinus nerve block on MSNA, SSNA, arterial pressure, and heart rate in a conscious human. 1, Control conditions. 2, The effects of atropine. Most notably, atropine caused heart rate to increase and blood pressure to rise slightly but had little impact on MSNA or SSNA. 3, Effect of unilateral glossopharyngeal nerve block on these variables. MSNA begins to rise, SSNA is unchanged, and arterial pressure continues to increase. 4, Bilateral nerve block is associated with a huge increase in MSNA, no change in SSNA, and a large increase in blood pressure. 5, This response is sustained. 6, The response “wears off” as the local anesthetic block recedes. This figure demonstrates that MSNA, but not SSNA, is under tonic baroreceptor control in humans. It also demonstrates the key role that baroreceptors play in suppressing sympathetic traffic and modulating mean arterial pressure. (From Fagius, J., Wallin, B. G., Sundlof, G., et al.: Sympathetic outflow in man after anaesthesia of the glossopharyngeal and vagus nerves. Brain 108:423, 1985, with permission.)

MSNA RESPONSES TO ORTHOSTATIC CHALLENGE One of the most common challenges to the autonomic nervous system in humans is assumption of the upright posture. As a result of gravity, there is a shift of approximately 0.5 to 1.0 L of blood from the heart, thorax, and abdomen to the dependent veins in the legs. Additionally, the head is now above heart level and the effective perfusion pressure to the carotid sinus and brain is reduced. In response to these changes, there is typically a 50% to 100% increase in MSNA that is sustained9 (Fig. 15–5). Additionally, if the orthostatic

stress is prolonged, there is a slow rise in MSNA above the immediate increase. This continues up to the point of syncope, when there is the sudden onset of sympathetic silence that occurs as the subject faints.64,84 An individual record of such a response is shown in Figure 15–6. The mechanisms that evoke the sympathetic silence with syncope are poorly understood. One common idea is that the heart becomes relatively empty, and that with vigorous contractions sympathoinhibitory afferents (normally active when the heart is “full”) are paradoxically stimulated and suppress sympathetic outflow. However, fainting can be seen in patients after heart transplantation who have

Sympathetic Nerves and Control of Blood Vessels to Human Limbs

Heart rate/min

327

denervated ventricles.23 It is also possible that hypotension in the carotid vessels or aortic vessels evokes paradoxical stimulation of arterial baroreceptor afferents and suppresses sympathetic outflow. However, experimental evidence for both mechanisms is far from clear-cut.

Bursts/min

120

100

80

80

60

60

40

40

20

20

0

MSNA RESPONSES TO CHEMORECEPTOR ACTIVATION

0 A

B

C

A

B

C

FIGURE 15–5 Effects of standing on heart rate and MSNA in conscious humans. Left, Heart rate increased from approximately 65 beats/min to approximately 80 to 90 beats/min. Right, The burst of MSNA essentially doubled from 30 to approximately 60/min. (From Burke, D., Sundlof, G., and Wallin, G.: Postural effects on muscle nerve sympathetic activity in man. J. Physiol. [Lond.] 272:399, 1977, with permission.)

Symp activity 200 mm Hg Blood pressure

100

Heart rate

0 90 Beats/min 60

Finger pleth 10 s

FIGURE 15–6 MSNA tracing from an individual subject during syncope. The tracing shows sympathetic activity, blood pressure, heart rate, and finger plethysmographic blood flow. The star indicates the time of the faint. The key observation is that, before the faint, there was a progressive rise of MSNA, which suddenly became silent at the time of the faint. This occurred concurrently with the fall in blood pressure and the marked bradycardia associated with vasovagal syncope. These observations demonstrate that a profound sympathetic silence is normally associated with standard vasovagal syncope. (From Wallin, B. G., and Sundlof, G.: Sympathetic outflow to muscles during vasovagal syncope. J. Auton. Nerv. Syst. 6:287, 1982, with permission.)

The peripheral chemoreceptors located in the carotid and aortic bodies are activated by a fall in O2, and to a lesser degree by a rise in CO2 or acidity.22 These receptors send signals to the medulla via cranial nerves IX and X, and, like the baroreceptor afferents, synapse initially in the nucleus of the tractus solitarius. The best known effect of peripheral chemoreceptor activation is hyperpnea, but these pathways also affect efferent sympathetic and parasympathetic outflow. Studies in animals have documented that exposure of the peripheral chemoreceptors to a fall in partial pressure of O2 results in a rise in sympathetic nerve activity that can be measured in peripheral sympathetic nerves and a bradycardia that is dependent on increased parasympathetic traffic to the heart. Likewise, studies in humans have observed increases in sympathetic vasoconstrictor outflow to muscle vascular beds during hypoxia.54,58,65 Figure 15–7 shows an example of the sympathetic responses to chemoreceptor activation. It should be noted that these primary chemoreflex responses are often modified, or overridden, by the effects of hyperpnea on autonomic outflow and cardiovascular function.57 For instance, in animals the bradycardia produced by peripheral chemoreceptor stimulation is often abolished, or even converted to tachycardia, by ongoing changes in ventilation.22 This may account for the rise in heart rate frequently reported during studies on hypoxia in humans.54,58,65 However, one series of studies suggests that increased ventilation or changes in breathing patterns in humans modify the pattern of sympathetic nerve activity within breaths but do not change the overall level of sympathetic outflow.60,61 This suggests that the sympathoexcitation produced by hypoxia in humans is a direct effect of peripheral chemoreflex activation.

Central Chemoreflexes and Autonomic Function The central chemoreceptors, located on the ventral aspect of the medulla, are activated by an increase in CO2 or acidity.4 The best known effects of central chemoreceptor activation are increases in ventilation. Indeed, studies have suggested that central chemoreceptor activation, depending on conditions, mediates 50% to 90% of the ventilatory response to hypercapnia.4 Under hyperoxic conditions, the response to hypercapnia is almost exclusively mediated by the central

328

Function of the Peripheral Nervous System

Control

8% Oxygen

Peroneal neurogram ECG Respiration Total MSNA 170 units

Burst frequency 17 min-1

Total MSNA 650 units

Burst frequency 50 min-1

Heart rate 52 beats min-1

Mean pressure 95 mm Hg

Heart rate 73 beats min-1

Mean pressure 94 mm Hg

FIGURE 15–7 Effects of systemic hypoxia, MSNA, heart rate, and respiration in a conscious human. Breathing 8% oxygen (right side) increases MSNA via the peroneal nerve. Heart rate and respiration also increased.

chemoreceptors.30 These pathways also affect the level of activity in the sympathetic outflow tracts, but this aspect of chemoreflex control is poorly understood. Studies in animals have documented that exposure of the central chemoreceptors to a rise in partial pressure of CO2 results in a rise in sympathetic nerve activity that can be measured in peripheral sympathetic nerves.23 Analogous studies in humans have shown that hyperoxic hypercapnia produces a rise in muscle sympathetic nerve activity and heart rate46,65; however, the vagal withdrawal that mediates the rise in heart rate may be secondary to changes in arterial pressure that were produced by hyperpnea.

MSNA AND AUTONOMIC RESPONSES TO EXERCISE Exercise is another frequent stressor that confronts the autonomic nervous system, and the increases in heart rate and blood pressure that occur with exercise are obvious examples of altered autonomic responses. In general, the increase in heart rate from resting to approximately 100 beats/min is largely due to withdrawal of vagal tone, with subsequent increases to maximum largely the result of increased sympathetic traffic.49,55,78,79 Three main mechanisms are thought to control the autonomic responses to exercise. These include central command, sensory feedback from contracting muscles, and baroreflex resetting. The behavior of MSNA and SSNA as contributors to these mechanisms is more complex. Central command is a term that describes a feedforward signal from the brain motor centers to the cardiovascular centers.81 This signal is thought to account for the nearly instantaneous rise in heart rate and arterial pressure that occurs with the onset of exercise. In this context, when conscious humans perform a static handgrip, there is usually no immediate rise in MSNA but there can be a rise in SSNA.41,80 These general observations suggest that central command is not a major regulator of MSNA in humans. Additionally, the initial burst in SSNA at the onset of contractions is consistent with the increased SSNA responses

seen with arousal stimuli. These observations again demonstrate that MSNA and SSNA behave differently and emphasize the idea that sympathetic outflow can be differentiated and selective in humans.

Sensory Feedback from Active Muscles Although MSNA does not rise at the onset of contractions if a static handgrip of sufficient force is maintained for a few minutes, there is a progressive rise in MSNA.41 This increase in MSNA can be maintained after the contractions if an arm cuff is inflated to supersystolic levels just before stopping the exercise. The general interpretation of this observation is that afferent nerves (i.e., ergoreceptors) in the active muscles that are sensitive to muscle metabolites can evoke a reflex increase in MSNA and arterial pressure. It does not appear that SSNA is under ergoreceptor control. Figure 15–8 details the MSNA responses to static handgripping and postexercise ischemia in humans. In general, the ergoreceptors are thinly myelinated (group III) or unmyelinated (group IV) afferents. In resting muscle, group III afferents are primarily mechanosensitive and group IV afferents chemosensitive. However, with contraction, the group III afferents become sensitized and respond to chemical stimuli. Studies in animals suggest that, although a variety of substances can stimulate the ergoreceptors, the factor that is most prominent in vivo is the H⫹ ion. This idea has been generally confirmed in humans, but it also appears that other unknown substances can stimulate the ergoreceptors when H⫹ ion is absent.34,78

Baroreflex Resetting during Exercise Blood pressure rises during exercise, and it appears that this increase in pressure is also facilitated by a resetting of the arterial baroreflexes so that the “set point” (Fig. 15–9) is elevated and a higher pressure is regulated. Baroreflex resetting is thought to be a second way that central command contributes to the autonomic responses to exercise.49,55,81 The neural substrates responsible for this

329

Sympathetic Nerves and Control of Blood Vessels to Human Limbs

10 sec. Mean voltage neurogram of MSA Control

Handgrip 1st min.

Handgrip 2nd min.

MIR 1st min.

MIR 2nd min.

Recovery

HR (beats/min)

76

88

83

67

63

67

MSA (burst/min X mean burst ampl.)

476

410

1182

1473

1386

603

MAP (mmHg)

100

114

127

122

122

107

FIGURE 15–8 Effects of isometric handgrip exercise and postexercise circulatory rest on MSNA responses to handgripping in humans. With the onset of contractions (central command), there is little increase in MSNA, but there are marked increases in heart rate and arterial pressure. As the contractions become fatiguing, MSNA slowly rises. When contractions stop but the metabolites associated with contractions are “trapped” in previously active muscles by inflation of an arm cuff (postexercise circulatory rest; MIR), the rise in sympathetic traffic seen at the end of exercise is maintained. These data demonstrate that central command is not associated with a large increase in MSNA at the onset of contractions and that, during exercise, MSNA is primarily under the control of “ergoreceptors” in the active muscles. These receptors appear to be sensitive to acidosis and the metabolic by-products of contraction. One idea is that they signal a “mismatch” between muscle metabolism and muscle blood flow, and the purpose of the increase in arterial pressure is to improve the blood flow to the “ischemic” contracting muscles. (From Mark, A. L., Victor, R. G., Nerhed, C., and Wallin, B. G.: Microneurographic studies of the mechanisms of sympathetic nerve responses to static exercise in humans. Circ. Res. 57:461, 1985, with permission.)

Sympathetic nerve activity or heart rate

Exercise

Rest

Arterial pressure

FIGURE 15–9 Schematic representation of arterial baroreflex “resetting” during exercise in humans. Exercise is associated with an increase in heart rate and arterial pressure. It is also associated with a “resetting” of baroreflex control of these variables. In general, the shape of the stimulus-response curve relating arterial pressure to either sympathetic nerve activity or heart rate is unchanged with exercise but the “set point” of the curve is shifted to the right. The mechanisms responsible for this are unknown, but one idea is that the “central command” associated with the motor effort needed to produce the contractions also stimulates the brainstem cardiovascular centers and causes the baroreflex resetting.

response in humans are poorly understood, but baroreflex resetting is thought to maintain or support the autonomic adjustments made by central command and ergoreceptors. To study central command and baroreflex resetting in humans, it is necessary to increase or decrease the effort associated with a given level of exercise and measure arterial baroreflex behavior during the various conditions. When humans are made weak with small doses of curarelike drugs, the effort associated with a given level of exercise is disproportionately increased. Under these conditions, when subjects are weak, the arterial pressure and heart rate responses to contraction are increased and there is clear resetting of arterial baroreflexes.49 The brief discussion above on the MSNA and autonomic response to exercise has generally used static handgrip exercise as a model to illustrate the role of central command, ergoreceptors, and baroreflex resetting. It should also be noted that MSNA is normally measured in resting muscle, but using a variety of approaches, it is clear that sympathetic activity increases in exercising muscle. The mechanisms that regulate MSNA during static handgripping also participate in the autonomic responses to whole-body dynamic exercise such as running or cycling. However, there is an additional challenge to consider during whole-body exercises55,62: how does the autonomic nervous system respond to the marked skeletal muscle vasodilatation seen in the active muscles?

330

Function of the Peripheral Nervous System

Interactions between Sympathetic Vasoconstriction and Metabolic Vasodilatation Metabolic vasodilatation in active muscles is one of the hallmarks of exercise, and blood flow to a small mass of active skeletal muscle can reach very high values (⬃300 mL/100 g tissue/min).1 Additionally, if only a modest fraction (⬃10 kg) of total skeletal muscle mass is “maximally” vasodilated during exercise, the ability of the heart to generate the cardiac output required to support mean arterial pressure would be surpassed.55 However, peak skeletal muscle blood flow is lower during large versus small muscle mass exercise, suggesting that blood flow to active skeletal muscles is “restrained” by the sympathetic nervous system so

that mean arterial pressure can be regulated. Figure 15–10 illustrates the importance of sympathetic restraint of blood flow to active muscles, showing that arterial pressure falls during supine (or 15-degree head-down) cycle exercise in a patient who had undergone multilevel surgical sympathectomy in the 1960s for the treatment of severe hypertension.42 In contrast to the argument above, which favors sympathetic restraint of metabolic vasodilatation, evidence from a variety of skeletal muscle preparations in animals demonstrates that the constrictor responses to either sympathetic nerve stimulation or administration of ␣-adrenergic agonists are blunted or eliminated during exercise-induced vasodilatation.6,28,29,51,70–74 This observation has been termed functional sympatholysis.51

Radial artery (mm. Hg) Supine leg exercise

200

100

0 Baseline

Radial artery (mm. Hg) 200

15 Head down tilt Supine leg exercise

100

0 Baseline

FIGURE 15–10 Effects of sympathetic denervation on blood pressure responses to exercise in conscious humans. This figure comes from a patient who underwent extensive surgical sympathectomy for the treatment of malignant hypertension around 1960. The key finding is that, with supine exercise, blood pressure falls as a result of unopposed metabolic vasodilatation in the active muscles (top). This fall in pressure occurs even when the subject is 15% head down to maximize venous return and cardiac output (bottom). This figure argues for the idea that the sympathetic nerves must restrain the metabolic vasodilatation that can occur during exercise so that arterial pressure can be regulated and blood flow directed appropriately to both vital organs and the contracting skeletal muscle. (From Marshall, R. J., Schirger, A., and Shepherd, J. T.: Blood pressure during supine exercise in idiopathic orthostatic hypotension. Circulation 24:76, 1961, with permission.)

Sympathetic Nerves and Control of Blood Vessels to Human Limbs

␣-Receptor Subtypes and Sympathetic Control of Blood Flow in Active Muscles In contrast to earlier ideas that ␣1 receptors were primarily postjunctional and vasoconstrictive while ␣2 receptors were primarily prejunctional and inhibited norepinephrine release, it is now clear that resistance vessels have both postsynaptic ␣1 and ␣2 receptors that are vasoconstricting.18,32,33,76 These postsynaptic ␣2 receptors are major contributors to limb vascular tone at rest in many species, including humans. For example, forearm vasodilatation in response to selective ␣1 blockade is less than that seen with nonselective blockade.18 Additionally, the vasoconstriction in the human forearm caused by various maneuvers is only partially blunted by brachial artery administration of selective ␣1 antagonists, but eliminated by nonselective ␣-antagonists.33 In this context, several observations have important implications for sympathetic control of blood flow in active muscles. First, experiments in animals demonstrate that contraction, hypoxia, and acidosis inhibit the ability of postsynaptic ␣2 receptors (and to a much lesser extent ␣1 receptors) to cause vasoconstriction in skeletal muscle resistance vessels.2,44 Second, there are relatively more postsynaptic ␣2 receptors in “fast-twitch” muscles, and in general there are relatively more ␣2-receptors in smaller versus larger skeletal muscle arterioles. These differences in receptor subtype distribution therefore favor “sympatholysis” because the postsynaptic ␣2 receptors are likely to be located near muscle fibers that become acidotic with moderate or heavy contractile activity.70 This distribution should also permit some ␣1-mediated vasoconstriction in the larger arterioles, which play a greater role in the control of muscle blood flow when the downstream vessels are dilated. The third line of evidence that clearly suggests a strong role for postsynaptic ␣2 receptors in “sympatholysis” comes from a series of studies in animals.70–74 These studies used sympathetic nerve stimulation along with selective ␣1 and ␣2 agonists and antagonists to show that skeletal muscle contraction clearly limits (or even eliminates) postsynaptic ␣2but not ␣1-mediated vasoconstriction in fast-twitch muscles from rodents. Additionally, studies with nitric oxide synthase inhibitors and nitric oxide–deficient animals strongly implicate locally produced NO as the major factor inhibiting locally produced ␣2-mediated vasoconstriction in the active muscles.73 Finally, in conscious instrumented dogs there is evidence of ongoing sympathetic restraint of blood flow to active muscles, but there is also evidence that postsynaptic vasoconstriction is blunted (especially ␣2 receptor–mediated constrictions) in a manner that is consistent with the studies discussed above.5–8,48 Taken together, it appears that sympathetic control of blood flow to active muscles is reduced during exercise, and that NO produced in the contracting muscles acting on vasoconstricting postsynaptic ␣2 receptors is the major site of “sympatholysis.”

331

Exercise: Summary MSNA and SSNA behave differently during exercise in humans. The main regulator of the increased MSNA seen in humans during static exercise appears to be ergoreceptors in the active muscles. During large muscle mass exercise, arterial baroreflexes probably contribute. In general, central command probably does not cause MSNA to rise but can influence SSNA. The idea that central command causes a resetting of arterial baroreflexes is now gaining wide acceptance. The way that increased sympathetic activity interacts with metabolic vasodilatation via postsynaptic ␣1 and ␣2 receptors is currently a major area of investigation.

AUTONOMIC AND MSNA RESPONSES TO MENTAL STRESS In humans, mental or emotional stress can evoke marked increases in arterial pressure and heart rate. These changes are also typically associated with measurable forearm vasodilatation.3,52 This has led to the concept that there might be sympathetic vasodilator nerves to skeletal muscle in humans. This concept is supported by animal data from a variety of species that provide neurophysiologic, pharmacologic, and histologic evidence for sympathetic dilator nerves to skeletal muscle.35 In general, these nerves are either sympathetic cholinergic nerves and/or nitroxidergic nerves. One idea is that acetylcholine released from sympathetic cholinergic nerves acts on the vascular endothelium to evoke NO release.12,43 Sympathetic vasodilator nerves to skeletal muscle in animals appear to be active during the so-called defense reaction, which occurs when certain brainstem areas are stimulated and evoke skeletal muscle vasodilatation in conjunction with hypertension and tachycardia.35,52 This response is thought to prepare the animal for a classic “fight or flight” response.

Neurally Mediated Dilatation in Human Skeletal Muscle A variety of lines of evidence suggest there might be neurally mediated skeletal muscle vasodilatation in humans. Some of the most obvious evidence comes from “mental stress” experiments conducted in the 1950s. In these studies, severe mental stress was associated with hypertension, tachycardia, and marked skeletal muscle vasodilatation.3 In a select number of subjects who had undergone surgical sympathectomy of the forearm, the vasodilatation was absent in the sympathectomized forearm. These and related studies were seen as roughly analogous to the defense reaction studies in animals and offered strong support for the existence of sympathetic dilator fibers to human muscle. However, in contrast to the older human studies, recent studies suggest that sympathetic vasodilator nerves to

332

Function of the Peripheral Nervous System

human skeletal muscle do not exist. First, in both primates and humans there is no histologic evidence for the existence of sympathetic cholinergic vasodilator nerves in muscle.35 Second, the vasodilator nerve responses to mental stress persist after various forms of nerve block, which should abolish sympathetic traffic to the forearm.27 Third, microneurographic records of sympathetic traffic to the forearm during mental stress demonstrate that there is sympathetic withdrawal during mental stress and not sympathetic activation.27 In other words, there is no neurophysiologic evidence for sympathetic dilator traffic. Although it is clear that the vasodilatation evoked by mental stress is at least partly dependent on NO, the current mechanism for the release of the NO appears to be sympathetic withdrawal, epinephrine-mediated activation of ␤2 receptors, and local mechanical factors evoking NO release.35,50

AGING, HYPERTENSION, AND CONGESTIVE HEART FAILURE When healthy older subjects who are free of disease (diabetes, hypertension, heart disease, etc.) are studied, it is clear that there is generalized increase in sympathetic outflow to many vascular beds and organ systems.59,67 Resting norepinephrine values are frequently higher, and norepinephrine spillover to many organ systems is increased. In this context, there is clear-cut evidence that sympathetic outflow to skeletal muscle, gut, brain, and perhaps the heart is increased with normal aging59 (Fig. 15–11). By contrast, sympathetic outflow to the kidney does not appear to increase with advancing age. Although baseline sympathetic outflow to many organ systems is increased with advancing age, the rise in sympathetic outflow above baseline during

A 20 yr old woman

24 yr old man

AP=122/78

BF

BI

11

19

17

29

25

38

38

66

AP=116/70

65 yr old woman

68 yr old man

AP=138/84

AP=116/68

5s

B 500

*

40

PNA (pg ml–1)

MSNA (bursts min–1)

50

*

30 * 20 * 10 0

400 300 200 100 0

YW

YM

OW

OM

YW

YM

OW

OM

FIGURE 15–11 A, Individual records of MSNA in older and younger men and women. B, MSNA and plasma noradrenaline levels in various groups. Young women (YW) have lower sympathetic traffic than young men (YM). Both younger groups have lower values than old women (OW). The values seen in older men (OM) are roughly twice those seen in their younger counterparts. These findings demonstrate that there is a progressive rise in sympathetic tone with aging. The mechanisms responsible for this rise are currently unknown. (From Seals, D. R., and Esler, M. D.: Human ageing and the sympathoadrenal system. J. Physiol. [Lond.] 528:407, 2000, with permission.)

Sympathetic Nerves and Control of Blood Vessels to Human Limbs

acute stress (standing, exercise, etc.) appears to be essentially normal.13–15,47,59,68 Currently the mechanisms responsible for the age-related changes in sympathetic outflow are poorly understood, but one idea is that increased subcortical central nervous sympathetic drive is responsible.39,59 Another possible contributor to the increased sympathetic outflow seen with aging is a progressive stiffening of the arterial blood vessels. If the vessels at the carotid and aortic receptors stiffen with age, a given change in arterial pressure will evoke less deformation and hence less baroreceptor inhibition of sympathetic outflow.11,31,69 In view of the aging population throughout the world, and the key role that the sympathetic nervous system plays in many disease states, further understanding of the autonomic responses to normal aging and aging in the context of disease are likely to become increasingly important. The pathophysiologic mechanisms that contribute to hypertension in humans are complex, and dysfunction in a host of key regulatory systems has been proposed. In this context, one important question is whether sympathetic outflow to skeletal muscle is chronically increased in individuals with essential hypertension. Although a detailed description of all the evidence for and against changes in MSNA as contributors to hypertension in humans is beyond the scope of this chapter, currently available evidence suggests that increases in MSNA are not major contributors to hypertension in most situations.67,83 The current idea is that baseline MSNA in individuals with essential hypertension is either normal or only slightly elevated. In contrast to hypertension, MSNA is clearly elevated in congestive heart failure. In many patients resting MSNA is roughly twofold greater than normal. Additionally, congestive heart failure is also associated with other signs of sympathetic activation, including increased resting norepinephrine levels.40 For many years the sympathetic activation and increase in MSNA with congestive heart failure were assumed to be due to the inability of the arterial cardiopulmonary receptors to inhibit sympathetic outflow as a result of poor “pump function” in people with congestive heart failure. More recent evidence from animal models suggests that congestive heart failure is associated with a complex series of events in the central nervous system that is associated with reduced centrally mediated inhibition of sympathetic outflow.85

AUTONOMIC RESPONSES TO BODY HEATING Increases in core temperature are sensed at the level of the hypothalamus and evoke widespread autonomic responses. Most notably, when the core temperature is increased by 0.5° to 1.0° C, there can be marked cutaneous vasodilatation. As core temperature increases between 1° and 2° C, skin blood flow can reach values

333

of 7 to 8 L/min. This means that skin blood flow can reach levels of 300 to 500 mL/100 g/min. This increase in flow is neurally mediated via sympathetic dilator fibers.53 Additionally, emerging evidence suggests that (as is the case with high skeletal muscle blood flows), when skin blood flow is very high, arterial baroreflexes act to either limit skin vasodilator nerve activity or restrain it via concurrent activation of vasoconstrictor activity.38

Cutaneous Vascular Responses to Changes in Temperature At rest, skin blood flow is probably on the order of 1 to 3 mL/100 g/min. When subjects are cooled, thermoregulatory reflexes are activated and there can be intense vasoconstriction of skin to protect core temperature.37,53 This vasoconstriction is mediated via noradrenergic sympathetic nerves. When core temperature increases initially, there is a withdrawal of vasoconstrictor sympathetic activity to skin and skin blood flow rises. As core temperature rises between 0.5° and 1.0° C, there is a marked increase in skin blood flow that occurs about the same time as the onset of sweating. As core temperature continues to rise, dilatation progresses. Temporary elimination of sympathetic outflow to the skin (nerve block) or surgical sympathectomy eliminates both the sweating response and the marked cutaneous vasodilator responses to body heating.53 In general, the sweating response is governed by sympathetic cholinergic nerves and can be blocked by administration of atropine. In contrast, the neurally mediated vasodilatation that can occur in skin can be blunted modestly by atropine administration but cannot be eliminated completely.53,63 This observation has led to the conclusion that acetylcholine released from the sympathetic cholinergic (sudomotor) nerves is not responsible for the marked cutaneous vasodilatation that occurs with body heating (Fig. 15–12). It also suggests that either there is a separate population of vasodilator nerve fibers that are independent from the sympathetic cholinergic fibers, or perhaps that some of the sympathetic cholinergic nerves release a vasodilating cotransmitter. This latter possibility has recently been supported in studies by Kellogg and colleagues.37 They demonstrated that, although muscarinic blockade had little response on the cutaneous vasodilator responses to body heating in humans, prejunctional inhibition of sympathetic cholinergic fibers eliminated the dilator response. They interpreted these findings as strong evidence that some substance cotransmitted with acetylcholine from the sympathetic cholinergic nerves is responsible for the marked cutaneous vasodilatation during body heating. Finally, for many years bradykinin was postulated to play a major role in the cutaneous vasodilator responses to body heating, but recent evidence using selective blockers of bradykinin clearly demonstrate that this substance is not obligatory for the response.36

334

Function of the Peripheral Nervous System

SUMMARY

A

Relative humidity (mV)

20

Control 15

10 Atropine

5 Botulinum 0

B

100 Control

Atropine

60

40 Cold stress

% Maximal CVC

80

20

Botulinum

ACKNOWLEDGMENTS Whole body heat stress

Local warming to 42 C

0 0

In the last 20 to 30 years, microneurographic and other sophisticated measurements have been used to probe the complexities of the autonomic nervous system in conscious humans. These studies have shown that sympathetic outflow and changes in sympathetic outflow to various organ systems and vascular beds can be highly selective and targeted. An increase in sympathetic outflow to one vascular bed or organ system does not necessarily mean a generalized increase in sympathetic outflow throughout the body. Additionally, these new techniques have permitted mechanistic studies on the reflex control of the cardiovascular system in humans and better understanding of how baroreflexes and chemoreflexes govern the cardiovascular system in health and disease. Finally, the autonomic responses to common stressors such as standing, exercise, and body heating have been explored and new insights of how the autonomic nervous system governs these responses have been provided. The success of these approaches demonstrates the continued power of integrative physiology in conscious humans in an era of molecular biology and genomics.

10

20

30

40

50

60

70

80

90 100

Minutes

FIGURE 15–12 A, The effect of whole-body heating on relative humidity of a small area of skin. The relative humidity is proportional to the sweat rate. Under normal conditions whole-body heating is associated with a large increase in sweating. This can be blocked by atropine, which postsynaptically limits the ability of the sudomotor cholinergic nerves to evoke the sweating response. Additionally, it can be inhibited presynaptically with botulinum toxin, which limits the release of acetylcholine from these nerves. B, The impact of these treatments on the skin blood flow responses to body heating. Normally during body heating there is intense activation of sympathetic vasodilator nerves. Whether these nerves are the same as or different from the sudomotor cholinergic nerves responsible for the sweating has been under lengthy and controversial investigation. B demonstrates the control response and also demonstrates that atropine blunts the cutaneous dilator response modestly but does not eliminate it. By contrast, botulinum toxin eliminated this response. This experiment argues that some substance co-released with acetylcholine by the sudomotor nerves is responsible for the marked cutaneous vasodilatation seen during body heating in humans. Despite more than 70 years of work on this topic, the substance responsible for this dilatation is currently unknown. (Adapted from Kellogg, D. L. Jr., Pergola, P. E., Piest, K. K., et al.: Cutaneous active vasodilation in humans is mediated by cholinergic nerve cotransmission. Circ. Res. 77:1222, 1995, with permission.)

The authors would like to thank Janet Beckman for her excellent secretarial assistance. They would also like to thank the many volunteer subjects who have participated in their studies over the years. This work was supported in part by grants HL 46493, NS 32352, HL 65305, and RR00585.

REFERENCES 1. Andersen, P., and Saltin, B.: Maximal perfusion of skeletal muscle in man. J. Physiol. (Lond.) 366:233, 1985. 2. Anderson, K. M., and Faber, J. E.: Differential sensitivity of arteriolar alpha 1- and alpha 2-adrenoceptor constriction to metabolic inhibition during rat skeletal muscle contraction. Circ. Res. 69:174, 1991. 3. Blair, D. A., Glover, W. E., Greenfield, A. D. M., and Roddie, I. C.: Excitation of cholinergic vasodilator nerves to human skeletal muscles during emotional stress. J. Physiol. (Lond.) 148:633, 1959. 4. Bruce, E. N., and Cherniack, N. S.: Central chemoreceptors. J. Appl. Physiol. 62:389, 1987. 5. Buckwalter, J. B., and Clifford, P. S.: Alpha-adrenergic vasoconstriction in active skeletal muscles during dynamic exercise. Am. J. Physiol. 277:H33, 1999. 6. Buckwalter, J. B., and Clifford, P. S.: The paradox of sympathetic vasoconstriction in exercising skeletal muscle. Exerc. Sport Sci. Rev. 29:159, 2001. 7. Buckwalter, J. B., Mueller, P. J., and Clifford, P. S.: Alpha 1-adrenergic-receptor responsiveness in skeletal muscle during dynamic exercise. J. Appl. Physiol. 85:2277, 1998.

Sympathetic Nerves and Control of Blood Vessels to Human Limbs 8. Buckwalter, J. B., Naik, J. S., Valic, Z., and Clifford, P. S.: Exercise attenuates alpha-adrenergic-receptor responsiveness in skeletal muscle vasculature. J. Appl. Physiol. 90:172, 2001. 9. Burke, D., Sundlof, G., and Wallin, G.: Postural effects on muscle nerve sympathetic activity in man. J. Physiol. (Lond.) 272:399, 1977. 10. Chapleau, M. W., and Abboud, F. M.: Contrasting effects of static and pulsatile pressure on carotid baroreceptor activity in dogs. Circ. Res. 61:648, 1987. 11. Chapleau, M. W., Cunningham, J. T., Sullivan, M. J., et al.: Structural versus functional modulation of the arterial baroreflex. Hypertension 26:341, 1995. 12. Davisson, R. L., Johnson, A. K., and Lewis, S. J.: Nitrosyl factors mediate active neurogenic hindquarter vasodilation in the conscious rat. Hypertension 23:962, 1994. 13. Davy, K. P., Jones, P. P., and Seals, D. R.: Influence of age on the sympathetic neural adjustments to alterations in systemic oxygen levels in humans. Am. J. Physiol. 273:R690, 1997. 14. Davy, K. P., Seals, D. R., and Tanaka, H.: Augmented cardiopulmonary and integrative sympathetic baroreflexes but attenuated peripheral vasoconstriction with age. Hypertension 32:298, 1998. 15. Davy, K. P., Tanaka, H., Andros, E. A., et al.: Influence of age on arterial baroreflex inhibition of sympathetic nerve activity in healthy adult humans. Am. J. Physiol. 275:H1768, 1998. 16. Delius, W., Hagbarth, K. E., Hongell, A., and Wallin, B. G.: General characteristics of sympathetic activity in human muscle nerves. Acta Physiol. Scand. 84:65, 1972. 17. Delius, W., Hagbarth, K. E., Hongell, A., and Wallin, B. G.: Manoeuvres affecting sympathetic outflow in human muscle nerves. Acta Physiol. Scand. 84:82, 1972. 18. Dinenno, F. A., Eisenach, J. H., Dietz, N. M., and Joyner, M. J.: Post-junctional alpha-adrenoceptors and basal limb vascular tone in healthy men. J. Physiol. (Lond.) 540:1103, 2002. 19. Ebert, T. J., and Cowley, A. W. Jr.: Baroreflex modulation of sympathetic outflow during physiological increases of vasopressin in humans. Am. J. Physiol. Heart Circ. Physiol. 262:H1372, 1992. 20. Eckberg, D. L., and Sleight, P.: Human Baroreflexes in Health and Disease. New York, Oxford University Press, 1992. 21. Fagius, J., Wallin, B. G., Sundlof, G., et al.: Sympathetic outflow in man after anaesthesia of the glossopharyngeal and vagus nerves. Brain 108:423, 1985. 22. Fitzgerald, R. S., and Lahiri, S.: Reflex responses to chemoreceptor stimulation. In Cherniack, N. S., and Widdicombe, J. G. (eds.): Handbook of Physiology: The Respiratory System. Vol. II: Control of Breathing. Washington, DC, American Physiological Society, p. 313, 1986. 23. Fitzpatrick, A. P., Banner, N., Cheng, A., et al.: Vasovagal reactions may occur after orthotopic heart transplantation. J. Am. Coll. Cardiol. 21:1132, 1993. 24. Hagbarth, K. E., Hallin, R. G., Hongell, A., et al.: General characteristics of sympathetic activity in human skin nerves. Acta Physiol. Scand. 84:164, 1972. 25. Hagbarth, K. E., and Vallbo, A. B.: Pulse and respiratory grouping of sympathetic impulses in human muscle nerves. Acta Physiol. Scand. 74:96, 1968.

335

26. Halliwill, J. R.: Segregated signal averaging of sympathetic baroreflex responses in humans. J. Appl. Physiol. 88:767, 2000. 27. Halliwill, J. R., Lawler, L. A., Eickhoff, T. J., et al.: Forearm sympathetic withdrawal and vasodilatation during mental stress in humans. J. Physiol. (Lond.) 504:211, 1997. 28. Hansen, J., Sander, M., and Thomas, G. D.: Metabolic modulation of sympathetic vasoconstriction in exercising skeletal muscle. Acta Physiol. Scand. 168:489, 2000. 29. Hansen, J., Sayad, D., Thomas, G. D., et al.: Exerciseinduced attenuation of alpha-adrenoceptor mediated vasoconstriction in humans: evidence from phase-contrast MRI. Cardiovasc. Res. 41:220, 1999. 30. Heeringa, J., Berkenbosch, A., de Goede, J., and Olievier, C. N.: Relative contribution of central and peripheral chemoreceptors to the ventilatory response to CO2 during hyperoxia. Respir. Physiol. 37:365, 1979. 31. Hunt, B. E., Farquhar, W. B., and Taylor, J. A.: Does reduced vascular stiffening fully explain preserved cardiovagal baroreflex function in older, physically active men? Circulation 103:2424, 1984. 32. Jie, K., van Brummelen, P., Vermey, P., et al.: Identification of vascular postsynaptic alpha 1- and alpha 2-adrenoceptors in man. Circ. Res. 54:447, 1984. 33. Jie, K., van Brummelen, P., Vermey, P., et al.: Postsynaptic alpha 1- and alpha 2-adrenoceptors in human blood vessels: interactions with exogenous and endogenous catecholamines. Eur. J. Clin. Invest. 17:174, 1987. 34. Joyner, M. J.: Muscle chemoreflexes and exercise in humans. Clin. Auton. Res. 2:201, 1992. 35. Joyner, M. J., and Halliwill, J. R.: Sympathetic vasodilatation in human limbs. J. Physiol. (Lond.) 526:471, 2000. 36. Kellogg, D. L. Jr., Liu, Y., McAllister, K., et al.: Bradykinin does not mediate cutaneous active vasodilation during heat stress in humans. J. Appl. Physiol. 93:1215, 2002. 37. Kellogg, D. L. Jr., Pergola, P. E., Piest, K. K., et al.: Cutaneous active vasodilation in humans is mediated by cholinergic nerve cotransmission. Circ. Res. 77:1222, 1995. 38. Kenney, W. L., Tankersley, C. G., Newswanger, D. L., and Puhl, S. M.: Alpha 1-adrenergic blockade does not alter control of skin blood flow during exercise. Am. J. Physiol. 260:H855, 1991. 39. Lambert, G. W., Kaye, D. M., Thompson, J. M., et al.: Internal jugular venous spillover of noradrenaline and metabolites and their association with sympathetic nervous activity. Acta Physiol. Scand. 163:155, 1998. 40. Leimbach, W. N. Jr., Wallin, B. G., Victor, R.G., et al.: Direct evidence from intraneural recordings for increased central sympathetic outflow in patients with heart failure. Circulation 73:913, 1986. 41. Mark, A. L., Victor, R. G., Nerhed, C., and Wallin, B. G.: Microneurographic studies of the mechanisms of sympathetic nerve responses to static exercise in humans. Circ. Res. 57:461, 1985. 42. Marshall, R. J., Schirger, A., and Shepherd, J. T.: Blood pressure during supine exercise in idiopathic orthostatic hypotension. Circulation 24:76, 1961. 43. Matsukawa, K., Shindo, T., Shirai, M., and Ninomiya, I.: Nitric oxide mediates cat hindlimb cholinergic vasodilation induced by stimulation of posterior hypothalamus. Jpn. J. Physiol. 43:473, 1993.

336

Function of the Peripheral Nervous System

44. McGillivray-Anderson, K. M., and Faber, J. E.: Effect of acidosis on contraction of microvascular smooth muscle by alpha 1- and alpha 2-adrenoceptors: implications for neural and metabolic regulation. Circ. Res. 66:1643, 1990. 45. Mifflin, S. W.: What does the brain know about blood pressure? News Physiol. Sci. 16:266, 2001. 46. Morgan, B. J., Crabtree, D. C., Palta, M., and Skatrud, J. B.: Combined hypoxia and hypercapnia evokes long-lasting sympathetic activation in humans. J. Appl. Physiol. 79:205, 1995. 47. Ng, A. V., Callister, R., Johnson, D. G., and Seals, D. R.: Sympathetic neural reactivity to stress does not increase with age in healthy humans. Am. J. Physiol. 267:H344, 1994. 48. O’Leary, D. S., Robinson, E. D., and Butler, J. L.: Is active skeletal muscle functionally vasoconstricted during dynamic exercise in conscious dogs? Am. J. Physiol. 272:R386, 1997. 49. Raven, P. B., Fadel, P. J., and Smith, S. A.: The influence of central command on baroreflex resetting during exercise. Exerc. Sport Sci. Rev. 30:39, 2002. 50. Reed, A. S., Tschakovsky, M. E., Minson, C. T., et al.: Skeletal muscle vasodilatation during sympathoexcitation is not neurally mediated in humans. J. Physiol. (Lond.) 525:253, 2000. 51. Remensnyder, J. P., Mitchell, J. H., and Sarnoff, S. J.: Functional sympatholysis during muscular activity. Circ. Res. 11:370, 1962. 52. Roddie, I. C.: Human responses to emotional stress. Ir. J. Med. Sci. 146:395, 1977. 53. Roddie, I. C.: Circulation to skin and adipose tissue. In Shepherd, J. T., and Abboud, F. M. (eds.): Handbook of Physiology, Sect. 2: The Cardiovascular System. Vol. III: Peripheral and Organ Blood Flow. Bethesda, MD, American Physiological Society, p. 285, 1983. 54. Rowell, L. B., Johnson, D. G., Chase, P. B., et al.: Hypoxemia raises muscle sympathetic activity but not norepinephrine in resting humans. J. Appl. Physiol. 66:1736, 1989. 55. Rowell, L. B., and O’Leary, D. S.: Reflex control of the circulation during exercise: chemoreflexes and mechanoreflexes. J. Appl. Physiol. 69:407, 1990. 56. Rudas, L., Crossman, A. A., Morillo, C. A., et al.: Human sympathetic and vagal baroreflex responses to sequential nitroprusside and phenylephrine. Am. J. Physiol. Heart Circ. Physiol. 276:H1691, 1999. 57. Rutherford, J. D., and Vatner, S. F.: Integrated carotid chemoreceptor and pulmonary inflation reflex control of peripheral vasoreactivity in conscious dogs. Circ. Res. 43:200, 1978. 58. Saito, M., Mano, T., Iwase, S., et al.: Responses in muscle sympathetic activity to acute hypoxia in humans. J. Appl. Physiol. 65:1548, 1988. 59. Seals, D. R., and Esler, M. D.: Human ageing and the sympathoadrenal system. J. Physiol. (Lond.) 528:407, 2000. 60. Seals, D. R., Suwarno, N. O., and Dempsey, J. A.: Influence of lung volume on sympathetic nerve discharges in normal humans. Circ. Res. 67:130, 1990. 61. Seals, D. R., Suwarno, N. O., Joyner, M. J., et al.: Respiratory modulation of muscle sympathetic nerve activity in intact and lung denervated humans. Circ. Res. 72:440, 1993. 62. Seals, D. R., and Victor, R. G.: Regulation of muscle sympathetic nerve activity during exercise in humans. Exerc. Sport Sci. Rev. 19:313, 1991.

63. Shastry, S., Minson, C. T., Wilson, S. A., et al.: Effects of atropine and L-NAME on cutaneous blood flow during body heating in humans. J. Appl. Physiol. 88:467, 2000. 64. Smith, M. L., Carlson, M. D., and Thames, M. D.: Naloxone does not prevent vasovagal syncope during simulated orthostasis in humans. J. Auton. Nerv. Syst. 45:1, 1993. 65. Somers, V. K., Mark, A. L., Zavala, D. C., and Abboud, F. M.: Influence of ventilation and hypocapnia on sympathetic nerve responses to hypoxia in normal humans. J. Appl. Physiol. 67:2095, 1989. 66. Sundlof, G., and Wallin, B. G.: The variability of muscle nerve sympathetic activity in resting recumbent man. J. Physiol. (Lond.) 272:383, 1977. 67. Sundlof, G., and Wallin, B. G.: Human muscle nerve sympathetic activity at rest: relationship to blood pressure and age. J. Physiol. (Lond.) 274:621, 1978. 68. Tanaka, H., Davy, K. P., and Seals, D. R.: Cardiopulmonary baroreflex inhibition of sympathetic nerve activity is preserved with age in healthy humans. J. Physiol. (Lond.) 515:249, 1999. 69. Tanaka, H., Dinenno, F. A., Monahan, K. D., et al.: Aging, habitual exercise, and dynamic arterial compliance. Circulation 102:1270, 2000. 70. Thomas, G. D., Hansen, J., and Victor, R. G.: Inhibition of alpha 2-adrenergic vasoconstriction during contraction of glycolytic, not oxidative, rat hindlimb muscle. Am. J. Physiol. 266:H920, 1994. 71. Thomas, G. D., Hansen, J., and Victor, R. G.: ATP-sensitive potassium channels mediate contraction-induced attenuation of sympathetic vasoconstriction in rat skeletal muscle. J. Clin. Invest. 99:2602, 1997. 72. Thomas, G. D., Sander, M., Lau, K. S., et al.: Impaired metabolic modulation of alpha-adrenergic vasoconstriction in dystrophin-deficient skeletal muscle. Proc. Natl. Acad. Sci. U. S. A. 95:15090, 1998. 73. Thomas, G. D., and Victor, R. G.: Nitric oxide mediates contraction-induced attenuation of sympathetic vasoconstriction in rat skeletal muscle. J. Physiol. (Lond.) 506:817, 1998. 74. Thomas, G. D., Zhang, W., and Victor, R. G.: Impaired modulation of sympathetic vasoconstriction in contracting skeletal muscle of rats with chronic myocardial infarctions: role of oxidative stress. Circ. Res. 88:816, 2001. 75. Vallbo, A. B., Hagbarth, K. E., Torebjork, H. E., and Wallin, B. G.: Somatosensory, proprioceptive, and sympathetic activity in human peripheral nerves. Physiol. Rev. 59:919, 1979. 76. van Brummelen, P., Jie, K., and van Zwieten, P. A.: Alphaadrenergic receptors in human blood vessels. Br. J. Clin. Pharmacol. 21:33S, 1986. 77. Vanhoutte, P. M., and Shepherd, J. T.: Autonomic nerves to the systemic blood vessels. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 301, 1984. 78. Victor, R. G., and Seals, D. R.: Reflex stimulation of sympathetic outflow during rhythmic exercise in humans. Am. J. Physiol. 257:H2017, 1989. 79. Victor, R. G., Seals, D. R., and Mark, A. L.: Differential control of heart rate and sympathetic nerve activity during

Sympathetic Nerves and Control of Blood Vessels to Human Limbs dynamic exercise: insight from direct intraneural recordings in humans. J. Clin. Invest. 79:508, 1987. 80. Vissing, S. F., and Hjortso, E. M.: Central motor command activates sympathetic outflow to the cutaneous circulation in humans. J. Physiol. (Lond.) 492:931, 1996. 81. Waldrop, T. G., Eldridge, F. L., Iwamoto, A., and Mitchell, J. H.: Central neural control of respiration and circulation during exercise. In Rowell, L. B., and Shepherd, J. T. (eds.): Handbook of Physiology, Sect. 12: Exercise: Regulation and Integration of Multiple Systems. New York, Oxford University Press, p. 333, 1996.

337

82. Wallin, B. G., and Fagius, J.: The sympathetic nervous system in man—aspects derived from microelectrode recordings. Trends Neurosci. 9:63, 1986. 83. Wallin, B. G., and Fagius, J.: Peripheral sympathetic neural activity in conscious humans. Annu. Rev. Physiol. 50:565, 1988. 84. Wallin, B. G., and Sundlof, G.: Sympathetic outflow to muscles during vasovagal syncope. J. Auton. Nerv. Syst. 6:287, 1982. 85. Zucker, I. H., Wang, W., Brandle, M., et al.: Neural regulation of sympathetic nerve activity in heart failure. Prog. Cardiovasc. Dis. 37:397, 1995.

16 Molecular Signaling in Schwann Cell Development RHONA MIRSKY AND KRISTJÁN R. JESSEN

Outline of Schwann Cell Development Differentiation Markers Associated with Embryonic Schwann Cell Development Brain-Specific Fatty Acid–Binding Protein Protein Zero Desert Hedgehog Oct6 Protein Growth-Associated Protein-43, CD9, Peripheral Myelin Protein 22, and Proteolipid Protein Neural Crest Origins of Schwann Cells Schwann Cell Precursors

Developmental Potential of Early Peripheral Glia Importance of Neuregulin-1 in Schwann Cell Development Neuregulin-1 Signaling Pathways in Schwann Cells Regulation of the Precursor–Schwann Cell Transition Signals That Control Schwann Cell Survival Cell Culture Studies on Schwann Cell Proliferation and Differentiation Transcription Factors Involved in Myelination

This chapter reviews glial development in embryonic and early neonatal nerve trunks. The origin of the Schwann cell lineage from the neural crest and the cell transformations that result in the emergence of immature Schwann cells are discussed. The extracellular signaling molecules and matrix receptors involved in controlling fate and lineage progression are considered, and the role of transcription factors that are known to be involved in regulating embryonic Schwann cell development and the onset of myelination is described. The myelinating and nonmyelinating Schwann cells of adult peripheral nerves are the two main peripheral glial cell types. Nevertheless, several other distinct categories of peripheral glia exist. These include satellite cells that surround neuronal cell bodies in sympathetic, parasympathetic, and sensory ganglia; the astrocyte-like enteric glial cells of the enteric nervous system; the teloglia (terminal Schwann cells) of somatic motor nerve terminals; and the specialized glia that associate with sensory terminals such as pacinian corpuscles and olfactory ensheathing cells.55,92,96,103,135,164,186,240,241,259,263,268,305,371 Evidence to date suggests that the molecular and morphologic differences

Signaling Pathways and Growth Factors That Regulate Myelination Neural Activity and Schwann Cell Development Extracellular Matrix, Matrix Receptors, and the Cytoskeleton in Schwann Cell Development Laminin Integrin Signaling in Schwann Cells and Interactions with the Actin Cytoskeleton Dystroglycan Signaling in Schwann Cells

between these various cells depend on the specific location and cellular environment in which they are found, and that the glial cells of the peripheral nervous system (PNS) retain unusual plasticity throughout life.

OUTLINE OF SCHWANN CELL DEVELOPMENT The myelinating and nonmyelinating Schwann cells of adult nerves originate from the neural crest in a process reviewed by several authors.136,184,208,209,277,336 Three main developmental transitions lie between migrating crest cells and mature Schwann cells (Fig. 16–1). The first involves the generation of Schwann cell precursors from neural crest cells. Next, the precursors transform into immature Schwann cells, which finally generate either myelinating or nonmyelinating cells. The first and last of these transition points involve a choice of fate, because neural crest cells give rise to several other cell types besides glia, and immature Schwann cells will become either myelinating 341

342

Neurobiology of the Peripheral Nervous System

FIGURE 16–1 The Schwann cell lineage in rat and mouse. There are three main transitions in the Schwann cell lineage: (1) the formation of Schwann cell precursors from undifferentiated migrating neural crest cells, (2) the precursor–Schwann cell transition, and (3) the formation of mature myelinating and nonmyelinating Schwann cells. The phenotypic changes that characterize each developmental step are summarized in Figure 16–2. The reversibility of postnatal Schwann cell development is indicated by stippled arrows. (From Jessen, K. R., and Mirsky, R.: Schwann cell development. In Lazzarini, R. A. [ed.]: Myelin Biology and Disorders, Vol. 1. San Diego, CA, Academic Press, p. 329, 2003, with permission.)

or nonmyelinating cells. These two transitions are also reversible, because myelinating and nonmyelinating cells revert to a phenotype comparable to that of immature Schwann cells, and Schwann cell precursors can be reprogrammed to generate other crest derivatives, at least in vitro. The transition from precursors to Schwann cells involves lineage progression rather than fate choice, because the only known fate of Schwann cell precursors in normal development is to become Schwann cells or die by programmed cell death. The extracellular signals that control fate choice in the lineage are not well understood. It is unclear what signals or mechanisms enable or direct crest cells to enter the glial lineage, and remarkably, the signals from larger diameter axons that are believed to instruct the Schwann cells associated with them to myelinate are unknown. Despite the lack of definitive information about what controls signaling at the choice points, two factors that regulate embryonic Schwann cell development have been identified by a combination of in vitro and in vivo experiments. These are neuregulin-1, which carries out a number of major functions in developing nerves, and endothelin, which is involved in the precursor–Schwann cell transition.26,99,136 (Neuregulin-1, also known as heregulin, ARIA, glial growth factor [GGF], and sensory neuron and motor–derived factor [SMDF], is a growth factor that, besides being involved in Schwann cell development, is also implicated in the progression of breast and other cancers.) Two transcription factors are known to be indispensable for Schwann cell development. The

first, Sox10, is essential for the generation of the earliest cells in the Schwann cell lineage, while the second, Krox20 (Egr2), is essential for myelination.27,336 (Sox10 is a member of the Sox family of the HMG [high mobility group] domain family of transcription factors. Another member of this family, SRY [sex-determining region of the Y chromosome], is essential for proper development of the male gonads, and other family members play crucial roles in development. Krox20 is one of four members of the EGR [early growth response] zinc finger transcription factor family. Another member of the family, Krox24 [Egr1], is essential for pituitary and ovarian development.) Additionally, the transcription factors Oct6 (SCIP, Tst1, POU3f1) and, to a lesser extent, Brn2 are important for the timing of myelination, while in vitro evidence suggests that the transcription factor nuclear factor-␬B (NF-␬B) is involved in the control of Oct6 expression.132,233,336 (Oct6 and Brn2 are both members of the POU domain family of transcription factors, which bend DNA and bind to DNA via an octamer sequence. They also contain a homeodomain and are involved in a variety of developmental and homeostatic processes. NF-␬B is a transcription factor involved in the regulation of many immune and inflammatory factors.) In vivo, Schwann cell precursors and immature Schwann cells proliferate rapidly, with a peak of DNA incorporation at the immature Schwann cell stage.30,313 Therefore, cell division is compatible with the differentiation processes that transform the phenotype of crest cells first to that of Schwann cell precursors and then to that of immature

Molecular Signaling in Schwann Cell Development

Schwann cells. Exit from the cell cycle occurs only at the final transition as immature Schwann cells differentiate into myelinating and nonmyelinating cells. A remarkable feature of the Schwann cell lineage is the rapid reversibility of this last step of Schwann cell development. Removal of Schwann cells from contact with axons, which initiates the reversal, can be achieved either by injuring nerves in vivo or by dissociating cells from adult nerves and placing them in culture without neurons. Both in vivo and in vitro, the process entails the developmental regression and dedifferentiation of individual Schwann cells and myelin breakdown. The characteristic molecular markers and structural features of myelinating and nonmyelinating cells are lost and the cells reenter the cell cycle and regress to a phenotype similar to that of immature Schwann cells. Such denervated Schwann cells redifferentiate readily on reassociation with axons during nerve regeneration.82,279 Are the other main transitions in the lineage also reversible? Reversal of the phenotype of immature Schwann cells leading to the re-formation of Schwann cell precursors has not yet been observed. Schwann cell precursors, in contrast, are likely to be more plastic. These cells have a phenotype that differentiates them unambiguously from both neural crest and early developing neurons, which indicates that they have been specified as glial cells. In their normal signaling environment in developing nerve trunks, they are also likely only to have a Schwann cell fate. As discussed further below, this does not imply that these cells are impotent to generate other cell types when challenged with alternative signals in an alien environment. It is possible that the entire cell population is not converted to the next stage at each of the three main transitions in Schwann cell development. For example, it has been suggested that, in addition to Schwann cell precursors, embryonic day (E) 14 rat nerves harbor a separate population of cells (10% to 15% of total cell number) that resembles neural crest cells.218,219 Evidence suggests that some precursor-like cells might be present in perinatal nerves21 and that even adult nerves may contain cells with an early phenotype corresponding to precursors or immature Schwann cells.267 Many other tissues, including the central nervous system (CNS), the enteric nervous system, muscle, and the hematopoietic system, retain cell populations with a phenotype and developmental potential reminiscent of early developing cells in their respective lineages.

DIFFERENTIATION MARKERS ASSOCIATED WITH EMBRYONIC SCHWANN CELL DEVELOPMENT The use of a set of well-defined molecular markers that characterize distinct stages in the development of myelinating and nonmyelinating cells has greatly facilitated analysis

343

of postnatal Schwann cell development and myelination. Comparable tools to study initial glial differentiation in the PNS are still relatively poorly developed despite the fact that the early development of neurons can be monitored using well-characterized molecular markers. These include neuron-specific tubulin (␤-III tubulin; TUJ1), neurofilament, peripherin, and early markers of distinct peripheral neuronal lineages such as neurogenins, Phox2, and Mash1.4,202,329,333 (Neurofilament and peripherin are intermediate filament proteins characteristic of neurons. Neurogenins 1 through 3 and Mash1 are members of the helix-loop-helix [HLH] family of transcription factors, while Phox2s are homeodomain transcription factors. All of these transcription factors play crucial roles in the specification and differentiation of distinct subsets of PNS and CNS neurons.) An orderly analysis of glial lineage progression in embryonic nerves has therefore not been straightforward, and markers that appear at a relatively late stage, in particular glial fibrillary acidic protein (GFAP) and S100, have often been used to monitor early events such as the emergence of the glial lineage from neural crest cells. (GFAP is an intermediate filament protein expressed by some astrocytes, enteric glia, and nonmyelinating Schwann cells. S100 is expressed by Schwann cells, astrocytes, neurons, and oligodendrocyte precursors, and is used in some circumstances as a glial marker.) Differentiation markers for embryonic Schwann cell development in rats and mice are, however, now appearing, and the task should become easier. Using stage specificity as the criterion, the markers shown in Figure 16–2 fall into four categories: 1. Markers that are expressed by embryonic PNS glia but do not differentiate between developmental stages, exemplified by the cell adhesion molecule L1 2. Markers that are expressed by crest cells and Schwann cell precursors but are expressed at very low levels on immature Schwann cells, namely N-cadherin (a cell adhesion molecule) and the transcription factor AP2␣ (a member of the AP2 family implicated in neural crest development) 3. Markers that differentiate Schwann cell precursors and immature Schwann cells from crest cells; three of these, brain-specific fatty acid–binding protein (B-FABP), protein zero (P0), and desert hedgehog (Dhh), are not found on early developing neurons and therefore distinguish between the neuronal and glial lineage at a very early stage 4. Markers that are expressed by immature Schwann cells that can be used to distinguish between them and Schwann cell precursors or crest cells, exemplified by S100 and GFAP Markers in the last two groups are particularly useful for monitoring the two main lineage transitions. Since the precursor–Schwann cell transition is relatively well characterized,136,209,210 the remarks below are limited to

344

Neurobiology of the Peripheral Nervous System

Neutal crest

Schwann-cell Precursors

Immature Schwann-cells

FIGURE 16–2 Markers of embryonic Schwann cell development. Neural crest cells, Schwann cell precursors, and Schwann cells are characterized by molecular markers as indicated in the graded gray boxes above the central diagrammatic scheme. The text below the boxes, graded from white to gray shades, summarizes some important differences in cellular properties between the three developmental stages. There are significant differences in survival properties between crest cells, precursors, and Schwann cells.364 Additional differences relate to the mitogenic response to FGF, because in rat cells FGF-2 is a mitogen for Schwann cells but not for precursors. Also precursors are highly motile in vitro in comparison to Schwann cells.69,134 (Modified from Jessen, K. R., and Mirsky, R.: The Schwann cell lineage. In Kettenmann, H., and Ransom, B. R. [eds.]: Neuroglia, 2nd ed. Oxford, UK, Oxford University Press, 2004.)

the third group, comprising markers that are likely to be useful for studying the important question of glial specification from the neural crest.

Brain-Specific Fatty Acid–Binding Protein This protein belongs to the fatty acid–binding protein family. It is closely related to the peripheral myelin protein P2 and the cellular retinoic acid–binding protein. In the

developing CNS, B-FABP is found in many regions, where it is expressed by radial glial cells in particular.85,159 Satellite cells in dorsal root sensory ganglia (DRGs) in embryonic and adult mice and glial cells in embryonic nerve trunks also express B-FABP, although adult Schwann cells do not.27,159 It is not expressed by migrating crest cells or developing neurons. It is first seen at thoracic levels of E10 mouse embryos, appearing first in DRGs and slightly later in limb nerves, and is seen in ganglia and

345

Molecular Signaling in Schwann Cell Development

nerves at all spinal levels at E11.27,364 This is consistent with the appearance of early glia, including Schwann cell precursors that have been identified in mouse nerves at E12 and E13.70 It has been suggested that B-FABP might characterize glia that are common precursors of Schwann and satellite cells.27,159 In the rat, although expression in the CNS appears comparable to that in the mouse at an equivalent developmental age, expression in both satellite cells and developing nerves at E13/14 is extremely low (K. R. Jessen and R. Mirsky, unpublished results, 2003). Therefore, in this species, it is unsuitable as a marker of early glial development.

Protein Zero Although P0 was initially identified as the major protein of Schwann cell myelin, it is now evident that P0 messenger RNA (mRNA) is expressed throughout the embryonic Schwann cell lineage in immature Schwann cells and Schwann cell precursors.52,168 P0 expression levels are, however, substantially lower than the axonally induced P0 expression in myelinating cells. In the chick, P0 protein is seen in a subpopulation of migrating crest cells in stage 19 embryos and subsequently in the earliest glia-associated axons in emerging nerves.19 In rats, P0 mRNA is first detected in a population of migrating crest cells located in a region ventral to condensing DRGs at E11, perhaps indicating differentiation into the glial lineage (see below).169 P0 expression is maintained in cells associating with the earliest motor axons emanating from the spinal cord and is subsequently seen in the large majority of cells in E14/15 rat nerves (i.e., Schwann cell precursors) and in most or all immature Schwann cells. P0 mRNA is not detectable in early neurons identified by the marker ␤-III tubulin (detected by TUJ1 antibodies), showing that P0 is not expressed even in very early cells of the neuronal lineage or at later stages (E. Calle and K. R. Jessen, unpublished observations, 1998).19,115,169 These findings suggest that P0 gene expression can be used as a very early marker of glial specification in neural crest development.19,168 Two comments should be made in relation to the acceptance of P0 as a marker specifically identifying cells that have just entered the glial lineage. First, in midgestation embryos, cells with detectable levels of P0 mRNA are, in the rat, seen in the notochord and in the otic vesicle and later on in cells in nonsensory regions of the ear, none of which contains, or gives rise to, PNS glia.169 The P0 gene is therefore transiently activated in restricted cell populations in other tissues in addition to its glial-restricted P0 expression within the crest sublineage. Second, it is clear that early crest-derived cells that express immunohistochemically detectable P0 are not irreversibly committed to a glial fate, because when these cells are exposed to differentiation signals in vitro, they can be induced to generate neurons and other crest

derivatives115,219,243 (see Developmental Potential of Early Peripheral Glia below). This raises an important general issue, that of the relationship between specification and developmental potential. These issues are discussed in the section on Developmental Potential of Early Peripheral Glia below. Briefly, the view that P0-positive Schwann cell precursors are specified glia, but retain the ability to be respecified when they are challenged by alternative signals in vitro away from their normal environment, accords well with current concepts and available data of developmental plasticity and transdifferentiation. At an equivalent developmental age in the mouse, it is more difficult to detect P0 mRNA in condensing DRGs and early nerves using digoxygenin-linked in situ hybridization than in the rat (R. Mirsky and K. R. Jessen, unpublished observations, 2003). Nevertheless, in E12 mouse nerves, P0 mRNA is readily detectable in Schwann cell precursors.253

Desert Hedgehog Dhh is a mammalian member of the hedgehog family of signaling molecules. It is secreted by Schwann cells and is crucial for the proper formation of the perineurium and epineurium around peripheral nerves.252 Dhh mRNA is present as early as E12 in mouse Schwann cell precursors but is not detectable in migrating crest cells or in developing PNS neurons20 (R. Mirsky and K. R. Jessen, unpublished observations, 2003).

Oct6 Protein Oct6, a member of the POU domain transcription factor family, controls the correct timing of Schwann cell myelination.15,212,336 In the perinatal period Oct6 protein is found in the nuclei of essentially all Schwann cells, with highest levels in promyelinating cells.7,280 It is also present at low levels earlier in development because it can be detected by reverse transcription–polymerase chain reaction and seen immunohistochemically in the nuclei of most or all late mouse Schwann cell precursors from E13 and in rat nerves from E16 at the time of the transition between precursors and Schwann cells.22 It is not seen in migrating neural crest cells or in developing peripheral neurons, although it is expressed by subsets of neurons in the CNS.336 Oct6 can therefore be used in vivo as a marker for late mouse Schwann cell precursors and in the rat for late precursors/early Schwann cells.171 It should be noted, however, that Oct6 expression depends in vivo on axonal signals280 and in vitro on factors that activate cyclic AMP (cAMP)-dependent pathways (e.g., forskolin or cAMP analogues). In vitro, therefore, neither Schwann cell precursors nor Schwann cells will express Oct6 unless cAMP pathways are activated.

346

Neurobiology of the Peripheral Nervous System

Growth-Associated Protein-43, CD9, Peripheral Myelin Protein 22, and Proteolipid Protein The first three of these markers are not expressed in migrating crest, although they are found on both rat Schwann cell precursors and DRG neurons at E14. Therefore, they cannot distinguish between early neurons and early glia, although they will distinguish these cells from the crest cells from which they originate. Growth-associated protein-43 (GAP-43; neuromodulin) is a phosphoprotein that has received much attention because it is expressed at high levels in growing neurites. Schwann cells at the neuromuscular junction and in nerve trunks also express detectable but lower levels of the protein.63,282,289,308 After nerve transection, GAP-43 expression is upregulated in Schwann cells in the distal stump of the nerve and at the denervated neuromuscular junction.254,365 In Schwann cell development, GAP-43 expression can be detected as early as the precursor stage at E14 by antibodies and immunohistochemistry.134 However, some of the currently available antibodies fail to detect GAP-43 at this early stage,216 while others detect GAP-43 at low levels or in a subpopulation of precursors (R. Mirsky and K. R. Jessen, unpublished observations, 2003). A possible explanation for this discrepancy could be that early glia may express a different form of the GAP-43 protein than that found in Schwann cells and neurons. CD9 is a tetraspan cell surface protein that is involved in Schwann cell migration in vitro. It has been described in Schwann cells as early as E18 using immunohistochemistry and tissue sections and is upregulated on myelination.5,10,142 CD9 is detectable immunohistochemically on E14 Schwann cell precursors and DRG neurons 2 to 3 hours after plating in vitro, although crest cells migrating from neural tubes are initially CD9 negative under identical culture conditions (K. R. Jessen, unpublished observations, 2003). In vitro, significant amounts of CD9 are shed from Schwann cell precursors onto the culture substrate. Peripheral myelin protein 22 (PMP22) is one of the important protein components of the PNS myelin sheath and, in common with P0, mutations or duplications in the gene frequently lead to peripheral neuropathy.229 It is present in some non-neural tissues in embryonic development,9 and is also expressed by some CNS and PNS neurons.115,250,251 PMP22 mRNA is present in the glia of E12 spinal nerves in the rat, and PMP22 mRNA and protein are expressed by E14 Schwann cell precursors. In contrast, the mRNA cannot be detected in migrating crest cells in E10 rat embryos.22,115 Proteolipid protein (PLP) is the major CNS myelin protein, but low quantities are also present in myelin sheaths in adult peripheral nerves. Null mutations in PLP can result in peripheral neuropathy.98 During

development, it is present in some non-neural tissues such as notochord and otic vesicle.304,332 In migrating mouse neural crest cells, mRNA can be detected as early as E9.5 in condensing sympathetic, trigeminal, and spinal ganglia and in peripheral nerves.304 In addition to glia, it may be expressed by neuroblasts or neurons at these early stages of development (B. Zalc, personal communication, 2003). It is therefore likely to detect neural cells at an early stage of their differentiation from the neural crest.

NEURAL CREST ORIGINS OF SCHWANN CELLS The neural crest is a transient group of cells that delaminates from the dorsal part of the neural tube during embryonic development. In the trunk region, cells of the neural crest give rise to glial cells; neurons of sensory, sympathetic, and parasympathetic ganglia; chromaffin cells; and melanocytes. The cardiac crest, situated in the most anterior part of the trunk, also gives rise to connective tissue and smooth muscle cells, while the cephalic crest, situated in the head region, generates in addition cartilage and bone cells.166 At the onset of migration, some neural crest cells may have already entered the glial lineage, while other cells appear to start glial development later as they migrate ventrally along the neural tube.121 The nature of the extrinsic signals that initiate glial development from crest cells is not established. As mentioned above, in embryonic nerves two growth factors, neuregulin-1 and endothelin,26,99 participate in the regulation of early Schwann cell development. At present it seems unlikely that these factors are required to trigger glial development from multipotent crest cells in vivo. The roles of neuregulin-1 and endothelin are discussed below in the section on Importance of Neuregulin-1 in Schwann Cell Development and the section on Regulation of the Precursor–Schwann Cell Transition, respectively. Cell culture studies indicate that bone morphogenetic proteins (BMPs), which are important for the generation of sympathetic neurons,283,290 may act as negative regulators of gliogenesis during crest development.290 Recent studies indicate that Notch activation may be involved in peripheral gliogenesis.95,97,155,218,346 (Notch is a transmembrane receptor activated by binding of membraneassociated ligands of the Delta family on neighboring cells. It is involved in cell fate determination in many different tissues.) This would be in line with observations on CNS glial development, because enforced Notch activation promotes the emergence of CNS glial cells in vivo.95,97,102,326 The evidence for Notch signaling in PNS gliogenesis is, however, less complete. It has been shown that, during division of avian crest cells, asymmetrical segregation of

Molecular Signaling in Schwann Cell Development

NUMB, a likely Notch antagonist, occurs. It has been suggested that, as a result, cells with higher levels of NUMB would be biased toward neuronal development within sensory ganglia.346 Indirect support for this prediction is provided by the finding that, in mice lacking NUMB, DRG neurons fail to form.374 Interestingly, in these animals sympathetic neurons are generated normally. Notch activation and neuregulin-1 share the ability to suppress neurogenesis from neural crest cells in vitro155,218,291,346 (see Importance of Neuregulin-1 in Schwann Cell Development below). In another parallel with neuregulin, it has been more difficult to demonstrate conclusively that Notch activation inductively triggers glial differentiation from neural crest cells. Clonal analysis experiments of migrating neural crest cells, treated with a soluble form of the Notch ligand Delta to activate Notch, reveal that Notch promotes glial differentiation, measured by an increase in the number of clones that contain GFAPpositive cells.155,218 In these experiments, lower levels of glial differentiation take place without Delta application, thus demonstrating that gliogenic signals are available to at least some of the cells without enforced Notch activation. Although this may be due to intrinsic Notch signaling, it is equally conceivable that the role of Notch activation in these experiments is to promote the action of other gliogenic signals. An established function of Notch is to block entry into nonglial lineages (see above). This could provide an indirect mechanism for an increase in glial numbers because it might well prolong the time during which crest cells are receptive to any instructive gliogenic signals present in the culture system. In conclusion, in the developing neural crest there is evidence that early developing neurons activate Notch in adjacent cells, which could be a mechanism for preventing excessive neurogenesis while allowing gliogenesis. It is still uncertain whether Notch activation also acts inductively to trigger gliogenesis or whether Notch promotes glial differentiation only indirectly. Experiments on cells from E14 rat nerves have raised the additional possibility that Notch signaling promotes the generation of Schwann cells from Schwann cell precursors (see Developmental Potential of Early Peripheral Glia below). The development of glia from crest cells depends on the transcription factor Sox10.27,157,243,253,302,303,353 In several other systems, Sox family members are also involved in regulating development.352 As the neural tube closes, Sox10 mRNA is expressed in restricted dorsal areas of the neural tube that give rise to neural crest cells.157 Sox10 mRNA is seen subsequently in migrating crest cells,27,257,303 and in vitro studies show that these cells all express Sox10 protein.243 In developing glia, both in ganglia and along nerve trunks, Sox10 expression persists, but in early developing neurons it is downregulated.157 During embryonic development the levels of Sox10 mRNA in peripheral nerves decline and levels in

347

Schwann cells in neonatal nerves are much lower than those in neural crest cells (U. Lange and K. R. Jessen, unpublished observations, 2002). The effects of Sox10 inactivation have been studied using both Sox10-null mice and dominant megacolon (Dom) mice, which have spontaneous nonsense or frameshift mutations in the Sox10 gene. In these mice, early peripheral glial cells are missing.27 Neither satellite cells within DRGs nor Schwann cell precursors in emerging peripheral nerves (identified in normal mice by expression of B-FABP) are present. In the mouse, B-FABP distinguishes early glial cells, which express B-FABP protein, from neural crest cells and early neurons, which are B-FABP negative (see Brain-Specific Fatty Acid–Binding Protein above). Despite the lack of B-FABP–positive glial precursors in DRGs in Sox10-deficient mice, sensory neurons, identified by expression of ␤-III tubulin, are initially generated in normal numbers, although they die later in development.302 This suggests that Sox10 has a particular role to play in glial development. One important function of Sox10 is likely to be the maintenance of ErbB3 expression. ErbB3, a key neuregulin-1 receptor, is expressed in crest cells as they emerge from the neural tube and in early glial cells. In Sox10-mutant mice, ErbB3 receptor mRNA appears on schedule in migrating crest cells, but as the crest cells migrate away from the dorsal neural tube and begin to differentiate into neurons and glia in the condensing DRGs, expression is not sustained, as it would be in normal mice.27 Although the survival of migrating crest cells in vivo does not appear to require neuregulin-1,28 the survival of Schwann cell precursors in nerve trunks depends on neuregulin-1 signaling and ErbB3 receptors (see Importance of Neuregulin-1 in Schwann Cell Development below). Neuregulin-1 is also a mitogen for these cells.28,69,70 Therefore, in Sox10 mutants, the predicted outcome of ErbB3 downregulation would be a selective reduction in the number of Schwann cell precursors in peripheral nerve trunks. This is in line with the finding that the number of Schwann cell precursors is severely reduced in nerve trunks of both Sox10 and ErbB3 mutants. Conversely, the principal reason for the absence of glia in Sox10-mutant DRGs cannot be due to the lack of neuregulin-1 signaling, because the development of DRG glia appears normal in mice in which neuregulin-1 signaling has been inactivated (see Importance of Neuregulin-1 in Schwann Cell Development below). Increased cell death and decreased proliferation of non-neuronal cells within condensing ganglia are seen in Sox10 mutants, although cell death in migrating crest cells prior to the condensation of the ganglia is normal.27,243,302 This would be explained if Sox10 were needed for survival signaling in DRG glia by factors other than neuregulin-1, such as insulin-like growth factors (IGFs) or fibroblast growth factors (FGFs).364 Another possible explanation for the lack of DRG glia

348

Neurobiology of the Peripheral Nervous System

might be that glial specification (i.e., the process by which crest cells change to become early glia) requires Sox10.27 Early DRGs of Sox10 mice, which lack B-FABP–positive glial cells (see above), contain, besides neurons, a cell population that appears to retain a neural crest cell phenotype, and a few such cells are present in nerve trunks.27,302 This is consistent with the idea that glial specification is blocked in the absence of Sox10, although crest cells thrive and can generate other cell types, including neurons. Cell culture experiments give indirect support to this idea. In Sox10 mutants, a proportion of cultured p75 low-affinity neurotrophin receptor (p75NTR)–positive non-neuronal cells from E13 mouse DRG, presumably undifferentiated neural crest cells,27 leaves the neural lineage and adopts a smooth muscle–like phenotype, one of the cell types in the crest repertoire, rather than the glial phenotype that cells from wild-type DRG adopt.243 The observation that enforced expression of Sox10 induces expression of the endogenous P0 gene, a marker of crest cell differentiation (see Protein Zero above), in the neuroblastoma cell line N2A provides a further link between Sox10 and glial differentiation.253 Thus at least two plausible mechanisms explain the effects of Sox10 mutations on early glial development, one relating to neuregulin-1 signaling and the other to glial specification. A wider role for Sox10 in the development of neural crest derivatives is indicated by deficiencies in pigmentation and in the enteric nervous system.352

SCHWANN CELL PRECURSORS The Schwann cell precursor defines the first stage of glial differentiation in peripheral nerve trunks.69,134,136 The large majority of cells in the limb nerves of E14/15 rats (E12/13 mice) are Schwann cell precursors. Analogous cells in dorsal and ventral roots and early ganglia may not have identical properties and are also likely to differ in the timing of appearance of glial differentiation markers.364 Initially, precursors appear mostly at the edge of nerves but are also found within more mature embryonic nerves. They are always in close apposition to axons. Their extensive sheetlike processes contact and form junctions with processes from neighboring cells, surrounding large groups of axons, which at this age are mostly of similar size (Fig. 16–3). In contrast to older nerves, in which there is abundant fibrous connective tissue containing blood vessels surrounding axon–Schwann cell units within a nerve fascicle, E14 rat or E12 mouse nerves possess no significant connective tissue or vessels. Examination of glia in rat hind limb nerves at various embryonic ages shows that, in E14 and E15 nerves, the large majority of cells have the phenotype of Schwann cell precursors, while by E18 most of these have

converted to cells with the distinct properties of immature Schwann cells. The precursor–Schwann cell transition therefore centers on the 16th embryonic day in these nerves. In mice the corresponding conversion occurs 2 days earlier.70 Early indications that there were significant differences between Schwann cells and Schwann cell precursors came from a study of cell survival and expression of the calciumbinding protein S100 in embryonic rat nerves. When cells from E14 and E18 nerves were compared, E14 cells showed very little cytoplasmic S100 immunoreactivity relative to older cells. Furthermore, while Schwann cells from E18 and older nerves survived at moderate or higher densities without added factors, the E14 cells died rapidly in vitro without added survival factors, irrespective of cell density. Autocrine survival circuits in Schwann cells were later revealed to be the cause of this difference in survival regulation. Schwann cell precursors do not possess autocrine survival circuits, relying instead on paracrine neuregulin-1 signaling from axons.134,200 Schwann cell precursors can now be differentiated from both the migrating crest cells they are derived from and the immature Schwann cells that they generate on the basis of a number of features (see Fig. 16–2). They are therefore specified glial cells. Presumably, during normal development they also have a single fate to become Schwann cells. This does not imply that Schwann cell precursors are determined or irreversibly committed to a glial fate (see Developmental Potential of Early Peripheral Glia below). The essential function of Schwann cell precursors is to generate Schwann cells, but an important additional role for these cells is suggested by studies on mutant mice lacking isoform III of neuregulin-1, the ErbB3 or ErbB2 neuregulin receptors, or the transcription factor Sox10. Schwann cell precursors are either missing or their number is severely reduced in all of these mutants. While DRG neurons and motor neurons are initially generated normally, by E14 and E18, respectively, at limb levels in the trunk the majority of these neurons have died. This cell death is most likely due to the absence of glialderived factors that are needed for the survival of sensory and motor neurons.27,266 Thus a key function of Schwann cell precursors and early Schwann cells may be to regulate the survival of discrete pools of embryonic CNS and PNS neurons. It is an interesting possibility that these cells control neuronal survival through the mechanism of back-signaling by the intracellular domain of neuregulin-1.11 If this were the case, neurons and glia in embryonic nerves would ensure mutual survival through a bidirectional effect of the same molecular interaction, the binding of neuregulin-1 to ErbB2/ErbB3 receptors. In the neuregulin mutants, loss of axonal contact with peripheral targets may also be an important contributor to sensory and motor neuron death.362

Molecular Signaling in Schwann Cell Development

349

FIGURE 16–3 Cells of the Schwann cell lineage as they appear during peripheral nerve development. Top, Four Schwann cell precursors (P1 through P4) in the hind limb nerves of a rat embryo at E15. The nuclei of three of the precursors (P1, P3, and P4) are visible. Schwann cell precursors form close contacts (arrows) with each other, and are either embedded between axons inside the nerve, as shown here, or found in close apposition to axons at the surface of the nerve. Because connective tissue spaces and extracellular matrix are essentially absent, these nerves are much more compact than older ones. The axons also have a smaller and more uniform diameter than those seen in mature nerves. (Courtesy of Y. Hashimoto.) Bar: 1 ␮m. Middle, Schwann cells in the sciatic nerve of a newborn mouse (NB). Immature Schwann cells (S**) communally envelop a group of axons of various diameters. Other cells (S) are at a very early stage of myelination, the promyelination stage, having formed a 1:1 relationship with large axons (A). S* has progressed slightly farther along the myelin lineage and is starting to form a compact sheath around the axon (A*). Part of a thin myelin sheath can be seen on the extreme right-hand side of the field (arrow). Note that, in contrast to the compact arrangement seen earlier in development, perinatal nerves contain considerable collagen (C) and extracellular space. Bar: 0.5 ␮m. Bottom, Mature Schwann cells in transverse section of adult rat sciatic nerve (AD). Left, A myelinating Schwann cell forms a compact multilayered sheath (M) around a single large-diameter axon (A*). Parts of other myelinated fibers are present. Bar: 1 ␮m. Right, Cross section through the nucleus of a nonmyelinating Schwann cell (N-M) that ensheathes 13 axons (e.g., A), each lying in a separate trough in the cell surface. Nonmyelinating Schwann cells can also ensheath singly. Myelin sheaths (M) of neighboring axons are visible, and the axon–Schwann cell units are surrounded by collagen-rich extracellular spaces (C). (Middle and bottom panels courtesy of R. M. King and P. K. Thomas.) Bar: 1 ␮m. (Modified from Jessen, K. R., and Mirsky, R.: Schwann cell development. In Lazzarini, R. A. [ed.]: Myelin Biology and Disorders, Vol. 1. San Diego, CA, Academic Press, p. 329, 2003.)

DEVELOPMENTAL POTENTIAL OF EARLY PERIPHERAL GLIA During normal development Schwann cell precursors are unlikely to generate cells other than Schwann cells, such as neurons or melanocytes. Nevertheless, it is clear that,

when cells from early nerves or DRGs of chick, rat, or mouse are exposed to growth factors in cell culture, they can be directed to generate other cell types. There could be two principal reasons for this (see below). Possibly, the early glia of peripheral nerves are not yet irreversibly committed to the glial lineage. If such

350

Neurobiology of the Peripheral Nervous System

cells are exposed in vitro to a signaling environment significantly different from that which they normally encounter in vivo, their potential to generate other crest derivatives could be revealed. The other possibility is that, in addition to specified glia, early nerves contain a small, undifferentiated cell population that remains similar to migrating neural crest cells in molecular phenotype and developmental potential. Studies on avian systems provide the earliest evidence that early nerve trunks can in vitro give rise to cells that are not normally generated in vivo. When cultured in complex media, nerve segments from E4-E6 chick embryos, presumed to consist largely of Schwann cells with a minimal number of fibroblasts, gave rise to pigmented cells of the melanocyte lineage.294 Melanocytes were not generated from nerves of older embryos in these studies, although recent experiments indicate that cells from adult mouse nerves can be diverted to melanocyte differentiation. Nerve injury triggers this melanogenesis, which is more extensive in mice containing a single copy of the neurofibromatosis type 1 (Nf1) gene than in wild-type mice.267 Experiments on early quail nerve segments treated with FGF-2 also showed that melanocytes can arise from cells that were P0 positive, and that were therefore presumed to be in the glial lineage. Thus, FGF-mediated signaling results in transdifferentiation or respecification of Schwann cell precursors that would normally give rise to Schwann cells.294,316 Two other factors, endothelin-3 and Steel factor, have also been implicated in the respecification.228 Cultures of sensory ganglia can also give rise to melanocytes.62,232 Taken together, these studies show that specified P0-positive avian Schwann cell precursors can be diverted to generate at least one other crest derivative (i.e., melanocytes), and that therefore they are not irreversibly committed to the Schwann cell lineage. Related observations have been made on early rat nerves.219 In these studies, cells dissociated from E14 rat nerves were divided into subpopulations by fluorescenceactivated cell sorting (FACS) prior to culturing, using antibodies to myelin protein P0 and p75NTR. Cells initially sorted as P0 positive and p75NTR positive were analyzed at clonal density after 2 weeks of culture in complex medium containing chick embryo extract and retinoic acid. Many of these cells had given rise to GFAP-positive Schwann cells, but also to neurons and/or cells that express smooth muscle actin. BMP2 promoted the development of neurons from many of the sorted cells, and neuregulin-1 blocked neurogenesis. These signals have similar effects on migrating crest cells.219 A comparable strategy was used to investigate the developmental potential of non-neuronal cells from E14 rat DRG. These cells coexpress p75NTR, P0, and PMP22 and are therefore likely to be cells that in normal unperturbed development would give rise to satellite glial cells. Nevertheless, in vitro they were able to generate non–glial crest derivatives, including neurons and smooth muscle

cells.115 Taken together, these experiments are consistent with the idea, derived from studies on avian nerves, that specified P0-positive early glial cells in the PNS retain some developmental plasticity, although there is some uncertainty about the FACS methodology used in the experiments of Morrison et al.219 (see below). As discussed previously, a distinct population of multipotent crest-like cells that could be diverted to nonglial lineages might be present in embryonic nerves, in addition to Schwann cell precursors. It has been suggested that these cells can be distinguished from Schwann cell precursors by absence of P0 expression and that this type of P0-negative cell accounts for approximately 15% of the total number of cells dissociated from E14 rat nerves.218,219,358 The evidence for this interesting possibility is not yet complete. Initially, it came from experiments in which cells from E14 rat nerves were fractionated using FACS (see above).155,218,219,358 Cells that were P0 negative and p75NTR positive, a molecular phenotype expected of migrating neural crest cells, were able to self-renew and generate neurons, glia, or “myofibroblasts” in response to BMPs, neuregulin-1, and transforming growth factor-␤ (TGF-␤), responses that are qualitatively similar to those of migrating crest cells. Subsequent studies demonstrated, however, that the E14 nerve-derived cells were about 10 times less sensitive to the neurogenic signal BMP2 than crest cells derived directly from neural tube cultures. Using an in vivo assay, the nerve-derived cells were also shown to be strongly biased toward generating glia, in comparison to tube-derived crest cells.358 Similarly, in vitro the P0-negative, p75NTR-positive E14 nerve-derived cells were more sensitive to the gliogenic actions of Delta-Notch signaling than migrating crest cells. This change may be caused by reduced expression of the putative Notch inhibitor NUMB and elevated expression of Notch in these cells.155 Thus cells that FACS sort from E14 nerves as P0 negative and p75NTR positive retain more than one developmental option, at least in vitro, but in comparison with migrating crest cells are strongly biased toward gliogenesis. As suggested by Morrison et al.,219 these cells may represent a distinct population of resident neural crest cell–like cells in embryonic nerves. Another interpretation of these results is possible, however, particularly since the methodology used for obtaining them raises some concern. These cells were FACS sorted as P0 negative and p75NTR positive using the P07 anti-P0 antibody described by Archelos et al.6 Two issues raise questions about relying on the P07 antibody to sort cells from embryonic nerves. Although it was made against an extracellular portion of P0, binding of this antibody to living, unfixed Schwann cells or their precursors cannot be detected immunohistochemically. Significant binding of this antibody to cell surfaces appears to require previous fixation. Second, another population of cells that can be sorted from

Molecular Signaling in Schwann Cell Development

the same tissue (i.e., P0-positive, p75NTR-negative cells) forms neither Schwann cells in the presence of neuregulin-1, as would be expected of cells expressing P0, nor neurons in the presence of BMP2/4.219 The absence of neuronal and glial differentiation from these cells in the presence of BMPs and neuregulin, respectively, as well as the population size (~20% of the total cell suspension derived from freshly dissected E14 nerves) and the lack of p75NTR expression, are consistent with the idea that these might be non-neural, mesenchymal cells adhering to dissected E14 nerves, as indicated by Morrison et al.219 In this case, the expression of P0 by these cells would be puzzling because such cells are not reported to express P0. The identity of the P0-negative, p75NTR-positive FACSsorted cells from E14 nerves as a distinct population of crest-like cells is therefore ambiguous. One alternative is that the cells in question represent Schwann cell precursors, a possibility that would rest on the presumption that FACS based on this P0 antibody is not meaningful. The observation that these cells can adopt more than one lineage in vitro does not illuminate the issue because, as discussed earlier (see Protein Zero above), plasticity may well be a property of both migrating crest cells and Schwann cell precursors. The likelihood that the P0-negative, p75NTR-positive cells sorted in these experiments are Schwann cell precursors would be consistent with their strong bias toward gliogenesis. It would also accord with the action of Notch activation in these cells, which is to induce irreversible commitment to the glial lineage.219 It is unlikely that such a commitment would occur as migrating crest cells generate Schwann cell precursors, because, as discussed above, precursors show lineage plasticity. Conversely, Schwann cells are much more lineage restricted.56,219 It seems more likely that the cells from E14 nerves that become irrevocably confined to glial differentiation on Notch activation are Schwann cell precursors. This implies that a function of Notch activation is to promote the generation of Schwann cells from Schwann cell precursors, an idea that has not been tested. To conclude, at least two plausible explanations exist for the phenotypic plasticity observed in cells from early embryonic nerves, and they are not mutually exclusive. First, although Schwann cell precursors are specified glial cells, they may not yet be determined (i.e., irreversibly committed) to a glial fate. Second, a minority population of cells with a developmental potential and phenotype similar to that of migrating crest cells might well be present in early nerves. New studies on stem cell development, respecification, and transdifferentiation of one cell type into another have challenged earlier views on commitment and phenotypic plasticity.93,339 Taken together, this work suggests that the differentiated state is less fixed than previously envisaged. The idea that specified precursor cells can be reprogrammed in vitro by a variety of external signals is no longer surprising

351

and is consistent with a large number of experiments. The oligodendrocyte lineage provides a parallel example to the Schwann cell precursor.153 This work shows that cells that are unambiguously specified as oligodendrocyte precursors, and are therefore in a glial lineage, can be converted into multipotential stem cells by extracellular signals. An important conclusion resulting from this and other work is that, during development, a significant time lag may occur between specification of a cell to a particular lineage and determination, a state that implies irreversible commitment. In Schwann cell development this is probably reflected in the observation that a broader spectrum of cells can be generated from rat nerves at E14/15, which contain Schwann cell precursors, than from E17 nerves, in which a majority of the cells are Schwann cells.219 Comparable observations have been made using chick DRGs and nerves.56

IMPORTANCE OF NEUREGULIN-1 IN SCHWANN CELL DEVELOPMENT Neuregulin-1 is involved in all stages of Schwann cell development from the neural crest stage through to fully differentiated Schwann cells and is the most important known cell-cell signal regulating the lineage.1,34,99,173 It has a long history of association with Schwann cells. In an impure form, it was initially isolated as one of the first mitogens for cultured rat Schwann cells, being later identified as the type II isoform of neuregulin-1, GGF.101,172,191 It is now considered to be the major axonally derived Schwann cell mitogen (see Cell Culture Studies on Schwann Cell Proliferation and Differentiation below). In recent years, three other members of the neuregulin gene family (neuregulins 2 through 4) have been identified.36,118,372 Neuregulin-2 and -3 are expressed in the nervous system, but their function in peripheral nerve development, if any, is unknown.1 As mentioned earlier (see Schwann Cell Precursors above), mice lacking either neuregulin-1 or the neuregulin receptors ErbB2 or ErbB3 have embryonic nerve trunks that are essentially devoid of Schwann cell precursors, although, interestingly, development of satellite glia within DRGs appears to be normal, as is the initial development of DRG neurons.99,204,216,266,361 Sympathetic ganglia, however, fail to develop normally in these animals.28 These striking phenotypes have provided essential clues for understanding the role of neuregulin signaling in embryonic development of the PNS. Because satellite glia within ganglia and Schwann cells along nerve trunks develop from neural crest cells, the normal occurrence of DRG satellite cells in neuregulin mutants suggests that neuregulin-1 signaling is not obligatory for glial differentiation from the neural crest in vivo. This accords with the observation that GFAP-positive glia develop readily from cultured neural crest cells with or without added

352

Neurobiology of the Peripheral Nervous System

neuregulin.291 Although neuregulin strongly represses the formation of neurons from cultured crest cells,291 the neuregulin mutants do not provide any obvious indication, such as overproduction of neurons, to support the idea that neuregulin-1 constrains neurogenesis in vivo. Nevertheless, it is possible that this effect of neuregulin is masked in vivo by alternative factors that are absent in the culture system. One important reason for the remarkable reduction or absence of Schwann cell precursors from nerve trunks in neuregulin mutants is likely to be connected to the role of neuregulin-1 as a survival factor for Schwann cell precursors and early Schwann cells.69,136,171 Unlike Schwann cells, Schwann cell precursors are unable to support their own survival by autocrine mechanisms and depend instead on paracrine signals from axons (see Regulation of the Precursor–Schwann Cell Transition, and Signals That Control Schwann Cell Survival, below).134,200 In developing nerves they are closely apposed to axons, and die rapidly when they are removed from axonal contact either by culture in the absence of axons or by nerve transection in vivo.57,69,134,360 Neuregulin-1 isoforms are found in vivo in embryonic DRGs and motor neurons and accumulate along axonal tracts.81,185,191,239 Furthermore, cell culture studies show that DRG neurons are a potent source of both axon-associated and secreted survival signals that have been identified as neuregulin-1 isoforms.69 Neuregulin-1 is therefore available at the right time and place to enable the survival of precursors in embryonic spinal nerves. This is further supported by observations on mutants lacking isoform III of neuregulin-1, the major isoform in peripheral nerves.362 In these mice, early Schwann cell precursors start populating nerves at E11, but by E14, a time when precursors are converting rapidly to Schwann cells, the number of cells in the nerves is severely depleted. This argues against the idea that defective migration of precursors into nerve trunks is the major reason for absence of precursors in the nerves of neuregulin mutants (see below). Neuregulin has also been found to support the survival of early glia in embryonic chick nerves.360 Taken together, these observations strongly support the view that a major function of neuregulin-1 in embryonic nerves is to ensure the survival of Schwann cell precursors.69,136 The other factor that may contribute to the lack of Schwann cell precursors in developing nerves of mice deficient in neuregulin signaling is reduced cell migration. Neuregulin-1 stimulates migration of postnatal Schwann cells in vitro and glial migration from newborn mice DRG explants.201,216 Furthermore, in mice lacking ErbB3 receptors, neural crest cell migration along the ventromedial pathway is impaired. This is thought to be a major factor in the failure to develop sympathetic ganglia.28 Finally, in cell culture, the migration of glia from E12 DRG of ErbB2mutant mice is reduced.216 However, neuregulin-1 has no

effect on migration of glia from E12 DRGs from normal mice in vitro, and neural crest cells migrate out normally from the neural tube, and from DRGs in appropriate locations, in mice lacking neuregulin-1 or the ErbB3 receptor. While the evidence relating to migration is therefore ambiguous, it is obviously possible that reduced migration is a contributory factor to the sparsity of precursors in nerves of mice with deficient neuregulin signaling. From studies on neuregulin-mutant mice, it has been concluded that neuregulin-1 isoform I (NDF, ARIA), isoform II (GGF), and, in particular, isoform III (SMDF, cysteine-rich domain [CRD]-NRG-1) are the most important isoforms for PNS and Schwann cell development.34,90,205 Animals with inactivation of the neuregulin-1 gene, and therefore of all three isoforms, or of the neuregulin receptors ErbB2 or ErbB3 have the fewest Schwann cell precursors in their peripheral nerve trunks. Mice with selective inactivation of the neuregulin-1 isoform III initially show a milder phenotype, although at later stages Schwann cell precursor numbers are severely depleted and Schwann cells are essentially absent from the distal portion of the nerves (see above).362 In contrast, normal glial development is seen in mice lacking neuregulin-1 isoforms I and II, although muscle spindle development is deficient.122,205 Therefore, of the three main neuregulin-1 isoforms, isoform III is most crucial for Schwann cell development. Isoforms I and II contain an immunoglobulin domain and bind heparin, and may therefore associate with cell surfaces,1 whereas isoform III is mostly expressed as a transmembrane protein.347 Membrane bound, but not soluble, isoform III induces S100 and Oct6 expression in P0-positive cultured non-neuronal cells from embryonic rat DRGs.171 This suggests, first, that membranebound forms may have functions not shared by soluble forms and, second, that one of the functions of neuregulin-1 is to accelerate the precursor–Schwann cell transition (see Regulation of the Precursor–Schwann Cell Transition).

NEUREGULIN-1 SIGNALING PATHWAYS IN SCHWANN CELLS Neuregulin signaling is important in cell types other than Schwann cells, and the intracellular signaling pathways that transduce neuregulin signals have been extensively studied. However, it is not known how many of the interacting proteins involved in the neuregulin signaling pathway identified in other cells are used by Schwann cells, as reviewed by several authors.33,99,165,175,234 In Schwann cells, neuregulin-1 induces proliferation, phosphorylation, and dimerization of the neuregulin receptors ErbB2 and ErbB3, processes potentiated by interaction with the transmembrane protein CD44, which also promotes neuregulin-1–induced cell survival.108,258,266,269,296,318,322,345,361 (CD44 is widely expressed. The extracellular matrix component hyaluronan binds to it, and it plays a role in migration, adhesion, and survival of

Molecular Signaling in Schwann Cell Development

some cell types.) In Schwann cells, as in other cell types, neuregulin-1 induces activation of phosphatidylinositol 3 (PI3) kinase and its target Akt, and mitogen-activated protein (MAP) kinase pp44/42. Inhibition of the PI3 kinase pathway inhibits DNA synthesis in Schwann cells in response to applied neuregulin-1 and contact with neurites.147,198 In these studies, inhibition of the MAP kinase pathway had no effect, suggesting that the PI3 kinase pathway is more important in the proliferative response. Schwann cell survival in response to neuregulin-1 also requires activation of PI3 kinase and its effector Akt and subsequent phosphorylation and inactivation of the proapoptotic protein BAD.70,179,198,200 In some studies, but not in all, inhibition of the MAP kinase pathway also caused considerable cell death.198,200,249 In Schwann cell precursors, inhibition of either the PI3 kinase or the MAP kinase pathway completely inhibits survival in neuregulin-1.70 Neuregulin-1 also regulates the Jun N-terminal kinase (JNK) and p38 stress kinase pathways, both of which are likely to be important in Schwann cell biology (see Transcription Factors Involved in Myelination below). In breast cancer cells and in Schwann cells, neuregulin-1 activates the p38 MAP kinase pathway, and inhibition of this pathway prevents neuregulin-1–induced proliferation230 (D. B. Parkinson, A. Bhaskaran, K. R. Jessen, and R. Mirsky, unpublished observations, 2003). The JNK pathway is also activated by neuregulin-1. Phosphorylation (activation) of c-Jun is required for Schwann cell proliferation in response to neuregulin,249 and is also required for TGF-␤–induced cell death (see Signals That Control Schwann Cell Survival below).248 Elevation of intracellular cAMP levels and protein kinase A (PKA) activation have profound effects on both proliferation and differentiation of Schwann cells in vitro (see Transcription Factors Involved in Myelination below). Neuregulin-induced Schwann cell proliferation involves and requires a delayed rise in intracellular cAMP levels and PKA activation.147 This response may account for the observation that, unlike most other growth factors, neuregulin-1 promotes DNA synthesis in the absence of other agents that stimulate cAMP pathways, although both forskolin (which elevates cAMP levels) and IGF potentiate the proliferative response to neuregulin-1.61,310 In Schwann cells, neuregulin-1 also activates pp70s6 kinase (a downstream effector of PI3 kinase; see below), PAK65 (a component of stress-activated signaling pathways), p95RSK2 (a cAMP response element [CREB] kinase activated by MAP kinase), and the transcription factor CREB by phosphorylation on serine 133. It also stimulates cyclin D1 expression and phosphorylation of retinoblastoma protein, both enhanced by addition of forskolin.258,324 (Cyclin D1 and retinoblastoma protein are both involved in regulation of the cell cycle.) There is also evidence that neuregulin-1 signaling

353

involves protein kinase C, another potential effector of PI3 kinase.276,369 Both neuregulin and autocrine-mediated survival of Schwann cells (see Signals That Control Schwann Cell Survival below) depend on the activity of Ets transcription factors, several of which are made by Schwann cells.249 In a related system, neuregulin-mediated stimulation of acetylcholine receptor synthesis at the neuromuscular junction also depends on an Ets transcription factor binding site.94 (Ets transcription factors bind to a specific DNA core sequence [GGAA/T] and cooperate with other transcription factors and cofactors to regulate proliferation, apoptosis, and development and differentiation of specific cell lineages.)

REGULATION OF THE PRECURSOR–SCHWANN CELL TRANSITION When Schwann cells are generated from Schwann cell precursors in vivo, a coordinated change in a number of disparate and apparently unrelated phenotypic features occurs. There are modifications in antigenic expression (see Fig. 16–2), survival regulation, response to mitogens, motility, and cell-cell interactions (see Schwann Cell Precursors above). Similar changes can be accomplished in vitro by culture of precursors from rat E14 nerves in the absence of neurons in chemically defined medium containing neuregulin-1 for 4 days (E14 ⫹ 4 ⫽ E18). When these cells are dissociated and re-plated, over 80% of them have a Schwann cell rather than precursor phenotype with respect to the presence of autocrine survival loops (see Signals That Control Schwann Cell Survival below), antigenic phenotype, and mitogenic responses.69 This demonstrates that neuregulin-1 alone is sufficient to support not only precursor survival but also lineage progression. A minority of precursors fail to convert to Schwann cells in these simplified conditions, and it is probable that the appropriate rate of Schwann cell generation in vivo involves multiple signals.26,171 Endothelin, acting through the endothelin B receptor, is likely to be one of these signals.26 The survival of Schwann cell precursors in vitro is supported by endothelins 1 through 3, and endothelins and endothelin receptors are present in embryonic nerves. Unlike neuregulin, endothelin does not induce precursor proliferation, and these two factors also affect lineage progression differently. In precursor cultures exposed to endothelin alone, Schwann cells are generated very slowly, but in the combined presence of endothelin and neuregulin-1, the rate of Schwann cell generation increases and is intermediate between that seen in endothelin alone or neuregulin-1 alone. This result suggests that neuregulin actively promotes the transition of precursors to Schwann cells. It

354

Neurobiology of the Peripheral Nervous System

also indicates that endothelins act to slow down Schwann cell generation, because Schwann cells are generated more slowly in endothelin plus neuregulin-1 than in neuregulin-1 alone. In vivo experiments on the spotting lethal rat confirm this idea. In these animals, endothelin B receptors are not functional and Schwann cell generation occurs ahead of schedule, as expected if, in normal nerves, endothelin B receptor activation acts to brake the rate of lineage progression.26 It is possible that FGF-2 may be a third regulatory signal in this system, because addition of FGF-2 accelerates Schwann cell generation in vitro in comparison with that seen in neuregulin-1 alone.70 This remains to be confirmed in vivo. The transition from precursors to Schwann cells involves systematic alterations in gene expression, implying, in turn, the existence of a coordinating gene control mechanism. At present, however, only one transcription factor, AP2␣, has been linked to this process.311 At the precursor–Schwann cell transition in vivo, both in rats and in mice, AP2␣ expression is sharply downregulated, and in vitro, enforced expression of AP2␣ in precursors delays their conversion to Schwann cells, suggesting that the rate of AP2␣ downregulation may be one of the factors that determines the rate of Schwann cell generation.311 There is evidence that the time course of differentiation of a related cell type, the oligodendrocyte progenitor cell, is controlled by a member of the Id family of HLH genes, Id4, the levels of which drop as the cell differentiates into an oligodendrocyte.152,348 Although Schwann cell precursors and Schwann cells coexpress Id1 through Id4, these factors are not obviously regulated during embryonic nerve development, and at present there is no evidence that Id proteins are involved in controlling the speed of the precursor–Schwann cell transition.

SIGNALS THAT CONTROL SCHWANN CELL SURVIVAL Axon-associated neuregulin-169,108,322,340 and autocrine Schwann cell signals51,200 are two signals that play a major role in promoting the survival of developing Schwann cells. Transection of neonatal sciatic nerves results in the death of all teloglia at the neuromuscular junction.340 Although most Schwann cells within the main nerve trunk survive, there is also increased Schwann cell death within the nerve, indicating a partial dependence on axonal signals. Since the cells can be rescued by application of exogenous neuregulin-1, it is likely that the death is caused, at least in part, by loss of contact with axon-associated neuregulin as axons degenerate after transection.108,322,340 After the first postnatal week, there is no comparable Schwann cell death immediately following nerve transection in rats, and in adult animals Schwann cells in the distal stumps survive for several months after nerve transection, although their number slowly decreases and they respond less to extrinsic signals.108,178,319,340 Significant Schwann cell death is, however, seen in regenerating adult nerves at 3 weeks postcrush at a time when there is a requirement to match the number of axons and Schwann cells.89 Schwann cell survival in the absence of axons is crucial for axonal regeneration after nerve injury, because Schwann cells provide both adhesive substrates and trophic factors that promote axonal growth. A major reason for Schwann cell survival in these conditions is likely to be the presence of autocrine circuits, which enable Schwann cells to support their own survival51,200 (Fig. 16–4). Schwann cell precursors, which fail to survive without axons, do not possess autocrine survival circuits and appear to be wholly dependent on axonal neuregulin-1 signaling for survival.70,200,360 Despite the reports that Schwann cells can make low levels of neuregulin-1,37,269 the

FIGURE 16–4 Survival regulation in the Schwann cell lineage. Both axon-derived and autocrine signals regulate survival of cells in the Schwann cell lineage. Schwann cell precursors rely exclusively on axonal signals, principally neuregulin-1, for survival, but as development proceeds and precursors generate Schwann cells, there is a shift to the establishment of autocrine loops, the main components of which are likely to include IGF-2, PDGF-BB, NT-3, LIF, and a LPA-like activity. (From Jessen, K. R., and Mirsky, R.: Schwann cell development. In Lazzarini, R. A. [ed.]: Myelin Biology and Disorders, Vol. 1. San Diego, CA, Academic Press, p. 329, 2003, with permission.)

Molecular Signaling in Schwann Cell Development

autocrine Schwann cell survival signal is unlikely to be neuregulin-1 because the autocrine signal is not mitogenic for precursors and does not support their survival.51,200 Major components of the autocrine loop have been identified as IGF-2, combined with neurotrophin-3 (NT-3) and plateletderived growth factor-BB (PDGF-BB).200 These factors support survival when applied together in very low concentrations to Schwann cells that are plated so sparsely that they would die without added survival signals. Schwann cells also express receptors for these factors in vivo and in vitro. Furthermore, using this assay, the survival activity present in Schwann cell–conditioned medium is blocked by antibodies to these proteins. IGF is also a potent Schwann cell survival factor in an alternative survival assay in which Schwann cell death is induced by abrupt serum deprivation. In these experiments, IGF acts via PI3 kinase and Akt to inhibit activation of JNK and prevent caspase-mediated apoptosis.35,49,67,323 Leukemia inhibitory factor (LIF) and lysophosphatidic acid (LPA) are other potential autocrine Schwann cell survival factors. Denervated Schwann cells secrete LIF, which can promote Schwann cell survival in the presence of other growth factors.71 LPA also promotes survival in the serum deprivation assay, and densely plated Schwann cells secrete an LPA-like activity into the medium. LPA signals through the Gi protein–coupled lpA1 receptor, which in turn activates the PI3 kinase pathway and Akt.354,355 Longer term survival in the presence of autocrine signals is enhanced by culture on a laminin substrate, although laminin alone does not support survival.200 As mentioned earlier, the activity of Ets transcription factors is required for Schwann cell survival mediated by neuregulin-1 or by Schwann cell–conditioned medium, whereas it is not needed for LPA-induced survival.249 In summary, the survival of Schwann cell precursors is acutely dependent on axonal neuregulin-1, and they do not possess autocrine survival circuits, whereas Schwann cells can save themselves from death by secretion of several survival factors that are likely to include IGF-2, PDGF-BB, NT-3, LIF, and LPA. An important consequence of this is that, in older nerves, Schwann cells can survive in the absence of axons for prolonged periods. In the perinatal period, axonal neuregulin-1 is likely to contribute additional survival support for Schwann cells in parallel with that provided by autocrine signals. In addition to positive signals that promote survival and proliferation, factors that actively promote apoptosis may also contribute to the regulation of Schwann cell numbers in developing nerves. There is evidence that two such factors, nerve growth factor (NGF) and TGF-␤, can act in this way.248,301 NGF, acting via p75NTR, promotes cell death in several systems, including Schwann cells. However, the effect is complex because, depending on conditions, NGF can promote either Schwann cell death or survival. This underlines the importance of the total cellular and signaling context in the balance between survival and death.145 There

355

is less Schwann cell death in nerves of p75NTR-null mice than in normal mice following neonatal nerve transection or in regenerating adult nerves 3 weeks after nerve crush.89,321 When deprived of serum and growth factors, cultured expanded neonatal Schwann cells from these mice survive better than normal Schwann cells,91,301,321 and antibodies to p75NTR block apoptosis in Schwann cells cultured from transected adult rat nerve distal stumps.123 Studies on a protein that interacts specifically with p75NTR, receptor-interacting protein (RIP)-2, highlight the complexity of the Schwann cell response to p75NTR activation. In the presence of RIP-2, Schwann cell death in response to NGF is prevented through activation of the transcription factor NF-␬B. RIP-2, a protein kinase that contains a caspase recruitment domain (CARD), is expressed by freshly isolated Schwann cells, but not by Schwann cells that have been cultured for a prolonged period. Schwann cells that do not express RIP-2 are sensitive to NGF-induced cell death, and transfection of constitutively active RIP-2 confers protection against death. Conversely, expression of dominant-negative RIP-2 in freshly cultured RIP-2–expressing Schwann cells confers sensitivity to NGF-induced cell death.145 Tumor necrosis factor (TNF) receptor–associated factor (TRAF)-6 is another factor that associates with p75NTR. This activates the JNK pathway in Schwann cells (see below), and may be responsible for the NGF-induced death that occurs in the absence of RIP-2.145,146 Expression of other p75NTR-associated proteins has been correlated with cell death in immortalized cell lines, but their involvement in Schwann cell death has not been demonstrated. Among these proteins are the zinc finger proteins neurotrophin receptor interacting factor (NRIF)-1 and -2, p75NTR-associated cell death executor (NADE), and neurotrophin receptor–interacting tumor suppressor MAGE (NRAGE), which also mediates cell cycle arrest in sympathetic neuroblasts.14,39,144,221,222,275 In addition, two other proteins that do not appear to be involved in cell death have been shown to bind to p75NTR. These are the zinc finger protein Schwann cell factor-1, which promotes growth arrest in Schwann cells, and Rho(A) GTPase, which promotes neurite elongation in PC12 cells, and which has been implicated in actin-based movement in many cell types.54,367 TGF-␤s, like NGF, can induce apoptotic cell death in neonatal Schwann cells both in vitro and in vivo.248 Freshly isolated neonatal Schwann cells cultured in serum-free media die by apoptosis in response to TGF-␤1 application, and increased cell death is seen when TGF-␤ is injected into transected, but not normal, neonatal nerves. In vitro, a combination of neuregulin-1 and the autocrine signal cocktail of IGF-2, PDGF-BB, and NT-3 completely blocks apoptosis. TGF-␤1 signals via JNK and phosphorylation of c-Jun, since adenoviral expression of dominant-negative c-Jun inhibits TGF-␤ induced apoptosis, while retroviral expression of constitutively active Jun (v-Jun) promotes cell death.

356

Neurobiology of the Peripheral Nervous System

Myelinating Schwann cells isolated from postnatal day 4 nerves are resistant to TGF-␤–induced cell death, and are unable to phosphorylate c-Jun in response to TGF-␤.248 In vitro, enforced expression of the myelin-related transcription factor Krox-20 is sufficient to make Schwann cells resistant to TGF-␤–induced death246,247 (see Transcription Factors Involved in Myelination below). In the presence of serum, a combination of TGF-␤ and TNF-␣, but neither alone, induces Schwann cell death. This combination also kills freshly isolated Schwann cells more effectively than TGF-␤ alone in serum-free conditions.245,299 In immortalized Schwann cells, TNF-␣ in combination with interferon-␥ also induces death via production of nitric oxide and ceramide.224 The actions of TGF-␤ are strikingly context dependent. In addition to inducing apoptosis in defined medium, it can promote Schwann cell DNA synthesis in the presence of serum or cAMP-elevating agents or both.75,264 Another potential of TGF-␤ is revealed in neuron–Schwann cell co-cultures, wherein TGF-␤ inhibits the Schwann cell proliferation induced by contact with neurites and blocks myelination.76,110 TGF-␤ also suppresses myelin gene expression in purified Schwann cell cultures.215 As described above, in perinatal nerves, the population of immature Schwann cells is exposed to survival signals (neuregulin-1 and autocrine factors), proliferative signals (principally neuregulin-1), and death signals (NGF and TGF-␤). These signals in combination are likely to provide a mechanism for matching the number of Schwann cells to axons. Two of the signals (TGF-␤ and neuregulin-1) can also suppress myelination (see Transcription Factors Involved in Myelination below) and could also play a role in timing the length of the promyelination phase of Schwann cell development.

CELL CULTURE STUDIES ON SCHWANN CELL PROLIFERATION AND DIFFERENTIATION Although neuronally derived neuregulin-1 (see Importance of Neuregulin-1 in Schwann Cell Development, and Neuregulin-1 Signaling Pathways in Schwann Cells, above) is usually thought to be the major mitogen in Schwann cell development, a variety of other growth factors present in neurons cause Schwann cell proliferation. These factors may therefore help to regulate proliferation in vivo.210 Two in vitro models have been employed to investigate proliferation and myelin-related differentiation in Schwann cells. The first model involves co-cultures of DRG neurons and Schwann cells. In this system, axon-induced proliferation and myelination occur sequentially in defined media with or without serum, and the effects of antibodies, growth factors, genetically engineered constructs, or intracellular pathway inhibitors (see Transcription Factors

Involved in Myelination below) can be studied. In particular, this model has established that neuregulin-1 is the primary axonal mitogen for Schwann cells and has established the importance of the basal lamina in myelination.31,32,78,177,220 In the second system, intracellular cAMP levels are elevated in purified cultures of Schwann cells, grown in media with or without identified growth factors or serum, to mimic the effects of axon-induced proliferation and differentiation. In this model, elevation of cAMP levels in the absence of growth factors (see below) strongly stimulates expression of myelin-related proteins and lipids, including P0, PMP22, periaxin, and galactocerebroside, without inducing DNA synthesis, thus mimicking events that occur when myelination is induced by axons. Alternatively, when cAMP levels are elevated in the presence of growth factors or serum, cell division is stimulated, and elevation of proteins such as P0 in the population as a whole is much reduced (reviewed by Jessen and Mirsky138,139,210). Results using these two models largely complement one another. Because so many different protocols have been used to study proliferation in purified Schwann cell cultures, it is difficult to strictly compare results from different laboratories, particularly when activation of intracellular signaling pathways is investigated. It is important to note that cells prepared by different methods, including freshly isolated immunopanned cells, serum-purified cells, and cells expanded in growth factors and forskolin for short or long periods, may show different responses to applied reagents. In addition, Schwann cells in co-cultures with neurons may also respond differently from purified Schwann cells cultured alone. Studies on the regulation of DNA synthesis using serumpurified Schwann cell cultures have identified several growth factors, including FGF-1 and -2, PDGF-BB, TGF-␤s, and Reg-2, that induce Schwann cell DNA synthesis in the presence or absence of serum, provided cAMP levels are elevated and type 1 IGF receptors are activated. None of these factors induces significant Schwann cell proliferation when used alone in the absence of cAMP elevation48,181,310 (for review, see Eccleston74 and Mirsky and Jessen210). As mentioned above, neuregulin-1 induces Schwann cell proliferation without the obligatory presence of cAMP-elevating agents, presumably as a result of the ability of neuregulin-1 to stimulate cAMP pathways (see Neuregulin-1 Signaling Pathways in Schwann Cells above). IGFs act in Schwann cells and other cells as progression factors and potentiate the effects of most if not all Schwann cell mitogens, including neuregulin-1.47,48,61,181,310,315 While neuregulin-1 and IGFs both promote cell cycle progression in Schwann cells, IGFs, but not neuregulin-1, also increase cell size in the G1 phase of the cell cycle.60,61 Uniquely, the mitogenic effect of hepatocyte growth factor (HGF), which stimulates Schwann cell proliferation in the presence of low levels of serum, is inhibited by cAMP elevation.154

Molecular Signaling in Schwann Cell Development

At present, no comprehensive picture of all the pathways involved in the cAMP-dependent stimulation of proliferation exists, but three different sets of experiments suggest some of the mechanisms that may be used. Early studies suggested that a major function of cAMP in Schwann cell proliferation was to upregulate synthesis of receptors for Schwann cell mitogens such as PDGF.356 Recent studies have confirmed this observation, but indicate that other mechanisms are involved at earlier stages in the response.148,149 Using the interaction between cAMP and PDGF-BB as a model, they show that PDGF acts upstream of cAMP as a competence factor in the G1 phase of the cell cycle. The primary effect of cAMP is to induce sustained elevation of the G1 phase–specific protein cyclin D1, which is induced only transiently by PDGF alone. Ectopic expression of cyclin D1 alleviates the G1 phase requirement for cAMP, while the tumor suppressor gene Nf1 antagonizes accumulation of cAMP and expression of cyclin D1. cAMP is likely to act in similar ways in concert with other Schwann cell mitogens. In cyclin D1– or cyclin D2–null mice, Schwann cell proliferation proceeds normally in the early postnatal period. In contrast, little proliferation is seen in serumpurified cultured cyclin D1–null Schwann cells in response to either PDGF plus forskolin or neuregulin-1 or after nerve transection in cyclin D1–null mice. Thus cyclin D1 is required for proliferation in the absence of axons, either in neonatal Schwann cell cultures or in transected adult nerves, whereas it is not required for axonally driven proliferation during development.8,148 Experiments on Schwann cells and thyroid follicular cells have implicated the p70 ribosomal S6 kinase (p70s6k) in cAMP-dependent proliferation, because cAMP-dependent proliferation in both cell types is inhibited by rapamycin, a specific blocker of this kinase.40 Further investigation of this mechanism in thyroid follicular cells suggests that cAMP-dependent proliferation is dependent on the convergence of two separate cAMP-activated pathways on p70s6k. One pathway involves PKA-dependent activation of p70s6k while the other involves PI3 kinase–dependent, PKA-independent, activation of p70s6k.41,42 Interestingly, in Schwann cell–neuron co-cultures, where proliferation is driven by contact with neurites, retroviral inhibition of cAMP-dependent PKA does not inhibit mitosis, whereas it does so in purified Schwann cells exposed to cAMP-elevating agents and mitogens. Neurite-induced proliferation is nevertheless inhibited by PKA inhibitors such as H89 and KT5780, which are presumably less specific than the retrovirally mediated vectors. This suggests that, in development, neurite-induced proliferation may involve pathways other than those activated by PKA127,147 (see below). The cyclin-dependent kinases (Cdks) 2, 4, and 6 are involved in cell cycle progression, and levels are high in Schwann cells cultured under proliferating conditions.195

357

Cdk2 levels, which are controlled transcriptionally, are low in growth-arrested cells in vitro, high in early postnatal nerves, and low in adult nerves.330 Levels of p27, a cell cycle–inhibitory protein, are higher in arrested Schwann cells than in proliferating cells,330 and p27 is elevated by enforced expression of the transcription factor Krox20, which also induces growth arrest.245 Schwann cell growth arrest can be induced by constitutive activation of Ras or Raf kinase, which results in activation of the MAP kinase pathway. Arrest occurs in the G1 phase of the cell cycle via the induction of p21Cip1 (but not p27Kip1), with subsequent inhibition of cyclin/Cdk activity.182,265

TRANSCRIPTION FACTORS INVOLVED IN MYELINATION Transcription factors that are expressed by Schwann cells or their precursors are summarized in Table 16–1. Those most relevant to myelination are discussed here. Myelination is the final stage of Schwann cell differentiation. On myelination, Schwann cells radically transform their phenotype in response to signals from the larger axons. This response represents one of the most striking examples of cell-cell interaction that is known, and results in the generation of one of the most highly specialized cell types in the body, the myelinating Schwann cell. This differentiation is accompanied by reciprocal changes in axonal ion channels and cytoskeleton that enable saltatory conduction (Fig. 16–5; see also Fig. 16–3). Although many experiments indicate that myelination and myelin maintenance depend on the presence of axons, and on close contact between axons and Schwann cells, no myelination-inducing signal from axons has been identified, nor have receptors that signal myelin formation been described in Schwann cells. A clear demonstration at the molecular level of cell-cell signaling leading to myelination is therefore lacking. Despite this, several other key steps in the process have been clarified. The importance of the basal lamina in enabling myelination is well established, and the role of extracellular matrix proteins and their receptors in myelination is discussed below (see Extracellular Matrix, Matrix Receptors, and the Cytoskeleton in Schwann Cell Development) and has been reviewed by several authors.32,53,256 Significant progress has also been made in the analysis of transcriptional mechanisms that control myelination. The crucial importance of the transcription factor Sox10 in early development of the glial lineage has already been outlined (see Neural Crest Origins of Schwann Cells above), but this factor also participates in later lineage events (see below). Two other transcription factors, the POU domain factor Oct6 and the zinc finger protein Krox20, have important roles in myelination15,133,175,213,337 (for reviews of these and other transcription factors in the Schwann cell lineage, see Mirsky

358

Neurobiology of the Peripheral Nervous System

Table 16–1. Transcription Factors Associated with Schwann Cell Precursors and/or Schwann Cells* Factor

DNA-Binding Domain

bZIP Factors AP2 CREB C/EBPa E12/47 c-Fos c-Jun Jun D Jun B

bZIP bZIP bZIP bZIP bZIP bZIP bZIP bZIP

Ets Factors ER81 ERM† Ets2 GABP-␣ Net

Ets Ets Ets Ets Ets winged HLH

HLH Factors Hey1 Id1, -2, -3, -4 REB

bHLH bHLH bHLH

POU Homeodomain Factors Brn2 Brn5 Oct1 Oct6 Zinc Finger Factors Krox20 Krox24 (NGF-1A/Egr1) KZF-1 SCp55 ZFP57 Other Factors Elongin A fkd6‡ NF-␬B NT3R Pax3 Sox10 Steroidogenic acute regulatory protein TFIIA large subunit

In Vivo

In Vitro

⫹ ⫹ ⫹ ⫹

⫹ ⫹ ⫹ ⫹

⫹ ⫹ /⫺ ⫹ /⫺

⫹ ⫹ ⫹

312 147, 170, 309, 324 18, 128 314 226 66, 213, 309 309 309

114, 242, 249

⫹ ⫹

⫹ ⫹

POU homeodomain POU homeodomain POU homeodomain POU homeodomain

⫹

⫹

⫹ ⫹

⫹ ⫹

Zn finger Zn finger Zn finger Zn finger Zn finger with KRAB domain

⫹ ⫹

⫹ ⫹

forkhead domain NF-␬B/Rel family Nuclear hormone receptor Paired box domain HMG domain START domain

⫹ ? ⫹ ⫹ ⫹

? ⫹ ? ⫹ ⫹

Four-helix bundle (FHB)

References

225 314, 328 314 132, 226, 298 366 22 15, 120, 133, 211–213, 357 22, 223, 337, 375, 376 158, 194, 226, 236, 335 226 161 72, 238 226 143 38, 233 12 22, 150 27, 157, 303 226 226

bZIP ⫽ basic leucine zipper motif; NF ⫽ nuclear factor; NT3R ⫽ nuclear triiodothyronine (T3) receptor; TF ⫽ transcription factor. *This list is almost certainly not comprehensive. The many gene array studies now underway using Schwann cells are likely to reveal additional important Schwann cell transcription factors in the coming years. † Note that this transcription factor is more highly expressed in neurons and satellite cells than in Schwann cell precursors or Schwann cells.114 It can nevertheless be detected by reverse transcription–polymerase chain reaction in Schwann cells.249 ‡ So far described only in zebra fish Schwann cells and satellite cells.143

and Jessen209 and Topilko and Meijer336). Myelination is severely delayed in Oct6-null mice,15,133 while Schwann cell myelination fails completely in Krox20-null mice.336,337 In both these mice, Schwann cells successfully achieve a 1:1 relationship with the larger axons and wrap

around them up to one and a half times before myelination is arrested temporarily (Oct6 null) or permanently (Krox20 null). In humans, Egr2 (Krox20) mutations are found associated with Charcot-Marie-Tooth disease, Déjérine-Sottas disease, and hereditary sensory and motor

Molecular Signaling in Schwann Cell Development

359

FIGURE 16–5 The main changes in protein and lipid expression that take place as the population of immature Schwann cells diverges to generate myelinating and nonmyelinating cells during development. Myelination involves a combination of downregulation and upregulation of key molecules, most of which are associated with formation of the myelin sheath. Dark gray box: Molecules that are strongly upregulated on myelination. Light gray box: Molecules associated with immature Schwann cells that are downregulated when cells myelinate. The generation of mature nonmyelinating cells from immature cells involves many fewer molecular changes, and most of the molecules typical of immature Schwann cells are also expressed by nonmyelinating cells, as indicated. Galactocerebroside (GalC) is expressed by both myelinating and nonmyelinating Schwann cells, but not by immature cells, while ␣1␤1 integrin is upregulated specifically on nonmyelinating cells as they mature. 04 antigen and S100 are expressed by all Schwann cells in peripheral nerves, although 04 is downregulated in the absence of axons. Hereditary demyelinating neuropathies that result from abnormal expression of some of the myelinassociated proteins are indicated. CMT ⫽ Charcot-Marie-Tooth neuropathy; PMD ⫽ Pelizaeus-Merzbacher disease. (From Jessen, K. R., and Mirsky, R.: The Schwann cell lineage. In Kettenmann, H., and Ransom, B. R. [eds.]: Neuroglia, 2nd ed. Oxford, UK, Oxford University Press, 2004, with permission.)

neuropathies, emphasizing the crucial importance of Krox20 in myelination.13,23,244,331,350,351,368 Schwann cells that have been signaled to myelinate express Krox20 selectively and at high levels (in the mouse from E16 onward).247,375,376 In contrast, Oct6 can be detected at low levels in late precursors/early Schwann cells from about E13 (mouse)/E15 (rat) onward. It is expressed by all Schwann cells in late embryogenesis and the early postnatal period, with highest expression in promyelin cells, declining to very low levels in adult

nerves.7,22 Schwann cells in peripheral nerves of Oct6-null mice do not express Krox20 at the appropriate time but do so after developmental delay of about 2 weeks just before they myelinate. The ability of Oct6-null Schwann cells to myelinate eventually is likely to be due to the activity of a related POU domain transcription factor, Brn2.132 Brn2 and Oct6 are expressed independently of each other but with a similar time course in developing nerves. While myelination is unaffected when Brn2 alone is knocked out, the myelination delay seen in Oct6-null nerves is much

360

Neurobiology of the Peripheral Nervous System

more severe when Brn2 is also missing. Conversely, overexpression of Brn2, achieved by driving Brn2 expression under the control of Oct6 Schwann cell enhancer (SCE) (see below), partially rescues the Oct6-null phenotype. Thus Brn2 can substitute for Oct6 in the initiation of myelination. It is intriguing that, even in the combined absence of Brn2 and Oct6, myelination is not permanently blocked, and that a large number of fibers eventually myelinate.104,132,298,336 Schwann cells in nerves of Krox20-null mice express Oct6 on time, but in the early postnatal period maintain expression at a higher than normal level. Thus Krox20 may participate in downregulating levels of Oct6. It may also regulate cell proliferation and death, because Schwann cells in these mice have elevated DNA synthesis and death rates.104,336,376 On nerve transection, Krox20 levels drop rapidly and both Oct6 and Krox20 are reexpressed in regenerating nerves and in cultured Schwann cells on elevation of cAMP, indicating that they are regulated by axons, like myelin protein genes.223,376 Gene array technology has been used to show that enforced expression of Krox20 in cultured Schwann cells is sufficient to induce a large set of diverse genes, including those involved in myelin protein and lipid synthesis, together with many other genes of unknown function in the Schwann cell.226 These and other studies245–247 demonstrate that enforced expression of Krox20 is sufficient to organize a large number of phenotypic changes in cultured Schwann cells, all of which, in vivo, are linked to the transition from immature proliferating cells to quiescent myelinating cells. Thus, in Schwann cells expressing Krox20, proliferation induced by the major axonal mitogen neuregulin-1 is blocked and cell death induced by TGF-␤ is inhibited. 137,245 The myelin proteins periaxin, myelin-associated glycoprotein, and P0 are activated and L1, a marker of nonmyelinating and immature Schwann cells, is downregulated, as occurs on myelination in vivo.192 Furthermore, Krox20 induces expression of periaxin and P0 and inhibits DNA synthesis and apoptosis in an unrelated fibroblastic cell type, the 3T3 cell,245 behavior that is reminiscent of the action of master regulatory genes such as MyoD, neural bHLH factors, or peroxisome proliferator–activated receptor (PPAR)-␥.65,167,183,272,334,376 (PPARs are involved in the regulation of lipid metabolism.) Examination of the molecular mechanisms used by Krox20 in Schwann cells has revealed an important interaction between Krox20 signaling and the JNK/c-Jun pathway.245 These studies indicate, first, that the JNK/c-Jun activity acts as a brake on myelin differentiation and promotes proliferation and death, and, second, that Krox20 expression is sufficient to inactivate this pathway. This would provide Krox20 with a mechanism for coordinated control of myelination, proliferation, and cell death in developing nerves.

The gene promoters of both the Oct6 and Krox20 genes contain SCE elements, which are controlled by axonally regulated transcription factors. (SCE sequences function to specifically regulate gene expression levels in Schwann cells.) The Oct6 SCE, which resides within a 4.3-kb sequence 12 kb downstream of the transcription initiation site, is sufficient to drive spatially and temporally correct expression of the gene both developmentally and during regeneration.104,190 The Krox20 promoter contains two widely separate elements. One of these, the immature SCE, situated upstream of the transcription start site, is active in immature Schwann cells from E15, presumably in promyelinating cells but not in actively myelinating cells. The other, the myelinating SCE, is active from E18 onward in myelinating cells. It is located within a 1.3-kb sequence positioned 35 kb downstream of the Krox20 open reading frame. It is dependent on Oct6 for activation and contains multiple Oct6 binding sites. In regenerating nerves sequential reexpression of both enhancers is observed.105 In addition to its function in promoting gliogenesis from the neural crest early in development (see Neural Crest Origins of Schwann Cells above), evidence suggests that Sox10 may also regulate later steps in the Schwann cell lineage. Sox10 can function synergistically with Oct6 to promote gene expression of reporter constructs containing recognition sites for both Sox10 and Oct6. Sox10 also interacts positively with Pax3 in studies using reporter constructs with Sox10 and Pax3 binding sites, but shows a negative interaction with Krox20 in comparable experiments.156 It positively regulates a P0 promoter construct in N2A neuroblastoma cells, and it functions synergistically with Krox20 to activate a gap junction connexin 32 (Cx32) promoter construct in N2A neuroblastoma cells and HeLa cells.24,253 Cx32 is expressed by myelinating Schwann cells and mutations in the Krox20/Sox10-binding region of the human Cx32 promoter are among the more than 80 mutations that result in X-linked Charcot-Marie-Tooth neuropathies. Furthermore, mutated forms of Sox10 and Krox20 identified in patients with Charcot-Marie-Tooth disease fail to transactivate the Cx32 promoter.24 Thus Sox10 has important functions in both the early and later stages of peripheral glial development. It should be noted that Sox10 is also required for terminal differentiation of oligodendrocytes, the myelinating cells of the CNS.317 Recently, a role for the transcription factor NF-␬B in enabling myelination has been shown.233 NF-␬B is highly expressed in premyelinating Schwann cells in sciatic nerves, but like Oct6 is downregulated in adult nerves. In myelinating neuron–Schwann cell co-cultures, NF-␬B is activated prior to Oct6 upregulation. Treatment with an inhibitor of NF-␬B or infection with a super-suppressor form of the NF-␬B inhibitory protein I␬B␣ prevents myelination, and Schwann cells fail to ensheath axons. Likewise, when Schwann cells null for the p65 subunit of

Molecular Signaling in Schwann Cell Development

NF-␬B are co-cultured with DRG neurons, myelination is severely reduced and cells fail to elongate properly along axons, making it likely that NF-␬B is required for Schwann cells to enter the promyelinating stage.233 HLH factors regulate both neuronal and muscle differentiation, and roles for HLH factors in regulating Schwann cell myelination have been suggested but not proved. The HLH genes expressed in Schwann cells include the B class HLH factor Mash2, which is expressed in a subset of Krox20-positive myelinating Schwann cells in adult nerves. (Mash 2 is required for development of the placenta.) Mash2 inhibits Schwann cell proliferation in culture and is expressed at lower levels after nerve transection. It regulates the expression of Krox24, the chemokine Mob1, and the chemokine receptor CXCR4 genes in vitro.160 In addition, the four HLH Id genes, which inhibit binding of B class HLH proteins to their A class binding partners such as REB and E12/47, are found in the Schwann cell lineage from the precursor stage onward.314 Id1 and -3 mRNAs are expressed at lower levels when peak myelination is occurring in 10-day nerves than in adult nerves. Both Id1 and -3 strongly repress myelin gene promoter activity, suggesting that the downregulation of myelin gene expression that occurs in adult nerves may involve these factors.328 The zinc finger transcription factor Krox24, which is closely related to Krox20 and binds to some of the same DNA sequences (see Topilko et al.335 and references therein), is expressed in mouse Schwann cell precursors from E11/12 and persists in nonmyelinating cells in adult nerves, being upregulated after nerve cut335 (but see Kury et al.160). In peripheral nerves of Krox24-null mice, myelination and regeneration appear to be normal, but after transection of neonatal nerves there is increased cell death compared with wild-type nerves.119,338

SIGNALING PATHWAYS AND GROWTH FACTORS THAT REGULATE MYELINATION As mentioned above, cell culture studies indicate that cAMP pathways are important in myelination. In neuron–Schwann cell co-cultures, retroviral inhibition of PKA inhibits myelination,127 and in purified Schwann cell cultures elevation of intracellular cAMP, particularly in nonproliferating conditions in the absence of growth factors, stimulates the expression of myelin-related molecules such as Oct6, Krox20, and myelin proteins and lipids. Thus elevation of cAMP partially mimics the axonal signals that initiate myelination in vivo.141,174,203,214,215,223,247,280,300 Furthermore, Schwann cells that express high levels of P0 in response to cAMP elevation also downregulate expression of GFAP, neural cell adhesion molecule, and p75NTR, all of which are downregulated in myelinating cells in vivo214 (see Fig. 16–4).

361

Proliferation is induced in Schwann cell cultures when cAMP-elevating agents and growth factors are present together, as discussed above (see Cell Culture Studies on Schwann Cell Proliferation and Differentiation). DNA synthesis is not induced, however, when cAMP pathways are activated in the absence of growth factors or serum.214 When cAMP-induced P0 expression in the presence (proliferating conditions) and absence (nonproliferating conditions) of growth factors is compared, much higher P0 levels are seen in the absence of growth factors.214 Furthermore, in individual cells, high P0 protein and mRNA levels are not seen in cells undergoing DNA synthesis.214,215,313 Therefore, cAMPinduced myelin-related differentiation in vitro is incompatible with proliferation, as is axonally induced myelination in vivo. Because cAMP elevates Krox20 levels in these cultures, both the upregulation of myelin proteins and cessation of proliferation are likely to be mediated by this transcription factor (see above). As mentioned before, IGFs promote cAMP-dependent Schwann cell proliferation in the presence of other growth factors (see Cell Culture Studies on Schwann Cell Proliferation and Differentiation above). In nonproliferating Schwann cell cultures, IGFs promote cAMP-induced P0 expression in the absence of growth factors310 and myelination in Schwann cell–DRG neuron co-cultures.47,48,271 They also promote the attachment and ensheathment of Schwann cells to axons, perhaps by enhancing Schwann cell motility.48,50 IGFs therefore have the potential to promote Schwann cell motility, proliferation, and differentiation. There is evidence that the PI3 kinase and Akt pathway, which is stimulated by IGFs, may be involved in early events of myelination in neuron–Schwann cell co-cultures, because pharmacologic block of PI3 kinase, but not MAP kinase, reversibly inhibits myelination, but not myelin maintenance, perhaps by interfering with Schwann cell–axon interactions that occur early in myelination.198 The rate of myelination can also be regulated by the neurotrophins brain-derived neurotrophic factor (BDNF) and NT-3. Reduction of BDNF signaling retards myelination during peripheral nerve regeneration and in neuron– Schwann cell co-cultures, while increasing BDNF levels enhances myelination.43,373 NT-3 acts in the opposite way and inhibits myelination.43 Progesterone increases the rate of myelination both in vivo and in vitro, although the evidence suggests that it does not affect the eventual extent of myelination44,151,237 (for reviews see Magnaghi et al.189 and Schumacher et al.285). In addition to availability from blood, both neurons and Schwann cells can synthesize progesterone from cholesterol, and there is evidence that both neurons and Schwann cells possess classic and nonclassic progesterone receptors, with higher expression in neurons45 (for review see Magnaghi et al.189 and Schumacher et al.285). Several studies suggest that exposure to progesterone elevates P0 and PMP22 mRNA in cultured Schwann cells and causes transient elevation of Krox20

362

Neurobiology of the Peripheral Nervous System

mRNA and protein.68,111,189,237,285 These experiments imply a direct action of progesterone on Schwann cells. Conversely, Chan and colleagues, using Schwann cell–DRG neuron co-cultures, provided evidence that progesterone synthesized by Schwann cells acts on progesterone receptors in neurons to activate neuronal target genes, which might in turn regulate Schwann cell myelin genes.45 Another growth factor, glial cell line–derived neurotrophic factor (GDNF), induces Schwann cell proliferation and promotes myelination of small-diameter axons that would normally not myelinate when it is administered exogenously to adult rats.125 It is argued that the direct effect of GDNF is to increase the size and axon diameter of the smaller neurons and that its effects on proliferation and myelination of Schwann cells are indirect. In the absence of axonal signals, growth-arrested Schwann cells do not myelinate, indicating that growth arrest alone is insufficient to trigger myelination. Conversely, myelinrelated differentiation can be inhibited without simultaneously stimulating DNA synthesis, as seen in experiments involving TGF-␤. This factor suppresses cAMP-induced P0, 04 antigen, and galactocerebroside induction in Schwann cell cultures, even at concentrations too low to induce Schwann cell proliferation,203,215,308 and has similar effects in myelinating neuron–Schwann cell cultures, suppressing DNA synthesis, P0 and galactocerebroside induction, and myelination.76,109 Thus TGF-␤s can override the myelin differentiation pathways activated either by axon-associated myelination signals or by elevation of intracellular cAMP, again suggesting that cAMP and axonal myelin signals activate common intracellular signaling mechanisms. Neuregulin-1 also inhibits myelination in DRG neuron– Schwann cell co-cultures and, unlike TGF-␤s, induces demyelination in cultures in which myelination has already taken place.370 In contrast, in transgenic mice, when neuregulin signaling is disrupted by controlled excision of ErbB2 driven by Krox20-Cre, hypomyelination is seen in peripheral nerves, suggesting that in vivo neuregulin signaling promotes rather than inhibits myelination.99,100,370

NEURAL ACTIVITY AND SCHWANN CELL DEVELOPMENT Neural activity may play a part in controlling glial development in peripheral nerves. Recent studies suggest that the phenotype of immature Schwann cells may be regulated by neuron-derived ATP, acting via P2-type purinergic receptors.3,59,187,188,199,306,307 (Purinergic receptors, which signal in response to extracellular adenine and uracil nucleotides, can be divided into the P2X family of ligandgated ion channel receptors and P2Y metabotropic G protein–coupled receptors.) In DRG neuron–Schwann cell co-cultures, electrical stimulation of neurons causes an

increase in Ca2⫹ concentration in the soma of individual neurons that is followed by delayed calcium elevation in Schwann cells associated with neurites of the same cell. This is caused by nonsynaptic ATP release from the neurons acting via P2 receptors.306 In response to ATP, cell proliferation is inhibited at the same time as Schwann cell differentiation, measured by 04 antigen expression and myelination, is delayed.306,307 Primary Schwann cells express purinergic receptors of the P2Y and P2X(7) subtypes,59,187 and P2U receptors are expressed by immortalized cells.17 The effect of ATP on proliferation in the co-culture system is probably caused by depression of adenylate cyclase activity induced by the action of P2Y receptors. As discussed earlier, in normal nerve development inhibition of proliferation is associated with the onset of myelination,30,313 and takes place well after 04 antigen is first expressed.70,207 Further studies will therefore be needed to determine how the effects of ATP signaling seen in these in vitro experiments integrate with other signals to generate the coordinated pattern of Schwann cell maturation, proliferation, and myelination seen in vivo. Related experiments using pure neuronal DRG cultures indicate that complicated patterns of neural activity lower neuronal expression of adhesion molecules such as N-cadherin and L1, which could affect interaction with Schwann cells.129,130 Low-frequency stimulation of neurons in DRG–Schwann cell co-cultures, which reduces axonal but not Schwann cell expression of L1, reduces myelination to one third of normal levels. Decreased axonal L1 also reduces Schwann cell adhesion to neurites in the cultures in a short-term assay.306,307 Some other studies also provide evidence for a role for L1 prior to myelination. Decreased L1 expression increases defasciculation both in vivo and in vitro,126,163,180 and function-blocking L1 antibodies in the neuron–Schwann cell co-culture system prevent Schwann cell alignment along axons, which in turn results in failure to myelinate.287,288,363 Despite this, in L1-null mice no major disturbance in development of either myelinated or unmyelinated fibers has been described, although disturbances in the maintenance of axon–Schwann cell relationships in sensory fibers occur in adult nerves, with subsequent axonal loss.64,117

EXTRACELLULAR MATRIX, MATRIX RECEPTORS, AND THE CYTOSKELETON IN SCHWANN CELL DEVELOPMENT Laminin Signals from the basal lamina and other extracellular matrix molecules are essential for the polarization and differentiation of many non–connective tissue cell types, including Schwann cells. In Schwann cells basal lamina forms just before myelination.31,193 The importance of this event for

Molecular Signaling in Schwann Cell Development

the myelination process was first demonstrated by Richard and Mary Bunge and their colleagues,31 and more recent experiments have built on this earlier work to identify the molecular interactions that occur between the major Schwann cell extracellular matrix molecules, their membrane receptors, and the intracellular signaling systems to which they are linked. The most important component of the basal lamina in this respect is laminin. Laminin exists in 11 different isoforms, each of which consists of different combinations of three subunits, designated ␣ (1 through 5), ␤ (1 through 3), and ␥ (1 and 2).256 The major isoform in the Schwann cell basal lamina is laminin 2 (merosin), consisting of ␣2, ␤1, and ␥1 chains, although smaller amounts of laminins 4 and 1 and other laminins are also present.131,176,227 In the dy/dy mouse there are mutations in the ␣2 chain, which is a constituent of laminins 2 and 4. These lead to defective laminin polymer formation and incomplete formation of the basal lamina. In addition to skeletal muscle atrophy, this mouse shows complex pathology of peripheral motor nerves, as reviewed by several authors.25,58,197,227 The most severe Schwann cell abnormalities are seen in the dorsal and ventral spinal roots, where undifferentiated Schwann cells surround but fail to penetrate naked axon bundles. More general defects seen not only in the roots but also within peripheral nerve bundles include a patchy basal lamina surrounding individual nerve fibers, hypomyelination, excessively wide nodes of Ranvier, and overlapping Schwann cells.2,343 The Schwann cell plasma membrane contains three major receptors for laminin 2. These are integrins ␣6 /␤1 and ␣6 /␤4, and ␣-dystroglycan, a member of the dystrophin complex of proteins.79,80,116,193,196 (Integrins are a large family of plasma membrane receptor molecules. They bind extracellular matrix components externally and cytoskeletal molecules internally, and thus link the extracellular matrix with the cytoskeleton. They consist of two subunits, ␣ and ␤.)

Integrin Signaling in Schwann Cells and Interactions with the Actin Cytoskeleton Schwann cells in vivo or in culture express a variety of integrins, including ␣1/␤1, ␣2 /␤1, ␣5 /␤1, ␣6 /␤1, ␣6 /␤4, an unidentified integrin (␣150 /␤1), ␤8 integrin, ␣v /␤3, and ␣v /␤1, plus very low levels of ␣4 and ␣5.76,87,206,314 The combinations of integrin receptors that are expressed by axons, Schwann cell precursors, and Schwann cells vary with the stage of peripheral nerve development and culture conditions (reviewed by Previtali et al.255). Conditional disruption of ␤1 integrin has demonstrated the importance of ␤1 integrin signaling in late embryonic and subsequent phases of peripheral nerve development in Schwann cells.83 Mice carrying a P0Cre transgene were crossed with mice carrying a floxed allele of ␤1 integrin,

363

resulting in excision of ␤1 integrin specifically from Schwann cells in peripheral nerves, starting between E13.5 and E14.5. Thus peripheral nerves can be populated by precursors and generate Schwann cells just before the ␤1 integrin is excised. The null Schwann cells survive and proliferate normally, but their relationships with axons are severely abnormal. Axons fail to segregate properly in many cases and fail to achieve the 1:1 relationship with the Schwann cell required for myelination. Despite this, a few axons are myelinated in adult mice, but even around these axons myelination is developmentally delayed. Aspects of the phenotype seen resemble those observed in the nerve roots of the laminin 2–deficient dy/dy mouse, and on this basis it is likely that the main partner for the ␤1 integrin is laminin 2, acting via the ␣6 /␤1 integrin receptor, which is highly expressed in embryonic Schwann cells.83,255 Absence of ␤1 integrin would then result in improper linkage between the laminin in the basal lamina with the Schwann cell cytoskeleton, a process that is required for axonal ensheathment.83 In addition to interactions between laminin and integrin receptors, there is considerable evidence that interactions between integrins and the cytoskeleton are important in Schwann cells. In myelinating Schwann cell–neuron cocultures, ␣6 /␤1 and ␣1/␤1 integrins are expressed by Schwann cells prior to myelination and antibodies to ␤1 block myelination, whereas antibodies to ␣1/␤1 do not, implying that ␣6 /␤1 is the receptor required for the onset of myelination.87,124 A block of myelination also occurs in response to drug-induced disruption of the actin cytoskeleton.86 Furthermore, in differentiating Schwann cell–neuron cocultures but not in pure Schwann cell cultures, ␤1 integrin co-precipitates in pull-down assays with focal adhesion kinase (FAK) and the actin-linked protein paxillin. (FAK is part of the signaling complex that forms when extracellular matrix components bind to integrins. This leads to clustering of the receptors, recruitment of actin filaments and signaling molecules to the cytoplasmic domain of the receptors, and attachment to the substrate.) As Schwann cells in co-cultures form basal lamina and differentiate, tyrosine phosphorylation of both FAK and paxillin also increases, suggesting that ␤1-mediated intracellular signaling is involved in the differentiation process.46 IGF-1, acting via a PI3 kinase pathway, activates FAK and promotes Schwann cell process extension and motility.50 Whether integrins take part in this process is not known, although it is likely given the co-localization of FAK, paxillin, and ␤1 integrin in filopodia.327 Another important Schwann cell protein, the tumor suppressor protein merlin/Schwannomin, also binds to paxillin, enabling it to anchor to the plasma membrane and associate with ␤1 integrin and ErbB2.88 Merlin is the product of the neurofibromatosis type 2 gene, and is a member of the ezrin/radixin/moesin family of proteins that links membrane proteins to the actin cytoskeleton in epithelia and other cell types.270,341 Mutations in merlin lead to an increased frequency of schwannomas, most sporadic schwannomas

364

Neurobiology of the Peripheral Nervous System

and meningiomas carry double mutations in merlin, and enforced expression of merlin in merlin-null cell lines inhibits cell proliferation, indicating a role for this gene in regulation of Schwann cell proliferation.284 In myelinating Schwann cells, it co-localizes at paranodal membranes with Rho(A), a protein heavily implicated in actin-based movement in other cell types, and interacts directly with the actin-binding protein ␤II spectrin and several other proteins, including HGF-regulated tyrosine kinase substrate, through which it regulates cell proliferation in schwannoma cell lines.278,286,320 Human schwannoma cells deficient in merlin have increased levels of activated Rac.140 (Rac is a Rho-GTPase associated with formation of lamellipodia in fibroblasts and other cell types.) This results in increased levels of activated JNK in the nucleus compared with normal human Schwann cells, although increased levels of phospho–c-Jun were not detected. This accords with earlier studies showing that merlin-deficient cells have increased membrane ruffling.292 While these observations indicate that merlin modifies the activity of Rac, a number of studies also showed that Rac controls the function of merlin.292,295 These effects of merlin loss are the converse of the effect of Krox20 expression in normal Schwann cells, namely downregulation of JNK, and of c-Jun and c-Jun activation (see Transcription Factors Involved in Myelination above). Perhaps related to this, merlin also interacts with the cytoplasmic tail of CD44 (see Importance of Neuregulin-1 in Schwann Cell Development above) to mediate contact inhibition of growth, while loss of merlin destabilizes adherens junctions.162,217,274 The phospholipid signaling molecule LPA (see Signals That Control Schwann Cell Survival above) also causes focal adhesion assembly of paxillin and vinculin, acting via the Rho/p160 Rho-associated kinase (ROCK) pathway; the rearrangement of the actin cytoskeleton into wreath-like structures; and induction of N-cadherin/catenin–mediated cell-cell contacts between adjacent Schwann cells.355 (Vinculin is a component of the FAK that binds actin. Catenins are multifunctional proteins that link cadherins to the actin cytoskeleton. They can also function as transcription factors.) Consistent with its distribution in vivo, N-cadherin mediates axon-aligned process growth of Schwann cells and Schwann cell–Schwann cell contacts in culture.297,349 Another actin-binding protein, dystonin, may also be involved in Schwann cell myelination. Schwann cells from mice with dystonin mutations have a severely disorganized cytoskeleton, fail to attach normally to laminin, and show abnormal myelination.16 Schwann cell migration, which is of importance in embryonic nerves and regeneration, is also likely to involve integrins. Antibodies to ␤1 integrin block Schwann cell migration on laminins 1 and 2 in response to neuregulin and forskolin. Antibodies to ␣6 /␤1 block migration on laminin 1 but not laminin 2, while RGD peptides block

migration on fibronectin, suggesting involvement of ␣v integrins.206 The MAP kinase pathway appears to be an important intracellular pathway used in migration on laminin,201 and integrins ␣3, ␣6, and ␤1 also associate with the tetraspan protein CD9, which has been implicated in Schwann cell adhesion, proliferation, and migration.5,113 In contrast to ␣6 /␤1, which is expressed earlier in development, the second main laminin receptor in Schwann cells, ␣6 /␤4 integrin, is strongly expressed in differentiated Schwann cells, including myelinating Schwann cells, although no peripheral nerve phenotype has been reported in ␤4 knockout mice77,83,84,87,235,344 (see Mirsky and Jessen208 for discussion of this point). Equally, strong ␤8 integrin immunolabeling is seen around myelin sheaths in mouse peripheral nerve, suggesting this protein might participate in signaling during myelination, but this has not been tested.206 Possible roles for another integrin, ␣1/␤1, are suggested by its in vivo distribution. In mature nerves, it is absent from myelinating cells but is found on mature nonmyelinating Schwann cells. It is expressed by all Schwann cells in adult nerve after transection, although it is not strongly expressed by Schwann cell precursors or Schwann cells in developing nerves. It might therefore be involved in the interaction of small-diameter axons with their enveloping Schwann cells, in the interaction of regenerating axons with Schwann cells, or in Schwann cell migration.314 Mice lacking ␣4 or ␣5 integrins provide evidence that these integrins play a role in the early embryonic development of the Schwann cell lineage.112 These experiments implicate ␣4 integrin in the survival of early developing glia, while ␣5 integrin is needed for early glial proliferation.

Dystroglycan Signaling in Schwann Cells The third main Schwann cell laminin receptor, dystroglycan, is involved in the assembly and maintenance of the basal lamina. When laminin binds to dystroglycan, it provides a scaffold for the assembly of a nascent basal lamina, which can then incorporate type IV collagen. In contrast, laminin interaction with ␤1 integrin induces a fibrillar matrix.342 In Schwann cells, ␣-dystroglycan binds to the transmembrane protein ␤-dystroglycan, which, in turn, binds to dystrophinrelated protein (DRP)-116 (an alternatively spliced form of dystrophin that is specific to Schwann cells), which lacks an actin-binding domain, or to utrophin, which has an actinbinding domain.273,293 In myelinating Schwann cells, dystroglycan can also bind to the recently described Schwann cell dystrophin isoform DRP-2. This in turn binds to the Schwann cell cytoskeletal PDZ-domain protein periaxin.293 (Postsynaptic density protein 95, Drosophila discs large tumor suppressor, zonula occludens-1 [PDZ] domains are commonly found in proteins that participate in signaling complexes.) Postnatally, periaxin is clearly restricted to myelinating cells, and in developing nerves, initial activation of the periaxin gene is seen in a subpopulation of early

Molecular Signaling in Schwann Cell Development

immature Schwann cells in embryonic nerves, where it appears to be an early indicator of myelin differentiation.247,281 Periaxin exists in two alternatively spliced forms, both of which contain an N-terminal PDZ domain. The large form of periaxin is associated with the plasma membrane in myelinating Schwann cells, where it forms clustered patched complexes with DRP-2, which are seen in close association with portions of the outermost myelin lamella. These clustered complexes are not formed in periaxin-null mice, although DRP-2 protein levels are normal.73,106,107,281,293 Mutations of periaxin result in severe hypomyelinating sensory neuropathy in mice and Charcot-Marie-Tooth neuropathy type 4F in humans.107,293,325,359 Although periaxin-null mice myelinate normally, it is clear that these complexes are essential to the ongoing stability of mature myelin.107,293 Interaction of Schwann cell–derived laminin 2 with a pathogenic bacterium is seen. In leprosy, dystroglycanlaminin interactions are perturbed. This disease, caused by Mycobacterium leprae, is characterized by infiltration and infection of the Schwann cells of sensory nerves of the skin and elsewhere. The infective agent in leprosy, the phenolic glycolipid-1 of the M. leprae cell wall, interacts specifically with the G domain of the laminin ␣2 chain of laminin 2. This interferes with the normal laminin interaction with ␣-dystroglycan.29,231,260,261,293 A recent study, using Schwann cell–DRG neuron co-cultures, demonstrated that M. leprae induces different responses in myelinating and nonmyelinating Schwann cells.262 It has been suggested that different signaling cascades might be initiated in the two Schwann cell types because myelinating Schwann cells express the DRP-2 dystrophin isoform, in addition to DRP-116 and utrophin expressed by both myelinating and nonmyelinating cells.29 In nonmyelinating cells the bacteria are taken up and sequestered with the Schwann cells, while in myelinating cells binding of the phenolic glycolipid induces rapid demyelination with relatively little uptake of bacteria.262 Whether this is relevant to the pathology of leprosy remains to be determined.

ACKNOWLEDGMENTS The work from our laboratory quoted in this review was supported by the Wellcome Trust. We thank Mrs. D. Bartram for extensive editorial assistance.

REFERENCES 1. Adlkofer, K., and Lai, C.: Role of neuregulins in glial cell development. Glia 29:104, 2000. 2. Aguayo, A. J., and Bray, G. M.: Developmental disorders of myelination in mouse mutants. In Sears, T. A. (ed.): Neuronal-Glial Cell Interrelationships. Berlin, SpringerVerlag, p. 57, 1982.

365

3. Amedee, T., and Despeyroux, S.: ATP activates cationic and anionic conductances in Schwann cells cultured from dorsal root ganglia of the mouse. Proc. R. Soc. Lond. B Biol. Sci. 259:277, 1995. 4. Anderson, D. J.: Genes, lineages and the neural crest: a speculative review. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355:953, 2000. 5. Anton, E. S., Hadjiargyrou, M., Patterson, P. H., and Matthew, W. D.: CD9 plays a role in Schwann cell migration in vitro. J. Neurosci. 15:584, 1995. 6. Archelos, J. J., Roggenbuck, K., Schneider-Schaulies, J., et al.: Production and characterization of monoclonal antibodies to the extracellular domain of P0. J. Neurosci. Res. 35:46, 1993. 7. Arroyo, E. J., Bermingham, J. R. Jr., Rosenfeld, M. G., and Scherer, S. S.: Promyelinating Schwann cells express Tst-1/SCIP/Oct-6. J. Neurosci. 18:7891, 1998. 8. Atanasoski, S., Shumas, S., Dickson, C., et al.: Differential cyclin D1 requirements of proliferating Schwann cells during development and after injury. Mol. Cell. Neurosci. 18:581, 2001. 9. Baechner, D., Liehr, T., Hameister, H., et al.: Widespread expression of the peripheral myelin protein-22 gene (PMP22) in neural and non-neural tissues during murine development. J. Neurosci. Res. 15:733, 1995. 10. Banerjee, S. A., and Patterson, P. H.: Schwann cell CD9 expression is regulated by axons. Mol. Cell. Neurosci. 6:462, 1995. 11. Bao, J., Wolpowitz, D., Role, L. W., and Talmage, D. A.: Back signaling by the Nrg-1 intracellular domain. J. Cell Biol. 161:1133, 2003. 12. Barakat-Walter, I., Duc, C., and Puymirat, J.: Changes in the nuclear 3,5,3'-triiodothyronine receptor expression in the rat dorsal root ganglia and sciatic nerve during development: comparison with regeneration. Eur. J. Neurosci. 5:319, 1993. 13. Bellone, E., Di Maria, E., Soriani, S., et al.: A novel mutation (D305V) in the early growth response 2 gene is associated with severe Charcot-Marie-Tooth type 1 disease. Hum. Mutat. 14:353, 1999. 14. Benzel, I., Barde, Y. A., and Casademunt, E.: Strain-specific complementation between NRIF1 and NRIF2, two zinc finger proteins sharing structural and biochemical properties. Gene 281:19, 2001. 15. Bermingham, J. R., Scherer, S. S., O’Connell, S., et al.: Tst-1/Oct-6/SCIP regulates a unique step in peripheral myelination and is required for normal respiration. Genes Dev. 10:1751, 1996. 16. Bernier, G., De Repentigny, Y., Mathieu, M., et al.: Dystonin is an essential component of the Schwann cell cytoskeleton at the time of myelination. Development 125:2135, 1998. 17. Berti-Mattera, L. N., Wilkins, P. L., Madhun, Z., and Suchovsky, D.: P2-purigenic receptors regulate phospholipase C and adenylate cyclase activities in immortalized Schwann cells. Biochem. J. 314:555, 1996. 18. Bharucha, V. A., Peden, K. W., Subach, B. R., et al.: Characterization of the cis-acting elements of the mouse myelin P2 promoter. J. Neurosci. Res. 36:508, 1993. 19. Bhattacharyya, A., Frank, E., Ratner, N., and Brackenbury, R.: P0 is an early marker of the Schwann cell lineage in chickens. Neuron 7:831, 1991.

366

Neurobiology of the Peripheral Nervous System

20. Bitgood, M. J., and McMahon, A. P.: Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev. Biol. 172:126, 1995. 21. Bixby, S., Kruger, G. M., Mosher, J. T., et al.: Cell-intrinsic differences between stem cells from different regions of the peripheral nervous system regulate the generation of neural diversity. Neuron 35:643, 2002. 22. Blanchard, A. D., Sinanan, A., Parmantier, E., et al.: Oct-6 (SCIP/Tst-1) is expressed in Schwann cell precursors, embryonic Schwann cells, and postnatal myelinating Schwann cells: comparison with Oct-1, Krox-20 and Pax-3. J. Neurosci. Res. 46:630, 1996. 23. Boerkoel, C. F., Takashima, H., Bacino, C. A., et al.: EGR2 mutation R359W causes a spectrum of Dejerine-Sottas neuropathy. Neurogenetics 3:153, 2001. 24. Bondurand, N., Girard, M., Pingault, V., et al.: Human connexin 32, a gap junction protein altered in the X-linked form of Charcot-Marie-Tooth disease, is directly regulated by the transcription factor SOX10. Hum. Mol. Genet. 10:2783, 2001. 25. Bönnemann, C. G., McNally, E. M., and Kunkel, L. M.: Beyond dystrophin: current progress in the muscular dystrophies. Curr. Opin. Pediatr. 8:569, 1996. 26. Brennan, A., Dean, C. H., Zhang, A. L., et al.: Endothelins control the timing of Schwann cell generation in vitro and in vivo. Dev. Biol. 227:545, 2000. 27. Britsch, S., Goerich, D. E., Riethmacher, D., et al.: The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev. 15:66, 2001. 28. Britsch, S., Li, L., Kirchhoff, S., et al.: The ErbB2 and ErbB3 receptors and their ligand, neuregulin-1, are essential for development of the sympathetic nervous system. Genes Dev. 12:1825, 1998. 29. Brophy, P. J.: Microbiology: subversion of Schwann cells and the leper’s bell. Science 296:862, 2002. 30. Brown, M. J., and Asbury, A. K.: Schwann cell proliferation in the postnatal mouse: timing and topography. Exp. Neurol. 74:170, 1981. 31. Bunge, R. P.: Expanding roles for the Schwann cell: ensheathment, myelination, trophism and regeneration. Curr. Opin. Neurobiol. 3:805, 1993. 32. Bunge, R. P., Bunge, M. B., and Eldridge, C. F.: Linkage between axonal ensheathment and basal lamina production by Schwann cells. Annu. Rev. Neurosci. 9:305, 1986. 33. Buonanno, A., and Fischbach, G. D.: Neuregulin and ErbB receptor signaling pathways in the nervous system. Curr. Opin. Neurobiol. 11:287, 2001. 34. Burden, S., and Yarden, Y.: Neuregulins and their receptors: a versatile signaling module in organogenesis and oncogenesis. Neuron 18:847, 1997. 35. Campana, W. M., Darin, S. J., and O’Brien, J. S.: Phosphatidylinositol 3-kinase and Akt protein kinase mediate IGF-I- and prosaptide-induced survival in Schwann cells. J. Neurosci. Res. 57:332, 1999. 36. Carraway, K. L. 3rd, Weber, J. L., Unger, M. J., et al.: Neuregulin-2, a new ligand of ErbB3/ErbB4-receptor tyrosine kinases. Nature 387:512, 1997. 37. Carroll, S. L., Miller, M. L., Frohnert, P. W., et al.: Expression of neuregulins and their putative receptors,

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

ErbB2 and ErbB3, is induced during Wallerian degeneration. J. Neurosci. 17:1642, 1997. Carter, B. D., Kaltschmidt, C., Kaltschmidt, B., et al.: Selective activation of NF-kappa B by nerve growth factor through the neurotrophin receptor p75. Science 272:542, 1996. Casademunt, E., Carter, B. D., Benzel, I., et al.: The zinc finger protein NRIF interacts with the neurotrophin receptor p75(NTR) and participates in programmed cell death. EMBO J. 18:6050, 1999. Cass, L. A., and Meinkoth, J. L.: Differential effects of cyclic adenosine 3⬘,5⬘-monophosphate on p70 ribosomal S6 kinase. Endocrinology 139:1991, 1998. Cass, L. A., and Meinkoth, J. L.: Ras signaling through PI3K confers hormone-independent proliferation that is compatible with differentiation. Oncogene 19:924, 2000. Cass, L. A., Summers, S. A., Prendergast, G. V., et al.: Protein kinase A-dependent and -independent signaling pathways contribute to cyclic AMP-stimulated proliferation. Mol. Cell. Biol. 19:5882, 1999. Chan, J. R., Cosgaya, J. M., Wu, Y. J., and Shooter, E. M.: Neurotrophins are key mediators of the myelination program in the peripheral nervous system. Proc. Natl. Acad. Sci. U.S.A. 98:14661, 2001. Chan, J. R., Phillips, L. J. 2nd, and Glaser, M.: Glucocorticoids and progestins signal the initiation and enhance the rate of myelin formation. Proc. Natl. Acad. Sci. U.S.A. 95:10459, 1998. Chan, J. R., Rodriguez-Waitkus, P. M., Ng, B. K., et al.: Progesterone synthesized by Schwann cells during myelin formation regulates neuronal gene expression. Mol. Biol. Cell 11:2283, 2000. Chen, L. M., Bailey, D., and Fernandez-Vallé, C.: Association of beta 1 integrin with focal adhesion kinase and paxillin in differentiating Schwann cells. J. Neurosci. 20:3776, 2000. Cheng, H. L., and Feldman, E. L.: Insulin-like growth factor-I (IGF-I) and IGF binding protein-5 in Schwann cell differentiation. J. Cell. Physiol. 171:161, 1997. Cheng, H. L., Russell, J. W., and Feldman, E. L.: IGF-I promotes peripheral nervous system myelination. Ann. N. Y. Acad. Sci. 883:124, 1999. Cheng, H. L., Steinway, M., Delaney, C. L., et al.: IGF-I promotes Schwann cell motility and survival via activation of Akt. Mol. Cell. Endocrinol. 170:211, 2000. Cheng, H. L., Steinway, M. L., Russell, J. W., and Feldman, E. L.: GTPases and phosphatidylinositol 3-kinase are critical for insulin-like growth factor-I-mediated Schwann cell motility. J. Biol. Chem. 275:27197, 2000. Cheng, L., Esch, F. S., Marchionni, M. A., and Mudge, A. W.: Control of Schwann cell survival and proliferation: autocrine factors and neuregulins. Mol. Cell. Neurosci. 12:141, 1998. Cheng, L., and Mudge, A. W.: Cultured Schwann cells constitutively express the myelin protein P0. Neuron 16:309, 1996. Chernousov, M. A., and Carey, D. J.: Schwann cell extracellular matrix molecules and their receptors. Histol. Histopathol. 15:593, 2000. Chittka, A., and Chao, M. V.: Identification of a zinc finger protein whose subcellular distribution is regulated by serum

Molecular Signaling in Schwann Cell Development

55. 56.

57.

58.

59.

60.

61. 62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

and nerve growth factor. Proc. Natl. Acad. Sci. U.S.A. 96:10705, 1999. Chuah, M. I., and West, A. K.: Cellular and molecular biology of ensheathing cells. Microsc. Res. Tech. 58:216, 2002. Ciment, G.: The melanocyte Schwann cell progenitor: a bipotent intermediate in the neural crest lineage. Comments Dev. Neurobiol. 1:207, 1990. Ciutat, D., Calderó, J., Oppenheim, R. W., and Esquerda, J. E.: Schwann cell apoptosis during normal development and after axonal degeneration induced by neurotoxins in the chick embryo. J. Neurosci. 16:3979, 1996. Colognato, H., and Yurchenco, P. D.: The laminin alpha2 expressed by dystrophic dy(2J) mice is defective in its ability to form polymers. Curr. Biol. 9:1327, 1999. Colomar, A., and Amedee, T.: ATP stimulation of P2X(7) receptors activates three different ionic conductances on cultured mouse Schwann cells. Eur. J. Neurosci. 14:927, 2001. Conlon, I., and Raff, M.: Differences in the way a mammalian cell and yeast cells coordinate cell growth and cell-cycle progression. J. Biol. 2:7, 2003. Conlon, I. J., Dunn, G. A., Mudge, A. W., and Raff, M. C.: Extracellular control of cell size. Nat. Cell Biol. 3:918, 2001. Cowell, L. A., and Weston, J. A.: An analysis of melanogenesis in cultured chick embryo spinal ganglia. Dev. Biol. 22:670, 1970. Curtis, R., Stewart, H. J., Hall, S. M., et al.: GAP-43 is expressed by nonmyelin-forming Schwann cells of the peripheral nervous system. J. Cell Biol. 116:1455, 1992. Dahme, M., Bartsch, U., Martini, R., et al.: Disruption of the mouse L1 gene leads to malformations of the nervous system. Nat. Genet. 17:346, 1997. Davis, R. L., Weintraub, H., and Lassar, A. B.: Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 51:987, 1987. De Felipe, C., and Hunt, S. P.: The differential control of c-Jun expression in regenerating sensory neurons and their associated glial cells. J. Neurosci. 14:2911, 1994. Delaney, C. L., Cheng, H. L., and Feldman, E. L.: Insulin-like growth factor-I prevents caspase-mediated apoptosis in Schwann cells. J. Neurobiol. 41:540, 1999. Desarnaud, F., Do Thi, A. N., Brown, A. M., et al.: Progesterone stimulates the activity of the promoters of peripheral myelin protein-22 and protein zero genes in Schwann cells. J. Neurochem. 71:1765, 1998. Dong, Z., Brennan, A., Liu, N., et al.: NDF is a neuron-glia signal and regulates survival, proliferation, and maturation of rat Schwann cell precursors. Neuron 15:585, 1995. Dong, Z., Sinanan, A., Parkinson, D., et al.: Schwann cell development in embryonic mouse nerves. J. Neurosci. Res. 56:334, 1999. Dowsing, B. J., Morrison, W. A., Nicola, N. A., et al.: Leukemia inhibitory factor is an autocrine survival factor for Schwann cells. J. Neurochem. 73:96, 1999. Durán Alonso, M. B., Zoidl, G., Taveggia, C., et al.: Identification and characterization of ZFP-57, a novel zinc finger transcription factor in the mammalian peripheral nervous system. J. Biol. Chem. 2004 in press. Dytrych, L., Sherman, D. L., Gillespie, C. S., and Brophy, P. J.: Two PDZ domain proteins encoded by the murine periaxin

74.

75.

76.

77.

78.

79.

80.

81.

82. 83.

84.

85.

86.

87.

88.

89.

90. 91.

367

gene are the result of alternative intron retention and are differentially targeted in Schwann cells. J. Biol. Chem. 273:5794, 1998. Eccleston, P. A.: Regulation of Schwann cell proliferation: mechanisms involved in peripheral nerve development. Exp. Cell Res. 199:1, 1992. Eccleston, P. A., Jessen, K. R., and Mirsky, R.: Transforming growth factor-beta and gamma-interferon have dual effects on growth of peripheral glia. J. Neurosci. Res. 24:524, 1989. Einheber, S., Hannocks, M.-J., Metz, C. N., et al.: Transforming growth factor-beta 1 regulates axon-Schwann cell interactions. J. Cell Biol. 129:443, 1995. Einheber, S., Milner, T., Giancotti, F., and Salzer, J.: Axonal regulation of Schwann cell integrin expression suggests a role for alpha6 beta4 in myelination. J. Cell Biol. 123:1223, 1993. Eldridge, C. F., Bunge, M. B., and Bunge, R. P.: Differentiation of axon-related Schwann cells in vitro: II. Control of myelin formation by basal lamina. J. Neurosci. 9:625, 1989. Engvall, E., Earwicker, D., Day, A., et al.: Merosin promotes cell attachment and neurite outgrowth and is a component of the neurite-promoting factor of RN22 schwannoma cells. Exp. Cell Res. 198:115, 1992. Ervasti, J. M., and Campbell, K. P.: A role for the dystrophinglycoprotein complex as a transmembrane linker between laminin and actin. J. Cell Biol. 122:809, 1993. Falls, D. L., Rosen, K. M., Corfas, G., et al.: ARIA, a protein that stimulates acetylcholine receptor synthesis, is a member of the neu ligand family. Cell 72:801, 1993. Fawcett, J. W., and Keynes, R. J.: Peripheral nerve regeneration. Annu. Rev. Neurosci. 13:43, 1990. Feltri, M. L., Graus Porta, D., Previtali, S. C., et al.: Conditional disruption of beta 1 integrin in Schwann cells impedes interactions with axons. J. Cell Biol. 156:199, 2002. Feltri, M. L., Scherer, S. S., Nemni, R., et al.: ␤4 integrin expression in myelinating Schwann cells is polarized, developmentally regulated and axonally dependent. Development 120:1287, 1994. Feng, L., Hatten, M. E., and Heintz, N.: Brain lipid-binding protein (BLBP): a novel signaling system in the developing mammalian CNS. Neuron 12:895, 1994. Fernandez-Vallé, C., Gorman, D., Gomez, A. M., and Bunge, M. B.: Actin plays a role in both changes in cell shape and gene-expression associated with Schwann cell myelination. J. Neurosci. 17:241, 1997. Fernandez-Vallé, C., Gwynn, L., Wood, P. M., et al.: Anti-␤1 integrin antibody inhibits Schwann cell myelination. J. Neurobiol. 25:1207, 1994. Fernandez-Vallé, C., Tang, Y., Ricard, J., et al.: Paxillin binds schwannomin and regulates its density-dependent localization and effect on cell morphology. Nat. Genet. 31:354, 2002. Ferri, C. C., and Bisby, M. A.: Improved survival of injured sciatic nerve Schwann cells in mice lacking the p75 receptor. Neurosci. Lett. 272:191, 1999. Fischbach, G. D., and Rosen, K. M.: ARIA: a neuromuscular junction neuregulin. Annu. Rev. Neurosci. 20:429, 1997. Frade, J. M., and Barde, Y. A.: Genetic evidence for cell death mediated by nerve growth factor and the neurotrophin receptor p75 in the developing mouse retina and spinal cord. Development 126:683, 1999.

368

Neurobiology of the Peripheral Nervous System

92. Franklin, R. J., and Barnett, S. C.: Olfactory ensheathing cells and CNS regeneration: the sweet smell of success? Neuron 28:15, 2000. 93. French, S. W., Hoyer, K. K., Shen, R. R., and Teitell, M. A.: Transdifferentiation and nuclear reprogramming in hematopoietic development and neoplasia. Immunol. Rev. 187:22, 2002. 94. Fromm, L., and Burden, S. J.: Neuregulin-1-stimulated phosphorylation of GABP in skeletal muscle cells. Biochemistry 40:5306, 2001. 95. Furukawa, T., Mukherjee, S., Bao, Z. Z., et al.: Rax, Hes1, and Notch1 promote the formation of Muller glia by postnatal retinal progenitor cells. Neuron 26:383, 2000. 96. Gabella, G.: Ultrastructure of the nerve plexuses of the mammalian intestine: the enteric glial cells. Neuroscience 6:425, 1981. 97. Gaiano, N., Nye, J. S., and Fishell, G.: Radial glial identity is promoted by Notch1 signaling in the murine forebrain. Neuron 26:395, 2000. 98. Garbern, J. Y., Cambi, F., Tang, X. M., et al.: Proteolipid protein is necessary in peripheral as well as central myelin. Neuron 19:205, 1997. 99. Garratt, A. N., Britsch, S., and Birchmeier, C.: Neuregulin, a factor with many functions in the life of a Schwann cell. Bioessays 22:987, 2000. 100. Garratt, A. N., Voiculescu, O., Topilko, P., et al.: A dual role of erbB2 in myelination and in expansion of the Schwann cell precursor pool. J. Cell Biol. 148:1035, 2000. 101. Gassman, M., and Lemke, G.: Neuregulins and neuregulin receptors in neural development. Curr. Opin. Neurobiol. 7:87, 1997. 102. Ge, W., Martinowich, K., Wu, X., et al.: Notch signaling promotes astrogliogenesis via direct CSL-mediated glial gene activation. J. Neurosci. Res. 69:848, 2002. 103. Gershon, M. D.: V. Genes, lineages, and tissue interactions in the development of the enteric nervous system. Am. J. Physiol. 275:G869, 1998. 104. Ghazvini, M., Mandemakers, W., Jaegle, M., et al.: A cell type-specific allele of the POU gene Oct-6 reveals Schwann cell autonomous function in nerve development and regeneration. EMBO J. 21:4612, 2002. 105. Ghislain, J., Desmarquet-Trin-Dinh, C., Jaegle, M., et al.: Characterisation of cis-acting sequences reveals a biphasic, axon-dependent regulation of Krox20 during Schwann cell development. Development 129:155, 2002. 106. Gillespie, C. S., Sherman, D. L., Blair, G. E., and Brophy, P. J.: Periaxin, a novel protein of myelinating Schwann cells with a possible role in axonal ensheathment. Neuron 12:497, 1994. 107. Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., et al.: Peripheral demyelination and neuropathic pain behavior in periaxin-deficient mice. Neuron 26:523, 2000. 108. Grinspan, J. B., Marchionni, M. A., Reeves, M., et al.: Axonal interactions regulate Schwann cell apoptosis in developing peripheral nerve: neuregulin receptors and the role of neuregulins. J. Neurosci. 16:6107, 1996. 109. Guenard, V., Gwynn, L. A., and Wood, P. M.: Transforming growth factor-beta blocks myelination but not ensheathment of axons by Schwann cells in vitro. J. Neurosci. 15:419, 1995.

110. Guenard, V., Rosenbaum, T., Gwynn, L. A., et al.: Effect of transforming growth factor-beta 1 and -beta 2 on Schwann cell proliferation on neurites. Glia 13:309, 1995. 111. Guennoun, R., Benmessahel, Y., Delespierre, B., et al.: Progesterone stimulates Krox-20 gene expression in Schwann cells. Brain Res. Mol. Brain Res. 90:75, 2001. 112. Haack, H., and Hynes, R. O.: Integrin receptors are required for cell survival and proliferation during development of the peripheral glial lineage. Dev. Biol. 233:38, 2001. 113. Hadjiargyrou, M., Kaprielian, Z., Kato, N., and Patterson, P. H.: Association of the tetraspan protein CD9 with integrins on the surface of S-16 Schwann cells. J. Neurochem. 67:2505, 1996. 114. Hagedorn, L., Paratore, C., Brugnoli, G., et al.: The Ets domain transcription factor Erm distinguishes rat satellite glia from Schwann cells and is regulated in satellite cells by neuregulin signaling. Dev. Biol. 219:44, 2000. 115. Hagedorn, L., Suter, U., and Sommer, L.: P0 and PMP22 mark a multipotent neural crest-derived cell type that displays community effects in response to TGF-beta family factors. Development 126:3781, 1999. 116. Hall, D. E., Reichardt, L. F., Crowley, E., et al.: The ␣1␤1 and ␣6␤1 integrin heterodimers mediate cell attachment to distinct sites on laminin. J. Cell Biol. 110:2175, 1990. 117. Haney, C. A., Sahenk, Z., Li, C., et al.: Heterophilic binding of L1 on unmyelinated sensory axons mediates Schwann cell adhesion and is required for axonal survival. J. Cell Biol. 146:1173, 1999. 118. Harari, D., Tzahar, E., Romano, J., et al.: Neuregulin-4: a novel growth factor that acts through the ErbB-4 receptor tyrosine kinase. Oncogene 18:2681, 1999. 119. Harris, B.: Zinc-finger transcription factors in the Schwann cell lineage. PhD thesis, University of London, 2001. 120. He, X., Gerrero, R., Simmons, D. M., et al.: Tst-1, a member of the POU domain gene family, binds the promoter of the gene encoding the cell surface adhesion molecule P0. Mol. Cell. Biol. 11:1739, 1991. 121. Henion, P. D., and Weston, J. A.: Timing and pattern of cell fate restrictions in the neural crest lineage. Development 124:4351, 1997. 122. Hippenmeyer, S., Shneider, N. A., Birchmeier, C., et al.: A role for neuregulin1 signaling in muscle spindle differentiation. Neuron 36:1035, 2002. 123. Hirata, H., Hibasami, H., Yoshida, T., et al.: Nerve growth factor signaling of p75 induces differentiation and ceramide-mediated apoptosis in Schwann cells cultured from degenerating nerves. Glia 36:245, 2001. 124. Hogervorst, F., Admiraal, L. G., Niessen, C., et al.: Biochemical characterization and tissue distribution of the A and B variants of the integrin alpha 6 subunit. J. Cell Biol. 121:179, 1993. 125. Hoke, A., Ho, T., Crawford, T. O., et al.: Glial cell line-derived neurotrophic factor alters axon Schwann cell units and promotes myelination in unmyelinated nerve fibers. J. Neurosci. 23:561, 2003. 126. Honig, M. G., Camilli, S. J., and Xue, Q. S.: Effects of L1 blockade on sensory axon outgrowth and pathfinding in the chick hindlimb. Dev. Biol. 243:137, 2002. 127. Howe, D. G., and McCarthy, K. D.: Retroviral inhibition of cAMP-dependent protein kinase inhibits myelination but

Molecular Signaling in Schwann Cell Development

128.

129.

130.

131.

132.

133.

134.

135.

136.

137. 138.

139.

140.

141.

142.

143.

144.

not Schwann cell mitosis stimulated by interaction with neurons. J. Neurosci. 20:3513, 2000. Huan, Y., Rutkowski, J. L., and Tennekoon, G. I.: The role of CCAAT/enhancer binding protein in Schwann cell differentiation (No. 2105). Presented at the 34th Annual Meeting of the American Society for Cell Biology, San Francisco, 1994. Itoh, K., Ozaki, M., Stevens, B., and Fields, R. D.: Activitydependent regulation of N-cadherin in DRG neurons: differential regulation of N-cadherin, NCAM, and L1 by distinct patterns of action potentials. J. Neurobiol. 33:735, 1997. Itoh, K., Stevens, B., Schachner, M., and Fields, R. D.: Regulated expression of the neural cell adhesion molecule L1 by specific patterns of neural impulses. Science 270:1369, 1995. Jaakkola, S., Savunen, O., Halme, T., et al.: Basement membranes during development of human nerve: Schwann cells and perineurial cells display marked changes in their expression profiles for laminin subunits and beta 1 and beta 4 integrins. J. Neurocytol. 22:215, 1993. Jaegle, M., Ghazvini, M., Mandemakers, W., et al.: The POU proteins Brn-2 and Oct-6 share important functions in Schwann cell development. Genes Dev. 17:1380, 2003. Jaegle, M., Mandemakers, W., Broos, L., et al.: The POU factor Oct-6 and Schwann cell differentiation. Science 273:507, 1996. Jessen, K. R., Brennan, A., Morgan, L., et al.: The Schwann cell precursor and its fate: a study of cell death and differentiation during gliogenesis in rat embryonic nerves. Neuron 12:509, 1994. Jessen, K. R., and Mirsky, R.: Astrocyte-like glia in the peripheral nervous system: an immunohistochemical study of enteric glia. J. Neurosci. 3:2206, 1983. Jessen, K. R., and Mirsky, R.: Schwann cells and their precursors emerge as major regulators of nerve development. Trends Neurosci. 22:402, 1999. Jessen, K. R., and Mirsky, R.: Signals that determine Schwann cell identity. J. Anat. 200:367, 2002. Jessen, K. R., and Mirsky, R.: Schwann cell development. In Lazzarini, R. A. (ed.): Myelin Biology and Disorders, Vol. 1. San Diego, CA, Academic Press, p. 329, 2003. Jessen, K. R., and Mirsky, R.: The Schwann cell lineage. In Kettenmann, H., and Ransom, B. R. (eds.): Neuroglia, 2nd ed. Oxford University Press, 2004. Kaempchen, K., Mielke, K., Utermark, T., et al.: Upregulation of the Rac1/JNK signaling pathway in primary human schwannoma cells. Hum. Mol. Genet. 12:1211, 2003. Kamholz, J., Sessa, M., Scherer, S., et al.: Structure and expression of proteolipid protein in the peripheral nervous system. J. Neurosci. Res. 31:231, 1992. Kaprielian, Z., Cho, K. O., Hadjiargyrou, M., and Patterson, P. H.: CD9, a major platelet cell surface glycoprotein, is a ROCA antigen and is expressed in the nervous system. J. Neurosci. 15:562, 1995. Kelsh, R. N., Dutton, K., Medlin, J., and Eisen, J. S.: Expression of zebrafish fkd6 in neural crest-derived glia. Mech. Dev. 93:161, 2000. Kendall, S., Goldhawk, D., Kubu, C., et al.: Expression analysis of a novel p75(NTR) signaling protein, which regulates cell cycle progression and apoptosis. Mech. Dev. 117:187, 2002.

369

145. Khursigara, G., Bertin, J., Yano, H., et al.: A prosurvival function for the p75 receptor death domain mediated via the caspase recruitment domain receptor-interacting protein 2. J. Neurosci. 21:5854, 2001. 146. Khursigara, G., Orlinick, J. R., and Chao, M. V.: Association of the p75 neurotrophin receptor with TRAF6. J. Biol. Chem. 274:2597, 1999. 147. Kim, H. A., DeClue, J. E., and Ratner, N.: cAMP-dependent protein kinase A is required for Schwann cell growth: interactions between the cAMP and neuregulin/tyrosine kinase pathways. J. Neurosci. Res. 49:236, 1997. 148. Kim, H. A., Pomeroy, S. L., Whoriskey, W., et al.: A developmentally regulated switch directs regenerative growth of Schwann cells through cyclin D1. Neuron 26:405, 2000. 149. Kim, H. A., Ratner, N., Roberts, T. M., and Stiles, C. D.: Schwann cell proliferative responses to cAMP and Nf1 are mediated by cyclin D1. J. Neurosci. 21:1110, 2001. 150. Kioussi, C., Gross, M. K., and Gruss, P.: Pax3: a paired domain gene as a regulator in PNS myelination. Neuron 15:553, 1995. 151. Koenig, H. L., Schumacher, M., Ferzaz, B., et al.: Progesterone synthesis and myelin formation by Schwann cells. Science 268:1500, 1995. 152. Kondo, T., and Raff, M.: The Id4 HLH protein and the timing of oligodendrocyte differentiation. EMBO J. 19:1998, 2000. 153. Kondo, T., and Raff, M.: Oligodendrocyte precursor cells reprogrammed to become multipotential CNS stem cells. Science 289:1754, 2000. 154. Krasnoselsky, A., Massay, M. J., DeFrances, M. C., et al.: Hepatocyte growth factor is a mitogen for Schwann cells and is present in neurofibromas. J. Neurosci. 14:7284, 1994. 155. Kubu, C. J., Orimoto, K., Morrison, S. J., et al.: Developmental changes in Notch1 and numb expression mediated by local cell-cell interactions underlie progressively increasing delta sensitivity in neural crest stem cells. Dev. Biol. 244:199, 2002. 156. Kuhlbrodt, K., Herbarth, B., Sock, E., et al.: Cooperative function of POU proteins and SOX proteins in glial cells. J. Biol. Chem. 273:16050, 1998. 157. Kuhlbrodt, K., Herbarth, B., Sock, E., et al.: Sox10, a novel transcriptional modulator in glial cells. J. Neurosci. 18:237, 1998. 158. Kuhn, R., Monuki, E. S., and Lemke, G.: The gene encoding the transcription factor SCIP has features of an expressed retroposon. Mol. Cell. Biol. 11:4642, 1991. 159. Kurtz, A., Zimmer, A., Schnutgen, F., et al.: The expression pattern of a novel gene encoding brain-fatty acid binding protein correlates with neuronal and glial cell development. Development 120:2637, 1994. 160. Kury, P., Greiner-Petter, R., Cornely, C., et al.: Mammalian achaete scute homolog 2 is expressed in the adult sciatic nerve and regulates the expression of Krox24, Mob-1, CXCR4, and p57kip2 in Schwann cells. J. Neurosci. 22:7586, 2002. 161. Labatut-Cazabat, I., Vekris, A., and Petry, K. G.: A protein with the characters of a zinc-finger is implicated in the differentiation of Schwann cells. Neuroreport 10:3037, 1999. 162. Lallemand, D., Curto, M., Saotome, I., et al.: NF2 deficiency promotes tumorigenesis and metastasis by destabilizing adherens junctions. Genes Dev. 17:1090, 2003.

370

Neurobiology of the Peripheral Nervous System

163. Landmesser, L., Dahm, L., Schultz, K., and Rutishauser, U.: Distinct roles for adhesion molecules during innervation of embryonic chick muscle. Dev. Biol. 130:645, 1988. 164. Landon, D. N, and Wiseman, O. J.: A pacinian corpuscle in the human bladder lamina propria. J. Neurocytol. 30:457, 2001. 165. Le, X. F., Varela, C. R., and Bast, Jr. R. C.: Heregulin-induced apoptosis. Apoptosis 7:483, 2002. 166. Le Douarin, N. M., and Kalcheim, C.: The Neural Crest. Cambridge, UK, Cambridge University Press, 1999. 167. Lee, J. E., Hollenberg, S. M., Snider, L., et al.: Conversion of Xenopus ectoderm into neurons by NeuroD, a basic helix-loop-helix protein. Science 268:836, 1995. 168. Lee, M.-J., Brennan, A., Blanchard, A., et al.: P0 is constitutively expressed in the rat neural crest and embryonic nerves and is negatively and positively regulated by axons to generate non-myelin-forming and myelin-forming Schwann cells, respectively. Mol. Cell. Neurosci. 8:336, 1997. 169. Lee, M.-J., Calle, E., Brennan, A., et al.: In early development of the rat mRNA for the major myelin protein P0 is expressed in nonsensory areas of the embryonic inner ear, notochord, enteric nervous system, and olfactory ensheathing cells. Dev. Dyn. 222:40, 2001. 170. Lee, M. M., Sato-Bigbee, C., and De Vries, G. H.: Schwann cells stimulated by axolemma-enriched fractions express cyclic AMP responsive element binding protein. J. Neurosci. Res. 46:204, 1996. 171. Leimeroth, R., Lobsiger, C., Lüssi, A., et al.: Membranebound neuregulin1 type III actively promotes Schwann cell differentiation of multipotent progenitor cells. Dev. Biol. 246:245, 2002. 172. Lemke, G.: Neuregulins in development. Mol. Cell. Neurosci. 7:247, 1996. 173. Lemke, G.: Glial control of neuronal development. Annu. Rev. Neurosci. 24:87, 2001. 174. Lemke, G., and Chao, M.: Axons regulate Schwann cell expression of the major myelin and NGF receptor genes. Development 102:499, 1988. 175. Lemke, G., Kuhn, R., Monuki, E. S., and Weinmaster, G.: Expression and activity of the transcription factor SCIP during glial differentiation and myelination. Ann. N. Y. Acad. Sci. 633:189, 1991. 176. Lentz, S. I., Miner, J. H., Sanes, J. R, and Snider, W. D.: Distribution of the ten known laminin chains in the pathways and targets of developing sensory axons. J. Comp. Neurol. 378:547, 1997. 177. Levi, A. D., Bunge, R. P., Lofgren, J. A., et al.: The influence of heregulins on human Schwann cell proliferation. J. Neurosci. 15:1329, 1995. 178. Li, H., Wigley, C., and Hall, S. M.: Chronically denervated rat Schwann cells respond to GGF in vitro. Glia 24:290, 1998. 179. Li, Y., Tennekoon, G. I., Birnbaum, M., et al.: Neuregulin signaling through a PI3K/Akt/Bad pathway in Schwann cell survival. Mol. Cell. Neurosci. 17:761, 2001. 180. Lin, D. M., and Goodman, C. S.: Ectopic and increased expression of Fasciclin II alters motoneuron growth cone guidance. Neuron 13:507, 1994. 181. Livesey, F. J., O’Brien, J. A., Li, M., et al.: A Schwann cell mitogen accompanying regeneration of motor neurons. Nature 390:614, 1997.

182. Lloyd, A. C., Obermuller, F., Staddon, S., et al.: Cooperating oncogenes converge to regulate cyclin/cdk complexes. Genes Dev. 11:663, 1997. 183. Lo, L., Tiveron, M. C., and Anderson, D. J.: MASH1 activates expression of the paired homeodomain transcription factor Phox2a, and couples pan-neuronal and subtype-specific components of autonomic neuronal identity. Development 125:609, 1998. 184. Lobsiger, C. S., Taylor, V., and Suter, U.: The early life of a Schwann cell. Biol. Chem. 383:245, 2002. 185. Loeb, J. A., Khurana, T. S., Robbins, J. T., et al.: Expression patterns of transmembrane and released forms of neuregulin during spinal cord and neuromuscular synapse development. Development 126:781, 1999. 186. Lubischer, J. L., and Thompson, W. J.: Neonatal partial denervation results in nodal but not terminal sprouting and a decrease in efficacy of remaining neuromuscular junctions in rat soleus muscle. J. Neurosci. 19:8931, 1999. 187. Lyons, S. A., Morell, P., and McCarthy, K. D.: Schwann cells exhibit P2Y purinergic receptors that regulate intracellular calcium and are up-regulated by cyclic AMP analogues. J. Neurochem. 63:552, 1994. 188. Lyons, S. A., Morell, P., and McCarthy, K. D.: Schwann cell ATP-mediated calcium increases in vitro and in situ are dependent on contact with neurons. Glia 13:27, 1995. 189. Magnaghi, V., Cavarretta, I., Galbiati, M., et al.: Neuroactive steroids and peripheral myelin proteins. Brain Res. Brain Res. Rev. 37:360, 2001. 190. Mandemakers, W., Zwart, R., Jaegle, M., et al.: A distal Schwann cell-specific enhancer mediates axonal regulation of the Oct-6 transcription factor during peripheral nerve development and regeneration. EMBO J. 19:2992, 2000. 191. Marchionni, M. A., Goodearl, A. D. J., Chen, M. S., et al.: Glial growth factors are alternatively spliced erbB2 ligands expressed in the nervous system. Nature 362:312, 1993. 192. Martini, R., Bollensen, E., and Schachner, M.: Immunocytological localization of the major peripheral nervous system glycoprotein P0 and the L2/HNK-1 and L3 carbohydrate structures in developing and adult mouse sciatic nerve. Dev. Biol. 129:330, 1988. 193. Masaki, T., Matsumura, K., Hirata, A., et al.: Expression of dystroglycan and the laminin-alpha 2 chain in the rat peripheral nerve during development. Exp. Neurol. 174:109, 2002. 194. Matheny, C., DiStephano, P. S., and Milbrandt, J.: Differential expression of NGF receptor and early response genes in neural crest-derived cells. Mol. Brain Res. 13:75, 1992. 195. Mathon, N. F., Malcolm, D. S., Harrisingh, M. C., et al.: Lack of replicative senescence in normal rodent glia. Science 291:872, 2001. 196. Matsumura, K., Chiba, A., Yamada, H., et al.: A role of dystroglycan in Schwannoma cell adhesion to laminin. J. Biol. Chem. 272:13904, 1997b. 197. Matsumura, K., Yamada, H., Saito, F., et al.: Peripheral nerve involvement in merosin-deficient congenital muscular dystrophy and dy mouse. Neuromuscul. Disord. 7:7, 1997a. 198. Maurel, P., and Salzer, J. L.: Axonal regulation of Schwann cell proliferation and survival and the initial events of myelination require PI 3-kinase activity. J. Neurosci. 20:4635, 2000.

Molecular Signaling in Schwann Cell Development 199. Mayer, C., Wachtler, J., Kamleiter, M., and Grafe, P.: Intracellular calcium transients mediated by P2 receptors in the paranodal Schwann cell region of myelinated rat spinal root axons. Neurosci. Lett. 224:49, 1997. 200. Meier, C., Parmantier, E., Brennan, A., et al.: Developing Schwann cells acquire the ability to survive without axons by establishing an autocrine circuit involving IGF, NT-3 and PDGF-BB. J. Neurosci. 19:3847, 1999. 201. Meintanis, S., Thomaidou, D., Jessen, K. R., et al.: The neuron-glia signal beta-neuregulin promotes Schwann cell motility via the MAPK pathway. Glia 34:39, 2001. 202. Memberg, S. P., and Hall, A. K.: Dividing neuron precursors express neuron-specific tubulin. J. Neurobiol. 27:26, 1995. 203. Mews, M., and Meyer, M.: Modulation of Schwann cell phenotype by TGF-beta 1: inhibition of P0 mRNA expression and downregulation of the low affinity NGF receptor. Glia 8:208, 1993. 204. Meyer, D., and Birchmeier, C.: Multiple essential functions of neuregulin in development. Nature 378:386, 1995. 205. Meyer, D., Yamaii, T., Garratt, A., et al.: Isoform-specific expression and function of neuregulin. Development 124:3575, 1997. 206. Milner, R., Wilby, M., Nishimura, S., et al.: Division of labor of Schwann cell integrins during migration on peripheral nerve extracellular matrix ligands. Dev. Biol. 185:215, 1997. 207. Mirsky, R., Dubois, C., Morgan, L., and Jessen, K. R.: 04 and A007-sulfatide antibodies bind to embryonic Schwann cells prior to the appearance of galactocerebroside; regulation of the antigen by axon-Schwann cell signals and cyclic AMP. Development 109:105, 1990. 208. Mirsky, R., and Jessen, K. R.: Schwann cell development, differentiation and myelination. Curr. Opin. Neurobiol. 6:89, 1996. 209. Mirsky, R., and Jessen, K. R.: The neurobiology of Schwann cells. Brain Pathol. 9:293, 1999. 210. Mirsky, R., and Jessen, K. R.: Embryonic and early postnatal development of Schwann cells. In Jessen, K. R., and Richardson, W. D. (eds.): Glial Cell Development: Basic Principles and Clinical Relevance, 2nd ed. Oxford, UK, Oxford University Press, p. 1, 2001. 211. Monuki, E. S., Kuhn, R., and Lemke, G.: Repression of the myelin P0 gene by the POU transcription factor SCIP. Mech. Dev. 42:15, 1993. 212. Monuki, E. S., Kuhn, R., Weinmaster, G., et al.: Expression and activity of the POU transcription factor SCIP. Science 249:1300, 1990. 213. Monuki, E. S., Weinmaster, G., Kuhn, R., and Lemke, G.: SCIP: a glial POU domain gene regulated by cyclic AMP. Neuron 3:783, 1989. 214. Morgan, L., Jessen, K. R., and Mirsky, R.: The effects of cAMP on differentiation of cultured Schwann cells: progression from an early phenotype (04⫹ ) to a myelin phenotype (P0⫹ , GFAP⫺, N-CAM⫺, NGF-receptor⫺) depends on growth inhibition. J. Cell Biol. 112:457, 1991. 215. Morgan, L., Jessen, K. R., and Mirsky, R.: Negative regulation of the P0 gene in Schwann cells: suppression of P0 mRNA and protein induction in cultured Schwann cells by FGF2 and TGF␤1, TGF␤2 and TGF␤3. Development 120:1399, 1994. 216. Morris, J. K., Lin, W., Hauser, C., et al.: Rescue of the cardiac defect in ErbB2 mutant mice reveals essential roles of ErbB2

217.

218.

219.

220.

221.

222.

223.

224.

225.

226.

227.

228.

229.

230.

231.

232.

371

in peripheral nervous system development. Neuron 23:273, 1999. Morrison, H., Sherman, L. S., Legg, J., et al.: The NF2 tumor suppressor gene product, merlin, mediates contact inhibition of growth through interactions with CD44. Genes Dev. 15:968, 2001. Morrison, S. J., Perez, S. E., Qiao, Z., et al.: Transient Notch activation initiates an irreversible switch from neurogenesis to gliogenesis by neural crest stem cells. Cell 101:499, 2000. Morrison, S. J., White, P. M., Zock, C., and Anderson, D. J.: Prospective identification, isolation by flow cytometry, and in vivo self-renewal of multipotent mammalian neural crest stem cells. Cell 96:737, 1999. Morrissey, T. K., Levi, A. D., Nuijens, A., et al.: Axon-induced mitogenesis of human Schwann cells involves heregulin and p185erbB2. Proc. Natl. Acad. Sci. U.S.A. 92:1431, 1995. Mukai, J., Hachiya, T., Shoji-Hoshino, S., et al.: NADE, a p75NTR-associated cell death executor, is involved in signal transduction mediated by the common neurotrophin receptor p75NTR. J. Biol. Chem. 275:17566, 2000. Mukai, J., Shoji, S., Kimura, M. T., et al.: Structure-function analysis of NADE: identification of regions that mediate nerve growth factor-induced apoptosis. J. Biol. Chem. 277:13973, 2002. Murphy, P., Topilko, P., Schneider-Manoury, S., et al.: The regulation of Krox-20 expression reveals important steps in the control of peripheral glial cell development. Development 122:2847, 1996. Nagano, S., Takeda, M., Ma, L., and Soliven, B.: Cytokineinduced cell death in immortalized Schwann cells: roles of nitric oxide and cyclic AMP. J. Neurochem. 77:1486, 2001. Nagarajan, R., Le, N., Mahoney, H., et al.: Deciphering peripheral nerve myelination by using Schwann cell expression profiling. Proc. Natl. Acad. Sci. U.S.A. 99:8998, 2002. Nagarajan, R., Svaren, J., Le, N., et al.: EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron 30:355, 2001. Nakagawa, M., Miyagoe-Suzuki, Y., Ikezoe, K., et al.: Schwann cell myelination occurred without basal lamina formation in laminin alpha2 chain-null mutant (dy3K/dy3K) mice. Glia 35:101, 2001. Nataf, V., and Le Douarin, N. M.: Induction of melanogenesis by tetradecanoylphorbol-13 acetate and endothelin 3 in embryonic avian peripheral nerve cultures. Pigment Cell Res. 13:172, 2000. Nave, K.-A.: Myelin-specific genes and their mutations in the mouse. In Jessen, K. R., and Richardson, W. D. (eds.): Glial Cell Development: Basic Principles and Clinical Relevance, 2nd ed. Oxford, UK, Oxford University Press, p. 177, 2001. Neve, R. M., Holbro, T., and Hynes, N. E.: Distinct roles for phosphoinositide 3-kinase, mitogen-activated protein kinase and p38 MAPK in mediating cell cycle progression of breast cancer cells. Oncogene 21:4567, 2002. Ng, V., Zanazzi, G., Timpl, R., et al.: Role of the cell wall phenolic glycolipid-1 in the peripheral nerve predilection of Mycobacterium leprae. Cell 103:511, 2000. Nichols, D. H., and Weston, J. A.: Melanogenesis in cultures of peripheral nervous tissue. I. The origin and prospective fate of cells giving rise to melanocytes. Dev. Biol. 60:217, 1977.

372

Neurobiology of the Peripheral Nervous System

233. Nickols, J. C., Valentine, W., Kanwal, S., and Carter, B. D.: Activation of the transcription factor NF-kappaB in Schwann cells is required for peripheral myelin formation. Nat. Neurosci. 6:161, 2003. 234. Niemann, C., Brinkmann, V., and Birchmeier, W.: Hepatocyte growth factor and neuregulin in mammary gland cell morphogenesis. Adv. Exp. Med. Biol. 480:9, 2000. 235. Niessen, C. M., Cremona, O., Daams, H., et al.: Expression of the integrin ␣6␤4 in peripheral nerves: localization in Schwann and perineural cells and different variants of the ␤4 subunit. J. Cell Sci. 107:543, 1994. 236. Nikam, S. S., Tennekoon, G. I., Christy, B. A., et al.: The zinc finger transcription factor Zif268/Egr-1 is essential for Schwann cell expression of the p75 NGF receptor. Mol. Cell. Neurosci. 6:337, 1995. 237. Notterpek, L., Snipes, G. J., and Shooter, E. M.: Temporal expression pattern of peripheral myelin protein 22 during in vivo and in vitro myelination. Glia 25:358, 1999. 238. Okazaki, S., Tanase, S., Choudhury, B. K., et al.: A novel nuclear protein with zinc fingers down-regulated during early mammalian cell differentiation. J. Biol. Chem. 269:6900, 1994. 239. Orr-Urtreger, A., Trakhtenbrot, L., Ben-Levy, R., et al.: Neural expression and chromosomal mapping of Neu differentiation factor to 8p12-p21. Proc. Natl. Acad. Sci. U.S.A. 90:1867, 1993. 240. Pannese, E.: The satellite cells of the sensory ganglia. Adv. Anat. Embryol. Cell Biol. 65:1, 1981. 241. Pannese, E.: Neurocytology: Fine Structure of Neurons, Nerve Processes, and Neuroglial Cells. Stuttgart, Thieme, 1994. 242. Paratore, C., Brugnoli, G., Lee, H. Y., et al.: The role of the Ets domain transcription factor Erm in modulating differentiation of neural crest stem cells. Dev. Biol. 250:168, 2002. 243. Paratore, C., Goerich, D. E., Suter, U., et al.: Survival and glial fate acquisition of neural crest cells are regulated by an interplay between the transcription factor Sox10 and extrinsic combinatorial signaling. Development 128:3949, 2001. 244. Pareyson, D., Taroni, F., Botti, S., et al.: Cranial nerve involvement in CMT disease type 1 due to early growth response 2 gene mutation. Neurology 54:1696, 2000. 245. Parkinson, D. B., Bhaskaran, A., Droggiti, A., et al.: Krox-20 inhibits Jun-NH2-terminal kinase/c-jun to control Schwann cell proliferation and death. J. Cell Biol. 164:385, 2004. 246. Parkinson, D. B., Dickinson, S., Bhaskaran, A., et al.: Krox 20 activates a set of complex changes in Schwann cells that characterize myelination. Glia Suppl. 1:S69, 2002. 247. Parkinson, D. B., Dickinson, S., Bhaskaran, A., et al.: Regulation of the myelin gene periaxin provides evidence for Krox-20 independent myelin-related signalling in Schwann cells. Mol. Cell. Neurosci. 23:13, 2003. 248. Parkinson, D. B., Dong, Z., Bunting, H., et al.: Transforming growth factor ␤ (TGF␤) mediates Schwann cell death in vitro and in vivo: examination of c-Jun activation, interactions with survival signals, and the relationship of TGF␤ mediated death to Schwann cell differentiation. J. Neurosci. 21:8572, 2001. 249. Parkinson, D. B., Langner, K., Namini, S. S., et al.: ␤-Neuregulin and autocrine mediated survival of Schwann

250.

251.

252.

253.

254.

255.

256.

257.

258.

259. 260.

261.

262.

263.

264.

265.

266.

267.

cells requires activity of Ets family transcription factors. Mol. Cell. Neurosci. 20:154, 2002. Parmantier, E., Braun, C., Thomas, J. L., et al.: PMP-22 expression in the central nervous system of the embryonic mouse defines potential transverse segments and longitudinal columns. J. Comp. Neurol. 378:159, 1997. Parmantier, E., Cabon, F., Braun, C., et al.: Peripheral myelin protein-22 is expressed in rat and mouse brain and spinal cord motoneurons. Eur. J. Neurosci. 7:1080, 1995. Parmantier, E., Lynn, B., Lawson, D., et al.: Schwann cell-derived Desert Hedgehog controls the development of peripheral nerve sheaths. Neuron 23:713, 1999. Peirano, R. I., Goerich, D. E., Riethmacher, D., and Wegner, M.: Protein zero gene expression is regulated by the glial transcription factor Sox10. Mol. Cell. Biol. 20:3198, 2000. Plantinga, L. C., Verhaagen, J., Edwards, P. M., et al.: The expression of B-50/GAP-43 in Schwann cells is upregulated in degenerating peripheral nerve stumps following nerve injury. Brain Res. 602:69, 1993. Previtali, S., Nodari, A., Dina, G., et al.: Cell-type specific knockouts in the study of glia-basement membrane interactions. Glia Suppl. 1:S7, 2002. Previtali, S. C., Feltri, M. L., Archelos, J. J., et al.: Role of integrins in the peripheral nervous system. Prog. Neurobiol. 64:35, 2001. Pusch, C., Hustert, E., Pfeifer, D., et al.: The SOX10/Sox10 gene from human and mouse: sequence, expression, and transactivation by the encoded HMG domain transcription factor. Hum. Genet. 103:115, 1998. Rahmatullah, M., Schroering, A., Rothblum, K., et al.: Synergistic regulation of Schwann cell proliferation by heregulin and forskolin. Mol. Cell. Biol. 18:6245, 1998. Raisman, G.: Olfactory ensheathing cells—another miracle cure for spinal cord injury? Nat. Rev. Neurosci. 2:369, 2001. Rambukkana, A., Salzer, J. L., Yurchenco, P. D., and Tuomanen, E. I.: Neural targeting of Mycobacterium leprae mediated by the G domain of the laminin-alpha2 chain. Cell 88:811, 1997. Rambukkana, A., Yamada, H., Zanazzi, G., et al.: Role of alpha-dystroglycan as a Schwann cell receptor for Mycobacterium leprae. Science 282:2076, 1998. Rambukkana, A., Zanazzi, G., Tapinos, N., and Salzer, J. L.: Contact-dependent demyelination by Mycobacterium leprae in the absence of immune cells. Science 296:927, 2002. Ramon-Cueto, A., and Santos-Benito, F. F.: Cell therapy to repair injured spinal cords: olfactory ensheathing glia transplantation. Restor. Neurol. Neurosci. 19:149, 2001. Ridley, A. J., Davis, J. B., Stroobant, P., and Land, H.: Transforming growth factors-beta 1 and beta 2 are mitogens for rat Schwann cells. J. Cell Biol. 109:3419, 1989. Ridley, A. J., Paterson, H. F., Noble, M., and Land, H.: Rasmediated cell cycle arrest is altered by nuclear oncogenes to induce Schwann cell transformation. EMBO J. 7:1635, 1988. Riethmacher, D., Sonnenberg-Riethmacher, E., Brinkmann, V., et al.: Severe neuropathies in mice with targeted mutations in the ErbB3 receptor. Nature 389:725, 1997. Rizvi, T. A., Huang, Y., Sidani, A., et al.: A novel cytokine pathway suppresses glial cell melanogenesis after injury to adult nerve. J. Neurosci. 22:9831, 2002.

Molecular Signaling in Schwann Cell Development 268. Robitaille, R.: Modulation of synaptic efficacy and synaptic depression by glial cells at the frog neuromuscular junction. Neuron 21:847, 1998. 269. Rosenbaum, C., Karyala, S., Marchionni, M. A., et al.: Schwann cells express NDF and SMDF/n-ARIA mRNAs, secrete neuregulin, and show constitutive activation of erbB3 receptors: evidence for a neuregulin autocrine loop. Exp. Neurol. 148:604, 1997. 270. Rouleau, G. A., Merel, P., Lutchman, M., et al.: Alteration in a new gene encoding a putative membrane-organizing protein causes neuro-fibromatosis type 2. Nature 363:515, 1993. 271. Russell, J. W., Cheng, H. L., and Golovoy, D.: Insulin-like growth factor-I promotes myelination of peripheral sensory axons. J. Neuropathol. Exp. Neurol. 59:575, 2000. 272. Sabourin, L. A., and Rudnicki, M. A.: The molecular regulation of myogenesis. Clin. Genet. 57:16, 2000. 273. Sadoulet-Puccio, H. M., and Kunkel, L. M.: Dystrophin and its isoforms. Brain Pathol. 6:25, 1996. 274. Sainio, M., Zhao, F., Heiska, L., et al.: Neurofibromatosis 2 tumor suppressor protein colocalizes with ezrin and CD44 and associates with actin-containing cytoskeleton. J. Cell Sci. 110:2249, 1997. 275. Salehi, A. H., Roux, P. P., Kubu, C. J., et al.: NRAGE, a novel MAGE protein, interacts with the p75 neurotrophin receptor and facilitates nerve growth factor-dependent apoptosis. Neuron 27:279, 2000. 276. Saunders, R. D., and DeVries, G. H.: Schwann cell proliferation is accompanied by enhanced inositol phospholipid metabolism. J. Neurochem. 50:876, 1988. 277. Scherer, S. S.: The biology and pathobiology of Schwann cells. Curr. Opin. Neurol. 10:386, 1997. 278. Scherer, S. S., and Gutmann, D. H.: Expression of the neurofibromatosis 2 tumor suppressor gene product, merlin, in Schwann cells. J. Neurosci. Res. 46:595, 1996. 279. Scherer, S. S., and Salzer, J. L.: Axon-Schwann cell interactions during peripheral nerve degeneration and regeneration. In Jessen, K. R., and Richardson, W. D. (eds.): Glial Cell Development: Basic Principles and Clinical Relevance, 2nd ed. Oxford, UK, Oxford University Press, p. 299, 2001. 280. Scherer, S. S., Wang, D.-Y., Kuhn, R., et al.: Axons regulate Schwann cell expression of the POU transcription factor SCIP. J. Neurosci. 14:1930, 1994. 281. Scherer, S. S., Xu, Y. T., Bannerman, P. G., et al.: Periaxin expression in myelinating Schwann cells: modulation by axon-glial interactions and polarized localization during development. Development 121:4265, 1995. 282. Scherer, S. S., Xu, Y. T., Roling, D., et al.: Expression of growth-associated protein-43 kD in Schwann cells is regulated by axon-Schwann cell interactions and cAMP. J. Neurosci. Res. 38:575, 1994. 283. Schneider, C., Wicht, H., Enderich, J., et al.: Bone morphogenetic proteins are required in vivo for the generation of sympathetic neurons. Neuron 24:861, 1999. 284. Schulze, K. M., Hanemann, C. O., Muller, H. W., and Hanenberg, H.: Transduction of wild-type merlin into human schwannoma cells decreases schwannoma cell growth and induces apoptosis. Hum. Mol. Genet. 11:69, 2002.

373

285. Schumacher, M., Guennoun, R., Mercier, G., et al.: Progesterone synthesis and myelin formation in peripheral nerves. Brain Res. Brain Res. Rev. 37:343, 2001. 286. Scoles, D. R., Huynh, D. P., Morcos, P. A., et al.: Neurofibromatosis 2 tumour suppressor schwannomin interacts with ␤11-spectrin. Nat. Gen. 18:354, 1998. 287. Seilheimer, B., Persohn, E., and Schachner, M.: Antibodies to the L1 adhesion molecule inhibit Schwann cell ensheathment of neurons in vitro. J. Cell Biol. 109:3095, 1989. 288. Seilheimer, B., and Schachner, M.: Studies of adhesion molecules mediating interactions between cells of peripheral nervous system indicate a major role for L1 in mediating sensory neuron growth on Schwann cells in culture. J. Cell Biol. 107:341, 1988. 289. Sensenbrenner, M., Lucas, M., and Deloulme, J. C.: Expression of two neuronal markers, growth-associated protein 43 and neuron-specific enolase, in rat glial cells. J. Mol. Med. 75:653, 1997. 290. Shah, N. M., Groves, A. K., and Anderson, D. J.: Alternative neural crest cell fates are instructively promoted by TGF␤ superfamily members. Cell 85:331, 1996. 291. Shah, N. M., Marchionni, M. A., Isaacs, I., et al.: Glial growth factor restricts mammalian neural crest stem cells to a glial fate. Cell 77:349, 1994. 292. Shaw, R. J., Paez, J. G., Curto, M., et al.: The Nf2 tumor suppressor, merlin, functions in Rac-dependent signaling. Dev. Cell 1:63, 2001. 293. Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J.: Specific disruption of a Schwann cell dystrophin-related protein complex in a demyelinating neuropathy. Neuron 30:677, 2001. 294. Sherman, L., Stocker, K. M., Morrison, R., and Ciment, G.: Basic fibroblast growth factor (bFGF) acts intracellularly to cause the transdifferentiation of avian neural crest-derived Schwann cell precursors into melanocytes. Development 118:1313, 1993. 295. Sherman, L. S., and Gutmann, D. H.: Merlin: hanging tumor suppression on the Rac. Trends Cell Biol. 11:442, 2001. 296. Sherman, L. S., Rizvi, T. A., Karyala, S., and Ratner, N.: CD44 enhances neuregulin signaling by Schwann cells. J. Cell Biol. 150:1071, 2000. 297. Shibuya, Y., Mizoguchi, A., Takeichi, M., et al.: Localization of N-cadherin in the normal and regenerating nerve fibers of the chicken peripheral nervous system. Neuroscience 67:253, 1995. 298. Sim, F., Zhao, C., Li, W., et al.: Expression of the POUdomain transcription factors SCIP/Oct-6 and Brn-2 is associated with Schwann cell but not oligodendrocyte remyelination of the CNS. Mol. Cell. Neurosci. 20:669, 2002. 299. Skoff, A. M., Lisak, R. P., Bealmear, B., and Benjamins, J. A.: TNF-alpha and TGF-beta act synergistically to kill Schwann cells. J. Neurosci. Res. 53:747, 1998. 300. Sobue, G., and Pleasure, D.: Schwann cell galactocerebroside induced by derivatives of adenosine 3',5'-monophosphate. Science 224:72, 1984. 301. Soilu-Hanninen, M., Ekert, P., Bucci, T., et al.: Nerve growth factor signaling through p75 induces apoptosis in Schwann cells via a Bcl-2-independent pathway. J. Neurosci. 19:4828, 1999.

374

Neurobiology of the Peripheral Nervous System

302. Sonnenberg-Riethmacher, E., Miehe, M., Stolt, C. C., et al.: Development and degeneration of dorsal root ganglia in the absence of the HMG-domain transcription factor Sox10. Mech. Dev. 109:253, 2001. 303. Southard-Smith, E. M., Kos, L., and Pavan, W. J.: Sox 10 mutation disrupts neural crest development in Dom Hirschsprung mouse model. Nat. Genet. 18:60, 1998. 304. Spassky, N., Goujet-Zalc, C., Parmantier, E., et al.: Multiple restricted origin of oligodendrocytes. J. Neurosci. 18:8331, 1998. 305. Spencer, P. S., and Schaumburg, H. H.: An ultrastructural study of the inner core of the pacinian corpuscle. J. Neurocytol. 2:217, 1973. 306. Stevens, B., and Fields, R. D.: Response of Schwann cells to action potentials in development. Science 287:2267, 2000. 307. Stevens, B., Tanner, S., and Fields, R. D.: Control of myelination by specific patterns of neural impulses. J. Neurosci. 18:9303, 1998. 308. Stewart, H. J., Curtis, R., Jessen, K. R., and Mirsky, R.: TGF-betas and cAMP regulate GAP-43 expression in Schwann cells and reveal the association of this protein with the trans-Golgi network. Eur. J. Neurosci. 7:1761, 1995. 309. Stewart, H. J. S.: Expression of c-Jun, Jun B, Jun D and cAMP response element binding protein by Schwann cells and their precursors in vivo and in vitro. Eur. J. Neurosci. 7:1366, 1995. 310. Stewart, H. J. S., Bradke, F., Tabernero, A., et al.: Regulation of rat Schwann cell P0 expression and DNA synthesis by insulin-like growth factors in vitro. Eur. J. Neurosci. 8:553, 1996. 311. Stewart, H. J. S., Brennan, A., Rahman, M., et al.: Developmental regulation and overexpression of the transcription factor AP-2, a potential regulator of the timing of Schwann cell generation. Eur. J. Neurosci. 14:363, 2001. 312. Stewart, H. J. S., Mirsky, R., and Jessen, K. R.: The Schwann cell lineage: embryonic and early postnatal development. In Jessen, K. R., and Richardson, W. D. (eds.): Glial Cell Development: Basic Principles and Clinical Relevance. Oxford, UK, Bios Scientific Publishers Ltd, p. 1, 1996. 313. Stewart, H. J. S., Morgan, L., Jessen, K. R., and Mirsky, R.: Changes in DNA synthesis rate in the Schwann cell lineage in vivo are correlated with the precursor-Schwann cell transition and myelination. Eur. J. Neurosci. 5:1136, 1993. 314. Stewart, H. J. S., Zoidl, G., Rossner, M., et al.: Helix-loophelix proteins in Schwann cells: a study of regulation and subcellular localization of Ids, REB and E12/47 during embryonic and postnatal development. J. Neurosci. Res. 50:684, 1997. 315. Stiles, C. D., Pledger, W. J., Tucker, R. W., et al.: Regulation of the Balb/c-3T3 cell cycle-effects of growth factors. J. Supramol. Struct. 13:489, 1980. 316. Stocker, K. M., Sherman, L., Rees, S., and Ciment, G.: Basic FGF and TGF-␤1 influence commitment to melanogenesis in neural crest-derived cells of avian embryos. Development 111:635, 1991. 317. Stolt, C. C., Rehberg, S., Ader, M., et al.: Terminal differentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev. 16:165, 2002.

318. Sudhalter, J., Whitehouse, L., Rusche, J. R., et al.: Schwann cell heparan sulfate proteoglycans play a critical role in glial growth factor/neuregulin signaling. Glia 17:28, 1996. 319. Sulaiman, O. A., and Gordon, T.: Transforming growth factorbeta and forskolin attenuate the adverse effects of long-term Schwann cell denervation on peripheral nerve regeneration in vivo. Glia 37:206, 2002. 320. Sun, C. X., Haipek, C., Scoles, D. R., et al.: Functional analysis of the relationship between the neurofibromatosis 2 tumor suppressor and its binding partner, hepatocyte growth factor-regulated tyrosine kinase substrate. Hum. Mol. Genet. 11:3167, 2002. 321. Syroid, D. E., Maycox, P. J., Soilu-Hanninen, M., et al.: Induction of postnatal Schwann cell death by the low-affinity neurotrophin receptor in vitro and after axotomy. J. Neurosci. 20:5741, 2000. 322. Syroid, D. E., Maycox, P. R., Burrola, P. G., et al.: Cell death in the Schwann cell lineage and its regulation by neuregulin. Proc. Natl. Acad. Sci. U.S.A. 93:9229, 1996. 323. Syroid, D. E., Zorick, T. S., Arbet-Engels, C., et al.: A role for insulin-like growth factor-I in the regulation of Schwann cell survival. J. Neurosci. 19:2059, 1999. 324. Tabernero, A., Stewart, H. J. S., Jessen, K. R., and Mirsky, R.: The neuron-glia signal ␤ neuregulin induces sustained CREB phosphorylation on Ser-133 in cultured rat Schwann cells. Mol. Cell. Neurosci. 10:309, 1998. 325. Takashima, H., Boerkoel, C. F., De Jonghe, P., et al.: Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51:709, 2002. 326. Tanigaki, K., Nogaki, F., Takahashi, J., et al.: Notch1 and Notch3 instructively restrict bFGF-responsive multipotent neural progenitor cells to an astroglial fate. Neuron 29:45, 2001. 327. Taylor, A. R., Geden, S. E., and Fernandez-Valle, C.: Formation of a beta1 integrin signaling complex in Schwann cells is independent of rho. Glia 41:94, 2003. 328. Thatikunta, P., Qin, W., Christy, B. A., et al.: Reciprocal Id expression and myelin gene regulation in Schwann cells. Mol. Cell. Neurosci. 14:519, 1999. 329. Thompson, M. A., and Ziff, E. B.: Structure of the gene encoding peripherin, an NGF-regulated neuronal-specific type III intermediate filament protein. Neuron 2:1043, 1989. 330. Tikoo, R., Zanazzi, G., Shiffman, D., et al.: Cell cycle control of Schwann cell proliferation: role of cyclin-dependent kinase-2. J. Neurosci. 20:4627, 2000. 331. Timmerman, V., De Jonghe, P., Ceuterick, C., et al.: Novel missense mutation in the early growth response 2 gene associated with Dejerine-Sottas syndrome phenotype. Neurology 52:1827, 1999. 332. Timsit, S. G., Bally-Cuif, L., Colman, D. R., and Zalc, B.: DM-20 mRNA is expressed during the embryonic development of the nervous system of the mouse. J. Neurochem. 58:1172, 1992. 333. Tiveron, M.-C., Hirsch, M.-R., and Brunet, J.-F.: The expression pattern of the transcription factor Phox2 delineates synaptic pathways of the autonomic nervous system. J. Neurosci. 16:7649, 1996. 334. Tontonoz, P., Hu, E., and Spiegelman, B. M.: Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipidactivated transcription factor. Cell 79:1147, 1994.

Molecular Signaling in Schwann Cell Development 335. Topilko, P., Levi, G., Merlo, G., et al.: Differential regulation of the zinc finger genes Krox-20 and Krox-24 (Egr-1) suggests antagonistic roles in Schwann cells. J. Neurosci. Res. 50:702, 1997. 336. Topilko, P., and Meijer, D.: Transcription factors that control Schwann cell development and myelination. In Jessen, K. R., and Richardson, W. D. (eds.): Glial Cell Development: Basic Principles and Clinical Relevance, 2nd ed. Oxford, UK, Oxford University Press, p. 223, 2001. 337. Topilko, P., Schneider-Manoury, S., Levi, G., et al.: Krox-20 controls myelination in the peripheral nervous system. Nature 371:796, 1994. 338. Topilko, P., Schneider-Maunoury, S., Levi, G., et al.: Multiple pituitary and ovarian defects in Krox-24 (NGFI-A, Egr-1)-targeted mice. Mol. Endocrinol. 12:107, 1998. 339. Tosh, D., and Slack, J. M.: How cells change their phenotype. Nat. Rev. Mol. Cell Biol. 3:187, 2002. 340. Trachtenberg, J. T., and Thompson, W. J.: Schwann cell apoptosis at developing neuromuscular junctions is regulated by glial growth factor. Nature 379:174, 1996. 341. Trofatter, J. A., MacCollin, M. M., Rutter, J. L., et al.: A novel moesin-, ezrin-, radixin-like gene is a candidate for the neurofibromatosis 2 tumor suppressor. Cell 72:791, 1993. 342. Tsiper, M. V., and Yurchenco, P. D.: Laminin assembles into separate basement membrane and fibrillar matrices in Schwann cells. J. Cell Sci. 115:1005, 2002. 343. Uziyel, Y., Hall, S., and Cohen, J.: Influence of laminin-2 on Schwann cell-axon interactions. Glia 32:109, 2000. 344. van der Neut, R., Krimpenfort, P., Calafat, J., et al.: Epithelial detachment due to absence of hemidesmosomes in integrin beta 4 null mice. Nat. Genet. 13:366, 1996. 345. Vartanian, T., Goodearl, A., Viehover, A., and Fischbach, G.: Axonal neuregulin signals cells of the oligodendrocyte lineage through activation of HER4 and Schwann cells through HER2 and HER3. J. Cell Biol. 137:211, 1997. 346. Wakamatsu, Y., Maynard, T. M., and Weston, J. A.: Fate determination of neural crest cells by NOTCH-mediated lateral inhibition and asymmetrical cell division during gangliogenesis. Development 127:2811, 2000. 347. Wang, J. Y., Miller, S. J., and Falls, D. L.: The N-terminal region of neuregulin isoforms determines the accumulation of cell surface and released neuregulin ectodomain. J. Biol. Chem. 276:2841, 2001. 348. Wang, S., Sdrulla, A., Johnson, J. E., et al.: A role for the helix-loop-helix protein Id2 in the control of oligodendrocyte development. Neuron 29:603, 2001. 349. Wanner, I. B., and Wood, P. M.: N-cadherin mediates axon-aligned process growth and cell-cell interaction in rat Schwann cells. J. Neurosci. 22:4066, 2002. 350. Warner, L. E., Garcia, C. A., and Lupski, J. R.: Hereditary peripheral neuropathies: clinical forms, genetics, and molecular mechanisms. Annu. Rev. Med. 50:263, 1999. 351. Warner, L. E., Mancias, P., Butler, I. J., et al.: Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet. 18:382, 1998. 352. Wegner, M.: From head to toes: the multiple facets of Sox proteins. Nucleic Acids Res. 27:1409, 1999.

375

353. Wegner, M.: Transcriptional control in myelinating glia: the basic recipe. Glia 29:118, 2000. 354. Weiner, J. A., and Chun, J.: Schwann cell survival mediated by the signaling phospholipid lysophosphatidic acid. Proc. Natl. Acad. Sci. U.S.A. 96:5233, 1999. 355. Weiner, J. A., Fukushima, N., Contos, J. J., et al.: Regulation of Schwann cell morphology and adhesion by receptormediated lysophosphatidic acid signaling. J. Neurosci. 21:7069, 2001. 356. Weinmaster, G., and Lemke, G.: Cell-specific cyclic AMP-mediated induction of the PDGF receptor. EMBO J. 9:915, 1990. 357. Weinstein, D. E., Burrola, P. G., and Lemke, G.: Premature Schwann cell differentiation and hypermyelination in mice expressing a targeted antagonist of the POU transcription factor SCIP. Mol. Cell. Neurosci. 6:212, 1995. 358. White, P. M., Morrison, S. J., Orimoto, K., et al.: Neural crest stem cells undergo cell-intrinsic developmental changes in sensitivity to instructive differentiation signals. Neuron 29:57, 2001. 359. Williams, A. C., and Brophy, P. J.: The function of the periaxin gene during nerve repair in a model of CMT4F. J. Anat. 200:323, 2002. 360. Winseck, A. K., Calderó, J., Ciutat, D., et al.: In vivo analysis of Schwann cell programmed cell death in the embryonic chick: regulation by axons and glial growth factor. J. Neurosci. 22:4509, 2002. 361. Woldeyesus, M. T., Britsch, S., Riethmacher, D., et al.: Peripheral nervous system defects in erbB2 mutants following genetic rescue of heart development. Genes Dev. 13:2538, 1999. 362. Wolpowitz, D., Mason, T. B., Dietrich, P., et al.: Cysteinerich domain isoforms of the neuregulin-1 gene are required for maintenance of peripheral synapses. Neuron 25:79, 2000. 363. Wood, P. M., Schachner, M., and Bunge, R. P.: Inhibition of Schwann cell myelination in vitro by antibody to the L1 adhesion molecule. J. Neurosci. 10:3635, 1990. 364. Woodhoo, A., Dean, C. H., Droggiti, A., et al.: The trunk neural crest and its early glial derivatives: a study of survival responses, developmental schedules and autocrine mechanisms. Mol. Cell. Neurosci. 25:40, 2004. 365. Woolf, C. J., Reynolds, M. L., Chong, M. S., et al.: Denervation of the motor endplate results in the rapid expression by terminal Schwann cells of the growth-associated protein GAP-43. J. Neurosci. 12:3999, 1992. 366. Wu, R., Jurek, M., Sundarababu, S., and Weinstein, D. E.: The POU gene Brn-5 is induced by neuregulin and is restricted to myelinating Schwann cells. Mol. Cell. Neurosci. 17:683, 2001. 367. Yamashita, T., Tucker, K. L., and Barde, Y. A.: Neurotrophin binding to the p75 receptor modulates Rho activity and axonal outgrowth. Neuron 24:585, 1999. 368. Yoshihara, T., Kanda, F., Yamamoto, M., et al.: A novel missense mutation in the early growth response 2 gene associated with late-onset Charcot-Marie-Tooth disease type 1. J. Neurol. Sci. 184:149, 2001. 369. Yoshimura, T., Goda, S., Kobayashi, T., and Goto, I.: Involvement of protein kinase C in the proliferation of cultured Schwann cells. Brain Res. 617:55, 1993.

376

Neurobiology of the Peripheral Nervous System

370. Zanazzi, G., Einheber, S., Westreich, R., et al.: Glial growth factor/neuregulin inhibits Schwann cell myelination and induces demyelination. J. Cell Biol. 152:1289, 2001. 371. Zelena, J.: Nerves and Mechanoreceptors. London, Chapman and Hall, 1994. 372. Zhang, D., Sliwkowski, M. X., Mark, M., et al.: Neuregulin-3 (NRG3): a novel neural tissue-enriched protein that binds and activates ErbB4. Proc. Natl. Acad. Sci. U.S.A. 94:9562, 1997. 373. Zhang, J. Y., Luo, X. G., Xian, C. J., et al.: Endogenous BDNF is required for myelination and regeneration of injured sciatic nerve in rodents. Eur. J. Neurosci. 12:4171, 2000.

374. Zilian, O., Saner, C., Hagedorn, L., et al.: Multiple roles of mouse Numb in tuning developmental cell fates. Curr. Biol. 11:494, 2001. 375. Zorick, T. S., Syroid, D. E., Arroyo, E., et al.: The transcription factors SCIP and Krox-20 mark distinct stages and cell fates in Schwann cell differentiation. Mol. Cell. Neurosci. 8:129, 1996. 376. Zorick, T. S., Syroid, D. E., Brown, A., et al.: Krox-20 controls SCIP expression, cell cycle exit and susceptibility to apoptosis in developing myelinating Schwann cells. Development 126:1397, 1999.

17 Neurotrophic Factors in the Peripheral Nervous System ANTHONY J. WINDEBANK AND ELIZABETH S. MCDONALD

Neurotrophin Families Nerve Growth Factor Brain-Derived Neurotrophic Factor Neurotrophin-3/Neurotrophin-4

Ciliary Neurotrophic Factor Glial Cell Line–Derived Neurotrophic Factor Insulin-like Growth Factor

Growth factors may be defined as soluble extracellular macromolecules that influence the proliferation, growth, or differentiation of target cells by a cell surface receptor– mediated mechanism. “Soluble” is included in the definition because the known growth factors are assumed to work in vivo by signaling from a target tissue (e.g., muscle) to a neuronal subpopulation (e.g., the anterior horn cell). This process involves local diffusion from the target to the neuron. However, it is probable that extracellular matrix proteins such as laminin, fibronectin, and collagen play important trophic roles for neurons, especially during peripheral nerve regeneration. Similarly, direct cell membrane–to–cell membrane contact by way of surface-bound neuronal cell adhesion molecules may influence neuronal growth and differentiation. For operational convenience, small molecules, such as neurotransmitters, are not included in the definition of growth factors because little is known about their roles in modulating growth and differentiation. It is possible that these small molecules play an important part in modulating or determining cell survival or specific aspects of cellular differentiation such as maintenance of synapse structure. Practically all of the neuronal growth factors studied to date are polypeptides that have biologically active monomeric subunits with molecular weights in the 10,000- to 30,000-Da range. Nonpolypeptide macromolecules have not been identified as growth factors, although glycosaminoglycans such as heparin may be important in modulating the action of growth factors in vivo.

Neurotrophin Signaling Clinical Applications of Neurotrophic Factors

Growth factors are presumably involved in many different stages of the development and maintenance of the final shape and connectivity of neurons. A simplified conceptual framework is illustrated in Figure 17–1. This type of scheme has been developed from the work of Hamburger and Keefe34 and others, who demonstrated that ablation of a target, such as a limb bud, or addition of extra target tissue during embryonic development resulted in a decrease or increase, respectively, in the number or size of neurons innervating that target. These observations led directly to the discovery of nerve growth factor (NGF). It is assumed that timing of interaction, especially during embryonic development, is critical. Thus a population of neurons may express specific growth factor receptors on their surface at one stage in their development and not at others. It is also assumed that these molecules work over short distances within tissues and that this local interaction may provide for some of the specificity of action. Understanding timing of receptivity and tissue localization becomes increasingly important as human recombinant growth factors become available as potential therapeutic agents. A final general issue to be considered is the naming of growth factors. Names in present conventional use reflect either the tissue from which the factor was originally derived (e.g., ciliary neurotrophic factor [CNTF], brain-derived neurotrophic factor [BDNF]) or the target cell upon which the factor was found to act in vitro (e.g., fibroblast growth factor). These original discoveries may not reflect the true functions of the factors in vivo. However, because none 377

378

Neurobiology of the Peripheral Nervous System

NGF NGF

NGF NGF Cell membrane

Y

Grb2 Y490 Shc SOS Gab1 Shp-2 Y

785

Ras Raf

PI3k PLCγ IP3 + DAG

GSK3

PKCδ

P* P*

MEK1/2

PDKs

Akt1/2

14-3-3 P*

MAPK1/2

p53

BAD ser136

IKKα

Forkhead P*

BAD Bcl-2

neurite outgrowth

Rsk NFκB

IκB P* NFκB CREB IκB Nucleus

Bcl-xL FasL Bax Bcl-2 IAPs (+) (+) (+) (+) Regulation of survival

FIGURE 17–1 Tyrosine receptor kinase A (TrkA) signal transduction pathways. Nerve growth factor binding induces autophosphorylation of the TrkA receptor. This recruits proteins to the TrkA phosphotyrosine residues that activate a variety of signaling pathways that aid neuron survival and differentiation. These include the phosphatidylinositol 3 kinase (PI3K)–Akt, extracellular signal-regulated kinase (ERK)–Ras, and phospholipase C-␥ 1 (PLC-␥ 1) pathways. Proteins serving an adaptor function are colored green, the small G protein is blue, kinases are purple, and transcription factors are red. (Data from Huang and Reichardt38 and Kaplan and Miller.41) See Color Plate

of these “true” actions is completely understood, it seems reasonable to retain for now the conventional names.

NEUROTROPHIN FAMILIES There are three families of characterized growth factors that promote differentiation and survival in the nervous system. The first is the classic neurotrophins: NGF, BDNF, and neurotrophins 3 through 7 (NT-3 through NT-7). The other two main families are glial cell line–derived neurotrophic factor (GDNF) and CNTF.13 Another family of growth factors, insulin-like growth factors (IGFs), have been traditionally studied as cytokine-like growth factors for non-neuronal cells but also play a role in neuronal development.92 The three-dimensional structures of the classic neurotrophins (NGF, BDNF, NT-3, and NT-4) have been determined through x-ray crystallography. They have

a common core structure of three disulfide bonds that join two pairs of antiparallel two-strand beta sheets (a “cysteine knot” structure). The GDNF family members also have the cysteine knot fold. The CNTF structure is different in that it has no cysteine knot but consists of four helices with several crossover loops.13

Nerve Growth Factor NGF, the first neurotrophin to be characterized, was discovered in the 1950s through the joint work of Viktor Hamburger and Rita Levi-Montalcini.20 NGF was originally studied because of a mouse sarcoma’s unexpected ability to support sympathetic and sensory spinal neurons in vitro. While trying to purify the supportive factor, it was discovered that snake venom and mouse submandibular salivary glands had an even more potent neurotrophic effect.51 The breakthrough that led to the isolation and characterization of the

Neurotrophic Factors in the Peripheral Nervous System

NGF protein was the finding that it was present in large quantities in the tubular portions of adult male mouse submandibular glands.12,51,78 The active form of NGF is a dimer of 13-kDa polypeptide chains that each have three intrachain disulfide bridges.79 The targets of NGF in the peripheral nervous system (PNS) are sympathetic and small-diameter, unmyelinated (C fiber) temperature and pain dorsal root ganglia (DRG) sensory neurons. These neurons express the NGF receptors tyrosine receptor kinase A (TrkA) and p75 low-affinity neurotrophin receptor (p75NTR).27,92 Cells that produce NGF include many targets of neurons during development, such as keratinocytes, melanocytes, vascular and smooth muscle cells, testis and ovarian cells, and cells of the pituitary, thyroid, parathyroid, and exocrine salivary glands.79 In the adult, NGF also plays a role in injury response. Cytokines released by macrophages after peripheral nerve injury cause Schwann cells and fibroblasts to synthesize NGF. Mast cells also synthesize and release NGF.38,52 NGF is part of a neurotrophin family that acts in an autocrine or paracrine fashion on cell surface receptors. The growth factors are present in limited quantities during development, when they function to promote neuronal survival and differentiation.39,51 In addition, neurotrophins have been shown to regulate cell proliferation, axonal and dendritic growth and remodeling, assembly of the cytoskeleton, membrane trafficking and fusion, and synapse formation and function.38 The four neurotrophins characterized in mammals (NGF, BDNF, NT-3, and NT-4)38 promote vertebrate nervous system development.48 Neurotrophin-deficient mutant mice die postnatally, with the exception of those deficient in NT-4.27 Unlike insulin or transforming growth factor-␤, these classic neurotrophins do not have homologues in Caenorhabditis elegans or Drosophila melanogaster.48 Two more neurotrophins, NT-6 and NT-7, have been isolated in fish and are not thought to have mammalian homologues.38 The neurotrophins are made as 30- to 35-kDa precursor proteins that are intracellularly cleaved to 118 to 120 amino acids. The precursor proteins have glycosylation and processing enzyme sites and a signal peptide. The mature proteins are 12- to 14-kDa C-terminal products. They are active as 26-kDa homodimers.16 NGF interacts with two receptors: p140 tyrosine receptor kinase (TrkA or p140TRK) and p75NTR. The TrkA receptor is responsible for the neurotrophic and neuroprotective actions of NGF.30,32 Cells that express the TrkA receptor include sympathetic neurons, DRG neurons, and cholinergic interneurons.45 Tyrosine receptor kinase (Trk) was first discovered as an oncogene in colon carcinoma. The extracellular TrkA kinase domain was fused to tropomyosin, which allows Trk constitutive tyrosine kinase activity. The proto-oncogene, TrkA, was expressed in the nervous system. Later researchers determined that neurotrophins bound to and activated this receptor, which was important in developmental neuronal death and neuronal differentiation.62,67

379

The Trk receptors have been well studied and have five domains: A cysteine-rich cluster constitutes domains one and three, three leucine-rich repeats form domain two, and the last two domains are immunoglobulin-like. The evidence so far supports the involvement of domain five in ligand binding.87 TrkA signaling primarily affects survival and differentiation of small-diameter pain and temperature sensory neurons and sympathetic neurons.62 Mice engineered to be deficient in TrkA do not make thermoceptive or nociceptive neurons.67 The mature form of the neurotrophin NGF interacts with the TrkA receptor with high affinity (KD ⫽ 2.3 ⫻ 10⫺11 M) and the p75NTR receptor with lower affinity (KD ⫽ 1.7 ⫻ 10⫺9 M).80 The immature or precursor protein, which is cleaved intracellularly to the active NGF homodimer, binds to the p75NTR fivefold stronger than mature NGF (10⫺10 M). The ratio of immature to mature NGF influences p75NTR activation. Thus proteolytic cleavage of neurotrophins may be a primary regulator of neuronal survival. The affinity of other immature neurotrophins for p75NTR is still under investigation.16,48 High-affinity binding to endogenous TrkA only occurs with concomitant expression of p75NTR.5,64 The binding of low-concentration NGF to TrkA is regulated by both TrkA and p75NTR levels.47 All of the neurotrophins bind p75NTR. In addition, BDNF and NT-4 bind specifically to tyrosine receptor kinase B (TrkB) and NT-3 specifically to tyrosine receptor kinase C (TrkC).39 NT-3 and adenosine, normally a G protein–coupled receptor agonist, can also activate TrkA and TrkB.46,67 TrkB and TrkC signaling primarily affects proprioceptive and light touch sensory neurons as well as motor neurons.62 p75NTR can prevent or promote cell death.6 Cells that express p75NTR include sympathetic, small-diameter sensory as well as alpha motor neurons and Schwann cells during development and injury.67 About 10 years before transport and endocytosis were studied in non-neuronal cells, the mechanism of NGF internalization and transport was initially defined. NGF is taken into the cell at the axon terminal by a receptordependent mechanism and transported along axons to the cell body in membrane vesicles by microtubule- and energy-dependent mechanisms. Eventually, the transport vesicles fuse with late endosomes at the cell bodies and then to lysosomes, where NGF is degraded.38,64 The neuronal cell body can be up to 1 m from the axon, and the speed of this transport mimics vesicular transport, with estimates ranging from 2 to 20 mm/h in peripheral neurons.57 The transport vesicles use dynein motor proteins to transport retrogradely along axonal microtubules.64 When axons in Campenot chambers are stimulated with neurotrophins, the Trk receptor quickly moves to the cell-body compartment. Some of this Trk is bound to neurotrophin, but there is also evidence that retrogradely transported Trk receptor signals without bound ligand.57 Recently, clathrin-coated vesicles retrogradely transporting NGF

380

Neurobiology of the Peripheral Nervous System

and TrkA were also found to contain activated proteins of the Ras–myelin-associated protein kinase (MAPK) pathway. These endosomes were able to activate a downstream target in the MAPK pathway.37 Autophosphorylated Trk movement to cell bodies is essential for NGF-mediated signal transduction of cyclic AMP–regulated enhancer binding protein (CREB) to neuronal nuclei.70

Brain-Derived Neurotrophic Factor The second neurotrophin, discovered in 1982, was purified from porcine brain and called brain-derived neurotrophic factor. Neurons responsive to BDNF include DRGs, spinal motor neurons, and nodose and trigeminal ganglion sensory neurons. BDNF helps the survival of motor neurons after nerve axotomy and can also prevent developmental motor neuron cell death. After nerve transection, BDNF messenger RNA (mRNA) increases in muscle and TrkB mRNA increases in motor neurons.92 BDNF and NT-3 regulate myelination in the PNS.15 BDNF can rescue about a third of postnatal DRG neurons in culture. Likewise, the BDNF null mutant mouse loses about a third of its DRG neurons after birth.27 BDNF may also have a role in memory and learning, because long-term potentiation is impaired in the mutant mouse.89 Neurotrophins are produced in target cells for presynaptic neurons. They are then retrogradely transported to the neuronal cell body. There is recent evidence that BDNF can act as an anterograde as well as a retrograde trophic factor. It is released at brain neuron synapse terminals and can also act on postsynaptic neurons.63 BDNF has been used intrathecally in Phase I/II65 and subcutaneously in Phase I through III clinical trials for amyotrophic lateral sclerosis (ALS)82 and in Phase II trials for diabetic polyneuropathy86,92 with no benefit demonstrated.

Neurotrophin-3/Neurotrophin-4 NT-3 is mainly expressed in muscle spindles, Merkel cells, and Golgi tendon organs. There is evidence that it aids differentiation of sympathetic cholinergic neurons and that its effects are antagonized by NGF.10 NT-3 mutant mice lose about 60% of their DRG neurons during embryogenesis between embryonic day (E) 10.5 and E13, before NGF or BDNF dependence has begun. Mice engineered to be deficient in the NT-3 receptor TrkC do not make proprioceptive neurons.67 When TrkC is experimentally removed, there is evidence that NT-3 can signal through the TrkA and TrkB receptors. NT-4 (also known as NT-5) regulates low-threshold hair afferent mechanoreceptors.27

Ciliary Neurotrophic Factor CNTF was first purified in 1979 from chick nerve and eye extracts that promoted survival of ciliary ganglionic neurons. It also supports dopaminergic, retinal rod, sympathetic, and

motor neurons. It may play a role in body weight homeostasis and PNS injury response.13,50,64 CNTF is a nonsecreted ␣-helical cytokine that first binds to ciliary neurotrophic factor receptor ␣ (CNTFR␣), followed by two other receptor subunits: gp130 and leukemia inhibitory factor receptor ␤ (LIFR␤). CNTFR␣ is tethered to the cell membrane with a glycosyl phosphatidylinositol anchor. It confers binding specificity. Gp130 is a nonspecific, high-affinity, signaltransducing subunit. The LIFR␤ and gp130 intracellular domains can bind STAT or JAK.13,50,64 A secreted ligand for CNTFR␣ was discovered in 2000: cardiotrophin-like cytokine (CLC), also identified as novel neurotrophin-1. CLC binds to the CNTF receptor after forming a complex with another secreted protein, cytokine-like factor-1 (CLF-1). CLF-1 was originally cloned as a cytokine receptor but functions in this case as a ligand.50

Glial Cell Line–Derived Neurotrophic Factor GDNF was first purified in 1993 from a glial cell line culture that promoted neuronal survival. It can support embryonic midbrain dopaminergic, sympathetic, and spinal motor neurons. It is required for development of the enteric nervous system, ureters, and kidneys.13,26,59,68,73,76 The GDNF family includes three other trophic factors: persephin (PSP), artemin (ART), and neurturin (NTN). They share a common tyrosine kinase receptor, c-Ret. In addition, they each have a receptor that allows ligand specificity. These receptors are named GDNF family receptor ␣ (GFR␣)-1 through -4 and are tethered to the cell membrane with a glycosyl phosphatidylinositol anchor. GDNF␣-1 preferentially binds GDNF, GDNF␣-2 preferentially binds NTN, GDNF␣-3 preferentially binds ART, and GDNF␣-4 preferentially binds PSP. Some GDNF family members can bind more than one “specific” receptor.64 This is similar to the crossover binding seen between classic neurotrophic factors and Trk receptors. The active GDNF family receptor is a heterohexameric complex made up of two Ret receptors and two GDNF ␣ receptors. It then signals through the phosphatidylinositol 3 kinase (PI3K) and MAPK pathways.13

Insulin-like Growth Factor The structure of IGF is about 50% homologous to insulin. The tyrosine kinase IGF-I receptor is also homologous to the insulin receptor and is expressed throughout the nervous system. The IGFs promote survival and neuronal outgrowth in developing sympathetic and sensory neurons. IGF-I promotes motor neuron survival of developing neurons after axotomy. IGF-I injection can also induce motor neuron axonal growth in adult mammals, and antiserum to IGF-I and IGF-II inhibits regeneration of motor axons after spinal root injury. IGF-I has been used in two Phase III clinical trials for ALS. The first showed a modest benefit and the second showed no benefit. A meta-analysis suggested a modest

Neurotrophic Factors in the Peripheral Nervous System

but unproven beneficial effect for IGF-I in the treatment of ALS.58 It has also been used in Phase II trials for diabetic polyneuropathy, chemotherapy-induced neuropathy, and postpolio syndrome.92 IGF-I is a potential therapeutic modality to promote neuronal regeneration.

NEUROTROPHIN SIGNALING The extracellular TrkA-d5 immunoglobulin-like domain proximal to the cell membrane is both necessary and sufficient for the binding of NGF to TrkA.88 After binding, the tyrosine kinase in the receptor cytoplasmic domain is phosphorylated and activated. The Trk receptors contain 10 conserved cytoplasmic domain tyrosines. Three of these (Y670, Y674, and Y675) form an autoregulatory loop in the receptor kinase domain that controls kinase activity. When the other tyrosine residues are phosphorylated, adapter proteins with Src homology 2 or phosphotyrosine-binding motifs can bind and activate different intracellular signaling cascades. The intracellular signaling cascades that are activated include PI3K-Akt, extracellular signal-regulated kinase (ERK)-Ras, and phospholipase C-␥ 1 (PLC-␥ 1)38 (see Fig. 17–1). PLC-␥ 1 is recruited to the membrane and phosphorylated by the TrkA kinase after phosphorylation of TrkA Y785. PLC-␥ 1 hydrolyses phosphatidylinositides to form inositol triphosphate and diacylglycerol (DAG). Inositol triphosphate stimulates release of Ca2⫹ into the cytoplasm from the endoplasmic reticulum. The cytoplasmic calcium then activates Ca2⫹-regulated isoforms of protein kinase C (PKC) and Ca2⫹-calmodulin–regulated protein kinases and phosphatases. DAG can activate DAG-regulated PKC isoforms.38,83 One DAG-dependent, Ca2⫹-independent isoform of PKC is PKC␦. PKC␦ activity is required in PC12 cells for neurite outgrowth and activation of ERK in response to NGF.19 Phosphorylation of the TrkA receptor on Y490 activates the Ras-MAPK pathway. The adaptor protein Shc is recruited to the TrkA receptor, and this allows subsequent binding of the adaptor protein Grb-2 and Ras exchange factor SOS. SOS, when recruited to the cell membrane, can activate Ras by allowing exchange of GDP for GTP. This activates the c-Raf–ERK and PI3K pathways, which are the main regulators of Ras- and neurotrophin-mediated survival.41,54 There are many important signaling targets of the serine-threonine kinases c-Raf and ERK, including ribosomal S6 kinase. Activation of these kinases can lead to the phosphorylation of CREB, a transcription factor that promotes neuron survival. Inhibition of CREB phosphorylation can trigger apoptosis.9 Upon activation, this transcription factor binds to a cyclic AMP response element inside the c-fos promoter, allowing transcription of genes needed for initiation and maintenance of differentiation.7,79

381

Ras activation of PI3K is mediated through the adaptor protein Grb2-associated binder 1 (Gab-1). The phosphatidylinositides formed by the PI3K activate the kinases PDK1 and PDK2, which in turn phosphorylate and activate the protein kinase Akt (protein kinase B ␣).41 Akt is a serine-threonine kinase that promotes cell survival. It acts through downstream effectors such as Bad, caspase-9, forkhead, nuclear factor-␬B (NF-␬B), and Pak.11,14,21,23,40,69,81 The phosphorylation of Bad causes association with 14-3-3 proteins that inhibit its pro-apoptotic function.21 Akt also phosphorylates the inhibitor of I␬B kinase (IKK␣) at threonine 23. NF-␬B is usually sequestered in the cytoplasm bound to the I␬B inhibitor protein. The phosphorylation of IKK␣ by both Akt and NF-␬B–inducing kinase (NIK) allows subsequent degradation of the I␬B complex (IKK␣ and IKK␤). NF-␬B can then translocate to the nucleus, where it acts as a survival-promoting transcription factor.24,66,71,91 Another Akt substrate, forkhead, can control the expression of genes involved in apoptosis, such as Fas ligand.11 The IAP caspase inhibitor family is also potentially activated by Akt.41 There is in vivo evidence that Akt is necessary for the neuroprotective effects of IGF-I.43 Mutations that activate Akt have been found in certain tumors.28 Activated Akt has also been shown to inhibit both apoptosis and changes in Bax expression induced by nitric oxide in primary hippocampal neurons.53 It also has the ability to prevent cytochrome c release from the mitochondria42 and delay the onset of p53-mediated apoptosis.72 p75NTR binds to all of the mature neurotrophins with a similar affinity. Receptor activation can both prevent and promote cell death6 as well as regulate axon growth and mediate neurotrophin retrograde transport.67 Neurotrophin activation of NF-␬B through p75NTR can promote neuronal survival.33 NGF-mediated activation of p75NTR allows association of a tumor necrosis factor receptor–associated factor (TRAF) adaptor protein.44 TRAF2, TRAF5, and TRAF6 activate NF-␬B.4,44 NIK is part of the signaling complex assembled with TRAF proteins after they are recruited to a surface receptor. NF-␬B is activated when NIK phosphorylates IKK. IKK then phosphorylates I␬B, which is degraded by proteases after ubiquination. This releases NF-␬B to translocate to the nucleus, where it can bind and activate promoters of responsive genes4 (Fig. 17–2). Activation of the p75NTR has been associated with apoptosis of sensory and motor neurons, sympathetic neurons, hippocampal neurons, and oligodendrocytes.56 p75NTR can cause apoptosis via activation of the Jun N-terminal kinase (JNK)–mediated signaling pathway, with subsequent activation of p53.2 Activation of the JNK pathway leads to Fas ligand upregulation.49 p53 can bind and transactivate the fas gene at a responsive element located within the first intron.61 There is evidence that neuronal death from neurotrophin withdrawal depends on the presence of p75NTR,6 p53,22,84 and Fas receptor.55 Thus death from NGF withdrawal probably

382

Neurobiology of the Peripheral Nervous System

p75NTR Cell membrane TRAF6

Traf 2/4

Ras TrkA RhoA

SC-1

NRIF NIK PI3K

Ceramide

NRAGE

P*

P* IκB P* p53 IκB

IKKα

Bax

NFκB

Cell death

Jun kinase pathway Nucleus NFκB

Apoptosis Cell survival

FIGURE 17–2 p75NTR signal transduction pathways. Neurotrophin binding causes the association of adaptor proteins that mediate prosurvival and prodeath pathways. Activation of the pro-apoptotic JNK pathway can be inhibited by concurrent TrkA activation. Activation of NF-␬B through the association of TRAF proteins leads to cell survival. Proteins serving an adaptor function are colored green, small G proteins are blue, kinases are purple, and transcription factors are red. (Data from Huang and Reichardt38 and Kaplan and Miller.41) See Color Plate

occurs through both a JNK-p53-Bax and a JNK-p53-Fas pathway. In the presence of NGF and TrkA signaling, the pro-apoptotic p75NTR pathways are suppressed while the prosurvival pathways are activated.41 If NGF is present in cells without TrkA (such as oligodendrocytes), then p75NTR activation will signal for apoptosis. If BDNF is present in cells without TrkB (such as sympathetic neurons), then p75NTR activation will signal for apoptosis. Thus Trk receptor expression regulates the pro- or antiapoptotic potential of p75NTR. Trk may inhibit p75NTR-mediated JNK activation via Ras and the PI3K-Akt pathway.56 p75NTR activation by NGF, in the absence of a TrkA receptor, induces sphingomyelin hydrolysis, resulting in the formation of the sphingolipid metabolite ceramide.25 Ceramide binds to Raf-1 and promotes its association with GTP-Ras without subsequent Raf-1 kinase activation. This inhibits the survival-promoting MAPK/ERK cascade.60 Ceramide also inhibits PI3K signaling.93,94 There are other proteins that interact with the p75NTR receptor, such as neurotrophin receptor interacting factor, neurotrophin

receptor-interacting MAGE (melanoma-associated antigen) homologue, and Schwann cell-1. These proteins may be involved in ceramide production or JNK activation.38 Activation of p75NTR may be partially responsible for injury-induced apoptosis, because p75NTR⫺/⫺ animals show less Schwann cell apoptosis after sciatic nerve axotomy. p75NTR is also induced in central nervous system neurons dying from excitotoxicity.56 NGF protects neurons from death resulting from various insults, including axotomy, ischemia, oxidative stress, and glutamate receptor–mediated excitotoxicity. NGF can also protect PC12 cells and DRG neurons from anoxia and glucose deprivation.8,35,36,79 The addition of NGF rescues PC12 cells from death after treatment with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP).77 Pretreatment with NGF can prevent nitric oxide cytotoxicity in PC12 cells85 as well as excitotoxic injury in vivo.1 NGF also protects sensory neurons from death after treatment with chemotherapeutic drugs such as cisplatin. Cisplatin causes apoptosis of DRG neurons.29 High-dose NGF protects against cisplatin-mediated apoptotic DRG

383

Neurotrophic Factors in the Peripheral Nervous System

death.31 NGF may protect neurons from cisplatin toxicity through NGF-mediated activation of Akt kinase and inhibition of Bax translocation. Cisplatin causes Bax translocation and subsequent death in DRG neurons. High-dose NGF prevents this translocation.55 DRG neurons that lack the gene for Bax have very little death when exposed to lethal doses of cisplatin. Activated Akt inhibits apoptosis and changes in Bax expression induced by nitric oxide in primary hippocampal neurons.53 Apoptosis of primary hippocampal neurons as a result of nitric oxide or hypoxia is dependent on the pro-apoptotic effects of p53. Exogenous Akt kinase suppresses p53-mediated transcriptional activation of Bax, caspase-3 activity, and cell death.90 Akt also has the ability to prevent cytochrome c release from the mitochondria42 and delay the onset of temperature-dependent p53-mediated apoptosis.72 There is in vitro and in vivo evidence that Akt is necessary for the neuroprotective effects of IGF-I.18,43 Akt may be necessary for the neuroprotective effects of NGF. The rescue of PC12 cells from death resulting from treatment with MPTP is prevented by co-treatment with a dominant negative form of Akt.77 The TrkA receptor inhibits p75NTR-mediated activation of the pro-apoptotic JNK pathway.41 Cisplatin can also activate the JNK pathway. Inhibition of this pathway decreases cisplatin-mediated cellular death.74,75 Thus TrkA activation could also be inhibiting cisplatin-mediated JNK activation and subsequent cell death.

CLINICAL APPLICATIONS OF NEUROTROPHIC FACTORS Neurotrophic factors have been successfully used to treat a variety of animal disease states, including motor neuron disease, toxic neuropathy, diabetic neuropathy, spinal cord and nerve trauma, stroke, and models of Alzheimer’s and Parkinson’s diseases. Unfortunately, when used in human trials, there have been few successes. NGF infusion has been tested for possible treatment of human immunodeficiency virus–associated neuropathy, diabetic neuropathy, and Alzheimer’s disease. Attempts to treat ALS have included clinical trials with BDNF, CNTF, and IGF-I.3 NGF, BDNF, NT-3, and IGF-I have all been tested in Phase II or III human clinical trials for various neurologic diseases. Adverse reactions have prevented GDNF from moving beyond Phase I clinical trials.92 Thus far, no Phase III trial of growth factor treatment has led to a new standard in disease therapy. Many of these trials have used systemically administered growth factors, which may not have access to neurons protected by a blood-nerve or blood-brain barrier. In addition, intermittent-bolus dosing has been used instead of more physiologic steady exposure. Systemic side effects have limited dosage below the amounts used in the successful animal trials even when effective delivery (such as to

sensory neurons exposed to systemic circulation) was possible.3 Thus it may be early to conclude that growth factors will never be used therapeutically. Targeted drug delivery and more physiologic dosing, which would also mitigate systemic reaction, may allow successful use of growth factors in the treatment of neurologic disease. Recently, an advance has been made with successful NT-3 gene transfer using herpes simplex virus. Endogenous herpes simplex virus targets peripheral sensory neurons, where it establishes a latent infection. Expression of NT-3 from a replication-incompetent viral vector protected DRG neurons from pyridoxine-mediated neuropathy.17 Viral vectors have also been used to successfully treat animal models of Parkinson’s disease, Huntington’s disease, and motor neuron disease.3 Thus viral vectors are one possibility to allow administration of neurotrophins to treat neurologic disease in both the peripheral and central nervous systems.

REFERENCES 1. Aloe, L.: Intracerebral pretreatment with nerve growth factor prevents irreversible brain lesions in neonatal rats injected with ibotenic acid. Biotechnology 5:1085, 1987. 2. Aloyz, R. S., Bamji, S. X., Pozniak, C. D., et al.: p53 is essential for developmental neuron death as regulated by the TrkA and p75 neurotrophin receptors. J. Cell Biol. 143:1691, 1998. 3. Apfel, S. C.: Is the therapeutic application of neurotrophic factors dead? Ann. Neurol. 51:8, 2002. 4. Arch, R. H., Gedrich, R. W., and Thompson, C. B.: Tumor necrosis factor receptor-associated factors (TRAFs)—a family of adapter proteins that regulates life and death. Genes Dev. 12:2821, 1998. 5. Barker, P. A., and Shooter, E. M.: Disruption of NGF binding to the low affinity neurotrophin receptor p75LNTR reduces NGF binding to TrkA on PC12 cells. Neuron. 13:203, 1994. 6. Barrett, G. L., and Bartlett, P. F.: The p75 nerve growth factor receptor mediates survival or death depending on the stage of sensory neuron development. Proc. Natl. Acad. Sci. U. S. A. 91:6501, 1994. 7. Berkowitz, L. A., Riabowol, K. T., and Gilman, M. Z.: Multiple sequence elements of a single functional class are required for cyclic AMP responsiveness of the mouse c-fos promoter. Mol. Cell. Biol. 9:4272, 1989. 8. Boniece, I. R., and Wagner, J. A.: Growth factors protect PC12 cells against ischemia by a mechanism that is independent of PKA, PKC, and protein synthesis. J. Neurosci. 13:4220, 1993. 9. Bonni, A., Brunet, A., West, A. E., et al.: Cell survival promoted by the Ras-MAPK signaling pathway by transcriptiondependent and -independent mechanisms. Science 286:1358, 1999. 10. Brodski, C., Schnurch, H., and Dechant, G.: Neurotrophin-3 promotes the cholinergic differentiation of sympathetic neurons. Proc. Natl. Acad. Sci. U. S. A. 97:9683, 2000. 11. Brunet, A., Bonni, A., Zigmond, M. J., et al.: Akt promotes cell survival by phosphorylating and inhibiting a forkhead transcription factor. Cell 96:857, 1999.

384

Neurobiology of the Peripheral Nervous System

12. Burdman, J. A., and Goldstein, M. N.: Synthesis and storage of a nerve growth protein in mouse submandibular glands. J. Exp. Zool. 160:183, 1965. 13. Butte, M. J.: Neurotrophic factor structures reveal clues to evolution, binding, specificity, and receptor activation. Cell. Mol. Life Sci. 58:1003, 2001. 14. Cardone, M. H., Roy, N., Stennicke, H. R., et al.: Regulation of cell death protease caspase-9 by phosphorylation. Science 282:1318, 1998. 15. Chan, J. R., Cosgaya, J. M., Wu, Y. J., and Shooter, E. M.: Neurotrophins are key mediators of the myelination program in the peripheral nervous system. Proc. Natl. Acad. Sci. U. S. A. 98:14661, 2001. 16. Chao, M. V., and Bothwell, M.: To cleave or not to cleave. Neuron. 33:9–12, 2001. 17. Chattopadhyay, M., Wolfe, D., Huang, S., et al.: In vivo gene therapy for pyridoxine-induced neuropathy by herpes simplex virus-mediated gene transfer of neurotrophin-3. Ann. Neurol. 51:19, 2002. 18. Cheng, H. L., Steinway, M., Delaney, C. L., et al.: IGF-I promotes Schwann cell motility and survival via activation of Akt. Mol. Cell. Endocrinol. 170:211, 2000. 19. Corbit, K. C., Foster, D. A., and Rosner, M. R.: Protein kinase C␦ mediates neurogenic but not mitogenic activation of mitogen-activated protein kinase in neuronal cells. Mol. Cell. Biol. 19:4209, 1999. 20. Cowan, W. M.: Viktor Hamburger and Rita Levi-Montalcini: the path to the discovery of nerve growth factor. Annu. Rev. Neurosci. 24:551, 2001. 21. Datta, S. R., Dudek, H., Tao, X., et al.: Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 91:231, 1997. 22. Deckwerth, T. L., Elliott, J. L., Knudson, C. M., et al.: BAX is required for neuronal death after trophic factor deprivation and during development. Neuron 17:401, 1996. 23. del Peso, L., Gonzalez-Garcia, M., Page, C., et al.: Interleukin-3-induced phosphorylation of BAD through the protein kinase Akt. Science 278:687, 1997. 24. Demarchi, F., Verardo, R., Varnum B., et al.: Gas6 antiapoptotic signaling requires NF-kappa B activation. J. Biol. Chem. 276:31738, 2001. 25. Dobrowsky, R. T., Jenkins, G. M., and Hannun, Y. A.: Neurotrophins induce sphingomyelin hydrolysis: modulation by co-expression of p75NTR with Trk receptors. J. Biol. Chem. 270:22135, 1995. 26. Enomoto, H., Araki, T., Jackman, A., et al.: GFR ␣1-deficient mice have deficits in the enteric nervous system and kidneys. Neuron 21:317, 1998. 27. Ernfors, P.: Local and target-derived actions of neurotrophins during peripheral nervous system development. Cell. Mol. Life Sci. 58:1036, 2001. 28. Evan, G. I., and Vousden, K. H.: Proliferation, cell cycle and apoptosis in cancer. Nature 411:342, 2001. 29. Fischer, S. J., McDonald, E. S., Gross, L., and Windebank, A. J.: Alterations in cell cycle regulation underlie cisplatin induced apoptosis of dorsal root ganglion neurons in vivo. Neurobiol. Dis. 8:1027, 2001. 30. Fischer, S. J., Podratz, J. L., and Windebank, A. J.: Nerve growth factor rescue of cisplatin neurotoxicity is mediated

31.

32.

33.

34.

35.

36.

37.

38.

39.

40. 41.

42.

43.

44.

45. 46.

through the high affinity receptor: studies in PC12 cells and p75 null mouse dorsal root ganglia. Neurosci. Lett. 308:1, 2001. Gill, J. S., and Windebank, A. J.: Cisplatin-induced apoptosis in rat dorsal root ganglion neurons is associated with attempted entry into the cell cycle. J. Clin. Invest. 101:2842, 1998. Greene, S. H., Rydel, R. E., Connolly, J. L., and Greene, L. A.: PC12 cell mutants that possess low- but not highaffinity nerve growth factor receptors neither respond to nor internalize nerve growth factor. J. Cell Biol. 102:830, 1986. Hamanoue, M., Middleton, G., Wyatt, S., et al.: p75-mediated NF-kappaB activation enhances the survival response of developing sensory neurons to nerve growth factor. Mol. Cell. Neurosci. 14:28, 1999. Hamburger, V., and Keefe, E. L.: The effects of peripheral factors on the proliferation and differentiation in the spinal cord of chick embryos. J. Exp. Zool. 96:223, 1944. Honma, H., Gross, L., and Windebank, A. J.: Hypoxia-induced apoptosis of dorsal root ganglion neurons is associated with DNA damage recognition and cell cycle disruption. Neurosci. Lett. 354:95, 2004. Honma, H., Podratz, J. L., and Windebank, A. J.: Acute glucose deprivation leads to apoptosis in a cell model of acute diabetic neuropathy. J. Peripher. Nerv. Syst. 8:65, 2003. Howe, C. L., Valletta, J. S., Rusnak, A. S., and Mobley, W. C.: NGF signaling from clathrin-coated vesicles: evidence that signaling endosomes serve as a platform for the Ras-MAPK pathway. Neuron 32:801, 2001. Huang, E. J., and Reichardt, L. F.: Neurotrophins: roles in neuronal development and function. Annu. Rev. Neurosci. 24:677, 2001. Johnson, J. E.: Neurotrophic factors. In Zigmund, M. J., Bloom, R. E., Landis, S. C., et al. (eds.): Fundamental Neuroscience. San Diego, Academic Press, p. 611, 1999. Kane, L. P., Shapiro, V. S., Stokoe, D., and Weiss, A.: Induction of NF-␬B by the Akt/PKB kinase. Curr. Biol. 9:601, 1999. Kaplan, D. R., and Miller, F. D.: Neurotrophin signal transduction in the nervous system. Curr. Opin. Neurobiol. 10:381, 2000. Kennedy, S. G., Kandel, E. S., Cross, T. K., and Hay, N.: Akt/protein kinase B inhibits cell death by preventing the release of cytochrome c from mitochondria. Mol. Cell. Biol. 19:5800, 1999. Kermer, P., Klöcker, N., Labes, M., and Bähr, M.: Insulin-like growth factor-I protects axotomized rat retinal ganglion cells from secondary death via PI3-K-dependent Akt phosphorylation and inhibition of caspase-3 in vivo. J. Neurosci. 20:722, 2000. Khursigara, G., Orlinick, J. R., and Chao, M. V.: Association of the p75 neurotrophin receptor with TRAF6. J. Biol. Chem. 274:2597, 1999. Korsching, S.: The neurotrophic factor concept: a reexamination. J. Neurosci. 13:2739, 1993. Le-Niculescu, H., Bonfoco, E., Kasuya, Y., et al.: Withdrawal of survival factors results in activation of the JNK pathway in neuronal cells leading to Fas ligand induction and cell death. Mol. Cell. Biol. 19:751, 1999.

Neurotrophic Factors in the Peripheral Nervous System 47. Lee, F. S., and Chao, M. V.: Activation of Trk neurotrophin receptors in the absence of neurotrophins. Proc. Natl. Acad. Sci. U. S. A. 98:3555, 2001. 48. Lee, F. S., Kim, A. H., Khursigara, G., and Chao, M. V.: The uniqueness of being a neurotrophin receptor. Curr. Opin. Neurobiol. 11:281, 2001. 49. Lee, R., Kermani, P., Teng, K. K., and Hempstead, B. L.: Regulation of cell survival by secreted proneurotrophins. Science 294:1945, 2001. 50. Lesser, S. S., and Lo, D. C.: CNTF II, I presume? Nat. Neurosci. 3:851, 2000. 51. Levi-Montalcini, R.: The nerve growth factor thirty-five years later. In Vitro Cell. Dev. Biol. 23:227, 1987. 52. Levi-Montalcini, R., Skaper, S. D., Dal Toso, R., et al.: Nerve growth factor: from neurotrophin to neurokine. Trends Neurosci. 19:514, 1996. 53. Matsuzaki, H., Tamatani, M., Mitsuda, N., et al.: Activation of Akt kinase inhibits apoptosis and changes in Bcl-2 and Bax expression induced by nitric oxide in primary hippocampal neurons. J. Neurochem. 73:2037, 1999. 54. McCormick, F.: Activators and effectors of ras p21 proteins. Curr. Opin. Genet. Dev. 4:71, 1994. 55. McDonald, E. S., and Windebank, A. J.: Cisplatin-induced apoptosis of DRG neurons involves Bax redistribution and cytochrome c release but not fas receptor signaling. Neurobiol. Dis. 9:220, 2002. 56. Miller, F. D., and Kaplan, D. R.: Neurotrophin signalling pathways regulating neuronal apoptosis. Cell. Mol. Life Sci. 58:1045, 2001. 57. Miller, F. D., and Kaplan, D. R.: On Trk for retrograde signaling. Neuron 32:767, 2001. 58. Mitchell, J. D., Wokke, J. H., and Borasio, G. D.: Recombinant human insulin-like growth factor I (rhIGF-I) for amyotrophic lateral sclerosis/motor neuron disease. Cochrane Database Syst. Rev. 3:CD002064, 2002. 59. Moore, M. W., Klein, R. D., Farinas, I., et al.: Renal and neuronal abnormalities in mice lacking GDNF. Nature 382:76, 1996. 60. Muller, G., Storz, P., Bourteele, S., et al.: Regulation of Raf-1 kinase by TNF via its second messenger ceramide and cross-talk with mitogenic signalling. EMBO J. 17:732, 1998. 61. Muller, M., Wilder, S., Bannasch, D., et al.: p53 activates the CD95 (APO-1/Fas) gene in response to DNA damage by anticancer drugs. J. Exp. Med. 188:2033, 1998. 62. Nakagawara, A.: Trk receptor tyrosine kinases: a bridge between cancer and neural development. Cancer Lett. 169:107, 2001. 63. Nawa, H., and Takei, N.: BDNF as an anterophin: a novel neurotrophic relationship between brain neurons. Trends Neurosci. 24:683, 2001. 64. Neet, K. E., and Campenot, R. B.: Receptor binding, internalization, and retrograde transport of neurotrophic factors. Cell. Mol. Life Sci. 58:1021, 2001. 65. Ochs, G., Penn, R. D., York, M., et al.: A Phase I/II trial of recombinant methionyl human brain derived neurotrophic factor administered by intrathecal infusion to patients with amyotrophic lateral sclerosis. Amyotroph. Lateral Scler. Other Motor Neuron Disord. 1:201, 2000.

385

66. Ozes, O. N., Mayo, L. D., Gustin, J. A., et al.: NF-␬B activation by tumour necrosis factor requires the Akt serine-threonine kinase. Nature 401:82, 1999. 67. Patapoutian, A., and Reichardt, L. F.: Trk receptors: mediators of neurotrophin action. Curr. Opin. Neurobiol. 11:272, 2001. 68. Pichel, J. G., Shen, L., Sheng, H. Z., et al.: Defects in enteric innervation and kidney development in mice lacking GDNF. Nature 382:73, 1996. 69. Rena, G., Guo, S., Cichy, S. C., et al.: Phosphorylation of the transcription factor forkhead family member FKHR by protein kinase B. J. Biol. Chem. 274:17179, 1999. 70. Riccio, A., Pierchala, B. A., Ciarallo, C. L., and Ginty, D. D.: An NGF-TrkA-mediated retrograde signal to transcription factor CREB in sympathetic neurons. Science 277:1097, 1997. 71. Romashkova, J. A., and Makarov, S. S.: NF-␬B is a target of AKT in anti-apoptotic PDGF signalling. Nature 401:86, 1999. 72. Sabatini, P., and McCormick, F.: Phosphoinositide 3-OH kinase (PI3K) and PKB/Akt delay the onset of p53-mediated, transcriptionally dependent apoptosis. J. Biol. Chem. 274:24263, 1999. 73. Sanchez, M. P., Silos-Santiago, I., Frisen, J., et al.: Renal agenesis and the absence of enteric neurons in mice lacking GDNF. Nature 382:70, 1996. 74. Sanchez-Perez, I., Martinez-Gomariz, M., Williams, D., et al.: CL100/MKP-1 modulates JNK activation and apoptosis in response to cisplatin. Oncogene 19:5142, 2000. 75. Sanchez-Perez, I., Murguia, J. R., and Perona, R.: Cisplatin induces a persistent activation of JNK that is related to cell death. Oncogene 16:533, 1998. 76. Schuchardt, A., D’Agati, V., Larsson-Blomberg, L., et al.: Defects in the kidney and enteric nervous system of mice lacking the tyrosine kinase receptor Ret. Nature 367:380, 1994. 77. Shimoke, K., and Chiba, H.: Nerve growth factor prevents 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-induced cell death via the Akt pathway by suppressing caspase-3-like activity using PC12 cells: relevance to therapeutical application for Parkinson’s disease. J. Neurosci. Res. 63:402, 2001. 78. Shooter, E. M.: Early days of the nerve growth factor proteins. Annu. Rev. Neurosci. 24:601, 2001. 79. Sofroniew, M. V., Howe, C. L., and Mobley, W. C.: Nerve growth factor signaling, neuroprotection, and neural repair. Annu. Rev. Neurosci. 24:1217, 2001. 80. Sutter, A., Riopelle, R. J., Harris-Warrick, R. M., and Shooter, E. M.: Nerve growth factor receptors: characterization of two distinct classes of binding sites on chick embryo sensory ganglia cells. J. Biol. Chem. 254:5972, 1979. 81. Tang, Y., Zhou, H., Chen, A., et al.: The Akt proto-oncogene links ras to pak and cell survival signals. J. Biol. Chem. 275:9106, 2000. 82. The BDNF Study Group (Phase III): A controlled trial of recombinant methionyl human BDNF in ALS. Neurology 52:1427, 1999. 83. Vetter, M. L., Martin-Zanca, D., Parada, L. F., et al.: Nerve growth factor rapidly stimulates tyrosine phosphorylation of phospholipase C-g1 by a kinase activity associated with the product of the trk protooncogene. Proc. Natl. Acad. Sci. U. S. A. 88:5650, 1991.

386

Neurobiology of the Peripheral Nervous System

84. Vogel, K. S., and Parada, L. F.: Sympathetic neuron survival and proliferation are prolonged by loss of p53 and neurofibromin. Mol. Cell. Neurosci. 11:19, 1998. 85. Wada, K., Okada, N., Yamamura, T., and Koizumi, S.: Nerve growth factor induces resistance of PC12 cells to nitric oxide cytotoxicity. Neurochem. Int. 29:461, 1996. 86. Wellmer, A., Misra, V. P., Sharief, M. K., et al.: A doubleblind placebo-controlled clinical trial of recombinant human brain-derived neurotrophic factor (rhBDNF) in diabetic polyneuropathy. J. Peripher. Nerv. Syst. 6:204, 2001. 87. Wiesmann, C., and de Vos, A. M.: Nerve growth factor: structure and function. Cell. Mol. Life Sci. 58:748, 2001. 88. Wiesmann, C., Ultsch, M. H., Bass, S. H., and de Vos, A. M.: Crystal structure of nerve growth factor in complex with the ligand-binding domain of the TrkA receptor. Nature 401:184, 1999. 89. Yamada, K., Mizuno, M., and Nabeshima, T.: Role for brainderived neurotrophic factor in learning and memory. Life Sci. 70:735, 2002.

90. Yamaguchi, A., Tamatani, M., Matsuzaki, H., et al.: Akt activation protects hippocampal neurons from apoptosis by inhibiting transcriptional activity of p53. J. Biol. Chem. 276:5256, 2001. 91. Yang, C. H., Murti, A., Pfeffer, S. R., et al.: Interferon ␣/␤ promotes cell survival by activating nuclear factor ␬B through phosphatidylinositol 3-kinase and Akt. J. Biol. Chem. 276:13756, 2001. 92. Yuen, E. C.: The role of neurotrophic factors in disorders of peripheral nerves and motor neurons. Phys. Med. Rehabil. Clin. N. Am. 12:293, 2001. 93. Zhou, H., Summers, S. A., Birnbaum, M. J., and Pittman, R. N.: Inhibition of Akt kinase by cell-permeable ceramide and its implications for ceramide-induced apoptosis. J. Biol. Chem. 273:16568, 1998. 94. Zundel, W., Swiersz, L. M., and Giaccia, A.: Caveolin 1-mediated regulation of receptor tyrosine kinase-associated phosphatidylinositol 3-kinase activity by ceramide. Mol. Cell. Biol. 20:1507, 2000.

18 Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease STEPHEN BRIMIJOIN

Properties and Mechanisms Overview Fast Anterograde Transport Fast Retrograde Transport Slow Transport Transport Motors Theoretical Implications of Transport Abnormalities Loading Stage Defects Anterograde Transport Defects Turnaround Defects

Defects of Retrograde Axonal Transport Potential Causes of Transport Defects Evidence for Transport Abnormalities in Human Neuropathy Axonal Transport in Experimental Neuropathies Overview of Toxicant-Induced Neuropathies Experimental Diabetes Acrylamide

PROPERTIES AND MECHANISMS Overview Neurons, because of their extended geometry and inability to synthesize proteins in axons, depend on special pathways for transport of macromolecules to their distal extremities. Waller’s original 1850 study of nerve transection233 identified the neuronal cell body as a “trophic center.” Fifty years later, Scott209 proposed that the trophic relationship between center and periphery involved actual nutrient flow the interruption of which was responsible for the degeneration Waller described. Though that seems self-evident today, Ranvier’s discussion of the mechanism of wallerian degeneration183 referred only to “vital energy” supplied by cell bodies, and Ramon y Cajal explicitly rejected the idea that soluble substances or enzymes move along axons.182 Axonal transport is now seen as a system of intracellular motility enabling nerve cells to deliver essential proteins and membrane components to peripheral sites, and to receive from these locations chemical signals and materials for disposal. Axonal transport is required to preserve cell viability, to elaborate and support axonal and dendritic arborization, to sustain impulse conduction, and to maintain the release

␤,␤⬘-Iminodipropionitrile Hexacarbons Zinc Pyridinethione p-Bromophenylacteylurea Genetically Based Animal Neuropathies Murine Dystrophy Related Dystrophies Other Genetically Based, Transgenic, and Cell Culture Models Conclusion

of transmitters at nerve endings. In addition, axonal transport provides a two-way line of chemical communication between neurons, on the one hand, and their synaptic targets and supporting cells, on the other. Anterograde transport delivers neurotrophic components to the nerve terminals, where their release affects muscles and other cells, and retrograde transport supplies the cell bodies with neurotrophic substances captured from the synaptic environment. The first part of this chapter outlines the basic phenomena of axonal transport, described in more detail in classic monographs by Grafstein and Forman87 and Sidney Ochs,168 and introduces recent advances in our understanding of the mechanisms involved. Axonal transport is a ubiquitous, ATP-driven process with multiple “components,” each moving at a distinctive macroscopic velocity. 239,240 The best defined components are (1) slow component a (SCa, averaging about 1 mm/day); (2) slow component b (SCb, averaging 2 to 10 mm/day); (3) fast anterograde transport (averaging 200 to 400 mm/day), and (4) fast retrograde transport (averaging 150 to 300 mm/day). These transport components each reflect the behavior of a particular subset of cargoes, such as the neurofilament proteins (slow moving) and synaptic 387

388

Neurobiology of the Peripheral Nervous System

vesicles (fast moving). It seems increasingly likely that all components are propelled by a common mechanism as Ochs suggested 30 years ago in his “unitary hypothesis,”167 which recent work largely confirms.49,211 From this perspective the differences in macroscopic transport rates might stem from different relative likelihoods of motion and stasis among cargoes whose instantaneous peak velocities are similar. Nonetheless, it is clear that transport involves at least two different families of transport motors, the kinesins and the dyneins, along with multiple adapter molecules that determine overall behavior. Before turning to motors and molecular mechanisms, however, we will consider the basic phenomena displayed by the different transport components, along with the identities of their characteristic cargoes.

Fast Anterograde Transport One useful way to study fast anterograde transport is to examine the outflow of proteins that have been radiolabeled by incorporation of a radioactive amino acid precursor. Typical data obtained with radiotracer methods are shown in Figure 18–1. A few hours after introducing 3H-leucine into a spinal ganglion, much radioactivity remains at the injection site but significant amounts are spread distally along the sciatic nerve. This actively transported label appears as a plateau rising distally to a crest, then descending steeply to background levels. The foot of the wave front progresses steadily along the nerve at a rate that precise

experiments determine to be near 400 mm/day at mammalian body temperature.163 Maximal rates are similar in myelinated and unmyelinated fibers, in motor and sensory nerves, in animals of sizes from mouse to goat, and at ages from perinatal to older adult.164,170 The rate varies with local temperature and exhibits a temperature coefficient (Q10) of 2.2.172 This high value means that transport slows markedly when nerves are cooled. Fast transport ceases altogether at a critical temperature near 11° C, probably because the cytoskeleton is disturbed, but it readily recovers after rewarming.37,45 Fast transport depends on local oxidative metabolism and is blocked within 30 minutes by incubation in a nitrogen atmosphere.140,165,169 Transport through any given region of nerve typically fails whenever oxidative metabolism is impaired to the point at which the level of highenergy phosphate (ATP ⫹ phosphocreatine) falls to about half of normal. This happens regardless of whether the impairment is due to pure anoxia, to ischemia, or to metabolic poisons such as dinitrophenol.192 Transport failure after metabolic insult is typically accompanied by a failure of impulse conduction. Both of these functional deficits are reversible if physiologic conditions are restored within 90 minutes. After anoxia or ischemia for up to a few hours, however, transport recovery can require days.140 Still longer periods of metabolic impairment lead to irreversible collapse and wallerian degeneration of the axon. Along with a supply of energy, microtubules are equally critical for fast axonal transport. Transport ceases when

FIGURE 18–1 Distribution of radioactivity in the dorsal root ganglia and sciatic nerves of five cats taken between 2 and 10 hours after injecting 3H-leucine into L7 ganglia (G). The activity present in consecutive 5-mm segments of roots, ganglia, and nerves is shown in counts per minute (CPM). Top left scale for 10-hour nerve, bottom left scale for 2-hour nerve (partial scales for 4-, 6-, and 8-hour nerves at the right). Arrows mark the wave fronts, defined as the points where the steeply sloping forward curves intersect the background. (Data from Ochs, S.: Fast transport of materials in mammalian nerve fibers. Science 176:252, 1972.)

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease

microtubules disaggregate, either because of treatment with antimitotic drugs such as colchicine, maytansine, or vinblastine,59,84,130 or because of exposure to temperatures below 11° C.45 The first studies using video-enhanced contrast–differential interference contrast microscopy (VECDIC) revealed fast-moving axonal particles following well-defined tracks that proved, in electron micrographs, to represent microtubules.3,27 Later it was possible to reconstruct similar motility in highly simplified cell-free systems consisting only of microtubules, membranous organelles, and a source of crude or purified kinesin. As a result of such studies, it became clear that microtubules represent the scaffolding or stationary elements along which molecular motors propel specific organelles in specific directions. Microtubules are composed of ␣- and ␤-tubulin dimers oriented end to end to form protofilaments, 13 of which combine to make up the tubule wall.70 Formed microtubules are dynamic structures from 1 to 100 ␮m in length, in which tubulin subunits are constantly being added to the “plus” ends, oriented away from the cell body, and being removed at the opposite or “minus” ends.51,107 Fast-transport cargoes were identified by years of painstaking observations with two-dimensional gel electrophoresis and autoradiography, video microscopy, and nerve ligation, among other methods (see Grafstein and Forman87). Early on, biochemical and electron microscopic data indicated that proteins, peptides, and neurotransmitters undergoing fast transport were associated with a membranous compartment that included synaptic vesicles, vesicle precursors, and a smooth reticular network of axonal membranes.60,68,69,149 The predominant entities whose motion can now be visualized directly by VEC-DIC microscopy are small membrane-bound organelles whose apparent dimensions are consistent with those of synaptic vesicles.2 When it comes to specific substances, a “stopflow” approach involving localized cooling and rewarming has been effective (Fig. 18–2). This method can directly demonstrate the fast transport of materials associated with neurotransmission, such as acetylcholinesterase (AChE),48 norepinephrine and its biosynthetic enzymes,35,36,46 and substance P.44 It is now widely agreed that all substances undergoing fast anterograde transport are associated with membrane-limited organelles, either as integral components or as passengers engaged in reversible associations.

Fast Retrograde Transport Retrograde transport closely resembles fast anterograde axonal transport. It returns many anterogradely transported proteins to the cell soma after their delivery to the periphery. Ligation of a nerve that is actively transporting radiolabeled protein results in an immediate accumulation of radioactivity on the “upstream” side and a delayed accumulation on the “downstream” side.15 The latter is driven by retrograde transport. Analysis of band patterns by gel elec-

389

FIGURE 18–2 Downflow of accumulated dopamine-␤-hydroxylase activity in vitro after rewarming of rabbit sciatic nerves locally cooled to 4° C for 90 minutes. Numbered arrows indicate duration of rewarming to 37° C, in hours. Mean values from four to eight nerves. The foot of the wave front progresses at the rate of 12.5 ⫾ 0.7 mm/h, or 300 mm/day. (Data from Brimijoin, S.: Stop-flow: a new technique for measuring axonal transport, and its application to the transport of dopamine-␤-hydroxylase. J. Neurobiol. 6:379, 1975.)

trophoresis and autoradiography shows that the proteins accumulating on both sides are similar.18,72,136,166 Many proteins must therefore reverse direction at some point, either in midpassage or after reaching the axon terminals. As with fast anterograde transport, most retrogradely transported proteins are probably associated with membrane-limited organelles. Thus, when transport is blocked by a local constriction or cooling, tubelike membranous structures accumulate above the blockade and large, multilamellar structures accumulate below.218,229 The latter are bigger, which probably explains why most of the motion resolved by VEC-DIC microscopy is directed toward the cell bodies. In any event, the speed of retrograde transport averages about 70% that of anterograde transport, whether determined optically or by stop-flow studies of marker proteins such as AChE and dopamine-␤-hydroxylase (DBH).48,83,108 Retrograde transport also carries exogenous components. Either along with endocytosis of nerve terminal membrane or following entry through special channels, materials in the immediate environment of the nerve terminals are taken up

390

Neurobiology of the Peripheral Nervous System

and transported retrogradely. Retrograde transport allows nerve growth factor (NGF) and other neurotrophins to play a major role in promoting the early development of sensory, sympathetic, and motor neurons.207 On the negative side, toxicity and neurologic disease can result when foreign proteins such as tetanus toxin74,181,206,208 and viruses such as herpes simplex and poliovirus are taken up by the nerve terminals and transported to the cell body.132,133 In experimental settings, the retrograde transport of dyes and markers is useful for the tracing of neuroanatomic pathways. Like anterograde transport, retrograde transport depends on local temperature, oxygen supply, and microtubular integrity. Because the motor for retrograde transport is not a kinesin, however, this component can be selectively inhibited, for example, by the agent 9-erythro2-(hydroxy-3-nonyl)adenine.82 Retrograde transport may play an important role in the pathologic reaction of chromatolysis (see Chapter 31), which follows an axonal interruption owing to trauma such as nerve crush, tear, or transection. Chromatolysis begins in the parent cell bodies after a lag time consistent with delivery of a signal by retrograde transport from the site of injury.131,225 The nature of the chromatolytic signal is still unclear. Possibilities are (1) a loss of some transported substance needed to sustain the normal state (e.g., a neurotrophin); (2) arrival of a “foreign” substance from the site of injury (e.g., a captured cytokine); or (3) excessive return of normally transported materials. Indications that retrograde transport does play a role in initiating chromatolysis come from experiments with the antimicrotubule drug colchicine. Local application of colchicine to the proximal stump of a transected nerve prevents at least one of the metabolic changes that precede full-blown chromatolysis, namely an increased uptake of 2-deoxyglucose by cell bodies.131 It is reasonable to conclude that the signal for this early reaction requires a transport process that depends on microtubular integrity.

Slow Transport Slow transport is classically considered to comprise two components, SCa and SCb. In the mid-1970s, Hoffman and Lasek extensively investigated SCa, the slower component moving at approximately 1 mm/day. Their studies demonstrated that SCa consists primarily of the three proteins making up the neurofilament triplet, plus the ␣- and ␤-tubulin subunits of microtubules.110 Impressed by the coherence of the radiolabeled wave associated with this phase of transport, these investigators proposed that SCa represented a slow and synchronous motion of fully formed cytoskeletal elements. This “structural hypothesis” generated years of controversy. Other scientists argued instead that the materials actually moving in SCa are soluble protein subunits of microtubules and neurofilaments, which must be locally incorporated into structural polymers after transport to distal sites.

Controversy continued even after photobleaching techniques were able to demonstrate the translocation of injected cytoskeletal proteins with fluorescent tags. Now newer methods with the power to reveal the behavior of individual microtubules and neurofilaments in live axons have gone a long way toward resolving the issue. The outcome, which seems obvious in hindsight, is at variance with both the “structural hypothesis” and the “local incorporation” proposal. Namely, it appears that both neurofilaments and microtubules move as formed polymeric structures rather than soluble subunits. However, such motion is anything but slow and synchronous; rather, it is rare, irregular, and surprisingly rapid when it does occur (for recent reviews, see Brown49 and Baas9). The most powerful and direct experiments in support of the new picture utilized neurofilaments tagged with green fluorescent protein, in sympathetic nerve axons with a sparse and discontinuous cytoskeletal array that permitted observation of individual polymers.187,234 Such work has revealed that neurofilaments move at peak speeds up to 3 ␮m/sec, well within the range for rapid axonal transport, although average speeds are 10- to 100-fold lower. However, periods of movement are interrupted by long pauses, so that the probability of motion at any one time is low (as little as 1%). This behavior is easily explained in terms of simple transitions between moving and stationary phases, as earlier envisioned by Ochs.167 The elucidation of microtubule transport has posed greater technical difficulties because of the dynamic nature of these organelles, which are constantly lengthening and shortening as they add and lose subunits. Nonetheless, rigorous studies with techniques capable of resolving individual fluorescencetagged polymers have now demonstrated that microtubules do move as formed elements, like neurofilaments.64,235 The current focus of research in this area is on identifying the molecular motors responsible for transport of cytoskeletal elements. In the case of microtubules, a likely candidate is cytoplasmic dynein, which is optimally designed to propel the anterograde movement of plus-ended structures.9 Less is known about SCb, a component of slow transport that includes a far more diverse collection of materials, many of which are soluble cytoplasmic proteins. In fact this component includes clathrin, actin, actin-binding proteins, and the enzymes of glycolysis. The unsolved question is whether such cargoes are propelled by direct interaction with specific molecular motors (and if so, which ones), or they move by piggybacking on faster moving carrier molecules.

Transport Motors As noted earlier, two major types of transport motor have been identified, the kinesins (a large and diverse family) and the dyneins (a smaller and less diverse family). The properties of the dyneins and kinesins are important

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease

because these proteins are potential targets of toxins and genetic defects that might lead to nerve disease. Indeed, certain inherited neuropathies, such as Charcot-MarieTooth disease type 2A, have already been shown to result from mutations in specific kinesins.246 For treatment of molecular motors in greater depth, the reader is referred to more specialized reviews by Goldstein and associates.4,86 An abbreviated outline follows. Kinesins share a similar architecture with a conserved motor domain located at the amino terminus of a heavy chain, which may be present as a homodimer or heterodimer. This basic unit is typically attached to a light chain (kinesin I) or to a regulator/adapter subunit. These constituents define the specificity of a kinesin for its cargo, either by associating directly with integral cargo proteins, or perhaps more often by recruiting other linker proteins that mediate the interaction (Fig. 18–3). Several of the partner proteins that are relevant for axonal transport have recently been identified. One is a transmembrane protein known as “Sunday driver” (Syd).23 The other is amyloid precursor protein, also a transmembrane protein.121 Both of these molecules interact directly with the light chain subunit of kinesin I through its tetratricopeptide repeat domain. Mammals express up to 100 different kinesin molecules, each with their own specificities. Among these motors, the monomeric kinesin “Unc104”176 appears to be of major importance for the fast axonal transport of synaptic vesicle precursors, even though its low “processivity” means that multiple motors per organelle are required.179 The diversity of kinesins is undoubtedly multiplied by a wide variety of adapter proteins, including many not yet

391

discovered. It therefore becomes possible to envision highly specific and individualized propulsion systems underlying the rapid anterograde transport of each of the various membrane elements found moving in axons. In the limit, these systems may be able to deliver particular proteins or functional complexes to precise destinations. A motor cannot move a vehicle unless it has something to push (or pull) against. All kinesins derive their motor functions from directional interactions with microtubules, the underlying scaffold for rapid axonal transport. Most kinesins interact specifically with the plus end of microtubules and thereby direct motion away from the cell body, generating anterograde transport. A few kinesins, however, are “minus-end directed.” The axonally expressed protein KIFC2 appears to be of this class and may therefore participate along with dynein in powering rapid retrograde transport of certain cargoes. Cytoplasmic dyneins are the main drivers of retrograde transport. The lower diversity among dyneins may be related to a lesser need for address-specific delivery, because many materials undergoing retrograde transport are destined only for lysosomal degradation. Be that as it may, dyneins are notable for their large size and complex subunit composition, which suggest multiple possibilities for regulatory control that have only begun to be investigated. Dyneins, like kinesins, are proteins that are themselves subject to rapid anterograde transport. An active delivery system is required to maintain their presence and function in distal regions of the axon. Therefore, a disorder that impairs anterograde transport will eventually impair retrograde transport as well.

FIGURE 18–3 Motors and accessory factors in a model for the mechanism of organelle movement. Motor enzymes (kinesin or cytoplasmic dynein) and accessory factors are both required for transport of the organelle. The complex formed by the motor, the accessory factor, and a motor receptor on the translocated organelle is termed the organelle translocation complex. It is likely that different subfamilies of kinesins and dyneins, and different accessory factors, are utilized to supportaddress specific transport of diverse organelles and protein complexes in the axon. (From Sheetz, M. R., Stener, E. R., and Schroer, T. A.: The mechanism and regulation of fast axonal transport. Trends Neurosci. 12:474, 1989, with permission.)

392

Neurobiology of the Peripheral Nervous System

THEORETICAL IMPLICATIONS OF TRANSPORT ABNORMALITIES Understanding of the physiologic role of axonal transport has generated persistent interest in the possibility that defects of motility underlie some pathologies of peripheral nerve. The role of interrupted transport in wallerian degeneration is well accepted. In addition, although outright failure of transport is not likely to be the root cause of any human neuropathy, subtle disturbances of axonal transport may contribute to a variety of disorders. A look at the steps in the life cycle of a transported protein makes clear why this is so. Each component of transport involves several stages. In rapid transport the following stages can be identified: (1) packaging of newly synthesized proteins into organelles and loading into the proximal axon; (2) distally directed, kinesin-driven motion (with occasional pauses and transient reversals); (3) arrival at the destination and temporary stasis; (4) turnaround, a directional reversal sometimes preceded by insertion into axolemmal membrane and later retrieval; (5) rapid dynein-driven retrograde transport (also with occasional pauses); and (6) lysosomal digestion in transit or after arrival in the cell body. Not every protein follows the entire sequence because metabolism or release from the nerve cell may intervene. Nevertheless, there are multiple points at which the transport system could be attacked by toxicants or disease, with varying outcomes.

Loading Stage Defects Many years ago Hammerschlag100 showed that rapid axonal transport is initiated by a loading phase that displays a distinctive pharmacology. As compared with motion along the axon, the initiation of axonal transport depends strongly on calcium and is sensitive to antimitotic drugs and monovalent ionophores.101–104 One can thus entertain the concept of selective diseases of transport loading. The pathophysiology of such a disorder would include a reduced axonal flow of synaptic vesicles, other formed organelles, and associated proteins, neuropeptides, and neurotransmitters. Many transported substances show distally declining concentration gradients, suggesting preferred distribution to proximal regions.143 A defect in the loading of such substances might especially hinder renewal and repair of surface membranes at the ends of long axons. Paradoxically, therefore, a loading defect in the cell body is a potential cause for a “dying-back” axonopathy that appears first in distal regions of nerve.54 However, another deleterious consequence could be that materials destined for transport would accumulate in the cell body, causing local swelling and structural dislocation. Defects in the Golgi apparatus need not affect proteins with alternative routes to the axon. Such proteins include many slowly transported materials, for example, soluble

enzymes and the components of microtubules, microfilaments, and neurofilaments. How these proteins reach the axon is still obscure, but there is evidence that their delivery can be selectively impaired by neurotoxins and neurologic disease. For example, as discussed later in more detail, the entry of neurofilament proteins into axons is specifically affected by ␤,␤⬘-iminodipropionitrile (IDPN), which causes ballooning of cell bodies and giant, filamentpacked swellings in the proximal axon.94

Anterograde Transport Defects One easily imagined sort of malfunction is retardation, or slowing. Though pathophysiologic studies have tended to look for that kind of malfunction, present evidence suggests that slowing of rapid anterograde transport is atypical. However, the slower components of transport appear more vulnerable to retardation, as will become apparent. Normal axons have surplus capacity for rapidly transported materials and can compensate for reduced velocity by increasing the concentration of material in transit.37,39,85 For this reason a small change in transport velocity is not in itself physiologically relevant. The meaningful variable is the distal flux of material (amount moved per unit time). To reduce this flux may require extreme slowing or a simultaneous loss of carrying capacity, for example, because microtubules are disorganized or kinesin is deficient. Another way to reduce delivery of materials to distant sites without greatly changing average velocity would be to lower the probability of onward transport from one region to the next. Causes could include premature reversal, accelerated degradation of cargo, or increased incorporation of transported material into fixed structures. Suppose that the probability of continued transport over a 1-cm distance were reduced only 3%, from 1.0 to 0.97. The cumulative probability of transport over a 1-m distance (typical for long sensory nerves in humans) would then drop to (0.97)100 ⫽ 0.05. Experimentally, such a transport defect would manifest itself as a steepened profile of transported radiolabel, with normal displacement of the wave front. Such effects have been seen with cytoskeletal proteins in the mutant diabetic mouse.244 A defect that reduced the anterograde transport of macromolecules by limiting supply would have important consequences. Failure to replenish neuropeptides, which must be synthesized in the cell body, would quickly alter the character of neurotransmission, and depletion of synaptic vesicles would cause persistent failure. Loss of membrane precursors and associated ion channels would compromise impulse conduction and the renewal of the axolemma. Reduced delivery of neurofilament proteins, which form skeletal elements that determine cross-sectional area,137 would produce axonal dwindling. Many such abnormalities are characteristic of advanced neuropathies.

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease

Turnaround Defects Bray and colleagues32 and Bisby15,16 showed decades ago that about half the protein rapidly transported to distal axons ultimately returns to nerve cell bodies by retrograde transport. This protein circulation involves directional reversal at the terminals, or wherever fast transport is interrupted (there is apparently no “turnaround” of slow components). An impaired turnaround would lead to swellings packed with the membranous organelles that normally undergo rapid anterograde transport. Terminal and preterminal axons would be likely sites for such lesions. There is ample evidence that turnaround defects contribute to distal axonopathy in one or more welldescribed disorders.

Defects of Retrograde Axonal Transport Rapid retrograde transport appears to reflect the movement of large particles, including multivesicular bodies and prelysosomal structures returning to the cell bodies for disposal, and pinocytotic vesicles carrying trophic signals from the periphery. The similar mechanisms for rapid retrograde and anterograde transport generate a shared sensitivity to antitubule drugs, the same temperature and ionic dependence, and identical energy requirements (for review of early work, see Kristensson132). However, the maximal velocity of retrograde transport is only about two-thirds that of anterograde transport, and different motors are involved, as indicated earlier. Furthermore, retrograde transport is differentially sensitive to certain chemical agents.81 These observations open the possibility of a selective block of retrograde transport by neurotoxins or disease. A selective block of retrograde transport would probably resemble impaired turnaround, except that the supply of extraneuronal trophic factors to nerve cell bodies would also be reduced. Moderate deficits of trophic factors would alter the patterns of gene expression and protein synthesis, and severe deficits could cause atrophy and even cell death.

Potential Causes of Transport Defects A disease could act in many ways to impair axonal transport. Potential mechanisms include local pressure, ischemia, inhibition of glycolysis or oxidative phosphorylation, disturbances in the amounts or organization of cytoskeletal proteins, interference with kinesin or dynein ATPases, and alterations of the ionic balance in axons. Each of these mechanisms deserves brief consideration. Local Pressure Local pressure is difficult to elevate without also preventing access to oxygen and nutrients, but sophisticated experiments suggest that high pressure by itself impedes axonal transport.98,99,188 Transport fails totally under pres-

393

sures of 200 mm Hg. If pressure is released within 2 hours, transport resumes within 1 to 3 days, but sustained pressure causes irreversible damage.188 In any event, pressures sufficient to cause outright collapse of the nerve, and mechanical obstruction of axons, are classic causes of compression neuropathies. A case of special interest in this connection is the so-called double-crush syndrome proposed by Upton and McComas.231 According to this concept, an initial compression lesion in the proximal course of a nerve sensitizes the patient to a subsequent compression at a more distal site, facilitating, for example, the development of an entrapment neuropathy. Animal experimentation provides some support for this mechanism,210 but its applicability to real-life conditions such as cervical radiculopathy remains doubtful.156 Ischemia Ischemia reduces the availability of oxygen and glucose together. Both are important in generating the ATP needed for rapid axonal transport.162,171,192 Actually, transport is surprisingly resistant to moderate hypoxia,159 though it is readily interrupted by complete anoxia even when highly localized.162 Nevertheless, ischemic impairment of transport has been demonstrated experimentally129 and is a potential contributor to the neuropathology of glaucoma,5 compression neuropathies (see Chapter 57), necrotizing angiopathic neuropathy,71 thromboangiitis obliterans, and advanced diabetes (see Chapter 85). Impaired Glycolysis or Oxidative Phosphorylation Any deficiency of energy metabolism, either from inherited enzyme defects or from exogenous toxins, tends to deplete cellular stores of ATP. Because fast transport depends on locally available ATP,125,169,171 it certainly would be impaired; other transport components would probably be affected as well. Especially interesting is the possibility that a toxic neuropathy might result from a localized ATP deficit with localized failure of transport. Cytoskeletal Disturbances Antimitotic drugs, including colchicine and the vinca alkaloids vinblastine and vincristine, induce the disaggregation of microtubules. Consistent with evidence for common mechanisms, these agents appear to block all components of transport.11,12,131 Such effects provided the initial evidence that microtubules are part of the machinery for transport. Now we recognize that microtubules serve to direct the propulsive forces provided by extrinsic motors and to determine transport capacity. The well-known peripheral neuropathy induced by vincristine is the limiting side effect in the use of vinca alkaloids to treat leukemia.237 It is highly likely that vincristine neuropathy results from a severe loss of transport capacity after microtubular disaggregation.26,136 Another agent that causes

Neurobiology of the Peripheral Nervous System

cytoskeletal disturbance is IDPN. As mentioned earlier, IDPN neuropathy involves marked retardation of the slow transport of neurofilament proteins, accompanied by giant, filament-rich swellings in the proximal axon. Disturbed Ionic Balance Fast transport depends, at least indirectly, upon appropriate axonal cations. The importance of Ca2⫹ for the loading stage of fast transport has already been mentioned. Studies with permeabilized cells and reconstituted systems of kinesin, microtubules, and isolated organelles show that Mg2⫹-ATP is sufficient to support transport per se.28,80 However, earlier indications that Ca2⫹ also affects transport173 are supported by some recent evidence for a regulatory role.33,151 Accordingly, long-term changes in serum electrolytes, particularly divalent cations, may be pathologically significant. Toxic Effects on Transport Motors Kinesins and dyneins are known to exhibit different sensitivities to inhibition by drugs.73,81,178 Therefore, differential toxicity for anterograde and retrograde transport motors becomes a reasonable topic of investigation. At another level, certain neurotoxic agents may eventually be found to impair individual motors, leading to disrupted transport of the corresponding cargoes. As yet, however, we know of no instances in which differential inhibition of particular transport motors contributes to the pathogenesis of neuropathy.

EVIDENCE FOR TRANSPORT ABNORMALITIES IN HUMAN NEUROPATHY If abnormalities of transport are probable causes of peripheral neuropathy, transport should be affected in actual diseases of human nerve. Although radiolabeling in vivo is not applicable to studies in humans, isolated axons continue transport in vitro. This transport can be studied by biochemical and optical methods. For biochemical experiments in human nerve, biopsy samples ligated at both ends are incubated in oxygenated physiologic salt solution. Later the distribution of rapidly transported enzyme or neurotransmitter is examined in consecutive short segments of the sample. So long as the total amounts of protein or transmitter are constant, an accumulation in the distal segment can be attributed to redistribution by anterograde axonal transport. The rate of accumulation is the transport flux, that is, the amount of material delivered per unit time. From the length of nerve that must contribute its content each hour to produce the accumulation, one calculates an overall average velocity of transport for a given marker. Interpretation of this measure is limited by uncertainties about the size of

the moving fraction.38 In favorable cases one can estimate the moving fraction by observing the clearance of marker from the middle of the sample or by identifying a break in the course of accumulation. Even when this is not possible, the average velocity of transport is useful because it is independent of the number of axons in the sample. A reduced average velocity of transport implies an actual slowing or an increased proportion of stationary material. Either would point to an abnormality of transport as opposed to a simple loss of fibers. Our initial studies of enzyme redistribution in human sural nerve utilized two enzyme markers, DBH and AChE, that catalyze the biosynthesis of norepinephrine and the hydrolysis of acetylcholine, respectively (Fig. 18–4). Observations of DBH and AChE redistribution indicated consistent reductions in transport in certain hereditary peripheral neuropathies. Transport abnormalities were especially pronounced when the sensory system was involved, alone or together with the motor system.40–42

Hours 5.0 4.0 3.5 3.0 2.5 DBH activity

394

2.0 1.5 1.0

Distance

0 Proximal 0

Distal 15

30

mm along nerve Ligature FIGURE 18–4 Redistribution of dopamine-␤-hydroxylase (DBH) activity in biopsy specimens of normal human sural nerve incubated in vitro. Each profile represents enzyme activity (measure of DBH protein) in consecutive 3-mm segments along an individual nerve ligated for the indicated number of hours. Accumulation proximal to the ligature occurs at a rate corresponding to transport at an average velocity of 2 mm/h. (From original data reported by Brimijoin, S., Capek, P., and Dyck P. J.: Axonal transport of dopamine-␤-hydroxylase by human sural nerves in vitro. Science 180:1295, 1973.)

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease

Dramatic deficits were seen in several cases of CharcotMarie-Tooth disease. That finding is consistent with the latest data. For example, it is now known that some varieties of this disorder reflect mutations in the light subunit of neurofilaments (NF-L),61 which disrupt neurofilament assembly and axonal transport.50 Of equal interest, another variety of the same disorder has been shown to involve mutations in the kinesin motor protein KIF1b␤.246 Another early study, by Behrens and co-workers,13 reported complete failure of AChE accumulation in a case of myotonic dystrophy, although enzyme content was more than 3 standard deviations above the mean from a series of normal nerves. In diabetic neuropathy, statistically significant reductions of enzyme content and of average transport velocity have been observed both with DBH and with AChE. We recognize that some of these findings might reflect end-stage axonal damage. However, the abnormalities of enzyme redistribution in neuropathic biopsies are also consistent with fundamental disturbances of transport. Optical methods are well suited to analysis of transport kinetics in neuropathy (Fig. 18–5). Nearly 20 years ago, DIC microscopy was applied to axonal transport in large myelinated sural nerve axons from cases of peroneal muscular

FIGURE 18–5 Video-enhanced contrast–differential interference contrast microscopy image of a single organelle undergoing rapid axonal transport in vitro. Digitized images of an axon were captured at 3-second intervals and electronically compiled while maintaining precise spatial registration. The transported organelle was digitally highlighted. Note that successive organelle displacements follow a consistent track but are not all equally long. Average velocity of this organelle was approximately 0.5 ␮m/s. In observations on nearly 300 particles from 19 axons, mean anterograde velocity was 2.2 ⫾ 0.6 ␮m/s and mean retrograde velocity was 1.4 ⫾ 0.4 ␮m/s. (D. Brat and S. Brimijoin, unpublished data, 1992.)

395

atrophy, Friedreich’s ataxia, and diabetic polyneuropathy.123 Visible particles were transported in some axons of all samples studied, and transport velocity varied considerably from case to case. Conclusions were limited because control biopsy specimens were not available. Another older optical study161 examined motor nerves in intercostal muscles of patients with amyotrophic lateral sclerosis (ALS). Estimated transport velocity was slowed by 30% in comparison with normal nerves from patients undergoing radical mastectomy. With the arrival of modern VEC-DIC microscopy and image processing, Breuer and colleagues were able to evaluate retrograde and anterograde transport in the motor branch of the median nerve in patients with ALS.33,34 Controls were various somatic nerves from patients undergoing limb amputation for removal of tumors, but at least one median nerve from a brain-dead donor was included. The design limitation imposed by the use of different types of nerve in patients and controls was mitigated by the absence of any known variation in the basic kinetics of transport in normal somatic nerves. Interestingly, the only significant kinetic abnormality in ALS nerves was a 57% increase in speed of anterograde transport. In contrast, the same analysis showed a 70% decrease in traffic density, or total flux expressed as the number of retrogradely moving organelles per second per square meter of axoplasm. This result is interesting because anterograde traffic density was only slightly reduced. The imbalance between anterograde and retrograde traffic suggests a defect in transport turnaround, and it may reflect an important biologic abnormality in ALS. Neurofilament dynamics are frankly disturbed in a number of neuropathic conditions. For example, axonal aggregations of neurofilamentous material are commonly observed in Alzheimer’s disease, in diabetic neuropathy, and in ALS.53,109,200,201 Such aggregations may very well reflect disruptions in the axonal transport of the underlying proteins or the neurofilaments themselves. Arguments for a primary transport disturbance are favored by observations in transgenic mouse models of ALS, which are discussed at the end of this chapter. Putting together the pathologic picture of neurofilament aggregation in disease states and the data from enzyme and optical studies, one concludes that many human peripheral nerve diseases probably do involve abnormalities of axonal transport.154 To determine whether the abnormalities cause the associated nerve damage or merely reflect structural damage from other causes, animal models of neuropathy must be investigated. The current status of research in this area is dealt with next. Meanwhile, it is worth emphasizing that even a secondary failure of transport could determine the course of neuropathy by inducing axonal atrophy, deficits in axolemmal renewal, impaired nerve conduction, failure of synaptic transmission, degeneration of nerve terminals, loss of trophic interactions, and reactive changes in the perikaryon.

396

Neurobiology of the Peripheral Nervous System

AXONAL TRANSPORT IN EXPERIMENTAL NEUROPATHIES Overview of Toxicant-Induced Neuropathies In 1969, Pleasure and associates reported marked reduction of slow axonal transport of radiolabeled protein in the central processes of sensory neurons of cats with acrylamide neuropathy.180 Since that original study, both fast and slow anterograde transport have been measured by various methods in many experimental neuropathies (Table 18–1). Minor reductions in fast-transport velocity have been encountered in the neuropathies induced by zinc pyridinethione (ZPT),147 acrylamide,26 and vincristine,26 although some studies have reported normal velocities in the same conditions.217 More substantial slowing of fast axonal transport, up to 36%, has been seen in hexacarbon neuropathy.152 Overall, however, one is struck by the extent to which fast anterograde transport seems to persist at full speed despite substantial axonal pathology. Few if any of the toxicant-induced neuropathies involve gross defects of fast anterograde transport. Some more pronounced effects on slow anterograde transport in several different toxic neuropathies are detailed subsequently.

Experimental Diabetes There has been disagreement about the status of rapid anterograde transport in diabetes (see Table 18–1). One early study reported slight reductions in the flux of AChE, a fast-transported enzyme.203 Others could neither confirm this observation nor detect changes in the velocity of fast-transported radiolabeled protein.10,17,116,238 A fair conclusion is that velocity and flux of fast anterograde transport are minimally affected in models of diabetes, whether induced by streptozotocin or alloxan. Experimental diabetes does seem to impair slow transport. This impairment is manifested as a reduction in the accumulation of enzymes such as choline acetyltransferase and as a retardation of the wave of radiolabeled protein in SCa.119,152,203,227,244 Certain of these defects are reversed or prevented by insulin, by myoinositol, or by aldose reductase inhibitors.147,226,228 Such beneficial effects suggest that the transport abnormalities arise specifically from the defect in carbohydrate metabolism. Conversely, inhibition of aldose reductase seems not to improve axonal transport of SCa, and there is conflicting information about the effect of insulin.135,227 Thus present information leaves room for doubt as to which transport effects represent streptozotocin neurotoxicity and which are truly metabolic. The retardation and reduced flux of neurofilament protein are nonetheless significant and might explain the modest but definite axonal dwindling in experimental diabetes.113,114 Turnaround and retrograde transport also appear to be impaired substantially in experimental diabetes. Jakobsen

and Sidenius found a significant reduction in the turnaround and retrograde transport of label associated with rapidly transported glycoconjugates in the sensory nerves of rats with streptozotocin diabetes.118 This reduction appeared within 1 day after onset of the diabetic state216 and persisted for at least 8 weeks.118 Significantly, turnaround was restored when blood sugar was normalized by insulin treatment.116 Hence the defect was probably not a primary neurotoxic effect of streptozotocin but a consequence of deranged carbohydrate metabolism. Unfortunately, fullblown neuropathy does not usually develop in animals with chemically induced diabetes, so the relevance of these observations to the neuropathology of human diabetes can still be questioned. We would argue nonetheless that diabetic neuropathy does involve deficits in transport reversal and retrograde transport. Additional work on this problem has emphasized the potential importance of deficits in the retrograde transport of neurotrophic factors from target organs. In 1981 our research team found that retrograde axonal transport of exogenous NGF was reduced in sciatic nerves of diabetic rats.116 This observation has been repeatedly confirmed.139,204 Such effects have been shown to persist in normoglycemic media in vitro.202 Hence they must reflect an abnormality of the nerve cell, not of its external environment. In vivo, deficits of NGF transport per se are compounded by a deficient expression of NGF in target tissues.78 Similar findings have been reported with regard to the related neurotrophin-3 (NT-3).79 Those observations take on added importance from the fact that NT-3 is a critical growth factor for many of the neurons that are most susceptible to diabetic neuropathy.75,77,126 It appears that deficits in neurotrophin transport are related in part to deficient uptake, owing to reduced expression of the relevant receptor proteins, among them tyrosine kinase A (TrkA) and p75.62,63,144 Other factors may also play a role, such as phosphatidylinositol 3-kinase/Akt signaling. This pathway, which is crucial for the internalization and retrograde transport of NGF in certain types of nerve,105,185 shows depressed activity in vagal neurons of diabetic rats.52

Acrylamide Until recently, opinion was divided as to whether acute administration of acrylamide was directly toxic to axonal transport and to what extent such toxicity could account for acrylamide neuropathy. A good deal of the evidence was negative. For example, in a careful study of proteins radiolabeled by 3H-leucine injection into dorsal root ganglia of the rat, Sidenius and Jakobsen found no significant change in the velocity of any transport component.217 Moreover, our own research group and that of Padilla found normal fluxes of transported vesicles in nerves and neurites observed by video microscopy, even when acrylamide was present in concentrations up to 1 mM for up to

Marked slowing (dorsal root only) No change — Large drop in vesicle flux 40% drop in vesicle flux at 1 h — — 25% slowing — — 40% less ChAT accumulation — 20% slowing — Normal velocity — — Normal velocity No change

— — SCa totally blocked — No change in ChAT accumulation — — NF transport speeds then slows No change 30% slowing No change — — —

Rat sciatic (sensory) Rat sciatic (sensory) Rat sciatic (motor) Rabbit sciatic Rat sciatic (cholinergic) Rat sciatic (sensory) Chicken sciatic Chicken sciatic Cat spinal roots Cat sciatic (sensory) Cat sciatic (sensory) Cat vagus Rat sciatic (motor) Rat sciatic (motor, sensory)

Slow Transport Effects

Cat spinal roots Cat sciatic Chicken oculomotor Cultured rat DRG neurons Mouse sciatic Rat sciatic (sensory) Rat sciatic (motor) Rat sciatic (sensory) Rat sciatic (motor) Neuroblastoma culture Rat sciatic (cholinergic) Rat sciatic (sensory) Rat sciatic (sensory) Rat sciatic (mixed) Rat sciatic (sensory) Rat ileal mesenteric Rat vagus (sensory) Rabbit vagus (sensory) Chicken sciatic Rat sural

Nerve

Reduced retrograde transport — — 10% slowing 25% slowing Irregular slowing Reduced AChE flux, antero- & retrograde Minor slowing, major block of turnaround

20% acceleration

Normal velocity, irregular block of retrograde transport

Focal blockade in swellings 10%–35% slowing; partial multifocal blockade No change

Normal AChE & DBH accumulation, impaired turnaround Normal velocity, reduced retrograde transport of NGF Reduced retrograde transport of NGF Reduced retrograde transport of NGF and NT-3 Normal velocity — 20% decreased velocity

Delayed onset, normal velocity No slowing, no deposition at sites of focal demyelination Delayed onset, normal velocity, impaired turnaround Accelerated onset, normal velocity, impaired turnaround Decreased anterograde and retrograde vesicle flux 20% reduction of AChE accumulation Delayed onset, normal velocity, impaired turnaround

— 10% slowing Focal retention in distal axon

Fast Transport Effects

AChE ⫽ acetylcholinesterase; BPAU ⫽ p-bromophenylacetylurea; ChAT ⫽ choline acetyltransferase; DBH ⫽ dopamine-␤-hydroxylase; DFP ⫽ diisopropylfluorophosphate; DRG ⫽ dorsal root ganglia; IDPN ⫽ ␤,␤⬘-imidodipropionitrile; NF ⫽ neurofilament; NGF ⫽ nerve growth factor; NT-3 ⫽ neurotrophin-3; SCa ⫽ slow component a; STZ ⫽ streptozotocin; TOCP ⫽ triorthocresylphosphate; — ⫽ not reported.

Acrylamide Acrylamide24 Acrylamide56,220 Acrylamide106 Acrylamide222 Adriamycin214 Antigalactocerebroside7 BPAU115 BPAU157,158 Cremophor EL31 Diabetes (STZ)203 Diabetes (STZ)215 Diabetes (STZ)120 Diabetes (STZ)116 Diabetes (STZ)17 Diabetes (STZ)204 Diabetes (STZ)138 Diabetes (alloxan)10 Diphtheria toxin122 Ethylene oxide160 Hexacarbons 2,5-Hexanedione96 Methyl n-butyl ketone152 IDPN93 Metals Aluminum141 Lead174 Methylmercury232 Organophosphates DFP, Mipafox, others155 DFP97 TOCP180 TOCP26 Vincristine26 Vincristine90 Vitamin E deficiency219 Zinc pyridinethione131

180

Agent (Reference)

Table 18–1. Experimental Neuropathy and Axonal Transport

398

Neurobiology of the Peripheral Nervous System

48 hours.30,177 By contrast, in an experimental system similar to that of Sidenius and Jacobsen, Sickles observed a substantial depression in the amount of radiolabeled protein undergoing fast anterograde transport.212 Such divergent results prompted a debate over factors that could spuriously affect measurements of transport. Over the past 10 years, however, a body of data has accumulated to show that acrylamide does impair axonal transport, and to provide an explanation for previous divergent findings. Much of the new evidence on acrylamide neuropathy comes from research by Sickles and colleagues. As summarized in a recent review,213 this work has established several important points. First, the effects of single exposures to acrylamide are limited in duration and easily overlooked when measurements are made more than a few hours after dosing, as happened in many earlier studies. Second, the amount of transported cargo is affected more severely than the maximal velocity. Third, acrylamide is capable of interacting directly with neural kinesins to impair transport capacity. Fourth, in different model systems based on cultured neurons and neural explants, acute exposure to acrylamide does reduce the traffic of rapidly transported, microscopically detected particles. Among the few older findings without obvious explanation to date are the observations of Brat and Brimijoin,30 which were obtained with neuron-like neuroblastoma cells. We now suppose that this discrepancy arises from a special feature, as yet unrecognized, of this tumor-derived cell line. When it comes to the severe clinical neuropathy that develops after repeated exposure to acrylamide, debate has focused not on the existence of transport abnormality but on its relationship to the neuropathology. Is it a cause or an epiphenomenon? It has long been recognized that at least one aspect of transport is severely disturbed in acrylamide neuropathy, as in experimental diabetes: turnaround and fast retrograde transport. Sahenk and Mendell197 obtained initial evidence for a turnaround defect, which was confirmed by Jakobsen and Sidenius.120 Droz and co-workers used elegant autoradiography in the ciliary ganglion of the chicken to clarify the structural basis of the defect. Application of this technique after incorporation of radioactive amino acid in cell bodies made it possible to show that acrylamide neuropathy was associated with abnormal retention of labeled material in the preterminal portions of axons.220 Thus the turnaround defect did not involve loss of proteins by metabolism or by release from the nerve terminal. Furthermore, electron microscope autoradiography demonstrated that the retained label coincided with accumulations of smooth endoplasmic reticulum.56 Because structurally similar membranes are chief among the elements undergoing rapid axonal transport, it is logical to infer that impaired turnaround directly causes the distal accumulation of membranous material in acrylamide neuropathy. Some investigators, however, have argued that the transport abnormalities in acrylamide neuropathy are nonetheless

epiphenomena.142 One reason for this contrary view is that large doses of acrylamide can cause behavioral signs of severe neuropathy, including foot splay and muscle weakness, without generalized axonopathy. Another is that the development of axonopathy is strongly influenced by the rate at which a given total dose is delivered. The pathology is greatest when administration is prolonged by repeated administration of small quantities, and least when a given dose is administered as a single large bolus.142 Such findings might suggest that axons are not the primary sites of acrylamide toxicity. This argument is not fully persuasive because a primary defect in transport might well manifest first in the nerve ending or the cell soma rather than the axon itself. Of course it is entirely possible that acrylamide neuropathy involves toxic actions on additional processes and proteins. Meanwhile, it remains reasonable to consider that a direct attack on the molecular motors for anterograde transport probably does represent an early and important step in the cascade of events leading to acrylamide neuropathy.

␤,␤⬘-Iminodipropionitrile IDPN neuropathy is still a prime example of a nerve disease in which impaired axonal transport is causative. The selective effect of IDPN on transport of neurofilament proteins in SCa93,128 provides a direct explanation for the giant filamentous swellings in the proximal axon and for the atrophy of the distal axon. The earliest effect of IDPN is a segregation of neurofilaments and microtubules. Local applications of IDPN cause neurofilaments to cluster in the subaxolemmal region, while an inner core of normally spaced microtubules continues to support fast axonal transport.91,95 These findings imply that IDPN does not attack transport per se but rather alters the properties of neurofilaments or their subunits so that they are no longer readily transported. Segregation of filaments in treated axons initially suggested an increased interaction among these structures (e.g., cross-linking) that does not directly involve the transport machinery. The distinction may be semantic because, no matter how it is initiated, transport blockade has a clear role in the expression of IDPN neurotoxicity. However, detailed biochemical studies of such phenomena are shedding new light on the general mechanism of filamentous neuropathies. In the case of IDPN, the underlying chemistry is still not clear, but there is reason to suspect that cyanoacetaldehyde or another reactive metabolic intermediate may covalently combine with lysine residues on the neurofilament subunits.199 The altered charge distribution on the protein surface would be a plausible basis for increased aggregation. Another plausible mechanism for altered intermolecular associations of neurofilaments in IDPN neuropathy was proposed by Eyer and associates.76 These workers found no covalent cross-linking but noticed increased autophosphorylation and in vitro gelation capacity of neurofilaments isolated from IDPN-treated rats. Similar effects

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease

occurred when IDPN was added directly to isolated neurofilaments but not when neurofilament-associated proteins were first removed by salt treatment. It was hypothesized that IDPN promotes interaction between neurofilament polymers through a preferential effect on associated proteins that form a minor but important part of the cytoskeletal network.

Hexacarbons A striking disturbance of fast anterograde axonal transport has been observed in the toxic neuropathy induced by methyl n-butyl ketone (MnBK) and its metabolite, 2,5hexanedione (2,5-HD). This disturbance bears a rough quantitative relation to the severity of neuropathy.152 Structural studies reveal that the neurotoxic hexacarbons cause multiple filament-packed axonal swellings, especially in the distal parts of long motor nerves.45 Mendell and coworkers, noting that transport was progressively impaired as increasingly long stretches of axon were examined, proposed that MnBK induced partial, multifocal blockade.152 Direct evidence for this concept was provided by Griffin and co-workers, who showed that rapidly transported proteins tend to accumulate in and proximal to the axonal swellings in 2,5-HD neuropathy.96 At present it can be argued that the cumulative effects of partial multifocal transport blockade, by impairing the delivery of fast-transported proteins to distal axons and terminals, partly account for dying-back degeneration. The pathogenic event, however, may be the appearance of filamentous swellings that represent deposits of the slowly transported neurofilament proteins. Two major explanations have been offered to account for focal failure of slow transport in 2,5-HD neuropathy. Sabri and Spencer have argued that local energy deficits are responsible.189–191,193,221 Their in vitro experiments show that MnBK and 2,5-HD are weak inhibitors of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and to a lesser extent, of phosphofructokinase (PFK).191 Both enzymes are essential for glycolysis and hence for generation of axonal ATP. A curious feature is the slow onset of the inhibition of GAPDH and PFK by neurotoxic hydrocarbons.193 This delay might explain why neuropathy develops only after weeks of daily dosing. Slow cumulative inhibition, with the target enzymes becoming critically compromised only as they reach the distal axon, also provides an explanation for the characteristic locus of pathology in hexacarbon neuropathy. However, new evidence suggests that hexacarbon axonopathy has more to do with the chemistry of neurofilaments than with axonal energy metabolism. It is now apparent that 2,5-HD reacts directly with free amino groups on proteins, especially neurofilaments, leading to pyrrole formation and, perhaps, extensive crosslinking.88,89 Such a mechanism could account for the

399

selective impairment of the transport of neurofilament components. It may also explain the location of the filamentous lesions. Thus analogues of the parent ␥-diketone can be shown to produce distal, proximal, or medial lesions, depending on their neurotoxic potency.6,92 Likewise, varying the dosage of toxicants that ordinarily induce proximal lesions (e.g., 3,4-dimethyl-2,5-hexanedione and IDPN) can lead to dramatic shifts in the location of axonal lesions.6,92 Direct evidence is still needed, but these results allow us to speculate that the locus of lesions in both cases is determined by the speed with which neurofilaments are chemically modified. Hexacarbon neuropathies exhibit a surprising acceleration of the transport of neurofilament proteins,29,150 which narrows the choice among putative pathogenic mechanisms. This effect is difficult to accommodate by hypothesizing energy deficits or toxicant-induced cross-linking, but it supports the idea of abnormal intermolecular interactions. Sayre and colleagues199 have proposed that the central mechanism of all filamentous neuropathies may be chemical alterations that disrupt the organization of neurofilaments, whether by cross-linking, pyrrole formation, or abnormal phosphorylation.

Zinc Pyridinethione Abnormalities in the turnaround of rapidly transported proteins were demonstrated first in the experimental neuropathy induced by ZPT. Using Bisby’s method of delayed nerve ligation after injecting radioactive amino acid into the spinal cord, Sahenk and Mendell196 showed a reduced return of labeled protein by retrograde transport in motor nerves of rats fed ZPT for 15 to 20 days. Further analysis indicated that ZPT caused both a delay in onset of retrograde transport and a reduction in the amount of returning protein.196 In parallel experiments, measurements of fast anterograde transport velocity failed to detect any consistent effect on this phase of transport, and the occasional reductions of velocity did not correlate with the symptomatic severity of neuropathy. Sahenk and Mendell concluded that ZPT selectively impaired the turnaround of protein from anterograde to retrograde transport. This impairment provided an explanation for the accumulation of branched tubular structures in distal axons.195 Later data implicated axonal proteases in transport turnaround and suggested that ZPT interferes with this crucial mechanism. Thus Schroer and colleagues showed that protease treatment of synaptic vesicles from squid axoplasm converts their in vitro transport from anterograde to retrograde.205 Martz and co-workers obtained evidence for proteolysis during turnaround in vivo, modifying the apparent mass of specific, vesicular proteins.145 Significant in the present context is the demonstration by Sahenk and Lasek that protease inhibitors profoundly impair transport turnaround when applied directly to axonal tips and also

400

Neurobiology of the Peripheral Nervous System

induce organellar accumulations that closely resemble the lesions in ZPT neuropathy.194 Whether ZPT is itself a protease inhibitor, and what its spectrum of activity might be, has never been established.

p-Bromophenylacetylurea One other disorder in which impaired turnaround has been intensively studied is an experimental neuropathy induced in rats by p-bromophenylacetylurea (BPAU). This distal axonopathy is convenient for analyzing the onset of peripheral nerve disease because it is produced by a single dose of compound and becomes symptomatic only after a latent period of several days.55,66 Work in the author’s laboratory showed that BPAU induces limb weakness accompanied by selective deficit of retrograde transport. When the sciatic nerve is ligated a few hours after injection of 3 H-leucine into dorsal root ganglia, accumulation of radiolabeled protein above the ligature is normal but accumulation below the ligature is less than half of control. There are several key features of this impairment in transport reversal. First, the defect occurs without any change in the velocity or total flux of labeled protein or enzymes moving by rapid anterograde axonal transport. Second, the abnormality appears in a matter of hours and precedes obvious structural damage. Third, in animals treated with BPAU, the decrease in retrograde accumulation correlates well with the clinical severity of the neuropathy115 and with the reduced amplitude of compound action potentials in the hind limb musculature.117 Because of the pronounced motor effects, further studies of axonal transport in BPAU neuropathy were performed in motor nerve after intraspinal injection of radioactive amino acid.157 Again the velocity of anterograde transport was unaffected while the recirculation of labeled protein by retrograde transport was reduced in a dose-dependent manner. The reduction was greatest during the later phases of accumulation. Therefore, the defect involves a net deficit in turnaround rather than a simple delay. Electron microscopy further strengthens the link between the transport defect and the characteristic tubulomembranous axonopathy of BPAU.19 Statistical analysis of autoradiograms shows that the lesions in BPAU-treated motor nerves are associated with rapidly transported, radiolabeled proteins. When examined at appropriate times after intraspinal radioisotope injection, most of the label is located over lesions and most of the lesions are heavily labeled.175 Thus the evidence in hand for BPAU neuropathy resembles that for acrylamide neuropathy. In both, the membranous axonal lesions appear to be local deposits of material trapped at the end stage of fast anterograde transport or at the onset of retrograde transport. The common pathogenic element may be local stasis of fast-transported organelles in the most distal reaches of the neuron.

Another interesting abnormality in BPAU neuropathy is a shortened lag between injection of protein precursor in ganglia and onset of transport in axons. In a structureactivity study of seven analogues, only the two neurotoxic agents, BPAU and its close relative p-chlorophenylacetylurea, were able to shorten transport lag.158 This information suggests that BPAU causes abnormal processing of proteins destined for transport. The possibility even arises that the neuropathy results from that effect. A working hypothesis for pathogenesis is as follows: (1) shortened lag time suggests accelerated transit of the Golgi apparatus; (2) accelerated transit could affect the posttranslational modification of proteins on the organellar surface; (3) organelles with abnormal surface chemistry may interact inappropriately with axonal translocation factors and be subject to local stasis; and (4) growing deposits of organelles could plug off distal axons and cause dying-back degeneration.

GENETICALLY BASED ANIMAL NEUROPATHIES A number of genetically mediated neuromuscular disorders of animals have attracted attention as potential models of the inherited neuropathies of humans. Chief among these are the muscular dystrophies occurring in mutant strains of mice, chickens, and hamsters, which resemble their human counterparts in many of their clinical and pathologic features.8,111,153 Axonal transport has been examined in all these disorders.

Murine Dystrophy Mice of the Jackson Laboratories ReJ/129 strain suffer from a hereditary autosomal recessive dystrophy with severe progressive limb weakness. There is minimal loss of axons but striking failure of myelination in the ventral and dorsal roots as well as in the proximal sciatic plexus.25 Early attempts to test the “neurogenic hypothesis” of muscular dystrophy148 indicated that the synthesis and axonal transport of proteins in dystrophic motor nerves may be abnormal. Bradley and Jaros, analyzing transported radioactivity by autoradiography, observed an increase in the amount of fast-transported label and a decrease in the amount of slow-transported label.24 Jablecki and Brimijoin112 noted a 40% decrease in the distal flux of choline acetyltransferase activity despite a 20% increase in the local content of enzyme activity. These findings are in reasonable accord because choline acetyltransferase is thought to move by slow axonal transport.71,198 However, in a later ligation study, dystrophic sciatic nerve also showed a deficient flux of the rapidly transported enzymes DBH and AChE.43

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease

Others reported conflicting data on the flow of labeled protein and lipid along the sciatic nerve after injection of radioactive precursors into dystrophic spinal cord or spinal ganglia. One group found normal velocities of transport but increased amounts of labeled material in dystrophic nerves.127,134,224 The largest difference was in the rapid component, which had three to five times the normal labeling, though the waves of radioactivity were indistinct and may have included blood-borne radioactivity. It was argued that murine dystrophy involves increased synthesis and rapid transport of selected proteins. Nearly opposite results were obtained by McLane and McClure, who used collection ligatures to distinguish transported label from local uptake.150 They noted a 40% decrease in the accumulation of rapidly transported protein labeled by injection of radioactive amino acid into the dystrophic spinal cord. When spinal ganglia were incubated with the same precursor in vitro, the results were similar and the impairment of transport progressively increased to a near-total block at 60 days of age. These conflicting observations have never been resolved. It should be worthwhile to determine whether the flux of rapidly transported proteins is actually increased or decreased. Reassessments of this problem should include controls for (1) amino acid pool sizes in the spinal cord, (2) blood-borne label, (3) alterations in the delay in initiation of transport, and (4) abnormalities of turnaround.

Related Dystrophies Axonal transport also has been examined in the dystrophic chicken and hamster. Although the neuromuscular abnormalities are superficially similar to those of the dystrophic mouse, there is no demyelination and transport seems unaffected. In dystrophic chickens, the anterograde and retrograde fluxes of AChE are normal, as is the pattern of this enzyme’s molecular forms.67 Likewise, the transport of 3H-leucine– or 3H-fucose–labeled protein in radial nerves does not differ in rate or amount from values observed in normal birds.223 These findings support the qualitative conclusions of an earlier study on retrograde labeling of motor neuron cell bodies in the spinal cord after injection of peroxidase into the latissimus dorsi muscles.65 In the dystrophic hamster, the anterograde flux of choline acetyltransferase and the velocity of fast transport of labeled protein along the sciatic nerve are apparently normal.21 Evidently, disturbances of axonal transport are not a universal feature of genetically mediated dystrophies.

Other Genetically Based, Transgenic, and Cell Culture Models There are fewer reports concerning axonal transport in other natural, genetically based neuromuscular diseases of

401

animals. Some early studies investigated axonal transport of radiolabeled protein in nerves of wobbler mice, which develop a motor neuron disease. Forelimb nerves in these mice, which are affected more severely than hind limb nerves, showed reduced fluxes of protein moving by fast and slow anterograde transport.14,47 Such results were in line with expectations for a model of a primary neuronal disorder. Studies on other natural animal models have investigated axonal transport in myelin-deficient mice such as the trembler mouse. These animals, in which cross-transplantation experiments localized the primary disorder to the Schwann cell,1 show severe demyelination in the sciatic nerve. An early study on trembler mice found that rapid axonal transport was at least grossly normal.20 This result was in line with expectations for a model in which the primary deficit was extrinsic to the neuron. The result was also consistent with observations of normal transport after focal demyelination by diphtheria toxin122 or antiserum to galactocerebroside.7 Surprisingly, however, more recent work has shown that a failure of myelination based solely on gene expression in glial cells can affect transport substantially. Thus a new study of trembler mice has found that the Schwann cell abnormality leads to a number of changes in the axons of peripheral nerve. These changes include a reduced density of axonal neurofilaments, a lowered stability of microtubules, and a slowing of axonal transport, among other phenomena.124 Effects in an opposite direction are seen in a model of central demyelination, the shiverer mouse, which is defective in myelin basic protein and expresses little compact myelin in the central nervous system. These animals show clear increases in the density of microtubules and the rate of slow axonal transport in affected brain regions.124 Such findings should be relevant to demyelinating peripheral neuropathies in humans, including those arising from genetic defects in glial cell protein expression and those arising from effects of toxicants and heavy metals on oligodendroglia. Studies of axonal transport in transgenic animals began relatively recently, but they already promise fresh insights into the mechanisms of neurologic disease in humans. Efforts to date have concentrated on degenerative disorders of central and alpha motor neurons. Mouse models of ALS have been particularly informative (for a recent review, see Rao and Nixon184). For example, it has been shown that transgenic mice expressing high levels of the human neurofilament light and heavy subunit proteins (NF-L and NF-H) develop an ALS-like pathology with filamentous aggregates in cell somata and proximal axons.58,243 These same mice exhibit marked retardation in the rates of transport of neurofilament protein and tubulin.57,146 A lesson may be that abnormal processing of cytoskeletal elements is central to ALS. Another likely key to ALS pathology, suggested by genetic studies of patients

402

Neurobiology of the Peripheral Nervous System

and their relations, is the enzyme superoxide dismutase 1 (SOD1), which is mutated in certain familial versions of ALS.186,242 Significantly, mice expressing mutant human SOD1 develop the classic pathology of motor neuron disease.22,230 As one might expect, the transport of neurofilament protein in these animals is disturbed.241,245 More surprisingly, reductions in fast anterograde transport and in axonal levels of kinesin precede the onset of pathology and motor weakness.236 A most recent innovation in the experimental approach to human neurologic disease is the use of Drosophila, an especially powerful model for studies of genetically based disorders. Insights into the neurodegeneration of Huntington’s disease have been obtained from transgenic Drosophila expressing a pathogenic version of the human huntingtin protein with an expanded polyglutamine motif. These flies were found to develop protein aggregates in distal regions of the nerve cell leading to disruption of axonal transport and a host of associated deficits, including loss of motor coordination and decreased lifespan.139

CONCLUSION The stages and major components of the system for axonal transport are indicated diagrammatically in Figure 18–6,

which also summarizes information about the sites where transport is vulnerable to attack by toxicants that induce neuropathy in humans or animals. One clear result of the past decade of studies is the increasing sophistication of our ideas about the relationship between axonal transport and peripheral nerve disease. Outright failure of transport would surely cause rapid degeneration and death of the nerve cell, but it is an unlikely cause of most forms of neuropathy. However, less dramatic abnormalities in the supply and disposition of transported materials occur early in certain nerve diseases and probably contribute to their evolution. Especially in toxic neuropathies, the pathogenic roles of transport abnormalities are beginning to be understood at the physiologic level. The most important pathology of rapid transport appears to involve defects operating in the most distal regions of the axon, during transport reversal or at stages just before or after. In the case of slow transport, there is an emerging consensus that chemical modifications of neurofilament subunits and deficits of filament transport contribute to the development of filamentous neuropathies. The challenge now is to develop fuller biochemical and molecular explanations of all these phenomena and link them to the rapidly expanding understanding of the diversity of transport motors and partner proteins that define the array of phenomena involved in axonal transport.

FIGURE 18–6 Stages in rapid axonal transport and suspected or proven sites of attack by agents capable of inducing peripheral neuropathy. Numbered arrows refer to the following stages: (1) loading of transported protein via the Golgi apparatus; (2) loading that bypasses the Golgi apparatus; (3) microtubule-based anterograde (kinesindriven) transport; (4) turnaround; (5) retrograde (dynein-driven) transport; and (6) lysosomal digestion. Gluc ⫽ glucose; Mitoch ⫽ mitochondrion; MT ⫽ microtubule; MVB ⫽ multivesicular body; NF ⫽ neurofilament; Pyr ⫽ pyruvate; SER ⫽ smooth endoplasmic reticulum; VES ⫽ synaptic vesicle.

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease

REFERENCES 1. Aguayo, A. J., Attiwell, M., and Trecarten, J.: Abnormal myelination in transplanted trembler mouse Schwann cells. Nature 265:73, 1977. 2. Allen, R. D., Travis, J. L., and Hayden, J. H.: Cytoplasmic transport: moving ultrastructural elements common to many cell types revealed by video-enhanced microscopy. Cold Spring Harb. Symp. Quant. Biol. 46:85, 1982. 3. Allen, R. D., Weiss, D. G., and Hayden, J. H.: Gliding movement of and bidirectional transport along single native microtubules from squid axoplasm: evidence for an active role of microtubules in cytoplasmic transport. J. Cell. Biol. 100:1736, 1985. 4. Almenar-Queralt, A., and Goldstein, L. S.: Linkers, packages and pathways: new concepts in axonal transport. Curr. Opin. Neurobiol. 11:550, 2001. 5. Anderson, D. R.: Axonal transport in the retina and optic nerve. In Glaser, J. S. (ed.): Neuro-Ophthalmology. St. Louis, C. V. Mosby, p. 140, 1977. 6. Anthony, D. C., Boekelheide, K., and Graham, D. G.: The effect of 3,4-dimethyl substitution on the neurotoxicity of 2,5-hexanedione. I. Accelerated clinical neuropathy is accompanied by more proximal axonal swellings. Toxicol. Appl. Pharmacol. 71:362, 1983. 7. Armstrong, R., Toews, A. D., and Morell, P.: Rapid axonal transport in focally demyelinated sciatic nerve. J. Neurosci. 7:4044, 1987. 8. Asmundson, V. S., Kratzer, F. H., and Julian, L. M.: Inherited myopathy in the chicken. Ann. N. Y. Acad. Sci. 138:49, 1966. 9. Baas, P. E.: Microtubule transport in the axon. Int. Rev. Cytol. 212:41, 2002. 10. Bajada, S., Sharma, A. K., and Thomas, P. K.: Axoplasmic transport in vagal afferent fibers in normal and alloxandiabetic rabbits. J. Neurol. Sci. 47:365, 1980. 11. Banks, P., Mayor, D., and Tomlinson, D.: Further evidence for the involvement of microtubules in the intra-axonal movement of noradrenaline storage vesicles. J. Physiol. (Lond.) 219:755, 1971. 12. Banks, P., and Till, R.: A correlation between the effects of antimitotic drugs on microtubule assembly in vitro and the inhibition of axonal transport in noradrenergic neurons. J. Physiol. (Lond.) 252:283, 1975. 13. Behrens, M. I., Torrealba, G., Court J., et al.: Axonal transport dysfunction in dystrophia myotonica. Acta Neuropathol. (Berl.) 62:157, 1983. 14. Bird, M. T., Shuttleworth, J. R., Koestner, A., and Reinglass, J.: The wobbler mouse mutant: an animal model of hereditary motor system disease. Acta Neuropathol. (Berl.) 19:39, 1971. 15. Bisby, M. A.: Orthograde and retrograde axonal transport of labeled protein in motoneurons. Exp. Neurol. 50:628, 1976. 16. Bisby, M. A.: Retrograde axonal transport of endogenous protein: differences between motor and sensory axons. J. Neurochem. 28:249, 1977. 17. Bisby, M. A.: Axonal transport of labelled protein and regeneration rate in nerves of streptozotocin-diabetic rats. Exp. Neurol. 69:74, 1980.

403

18. Bisby, M. A.: Retrograde axonal transport. Adv. Cell Neurobiol. 1:69, 1980. 19. Blakemore, W. F., and Cavanagh, J. B.: “Neuroaxonal dystrophy” occurring in the experimental “dying back” process in the rat. Brain 92:789, 1969. 20. Boegman, R. J., Aguayo, A. J., and Bray, G. M.: Axoplasmic transport in (trembler mouse) nerves with a widespread disorder of myelination. J. Neuropathol. Exp. Neurol. 36:590, 1977. 21. Boegman, R. J., and Wood, P. L.: Axonal transport in dystrophic hamsters. Can. J. Physiol. 59:202, 1981. 22. Borchelt, D. R., Wong, P. C., Becher, M. W., et al.: Axonal transport of mutant superoxide dismutase 1 and focal axonal abnormalities in the proximal axons of transgenic mice. Neurobiol. Dis. 5:27, 1998. 23. Bowen, D. M., Smith, C. B., White, P., and Davison, A. N.: Neurotransmitter-related enzymes and indices of hypoxia in senile dementia and other abiotrophies. Brain 99:459, 1976. 24. Bradley, W. G., and Jaros, E.: Axoplasmic flow in axonal neuropathies. II. Axoplasmic flow in mice with motor neuron: disease and muscular dystrophy. Brain 96:247, 1973. 25. Bradley, W. G., and Jenkinson, M.: Abnormalities of peripheral nerves in murine muscular dystrophy. J. Neurol. Sci. 18:227, 1973. 26. Bradley, W. G., and Williams, M. H.: Axoplasmic flow in axonal neuropathies. I. Axoplasmic flow in cats with toxic neuropathies. Brain 96:235, 1973. 27. Brady, S. T., and Lasek, R. J.: Fast axonal transport in extruded axoplasm from squid giant axon. Science 218:1129, 1982. 28. Brady, S. T., Lasek, R. J., and Allen, R. D.: Video microscopy of fast axonal transport in extruded axoplasm: A new model for study of molecular mechanisms. Cell Motil. 5:81, 1985. 29. Braendgaard, H., and Sidenius, P.: The retrograde fast component of axonal transport in motor and sensory nerves of the rat during administration of 2,5-hexanedione. Brain Res. 378:1, 1986. 30. Brat, D., and Brimijoin, S.: Acrylamide and glycidamide impair neurite outgrowth in differentiating N1E.115 neuroblastoma without disturbing rapid bidirectional transport of organelles observed by video microscopy. J. Neurochem. 60:2145, 1993. 31. Brat, D., Windebank, A., and Brimijoin, S.: Emulsifier for intravenous cyclosporin inhibits neurite outgrowth, causes deficits in rapid axonal transport and leads to structural abnormalities in differentiating N1E.115 neuroblastoma. J. Pharmacol. Exp. Ther. 261:803, 1992. 32. Bray, J. J., Kon, E. M., and Breckenridge, B. M.: Reversed polarity of rapid axonal transport in chicken motoneurons. Brain Res. 33:560, 1971. 33. Breuer, A. C., and Atkinson, M. B.: Fast axonal transport alterations in amyotrophic lateral sclerosis (ALS) and in parathyroid hormone (PTH)-treated axons. Cell Motil. Cytoskeleton 10:321, 1988. 34. Breuer, A. C., Lynn, M. B., and Atkinson, M. B., et al.: Fast axonal transport in amyotrophic lateral sclerosis: an intraaxonal organelle traffic analysis. Neurology 37:738, 1987. 35. Brimijoin, S.: Stop-flow: a new technique for measuring axonal transport, and its application to the transport of dopamine-␤-hydroxylase. J. Neurobiol. 6:379, 1975.

404

Neurobiology of the Peripheral Nervous System

36. Brimijoin, S.: A histofluorescence study of events accompanying accumulation and migration of norepinephrine within locally cooled nerves. J. Neurobiol. 8:251, 1977. 37. Brimijoin, S.: On the kinetics and maximal capacity of the system for rapid axonal transport in mammalian neurones. J. Physiol.(Lond.) 292:325, 1979. 38. Brimijoin, S.: Axonal transport in autonomic nerves: views on its kinetics. In Kalsner, S. (ed.): Trends in Autonomic Pharmacology. Baltimore, Urban and Schwarzenberg, p. 17, 1982. 39. Brimijoin, S.: Microtubules and the capacity of the system for rapid axonal transport. Fed. Proc., 41:2312, 1982. 40. Brimijoin, S., Capek, P., and Dyck, P. J.: Axonal transport of dopamine-beta-hydroxylase by human sural nerves in vitro. Science 180:1295, 1973. 41. Brimijoin, S., and Dyck, P. J.: Axonal transport of dopamine-beta-hydroxylase and acetylcholinesterase in human peripheral neuropathy. Exp. Neurol. 66:467, 1979. 42. Brimijoin, S., Dyck, P. J., Jakobsen, J., and Lambert, E. H.: Axonal transport in human nerve disease and in the experimental neuropathy induced by p-bromophenylacetylurea. In Weiss, D.G., and Gorio, A. (eds.): Axoplasmic Transport in Physiology and Pathology. Berlin, Springer-Verlag, p. 124, 1982. 43. Brimijoin, S., and Jablecki, C. K.: Reduced axonal transport of dopamine-beta-hydroxylase in dystrophic mice: evidence for abnormality of adrenergic nerve cells. Exp. Neurol. 53:454, 1976. 44. Brimijoin, S., Lundberg, J. M., Brodin, E., et al.: Axonal transport of substance P in the vagus and sciatic nerves of the guinea pig. Brain Res., 191:443, 1980. 45. Brimijoin, S., Olsen, J., and Rosenson, R.: Comparison of the temperature-dependence of rapid axonal transport and microtubules in nerves of the rabbit and bullfrog. J. Physiol. (Lond.), 287:303, 1979. 46. Brimijoin, S., and Wiermaa, M. J.: Direct comparison of the rapid axonal transport of norepinephrine and dopaminebeta-hydroxylase activity. J. Neurobiol. 8:239, 1977. 47. Brimijoin, S., and Wiermaa, M. J.: Rapid axonal transport of tyrosine hydroxylase in rabbit sciatic nerves. J. Physiol. (Lond.) 120:77, 1977. 48. Brimijoin, S., and Wiermaa, M. J.: Rapid orthograde and retrograde axonal transport of acetylcholinesterase as characterized by the stop-flow technique. J Physiol (Lond.) 285:129, 1978. 49. Brown, A.: Axonal transport of membranous and nonmembranous cargoes: a unified perspective. J. Cell Biol. 160:817, 2003. 50. Brownlees, J., Ackerley, S., Grierson, A. J., et al.: CharcotMarie-Tooth disease neurofilament mutations disrupt neurofilament assembly and axonal transport. Hum. Mol. Genet. 11:2837, 2002. 51. Burton, P. R., and Paige, J. L.: Polarity of axoplasmic microtubules in the olfactory nerve of the frog. Proc. Natl. Acad. Sci. U. S. A. 78:3269, 1981. 52. Cai, F., and Helke, C. J.: Abnormal PI3 kinase/Akt signal pathway in vagal afferent neurons and vagus nerve of streptozotocin-diabetic rats. Mol. Brain. Res. 110:234, 2003. 53. Carpenter, S.: Proximal axonal enlargement in motor neuron disease. Neurology, 18:841, 1968.

54. Cavanagh, J. B.: The significance of the “dying back” process in experimental and human neurological diseases. Int. Rev. Exp. Pathol. 3:219, 1964. 55. Cavanagh, J. B., Chen, F. C. K., Kyu, M., H., and Ridley, A.: The experimental neuropathy in rats caused by p-bromophenylacetylurea. J. Neurol. Neurosurg. Psychiatry 31:471, 1968. 56. Chretien, M., Patey, G., Souyri, F., and Droz, B.: “Acrylamide-induced” neuropathy and impairment of axonal transport of proteins. II. Abnormal accumulations of smooth endoplasmic reticulum at sites of focal retention of fast transported proteins: electron microscope radioautographic study. Brain Res. 205:15, 1981. 57. Collard, J. F., Cote, F., and Julien, J. P.: Defective axonal transport in a transgenic mouse model of amyotrophic lateral sclerosis. Nature 375:61, 1995. 58. Cote, F., Collard, J. F., and Julien, J. P.: Progressive neuronopathy in transgenic mice overexpressing the human neurofilament heavy gene: a mouse model of amyotropic lateral sclerosis. Cell 73:35, 1993. 59. Dahlström, A.: Effect of colchicine on transport of amine storage granules in sympathetic nerves of the rat. Eur. J. Pharmacol. 5:111, 1968. 60. Dahlström, A.: Axoplasmic transport (with particular respect to adrenergic neurons). Philos. Trans. R. Soc. Lond. [Biol.] 261:325, 1971. 61. De Jonghe, P., Mersivanova, I., Nelis, E., et al.: Further evidence that neurofilament light chain gene mutations can cause Charcot-Marie-Tooth disease type 2E. Ann. Neurol. 49:245, 2001. 62. Delcroix, J. D., Michael, G. J., Priestley, J. V., et al.: Effect of nerve growth factor treatment on p75NTR gene expression in lumbar dorsal root ganglia of streptozocin-induced diabetic rats. Diabetes 47:1779, 1998. 63. Delcroix, J. D., Tomlinson, D. R., and Fernyhough, P.: Diabetes and axotomy-induced deficits in retrograde axonal transport of nerve growth factor correlate with decreased levels of p75lntr protein in lumbar dorsal root ganglia. Mol. Brain Res. 51:82, 1997. 64. Dent, W. W., Calloway, J. L., Szebenyi, G., et al.: Reorganization and movement of microtubules in axonal growth cones and developing interstitial branches. J. Neurosci. 19:8894, 1999. 65. DeSantis, M., Hoekman, T., and Limwongse, V.: Retrograde transport of peroxidase in motor neurons innervating slow and fast twitch muscles: absence of influence between normal and dystrophic chickens. Brain Res. 119:454, 1977. 66. Diezel, P. B., and Quadbeck, G.: Nervenschudigung durch p-Bromophenylacetyl-Harnstoff. Naunyn Schmiedebergs Arch. Exp. Pathol. Pharmakol. 238:534, 1960. 67. DiGiamberardino, L., Couraud, J. Y., and Barnard, E. A.: Normal axonal transport of acetylcholinesterase forms in peripheral nerves of dystrophic chickens. Brain Res. 160:196, 1979. 68. Droz, B., Brunetti, M., and Giamberardino, L. D.: Selective distribution of axonally transported phospholipids to nerve endings and/or myelin: the case of ethanolamine glycerophospholipids. In Horrocks, L. A. (ed.): Phospholipids in the Nervous System. New York, Raven Press, p. 315, 1985.

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease 69. Droz, B., Giamberardino, L. D., and Koenig, H. L.: Contribution of axonal transport to the renewal of myelin phospholipids in peripheral nerves. I. Quantitative radioautographic study. Brain Res. 219:57, 1981. 70. Dustin, P., Microtubules. In Dustin, P. (ed.): Microtubules, 2nd ed. New York, Springer-Verlag, p. 1, 1984. 71. Dyck, P. J., Conn, D. L., and Okazaki, H.: Necrotizing angiopathic neuropathy: Three dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972. 72. Edstrom, A., and Hanson, M.: Retrograde axonal transport of proteins in vitro in frog sciatic nerves. Brain Res. 6:311, 1973. 73. Edström, A., Kanje, M., and Rusovan, A.: Orthograde and retrograde axonal transport in the regenerating frog sciatic nerve show different sensitivities to vanadate. Acta Physiol. Scand. 134:437, 1988. 74. Erdmann, G., Wiegand, H., and Wellhoner, H. H.: Intraaxonal and extraaxonal transport of 125I-tetanus toxin in early local tetanus. Naunyn Schmiedebergs Arch. Pharmacol. 290:357, 1975. 75. Ernfors, P., Lee, K. F., Kucera, J., and Jaenisch, R.: Lack of neurotrophin-3 leads to deficiencies in the peripheral nervous system and loss of limb proprioceptive afferents. Cell 77:503, 1994. 76. Eyer, J., McLean, W. G., and Leterrier, J.-F.: Effect of a single dose of ␤,␤⬘-iminodipropionitrile in vivo on the properties of neurofilaments in vitro: comparison with the effect of iminodipropionitrile added directly to neurofilaments in vitro. J. Neurochem. 52:1759, 1989. 77. Fariñas, I., Jones, K. R., Backus, C., et al.: Severe sensory and sympathetic deficits in mice lacking neurotrophin-3. Nature 369:658, 1994. 78. Fernyhough, P., Diemel, L. T., Brewster, W. J., and Tomlinson, D. R.: Deficits in sciatic nerve neuropeptide content coincide with a reduction in target tissue nerve growth factor mRNA in streptozotocin-diabetic rats: effects of insulin treatment. Neuroscience 62:337, 1994. 79. Fernyhough, P., Diemel, L. T., and Tomlinson, D. R.: Target tissue production and axonal transport of neurotrophin-3 are reduced in streptozotocin-diabetic rats. Diabetologia 41:300, 1998. 80. Forman, D. S.: A permeabilized cell model of saltatory organelle movement. J. Cell Biol. 91:414a, 1981. 81. Forman, D. S., Brown, K. J., and Livengood, D. R.: Fast axonal transport in permeabilized lobster giant axons is inhibited by vanadate. J. Neurosci. 3:1279, 1983. 82. Forman, D. S., Brown, K. J., and Promersberger, M. E.: Selective inhibition of retrograde axonal transport by erythro9[3-(2-hydroxy-nonyl)]adenine. Brain Res. 272:194, 1983. 83. Forman, D. S., Padjen, A. L., and Siggins, G. R.: Axonal transport of organelles visualized by light microscopy: cinemicrographic and computer analysis. Brain Res. 136:197, 1977. 84. Ghetti, B., and Ochs, S.: On the relation between microtubule density and axoplasmic transport in nerves treated with maytansine. In Canal, N., and Pozza, G. (eds.): Peripheral Neuropathies. Amsterdam, Elsevier, p. 177, 1978. 85. Goldberg, D. J., Schwartz, J. H., and Sherbany, A. A.: Kinetic properties of normal and perturbed axonal transport of serotonin in a single identified axon. J. Physiol. (Lond.) 281:559, 1978.

405

86. Goldstein, S. B., and Yang, Z. H.: Microtubule-based transport systems in neurons: the roles of kinesins and dyneins. Annu. Rev. Neurosci. 23:39, 2000. 87. Grafstein, B., and Forman, D.: Intracellular transport in neurons. Physiol. Rev. 60:1167, 1980. 88. Graham, D. G., Anthony, D. C., and Boekelheide K.: Studies of the molecular pathogenesis of hexane neuropathy II. Evidence that pyrrole derivatization of lysyl residues leads to protein crosslinking. Toxicol. Appl. Pharmacol. 64:629, 1982. 89. Graham, D. G., Szakal-Quin, G., Priest, J. W., and Anthony, D. C.: In vitro evidence that covalent crosslinking of neurofilaments occurs in gamma-diketone neuropathy. Proc. Natl. Acad. Sci. U. S. A. 81:4979, 1984. 90. Green, L. S., Donoso, J. A., Heller-Bettinge, I. E., and Samson, F. E.: Axonal transport disturbances in vincristineinduced peripheral neuropathy. Ann. Neurol. 1:255, 1977. 91. Griffin, J. W., Pahnestock, K. E., Price, D. L., and Hoffman, P. N.: Microtubule-neurofilament segregation produced by ␤,␤⬘-iminodipropionitrile: evidence for association of fast axonal transport with microtubules. J. Neurosci. 3:557, 1983. 92. Griffin, J. W., Gold, B. G., and Cork, L. C.: IDPN neuropathy in the cat: coexistence of proximal and distal axonal swellings. Neuropathol. Appl. Neurobiol. 8:351, 1982. 93. Griffin, J. W., Hoffman, P. N., and Clark, A. W.: Slow axonal transport of neurofilament proteins: impairment by ␤,␤⬘iminodipropionitrile administration. Science 202:633, 1978. 94. Griffin, J. W., and Price, D. L.: Proximal axonopathies induced by toxic chemicals. In Spencer, P. S., and Schaumberg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 161, 1980. 95. Griffin, J. W., Price, D. L., Hoffman, P. N., and Cork, L. C.: The axonal cytoskeleton: alterations of organization and axonal transport in models of neurofibrillary pathology [abstract]. J. Neuropathol. Exp. Neurol. 40:316, 1981. 96. Griffin, J. W., Price, D. L., and Spencer, P. S.: Fast axonal transport through giant axonal swellings in hexacarbon neuropathy. J. Neuropathol. Exp. Neurol. 36:603, 1977. 97. Gupta, R. P., Abdel-Rahman, A., Wilmarth, K. W., and Abou-Donia, M. B.: Alteration in neurofilament axonal transport in the sciatic nerve of the diisopropyl phosphorofluoridate (DFP)-treated hen. Biochem. Pharmacol. 53:1799, 1997. 98. Hahnenberger, R. W.: Effects of pressure on fast axoplasmic flow: an in vitro study in the vagus nerve of rabbits. Acta Physiol. Scand. 104:299, 1978. 99. Hahnenberger, R. W.: Effect of a pressure barrier on retrograde axoplasmic transport in vitro: a study in the motor neurons of the rabbit vagus. Acta Physiol. Scand. 108:133, 1980. 100. Hammerschlag, R.: The role of calcium in the initiation of fast axonal transport. Fed. Proc. 39:2809, 1980. 101. Hammerschlag, R., Bakhit, C., and Chiu, A. Y.: Role of calcium in the initiation of fast axonal transport of protein-effects of divalent cations. J. Neurobiol. 8:439, 1977. 102. Hammerschlag, R., Bolen, F. A., and Stone, G. C.: Inhibition of fast axonal transport by a sodium ionophore. Trans. Am. Soc. Neurochem. 11:143, 1980. 103. Hammerschlag, R., Chiu, A. Y., and Dravid, A. R.: Inhibition of fast axonal transport of [3H]-protein by cobalt ions. Brain Res. 114:353, 1976.

406

Neurobiology of the Peripheral Nervous System

104. Hammerschlag, R., Stone, G. C., and Bolen, F. A.: Evidence that all newly synthesized proteins destined for fast axonal transport pass through the Golgi apparatus. J. Cell Biol. 93:568, 1982. 105. Hansen, T., Andersen, C. B., Echwald, S. M., et al.: Identification of a common amino acid polymorphism in the p85alpha regulatory subunit of phosphatidylinositol 3-kinase: effects on glucose disappearance constant, glucose effectiveness, and the insulin sensitivity index. Diabetes 46:494, 1997. 106. Harris, C. H., Gulati, A. K., Friedman, M. A., and Sickles, D. W.: Toxic neurofilamentous axonopathies and fast anterograde axonal transport. Part V. Reduced bi-directional vesicle transport in cultured neurons by acrylamide and glycidamide. J. Toxicol. Environ. Health 42:343, 1994. 107. Heidemann, S. R., and McIntosh, J. R.: Visualization of the structural polarity of microtubules. Nature 286:517, 1980. 108. Helland, L.: Rapid retrograde transport of dopamine␤-hydroxylase as examined by the stop-flow technique. Brain Res. 102:217, 1976. 109. Hirano, A., Donnenfeld, H., Sasaki, S., and Nakano, I.: Fine structural observations of neurofilamentous changes in amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 43:461, 1984. 110. Hoffman, P. N., and Lasek, R. J.: The slow component of axonal transport: identification of major structural polypeptides of the axon and their generality among mammalian neurons. J. Cell Biol. 66:351, 1975. 111. Homburger, F., Nixon, C. W., Eppenberger, M., and Baker, J.: Hereditary myopathy in the Syrian hamster: studies on pathogenesis. Ann. N. Y. Acad. Sci. 138:14, 1966. 112. Jablecki, C. K., and Brimijoin, S.: Reduced axoplasmic transport of choline acetyltransferase activity in dystrophic mice. Nature 250:151, 1974. 113. Jakobsen, J.: Axonal dwindling in early experimental diabetes. Diabetologia 12:539, 1976. 114. Jakobsen, J.: Axonal dwindling in early experimental diabetes. II. A study of isolated nerve fibers. Diabetologia 12:547, 1976. 115. Jakobsen, J., and Brimijoin, S.: Axonal transport of enzymes and labeled proteins in experimental axonopathy induced by p-bromophenylacetylurea. Brain Res. 229:103, 1981. 116. Jakobsen, J., Brimijoin, S., Skau, K., and Wells, D.: Retrograde axonal transport of transmitter enzymes, fucose labeled proteins and nerve growth factor in streptozotocindiabetic rats. Diabetes 30:797, 1981. 117. Jakobsen, J., Lambert, E., Carlson, G., and Brimijoin, S.: Clinical and electrophysiological characteristics of the experimental neuropathy caused by p-bromophenylacetylurea. Exp. Neurol. 75:158, 1982. 118. Jakobsen, J., and Sidenius, P.: Decreased axonal flux of retrogradely transported glycoproteins in early experimental diabetes. J. Neurochem. 33:1055, 1979. 119. Jakobsen, J., and Sidenius, P.: Decreased axonal transport of structural proteins in streptozotocin diabetic rats. J. Clin. Invest. 66:292, 1980. 120. Jakobsen, J., and Sidenius, P.: Early and dose-dependent decrease in retrograde axonal transport in acrylamide-intoxicated rats. J. Neurochem. 40:447, 1983.

121. Kamal, A., Stokin, G. B., Yang, Z., et al.: Axonal transport of amyloid precursor protein is mediated by direct binding to the kinesin light chain subunit of kinesin-1. Neuron 28:449, 2000. 122. Kidman, A., Hanwell, M., and Cooper, N.: Failure of diphtheritic demyelination to block slow axonal transport in chicken sciatic nerve. J. Neurochem. 33:357, 1979. 123. Kirkpatrick, J. B., and Stern, L. Z.: Axoplasmic flow in human sural nerve. Arch. Neurol. 28:308, 1973. 124. Kirkpatrick, L. L., Witt, A. S., Payne, H. R., et al.: Changes in microtubule stability and density in myelin-deficient shiverer mouse CNS axons. J. Neurosci. 21:2288, 2001. 125. Kirpekar, S. M., Prat, J. C., and Wakade, A. R.: Metabolic requirements for the intraaxonal transport of noradrenaline in the cat hypogastric nerve. J. Physiol. (Lond.) 228:173, 1973. 126. Klein, R., Silos-Santiago, I., and Smeyne, R. J.: Disruption of the neurotophin-3 receptor gene TrkC eliminates 1a muscle afferents and results in abnormal movements. Nature 368:249, 1994. 127. Komiya, Y., Cooper, N. A., and Kidman, A. D.: The long-term effects of a single injection of ␤,␤⬘-iminodipropionitrile on slow axonal transport in the rat. J. Biochem. 100:1241, 1986. 128. Komiya, Y., and Austin, L.: Axoplasmic flow of protein in the sciatic nerve of normal and dystrophic mice. Exp. Neurol. 43:1, 1974. 129. Korthals, J. K., Korthals, M. A., and Wisniewski, H. M.: Peripheral nerve ischemia: part 2. Accumulation of organelles. Ann. Neurol. 4:487, 1978. 130. Kreutzberg, G. W.: Neuronal dynamics and axonal flow IV. Blockage of intra-axonal enzyme transport by colchicine. Proc. Natl. Acad. Sci. U. S. A. 62:722, 1969. 131. Kreutzberg, G. W., and Emmert, H.: Glucose utilization of motor nuclei during regeneration: a [14C] 2-deoxyglucose study. Exp. Neurol. 70:712, 1980. 132. Kristensson, K.: Retrograde transport of macromolecules in axons. Ann. Rev. Pharmacol. Toxicol. 18:97, 1978. 133. Kristensson, K., Lycke, E., and Sjostrand, J.: Spread of herpes simplex virus in peripheral nerves. Acta Neuropathol. (Berl.) 19:44, 1971. 134. Kuffer, A. D., Komiya, Y., and Austin, L.: Proteins of fast axoplasmic transport in the sciatic nerve of the dystrophic mouse. Exp. Neurol. 55:74, 1977. 135. Larsen, J. R., and Sidenius, P.: Slow axonal transport of structural polypeptides in rat, early changes in streptozocin diabetes, and effect of insulin treatment. J. Neurochem. 52:390, 1989. 136. Lasek, R. J.: Bidirectional transport of radioactively labelled axoplasmic components. Nature 216:1212, 1967. 137. Lasek, R. J., and Hoffman, P. N.: The neuronal cytoskeleton, axonal transport and axonal growth. In Goldman, R., Pollard, T., and Rosenbaum, J. (eds.): Cell Motility (Cold Spring Harbor Conferences on Cell Proliferation Series). Cold Spring Harbor, NY, Cold Spring Harbor Press, p. 1021, 1976. 138. Lee, P. G., Hohman, T. C., Cai, F., et al.: Streptozotocininduced diabetes causes metabolic changes and alterations in neurotrophin content and retrograde transport in the cervical vagus nerve. Exp. Neurol. 170:149, 2001. 139. Lee, W. C., Yoshihara, M., and Littleton, J. T.: Cytoplasmic aggregates trap polyglutamine-containing proteins and block axonal transport in a Drosophila model of Huntington’s disease. Proc. Natl. Acad. Sci. U. S. A. 101:3224, 2004.

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease 140. Leone, J., and Ochs, S.: Anoxic block and recovery of axoplasmic transport and electrical excitability of nerve. J. Neurobiol. 9:229, 1978. 141. Liwnicz, B. H., Kristensson, I., and Wisniewski, H. M.: Observations on axoplasmic transport in rabbits with aluminum-induced neurofibrillary tangles. Brain Res. 80:413, 1974. 142. LoPachin, R. M.: The role of fast axonal transport in acrylamide pathophysiology: mechanism of epiphenomenon? Neurotoxicology 23:253, 2002. 143. Lubinska, L.: Axoplasmic streaming in regenerating and in normal nerve fibres. Prog. Brain Res. 13:1, 1964. 144. Maeda, K., Fernyhough, P., and Tomlinson, D. R.: Regenerating sensory neurones of diabetic rats express reduced levels of mRNA for GAP-43, gamma-preprotachykinin and the nerve growth factor receptors, trkA and p75NGFR. Mol. Brain Res. 37:166, 1996. 145. Martz, D., Gamer, J., and Lasek, R. J.: Protein changes during anterograde-to-retrograde conversion of axonally transported vesicles. Brain Res. 476:199, 1989. 146. Marszalek, J. R., Williamson, T. L., Lee, M. K., et al.: Neurofilament subunit H modulates axonal diameter by affecting the rate of neurofilament transport. J. Cell Biol. 135:711, 1996. 147. Mayer, J. H., and Tomlinson, D. R.: Prevention of defects of axonal transport and nerve conduction velocity by oral administration of myo-inositol or an aldose reductase inhibitor in streptozotocin-diabetic rats. Diabetologia 25:433, 1983. 148. McComas, A. J., Sica, R. E. P., and Currie, S.: Muscular dystrophy: evidence for a neural factor. Nature 226:1263, 1970. 149. McEwen, B. S., Forman, D. S., and Grafstein, B.: Components of fast and slow axonal transport in the goldfish optic nerve. J. Neurobiol. 2:361, 1971. 150. McLane, J., and McClure, W. O.: Rapid axoplasmic transport in dystrophic mice. J. Neurochem. 29:863, 1977. 151. McNiven, M. A., and Ward, J. B.: Calcium regulation of pigment transport in vitro. J. Cell Biol. 106:111, 1988. 152. Mendell, J. R., Sahenk, Z., and Saida, K.: Alterations of fast axoplasmic transport in experimental methyl n-butyl ketone neuropathy. Brain Res. 133:107, 1977. 153. Michelson, A. M., Russell, E. S., and Pinckney, J. H.: Dystrophia muscularis: a hereditary primary myopathy in the house mouse. Proc. Natl. Acad. Sci. U. S. A. 41:1079, 1955. 154. Miller, C. C., Ackerley, S., Brownlees, J., et al.: Axonal transport of neurofilaments in normal and disease states. Cell. Mol. Life Sci. 59:323, 2002. 155. Moretto, A., Capodicasa, E., Peraica, M., and Lotti, M.: Age sensitivity to organophosphate-induced delayed polyneuropathy: Biochemical and toxicological studies in developing chicks. Biochem. Pharmacol. 41:1497, 1991. 156. Morgan, G., and Wilbourn, A. J.: Cervical radiculopathy and coexisting distal entrapment neuropathies: doublecrush syndromes? Neurology 50:78, 1998. 157. Nagata, H., and Brimijoin, S.: Axonal transport in the motor neurons of rats with neuropathy induced by p-bromophenylurea. Ann. Neurol. 19:458, 1986. 158. Nagata, H., and Brimijoin, S.: Neurotoxicity of halogenated pheylacetylureas is linked to abnormal onset of rapid axonal transport. Brain Res. 385:136, 1986.

407

159. Nagata, H., Brimijoin, S., Low, P., and Schmelzer, J. D.: Slow axonal transport in experimental hypoxia and in neuropathy induced by p-bromophenylacetylurea. Brain Res. 422:319, 1987. 160. Nagata, H., Ohkoshi, N., Kanazawa, J., et al.: Rapid axonal transport velocity is reduced in experimental ethylene oxide neuropathy. Mol. Chem. Neuropathol. 17:209, 1992. 161. Nordberg, A., Hellstrom-Lindahl, E., Lee, M., et al.: Chronic nicotine treatment reduces b-amyloidosis in the brain of a mouse model of Alzheimer’s disease (APPsw). J. Neurochem. 81:655, 2002. 162. Ochs, S.: Local supply of energy to the fast axoplasmic transport mechanism. Proc. Natl. Acad. Sci. U. S. A. 68:1279, 1971. 163. Ochs, S.: Rate of fast axoplasmic transport in mammalian nerve fibres. J. Physiol. (Lond.) 227:627, 1972. 164. Ochs, S.: Effect of maturation and aging on the rate of fast axoplasmic transport in mammalian nerve. Prog. Brain Res. 40:349, 1973. 165. Ochs, S.: Axoplasmic transport—energy metabolism and mechanism. In Hubbard, J. I. (ed.): The Vertebrate Peripheral Nervous System. New York, Plenum Press, p. 47, 1974. 166. Ochs, S.: Retention and redistribution of proteins in mammalian nerve fibres by axoplasmic transport. J. Physiol. (Lond.) 253:459, 1975. 167. Ochs, S.: A unitary concept of axoplasmic transport based on the transport filament hypothesis. In Bradley, W. G., Gardner-Medwin, D., and Walton, J. N., (eds.): Third International Congress on Muscle Diseases. Amsterdam, Excerpta Medica, p. 189, 1975. 168. Ochs, S.: Axoplasmic Transport and Its Relation to Other Nerve Functions. New York, Wiley Interscience, 1982. 169. Ochs, S., and Hollingsworth, D.: Dependence of fast axoplasmic transport in nerve on oxidative metabolism. J. Neurochem. 18:107, 1971. 170. Ochs, S., and Jersild, R. A. Jr.: Fast axoplasmic transport in nonmyelinated mammalian nerve fibers shown by electron microscopic radioautography. J. Neurobiol. 5:373, 1974. 171. Ochs, S., and Smith, C. B.: Fast axoplasmic transport in mammalian nerve in vitro after block of glycolysis with iodoacetic acid. J. Neurochem. 18:833, 1971. 172. Ochs, S., and Smith, C. B.: Low temperature slowing and cold-block of fast axoplasmic transport in mammalian nerves in vitro. J. Neurobiol. 6:85, 1975. 173. Ochs, S., Worth, R. M., and Chan, S. Y.: Calcium requirement for axoplasmic transport in mammalian nerve. Nature 270:748, 1977. 174. Ohnishi, A., Schilling, K., Brimijoin, S., et al.: Lead neuropathy. I. Morphometry, nerve conduction and choline acetyltransferase transport: new finding of endoneurial edema associated with segmental demyelination. J. Neuropathol. Exp. Neurol. 36:499, 1977. 175. Oka, N., and Brimijoin, S.: Tubulomembranous lesions in BPAU neuropathy reflect local stasis of fast axonal transport: evidence from electronmicroscope autoradiography. Mayo Clin. Proc. 67:341, 1992. 176. Otsuka, A. J., Jeyaprakash, A., Garcia-Anoveros, J., et al.: The C. elegans unc-104 gene encodes a putative kinesin heavy chain-like protein. Neuron 6:113, 1991. 177. Padilla, S., Atkinson, M. B., and Breuer, A. C.: Direct measurement of fast axonal organelle transport in the sciatic

408

178.

179.

180.

181.

182.

183. 184.

185.

186.

187.

188.

189.

190.

191.

192.

193.

194.

195.

Neurobiology of the Peripheral Nervous System nerve of rats treated with acrylamide. J. Toxicol. Environ. Health 39:429, 1993. Paschal, B. M., and Vallee, R. B.: Retrograde transport by the microtubule associated protein MAP IC. Nature 330:181, 1987. Pierce, D. W., Hom-Booher, N., Otsuka, A. J., and Vale, R. D.: Single-molecule behavior of monomeric and heteromeric kinesins. Biochemistry 38:5412, 1999. Pleasure, D. E., Mishler, K. C., and Engel, W. K.: Axonal transport of proteins in experimental neuropathies. Science 166:524, 1969. Price, D. L., Griffin, J., and Young, A.: Tetanus toxin: direct evidence for retrograde intraaxonal transport. Science 88:945, 1975. Ramon y Cajal, S.: Degeneration and Regeneration of the Nervous System (R. M. May, transl.). London, Oxford University Press, 1928. Ranvier, M. L. L.: Leçons sur I’Histologie du Systeme Nerveux. Paris, Librairie F. Savy, p. 68, 1878. Rao, M. V., and Nixon, R. A.: Defective neurofilament transport in mouse models of amyotrophic lateral sclerosis: a review. Neurochem. Res. 28:1041, 2003. Reynolds, A. J., Bartlett, S. E., and Hendry, I. A.: Signaling events regulating the retrograde axonal transport of 125I-beta nerve growth factor in vivo. Brain Res. 798:67, 1998. Rosen, D. R., Siddique, T., Patterson, D., et al.: Mutation in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 362:59, 1993. Roy, S., Coffee, P., Smith, G., et al.: Neurofilaments are transported rapidly but intermittently in axons: implications for slow axonal transport. J. Neurosci. 20:6849, 2000. Rydevik, B., McLean, W. G., Sjöstrand, J., and Lundborg, G.: Blockage of axonal transport induced by acute graded compression of the rabbit vagus nerve. J. Neurol. Neurosurg. Psychiatry 43:690, 1980. Sabri, M. I.: Further observations on in vitro and in vivo effects of 2,5-hexanedione on giyceraldehyde-3-phosphate dehydrogenase. Arch. Toxicol. 55:191, 1984. Sabri, M. I.: In vitro effect of n-hexane and its metabolites on selected enzymes in glycolysis, pentose phosphate pathway and citric acid cycle. Brain Res. 297:145, 1984. Sabri, M. I., Moore, C., and Spencer, P. S.: Studies on the biochemical basis of distal axonopathies. I. Inhibition of glyceraldehyde-3-phosphate dehydrogenase by neurotoxic hexacarbon compounds. J. Neurochem. 32:683, 1979. Sabri, M. I., and Ochs, S.: Relation of ATP and creatine phosphate to fast axoplasmic transport in mammalian nerve. J. Neurochem. 19:2821, 1972. Sabri, M. I., and Spencer, P. S.: Toxic distal axonopathy: biochemical studies and hypothetical mechanisms. In Spencer, P. S., and Schaumberg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 206, 1980. Sahenk, Z., and Lasek, R. J.: Inhibition of proteolysis blocks anterograde-to-retrograde conversion of axonally transported vesicles. Brain Res. 460:199, 1988. Sahenk, Z., and Mendell, J. R.: Ultrastructural study of zinc pyridinethione-induced peripheral neuropathy. J. Neuropathol. Exp. Neurol. 38:532, 1979.

196. Sahenk, Z., and Mendell, J. R.: Axoplasmic transport in zinc pyridinethione neuropathy: evidence for an abnormality in distal turn-around. Brain Res. 186:343, 1980. 197. Sahenk, Z., and Mendell, J. R.: Acrylamide and 2,5-hexanedione neuropathies: abnormal bidirectional transport rate in distal axons. Brain Res. 219:397, 1981. 198. Saunders, N. R., Dziegielewska, K., Haggendal, C. J., and Dahlstrom, A. B.: Slow accumulation of choline acetyltransferase in crushed sciatic nerves of the rat. J. Neurobiol. 4:95, 1973. 199. Sayre, L. M., Autilio-Gambetti, L., and Gambetti, P.-L.: Pathogenesis of experimental giant neurofilamentous axonopathies: a unified hypothesis based on chemical modification of neurofilaments. Brain Res. 357:69, 1985. 200. Schmidt, M. L., Martin, J. A., Lee, V. M., and Trojanowski, J. Q.: Convergence of Lewy bodies and neurofibrillary tangles in amygdala neurons of Alzheimer’s disease and Lewy body disorders. Acta Neuropathol (Berl.) 91:475, 1996. 201. Schmidt, R. E., Beaudet, L. N., Plurad, S. B., and Dorsey, D. A.: Axonal cytoskeletal pathology in aged and diabetic human sympathetic autonomic ganglia. Brain Res. 769:375, 1997. 202. Schmidt, R. E., Grabau, G. G., and Yip, H. K.: Retrograde axonal transport of [125I] nerve growth factor in ileal mesenteric nerves in vitro: effect of streptozotocin diabetes. Brain Res. 378:325, 1986. 203. Schmidt, R. E., Matschinsky, F. M., and Godfrey, D. A.: Fast and slow axoplasmic flow in sciatic nerve of diabetic rats. Diabetes 24:1081, 1975. 204. Schmidt, R. E., Plurad, S. G., and Saffitz, J. E.: Retrograde axonal transport of [l25I]-nerve growth factor in rat ileal mesenteric nerves: effect of streptozotocin diabetes. Diabetes 34:1230, 1985. 205. Schroer, T. A., Brady, S. T., and Kelly, R. B.: Fast axonal transport of foreign synaptic vesicles in squid axoplasm. J. Cell Biol. 101:568, 1985. 206. Schwab, M. E., Agid, Y., Glowinski, J., and Thoenen, H.: Retrograde axonal transport of 125I-tetanus toxin as a tool for tracing fiber connections in the central nervous system: connections of the rostral part of the rat neostriatum. Brain Res. 126:211, 1977. 207. Schwab, M. E., Heumann, R., and Thoenen, H.: Communication between target organs and nerve cells: retrograde axonal transport and site of action of nerve growth factor. Cold Spring Harb. Symp. Quant. Biol. 46:125, 1982. 208. Schwab, M. E., and Thoenen, H.: Selective binding uptake and retrograde transport of tetanus toxin by nerve terminals in the rat iris. J. Cell Biol. 77:3, 1978. 209. Scott, F. H.: On the relation of nerve cells to fatigue of their nerve fibers. J. Physiol. (Lond.) 34:145, 1906. 210. Seiler, W. A., Schelgel, R., Mackinnon, S., and Dellon, A. L.: Double crush syndrome: experimental model in the rat. Surg. Forum 34:596, 1983. 211. Shah, J. V., and Cleveland, D. W.: Slow axonal transport: fast motors in the slow lane. Curr. Opin. Cell Biol. 14:58, 2002. 212. Sickles, D. W.: Toxic neurofilamentous axonopathies and fast anterograde axonal transport. Part I. The effects of single doses of acrylamide on the rate and capacity of transport. Neurotoxicology 10:91, 1989.

Axonal Transport: Properties, Mechanisms, and Role in Nerve Disease 213. Sickles, D. W., Stone, J. D., and Friedman, M. A.: Fast axonal transport: a site of acrylamide neurotoxicity. Neurotoxicology 23:223, 2002. 214. Sidenius, P.: The effect of doxorubicin on slow and fast components of the axonal transport system in rats. Brain 109:885, 1986. 215. Sidenius, P., and Jakobsen, J.: Axonal transport in early experimental diabetes. Brain Res. 173:315, 1979. 216. Sidenius, P., and Jakobsen, J.: Retrograde axonal transport: a possible role in the development of neuropathy. Diabetologia 20:110, 1981. 217. Sidenius, P., and Jakobsen, J.: Anterograde axonal transport in rats during intoxication with acrylamide. J. Neurochem. 40:697, 1983. 218. Smith, R. S.: The short term accumulation of axonally transported organelles in the region of localized lesions of single myelinated axons. J. Neurocytol. 9:39, 1980. 219. Southam, E., Thomas, P. K., King, R. H., et al.: Experimental vitamin E deficiency in rats: morphological and functional evidence of abnormal axonal transport secondary to free radical damage. Brain 114:915, 1991. 220. Souyri, F., Chretien, M., and Droz, B.: “Acrylamideinduced” neuropathy and impairment of axonal transport of proteins. I. Multifocal retention of fast transported proteins at the periphery of axons as revealed by light microscope radioautography. Brain Res. 205:1, 1981. 221. Spencer, P. S., Sabri, M. I., Schaumburg, H. H., and Moore, C.: Does a defect in energy metabolism in the nerve fiber cause axonal degeneration in polyneuropathies? Ann. Neurol. 5:501, 1979. 222. Stone, J. D., Peterson, A. P., Eyer, J., et al.: Axonal neurofilaments are non-essential elements of toxicant-induced reductions in fast axonal transport. Part II. Video-enhanced differential interference microscopy in PNS axons. Toxicol. Appl. Pharmacol. 161:50, 1999. 223. Stromska, D., and Ochs, S.: Patterns of slow transport in the sensory nerves. J. Neurobiol. 12:441, 1981. 224. Tang, B. Y., Komiya, Y., and Austin, L.: Axoplasmic flow of phospholipids and cholesterol in the sciatic nerve of normal and dystrophic mice. Exp. Neurol. 43:13, 1974. 225. Tetzlaff, W., and Kreutzberg, G. W.: Ornithine decarboxylation in motoneurons during regeneration. Exp. Neurol. 89:679, 1985. 226. Tomlinson, D. R., and Mayer, J. H.: Reversal of deficits in axonal transport and nerve conduction velocity by treatment of streptozotocin-diabetic rats with myo-inositol. Exp. Neurol. 89:420, 1985. 227. Tomlinson, D. R., Sidenius, P., and Larsen, J. R.: Slow component-a of axonal transport, nerve myo-inositol, and aldose reductase inhibition in streptozocin-diabetic rats. Diabetes 35:398, 1985. 228. Tomlinson, D. R., Townsend, J., and Fretten, P.: Prevention of defective axonal transport in streptozocin-diabetic rats by treatment with “Statil” (ICI 128436), an aldose reductase inhibitor. Diabetes 34:970, 1985. 229. Tsukita, S., and Ishikawa, H.: The movement of membranous organelles in axons: electron microscopic identification of anterogradely and retrogradely transported organelles. J. Cell Biol. 84:513, 1980.

409

230. Tu, P.-H., Raju, P., Robinson, K. A., et al.: Transgenic mice carrying a human mutant superoxide dismutase transgene develop neuronal cytoskeletal pathology resembling human amyotrophic lateral sclerosis lesions. Proc. Natl. Acad. Sci. U. S. A. 93:3155, 1996. 231. Upton, A. R., and McComas, A. J.: The double crush in nerve entrapment syndromes. Lancet 2:359, 1973. 232. Wakabayashi, M., Araki, K., and Takahashi, Y.: Increased rate of fast axonal transport in methylmercury-induced neuropathy. Brain Res. 117:524, 1976. 233. Waller, A.: Experiments on the section of glossopharyngeal and hypoglossal nerves of the frog and observations of the alternatives produced thereby in the structure of their primitive fibers. Philos. Trans. R. Soc. Lond. 140:423, 1850. 234. Wang, L., and Brown, A.: Rapid intermittent movement of axonal neurofilaments observed by fluorescence photobleaching. Mol. Biol. Cell 12:3257, 2001. 235. Wang, L., and Brown, A.: Rapid movement of microtubules in axons. Curr. Biol. 12:1496, 2002. 236. Warita, H., Itoyama, Y., and Abe, K.: Selective impairment of fast anterograde axonal transport in the peripheral nerves of asymptomatic transgenic mice with a G93A mutant SOD1 gene. Brain Res. 819:120, 1999. 237. Weiss, H. D., Walker, M. D., and Wiernik, P. H.: Neurotoxicity of commonly used antineoplastic drugs. N Engl. J. Med. 291:127, 1974. 238. Whitely, S. J., Townsend, J., Tomlinson, D. R., and Brown, A. M.: Fast orthograde axonal transport in sciatic motoneurones and nerve temperature in streptozotocindiabetic rats. Diabetologia 28:8471, 1985. 239. Willard, M., Cowan, W. M., and Vagelos, P. R.: The polypeptide composition of intra-axonally transported proteins: evidence for four transport velocities. Proc. Natl. Acad. Sci. U. S. A., 71:2183, 1974. 240. Willard, M. B., and Hulebak, K. L.: The intra-axonal transport of polypeptide H: evidence for a fifth (very slow) group of transported proteins in the retinal ganglion cells of the rabbit. Brain Res. 36:289, 1977. 241. Williamson, T. L., and Cleveland, D. W.: Slowing of axonal transport is a very early event in the toxicity of ALS-linked SOD1 mutants to motor neurons. Nat. Neurosci. 2:50, 1999. 242. Wong, P. C., Pardo, C. A., Borchelt, D. R., et al.: An adverse property of a familial ALS-linked SOD1 mutation causes motor neuron disease characterized by vacuolar degeneration of mitochondria. Neuron 14:1104, 1995. 243. Xu, Z., Cork, L. C., Griffin, J. W., and Cleveland, D. W.: Increased expression of neurofilament subunit NF-L produces morphological alterations that resemble the pathology of human motor neuron disease. Cell 73:23, 1993. 244. Yitadello, M., Filliatreau, G., and Dupont, J. L.: Altered axonal transport of cytoskeletal proteins in the mutant diabetic mouse. J. Neurochem. 45:860, 1985. 245. Zhang, F., Eckman, C., Younkin, S., et al.: Increased susceptibility to ischemic brain damage in transgenic mice overexpressing the amyloid precursor protein. J. Neurosci. 17:7655, 1997. 246. Zhao, C., Takita, J., Tanaka, Y., et al.: Charcot-Marie-Tooth disease type 2A caused by mutation in a microtubule motor KIF1B␤. Cell 105:587, 2001.

19 Myelination UELI SUTER AND RUDOLF MARTINI

Morphology of Peripheral Nervous System Myelination Establishing the Myelin Internode Architecture of the Node of Ranvier

Molecular Biology of PNS Myelination Proteins Involved in Myelination Regulation of Myelin Genes Regulation of Myelin Protein Biosynthesis

MORPHOLOGY OF PERIPHERAL NERVOUS SYSTEM MYELINATION Establishing the Myelin Internode Myelination is a pivotal prerequisite for the rapid and saltatory conduction of action potentials, but also for the maintenance of the axonal structure.109 Two principle myelinating glial cells are found in the nervous system of higher vertebrates, the oligodendrocytes in the central nervous system (CNS) and the Schwann cells in the peripheral nervous system (PNS). Being predominantly derivatives of the neural crest,75,90,127 embryonic Schwann cells travel along axon bundles during development.16,17 These cells are most probably Schwann cell precursors, that is, prospective Schwann cells that still lack the ability to survive in the absence of axons.117 In the rat, Schwann cell precursors have been identified up to embryonic days 14 and 15 and do not yet express significant levels of the glial marker S100. Surprisingly, only scant attention has been paid so far to the morphologic characteristics and developmental changes of these cells, although they have been viewed as pivotal cellular components for axon survival.148 A role in guiding axons to their targets during development has been proposed earlier.120 This appears less probable because, during development, Schwann cell precursors follow pioneer axons in the embryo.16,17,154 In addition, Schwann cell precursors cannot survive when contact with axons is precluded,117 and axons find their targets in the absence of Schwann cell precursors.148

Role of Lipids in Myelination Role of the Extracellular Matrix, Its Cellular Receptors, and the Schwann Cell Cytoskeleton in Myelination

Schwann cell precursor cells are characteristically devoid of a basal lamina, intermingling and extending processes between the axon bundles (Fig. 19–1A). These cells are additionally found at the margin of the prospective nerve, facing the mesenchyme with their abaxonal surface (Fig. 19–1A).12,17,34,35,108,135,209 When development proceeds, Schwann cell precursors start to form a basal lamina, proliferate, and collectively ensheathe fasciculating axons, forming so-called Schwann cell families as initially described by Webster and colleagues.136,195 By then such glial cells most probably have reached the stage of immature Schwann cells, a term characterized by the ability to survive independently as a result of an autocrine survival mechanism.99,117 In the rat, all glial cells of the sciatic nerve are immature Schwann cells by embryonic day 17, whereas in the mouse, immature Schwann cells are the main population by embryonic day 15.43,117 The next step in development is related to an increase in caliber of some axons before becoming myelinated. They acquire a “peripheral” position within the axon bundles and achieve a so-called 1:1 ratio with immature Schwann cells (Fig. 19–1B). This stage of Schwann cell development is called “promyelin,” a morphologic starting point for myelin formation. In rodents, this sorting of many larger caliber axons into the promyelin stage occurs at around birth. However, myelin formation in the PNS is not a highly synchronized event. Axons with large calibers become myelinated earlier than those with smaller diameters, so that some promyelin stages can still be found in mice of approximately 3 weeks of age. 411

412

Neurobiology of the Peripheral Nervous System

B

A

FIGURE 19–1 Developing peripheral nerve of the mouse. A, Sciatic nerve of a mouse at embryonic day 13. Schwann cell precursors (SP) are found at the periphery of the nerve and protrude slender processes in between the embryonic axon bundles (A). A Schwann cell precursor in a more central position of the nerve is also visible (SP). Note the absence of Schwann cell basal lamina. M ⫽ mesenchymal cell. B, Femoral nerve of a 4-day-old mouse. Three axons of larger caliber display a thin sheath of compacted myelin. The upper fiber shows a redundant myelin profile (arrow). The axon demarcated by an asterisk is positioned at the periphery of a bundle of prospective nonmyelinating axons and is separated from them by a process of the immature Schwann cell. Double asterisks mark a similar situation at the bottom of the micrograph. A1 and A2 indicate promyelin stages. The Schwann cell process of A1 shows a more progressed stage of spiraling than A2. Bars: 1 ␮m.

The promyelin stage is followed by spiral formation of one of the Schwann cell processes engulfing the axon (Fig. 19–2; see also Fig. 19–1B). The apposition of Schwann cell membranes at the inner, axon-related side of the Schwann cell process is called the “inner mesaxon,” whereas the contact of Schwann cell membranes at the endoneurial side is termed the “outer mesaxon” (Fig. 19–2). These cell contacts are characterized by the formation of adherens (desmosomelike) junctions (Fig. 19–2). In an elegant study, Bunge and colleagues25 investigated the question of which end of the Schwann cell, the inner or outer one, turns around the axon during myelination. For this purpose, living rat Schwann cells co-cultured with dorsal root ganglion neurons were first investigated at the light microscopic level and the movements of the Schwann cell nuclei were recorded. After having monitored the behavior of the Schwann cells for up to 70 hours, the cultures were fixed and the axon–Schwann cell units in question were examined by electron microscopy (Fig. 19–3). These studies revealed that it is basically the inner, axon-related Schwann cell process that turns around the axon, and that the cell soma containing the nucleus is dragged behind at a much slower rate than the inner lip of the axon-related end of the Schwann cell process. This view is in line with older models stating that insertion of myelin components (e.g., radiolabeled lipids)136 occurs along the

entire extension of the developing spiral. Thus relatively rapid turning of the inner lip of a relatively slow-moving Schwann cell body and the simultaneous overall insertion of myelin membrane components appear to be characteristic events during myelin formation in the PNS. In vivo studies revealed that the myelinating process appears more complicated than just a spiral formation of the inner Schwann cell process around the axon. For instance, irregular forms with thin, redundant myelin loops are often observed (see Fig. 19–1B). In addition, the myelinating process in the region of the Schwann cell nucleus turns at a higher rate around the axon than the same process at the level of the Schwann cell edges (i.e., near the prospective node of Ranvier). A comprehensive overview of the morphologic events during myelin formation is summarized by Webster and colleagues.136,195 A unique subcellular feature of myelinating glial cells is the compaction of the turning membranes. In the case of rodent Schwann cells, this occurs when the myelinating process has turned around the axon a few times. Two processes occur simultaneously: (1) the narrowing of the spiraling surface membranes from approximately 12 to approximately 2 nm, and (2) the “squeezing out” of cytoplasm. The collapsed cytoplasmic sites of the Schwann cell membranes fuse and form a 3.5-nm-wide electron-dense

Myelination

413

A

FIGURE 19–2 Internodal segment of the mature myelin sheath. A, The majority of this internodal segment consists of compact myelin (M), whereas cytoplasmic aspects are sparse. Inner (arrow) and outer mesaxons (arrowhead) are indicated. A ⫽ axon. B, Highpower micrograph of A focusing on the inner mesaxon (arrow). Note adherens-like junction and layers of compact myelin consisting of major dense lines and intraperiod lines. Bars: 0.5 ␮m (A); 0.25 ␮m (B).

B

band called the “major dense line”; the membrane leaflets facing the extracellular space of the spiral form the “intraperiod line,” which is double-layered as a result of the 2-nm gap separating the extracellular leaflets (see Fig. 19–2B).136 Immature myelin sheaths are characterized by relatively expanded sites of still uncompacted myelin containing cytoplasm. During maturation, however, uncompacted myelin becomes restricted to distinct sites, such as the periaxonal collar (see Fig. 19–2), the outer Schwann cell loop, the paranodal loops, and the Schmidt-Lanterman incisures. The

latter are funnel-like clefts traversing the internodal myelin sheath and forming a helical cytoplasmic band that connects the periaxonal with the perinuclear (abaxonal) Schwann cell cytoplasm. In longitudinal sections of osmium tetroxide– fixed tissue, they can be identified on light microscopy as slim, bright lines obliquely traversing the myelin. In teased fiber preparations labeled for markers of noncompacted myelin (e.g., myelin-associated glycoprotein [MAG], connexin 32 [Cx32]), Schmidt-Lanterman incisures appear as funnel-like profiles.5,6,159 Electron microscopy reveals that

414

Neurobiology of the Peripheral Nervous System

FIGURE 19–3 Light microscopy (A) and electron micrographs (B and C) of a profile of a myelinated nerve fiber grown in vitro. A, Three internodes (1 through 3) are visible. These internodes have been investigated for 34 hours and nuclear movements have been recorded. B and C, The insets show the position of the nuclei of internodes 1 (B) and 3 (C) at distinct time points. The electron micrographs show the direction of the inner lip of the myelinating Schwann cells (arrows). Note that movement of Schwann cell nuclei (insets) and of the corresponding inner lip (electron micrograph) are always in the same direction. Bars: 10 ␮m (A), 0.5 ␮m (B and C). (From Bunge, R. P., Bunge, M. B., and Bates, M.: Movements of the Schwann cell nucleus implicate progression of the inner (axon-related) Schwann cell process during myelination. J. Cell Biol. 109:273, 1989, with copyright permission of The Rockefeller University Press.)

they consist of slender pockets of cytoplasm that form a helical funnel. Based on the observation that these cytoplasmic domains are Cx32 positive, it has been speculated that they form a rapid, radial cytoplasmic pathway for ions and small molecules between the adaxonal and abaxonal Schwann cell cytoplasm.160 Interestingly, recent immunohistochemical studies on teased fiber preparations revealed that Caspr1/paranodin and the voltage-gated K⫹ channels Kv1.1 and 1.2 demarcate the axonal domain underlying the inner loop of the Schmidt-Lanterman incisures.5,6,159 An interesting question is the regulation of myelin thickness. In particular, a striking feature is the positive correlation between axonal caliber and myelin thickness. This is

clearly reflected by the observation that the quotient between axon diameter and fiber diameter (taken at the outer aspect of the myelin sheath) is generally constant when axons of different diameters are compared.64 A similarly constant interrelationship exists between internodal length and axon diameter in the adult nerve. Quantitative studies in various species have revealed that a normal internode of a mature peripheral nerve is usually 100 times as long as the diameter of the corresponding axon.64 However, developmental studies in nerves with a robust growth in length during maturation in the absence of Schwann cell proliferation modified this simple view, as revealed by studies in phrenic nerves in rabbits and several

Myelination

peripheral nerves in humans.64,165 In young rabbits, myelin sheaths are relatively thin for axon diameter, followed by an increase in myelin thickness.64 An extreme case is found in various human peripheral nerves. Although radial growth of axons is completed at 5 years of age, the corresponding myelin continues to grow in thickness until the age of 17.165 In tibial nerves of rats, maximal growth in axonal diameter occurs between 3 weeks and 3 months. Here, the developmental change in relative myelin thickness is not uniform because myelin sheaths of fibers of different caliber behave in a slightly different way during development.60

mental in rapid restoration of the resting potential after an action potential has occurred. This chapter focuses on the structural characteristics of the node of Ranvier. The molecular components of the node of Ranvier and their functional roles are discussed below and in Chapter 24. The outer cytoplasmic aspect of the Schwann cells ends in the nodal region with microvilli-like protrusions that abut against the nodal plasmalemma (Fig. 19–4A). These protrusions contain actin filaments189 and typical microvilli-related cytoskeletal components (i.e., ezrin, radixin, and moesin, the ERM proteins).114,161 The ERM proteins may bind to the actin filaments linking the cytoskeleton to integral membrane components, possibly adhesion molecules. The functional role of the microvilli is not yet known, but it is assumed that they organize or stabilize distinct domains of the axolemma via their ERM-linked adhesion molecules in a trans-specific manner.11,114,161 The nodal microvilli and the nodal axolemma are surrounded by an extracellular matrix consisting of tenascin-C and the proteoglycan NG2.5,6,107,134 The nodal axolemma is characterized by a typical, electron-dense undercoating (see Fig. 19–4A). In this region, cytoskeletal components such as ankyrin G and spectrin are highly enriched and may organize the accumulation of the adhesion molecules Nr-CAM and neurofascin 186 and voltage-gated Na⫹ channels.11,159 It is thus plausible to assume that the accumulation of cytoskeletal elements is at

Architecture of the Node of Ranvier The node of Ranvier is highly organized both structurally and molecularly. Ontogenetically, it develops from the cell borders of neighboring Schwann cells that form the insulating myelin sheath around axons of larger caliber.14 From the physiologic point of view, the node of Ranvier is the component of the fiber responsible for the generation and propagation of action potentials. Two important prerequisites for this functional role are of note: (1) the accessibility of the axolemma for currents and (2) the presence of a high density of voltage-gated sodium channels that allow the generation of the action potential. In addition, voltagegated potassium channels and Na⫹,K⫹-ATPases are instru-

A

415

B

FIGURE 19–4 Nodal complex of the peripheral nerve of the adult mouse. A, Cross section of a node of Ranvier. Note nodal microvilli (arrows) and electron-dense undercoating of the axolemma (arrowheads). B, Longitudinal section of a paranodal region of a smaller caliber nerve fiber. Note “regular” organization of paranodal loops and that each loop abuts the axolemma. Septate-like junctions are indicated by arrows. Figure continued on following page

416

Neurobiology of the Peripheral Nervous System

C

D FIGURE 19–4 Continued C and D, Longitudinal section of a nodal region of a smaller (C) and a larger (D) caliber nerve fiber. In D, the organization of paranodal loops appears less regular than in C. Also, the juxtaparanodal aspects are conspicuous, as is typical for larger caliber axons, with interdigitations between inner Schwann cell loops and protruding axolemma (arrows). Bars: 0.5 ␮m (A), 0.25 ␮m (B), 1 ␮m (C and D).

least partially related to the electron-dense undercoating of the nodal axolemma. Other cytoskeletal components are neurotubules and neurofilaments. Characteristically, the latter components are of low phosphorylation stage as opposed to their internodal counterparts.5 The low phosphorylation level may be related to the diminution of axon diameter at the nodal region.39

Other striking compartments of the nodal region are the paranodes. In longitudinal sections, myelin lamellae end as cytoplasmic pockets that abut the paranodal axolemma (Fig. 19–4B and C). Typically, those myelin lamellae that are next to the axon end earlier than those closer to the node (Fig. 19–4B and C). Similar to the SchmidtLanterman incisures, they form a helical cytoplasmic band

417

Myelination

that connects the periaxonal with the perinuclear (abaxonal) Schwann cell cytoplasm. Forming reflexive Cx32positive gap junctions, both paranodal pockets and Schmidt-Lanterman incisures may function as rapid pathways between periaxonal and abaxonal cytoplasm.8 The paranodal pockets appear to be closely associated with the paranodal axolemma. With conventional aldehyde fixation, the axolemma appears “scalloped” as a result of a slight protrusion of the convex axon-associated surfaces of the paranodal pockets into the axon surface (see Fig. 19–4B). In addition, septate junctions form from the axonal surface and abut the paranodal Schwann cell membrane (Fig. 19–4B). These septate-like junctions are the place where Neurexin-Caspr-Paranodin is expressed and are involved in the molecular organization of the nodal environment.48,115,133 A detailed description of the functional role of paranodal molecules is found in Chapter 24. Other subcellular components of the paranodal loops are microtubules, which appear crosscut in longitudinal nerve sections as a result of their spiral orientation with respect to the longitudinal axis of the axon. Other compartments of interest are adherens (desmosome-like) junctions, which connect the paranodal pockets at their lateral sites. These junctions are characterized by the presence of the Ca2⫹-dependent adhesion molecule E-cadherin, which might be involved in their formation (see below).50 Depending on the size of the axon, there are in principle two different forms of paranodal region14,136,137 (see Fig. 19–4C and D). In smaller fibers, the cytoplasmic pockets approach the axon tangentially, and each pocket reaches the axolemma (Fig. 19–4B and C). The general appearance is “orderly.” In larger caliber fibers, the pockets approximate the nodal axolemma in a steep fashion and not all of them reach the axolemma (Fig. 19–4D). They are usually smaller and often appear much more electron dense than the pockets of the smaller axons. As a result of these characteristic appearances and the fact that myelin structures of larger fibers are more prone to show fixation artifacts, the paranodes of thick fibers appear less organized (Fig. 19–4D). Another issue of interest is the node-related compartment proximal to the paranode, the juxtaparanodal region. Some authors add this region to the paranodal compartment,14 but based on recent molecular findings, we prefer to treat this region as a separate compartment. It is particularly interesting from the physiologic point of view because this is the expression site of the axon- and Schwann cell–related, voltage-dependent K⫹ channels (see Chapter 24). From the morphologic point of view, the juxtaparanodal zone is particularly conspicuous in larger caliber fibers. As a result of a tapering of larger caliber axons, the myelin sheath appears fluted. In cross sections, the myelin sheath and the axon are cross-shaped or acquire the form of a trefoil. Here, the outer Schwann cell cytoplasm “fills” the grooves in the folded myelin and shows accumulations of mitochondria. The axon–Schwann cell

apposition often shows a complicated organization related to a profound interdigitation of inner Schwann cell loop and axolemma, called the axon–Schwann cell network (formerly paranodal network) (see Fig. 19–4D). In longitudinal sections, a striking asymmetry is detectable in that the axon–Schwann cell network is much more conspicuous at the distal side of the nodal region than at the proximal side. In addition, axonal lysosomes containing acid phosphatase are more abundant in the distal regions.68

MOLECULAR BIOLOGY OF PNS MYELINATION Proteins Involved in Myelination (Fig. 19–5 and Table 19–1) P0 P0 (myelin protein zero [MPZ]) is the major protein of the peripheral nerve myelin sheath, accounting for 50% to 60% of myelin proteins.93,94 It is a 30-kDa adhesion molecule belonging to the immunoglobulin (Ig) superfamily. Its single Ig domain bears a glycosylation site that can carry the HNK-1 carbohydrate epitope that is common to many cell adhesion and recognition molecules of the nervous system.156 In vitro studies suggested a pivotal function in myelin compaction as a result of its homophilic adhesive properties when transgenically expressed in various cell types.45,58,164 This view received support from the determination of the three-dimensional structure of the extracellular domain of P0 by x-ray crystallography.167 A simplistic model, however, does not, at present, explain all aspects of P0-mediated adhesion, because the intracellular domain of the molecule appears to be crucial for the adhesive properties of the extracellular domain,66,198 possibly by providing phosphorylation sites for protein kinase C.200 In addition, the intracellular domain of P0 contains predominantly basic residues, which have been suggested to interact with negatively charged phospholipids of the adjacent cytoplasmic parts of the Schwann cell membrane, leading to the formation of the major dense line.41,85,95 Expression of antisense-P0 messenger RNA (mRNA) blocks myelination in vitro at the promyelin stage,123 and studies in P0-deficient mice clearly demonstrate that P0 has a pivotal function during myelination, such as spiral formation, compaction, and myelin maintenance (see Chapter 24). Although P0 has been identified as a major component of peripheral myelin, it is expressed much earlier than initially anticipated. Its first expression has been found on neural crest cells of chicken and rat,15,32,75,92,127 but the functional roles at these early developmental stages are still elusive. Upregulation of P0 expression is temporally linked to the myelination process, and axon–Schwann cell units having acquired the promyelin stage show some weak but substantial immunoreactivity.110 The majority of P0 immunoreactivity, however, is found in compact

FIGURE 19–5 Schematic representation of various myelin-related proteins of the peripheral nerve. Note that distinct domains of myelin (e.g., compact vs. noncompact myelin) are associated with their specific set of molecules. For the specific functional roles of the molecules, see text.

419

Myelination

Table 19–1. Proteins of the Peripheral Nerve Myelin Sheath Molecule

Structure

Localization in Myelin

Function

Comments

P0, MPZ

Adhesion molecule with one Ig-like domain

Compact myelin (minor expression in neural crest cells)

Extra- and intracellular compaction, promotion of myelination

PMP22

Membrane protein of the EMP family, tetraspan topology questionable Cytoplasmic, membrane-associated protein

Compact myelin (minor expression in neural crest cells) Compact myelin (minor expression in neural crest cells) Noncompacted myelin

Cell cycle proliferation, apoptosis, spreading; regulation of myelination Supports P0 in intracellular compaction

Major PNS myelin protein, mutations lead to inherited myelin disorders Mutations lead to inherited myelin disorders

MBP

MAG

Adhesion molecule with 5 Ig-like domains

Cx32

Tunnel protein of gap junctions

Noncompacted myelin

E-cadherin

Ca2⫹-dependent adhesion molecule

Neurofascin 155

Adhesion molecule with 6 Ig-like domains

Adherens junctions of noncompacted myelin and of inner and outer mesaxon Adaxonal site of paranodal loops

TAG-1

Adhesion molecule with 6 Ig-like domains (GPI-linked) Tetraspan protelipid

MAL Periaxin

Cytoskeleton-associated protein with PDZ motif; L-periaxin, S-periaxin

Juxtaparanodal myelin

Compact myelin, nonmyelinating fibers Periaxonal cytoplasm during development (mainly S-form), abaxonal in mature fibers (mainly L-form)

myelin.110,188,205 Interestingly, P0 expression on nonmyelinating Schwann cells appears to be actively suppressed by the corresponding small-caliber axons.92 Peripheral Myelin Protein 22 Peripheral myelin protein 22 (kDa) (PMP22), the first identified culprit gene for inherited neuropathies of CharcotMarie-Tooth (CMT) type, is an integral membrane glycoprotein containing a single N-linked carbohydrate moiety. It is expressed mainly in compact myelin of peripheral nerves.177 It accounts for approximately 2% to 5% of myelin proteins. In contrast to P0, expression in other neural structures and even non-neural organs and tissues has been described.7,121,130,178 Initially, PMP22 was believed to be a tetraspan membrane molecule, but recent studies suggest that an alternative organization is conceivable.183 By biochemical approaches, PMP22 and P0 could be co-purified from peripheral nerve myelin, suggesting that they form complexes in myelin.46 In vitro experiments suggest a func-

Maintenance of myelin and axonal integrity Communication of non-compacted myelin domains Stabilization of adherens junctions

Glial partner for NCP-1

Lack of MBP leads to minor abnormalities in PNS In vitro experiments suggested involvement in myelin formation Mutations lead to inherited myelin disorders Downregulation in P0 mutants is associated with abnormal junctions Dysregulated in mice with abnormal NCP-1 expression

May mediate distribution of K⫹ channels Component of lipid rafts L-form interacts indirectly with laminin of SC basal lamina via DRP2–␣␤–dystroglycan complex

Mutations lead to inherited myelin disorders

tion of PMP22 in cell growth, differentiation, and apoptosis,22,210 but mice lacking PMP22 or overexpressing the gene show a normal number of Schwann cells perinatally.153 Although it is predominantly a component of compact myelin, PMP22 expression in rat embryos starts long before the onset of myelination, in neural crest–derived progenitor cells at embryonic day 14. By contrast, such an early expression has not been found in the mouse, so the role of the early expression remains elusive.127 PMP22 appears to play pivotal roles during myelination, because in the absence of the gene, myelination is severely affected.2 At present the exact role of PMP22 during myelination is difficult to derive from the pathologic changes in nerves of knockout mutants (see Chapter 24). Myelin Basic Protein Myelin basic proteins (MBPs) are products from a large gene that encodes two families of proteins, the Golli proteins and the “classic” (i.e., myelin-related) MBP isoforms.27 The

420

Neurobiology of the Peripheral Nervous System

latter appear to be confined to differentiated myelinforming cells, whereas Golli proteins are more widely distributed. Substantial levels are expressed in neurons and oligodendrocyte precursors at embryonic stages as well as in the mature immune system.27,76,89 Five isoforms of the classic proteins ranging from 14 to 21.5 kDa are expressed in myelin-forming cells, but the 17.5-kDa isoform alone can obviously fulfill the functional tasks of all isoforms in the nervous system.84 In the PNS, MBP accounts for 5% to 15% of myelin proteins. As a component of the cytoplasm-related aspect of the compact myelin sheath (i.e., the major dense line), the highly positively charged domains of the protein interact with the negatively charged phospholipids and may mediate compaction.173 Morphologic evidence for this hypothesis is the absence of major dense lines in CNS myelin of shiverer mice,144,150 a spontaneous functional null mutant for classic MBPs.146,149 In the PNS of shiverer mice, absence of MBP does not lead to loss of major dense lines.85,151 This is due to the compensatory function of the cytoplasmic part of the P0 protein.41,111 However, subtle changes have been noted in the PNS of shiverer mice, such as an increase of Schmidt-Lanterman incisures and posttranscriptional modulation of Cx32 protein expression.72,171 Myelin-associated Glycoprotein MAG is a minor component of myelin, accounting for only 0.1% of other PNS myelin proteins. It is a transmembrane glycoprotein with five Ig-like extracellular domains, a transmembrane domain, and an intracellular part. Two isoforms differing in their intracellular domains have been found: a 67-kDa (S-MAG) and a 72-kDa (L-MAG) isoform.88,192 The major form of the PNS is S-MAG. The glycoprotein is a typical adhesion molecule with plasma membrane binding partners.140,203 Recently, the neurotrophin receptor p75 and the Nogo66 receptor have been identified as axon-related partners that transmit the growth-inhibitory activities of MAG to CNS neurons.42,202 Additionally, there is compelling evidence that sialic acid–containing glycoproteins and complex gangliosides are good candidates for axonal MAG receptors.38,203 Thus MAG can also been viewed as a member of the siglecs.83 As opposed to P0, PMP22, and MBP, MAG is confined to noncompacted myelin. Because of its early expression during myelination and its strategic location at the axon–Schwann cell interface,112,190,191 and from perturbation experiments using MAG antisense mRNA,124 the molecule has been considered to play pivotal roles during myelination. Mice deficient in MAG showed an unexpectedly normal myelination in the PNS and MAG-deficient Schwann cells myelinated normally in vitro (see Chapter 24).29,96,118 However, there was prominent axon and myelin degeneration in motor nerves.28,65,204 Thus

MAG is dispensable for myelin formation but is vital for the maintenance of the axon–Schwann cell entity. Connexins 32 and 29 Similar to MAG, Cx32 is another component of noncompacted regions of peripheral nerve myelin,160 but its structure is completely different from that of MAG. Cx32 is a tunnel protein with four transmembrane domains. It belongs to a still-enlarging multigene family comprising more than 15 other gap junction components.169 Cx32 is regulated similarly to a typical myelin gene in that its expression in Schwann cells is dependent on contact with larger caliber axons.160 In the mature myelin sheath, Cx32 is considered to connect each individual noncompacted loop by diffusion channels, thus creating a radial pathway between the periaxonal cytoplasmic collar and the outer Schwann cell cytoplasm.8,160 The blockade of this pathway by the engineered null mutation in the mouse or by various spontaneous mutations in humans1,147 leads to typical pathologic hallmarks such as widened periaxonal collars and abnormal noncompacted zones.4,163 A recently identified myelin-related connexin is connexin 29 (Cx29).3,97,172 Its distribution overlaps with, but is also distinct from, that of Cx32. In addition to the paranodal region and the Schmidt-Lanterman incisures, Cx29 is primarily found at the innermost aspect of the myelin sheath and in the juxtaparanodal regions (i.e., in the immediate neighborhood of K⫹ channels).3 Although its functional role remains to be established, it is plausible to assume that it contributes to the reflexive junctions between noncompacted myelin loops. This would explain why, in Cx32deficient mice, dye-coupling between abaxonal and adaxonal Schwann cell cytoplasm is not interrupted.3,8 E-cadherin E-cadherin is a member of the very large group of Ca2⫹dependent adhesion molecules comprising more than 80 members in mammals.201 It is expressed in many different tissues of epithelial organization.181,182 In the myelin sheath, E-cadherin is associated with autotypic adherens junctions of the inner and outer mesaxons, the SchmidtLanterman incisures, and the paranodal loops.50 Indirect evidence for the role of E-cadherin in myelinating Schwann cells comes from studies in P0-deficient mice in which E-cadherin is downregulated and adherens junctions are only sparsely developed.116 However, tissuespecific gene ablation in Schwann cells revealed that E-cadherin is not required for the integrity of myelinated PNS fibers but does affect the correct formation of autotypic adherens junctions of the outer mesaxon.206 Neurofascin 155 and TAG-1 The neurofascins are members of the Ig superfamily closely related to L1-like adhesion molecules.155,156 In myelinated fibers, the larger isoform, designated neuro-

421

Myelination

fascin 186, is a molecular component of the nodal axolemma, whereas the smaller isoform, neurofascin 155, has been detected as an adaxonal, paranodal membrane component of oligodendrocytes and Schwann cells.21,139,180 Although direct proof is lacking, it is probable that neurofascin 155 is the trans-interacting partner of the paranodal, axon-related NCP1-contactin complex. Indirect evidence for this is the observation that, in mice with an abnormal NCP1-contactin complex expression, neurofascin 155 is altered in its distribution21,44,139 (see also Chapter 24). TAG-1, a glycosylphosphatidylinositol-linked member of the Ig superfamily, has recently been localized at the juxtaparanodal region of developing and mature myelinated fibers. Its role at that position is still obscure, but the co-localization with voltage-gated K⫹ channels suggests participation in the distribution of these channels.187 Myelin and Lymphocyte Protein and Plasmolipin The myelin and lymphocyte protein MAL is a tetraspan proteolipid component of both PNS and CNS myelin.61 In the cell membrane, it is associated with glycosphingolipids possibly forming microdomains in the form of lipid rafts. Apart from its expression in myelin, MAL is expressed in various epithelial tissues, such as kidney and stomach, but also by immune cells.61 In the peripheral nervous system, it is found in compact myelin, but also on nonmyelinating and immature Schwann cells. The early expression argues in favor of a functional role during segregation of axons leading to the promyelin stage. Interestingly, overexpression of MAL leads to an unusual separation of nonmyelinated axons, which might be an argument in favor of a segregating function of MAL in the PNS.62 A MAL-related and myelin-associated proteolipid protein is plasmolipin, which shares approximately 30% amino acid identity with MAL.61,78,102 Its functional roles in the myelinating process are presently not known. Periaxin Periaxin is a Schwann cell protein with two different sizes: 147 kDa (L-periaxin) and 16 kDa (S-periaxin). As a product of a single gene, each form contains a PDZ (postsynaptic density protein 95, Drosophila discs large tumor suppressor, zonula occludens-1) motif, suggesting involvement in cell-cell contact and cell signaling.47,168 The name of the protein is derived from its localization within the periaxonal Schwann cell loop in young mice,70 whereas the expression of the predominantly large form of the protein shifts to the abaxonal compartment in mature nerves.162 The function of periaxin has been investigated by the generation of null mutants. A striking feature was the formation of myelin tomacula at 6 weeks of age followed by myelin degeneration.71 Thus the protein is a nonredundant component of the myelinating Schwann cell stabilizing the myelinated axon–Schwann cell unit. This stabilization is

most probably mediated by the interaction of L-periaxin with dystroglycan-dystrophin–related protein 2 (DRP2), which in turn forms a complex with ␣,␤-dystroglycan that interacts with laminin in the Schwann cell basal lamina.168 Thus periaxin and DRP2 form an essential link between the extracellular matrix of the Schwann cell and the Schwann cell itself, which eventually stabilizes the mature myelin sheath. Mutations in the periaxin gene lead to distinct forms of inherited peripheral neuropathies in humans.18,74

Regulation of Myelin Genes Myelination during development and after demyelination demands an extremely high synthesis rate of myelin proteins and lipids within a short period of time. To accomplish this precisely regulated task, the myelination process is guided by the coordinated expression of genes that encode myelin components.185 The principal control of the system is governed by transcription factors. Some of these are present ubiquitously in many cell types, including Schwann cells, whereas others are more specifically expressed. Together, these factors are responsible for the cell type–specific and differentiation stage–specific gene expression, including the regulation of myelination.197 In addition to the pivotal roles of transcription factors in orchestrating the myelination process, additional regulation at the posttranscriptional level by altering mRNA stability is likely, but this possibility has not yet been thoroughly explored experimentally in the PNS. Recent evidence suggests a particularly important role for the close interaction between neurons (axons) and the ensheathing Schwann cells in the regulatory network governing myelination.104,157 This finding may be part of the reason why the use of transcription factors by Schwann cells is quite different from that by oligodendrocytes, the myelinating counterpart in the CNS.197 Several parallel strategies have been employed to examine myelin gene regulation in the PNS. Major efforts have focused on the identification of transcription factors expressed by Schwann cells that might be master regulators of myelin gene expression, as has been found during muscle development (MyoD).20 This hunt has proved to be difficult, but some candidates have been identified. In particular, suppressed cyclic AMP (cAMP)–inducible POU protein (SCIP, Tst-1, Oct-6) plays a critical role in the regulation of PNS myelination because SCIP-deficient mice are characterized by severe congenital hypomyelination of peripheral nerves. However, the defect is transient and the nerves are close to normal at 3 months of age.81 Similarly, the zinc-finger family protein Egr2 (Krox20) appears to be required for the normal development of the myelinating Schwann cell phenotype. Transgenic mice carrying a null mutation in the egr2 gene display severe defects in Schwann cell development resulting in hyperproliferation

422

Neurobiology of the Peripheral Nervous System

and presumed differentiation arrest at the premyelination stage.186 Egr2 also regulates the expression of genes involved in PNS myelination.119 In line with these findings, different mutations affecting Egr2 are associated with CMT disease type 1, the Déjèrine-Sottas syndrome, and congenital hypomyelinating neuropathy (see Chapter 72).119,194 Additional transcription factors that may affect late steps in the differentiation of Schwann cells include Pax3, Krox24, and Brn5.157,199 Of particular importance is the transcription factor Sox10, which has been identified as a common transcriptional modulator for SCIP, Krox20, and Pax3. It was suggested that Sox10 is responsible for celltype specificity by interacting with these transcription factors in developing and mature glia.87 Furthermore, a crucial role of Sox10 in early Schwann cell development has been demonstrated23,126 and reviewed by Lobsiger and colleagues.100 With regard to PNS myelination, Sox10 regulates MPZ (P0) and GJB1 (Cx32) gene expression.19,132 Mutations in each of these genes are responsible for distinct forms of CMT disease (see Chapter 71). Sox10 mutations are associated with Shah-Wardenburg syndrome (WS4), a neurocristopathy with intestinal aganglionosis, pigmentation defects, sensorineural deafness, and, in specific cases, alterations in myelination in the PNS and CNS.80,125,138 These Sox10 mutants are unable to activate the Cx32 promoter and, conversely, a specific mutation in the Cx32 promoter, previously described in a patient with CMT disease, impairs Sox10 function directly.19 This elegant series of studies provides a direct genetic link between the regulation of myelin gene expression by the transcription factors Egr2 and Sox10 and myelin deficiencies in inherited peripheral neuropathies. It suggests that factors involved in the regulation of myelin gene expression should be generally considered as candidates involved in the etiology of peripheral neuropathies. Furthermore, a general concept emerges that correct myelination, myelin maintenance, and axonal maintenance are crucially dependent on the correct expression levels of myelin proteins.13,104 Besides studying the role of transcription factors and their modulators, myelination-related, cell type–specific control can also be elucidated by the identification and characterization of cis-acting control elements of genes encoding myelin components.196 Transfection experiments in cultured Schwann cells have been used as a standard assay in this regard, followed by classic footprinting and bandshift assays to identify the sequences that are bound by transcription factors.24 However, there are limitations with this assay system because Schwann cells that have been kept in culture do not express myelin genes up to the rates that are observed in vivo during development and in regeneration. Co-culturing with neurons is required for myelination, but because of technical restrictions, this system is not well suited for gene transfer analysis. Transgenic mice, however, provide an excellent assay system to examine myelin gene regulation because numerous developmental

and physiologic signals for correct interactions are present. Such a strategy has recently led to the identification of Schwann cell–specific enhancers that mediate axonal regulation in SCIP,106 MBP,59 Egr2,69 and PMP22103 genes. Initial results of studies with similar aims have been reported for the 2⬘,3⬘-cyclic nucleotide 3⬘-phosphodiesterase gene,30,73 the P0 gene,52 and the proteolipid protein gene.105 Apart from providing important insights into the molecular basis of the regulation of myelin genes, these studies have also generated important tools to target transgenic expression of genes of choice to Schwann cells.53,143 PMP22, the dosage-sensitive culprit gene of CMT disease type 1A and hereditary neuropathy with liability to pressure palsies, was reviewed by Suter and Snipes.176 Studies on PMP22 gene regulation during myelination may also provide the basis for future therapies.77,207 What are the signal transduction events that regulate Schwann cell myelination, and by which transcriptional effectors? Examination of the role of different signaling pathways in Schwann cell differentiation using Schwann cell–neuron co-cultures revealed that, at early stages, inhibition of phosphatidylinositol 3 (PI3)-kinase, but not myelin-associated protein (MAP) kinase, blocked Schwann cell elongation and subsequent myelination.113 After Schwann cells established a one-to-one relationship with axon segments, inhibition of PI3-kinase did not block myelin formation, but the myelin segments were shorter and the rate of myelin protein accumulation was decreased. PI3-kinase inhibition had no detectable effect on the maintenance of myelin sheaths in mature myelinated co-cultures. Interestingly, glial growth factor (GGF), a neuregulin-1 isoform, significantly inhibited myelination in the same system by preventing axonal segregation and ensheathment.208 Treatment of established myelinated cultures with GGF resulted in striking demyelination. The neuregulin receptors ErbB2 and ErbB3 are expressed on ensheathing and myelinating Schwann cells and are rapidly activated by GGF treatment. GGF treatment of myelinating cultures also induced phosphorylation of PI3-kinase, MAP kinase, and a 120-kDa protein. Thus neuronal mitogens, including neuregulins, may inhibit myelination during development, and activation of mitogen signaling pathways may contribute to the initial demyelination and subsequent Schwann cell proliferation observed in various pathologic conditions. Surprisingly, functional information about the downstream transcriptional effectors mediating the events described above is still scarce.196 Several AP-1 (dimeric transcription factors of the Jun, Fos, and ATF family of basic leucine zipper proteins) binding sites have been found in putative regulatory regions of myelin genes. However, although c-Jun is expressed by Schwann cells, a direct functional role of AP-1 transcription factors in regulating PNS myelination remains unclear.174 Similarly, the transcription factor CREB (which becomes activated

423

Myelination

through phosphorylation) is found in Schwann cells, and a protein kinase A–dependent increase of CREB phosphorylation was observed after axonal stimulation and the elevation of cAMP, intracellular calcium, platelet-derived growth factor, or endothelins. Furthermore, ␤-neuregulin treatment causes sustained CREB phosphorylation, and this effect appears to be mainly dependent on the MAP kinase pathway.179 Although this is strong evidence for a function of CREB in the regulation of Schwann cells, a direct link to myelination has not yet been described. Such a link has been established for progesterone that promotes the myelination of sciatic nerves during regeneration after a cryolesion.86 Progesterone triggers a strong upregulation of Egr1 (Krox24), Egr2, Egr3, and FosB in cultured Schwann cells, most likely at the transcriptional level via the interaction of the hormone with its cognate receptor. Furthermore, neuroactive steroids are able to upregulate the mRNA levels of PMP22 and P0 and, thus, specific receptor ligands or antagonists may be promising candidates for therapeutic approaches in demyelinating inherited neuropathies.101,166

Regulation of Myelin Protein Biosynthesis Relatively little is known about the regulation of myelin proteins at the posttranslational level, although a thorough understanding of intracellular transport mechanisms, protein turnover regulation, and membrane and myelin insertion processes are of utmost importance to the regulation of myelination in health and disease. The PMP22 protein provides an illustrative example. Schwann cells express low levels of PMP22 in the absence of neurons. Only if myelin is formed do these levels increase significantly. The exact amount of PMP22 protein produced is critical, because peripheral neuropathies result from either its underexpression or its overexpression. Most of the newly synthesized PMP22 in Schwann cells is rapidly degraded in the endoplasmic reticulum.128,129 Only a small proportion of the total PMP22 acquires complex glycosylation and accumulates in the Golgi compartment. This material is translocated to the Schwann cell membrane in detectable amounts only when axonal contact and myelination occur. However, myelination does not alter the rapid turnover of PMP22 in Schwann cells. Inhibition of the proteosome pathway results in marked accumulation of PMP22 in the perinuclear cytoplasm, forming unique intracellular inclusions called aggresomes.122,152 Moreover, overexpression of PMP22 in Schwann cells can induce the perinuclear accumulation of the protein. Thus degradation of PMP22 by the proteosome pathway might be critical for the regulation of PMP22 protein levels and may be involved in the pathogenesis of PMP22-associated peripheral neuropathies. Recently, association of the chaperone calnexin with PMP22 has also been reported.40 To understand the dynamics of the biosynthesis of PMP22 and other myelin

protein further, sophisticated novel techniques need to be developed that allow direct observation of how proteins travel in living myelinating Schwann cells.26,131

Role of Lipids in Myelination Myelinogenesis requires precise coordination not only of gene expression of the proteins involved in myelination but also of those enzymes associated with the synthesis of myelin lipids. This is not surprising given that myelin membranes contain 70% lipids and 30% proteins, both of which are highly organized within the myelin sheath.67 Myelin membranes resemble apical membranes in their lipid composition and contain high levels of cholesterol and glycosphingolipids such as galactocerebroside and sulfatide.175 In addition, myelin-forming cells are also polarized cells with distinct sorting pathways. Thus it has been speculated that the model of sphingolipid-cholesterol rafts to generate distinct transport and sorting pathways, particular to apical membranes, may also apply to myelinating cells.170 In the CNS, a number of studies have provided support for such a model, and membrane proteins appear to be delivered to the myelin sheath of an oligodendrocyte on rafts with a distinctive lipid composition.91 Not much of this important process has been explored with regard to PNS myelination. Recent data suggest that the myelin proteolipid MAL, which has been shown in other systems to be required for apical transport, and the glycosylphosphatidylinositol-anchored protein CD59 are specific components of PNS myelin rafts.49 The structural myelin proteins PMP22 and P0 are, at least partially, associated with glycosphingolipid rafts,49,79 but the functional role of the association remains to be determined. It appears that the role of rafts in myelination is likely to be more complex than in epithelial cells. Myelinating cells establish several structurally and functionally distinct membrane compartments within the myelin sheath, and each of these myelin compartments is characterized by a unique protein and lipid composition.159 This diversity may preclude the simple assignment of an apical compartment in myelinating cells. Sorting of proteins by different types of rafts may mediate different transport pathways to direct specific proteins to their respective target membranes within the myelin sheath. This process may also guide the assembly of microdomains in myelin membranes that might be important for signal transduction. The role of galactolipids in PNS myelination has recently been examined using mice deficient in the UDPgalactose:ceramide:galactosyltransferase gene.44 These mice lack galactolipids. In contrast to the CNS, PNS myelin sheaths appear normal with regard to myelin stability, thickness, and compaction. The paranodal morphology is also not disrupted in a major way, but transverse bands are absent. This is accompanied by compromised function of the axon–Schwann cell junctions because

424

Neurobiology of the Peripheral Nervous System

potassium channels that are normally shielded by myelin membrane are now exposed to the extracellular milieu. These observations and other recently published data suggest that galactolipids may play an important role in axon–Schwann cell interactions and node of Ranvier formation.141 A potential molecular explanation for these findings might be that glycosphingolipids are important components of intracellular lipid rafts. Thus alteration of these lipids may alter protein trafficking and result in a disruption of the highly ordered cellular distribution of membrane proteins and lipids that are crucial for myelinated nerves. In general, CNS nerves appear to be much more dependent on galactolipids than PNS nerves. The basis for this observation is currently not clear but may be related to the fact that myelinated nerve fibers are stabilized and protected by basal laminae. Recent evidence using genetically altered mice suggests that sialylated glycosphingolipids (gangliosides) are critical components of myelinated peripheral nerves. Mice lacking complex gangliosides show axonal degeneration and demyelination in the PNS.193 This phenotype is similar to MAG-deficient mice and, indeed, MAG is downregulated in complex ganglioside-deficient mouse nerves. The proposal that this is because complex gangliosides are MAG ligands in the nerve remains to be substantiated because an alternative explanation, such as altered MAG trafficking to the cell membrane resulting from altered lipid composition, remains a possibility. The first step of sphingolipid synthesis involves the enzyme serine palmitoyltransferase.98 In mammals, this enzyme is composed of two subunits, encoded by SPTLC1 and SPTLC2, both of which are necessary for activity. Mutations in SPTLC1 are associated with hereditary sensory neuropathy type 1.10,37 Furthermore, a ganglioside-induced differentiation-associated gene (GDAP1) is mutated in the demyelinating peripheral neuropathy CMT type 4A.9,36 In both cases, the corresponding disease mechanisms remain to be determined. Finally, aberrant sphingolipid turnover can also affect myelinated peripheral nerves considerably.184 These diseases are described in Chapter 69.

Role of the Extracellular Matrix, Its Cellular Receptors, and the Schwann Cell Cytoskeleton in Myelination Proper interactions between Schwann cells and axons leading to the formation of myelin are crucially dependent on the extracellular matrix. This has first become evident in classic in vitro myelination studies by Bunge’s laboratory demonstrating that, if the generation of a proper basal lamina is prevented, myelination and the accompanying expression of myelin proteins do not take place.54 The concept emerged that Schwann cells are polarized cells that separate

themselves from mesenchyme by forming a basal lamina. The adaxonal (apical) surface of the Schwann cell contacts the axon while the abaxonal (basal) side is in contact with the basal lamina, likely to provide mechanical support during the myelination process and to permit signal transduction to the cytoskeleton and back. This regulation is critically important to allow the major Schwann cell cytoskeletal reorganization that is required during myelination.55 The importance of the basal lamina in myelination is further highlighted by the mouse mutant dystrophia muscularis and merosin-deficient patients with congenital muscular dystrophy. Both are affected by muscular dystrophy and a dysmyelinating peripheral neuropathy resulting from mutations in the Lama2 gene resulting in defective laminin-2.82 The connection between myelinating Schwann cells and the basal lamina is molecularly maintained by interactions with laminin receptors. There are mainly two sorts of laminin receptors. First, two dystroglycan complexes link the basal lamina to the actin cytoskeleton of Schwann cells through dystrophins or utrophin. A third dystroglycan complex is found exclusively in myelinating Schwann cells and contains DRP2 instead of dystrophin and utrophin.168 Disruption of this complex by mutation in the intracellularly associated periaxin protein, the culprit gene of CMT type 4F, causes demyelination of both mouse and human peripheral nerves.13,71 Intriguingly, Schwann cell dystroglycan complexes also act as receptors for Mycobacterium leprae, which invades nonmyelinating Schwann cells by binding of the M. leprae glycolipid PGL-1 to dystroglycan receptors in the Schwann cell plasma membrane.145 Simultaneously, the bacillus also attacks myelinating Schwann cells and induces demyelination. Dedifferentiation and proliferation of formerly myelinating Schwann cells provides additional carrier cells for colonization by M. leprae, while the loss of myelin sheaths appears to cause axonal degeneration as has been observed in several demyelinating neuropathies.104,109 A second set of laminin receptors found on Schwann cells is integrins.33,142 On myelinating Schwann cells, these include mainly ␣6/␤1 and ␣6/␤4 integrin dimers. Functionblocking anti–␤1 integrin antibodies block Schwann cell attachment to laminin as well as myelination in vitro.56 Furthermore, gene ablation of ␤1 in Schwann cells (␤1 was eliminated at E17.5 in the mouse) leads to markedly delayed myelination. This appears to be due to a partial failure of Schwann cells to subdivide axons as well as impaired adhesion of the promyelinating Schwann cells to their basal laminae.53 However, myelination per se is possible in these mutants, albeit with a long delay. One possible interpretation is that Schwann cells switch their laminin integrin receptors during development. Immature and promyelinating Schwann cells express ␣6/␤1 integrin while myelinating Schwann cells predominantly express ␣6/␤4 integrin.51 Analysis of ␤4 integrin–deficient mice has shown that initial steps of myelination are not dependent on the expression of the ␤4 subunit and in vitro myelin-

Myelination

ation can proceed without ␤4 integrin expression.63 Alternatively, dystroglycan receptors might be predominantly important for myelination.158 Only limited information is currently available with regard to the signal transduction mechanisms related to integrin signaling in Schwann cell myelination. Initial experiments have demonstrated that ␤1 integrin, focal adhesion kinase, paxillin, and fyn kinase form an actin-associated complex in Schwann cells adhering to basal lamina in the presence of axons.31 The adaptor protein paxillin also binds directly to schwannomin, and this interaction mediates the membrane localization of schwannomin to the plasma membrane, where it associates with ␤1 integrin and ErbB2.57 Given the complex structural changes that Schwann cells undergo during myelination, the further elucidation of potential integrin-mediated inside-out and outside-in signaling is likely to contribute significantly to our understanding of the myelination process.

REFERENCES 1. Abrams, C. K., Oh, S., Ri, Y., and Bargiello, T. A.: Mutations in connexin 32: the molecular and biophysical bases for the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Rev. 32:203, 2000. 2. Adlkofer, K., Martini, R., Aguzzi, A., et al.: Hypermyelination and demyelinating peripheral neuropathy in Pmp22-deficient mice. Nat. Genet. 11:274, 1995. 3. Altevogt, B. M., Kleopa, K. A., Postma, F. R., et al.: Connexin29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J. Neurosci. 22:6458, 2002. 4. Anzini, P., Neuberg, D. H. H., Schachner, M., et al.: Structural abnormalities and deficient maintenance of peripheral nerve myelin in mice lacking the gap junction protein connexin 32. J. Neurosci. 17:4545, 1997. 5. Arroyo, E. J., and Scherer, S. S.: On the molecular architecture of myelinated fibers. Histochem. Cell Biol. 113:1, 2000. 6. Arroyo, E. J., Xu, Y.-T., Zhou, L., et al.: Myelinating Schwann cells determine the internodal localization of Kv1.1, Kv1.2, Kv␤2, and Caspr. J. Neurocytol. 28:333, 1999. 7. Baechner, D., Liehr, T., Hameister, H., et al.: Widespread expression of the peripheral myelin protein-22 gene (pmp22) in neural and non-neural tissues during murine development. J. Neurosci. Res. 42:733, 1995. 8. Balice-Gordon, R. J., Bone, L. J., and Scherer, S. S.: Functional gap junctions in the Schwann cell myelin sheath. J. Cell Biol. 142:1095, 1998. 9. Baxter, R. V., Ben Othmane, K., Rochelle, J. M., et al.: Ganglioside-induced differentiation-associated protein-1 is mutant in Charcot-Marie-Tooth disease type 4A/8q21. Nat. Genet. 30:21, 2002. 10. Bejaoui, K., Wu, C., Scheffler, M. D., et al.: SPTLC1 is mutated in hereditary sensory neuropathy, type 1. Nat. Genet. 27:261, 2001.

425

11. Bennett, V., Lambert, S., Davis, J. Q., and Zhang, X.: Molecular architecture of the specialized axonal membrane at the node of Ranvier. Soc. Gen. Physiol. Ser. 52:107, 1997. 12. Bentley, C. A., and Lee, K. F.: p75 is important for axon growth and Schwann cell migration during development. J. Neurosci. 20:7706, 2000. 13. Berger, P., Young, P., and Suter, U.: Molecular cell biology of Charcot-Marie-Tooth disease. Neurogenetics 4:1, 2002. 14. Berthold, C. H.: Development of nodes of Ranvier in feline nerves: an ultrastructural presentation. Microsc. Res. Tech. 34:399, 1996. 15. Bhattacharyya, A., Frank, E., Ratner, N., and Brackenbury, R.: P0 is an early marker of the Schwann cell lineage in chickens. Neuron 7:831, 1991. 16. Billings-Gagliardi, S.: Mode of locomotion of Schwann cells migrating in vivo. Am. J. Anat. 150:73, 1977. 17. Billings-Gagliardi, S., Webster, H. d., and O’Connel, M. F.: In vivo and electron microscopic observations on Schwann cells in developing tadpole nerve fibers. Am. J. Anat. 141:375, 1974. 18. Boerkoel, C., Takashima, H., Stankiewicz, P., et al.: Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am. J. Hum. Genet. 68:325, 2001. 19. Bondurand, N., Girard, M., Pingualt, V., et al.: Human connexin 32, a gap junction protein altered in CharcotMarieTooth disease, is directly regulated by the transcription factor SOX10. Hum. Mol. Genet. 10:2783, 2001. 20. Borycki, A. G., and Emerson, C. P.: Muscle determination: another key player in myogenesis? Curr. Biol. 7:R620, 1997. 21. Boyle, M. E., Berglund, E. O., Murai, K. K., et al.: Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30:385, 2001. 22. Brancolini, C., Edomi, P., Marzinotto, S., and Schneider, C.: Exposure at the cell surface is required for gas3/PMP22 to regulate both cell death and cell spreading: implication for the Charcot-Marie-Tooth type 1A and Dejerine-Sottas diseases. Mol. Biol. Cell 11:2901, 2000. 23. Britsch, S., Goerich, D. E., Riethmacher, D., et al.: The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev. 15:66, 2001. 24. Brown, A. M., and Lemke, G.: Multiple regulatory elements control transcription of the peripheral myelin protein zero gene. J. Biol. Chem. 272:28939, 1997. 25. Bunge, R. P., Bunge, M. B., and Bates, M.: Movements of the Schwann cell nucleus implicate progression of the inner (axon-related) Schwann cell process during myelination. J. Cell Biol. 109:273, 1989. 26. Caduff, J., Sansano, S., Bonnet, A., et al.: Characterization of GFP-MAL expression and incorporation in rafts. Microsc. Res. Tech. 52:645, 2001. 27. Campagnoni, A. T., and Skoff, R. P.: The pathobiology of myelin mutants reveal novel biological functions of the MBP and PLP genes. Brain Pathol. 11:74, 2001. 28. Carenini, S., Montag, D., Cremer, H., et al.: Absence of myelin-associated glycoprotein (MAG) and the neural cell adhesion molecule (N-CAM) interferes with the maintenance, but not with the formation of peripheral myelin. Cell Tissue Res. 287:3,1997.

426

Neurobiology of the Peripheral Nervous System

29. Carenini, S., Montag, D., Schachner, M., and Martini, R.: MAG-deficient Schwann cells myelinate dorsal root ganglion cells in culture. Glia 22:213, 1998. 30. Chandross, K. J., Cohen, R. I., Paras, P. Jr., et al.: Identification and characterization of early glial progenitors using a transgenic selection strategy. J. Neurosci. 19:759, 1999. 31. Chen, L. M., Bailey, D., and Fernandez-Valle, C.: Association of beta 1 integrin with focal adhesion kinase and paxillin in differentiating Schwann cells. J. Neurosci. 20:3776, 2000. 32. Cheng, L., and Mudge, A. W.: Cultured Schwann cells constitutively express the myelin protein P0. Neuron 16:309, 1996. 33. Chernousov, M. A., and Carey, D. J.: Schwann cell extracellular matrix molecules and their receptors. Histol. Histopathol. 15:593, 2000. 34. Ciutat, D., Caldero, J., Oppenheim, R. W., and Esquerda, J. E.: Schwann cell apoptosis during normal development and after axonal degeneration induced by neurotoxins in the chick embryo. J. Neurosci. 16:3979, 1996. 35. Cravioto, H.: The role of Schwann cells in the development of human peripheral nerves: an electron microscopic study. J. Ultrastruct. Res. 12:634, 1965. 36. Cuesta, A., Pedrola, L., Sevilla, T., et al.: The gene encoding ganglioside-induced differentiation-associated protein 1 is mutated in axonal Charcot-Marie-Tooth type 4A disease. Nat. Genet. 30:22, 2002. 37. Dawkins, J. L., Hulme, D. J., Brahmbhatt, S. B., et al.: Mutations in SPTLC1, encoding serine palmitoyltransferase, long chain base subunit-1, cause hereditary sensory neuropathy type I. Nat. Genet. 27:309, 2001. 38. De Bellard, M. E., and Filbin, M. T.: Myelin-associated glycoprotein, MAG, selectively binds several neuronal proteins. J. Neurosci. Res. 56:213, 1999. 39. de Waegh, S. M., Lee, V. M. Y., and Brady, S. T.: Local modulation of neurofilament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68:451, 1992. 40. Dickson, K. M., Bergeron, J. J., Shames, I., et al.: Association of calnexin with mutant peripheral myelin protein-22 ex vivo: a basis for “gain-of-function” ER diseases. Proc. Natl. Acad. Sci. U. S. A. 99:9852, 2002. 41. Ding, Y., and Brunden, K. R.: The cytoplasmic domain of myelin glycoprotein P0 interacts with negatively charged phospholipid bilayers. J. Biol. Chem. 269:10764, 1994. 42. Domeniconi, M., Cao, Z., Spencer, T., et al.: Myelin-associated glycoprotein interacts with the Nogo66 receptor to inhibit neurite outgrowth. Neuron 35:283, 2002. 43. Dong, Z., Sinanan, A., Parkinson, D., et al.: Schwann cell development in embryonic mouse nerves. J. Neurosci. Res. 56:334, 1999. 44. Dupree, J. L., and Popko, B.: Genetic dissection of myelin galactolipid function. J. Neurocytol. 28:271, 1999. 45. D’Urso, D., Brophy, P. J., Staugaitis, S. M., et al.: Protein zero of peripheral nerve myelin: biosynthesis, membrane insertion, and evidence for homotypic interaction. Neuron 2:449, 1990. 46. D’Urso, D., Ehrhardt, P., and Müller, H. W.: Peripheral myelin protein 22 and protein zero: a novel association in peripheral nervous system myelin. J. Neurosci. 19:3396, 1999.

47. Dytrych, L., Sherman, D. L., Gillespie, C. S., and Brophy, P. J.: Two PDZ domain proteins encoded by the murine periaxin gene are the result of alternative intron retention and are differentially targeted in Schwann cells. J. Biol. Chem. 273:5794, 1998. 48. Einheber, S., Zanazzi, G., Ching, W., et al.: The axonal membrane protein Caspr, a homologue of neurexin IV, is a component of the septate-like paranodal junctions that assemble during myelination. J. Cell Biol. 139:1495, 1997. 49. Erne, B., Sansano, S., Frank, M., and Schaeren-Wiemers, N.: Rafts in adult peripheral nerve myelin contain major structural myelin proteins and myelin and lymphocyte protein (MAL) and CD59 as specific markers. J. Neurochem. 82:550, 2002. 50. Fannon, A. M., Sherman, D. L., Ilynia, G., et al.: Novel E-Cadherin-mediated adhesion in peripheral nerve: Schwann cell architecture is stabilized by autotypic adherens junctions. J. Cell Biol. 129:189, 1995. 51. Feltri, L., Scherer, S. S., Nemni, R., et al.: ␤4 integrin expression in myelinating Schwann cells is polarized, developmentally regulated and axonally dependent. Development 120:1287, 1994. 52. Feltri, M. L., D’Antonio, M., Quattrini, A., et al.: A novel P0 glycoprotein transgene activates expression of lacZ in myelin-forming Schwann cells. Eur. J. Neurol. 11:1577, 1999. 53. Feltri, M. L., Graus Porta, D., Previtali, S. C., et al.: Conditional disruption of beta 1 integrin in Schwann cells impedes interactions with axons. J. Cell Biol. 156:199, 2002. 54. Fernandez-Valle, C., Fregien, N., Wood, P. M., and Bunge, M. B.: Expression of the protein zero myelin gene in axonrelated Schwann cells is linked to basal lamina formation. Development 119:867, 1993. 55. Fernandez-Valle, C., Gorman, D., Gomez, A. M., and Bunge, M. B.: Actin plays a role in both changes in cell shape and gene expression associated with Schwann cell myelination. J. Neurosci. 17:241, 1997. 56. Fernandez-Valle, C., Gwynn, L., Wood, P. W., et al.: Anti␤1 integrin antibody inhibits Schwann cell myelination. J. Neurobiol. 25:1207, 1994. 57. Fernandez-Valle, C., Tang, Y., Ricard, J., et al.: Paxillin binds schwannomin and regulates its density-dependent localization and effect on cell morphology. Nat. Genet. 31:354, 2002. 58. Filbin, M. T., Walsh, F. S., Trapp, B. D., et al.: Role of myelin P0 protein as a homophilic adhesion molecule. Nature 344:871, 1990. 59. Forghani, R., Garofalo, L., Foran, D. R., et al.: A distal upstream enhancer from the myelin basic protein gene regulates expression in myelin-forming Schwann cells. J. Neurosci. 21:3780, 2001. 60. Fraher, J. P., O’Leary, D., Moran, M. A., et al.: Relative growth and maturation of axon size and myelin thickness in the tibial nerve of the rat. 1. Normal animals. Acta Neuropathol. 79:364, 1990. 61. Frank, M.: MAL, a proteolipid in glycosphingolipid enriched domains: functional implications in myelin and beyond. Prog. Neurobiol. 60:531, 2000. 62. Frank, M., Atanasoski, S., Sancho, S., et al.: Progressive segregation of unmyelinated axons in peripheral nerves,

Myelination

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

myelin alterations in the CNS, and cyst formation in the kidneys of myelin and lymphocyte protein-overexpressing mice. J. Neurochem. 75:1927, 2000. Frei, R., Dowling, J., Carenini, S., et al.: Myelin formation by Schwann cells in the absence of ␤4 integrin. Glia 27:269, 1999. Friede, R. L., and Beuche, W.: A new approach toward analyzing peripheral nerve fiber populations. I. Variance in sheath thickness corresponds to different geometric proportions of the internodes. J. Neuropathol. Exp. Neurol. 44:60, 1985. Fruttiger, M., Montag, D., Schachner, M., and Martini, R.: Crucial role for the myelin-associated glycoprotein in the maintenance of axon-myelin integrity. Eur. J. Neurosci. 7:511, 1995. Gao, Y., Li, W., and Filbin, M. T.: Acylation of myelin P0 protein is required for adhesion. J. Neurosci. Res. 60:704, 2000. Garbay, B., Heape, A. M., Sargueil, F., and Cassagne, C.: Myelin synthesis in the peripheral nervous system. Prog. Neurobiol. 61:267, 2000. Gatzinsky, K. P.: Node-paranode regions as local degradative centres in alpha-motor axons. Microsc. Res. Tech. 34:492, 1996. Ghislain, J., Desmarquet-Trin-Dinh, C., Jaegle, M., et al.: Characterisation of cis-acting sequences reveals a biphasic, axon-dependent regulation of Krox20 during Schwann cell development. Development 129:155, 2002. Gillespie, C. S., Sherman, D. L., Blair, G. E., and Brophy, P. J.: Periaxin, a novel protein of myelinating Schwann cells with a possible role in axon ensheathment. Neuron 12:497, 1994. Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., et al.: Peripheral demyelination and neuropathic pain behavior in periaxin-deficient mice. Neuron 26:523, 2000. Gould, R. M., Byrd, A. L., and Barbarese, E.: The number of Schmidt-Lanterman incisures is more than doubled in shiverer PNS myelin sheaths. J. Neurocytol. 24:85, 1995. Gravel, M., Di Polo, A., Valera, P. B., and Braun, P. E.: Four-kilobase sequence of the mouse CNP gene directs spatial and temporal expression of lacZ in transgenic mice. J. Neurosci. Res. 53:393, 1998. Guilbot, A., Williams, A., Ravise, N., et al.: A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum. Mol. Genet. 10:415, 2001. Hagedorn, L., Suter, U., and Sommer, L.: P0 and PMP22 mark a multipotent neural crest-derived cell type that displays community effects in response to TGF-␤family factors. Development 126:3781, 1999. Hahn, A. F., Whitaker, J. N., Kachar, B., and Webster, H. d.: P2, P1, and P0 myelin protein expression in developing rat sixth nerve: a quantitative immunocytochemical study. J. Comp. Neurol. 260:501, 1987. Hai, M., Bidichandani, S. I., Hogan, M. E., and Patel, P. I.: Competitive binding of triplex-forming oligonucleotides in the two alternate promoters of the PMP22 gene. Antisense Nucleic Acid Drug Dev. 11:233, 2001. Hamacher, M., Pippirs, U., Köhler, A., et al.: Plasmolipin: genomic structure, chromosomal localization, protein

79.

80.

81. 82.

83.

84.

85.

86.

87.

88.

89.

90. 91. 92.

93. 94. 95.

96.

97.

427 expression pattern, and putative association with BardetBiedl syndrome. Mamm. Genome 12:933, 2001. Hasse, B., Bosse, F., and Muller, H. W.: Proteins of peripheral myelin are associated with glycosphingolipid/cholesterol-enriched membranes. J. Neurosci. Res. 69:227, 2002. Inoue, K., Tanabe, Y., and Lupski, J. R.: Myelin deficiencies in both the central and the peripheral nervous systems associated with a SOX10 mutation. Ann. Neurol. 46:313, 1999. Jaegle, M., and Meijer, D.: Role of Oct-6 in Schwann cell differentiation. Microsc. Res. Tech. 41:372, 1998. Jones, K. J., Morgan, G., Johnston, H., et al.: The expanding phenotype of laminin alpha2 chain (merosin) abnormalities: case series and review. J. Med. Genet. 38:649, 2001. Kelm, S., Schauer, R., and Crocker, P. R.: The sialoadhesins—a family of sialic acid-dependent cellular recognition molecules within the immunoglobulin superfamily. Glycoconj. J. 13:913, 1996. Kimura, M., Sato, M., Akatsuka, A., et al.: Overexpression of a minor component of myelin basic protein isoform (17.2 kDa) can restore myelinogenesis in transgenic shiverer mice. Brain Res. 785:245, 1998. Kirschner, D. A., and Ganser, A. L.: Compact myelin exists in the absence of basic protein in the shiverer mutant mouse. Nature 283:207, 1980. Koenig, H. L., Schumacher, M., Ferzaz, B., et al.: Progesterone synthesis and myelin formation by Schwann cells. Science 268:1500, 1995. Kuhlbrodt, K., Herbarth, B., Sock, E., et al.: Sox10, a novel transcriptional modulator in glial cells. J. Neurosci. 18:237, 1998. Lai, C., Brow, M. A., Nave, K. A., et al.: Two forms of 1B236/myelin-associated glycoprotein, a cell adhesion molecule for postnatal neural development, are produced by alternative splicing. Proc. Natl. Acad. Sci. U. S. A. 84:4337, 1987. Landry, C. F., Ellison, J. P., Skinner, E., and Campagnoni, A. T.: Golli-MBP proteins mark the earliest stages of fiber extension and terminal arboration in the mouse peripheral nervous system. J. Neurosci. Res. 50:265, 1997. Le Douarin, N., Dulac, C., Dupin, E., and Cameron-Curry, P.: Glia cell lineages in the neural crest. Glia 4:175, 1991. Lee, A. G.: Myelin: delivery by raft. Curr. Biol. 11:R60, 2001. Lee, M. J., Brennan, A., Blanchard, A., et al.: P0 is constitutively expressed in the rat neural crest and embryonic nerves and is negatively and positively regulated by axons to generate non-myelin forming and myelin forming Schwann cells, respectively. Mol. Cell. Neurosci. 8:336, 1997. Lemke, G.: Unwrapping the genes of myelin. Neuron 1:535, 1988. Lemke, G.: The molecular genetics of myelination: an update. Glia 7:263, 1993. Lemke, G., Lamar, E., and Patterson, J.: Isolation and analysis of the gene encoding peripheral myelin protein zero. Neuron 1:73, 1988. Li, C., Tropak, M. B., Gerial, R., et al.: Myelination in the absence of myelin-associated glycoprotein. Nature 369:747, 1994. Li, X., Lynn, B. D., Olson, C., et al.: Connexin29 expression, immunocytochemistry and freeze-fracture replica immunogold labelling (FRIL) in sciatic nerve. Eur. J. Neurosci. 16:795, 2002.

428

Neurobiology of the Peripheral Nervous System

98. Linn, S. C., Kim, H. S., Keane, E. M., et al.: Regulation of de novo sphingolipid biosynthesis and the toxic consequences of its disruption. Biochem. Soc. Trans. 29:831, 2001. 99. Lobsiger, C. S., Schweitzer, B., Taylor, V., and Suter, U.: Platelet-derived growth factor-BB supports the survival of cultured rat Schwann cell precursors in synergy with neurotrophin-3. Glia 30:290, 2000. 100. Lobsiger, C. S., Taylor, V., and Suter, U.: The early life of a Schwann cell. Biol. Chem. 383:245, 2002. 101. Magnaghi, V., Cavarretta, I., Galbiati, M., et al.: Neuroactive steroids and peripheral myelin proteins. Brain Res. Brain Res. Rev. 37:360, 2001. 102. Magyar, J., Ebensperger, C., Schaeren-Wiemers, N., and Suter, U.: Myelin and lymphocyte protein (MAL/ MVP17/VIP17) and plasmolipin are members of an extended gene family. Gene 189:269, 1997. 103. Maier, M., Berger, P., Nave, K. A., and Suter, U.: Identification of the regulatory region of the peripheral myelin protein 22 (PMP22) gene that directs temporal and spatial expression in development and regeneration of peripheral nerves. Mol. Cell. Neurosci. 20:93, 2002. 104. Maier, M., Berger, P., and Suter, U.: Understanding Schwann cell-neurone interactions: the key to CharcotMarie-Tooth disease? J. Anat. 200:357, 2002. 105. Mallon, B. S., Shick, H. E., Kidd, G. J., and Macklin, W. B.: Proteolipid promoter activity distinguishes two populations of NG2-positive cells throughout neonatal cortical development. J. Neurosci. 22:876, 2002. 106. Mandemakers, W., Zwart, R., Jaegle, M., et al.: A distal Schwann cell-specific enhancer mediates axonal regulation of the Oct-6 transcription factor during peripheral nerve development and regeneration. EMBO J. 19:2992, 2000. 107. Martin, S., Levine, A. K., Chen, Z. J., et al.: Deposition of the NG2 proteoglycan at nodes of Ranvier in the peripheral nervous system. J. Neurosci. 21:8119, 2001. 108. Martini, R.: Expression and functional roles of neural cell surface molecules and extracellular matrix components during development and regeneration of peripheral nerves. J. Neurocytol. 23:1, 1994. 109. Martini, R.: The effect of myelinating Schwann cells on axons. Muscle Nerve 24:456, 2001. 110. Martini, R., Bollensen, E., and Schachner, M.: Immunocytological localization of the major peripheral nervous system glycoprotein P0 and the L2/HNK-1 and L3 carbohydrate structures in developing and adult mouse sciatic nerve. Dev. Biol. 129:330, 1988. 111. Martini, R., Mohajeri, M. H., Kasper, S., et al.: Mice doubly deficient in the genes for P0 and myelin basic protein show that both proteins contribute to the formation of the major dense line in peripheral nerve myelin. J. Neurosci. 15:4488, 1995. 112. Martini, R., and Schachner, M.: Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and MAG) and their shared carbohydrate epitope and myelin basic protein in developing sciatic nerve. J. Cell Biol. 103:2439, 1986. 113. Maurel, P., and Salzer, J. L.: Axonal regulation of Schwann cell proliferation and survival and the initial events of myelin-

114.

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

ation requires PI 3-kinase activity. J. Neurosci. 20:4635, 2000. Melendez-Vasquez, C. V., Rios, J. C., Zanazzi, G., et al.: Nodes of Ranvier form in association with ezrin-radixin-moesin (ERM)-positive Schwann cell processes. Proc. Natl. Acad. Sci. U. S. A. 98:1235, 2001. Menegoz, M., Gaspar, P., Le Bert, M., et al.: Paranodin, a glycoprotein of neuronal paranodal membranes. Neuron 19:319, 1997. Menichella, D. M., Arroyo, E. J., Awatramani, R., et al.: Protein zero is necessary for E-cadherin-mediated adherens junction formation in Schwann cells. Mol. Cell. Neurosci. 18:606, 2001. Mirsky, R., Jessen, K. R., Brennan, A., et al.: Schwann cells as regulators of nerve development. J. Physiol. Paris 96:17, 2002. Montag, D., Giese, K. P., Bartsch, U., et al.: Mice deficient for the myelin-associated glycoprotein show subtle abnormalities in myelin. Neuron 13:229, 1994. Nagarajan, R., Svaren, J., Le, N., et al.: EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron 30:355, 2001. Noakes, P. G., and Bennett, M. R.: Growth of axons into developing muscles of the chick forelimb is preceded by cells that stain with Schwann cell antibodies. J. Comp. Neurol. 259:330, 1987. Notterpek, L., Roux, K. J., Amici, S. A., et al.: Peripheral myelin protein 22 is a constituent of intercellular junctions in epithelia. Proc. Natl. Acad. Sci. U. S. A. 98:14404, 2001. Notterpek, L., Ryan, M. C., Tobler, A. R., and Shooter, E. M.: PMP22 accumulation in aggresomes: implications for CMT1A pathology. Neurobiol. Dis. 6:450, 1999. Owens, G. C., and Boyd, C. J.: Expression antisense P0 RNA in Schwann cells perturbs myelination. Development 112:639, 1991. Owens, G. C., and Bunge, R. P.: Schwann cells infected with a recombinant retrovirus expressing myelin-associated glycoprotein antisense RNA do not form myelin. Neuron 7:565, 1991. Paratore, C., Eichenberger, C., Suter, U., and Sommer, L.: Sox10 haploinsufficiency affects maintenance of progenitor cells in a mouse model of Hirschsprung disease. Hum. Mol. Genet. 11:3075, 2002. Paratore, C., Goerich, D. E., Suter, U., et al.: Survival and glial fate acquisition of neural crest cells are regulated by an interplay between the transcription factor Sox10 and extrinsic combinatorial signaling. Development 128:3949, 2001. Paratore, C., Hagedorn, L., Floris, J., et al.: Cell-intrinsic and cell-extrinsic cues regulating lineage decisions in multipotent neural crest-derived progenitor cells. Int. J. Dev. Biol. 46:193, 2002. Pareek, S., Notterpek, L., Snipes, G. J., et al.: Neurons promote translocation of peripheral myelin protein 22 into myelin. J. Neurosci. 17:7754, 1997. Pareek, S., Suter, U., Snipes, G. J., et al.: Detection and processing of peripheral myelin protein PMP22 in cultured Schwann cells. J. Biol. Chem. 268:10372, 1993. Parmantier, E., Cabon, F., Braun, C., et al.: Peripheral myelin protein-22 is expressed in rat and mouse brain and spinal cord motoneurons. Eur. J. Neurosci. 7:1080, 1995.

Myelination 131. Pedraza, L., and Colman, D. R.: Fluorescent myelin proteins provide new tools to study the myelination process. J. Neurosci. Res. 60:697, 2000. 132. Peirano, R. I., Goerich, D. E., Riethmacher, D., and Wegner, M.: Protein zero gene expression is regulated by the glial transcription factor Sox10. Mol. Cell. Biol. 20:3198, 2000. 133. Peles, E., Nativ, M., Lustig, M., et al.: Identification of a novel contactin-associated transmembrane receptor with multiple domains implicated in protein-protein interactions. EMBO J. 16:978, 1997. 134. Peles, E., and Salzer, J. L.: Molecular domains of myelinated axons. Curr. Opin. Neurobiol. 10:558, 2000. 135. Peters, A., and Muir, A. R.: The relationship between axons and Schwann cells during development of peripheral nerves in the rat. Q. J. Exp. Physiol. 44:117, 1959. 136. Peters, A., Palay, S. L., and Webster, H. d.: The Fine Structure of the Nervous System. New York, Oxford University Press, 1991. 137. Phillips, D. D., Hibbs, R. G., Ellison, J. P., and Shapiro, H.: An electron microscopic study of central and peripheral nodes of Ranvier. J. Anat. 111:229, 1972. 138. Pingault, V., Guiochon-Mantel, A., Bondurand, N., et al.: Peripheral neuropathy with hypomyelination, chronic intestinal pseudo-obstruction and deafness: a developmental “neural crest syndrome” related to a SOX10 mutation. Ann. Neurol. 48:671, 2000. 139. Poliak, S., Gollan, L., Salomon, D., et al.: Localization of Caspr2 in myelinated nerves depends on axon-glia interactions and the generation of barriers along the axon. J. Neurosci. 21:7568, 2001. 140. Poltorak, M., Sadoul, R., Keilhauer, G., et al.: Myelinassociated glycoprotein, a member of the L2/HNK-1 family of neural cell adhesion molecules, is involved in neuronoligodendrocyte and oligodendrocyte-oligodendrocyte interaction. J. Cell Biol. 105:1893, 1987. 141. Popko, B.: Myelin galactolipids: mediators of axon-glial interactions? Glia 29:149, 2000. 142. Previtali, S. C., Feltri, M. L., Archelos, J. J., et al.: Role of integrins in the peripheral nervous system. Prog. Neurobiol. 64:35, 2001. 143. Previtali, S. C., Quattrini, A., Fasolini, M., et al.: Epitopetagged P(0) glycoprotein causes Charcot-Marie-Tooth-like neuropathy in transgenic mice. J. Cell Biol. 151:1035, 2000. 144. Privat, A., Jacque, C., Bourre, J. M., et al.: Absence of the major dense line in the myelin of the mutant mouse “shiverer.” Neurosci. Lett. 12:107, 1979. 145. Rambukkana, A., Zanazzi, G., Tapinos, N., and Salzer, J. L.: Contact-dependent demyelination by Mycobacterium leprae in the absence of immune cells. Science 296:927, 2002. 146. Readhead, C., Popko, B., Takahashi, N., et al.: Expression of a myelin basic protein gene in transgenic shiverer mice: correction of the dysmyelinating phenotype. Cell 48:703, 1987. 147. Ressot, C., and Bruzzone, R.: Connexin channels in Schwann cells and the development of the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Rev. 32:192, 2000. 148. Riethmacher, D., Sonnenberg-Rietmacher, E., Brinkmann, V., et al.: Severe neuropathies in mice with targeted mutations in the ErbB3 receptor. Nature 389:725, 1997. 149. Roach, A., Boylan, K., Horvath, S., et al.: Characterization of cloned cDNA representing rat myelin basic protein:

150. 151. 152.

153.

154.

155. 156.

157. 158. 159.

160.

161.

162.

163.

164.

165.

166.

167.

168.

429 absence of expression in brain of shiverer mutant mice. Cell 34:799, 1983. Rosenbluth, J.: Central myelin in the mouse mutant shiverer. J. Comp. Neurol. 194:639, 1980. Rosenbluth, J.: Peripheral myelin in the mouse mutant shiverer. J. Comp. Neurol. 193:729, 1980. Ryan, M. C., Shooter, E. M., and Notterpek, L.: Aggresome formation in neuropathy models based on peripheral myelin protein 22 mutations. Neurobiol. Dis. 10:109, 2002. Sancho, S., Young, P., and Suter, U.: Regulation of Schwann cell proliferation and apoptosis in PMP22-deficient mice and mouse models of Charcot-Marie-Tooth disease type 1A. Brain 124:2177, 2001. Sanes, J. R., and Lichtman, J. W.: Development of the vertebrate neuromuscular junction. Annu. Rev. Neurosci. 22:389, 1999. Schachner, M.: Neural recognition molecules in disease and regeneration. Curr. Opin. Neurobiol. 4:726, 1994. Schachner, M., and Martini, R.: Glycans and the modulation of neural recognition molecule function. Trends Neurosci. 18:183, 1995. Scherer, S. S.: The biology and pathobiology of Schwann cells. Curr. Opin. Neurol. 10:386, 1997. Scherer, S. S.: Myelination: some receptors required. J. Cell. Biol. 156:13, 2002. Scherer, S. S., and Arroyo, E. J.: Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Syst. 7:1, 2002. Scherer, S. S., Deschênes, S. M., Xu, Y. T., et al.: Connexin32 is a myelin-related protein in the PNS and CNS. J. Neurosci. 15:8281, 1995. Scherer, S. S., Xu, T., Crino, P., et al.: Ezrin, radixin, and moesin are components of Schwann cell microvilli. J. Neurosci. Res. 65:150, 2001. Scherer, S. S., Xu, Y. T., Bannerman, P. G. C., et al.: Periaxin expression in myelinating Schwann cells: modulation by axon-glial interactions and polarized localization during development. Development 121:4265, 1995. Scherer, S. S., Xu, Y.-T., Nelles, E., et al.: Connexin32-null mice develop demyelinating peripheral neuropathy. Glia 24:8, 1998. Schneider-Schaulies, J., von Brunn, A., and Schachner, M.: Recombinant peripheral myelin protein P0 confers both adhesion and neurite outgrowth-promoting properties. J. Neurosci. Res. 27:286, 1990. Schröder, J. M., Bohl, J., and von Bardeleben, U.: Changes of the ratio between myelin thickness and axon diameter in human developing sural, femoral, ulnar, facial, and trochlear nerves. Acta Neuropathol. 76:471, 1988. Schumacher, M., Guennoun, R., Mercier, G., et al.: Progesterone synthesis and myelin formation in peripheral nerves. Brain Res. Brain Res. Rev. 37:343, 2001. Shapiro, L., Doyle, J. P., Hensley, P., et al.: Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17:435, 1996. Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J.: Specific disruption of a Schwann cell dystrophin-related protein complex in a demyelinating neuropathy. Neuron 30:677, 2001.

430

Neurobiology of the Peripheral Nervous System

169. Shibata, Y., Kumai, M., Nishii, K., and Nakamura, K.: Diversity and molecular anatomy of gap junctions. Med. Electron Microsc. 34:153, 2001. 170. Simons, K., and Ikonen, E.: Functional rafts in cell membranes. Nature 387:569, 1997. 171. Smith-Slatas, C., and Barbarese, E.: Myelin basic protein gene dosage effects the PNS. Mol. Cell. Neurosci. 15:343, 2000. 172. Sohl, G., Eiberger, J., Jung, Y. T., et al.: The mouse gap junction gene connexin29 is highly expressed in sciatic nerve and regulated during brain development. Biol. Chem. 382:973, 2001. 173. Staugaitis, S. M., Colman, D. R., and Pedraza, L.: Membrane adhesion and other functions for the myelin basic proteins. BioEssays 18:13, 1995. 174. Stewart, H. J.: Expression of c-Jun, Jun B, Jun D and cAMP response element binding protein by Schwann cells and their precursors in vivo and in vitro. Eur. J. Neurosci. 7:1366, 1995. 175. Stoffel, W., and Bosio, A.: Myelin glycolipids and their functions. Curr. Opin. Neurobiol. 7:654, 1997. 176. Suter, U., and Snipes, G. J.: Biology and genetics of hereditary motor and sensory neuropathies. Annu. Rev. Neurosci. 18:45, 1995. 177. Suter, U., and Snipes, G. J.: Peripheral myelin protein 22: facts and hypotheses. J. Neurosci. Res. 40:145, 1995. 178. Suter, U., Snipes, G. J., Schoener-Scott, R., et al.: Regulation of tissue-specific expression of alternative peripheral myelin protein-22 (PMP22) gene transcripts by two promoters. J. Biol. Chem. 269:25795, 1994. 179. Tabernero, A., Stewart, H. J. S., Jessen, K. R., and Mirsky, R.: The neuron-glia signal beta neuregulin induces sustained CREB phosphorylation on Ser-133 in cultured rat Schwann cells. Mol. Cell. Neurosci. 10:309, 1998. 180. Tait, S., Gunn-Moore, F., Collinson, J. M., et al.: An oligodendrocyte adhesion molecule at the site of assembly of the paranodal axo-glial junction. J. Cell Biol. 150:657, 2000. 181. Takeichi, M.: The cadherins: cell-cell adhesion molecules controlling animal morphogenesis. Development 102:639, 1988. 182. Takeichi, M.: Cadherins: a molecular family important in selective cell-cell adhesion. Annu. Rev. Biochem. 59:237, 1990. 183. Taylor, V., Zgraggen, C., Naef, R., and Suter, U.: Membrane topology of peripheral myelin protein 22. J. Neurosci. Res. 62:15, 2001. 184. Tifft, C. J., and Proia, R. L.: Stemming the tide: glycosphingolipid synthesis inhibitors as therapy for storage diseases. Glycobiology 10:1249, 2000. 185. Toews, A. D., Hostettler, J., Barrett, C., and Morell, P.: Alterations in gene expression associated with primary demyelination and remyelination in the peripheral nervous system. Neurochem. Res. 22:1271, 1997. 186. Topilko, P., Schneider-Maunoury, S., Levi, G., et al.: Krox20 controls myelination in the peripheral nervous system. Nature 371:796, 1994. 187. Traka, M., Dupree, J. L., Popko, B., and Karagogeos, D.: The neuronal adhesion protein TAG-1 is expressed by Schwann cells and oligodendrocytes and is localized to the

188.

189.

190.

191.

192.

193.

194.

195.

196. 197. 198.

199.

200.

201.

202.

203.

204.

juxtaparanodal region of myelinated fibers. J. Neurosci. 22:3016, 2002. Trapp, B.: Distribution of the myelin-associated glycoprotein and P0 protein during myelin compaction in quaking mouse peripheral nerve. J. Cell Biol. 107:675, 1988. Trapp, B. D., Andrews, S. B., Wong, A., et al.: Co-localization of the myelin-associated glycoprotein and the microfilament components, F-actin and spectrin, in Schwann cells of myelinated nerve fibres. J. Neurocytol. 18:47, 1989. Trapp, B. D., O’Connell, M. F., and Andrews, S. B.: Ultrastructural immunolocalization of MAG and P0 proteins in cryosections of peripheral nerve. J. Cell Biol. 1035:228a, 1986. Trapp, B. D., and Quarles, R. H.: Presence of the myelinassociated glycoprotein correlates with alterations in the periodicity of peripheral myelin. J. Cell Biol. 92:877, 1982. Tropak, M. B., Johnson, P. W., Dunn, R. J., and Roder, J. C.: Differential splicing of MAG transcripts during CNS and PNS development. Brain Res. 464:143, 1988. Vyas, A. A., and Schnaar, R. L.: Brain gangliosides: functional ligands for myelin stability and the control of nerve regeneration. Biochimie 83:677, 2001. Warner, L. E., Mancias, P., Butler, I. J., et al.: Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet. 18:382, 1998. Webster, H. d.: The geometry of peripheral myelin sheaths during their formation and growth in rat sciatic nerves. J. Cell Biol. 48:348, 1971. Wegner, M.: Transcriptional control in myelinating glia: flavors and spices. Glia 31:1, 2000. Wegner, M.: Transcriptional control in myelinating glia: the basic recipe. Glia 29:118, 2000. Wong, M. H., and Filbin, M. T.: Dominant-negative effect on adhesion by myelin P0 protein truncated in its cytoplasmic domain. J. Cell Biol. 134:1531, 1996. Wu, R., Jurek, M., Sundarababu, S., and Weinstein, D. E.: The POU gene Brn-5 is induced by neuregulin and is restricted to myelinating Schwann cells. Mol. Cell. Neurosci. 17:683, 2001. Xu, W., Shy, M., Kamholz, J., et al.: Mutations in the cytoplasmic domain of P0 reveal a role for PKC-mediated phosphorylation in adhesion and myelination. J. Cell Biol. 155:439, 2001. Yagi, T., and Takeichi, M.: Cadherin superfamily genes: functions, genomic organization, and neurologic diversity. Genes Dev. 14:1169, 2000. Yamashita, T., Higuchi, H., and Tohyama, M.: The p75 receptor transduces the signal from myelin-associated glycoprotein to Rho. J. Cell Biol. 157:565, 2002. Yang, L. J. S., Zeller, C. B., Shaper, N. L., et al.: Gangliosides are neural ligands for myelin-associated glycoprotein. Proc. Natl. Acad. Sci. U. S. A. 93:814, 1996. Yin, X., Crawford, T. O., Griffin, J. W., et al.: Myelinassociated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J. Neurosci. 18:1953, 1998.

Myelination 205. Yin, X., Kidd, G. J., Wrabetz, L., et al.: Schwann cell myelination requires timely and precise targeting of P0 protein. J. Cell Biol. 148:1009, 2000. 206. Young, P., Boussadia, O., Berger, P., et al.: E-cadherin is required for the correct formation of autotypic adherens junctions of the outer mesaxon but not for the integrity of myelinated fibers of peripheral nerves. Mol. Cell. Neurosci. 21:341, 2002. 207. Young, P., and Suter, U.: Disease mechanisms and potential therapeutic strategies in Charcot-Marie-Tooth disease. Brain Res. Rev. 36:213, 2001.

431

208. Zanazzi, G., Einheber, S., Westreich, R., et al.: Glial growth factor/neuregulin inhibits Schwann cell myelination and induces demyelination. J. Cell Biol. 152:1289, 2001. 209. Ziskind-Conhaim, L.: Physiological and morphological changes in developing peripheral nerves of rat embryos. Dev. Brain Res. 42:15, 1988. 210. Zoidl, G., Blass-Kampmann, S., D’Urso, D., et al.: Retroviral-mediated gene transfer of the peripheral myelin protein PMP22 in Schwann cells: modulation of cell growth. EMBO J. 14:1122, 1995.

20 The Control of Axonal Caliber JOHN W. GRIFFIN AND AHMET HÖKE

Cytology of the Axon Extrinsic Factors That Influence Axonal Caliber Myelination: Alteration of the Axonal Cytoskeleton Signaling Pathways from Myelinating Schwann Cell to Axonal Cytoskeleton Myelin-Associated Glycoprotein: Involvement in Signaling to the Axon

Does Axonal Caliber Influence Myelination? Intrinsic Factors That Influence Axonal Caliber Neurofilaments and Control of Axonal Diameter in Large-Caliber Myelinated Nerve Fibers Neurofilament Transport

One of the cardinal phenotypic features of a neuron is the caliber of its axon. Axonal caliber in the peripheral nervous system (PNS) is precisely regulated, as reflected in the predictable association of fiber sizes with functional categories. The range of normal axonal calibers is noteworthy; a factor of 100 separates the cross-sectional areas of the smallest unmyelinated axons in the human PNS from the largest myelinated ones.7 After maturation caliber can undergo only modest variation, even in disease. For example, an axon 10 ␮m in diameter never withers to 2 ␮m, nor does the reverse occur. Axonal caliber has important functional correlates. In myelinated fibers, the conduction velocity is linearly related to nerve fiber diameter.29 Consequently, in the case of the simplest neuronal circuit, the monosynaptic reflex, the latency (duration) of the reflex changes when axonal caliber is altered.15 The order of recruitment of fibers also correlates with their axonal calibers.44 In addition to these physiologic measures, axonal caliber has obvious structural influences on non-neuronal cells, including the ensheathing Schwann cells. For example, in myelinated fibers myelin sheath thickness and total myelin volume are influenced in some fashion by axonal caliber, and some data suggest that even the initial formation of a myelin sheath may be influenced by caliber.118 Finally, alterations in axonal caliber are prominent and frequent pathologic abnormalities in peripheral nerve diseases and may contribute to such late consequences as secondary demyelination.42

Alterations in Velocity of Neurofilament Transport: Influence on Axonal Neurofilament Content and Axonal Caliber Neurofilament Gene Expression Influences Axonal Caliber Regulation of Neurofilament Gene Expression Conclusions

This chapter considers the mechanisms by which axonal caliber is determined. The influences on axonal caliber are conveniently divided into those extrinsic to the neuron and those intrinsic to it. For example, the synthesis and transport of axonal constituents are considered as intrinsic factors affecting or determining axonal size. However, at some stage of development it is likely that the “decision” of the neuron to grow to its adult proportions is determined by the action of extrinsic factors. Conversely, the presence or absence of myelination around a given axonal segment influences caliber. This “extrinsic” influence is now known to alter posttranslational modifications of intrinsic axonal cytoskeletal proteins.

CYTOLOGY OF THE AXON As viewed in electron micrographs, axoplasm is deceptively simple, containing only a few particulate organelles, mitochondria, and the cytoskeletal elements. Rotaryshadowed, deep-etched preparations give a strikingly different image by emphasizing the innumerable side arms that appear to interlace the cytoskeletal elements to each other as well as to the particulate organelles.12,45,46,116 In both preparations it is apparent that most of the space of the axoplasm appears empty and is occupied by the cytosol. 433

434

Neurobiology of the Peripheral Nervous System

As in other cells, the organization of the cytoplasm appears to be a function of the cytoskeleton.32,62,105 The axonal cytoskeleton includes three elements: microfilaments (MFs), microtubules (MTs), and neurofilaments (NFs). Microfilaments, which are poorly preserved in routine electron micrographs, are prominent in the subaxolemmal cytoskeleton.16,64,68,115 In squid axoplasm, MFs of the inner cytoskeleton are organized in longitudinally oriented columns that are surrounded by a dense matrix and associated with the MT domains.16 Microtubules are present throughout the axoplasm; these tubulin polymers are relatively long, with lengths estimated to be 350 to 750 ␮m.114 Neurofilaments, the major intermediate filaments (IFs) of neurons, deserve special attention in this chapter because of their relationship to axonal caliber. IFs as a class are 10-nm filaments found in the cytoskeletons of eukaryotic cells. Unlike actin filaments and MTs, which are composed of highly conserved protein subunits, the protein composition of IFs varies in different cell types. There are five generally accepted classes of IF that can be

distinguished based on differences in sequence and gene structure.17,32 Class I and class II are composed of type I (acidic) and type II (basic) keratins, respectively, which are expressed in epithelial cells. Class III comprises a diverse group of proteins including vimentin, expressed in mesenchymal cells and embryonic neurons108; glial fibrillary acidic protein, expressed in glial cells97,130; and desmin, expressed in muscle cells. A class III IF especially relevant to the PNS is peripherin, expressed in small sensory and autonomic neurons.82 Class IV comprises the neuronspecific NF proteins, which are reviewed in detail later.97,130 Finally, class V comprises the nuclear lamins. The NFs are composed of three distinct protein subunits, neurofilament light (NF-L, 68 kDa), neurofilament medium (NF-M, 145 kDa), and neurofilament heavy (NF-H, 200 kDa), collectively referred to as the NF triplet.54,67,92 Like all IF proteins, each NF subunit contains a central, structurally conserved, ␣-helical coiled-coil domain divided into three regions (Fig. 20–1).17,32,105 The presence of this conserved rod domain suggests that each of these subunits is associated with the NF core. A unique carboxy-terminal

FIGURE 20–1 Comparison of the structural domains of NF-L, NF-M, and NF-H polypeptides. All three contain an amino-terminal domain approximately 100 amino acids in length and a structurally conserved ␣-helical rod domain, which is approximately 310 amino acids in length and composed of three coiled-coil regions, designated 1a, 1b, and 2. The greatest divergence among NF-L, NF-M, and NF-H resides in the carboxy-terminal tail domains, which in mouse are 140, 435, and 677 amino acids long, respectively.59 The carboxy-terminal tails of NF-M and NF-H contain both glutamic acid–rich regions and multiple repeats (0 to 5 for NF-M and approximately 50 for NF-H) of the KSP tripeptide. The KSP serine residues are NF-specific phosphorylation sites. Locations of mutations in NFs in human disorders are shown by arrows. ALS ⫽ amyotrophic lateral sclerosis; CMT ⫽ Charcot-Marie-Tooth disease. (Adapted from Al-Chalabi, A., and Miller, C. C. J.: Neurofilaments and neurological disease. Bioessays 24:346, 2003.)

The Control of Axonal Caliber

tail contains multiple repeats of the lysine-serine-proline (KSP) sequence. There are 0 to 7 repeats, depending on the species, for NF-M and approximately 50 repeats for NF-H.30,59,63,65,66,74 There are also head domain serine phosphorylation sites.102 Immunocytochemical and in vitro reassembly studies suggest that the tail domain of NF-M and NF-H is associated with side arms that extend from the NF core.30,31,46,47,60,96 As discussed later, these side arms are likely to influence interfilament spacing.

EXTRINSIC FACTORS THAT INFLUENCE AXONAL CALIBER Myelination: Alteration of the Axonal Cytoskeleton Of the extrinsic factors that have been shown to affect axonal caliber, the extent of Schwann cell ensheathment and myelination is the most important in normal nerves. In cultured sensory neurites ensheathment by Schwann cells results in some enlargement of the axons.121 Myelination provides another and greater increment in caliber. This myelination-mediated effect on axonal size is segmental, affecting only the axon beneath the sheath.58 For example, in the unmyelinated initial segment and stem process of dorsal root ganglion neurons, the axon is smaller. It increases in caliber at the first myelinated heminode

435

(Fig. 20–2). Caliber decreases again at the nodes of Ranvier. Demyelination of one or a few segments of a myelinated fiber is associated with loss of axonal caliber in the demyelinated segments,2,86 as initially recognized by Gombault in the 1870s.37 In individual nerve fibers, the myelinated segments have different cytoskeletal organization than the unmyelinated segments, such as the stem process, nodes of Ranvier, and axon terminals. The greater caliber in myelinated segments correlates with greater spacing of neighboring NFs and with greater phosphorylation of NF-H and NF-M.11,58,131 In the past the predominant hypothesis was that negative charges on the phosphorylated carboxy-terminal tails of NF-H molecules might be repulsive, leading to greater separation of filaments, thereby reducing NF density and hence axonal caliber. Many data seemed to confirm this relationship. Heavy phosphorylation of NF-H is not seen in the neuron cell bodies, the dendrites, the initial segment, or the nodes of Ranvier,106 and in these sites the interfilament distances are less than in myelinated segments.46 These data predict that elimination of the KSP repeats on NF-H would prevent widely spaced NFs. Surprisingly, removal by genetic engineering of the 150 carboxy-terminal amino acids of NF-H, thus removing all of the KSP sites, had no effect on NF spacing, radial axonal growth, ultimate axonal caliber, or conduction velocity.87,88 In contrast, eliminating the seven KSP repeats on the murine NF-M molecule produced partially collapsed neurofilament

FIGURE 20–2 Myelination results in increased phosphorylation of neurofilaments at the first heminode. A, Schematic diagram of a primary sensory neuron showing the perikaryon, the glomerular stem process, the myelinated internode, and the bifurcation into central and peripheral processes. The dotted line describes a plane of section similar to that shown in B, with the perikaryon (P), nonmyelinated stem process (S), and myelinated internode (m) of the same neuron shown in a single section. C, Graph of the ratio of phosphorylated to phosphorylation-independent epitopes in the stem process compared to the myelinated axon. Myelination is associated with a marked increase in the proportion of phosphorylated epitopes. (Modified from Hsieh, S.-T., Kidd, G. J., Crawford, T. O., et al.: Regional modulation of neurofilament organization by myelination in normal axons. J. Neurosci. 14:6392, 1994.)

436

Neurobiology of the Peripheral Nervous System

spacing, slowed growth and reduced calibers, and slower conduction velocities.28 These results appear to exclude simple charge repulsion, because in that model the effect of NF-H phosphorylation would necessarily be greater than that of NF-M.28 Instead, a more rigid side arm may result from phosphorylation of NF-M.28

Signaling Pathways from Myelinating Schwann Cell to Axonal Cytoskeleton The mechanisms by which myelination signals NF phosphorylation involve axonal kinases. Current data suggest that the most important is extracellular signal–related protein kinase 1/2 (ERK1/2). Cdk5, a member of the cyclindependent protein kinase family, can also phosphorylate NF-H and NF-M in vitro. However, the role of Cdk5 in vivo is complex. In mice lacking cdk5 or its activator (p35), the neurofilament proteins, especially NF-M, are hyperphosphorylated, not hypophosphorylated.95 Normally, Cdk/p35 phosphorylates mitogen-activated protein kinase-1 (MEK-1), resulting in a reduction in MEK-1 activity. A recent hypothesis suggests that the myelination-associated signal from Schwann cell to axon may activate the small GTPase Raf, activating MEK-1 and in turn activating ERK1/2, leading to phosphorylation of NFs but also upregulating the Cdk activator p35, resulting in increased p35/Cdk activity and thus feedback inhibition of MEK-1.95 Thus Cdk5 may be more important in limiting than in promoting NF phosphorylation.95 This hypothesis is further complicated by a recent study demonstrating that overexpression of Cdk5 in chick dorsal root ganglion neurons leads to increased NF phosphorylation and reduced axonal transport of NFs.98 Conversely, inhibition of Cdk5 activity reduces NF phosphorylation and enhances NF axonal transport. These findings suggest that Cdk5 activity is tightly linked to NF phosphorylation, transport, and distribution within the axoplasm.

Myelin-Associated Glycoprotein: Involvement in Signaling to the Axon How does the myelin-forming Schwann cell signal to these enzymes? At least part of this signaling involves myelinassociated glycoprotein (MAG). MAG is an intrinsic membrane glycoprotein that is not found in compact myelin, but is found in myelinated internodes. It is localized in several Schwann cell membranes, including the inner and outer mesaxons, the terminal loops, and the site relevant to the present discussion, the adaxonal Schwann cell. MAG is not necessary for myelination,79 but myelinated internodes in MAG⫺/⫺ mice have the smaller caliber, lower NF phosphorylation, and lesser NF spacing that are characteristic of unmyelinated axons.61,132 Thus MAG may represent at least one Schwann cell ligand capable of initiating ERK1/2 activation and thus NF phosphorylation.

The nature of the MAG receptor on the axon has been controversial. The gangliosides GT1b and GD1a appear to interact with MAG,129 and mice lacking complex gangliosides have many of the axonal features of the MAG⫺/⫺ mice, including closer NF spacing.99 However, increasing evidence suggests that there is also a protein receptor, or coreceptor. This might be the p75 low-affinity neurotrophin receptor119,124,128 and the gangliosides may be organized in membrane rafts, so that they cooperate in facilitating MAG signaling and activation of ERK1/2.128 It is likely that Schwann cells also have MAGindependent signaling mechanisms, but the sequence outlined above represents a satisfying if incomplete synthesis of the available data regarding NF spacing. To complete the story of the relationship of myelination to axonal caliber, it is necessary to add that myelinated internodes may contain more NFs than unmyelinated segments. This in turn appears to reflect greater retention of transported NF proteins in myelinated segments. In this aspect of caliber control, phosphorylation of NF-H may play a role; several data suggest that NF-H phosphorylation slows the aggregate transport of NF proteins.1 In contrast, removing the phosphorylation sites of NF-M has no effect on NF transport.87 In sum, an attractive hypothesis is that MAG and perhaps other Schwann cell membrane ligands interact with p75 neurotrophin receptor in ganglioside-containing membrane rafts of the axolemma, initiating the signaling cascade that leads to ERK1/2-mediated phosphorylation of NF-M and NF-H carboxy-terminal domains. NF-M phosphorylation promotes an extension and perhaps a “stiffening” of the NF side arms and thus greater NF spacing. NF-H phosphorylation promotes increased retention of NFs within the myelinated segments, and thus greater NF numbers. Taken together, these processes substantially increase the caliber of myelinated axonal segments. This is reflected in greater conduction velocity. The magnitude of these effects is modeled in Figure 20–3.

DOES AXONAL CALIBER INFLUENCE MYELINATION? The focus so far has been on the effect of myelination on axonal caliber. Conversely, both classic and modern investigators have devoted attention to the effect of axonal caliber on myelination. Two related questions have been the subjects of analyses: to what extent does axonal caliber determine whether Schwann cells initiate myelination of an axon, and to what degree is the amount of myelin formed influenced by axonal caliber? These possibilities are often contrasted with the alternative that specific molecular signals that trigger myelination are present in some axons, and absent in those that do not become myelinated. At the moment much of the available evidence,

The Control of Axonal Caliber

437

FIGURE 20–3 The effects of modification of neurofilament organization by myelination on nerve fiber diameters and conduction velocities. The unmyelinated axon at the top represents the mean diameter of stem processes in L5 dorsal root ganglion of the rat. The open arrow to the left illustrates the consequence of simply adding a myelin sheath to that stem process, with values calculated so that the axonal diameter/fiber diameter ratio (the G ratio) equals 0.6. The conduction velocity (in meters per second) was calculated as the total fiber diameter (axon ⫹ myelin sheath) ⫻ 6.109,120 The solid arrow to the right illustrates the features that were observed by electron microscopic morphometry.58 Note that myelination is associated with increased neurofilament spacing and number, so that the diameter of the axon is 6 ␮m, the fiber diameter is 10 ␮m, and the calculated conduction velocity is 2.7-fold greater than without the modification of the axon by myelination. (Adapted from Hsieh S.-T., Kidd, G. J., Crawford, T. O., et al.: Regional modulation of neurofilament organization by myelination in normal axons. J. Neurosci. 14:6392, 1994.)

reviewed below, can be interpreted in favor of either hypothesis, so that the influence of axonal caliber on myelination should be regarded as an unresolved issue. The onset of myelin formation correlates with radial growth of the axon. This was an early argument in favor of the “caliber” hypothesis—that a critical diameter triggers myelination. This concept has been supported20,118 and critiqued43 (see Chapter 19). Two recent studies examined models in which normally unmyelinated axons were found to myelinate. In the first study, a murine salivary gland was partially denervated by a partial injury of the external carotid branch of the superior cervical ganglion. As a result, the uninjured, normally unmyelinated axons in this branch sprouted to reinnervate denervated salivary gland cells, and the individual unmyelinated axons thereby took on extended fields of innervation.118 The unmyelinated axons in this branch were observed to increase in caliber. An attractive speculation is that this axonal enlargement occurred because the same amount of target-derived growth factors were being retrogradely

transported to fewer nerve cells, so that each neuron had a relative increase in its trophic support. The pivotal finding was that the enlarged unmyelinated axons became segregated into a 1:1 relationship with their Schwann cells, and many subsequently myelinated. This result is consistent with the possibility that increasing axonal caliber may be sufficient to initiate myelin formation. However, it is possible that the same stimulus that induced axonal hypertrophy also changed the neuronal phenotype in other ways, so that a new molecular signal for myelination was expressed. The second study involved exogenous administration of large doses of the growth factor glial cell line–derived neurotrophic factor (GDNF).56 A population of unmyelinated axons, including some C-fiber nociceptor axons, normally express the low-affinity GDNF receptor GFR␣-1 and the high-affinity receptor c-Ret. Giving high daily doses of GDNF resulted in proliferation of Schwann cells (Fig. 20–4), followed by segregation of some unmyelinated axons into 1:1 relationships with Schwann cells and subsequent

438

Neurobiology of the Peripheral Nervous System

FIGURE 20–4 Administration of exogenous glial cell line–derived neurotrophic factor (GDNF) results in proliferation of Schwann cells. Toluidine blue–stained 1-␮m-thick sections of intact sciatic nerves of control vehicle (A) and high-dose recombinant human GDNF (rhGDNF)–injected (B) rats. In rhGDNFinjected animals the myelinated fibers were widely separated and the intervening space contained an increased number of unmyelinated Schwann cells. Examples of Schwann cell nuclei are marked by arrows in B. Mitotic figure (arrow) in a Schwann cell in a rhGDNF-injected animal is shown among other Schwann cell nuclei (asterisks) in the inset (original magnification, ⫻1000). (From Hoke, A., Ho, T., Crawford, T. O., et al.: Glial cell line-derived neurotrophic factor alters axon Schwann cell units and promotes myelination in unmyelinated nerve fibers. J. Neurosci. 23:561, 2003.)

myelination (Fig. 20–5). In GDNF-treated animals there was abundant myelination of very small fibers, and some fibers were clearly undergoing early stages of myelination, with only a few wraps of noncompacted myelin. That these fibers were originally axons of unmyelinated fibers was suggested by their presence within a single continuous basal lamina that also encircled unmyelinated fibers; no such profiles occur normally. Notably, these myelinating profiles occur in intact nerves with neither nerve fiber degeneration nor, more importantly, axonal sprouting. The segregated axons were larger, consistent with the possibility that the process was driven by axonal hypertrophy. However, the same caveat—that GDNF might have altered the neuronal phenotype in more than one fashion—remains. In addition, the possibility of a direct effect of GDNF on Schwann cells cannot be excluded; although the Schwann cells lacked c-Ret, GDNF signaling can occur though GFR␣-1.84,113 Nevertheless, the role of caliber in myelination is very complex. In tissue culture, Windebank et al.121 found that axons increase in size as they segregate into a 1:1 relationship, and there is an additional increase in caliber as they myelinate. Nonmyelinated segments of axons are smaller, have fewer neurofilaments, and have lower levels of phosphorylation of NF-H than myelinated segments of the same axons.58 Demyelination results in loss of caliber and NF-H phosphorylation.11 Does axonal caliber influence the amount of internodal myelin? In a given nerve, the ratio of axonal diameter to total nerve fiber diameter (the outer surface of the myelin sheath) is relatively constant.35 This ratio, termed the G ratio, is typically around 0.6 (see Fig. 20–3). However, this value is not invariant normally,19,102 and it is altered in a variety of pathologic conditions. In remyelinated or regenerated

fibers, it is low (the myelin is relatively thin).18,22,90,93,101 In axonal atrophy the G ratio is higher, and in axonal swelling it is reduced.24,25 It is clear that axonal circumference does not in itself dictate the number of myelin lamellae or the total myelin volume, although it appears to have an influence. Although differing in details, two lines of argument have shown that internodal distance is also an important factor. Thus internodes on regenerated and remyelinated fibers are relatively short, as are internodes near the dorsal root ganglion perikarya and in other areas where longitudinal growth is not as great (e.g., the sciatic nerve).21,23,101 These short internodes have relatively thin sheaths. Thus Smith and colleagues concluded that the determinant of total myelin surface area or myelin volume is that total axolemma surface area covered by the internode.101 There are no data indicating how this measure is read by the myelinating Schwann cell, or how signaling to alter myelin production might occur. The changes in G ratio that occur with atrophy or swelling of axons have been interpreted to reflect “slippage” of myelin lamellae, such that, in swelling, lamellae slide over each other radially to produce a thinner sheath, and in atrophy, the sliding produces a thicker sheath.25 However, the data are better explained by constancy of number of lamellae as the axon changes, with “passive” changes in the G ratio, except in paranodal regions. In these areas lamellae appear to slide over each other longitudinally, so that the sheath gets progressively thin as it approaches the node. This pattern reflects the slippage of the innermost myelin terminal loops into the old internode, so that only the outermost myelin lamellae still reach their original attachment sites next to the node (see Fig. 20–7C later).

The Control of Axonal Caliber

439

FIGURE 20–5 Segregation of axon–Schwann cell units and myelination of unmyelinated axons. A, Electron micrograph of an intact sciatic nerve of a control rat at 4 weeks. An example of an unmyelinated fiber is outlined by arrowheads on the basal lamina. This fiber contains 14 axons (examples identified by “a”; SC ⫽ Schwann cell nucleus). (Original magnification, ⫻5000.) B, Electron micrograph of an intact sciatic nerve of a high-dose GDNF-treated rat at 4 weeks. Note the marked decrease in the axon–Schwann cell ratio and numerous Remak bundles with one or two axons per Schwann cell basal lamina (examples outlined by arrowheads). (Original magnification, ⫻5000.) C, Low-power electron micrograph of the sciatic nerve of GDNF-treated rat shows examples of newly myelinated small axons (original magnification, ⫻3000). D, Some of the small myelinated fibers were still within Remak bundles that also contained unmyelinated axons. Arrowheads point to portions of the basal lamina that are continuous between a myelinated axon and an unmyelinated one. (Original magnification, ⫻25,000.) E, Histograms of the number of axons within a single Schwann cell basal lamina showed a dramatic shift to a higher percentage of Remak bundles attaining a 1:1 relationship when the rats were treated with rhGDNF for 4 weeks (top) compared to controls (bottom). (Adapted from Höke, A., Ho, T., Crawford, T. O., et al.: Glial cell line-derived neurotrophic factor alters axon Schwann cell units and promotes myelination in unmyelinated nerve fibers. J. Neurosci. 23:561, 2003.)

INTRINSIC FACTORS THAT INFLUENCE AXONAL CALIBER Neurofilaments and Control of Axonal Diameter in Large-Caliber Myelinated Nerve Fibers Several lines of evidence indicate that, in large fibers, NF content is closely related to axonal caliber. During radial growth and maturation in normal postnatal development, the composition of the axonal cytoskeleton changes to

include a higher proportion of NFs. In early development axons are thin (⬍1 ␮m in diameter), their cytoskeletons have high proportions of MTs and peripherin filaments but few NFs, and a greater proportion of their cross-sectional area is occupied by membranous elements, including mitochondria and agranular reticulum. The relatively few IFs in these axons are often organized in discrete bundles.6,83,107 Even in adults the cross-sectional areas of unmyelinated axons correlate closely with the content of MTs and membranous elements.107 With radial growth in development, NFs become the predominant cytoskeletal

440

Neurobiology of the Peripheral Nervous System

organelle, outnumbering MTs by more than 10:1 in large axons.55 In large myelinated fibers NFs are the most numerous cytoskeletal elements, and axonal crosssectional area is linearly related to their number.26,27 In large axons MTs and membranous organelles occupy only a small fraction of axonal cross-sectional area.6 The extent to which the content of MTs and membranous organelles correlates with caliber represents a major difference between myelinated and unmyelinated axons. The relationship between NF number and axon crosssectional area could be interpreted to suggest that NF density is fixed and immutable. In fact, even in normal nerves, modest differences in axonal NF density (approximately 25%) have been reported among different populations of neurons85 and in different regions of small-caliber nerve fibers.75,80 During developmental growth, as well as in injury-related atrophy of axons (see later), NF densities stay relatively constant.51,52 In contrast, in some types of axonal swelling, marked increases in NF densities can be found (Fig. 20–6). A number of genetically engineered models have shown that, for altered NF protein expression to alter axonal caliber, the normal stoichiometric relationships between NF-L, NF-H, and NF-M need to be maintained.34,72,123 For example, the content of the NF core protein, NF-L, can be changed substantially without affecting axonal caliber. NF density can increase two- to threefold without compensatory increases in axonal

caliber in transgenic mice overexpressing NF-L alone.72 Because this increase in NF-L occurs in the absence of comparable increases in either NF-M or NF-H, increased NF density in these transgenic axons correlates with reductions in the relative proportions of NF-M and NF-H, which presumably reflect a decrease in the relative number of NF side arms. In mice in which NF entry into the axons is defective, marked reductions in caliber are found,14 and a similar finding is present in Japanese quail deficient in NF proteins.126,127

Neurofilament Transport Because axons lack cytoplasmic ribosomes, most axonal proteins are synthesized in the neuron cell body (soma). However, some pol(A) messenger RNA (mRNA) is present in axons, including message for NF proteins. The total contribution of local axonal synthesis to the total protein economy of the axon has been estimated to be 0.5% to 5%,57 so that perikaryally derived proteins must represent most of the axonal proteins. Pulse-labeled cytoskeletal proteins, including the NF triplet, are slowly transported somatofugally within the axon at rates of several millimeters per day.54 The mechanism of slow transport is incompletely understood, but appears to be MT and kinesin based. The slow aggregate rate represents stationary periods. The state of polymerization of the transported NF proteins remains to be determined.4,78

FIGURE 20–6 Electron micrographs illustrating factors that influence NF density. A, Axoplasm in the sciatic nerve of a normal mouse; this nontransgenic mouse was a littermate of the transgenic mouse whose axoplasm is shown in B. B, Axoplasm from a transgenic mouse overexpressing NF-L. The relative overabundance of NF-L (compared to NF-M and NF-H) correlates with an increase in NF density and greater variability in the orientation of NF. C, An even greater increase in NF density is seen in axons of rat sciatic nerve after systemic intoxication with IDPN. Also note that MT and membranous organelles have become segregated from regions of axoplasm highly enriched in NF. Bars: 0.5 ␮m.

The Control of Axonal Caliber

NF synthesis or the amount of newly synthesized NF protein entering the axon.81 After IDPN intoxication, newly synthesized NFs are unable to move distally after entering the axon. This leads to massive accumulation of NFs in the proximal swellings8–10,39,40 (Fig. 20–7; see also Fig. 20–6). The rapidity with which these axonal swellings form is related to velocity of NF transport,

Alterations in Velocity of Neurofilament Transport: Influence on Axonal Neurofilament Content and Axonal Caliber Systemic intoxication with the neurotoxin ␤,␤⬘iminodipropionitrile (IDPN) severely impairs the transport of NF in axons41 without altering either the level of

A

441

B

FIGURE 20–7 Diagrammatic representation of the cell body and proximal axon of motor neurons in normal (A) and ␤,␤⬘-iminodipropionitrile (IDPN)–intoxicated (B) animals. A, The normal motor neuron is shown giving rise to a thin axon that is myelinated over the first three internodes by oligodendroglia; the last two internodes diagrammed are ensheathed by Schwann cells. The subpial astrocytes (stippled) give rise to a basal lamina, the glial limitans, which is continuous with that of the Schwann cells (dashed line) in the roots. A few pial cells are shown outside the glial limitans. B, In the early stage of IDPN intoxication, the first few internodes undergo fusiform enlargement. The initial segment and the nodes of Ranvier are normal or only moderately enlarged. The internodal axoplasm, shown in cross section, contains a central channel surrounded by whorled NF. The axonal swelling results in attenuation of the internodal myelin sheath and paranodal demyelination. The subpial astrocytes show proliferation of glial filaments. (From Griffin, J. W., Cork, L. C., Hoffman, P. N., and Price, D. L.: Experimental models of motor neuron degeneration. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. [eds.]: Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 621, 1984.) C, Slippage of paranodal myelin lamella in IDPN-induced swelling of the axon. Ventral roots from IDPN-intoxicated animals were teased and immunostained with anti-Caspr antibody. Arrows point to retraction of the terminal myelin loops from the paranodal region. Note the larger size of the ventral roots from IDPN-treated animals demonstrating axonal swelling. (Original magnification ⫻630.) (Courtesy of Thien Nguyen.)

442

Neurobiology of the Peripheral Nervous System

which declines with age. Swellings form more rapidly in young than in old animals. No swellings occur in the NF-deficient Japanese quail when given IDPN,69 and the degree of IDPN-induced axonal swelling is substantially less in the proximal stumps of axotomized nerves than in control nerves (as described later, axotomy reduces the level of NF gene expression). These data suggest that the amount of new NF synthesis by neurons also affects their vulnerability to IDPN. The greater the level of NF gene expression and the greater the velocity of NF transport, the greater the size of the neurofilamentous masses induced by IDPN. Neurofilament-rich axonal swellings, similar to those produced by IDPN intoxication, are also found in a number of other pathologic situations. Examples include the human motor neuron disease amyotrophic lateral sclerosis.7 Structurally similar swellings occur with a different distribution in the heritable disorder giant axonal neuropathy,5 and in the toxic neuropathies caused by 2,5-hexanedione,103,104 carbon disulfide,71 aluminum,33,110 and paclitaxel (Taxol).89 In these disorders the swellings tend to begin in distal regions of the nerves.103,104 In experimental 2,5-hexanedione neuropathy the swellings have been correlated with a speeding of NF transport.70,71

Neurofilament Gene Expression Influences Axonal Caliber The level of NF gene expression in mature neurons correlates with axonal NF content and axonal caliber. Similarly, small spinal sensory neurons contain low levels of NF mRNAs and give rise to unmyelinated axons.50 In contrast, neurons that contain relatively high levels of NF mRNAs (e.g., large spinal sensory neurons) give rise to largecaliber myelinated nerve fibers, which are highly enriched in NF.50 Given the correlation between the level of NF gene expression and axonal caliber in mature neurons, it is not surprising that postnatal increases in NF gene expression correlate closely with the radial growth and myelination of developing axons.49,73,91 In developing sensory neurons, postnatal increases in NF gene expression also correlate with increases in perikaryal size.73 Studies of NF gene expression in axotomized neurons also indicate that axonal caliber correlates closely with the level of NF gene expression. Selective reductions in the abundance of NF mRNAs in axotomized neurons50 correlate with declines in (1) NF synthesis (as reflected by reduced amino incorporation),38 (2) the amount of pulselabeled NF protein undergoing axonal transport,55 (3) axonal NF content,52 and (4) axonal caliber in the proximal stumps of transected nerve fibers.52 These reductions in caliber start proximally near the cell body (soma) and proceed somatofugally along nerve fibers at a rate equal to the velocity of NF transport, a process referred to as somatofugal axonal atrophy.48,52,53 Thus alterations in NF

gene expression have a profound influence on both axonal NF content and axonal caliber. Similar reductions in NF content also occur in the axonal atrophies associated with hereditary motor and sensory neuropathy type I in humans76 and streptozotocin-induced diabetic neuropathy in experimental animals.125 It should be noted that axotomy-induced reductions in axonal caliber are confined primarily to large-caliber myelinated nerve fibers (e.g., alpha motor fibers). There is little, if any, change after axotomy in the calibers of relatively small myelinated axons (e.g., gamma motor fibers).52 In fact, substantial changes in axonal NF content are associated with only modest changes in the crosssectional areas of small-caliber myelinated axons such as those in the optic nerve,75,80 indicating that factors other than NF content play important roles in the control of axonal caliber in these nerve fibers. Neurofilaments, class IV IF proteins, appear to play a unique role in the control of axonal caliber. The temporal expression and distribution of peripherin, a class III IF protein expressed in neurons,17,32,82,105 suggest that its role is different from that of NF. In contrast to NF, peripherin is expressed at relatively high levels (in comparison to those in mature neurons) during early development94,111 and axon regeneration.77,112 These findings indicate that the radial growth of sensory axons during postnatal development correlates with a decline in the abundance of peripherin in large sensory neurons, and the somatofugal atrophy of regenerating axons is associated with increased expression of peripherin. Furthermore, unlike NF, which is expressed at highest levels in large sensory neurons, peripherin is expressed preferentially in small sensory neurons.13,111

Regulation of Neurofilament Gene Expression Although the factors that regulate NF gene expression are largely unknown, the available evidence suggests that interactions with target cells may play an important role in this process. Disconnection of neurons from their targets by axotomy results in downregulation of NF gene expression. Reconnection of regenerating axons with targets appears to be a prerequisite for the return of gene expression to preaxotomy levels.50,52 These observations have led to the suggestion that tropic molecules derived from target cells and retrogradely transported to the neuron cell body may play an important role in this process.50 Although the ability of exogenous nerve growth factor (NGF) to induce NF expression in sensory neurons with high-affinity NGF receptors is controversial,117,122 exogenous NGF does appear to prevent, at least in part, axotomy-induced reductions in the calibers of sensory axons.36 A role for target cells in the control of NF expression may also be inferred from the observation that the radial growth and subsequent myelination of sympathetic axons (which are

The Control of Axonal Caliber

not normally myelinated) can be stimulated by forcing these neurons to innervate an increased number of target cells.118 14.

CONCLUSIONS There is an ongoing dialogue between Schwann cells and the neurofilamentous cytoskeleton. The state of myelination of the axon by the Schwann cell influences NF spacing and numbers, and thus axonal caliber. Conversely, some evidence supports the concept that the radial growth of axons is at least one stimulus for myelin formation, and that axonal caliber may be one influence, along with internodal length, on the volume of myelin formed.

REFERENCES 1. Ackerley, S., Thornhill, P., Grierson, A. J., et al.: Neurofilament heavy chain side arm phosphorylation regulates axonal transport of neurofilaments. J. Cell Biol. 161:489, 2003. 2. Aguayo, A. J., Attiwell, M., Trecarten, J., et al.: Abnormal myelination in transplanted trembler mouse Schwann cells. Nature 265:73, 1977. 3. Al-Chalabi, A., and Miller, C. C. J.: Neurofilaments and neurological disease. Bioessays 24:346, 2003. 4. Angelides, K. J., Smith, K. E., and Takeda, M.: Assembly and exchange of intermediate filament proteins of neurons: neurofilaments are dynamic structures. J. Cell Biol. 108:1495, 1989. 5. Asbury, A. K., Gale, M. K., Cox, S. C., et al.: Giant axonal neuropathy: a unique case with segmental neurofilamentous masses. Acta Neuropathol. 20:237, 1972. 6. Berthold, C. H.: Morphometry of normal peripheral axons. In Waxman, S. G. (ed.): Physiology and Pathobiology of Axons. New York, Raven Press, p. 3, 1978. 7. Carpenter, S.: Proximal axonal enlargement in motor neuron disease. Neurology 18:841, 1968. 8. Chou, S.-M., and Hartman, H. A.: Axonal lesions and waltzing syndrome after IDPN administration in rats: with a concept—“axostasis.” Acta Neuropathol. 3:428, 1964. 9. Chou, S.-M., and Hartman, H. A.: Electron microscopy of focal neuroaxonal lesions produced by ␤,␤⬘-iminodipropionitrile (IDPN) in rats. Acta Neuropathol. 39:590, 1965. 10. Clark, A. W., Griffin, J. W., and Price, D. L.: The axonal pathology in chronic IDPN intoxication. J. Neuropathol. Exp. Neurol. 39:42, 1980. 11. deWaegh, S. M., Lee, V. M. Y., and Brady, S. T.: Local modulation of neurofilament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68:451, 1992. 12. Ellisman, M. H., and Porter, K. R.: Microtrabecular structure of the axoplasmic matrix: visualization of cross-linking structures and their distribution. J. Cell Biol. 87:464, 1980. 13. Escurat, M., Djabali, K., Gumpel, M., et al.: Differential expression of two neuronal intermediate filament proteins, peripherin and the low-molecular-mass neurofilament

15. 16.

17.

18.

19.

20.

21.

22.

23. 24.

25.

26.

27.

28.

29.

30.

31.

443

protein (NF-L), during the development of rat. J. Neurosci. 10:764, 1990. Eyer, J., and Peterson, A.: Neurofilament-deficient axons and perikaryal aggregates in viable transgenic mice expressing a neurofilament–␤-galactosidase fusion protein. Neuron 12:389, 1994. Farel, P. B.: Reflex activity of regenerating frog spinal motorneurons. Brain Res. 158:331, 1978. Fath, K. R., and Lasek, R. J.: Two classes of actin microfilaments are associated with the inner cytoskeleton of axons. J. Cell Biol. 107:613, 1988. Franke, W. W.: Nuclear lamins and cytoskeletal intermediate filament proteins: a growing multigene family. Cell 48:3, 1987. Fried, K., Hildebrand, C., and Eldelyi, G.: Myelin thickness and internodal length of nerve fibers in developing feline inferior alveolar nerve. J. Neurol. Sci. 54:47, 1982. Friede, R., Meier, T., and Diem, M.: How is the exact length of an internode determined? J. Neurol. Sci. 50:217, 1981. Friede, R. L.: Control of myelin formation by axonal caliber with a model of the control mechanism. J. Comp. Neurol. 144:233, 1972. Friede, R. L.: Variance in relative internodal length (l/d) in the rat and its presumed significance for the safety factor and neuropathy. J. Neurol. Sci. 60:89, 1983. Friede, R. L., and Beuche, W.: A new approach toward analyzing peripheral nerve fiber populations. I. Variance in sheath thickness corresponds to different geometric proportions of the internodes. J. Neuropathol. Exp. Neurol. 44:60, 1985. Friede, R. L., and Bischhausen, R.: The precise geometry of large internodes. J. Neurol. Sci. 48:367, 1980. Friede, R. L., and Martinez, A. J.: Analysis of axon-sheath relations during early wallerian degeneration. Brain Res. 19:199, 1970. Friede, R. L., and Martinez, A. J.: Analysis of the process of sheath expansion in swollen nerve fibers. Brain Res. 19:165, 1970. Friede, R. L., and Miyaghishi, T.: Axon caliber, neurofilaments, microtubules, sheath thickness and cholesterol in cat optic nerve fiber. Anat. Rec. 108:365, 1971. Friede, R. L., and Samorajski, T.: Axonal caliber related to neurofilaments and microtubules in sciatic nerve fibers of rats and mice. Anat. Rec. 167:379, 1971. Garcia, M. L., Lobsiger, C. S., Shah, S. B., et al.: NF-M is an essential target for the myelin-directed “outside-in” signaling cascade that mediates radial axonal growth. J. Cell Biol. 163:1011, 2003. Gasser, H. S., and Grundfest, H.: Axon diameters in relation to the spike dimensions and the conduction velocity in mammalian A fibers. Am. J. Physiol. 127:393, 1939. Geisler, N., Kaufmann, E., Fischer, S., et al.: Neurofilament architecture combines structural principles of intermediate filaments with carboxy-terminal extensions increasing in size between triplet proteins. EMBO J. 2:1295, 1983. Geisler, N., and Weber, K.: Self-assembly in vitro of the 68,000 molecular weight component of the mammalian neurofilament triplet proteins into intermediate-size filaments. J. Mol. Biol. 151:565, 1981.

444

Neurobiology of the Peripheral Nervous System

32. Geisler, N., and Weber, K.: Structural aspects of intermediate filaments. In Shay, J. W. (ed.): Cell and Molecular Biology of the Cytoskeleton. New York, Plenum Publishing, p. 41, 1986. 33. Gilbert, M. R., Harding, B. L., Hoffman, P. N., et al.: Aluminum-induced neurofilamentous changes in cultured rat dorsal root ganglia explants. J. Neurosci. 12:1763, 1992. 34. Gill, S. R., Wong, P. C., Monteiro, M. J., et al.: Assembly properties of dominant and recessive mutations in the small mouse neurofilament (NF-L) subunit. J. Cell Biol. 111:2005, 1990. 35. Gillespie, M. J., and Stein, R. B.: The relationship between axon diameter, myelin thickness and conduction velocity during atrophy of mammalian peripheral nerves. Brain Res. 259:41, 1983. 36. Gold, B. G., Mobley, W. C., and Matheson, S. F.: Regulation of axonal caliber, neurofilament content, and nuclear localization in mature sensory neurons by nerve growth factor. J. Neurosci. 11:943, 1991. 37. Gombault, F. A. A.: Contribution a l’histoire anatomique de l’atrophie musculaire saturnine. Arch. Physiol. V:592, 1873. 38. Greenberg, S. G., and Lasek, R. J.: Neurofilament protein synthesis in DRG neurons decreases more after peripheral axotomy than after central axotomy. J. Neurosci. 8:1739, 1988. 39. Griffin, J. W., Cork, L. C., Hoffman, P. N., et al.: Experimental models of motor neuron degeneration. In Dyck, P. J., Thomas, P. K., Lambert, E. H., et al. (eds.): Peripheral Neuropathy, 3rd ed., Vol. I. Philadelphia, W. B. Saunders, p. 621, 1984. 40. Griffin, J. W., Gold, B. G., Cork, L. C., et al.: IDPN neuropathy in the cat: coexistence of proximal and distal axonal swellings. Neuropathol. Appl. Neurobiol. 8:351, 1982. 41. Griffin, J. W., Hoffman, P. N., Clark, A. W., et al.: Slow axonal transport of neurofilament proteins: impairment by ␤,␤⬘-iminodipropionitrile administration. Science 202:633, 1978. 42. Griffin, J. W., and Price, D. L.: Demyelination in experimental IDPN and hexacarbon neuropathies: evidence for an axonal influence. Lab. Invest. 45:130, 1981. 43. Griffin, J. W., Rosenfeld, J., Hoffman, P. N., et al.: The axonal cytoskeleton: influences on nerve fiber form and Schwann cell behavior. In Lasek, R. (ed.): Intrinsic Determinants of Neuronal Form and Function. New York, Alan R. Liss, p. 403, 1988. 44. Henneman, E., Somjen, G., and Carpenter, D. O.: Excitability and inhibitability of motoneurons of different sizes. J. Neurophysiol. 28:599, 1965. 45. Hirokawa, N.: Cross-linker system between neurofilaments, microtubules, and membranous organelles in frog axons revealed by the quick-freeze, deep-etching method. J. Cell Biol. 94:129, 1982. 46. Hirokawa, N., Glicksman, M. A., and Willard, M. B.: Organization of mammalian neurofilament polypeptides within the neuronal cytoskeleton. J. Cell Biol. 98:1523, 1984. 47. Hisanaga, S.-I., and Hirokawa, N.: Structure of the peripheral domains of neurofilaments revealed by low angle rotary shadowing. J. Mol. Biol. 202:297, 1988. 48. Hoffman, P. N.: Distinct roles of neurofilament and tubulin gene expression in axonal growth. Ciba Found. Symp. 138:192, 1988.

49. Hoffman, P. N.: Expression of GAP-43, a rapidly transported protein, and class II beta tubulin, a slowly transported protein, are coordinated in regenerating neurons. J. Neurosci. 9:893, 1989. 50. Hoffman, P. N., Cleveland, D. W., Griffin, J. W., et al.: Neurofilament gene expression: a major determinant of axonal caliber. Proc. Natl. Acad. Sci. U. S. A. 84:3472, 1987. 51. Hoffman, P. N., Griffin, J. W., Cold, B. G., et al.: Slowing of neurofilament transport and the radial growth of developing nerve fibers. J. Neurosci. 5:2920, 1985. 52. Hoffman, P. N., Griffin, J. W., and Price, D. L.: Control of axonal caliber by neurofilament transport. J. Cell Biol. 99:705, 1984. 53. Hoffman, P. N., Koo, E. H., Muma, N. A., et al.: Role of neurofilaments in the control of axonal caliber in myelinated nerve fibers. In Lasek, R. J., and Black, M. M. (eds.): Intrinsic Determinants of Neuronal Form and Function. New York, Alan R. Liss, p. 389, 1988. 54. Hoffman, P. N., and Lasek, R. J.: The slow component of axonal transport: identification of major structural polypeptides of the axon and their generality among mammalian neurons. J. Cell Biol. 66:351, 1975. 55. Hoffman, P. N., Thompson, G. W., Griffin, J. W., et al.: Changes in neurofilament transport coincide temporally with alterations in the caliber of axons in regenerating motor fibers. J. Cell Biol. 101:1332, 1985. 56. Hoke, A., Ho, T., Crawford, T. O., et al.: Glial cell linederived neurotrophic factor alters axon Schwann cell units and promotes myelination in unmyelinated nerve fibers. J. Neurosci. 23:561, 2003. 57. Hollenbeck, P. J., and Hollenbeck, P. J.: Organization and translation of mRNA in sympathetic axons. J. Cell Sci. 116:4467, 2003. 58. Hsieh, S.-T., Kidd, G. J., Crawford, T. O., et al.: Regional modulation of neurofilament organization by myelination in normal axons. J. Neurosci. 14:6392, 1994. 59. Julien, J.-P., Cote, F., Beaudet, L., et al.: Sequence and structure of the mouse gene coding for the largest neurofilament subunit. Gene 68:307, 1988. 60. Julien, J.-P., and Mushynski, W. E.: The distribution of phosphorylation sites among identified proteolytic fragments of mammalian neurofilaments. J. Biol. Chem. 258:4019, 1983. 61. Kumar, S., Yin, X., Trapp, B. D., et al.: Role of long-range repulsive forces in organizing axonal neurofilament distributions: evidence from mice deficient in myelin-associated glycoprotein. J. Neurosci. Res. 68:681, 2002. 62. Lasek, R. J.: Studying the intrinsic determinants of neuronal form and function. In Lasek, R. J. (ed.): Intrinsic Determinants of Neuronal Form and Function. New York, Alan R. Liss, 1988. 63. Lees, J. F., Shneidman, P. S., Skuntz, S. F., et al.: The structure and organization of the human heavy neurofilament subunit (NF-H) and the gene encoding it. EMBO J. 7:1947, 1988. 64. Letourneau, P.: Differences in the organization of actin in the growth cones compared with the neurites of cultured neurons from chick embryos. J. Cell Biol. 97:963, 1983. 65. Levy, E., Liem, R. K. H., D’Eustachio, P., et al.: Structure and evolutionary origin of the gene encoding mouse NF-M,

The Control of Axonal Caliber

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

the middle-molecular-mass neurofilament protein. Eur. J. Biochem. 166:71, 1987. Lewis, S. A., and Cowan, N. J.: Anomalous placement of introns in a member of the intermediate filament multigene family: an evolutionary conundrum. Mol. Cell Biol. 6:1529, 1986. Liem, R. K. H., Yen, S.-H., Salomon, G. D., et al.: Intermediate filaments in nervous tissue. J. Cell Biol. 79:637, 1978. Metuzals, J., and Tasaki, I.: Subaxolemmal filamentous network in the giant nerve fiber of the squid (Loligo pealei) and its possible role in excitability. J. Cell Biol. 78:597, 1978. Mitsuishi, K., Takahashi, A., Mizutani, M., et al.: ␤,␤⬘Iminodipropionitrile toxicity in normal and congenitally neurofilament-deficient Japanese quails. Acta Neuropathol. (Berl.) 86:578, 1993. Monaco, S., Autilio-Gambetti, L., Zabel, D., et al.: Giant axonal neuropathy: acceleration of neurofilament transport in optic axons. Proc. Natl. Acad. Sci. U. S. A. 82:920, 1985. Monaco, S., Jacob, J., Jenich, H., et al.: Axonal transport of neurofilament is accelerated in peripheral nerve during 2,5-hexanedione intoxication. Brain Res. 491:328, 1989. Monteiro, M. J., Hoffman, P. N., Gearhart, J. D., et al.: Expression of NF-1 in both neuronal and nonneuronal cells of transgenic mice: increased neurofilament density in axons without affecting caliber. J. Cell Biol. 111:1543, 1990. Muma, N. A., Slunt, H. H., and Hoffman, P. N.: Postnatal increase in neurofilament gene expression and the radial growth of axons. J. Neurocytol. 20:844, 1991. Myers, M. W., Lazzarini, R. A., Lee, V. M. Y., et al.: The human mid-size neurofilament subunit: a repeated protein sequence and the relationship of its gene to the intermediate filament gene family. EMBO J. 6:1617, 1987. Nixon, R. A., and Logvinenko, K. B.: Multiple fates of newly synthesized neurofilament proteins: evidence for a stationary neurofilament network distributed nonuniformly along axons of retinal ganglion cell neurons. J. Cell Biol. 102:647, 1986. Nukada, H., and Dyck, P. J.: Decreased axonal caliber and neurofilaments in hereditary motor and sensory neuropathy, type I. Ann. Neurol. 16:238, 1984. Oblinger, M. M., Wong, J., and Parysek, L. M.: Axotomyinduced changes in the expression of a type III neuronal intermediate filament. J. Neurosci. 9:3766, 1989. Ochs, S., Jersild, A. Jr., and Li, J.-M.: Slow transport of freely movable cytoskeletal components shown by beading partition of nerve fibers in the cat. Neuroscience 33:421, 1989. Owens, G. C., and Bunge, R. P.: Schwann cells infected with a recombinant retrovirus expressing myelin-associated glycoprotein antisense RNA do not form myelin. Neuron 7:565, 1991. Parhad, I. M., Clark, A. W., and Griffin, J. W.: The effect of changes in neurofilament content on caliber of axons: the ␤,␤⬘-iminodipropionitrile model. J. Neurosci. 7:2256, 1987. Parhad, I. M., Swedberg, E. A., Hoar, D. I., et al.: Neurofilament gene expression following ␤,␤⬘-iminodipropionitrile (IDPN) intoxication. Mol. Brain Res. 4:293, 1988. Parysek, L. M., and Goldman, R. D.: Distribution of a novel 57 kDa intermediate filament (IF) protein in the nervous system. J. Neurosci. 8:555, 1988.

445

83. Peters, A., and Vaughn, J. E.: Microtubules and filaments in the axons and astrocytes of early postnatal rat optic nerves. J. Cell Biol. 32:113, 1967. 84. Poteryaev, D., Titievsky, A., Sun, Y. F., et al.: GDNF triggers a novel ret-independent Src kinase family-coupled signaling via a GPI-linked GDNF receptor alpha1. FEBS Lett. 463:63, 1999. 85. Price, R. L., Paggi, P., Lasek, R. J., et al.: Neurofilaments are spaced randomly in the radial dimensions of axons. J. Neurocytol. 17:55, 1988. 86. Raine, C. S., Wisniewski, H., and Prineas, J.: An ultrastructural study of experimental demyelination and remyelination. Lab. Invest. 21:316, 1969. 87. Rao, M. V., Campbell, J., Yuan, A., et al.: The neurofilament middle molecular mass subunit carboxyl-terminal tail domain is essential for the radial growth and cytoskeletal architecture of axons but not for regulating neurofilament transport rate. J. Cell Biol. 163:1021, 2003. 88. Rao, M. V., Garcia, M. L., Miyazaki, Y., et al.: Gene replacement in mice reveals that the heavily phosphorylated tail of neurofilament heavy subunit does not affect axonal caliber or the transit of cargoes in slow axonal transport. J. Cell Biol. 158:681, 2002. 89. Roytta, M., Horwitz, S. B., and Raine, C. S.: Taxol-induced neuropathy: short-term effects of local injection. J. Neurocytol. 13:685, 1984. 90. Sanders, F. K.: The thickness of myelin sheaths of normal and regenerating peripheral nerve fibers. Proc. Natl. Acad. Sci. Lond. 135:323, 1948. 91. Schlaepfer, W. W., and Bruce, J.: Simultaneous up-regulation of neurofilament proteins during the postnatal development of the rat nervous system. J. Neurosci. 25:39, 1990. 92. Schlaepfer, W. W., and Freeman, L. A.: Neurofilament proteins of rat peripheral nerve and spinal cord. J. Cell Biol. 78:653, 1978. 93. Schroder, J. M.: Alteration between axonal diameter and myelin sheath thickness and regenerated fibers. Brain Res. 35:49, 1972. 94. Seymour, M. A., and Oblinger, M. M.: Differential regulation of type III and IV intermediate filament genes and tubulin genes in developing hamster brain. J. Cell Biol. 107:462a, 1988. 95. Sharma, P., Veeranna, Sharma, M., et al.: Phosphorylation of MEK1 by cdk5/p35 down-regulates the mitogen-activated protein kinase pathway. J. Biol. Chem. 277:528, 2002. 96. Sharp, G. A., Shaw, G., and Weber, K.: Immunoelectronmicroscopical localization of the three neurofilament triplet proteins along neurofilaments of cultured dorsal root ganglion cell neurones. Exp. Cell Res. 137:403, 1982. 97. Shaw, G., Osborn, M., and Weber, K.: An immunofluorescence microscopical study of neurofilament triplet proteins, vimentin, and glial fibrillary acidic protein within the adult rat brain. J. Cell Biol. 26:68, 1981. 98. Shea, T. B., Yabe, J. T., Ortiz, D., et al.: Cdk5 regulates axonal transport and phosphorylation of neurofilaments in cultured neurons. J. Cell Sci. 117:933, 2004. 99. Sheikh, K. A., Sun, J. L. Y., Kawai, H., et al.: Mice lacking complex gangliosides develop wallerian degeneration and myelination defects. Proc. Natl. Acad. Sci. U. S. A. 96:7532, 1999.

446

Neurobiology of the Peripheral Nervous System

100. Sheng, J. G., Zhu, S. G., Jones, R. A., et al.: Interleukin-1 promotes expression and phosphorylation of neurofilament and tau protein in vivo. Exp. Neurol. 163:388, 2000. 101. Smith, K. J., Blakemore, W. F., Murray, J. A., et al.: Internodal myelin volume and axon surface area: a relationship determining myelin thickness? J. Neurol. Sci. 55:231, 1982. 102. Spencer, P. S., Raine, C. S., and Wisniewski, H.: Axon diameter and myelin thickness: unusual relationships in dorsal root ganglia. Anat. Rec. 176:225, 1973. 103. Spencer, P. S., and Schaumburg, H. H.: Central-peripheral distal axonopathy—the pathogenesis of dying-back polyneuropathies. Prog. Neuropathol. 3:253, 1976. 104. Spencer, P. S., and Schaumburg, H. H.: Ultrastructural studies of the dying back process. III. The evolution of experimental peripheral giant axonal degeneration. J. Neuropathol. Exp. Neurol. 36:276, 1977. 105. Steinert, P. M., and Roop, D. R.: Molecular and cellular biology of intermediate filaments. Annu. Rev. Biochem. 57:593, 1988. 106. Sternberger, L. A., and Sternberger, N. H.: Monoclonal antibodies distinguish phosphorylated and nonphosphorylated forms of neurofilaments in situ. Proc. Natl. Acad. Sci. U. S. A. 80:6126, 1983. 107. Stevens, J. K., Trogadis, J., and Jacobs, J. R.: Development and control of axial neurite form: a serial electron microscopic analysis. In Lasek, R. J., and Black, M. M. (eds.): Intrinsic Determinants of Neuronal Form and Function. New York, Alan R. Liss, p. 115, 1988. 108. Tapscott, S. J., Bennett, G. S., Toyama, Y., et al.: Intermediate filament proteins in the developing chick spinal cord. Dev. Biol. 86:40, 1981. 109. Thomas, P. K., Barthold, C.-H., and Ochoa, J.: Microscopic anatomy of the peripheral nervous system. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 28, 1993. 110. Troncoso, J. C., Price, D. L., Griffin, J. W., et al.: Neurofibrillary axonal pathology in aluminum intoxication. Ann. Neurol. 12:278, 1982. 111. Troy, C. M., Brown, K., Greene, L. A., et al.: Ontogeny of the neuronal intermediate filament protein, peripherin, in the mouse embryo. J. Neurosci. 36:217, 1990. 112. Troy, C. M., Muma, N. A., Greene, L. A., et al.: Contrasting regulation of peripherin and neurofilament expression in motor neuron regeneration. Brain Res. 529:232, 1990. 113. Trupp, M., Scott, R., Whittemore, S. R., et al.: Ret-dependent and -independent mechanisms of glial cell line-derived neurotrophic factor signaling in neuronal cells. J. Biol. Chem. 274:20885, 1999. 114. Tsukita, S., and Ishikawa, H.: The cytoskeleton in myelinated axons: serial section study. Biomed. Res. 2:424, 1981. 115. Tsukita, S., Tsukita, S., Kobayashi, T., and Matsumoto, G.: Subaxolemmal cytoskeleton in squid giant axon. II. Morphological identification of microtubule- and microfilament-associated domains of axolemma. J. Cell Biol. 102:1710, 1986.

116. Tsukita, S., Usukura, J., and Ishikawa, H.: The cytoskeleton in myelinated axons: a freeze-etch replica study. Neuroscience 7:2135, 1982. 117. Verge, V. M. K., Tetzlaff, W., Bisby, M. A., et al.: Influence of nerve growth factor on neurofilament gene expression in mature primary sensory neurons. J. Neurosci. 10:2018, 1990. 118. Voyvodic, J. T.: Target size regulates calibre and myelination of sympathetic axons. Nature 342:430, 1989. 119. Wang, K. C., Kim, J. A., Sivasankaran, R., et al.: p75 interacts with the Nogo receptor as a co-receptor for Nogo, MAG and OMgp. Nature 420:748, 2002. 120. Waxman, S. G.: Determinants of conduction velocity in myelinated nerve fiber. Muscle Nerve 3:141, 1980. 121. Windebank, A. J., Word, P., Bunge, R. P., et al.: Myelination determines the caliber of dorsal root ganglion neurons in culture. J. Neurosci. 6:1563, 1985. 122. Wong, J., and Oblinger, M. M.: NGF rescues substance P expression but not neurofilament or tubulin gene expression in axotomized sensory neurons. J. Neurosci. 11:543, 1991. 123. Wong, P. C., and Cleveland, D. W.: Characterization of dominant and recessive assembly-defective mutations in mouse neurofilament NF-M. J. Cell Biol. 111:1987, 1990. 124. Wong, S. T., Henley, J. R., Kanning, K. C., et al.: A p75(NTR) and Nogo receptor complex mediates repulsive signaling by myelin-associated glycoprotein. Nat. Neurosci. 5:1302, 2002. 125. Yagihashi, S., Kamijo, M., and Watanabe, K.: Reduced myelinated fiber size correlates with loss of axonal neurofilaments in peripheral nerve of chronically streptozotocin diabetic rats. Am. J. Pathol. 136:1365, 1990. 126. Yamasaki, H., Bennett, G. S., Itakura, C., et al.: Defective expression of neurofilament protein subunits in hereditary hypotrophic axonopathy of quail. Lab. Invest. 66:734, 1992. 127. Yamasaki, H., Itakura, C., and Mizutani, M.: Hereditary hypotrophic axonopathy with neurofilament deficiency in a mutant strain of Japanese quail. Acta Neuropathol. (Berl.) 82:427, 1991. 128. Yamashita, T., Higuchi, H., and Tohyama, M.: The p75 receptor transduces the signal from myelin-associated glycoprotein to Rho. J. Cell Biol. 157:565, 2002. 129. Yang, L. J. S., Zeller, C. B., Shaper, N. L., et al.: Gangliosides are neuronal ligands for myelin-associated glycoprotein. Proc. Natl. Acad. Sci. U. S. A. 93:814, 1996. 130. Yen, S.-H., and Fields, K. L.: Antibodies to neurofilament, glial filament and fibroblast intermediate filament proteins bind to different cell types in the nervous system. J. Cell Biol. 88:115, 1981. 131. Yeung, K. B., Thomas, P. K., King, R. H. M., et al.: The clinical spectrum of peripheral neuropathies associated with benign monoclonal IgM, IgG and IgA paraproteinemia: comparative clinical, immunological and nerve biopsy findings. J. Neurol. 238:383, 1991. 132. Yin, X., Crawford, T. O., Griffin, J. W., et al.: Myelinassociated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J. Neurosci. 18:1953, 1998.

21 Guidance of Axons to Targets in Development and in Disease HARALD WITTE AND FRANK BRADKE

Growth Cones and Cytoskeletal Dynamics Neurite Outgrowth and Steering Actin Dynamics Axon Guidance in Development Ephrin/Eph

Slit/Robo Netrin/DCC/UNC-5 Semaphorin/Plexin Axon Guidance in Disease Epilepsy

Neurons must connect with the proper partners to establish functional circuits. How do neuronal axons grow to their specific targets? Ramon y Cajal and Speidel proposed one of the first theories to explain the amazing wiring capacity of the nervous system,294,351 suggesting that axonal growth was directional and that axons were specifically growing on their routes to their specific target. Sperry’s later experiments provided support for this hypothesis, showing that regenerating retinal ganglion cells correctly target their axons to the optic tectum.352 These experiments led Sperry to formulate his “chemoaffinity theory” postulating that specific ligands and receptors would allow the axons to find and connect to the appropriate target. The molecular identity of these signals, however, remained elusive until recently. With the identification of four main classes of axon guidance ligand-receptor systems (ephrin/Eph, Slit/Robo, netrin/DCC/UNC-5, and semaphorin/plexin), the last 15 years have seen explosive progress in elucidating the mechanisms of wiring within the nervous system. Here we describe these systems in detail, giving special attention to the signaling cascades and molecular downstream targets associated with them. For a better understanding of the signaling events involved, we start the chapter with a general description of the underlying cell biologic mechanisms regulating axonal growth. Recent studies have also revealed that molecules involved in axon guidance—either at the level of ligand-receptor interactions or in the downstream signaling—account for

Schizophrenia CRASH Syndrome/MASA Syndrome/ L1 Syndrome Kallmann’s Syndrome Outlook

various diseases. Progress in unveiling the mechanisms behind these syndromes is discussed in the last part of the chapter.

GROWTH CONES AND CYTOSKELETAL DYNAMICS In this section we introduce the reader to the growth cone, its steering mechanisms, and the molecules involved in axonal growth or halt. Many of the molecular mechanisms underlying cellular locomotion have been elucidated using the migrating fibroblast system. Migrating fibroblasts have also been studied through the years as a more accessible alternative to neuronal growth cones for understanding axonal growth and locomotion. Despite the many differences between these two systems, it has been astonishing to find many similarities on a molecular level. Hence, to allow a better overall view, we include current research development on actin dynamics in migrating fibroblasts in this section.

Neurite Outgrowth and Steering The Growth Cone: The Steering Apparatus of the Axon A unique structure at the leading edge of growing axons is the growth cone, first described by Ramon y Cajal in 1890.294 Growth cones are equipped with a large set of receptors to react to a multitude of guidance signals. These guidance cues are integrated in growth cones, allowing the 447

448

Neurobiology of the Peripheral Nervous System

shaft into the central domain. The peripheral domain contains very few microtubules but instead has a dense network of cross-linked, dynamic actin filaments. The actin filaments of the periphery form filopodia, thin membrane extensions emerging from the leading edge of the growth cone, and lamellipodia, fan-shaped structures. Filopodia and lamellipodia explore the extracellular environment (see Fig. 21–1A) with dynamic advance and retraction movements, and mediate guidance responses to external cues.76 Upon depolymerization of the actin filaments in filopodia and lamellipodia, axon growth is disoriented.24,61 Apart from their role in guiding the axons, actin organization and dynamics appear to play a major role in restraining microtubules from entering the peripheral zone. The dense actin meshwork excludes particles with a diameter of 24 nm from the peripheral zone, including organelles and macromolecules.229 Thus microtubules may be

growing axon to correctly navigate to its ultimate target over long distances. Growth cones contain the machinery required for both translation and protein degradation48 and are able to adjust the receptor composition on their surface according to the local environment. Guidance cues can be attractive or repulsive; that is, an axon will grow toward or away from the source of the guidance signal, respectively. Guidance cues act either by direct contact (e.g., between the axonal growth cone and a repelling environment) or as diffusible factors that are secreted by the target region. Thus they act as short-range or long-range cues. Growth Cone Cytoskeleton Dynamics A growth cone consists of three distinct domains: the central domain, the peripheral domain, and the transition zone in between these two108,178 (Fig. 21–1A). Bundles of microtubules and organelles emanate from the axonal B

A

C

FIGURE 21–1 Growth cone structure and function. A, The central domain of the growth cone is rich in organelles and stable microtubules, while the peripheral domain is characterized by a dense network of cross-linked, highly mobile actin filaments and only few, individual microtubules (for clarity, only a few actin filaments are shown). Actin filaments and microtubules are subject to disassembly and turnover in the transition zone. Filopodia and lamellipodia emanating from the peripheral domain probe the environment of the growth cone and mediate the response to guidance cues. B, Actin dynamics cause highly dynamic advance and retraction of filopodia. G-actin molecules add to the distally oriented plus (⫹) ends (barbed ends) of preexisting actin fibers. These fibers undergo myosin-dependent retrograde transport. Attractive guidance cues stimulate F-actin assembly at the tips of filopodia and reduce the rate of retrograde F-actin flow, while repulsive guidance cues act conversely. C, In the transition zone, actin filaments are depolymerized. At the minus (⫺) end (pointed end), actin monomers dissociate from the filaments. The released actin monomers repolymerize at the tips of filopodia.

Guidance of Axons to Targets in Development and in Disease

restrained from entering the peripheral zone by steric hindrance.108 Indeed, depletion of actin filaments using actindepolymerizing drugs allows microtubules to protrude into the peripheral zone of the growth cone in neurons of the sea slug Aplysia.108 Moreover, destabilization of actin filaments allows hippocampal neurons in cell culture to form supernumerary axons.35,36 Microtubule protrusion into the peripheral domain may—alternatively or additionally—be prevented by signaling between actin and microtubulebinding proteins.390,418 Despite the dense peripheral actin network, individual microtubules explore the peripheral domain along actin filaments326 (see Fig. 21–1A) in a process of alternation of polymerization and depolymerization phases, a state referred to as dynamic instability. Actin filaments also undergo treadmilling: actin monomers are added to the plus (barbed) ends of existing filaments at the leading edge (Fig. 21–1B), and they dissociate from the minus (pointed) ends (Fig. 21–1C). Retrograde F-actin flow driven by myosin motors mediates the transport of actin filaments

449

centripetally away from the leading edge (peripheral domain) (see Fig. 21–1B) toward the transition zone where actin bundles are depolymerized108,227 (see Fig. 21–1C). Subsequently, the actin monomers diffuse to the leading edge, where they re-form actin filaments by polymerization. The growth of microtubules toward the periphery of the growth cone is opposed by this retrograde actin flow. Along with actin filaments, microtubules are transported toward the transition zone, where microtubule turnover occurs.326 Cytoskeletal dynamics are subject to tight regulation during axon growth and guidance. As an example, contact between beads coated with Aplysia cell adhesion molecules and cultured Aplysia neuronal growth cones induces growth cone turning toward the site to which the bead is attached.361 On the level of the cytoskeleton, actin polymerizes at the site of bead attachment and retrograde actin filament flow is locally attenuated. Part of the peripheral domain adjacent to the bead thus becomes devoid of actin filaments. Protruding microtubules from the central domain invade this area, resulting in a net turning response (Fig. 21–2).

FIGURE 21–2 Remodeling of the cytoskeleton during axon guidance (all images refer to the same growth cone). A, Upon contact between a bead coated with the anti–Aplysia cell adhesion molecule (apCAM) antibody (4E8) and the peripheral domain of an Aplysia bag cell growth cone, the central domain of the growth cone extends towards the bead (arrowhead marks initial boundary of the central domain) (B). C and D, Rhodamine-phalloidin labeling of F-actin (C) and ␤-tubulin immunostaining (D) at the 5-minute time point. (Arrowhead marks the position of the bead, visible as a ring because the protein A coating of the bead also binds the secondary antibody). Microtubules (visualized by video-enhanced contrast–differential-interference contrast microscopy) protrude into the peripheral domain as the F-actin network is destabilized in the direction of the interaction axis, that is, toward the guidance signal. E, Pseudocolor overlay of F-actin (red) and tubulin (green) stainings. Note that the actin network in the remaining peripheral domain almost completely excludes microtubules. (From Suter, D. M., Errante, L. D., Belotserkovsky, V., and Forscher, P.: The Ig superfamily cell adhesion molecule, apCAM, mediates growth cone steering by substrate-cytoskeletal coupling. J. Cell Biol. 141:227, 1998 by copyright permission of The Rockefeller University Press and Paul Forscher, Yale University.) See Color Plate

450

Neurobiology of the Peripheral Nervous System

Summary Guidance cues regulate cytoskeletal polymerization, depolymerization, and the retrograde F-actin flow to modulate the turning behavior of a growing axon (see Fig. 21–1B). A major challenge in recent years was to decipher the signaling events that lead to downstream cytoskeletal changes upon ligand receptor binding. Because changes in the actin dynamics seem to be one of the major forces to orchestrate axonal growth and guidance, we next discuss the molecules that are involved in changing actin dynamics by modulating the polymerizing, depolymerizing, or treadmilling rate of actin filaments.

Actin Dynamics Rho GTPases: Key Modulators of Actin Dynamics The role of members of the Rho family of small GTPases (Rho GTPases) in actin dynamics has been studied extensively in non-neuronal cells, in which they control various actin-dependent cellular processes, including migration, phagocytosis, secretion, cell division, and polarity phenomena (for review, see Etienne-Manneville and Hall98 and Raftopoulou and Hall291). Importantly, there is also broad evidence for their involvement in the regulation of cytoskeletal dynamics in growth cones (for review, see Dickson85 and Luo230). Rho GTPases cycle between a GDP-bound, inactive state and a GTP-bound, active form, which functions by interacting with a broad variety of downstream effectors such as kinases or scaffold proteins28 (Fig. 21–3). The activity of Rho GTPases is controlled by three groups of regulatory proteins (see Fig. 21–3): guanine nucleotide exchange factors (GEFs),60,328 guanine nucleotide dissociation inhibitors,276 and GTPase-activating proteins (GAPs).283 Some of these regulators show specificity for one or two Rho GTPases, whereas others are less specific (Fig. 21–4). As is discussed later, axon guidance receptors act, either directly or indirectly, as regulators of the Rho GTPases. The Rho GTPases most comprehensively analyzed are Cdc42, Rac, and RhoA. In cells from the Swiss 3T3 fibroblast cell line, microinjection of constitutively active Cdc42 triggers the formation of filopodia,202,272 constitutively active Rac promotes the formation of lamellipodia,300 and constitutively active RhoA induces the formation of contractile actin/myosin filaments, the stress fibers.299 Similarly, inactivation of Rho induces loss of actin filaments and enhances neurite outgrowth in PC12 cells (derived from adrenal gland tumors), N1E-115 neuroblastoma cells, and cerebellar and hippocampal neurons, whereas presumed activation of Rho by lysophosphatidic acid causes growth cone collapse and neurite retraction.29,74,175,176,269 In contrast, Rac and Cdc42 appear to be positive regulators of axonal growth.42,185,231,232 Interestingly,

FIGURE 21–3 Regulation of Rho GTPase activity. Rho GTPases cycle between an inactive GDP-bound and an active GTP-bound state. Activity of Rho GTPases is controlled by GTPase-activating proteins (GAPs), guanine nucleotide dissociation inhibitors (GDIs), and guanine nucleotide exchange factor (GEFs), which are themselves subject to regulation by guidance signals.

Rac and Cdc42 negatively regulate Rho in cells from the NIH 3T3 fibroblast cell line.319 If a similar correlation exists in neurons, one could assume that the migrating growth cone of a growing axon with active filopodia and lamellipodia contains more active Cdc42 and Rac and less active Rho, reflected by high Cdc42-GTP/Cdc42-GDP and RacGTP/Rac-GDP ratios as well as a low Rho-GTP/Rho-GDP ratio. An axon that has stopped growing should show the opposite. The Rho-GTPases act on the actin cytoskeleton via various downstream targets. These actin-associated molecules are discussed next. Actin Polymerization and Nucleation Actin-modulating proteins change polymerization rate, depolymerization rate, and motility of actin filaments. We first discuss molecules involved in actin polymerization and nucleation.

Guidance of Axons to Targets in Development and in Disease

451

FIGURE 21–4 Downstream effectors of Rho GTPases. Rho GTPases modulate cytoskeletal dynamics by multiple downstream effectors (for details see text). Actin polymerization and depolymerization and retrograde flow of F-actin are the main parameters influenced by Rho GTPases to regulate growth cone dynamics. Generally, effectors of Cdc42 and Rac mediate growth cone attraction and neurite growth, whereas RhoA targets cause growth cone repulsion and collapse (for details see text).

Profilin, Thymosin, and ROCK. The actin-binding proteins profilin and thymosin ␤4 are key regulators of actin polymerization because they control the addition of actin monomers to growing (plus) ends of actin filaments.51,53,183 Actin monomers are sequestered by the actin monomer–binding protein thymosin ␤4 to prevent their spontaneous polymerization.53,313 Like thymosin ␤4, profilin also associates with monomeric actin. It has been suggested that profilin releases actin monomers from thymosin ␤4 and allows their assembly onto the plus ends of actin filaments, thereby enhancing actin polymerization.183 Regulation of profilin activity is a means for Rho GTPases to influence actin assembly at the plus ends of actin filaments. Rho kinase (ROCK), a downstream effector kinase of activated Rho, complexes with and phosphorylates the brain-specific profilin II (see Fig. 21–4). Activation of this pathway increases actin stability and inhibits neuritogenesis74,401 (compare with

discussion in Rho GTPases: Key Modulators of Actin Dynamics earlier). Nucleation of Actin Filaments by Members of the Formin Family. In contrast to the polymerization of actin filaments, its initial step, the nucleation of the filaments de novo, is not well characterized yet. Only recently has the identification of formin-1 as a nucleator of actin filaments in yeast shed some light on the process of actin nucleation.99,315 Formins recruit actin monomers via their proline-rich, profilin-bound, G-actin–binding domain.196,316 Mammalian diaphanous (mDia), a member of the formin protein family that also interacts with the insulin receptor tyrosine kinase substrate protein (IRSp53),115 provides a link between profilin and the GTPase Rho in fibroblasts.204,389 mDia as part of a Rho/mDia complex might have a role similar to that of formin-1 in Rho-mediated actin polymerization (see Fig. 21–4). Indeed, mDia is necessary

452

Neurobiology of the Peripheral Nervous System

for stromal cell–derived factor-1␣–mediated axon formation in cerebellar granule cells6 (compare with Fig. 21–5 later in the chapter). Activation of Arp2/3 by WASP/Scar Family Proteins. The Arp2/3 complex nucleates branching of actin filaments upon activation by nucleation-promoting factors of the Wiskott-Aldrich syndrome protein (WASP)/Scar family, including neuronal WASP (N-WASP) and Scar/WASP-family verprolin homologous protein (WAVE)31,391 (see Fig. 21–4). N-WASP and Scar are effectors of the Rho-GTPases Cdc42 and Rac, respectively. N-WASP and Cdc42 associate directly via the GTPase binding domain/Cdc42/Rac interactive binding region (CRIB) of N-WASP,1,311 and IRSp53 mediates binding of Scar/WAVE to Rac.248,249 In conjunction with Cdc42 and Rac, respectively, N-WASP and Scar/WAVE act upon profilin and the Arp2/3 complex, which function synergistically to facilitate nucleation and branching of actin filaments.235,236,247,302,357 Promotion of Actin Polymerization by Mena Anticapping Function. Proteins of the Ena/VASP family, including enabled (Ena) of the fruit fly Drosophila,124 vasodilator-stimulated phosphoprotein (VASP),140,141 and mammalian Ena (Mena),125 are important modulators of actin polymerization in processes such as axon guidance, cellular migration, T-cell activation, and phagocytosis (for review, see Krause and colleagues203). Capping proteins associate with the plus ends of actin filaments and terminate actin polymerization.170 Ena/VASP proteins also interact with plus ends and prevent the binding of capping proteins, thereby stimulating actin polymerization by their anti-capping function.22 The Rho GTPase Cdc42 interacts with IRSp53 to trigger the formation of an IRSp53/Mena complex.205 Mena in turn promotes the release of capping proteins from the plus ends of actin filaments, leading to a Cdc42-mediated increase of actin polymerization in fibroblasts22,205 (see Fig. 21–4). Mena is localized at the leading edge of neuronal growth cones, suggesting a similar Mena/VASP-dependent mechanism in neurons.216 Actin Depolymerization In addition to actin polymerization, Rho GTPases also control growth cone dynamics by modulating the depolymerization rate of actin filaments. Members of the actin depolymerizing factor (ADF)/cofilin family, which are key regulators of actin dynamics, have the ability to sever actin filaments and increase filament turnover by enhancing the dissociation rate of actin subunits from the minus end of actin filaments.15,206,321 The F-actin severing protein cofilin,268 which makes up the majority of ADF/cofilin proteins in the human central nervous system (CNS), predominantly localizes to neuronal somata and growth cones.14,252 Cofilin activity is subject to phosphoregulation by the cytosolic serine/threonine kinase LIM kinase (LIMK) and

the phosphatase slingshot7,96,271 (see Fig. 21–4). LIMK and slingshot decrease and increase cofilin activity, respectively. Although upstream regulators of slingshot phosphatase have not been identified yet, modulators of LIMK activity are well described. p21-Activated kinase (PAK) and ROCK both activate LIMK by phosphorylation, thereby providing a link to Rho GTPases.7,92,237,273,358,411 PAK is a downstream effector kinase of Cdc42 and Rac,239 and ROCK is activated by Rho174,221 (see Fig. 21–4). Activation of Rho GTPases leads to phosphorylation and thereby activation of LIMK via PAK or ROCK; activated LIMK in turn promotes an increase in F-actin levels by inactivating the depolymerization activity of cofilin. Multiple lines of experiments suggest that cofilin has an important role in axonal growth and guidance. Overexpression of ADF in cortical neurons causes both increased actin-driven growth cone dynamics and enhanced axonal growth.243 Similarly, overexpression of an active, nonphosphorylatable form of cofilin in dorsal root ganglion (DRG) neurons increases axonal growth rates.96 Moreover, semaphorin 3A (Sema3A)–induced growth cone collapse results in a change in phosphorylation levels of cofilin.3 Together, cell culture studies showed that cofilin and its regulators appear to be central modulators of actin dynamics in axonal growth. It will be interesting now to see whether the respective knockout mice, assuming nonlethality, show changes in axonal growth and guidance. Retrograde F-Actin Flow Retrograde flow of actin filaments driven by nonmuscle myosins is another crucial process in growth cone steering in addition to actin assembly and disassembly.227,361 In this respect, control of myosin activity and actin/myosinmediated contractility provides a means for Rho GTPases to regulate growth cone progression or retraction. The activity of myosin is dependent on the phosphorylation state of the myosin light chain (MLC).182 Phosphorylated myosin shows higher ATPase activity resulting in higher contractility. MLC phosphorylation is controlled by ROCK, MLC kinase (MLCK) and MLC phosphatase (MLCP), which are themselves subject to regulation by Rho GTPases (see Fig. 21–4). Rho interacts with MLCP and triggers its inactivation via phosphorylation by ROCK.191 Inactivation of MLCP increases the fraction of phosphorylated, active myosin and leads to an increase in contractility, thus promoting growth cone retraction or collapse. In addition to inactivating MLCP, ROCK directly phosphorylates MLC, also contributing to myosin activation.4 Cdc42 and Rac signaling has the opposite effect on myosin activity. PAK, their shared downstream effector kinase, phosphorylates MLCK, which in turn leads to its inactivation320 (see Fig. 21–4). Consequently, myosin phosphorylation drops and the fraction of dephosphorylated, inactive myosin rises; therefore, Cdc42 and Rac inhibit growth cone retraction.

Guidance of Axons to Targets in Development and in Disease

Summary Actin dynamics control axonal growth and guidance events. We have discussed the key molecular players orchestrating actin dynamics, both those within the growth cone and migrating fibroblasts (including Rho-GTPases), the factors involved in actin nucleation and polymerization (formins/mDia, Arp2/3, Ena/VASP family members, profilin), proteins that depolymerize actin filaments (including members of the ADF/cofilin family), and myosins that generate the molecular force creating retrograde flow of actin filaments within the growth cone. In the next section we present the different ligand-receptor systems involved in axon guidance. The discussion reveals that, despite many differences in the structure and action of these systems, they merge onto similar downstream molecular targets that regulate the actin cytoskeleton. Thus the described actin modulators will reoccur as downstream targets of the receptor signaling.

AXON GUIDANCE IN DEVELOPMENT In this section we present the four main classes of ligandreceptor systems that are currently well described in axon guidance. The vast expansion of this field does not allow coverage of all the receptor and ligand subtypes in a single chapter. We therefore refer the reader to recent reviews for further information. We concentrate here on major and welldescribed signaling events leading to axonal guidance and growth. Other ligand-receptor systems that show changes in pathologic conditions, including the Wnt signaling pathway and cell adhesion receptors such as L1 and N-CAM, are discussed in Axon Guidance in Disease later in this chapter.

Ephrin/Eph The ephrin/Eph family is largely known for its key role in setting up the correct connections in the retinal tectal system, which has been fascinating scientists for over half a century. Axons of retinal ganglion neurons project from the retina to the optical tectum. Depending on the location of the neurons within the retina, their axons arborize at different locations within the optical tectum. Axons from the nasal and temporal retina project to the posterior and anterior tectum, respectively, and axons from the dorsal and ventral retina project to the lateral and medial tectum, respectively. Characterization of the distribution of Eph receptors and ephrin ligands in retinal ganglion cells and the optical tectum showed that ephrin ligands and Eph receptors are present in a gradient and that the concentrations of ligands and receptors give positional information to retinal ganglion axons in the tectum (for review, see Drescher and co-workers,88 Flanagan and Vanderhaeghen,105 and Knoll and Drescher194). Recently, ephrin/Eph signaling was also shown to be involved in the formation of other topographic maps,

453

including the vomeronasal olfactory system, which is necessary for the detection of pheromones.195 Members of the ephrin/Eph family mediate axon pathfinding in other contexts as well, including the guidance of forebrain commissural axons and commissural neurons in the spinal cord.154,208,209,211 There they act as a midline repulsion signal. Receptor Complexes The Eph Receptor Family. With 13 members, the family of Eph tyrosine kinases, receptors for the ephrins, constitutes the largest family of receptor tyrosine kinases in the mammalian genome (reviewed by Kullander and Klein210 and Wilkinson395). Depending on their ligand specificity and sequence similarities, the Eph receptors are divided into two subclasses, EphA and EphB receptors. EphA receptors (EphA1 through EphA8) bind glycosylphosphatidylinositol (GPI)-linked ephrin-As (ephrin-A1 through ephrin-A5), and EphB receptors (EphB1 through EphB4 and EphB6) bind transmembrane ephrin-Bs (ephrin-B1 through ephrin-B3). One exception is EphA4, which binds both A-type and most B-type ephrins. Structure of Eph Receptors. The extracellular part of the Eph receptors consists of a ligand-binding globular domain, a cysteine-rich region, and two fibronectin type III (FNIII) repeats. The cytoplasmic part contains a juxtamembrane domain (with two conserved tyrosine residues), a protein tyrosine kinase domain, a sterile ␣-motif (SAM) domain, and a C-terminal PDZ-binding motif (protein-protein interaction domain mediating interaction with C-terminal polypeptides or other postsynaptic density protein 95, Drosophila discs large tumor suppressor, zonula occludens-1 [PDZ] domains). In conjunction with their ligands, the ephrins, Eph receptors dimerize and form higher order clusters possibly enhanced by their SAM domain.23,155 With both Eph receptors and ephrin ligands being membrane attached, intercellular contact is required for ephrin signaling. In this respect binding by clustered, membrane-bound ephrins, which may facilitate the formation of ephrin-Eph clusters, was shown to be necessary for Eph receptor activity.78 Ephrin Forward Signaling One interesting feature of Eph/ephrin signaling is its bidirectionality. That is, in addition to regular downstream or “forward” signaling (into the Eph receptor–expressing cell), there is also “reverse” signaling (into the ephrin ligand–expressing cell)44,159 (Fig. 21–5). In Eph receptors the nonphosphorylated juxtamembrane domain interacts with the kinase domain and prevents it from adopting its catalytically active conformation.407 Upon ligand binding, autophosphorylation of the receptor dimer leads to phosphorylation of intracellular tyrosine residues within the juxtamembrane domain and the kinase domain activation segment, enabling the Eph receptors to exhibit their biologic and catalytic activity.27,180

454

Neurobiology of the Peripheral Nervous System

FIGURE 21–5 Bidirectional ephrin/Eph signaling. Both ephrin ligands and Eph receptors are linked to downstream effectors that modulate actin dynamics and cellular adhesion. The Eph receptor tyrosine kinases mediate forward signaling and promote axon repulsion. GPI-anchored ephrin-As and transmembrane ephrin-Bs are mediators of reverse signaling, for which ephrin-As require other transmembrane proteins. Ephrin reverse signaling can promote both attractive and repulsive responses.

In line with the broad variety of biologic functions they control, the activated Eph receptors associate with a large number of proteins (reviewed by Kullander and Klein210 and Wilkinson395) (see Fig. 21–5, forward signaling). We focus here on those relevant for axon guidance. Regulation of Rho GTPases by Eph-Associated GEF Ephexin. In a yeast two-hybrid screen for interactors of the cytoplasmic domain of EphA receptors, a novel GEF termed ephexin (Eph-interacting exchange protein) was identified.336 The Dbl-homology/pleckstrin-homology region of ephexin interacts with the C-terminal lobe of the EphA4 kinase domain, and the activity of ephexin is regulated by ephrin-A signaling. Ephrin-A–induced activation of ephexin leads to growth cone collapse in rat

retinal ganglion cells.336 The identification of ephexin provides an important link between EphA receptors and the cytoskeleton in both functional and mechanistic respects. Without ephrin stimulation, the GEF activity of ephexin activates RhoA, Cdc42, and Rac.336 Upon stimulation with ephrin-A, however, the EphA receptor inhibits ephexin activation of Cdc42 and Rac.336 This results in a shift of ephexin activity toward RhoA, which modulates actin dynamics in conjunction with its downstream kinase ROCK to cause growth cone collapse336,385 (see Fig. 21–5; compare with Fig. 21–4). Abl and Arg Kinase. Another link to the actin cytoskeleton is provided by the interaction of EphB2 and EphA4 receptors with Abelson kinase (Abl) (also see Phosphorylated

Guidance of Axons to Targets in Development and in Disease

Ephrin-B: Association with Grb4 later) and Abl-related gene kinase (Arg), two nonreceptor tyrosine kinases of the Abl kinase family. Eph receptors associate directly with Abl and Arg in both a phosphorylation-dependent and phosphorylation-independent manner as well as indirectly via a so far unidentified bridging protein.414 Abl and Arg both possess F-actin binding domains,242,414 and Abl localizes to growth cones in the CNS.70 In addition, Abl tyrosinephosphorylates Mena,367 which in turn stimulates actin polymerization by releasing actin-capping proteins22 (see Promotion of Actin Polymerization by Mena Anti-capping Function, earlier, and Fig. 21–4). FAK and Cas. Another downstream target of Eph receptors that acts on the actin cytoskeleton is focal adhesion kinase (FAK), a regulator of integrin-mediated cell adhesion (see Fig. 21–5, forward signaling). In fibroblasts, stimulation with ephrin-A1 induces cytoskeletal reorganization, which results in enhanced spreading and adhesion.52 Both FAK and Crk-associated tyrosine kinase substrate (Cas) are required for this process. Cas is also implicated in actin filament assembly and integrin-dependent processes such as migration, cell cycle control, and transformation.161,277,324,362 In the absence of ephrin, EphA2 receptors are associated with FAK.245,280 Upon stimulation with ephrin-A1, the protein tyrosine phosphatase Shp2, a regulator of FAK phosphorylation, is recruited to the EphA2 receptor with delayed kinetics (compare with discussion in Tyrosine Phosphatase Shp-2 Binding of UNC-5 later). Subsequently, FAK is rapidly dephosphorylated and thereby inactivated and dissociates from the EphA2 receptor, which may result in the termination of FAK-mediated adhesion. Promotion of Neurite Retraction by Inactivation of the MAPK Pathway. Ephrin-mediated neurite retraction can also be triggered by inactivation of the mitogen-activated protein kinase (MAPK) pathway95,246 (compare with discussion in Promotion of Local Protein Synthesis and Degradation by Netrin Signaling later). As mentioned earlier, activation of the EphB2 receptor leads to autophosphorylation of tyrosine residues within the juxtamembrane domain. Subsequent downregulation of Ras activity may be promoted by Src homology 2 (SH2) domain–mediated binding of p120-RasGAP to EphB2. A functional link between Ras activity and neurite retraction might be the cross talk between the Ras and Rho signaling pathways. In fibroblasts, MAPK signaling activated by Ras mediates inhibition of the Rho target ROCK, leading to increased cell motility.317 Ras inactivated by ephrin signaling no longer activates the MAPK pathway; therefore, ROCK inhibition is released, which will contribute to neurite retraction upon stimulation with ephrins (see Fig. 21–5). In the context of EphB1-dependent cell migration, however, Eph/ephrin signaling activates MAPKs.382 Activated EphB1 recruits the adaptor proteins Grb2 and p52Shc.

455

Subsequently, the tyrosine kinase c-Src as well as the extracellular signal-regulated kinase (ERK) MAPKs are activated, which promotes chemotaxis in human renal microvascular endothelial cells. Ephrin Reverse Signaling In contrast to other signaling pathways, in the case of ephrin signaling not only the receptors (Ephs) but also the ligands (ephrins) have both phosphorylation-dependent and phosphorylation-independent signaling capacity. For example, ephrin-B reverse signaling causes forebrain commissural axons to turn away from EphB-expressing regions.154 Somewhat unexpectedly, in addition to the transmembrane ephrin-Bs, the GPI-anchored ephrin-As also possess this reverse signaling capacity, indicating that ephrin-A reverse signaling functions via so far unidentified coreceptors.79 Recent work on the vomeronasal system suggests that vomeronasal axons receive guidance signals through their ephrin-A “ligands.”195 Phosphorylation of Ephrins by Tyrosine Kinases. Three conserved tyrosine residues within the cytoplasmic domain of ephrin-B ligands can be phosphorylated by Src family tyrosine kinases (SFKs), fibroblast growth factor (FGF) receptor tyrosine kinases, or platelet-derived growth factor receptor tyrosine kinases.63,180,279 Upon Eph/ephrin engagement, the protein-tyrosine phosphatase PTP-BL is recruited to ephrins in a PDZ domain–dependent manner with delayed kinetics279 (see Fig. 21–5, reverse signaling; compare with discussion in Phosphorylation-Independent, PDZ Domain–Mediated Interactions of B-Type Ephrins later). PTP-BL subsequently dephosphorylates ephrin-Bs and thereby antagonizes the phosphorylation by tyrosine kinases. Phosphorylated Ephrin-B: Association with Grb4. The SH2–Src homology 3 (SH3) adaptor protein Grb4, an adaptor protein for multiple regulators of actin dynamics, binds the phosphorylated tyrosine residues of ephrin-Bs by virtue of its SH2 domain, and it also interacts via its SH3 domains with other proteins implicated in cytoskeletal dynamics.72 These Grb4 effectors include Cbl-associated protein (CAP)/ponsin, a regulatory protein in the formation of actin stress fibers and focal adhesions298; the Abl-interacting protein-1 (Abi-1), an interaction partner of Abl tyrosine kinase70,338; dynamin, a GTPase implicated in endocytosis and actin dynamics73,318; PAK1, the effector kinase of Cdc42 and Rac239 (see Fig. 21–4); heterogeneous ribonucleoprotein K, an interactor of the Rac GEF Vav158,165; and axin, a scaffold protein in the Wnt signaling pathway, a pathway recently found to be involved in guidance of commissural axons.143,234,417 What is more, Grb4 can also interact with Abl71 (see Fig. 21–5, reverse signaling; compare with discussion in Abl and Arg Kinase earlier), thereby providing another

456

Neurobiology of the Peripheral Nervous System

link to actin dynamics. Abi-1 promotes the Abl-mediated tyrosine phosphorylation of Mena,367 which in turn modulates actin polymerization by release of actin filament capping proteins22 (see Promotion of Actin Polymerization by Mena Anti-capping Function, earlier, and Fig. 21–4). Phosphorylation-Independent, PDZ Domain–Mediated Interactions of B-Type Ephrins. In addition to phosphorylation-dependent interactions via their SH2 motif, B-type ephrins can also undergo phosphorylationindependent interactions mediated by their PDZ domain (see Fig. 21–5, reverse signaling). Such interactors include glutamate receptor–interacting proteins GRIP1 and GRIP2,43 syntenin,135 the tyrosine phosphatase PTP-BL,153 protein kinase C–interacting protein-1 (Pick1),374 and PDZRGS3.228 PTP-BL releases the phosphorylation of tyrosine residues within the cytoplasmic domain of ephrin-Bs,279 antagonizing the tyrosine phosphorylation by SFKs and thereby the phosphorylation-dependent binding of Grb4 and its interactors. GPI-Linked A-Type Ephrins: Capacity for Reverse Signaling. Although reverse signaling by transmembrane ephrin-B ligands is more evident, GPI-linked ephrin-A ligands also have the capacity for reverse signaling. However, the mechanism by which they execute this effect is not yet understood. GPI-anchored proteins are clustered in membrane microdomains (rafts) via their GPI anchors.5,40 In contrast to EphA2 receptor signaling, which leads to FAK dephosphorylation and reduced integrin-mediated adhesion245 (see FAK and Cas earlier), compartmentalized ephrin-A5 signaling increases phosphorylation of FAK and cellular adhesion in fibroblasts and neuroblastoma cells. In addition, an unidentified 120-kDa membrane protein was also phosphorylated upon ephrin-A5 signaling (see Fig. 21–5, reverse signaling). These ephrin-mediated effects depend on ␤1 integrin and activation of lipid raft–associated Fyn tyrosine kinase, a member of the Src kinase family that is one of the kinases upstream of FAK.79,80,134,168 Modulation of Ephrin/Eph Signaling Three mechanisms have been identified so far that enable targets to modulate their responsiveness to ephrin/Eph signaling. The ephrin-A2 ligand forms a complex with the metalloprotease Kuzbanian, which specifically and locally cleaves ephrin-A2 upon Eph receptor binding. A mutation in ephrin-A2 that inhibits cleavage delays axon withdrawal. Proteolytic cleavage is a common theme in the regulation of response to guidance cues (compare with discussion in Receptor Degradation or Proteolytic Cleavage of Robo and in Proteolytic Cleavage of the DCC Receptor later). It therefore provides a mechanism for efficient control and possibly termination of ephrin-A2 signaling.148 Local protein synthesis provides another means to adjust the surface receptor composition and thereby the

responsiveness of the growth cone to its environment. EphA2 receptor messenger RNA (mRNA) is locally translated in distal segments of axons after they have reached their intermediate targets in the midline. The resulting increase in Eph receptors on the growth cone surface changes the responsiveness to ephrin signals.37 Endocytosis constitutes another possibility to clear ligandreceptor complexes from the growth cone surface. In fibroblasts and endothelial cells, the internalization of EphB4/ephrin-B2 complexes and the subsequent cell retraction require Rac-dependent actin polymerization within the receptor-expressing cells.240 Upon binding of ephrin-B1 to the EphB2 receptor, growth cones of cultured telencephalic or hippocampal neurons detach and withdraw. Subsequently, they internalize the entire ligandreceptor complex.420 Like ephrin signaling, endocytosis is bidirectional, that is, both ligand- and receptor-expressing cell take up the ligand-receptor complex. In vivo this process is likely to be regulated by additional signals, because the C-terminal part of ephrin and the kinase domain of the Eph receptor are necessary for the uptake of the complex into the ephrin- and Eph-expressing cell, respectively. Bidirectional endocytosis of ephrin/Eph complexes represents an efficient mechanism to terminate cellular adhesion and allow growth cone retraction.

Slit/Robo Receptor Complexes Roundabout (Robo), a single-pass transmembrane protein of the immunoglobulin (Ig) superfamily, is the cell surface receptor mediating the repulsive effects of the guidance cue molecule Slit.39,190,223 Initially, Robo was discovered in a genetic screen for mutations that affect the development of CNS axon pathways in Drosophila.332 Axons that are prevented from crossing the midline in wild-type flies can cross and recross in Robo mutants, suggesting that Robo is a component of a repulsive signaling system at the midline, whereas Slit is the repulsive cue located at the midline. Robo Structure and Differential Expression Patterns. The extracellular part of Robo consists of five Ig domains and three FNIII repeats. The cytoplasmic region of Robo contains four conserved motifs (CC0, CC1, CC2, and CC3), and each of these is crucial for certain aspects of Robo signaling.18,354 In Drosophila, the pattern of overlapping expression of its three Robo receptors (Robo-1, -2, and -3, with both Robo-2 and -3 lacking the CC2 and CC3 motifs) in the CNS defines a pattern of differential Slit sensitivity, the Robo code.292,341 These expression domains make up medial, intermediate, and lateral zones within the longitudinal pathways that allow a well-defined medial-to-lateral distribution of longitudinal axon bundles in conjunction with cell adhesion molecules such as fasciclin and neural cadherin (N-cadherin).94,292,297,341,344

Guidance of Axons to Targets in Development and in Disease

Robo receptors seem to have differential expression patterns in vertebrates as well, as recently shown for Robo-1 and -2 in the developing mouse nervous system.360 Robo-1/2 expression in the mouse spinal cord corresponds with the Robo functions proposed for Drosophila; however, the more complex patterns in murine brain suggest additional roles for Robo in vertebrate neural development. Slit Proteins: Ligands of the Robo Receptor. Mutations in the slit genes were first known to cause midline defects in the Drosophila CNS308; later on, Slit proteins were identified as repulsive guidance cue molecules signaling through the Robo receptor.39,190,223 Slit proteins constitute a family of guidance molecules structurally conserved between species as distant as insects, nematodes, and mammals. In mammals, three Slit proteins (Slit-1, -2, and -3) have been identified to date. Similar to Drosophila, mammalian Slit is expressed in the midline.39 Mice deficient for both Slit-1 and Slit-2 have defects in the formation of the optic chiasm.287 However, in contrast to Drosophila, these mice show normal midline guidance in the spinal cord. This different effect in midline guidance may be due to the redundancy of Slit-3 present at the midline of the spinal cord but not in the optic chiasm during development.287 Similarly, astray/robo2 mutant zebra fish show axon guidance defects in retinal axons at the midline.112 Moreover, Slit proteins act as axon guidance cues for corticofugal, callosal, thalamocortical, and olfactory neurons,13,265 and they regulate branching of axons and dendrites.386,394 Slit Structure and Proteolytic Cleavage. Members of the Slit family share a typical conserved domain structure. They contain an N-terminal signal peptide targeting them for secretion, four leucine-rich repeats mediating receptor interaction, seven (Drosophila) to nine (vertebrates) glutamic acid–glycine-phenylalanine (EGF) repeats, a lamin G domain, and a C-terminal cysteine knot.21,58,307–309 Slit proteins are subject to proteolytic cleavage (in the case of human Slit-2, between EGF repeats 5 and 6); a so far unidentified protease generates two fragments originating from the N- and C-terminus (Slit-N and Slit-C, respectively).39,386 The function of both Slit-C and the C-terminal part of uncleaved Slit is not well defined yet. Binding of Slit to Robo and Robo/Slit-mediated axon guidance depend on heparan sulfate proteoglycans,166,353 and Slit-C has higher affinity to heparan sulfate proteoglycans than Slit-N.226,305 There is also indirect evidence for a role of the C-terminal part of Slit in the regulation of Slit diffusion58; however, it is dispensable for receptor interaction and signaling.21,58,263 Uncleaved Slit and Slit-N act as a repulsive cue for axons21,58,263; however, depending on the type of neuron, there are differences in the response. For example, both

457

Slit-N and uncleaved Slit promote growth cone collapse in retinal ganglion cells, but only Slit-N leads to growth cone collapse in olfactory bulb neurons,263 and uncleaved Slit antagonizes the branch-promoting activity of Slit-N in DRG neurons.263,386 Robo/Slit Signaling Abl and Mena: Implementation of Axon Repulsion Mediated by Robo/Slit Signaling. Studies in Drosophila have shed some light on the mechanisms by which Robo/Slit signaling mediates axon repulsion. A similar mechanism likely exists in vertebrates. Both Abl (compare with discussion in Abl and Arg Kinase earlier) and Ena (see Promotion of Actin Polymerization by Mena Anti-capping Function earlier), one of its downstream targets123 involved in actin dynamics, associate with Robo (Fig. 21–6). Abl interacts with the proline-rich CC3 motif of Robo via its SH3 domain and phosphorylates Robo in the conserved CC1 motif.19 Tyrosine phosphorylation serves as a negative regulatory switch for both Ena and Robo activity. The Robo CC2 motif contains a consensus binding site for the EnaVASP homology (EVH1) domain of Ena, and Ena indeed binds to Robo, via interactions with its CC1 and CC2 motifs.19 Axon repulsion in the nematode Caenorhabditis elegans involves signaling through SLT/SAX-3, the homologues of Slit/Robo. This signaling pathway depends on the Ena homologue UNC-34,415 which modulates actin dynamics in the growth cone (compare with discussion in Promotion of Actin Polymerization by Mena Anti-capping Function, earlier, and Fig. 21–4). Abl Target Capulet and Robo/Slit Signaling. Yet another Abl target, Capulet (Capt), a homologue of yeast adenylyl cyclase–associated protein (CAP), seems to be involved in Robo/Slit signaling (see Fig. 21–6). Abl and Capt interact and function in midline axon guidance.398 CAP can interact directly with actin,111 and its function is also linked to profilin.139,383 Capt lacking the adenylyl cyclase binding domain still rescues capt mutations; however, this rescue is only partial in comparison to fulllength Capt.398 Adenylyl cyclase and modulation of intracellular cyclic AMP (cAMP) levels may play a role in Robo/Slit signaling (see Modulation of Robo/Slit Signaling by Intracellular cGMP Levels later). Protein Tyrosine Phosphatases: Role in Antagonizing Abl Tyrosine Kinase. In contrast to Ena and Capt, the targets of the tyrosine kinase Abl, it is intriguing that Ptp10D and Ptp69D, two Drosophila PTPs, are positive regulators of Robo/Slit repulsive signaling.359 Dlar (Drosophila leukocyte antigen–related-like), another Drosophila PTP, is involved in motor axon guidance in conjunction with Abl and Ena397; it is therefore likely that the role of Ptp10D and Ptp69D in Robo/Slit signaling is to antagonize Abl kinase (see Fig. 21–6).

458

Neurobiology of the Peripheral Nervous System

FIGURE 21–6 Slit/Robo signaling. The tyrosine kinase Abl and its targets are important players in mediating the repulsive effects of Slit/Robo signaling. Their function is likely to be antagonized by protein tyrosine phosphatases. The Robo receptor associates with members of a subfamily of GAPs, the Slit/Robo GAPs (srGAPs). srGAPs decrease Cdc42 activity, thus also promoting Slit-mediated axon repulsion. Robo is able to form heteromultimeric DCC/Robo receptor complexes, which interfere with netrin-mediated attraction.

Although extensively studied in Drosophila, the roles of Abl, Ena/Mena, and Capt in vertebrate Robo/Slit signaling are not yet defined. Slit/Robo GAPs: Linking Robo/Slit Signaling to Rho GTPases and Actin Dynamics. The activity of Rho GTPases involved in Robo/Slit signaling is regulated by a subfamily of Rho GAPs, the Slit/Robo GAPs (srGAPs), which bind directly to the proline-rich CC3 motif of Robo via their SH3 domain405 (see Fig. 21–6). As discussed, GAPs stimulate the conversion of active, GTP-bound Rho GTPases to their inactive, GDP-bound form (see Rho GTPases: Key Modulators of Actin Dynamics, earlier, and Fig. 21–3). srGAP1 interacts with Cdc42, RhoA, and Robo; the interaction with Robo is SH3 dependent. Stimulation with Slit increases the binding of srGAP1 to Robo and Cdc42 and subsequently decreases Cdc42 activity (see Fig. 21–6). Conversely, Slit reduces srGAP-RhoA binding, yet the effect (a slight increase of RhoA-activity) is modest compared to Cdc42. A constitutively active Cdc42 overrides the repulsive effect of Slit on migrating cells from

the anterior subventricular zone,405 further supporting a role of srGAPs and Cdc42 in Robo/Slit signaling. Involvement of the Rho GTPase Rac in Robo/Slit Signaling. Another study in Drosophila implicates Rac1 in Robo/Slit signaling.102 Mutations in Rac, the SH2-SH3 adaptor protein dreadlocks (Dock)/Nck, or the serine/threonine kinase Pak, the Cdc42/Rac downstream effector (see Fig. 21–4), interfered with Robo/Slit signaling. Stimulation with Slit recruits a complex of Dock and Pak to the Robo receptor and mediates an increase in Rac1 activity (see Fig. 21–6). Rac functions in lamellipodia formation and attractive guidance; its precise role in repulsion is not yet understood. Rac may be able to trigger both attraction and repulsion. Modulation of Robo/Slit Signaling Receptor Degradation or Proteolytic Cleavage of Robo. Robo/Slit signaling is tightly regulated both to restrain ipsilaterally projecting axons from crossing the midline and to prevent contralaterally projecting axons from recrossing.

Guidance of Axons to Targets in Development and in Disease

In Drosophila, surface levels of Robo are controlled by commissureless (Comm),369 a sorting receptor for Robo. Via its conserved sorting signal (LPSY), Comm sorts Robo directly from the trans-Golgi network into vesicles targeted to late endosomes and lysosomes. This leads to degradation of both Robo and Comm.187 This sorting mechanism also involves active Nedd4 ubiquitin ligase. Nedd4 conjugates Comm to ubiquitin, thereby marking it for protein degradation by the ubiquitin-proteasome system.258 An alternative mechanism such as direct proteolytic cleavage, as observed for ephrin/Eph signaling (see Modulation of Ephrin/Eph Signaling earlier), may also contribute to control Robo levels. Usually Robo is not present on neurons in commissural tracts; however, Drosophila commissural axons expressing a dominant negative form of the metalloprotease Kuzbanian fail to clear Robo from the cell surface and show midline crossing defects.327 Modulation of Robo/Slit Signaling by Intracellular cGMP Levels. Intracellular levels of cyclic nucleotides are known to modulate turning responses to several guidance cues 348 (e.g., semaphorin 347 or netrin 254 ) (see also Modulation of Responses to Semaphorin by Intracellular cGMP Levels, and Modulation of Response to Netrin Signaling by Intracellular cAMP Levels, later). The possible implication of capt, the homologue of yeast CAP, in Robo/Slit signaling (see Abl Target Capulet and Robo/Slit Signaling earlier) may provide a link to regulation by cyclic nucleotide levels. The level of intracellular cGMP modulates the response to Slit in DRG neurons,264 and for netrin signaling the ratio of cAMP to cGMP proved to be important in the turning response.270 Capt may associate with adenylyl cyclase to alter intracellular cAMP levels in response to Slit stimulation (see Fig. 21–6); however, an actual involvement of adenylyl cyclase in Robo/Slit signaling remains to be shown. Cross Talk between Robo/Slit and DCC/Netrin Signaling. Studies in C. elegans415 and embryonic Xenopus spinal neurons354 suggest cross talk between Robo/Slit and deleted in colorectal cancer (DCC)/netrin signaling (compare with discussion in Silencing of NetrinMediated Attraction by Formation of Robo/DCC Complexes later). In Xenopus, Robo and the netrin receptor DCC associate via interactions of their respective CC1 and P3 cytoplasmic domains to form a heteromultimeric receptor complex. In this complex, activation of the Robo receptor by Slit silences netrinmediated attraction; however, the growth-stimulatory effect of netrin is not impaired354 (see Fig. 21–6). By this means, growing axons can be attracted to the midline by netrin but are not trapped there because upregulation of Robo activity at the midline silences netrin signaling. Subsequently, Slit-mediated repulsion allows axons to leave the midline again.187,339,354

459

Unlike in Xenopus, Robo/Slit-mediated repulsion in C. elegans may not be opposing signaling by netrin receptor UNC-40/DCC. Instead a netrin-independent function of UNC-40/DCC potentiates repulsion from Slit, possibly mediated by an UNC-40/Robo heteromeric receptor complex.415

Netrin/DCC/UNC-5 The netrins have been independently discovered by biochemical purification of a diffusible outgrowth-promoting activity for axons in the spinal cord335,371 and by genetic analysis of the C. elegans mutant UNC-6 showing abnormal axonal growth toward the midline151,172 (for review, see Kennedy188 and Merz and Culotti244). The netrin family consists of phylogenetically conserved secreted proteins sharing similarity with members of the laminin family. Netrins can function in both an attracting and repelling manner; netrin-1, for example, attracts spinal cord commissural axons but repels trochlear motor axons.65,189,334 Receptor Complexes Receptors for Netrins: DCC, UNC-5, and A2b Receptor. Netrin function is mediated by two different classes of transmembrane receptors, homologues to C. elegans UNC-40 and UNC-5. Vertebrate DCC and neogenin104,186,379 are homologous to C. elegans UNC40151,152 and Drosophila frazzled.200 Vertebrate UNC-5H-1, -2, -3, and -42,97,220 are homologous to C. elegans UNC-5.222 DCC appears to be involved in attraction, whereas UNC-5 appears to mediate repulsion in both vertebrates and invertebrates.56,82,104,144,186,222 In addition to DCC and UNC5, other coreceptors may be involved in netrin signaling, for example, the adenosine A2b receptor, a G protein–coupled receptor inducing cAMP accumulation upon adenosine binding.293 Before the direct netrin-DCC interaction had been established, one study found an interaction of the A2b receptor with both netrin and DCC in a yeast two-hybrid screen and in 293T cells, suggesting that the A2b receptor was a netrin receptor that mediates cAMP elevation upon netrin binding.69 However, a conflicting report later showed the direct netrin-DCC interaction without the need for the A2b receptor in netrin-induced axon outgrowth and attraction.355 The precise role of the A2b receptor in netrin signaling therefore awaits further clarification. Structure of DCC and UNC-5. The extracellular part of DCC receptors includes several Ig domains and FNIII repeats, and three motifs (P1, P2, and P3) conserved from insects to nematodes and mammals200 characterize the intracellular moiety of DCC. Upon ligand stimulation DCC multimerizes via its P3 domains, and this multimerization is necessary for netrin-mediated attraction.355 The organization of UNC-5 diverges substantially from DCC receptors. The extracellular part contains some

460

Neurobiology of the Peripheral Nervous System

Ig and thrombospondin domains, and the cytosolic part includes several other conserved motifs. First it contains a ZU-5 domain, also present in zonula occludens-1, a component of gap junctions. It further contains a death domain, a protein-protein interaction module present in many proteins involved in apoptosis, and a DCC-binding domain necessary for interaction with DCC. Association of DCC and UNC-5 via their cytoplasmic domains converts netrin/DCC-mediated attraction to netrin/DCC/UNC5–driven repulsion in Xenopus spinal axons.162 Netrin Signaling How do netrin receptors function to regulate axon growth and guidance? Several studies suggest that they modulate the actin cytoskeleton via Rho GTPases in response to netrin. Again, although many signaling events are very similar across species, there are also distinct differences. In C. elegans, several mutations suppress a gain-of-function mutation of the netrin receptor UNC-40/DCC, including CED-10, a Rac homologue; UNC-115, a protein homologous to the vertebrate actin-binding protein AbLIM; and

UNC-34, a homologue of Mena (see Promotion of Actin Polymerization by Mena Anti-capping Function earlier), which is an important modulator of actin polymerization.129 Coupling of the Rho GTPase Rac and Its Effector Kinase PAK to Netrin Signaling by the Adaptor Protein Nck. CED-10/Rac and UNC-115/AbLIM act in one pathway dependent on the conserved P2 domain of UNC40/DCC.129 It is interesting to note that the adaptor protein Nck associates with DCC in rat commissural neurons and that Nck is required for the activation of Rac by netrin signaling224 (Fig. 21–7). In Drosophila, the Nck Drosophila homologue Dock as well as the GEF trio are necessary for photoreceptor and motor axon guidance.20,119,262 Both Rac and Cdc42 activity are required for netrin-induced neurite outgrowth in a neuroblastoma cell line, and Rac activity is increased by netrin stimulation.225 Taking into account its homology with Drosophila Dock, vertebrate Nck is likely to associate with a GEF to activate Rac upon netrin stimulation.224 Nck also binds the Rac effector kinase PAK32,117 (see Fig. 21–4), which could be

FIGURE 21–7 Netrin/DCC/UNC-5. Depending on the receptor complex, netrin can trigger both axon attraction and repulsion. Upon ligand stimulation, the DCC receptor multimerizes (for clarity, only two DCC receptors are shown) and promotes netrin-mediated axon attraction and outgrowth via modulation of actin dynamics, transcription, and local protein synthesis. In contrast, a heteromeric DCC/UNC-5 receptor complex triggers netrin-mediated repulsion. The DCC/Robo receptor complex, which interferes with netrin-mediated attraction, is shown in Figure 21–6.

Guidance of Axons to Targets in Development and in Disease

locally activated by Rac to mediate netrin-induced neurite outgrowth by activation of cofilin (see Actin Depolymerization earlier) and downregulation of retrograde F-actin flow (see Retrograde F-Actin Flow earlier). Function of UNC-34/Mena in Netrin Signaling. Caenorhabditis elegans UNC-34, the aforementioned Mena homologue (see Promotion of Actin Polymerization by Mena Anti-capping Function, earlier, and Fig. 21–4), suppresses a gain-of-function mutation of the netrin receptor UNC40/DCC, suggesting a role for UNC-34/Mena in netrin signaling.129 Being homologous to members of the Ena/VASP family, UNC-34 probably functions in netrin signaling via modulating actin polymerization (compare with discussion in Promotion of Actin Polymerization by Mena Anti-capping Function, earlier, and Fig. 21–4). UNC-34 was initially identified as a suppressor of UNC-5–mediated growth cone steering.66 UNC-34 function depends on the conserved P1 motif of UNC-40/DCC (see Fig. 21–7). The UNC-34/DCC interaction is likely to be indirect because the P1 motif of UNC-40/DCC does not contain any known Ena-binding motifs.129 Involvement of MAX-1 in a DCC-Independent Netrin Signaling Pathway. Further genetic analysis in C. elegans mutants suggests a UNC-5–dependent pathway for netrin-mediated axon repulsion independent of UNC40/DCC.169 MAX-1 (required for motor neuron axon guidance), a cytosolic protein containing two pleckstrin homology domains, a myosin tail homology 4 domain, and a Band4.1/ezrin/radixin/moesin domain, genetically interact with UNC-5 and UNC-6/netrin, but not UNC40/DCC. Because of its characteristic domain structure, MAX-1 may be membrane localized and involved in phosphatidylinositol phospholipid signaling. Such a role would be in line with the observation that phospholipase C-␥ and phosphoinositide 3-kinase are required for the netrin turning response.253 The precise role of MAX-1 in netrin signaling, however, needs to be further scrutinized. Tyrosine Phosphatase Shp-2 Binding of UNC-5. Netrin stimulation increases UNC-5 tyrosine phosphorylation, possibly via Src kinase. The PTP Shp-2 subsequently binds to the phosphorylated ZU-5 motif of UNC-5373 (see Fig. 21–7). Because Shp2 regulates RhoA activity,329 migration, and cellular adhesion,413 this interaction provides a possible explanation for netrin-controlled regulation of adhesion and actin dynamics (compare with discussion in FAK and Cas earlier). Netrin Signaling at the Transcriptional Level NFAT-Mediated Transcription. Nuclear factor of activated T-cells (NFAT) plays a role in netrin-mediated embryonic axon outgrowth.133 The serine/threonine phosphatase calcineurin, which is activated by an increase

461

of intracellular Ca 2⫹ levels, dephosphorylates the cytoplasmic NFATc subunits and promotes their import into the nucleus, where they assemble with nuclear NFAT subunits to the active transcription factor.64,106,192 Netrins, which also modulate Ca2⫹ influx in growth cones,163 stimulate this calcineurin-dependent nuclear localization of NFAT subunits, thereby activating NFAT-mediated gene transcription (see Fig. 21–7). In contrast to netrinmediated axon outgrowth, netrin-induced axon turning is NFAT independent. Promotion of Local Protein Synthesis and Degradation by Netrin Signaling. Netrin stimulates both local protein synthesis and protein degradation in growth cones, the latter via the ubiquitin-proteasome system. These events are required for both attraction and repulsion in response to netrin.48,49 In this respect, the netrin response diverges from the response to semaphorin, for which protein synthesis but not degradation is necessary to mediate its chemotropic effect48 (see Requirement of Local Protein Synthesis for Sema3A Signaling later). The MAPK pathway, which is also essential for netrin signaling107 (see Fig. 21–7) is one of the major pathways controlling protein synthesis in response to growth factors.346 Upon netrin binding, DCC recruits ERK, which mediates activation of MAPK signaling and phosphorylation of the translation initiation factor eIF-4E via MAP kinase-interacting kinase.372,388 Phosphorylation of eIF-4E, observed in growth cones upon netrin stimulation, increases eIF-4E affinity for the mRNA cap, thereby leading to stimulation of protein synthesis127 and facilitating netrin-mediated growth cone steering.48 Modulation of Netrin Signaling Modulation of Response to Netrin Signaling by Intracellular cAMP Levels. Decreased cAMP levels convert netrin-mediated attraction to repulsion in Xenopus spinal neurons, suggesting a role for cyclic nucleotide levels in the modulation of netrin signaling.254 As discussed before, the regulation of intracellular levels of cyclic nucleotides is one means to adjust the turning response to guidance cues (compare with discussion in Modulation of Robo/Slit Signaling by Intracellular cGMP Levels earlier and Modulation of Responses to Semaphorin by Intracellular cGMP Levels later). The ratio of cAMP to cGMP turned out to be important in the turning response, with high ratios favoring attraction and low ratios favoring repulsion.270 Netrin itself may influence intracellular cAMP levels because it associates with both DCC and the adenosine A2b receptor, which activates cAMP production69 (see Receptors for Netrins: DCC, UNC-5, and A2b Receptor earlier). Reflecting intracellular cAMP levels, differences in cAMP-dependent protein kinase A activity also result in a modulation of the response to netrin and other guidance cues.348

462

Neurobiology of the Peripheral Nervous System

Sequestration of Netrin by Netrin Receptors. In some contexts netrin receptors themselves may influence the response to netrin. During the guidance of Drosophila photoreceptor axons for retinal projections, the DCC orthologue frazzled is not required in growing axons.130 Instead, frazzled is necessary in the lamina target, where it captures netrin and mediates its active relocation.156 This netrin rearrangement by frazzled/DCC creates positional information and allows the presentation of netrin to other growth cones and possibly its recognition by other receptors. Silencing of Netrin-Mediated Attraction by Formation of Robo/DCC Complexes. Netrin-mediated attraction can also be silenced by other signaling pathways. As discussed for Robo/Slit signaling (see Cross Talk between Robo/Slit and DCC/Netrin Signaling, earlier, and Fig. 21–6), DCC and the Slit-receptor Robo form a heteromultimeric receptor complex. Activation of this complex by Slit silences netrin-mediated attraction without affecting the growthstimulatory effect of netrin.354 Proteolytic Cleavage of the DCC Receptor. Similar to the cleavage of ephrin-A2 in ephrin/Eph signaling (compare with discussion in Modulation of Ephrin/Eph Signaling earlier), the number of functional DCC receptors present on the cell surface is regulated by metalloproteases, which promote shedding of the extracellular domain.118 Specific metalloprotease inhibitors enhanced netrin-mediated axon outgrowth and increased the DCC protein levels in rat spinal cord explants. Proteolytic cleavage thereby provides another means to modulate the activity of netrin signaling.

Semaphorin/Plexin Similar to the netrins, the semaphorins also were discovered in a parallel search for axon guidance molecules in both vertebrates and invertebrates. The first semaphorin identified was grasshopper semaphorin 1a (fasciclin IV), shown to be expressed in stripes in the epidermis of the leg. Upon antibody treatment, pioneer sensory axons grow misdirected.199 The first vertebrate semaphorin was discovered by biochemical purification and testing of the fractions for their ability to collapse growth cones.233 Semaphorins are a large family of phylogenetically conserved axon guidance molecules. They are categorized into eight subclasses on the basis of structure, sequence similarities, and species of origin333 (reviewed by Pasterkamp and Kolodkin281 and Raper295). The semaphorin family contains both secreted and membrane-associated proteins, with subclasses 1 and 2 found in invertebrates, subclasses 3 through 7 found in vertebrates, and subclass V being of viral origin. A hallmark all semaphorins share is their conserved Sema domain at the N-terminus.

Many semaphorins are potent chemorepellents,57,233,408 but, as discussed later, there is also evidence that some semaphorins function in a chemoattractive manner.81,282,288 Receptor Complexes The semaphorin receptor complexes for the different semaphorin subclasses, or even for different members of the subclasses, vary substantially in their composition. In summary, most semaphorin holoreceptor complexes include members of the plexin or neuropilin receptor families and receptors such as Otk (off-track), L1, Met, CD72, or Tim-2 (T-cell immunoglobulin domain and mucin domain). An exception is Sema7A, which associates with integrins in vivo282 (see Promotion of Axon Outgrowth by Sema7A via Integrin Signaling later). Semaphorin Receptors: Plexins and Other Components. Plexins, components of most semaphorin receptor complexes (Fig. 21–8A, B, and D), constitute a family of transmembrane receptors with at least nine members in four subfamilies (plexin-A, -B, -C, and -D) identified so far in the mammalian genome (reviewed by Fujisawa and Kitsukawa114 and Tamagone and Comoglio366). Like semaphorins, plexins contain conserved Sema domains. A direct semaphorin-plexin interaction has been shown for invertebrate semaphorins, vertebrate membrane-associated semaphorins, and viral semaphorins.67,365,399 Plexins are also part of the receptor complex for secreted class 3 semaphorins (see Fig. 21–8A). In this complex, neuropilins are additional receptor components149,198 that mediate Sema3 binding, and plexins constitute the signaling component303 (see Fig. 21–8A). In addition, L1, a cell adhesion molecule of the Ig superfamily, represents a part of an L1/neuropilin/plexin receptor complex mediating Sema3A (but not Sema3B, 3C, or 3E) signaling54,55 (see Fig. 21–8A; also see L1, the Semaphorin Receptor, and Modulation of Semaphorin Signaling later). Like neuropilins, Otk and Met also form receptor complexes with plexins. The receptor tyrosine kinase Met (scatter factor 1/hepatocyte growth factor receptor) transduces Sema4D signaling in association with plexin-B1 (see Fig. 21–8B) and promotes invasive growth of epithelial cells.128 Drosophila transmembrane protein Otk mediates Sema1A repulsive functions required for correct projections of motor neurons to their muscle targets. Otk, which forms a receptor complex with plexin-A (see Fig. 21–8D), resembles receptor tyrosine kinases but has a catalytically inactive kinase domain.400 Other Semaphorin Receptors. Cellular processes distinct from axon guidance may require their own semaphorin receptors. In lymphoid tissues CD72, a type 2 transmembrane protein belonging to the calciumdependent C-type lectin superfamily, functions as a receptor for the class 4 semaphorin CD100 to regulate

FIGURE 21–8 Semaphorin/plexin signaling. A, Secreted class 3 semaphorins are potent chemorepellents. Sema3A dimerizes in order to induce growth cone repulsion or collapse.193 Neuropilin, A-type plexins, and the cell adhesion molecule L1 form the receptor complex for Sema3A. This complex recruits a broad variety of downstream effectors to modulate actin and tubulin dynamics. B, Class 4 semaphorins mediate growth cone turning and repulsion by sequestration of Rac and activation of RhoA. C, GPI-anchored class 7 semaphorins function via their arginine-glycine–aspartic acid (RGD)-mediated interaction with ␤-integrins. Activation of focal adhesion kinase (FAK) and mitogen-activated protein kinase (MAPK) promotes axonal outgrowth. D, In Drosophila, the association of class 1 semaphorins with MICAL (molecule associated with CasL) causes axon repulsion. MICAL is a putative flavoprotein monooxygenase, pointing to a possible role of redox signaling in semaphorin-mediated axon repulsion. A similar mechanism may exist in vertebrates as well.

464

Neurobiology of the Peripheral Nervous System

B-cell responsiveness.213 Also implicated in Sema4A signaling, Tim-2, a member of the family of T-cell immunoglobulin domain and mucin domain (Tim) proteins, enhances the generation of antigen-specific T-cells in response to Sema4D/CD100.212 Semaphorin Signaling Rho GTPases and Semaphorin Signaling. Like other axon guidance signaling pathways, semaphorin signaling exerts its effects on axon guidance via modulation of cytoskeletal dynamics in the growth cone upon receptor activation.100,101,113 There is evidence for direct and indirect interactions between plexins and Rho GTPases, the key regulators of the actin cytoskeleton. The cytoplasmic domain of plexins shares significant sequence similarity with the GAP domain of Ras GAPs.304 GAP activity of plexins might allow a direct regulation of Rho GTPases (compare with discussion in Rho GTPases: Key Modulators of Actin Dynamics, earlier, and Fig. 21–3); however, such activity and putative targets remain to be shown. In vitro studies with chick DRG neurons implicated Rac1 in semaphorin signaling. DRG neurons transfected with dominant negative or constitutively active Rac1 decreased or enhanced Sema3A-induced growth cone collapse, respectively.177 Rac is implicated in growth cone advance (see Actin Depolymerization, and Retrograde F-Actin Flow, earlier, and Fig. 21–4), so its role in Sema3A-mediated growth cone collapse is not self-evident. There is also direct and specific interaction between GTP-bound active Rac and plexinB1.90,381 Plexin-B1 clustering in fibroblasts does not induce lamellipodia formation as one would expect in the case when Rac is activated. Instead, actin-myosin fibers are formed and the cell retracts, suggesting that Rho might be activated.90 Plexin-B1 sequesters active Rac from its downstream target PAK, thereby inhibiting Rac-induced PAK activation380 (see Fig. 21–8B). Rac mediates the inactivation of the actindepolymerizing factor cofilin, so if Rac is sequestered, cofilin remains active, which promotes semaphorin-induced growth cone collapse (see Actin Depolymerization, earlier, and Fig. 21-4). In Drosophila, a direct interaction between plexin-B and RhoA was found that leads to RhoA activation.167 In vertebrates, the plexin-B–RhoA interaction is indirect: Two Rho-specific GEFs, PDZ-Rho-GEF and LARG, interact with the C-terminal PDZ recognition sequence of plexin-B via their PDZ domains. These GEFs activate RhoA, which in turn mediates growth cone collapse10,62,157,286,363 (see Fig. 21–4). Unlike plexin-B, other plexin subfamilies do not contain the CRIB motif or the C-terminal PDZ domain, which suggests a different mode of regulation of cytoskeletal dynamics. Rnd1 and RhoD, members of the Rho GTPase family implicated in spine formation173 and motility of early endo-

somes,121 respectively, bind the cytoplasmic domain of plexin-A1 and compete for the same binding site416 (see Fig. 21–8A). Rnd1 increases the GAP activity of a Rhospecific GAP, p190GAP,392 so the antagonism of Rnd1 and RhoD binding might regulate plexin-A1 activity and thereby Sema3A signaling. However, another group arrived at a different result, showing that Rnd1 interacts directly with plexin-B1 but not plexin-A1, suggesting that Rnd1 might activate RhoA via PDZ-RhoGEF.274 Kinases Downstream of Semaphorin Signaling. Fes tyrosine kinase binds to the intracellular domain of plexinA1 and recruits the collapsin response mediator protein (CRMP)/CRMP-associated molecule (CRAM) complex132,255 (see Fig. 21–8A). In the absence of Sema3A, neuropilin-1 (Npn-1) blocks the association of plexin-A and Fes. Upon stimulation with Sema3A, which uses Npn-1 as receptor, Fes binds plexin-A and phosphorylates plexin-A1, CRMP, and CRAM.255 CRMP binds tubulin heterodimers directly.116 CRMP is also a target of ROCK8 and alters the responses to Rho and Rac signaling.142 Taken together, these results implicate CRMP/CRAM in the regulation of actin and microtubule dynamics in semaphorin signaling. Sema3A-induced growth cone collapse can be prevented by kinase inhibitors specific for tyrosine kinases or cyclindependent kinases; it was therefore suggested that plexin-A signaling depends on both the Src family tyrosine kinase Fyn and the serine/threonine kinase cyclin-dependent kinase 5 (Cdk5).323 Fyn kinase associates with plexin-A1 and -A2 and phosphorylates their cytoplasmic domains (see Fig. 21–8A). Additionally, plexin-A2 associates with Cdk5 via an interaction with active Fyn, in which Fyn phosphorylates and thereby activates Cdk5. The microtubule-associated protein (MAP) tau is one of the targets of Cdk5. MAPs are the phosphoregulated key modulators of microtubule stability and dynamics (for review, see Drewes and colleagues89), so tau phosphorylation by Cdk5 upon Sema3A-stimulation might provide a link to microtubule dynamics.59 Cdk5 and its neuron-specific regulator p35 associate with GTP-bound Rac and its effector kinase PAK (see Fig. 21–8A). In this complex, hyperphosphorylation of PAK by Cdk5/p35 downregulates PAK activity,267 thereby modulating dynamics of the actin cytoskeleton to allow repulsion and growth cone collapse (compare Fig. 21–4). An inactive, phosphorylated pool of glycogen synthase kinase-3 (GSK-3), another kinase able to phosphorylate tau,145 co-localizes with F-actin in filopodia and at the leading edge of lamellipodia in growth cones of DRG neurons. Stimulation with Sema3A activates this GSK-3 pool at the leading edge of neuronal growth cones93 (see Fig. 21–8A). There is also evidence for cross talk between GSK-3 and Rho signaling. G␣13, an ␣-subunit of heterotrimeric G proteins, interacts with the Rho-specific GEF p115RhoGEF,

465

Guidance of Axons to Targets in Development and in Disease

and GSK-3 activation by G␣13 is Rho mediated.201,325 However, further studies will be necessary to elucidate the involvement of GSK-3 in Sema3A signaling.

Sema3A as a result of asymmetrical localization of soluble guanylate cyclase,288 so cGMP levels may be an endogenous regulator of Sema3A signaling.

Promotion of Axon Outgrowth by Sema7A via Integrin Signaling. In contrast to many other semaphorins, Sema7A promotes axon outgrowth. In vitro studies suggested plexin-C1 to be the binding partner of Sema7A.365 The growth-promoting effect of Sema7A, however, is independent of plexin-C1.282 In vivo, Sema7A associates with integrins by virtue of its arginine-glycine– aspartic acid motif. Integrins activate FAK and subsequently the MAPKs ERK 1 and 2, thereby mediating the growth-promoting effect of Sema7A signaling in neurons282 (see Fig. 21–8C).

Requirement of Local Protein Synthesis for Sema3A Signaling. Inhibition of protein synthesis prevented Sema3A-induced growth cone turning or collapse. Upon Sema3A stimulation, local protein synthesis in growth cones was rapidly increased by inactivation of the translation repressor eIF-4EBP1 and subsequent activation of the translation initiation factor eIF-4E. Although not shown directly, local protein synthesis may alter the receptor composition on the growth cone surface, a mechanism similarly found in ephrin signaling (see Modulation of Ephrin/Eph Signaling earlier). In contrast, inhibition of the ubiquitin-proteasome system did not alter the chemotropic response to Sema3A.48 Therefore, local protein synthesis but not degradation seems to be required for Sema3A signaling.

A Putative Role for Redox Signaling in SemaphorinMediated Axon Repulsion. Drosophila MICAL (molecule associated with CasL), a multidomain cytosolic protein expressed in axons that also contains a flavoprotein monooxygenase domain, was found to interact with plexinA.370 MICAL turned out to be necessary for repulsive axon guidance of motor neurons mediated by Sema1A/plexin-A (see Fig. 21–8D). Vertebrate Sema6A is homologous to insect class 1 semaphorins.419 Vertebrate MICALs show specific neuronal and non-neuronal expression patterns during development, and human MICAL-1 and mouse MICAL-2 seem to specifically interact with human plexinA3 and mouse plexin-A4, respectively.370 In cultured DRG neurons, inhibition of 12/15-lipoxygenase blocks Sema3Ainduced growth cone collapse, which can be induced by exogenously applied product of 12/15-lipoxygenase.250 These results point to an involvement of redox signaling in semaphorin-mediated axon repulsion; however, its exact role remains to be determined. MICAL interactions may provide a link to cytoskeletal dynamics because MICAL associates with intermediate filaments and the adaptor protein CasL (compare with discussion in FAK and Cas earlier), a member of the Cas family, which is implicated in actin filament assembly and integrin-dependent processes such as migration, cell-cycle control, and transformation.161,277,324,362 Modulation of Semaphorin Signaling Modulation of Responses to Semaphorin by Intracellular cGMP Levels. Differences in intracellular cAMP levels are known to result in opposite turning of growth cones in response to the same guidance cue.348 As in netrin and Slit signaling (see Modulation of Response to Netrin Signaling by Intracellular cAMP Levels, and Modulation of Robo/Slit Signaling by Intracellular cGMP Levels earlier), the response to semaphorin is also subject to modulation by cyclic nucleotides. The collapse of cultured rat sensory growth cones upon Sema3A stimulation can be inhibited by activation of the cGMP pathway.347 Axons and dendrites differently respond to

AXON GUIDANCE IN DISEASE Surprisingly, the rapid progress in identifying and characterizing the large number of key molecules involved in axon guidance during development has revealed the involvement of these same molecules in only a small number of diseases. However, considering the complex process of brain development, one explanation may be the devastating consequences of mutations in main guidance systems on neural development. A fetus carrying a severe mutation is not likely to reach peri- or postnatal stages. Another explanation may be a certain functional redundancy between guidance molecules, which may lead to compensation for less severe mutations. The potential role of axon guidance molecules as inhibitors of the axonal regrowth after CNS traumatic lesions is discussed in recent reviews.83,197 We concentrate here on the role of axon guidance molecules in diseases, specifically epilepsy, schizophrenia, CRASH syndrome, and Kallmann’s syndrome.

Epilepsy Background Epilepsy is a neurologic disorder characterized by a tendency to recurrent seizures that can cause temporal impairment or loss of consciousness, abnormal motor phenomena such as convulsions, psychic or sensory disturbances, and perturbation of the autonomic nervous system (reviewed by Riviello301 and Shneker and Fountain340). Approximately 1% to 3% of the population worldwide is affected by epilepsy. Seizures are the result of excessive electrical discharges in the brain, which usually occur briefly and suddenly. If a seizure arises from one or more localized areas of the brain

466

Neurobiology of the Peripheral Nervous System

and does not cause the loss of consciousness, it is called a partial or focal seizure. A generalized seizure, in contrast, brings about a loss of consciousness and involves the entire brain. A state referred to as “status epilepticus” is characterized by frequent seizures without recovery of consciousness between the episodes. Mossy Fiber Sprouting One of the pathophysiologic hallmarks of temporal lobe epilepsy is the hyperexcitability of the hippocampal formation, which may be caused by a phenomenon called mossy fiber sprouting (MFS).11,343 Mossy fibers, the axons of dentate granule cells, normally innervate CA3 pyramidal cells and hilar cells. Upon loss of their proper targets, however, mossy fibers start to sprout into the inner molecular layer of the dentate gyrus, where they form functional synapses with dendrites of granule cells in animal models.275,368,393 This aberrant innervation may contribute to the formation of recurrent excitatory circuits, which may in turn facilitate the spontaneous recurrence of partial seizures and propagate the epileptic state12 (for review, see Nadler259). Loss of hippocampal tissue may be due to an initial precipitating injury, which may cause hippocampal sclerosis and trigger MFS.241 Glycosaminoglycans, Proteoglycans, and N-Cadherin Axon targeting during formation of the hippocampus is regulated by multiple guidance signals. In the adult brain, inhibitory guidance cues stabilize the complex, balanced neuronal network, which is very sensitive to further axon outgrowth and wiring. The levels of several guidance molecules are altered and may thereby facilitate MFS in the epileptic brain. Glycosaminoglycans (e.g., hyaluronic acid and chondroitin sulfate) and proteoglycans (proteins covalently linked to glycosaminoglycans) are elements of the extracellular matrix. They function as coreceptors for growth factors in the nervous system and regulate a broad variety of physiologic processes, including cell migration, neurite outgrowth, axonal pathfinding, synaptogenesis, and structural plasticity (for review, see Bandtlow and Zimmerman16 and Bovolenta and Fernaud-Espinosa34). Alterations in the levels of the glycosaminoglycans chondroitin sulfate and hyaluronic acid as well as receptor PTP ␨/␤, a chondroitin sulfate proteoglycan implicated in axonal sprouting, suggest a perturbance of guidance signals in the epileptic brain.260,285,345 Similarly, N-cadherin is upregulated in the inner molecular layer of the dentate gyrus, corresponding to MFS in this hippocampal region.337 N-cadherin is involved in synapse formation during development103,375; in line with this, N-cadherin localizes to newly formed synapses between mossy fibers and granule cell dendrites.337 Semaphorin Signaling Semaphorin signaling also seems to play a role in MFS. Mossy fibers express plexin-A and neuropilin,257 which

form the receptor complex for the chemorepellent Sema3A (see Semaphorin Receptors: Plexins and Other Components, and Fig. 21–8A earlier). Sema3A, which is secreted by cells of the entorhinal cortex,126 diffuses into the adjacent dentate gyrus, thereby repelling mossy fibers harboring the Sema3A receptor. In a rat model, Sema3A mRNA was transiently downregulated in the entorhinal cortex after induction of status epilepticus by electrical stimulation.160 This study could not show a reduction of Sema3A at the protein level; therefore, further inquiries are necessary to support the relevance of this finding. A possible cutback of Sema3A-mediated repulsion (see Semaphorin/Plexin earlier) plausibly explains sprouting of mossy fibers into the inner molecular layer of the dentate gyrus, which no longer represents a repelling environment. Further support for a role of semaphorin signaling in epilepsy comes from the phenotype of plexin-A3 knockout mice, which are susceptible to kainate-induced seizures (H. J. Cheng, M. Tessier-Lavigne, and S. Pleasure, personal communication, 2004). Plexin-A3 is the receptor for Sema3A; therefore, impaired Sema3A signaling (as during Sema3A downregulation) may contribute to MFS and the formation of additional neuronal circuits, which would account for the hyperexcitability of the hippocampal formation in patients with epilepsy. Summary Disturbances of the levels of guidance cues contribute to aberrant axon sprouting and synapse formation in the epileptic brain. This may facilitate the formation of recurrent excitatory circuits that in turn promote seizures and propagate the epileptic state.

Schizophrenia Background Schizophrenia is a severe mental disorder that is generally characterized by fundamental and characteristic distortions of thinking and perception, such as delusions and hallucinations, and by inappropriate or blunted affect.406 The lifetime prevalence of schizophrenia worldwide has been estimated to be about 1%, with the average onset occurring from adolescence through early adulthood. Several anatomic abnormalities, including ventricular enlargement and decreased cerebral (cortical and hippocampal) volume, are hallmarks of schizophrenia.146 The causes of schizophrenia are still largely obscure; however, it is now accepted that genetic factors play a role in forming a predisposition to developing the illness,50 and external factors influence the timing of its possible onset. Multiple neuropathologic studies suggest an abnormal early neurodevelopment of patients with schizophrenia, especially a disturbance in neuronal migration, synaptogenesis, and process formation.9,68,146,384

Guidance of Axons to Targets in Development and in Disease

Wnt Signaling The Wnt signaling pathway is crucial for a large number of developmental processes, including embryonic patterning, cell fate determination, and the regulation of cell proliferation, polarity, and lymphocyte development.41,376,402 Additionally, Wnt signaling is implicated in anterior-posterior guidance of commissural axons234 and organization of the patterning during cortical development.136 Frizzled proteins constitute a large family of seven-pass transmembrane proteins, some of which serve as receptors for Wnt ligands, a family of secreted glycoproteins.25,412 One member of the frizzled family, frizzled-3 (FZD3), functions in development of major fiber tracts in the rostral CNS387 and formation of the neural crest in Xenopus.84 Several studies point to an association of the FZD3 locus and the susceptibility to schizophrenia.184,410 The molecular basis for this correlation is not yet elucidated; however, mutations in FZD3 might contribute to defects in axon guidance and neural connection, thus elevating susceptibility to schizophrenia. DISC-1/NUDEL/LIS-1 Genetic studies using family analysis were able to link the locus for DISC-1 (disrupted in schizophrenia-1) to susceptibility to schizophrenia.30,251 The mutant form of DISC-1 occurring in these families is C-terminally truncated and fails to bind NudE-like (NUDEL),278 a DISC-1 interaction partner involved in nucleokinesis and neuronal migration.147,215 NUDEL also forms a complex with LIS-1,266,322 a factor implicated in lissencephaly. Lissencephaly is a human brain malformation exhibiting a smooth cerebral cortex and abnormal neuronal migration.296 The NUDEL/LIS-1 complex participates in organizing the microtubule cytoskeleton during nuclear translocation and neurite outgrowth.306 In line with these results, another study showed a direct implication of DISC-1 in neurite extension.256 Upon overexpression of DISC-1, PC12 cells show longer neurites and DISC-1 localizes in growth cones. In rats, DISC-1 is highly expressed in the cerebral cortex and the hippocampus in late development278; these are two regions shown to be affected in schizophrenia. Mutations in DISC-1 may prevent an adequate response of migrating neurons or growing axons to guidance cues in development, thereby raising the susceptibility to schizophrenia. Semaphorin Signaling Sema3A, a potent chemorepellant in axon guidance (see Semaphorin/Plexin earlier), was recently shown to be upregulated in the cerebellum of adult schizophrenia patients.91 Along with the Sema3A upregulation, a decrease of reelin expression was found,91 for both cerebellum and prefrontal cortex.137 The reelin signaling pathway is involved in neuronal migration, and mutations in reelin cause an autosomal recessive form of lissencephaly.164 Moreover, reelin mutant mice (reeler mice) show

467

decreased axon branching of entorhinal axons targeting the hippocampus, altered hippocampal synaptogenesis, and misrouted fibers.33 Alteration in Sema3A and reelin levels in patients with schizophrenia might impair normal connectivity and synaptogenesis during development, thus contributing to predisposition for schizophrenia. NCAM and L1: Immunoglobulin Superfamily Members Molecules of the Ig superfamily such as neural cell adhesion molecule (NCAM) and L1 are involved in axon guidance, synapse stability, cell migration, and maintenance and function of neuronal networks.310,342 In the cerebrospinal fluid of patients with schizophrenia, levels of L1 and a cleaved isoform of NCAM are decreased and elevated, respectively. Additionally, the cleaved form of NCAM is increased in the hippocampus and the prefrontal cortex.289,377 Augmented cleavage of Ig proteins may be due to abnormal proteolysis in schizophrenia, either by enhanced proteolytic activity or decreased resistance to cleavage. Removal of polysialic acid from NCAM by neuraminidase may account for decreased protease resistance,377 and polysialylated NCAM is decreased in hippocampi of patients with schizophrenia.17 Genetic and twin studies indicate that changes of Ig protein levels are not likely to be caused by a genetic predisposition resulting from structural alterations of the genes,290,378 but might rather be a consequence of schizophrenia onset. The relevance of altered Ig protein levels for onset and progression of schizophrenia therefore awaits further clarification.

CRASH Syndrome/MASA Syndrome/L1 Syndrome Background The NCAM L1 (compare with discussion in Semaphorin Receptors: Plexins and Other Components, earlier, and Fig. 21–8A), a single-pass transmembrane glycoprotein of the Ig superfamily, is involved in the regulation of neurite growth, axon guidance, cell migration, the activation of second messenger signaling cascades, and possibly also cytoskeletal rearrangements.45,47,77,181 Mutations in the X-chromosomal gene for L1 cause several X-linked mental retardation syndromes, which were first considered different diseases, including X-linked hydrocephalus (hydrocephalus resulting from stenosis of the aqueduct of Sylvius), MASA syndrome26 (an acronym for mental retardation, aphasia, shuffling gait, and adducted thumbs), X-linked spastic paraplegia/paraparesis, and X-linked agenesis of the corpus callosum. The clinical syndrome of L1 mutation is now sometimes also referred to as CRASH syndrome, owing to its main clinical features (corpus callosum agenesis, mental retardation, adducted thumbs, spastic paraplegia, and hydrocephalus).110 Given the widespread expression of L1,

468

Neurobiology of the Peripheral Nervous System

alterations in brain morphology of patients with CRASH syndrome are surprisingly modest. The most prominent consequences of mutations in L1 are reduction or lack of two long axonal tracts, the aforementioned corpus callosum, which connects the cerebral hemispheres, and the corticospinal tract, which is involved in controlling voluntary motor functions. Reduction of the corpus callosum often coincides with mental retardation, and malformation in the corticospinal tract can account for the spasticity syndromes. A large number of mutations giving rise to defects in L1 function has been described by now, causing a broad range of clinical symptoms (for a list of L1 mutations and further links, see www.uia.ac.be/dnalab/l1/). These L1 mutations can be classified into three categories: class 1 includes mutations in the cytoplasmic domain, class 2 exhibits point mutations in the extracellular domain, and class 3 is characterized by a premature stop codon in the extracellular domain. Because of the many physiologic processes L1 is involved in, severity of the symptoms can vary widely and clearly correlates with the site of mutation, depending on the function influenced. Patients with class 1 mutations are least affected, and class 3 mutations cause the most severe symptoms.409 Fibroblast Growth Factor Receptor and Src Kinase L1 itself does not possess any catalytic domains; however, it engages several binding partners to activate signaling pathways involved in neurite growth and guidance. Surfacebound L1 constitutes a substrate that promotes neurite outgrowth in vitro. This provides an assay to test L1-function. Neurons from L1 knockout mice lose the ability to extend neurites on L1 substrate, suggesting that a homophilic interaction between cellular and extracellular L1 is necessary to promote neurite outgrowth.75,219 This effect is not simply due to cellular adhesion, because a soluble form of L1 can also trigger neurite extension.87 A putative receptor for L1 is fibroblast growth factor receptor (FGFR). FGFR signaling not only promotes survival and proliferation but also is involved in guiding axons of retinal ganglion cells to the optic fissure.38 Activation of FGFR is necessary for L1-mediated axon outgrowth,396 and antibody-mediated inactivation of FGFR396 or expression of dominant negative FGFR314 abolishes neurite extension triggered by L1. The tyrosine kinase Src may be part of this signaling pathway, because neurons deficient for Src show a diminished rate of neurite extension on L1 substrate.171 Axonin-1 and Associated Kinases Heterophilic association of L1 with axon-associated cell adhesion molecule axonin-1 effectively promotes neurite outgrowth of embryonic DRG neurons.207,356 Axonin-1 interacts with the tyrosine kinase Fyn,214 which provides a link for L1 signaling to cellular adhesion and cytoskeleton dynamics (compare with discussion in Kinases

Downstream of Semaphorin Signaling, and GPI-Linked A-Type Ephrins: Capacity for Reverse Signaling earlier). Serine/threonine casein kinase II,403 serine/threonine ribosomal S6 kinase (rsk),404 and the Eph receptor tyrosine kinase homologue chicken embryo kinase 5421 phosphorylate the cytoplasmic domain of L1 and are therefore likely to modulate L1 function. Indeed, cell-cell contacts induce L1–axonin-1 clustering, which modulates the activity of axonin-associated Fyn kinase and L1-associated casein kinase II- and rsk-related activity.214 Cytoskeleton Association via Ankyrin The highly conserved cytoplasmic domain of L1 binds to ankyrin, thereby associating with the spectrin-based membrane cytoskeleton.77 Tyrosine phosphorylation of the intracellular L1 domain abolishes ankyrin binding and increases the lateral mobility of L1. Phosphorylation by Src, Fyn, or an Eph receptor kinase might therefore regulate adhesive properties of L1-expressing cells during neuronal migration or axon outgrowth.120 L1, the Semaphorin Receptor, and Modulation of Semaphorin Signaling Cortical axons from L1-deficient mice fail to respond to a repulsive signal from the spinal cord, which could be identified as the secreted chemorepellent Sema3A. L1 and Npn-1 form a Sema3A-binding complex via direct interaction of their extracellular domains.54 Together with plexinA, this complex constitutes the receptor that triggers Sema3A signaling364,365 (see Semaphorin Receptors: Plexins and Other Components, earlier, and Fig. 21–8A), implicating L1 in semaphorin-mediated long-distance axon guidance. A soluble form of L1 interacting with Npn-1 converts Sema3A-induced chemorepulsion to attraction by activation of neuronal nitric oxide (NO) synthase. Subsequent NO synthesis stimulates cGMP guanylyl cyclase, which increases internal cGMP levels, thereby mediating the switch from repulsion to attraction54 (compare with discussion in Modulation of Robo/Slit Signaling by Intracellular cGMP Levels, Modulation of Response to Netrin Signaling by Intracellular cAMP Levels, and Modulation of Responses to Semaphorin by Intracellular cGMP Levels earlier). In vivo, L1 is subject to proteolytic cleavage by the serine protease plasmin261 and the metalloprotease ADAM 10,138 a vertebrate orthologue to Drosophila Kuzbanian. Proteolytic cleavage might modulate the response to Sema3A by producing a soluble L1 cleavage product or by disrupting the L1/Npn-1/plexin-A receptor complex.

Kallmann’s Syndrome Background The main characteristics of the congenital Kallmann’s syndrome (KS) are hypogonadotropic hypogonadism, that is, defective gonadal development or function resulting from

Guidance of Axons to Targets in Development and in Disease

gonadotropic hormone deficiency, and anosmia or hyposmia (the absence or reduction of the sense of smell). These symptoms were already described in 1856 by Maestre de San Juan,238 and further characterized as a clinical entity by Kallmann in 1944.179 Different forms of KS are inherited in an X-chromosomal or autosomal manner, with varying associated anomalies, including upper body mirror movements, renal agenesis, and midline/craniofacial abnormalities such as harelip or cleft palate. The prevalence of KS is about five times higher in males (~1:10,000) than in females. The anosmia observed in KS is due to reduction or absence of the olfactory bulb. Early innervation of the olfactory bulb primordium by axons from the olfactory epithelium is necessary for induction of bulb development. Because migration and targeting of axons from the olfactory epithelium is impaired in KS, the olfactory bulb does not form properly.131 Hypogonadotropic hypogonadism in KS is secondary to a lack of gonadotropin-releasing hormone (GnRH). Neurons secreting GnRH use axons of olfactory receptor neurons as guiding tracks on their way from the olfactory placode to the brain.331 Deprived of their guides, GnRHsynthesizing neurons fail to reach their targets, which results in hypogonadism in KS.330 Anosmin-1, Gene Product of Kal-1, and Its Dual Branch-Promoting and Guidance Activity Kal-1 was the first gene to be discovered that is responsible for the X-linked form of KS.109,217 Kal-1 codes for anosmin-1, an extracellular matrix protein containing four FNIII domains and a whey acidic protein domain, often found in proteinase inhibitors. Anosmin-1, which is secreted by the olfactory cortex, has a dual branchpromoting and guidance activity. First, it promotes axonal branching of cultured rat olfactory bulb neurons, resulting in collateral branches.349 Second, it functions as an axonal guidance cue, which allows the olfactory cortex to attract these collaterals. Anosmin-1, Heparan Sulfate Proteoglycans, and FGFR Activity of anosmin-1 is regulated in embryonic development, and seems to depend on an additional factor.349 Heparan sulfate proteoglycans (see Glycosaminoglycans, Proteoglycans, and N-Cadherin earlier) are good cofactor candidates because they regulate processes such as cell migration or axonal pathfinding16,34 and are likely to tether anosmin-1 to the cellular membrane.350 In line with this, mutations in heparan-6-O-sulfotransferase were able to suppress the axon-branching and misrouting phenotype of Kal-1 mutations in C. elegans.46 Anosmin-1 functions in C. elegans and vertebrates are comparable because they show high functional conservation: Similarly to Kal-1 in vertebrates, C. elegans Kal-1 seems to control axon branching and guidance, because overexpression of Kal-1 in C. elegans

469

promotes enhanced axon branching and faulty axon guidance.46 The high conservation is also stressed by the fact that human Kal-1 can rescue C. elegans Kal-1 mutations.312 In addition to their interaction with anosmin-1, heparan sulfate proteoglycans likely promote the dimerization of the binary complex of FGF and its receptor (FGFR)284 (also see Fibroblast Growth Factor Receptor and Src Kinase earlier). Interestingly, an autosomal dominant variant of KS is caused by loss-of-function mutations in FGFR,86 and FGFR is required for the formation of the olfactory bulb.150 These observations suggest a functional link between anosmin-1 and FGFR, because mutations in both impair olfactory bulb and gonad development. Anosmin-1 may be implicated in FGF signaling, resulting in the symptoms of KS.

OUTLOOK Although recent years have seen the rapid identification and characterization of the four major axon guidance systems, ephrin/Eph, Slit/Robo, Netrin/DCC/UNC-5, and semaphorin/plexin, we are still far from understanding how the brain is wired. First, although the identified axon guidance ligands and receptors give basic insights into how axons find their appropriate target, they only account for a minute fraction of guidance events that occur within the nervous system. Thus other axon guidance cues wait to be discovered. Genetic approaches in which a gene is silenced or mutated together with expression of a reporter within the affected neurons should lead to such discoveries.218 Second, although numerous signaling components downstream of the receptors are now known, our knowledge is still fragmentary. Improved detection of protein-protein interaction and mass spectroscopic analysis of whole signaling complexes122 will help to reveal the mechanisms underlying guidance phenomena. Finally, better understanding of the axon guidance cues and signaling will allow us to correlate more diseases and defective axon guidance molecules, which may in some cases lead to improved treatment and perspectives for affected patients.

ACKNOWLEDGMENTS We are very grateful to Christel Bauereiss and Robert Schorner for help with the figure preparation. We would like to thank Drs. Boyan Garvalov, Rüdiger Klein, Edgar Meinl, Brian Storrie, and Gaia Tavosanis for reading the manuscript. We are also indebted to Drs. Hwai-Jong Cheng, Samuel J. Pleasure, and Marc Tessier-Lavigne for sharing unpublished results with us. We are very grateful to Dr. Paul Forscher (Yale University) for granting permission to reproduce Figure 21–2. Frank Bradke is a recipient of a Career Development Award from the Human Frontier Science Program.

470

Neurobiology of the Peripheral Nervous System

REFERENCES 1. Abdul-Manan, N., Aghazadeh, B., Liu, G. A., et al.: Structure of Cdc42 in complex with the GTPase-binding domain of the ‘Wiskott-Aldrich syndrome’ protein. Nature 399:379, 1999. 2. Ackerman, S. L., Kozak, L. P., Przyborski, S. A., et al.: The mouse rostral cerebellar malformation gene encodes an UNC-5-like protein. Nature 386:838, 1997. 3. Aizawa, H., Wakatsuki, S., Ishii, A., et al.: Phosphorylation of cofilin by LIM-kinase is necessary for semaphorin 3A-induced growth cone collapse. Nat. Neurosci. 4:367, 2001. 4. Amano, M., Ito, M., Kimura, K., et al.: Phosphorylation and activation of myosin by Rho-associated kinase (Rho-kinase). J. Biol. Chem. 271:20246, 1996. 5. Anderson, R. G.: Plasmalemmal caveolae and GPI-anchored membrane proteins. Curr. Opin. Cell Biol. 5:647, 1993. 6. Arakawa, Y., Bito, H., Furuyashiki, T., et al.: Control of axon elongation via an SDF-1alpha/Rho/mDia pathway in cultured cerebellar granule neurons. J. Cell Biol. 161:381, 2003. 7. Arber, S., Barbayannis, F. A., Hanser, H., et al.: Regulation of actin dynamics through phosphorylation of cofilin by LIM-kinase. Nature 393:805, 1998. 8. Arimura, N., Inagaki, N., Chihara, K., et al.: Phosphorylation of collapsin response mediator protein-2 by Rho-kinase: evidence for two separate signaling pathways for growth cone collapse. J. Biol. Chem. 275:23973, 2000. 9. Arnold, S. E.: Neurodevelopmental abnormalities in schizophrenia: insights from neuropathology. Dev. Psychopathol. 11:439, 1999. 10. Aurandt, J., Vikis, H. G., Gutkind, J. S., et al.: The semaphorin receptor plexin-B1 signals through a direct interaction with the Rho-specific nucleotide exchange factor, LARG. Proc. Natl. Acad. Sci. U. S. A. 99:12085, 2002. 11. Babb, T. L., Lieb, J. P., Brown, W. J., et al.: Distribution of pyramidal cell density and hyperexcitability in the epileptic human hippocampal formation. Epilepsia 25:721, 1984. 12. Babb, T. L., Pretorius, J. K., Mello, L. E., et al.: Synaptic reorganizations in epileptic human and rat kainate hippocampus may contribute to feedback and feedforward excitation. Epilepsy Res. Suppl. 9:193, 1992. 13. Bagri, A., Marin, O., Plump, A. S., et al.: Slit proteins prevent midline crossing and determine the dorsoventral position of major axonal pathways in the mammalian forebrain. Neuron 33:233, 2002. 14. Bamburg, J. R., and Bray, D.: Distribution and cellular localization of actin depolymerizing factor. J. Cell Biol. 105:2817, 1987. 15. Bamburg, J. R., McGough, A., and Ono, S.: Putting a new twist on actin: ADF/cofilins modulate actin dynamics. Trends Cell Biol. 9:364, 1999. 16. Bandtlow, C. E., and Zimmermann, D. R.: Proteoglycans in the developing brain: new conceptual insights for old proteins. Physiol. Rev. 80:1267, 2000. 17. Barbeau, D., Liang, J. J., Robitalille, Y., et al.: Decreased expression of the embryonic form of the neural cell adhesion molecule in schizophrenic brains. Proc. Natl. Acad. Sci. U. S. A. 92:2785, 1995. 18. Bashaw, G. J., and Goodman, C. S.: Chimeric axon guidance receptors: the cytoplasmic domains of slit and netrin receptors specify attraction versus repulsion. Cell 97:917, 1999.

19. Bashaw, G. J., Kidd, T., Murray, D., et al.: Repulsive axon guidance: Abelson and Enabled play opposing roles downstream of the roundabout receptor. Cell 101:703, 2000. 20. Bateman, J., Shu, H., and Van Vactor, D.: The guanine nucleotide exchange factor trio mediates axonal development in the Drosophila embryo. Neuron 26:93, 2000. 21. Battye, R., Stevens, A., Perry, R. L., and Jacobs, J. R.: Repellent signaling by Slit requires the leucine-rich repeats. J. Neurosci. 21:4290, 2001. 22. Bear, J. E., Svitkina, T. M., Krause, M., et al.: Antagonism between Ena/VASP proteins and actin filament capping regulates fibroblast motility. Cell 109:509, 2002. 23. Behlke, J., Labudde, D., and Ristau, O.: Self-association studies on the EphB2 receptor SAM domain using analytical ultracentrifugation. Eur. Biophys. J. 30:411, 2001. 24. Bentley, D., and Toroian-Raymond, A.: Disoriented pathfinding by pioneer neurone growth cones deprived of filopodia by cytochalasin treatment. Nature 323:712, 1986. 25. Bhanot, P., Brink, M., Samos, C. H., et al.: A new member of the frizzled family from Drosophila functions as a Wingless receptor. Nature 382:225, 1996. 26. Bianchine, J. W., and Lewis, R. C. Jr.: The MASA syndrome: a new heritable mental retardation syndrome. Clin. Genet. 5:298, 1974. 27. Binns, K. L., Taylor, P. P., Sicheri, F., et al.: Phosphorylation of tyrosine residues in the kinase domain and juxtamembrane region regulates the biological and catalytic activities of Eph receptors. Mol. Cell. Biol. 20:4791, 2000. 28. Bishop, A. L., and Hall, A.: Rho GTPases and their effector proteins. Biochem. J. 348(Pt. 2):241, 2000. 29. Bito, H., Furuyashiki, T., Ishihara, H., et al.: A critical role for a Rho-associated kinase, p160ROCK, in determining axon outgrowth in mammalian CNS neurons. Neuron 26:431, 2000. 30. Blackwood, D. H., Fordyce, A., Walker, M. T., et al.: Schizophrenia and affective disorders—cosegregation with a translocation at chromosome 1q42 that directly disrupts brain-expressed genes: clinical and P300 findings in a family. Am. J. Hum. Genet. 69:428, 2001. 31. Blanchoin, L., Amann, K. J., Higgs, H. N., et al.: Direct observation of dendritic actin filament networks nucleated by Arp2/3 complex and WASP/Scar proteins. Nature 404:1007, 2000. 32. Bokoch, G. M., Wang, Y., Bohl, B. P., et al.: Interaction of the Nck adapter protein with p21-activated kinase (PAK1). J. Biol. Chem. 271:25746, 1996. 33. Borrell, V., Del Rio, J. A., Alcantara, S., et al.: Reelin regulates the development and synaptogenesis of the layer-specific entorhino-hippocampal connections. J. Neurosci. 19:1345, 1999. 34. Bovolenta, P., and Fernaud-Espinosa, I.: Nervous system proteoglycans as modulators of neurite outgrowth. Prog. Neurobiol. 61:113, 2000. 35. Bradke, F., and Dotti, C. G.: The role of local actin instability in axon formation. Science 283:1931, 1999. 36. Bradke, F., and Dotti, C. G.: Differentiated neurons retain the capacity to generate axons from dendrites. Curr. Biol. 10:1467, 2000.

Guidance of Axons to Targets in Development and in Disease 37. Brittis, P. A., Lu, Q., and Flanagan, J. G.: Axonal protein synthesis provides a mechanism for localized regulation at an intermediate target. Cell 110:223, 2002. 38. Brittis, P. A., Silver, J., Walsh, F. S., and Doherty, P.: Fibroblast growth factor receptor function is required for the orderly projection of ganglion cell axons in the developing mammalian retina. Mol. Cell. Neurosci. 8:120, 1996. 39. Brose, K., Bland, K. S., Wang, K. H., et al.: Slit proteins bind Robo receptors and have an evolutionarily conserved role in repulsive axon guidance. Cell 96:795, 1999. 40. Brown, D. A., and Rose, J. K.: Sorting of GPI-anchored proteins to glycolipid-enriched membrane subdomains during transport to the apical cell surface. Cell 68:533, 1992. 41. Brown, J. D., and Moon, R. T.: Wnt signaling: why is everything so negative? Curr. Opin. Cell Biol. 10:182, 1998. 42. Brown, M. D., Cornejo, B. J., Kuhn, T. B., and Bamburg, J. R.: Cdc42 stimulates neurite outgrowth and formation of growth cone filopodia and lamellipodia. J. Neurobiol. 43:352, 2000. 43. Bruckner, K., Pablo Labrador, J., Scheiffele, P., et al.: EphrinB ligands recruit GRIP family PDZ adaptor proteins into raft membrane microdomains. Neuron 22:511, 1999. 44. Bruckner, K., Pasquale, E. B., and Klein, R.: Tyrosine phosphorylation of transmembrane ligands for Eph receptors. Science 275:1640, 1997. 45. Brummendorf, T., and Rathjen, F. G.: Structure/function relationships of axon-associated adhesion receptors of the immunoglobulin superfamily. Curr. Opin. Neurobiol. 6:584, 1996. 46. Bulow, H. E., Berry, K. L., Topper, L. H., et al.: Heparan sulfate proteoglycan-dependent induction of axon branching and axon misrouting by the Kallmann syndrome gene kal-1. Proc. Natl. Acad. Sci. U. S. A. 99:6346, 2002. 47. Burden-Gulley, S. M., Pendergast, M., and Lemmon, V.: The role of cell adhesion molecule L1 in axonal extension, growth cone motility, and signal transduction. Cell Tissue Res. 290:415, 1997. 48. Campbell, D. S., and Holt, C. E.: Chemotropic responses of retinal growth cones mediated by rapid local protein synthesis and degradation. Neuron 32:1013, 2001. 49. Campbell, D. S., and Holt, C. E.: Apoptotic pathway and MAPKs differentially regulate chemotropic responses of retinal growth cones. Neuron 37:939, 2003. 50. Cannon, T. D., Kaprio, J., Lonnqvist, J., et al.: The genetic epidemiology of schizophrenia in a Finnish twin cohort: a population-based modeling study. Arch. Gen. Psychiatry 55:67, 1998. 51. Carlsson, L., Nystrom, L. E., Sundkvist, I., et al.: Actin polymerizability is influenced by profilin, a low molecular weight protein in non-muscle cells. J. Mol. Biol. 115:465, 1977. 52. Carter, N., Nakamoto, T., Hirai, H., and Hunter, T.: EphrinA1-induced cytoskeletal re-organization requires FAK and p130(cas). Nat. Cell Biol. 4:565, 2002. 53. Cassimeris, L., Safer, D., Nachmias, V. T., and Zigmond, S. H.: Thymosin beta 4 sequesters the majority of G-actin in resting human polymorphonuclear leukocytes. J. Cell Biol. 119:1261, 1992. 54. Castellani, V., Chedotal, A., Schachner, M., et al.: Analysis of the L1-deficient mouse phenotype reveals cross-talk

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

471

between Sema3A and L1 signaling pathways in axonal guidance. Neuron 27:237, 2000. Castellani, V., De Angelis, E., Kenwrick, S., and Rougon, G.: Cis and trans interactions of L1 with neuropilin-1 control axonal responses to semaphorin 3A. EMBO J. 21:6348, 2002. Chan, S. S., Zheng, H., Su, M. W., et al.: UNC-40, a C.elegans homolog of DCC (Deleted in Colorectal Cancer), is required in motile cells responding to UNC-6 netrin cues. Cell 87:187, 1996. Chedotal, A., Del Rio, J. A., Ruiz, M., et al.: Semaphorins III and IV repel hippocampal axons via two distinct receptors. Development 125:4313, 1998. Chen, J. H., Wen, L., Dupuis, S., et al.: The N-terminal leucine-rich regions in Slit are sufficient to repel olfactory bulb axons and subventricular zone neurons. J. Neurosci. 21:1548, 2001. Cheng, Q., Sasaki, Y., Shoji, M., et al.: Cdk5/p35 and Rhokinase mediate ephrin-A5-induced signaling in retinal ganglion cells. Mol. Cell. Neurosci. 24:632, 2003. Cherfils, J., and Chardin, P.: GEFs: structural basis for their activation of small GTP-binding proteins. Trends Biochem. Sci. 24:306, 1999. Chien, C. B., Rosenthal, D. E., Harris, W. A., and Holt, C. E.: Navigational errors made by growth cones without filopodia in the embryonic Xenopus brain. Neuron 11:237, 1993. Chikumi, H., Fukuhara, S., and Gutkind, J. S.: Regulation of G protein-linked guanine nucleotide exchange factors for Rho, PDZ-RhoGEF, and LARG by tyrosine phosphorylation: evidence of a role for focal adhesion kinase. J. Biol. Chem. 277:12463, 2002. Chong, L. D., Park, E. K., Latimer, E., et al.: Fibroblast growth factor receptor-mediated rescue of x-ephrin B1induced cell dissociation in Xenopus embryos. Mol. Cell. Biol. 20:724, 2000. Clipstone, N. A., and Crabtree, G. R.: Identification of calcineurin as a key signalling enzyme in T-lymphocyte activation. Nature 357:695, 1992. Colamarino, S. A., and Tessier-Lavigne, M.: The axonal chemoattractant netrin-1 is also a chemorepellent for trochlear motor axons. Cell 81:621, 1995. Colavita, A., and Culotti, J. G.: Suppressors of ectopic UNC-5 growth cone steering identify eight genes involved in axon guidance in Caenorhabditis elegans. Dev. Biol. 194:72, 1998. Comeau, M. R., Johnson, R., DuBose, R. F., et al.: A poxvirus-encoded semaphorin induces cytokine production from monocytes and binds to a novel cellular semaphorin receptor, VESPR. Immunity 8:473, 1998. Conrad, A. J., and Scheibel, A. B.: Schizophrenia and the hippocampus: the embryological hypothesis extended. Schizophr. Bull. 13:577, 1987. Corset, V., Nguyen-Ba-Charvet, K. T., Forcet, C., et al.: Netrin-1-mediated axon outgrowth and cAMP production requires interaction with adenosine A2b receptor. Nature 407:747, 2000. Courtney, K. D., Grove, M., Vandongen, H., et al.: Localization and phosphorylation of Abl-interactor proteins, Abi-1 and Abi-2, in the developing nervous system. Mol. Cell. Neurosci. 16:244, 2000.

472

Neurobiology of the Peripheral Nervous System

71. Coutinho, S., Jahn, T., Lewitzky, M., et al.: Characterization of Grb4, an adapter protein interacting with Bcr-Abl. Blood 96:618, 2000. 72. Cowan, C. A., and Henkemeyer, M.: The SH2/SH3 adaptor Grb4 transduces B-ephrin reverse signals. Nature 413:174, 2001. 73. da Costa, S. R., Okamoto, C. T., and Hamm-Alvarez, S. F.: Actin microfilaments et al.—the many components, effectors and regulators of epithelial cell endocytosis. Adv. Drug Deliv. Rev. 55:1359, 2003. 74. Da Silva, J. S., Medina, M., Zuliani, C., et al.: RhoA/ROCK regulation of neuritogenesis via profilin IIa-mediated control of actin stability. J. Cell Biol. 162:1267, 2003. 75. Dahme, M., Bartsch, U., Martini, R., et al.: Disruption of the mouse L1 gene leads to malformations of the nervous system. Nat. Genet. 17:346, 1997. 76. Davenport, R. W., Dou, P., Rehder, V., and Kater, S. B.: A sensory role for neuronal growth cone filopodia. Nature 361:721, 1993. 77. Davis, J. Q., and Bennett, V.: Ankyrin binding activity shared by the neurofascin/L1/NrCAM family of nervous system cell adhesion molecules. J. Biol. Chem. 269:27163, 1994. 78. Davis, S., Gale, N. W., Aldrich, T. H., et al.: Ligands for EPH-related receptor tyrosine kinases that require membrane attachment or clustering for activity. Science 266:816, 1994. 79. Davy, A., Gale, N. W., Murray, E. W., et al.: Compartmentalized signaling by GPI-anchored ephrin-A5 requires the Fyn tyrosine kinase to regulate cellular adhesion. Genes Dev. 13:3125, 1999. 80. Davy, A., and Robbins, S. M.: Ephrin-A5 modulates cell adhesion and morphology in an integrin-dependent manner. EMBO J. 19:5396, 2000. 81. de Castro, F., Hu, L., Drabkin, H., et al.: Chemoattraction and chemorepulsion of olfactory bulb axons by different secreted semaphorins. J. Neurosci. 19:4428, 1999. 82. de la Torre, J. R., Hopker, V. H., Ming, G. L., et al.: Turning of retinal growth cones in a netrin-1 gradient mediated by the netrin receptor DCC. Neuron 19:1211, 1997. 83. de Wit, J., and Verhaagen, J.: Role of semaphorins in the adult nervous system. Prog. Neurobiol. 71:249, 2003. 84. Deardorff, M. A., Tan, C., Saint-Jeannet, J. P., and Klein, P. S.: A role for frizzled 3 in neural crest development. Development 128:3655, 2001. 85. Dickson, B. J.: Rho GTPases in growth cone guidance. Curr. Opin. Neurobiol. 11:103, 2001. 86. Dode, C., Levilliers, J., Dupont, J. M., et al.: Loss-of-function mutations in FGFR1 cause autosomal dominant Kallmann syndrome. Nat. Genet. 33:463, 2003. 87. Doherty, P., Williams, E., and Walsh, F. S.: A soluble chimeric form of the L1 glycoprotein stimulates neurite outgrowth. Neuron 14:57, 1995. 88. Drescher, U., Bonhoeffer, F., and Muller, B. K.: The Eph family in retinal axon guidance. Curr. Opin. Neurobiol. 7:75, 1997. 89. Drewes, G., Ebneth, A., and Mandelkow, E. M.: MAPs, MARKs and microtubule dynamics. Trends Biochem. Sci. 23:307, 1998. 90. Driessens, M. H., Hu, H., Nobes, C. D., et al.: Plexin-B semaphorin receptors interact directly with active Rac and

91.

92.

93.

94.

95.

96.

97.

98. 99.

100.

101.

102.

103.

104.

105.

106.

107.

regulate the actin cytoskeleton by activating Rho. Curr. Biol. 11:339, 2001. Eastwood, S. L., Law, A. J., Everall, I. P., and Harrison, P. J.: The axonal chemorepellant semaphorin 3A is increased in the cerebellum in schizophrenia and may contribute to its synaptic pathology. Mol. Psychiatry 8:148, 2003. Edwards, D. C., Sanders, L. C., Bokoch, G. M., and Gill, G. N.: Activation of LIM-kinase by Pak1 couples Rac/Cdc42 GTPase signalling to actin cytoskeletal dynamics. Nat. Cell Biol. 1:253, 1999. Eickholt, B. J., Walsh, F. S., and Doherty, P.: An inactive pool of GSK-3 at the leading edge of growth cones is implicated in Semaphorin 3A signaling. J. Cell Biol. 157:211, 2002. Elkins, T., Hortsch, M., Bieber, A. J., et al.: Drosophila fasciclin I is a novel homophilic adhesion molecule that along with fasciclin III can mediate cell sorting. J. Cell Biol. 110:1825, 1990. Elowe, S., Holland, S. J., Kulkarni, S., and Pawson, T.: Downregulation of the Ras-mitogen-activated protein kinase pathway by the EphB2 receptor tyrosine kinase is required for ephrin-induced neurite retraction. Mol. Cell. Biol. 21:7429, 2001. Endo, M., Ohashi, K., Sasaki, Y., et al.: Control of growth cone motility and morphology by LIM kinase and Slingshot via phosphorylation and dephosphorylation of cofilin. J. Neurosci. 23:2527, 2003. Engelkamp, D.: Cloning of three mouse unc5 genes and their expression patterns at mid-gestation. Mech. Dev. 118:191, 2002. Etienne-Manneville, S., and Hall, A.: Rho GTPases in cell biology. Nature 420:629, 2002. Evangelista, M., Pruyne, D., Amberg, D. C., et al.: Formins direct Arp2/3-independent actin filament assembly to polarize cell growth in yeast. Nat. Cell Biol. 4:260, 2002. Fan, J., Mansfield, S. G., Redmond, T., et al.: The organization of F-actin and microtubules in growth cones exposed to a brain-derived collapsing factor. J. Cell Biol. 121:867, 1993. Fan, J., and Raper, J. A.: Localized collapsing cues can steer growth cones without inducing their full collapse. Neuron 14:263, 1995. Fan, X., Labrador, J. P., Hing, H., and Bashaw, G. J.: Slit stimulation recruits Dock and Pak to the roundabout receptor and increases Rac activity to regulate axon repulsion at the CNS midline. Neuron 40:113, 2003. Fannon, A. M., and Colman, D. R.: A model for central synaptic junctional complex formation based on the differential adhesive specificities of the cadherins. Neuron 17:423, 1996. Fazeli, A., Dickinson, S. L., Hermiston, M. L., et al.: Phenotype of mice lacking functional Deleted in colorectal cancer (Dcc) gene. Nature 386:796, 1997. Flanagan, J. G., and Vanderhaeghen, P.: The ephrins and Eph receptors in neural development. Annu. Rev. Neurosci. 21:309, 1998. Flanagan, W. M., Corthesy, B., Bram, R. J., and Crabtree, G. R.: Nuclear association of a T-cell transcription factor blocked by FK-506 and cyclosporin A. Nature 352:803, 1991. Forcet, C., Stein, E., Pays, L., et al.: Netrin-1-mediated axon outgrowth requires deleted in colorectal cancer-dependent MAPK activation. Nature 417:443, 2002.

Guidance of Axons to Targets in Development and in Disease 108. Forscher, P., and Smith, S. J.: Actions of cytochalasins on the organization of actin filaments and microtubules in a neuronal growth cone. J. Cell Biol. 107:1505, 1988. 109. Franco, B., Guioli, S., Pragliola, A., et al.: A gene deleted in Kallmann’s syndrome shares homology with neural cell adhesion and axonal path-finding molecules. Nature 353:529, 1991. 110. Fransen, E., Lemmon, V., Van Camp, G., et al.: CRASH syndrome: clinical spectrum of corpus callosum hypoplasia, retardation, adducted thumbs, spastic paraparesis and hydrocephalus due to mutations in one single gene, L1. Eur. J. Hum. Genet. 3:273, 1995. 111. Freeman, N. L., Chen, Z., Horenstein, J., et al.: An actin monomer binding activity localizes to the carboxyl-terminal half of the Saccharomyces cerevisiae cyclase-associated protein. J. Biol. Chem. 270:5680, 1995. 112. Fricke, C., Lee, J. S., Geiger-Rudolph, S., et al.: Astray, a zebrafish roundabout homolog required for retinal axon guidance. Science 292:507, 2001. 113. Fritsche, J., Reber, B. F., Schindelholz, B., and Bandtlow, C. E.: Differential cytoskeletal changes during growth cone collapse in response to hSema III and thrombin. Mol. Cell. Neurosci. 14:398, 1999. 114. Fujisawa, H., and Kitsukawa, T.: Receptors for collapsin/semaphorins. Curr. Opin. Neurobiol. 8:587, 1998. 115. Fujiwara, T., Mammoto, A., Kim, Y., and Takai, Y.: Rho small G-protein-dependent binding of mDia to an Src homology 3 domain-containing IRSp53/BAIAP2. Biochem. Biophys. Res. Commun. 271:626, 2000. 116. Fukata, Y., Itoh, T. J., Kimura, T., et al.: CRMP-2 binds to tubulin heterodimers to promote microtubule assembly. Nat. Cell Biol. 4:583, 2002. 117. Galisteo, M. L., Chernoff, J., Su, Y. C., et al.: The adaptor protein Nck links receptor tyrosine kinases with the serinethreonine kinase Pak1. J. Biol. Chem. 271:20997, 1996. 118. Galko, M. J., and Tessier-Lavigne, M.: Function of an axonal chemoattractant modulated by metalloprotease activity. Science 289:1365, 2000. 119. Garrity, P. A., Rao, Y., Salecker, I., et al.: Drosophila photoreceptor axon guidance and targeting requires the dreadlocks SH2/SH3 adapter protein. Cell 85:639, 1996. 120. Garver, T. D., Ren, Q., Tuvia, S., and Bennett, V.: Tyrosine phosphorylation at a site highly conserved in the L1 family of cell adhesion molecules abolishes ankyrin binding and increases lateral mobility of neurofascin. J. Cell Biol. 137:703, 1997. 121. Gasman, S., Kalaidzidis, Y., and Zerial, M.: RhoD regulates endosome dynamics through Diaphanous-related Formin and Src tyrosine kinase. Nat. Cell Biol. 5:195, 2003. 122. Gavin, A. C., Bosche, M., Krause, R., et al.: Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415:141, 2002. 123. Gertler, F. B., Comer, A. R., Juang, J. L., et al.: Enabled, a dosage-sensitive suppressor of mutations in the Drosophila Abl tyrosine kinase, encodes an Abl substrate with SH3 domain-binding properties. Genes Dev. 9:521, 1995. 124. Gertler, F. B., Doctor, J. S., and Hoffmann, F. M.: Genetic suppression of mutations in the Drosophila abl protooncogene homolog. Science 248:857, 1990.

473

125. Gertler, F. B., Niebuhr, K., Reinhard, M., et al.: Mena, a relative of VASP and Drosophila Enabled, is implicated in the control of microfilament dynamics. Cell 87:227, 1996. 126. Giger, R. J., Pasterkamp, R. J., Heijnen, S., et al.: Anatomical distribution of the chemorepellent semaphorin III/collapsin-1 in the adult rat and human brain: predominant expression in structures of the olfactory-hippocampal pathway and the motor system. J. Neurosci. Res 52:27, 1998. 127. Gingras, A. C., Raught, B., and Sonenberg, N.: eIF4 initiation factors: effectors of mRNA recruitment to ribosomes and regulators of translation. Annu. Rev. Biochem. 68:913, 1999. 128. Giordano, S., Corso, S., Conrotto, P., et al.: The semaphorin 4D receptor controls invasive growth by coupling with Met. Nat. Cell Biol. 4:720, 2002. 129. Gitai, Z., Yu, T. W., Lundquist, E. A., et al.: The netrin receptor UNC-40/DCC stimulates axon attraction and outgrowth through enabled and, in parallel, Rac and UNC115/AbLIM. Neuron 37:53, 2003. 130. Gong, Q., Rangarajan, R., Seeger, M., and Gaul, U.: The netrin receptor frazzled is required in the target for establishment of retinal projections in the Drosophila visual system. Development 126:1451, 1999. 131. Gong, Q., and Shipley, M. T.: Evidence that pioneer olfactory axons regulate telencephalon cell cycle kinetics to induce the formation of the olfactory bulb. Neuron 14:91, 1995. 132. Goshima, Y., Nakamura, F., Strittmatter, P., and Strittmatter, S. M.: Collapsin-induced growth cone collapse mediated by an intracellular protein related to UNC-33. Nature 376:509, 1995. 133. Graef, I. A., Wang, F., Charron, F., et al.: Neurotrophins and netrins require calcineurin/NFAT signaling to stimulate outgrowth of embryonic axons. Cell 113:657, 2003. 134. Grant, S. G., Karl, K. A., Kiebler, M. A., and Kandel, E. R.: Focal adhesion kinase in the brain: novel subcellular localization and specific regulation by Fyn tyrosine kinase in mutant mice. Genes Dev. 9:1909, 1995. 135. Grootjans, J. J., Zimmermann, P., Reekmans, G., et al.: Syntenin, a PDZ protein that binds syndecan cytoplasmic domains. Proc. Natl. Acad. Sci. U. S. A. 94:13683, 1997. 136. Grove, E. A., Tole, S., Limon, J., et al.: The hem of the embryonic cerebral cortex is defined by the expression of multiple Wnt genes and is compromised in Gli3-deficient mice. Development 125:2315, 1998. 137. Guidotti, A., Auta, J., Davis, J. M., et al.: Decrease in reelin and glutamic acid decarboxylase67 (GAD67) expression in schizophrenia and bipolar disorder: a postmortem brain study. Arch. Gen. Psychiatry 57:1061, 2000. 138. Gutwein, P., Oleszewski, M., Mechtersheimer, S., et al.: Role of Src kinases in the ADAM-mediated release of L1 adhesion molecule from human tumor cells. J. Biol. Chem. 275:15490, 2000. 139. Haarer, B. K., Petzold, A. S., and Brown, S. S.: Mutational analysis of yeast profilin. Mol. Cell. Biol. 13:7864, 1993. 140. Haffner, C., Jarchau, T., Reinhard, M., et al.: Molecular cloning, structural analysis and functional expression of the proline-rich focal adhesion and microfilament-associated protein VASP. EMBO J. 14:19, 1995.

474

Neurobiology of the Peripheral Nervous System

141. Halbrugge, M., and Walter, U.: Purification of a vasodilatorregulated phosphoprotein from human platelets. Eur. J. Biochem. 185:41, 1989. 142. Hall, C., Brown, M., Jacobs, T., et al.: Collapsin response mediator protein switches RhoA and Rac1 morphology in N1E-115 neuroblastoma cells and is regulated by Rho kinase. J. Biol. Chem. 276:43482, 2001. 143. Hamada, F., Tomoyasu, Y., Takatsu, Y., et al.: Negative regulation of Wingless signaling by D-axin, a Drosophila homolog of axin. Science 283:1739, 1999. 144. Hamelin, M., Zhou, Y., Su, M. W., et al.: Expression of the UNC-5 guidance receptor in the touch neurons of C. elegans steers their axons dorsally. Nature 364:327, 1993. 145. Hanger, D. P., Hughes, K., Woodgett, J. R., et al.: Glycogen synthase kinase-3 induces Alzheimer’s disease-like phosphorylation of tau: generation of paired helical filament epitopes and neuronal localisation of the kinase. Neurosci. Lett. 147:58, 1992. 146. Harrison, P. J.: The neuropathology of schizophrenia: a critical review of the data and their interpretation. Brain 122(Pt. 4):593, 1999. 147. Hatten, M. E.: New directions in neuronal migration. Science 297:1660, 2002. 148. Hattori, M., Osterfield, M., and Flanagan, J. G.: Regulated cleavage of a contact-mediated axon repellent. Science 289:1360, 2000. 149. He, Z., and Tessier-Lavigne, M.: Neuropilin is a receptor for the axonal chemorepellent Semaphorin III. Cell 90:739, 1997. 150. Hebert, J. M., Lin, M., Partanen, J., et al.: FGF signaling through FGFR1 is required for olfactory bulb morphogenesis. Development 130:1101, 2003. 151. Hedgecock, E. M., Culotti, J. G., and Hall, D. H.: The unc-5, unc-6, and unc-40 genes guide circumferential migrations of pioneer axons and mesodermal cells on the epidermis in C. elegans. Neuron 4:61, 1990. 152. Hedgecock, E. M., Culotti, J. G., Hall, D. H., and Stern, B. D.: Genetics of cell and axon migrations in Caenorhabditis elegans. Development 100:365, 1987. 153. Hendriks, W., Schepens, J., Bachner, D., et al.: Molecular cloning of a mouse epithelial protein-tyrosine phosphatase with similarities to submembranous proteins. J. Cell. Biochem. 59:418, 1995. 154. Henkemeyer, M., Orioli, D., Henderson, J. T., et al.: Nuk controls pathfinding of commissural axons in the mammalian central nervous system. Cell 86:35, 1996. 155. Himanen, J. P., Rajashankar, K. R., Lackmann, M., et al.: Crystal structure of an Eph receptor-ephrin complex. Nature 414:933, 2001. 156. Hiramoto, M., Hiromi, Y., Giniger, E., and Hotta, Y.: The Drosophila Netrin receptor Frazzled guides axons by controlling Netrin distribution. Nature 406:886, 2000. 157. Hirotani, M., Ohoka, Y., Yamamoto, T., et al.: Interaction of plexin-B1 with PDZ domain-containing Rho guanine nucleotide exchange factors. Biochem. Biophys. Res. Commun. 297:32, 2002. 158. Hobert, O., Jallal, B., Schlessinger, J., and Ullrich, A.: Novel signaling pathway suggested by SH3 domain-mediated p95vav/heterogeneous ribonucleoprotein K interaction. J. Biol. Chem. 269:20225, 1994.

159. Holland, S. J., Gale, N. W., Mbamalu, G., et al.: Bidirectional signalling through the EPH-family receptor Nuk and its transmembrane ligands. Nature 383:722, 1996. 160. Holtmaat, A. J., Gorter, J. A., De Wit, J., et al.: Transient downregulation of Sema3A mRNA in a rat model for temporal lobe epilepsy: a novel molecular event potentially contributing to mossy fiber sprouting. Exp. Neurol. 182:142, 2003. 161. Honda, H., Oda, H., Nakamoto, T., et al.: Cardiovascular anomaly, impaired actin bundling and resistance to Srcinduced transformation in mice lacking p130Cas. Nat. Genet. 19:361, 1998. 162. Hong, K., Hinck, L., Nishiyama, M., et al.: A ligand-gated association between cytoplasmic domains of UNC5 and DCC family receptors converts netrin-induced growth cone attraction to repulsion. Cell 97:927, 1999. 163. Hong, K., Nishiyama, M., Henley, J., et al.: Calcium signalling in the guidance of nerve growth by netrin-1. Nature 403:93, 2000. 164. Hong, S. E., Shugart, Y. Y., Huang, D. T., et al.: Autosomal recessive lissencephaly with cerebellar hypoplasia is associated with human RELN mutations. Nat. Genet. 26:93, 2000. 165. Hornstein, I., Alcover, A., and Katzav, S.: Vav proteins, masters of the world of cytoskeleton organization. Cell. Signal. 16:1, 2004. 166. Hu, H.: Cell-surface heparan sulfate is involved in the repulsive guidance activities of Slit2 protein. Nat. Neurosci. 4:695, 2001. 167. Hu, H., Marton, T. F., and Goodman, C. S.: Plexin B mediates axon guidance in Drosophila by simultaneously inhibiting active Rac and enhancing RhoA signaling. Neuron 32:39, 2001. 168. Huai, J., and Drescher, U.: An ephrin-A-dependent signaling pathway controls integrin function and is linked to the tyrosine phosphorylation of a 120-kDa protein. J. Biol. Chem. 276:6689, 2001. 169. Huang, X., Cheng, H. J., Tessier-Lavigne, M., and Jin, Y.: MAX-1, a novel PH/MyTH4/FERM domain cytoplasmic protein implicated in netrin-mediated axon repulsion. Neuron 34:563, 2002. 170. Hug, C., Jay, P. Y., Reddy, I., et al.: Capping protein levels influence actin assembly and cell motility in Dictyostelium. Cell 81:591, 1995. 171. Ignelzi, M. A. Jr., Miller, D. R., Soriano, P., and Maness, P. F.: Impaired neurite outgrowth of src-minus cerebellar neurons on the cell adhesion molecule L1. Neuron 12:873, 1994. 172. Ishii, N., Wadsworth, W. G., Stern, B. D., et al.: UNC-6, a laminin-related protein, guides cell and pioneer axon migrations in C. elegans. Neuron 9:873, 1992. 173. Ishikawa, Y., Katoh, H., and Negishi, M.: A role of Rnd1 GTPase in dendritic spine formation in hippocampal neurons. J. Neurosci. 23:11065, 2003. 174. Ishizaki, T., Maekawa, M., Fujisawa, K., et al.: The small GTP-binding protein Rho binds to and activates a 160 kDa Ser/Thr protein kinase homologous to myotonic dystrophy kinase. EMBO J. 15:1885, 1996. 175. Jalink, K., Eichholtz, T., Postma, F. R., et al.: Lysophosphatidic acid induces neuronal shape changes via a novel, receptor-mediated signaling pathway: similarity to thrombin action. Cell Growth Differ. 4:247, 1993.

Guidance of Axons to Targets in Development and in Disease 176. Jalink, K., van Corven, E. J., Hengeveld, T., et al.: Inhibition of lysophosphatidate- and thrombin-induced neurite retraction and neuronal cell rounding by ADP ribosylation of the small GTP-binding protein Rho. J. Cell Biol. 126:801, 1994. 177. Jin, Z., and Strittmatter, S. M.: Rac1 mediates collapsin-1induced growth cone collapse. J. Neurosci. 17:6256, 1997. 178. Kabir, N., Schaefer, A. W., Nakhost, A., et al.: Protein kinase C activation promotes microtubule advance in neuronal growth cones by increasing average microtubule growth lifetimes. J. Cell Biol. 152:1033, 2001. 179. Kallmann, F. J., Schoenfeld, W. A., and Barrera, S. E.: The genetic aspects of primary eunuchoidism. Am. J. Ment. Defic. XLVIII:203, 1944. 180. Kalo, M. S., and Pasquale, E. B.: Multiple in vivo tyrosine phosphorylation sites in EphB receptors. Biochemistry 38:14396, 1999. 181. Kamiguchi, H., Hlavin, M. L., Yamasaki, M., and Lemmon, V.: Adhesion molecules and inherited diseases of the human nervous system. Annu. Rev. Neurosci. 21:97, 1998. 182. Kamm, K. E., and Stull, J. T.: Myosin phosphorylation, force, and maximal shortening velocity in neurally stimulated tracheal smooth muscle. Am. J. Physiol. 249:C238, 1985. 183. Kang, F., Purich, D. L., and Southwick, F. S.: Profilin promotes barbed-end actin filament assembly without lowering the critical concentration. J. Biol. Chem. 274:36963, 1999. 184. Katsu, T., Ujike, H., Nakano, T., et al.: The human frizzled-3 (FZD3) gene on chromosome 8p21, a receptor gene for Wnt ligands, is associated with the susceptibility to schizophrenia. Neurosci. Lett. 353:53, 2003. 185. Kaufmann, N., Wills, Z. P., and Van Vactor, D.: Drosophila Rac1 controls motor axon guidance. Development 125:453, 1998. 186. Keino-Masu, K., Masu, M., Hinck, L., et al.: Deleted in Colorectal Cancer (DCC) encodes a netrin receptor. Cell 87:175, 1996. 187. Keleman, K., Rajagopalan, S., Cleppien, D., et al.: Comm sorts robo to control axon guidance at the Drosophila midline. Cell 110:415, 2002. 188. Kennedy, T. E.: Cellular mechanisms of netrin function: long-range and short-range actions. Biochem. Cell Biol. 78:569, 2000. 189. Kennedy, T. E., Serafini, T., de la Torre, J. R., and TessierLavigne, M.: Netrins are diffusible chemotropic factors for commissural axons in the embryonic spinal cord. Cell 78:425, 1994. 190. Kidd, T., Bland, K. S., and Goodman, C. S.: Slit is the midline repellent for the robo receptor in Drosophila. Cell 96:785, 1999. 191. Kimura, K., Ito, M., Amano, M., et al.: Regulation of myosin phosphatase by Rho and Rho-associated kinase (Rho-kinase). Science 273:245, 1996. 192. Klee, C. B., Crouch, T. H., and Krinks, M. H.: Calcineurin: a calcium- and calmodulin-binding protein of the nervous system. Proc. Natl. Acad. Sci. U. S. A. 76:6270, 1979. 193. Klostermann, A., Lohrum, M., Adams, R. H., and Puschel, A. W.: The chemorepulsive activity of the axonal guidance signal semaphorin D requires dimerization. J. Biol. Chem. 273:7326, 1998. 194. Knoll, B., and Drescher, U.: Ephrin-As as receptors in topographic projections. Trends Neurosci. 25:145, 2002.

475

195. Knoll, B., Zarbalis, K., Wurst, W., and Drescher, U.: A role for the EphA family in the topographic targeting of vomeronasal axons. Development 128:895, 2001. 196. Kobielak, A., Pasolli, H. A., and Fuchs, E.: Mammalian formin-1 participates in adherens junctions and polymerization of linear actin cables. Nat. Cell Biol. 6:21, 2004. 197. Koeberle, P. D., and Bahr, M.: Growth and guidance cues for regenerating axons: where have they gone? J. Neurobiol. 59:162, 2004. 198. Kolodkin, A. L., Levengood, D. V., Rowe, E. G., et al.: Neuropilin is a semaphorin III receptor. Cell 90:753, 1997. 199. Kolodkin, A. L., Matthes, D. J., O’Connor, T. P., et al.: Fasciclin IV: sequence, expression, and function during growth cone guidance in the grasshopper embryo. Neuron 9:831, 1992. 200. Kolodziej, P. A., Timpe, L. C., Mitchell, K. J., et al.: Frazzled encodes a Drosophila member of the DCC immunoglobulin subfamily and is required for CNS and motor axon guidance. Cell 87:197, 1996. 201. Kozasa, T., Jiang, X., Hart, M. J., et al.: p115 RhoGEF, a GTPase activating protein for Galpha12 and Galpha13. Science 280:2109, 1998. 202. Kozma, R., Ahmed, S., Best, A., and Lim, L.: The Ras-related protein Cdc42Hs and bradykinin promote formation of peripheral actin microspikes and filopodia in Swiss 3T3 fibroblasts. Mol. Cell. Biol. 15:1942, 1995. 203. Krause, M., Dent, E. W., Bear, J. E., et al.: Ena/VASP proteins: regulators of the actin cytoskeleton and cell migration. Annu. Rev. Cell Dev. Biol. 19:541, 2003. 204. Krebs, A., Rothkegel, M., Klar, M., and Jockusch, B. M.: Characterization of functional domains of mDia1, a link between the small GTPase Rho and the actin cytoskeleton. J. Cell Sci. 114:3663, 2001. 205. Krugmann, S., Jordens, I., Gevaert, K., et al.: Cdc42 induces filopodia by promoting the formation of an IRSp53:Mena complex. Curr. Biol. 11:1645, 2001. 206. Kuhn, T. B., Meberg, P. J., Brown, M. D., et al.: Regulating actin dynamics in neuronal growth cones by ADF/cofilin and rho family GTPases. J. Neurobiol. 44:126, 2000. 207. Kuhn, T. B., Stoeckli, E. T., Condrau, M. A., et al.: Neurite outgrowth on immobilized axonin-1 is mediated by a heterophilic interaction with L1(G4). J. Cell Biol. 115:1113, 1991. 208. Kullander, K., Butt, S. J., Lebret, J. M., et al.: Role of EphA4 and EphrinB3 in local neuronal circuits that control walking. Science 299:1889, 2003. 209. Kullander, K., Croll, S. D., Zimmer, M., et al.: Ephrin-B3 is the midline barrier that prevents corticospinal tract axons from recrossing, allowing for unilateral motor control. Genes Dev. 15:877, 2001. 210. Kullander, K., and Klein, R.: Mechanisms and functions of Eph and ephrin signalling. Nat. Rev. Mol. Cell. Biol. 3:475, 2002. 211. Kullander, K., Mather, N. K., Diella, F., et al.: Kinasedependent and kinase-independent functions of EphA4 receptors in major axon tract formation in vivo. Neuron 29:73, 2001. 212. Kumanogoh, A., Marukawa, S., Suzuki, K., et al.: Class IV semaphorin Sema4A enhances T-cell activation and interacts with Tim-2. Nature 419:629, 2002.

476

Neurobiology of the Peripheral Nervous System

213. Kumanogoh, A., Watanabe, C., Lee, I., et al.: Identification of CD72 as a lymphocyte receptor for the class IV semaphorin CD100: a novel mechanism for regulating B cell signaling. Immunity 13:621, 2000. 214. Kunz, S., Ziegler, U., Kunz, B., and Sonderegger, P.: Intracellular signaling is changed after clustering of the neural cell adhesion molecules axonin-1 and NgCAM during neurite fasciculation. J. Cell Biol. 135:253, 1996. 215. Lambert de Rouvroit, C., and Goffinet, A. M.: Neuronal migration. Mech. Dev. 105:47, 2001. 216. Lanier, L. M., Gates, M. A., Witke, W., et al.: Mena is required for neurulation and commissure formation. Neuron 22:313, 1999. 217. Legouis, R., Hardelin, J. P., Levilliers, J., et al.: The candidate gene for the X-linked Kallmann syndrome encodes a protein related to adhesion molecules. Cell 67:423, 1991. 218. Leighton, P. A., Mitchell, K. J., Goodrich, L. V., et al.: Defining brain wiring patterns and mechanisms through gene trapping in mice. Nature 410:174, 2001. 219. Lemmon, V., Farr, K. L., and Lagenaur, C.: L1-mediated axon outgrowth occurs via a homophilic binding mechanism. Neuron 2:1597, 1989. 220. Leonardo, E. D., Hinck, L., Masu, M., et al.: Vertebrate homologues of C. elegans UNC-5 are candidate netrin receptors. Nature 386:833, 1997. 221. Leung, T., Chen, X. Q., Manser, E., and Lim, L.: The p160 RhoA-binding kinase ROK alpha is a member of a kinase family and is involved in the reorganization of the cytoskeleton. Mol. Cell. Biol. 16:5313, 1996. 222. Leung-Hagesteijn, C., Spence, A. M., Stern, B. D., et al.: UNC-5, a transmembrane protein with immunoglobulin and thrombospondin type 1 domains, guides cell and pioneer axon migrations in C. elegans. Cell 71:289, 1992. 223. Li, H. S., Chen, J. H., Wu, W., et al.: Vertebrate slit, a secreted ligand for the transmembrane protein roundabout, is a repellent for olfactory bulb axons. Cell 96:807, 1999. 224. Li, X., Meriane, M., Triki, I., et al.: The adaptor protein Nck1 couples the netrin-1 receptor DCC (deleted in colorectal cancer) to the activation of the small GTPase Rac1 through an atypical mechanism. J. Biol. Chem. 277:37788, 2002. 225. Li, X., Saint-Cyr-Proulx, E., Aktories, K., and LamarcheVane, N.: Rac1 and Cdc42 but not RhoA or Rho kinase activities are required for neurite outgrowth induced by the Netrin-1 receptor DCC (deleted in colorectal cancer) in N1E-115 neuroblastoma cells. J. Biol. Chem. 277:15207, 2002. 226. Liang, Y., Annan, R. S., Carr, S. A., et al.: Mammalian homologues of the Drosophila slit protein are ligands of the heparan sulfate proteoglycan glypican-1 in brain. J. Biol. Chem. 274:17885, 1999. 227. Lin, C. H., Espreafico, E. M., Mooseker, M. S., and Forscher, P.: Myosin drives retrograde F-actin flow in neuronal growth cones. Neuron 16:769, 1996. 228. Lu, Q., Sun, E. E., Klein, R. S., and Flanagan, J. G.: Ephrin-B reverse signaling is mediated by a novel PDZRGS protein and selectively inhibits G protein-coupled chemoattraction. Cell 105:69, 2001. 229. Luby-Phelps, K., and Taylor, D. L.: Subcellular compartmentalization by local differentiation of cytoplasmic structure. Cell Motil. Cytoskeleton 10:28, 1988.

230. Luo, L.: Actin cytoskeleton regulation in neuronal morphogenesis and structural plasticity. Annu. Rev. Cell Dev. Biol. 18:601, 2002. 231. Luo, L., Hensch, T. K., Ackerman, L., et al.: Differential effects of the Rac GTPase on Purkinje cell axons and dendritic trunks and spines. Nature 379:837, 1996. 232. Luo, L., Liao, Y. J., Jan, L. Y., and Jan, Y. N.: Distinct morphogenetic functions of similar small GTPases: Drosophila Drac1 is involved in axonal outgrowth and myoblast fusion. Genes Dev. 8:1787, 1994. 233. Luo, Y., Raible, D., and Raper, J. A.: Collapsin: a protein in brain that induces the collapse and paralysis of neuronal growth cones. Cell 75:217, 1993. 234. Lyuksyutova, A. I., Lu, C. C., Milanesio, N., et al.: Anteriorposterior guidance of commissural axons by Wnt-frizzled signaling. Science 302:1984, 2003. 235. Machesky, L. M., and Insall, R. H.: Scar1 and the related Wiskott-Aldrich syndrome protein, WASP, regulate the actin cytoskeleton through the Arp2/3 complex. Curr. Biol. 8:1347, 1998. 236. Machesky, L. M., Mullins, R. D., Higgs, H. N., et al.: Scar, a WASp-related protein, activates nucleation of actin filaments by the Arp2/3 complex. Proc. Natl. Acad. Sci. U. S. A. 96:3739, 1999. 237. Maekawa, M., Ishizaki, T., Boku, S., et al.: Signaling from Rho to the actin cytoskeleton through protein kinases ROCK and LIM-kinase. Science 285:895, 1999. 238. Maestre de San Juan, A.: Falta total de los nervios olfactorios con anosmia en un individuo en quien existia una atrofia congénita de los testiculos y miembro viril. Siglo Med. 131:211, 1856. 239. Manser, E., Leung, T., Salihuddin, H., et al.: A brain serine/threonine protein kinase activated by Cdc42 and Rac1. Nature 367:40, 1994. 240. Marston, D. J., Dickinson, S., and Nobes, C. D.: Racdependent trans-endocytosis of ephrinBs regulates Ephephrin contact repulsion. Nat. Cell Biol. 5:879, 2003. 241. Mathern, G. W., Babb, T. L., Leite, J. P., et al.: The pathogenic and progressive features of chronic human hippocampal epilepsy. Epilepsy Res. 26:151, 1996. 242. McWhirter, J. R., and Wang, J. Y.: An actin-binding function contributes to transformation by the Bcr-Abl oncoprotein of Philadelphia chromosome-positive human leukemias. EMBO J. 12:1533, 1993. 243. Meberg, P. J., and Bamburg, J. R.: Increase in neurite outgrowth mediated by overexpression of actin depolymerizing factor. J. Neurosci. 20:2459, 2000. 244. Merz, D. C., and Culotti, J. G.: Genetic analysis of growth cone migrations in Caenorhabditis elegans. J. Neurobiol. 44:281, 2000. 245. Miao, H., Burnett, E., Kinch, M., et al.: Activation of EphA2 kinase suppresses integrin function and causes focal-adhesion-kinase dephosphorylation. Nat. Cell Biol. 2:62, 2000. 246. Miao, H., Wei, B. R., Peehl, D. M., et al.: Activation of EphA receptor tyrosine kinase inhibits the Ras/MAPK pathway. Nat. Cell Biol. 3:527, 2001. 247. Miki, H., Suetsugu, S., and Takenawa, T.: WAVE, a novel WASP-family protein involved in actin reorganization induced by Rac. EMBO J. 17:6932, 1998.

Guidance of Axons to Targets in Development and in Disease 248. Miki, H., and Takenawa, T.: WAVE2 serves a functional partner of IRSp53 by regulating its interaction with Rac. Biochem. Biophys. Res. Commun. 293:93, 2002. 249. Miki, H., Yamaguchi, H., Suetsugu, S., and Takenawa, T.: IRSp53 is an essential intermediate between Rac and WAVE in the regulation of membrane ruffling. Nature 408:732, 2000. 250. Mikule, K., Gatlin, J. C., de la Houssaye, B. A., and Pfenninger, K. H.: Growth cone collapse induced by semaphorin 3A requires 12/15-lipoxygenase. J. Neurosci. 22:4932, 2002. 251. Millar, J. K., Wilson-Annan, J. C., Anderson, S., et al.: Disruption of two novel genes by a translocation co-segregating with schizophrenia. Hum. Mol. Genet. 9:1415, 2000. 252. Minamide, L. S., Striegl, A. M., Boyle, J. A., et al.: Neurodegenerative stimuli induce persistent ADF/cofilinactin rods that disrupt distal neurite function. Nat. Cell Biol. 2:628, 2000. 253. Ming, G., Song, H., Berninger, B., et al.: Phospholipase C-gamma and phosphoinositide 3-kinase mediate cytoplasmic signaling in nerve growth cone guidance. Neuron 23:139, 1999. 254. Ming, G. L., Song, H. J., Berninger, B., et al.: cAMPdependent growth cone guidance by netrin-1. Neuron 19:1225, 1997. 255. Mitsui, N., Inatome, R., Takahashi, S., et al.: Involvement of Fes/Fps tyrosine kinase in semaphorin3A signaling. EMBO J. 21:3274, 2002. 256. Miyoshi, K., Honda, A., Baba, K., et al.: Disrupted-InSchizophrenia 1, a candidate gene for schizophrenia, participates in neurite outgrowth. Mol. Psychiatry 8:685, 2003. 257. Murakami, Y., Suto, F., Shimizu, M., et al.: Differential expression of plexin-A subfamily members in the mouse nervous system. Dev. Dyn. 220:246, 2001. 258. Myat, A., Henry, P., McCabe, V., et al.: Drosophila Nedd4, a ubiquitin ligase, is recruited by Commissureless to control cell surface levels of the roundabout receptor. Neuron 35:447, 2002. 259. Nadler, J. V.: The recurrent mossy fiber pathway of the epileptic brain. Neurochem. Res. 28:1649, 2003. 260. Naffah-Mazzacoratti, M. G., Arganaraz, G. A., Porcionatto, M. A., et al.: Selective alterations of glycosaminoglycans synthesis and proteoglycan expression in rat cortex and hippocampus in pilocarpine-induced epilepsy. Brain Res. Bull. 50:229, 1999. 261. Nayeem, N., Silletti, S., Yang, X., et al.: A potential role for the plasmin(ogen) system in the posttranslational cleavage of the neural cell adhesion molecule L1. J. Cell Sci. 112 (Pt 24):4739, 1999. 262. Newsome, T. P., Schmidt, S., Dietzl, G., et al.: Trio combines with dock to regulate Pak activity during photoreceptor axon pathfinding in Drosophila. Cell 101:283, 2000. 263. Nguyen Ba-Charvet, K. T., Brose, K., Ma, L., et al.: Diversity and specificity of actions of Slit2 proteolytic fragments in axon guidance. J. Neurosci. 21:4281, 2001. 264. Nguyen-Ba-Charvet, K. T., Brose, K., Marillat, V., et al.: Sensory axon response to substrate-bound Slit2 is modulated by laminin and cyclic GMP. Mol. Cell. Neurosci. 17:1048, 2001.

477

265. Nguyen-Ba-Charvet, K. T., Plump, A. S., Tessier-Lavigne, M., and Chedotal, A.: Slit1 and Slit2 proteins control the development of the lateral olfactory tract. J. Neurosci. 22:5473, 2002. 266. Niethammer, M., Smith, D. S., Ayala, R., et al.: NUDEL is a novel Cdk5 substrate that associates with LIS1 and cytoplasmic dynein. Neuron 28:697, 2000. 267. Nikolic, M., Chou, M. M., Lu, W., et al.: The p35/Cdk5 kinase is a neuron-specific Rac effector that inhibits Pak1 activity. Nature 395:194, 1998. 268. Nishida, E., Maekawa, S., and Sakai, H.: Cofilin, a protein in porcine brain that binds to actin filaments and inhibits their interactions with myosin and tropomyosin. Biochemistry 23:5307, 1984. 269. Nishiki, T., Narumiya, S., Morii, N., et al.: ADP-ribosylation of the rho/rac proteins induces growth inhibition, neurite outgrowth and acetylcholine esterase in cultured PC-12 cells. Biochem. Biophys. Res. Commun. 167:265, 1990. 270. Nishiyama, M., Hoshino, A., Tsai, L., et al.: Cyclic AMP/GMP-dependent modulation of Ca2⫹ channels sets the polarity of nerve growth-cone turning. Nature 424:990, 2003. 271. Niwa, R., Nagata-Ohashi, K., Takeichi, M., et al.: Control of actin reorganization by Slingshot, a family of phosphatases that dephosphorylate ADF/cofilin. Cell 108:233, 2002. 272. Nobes, C. D., and Hall, A.: Rho, rac, and cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. Cell 81:53, 1995. 273. Ohashi, K., Nagata, K., Maekawa, M., et al.: Rho-associated kinase ROCK activates LIM-kinase 1 by phosphorylation at threonine 508 within the activation loop. J. Biol. Chem. 275:3577, 2000. 274. Oinuma, I., Katoh, H., Harada, A., and Negishi, M.: Direct interaction of Rnd1 with Plexin-B1 regulates PDZRhoGEF-mediated Rho activation by Plexin-B1 and induces cell contraction in COS-7 cells. J. Biol. Chem. 278:25671, 2003. 275. Okazaki, M. M., Evenson, D. A., and Nadler, J. V.: Hippocampal mossy fiber sprouting and synapse formation after status epilepticus in rats: visualization after retrograde transport of biocytin. J. Comp. Neurol. 352:515, 1995. 276. Olofsson, B.: Rho guanine dissociation inhibitors: pivotal molecules in cellular signalling. Cell. Signal. 11:545, 1999. 277. O’Neill, G. M., Fashena, S. J., and Golemis, E. A.: Integrin signalling: a new Cas(t) of characters enters the stage. Trends Cell Biol. 10:111, 2000. 278. Ozeki, Y., Tomoda, T., Kleiderlein, J., et al.: Disrupted-inSchizophrenia-1 (DISC-1): mutant truncation prevents binding to NudE-like (NUDEL) and inhibits neurite outgrowth. Proc. Natl. Acad. Sci. U. S. A. 100:289, 2003. 279. Palmer, A., Zimmer, M., Erdmann, K. S., et al.: EphrinB phosphorylation and reverse signaling: regulation by Src kinases and PTP-BL phosphatase. Mol. Cell 9:725, 2002. 280. Parsons, J. T.: Focal adhesion kinase: the first ten years. J. Cell Sci. 116:1409, 2003. 281. Pasterkamp, R. J., and Kolodkin, A. L.: Semaphorin junction: making tracks toward neural connectivity. Curr. Opin. Neurobiol. 13:79, 2003.

478

Neurobiology of the Peripheral Nervous System

282. Pasterkamp, R. J., Peschon, J. J., Spriggs, M. K., and Kolodkin, A. L.: Semaphorin 7A promotes axon outgrowth through integrins and MAPKs. Nature 424:398, 2003. 283. Peck, J., Douglas, G. T., Wu, C. H., and Burbelo, P. D.: Human RhoGAP domain-containing proteins: structure, function and evolutionary relationships. FEBS Lett. 528:27, 2002. 284. Pellegrini, L.: Role of heparan sulfate in fibroblast growth factor signalling: a structural view. Curr. Opin. Struct. Biol. 11:629, 2001. 285. Perosa, S. R., Porcionatto, M. A., Cukiert, A., et al.: Glycosaminoglycan levels and proteoglycan expression are altered in the hippocampus of patients with mesial temporal lobe epilepsy. Brain Res. Bull. 58:509, 2002. 286. Perrot, V., Vazquez-Prado, J., and Gutkind, J. S.: Plexin B regulates Rho through the guanine nucleotide exchange factors leukemia-associated Rho GEF (LARG) and PDZRhoGEF. J. Biol. Chem. 277:43115, 2002. 287. Plump, A. S., Erskine, L., Sabatier, C., et al.: Slit1 and Slit2 cooperate to prevent premature midline crossing of retinal axons in the mouse visual system. Neuron 33:219, 2002. 288. Polleux, F., Morrow, T., and Ghosh, A.: Semaphorin 3A is a chemoattractant for cortical apical dendrites. Nature 404:567, 2000. 289. Poltorak, M., Khoja, I., Hemperly, J. J., et al.: Disturbances in cell recognition molecules (N-CAM and L1 antigen) in the CSF of patients with schizophrenia. Exp. Neurol. 131:266, 1995. 290. Poltorak, M., Wright, R., Hemperly, J. J., et al.: Monozygotic twins discordant for schizophrenia are discordant for N-CAM and L1 in CSF. Brain Res. 751:152, 1997. 291. Raftopoulou, M., and Hall, A.: Cell migration: Rho GTPases lead the way. Dev. Biol. 265:23, 2004. 292. Rajagopalan, S., Vivancos, V., Nicolas, E., and Dickson, B. J.: Selecting a longitudinal pathway: Robo receptors specify the lateral position of axons in the Drosophila CNS. Cell 103:1033, 2000. 293. Ralevic, V., and Burnstock, G.: Receptors for purines and pyrimidines. Pharmacol. Rev. 50:413, 1998. 294. Ramon y Cajal, S.: À quelle époque apparaissent les expansions des cellules nerveuses de la moëlle épinière du poulet? Anat. Anzeiger 21–22:609, 1890. 295. Raper, J. A.: Semaphorins and their receptors in vertebrates and invertebrates. Curr. Opin. Neurobiol. 10:88, 2000. 296. Reiner, O., Carrozzo, R., Shen, Y., et al.: Isolation of a Miller-Dieker lissencephaly gene containing G protein beta-subunit-like repeats. Nature 364:717, 1993. 297. Rhee, J., Mahfooz, N. S., Arregui, C., et al.: Activation of the repulsive receptor Roundabout inhibits N-cadherin-mediated cell adhesion. Nat. Cell Biol. 4:798, 2002. 298. Ribon, V., Herrera, R., Kay, B. K., and Saltiel, A. R.: A role for CAP, a novel, multifunctional Src homology 3 domaincontaining protein, in formation of actin stress fibers and focal adhesions. J. Biol. Chem. 273:4073, 1998. 299. Ridley, A. J., and Hall, A.: The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 70:389, 1992. 300. Ridley, A. J., Paterson, H. F., Johnston, C. L., et al.: The small GTP-binding protein rac regulates growth factor-induced membrane ruffling. Cell 70:401, 1992.

301. Riviello, J. J.: Classification of seizures and epilepsy. Curr. Neurol. Neurosci. Rep. 3:325, 2003. 302. Rohatgi, R., Ma, L., Miki, H., et al.: The interaction between N-WASP and the Arp2/3 complex links Cdc42dependent signals to actin assembly. Cell 97:221, 1999. 303. Rohm, B., Ottemeyer, A., Lohrum, M., and Puschel, A. W.: Plexin/neuropilin complexes mediate repulsion by the axonal guidance signal semaphorin 3A. Mech. Dev. 93:95, 2000. 304. Rohm, B., Rahim, B., Kleiber, B., et al.: The semaphorin 3A receptor may directly regulate the activity of small GTPases. FEBS Lett. 486:68, 2000. 305. Ronca, F., Andersen, J. S., Paech, V., and Margolis, R. U.: Characterization of Slit protein interactions with glypican-1. J. Biol. Chem. 276:29141, 2001. 306. Ross, M. E., and Walsh, C. A.: Human brain malformations and their lessons for neuronal migration. Annu. Rev. Neurosci. 24:1041, 2001. 307. Rothberg, J. M., and Artavanis-Tsakonas, S.: Modularity of the slit protein: characterization of a conserved carboxyterminal sequence in secreted proteins and a motif implicated in extracellular protein interactions. J. Mol. Biol. 227:367, 1992. 308. Rothberg, J. M., Hartley, D. A., Walther, Z., and ArtavanisTsakonas, S.: Slit: an EGF-homologous locus of D. melanogaster involved in the development of the embryonic central nervous system. Cell 55:1047, 1988. 309. Rothberg, J. M., Jacobs, J. R., Goodman, C. S., and Artavanis-Tsakonas, S.: Slit: an extracellular protein necessary for development of midline glia and commissural axon pathways contains both EGF and LRR domains. Genes Dev. 4:2169, 1990. 310. Rougon, G., and Hobert, O.: New insights into the diversity and function of neuronal immunoglobulin superfamily molecules. Annu. Rev. Neurosci. 26:207, 2003. 311. Rudolph, M. G., Bayer, P., Abo, A., et al.: The Cdc42/Rac interactive binding region motif of the Wiskott Aldrich syndrome protein (WASP) is necessary but not sufficient for tight binding to Cdc42 and structure formation. J. Biol. Chem. 273:18067, 1998. 312. Rugarli, E. I., Di Schiavi, E., Hilliard, M. A., et al.: The Kallmann syndrome gene homolog in C. elegans is involved in epidermal morphogenesis and neurite branching. Development 129:1283, 2002. 313. Safer, D., Elzinga, M., and Nachmias, V. T.: Thymosin beta 4 and Fx, an actin-sequestering peptide, are indistinguishable. J. Biol. Chem. 266:4029, 1991. 314. Saffell, J. L., Williams, E. J., Mason, I. J., et al.: Expression of a dominant negative FGF receptor inhibits axonal growth and FGF receptor phosphorylation stimulated by CAMs. Neuron 18:231, 1997. 315. Sagot, I., Klee, S. K., and Pellman, D.: Yeast formins regulate cell polarity by controlling the assembly of actin cables. Nat. Cell Biol. 4:42, 2002. 316. Sagot, I., Rodal, A. A., Moseley, J., et al.: An actin nucleation mechanism mediated by Bni1 and profilin. Nat. Cell Biol. 4:626, 2002. 317. Sahai, E., Olson, M. F., and Marshall, C. J.: Cross-talk between Ras and Rho signalling pathways in transformation favours proliferation and increased motility. EMBO J. 20:755, 2001.

Guidance of Axons to Targets in Development and in Disease 318. Salazar, M. A., Kwiatkowski, A. V., Pellegrini, L., et al.: Tuba, a novel protein containing bin/amphiphysin/Rvs and Dbl homology domains, links dynamin to regulation of the actin cytoskeleton. J. Biol. Chem. 278:49031, 2003. 319. Sander, E. E., ten Klooster, J. P., van Delft, S., et al.: Rac downregulates Rho activity: reciprocal balance between both GTPases determines cellular morphology and migratory behavior. J. Cell Biol. 147:1009, 1999. 320. Sanders, L. C., Matsumura, F., Bokoch, G. M., and de Lanerolle, P.: Inhibition of myosin light chain kinase by p21-activated kinase. Science 283:2083, 1999. 321. Sarmiere, P. D., and Bamburg, J. R.: Regulation of the neuronal actin cytoskeleton by ADF/cofilin. J. Neurobiol. 58:103, 2004. 322. Sasaki, S., Shionoya, A., Ishida, M., et al.: A LIS1/NUDEL/ cytoplasmic dynein heavy chain complex in the developing and adult nervous system. Neuron 28:681, 2000. 323. Sasaki, Y., Cheng, C., Uchida, Y., et al.: Fyn and Cdk5 mediate semaphorin-3A signaling, which is involved in regulation of dendrite orientation in cerebral cortex. Neuron 35:907, 2002. 324. Sattler, M., Salgia, R., Shrikhande, G., et al.: Differential signaling after beta1 integrin ligation is mediated through binding of CRKL to p120(CBL) and p110(HEF1). J. Biol. Chem. 272:14320, 1997. 325. Sayas, C. L., Avila, J., and Wandosell, F.: Glycogen synthase kinase-3 is activated in neuronal cells by Galpha12 and Galpha13 by Rho-independent and Rho-dependent mechanisms. J. Neurosci. 22:6863, 2002. 326. Schaefer, A. W., Kabir, N., and Forscher, P.: Filopodia and actin arcs guide the assembly and transport of two populations of microtubules with unique dynamic parameters in neuronal growth cones. J. Cell Biol. 158:139, 2002. 327. Schimmelpfeng, K., Gogel, S., and Klambt, C.: The function of leak and kuzbanian during growth cone and cell migration. Mech. Dev. 106:25, 2001. 328. Schmidt, A., and Hall, A.: Guanine nucleotide exchange factors for Rho GTPases: turning on the switch. Genes Dev. 16:1587, 2002. 329. Schoenwaelder, S. M., Petch, L. A., Williamson, D., et al.: The protein tyrosine phosphatase Shp-2 regulates RhoA activity. Curr. Biol. 10:1523, 2000. 330. Schwanzel-Fukuda, M., Bick, D., and Pfaff, D. W.: Luteinizing hormone-releasing hormone (LHRH)-expressing cells do not migrate normally in an inherited hypogonadal (Kallmann) syndrome. Brain Res. Mol. Brain Res. 6:311, 1989. 331. Schwanzel-Fukuda, M., and Pfaff, D. W.: Origin of luteinizing hormone-releasing hormone neurons. Nature 338:161, 1989. 332. Seeger, M., Tear, G., Ferres-Marco, D., and Goodman, C. S.: Mutations affecting growth cone guidance in Drosophila: genes necessary for guidance toward or away from the midline. Neuron 10:409, 1993. 333. Semaphorin Nomenclature Committee: Unified nomenclature for the semaphorins/collapsins. Semaphorin Nomenclature Committee. Cell 97:551, 1999. 334. Serafini, T., Colamarino, S. A., Leonardo, E. D., et al.: Netrin-1 is required for commissural axon guidance in the developing vertebrate nervous system. Cell 87:1001, 1996.

479

335. Serafini, T., Kennedy, T. E., Galko, M. J., et al.: The netrins define a family of axon outgrowth-promoting proteins homologous to C. elegans UNC-6. Cell 78:409, 1994. 336. Shamah, S. M., Lin, M. Z., Goldberg, J. L., et al.: EphA receptors regulate growth cone dynamics through the novel guanine nucleotide exchange factor ephexin. Cell 105:233, 2001. 337. Shan, W., Yoshida, M., Wu, X. R., et al.: Neural (N-) cadherin, a synaptic adhesion molecule, is induced in hippocampal mossy fiber axonal sprouts by seizure. J. Neurosci. Res. 69:292, 2002. 338. Shi, Y., Alin, K., and Goff, S. P.: Abl-interactor-1, a novel SH3 protein binding to the carboxy-terminal portion of the Abl protein, suppresses v-abl transforming activity. Genes Dev. 9:2583, 1995. 339. Shirasaki, R., Katsumata, R., and Murakami, F.: Change in chemoattractant responsiveness of developing axons at an intermediate target. Science 279:105, 1998. 340. Shneker, B. F., and Fountain, N. B.: Epilepsy. Dis. Mon. 49:426, 2003. 341. Simpson, J. H., Kidd, T., Bland, K. S., and Goodman, C. S.: Short-range and long-range guidance by slit and its Robo receptors: Robo and Robo2 play distinct roles in midline guidance. Neuron 28:753, 2000. 342. Skaper, S. D., Moore, S. E., and Walsh, F. S.: Cell signalling cascades regulating neuronal growth-promoting and inhibitory cues. Prog. Neurobiol. 65:593, 2001. 343. Sloviter, R. S.: The functional organization of the hippocampal dentate gyrus and its relevance to the pathogenesis of temporal lobe epilepsy. Ann. Neurol. 35:640, 1994. 344. Snow, P. M., Bieber, A. J., and Goodman, C. S.: Fasciclin III: a novel homophilic adhesion molecule in Drosophila. Cell 59:313, 1989. 345. Snyder, S. E., Li, J., Schauwecker, P. E., et al.: Comparison of RPTP zeta/beta, phosphacan, and trkB mRNA expression in the developing and adult rat nervous system and induction of RPTP zeta/beta and phosphacan mRNA following brain injury. Brain Res. Mol. Brain Res. 40:79, 1996. 346. Sonenberg, N., and Gingras, A. C.: The mRNA 5⬘ capbinding protein eIF4E and control of cell growth. Curr. Opin. Cell Biol. 10:268, 1998. 347. Song, H., Ming, G., He, Z., et al.: Conversion of neuronal growth cone responses from repulsion to attraction by cyclic nucleotides. Science 281:1515, 1998. 348. Song, H. J., Ming, G. L., and Poo, M. M.: cAMP-induced switching in turning direction of nerve growth cones. Nature 388:275, 1997. 349. Soussi-Yanicostas, N., de Castro, F., Julliard, A. K., et al.: Anosmin-1, defective in the X-linked form of Kallmann syndrome, promotes axonal branch formation from olfactory bulb output neurons. Cell 109:217, 2002. 350. Soussi-Yanicostas, N., Hardelin, J. P., Arroyo-Jimenez, M. M., et al.: Initial characterization of anosmin-1, a putative extracellular matrix protein synthesized by definite neuronal cell populations in the central nervous system. J. Cell Sci. 109(Pt. 7):1749, 1996. 351. Speidel, C. C.: Studies of living nerves II. Activities of ameboid growth cones, sheath cells, and myelin segments, as revealed by prolonged observation of individual nerve fibers in frog tadpoles. Am. J. Anat. 52:1, 1933.

480

Neurobiology of the Peripheral Nervous System

352. Sperry, R. W.: Chemoaffinity in the orderly growth of nerve fiber patterns and connections. Proc. Natl. Acad. Sci. U. S. A. 50:703, 1963. 353. Steigemann, P., Molitor, A., Fellert, S., et al.: Heparan sulfate proteoglycan syndecan promotes axonal and myotube guidance by slit/robo signaling. Curr. Biol. 14:225, 2004. 354. Stein, E., and Tessier-Lavigne, M.: Hierarchical organization of guidance receptors: silencing of netrin attraction by slit through a Robo/DCC receptor complex. Science 291:1928, 2001. 355. Stein, E., Zou, Y., Poo, M., and Tessier-Lavigne, M.: Binding of DCC by netrin-1 to mediate axon guidance independent of adenosine A2B receptor activation. Science 291:1976, 2001. 356. Stoeckli, E. T., Kuhn, T. B., Duc, C. O., et al.: The axonally secreted protein axonin-1 is a potent substratum for neurite growth. J. Cell Biol. 112:449, 1991. 357. Suetsugu, S., Miki, H., and Takenawa, T.: The essential role of profilin in the assembly of actin for microspike formation. EMBO J. 17:6516, 1998. 358. Sumi, T., Matsumoto, K., Takai, Y., and Nakamura, T.: Cofilin phosphorylation and actin cytoskeletal dynamics regulated by rho- and Cdc42-activated LIM-kinase 2. J. Cell Biol. 147:1519, 1999. 359. Sun, Q., Bahri, S., Schmid, A., et al.: Receptor tyrosine phosphatases regulate axon guidance across the midline of the Drosophila embryo. Development 127:801, 2000. 360. Sundaresan, V., Mambetisaeva, E., Andrews, W., et al.: Dynamic expression patterns of Robo (Robo1 and Robo2) in the developing murine central nervous system. J. Comp. Neurol. 468:467, 2004. 361. Suter, D. M., Errante, L. D., Belotserkovsky, V., and Forscher, P.: The Ig superfamily cell adhesion molecule, apCAM, mediates growth cone steering by substratecytoskeletal coupling. J. Cell Biol. 141:227, 1998. 362. Suzuki, T., Nakamoto, T., Ogawa, S., et al.: MICAL, a novel CasL interacting molecule, associates with vimentin. J. Biol. Chem. 277:14933, 2002. 363. Swiercz, J. M., Kuner, R., Behrens, J., and Offermanns, S.: Plexin-B1 directly interacts with PDZ-RhoGEF/LARG to regulate RhoA and growth cone morphology. Neuron 35:51, 2002. 364. Takahashi, T., Fournier, A., Nakamura, F., et al.: Plexinneuropilin-1 complexes form functional semaphorin-3A receptors. Cell 99:59, 1999. 365. Tamagnone, L., Artigiani, S., Chen, H., et al.: Plexins are a large family of receptors for transmembrane, secreted, and GPI-anchored semaphorins in vertebrates. Cell 99:71, 1999. 366. Tamagnone, L., and Comoglio, P. M.: Signalling by semaphorin receptors: cell guidance and beyond. Trends Cell Biol. 10:377, 2000. 367. Tani, K., Sato, S., Sukezane, T., et al.: Abl interactor 1 promotes tyrosine 296 phosphorylation of mammalian enabled (Mena) by c-Abl kinase. J. Biol. Chem. 278:21685, 2003. 368. Tauck, D. L., and Nadler, J. V.: Evidence of functional mossy fiber sprouting in hippocampal formation of kainic acid-treated rats. J. Neurosci. 5:1016, 1985. 369. Tear, G., Harris, R., Sutaria, S., et al.: Commissureless controls growth cone guidance across the CNS midline in Drosophila and encodes a novel membrane protein. Neuron 16:501, 1996.

370. Terman, J. R., Mao, T., Pasterkamp, R. J., et al.: MICALs, a family of conserved flavoprotein oxidoreductases, function in plexin-mediated axonal repulsion. Cell 109:887, 2002. 371. Tessier-Lavigne, M., Placzek, M., Lumsden, A. G., et al.: Chemotropic guidance of developing axons in the mammalian central nervous system. Nature 336:775, 1988. 372. Tomek, W., Melo Sterza, F. A., Kubelka, M., et al.: Regulation of translation during in vitro maturation of bovine oocytes: the role of MAP kinase, eIF4E (cap binding protein) phosphorylation, and eIF4E-BP1. Biol. Reprod. 66:1274, 2002. 373. Tong, J., Killeen, M., Steven, R., et al.: Netrin stimulates tyrosine phosphorylation of the UNC-5 family of netrin receptors and induces Shp2 binding to the RCM cytodomain. J. Biol. Chem. 276:40917, 2001. 374. Torres, R., Firestein, B. L., Dong, H., et al.: PDZ proteins bind, cluster, and synaptically colocalize with Eph receptors and their ephrin ligands. Neuron 21:1453, 1998. 375. Uchida, N., Honjo, Y., Johnson, K. R., et al.: The catenin/cadherin adhesion system is localized in synaptic junctions bordering transmitter release zones. J. Cell Biol. 135:767, 1996. 376. van de Wetering, M., de Lau, W., and Clevers, H.: WNT signaling and lymphocyte development. Cell 109(Suppl.):S13, 2002. 377. Vawter, M. P., Usen, N., Thatcher, L., et al.: Characterization of human cleaved N-CAM and association with schizophrenia. Exp Neurol 172:29, 2001. 378. Vicente, A. M., Macciardi, F., Verga, M., et al.: NCAM and schizophrenia: genetic studies. Mol Psychiatry 2:65, 1997. 379. Vielmetter, J., Kayyem, J. F., Roman, J. M., and Dreyer, W. J.: Neogenin, an avian cell surface protein expressed during terminal neuronal differentiation, is closely related to the human tumor suppressor molecule deleted in colorectal cancer. J Cell Biol 127:2009, 1994. 380. Vikis, H. G., Li, W., and Guan, K. L.: The plexin-B1/Rac interaction inhibits PAK activation and enhances Sema4D ligand binding. Genes Dev 16:836, 2002. 381. Vikis, H. G., Li, W., He, Z., and Guan, K. L.: The semaphorin receptor plexin-B1 specifically interacts with active Rac in a ligand-dependent manner. Proc. Natl. Acad. Sci. U. S. A. 97:12457, 2000. 382. Vindis, C., Cerretti, D. P., Daniel, T. O., and Huynh-Do, U.: EphB1 recruits c-Src and p52Shc to activate MAPK/ERK and promote chemotaxis. J. Cell Biol. 162:661, 2003. 383. Vojtek, A., Haarer, B., Field, J., et al.: Evidence for a functional link between profilin and CAP in the yeast S. cerevisiae. Cell 66:497, 1991. 384. Waddington, J. L., Lane, A., Scully, P. J., et al.: Neurodevelopmental and neuroprogressive processes in schizophrenia: antithetical or complementary, over a lifetime trajectory of disease? Psychiatr. Clin. North Am. 21:123, 1998. 385. Wahl, S., Barth, H., Ciossek, T., et al.: Ephrin-A5 induces collapse of growth cones by activating Rho and Rho kinase. J. Cell Biol. 149:263, 2000. 386. Wang, K. H., Brose, K., Arnott, D., et al.: Biochemical purification of a mammalian slit protein as a positive regulator of sensory axon elongation and branching. Cell 96:771, 1999. 387. Wang, Y., Thekdi, N., Smallwood, P. M., et al.: Frizzled-3 is required for the development of major fiber tracts in the rostral CNS. J. Neurosci. 22:8563, 2002.

Guidance of Axons to Targets in Development and in Disease 388. Waskiewicz, A. J., Flynn, A., Proud, C. G., and Cooper, J. A.: Mitogen-activated protein kinases activate the serine/threonine kinases Mnk1 and Mnk2. EMBO J. 16:1909, 1997. 389. Watanabe, N., Madaule, P., Reid, T., et al.: p140mDia, a mammalian homolog of Drosophila diaphanous, is a target protein for Rho small GTPase and is a ligand for profilin. EMBO J. 16:3044, 1997. 390. Waterman-Storer, C. M., and Salmon, E. D.: Actomyosinbased retrograde flow of microtubules in the lamella of migrating epithelial cells influences microtubule dynamic instability and turnover and is associated with microtubule breakage and treadmilling. J. Cell Biol. 139:417, 1997. 391. Welch, M. D., Iwamatsu, A., and Mitchison, T. J.: Actin polymerization is induced by Arp2/3 protein complex at the surface of Listeria monocytogenes. Nature 385:265, 1997. 392. Wennerberg, K., Forget, M. A., Ellerbroek, S. M., et al.: Rnd proteins function as RhoA antagonists by activating p190 RhoGAP. Curr. Biol. 13:1106, 2003. 393. Wenzel, H. J., Woolley, C. S., Robbins, C. A., and Schwartzkroin, P. A.: Kainic acid-induced mossy fiber sprouting and synapse formation in the dentate gyrus of rats. Hippocampus 10:244, 2000. 394. Whitford, K. L., Marillat, V., Stein, E., et al.: Regulation of cortical dendrite development by Slit-Robo interactions. Neuron 33:47, 2002. 395. Wilkinson, D. G.: Multiple roles of EPH receptors and ephrins in neural development. Nat. Rev. Neurosci. 2:155, 2001. 396. Williams, E. J., Furness, J., Walsh, F. S., and Doherty, P.: Activation of the FGF receptor underlies neurite outgrowth stimulated by L1, N-CAM, and N-cadherin. Neuron 13:583, 1994. 397. Wills, Z., Bateman, J., Korey, C. A., et al.: The tyrosine kinase Abl and its substrate enabled collaborate with the receptor phosphatase Dlar to control motor axon guidance. Neuron 22:301, 1999. 398. Wills, Z., Emerson, M., Rusch, J., et al.: A Drosophila homolog of cyclase-associated proteins collaborates with the Abl tyrosine kinase to control midline axon pathfinding. Neuron 36:611, 2002. 399. Winberg, M. L., Noordermeer, J. N., Tamagnone, L., et al.: Plexin A is a neuronal semaphorin receptor that controls axon guidance. Cell 95:903, 1998. 400. Winberg, M. L., Tamagnone, L., Bai, J., et al.: The transmembrane protein Off-track associates with Plexins and functions downstream of Semaphorin signaling during axon guidance. Neuron 32:53, 2001. 401. Witke, W., Podtelejnikov, A. V., Di Nardo, A., et al.: In mouse brain profilin I and profilin II associate with regulators of the endocytic pathway and actin assembly. EMBO J. 17:967, 1998. 402. Wodarz, A., and Nusse, R.: Mechanisms of Wnt signaling in development. Annu. Rev. Cell Dev. Biol. 14:59, 1998. 403. Wong, E. V., Schaefer, A. W., Landreth, G., and Lemmon, V.: Casein kinase II phosphorylates the neural cell adhesion molecule L1. J. Neurochem. 66:779, 1996. 404. Wong, E. V., Schaefer, A. W., Landreth, G., and Lemmon, V.: Involvement of p90rsk in neurite outgrowth mediated by the cell adhesion molecule L1. J. Biol. Chem. 271:18217, 1996.

481

405. Wong, K., Ren, X. R., Huang, Y. Z., et al.: Signal transduction in neuronal migration: roles of GTPase activating proteins and the small GTPase Cdc42 in the Slit-Robo pathway. Cell 107:209, 2001. 406. World Health Organization: The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva, World Health Organization, 1992. 407. Wybenga-Groot, L. E., Baskin, B., Ong, S. H., et al.: Structural basis for autoinhibition of the Ephb2 receptor tyrosine kinase by the unphosphorylated juxtamembrane region. Cell 106:745, 2001. 408. Xu, X. M., Fisher, D. A., Zhou, L., et al.: The transmembrane protein semaphorin 6A repels embryonic sympathetic axons. J. Neurosci. 20:2638, 2000. 409. Yamasaki, M., Thompson, P., and Lemmon, V.: CRASH syndrome: mutations in L1CAM correlate with severity of the disease. Neuropediatrics 28:175, 1997. 410. Yang, J., Si, T., Ling, Y., et al.: Association study of the human FZD3 locus with schizophrenia. Biol. Psychiatry 54:1298, 2003. 411. Yang, N., Higuchi, O., Ohashi, K., et al.: Cofilin phosphorylation by LIM-kinase 1 and its role in Rac-mediated actin reorganization. Nature 393:809, 1998. 412. Yang-Snyder, J., Miller, J. R., Brown, J. D., et al.: A frizzled homolog functions in a vertebrate Wnt signaling pathway. Curr. Biol. 6:1302, 1996. 413. Yu, D. H., Qu, C. K., Henegariu, O., et al.: Protein-tyrosine phosphatase Shp-2 regulates cell spreading, migration, and focal adhesion. J. Biol. Chem. 273:21125, 1998. 414. Yu, H. H., Zisch, A. H., Dodelet, V. C., and Pasquale, E. B.: Multiple signaling interactions of Abl and Arg kinases with the EphB2 receptor. Oncogene 20:3995, 2001. 415. Yu, T. W., Hao, J. C., Lim, W., et al.: Shared receptors in axon guidance: SAX-3/Robo signals via UNC-34/Enabled and a Netrin-independent UNC-40/DCC function. Nat. Neurosci. 5:1147, 2002. 416. Zanata, S. M., Hovatta, I., Rohm, B., and Puschel, A. W.: Antagonistic effects of Rnd1 and RhoD GTPases regulate receptor activity in Semaphorin 3A-induced cytoskeletal collapse. J. Neurosci. 22:471, 2002. 417. Zeng, L., Fagotto, F., Zhang, T., et al.: The mouse Fused locus encodes Axin, an inhibitor of the Wnt signaling pathway that regulates embryonic axis formation. Cell 90:181, 1997. 418. Zhang, X. F., Schaefer, A. W., Burnette, D. T., et al.: Rhodependent contractile responses in the neuronal growth cone are independent of classical peripheral retrograde actin flow. Neuron 40:931, 2003. 419. Zhou, L., White, F. A., Lentz, S. I., et al.: Cloning and expression of a novel murine semaphorin with structural similarity to insect semaphorin I. Mol. Cell. Neurosci. 9:26, 1997. 420. Zimmer, M., Palmer, A., Kohler, J., and Klein, R.: EphBephrinB bi-directional endocytosis terminates adhesion allowing contact mediated repulsion. Nat. Cell Biol. 5:869, 2003. 421. Zisch, A. H., Stallcup, W. B., Chong, L. D., et al.: Tyrosine phosphorylation of L1 family adhesion molecules: implication of the Eph kinase Cek5. J. Neurosci. Res. 47:655, 1997.

22 Aging in the Peripheral Nervous System T. COWEN, B. ULFHAKE, AND R. H. M. KING

Nature of Aging in the PNS Primary Sensory Neurons Nerve Trunks Autonomic Neurons Mechanisms of Aging in the PNS

Mechanisms Underlying Sensory Impairments during Aging Mechanisms Underlying Age-Related Changes in Nerve Trunks

A characteristic of aging in the peripheral nervous system (PNS), as in other areas of the nervous system, is the locally specific nature of the changes that occur. ‘Selective vulnerability’ in aging, as this phenomenon has been called, seems to characterize the effects of aging in all areas of the PNS, to a greater or lesser extent. Long, myelinated nerve fibers are more vulnerable to age-related change compared with shorter fibers. Large sensory neurons appear to be more vulnerable than small ones. Sympathetic neurons supplying some target tissues are more vulnerable than those supplying others. Enteric neurons are particularly vulnerable to age-related neuronal cell death in which diet plays an important role. All these groups of neurons undergo dying back of peripheral axon terminals and also probably of fine dendritic branches to some extent. Interestingly, in peripheral as in central neurons, the definitive demonstration of neuron loss during aging is relatively rare, at least in the mammalian nervous system, whereas neuronal atrophy and dying back of fine terminal axonal and dendritic fibers are much more common. In at least some of these instances, disturbance of trophic interactions between peripheral nerve fibers and their target sensory receptors or effectors is thought to contribute to age-related selective vulnerability.

Mechanisms Underlying Age-Related Impairments in Autonomic Neurons Conclusions

NATURE OF AGING IN THE PNS Primary Sensory Neurons Functional Impairments in Aging Sensory Neurons Disturbances of gait cycle and balance control as well as increased thresholds for exteroceptive, proprioceptive, and vibratory sensations are some of the most common problems seen in the elderly.21,30,51,54,68,71,84,104,127,137,144,177,183,212,222 These impairments contribute to characteristic traits of aging such as an increased frequency of falls and fractures, impediment of the activities of daily life, and loss of independence.136 Falls and their complications constitute a major cause of hospitalization in senescent individuals, and a significant correlation has been found between the occurrence of falls in the elderly and decreased proprioceptive function.30 Structural Impairments in Aging Sensory Neurons Age-related changes are more conspicuous in myelinated than in unmyelinated axons, and there appears to be a good agreement between the severity of sensory nerve lesions and functional sensory deficits. Numbers of myelinated 483

484

Neurobiology of the Peripheral Nervous System

sensory axons are either unaltered or moderately decreased in peripheral nerves and spinal roots of aged subjects.126,129,157,180,187,194,197,213 It is at present unclear whether unmyelinated sensory axons are affected to the same extent. Consistent with a biased distribution of sensory deficits, lesions in both humans (bipedal) and rodents (quadripedal) are more widespread in lumbar than in cervical nerves.17,32,182,212,213,222 Despite relatively restricted losses of preterminal sensory nerve fibers and sensory neurons, there are numerous and selective losses of sensory receptors and nerve endings in the different target tissues of aged individuals. A major specific effect of aging on tactile skin receptors has been demonstrated in both rodents and humans. For example, human finger and footpads show a pronounced decline in the number of pacinian corpuscles, Meissner corpuscles, and Merkel cell–neurite complexes in senescence.26,34,36,192,225 Consistent with these morphologic changes, Gescheider and collaborators84 found that the thresholds for vibrotactile stimuli, mediated by pacinian, Meissner, and Ruffini corpuscles and Merkel endings, increased with age, with the threshold of pacinian corpuscles the most affected. Moreover, tests of tactile responses in the finger pulp194 indicate that the distal axon, including the endings, seems to be more severely affected than the proximal axon. The receptors most vulnerable in aging are therefore primarily mechanoreceptors in skin and muscle whose axons are the large myelinated A␣ and A␤ fibers. These peripheral fibers originate in the large, pale neurons of the dorsal root ganglia (DRG)140 and terminate either in laminae III through VI or, for some muscle afferents, in laminae VII and IX of the deep dorsal horn of the spinal cord.224 Taken together, it seems as if the decrease in tactile sensitivity with advancing age may be explained by a decreased receptor density resulting from a degenerative process that probably involves mainly the peripheral receptor/nerve ending rather than more proximal regions of the parent sensory neuron. In addition, the possibility cannot be excluded that conventional morphologic techniques are less sensitive than functional tests in detecting age-related aberrations in mechanoreception. Further evidence of functional impairments of aging sensory neurons comes from studies showing that conduction velocity declines, albeit at a slower rate and more uniformly than in motor nerve fibers (see Functional Impairments in Axons below). In the median nerve, sensory axon conduction velocity declines at a rate of 2 m/s per decade, whereas in the ulnar nerve the decrease is 1m/s per decade between 20 and 55 years of age, increasing to 3 m/s per decade after 55 years of age.31 In addition to altered conduction velocity, the amplitude of nerve action potentials in sensory nerves from humans at 70 years of age has declined to half the value at 20 years. Whereas data on the effects of age in peripheral sensory nerves are abundant, much less is known about how these

processes affect the central terminations of primary sensory neurons.20 Taking advantage of tracer substances with differing affinities for different classes of primary sensory neurons, evidence was obtained for a dramatic and selective decrease in the density of myelinated, but not unmyelinated, primary afferents terminating in the dorsal horn of aged rodents (Fig. 22–1).22 The substantial reduction in mechanoreceptive input to the central nervous system (CNS) includes muscle spindle and Golgi tendon organ afferents terminating in the deep dorsal horn and the spinal motor nucleus. Furthermore, the extent of the loss of mechanosensory input to the spinal cord agreed with the degree as well as the distribution of sensorimotor impairment of the animals. Although the major causes of age-related behavioral changes remain to be determined, loss of neurons has often been considered to be a prime cause of functional deficits during aging. However, evidence from a growing number of studies shows that cell loss is far less prominent than previously thought.4,39,43,145,159,191,223 Unbiased estimates of primary sensory neurons in rodents show only a small decrease (⬍15%) of both large neurons with myelinated processes and small neurons with unmyelinated processes at advanced age. More importantly, there is no clear association between the degree of cell loss and sensory deficits among the individual animals (Fig. 22–2).21 Earlier studies of numbers of DRG neurons are contradictory, with some showing a notable loss of neurons during aging,78,161 whereas others report no change.135,169 Part of the inconsistency in these studies can probably be attributed to the fact that they were performed using indirect counting techniques based on two-dimensional probes, thereby producing biased results.38 Signaling Molecules in Aging Sensory Neurons The loss of epidermal and dermal innervation involves both sensory and autonomic components (see sections on Myelinated Axons and on Selective Vulnerability of Autonomic Neurons to Age-Related Neurodegeneration below). Aminergic, cholinergic, and peptidergic nerve fibers are decreased in number, with a decreased innervation of blood vessels, subepidermis, and sweat glands.2,45,48,59,75,155,164 Impaired innervation of sweat glands and skin blood vessels will have a negative impact on wound healing and body temperature control118 (see below). In this context it is important to consider that the absence of signaling molecules such as neuropeptides in peripheral nerves does not necessarily reflect loss of the axons. At the cellular level, primary sensory neurons show a complex pattern of changes during aging. Besides the axon aberrations and the small unselective loss of neurons (see above), there is evidence for a selective cell body atrophy (large pale cell bodies)21,171,180 and phenotypic alterations in the expression pattern of signaling molecules, receptors, and cytoskeletal proteins (reviewed by Ulfhake et al.211). Primary sensory neurons

Aging in the Peripheral Nervous System

FIGURE 22–1 Images showing cholera toxin B subunit (CTB) labeling in the lower lumbar spinal cord dorsal horns (A and B), ventral horn (C and D), and dorsal root ganglia (E and F) 3 days after injection of CTB into the tibial nerve of young adult (Ad; A, C, and E) or aged (Ag; B, D, and F) rats. Tibial nerve CTB injections in young adult rats resulted in a moderate to dense labeling of myelinated primary afferent axons terminating in the medial part of the deep dorsal horn (laminae III through VI; A) and in the ventral horn, including its motor nucleus (VII and IX; C) as well as in motoneurons (C). Aged rats disclosed a dramatic decrease in CTB labeling both in the deep dorsal horn (B) and in the ventral horn (D), whereas the labeling appeared unaffected in motoneurons (D). CTB labeling in the corresponding dorsal root ganglia was confined mainly to large neurons (E and F), with no difference observed between young adult and aged individuals. Insets in C and D show motoneurons at higher magnification. Scale bar: 150 ␮m (A–D); 75 ␮m (E and F); 50 ␮m (insets C and D). (From Bergman, E., and Ulfhake, B.: Evidence for loss of myelinated input to the spinal cord in senescent rats. Neurobiol. Aging 23:271, 2002.)

485

486

Neurobiology of the Peripheral Nervous System

20 103

Total number of DRG neurons

18 103

16 103 L4

14 103

12 103

C5

10 103

8 103

Young adult

Stage I

Stage II

Stage III

FIGURE 22–2 Graphic representation of the relation between the total number of neurons in the fifth cervical (C5; below hatched line) and fourth lumbar (L4; above hatched line) dorsal root ganglia (DRG), age (young adult: open symbols; aged: filled symbols), and stage of sensory impairment (stages I through III, in which stage I represents symptoms of a mild degree of deficits and stage III corresponds to symptoms of advanced deficits). Male and female rats are indicated with squares and circles, respectively. Note the lack of correlation between neuron loss and symptoms among the aged rats at both DRG levels studied. (From Bergman, E., and Ulfhake, B.: Loss of primary sensory neurons in the very old rat: neuron number estimates using the disector method and confocal optical sectioning. J. Comp. Neurol. 396:211, 1998.)

synthesize neuropeptides, which act as transmitters and/or modulators of sensory transmission at the central termination of these neurons.102,139,224 Consistent with the decrement of neuropeptides in peripheral sensory nerves and endings, there is a decreased expression of both calcitonin gene–related peptide and substance P in aged DRG neurons.18,75 In addition to their central effects, these neuropeptides exert effects in the periphery, where they are thought to be involved in inflammatory responses and wound healing.94,103,119,165 The lower level of neuropeptide expression in subsets of sensory neurons during aging provides a molecular and cellular substrate for the previously described diminished axon reflex in senescence.98,120

Nerve Trunks Functional Impairments in Axons As previously stated (see Functional Impairments in Aging Sensory Neurons above), the main clinical signs of functional impairment of peripheral nerves in the elderly are

loss of ankle jerks and impaired vibration sense starting in the feet, combined with pain, temperature, or light touch impairment in 25% of patients between 70 and 85 years of age.51,104 The pain and touch impairments were ascribed to peripheral nerve degeneration.51,104 Although not all the earlier studies excluded impaired glucose tolerance, several reliable studies have concluded that there is a significant decline in the threshold of vibratory sensation in healthy people after the age of 50 years.88,173,203 This correlates with evidence of loss of mechanoreceptors (see Structural Impairments in Aging Sensory Neurons above) and axonal loss affecting predominantly large myelinated fibers (including those involved in mechanoreception) in older people over the age of 60 years, together with abnormalities in the vasa nervorum.207 However, age-related functional deficits in peripheral nerves are not a consistent observation, and differentiating clinically between the effects of age on central versus peripheral neurons and on neurons versus the muscles or receptors they innervate is difficult. Thus a recent, exhaustive study of 200 people with a mean age of 80 years found only 3 people with absent ankle jerks that could not be explained by disease, trauma, or other factors.105 This compares well with a study on elderly patients with chronic inflammatory demyelinating polyneuropathy in whom only 2% of the normal controls lacked sural nerve action potentials.219 Effects on motor nerves are more marked, as previously stated, with conduction velocity declining with age by up to 30% in humans and animals.167,220,221 In sensory nerves, the decrease in conduction velocity is less than in motor nerves, although recording techniques do not allow reliable distinction between responses in large myelinated sensory and motor axons (see sections on Myelinated Axons and on Unmyelinated Axons below). In another study, sural responses were unobtainable in 10% of subjects age 80 and over, indicating the lack of consistency between different studies of similar nerve populations. Despite their shorter lifespan, experimental animals also show an increasing number of abnormalities in their peripheral nerves with advancing age.9,28,29,126,131,205 Electrophysiological studies in animals indicate a functional decline similar to that seen in humans. For example, a study in old cats showed a decrease in motor nerve conduction velocity with age and myelin abnormalities.3 Although conduction velocity in the nerve to the flexor hallucis longus muscle of aged rats was reported to be unchanged,221 it was reduced in the muscle spindles.158 These fibers are among the largest, indicating that the size of the fibers may be important in determining which nerves are vulnerable to aging, rather than whether they are motor or sensory. Structural Impairments in Axons Examination of biopsy or necropsy specimens helps to define the morphological changes that underlie functional impairments. Few studies have examined all the different

Aging in the Peripheral Nervous System

parts of the PNS, including primary sensory neurons, peripheral motor nerve fibers, and autonomic neurons. Despite the difficulties associated with obtaining wellpreserved human material for light and electron microscopy, several earlier as well as more recent studies in humans, and also in experimental animals, have consistently described age-related axonal loss in spinal roots and peripheral nerves.65,168 There are indications that the loss of axons does not affect all areas to the same extent; a preferential loss of nerve fibers in the hind limbs of old rats has been demonstrated compared with those supplying the forelimbs.97 Neurogenic atrophy in human skeletal muscle has been reported as more marked in muscles of the legs than in the arms, with intermediate findings in the trunk.208 Both observations suggest that the larger and longer axons supplying the muscles of the hind limbs are more vulnerable to aging than the smaller and shorter axons supplying the forelimbs. Age-related abnormalities of motor end plates in the middle portion of thigh muscles, consistent with axonal atrophy, appeared different from those produced by dying-back neuropathies such as are caused by acrylamide.72 Axonal loss may, of course, be due to neuronal death. However, relatively few studies prove this correlation. Axonal degeneration in the cochlear nerves was observed at 6 months in Sprague-Dawley rats, while axonal atrophy was not seen before 35 months.101 These changes are probably related to the loss of spiral ganglion cells.116 Changes in the plantar nerves of Wistar rats were also first evident at 6 months, and by 24 months 55% of teased nerve fibers were abnormal.197 Both axonal degeneration and regeneration and segmental demyelination occurred, but the decrease in total fiber counts was minimal and the fate of the associated neurons was not examined. In the same animals, changes in the tibial nerve were not found until after 18 months of age and by 24 months only involved 30% of fibers.197 Teased fiber studies preferentially examine the larger fibers, so these figures cannot be extrapolated to the whole fiber population. No significant axonal loss has been found in the inferior alveolar nerve of the rat (C. S. Johansson, personal communication, 2002). Possible explanations for the variation in changes found between different sites could be the effects of local trauma on the foot nerves as suggested by work on guinea pigs.74 The high proportion (72%) of degenerating fibers in the distal tail nerves of 2-year-old rats may also be partly the result of damage produced by handling the animals by the tail123,125 (Fig. 22–3). Alternatively, damage to foot and tail nerves may result from their greater axonal length, as suggested above. This would not, however, explain the extensive degeneration and fiber loss found in the cochlear nerve,101 where distal axonal atrophy may be involved. Besides the apparent loss of axons, major stigmata of aging in peripheral axons are axon (neuroaxonal) dystrophy, axon atrophy, and disturbed myelination. Axon dystrophy

487

and atrophy result in a loss of peripheral target connections both in somatosensory (see Primary Sensory Neurons above) and autonomic pathways193 (i.e., a loss of innervation). Disturbed myelination will affect not only axon conduction (impulse propagation as well as propagation velocity) but also other aspects of axon biology, including the structural integrity of the axon. Axonal dystrophy is a process initially confined to the terminal part of the axon, which again involves certain axons but not others. For example, the extended projections of myelinated primary sensory neurons terminating in the dorsal column of the spinal cord are among the most severely affected.73,111 Axon atrophy, that is, a reduction in diameter of axons resulting in abnormally thick myelin, has been reported in both roots and peripheral nerves,115,126,184 but unequivocal evidence of dying-back atrophy is still lacking. Within aging axons, light microscopic and ultrastructural studies on 24- to 30-month-old Sprague-Dawley rats show abnormalities such as axonal glycogenosomes, degenerate mitochondria, dense membranous bodies, polyglucosan or Lafora bodies, and decorated particles (Fig. 22–4). These abnormalities occur with increasing frequency distally in the tibial and plantar nerves. These abnormalities can, however, also be found close to the nerve cell body in the DRG (Fig. 22–5). Glycogen granules are first seen during aging in axonal mitochondria. These develop into glycogenosomes and are then transformed into polyglucosan bodies (Fig. 22–6).174,214 Active axonal degeneration, indicated by bands of Büngner, and demyelination are rarely prominent and are only seen with electron microscopy. Schwann cell infoldings into the axon are sometimes found, as are Hirano bodies (eosinophilic fusiform bodies with a paracrystalline fine structure100) in the adaxonal Schwann cell cytoplasm.123 It has been suggested that trauma, particularly chronic pressure, may be a factor in the changes found in the plantar nerves, especially because these changes were not seen in branches of the tibial nerve to the gastrocnemius muscle.92 However, the plantar nerves are also among the longest nerves, and the occurrence of loss, dystrophy, and atrophy of other extended axons in nerves with more protected paths, such as visceral autonomic nerves (see Autonomic Neurons below), the cochlear nerve (see below), and the distal tail nerves,123,125 indicates that axon length is more strongly associated with selective vulnerability in aging than trauma. Myelinated Axons In addition to axonal degeneration, peripheral nerves and spinal nerve roots of aged animals show evidence of abnormal myelin, including demyelination and the formation of large intramyelinic swellings.17,32,41,85,126,156,180,197,205,213 Myelin abnormalities occur relatively early in life, appearing soon after 6 months in rats,186 suggesting that nerves have barely finished growing before they start to exhibit signs of regression.

488

Neurobiology of the Peripheral Nervous System

A

B

Myelinated fiber profiles become more irregular with age, with extensive infoldings and outpouchings of the myelin sheath. Such appearances are particularly prominent in the spinal roots and may be explained by axonal shrinkage.126,130 In support of the idea that large myelinated fibers are more vulnerable to the effects of aging compared with small unmyelinated nerves, morphological and teased fiber studies

FIGURE 22–3 A, Tail nerve from an old rat. There is considerable loss of myelinated fibers, many of which have been replaced by clumps of Schwann cell processes (black arrow) distinguishable by their lack of axons from normal Remak fibers containing unmyelinated axons (yellow arrow) (shown in more detail in inset). Most of the myelinated fibers have inappropriately thin myelin for axon diameter. There is no evidence of regeneration or demyelination. Resin section stained with thionin and acridine orange. Bar: 20 ␮m. B, Tail nerve from a 6-month-old normal rat. The myelinated fibers are closely packed with very little collagen between them. Resin section stained with thionin and acridine orange. Bar: 20 ␮m.

showed a reduction in numbers of larger myelinated fibers and an increased variability in internodal length. These changes only appeared after 60 years of age in humans.138 In contrast, several studies found little change in the population of small myelinated fibers.185,199,207 The increased variability in internodal length may result from demyelination and remyelination, or from axonal degeneration and regeneration.

Aging in the Peripheral Nervous System

489

A FIGURE 22–4 A, Plantar nerve from a 24-month-old male Sprague-Dawley rat showing a reduction in myelinated fiber density. There are round inclusions in the axons of many of the fibers. The yellow inclusions are polyglucosan bodies (red arrows); the blue-gray ones are glycogenosomes (black arrows). Some myelin sheaths are inappropriately thin, suggesting remyelination (green arrow). There are occasional macrophages (asterisk). Even the apparently normal myelin sheaths have a more irregular profile with numerous infoldings and outpouchings. Resin section stained with thionin and acridine orange. Bar: 20 ␮m. B, Plantar nerve from a normal 6-month-old rat. The fibers are closely packed. Some have a crenated appearance because they have been sectioned through the paranode (arrow). Resin section stained with thionin and acridine orange. Bar: 20 ␮m.

B

When teased fibers are examined, regenerated fibers show many consecutive, similarly myelinated internodes of the same short length, whereas remyelinated fibers have differing internodal lengths and myelin thickness. Both regeneration and remyelination result in an increased variability in the thickness of the myelin sheath relative to the diameter of its

axon.109 The most dramatic change seen in nerves from older animals is the formation of the large intramyelinic edematous spaces, commonly referred to as myelin bubbles or balloons, as previously mentioned. These appear from 6 months onward and are largest in the spinal roots (Fig. 22–7). Myelin changes are not necessarily associated with axon atrophy,

490

Neurobiology of the Peripheral Nervous System

FIGURE 22–5 DRG from a 26-month-old SpragueDawley rat. Several of the axons contain inclusions. Yellowish ones are polyglucosan bodies (red arrow) and the bluish ones are glycogenosomes (black arrows). The very thinly myelinated fibers are probably remyelinated (green arrows). Resin section stained with thionin and acridine orange. Bar: 20 ␮m.

being found more frequently in the spinal roots, whereas axonal abnormalities are more prominent in the peripheral nerves. The often dense and shrunken appearance of axons running through the myelin bubbles could be due to compression by the edematous fluid that fills them. Segmental demyelination has been reported from a relatively early age in the hind limb nerves of experimental animals. Abnormalities of both axon and myelin are far fewer and appear later in life in nerves such as the phrenic that are relatively protected from trauma and have less extended axons. Focal intramyelinic edema131,156 is more prominent in the spinal roots and phrenic nerve than in the sciatic nerve and its branches but, again, only in older animals. Myelin bubbles are most conspicuous in lumbar spinal roots from approximately 2 years of age in rats.17 This leads to demyelination (Fig. 22–8), and repeated episodes may produce onion bulb formation.64 The axons are frequently unaffected apart from reduction in diameter.131 Extensive axonal degeneration seems to be more common in Charles River rats compared with the SpragueDawley and Wistar strains.41,85 Rarely, severe cystic changes may be found in spinal roots, with axonal degeneration and sprouting, deposition of cholesterol crystals, and macrophage infiltration.32,123 These may result from unrecognized pathologies of the vertebrae and intervertebral discs.32 The cause of focal intramyelinic edema (Fig. 22–8) is still unclear. It was first reported in rats proximal to neuromas, suggesting a relationship with axonal atrophy,202 although axonal atrophy more commonly produces small, dense axons with an abnormally thick myelin sheath. The animals in question were 18 months old. The myelin abnormalities may therefore be due to age or to an age-related alteration in the reaction to injury. Two independent processes may affect myelination: one primary, involving alterations in Schwann cells, and the other secondary to axonal shrinkage. It is generally considered that the primary process is more important. Axonal atrophy

FIGURE 22–6 High-power view of a longitudinally sectioned myelinated nerve fiber in the median plantar nerve of a 24-month-old rat. There are several glycogenosomes in the axon (arrows). The myelin sheath appears normal (asterisk). Ultrathin section contrasted with uranyl acetate and lead citrate. Bar: 1 ␮m.

Aging in the Peripheral Nervous System

491

A

FIGURE 22–7 A, Ventral root from a 24-month-old rat containing many myelinated fibers affected by intramyelinic edema (asterisks). There is little axonal abnormality. Resin section stained with thionin and acridine orange. Bar: 50 ␮m. B, Ventral root from a 4-month-old rat showing closely packed normal myelinated fibers. Resin section stained with thionin and acridine orange. Bar: 50 ␮m.

B

leads to paranodal demyelination, seen in teased fiber preparations as intercalated short, thinly myelinated segments alternating in a regular pattern with segments of normal thickness and nearly normal length. Recovery may occur following this form of demyelination in animals,69 but the situation in human myelinated nerves where this has been studied is less clear.67 Changes occur with age in the proportions of the various myelin proteins in the sciatic nerve and spinal roots of old rats. The relative amounts of P0 and P1 decrease similarly at both sites.210 The relationship of these changes to the intramyelinic edema is uncertain because myelin balloons (or bubbles; see above) are much more prominent in the spinal roots than in the sciatic nerve and thus do not match the pattern of alterations in myelin proteins.

Unmyelinated Axons Relatively few studies have attempted to quantify the effects of age on the unmyelinated fiber population in peripheral nerves. The most obvious changes are in their associated Schwann cells. A unit of Schwann cell processes and associated unmyelinated axons can be conveniently called a Remak fiber after the original description.181 A reduction in the fiber population is seen as the occurrence of Schwann cell processes unassociated with axons. The density of these correlates with age in normal subjects.113 The morphological picture resulting from unmyelinated axon loss is rather different from that associated with myelinated fibers. When unmyelinated axons degenerate, the surrounding Schwann cell processes collapse inward, forming flattened sheets.204 These stacks of flattened

492

Neurobiology of the Peripheral Nervous System

ial,109,113,168 suggesting that the difference was not the result of poor preservation or other artifact.113 Regarding fiber size, one study found that size distribution of unmyelinated axon diameters was unimodal in younger subjects but developed a lower second peak representing smaller fibers from the fifth decade,109 whereas another found a defined age-related bimodal distribution of fiber size. The increase in proportion of very small axons may result from degeneration of the larger axons and their subsequent replacement by regenerative sprouts of smaller diameter. Contamination can also occur, with unmyelinated axon sprouts arising after myelinated fiber degeneration,168 but these will probably not be as small as those resulting from regeneration of unmyelinated fibers per se.

FIGURE 22–8 Electron micrograph of a dorsal root from a 24-month-old rat demonstrating demyelinated fibers (black arrow), thinly remyelinated ones (white arrows), and intramyelinic edema (asterisk). Section contrasted with lead citrate and uranyl actetate. Bar: 5 ␮m.

processes do not resemble bands of Büngner left by myelinated fiber degeneration, which consist of rounded cell processes often within the original Schwann cell basal lamina. Collections of Schwann cell sheets were reported in studies of changes of the aging human sural nerve occurring after the fourth decade, where they were interpreted to indicate loss of axons.168 A study of younger subjects between 17 and 47 years of age reported no difference in incidence of Schwann cell sheets,14 probably because they were too young, since Schwann cell sheets were numerous in older subjects aged 83 and 88 years.196 In addition, Schwann cell processes encircling bundles of collagen fibrils, apparently in place of axons, may be found, although the association with axon loss is not proven (see Chapter 3). Morphometric studies of age changes in unmyelinated axon size and diameter give varying results between animals and humans. In the pelvic nerve of rats aged 30 to 37 months, the number of small fibers (⬍0.7 ␮m) was selectively decreased while larger ones were unaffected. The distributions, however, remained unimodal.162 In the mouse tibial nerve, the smallest unmyelinated fibers also declined by 50% at 27 months of age, with the distribution again remaining unimodal.35 These results contrast with a lack of any effect of age on density of unmyelinated axons in the human sural nerve in biopsy and postmortem mater-

Vascular Changes Age-related changes are observed in several features of the blood vessels supplying nerve trunks. For example, endoneurial blood vessels in human nerves often show a thickening of their basal lamina ensheathment during aging,109 which resembles that seen at earlier ages in patients with diabetic neuropathy,24 and may result from the accumulation of less degradable glycated basal laminal proteins.124 Although not confirmed by animal studies, this may contribute to age-related episodes of local ischemia sufficient to damage the associated nerve trunk. The formation and accumulation of advanced glycation end products (AGEs) during aging may be particularly important because they reduce nitric oxide–dependent vasodilatation as well as altering the physical properties of vascular connective tissue with significant pathological effects. In support of this possibility, administration of AGE-modified albumin to normal rats and rabbits produced an increase in vascular permeability and a decrease in vasodilator responses.217 Other changes involving peripheral blood vessels include a decreased axon reflex associated with impairments in inflammatory responses and wound healing such as are involved in aging.118,146,172 In addition, in 24-month-old Sprague-Dawley rats there is altered sensitivity of perineurial blood vessels to the vasoconstrictors endothelin-1, noradrenaline, and angiotensin II, possibly as a result of alterations in the function or expression of their receptors.121,122 The same authors showed reduced vasoconstriction in aged rats in response to vasopressin.121,122

Autonomic Neurons Selective Vulnerability of Autonomic Neurons to Age-Related Neurodegeneration Of the different branches of the autonomic nervous system, the sympathetic and enteric systems have been most widely studied in relation to aging. The majority of these studies are in rodents. Like other peripheral neurons, sympathetic neurons exhibit the phenomenon of selective vulnerability.42

493

Aging in the Peripheral Nervous System

FIGURE 22–9 Selective age-related axon loss in sympathetic neurons. Immunofluorescence staining for the pan-neuronal marker PGP 9.5 shows a substantial and obvious loss of axons in cerebral blood vessels taken from 24-month-old (B) compared with 2-month-old (A) Sprague-Dawley rats.

A

Neighboring groups of nerve cells within the same ganglion undergo markedly different trajectories of change during their life history. An example is the superior cervical ganglion from which neurons project to target tissues in the head and neck. Superior cervical ganglion neurons projecting to cerebral blood vessels,206 the pineal gland,132 sweat glands of the footpad, skin, and dermal blood vessels2 all exhibit age-related axonal atrophy (Fig. 22–9). In one of these instances,132 the axon loss has been characterized as resulting from loss of whole axon branches with the associated varicosities—the sites of neurotransmitter release. In contrast, neurons projecting to the iris,83 submandibular gland, and other tissues appear to maintain their axonal structure and even continue to grow in old age. Although studied in less detail, sympathetic neurons of the coeliac ganglion also exhibit selective vulnerability during aging.10,45 Axons of neurons projecting to the gut wall10 and renal blood vessels45 exhibit loss of branches and/or reduced neurotransmitter levels during aging, whereas those projecting to the mesenteric vasculature do not. The majority of peripheral autonomic and sensory neurons exhibit continued growth and retraction on a day-to-day basis during development and in early adult life, as was elegantly demonstrated by Purves and co-workers.96,175 This dynamic state extends into adulthood (see Signaling Molecules in Aging Sensory Neurons above). Vulnerability to age-related neuronal atrophy may therefore be the result of a negative balance between ongoing nerve growth and retraction. In the main, studies of this kind are difficult to reproduce in human tissues. Therefore, the question of the adequacy of rodent models for understanding human aging processes as they affect the nervous system remains unresolved. However, in one of the few cases in which direct comparisons

B have been made, losses of axons projecting to human cerebral blood vessels have been observed25 that partly resemble those observed in the aging rat.206 Losses of peripheral fibers were accentuated in tissues from patients with Alzheimer’s disease compared with those from healthy aged individuals.25 In some cases, axon loss has been shown to be matched by age-related loss of dendritic branches of the same neurons. Thus superior cervical ganglion neurons projecting to cerebral blood vessels exhibit loss of dendrites while those projecting to the iris do not (Fig. 22–10).8 Furthermore, the preganglionic neurons of peripheral autonomic neurons may be similarly affected. Thus both pre- and postganglionic sympathetic neurons associated with spinal cord projections to the pelvic ganglia in rats exhibit age-related atrophy of dendrites and loss of synaptic contacts and inputs.58,190 These declines are associated with selective loss of sympathetic, but not parasympathetic, postganglionic neurons in the pelvic ganglia. This is one of the only sites where sympathetic neuron loss has been clearly demonstrated. Studies of the rat vagus nerve indicate that the numbers of parasympathetic and sensory fibers are unaffected by aging.188 There have been few systematic investigations of the effects of age on autonomic neuroeffector junctions. However, electron microscopic studies of aging cerebral blood vessels have demonstrated a proportional loss of nerve axons with age,106 which implies a reduction of functional nerve supply to the vascular smooth muscle. Apart from the major pelvic ganglia (see above), no widespread loss of sympathetic neurons has been demonstrated in other sympathetic ganglia such as the superior cervical and coeliac ganglia of the aging rat. Despite the absence of neuron loss, there is evidence of decreased sympathetic

494

Neurobiology of the Peripheral Nervous System

Young: Iris. Aged: Iris.

*

* *

Young: Middle cerebral artery

Aged: Middle cerebral artery.

*

*

*

FIGURE 22–10 Selective age-related loss of dendrites in sympathetic neurons. Intracellular injection and immunolabeling demonstrate loss of length and complexity in dendritic arborizations of sympathetic neurons projecting to cerebral blood vessels, but not in neurons projecting to the iris, showing that sympathetic neurons vulnerable to age-related degeneration exhibit loss of both axons (see Fig. 22–9) and dendrites. (From Andrews, T. J., Thrasivoulou, C., Nesbit, W., and Cowen, T.: Target-specific differences in the dendritic morphology and neuropeptide content of neurons in the rat SCG during development and aging. J. Comp. Neurol. 368:33, 1996.)

innervation and functional control of the intestines10,11 suggestive of axon atrophy. However, because of the technical difficulties in identifying neuron loss in small groups of neurons, the possibility cannot be discounted that loss of small subpopulations of sympathetic neurons may occur in these and other peripheral ganglia during aging. The only major branch of the autonomic nervous system where large-scale neuron loss has been observed is the enteric nervous system. Losses of up to 50% of neurons have been observed in the myenteric plexus of rats (Fig. 22–11)47,189 and guinea pigs.76 These losses appear to be selective, affecting principally the cholinergic population,47 although this group is not homogeneous and comprises a large proportion of the total. The timing of age-related neuronal impairment remains poorly understood, largely because many studies

rely on comparisons between ‘young’ and ‘old’ rather than examining samples over a wider age range. However, the timing of age-related neuronal loss in the rat enteric nervous system has been examined and found to begin unexpectedly early in the lifespan of the Sprague-Dawley rat, at around 13 months, and to be complete by approximately 16 months.47 Thus neuron loss occurs long before the more obvious indicators of senescence, such as increased incidence of cancer, cataract, and changes of hair color, which are seen between 18 and 21 months in SpragueDawley rats and rather later in Wistar and other strains. Neurodegeneration and Loss of Function As we have stated, it is often difficult to establish a clear link between age-related neurodegeneration and loss of function either in the innervated end organ or in the

Aging in the Peripheral Nervous System

495

FIGURE 22–11 Age-related loss of myenteric neurons. Immunofluorescence staining with the panneuronal marker PGP 9.5 demonstrates substantial loss of neurons in 24-month-old (B) compared with 3-month-old (A) SpragueDawley rats. Note large holes (arrows, B) in aged myenteric ganglion indicating cell loss. (From Cowen, T., and Thrasivoulou, C.: Cerebrovascular nerves in old rats show reduced accumulation of 5-hydroxytryptamine and loss of nerve fibers. Brain Res. 513:237, 1990.)

neurons themselves. In a few situations, such a correlation has been attempted or has been made possible by parallel functional and morphological studies in autonomic neurons. For example, studies of human skin have shown loss of sweating responses to cholinergic agonists during aging, which correlate with observed patterns of neurodegeneration.1 Similar losses of nerves have been demonstrated in aging rat skin and sweat glands.2 In humans, loss of motility in the large intestine—something that commonly affects the elderly95—correlates with observations of agerelated neuron loss in the myenteric plexus of large90 and small55 intestines similar to those observed in the gastrointestinal tract of aging rodents.47 These examples suggest that neurodegeneration is an underlying cause of age-related impairments in neural function in animals and humans.

MECHANISMS OF AGING IN THE PNS Mechanisms Underlying Sensory Impairments during Aging Age-related loss and degeneration of cutaneous receptors and sensory nerve terminals is extensive enough to explain most of the sensory deficits observed in elderly individuals. Thus sensory impairments seem to reflect a process starting in the distal axon domain, possibly instigated by changes in nerve-target interactions22,23,31,75,82,109,206 (see Mechanisms Underlying Age-Related Impairments in Autonomic Neurons below). Incapacitation of the intrinsic machinery that maintains neuron integrity may also contribute to age-related neuronal dysfunction. Given that neuronal longevity in the main matches that of the host organism, postmitotic cells of the nervous system may be especially susceptible to the cumulative damaging effects of aging. Mechanical ‘wear and tear’ probably plays a role23,66; however, it is clearly not the only factor.75,205 Several lines of evidence suggest that aging has a more deleterious effect on myelinated primary afferents than their unmyelinated

counterparts. Neuronal atrophy, axonal lesions, and loss of peripheral nerve endings and receptor organs as well as of centrally projecting nerve terminals preferentially affect large myelinated primary afferent neurons. Although the mere difference in neuronal volume between myelinated and unmyelinated primary afferents may explain the selective vulnerability of the former, perhaps because of the greater metabolic load, it is highly likely that other factors contribute. One possible explanation for the difference in vulnerability may be sought in age-related changes in neurotrophin signaling. The development of the nervous system is characterized by the establishment of neuronal numbers, neuronal connectivity, and a differentiated phenotype through close interactions with target cells. In the adult animal, extension of developmental processes enables the nervous system to adapt to new or altered demands or experiences through plasticity, which includes differential growth of neurons and their axonal and dendritic arborizations.176 In this context, old age may be considered as the final stage of development, wherein neurons, with greater or lesser success, continue to respond to altered functional demands. An inability to maintain appropriate neuronal function in senescence may result from a disturbance in the trophic signaling between neurons and their target cell.43,211,212 This possibility is supported by evidence for a decreased capacity of peripheral target tissues such as skeletal muscle and skin23,154 to synthesize neurotrophic factors of the nerve growth factor (NGF) family of neurotrophins. In parallel, there is a lowered expression of the cognate (i.e., trk) receptors in primary sensory neurons.19,154 Neurotrophins have been shown to regulate the expression of neurofilaments,89 thereby affecting neuronal plasticity. Expression of neurofilaments, the major determinants of axon caliber, is depressed in primary sensory neurons of aged rats,133,171 which may contribute to the impaired capacity of, in particular, large primary sensory neurons to maintain their axonal arborizations in senescence. Neurotrophins also influence the capacity of neurons to withstand the

496

Neurobiology of the Peripheral Nervous System

damaging effects of free radicals166 (see Role of Changes in Autonomic Neurons below). Moreover, by virtue of its capacity to regulate expression of neuropeptides and other neurotransmitter substances, attenuated neurotrophin signaling may influence important functional aspects of primary sensory neurons.18,20,216 Changes in neurotrophin signaling may therefore influence several aspects of the aging process as it affects sensory neurons. In contrast to the decrease in neurotrophin signaling, glialderived neurotrophic factor (GDNF) signaling increases with advancing age.19,153 GDNF has been shown to protect the phenotype of nociceptive primary sensory neurons, as well as preserving the conduction velocity of unmyelinated, but not myelinated, axons.16,160 Thus increased GDNF signaling during senescence, in parallel with decreased neurotrophin signaling, may explain the preserved phenotype and the lack of cell body atrophy and loss of terminals among small unmyelinated primary afferents.22,75,212 An important unresolved issue concerns whether alterations in target tissues, nerve-target interactions, or neuronal/axonal aberrations represent the primary event underlying the changes observed in sensory neurons with advancing age. Recent studies on two sensory pathways, both relying on peripheral receptor cells, may shed some light on this issue. The primary event during aging of the Merkel cell–neurite complex seems to be failure in the maintenance of the Merkel cells and their production of neurotrophin 3 (NT3). In contrast, the Merkel cell axons, although reduced in number and showing decreased expression of trkC (the NT3 receptor), show signs of regenerative growth toward alternative targets that exhibit preserved expression of NT3.23 Thus the loss of Merkel cell–neurite complexes during aging appears to be a peripheral process, starting with the Merkel cell itself, possibly caused by mechanical wear and tear.23 Another interesting model is the hair cells of the cochlea and the sensory neurons of the spiral ganglion. Hearing impairment is common in senescence, and a reduction in both hair cells and in their innervation from the spiral ganglion has been noted. The spiral ganglion cells rely on target-derived BDNF (type II cells) and NT3 (type I cells) for their development. It was recently shown that the process causing loss or damage of hair cells might differ from the mechanism inducing damage or loss of the innervating neurons in the spiral ganglion. Thus the initial process involves damage to the hair cells and their contact with auditory neurons resulting from glutamate–N-methyl- D -aspartate (NMDA) receptor mechanisms that can be blocked by MK801. As a result of this loss of contact, spiral ganglion neurons are deprived of NT3 produced by the target hair cells and are consequently lost, although they can be rescued by treatment with NT3.12,61,70 In both of these examples, the initial process appears to take place in the target Merkel or hair cells, which affects their contacts with the

innervating neurons. As a result of the breakdown of target-neuron contact, the target becomes unable to sustain the sensory innervation, which consequently exhibits secondary degeneration.

Mechanisms Underlying Age-Related Changes in Nerve Trunks The causes of the observed morphological and electrophysiological changes reported in peripheral nerves from aging humans are still unclear. There are several possible mechanisms that could have a deleterious effect on nerve axons or their Schwann cells. These include changes in uptake of metabolites, reduction in protein synthesis by the cell body, disturbances of fast and slow axonal transport, and defects in ‘turnaround’ at the axonal end organ (where molecular cargoes are delivered by anterograde axonal transport and collected by retrograde axonal transport). Defects in turnaround create an imbalance between anterograde and retrograde flow, resulting in dying-back neuropathies such as that induced by p-bromophenylacetylurea.110 In addition, it has been suggested that age changes are the result of repeated minor episodes of trauma or ischemia.109 Ischemia-induced reductions in metabolism or in axonal transport may also contribute to dying-back neuropathy. The electrophysiological studies already discussed provide partial support for this view.31 Alternatively, axonal loss may result from neuronal death, as suggested by reports of loss of anterior horn cells in older subjects,143 although these observations have not been confirmed using unbiased counting methods. Studies of age-related changes in axonal transport have produced contradictory results. For example, axoplasmic transport of cholinesterase was reduced in sciatic nerves of old rats,148 whereas a study in aging mice failed to find a reduction in axonal transport rates despite demonstrating a reduction in Na⫹K⫹-ATPase in sciatic nerve and dorsal roots.184 Because ATPases are transported by fast anterograde transport from the neuron, the reduction found by these authors could indicate a decrease in protein synthesis in the neuron, which might also contribute to the reported slowing of nerve conduction velocity.184 Confusingly, other studies on axons closer to the cell bodies in the DRG suggested that a reduced rate of slow axonal transport may result in an increase in axon diameter and, therefore, an increase in conduction velocity.128,149 Failure of the turnaround system could contribute to atrophy of axon terminals or, alternatively, to an abnormal accumulation of material. Neurons with extended axon arborizations are likely to be most sensitive to age-related impairments of axonal transport and/or of the turnaround system, perhaps explaining the greater vulnerability of neurons of this character. A reduction in medium-molecularweight neurofilaments (NF-M) compared with the light

Aging in the Peripheral Nervous System

form (NF-L) has been demonstrated.209 This may reduce the neurofilament packing density, and thus the diameter of the axon. This is supported by morphometric studies showing that age-related reductions in slow axoplasmic transport are associated with reduced spacing rather than numbers of cytoskeletal elements.33 Dystrophy and Atrophy of Axons, Demyelination, and Selective Vulnerability As discussed above (see Structural Impairments in Axons), large myelinated primary afferents are selectively vulnerable to axon dystrophy. Large myelinated primary afferents use glutamate as the fast neurotransmitter. Studies in the CNS have shown a high incidence of axon dystrophy in glutamatergic neurons, and that the induction of axon dystrophy may relate to oxidative stress in excitatory (glutamatergic) pathways (see Ramirez-Leon et al.179 and references therein). Glutamate signaling can initiate oxidative challenge through stimulation of NMDA receptors, which are also expressed presynaptically141,150; thus glutamate may be detrimental to the glutamatergic terminals themselves.147 Further evidence for this autolytic hypothesis comes from observations that the total content of glutathione, a major scavenger of reactive oxygen species (ROS) expressed by neurons, is increased in dystrophic axons.179 Moreover, vitamin E deficiency has been shown to accelerate the emergence and extent of dystrophic axon lesions in sensory but not autonomic pathways195,201 (see, however, Johnson et al.112). The major contribution to ROS in neurons derives from cell respiration. Thus the metabolic demand on neurons with large and/or extensive axon arbors, such as myelinated sensory fibers to the distal hind limb, may increase the risk for oxidative damage compared with smaller neurons with a less extensive axon arbor. It cannot be excluded that glutamate and ROS act in concert to make some neurons vulnerable to axon dystrophy. Another, not mutually exclusive, mechanism (discussed in the sections on Mechanisms Underlying Sensory Impairments during Aging above and on Mechanisms Underlying Age-Related Impairments in Autonomic Neurons below) is that peripherally projecting neurons may face reduced access to target-derived neurotrophic factors in senescence (for reviews, see Ulfhake et al.211,212 and references below). This is highly relevant in the context of oxidative stress because the available evidence indicates that neurotrophic factors increase neuronal resistance to ROS challenge (see Role of Changes in Autonomic Neurons below). Moreover, deprivation of target-derived neurotrophic factors may decrease the capacity of dependent neurons to respond to ongoing adaptive demands, which include maintaining contact with the target. It is conceivable that axon dystrophy as well as axon atrophy may be induced by this mechanism. This notion is supported by observations that chronic axotomy,

497

a situation in which the axon faces permanent disruption of contact with its target, induces axon atrophy.69 Axon dystrophy and atrophy appear to be primarily cellautonomously regulated, as suggested by the evidence that early signs of axon dystrophy seen in the dorsal column system seem to be confined to the terminals.73 However, myelinated axons are also dependent on Schwann cells for their integrity. Thus axon dystrophy, as well as atrophy, is associated with demyelination.73,111 Studies on multiple sclerosis and experimentally induced encephalomyelitis have shown that axon dystrophy appears in the wake of demyelination, suggesting that demyelination can induce axon dystrophy.86,178,226 A possible explanation for this is that compact myelin or one of its components is necessary for neurofilament organization in the axonal cytoskeleton.27,56,227 Subsequently, demyelination may interfere with axonal transport. Despite controversy regarding the changes that take place in the different components of axonal transport with age, there is evidence suggesting that the rate of slow anterograde transport, which is responsible for the movement of cytoskeleton and enzymes of intermediary metabolism, may be decreased.108,128,149 This has implications for axon plasticity and may reduce activity at the axon terminal, resulting in neurofilament accumulation and subsequent distortion of the axon architecture. Moreover, interactions between compact myelin and the axon are thought to influence neurofilament organization through activation of kinases and phosphatases.27 Hence, it is possible that demyelination itself, by disrupting the compact myelin, may result in abnormally phosphorylated epitopes that are more resistant to ubiquitin-proteasomal degradation and that subsequently can accumulate in the distal axon. However, the evidence regarding the extent and specificity of phosphorylation in the different groups of neurofilaments in aging axons is at present contradictory. Indeed, there is evidence for increased phosphorylation of neurofilaments in senescence.91,209 Interestingly, overload or disturbance of the ubiquitin-proteasomal pathway, which is responsible for the cellular handling of damaged or toxic proteins, has been implicated in the progress of neurodegenerative as well as demyelinating diseases.37,86 Even if circumstantial evidence implies that axon changes are the primary event during aging, altered Schwann cell function, including the production and properties of myelin, cannot be excluded as a potential influence. ROS challenge may cause oxidative damage in the PNS in a manner similar to that affecting the white matter of the CNS during aging,198 in which oxidative damage induces lipid peroxidation and altered membrane properties13 as well as generation of cytotoxic compounds such as 4-hydroxynonenal.77 Another possible cause of myelin sheath disruption during aging is suggested by evidence for an age-related increase in the length of very-long-chain fatty acids (VLCFA) as well as an increased proportion of unsaturated

498

Neurobiology of the Peripheral Nervous System

compared with saturated VLCFA.87 This pattern resembles the situation in adrenoleukodystrophy, an X-linked inherited demyelinating disease in which VLCFA accumulate as a result of disturbed transport into the peroxisomes and impaired ␤-oxidation of VLCFAs. The accumulation of VLCFA is thought to destabilize the myelin and cause an activation of macrophages.62 Recovery after Injury Failure of neuronal protein synthesis may underlie agerelated alterations in motor nerve regeneration. In young animals, the latent period before reinnervation of the nerve stump after transection is only approximately 1 day. However, in aged rats, this period may extend to 8 days. Once this phase is completed, there is little difference between young and old animals in rate of growth of the regenerating axons. Autoradiographic labeling showed that 3 days posttransection, proteosynthesis in the motor neurons increased considerably in young rats, whereas the rate did not increase in 28-month-old rats, perhaps explaining the extended latent period in older animals.93 Regeneration in rat sciatic nerves sectioned at 15 months of age also differed from that in nerves sectioned at 3 months, and the retrograde changes were more marked. Changes in growth factor signaling in older animals may underlie the observed age differences in axonal atrophy and neuronal vacuolation.117 Although investigated less extensively than myelinated axons, regeneration of unmyelinated axons in older animals seems to be slower and less complete than in younger animals.163,215

Mechanisms Underlying Age-Related Impairments in Autonomic Neurons The underlying causes of nonpathological changes in the aging autonomic nervous system remain obscure. Unanswered questions include problems similar to those addressed above in relation to sensory neurons; for example, are neuronal changes secondary to changes in the innervated target tissue? Also, a related question, are neuronal changes secondary adaptations to altered functional demand in the aging organism, or are neuronal changes partly the result of intrinsic changes in the capacity of neurons to cope with the stresses of aging? Neuron-Target Interactions in the Aging Autonomic Nervous System As already stated, neurotrophic theory53,176 provides a model with powerful credentials to help in understanding neuronal aging. Neurotrophic theory states that neurons depend for their survival, and perhaps also for their growth and regulation of other aspects of phenotype, on neurotrophic factors derived from their target tissues. In the autonomic nervous system, ‘target tissues’ may include

peripheral effectors such as muscle and glands, or other neurons. The archetypal neurotrophic factor is NGF.142 However, in recent years a family of relatives of NGF, the neurotrophins, have been discovered with effects in many different regions of the CNS and PNS.15 In addition, a large number of cytokines and other growth factors have been identified with neurotrophic activities on central as well as peripheral neurons (see, e.g., Airaksinen and Saarma5). Based on the view that neurons require trophic interactions with their target tissues, the hypothesis has been advanced that availability of target-derived neurotrophic factors and/or expression of the appropriate receptors may continue to determine neuronal growth, survival, and phenotype during adult life and, perhaps, on into old age.82 In order to test this hypothesis, transplants of target tissues exhibiting age-related changes in the density of innervating sympathetic nerve fibers were taken from 6- or 24-month-old donor rats and placed onto the iris in the anterior eye chamber of young host animals. In this situation, the sympathetic and other nerves innervating the host iris extend over the implanted target tissue, allowing the interaction between host nerves and implanted target tissues of different ages to be studied. Donor blood vessels,206 sweat glands,49 and other tissues became organotypically reinnervated by irideal nerves from the host, testifying to the capacity of the target tissues to organize ingrowing nerves into an appropriate pattern. Cerebral blood vessels transplanted in this way from aged donors exhibited a 50% lower density of reinnervation by young host nerves compared with young transplanted vessels. The pattern as well as the density of reinnervation closely resembled that seen in vivo. It was concluded that target tissues influence both pattern and density of innervation during adult life, including the reductions of nerve density seen in old age. Examination of the responses of host and transplanted neurons revealed that age changes in neuronal plasticity also contribute to the outcome of trophic interactions between nerve and target tissue. Old and young sympathetic ganglia transplanted in oculo in young hosts extended new innervation over the denervated host iris with equal success, indicating that aged sympathetic neurons were unimpaired in their capacity to respond to the axotomy caused by transplanting the ganglia.81 However, examination of the responses of the iris-projecting sympathetic neurons of old hosts to implanted target tissues of different ages revealed that aged host nerves were impaired in their capacity to reinnervate all target tissues, irrespective of age,79 suggesting that intrinsic changes in neurons, as well as alterations in targets, contribute to age-related changes in neuronal plasticity. Treatment of the transplants with NGF failed to counteract the age-related loss of plasticity of aging sympathetic neurons in the heterochronic transplantation model. These apparently contradictory results are likely to reflect the differential effects of age on

499

Aging in the Peripheral Nervous System

NGF mRNA (RT-PCR)

fg/pg GAP DH

30

8–12wk 24m

20 10 0 CV

B

Iris

NGF protein (ELISA) 6

fpg/mg wet wt.

Role of Changes in Target Tissues Reduced expression of neurotrophic factors has been shown in some regions of the brain228 and spinal cord.154 In addition, there is evidence from sensory systems of peripheral changes in neurotrophic factors (see Mechanisms Underlying Sensory Impairments during Aging above). Other studies in the CNS have shown no changes,52 leading to the view that there is no clear association between altered expression of neurotrophic factors and vulnerability to age-related neurodegeneration. Models using sympathetic neurons and their targets, which often consist of regionally distinct tissues, therefore provide an advantage in analyzing and distinguishing the contributions of neurons versus targets to neuronal aging processes. Levels of expression of neurotrophic factors have been examined in target tissues that are associated with sympathetic neurons that are vulnerable to, or protected from, age-related atrophy or degeneration.44,132 These studies failed to show any age-related reduction in neurotrophic factors in those tissues, such as the rat pineal and cerebral blood vessels, that are innervated by neurons exhibiting atrophy during aging. There is also no change in expression of neurotrophic factors in tissues such as the rat iris, which is innervated by neurons that do not exhibit neurodegeneration. However, it has recently been shown that target tissues innervated by neurons vulnerable to age-related atrophy exhibit much lower levels of expression of neurotrophic factors throughout life compared with target tissues supplied by neurons that are not so vulnerable (Fig. 22–12).50 Thus the lifelong level of availability of target-derived neurotrophic factors to the innervating neurons may enhance the capacity of neurons to cope with age-related stressors. This model suggests that capacity to cope with the stresses of aging results partly from the developmental program which leads to functional differentiation of neuronal phenotype. For a fuller discussion of this model, the reader is referred to Cowen et al.50 Further evidence that neurotrophic factors remain important in the aging nervous system comes from experiments involving treatment of different groups of neurons with neurotrophic factors. Sympathetic nerves supplying cerebral blood vessels have been shown to sprout new fibers in response to treatment with NGF.7,106 This growth response is retained in old age6,106 when high doses of NGF are used. However, growth responses to NGF have been shown to be dose dependent,107 and one study has demonstrated that sympathetic nerves in aging rats exhibit

A

8–12wk 24m

4 2 0 CV

C

Iris

NT3mRNA (RT-PCR) 50

tg/pg GAP DH

regeneration of sympathetic neurons, which is unaffected by age, versus the capacity for collateral sprouting, which is impaired with age.43,60 In light of this evidence that changes in targets as well as in neurons contribute to age-related neurodegeneration, it is appropriate to enquire as to the nature of these changes and whether they occur independently of each other or as part of an ‘aging program’ affecting neurons.

8–12wk 24m

40 30 20 10 0 CV

Iris

FIGURE 22–12 Age changes in expression of neurotrophic factors. Reverse transcriptase–polymerase chain reaction for mRNA (A and C) and enzyme-linked immunosorbent assay of protein (B) for NGF (A and B) and for NT3 (C) demonstrate lack of age-related changes of neurotrophin expression in cerebral blood vessels (CV) or iris from 6-month-old and 24-month-old Sprague-Dawley rats. However, neurotrophin expression (mRNA and protein) is markedly higher in iris, where innervating neurons are unaffected by age, compared with CV, where neurons exhibit vulnerability to age-related atrophy of axons (see Fig. 22–9) and dendrites (see Fig. 22–10). This could indicate a protective effect of lifelong exposure to high levels of neurotrophin expression. (From Cowen, T., Woodhoo, A., Sullivan, C. D., et al.: Reduced age-related plasticity of neurotrophin receptor expression in sympathetic neurones of the rat. Aging Cell 2:59, 2003.)

decreased growth responsiveness to NGF compared with neurons from young animals.83 The extracellular matrix (ECM) may make a significant contribution to the progress of neurodegeneration during aging. Laminin is a key component of the ECM with wellknown nerve growth and survival-promoting capabilities. In vitro studies of sympathetic neurons show that laminin and NGF act synergistically to promote neurite outgrowth in mature as well as in aged neurons.46 Examination of the

500

Neurobiology of the Peripheral Nervous System

levels and distribution of laminin in cerebral blood vessels from aged rats shows that high levels of laminin are found in the basal lamina interposed between nerve and the target muscle cell at all ages.80 Laminin levels are reduced in the aged blood vessel wall, making the target tissue less attractive for the maintenance of nerve fibers in that area. Assuming that autonomic nerves undergo continual remodeling throughout life (see Selective Vulnerability of Autonomic Neurons to Age-Related Neurodegeneration above), laminin is likely to be a key player in determining neuronal vulnerability to aging, perhaps through its interactions with neurotrophin signaling. Role of Changes in Autonomic Neurons There is strong evidence that neuronal responsiveness to neurotrophic factors alters during aging and that this is likely to contribute in some way to age-related neurodegeneration. Evidence includes reduced uptake or retrograde axonal transport of NGF in aged rat sympathetic nerves,44 including those projecting to cerebral blood vessels.134 Similar observations have been made in NGFresponsive neurons of the aged rat basal forebrain.40 A decrease in neurofilament light (NF-L) gene expression in the superior cervical ganglion of aging rats could be related to altered axonal transport, as well as to axonal hypotrophy or atrophy.133 Furthermore, as stated above, there is evidence of reduced growth responsiveness to NGF in aging rat sympathetic neurons. Responsiveness to the neurotrophin family of neurotrophic factors is mediated by the trk family of receptor tyrosine kinases, each with specific binding for particular neurotrophins, and by the pan-specific p75 receptor (a member of the tumor necrosis factor receptor family), which binds all neurotrophins.114 The role of p75 includes a role in apoptotic neuronal cell death when expressed alone.151 However, when it is coexpressed with trkA, as in the majority of sympathetic neurons, p75 is generally considered to enhance responsiveness to the neurotrophins.99 Reduced levels of expression of the p75 NGF receptor and the inability to upregulate receptor expression in response to NGF characterize those sympathetic neurons vulnerable to age-related neurodegeneration.50 During development and early postnatal life, neurotrophic factors regulate survival as well as growth of sympathetic, enteric, and other neuron types. These responses are reinforced by an upregulation of expression of the appropriate receptors, notably p75 in the case of sympathetic neurons,152 and this is considered to provide a mechanism by which growing target tissues can influence their innervation. During maturation, however, changes occur in neuronal responsiveness to neurotrophic factors. Sympathetic and NGF-responsive sensory neurons become relatively independent of NGF for their survival57,170,218 while retaining their responsiveness for growth, a change of obvious significance for the functional

preservation of mature neurons. Intrinsic downregulation of neurotrophin receptors and impaired regulation of receptor expresssion50 are therefore the likely mechanisms underlying reduced growth responsiveness of aging autonomic neurons (see above). Neurotrophin signaling may impact on neuronal aging processes in at least one further way. ROS play an important part in aging processes in general, having well-established damaging effects on the structure and function of many cells and organ systems, as previously discussed (see Mechanisms Underlying Sensory Impairments during Aging above).13 Neurons are particularly vulnerable to ROS-induced damage. A novel role for neurotrophic factors in antioxidant defense in neurons, including in sympathetic and enteric neurons of the autonomic nervous system, has recently been discovered.63 In enteric neurons, ROS levels increase with age. Dietary restriction, which increases longevity in rodents and many other species,200 also protects neurons against the age-related increase in ROS levels, apparently by enhancing the antioxidant role of neurotrophic factors. It therefore seems likely that intrinsic changes in neuronal gene expression, including the capacity to express neurotrophin receptors, are an essential part of an antioxidant defense mechanism, which may be adversely influenced by ongoing, age-related increases in intracellular ROS.

CONCLUSIONS Old age exerts highly selective effects on peripheral as on central areas of the nervous system. In the periphery, large myelinated sensory neurons, their axons, and their central connections appear to be particularly vulnerable during aging. The effects of age on the structure of these particular neurons correlates with some of the principal signs of aging in the elderly, which include a range of functional deficits affecting, notably, mechanoreceptive systems. We present evidence that the sensory receptors that these neurons innervate also degenerate with age and that the distal axons appear to be affected before the more proximal parts of the neuron. Based on this evidence, the view is advanced that interactions between sensory neurons and their receptor targets are crucial initiators of an agerelated nervous deterioration, which extends centripetally from the periphery to affect whole neuronal circuits, including central connections in the spinal cord. Defects in neurotrophin signaling between the neuron and its target, associated with impaired axonal transport and disturbed myelination, may provide the mechanism underlying functional and structural impairments of peripheral neurons. Other groups of sensory and motor neurons are affected by age, but in a less obvious way than in the large myelinated axons of sensory neurons. In autonomic neurons, defects in neurotrophin signaling also contribute significantly to the selective impairments that

Aging in the Peripheral Nervous System

affect these neurons in old age. Lifelong exposure to high or low levels of neurotrophic factors respectively provides protection or renders neurons vulnerable. As in sensory neurons, failure of axon transport probably contributes to vulnerability in aging. However, in other respects, autonomic neurons are different from sensory and somatic motor neurons. For example, larger autonomic neurons seem to cope with the stresses of aging better than their smaller relatives, unlike sensory neurons, in which the reverse is the case. Important clues are surfacing that neurotrophic factors, in addition to their well-established role in relation to neuronal growth and survival, help to defend autonomic neurons against free radical damage. ROS-induced damage is also implicated in age-related sensory deficits, as well as in damage to neurons of the CNS. In both areas, glutamatergic neurotransmission appears to confer particular vulnerability to ROS. The strengthening connection between aging and ongoing free radical damage therefore links agerelated neurodegeneration in central, peripheral, and autonomic nervous systems.

REFERENCES 1. Abdel-Rahman, T. A., Collins, K. J., Cowen, T., and Rustin, M.: Immunohistochemical, morphological and functional changes in the peripheral sudomotor neuro-effector system in elderly people. J. Auton. Nerv. Syst. 37:187, 1992. 2. Abdel-Rahman, T. A., and Cowen, T.: Neurodegeneration in sweat glands and skin of aged rats. J. Auton. Nerv. Syst. 46:55, 1993. 3. Adinolfi, A., Yamuy, J., Morales, F. R., and Chase, M. H.: Segmental demyelination in peripheral nerves of old cats. Neurobiol. Aging 12:173, 1991. 4. Ahmad, A., and Spear, P. D.: Effects of aging on the size, density, and number of rhesus monkey lateral geniculate neurons. J. Comp. Neurol. 334:631, 1993. 5. Airaksinen, M. S., and Saarma, M.: The GDNF family: signalling, biological functions and therapeutic value. Nat. Rev. Neurosci. 3:383, 2002. 6. Andrews, T. J., and Cowen, T.: In vivo infusion of NGF induces the organotypic regrowth of perivascular nerves following their atrophy in aged rats. J. Neurosci. 14:3048, 1994. 7. Andrews, T. J., and Cowen, T.: Nerve growth factor enhances the dendritic arborization of sympathetic ganglion cells undergoing atrophy in aged rats. J. Neurocytol. 23:234, 1994. 8. Andrews, T. J., Thrasivoulou, C., Nesbit, W., and Cowen, T.: Target-specific differences in the dendritic morphology and neuropeptide content of neurons in the rat SCG during development and aging. J. Comp. Neurol. 368:33, 1996. 9. Ansved, T., and Larsson, L.: Quantitative and qualitative morphological properties of the soleus motor nerve and the L5 ventral root in young and old rats. J. Neurol. Sci. 96:269, 1990. 10. Baker, D. M., and Santer, R. M.: A quantitative study of the effects of age on the noradrenergic innervation of Auerbach’s plexus in the rat. Mech. Ageing Dev. 42:147, 1988.

501

11. Baker, D. M., Watson, S. P., and Santer, R. M.: Evidence for a decrease in sympathetic control of intestinal function in the aged rat. Neurobiol. Aging 12:363, 1991. 12. Basile, A. S., Huang, J. M., Xie, C., et al.: N-methyl-Daspartate antagonists limit aminoglycoside antibioticinduced hearing loss. Nat. Med. 2:1338, 1996. 13. Beckman, K. B., and Ames, B. N.: The free radical theory of aging matures. Physiol. Rev. 78:547, 1998. 14. Behse, F., Buchthal, F., Carlsen, F., and Knappeis, G. G.: Unmyelinated fibers and Schwann cells of sural nerve in neuropathy. Brain 98:493, 1975. 15. Bennet, M. R., Gibson, W. G., and Lemon, G.: Neuronal cell death, nerve growth factor and neurotrophic models: 50 years on. Auton. Neurosci. 95:1, 2002. 16. Bennett, D. L., Michael, G. J., Ramachandran, N., et al.: A distinct subgroup of small DRG cells express GDNF receptor components and GDNF is protective for these neurons after nerve injury. J. Neurosci. 18:3059, 1998. 17. Berg, B. N., Wolf, A., and Simms, H. S.: Degenerative lesions of spinal roots and peripheral nerves in aging rats. Gerontologia 6:72, 1962. 18. Bergman, E., Carlsson, K., Liljeborg, A., et al.: Neuropeptides, nitric oxide synthase and GAP-43 in B4binding and RT97 immunoreactive primary sensory neurons: normal distribution pattern and changes after peripheral nerve transection and aging. Brain Res. 832:63, 1999. 19. Bergman, E., Fundin, B. T., and Ulfhake, B.: Effects of aging and axotomy on the expression of neurotrophin receptors in primary sensory neurons. J. Comp. Neurol. 410:368, 1999. 20. Bergman, E., Johnson, H., Zhang, X., et al.: Neuropeptides and neurotrophin receptor mRNAs in primary sensory neurons of aged rats. J. Comp. Neurol. 375:303, 1996. 21. Bergman, E., and Ulfhake, B.: Loss of primary sensory neurons in the very old rat: neuron number estimates using the disector method and confocal optical sectioning. J. Comp. Neurol. 396:211, 1998. 22. Bergman, E., and Ulfhake, B.: Evidence for loss of myelinated input to the spinal cord in senescent rats. Neurobiol. Aging 23:271, 2002. 23. Bergman, E., Ulfhake, B., and Fundin, B. T.: Regulation of NGF-family ligands and receptors in adulthood and senescence: correlation to degenerative and regenerative changes in cutaneous innervation. Eur. J. Neurosci. 12:2694, 2000. 24. Bischoff, A.: Die Ultrastruktur peripherer Nerven bei der diabetischen Neuropathie. Dtsch. Ges. Inn. Med. 18:1138, 1967. 25. Bleys, R. L., Cowen, T., Groen, G. J., and Hillen, B.: Perivascular nerves of the human basal cerebral arteries: II. Changes in aging and Alzheimer’s disease. J. Cereb. Blood Flow Metab. 16:1048, 1996. 26. Bolton, C. F., Winkelmann, R. K., and Dyck, P. J.: A quantitative study of Meissner’s corpuscles in man. Neurology 16:1, 1966. 27. Brady, S. T., Witt, A. S., Kirkpatrick, L. L., et al.: Formation of compact myelin is required for maturation of the axonal cytoskeleton. J. Neurosci. 19:7278, 1999. 28. Braund, K. G., McGuire, K. A., and Lincoln, C. E.: Agerelated changes in the peripheral nerves of the dog. I. A morphologic and morphometric study of single-teased fibers. Vet. Pathol. 19:365, 1982.

502

Neurobiology of the Peripheral Nervous System

29. Braund, K. G., McGuire, K. A., and Lincoln, C. E.: Agerelated changes in the peripheral nerves of the dog. II. A morphologic and morphometric study of cross-sectioned nerve. Vet. Pathol. 19:398, 1982. 30. Brocklehurst, J. C., Robertson, D., and James-Groom, P.: Clinical correlates of sway in old age-sensory modalities. Age Aging 11:1, 1982. 31. Buchthal, F., Rosenfalck, A., and Behse, F.: Sensory potentials of normal and diseased human nerves. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 442, 1975. 32. Burek, J. D., Van der Kogel, A. J., and Hollander, C. F.: Degenerative myelopathy in three strains of aging rats. Vet. Pathol. 13:321, 1976. 33. Caselli, U., Bertoni-Freddari, C., Paoloni, R., et al.: Morphometry of axon cytoskeleton at internodal regions of rat sciatic nerve during aging. Gerontologist 45:307, 1999. 34. Cauna, N.: The effects of aging on the receptor organs of the human dermis. In Montagna, W. (ed.): Advances in the Biology of the Skin. New York, Pergamon Press, p. 63, 1965. 35. Ceballos, D., Cuadras, J., Verdu, E., and Navarro, X.: Morphometric and ultrastructural changes with ageing in mouse peripheral nerve. J. Anat. 195:563, 1999. 36. Cerimele, D., Celleno, L., and Serri, F.: Physiological changes in ageing skin. Br. J. Dermatol. 122:13, 1990. 37. Chung, K. K., Dawson, V. L., and Dawson, T. M.: The role of the ubiquitin-proteasomal pathway in Parkinson’s disease and other neurodegenerative disorders. Trends Neurosci. 24:S7, 2001. 38. Coggeshall, R. E.: A consideration of neural counting methods. Trends Neurosci. 15:9, 1992. 39. Coleman, P. D., and Flood, D. G.: Neuron numbers and dendritic extent in normal aging and Alzheimer’s disease. Neurobiol. Aging 8:521, 1987. 40. Cooper, J. D., Lindholm, D., and Sofroniew, M. V.: Reduced transport of [125I]nerve growth factor by cholinergic neurons and down-regulated TrkA expression in the medial septum of aged rats. Neuroscience 62:625, 1994. 41. Cotard-Bartley, M. P., Secchi, J., Glomot, R., and Cavanagh, J. B.: Spontaneous degenerative lesions of peripheral nerves in aging rats. Vet. Pathol. 18:110, 1981. 42. Cowen, T.: Selective vulnerability in adult and aging mammalian neurons. Auton. Neurosci. 96:20, 2002. 43. Cowen, T., and Gavazzi, I.: Plasticity in adult and aging sympathetic neurons. Prog. Neurobiol. 54:249, 1998. 44. Cowen, T., Gavazzi, I., Weingartner, J., and Crutcher, K. A.: Levels of NGF protein do not correlate with changes in innervation of the rat iris in old age. Neuroreport 7:2216, 1996. 45. Cowen, T., Haven, A. J., Wen Qin, C., et al.: Development and ageing of perivascular adrenergic nerves in the rabbit: a quantitative fluorescence histochemical study using image analysis. J. Auton. Nerv. Syst. 5:317, 1982. 46. Cowen, T., Jenner, C., Song, G. X., et al.: Responses of mature and aged sympathetic neurons to laminin and NGF: an in vitro study. Neurochem. Res. 22:1003, 1997. 47. Cowen, T., Johnson, R. J. R., Soubeyre, V., and Santer, R. M.: Restricted diet rescues rat enteric motor neurones from age related cell death. Gut 47:653, 2000.

48. Cowen, T., and Thrasivoulou, C.: Cerebrovascular nerves in old rats show reduced accumulation of 5-hydroxytryptamine and loss of nerve fibers. Brain Res. 513:237, 1990. 49. Cowen, T., Thrasivoulou, C., Shaw, S. A., and AbdelRahman, T. A.: Transplanted sweat glands from mature and aged donors determine cholinergic phenotype and altered density of host sympathetic nerves. J. Auton. Nerv. Syst. 60:215, 1996. 50. Cowen, T., Woodhoo, A., Sullivan, C. D., et al.: Reduced age-related plasticity of neurotrophin receptor expression in sympathetic neurones of the rat. Aging Cell 2:59, 2003. 51. Critchley, M.: The neurology of old age. Lancet 1:1221, 1931. 52. Crutcher, K. A., and Weingartner, J.: Hippocampal NGF levels are not reduced in the aged Fischer-344 rat. Neurobiol. Aging 12:449, 1991. 53. Davies, A. M.: The neurotrophic hypothesis: where does it stand? Philos. Trans. R. Soc. Lond. B 351:389, 1996. 54. de Neeling, J. N., Beks, P. J., Bertelsmann, F. W., et al.: Sensory thresholds in older adults: reproducibility and reference values. Muscle Nerve 17:454, 1994. 55. de Souza, R. R., Moratelli, H. B., Borges, N., and Liberti, E. A.: Age-induced nerve cell loss in the myenteric plexus of the small intestine in man. Gerontology 39:183, 1993. 56. De Waegh, S. M., Lee, V. M., and Brady, S. T.: Local modulation of neurofilament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68:451, 1992. 57. Deckwerth, T. L., and Johnson, E. M.: Temporal analysis of events associated with programmed cell death (apoptosis) of sympathetic neurons deprived of nerve growth factor. J. Cell Biol. 123:1207, 1993. 58. Dering, M. A., Santer, R. M., and Watson, A. H.: Agerelated changes in the morphology of preganglionic neurons projecting to the rat hypogastric ganglion. J. Neurocytol. 25:555, 1996. 59. Dhall, U., Cowen, T., Haven, A. J., and Burnstock, G.: Perivascular noradrenergic and peptide-containing nerves show different patterns of change during development and ageing in the guinea-pig. J. Auton. Nerv. Syst. 16:109, 1986. 60. Diamond, J., Foerster, A., Holmes, M., and Coughlin, M.: Sensory nerves in adult rats regenerate and restore sensory function to the skin independently of endogenous NGF. J. Neurosci. 12:1467, 1992. 61. Duan, M., Agerman, K., Ernfors, P., and Canlon, B.: Complementary roles of neurotrophin 3 and a N-methylD-aspartate antagonist in the protection of noise and aminoglycoside-induced ototoxicity. Proc. Natl. Acad. Sci. U. S. A. 97:7597, 2000. 62. Dubois-Dalcq, M., Feigenbaum, V., and Aubourg, P.: The neurobiology of X-linked adrenoleukodystrophy, a demyelinating peroxisomal disorder. Trends Neurosci. 22:4, 1999. 63. Dugan, L. L., Creedon, D. J., Johnson, E. M., and Holtzman, D. M.: Rapid suppression of free radical formation by nerve growth factor involves the mitogen-activated protein kinase pathway. Proc. Natl. Acad. Sci. U. S. A. 94:4086, 1997. 64. Duncan, I. D.: Age-related neuropathy in animals. In Thomas, P. K. (ed.): Peripheral Nerve Changes in the Elderly. Chichester, UK, John Wiley, p. 213, 1988. 65. Dunn, E. H.: The influence of age, sex, weight and relationship upon the number of medullated nerve fibers and

Aging in the Peripheral Nervous System

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76. 77.

78. 79.

80.

81.

82.

83.

on the size of the largest fibers in the ventral root of the second cervical nerve of the albino rat. J. Comp. Neurol. 22:131, 1912. Dyck, P. J., Classen, S. M., Stevens, J. C., and O’Brien, P. C.: Assessment of nerve damage in the feet of long-distance runners. Mayo Clin. Proc. 62:568, 1987. Dyck, P. J., Johnson, W. J., Lambert, E. H., and O’Brien, P. C.: Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin. Proc. 46:400, 1971. Dyck, P. J., Karnes, J., O’Brien, P. C., and Zimmerman, I.: Detection thresholds of cutaneous sensations in humans. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 1103, 1984. Dyck, P. J., Lais, A. C., Karnes, J. L., et al.: Permanent axotomy, a model of axonal atrophy and secondary segmental demyelination and remyelination. Ann. Neurol. 9:575, 1981. Ernfors, P., Duan, M. L., ElShamy, W. M., and Canlon, B.: Protection of auditory neurons from aminoglycoside toxicity by neurotrophin-3. Nat. Med. 2:463, 1996. Ferrell, W. R., Crighton, A., and Sturrock, R. D.: Agedependent changes in position sense in human proximal interphalangeal joints. Neuroreport 3:259, 1992. Fujisawa, K.: Some observations on the skeletal musculature of aged rats. III. Abnormalities of terminal axons found in motor end-plates. Exp. Gerontol. 11:43, 1976. Fujisawa, K., and Shiraki, H.: Study of axonal dystrophy. II. Dystrophy and atrophy of the presynaptic boutons: a dual pathology. Neuropathol. Appl. Neurobiol. 6:387, 1980. Fullerton, P. M., and Gilliatt, R. W.: Pressure neuropathy in the hind foot of the guinea pig. J. Neurol. Neurosurg. Psychiatry 30:18, 1967. Fundin, B. T., Bergman, E., and Ulfhake, B.: Alterations in mystacial pad innervation in the aged rat. Exp. Brain Res. 117:324, 1997. Gabella, G.: Fall in the number of myenteric neurons in aging guinea pigs. Gastroenterology 96:1487, 1989. Gard, A. L., Solodushko, V. G., Waeg, G., and Majic, T.: 4-Hydroxynonenal, a lipid peroxidation byproduct of spinal cord injury, is cytotoxic for oligodendrocyte progenitors and inhibits their responsiveness to PDGF. Microsc. Res. Tech. 52:709, 2001. Gardener, E.: Decrease in human neurones with age. Anat. Rec. 77:529, 1940. Gavazzi, I.: Collateral sprouting and responsiveness to nerve growth factor of ageing neurons. Neurosci. Lett. 189:47, 1995. Gavazzi, I., Boyle, K. S., Edgar, D., and Cowen, T.: Reduced laminin immunoreactivity in the blood vessel wall of aging rats correlates with reduced innervation in vivo and following transplantation. Cell Tissue Res. 281:23, 1995. Gavazzi, I., and Cowen, T.: Axonal regeneration from transplanted sympathetic ganglia is not impaired by age. Exp. Neurol. 122:57, 1993. Gavazzi, I., and Cowen, T.: Can the neurotrophic hypothesis explain degeneration and loss of plasticity in mature and ageing autonomic nerves? J Auton. Nerv. Syst. 58:1, 1996. Gavazzi, I., Railton, K. L., Ong, E., and Cowen, T.: Responsiveness of sympathetic and sensory irideal nerves

84.

85. 86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

503

to NGF treatment in young and aged rats. Neurobiol. Aging 22:287, 2001. Gescheider, G. A., Beiles, E. J., Checkosky, C. M., et al.: The effects of aging on information-processing channels in the sense of touch: II. Temporal summation in the P channel. Somatosens. Mot. Res. 11:359, 1994. Gilmore, S. A.: Spinal nerve root degeneration in aging laboratory rats: a light microscopic study. Anat. Rec. 174:251, 1972. Giordana, M. T., Richiardi, P., Trevisan, E., et al.: Abnormal ubiquitination of axons in normally myelinated white matter in multiple sclerosis brain. Neuropathol. Appl. Neurobiol. 28:35, 2002. Giusto, N. M., Roque, M. E., and Ilincheta de Boschero, M. G.: Effects of aging on the content, composition and synthesis of sphingomyelin in the central nervous system. Lipids 27:835, 1992. Goff, G. D., Rosner, B. S., Detre, T., and Kennard, D.: Vibration perception in normal man and medical patients. J. Neurol. Neurosurg. Psychiatry 28:503, 1965. Gold, B. G., Mobley, W. C., and Matheson, S. F.: Regulation of axonal caliber, neurofilament content, and nuclear localization in mature sensory neurons by nerve growth factor. J. Neurosci. 11:943, 1991. Gomes, O. A., de Souza, R. R., and Liberti, E. A.: A preliminary investigation of the effects of aging on the nerve cell number in the myenteric ganglia of the human colon. Gerontology 43:210, 1997. Gou, J. P., Eyer, J., and Leterrier, J. F.: Progressive hyperphosphorylation of neurofilament heavy subunits with aging: possible involvement in the mechanism of neurofilament accumulation. Biochem. Biophys. Res. Commun. 215:368, 1995. Grover-Johnson, N., and Spencer, P. S.: Peripheral nerve abnormalities in aging rats. J. Neuropathol. Exp. Neurol. 40:155, 1981. Gutmann, E., Jakoubek, B., Hájek, I., et al.: Effect of age on proteosynthesis in spinal motoneurons following nerve interruption as shown by autoradiography of S35 labelled methionine. Physiol. Bohemoslov. 11:437, 1962. Haegerstrand, A., Dalsgaard, C. J., Jonson, B., et al.: Calcitonin gene-related peptide stimulates proliferation of human endothelial cells. Proc. Natl. Acad. Sci. U. S. A. 87:3299, 1990. Hall, K. E.: Aging and neural control of the GI tract. II. Neural control of the aging gut: can an old dog learn new tricks? Am. J. Physiol. Gastrointest. Liver Physiol. 283:G827, 2002. Harris, L. W., and Purves, D.: Rapid remodelling of sensory endings in the corneas of living mice. J. Neurosci. 9:2210, 1989. Hashizume, K., and Kanda, K.: Differential effects of aging on motoneurons and peripheral nerves innervating the hindlimb and forelimb muscles of rats. Neurosci. Res. 22:189, 1995. Helme, R. D., and McKernan, S.: Effects of age on the axon reflex response to noxious chemical stimulation. Clin. Exp. Neurol. 22:57, 1985. Hempstead, B., Martin-Zanca, D., Kaplan, D. R., et al.: High-affinity NGF binding requires coexpression of the trk proto-oncogene and the low affinity NGf receptor. Nature 350:678, 1991.

504

Neurobiology of the Peripheral Nervous System

100. Hirano, A.: Hirano bodies and related neuronal inclusions. Neuropathol. Appl. Neurobiol. 20:3, 1994. 101. Hoeffding, V., and Feldman, M. L.: Changes with age in the morphology of the cochlear nerve in rats: light microscopy. J. Comp. Neurol. 276:537, 1988. 102. Hökfelt, T.: Neuropeptides in perspective: the last ten years. Neuron 7:867, 1991. 103. Holzer, P.: Local effector functions of capsaicin-sensitive sensory nerve endings: involvement of tachykinins, calcitonin gene-related peptide and other neuropeptides. Neuroscience 24:739, 1988. 104. Howell, T. H.: Senile deterioration of the central nervous system: a clinical study. Br. Med. J. 1:56, 1949. 105. Impallomeni, M., Kenny, R. A., Flynn, M. D., et al.: The elderly and their ankle jerks. Lancet 1:670, 1984. 106. Isaacson, L. G., and Crutcher, K. A.: Uninjured aged sympathetic neurons sprout in response to exogenous NGF in vivo. Neurobiol. Aging 19:333, 1998. 107. Isaacson, L. G., Mareska, M., Nixdorf, W., and Oris, J. T.: Dose-dependent response of mature cerebrovascular axons in vivo following intracranial infusion of nerve growth factor. Neurosci. Lett. 222:21, 1997. 108. Jacob, J. M.: Fast axonal transport rates are unchanged in 6- and 24-month F344 rats. Brain Res. 699:154, 1995. 109. Jacobs, J. M., and Love, S.: Qualitative and quantitative morphology of human sural nerve at different ages. Brain 108:897, 1985. 110. Jakobsen, J., and Brimijoin, S.: Axonal transport of enzymes and labeled proteins in experimental axonopathy induced by p-bromophenylacetylurea. Brain Res. 229:103, 1981. 111. Jellinger, K., and Jirasek, A.: Neuroaxonal dystrophy in man: character and natural history. Acta Neuropathol. (Berl.) Suppl. 5:16, 1971. 112. Johnson, J. E., Mehler, W. R., and Miquel, J.: A fine structural study of degenerative changes in the dorsal column nuclei of aging mice: lack of protection by vitamin E. J. Gerontol. 30:395, 1975. 113. Kanda, T., Tsukagoshi, H., Oda, M., et al.: Morphological changes in unmyelinated nerve fibres in the sural nerve with age. Brain 114:585, 1991. 114. Kaplan, D. R., and Miller, F. D.: Neurotrophin signal transduction in the nervous system. Curr. Opin. Neurobiol. 10:381, 2000. 115. Kazui, H., and Fujisawa, K.: Radiculoneuropathy of aging rats: a quantitative study. Neuropathol. Appl. Neurobiol. 14:137, 1988. 116. Keithley, E. M., Ryan, A. F., and Feldman, M. L.: Cochlear degeneration in aged rats of four strains. Hear. Res. 59:171, 1992. 117. Kerezoudi, E., King, R. H. M., Muddle, J. R., et al.: Influence of age on the late retrograde effects of sciatic nerve section in the rat. J. Anat. 187:27, 1995. 118. Khalil, Z., and Helme, R.: Sensory peptides as neuromodulators of wound healing in aged rats. J. Gerontol. A Biol. Sci. Med. Sci. 51:354, 1996. 119. Khalil, Z., and Helme, R. D.: Sequence of events in substance P plasma extravasation in rat skin. Brain Res. 500:256, 1989. 120. Khalil, Z., Ralevic, V., Bassirat, M., et al.: Effects of ageing on sensory nerve function in rat skin. Brain Res. 641:265, 1994.

121. Kihara, M., Nakasaka, Y., Mitsui, Y., et al.: Aging differentially modifies sensitivity of nerve blood flow to vasocontractile agents (endothelin-1, noradrenaline and angiotensin II) in sciatic nerve. Mech. Ageing Dev. 114:5, 2000. 122. Kihara, M., Shioyama, M., Okuda, K., and Takahashi, M.: The impact of aging on vasa nervorum, nerve blood flow and vasopressin responsiveness. Can. J. Neurol. Sci. 29:164, 2002. 123. King, R. H. M.: Age changes in the peripheral nervous system. In Mohr, U., Dungworth, D. L., and Capen, C. C. (eds.): Pathobiology of the Aging Rat. Washington, DC, ILSI Press, p. 35, 1994. 124. King, R. H. M., Llewelyn, J. G., Thomas, P. K., et al.: Diabetic neuropathy: abnormalities of Schwann cell and perineurial basal laminae. Implications for diabetic vasculopathy. Neuropathol. Appl. Neurobiol. 15:339, 1989. 125. King, R. H. M., and Thomas, P. K.: Distal axonal degeneration in aging rats. Neuropathol. Appl. Neurobiol. 9:73, 1983. 126. Knox, C. A., Kokmen, E., and Dyck, P. J.: Morphometric alteration of rat myelinated fibers with aging. J. Neuropathol. Exp. Neurol. 48:119, 1989. 127. Kokmen, E., Bossemeyer, R. W. J., Barney, J., and Williams, W. J.: Neurological manifestations of aging. J. Gerontol. 32:411, 1977. 128. Komiya, Y.: Slowing with age of the rate of slow axon flow in bifurcating axons of rat dorsal root ganglion cells. Brain Res. 183:477, 1980. 129. Krinke, G.: Spinal radiculoneuropathy in aging rats: demyelination secondary to neuronal dwindling? Acta Neuropathol. (Berl.) 59:63, 1983. 130. Krinke, G., Froehlich, E., Herrmann, M., et al.: Adjustment of the myelin sheath to axonal atrophy in the rat spinal root by the formation of infolded myelin loops. Acta Anat. 131:182, 1988. 131. Krinke, G., Suter, J., and Hess, R.: Radicular myelinopathy in aging rats. Vet. Pathol. 8:335, 1981. 132. Kuchel, G. A., Crutcher, K. A., Naheed, U., et al.: NGF expression in the aged rat pineal gland does not correlate with loss of sympathetic axonal branches and varicosities. Neurobiol. Aging 20:685, 1999. 133. Kuchel, G. A., Poon, T., Irshad, K., et al.: Decreased neurofilament gene expression is an index of selective axonal hypotrophy in aging. Neuroreport 8:799, 1996. 134. Kudwa, A., Shoemaker, S., Crutcher, K., and Isaacson, L.: Evidence for reduced accumulation of exogenous neurotrophin by aged sympathetic neurons. Brain Res. 948:24, 2002. 135. La Forte, R. A., Melville, S., Chung, K., and Coggeshall, R. E.: Absence of neurogenesis of adult rat dorsal root ganglion cells. Somatosens. Mot. Res. 8:3, 1991. 136. Lamberts, S. W. J., van den Beld, A. W., and van der Lely, A. J.: The endocrinology of aging. Science 278:419, 1997. 137. Larish, D. D., Martin, P. E., and Mungiole, M.: Characteristic patterns of gait in the healthy old. Ann. N. Y. Acad. Sci. 515:18, 1988. 138. Lascelles, R. G., and Thomas, P. K.: Changes due to age in internodal length in the sural nerve in man. J. Neurol. Neurosurg. Psychiatry 29:40, 1966. 139. Lawson, S. L.: Morphological and biochemical cell types of sensory neurons. In Scott, S. E. (ed.): Sensory Neurons:

Aging in the Peripheral Nervous System

140.

141. 142. 143.

144. 145.

146. 147. 148.

149.

150.

151.

152.

153.

154.

155.

156.

157.

158.

Diversity, Development and Plasticity. New York, Oxford University Press, p. 27, 1992. Lawson, S. N., and Waddell, P. J.: Soma neurofilament immunoreactivity is related to cell size and fiber conduction velocity in rat primary sensory neurons. J. Physiol. (Lond.) 435:41, 1991. Leist, M., and Nicotera, P.: Apoptosis, excitotoxicity, and neuropathology. Exp. Cell Res. 239:183, 1998. Levi-Montalcini, R.: The nerve growth factor: 35 years later. Science 237:1154, 1987. Low, P. A., Okazaki, H., and Dyck, P. J.: Splanchnic preganglionic neurons in man. I. Morphometry of preganglionic cytons. Acta Neuropathol. (Berl.) 40:55, 1977. Macintosh, S. R., and Sinclair, D. C.: Age-related changes in the innervation of the rat snout. J. Anat. 125:149, 1978. Madeira, M. D., Sousa, N., Santer, R. M., et al.: Age and sex do not affect the volume, cell numbers, or cell size of the suprachiasmatic nucleus of the rat: an unbiased stereological study. J. Comp. Neurol. 361:585, 1995. Makinodan, T., and Kay, M. M. B.: Age influence on the immune system. Adv. Immunol. 29:287, 1989. Mattson, M. P., Keller, J. N., and Begley, J. G.: Evidence for synaptic apoptosis. Exp. Neurol. 153:35, 1998. McMartin, D. N., and O’Connor, J. A. J.: Effect of age on axoplasmic transport of cholinesterase in rat sciatic nerves. Mech. Ageing Dev. 10:241, 1979. McQuarrie, I. G., Brady, S. T., and Lasek, R. J.: Retardation of the rate of slow axonal transport of cytoskeletal elements during maturation and aging. Neurobiol. Aging 10:359, 1989. Michaelis, E. K.: Molecular biology of glutamate receptors in the central nervous system and their role in excitotoxicity, oxidative stress and aging. Prog. Neurobiol. 54:369, 1998. Miller, F. D., and Kaplan, D. R.: Neurotrophin signaling pathways regulating neuronal apoptosis. Cell. Mol. Life Sci. 58:1045, 2001. Miller, F. D., Speelman, A., Mathew, T. C., et al.: Nerve growth factor derived from terminals selectively increases the ratio of p75 to trkA NGF receptors on mature sympathetic neurons. Dev. Biol. 161:206, 1994. Ming, Y., Bergman, E., Edström, E., and Ulfhake, B.: Evidence for increased GDNF signaling in aged sensory and motor neurons. Neuroreport 10:1529, 1999. Ming, Y., Bergman, E., Edström, E., and Ulfhake, B.: Reciprocal changes in the expression of neurotrophin mRNAs in target tissues and peripheral nerves of aged rats. Neurosci. Lett. 273:187, 1999. Mione, M. C., Dhital, K. K., Amenta, F., and Burnstock, G.: An increase in the expression of neuropeptidergic vasodilator, but not vasoconstrictor, cerebrovascular nerves in aging rats. Brain Res. 460:103, 1988. Mitsumori, K., Maita, K., and Shirasu, Y.: An ultrastructural study of spinal nerve roots and dorsal root ganglia in aging rats with spontaneous radiculoneuropathy. Vet. Pathol. 18:714, 1981. Mittal, K. R., and Logmani, F. H.: Age-related reduction in 8th cervical ventral nerve root myelinated fibre diameters and numbers in man. J. Gerontol. 42:8, 1987. Miwa, T., Miwa, Y., and Kanda, K.: Dynamic and static sensitivities of muscle spindle primary endings in aged rats to ramp stretch. Neurosci. Lett. 201:179, 1995.

505

159. Monji, A., Morimoto, N., Okuyama, I., et al.: The number of noradrenergic and adrenergic neurons in the brain stem does not change with age in male Sprague-Dawley rats. Brain Res. 641:171, 1994. 160. Munson, J. B., and McMahon, S. B.: Effects of GDNF on axotomized sensory and motor neurons in adult rats. Eur. J. Neurosci. 9:1126, 1997. 161. Nagashima, K., and Oota, K.: A histopathological study of the human spinal ganglia. 1. Normal variations in aging. Acta Pathol. Jpn. 24:333, 1974. 162. Nakayama, H., Noda, K., Hotta, H., et al.: Effects of aging on numbers, sizes and conduction velocities of myelinated and unmyelinated fibers of the pelvic nerve in rats. J. Auton. Nerv. Syst. 69:148, 1998. 163. Navarro, X., Kamei, H., and Kennedy, W. R.: Effect of age and maturation on sudomotor nerve regeneration in mice. Brain Res. 447:133, 1988. 164. Navarro, X., and Kennedy, W. R.: Changes in sudomotor nerve territories with aging in the mouse. J. Auton. Nerv. Syst. 31:101, 1990. 165. Nilsson, J., von Euler, A. M., and Dalsgaard, C. J.: Stimulation of connective tissue cell growth by substance P and substance K. Nature 315:61, 1985. 166. Nistico, G., Ciriolo, M. R., Fiskin, K., et al.: NGF restores decreases in catalase activity and increases superoxide dismutase and glutathione peroxidase activity in the brain of aged rats. Free Radical Biol. Med. 12:177, 1992. 167. Norris, A. H., Shock, N. W., and Wagman, I. H.: Age changes in the maximum conduction velocity of motor fibres of human ulnar nerves. J. Appl. Physiol. 5:589, 1953. 168. Ochoa, J., and Mair, W. G.: The normal sural nerve in man. II. Changes in the axons and Schwann cells due to ageing. Acta Neuropathol. 13:217, 1969. 169. Ohta, M., Offord, K., and Dyck, P. J.: Morphometric evaluation of first sacral ganglia of man. J. Neurol. Sci. 22:73, 1974. 170. Orike, N., Thrasivoulou, C., Wrigley, A., and Cowen, T.: Differential regulation of survival and growth in adult sympathetic neurons: an in vitro study of neurotrophin responsiveness. J. Neurobiol. 47:295, 2001. 171. Parhad, I. M., Scott, J. N., Cellars, L. A., et al.: Axonal atrophy in aging is associated with a decline in neurofilament gene expression. J. Neurosci. Res. 41:355, 1995. 172. Parkhouse, N., and LeQuesne, P. M.: Impaired neurogenic vascular response in patients with diabetes and neuropathic foot lesions. N Engl. J. Med. 318:1306, 1988. 173. Perret, E., and Regli, F.: Age and the perceptual threshold for vibratory stimuli. Eur. Neurol. 4:65, 1970. 174. Powell, H. C., Ward, H. W., Garrett, R. S., et al.: Glycogen accumulation in the nerves and kidney of chronically diabetic rats. J. Neuropathol. Exp. Neurol. 38:114, 1979. 175. Purves, D., Hadley, R. D., and Voyvodic, J. T.: Dynamic changes in the dendritic geometry of individual neurons visualized over periods of up to three months in the superior cervical ganglion of living mice. J. Neurosci. 6:1051, 1986. 176. Purves, D., Snider, W. D., and Voyvodic, J. T.: Trophic regulation of nerve cell morphology and innervation in the autonomic nervous system. Nature 336:123, 1988. 177. Quoniam, C., Hay, L., Roll, J. P., and Harlay, F.: Age effects on reflex and postural responses to propriomuscular inputs

506

178.

179.

180.

181.

182.

183.

184.

185.

186.

187. 188.

189.

190.

191.

192.

193.

194.

195.

Neurobiology of the Peripheral Nervous System generated by tendon vibration. J. Gerontol. A Biol. Sci. Med. Sci. 50:155, 1995. Raine, C. S., and Cross, A. H.: Axonal dystrophy as a consequence of long-term demyelination. Lab. Invest. 60:714, 1989. Ramirez-Leon, V., Kullberg, S., Hjelle, O. P., et al.: Increased glutathione levels in neurochemically identified fibre systems in the aged rat lumbar motor nuclei. Eur. J. Neurosci. 11:2935, 1999. Rao, R. S., and Krinke, G.: Changes with age in the number and size of myelinated axons in the rat L4 dorsal spinal root. Acta Anat. 117:187, 1983. Remak, R.: Observationes Anatomicae et Microscopicae de Systematis Nervosi Structura. Berlin, Sumptibus et Formis Reimerianis, 1838. Rexed, B.: Contributions to the knowledge of the postnatal development of the peripheral nervous system in man. Acta Psychiat. Neurol. Suppl. 33:1, 1944. Robbins, S., Waked, E., and McClaran, J.: Proprioception and stability: foot position awareness as a function of age and footwear. Age Ageing 24:67, 1995. Robertson, A., Day, B., Pollock, M., and Collier, P.: The neuropathy of elderly mice. Acta Neuropathol. (Berl.) 86:163, 1993. Rosenberg, S. I., Malmgren, L. T., and Woo, P.: Agerelated changes in the internal branch of the rat superior laryngeal nerve. Arch. Otolaryngol. Head Neck Surg. 115:78, 1989. Saitua, F., and Alvarez, J.: Do axons grow during adulthood? A study of the caliber and microtubules of sural nerve axons in young, mature, and aging rats. J. Comp. Neurol. 269:203, 1988. Samorajski, T.: Age differences in the morphology of posterior tibial nerves of mice. J. Comp. Neurol. 157:439, 1974. Santer, R. M.: Morphological evidence for the maintenance of the cervical sympathetic system in aged rats. Neurosci. Lett. 130:248, 1991. Santer, R. M., and Baker, D. M.: Enteric neuron numbers and sizes in Auerbach’s plexus in the small and large intestine of young adult and aged rats. J. Auton. Nerv. Syst. 25:59, 1988. Santer, R. M., Dering, M. A., Ranson, R. N., et al.: Differential susceptibility to ageing of rat preganglionic neurones projecting to the major pelvic ganglion and of their afferent inputs. Auton. Neurosci. 96:73, 2002. Satorre, J., Cano, J., and Reinoso-Suarez, F.: Stability of the neuronal population of the dorsal lateral geniculate nucleus (LGNd) of aged rats. Brain Res. 339:375, 1985. Schimrigk, K., and Ruttinger, H.: The touch corpuscles of the plantar surface of the big toe: histological and histometrical investigations with respect to age. Eur. J. Neurosci. 19:49, 1980. Schmidt, R. E., Beaudet, L., Plurad, S. B., et al.: Pathologic alterations in pre- and postsynaptic elements in aged mouse sympathetic ganglia. J. Neurocytol. 24:189, 1995. Schmidt, R. E., Chae, H. Y., Parvin, C. A., and Roth, K. A.: Neuroaxonal dystrophy in aging human sympathetic ganglia. Am. J. Pathol. 136:1327, 1990. Schmidt, R. E., Coleman, B. D., and Nelson, J. S.: Differential effect of chronic vitamin E deficiency on the development of neuroaxonal dystrophy in rat gracile/cuneate

196.

197. 198.

199.

200. 201.

202.

203. 204.

205.

206.

207.

208.

209.

210.

211.

212.

213. 214.

nuclei and prevertebral sympathetic ganglia. Neurosci. Lett. 123:102, 1991. Schroder, J. M., and Gibbels, E.: Marklöse Nervernfasern im Senium und Spätstadium der Thalidomid-polyneuropathie: quantitativ-electronen-mikroscopische Untersuchungen. Acta Neuropathol. 39:271, 1977. Sharma, A. K., Bajada, S., and Thomas, P. K.: Age changes in the tibial and plantar nerves of the rat. J. Anat. 130:417, 1980. Sloane, J. A., Hollander, W., Moss, M. B., et al.: Increased microglial activation and protein nitration in white matter of the aging monkey. Neurobiol. Aging 20:395, 1999. Smith, D. O., and Rosenheimer, J. L.: Factors governing speed of action potential conduction and neuromuscular transmission in aged rats. Exp. Neurol. 83:358, 1984. Sohal, R. S., and Weindruch, R.: Oxidative stress, caloric restriction, and aging. Science 273:59, 1996. Southam, E., Thomas, P. K., King, R. H. M., et al.: Experimental vitamin E deficiency in rats: morphological and functional evidence of abnormal axonal transport secondary to free radical damage. Brain 114:915, 1991. Spencer, P. S., and Thomas, P. K.: The examination of isolated nerve fibres by light and electron microscopy, with observations on demyelination proximal to neuromas. Acta Neuropathol. (Berl.) 16:177, 1970. Steiness, I. B.: Vibratory perception in normal subjects: a biothesiometric study. Acta Med. Scand. 198:315, 1957. Thomas, P. K., and King, R. H. M.: The degeneration of unmyelinated axons following nerve section: an ultrastructural study. J. Neurocytol. 3:497, 1974. Thomas, P. K., King, R. H. M., and Sharma, A. K.: Changes with age in the peripheral nerves of the rat: an ultrastructural study. Acta Neuropathol. (Berl.) 52:1, 1980. Thrasivoulou, C., and Cowen, T.: Regulation of rat sympathetic nerve density by target tissues and NGF in maturity and old age. Eur. J. Neurosci. 7:381, 1995. Toghi, H., Tsukagoshi, H., and Toyokura, Y.: Quantitative changes with age in normal sural nerves. Acta Neuropathol. (Berl.) 38:213, 1977. Tomonaga, M.: Histochemical and ultrastructural changes in senile human skeletal muscle. J. Am. Geriatr. Soc. 25:125, 1977. Uchida, A., Yorifuji, H., Lee, V. M., et al.: Neurofilaments of aged rats: the strengthened interneurofilament interaction and the reduced amount of NF-M. J. Neurosci. Res. 58:337, 1999. Uchida, Y., Tomonaga, M., and Nomura, K.: Age-related changes of myelin proteins in the rat peripheral nervous system. J. Neurochem. 46:1376, 1986. Ulfhake, B., Bergman, E., Edstrom, E., et al.: Regulation of neurotrophin signaling in aging sensory and motoneurons: dissipation of target support? Mol. Neurobiol. 21:109, 2000. Ulfhake, B., Bergman, E., and Fundin, B. T.: Impairment of peripheral sensory innervation in senescence. Auton. Neurosci. 96:43, 2002. van Steenis, G., and Kroes, R.: Changes in the nervous system and musculature of old rats. Vet. Pathol. 8:320, 1971. Vanneste, J., and van den Bosch de Aguilar, P.: Mitochondrial alterations in the spinal ganglion neurons in ageing rats. Acta Neuropathol. 54:83, 1981.

Aging in the Peripheral Nervous System 215. Verdu, E., Buti, M., and Navarro, X.: Functional changes of the peripheral nervous system with aging in the mouse. Neurobiol. Aging 17:73, 1996. 216. Verge, V. M., Gratto, K. A., Karchewski, L. A., and Richardson, P. M.: Neurotrophins and nerve injury in the adult. Philos. Trans. R. Soc. B 351:423, 1996. 217. Vlassara, H., Fuh, H., Makita, Z., et al.: Exogenous advanced glycosylation end products induce complex vascular dysfunction in normal animals: a model for diabetic and aging complications. Proc. Natl. Acad. Sci. U. S. A. 89:12043, 1992. 218. Vogelbaum, M. A., Tong, J. X., and Rich, K. M.: Developmental regulation of apoptosis in dorsal root ganglion neurons. J. Neurosci. 18:8928, 1998. 219. Vrancken, A. F. J. E., Franssen, H., Wokke, J. H. J., et al.: Chronic idiopathic axonal polyneuropathy and successful aging of the peripheral nervous system in elderly people. Arch. Neurol. 59:533, 2002. 220. Wagman, I. H., and Lesse, H.: Maximum conduction studies of motor fibers of ulnar nerve in human subjects of various ages and sizes. J. Neurophysiol. 15:235, 1952. 221. Wayner, M. J. J., and Emmers, R.: Spinal synaptic delay in young and aged rats. Am. J. Physiol. 19:403, 1958.

507

222. Whanger, A. D., and Wang, H. S.: Clinical correlates of the vibratory sense in elderly psychiatric patients. J. Gerontol. 29:39, 1974. 223. Wickelgren, I.: Is hippocampal cell death a myth? Science 271:1229, 1996. 224. Willis, W. D., and Coggeshall, R. E.: Sensory Mechanisms of the Spinal Cord. New York, Plenum Press, 1991. 225. Winkelmann, R. K.: Nerve changes in the aging skin. In Montagna, W. (ed.): Advances in Biology of Skin, Vol. 6: Aging: Symposium on the Biology of Skin. Oxford, Pergamon Press, p. 51, 1965. 226. Woodruff, R. H., and Franklin, R. J.: Demyelination and remyelination of the caudal cerebellar peduncle of adult rats following stereotaxic injections of lysolecithin, ethidium bromide, and complement/anti-galactocerebroside: a comparative study. Glia 25:216, 1999. 227. Yin, X. H., Crawford, T. O., Griffin, J. W., et al.: Myelinassociated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J. Neurosci. 18:1953, 1998. 228. Yurek, D. M., and Fletcher-Turner, A.: Differential expression of GDNF, BDNF, and NT-3 in the aging nigrostriatal system following a neurotoxic lesion. Brain Res. 891:228, 2001.

23 Oxidative Stress and Excitatory Neurotoxins in Neuropathy PHILLIP A. LOW

Overview Free Radical Biology Biologic Generators Protective Mechanisms Oxidative Stress Oxidative Stress and Peripheral Nerve Disease Diabetic Neuropathy Evidence of Oxidative Stress in Non-neural Tissue Integrative View of Oxidative Mechanisms of Diabetic Neuropathy

Evidence of Oxidative Stress in Diabetic Peripheral Nerve Antioxidants Molecular Pathogenesis of Endothelial Dysfunction Molecular Pathogenesis of Sensory Neuropathy: Pivotal Role of the Mitochondrion Molecular Pathogenesis of Schwann Cell Injury Oxidative Stress Related to Nerve Ischemia

OVERVIEW The pivotal role of oxidative stress and injury to peripheral nerve has become increasingly appreciated in the past decade. A number of interacting mechanisms cause oxidative injury. Especially important are the roles of ischemia, ischemia-reperfusion, the inflammatory response, and hyperglycemia. All these mechanisms cause oxidative stress. Oxidative DNA damage in turn can cause cellular dysfunction and apoptosis. Among the disorders, the role of oxidative stress has been best studied in diabetic neuropathy and, to a lesser degree, in ischemic neuropathy. It is likely also important in the toxic and inflammatory neuropathies, but little information is available at this time. This chapter focuses on oxidative stress as applied to diabetic neuropathy and ischemic injury. To a lesser extent we explore the role of excitotoxic injury, especially as it applies to painful neuropathies.

FREE RADICAL BIOLOGY Reactive oxygen species (ROS), or free radicals, comprise unpaired electrons that are highly reactive. They

Nerve Ischemia Reperfusion Injury Neuropathology Inflammatory Response Role of Cytokines Genetic Susceptibility Molecular Pathogenesis Neuroprotection of Ischemia-Reperfusion Injury N-Methyl-D-Aspartate and Glutamate Receptors

are generated by a number of chemical reactions, and the flow of electrons is shown in Equations 1 through 536: enz-H2 ⫹ 2O2 : enz ⫹ 2O2⫺ ⫹ 2H⫹

(1)

enz-H2 ⫹ O2 : enz ⫹ H2O2

(2)

O2⫺ ⫹ O2⫺ ⫹ 2H⫹ : H2O2 ⫹ O2

(3)

Me(n⫹)chelate ⫹ O2 : Me(n⫺1)chelate ⫹ O2

(4)

Me(n⫺1)chelate ⫹ H2O2 :Me(n⫹)chelate ⫹ OH ⫹ OH•

(5)

The enzyme (enz) can reduce O2 univalently to yield the superoxide anion (O2 : O2⫺) in Equation 1 (often called the univalent pathway) or divalently to yield hydrogen peroxide (O2 : H2O2) in Equation 2 (divalent pathway). The superoxide anion can undergo spontaneous dismutation (Equation 3), yielding hydrogen peroxide.113 Another fate of the superoxide anion is the reduction of metal (Me) chelates (Equation 4). The H2O2 may react with reduced 509

510

Neurobiology of the Peripheral Nervous System

metal chelates by the Haber-Weiss reaction (Equation 5) to generate hydroxyl radical (OH•). OH• reacts with several compounds, including lipids.131 Carbon-centered lipid radical (L•) is formed from polyunsaturated fatty acid by hydrogen abstraction (Fig. 23–1). Molecular rearrangement may occur to yield conjugated diene (a relatively stable footprint of lipid peroxidation). This compound reacts with oxygen rapidly to give lipid peroxy radical (LO2•). This radical attacks another lipid molecule and abstracts a hydrogen atom to give lipid hydroperoxide. At the same time it attaches another lipid radical and starts the process all over again (propagation). Molecular rearrangement yields lipid endoperoxide and fragmentation yields malondialdehyde, a frequently measured but relatively nonspecific footprint of lipid peroxidation. These small organic radicals are reactive and therefore short lived.151 The hydroxyl radical is so reactive that it reacts within 1 to 5 molecular diameters. The half-life of the hydroxyl radical has been estimated to be about 10⫺9 seconds.

Biologic Generators There are a number of biologic generators of free radicals.57 One source is sympathetic adrenergic postganglionic terminals. For instance, in response to ischemia, catecholamines are released from these nerve terminals and undergo oxidation, releasing free radicals.12 However, this contribution is quantitatively probably small.57 A more important source is the mitochondrion (mitochondrial leak mechanism). In normal mitochondria 1% of electron flow results in O2,181 thought to occur at the NADH

dehydrogenase step and near the ubiquinone component.11 This mechanism is enhanced in the diabetic state and in aging, in which free radical leakage is increased. A third source of ROS is the leukocyte. Leukocytes can produce large amounts of O2. Indeed, as much as 70% of O2 consumed by the activity of leukocytes may be converted to O2.186 This mechanism is an important source of ROS in reperfusion injury following ischemia to nerve. It is also an important source in the inflammatory neuropathies.

Protective Mechanisms A number of mechanisms reduce the toxicity of ROS.143 The toxicity of mitochondrial superoxide leak is greatly attenuated by the composition of the electron transport chain. Electrons and oxidant species so formed are tightly bound so that they do not participate in redox reactions. The electronic structure of free oxygen is also protective. Oxygen is a bi-radical that contains two separate orbitals each housing an electron with identical spin. Most molecules have two electrons with opposite spin. This difference in rotation creates a spin restriction that limits the ability of O2 to accept electrons directly. For oxygen to accept both electrons, an energy-requiring spin inversion must take place first. Toxic transitional metals are rendered nontoxic by certain proteins, yielding nontoxic proteins such as ferritin, transferrin, and ceruloplasmin. A series of antioxidants (Table 23–1) reduce chain initiation and suppress free radical generation, hence these are sometimes referred to as preventive antioxidants.131 Examples (with the relevant radical in parentheses)

Fatty acid with 3 double bonds Hydrogen abstraction

–H

Lipid radical Molecular rearrangement

Conjugated diene O2 uptake

Peroxy radical RH

OOH +R

Lipid hydroperoxide

Lipid endoperoxide Fragmentation

Malondialdehyde

FIGURE 23–1 Molecular steps of lipid peroxidation from polyunsaturated free fatty acids.

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

Table 23–1. Some Free Radical Defenses of Peripheral Nerve Cytosolic

Membrane

Ascorbate, urate, cysteine, glutathione, transferrin, albumin; ␤-carotene, ceruloplasmin, reduced glutathione, glutathione peroxidase, catalase, superoxide dismutase, glutathione reductase ␣-Tocopherol

are glutathione peroxidase (hydroperoxides), catalase (H2O2), transferrin (Fe2⫹), albumin (Cu2⫹), ␤-carotene (singlet oxygen), and ceruloplasmin (Fe2⫹). Chain-breaking antioxidants suppress free radical chain oxidation by molecular oxygen. These are often typically divided into water-soluble and lipid-soluble antioxidants. Examples of water-soluble antioxidants are ascorbate, urate, cysteine, and glutathione. The only significant lipid-soluble chain-breaking antioxidant is ␣-tocopherol.

Oxidative Stress There are a number of factors that increase free radical activity in tissue. The first is the increased generation of ROS by mechanisms such as ischemia, hyperglycemia, and inflammation. The second mechanism of increased free radical activity is reduced scavenging capacity. Free radical scavengers include superoxide dismutase (SOD), catalase, reduced glutathione (GSH), glutathione S-transferase, glutathione peroxidase, glutathione reductase, and ␣-tocopherol. Another mechanism leading to oxidative stress is an alteration of pro-oxidant status of tissue, increasing susceptibility of the cell to ROS. These include an accumulation of free iron,39 reduction in pH,4 and increased lipolysis of phospholipids, resulting in increased release of polyunsaturated fatty acids.48,152,196

OXIDATIVE STRESS AND PERIPHERAL NERVE DISEASE Oxidative stress is likely to play major pathogenetic roles in several types of neuropathy. These include neuropathies associated with ischemia (e.g., angiopathy, diabetes, peripheral vascular disease), inflammation (e.g., infections, inflammatory-demyelinating, sarcoid, paraneoplastic disease), neurotoxicity (e.g., heavy metals, cisplatin), and hyperglycemia. In this chapter we focus on diabetic and ischemic neuropathy. Earlier studies have focused on measurements of footprints of oxidative stress, typically in plasma. Subsequently measurements in nerve have been done. More recently, emphasis has shifted to measuring

511

indices that indicate tissue and especially DNA damage. Immunohistochemistry of 8-hydroxydeoxyguanosine and hydroxynonenal (HNE) permit simultaneous cellular localization and indexing of DNA damage.

DIABETIC NEUROPATHY With the combination of macrovascular and microvascular disease and chronic hyperglycemia, it is not surprising that evidence of oxidative stress has accumulated for multiple tissues. Much of the early data derived from non-neural tissue. We review some of this information here, followed by data on neural tissue. A number of major pathogenetic mechanisms have been postulated. These mechanisms converge in a final common pathway of oxidative injury. Finally we provide molecular schemes for the pathogenesis of diabetic sensory neuropathy, diabetic endothelial dysfunction, and Schwann cell apoptosis.

Evidence of Oxidative Stress in Non-neural Tissue Human Diabetes Plasma levels of lipid peroxide are increased in human diabetes.1,72,111,159 The highest levels were found in patients with microvascular angiopathy, manifested as retinopathy or microalbuminuria, and the lowest in those diabetic patients without angiopathy27,159; levels were normal in patients with well-controlled diabetes.159 The relationship may relate to the observation that low-density lipoproteins of diabetic patients are significantly more oxidizable than those of controls, an abnormality that is correctable by 6 weeks’ treatment with the antioxidant probucol.3 Presumably diabetics with angiopathy, who have higher levels of low-density lipoproteins, will have the greatest lipid peroxidation. GSH is reduced in erythrocytes from patients with type 2 diabetes mellitus, and there is a corresponding increase in oxidized glutathione (GSSG).122 In subsequent studies, these workers additionally demonstrated reductions in the glutathione-synthesizing enzyme ␥-glutamylcysteine synthetase, and in transport of the thiol [S-(s,4-dintrophenyl)glutathione] in erythrocytes of these patients. These abnormalities were reversible with improved glycemic control. They also demonstrated that a high-glucose medium augments the toxicity of xenobiotics on K562 cells, associated with reduction in both the enzyme and its messenger RNA (mRNA).197 Erythrocyte cuprozinc SOD is reduced in patients with type 2 diabetes.111 This reduction is suggested to be mediated by the accumulation of intracellular H2O2.99 ␣-Tocopherol levels are reported to be reduced in the platelets78,99 and erythrocytes but not plasma of insulindependent diabetes mellitus patients.

512

Neurobiology of the Peripheral Nervous System

Wolff 190 has emphasized the role of decompartmentalized transitional metals in diabetic patients in addition to the effects of hyperglycemia in producing auto-oxidative lipid peroxidation. Particular emphasis has been placed on copper and iron. Copper levels have been reported to be higher in diabetics than in normal subjects and are highest in those with angiopathy.110,134 Experimental Diabetes Oxidative stress occurs in experimental diabetes induced by streptozotocin and alloxan, and in the diabetic BB Wistar rat. Lipid peroxidation in chemical diabetes appears to be due to hyperglycemia and not the agent, because the pattern of changes is not agent specific. Streptozotocin and alloxan cause similar changes, and additionally the spontaneously diabetic BB Wistar rat shows virtually identical patterns of tissue antioxidant enzyme changes.50 Furthermore, these alterations are preventable by insulin treatment.10,50,188,189 The common mechanism of increased oxidative stress is, therefore, diabetes (hyperglycemia) and not its mode of induction. Plasma and liver lipid peroxides are increased in streptozotocin- or alloxan-induced diabetes111,150 and improved by ␣-tocopherol supplementation.150 Tissue levels of lipid peroxide, estimated by the thiobarbituric acid method, were increased in kidney and retina and accompanied by a reduction in fat-soluble antioxidants as determined by the ferric chloride–bipyridyl reaction. These changes were eliminated by insulin treatment.132 Complex patterns of changes in antioxidant enzymes have been described in different tissues in streptozocin diabetes.188,189 Liver and kidney have reduced catalase and SOD. Glutathione peroxidase and GSH are reduced in liver; glutathione peroxidase is increased in kidney. Catalase and glutathione reductase are increased in heart and pancreas and SOD is additionally increased in pancreas. One of the most common alterations is a reduction in cuprozinc SOD. This reduction has been reported in numerous tissues, including erythrocyte32,43,100,111,201; liver8,111; retina and kidney43,100; and spleen, heart, testis, pancreas, and skeletal muscle in rats with streptozotocin and/or alloxan diabetes.8,43,100,111 The loss of SOD appears to be a function of duration and severity of diabetes.85,99 ␣-Tocopherol is reported to be reduced in streptozotocin diabetes.85,99 Our observations are that plasma ␣-tocopherol is very variable and is greatly dependent on dietary intake; it can be increased in experimental diabetes because of polyphagia.130

Integrative View of Oxidative Mechanisms of Diabetic Neuropathy A major area of research in diabetic neuropathy is on oxidative stress. Earlier studies have focused on the demonstration of the presence of footprints of oxidative stress in serum

and in nerve tissue. Subsequent studies have focused on cellular localization of oxidative DNA injury and on the pathogenetic role of oxidative stress leading to peripheral neuronal injury and apoptosis. A number of potential pathogenetic mechanisms of diabetic neuropathy have been identified. These include ischemia/hypoxia, auto-oxidative lipid peroxidation, advanced glycation end products (AGEs), polyol pathway overactivity, protein kinase C (PKC) overactivity, excessive lipolysis, growth factor deficiency, and the inflammatory response. Of interest is that each of these can cause oxidative stress. Ischemia/Hypoxia in Experimental Diabetic Neuropathy A nerve blood flow deficit of 50% in experimental diabetic neuropathy (EDN)20,126,180 results in the generation of hypoxanthine from ATP, NADPH (the cofactor), and conversion of the inactive enzyme to xanthine oxidase.104,113 Figure 23–2 shows a reduction in nerve blood flow by 50% and a dosedependent prevention by treatment with ␣-lipoic acid. There is a concomitant reduction in GSH, the best index of current oxidative stress. Concomitantly, GSSG increases. Treatment with the antioxidant ␣-lipoic acid results in a dose-dependent normalization of GSH (Fig. 23–3). The diabetic state and endoneurial ischemia increase lipolysis, resulting in an increase in ␻-6 fatty acids such as linoleic acid and arachidonic acid,196 whose peroxidation results in 4-HNE, and causes prominent cytotoxic effects in cultured endothelial cells manifested by morphologic changes, diminished cellular viability, and impaired endothelial barrier function.60 HNE has relatively long half-lives within cells (minutes to hours), allowing for multiple interactions with cellular components.82 It impairs glutamate transport and mitochondrial function in neurons148 and mediates oxidative stress–induced neuronal apoptosis.95 Immunohistochemical labeling of HNE is a reliable indicator of oxidative injury to cells, including dorsal root ganglion (DRG) neurons and Schwann cells179 (see Figure 23–10 later). Auto-oxidative Lipid Peroxidation Hyperglycemia, by a process of auto-oxidation in the presence of decompartmentalized redox-active trace transitional metals, can generate highly reactive oxidants and result in lipid peroxidation.191 We have demonstrated, using an in vitro lipid peroxidation model (ascorbateiron–ethylenediaminetetraacetic acid [EDTA] preparation), that a high-glucose medium will result in lipid peroxidation, in vitro, of brain and sciatic nerve. The addition of 20-mM glucose to the incubation medium increased lipid peroxidation fourfold, confirming rapid and marked glucose-mediated auto-oxidative lipid peroxidation.129 These studies confirm the observation of auto-oxidative glycation/oxidation in plasma.7,190 The relationship between oxidative glycation and free radical production has been explored.66 Glucose auto-oxidation

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

513

FIGURE 23–2 Nerve blood flow (NBF) and nerve vascular resistance of controls (Con), animals with streptozotocin diabetic neuropathy (STZ), and animals supplemented with lipoic acid at doses of 20 mg/kg (STZ20), 50 mg/kg (Con50; STZ50), and 100 mg/kg (STZ100). Lipoic acid supplementation results in normal NBF and nerve vascular resistance. Significance of difference, a, P ⬍ .05. (From Nagamatsu, M., Nickander, K. K., Schmelzer, J. D., et al.: Lipoic acid improves nerve blood flow, reduces oxidative stress, and improves distal nerve conduction in experimental diabetic neuropathy. Diabetes Care 18:1160, 1995, with permission.)

results in the production of protein-reactive ketoaldehydes, hydrogen peroxide, and other highly reactive oxidants and the fragmentation of proteins (indicative of free radical mechanisms). Glycation and oxidation are simultaneous and were considered to be inextricably linked.66 There are numerous indices of oxidative stress. These include malondialdehyde, lipid hydroperoxide, and conju-

gated dienes. Malondialehyde is rather nonspecific and lipid hydroperoxides are somewhat unstable.101 The formation of the double bond in conjugated dienes (see Fig. 23–1) confers stability, and its measurement has been a preferred index of oxidative stress. Sciatic nerve conjugated dienes were significantly increased in diabetic rats that had been diabetic for 1, 4, or 12 months (Fig. 23–4).

FIGURE 23–3 Sciatic nerve reduced glutathione (GSH) concentrations in controls (Con) and in animals with restricted caloric intake (Con[R]), streptozotocin diabetes (STZ), ␣-tocopherol depletion (⫺), and supplementation with lipoic acid at 20, 50, and 100 mg/kg. Lipoic acid supplementation resulted in a dose-dependent prevention of GSH. Significance of difference versus control, *P ⬍ .05; ***P ⬍ .001; a, P ⬍ .05; t, P ⬍ .001. (From Nagamatsu, M., Nickander, K. K., Schmelzer, J. D., et al.: Lipoic acid improves nerve blood flow, reduces oxidative stress, and improves distal nerve conduction in experimental diabetic neuropathy. Diabetes Care 18:1160, 1995, with permission.)

514

Neurobiology of the Peripheral Nervous System

Protein Kinase C There is some uncertainty as to whether activity of PKC is increased in nerve and as to which subtype is altered. Our own studies found an increase but did not study specific isoforms.90 It is known that inhibition of PKC-␤ will reduce oxidative stress109 and will normalize the deficits in blood flow and nerve conduction.19 In endothelial cells, high glucose causes NF-␬B activation. Co-incubation with a selective PKC inhibitor, calphostin C, produced a concentration-dependent inhibition of glucose-induced NF-␬B activation, suggesting that PKC is important at the endothelial cell level in the activation of adhesion molecules and generation of ROS. FIGURE 23–4 Sciatic nerve conjugated dienes at 1, 4, and 12 months of diabetes. (From Low, P. A., and Nickander, K. K.: Oxygen free radical effects in sciatic nerve in experimental diabetes. Diabetes 40:873, 1991, with permission.)

AGE, Its Receptor, and Nuclear Factor-␬B Glycation and the formation of AGE is followed by binding with its receptor (RAGE), the activation of nuclear factor-␬B (NF-␬B),164 and the generation of ROS and an inflammatory response.83 There is induction of specific DNA binding activity for NF-␬B in the vascular cell adhesion molecule-1 promoter region. The necessity for RAGE and the role of ROS was demonstrated by a block of this induction by antiRAGE immunoglobulin G or N-acetylcysteine (GSH donor). The application of this finding to humans is supported by the finding that peripheral blood mononuclear cells isolated from patients with diabetic nephropathy show increased activation of NF-␬B.64 There is a vicious cycle involving AGE-producing superoxide, superoxide-accelerating AGE generation, and AGE-quenching nitric oxide (NO).14 Generation of ROS by Polyol Pathway Overactivity Polyol pathway overactivity generates ROS in a number of ways. Depletion in NADPH results in NO deficiency and an increase in leukocyte superoxide anion generation. Because NADPH is also required for the regeneration of GSH, its depletion results in a reduction of GSH.96 Sorbitol dehydrogenase, the second enzyme in the polyol pathway that converts sorbitol to fructose, also contributes to oxidative stress most likely because depletion of its cofactor NAD⫹ leads to more glucose being channeled through the polyol pathway.96 Oxidative stress results in poly(ADP-ribose) polymerase (PARP) activation. In conditions with reduced NAD⫹, PARP activation leads to caspase-3 activation and apoptosis.172 Polyol pathway overactivity results in reductions in myoinositol and taurine. The latter is a potent antioxidant, and its depletion174 further contributes to oxidative stress.

Excessive Lipolysis The diabetic state and endoneurial ischemia increase lipolysis, resulting in an increase in ␻-6 fatty acids such as linoleic acid or arachidonic acid,196 whose peroxidation results in 4-HNE60 (discussed earlier). Malondialdehyde is also generated in the cyclooxygenase pathway. Growth Factor Deficit Lipid peroxidation is aggravated by a reduction in nerve growth factor.58 Nerve growth factor reduction will reduce, and its administration will restore, glutathione peroxidase and catalase.133,157 Inflammatory Response The main sources of ROS in mammals are the leukocyte and the mitochondrion. The quantitative role of the leukocyte, cytokines, and catecholamine oxidation in diabetic nerve is uncertain. Leukocytic infiltration is not a feature in most cases of EDN. In human diabetic neuropathy, there is some evidence of an immune-mediated process as suggested by the presence of iritis and inflammatory infiltrates in sympathetic ganglia.41,103 Some forms of diabetic neuropathy, such as acute autonomic neuropathy and the subacute radiculoplexus neuropathies, might be associated with prominent round cell infiltration.44,147 A role of associated ROS and oxidative stress should be considered.

Evidence of Oxidative Stress in Diabetic Peripheral Nerve Oxidative stress in diabetic nerves can develop by (1) an impairment of free radical defenses, (2) an increase in pro-oxidant status, and (3) increased free radical generation. Evidence for changes in all three areas are present in diabetic nerves and combine to generate oxidative stress and injury to cause diabetic peripheral neuropathy. Reduction of Free Radical Defenses in Diabetic Peripheral Nerve Peripheral nerve has a number of cytosolic and lipophilic antioxidant defenses, as described earlier (see Table 23–1).

515

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

Table 23–2. Enzymatic Free Radical Scavenger Activity in Nerve Relative to Brain and Liver Activity† Antioxidant* GSH-GSSG GSH-Px (: H2O2) GSH-Px (: t-BOOH) GSSG Reductase GST (: CDNB) GST (: 4-HNE) DT-diaphorase SO

Nerve

% of Brain

Brain

261.0 ⫾ 24.0 6.4 ⫾ 2.1 4.6 ⫾ 1.4 2.7 ⫾ 0.3 9.4 ⫾ 2.4 5.4 ⫾ 1.2 9.9 ⫾ 2.2 93.8 ⫾ 12.4

10% 13% 9% 13% 4% 5%

2620.0 ⫾ 124.0 48.5 ⫾ 5.5 50.3 ⫾ 1.4 20.5 ⫾ 0.6 232.5 ⫾ 14.0 116.5 ⫾ 10.4

Liver 74.9 ⫾ 7.1 144.0 ⫾ 16.9 56.3 ⫾ 6.1 171.3 ⫾ 55.4 345.2 ⫾ 92.0 208.0 ⫾ 49.9 171.3 ⫾ 55.4

GSH-GSSG ⫽ total glutathione; GSH-Px ⫽ glutathione peroxidase; t-BOOH ⫽ tert-butyl-hydroperoxide; GSSG Reductase ⫽ gluathione disulfide reductase; GST ⫽ glutathione transferase; CDNB ⫽ 1-chloro-2,4-dinitrobenzene; 4-HNE ⫽ 4-hydroxy-2,3-trans-nonenol; SO ⫽ superoxide dismutase. † Results are expressed as nanomoles per minute per milligram of protein, ⫾SEM. *

The cytosolic and membrane antioxidants work in concert in a well-organized interacting chain. Activity of all components is needed to maintain these antioxidants in their reduced state. There is an enormous literature on free radical biology, but information on peripheral nerve, in particular diabetic peripheral nerve, is quite limited. Free radical defenses in nerve are selectively reduced. Compared with that of brain, SOD activity is as high,153,154 but GSH and glutathione-containing enzymatic scavengers (glutathione peroxidase and reductase) are only about 10% that of brain61 (Table 23–2). Under conditions of increased oxidative stress, diabetic nerves undergo a reduction in GSH and neuropathy. The diabetic state results in additional alterations in these defenses (Table 23–3). Cuprozinc SOD is reduced in sciatic nerve in EDN, and this reduction is improved by insulin treatment.102 Glutathione peroxidase is further reduced in EDN.61,101 These workers were able to regress glutathione peroxidase activity against blood glucose concentration (y ⫽ 69.3 ⫺ 0.9x, where y is enzyme activity measured as nanomoles per milligram of protein per minute and x is blood glucose in millimoles per liter; r ⫽ .9). Plasma and leukocyte ascorbic acid are reduced and oxidation of this

antioxidant is increased.68,171 GSH is reduced in diabetic nerves.126,130 SOD,102 GSH,130 glutathione peroxidase,61 and catalase and total quinone reductase53,141 are reduced in diabetic rodent nerves. Increased Free Radical Generation in EDN Potential sources of free radical generation in diabetes include hyperglycemia, ischemia, increased mitochondrial leak, catecholamine oxidation, and leukocytes. A nerve blood flow deficit of 50% in EDN20,162,165,180 results in the generation of hypoxanthine from ATP and NADPH (the cofactor), and conversion of the inactive enzyme to xanthine oxidase.101,113 Increased Diabetic Pro-oxidant Status Diabetic pro-oxidant status is increased because diabetic sciatic nerve has increased polyunsaturated fatty acids with excessive lipolysis.196 It has also been suggested that metals, especially copper, might be decompartmentalized or increased.190 Although there is good evidence that oxidative stress occurs in the diabetic state, the demonstration of lipid peroxidation alone is insufficient evidence for a free

Table 23–3. Changes in Antioxidant Defenses in EDN Antioxidant

Change

Reference(s)

Superoxide dismutase (cuprozinc) Glutathione peroxidase Ascorbic acid content

Reduced Reduced Reduced

Ascorbic oxidation

Increased

Reduced glutathione

Reduced

Low and Nickander (1991)102 Hermenegildo et al. (1993)61 Sodhi et al. (1989),170 Greene et al. (1989)55 Sodhi et al. (1989),170 Greene et al. (1989)55 Nickander et al. (1994)130 Nagamatsu et al. (1995)126

516

Neurobiology of the Peripheral Nervous System

radical role in the etiopathogenesis of neuropathy. We propose that the following four criteria should be satisfied before accepting that neuropathy is due to excessive lipid peroxidation: 1. An increase in lipid peroxidation in previously normal nerves results in neuropathy. 2. An increase in lipid peroxidation in diabetic neuropathic nerves further worsens function. 3. Antioxidant improves or prevents neuropathy. 4. This improvement or prevention is associated with an improvement in the indices of lipid peroxidation of peripheral nerve. When weanling rats were fed an ␣-tocopherol–deficient diet, plasma ␣-tocopherol became unmeasurable.130 Endoneurial oxidative stress developed, as indicated by a reduction in GSH and increased lipid peroxidation (conjugated dienes, lipid hydroperoxides). These findings were associated with the development of a sensory neuropathy in previously normal nerves.130 A second line of evidence derives from the worsening of neuropathy in EDN130 associated with an increase in lipid peroxidation. The pro-oxidant primaquine caused conduction deficit in the hind limb nerves of previously normal rats, a reduction in sciatic nerve blood flow, and endoneurial hypoxia; these deficits were prevented by treatment with the antioxidant probucol.18 Evidence of increased nerve lipid peroxidation is now available. Diabetic peripheral nerve has increased conjugated dienes,102,130 reduced GSH,130 and reduced glutathione peroxidase.61,101 Among the indices of increased oxidative stress in a chronic in vivo situation, malondialdehyde, conjugated dienes, and lipid hydroperoxides are increased in peripheral nerve in EDN,33,130,162 and the increase is more consistent in lumbar dorsal root and superior cervical ganglia.130 The most reliable index of increased oxidative stress is reduction in GSH.130

Antioxidants Several antioxidants have shown promise in the treatment of EDN (Table 23–4). Probucol, a powerful free radical scavenger,73 normalizes both nerve blood flow and electrophysiology.18,77 These workers also reported an improvement in nerve perfusion with a 1% vitamin E diet.77 ␣-Lipoic acid is a powerful lipophilic antioxidant.16,145 It has a number of additional properties relevant to neuropathy. It stimulates nerve growth factor123 and promotes fiber regeneration in a culture system.40 ␣-Lipoic acid will prevent the deficit in nerve blood flow (see Fig. 23–2) and the distal electrophysiologic changes in EDN.126 These benefits are associated with improved indices of lipid peroxidation, the most sensitive of which is a prevention of the reduction in GSH (see Fig. 23–3). Mild improvement in EDN has been reported following treatment with GSH.13 One percent butylated hydroxytoluene for 2 months completely prevented the conduction deficit in EDN.21 Carvedilol, an antioxidant and vasodilator, is reported to be efficacious in preventing the deficits in both blood flow and conduction.29 Cameron and Cotter17 have reported improvement in both nerve blood flow and nerve conduction in EDN following treatment with the iron chelator deferoxamine. We have also found that modified SOD will prevent the deficit in nerve blood flow. However, a significant deficiency in most antioxidant studies in EDN is the lack of measurements of oxidative stress, so that it is not known if antioxidants, all of which have multiple mechanisms of action, improve neuropathy by their antioxidant or other properties. We evaluated the gene expression of glutathione peroxidase, catalase, and SOD (cuprozinc and manganese separately) in L4/5 DRG and superior cervical ganglion, as well as enzyme activity of glutathione peroxidase in DRG and sciatic nerve, in EDN of 3-month and 12-month durations. Gene expression of glutathione peroxidase, catalase, cuprozinc SOD, and manganese SOD was not reduced in EDN at either 3 or 12 months. Catalase mRNA was

Table 23–4. Antioxidants in the Treatment of Experimental Diabetic Neuropathy Antioxidant

Lipid Peroxidation

NBF Deficit

NCS Deficit

Reference(s)

Probucol ␣-Tocopherol

Not studied Not studied

Prevented Prevented

Prevented Prevented

␣-Lipoic acid GSH Butylated hydroxytoluene Carvedilol Deferoxamine SOD ␤-Carotene Ascorbic acid

Improved Not studied Not studied Not studied Not studied Improved Not studied Not studied

Prevented Not studied Not studied Prevented Prevented Prevented Prevented Partial prevention

Prevented Partial prevention Prevented Prevented Prevented Partial prevention Prevented Partial prevention

Cameron et al. (1994)18 Cameron et al. (1994),18 Cotter et al. (1995),30 Karasu et al. (1995)77 Nagamatsu et al. (1995)126 Bravenboer et al. (1992)13 Cameron et al. (1993)21 Cotter and Cameron (1995)29 Cotter and Cameron (1995)28 J. F. Poduslo, unpublished Cotter et al. (1995)30 Cotter et al. (1995)30

GSH ⫽ reduced glutathione; NBF ⫽ nerve blood flow; NCS ⫽ nerve conduction study; SOD ⫽ superoxide dismutase.

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

significantly increased in EDN at 12 months. Glutathione peroxidase enzyme activity was normal in sciatic nerve. We concluded that gene expression is not reduced in peripheral nerve tissues in very chronic EDN. Changes in enzyme activity may be related to duration of diabetes or due to posttranslational modifications.89 Experimental depletion of tissue ␣-tocopherol results in depletion in GSH, increased lipid peroxidation, and the development of a sensory neuropathy in previously normal nerves and worsening of neuropathy in diabetic nerves.130 As a result, diabetic peripheral nerve has increased malondialdehyde, conjugated dienes (see Fig. 23–4), and lipid hydroperoxides33,130,162; the increase is more consistent in lumbar dorsal root and superior cervical ganglia.130,162 The most reliable index of increased oxidative stress is reduction in GSH.130,162 Finally, oxidative stress can cause neuropathy, because experimental depletion of tissue ␣-tocopherol results in a reduction in GSH, increased lipid peroxidation,

517

and the development de novo of a distal sensory neuropathy, and worsens EDN.130 Recent emphasis is on immunocytochemical labeling of products of oxidative stress that are known to be neurotoxic.81 Ischemia activates phospholipase A2 and the arachidonic acid cascade, resulting in superoxide anion and 4-HNE, known to cause neuronal apoptosis and mitochondrial damage.81,148 Additional support for a role of oxidative stress comes from improvement in neuropathy following treatment with probucol,18,73 vitamin E,77 and especially ␣-lipoic acid. With increasing age, there are myelin changes in ventral root.92 More recently, advanced changes in both dorsal and ventral roots have been described in EDN with a duration beyond 6 months.178 We demonstrated pathologic alteration in spinal roots and DRGs, showing demyelination, vacuolar degeneration, and pigmentary changes158 (Fig. 23–5). These observations have been furthered by recent pathophysiologic

FIGURE 23–5 Electron micrographs of representative ventral root fibers showing a progression of myelinopathy in experimental diabetes. A, Myelin decompaction. B, A rim of intact myelin surrounds degenerating myelin, with early myelin balls and a denuded atrophic axon. C, An atrophic axon is surrounded by myelin showing residual rims separated by myelin degeneration assuming a prominent honeycombed appearance. D, A completely demyelinated axon. (From Sasaki, H., Schmelzer, J. D., Zollman, P. J., and Low, P. A.: Neuropathology and blood flow of nerve, spinal roots and dorsal root ganglia in longstanding diabetic rats. Acta Neuropathol. 93:118, 1997, with permission.)

518

Neurobiology of the Peripheral Nervous System

studies supporting the sensory neuron as a target, in particular with the mitochondrion as the most relevant organelle. In a series of studies, the Michigan group has confirmed ballooning of mitochondria and disruption of the internal cristae as a result of the neurotoxic effects of glucose. They demonstrated apoptosis with sequential steps, including a reduction of mitochondrial membrane potential, leakage of cytochrome c, and caspase-3 activation.45,54,146 The molecular pathogenesis of diabetic neuropathy has focused on oxidative DNA damage to three targets. Most of the interest has focused on the endothelial cell and DRG, and to a lesser degree the Schwann cell.

Molecular Pathogenesis of Endothelial Dysfunction A simplified model of the role of oxidative stress in the pathogenesis of endothelial dysfunction and reduced nerve blood flow is shown in Figure 23–6. Nerve blood flow is impaired as a result of reduced action of vasodilators (including nitric oxide and prostacyclin) and enhanced activity of vasoconstrictors (including endothelin). The roles of endothelin, vasopressin, and prostaglandins are all important but are not covered here. Hyperglycemia in vivo and in vitro generates ROS and lipid peroxidation129 and single-strand DNA breaks. This in turn activates PARP, a profuse nuclear enzyme that has been implicated in response to DNA injury.172 PARP initiates an energy-consuming cycle by transferring ADP-ribose units from NAD⫹ to nuclear proteins, resulting in a depletion of the intracellular NAD⫹ and ATP pools, which impairs glycolysis and mitochondrial respiration leading cellular dysfunction.175 PARP-172 and polyol pathway–induced

FIGURE 23–6 Suggested pathogenesis of endothelial dysfunction.

reduction in NADPH results in reduced endothelial nitric oxide synthase (eNOS) activity, an observation that has consistently been found in diabetic nerves.84,112,200 NADPH depletion also results in a reduction in GSH and escalating oxidative stress. PARP is implicated in the process of pro-inflammatory gene expression.176 NF-␬B activation, intracellular adhesion molecule-1 (ICAM-1), tumor necrosis factor-␣ (TNF-␣), and nitric oxide synthase (NOS) have been implicated in the pathogenesis of diabetic endothelial dysfunction.5,120,149 Support for this notion derives from the observation that, in macrophages from PARP-deficient animals, there is a reduction in the activation of NF-␬B, with subsequent suppression of pro-inflammatory gene expression, ICAM-1,199 TNF-␣,144 and inducible NOS176 in PARP-deficient cells. Treatment with the PARP inhibitors reversed endothelial dysfunction.172 The pathogenesis of ROS-mediated endothelial dysfunction is shown in Figure 23–6. Hyperglycemia induces ROS, especially superoxide anion. ROS, especially hydroxyl and peroxynitrite radicals, cause DNA strand breaks and PARP generation. PARP activation depletes NAD⫹, NADPH, and ATP, resulting directly and indirectly (by reducing eNOS) in endothelial dysfunction. Hyperglycemia and PARP activate NF-␬B, resulting in pro-inflammatory mediators and endothelial dysfunction. The net effect of PARP activation, therefore, results in energy depletion, activation of adhesion molecules, an inflammatory response, NO depletion and endoneurial ischemia/hypoxia, and escalating oxidative stress.

Molecular Pathogenesis of Sensory Neuropathy: Pivotal Role of the Mitochondrion Diabetic neuropathy disproportionately affects sensory nerve fibers. The most consistent findings in chronic EDN in our hands is conduction slowing of distal sensory fibers.126 The most convincing pathologic alterations are in the nerve roots, with prominent myelin and axonal changes that develop in chronic (ⱖ6 months) experimental diabetes.158,178 These alterations in ventral root can occur in rats older than 12 months,92 although the changes in diabetes are more advanced and occur earlier; the changes in dorsal root are specific to the disorder. These findings, coupled with the recognition of the pivotal role of oxidative stress in diabetes and the high density of mitochondria in DRG neurons, led to the hypothesis that much of the sensory neuropathy is due to the effect of oxidative injury to the DRG and in particular the mitochondrion. The suggested pathogenesis is shown in Figure 23–7. Hyperglycemia results in oxidative stress by numerous pathways described earlier. Of particular importance to mitochondrial function is the reduction in GSH, reduction in NADPH, and increase in NADH. The generation of

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

519

depletion of energy substrates blocks necrosis and commits the cell to apoptosis.71 We evaluated 8-hydroxydeoxyguanosine and caspase-3 expression using immunohistochemistry and terminal deoxynucleotidyl transferase–mediated dUTP nick end-labeling (TUNEL) staining in L5 DRGs of streptozotocin diabetic rats. Duration of diabetes was 1 month, 3 months, and 12 months. Oxidative injury began early and was sustained, with 12% of neurons showing TUNEL positivity at 12 months. Morphometric analysis showed a modest reduction of neurons, with a selective loss of 43% of the largest neurons.88

Molecular Pathogenesis of Schwann Cell Injury

FIGURE 23–7 Suggested pathogenesis of sensory neuropathy.

ROS, especially superoxide, and peroxynitrite results in single-strand breaks of DNA. DRG neuronal labeling with 8-hydroxydeoxyguanosine, indicative of oxidative DNA injury to the neuron, is consistently seen in chronic EDN (Fig. 23–8). Pivotal to the pathogenesis to apoptosis of DRG neurons is the release of cytochrome c from outer mitochondrial membrane. The egress of cytochrome c requires the opening of mitochondrial pores. There is good evidence for this event. Mitochondrial membrane potential falls in the hyperglycemic mitochondrion173 associated with swelling of the organelle. The precise mechanism of pore opening is not known. The most likely mechanisms are a combination of an alteration in the balance between the protective and pro-apoptotic components of the Bcl family and a reduction in GSH.70 Cytochrome c release has been demonstrated,173 and this binds to Apaf 1, which then recruits procaspase-9 (aptosome complex).59 This oligomerization and autoactivation of procaspase-9 to caspase-9 is followed by proteolytic cleavage of procaspase-3 (and other executioner procaspases), activating the executioner caspase-3. It is consistently increased in experimental diabetes.45,156 Caspase-3 will activate PARP with several important consequences. It promotes the inflammatory response by increasing NF-␬B,59 depletes energy substrates, and starts the cycles of escalating oxidative injury. The net effect is apoptosis of DRG neurons (see Fig. 23–7). The

Limited information is available on this subject. Most of the studies on the Schwann cell have been subsidiary to studies on DRG neurons. Changes such as chromatin condensation and vacuolar mitochondrial changes have been described.37,146 These observations of Schwann cells in DRGs have been supplemented by in vitro studies. Overexpression of Bcl-XL, or insulin-like growth factor, signaling via phosphatidylinositol 3-kinase, was reported to protect Schwann cells from glucose-mediated apoptosis in vitro.37 Nukada and colleagues have studied morphologic alterations that occur in Schwann cells subjected to reperfusion following ischemic injury to nerve.140 Additional studies have been done on the unusual susceptibility of diabetic Schwann cells to ischemia and reperfusion injury. For convenience, these studies are described under Neuropathology in the next section.

OXIDATIVE STRESS RELATED TO NERVE ISCHEMIA Nerve Ischemia There have been special problems in the study of peripheral nerve ischemia. Unlike brain and heart, in which ischemic effects are readily produced, peripheral nerve is relatively resistant because of its low energy needs and extensive anastomoses.105 Thus early attempts to produce experimental ischemia were largely unsuccessful. Models of ischemia have, however, been subsequently developed using ligation or infusion of microemboli of appropriate size to occlude microvessels.42,62,93,146 To study reperfusion injury, we used the model of ligation of supplying arteries (and collaterals) for a specified duration, followed by release of ligatures for reperfusion.117,163 This model reliably permits the separate evaluation of the effects of ischemia and those of reperfusion. The effects of ischemia and reperfusion have been studied on the blood-nerve barrier (BNB) and then on endoneurial contents.

520

Neurobiology of the Peripheral Nervous System

Diabetic

TUNEL

8-OHdG

Control

FIGURE 23–8 L5 dorsal root ganglion neuron of chronic experimental diabetic neuropathy (duration 12 months) shows labeling with 8-hydroxydeoxyguanosine (top) and TUNEL positivity (bottom). Age-matched control neurons are negative. See Color Plate

The Blood-Nerve Barrier Ischemia results in depletion of energy metabolites, an increase in tissue reducing equivalents,104,128 and an increase in tissue xanthine.69 Ischemia activates a calciumdependent protease that converts cytosolic xanthine dehydrogenase to xanthine oxidase6 in capillary endothelium.67 The above-mentioned conditions create an oxygen free radical–generating system. Furthermore, reperfusion accelerates the formation of ROS by introducing O2 into a system primed and generating ROS.113 ROS regularly increase microvascular permeability,36,94 and disruption of the BNB is an important first step in oxidative injury to nerve. Reperfusion-mediated BNB breakdown was clearly shown in direct measurements of permeability of rat sciatic nerve. We found that ischemia alone for 1 hour did not cause BNB alteration, but reperfusion resulted in reduced reflow and an increase in the permeability coefficient, indicating disruption of the interface.163 We measured both nerve blood flow and the permeability–surface area product. The permeability–surface area product may be increased as a result of an increase in permeability or an increase in the surface area (recruitment of capillaries). In this study, nerve blood flow was found to be persistently reduced with reperfusion. Therefore, the progressive increase in the permeability–surface area product must

be due to an increase in permeability coefficient. Reduced reflow is due to swelling of endothelial cells and pericytes with adhesion of leukocytes,139 resulting from activation of adhesion molecules, especially ICAM-1.140 Adrenergic Contribution Ischemia results in the generation of ROS from several sources. There is release of norepinephrine from sympathetic nerve terminals.12 Catecholamine release results in the increased synthesis and release of prostacyclin metabolites49,52 by ␣ receptor–mediated and calcium/calmodulindependent mechanisms. Calmodulin is known to activate phospholipase A2,192 resulting in a breakdown of membrane phospholipids and activation of the arachidonic acid cascade,15 generating leukotrienes (further damaging endothelial cells) and prostaglandins. Prostacyclin, localized in vascular endothelial cells,106 is the major vasodilator and inhibitor of platelet aggregation described, acting by stimulating platelet adenylate cyclase.51 Prostacyclin synthetase is rich in vascular endothelial cells119,187 and inhibited by lipid endoperoxides.118 The ratio of prostacyclin to thromboxane A2 is considered to be important in the maintenance of vascular tone118 and is reduced in ischemia-reperfusion and experimental diabetes.101

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

Endoneurial Contents The effects of oxidative stress depend on a balance among the severity of oxidative stress, pro-oxidant status, and free radical defenses. Free radical defenses include the enzymatic scavengers, especially SOD, glutathione peroxidase, glutathione reductase, GSH, ascorbate, and ␣-tocopherol. We and others have demonstrated changes in all three areas in ischemic peripheral nerve. An increase in free radical generation occurs in ischemia-reperfusion, indicated by the breakdown of the blood-nerve interface,163 endoneurial edema, increase in hydroperoxides, and ischemic fiber degeneration.127 Altered pro-oxidant status occurs with an increase in polyunsaturated fatty acids, increasing arachidonic acid.196 Free radical defenses in nerve are selectively reduced. GSH and glutathione-containing enzymatic scavengers are only about 10% that of brain.153 With ischemia, superoxide anion is converted to H2O2, but its further decomposition, mediated by glutathione peroxidase,26 may be compromised if the low content of this enzyme is further reduced. 4-HNE, an aldehydic product of membrane lipid peroxidation, is especially pertinent in that ischemia, by increasing arachidonic acid, provides substrate for HNE,80 which mediates neuronal apoptosis.95 GSH is an especially sensitive indicator of ongoing oxidative stress.126 Similarly, Wagner et al.183 reported that the frequency of fiber degeneration is inversely proportional to GSH. ␣-Lipoic acid, a powerful lipophilic ROS scavenger,145 normalized the perfusion and conduction deficits in experimental diabetes126 and neuroprotected nerve subjected to ischemia-reperfusion.115 Lipid peroxidation generates HNE, which selectively inhibits ␣-ketoglutarate and pyruvate dehydrogenases and depletes these two multienzyme complexes of ␣-lipoic acid,65 rendering treatment with this antioxidant particularly relevant.

Reperfusion Injury The effects of ischemia in several tissues are amplified during reperfusion, a phenomenon referred to as reperfusion injury.57,113 Applied to peripheral nerves, there are a number of important observations. Peripheral nerve has low energy requirements and extensive collateral flow. One implication is its resistance to both ischemia and reperfusion injury. We demonstrated that reperfusion injury does occur in nerve, but only if the ischemia is near total and lasts at least 60 minutes.34,163 Ischemia followed by reperfusion results in disruption of the BNB and the activation of an immediate response involving adrenergic and protease activation, followed later by ROS associated with the inflammatory response and cytokines. The role of norepinephrine has been described earlier, and the inflammatory and cytokine responses are described later. Reperfusion following ischemia results in protease activation and a massive increase in intracellular calcium,25 leading to calcium-mediated activation of phospholipases and the production of free fatty acids and lysophospholipids.

521

There is some indirect evidence that similar ischemic mechanisms are operative in peripheral nerve. Calcium ionophore will cause vesicular disruption of myelin161,169 and axonal degeneration,160 and nerve reconnection in a calcium-free medium resulted in improved functional recovery. 35 The vasoconstriction and microvascular ischemia/hypoxia in disorders such as diabetes have been in part ascribed to perturbations of prostaglandins and ROS generation. Lipid hydroperoxides are increased and inhibit prostacyclin synthetase activity, resulting in a reduced prostacyclin:thromboxane ratio, and vasoconstriction and platelet aggregation.198 Footprints of Oxidative Injury Further support for the role of ROS comes from measurement of footprints of oxidative stress. We evaluated the effect of ischemia caused by the ligation of the supplying arteries to sciatic and tibial nerve for 3 hours, followed by reperfusion.127 Reperfusion resulted in an increase in nerve lipid hydroperoxides, greatest at 3 hours, followed by a gradual decline over the next month. Nerve edema and ischemic fiber degeneration consistently became more severe with reperfusion, indicating that oxidative stress impairs the BNB, resulting in edema, and causes ischemic fiber degeneration. Reduced reperfusion was greatest over distal sciatic nerve and mid-tibial nerve at day 7. The most ischemic segment (mid-tibial) of non-reperfused ischemic nerves (duration of ischemia, 3 hours) underwent both edema and ischemic fiber degeneration that was as pronounced as those of other segments after reperfusion, and underwent a smaller increase with reperfusion, suggesting that ischemia alone can also cause ischemic fiber degeneration and edema. The type of fiber degeneration was that of axonal degeneration. These indices indicate the presence of oxidative stress but do not indicate oxidative injury. Hence, recent focus has shifted to immunohistochemical indices that provide cellular localization and indicate DNA damage. These findings are detailed next.

Neuropathology The neuropathology of ischemic fiber degeneration has been well described.86,127,137 We have related the severity of nerve ischemic fiber degeneration to that of ischemic stress (dose of emboli or number of vessels ligated).86 The severity of fiber degeneration regresses linearly with the severity of ischemia (reduction in nerve blood flow).86 Ischemic demyelination is well recognized, but its frequency and pathophysiology have been largely unknown. Its presence has been emphasized by Nukada and colleagues,137,139 especially in the context of ischemia-reperfusion, with a perivascular distribution and in association with endoneurial edema, intramyelinic edema, and endothelial swelling.139 It has also been noted by others.62,127 Nukada and Dyck138 provided convincing evidence, based on a serial-section reconstruction of 55 single myelinated fibers showing ischemic fiber

522

Neurobiology of the Peripheral Nervous System

degeneration, that demyelination can be secondary to changes in axonal caliber. However, Schwann cell and myelin pathology, with intramyelinic edema, suggests a separate effect of oxidative injury associated with ischemia, and especially with ischemia followed by reperfusion, on the Schwann cell.139 Ischemic demyelination takes on greater importance with evidence of ischemia-reperfusion–induced Schwann cell oxidative injury, activated caspase-3 expression, and apoptosis. This mechanism of injury is especially pronounced in the chronically diabetic nerve, in which oxidative stress is already increased. Hyperglycemic nerves, although resistant to ischemic conduction failure,104 are, if ischemic stress is maintained, excessively susceptible to ischemic fiber degeneration.135,136 The resistance to ischemic conduction failure is due to a combination of low metabolic rate and increased energy substrate stores, so that anaerobic metabolism could provide sufficient ATP to maintain conduction.104 In an elegant study of rat peripheral nerve, intracellular H⫹ and calcium were measured.183 Nerve was rendered hypoxic for 3 to 6 hours in a 25- or 5-mM glucose medium. These authors found that hyperglycemia resulted in intracellular acidosis and an increase in intracellular glucose. Similar intracellular calcium levels were found in normal and highglucose media. The investigators concluded that ischemia increased intracellular calcium, and hyperglycemia compounded the ischemic insult with intracellular acidosis. A high-glucose medium causes apoptosis of DRG neurons; in vivo confirmation has been reported.155 Nukada made the observation that diabetic nerve is unusually susceptible to ischemia, but only after a duration of diabetes of at least 4 months.135,136 This mechanism is especially pronounced if diabetic nerves are subjected to ischemia-reperfusion. Figure 23–9 shows results in rats with EDN (duration of diabetes, 4 months), and age- and gender-matched controls, subjected to nerve ischemia for 3 hours followed by reperfusion. There was a lack of tibial nerve pathology in controls subjected to ischemic stress, whereas corresponding diabetic nerve had severe axonal pathology. The lack of caspase-3 activation in controls but prominent Schwann cell caspase-3 activation in diabetic nerve indicates that the Schwann cell is committed to the efferent limb of the apoptotic pathway. Finally, there was prominent Schwann cell expression of 8-hydroxydeoxyguanosine and 4-HNE in diabetic Schwann cells, indicative of oxidative DNA damage. These observations taken together lead to the appreciation that ischemic demyelination is different pathophysiologically from ischemic fiber degeneration. There is some evidence that demyelination occurs secondary to the inflammatory response with injury from cytokines and ROS in concert with axon-glia interactions, including changes in axonal caliber. Some of these changes are described next. Histochemical detection of oxidative stress in ischemic nerve was undertaken by Anderson et al.2 These workers labeled carbonyl compounds by applying naphthoic acid hydrazide (NAH) and Fast Blue B (FBB) and studied sciatic,

tibial, and peroneal nerves. NAH-FBB reactivity was not seen in ischemic nerve but appeared in vessels of nerves subjected to reperfusion following ischemia. Positively stained epi-, peri-, and endoneurial vessels were invariably observed after 2 hours of reperfusion at all levels examined. Pretreatment with ␣-tocopherol prevented NAH-FBB staining. In a subsequent study, He et al.56 measured carbonyl formation using a sensitive enzymelinked immunosorbent assay. Protein carbonyl content was unaffected by ischemia alone, but increased by 55% after 12 to 18 hours of reperfusion, correlating with the onset of nerve pathology. Morphometry in endoneurial vessels showed an increase in endothelial cell area that was prevented in ␣-tocopherol–treated reperfused nerves. Ischemia-reperfusion results in ICAM-1 expression on endothelial cells, followed by endoneurial edema and the inflammatory response with demyelination.140

Inflammatory Response Recently Nukada et al.140 studied the inflammatory response following total ischemia of 5 hours’ duration and varying durations of reperfusion, using immunohistochemistry, light, and electron microscopy. They evaluated the role of acute inflammation in the development of demyelination in reperfused rat sciatic, tibial, and peroneal nerves. They studied the time course of expression of ICAM-1, polymorphonuclear neutrophils, and macrophage migration (Fig. 23–10). After 18 hours of reperfusion, there was maximal ICAM-1 expression on endoneurial vessels, and polymorphonuclear neutrophil accumulation was then prominent, reaching a peak 24 hours after reperfusion (Fig. 23–10). Endoneurial mononuclear macrophages increased nearly fourfold after 48 to 72 hours of reperfusion. Macrophages were observed invading Schwann cells and myelinated lamellae, with associated demyelination. They concluded that the macrophage was most closely associated with demyelination after reperfusion.

Role of Cytokines Associated with the inflammatory response is the activation of cytokines, and much of the oxidative injury is due to the role of cytokines. Recently, we demonstrated that ischemia induced by ischemia-reperfusion114,127 results in an increase in gene expression of TNF-␣, interleukin (IL)–1␤, IL-6, and IL-10 (Fig. 23–11). In the studies of ischemia-reperfusion of nerve, we found that TNF-␣ and IL-1␤ expression occurred in ischemic-reperfused nerves but not following ischemia alone. Furthermore, their expression required ischemia of sufficient severity to cause ischemic fiber degeneration and demyelination. These cytokines, derived mainly from macrophages, reflect a hierarchy with TNF-␣ and IL-1 assuming particular importance.46 Cytokine message and protein expression in rat peripheral nerve is reported to be

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

Control

523

Diabetic

FIGURE 23–9 Experimental diabetic neuropathy model (duration 12 months) subjected to moderate nerve ischemia for 3 hours followed by 7 days of reperfusion. Note the lack of tibial nerve pathology in controls subjected to the same ischemic stress (top left), whereas corresponding diabetic nerve has severe axonal pathology. Top right, Lack of caspase-3 activation in controls but prominent Schwann cell caspase-3 activation in diabetic nerve. Bottom, Prominent Schwann cell expression of 8-hydroxydeoxyguanosine (left) and 4-hydroxynonenal (right) in diabetic Schwann cell, indicative of oxidative damage. See Color Plate

concordant (e.g., Wagner and Myers185). Some cytokines, especially TNF-␣, will activate adhesion molecules and leukocyte adherence.121 These cytokines are functionally interactive but have specific functions. For instance, IL-10 is neuroprotective.63 Administering a single dose of IL-10 at the site of a chronic constriction injury to nerve attenuated the inflammatory response and hyperalgesia.183 In IL-10–treated nerves, macrophage recruitment and cell profiles immunoreactive for TNF-␣ were also reduced. IL-6 is reported to be deleterious but is also, at least in part, neuroprotective in cerebral ischemia.98 Subperineurial administration of TNF-␣ will result in both demyelination

and ischemic fiber degeneration, and hyperalgesia.184 Schwann cells play an important role in the inflammatory response; they express major histocompatibility complex (MHC) class I and II, TNF-␣, and IL-6 mRNA.124

Genetic Susceptibility There has been no attention to the importance of genetic susceptibility. Ischemic susceptibility may be dependent on MHC class of the mouse, which influences host inflammatory response. MHC class II region and CD4⫹ T cells may synergize with other genes to preferentially produce high

524

Neurobiology of the Peripheral Nervous System

release of cytochrome c,91 which binds to Apaf-1, triggering activation of caspase-9,167 which activates downstream caspases, especially caspase-3, the principal effector in the apoptotic pathway in neurons and support cells.79 This hypothesis takes on special importance with the availability of potent antioxidants, anticaspases, and anticytokine treatment. If the inflammatory response is important, the severity of demyelination should be greater with ischemiareperfusion than ischemia alone.

Neuroprotection of Ischemia-Reperfusion Injury

FIGURE 23–10 Time course of densities (mean ⫾ 6 SD) of intercellular adhesion molecule-1 (ICAM-1)–positive vessels (top), polymorphonuclear neutrophils expressed by HIS48 (middle), and mononuclear macrophages visualized by 1C7 (bottom) in the endoneurium of ischemic-reperfused rat sciatic nerve at the thigh level.

levels of pro-inflammatory cytokines such as TNF-␣.142 Differences in genetic background with different cytokine responses could result in different susceptibilities to ischemia-reperfusion–related oxidative injury.

Molecular Pathogenesis On the efferent limb, caspases have been demonstrated to be important in delayed neuronal death in cerebral ischemia.24 Ischemia is proposed as being especially damaging to mitochondria-rich cells such as Schwann cells, causing mitochondrial membrane depolarization with

Mitsui et al.116 undertook a detailed study of hypothermic neuroprotection of ischemia-reperfusion injury of nerve. They compared an ischemia duration of 3 versus 5 hours and temperatures of 37°, 32°, and 28° C. Electrophysiologic, behavioral, and histologic end points (edema grade and ischemic fiber degeneration grade) were evaluated. The groups treated at 37° C underwent marked fiber degeneration, associated with a reduction in action potential and impairment in behavioral score. The groups treated at 28° C (for both 3 and 5 hours) showed significantly less (P ⫽ .01; analysis of variance, Bonferoni post hoc test) reperfusion injury for all indices (behavioral score, electrophysiology, and neuropathology), and the groups treated at 32° C had scores intermediate between the groups treated at 36° to 37° C and at 28° C. Figure 23–12 shows the graded neuroprotection as a function of degree of hypothermia. Our results showed that cooling the limbs dramatically protects the peripheral nerve from ischemia-reperfusion injury. Antioxidant pretreatment will prevent histochemical evidence of oxidative injury. ␣-Tocopherol pretreatment resulted in less carbonyl staining (as an index of oxidative injury) and the area of endothelial cells was also reduced, indicating amelioration of endothelial swelling.2 Pretreatment with the xanthine oxidase inhibitor allopurinol prevented this protein carbonyl formation.56 In a more formal study, Mitsui et al.115 evaluated whether racemic ␣-lipoic acid, a potent antioxidant, would protect peripheral nerve from reperfusion injury. ␣-Lipoic acid was given intraperitoneally daily for 3 days both pre- and postsurgery. Distal sensory conduction (amplitude of sensory action potential and sensory conduction velocity of digital nerve) was significantly improved in the 3-hour ischemia group treated with ␣-lipoic acid (P ⬍ .05) but not if the duration of ischemia was longer (5-hour ischemia). Figure 23–13 shows the effects of lipoic acid–mediated neuroprotection.

N-METHYL-D-ASPARTATE AND GLUTAMATE RECEPTORS Most of the interest in N-methyl-D-aspartate (NMDA) receptors has revolved around their contribution to the pain state. NMDA receptors are present in several sites.

525

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

A: TNF-α 3

Sciatic Nerve

Arbitrary units

Tibial Nerve

β

2

β

α α

α

1

α

0 SHAM

3I

3I3R

3I12R

3I24R

24I

3I48R

3I7dR

B: IL1-β 30

Sciatic Nerve Tibial Nerve

20

γ

10

α

α

α

β

α

0 SHAM

3I

3I3R

3I12R

3I24R

24I

3I48R

3I7dR

FIGURE 23–11 Time course of messenger RNA (mRNA) expression of tumor necrosis factor-␣ (TNF-␣) and interleukin-1␤ (IL-1␤). mRNA expression of TNF-␣ peaked at 24 hours of reperfusion and remained elevated even at 7 days. In contrast, IL-1␤ peaked at 12 hours of reperfusion. Tibial versus sciatic nerves: *, P ⬍ .05; ***, P ⬍ .001; tibial nerves versus sham: ␣, P ⬍ .05; ␤, P ⬍ .01; ␥, P ⬍ .001. (From Mitsui, Y., Okamoto, K., Martin, D. P., et al.: The expression of proinflammatory cytokine mRNA in the sciatic-tibial nerve of ischemia-reperfusion injury. Brain Res. 844:192, 1999, with permission.)

These include the nerve root entry zone, primary afferent endings, and sympathetic efferents.23 These receptors appear to be vitally involved in the pathophysiology of painfulness.38,194 Three closely interlinked sites of sensitization (nerve axon microenvironment, DRG, and nerve root entry zone) may together be operative in producing hyperalgesia. There is interest in the effect of inflammation on glutamate receptors. Carlton and Coggeshall23 demonstrated that the number of peripheral primary afferent axons expressing NMDA, ␣-aminohydroxy-5-methyl4-isoxazolepropionic acid (AMPA), or kainate inotropic glutamate receptors were significantly increased during complete Freund’s adjuvant–induced inflammation of the hind paw. This suggests an increased role for glutamate in the pain that accompanies inflammation. Glutamate may also regulate sympathetic efferents locally because inotropic receptors are expressed by peripheral postganglionic sympathetic fibers.22 In a subsequent paper, Carlton and Coggeshall23 demonstrated that, following inflammation, there was a 2-fold increase in postgan-

glionic adrenergic axons expressing NMDA receptors and a 10-fold increase in axons expressing AMPA or kainate receptors. These data suggest that postganglionic activity may be enhanced by glutamate receptor activation during inflammation. Increased activity in postganglionic fibers could lead to an increased release of norepinephrine and other substances in postganglionic efferents, such as prostaglandins, which in turn could enhance nociceptor activity. This change in glutamate receptor organization offers a possible site of pharmacologic intervention for the maladaptive symptoms that often arise following peripheral inflammation. The chronic constriction model results in increased norepinephrine, neuropeptide Y (NPY), and calcitonin gene–related peptide (CGRP) at the nerve lesion. Nerve blood flow is increased within the lesion but reduced at the lesion edge by more than 50%, a deficit that is sufficient to cause fiber degeneration.86,166 This increase in CGRP in the lesion is associated with reduced pain threshold. Second, we found a large increase of NPY and CGRP by radioimmunoassay and immunocytochemistry

526

Neurobiology of the Peripheral Nervous System

37˚C

32˚C

28˚C

FIGURE 23–12 Representative transverse sections of the mid-tibial nerve after 3 hours of ischemia-reperfusion. Most fibers degenerated when treated at 37° C; fewer degenerated fibers were seen after treatment at 32° C. Most fibers were normal after treatment at 28° C. Bar: 100 ␮m. (From Mitsui, Y., Schmelzer, J. D., Zollman, P. J., et al.: Hypothermic neuroprotection of peripheral nerve of rats from ischaemia-reperfusion injury. Brain 122:161, 1999, with permission.)

Isch

Isch- LA

FIGURE 23–13 Light microscopic findings of transverse sections of the mid-tibial nerve after 3 hours of ischemia followed by 1-week reperfusion. Degenerated fibers were predominant for the nontreatment group, and a small number of degenerated fibers were seen in the lipoic acid group. Bar: 100 ␮m. Isch ⫽ ischemiareperfusion; Isch-LA ⫽ ischemia-reperfusion ⫹ lipoic acid treatment. (From Mitsui, Y., Schmelzer, J. D., Zollman, P. J., et al.: Alphalipoic acid provides neuroprotection from ischemia-reperfusion injury of peripheral nerve. J. Neurol. Sci. 163:11, 1999, with permission.)

Oxidative Stress and Excitatory Neurotoxins in Neuropathy

in ipsilateral L5 DRGs but not in sympathetic ganglia. For instance, central sensitization requires peripheral inputs and can be prevented if peripheral input is blocked or absent.38 Rats with a chronic constriction injury to the sciatic nerve have been found to have small- to medium-sized, pyknotic, and hyperchromatic neurons (“dark neurons”) in spinal dorsal horn laminae I through III.125 These authors proposed that these hyperchromatic neurons are produced by an excitotoxic insult involving NMDA receptor activation subsequent to ectopic nociceptor discharge, and that at least some dark neurons are inhibitory interneurons whose functional impairment or death contributes to a central state of hyperexcitability that underlies neuropathic hyperalgesia and allodynia. NMDA antagonists will prevent or reverse allodynia.168,177 Pain models include the chronic constriction nerve trunk model of Bennett and Xie,9 the root constriction model of Kim and Chung,87 and the paw formalin injection model. A sustained afferent barrage from small nerve fibers is associated with spinal release of glutamate, prostaglandins, and aspartate,107 and results in a hyperalgesic state. There are close interactions among spinal NMDA, neurokinin-1 (NK), and ␣ adrenoreceptors. There is also NO modulation and mediation by the spinal release of cyclooxygenase products.195 The hyperalgesic state can be prevented by antinociceptive doses of cyclooxygenase inhibitors,108 the ␣2 adrenoceptor agonist clonidine, or NMDA antagonists (such as MK-801). Intrathecal clonidine may act to diminish sympathetic outflow, whereas MK-801 blocks the NMDA receptor. The two separate mechanisms may account for the powerful synergy observed by Lee and Yaksh.97 This hyperalgesic component appears to be initiated by the activation of a spinal NMDA receptor that, through the generation of NO, leads to the observed augmented processing of afferent input and the associated hyperalgesic component of the subsequent pain behavior.107 Hyperalgesia is prevented by pretreatment with agents known to block afferent input (local anesthetics) or C-fiber transmitter release (opiates) or to act at one of several links to block a complex spinal cascade involving the NMDA receptor, NOS, and cyclooxygenase.193 Neuropathic animals that have thermal hyperalgesia had an ipsilateral decrease in substance P (SP) staining density without an accompanying change in CGRP staining density. MK-801–treated animals showed a dose-dependent attenuation of the thermal hyperalgesia, and the expected ipsilateral decrease in SP was prevented. MK-801 treatment in naive rats caused a global increase in both SP and CGRP staining in the dorsal horn. The results suggest a functional interaction between excitatory amino acids (EAAs) and SP, with activation of NMDA receptors mediating depletion of SP in neuropathic animals. It is suggested that SP-containing interneurons are a target of the EAAs in the dorsal horn.47

527

Although NO-NMDA interactions can result in hyperalgesia, NO also produces a dose-dependent antinociceptive response in diabetic mice.74 The antinociceptive effects were significantly antagonized by subcutaneous administration of naltrindole, a selective ␦ opioid receptor antagonist, suggesting the involvement of activation of ␦ opioid receptors. The hyperalgesic state in experimental diabetic rats is associated with increased spinal release,76 reduced spinal levels of SP, and a significant increase in the number of binding sites for SP in dorsal spinal cord.75 RP-67580, a specific tachykinin NK1 receptor antagonist, relieves chronic hyperalgesia in diabetic rats.31

REFERENCES 1. Ahlskog, J. E., Uitti, R. J., Low, P. A., et al.: No evidence for systemic oxidant stress in Parkinson’s or Alzheimer’s disease. Mov. Disord. 10:566, 1995. 2. Anderson, G. M., Nukada, H., and McMorran, P. D.: Carbonyl histochemistry in rat reperfusion nerve injury. Brain Res. 772:156, 1997. 3. Babiy, A. V., Gebicki, J. M., Sullivan, D. R., and Willey, K.: Increased oxidizability of plasma lipoproteins in diabetic patients can be decreased by probucol therapy and is not due to glycation. Biochem. Pharmacol. 43:995, 1992. 4. Baker, M. S., and Gebicki, J. M.: The effect of pH on the conversion of superoxide to hydroxyl free radicals. Arch. Biochem. Biophys. 234:258, 1984. 5. Barouch, F. C., Miyamoto, K., Allport, J. R., et al.: Integrin-mediated neutrophil adhesion and retinal leukostasis in diabetes. Invest. Ophthalmol. Vis. Sci. 41:1153, 2000. 6. Battelli, M. G., Corte, E. D., and Stirpe, F.: Xanthine oxidase type d (dehydrogenase) in the intestine and other organs of the rat. Biochem. J. 126:747, 1972. 7. Baynes, J. W.: Role of oxidative stress in the development of complications in diabetes. Diabetes 40:405, 1991. 8. Benarroch, E. E., Opfer-Gehrking, T. L., and Low, P. A.: Use of the photoplethysmographic technique to analyze the Valsalva maneuver in normal man. Muscle Nerve 14:1165, 1991. 9. Bennett, G. J. and Xie, Y. K.: A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain 33:87, 1988. 10. Bhimji, S., Godin, D. V., and McNeill, J. H.: Insulin reversal of biochemical changes in hearts of diabetic rats. Am. J. Physiol. 251:H670, 1986. 11. Boveris, A., Cardenas, E., and Stoppani, A. O.: Role of ubiquinone in the mitochondrial generation of hydrogen peroxide. Biochem. J. 156:435, 1976. 12. Boveris, A., and Chance, B.: The mitochondrial generation of hydrogen peroxide: general properties and effect of hyperbaric oxygen. Biochem. J. 134:707, 1973. 13. Bravenboer, B., Kappelle, A. C., Hamers, F. P., et al.: Potential use of glutathione for the prevention and treatment of diabetic neuropathy in the streptozotocin-induced diabetic rat. Diabetologia 35:813, 1992.

528

Neurobiology of the Peripheral Nervous System

14. Bucala, R., Tracey, K. J., and Cerami, A.: Advanced glycosylation products quench nitric oxide and mediate defective endothelium-dependent vasodilatation in experimental diabetes. J. Clin. Invest. 87:432, 1991. 15. Burton, N. K., and Aherne, G. W.: Sensitive measurement of glutathione using isocratic high-performance liquid chromatography with fluorescence detection. J. Chromatogr. 382:253, 1986. 16. Busse, E., Zimmer, G., Schopohl, B., and Kornhuber, B.: Influence of alpha-lipoic acid on intracellular glutathione in vitro and in vivo. Arzneimittelforschung 42:829, 1992. 17. Cameron, N. E., and Cotter, M. A.: Neurovascular dysfunction in diabetic rats: potential contribution of autoxidation and free radicals examined using transition metal chelating agents. J. Clin. Invest. 96:1159, 1995. 18. Cameron, N. E., Cotter, M. A., Archibald, V., et al.: Antioxidant and pro-oxidant effects on nerve conduction velocity, endoneurial blood flow and oxygen tension in non-diabetic and streptozotocin-diabetic rats. Diabetologia 37:449, 1994. 19. Cameron, N. E., Cotter, M. A., Jack, A. M., et al.: Protein kinase C effects on nerve function, perfusion, Na( ⫹ ), K( ⫹ )ATPase activity and glutathione content in diabetic rats. Diabetologia 42:1120, 1999. 20. Cameron, N. E., Cotter, M. A., and Low, P. A.: Nerve blood flow in early experimental diabetes in rats: relation to conduction deficits. Am. J. Physiol. 261:E1, 1991. 21. Cameron, N. E., Cotter, M. A., and Maxfield, E. K.: Antioxidant treatment prevents the development of peripheral nerve dysfunction in streptozotocin-diabetic rats. Diabetologia 36:299, 1993. 22. Carlton, S. M., Chung, K., Ding, Z., and Coggeshall, R. E.: Glutamate receptors on postganglionic sympathetic axons. Neuroscience 83:601, 1998. 23. Carlton, S. M., and Coggeshall, R. E.: Inflammation-induced changes in peripheral glutamate receptor populations. Brain Res. 820:63, 1999. 24. Chen, J., Nagayama, T., Jin, K., et al.: Induction of caspase3-like protease may mediate delayed neuronal death in the hippocampus after transient cerebral ischemia. J. Neurosci. 18:4914, 1998. 25. Cheung, J. Y., Bonventre, J. V., Malis, C. D., and Leaf, A.: Calcium and ischemic injury. N. Engl. J. Med. 314:1670, 1986. 26. Cohen, G., and Hochstein, P.: Glutathione peroxidase: the primary agent for the elimination of hydrogen peroxide in erythrocytes. Biochemistry 2:1420, 1963. 27. Collier, A., Leach, J. P., McLellan, A., et al.: Plasma endothelin-like immunoreactivity levels in IDM patients with microalbuminuria. Diabetes Care 15:1038, 1992. 28. Cotter, M. A., and Cameron, N. E.: Correction of impaired sciatic nerve perfusion by desferrioxamine in diabetic rats. Diabetologia 38:898, 1995. 29. Cotter, M. A., and Cameron, N. E.: Neuroprotective effects of carvedilol in diabetic rats: prevention of defective peripheral nerve perfusion and conduction velocity. Naunyn Schmiedebergs Arch. Pharmacol. 351:630, 1995. 30. Cotter, M. A., Love, A., Watt, M. J., et al.: Effects of natural free radical scavengers on peripheral nerve and neurovascular function in diabetic rats. Diabetologia 38:1285, 1995.

31. Courteix, C., Lavarenne, J., and Eschalier, A.: RP-67580, a specific tachykinin NK1 receptor antagonist, relieves chronic hyperalgesia in diabetic rats. Eur. J. Pharmacol. 241:267, 1993. 32. Crouch, R., Kimsey, G., Priest, D. G., et al.: Effect of streptozotocin on erythrocyte and retinal superoxide dismutase. Diabetologia 15:53, 1978. 33. Day, T. J., Lagerlund, T. D., and Low, P. A.: Analysis of H2 clearance curves used to measure blood flow in rat sciatic nerve. J. Physiol. (Lond.) 414:35, 1989. 34. Day, T. J., Schmelzer, J. D., and Low, P. A.: Aortic occlusion and reperfusion and conduction, blood flow, and the bloodnerve barrier of rat sciatic nerve. Exp. Neurol. 103:173, 1989. 35. de Medinaceli, L., Wyatt, R. J., and Freed, W. J.: Peripheral nerve reconnection: mechanical, thermal, and ionic conditions that promote the return of function. Exp. Neurol. 81:469, 1983. 36. Del Maestro, R. F., Bjork, J., and Arfors, K. E.: Increase in microvascular permeability induced by enzymatically generated free radicals. II. Role of superoxide anion radical, hydrogen peroxide, and hydroxyl radical. Microvasc. Res. 22:255, 1981. 37. Delaney, C. L., Russell, J. W., Cheng, H. L., and Feldman, E. L.: Insulin-like growth factor-I and over-expression of Bcl-xL prevent glucose-mediated apoptosis in Schwann cells. J. Neuropathol. Exp. Neurol. 60:147, 2001. 38. Dickenson, A. H.: NMDA receptor antagonists as analgesics. In Fields, H. L., and Liebeskind, J. C. (eds.): Progress in Pain Research and Management, Vol. 1. Seattle, IASP Press, p. 173, 1994. 39. Dietrich, R. B., and Bradley, W. G. Jr.: Iron accumulation in the basal ganglia following severe ischemic-anoxic insults in children. Radiology 168:203, 1988. 40. Dimpfel, W., Spuler, M., Pierau, F. K., and Ulrich, H.: Thioctic acid induces dose-dependent sprouting of neurites in cultured rat neuroblastoma cells. Dev. Pharmacol. Ther. 14:193, 1990. 41. Duchen, L. W., Anjorin, A., Watkins, P. J., and Mackay, J. D.: Pathology of autonomic neuropathy in diabetes mellitus. Ann. Intern. Med. 92:301, 1980. 42. Dyck, P. J., Karnes, J., O’Brien, P., et al.: Spatial pattern of nerve fiber abnormality indicative of pathologic mechanisms. Am. J. Pathol. 117:225, 1984. 43. Dyck, P. J., Low, P. A., Windebank, A. J., et al.: Plasma exchange in polyneuropathy associated with monoclonal gammopathy of undetermined significance. N. Engl. J. Med. 325:1482, 1991. 44. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999. 45. Feldman, E. L., Russell, J. W., Sullivan, K. A., and Golovoy, D.: New insights into the pathogenesis of diabetic neuropathy. Curr. Opin. Neurol. 12:553, 1999. 46. Firestein, G. S., and Zvaifler, N. J.: Anticytokine therapy in rheumatoid arthritis. N. Engl. J. Med. 337:195, 1997. 47. Garrison, C. J., Dougherty, P. M., and Carlton, S. M.: Quantitative analysis of substance P and calcitonin generelated peptide immunohistochemical staining in the dorsal horn of neuropathic MK-801-treated rats. Brain Res. 607:205, 1993.

Oxidative Stress and Excitatory Neurotoxins in Neuropathy 48. Gaudet, R. J., and Levine, L.: Transient cerebral ischemia and brain prostaglandins. Biochem. Biophys. Res. Commun. 86:893, 1979. 49. Gilmore, N., Vane, J. R., and Wyllie, J. H.: Prostaglandins released by the spleen. Nature 218:1135, 1968. 50. Godin, D. V., Wohaieb, S. A., Garnett, M. E., and Goumeniouk, A. D.: Antioxidant enzyme alterations in experimental and clinical diabetes. Mol. Cell. Biochem. 84:223, 1988. 51. Gorman, R. R., Bunting, S., and Miller, O. V.: Modulation of human platelet adenylate cyclase by prostacyclin (PGX). Prostaglandins 13:377, 1977. 52. Greenberg, R.: The neuronal origin of prostaglandin released from the rabbit portal vein in response to electrical stimulation. Br. J. Pharmacol. 63:79, 1978. 53. Greene, D. A., Cao, X., Van Huysen, C., and Obrosova, I.: Effects of DL-alpha-lipoic acid (LA) on diabetic nerve function, bioenergetics and antioxidant defense. In Abstracts of the Workshop on Oxidative Stress in Diabetes and Its Complications, Santa Barbara, CA, February 5–8, p. 80, 1998. 54. Greene, D. A., Stevens, M. J., Obrosova, I., and Feldman, E. L.: Glucose-induced oxidative stress and programmed cell death in diabetic neuropathy. Eur. J. Pharmacol. 375:217, 1999. 55. Greene, R. M., Winkelmann, R. K., Opfer-Gehrking, T. L., and Low, P. A.: Sweating patterns in atopic dermatitis patients. Arch. Dermatol. Res. 281:373, 1989. 56. He, K., Nukada, H., McMorran, P. D., and Murphy, M. P.: Protein carbonyl formation and tyrosine nitration as markers of oxidative damage during ischaemia-reperfusion injury to rat sciatic nerve. Neuroscience 94:909, 1999. 57. Hearse, D. J., Manning, A. S., Downey, J. M., and Yellon, D. M.: Xanthine oxidase: a critical mediator of myocardial injury during ischemia and reperfusion? Acta Physiol. Scand. Suppl. 548:65, 1986. 58. Hellweg, R., and Hartung, H. D.: Endogenous levels of nerve growth factor (NGF) are altered in experimental diabetes mellitus: a possible role for NGF in the pathogenesis of diabetic neuropathy. J. Neurosci. Res. 26:258, 1990. 59. Hengartner, M. O.: The biochemistry of apoptosis. Nature 407:770, 2000. 60. Herbst, U., Toborek, M., Kaiser, S., et al.: 4-Hydroxynonenal induces dysfunction and apoptosis of cultured endothelial cells. J. Cell. Physiol. 181:295, 1999. 61. Hermenegildo, C., Raya, A., Roma, J., and Romero, F. J.: Decreased glutathione peroxidase activity in sciatic nerve of alloxan-induced diabetic mice and its correlation with blood glucose levels. Neurochem. Res. 18:893, 1993. 62. Hess, K., Eames, R. A., Darveniza, P., and Gilliatt, R. W.: Acute ischaemic neuropathy in the rabbit. J. Neurol. Sci. 44:19, 1979. 63. Ho, A. S., and Moore, K. W.: Interleukin-10 and its receptor. Ther. Immunol. 1:173, 1994. 64. Hofmann, M. A., Schiekofer, S., Isermann, B., et al.: Peripheral blood mononuclear cells isolated from patients with diabetic nephropathy show increased activation of the oxidative-stress sensitive transcription factor NF-kappaB. Diabetologia 42:222, 1999. 65. Humphries, K. M., and Szweda, L. I.: Selective inactivation of alpha-ketoglutarate dehydrogenase and pyruvate

66.

67.

68.

69. 70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

529

dehydrogenase: reaction of lipoic acid with 4-hydroxy2-nonenal. Biochemistry 37:15835, 1998. Hunt, J. V., and Wolff, S. P.: Oxidative glycation and free radical production: a causal mechanism of diabetic complications. Free Radic. Res. Commun. 12–13:115, 1991. Jarasch, E. D., Grund, C., Bruder, G., et al.: Localization of xanthine oxidase in mammary-gland epithelium and capillary endothelium. Cell 25:67, 1981. Jennings, P. E., Chirico, S., Lunec, J., and Barnett, A. H.: Vitamin C metabolites and microangiopathy in diabetes mellitus. Diabetes Res. 6:151, 1987. Jennings, R. B., and Reimer, K. A.: Lethal myocardial ischemic injury. Am. J. Pathol. 102:241, 1981. Ji, C., Amarnath, V., Pietenpol, J. A., and Marnett, L. J.: 4-Hydroxynonenal induces apoptosis via caspase-3 activation and cytochrome c release. Chem. Res. Toxicol. 14:1090, 2001. Jiang, C., Wang, Z., Ganther, H., and Lu, J.: Caspases as key executors of methyl selenium-induced apoptosis (anoikis) of DU-145 prostate cancer cells. Cancer Res. 61:3062, 2001. Kaji, H., Kurasaki, M., Ito, K., et al.: Increased lipoperoxide value and glutathione peroxidase activity in blood plasma of type 2 (non-insulin dependent) diabetic women. Klin. Wochenschr. 63:765, 1985. Kalyanaraman, B., Darley-Usmar, V. M., Wood, J., et al.: Synergistic interaction between the probucol phenoxyl radical and ascorbic acid in inhibiting the oxidation of low density lipoprotein. J. Biol. Chem. 267:6789, 1992. Kamei, J., Iwamoto, Y., Misawa, M., et al.: Antinociceptive effect of L-arginine in diabetic mice. Eur. J. Pharmacol. 254:113, 1994. Kamei, J., Ogawa, M., and Kasuya, Y.: Development of supersensitivity to substance P in the spinal cord of the streptozotocin-induced diabetic rats. Pharmacol. Biochem. Behav. 35:473, 1990. Kamei, J., Ogawa, Y., Ohhashi, Y., and Kasuya, Y.: Alterations in the potassium-evoked release of substance P from the spinal cord of streptozotocin-induced diabetic rats in vitro. Gen. Pharmacol. 22:1093, 1991. Karasu, C., Dewhurst, M., Stevens, E. J., and Tomlinson, D. R.: Effects of anti-oxidant treatment on sciatic nerve dysfunction in streptozotocin-diabetic rats: comparison with essential fatty acids. Diabetologia 38:129, 1995. Karpen, C. W., Cataland, S., O’Dorisio, T. M., and Panganamala, R. V.: Interrelation of platelet vitamin E and thromboxane synthesis in type I diabetes mellitus. Diabetes 33:239, 1984. Keane, R. W., Srinivasan, A., Foster, L. M., et al.: Activation of CPP32 during apoptosis of neurons and astrocytes. J. Neurosci. Res. 48:168, 1997. Keller, J. N., Hanni, K. B., and Markesbery, W. R.: 4-Hydroxynonenal increases neuronal susceptibility to oxidative stress. J. Neurosci. Res. 58:823, 1999. Keller, J. N., Mark, R. J., Bruce, A. J., et al.: 4-Hydroxynonenal, an aldehydic product of membrane lipid peroxidation, impairs glutamate transport and mitochondrial function in synaptosomes. Neuroscience 80:685, 1997. Keller, J. N., and Mattson, M. P.: Roles of lipid peroxidation in modulation of cellular signaling pathways, cell dysfunction, and death in the nervous system. Rev. Neurosci. 9:105–116, 1998.

530

Neurobiology of the Peripheral Nervous System

83. Khechai, F., Ollivier, V., Bridey, F., et al.: Effect of advanced glycation end product-modified albumin on tissue factor expression by monocytes: role of oxidant stress and protein tyrosine kinase activation. Arterioscler. Thromb. Vasc. Biol. 17:2885–2890, 1997. 84. Kihara, M., and Low, P. A.: Impaired vasoreactivity to nitric oxide in experimental diabetic neuropathy. Exp. Neurol. 132:180–185, 1995. 85. Kihara, M., Weerasuriya, A., and Low, P. A.: Endoneurial blood flow in rat sciatic nerve during development. J. Physiol. (Lond.) 439:351–360, 1991. 86. Kihara, M., Zollman, P. J., Schmelzer, J. D., and Low, P. A.: The influence of dose of microspheres on nerve blood flow, electrophysiology, and fiber degeneration of rat peripheral nerve. Muscle Nerve 16:1383, 1993. 87. Kim, S. H., and Chung, J. M.: An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 50:355, 1992. 88. Kishi, M., Tanabe, J., Schmelzer, J. D., and Low, P. A.: Morphometry of dorsal root ganglion in chronic experimental diabetic neuropathy. Diabetes 51:819, 2002. 89. Kishi, Y., Nickander, K. K., Schmelzer, J. D., and Low, P. A.: Gene expression of antioxidant enzymes in experimental diabetic neuropathy. J. Peripher. Nerv. Syst. 5:3, 2000. 90. Kishi, Y., Schmelzer, J. D., Yao, J. K., et al.: ␣-Lipoic acid: effect on glucose uptake, sorbitol pathway, and energy metabolism in experimental diabetic neuropathy. Diabetes 48:2045, 1999. 91. Kluck, R. M., Bossy-Wetzel, E., Green, D. R., and Newmeyer, D. D.: The release of cytochrome c from mitochondria: a primary site for Bcl-2 regulation of apoptosis. Science 275:1132, 1997. 92. Knox, C. A., Kokmen, E., and Dyck, P. J.: Morphometric alteration of rat myelinated fibers with aging. J. Neuropathol. Exp. Neurol. 48:119, 1989. 93. Korthals, J. K., and Wisniewski, H. M.: Peripheral nerve ischemia. Part 1. Experimental model. J. Neurol. Sci. 24:65, 1975. 94. Korthuis, R. J., Granger, D. N., Townsley, M. I., and Taylor, A. E.: The role of oxygen-derived free radicals in ischemiainduced increases in canine skeletal muscle vascular permeability. Circ. Res. 57:599, 1985. 95. Kruman, I., Bruce-Keller, A. J., Bredesen, D., et al.: Evidence that 4-hydroxynonenal mediates oxidative stress-induced neuronal apoptosis. J. Neurosci. 17:5089, 1997. 96. Lee, A. Y., and Chung, S. S.: Contributions of polyol pathway to oxidative stress in diabetic cataract. FASEB J. 13:23, 1999. 97. Lee, Y. W., and Yaksh, T. L.: Analysis of drug interaction between intrathecal clonidine and MK-801 in peripheral neuropathic pain rat model. Anesthesiology 82:741, 1995. 98. Loddick, S. A., Turnbull, A. V., and Rothwell, N. J.: Cerebral interleukin-6 is neuroprotective during permanent focal cerebral ischemia in the rat. J. Cereb. Blood Flow Metab. 18:176, 1998. 99. Loven, D. P., and Oberley, L. W.: Free radicals, insulin action and diabetes. In Oberley, L. W. (ed.) Superoxide Dismutase, Vol. III. Disease States. Boca Raton, FL, CRC Press, p. 151, 1985.

100. Loven, D. P., Schedl, H. P., Oberley, L. W., et al.: Superoxide dismutase activity in the intestine of the streptozotocindiabetic rat. Endocrinology 111:737, 1982. 101. Low, P. A., Lagerlund, T. D., and McManis, P. G.: Nerve blood flow and oxygen delivery in normal, diabetic, and ischemic neuropathy. Int. Rev. Neurobiol. 31:355, 1989. 102. Low, P. A., and Nickander, K. K.: Oxygen free radical effects in sciatic nerve in experimental diabetes. Diabetes 40:873, 1991. 103. Low, P. A., Schmelzer, J. D., Ward, K. K., et al.: Effect of hyperbaric oxygenation on normal and chronic streptozotocin diabetic peripheral nerves. Exp. Neurol. 99:201, 1988. 104. Low, P. A., Ward, K., Schmelzer, J. D., and Brimijoin, S.: Ischemic conduction failure and energy metabolism in experimental diabetic neuropathy. Am. J. Physiol. 248:E457, 1985. 105. Lundborg, G.: Structure and function of the intraneural microvessels as related to trauma, edema formation, and nerve function. J. Bone Joint Surg. Am. 57:938, 1975. 106. MacIntyre, D. E., Pearson, J. D., and Gordon, J. L.: Localisation and stimulation of prostacyclin production in vascular cells. Nature 271:549, 1978. 107. Malmberg, A. B., and Yaksh, T. L.: Spinal nitric oxide synthesis inhibition blocks NMDA-induced thermal hyperalgesia and produces antinociception in the formalin test in rats. Pain 54:291, 1993. 108. Malmberg, A. B., and Yaksh, T. L.: Cyclooxygenase inhibition and the spinal release of prostaglandin E2 and amino acids evoked by paw formalin injection: a microdialysis study in unanesthetized rats. J. Neurosci. 15:2768, 1995. 109. Martinez-Blasco, A., Bosch-Morell, F., Trenor, C., and Romero, F. J.: Experimental diabetic neuropathy: role of oxidative stress and mechanisms involved. Biofactors 8:41, 1998. 110. Mateo, M. C. M., Bustamante, J. B., and Cantalapiedra, M. A. G.: Serum zinc, copper and insulin in diabetes mellitus. Biomedicine 29:56, 1978. 111. Matkovics, B., Varga, S. I., Szabo, L., and Witas, H.: The effect of diabetes on the activities of the peroxide metabolism enzymes. Horm. Metab. Res. 14:77, 1982. 112. Maxfield, E. K., Cameron, N. E., and Cotter, M. A.: Effects of diabetes on reactivity of sciatic vasa nervorum in rats. J. Diabetes Complications 11:47, 1997. 113. McCord, J. M.: Oxygen-derived free radicals in postischemic tissue injury. N. Engl. J. Med. 312:159, 1985. 114. Mitsui, Y., Okamoto, K., Martin, D. P., et al.: The expression of proinflammatory cytokine mRNA in the sciatic-tibial nerve of ischemia-reperfusion injury. Brain Res. 844:192, 1999. 115. Mitsui, Y., Schmelzer, J. D., Zollman, P. J., et al.: Alpha-lipoic acid provides neuroprotection from ischemia-reperfusion injury of peripheral nerve. J. Neurol. Sci. 163:11, 1999. 116. Mitsui, Y., Schmelzer, J. D., Zollman, P. J., et al.: Hypothermic neuroprotection of peripheral nerve of rats from ischaemiareperfusion injury. Brain 122:161, 1999. 117. Mitsui, Y., Schmelzer, J. D., Zollman, P. J., et al.: Hypothermic neuroprotection of peripheral nerve of rats from ischemiareperfusion injury: intraischemic vs. reperfusion hypothermia. Brain Res. 827:63, 1999. 118. Moncada, S.: Eighth Gaddum Memorial Lecture, University of London Institute of Education, December 1980. Biological importance of prostacyclin. Br. J. Pharmacol. 76:3, 1982.

Oxidative Stress and Excitatory Neurotoxins in Neuropathy 119. Moncada, S., Herman, A. G., Higgs, E. A., and Vane, J. R.: Differential formation of prostacyclin (PGX or PGI2) by layers of the arterial wall: an explanation for the antithrombotic properties of vascular endothelium. Thromb. Res. 11:323, 1977. 120. Morigi, M., Angioletti, S., Imberti, B., et al.: Leukocyteendothelial interaction is augmented by high glucose concentrations and hyperglycemia in a NF-kB-dependent fashion. J. Clin. Invest. 101:1905, 1998. 121. Munoz-Fernandez, M. A., and Fresno, M.: The role of tumour necrosis factor, interleukin 6, interferon-gamma and inducible nitric oxide synthase in the development and pathology of the nervous system. Prog. Neurobiol. 56:307, 1998. 122. Murakami, K., Kondo, T., Ohtsuka, Y., et al.: Impairment of glutathione metabolism in erythrocytes from patients with diabetes mellitus. Metabolism 38:753, 1989. 123. Murase, K., Hattori, A., Kohno, M., and Hayashi, K.: Stimulation of nerve growth factor synthesis/secretion in mouse astroglial cells by coenzymes. Biochem. Mol. Biol. Int. 30:615, 1993. 124. Murwani, R., Hodgkinson, S., and Armati, P.: Tumor necrosis factor alpha and interleukin-6 mRNA expression in neonatal Lewis rat Schwann cells and a neonatal rat Schwann cell line following interferon gamma stimulation. J. Neuroimmunol. 71:65, 1996. 125. Nachemson, A. K., and Bennett, G. J.: Does pain damage spinal cord neurons? Transsynaptic degeneration in rat following a surgical incision. Neurosci. Lett. 162:78, 1993. 126. Nagamatsu, M., Nickander, K. K., Schmelzer, J. D., et al.: Lipoic acid improves nerve blood flow, reduces oxidative stress, and improves distal nerve conduction in experimental diabetic neuropathy. Diabetes Care 18:1160, 1995. 127. Nagamatsu, M., Schmelzer, J. D., Zollman, P. J., et al.: Ischemic reperfusion causes lipid peroxidation and fiber degeneration. Muscle Nerve 19:37, 1996. 128. Neely, J. R., and Feuvray, D.: Metabolic products and myocardial ischemia. Am. J. Pathol. 102:282, 1981. 129. Nickander, K. K., McPhee, B. R., Low, P. A., and Tritschler, H. J.: Alpha-lipoic acid: antioxidant potency against lipid peroxidation of neural tissue in vitro and implications for diabetic neuropathy. Free Radic. Biol. Med. 21:631, 1996. 130. Nickander, K. K., Schmelzer, J. D., Rohwer, D. A., and Low, P. A.: Effect of ␣-tocopherol deficiency on indices of oxidative stress in normal and diabetic peripheral nerve. J. Neurol. Sci. 126:6, 1994. 131. Niki, E.: Antioxidants in relation to lipid peroxidation. Chem. Phys. Lipids 44:227, 1987. 132. Nishimura, C., and Kuriyama, K.: Alteration of lipid peroxide and endogenous antioxidant contents of retina of streptozotocin-induced diabetic rats: effect of vitamin A administration. Jpn. J. Pharmacol. 37:365, 1985. 133. Nistico, G., Cirolo, M. R., Fiskin, K., et al.: NGF restores decrease in catalase activity and increases superoxide dismutase and glutathione peroxidase activity in the brain of aged rats. Free Radic. Biol. Med. 12:177, 1992. 134. Noto, R., Alicata, R., and Sfogliano, L. A.: A study of cupremia in a group of elderly diabetics. Acta Diabetol. Lat. 20:81, 1983. 135. Nukada, H.: Increased susceptibility to ischemic damage in streptozocin-diabetic nerve. Diabetes 35:1058, 1986.

531

136. Nukada, H.: Mild ischaemia causes severe pathological changes in experimental diabetic nerve. Muscle Nerve 15:1116, 1992. 137. Nukada, H., and Dyck, P. J.: Microsphere embolization of nerve capillaries and fiber degeneration. Am. J. Pathol. 115:275, 1984. 138. Nukada, H., and Dyck, P. J.: Acute ischemia causes axonal stasis, swelling, attenuation, and secondary demyelination. Ann. Neurol. 22:311, 1987. 139. Nukada, H., and McMorran, P. D.: Perivascular demyelination and intramyelinic oedema in reperfusion nerve injury. J. Anat. 185:259, 1994. 140. Nukada, H., McMorran, P. D., and Shimizu, J.: Acute inflammatory demyelination in reperfusion nerve injury. Ann. Neurol. 47:71, 2000. 141. Obrosova, I. G., Van Huysen, C., Fathallah, L., et al.: Evaluation of alpha(1)-adrenoceptor antagonist on diabetesinduced changes in peripheral nerve function, metabolism, and antioxidative defense. FASEB J. 14:1548, 2000. 142. Oksenberg, J. R.: Immunogenetics and heterogeneity in multiple sclerosis. Ann. Neurol. 40:557, 1996. 143. Olanow, C. W.: An introduction to the free radical hypothesis in Parkinson’s disease. Ann. Neurol. 32:S2, 1992. 144. Oliver, F. J., Menissier-de Murcia, J., Nacci, C., et al.: Resistance to endotoxic shock as a consequence of defective NF-kappaB activation in poly (ADP-ribose) polymerase-1 deficient mice. EMBO J. 18:4446, 1999. 145. Packer, L., Witt, E. H., and Tritschler, H. J.: alpha-Lipoic acid as a biological antioxidant. Free Radic. Biol. Med. 19:227, 1995. 146. Parry, G. J., and Brown, M. J.: Selective fiber vulnerability in acute ischemic neuropathy. Ann. Neurol. 11:147, 1982. 147. Pascoe, M. K., Low, P. A., Windebank, A. J., and Litchy, W. J.: Subacute diabetic proximal neuropathy. Mayo Clin. Proc. 72:1123, 1997. 148. Pedersen, W. A., Cashman, N. R., and Mattson, M. P.: The lipid peroxidation product 4-hydroxynonenal impairs glutamate and glucose transport and choline acetyltransferase activity in NSC-19 motor neuron cells. Exp. Neurol. 155:1, 1999. 149. Pieper, G. M., and Riaz-ul-Haq, G.: Activation of nuclear factor-kappaB in cultured endothelial cells by increased glucose concentration: prevention by calphostin C. J. Cardiovasc. Pharmacol. 30:528, 1997. 150. Pritchard, K. A. Jr., Patel, S. T., Karpen, C. W., et al.: Triglyceride-lowering effect of dietary vitamin E in streptozocin-induced diabetic rats: increased lipoprotein lipase activity in livers of diabetic rats fed high dietary vitamin E. Diabetes 35:278, 1986. 151. Pryor, W. A.: Oxy-radicals and related species: their formation, lifetimes, and reactions. Annu. Rev. Physiol. 48:657, 1986. 152. Rehncrona, S., Westerberg, E., Akesson, B., and Siesjo, B. K.: Brain cortical fatty acids and phospholipids during and following complete and severe incomplete ischemia. J. Neurochem. 38:84, 1982. 153. Romero, F. J., Monsalve, E., Hermenegildo, C., et al.: Oxygen toxicity in the nervous tissue: comparison of the antioxidant defense of rat brain and sciatic nerve. Neurochem. Res. 16:157, 1991.

532

Neurobiology of the Peripheral Nervous System

154. Romero, F. J., Segura-Aguilar, J., Monsalve, E., et al.: Antioxidant and glutathione-related enzymatic activities in rat sciatic nerve. Neurotoxicol. Teratol. 12:603, 1990. 155. Russell, J. W., Hermann, D. N., Sullivan, K. A., and Feldman, E. L.: Hyperglycemia induces programmed cell death in sensory neurons in vivo and in vitro. Endocr. Soc. Abstr. P1–475:218, 1998. 156. Russell, J. W., Sullivan, K. A., Windebank, A. J., et al.: Neurons undergo apoptosis in animal and cell culture models of diabetes. Neurobiol. Dis. 6:347, 1999. 157. Sampath, D., Jackson, G. R., Werrbach-Perez, K., and Perez-Polo, J. R.: Effects of nerve growth factor on glutathione peroxidase and catalase on PC12 cells. J. Neurochem. 62:2476, 1994. 158. Sasaki, H., Schmelzer, J. D., Zollman, P. J., and Low, P. A.: Neuropathology and blood flow of nerve, spinal roots and dorsal root ganglia in longstanding diabetic rats. Acta Neuropathol. (Berl.) 93:118, 1997. 159. Sato, Y., Hotta, N., Sakamoto, N., et al.: Lipid peroxide level in plasma of diabetic patients. Biochem. Med. 21:104, 1979. 160. Schlaepfer, W. W.: Structural alterations of peripheral nerve induced by the calcium ionophore A23187. Brain Res. 136:1, 1977. 161. Schlaepfer, W. W.: Vesicular disruption of myelin simulated by exposure of nerve to calcium ionophore. Nature 265:734, 1977. 162. Schmelzer, J. D., and Low, P. A.: The effect of hyperbaric oxygenation and hypoxia on the blood-nerve barrier. Brain Res. 473:321, 1988. 163. Schmelzer, J. D., Zochodne, D. W., and Low, P. A.: Ischemic and reperfusion injury of rat peripheral nerve. Proc. Natl. Acad. Sci. U. S. A. 86:1639, 1989. 164. Schmidt, A. M., Hori, O., Chen, J. X., et al.: Advanced glycation endproducts interacting with their endothelial receptor induce expression of vascular cell adhesion molecule-1 (VCAM-1) in cultured human endothelial cells and in mice: a potential mechanism for the accelerated vasculopathy of diabetes. J. Clin. Invest. 96:1395, 1995. 165. Shupeck, M., Ward, K. K., Schmelzer, J. D., and Low, P. A.: Comparison of nerve regeneration in vascularized and conventional grafts: nerve electrophysiology, norepinephrine, prostacyclin, malondialdehyde, and the blood-nerve barrier. Brain Res. 493:225, 1989. 166. Sladky, J. T., Greenberg, J. H., and Brown, M. J.: Regional perfusion in normal and ischemic rat sciatic nerves. Ann. Neurol. 17:191, 1985. 167. Slee, E. A., Harte, M. T., Kluck, R. M., et al.: Ordering the cytochrome c-initiated caspase cascade: hierarchical activation of caspases-2, -3, -6, -7, -8, and -10 in a caspase-9-dependent manner. J. Cell Biol. 144:281, 1999. 168. Smith, G. D., Wiseman, J., Harrison, S. M., et al.: Pre treatment with MK-801, a non-competitive NMDA antagonist, prevents development of mechanical hyperalgesia in a rat model of chronic neuropathy, but not in a model of chronic inflammation. Neurosci. Lett. 165:79, 1994. 169. Smith, K. J., Hall, S. M., and Schauf, C. L.: Vesicular demyelination induced by raised intracellular calcium. J. Neurol. Sci. 71:19, 1985.

170. Sodhi, N., Camilleri, M., Camoriano, J. K., et al.: Autonomic function and motility in intestinal pseudoobstruction caused by paraneoplastic syndrome. Dig. Dis. Sci. 34:1937, 1989. 171. Som, S., Basu, S., Mukherjee, D., et al.: Ascorbic acid metabolism in diabetes mellitus. Metabolism 30:572, 1981. 172. Soriano, F. G., Virag, L., Jagtap, P., et al.: Diabetic endothelial dysfunction: the role of poly(ADP-ribose) polymerase activation. Nat. Med. 7:108, 2001. 173. Srinivasan, S., Stevens, M., and Wiley, J. W.: Diabetic peripheral neuropathy: evidence for apoptosis and associated mitochondrial dysfunction. Diabetes 49:1932, 2000. 174. Stevens, M. J., Lattimer, S. A., Kamijo, M., et al.: Osmoticallyinduced nerve taurine depletion and the compatible osmolyte hypothesis in experimental diabetic neuropathy in the rat. Diabetologia 36:608, 1993. 175. Szabo, C., Cuzzocrea, S., Zingarelli, B., et al.: Endothelial dysfunction in a rat model of endotoxic shock: importance of the activation of poly (ADP-ribose) synthetase by peroxynitrite. J. Clin. Invest. 100:723, 1997. 176. Szabo, C., Virag, L., Cuzzocrea, S., et al.: Protection against peroxynitrite-induced fibroblast injury and arthritis development by inhibition of poly(ADP-ribose) synthase. Proc. Natl. Acad. Sci. U. S. A. 95:3867, 1998. 177. Tal, M., and Bennett, G. J.: Dextrorphan relieves neuropathic heat-evoked hyperalgesia. Neurosci. Lett. 151:107, 1993. 178. Tamura, E., and Parry, G. J.: Severe radicular pathology in rats with longstanding diabetes. J. Neurol. Sci. 127:29, 1994. 179. Toyokuni, S., Miyake, N., Hiai, H., et al.: The monoclonal antibody specific for the 4-hydroxy-2-nonenal histidine adduct. FEBS Lett. 359:189, 1995. 180. Tuck, R. R., Schmelzer, J. D., and Low, P. A.: Endoneurial blood flow and oxygen tension in the sciatic nerves of rats with experimental diabetic neuropathy. Brain 107:935, 1984. 181. Turrens, J. F., and Boveris, A.: Generation of superoxide anion by the NADH dehydrogenase of bovine heart mitochondria. Biochem. J. 191:421, 1980. 182. Wachtler, J., Mayer, C., Rucker, F., and Grafe, P.: Glucose availability alters ischaemia-induced changes in intracellular pH and calcium of isolated rat spinal roots. Brain Res. 725:30, 1996. 183. Wagner, R., Janjigian, M., and Myers, R. R.: Antiinflammatory interleukin-10 therapy in CCI neuropathy decreases thermal hyperalgesia, macrophage recruitment, and endoneurial TNF-alpha expression. Pain 74:35, 1998. 184. Wagner, R., and Myers, R. R.: Endoneurial injection of TNF-alpha produces neuropathic pain behaviors. Neuroreport 7:2897, 1996. 185. Wagner, R., and Myers, R. R.: Schwann cells produce tumor necrosis factor alpha: expression in injured and non-injured nerves. Neuroscience 73:625, 1996. 186. Weening, R. S., Wever, R., and Roos, D.: Quantitative aspects of the production of superoxide radicals by phagocytizing human granulocytes. J. Lab. Clin. Med. 85:245, 1975. 187. Weksler, B. B., Marcus, A. J., and Jaffe, E. A.: Synthesis of prostaglandin I2 (prostacyclin) by cultured human and bovine endothelial cells. Proc. Natl. Acad. Sci. U. S. A. 74:3922, 1977. 188. Wohaieb, S. A., and Godin, D. V.: Alterations in free radical tissue-defence mechanisms in streptozotocin-induced diabetes in the rat. Diabetes 36:1014, 1987.

Oxidative Stress and Excitatory Neurotoxins in Neuropathy 189. Wohaieb, S. A., and Godin, D. V.: Alterations in tissue antioxidant systems in the spontaneously diabetic (BB Wistar) rat. Can. J. Physiol. Pharmacol. 65:2191, 1987. 190. Wolff, S. P.: Diabetes mellitus and free radicals. Br. Med. Bull. 49:642, 1993. 191. Wolff, S. P., Jiang, Z. Y., and Hunt, J. V.: Protein glycation and oxidative stress in diabetes mellitus and ageing. Free Radic. Biol. Med. 10:339, 1991. 192. Wong, P. Y., and Cheung, W. Y.: Calmodulin stimulates human platelet phospholipase A2. Biochem. Biophys. Res. Commun. 90:473, 1979. 193. Yaksh, T. L.: The spinal pharmacology of facilitation of afferent processing evoked by high-threshold afferent input of the postinjury pain state. Curr. Opin. Neurol. Neurosurg. 6:250, 1993. 194. Yaksh, T. L., Chaplan, S. R., and Malmberg, A. B.: Future directions in the pharmacological management of hyperalgesic and allodynic pain states: the NMDA receptor. NIDA Res. Monogr. 147:84, 1995. 195. Yaksh, T. L., and Malmberg, A. B.: Spinal actions of NSAIDs in blocking spinally mediated hyperalgesia: the role of cyclooxygenase products. Agents Actions Suppl. 41:89, 1993.

533

196. Yao, J. K., and Low, P. A.: Improvement of endoneurial lipid abnormalities in experimental diabetic neuropathy by oxygen modification. Brain Res. 362:362, 1986. 197. Yoshida, K., Kirokawa, J., Tagami, S., et al.: Weakened cellular scavenging activity against oxidative stress in diabetes mellitus: regulation of glutathione synthesis and efflux. Diabetologia 38:201, 1995. 198. Ziboh, V. A., Maruta, H., Lord, J., et al.: Increased biosynthesis of thromboxane A2 by diabetic platelets. Eur. J. Clin. Invest. 9:223, 1979. 199. Zingarelli, B., Salzman, A. L., and Szabo, C.: Genetic disruption of poly (ADP-ribose) synthetase inhibits the expression of P-selectin and intercellular adhesion molecule-1 in myocardial ischemia/reperfusion injury. Circ. Res. 83:85, 1998. 200. Zochodne, D. W., Verge, V. M., Cheng, C., et al.: Nitric oxide synthase activity and expression in experimental diabetic neuropathy. J. Neuropathol. Exp. Neurol. 59:798, 2000. 201. Zollman, P. J., Awad, O., Schmelzer, J. D., and Low, P. A.: Effect of ischemia and reperfusion in vivo on energy metabolism of rat sciatic-tibial and caudal nerves. Exp. Neurol. 114:315, 1991.

24 Transgenic Models of Nerve Degeneration RUDOLF MARTINI

Transgenic Models Related to Myelin Sheath Organization and Maintenance Mice Homozygously Deficient for P0 Mice Heterozygously Deficient for P0 Mice Overexpressing P0 Mice Deficient in PMP22 Mice and Rats Overexpressing PMP22 Mice Deficient in the Tunnel Protein Cx32 Mice Deficient in Periaxin and Dystroglycan

Mice Deficient in ␤1 and ␤4 Integrins Mice Deficient in MAG Transgenic Models Primarily Related to the Axon Mice Deficient in the Complex Gangliosides GD1a and GT1b Mice Deficient in the Cell Adhesion Molecule L1 Mice Deficient in NF Genes Transgenic Models Related to the Organization of the Node of Ranvier

As a result of its relatively simple organization, the peripheral nerve is an excellent microcosmos to study axon-glia interactions, myelination, and disease mechanisms in myelin disorders. The principle neural partners, the axons and the Schwann cells, interact in a specific way and mutually influence each other. For instance, axonal properties trigger Schwann cell survival, differentiation, and myelination, whereas Schwann cell characteristics modulate axonal features such as axonal diameter, axonal transport, and survival. One instrumental approach to dissect these mechanisms is to establish appropriate co-culture techniques. Among many topics,17 such techniques were helpful in the resolution of Schwann cell behavior during myelination,18 in the investigation of myelin-axon communication,134 and in the investigation of the role of extracellular matrix components,39,40 cell adhesion molecules,45,46,88,89,136 and neurotrophic factors.23 In the intact organism, the influence of axons on Schwann cell phenotype was monitored by surgical cross-anastomosis of different nerve types. Larger caliber axons from sciatic nerve were allowed to grow into bundles of Remak fibers of the sympathetic system in which the sciatic axons induced myelination, reflecting the dependency of Schwann cell fate on axons.4,5,131 Thus it was

Mice Deficient in Caspr Mice Deficient in Contactin and CD9 Mice Lacking Galactocerebroside and Sulfatide Transgenic Models with Peripheral Neuropathy Not Directly Related to Neural Genes Mice Deficient in Porphobilinogen Deaminase

concluded that it is predominantly the axon that determines whether a Schwann cell forms myelin. The reciprocal influence—the effect of Schwann cells on axonal properties—has also been investigated by transplantation studies. Such experiments revealed that mutant Schwann cells can modify axonal caliber, phosphorylation of neurofilaments, and slow axonal transport.3,29,102,103 Another and more modern method to understand axon–Schwann cell interaction and myelination was to create mutants deficient in or overexpressing distinct myelin components. By such approaches, the function of several molecular players involved in myelin formation could be identified in the intact organism.75,79,97,109,132,143 In addition, the effect of impaired myelination on axon structure could be analyzed.76 Another lesson was that, in some myelin-related mutants, phenotypic alterations were unexpectedly mild. This suggests that nontarget molecules might have compensated for the roles of the inactivated genes.20–22,77,85,86 A very challenging perspective was generated when some mutants could be viewed as models for distinct genetically mediated myelin disorders. This made possible the study of pathogenesis not only in the context of axon–Schwann cell interaction, but also in relation to 535

536

Neurobiology of the Peripheral Nervous System

implications for the entire organism. For instance, the aforementioned consequences of demyelination on axonal fate and target organs, the effect of nerve length on severity of nerve pathology, and the initially unexpected diseasemodifying impact of the immune system are very important issues that have to be considered in the context of understanding pathogenesis and disease treatment.75,143 To understand the cross-talk between axons and Schwann cells, it was also of interest to generate mutants related to axonal genes. Therefore, a mutant deficient in complex gangliosides, the putative receptor molecules for myelin-associated glycoprotein (MAG), was created, resulting in a pathologic phenotype similar to that seen in MAG-deficient mice.120 Moreover, mice deficient in distinct neurofilament genes displayed not only the expected axon-related abnormalities, but also secondary defects related to myelination.61 A recent track in peripheral nervous system (PNS) research is focusing on the molecular organization of the node of Ranvier and its neighboring domains, the paranode and the juxtaparanode.54,92,95a,112 Mice deficient in two closely associated axonal components of the paranode show a very similar phenotype with disrupted paranodal organization.12,16 Interestingly, a mutant deficient in nearly ubiquitous myelin lipids developed almost identical paranodal abnormalities.32,60 This chapter gives an introduction into the phenotype of various engineered mutants related to the PNS, with the aim to provide an understanding of axon-glia interactions and the consequences when these recognition processes are disturbed in the intact organism. Furthermore, it is demonstrated that some of the mutants are instrumental in understanding pathogenesis of inherited peripheral neuropathies, which is an important prerequisite for developing novel treatment strategies for the still untreatable disorders.

TRANSGENIC MODELS RELATED TO MYELIN SHEATH ORGANIZATION AND MAINTENANCE Mice Homozygously Deficient for P0 P0 (also designated myelin protein zero) is a 30-kDa adhesion molecule belonging to the immunoglobulin (Ig) superfamily. Based on in vitro studies, it has been speculated to mediate compaction of myelin as a result of its homophilic adhesion properties (reviewed by Martini and Schachner79), leading to the close apposition of the extracellular aspects of the spiraling Schwann cell membrane forming the intraperiod line. This model received strong support from the determination of the three-dimensional structure of the extracellular domain of P0 by x-ray crystallography.118 The intracellular domain of P0 contains predominantly basic residues, which have been suggested to

interact with negatively charged phospholipids of the adjacent cytoplasmic parts of the Schwann cell membrane, leading to the formation of the major dense line.30,63,68 Interestingly, mutations in the P0 gene affecting the intracellular domain of the protein lead to reduced extracellular adhesion.135,139 P0: The Major Mediator of Myelin Compaction in Peripheral Nerves Mice homozygously deficient in P0 have been generated by targeted disruption of the gene in embryonic stem cells. This was the first engineered mouse mutant deficient in an adhesion molecule. Corroborating previous in vitro experiments, myelin compaction was substantially affected in the absence of P0 (Fig. 24–1A)51 (see Martini and Schachner 79 for review). In addition, the abnormal myelin sheaths were not stable but rather prone to degenerate, as reflected by a steady increase of demyelinated axons in mutants older than 4 weeks.79 Although myelin compaction was impaired, major dense lines were still preserved in approximately 60% of the axon–Schwann cell units of 4-week-old mutants (see Fig. 24–1A). This was unexpected, because formation of the major dense line has been speculated to be mediated by the intracellular part of the molecule.30 Based on the fact that, in the central nervous system (CNS), myelin basic protein (MBP) is a mediator of major dense line formation,99 it was proposed that this protein acts similarly in peripheral nerves of P0-deficient mice. Indeed, immunoelectron microscopy revealed expression of MBP at those sites of abnormal P0-deficient myelin that contained major dense lines.79 The hypothesis was then tested by crossbreeding P0-deficient mice with a spontaneous mutant deficient in MBP, the shiverer mouse. These MBP-deficient mice show almost normal peripheral nerves (i.e., in the presence of P0), but severely impaired CNS myelin with missing major dense lines.99 The double mutants deficient in both P0 and MBP showed a complete absence of major dense lines in peripheral nerves (Fig. 24–1B)77 thus proving that P0 and MBP can fulfill interchangeable roles in the PNS. Based on the fact that MBP can partially compensate for the loss of P0, the hypothesis of whether other adhesion molecules upregulated in peripheral nerves of P0-deficient mice51,138 had functional significance was tested. In analogy to the experiment with MBP-deficient mutants, P0deficient mice were crossbred with other mutants deficient in molecules that are upregulated in the P0 single mutant. Double-mutant mice deficient in P0 and either MAG, neural cell adhesion molecule (NCAM), or peripheral myelin protein 22 (kDa) (PMP22), revealed only subtle changes in myelin phenotype when compared with P0 single mutants.21,22 Thus P0 is the pivotal mediator of myelin compaction in the PNS, sharing some functional properties for intracellular compaction with MBP.

Transgenic Models of Nerve Degeneration

537

B

A

FIGURE 24–1 Myelin sheath in femoral nerve of a P0⫺⫺ mouse (A) and of a P0⫺⫺/MBP⫺⫺ (shiverer) double mutant (B). Note that, in the absence of P0 alone, most Schwann cells form an abnormal myelin sheath comprising both noncompacted and partially compacted myelin. Myelin sheaths of the double mutants are completely uncompacted. Bars: 1 ␮m.

P0-Deficient Mice: A Genetic Model for Myelin-Related Axonopathy Electrophysiologic investigations in P0-deficient mice revealed features indicative not only of abnormal myelination, but also of axonal impairment, because the amplitudes of compound muscle action potentials (CMAPs) were strongly reduced.80,145 Morphometric analysis of peripheral nerves of P0 mutants revealed a significant reduction of axonal diameters in proximal regions of femoral and in facial nerves. In addition, an increase in neurofilament (NF) density was detected.48 Even more striking was the loss of approximately 75% of the distal portions of myelin-competent axons and features indicative of Wallerian degeneration in the toes of the myelin mutants.48 Immunolabeling of footpads with antibodies to cytokeratin 20 revealed a 75% loss of Merkel cells, suggesting that survival of these cells is dependent on the presence or maintenance of their innervating myelinated axons.48 A similar loss of axons was found in the plantar nerve of P0-deficient mice.106 Strikingly, the most prominent rate of axon loss was found within the first 3 months (Fig. 24–2). To investigate whether myelin-related axonal loss at least partially shares the mechanisms underlying Wallerian degeneration, P0-deficient mice were crossbred with the spontaneous mutant C57BL / Wlds, which typically shows protection from Wallerian degeneration as a result of fusion of the genes for ubiquitination factor E4B and nicotinamide mononucleotide adenylyltransferase.71 The double mutants showed a significantly delayed myelin-related axonal loss,106 and retrograde

labeling of plantar nerves revealed that especially motor axons were preserved by the Wlds mutation (see Fig. 24–2). Most interestingly, the surviving axons appeared functionally active because both the amplitude of CMAP and the muscle strength were significantly increased.106 Thus myelin-related axonal loss is a process similar to Wallerian degeneration and can be delayed by introducing the Wlds gene.

Mice Heterozygously Deficient for P0 In contrast to mice homozygously deficient for P0, the reduction of P0 dosage by 50% in heterozygous mutants does not strongly impair myelination. Rather, myelin can form almost normally for approximately 4 months (Fig. 24–3A and B). Then a progressive demyelinating neuropathy is detectable in motor but not sensory nerves.74,80,105,122 Typical pathologic features are demyelinated axons, unusually thin myelin reflecting incomplete remyelination, and supernumerary Schwann cells reminiscent of onion bulbs in human neuropathies (Fig. 24–3C). In line with the demyelinating phenotype was the prolonged F-wave latency of the CMAP of these mice.80 Partial Mediation of Demyelination by Immune Cells A striking and initially unexpected feature was the occurrence of CD8-positive T lymphocytes and macrophages in the endoneurium of these mutants.116,122 The number of these immune cells increased with time and progressing demyelinating neuropathy, suggesting that they might be

538

Neurobiology of the Peripheral Nervous System

A

PO++/WT PO--/WT

1800 ***

1600

***

*

*

Number of axons

1400 1200 1000 800 600 400 200 0 1 month 1.5 months 3 months 12 months n= 3

4

2

3

3

4

Number of labelled motoneurons

B

2

2

PO++/Wlds PO++/WT PO--/WT PO--/Wlds-hom/het

250 200

* *

150 100 50 0 n=

3

2

4

4

FIGURE 24–2 Progressive axonal loss in plantar nerves of P0⫺⫺ mice (A) and delay of plantar motor axon degeneration resulting from cross-breeding of P0⫺⫺ mice with Wlds mice (B). A, Note substantial axonal loss in plantar nerve, particularly at younger ages. Axon numbers have been determined on ultrathin sections using electron microscopy. B, Number of plantar motor axons is reduced in 3-month-old P0⫺⫺ mice. At this age, motor axon loss is not vivsible in P0⫺⫺/ Wlds mutants. Values have been obtained by counting spinal motor neurons after retrograde labeling of plantar nerves using fluorogold.

functionally related to the disorder. To investigate the functional roles of these cells, the myelin mutants were crossbred with mice deficient in mature T and B lymphocytes (i.e., recombinant activating gene 1 [RAG-1]–deficient mice).84 Histologically, the double mutants showed a less severe myelin degeneration in the absence of lymphocytes (Fig. 24–4). This improvement in myelin maintenance was reflected by an amelioration of nerve conduction properties116 and was experimentally reversible, because P0⫹⫺/RAG-1⫺⫺ mutants showed an aggravated phenotype when reconstituted with bone marrow from wild-type mice (Fig. 24–4),81 proving that T lymphocytes are involved in the primarily genetically mediated demyelination.

The role of macrophages in demyelination was also investigated in P0 mutants. Electron microscopy and immunoelectron microscopy revealed that, in the P0 mutants, some macrophages had entered the endoneurial tubes, contacting either still morphologically intact myelin or demyelinated axons (Fig. 24–5A and B). Because this apposition of macrophages with endoneurial tubes was highly suggestive for an involvement in degeneration and resembled macrophage-myelin interaction in GuillainBarré syndrome,9,57,67 myelin mutants were crossbred with spontaneous mutants deficient in macrophage colonystimulating factor (M-CSF), hence displaying impaired macrophage activation. In the P0-deficient mutants also deficient in M-CSF, the numbers of macrophages were not elevated in the demyelinating nerves. The demyelinating phenotype was much less severe than in the P0 single mutants, proving that macrophages are functionally involved in genetic demyelination (Fig. 24–5C and D).19 These examples clearly demonstrate that inherited demyelination can be mediated and modulated by the immune system. The mechanisms underlying this involvement of the immune system are not yet understood, but our observations may have putative impact on future treatment strategies for inherited demyelination using immunomodulators. Sensory Nerve Preservation from Demyelination but Impairment of Function on Mechanical and Thermal Stimuli Another striking feature of mice heterozygously deficient in P0 is the relative preservation of sensory nerves and dorsal roots.74,75,80,105,122 This is reflected by the absence of supernumerary Schwann cells and other features indicative of demyelination. Moreover, neither T lymphocytes nor macrophages are elevated in the sensory nerves.19,116 Unexpectedly, a behavioral test revealed that some sensory functions were slightly impaired, as reflected by raised withdrawal thresholds to mechanical and thermal stimuli, whereas behavioral signs of a painful neuropathy were not detectable.105 On electron microscopy of longitudinal sections of sensory nerves, many nodes of Ranvier were abnormally formed and displayed enlarged nodal gaps with poorly developed nodal Schwann cell microvilli.105 These alterations might be causally linked to the sensory deficits in the absence of profound myelin degeneration in the sensory nerves of the mutants. It is not yet known how reduction in gene dosage of P0 leads to structural alteration of nodes of Ranvier. In summary, mice homo- and heterozygously deficient in P0 have been instrumental in the analysis of the functional roles of distinct myelin components. These mutants also revealed that loss of an important myelin component leads to many secondary reactions by the Schwann cells, such as upregulation of nontarget myelin components with partially compensatory functions. In addition, abnormal myelin sheaths are not stable but are prone to degenerate.

Transgenic Models of Nerve Degeneration

539

B

A

C FIGURE 24–3 Myelinated axons of 10-day-old wild-type (A), 10-day-old P0⫹⫺ (B), and 8-month-old P0⫹⫺ mice (C). Note that in young P0⫹⫺ mice myelination is normal (compare A with B). In adult mice, features indicative of demyelination, such as thin myelin and supernumerary Schwann cells (arrows), are striking (C). Bars: 1 ␮m.

Moreover, the axonal partners are influenced by the loss of the myelin protein, as reflected by a reduction in diameter and eventually degeneration. A striking observation was the involvement of the immune system in the degener-

ation of unstable myelin sheaths as a secondary reaction to reduction in gene dosage. Thus the cellular, subcellular, and molecular consequences of P0 disruption are not confined to the targeted organelle (the myelin) or the

540

Neurobiology of the Peripheral Nervous System

A

B

C

D

FIGURE 24–4 Semithin sections of femoral quadriceps nerves of P0⫹⫺ mice (A), of P0⫹⫺/RAG-1⫺⫺ mice (B), of P0⫹⫺/RAG-1⫺⫺ mice that received bone marrow from wild-type mice (C), and of P0⫹⫺/RAG-1⫺⫺ mice that received bone marrow from RAG-1 mice (D). Note that the presence of immune-competent wild-type bone marrow (A and C) leads to demyelination in P0⫹⫺ mice. All mice were investigated at the age of 1 year. Bars: 20 ␮m.

target cell (the Schwann cell), but involve the neuronal partners and even cells of the immune system. Most interestingly, both features are not confined to P0 mutants, because axonal injury and the involvement of immune cells have also been described in another myelin mutant, the connexin 32 (Cx32)–deficient mouse64 (see below). These common pathologic features of different myelin mutants are of particular interest when the mutants are viewed as animal models for inherited neuropathies.

Mice Overexpressing P0 Based on the fact that other myelin components, such as PMP22 and proteolipid protein (PLP), cause myelinopathies when expressed at unusually high levels,132,143 transgenic mice overexpressing P0 have been generated by introducing additional gene copies. The vector used consisted of the whole P0 gene of the mouse, including 6 kb of the promoter, all exons and introns, and the polyadenylation site.43,137,142

Transgenic Models of Nerve Degeneration

A

B

C

D FIGURE 24–5 A and B, Immunoelectron microscopic localization of macrophages in peripheral nerves of 6-month-old P0⫹⫺ mice. Note that F4/80-positive macrophages have penetrated the Schwann cell basal laminae. A, A macrophage contacts morphologically intact myelin (ruptures in myelin are the consequence of mild fixation necessary for immunoelectron microscopy). B, Myelin is already phagocytosed and the macrophage directly contacts the demyelinated axon. Arrows indicate electron-dense immunoreaction product identifying macrophages. S ⫽ Schwann cell. C and D, Ventral spinal roots of P0⫹⫺ mice (C) and of P0⫹⫺ mice additionally deficient in M-CSF showing an impaired macrophage activation (D). Note that demyelinating phenotype in the M-CSF–deficient myelin mutants is much less severe than in the P0 single mutants. Bars: 1 ␮m. (A and B); 5 ␮m (C and D).

541

542

Neurobiology of the Peripheral Nervous System

The transgenic mice that contained extra copies of P0 showed a dosis-dependent dysmyelinating neuropathy. Mice with mild overexpression showed a transient perinatal formation of too-thin myelin; mice with stronger elevation of P0 showed an arrest of myelination at the stage of the 1:1 ratio between Schwann cells and myelin-competent axons.137,142 Most interestingly, very high expression led to an unexpected impaired sorting of larger caliber axons (Fig. 24–6).137,142 To prove that the dysmyelination is the result of the gene dosage and not the result of a structural effect of the transgene P0 protein, the overexpressing mice were crossbred to P0⫺ mice. When crossbreeding of the transgenic mice and the knockouts was designed in a way that the gene dosage was brought to wild-type levels, the phenotype could be corrected.137 To get insight into the molecular mechanisms that underlie the impaired myelination when P0 is overexpressed, immunoelectron microscopy in peripheral nerves of mice highly overexpressing P0 has been performed. The most striking finding was that those fibers arrested in myelination showed a misdirected trafficking of P0 with abaxonal P0 expression. Moreover, P0 was unusually strongly expressed at the axon–Schwann cell interface and at the apposing Schwann cell membranes of the mesaxon. This led to an abnormally tight membrane apposition reminiscent of that seen in compacted myelin where P0 is the mediator of compaction. Ectopic expression of P0 may thus

A

block spiraling of the myelinating Schwann cells as a result of its adhesive properties in membranes that should be able to be highly mobile. Another mechanism that might exist in overexpressing mice is a disturbed stoichiometry in the expression of myelin components. Indeed, D’Urso and collaborators found evidence for the possibility that P0 and PMP22 interact with each other.34 However, it is not yet known how the misbalanced relationship of the two components led to dysmyelination.

Mice Deficient in PMP22 PMP22 was the first identified culprit in inherited peripheral neuropathies. A 1.5-megabase DNA duplication of chromosome 17p11.2-p12 containing the PMP22 gene has been shown to be the cause of the most frequent form of Charcot-Marie-Tooth (CMT) disease, CMT1A.125 The reciprocal event, the heterozygous deletion of the same chromosomal region as that causing CMT1A, leads to another inherited disorder, hereditary neuropathy with liability to pressure palsies, when duplicated.125 In Chapter 66, the use of conventional and conditional transgenic mice and rats overexpressing PMP22 and heterozygous knockout mice as models for the respective disorders is thoroughly discussed. This chapter briefly deals with mice homozygously deficient in

B FIGURE 24–6 Myelination in 10-day-old transgenic mutants overexpressing P0. Note fasciculating large-caliber axons surrounded by processes of a nonmyelinating Schwann cell (left side) (A) and a myelinating Schwann cell contacting a supernumerary larger caliber axon (asterisk) at the site of the outer mesaxon (B). This situation and fasciculating larger caliber axons reflect deficient sorting of myelin-competent axons. Bars: 1 ␮m.

Transgenic Models of Nerve Degeneration

PMP22, trying to get insight into some functional roles of the molecule. PMP22 is an integral membrane glycoprotein mainly expressed in compact myelin in peripheral nerves. Initially, PMP22 was belived to be a tetraspan membrane molecule, but recent studies suggest that alternative arrangements of the intramembrane domains are conceivable.127 There are several lines of evidence that the molecule is involved in myelin formation and maintenance, but a simplistic model such as cell adhesion, as discussed for P0, is presently not available. Biochemical approaches revealed that PMP22 and P0 can be copurified from peripheral nerve myelin, suggesting that they form complexes in the myelin membrane.34 In vitro experiments suggest a function in Schwann cell growth and differentiation,146 but mice lacking PMP22 or overexpressing the gene show a normal number of Schwann cells perinatally, and altered (increased) numbers of Schwann cells are detected only at later developmental stages.108 In the absence of PMP22, myelination is severely affected. A significant delay in myelination of most profiles, as well as some fibers with hypermyelinated myelin sheaths, are the striking abnormalities visible in peripheral nerves of 4-day-old knockout mice.2 At approximately 3 weeks of age, myelination proceeded in an unusual way in that myelin tomacula formed around almost every myelincompetent axon, as revealed by single fiber preparations. Typically, most myelin tomacula were found at paranodal aspects. In 10-week-old mice, the majority of the tomacula had disappeared in favor of features indicative of demyelination, such as demyelinated axons and supernumerary Schwann cell profiles reminiscent of onion bulbs. Thus the PMP22-deficient myelin sheaths appear to be unstable and prone to degenerate.2 Another striking feature was the involvement of axonal properties in the absence of PMP22. Mice homozygously deficient for PMP22 showed a significant reduction of axonal diameters. Moreover, axonal loss in both lumbar ventral roots and quadriceps nerves was found. In the lumbar roots, approximately 25% of the axons degenerated, whereas in more distal regions of the femoral nerve, axonal loss was even more pronounced, with 40% of axons degenerating.107 Thus, similarly to P0-deficient mice, PMP22 knockout mice show a distally pronounced axonal loss in peripheral nerves. Taken together, these observations suggest multiple functions of PMP22. The molecule might be involved in the turning of Schwann cell loops, the determination of myelin thickness, and in the maintenance of axon–Schwann cell integrity. How these functions are achieved and how other myelin components are involved, such as complex formation with P0, is not yet known.

Mice and Rats Overexpressing PMP22 Transgenic mice and rats have been generated to mimic the PMP22 duplication that is the culprit of the majority of

543

CMT1A cases.58,59,72,93,117 As in transgenic mice overexpressing P0, there was a clear positive correlation between the number of the extra gene copies and the severity of the neuropathy. A more detailed description is given in Chapter 66.

Mice Deficient in the Tunnel Protein Cx32 The discovery Cx32 as a component of peripheral nerve myelin was initially unexpected and originated from the linkage of the X-chromosomal dominant form of Charcot-MarieTooth disorder (CTMX) to mutations in Cx32.11 Detailed immuncytologic investigations113 and the demonstration of functional connexin channels in teased nerve fibers8 established the molecule as a true component of noncompacted membranes of the Schwann cell–related myelin sheath. Cx32 is thought to form so-called reflexive junctions connecting cytoplasmic domains (paranodal loops, SchmidtLanterman incisures) of the myelin spiral, thus forming a radial rapid pathway for ions and small molecules from the periaxonal collar to the Schwann cell body and vice versa.8,110 Several mutations causing CMTX have been investigated in vitro with respect to their influence on channel expression, channel properties, and trafficking of the mutant and wildtype proteins in the cell.1,100 This chapter focuses on the pathologic changes seen in null mutants with the aim to try to understand the functional role of the molecule in myelination and myelin maintenance. In addition, evidence is provided that, as is the case with reduced P0 expression, Cx32 deficiency leads to an activation of immune cells that fosters demyelination and axonopathic changes. Similar to mice heterozygously deficient in P0 or homozygously deficient in MAG (see above and subsequent paragraphs), Cx32-knockout mice show initially normal myelin formation followed by demyelination from postnatal month 4 onward.5a The features indicative of demyelination, such as thinly remyelinated axons and supernumerary Schwann cells, are very similar to those in P0⫹⫺ mice. Another similarity to these mutants is that predominantly motor nerves are affected by the mutations. However, a few features are unique to Cx32 deficiency. A characteristic hallmark is abnormally enlarged periaxonal collars5a (Fig. 24–7A). Based on the fact that periaxonal Schwann cell cytoplasm is the most distant cytoplasmic domain from the Schwann cell body, it is plausible to assume that this aspect suffers most from an interruption of the rapid pathway resulting from Cx32 deficiency. In this contex, it is interesting that absence of Cx32 strongly accelerates demyelination in P0⫹⫺ mice, which could reflect the dependency of myelin in P0⫹⫺ mice on radial pathways.86 Another specific feature of Cx32-deficient mice is axonal damage and resprouting. How this damage is mediated is not known, but it is worthwhile to mention in this context that neurons from Cx32-deficient mice are particularly vulnerable when injured.87

544

Neurobiology of the Peripheral Nervous System

Based on the fact that Cx32 is linked to the X chromosome, Scherer and colleagues115 investigated whether Cx32 is subject to X-inactivation resulting in demyelination of a subpopulation of Schwann cells. In a first step, labeling of teased fibers from Cx32⫹⫺ female mice was perfomed with Cx32-specific antibodies. Indeed, a mosaic labeling was found in that some Schwann cells displayed Cx32 immunoreactivity at their typical compartments, while other myelinating Schwann cells remained entirely negative. In a second step, pathologic changes of teased fibers from heterozygous females were compared with fibers from wild-type mice. In these preparations Cx32⫹⫺ mice, but not their wild-type littermates, showed pathologic features. These investigations suggest that the mild demyelinating neuropathy in heterozygous females is the result of lack of Cx32 in individual Schwann cells.115 Based on the observation that T lymphocytes and macrophages foster demyelination in mice heterozygously deficient in P0, the same hypothesis was investigated in homo-or hemizygous Cx32-deficient mice. Increased numbers of T lymphocytes and macrophages could be detected in demyelinating nerves of the Cx32 mutants. In addition, macrophages were found in apposition to degenerating myelin, reminiscent of a macrophage-mediated demyelinating neuropathy.65 Crossbreeding of Cx32-deficient mice with RAG-1⫺⫺ mice led not only to reduced numbers of endoneurial macrophages, but also to a substantial mitigation of features indicative of myelin degeneration and axonopathic changes (Fig. 24–7B to E).64 However, hallmarks for Cx32 deficiency, such as enlarged periaxonal Schwann cell collars, were not reduced (see Fig. 24–7D and E).64 Thus the immune system appears to be a possibly widespread modulator of the severity of pathologic changes in myelin mutants. This observation might have an important impact on treatment strategies of the corresponding human disorders.

hypermyelinated fibers showing tomacula-like morphology at 6 weeks of age (Fig. 24–8).53 As with the unstable tomacula in PMP22-deficient mice, substantial degeneration of myelin occurred at 6 months of age, as reflected by production of supernumerary Schwann cells and thinly myelinated fibers. The demyelinating features were accompanied by impaired conduction properties of peripheral nerves. Another feature is impaired remyelination of regrowing axons following peripheral nerve injury.133 Thus the progression of the disease, with initially normal myelination followed by degenerative changes, is comparable to that in MAG⫺⫺ and Cx32⫺⫺ mice and mutants heterozygously deficient in P0. However, there is a striking difference between the aforementioned myelin mutants and the periaxin mutants. Whereas in MAG⫺⫺, Cx32⫺⫺, and heterozygous P0 mutants the sensory nerves were mostly preserved from degeneration, these nerves were highly affected in the periaxin mutants. Moreover, in periaxin mutants, mechanical allodynia and thermal hyperalgesia were detectable, as revealed by decreased retraction threshold on mechanical stimuli and shortened retraction latency on heat stimuli, respectively.53 This behavior was in marked contrast to that in heterozygous

Mice Deficient in Periaxin and Dystroglycan Periaxin is a Schwann cell membrane–related and cytoskeleton-associated protein. Two proteins of different sizes (147 kDa [L-periaxin] and 16 kDa [S-periaxin]) are encoded by the corresponding gene, and each form contains a PDZ (postsynaptic density protein 95, Drosophila discs large tumor suppressor, zonula occludens-1) motif, suggesting an involvement in cell-cell contact and cell signaling.35,121 The name of the protein is derived from its early localization within the periaxonal Schwann cell loop in young mice.52 Studies focusing on later stages revealed that the expression of the protein shifts from the periaxonal to the abaxonal compartment of the Schwann cell, where it is predominantly expressed in its large splice form.114 To evaluate the functional roles of periaxin, the periaxin gene was inactivated by homologous recombination in embryonic stem cells. The homozygous null mutants displayed almost normal myelin formation, with a few

A FIGURE 24–7 Pathologic features in Cx32-deficient mice. A, Electron micrograph of a myelinated axon in the femoral quadriceps nerve of a 13-month-old Cx32⫺⫺ mouse. Note the thin myelin sheath, supernumerary Schwann cells (arrows), and typically enlarged periaxonal collar (arrowheads). Figure continued on opposite page

Transgenic Models of Nerve Degeneration

545

C

B

D

E

FIGURE 24–7 Continued B through E, Light (B and C) and electron (D and E) microscopy of ventral spinal roots of 13-month-old Cx32⫺⫺ mice with intact immune system (B and D) and of Cx32⫺⫺ littermates deficient in RAG-1 lacking mature lymphocytes (C and E). Note that degenerative features, such as thin or missing myelin (asterisk) and axonopathic changes (double asterisks), are strongly reduced in myelin mutants deficient in RAG-1. However, the typical hallmarks for Cx32 deficiency, such as enlarged periaxonal collars (arrowheads in D and E), are not influenced by immune deficiency. BV, blood vessel. Bars: 1 ␮m (A, D, and E); 10 ␮m (B and C).

P0-deficient mice, which showed the opposite behavior as reflected by raised withdrawal thresholds to mechanical and thermal stimuli, possibly as a result of malformation of nodes of Ranvier.105

The demyelinating phenotype of periaxin-deficient mice clearly demonstrates that the protein is a nonredundant component of the myelinating Schwann cell in stabilizing the myelinated axon–Schwann cell unit. This stabilization

546

Neurobiology of the Peripheral Nervous System

B

A

FIGURE 24–8 Electron microscopy of a sciatic nerve of a periaxin-deficient knockout mouse. A, A tomaculum-like profile is visible. B, Features of demyelination, remyelination, and possibly axonal resprouting (arrows) reflecting instability in myelin and axon maintenance in the absence of periaxin. Note abundant supernumerary Schwann cells forming onion bulbs. Bars: 2 ␮m. (Courtesy of Drs. Diane Sherman and Peter Brophy.)

is most probably mediated by the interaction of periaxin with dystroglycan-dystrophin–related protein 2 (DRP2), which in turn forms a complex with ␣␤-dystroglycan that interacts with laminin of the Schwann cell basal lamina.121 Thus periaxin and DRP2 form an essential link between the extracellular matrix of the Schwann cell and the Schwann cell itself, stabilizing the mature myelin sheath. In line with this pivotal function is the finding that mutations in the periaxin gene lead to distinct forms of inherited peripheral neuropathies in humans.13,55 In a recent study, the Schwann cell–specific deletion of dystroglycan using the Cre-loxP system led to misfolding of peripheral myelin as similarly observed in periaxin-deficient mice.104 The abnormally folded myelin was related to the strong downregulation of DRP2, an alteration that has been similarly described for the periaxin mutants. In addition, nodal sodium channels were significantly more weakly expressed than in wild-type mice, and mild ultrastructural changes in the organization of nodal microvilli were described, suggesting a role of dystroglycan in the cytoarchitecture of nodes of Ranvier.

Mice Deficient in ␤1 and ␤4 Integrins Other molecules linking the Schwann cell surface and the Schwann cell basal lamina are integrin heterodimers such as ␣6 /␤1 and ␣6 /␤4, which bind to laminin98,111 (see also

Chapter 19). Using the Cre-loxP system, Feltri and colleagues44 inactivated ␤1 integrin exclusively in Schwann cells, which was necessary to circumvent early embryonic death noted to occur in conventional ␤1 integrin knockout mice.41,124 The Schwann cell–specific gene inactivation of ␤1 integrin led to a severe dysmyelinating phenotype with impaired axonal sorting.44 Although regulated similarly to myelin proteins,36,42 ␤4 integrin appears to be redundant during myelination. ␤4 integrin–deficient mice that die perinatally show myelin formation in peripheral nerves similar to that of wild-type mice.47 In addition, Schwann cells of dorsal root ganglion explants from these mutants develop a capacity for myelination in vitro similar to that in explants from wildtype mice.47

Mice Deficient in MAG MAG is a transmembrane glycoprotein with five Ig-like extracellular domains, a transmembrane domain, and an intracellular portion. As a result of alternative splicing, two isoforms of 67 kDa (S-MAG) and 72 kDa (L-MAG) have been found in the deglycosylated stage that also differ in their intracellular domains.66,130 The major form in the PNS is S-MAG. This glycoprotein is a typical adhesion molecule with plasma membrane binding partners.96,140 In terms of axon–Schwann cell interactions, knowledge of the

Transgenic Models of Nerve Degeneration

axonal partner(s) might be very interesting, but these have not yet been identified. However, there is accumulating evidence that sialic acid–containing glycoproteins and complex gangliosides are good candidates for axonal MAG receptors.28,140 Thus MAG can also been viewed as a member of the siglecs.62 As opposed to P0 and PMP22, MAG is a constituent not of compact, but of noncompacted myelin. As a result of its early expression during myelination and its strategic location at the axon–Schwann cell interface,78,128,129 the molecule has been thought to play pivotal roles during myelination. However, mice deficient in MAG showed, as opposed to the CNS,109 an unexpected normal myelination in the PNS (Fig. 24–9A).69,85 Interestingly, the initially normally shaped axon–Schwann cell units were not stable in the absence of MAG. At 6 to 8 months of age, features indicative of axon and myelin disruption were visible (Fig. 24–9B) and predominantly associated with motor nerves rather than with sensory nerves.20,49,141 Reminiscent of Wallerian degeneration, myelin ovoids and upregulation of tenascin C were typical alterations in nerves of aging MAG-deficient mice.20,49 Moreover, tomacula-like structures were found predominantly at paranodal aspects of motor nerve fibers.20,141 In these regions, axonal diameters appeared strongly reduced. Detailed morphometric analysis in MAG-deficient mice revealed a dysregulation of axonal calibers already present at 3 months of age. This

A

547

reduction in axonal diameter was accompanied by an increase in NF density and by a reduction of phosphorylation of NF in the mutants.141 One interpretation of these alterations was that the too-thick myelin resulting from tomacula formation was a result of axonal shrinkage and that MAG regulates axonal caliber. However, formal proof is still missing that reduced axonal diameters are directly caused by MAG deficiency rather than by constriction of the axons resulting from poorly controlled growth of MAG-deficient myelin. Indeed, paranodal tomacula-like structures and myelin folding with reduced axonal caliber have been described in other myelin mutants, including PMP22- and periaxin-deficient mice (see above). It is therefore possible that the highly organized paranodal structures, including the complicated axon–Schwann cell apposition in the form of the paranodal network,50 are particularly susceptible to molecular alterations leading to abnormal myelin folding and axonal constriction. In summary, studies in mice deficient in MAG revealed that, in motor nerves, the myelin-related axon–Schwann cell unit cannot persist in the absence of MAG. For myelin formation, however, MAG appears to be dispensable. Based on the observation that the related cell adhesion molecule NCAM is slightly upregulated periaxonally in MAG-deficient mice, it was tempting to speculate that NCAM functionally compensates for the lack of MAG during myelin formation. Therefore, mice doubly deficient

B FIGURE 24–9 A, Normal myelin sheath of a 4-week-old MAG⫺⫺ mouse demonstrating that myelin formation is possible in the absence of MAG. B, Axon–Schwann cell units are unstable in the absence of MAG. This profile reflects axon degeneration in an 8-month-old MAG⫺⫺ mouse. Bars: 1 ␮m.

548

Neurobiology of the Peripheral Nervous System

in MAG and NCAM have been generated by crossbreeding the respective single mutants. In the absence of both adhesion molecules, myelin formation was still normal.20 However, tomacula formation and degeneration of axon–Schwann cell units was protracted by several months.20 Thus NCAM is a compensatory molecule in the absence of MAG, preventing axon–Schwann cell units from degenerating for a limited time period.

TRANSGENIC MODELS PRIMARILY RELATED TO THE AXON Mice Deficient in the Complex Gangliosides GD1a and GT1b Based on the observation that MAG-transfected COS cells adhere to microwell plates to which the gangliosides GD1a and GT1b have been adsorbed, it was proposed that GD1a and GT1b are putative axonal receptors for MAG.140 The finding that the GT1b-binding tetanus toxin C decorates the axolemma was further support for this hypothesis.119 If this view is correct, one would expect that mice deficient in GD1a and GT1b would develop neuropathologic changes similar to those of MAG-deficient mice. Therefore, mice deficient in the complex ganglioside synthesizing enzyme UDP-N-acetyl-D-galactosamine:GM3 /GD3 N-acetyl-Dgalactoaminyltransferase have been generated.120 In 12- to 16-week-old GD1a- and GT1b-deficient mutants, the pathologic changes were indeed very similar to the alterations seen in MAG-deficient mutants. In the PNS, particularly striking features were axonal collapse, myelin destruction, production of supernumerary Schwann cells, and the presence of axonal sprouts. In addition, behavioral and electrophysiologic features were abnormal.24 Based on the electron micrographs presented by Sheikh et al.,120 it appears, however, that the degenerative features are of earlier onset than in the MAG-deficient mutants. This may indicate that GD1a and GT1b are additional ligands for other Schwann cell–related molecules with functions similar to those of MAG.

Mice Deficient in the Cell Adhesion Molecule L1 L1 is an adhesion molecule of the Ig superfamily. It is expressed by immature axons and Schwann cells and is downregulated to undetectable levels when myelination starts with the turning of Schwann cell loops around segregated, prospective myelinated axons.78 Although the molecule is expressed both by neurons and Schwann cells, the corresponding null mutant is regarded as axon related, because the major phenotype in peripheral nerves appears to be caused predominantly by axonal defects (see below).

L1-deficient mice suffer from many severe abnormalities of the nervous system, including enlarged ventricles, misguiding of axons of the corticospinal tract, and abnormal organization of peripheral nonmyelinated fibers where L1 is constitutively expressed in wild-type mice.26,27 In wild-type mice, the latter structures consist of thin axonal profiles, most of which are ensheathed and separated from one another by slender Schwann cell processes. In the mutants, however, nonmyelinating Schwann cells formed not only such ensheathing processes, but also additional processes not associated with axons and protruding into the endoneurial space (Fig. 24–10). Another abnormality was a strong reduction of ensheathment in some nonmyelinating axon–Schwann cell units leading to extensive fasciculation of individual axonal profiles without intervening Schwann cell processes. In addition, the nonmyelinating Schwann cells in the mutants were associated with a much lower number of axonal profiles as compared with the wild types.27 In an elegant series of experiments, Haney and colleagues56 tested the possiblity that the disturbed axon ensheathment and the reduced number of axons is the result of the lack of L1 on the Schwann cell site, the axon site, or both. For this purpose, sciatic nerve stumps from wild-type mice were sutured into transected nerves from L1-deficient mice and vice versa. The principle finding was that axonal rather than Schwann cell L1 is required for correct ensheathment and axon survival.56

Mice Deficient in NF Genes Neurofilaments are composed of three subunits (NF-L, NF-M, and NF-H), each of which is the product of a separate gene.61 The NF-H component, with its frequent lysine-serine-proline repeats, provides potential phosphorylation sites that contribute to the side arms of neurofilaments, but NF-M can be phosphorylated as well.61 Highly phosphorylated side arms of NFs are believed to be repelled from the filamentous cores by negative charges, resulting in an increase in space between the NFs that leads to an increase in axon caliber. Deletion of the NF-L gene resulted in severe axonal atrophy and impaired maturation of regenerating axons, most probably the result of the fact that, in the absence of NF-L, NF-M and NF-H cannot form filaments.61,144 Deficiency in NF-H leads to a surprisingly modest reduction in axonal diameter, particularly in the C57/B16 mouse strains, whereas in 129J strain the axon size reduction was stronger. NF-M deficiency showed a more robust and uniform decrease in axonal diameters. Surprisingly, all mutants do not show a significant defect in the organization of the nervous system in general. However, a significant delay in myelination of regenerating axons has been documented in NF-L–deficient mice. In addition, NF-L and NF-M mutants appear to show a decrease in axon number of 10% to 20% in peripheral

Transgenic Models of Nerve Degeneration

A

549

B FIGURE 24–10 Nonmyelinated fibers in saphenous nerves of wild-type (A) and L1–deficient knockout (B) mice. Note abnormal Schwann cell processes of nonmyelinating Schwann cells protruding into the endoneurial space of the mutant nerves. This feature probably reflects loss of thin-caliber axons in the mutants. In the wild-type mice, thin-caliber axons are well ensheathed by the nonmyelinating Schwann cells and abnormal Schwann cell processes are not visible. Bars: 1 ␮m.

nerves. In a study focusing on axonal changes in aging NF-L, NF-M, and NF-H knockout mutants, Elder and colleagues38 showed a progressing axonal atrophy in lumbar roots of NF-M and NF-M/H mutants. Substantial axonal loss in aged animals, however, was not seen at the level of the roots. Investigation of more distal aspects of the peripheral nerves, where axonal loss is often more pronounced, has not been carried out. This aspect is of particular interest because mutations in NF-L have been shown to cause an axonal form of CMT with substantial axonal loss in lower nerves.83

TRANSGENIC MODELS RELATED TO THE ORGANIZATION OF THE NODE OF RANVIER In recent years, substantial progress in the knowledge of the sophisticated molecular and cytologic architecture of the node of Ranvier has been achieved.6,7,92,95a,112 In the peripheral nerve, the nodal membrane is contacted by Schwann cell microvilli. The molecular characteristics of the nodal membrane are the clustered voltage-gated Na⫹ channels and the cell adhesion molecules Nr-CAM and

neurofascin 186, which form cytoskeleton-linked complexes via ankyrin G with the Na⫹ channels. In addition, tenascin C and the proteoglycan NG2 characterize the extracellular matrix milieu of the nodal gap.6,7,54,73,92 A recent study showed that inactivation of the dystroglycan gene leads to mild abnormalities in the nodal architecture (see above). Also of particular interest is the paranodal zone, which is characterized by the paranodal or lateral loops that abut the axolemma and cause the typical “scalloping” of the axon membrane (see Chapter 19). A characteristic subcellular element is the septate junction–like transverse bands that form a partially sealing barrier between the nodal gap and the periaxonal space of the juxtaparanode and the internode. Recent studies have focused on the molecular composition of these bands and their putative functional roles. Two groups have independently identified a novel axonal component of these bands initially designated contactin-associated protein (Caspr) or Paranodin.37,82,91 As a result of the homology to the septate junction–related PDZ-related protein neurexin-IV in Drosophila,90 the molecule has also been designated NCP1 (Neurexin-Caspr-Paranodin12). Caspr2 is another axolemmal component, but found at the juxtaparanodal region together with the K⫹ channels Kv1.1 and Kv1.2.94

550

Neurobiology of the Peripheral Nervous System

During development, Caspr2 and the K⫹ channels are first expressed at the paranodal site and then shift to their final destination when Caspr is upregulated.95

Mice Deficient in Caspr The functional role of the paranodal component Caspr has recently been investigated in the corresponding knockout mutant.12 In these mice, paranodal regions of peripheral nerves lack transverse bands and the scalloping of the axonal membrane is mostly absent. Occasionally, protrusion of nodal microvilli below the loops was observed.12 Another abnormality was the absence of the putative cisrelated binding partner of Caspr, contactin, in peripheral nerves, whereas contactin was abnormally distributed in the CNS. Na⫹ channels are mainly organized similar to their organization in wild-type mice, whereas K⫹ channels, which are normally localized at the juxtaparanodal regions, directly neighbored or even partially overlapped with the nodal Na⫹ channels in the mutants. This finding may argue for a role of Caspr in sorting or separating nodal/paranodal ion channels.95 The abnormal morphologic and molecular organization of the paranode of Casprdeficient mutants was paralleled by reduced conduction velocities.12 Although Caspr deficiency produced a clearly altered paranodal organization, barely any degenerative features were detectable. This may be the result of the limited age the mutants reach, possibly because of abnormal cerebellar development (M. Bhat, personal communication, 2001).

Mice Deficient in Contactin and CD9 Contactin, the putative cis-related binding partner of Caspr, has also been inactivated in mice.10 Similar to Caspr mutants, the neurologic phenotype is strong and limits lifespan to approximately 3 months. This is most probably the result of abnormal cerebellar development, as has been suggested for the Caspr mutants.10 More detailed investigations on the nodal domains in peripheral nerves revealed alterations very similar to those seen in Caspr mutants.16 Striking features were the loss of paranodal transverse bands, the abnormally increased space between paranodal loops and the axolemma, and the loss of axolemmal scalloping. An additional finding was the absence of Caspr at paranodal regions but an accumulation in neuronal somata, indicating that presence of contactin is necessary for a correct trafficking of Caspr, possibly as molecular complexes.101 The putative paranodal trans-related binding partner of contactin, neurofascin 155,126 was expressed at the paranodes, but its detectability was more variable than in the wild-type mice. A more striking feature was the abnormal expression of shaker-like K⫹ channels. Similarly to Caspr

mutants, they directly neighbored or even partially overlapped with the nodal Na⫹ channels. Again, these morphologic and molecular alterations in the mutants led to a substantially impaired nerve conduction. A similar, yet less constant, phenotype has recently been described for mice deficient in the glial paranodal component CD9.60a

Mice Lacking Galactocerebroside and Sulfatide Lack of transverse band structures and mislocalization of K⫹ channels is not confined to the Caspr and contactin mutants. Mice deficient in the UDP-galactose:ceramide galactosyltransferase (CGT), which are unable to synthesize galactocerebroside and sulfatide, show a very similar disorganization of paranodal structures.32,33 In analogy to the Caspr and contactin mutants, this was accompanied by a lack of the paranodal proteins Caspr and contactin, and neurofascin 155 was only occasionally detectable and then diffusely distributed in the region of the paranodal Schwann cell loops.33,95 The unchanged strong expression of this glial molecule in Schmidt-Lanterman incisures clearly reflected that the downregulation of neurofascin 155 in the mutants was paranode specific.95 Similarly, as seen in the contactin- and Caspr-deficient mutants, the K⫹ channels and the Caspr2 protein remained paranodal, where the molecules are normally only transiently expressed during development, rather than being translocated to the juxtaparanodal compartment as is the case in adult wild-type mice,95 possibly as a result of the lack of Caspr in paranodal domains. Initially, altered electrophysiologic parameters in the PNS, such as reduced conduction velocities and CMAP amplitudes, have been interpreted as reduced insulation properties of the compact myelin sheaths lacking galactocerebroside or sulfatide.14,15,25 However, it is plausible that these alterations are at least partially caused by the abnormal organization of the paranodes. In contrast to mice deficient in Caspr and contactin, the CGT-deficient mice showed a severe dysmyelination in the ventral white matter of the spinal cord, while the dorsal columns were preserved.25,31 The reason for this locally restricted myelin deficiency is not known but certainly reflects impaired axon-glia interaction in the CNS.97 It is possible that the myelin deficiency in the ventral columns is related to the role of galactocerebroside and sulfatide in the incorporation of PLP into lipid rafts of the oligodendrocyte membrane so that the CGT-deficient mice suffer from a missorting of PLP123 (see also Chapter 19). Recently, it has been found that inactivation of the galactosylceramide sulfotransferase gene (GST), causing deficiency of sulfatide, leads to a very similar disorganization of the nodal region, including paranodal clustering of K⫹ channels and diffuse localization of Caspr.60 Figure 24–11 summarizes the molecular changes in node-related mutants.

551

Transgenic Models of Nerve Degeneration

A

WILD TYPE MICE Perinodal astrocyte

Juxtaparanode

Microvilli

Nav1.6 ankG

CNS/Oligodendrocytes

B

NF155

PNS/Schwann cells

CGT, GST, CANTACTIN, CASPR-I-MICE Perinodal astrocyte

FIGURE 24–11 Synoptic view of the consequences for cellular and molecular organization of the nodal complex resulting from the absence of CGT, GST, contactin, and Caspr. For comparison, consequences for myelinated fibers of the CNS (left side) are also indicated, reflecting principally similar functional roles of the molecules in PNS and CNS. (From Scherer, S. S., and Arroyo, E. J.: Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Syst. 7:1, 2002, with permission.)

Incisure

Kv1.1 Kv1.2 Kvβ2 Caspr2

Node

Internode

Basal lamina

Internode

Microvilli

Caspr 1.1 Nav1.6 1.2 Juxtaparanode para- Node node ank G

CNS/Oligodendrocytes

TRANSGENIC MODELS WITH PERIPHERAL NEUROPATHY NOT DIRECTLY RELATED TO NEURAL GENES Mice Deficient in Porphobilinogen Deaminase The mutant deficient in the heme biosynthesis enzyme porphobilinogen deaminase has been created to obtain an animal model for acute intermittent porphyria manifesting in a neurologic syndrome that includes autonomic neuropathy, vomiting, hypertension, tachycardia, and motor weakness associated with a peripheral neuropathy.70 In this context, it was of interest that the mouse developed an axonopathy in

NF155

Basal lamina

1.1 Caspr 1.2 para- Juxtaparanode node

Incisure

Internode

PNS/Schwann cells

motor but not sensory nerves. Particularly, the fibers larger than 8 ␮m degenerated from postnatal month 6 onward, resulting in a decrease of 90% by the age of 17 months.70 The pathologic features resembled hallmarks of Wallerian degeneration, including disruption of axonal cytoskeleton, contorted myelin sheaths devoid of axons, and the formation of Schwann cell profiles reminiscent of bands of Büngner. These degenerative events were reflected by a significant decrease of sciatic nerve–related CMAP amplitude in small foot muscles, whereas features indicative of primary demyelination were mostly absent or only mildly expressed. Major electromyographic changes were a severe and chronic neurogenic pattern with decreased motor unit action potential recruitment, increased motor unit action potential, and polyphasia.70

552

Neurobiology of the Peripheral Nervous System

The mechanisms leading to axonal damage in the mutants are still unresolved. However, based on the observation that axonal loss in these mice occurs at almost normal levels of ␦-aminolevulinic acid, one can exclude this neurotoxic heme precursor as a cause for the neuropathy, although it is typically found in plasma and urine of human patients. Rather, a decrease in heme proteins such as mitochondrial cytochromes could damage the motor neurons and/or their processes, leading to neuropathy.70

ACKNOWLEDGMENTS The author’s laboratory is supported by the Deutsche Forschungsgemeinschaft (SFB 581, special research program on “Microglia”), the Gemeinnützige Hertie-Stiftung, the State of Bavaria, and local research funds of the University of Würzburg. I am grateful to Klaus V. Toyka for stimulating discussions and support, and to all the members of the laboratory and external colleagues for their valuable contributions during scientific collaboration. I am particularly grateful to Steve Scherer, Diane Sherman, and Peter Brophy for supplying their figures.

REFERENCES 1. Abrams, C. K., Oh, S., Ri, Y., and Bargiello, T. A.: Mutations in connexin 32: the molecular and biophysical bases for the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Rev. 32:203, 2000. 2. Adlkofer, K., Martini, R., Aguzzi, A., et al.: Hypermyelination and demyelinating peripheral neuropathy in Pmp22deficient mice. Nat. Genet. 11:274, 1995. 3. Aguayo, A. J., Attiwell, M., Trecarten, J., et al.: Abnormal myelination in transplanted trembler mouse Schwann cells. Nature 265:73, 1977. 4. Aguayo, A. J., Charron, L., and Bray, G. M.: Potential of Schwann cells from unmyelinated nerves to produce myelin: a quantitative ultrastructural and radiographic study. J. Neurocytol. 5:565, 1976. 5. Aguayo, A. J., Epps, J., Charron, L., and Bray, G. M.: Multipotentiality of Schwann cells in cross-anastomosed and grafted myelinated and unmyelinated nerves: quantitative microscopy and radioautography. Brain Res. 104:1, 1976. 5a. Anzini, P., Neuberg, D. H. H., Schachner, M., et al.: Structural abnormalities and deficient maintenance of peripheral nerve myelin in mice lacking the gap junction protein connexin 32. J. Neurosci. 17:4545, 1997. 6. Arroyo, E. J., and Scherer, S. S.: On the molecular architecture of myelinated fibers. Histochem. Cell Biol. 113:1, 2000. 7. Arroyo, E. J., Xu, Y.-T., Zhou, L., et al.: Myelinating Schwann cells determine the internodal localization of Kv1.1, Kv1.2, Kv␤2, and Caspr. J. Neurocytol. 28:333, 1999. 8. Balice-Gordon, R. J., Bone, L. J., and Scherer, S. S.: Functional gap junctions in the Schwann cell myelin sheath. J. Cell Biol. 142:1095, 1998.

9. Ballin, R. H., and Thomas, P. K.: Electron microscope observations on demyelination and remyelination in experimental allergic neuritis. I. Demyelination. J. Neurol. Sci. 8:1, 1969. 10. Berglund, E. O., Murai, K. K., Fredette, B., et al.: Ataxia and abnormal cerebellar microorganization in mice with ablated contactin gene expression. Neuron 24:739, 1999. 11. Bergoffen, J., Scherer, S. S., Wang, S., et al.: Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262:2039, 1993. 12. Bhat, M. A., Rios, J. C., Lu, Y., et al.: Axon-glia interactions and the domain organization of myelinated axons requires neurexin IV/Caspr/Paranodin. Neuron 30:369, 2001. 13. Boerkoel, C., Takashima, H., Stankiewicz, P., et al.: Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am. J. Hum. Genet. 68:325, 2001. 14. Bosio, A., Binczek, E., Haupt, W. F., and Stoffel, W.: Composition and biophysical properties of myelin lipid define the neurological defects in galactocerebroside- and sulfatide-deficient mice. J. Neurochem. 70:308, 1998. 15. Bosio, A., Binczek, E., and Stoffel, W.: Functional breakdown of the lipid bilayer of the myelin membrane in central and peripheral nervous system by disrupted galactocerebroside synthesis. Proc. Natl. Acad. Sci. U. S. A. 93:13280, 1996. 16. Boyle, M. E., Berglund, E. O., Murai, K. K., et al.: Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30:385, 2001. 17. Bunge, R. P.: Expanding roles for the Schwann cell: ensheathment, myelination, trophism and regeneration. Curr. Opin. Neurobiol. 3:805, 1993. 18. Bunge, R. P., Bunge, M. B., and Bates, M.: Movements of the Schwann cell nucleus implicate progression of the inner (axon-related) Schwann cell process during myelination. J. Cell Biol. 109:273, 1989. 19. Carenini, S., Mäurer, M., Werner, A., et al.: The role of macrophages in demyelinating peripheral nervous system of mice heterozygously deficient in P0. J. Cell Biol. 152:301, 2001. 20. Carenini, S., Montag, D., Cremer, H., et al.: Absence of myelin-associated glycoprotein (MAG) and the neural cell adhesion molecule (N-CAM) interferes with the maintenance, but not with the formation of peripheral myelin. Cell Tissue Res. 287:3, 1997. 21. Carenini, S., Montag, D., Schachner, M., and Martini, R.: Subtle roles of neural cell adhesion molecule and myelinassociated glycoprotein during Schwann cell spiralling in P0-deficient mice. Glia 27:203, 1999. 22. Carenini, S., Neuberg, D., Schachner, M., et al.: Localization and functional roles of PMP22 in peripheral nerves of P0-deficient mice. Glia 28:256, 1999. 23. Chan, J. R., Cosgaya, J. M., Wu, Y. J., and Shooter, E. M.: Neurotrophins are key mediators of the myelination program in the peripheral nervous system. Proc. Natl. Acad. Sci. U. S. A. 98:14661, 2001. 24. Chiavegatto, S., Sun, J., Nelson, R. J., and Schnaar, R. L.: A functional role for complex gangliosides: motor deficits in GM2/GD2 synthase knockout mice. Exp. Neurol. 166:227, 2000.

Transgenic Models of Nerve Degeneration 25. Coetzee, T., Fujita, N., Dupree, J., et al.: Myelination in the absence of galactocerebroside and sulfatide: normal structure with abnormal function and regional instability. Cell 86:209, 1996. 26. Cohen, N. R., Taylor, J. S., Scott, L. B., et al.: Errors in corticospinal axon guidance in mice lacking the neural cell adhesion molecule L1. Curr. Biol. 8:26, 1998. 27. Dahme, M., Bartsch, U., Martini, R., et al.: Disruption of the mouse L1 gene leads to malformations of the nervous system. Nat. Genet. 17:346, 1997. 28. De Bellard, M. E., and Filbin, M. T.: Myelin-associated glycoprotein, MAG, selectively binds several neuronal proteins. J. Neurosci. Res. 56:213, 1999. 29. de Waegh, S. M., Lee, V. M. Y., and Brady, S. T.: Local modulation of neurofilament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell 68:451, 1992. 30. Ding, Y., and Brunden, K. R.: The cytoplasmic domain of myelin glycoprotein P0 interacts with negatively charged phospholipid bilayers. J. Biol. Chem. 269:10764, 1994. 31. Dupree, J. L., Coetzee, T., Suzuki, K., and Popko, B.: Myelin abnormalities in mice deficient in galactocerebroside and sulfatide. J. Neurocytol. 27:649, 1998. 32. Dupree, J. L., Girault, J.-A., and Popko, B.: Axo-glial interactions regulate the localization of axonal paranodal proteins. J. Cell Biol. 147:1145, 1999. 33. Dupree, J. L., and Popko, B.: Genetic dissection of myelin galactolipid function. J. Neurocytol. 28:271, 1999. 34. D’Urso, D., Ehrhardt, P., and Müller, H. W.: Peripheral myelin protein 22 and protein zero: a novel association in peripheral nervous system myelin. J. Neurosci. 19:3396, 1999. 35. Dytrych, L., Sherman, D. L., Gillespie, C. S., and Brophy, P. J.: Two PDZ domain proteins encoded by the murine periaxin gene are the result of alternative intron retention and are differentially targeted in Schwann cells. J. Biol. Chem. 273:5794, 1998. 36. Einheber, S., Milner, T. A., Giancotti, F., and Salzer, J. L.: Axonal regulation of Schwann cell integrin expression suggests a role for alpha6 beta4 in myelination. J. Cell Biol. 123:1223, 1993. 37. Einheber, S., Zanazzi, G., Ching, W., et al.: The axonal membrane protein Caspr, a homologue of neurexin IV, is a component of the septate-like paranodal junctions that assemble during myelination. J. Cell Biol. 139:1495, 1997. 38. Elder, G. A., Friedrich, V. L. Jr., Margita, A., and Lazzarini, L. A.: Age-related atrophy of motor axons in mice deficient in the mid-sized neurofilament subunit. J. Cell Biol. 146:181, 1999. 39. Eldridge, C. F., Bunge, M. B., and Bunge, R. P.: Differentiation of axon-related Schwann cells in vitro: II. Control of myelin formation by basal lamina. J. Neurosci. 9:625, 1989. 40. Eldridge, C. F., Bunge, M. B., Bunge, R. P., and Wood, P. M.: Differentation of axon-related Schwann cells in vitro. I. Ascorbic acid regulates basal lamina assembly and myelin formation. J. Cell Biol. 105:1023, 1987. 41. Fassler, R., and Meyer, M.: Consequences of lack of beta 1 integrin gene expression in mice. Genes Dev. 9:1896, 1995.

553

42. Feltri, L., Scherer, S. S., Nemni, R., et al.: ␤4 Integrin expression in myelinating Schwann cells is polarized, developmentally regulated and axonally dependent. Development 120:1287, 1994. 43. Feltri, M. L., D’Antonio, M., Quattrini, A., et al.: A novel P0 glycoprotein transgene activates expression of lacZ in myelinforming Schwann cells. Eur. J. Neurol. 11:1577, 1999. 44. Feltri, M. L., Graus Porta, D., Previtali, S. C., et al.: Conditional disruption of beta 1 integrin in Schwann cells impedes interactions with axons. J. Cell Biol. 156:199, 2002. 45. Fernandez-Valle, C., Fregien, N., Wood, P. M., and Bunge, M. B.: Expression of the protein zero myelin gene in axon-related Schwann cells is linked to basal lamina formation. Development 119:867, 1993. 46. Fernandez-Valle, C., Gwynn, L., Wood, P. W., et al.: Anti␤1 integrin antibody inhibits Schwann cell myelination. J. Neurobiol. 25:1207, 1994. 47. Frei, R., Dowling, J., Carenini, S., et al.: Myelin formation by Schwann cells in the absence of ␤4 integrin. Glia 27:269, 1999. 48. Frei, R., Mötzing, S., Kinkelin, I., et al.: Loss of distal axons and sensory Merkel cells and features indicative of muscle denervation in hindlimbs of P0-deficient mice. J. Neurosci. 19:6058, 1999. 49. Fruttiger, M., Montag, D., Schachner, M., and Martini, R.: Crucial role for the myelin-associated glycoprotein in the maintenance of axon-myelin integrity. Eur. J. Neurosci. 7:511, 1995. 50. Gatzinsky, K. P., Persson, G. H., and Berthold, C. H.: Removal of retrogradely transported material from rat lumbosacral alpha-motor axons by paranodal axonSchwann cell networks. Glia 20:115, 1997. 51. Giese, K. P., Martini, R., Lemke, G., et al.: Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71:565, 1992. 52. Gillespie, C. S., Sherman, D. L., Blair, G. E., and Brophy, P. J.: Periaxin, a novel protein of myelinating Schwan cells with a possible role in axon ensheathment. Neuron 12:497, 1994. 53. Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., et al.: Peripheral demyelination and neuropathic pain behavior in periaxin-deficient mice. Neuron 26:523, 2000. 54. Girault, J. A., and Peles, E.: Development of nodes of Ranvier. Curr. Opin. Neurobiol. 12:476, 2002. 55. Guilbot, A., Williams, A., Ravise, N., et al.: A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum. Mol. Genet. 10:415, 2001. 56. Haney, C. A., Sahenk, Z., Li, C., et al.: Heterophilic binding of L1 on unmyelinated sensory axons mediates Schwann cell adhesion and is required for axonal survival. J. Cell Biol. 146:1173, 1999. 57. Ho, T. W., McKhann, G. M., and Griffin, J. W.: Human autoimmune neuropathies. Annu. Rev. Neurosci. 21:187, 1998. 58. Huxley, C., Passage, E., Manson, A., et al.: Construction of a mouse model of Charcot-Marie-Tooth disease type 1A by pronuclear injection of human YAC DNA. Hum. Mol. Genet. 5:563, 1996.

554

Neurobiology of the Peripheral Nervous System

59. Huxley, C., Passage, E., Robertson, A. M., et al.: Correlation between varying levels and the degree of demyelination and reduction in nerve conduction velocity in transgenic mice. Hum. Mol. Genet. 7:449, 1998. 60. Ishibashi, T., Dupree, J. L., Ikenaka, K., et al.: A myelin galactolipid, sulfatide, is essential for maintenance of ion channels on myelinated axon but not essential for initial cluster formation. J. Neurosci. 22:6507, 2002. 60a. Ishibashi, T., Ding, L., Ikenaka, K., et al.: Tetraspanin protein CD9 is a novel paranodal component regulating paranodal junctional formation. J. Neurosci. 24:96, 2004. 61. Julien, J.-P.: Neurofilament functions in health and disease. Curr. Opin. Neurobiol. 9:554, 1999. 62. Kelm, S., Schauer, R., and Crocker, P. R.: The sialoadhesins—a family of sialic acid-dependent cellular recognition molecules within the immunoglobulin superfamily. Glycoconj. J. 13:913, 1996. 63. Kirschner, D. A., and Ganser, A. L.: Compact myelin exists in the absence of basic protein in the shiverer mutant mouse. Nature 283:207, 1908. 64. Kobsar, I., Berghoff, M., Samsam, M., et al.: Preserved myelin integrity and reduced axonopathy in connexin32 deficient mice lacking the recombination activation gene1. Brain 126:804, 2003. 65. Kobsar, I., Mäurer, M., Ott, T., and Martini, R.: Macrophage-related demyelination in peripheral nerves of mice deficient in the gap junction protein connexin 32. Neurosci. Lett. 320:17, 2002. 66. Lai, C., Brow, M. A., Nave, K. A., et al.: Two forms of 1B236/myelin-associated glycoprotein, a cell adhesion molecule for postnatal neural development, are produced by alternative splicing. Proc. Natl. Acad. Sci. U. S. A. 84:4337, 1987. 67. Lampert, P. W.: Mechanism of demyelination in experimental allergic neuritis. Lab. Invest. 20:127, 1969. 68. Lemke, G., Lamar, E., and Patterson, J.: Isolation and analysis of the gene encoding peripheral myelin protein zero. Neuron 1:73, 1988. 69. Li, C., Tropak, M. B., Gerial, R., et al.: Myelination in the absence of myelin-associated glycoprotein. Nature 369:747, 1994. 70. Lindberg, R. L. P., Martini, R., Baumgartner, M., et al.: Motor neuropathy in porphobilinogen deaminase-deficient mice imitiates the peripheral neuropathy of human acute porphyria. J. Clin Invest. 103:1127, 1999. 71. Mack, T. G. A., Reiner, M., Beurowski, B., et al.: Wallerian degeneration of injured axons and synapses is delayed by a Ube4b/Nmnat chimeric gene. Nat. Neurosci. 4:1199, 2001. 72. Magyar, J. P., Martini, R., Ruelicke, T., et al.: Impaired differentiation of Schwann cells in transgenic mice with increased PMP22 gene dosage. J. Neurosci. 16:5351, 1996. 73. Martin, S., Levine, A. K., Chen, Z. J., et al.: Deposition of the NG2 proteoglycan at nodes of Ranvier in the peripheral nervous system. J. Neurosci. 21:8119, 2001. 74. Martini, R.: Animal models for inherited peripheral neuropathies. J. Anat. 191:321, 1997. 75. Martini, R.: Animal models for inherited peripheral neuropathies: chances to find treatment strategies? J. Neurosci. Res. 61:244, 2000.

76. Martini, R.: The effect of myelinating Schwann cells on axons. Muscle Nerve 24:456, 2001. 77. Martini, R., Mohajeri, M. H., Kasper, S., et al.: Mice doubly deficient in the genes for P0 and myelin basic protein show that both proteins contribute to the formation of the major dense line in peripheral nerve myelin. J. Neurosci. 15:4488, 1995. 78. Martini, R., and Schachner, M.: Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and MAG) and their shared carbohydrate epitope and myelin basic protein in developing sciatic nerve. J. Cell Biol. 103:2439, 1986. 79. Martini, R., and Schachner, M.: Molecular bases of myelin formation as revealed by investigations on mice deficient in glial cell surface molecules. Glia 19:298, 1997. 80. Martini, R., Zielasek, J., Toyka, K. V., et al.: Protein zero (P0)-deficient mice show myelin degeneration in peripheral nerves characteristic of inherited human neuropathies. Nat. Genet. 11:281, 1995. 81. Mäurer, M., Schmid, C. D., Bootz, F., et al.: Bone marrow transfer from wild type mice reverts the beneficial effect of genetically-mediated immune deficiency in myelin mutants. Mol. Cell. Neurosci. 17:1094, 2001. 82. Menegoz, M., Gaspar, P., Le Bert, M., et al.: Paranodin, a glycoprotein of neuronal paranodal membranes. Neuron 19:319, 1997. 83. Mersiyanova, I. V., Perepelov, A. V., Polyakov, A. V., et al.: A new variant of Charcot-Marie-Tooth disease type 2 is probably the result of a mutation in the neurofilamentlight gene. Am. J. Hum. Genet. 67:37, 2000. 84. Mombaerts, P., Iacomini, J., Johnson, R. S., et al.: RAG-1deficient mice have no mature B and T lymphocytes. Cell 68:869, 1992. 85. Montag, D., Giese, K. P., Bartsch, U., et al.: Mice deficient for the myelin-associated glycoprotein show subtle abnormalities in myelin. Neuron 13:229, 1994. 86. Neuberg, D. H.-H., Carenini, S., Schachner, M., et al.: Accelerated demyelination of peripheral nerves in mice deficient in connexin 32 and protein zero. J. Neurosci. Res. 53:542, 1998. 87. Oguro, K., Jover, T., Tanaka, H., et al.: Global ischemiainduced increases in the gap junctional proteins connexin 32 (Cx32) and Cx36 in hippocampus and enhanced vulnerability of Cx32 knock-out mice. J. Neurosci. 21:7534, 2001. 88. Owens, G. C., Boyd, C. J., Bunge, R. P., and Salzer, J. L.: Expression of recombinant myelin-associated glycoprotein in primary Schwann cells promotes the initial investment of axons by myelinating Schwann cells. J. Cell Biol. 111:1171, 1990. 89. Owens, G. C., and Bunge, R. P.: Schwann cells infected with a recombinant retrovirus expressing myelin-associated glycoprotein antisense RNA do not form myelin. Neuron 7:565, 1991. 90. Peles, E., Joho, K., Plowman, G., and Schlessinger, J.: Close similarity between Drosophila neurexin IV and mammalian Caspr protein suggests a conserved mechanism for cellular interactions. Cell 88:745, 1997. 91. Peles, E., Nativ, M., Lustig, M., et al.: Identification of a novel contactin-associated transmembrane receptor with

Transgenic Models of Nerve Degeneration

92. 93.

94.

95.

95a.

96.

97. 98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

multiple domains implicated in protein-protein interactions. EMBO J. 16:978, 1997. Peles, E., and Salzer, J. L.: Molecular domains of myelinated axons. Curr. Opin. Neurobiol. 10:558, 2000. Perea, J., Robertson, A., Tolmachova, T., et al.: Induced myelination and demyelination in a conditional mouse model of Charcot-Marie-Tooth disease type 1A. Hum. Mol. Genet. 10:1007, 2001. Poliak, S., Gollan, L., Martinez, R., et al.: Caspr2, a new member of the neurexin superfamily, is localized at juxtaparanodes of myelinated axons and associates with K⫹ channels. Neuron 24:1037, 1999. Poliak, S., Gollan, L., Salomon, D., et al.: Localization of Caspr2 in myelinated nerves depends on axon-glia interactions and the generation of barriers along the axon. J. Neurosci. 21:7568, 2001. Poliak, S., and Peles, E.: The local differentiation of myelinated axons at nodes of Ranvier. Nat. Rev. Neurosci. 4:968, 2003. Poltorak, M., Sadoul, R., Keilhauer, G., et al.: Myelinassociated glycoprotein, a member of the L2/HNK-1 family of neural cell adhesion molecules, is involved in neuronoligodendrocyte and oligodendrocyte-oligodendrocyte interaction. J. Cell Biol. 105:1893, 1987. Popko, B.: Myelin galactolipids: mediators of axon-glial interactions? Glia 29:149, 2000. Previtali, S. C., Feltri, M. L., Archelos, J. J., et al.: Role of integrins in the peripheral nervous system. Prog. Neurobiol. 64:35, 2001. Privat, A., Jacque, C., Bourre, J. M., et al.: Absence of the major dense line in the myelin of the mutant mouse “shiverer.” Neurosci. Lett. 12:107, 1979. Ressot, C., and Bruzzone, R.: Connexin channels in Schwann cells and the development of the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Rev. 32:192, 2000. Rios, J. C., Melendez-Vasquez, C. V., Einheber, S., et al.: Contactin-associated protein (Caspr) and contactin form a complex that is targeted to the paranodal junctions during myelination. J. Neurosci. 20:8354, 2000. Sahenk, Z., and Chen, L.: Abnormalities in the axonal cytoskeleton induced by a connexin32 mutation in nerve xenografts. J. Neurosci. Res. 51:174, 1998. Sahenk, Z., Chen, L., and Mendell, J. R.: Effects of PMP22 duplication and deletions on the axonal cytoskeleton. Ann. Neurol. 45:16, 1999. Saito, F., Moore, S. A., Barresi, R., et al.: Unique role of dystroglycan in peripheral nerve myelination, nodal structure, and sodium channel stabilization. Neuron 38:747, 2003. Samsam, M., Frei, R., Marziniak, M., et al.: Impaired sensory function in heterozygous P0 knockout mice is associated with nodal changes in sensory nerves. J. Neurosci. Res. 67:167, 2002. Samsam, M., Mi, W., Wessig, C., et al.: The wlds mutation delays robust loss of motor and sensory axons in a genetic model for myelin-related axonopathy. J. Neurosci. 23:2833, 2003. Sancho, S., Magyar, J. P., Aguzzi, A., and Suter, U.: Distal axonopathy in peripheral nerves of PMP22-mutant mice. Brain 122:1563, 1999.

555

108. Sancho, S., Young, P., and Suter, U.: Regulation of Schwann cell proliferation and apoptosis in PMP22-deficient mice and mouse models of Charcot-Marie-Tooth disease type 1A. Brain 124:2177, 2001. 109. Schachner, M., and Bartsch, U.: Multiple functions of the myelin-associated glycoprotein MAG (siglec-4a) in formation and maintenance of myelin. Glia 29:154, 2000. 110. Scherer, S. S.: Molecular genetics of demyelination: new wrinkles on an old membrane. Neuron 18:13, 1997. 111. Scherer, S. S.: Myelination: some receptors required. J. Cell Biol. 156:13, 2002. 112. Scherer, S. S., and Arroyo, E. J.: Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Syst. 7:1, 2002. 113. Scherer, S. S., Deschênes, S. M., Xu, Y. T., et al.: Connexin32 is a myelin-related protein in the PNS and CNS. J. Neurosci. 15:8281, 1995. 114. Scherer, S. S., Xu, Y. T., Bannerman, P. G. C., et al.: Periaxin expression in myelinating Schwann cells: modulation by axon-glial interactions and polarized localization during development. Development 121:4265, 1995. 115. Scherer, S. S., Xu, Y.-T., Nelles, E., et al.: Connexin32-null mice develop demyelinating peripheral neuropathy. Glia 24:8, 1998. 116. Schmid, C. D., Stienekemeier, M., Oehen, S., et al.: Immune deficiency in mouse models for inherited peripheral neuropathies leads to improved myelin maintenance. J. Neurosci. 20:729, 2000. 117. Sereda, M., Griffiths, I., Pühlhofer, A., et al.: A transgenic rat model of Charcot-Marie-Tooth disease. Neuron 16:1049, 1996. 118. Shapiro, L., Doyle, J. P., Hensley, P., et al.: Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17:435, 1996. 119. Sheikh, K. A., Deerinck, T. J., Ellisman, M. H., and Griffin, J. W.: The distribution of ganglioside-like moieties in peripheral nerves. Brain 122:449, 1999. 120. Sheikh, K. A., Sun, J., Liu, Y., et al.: Mice lacking complex gangliosides develop Wallerian degeneration and myelination defects. Proc. Natl. Acad. Sci. U. S. A. 96:7532, 1999. 121. Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J.: Specific disruption of a Schwann cell dystrophin-related protein complex in a demyelinating neuropathy. Neuron 30:677, 2001. 122. Shy, M. E., Arroyo, E., Sladky, J., et al.: Heterozygous P0 knock-out mice develop a peripheral neuropathy that resembles chronic inflammatory demyelinating polyneuropathy. J. Neuropathol. Exp. Neurol. 56:811, 1997. 123. Simons, M., Kramer, E. M., Thiele, C., et al.: Assembly of myelin by association of proteolipid protein with cholesterol- and galactosylceramide-rich membrane domains. J. Cell Biol. 151:143, 2000. 124. Stephens, L. E., Sutherland, A. E., Klimanskaya, I. V., et al.: Deletion of beta 1 integrins in mice results in inner cell mass failure and peri-implantation lethality. Genes Dev. 9:1883, 1995. 125. Suter, U., and Snipes, G. J.: Biology and genetics of hereditary motor and sensory neuropathies. Annu. Rev. Neurosci. 18:45, 1995.

556

Neurobiology of the Peripheral Nervous System

126. Tait, S., Gunn-Moore, F., Collinson, J. M., et al.: An oligodendrocyte adhesion molecule at the site of assembly of the paranodal axo-glial junction. J. Cell Biol. 150:657, 2000. 127. Taylor, V., Zgraggen, C., Naef, R., and Suter, U.: Membrane topology of peripheral myelin protein 22. J. Neurosci. Res. 62:15, 2001. 128. Trapp, B. D., O’Connell, M. F., and Andrews, S. B.: Ultrastructural immunolocalization of MAG and P0 proteins in cryosections of peripheral nerve. J. Cell Biol. 135:228a, 1986. 129. Trapp, B. D., and Quarles, R. H.: Presence of the myelinassociated glycoprotein correlates with alterations in the periodicity of peripheral myelin. J. Cell Biol. 92:877, 1982. 130. Tropak, M. B., Johnson, P. W., Dunn, R. J., and Roder, J. C.: Differential splicing of MAG transcripts during CNS and PNS development. Brain Res. 464:143, 1988. 131. Weinberg, H. J., and Spencer, P. S.: Studies on the control of myelinogenesis. I. Myelination of regenerating axons after entry into a foreign unmyelinated nerve. J. Neurocytol. 4:395, 1975. 132. Werner, H., Jung, M., Klugmann, M., et al.: Mouse models of myelin disease. Brain Pathol. 8:771, 1998. 133. Williams, A. C., and Brophy, P. J.: The function of the periaxin gene during nerve repair in a model of CMT4F. J. Anat. 200:323, 2002. 134. Windebank, A. J., Wood, P., Bunge, R. P., and Dyck, P. J.: Myelination determines the caliber of dorsal root ganglions in culture. J. Neurosci. 56:1563, 1985. 135. Wong, M. H., and Filbin, M. T.: Dominant-negative effect on adhesion by myelin P0 protein truncated in its cytoplasmic domain. J. Cell Biol. 134:1531, 1996. 136. Wood, P. M., Schachner, M., and Bunge, R. P.: Inhibition of Schwann cell myelination in vitro by antibody to the L1 adhesion molecule. J. Neurosci. 111:3635, 1990.

137. Wrabetz, L., Feltri, M. L., Quattrini, A., et al.: P0 glycoprotein overexpression causes congenital hypomyelination of peripheral nerves. J. Cell Biol. 148:1021, 2000. 138. Xu, W., Menichella, D., Jiang, H., et al.: Absence of P0 leads to the dysregulation of myelin gene expression and myelin morphogenesis. J. Neurosci. Res. 60:714, 2000. 139. Xu, W., Shy, M., Kamholz, J., et al.: Mutations in the cytoplasmic domain of P0 reveal a role for PKC-mediated phosphorylation in adhesion and myelination. J. Cell Biol. 155:439, 2001. 140. Yang, L. J. S., Zeller, C. B., Shaper, N. L., et al.: Gangliosides are neural ligands for myelin-associated glycoprotein. Proc. Natl. Acad. Sci. U. S. A. 93:814, 1996. 141. Yin, X., Crawford, T. O., Griffin, J. W., et al.: Myelinassociated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J. Neurosci. 18:1953, 1998. 142. Yin, X., Kidd, G. J., Wrabetz, L., et al.: Schwann cell myelination requires timely and precise targeting of P0 protein. J. Cell Biol. 148:1009, 2000. 143. Young, P., and Suter, U.: Disease mechanisms and potential therapeutic strategies in Charcot-Marie-Tooth disease. Brain Res. Rev. 36:213, 2001. 144. Zhu, Q., Couillard-Despres, S., and Julien, J.-P.: Delayed maturation of regenerating myelinated axons in mice lacking neurofilaments. Exp. Neurol. 148:299, 1997. 145. Zielasek, J., Martini, R., and Toyka, K. V.: Functional abnormalities in P0-deficient mice resemble human hereditary neuropathies linked to P0 gene mutations. Muscle Nerve 19:946, 1996. 146. Zoidl, G., Blass-Kampmann, S., D’Urso, D., et al.: Retroviral-mediated gene transfer of the peripheral myelin protein PMP22 in Schwann cells: modulation of cell growth. EMBO J. 14:1122, 1995.

25 Introduction to Immune Reactions in the Peripheral Nervous System HANS-PETER HARTUNG, BERND C. KIESEIER, RALF GOLD, GUIDO STOLL, AND KLAUS V. TOYKA

Immunocompetent Cellular Components of the PNS Macrophages T Lymphocytes

B Lymphocytes Mast Cells Schwann Cells Tolerance and Autoimmunity

The peripheral nervous system (PNS) has traditionally been considered as “immunologically privileged.” This view has undergone revision within recent years, and nowadays differences from other organs appear to be more quantitative than qualitative. The PNS is separated from the external environment by the blood-nerve barrier (BNB), which does restrict access of immune cells and soluble mediators to a certain degree; however, this restriction is not complete, either anatomically (it is absent or relatively deficient at the roots, in the ganglia, and in the motor terminals) or functionally. The PNS, as are most organs, is subject to immune surveillance. Activated T lymphocytes can cross the BNB irrespective of their antigen specificity, patrolling the PNS scanning for non–self-antigens, and abundant antigen-presenting cells (APCs) can be detected in peripheral nerve tissue. Molecules required for initiation of a cognate or innate immune response can be induced. Hence a local immune circuitry is in place to defend the integrity of the PNS or, when regulatory mechanisms fail to interact with the extraneural immune system, to cause its destruction. Various cell types armed with immunocompetent properties and found in normal PNS play distinct roles in immune reactions in the PNS, and their interplay is discussed in this chapter (Fig. 25–1).

B-Cell Tolerance T-Cell Tolerance Breakdown of Tolerance Termination of the Immune Response

IMMUNOCOMPETENT CELLULAR COMPONENTS OF THE PNS Macrophages A considerable number of local macrophages reside within the endoneurium of peripheral nerve. First recognized by Arvidson because of their phagocytosing capacity, resident macrophages are known to comprise up to 9% of the cellular components of normal peripheral nerve.7,8,34 Resident endoneurial macrophages of normal nerve are elongated cells stretching along the longitudinal axis of peripheral nerves and show small ramifications with two or three terminal branches at their ends. They are found close to endoneurial blood vessels but also scattered throughout the endoneurium. Moreover, these cells usually lie outside the basal lamina of the blood vessels and have a dendritic appearance.54 Their perivascular distribution at the bloodnerve interface makes them uniquely suited to act as APCs in the PNS.37 Many of these resident macrophages constitutively express major histocompatibility complex (MHC) class II molecules, complement receptor type 3 (CR3), and CD4, but the level of expression of these and additional co-stimulatory molecules can be enhanced greatly in inflammatory conditions, including Guillain-Barré syndrome (GBS) and chronic inflammatory demyelinating 559

560

Neuroimmunology of the Peripheral Nervous System

Tissue resident/endoneurial macrophage -MHC I/II -B7-1/2 Perivascular macrophage -MHC I/II -B7-1/2 T

Cytokines NO proteases

Blood-nerve-barrier Schwann cell

Damaged neurons -MHC I

Antigen presentation (MHC I/II) ICAM, cytokines, NO, proteases

polyradiculoneuropathy (CIDP).52,54 Studies suggest that endoneurial macrophages might act as sensors of pathology at early disease stages much like their central nervous system counterparts, the microglial cells.63 Macrophages are likely to be involved in virtually all steps of an immune reaction within the PNS. From early immune surveillance, antigen presentation, and activation of the cellular immune cascade throughout the evolving immune response, they serve a multitude of functions both protective and injurious: antigen-specific demyelination and axonal damage, nonspecific secondary tissue destruction, removal of debris, and regeneration.34,54 It has been well established that macrophage-mediated segmental demyelination is the pathologic hallmark of autoimmune demyelinating polyneuropathies, including classic GBS and CIDP (see Chapters 27, 98, and 99).42 Histopathologically, macrophages invade the Schwann cell basal lamina, penetrate myelin lamellae, strip myelin lamellae, and phagocytose both damaged and apparently intact myelin.7 Macrophages may be derived from circulating monocytes that invade or reside in the PNS. Hematogenous macrophages find their way into the peripheral nerve en passant with T cells following a concerted action of adhesion molecules,4 matrix metalloproteinases,55 and chemotactic signals.56 The immunologic processes leading to macrophage-mediated segmental demyelination have only partially been explored. Macrophages, in contrast to T cells, do not act in an antigen-specific manner and need to be targeted by additional mechanisms. There is good evidence that antibodies, through binding to their Fc receptors, may

FIGURE 25–1 The local immune circuitry of the peripheral nervous system (PNS). The blood-nerve barrier restricts access of immune cells and soluble mediators to a certain degree; however, activated T cells (T) can cross this barrier and scan the nervous system for non–self-antigens. Perivascular as well as tissue-resident macrophages, the latter abundantly found within the endoneurium, can act as antigen-presenting cells (APC), expressing MHC class I and II molecules and co-stimulatory molecules, such as B7-1 or B7-2. Schwann cells and damaged neurons are able to present antigens as well. Inflammatory mediators (e.g., cytokines, proteases, nitric oxide [NO]), can be released by macrophages and Schwann cells and perpetuate local inflammation within the PNS. See Color Plate

direct macrophages toward their myelin or axonal targets in autoimmune neuropathies.54 Activation of Fc receptors on macrophages results in the release of toxic mediators (oxygen radicals, nitric oxide [NO] metabolites, and proteases) that can damage the myelin sheath. Activated macrophages also strip myelin lamellae and phagocytose off both damaged and apparently intact myelin. Once within the nerve, macrophages promote inflammation by releasing pro-inflammatory cytokines including interleukin (IL)-1 and IL-6, as well as tumor necrosis factor-␣ (TNF-␣).61,75,108 They thus not only act as chief effector cells in demyelination and tissue destruction but are also intimately involved in the control of the pathogenetic process. Conversely, macrophages also contribute to the termination of the immuno-inflammatory attack through the induction of T-cell apoptosis28 and the release of antiinflammatory cytokines, including transforming growth factor (TGF)-␤1 and IL-10.53 It can be speculated that macrophages support the normal intact microenvironment of the PNS, but when disturbed by autoimmunity, they escape from normal control and develop their destructive potential as a result of the misdirected specific immune response following, for example, a preceding infection as in GBS.

T Lymphocytes T cells recognize via the T-cell receptor short linear peptide fragments in the context of MHC molecules on the cell surface. Each T cell carries its own unique receptor. Part of the

561

Introduction to Immune Reactions in the Peripheral Nervous System

T-cell receptor recognizes the foreign peptide, and part of it recognizes the self MHC molecule.44 Only a minority of T lymphocytes are able to perform this task, and a rigorous process of selection is necessary to determine those T cells that form the pool of lymphocytes to be exported from the thymus during T-cell development. This process is termed thymic education and involves both positive and negative selection.76,80,85,109 T-cell receptors are produced as transmembrane molecules; they are composed as heterodimers and consist of an ␣ and a ␤ or a ␥ and a ␦ chain, with each chain containing a variable and a constant domain. Eighty-five percent to 90% of the T cells carry the ␣/␤ T-cell receptor and 5% to 15% have ␥/␦ receptors. Whereas ␣/␤ T-cell receptors recognize a complex formed by a peptide seated within the groove of an MHC molecule, most ␥/␦ T cells do not recognize antigen in the form of peptide-MHC complexes. ␥/␦ T cells may be A

α– β–

triggered by infectious agents.68 T-cell responses can also be mounted against lipids and glycolipids, which are presented by CD1 molecules (CD1a, CD1b, CD1c, and CD1d). The CD1 proteins are distantly related to MHC class I and class II molecules and represent a small to moderate-sized family of ␣2-microglobulin–associated transmembrane proteins.13,24 Expression of CD1a and CD1b has been detected in endoneurial macrophages in inflammatory neuropathies (Fig. 25–2).51,95 T lymphocytes can be divided in two major subclasses: CD4⫹ and CD8⫹ T cells. CD4⫹ T cells usually act as helper T (Th) cells and recognize preferentially extracellular peptide antigens presented by MHC class II molecules, whereas CD8⫹ T cells are usually cytotoxic and recognize cytosolic antigens presented by MHC class I molecules. The latter are expressed on all nucleated cells. Thus any infected cell can signal to CD8⫹ T cells to get destroyed,

chain Variable region

Disulfide bonds Cell membrane

Bone marrow

Prothymocyte

Constant region

α0β0 γ 0δ0 r γ/δ r D-Jβ γ –δ–

CD3+ CD4+ CD8+

α+β+

CD3+ CD4– CD8–

γ +δ+

CD3+ CD4– CD8–

γ +δ+

Thymus

CD3– CD4– CD8–

CD3+ CD4– CD8+

α+β+

CD3+ CD4+ CD8–

α+β+

Periphery

Selection

B FIGURE 25–2 Structure of the T-cell receptor and its distribution during T-cell development. A, The T-cell receptor, expressed on the cellular surface only, is a heterodimer composed of two transmembrane glycoprotein chains, the ␣ and the ␤ chain. Both consist of a variable and a constant domain, and carry carbohydrate side chains. An alternative type of T-cell receptor is made up of different polypeptides designated ␥ and ␦. B, During T-cell development various types of T cells are selected. See Color Plate

562

Neuroimmunology of the Peripheral Nervous System

Immature effector T cell – CD4+

removing sites of pathogen replication. MHC class II molecules are expressed on the surface of professional APCs only. Subtypes of both CD4⫹ and CD8⫹ T cells engage in immunoregulatory actions and are instrumental in checking T-cell activities and in maintaining peripheral tolerance. The CD4⫹CD25⫹ regulatory T cell (Treg) population appears particularly responsive to self-antigens and functionally prevents under normal circumstances emergence or expansion of autoreactive T-cell clones.49,81,99 Once a naive CD4⫹ T lymphocyte encounters a specific antigen on the surface of a professional APC in the context of co-stimulatory molecules, such as CD80 or CD86, it gets activated, proliferates, and differentiates into an appropriately armed effector Th lymphocyte.1 A new serial encounter model of antigen recognition invokes a dynamic physical contact with T cells literally crawling over and scanning the surface of multiple APCs.45,104 Three types of such effector Th cells can be distinguished. Th1 cells manufacture effector molecules that activate macrophages, and Th2 cells generate B-cell activating effector molecules. The third group, so-called Th0 cells, from which both these functional classes of Th cells derive, also secrete molecules of both Th1 and Th2 and may therefore have a distinct effector function (Fig. 25–3).

Mature effector T cell – CD4+

+ IFN-γ IL-12 _ IL-4 IL-10

In normal nerves, T lymphocytes can only rarely be identified, although they normally traffic into and through the endoneurium of the PNS. Local activation of T cells requires antigen presentation in the context of MHC class II antigens, which can be found on a restricted number of cells in the peripheral nerves. In immune-mediated disorders the number of T cells within the PNS increases dramatically by invasion and clonal expansion in situ, underlining the important role of this cell type in the local immune response. A central role of T cells in disease pathogenesis was established by demonstrating that transfer of autoreactive neuritogenic T-cell lines into healthy recipient animals can induce the clinical, electrophysiologic, and morphologic features of classic experimental autoimmune neuritis (EAN) (see Chapter 27), a model of the human GBS.42 The pivotal role of T cells is also underscored by the preventive and suppressive effects of manipulations that eliminate or silence T lymphocytes in EAN (see Chapter 27). T-cell–derived immune responses are initiated and activated in the systemic immune compartment by the trimolecular interaction between APCs, MHC molecules, and a specific T-cell receptor, and fine tuning is done by the simultaneous perception of co-stimulatory signals,

TH0

IL-4 + IL-10 IFN-γ _ IL-12

TH1

TH2

Macrophageactivating molecules

Others

B cellactivating molecules

Others

IFN-γ TNF-α GM-CSF

IL-2 IL-3 TNF-β

IL-4 IL-5

IL-3 IL-10 TGF-β

FIGURE 25–3 T-cell activation. Once a naive CD4⫹ T cell encounters its specific antigenic epitope, displayed in the context of MHC class II gene products and sufficient levels of co-stimulatory molecules on an appropriate antigen-presenting cell, it starts to proliferate and differentiates into an immature effector cell, termed Th0. This cell type carries the potential to become either a Th1 or a Th2 cell, which differ in their spectrum of cytokine production. See Color Plate

Introduction to Immune Reactions in the Peripheral Nervous System

such as B7-1 (CD80) and B7-2 (CD86), on the cell surface of APCs.36 The latter ligands bind to the CD28 or CTLA-4 receptors on T cells (Fig. 25–4).82,87 Genesis of inflammatory lesions in the peripheral nerve requires that activated T cells cross the BNB and enter the PNS. The complex process of their homing, adhesion, and transmigration64 has been studied in EAN in great detail (see Chapter 27).28 Although breakdown of the BNB is one of the earliest morphologically demonstrable events in lesion development, a major disruption of the BNB may not be necessary. Thus it has been postulated that activated T cells, irrespective of their antigenic specificity, can traverse a structurally intact barrier to execute immune surveillance. As in the systemic immune compartment, also in the PNS T lymphocytes also encounter their target antigen, recognize appropriate MHC molecules, and perceive additional co-stimulatory signals.28 This complex interaction prompts antigen-specific T cells to divide and to undergo clonal expansion. These locally expanded T lymphocytes then may exert various effects in peripheral nerve. First, CD4⫹ T cells of the Th1 inflammatory phenotype can damage myelin by secreting pro-inflammatory and myelinotoxic cytokines and operate by recruiting and instructing macrophages to produce and release an array of toxic molecules or to engage in increased phagocytotic activity. Second, CD4⫹ T cells of the Th2 phenotype may cause B-cell proliferation and transformation into plasma cells that manufacture antibodies against peripheral

APC CD40L

B7

MHC

Antigen TCR

CD28

CD40 T cell

FIGURE 25–4 Trimolecular complex. T-cell activation requires cognate dual recognition of peptide-MHC complex and co-stimulatory molecules displayed on antigen-presenting cells (APCs). T cell–APC interaction is further strengthened by reciprocal recognition of accessory molecules, such as B7-CD28 and CD40L-CD40. See Color Plate

563

myelin components. Finally, activated CD8⫹ or perhaps a subset of CD4⫹ T lymphocytes may be directly cytotoxic to Schwann cells. Conversely, specialized subpopulations of T cells may terminate the acute immuno-inflammatory process by the secretion of anti-inflammatory cytokines, such as IL-10 and TGF-␤ or other molecules (Fig. 25–5 and Table 25–1).17,33,37,70,72,98

B Lymphocytes The unique feature of B lymphocytes is their ability to express and secrete antibodies. Antibodies specifically bind antigens both in the recognition and the effector phase of a humoral immune response. Antibodies or immunoglobulins consist of two identical heavy chains and two identical light chains that are held together by disulfide bonds. The amino terminus of each chain possesses a variable domain that binds antigen through three hypervariable complementarity-determining regions. The C-terminal domains of the heavy and the light chains form the constant regions, which define the class and subclass of the antibody and govern whether the light chain is of the ␣ or ␤ type. Moreover, the Fc part governs the binding and activation of complement, and is responsible for binding of antibodies to macrophages (discussed subsequently). In the central part of the antibody molecule is the hinge region with variable glycoconjugate side chains that may be important for some of the non–antigen-specific physicochemical properties of antibodies, including hydrophobic bonds to membrane constituents.35,78 Five different classes of immunoglobulin (Ig) antibodies (IgD, IgM, IgG, IgA, and IgE), as well as four subclasses of IgG and two subclasses of IgA, are known, each of which exhibits different functional properties. Each type of antibody can be produced as a soluble circulating immunoglobulin molecule or as a cell surface molecule anchored through a transmembrane domain in the B-cell membrane, where it acts as the B-cell receptor. Activation of B cells to produce antibodies requires the help of CD4⫹ T cells if the antigens are proteins.62 T-cell help is not required if the antigens are polysaccharides and lipids. On the interaction with CD4⫹ Th cells, and only then, B cells undergo heavy chain isotype switching, which results in the production of antibodies with heavy chains of different classes. Besides the increase in number of antibodies secreted, the affinity of individual antibodies against the stimulating antigen will increase over time, as B cells expressing higher affinity receptors on their surface are selectively activated (Fig. 25–6).40 Autoantibodies targeting structures within the PNS have been implicated in the pathogenesis of immune-mediated diseases of the peripheral nerve (see Chapter 26). An example of the role for antibodies is provided by the production of an axonal neuropathy in rabbit EAN through immunization with GM1 gangliosides.106 Antibodies can conceivably

564

Neuroimmunology of the Peripheral Nervous System

Recognition

Activation

Effector phase

APC

Macrophage activation

APC Costimulation

B cell activation, antibody production

MHC TCR CD4+

CD4+

Memory T cell

IL-2R Cytokines e.g. IL-2

Inflammation

CD4+ T cell mediated cytolysis

CD8+

Infected target cell

FIGURE 25–5 Different stages of T-cell responses. First, T-cell activation requires contact with processed antigenic epitopes associated with MHC class II molecules displayed on antigen-presenting cells (APC) and perception of additional co-stimulatory molecules. Upregulation of receptors, such as that for interleukin-2 (IL-2), occurs, and IL-2 drives T cells into clonal proliferation during the activation stage. The effector stage is diverse. Different pathways lead to activation of macrophages and B cells. CD8⫹ cells cause MHC class II–restricted target cell lysis. CD4-driven responses can culminate in inflammatory tissue injury. See Color Plate

induce myelin damage by three mechanisms: (1) on binding to the Fc receptor of macrophages, they can direct these to the putative (auto)antigenic structures and induce so-called antibody-dependent cellular cytotoxicity; (2) by opsonizing target structures, they can promote their internalization by macrophages; and (3) on binding to the antigenic epitopes, they can activate the classic complement pathway with subsequent assembly of the terminal complement complex (C5b-9). This results in pore formation allowing calcium influx, which triggers myelin-integrated proteases that degrade the myelin sheath. The complement system is the major effector arm of humoral immunity. An important role of complement in the pathogenesis of inflammatory demyelination was underlined by the observation that decomplementation of animals with cobra venom factor or

inactivation by soluble complement receptor type 1 (CR1) partly suppressed EAN. Current knowledge suggests that complement may be important in recruiting macrophages into the endoneurium, in opsonizing myelin for phagocytosis, in amplifying ongoing inflammatory reactions, and in disintegrating the myelin sheath.41,75 Besides mediating structural damage, antibodies may impair nerve impulse propagation and neuromuscular transmission when binding at or close to the node of Ranvier or at the motor terminals.15,23,28,47,59

Mast Cells There is an extensive mast cell population within the PNS; however, their physiologic role remains largely unexplained.

Introduction to Immune Reactions in the Peripheral Nervous System

565

Table 25–1. Major Cytokines and Their Function Cytokine

Producer Cell

Actions

IFN-␥

T cells, NK cells

TNF-␣ TNF-␤ TGF-␤ IL-1 IL-2 IL-4 IL-5 IL-10 IL-12

Macrophages, NK cells, T cells T cells, B cells Chondrocytes, monocytes, T cells Macrophages, epithelial cells T cells T cells, mast cells T cells, mast cells T cells, macrophages B cells, macrophages

IL-13

T cells

IL-18

Activated macrophages

IL-21

T-helper cells

IL-23

Macrophages, dendritic cells

Macrophage activation, increased expression of MHC molecules and antigen processing components, Ig class switching Local inflammation, endothelial activation Killing, endothelial activation Inhibits cell growth, anti-inflammatory Fever, T-cell activation, macrophage activation T-cell proliferation B-cell activation, IgE switch suppresses Th1 cells Eosinophil growth, differentiation Macrophage suppression Activates NK cells, induces CD4⫹ T-cell differentiation to Th1 cells B-cell growth and differentiation, inhibits macrophage inflammatory cytokine production and Th1 cells Induces IFN-␥ production by T cells and NK cells, favors Th1 induction Enhances CD8⫹ T-cell response, regulates B-cell–mediated humoral immunity Induces IFN-␥ production and CD4⫹ T-cell differentiation to Th1 cells

IFN ⫽ interferon; Ig ⫽ immunoglobulin; IL ⫽ interleukin; MHC ⫽ major histocompatibility complex; NK cell ⫽ natural killer cell; Th1 ⫽ helper T lymphocyte type 1; TGF ⫽ transforming growth factor; TNF ⫽ tumor necrosis factor.

Immunoglobulin-free light chains are known to sensitize mast cells, such that a second encounter with the appropriate antigen results in mast cell activation,79 one of the putative scenarios that might be critical in the inflamed peripheral nerve. Through degranulation, endoneurial mast cells contribute to the genesis of immune-mediated demyelination by releasing vasoactive amines and arachidonic acid–derived metabolites that augment vascular permeability and disturb BNB integrity and nerve conduction.26,28 Pharmacologic experiments in which mast cell–stabilizing drugs such as reserpine and nedocromil prevented or attenuated the animal model support the concept of a pathogenic role of mast cell degranulation in PNS inflammation.14,86

Schwann Cells The contribution of Schwann cells to the initiation and termination of an immune response in the PNS is still a matter of debate.5 In principle it has been shown that they possess all immune molecules of the trimolecular complex as a basic prerequisite to interact with invading T cells. Schwann cells in vitro constitutively express at low levels MHC class I but no significant numbers of MHC class II molecules. Upregulation of MHC class I or expression of MHC class II can be induced by stimulation with interferon-␥ (IFN-␥) or upon co-culture with

activated T cells.6 Interestingly, TNF-␣ synergizes with IFN-␥ and further increases MHC expression, which also has functional importance.30 Moreover, adhesion molecules such as intracellular adhesion molecule-1 (ICAM-1/ CD54), which are constitutively expressed on cultured Schwann cells, are upregulated by these cytokines and may exert co-stimulatory functions. Cytokine-activated Schwann cells have been shown to process exogenous P2 protein or its neuritogenic peptide for antigen presentation.30 Under certain conditions they can even present endogenous myelin proteins such as myelin basic protein to autoaggressive lymphocytes in an MHC class II–restricted manner.101 Although the relevance of MHC class II expression on Schwann cells has been questioned in EAN (discussed subsequently),84 there is suggestive evidence that Schwann cells are capable of directly engaging in this immune function in vivo.73 A possible explanation for this discrepancy between the in vitro and in vivo situation may be provided by the finding that viable, electrically active neurons exert a regulatory function on glial cells by suppressing MHC expression. Schwann cells also express molecules that can terminate T-cell inflammation by induction of apoptosis (discussed subsequently) and downregulate immune functions. Fas and its ligand are central molecules of a family of death factors that govern T-cell survival in the immune system. Schwann cells apparently do not express Fas (CD95) or Fas

566

Neuroimmunology of the Peripheral Nervous System

A

N-terminus

N-terminus VH

Variable region

VL

CH1

Fab

CH2

Constant region

Bone marrow

CL

B0

Fc CH3

Ag B′

Ab P

M

Periphery

Lymphatic organs

C-terminus

B FIGURE 25–6 B-cell development. A, Basic structure of an immunoglobulin molecule. Each immunoglobulin molecule consists of two heavy and two light chains, linked by disulfide bonds. The amino-terminal domain of each chain is variable in sequence (Fab fragment), whereas the remaining domains are constant (Fc fragment). Antibodies recognize conformational epitopes via the Fab fragment, whereas the Fc fragment mediates binding to Fc receptors on immuno-inflammatory cells. B, B cells develop in the bone marrow (B0), where potentially self-reactive cells are eliminated soon after their antigen receptor is first expressed. In the lymph node the heterogeneous repertoire of mature B cells (B⬘) encounters antigens. Upon interaction with an antigen and specific helper T cells, B cells are activated to divide. Selected B lymphocytes differentiate into plasma cells (P), which secrete large amounts of antibodies, or into long-lived memory cells (M), which contribute to lasting protective immunity. See Color Plate

ligand (FasL) constitutively but can be induced to do so by pro-inflammatory Th1 cytokines within 48 hours.12,103 Cross-linking of Fas molecules on invading T cells by membrane-bound or secreted FasL could eliminate the autoaggressive immune effectors. This could explain T-cell apoptosis observed during the natural disease course of EAN (discussed subsequently). Second, expression of Fas on Schwann cell membranes could render them susceptible to T-cell attack. Apoptotic elimination of Schwann cells is observed during EAN and may further augment demyelination in the PNS.100 Probably local secretion of TNF-␣ is involved in regulation of apoptotic Schwann cell death.

Schwann cells produce IL-1, a potent cytokine that promotes T-cell activation and proliferation,89 and also IL-6, which may bias the local cytokine milieu to a Th2 type of reaction.11 Furthermore, Schwann cells can be induced in vitro to secrete prostaglandin E2 and thromboxane A2.20 These immunomodulators may inhibit or stimulate T cells, depending on their level of production. Schwann cells are also endowed with a cytokine-inducible nitric oxide synthase that is rapidly upregulated after simultaneous treatment with IFN-␥ and TNF-␣.31 Messenger RNA was detected within 12 hours, and nitrite secretion as a measure of NO production was detectable after 24 hours. Secretion of NO by

Introduction to Immune Reactions in the Peripheral Nervous System

Schwann cells exerts a strong suppressive effect on T-cell activation in a co-culture model. Schwann cell–derived NO intermediates have the potential to limit inflammatory demyelination, unless T cells are rescued by exogenous IL-2 (discussed subsequently). Another mediator of steroid hormones, lipocortin-1 (annexin-1), may also aid in the downregulation of inflammatory reactions.29 Schwann cells carry on their surface a number of regulatory complement proteins such as CR1 (CD35), decay accelerating factor CD55, and membrane cofactor proteins CD46 and CD59.58,83,96,97 This ensemble of proteins serves to attenuate the pro-inflammatory and demyelinating properties of activated complement.40,90,94 Conversely, although macrophages constitute the major source of complement at inflammatory foci, Schwann cells can also be induced by Th1 cytokines to generate the central complement component C3.22

TOLERANCE AND AUTOIMMUNITY The random generation of a highly diverse repertoire of B and T lymphocytes carrying specific receptors allows the immune system to recognize virtually any antigen, including autoantigens. These are recognized by selfreactive T cells or autoantibodies. Tolerance is the process that eliminates or downregulates such autoreactive cells. Consequently, a breakdown in this system can cause an autoimmune response or even autoimmune disease.

B-Cell Tolerance Autoantibodies are characteristic for several autoimmune diseases of the nervous system. They bind to surface molecules or receptors, leading to functional impairment of the affected cell. They can also bind to intracellular antigens and may thereby cause disease.60 The immune system has several mechanisms available to eliminate autoreactive B lymphocytes out of the B-cell pool: (1) by clonal deletion of immature B cells in the bone marrow67; (2) by deletion of autoreactive B lymphocytes in the T-cell zones of secondary lymphoid organs, such as lymph nodes or the spleen77; (3) by induction of cellular anergy (“functional inactivation”)32; and (4) by a process called receptor editing, which changes the receptor specificity once an autoantigen has been encountered.66 At present the relative contribution of these mechanisms in preventing autoimmune disorders in the PNS remains unclear.

T-Cell Tolerance The basic principle of T-cell tolerance is the deletion of self-reactive T lymphocytes in the thymus. During the developmental process of thymic education, self-reactive

567

T lymphocytes are usually eliminated by complex mechanisms consisting of positive and negative selection. To delete all self-reactive cells, the presence of all autoantigens is required in the thymus. However, this is not always the case.50 Further, some autoreactive T cells escape thymic education and enter the systemic immune compartment. Several mechanisms are required to keep T cells in check and maintain peripheral tolerance. Immunologic Ignorance Several mechanisms can cause immunologic ignorance. Antigens may be physically isolated from self-reactive T cells as a result of their separation in different biologic compartments (e.g., by the blood-brain barrier, marking the central nervous system as a potential immunologically privileged site).10 Also, the level of antigen expressed might be below a certain threshold required for T-cell activation,2,25 or the physical encounters with APCs are not sufficiently frequent. If tissues lack professional antigen presenters, resident antigens fail to activate T cells and are therefore ignored. Importantly, autoreactive T cells in such a setting remain functionally intact. Peripheral Clonal Deletion If abundant self-antigens in the periphery continually stimulate autoreactive T cells, these may succumb to activation-induced cell death through apoptosis. The lack of growth factors may also lead to elimination of autoreactive T cells. Inhibition T cells require the presence of co-stimulatory molecules when detecting antigens presented by an MHC molecule. CD152 (or CTLA-4) on T cells binds the co-stimulatory molecules CD80 (or B7-1) and CD86 (or B7-2) with a higher affinity than the co-stimulatory receptor CD28. Thus CD152 inhibits T-cell activation16 and results in anergy (i.e., functional unresponsiveness). Anergy If T cells recognize antigens presented without adequate levels of co-stimulatory molecules, they are rendered unresponsive and cannot even launch a response when restimulated by antigen presenters expressing sufficient amounts of co-stimulators. This state is termed anergy. T-cell anergy can also be induced by altered peptide ligands that contain modified T-cell receptor contact residues. Suppression Treg cells, by the production of various cytokines, inhibit or suppress the activation of T lymphocytes. Collective evidence suggests that these CD4⫹CD25⫹ lymphocytes (Treg cells) contribute notably to endogenous mechanisms that actively regulate induced autoimmune-mediated diseases of the nervous system (Fig. 25–7).18,21,74,88,99

568

Neuroimmunology of the Peripheral Nervous System

Immune reaction

Immunologic ignorance

Inhibition

Suppression

Deletion

Antigen

APC

APC

B7-1

MHC

CD28

TCR

APC

APC

APC FAS ligand FAS

CD152 T

T

IL-10 TGF-β

T

T

T

BNB

T

T

T

T

T cell activation

No cell activation

No cell activation

No cell activation

No cell activation

FIGURE 25–7 Peripheral T-cell tolerance. Because self-reactive T lymphocytes can escape thymic education, various mechanisms control the activity of these potentially autoreactive cells in the periphery. See text for details. BNB ⫽ blood-nerve barrier. See Color Plate

BREAKDOWN OF TOLERANCE If one of the regulatory mechanisms outlined earlier fails, the specific immune response is mounted against self-antigens, which leads to expansion of autoreactive effector T cells, generation of autoantibodies through T-cell help, or both, and may give rise to severe tissue damage9—a scenario that Paul Ehrlich termed horror autotoxicus over a century ago. Autoaggressive responses eventuate autoimmune disease. The autoimmune response persists because the immune system is not able to remove the offending autoantigen from the body; even worse, new hitherto hidden autoantigens could be released to amplify the response and broaden its epitope specificity, a process termed epitope spreading. The mechanisms of tissue damage in autoimmune diseases are essentially the same as those encountered in other inflammatory disorders. One hypothesis suggests that lymphopenia, often observed in autoimmune disorders and

caused by viral infections, may provoke a compensatory homeostatic expansion of T cells, forming self-reactive immune cell populations.57 Various mechanisms are operative to promote or prevent tissue damage once an autoimmune reaction has started. Recruitment of large numbers of host monocytes/macrophages and T cells and the release of cytotoxic cytokines and chemokines would serve to augment the injurious tissue reaction. Autoantibodies, either directly with the mediation of complement or via antibody-dependent cellular cytotoxicity, are also active partners. Conversely, a suppressive/inhibitory local environment with only few or not fully competent APCs, anatomic distance of autoreactive cells to target structures, anti-inflammatory cytokines, or inappropriate antibody concentrations may help to defend against autoimmune tissue destruction. Finally, induction of apoptosis in activated T cells may eliminate the autoreactive T cells before the full inflammatory cascade is set in motion.

Introduction to Immune Reactions in the Peripheral Nervous System

In none of the disorders of the PNS thought to be autoimmune in nature has the ultimate cause or the precise sequence of pathogenic events been unequivocally established. Structural similarities between microbial and self-antigens could activate autoreactive T or B cells, a mechanism termed molecular mimicry.3,19,69,105 T lymphocytes can recognize microbial as well as self-peptides with similar amino acid sequence.27,48 In contrast, a single T-cell receptor can recognize several peptides with various degrees in sequence homology.39 The principle of molecular mimicry has been invoked in the pathogenesis of GBS following infection with Campylobacter jejuni, Mycoplasma pneumoniae, or certain viruses such as cytomegalovirus (see Chapter 26).38,43,91,92,102 Another possible role in the genesis of autoimmunity has been attributed to a group of peptides derived from viral and bacterial pathogens, so-called superantigens. They have a distinct mode of binding to MHC molecules outside the groove that enables them to activate large numbers of T or B cells.71 Bystander activation, implicating the “accidental” activation of effector immune cells, may also play an important role.46,93

TERMINATION OF THE IMMUNE RESPONSE Termination of ongoing PNS inflammation can be mediated by continuous downregulation of the vicious cycle of the amplification-effector pathways. One such mode is silencing of macrophages, which appears to be mediated via the secretion of anti-inflammatory Th2 cytokines. These may also prevail over Th1 cytokines and their cellular sources. Another mechanism is apoptosis of autoimmune T cells, which occurs during the natural disease course of EAN.107 Although the vital mediators of T-cell apoptosis during neuritis have not been identified, various effector molecules, including NO, TNF-␣, and lipocortins, may act in concert to mediate T-cell death in the inflamed PNS. Accumulating evidence suggests that molecules of the innate immune system, such as complement components and pentraxins, have a role in the removal of apoptotic cells.65

REFERENCES 1. Abbas, A. K., and Sharpe, A. H.: T-cell stimulation: an abundance of B7s. Nat. Med. 5:1345, 1999. 2. Akkaraju, S., Ho, W. Y., Leong, D., et al.: A range of CD4 T cell tolerance: partial inactivation to organ-specific antigen allows nondestructive thyroiditis or insulitis. Immunity 7:255, 1997. 3. Albert L. J., and Inman, R. D.: Molecular mimicry and autoimmunity. N. Engl. J. Med. 341:2068, 1999.

569

4. Archelos, J. J., Previtali, S. C., and Hartung, H.-P.: The role of integrins in immune-mediated diseases of the nervous system. Trends Neurosci. 22:30, 1999. 5. Armati, P. J., and Pollard, J. D.: Immunology of the Schwann cell. Baillieres Clin. Neurol. 5:47, 1996. 6. Armati, P. J., Pollard, J. D., and Gatenby, P.: Rat and human Schwann cells in vitro can synthesize and express MHC molecules. Muscle Nerve 13:106, 1990. 7. Arnason, B. G. W., and Soliven, B.: Acute inflammatory demyelinating polyradiculoneuropathy. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, Vol. 3. Philadelphia, W. B. Saunders, p. 1437, 1993. 8. Arvidson, B.: Cellular uptake of exogenous horseradish peroxidase in mouse peripheral nerve. Acta Neuropathol. (Berl.) 37:35, 1977. 9. Bach, J. F.: Autoimmune diseases as the loss of active “self-control.” Ann. N. Y. Acad. Sci. 998:161, 2003. 10. Barker, C. F., and Billingham, R. E.: Immunologically privileged sites. Adv. Immunol. 25:1, 1977. 11. Bolin, L. M., Verity, A. N., Silver, J. E., et al.: Interleukin-6 production by Schwann cells and induction in sciatic nerve injury. J. Neurochem. 64:850, 1995. 12. Bonetti, B., Valdo, P., Ossi, G., et al.: T-cell cytotoxicity of human Schwann cells: TNFalpha promotes fasL-mediated apoptosis and IFN gamma perforin-mediated lysis. Glia 43:141, 2003. 13. Brigl, M., and Brenner, M. B.: CD1: Antigen presentation and T cell function. Annu. Rev. Immunol. 22:817, 2004. 14. Brosnan, C. F., and Tansey, F. A.: Delayed onset of experimental allergic neuritis in rats treated with reserpine. J. Neuropathol. Exp. Neurol. 43:84, 1984. 15. Buchwald, B., Weishaupt, A., Toyka, K. V., and Dudel, J.: Pre- and postsynaptic blockade of neuromuscular transmission by Miller-Fisher syndrome IgG at mouse motor nerve terminals. Eur. J. Neurosci. 10:281, 1998. 16. Chambers, C. A., and Allison, J. P.: Costimulatory regulation of T cell function. Curr. Opin. Cell Biol. 11:203, 1999. 17. Chen, W., and Wahl, S. M.: TGF-beta: receptors, signaling pathways and autoimmunity. Curr. Dir. Autoimmun. 5:62, 2002. 18. Chen, Y., Kuchroo, V. K., Inobe, J.-L., et al.: Regulatory T cell clones induced by oral tolerance: suppression of autoimmune encephalomyelitis in anti-myelin basic protein T cell receptor transgenic mice. J. Exp. Med. 265:1237, 1994. 19. Christen, U., and von Herrath, M. G.: Induction, acceleration or prevention of autoimmunity by molecular mimicry. Mol. Immunol. 40:1113, 2004. 20. Constable, A. L., Armati, P. J., Toyka, K. V., and Hartung, H. P.: Production of prostanoids by Lewis rat Schwann cells in vitro. Brain Res. 635:75, 1994. 21. D’Ambrosio, D., Sinigaglia, F., and Adorini, L.: Special attractions for suppressor T cells. Trends Immunol. 24:122, 2003. 22. Dashiell, S. M., Vanguri, P., and Koski, C. L.: Dibutyryl cyclic AMP and inflammatory cytokines mediate C3 expression in Schwann cells. Glia 20:308, 1997. 23. Dilley, A., Gregson, N. A., Hadden, R. D., and Smith, K. J.: Effects on axonal conduction of anti-ganglioside sera and sera from patients with Guillain-Barre syndrome. J. Neuroimmunol. 139:133, 2003.

570

Neuroimmunology of the Peripheral Nervous System

24. Dutronc, Y., and Porcelli, S. A.: The CD1 family and T cell recognition of lipid antigens. Tissue Antigens 60:337, 2002. 25. Ferber, I., Schörich, G., Schenkel, J., et al.: Levels of peripheral T cell tolerance induced by different doses of tolerogen. Science 263:674, 1994. 26. Frossi, B., De Carli, M., and Pucillo, C.: The mast cell: an antenna of the microenvironment that directs the immune response. J. Leukoc. Biol. 75:579, 2004. 27. Fujinami, R. S., and Oldstone, M. B.: Amino acid homology between the encephalitogenic site of myelin basic protein and virus: mechanism for autoimmunity. Science 230:1043, 1985. 28. Gold, R., Archelos, J. J., and Hartung, H.-P.: Mechanisms of immune regulation in the peripheral nervous system. Brain Pathol. 9:343, 1999. 29. Gold, R., Oelschlager, M., Pepinsky, R. B., et al.: Increased lipocortin-1 (annexin-1) expression in the sciatic nerve of Lewis rats with experimental autoimmune neuritis. Acta Neuropathol. (Berl.) 98:583, 1999. 30. Gold, R., Toyka, K. V., and Hartung, H. P.: Synergistic effect of IFN-gamma and TNF-alpha on expression of immune molecules and antigen presentation by Schwann cells. Cell. Immunol. 165:65, 1995. 31. Gold, R., Zielasek, J., Kiefer, R., et al.: Secretion of nitrite by Schwann cells and its effect on T-cell activation in vitro. Cell. Immunol. 168:69, 1996. 32. Goodnow, C. C., Crosbie, J., Adelstein, S., et al.: Altered immunoglobulin expression and functional silencing of self-reactive B lymphocytes in transgenic mice. Nature 334:676, 1988. 33. Gorelik, L., and Flavell, R. A.: Transforming growth factorbeta in T-cell biology. Nat. Rev. Immunol. 2:46, 2002. 34. Griffin, J. W., George, R., and Ho, T.: Macrophage system in peripheral nerves: a review. J. Neuropathol. Exp. Neurol. 52:553, 1993. 35. Harris, L. J., Larson, S. B., and McPherson, A.: Comparison of intact antibody structures and the implications for effector function. Adv. Immunol. 72:191, 1999. 36. Hartung, H.-P.: Pathogenesis of inflammatory demyelination: implications for therapy. Curr. Opin. Neurol. 8:191, 1995. 37. Hartung, H.-P., Pollard, J. D., Harvey, G. K., and Toyka, K. V.: Invited review–immunopathogenesis and treatment of the Guillain-Barré syndrome. Parts I and II. Muscle Nerve 18:137, 1995. 38. Hartung, H.-P., van der Meché, F. G. A., and Pollard, J. D.: Guillain-Barré syndrome, CIDP and other chronic immunemediated neuropathies. Curr. Opin. Neurol. 11:497, 1998. 39. Hemmer, B., Vergelli, M., Pinilla, C., et al.: Probing degeneracy in T-cell recognition using peptide combinatorial libraries. Immunol. Today 19:163, 1998. 40. Heyman, B.: Regulation of antibody responses via antibodies, complement, and Fc receptors. Annu. Rev. Immunol. 18:709, 2000. 41. Hila, S., Soane, L., and Koski, C. L.: Sublytic C5b-9-stimulated Schwann cell survival through PI 3-kinase-mediated phosphorylation of BAD. Glia 36:58, 2001. 42. Ho, T. W., McKhann, G. M., and Griffin, J. W.: Human autoimmune neuropathies. Annu. Rev. Neurosci. 21:187, 1998.

43. Hughes, R. A. C., Hadden, R. D. M., Gregson, N. A., and Smith, K. J.: Pathogenesis of Guillain-Barré syndrome. J. Neuroimmunol. 100:74, 1999. 44. Huppa, J. B., and Davis, M. M.: T-cell-antigen recognition and the immunological synapse. Nat. Rev. Immunol. 3:973, 2003. 45. Iezzi, G., Scheidegger, D., and Lanzavecchia, A.: Migration and function of antigen-primed nonpolarized T lymphocytes in vivo. J. Exp. Med. 193:987, 2001. 46. Infante-Duarte, C., Horton, H. F., Byrne, M. C., and Kamradt, T.: Microbial lipopeptides induce the production of IL-17 in Th cells. J. Immunol. 165:6107, 2000. 47. Jacobs, B. C., O’Hanlon, G. M., Bullens, R. W., et al.: Immunoglobulins inhibit pathophysiological effects of antiGQ1b-positive sera at motor nerve terminals through inhibition of antibody binding. Brain 126:2220, 2003. 48. Jahnke, U., Fischer, E. H., and Alvord, E. C. J.: Sequence homology between certain viral proteins and proteins related to encephalomyelitis and neuritis. Science 229:282, 1985. 49. Jonuleit, H., and Schmitt, E.: The regulatory T cell family: distinct subsets and their interrelations. J. Immunol. 171:6323, 2003. 50. Kamradt, T., and Mitchison, N. A.: Tolerance and autoimmunity. N. Engl. J. Med. 334:655, 2001. 51. Khalili-Shirazi, A., Gregson, N. A., Londei, M., et al.: The distribution of CD1 molecules in inflammatory neuropathy. J. Neurol. Sci. 158:154, 1998. 52. Kiefer, R., Dangond, F., Mueller, M., et al.: Enhanced B7 costimulatory molecule expression in inflammatory human sural nerve biopsies. J. Neurol. Neurosurg. Psychiatry 69:362, 2000. 53. Kiefer, R., Funa, K., Schweitzer, T., et al.: Transforming growth factor-beta 1 in experimental autoimmune neuritis: cellular localization and time course. Am. J. Pathol. 148:211, 1996. 54. Kiefer, R., Kieseier, B. C., Stoll, G., and Hartung, H. P.: The role of macrophages in immune-mediated damage to the peripheral nervous system. Prog. Neurobiol. 64:109, 2001. 55. Kieseier, B. C., Clements, J. M., Pischel, H. B., et al.: Matrix metalloproteinases MMP-9 and MMP-7 are expressed in experimental autoimmune neuritis and the Guillain-Barre syndrome. Ann. Neurol. 43:427, 1998. 56. Kieseier, B. C., Tani, M., Mahad, D., et al.: Chemokines and chemokine receptors in inflammatory demyelinating neuropathies: a central role for IP-10. Brain 125:823, 2002. 57. King, C., Ilic, A., Koelsch, K., and Sarvetnick, N.: Homeostatic expansion of T cells during immune insufficiency generates autoimmunity. Cell 117:265, 2004. 58. Koski, C. L., Estep, A. E., Sawant-Mane, S., et al.: Complement regulatory molecules on human myelin and glial cells: differential expression affects the deposition of activated complement proteins. J. Neurochem. 66:303, 1996. 59. Krampfl, K., Mohammadi, B., Buchwald, B., et al.: IgG from patients with Guillain-Barre syndrome interacts with nicotinic acetylcholine receptor channels. Muscle Nerve 27:435, 2003. 60. Matsumoto, I., Staub, A., Benoist, C., and Mathis, D.: Arthritis provoked by linked T and B cell recognition of a glycolytic enzyme. Science 286:1732, 1999.

Introduction to Immune Reactions in the Peripheral Nervous System 61. Maurer, M., Toyka, K. V., and Gold, R.: Cellular immunity in inflammatory autoimmune neuropathies. Rev. Neurol. (Paris) 158:S7, 2002. 62. Mitchison, N. A.: T-cell–B-cell cooperation. Nat. Rev. Immunol. 4:308, 2004. 63. Mueller, M., Wacker, K., Ringelstein, E. B., et al.: Rapid response of identified resident endoneurial macrophages to nerve injury. Am. J. Pathol. 159:2187, 2001. 64. Muller, W. A.: Leukocyte–endothelial-cell interactions in leukocyte transmigration and the inflammatory response. Trends Immunol. 24:327, 2003. 65. Nauta, A. J., Daha, M. R., van Kooten, C., and Roos, A.: Recognition and clearance of apoptotic cells: a role for complement and pentraxins. Trends Immunol. 24:148, 2003. 66. Nemazee, D.: Receptor selection in B and T lymphocytes. Annu. Rev. Immunol. 18:19, 2000. 67. Nemazee, D. A., and Burki, K.: Clonal deletion of B lymphocytes in a transgenic mouse bearing anti-MHC class I antibody genes. Nature 337:562, 1989. 68. Nikolich-Zugich, J., Slifka, M. K., and Messaoudi, I.: The many important facets of T-cell repertoire diversity. Nat. Rev. Immunol. 4:123, 2004. 69. Olson, J. K., Ludovic Croxford, J., and Miller, S. D.: Innate and adaptive immune requirements for induction of autoimmune demyelinating disease by molecular mimicry. Mol. Immunol. 40:1103, 2004. 70. O’Shea, J. J., Ma, A., and Lipsky, P.: Cytokines and autoimmunity. Nat. Rev. Immunol. 2:37, 2002. 71. Perron, H., Garson, J. A., Bedin, F., et al.: Molecular identification of a novel retrovirus repeatedly isolated from patients with multiple sclerosis. Proc. Natl. Acad. Sci. U. S. A. 94:7583, 1997. 72. Pestka, S., Krause, C. D., Sarkar, D., et al.: Interleukin-10 and related cytokines and receptors. Annu. Rev. Immunol. 22:929, 2004. 73. Pollard, J. D., Baverstock, J., and McLeod, J. G.: Class II antigen expression and inflammatory cells in the GuillainBarre syndrome. Ann. Neurol. 21:337, 1987. 74. Powrie, F., Carlino, J., Leach, M. W., et al.: A critical role for transforming growth factor-beta but not interleukin 4 in suppression of T helper type-1 mediated colitis by CD45RB(low) CD4⫹ T cells. J. Exp. Med. 183:2669, 1996. 75. Putzu, G. A., Figarella-Branger, D., Bouvier-Labit, C., et al.: Immunohistochemical localization of cytokines, C5b-9 and ICAM-1 in peripheral nerve of Guillain-Barre syndrome. J. Neurol. Sci. 174:16, 2000. 76. Rathmell, J. C., Thompson, C. B.: The central effect of cell death in the immune system. Annu. Rev. Immunol. 17:781, 1999. 77. Rathmell, J. C., Townsend, S. E., Xu, J. C., et al.: Expansion or elimination of B cell in vivo: dual roles for CD40- and Fas (CD95)-ligands modulated by the B cell antigen receptor. Cell 87:319, 1996. 78. Ravetch, J. V., and Bolland, S.: IgG Fc receptors. Annu. Rev. Immunol. 19:275, 2001. 79. Redegeld, F. A., Van Der Heijden, M. W., Kool, M., et al.: Functional role for Ig free light chains in immediate and delayed hypersensitivity responses. Inflamm. Res. 53(Suppl. 1):S6, 2004.

571

80. Sakaguchi, S.: Policing the regulators. Nat. Immunol. 2:283, 2001. 81. Sakaguchi, S.: Naturally arising CD4⫹ regulatory T cells for immunologic self-tolerance and negative control of immune responses. Annu. Rev. Immunol. 22:531, 2004. 82. Salomon, B., and Bluestone, J. A.: Complexities of CD28/B7: CTLA-4 costimulatory pathways in autoimmunity and transplantation. Annu. Rev. Immunol. 19:225, 2001. 83. Sawant-Mane, S., Piddlesden, S. J., Morgan, B. P., et al.: CD59 homologue regulates complement-dependent cytolysis of rat Schwann cells. J. Neuroimmunol. 69:63, 1996. 84. Schmidt, B., Toyka, K. V., Kiefer, R., et al.: Inflammatory infiltrates in sural nerve biopsies in Guillain-Barré syndrome and chronic inflammatory demyelinating neuropathy. Muscle Nerve 19:474, 1996. 85. Sebzda, E., Mariathasan, S., Ohteki, T., et al.: Selection of the T cell repertoire. Annu. Rev. Immunol. 17:829, 1999. 86. Seeldrayers, P. A., Yasui, D., Weiner, H. L., and Johnson, D.: Treatment of experimental allergic neuritis with nedocromil sodium. J. Neuroimmunol. 25:221, 1989. 87. Sharpe, A. H., and Freeman, G. J.: The B7-CD28 superfamily. Nat. Rev. Immunol. 2:116, 2002. 88. Shevach, E. M.: Regulatory T cells in autoimmunity. Annu. Rev. Immunol. 18:423, 2000. 89. Skundric, D. S., Lisak, R. P., Rouhi, M., et al.: Schwann cell-specific regulation of IL-1 and IL-1Ra during EAN: possible relevance for immune regulation at paranodal regions. J. Neuroimmunol. 116:74, 2001. 90. Song, W.C.: Membrane complement regulatory proteins in autoimmune and inflammatory tissue injury. Curr. Dir. Autoimmun. 7:181, 2004. 91. Susuki, K., Nishimoto, Y., Yamada, M., et al.: Acute motor axonal neuropathy rabbit model: immune attack on nerve root axons. Ann. Neurol. 54:383, 2003. 92. Susuki, K., Odaka, M., Mori, M., et al.: Acute motor axonal neuropathy after Mycoplasma infection: evidence of molecular mimicry. Neurology 62:949, 2004. 93. Tough, D. F., Sun, S., and Sprent, J.: T cell stimulation in vivo by lipopolysaccharide (LPS). J. Exp. Med. 185:2089, 1997. 94. Tsokos, G. C., and Fleming, S. D.: Autoimmunity, complement activation, tissue injury and reciprocal effects. Curr. Dir. Autoimmun. 7:149, 2004. 95. Van Rhijn, I., Van den Berg, L. H., Bosboom, W. M., et al.: Expression of accessory molecules for T-cell activation in peripheral nerve of patients with CIDP and vasculitic neuropathy. Brain 123(Pt. 10):2020, 2000. 96. Vedeler, C. A., Conti, G., Fujioka, T., et al.: The expression of CD59 in experimental allergic neuritis. J. Neurol. Sci. 165:154, 1999. 97. Vedeler, C., Ulvestad, E., Bjorge, L., et al.: The expression of CD59 in normal human nervous tissue. Immunology 82:542, 1994. 98. Walker, L. S.: CD4⫹ CD25⫹ Treg: divide and rule? Immunology 111:129, 2004. 99. Walker, L. S., and Abbas, A. K.: The enemy within: keeping self-reactive T cells at bay in the periphery. Nat. Rev. Immunol. 2:11, 2002. 100. Weishaupt, A., Bruck, W., Hartung, T., et al.: Schwann cell apoptosis in experimental autoimmune neuritis of the Lewis

572

101.

102.

103.

104.

Neuroimmunology of the Peripheral Nervous System rat and the functional role of tumor necrosis factor-alpha. Neurosci. Lett. 306:77, 2001. Wekerle, H., Schwab, M., Linington, C., and Meyermann, R.: Antigen presentation in the peripheral nervous system: Schwann cells present endogenous myelin autoantigens to lymphocytes. Eur. J. Immunol. 16:1551, 1986. Wim Ang, C., Jacobs, B. C., and Laman, J. D.: The GuillainBarre syndrome: a true case of molecular mimicry. Trends Immunol. 25:61, 2004. Wohlleben, G., Ibrahim, S. M., Schmidt, J., et al.: Regulation of Fas and FasL expression on rat Schwann cells. Glia 30:373, 2000. Wolf, K., Muller, R., Borgmann, S., et al.: Amoeboid shape change and contact guidance: T-lymphocyte crawling through fibrillar collagen is independent of matrix remodeling by MMPs and other proteases. Blood 102:3262, 2003.

105. Wucherpfennig, K. W., and Strominger, J. L.: Molecular mimicry in T-cell mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein. Cell 80:695, 1995. 106. Yuki, N., Yamada, M., Koga, M., et al.: Animal model of axonal Guillain-Barré syndrome induced by sensitization with GM1 gangliosides. Ann. Neurol. 49:712, 2001. 107. Zettl, U. K., Gold, R., Toyka, K. V., and Hartung, H. P.: In situ demonstration of T cell activation and elimination in the peripheral nervous system during experimental autoimmune neuritis in the Lewis rat. Acta Neuropathol. (Berl.) 91:360, 1996. 108. Zhu, J., Mix, E., and Link, H.: Cytokine production and the pathogenesis of experimental autoimmune neuritis and Guillain-Barre syndrome. J. Neuroimmunol. 84:40, 1998. 109. Zuniga-Pflucker, J. C.: T-cell development made simple. Nat. Rev. Immunol. 4:67, 2004.

26 Peripheral Nerve Antigens HUGH J. WILLISON, NORMAN A. GREGSON, GRAHAM M. O’HANLON, AND RICHARD A. C. HUGHES

Introduction Immune Responses in Peripheral Nerve Molecular Composition of Peripheral Nerve Cell Types, Subcellular Topographic Compartments, and the Blood-Nerve Barrier Schwann Cell and Myelin Antigens Neuronal Antigens Neurofilament Antigens Gangliosides Ganglioside Metabolism and Function

Voltage-Gated Calcium and Potassium Channels Molecular Mimicry and Microbial Antigens Campylobacter Infection Other Infections Molecular Mimicry Pathophysiologic Relevance of Autoantigens in Animal Models Galactocerebroside-Induced EAN MAG-Induced Neuropathy Models Ganglioside-Induced Neuropathy Models

INTRODUCTION The peripheral nervous system (PNS) is composed of a diverse array of cell types within various specialized compartments. It is an immunologically complex structure, and many sites within the PNS are subject to a wide range of inflammatory and autoimmune responses. Chapter 25, on immunologic responses, describes in detail the underlying immunologic principles pertaining to the PNS, and Chapter 97 describes the clinical immunology of PNS disease. The present chapter focuses particularly on a description of the PNS antigens that are capable of acting as autoimmune targets, and readers are referred to other chapters for further immunologic and clinical information. Some immunologic issues of special relevance to peripheral nerve, particularly innate immune mechanisms and molecular mimicry, are also included herein. The PNS contains many molecules potentially capable of acting as antigens that are either unique to the PNS, such as some of the myelin proteins, or highly enriched in it, such as gangliosides. Additionally, the PNS contains a large number of molecules that are common to other sites, both in the central nervous system (CNS), such as the

Pathophysiologic Relevance of Autoantigens in Human Neuropathies Neuronopathies Axonopathies Inflammatory Demyelinating Polyradiculoneuropathies Ganglioside Antibody–Related Diseases Galactocerebroside Charcot-Marie-Tooth Disease Paraproteinemic Neuropathies Protein Antigens Channelopathies

neuronal antigens Hu and CV2, and in non-neural regions of the body, such as basement membrane antigens. However, out of the many thousands of candidate antigens, only a limited number of molecular structures have been convincingly demonstrated to behave as peripheral nerve autoantigens. This chapter includes a description of these antigens, along with the relevant morphologic, pathophysiologic, and immunologic background, and the features of the antigenic repertoire within the PNS that make it an important site for both B- and T-cell responses to its constituent components.

IMMUNE RESPONSES IN PERIPHERAL NERVE The role of the immune system is to recognize the presence of foreign, usually microbial, material within an organism and to provide protection by the destruction and removal of this material. In autoimmune neuropathy, one of the major principles underlying nerve injury is a breakdown of this system, such that elements of the nerve are inadvertently seen as foreign. The provision of resistance 573

574

Neuroimmunology of the Peripheral Nervous System

to infection by the immune system involves the correct functioning of a large number of genes whose dysregulation can result in autoimmunity.294 In higher vertebrates the immune system can be divided into the innate and the adaptive subsystems, both being required for effective protection against infection and for aberrant development of autoimmune reactions. Central to this function is selftolerance, an immunologic awareness of self and the avoidance of a response against healthy self. This process is more complex than an intrinsic inability to react to selfantigens, since clearly this occurs both naturally and normally for peripheral nerve antigens. Antigen receptors of T and B cells show a theoretically infinite variability. T cells with stable receptors are controlled by exposure to selfantigens presented by major histocompatibility complex (MHC) molecules on the surface of thymic epithelia, dendritic cells, or macrophages. After maturation, cells showing strong binding or no binding with self-peptides and relevant MHC become apoptotic or anergic. Cells showing intermediate reactivity are preserved to form the population of “naïve” cells.203 B-cell development differs in that immunoglobulin gene rearrangement occurs both before and after exposure to antigen, and cells within the lymph nodes stimulated by antigen and T helper cells undergo further gene rearrangement and V-gene hypermutation, leading to antibody isotype switching and affinity maturation. Cells reactive with self-antigens normally fail to receive sufficient co-stimulation and become apoptotic or anergic. However, some B cells, particularly CD5 cells, do produce antibodies with broad reactivity that can react with certain autoantigens, including peripheral nerve glycolipids. For a new antigen to selectively stimulate and clonally expand T and B cells, the antigen must be presented under conditions that will optimize the interaction. The rapid recognition of foreign antigens by the molecular and cellular components of the innate system promotes inflammation and antigen processing and so promotes the adaptive response. The generation of a mature immune response takes place in the lymphoid organs, but the restimulation of memory cells and their rapid expansion can also occur in the peripheral tissues. The PNS, like the vascular system, penetrates all parts of the body and tissues, including the blood vessels. It therefore might be expected to have a high risk of trauma and infection. Nonetheless, like the CNS, it is relatively isolated from the immune system. The endoneurial vascular bed is provided by penetrating arterioles that feed an extensive pericyte-covered capillary network leading to small collecting venules. There are no large endoneurial veins. Tight junctions seal the endoneurial blood vessels and the perineurium, so that serum proteins, including antibodies, and passively migrating cells cannot access the nerve, except at peripheral nerve terminals and the dorsal root ganglia (DRGs). As in the CNS, the basal level of adhesion molecule expression on the endoneurial

endothelia is sufficient to allow some activated T cells to enter the nerve irrespective of their antigen specificity.250 Also like the CNS, peripheral nerves do not contain lymphatic vessels, and so the passage of immunocompetent cells from the endoneurium to the lymphatic organs is indirect via the blood.152 These structural features, and the absence of dendritic cells, suggest that the development of a primary immune response arising from self or foreign endoneurial antigens would be unlikely to occur. However, autoimmune neuritis can be induced experimentally by the transfer of activated T cells specific for certain nerve antigens into healthy syngeneic animals, implying that there is always some autoantigen presentation within the endoneurial compartment. It is not known which cells may be responsible for endoneurial presentation of antigens. Unlike the CNS, peripheral nerves are only sparingly populated by myeloid cells, and these are normally downregulated. Schwann cells do not normally express MHC molecules, but they clearly do so in certain infections such as leprosy, and they can also express the related molecules of the CD1 family.141,326 It is possible that the perivascular pericytes are the antigen-presenting cells in normal resting nerve. Both macrophages and mast cells are found in healthy nerves, and endoneurial macrophages are downregulated, although not to the same extent as brain microglia. Macrophage numbers increase markedly on injury and during inflammation, and there is even an increase in both number and level of activation of macrophages in the normal contralateral nerve after crushing one sciatic nerve.244 Nonetheless, the induced response is monophasic and brief, suggesting efficient suppression of unwanted autoimmune responses. Active induction of autoimmune neuritis is possible in mice and rats with the peripheral myelin proteins P2, P0, and PMP22 (see below), but this is highly strain dependent, indicating a strong genetic element in the susceptibility. The functioning of the innate immune system depends on soluble and cell surface receptors and effector molecules. Some of these are homologues of proteins found in invertebrates, where they may have similar functions. In general, mammals show an increase in diversity of these receptors and molecules, often as a result of gene duplication. The receptors, assisted by soluble factors such as lectins, complement, and antibodies, detect a variety of products from bacteria and viruses and result in the activation of the effector cells of the innate system, including macrophages, polymorphonuclear cells, natural killer (NK) cells, dendritic cells, mast cells, some cytotoxic T cells, and some B cells. Most of these cells bear more than one type of receptor, and the pattern of ligand binding to these can provide a pathogen-specific response, particularly demonstrated by the Toll receptors.319 The full interaction of myelomonocytic cells with other cells and microorganisms, as with the interactions of T cells and NK

575

Peripheral Nerve Antigens

cells, requires a clustering of a variety of receptors to facilitate phosphorylation of their cytoplasmic domains and further signaling.20,316 The effective functioning of the immune system in infection is further diversified by the fact that so many pathogens have acquired and evolved defensive mechanisms that often rely on functional mimicry of receptors in both the innate and adaptive systems (see below). Although certain receptors/ligands may be defined as critical for responsiveness, full function requires additional molecules. Through the activation of complement, many of the soluble factors, such as lectins and C-reactive protein, increase opsonization and facilitate phagocytosis by myeloid and polymorphonuclear cells. The blood-nerve barrier would normally exclude these proteins from the endoneurium, along with antibodies and complement proteins. The innate immune system strongly promotes the inflammatory process and the adaptive response. In autoimmune disease, the target tissue is usually defined by antibody or T cells but the development of inflammation in the tissue, and hence often the symptoms, is highly dependent on the activity of components of the innate system.108 Thus the extent of inflammation seen in acute and chronic anti-GM1 antibody–associated syndromes varies considerably, being high in the former (Guillain-Barré syndrome [GBS] associated with highaffinity immunoglobulin G [IgG]) and low in the latter (multifocal motor neuropathy [MMN] associated with lowaffinity immunoglobulin M [IgM]). Similarly, in chronic demyelination associated with anti–myelin-associated glycoprotein (MAG) IgM paraproteinemia, both antibody and complement are bound to myelin sheaths,320 but there is little or no inflammation and the development of the disease is very slow.92 The basis of self-tolerance in cells of the innate immune system is not indifference to self; rather, self-tolerance results from the inhibitory influence of a number of receptors reacting with surface molecules expressed normally in the tissues. Molecules that are currently considered to provide a “self ” signature for macrophages, NK cells, and other members of the innate system are CD200 (Ox-2), CD47, MHC class I molecules, and a variety of sialic acid–bearing glycoproteins and glycolipids. These molecules are expressed widely throughout all tissues, while their cognate receptors, CD200R, signal-regulatory protein (CD172a), NK inhibitory receptors, and the family of siglecs, are expressed by the innate effector cells.107,261 CD47 binding affects phagocytosis as well as dendritic cell maturation and polymorphonuclear cell function.60,180,235 Both CD200 and CD47 are prominent in the nervous system and found mainly on neurons, and their presence in the CNS contributes to the characteristic downregulated state of the endogenous myeloid cells, the microglia. There is little information on the status of these molecules in the peripheral nerves and ganglia.

The siglec family, which binds sialic acid–containing glycoconjugates, has at least 10 members, including sialoadhesin, CD22, CD33, and MAG,52 and may play a role in adjusting the activation threshold of immune cells. In quiescent cells they are mainly cis-engaged, but they become released by activation signals and reengaged by sialyl conjugates on neighboring normal cells. If neighboring cells do not express normal sialic acid conjugates (e.g., after exposure to bacterial or viral sialidase), then reengagement is not possible and activation becomes more probable. As described below, some infective agents that have acquired surface sialic acid may have subverted this function. It is thus clear that many properties of the innate immune system have an important bearing on any consideration of the regulation of peripheral nerve antigen responses, along with the more traditional considerations of adaptive T- and B-cell responses.

MOLECULAR COMPOSITION OF PERIPHERAL NERVE Cell Types, Subcellular Topographic Compartments, and the Blood-Nerve Barrier The PNS can be subdivided into functional, structural, or regional compartments that are all relevant to a consideration of antigenic components (Fig. 26–1).346 In functional terms, the PNS comprises motor, sensory, and autonomic components. In the case of motor neurons, cell bodies within the spinal cord extend axonal projections via the nerve trunks, to reach motor nerve terminals up to 1 m away. The cell body, myelin, and axon compartments are each relatively protected by the blood-brain and bloodnerve barriers and by the basal laminae and membranes of myelinating Schwann cells. Only the motor nerve terminal lies entirely outside the blood-nerve barrier, although it is itself protected by a cap of perisynaptic Schwann cells. Immunologic attack may be primarily directed at any of these sites to create distinctive phenotypes, and the wide range of motor nerve syndromes offers appropriate examples of this segregation. For example, certain autoimmune neuropathies principally target motor nerve myelin, such as MMN, or motor nerve axons, such as acute motor axonal neuropathy (AMAN). Furthermore, motor nerve involvement may be topographically restricted to particular anatomic sites, such as the restriction of the motor deficit to cranially innervated muscles in Miller Fisher syndrome (MFS). In the sensory arm of the PNS, the diversity of sensory fiber types and sizes, the different types of sensory ending, and the presence of the DRG with peripheral axons and central projections into the dorsal horn of the spinal cord contribute to an equal, or indeed greater, level of complexity than that seen in the motor arm. Autonomic nerves

576

Neuroimmunology of the Peripheral Nervous System

PERIPHERAL NERVE ANTIGEN COMPARTMENTS Dorsal root

Dorsal root ganglion

Nerve bundles Satellite cell Neuronal cell bodies Schwann cell basal lamina

Ventral root

Axon Adaxonal membrane Compact myelin

Nerve terminal

Periaxonal space

Internode

Abaxonal membrane

Node

Para node

Perisynaptic Schwann cell

Internode

Pre-synaptic membrane Post-synaptic membrane

Paranodal Schmidt-Lanterman loops incisure

Muscle Active zone

are also regarded as PNS components. Thus a multitude of clinical syndromes may selectively involve particular sites, based on antigen distribution, the accessibility to circulating immunopathogenic factors, and the innate immune considerations described above. In addition to the peripheral nerve terminals, areas of relative vulnerability to circulating immune factors are the dorsal and ventral roots and the DRG, owing to the partial deficiency of the bloodnerve barrier at these sites. Thus there may be a predilection for immune injury in the nerve roots that is independent of antigen composition, in that there is no evidence that the antigenic profile of the roots differs very substantially from more distal portions of the equivalent peripheral nerve trunk. Within this broad architecture, a number of topographic compartments exist. The PNS contains abundant blood vessels with microvascular endothelial cells that form the tight junctions characteristic of the blood-nerve barrier described above. The integrity of the endothelial layer is supported by endoneurial pericytes that form the basement membrane. Clearly such vascular structures are vital to the immunopathology of peripheral nerve, and may also be relevant to peripheral nerve vasculitis, but antigens expressed in these sites have not been studied in any detail. Anti–endothelial cell antibodies have been found to

FIGURE 26–1 Schematic diagram of the main antigen compartments in the peripheral nervous system. Regions that are potentially vulnerable to autoimmune attack and locations at which differential motor or sensory symptoms could be generated are indicated. (Adapted from Willison H. J., and O’Hanlon, G. M.: Anti-glycosphingolipid antibodies and Guillain-Barré syndrome. In Nachamkin, I., and Blaser, M. J. [eds.]: Campylobacter. Washington, DC, ASM Press, p. 259, 2000.)

a greater extent in rheumatoid arthritis complicated by peripheral neuropathy than in uncomplicated rheumatoid arthritis.270 However, a direct role for antigen-specific cellular or humoral mechanisms in the formation of vasculitic lesions in endoneurial blood vessels may be minor. The endoneurial compartment, containing intrafascicular nerve fibers, comprises connective tissue rich in collagen and extracellular matrix material, and is surrounded by the perineurium. Within the endoneurial compartment lie the nerve axons enveloped by Schwann cells and their basal laminae. The basal lamina is highly enriched in laminin-2, which binds dystroglycans and 62 integrin.318 Although basement membrane antigens, including various laminins, integrins, glycosaminoglycans, and collagens, are recognized as important autoantigens in several organ-specific autoimmune diseases, they have not been widely identified as such in autoimmune neuropathy syndromes. A single study has demonstrated the presence of antibodies to heparan sulfate glycosaminoglycans in up to one third of patients with GBS and other inflammatory neuropathies.246 Among the different fiber types, it is clear that there are features of the molecular organization common to all, and also fiber type specializations, that may both be of relevance to autoimmune neuropathy. All myelinated fibers contain

Peripheral Nerve Antigens

very high concentrations of ion channels at the node of Ranvier, particularly voltage-gated sodium channels in the nodal axolemma and rectifying potassium channels in juxtaparanodes, and the extent to which these channels act as antigens is a subject of much current research. Gangliosides are also believed to be highly expressed in nodal and paranodal regions, as discussed below.

Schwann Cell and Myelin Antigens Peripheral nerve myelin membranes are synthesized and maintained by the Schwann cells that enwrap large-caliber axons. These specialized membranes are key sites for the location of neuropathy-specific antigens, as described in detail below. Both myelinating and nonmyelinating Schwann cells are highly polarized, containing an adaxonal membrane that is separated from the axonal membrane by the periaxonal space, and an abaxonal membrane surrounded by the basal lamina. The myelin sheath also develops specializations associated with noncompact myelin at paranodal regions, Schmidt-Lanterman incisures, and the inner and outer mesaxons. The apposed cytoplasmic and extracellular surfaces of myelin-forming regions of the membrane compress to extrude almost all cytoplasmic components and extracellular fluid, resulting in the multilamellar structure that is characteristic of the internodal myelin sheath. This highly compacted membrane forms the bulk of the peripheral nerve myelin and has been the subject of extensive biochemical, metabolic, and immunologic investigation. It is believed that the molecular composition of compact myelin is common to all fiber types, although it remains possible that there could be subtle differences in the content of particular antigens, including gangliosides, as described below. The compact internodal myelin membrane is not directly accessible to antibody present in the extracellular space because it is in direct continuity with, and completely enveloped by, the Schwann cell plasma membrane and its mesaxonal and paranodal specializations. It is these latter membranes that form the outermost surface of the myelin sheath and are directly accessible to antibody. Penetration of soluble molecules from the extracellular fluid into the apposed membranes of compact myelin is limited. However, the intraperiod line is quite mobile and the extracellular surfaces can be separated by changes in pH and ionic composition, and it is clear that under some circumstances, such as that seen in the IgM paraproteinemic neuropathy associated with anti-MAG antibodies, antibody can penetrate into this compartment in surprisingly large amounts.192 Thus the concept of cryptic sites and the compartmentalization of antigens as a majorly limiting factor, even in compact myelin, should be viewed with some caution. The biochemical composition of isolated peripheral nerve myelin is remarkably simple, but also unique, having

577

a high proportion of lipid and relatively few individual protein components. This high lipid content is responsible for the low buoyant density of myelin, a property that has been exploited to prepare highly purified multilamellar myelin by a combination of density gradient and differential centrifugation. Protein accounts for 25% to 38% of the dry weight of peripheral nerve myelin, with the remaining approximately 70% being lipid with a high proportion of cholesterol and glycolipid, indicating the unique nature of the membrane. The importance of PNS myelin lipid, and in particular glycolipids and gangliosides, as antigens has been a major focus of research in recent years. The major myelin lipids are cholesterol, ethanolamine glycerophosphatide, sphingomyelin, and galactocerebroside. Ethanolamine plasmalogens, serine and choline glycerophosphatides, and sulfatides are also present but in smaller amounts. Isolated peripheral nerve myelin also contains low levels of several different gangliosides and complex neutral glycolipids that have been implicated as antigens in autoimmune neuropathy. These are discussed extensively below. Polyacrylamide gel electrophoresis (PAGE) of myelin proteins reveals one major band, the glycoprotein P0 (molecular weight [MW] 28 kDa). There are two less prominent bands, myelin basic protein (MBP), also termed P1 (MW 17 kDa), and P2 (MW 15 kDa), so called because of their position of migration in PAGE (Fig. 26–2). Several minor bands of immunologic interest can also be revealed by detailed separative techniques, including MAG (MW ~ 110 kDa) and peripheral myelin protein 22 (PMP22), so named because of its MW (22 kDa). Each of these proteins is capable of inducing an immune response following injection into experimental animals, and antibody and T-cell responses to some of these proteins have been reported in human inflammatory neuropathy. The P0 and PMP22 proteins are confined to compact myelin, but they are transmembrane proteins with extracellular domains that may be accessible to antibodies and immune cells (Fig. 26–3).274 MBP and P2 protein are relatively inaccessible, being confined to the major dense line of compact myelin corresponding to the apposed intracellular surfaces of the Schwann cell membrane. MAG is a transmembrane protein with a large, accessible extracellular domain, and is located in areas of the myelin sheath where Schwann cell cytoplasm is present. P0 Glycoprotein The P0 glycoprotein accounts for over 50% of peripheral nerve myelin protein and is absent from CNS myelin. Its primary sequence of 219 amino acid residues is highly conserved among species.269 It is a primitive member of the immunoglobulin supergene family, having a single IgG-like extracellular domain with a single glycosylation site, a very hydrophobic transmembrane domain, and a basic cytoplasmic domain.178 The extracellular domain is

578

Neuroimmunology of the Peripheral Nervous System

Bovine myelin CNS proteins

CNS

PNS

Rat myelin CNS

PNS

PNS proteins

MAG*

MAG*

P0 PLP 21.5K MBP

21.5K MBP 18.5K P1

18.5K MBP 17.0K MBP

17.0K MBP 14.0K MBP, PR P2

14.0K MBP

glycosylated at Asn93 with an N-linked nonasaccharide that contains the HNK-1 epitope, also present on PMP22 and MAG.36 There are three differences in the amino acid sequence of the extracellular domain of the rat P0 protein

FIGURE 26–2 Polyacrylamide gel electrophoresis (PAGE) in the presence of sodium dodecyl sulfate of bovine and rat central and peripheral nervous system myelin proteins. The proteins are each identified by size or name or both. The PAGE was performed according to the procedure of Laemmli.169 The approximate location of myelin-associated glycoprotein is also shown.

from that of the human, which would be expected to alter the tertiary and quaternary structure.29 The crystal structure of purified extracellular domain of P0 is consistent with the formation of tetramers at the extracellular surface

Extracellular Po

Intracellular

PMP22

MAG S MBP Gal-C S Sulfatide

Non-compact myelin

Compact myelin

FIGURE 26–3 Schematic depiction of the localization of peripheral nervous system myelin antigens. The dispositions of protein P0 tetramers, PMP22 dimers, and MBP monomers, as well as the glycolipids galactocerebroside and sulfatide, are shown. (Adapted from Scherer, S. S., and Arroyo, E. J.: Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Syst. 7:1, 2002.)

Peripheral Nerve Antigens

that associate with tetramers on the opposing surface (see Fig. 26–3). Both homophilic interactions between the proteins and interaction of protruding tryptophan residues and negatively charged lipids govern the compaction of the extracellular surfaces of myelin and maintain its spacing.281 The recent successful expression of the human protein in Escherichia coli and improvements in purification will assist future investigations of the antigenic properties of P0.29 Mice carrying a null mutation and lacking P0 do not form compact myelin, and mice lacking one P0 gene develop a late-onset demyelinating neuropathy, suggesting that the appropriate amount of P0 is important for myelin maintenance.276 Mutations of the P0 gene in humans cause hereditary motor and sensory neuropathy whose severity depends on the position of the mutation. The resulting clinical picture may be congenital hypomyelinating neuropathy, childhood-onset Déjérine-Sottas disease, or later onset Charcot-Marie-Tooth (CMT) disease type 1b (see Chapter 71).178 There is evidence that immune responses are important in controlling the phenotype of P0-deficient mice. Martini and colleagues crossed mice heterozygously deficient for P0 with null mutants for the recombinant activating gene 1 or tumor necrosis factor-a receptor. Both these mutants had severely impaired T-cell responses and less severe neuropathy.200 PMP22 Protein PMP22 is a minor peripheral nerve myelin protein of great importance because its gene is duplicated in the most common form of CMT disease and deleted in hereditary neuropathy with liability to pressure palsies.293 The C- and N-terminal domains are thought to be intracellular, and it is usually considered as a transmembrane protein with four membrane-spanning regions and two extracellular domains, although alternative structures in which the second and third hydrophobic domains are also extracellular have been proposed.309 The first extracellular domain has 38 amino acid residues and is variably glycosylated, with the carbohydrate portion including the HNK-1 epitope shared by P0 and MAG, suggesting a cell adhesion function.292 The second extracellular domain has 14 amino acid residues and is not glycosylated.69,178 In the rat, expression of PMP22 begins soon after birth and reaches adult quantities after 3 weeks.293 It is glycosylated at Asn41 of the first extracellular domain, and is then translocated to the cell membrane.69 Nonglycosylated protein does not reach the cell membrane and is broken down. Its function is not entirely clear; mice carrying null mutations for PMP22 do myelinate, but myelination is delayed.2 The protein is therefore likely to be important in stabilizing myelin. It has been shown to complex with P0 at the cell surface in HeLa cells expressing both proteins, an interaction that was not mediated by carbohydrate.68 In addition, PMP22 probably plays a role in regulating

579

Schwann cell proliferation because synthesis coincides with cessation of Schwann cell division.214 The gene for PMP22 was originally described as the growth arrest– specific gene (gas-3), which is expressed by resting but not proliferating mouse fibroblasts.295 The messenger RNA (mRNA) for PMP22 is present in many tissues, although the amounts are lower than in myelin and the promoter driving its transcription in myelin differs from that in other tissues.31 The PMP22 protein is a component of tight junctions between epithelial cells, where it is closely associated with the transmembrane junctional proteins zonula occludens 1 and occludin.226 It might therefore be important in the function of Schwann cell tight junctions. P2 Protein The P2 protein, forming 2% to 15% of myelin protein, is a basic protein confined to peripheral nerve myelin, and of its 131 amino acids, there are large sequence differences at nine positions among different mammalian species. It is situated in the period line of myelin corresponding to the apposed intracellular surfaces of the Schwann cell. Xray crystallography and modeling predict a barrel-shaped molecule composed of beta pleated sheets surrounding a space that might contain fatty acid, and it may therefore be involved in fatty acid transport.132 There are no reports of mutants lacking P2 or of mutation causing animal or human neuropathy. Its intracellular location renders it relatively inaccessible to the immune system, but if the myelin sheath is disrupted and phagocytosed, P2 might be processed by antigen-presenting cells and become accessible to the T-cell receptor. P2 has been studied extensively as an antigen in relation to experimental allergic neuritis (EAN), as described in Chapter 27. Myelin Basic Protein PNS myelin contains another basic protein, previously called P1, that was found to have the same amino acid sequence as MBP in the CNS181 and accounts for 5% to 15% of PNS myelin protein, about half of that in the CNS. The major form has a molecular weight of 18 kDa, but there are several isoforms with molecular weights ranging from 14 to 21 kDa, arising from differences in splicing of the primary transcript and variations in glycosylation and phosphorylation. Partial loss of the MBP gene in the shiverer mutant mouse causes failure to synthesize compact CNS myelin but does not prevent the formation of normal peripheral nerve myelin.39 Like P2, MBP is located in the period line, where it may play a role in myelin compaction (see Fig. 26–3), but the normality of peripheral nerve myelin in shiverer mice suggests that this role is redundant or readily replaced by another molecule, perhaps P2. It would not be expected to be accessible to the immune system unless the integrity of the Schwann cell basal lamina and Schwann cell membranes was disturbed.

580

Neuroimmunology of the Peripheral Nervous System

peripheral neuropathy in adult life. This finding indicates that, in as yet unidentified ways, periaxin is essential for the maintenance of a normal myelin sheath.285 Mutations in the periaxin gene have been identified as causing autosomal recessive demyelinating CMT disease (CMT4F) and Déjérine-Sottas disease.27,94,305 There is a single report of polyclonal IgM antibodies to periaxin being discovered in the serum of a small proportion of patients with IgG paraprotein-associated demyelinating neuropathy and with diabetes mellitus type 2 and neuropathy.174

Myelin-Associated Glycoprotein MAG exists in large and small forms (L- and S-MAG) as a result of alternative splicing of the C-terminal sequences, with unglycosylated weights of 67 and 72 kDa and, when glycosylated, weights of about 100 and 110 kDa, respectively. L-MAG and S-MAG predominate in myelinating and mature Schwann cells, respectively. MAG is situated in the periaxonal space of noncompacted myelin, at paranodes, mesaxons, and the Schmidt-Lanterman incisures (see Fig. 26–3).314 MAG is a transmembrane protein and a member of the immunoglobulin superfamily, with five extracellular immunoglobulin-like domains and an intracellular C-terminal tail (Fig. 26–4).209 The extracellular domains have N-linked glycosylation sites to which the HNK-1 epitope may be attached, and the large extracellular domains may interact homophilically with identical molecules on the opposing surface to maintain the periaxonal space. The intracellular domain of L-MAG contains a phosphorylation site that may act as a target for tyrosine kinases. It belongs to the siglec family, having a sialic acid binding site between the first and second extracellular domains, and plays a role in cell adhesion and Schwann cell axon signaling.357 The sialic acid binding site binds to ganglioside GD1a and other gangliosides with an 2–3 sialic acid linkage.330 In culture, MAG functions as a cell adhesion molecule, promotes or inhibits neurite outgrowth depending on conditions, and binds to collagens I and G.192,272

Myelin Glycolipids Peripheral nerve myelin is enriched with a wide range of glycolipids and some gangliosides that are important antigens for humoral responses. Gangliosides are principally enriched in neuronal membranes and are described in detail in the next section; the glycoconjugate structures that are predominantly localized to human PNS myelin are described here. Structural diagrams of the major glycolipids are shown in Table 26–1. Galactocerebroside (GalC), one of the simplest glycosphingolipids, comprising galactose linked to ceramide (monogalactosylceramide), was the first glycolipid identified as an important antigen in experimental autoimmune encephalitis and EAN, in which it is the principal target for complement-fixing antibodies. Galactosulfatide, also called sulfatide, is GalC sulfated on the third carbon of galactose. GalC and sulfatide are very typical and prominent lipids in myelin from both the CNS and PNS. The arrangement of GalC and sulfatide within the lipid bilayer is asymmetrical, with the glycolipids being localized to the outer leaflet, where they are available for antibody binding, as opposed to the phospholipids, which are principally localized to the inner leaflet. The importance of GalC in maintenance of myelin membranes has been demonstrated by targeted disruption

Periaxin Periaxin, a component of the dystroglycan-dystrophin– related protein-2 complex, has recently been described as a possible peripheral nerve antigen (see below). This complex acts to link the Schwann cell cytoskeleton to the extracellular matrix. Periaxin knockout mice myelinate normally, but then go on to develop a demyelinating

S S

S S

S S

S S

S S

L-MAG N

P C

C

P = Phosphorylation site

= Sialic acid binding site

S S

S S

S S

S S

S S

S-MAG N

= Glycosylation site

FIGURE 26–4 Diagram of the molecular structure of the long (L-MAG) and short (S-MAG) forms of myelinassociated glycoprotein. The glycosylation and sialic acid binding sites are shown. (After a diagram in Milner, R. J., Lai, C., Nave, K.-A., et al.: Organisation of myelin protein genes: myelin-associated glycoprotein. In Duncan, I., Skoff, R. P., and Colman, D. [eds.]: Myelination and Demyelination. New York, New York Academy of Sciences, p. 254, 1990.)

Gangliosides with an (2-8) linked disialosyl group

Terminal Gal(1-3)GalNAcconfigured gangliosides

Ganglioside and Glycolipid Antigenic Motifs

GQ1b

GD3

GD2

GD1b

GT1b

GM1

Structure

Table 26–1. Structural Diagrams of the Major Glycolipids*

asialo-GM1

Table continued on following page

Glial Structures Members of this group are present in compact myelin,49 especially small sensory fibers.230 GM1 is more abundant in VR than DR.45,229 Areas selectively bound by ligands include Schmidt-Lanterman incisures,211 paranodal end segments of the myelin sheath,49,82,83,161,165,211 nodal gap,273 and abaxonal Schwann cell cytoplasm.211 Neuronal Structures Ligand binding was observed with motor neurons,50,195 at the nodal axonal surface,82,243 at the neuromuscular junction,172,173,230,231,275,313 and to DRG neurons.165,195,230,231 Other Ligand binding was observed to muscle spindles231 and a series of Gal-GalNAc–bearing glycoproteins.9,49,191,132 Sensitization with GM1 ganglioside led to high anti-GM1 antibody titers and the development of acute-onset flaccid paralysis in rabbits.365 Endothelial cells express ganglioside antigens, including GM1 and GD1b. Circulating antiganglioside antibodies may damage cell-to-cell attachments, thus weakening the blood-nerve barrier and contributing to development of autoimmune demyelinating neuropathy.136 Glial Structures GD1b and GQ1b are enriched in the paranodal myelin.44,156,161,165,273 Anti-GD2 antibodies label the myelin sheath.366 GQ1b is enriched in the cranial nerves serving the extraocular muscle.45 GQ1b/GD3-reactive antibodies bind to cytoplasmic channels on the surface of myelinating Schwann cells, and to perisynaptic Schwann cells at the NMJ.87 Neuronal Structures Abundant in cultured DRG neurons.37 Antibodies bind DRG neurons,165,195,230,231,329 and are able to lyse them.234 Anti-GD1a antibodies induce ataxic neuropathy in rabbits.104,162 GT1b is confined to the axolemma.282 Thin-layer chromatography with immunostaining showed that GT1a is present in human oculomotor and lower cranial nerves.146 Complex gangliosides are present at the NMJ but are redundant for normal synaptic function.35 Antibodies bind the NMJ and can disrupt normal structure and function.34,87,232,249,347 Others Muscle spindles are stained strongly by an antibody to polysialylated gangliosides.347

Anatomic Localization and Immunopathologic Features

GalC

Fucosyl gangliosides

Chol-1 antigens

GM2

Terminal GalNAc-Gal-NeuNAcconfigured gangliosides

Fuc-GM1

GM1a

GM1b

Fuc-GD1b

GQ1ba

GalNAc-GM1b

GalNAc-GD1a

GD1a

Structure

Terminal NeuNAc(2-3)Galconfigured gangliosides

Ganglioside and Glycolipid Antigenic Motifs

Table 26–1. Structural Diagrams of the Major Glycolipids*—Continued

GalC is a major glycolipid component of myelin298 and may also be expressed by nonmyelinating Schwann cells.129 Intraneurally injected antibodies accumulated on the Schwann cell, particularly at the paranodal areas and Schmidt-Lanterman clefts.301 Sera from rabbits with EAN induced by sensitization with GalC binds to cultured Schwann cells.186

Anti–Fuc-GM1 antibodies bind to some small DRG neurons and their surrounding satellite cells.163,164

A series of at least six very minor gangliosides that are specific to cholinergic neurons.5,102,120,259 At least some of these are expressed on the cell body of motorneurons227,339 and at the NMJ.61 Proteins with cross-reactive carbohydrate sequences may also be present.62

Suggested as antigenic targets in autoimmune motor neuropathy.117,118,160,363 In biochemical studies, GM2 and GalNAc-GD1a were present in human peripheral nerves,117 but were undetected by immunolocalization studies.233

GD1a is a significant component of nerve in biochemical studies.45,229 High-titer IgG anti-GD1a antibodies selectively bind to motor, but not sensory, nerve nodes of Ranvier.58 GM1b has been detected in the brain but not extensively studied in the perpheral nervous system.77,151

Anatomic Localization and Immunopathologic Features

Gal

NeuNAc GalNAc

SO3

Glu

Fuc

SGLPG

GlcNAc

Glc UA

Ceramide

SGPG and SGLPG are found by biochemcial analysis in the myelin and axons of peripheral nerve.13,46,148 MAG is localized to regions of the myelin sheath in which the membranes are uncompacted, including the periaxonal membrane, the paranode, and Schmidt-Lanterman incisures.199,315 Non–MAG-reactive antibodies bind at the outer surface of myelin sheaths. Faint staining is also visible at the axolemmal-myelin interface, but compact myelin was not stained.355

Neuropathy-associated antibodies cross-react with MAG252 and the myelin proteins P0 and HNK-1.28,36 Sural biopsies from patient with these antibodies show deposits within the myelin sheath and signs of demyelination.317

Glial and Neuronal Structures In peripheral nerve, the most common sites in antibody binding studies include axons, resident macrophages and Schwann cytoplasm, especially the Schmidt-Lanterman incisures, nodes of Ranvier, and perineuronal sheath of satellite cells in DRG.190,219,254 Sural biopsies from patients with these antibodies show IgM and complement product C3d bound to the myelin sheaths of almost all fibers.76 A significant component of both sensory and motor nerves.45,229 Absent from several cranial nerves displaying CNS characteristics (e.g., optic nerve).45

CNS  central nervous system; DR  dorsal root; DRG  dorsal root ganglion; GalC  galactocerebroside; IgG  immunoglobulin G; LM1  sialosylneolactotetraosylceramide; MAG  myelin-associated glycoprotein; NMJ  neuromuscular junction; SGLPG  sulfated glucuronyllactosaminyl paragloboside; SGPG  sulfated glucuronyl paragloboside; VR  sulfated glucuronyl paragloboside; VR  ventral root. * The groupings described in this table are illustrative of cross-reactive epitopes reported in the literature, but are not intended as a definitive list. Gangliosides share common structural motifs outside of those presented here, and even within the presented groupings, it should be noted that an individual ganglioside can fall within two or more groups (e.g., GD1b). Equally, the fact that two ganglioside species appear in the same group does not necessarily imply that an antibody to one will cross-react with the other.231 Glycolipids that appear dissimilar when presented as shown here may adopt tertiary structures that display common motifs, and through the same process similar-looking structures may be antigenically different. † Gal  galactose; NeuNAc  neuraminic acid; GalNAc  N-acetylgalactosamine; Glu  glucose; Fuc  fucose; GluNAc  N-acetylglucosamine; Glc  glucosamine.

Structural components†

SGPG

SGPG/SGLPG

SO3

LM1

SO3

LM1

Sulfatide

584

Neuroimmunology of the Peripheral Nervous System

of UDP-galactose ceramide galactosyltransferase (CGT), the key enzyme in galactolipid biosynthesis. Such mice have a phenotype marked by slow conduction and delayed myelin breakdown that has been characterized in detail.47 Sulfatide has also been identified as an important antigen in autoimmune neuropathy, and it has been possible to delete the galactosylceramide sulfotransferase gene that specifically inhibits the synthesis of sulfatide, while maintaining GalC. These modifications may have a much less severe phenotype than the CGT knockout,109 but they exhibit ultrastructural abnormalities in the paranodal regions, and it appears that sulfatide may play a critical role in the localization and maintenance of ion channels in the underlying axonal membrane.122 Peripheral nerve has an abundance of neolacto-series gangliosides that are localized mainly in myelin.228 The term LM1 is used for the sialosylneolactotetraosylceramide, which is also known as sialosyl paragloboside, and Hex-LM1 is used for sialosyllactosaminyl paragloboside. LM1 and Hex-LM1 are major monosialosyl glycolipids in human peripheral nerve myelin and have been identified as autoantigens in both acute and chronic autoimmune neuropathies. The function of LM1 and Hex-LM1 has not been elucidated, although they are very widely distributed in human peripheral nerve myelin, as well as other tissues and cells such as the red blood cell. Another important class of glycolipid antigens is sulfated glucuronyl paragloboside (SGPG) and its higher lactosaminyl homologue, sulfated glucuronyllactosaminyl paragloboside (SGLPG).42,171,253 Both were discovered directly as a result of studying IgM paraproteins reactive to MAG, with which they share immunologically crossreactive structures, from patients with chronic polyneuropathy.13 SGPG and SGLPG also have structures similar to that of LM1, except for a 3-sulfated glucuronic acid instead of sialic acid on the terminal saccharide chain. The subcellular distribution of these sulfated glucuronyl glycolipids has been studied in myelin- and axon-enriched fractions of motor and sensory nerves from human subjects. SGPG is slightly more abundant in sensory nerve axolemmal fractions than motor nerve fractions, which may account for the relative predominance of sensory features in patients with anti-SGPG antibodies. However, it should also be recognized that these antibodies crossreact with MAG and other myelin proteins that may not have such a distribution.

Neuronal Antigens A number of molecules that are principally localized to the neuronal and axonal compartments of peripheral nerve have also been identified as relevant to the pathogenesis of human autoimmune neuropathy. These antigens include the paraneoplastic antigens Hu, CV2, and amphiphysin, which are each associated with the development of para-

neoplastic sensory neuronopathies. In addition, axonal neurofilaments, several ion channels, and numerous gangliosides are all reported as important antigens in diverse neuropathy syndromes, principally manifested by neuronal or axonal dysfunction. Neurologic paraneoplastic syndromes provide some of the clearest examples of autoimmune disease in the nervous system, affecting diverse central and peripheral structures, and are recognized by the presence of high-affinity antibodies against specific and, in many cases well-characterized, neural antigens. The origin of the immune response is presumed to be against the tumor cells in which the neural antigens are inappropriately expressed. In syndromes in which the antigens are expressed on the plasma membrane, the immunopathology is primarily dependent on antibody and antibody-mediated processes such as complement fixation. However, in those paraneoplastic syndromes involving an intracellular antigen, the mechanism of pathogenesis is not clear, but there is evidence for the involvement of cytotoxic T cells.3,306,307 In most of these syndromes the CNS is the predominant site of disease, and these are not discussed here. Paraneoplastic Neuronal Antigens: Hu, CV2, and Amphiphysin The paraneoplastic subacute sensory neuronopathy (PSSN) associated with anti-Hu antibodies,56 and less commonly anti-amphiphysin antibodies, and the more recently recognized subacute polyneuropathy associated with anti-CV2 antibodies7 are the three paraneoplastic syndromes in which peripheral sensory neurons are most clearly involved. The autoimmune nature of sensory neuropathy associated with small cell lung carcinoma (SCLC) was originally detected by complement fixation,53,342 and the antibodies were shown to react preferentially with neuronal nuclei and a small group of proteins of 32 to 42 kDa, including the Hu antigen.4,90 The Hu gene is located on the human chromosome 1 at 1p34,215,279 and the antigen has several forms generated by alternative splicing of four exons, the most prominent product being HuD (Fig. 26–5).303 Additional closely related gene products HuC and Hel-N1a have also been recognized.187,268 The proteins characteristically contain three RNA recognition regions, the first two in tandem separated from the third domain by a basic segment. Sequence homology with the Drosophila proteins ELAV and Sex-Lethal suggests that Hu proteins bind specifically to AU-rich elements in the untranslated 3 regions of some classes of mRNA,154,187 and this has been confirmed by direct analysis.187,334 Those mRNAs with AU-rich regions are frequently short-lived species, but the binding of proteins such as HuD increases their stability and lifespan. The human anti-HuD IgG binds to those regions of the molecule containing the first two RNA binding domains. At least two different epitopes exist, one of which is in the

Peripheral Nerve Antigens

FIGURE 26–5 Diagrammatic view of HuD. The molecule contains three globular regions, a, b, and c, with homologies to RNA-binding proteins. These contain -helical and random coil structures (gray) and -structures (black) that contain the major RNA-binding residues. Serum from patients with paraneoplastic disease contains antibodies directed against discrete isotopes in the globular regions a and b that do not interfere with their RNA-binding property. (Data from Wang X., and Tanaka Hall, T. M.: Structural basis for recognition of AU-rich element RNA by the HuD protein. Nat. Struct. Biol. 8:141, 2001.)

vicinity of each of the RNA binding domains.197 However, the binding of the antibodies does not appear to interfere with the binding to RNA.154 Although the Hu proteins are highly characteristic and specific to all central, peripheral, and enteric neurons, they are also expressed by most SCLC cell lines,197,279 as well as cells of other tumors associated with PSSN. The presence of the antigen in the associated tumor satisfies the hypothesis that the autoimmunity arises during an immune response against the tumor. One explanation for the presence of an immune response against the protein could be that the tumor expresses a mutated form of HuD. However, this was not found to be the case in a study of 26 lung cancer cell lines, three from patients with anti-Hu antibodies, of whom one had a paraneoplastic disorder.279 Low levels of anti-Hu antibodies have been found in cases of SCLC in the absence of any neurologic disease.204 Histopathologic examination of SCLC tumors has indicated that paraneoplastic disease is associated with strong expression of MHC antigens by the tumor cells.55

585

However, these observations were clearly made after the initiation of the immune response, and the expression of MHC antigens could reflect the strong inflammatory nature of the response in PSSN. Although high-titer IgG antibodies against Hu and related antigens occur in the circulation and cerebrospinal fluid (CSF) in PSSN, it has not been possible to demonstrate any cytopathic effect of this antibody on neurons or SCLC cells in culture, even though the antibody can be demonstrated in the nuclei of the exposed cells.113 Similarly, it has not been possible to demonstrate the development of any neurologic syndrome in animals following passive transfer of patient IgG,63 or by active immunization against the Hu antigens.289 High titers of anti-Hu antibodies have been found in association with inflammation of the gut and loss of enteric neurons in the absence of a tumor and without any other peripheral or central signs.291 T cells are prominent in the cellular infiltrates in the nervous system, and these are predominantly CD8 T cells.91 Some circulating CD4 T cells have been shown to respond to recombinant HuD,24 and more recently CD8 cytotoxic T cells have been found to kill syngeneic fibroblasts containing HuD and expressing MHC class I molecules.306 In tissue culture, neurons can be induced to express MHC class I molecules on their surface in the presence of interferon-, particularly when the neurons are electrically silent.221 Greater MHC class I expression has been seen as a result of neural activity.51 The expression of MHC class I molecules could well provide the means of targeting the specific cytotoxic CD8 T cells and explain the neuronal killing in this disease. However, neuronal expression of human leukocyte antigen has not been demonstrated in pathologic material,91 whereas in the DRG, satellite cells may normally express both MHC class I and II molecules.89 Even if neurons can be induced to express MHC class I antigens, the problem remains of how such an induction is initiated if it requires the presence of inflammation. The CV2 antigen was first identified as a paraneoplastic antigen when serum and CSF of some patients with neuropathy associated with cerebellar ataxia and cancer were found to react with a subpopulation of oligodendrocytes in the rat brainstem, cerebellum, and spinal cord. The IgG antibodies bound to a 66-kDa protein found in the newborn rat brain.111 Cloning and Northern blot analysis demonstrated a 3.8-kb mRNA that was brain specific and increased from rat E17 to reach a maximum at P7, after which it declined to negligible levels except in the spinal cord and cerebellum.251 The sequence indicated that the antigen was an Unc-like phosphoprotein (Ulip), and the CV2 antigen was assigned to the Ulip-3 group, which in the human and rat corresponds to collapsin response mediator protein-1 (CRMP-1). Collapsins are receptors for the extracellular semaphorins that induce growth cone collapse and are therefore also important in axonal growth and plasticity.

586

Neuroimmunology of the Peripheral Nervous System

Sera from patients bind variously to CRMP-1, -2, and -3 prepared as recombinant protein in HeLa cells, but all bind to CRMP-5.110 CRMP-5 shows only 50% sequence identity to the other four CRMPs, which share 68% to 74% identity with each other. There is therefore a degree of uncertainty as to the identity of the “antigen” complementary to the anti-CV2 antibodies, but on the basis of a reported screen, CRMP-5 appears to be the likely candidate.110 However, the possibility must remain that more than one antibody specificity exists. The identity of the anti-CV2 antigen as CRMP-5 is compatible with the fact that it is also present in oligodendrocytes, and that CV2positive sera do bind to retinal neurons, neurons in the dentate gyrus, and the rostral band of olfactory neurons as well as peripheral axons.8,110 The putative antigen for CV2 antibodies is known to be concerned with axonal growth, and the antibodies do react with peripheral axons,7 but again CRMP proteins are intracellular and the question of access remains unresolved. There is no information about whether such antibodies are pathogenic, or about whether patients have T-cell responses to CRMP. CRMP-5 has been demonstrated to occur in SCLC cells, satisfying the requirement for tumor expression of the neural antigen.359 It is not uncommon for antibodies against more than one neural antigen to occur in paraneoplastic syndromes. Thus, as indicated above, anti–CRMP-3 antibodies have been found in association with anti-Hu but also antibodies against voltage-gated Ca2 channels176 and, less commonly, amphiphysin.67 Amphiphysin-1 was originally noted as a neural target in patients with stiff-man syndrome and breast cancer.57 The 128-kDa amphiphysin-1 and -2 play a role in clathrin-mediated endocytosis, forming a heterodimer interacting with the GTPase dynamin to act in vesicle budding from the membrane.341 Amphiphysin is localized predominantly at synaptic terminals and is passed anterogradely along axons by fast axonal transport, but only about a third returns by retrograde flow, suggesting local breakdown.179 Antibodies against amphiphysin are mostly directed against the C-terminal half of the molecule and have been found in PSSN,67 opsoclonus-myoclonus,21 and a small proportion of patients with SCLC, irrespective of the presence of paraneoplastic disease.267 They have also been found in a patient with a sensorimotor axonopathy.245 The occurrence of the antibodies in the absence of neurologic symptoms raises questions as to their pathogenic significance.267 There is no record of a model neuropathy inducible by hyperimmunization of a laboratory animal.

Neurofilament Antigens Neurofilaments (NFs) belong to the class of intermediate filament proteins and are prominent structures of neurons and axons. The filaments contain three proteins of 200, 160, and 68 kDa (NF-H, NF-M, and NF-L, respectively); their proportions and phosphorylation vary in different

parts of the neuron. In axons NFs are highly phosphorylated, and NF-L forms the central core of the structure. Antibodies against NF have been found in a variety of neurologic conditions and in healthy control subjects.19,288,297,310 It is possible that in healthy controls such antibodies represent part of the “natural antibody” repertoire, but it is also likely that in neurodegeneration these are upregulated and diversified as a response to the release of antigen.59 In relation to natural antibodies, 4 of 75 patients with paraprotein-associated neuropathy had an IgM antibody against NF-H,224 and a monoclonal immunoglobulin A (IgA) from a patient with amyotrophic lateral sclerosis (ALS) was reactive with the 200-kDa NF.263 There is little indication that such antibodies are pathogenic, although the monoclonal IgA from the patient with ALS was found to cross-react with an undefined neuronal surface protein.263 Once again, it is difficult to see how antibodies against intracellular antigens could be strongly immunopathogenic. However, both IgG and IgM can be endocytosed by neurons with peripheral processes and retrogradely transported to the perikaryon.72–74,194 Anti-NF antibodies have been detected in patients with paraneoplastic optic neuropathy, and NF was found in the associated tumors.150 Other studies have found that NF expression in SCLC tumors is not common, unlike that of HuD.325 However, neurofilaments of 160 and 68 kDa have been demonstrated in the cells of thymic cortical tumors in association with myasthenia gravis.201,202,277 It has been suggested that the antiaxonal antibodies in these patients represent antibodies against the NF-M expressed in these tumors.202,277 Epitope mapping showed that the antibodies reacted with the sequence KSPVEEK, which was repeated five times in the NF molecule, and also with a cytoplasmic epitope (KSAIGIK) of the a subunit of acetylcholine receptor.277 The immunity against the NF-M expressed by the tumor might account for the anti–acetylcholine receptor antibodies in myasthenia patients with cortical thymic tumors. However, it has not been established by experiment that such an immune response can induce myasthenia gravis.

Gangliosides Structure and Nomenclature The term ganglioside refers to the large family of glycosphingolipids that contain sialic acid linked to the oligosaccharide core. In humans, most ganglioside sialic acid is in the N-acetyl form, as opposed to the N-glycosyl configuration that is common to many other species. The nomenclature of Svennerholm is usually used for gangliosides.43,302 The designations are based on the four major brain gangliosides with a ganglio-series tetraose chain of neutral sugars (i.e., asialo-GM1) that are sialylated in different positions. However, there are around 100 structurally distinct gangliosides, and it has become difficult to create a single, easily workable nomenclature.

587

Peripheral Nerve Antigens

In the initial classification, four gangliosides, GM1, GD1a, GD1b, and GT1b, were designated to belong to the G1 series, wherein “G” stands for ganglio-series ganglioside. The four major gangliosides differ with regard to the number and position of their sialic acids, with M, D, and T indicating mono-, di-, and trisialosyl groups, respectively. Thus there are two disialosylgangliosides, GD1a and GD1b. Although “b” is normally used to designate gangliosides with a disialosyl group attached to the internal galactose (so-called b-series gangliosides), the term GM1b is used for the monosialosyl gangliotetraosyl ceramide, in which the sialosyl group is attached to the terminal galactose, in contrast to GM1a (normally more simply referred to as GM1), in which the sialic acid is on the internal galactose. When three sialic acids link to the internal galactose of the ganglioside, they are designated to belong to the “c” series. Now that the biosynthetic pathway of gangliosides of the ganglio-series has been in large part elucidated, it is evident that this early description of the “a,” “b,” and “c” series predicted the crucial role of the sialyltransferases in ganglioside biosynthesis, as shown in Figure 26–6. Gangliosides lacking the terminal galactose, preterminal galactosyl-N-acetylgalactosamine, or internal galactose are assigned the number 2, 3, or 4, respectively.

Ganglioside Metabolism and Function Gangliosides are present throughout the body but are very highly concentrated in the nervous system. They are particularly enriched in neuronal membranes, but are also minor

constituents of myelin.175,311,358 Gangliosides are amphipathic molecules comprising a hydrophobic ceramide moiety that is embedded in the lipid membrane and a hydrophilic oligosaccharide moiety, which is exposed to the cytosolic compartment or to the extracellular space. Thus gangliosides are highly associated with membrane structures both in the plasma membrane, where their density is high, and in a wide variety of intracellular membrane compartments, including nuclear and mitochondrial membranes, endosomes, lysosomes, and the Golgi apparatus. Gangliosides cycle through these various intracellular compartments to and from the plasma membrane as small vesicles and through endosomal sorting. Gangliosides are synthesized in the Golgi apparatus by the stepwise addition of saccharides in reactions catalyzed by glycosyltransferases and sialyltransferases, and are degraded in a reverse fashion by the stepwise removal of saccharides in lysosomes. They can accumulate abnormally in lysosomal storage diseases as a result of mutations in the enzymes or transport proteins involved in these degradative pathways. Targeted disruption of genes involved in ganglioside biosynthesis has demonstrated the importance of gangliosides in a wide range of neural processes, including the maintenance of peripheral nerve integrity and function. Thus null-mutant mice lacking the gene coding for the glycosyltransferase 1,4-GalNac-transferase (1,4-GalNac-T; GM2 /GD2 synthase) lack complex gangliosides, and consequently neuronal membranes only bear the simple gangliosides GD3 and GM3, whose expression is upregulated.304 Such animals develop age-related axonal degeneration in

CER

Glucose

Galactose

GaINAc

NeuAc

CER

CER

CER LacCer

CER

CER GA2

FIGURE 26–6 Schematic representation of ganglioside biosynthesis. Ganglioside nomenclature is according to Svennerholm.302 Gangliosides are synthesized through addition of monosaccharides in a stepwise fashion by glycosyltransferases and sialyltransferases (arrows). CER  ceramide; GalNAc  N-acetylgalactosamine; LacCer  lactosyl ceramide; NeuAc  neuraminic acid or sialic acid. (Adapted from Bullens, R. W. M., O’Hanlon, G. M., Wagner, E. R., et al.: Complex gangliosides at the neuromuscular junction are essential receptors for autoantibodies and botulinum neurotoxin but redundant for normal synaptic function. J. Neurosci. 22:6876, 2002.)

CER GM3

CER

CER GM2

CER GA1

CER

CER

GD1a

CER

“a-series”

GT1c CER

GT1b CER

GT1a

GT2

GD1b CER

CER GD1c

CER

CER

CER

GT3

GD2

GM1

GM1b

CER GD3

GQ1c CER

GQ1b “b-series”

GP1c “c-series”

588

Neuroimmunology of the Peripheral Nervous System

the CNS and PNS, and also develop demyelination in peripheral nerves with pathologic features resembling those seen in MAG knockout mice, MAG being a known ligand for complex brain gangliosides.330 In double-knockout mice that lack both GM2/GD2 and GD3 synthase, expressing only GM3, peripheral nerve degeneration has also been observed, combined with mutilating skin lesions and nerve fiber proliferation in the skin, suggesting that reduced sensory function is present.119 These types of studies are continually revealing important facets of ganglioside biology in peripheral nerve. With respect to their behavior as antigen targets, the predominant relevant ganglioside localization is the plasma membrane, where the oligosaccharide portion of gangliosides is extracellularly sited. Gangliosides are not uniformly distributed in plasma membranes but are concentrated in association with cholesterol and phospholipids in clusters referred to as functional or lipid rafts.290,298 Such rafts are important sites for signal transduction, being highly enriched in glycosylphosphatidylinositol-anchored proteins on the extracellular face and Src kinases on the cytoplasmic face. Rafts are also associated with small plasma membrane invaginations called caveoli, which are sites of endocytosis. The high concentration and close proximity of gangliosides in lipid rafts may enhance their ability to bind antibody with high avidity or affinity. The pathogenic effects of antiganglioside antibodies are likely to depend not only on how avidly they bind gangliosides but the extent to which the target gangliosides are intimately involved in modulating neuronal function. Additionally, activated complement components may in turn affect the normal functioning of ganglioside raft–associated proteins. The relative contribution of these factors may vary from site to site, and among antiganglioside antibodies of differing reactivity. Relatively little is currently known about the distribution of gangliosides in different membrane and raft components throughout the nervous system, and the relevance that this distribution may have to their role as antigens in autoimmune neuropathy remains unclear. However, when considering the pathogenic relationship between the presence of an antibody and neuropathy, it is clearly important to have detailed knowledge of the glycosphingolipid composition and distribution within the PNS both in humans and in experimental animals, and furthermore in species from which gangliosides are purified for experimental and diagnostic use. These issues are not straightforward because the regulation of ganglioside expression and methods for the analysis and purification of gangliosides are complex. Gangliosides are developmentally regulated and spatially segregated, varying among different peripheral nerve fiber types and among different species. A complete map of the ganglioside composition of human nerves, and a comparison among species used for experimental modeling, would be a valuable resource but might still have limitations.

The biochemical and immunohistologic approaches to establishing the distribution of gangliosides each have their merits and limitations. Biochemical analysis has been useful to identify significant differences in the ganglioside composition of different nerves and can reveal subtle differences in the overall lipid composition. However, this approach is limited by the pleomorphic composition of tissue and the lack of information about microanatomic distribution. In such circumstances the second approach, that of immunohistology or other in situ ligand-binding studies (e.g., using ganglioside-binding bacterial toxins such as cholera toxin) can reveal fine structural detail about ganglioside distribution at the cellular and subcellular level (Fig. 26–7). For these studies, high-quality reagents such as affinity-purified antisera or monoclonal antibodies are essential. Many antiganglioside antibodies are not monospecific but may crossreact with structurally similar gangliosides and other glycoconjugate antigens, making extrapolation of results to ganglioside localization difficult. Biochemical and immunohistologic approaches characterized by homogenization and tissue sectioning, respectively, may expose gangliosides that normally occupy cryptic sites (e.g., within compact myelin). Gangliosides may also be complexed with intrinsic ligands such as siglecs, as discussed above. Thus diverse factors may misrepresent the ganglioside array that would be visible to circulating antibodies in physiologic environments. Furthermore, gangliosides can be heterogeneously distributed within a membrane, as in functional rafts as described above. The antigen density and the surrounding lipid environment can also markedly influence the ability of antiganglioside antibodies to bind; thus failure to detect a ganglioside by immunohistology does not necessarily indicate its biochemical absence.188 A high local concentration of ganglioside may allow for good immunohistologic detection, whereas a ganglioside that is evenly distributed throughout a membrane may have the same total tissue concentration in biochemical evaluation, yet not be detectable by immunohistology. Interpretation of both biochemical and immunohistologic studies thus requires caution. The accessibility of PNS gangliosides to circulating antibodies, being protected in their neural environment by the blood-nerve barrier and other factors, is important.188 One explanation for the lack of CNS involvement in antiglycolipid antibody–associated neuropathy, despite the wide distribution of gangliosides in the CNS, would thus be the protection from autoimmune attack afforded by the blood-brain barrier. Blood-nerve barrier injury could be mediated by antiganglioside antibodies reacting with glycolipid antigens expressed by intraneural microvascular endothelial cells.137 Ganglioside Localization to Specific Nerve Sites and Fiber Types The simplest explanation for particular antiganglioside antibody–associated neuropathies being confined to a motor or

Peripheral Nerve Antigens

589

FIGURE 26–7 Immunofluorescent labeling of mouse tissue using antiganglioside antibodies. The monoclonal IgM antibodies Ha1 (human347) and CGM3 (mouse87) are both reactive with gangliosides possessing a disialosyl moiety, including GQ1b, GD1b, and GD3. Disialosyl gangliosides are detected throughout the cell bodies of dorsal root ganglion neurons (A), as delineated by antineurofilament (NF) antibodies (B). Disialosyl gangliosides are largely confined to the nerve terminal and associated nonmyelinating Schwann cells (C) at the neuromuscular junction, as defined by NF and -bungarotoxin (BTx) labeling (D). At the nodes of Ranvier, they occupy a focal nodal position suggestive of an axonal distribution (E) compared to the more diffuse paranodal distribution of ganglioside GM1, as detected by cholera toxin B subunit (CT) binding (F). Bars: 20 m.

sensory clinical phenotype is that the two systems contain different gangliosides; this issue has been addressed in many experimental studies, including immunohistologic analyses.86 A comparison of total ganglioside composition of human spinal roots showed that GM1 (associated with antibodies in motor neuropathy) is relatively enriched in the ventral roots compared with the dorsal roots.229 Similarly, the cranial motor nerves supplying the extraocular muscles contain higher concentrations of GQ1b, the ganglioside antigen associated with ophthalmoplegia.45 However, from these studies it is also apparent that key gangliosides are also present at sites unaffected by the disease process. Thus in many cases the absolute tissue distribution of gangliosides is an insufficient explanation for the regional localization of the clinical pathology. The DRG is a particularly interesting site to consider in these respects and illustrates some of the anomalies described above. Functional subpopulations of DRG neurons can be distinguished by the expression of lactoseries and globo-series carbohydrates that correlate with

their peptide and enzymatic phenotype.64,65 It is very likely that ganglio-series antigens are also selectively distributed and that such differences may underlie the specific nature of sensory deficits. In the rodent DRG, the GM1 ligand cholera toxin B subunit and anti-GM1 antibodies selectively identify a subset of DRG neurons. In contrast, cholera toxin and anti-GM1 antibodies bind the majority of neurons in the human DRG.230,231 Thus GM1 is clearly present in the human DRG in abundance, yet the clinical phenotype of anti-GM1–associated antibody neuropathy is strikingly devoid of sensory features. However, anti-GD1b antibodies are highly associated with sensory neuropathy and bind to the vast majority of DRG neurons.161,165,195 Disialosylated gangliosides, including GD1b, GT1b, GQ1b, GD3, and GD2, are prominent gangliosides in cultured rodent DRG neurons, and antidisialosyl antibodies can lyse these.234 These observations also suggest that care should be exercised when attempting to establish an animal model of a human disease.

590

Neuroimmunology of the Peripheral Nervous System

The node of Ranvier is another key site of injury in autoimmune neuropathy. Immune attack directed at antigenic determinants located at the paranodal Schwann cell surface may lead to paranodal demyelination, whereas antigens targeted on the exposed axolemma may result in axonal degeneration, both of which would result in conduction failure. Ligand-binding studies have suggested that GM1, GD1b, and polysialosylated gangliosides are enriched in the paranodal myelin loops of peripheral nerve (see Figs. 26–1 and 26–7).165,273 In the ocular motor nerves affected in MFS, GQ1b is particularly enriched at nodes of Ranvier.44 With respect to the neuronal components of the node of Ranvier, toxin- and antibody-binding studies have identified gangliosides on paranodal and internodal axolemma49,82,83,282 and the adaxonal membrane.211 Similarly, an antibody reactive with disialosylated gangliosides has been shown to bind to internodal axolemma and/or adaxonal Schwann cell cytoplasm.347 Antibodies to GD1a are associated with pure motor axonal neuropathy, and preferentially stain ventral root axons in comparison with dorsal root axons, indicating a good correlation between ganglioside localization and phenotype in this example.86,283 With some clinical justification and for a variety of hypothetical reasons, the presynaptic neuromuscular junction (NMJ) has recently been considered a potential target vulnerable to autoimmune attack in GBS. First, it lacks a blood-nerve barrier, thereby readily allowing access to circulating autoantibodies. Second, it is the site for other paralytic antibody-mediated diseases, including myasthenia gravis and Lambert-Eaton myasthenic syndrome. Third, it is rich in gangliosides, including GQ1b, GM1, and GD1a.34,87,173,230,231,249,275,313,347 Fourth, it is the binding site for a wide range of bacterial toxins that also use gangliosides as ectoacceptors.344 In particular, cholera and tetanus toxins are readily taken up into nerve terminals, loaded into synaptic vesicles, and ultimately transported back to the motor neuron cell body.103,333 As a result of this property, enzymic conjugates of cholera toxin, or its binding B subunit, are frequently used as retrograde neuronal markers. As expected, histologic analysis has demonstrated cholera toxin and anti-GM1/GD1b antibody binding to the NMJ.172,173,230,275 Antibodies reactive to polysialosylated gangliosides also bind to the NMJ (see Fig. 26–7). Some of the -series gangliosides specific to cholinergic neurons (Chol-1 antigens) are also expressed at the mature NMJ, making them potential targets for autoantibodies.63

Voltage-Gated Calcium and Potassium Channels Voltage-gated ion channels are glycosylated multimeric proteins forming a potential pore that opens and closes according to the membrane potential, leading to changes

in ion flux across the membrane. Voltage-gated potassium (K) channels have four  subunits that form the pore and a variety of  accessory subunits. There are many different K channels, and mutations of them cause episodic ataxia or epileptic syndromes.153 Antibodies against one of them, the dendrotoxin-sensitive fast potassium channel, cause Isaacs’ syndrome or acquired neuromyotonia (see below). In voltage-gated calcium (Ca2) channels, the pore is formed by a single  subunit with four domains each having six transmembrane segments and thus resembling the K channel. Different mutations of one Ca2 channel cause familial hemiplegic migraine, episodic ataxia, and spinocerebellar atrophy. Autoantibodies to presynaptic calcium channels present at motor nerve terminals are found in Lambert-Eaton myasthenic syndrome.

MOLECULAR MIMICRY AND MICROBIAL ANTIGENS Infection has long been considered to initiate autoimmune disease and particularly autoimmune neuropathy. Infective agents provide a large number of antigenic peptides and other structures that may stimulate naïve and memory T and B cells. The high levels of cross-reactivity shown by T cells means that some cells are likely to be stimulated that will also react with self-peptides.203 Infection can lead to an intense inflammatory activation, which increases the possibility of reactivating anergic autoreactive lymphocytes. Nonetheless, the situations in which infection stimulates autoimmunity that have been demonstrated are few and specific. In autoimmune neuropathies, GBS is the clearest example, but infection might also be a factor in chronic inflammatory neuropathies. An association between acute cytomegalovirus (CMV) infection and antiMAG paraproteinemic polyneuropathy was suggested362 but not confirmed.121,193 The symptoms of GBS usually become apparent some weeks after an infective episode, and thus the identification of the pathogen by its isolation is difficult. As a result, most of the information relating to pathogen identity is derived from testing for serum antibodies, but there are some problems with this approach. The serologic tests used are not always absolutely specific, and different centers use different assays, so that detailed comparisons are difficult. Even so, certain specific organisms are particularly strongly associated with the onset of the neuropathy. Of these, Campylobacter jejuni has been found to be the most prevalent agent, followed by CMV and Mycoplasma pneumoniae.95,127,145,213,322,324,352 An association with Haemophilus influenzae is currently controversial but seems likely. It is possible that the level of association of individual organisms with GBS differs according to geographic region and their general incidence or genetic subtype.

591

Peripheral Nerve Antigens

Campylobacter Infection As already indicated, antiganglioside antibodies are autoantibodies frequently found to be associated with GBS, and are strongly associated with prior C. jejuni enteritis.10,257,332,367 Several studies have also demonstrated that the AMAN/motor form of GBS is more frequent following C. jejuni infection.105,205,256,257,328,352 Campylobacter jejuni is a spiral, flagellated bacillus with a large genetic and phenotypic diversity. It is widespread in the environment, including fresh water. In humans, particularly in the northern hemisphere, it is now probably the most common agent of sporadic enteritis and appears to be a normal component of avian gut flora. The early finding that the lipopolysaccharide (LPS) fraction from C. jejuni isolates from GBS patients bears carbohydrate sequences identical to gangliosides suggests that this is the antigenic stimulus of the antiganglioside antibodies frequently found in patients.15–17,93,126,271,364 Immunization of experimental animals with LPS derived from isolates from GBS and MFS leads to the production of antiganglioside antibodies.6,87 However, neither the type of GBS nor the occurrence of antiganglioside antibodies is absolutely associated with C. jejuni infection. Recent genotyping studies on C. jejuni, using multienzyme electrophoresis, ribotype and flagellin gene (flaA) polymorphism, amplified fragment length polymorphism, and microarray, have all confirmed the high degree of genetic diversity.66,70,217 These studies have so far failed to find any clonal clustering of GBS isolates. In Japan and China serotype O:19,155,222 and in South Africa serotype O:41,170 are common isolates from GBS/MFS but not from uncomplicated enteritis patients. It is likely that some O:19 isolates are clonal and share an ancestor with Penner O:41 strains.217 However, the full extent of the ancestral relationships between GBS-associated strains of C. jejuni remains unknown. Endogenous synthesis of N-acetylneuraminic acid (sialic acid) from N-acetylmannosamine in C. jejuni is encoded by three neuB genes, only one of which, neuB1, affects LPS synthesis.185 Both (2-3)- and (2-8)-sialyltransferases are present,84,321 enabling the synthesis of mono-, disialo-, and trisialo-oligosaccharides. The genes encoding these transferases are widely distributed in isolates from both GBS/MFS and enteritis patients. The gene (wlaN) encoding the 1,3-galactosyltransferase is essential for the terminal galactose to produce GM1- and higher di- and trisialo ganglioside-like structures, and contains a homopolymeric sequence that introduces phase-variable behavior in the expression of this enzyme.184 This means that the expression of the GM1 epitopes on LPS will vary according to a single base insertion/deletion, and most isolates are likely to contain organisms with and without expression of this enzyme. Which members of the population become dominant

depends on growth conditions. Without wlaN expression, the LPS expresses a GM2 ganglioside epitope, which has been demonstrated to be immunogenic.260 Thus, although the production of antiganglioside antibodies is dependent on the presence and expression of these genes, the genes are sufficiently common in isolates of Campylobacter that they do not typify neurogenic strains.

Other Infections Antiganglioside antibodies also occur in GBS cases associated with other infections. Haemophilus influenzae has been associated with anti-GQ1b/GT1a antibodies in MFS147 and anti-GM1 antibodies in GBS.213 The LPSs of H. influenzae and H. ducreyi are known to be sialylated,98,196,198 and the organisms have multiple sialyltransferase genes.112,131 However, unlike Campylobacter, Haemophilus does not synthesize sialic acid but uses host sialic acid to sialylate its LPS.327 Antiganglioside antibodies are also found in some patients with serologic evidence of recent CMV and M. pneumoniae infection.95,97,101,128,143,158,362 These two organisms do not have the genes for glycolipid synthesis but incorporate molecules from the host cell into their own membranes. Interestingly, CMV is particularly associated with antibodies reactive to GM2 ganglioside,140 which is a very common ganglioside of tissues outside the nervous system. The occurrence of ganglioside structures is very prominent in the most common organisms associated with GBS/MFS. Also, the presence of surface sialic acid is considered to enable the adsorption of factor H and the inhibition of complement activation.

Molecular Mimicry The production of antiganglioside antibodies in response to the expression of ganglioside oligosaccharide sequences by a pathogen is most frequently referred to as “molecular mimicry.” This is a convenient shorthand phrase in autoimmunity to describe the similarity in structural sequences between a pathogen and a relevant self-antigen that might generate self-reactive T or B cells. However, in biologic systems, mimicry traditionally implies that the mimic obtains a selective advantage by expressing some phenotypic feature of another unrelated species. In fact, molecular mimicry is a prominent and familiar feature of pathogens and is related to their success. In most of the studied instances, the molecular mimics are “functional” mimics, lacking both sequential and structural homology to the host proteins, the function of which is subverted.296 Often the pathogen mimicry appears to have arisen by a process of convergent evolution, but some mimicry involves pathogen expression of homologues of host

592

Neuroimmunology of the Peripheral Nervous System

proteins, the genes for which are likely to have been acquired by horizontal transfer. In many cases the mimicry involves proteins that disrupt the immune response and can target both the innate and adaptive immune systems. The presence of sialic acid at the surface of microorganisms has been suggested to suppress serologic responses by an unknown mechanism. The discovery of the family of sialic acid–binding proteins on immune cells may offer a partial explanation. Thus the NK cell inhibitory receptor siglec-7 has been shown to bind to disialogangliosides.124,354 This receptor is also found on monocytes, dendritic cells, and a CD8 T-cell subset.52 Thus the expression of disialogangliosides by C. jejuni and other organisms could suppress the innate immune response and therefore possibly the adaptive response, with obvious selective advantage to the organism. How monosialoganglioside might have a similar effect is still not known. It is not known for how long C. jejuni has been a human pathogen, but it is quite possible that any displayed mimicry is an adaptation to the avian immune system. Nonetheless, it appears that the mimicry is successful in humans because, in most cases of C. jejuni enteritis, high titers of antiganglioside antibodies are not found. Similarly, in the majority of cases, the repeated injection of gangliosides into human subjects does not lead to a marked humoral response.88,177,255 This suggests that the development of high-titer antibodies in GBS/MFS may be due to either the host failing to respond appropriately, or to some strains of C. jejuni carrying an additional factor that defeats the subversion of the immune response. There is little direct evidence that C. jejuni provokes a poor immune response, or that it is particularly resistant to the innate immune system. Avian peritoneal macrophages are able to phagocytose and effectively kill the organism,216 although chickens produce weak B-cell responses against thymus independent type-1 antigens.130 In humans as few as 500 organisms can be infective,331 yet most laboratory animals are resistant and no good laboratory model of the human enteritis exists. Mice clear the organism quickly, although there are some strain differences.242 Campylobacter jejuni is susceptible to human complement,75 as is C. coli, but C. fetus (which causes bacteremia more commonly) is more resistant.26 Human macrophages have been shown to be able to phagocytose and kill C. jejuni efficiently independent of strain, but, interestingly, macrophages from about 10% of donors, although able to phagocytose the organisms, did not kill them.336 Plasmid involvement in virulence has been largely discounted because plasmids have been found in less than 10% of isolates. However, at least one strain has been shown to carry a plasmid that contains possible virulence genes.18 Similarly the occurrence of strains producing an enterotoxin with properties similar to cholera toxin and binding to ganglioside has been suggested, but consistent evidence is lacking.335 Nonetheless, a putative enterotoxigenic strain

has been shown to suppress the murine humoral response to red blood cells.241 An increasing incidence of GBS with a high association with enteritis and C. jejuni in Curacao could be due to a recent introduction of a new strain or increased exposure.323 The immune response against ganglioside oligosaccharide may not be just humoral but may also involve T cells. The preponderance of IgG1/IgG3 antiganglioside antibodies93,350 suggests a close T-cell involvement with B cells, which may arise from the antiganglioside B cells acting to present a C. jejuni–derived protein antigen.93,344 In addition, the neuropathy associated with antecedent C. jejuni infection is not always AMAN. Acute inflammatory demyelinating polyradiculoneuropathy (AIDP) is relatively frequent105,258,322 and occurs in the presence of antiganglioside antibodies.284 T cells responsive to C. jejuni antigens derived from nerve biopsy from a GBS patient were found to be predominantly of the / receptor type,23 and a cell line that was developed was shown to have an unusual restricted  gene usage.48 The number of circulating / cells expressing V 1 is increased in GBS.30 A similar expansion has been noted following a number of infections, including human immunodeficiency virus (HIV).32 The / T cells are very important in bacterial and viral infections and in particular for downregulating the / T-cell response and the concomitant inflammation.99 The / T cells have more resemblance to cells of the innate system, carrying a number of NK cell–type receptors and having wide antigen specificities that are not MHC restricted. Although it has been suggested that they respond to lipid antigens, this has not been clearly demonstrated. However, lipid antigens, including glycosphingolipids, are presented by the invariant MHClike molecule CD1d.212,218 There are four CD1 molecules in humans, CD1a through CD1d. They typically present antigen to stimulate / T cells, including CD4 /CD8

cells as well as CD4 and CD8 cells. The CD1 molecules are expressed on myeloid cells, in particular dendritic or predendritic cells, and occur in the peripheral nerve in inflammatory neuropathy.141 CD1d is able to present -galactosylceramide but not -galactosylceramide in humans.138 A CD1b-dependent stimulation of CD8 / T cells by glycolipids has been shown in multiple sclerosis patients.280 However, CD1b is known to present bacterial glycolipids by binding to fatty acid components and is not specific for the carbohydrate component. There thus remains the possibility that oligosaccharide haptens may be directly involved in T-cell activation, but neither T-cell receptor specificity nor the ability of a CD1 molecule to present LPS has been clearly demonstrated. Both CD1 and the / T cell are clearly important factors in the innate response to infection and may be implicated in the variant response shown by GBS patients to certain infective agents.

Peripheral Nerve Antigens

PATHOPHYSIOLOGIC RELEVANCE OF AUTOANTIGENS IN ANIMAL MODELS Galactocerebroside-Induced EAN It is difficult to induce immune responses to GalC in rodents, but repeated immunization of rabbits produces high titers of complement-fixing antibody and a subacute neuropathy. The neuropathy is characterized by macrophage-associated demyelination in the absence of T-cell infiltration, a point of difference from the EAN induced by proteins (discussed below).265,299 The presumption is that the disease is produced by a direct action of the antibodies on the myelin with complement fixation, with the resulting opsonization leading to vesicular dissolution of myelin and macrophage invasion without the detected participation of T cells. Whether anti-GalC is expressed on the plasma membrane of Schwann cells and leads to their lysis is not clear. In support of this hypothesis, anti-GalC serum induces demyelination when applied to myelinated tissue cultures,266 and following intraneural injection into the rat sciatic nerve.116,264,300 When antibody to GalC is injected into rats with adoptive transfer EAN induced by P2, the severity of the disease and amount of demyelination are increased.96 Galactocerebroside is abundant in CNS as well as peripheral nerve myelin. Therefore, it is not surprising that rabbits immunized with GalC and carrying high titers of circulating antibody experienced an inflammatory response–induced demyelination of the optic nerve.85 P0 EAN Active immunization of Lewis rats with bovine P0 produces an EAN whose clinical and light microscopic features resemble those induced by whole myelin, although detailed morphologic comparisons have not been made.208 With immunization using either of two synthetic peptides of P0 protein, the resulting EAN was more severe and prolonged with the cytoplasmic domain peptide 180–199 than with the extracellular domain peptide 56–71.368 Adoptive transfer of CD4 T-cell lines directed against the same peptides also induced EAN.183 Peptide 180–199 was immunodominant and was strongly recognized by T cells following immunization with whole myelin, whereas peptide 56–71 was not recognized. Antibodies to P0 appear early in the course of EAN induced by myelin and then subside. By contrast, antibodies to P2 appear later and persist, which argues for P0 being more important than P2 in producing myelin-induced disease.11,12 The relative importance of P0 was also demonstrated by showing that tolerance induced by nasal administration of P0 peptides 56–71 and 180–199 significantly reduced the severity of EAN induced with whole bovine myelin.369 Antibodies to P0 induce demyelination upon intraneural injection into rat sciatic nerves.116,356 However, it may be speculated that

593

antibodies directed against any extracellularly expressed antigen would induce demyelination if they were brought into direct contact with the external surface of the myelin sheath. Similarity between viral and myelin proteins might cause a viral infection to initiate a cross-reactive immune response, which would explain the pathogenesis of some cases of GBS. Attempting to further this hypothesis, Adelmann and Linington searched a protein sequence database for homologies between P0 and microbial proteins.1 They identified identical pentapeptides in P0 and proteins from Epstein-Barr virus, CMV, varicella, and HIV, all viruses that are recognized as precipitating GBS. Synthetic peptides from some of these proteins did not stimulate neuritogenic T-cell lines directed against P0 and did not induce EAN after injection into Lewis rats. PMP22 EAN There is a single report of the induction of EAN with recombinant rat PMP22 in Lewis rats.81 The disease induced was mild, but the histologic changes included perivascular mononuclear cell infiltrates and demyelinated nerve fibers, similar to the findings in AIDP. There is also a report of the detection of antibodies to PMP22 during the course of EAN induced in rats by myelin.144 Difficulties in purifying or expressing sufficient quantities of PMP22 have inhibited further investigation of its role in EAN.278 P2 EAN P2 was the first purified myelin protein to be identified as a neuritogenic antigen, and active immunization with it readily induces EAN in Lewis rats.114,133,134 Residues 61 through 70 form the longest amphipathic -helical domain in the P2 molecule, and represent the minimum sequence of peptide residues necessary to induce disease.238,239 This is consistent with the observation that such domains are immunodominant T-cell epitopes. Actively induced disease can be replicated in Lewis rats by adoptive transfer of CD4 T cells from rats immunized with P2 protein in a dose-dependent manner.182 Small numbers of cells induce mild disease in which the predominant change is demyelination, whereas larger numbers induce axonal degeneration and cause more severe disease.125,340 As in actively induced disease and also in GBS, the spinal cord is notably spared, although there is prominent meningeal infiltration surrounding the lumbosacral cord in EAN. The active and passive transfer models of P2-induced EAN in the Lewis rat have been enormously helpful in the dissection of the mechanisms of EAN and in studying potential therapies for autoimmune disease. For instance, intravenous administration of large amounts of recombinant human P2 ameliorated EAN, which had been induced either actively or passively by a synthetic peptide resembling bovine P2 residues 53 through 78.338 The likely explanation of this therapeutic effect is that antigen therapy

594

Neuroimmunology of the Peripheral Nervous System

induced T-cell apoptosis in the inflammatory lesions. More details of the immune responses in EAN and approaches to treatment are given in Chapter 27.

MAG-Induced Neuropathy Models The direct induction of EAN with MAG has not been reported. However, adoptive transfer of four T-cell lines directed against different peptide sequences predicted to be T-cell epitopes from the rat MAG extracellular domain induced mild inflammation of the central nervous system in Lewis rats.337 Two of the lines, one directed against residues 20 through 40 and another directed against residues 354 through 377, also induced mild inflammation but not demyelination in the dorsal roots and sciatic nerves. The presence of inflammation in both the PNS and CNS is not surprising because MAG is present in both. The experiment further illustrates that immunization of appropriate animals with many CNS or PNS myelin antigens will induce inflammation in neural tissue in which that epitope is expressed. Injection of human serum containing anti-MAG antibodies has induced demyelination in animal nerve using species that contain MAG expressing the HNK-1 carbohydrate epitope.100,349 Repeated injection of purified human anti-MAG IgM induced demyelination and widely spaced myelin in chick nerves.308 These experimental observations, and the demonstration of complement components and IgM on myelin sheaths in a location similar to the widely spaced myelin, make it very likely that the antibody is causative.

Ganglioside-Induced Neuropathy Models GM1 Model Despite the failure of several previous attempts, Yuki et al.365 recently succeeded in inducing an anti-GM1– mediated experimental neuropathy in rabbits by repeated immunization with a ganglioside mixture. All of the injected animals developed anti-GM1 antibodies, which switched class from IgM to IgG, and all suffered an acute monophasic neuropathy. The histologic appearances were those of an axonal neuropathy with wallerian degeneration but no lymphocytic infiltration or primary demyelination. Immunoglobulin deposition was demonstrated on peripheral nerve axons. A similar disease was also induced by immunization with purified GM1.365 This model is similar to the acute motor axonal neuropathy form of GBS. GD1b Model In a seminal study, another interesting model comprising an ataxic neuropathy was induced in rabbits with purified ganglioside GD1b.167 The affected rabbits had a dorsal root ganglionopathy with loss of DRG cells, degeneration of sensory axons and the dorsal column, but no inflammation

or involvement of the motor axons. Immunized rabbits all developed anti-GD1b antibodies, but only half developed the neuropathy. However, the demonstration that a monoclonal antibody directed against GD1b labeled about 50% of rabbit DRG neurons added strength to the argument that the antibody response was responsible for the neuropathy.

PATHOPHYSIOLOGIC RELEVANCE OF AUTOANTIGENS IN HUMAN NEUROPATHIES This chapter has been restricted to a description of antigens that have been shown to be or suspected as relevant to the pathogenesis of human neuropathy. Many of the specific clinical and immunologic phenotypes associated with these antigens are principally covered in Part N, Chapters 99 through 108, and the present section only briefly summarizes these findings in relation to the preceding text.

Neuronopathies Paraneoplastic neuropathies are often associated with antineuronal antibodies, and the most characteristic syndrome is a subacute sensory neuronopathy associated with SCLC and antibodies against Hu, a neuronal RNA-binding protein that is also present in SCLC (see above). The clinical picture is usually that of an asymmetrical proximal painful sensory syndrome often associated with other features, especially limbic encephalopathy, cerebellar, brainstem, and autonomic dysfunction; and weakness.38,56 Electrophysiologic testing also shows subclinical motor nerve conduction abnormalities.38 The finding of anti-Hu antibodies is highly specific for this condition and is usually associated with a SCLC, although a wide variety of other neoplasms have been reported.41,56 Antibodies against voltage-regulated Ca2 channels have been detected as co-occurring in patients with anti-Hu antibodies.176 A small number of patients with an acute sensory neuronopathy and antibodies to ganglioside GD1b and other gangliosides have been reported,240,343 consistent with the model described by Kusunoki et al. above.167 However, the association of this antibody with acute sensory neuropathy is not absolute. Another Japanese study identified 9 patients of 445 with GBS who had antibodies to ganglioside GD1b but no other glycolipids. Their clinical features were those of AIDP, but all had sensory changes.210

Axonopathies Paraneoplastic neuropathies associated with the antineuronal antibodies described in the previous paragraph, and others detailed below, may also present as distal sensory or

Peripheral Nerve Antigens

sensory and motor neuropathies caused by axonopathy. Antibodies to CV2 or CRMP are associated with a subacute sensorimotor neuropathy and frequently also with cerebellar symptoms.7,359 Endoneurial inflammation and demyelination have been found in some patients.359 However, antibodies to this group of antigens are not always associated with peripheral neuropathy. Of 116 patients with anti–CRMP-5 antibodies, only 47% had signs of peripheral neuropathy.359 As might be expected, the success or otherwise of treatment is dependent on the stage and severity of the disease. Removal of the tumor and immunosuppressive treatment while the neurologic impairment was not severe led to good recovery,79 but in patients with severe neurologic impairment, immunosuppression was not successful.22,25,56 However, if T cells are the agents of neuronal loss, possibly specific T-cell directed therapy might be more successful.3 Sensory neuropathy or sensory and motor neuropathy is also associated with other neoplasms and other antineuronal antibodies. For instance, the cerebellar degeneration associated with anti-Yo antibodies and ovarian carcinoma and other tumors may be associated with peripheral neuropathy.248 Another example of an immunologically mediated axonopathy is the AMAN associated with antibodies to gangliosides GM1, GD1a, and other gangliosides to be discussed in the next section. There have been several reports of antibodies to sulfatide being associated with sensory neuropathy.220,247 Antibodies to sulfatide stained a population of DRG neurons.254 In one report, IgM antibodies to sulfatide in the absence of a paraprotein were associated with a sensory axonopathy, and the antibodies stained axons.189 In the same report, antibodies to sulfatide in the presence of an IgM paraprotein were associated with a motor and sensory demyelinating neuropathy, and the sera stained myelin. However, in a conflicting report, of five patients with IgM paraproteins reactive to sulfatide alone, three had a sensory axonal neuropathy and two a predominantly motor demyelinating neuropathy.71 It is apparent that further research is needed to clarify the role and antigen specificity of antibodies to sulfatide in causing peripheral neuropathy.

Inflammatory Demyelinating Polyradiculoneuropathies P0 Antibodies Antibodies to P0 were found in 16% to 29% of patients with chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) in two studies139,356 but not in a third.168 Sera containing antibodies to P0 have been shown to produce demyelination following injection into rat sciatic nerve.116,356 Antibodies to a 35-kDa isoform of P0 have been found in CIDP but are not specific, being also found in motor neuron disease and other conditions.123,207,225 The

595

nature of this isoform of P0 is unclear, but it does not appear to represent a variation in glycosylation.206 Most effort has been directed toward detecting antibody responses to P0 in human neuropathies, despite the clear evidence that in EAN the T-cell responses are fundamentally important. However, Khalili-Shirazi et al. identified T-cell responsiveness to purified human P0 protein, judged by tritiated thymidine incorporation of blood mononuclear cells in the presence of antigen, in 6 of 19 patients, and to synthetic rat P0 peptides in 3 others.142 Responses in controls were rare. In the only other paper to address this issue,54 patients with GBS developed interleukin-4 (IL-4) responses to synthetic human peptides corresponding to those previously found to induce EAN.183 However, the responses were delayed, and IL-4 responses are characteristic of helper T cell type 2 responses, which are more associated with recovery from than initiation of autoimmune disease. PMP22 Antibodies There are no reports about T-cell responses to PMP22 in humans and conflicting reports concerning the presence of antibodies. Gabriel et al. identified IgG or IgM antibodies by enzyme-linked immunosorbent assay (ELISA) to synthetic peptides representing either the first or second extracellular domains in 58% of 19 GBS patients, 41% of 17 CIDP patients, 10% of 30 other neuropathy controls, and 4% of 51 normal subjects.80 The presence of antibodies was confirmed by identifying an appropriate 22-kDa band on a Western blot of human cauda equina in most of the GBS and CIDP patients but not usually in the others. However, two other groups have failed to identify such a clear relationship between antibodies to PMP22 and inflammatory demyelinating polyradiculoneuropathy. One study identified antibodies to recombinant PMP22 produced in E. coli in 23% of normal subjects and in 23% to 75% of patients with various sorts of neuropathy.262 However, these authors could not identify staining of PMP22-transfected cells with any of the positive sera. The other study did not identify antibody to PMP22 expressed in Chinese hamster ovary cells in any of 24 patients with CIDP and 25 patients with GBS.168 Therefore, the relevance of immunity to PMP22 in this disorder remains unclear, and resolution of these discrepant results will require further research. P2 Antibodies Because P2 was the first protein to be shown capable of inducing EAN, there have been several attempts to find antibodies to P2 in GBS and CIDP patients, and most have been negative,115 apart from one report in which a minority of patients had very high titers of IgM antibodies to bovine P2 protein.141 T-cell responses to P2 were sought in two studies. Khalili-Shirazi et al. used the mononuclear cell transformation technique with bovine P2 protein and

596

Neuroimmunology of the Peripheral Nervous System

identified responses in 6 of 19 GBS and 5 of 13 CIDP patients but also in occasional control subjects.142 These experiments would be worth repeating with human P2 protein. Dahle et al., using the ELISPOT technique, found that synthetic peptides of human P2 induced exaggerated production of interferon- by lymphocytes from one of seven patients with GBS.56

Ganglioside Antibody–Related Diseases Antibodies to gangliosides are found in both acute and chronic neuropathy syndromes. In the former, they tend to be of the IgG class, arise transiently after preceding infections, and disappear concomitant with clinical recovery. As described above, molecular mimicry is believed to be one major mechanism by which they arise. In chronic neuropathy syndromes, the antibodies tend to be of the IgM class and are persistently present over many years, often as IgM paraproteins, and in this situation the relation to molecular mimicry is less clear. Anti-GM1 IgM antibodies and their related clinical syndromes have been thoroughly studied since first being identified in MMN with demyelinating conduction block.149 IgM antibodies to GM1 are found in about 50% of MMN cases, but the figures vary depending upon the assay methodology used. Patients with MMN have a stereotyped clinical picture comprising a chronic asymmetrical motor syndrome, often with distal onset in an upper limb. Atypical cases may occur without focal conduction block in which the phenotype resembles chronic spinal muscular atrophy. Despite evidence of minor sensory involvement in some reported cases, anti-GM1 antibodies are clearly a marker for predominantly motor nerve syndromes. Antibodies to a wide range of glycolipids, including GM 1 , GM 1 (NeuGc), GM 1b , GalNAc-GM 1b , GD 1a , GalNAc-GD1a, GD1b, 9-O-acetyl-GD1b, GD3, GT1a, GT1b, GQ1b, GQ1ba, LM1, GalC, and SGPG, have now been reported either as case reports or larger clinical series in patients with GBS.237,253,351,360 The strongest associations are found in patients with AMAN and MFS. The prevalence of anti-GM1 antibodies in AMAN and AIDP varies from study to study, ranging from 0 to 80%, reflecting diverse variables including the prevalence of C. jejuni infection in the study population and the methodology used for antibody measurement.40,360 In the United States and Europe, approximately 20% of GBS cases have such antibodies. Anti-GD1a antibodies are also found in AMAN to a much greater extent than in AIDP. In a series of 138 AMAN and AIDP GBS patients from China, 60% of AMAN but only 4% of AIDP cases had anti-GD1a antibodies.106 Antibodies to GM1b and GalNAc-GD1a are also present in AMAN cases.135,157,361 In AIDP, antibodies to gangliosides and other glycolipids have been less consistently found, although they are certainly present in a proportion of cases. GBS occurring

in association with CMV infection has been linked with anti-GM2 antibodies.328 Mycoplasma pneumoniae infection preceding GBS has been reported in association with anti-GalC antibodies as described below. Anti-LM1 and anti-SGPG antibodies have also been reported in AIDP.253,353,360 MFS is very strongly associated with anti-GQ1b ganglioside antibodies.159 These antibodies are a very sensitive and specific marker for syndromes characterized by ophthalmoplegia.237,345 In MFS, anti-GQ1b IgG antibodies are elevated in over 90% of cases during the acute phase but may disappear rapidly, often being absent during convalescence. Diagnostic testing should therefore be conducted on serum samples drawn early in the course of the disease. Anti-GQ1b antibodies invariably cross-react with GT1a and in 50% of cases with GD3 and/or GD1b.

Galactocerebroside Antibodies to GalC have been found in GBS associated with M. pneumoniae infections,158,166 and the antibody activity could be absorbed by M. pneumoniae. Additionally, antibodies to GalC have been found in M. pneumoniae– associated encephalitis,223 providing strong evidence of molecular mimicry between glycolipid epitopes in M. pneumoniae and GalC.

Charcot-Marie-Tooth Disease Ritz et al. reported antibodies detected by ELISA against the whole recombinant human PMP22 molecule in 70% of patients with CMT1 and 60% with CMT2 as well as 23% of normal subjects.262 Gabriel et al. also found antibodies to PMP22 extracellular domain peptides by ELISA in 28% of 53 patients with CMT1a compared with only 10% of 30 patients with other neuropathies and 4% of 51 normal subjects. Of the patients with CMT1A, 12 had a stepwise progressive course, and of these four had antibodies.78 It was suggested that immune responses to myelin proteins, especially one that is overexpressed or mutated, might contribute to producing nerve damage in a subset of patients with hereditary neuropathy.

Paraproteinemic Neuropathies Myelin-Associated Glycoprotein About half of patients with paraproteinemia and demyelinating neuropathy have an IgM paraprotein that has antibody activity directed against MAG. The associated neuropathy is variable but is most commonly an insidiously progressive predominantly sensory neuropathy associated with a tremor in about a third of cases.224 The fine specificity of the antibody activity varies from patient to patient, with the paraproteins from many patients reacting with the HNK-1 epitope that is also present on P0, although the

597

Peripheral Nerve Antigens

fine structure of the oligosaccharides on the two proteins is different. The epitope is present only on tri- and quatroantennary oligosaccharides on P0, but on all types of oligosaccharide on MAG.36 As a consequence of the differences in fine specificity, the pattern of binding of sera with anti-MAG antibodies to peripheral nerve is variable. Most sera bind noncompact myelin in the periaxonal and abaxonal Schwann cell membranes and SchmidtLanterman incisures where MAG is located, but a few sera bind axons.189 In biopsied nerves from patients, bound IgM is found throughout the myelin sheath. Other Sulfated Antigens Most sera that react with MAG also react with SGPG, a myelin glycolipid that shares an epitope with MAG.71 Of patients with antibodies reactive with SGPG and not MAG or sulfatide, one had a slowly progressive sensory and motor demyelinating neuropathy and one a predominantly motor demyelinating neuropathy.71 There are a small number of reports of antibody activity against chondroitin sulfate in the serum of patients with IgM paraproteinemia and sensory axonal neuropathy.220,286 Chondroitin sulfate is a glycosaminoglycan present in axonal membranes and not myelin. In about half these cases the serum also shows antibody activity against sulfatide, and so may be directed against a shared galactosesulfatide epitope. GD1b Ganglioside An uncommon sensory ataxic neuropathy syndrome first described in 1985 is a variant of IgM paraproteinemic neuropathy in which the paraprotein reacts with NeuNAc(2-8)NeuNAc(2-3)-configured disialylated gangliosides, including GD1b, GD3, GD2, GT1a, GT1b, and GQ1b. In some publications this syndrome has been referred to as CANOMAD (chronic ataxic neuropathy with ophthalmoplegia, M protein, agglutination, and disialosyl antibodies).236,348 The clinical pattern comprises profound loss of limb kinesthesia (proprioception and vibration sense) with relative preservation of limb muscle strength. The patients therefore present with gait and upper limb ataxia. In addition, they often have craniobulbar motor involvement, either as a fixed set of symptoms or as relapsing-remitting symptoms; these comprise diplopia, ptosis, dysphagia, and dysarthria. The clinical course is chronic, often extending over more than 10 years, and is interspersed with episodes of relapse particularly affecting the craniobulbar and respiratory motor system. Patients are generally within the typical age group for paraproteinemic neuropathy (60 years) but may be younger. Electrophysiologically, motor studies show normal or moderately reduced conduction velocities and normal or prolonged distal motor latencies and F-wave latencies. Denervation changes are variable. Sensory nerve action potentials are absent or markedly reduced, and H reflexes are absent. Where performed, sural nerve biopsies

have shown varying degrees of demyelination and large axonal loss with relative preservation of small and unmyelinated fiber bundles. A partial response to intravenous immunoglobulin and other treatments has been reported in some cases.

Protein Antigens There are also a number of isolated reports of antibody activity directed against a variety of different proteins in the sera of patients with paraproteins. For instance, one patient with an axonal neuropathy had an IgM paraprotein and antibody activity against the 200-kDa neurofilament protein.33 However, this antibody also reacted with ribosomal proteins, and low-titer antibodies to this neurofilament protein are not uncommon in normal sera.297 Antibodies against the 170-kDa Schwann cell antigen L-periaxin have been detected in 3 of 17 patients with neuropathy and an IgG paraproteinemia.174 These antibodies were, however, IgM and polyclonal and could reflect a secondary response. Similar antibodies (but possibly monoclonal) were also found in some patients with type 2 diabetes and an associated neuropathy. No similar antibodies were found in nine patients with an IgM paraprotein and neuropathy. Intraneural injection of periaxinpositive sera with added complement produced sensory nerve conduction changes and structural indications of demyelination and axonal changes.

Channelopathies Isaacs’ Syndrome Isaacs’ syndrome consists of muscle cramps, slow relaxation of muscles following contraction, and hyperhidrosis. The electrophysiologic characteristics are spontaneous discharges of motor units as doublets, triplets, or multiplets at 5 to 150 Hz (myokymia) or faster (neuromyotonia). The serum of many patients contains antibodies directed against dendrotoxin-sensitive fast voltage-gated potassium channels, which are present at paranodes and internodes. The antibodies probably gain access to the motor nerve fibers in the anterior roots or at the motor nerve terminals, where the blood-nerve barrier is relatively deficient, and thus trigger the spontaneous discharges.14,287 Lambert-Eaton Syndrome Lambert-Eaton syndrome is a well-recognized syndrome often associated with a SCLC. The main clinical features are weakness, fatigue, and autonomic symptoms. It is due to antibodies directed against the P/Q type of voltagegated calcium channels that lead to the release of acetylcholine vesicles from the motor end plate following arrival of an action potential.176 It responds well to treatments

598

Neuroimmunology of the Peripheral Nervous System

that lower the antibody titer, including plasma exchange, intravenous immunoglobulin, and immunosuppression.

16.

REFERENCES 1. Adelmann, M., and Linington, C.: Molecular mimicry and the autoimmune response to the peripheral nerve myelin P0 glycoprotein. Neurochem. Res. 17:887, 1992. 2. Adlkofer, K., Martini, R., Aguzzi, A., et al.: Hypermyelination and demyelinating peripheral neuropathy in Pmp22deficient mice. Nat. Genet. 11:274, 1995. 3. Albert, M. L., Austin, L. M., and Darnell, R. B.: Detection and treatment of activated T cells in the cerebrospinal fluid of patients with paraneoplastic cerebellar degeneration. Ann. Neurol. 47:9, 2000. 4. Alderuccio, F., Rolland, J. M., Toner, G. C., et al.: Autoantibodies to neurons and to the cytoskeleton in small cell carcinoma with paraneoplastic sensory neuropathy. Autoimmunity 5:115, 1989. 5. Ando, S., Hirabayashi, Y., Kon, K., et al.: A trisialoganglioside containing a sialyl 2-6 N-acetylgalactosamine residue is a cholinergic-specific antigen, Chol-1a. J. Biochem. 111:287, 1992. 6. Ang, C. W., de Klerk, M. A., Endtz, H. P., et al.: GuillainBarré syndrome- and Miller Fisher syndrome-associated Campylobacter jejuni lipopolysaccharides induce anti-GM1 and anti-GQ1b antibodies in rabbits. Infect. Immun. 69:2462, 2001. 7. Antoine, J. C., Honnorat, J., Camdessanche, J. P., et al.: Paraneoplastic anti-CV2 antibodies react with peripheral nerve and are associated with a mixed axonal and demyelinating peripheral neuropathy. Ann. Neurol. 49:214, 2001. 8. Antoine, J. C., Honnorat, J., Vocanson, C., et al.: Posterior uveitis, paraneoplastic encephalomyelitis and auto-antibodies reacting with developmental protein of brain and retina. J. Neurol. Sci. 117:215, 1993. 9. Apostolski, S., Sadiq, S. A., Hays, A., et al.: Identification of Gal(1-3)GalNAc bearing glycoproteins at the nodes of Ranvier in peripheral nerve. J. Neurosci. Res. 38:134, 1994. 10. Arasaki, K., Kusonoki, S., Kudo, N., and Kanazawa, I.: Acute conduction block in vitro following exposure to antiganglioside sera. Muscle Nerve 16:587, 1993. 11. Archelos, J., Toyka, K., and Hartung, H.: B cell responses to the PNS protein P0 in experimental autoimmune neuritis. J. Neurol. Sci. 128:111, 1995. 12. Archelos, J. J., Roggenbuck, K., Schneider-Schaulies, J., et al.: Detection and quantification of antibodies to the extracellular domain of P0 during experimental allergic neuritis. J. Neurol. Sci. 117:197, 1993. 13. Ariga, T., Kohriyama, T., and Freddo, L.: Characterization of sulfated glucuronic acid containing glycolipids reacting with IgM M-proteins in patients with neuropathy. J. Biol. Chem. 262:848, 1987. 14. Arimura, K., Sonoda, Y., Watanabe, O., et al.: Isaacs’ syndrome as a potassium channelopathy of the nerve. Muscle Nerve 25(Suppl. 11):S55, 2002. 15. Aspinall, G. O., Fujimoto, S., McDonald, A. G., et al.: Lipopolysaccharides from Campylobacter jejuni associated

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

with Guillain Barre syndrome patients mimic human gangliosides in structure. Infect. Immun. 62:2122, 1994. Aspinall, G. O., McDonald, A. G., Pang, H., et al.: Lipopolysaccharides of Campylobacter jejuni serotype O:19: structure of core oligosaccharide regions from the serostrain and two bacterial isolates from patients with the Guillain-Barré syndrome. Biochemistry 33:241, 1994. Aspinall, G. O., McDonald, A. G., Raju, T. S., et al.: Chemical structures of the core regions of Campylobacter jejuni serotypes O:1, O:4, O:23, and O:36 lipopolysaccharides. Eur. J. Biochem. 213:1017, 1993. Bacon, D. J., Alm, R. A., Burr, D. H., et al.: Involvement of a plasmid in virulence of Campylobacter jejuni 81-176. Infect. Immun. 68:4384, 2000. Bahmanyar, S., Moreau-Dubois, M.-C., Brown, P., et al.: Serum antibodies to neurofilament antigens in patients with neurological and other diseases and in healthy controls. J. Neuroimmunol. 5:191, 1983. Barclay, A. N., Wright, G. J., Brooke, G., and Brown, M. H.: CD200 and membrane protein interactions in the control of myeloid cells. Trends Immunol. 23:285, 2002. Bataller, L., Graus, F., Saiz, A., et al.: Clinical outcome in adult onset idiopathic or paraneoplastic opsoclonus-myoclonus. Brain 124:437, 2001. Batchelor, T. T., Platten, M., and Hochberg, F. H.: Immunoadsorption therapy for paraneoplastic syndromes. J. Neurooncol. 40:131, 1998. Ben-Smith, A., Gaston, J. S., Barber, P. C., and Winer, J. B.: Isolation and characterisation of T lymphocytes from sural nerve biopsies in patients with Guillain-Barré syndrome and chronic inflammatory demyelinating polyneuropathy. J. Neurol. Neurosurg. Psychiatry 61:362, 1996. Benyahia, B., Liblau, R., Merle-Beral, H., et al.: Cell-mediated autoimmunity in paraneoplastic neurological syndromes with anti-Hu antibodies. Ann. Neurol. 45:162, 1999. Blaes, F. Immunotherapeutic approaches to paraneoplastic neurological disorders. Expert Opin. Invest. Drugs 9:727, 2000. Blaser, M. J., Smith, P. F., and Kohler, P. F.: Susceptibility of Campylobacter isolates to the bactericidal activity of human serum. J. Infect. Dis. 151:227, 1985. Boerkoel, C. F., Takashima, H., Stankiewicz, P., et al.: Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am. J. Hum. Genet. 68:325, 2001. Bollensen, E., Steck, A. J., and Schachner, M.: Reactivity with the peripheral myelin glycoprotein P(0) in serum from patients with monoclonal IgM gammopathy and polyneuropathy. Neurology 38:1266, 1988. Bond, J. P., Saavedra, R. A., and Kirschner, D. A.: Expression and purification of the extracellular domain of human myelin protein zero. Protein Expr. Purif. 23:398, 2001. Borsellino, G., Koul, O., Placido, R., et al.: Evidence for a role of gammadelta T cells in demyelinating diseases as determined by activation states and responses to lipid antigens. J. Neuroimmunol. 107:124, 2000. Bosse, F., Zoidl, G., Wilms, S., et al.: Differential expression of two mRNA species indicates a dual function of peripheral myelin protein, PMP-22, in cell growth and myelination. J. Neurosci. Res. 37:529, 1994.

Peripheral Nerve Antigens 32. Boullier, S., Cochet, M., Poccia, F., and Gougeon, M. L.: CDR3-independent gamma delta V delta 1 T cell expansion in the peripheral blood of HIV-infected persons. J. Immunol. 154:1418, 1995. 33. Brindel, I., Preud’homme, J. L., Diaz, J. J., et al.: A human monoclonal IgM lambda specific for an epitope shared by the 200 kDa neurofilament protein, histones and ribosomal proteins. J. Autoimmun. 8:915, 1995. 34. Buchwald, B., Weishaupt, A., Toyka, K. V., and Dudel, J.: Pre- and postsynaptic blockade of neuromuscular transmission by Miller-Fisher syndrome IgG at mouse motor nerve terminals. Eur. J. Neurosci. 10:281, 1998. 35. Bullens, R. W. M., O’Hanlon, G. M., Wagner, E. R., et al.: Complex gangliosides at the neuromuscular junction are essential receptors for autoantibodies and botulinum neurotoxin but redundant for normal synaptic function. J. Neurosci. 22:6876, 2002. 36. Burger, D., Simon, M., Peruisseau, G., and Steck, A. J.: The epitope(s) recognised by HNK-1 antibody and IgM paraprotein in neuropathy is present on several N-linked oligosaccharide structures on human P0 and myelin-associated glycoprotein. J. Neurochem. 54:1569, 1990. 37. Calderon, R. O., Attema, B., and DeVries, G. H.: Lipid composition of neuronal cell bodies and neurites from cultured dorsal root ganglia. J. Neurochem. 64:424, 1995. 38. Camdessanche, J. P., Antoine, J. C., Honnorat, J., et al.: Paraneoplastic peripheral neuropathy associated with antiHu antibodies: a clinical and electrophysiological study of 20 patients. Brain 125:166, 2002. 39. Campagnoni, A. T., and Macklin, W. B.: Cellular and molecular aspects of myelin protein gene expression. Mol. Neurobiol. 2:41, 1988. 40. Carpo, M., Pedotti, R., Allaria, S., et al.: Clinical presentation and outcome of Guillain-Barre and related syndromes in relation to anti-ganglioside antibodies. J. Neurol. Sci. 168:78, 1999. 41. Chalk, C. H., Lennon, V. A., Stevens, J. C., and Windebank, A. J.: Seronegativity for type 1 antineuronal nuclear antibodies (“anti-Hu”) in subacute sensory neuronopathy patients without cancer. Neurology 43:2209, 1993. 42. Chassande, B., Leger, J. M., Younes-Chennoufi, A. B., et al.: Peripheral neuropathy associated with IgM monoclonal gammopathy: correlations between M-protein antibody activity and clinical/electrophysiological features in 40 cases. Muscle Nerve 21:55, 1998. 43. Chester, M. A., for the IUPAC-IUB Joint Commission on Biochemical Nomenclature (JCBN): Nomenclature of glycolipids—recommendations 1997. Eur. J. Biochem. 257:293, 1998. 44. Chiba, A., Kusunoki, S., Obata, H., et al.: Serum anti-GQ1b IgG antibody is associated with ophthalmoplegia in Miller Fisher syndrome and Guillain-Barre syndrome: clinical and immunohistochemical studies. Neurology 43:1911, 1993. 45. Chiba, A., Kusunoki, S., Obata, H., et al.: Ganglioside composition of the human cranial nerves, with special reference to pathophysiology of Miller Fisher syndrome. Brain Res. 745:32, 1997. 46. Chou, D. K. H., Ilyas, A. A., and Evans, J. E.: Structure of sulfated glucuronyl glycolipids in the nervous system react-

47.

48.

49.

50.

51.

52. 53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

599

ing with HNK-1 antibody and some IgM paraproteins in neuropathy. J. Biol. Chem. 261:11717, 1986. Coetzee, T., Fujita, N., Dupree, J., et al.: Myelination in the absence of galactocerebroside and sulfatide: normal structure with abnormal function and regional instability. Cell 86:209, 1996. Cooper, J. C., Ben-Smith, A., Savage, C. O. S., and Winer, J. B.: Unusual T cell receptor phenotype V gene usage of / T cells in a line derived from the peripheral nerve of a patient with Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 69:522, 2000. Corbo, M., Quattrini, A., Latov, N., and Hays, A. P.: Localization of GM1 and Gal(1-3)GalNAc antigenic determinants in peripheral nerve. Neurology 43:809, 1993. Corbo, M., Quattrini, A., Lugaresi, A., et al.: Patterns of reactivity of human anti-GM1 antibodies with spinal-cord and motor neurons. Ann. Neurol. 32:487, 1992. Corriveau, R. A., Huh, G. S., and Shatz, C. J.: Regulation of class I MHC gene expression in the developing and mature CNS by neural activity. Neuron 21:505, 1998. Crocker, P. R., and Varki, A.: Siglecs, sialic acids and innate immunity. Trends Immunol. 22:337, 2001. Croft, P. H., Henson, R. A., Urich, H., and Wilkinson, P. C.: Sensory neuropathy with bronchial carcinoma: a study of four cases showing serological abnormalities. Brain 87:501, 1964. Dahle, C., Ekerfelt, C., Vrethem, M., et al.: T helper type 2 like cytokine responses to peptides from P0 and P2 myelin proteins during the recovery phase of Guillain-Barré syndrome [abstract]. J. Neurol. Sci. 153:54, 1997. Dalmau, J., Graus, F., Cheung, N.-K. V., et al.: Major histocompatibility proteins, anti-Hu antibodies, and paraneoplastic encephalomyelitis in neuroblastoma and small cell lung cancer. Cancer 75:99, 1995. Dalmau, J., Graus, F., Rosenblum, M. K., and Posner, J. B.: Anti-Hu–associated paraneoplastic encephalomyelitis/sensory neuronopathy: a clinical study of 71 patients. Medicine (Baltimore) 71:59, 1992. David, C., Solimena, M., and De Camilli, P.: Autoimmunity in stiff-man syndrome with breast cancer is targeted to the C-terminal region of human amphiphysin, a protein similar to the yeast proteins, Rvs167 and Rvs161. FEBS Lett. 351:73, 1994. De Angelis, M. V., Di Muzio, A., Lupo, S., et al.: Anti-GD1a antibodies from an acute motor axonal neuropathy patient selectively bind to motor nerve fiber nodes of Ranvier. J. Neuroimmunol. 121:79, 2001. Dehaut, F., Haddad, K., Alhayek, G., and PouplardBarthelaix, A.: Autoantibodies against H- and M-subunits of neurofilaments are induced by PC12 cell grafts or lesions into different sites of rat brain. Neurosci. Lett. 165:59, 1994. Demeure, C. E., Tanaka, H., Mateo, V., et al.: CD47 engagement inhibits cytokine production and maturation of human dendritic cells. J. Immunol. 164:2193, 2000. Derrington, E. A., and Borroni, E.: The developmental expression of the cholinergic-specific antigen Chol-1 in the central and peripheral nervous system of the rat. Dev. Brain Res. 52:131, 1990. Derrington, E. A., Kelic, S., and Whittaker, V. P.: A novel cholinergic-specific antigen (Chol-2) in mammalian brain. Brain Res. 620:16, 1993.

600

Neuroimmunology of the Peripheral Nervous System

63. Dick, D. J., Harris, J. B., Falkous, G., et al.: Neuronal antinuclear antibody in paraneoplastic sensory neuronopathy. J. Neurol. Sci. 85:1, 1988. 64. Dodd, J., and Jessell, T. M.: Lactoseries carbohydrates specify subsets of dorsal root ganglion neurons projecting to the superficial dorsal horn of rat spinal cord. J. Neurosci. 5:3278, 1985. 65. Dodd, J., Solter, D., and Jessell, T. M.: Monoclonal antibodies against carbohydrate differentiation antigens identify subsets of primary sensory neurones. Nature 311:469, 1984. 66. Dorrell, N., Mangan, J. A., Laing, K. G., et al.: Whole genome comparison of Campylobacter jejuni human isolates using a low-cost microarray reveals extensive genetic diversity. Genome Res. 11:1706, 2001. 67. Dropcho, E. J.: Antiamphiphysin antibodies with small-cell lung carcinoma and paraneoplastic encephalomyelitis. Ann. Neurol. 39:659, 1996. 68. D’Urso, D., Ehrhardt, P., and Muller, H. W.: Peripheral myelin protein 22 and protein zero: a novel association in peripheral nervous system myelin. J. Neurosci. 19:3396, 1999. 69. D’Urso, D., and Muller, H. W.: Ins and outs of peripheral myelin protein-22: mapping transmembrane topology and intracellular sorting. J. Neurosci. Res. 49:551, 1997. 70. Engberg, J., Nachamkin, I., Fussing, V., et al.: Absence of clonality of Campylobacter jejuni in serotypes other than HS:19 associated with Guillain-Barre syndrome and gastroenteritis. J. Infect. Dis. 184:215, 2001. 71. Eurelings, M., Moons, K. G., Notermans, N. C., et al.: Neuropathy and IgM M-proteins: prognostic value of antibodies to MAG, SGPG, and sulfatide. Neurology 56:228, 2001. 72. Fabian, R. H.: Uptake of antineuronal IgM by CNS neurons: comparison with antineuronal IgG. Neurology 40:419, 1990. 73. Fabian, R. H.: Retrograde axonal transport and transcytosis of immunoglobulins: implications for the pathogenesis of autoimmune motor neuron disease [review]. Adv. Neurol. 56:433, 1991. 74. Fabian, R. H., and Petroff, G.: Intraneuronal IgG in the central nervous system: uptake by retrograde axonal transport. Neurology 37:1780, 1987. 75. Fernandez, H., Giusti, G., and Bertoglio, J. C.: Effect of the complement system on the sensitivity of Campylobacter jejuni and Campylobacter coli to human blood serum. Braz. J. Med. Biol. Res. 28:227, 1995. 76. Ferrari, S., Morbin, M., Nobile-Orazio, E., et al.: Antisulfatide polyneuropathy: antibody-mediated complement attack on peripheral myelin. Acta Neuropathol. 96:569, 1998. 77. Furuya, S., Hashikawa, T., Irie, F., et al.: Neuronal expression of a minor monosialosyl ganglioside GM1b in rat brain: immunochemical characterization using a specific monoclonal antibody. Neurosci. Res. 22:411, 1995. 78. Gabriel, C., Gregson, N., Wood, N. W., and Hughes, R. A. C.: An immunological study of hereditary motor and sensory neuropathy type 1a (HSMN1a). J. Neurol. Neurosurg. Psychiatry 72:230, 2002. 79. Gabriel, C. M., Gregson, N. A., and Hughes, R. A. C.: Sensory neuropathy and anti-Hu antibodies in a patient with seminoma. Eur. J. Neurol. 3:471, 1996.

80. Gabriel, C. M., Gregson, N. A., and Hughes, R. A. C.: Antibodies to PMP22 in patients with inflammatory neuropathies. J. Neurol. Neurosurg. Psychiatry 66:260, 1999. 81. Gabriel, C. M., Hughes, R. A. C., Moore, S. E., et al.: Induction of experimental neuritis with peripheral myelin protein 22. Brain 121:1895, 1998. 82. Ganser, A. L., and Kirschner, D. A.: Differential expression of gangliosides on the surfaces of myelinated nerve fibers. J. Neurosci. Res. 12:245, 1984. 83. Ganser, A. L., Kirschner, D. A., and Willinger, M.: Ganglioside localization on myelinated nerve fibres by cholera toxin binding. J. Neurocytol. 12:921, 1983. 84. Gilbert, M., Brisson, J. R., Karwaski, M. F., et al.: Biosynthesis of ganglioside mimics in Campylobacter jejuni OH4384: identification of the glycosyltransferase genes, enzymatic synthesis of model compounds, and characterization of nanomole amounts by 600-mHz (1)H and (13)C NMR analysis. J. Biol. Chem. 275:3896, 2000. 85. Goban, Y., Saida, T., Saida, K., et al.: Ultrastructural study of central nervous system demyelination in galactocerebroside sensitized rabbits. Lab. Invest. 55:86, 1986. 86. Gong, Y., Tagawa, Y., Lunn, M. P. T., et al.: Localization of major gangliosides in the PNS: implications for immune neuropathies. Brain 125:2491, 2002. 87. Goodyear, C. S., O’Hanlon, G. M., Plomp, J. J., et al.: Monoclonal antibodies raised against Guillain Barré syndrome-associated Campylobacter jejuni lipopolysaccharides react with neuronal gangliosides and paralyse nerve muscle preparations. J. Clin. Invest. 104:697, 1999. 88. Granieri, E., Casetta, I., Govoni, V., et al.: Ganglioside therapy and Guillain-Barre syndrome: a historical cohort study in Ferrara, Italy, fails to demonstrate an association. Neuroepidemiology 10:161, 1991. 89. Graus, F., Campo, E., Cruz Sanchez, F., et al.: Expression of lymphocyte, macrophage and class I and II major histocompatibility complex antigens in normal human dorsal root ganglia. J. Neurol. Sci. 98:203, 1990. 90. Graus, F., Cordon-Cardo, C., and Posner, J.: Neuronal antinuclear antibody in sensory neuronopathy from lung cancer. Neurology 35:538, 1985. 91. Graus, F., Ribalta, T., Campo, E., et al.: Immunohistochemical analysis of the immune reaction in the nervous system in paraneoplastic encephalomyelitis. Neurology 40:219, 1990. 92. Gregson, N. A., and Leibowitz, S.: IgM paraproteinaemia, polyneuropathy and myelin-associated glycoprotein (MAG). Neuropathol. Appl. Neurobiol. 11:329, 1985. 93. Gregson, N. A., Rees, J. H., and Hughes, R. A. C.: Reactivity of serum IgG anti-GM1 antibodies with the lipopolysaccharide fractions of Campylobacter jejuni isolates from patients with Guillain Barre syndrome (GBS). J. Neuroimmunol. 73:28, 1997. 94. Guilbot, A., Williams, A., Ravise, N., et al.: A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum. Mol. Genet. 10:415, 2001. 95. Hadden, R. D., Karch, H., Hartung, H. P., et al.: Preceding infections, immune factors, and outcome in Guillain-Barre syndrome. Neurology 56:758, 2001. 96. Hahn, A. F., Feasby, T. E., Wilkie, L., and Lovgren, D.: Antigalactocerebroside antibody increases demyelination in

Peripheral Nerve Antigens

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

adoptive transfer experimental allergic neuritis. Muscle Nerve 16:1174, 1993. Hao, Q., Saida, T., Kuroki, S., et al.: Antibodies to gangliosides and galactocerebroside in patients with Guillain-Barre syndrome with preceding Campylobacter jejuni and other identified infections. J. Neuroimmunol. 81:116, 1998. Harvey, H. A., Swords, W. E., and Apicella, M. A.: The mimicry of human glycolipids and glycosphingolipids by the lipooligosaccharides of pathogenic Neisseria and Haemophilus. J. Autoimmun. 16:257, 2001. Hayday, A. C.:  Cells: a right time and a right place for a conserved third way of protection. Annu. Rev. Immunol. 18:975, 2000. Hays, A. P., Latov, N., Takatsu, M., and Sherman, W. H.: Experimental demyelination of nerve induced by serum of patients with neuropathy and an anti MAG IgM M protein. Neurology 37:242, 1987. Heckmann, J. G., Sommer, J. B., Druschky, A., et al.: Acute motor axonal neuropathy associated with IgM anti-GM1 following Mycoplasma pneumoniae infection. Eur. Neurol. 41:175, 1999. Hirabayashi, Y., Nakao, T., Irie, F., et al.: Structural characterization of a novel cholinergic neuron-specific ganglioside in bovine brain. J. Biol. Chem. 267:12973, 1992. Hirakawa, M., McCabe, T., and Kawata, M.: Time-related changes in the labeling pattern of motor and sensory neurons innervating the gastrocnemius muscle, as revealed by the retrograde transport of the cholera toxin B subunit. Cell Tissue Res. 267:419, 1992. Hitoshi, S., Kusunoki, S., Murayama, S., et al.: Rabbit experimental sensory ataxic neuropathy: anti-GD1b antibody-mediated trkC downregulation of dorsal root ganglia neurons. Neurosci. Lett. 260:157, 1999. Ho, T. W., Mishu, B., Li, C. Y., et al.: Guillain-Barre syndrome in northern China: relationship to Campylobacter jejuni infection and anti-glycolipid antibodies. Brain 118:597, 1995. Ho, T. W., Willison, H. J., Nachamkin, I., et al.: Anti-GD1a antibody is associated with axonal but not demyelinating forms of Guillain-Barre syndrome. Ann. Neurol. 45:168, 1999. Hoek, R. M., Ruuls, S. R., Murphy, C. A., et al.: Downregulation of the macrophage lineage through interaction with OX2 (CD200). Science 290:1768, 2000. Hong, J., Ohmura, K., Mahmood, U., et al.: Arthritis critically dependent on innate immune system players. Immunity 16:157, 2002. Honke, K., Hirahara, Y., Dupree, J., et al.: Paranodal junction formation and spermatogenesis require sulfoglycolipids. Proc. Natl. Acad. Sci. U. S. A. 99:4227, 2002. Honnorat, J., Antoine, J. C., and Belin, M. F.: Are the “newly discovered” paraneoplastic anticollapsin responsemediator protein 5 antibodies simply anti-CV2 antibodies? Ann. Neurol. 50:688, 2001. Honnorat, J., Antoine, J. C., Derrington, E., et al.: Antibodies to a subpopulation of glial cells and a 66 kDa developmental protein in patients with paraneoplastic neurological syndromes. J. Neurol. Neurosurg. Psychiatry 61:270, 1996. Hood, D. W., Cox, A. D., Gilbert, M., et al.: Identification of a lipopolysaccharide alpha-2,3-sialyltransferase from Haemophilus influenzae. Mol. Microbiol. 39:341, 2001.

601

113. Hormigo, A., and Lieberman F.: Nuclear localization of anti-Hu antibody is not associated with in vitro cytotoxicity. J. Neuroimmunol. 55:205, 1994. 114. Hughes, R. A. C.: Guillain-Barré Syndrome. Heidelberg, Springer-Verlag, 1990. 115. Hughes, R. A. C., Gregson N. A., Hadden R. D. M., and Smith K. J.: Pathogenesis of Guillain-Barré syndrome. J. Neuroimmunol. 100:74, 1999. 116. Hughes, R. A. C., Powell H. C., Braheny S. L., and Brostoff S. W.: Endoneurial injection of antisera to myelin antigens. Muscle Nerve 8:516, 1985. 117. Ilyas, A. A., Li, S. C., Chou, D. K. H., et al.: Gangliosides GM2, IV4GalNAcGM1b, and IV4GalNAcGD1a as antigens for monoclonal immunoglobulin M in neuropathy associated with gammopathy. J. Biol. Chem. 263:4369, 1988. 118. Ilyas, A. A., Willison, H. J., Dalakas, M. C., et al.: Identification and characterization of gangliosides reacting with IgM paraproteins in three patients with neuropathy associated with biclonal gammopathy. J. Neurochem. 51:851, 1988. 119. Inoue, M., Fujii, Y., Furukawa, K., et al.: Refractory skin injury in complex knock-out mice expressing only the GM3 ganglioside. J. Biol. Chem. 277:29881, 2002. 120. Irie, F., Kurono, S., Li, Y.-T., et al.: Isolation of three novel cholinergic neuron-specific gangliosides from bovine brain and their in vitro syntheses. Glycoconj. J. 13:177, 1996. 121. Irie, S., Kanazawa, N., Ogino, M., et al.: No cytomegalovirus DNA in sera from patients with anti-MAG/SGPG antibodyassociated neuropathy. Ann. Neurol. 47:274, 2000. 122. Ishibashi, T., Dupree, J. L., Ikenaka, K., et al.: A myelin galactolipid, sulfatide, is essential for maintenance of ion channels on myelinated axon but not essential for initial cluster formation. J. Neurosci. 22:6507, 2002. 123. Ishida, K., Takeuchi, H., Takahashi, R., et al.: A possible novel isoform of peripheral myelin PO protein: a target antigen recognized by an autoantibody in a patient with malignant lymphoma and peripheral neuropathy. J. Neurol. Sci. 188:43, 2001. 124. Ito, A., Handa, K., Withers, D. A., et al.: Binding specificity of siglec7 to disialogangliosides of renal cell carcinoma: possible role of disialogangliosides in tumor progression. FEBS Lett. 504:82, 2001. 125. Izumo, S., Linington, C., Wekerle, H., and Meyermann, R.: Morphological study on EAN mediated by T-cell line specific for bovine P2 protein in Lewis rats. Lab. Invest. 53:209, 1985. 126. Jacobs, B. C., Hazenberg, M. P., van Doorn, P. A., et al.: Cross-reactive antibodies against gangliosides and Campylobacter jejuni lipopolysaccharides in patients with Guillain-Barre or Miller Fisher syndrome. J. Infect. Dis. 175:729, 1997. 127. Jacobs, B. C., Rothbarth, P. H., van der Meché, F. G. A., et al.: The spectrum of antecedent infections in Guillain-Barré syndrome: a case control study. Neurology 51:1110, 1998. 128. Jacobs, B. C., van Doorn, P. A., Groeneveld, J. H., et al.: Cytomegalovirus infections and anti-GM2 antibodies in Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 62:641, 1997. 129. Jessen, K. R., Morgan, L., Brammer, M., and Mirsky, R.: Galactocerebroside is expressed by non-myelin-forming Schwann cells in situ. J. Cell Biol. 101:1135, 1985.

602

Neuroimmunology of the Peripheral Nervous System

130. Jeurissen, S. H., Janse, E. M., van Rooijen, N., and Claassen, E.: Inadequate anti-polysaccharide antibody responses in the chicken. Immunobiology 198:385, 1998. 131. Jones, P. A., Samuels, N. M., Phillips, N. J., et al.: Haemophilus influenzae type b strain A2 has multiple sialyltransferases involved in lipooligosaccharide sialylation. J. Biol. Chem. 277:14598, 2002. 132. Jones, T. A., Bergfors, T., Sedzik, J., and Unge, T.: The threedimensional structure of P2 myelin protein. EMBO J. 7:1597, 1988. 133. Kadlubowski, M., and Hughes, R. A. C.: Identification of the neuritogen responsible for experimental allergic neuritis. Nature 277:140, 1979. 134. Kadlubowski, M., Hughes, R. A. C., and Gregson, N. A.: Experimental allergic neuritis in the Lewis rat: characterisation of the activity of peripheral myelin and its major basic protein P2. Brain Res. 184:439, 1980. 135. Kaida, K., Kusunoki S., Kamakura K., et al.: Guillain-Barré syndrome with antibody to a ganglioside, N-acetylgalactosaminyl GD1a. Brain 123:116, 2000. 136. Kanda, T., Yamawaki, M., Iwasaki, T., and Mizusawa, H.: Glycosphingolipid antibodies and blood-nerve barrier in autoimmune demyelinative neuropathy. Neurology 54:1459, 2000. 137. Kanda, T., Yoshino, H., Ariga, T., et al.: Glycosphingolipid antigens in cultured microvascular bovine brain endothelial cells: sulfoglucuronosyl paragloboside as a target of monoclonal IgM in demyelinative neuropathy. J. Cell Biol. 126:235, 1994. 138. Kawano, T., Cui, J., Koezuka, Y., et al.: CD1d-restricted and TCR-mediated activation of valpha14 NKT cells by glycosylceramides. Science 278:1626, 1997. 139. Khalili-Shirazi, A., Atkinson, P., Gregson, N., and Hughes, R. A. C.: Antibody responses to P0 and P2 myelin proteins in Guillain-Barré syndrome and chronic idiopathic demyelinating polyradiculoneuropathy. J. Neuroimmunol. 46:245, 1993. 140. Khalili-Shiraz, A., Gregson, N., Gray, I., et al.: Antiganglioside antibodies in Guillain-Barré syndrome after a recent cytomegalovirus infection. J. Neurol. Neurosurg. Psychiatry 66:376, 1999. 141. Khalili-Shiraz, A., Gregson, N. A., Londei, M., et al.: The distribution of CD1 molecules in inflammatory neuropathy. J. Neurol. Sci. 158:154, 1998. 142. Khalili-Shirazi, A., Hughes, R. A. C., Brostoff, S., et al.: T cell response to myelin proteins in Guillain-Barré syndrome. J. Neurol. Sci. 111:200, 1992. 143. Kitazawa, K., Tagawa, Y., Honda, A., and Yuki, N.: GuillainBarré syndrome associated with IgG anti-GM1b antibody subsequent to Mycoplasma pneumoniae infection. J. Neurol. Sci. 156:99, 1998. 144. Koehler, N. K. U., Martin, R., and Wiethölter, H.: The antibody repertoire in experimental allergic neuritis: evidence for PMP-22 as a novel neuritogen. J. Neuroimmunol. 71:179, 1996. 145. Koga, M., Ang, C. W., Yuki, N., et al.: Comparative study of preceding Campylobacter jejuni infection in Guillain-Barré syndrome in Japan and the Netherlands. J. Neurol. Neurosurg. Psychiatry 70:693, 2001. 146. Koga, M., Yoshino, H., Morimatsu, M., and Yuki, N.: AntiGT1a IgG in Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 72:767, 2002.

147. Koga, M., Yuki, N., Tai, T., and Hirata, K.: Miller Fisher syndrome and Haemophilus influenzae infection. Neurology 57:686, 2001. 148. Kohriyama, T., Kusunoki, S., and Ariga, T.: Subcellular localization of sulfated glucuronic acid-containing glycolipids reacting with anti-myelin-associated glycoprotein antibody. J. Neurochem. 48:1516, 1987. 149. Kornberg, A. J., and Pestronk, A.: Chronic motor neuropathies: diagnosis, therapy, and pathogenesis. Ann. Neurol. 37(Suppl. 1):S43, 1995. 150. Kornguth, S. E., Kalinke, T., Grunwald, G. B., et al.: Antineurofilament antibodies in the sera of patients with small cell carcinoma of the lung and with visual paraneoplastic syndrome. Cancer Res. 46:2588, 1986. 151. Kotani, M., Kawashima, I., Ozawa, H., et al.: Immunohistochemical localization of minor gangliosides in the rat central nervous system. Glycobiology 4:855, 1994. 152. Kuhlmann, T., Bitsch, A., Stadelmann, C., et al.: Macrophages are eliminated from the injured peripheral nerve via local apoptosis and circulation to regional lymph nodes and the spleen. J. Neurosci. 21:3401, 2001. 153. Kullmann, D. M.: The neuronal channelopathies. Brain 125:1177, 2002. 154. Kumagai, T., Kitagawa, Y., Hirose, G., and Sakai, K.: Antibody recognition and RNA binding of a neuronal nuclear autoantigen associated with paraneoplastic neurological syndromes and small cell lung carcinoma. J. Neuroimmunol. 93:37, 1999. 155. Kuroki, S., Saida, T., Nukina, M., et al.: Campylobacter jejuni strains from patients with Guillain-Barré syndrome belong mostly to Penner serogroup 19 and contain betaN-acetylglucosamine residues. Ann. Neurol. 33:243, 1993. 156. Kusunoki, S.: Antiglycolipid antibody in inflammatory neuropathy. Clin. Neurol. 35:1370, 1995. 157. Kusunoki, S.: Antiglycolipid antibodies in Guillain-Barré syndrome and autoimmune neuropathies. Am. J. Med. Sci. 319:234, 2000. 158. Kusunoki, S., Chiba, A., Hitoshi, S., et al.: Anti-Gal-C antibody in autoimmune neuropathies subsequent to mycoplasma infection. Muscle Nerve 18:409, 1995. 159. Kusunoki, S., Chiba, A., and Kanazawa, I.: Anti-GQ1b IgG antibody is associated with ataxia as well as ophthalmoplegia. Muscle Nerve 22:1071, 1999. 160. Kusunoki, S., Chiba, A., Kon, K., et al.: N-acetylgalactosaminyl GD1a is a target molecule for serum antibody in Guillain-Barre syndrome. Ann. Neurol. 35:570, 1994. 161. Kusunoki, S., Chiba, A., Tai, T., and Kanazawa, I.: Localization of GM1 and GD1b antigens in the human peripheral nervous system. Muscle Nerve 16:752, 1993. 162. Kusunoki, S., Hitoshi, S., Kaida, K., et al.: Monospecific anti-GD1b IgG is required to induce rabbit ataxic neuropathy. Ann. Neurol. 45:400, 1999. 163. Kusunoki, S., Inoue, K., Iwamori, M., et al.: Discrimination of human dorsal root ganglion cells by anti-fucosyl GM1 antibody. Brain Res. 494:391, 1989. 164. Kusunoki, S., Inoue, K., Iwamori, M., et al.: Fucosylated glycoconjugates in human dorsal root ganglion cells with unmyelinated axons. Neurosci. Lett. 126:159, 1991. 165. Kusunoki, S., Mashiko, H., Mochizuki, N., et al.: Binding of antibodies against GM1 and GD1b in human peripheral nerve. Muscle Nerve 20:840, 1997.

Peripheral Nerve Antigens 166. Kusunoki, S., Shiina, M., and Kanazawa, I.: Anti-Gal-C antibodies in GBS subsequent to mycoplasma infection: evidence of molecular mimicry. Neurology 57:736, 2001. 167. Kusunoki, S., Shimizu, J., Chiba, R., et al.: Experimental sensory neuropathy induced by sensitisation with ganglioside GD1b. Ann. Neurol. 39:324, 1996. 168. Kwa, M. S., van Schaik, I. N., Brand, A., et al.: Investigation of serum response to PMP22, connexin 32 and P(0) in inflammatory neuropathies. J. Neuroimmunol. 116:220, 2001. 169. Laemmli, U. K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680, 1970. 170. Lastovica, A. J., Goddard, E. A., and Argent, A. C.: Guillain-Barré syndrome in South Africa associated with Campylobacter jejuni O:41 strains. J. Infect. Dis. 176:s139, 1997. 171. Latov, N.: Antibodies to glycoconjugates in neuropathy and motor-neuron disease. Proc. Brain Res. 101:295, 1994. 172. Latov, N., Hays, A. P., Donofrio, P. D., et al.: Monoclonal IgM with unique specificity to gangliosides GM1 and GD1b and to lacto-N-tetraose associated with human motor neuron disease. Neurology 38:763, 1988. 173. Latov, N., Hays, A. P., Yu, R. K., et al.: Antibodies to glycoconjugates in human motor neuron disease. Neurochem. Pathol. 8:181, 1988. 174. Lawlor, M. W., Richards, M. P., De Vries, G. H., et al.: Antibodies to L-periaxin in sera of patients with peripheral neuropathy produce experimental sensory nerve conduction deficits. J. Neurochem. 83:592, 2002. 175. Ledeen, R. W.: Gangliosides of the neuron. Trends Neurosci. 10:169, 1985. 176. Lennon, V. A., Kryzer, T. J., Griesmann, G. E., et al.: Calcium-channel antibodies in the Lambert-Eaton syndrome and other paraneoplastic syndromes. N. Engl. J. Med. 332:1467, 1995. 177. Levitt, D., Griffin, N. B., and Egan, M. L.: Mitogeninduced plasma cell differentiation in patients with multiple sclerosis. J. Neuroimmunol. 124:2117, 1980. 178. Lewis, R. A., Sumner, A. J., and Shy, M. E.: Electrophysiological features of inherited demyelinating neuropathies: a reappraisal in the era of molecular diagnosis. Muscle Nerve 23:1472, 2000. 179. Li, J. Y., De Camilli, P., and Dahlstrom, A.: Intraneuronal trafficking and distribution of amphiphysin and synaptojanin in the rat peripheral nervous system and the spinal cord. Eur. J. Neurosci. 9:1864, 1997. 180. Lindberg, F. P., Bullard, D. C., Caver, T. E., et al.: Decreased resistance to bacterial infection and granulocyte defects in IAP-deficient mice. Science 274:795, 1996. 181. Linington, C., and Brostoff, S. W.: Peripheral nerve antigens. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 404, 1993. 182. Linington, C., Izumo, S., Suzuki, M., et al.: A permanent rat T cell line that mediates experimental allergic neuritis in the rat in vivo. J. Immunol. 133:1946, 1984. 183. Linington, C., Lassmann, H., Ozawa, K., et al.: Cell adhesion molecules of the immunoglobulin supergene family as tissue-specific autoantigens: induction of experimental allergic neuritis (EAN) by P0 protein-specific T cell lines. Eur. J. Immunol. 22:1813, 1992.

603

184. Linton, D., Gilbert, M., Hitchen, P. G., et al.: Phase variation of a -1,3 galactosyltransferase involved in generation of the ganglioside GM1-like lipo-oligosaccharide of Campylobacter jejuni. Mol. Microbiol. 37:501, 2000. 185. Linton, D., Karlyshev, A. V., Hitchen, P. G., et al.: Multiple N-acetyl neuraminic acid synthetase (neuB) genes in Campylobacter jejuni: identification and characterization of the gene involved in sialylation of lipo-oligosaccharide. Mol. Microbiol. 35:1120, 2000. 186. Lisak, R. P., Saida, T., and Kennedy, P. G. E.: EAE, EAN and galactocerebroside sera bind to oligodendrocytes and Schwann cells. J. Neurol. Sci. 48:287, 1980. 187. Liu, J., Dalmau, J., Szabo, A., et al.: Paraneoplastic encephalomyelitis antigens bind to the Au-rich elements of mRNA. Neurology 45:544, 1995. 188. Lloyd, K. O., Gordon, C. M., Thampoe, I. J., and DiBenedetto, C.: Cell surface accessibility of individual gangliosides in malignant melanoma cells to antibodies is influenced by the total ganglioside composition of the cells. Cancer Res. 52:4948, 1992. 189. Lopate, G., Kornberg, A. J., Yue, J., et al.: Anti-myelin associated glycoprotein antibodies: variability in patterns of IgM binding to peripheral nerve. J. Neurol. Sci. 188:67, 2001. 190. Lopate, G., Pestronk, A., Kornberg, A. J., et al.: IgM antisulfatide autoantibodies: patterns of binding to cerebellum, dorsal root ganglion and peripheral nerve. J. Neurol. Sci. 151:189, 1997. 191. Lugaresi, A., Corbo, M., Thomas, F. P., et al.: Identification of glycoconjugates which are targets for anti-Gal(-1-3)GalNAc autoantibodies in spinal motor neurons. J. Neuroimmunol. 34:69, 1991. 192. Lunn, M. P., Crawford, T. O., Hughes, R. A., et al.: Antimyelin-associated glycoprotein antibodies alter neurofilament spacing. Brain 125:904, 2002. 193. Lunn, M. P., Muir, P., Brown, L. J., et al.: Cytomegalovirus is not associated with IgM anti-myelin-associated glycoprotein/sulphate-3-glucuronyl paragloboside antibody-associated neuropathy. Ann. Neurol. 46:267, 1999. 194. Madden, K. S., Sanders, V. M., and Felten, D. L.: Catecholamine influences and sympathetic neural modulation of immune responsiveness [review]. Annu. Rev. Pharmacol. Toxicol. 35:417, 1995. 195. Maehara, T., Ono, K., Tsutsui, K., et al.: A monoclonal antibody that recognizes ganglioside GD1b in the rat central nervous system. Neurosci. Res. 29:9, 1997. 196. Mandrell, R. E., McLaughlin, R., Aba Kwaik, Y., et al.: Lipooligosaccharides (LOS) of some Haemophilus species mimic human glycosphingolipids, and some LOS are sialylated. Infect. Immun. 60:1322, 1992. 197. Manley, G. T., Smitt, P. S., Dalmau, J., and Posner, J. B.: Hu antigens: reactivity with Hu antibodies, tumour expression, and major immunogenic sites. Ann. Neurol. 38:102, 1995. 198. Mansson, M., Hood, D. W., Li, J., et al.: Structural analysis of the lipopolysaccharide from nontypeable Haemophilus influenzae strain 1003. Eur. J. Biochem. 269:808, 2002. 199. Martini, R., and Schachner, M.: Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and MAG) and their shared carbohydrate epitope and myelin basic protein in developing sciatic nerve. J. Cell Biol. 103:2439, 1986.

604

Neuroimmunology of the Peripheral Nervous System

200. Martini, R., Zielasek, J., Toyka, K. V., et al.: Protein zero (P0)-deficient mice show myelin degeneration in peripheral nerves characteristic of inherited human neuropathies. Nat. Genet. 11:281, 1995. 201. Marx, A., Kirchner, T., Greiner, A., et al.: Neurofilament expression in thymic epithelial tumors and anti-axonal autoantibodies in myasthenia gravis: a model for autoimmunity by abnormal T cell selection [in German]. Verh. Dtsch. Ges. Pathol. 76:256, 1992. 202. Marx, A., Wilisch, A., Schultz, A., et al.: Expression of neurofilaments and of a titin epitope in thymic epithelial tumors: implications for the pathogenesis of myasthenia gravis. Am. J. Pathol. 148:1839, 1996. 203. Mason, D.: A very high level of crossreactivity is an essential feature of the T-cell receptor. Immunol. Today 19:395, 1998. 204. Mason, W. P., Verschuuren, J., Graus, F., et al.: Anti-Hu antibodies in patients with small-cell lung cancer but no paraneoplastic disorder. Ann. Neurol. 38:341, 1995. 205. McKhann, G. M., Cornblath, D. R., Griffin, J. W., et al.: Acute motor axonal neuropathy: a frequent cause of acute flaccid motor paralysis in China. Ann. Neurol. 33:333, 1993. 206. Meléndez-Vásquez, C., and Gregson, N. A.: Characterization and partial purification of a novel 36 kDa peripheral myelin protein recognized by the sera of patients with neurological disorders. J Neuroimmunol 91:10, 1998. 207. Meléndez-Vásquez, C., Redford, J., Choudhary, P. P., et al.: Immunological investigation of chronic inflammatory demyelinating polyradiculoneuropathy. J. Neuroimmunol. 73:124, 1997. 208. Milner, P., Lovelidge, C. A., Taylor, W. A., and Hughes, R. A. C.: P0 myelin protein produces experimental allergic neuritis in Lewis rats. J. Neurol. Sci. 79:275, 1987. 209. Milner, R. J., Lai, C., Nave, K.-A., et al.: Organisation of myelin protein genes: myelin-associated glycoprotein. In Duncan, I., Skoff, R. P., and Colman, D. (eds.): Myelination and Demyelination. New York, New York Academy of Sciences, p. 254, 1990. 210. Miyazaki, T., Kusunoki, S., Kaida, K., et al.: Guillain-Barre syndrome associated with IgG monospecific to ganglioside GD1b. Neurology 56:1227, 2001. 211. Molander, M., Berthold, C.-H., Persson, H., et al.: Monosialoganglioside (GM1) immunofluorescence in rat spinal roots studied with a monoclonal antibody. J. Neurocytol. 26:101, 1997. 212. Moody, D. B., Besra, G. S., Wilson, I. A., and Porcelli, S. A.: The molecular basis of CD1-mediated presentation of lipid antigens. Immunol. Rev. 172:285, 1999. 213. Mori, M., Kuwabara, S., Miyake, M., et al.: Haemophilus influenzae infection and Guillain-Barre syndrome. Brain 123:2171, 2000. 214. Muller, H. W.: Tetraspan myelin protein PMP22 and demyelinating peripheral neuropathies: new facts and hypotheses. Glia 29:182, 2000. 215. Muresu, R., Baldini, A., Gress, T., et al.: Mapping of the gene coding for a paraneoplastic encephalomyelitis antigen (HuD) to human chromosome site 1p34. Cytogenet. Cell Genet. 65:177, 1994. 216. Myszewski, M. A., and Stern, N. J.: Phagocytosis and intracellular killing of Campylobacter jejuni by elicited chicken peritoneal macrophages. Avian Dis. 35:750, 1991.

217. Nachamkin, I., Engberg, J., Gutacker, M., et al.: Molecular population genetic analysis of Campylobacter jejuni HS:19 associated with Guillain-Barre syndrome and gastroenteritis. J. Infect. Dis. 184:221, 2001. 218. Naidenko, O. V., Koezuka, Y., and Kronenberg, M.: CD1mediated antigen presentation of glycosphingolipids. Microbes Infect. 2:621, 2000. 219. Nardelli, E., Bassi, A., Mazzi, G., et al.: Systemic passive transfer studies using IgM monoclonal antibodies to sulfatide. J. Neuroimmunol. 63:29, 1995. 220. Nemni, R., Fazio, R., Quattrini, A., et al.: Antibodies to sulfatide and to chondroitin sulfate C in patients with chronic sensory neuropathy. J. Neuroimmunol. 43:79, 1993. 221. Neumann, H., Cavalie, A., Jenne, D. E., and Wekerle, H.: Induction of MHC class I genes in neurons. Science 269:549, 1995. 222. Nishimura, M., Nukina, M., Kuroki, S., et al.: Characterization of Campylobacter jejuni isolates from patients with GuillainBarré syndrome. J. Neurol. Sci. 153:91, 1997. 223. Nishimura, M., Saida, T., Kuroki, S., et al.: Post-infectious encephalitis with anti-galactocerebroside antibody subsequent to Mycoplasma pneumoniae infection. J. Neurol. Sci. 140:91, 1996. 224. Nobile-Orazio, E., Manfredini, E., Carpo, M., et al.: Frequency and clinical correlates of anti-neural IgM antibodies in neuropathy associated with IgM monoclonal gammopathy. Ann. Neurol. 36:416, 1994. 225. Nobile-Orazio, E., Manfredini, E., Sgarzi, M., et al.: Serum IgG antibodies to a 35-kDa P0-related glycoprotein in motor neuron disease. J. Neuroimmunol. 53:143, 1994. 226. Notterpek, L., Roux, K. J., Amici, S. A., et al.: Peripheral myelin protein 22 is a constituent of intercellular junctions in epithelia. Proc. Natl. Acad. Sci. U. S. A. 98:14404, 2001. 227. Obrocki, J., and Borroni, E.: Immunocytochemical evaluation of a cholinergic-specific ganglioside antigen (Chol-1) in the central nervous system of the rat. Exp. Brain Res. 72:71, 1988. 228. Ogawa-Goto, K., and Abe, T.: Gangliosides and glycosphingolipids of peripheral nervous system myelins—a minireview. Neurochem. Res. 23:305, 1998. 229. Ogawa-Goto, K., Funamoto, N., Ohta, Y., et al.: Myelin gangliosides of human peripheral nervous system: an enrichment of GM1 in the motor nerve myelin isolated from cauda equina. J. Neurochem. 59:1844, 1992. 230. O’Hanlon, G. M., Paterson, G. J., Veitch, J., et al.: Mapping immunoreactive epitopes in the human peripheral nervous system using human monoclonal anti-GM1 ganglioside antibodies. Acta Neuropathol. 95:605, 1998. 231. O’Hanlon, G. M., Paterson, G. J., Wilson, G., et al.: AntiGM1 ganglioside antibodies cloned from autoimmune neuropathy patients show diverse binding patterns in the rodent nervous system. J. Neuropathol. Exp. Neurol. 55:184, 1996. 232. O’Hanlon, G. M., Plomp, J. J., Chakrabarti, M., et al.: AntiGQ1b ganglioside antibodies mediate complementdependent destruction of the motor nerve terminal. Brain 124:893, 2001. 233. O’Hanlon, G. M., Veitch, J., Gallardo, E., et al.: Peripheral neuropathy associated with anti-GM2 ganglioside antibodies:

Peripheral Nerve Antigens

234.

235.

236.

237.

238.

239.

240.

241.

242.

243.

244.

245.

246.

247.

248.

249.

250.

clinical and immunopathological studies. Autoimmunity 32:133, 2000. Ohsawa, T., Miyatake, T., and Yuki, N.: Anti-B-series ganglioside-recognizing autoantibodies in an acute sensory neuropathy patient cause cell death of rat dorsal root ganglion neurons. Neurosci. Lett. 157:167, 1993. Oldenborg, P. A., Zheleznyak, A., Fang, Y. F., et al.: Role of CD47 as a marker of self on red blood cells. Science 288:2051, 2000. O’Leary, C. P., and Willison, H. J.: Autoimmune ataxic neuropathies (sensory ganglionopathies). Curr. Opin. Neurol. 10:366, 1997. O’Leary, C.P., and Willison, H. J.: The role of antiglycolipid antibodies in peripheral neuropathies. Curr. Opin. Neurol. 13:583, 2000. Olee, T., Powell H. C., and Brostoff S. W.: New minimum length requirement for a T cell epitope for experimental allergic neuritis. J. Neuroimmunol. 27:187, 1990. Olee, T., Weise, M., Powers, J., and Brostoff, S. W.: A T cell epitope for experimental allergic neuritis is an amphipathic alpha-helical structure. J. Neuroimmunol. 21:235, 1989. Pan, C. L., Yuki, N., Koga, M., et al.: Acute sensory ataxic neuropathy associated with monospecific anti-GD1b IgG antibody. Neurology 57:1316, 2001. Pancorbo, P.L., de Pablo, M.A., Ortega, E., et al.: Potential intervention of Campylobacter jejuni in the modulation of murine immune response. Curr. Microbiol 43:209, 2001. Pancorbo, P. L., Gallego, A. M., de Pablo, M., et al.: Inflammatory and phagocytic response to experimental Campylobacter jejuni infection in mice. Microbiol. Immunol. 38:89, 1994. Paparounas, K., O’Hanlon, G. M., O’Leary, C. P., et al.: Anti-ganglioside antibodies can bind peripheral nerve nodes of Ranvier and activate the complement cascade without inducing acute conduction block in vitro. Brain 122:807, 1999. Paul, J. A., and Gregson, N. A.: An immunohistochemical study of phospholipase A2 in peripheral nerve during Wallerian degeneration. J. Neuroimmunol. 39:31, 1992. Perego, L., Previtali, S. C., Nemni, R., et al.: Autoantibodies to amphiphysin I and amphiphysin II in a patient with sensory-motor neuropathy. Eur. Neurol. 47:196, 2002. Pestronk, A., Choksi, R., Yee, W. C., et al.: Serum antibodies to heparan sulfate glycosaminoglycans in Guillain-Barre syndrome and other demyelinating polyneuropathies. J. Neuroimmunol. 91:204, 1998. Pestronk, A., Li, F., Griffin, J., et al.: Polyneuropathy syndromes associated with serum antibodies to sulfatide and myelin-associated glycoprotein. Neurology 41:357, 1991. Peterson, K., Rosenblum, M. K., Kotanides, H., and Posner, J. B.: Paraneoplastic cerebellar degeneration. I. A clinical analysis of 55 anti-Yo antibody-positive patients. Neurology 42:1931, 1992. Plomp, J. J., Molenaar, P. C., O’Hanlon, G. M., et al.: Miller Fisher anti-GQ1b antibodies: -latrotoxin-like effects on motor end plates. Ann. Neurol. 45:189, 1999. Pollard, J. D., Westland, K. W., Harvey, G. K., et al.: Activated T cells of nonneural specificity open the bloodnerve barrier to circulating antibody. Ann. Neurol. 37:467, 1995.

605

251. Quach, T. T., Rong, Y., Belin, M. F., et al.: Molecular cloning of a new unc-33-like cDNA from rat brain and its relation to paraneoplastic neurological syndromes. Brain Res. Mol. Brain Res. 46:329, 1997. 252. Quarles, R. H., Ilyas, A. A., and Willison, H. J.: Antibodies to glycolipids in demyelinating diseases of the human peripheral nervous system. Chem. Phys. Lipids 42:235, 1986. 253. Quarles, R. H., and Weiss, M. D.: Autoantibodies associated with peripheral neuropathy. Muscle Nerve 22:800, 1999. 254. Quattrini, A., Corbo, M., Dhaliwal, S. K., et al.: Antisulfatide antibodies in neurological disease: binding to rat dorsal root ganglia neurons. J. Neurol. Sci. 112:152, 1992. 255. Raschetti, R., Maggini, M., Popoli, P., et al.: Gangliosides and Guillain-Barre syndrome. J. Clin. Epidemiol. 48:1399, 1995. 256. Rees, J. H., Gregson, N. A., Griffiths, P. L., and Hughes, R. A. C.: Campylobacter jejuni and Guillain-Barre syndrome. Q. J. Med. 86:623, 1993. 257. Rees, J. H., Gregson, N. A., and Hughes, R. A. C.: Antiganglioside GM1 antibodies in Guillain-Barré syndrome and their relationship to Campylobacter jejuni infection. Ann. Neurol. 38:809, 1995. 258. Rees, J. H., Soudain, S. A., Gregson, N. A., and Hughes, R. A. C.: Campylobacter jejuni infection and Guillain-Barré syndrome. N. Engl. J. Med. 333:1374, 1995. 259. Richardson, P. J., Walker, J. H., Jones, R. T., and Whittaker, V. P.: Identification of a cholinergic-specific antigen Chol-1 as a ganglioside. J. Neurochem. 38:1605, 1982. 260. Ritter, G., Fortunato, S. R., Cohen, L., et al.: Induction of antibodies reactive with GM2 ganglioside after immunization with lipopolysaccharides from Campylobacter jejuni. Int. J. Cancer 66:184, 1996. 261. Ritz, J.: The role of NK cells in immune surveillance. N. Engl. J. Med. 320:1748, 1989. 262. Ritz, M. F., Lechner-Scott, J., Scott, R. J., et al.: Characterisation of autoantibodies to peripheral myelin protein 22 in patients with hereditary and acquired neuropathies. J. Neuroimmunol. 104:155, 2000. 263. Sadiq, S. A., Van den Berg, L. H., Thomas, F. P., et al.: Human monoclonal antineurofilament antibody crossreacts with a neuronal surface protein. J. Neurosci. Res. 29:319, 1991. 264. Saida, K., Saida, T., and Brown, M. J.: In vitro demyelination induced by intraneural injection of antigalactocerebroside serum: a morphological study. Am. J. Pathol. 95:99, 1979. 265. Saida, T., Saida, K., and Dorfman, S. H.: Experimental allergic neuritis induced by sensitisation with galactocerebroside. Science 204:1103, 1979. 266. Saida, T., Saida, K., and Silberberg, D. H.: Demyelination produced by experimental allergic neuritis serum and antigalactocerebroside antiserum in central nervous system cultures: an ultrastructural study. Acta Neuropathol. 48:19, 1979. 267. Saiz, A., Dalmau, J., Butler, M. H., et al.: Anti-amphiphysin I antibodies in patients with paraneoplastic neurological disorders associated with small cell lung carcinoma. J. Neurol. Neurosurg. Psychiatry 66:214, 1999.

606

Neuroimmunology of the Peripheral Nervous System

268. Sakai, K., Gofuku, M., Kitagawa, Y., et al.: A hippocampal protein associated with paraneoplastic neurologic syndrome and small cell lung carcinoma. Biochem. Biophys. Res. Commun. 199:1200, 1994. 269. Sakamoto, Y., Kitamura, K., Yoshimura, K., et al.: Complete amino acid sequence of P0 protein in bovine peripheral nerve myelin. J. Biol. Chem. 262:4208, 1987. 270. Salih, A. M., Nixon, N. B., Dawes, P. T., and Mattey, D. L.: Soluble adhesion molecules and anti-endothelial cell antibodies in patients with rheumatoid arthritis complicated by peripheral neuropathy. J. Rheumatol. 26:551, 1999. 271. Salloway, S., Mermel, L. A., Seamans, M., et al.: MillerFisher syndrome associated with Campylobacter jejuni bearing lipopolysaccharide molecules that mimic human ganglioside GD3. Infect. Immun. 64:2945, 1996. 272. Schachner, M., and Bartsch, U.: Multiple functions of the myelin-associated glycoprotein MAG (siglec-4a) in formation and maintenance of myelin. Glia 29:154, 2000. 273. Scherer, S. S.: Molecular specializations at nodes and paranodes in peripheral nerve. Microsc. Res. Tech. 34:452, 1996. 274. Scherer, S. S., and Arroyo, E. J.: Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Syst. 7:1, 2002. 275. Schluep, M., and Steck, A. J.: Immunostaining of motor nerve terminals by IgM M protein with activity against gangliosides GM1 and GD1b from a patient with motor neuron disease. Neurology 38:1890, 1988. 276. Schmid, C. D., Stienekemeier, M., Oehen, S., et al.: Immune deficiency in mouse models for inherited peripheral neuropathies leads to improved myelin maintenance. J. Neurosci. 20:729, 2000. 277. Schultz, A., Hoffacker, V., Wilisch, A., et al.: Neurofilament is an autoantigenic determinant in myasthenia gravis. Ann. Neurol. 46:167, 1999. 278. Sedzik, J., Kotake, Y., and Uyemura, K.: Purification of PASII/PMP22—an extremely hydrophobic glycoprotein of PNS myelin membrane. Neuroreport 9:1595, 1998. 279. Sekido, Y., Bader, S. A., Carbone, D. P., et al.: Molecular analysis of the HuD gene encoding a paraneoplastic encephalomyelitis antigen in human lung cancer. Cancer Res. 54:4988, 1994. 280. Shamshiev, A., Donda, A., Carena, I., et al.: Self glycolipids as T-cell autoantigens. Eur. J. Immunol. 29:1667, 1999. 281. Shapiro, L., Doyle, J. P., Hensley, P., et al.: Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17:435, 1996. 282. Sheikh, K. A., Deerinck, T. J., Ellisman, M. H., and Griffin, J. W.: The distribution of ganglioside-like moieties in peripheral nerves. Brain 122:449, 1999. 283. Sheikh, K. A., Gong, Y., Schnaar, R. L., and Griffin, J. W.: Anti-ganglioside antibody mediated axonal degeneration. J. Peripher. Nerv. Syst. 6:175, 2001. 284. Sheikh, K. A., Nachamkin, I., Ho, T. W., et al.: Campylobacter jejuni lipopolysaccharides in Guillain-Barre syndrome: molecular mimicry and host susceptibility. Neurology 51:371, 1998. 285. Sherman, D. L., Fabrizi, C., Gillespie, C. S., and Brophy, P. J.: Specific disruption of a Schwann cell dystrophin-related protein complex in a demyelinating neuropathy [see comment]. Neuron 30:677, 2001.

286. Sherman, W. H., Latov, N., Hays, A. P., et al.: Monoclonal IgMk antibody precipitating with chondroitin sulfate C from patients with axonal polyneuropathy and epidermolysis. Neurology 33:192, 1983. 287. Shillito, P., Molenaar, P. C., Vincent, A., et al.: Acquired neuromyotonia: evidence for autoantibodies directed against K channels of peripheral nerves. Ann. Neurol. 38:714, 1995. 288. Silber, E., Semra, Y. K., Gregson, N. A., and Sharief, M. K.: Patients with progressive multiple sclerosis have elevated antibodies to neurofilament subunit. Neurology 58:1372, 2002. 289. Sillevis Smitt, P. A. E., Manley, G. T., and Posner, J. B.: Immunization with the paraneoplastic encephalomyelitis antigen HuD does not cause neurologic disease in mice. Neurology 45:1873, 1995. 290. Simons, K., and Ikonen, E.: Functional rafts in cell membranes. Nature 387:569, 1997. 291. Smith, V. V., Gregson, N., Foggensteiner, L., et al.: Acquired intestinal aganglionosis and circulating autoantibodies without neoplasia or other neural involvement. Gastroenterology 112:1366, 1997. 292. Snipes, G. J., Suter, U., and Shooter, E. M.: Human peripheral myelin protein-22 carries the L2/HNK-1 carbohydrate adhesion epitope. J. Neurochem. 61:1961, 1993. 293. Snipes, G. J., Suter, U., Welcher, A. A., and Shooter, E. M.: Characterization of a novel peripheral nervous system myelin protein (PMP-22 /SR13). J. Cell Biol. 117:225, 1992. 294. Sorensen, T. I., Nielsen, G. G., Andersen, P. K., and Teasdale, T. W.: Genetic and environmental influences on premature death in adult adoptees. N. Engl. J. Med. 318:727, 1988. 295. Spreyer, P., Kuhn, G., Hanemann, C. O., et al.: Axonregulated expression of a Schwann cell transcript that is homologous to a “growth arrest-specific” gene. EMBO J. 10:3661, 1991. 296. Stebbins, E. C., and Galán, J. E.: Structural mimicry in bacterial virulence. Nature 412:701, 2001. 297. Stefansson, K., Marton, L. S., Dieperink, M. E., et al.: Circulating autoantibodies to the 200,000-dalton protein of neurofilaments in the serum of healthy individuals. Science 228:1117, 1985. 298. Stoffel, W., and Bosio, A.: Myelin glycolipids and their functions. Curr. Opin. Neurobiol. 7:654, 1997. 299. Stoll, G., Schwendemann, G., Heininger, K., et al.: Relation of clinical, serological, morphological and electrophysiological findings in galactocerebroside induced experimental allergic neuritis. J. Neurol. Neurosurg. Psychiatry 49:258, 1986. 300. Sumner, A.: Electrophysiological and morphological effects of the injection of Guillain-Barré sera in the sciatic nerve of the rat. Rev. Neurol. (Paris) 138:17, 1982. 301. Sumner, A. J., Saida, K., and Saida, T.: Acute conduction block associated with experimental antiserum-mediated demyelination of peripheral nerve. Ann. Neurol. 11:469, 1982. 302. Svennerholm, L.: Designation and schematic structure of gangliosides and allied glycosphingolipids. Proc. Brain Res. 101:xi, 1994. 303. Szabo, A., Dalmau, J., Manley, G., et al.: HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding

Peripheral Nerve Antigens

304.

305.

306.

307.

308.

309.

310.

311. 312.

313.

314.

315.

316.

317.

318.

319.

320.

domains and is homologous to Elav and Sex-lethal. Cell 67:325, 1991. Takamiya, K., Yamamoto, A., Furukawa, K., et al.: Mice with disrupted GM2/GD2 synthase gene lack complex gangliosides but exhibit only subtle defects in their nervous system. Proc. Natl. Acad. Sci. U. S. A. 93:10662, 1996. Takashima, H., Boerkoel, C. F., De Jonghe P., et al.: Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51:709, 2002. Tanaka, K., Tanaka M., Inuzuka, T., et al.: Cytotoxic T lymphocyte-mediated cell death in paraneoplastic sensory neuronopathy with anti-Hu antibody. J. Neurol. Sci. 163:159, 1999. Tanaka, M., Tanaka, K., Tsuji, S., et al.: Cytotoxic T cell activity against the peptide, AYRARALEL, from Yo protein of patients with the HLA A24 or B27 supertype and paraneoplastic cerebellar degeneration. J. Neurol. Sci. 188:61, 2001. Tatum, A. H.: Experimental paraprotein neuropathy: demyelination by passive transfer of human IgM antimyelin-associated glycoprotein. Ann. Neurol. 33:502, 1993. Taylor, V., Zgraggen, C., Naef, R., and Suter, U.: Membrane topology of peripheral myelin protein 22. J. Neurosci. Res. 62:15, 2000. Terryberry, J., Thor, G., and Peter, J. B.: Autoantibodies in neurodegenerative diseases: antigen-specific frequencies and intrathecal synthesis. Neurobiol. Aging 19:205, 1998. Tettamanti, G., and Riboni, L.: Gangliosides and modulation of the function of neural cells. Adv. Lipid Res. 25:235, 1993. Thomas, F. P., Adapon, P. H., Goldberg, G. P., et al.: Localization of neural epitopes that bind to IgM monoclonal autoantibodies (M-proteins) from two patients with motor neuron disease. J. Neuroimmunol. 21:31, 1989. Thomas, F. P., Lee, A. M., Romas, S. N., and Latov, N.: Monoclonal IgMs with anti-Gal(1–3)GalNAc activity in lower motor neuron disease: identification of glycoprotein antigens in neural tissue and cross-reactivity with serum immunoglobulins. J. Neuroimmunol. 23:167, 1989. Trapp, B. D.: Distribution of the myelin-associated glycoprotein and P0 protein during myelin compaction in quaking mouse peripheral nerve. J. Cell Biol. 107:675, 1988. Trapp, B. D., Andrews, S. B., Wong, A., et al.: Co-localization of the myelin-associated glycoprotein and the microfilament components, F-actin and spectrin, in Schwann cells of myelinated nerve fibers. J. Neurocytol. 18:47, 1989. Triantafilou, M., and Triantafilou, K.: Lipopolysaccharide recognition: CD14, TLRs and the LPS activation cluster. Trends Immunol. 23:301, 2002. Trojaborg, W., Hays, A. P., van den Berg, L., et al.: Motor conduction parameters in neuropathies associated with anti-MAG antibodies and other types of demyelinating and axonal neuropathies. Muscle Nerve 18:730, 1995. Tsiper, M. V., and Yurchenco, P. D.: Laminin assembles into separate basement membrane and fibrillar matrices in Schwann cells. J. Cell Sci. 115:1005, 2002. Underhill, D. M., and Ozinsky, A.: Toll-like receptors: key mediators of microbe detection. Curr. Opin. Immunol. 14:103, 2002. Vallat, J. M., Leboutet, M. J., Jauberteau, M. O., et al.: Widenings of the myelin lamellae in a typical Guillain-Barré syndrome. Muscle Nerve 17:378, 1994.

607

321. van Belkum, A., Van Den Braak, N., Godschalk, P., et al.: A Campylobacter jejuni gene associated with immunemediated neuropathy. Nat. Med. 7:752, 2001. 322. van der Meché, F. G. A., Visser, L. H., Jacobs, B. C., et al.: Guillain-Barre syndrome: multifactorial mechanisms versus defined subgroups. J. Infect. Dis. 176(Suppl. 2):S99, 1997. 323. van Koningsveld, R., Rico, R., Gerstenbluth, I., et al.: Gastroenteritis-associated Guillain-Barre syndrome on the Caribbean island Curacao. Neurology 56:1467, 2001. 324. van Koningsveld, R., van Doorn, P. A., Schmitz, P. I., et al.: Mild forms of Guillain-Barre syndrome in an epidemiologic survey in The Netherlands. Neurology 54:620, 2000. 325. van Muijen, G. N., Ruiter, D. J., van Leeuwen, C., et al.: Cytokeratin and neurofilament in lung carcinomas. Am. J. Pathol. 116:363, 1984. 326. Van Rhijn, I., Van den Berg, L. H., Bosboom, W. M., et al.: Expression of accessory molecules for T-cell activation in peripheral nerve of patients with CIDP and vasculitic neuropathy. Brain 123:2020, 2000. 327. Vimr, E., Lichtensteiger, C., and Steenbergen, S.: Sialic acid metabolism’s dual function in Haemophilus influenzae. Mol. Microbiol. 36:1113, 2000. 328. Visser, L. H., van der Meché, F. G. A., Meulstee, J., et al.: Cytomegalovirus infection and Guillain-Barre syndrome: the clinical, electrophysiologic, and prognostic features. Dutch Guillain-Barre Study Group [see comments]. Neurology 47:668, 1996. 329. Vriesendorp, F. J.: Insights into Campylobacter jejuniinduced Guillain-Barre syndrome from the Lewis rat model of experimental allergic neuritis. J. Infect. Dis. 176:S164, 1997. 330. Vyas, A. A., Patel, H. V., Fromholt, S. E., et al.: Gangliosides are functional nerve cell ligands for myelin-associated glycoprotein (MAG), an inhibitor of nerve regeneration [see comments.]. Proc. Natl. Acad. Sci. U. S. A. 99:8412, 2002. 331. Walker, R. I., Caldwell, M. B., Lee, E. C., et al.: Pathophysiology of Campylobacter enteritis. Microbiol. Rev. 50:81, 1986. 332. Walsh, F. S., Cronin, M., Koblar, S., et al.: Association between glycoconjugate antibodies and Campylobacter infection in patients with Guillain-Barre syndrome. J. Neuroimmunol. 34:43, 1991. 333. Wan, X. C. S., Trojanowski, J. Q., and Gonatas, J. O.: Cholera toxin and wheat germ agglutinin conjugates as neuroanatomical probes: their uptake and clearance, transganglionic and retrograde transport and sensitivity. Brain Res. 243:215, 1982. 334. Wang, X., and Tanaka Hall, T. M.: Structural basis for recognition of AU-rich element RNA by the HuD protein. Nat. Struct. Biol. 8:141, 2001. 335. Wassenaar, T. M.: Toxin production by Campylobacter spp. Clin. Microbiol. Rev. 10:466, 1997. 336. Wassenaar, T. M., Engelskirchen, M., Park, S., and Lastovica, A.: Differential uptake and killing potential of Campylobacter jejuni by human peripheral monocytes/macrophages. Med. Microbiol. Immunol. (Berl.) 186:139, 1997. 337. Weerth, S., Berger, T., Lassman, H., and Linington, C.: Encephalitogenic and neuritogenic T cell responses to the

608

338.

339. 340.

341.

342. 343.

344.

345.

346.

347.

348.

349.

350.

351. 352.

353.

354.

Neuroimmunology of the Peripheral Nervous System myelin-associated glycoprotein (MAG) in the Lewis rat. J. Neuroimmunol. 95:157, 1999. Weishaupt, A., Gold, R., Gaupp, S., et al.: Antigen therapy eliminates T cell inflammation by apoptosis: effective treatment of experimental autoimmune neuritis with recombinant myelin protein P2. Proc. Natl. Acad. Sci. U. S. A. 94:2164, 1997. Whittaker, V. P., and Kelic, S.: Cholinergic-specific glycoconjugates. Neurochem. Res. 20:1377, 1995. Wiethölter, H., Hulser, P.-J., Linington, C., et al.: Electrophysiological follow-up of experimental allergic neuritis mediated by a permanent T cell line in rats. J. Neurol. Sci. 83:1, 1988. Wigge, P., and McMahon, H. T.: The amphiphysin family of proteins and their role in endocytosis at the synapse. Trends Neurosci. 21:339, 1998. Wilkinson, P. C.: Serological findings in carcinomatous neuropathy. Lancet 1:1301, 1964. Willison, H. J., Almemar, A., Veitch, J., and Thrush, D.: Acute ataxic neuropathy with cross-reactive antibodies to GD1b and GD3 gangliosides. Neurology 44:2395, 1994. Willison, H. J., and Kennedy, P. G. E.: Gangliosides and bacterial toxins in Guillain-Barre-syndrome. J. Neuroimmunol. 46:105, 1993. Willison, H. J., and O’Hanlon, G. M.: The immunopathogenesis of Miller Fisher syndrome. J. Neuroimmunol. 100:3, 1999. Willison, H. J., and O’Hanlon, G. M.: Anti-glycosphingolipid antibodies and Guillain-Barre syndrome. In Nachamkin, I., and Blaser, M. J. (eds.): Campylobacter. Washington, DC, ASM Press, p. 259, 2000. Willison, H. J., O’Hanlon, G. M., Paterson, G. J., et al.: A somatically mutated human antiganglioside IgM antibody that induces experimental neuropathy in mice is encoded by the variable region heavy chain gene, V1–18. J. Clin. Invest. 97:1155, 1996. Willison, H. J., O’Leary, C. P., Veitch, J., et al.: The clinical and laboratory features of chronic sensory ataxic neuropathy with anti-disialosyl IgM antibodies. Brain 124:1968, 2001. Willison, H. J., Trapp, B. D., Bacher, J. D., et al.: Demyelination induced by intraneural injection of human anti MAG antibodies. Muscle Nerve 11:1169, 1988. Willison, H. J., and Veitch, J.: Immunoglobulin subclass distribution and binding characteristics of anti-GQ1b antibodies in Miller Fisher syndrome. J. Neuroimmunol. 50:159, 1994. Willison, H. J., and Yuki, N.: Peripheral neuropathies and anti-glycolipid antibodies. Brain 125:1, 2002. Winer, J. B., Hughes, R. A. C., Anderson, M. J., et al.: A prospective study of acute idiopathic neuropathy. II: Antecedent events. J. Neurol. Neurosurg. Psychiatry 51:613, 1988. Yako, K., Kusunoki, S., and Kanazawa, I.: Serum antibody against a peripheral nerve myelin ganglioside, LM1, in Guillain-Barre syndrome. J. Neurol. Sci. 168:85, 1999. Yamaji, T., Teranishi, T., Alphey, M. S., et al.: A small region of the natural killer cell receptor, Siglec-7, is responsible for

355.

356.

357.

358.

359.

360. 361.

362. 363.

364.

365.

366.

367.

368.

369.

its preferred binding to alpha 2,8-disialyl and branched alpha 2,6-sialyl residues: a comparison with Siglec-9. J. Biol. Chem. 277:6324, 2002. Yamawaki, M., Ariga, T., Bigbee, J. W., et al.: Generation and characterization of anti-sulfoglucuronosyl paragloboside monoclonal antibody NGR50 and its immunoreactivity with peripheral nerve. J. Neurosci. Res. 44:586, 1996. Yan, W. X., Archelos, J. J., Hartung, H. P., and Pollard, J. D.: P0 protein is a target antigen in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:286, 2001. Yinchang, Y., Crawford, T. O., Griffin, J. W., et al.: Myelinassociated glycoprotein is a myelin signal that modulates the caliber of myelinated axons. J. Neurosci. 18:1953, 1998. Yu, R. K., and Ariga, T.: The role of glycosphingolipids in neurological disorders—mechanisms of immune action. Ann. N. Y. Acad. Sci. 845:285, 1998. Yu, Z., Kryzer, T. J., Griesmann, G. E., et al.: CRMP-5 neuronal autoantibody: marker of lung cancer and thymoma-related autoimmunity. Ann. Neurol. 49:146, 2001. Yuki, N.: Anti-ganglioside antibody and neuropathy: review of our research. J. Peripher. Nerv. Syst. 3:3, 1998. Yuki, N., Ho, T. W., Tagawa, Y., et al.: Autoantibodies to GM1b and GalNAc-GD1a: relationship to Campylobacter jejuni infection and acute motor axonal neuropathy in China. J. Neurol. Sci. 164:134, 1999. Yuki, N., and Tagawa, Y.: Acute cytomegalovirus infection and IgM anti-GM2 antibody. J. Neurol. Sci. 154:14, 1998. Yuki, N., Taki, T., and Handa, S.: Antibody to GalNAc-GD1a and GalNAc-GMlb in Guillain-Barre syndrome subsequent to Campylobacter jejuni enteritis. J. Neuroimmunol. 71:155, 1996. Yuki, N., Taki, T., Inagaki, F., et al.: A bacterium lipopolysaccharide that elicits Guillain-Barre syndrome has a GM1 ganglioside-like structure. J. Exp. Med. 178:1771, 1993. Yuki, N., Yamada, M., Koga, M., et al.: Animal model of axonal Guillain-Barre syndrome induced by sensitization with GM1 ganglioside [see comments]. Ann. Neurol. 49:712, 2001. Yuki, N., Yamada, M., Tagaawa, Y., and Takahashi, H.: Pathogenesis of the neurotoxicity caused by anti-GD2 antibody therapy. J. Neurol. Sci. 149:127, 1997. Yuki, N., Yoshino, H., Sato, S., and Miyatake, T.: Acute axonal polyneuropathy associated with anti-GM1 antibodies following Campylobacter enteritis [see comments]. Neurology 40:1900, 1990. Zhu, J., Pelidou, S. H., Deretzi, G., et al.: P0 glycoprotein peptides 56–71 and 180–199 dose-dependently induce acute and chronic experimental autoimmune neuritis in Lewis rats associated with epitope spreading. J. Neuroimmunol. 114:99, 2001. Zou, L. P., Ma, D.-H., Levi, M., et al.: Antigen-specific immunosuppression: nasal tolerance to P0 protein peptides for the prevention and treatment of experimental autoimmune neuritis in Lewis rats. J. Neuroimmunol. 94:109, 1999.

27 Experimental Autoimmune Neuritis RALF GOLD, GUIDO STOLL, BERND C. KIESEIER, HANS-PETER HARTUNG, AND KLAUS V. TOYKA

Immunologic Principles Categories of the Immune Response Tolerance and Autoimmunity The Local Immune Circuit The EAN Models Actively Induced EAN Adoptive Transfer EAN

Chronic (Relapsing) EAN: Mode of Immunization and Species-Specific Aspects Experimental Treatments in EAN Induction Phase: The Role of Antigen Presentation and Co-stimulation Transmigration and Early Effector Phase: Adhesion Molecules and MMPs

IMMUNOLOGIC PRINCIPLES Categories of the Immune Response The immune system is a multifaceted system of cells and molecules with specialized tasks in defending the organism from external agents. In addition, it plays a central role in maintaining antigenic homeostasis in the body. Two types of responses to invading organisms can occur: an acute response initiated within hours, mediated by the so-called innate immune system, and a delayed response occurring within days delivered by the adaptive or acquired immune system. The main distinctions between these two systems relate to the mechanisms and receptors used for immune recognition and to the maturational processing induced by somatic mutations of highly specific antigen receptors. Adaptive and innate immunity are functionally connected, allowing for intensive interactions (see Chapter 98).17,104 The Innate Immune System Innate immune responses consist of all the immune defense mechanisms that do not require antigen-specific immunologic memory. The innate immune system forms a fast first-line defense. The strategy of the innate immune response preferentially focuses on a few highly conserved structures present in a large variety of microorganisms. These structures are referred to as pathogen-associated molecular patterns, and the corresponding receptors of the

Effector Phase: T-Cell–Directed Treatments Effector Phase: Macrophage- and Antibody-Directed Treatments Termination of the Immune Response Survival Factors in the PNS Future Perspectives for Treatment

innate immune system are called pattern-recognition receptors.66 These are expressed on many effector cells of the innate immune system, most importantly on macrophages, dendritic cells, and B lymphocytes—collectively termed professional antigen-presenting cells (APCs). The total number of receptors involved in the innate immune response is thought to be on the order of hundreds, in contrast to the approximately 1015 somatically generated immunoglobulins and T-cell receptors (TCRs) of the adaptive (antigen-specific) immune response. Recent evidence shows that this recognition can mainly be attributed to the family of TOLL-like receptors (TLRs). Binding of pathogenassociated molecular patterns to TLRs induces the production of reactive oxygen and nitrogen intermediates and the proinflammatory cytokines, and upregulates expression of co-stimulatory molecules, subsequently initiating adaptive immunity.135 The Adaptive Immune System The adaptive immune response is based on two classes of highly specialized cells, T and B lymphocytes. Each of these cells usually expresses a single kind of a structurally unique receptor, resulting in a broad and extremely diverse repertoire of antigen recognition, usually with high affinity and avidity. Both B and T lymphocytes are derived from primordial stem cells in primary lymphoid tissues such as bone marrow 609

610

Neuroimmunology of the Peripheral Nervous System

and fetal liver. Their early phase of development is not dependent on the presence of any antigen, but once these cells express a mature antigen receptor, their further differentiation, survival, and programmed cell death (apoptosis) become antigen dependent.

Tolerance and Autoimmunity The random generation of a highly diverse repertoire of B- and T-cell receptors allows the adaptive immune system to recognize virtually any antigen. T cells recognizing autoantigens are termed self-reactive or autoreactive T cells, while B cells specific for autoantigens express autoantibodies on their surface. Tolerance is the process that eliminates (deletes) or downregulates such autoreactive cells. Consequently, a breakdown in this system can cause an autoimmune response or even autoimmune disease.73 B-Cell Tolerance The immune system has several mechanisms to filter autoreactive B lymphocytes out of the B-cell pool: (1) by clonal deletion of immature B cells in the bone marrow103; (2) by deletion of autoreactive B lymphocytes in the T-cell zones of secondary lymphoid organs, such as lymph nodes or the spleen110; (3) by induction of cellular anergy (“functional inactivation”); and (4) by a process called “receptor editing,” which changes the receptor specificity once an autoantigen has been encountered.103 At present it remains unclear to what extent these mechanisms are of relevance in maintaining tolerance and preventing autoimmune disorders in the human nervous system. T-Cell Tolerance The principle mechanism of T-cell tolerance is the deletion of self-reactive T lymphocytes during their development in the thymus.141 To delete all autoreactive cells, the presence of all autoantigens is required in the thymus. However, this is not the case.62,73,86 Some autoreactive T cells escape thymic education and enter the systemic immune compartment. In the mature immune system, several mechanisms are required to keep T cells in check and maintain peripheral tolerance, such as immunologic ignorance, peripheral deletion of specific T-cell clones, T-cell anergy, and suppression of T-cell activity. All these processes depend on the “transmission” of signals between T lymphocytes and adjacent APCs in addition to antigen recognition through a series of cell surface receptors that interact with their counterligands on the neighboring cell via cell-cell contact to provide co-stimulatory signals (see Experimental Treatments in EAN below). The site of cell-cell contact has also been termed the immunologic synapse. Breakdown of Tolerance If one of the regulatory mechanisms fails, a specific immune response is mounted against self-antigens. This

leads to expansion of autoreactive effector T cells and generation of autoantibodies through T-cell help, and may give rise to tissue damage, a scenario that Paul Ehrlich termed horror autotoxicus. Autoaggressive responses may initiate autoimmune disease. The autoimmune response may persist because the immune system is not able to remove the autoantigen from the body, in contrast to foreign antigens; even worse, new hitherto hidden autoantigens can be released to amplify the immune response and broaden its epitope specificity, a process termed epitope spreading.91 The effector mechanisms leading to tissue damage in autoimmune diseases are essentially the same as those operative in other inflammatory disorders. Various mechanisms are operative to promote or prevent tissue damage once an autoimmune reaction has been initiated. Recruitment of large numbers of monocytes/macrophages and T cells and the release of cytotoxic cytokines and chemokines would serve to augment the injurious tissue reaction. Autoantibodies are also active partners in tissue damage or dysfunction. Conversely, a suppressive/inhibitory local environment with only a few or not fully competent APCs and anti-inflammatory cytokines, inappropriate antibody concentrations, lack of complement activation, or merely anatomic barriers between autoreactive immune cells and target structures may all help to prevent autoimmune tissue destruction. Finally, induction of apoptosis in activated T cells may eliminate the autoreactive T cells before the full inflammatory cascade is set in motion.

The Local Immune Circuit The peripheral nervous system (PNS) has traditionally been considered as “immunologically privileged,” yet not as strictly as the central nervous system (CNS).152 This view has undergone revision within recent years, and now differences from other organs appear to be more quantitative than qualitative. The PNS is separated from the external environment by the blood-nerve barrier (BNB), which does restrict access of immune cells and soluble mediators to a certain degree; however, this restriction is not complete, either anatomically or functionally. The BNB is practically absent at nerve roots, in dorsal root ganglia, and at nerve terminals. Immune surveillance, as found in most organs, is present in the PNS as well: activated T and B lymphocytes can cross the BNB irrespective of their antigen specificity, and APCs, such as macrophages, can abundantly be detected in peripheral nerve tissue. Schwann cells also can function as “nonprofessional” APCs.153 Most of our present knowledge on the relevance of putative autoantigens and on the mechanisms involved in the pathogenesis of immune-mediated demyelination has been obtained in the animal model known as experimental autoimmune neuritis (EAN).

611

Experimental Autoimmune Neuritis

THE EAN MODELS Actively Induced EAN Mode of Immunization and Antigens Two decades after the description of experimental autoimmune encephalomyelitis (EAE), the first successful attempt to induce EAN was undertaken by Byron Waksman, who immunized rabbits with homogenized PNS tissue in adjuvant.145 Since then, investigators have studied EAN in various species using a panel of PNS autoantigens (see Chapter 26). As in EAE, many molecularly defined neuritogenic components of PNS myelin were identified, such as protein P2,71 the P2 peptides spanning the neuritogenic epitope (amino acids 61 to 70) of P2,106 protein P0,96 peripheral myelin protein 22 (PMP22),30 and myelin-associated glycoprotein (MAG).147 Susceptible animal species, including rabbit, monkey, rat, and mouse (Table 27–1), are inoculated subcutaneously with the neuritogen emulsified in incomplete Freund’s adjuvant

enriched with homogenized Mycobacterium tuberculosis (complete Freund’s adjuvant [CFA]). This mode of immunization was more specifically named actively induced EAN. The mycobacterial component of CFA with the included heat shock proteins may lead to severe arthritis,146 which is the basis of experimental arthritis models,155 but represents an unwanted side effect in EAN induction. Yet the use of CFA still represents the international standard in T-cell–mediated EAN and EAE models. Modern adjuvants such as Titermax (lacking mineral oil) are optimized mainly for B-cell responses and induce a weaker and more variable EAN (S. Jung and R. Gold, unpublished observations, 1998). In our experience in the rat, the following procedure minimizes unwanted side effects when using CFA: (1) preparation of the emulsion with 1 mg/mL or less final mycobacterial concentration, (2) subcutaneous injection of less than 50 ␮L of the immunogen at only one rat hind limb footpad, and (3) distribution of the remaining adjuvant at the tail base. In mouse models, injection at the footpad is avoided and the adjuvant is distributed along the flanks and the tail base. This strategy

Table 27–1. Animal Models of Immune-Mediated Neuropathies Animal Model Lewis Rat Active EAN

Adoptive transfer EAN

Antigen (First Description)

Comments

• PNS myelin, P2 (Kadlubowski and Hughes71) P0 (Linington et al.96) • ⫹ cyclosporine A (Pender et al.108) • PMP22 (Gabriel et al.30), MAG (Weerth et al.147) • P2 (Linington et al.95) P2 (aa 61–70) (Olee et al.106), P0 (aa 180–199) (Linington et al.96)

• Reliable model, commonly used • Chronic relapsing course, not robust • Only mild disease course • Homogeneous course, rapid onset, chronic relapsing after repeated T-cell transfer

Brown Norway rat

• P2 adoptive transfer (Linington et al.97)

Rabbits

• GalC (Saida et al.116) • Bovine brain ganglioside mixture, GM1 (Yuki et al.157)

• Chronic course • Axonal pathology

• P2 (Taylor and Hughes137) • P2 ⫹ PT ⫹ IL-12 (Calida15) • P0 aa 180–199 ⫹ PT ⫹ anti–CTLA-4 (Zhu et al.162)

• Mild disease • Severe course • Multimodal manipulation of immune system necessary

• MBP (18.5-, 21-KDa isoforms) (Abromsom-Leeman1)

• Peripheral and central demyelination

Knockout Mice Spontaneous inflammatory neuropathy

• NOD/B7-2 deficient (Salomon et al.117)

Superimposed neuropathy

• P0 ⫹/⫺ (Schmid120), CX32 ⫺/⫺ (Kobsar et al.83)

• Spontaneous chronic autoimmune peripheral neuropathy, mimics human CIDP • Macrophage mediated, secondary chronic immune response

Murine EAN Active SJL mice C57/B16 mice

Adoptive transfer Balb/C

aa ⫽ amino acids; CIDP ⫽ chronic inflammatory demyelinating neuropathy; CTLA-4 ⫽ cytotoxic T lymphocyte–associated antigen-4; CX32 ⫽ connexin 32; EAN ⫽ experimental autoimmune neuritis; GalC ⫽ galactocerebrosidase; IL-12 ⫽ interleukin-12; MAG ⫽ myelin-associated glycoprotein; MBP ⫽ myelin basic protein; PMP22 ⫽ peripheral myelin protein 22; PNS ⫽ peripheral nervous system; PT ⫽ pertussis toxin.

612

Neuroimmunology of the Peripheral Nervous System

guarantees a high efficacy of disease induction without producing concomitant severe arthritis. Ultimately these measures will also help to reduce the size of experimental groups. EAN in the Lewis rat is still the most commonly studied monophasic model because the disease course is predictable and relatively homogeneous. This allows for small group sizes of five to six rats in therapeutic or pathophysiologic studies. Typically, female rats at the age of 6 to 8 weeks are used for experimentation. Although the spectrum of available neuritogens has been expanded, immunization with bovine peripheral myelin (BPM) or the neuritogenic P2 peptide aa 53–7869 still gives the most reproducible results in the Lewis rat. Disease severity depends on the dose of the neuritogen inoculated, and severe disease may lead not only to demyelination, but also to axonal damage (Fig. 27–1). In this case, residual clinical signs such as gait ataxia may persist into remission. There are also differences in disease expression depending on the type and dosage of neuritogen. Whereas myelin immunization in the Lewis rat causes widespread perivascular demyelination resembling the pathology of Guillain-Barré syndrome (GBS) and chronic inflammatory demyelinating polyradiculoneuropathy (CIDP), with P2 peptide, wallerian degeneration indicative of axonal damage also is observed. With P0, and more so with PMP22 and MAG, the disease course is much milder at roughly equivalent molar doses. Clinical signs may even be lacking, and often inflammatory demyelination is only evident by histologic or electrophysiologic studies. In contrast to Lewis rats, the Brown Norway (BN) strain is resistant to active EAN. In principle EAN can also be studied in monkeys,26 yet primate models have been largely abandoned in EAN research. EAN has been induced in SJL mice137 and C57BL/6 mice,162 but the currently available models are characterized by rather mild disease in contrast to EAE. EAN in the mouse allows the use of genetically modified strains obtained by crossbreeding in various gene mutations. One example is knockout mice that lack P0,33 leading to dysmyelination and secondary autoimmunity.120 These mice were originally generated to better understand functional consequences of myelin deficiency in genetic models for Charcot-MarieTooth disease. By reverse transcription–polymerase chain reaction (RT-PCR) and Western blot, P0 was also found to be expressed in thymic epithelial cells.100,142 In P0 knockout mice on a C57BL /6 genetic background, tolerance to P0 is lacking and can be reconstituted by thymic transplants from wild-type mice100 or partly by generating bone marrow chimeras.142 Using the P0 knockout model, nontolerized or cryptic epitopes located in the extracellular P0 domain could be described, which may allow for development of better mouse EAN models in the future. All these models depend on immunization with a neuritogenic autoantigen. Recently, some mouse models have become available that develop spontaneous autoimmune diseases. Among these are genetically engineered mice or spon-

taneous mutants like the nonobese diabetic (NOD) mouse. In the NOD model, the co-stimulatory B7-1 and B7-2 molecules on immune cells have been shown to play distinct roles. Elimination of B7-2 expression by breeding NOD mice onto the B7-2–deficient background prevents diabetes, but leads to the development of a spontaneous autoimmune peripheral polyneuropathy (SAPP).118 Morphologic analysis revealed significant demyelination, with a mononuclear cellular infiltrate composed of dendritic cells, CD4⫹ T cells, and CD8⫹ T cells. Adoptive transfer of reactivated T cells from NOD mice with SAPP induced a severe peripheral neuropathy in the recipient crossbred NOD/severe combined immunodeficiency mice, thus formally fulfilling the traditional KochWitebsky’s postulates for an autoimmune pathogenesis of this mouse disease, as reviewed by Rose and Bona.113 Most EAN models feature combined demyelinating and axonal pathology. In rabbits, but not in rodent species, one recently described specific EAN model affects principally axons. On repetitive sensitization with a bovine brain ganglioside mixture, injected rabbits developed flaccid limb weakness of acute onset.157 Pathology showed predominant wallerian-like degeneration, with neither lymphocytic infiltration nor demyelination, indicating primarily axonal damage. The injected rabbits developed high titers of immunoglobulin G (IgG) antibodies to the monosialoganglioside GM1, and these antibodies were deposited on axons of the anterior roots and of peripheral nerves. Sensitization with purified GM1 also induced axonal neuropathy, indicating that GM1 was the likely immunogen in the ganglioside mixture. This model may help to clarify the molecular pathogenesis of axonal GBS and to develop appropriate treatment strategies. Disease Characteristics and Electrophysiology In EAN inflammation is multifocal, affecting nerve roots and peripheral nerves at multiple sites with, for unknown reasons, predominance in the lumbosacral region. Consequently, loss of tail tone occurs at disease onset. Disease progression is monitored by using a 10-point scale52,82 ranging from 0 ⫽ normal to 10 ⫽ death. Some investigators even prefer a 20-point EAN scale, which may be more sensitive in specific experimental settings.29 To obtain an objective scoring, the rater must be blinded as to the experimental groups. Grades 9 and 10 of 10 are usually not observed in EAN because, in contrast to EAE, the disease does not afflict the brainstem with its cardiovascular centers. All scoring is done on the same platform, which should be flat and not slippery. In case of painful adjuvant arthritis, scoring of gait is more difficult and prone to misclassification. In the Lewis rat, actively induced EAN typically starts around day 12 and then progresses for another 4 to 6 days to reach the peak of disease. At scores above 6, care must be taken that animals still have access to food and water. In full-blown EAN, concomitant weight loss of up to 15% of the initial weight is a common finding, probably reflecting

Experimental Autoimmune Neuritis

FIGURE 27–1 Antigen dose determines the severity of myelin-induced EAN. Lewis rats were immunized with increasing doses of bovine myelin in complete Freund’s adjuvant. Two representative cross sections taken at the S1 nerve root level are shown from a rat immunized with 3 mg of myelin (A) and from a rat immunized with 5 mg of myelin (B). Note predominant demyelination in a perivascular pattern in A versus widespread demyelination and admixed axonal changes in B. Bar: 20 ␮m. C, Representative electron micrograph with a demyelinated axon exhibiting mild degenerative changes adjacent to an axon with severe myelin disruption (long arrow). A ⫽ axon; M ⫽ mitochondria; SC ⫽ Schwann cell. (Electron micrograph provided by Dr. B. Schäfer.)

613

614

Neuroimmunology of the Peripheral Nervous System

cytokine and leptin stress responses similar to those in human disease.7 Rats with mild EAN may present without clinical signs, while histology clearly demonstrates inflammatory infiltrates in the PNS. Similar to studies in human patients, electrophysiologic studies can be performed under general anesthesia (neu-

roleptanalgesia) to define the functional consequences of PNS pathology. Thus a number of relevant nerve functions can be studied in serial electrophysiologic recordings (Fig. 27–2). As first signs of affliction, F-wave latencies become markedly prolonged and the responses are dispersed at disease onset.47,57 Progression of the disease

Day 0–3 Day 4

Time after transfer

15 ∞

18 ∞

1mV 2ms

21 ∞

24 ∞

Day 5 3 ∞

15 ∞ 0 2 4 6 8 10 ms

0 2 4 6 8 10 12 14 ms

Proximal

Distal

A

Time after transfer

5 weeks

50μV 5ms

6 weeks

100μV 5ms

7 weeks

500μV 2ms

9 weeks

1mV 2ms

10 weeks

1mV 2ms Proximal

Distal

B FIGURE 27–2 Representative recording of a serial nerve conduction study in AT-EAN. A, Time course of proximally and distally evoked compound muscle action potentials and F waves in a rat developing fulminant neuritis with complete paraplegia 5 days after injection of a high cell dose of P2-specific CD4⫹ T cells. B, Five weeks later, the first baseline deflections (compound muscle action potentials in statu nascendi) can be seen after distal stimulation, indicating earliest signs of regeneration. Over the following 10 weeks, functional restitution to virtually normal conduction occurred. (From Heininger, K., Stoll, G., Linington, C., et al.: Conduction failure and nerve conduction slowing in experimental allergic neuritis induced by P2-specific T-cell lines. Ann. Neurol. 19:44, 1986, with permission.)

615

Experimental Autoimmune Neuritis

is paralleled by slowing of motor nerve conduction velocity by up to 40%, decreased afferent nerve conduction velocity, and marked lowering of proximal and distal compound muscle action potential amplitudes. Conduction failure or conduction block may be observed in 80% to 100% of rats with EAN at disease maximum, and also F waves disappear in up to 80% of rats at this stage of the disease. Depending on disease severity and the degree of admixed axonal damage, these abnormalities may only partly improve during recovery, thus leading to irreversible electrophysiologic changes similar to those in patients with GBS.

revealed that they strip off myelin lamellae, induce vesicular disruption of the myelin sheath, and phagocytose damaged myelin.114 Interestingly, macrophages and not Schwann cells are the primary major histocompatibility complex (MHC) class II–expressing cell type in the inflamed PNS.121 Depending on the amount of myelin or neuritogenic peptide used for immunizations, and thus on the intensity of inflammation, an increasing degree of admixed axonal damage may occur.45,47,114 In particular, in T-cell line–mediated adoptive transfer (AT-) EAN (see below), the neuritis may take a fulminant course as a result of prominent endoneurial edema and ischemia that lead to severe axonal damage.

Pathology The inflammatory infiltrate in EAN is largely made up of lymphocytes and macrophages. In the rat, degranulated mast cells also are present. This leads to focal demyelination of nerves predominantly around venules. A semiquantitative and robust 4-point score has been proposed that facilitates quantification of inflammatory and demyelinating lesions in semithin sections52 (Fig. 27–3). As the inflammatory lesion proceeds, macrophages emerge as the predominant cells. These macrophages may be derived from blood-borne monocytes and endoneurial macrophages.76,102 Electron microscopic studies have

Cellular and Molecular Disease Mechanisms in EAN Entry of Inflammatory Cells: The Role of Cell Adhesion Molecules, Chemokines, and Matrix Metalloproteinases. In order to mediate a local immune response within the peripheral nerve, activated immunocompetent cells need to cross the BNB, an anatomically tight interface separating the systemic immune compartment from the PNS109 (Fig. 27–4). This mechanism of transendothelial migration is a multistep process occurring in an ordered sequence. In the first step, cellular adhesion molecules (CAMs) are expressed on leukocytes and vascular endothelium, resulting in a

FIGURE 27–3 Semiquantitative, categorical histologic score to assess inflammation and demyelination. Semithin sections are cut and stained with toluidine blue. At least 50 perivascular areas of cross sections are examined by a blinded observer as described by Hartung et al.52 A, Grade 0 ⫽ normal perivascular area. B, Grade 1 ⫽ mild cellular infiltration adjacent to a vessel. C, Grade 2 ⫽ cellular infiltrate plus demyelinating fibers adjacent to a vessel. D, Grade 3 ⫽ cellular infiltrate plus demyelination around a vessel and at more distant sites and minor axonal damage. (Magnification approximately ⫻450.)

A

B

C

D

616

Neuroimmunology of the Peripheral Nervous System

A

RBC 3

1 2

B

slowing and attachment of the circulating immune cells along the vessel wall. Normally, the flowing blood quickly dislodges cells that touch the vessel wall; thus adhesion molecules must act as mechanical anchors, but also function as tissue-specific recognition molecules. Based on structural differences, CAMs can be categorized into four groups: the immunoglobulin superfamily, selectins, integrins, and cadherins, all of which are involved in lymphocyte recruitment and extravasation.5 The specific function of individual CAMs has been elucidated by blocking their action with specific monoclonal antibodies in various animal models (see Transmigration and Early Effector Phase: Adhesion Molecules and MMPs below) or by generat-

FIGURE 27–4 Pathology at the blood-nerve barrier in EAN. Electron micrographs illustrating perivenular inflammation and endothelial morphology in rat sciatic nerve after immunization with the P2 aa 60–70 peptide of basic P2 protein. A, The perivenular interstitium contains many mononuclear cells and a single erythrocyte while an intravascular lymphoid cell is adherent to the adluminal endothelial surface. Intercellular tight junctions appear normal. (⫻7000.) B, Endothelial abnormalities in a vessel surrounded by inflammatory cells. Endothelial cell processes labeled 1, 2, and 3 exhibit intercellular dissociation with loss of tight junctions. A wide intercellular space has opened between endothelial cells 1 and 2, and a platelet overlies but does not block the space. Lower left inset, Electron-dense material appears to accumulate next to the basal lamina in the subendothelial space (arrowhead) (⫻14,600). Loss of tight junctions and endothelial separation appears between endothelial cells 2 and 3 (arrowhead). Lower right inset, Separation between endothelial cells 2 and 3 at higher magnification (⫻14,600). (From Powell, H. C., Olee, T., Brostoff, S. W., and Mizisin, A. P.: Comparative histopathology of experimental allergic neuritis induced with minimum length neuritogenic peptides by adoptive transfer with sensitized cells or direct sensitization. J. Neuropathol. Exp. Neurol. 50:658, 1991, with permission.)

ing knockout animals for the corresponding gene of a particular CAM. The extravasation of leukocytes into the CNS parenchyma is facilitated by the expression of CAMs and counterligands on both cerebral vascular endothelial cells and leukocytes, which strengthens their adhesive interaction.5 The initially unstable interaction provided by some CAMs is strengthened through the interaction of other CAMs, such as intracellular adhesion molecule-1 (ICAM-1) with lymphocyte function–associated antigen-1 (LFA-1), and vascular cell adhesion molecule-1 (VCAM-1) with very late antigen-4 (VLA-4), as demonstrated in EAN.3,24 Within the inflammatory process, expression of CAMs is upregulated, thus

Experimental Autoimmune Neuritis

augmenting attachment of immune cells. After interaction with the corresponding ligands, various CAMs are shed or cleaved from the cell surface.32 They circulate within body fluids and are thought to regulate cellular interactions and to promote de-adhesion. In a second step, chemokines come into play, providing signals to direct leukocyte migration into and within the extravascular space. Because lymphocytes must be positioned correctly to interact with other cells, the pattern and anatomic distribution of chemokines within the target tissue as well as the types of chemokine receptors expressed on the cell surface become critically important in orchestrating the ongoing immune responses.16,101 In EAN, a differential pattern of chemotactic signals with peak expression levels prior to or coincident with maximum clinical disease activity has been defined, involving different chemokines such as CXCL10 (IP-10), CCL2 (MCP-1), CCL3 (MIP-1␣), and CCL5 (RANTES).81 Corresponding chemokine receptors were detected in the inflamed PNS, with CCR1 and CCR5 primarily expressed by endoneurial macrophages, and CCR2, CCR4, and CXCR3 by invading T lymphocytes. Quantitative analysis revealed that CXCR3 is highest in infiltrating T cells compared to the other receptors. Its ligand CXCL10 mirrors the distribution of the cognate receptor within the inflamed PNS, and delineates endothelial cells as the primary cellular source of CXCL10, thus pointing to a pathogenic role for specific chemokine receptors and IP-10 in the generation of inflammatory demyelinating neuropathies.81 Finally, in a third step, matrix metalloproteinases (MMPs) are secreted by leukocytes in order to disrupt the BNB, thus facilitating transmigration of cells and of plasma-derived macromolecules (immunoglobulins and complement components) into the parenchyma of the nervous system. A differential expression pattern of various MMPs, specifically MMP-7 and MMP-9, can be depicted in the endoneurium and epineurium of the inflamed peripheral nerve (Fig. 27–5).79,92 The therapeutic potential of blocking chemokines and MMPs is described below (see Transmigration and Early Effector Phase: Adhesion Molecules and MMPs). Amplification and Termination of the Local Immune Response: The Role of Cytokines. Concomitant with cellular infiltration, a broad range of cytokines is induced in EAN nerves.34,77 Cytokines are mainly expressed by T cells and macrophages, but also to some extent by Schwann cells. At disease onset, T cells and a subpopulation of macrophages express interferon-␥ (IFN-␥), the prototype of a proinflammatory cytokine122 (Fig. 27–6). IFN-␥ is inducible by interleukin (IL)-18, a key mediator of innate immunity. In EAN, infiltrating macrophages strongly express IL-18. It is of note that patients with GBS may show increased IL-18 serum levels during the acute phase of their disease.65 IL-18 probably contributes to the

617

FIGURE 27–5 Immunohistochemistry for matrix metalloproteinases (MMPs) in the sciatic nerve of a Lewis rat with experimental autoimmune neuritis with clinically active disease. Arrowheads point to immunoreactivity for MMP-9, which can be localized around blood vessels, and to invading mononuclear cells, which can be identified as infiltrating T lymphocytes. Localization and distribution of the signal suggest that MMPs, such as MMP-9, are involved in the disruption of the blood-nerve barrier as well as in the process of migration of immunocompetent cells within the inflamed peripheral nerve. (Original magnification: ⫻200.) See Color Plate

amplification of the initial inflammatory response. In keeping with a key role in disease initiation, neutralization of IFN-␥ by antibodies attenuated clinical EAN53 (Fig. 27–7). IFN-␥ induces the expression of MHC class II molecules on monocytes/macrophages and dendritic cells, which are indispensable components of the trimolecular antigen complex in specific antigen recognition by autoreactive T cells. IFN-␥, moreover, activates macrophages to release tumor necrosis factor-␣ (TNF-␣) and other potentially myelinotoxic substrates. Indeed, in EAN macrophages adhering to nerve fibers expressed TNF-␣,130 thus providing the appropriate local milieu for myelin damage. Of note, in GBS patients, levels of circulating TNF-␣ correlated with electrophysiologic abnormalities.125 Injections of TNF-␣ into nerves may cause inflammation, demyelination, and damage to endothelial cells,111,115 although this mechanism has been debated by others.140 In further support of a detrimental role of TNF-␣ in EAN, experimental neutralization of TNF-␣ ameliorated demyelination.130 Seemingly in contradiction to this hypothesis, TNF-␣ also diminishes inflammation in nerves by inducing T-cell apoptosis. Apoptosis of T cells is an important mechanism for termination of immune responses (see below). Accordingly, neutralization of TNF-␣ by antibodies led to a prolonged persistence of T cells in

618

Neuroimmunology of the Peripheral Nervous System

FIGURE 27–6 Interferon-␥ (IFN-␥) immunoreactivity in EAN. Ventral roots of a rat with experimental autoimmune neuritis 12 days after active immunization. Cryosections of 1 ␮m thickness were labeled for IFN-␥ immunoreactivity by the DB-1 monoclonal antibody (A) and W3/13 antigen expressed on T cells and some neutrophils (B). Arrows denote cells staining positive for W3/13 and IFN-␥, respectively. ax ⫽ axon; my ⫽ myelin sheath; V ⫽ venule. Counterstaining was with hematoxylin. T cells, neutrophils, and macrophages are stained for IFN-␥. However, IFN-␥ was present in nerves only transiently between days 12 and 14 after immunization. Bar: 10 ␮m. (See Schmidt et al.122)

EAN nerves.151 At present, it appears that in EAN neutralization of TNF-␣–mediated myelin damage outweighs the ameliorating effects through T-cell removal, because animals treated with antibodies against TNF-␣ showed a modest overall clinical benefit.130 In addition to IL-18, IL-12 is a key cytokine of the innate immune system produced by type 1 T-helper (Th) cells and macrophages.139 IL-12 exerts proinflammatory actions as a p35/40 heterodimer, but is immunosuppressive in the form of its IL-12 p40 homodimers. IL-12 messenger RNA (mRNA)–positive cells have been described in EAN by in situ hybridization at onset of overt disease.160 However, IL12 p35 and IL-12 p40 have not been investigated separately. Another study using RT-PCR found no modification

of constitutive IL-12 p35 mRNA levels during EAN, but marked induction of IL-12 p40 mRNA that was delayed to the recovery phase.34 Functionally, IL-12 injected into healthy peripheral nerve provoked inflammation and caused marked demyelination.107 Thus it is conceivable that macrophage-derived IL-12 might promote inflammation at onset of EAN, while IL-12 p40 homodimers may have immunosuppressive effects during recovery. Cellular infiltration in EAN apparently has an intrinsic anti-inflammatory component in part mediated by downregulatory cytokines. IL-10, a potent downregulating Th2 cytokine, is already induced on a subpopulation of infiltrating macrophages at disease onset and parallels the initially strong expression of proinflammatory cytokines.34,64 When therapeutically administered from the start of immunization, IL-10 effectively suppressed and shortened clinical EAN. Even when given after day 12 postimmunization, after clinical EAN had been established, IL-10 also effectively suppressed the severity of EAN.6 Both prophylactic and therapeutic treatment with fusidin (4 mg/day intraperitoneally per rat) markedly ameliorated the clinical course of the disease compared to vehicle-treated animals. The beneficial effects were associated with profound modifications of the capacity of these rats to produce and release the aforementioned pro- and anti-inflammatory cytokines.21 Transforming growth factor-␤1 (TGF-␤1), another antiinflammatory cytokine, reached peak levels later, but shortly before clinical recovery.75 In support of a diseasemitigating effect, external administration of TGF-␤1 to EAN animals reduced inflammation and clinical severity.40 In conclusion, analysis of cytokine responses in autoimmune disorders such as EAN and GBS helped to disclose some of the regulatory mechanisms operative in immunemediated demyelination. Effector Mechanisms of Myelin Destruction. T-cell infiltration is the initial event in EAN. However, the mere presence of autoreactive T cells in nerves is not sufficient to induce clinical disease.54 Recruitment of macrophages is a crucial secondary step. The essential requirement of macrophages for disease induction in EAN has been highlighted by depletion experiments. Animals received intraperitoneal injections of silica quartz dust to divert monocytes to the peritoneal cavity. Macrophage depletion prevented all clinical, electrophysiologic, and histologic signs of EAN.19,52,58,136 There are two ways by which macrophages can destroy nerve tissue: by diffuse release of toxic factors or, more specifically, by adhering to and attacking the myelin sheath. MACROPHAGE-MEDIATED MECHANISMS: TOXIC FACTORS. Besides complement and cytokines, macrophages are a major source of toxic factors such as arachidonic acid metabolites, nitric oxide, oxygen radicals, and proteases. Their functional significance in disease progression has been

619

Experimental Autoimmune Neuritis

Day 0

Day 4

5mV 1ms

Day 8

SHAM

+anti-IFN γ

+IFN γ

FIGURE 27–7 Effect of IFN-␥ in EAN induced by adoptive transfer of P2-specific T-cell lines. Serial motor nerve conduction studies at the sciatic nerve. After T-cell transfer, animals received a monoclonal antibody to IFN-␥ (DB-1) on days 0, 3, and 7, leading to amelioration of disease activity. Conversely, rats treated with recombinant rat IFN-␥ showed more pronounced functional deficit. (From Hartung, H. P., Schäfer, B., Van der Meide, P. H., et al.: The role of interferon-gamma in the pathogenesis of experimental autoimmune disease of the peripheral nervous system. Ann. Neurol. 27:247, 1990, with permission.)

elucidated by pharmacologic blocking experiments (see Effector Phase: Macrophage- and Antibody-Directed Treatments below). Arachidonic acid metabolites, possibly synergizing with activated complement, could function as chemoattractants and secretagogues for inflammatory cells or could enhance vascular permeability and breach the BNB. Reactive oxygen species such as superoxide anion, hydrogen peroxide, and hydroxyl radicals could inflict peroxidation injury on myelin, but may also act by generating chemotactic signals and by exerting cytotoxic effects on endothelial cells.85 Macrophage-derived neutral proteases and phospholipases induce myelin damage in vitro. Accordingly, microinjection of a proteinase into rat sciatic nerve produced inflammatory demyelination, and treatment of EAN rats154 with proteinase inhibitors delayed development of clinical disease.119 B-CELL AND MACROPHAGE-MEDIATED MECHANISMS: ANTI-MYELIN ANTIBODIES AND COMPLEMENT. The hallmark of actively induced EAN and of the majority of human GBS cases is segmental demyelination that cannot be explained by diffuse release of toxic mediators alone. It is a basic morphologic observation that, in EAN and GBS, macrophages adhere to nerve fibers at an early stage of disease development and often normally appearing nerve

fibers are attacked.41,88 In contrast, after axotomy or nerve crush, macrophages selectively infiltrate nerve fibers undergoing wallerian degeneration, but they spare intact fibers.129 In EAN, macrophages strip off myelin lamellae from the axons, resulting in demyelination. The mechanisms by which initial T-cell infiltration leads to macrophage adherence on the surface of individual myelin sheaths are largely unknown. In analogy to human disease,11,156 it is likely that humoral autoantibodies to myelin components can play a role in the pathogenesis of EAN. These antibodies may block nerve excitation or neuromuscular transmission or may enhance T-cell–mediated demyelination by complement-mediated lysis and antibody-dependent cellular cytotoxicity effector mechanisms (see Chapter 26). Polyvalent antibodies to various components of the PNS have been detected in hyperimmune serum of rabbits and guinea pigs with EAN. However, these autoantibodies are unable to trigger demyelination without prior disruption of the BNB. To overcome this obstacle, the AT-EAN model can be used in co-transfer studies to examine the putative demyelinating activity of sera from patients with GBS or CIDP.42,54,138 Following an intravenous injection of a subneuritogenic cell dose, breakdown of the BNB is observed.43 Subsequent

620

Neuroimmunology of the Peripheral Nervous System

intravenous injection of purified IgG fractions from GBS/EAN then allows for examination of demyelinating activity by electrophysiology and/or histology (see above). In EAN increased antibody titers against peripheral myelin components could be measured in the blood stream.84,161 They can gain access to nerves by T-cell–mediated disruption of the BNB.126 Accordingly, a rapid increase in the passage of immunoglobulin into spinal roots, together with endoneurial infiltration of T lymphocytes and polymorphonuclear leukocytes, was demonstrated in AT-EAN. Accumulation of immunoglobulin was maximal during the worsening of neurologic deficit, and declined rapidly before the onset of neurologic recovery.43 In support of a decisive role of complement in the pathogenesis, decomplementation of animals with cobra venom factor partly suppressed EAN.27,144 Moreover, in myelin-induced EAN, macrophages were concentrated in areas of strong terminal complement complex (TCC) accumulation on myelin sheaths and Schwann cells131 (Fig. 27–8). Functionally, TCC deposition on the surface of myelin sheaths may promote an influx of calcium, which is considered an important initial step in myelin breakdown and cytotoxicity in general. Additionally, macrophages in culture respond to TCC by liberating chemotactically active and proinflammatory eicosanoids that have been found to contribute to the pathogenesis of EAN.47,52 Alternatively, TCC formation in myelin membranes may cause activation

of myelin-associated neutral proteinases with subsequent hydrolysis of myelin proteins. It is conceivable that TCCtargeted myelin is selectively attacked by macrophages. In support of this notion, Hafer-Macko and colleagues44 showed TCC deposits on the outer surface of myelin sheaths in nerve specimens from demyelinating GBS, while in axonal GBS, TCC was deposited axonally at paranodes. In both settings, macrophages adhered to nerve fibers at the location of TCC deposits. Schwann cells appear to escape TCC attacks, as shown in vitro by induction of cell cycle activation, proliferation, and rescue from apoptotic cell death.60 It was concluded from these in vitro studies that sublytic C5b-9 detected on Schwann cells in vivo during inflammatory neuropathies may even facilitate survival of Schwann cell to ensure remyelination. The Role of Mast Cells and Polymorphonuclear Leukocytes. Mast cells are part of inflammatory infiltrates in EAN,99 but they decrease in number in the course of EAN9 and degranulate possibly as a consequence of T-cell–mediated delayed-type hypersensitivity63; alternatively, their activation may be induced by myelin as shown in vitro. In turn, mast cell–derived proteases can degrade myelin in vitro. Mast cells are important sources of vasoactive amines (5-hydroxytryptamine, histamine) and arachidonic acid metabolites that both could augment vascular permeability and thereby induce a leaky BNB. In support of a pathogenic role, mast cell–stabilizing drugs such as reserpine and nedocromil prevented or attenuated the disease.10 Polymorphonuclear leukocytes that have an enormous proinflammatory potential can be detected in EAN lesions, but their pathophysiologic role is not well defined.43,88,122

Adoptive Transfer EAN

FIGURE 27–8 Complement deposition in EAN. Ventral root section from an animal with EAN 11 days after immunization (i.e., before overt clinical disease), immunostained for the terminal complement complex (C5b-9). The reaction product is localized on the surface of myelin sheaths (arrows) and on myelinating Schwann cells (asterisk). Staining appears preferentially around a venule (v), while other fibers are spared. Terminal complement complex is deposited only transiently for a few days. Bar: 10 ␮m. (See Stoll et al.131)

AT-EAN and Disease Course The generation of antigen-specific autoaggressive T-cell lines was an important step toward a better understanding of experimental paradigms in EAN. AT-EAN has also provided conclusive proof of the pivotal pathogenic role of T cells in experimental neuritis, thus fulfilling the Koch-Witebsky criteria for EAN as an (experimental) autoimmune disorder.113 Following the classic methods developed for EAE,8 T-cell lines can be established in vitro from lymph node cells of rats immunized with P2.95,96 These primary cell cultures are propagated by repetitive, antigen-specific activation cycles in turn with IL-2–mediated expansion to generate T-cell lines. Also, T-cell clones can be generated from immunized Lewis rats.89 By intravenous transfer of defined numbers of T lymphoblasts, ranging from 1 to 20 million cells, monophasic AT-EAN is induced. In analogy to actively induced EAN, T-cell lines

Experimental Autoimmune Neuritis

reactive with P2 and P0 were first found to transfer the disease.96 Epitope specificity of autoreactive T-cell lines that can produce disease has been examined and found to reside primarily in the amino acid sequences 61 to 70 of P2105 or 180 to 199 of P0.96 AT-EAN of the Lewis rat begins within 3 to 4 days after cell injection and typically achieves its maximum between days 6 and 10: High T-cell numbers cause earlier and more severe disease than do lower numbers.59 Disease incidence reaches up to 100% going in parallel with a very low interindividual variation of clinical disease; this is ideal to study smaller groups of animals, which is of value in therapeutic settings with restricted access to precious reagents. Although BN rats have a very low susceptibility to overt EAN induced by active immunization, T-cell lines specifically reactive to bovine P2 could be isolated from these normal-looking rats. Upon transfer to syngeneic BN recipients, these T cells were capable of evoking EAN.97 This indicates that autoreactive T-cell clones are also contained in the T-cell repertoire of this nonresponder strain. Non-neural antigens such as ovalbumin are capable of inducing disease by using another paradigm of generating AT-EAN. In a first step, T-cell lines are generated that react specifically to ovalbumin, a foreign (nonself) protein in the rat. Following intraneural microinjection of ovalbumin into one sciatic nerve, ovalbumin-specific T-cell lines were transferred that led to EAN-like inflammatory lesions in the nerve injected with ovalbumin, but not at the contralateral side that received casein as control antigen.54 These observations allow us to conclude that, in principle, any foreign protein achieving access to the PNS may become subject to immune surveillance. Applied to the situation in GBS, it is conceivable that preceding infections with bacteria or viruses may initiate the autoaggressive assault once a specific T-cell repertoire has been generated and antigens of the infectious agent are deposited in nerve tissue. Alternatively, an immune reaction may occur with infectious agents such as Campylobacter jejuni that show molecular similarities with nerve antigens (molecular mimicry).23 Pathology of AT-EAN In contrast to active EAN, the T-cell response is predominant in AT-EAN, with a negligible contribution of B-cell–derived autoantibodies. Macrophage-derived effector mechanisms are also operative in AT-EAN.58,63 Usually, pathogenic changes are initiated by vasogenic edema resulting from perivascular influx of autoaggressive lymphocytes, followed by increased endothelial permeability to serum proteins. In full-blown EAN, the BNB breaks down completely and allows rapid immigration of numerous inflammatory cells.58,63 Thus endoneurial fluid pressure and edema increase rapidly, adding superimposed ischemichypoxic damage. This is associated with rapid deterioration

621

of nerve conduction up to complete and long-lasting conduction failure in severe disease, indicating massive destruction of axons. The electrophysiologic findings resemble hyperacute GBS.59

Chronic (Relapsing) EAN: Mode of Immunization and Species-Specific Aspects Chronic EAN has been developed as a model for human CIDP. The only myelin antigen for which there is clear evidence for a purely demyelinating response is galactocerebroside (GalC; see Chapter 26). Rabbits immunized with GalC develop a slowly progressive disease starting with trembling and then progressing to tetraplegia (Fig. 27–9).116,132 High titers of circulating antibodies against GalC could be detected.132 Intraneural injections of antiserum from afflicted rabbits with high titers of these antibodies caused demyelination in recipient rat sciatic nerve in vivo87 and also could be seen in vitro in myelin preparations. This is remarkable because the antiserum recognizes epitopes involving the galactose moiety of the glycolipid and does not bind to micellar dispersions of the purified lipid. Surprisingly and still unexplained, purified IgG fractions containing high-titer anti-GalC antibodies do not cause demyelination upon passive transfer, nor can they inhibit normal (re-) myelination after experimental nerve damage in rabbits with ongoing GalC-EAN.128,138 In the absence of complement, binding of the anti-GalC serum to myelin causes swelling and separation of compacted myelin lamellae. The usefulness of GalC-EAN as a disease model is limited because the rabbit is the only species producing overt disease, while mice and rats are poor responders to an immune challenge with GalC. Chronic relapsing EAN can be induced in the Lewis rat by treatment with low-dose cyclosporine A after active immunization with myelin.108 First, cyclosporine A prevents the development of EAN. Upon its withdrawal, chronic relapsing EAN with severe inflammation and demyelination of nerve roots, ganglia, and spinal nerves is observed. This chronic EAN also goes along with signs of remyelination, yet this model is not very robust. DA rats show mild chronic EAN upon immunization with BPM, which makes this model more suitable for testing new strategies of therapeutic intervention. Repeated adoptive transfer of P2-specific T-cell lines has also been used to produce chronic relapsing EAN resembling relapsing GBS but not CIDP.90 It is plausible that in chronic EAN, similar to the situation in EAE, chronic and ongoing destruction of myelin causes sensitization to other components of the myelin sheath as well. This paradigm of “epitope spreading” as shown in EAE91 has not conclusively been demonstrated in EAN (see Table 27–1).

622

Neuroimmunology of the Peripheral Nervous System

Pre-immunization 2mV 2ms 4 months 10 months 18 months

C

Proximal

Distal

FIGURE 27–9 Galactocerebroside-induced chronic EAN in the rabbit. A, Dorsal root of a rabbit 20 weeks after primary immunization with galactocerebroside, methylated bovine serum albumin, and complete Freund’s adjuvant. Numerous axons (A) close to the venule (v) are completely demyelinated (arrows), and macrophages abound around the vessel. B, Electron micrograph from the same animal showing a macrophage with an activated nucleus (N) loaded with myelin debris and encircling demyelinated axons (A), one of which is enveloped by its Schwann cell cytoplasm. C, Representative example of a serial nerve conduction study. Compound muscle action potentials evoked after stimulation of the sciatic nerve at proximal and distal sites are recorded over the small foot muscles. Note the parallel decrease of potential amplitudes after proximal and distal stimulation at months 4 and 10 and the prolonged latencies of the M and F responses. After 18 months, following clinical recovery, nerve conduction velocities reached nearly preimmunization values. (From Stoll, G., Schwendemann, G., Heininger, K., et al.: Relation of clinical, serological, morphological, and electrophysiological findings in galactocerebrosideinduced experimental allergic neuritis. J. Neurol. Neurosurg. Psychiatry 49:258, 1986, with permission.)

623

Experimental Autoimmune Neuritis

result of downregulation of effector molecules and T-cell apoptosis (Table 27–2 and Fig. 27–10).

EXPERIMENTAL TREATMENTS IN EAN The immune response in EAN can be divided into three phases: induction, amplification, and effector phases. In the induction phase, the injected autoantigen is presented to “naive” T cells by professional APCs such as macrophages or dendritic cells, resulting in T-cell activation (see Immunologic Principles above). Two external signals are crucially required for effective T-cell activation by antigen presentation: the antigen-specific signal provided by the immunogenic peptide and presented in the context of MHC molecules on APCs, and the antigen-independent signal called co-stimulation (see Chapter 98). Co-stimulation is mediated by adhesion molecules of the integrin family, such as ␣L␤2 (LFA-1, CD11a/CD18) and ␣4␤1 (VLA-4, CD49d/CD29), and by adhesion molecules of the immunoglobulin superfamily, such as ICAM-1 (CD54), VCAM-1 (CD106),5 and, of special functional relevance, cytotoxic T lymphocyte–associated antigen-4 (CD152), B7-1 (CD80), and B7-2 (CD86), expressed on both T cells and APCs117 (see Chapter 26). Activated T cells then circulate in the blood, attach to the venular endothelium in the PNS, and penetrate the BNB through activation of MMPs.35 This transendothelial migration gives rise to the effector phase of the immune response in EAN. In the PNS, the autoantigen is then presented to T cells by macrophages. At the same time, the reactivated CD4⫹ T cells augment the immune response by recruiting further T cells and macrophages via chemokines and cytokines (amplification phase). The resulting breakdown of the BNB then allows the passage of circulating autoantibodies and complement components (a complementary pathway of the effector phase). Finally, the termination of the ongoing immune response is the

Induction Phase: The Role of Antigen Presentation and Co-stimulation The first important step in the pathogenesis of EAN is the presentation of autoantigen in the induction phase. This results in the physiologic activation of disease-inducing T cells. Several tools are available to modulate immunoregulatory mechanisms. Using the paradigm of myelinspecific oral tolerance,148 Gaupp et al. have shown that active EAN can be prevented or ongoing disease be ameliorated by feeding with oral myelin.31 This effect can be further modulated by adjuvants such as cholera toxin.67 Active suppression of effector T cells rather than generation of Th2-type T cells was identified as the putative underlying mechanism of oral tolerance. Therapy targeted against the ␣/␤ TCR by using the monoclonal antibody (mAb) R73 showed therapeutic and preventive potential in different EAN models.69 The underlying molecular mechanism was impairment of antigen recognition and T-cell function by TCR occupancy. Because the frequency of TCR-positive cells was transiently reduced by only 50% and returned to normal within 10 days as shown by fluorescence-activated cell sorter analysis, TCR-mediated T-cell downregulation rather than full depletion of lymphocytes appears to be decisive. A more selective way to inhibit proliferating T cells is aimed at blockade of the IL-2 receptor, which is expressed on this subset of activated T cells. In AT-EAN, prophylactic treatment with the mAb ART18, directed at the IL-2 receptor, mitigated disease severity when given

Table 27–2. Recent Experimental Treatments of Immune-Mediated Neuropathies Experimental Strategy

Target

Blockade of adhesion molecules

• • • • • • • • • • • • • • • •

Tissue invasion Induction of T-cell apoptosis Cytokine modulation Induction of T-cell anergy

Induction of Th2 response or regulatory T cells Mitigation of macrophage function

ICAM-1/LFA-1 VCAM-1/VLA-4 L-selectin Matrix metalloproteinases Antigen therapy Steroid pulse therapy IL-2 receptor IL-10 T-cell receptor ␣/␤ CD2 molecule T-cell activation Anti-CD28 Oral tolerance Cyclooxygenase inhibitors Oxygen radical scavengers Liposome-encapsulated dichlormethylene

Reference 4 24 2 61 150, 158 6, 50 69 70 127 122, 134 31 46, 51 68

624

Neuroimmunology of the Peripheral Nervous System

VENULE

PERIPHERAL NERVE

APC

BNB

MΦ T

TNF OH¯ PGE

Res. MΦ

T

TH

Axon B

C6

CAMs T

Chemokines T

Abs

MMPs T

PLT

NO TNF MMPs

LT TNF

C5b-9

MΦ T SC

CD40L MΦ

LTC4 MMPs ROS OH¯ IL-1

Apoptosis

MC

FIGURE 27–10 Simplified scheme depicting the hypothetical sequence of major immune mechanisms in EAN (see sections on Immunologic Principles and The EAN Models in text). Circulating lymphocytes are activated by an as yet unknown antigen (e.g., P2 or P0 in EAN). Activated lymphocytes adhere to the endothelial layer of intraneural venules and migrate intraneurally. Adhesion, rolling, and migration are governed by a number of cell adhesion molecules. Intraneurally, macrophages are activated and produce cytokines, reactive oxygen species, prostaglandins, and complement factors, which in turn activate further inflammatory cells that are resident or have meanwhile transmigrated through an open bloodnerve barrier. Here, metalloproteinases, leukotrienes, and interleukin-1 play a role. Moreover, mast cells may also induce increased vascular permeability and inflammatory reactions. Platelets (PLT) may stimulate antigen-presenting cells (APC) through the CD40 ligand pathway. Activated macrophages may also function as antigen-presenting cells and activate specific and bystander T-cells. Moreover, macrophages may directly damage myelin sheaths and axons by cytotoxic mechanisms including oxygen radicals. Transmigrated T-helper cells activate transmigrated B cells to produce myelin-specific antibodies that can attach to the myelin sheath or to the axon and damage nerve fibers with the mediation of complement. Macrophages may also be targeted to nerve fibers after being armed with circulating antibodies (not shown) following the mechanism of antibody-dependent cellular cytotoxicity. Finally, antibodies may directly block axonal excitation at the nerve terminal. It is not clear whether Schwann cells per se are also attacked by antibodies and T cells, and some may undergo programmed cell death (apoptosis). The inflammatory reaction can be terminated by T-cell apoptosis induced by repeated antigenic stimulation and by the Fas/Fas-ligand system, in which Schwann cells may also play an important part. Inset, Electron micrograph of an apoptotic Schwann cell. T-cells may also be downregulated by the cytokine TGF-␤ (not shown). Abs ⫽ antibodies; APC ⫽ antigen-presenting cell; BNB ⫽ blood-nerve barrier; CAM ⫽ cellular adhesion molecule; C ⫽ complement components; IL ⫽ interleukin; LTC ⫽ leukotriene; MC ⫽ mast cell; MMP ⫽ matrix metalloproteinases; M⌽ ⫽ macrophage; NO ⫽ nitric oxide; OH ⫽ hydroxyl radicals; PGE ⫽ prostaglandin E; PLT ⫽ platelet; SC ⫽ Schwann cell; Th ⫽ T-helper cell; TNF ⫽ tumor necrosis factor. See Color Plate

Experimental Autoimmune Neuritis

625

during the latent induction phase but not after onset of clinical symptoms.50 The most potent co-stimulatory signal known to date is mediated by the T-cell surface receptor CD28 interacting with CD80 (B7-1) or CD86 (B7-2) on APCs such as macrophages or dendritic cells. Normally, a CD28 signal alone does not lead to activation of resting T cells. In apparent contradiction to this paradigm, CD28-specific mAbs have been identified recently in several experimental models that induce proliferation and/or cytokine secretion of T cells in the absence of TCR engagement. In the rat system, the mAb JJ316 activates all resting T cells to proliferate without a TCR signal in vitro and in vivo133 and is therefore called “superagonistic.” In keeping with the Th2-promoting effect of CD28 signals in co-stimulation, it was shown that T-cell activation with the mAb JJ316 primed CD4⫹ T cells for Th2 differentiation in vitro, and induced IL-4 and IL-10 expression along with Th2dependent immunoglobulin isotypes in vivo. Treatment with JJ316 during the induction phase of active EAN and AT-EAN dramatically reduced disease severity and improved nerve function as revealed by electrophysiology. In addition, JJ316 given 1 week before immunization had a preventive effect.123 Morphologically, JJ316 markedly reduced T-cell infiltration of the sciatic nerve without induction of apoptosis. These results suggest that the superagonistic mAb preferentially activates regulatory (formerly called suppressor) T cells, thus providing a model for novel antibody-mediated treatments as an alternative to conventional immunosuppressive strategies.

ameliorating EAN24; this integrin seems to be the adhesion molecule that is most important in transendothelial migration of T cells in rodent EAN and hence the most promising target candidate for future therapeutic intervention in GBS. Surprisingly, blockade of VLA-4 and VCAM-1 blockade are also effective in enhancing T-cell apoptosis in the inflammatory lesion (see Termination of the Immune Response below).93 In vivo, inhibition of other coaccessory molecules such as CD2 turned out also to be effective in various EAN models.70 Since the anti-CD2 antibody did not exert its effect by inhibition of T-cell activation, induction of anergy, or depletion of T cells, the most probable mechanism is thought to be impairment of T-cell migration across the BNB. Transmigration of T cells through the BNB is mediated by upregulation of a repertoire of enzymes. Among these are MMPs, a family of calcium-dependent zinc endopeptidases that may be implicated in the pathogenesis of inflammatory demyelinating disorders.48 MMPs are also involved in the processing of TNF-␣. Among the MMPs, MMP-7 and MMP-9 are upregulated in inflamed muscle or nerve specimens in human nerve biopsies and in EAN nerves,18,20,78,80 as shown by immunocytochemistry, RT-PCR, and zymography. Indeed, the inhibition of these proinflammatory enzymes by the broad-spectrum MMP inhibitor BB-1101 exerted a strong preventive and also therapeutic effect in EAN.112

Transmigration and Early Effector Phase: Adhesion Molecules and MMPs

By transmigration and damage to the separating endothelial barrier, T cells breach the way for the passage of other humoral and cellular components of the inflammatory cascade from the blood stream to the nerve parenchyme. Intraneurally, CD4- or CD8-positive T cells meet their target antigen presented by MHC class II or class I molecules on infiltrating or resident macrophages102,121 or on nonprofessional APCs such as Schwann cells37 and support the amplification of the immune response. Recently, antigen-specific therapy has been developed as a novel tool for the treatment of experimental autoimmune disorders (Fig. 27–11). This is based on the observation that TCR reengagement at an appropriate stage of the T-cell cycle eventually leads to T-cell apoptosis by activationinduced cell death.94 In EAN, both actively induced and adoptively transferred, the intravenous administration of recombinant P2 prevented and ameliorated disease in the Lewis rat, and this was associated with a profound increase in T-cell apoptosis in peripheral immune organs as well as in the sciatic nerve.150 TNF-␣ has been identified as a principal mediator of the beneficial effects of antigen therapy in situ.151 The intracellular signaling mechanisms that are associated with high-dose antigen therapy have not yet been elucidated.

More than a decade ago it was shown that activated T lymphocytes are capable of traversing endothelial barriers that separate the nervous system from the blood stream.152 CAMs are essential for effective antigen presentation and also mediate cell-cell communication in vivo and in vitro in rodents and humans. Interaction of immune cells or extracellular matrix via these molecules profoundly influences a variety of biologic functions, including homing and transmigration of immune cells through the BNB. In the acute phase of EAN, upregulation of CAMs such as ICAM-1 and VCAM-1 at the endothelial tight junctions of lesion-associated blood vessels (i.e., part of the BNB) parallels disease activity and parenchymal infiltration.24 This points to a common pathogenic role of adhesion molecules in the initiation of tissue inflammation in EAN. This notion was reinforced by therapeutic manipulation with mAbs directed at these molecules in vivo. First, the involvement of ICAM-1 in transendothelial migration was delineated in EAN,3 followed by that of LFA-14 and L-selectin.2 This was later extended to VLA-4 and its ligand VCAM-1. Blockade of VLA-4 (␣4␤1) by mAbs in vivo was effective in

Effector Phase: T-Cell–Directed Treatments

626

Neuroimmunology of the Peripheral Nervous System

DISEASE COURSE

WEIGHT

10 200

9

6 5 4 3 2 1

160 140 120

Daily i.v. treatment

Daily i.v. treatment

100

0 0

A

180

Injection

Mean weight ± SD

7

Injection

Mean score ± SD

8

10

0

20

B

Time (days)

10

20

Time (days)

Ovalbumin control group 2x 500 μg recombinant P2 FIGURE 27–11 Preventive antigen-specific therapy of AT-EAN. In groups of six rats, AT-EAN was induced by intravenous transfer of neuritogenic P2-specific T cells. Starting on day 1, rats received either the specific antigen P2 (filled squares) or ovalbumin as control antigen (filled circles) by daily intravenous injection. Graphs show the typical disease course (A) and body weight (B); note that in B the y axis starts at 100 g. In the control group, disease started at day 3, reaching its maximum at day 7 with paraparesis (A) and 15% weight loss (B). In contrast, the group receiving high-dose antigen was completely protected. See Color Plate

To further modulate the P2-directed immune response in the antigen-specific treatment paradigm, a polypeptide oligomer (P2-16mer) harboring 16 repeats of the neuritogenic epitope (aa 58 to 73) of the P2 protein interspersed by peptides as spacers was designed and synthesized. T-cell epitope oligomers of this type have been shown to be particularly effective in the specific suppression of EAE by the mechanism of inducing high-zone tolerance. In contrast to other broadly immunosuppressive mechanisms, this new therapeutic approach for prevention and therapy of autoimmune diseases aimed more specifically at the antigenspecific elimination of autoreactive T cells. Not only were these “designer polypeptides” useful for prevention and therapy but, more importantly, a tolerizing effect was achieved (labeled “vaccination”) that lasted at least 4 months.127,151

Effector Phase: Macrophage- and Antibody-Directed Treatments In parallel with local T-cell activation, secretion of cytokines upregulates local chemokine expression, which leads to attraction and local activation of monocytic cells, thus resulting in amplification of the immune response52 (see Immunologic Principles above; see also Chapter 98).

Macrophages, the principal APCs in the PNS, are armed with a repertoire of secreted proinflammatory mediators, including free reactive oxygen metabolites (“oxygen radicals”). They also release free NO, which may further augment tissue damage and block impulse propagation, because NO radicals were shown to interfere with conduction properties of nerve fibers and may induce permanent nerve damage if applied ex vivo.112 In addition, macrophages bind via Fc receptors to sites of immunoglobulin and complement deposition. Thus macrophages not only serve as phagocytic cells to clear tissue debris, but in addition receive stimulatory signals that augment their proinflammatory function and thereby initiate a vicious circle of immune activation. These effector functions of macrophages can be inhibited by administration of steroids, cyclooxygenase inhibitors, or eicosanoids, ameliorating the disease course of EAN.49,51,52 (Fig. 27–12). Thus macrophage-derived proinflammatory arachidonic acid metabolites significantly contribute to functional and tissue damage in EAN. To more selectively eliminate phagocytic cells, liposomeencapsulated dichlormethylene diphosphate was studied in active EAN and AT-EAN.68 Efficient suppression of active EAN and of AT-EAN was observed, thus confirming the decisive role of macrophages as effector cells during EAN. Similar to experimental studies in EAE,124

627

Experimental Autoimmune Neuritis FIGURE 27–12 Serial motor nerve conduction studies in EAN: therapeutic effects of oxygen radical scavengers. Representative redrawn original recording of compound muscle action potentials obtained from the small foot muscles after electrical stimulation of the sciatic nerve distally (above the ankle; upper trace) and proximally (sciatic notch; lower trace). Lewis rats were immunized with 3 mg of myelin and treated with either catalase or superoxide dismutase (SOD) twice daily from day 7 after immunization until sacrifice. F-wave failure, conduction block, and conduction slowing evolve over time in shamtreated EAN rats, designated “myelin,” whereas only minor changes are seen in enzyme-treated animals. Arrows indicate stimulation artifacts. (From Hartung, H. P., Schäfer, B., Heininger, K., and Toyka, K. V.: Suppression of experimental autoimmune neuritis by the oxygen radical scavengers superoxide dismutase and catalase. Ann. Neurol. 23:453, 1988, with permission.)

Myelin

+ Catalase

+ Sod

Day 1

Day 13

Day 21

liposome-encapsulated glucocorticosteroids may also be of superior therapeutic efficacy in EAN because they specifically target inflammatory phagocytes and were able to reduce macrophage infiltration. Along with other proinflammatory mediators, NO may also be involved in orchestrating the local immune reaction. Interestingly, not only T cells but also glial cells as facultative APCs can produce NO in response to proinflammatory cytokines.38 Because NO seems to be involved in a complex network of pro- and anti-inflammatory signaling, it is not surprising that the net effect of NO synthase inhibition is often unpredictable and such inhibition has indeed been shown to ameliorate active EAN.72,163 As important players in the effector phase, antibodies and complement are other targets of experimental treatments. Plasma exchange, which is an established treatment in human inflammatory neuropathies, has also been shown to be therapeutically effective in EAN.56 Similar findings were obtained by complement depletion via cobra venom factor, which reduced inflammation and demyelination in EAN.143 Treatment of ongoing EAN was also achieved through neutralization of hyperimmune EAN serum containing polyvalent antibodies by plasma infusions from healthy animals,55 whereas in rat EAN, human intravenous immunoglobulin was only marginally 29 or not at all25 effective. In GBS, serum IgG antibodies directed to gangliosides (see Chapter 26) have been identified that functionally and sometimes irreversibly block neuromuscular transmission ex vivo.13 With EAN serum, this block is fulminant and irreversible.14 This blockade can be neutralized by polyvalent human IgG and its Fab fraction.12

Termination of the Immune Response The termination of ongoing PNS inflammation can be induced by continuous downregulation of the vicious circle

5mV 1ms

of the amplification-effector pathways. One such method is silencing of macrophages (see Effector Phase: Macrophageand Antibody-Directed Treatments above); another is apoptosis of autoimmune T cells, which occurs during the natural disease course of EAN.159 Glucocorticosteroids are among the most efficient antiinflammatory drugs and can eradicate infiltrating T cells through augmenting apoptosis (Fig. 27–13). Based on the use of high-dose glucocorticosteroids in human disease, steroid “pulse therapy” (10 mg/kg body weight) in rat EAN led to a four- to fivefold increase of T-cell apoptosis in the inflamed sciatic nerve.158 It is assumed that, at high doses, glucocorticosteroids can directly promote cell death, and this is thought to reflect a nongenomic, physicochemical action of glucocorticosteroids.36 Thus high-dose steroid hormones may accelerate the termination of the inflammatory assault and thereby mitigate the effector phase of the disease.

Survival Factors in the PNS As a counterbalance to the autoimmune attack, endogenous tissue components may serve to attenuate the proinflammatory and destructive properties of activated complement factors. For example, Schwann cells are endowed on their membrane with a number of regulatory complement proteins such as CR1 (CD35), decay-accelerating factor CD55, and membrane cofactor proteins CD46 and CD59.35 Another possibility is through the expression of neurotrophic factors, which have been shown to protect glial cells from the destructive effect of TNF-␣.98 To further support survival of local glial cells that undergo apoptosis during EAN,149 therapeutic studies with neurotrophins were undertaken. Neither administration of ciliary neurotrophic factor39 nor of brain cell–derived neurotrophic factor28 ameliorated the disease course. Part of the treatment failure may

628

Neuroimmunology of the Peripheral Nervous System

obstacles not the least because these trials are very costly and time consuming. There is an obvious bias toward newly developed drugs and compounds provided by the pharmaceutical industry that are successfully explored in experimental models, and then enter the stage of controlled clinical trials. To test existing drugs, in particular in combination with already available effective treatments, investigator-initiated trials are needed, which require funding by institutional grants and larger international consortiums such as the Immune Neuropathies Cause and Treatment (INCAT) Group in Europe or the intramurally funded trials at the National Institutes of Health in the United States.

A

REFERENCES

B FIGURE 27–13 T-cell apoptosis in EAN. A, Inflammatory infiltrate in the sciatic nerve of an EAN rat after glucocorticosteroid therapy. T cells are labeled in red, and apoptotic cells appear black. B, At higher magnification some apoptotic cells are still labeled red for a T-cell membrane antigen (arrows). At late stages of T-cell apoptosis, only the black signal for fragmented DNA is left, and the membrane signal has been lost (arrowheads). Bar in B: 20 ␮m. See Color Plate

be due to reduced bioavailability of the neurotrophin at the inflammatory site.22 In analogy to the CNS,74 immune cells themselves could participate in the downregulation or termination of the inflammatory process, by means of “protective” immunity. One mechanism that may be involved in this process is local release of neurotrophic factors.

Future Perspectives for Treatment In conclusion, the field of modern immunotherapy in inflammatory neuropathies is rapidly moving forward, especially as we recognize targets of T-cell signaling and as newly developed designer molecules become available as more specific tools. The enlarging gap between our deeper understanding of molecular mechanisms in the animal models and the increased demands on the scientific quality standards of treatment trials pose increasing practical

1. Abromson-Leeman, S., Bronson, R., and Dorf, M. E.: Experimental autoimmune peripheral neuritis induced in BALB/c mice by myelin basic protein-specific T cell clones. J. Exp. Med. 182:587, 1995. 2. Archelos, J. J., Fortwangler, T., and Hartung, H. P.: Attenuation of experimental autoimmune neuritis in the Lewis rat by treatment with an antibody to L-selectin. Neurosci. Lett. 235:9, 1997. 3. Archelos, J. J., Maurer, M., Jung, S., et al.: Suppression of experimental allergic neuritis by an antibody to the intracellular adhesion molecule ICAM-1. Brain 116:1043, 1993. 4. Archelos, J. J., Maurer, M., Jung, S., et al.: Inhibition of experimental autoimmune neuritis by an antibody to the lymphocyte function-associated antigen-1. Lab. Invest. 70:667, 1994. 5. Archelos, J. J., Previtali, S. C., and Hartung, H. P.: The role of integrins in immune-mediated diseases of the nervous system. Trends Neurosci. 22:30, 1999. 6. Bai, X. F., Zhu, J., Zhang, G. X., et al.: IL-10 suppresses experimental autoimmune neuritis and down- regulates TH1-type immune responses. Clin. Immunol. Immunopathol. 83:117, 1997. 7. Batocchi, A. P., Rotondi, M., Caggiula, M., et al.: Leptin as a marker of multiple sclerosis activity in patients treated with interferon-beta. J. Neuroimmunol. 139:150, 2003. 8. Ben-Nun, A., Wekerle, H., and Cohen, I. R.: The rapid isolation of clonable antigen-specific T lymphocyte lines capable of mediating autoimmune encephalomyelitis. Eur. J. Immunol. 11:195, 1981. 9. Brosnan, C. F., Lyman, W. D., Tansey, F. A., and Carter, T. H.: Quantitation of mast cells in experimental allergic neuritis. J. Neuropathol. Exp. Neurol. 44:196, 1985. 10. Brosnan, C. F., and Tansey, F. A.: Delayed onset of experimental allergic neuritis in rats treated with reserpine. J. Neuropathol. Exp. Neurol. 43:84, 1984. 11. Buchwald, B., Ahangari, R., and Toyka, K. V.: Differential blocking effects of the monoclonal anti-GQ1b IgM antibody and alpha-latrotoxin in the absence of complement at the mouse neuromuscular junction. Neurosci. Lett. 334:25, 2002.

Experimental Autoimmune Neuritis 12. Buchwald, B., Ahangari, R., Weishaupt, A., and Toyka, K. V.: Intravenous immunoglobulins neutralize blocking antibodies in Guillain-Barre syndrome. Ann. Neurol. 51:673, 2002. 13. Buchwald, B., Toyka, K. V., Zielasek, J., et al.: Neuromuscular blockade by IgG antibodies from patients with Guillain-Barre syndrome: a macro-patch-clamp study. Ann. Neurol. 44:913, 1998. 14. Buchwald, B., Weishaupt, A., Toyka, K. V., and Dudel, J.: Immunoglobulin G from a patient with Miller-Fisher syndrome rapidly and reversibly depresses evoked quantal release at the neuromuscular junction of mice. Neurosci. Lett. 201:263, 1995. 15. Calida, D. M., Kermlev, S. G., Fujioka, T., et al.: Experimental allergic neuritis in the SJL/J mouse: induction of severe and reproducible disease with bovine peripheral nerve myelin and pertussis toxin with or without interleukin-12. J. Neuroimmunol. 107:1, 2000. 16. Campbell, J. J., Hedrick, J., Zlotnik, A., et al.: Chemokines and the arrest of lymphocytes rolling under flow conditions. Science 279:381, 1998. 17. Carroll, M. C., and Prodeus, A. P.: Linkages of innate and adaptive immunity. Curr. Opin. Immunol. 10:36, 1998. 18. Choi, Y. C., and Dalakas, M. C.: Expression of matrix metalloproteinases in the muscle of patients with inflammatory myopathies. Neurology 54:65, 2000. 19. Craggs, R. I., Brosnan, J. V., King, R. H., and Thomas, P. K.: Chronic relapsing experimental allergic neuritis in Lewis rats: effects of thymectomy and splenectomy. Acta Neuropathol. (Berl.) 70:22, 1986. 20. Dalakas, M. C., and Quarles, R. H.: Autoimmune ataxic neuropathies (sensory ganglionopathies): are glycolipids the responsible autoantigens? Ann. Neurol. 39:419, 1996. 21. Di Marco, R., Khademi, M., Wallstrom, E., et al.: Amelioration of experimental allergic neuritis by sodium fusidate (fusidin): suppression of IFN-gamma and TNFalpha and enhancement of IL-10. J Autoimmun. 13:187, 1999. 22. Dittrich, F., Thoenen, H., and Sendtner, M.: Ciliary neurotrophic factor: pharmacokinetics and acute-phase response in the rat. Ann. Neurol. 35:151, 1994. 23. Enders, U., Karch, H., Toyka, K. V., et al.: The spectrum of immune responses to Campylobacter jejuni and glycoconjugates in Guillain-Barre syndrome and in other neuroimmunological disorders. Ann. Neurol. 34:136, 1993. 24. Enders, U., Lobb, R., Pepinsky, R. B., et al.: The role of the very late antigen-4 (VLA-4) and its counterligand vascular cell adhesion molecule-1 (VCAM-1) in the pathogenesis of experimental autoimmune neuritis (EAN) of the Lewis rat. Brain 121:1257, 1998. 25. Enders, U., Toyka, K. V., Hartung, H. P., and Gold, R.: Failure of intravenous immunoglobulin (IVIg) therapy in experimental autoimmune neuritis (EAN) of the Lewis rat. J. Neuroimmunol. 76:112, 1997. 26. Eylar, E. H., Toro Goyco, E., Kessler, M. J., and Szymanska, I.: Induction of allergic neuritis in rhesus monkeys. J. Neuroimmunol. 3:91, 1982. 27. Feasby, T. E., Gilbert, J. J., Hahn, A. F., and Neilson, M.: Complement depletion suppresses Lewis rat experimental allergic neuritis. Brain Res. 419:97, 1987.

629

28. Felts, P. A., Smith, K. J., Gregson, N. A., and Hughes, R. A. C.: Brain-derived neurotrophic factor in experimental autoimmune neuritis. J. Neuroimmunol. 124:62, 2002. 29. Gabriel, C. M., Gregson, N. A., Redford, E. J., et al.: Human immunoglobulin ameliorates rat experimental autoimmune neuritis. Brain 120:1533, 1997. 30. Gabriel, C. M., Hughes, R. A. C., Moore, S. E., et al.: Induction of experimental autoimmune neuritis with peripheral myelin protein-22. Brain 121:1895, 1998. 31. Gaupp, S., Hartung, H. P., Toyka, K., and Jung, S.: Modulation of experimental autoimmune neuritis in Lewis rats by oral application of myelin antigens. J. Neuroimmunol. 79:129, 1997. 32. Gearing, A. J. H., and Newman, W.: Circulating adhesion molecules in disease. Immunol. Today 14:506, 1993. 33. Giese, K. P., Martini, R., Lemke, G., et al.: Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71:565, 1992. 34. Gillen, C., Jander, S., and Stoll, G.: Sequential expression of mRNA for proinflammatory cytokines and interleukin-10 in the rat peripheral nervous system: comparison between immune-mediated demyelination and wallerian degeneration. J. Neurosci. Res. 51:489, 1998. 35. Gold, R., Archelos, J. J., and Hartung, H. P.: Mechanisms of immune regulation in the peripheral nervous system. Brain Pathol. 9:343, 1999. 36. Gold, R., Buttgereit, F., and Toyka, K. V.: Mechanism of action of glucocorticosteroid hormones: possible implications for therapy of neuroimmunological disorders. J. Neuroimmunol. 117:1, 2001. 37. Gold, R., Toyka, K. V., and Hartung, H. P.: Synergistic effect of IFN-gamma and TNF-alpha on expression of immune molecules and antigen presentation by Schwann cells. Cell. Immunol. 165:65, 1995. 38. Gold, R., Zielasek, J., Kiefer, R., et al.: Secretion of nitrite by Schwann cells and its effect on T-cell activation in vitro. Cell. Immunol. 168:69, 1996. 39. Gold, R., Zielasek, J., Schroder, J. M., et al.: Treatment with ciliary neurotrophic factor does not improve regeneration in experimental autoimmune neuritis of the Lewis rat. Muscle Nerve 19:1177, 1996. 40. Gregorian, S. K., Lee, W. P., Beck, L. S., et al.: Regulation of experimental autoimmune neuritis by transforming growth factor-beta 1. Cell. Immunol. 156:102, 1994. 41. Griffin, J. W., Stoll, G., Li, C. Y., et al.: Macrophage responses in inflammatory demyelinating neuropathies. Ann. Neurol. 27(Suppl.):S64, 1990. 42. Hadden, R. D. M., Gregson, N. A., Gold, R., et al.: Guillain-Barre syndrome serum and anti-Campylobacter antibody do not exacerbate experimental autoimmune neuritis. J. Neuroimmunol. 119:306, 2001. 43. Hadden, R. D. M., Gregson, N. A., Gold, R., et al.: Accumulation of immunoglobulin across the ‘blood-nerve barrier’ in spinal roots in adoptive transfer experimental autoimmune neuritis. Neuropathol. Appl. Neurobiol. 28:489, 2002. 44. Hafer-Macko, C., Hsieh, S. T., Li, C. Y., et al.: Acute motor axonal neuropathy: an antibody-mediated attack on axolemma. Ann. Neurol. 40:635, 1996.

630

Neuroimmunology of the Peripheral Nervous System

45. Hahn, A. F., Feasby, T. E., Steele, A., et al.: Demyelination and axonal degeneration in Lewis rat experimental allergic neuritis depend on the myelin dosage. Lab. Invest. 59:115, 1988. 46. Hartung, H. P., Heininger, K., Schafer, B., and Toyka, K. V.: Substance P and astrocytes: stimulation of the cyclooxygenase pathway of arachidonic acid metabolism. FASEB J. 2:48, 1988. 47. Hartung, H. P., Heininger, K., Schäfer, B., et al.: Immune mechanisms in inflammatory polyneuropathy. Ann. N. Y. Acad. Sci. 540:122, 1988. 48. Hartung, H. P., and Kieseier, B. C.: The role of matrix metalloproteinases in autoimmune damage to the central and peripheral nervous system. J. Neuroimmunol. 107:140, 2000. 49. Hartung, H. P., Pollard, J. D., Harvey, G. K., and Toyka, K. V.: Immunopathogenesis and treatment of the Guillain-Barré syndrome. Part I. Muscle Nerve 18:137, 1995. 50. Hartung, H. P., Schäfer, B., Diamantstein, T., et al.: Suppression of P2-T cell line-mediated experimental autoimmune neuritis by interleukin-2 receptor targeted monoclonal antibody ART 18. Brain Res. 489:120, 1989. 51. Hartung, H. P., Schäfer, B., Heininger, K., and Toyka, K. V.: Suppression of experimental autoimmune neuritis by the oxygen radical scavengers superoxide dismutase and catalase. Ann. Neurol. 23:453, 1988. 52. Hartung, H. P., Schäfer, B., Heininger, K., et al.: The role of macrophages and eicosanoids in the pathogenesis of experimental allergic neuritis: serial clinical, electrophysiological, biochemical and morphological observations. Brain 111:1039, 1988. 53. Hartung, H. P., Schäfer, B., Van der Meide, P. H., et al.: The role of interferon-gamma in the pathogenesis of experimental autoimmune disease of the peripheral nervous system. Ann. Neurol. 27:247, 1990. 54. Harvey, G. K., Gold, R., Toyka, K. V., and Hartung, H. P.: Nonneural-specific T lymphocytes can orchestrate inflammatory peripheral neuropathy. Brain 118:1263, 1995. 55. Harvey, G. K., Pollard, J. D., Schindhelm, K., and McLeod, J. G.: Experimental allergic neuritis: effect of plasma infusions. Clin. Exp. Immunol. 76:452, 1989. 56. Harvey, G. K., Schindhelm, K., Antony, J. H., and Pollard, J. D.: Membrane plasma exchange in experimental allergic neuritis: effect on antibody levels and clinical course. J. Neurol. Sci. 88:207, 1988. 57. Heininger, K., Fierz, W., Schafer, B., et al.: Electrophysiological investigations in adoptively transferred experimental autoimmune encephalomyelitis in the Lewis rat. Brain 112:537, 1989. 58. Heininger, K., Schäfer, B., Hartung, H. P., et al.: The role of macrophages in experimental autoimmune neuritis induced by a P2-specific T-cell line. Ann. Neurol. 23:326, 1988. 59. Heininger, K., Stoll, G., Linington, C., et al.: Conduction failure and nerve conduction slowing in experimental allergic neuritis induced by P2-specific T-cell lines. Ann. Neurol. 19:44, 1986. 60. Hila, S., Soane, L., and Koski, C. L.: Sublytic C5b-9stimulated Schwann cell survival through PI 3-kinasemediated phosphorylation of BAD. Glia 36:58, 2001.

61. Hughes, P. M., Wells, G. M. A., Clements, J. M., et al.: Matrix metalloproteinase expression during experimental autoimmune neuritis. Brain 121:481, 1998. 62. Huseby, E. S., Sather, B., Huseby, P. G., and Goverman, J.: Age-dependent T cell tolerance and autoimmunity to myelin basic protein. Immunity 14:471, 2001. 63. Izumo, S., Linington, C., Wekerle, H., and Meyermann, R.: Morphologic study on experimental allergic neuritis mediated by T cell line specific for bovine P2 protein in Lewis rats. Lab. Invest. 53:209, 1985. 64. Jander, S., Pohl, J., Gillen, C., and Stoll, G.: Differential expression of interleukin-10 mRNA in wallerian degeneration and immune-mediated inflammation of the rat peripheral nervous system. J. Neurosci. Res. 43:254, 1996. 65. Jander, S., and Stoll, G.: Interleukin-18 is induced in acute inflammatory demyelinating polyneuropathy. J. Neuroimmunol. 114:253, 2001. 66. Janeway, C. A. J.: The immune system evolved to discriminate infectious nonself from noninfectious self. Immunol. Today 13:11, 1992. 67. Jung, S., Gaupp, S., Hartung, H. P., and Toyka, K. V.: Oral tolerance in experimental autoimmune neuritis (EAN) of the Lewis rat – II. Adjuvant effects and bystander suppression in P2 peptide-induced EAN. J. Neuroimmunol. 116:21, 2001. 68. Jung, S., Huitinga, I., Schmidt, B., et al.: Selective elimination of macrophages by dichlormethylene diphosphonatecontaining liposomes suppresses experimental autoimmune neuritis. J. Neurol. Sci. 119:195, 1993. 69. Jung, S., Kramer, S., Schluesener, H. J., et al.: Prevention and therapy of experimental autoimmune neuritis by an antibody against T cell receptors-alpha/beta. J. Immunol. 148:3768, 1992. 70. Jung, S., Toyka, K., and Hartung, H. P.: T cell directed immunotherapy of inflammatory demyelination in the peripheral nervous system—potent suppression of the effector phase of experimental autoimmune neuritis by anti-CD2 antibodies. Brain 119:1079, 1996. 71. Kadlubowski, M., and Hughes, R. A. C.: Identification of the neuritogen for experimental allergic neuritis. Nature 277:140, 1979. 72. Kahl, K. G., Zielasek, J., Uttenthal, L. O., et al.: Protective role of the cytokine-inducible isoform of nitric oxide synthase induction and nitrosative stress in experimental autoimmune encephalomyelitis of the DA rat. J. Neurosci. Res. 73:198, 2003. 73. Kamradt, T., and Mitchison, N. A.: Tolerance and autoimmunity. N. Engl. J. Med. 334:655, 2001. 74. Kerschensteiner, M., Gallmeier, E., Behrens, L., et al.: Activated human T cells, B cells, and monocytes produce brain-derived neurotrophic factor in vitro and in inflammatory brain lesions: a neuroprotective role of inflammation? J. Exp. Med. 189:865, 1999. 75. Kiefer, R., Funa, K., Schweitzer, T., et al.: Transforming growth factor-beta 1 in experimental autoimmune neuritis: cellular localization and time course. Am. J. Pathol. 148:211, 1996. 76. Kiefer, R., Kieseier, B. C., Brück, W., et al.: Macrophage differentiation antigens in acute and chronic autoimmune polyneuropathies. Brain 121:469, 1998.

Experimental Autoimmune Neuritis 77. Kiefer, R., Kieseier, B. C., Stoll, G., and Hartung, H. P.: The role of macrophages in immune-mediated damage to the peripheral nervous system. Prog. Neurobiol. 64:109, 2001. 78. Kieseier, B. C., Clements, J. M., Pischel, H. B., et al.: Matrix metalloproteinases MMP-9 and MMP-7 are expressed in experimental autoimmune neuritis and the Guillain-Barré syndrome. Ann. Neurol. 43:427, 1998. 79. Kieseier, B. C., Seifert, T., Giovannoni, G., and Hartung, H. P.: Matrix metalloproteinases in inflammatory demyelination— targets for treatment. Neurology 53:20, 1999. 80. Kieseier, B. C., Storch, M. K., Archelos, J. J., et al.: Effector pathways in immune mediated central nervous system demyelination. Curr. Opin. Neurol. 12:323, 1999. 81. Kieseier, B. C., Tani, M., Mahad, D., et al.: Chemokines and chemokine receptors in inflammatory demyelinating neuropathies: a central role for IP-10. Brain 125:823, 2002. 82. King, R. H. M., Craggs, R. I., Cross, M. L. P., and Thomas, P. K.: Effects of glucocorticoids on experimental allergic neuritis. Exp. Neurol. 87:9, 1985. 83. Kobsar, I., Berghoff, M., Samsam, M., et al.: Preserved myelin integrity and reduced axonopathy in connexin32-deficient mice lacking the recombination activating gene-1. Brain 126:804, 2003. 84. Koehler, N. K., Martin, R., and Wietholter, H.: The antibody repertoire in experimental allergic neuritis: evidence for PMP-22 as a novel neuritogen. J. Neuroimmunol. 71:179, 1996. 85. Konat, G. W., and Wiggins, R. C.: Effect of reactive oxygen species on myelin membrane proteins. J. Neurochem. 45:1113, 1985. 86. Kyewski, B., Derbinski, J., Gotter, J., and Klein, L.: Promiscuous gene expression and central T-cell tolerance: more than meets the eye. Trends Immunol. 23:364, 2002. 87. Lafontaine, S., Rasminsky, M., Saida, T., and Sumner, A. J.: Conduction block in rat myelinated fibres following acute exposure to anti-galactocerebroside serum. J. Physiol. (Lond.) 323:287, 1982. 88. Lampert, P. W.: Mechanism of demyelination in experimental allergic neuritis: electron microscopic studies. Lab. Invest. 20:127, 1969. 89. Lannes-Viera, J., Goudable, B., Drexler, K., et al.: Encephalitogenic, myelin basic protein-specific T cells from naive rat thymus: preferential use of the T cell receptor gene V beta 8.2 and expression of the CD4⫺CD8⫺ phenotype. Eur. J. Immunol. 25:611, 1995. 90. Lassmann, H., Fierz, W., Neuchrist, C., and Meyermann, R.: Chronic relapsing experimental allergic neuritis induced by repeated transfer of P2-protein reactive T cell lines. Brain 114:429, 1991. 91. Lehmann, P. V., Forsthuber, T., Miller, A., and Sercarz, E. E.: Spreading of T-cell autoimmunity to cryptic determinants of an autoantigen. Nature 358:155, 1992. 92. Leppert, D., Hughes, P., Huber, S., et al.: Matrix metalloproteinase upregulation in chronic inflammatory demyelinating polyneuropathy and nonsystemic vasculitic neuropathy. Neurology 53:62, 1999. 93. Leussink, V. I., Zettl, U. K., Jander, S., et al.: Blockade of signaling via the very late antigen (VLA-4) and its counterligand vascular cell adhesion molecule-1 (VCAM-1) causes increased T cell apoptosis in experimental autoimmune neuritis. Acta Neuropathol. (Berl.) 103:131, 2002.

631

94. Liblau, R., Tisch, R., Bercovici, N., and McDevitt, H. O.: Systemic antigen in the treatment of T-cell-mediated autoimmune diseases. Immunol. Today 18:599, 1997. 95. Linington, C., Izumo, S., Suzuki, M., et al.: A permanent rat T cell line that mediates experimental allergic neuritis in the Lewis rat in vivo. J. Immunol. 133:1946, 1984. 96. Linington, C., Lassmann, H., Ozawa, K., et al.: Cell adhesion molecules of the immunoglobulin supergene family as tissue-specific autoantigens: induction of experimental allergic neuritis (EAN) by P0 protein-specific T cell lines. Eur. J. Immunol. 22:1813, 1992. 97. Linington, C., Mann, A., Izumo, S., et al.: Induction of experimental allergic neuritis in the BN rat: P2 proteinspecific T cells overcome resistance to actively induced disease. J. Immunol. 137:3826, 1986. 98. Louis, J. C., Magal, E., Takayama, S., and Varon, S.: CNTF protection of oligodendrocytes against natural and tumor necrosis factor-induced death. Science 259:689, 1993. 99. McCombe, P. A., van der Kreek, S. A., and Pender, M. P.: Neuropathological findings in chronic relapsing experimental allergic neuritis induced in the Lewis rat by inoculation with intradural root myelin and treatment with low dose cyclosporin A. Neuropathol. Appl. Neurobiol. 18:171, 1992. 100. Miyamoto, K., Miyake, S., Schachner, M., and Yamamura, T.: Heterozygous null mutation of myelin P0 protein enhances susceptibility to autoimmune neuritis targeting P0 peptide. Eur. J. Immunol. 33:656, 2003. 101. Moser, B., and Loetscher, P.: Lymphocyte traffic control by chemokines. Nat. Immunol. 2:123, 2001. 102. Mueller, M., Wacker, K., Ringelstein, E. B., et al.: Rapid response of identified resident endoneurial macrophages to nerve injury. Am. J. Pathol. 159:2187, 2001. 103. Nemazee, D.: Receptor selection in B and T lymphocytes. Annu. Rev. Immunol. 18:19, 2000. 104. Ochsenbein, A. F., and Zinkernagel, R. M.: Natural antibodies and complement link innate and acquired immunity. Immunol. Today 21:624, 2000. 105. Olee, T., Powell, H. C., and Brostoff, S. W.: New minimum length requirement for a T cell epitope for experimental allergic neuritis. J. Neuroimmunol. 27:187, 1990. 106. Olee, T., Weise, M., Powers, J., and Brostoff, S.: A T cell epitope for experimental allergic neuritis is an amphipathic alpha-helical structure. J. Neuroimmunol. 21:235, 1989. 107. Pelidou, S. H., Deretzi, G., Zou, L. P., et al.: Inflammation and severe demyelination in the peripheral nervous system induced by the intraneural injection of recombinant mouse interleukin-12. Scand. J. Immunol. 50:39, 1999. 108. Pender, M. P., Stanley, G. P., Yoong, G., and Nguyen, K. B.: The neuropathology of chronic relapsing experimental allergic encephalomyelitis induced in the Lewis rat by inoculation with whole spinal cord and treatment with cyclosporin A. Acta Neuropathol. (Berl.) 80:172, 1990. 109. Powell, H. C., Olee, T., Brostoff, S. W., and Mizisin, A. P.: Comparative histopathology of experimental allergic neuritis induced with minimum length neuritogenic peptides by adoptive transfer with sensitized cells or direct sensitization. J. Neuropathol. Exp. Neurol. 50:658, 1991. 110. Rathmell, J. C., Townsend, S. E., Xu, J. C., et al.: Expansion or elimination of B cells in vivo: dual roles for. Cell 87:319, 1996.

632

Neuroimmunology of the Peripheral Nervous System

111. Redford, E. J., Hall, S. M., and Smith, K. J.: Vascular changes and demyelination induced by the intraneural injection of tumour necrosis factor. Brain 118:869, 1995. 112. Redford, E. J., Smith, K. J., Gregson, N. A., et al.: A combined inhibitor of matrix metalloproteinase activity and tumour necrosis factor-a processing attenuates experimental autoimmune neuritis. Brain 120:1895, 1997. 113. Rose, N. R., and Bona, C.: Defining criteria for autoimmune diseases (Witebsky’s postulates revisited). Immunol. Today 14:426, 1993. 114. Rosen, J. L., Brown, M. J., Hickey, W. F., and Rostami, A.: Early myelin lesions in experimental allergic neuritis. Muscle Nerve 13:629, 1990. 115. Said, G., and Hontebeyrie Joskowicz, M.: Nerve lesions induced by macrophage activation. Res. Immunol. 143:589, 1992. 116. Saida, K., Saida, T., Brown, M. J., et al.: Antiserum-mediated demyelination in vivo: a sequential study using intraneural injection of experimental allergic neuritis serum. Lab. Invest. 39:449, 1978. 117. Salomon, B., and Bluestone, J. A.: Complexities of CD28/B7: CTLA-4 costimulatory pathways in autoimmunity and transplantation. Annu. Rev. Immunol. 19:225, 2001. 118. Salomon, B., Rhee, L., Bour-Jordan, H., et al.: Development of spontaneous autoimmune peripheral polyneuropathy in B7-2-deficient NOD mice. J. Exp. Med. 194:677, 2001. 119. Schabet, M., Whitaker, J. N., Schott, K., et al.: The use of protease inhibitors in experimental allergic neuritis. J. Neuroimmunol. 31:265, 1991. 120. Schmid, C. D., Stienekemeier, M., Oehen, S., et al.: Immune deficiency in mouse models for inherited peripheral neuropathies leads to improved myelin maintenance. J. Neurosci. 20:729, 2000. 121. Schmidt, B., Stoll, G., Hartung, H. P., et al.: Macrophages but not Schwann cells express Ia antigen in experimental autoimmune neuritis. Ann. Neurol. 28:70, 1990. 122. Schmidt, B., Stoll, G., van der Meide, P., et al.: Transient cellular expression of gamma-interferon in myelin-induced and T-cell line-mediated experimental autoimmune neuritis. Brain 115:1633, 1992. 123. Schmidt, J., Elflein, K., Stienekemeier, M., et al.: Treatment and prevention of experimental autoimmune neuritis with superagonistic CD28-specific monoclonal antibodies. J. Neuroimmunol. 140:143, 2003. 124. Schmidt, J., Metselaar, J. M., Wauben, M. H., et al.: Drug targeting by long-circulating liposomal glucocorticosteroids increases therapeutic efficacy in a model of multiple sclerosis. Brain 126:1895, 2003. 125. Sharief, M. K., Ingram, D. A., and Swash, M.: Circulating tumor necrosis factor-alpha correlates with electrodiagnostic abnormalities in Guillain-Barre syndrome. Ann. Neurol. 42:68, 1997. 126. Spies, J. M., Pollard, J. D., Bonner, J. G., et al.: Synergy between antibody and P2-reactive T cells in experimental allergic neuritis. J. Neuroimmunol. 57:77, 1995. 127. Stienekemeier, M., Falk, K., Rotzschke, O., et al.: Vaccination, prevention, and treatment of experimental autoimmune neuritis (EAN) by an oligomerized T cell epitope. Proc. Natl. Acad. Sci. U. S. A. 98:13872, 2001.

128. Stoll, G., Reiners, K., Schwendemann, G., et al.: Normal myelination of regenerating peripheral nerve sprouts despite circulating antibodies to galactocerebroside in rabbits. Ann. Neurol. 19:189, 1986. 129. Stoll, G., Griffin, J. W., Li, C. Y., and Trapp, B. D.: Wallerian degeneration in the peripheral nervous system: participation of both Schwann cells and macrophages in myelin degradation. J. Neurocytol. 18:671, 1989. 130. Stoll, G., Jung, S., Jander, S., et al.: Tumor necrosis factoralpha in immune-mediated demyelination and wallerian degeneration of the rat peripheral nervous system. J. Neuroimmunol. 45:175, 1993. 131. Stoll, G., Schmidt, B., Jander, S., et al.: Presence of the terminal complement complex (C5b-9) precedes myelin degradation in immune-mediated demyelination of the rat peripheral nervous system. Ann. Neurol. 30:147, 1991. 132. Stoll, G., Schwendemann, G., Heininger, K., et al.: Relation of clinical, serological, morphological, and electrophysiological findings in galactocerebroside-induced experimental allergic neuritis. J. Neurol. Neurosurg. Psychiatry 49:258, 1986. 133. Tacke, M., Clark, G. J., Dallman, M. J., and Hünig, T.: Cellular distribution and costimulatory function of rat CD28: regulated expression during thymocyte maturation and induction of cyclosporin A sensitivity of costimulated T cell responses by phorbol ester. J. Immunol. 154:5121, 1995. 134. Tacke, M., Hanke, G., Hanke, T., and Hunig, T.: CD28mediated induction of proliferation in resting T cells in vitro and in vivo without engagement of the T cell receptor: evidence for functionally distinct forms of CD28. Eur. J. Immunol. 27:239, 1997. 135. Takeda, K., Kaisho, T., and Akira, S.: Toll-like receptors. Annu. Rev. Immunol. 21:335, 2003. 136. Tansey, F. A., and Brosnan, C. F.: Protection against experimental allergic neuritis with silica quartz dust. J. Neuroimmunol. 3:169, 1982. 137. Taylor, W. A., and Hughes, R. A.: Experimental allergic neuritis induced in SJL mice by bovine P2. J. Neuroimmunol. 8:153, 1985. 138. Toyka, K. V., and Heininger, K.: Humoral factors in peripheral nerve disease. Muscle Nerve 10:222, 1987. 139. Trinchieri, G.: Interleukin-12: a proinflammatory cytokine with immunoregulatory functions that bridge innate resistance and antigen-specific adaptive immunity. Annu. Rev. Immunol. 13:251, 1995. 140. Uncini, A., Di Muzio, A., Di Guglielmo, G., et al.: Effect of rhTNF-alpha injection into rat sciatic nerve. J. Neuroimmunol. 94:88, 1999. 141. Van Parijs, L., Biuckians, A., and Abbas, A. K.: Functional roles of Fas and Bcl-2-regulated apoptosis of T lymphocytes. J. Immunol. 160:2065, 1998. 142. Visan, L. A., Visan, A., Weishaupt, H. H., et al.: Tolerance induction by intrathymic expression of myelin P0 protein. J. Immunol. 172:1364, 2004. 143. Vriesendorp, F. J., Flynn, R. E., Malone, M. R., and Pappolla, M. A.: Systemic complement depletion reduces inflammation and demyelination in adoptive transfer experimental allergic neuritis. Acta Neuropathol. (Berl.) 95:297, 1998. 144. Vriesendorp, F. J., Flynn, R. E., Pappolla, M. A., and Koski, C. L.: Complement depletion affects demyelination and

Experimental Autoimmune Neuritis

145.

146.

147.

148.

149.

150.

151.

152.

153.

inflammation in experimental allergic neuritis. J. Neuroimmunol. 58:157, 1995. Waksman, B. H., and Adams, R. D.: Allergic neuritis: an experimental disease of rabbits induced by the injection of peripheral nervous tissue and adjuvants. J. Exp. Med. 102:213, 1955. Wauben, M. H. M., Hoedemaekers, A. C. W. E., Graus, Y. M. F., et al.: Coimmunization of MHC class II competitor peptides during experimental autoimmune myasthenia gravis induction resulted not only in a suppressed, but also in an altered immune response. Ann. N. Y. Acad. Sci. 841:338, 1998. Weerth, S., Berger, T., Lassmann, H., and Linington, C.: Encephalitogenic and neuritogenic T cell responses to the myelin-associated glycoprotein (MAG) in the Lewis rat. J. Neuroimmunol. 95:157, 1999. Weiner, H. L., Friedman, A., Miller, A., et al.: Oral tolerance: immunologic mechanisms and treatment of animal and human organ-specific autoimmune diseases by oral administration of autoantigens. Annu. Rev. Immunol. 12:809, 1994. Weishaupt, A., Bruck, W., Hartung, T., et al.: Schwann cell apoptosis in experimental autoimmune neuritis of the Lewis rat and the functional role of tumor necrosis factor-alpha. Neurosci. Lett. 306:77, 2001. Weishaupt, A., Gold, R., Gaupp, S., et al.: Antigen therapy eliminates T cell inflammation by apoptosis: effective treatment of experimental autoimmune neuritis with recombinant myelin protein P2. Proc. Natl. Acad. Sci. U. S. A. 94:1338, 1997. Weishaupt, A., Gold, R., Hartung, T., et al.: Role of TNFalpha in high-dose antigen therapy in experimental autoimmune neuritis: inhibition of TNF-alpha by neutralizing antibodies reduces T-cell apoptosis and prevents liver necrosis. J. Neuropathol. Exp. Neurol. 59:368, 2000. Wekerle, H., Linington, C., Lassmann, H., and Meyermann, R.: Cellular immune reactivity within the CNS. Trends Neurosci. 9:271, 1986. Wekerle, H., Schwab, M., Linington, C., and Meyermann, R.: Antigen presentation in the peripheral nervous system: Schwann cells present endogenous myelin autoantigens to lymphocytes. Eur. J. Immunol. 16:1551, 1986.

633

154. Westland, K., and Pollard, J. D.: Proteinase induced demyelination: an electrophysiological and histological study. J. Neurol. Sci. 82:41, 1987. 155. Whitehouse, D. J., Whitehouse, M. W., and Pearson, C. M.: Passive transfer of adjuvant-induced arthritis and allergic encephalomyelitis in rats using thoracic duct lymphocytes. Nature 224:1322, 1969. 156. Yan, W. X., Archelos, J. J., Hartung, H. P., and Pollard, J. D.: P0 protein is a target antigen in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:286, 2001. 157. Yuki, N., Yamada, M., Koga, M., et al.: Animal model of axonal Guillain-Barre syndrome induced by sensitization with GM1 ganglioside. Ann. Neurol. 49:712, 2001. 158. Zettl, U. K., Gold, R., Toyka, K. V., and Hartung, H. P.: Intravenous glucocorticosteroid treatment augments apoptosis of inflammatory T cells in experimental autoimmune neuritis (EAN) of the Lewis rat. J. Neuropathol. Exp. Neurol. 54:540, 1995. 159. Zettl, U. K., Gold, R., Toyka, K. V., and Hartung, H. P.: In situ demonstration of T cell activation and elimination in the peripheral nervous system during experimental autoimmune neuritis in the Lewis rat. Acta Neuropathol. (Berl.) 91:360, 1996. 160. Zhu, J., Bai, X. F., Mix, E., and Link, H.: Cytokine dichotomy in peripheral nervous system influences the outcome of experimental allergic neuritis: dynamics of mRNA expression for IL-1b, IL-6, IL-10, IL-12, TNF-a, TNF-b, and cytolysin. Clin. Immunol. Immunopathol. 84:85, 1997. 161. Zhu, J., Link, H., Weerth, S., et al.: The B cell repertoire in experimental allergic neuritis involves multiple myelin proteins and GM1. J. Neurol. Sci. 125:132, 1994. 162. Zhu, Y., Ljunggren, H. G., Mix, E., et al.: CD28-B7 costimulation: a critical role for initiation and development of experimental autoimmune neuritis in C57BL/6 mice. J. Neuroimmunol. 114:114, 2001. 163. Zielasek, J., Jung, S., Gold, R., et al.: Administration of nitric oxide synthase inhibitors in experimental autoimmune neuritis and experimental autoimmune encephalomyelitis. J. Neuroimmunol. 58:81, 1995.

28 Principles of Immunotherapy JOHN D. POLLARD, HANS-PETER HARTUNG, AND RICHARD A. C. HUGHES

Immunosuppressive Agents Glucocorticoids Alkylating Agents

Immunophilin Binding Agents Azathioprine Mycophenolate Mofetil

Immunotherapy for immune-mediated disorders of the peripheral nervous system (PNS) is currently under rapid development. The aim of an effective therapy is to reduce severity and duration of clinical deterioration, as in monophasic diseases such as Guillain-Barré syndrome (GBS), and to reduce frequency of relapses and to prevent disability from disease progression, as in chronic diseases such as chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). When choosing treatment with short- and/or long-term effects, benefits and risks of the therapeutic approach need to be weighed carefully in each individual patient. Unfortunately, our therapeutic armamentarium, consisting of glucocorticoids (GCs), alkylating agents, plasma exchange (PE), intravenous immunoglobulin (IVIg), and others, is not always effective enough to stop ongoing disease. In GBS, for example, approximately 50% to 60% of patients do respond to plasma exchange or high-dose IVIg, but up to 40% develop significant long-term disability (see Chapter 98). Thus there is a pressing need to provide more effective therapeutic strategies. Large efforts have been undertaken to further increase our current understanding of the pathogenetic role of various factors in immune-mediated peripheral neuropathies. New approaches including the use, for example, of more effective immunomodulators, as tested already in the animal model experimental autoimmune neuritis (EAN),67 or compounds that target the transmigration of immunocompetent cells across the blood-nerve barrier by blocking specific chemokine receptors, adhesion molecules, or proteases,4,63,64 are underway. However, whether these encouraging approaches, effective in the animal model, will

Immunomodulatory Agents Therapeutic Plasma Exchange Intravenous Immunoglobulin

translate into efficacious clinical therapy still needs to be evaluated. At present all therapeutic approaches are aimed to modulate or to suppress the immune system in order to reduce the amount of damage to the myelin sheath and/or the axon. Several lines of research are trying to identify strategies that promote remyelination and axonal regeneration. Neurotrophic factors, at least in theory, could represent a group of proteins that might promote regeneration. In a small, placebo-controlled pilot trial of 10 GBS patients, the potential role for subcutaneous brain-derived neurotrophic factor was examined. However, no difference in the outcome between the groups was observed.13 Similarly, no striking effects have been seen with different neurotrophins in EAN.38,71 Although present data are not encouraging, the favorable effects of neurotrophic factors on nerve regeneration define them as interesting molecules that might be clinically helpful in the near future. Furthermore, recent evidence indicates that immunocompetent cells can produce neurotrophins, suggesting that the inflammatory reaction might exhibit some beneficial or even neuroprotective effects.61 Nevertheless, a concept of a “neuroprotective immunity” cannot fully be established at present.59 It is unlikely that a single treatment will be effective in all patients as long as the pathogenesis of the underlying immune-mediated disease is not precisely understood. The principle of combination therapy has found utility in cancer treatment and is evolving as a strategic principle for immune-mediated peripheral neuropathies. Rationales for combining different mechanistic approaches in order to target a broader spectrum of critical immune mechanisms 635

636

Neuroimmunology of the Peripheral Nervous System

exist. Studies in animal models support the concept that different treatment strategies can act synergistically, as demonstrated with GCs and the administration of a specific antigen in EAN.117 Immunomodulation may find utility in conjunction with specific forms of therapy. As it becomes clearer which mechanistic strategies work best, more rational combinations can be designed. In chronic immune-mediated disorders a major problem for the treating neurologist is to predict the clinical course of the disease. In an individual patient it may be difficult to determine whether the chosen treatment regimen is having a therapeutic effect unless the patient improves rapidly during therapy. In some instances it becomes clear that a patient is either a responder or a nonresponder to a particular treatment; however, this clear dichotomy remains difficult to establish in many other patients. Thus surrogate markers that are linked to the underlying immunopathogenesis might be a valuable tool to objectively measure ongoing disease activity. Such markers are not fully established yet for immune-mediated neuropathies of the PNS; however, various markers are under critical investigation at present. It may be of great practical value to have useful surrogate markers in predicting the clinical response of patients at hand, because it may be helpful to stratify affected patients into different therapeutic regimens. Moreover, such markers may facilitate the administration of certain treatments at critical checkpoints during the clinical course of the disease, optimizing therapeutic effects and minimizing side effects. At present, the treating neurologist must make decisions on therapy based on clinical assessment and accumulation of disability. Moreover, he or she has to define whether a patient has responded to therapy or additional treatment should be given. This chapter discusses the available therapeutic arsenal for immunotherapy in disorders of the PNS.

IMMUNOSUPPRESSIVE AGENTS Glucocorticoids Glucocorticoids have powerful anti-inflammatory and immunosuppressive effects that have led to their use in a large number of autoimmune disorders and in transplant rejection. They are known to be associated with serious side effects that depend on dose and duration of treatment. Mechanism of Action Glucocorticoids act directly and indirectly by affecting gene transcription. They diffuse across all membranes by means of their high lipid solubility. Within the cytoplasm they bind to specific glucocorticoid receptors (GCRs). The cytosolic GCR is associated with immunophilin and heat shock proteins and, upon ligand binding, a conformational

change occurs with disassociation of the GC-GCR complex from the associated proteins. The complex is then transported from the cytoplasm to the nucleus, where it binds to specific regions of DNA known as glucocorticoid response elements (GREs). Genes with GREs can be directly influenced by GCs.8 Other genes are affected indirectly by means of GC effects on transcription factors. For instance, GC-GCR complexes may bind to nuclear factor-␬B (NF-␬B), which mediates upregulation of cytokine gene transcription. Binding of GC-GCR to NF-␬B inhibits the latter from binding to its DNA binding site, thus inhibiting pro-inflammatory gene transcription.1,8 Glucocorticoids can also alter gene transcription by upregulating the inhibitory molecule I␬B␣, which inactivates NF-␬B, binding to it before it can enter the nucleus.6 Glucocorticoids have also been shown to alter posttranscriptional mechanisms, including messenger RNA translation and protein synthesis. At the cellular level GCs have been shown to inhibit the production of a number of pro-inflammatory cytokines, including interleukin (IL)-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-13, granulocyte-macrophage colony-stimulating factor, and tumor necrosis factor (TNF)-␣. Glucocorticoids decrease the number of inflammatory cells entering a site of inflammation, and this effect is thought to be an indirect one because upregulation of those adhesion molecules involved in transmigration (e.g., vascular cell adhesion molecule-1, intracellular adhesion molecule [ICAM]-1) is dependent on cytokine production.107 Glucocorticoids inhibit phagocytosis by the downregulation of Fc␥ receptors on macrophages and monocytes5 and can inhibit the release of lysosomal enzymes from monocytes. B lymphocytes are relatively resistant to the effects of GCs, but immunoglobulin levels may be reduced by highdose therapy. Glucocorticoids can also inhibit lipid mediators of inflammation, including the cyclooxygenase metabolites prostaglandin D2, thromboxane B2, prostaglandin F2, and the leukotrienes. Thus GCs influence the movement and distribution of leukocytes, their functional properties, the synthesis and secretion of cytokines and other immune mediators, and microvascular permeability. Pharmacokinetics Glucocorticoids are lipophilic and rapidly and relatively completely absorbed when given orally, but because they may undergo presystemic metabolism, they may be incompletely bioavailable by the oral route. Intravenous steroids may be given as the water-soluble esters (e.g., succinates and phosphates), and these esters are rapidly hydrolyzed in the circulation so that the steroid is present as the corresponding free alcohol.12 Glucocorticoids bind to plasma proteins (␣-globulin), and it is the free steroid that is responsible for the biochemical and physiologic effects. Elimination occurs almost completely by way of metabolism. The II-oxo steroids cortisone and prednisone are reduced reversibly to their II-hydroxy

637

Principles of Immunotherapy

analogues, cortisol and prednisolone, their biologically active forms. The phase I metabolites undergo glucuronide or sulfate conjugation before secretion in the urine. Peak plasma concentrations occur between 1 and 2 hours following prednisone ingestion. The half-life of the GC is relatively short, in the range of 2 to 4 hours.90 Corticosteroid metabolism is altered by numerous other drugs, particularly the anticonvulsants phenytoin, carbamazepine, and phenobarbital, which induce the monooxygenase system. These agents increase the elimination of prednisone and methylprednisolone to a very significant degree.10 Ketoconazole, the macrolide antibiotics, and oral contraceptives can delay the elimination of the glucocorticoids. Adverse Effects Glucocorticoids are used therapeutically mainly for their anti-inflammatory actions. However, they have diverse metabolic effects that may constitute unwanted and adverse reactions. Because all nucleated cells have the same GCR, all cells are susceptible to the development of adverse events. The metabolic effects of steroids are mediated largely by their binding to GREs on DNA, whereas the anti-inflammatory effects result mainly from inhibitors of transcription factors. It may therefore be possible ultimately to design a GC with anti-inflammatory effects but without metabolic complications. The adverse effects of steroids are well known and include endocrine (adrenal suppression, growth suppression in children, increased weight and cushingoid appearance, hyperglycemia, glycosuria, hypokalemia, hyperlipidemia); musculoskeletal (myopathy, osteoporosis, aseptic necrosis of bone); gastrointestinal (dyspepsia, peptic ulceration); psychological (mood swings, psychosis, insomnia); ophthalmologic (cataract formation, glaucoma); cardiovascular (hypertension, arteriosclerosis); dermatologic (increased fragility of skin, ecchymoses, hirsutism, acne, striae); and immunologic (lymphopenia, susceptibility to opportunistic infection, delayed wound healing). The most common of these in patients on long-term therapy are osteoporosis, hypertension, diabetes, ulcer formation, cataracts, and increased weight. Use in Neuropathy Based on the mechanisms of action described earlier and current understanding of the pathogenesis of the various immune-mediated neuropathies, the efficacy of GCs in these neuropathies is fairly predictable. Glucocorticoids do not improve patients with GBS even when given by high-dose pulse therapy.52,54,113 Pathogenic antibodies to various gangliosides have been demonstrated at disease presentation in certain subtypes (Miller Fisher syndrome, acute motor axonal neuropathy), and these would not be altered by GCs. Similarly, neuropathies associated with immunoglobulin M (IgM) antibodies (i.e., anti–myelinassociated glycoprotein and anti-GM1 antibodies), such as

multifocal motor neuropathy (MMN), do not benefit from GCs.76 IgM antibodies tend to be T-cell independent, and B cells are relatively resistant to GCs. However, CIDP probably does respond to GCs.33,50,80 Corticosteroids are usually given in a daily oral dose but have been administered as pulsed doses of intravenous methylprednisone given over 5 days each month.68 The difference in treatment response suggests significant differences in pathogenic mechanisms in these acute and chronic disorders. The use of GCs in CIDP has largely been relegated to supportive treatment to IVIg and plasmapheresis because of their adverse affects. Similarly, even in the absence of randomized trials, experts regard GCs as an essential part of the therapy for vasculitic neuropathy,39 a disorder in which T cells are prominent in the pathogenesis. In those cases associated with systemic vasculitis, GCs are usually combined with cyclophosphamide.36 Glucocorticoids are also indicated, according to expert opinion, in high oral doses (60 to 100 mg/day) in the treatment of type 1 and 2 reactions of leprosy, which are characterized by an increase in cell-mediated immunoreactivity.95 GCs have not been found beneficial in other forms of generalized peripheral neuropathy. Anecdotal reports of their successful use in paraneoplastic neuropathy and subacute sensory neuronopathy are the exception rather than the rule, and randomized trials have not been performed. Their use in Bell’s palsy is controversial because of a lack of adequate randomized trials.96 The injection of GCs into the wrist of patients with carpal tunnel syndrome significantly ameliorates symptoms for up to 1 month, but the duration of the effect and the extent to which such injections avoid or delay the need for operation are unclear.79

Alkylating Agents Cyclophosphamide, chlorambucil, and melphalan are the most commonly used alkylating agents, the first of which to be used in clinical practice was nitrogen mustard. Although their use has greatly improved the outlook for patients with certain disorders such as vasculitis and lupus nephritis, severe toxicity has restricted their more general use. Mechanism of Action Alkylating agents act largely by cross-linking DNA and RNA and hence inhibit transcription and translation in both proliferating and resting cells. Cyclophosphamide and chlorambucil are rapidly metabolized to their active forms, phosphoramide mustard and phenylacetic acid mustard, respectively. Cyclophosphamide affects both B and T cells, and hence reduces both cell-mediated and antibodymediated immune reactions.123 The immunomodulatory effect depends on the dose and duration of treatment. Low-dose effects are attributed to deletion of suppressor cell precursors.45 Moderate doses deplete B cells without significant effects on T-cell–mediated immune responses.

638

Neuroimmunology of the Peripheral Nervous System

Higher doses inhibit T helper cells, and very high doses reduce cytotoxic T cells and natural killer cells.88 Pharmacokinetics Cyclophosphamide can be given orally or intravenously. It is metabolized in the liver to its active metabolites, and about 95% is excreted by the kidney. Chlorambucil is usually given orally and is also metabolized to its active metabolite, phenylacetic acid mustard, in the liver. Excretion is exclusively by the kidney. Adverse Effects Unwanted side effects are often dose related. Common adverse effects include alopecia, nausea, and vomiting, but the more important ones are bone marrow suppression, bladder toxicity, cancer, and infertility. Suppression of bone marrow is dose dependent. Following intravenous pulse treatment, minimum white blood cell (WBC) counts occur at between 1 and 2 weeks, and dose change is necessary if the leukocyte count is less than 1.5 ⫻ 106 WBCs/L. Regular monitoring of the WBC count is needed in patients on daily oral therapy, and the dose should be changed if the count falls below 3 ⫻ 106 WBCs/L. Hemorrhagic cystitis is more common in patients given daily oral cyclophosphamide, probably because of continued exposure to the toxic metabolite acrolein; however, its incidence has been reduced by encouraging adequate hydration. Patients given pulse intravenous therapy should be given pre- and postinfusion hydration and coadministration of mesna (2-mercaptoethane sulfonate), which has been reported to bind and neutralize the toxic metabolite acrolein.56 There is a greater risk of bladder and hematologic malignancy (leukemia and lymphoma) in patients given a total dose greater than 80 g. Monthly urinalysis is indicated during the period of therapy and at 6-month intervals thereafter. Cystoscopy is required in any patient found to have hematuria. The incidence of infertility increases with cumulative dose and patient age. The adverse effects of chlorambucil are similar to those of cyclophosphamide except that it does not produce bladder toxicity and is more prone to induce hematologic malignancy. Use in Neuropathy Despite the absence of randomized trials, we regard cyclophosphamide, given usually in association with corticosteroids, as the treatment of choice in patients with neuropathy caused by systemic vasculitis. It may be given by daily oral therapy at a dose of 2 mg/kg/day, with dose adjustments as necessary to keep the leukocyte count greater than 3.0 ⫻ 106 WBCs/L, together with prednisone 1 mg/kg/day with subsequent tapering. Hemorrhagic cystitis is less common when the drug is given by monthly intravenous therapy.56 Pulses are given in a dose of 0.5 to 1 g/m2 body surface area monthly for 6 months and thereafter at

3-month intervals for 1 to 2 years. Adequate hydration with frequent urination is particularly important for patients on daily therapy. Cyclophosphamide has also been recommended in patients with MMN. Although most of these patients respond to IVIg, they may continue to slowly deteriorate, and cyclophosphamide may allow disease arrest and even regression. Clinical improvement in association with reduction of anti-GM1 antibodies has been reported following oral cyclophosphamide.37 Pestronk and colleagues89 reported a superior response after high-dose intravenous induction therapy followed by intermittent monthly courses. A supplemental dose of oral cyclophosphamide (1 to 3 mg/kg) decreased the required frequency of IVIg in a further study.82 Randomized trials have not been performed.110 Both cyclophosphamide and chlorambucil have been used in patients with progressive neuropathy associated with IgM paraproteinemia85 but only in uncontrolled studies in small numbers of patients.76 Immunosuppression with cyclophosphamide in association with PE has been suggested by others. Cyclophosphamide was reported to benefit 80% of CIDP cases in an uncontrolled series when given with or without prednisone as pulse therapy in a dose of 1 g/m2 each month for 3 to 6 months. However, 30% of the patients relapsed and needed further treatment in 6 to 36 months.40 Once again, randomized trials have not been performed.53

Immunophilin Binding Agents The immunophilins are a large family of highly conserved, ubiquitously expressed proteins initially characterized by their ability to bind a number of immunosuppressive agents, including cyclosporin A (CsA), tacrolimus (FK506), and sirolimus (rapamycin). Three groups are recognized: cyclophilins, which bind CsA; FK506 binding proteins (FKBPs), which bind tacrolimus and sirolimus; and the parvulins. It has become clear that immunophilins bind a number of other molecules besides immunosuppressive agents, including heat shock proteins and cell cycle regulatory factors. They share the capacity to catalyze the isomerization of protein residues between their cis and trans configurations and are thought to play a role in protein folding.106 Immunophilins are currently considered to play a role in immunosuppression, protein folding, calcium homeostasis, and the regulation of RNA splicing. Mechanism of Action The immunosuppressive capacity of CsA and FK506 depends on the ability of the complex formed between CsA and its cytophilin and that between tacrolimus and its FKBP to inhibit calcineurin. Calcineurin is a calcium/calmodulin serine/threonine phosphatase that is an essential intermediate in T-cell signaling from the T-cell

639

Principles of Immunotherapy

antigen receptor.75 Calcineurin activity is necessary for transcriptional activation of a number of cytokine and other genes. Calcineurin binds to and dephosphorylates the transcriptional factor nuclear factor of activated T cell (NFAT) in the cytoplasm. Dephosphorylation of NFAT results in its translocation to the nucleus, where it binds to its target sequence on the promotor region of many genes, including the cytokine genes IL-2, IL-3, IL-4, IL-12, and TNF, resulting in their activation.93 NF-␬B is also activated by calcineurin.108 Hence, following stimulation of T cells via the T-cell receptor, a number of intracellular signaling events are initiated, including the activation of protein kinase C in addition to an increase in intracellular calcium levels. Calcium and calmodulin are recruited to activate calcineurin. In the presence of a complex of CsA with its cytophilin or tacrolimus with its FKBP, the activation of calcineurin is inhibited and hence also cytokine activation. The effects of cyclosporin are seen mainly on T cells, with B cells and macrophages apparently unaffected. Cyclosporin and calcineurin also regulate the production of transforming growth factor-␤, the increased production of which correlates with renal fibrosis and nephrotoxicity.24,62 Although sirolimus (rapamycin) resembles tacrolimus in structure and binds to the same family of immunophilins, it has a different molecular target. It has been shown to function only when bound to an FKBP in an intracellular complex that does not bind calcineurin. This target protein is termed the mammalian target of rapamycin (sirolimus) (mTOR), and is a member of the DNA PK family.44 Sirolimus is more effective than CsA in inhibiting T-cell proliferation, although it does not alter IL-2 or IL-2 receptor expression.31 Sirolimus, unlike CsA and FK506, has been shown to inhibit IL-2–dependent T-cell proliferation following binding of the cytokine to its receptor. The mTOR appears to play an important role in the control of translation and amino acid transport for a wide variety of cells. Binding of the sirolimus complex to mTOR inhibits the phosphorylation of p70 S6 kinase, the subsequent phosphorylation of S6 ribosomal protein, and protein synthesis.15 Amino acid transport and an inhibitor of translational initiation (4E-BP1) are also sensitive to sirolimus.14 Cyclosporin and tacrolimus are approved for use in organ and stem cell transplantation and sirolimus in the prevention of acute renal transplant rejection. Their use in autoimmune disorders is currently being explored, and there is already considerable experience for the use of cyclosporin. Adverse Effects The unwanted side effects of cyclosporin and tacrolimus are very similar and consist most commonly of nephrotoxicity, hypertension, and neurotoxicity. Nephrotoxicity is dose dependent, and it is important to monitor plasma CsA and serum creatinine levels during therapy. Should

there be any rise in creatinine levels, creatinine clearance should be measured. Less severe side effects include hepatotoxicity, lethargy, anorexia, hirsutism, and gum hypertrophy. Neurotoxicity may be manifested by headache, tremor, sleep disturbance, or a reversible, mainly posterior encephalopathy. These complications indicate the need for dose reduction or the drug’s discontinuation. As may be predicted from the differences described earlier between molecular targets of sirolimus and cyclosporin, the toxicity profile of sirolimus differs from that of tacrolimus and cyclosporin. Nephrotoxicity, hypertension, and neurotoxicity are not prominent but hyperlipidemia, leukopenia, and thrombocytopenia are more frequent. Use in Neuropathy There are no controlled trials of these agents in neuropathy.53,110 Cyclosporin has been used in refractory CIDP9,46 and vasculitis of nerve, but the side effect profile prevents its use as a first-line agent. The efficacy of tacrolimus and sirolimus in human autoimmune neuropathy has yet to be explored in appropriate trials. Beneficial response to tacrolimus has been reported in two patients with CIDP.2,84 Tacrolimus and sirolimus and their derivatives have been reported to stimulate nerve regeneration, a property that would be of value in the treatment of immune-mediated neuropathy.104

Azathioprine Azathioprine (AZA) was first synthesized in 1957, and has been an effective immunosuppressive agent in the treatment of autoimmune and neuroimmunologic disorders but particularly in immunomodulation of transplant recipients. Azathioprine has particular efficacy in the treatment of myasthenia gravis. Mechanism of Action Azathioprine is converted to 6-mercaptopurine (6MP), a purine analogue of hypoxanthine, and guanine, which in turn are converted to the active metabolites 6-thionosinic acid and 6-thioguanilic acid. These metabolites interfere with purine biosynthesis and reduce RNA and DNA synthesis. Azathioprine decreases T and B lymphocytes and antibody production and B cell proliferation.3,49 Natural killer cell activity and neutrophil migration are also decreased. Pharmacokinetics AZA is absorbed from the alimentary tract and distributed throughout the body, 30% being bound to plasma protein. It has a half-life of 4 to 5 hours, whereas that of 6MP is only 20 to 50 minutes. The majority of AZA is converted by xanthine oxidase to the inactive metabolite 6-thiouric acid, which is excreted by the kidney. Patients receiving treatment

640

Neuroimmunology of the Peripheral Nervous System

with allopurinol are at increased risk of toxicity because it suppresses xanthine oxidase, resulting in increased levels of AZA and its metabolites. In such patients the dose of AZA should be reduced by 50% to 70% and the blood count carefully monitored. Adverse Effects The incidence of serious adverse events with AZA is relatively low. The most frequent side effects reported in a group of 104 patients with severe myasthenia were reversible marrow depression, with leukopenia, gastrointestinal complications, infections, and transient elevation of liver enzymes.47 In 546 patients with rheumatoid arthritis, 60% of therapy terminations were due to gastrointestinal problems.103 These include nausea, vomiting, diarrhea, and abdominal pain. Hepatotoxicity manifests by cholestasis, hypersensitivity, or hepatic necrosis and has been reported in up to 10% of cases, but pancreatitis is a relatively rare event. Cessation of therapy is indicated in the presence of hepatotoxicity or pancreatitis. Bone marrow suppression, although more common, is dose dependent and usually responds to dose reduction or cessation of therapy. About 10% of the population is deficient in the enzyme thiopurine5-methyltransferase, with resultant reduced AZA metabolism and increased toxicity. One in 300 is homozygous for deficiency of this enzyme and at severe risk of major side effects from a single dose. Levels of this enzyme can now be measured, and measurement before starting treatment is helpful for selecting the correct dose.48 Mild intestinal symptoms are relatively common and can usually be minimized by dividing the dose and taking it after meals. Elevation of liver enzymes of mild degree, up to three times normal, is also commonly seen and may respond to dose reduction. Leukopenia induced by AZA increases the risk of opportunistic infections. There is an increased risk of lymphoma in renal transplant patients treated with AZA and corticosteroids, and also an increased risk of skin cancers. The latter are seen in patients treated for myasthenia gravis, but the risk of other malignancies appears to be less than in transplant patients.118 Use in Neuropathy Azathioprine is usually given in a dose of 2.5 mg/kg daily. In the treatment of myasthenia, a stepwise dose reduction to a maintenance dose of 0.8 mg/kg/day is usually advised once improvement has been maintained for 6 to 9 months. There is no information in the neuropathy literature on such changes to maintenance dosage. Complete blood counts and liver function tests should be monitored every 2 weeks for the first 8 weeks and then once every 4 weeks. If the WBC count falls below 3.0 ⫻ 106 cells/L, AZA should be discontinued for a few days and treatment introduced at a lower dose when the count rises above 3.5 ⫻ 106.

Azathioprine is widely used in patients with CIDP28 to lessen the need for IVIg and PE,42,92 although its efficacy has never been proven by randomized controlled trials. In the only controlled study reported, AZA (2.0 mg/kg) was not shown to be of additional benefit when combined with corticosteroid in a small study of CIDP.34,53

Mycophenolate Mofetil This relatively recently introduced compound was isolated from Penicillium mold cultures in 1913, but was developed for use in transplantation only in the 1980s. It appears to be marginally more effective than AZA in transplantation medicine.74 Mechanism of Action Mycophenolate mofetil is hydrolyzed in the liver to its active parent compound, mycophenolic acid, which is a noncompetitive inhibitor of inositol monophosphate dehydrogenase, an enzyme central to the synthesis of guanosine nucleotides and hence purines. Proliferating lymphocytes depend primarily on the de novo pathway for purine synthesis, whereas most other cells can use both the de novo and salvage pathways. Thus mycophenolate has a selective immunosuppressive effect on proliferating lymphocytes. Mycophenolate inhibits T-cell and B-cell proliferation to both T-cell–dependent and T-cell–independent mitogens.18 The incorporation of mannose and fucose into membrane glycoproteins such as adhesion molecules appears to be dependent on guanosine levels, which may explain the demonstrated effect of mycophenolate to reduce the adherence of leukocytes to endothelial cells. Pharmacokinetics Following absorption from the gastrointestinal tract, mycophenolate is completely hydrolyzed to the active form, mycophenolic acid. Peak serum concentration occurs at 1 hour, with a secondary peak 8 hours later. The biologic half-life is 17 hours and the bioavailability 94%. Mycophenolic acid is metabolized by the liver to mycophenolic glucuronide, which is excreted in the urine. Adverse Effects The most common side effects are gastrointestinal, including nausea, vomiting, diarrhea, cramping, and abdominal pain. Gastrointestinal hemorrhage, pancreatitis, leukopenia, and thrombocytopenia have been reported. Hepatotoxicity appears to be low compared to AZA.74 Side effects are mostly dose dependent and respond to lowering the dose or stopping the drug. From relatively brief experience with the use of this drug in rheumatoid arthritis and

Principles of Immunotherapy

transplantation, no increased incidence of carcinoma has been reported. Use in Neuropathy Mycophenolate is usually commenced in a daily divided dose of 2 g, but may be increased to 4 or 5 g. The dose should be reduced or the drug discontinued in the presence of neutropenia. Mycophenolate was shown to be slightly superior to AZA in a multicenter randomized trial in heart transplant patients in reducing mortality and acute rejection.65 However, there have been no controlled trials to date of mycophenolate in autoimmune neurologic disease, including neuropathy. Small case series have claimed both benefit and the lack of it when given to patients with CIDP.19,53,84 Blood counts should be monitored weekly for the first month, then twice monthly for 2 months, and then monthly.

IMMUNOMODULATORY AGENTS Therapeutic Plasma Exchange Therapeutic PE (plasmapheresis) was first shown to have clinical benefit in 1960 when Schwab and Fahey demonstrated alleviation of the symptoms of hyperviscosity in patients with Waldenström’s macroglobulinemia. Application to autoimmune disorders characterized by the presence of pathogenic antibodies or other serum components followed in the next two decades. The common method of PE utilizes cell separators, which are automated centrifugation devices. Blood is withdrawn from the patient, separated into its components, and returned following removal of one or more of its components. Later techniques employed a membrane filtration system to remove plasma from whole blood during extracorporeal perfusion. Most recently, selective immunoabsorption techniques utilizing tryptophanlinked polyvinyl alcohol have been developed. Whereas the former techniques require protein replacement (i.e., plasma or albumen), this is not required with selective immunoabsorption. The schedule of PE varies among different centers and often within a single center for different diseases. If the amount of a substance catabolized during exchange is equal to the sum of the amount synthesized and the amount shifted from the extravascular to the intravascular compartment, each procedure in which the volume exchanged approaches the patient’s plasma volume will remove approximately 50% to 60% of an intravascular substance. It is commonly believed that four to five such exchanges over 7 to 10 days constitutes adequate short-term therapy.102 Hence, according to a GBS study group in 1985, in the major GBS trials a total of 200 to 250 mL of plasma per kilogram of body weight was exchanged over 7 to 10 days.94 Because the efficiency of removing a particular serum

641

factor is greatest when the serum level is highest, some centers prefer to exchange smaller volumes on a more frequent basis (i.e., 2-L exchanges performed 10 times over 10 to 14 days).91 Unless the pathogenic factor is known, no ideal regimen of PE can be devised because plasma levels cannot be measured. PE may be of therapeutic benefit in those conditions in which a plasma constituent is responsible for disease manifestations. Such factors include autoantibodies (i.e., antibody to the acetylcholine receptor), toxins (i.e., phytanic acid in Refsum’s disease), and monoclonal antibodies (contributing to hyperviscosity in Waldenström’s macroglobulinemia). Other plasma components that may contribute to the disease pathogenesis, including cytokines and complement, are also removed by PE. The first application of PE in neurologic disease was in myasthenia gravis, an autoimmune disorder in which pathogenic antibodies play a central role. There is now substantial evidence of autoantibody involvement in several neuropathies, including GBS and its subsets, CIDP, IgM paraproteinemic-associated neuropathies, and possibly MMN (see Chapters 98, 99, and 101). However, the presence of pathogenic antibodies does not guarantee a response to PE, because antibodies may induce irreversible damage (i.e., neuronal loss). Adverse Events Potential adverse events identified by the National Institutes of Health Consensus Development Conference Statement (1960) include • Allergic reactions leading to anaphylaxis • Replacement with fluids depleted of coagulation factors, proteins, or electrolytes • Citrate-induced hypocalcemia • Replacement fluids containing plasma, which have the potential to transmit infection (e.g., human immunodeficiency virus [HIV] or hepatitis). Most centers now employ albumin or artificial plasma replacement to avoid risks associated with plasma. • Hemorrhage secondary to the use of systemic anticoagulants or dilution of coagulation factors • Activation of coagulation, complement, the fibrinolytic cascade, and/or the aggregation of platelets. Hence thrombophlebitis, thromboses, and pulmonary embolism have been observed. • Fluid imbalance • Vessel perforation and, in patients given arteriovenous shunts and central venous catheters, thrombosis and infection (bacterial endocarditis) may occur. Elderly patients with vascular instability carry an increased risk of severe complications. However, in centers expert in its use, PE is a safe procedure with fewer long-term side effects than corticosteroids and immunosuppressive agents. Complications were reported in 17% of 381 procedures in one series.22

642

Neuroimmunology of the Peripheral Nervous System

Use in Neuropathy Three large multicenter randomized controlled trials and a systematic review have shown that PE is effective therapy in GBS if started within the first 4 weeks of neurologic symptoms94 (see Chapter 98). The efficacy of PE in CIDP has also been shown by randomized trials32,43 and case series.20 Although the cause of GBS and CIDP remains uncertain, there is mounting evidence of a role for antibody and complement in certain subtypes of GBS (Miller Fisher syndrome and acute motor axonal neuropathy; see Chapter 26) and in some cases of CIDP,120 thus providing a rationale for PE therapy. PE has not proven of value in MMN, and it may aggravate this condition.86 Patients with neuropathy associated with immunoglobulin A (IgA) and immunoglobulin G (IgG) paraproteins respond to treatment with PE, whereas those with neuropathies associated with IgM paraproteins did not benefit in the one controlled study of PE in neuropathy associated with monoclonal gammopathy of uncertain significance (MGUS).32 Several uncontrolled studies have shown improvement when the IgM level is decreased by 50%,101 but because IgM levels rise rapidly following PE, it needs to be accompanied by some form of immunosuppression.85 PE together with dietary control remains essential in the maintenance treatment of Refsum’s disease (see Chapter 79). PE therapy has largely been replaced by IVIg therapy in patients with inflammatory demyelinating neuropathy on the grounds of convenience, avoidance of the necessity for specialized units and staff, greater ease of administration, and lesser occurrence of adverse events.

Intravenous Immunoglobulin The response of coincidental thrombocytopenia to treatment with IVIg being given for immune deficiency led to the introduction of IVIg for autoimmune thrombocytopenia.57 By a similar serendipitous observation, IVIg given for thrombocytopenia benefited a patient with CIDP115 and was introduced for the treatment of this and other potentially autoimmune diseases. The neurologic and neuromuscular disease indications for the use of IVIg have been reviewed by experts and expert panels25,70,119 and national guidelines have been promulgated.7 The immunoglobulin used in IVIg is prepared from the pooled plasma of several thousand donors.25,60 The method of preparation varies from product to product. The basic principle is that the gamma globulin is precipitated from the plasma by cold ethanol. The precipitate is gently treated at low pH with enzymes to avoid aggregation and remove impurities while preserving antibody activity. The final product consists of IgG monomer with some dimer and variable but small amounts of IgA (less than 2.5%) and traces of IgM. Some brands have particularly low levels of IgA. Viral safety is secured by screening the donors for risk

factors and testing their plasma for viral antigens and antibodies to eliminate hepatitis B, HIV, and other viruses. Viral elimination steps are also built into the manufacturing process so that the transmission of viral infection, although still a theoretical risk and the subject of active surveillance, has not been reported with current products. The final product is presented freeze dried or already dissolved as a solution with a stabilizer such as sucrose, maltose, or glycine. The distribution of IgG subclasses is approximately the same as that in plasma, and the full range of antibody activities is preserved. The inclusion of such a large donor population is thought to enhance the likelihood of the pool containing therapeutically useful anti-idiotypic antibodies that will inhibit the action of harmful autoantibodies. Administration The standard course first used for autoimmune thrombocytopenia was 0.4 g/kg daily for 5 days, and this has been adopted in neuromuscular diseases. However, for recurrent courses we and many others use 2.0 g/kg as a continuous infusion or as infusions of 1.0 g/kg on two consecutive days, thus avoiding the need for an overnight stay. We adhere to the manufacturers’ guidelines for the rate of infusion. In a recently reported series, patients preferred a faster rate of infusion (up to a maximum of 800 mL/hr instead of the usual 200 mL/hr) despite the increased risk of adverse events.41 If the patient is known to develop possibly allergic side effects during IVIg infusion and, as an exception, it has been decided to give IVIg again, we give chlorpheniramine and hydrocortisone intravenously 30 minutes before starting the infusion. During the infusion the patient should be under observation, especially during the first 30 minutes when the temperature, pulse, respiratory rate, and blood pressure should be monitored. Observations should be maintained throughout the infusion. If the patient develops side effects, we stop the infusion until the side effects subside and it may be possible to restart at a slower rate. Provision for treating anaphylaxis should be immediately available. Because many of the neurologic uses of IVIg are unlicensed, we always obtain signed informed consent before administering the treatment. Because of the theoretical risks, we also obtain such consent from patients with GBS, for whom the treatment is licensed. Our information sheet mentions the theoretical risks of transmission of viral and other infectious agents, including Creutzfeldt-Jakob disease. Mechanisms of Action Many different mechanisms of action have been proposed for IVIg (Table 28–1). It is likely that the actual mechanisms are multiple and the principal mechanism varies from disease to disease.60 Most of the evidence for these mechanisms comes from non-neurologic diseases, and only a few mechanisms have been investigated in peripheral nerve disorders.26

Principles of Immunotherapy

Table 28–1. Mechanisms of Action of Intravenous Immunoglobulin Autoantibodies Anti-idiotype blockade Increased breakdown of autoantibodies by blockade of endothelial FcRn transport receptors Suppression of antibody production Fc receptor Blockade of Fc␥ receptor III Stimulation of inhibitory Fc␥ receptor II Complement Prevention of membrane attack complex formation Cytokines Suppression of cytokine release Neutralization of circulating cytokines T cells Modulation of T-cell function Remyelination Possible enhancement of remyelination

Anti-idiotypic binding of IgG to autoantibodies has been shown to inhibit the action of autoantibodies against several systemic autoantigens and also against ganglioside GM1 in MMN,77,122 GQ1b in Miller Fisher syndrome, and neuroblastoma cells in CIDP.112 Immunoglobulin has also been shown to inhibit the neuromuscular blockade produced by GBS serum in a mouse phrenic nerve–diaphragm motor end-plate macropatch-clamp preparation.16 In a negative feedback loop, IVIg may engage Fc receptors on the B-cell surface and downregulate B-cell activation.29 Additionally, IVIg may saturate a protective endosomal receptor, FcRn, in endothelial cells that normally sequesters and protects IgG. If the receptor sites are blocked by the excess of IgG, then autoantibodies may be more exposed to catabolism in the lysosomes.121 IVIg is effective in idiopathic thrombocytopenia by blockade of the Fc receptor on macrophages. Antibodies to the Fc␥ receptor III or the Fc fraction of IgG are also effective. In addition, IgG has been shown to block the Fc receptors on human monocytes in vitro.69 In a murine model of autoimmune thrombocytopenia, the efficacy of IVIg in treating the lowered platelet count was shown in a knockout model to depend on stimulation of an inhibitory Fc␥ receptor II.97 However Fc␥ receptor III blockade on macrophages would inhibit macrophage-associated demyelination and provide a plausible explanation for the efficacy of IVIg in inflammatory neuropathy. Consistent with this hypothesis, there is a modulation of Fc␥II/Fc␥III receptors on the surface of circulating monocytes following IVIg23 such that the proportion of inhibitory Fc␥II receptors is increased relative to the active Fc␥III receptors. In dermatomyositis IVIg prevents the formation of complement membrane attack complex, possibly by binding of C3b to the Fc on IgG.11 Because complement-fixing

643

antibodies to myelin and deposition of complement at the myelin surface have been found in acute inflammatory demyelinating polyradiculoneuropathy and CIDP, this property of IVIg may also be important in these conditions. Human immune globulin contains antibodies against many cytokines.105 These may be partly responsible for the fall in levels of circulating cytokines after the administration of IVIg.26,60 In the acute stage of GBS the serum concentrations of TNF-␣ are increased but they fall following the administration of IVIg.99,100 There is also evidence that IVIg regulates T-cell function. Human immune globulin contains antibodies against the T-cell receptor that may be partly responsible.78 It also contains antibodies to superantigens that directly stimulate T cells. Binding to staphylococcal superantigen is the probable mechanism of action in Kawasaki disease.72 The expression of the important adhesion molecule ICAM on the surface of T cells is downregulated following IVIg, which may reduce their passage into the endoneurium.23 Indications Guillain-Barré Syndrome. According to a Cochrane systematic review, there are no adequate trials comparing IVIg with placebo, but IVIg has equivalent efficacy to PE in hastening recovery from severe GBS when treated within 2 weeks from onset of the disease.50 PE costs about the same, has more side effects, and is obviously less convenient. This led the American Academy of Neurology Practice Parameter Group to recommend consideration of IVIg for patients within 2 weeks and probably within 4 weeks after the onset of symptoms.51 It is not known whether a second course of IVIg is effective in patients who remain severely affected 2 or 3 weeks after the first. It is also not known whether IVIg is effective in children and in patients who have atypical forms of GBS, such as Miller Fisher syndrome, although it is often used in these situations. About 10% of patients relapse between 2 and 10 weeks after their first course of IVIg, an important consideration when patients are discharged to rehabilitation, and a second course is often given. Chronic Inflammatory Demyelinating Polyradiculoneuropathy. A Cochrane systematic review concluded that randomized controlled trials show that IVIg improves disability for at least 2 to 6 weeks compared with placebo.114 This is consistent with the general experience that between 60% and 80% of patients respond to IVIg.42,81 One trial showed that it has similar efficacy to PE,32 and another that it has similar efficacy to oral prednisolone.50 Because these three treatments seem to be equally effective, it is uncertain which of them should be the first choice. The disadvantages of IVIg are that it has to be repeated and 60% of patients require treatment for more than 6 months and 50% for more than a year. The frequency of repeated treatments ranges from 2 to 12 weeks. Despite the strength of the evidence and

644

Neuroimmunology of the Peripheral Nervous System

the conclusions of the Cochrane review, IVIg is not currently licensed for use in CIDP in the United States or the United Kingdom. Multifocal Motor Neuropathy or Pure Motor CIDP. Three small randomized trials have shown that MMN responds to treatment with IVIg in terms of an increase in strength, but it is not clear that there is a commensurate improvement in disability.111 Because steroids and PE are ineffective and immunosuppressant agents unproven, IVIg is the current treatment of choice. About two thirds of patients respond and require treatment every 2 to 12 weeks, which is inconvenient and expensive. The pure motor form of CIDP responds to IVIg and may be made worse by steroids,30 so IVIg is also the preferred first-line treatment. Paraproteinemic Demyelinating Neuropathy. Patients with IgA or IgG MGUS and demyelinating neuropathy often resemble those with CIDP and show a similar response to IVIg, although no randomized trial has ever been done. In two randomized controlled trials only a minority of patients with IgM MGUS and demyelinating neuropathy responded to IVIg.76 Other Neuropathies. There are anecdotal reports of improvement following IVIg in vasculitic neuropathy,58 diabetic proximal motor neuropathy, acute sensory neuronopathy,83 and paraneoplastic sensory neuropathy.109 No controlled trials or substantial series have been reported, and it is likely that such responses are rare and possible that they are coincidental. Adverse Events Minor side effects are common but serious adverse events are rare (Table 28–2).25,35,55,66,70,119 During standard rates of infusion, at least 10% of patients experience headache, back and limb pain, chest pain, flushing, or nausea, but these symptoms are usually tolerable or respond to slowing the infusion rate. The treatment may trigger migraine that can sometimes be prevented with propranolol,21 and in one series of 54 patients aseptic meningitis was reported in Table 28–2. Risks of Intravenous Immunoglobulin Common (5%–25%)

Uncommon (⬍5%)

Rare (⬍1%)

Very rare (⬍0.01%)

Headache, malaise, fever, myalgia, hypotension, lymphopenia, increased liver enzymes Meningism, urticaria, pompholyx, eczema (especially on the hands), neutropenia, renal failure Cerebral vasospasm, stroke, myocardial infarction, pulmonary embolism, pancytopenia, hemolytic anemia, alopecia Anaphylaxis, hepatic veno-occlusive disease, multiorgan failure

11%.98 The frequency of aseptic meningitis was not related to the brand of IVIg but was greater in those patients with a history of migraine. It subsided spontaneously after between 3 and 5 days. In one case aseptic meningitis was associated with reversible cerebral vasospasm causing ischemia116; the cause is not clear. With faster rates of infusion (up to 800 mL/hr), adverse effects are more common. They occurred in 38 of 50 consecutive patients (26% of 341 infusions) treated with a 5% solution given at up to 800 mL/hr.41 Eleven (22%) had major events but all recovered completely. Various skin reactions occur with IVIg and affected 7 (6%) of 120 patients in one series. During the infusion, flushing and urticaria may occur. These reactions may be related to particular batches. After the infusion, a vesicular rash on the hands and eczema have been reported. Renal failure has also been described.25 It has been attributed to sucrose in the IVIg preparation causing osmotic damage to the renal tubules. However, 8 (6.7%) of 119 patients (287 courses) developed renal impairment, irreversible in 2, and the occurrence of this side effect was independent of the sucrose content of the preparation.73 Patients should be checked for renal impairment before treatment and IVIg should be withheld if present. If it is given despite the presence of renal impairment, a non–sucrose-containing n brand should be used and renal function carefully monitored. Rare cases of stroke, myocardial infarction, and pulmonary embolism have been reported after IVIg. Their cause might be related to the increase in serum viscosity induced by IVIg.17,25,27,87 These occurrences have led to recommendations that extra caution should be exercised for patients who are suspected to have raised serum viscosity or cardiovascular or cerebrovascular disease or who are at risk of pulmonary embolism, but there is no consensus about the definition of these risk factors. Dalakas and Clark27 recommend caution in the presence of paraproteinemia, hyperfibrinogenemia, and hypercholesterolemia. Common sense dictates the use of a slower infusion rate, such as 0.4 g/kg over 5 days. Measuring plasma viscosity to identify patients who are at risk and lowering the viscosity if it is raised has been proposed.27 The use of prophylactic antiplatelet agents or low-dose heparin in patients with a history of stroke, myocardial infarction, or deep vein thrombosis has not been evaluated.

REFERENCES 1. Adcock, I. M., Brown, C. R., Gelder, C. M., et al.: Effects of glucocorticoids on transcription factor activation in human peripheral blood mononuclear cells. Am. J. Physiol. Cell Physiol. 268:C331, 1995. 2. Ahlmen, J., Andersen, O., Hallgren, G., and Peilot, B.: Positive effects of tacrolimus in a case of CIDP. Transplant. Proc. 30:4194, 1998.

Principles of Immunotherapy 3. Anstey, A., and Lear, J. T.: Azathioprine: Clinical pharmacology and current indications in autoimmune disorders. Biodrugs 33, 1998. 4. Archelos, J. J., Previtali, S. C., and Hartung, H. P.: The role of integrins in immune-mediated diseases of the nervous system. Trends Neurosci. 22:30, 1999. 5. Atkinson, J. P., and Frank, M. M.: Complement-independent clearance of IgG-sensitized erythrocytes: inhibition by cortisone. Blood 44:629, 1974. 6. Auphan, N., DiDonato, J. A., Rosette, C., et al.: Immunosuppression by glucocorticoids: inhibition of NF-kappa B activity through induction of I kappa B synthesis. Science 270:286, 1995. 7. Barnes, P., Hughes, R., Lecky, B., et al.: Association of British Neurologists Guidelines for the Use of Intravenous Immunoglobulin in Neurological Diseases. London, Association of British Neurologists, 2002. 8. Barnes, P. J., Greening, A. P., and Crompton, G. K.: Glucocorticoid resistance in asthma. Am. J. Respir. Crit. Care Med. 152:S125, 1995. 9. Barnett, M. H., Pollard, J. D., Davies, L., and McLeod, J. G.: Cyclosporin A in resistant chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 21:454, 1998. 10. Bartoszek, M., Brenner, A. M., and Szefler, S. J.: Prednisolone and methylprednisolone kinetics in children receiving anticonvulsant therapy. Clin. Pharmacol. Ther. 42:424, 1987. 11. Basta, M., and Dalakas, M. C.: High-dose intravenous immunoglobulin exerts its beneficial effect in patients with dermatomyositis by blocking endomysial deposition of activated complement fragments. J. Clin. Invest. 94:1729, 1994. 12. Begg, E. J., Atkinson, H. C., and Gianarakis, N.: The pharmacokinetics of corticosteroid agents. Med. J. Aust. 146:37, 1987. 13. Bensa, S., Hadden, R. D., Hahn, A., et al.: Randomized controlled trial of brain-derived neurotrophic factor in Guillain-Barre syndrome: a pilot study. Eur. J. Neurol. 7:423, 2000. 14. Berretta, L., Gingras, A. C., Svithin, Y. V., et al.: Rapamycin blocks the phosphorylation of 4E-BPI and inhibits capdependent initiation of translation. EMBO J. 15:658, 1996. 15. Brown, E. J., Beal, P. A., Keith, C. T., et al.: Control of p70 s6 kinase by kinase activity of FRAP in vivo. Nature 377:441, 1995. [erratum appears in Nature 378:644, 1995.] 16. Buchwald, B., Ahangari, R., Weishaupt, A., and Toyka, K. V.: Intravenous immunoglobulins neutralize blocking antibodies in Guillain-Barre syndrome. Ann. Neurol. 51:673, 2002. 17. Caress, J. B., Cartwright, M. S., Donofrio, P. D., and Peacock, J. E. Jr.: The clinical features of 16 cases of stroke associated with administration of IVIg. Neurology 60:1822, 2003. 18. Chang, C. C., Aversa, G., Punnonen, J., et al.: Brequinar sodium, mycophenolic acid, and cyclosporin A inhibit different stages of IL-4- or IL-13-induced human IgG4 and IgE production in vitro. Ann. N. Y. Acad. Sci. 696:108, 1993. 19. Chaudhry, V., Cornblath, D. R., Griffin, J. W., et al.: Mycophenolate mofetil: a safe and promising immunosuppressant in neuromuscular diseases. Neurology 56:94, 2001. 20. Choudhary, P. P., and Hughes, R. A.: Long-term treatment of chronic inflammatory demyelinating polyradiculoneuropathy with plasma exchange or intravenous immunoglobulin. QJM 88:493, 1995.

645

21. Constantinescu, C. S., Chang, A. P., and McCluskey, L. F.: Recurrent migraine and intravenous immune globulin therapy. N. Engl. J. Med. 329:583, 1993. 22. Couriel, D., and Weinstein, R.: Complications of therapeutic plasma exchange: a recent assessment. J. Clin. Apheresis 9:1, 1994. 23. Creange, A., Gregson, N. A., and Hughes, R. A.: Intravenous immunoglobulin modulates lymphocyte CD54 and monocyte FcgammaRII expression in patients with chronic inflammatory neuropathies. J. Neuroimmunol. 135:91, 2003. 24. Cuhaci, B., Kumar, M. S., Bloom, R. D., et al.: Transforming growth factor-beta levels in human allograft chronic fibrosis correlate with rate of decline in renal function. Transplantation 68:785, 1999. 25. Dalakas, M. C.: Intravenous immunoglobulin in the treatment of autoimmune neuromuscular diseases: present status and practical therapeutic guidelines. Muscle Nerve 22:1479, 1999. 26. Dalakas, M. C.: Mechanisms of action of IVIg and therapeutic considerations in the treatment of acute and chronic demyelinating neuropathies. Neurology 59:S13, 2002. 27. Dalakas, M. C., and Clark, W. M.: Strokes, thromboembolic events, and IVIg: rare incidents blemish an excellent safety record. Neurology 60:1736, 2003. 28. Dalakas, M. C., and Engel, W. K.: Chronic relapsing (dysimmune) polyneuropathy: pathogenesis and treatment. Ann. Neurol. 9(Suppl.):134, 1981. 29. Diegel, M. L., Rankin, B. M., Bolen, J. B., et al.: Cross-linking of Fc gamma receptor to surface immunoglobulin on B cells provides an inhibitory signal that closes the plasma membrane calcium channel. J. Biol. Chem. 269:11409, 1994. 30. Donaghy, M., Mills, K. R., Boniface, S. J., et al.: Pure motor demyelinating neuropathy deterioration after steroid therapy and improvement with intravenous immunoglobulin. J. Neurol. Neurosurg. Psychiatry 57:778, 1994. 31. Dumont, F. J., Staruch, M. J., Koprak, S. L., et al.: Distinct mechanisms of suppression of murine T cell activation by the related macrolides FK-506 and rapamycin. J. Immunol. 144:251, 1990. 32. Dyck, P. J., Low, P. A., Windebank, A. J., et al.: Plasma exchange in polyneuropathy associated with monoclonal gammopathy of undetermined significance. N. Engl. J. Med. 325:1482, 1991. 33. Dyck, P. J., O’Brien, P. C., Oviatt, K. F., et al.: Prednisone improves chronic inflammatory demyelinating polyradiculoneuropathy more than no treatment. Ann. Neurol. 11:136, 1982. 34. Dyck, P. J., O’Brien, P. C., Swanson, C., et al.: Combined azathioprine and prednisone in chronic inflammatorydemyelinating polyneuropathy. Neurology 35:1173, 1985. 35. Eijkhout, H. W., and van Aken, W. G.: Blood, blood components and plasma products. In Side Effects of Drugs Annual. Amsterdam, Elsevier, p. 398, 2004. 36. Fauci, A. S., Katz, P., Haynes, B. F., and Wolff, S. M.: Cyclophosphamide therapy of severe systemic necrotizing vasculitis. N. Engl. J. Med. 301:235, 1979. 37. Feldman, E. L., Bromberg, M. B., Albers, J. W., and Pestronk, A.: Immunosuppressive treatment in multifocal motor neuropathy. Ann. Neurol. 30:397, 1991.

646

Neuroimmunology of the Peripheral Nervous System

38. Felts, P. A., Smith, K. J., Gregson, N. A., and Hughes, R. A.: Brain-derived neurotrophic factor in experimental autoimmune neuritis. J. Neuroimmunol. 124:62, 2002. 39. Frohnert, P. P., and Sheps, S. G.: Long-term follow-up study of periarteritis nodosa. Am. J. Med. 43:8, 1967. 40. Good, J. L., Chehrenama, M., Mayer, R. F., and Koski, C. L.: Pulse cyclophosphamide therapy in chronic inflammatory demyelinating polyneuropathy. Neurology 51:1735, 1998. 41. Grillo, J. A., Gorson, K. C., Ropper, A. H., et al.: Rapid infusion of intravenous immune globulin in patients with neuromuscular disorders. Neurology 57:1699, 2001. 42. Hahn, A. F., Bolton, C. F., Pillay, N., et al.: Plasma-exchange therapy in chronic inflammatory demyelinating polyneuropathy: a double-blind, sham-controlled, cross-over study. Brain 119:1055, 1996. 43. Hahn, A. F., Bolton, C. F., Zochodne, D., and Feasby, T. E.: Intravenous immunoglobulin treatment in chronic inflammatory demyelinating polyneuropathy: a double-blind, placebocontrolled, cross-over study. Brain 119:1067, 1996. 44. Heitman, J., Movva, N. R., and Hall, M. N.: Targets for cell cycle arrest by the immunosuppressant rapamycin in yeast. Science 253:905, 1991. 45. Hengst, J. C., and Kempf, R. A.: Immunomodulation by cyclophosphamide. Clin. Immunol. Allergy 4:199, 1984. 46. Hodgkinson, S. J., Pollard, J. D., and McLeod, J. G.: Cyclosporin A in the treatment of chronic demyelinating polyradiculoneuropathy. J. Neurol. Neurosurg. Psychiatry 53:327, 1990. 47. Hohlfeld, R., Michels, M., Heininger, K., et al.: Azathioprine toxicity during long-term immunosuppression of generalized myasthenia gravis. Neurology 38:258, 1988. 48. Holme, S. A., Duley, J. A., Sanderson, J., et al.: Erythrocyte thiopurine methyl transferase assessment prior to azathioprine use in the UK. QJM 95:439, 2002. 49. Hortelano, S., and Bosca, L.: 6-Mercaptopurine decreases the Bcl-2/Bax ratio and induces apoptosis in activated splenic B lymphocytes. Mol. Pharmacol. 51:414, 1997. 50. Hughes, R., Bensa, S., Willison, H., et al.: Randomized controlled trial of intravenous immunoglobulin versus oral prednisolone in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:195, 2001. 51. Hughes, R., Wijdicks, E., Barohn, R. J., et al.: Practice parameter: immunotherapy for Guillain-Barre syndrome. Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 61:736, 2003. 52. Hughes, R. A.: Treatment of Guillain-Barre syndrome with corticosteroids: lack of benefit? Lancet 363:181, 2004. 53. Hughes, R. A., Swan, A. V., and van Doorn, P. A.: Cytotoxic drugs and interferons for chronic inflammatory demyelinating polyradiculoneuropathy. Cochrane Database Syst. Rev. CD003280, 2003. 54. Hughes, R. A., and van der Meche, F. G.: Corticosteroids for treating Guillain-Barre syndrome. Cochrane Database Syst. Rev. CD001446, 2000. [update in Cochrane Database Syst. Rev. 3:CD001446; 2000.] 55. Hughes, R. A. C.: Guidelines for the Use of Beta Interferons and Glatiramer Acetate in Multiple Sclerosis. London, Association of British Neurologists, p. 1, 2001. 56. Hull, K. M., and Boumpas, D. T.: Immunomodulating pharmaceuticals. In Rich, R. R., Fleisher, T. A., Shearer, W. T.,

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68. 69.

70.

71.

72.

73.

et al. (eds.): Clinical Immunology Principles and Practice, 2nd ed. St. Louis, C. V. Mosby, p. 1, 2001. Imbach, P., Barandun, S., d’Apuzzo, V., et al.: High-dose intravenous gammaglobulin for idiopathic thrombocytopenic purpura in childhood. Lancet 1:1228, 1981. Jayne, D. R., Davies, M. J., Fox, C. J., et al.: Treatment of systemic vasculitis with pooled intravenous immunoglobulin. Lancet 337:1137, 1991. Jones, T. B., Ankeny, D. P., Guan, Z., et al.: Passive or active immunization with myelin basic protein impairs neurological function and exacerbates neuropathology after spinal cord injury in rats. J. Neurosci. 24:3752, 2004. Kazatchkine, M. D., and Kaveri, S. V.: Immunomodulation of autoimmune and inflammatory diseases with intravenous immune globulin. N. Engl. J. Med. 345:747, 2001. Kerschensteiner, M., Stadelmann, C., Dechant, G., et al.: Neurotrophic cross-talk between the nervous and immune systems: implications for neurological diseases. Ann. Neurol. 53:292, 2003. Khanna, A. K., Cairns, V. R., Becker, C. G., and Hosenpud, J. D.: Transforming growth factor (TGF)-beta mimics and anti-TGF-beta antibody abrogates the in vivo effects of cyclosporine: demonstration of a direct role of TGF-beta in immunosuppression and nephrotoxicity of cyclosporine. Transplantation 67:882, 1999. Kieseier, B. C., Seifert, T., Giovannoni, G., and Hartung, H. P.: Matrix metalloproteinases in inflammatory demyelination: targets for treatment. Neurology 53:20, 1999. Kieseier, B. C., Tani, M., Mahad, D., et al.: Chemokines and chemokine receptors in inflammatory demyelinating neuropathies: a central role for IP-10. Brain 125:823, 2002. Kobashigawa, J., Miller, L., Renlund, D., et al.: A randomized active-controlled trial of mycophenolate mofetil in heart transplant recipients. Mycophenolate Mofetil Investigators. Transplantation 66:507, 1998. Koch, C.: Blood, blood components, plasma and plasma products. In Dukes, M. N. G., and Aronson, J. K. (eds.): Meyler’s Side Effects of Drugs. Amsterdam, Elsevier, p. 1123, 2000. Korn, T., Toyka, K., Hartung, H. P., and Jung, S.: Suppression of experimental autoimmune neuritis by leflunomide. Brain 124:1791, 2001. Koski, C. L.: Therapy of CIDP and related immune-mediated neuropathies. Neurology 59:S22, 2002. Kurlander, R. J.: Reversible and irreversible loss of Fc receptor function of human monocytes as a consequence of interaction with immunoglobulin G. J. Clin. Invest. 66:773, 1980. Latov, N., Chaudhry, V., Koski, C. L., et al.: Use of intravenous gamma globulins in neuroimmunologic diseases. J. Allergy Clin. Immunol. 108:S126, 2001. Laura, M., Gregson, N. A., Curmi, Y., and Hughes, R. A.: Efficacy of leukemia inhibitory factor in experimental autoimmune neuritis. J. Neuroimmunol. 133:56, 2002. Leung, D. Y., Burns, J. C., Newburger, J. W., and Geha, R. S.: Reversal of lymphocyte activation in vivo in the Kawasaki syndrome by intravenous gammaglobulin. J. Clin. Invest. 79:468, 1987. Levy, J. B., and Pusey, C. D.: Nephrotoxicity of intravenous immunoglobulin. QJM 93:751, 2000.

Principles of Immunotherapy 74. Lipsky, J. J.: Mycophenolate mofetil. Lancet 348:1357, 1996. 75. Liu, J., Farmer, J. D. Jr., Lane, W. S., et al.: Calcineurin is a common target of cyclophilin-cyclosporin A and FKBPFK506 complexes. Cell 66:807, 1991. 76. Lunn, M. P., and Nobile-Orazio, E.: Immunotherapy for IgM anti-myelin-associated glycoprotein paraproteinassociated peripheral neuropathies. Cochrane Database Syst. Rev. CD002827, 2003. 77. Malik, U., Oleksowicz, L., Latov, N., and Cardo, L. J.: Intravenous gamma-globulin inhibits binding of anti-GM1 to its target antigen. Ann. Neurol. 39:136, 1996. 78. Marchalonis, J. J., Kaymaz, H., Dedeoglu, F., et al.: Human autoantibodies reactive with synthetic autoantigens from T-cell receptor beta chain. Proc. Natl. Acad. Sci. U. S. A. 89:3325, 1992. 79. Marshall, S., Tardif, G., and Ashworth, N.: Local corticosteroid injection for carpal tunnel syndrome. Cochrane Database Syst. Rev. CD001554, 2002. [update in Cochrane Database Syst. Rev. 4:CD001554, 2000.] 80. Mehndiratta, M. M., and Hughes, R. A.: Corticosteroids for chronic inflammatory demyelinating polyradiculoneuropathy. Cochrane Database Syst. Rev. CD002062, 2002. [update in Cochrane Database Syst. Rev. 3:CD002062, 2001.] 81. Mendell, J. R., Barohn, R. J., Freimer, M. L., et al.: Randomized controlled trial of IVIg in untreated chronic inflammatory demyelinating polyradiculoneuropathy. Neurology 56:445, 2001. 82. Meucci, N., Cappellari, A., Barbieri, S., et al.: Long term effect of intravenous immunoglobulins and oral cyclophosphamide in multifocal motor neuropathy. J. Neurol. Neurosurg. Psychiatry 63:765, 1997. 83. Molina, J. A., Benito-Leon, J., Bermejo, F., et al.: Intravenous immunoglobulin therapy in sensory neuropathy associated with Sjogren’s syndrome. J. Neurol. Neurosurg. Psychiatry 60:699, 1996. 84. Mowzoon, N., Sussman, A., and Bradley, W. G.: Mycophenolate (CellCept) treatment of myasthenia gravis, chronic inflammatory polyneuropathy and inclusion body myositis. J. Neurol. Sci. 185:119, 2001. 85. Nobile-Orazio, E.: Neuropathies associated with anti MAG antibodies and IgM monoclonal gammopathies. In Latov, N., Wokke, J. H., and Kelly, J. I. (eds.): Immunological and Infectious Diseases of the Peripheral Nerves. Cambridge, UK, Cambridge University Press, p. 168, 1998. 86. Nobile-Orazio, E.: Multifocal motor neuropathy. J. Neuroimmunol. 115:4, 2001. 87. Okuda, D., Flaster, M., Frey, J., and Sivakumar, K.: Arterial thrombosis induced by IVIg and its treatment with tPA. Neurology 60:1825, 2003. 88. Ozer, H., Cowens, J. W., Colvin, M., et al.: In vitro effects of 4-hydroperoxycyclophosphamide on human immunoregulatory T subset function. I. Selective effects on lymphocyte function in T-B cell collaboration. J. Exp. Med. 155:276, 1982. 89. Pestronk, A., Cornblath, D. R., Ilyas, A. A., et al.: A treatable multifocal motor neuropathy with antibodies to GM1 ganglioside. Ann. Neurol. 24:73, 1988. 90. Pickup, M. E.: Clinical pharmacokinetics of prednisone and prednisolone. Clin. Pharmacokinet. 4:111, 1979.

647

91. Pollard, J. D.: Chronic inflammatory demyelinating polyradiculoneuropathy. Baillieres Clin. Neurol. 3:107, 1994. 92. Pollard, J. D., and Spies, J. M.: The Guillain Barre syndrome. In Asbury, A. K., McKhain, G. M., McDonald, W. I., et al. (eds.): Diseases of the Nervous System. Cambridge, UK, Cambridge University Press, p. 1110, 2002. 93. Rao, A., Luo, C., and Hogan, P. G.: Transcription factors of the NFAT family: regulation and function. Ann. Rev. Immunol. 15:707, 1997. 94. Raphael, J. C., Chevret, S., Hughes, R. A., and Annane, D.: Plasma exchange for Guillain-Barre syndrome. Cochrane Database Syst. Rev. CD001798, 2001. [update in Cochrane Database Syst. Rev. 2:CD001798, 2002.] 95. Sabin, T. D., Swift, T. R., and Jacobson, T. R.: Leprosy. In Dyck, P. J., Thomas, P. K., Griffin, J., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1354, 1994. 96. Salinas, R. A., Alvarez, G., Alvarez, M. I., and Ferreira, J.: Corticosteroids for Bell’s palsy (idiopathic facial paralysis). Cochrane Database Syst. Rev. CD001942, 2002. 97. Samuelsson, A., Towers, T. L., and Ravetch, J. V.: Antiinflammatory activity of IVIG mediated through the inhibitory Fc receptor. Science 291:484, 2001. 98. Sekul, E. A., Cupler, E. J., and Dalakas, M. C.: Aseptic meningitis associated with high-dose intravenous immunoglobulin therapy: frequency and risk factors. Ann. Intern. Med. 121:259, 1994. 99. Sharief, M. K., Ingram, D. A., Swash, M., and Thompson, E. J.: IV immunoglobulin reduces circulating proinflammatory cytokines in Guillain-Barre syndrome. Neurology 52:1833, 1999. 100. Sharief, M. K., McLean, B., and Thompson, E. J.: Elevated serum levels of tumor necrosis factor-alpha in Guillain-Barre syndrome. Ann. Neurol. 33:591, 1993. 101. Sherman, W. H., Olarte, M. R., McKiernan, G., et al.: Plasma exchange treatment of peripheral neuropathy associated with plasma cell dyscrasia. J. Neurol. Neurosurg. Psychiatry 47:813, 1984. 102. Shumak, K. H., and Rock, G. A.: Therapeutic plasma exchange. N. Engl. J. Med. 310:762, 1984. 103. Singh, G., Fries, J. F., Spitz, P., and Williams, C. A.: Toxic effects of azathioprine in rheumatoid arthritis: a national post-marketing perspective. Arthritis Rheum. 32:837, 1989. 104. Steiner, J. P., Connolly, M. A., Valentine, H. L., et al.: Neurotrophic actions of nonimmunosuppressive analogues of immunosuppressive drugs FK506, rapamycin and cyclosporin A. Nat. Med. 3:421, 1997. 105. Svenson, M., Hansen, M. B., Ross, C., et al.: Antibody to granulocyte-macrophage colony-stimulating factor is a dominant anti-cytokine activity in human IgG preparations. Blood 91:2054, 1998. 106. Takahashi, N., Hayano, T., and Suzuki, M.: Peptidyl-prolyl cis-trans isomerase is the cyclosporin A-binding protein cyclophilin. Nature 337:473, 1989. 107. Thornhill, M. H., Kyan-Aung, U., and Haskard, D. O.: IL-4 increases human endothelial cell adhesiveness for T cells but not for neutrophils. J. Immunol. 144:3060, 1990. 108. Trushin, S. A., Pennington, K. N., Algeciras-Schimnich, A., and Paya, C. V.: Protein kinase C and calcineurin synergize to

648

109.

110.

111.

112.

113.

114.

115.

Neuroimmunology of the Peripheral Nervous System activate IkappaB kinase and NF-kappaB in T lymphocytes. J. Biol. Chem. 274:22923, 1999. Uchuya, M., Graus, F., Vega, F., et al.: Intravenous immunoglobulin treatment in paraneoplastic neurological syndromes with antineuronal autoantibodies. J. Neurol. Neurosurg. Psychiatry 60:388, 1996. Umapathi, T., Hughes, R. A., Nobile-Orazio, E., and Leger, J. M.: Immunosuppressive treatment for multifocal motor neuropathy. Cochrane Database Syst. Rev. CD003217, 2002. Van den Berg-Vos, R. M., Franssen, H., Wokke, J. H., and Van den Berg, L. H.: Multifocal motor neuropathy: long-term clinical and electrophysiological assessment of intravenous immunoglobulin maintenance treatment. Brain 125:1875, 2002. van Doorn, P. A., Brand, A., and Vermeulen, M.: Antineuroblastoma cell line antibodies in inflammatory demyelinating polyneuropathy: inhibition in vitro and in vivo by IV immunoglobulin. Neurology 38:1592, 1988. van Koningsveld, R., Schmitz, P. I., Meche, F. G., et al.: Effect of methylprednisolone when added to standard treatment with intravenous immunoglobulin for GuillainBarre syndrome: randomised trial. Lancet 363:181, 2004. van Schaik, I. S., Winer, J. B., de Haan, R., and Vermeulen, M.: Intravenous immunoglobulin for chronic inflammatory demyelinating polyneuropathy (Cochrane Review). In The Cochrane Library. Chichester, UK, John Wiley, Issue 2, 2003. Vermeulen, M., van der Meche, F. G., Speelman, J. D., et al.: Plasma and gamma-globulin infusion in chronic inflammatory polyneuropathy. J. Neurol. Sci. 70:317, 1985.

116. Voltz, R., Rosen, F. V., Yousry, T., et al.: Reversible encephalopathy with cerebral vasospasm in a Guillain-Barre syndrome patient treated with intravenous immunoglobulin. Neurology 46:250, 1996. 117. Weishaupt, A., Schonrock, L. M., Stienekemeier, M., et al.: Glucocorticosteroids modulate antigen-induced T cell apoptosis in experimental autoimmune neuritis and cause T cell proliferation in situ. Acta Neuropathol. (Berl.) 102:75, 2001. 118. Wessel, G., Abendroth, K., and Wisheit, M.: Malignant transformation during immunosuppressive therapy (azathioprine) of rheumatoid arthritis and systemic lupus erythematosus: a retrospective study. Scand. J. Rheumatol. Suppl. 67:73, 1987. 119. Wiles, C. M., Brown, P., Chapel, H., et al.: Intravenous immunoglobulin in neurological disease: a specialist review. J. Neurol. Neurosurg. Psychiatry 72:440, 2002. 120. Yan, W. X., Archelos, J. J., Hartung, H. P., and Pollard, J. D.: P0 protein is a target antigen in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:286, 2001. 121. Yu, Z., and Lennon, V. A.: Mechanism of intravenous immune globulin therapy in antibody-mediated autoimmune diseases. N. Engl. J. Med. 340:227, 1999. 122. Yuki, N., and Miyagi, F.: Possible mechanism of intravenous immunoglobulin treatment on anti-GM1 antibody-mediated neuropathies. J. Neurol. Sci. 139:160, 1996. 123. Zhu, L. P., Cupps, T. R., Whalen, G., and Fauci, A. S.: Selective effects of cyclophosphamide therapy on activation, proliferation, and differentiation of human B cells. J. Clin. Invest. 79:1082, 1987.

29 Blood-Nerve Interface and Endoneurial Homeostasis ANANDA WEERASURIYA

Permeability of the BNI Dynamics of Blood-Nerve Exchange Relative Contributions of Perineurium and Microvessels Two Routes of Transcapillary Permeation Route of Transperineurial Permeation Transporters at the BNI The Perineurium: An Epithelium? The Epineurium: Not a Diffusion Barrier Assessment of Blood-Nerve Exchange of Solutes Endoneurial Fluid Dynamics Continuity between Cerebrospinal Fluid and Endoneurial Fluid

Convective Endoneurial Fluid Flow Driven by a Proximodistal Gradient Driving Force for Proximodistal Fluid Flow Daily Endoneurial Fluid Turnover (About 30%) Effect of Endoneurial Protein Concentration on EHP Decrease in Perineurial Compliance with Age Implications for Entrapment Neuropathies Endoneurial Homeostasis Controlled Variables of Endoneurial Homeostasis

Vertebrate peripheral axons and their associated glial cells function within a specialized milieu intérieur: the endoneurial microenvironment. Exchange of material between this intrafascicular physiologic space (accounting for about 20% to 25% of the fascicular volume) and the general extracellular space is restricted and regulated by the blood-nerve interface (BNI). This interface, also known as the bloodnerve barrier (BNB), consists of the endoneurial vascular endothelium and the multilayered investing sheath, the perineurium. The relative impermeability of the BNI to bloodborne material protects the endoneurial microenvironment from potentially harmful plasma constituents and rapid fluctuations of plasma solute concentrations that influence glial and axonal functions. The absence of lymphatic drainage in the endoneurial space79 further emphasizes the protective nature of the BNI. The composition of the endoneurial fluid and its physical properties are regulated by homeostatic mechanisms located at the BNI and in the endoneurial cellular elements. Highlighting the need to regulate these variables, the concept of the endoneurial microenvironment

Endoneurial Homeostasis in Lead (Pb) Neuropathy Endoneurial Homeostasis in Nerve Degeneration and Regeneration Endoneurial Homeostasis during Development Factors Regulating BNI Permeability Neurovascular Unit in Nerve Cellular Control of BNI Permeability Molecular Basis of BNI Permeability Regulation Therapeutic Implications for Peripheral Neuropathies Conclusion

as a regulated physiologic space was first proposed about 25 years ago.101,102 Early morphologic studies with several tracers described the permeability of the BNI as either leaky or impermeant and thus downplayed the BNI’s graded permeability to solutes of different sizes, which has been shown by subsequent physiologic studies. In reporting the results of morphologic studies of BNI alterations associated with neuropathies, it is unfortunately still common for authors to use terms such as functional/nonfunctional, competent/noncompetent, integrity/lack of integrity, and intact/breakdown to describe a change in permeability of endoneurial microvessels to morphologic tracers. Characterization of these permeability changes in these judgmental terms is not supported by the available literature. Such terms have limited usefulness in describing changes in endoneurial endothelial cells caused by crush, other trauma, severe toxic insults, and the like. In these cases, there is compelling evidence of endothelial cell injury or death. Therefore, when interpreting results 651

652

Functional Compartments and Nerve-Fiber Environment

from morphologic studies on BNI permeability, it is more prudent to limit the description to what is observed (e.g., a change in permeability to a particular tracer) rather than to misleadingly postulate a “breakdown” of the BNI, or BNB. Numerous studies have demonstrated increases in endoneurial microvessel permeability without any accompanying ultrastructural evidence of endothelial cell damage or death. At the same time, physiologic studies have measured initial increases and subsequent decreases in endoneurial microvessel and perineurial permeability. Taken together, these results reflect a functional complexity of the BNI much greater than that of a simple and passive restrictive barrier impeding the access of hydrophilic solutes to the endoneurial microenvironment. In fact, they are consistent with the dynamic aspects of a “blood-nerve interface” and reflect the operation of endoneurial homeostatic mechanisms, much as in the blood-tissue interfaces of other mammalian organs. For these reasons, and in anticipation of the increasing significance of emerging concepts such as the neurovascular unit and immunologically driven dynamic permeability changes, it is proposed that the BNI represents a more powerful and rigorous description of the regulatory dynamics governing the blood-nerve exchange of various solutes and water.

PERMEABILITY OF THE BNI Morphologic tracers (Evans blue–albumin, horseradish peroxidase [HRP], ferritin, etc.) have not been detected in the endoneurial space following intravascular or perifascicular administration of these substances.51,83 Furthermore, lymphatics have not been described in the endoneurium.79 This has led to the belief that the endoneurial extracellular space, like the cerebral microenvironment, is a protected space relatively free of plasma proteins. Subsequent quantitative studies with radiotracers confirmed the limited permeability of the endoneurial endothelium and perineurium.8,62,94,100,101,106 The morphologic correlates of this restricted blood-nerve exchange of small solutes as well as plasma macromolecules are the belts of tight junctions (zonulae occludentes); these surround the cells of the endoneurial vascular endothelium and inner layers of the perineurium.4,52,61,72 The calculated permeability coefficients demonstrate that only the cerebral capillaries are less permeable than the endoneurial counterparts. These studies have clearly and definitively established the specialized and regulated nature of the endoneurial microenvironment. The perineurium, which is the less permeable component of the BNI,24,64,100,101 defines and circumscribes the endoneurial microenvironment and separates it from the epineurial perifascicular fluid. This passive and nonvectorial role for the perineurium is consistent with its nonpolarized cytologic ultrastructure: basal membrane on both sides and

absence of a histologic distinction between apical and basal membranes,51,72 and the absence of transperineurial transport of ions102 and L-phenylalanine.87 The endoneurial vascular endothelium, in contrast, exerts an active role in regulating the immediate environs of axons and glia through its direct mediation of blood-nerve exchange. It exhibits clear cytologic polarization,37 and demonstrates facilitated transport of D-glucose62 and L-phenylalanine.87 Endoneurial capillaries, though leakier than the very “tight” epithelia,56 are nevertheless less permeable than all other capillaries examined so far except cerebral capillaries.10,11,48

DYNAMICS OF BLOOD-NERVE EXCHANGE Two routes are available for blood-nerve exchange of material. One pathway, the direct one, is across the endothelium of the endoneurial microvasculature. The other is an indirect route requiring passage of material through a third compartment interposed between the vascular space and the endoneurial extracellular space. This compartment is the perifascicular fluid that exchanges directly with the vascular compartment through the relatively leaky epineurial capillaries. Evidence from various physiologic and morphologic studies strongly indicates that blood-nerve exchange occurs predominantly via endoneurial capillaries and that transperineurial exchange is a minor contributor to this process. Hence, the terms BNI permeability and endoneurial capillary permeability are used interchangeably. Strictly speaking, BNI permeability to a given solute is slightly higher than the endoneurial capillary permeability to that same solute. The difference is due to the minor contribution of the relatively impermeable perineurium to BNI permeability.

Relative Contributions of Perineurium and Microvessels Theoretically, blood-borne substances can reach the endoneurial extracellular space either by traversing the endoneurial vascular endothelium or by crossing the multilayered perineurium after gaining access to the perineurial extracellular space. The following evidence favors the former pathway as the major route of blood-nerve exchange: 1. The perineurium is impermeable to ionic lanthanum, whereas the endoneurial capillaries are not.32,72 2. Intravascular perfusion with a hyperkalemic solution inactivates peripheral nerve much more rapidly than when the nerve is bathed by the same hyperkalemic solution.24 3. Histamine increases the permeability of endoneurial capillaries to macromolecular tracers, whereas it has no such effect on the perineurium.49,76

Blood-Nerve Interface and Endoneurial Homeostasis

4. In leprosy, the endoneurial blood vessels become permeable to ferritin, whereas the perineurium remains impermeable to this tracer.7 5. In the frog sciatic nerve, where perineurial permeability has been measured independently, the endoneurial capillaries are more permeable.89,100–102 This relationship has also been confirmed for the rat tibial63 and sciatic nerve.91 6. During the second to sixth week of wallerian degeneration, while the perineurial permeability increases about fourfold, the permeability coefficient–surface area product (PS) of the frog sciatic nerve decreases by more than 60%, reflecting the greater sensitivity to PS to permeability (P) of the capillaries than that of the perineurium.90 7. Calculations from the morphometric studies of Bell and Weddell6 reveal that the perimeter of perineurium in rat sciatic nerve cross section is approximately equal to the sum of the perimeters of all the endoneurial vessels. 8. PS of the adult rat sciatic nerve perineurium to [125I]albumin was measured to be 1.48 ⫾ 0.28 ⫻ 10⫺7 mL ⭈ g⫺1 ⭈ s⫺1 (n ⫽ 6) (A. Weerasuriya, unpublished observations). This is about two orders of magnitude less than the corresponding value for the endoneurial vessels.94 All these studies clearly emphasize the much more restrictive diffusion barrier properties of the perineurium. As far as solute concentrations in the endoneurial fluid are concerned, permeability of the BNI to a solute is an accurate and adequate measure of the blood-nerve exchange of that solute.

Two Routes of Transcapillary Permeation Before adequate structural evidence was available, physiologists, on the basis of permeability measurement to hydrophilic solutes of various sizes, postulated that the capillary endothelium contained a set of large and another set of small hydrophilic pores.19,53 The postulated small pores could be either cylindrical channels with a diameter of 7 to 9 nm or slitlike pores with a width of about 6 nm. They are considered to occupy about 0.01% of the capillary surface area, and are permeable to ions, amino acids, glucose, and the like but not to macromolecules such as albumin. These bigger molecules traverse the capillary wall via the larger pores, which are postulated to have a diameter of 70 to 90 nm. As would be expected, these large pores are rare, and the ratio of large to small pores, though variable, seems to be in the range of about 1:30,000.22,66 Evidence from electron microscopic studies suggest that the small pores are the interendothelial clefts, which have a width of about 20 nm and occupy about 0.4% of the capillary surface area.9 Because of their rarity, the identity of large pores is less certain. Potential candidates are (1) widened intercellular

653

junctions, (2) transendothelial channels formed by the fusion of plasmalemmal invaginations, and (3) fenestrae. Properties of the small- and large-pore pathways are exhaustively reviewed by Crone and Levitt,12 Taylor and Granger,80 and Rippe and Haraldsson.67 Endoneurial capillaries also have separate small- and large-pore pathways for hydrophilic solutes of different sizes.

Route of Transperineurial Permeation The perineurium is a multilayered structure of flattened cells derived from fibroblasts. Each layer of the perineurium is invested by basement membrane on both sides. Hence, in contrast to classic epithelial tissue, the perineurium does not display an apicobasal polarity. The major route for transperineurial passage of material, then, is a paracellular pathway. This paracellular pathway consists of a number of belts of intercellular tight junctions arranged in series due to the multilayered nature of the perineurium. The relative impermeance of this route is illustrated by the fact that local anesthetics administered for nerve blocks are injected via a needle that penetrates the perineurial sheath. The number of layers in the perineurium decreases proximodistally. Thus the fine terminals of sensory and motor nerves have only one or two layers of perineurial cells. In comparison to the central nervous system, the structure of the perineurium combines features of the dura mater in terms of mechanical strength and the arachnoid with respect to impermeance. A recent report of the presence of vascular endothelial cadherin in perineurial cells73 suggests that perineurial permeability is regulated by mechanisms similar to those operating on the vascular endothelial cells of the endoneurium. This possibility is further strengthened by the presence of tight junction proteins (occludin and zonula occludens-1) in perineurial and nerve endothelial cells.85 Attempts to examine the consequences of perineurial removal are confounded by a compromise of the endoneurial vasculature, which receives nutrient branches from transperineurial arteries.40,44,46,75,81 However, this remains an important question especially with regard to perineurial regeneration associated with nerve repair. Thomas and Bhagat,82 studying perineurial regeneration and reorganization following removal of endoneurial contents, suggested that “neural structures may be responsible for the development and maintenance of the structural organization of the perineurium.”

Transporters at the BNI Given the relative impermeability of the perineurium and endoneurial capillary wall, it is not surprising that bloodnerve exchange of several solutes depends on the presence of specific transporter molecules straddling the interface. This condition is analogous to that of the cerebral microvasculature, which is a dynamic capillary interface with the

654

Functional Compartments and Nerve-Fiber Environment

least known permeability and minimal pinocytotic traffic coupled with a whole host of transporters and sensors responsible for the moment-to-moment, active and immediate regulation of the cerebral microenvironment. The endoneurial microenvironment, unlike the cerebral microenvironment, does not require a moment-to-moment regulation of nutrients and oxygen. Several transporters designed to meet the metabolic requirements of endoneurial constituents have been described. The earlier radioisotopic demonstration of facilitated transport of D-glucose62 has been complemented by reports on the presence of GLUT-1 in endothelial cells and perineurial cells.15,16,33 In keeping with the earlier postulate that the perineurium is a specialized connective tissue, these studies did not demonstrate an apicobasal polarity of GLUT-1 transporters in perineurial cells. Hence the role of perineurial GLUT-1 transporters appears to be the nourishment of perineurial cells.

The Perineurium: An Epithelium? The earliest ultrastructural studies of the perineurium concluded that it was an epithelium.71 It was postulated that perineurial cells with active metabolically driven pumps contributed to the composition of the fluid surrounding axons and glial cells. An analogy was drawn to the arachnoid epithelium investing the central nervous system. Subsequent ultrastructural studies have not confirmed these suppositions, and in fact a somewhat different picture has emerged. Frank Low and colleagues20,21,27,35 concluded that the perineurium is a specialized type of connective tissue, and the subsequent demonstration of the embryologic origin of the perineurium from fibroblasts has validated this conclusion. These studies showed that perineurial cells were thin, flattened, interconnected cells and that the layers of perineurial cells had basement membrane on both sides. This is a clear distinction from layers of epithelial cells, which have basement membrane only on one side—the basolateral side. The other surface of the epithelial layer constitutes the apical side and is not covered by basement membrane. A cardinal feature of epithelial layers is the apicobasal polarity of functional organization. This is reflected in the asymmetrical distribution of transporters, channels, and pumps across the apical and basolateral membranes. In addition, the functional correlate is a transepithelial flux of solutes and most often a transepithelial potential difference. Despite several attempts, active transperineurial fluxes of solutes and a transperineurial potential difference have not been discovered.45,102,104 Thus the current consensus is that the fibroblastderived perineurium is a specialized laminated connective tissue.27,39 Even though perineurial cells express GLUT-1, transperineurial transport of glucose is yet to be demonstrated. There is as yet no evidence of sodium-driven transperineurial cotransport of glucose and amino acids.

Whether perineurial cells participating in microfasciculation associated with nerve regeneration acquire specialized functions remains an open and intriguing question.

The Epineurium: Not a Diffusion Barrier Several fascicles are loosely bound together by the epineurium to form a peripheral nerve. The function of the epineurium is that of a connective tissue. Separation of the bundles by disruption of the epineurium does not seem to disturb the integrity of the nerve fascicles. Morphologic tracer studies52,72 demonstrated the easy penetrance of tracer macromolecules through the epineurium and even some of the outer layers of the perineurium. It is only the innermost layers of the perineurium that are responsible for the diffusion barrier properties to the nerve sheath.

Assessment of Blood-Nerve Exchange of Solutes Concentrations of endoneurial constituents are affected by exchange across endoneurial capillary interface and the barrier imposed the inner layers of the perineurium, and the metabolic processes of endoneurial cells. Permeabilities of the two exchange sites are assessed by morphologic and physiologic methods. Both techniques have their advantages and limitations. Historically, morphologic techniques establish the initial broad parameters and then physiologic methods provide quantitative measurements of blood-nerve transfer coefficients. The major advantage of morphologic techniques is the localization of barriers to the penetrance of markers and the array of tracers of varying molecular weights, charges, and sizes. Conversely, these methods are limited by the sensitivity of the histologic staining techniques. For example, histologic methods have not detected blood-nerve transfer of macromolecules, but physiologic methods, with radiotracers, estimate blood-nerve transfer of albumin at about 10⫺5 cm/s. Thus physiologic methods not only allow a quantification of transfer rates, but also are more sensitive, especially to macromolecules and other larger species. Furthermore, only transfer rates of amino acids, sugars, and other small molecules can be studied by physiologic radioisotopic methods; these techniques do not allow localization of the sites of transfer or transport. Immunostaining techniques and in situ hybridization complement these transport studies by localizing the putative transporters and diffusion sites. Technical details of measurement of blood-nerve transfer rates are described elsewhere.64,94 The theoretical foundations of blood-tissue transfer studies and their limitations are elegantly described by Smith and Allen.74

655

Blood-Nerve Interface and Endoneurial Homeostasis

The extracellular space within a nerve fascicle is about 20% to 25% of the total intrafascicular volume. The endoneurial hydrostatic pressure (EHP) exerted by this fluid is about 2 to 3 mm Hg. In terms of fluid dynamics, the endoneurial contents consist of a noncompressible aqueous solution, and a somewhat compressible cellular content constrained by an elastic sheath, the perineurium. Hence, the EHP will be determined by the volume of the endoneurial contents and the compliance of the perineurial sheath. Fluid enters and leaves a given segment of the endoneurial space across the walls of the endoneurial microvessels, and by convective endoneurial proximodistal fluid flow. Given the slightly positive tissue hydrostatic pressure of the endoneurial space, fluid does not normally enter this space across the perineurium. Endoneurial extracellular fluid exchanges material with blood directly across the endoneurial vascular endothelium, and indirectly across the multilayered perineurium, and is turned over by convective endoneurial fluid flow (EFF). The kinetics of this exchange are governed by (1) the hydrostatic and osmotic pressures in the different compartments, (2) the ion concentrations in plasma and endoneurial fluid, (3) the distribution volumes of the relevant solutes in the endoneurium, (4) the PSs of the BNI to these solutes, and (5) the rate of EFF.

Continuity between Cerebrospinal Fluid and Endoneurial Fluid The meninges of the central nervous system are continuous with sheaths of peripheral nerve.21 Whereas the epineurial connective tissue layers are continuous with the dura mater at the central ends (dorsal and ventral roots) of peripheral nerves, the relationship between the other meninges (pia mater and arachnoid layer) and perineurium is more complex. At the subarachnoid angle of nerve roots, there appears to be continuity of the subarachnoid space and endoneurial space, thus providing for continuity between cerebrospinal fluid (CSF) and endoneurial fluid. This, then, is the conduit through which material passes from CSF to endoneurial fluid.69 The embryologic aspects of this continuity do not appear to have been investigated.

injections of dyes, crystals, and radioactive mineral salts. Their conclusions were based on comparisons of proximodistal spread of the indicators in dead and living tissue. The two major limitations of that study were (1) the use of injection volumes that were no smaller than 100 ␮L and (2) the use of small hydrophilic tracers that could have migrated down the nerve by entering the vascular compartment and later reentering the endoneurial space. Therefore, they were unable to calculate a precise rate of convective fluid flow from their data and suggested an approximate rate of 3 mm/hr. This was supported by Mellick and Cavanagh36 from the results of their studies on traumatized chicken sciatic nerves. Low,28 injecting 10 ␮L of tetrodotoxin into the endoneurium and monitoring the rate and spread of inactivation, concluded that convective fluid flow is about 4 to 8 mm/hr. Morphologic studies21,35 demonstrated a continuity between the spinal subarachnoid space and the endoneurial space of spinal nerve roots; this strongly supports the possibility that CSF contributes to endoneurial fluid. Earlier, Waksman88 had suggested that diphtheria toxin enters the central nervous system through peripheral nerves, probably exploiting this pathway. It is not unlikely that, in diabetic neuropathy, disturbances in convective EFF play a role in the greater fiber loss and demyelination seen more distally than proximally. The data in Figure 29–1 are consistent with a convective flow of endoneurial fluid. The shift of the curve to the right (proximal to distal) from the first to the second hour is clearcut, but that from the second to the fourth hour is somewhat more subtle, although nevertheless evident upon closer inspection of the two curves. Whereas both curves seem to

900 800 Activity (cpm) per mg of nerve

ENDONEURIAL FLUID DYNAMICS

700

1 hour 2 hours 4 hours

600 500 400 300 200 100

Proximal

Distal

0 −20

−15

−10

−5

0

5

10

15

20

Distance along nerve (mm)

Convective Endoneurial Fluid Flow Driven by a Proximodistal Gradient In addition to the blood-nerve exchange discussed earlier, another putative source of input/output to the endoneurium is convective EFF. In an elegant series of experiments, Weiss and colleagues105 demonstrated the presence of a proximodistal flow of fluid in rat and guinea pig sciatic nerve. As indices of fluid movement, they used endoneurial

FIGURE 29–1 Rate of endoneurial convective fluid flow of 22Na as a function of time. In all three experiments, 70 nL of saline with 22Na were microinjected into a rat sciatic nerve. Then, 1, 2, or 4 hours later, the nerves were harvested, cut into 3-mm segments, counted for 22 Na activity, dried, and weighed. Negative numbers indicate distances proximal to the site of injection and positive numbers represent distances distal to the site of injection. (A. Weerasuriya, unpublished data.)

656

Functional Compartments and Nerve-Fiber Environment

peak at the fifth millimeter, the 2-hour curve has a bigger shoulder at lengths less than 5 mm, and the 4-hour curve has a shoulder at lengths greater than 5 mm. Thus the pattern of 22 Na distribution along the length of the nerve at the three different survival times clearly demonstrates a progressive asymmetrical proximodistal movement of the isotope.

Driving Force for Proximodistal Fluid Flow There does not appear to have been any mechanistic studies addressing the issue of the driving force for proximodistal fluid flow. The descriptive studies on EHP57 are consistent with the CSF pressure in the spinal cord being the pressure head for the proximodistal flow of endoneurial fluid. The hydrostatic pressure in the cord is about 10 mm Hg, that in the peripheral nerves about 1 to 2 mm Hg, and that in the dorsal root ganglia about 3 to 5 mm Hg.43 The presence of this pressure drop from the interstitial spaces of the spinal cord to the endoneurial interstitium of peripheral nerve supports the contention that the CSF pressure is the main propulsive force of the endoneurial fluid. Unfortunately, this postulate is not without theoretical limitations. For example, what would be the fate of this CSF pressure–driven endoneurial fluid when it reaches the distal ends of sensory nerves, which do not have open perineurial sleeves?27 Is the perineurium in such nerves more permeable distally than proximally to allow for a transperineurial dissipation of endoneurial fluid? If not, is there a slower turnover of endoneurial fluid at the distal end of a sensory nerve? Would this then contribute to the greater vulnerability of sensory nerves to pyridoxine toxicity?

Daily Endoneurial Fluid Turnover (About 30%) The albumin content of desheathed human sural nerve was reported to be 8.7 ␮g/mg of dry weight.54 If it is assumed that (1) the endoneurial wet/dry weight ratio is 3.0, (2) endoneurial albumin is extracellular and free, and (3) endoneurial extracellular space is about 25%, then this value can be converted to an albumin concentration of 11.6 mg/mL in the endoneurial fluid. Plasma albumin concentration is 33.1 mg/mL.54 From these two concentration terms and the PS to albumin,94 the calculated rate of blood-nerve albumin transfer is about 1.2 mg ⭈ g⫺1 ⭈ day⫺1. At this rate of transfer, and assuming relatively constant albumin concentrations in endoneurium and plasma, about 30% of the endoneurial albumin is turned over each day. By comparison, CSF (and its constituents) is turned over about four times each day60 and, in anesthetized rats, daily turnover of interstitial albumin in hind limb muscles and skin is about 65%.65 The latter value increases about fourfold in awake, freely moving rats. The rate of removal of albumin from the endoneurium is also about 1.2 mg ⭈ g⫺1 ⭈ day⫺1. In muscle, skin, gut, and

similar tissues, lymphatic drainage plays a role in clearing interstitial albumin, and in the central nervous system the cerebrospinal fluid and perivascular drainage are the sinks for extracellular albumin.3 In peripheral nerve, in the absence of both lymphatics and an active CSF circulation, the route of removal of albumin and other macromolecules remains to be identified. The metabolic breakdown of albumin to supply amino acids to axons and glia is one possibility; another is the long-suspected proximodistal convective flow of endoneurial fluid.

Effect of Endoneurial Protein Concentration on EHP From the equation proposed by Landis and Pappenheimer,53 endoneurial albumin can be expected to exert an interstitial oncotic pressure of about 3 mm Hg. The recorded EHP of 2 to 3 mm Hg will oppose the endoneurial oncotic pressure and thus minimize net fluid filtration from the endoneurial vasculature. However, this balance of forces can be disturbed if the capillary permeability to albumin increases slightly, allowing the endoneurial albumin concentration to rise and thus draw more fluid from the vascular compartment into the endoneurial interstitium. The resulting endoneurial edema together with the low compliance and hydraulic conductivity of the perineurium will elevate EHP. The net fluid gain by the endoneurial space will cease when a new equilibrium is established among the hydrostatic and oncotic pressures of the endoneurial and intravascular compartments. It is quite likely that this is the mechanism of edema formation observed in experimental diabetic and lead neuropathies, as well as early wallerian degeneration. In contrast, the edema present in galactose-induced neuropathy is likely to be due to the presence of an excess of non–vascularly derived osmolytes in the endoneurium, corroborated by the absence of change in the PS of the BNI to sucrose in galactose-intoxicated rats.38

Decrease in Perineurial Compliance with Age Developmental increase of EHP has two plateaus.13 The first one is from about 3 to 13 weeks of age, and the second one is from 6 months onward. Elevations of EHP can be produced either by increased EFF or by reduced perineurial compliance. The results of the study by Crowley and colleagues13 demonstrated that reduced perineurial compliance with aging contributes to the second increase in EHP.

Implications for Entrapment Neuropathies Extrafascicular mechanical compression and the resultant ischemia certainly contribute to the symptomatology of entrapment neuropathies. But, are there other factors more closely tied to the endoneurial microenvironment that affect the susceptibility and evolution of the nerve injury? It is

Blood-Nerve Interface and Endoneurial Homeostasis

hypothesized that the initial event in the evolution of an entrapment neuropathy is the limitation and reduction of EFF as a result of externally applied mechanical forces. Second, elevation of EHP as a result of obstruction of EFF and continued application of these external forces leads to endoneurial ischemia and its attendant pathology. Additionally, the reduced compliance of the aged nerve13,99 makes it more susceptible to externally applied mechanical pressures causing an elevation of EHP. This hypothesis is consistent with (1) the paucity of carpal tunnel syndrome (CTS) in teenagers who play arcade games, (2) the higher incidence of CTS in conditions with increased tissue water content or edema, such as pregnancy and diabetes, and (3) the presence of symptoms quite proximal to the site of entrapment. However, this hypothesis does not explain the higher incidence of CTS in females or the varying patterns of recovery seen after resection of the flexor retinaculum.

ENDONEURIAL HOMEOSTASIS Much as in other tissues, the various compartments of the nervous system employ homeostatic mechanisms to regulate their respective internal environments. The relevant physiologic functional unit for a vertebrate peripheral nerve is a fasciculus. This cylindrical structural entity, encompassing axons, glial cells, and other attendant satellite cells, has its outer limits defined by the perineurium. This physical limit is also a physiologic barrier and limits the exchange of material between the endoneurial space and the general extracellular space surrounding a peripheral nerve.

Controlled Variables of Endoneurial Homeostasis The specialized microenvironment of peripheral nerve fibers is maintained with the assistance of the BNI. Regulated blood-nerve exchange across the BNI and turnover of endoneurial fluid by convective fluid flow are vital for the maintenance of endoneurial physiologic parameters (blood flow, oxygen tension, pH, oncotic pressure, hydrostatic pressure, ion concentrations, etc.) within the normal range necessary for the proper functioning of nerve fibers. There are independent transendothelial pathways for the movement of ions and macromolecules. Some of these physiologic parameters have been measured: PS to [14C]sucrose,63,70 to [14C]glucose,62 to [125I]albumin,94 and to 22 Na,91 as well as endoneurial blood flow,31,68 EHP,30,41 and ion concentration in endoneurial fluid.28,93 Other parameters have been described only qualitatively or not at all (endoneurial concentration of hydrogen and calcium, rate of convective EFF; volume of endoneurial extracellular space, perineurial permeability to macromolecules, tortuosity of the endoneurial extracellular contents, etc.). To establish that a particular variable is the controlled variable of a

657

homeostatic loop, it is necessary to establish the relative stability of that variable within a physiologic range in the face of environmental stressors and perturbations. This list of variables is not exhaustive, nor does it relate to all the pathophysiologic alterations associated with the endoneurial microenvironment. However, it is hoped that discussions regarding these few variables with available data in a few clinical scenarios will emphasize the relevance of considering pathophysiologic alterations as perturbations of endoneurial homeostasis. This promotes the view that therapeutic strategies are attempts to assist endoneurial mechanisms to restore the normal homeostatic balance of the endoneurium.

Endoneurial Homeostasis in Lead (Pb) Neuropathy The major effects of Pb on peripheral nerve are endoneurial edema, nuclear inclusion bodies, demyelination, elevation of EHP, and increased permeability of the BNI.108 Of these alterations, the nuclear inclusion bodies are observed first, followed by endoneurial edema. It had been suggested earlier29,42 that an increase in the permeability of endoneurial capillaries was the primary pathologic event leading to subsequent Schwann cell damage and segmental demyelination. Later studies on nerve pathology and with morphologic tracers indicated that accumulation of lead in the endoneurium and nerve edema precede qualitative changes in the permeability of endoneurial capillaries.59,109 The BNI index to albumin (a measure of the rate of albumin entry and removal from the endoneurium) starts to increase only at 6 weeks in lead-intoxicated rats, and suggests that the change in BNI permeability is subsequent to the direct toxic effect of lead on Schwann cells.47 A subsequent study96 provided clear, quantitative evidence that endoneurial pathology precedes an increase in permeability of the BNI, and furthermore, that the increase in permeability is about threefold. It is further implied that the increase of BNI permeability is not a consequence of disruption of its barrier properties, but more in the nature of an adaptive response to aid in the clearing of myelin debris from the endoneurium. The following scheme (Fig. 29–2), which is an extension of earlier hypotheses, attempts to delineate the causal relationships among the changes in various components of the endoneurium in leadinduced neuropathy. The primary event is the increase of plasma lead concentration. From the calculated PS of the BNI to ions and small nonelectrolytes, the PS of the BNI to lead is predicted to be around 1 ⫻ 10⫺4 mL ⭈ g⫺1 ⭈ s⫺1 and the half-time to be around 30 minutes. Therefore, in the absence of rapid fluctuations of blood lead levels, there is likely to be approximate equilibration between the lead concentrations in plasma and endoneurial fluid. Endoneurial lead accumulates among the myelin lamellae and in nuclear inclusion bodies in Schwann cells. The sequestration of lead in intranuclear bodies is a

658

Functional Compartments and Nerve-Fiber Environment

↑ Endoneurial Pb Schwann cells

↑ Permeability of BNI

Demyelination Endothelium

↑ Endoneurial albumin content

Endoneurial Homeostasis in Nerve Degeneration and Regeneration

Toxic effects on Schwann cells

↑ Endoneurial osmolytes ↑ Endoneurial water content and edema

Relatively unchanged perineurial permeability ↑ Endoneurial hydrostatic pressure

FIGURE 29–2 Proposed sequence of events leading to demyelination and endoneurial edema in lead neuropathy. (Data from Weerasuriya, A., Curran, G. L., and Poduslo, J. F.: Physiological changes in the sciatic nerve endoneurium of lead intoxicated rats: a model of endoneurial homeostasis. Brain Res. 517:1, 1990.)

protective mechanism to minimize elevations of cytosolic lead concentration. Myelin by-products increase the osmolality of the endoneurial fluid and cause a net movement of water from the vascular lumen to the endoneurium, leading to endoneurial edema. In the presence of a relatively unchanged permeability of the perineurium, the increased water content will cause an elevation of EHP. Endothelial cells also accumulate lead in intranuclear bodies, and the increased permeability of the BNI could be a consequence of endothelial damage, a Schwann cell–initiated mechanism to clear myelin debris from the endoneurium, or both. In either case, because interendothelial HRP leakage is not profuse and endothelial cell changes are more of a reactive nature rather than an expression of outright injury, it is suggested that the change in BNI permeability is an adaptive response of the vascular endothelium to restore/maintain the homeostasis of the endoneurial microenvironment. The occurrence of remyelination in chronically leadintoxicated rats supports the previous contention of an endoneurial microenvironment conducive for repair and regeneration. An intriguing but unresolved clinical issue is that lead toxicity predominantly leads to an encephalopathy in children, whereas in adults the major manifestation is a neuropathy. A greater vulnerability of the juvenile cerebral neurovascular unit to lead is a likely but as yet unproven hypothesis.

Wallerian degeneration is probably the most drastic reorganization of the endoneurial architecture.77,84 The fact that perineurial permeability increases only about twofold under these conditions97 argues against a “breakdown” of the barrier properties of this structure. An alternative interpretation of a dynamic response of the perineurium to maintain endoneurial homeostasis seems more plausible. Morphologically, perineurial cells in degenerating nerve show no evidence of disruption.25,107 On the contrary, they proliferate, hypertrophy, and increase their organelle content.25 The initial peak of perineurial permeability, lasting only a few days,97 is most likely related to the acute inflammatory response triggered by the trauma of nerve section, and the proliferative response of the perineurium25 (Fig. 29–3). Sectioned degenerating nerves display an increased number of mast cells.14,25,57 Degranulation of these mast cells releases histamine, a biogenic amine with potent inflammatory properties. The later sustained increase is probably a component of endoneurial homeostasis related to prevention of elevation of EHP (Fig. 29–4) and clearance of myelin debris from the endoneurium. The increased permeability of both components of the BNI is likely to facilitate the entry of monocytes into the endoneurium for myelin phagocytosis.78 Although lipid droplets and proteinaceous material are present among perineurial layers in degenerating nerves, the cells themselves are not disrupted.25,107 Also, it is unlikely that the postulated diffusible factor plays a role in the short-lasting acute increase of perineurial permeability. 175 Change from normal (%)

↑ Blood Pb >0.8 ug/mL

Perin. perm.

150 125 100

Endon. volume

75 50

Perin. area

25 0 0

2

4

6

8

10

12

14

16

Time (wk)

FIGURE 29–3 Relative changes in perineurial permeability to 22Na, endoneurial volume, and perineurial area during wallerian degeneration. Endon. Volume ⫽ endoneurial volume; Perin. Area ⫽ perineurial area; Perin. perm. ⫽ perineurial permeability. (Data from Weerasuriya, A., and Hockman, C. H.: Perineurial permeability to sodium during wallerian degeneration in rat sciatic nerve. Brain Res. 587:327, 1992.)

659

10

150 Pressure 125

8

100

6

75 Permeability

4

50

2

25

0

0 0

4

8

12

20

16

21

28

Time (d)

Perineural permeability—change from normal (%)

Endoneurial pressure (cm H2O)

Blood-Nerve Interface and Endoneurial Homeostasis

FIGURE 29–4 Endoneurial hydrostatic pressure and perineurial permeability to 22 Na during the first 4 weeks of wallerian degeneration. (Data from Weerasuriya, A., and Hockman, C. H.: Perineurial permeability to sodium during wallerian degeneration in rat sciatic nerve. Brain Res. 587:327, 1992.)

Endoneurial Homeostasis during Development During normal development, results from studies employing morphologic tracers indicate that the BNI of juvenile animals is more permeable to macromolecules than that of adult animals.23,34 This decrease in the PS of the BNI during development has been quantified with radiotracer studies.95 An intriguing issue arising out of these observations is how developing nerve, with its highly permeable BNI, avoids the

Wet weight to dry weight ratio

BNI PS to 22Na (mL . g−1 . sec−1) ⫻ 104

Delineation of the properties of the BNI during nerve regeneration provides a more comprehensive picture.92 From fourth day up to the eighth week after the crush lesion, BNI PS to 22Na was significantly greater than the normal value, but by the 18th week after the crush, PS was not significantly different from the normal value. The data in Figure 29–5 demonstrate the easier access of water into the endoneurial compartment for up to at least 8 weeks after the crush lesion. This increased ability of water to move into the endoneurial space is compared with the

measured endoneurial water content at the same time points (Fig. 29–6). The endoneurial water content, calculated as the wet weight–to–dry weight ratio, increased from the fourth day and remained elevated during the entire experimental period. Its peak at the second week postcrush corresponds to a period of increased PS of the BNI to 22Na. However, at 18 weeks postcrush, when the BNI PS to 22Na is normal, the nerve is still edematous. What, then, is the driving force for the elevated endoneurial water content? Although an increased BNI PS to plasma macromolecules could elevate the osmolality of the endoneurial fluid and thus draw water into this space, it is contended that a decrease in the endoneurial fluid turnover is also a factor in maintaining the increased water content of the regenerated nerve. Conversely, the elevated water content is not associated with nerve edema; at 12 weeks after a crush injury the EHP of the regenerating nerve has returned to normal.98 It is postulated that an altered compliance of the regenerated nerve helps to maintain EHP in the normal range.

14 12 10 8 6 4 2 0

7 6 5 4 3 0

0

2

4

6

8

10

12

14

16

18

Time (wk)

FIGURE 29–5 BNI permeability to 22Na during regeneration after a crush lesion. Time after the lesion is represented on the x axis. There were four animals at the 4-day and 1-week time points, and five animals at all other time points. (Data from Weerasuriya, A.: Changes in the permeability coefficient-surface area product [PS] of the bloodnerve interface [BNI] to sodium in degenerating and regenerating rat sciatic nerve. Paper presented at the 11th Biennial Meeting of the Peripheral Nerve Study Group, Aachen, Germany, July 1993.)

2

4

6

8

10

12

14

16

18

Time (wk)

FIGURE 29–6 Changes in endoneurial wet weight-to-dry weight ratio in desheathed rat sciatic nerve following a crush injury. Data are presented as mean ⫾ SEM. The SEMs of normal controls and of injured nerve at 6, 8, and 18 weeks are smaller than the size of the symbol. (Data from Weerasuriya, A.: Changes in the permeability coefficient-surface area product [PS] of the blood-nerve interface [BNI] to sodium in degenerating and regenerating rat sciatic nerve. Paper presented at the 11th Biennial Meeting of the Peripheral Nerve Study Group, Aachen, Germany, July 1993.)

660

Functional Compartments and Nerve-Fiber Environment

pathology observed in an adult nerve with a comparably permeable BNI. For example, in the adult an elevated PS causing BNI extravasation of plasma proteins inevitably leads to endoneurial edema and increased EHP. The primary event in that pathology is the increase of endoneurial albumin content elevating the oncotic pressure in the endoneurial interstitial space. Because of the operation of Starling forces, water is drawn into the endoneurial interstitium from the vascular compartment. This leads to an elevated endoneurial water content and, together with the low hydraulic conductivity and compliance of the perineurium, causes an increase in the EHP. This elevated EHP can compromise the intrafascicular microcirculation, leading to ischemia and its accompanying pathology. By contrast, a highly permeable BNI during the first weeks of development does not lead to an elevated EHP and edema because the endoneurial accumulation of fluids and osmolytes is prevented by a more permeable perineurium, allowing for the clearance of this material by the epineurial lymphatics. Additionally, a higher compliance of the juvenile endoneurial tissue mass could also counter a tendency for an increased EHP. Therefore, in the developing nerve a much more permeable endoneurial microvasculature (which is probably necessary for greater blood-nerve exchange of material to support the more active metabolism, and is also a consequence of tight junction dissolution and reformation during endothelial proliferation) does not lead to pathologic consequences because of the combination of a relatively permeable perineurium and epineurial lymphatics, and metabolic clearance of plasma-derived macromolecules.

FACTORS REGULATING BNI PERMEABILITY Microvessels in the endoneurium are the least permeable capillaries except for cerebral capillaries. The investing perineurium is even less permeable than the endoneurial capillaries. Hence, cellular components of the endoneurium secrete soluble factors responsible for maintaining the integrity of the tight junctional complexes and associated actin cytoskeletal components of endoneurial vascular endothelial cells and of cells composing the innermost layers of the perineurium. It should be recognized that endoneurial elements not only expend metabolic energy to maintain the junctional integrity of the endoneurium’s barrier tissues, but are at the same time obliged to devote metabolic energy to maintain the pumps and transporters for blood-nerve exchange of critical material needed by the endoneurial cellular elements. Thus the dynamically selective per-

meability properties of the BNI extract a considerable metabolic cost.

Neurovascular Unit in Nerve When considering the homeostatic mechanisms of the cerebral microenvironment, an important concept that is emerging with increasing validity and support from experimental data is that of the neurovascular unit.2,26 The structural components of the neurovascular unit are the cerebral endothelium, the neurons and astrocytes that make intimate contact with the endothelial cells, pericytes, and the extracellular matrix investing this functional unit. Functionally this unit subserves dynamic signaling among the vascular, glial, neuronal, and matrix elements of the cerebral microenvironment. There is direct synaptic and perisynaptic contact among these elements to effect changes extremely rapidly over very short distances. The modular organization of this unit and its microscopic size imply that it is repeated many times to limit control and regulation of the cerebral microvasculature to discrete small volumes of cerebral parenchyma. Signaling in the neurovascular unit is mostly directed at the cerebral endothelium to influence two important physiologic parameters of the cerebral microcirculation: localized cerebral blood flow and capillary permeability. Thus the neurovascular unit plays a crucial role in maintaining an adequate supply of oxygen and nutrients to neurons and glial cells. It plays an even more critical role in cerebral homeostasis by responding to changes in nutrient needs imposed by both physiologic and potentially pathologic stressors. The current concept of the neurovascular unit emphasizes the capillary endothelium, but it is expected that this concept will evolve to encompass both cerebral arterioles and venules because these components of the microcirculation play significant roles in regulating vascular resistance and blood-brain water exchange. What is the relevance of the neurovascular unit for the endoneurial microenvironment? Functionally, axons and Schwann cells can tolerate limited hypoxia and metabolic insults for brief periods of time (seconds to perhaps a few minutes) without suffering irreversible damage. Structurally, only pericytes are within a few microns of the endothelial cells and invested by a common basal lamina. Nevertheless, focal ischemic lesions in peripheral nerve are probably indicative of the presence of functional neurovascular units in the endoneurial microenvironment, with two important differences from their cerebral counterparts. In peripheral nerve, given the greater separation between axons and endothelial cells and between Schwann cells and endothelial cells, signaling molecules have to diffuse through greater distances to reach their targets. Hence the neurovascular unit in nerve, compared to that in cerebrum, represents a functional unit that operates in a larger volume and respond more slowly to perturbations.

Blood-Nerve Interface and Endoneurial Homeostasis

Cellular Control of BNI Permeability The perineurium and the endoneurial microvasculature form the BNI, which isolates the endoneurium by virtue of the “tight” intercellular junctions present in these two structures. Perineurial microvessels that are permeable to plasma macromolecules acquire tight intercellular junctions upon penetrating the innermost layers of the perineurium to enter the endoneurium.50 Thus one or more diffusible factors present in the endoneurium is responsible for the induction of tight intercellular junctions in barrier tissues circumscribing the endoneurium. Such an inducer is likely to be produced by axons, Schwann cells, or both, and regulate synthesis and elaboration of intercellular junctional proteins by perineurial and endothelial cells. In chronically transected nerves, the near-normal PS of the endoneurial endothelium to small nonelectrolytes,90 and the presence of “tight” intercellular junctions in the endothelium,25 strongly argue against axons as being the source of a regulatory factor for the maintenance of the tight junctions between vascular endothelial cells. However, the increased permeability of the perineurium accompanying axonal degeneration25,86,103 strongly suggests that axonally derived factors might be responsible for maintaining the relatively tight barrier properties of the perineurium. This working hypothesis on a dual source of regulatory proteins for modulating the permeability properties of the endoneurial vascular endothelium and perineurium needs to be rigorously tested by examining BNI permeability when either axons or Schwann cells are preferentially damaged by endoneurial injection of toxins. A corollary of this hypothesis is that Schwann cells have several phenotypes, and a major function of some of these subsets is regulation of endoneurial vascular permeability. This is accomplished through soluble factors released into the endoneurial extracellular space and diffusing less than 100 ␮ to reach the intended targets, the endothelial cells. This would be a form of paracrine signaling within a confined and specialized extracellular space. Conversely, based on reasons outlined earlier, perineurial junctional integrity appears to depend primarily on soluble factors secreted by axons and perhaps aided by Schwann cell–derived signaling molecules. A role for Schwann cells is postulated because of the expression of vascular endothelial cadherins by perineurial cells.73 Hence it is hypothesized that the baseline resting permeability of the two components of the BNI are regulated by axons and Schwann cells. An integral aspect of endoneurial homeostasis is the ability of the BNI to adaptively alter its permeability properties to meet the changing needs of endoneurial elements. Such changes may be dictated by programmed and gradual changes associated with growth and maturation, and microenvironmental disturbances precipitated by disease and trauma. How might these adaptive alterations in permeability of the BNI be initiated and effected?

661

Recent evidence strongly implicates the immune system as an active modulator of BNI permeability in a whole host of conditions ranging from trauma to metabolic neuropathies to vascular disorders.18,55,58,110 Thus it is postulated that hematogenous elements interacting directly with the endoneurial vascular elements effect increases in capillary permeability. The perineurium appears to be unusually resistant to inflammatory mediators.1 Hence immunomodulation of the BNI is expected to be limited to the nerve endothelial cells.

Molecular Basis of BNI Permeability Regulation Even though the molecular biology of junctional complexes of the BNI have not been investigated in detail, it is reasonable to assume that they are essentially similar to those of the cerebral microvasculature.5,17 Thus mediators of BNI permeability alterations have to utilize one or more of the following mechanisms: (1) modification of the actin cytoskeleton responsible for anchoring the junctional complexes to the intracellular scaffolding; (2) posttranslational modification of the constituent proteins (occludins, family of zonula occludens proteins, cadherins, claudins, catenins, etc.) of the junctional complexes; and (3) proteolytic degradation of the junctional components. Whereas the first two modes require signaling molecules to gain intracellular access, the last mode may be initiated and completed extracellularly at the junction itself. The classic cytokines and interleukins mediate inflammation-related changes in BBB. They are implicated to play a role in disease-related BNI changes as well. What are the events that initiate permeability changes and how are they sensed? It is not know whether hypoxia inducible factor plays an active role in endoneurial response to ischemia and hypoxia. Does the endoneurium contain angiotensin type 1 receptors that are activated by mechanical stress?111 They would be very attractive candidates for detecting changes in endoneurial pressure.

THERAPEUTIC IMPLICATIONS FOR PERIPHERAL NEUROPATHIES Some neuropathies are a direct outcome of microcirculatory failure in peripheral nerve, while other forms of neuropathy are not associated with microangiopathic failures, and a few exhibit a microcirculatory insufficiency which does not appear to have been the precipitating factor. What role can modification of the BNI play in the treatment of peripheral nerve diseases and trauma? There are at least three strategies. First, an increase in BNI permeability could deliver a higher concentration of therapeutics molecules to their targets. This could be

662

Functional Compartments and Nerve-Fiber Environment

achieved by modifying the drug molecules to increase their passive permeance across the BNI or utilize BNI transporters to gain better access to axons, Schwann cells, and other cells in the endoneurial compartment. Secondly, increased angiogenesis and endoneurial blood flow will directly address those clinical entities with microcirculatory failure. This would also indirectly facilitate the previous strategy. Finally, gene transfer has been successfully employed to treat a variety of nerve disorders in animal models. Therapeutic concentrations of several neurotrophic factors have been delivered with the following methods: intratissue injection of naked DNA, adenoviral gene transfer system. T-lymphocytes transduced with a recombinant retrovirus, and a herpes simplex-based vector. With accelerating advances in gene transfer technology, this last strategy is bound to play an increasing role in the treatment of peripheral neuropathies.

CONCLUSION It is advocated that the term blood-nerve interface is preferable to blood-nerve barrier. The former term encompasses the regulatory nature as well as the passive barrier properties of the perineurium and the endoneurial microvasculature. Historically, the term blood-nerve barrier, derived from the related term for the cerebral vasculature (blood-brain barrier), owes its origin to morphologic demonstrations of the inability of certain tracers to enter the immediate interstitial spaces of axons and neurons, and hence implied an all-or-none nature of the permeability of this barrier to tracers of various sizes. The original scientific term was, in German, Blut-Hirn Schrank, and a literal translation of that term would be “blood-brain cabinet,” with a strong emphasis on its passive barrier properties! As we learn more about the structure and functions of the perineurium and endoneurial microvasculature, especially the graded nature of its permeability and its dynamic adaptive responses to alterations in the endoneurial microenvironment, it seems more appropriate to refer to them as the blood-nerve interface. This conceptual thrust is likely to play a significant role in our understanding of peripheral nerve disorders and the evolution of therapeutic strategies and paradigms for the treatment of these disorders.

ACKNOWLEDGMENTS Experiments reported herein from the author’s laboratory were partially supported by grants RO1–NS30197 (National Institutes of Health), IBN-9420525 (National Science Foundation), and 23750 (MedCen Foundation). This chapter is dedicated to the memory of KNS (1929–1987) and KJN (1927–2001). Ms. Marianne Watkins

provided indispensable secretarial assistance during the preparation of this manuscript. My sincerest apologies are extended to colleagues whose publications are inadequately cited due to constraints of time and space.

REFERENCES 1. Abbott, N. J.: Inflammatory mediators and modulation of blood-brain barrier permeability. Cell. Mol. Neurobiol. 20:131, 2000. 2. Abbott, N. J.: Astrocyte-endothelial interactions and bloodbrain barrier permeability. J. Anat. 200:629, 2002. 3. Abbott, N. J.: Evidence for bulk flow of brain interstitial fluid: significance for physiology and pathology. Neurochem Int. 45:545, 2004. 4. Akert, K., Sandri, C., Weibel, E. R., et al.: The fine structure of the perineurial endothelium. Cell Tissue Res. 165:281, 1976. 5. Anderson, C. M., and Nedergaard, M.: Astrocyte-mediated control of cerebral microcirculation. Trends Neurosci. 26:340, 2003. 6. Bell, M. A., and Weddell, A. G. M.: A morphometric study of intrafascicular vessels in mammalian sciatic nerve. Muscle Nerve 7:524, 1984. 7. Boddingius, J.: Ultrastructural and histophysiological studies on the blood-nerve barrier and perineurial barrier in leprosy neuropathy. Acta Neuropathol. (Berl.) 64:282, 1984. 8. Bradbury, M. W. B., and Crowder, J.: Compartments and barriers in the sciatic nerve of the rabbit. Brain Res. 103:515, 1976. 9. Bundgaard, M., and Frokjaer-Jensen, J.: Functional aspects of the ultrastructure of terminal blood vessels: a quantitative study on consecutive segments of the frog mesenteric microvasculature. Microvasc. Res. 23:1, 1982. 10. Butt, A. M., Jones, H. C., and Abbott, N. J.: Electrical resistance across the blood-brain barrier in anaesthetized rats: a developmental study. J. Physiol. (Lond.) 429:47, 1990. 11. Crone, C.: The blood-brain barrier: a modified tight epithelium. In Suckling, A. J., Rumsby, M. G., and Bradbury, M. W. B. (eds.): The Blood-Brain Barrier in Health and Disease. Chichester, UK, Ellis Harwood Ltd, p. 17, 1986. 12. Crone, C., and Levitt, D. G.: Capillary permeability to small solutes. In Renkin, E. M., and Michel, C. C. (eds.): Handbook of Physiology, Section 2. The Cardiovascular System, Vol. IV. Microcirculation. Bethesda, MD, American Physiological Society, p. 411, 1984. 13. Crowley, N. R., Nelson, S. L., Weerasuriya, A., and Black, A. C.: Endoneurial hydrostatic pressure in rat sciatic nerve during development. Soc. Neurosci. Abstr. 22:772, 1996. 14. Enerback, L., Olsson, Y., and Sourander, P.: Mast cells in normal and sectioned peripheral nerve. Z. Zellforsch. 66:596, 1965. 15. Froehner, S. C., Davies, A., Baldwin, S. A., and Lienhard, G. E.: The blood-nerve barrier is rich in glucose transporter. J. Neurocytol. 17:173, 1988.

Blood-Nerve Interface and Endoneurial Homeostasis 16. Gerhart, D. Z., and Drewes, L. R.: Glucose transporters at the blood-nerve barrier are associated with perineurial cells and endoneurial microvessels. Brain Res. 508:46, 1990. 17. Gloor, S. M., Wachtel, M., Bolliger, M. F., et al.: Molecular and cellular permeability control at the blood-brain barrier. Brain Res. Rev. 36:258, 2001. 18. Griffin, J. W.: Vasculitic neuropathies. Rheum. Dis. Clin. North Am. 27:751, 2001. 19. Grotte, G.: Passage of dextran molecules across the bloodlymph barrier. Acta Chir. Scand. 211:1, 1956. 20. Haller, F. R., Haller, C., and Low, F. N.: The fine structure of cellular layers and connective tissue space at spinal nerve root attachment in the rat. Am. J. Anat. 133:109, 1972. 21. Haller, F. R., and Low, F. N.: The fine structure of the peripheral nerve root sheath in the subarachnoid space in the rat and other laboratory animals. Am. J. Anat. 131:1, 1971. 22. Karnovsky, M. J.: Morphology of capillaries with special reference to muscle capillaries. In Crone, C., and Lassen, N. A. (eds.): Capillary Permeability. Copenhagen, Munksgaard, p. 341, 1970. 23. Kristensson, K., and Olsson, Y.: The perineurium as a diffusion barrier to protein tracers: differences between mature and immature animals. Acta Neuropathol. (Berl.) 17:127, 1971. 24. Krnjevic, K.: Some observations on perfused frog sciatic nerves. J. Physiol. (Lond.) 123:338, 1954. 25. Latker, C. H., Wadhwani, K. C., Balbo, A., and Rapoport, S. I.: Blood-nerve barrier in the frog during wallerian degeneration: are axons necessary for maintenance of barrier function? J. Comp. Neurol. 309:650, 1991. 26. Lo, E. H., Broderick, J. P., and Moskowitz, M. A.: tPA and proteolysis in the neurovascular unit. Stroke 35:354, 2004. 27. Low, F. N.: The perineurium and connective tissue of peripheral nerve. In Landon, D. N. (ed.): The Peripheral Nerve. London, Chapman & Hall, p. 159, 1976. 28. Low, P. A.: Endoneurial potassium is increased and enhances spontaneous activity in regenerating mammalian nerve fibers: implications for neuropathic positive symptom. Muscle Nerve 8:27, 1985. 29. Low, P. A., and Dyck, P. J.: Increased endoneurial fluid pressure in experimental lead neuropathy. Nature 269:427, 1977. 30. Low, P. A., Marchand, G., Knox, F., and Dyck, P. J.: Measurement of endoneurial fluid pressure with polyethylene matrix capsule. Brain Res. 122:373, 1977. 31. Low, P. A., and Tuck, R. R.: Effects of changes in blood pressure, respiratory acidosis and hypoxia on blood flow in the sciatic nerve of the rat. J. Physiol. (Lond.) 347:513, 1984. 32. Mackenzie, M. L., Ghabriel, M. N., and Allt, G.: The bloodnerve barrier: an in vivo lanthanum tracer study. J. Anat. 154:27, 1987. 33. Magnani, P., Cherian, P. V., Gould, G. W., et al.: Glucose transporters in rat peripheral nerve: paranodal expression of GLUT1 and GLUT3. Metabolism 45:1466, 1996. 34. Malmgren, L. T., and Brink, J. J.: Permeability barriers to cytochrome-C in nerves of adult and immature rats. Anat. Rec. 181:755, 1975. 35. McCabe, J. S., and Low, F. N.: The subarachnoid angle: an area of transition in peripheral nerve. Anat. Rec. 164:15, 1969.

663

36. Mellick, R. S., and Cavanagh, J. B.: Longitudinal movement of radio-iodinated albumin within extravascular spaces of peripheral nerves following three systems of trauma. J. Neurol. Neurosurg. Psychiatry 30:458, 1967. 37. Michel, M. E., Shinowara, N. L., and Rapoport, S. I.: Presence of a blood-nerve barrier within blood vessels of frog sciatic nerve. Brain Res. 299:25, 1984. 38. Mizisin, A. P., and Kalichman, M. W.: Permeability and surface area of the blood-nerve barrier in galactose intoxication. Brain Res. 618:109, 1993. 39. Morris, J. H., Hudson, A. R., and Weddell, G.: A study of degeneration and regeneration in the divided rat sciatic nerve based on electron microscope observations. Parts I–III. Z. Zellforsch. 123:76, 1972. 40. Myers, R. R., Heckman, H. M., Galbraith, J. A., and Powell, H. C.: Subperineurial demyelination associated with reduced nerve blood flow and oxygen tension after epineurial vascular stripping. Lab. Invest. 65:41, 1991. 41. Myers, R. R., Powell, H. C., Costello, M. L., et al.: Endoneurial fluid pressure: direct measurement with micropipettes. Brain Res. 148:510, 1978. 42. Myers, R. R., Powell, H. C., Shapiro, H. M., et al.: Changes in endoneurial fluid pressure, permeability and peripheral nerve ultrastructure in experimental lead neuropathy. Ann. Neurol. 8:392, 1980. 43. Myers, R. R., Rydevik, B. L., Heckman, H. M., and Powell, H. C.: Proximodistal gradient in endoneurial fluid pressure. Exp. Neurol. 102:368, 1988. 44. Nesbitt, J. A., and Acland, R. D.: Histopathological changes following removal of the perineurium. J. Neurosurg. 53:233, 1980. 45. Nicely, M.: Measurement of the potential difference across the connective tissue sheath of frog sciatic nerve. Experientia 8:199, 1955. 46. Nukada, H., Powell, H. C., and Myers, R. R.: Perineurial window: demyelination in nonherniated endoneurium with reduced nerve blood flow. J. Neuropathol. Exp. Neurol. 51:523, 1992. 47. Ohi, T., Poduslo, J. F., Curran, G. L., and Dyck, P. J.: Quantitative methods for detection of blood-nerve barrier alterations in experimental animal models of neuropathy. Exp. Neurol. 90:365, 1985. 48. Ohno, K., Pettigrew, K., and Rapoport, S. I.: Lower limits of cerebrovascular permeability to nonelectrolytes in the conscious rat. Am. J. Physiol. Heart Circ. Physiol. 235:H299, 1978. 49. Olsson, Y.: Studies on vascular permeability in peripheral nerve I. Acta Neuropathol. (Berl.) 7:1, 1966. 50. Olsson, Y.: Studies on vascular permeability in peripheral nerve II. Acta Physiol. Scand. 284:1, 1966. 51. Olsson, Y.: Vascular permeability in the peripheral nervous system. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. P. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 579, 1984. 52. Olsson, Y., and Reese, T. S.: Permeability of vasa nervorum and perineurium in mouse sciatic nerve studied by fluorescence and electron microscopy. J. Neuropathol. Exp. Neurol. 30:105, 1971. 53. Pappenheimer, J. R.: Passage of molecules through capillary walls. Physiol. Rev. 33:387, 1953.

664

Functional Compartments and Nerve-Fiber Environment

54. Poduslo, J. F., Curran, G. L., and Dyck, P. J.: Increase in albumin, IgG and IgM blood-nerve barrier indices in human diabetic neuropathy. Proc. Natl. Acad. Sci. U. S. A. 88:4879, 1988. 55. Polydefkis, M., Griffin, J. W., and McArthur, J.: New insights into diabetic polyneuropathy. JAMA 290;1371, 2003. 56. Powell, D. W.: Barrier function of epithelia. Am. J. Physiol. Gastrointest. Liver Physiol. 241:G275, 1981. 57. Powell, H. C., and Myers, R. R.: The blood-nerve barrier and the pathological significance of nerve edema. In Neuwelt, E. A. (ed.): Implications of the Blood-Brain Barrier and Its Manipulation, Vol. 1. New York, Plenum Press, p. 199, 1989. 58. Powell, H. C., and Myers, R. R.: Impact of inflammatory disease on the nerve microenvironment. J. Neurol. Sci. 220:131, 2004. 59. Powell, H. C., Myers, R. R., and Lampert, P. W.: Changes in Schwann cells and vessels in lead neuropathy. Am. J. Pathol. 109:193, 1982. 60. Rapoport, S. I.: Blood-Brain Barrier in Physiology and Medicine. New York, Raven Press, 1976, p. 45. 61. Reale, E., Luciano, L., and Spitznas, M.: Freeze-fracture faces of the perineurial sheath of the rabbit sciatic nerve. J. Neurocytol. 4:261, 1975. 62. Rechthand, E., Smith, Q. R., and Rapoport, S. I.: Facilitated transport of glucose from blood into peripheral nerve. J. Neurochem. 45:957, 1985. 63. Rechthand, E., Smith, Q. R., and Rapoport, S. I.: Transfer of nonelectrolytes from blood into peripheral nerve endoneurium. Am. J. Physiol. Heart Circ. Physiol. 252:H1175, 1987. 64. Rechthand, E., Smith, Q. R., and Rapoport, S. I.: A compartmental analysis of solute transfer and exchange across blood-nerve barrier. Am. J. Physiol. 255:317, 1988. 65. Reed, R. K., Johansen, S., and Noddeland, H.: Turnover rate of interstitial albumin in rat skin and skeletal muscle: effect of limb movements and motor activity. Acta Physiol. Scand. 125:711, 1985. 66. Renkin, E. M.: Multiple pathways of capillary permeability. Circ. Res. 41:735, 1977. 67. Rippe, B., and Haraldsson, B.: How are macromolecules transported across the capillary wall? News Physiol. Sci. 2:135, 1987. 68. Rundquist, I., Smith, Q. R., Michael, M. E., et al.: Sciatic nerve blood flow measured by laser Doppler flowmetry and [14C] iodoantipyrine. Am. J. Physiol. Heart Circ. Physiol. 248:H311, 1985. 69. Rydevik, B. L., Kwan, M. K., Myers, R. R., et al.: Effects of acute stretching on rabbit tibial nerve: an in vitro mechanical and histological study. J. Orthop. Res. 8:694, 1990. 70. Schmelzer, J. D., and Low, P. A.: The effects of hyperbaric oxygenation and hypoxia on the blood-nerve barrier. Brain Res. 473:321, 1988. 71. Shantha, T. S., and Bourne, G. H.: The “perineurial epithelium,” a metabolically active, continuous, protoplasmic cell barrier surrounding peripheral nerve fasciculi. J. Anat. 96:527, 1962. 72. Shinowara, N. L., Michel, M. E., and Rapoport, S. I.: Morphological correlates of permeability in the frog

73.

74. 75.

76. 77.

78.

79. 80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

perineurium: vesicles and transcellular channels. Cell Tissue Res. 227:11, 1982. Smith, M. E., Jones, T. A., and Hilton, D.: Vascular endothelial cadherin is expressed by perineurial cells of peripheral nerve. Histopathology 32:411, 1998. Smith, Q. R., and Allen, D. D.: In situ brain perfusion technique. Methods Mol. Med. 89:209, 2003. Spencer, P. S., Weinberg, H. J., and Raines, C. S.: The perineurial window: a new model of focal demyelination and remyelination. Brain Res. 96:323, 1975. Soderfeldt, B.: The perineurium as diffusion barrier to protein tracers. Acta Neuropathol. (Berl.) 27:55, 1974. Stoll, G., Jander, S., and Myers, R. R.: Degeneration and regeneration in the peripheral nervous system: from August Waller’s observation to neuroinflammation. J. Peripher. Nerv. Syst. 7:13, 2002. Stoll, G., Trapp, B. D., and Griffin, J. W.: Macrophage function during wallerian degeneration of rat optic nerve: clearance of degenerating myelin and Ia expression. J. Neurosci. 9:2327, 1989. Sunderland, S.: Nerves and Nerve Injury. Edinburgh, Livingstone, p. 55, 1978. Taylor, A. E., and Granger, D. N.: Exchange of macromolecules across the microcirculation. In Renkin, E. M., and Michel, C. C. (eds.): Handbook of Physiology: Section 2. The Cardiovascular System, Vol. IV. Microcirculation. Bethesda, MD, American Physiological Society, p. 467, 1984. Terho, P. M., Vuorinen, V. S., and Roytta, M.: The endoneurial response to microsurgically removed epi- and perineurium. J. Peripher. Nerv. Syst. 7:155, 2002. Thomas, P. K., and Bhagat, S.: The effect of extraction of the intrafascicular contents of peripheral nerve trunks on perineurial structure. Acta Neuropathol. (Berl.) 43:135, 1978. Thomas, P. K., Ochoa, J., Berthold, C.-H., et al.: Microscopic anatomy of the peripheral nervous system. In Dyck, P. J., and Thomas, P. K. (eds): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 28, 1993. Tsao, J. W., George, E. B., and Griffin, J. W.: Temperature modulation reveals three distinct stages of wallerian degeneration. J. Neurosci. 19:4718, 1999. Tserentsoodol, N., Shin, B. C., Koyama, H., et al.: Immunolocalization of tight junction proteins, occludin and ZO-1, and GLUT1 in the cells of the blood-nerve barrier. Arch. Histol. Cytol. 62:459, 1999. Wadhwani, K. C., Latker, C. H., Balbo, A., and Rapoport, S. I.: Perineurial permeability and endoneurial edema during wallerian degeneration of the frog peripheral nerve. Brain Res. 493:231, 1989. Wadhwani, K. C., Smith, Q. R., and Rapoport, S. I.: Facilitated transport of L-phenylalanine across the bloodnerve barrier of rat sciatic nerve. Am. J. Physiol. Regul. Integr. Comp. Physiol. 258:R1436, 1990. Waksman, B. H.: Experimental study of diphtheritic polyneuritis in the rabbit and guinea pig. III. The bloodnerve barrier in the rabbit. J. Neuropathol. Exp. Neurol. 20:35, 1961. Weerasuriya, A.: Permeability of endoneurial capillaries to K, Na, and Cl and its relation to peripheral nerve excitability. Brain Res. 419:188, 1987.

Blood-Nerve Interface and Endoneurial Homeostasis 90. Weerasuriya, A.: Patterns of change in endoneurial capillary permeability and vascular space during wallerian degeneration. Brain Res. 445:181, 1988. 91. Weerasuriya, A.: Permeabilities of endoneurial capillaries and perineurium of rat sciatic nerve to 22Na during development. Physiologist 33:A100, 1990. 92. Weerasuriya, A.: Changes in the permeability coefficientsurface area product (PS) of the blood-nerve interface (BNI) to sodium in degenerating and regenerating rat sciatic nerve. Paper presented at the 11th Biennial Meeting of the Peripheral Nerve Study Group, Aachen, Germany, July 1993. 93. Weerasuriya, A., Coath, G., and Crowley, N.: Endoneurial Na concentration and hydrostatic pressure in galactose neuropathy. FASEB J. 10:A764, 1996. 94. Weerasuriya, A., Curran, G. L., and Poduslo J. F.: Blood-nerve transfer of albumin and its implications for the endoneurial microenvironment. Brain Res. 494:114, 1989. 95. Weerasuriya, A., Curran, G. L., and Poduslo, J. F.: Developmental changes in blood-nerve transfer of albumin and endoneurial albumin content in rat sciatic nerve. Brain Res. 521:40, 1990. 96. Weerasuriya, A., Curran, G. L., and Poduslo, J. F.: Physiological changes in the sciatic nerve endoneurium of lead intoxicated rats: a model of endoneurial homeostasis. Brain Res. 517:1, 1990. 97. Weerasuriya, A., and Hockman, C. H.: Perineurial permeability to sodium during wallerian degeneration in rat sciatic nerve. Brain Res. 587:327, 1992. 98. Weerasuriya, A., Nelson, S. L., and Crowley, N. R.: Evidence for endoneurial fluid flow from endoneurial hydrostatic pressure measurements in transected and crushed rat sciatic nerves. Soc. Neurosci. Abstr. 24:267, 1998. 99. Weerasuriya, A., Nelson, S. L., and Crowley, N. R.: Development of an entrapment neuropathy is a two-stage process. Paper presented at the Peripheral Nerve Society Meeting, San Diego, July 1999.

665

100. Weerasuriya, A., and Rapoport, S. I.: Endoneurial capillary permeability to [14C] sucrose in frog sciatic nerve. Brain Res. 375:150, 1986. 101. Weerasuriya, A., Rapoport, S. I., and Taylor, R. E.: Modification of permeability of frog perineurium to [14C]sucrose by stretch and hypertonicity. Brain Res. 173:503, 1979. 102. Weerasuriya, A., Rapoport, S. I., and Taylor, R. E.: Ionic permeabilities of the frog perineurium. Brain Res. 191:405, 1980. 103. Weerasuriya, A., Rapoport, S. I., and Taylor, R. E.: Perineurial permeability increases during wallerian degeneration. Brain Res. 192:581, 1980. 104. Weerasuriya, A., Spangler, R. A., Rapoport, S. I., and Taylor, R. E.: AC impedance of the perineurium of the frog sciatic nerve. Biophys. J. 46:167, 1984. 105. Weiss, P., Wang, H., Taylor, A. C., and Edds, M. V.: Proximodistal fluid convection in the endoneurial spaces of peripheral nerves, demonstrated by colored and radioactive (isotope) tracers. Am. J. Physiol. 143:521, 1945. 106. Welch, K., and Davson, H.: The permeability of capillaries of the sciatic nerve of the rabbit to several materials. J. Neurosurg. 36:21, 1972. 107. Williams, P. L., and Hall, S. M.: Chronic wallerian degeneration: an in vivo and ultrastructural study. J. Anat. 109:487, 1971. 108. Windebank, A. J.: Metal neuropathy. In Dyck, P. J., and Thomas, P. K. (eds.): Peripheral Neuropathy. 3rd ed. Philadelphia, W. B. Saunders, p. 1549, 1993. 109. Windebank, A. J., McCall, J. T., Hunder, H. G., and Dyck, P. J.: The endoneurial content of lead related to the onset and severity of segmental demyelination. J. Neuropathol. Exp. Neurol. 39:692, 1980. 110. Zhou, L., and Griffin, J. W.: Demyelinating polyneuropathies. Curr. Opin. Neurol. 16:307, 2003. 111. Zou, Y., Akazawa, H., Qin, Y., et al.: Mechanical stress activates angiotensin II type 1 receptor without the involvement of angiotensin II. Nature Cell Biol. 6:499, 2004.

30 Nerve Blood Flow and Microenvironment PHILIP G. MCMANIS, PHILLIP A. LOW, AND TERRENCE D. LAGERLUND

Overview Unique Aspects of Special Physiology of Nerve Microvasculature Poor Autoregulation of Peripheral Nerve Peripheral Nerve as a NutritiveCapacitance Microvascular System

Peripheral Nerve Adaptation to Hypoxic Stress Regulation of Nerve Blood Flow Intrinsic Mechanisms Extrinsic Mechanisms Experimental Nerve Ischemia Models of Nerve Ischemia

OVERVIEW Peripheral nerve is unique in a number of respects. It is dependent on its communications with its parent cell body situated a great distance away. Indeed, the length:diameter ratio exceeds 109, so that export of nutrients and structural proteins is very inefficient. The function and indeed the survival of peripheral nerve depend also on its interactions with its microenvironment. The focus of this chapter is on experimental nerve ischemia as a particular component of this microenvironment. Because of space limitations, only cursory coverage is provided to other aspects of nerve microenvironment.

UNIQUE ASPECTS OF SPECIAL PHYSIOLOGY OF NERVE MICROVASCULATURE Peripheral nerve axon, especially its distal portion, is extremely long relative to its diameter and is a great distance from its parent cell body. The regulation and maintenance of nerve depend on the bidirectional axonal transport of structural proteins and growth factors from its neuron. It is exquisitely dependent on nerve microenvironment for its blood supply, oxygenation, and nutrition, and for the removal of toxic metabolic products. The phys-

Factors that Modulate the Effects of Nerve Ischemia Influence of Diabetic State on Nerve Ischemia Reperfusion Injury Pathogenesis of Nerve Ischemia

iology of nerve microenvironment has a number of unique characteristics that are discussed in this section.

Poor Autoregulation of Peripheral Nerve Autoregulation refers to the maintenance of constant blood flow, within a range of blood pressures (BPs) (the autoregulated range), when blood pressure is changed. Autoregulation is achieved by varying arteriolar tone and is achieved by myogenic rather than neurogenic mechanisms in most tissues. We demonstrated that peripheral nerve did not autoregulate.56 We found a curvilinear relationship of nerve blood flow (NBF) to mean BP, a characteristic of a passive system. Thus mammalian peripheral nerve autoregulates poorly, if at all,56 a finding that has subsequently been confirmed in anesthetized90 and nonanesthetized rats.99 When the rate of exsanguination is altered, the shape of the NBF:BP curve also changes. With steady-state recordings of blood flow using hydrogen (H2) polarography, a curvilinear relationship occurs.56 When exsanguination is rapid and blood flow is monitored in real time using laser Doppler velocimetry, a linear relationship occurs.100 The difference may relate to the time-dependent response of arterioles to intravascular pressure alterations.11 The concept of autoregulating and nonautoregulating systems is artificial because different organs autoregulate to 667

668

Functional Compartments and Nerve-Fiber Environment

different degrees. Furthermore, for intensively studied tissues such as brain, there is regional, temporal, and segmental heterogeneity of autoregulation. This range of autoregulatory capacities means that there is a momentto-moment distribution and redistribution of blood in such a way that tissues that have the greatest metabolic needs will have relatively the most constant blood flow, whereas tissues with the least needs will have blood flow that varies the most. The answer to the question as to whether a tissue is autoregulated or not should be: relative to what tissue? There is indeed heterogeneity of innervation of microvessels. For instance, dense noradrenergic and peptidergic innervation of vasa nervorum has been demonstrated,3 and dense innervation of small arteries is well known.15,16 Adrenergic innervation is reportedly increased in experimental diabetic neuropathy.25 These apparently disparate observations may relate to the regional and segmental heterogeneity of blood flow regulation described in other tissues.5,9 Regulation of blood flow in these tissues is largely neurogenic in vessels between 50 to 200 ␮m and largely myogenic in vessels less than 50 ␮m.29 In brain, nutritive vessel regulation appears to be metabolic and has been suggested to be due to adenosine.9 There is also considerable variation from one vessel bed to another. For instance, reactivity of carotid and mesenteric vessels is quite different. There is a characteristic NBF:BP curve, flat within the autoregulated range and sloping at either end.

Peripheral Nerve as a Nutritive-Capacitance Microvascular System Nerve microvasculature is physiologically unique in that the capillaries are much larger in caliber than those of the muscle bed so that red cells can pass unimpeded.7 These large capillaries are in continuity with venules and comprise a nutritive capacitance system.100 Such a system is disadvantaged because a small change in blood volume results in a disproportionate change in NBF. The morphologic basis of this system appears to be the large-diameter capillaries, with a median diameter of 8 to 9 ␮m, considerably larger than that of muscular capillaries7 and poorly developed arteriolar smooth muscle.6

Peripheral Nerve Adaptation to Hypoxic Stress Peripheral nerve is metabolically unique, being able to function relatively well on anaerobic metabolism and having powerful adaptive mechanisms. Compared to rat brain, nerve has about 10% of brain’s oxygen requirements but similar energy stores.57,98 When maximally active, nerve increases its energy demands by less than 100%, whereas brain increases severalfold.57 The relatively large energy stores and the low resting and maximal energy expenditure

enable nerve to function quite well on anaerobically generated high-energy phosphate. We have demonstrated that nerve will conduct impulses for many additional minutes when energy substrates are increased, as in diabetes.57 Indeed, Fink and Cairns28 demonstrated that mammalian peripheral nerve will conduct impulses for hours when provided with a limitless supply of glucose. Another strategy of resistance to ischemic conduction failure is a further downregulation of energy-requiring enzymes, a situation that occurs in aging54 and in chronic hypoxia.55 There is also a suggestion that acute hypovolemic stress also results in adaptive mechanisms; peripheral nerve responds by reducing its oxygen consumption acutely.100

REGULATION OF NERVE BLOOD FLOW The regulation of nerve blood flow is relatively well defined by predictions of the Poiseuille equation, which defines flow as a function of its diameter, pressure gradient, capillary length, and viscosity. However, consideration needs to be additionally focused on the status of the perineurium. Of great interest is the role of agents that dynamically change the caliber of the ends of the microvessel (arteriole and venule).

Intrinsic Mechanisms Microvascular flow is predictable from the Poiseuille equation: NBF ⫽ (P1 ⫺ P2) ⫻ (r4/8) ⫻ (␲/L), where NBF varies directly with the fourth power of its radius (r), the viscosity of blood (␲), capillary length (L), and the pressure gradient (P1 ⫺ P2) between the venular end (P2) and the arteriolar end (P1). Strictly speaking, the Poiseuille equation applies only to Newtonian fluids, and blood flow is non-Newtonian, so that the formulation is semiquantitative. The ultimate purpose of microvasculature is to deliver oxygen and remove metabolic waste products. That the Poiseuille equation has practical application is supported by detailed simulation studies on oxygen delivery to nerve.51 In a mathematical simulation of the release, diffusion, and consumption of oxygen in the capillaries and surrounding tissue of peripheral nerve under steady-state conditions, the Krogh-Erlang equation was used to calculate oxygen tension in tissue, and numerical solution of the differential equation governing oxygen release from hemoglobin and diffusion was used to calculate oxygen tension in the capillary. Using average measured values for the parameters of oxygen solubility, diffusion coefficient, capillary diameter, capillary density, NBF, oxygen consumption rate, and arterial oxygen tension in rat peripheral nerve, we calculated the endoneurial oxygen tension as a function of distance from the nearest capillary and distance along the capillary from the arterial end to the venous end (Fig. 30–1). Oxygen tension is highest at the

669

Nerve Blood Flow and Microenvironment

Capillary Length The length of the capillary influences the drop in hydrostatic pressure along the vessel, although the largest dissipation of pressure is along small arteries and terminal arterioles.38 In nearly all tissues studied, 40% to 60% of the reduction in pressure occurs above the first-order arterioles. The second major drop is across terminal arterioles, with little evidence for precapillary sphincters in most tissues.

FIGURE 30–1 Oxygen delivery based on capillary model using Krogh-Erlang equation. (From Lagerlund, T. D., and Low, P. A.: A mathematical simulation of oxygen delivery in rat peripheral nerve. Microvasc. Res. 34:211, 1987, with permission.)

arterial end and lower at the venous end. It also falls off with distance from the capillary. There is a point (lethal corner) comprising the furthest distance from the arteriolar end of the capillary where oxygenation is poorest (see Fig. 30–1). The range of calculated values agreed with experimental measurements obtained from the sciatic nerves of rats. Conditions that adversely affected oxygen delivery include reduced capillary diameter, increased intercapillary distance, reduced blood flow, and reduced arterial oxygen tension. The lower experimentally obtained oxygen tensions in sciatic nerves of diabetic rats could be accounted for reasonably by this model on the basis of a 33% reduction in NBF (consistent with previously measured flow reduction). Viscosity of Blood Hemorheologic mechanisms also have major influences on blood flow and include the concentration of hemoglobin, serum proteins, the presence of abnormal hemoglobins, and factors affecting the coagulability of blood. Blood hematocrit is the major determinant of whole blood viscosity, but erythrocyte aggregability and deformability are also important contributors to blood viscosity.26,81 Capillary Radius A small change in radius results in a major change in NBF because NBF varies with the fourth power of the radius. Capillary diameter is altered in certain peripheral neuropathies. For instance, in diabetic neuropathy, there is endothelial cell hypertrophy and hyperplasia, and intimal and smooth muscle cell proliferation leading to a reduction in lumen diameter,37 as well as capillary thrombosis101 and closure.27

Pressure Gradient A fall in systemic BP will reduce P1 and an increase in venous pressure will increase P2. Particularly in a tissue that autoregulates poorly, P1 is an important regulator of NBF. P1 is altered dynamically by arteriolar tone and passively by blood volume. Less focus has been given to P2, the venular end, which is also vasoregulated, and the status of both P1 and P2 are important. NBF has been found by all recent studies to regress against systemic BP56,90,99,100 in anesthetized and decerebrate nonanesthetized rats.99 Because nerve microvasculature is a nutritive capacitance system, a small reduction in blood volume, as might occur in hypovolemia or exsanguination, results in a disproportionate reduction in NBF100 and precedes the reduction in systemic BP.

Extrinsic Mechanisms Perineurial Pinch and Endoneurial Edema Nerve is supplied by an intrinsic interconnecting system of microvessels. Feeding into the intrinsic system is an epineurial system, the extrinsic system. The two systems are interconnected via arterioles, which traverse nerve perineurium. What is not known is whether the interconnections are so diffuse and uniform that endoneurial ischemia results only when many perforating arterioles are occluded, or whether each segment of nerve has a major controlling influence on its underlying intrinsic nerve blood flow. The former model would be one in which nerve ischemia occurs by a hemodynamic mechanism. The latter would be at least in part a local and regional supply and ischemic mechanism. Myers et al.71 suggested that perineurial distortion, as might occur in nerve edema, may result in distortion of perineurium creating a perineurial pinch mechanism, and provided mathematical modeling data to demonstrate its feasibility. In earlier experimental studies in which peripheral nerve was undercut (to devascularize nerve) for long stretches or the regional supply was ligated, nerve fiber degeneration did not occur.1,10,24,60 However, the focus of these studies had been on the effect of arterial rather than arteriolar occlusion. Additionally, the end point of fiber degeneration may be too crude because hypoxic/ischemic effects, including conduction failure, occur well before fiber degeneration. Epineurial vasoconstriction induced by norepinephrine over a short

670

Functional Compartments and Nerve-Fiber Environment

segment of nerve (2 cm) on the underlying endoneurial blood flow, measured simultaneously in the subperineurial and centrifascicular sites using microelectrode H2 polarography, resulted in a dramatic reduction in NBF in both sites at concentrations as low as 10⫺7 M.40 We excluded a systemic effect of norepinephrine by demonstrating that the local ischemia was unassociated with NBF reduction in the contralateral sciatic nerve or with an increase in plasma norepinephrine above baseline (M. Kihara and P. A. Low, unpublished observations). We interpret our data as indicating that there is local vasoregulation of NBF by epineurial arterioles, but that the small residual blood flow is sufficient to prevent fiber degeneration. Endoneurial edema, by perineurial distortion, may contribute to the NBF reduction demonstrated in experimental galactose neuropathy. Vasoregulation of Peripheral Nerve: Vasoconstrictors Regional NBF is regulated by systemic BP56 and the balance between neural vasoconstrictors and vasodilators (Table 30–1). The major neurotransmitter of mammalian peripheral nerve is norepinephrine,3,85,104 which is confined to the microvasculature of epineurium and is essentially absent in endoneurium.86 Other vasoconstrictors are endothelin and vasopressin. In a study of the role of ␣-adrenergic innervation in nerve vasoregulation,111 NBF was monitored using a number of techniques. We used laser Doppler velocimetry, endoneurial measurement by microelectrode H2 polarography, and epineurial vasoreactivity by computerized videoangiology. Both ␣ agonists and antagonists were applied locally and systemically. Vasoconstriction mediated by ␣ agonists was regularly produced and was blocked by prior treatment with ␣ antagonists. The videoangiologic recordings showed markedly heterogeneous vasoreactivity to norepinephrine along segments of arterioles, suggesting a segregation of ␣ receptors. Additionally, there is also a local noradrenergic regulation of underlying endoneurial NBF by epineurial arterioles. Sympathetic stimulation results in a reduction

in NBF and chemical sympathectomy results in an increase in NBF. Vasoconstriction is known to be mediated by epineurial ␣-adrenergic receptors40,111 including vasopressin, 5-hydroxytryptamine, and endothelin receptors. 110 5-Hydroxytryptamine reduces NBF with a predominant effect on arteriovenous flow.21 In a previous study, we reported an EC50 (estimated concentration causing 50% of maximal constriction) of 10⫺4.9 M based on a doseresponse study of the epineurial application of norepinephrine, a method that generates EC50 values about two orders of magnitude lower than intra-arterial methods.40 The EC50 for endothelin, using identical methodology, was about three orders of magnitude lower at 10⫺8 M.110 An identical value of EC50 for endothelin was reported by Zochodne et al.110 We undertook a dose-effect study of vasopressin on NBF, its interactions with ␣ adrenoreceptors, and its effect on ischemic conduction failure. Topical epineurial application of vasopressin caused a concentrationdependent reduction of NBF (EC50 ⫽ 6.6 ⫻ 10⫺9 M; asymptote ⫽ 73.9% NBF reduction). The topical application of subthreshold concentrations of vasopressin and norepinephrine alone resulted in no change in NBF, but combined application resulted in a dramatic reduction (72.3%), suggesting synergistic action. We also demonstrated that vasoconstriction caused by combined vasopressin and norepinephrine can produce partial conduction block of sciatic-tibial nerve.91 Vasoregulation of Peripheral Nerve: Vasodilators These potent vasoconstrictor actions are balanced by vasodilation and mediated by calcitonin gene–related peptide (CGRP),108 substance P (SP),108 nitric oxide,18,41 and the prostaglandins.42 CGRP fibers are more dense in nerve than SP, and CGRP is a more potent vasodilator.108 In addition to adrenergic innervation, there is also prominent peptidergic innervation of vasa nervorum.3 Prostacyclin (largely confined to endothelial cells) is the major vasodilator and inhibitor of platelet aggregation of microvessels, and thromboxane A2 (largely confined to platelets) has the

Table 30–1. Vasoregulation of Peripheral Nerve Site of Action

Reference(s)

Vasoconstrictors Norepinephrine Endothelin Vasopressin

Epineurial ␣ adrenoreceptors Endothelial cell Epineurial; ? endothelial

Zochodne and Low,111 Kihara and Low40 Zochodne et al.,110 Cameron et al.18 Sasaki and Low91

Vasodilators CGRP SP Nitric oxide Prostaglandins

Endothelium Endothelium Endothelium Endothelium

Zochodne and Ho108 Zochodne and Ho108 Kihara and Low,41 Cameron et al.18 Kihara and Low42

CGRP = calcitonin gene–related peptide; SP = subtance P.

Nerve Blood Flow and Microenvironment

opposite effects. The ratio of prostacyclin to thromboxane A2 is considered to be important in the maintenance of vascular tone.70 We demonstrated that nerve biosynthesis of 6-keto prostaglandin F1␣ (6KPGF1␣), the stable metabolite of prostacyclin, was largely confined to nerve sheath,104 suggesting another mechanism of epineurial regulation of endoneurial nerve blood flow. Observations similar to ours were made in diabetic rat aorta89 and heart, in which altered regulation of phospholipase activity was suggested as the mechanism of reduced endogenous 6KPGF1␣.88 One possible mechanism of reduced prostacyclin biosynthesis is the reduction of nerve norepinephrine. Norepinephrine release results in the increased synthesis and release of prostaglandin I2 (PGI2) metabolites30 by ␣ receptor–mediated and calcium/calmodulin-dependent mechanisms. Calmodulin is known to activate phospholipase A2,105 resulting in a breakdown of membrane phospholipids to generate arachidonic acid,17 whose availability is the rate-limiting step in prostaglandin synthesis.34 Norepinephrine may also be reduced by the generation of reduced oxygen species, which is thought to be increased in chronic experimental diabetes39 and may be generated from norepinephrine.12 Effect of Maturation and Aging on Nerve Blood Flow We have studied the effects of development (ages 2 to 12 weeks)46 and aging (to 30 months)44 in the sciatic nerve of rats using microelectrode H2 polarography. NBF was highest in 2-week-old rats and progressively declined during development. There was an inverse relationship between NBF and mean arterial pressure, which was lowest at 2 weeks of age, gradually increasing through the next 4 weeks and remaining relatively constant thereafter. The progressive decrease in NBF, associated with an increase in BP, occurs because of a progressive increase in endoneurial vascular resistance. This progressive increase in nerve vascular resistance is due to an increase in plasma viscosity and hematocrit, an increase in innervation of arterioles and small arteries, and a decrease in capillary density.46 The higher NBF in immature rats is likely to be a developmentally adaptive mechanism subserving the special metabolic and functional needs of immature nerves. It permits greater blood-nerve exchange of materials to accommodate the greater metabolic needs of rapidly elongating and myelinating axons and proliferating Schwann cells. Rat peripheral nerve is progressively more resistant to ischemic/anoxic conduction failure with increasing age, a phenomenon that is lowest in immature nerves.54 This resistance was paralleled by a progressive age-related decline in endoneurial oxygen consumption. Endoneurial ATP and creatine phosphate concentrations and expenditure were also progressively reduced with age. Anoxia resulted in progressively smaller reductions in these nucleotide phosphates with increasing age to 8 months,

671

after which time little further change occurred. These findings indicate that the major mechanism of resistance to ischemic conduction failure is a progressive decline in energy requirements. At no stage in development was there any compelling evidence of autoregulation of NBF. Adult peripheral nerve, with increasing age, undergoes a progressive increase in resistance to ischemia and reduction in energy utilization. Nerve conduction velocity and amplitude of nerve action potential increased progressively with age to 8 months, after which time no further increases were demonstrated in the rat.54 This increase was associated with greater resistance to ischemic/anoxic conduction failure and a progressive age-related decline in endoneurial consumption of oxygen, ATP, and creatine phosphate. NBF regressed negatively with increasing age, and this decline was associated with an increase in nerve vascular resistance. Twenty minutes of nerve stimulation resulted in an increase in blood flow by about 50% in adult animals, and this response did not decline with increasing age. As indices of oxygen free radical activity, we measured conjugated dienes, hydroperoxides, and norepinephrine from 2 to 30 months. There was a gradual decline of all indices with increasing age. We concluded that NBF declines with aging as a result of reduced microvascular caliber. These vessels retain their hyperemic response, and oxygen free radical activity is less with increasing age.44 These findings, taken together, suggest that increasing age is associated with reduced utilization of high-energy phosphate.

EXPERIMENTAL NERVE ISCHEMIA Models of Nerve Ischemia Unlike richly perfused tissues such as brain, kidney, and heart, in which ischemic injury is readily produced, peripheral nerve is remarkably resistant to both ischemic conduction failure and ischemic fiber degeneration because of its low energy needs and extensive anastomoses.61 Early attempts to produce experimental ischemia by ligating supplying arteries or undercutting nerve were largely unsuccessful. Models of ischemia have, however, subsequently been developed. These include ligation of supplying arteries to the limb nerves, resulting in greater degree of ischemic fiber degeneration at the level of the tibial nerve and less severe alterations of sciatic nerve.32,49,77,83 The severity of ischemic stress can be increased by the use of agents that induce vasoconstriction, such as arachidonic acid,82 norepinephrine,111 and endothelin.110 A highly successful model of nerve ischemia can be induced by the intra-arterial injection of microemboli of sufficient size to occlude supplying capillaires.77 The severity of reduction of NBF and of ischemic fiber degeneration is linearly related to the dose of microemboli.47 For physiologic studies of nerve ischemia and the effects of reperfusion injury, it was

672

Functional Compartments and Nerve-Fiber Environment

necessary to develop a model of near-total yet reversible ischemia. Our model of severe nerve ischemia, produced by ligating collateral arteries followed by the temporary occlusion of the abdominal aorta and both iliac arteries, consistently occluded NBF.94

Factors that Modulate the Effects of Nerve Ischemia The severity of ischemic fiber degeneration is dependent on the severity and duration of the ischemic stress. Nukada and Dyck77 used polystyrene microspheres of a size (14 ␮m) sufficient to plug capillaries and precapillaries. These were injected into the arterial supply of rat sciatic nerves, producing widespread segmental occlusion of capillaries in lower limb nerves. The clinical and pathologic effect was dose related. One million microspheres produced selective capillary occlusion but no nerve fiber degeneration; approximately 6 million microspheres also produced selective capillary occlusion and associated foot and leg weakness, sensory loss, and fiber degeneration beginning in a central core of the distal sciatic nerve; and 30 million microspheres caused both capillary and arterial occlusion and a greater neuropathologic deficit. From these observations, it is inferred that (1) occlusion of isolated precapillaries and capillaries does not produce ischemic fiber degeneration; (2) occlusion of many microvessels results in central fascicular fiber degeneration, indicating that these cores are watershed regions of poor perfusion; and (3) stereotyped pathologic alterations of nerve fibers and Schwann cells are related to dose, anatomic site, and time elapsed since injection. We evaluated the relationship of microsphere dose and blood flow. The dose of microspheres regressed with the degree of hind limb paresis, reduction in NBF, degree of fiber pathology, and ischemic conduction failure of the tibial nerve.47 That ischemic fiber degeneration is due to ischemia is supported by the demonstration that the severity of fiber degeneration regresses inversely with NBF (Fig. 30–2).47 All nerves with severe fiber degeneration had flows of less than 3 mL/100 g/min. Intercapillary distance is a determinant of NBF, and the increased intercapillary distance associated with endoneurial edema results in a corresponding reduction in NBF, as predicted by the Poiseuille equation.64 In experimental nerve ischemia, centrifascicular fiber degeneration is typical, with subperineural sparing.77 The difference in intercapillary distance in centrifascicular regions versus the subperineurial area has been proposed as being responsible,53,63 although diffusion of oxygen from surrounding areas is likely more important.63 A major determinant of the severity of ischemic fiber degeneration is the temperature of nerve during ischemia. Because peripheral nerve has a large ischemic safety factor, hypothermia (by reducing metabolic demands) greatly

FIGURE 30–2 Relationship between severity of ischemic fiber degeneration and nerve blood flow. Severity of fiber degeneration is inversely related to measured perfusion.

reduces the severity of ischemic fiber degeneration. We evaluated the influence of temperature on the effect of ischemia to the sciatic nerve in the rat. Ischemia was produced by embolization of microspheres (diameter 14 ␮m) into the supplying arteries at three temperatures (37°, 32°, and 28° C) and maintaining the nerve at the defined temperature for an additional 4 hours. End points (evaluated 7 days after embolization) of function, electrophysiology, NBF (in mL/100 g/min), and severity of ischemic fiber degeneration demonstrated a close relationship between temperature and severity of nerve fiber degeneration.45 Hypothermia at 28° C dramatically reduced the severity of ischemic fiber degeneration. The timing of hypothermia is important. Its neuroprotective value is much greater during the ischemic insult than during reperfusion, although the latter was partially effective and ameliorated the development of endoneurial edema, presumably by reducing the severity of blood-nerve barrier breakdown.68

Influence of Diabetic State on Nerve Ischemia A consistent finding in experimental diabetic neuropathy is a reduction in NBF. Decreased endoneurial perfusion has been found repetitively (about 100 reports) by 12 programs, although it has been questioned by two investigators.20,109 Table 30–2 summarizes the current status of NBF experimental results. Cameron et al.19 reported that there is a tight relationship between NBF and nerve conduction in experimental diabetic neuropathy (Fig. 30–3). The reduction in NBF is associated with a reduction in endoneurial oxygen tension.102 Chronic hyperglycemia results in a large deficit in NBF. Both auto-oxidative and ischemia-induced lipid peroxidation occurs with resultant peripheral sensory

673

Nerve Blood Flow and Microenvironment

Table 30–2. Results of Studies on Nerve Blood Flow in Experimental Diabetes Investigator/Program

Method

Low, Mayo (Rochester, USA) Powell/Myers (San Diego, USA) Greene (Michigan, USA) Cameron/Cotter (Scotland, UK) Gispen (Netherlands, UK) Hotta (Nagoya, Japan) Stevens (Michigan, USA) Tomlinson (London, UK) Ueno (Kanagawa, Japan) Yasuda (Ohtsu, Japan) Wright/Nukada (New Zealand) Yorek (Wisconsin, USA) Kihara (Nagoya, Japan) Nakamura (Nagoya, Japan) Schratzberger (Boston, USA) Ueno (Japan) Yamamoto (Osaka, Japan) Van Dam (Utrect, Netherlands) Van Buren (Utrect, Netherlands) Zochodne (Calgary, Canada) Williamson (USA)

H2; LDF; iodo-AP LDF H2 H2; LDF LDF LDF; H2 H2 LDF LDF LDF LDF H2 H2 H2 LDF, imaging LDF LDF LDF LDF LDF; H2 Microemboli

Results b b b b b b b b b b b b b b b b b b b ? c

H2 ⫽ hydrogen polarography; iodo-AP ⫽ iodoantipyrine radiography; LDF ⫽ laser Doppler flow velocimetry.

neuropathy in streptozotocin-induced diabetes in the rat. Free radical defenses, especially involving antioxidant enzymes, have been suggested to be reduced, but scant information is available on these defenses in chronic hyperglycemia. We evaluated the gene expression of glu-

FIGURE 30–3 Relationship between nerve blood flow and conduction velocity in experimental diabetic neuropathy.

tathione peroxidase, catalase, and superoxide dismutase (cuprozinc and manganese separately) in L4/5 dorsal root ganglion (DRG) and superior cervical ganglion, as well as enzyme activity of glutathione peroxidase in DRG and sciatic nerve in experimental diabetic neuropathy of 3-month and 12-month durations. We also evaluated nerve electrophysiology of caudal, sciatic-tibial, and digital nerves. A nerve conduction deficit was seen in all nerves in experimental diabetic neuropathy at both 3 and 12 months. Gene expression of glutathione peroxidase, catalase, cuprozinc superoxide dismutase, and manganese superoxide dismutase was not reduced in experimental diabetic neuropathy at either 3 or 12 months. Catalase messenger RNA (mRNA) was significantly increased in experimental diabetic neuropathy at 12 months. Glutathione peroxidase enzyme activity was normal in sciatic nerve. We conclude that gene expression is not reduced in peripheral nerve tissues in very chronic experimental diabetic neuropathy. Changes in enzyme activity may be related to duration of diabetes or due to posttranslational modifications.48 Excessive Susceptibility Nukada first demonstrated excessive vulnerability of diabetic nerves to ischemia in rats that had been diabetic for 4 months.74 He ligated the supplying arteries to diabetic (duration 4 months) rat sciatic nerve and demonstrated

674

Functional Compartments and Nerve-Fiber Environment

FIGURE 30–4 Ischemia-reperfusion injury (3 hours of ischemia, 7 days of reperfusion) to an experimental diabetic neuropathy model showing excessive fiber degeneration and immunolabeling of activated caspase-3, 8-hydroxydeoxyguanosine, and 4-hydroxynonenal in diabetic tibial nerve compared with age-matched controls.

excessive fiber degeneration. Of interest is that, whereas resistance to ischemic fiber degeneration develops early, the excessive vulnerability to ischemic fiber degeneration is time dependent, being seen at 4 months and not apparently present at 2 weeks.75 In Figure 30–4, nerves in a model with experimental diabetic neuropathy (duration of diabetes of 4 months) underwent extensive fiber degeneration in response to 3 hours of ischemia, whereas corresponding control nerve was relatively unaffected. Schwann cells demonstrated prominent immunolabeling to 8-hydroxydeoxyguanosine and hydroxynonenal, indicating oxidative injury (Fig. 30–4). Caspase-3 immunoreactivity indicates commitment to the efferent limb of the apoptotic pathway. Resistance to ischemic conduction failure is due largely to the ability of diabetic nerves to continue conducting longer than normal nerves because of the large increase in energy stores,58 and, because diabetic nerve is partially depolarized,14 there is reduced energy cost involved in impulse generation.87 The pathophysiology of excessive susceptibility to ischemic fiber degeneration is not fully elucidated. Areas of focus have been on the role of vasoconstrictors (increased) and vasodilators (reduced) and the role of oxidative injury and the inflammatory response. Nitric oxide synthase activity is reduced in sciatic nerve of experimental diabetic neuropathy41,80 and is reversed by insulin treatment early on. Experimental diabetes is associated with denervation supersensitivity to vasoconstrictors.107 Application of topical endothelin-1 to sciatic nerve trunk resulted in axonal damage that exceeded that of nondiabetics by a factor greater than 5. The pattern of axonal degeneration was multifocal but not centrofascicular and did not vary with fascicular

area.107 There are focal electrophysiologic alterations in these rats. There is some uncertainty as to whether there is receptor supersensitivity to endothelin107,110 or a change in the balance of nerve vasodilators to vasoconstrictors,52 because some studies have not found, at least in in vitro experiments, supersensitivy to endothelin.41 Limited information is available on the mechanisms of this susceptibility. Using an in vitro model, Wachtler et al.103 studied the role of intracellular pH (pHi) and intracellular calcium concentrations ([Ca2⫹]i). Isolated rat spinal roots were preincubated for 3 to 6 hours in either 5- or 25-mM D-glucose before transient exposure to gaseous hypoxia or cyanide. Intracellular pH and Ca2⫹ concentrations were measured photometrically by means of the fluorescent dyes carboxy-SNARF-1 and a combination of Calcium Green-1 and Fura Red, respectively. They observed that 25-mM D-glucose resulted in stronger intracellular acidification and much slower posthypoxic recovery of pHi as compared to 5-mM D-glucose. From these and the calcium experiments, they concluded that hypoxia-induced nerve acidification rather than a rise in [Ca2⫹]i might contribute to ischemic lesions found in diabetic neuropathy.

Reperfusion Injury Although ischemia alone will cause ischemic fiber degeneration, reperfusion will aggravate ischemic injury to the axon, the Schwann cell, and the blood-nerve barrier by oxidative injury. In rat sciatic-tibial nerve subjected to ischemia sufficient to cause conduction failure (1 or 3 hours of complete ischemia), two interesting observations were found with reperfusion. First, with reperfusion NBF was restored to only 55% and 45% of resting values

Nerve Blood Flow and Microenvironment

following 1 and 3 hours, respectively, indicating reduced reperfusion94 as a result of swelling of endothelial cells.2 Immunolabeling for 8-hydroxydeoxyguanosine, a marker of oxidative DNA injury, is demonstrable for endothelial cells (see Fig. 30–4). Second, the blood-nerve barrier breakdown occurred during reperfusion as a result of oxidative injury.94 We subsequently evaluated the effect of ischemic reperfusion of rat sciatic-tibial nerve seeking biochemical and pathologic evidence of blood-nerve barrier disruption and lipid peroxidation. Ischemia caused by the ligation of the supplying arteries to sciatic-tibial nerve was maintained for 3 hours, followed by reperfusion. Reperfusion resulted in an increase in nerve lipid hydroperoxides, greatest at 3 hours, followed by a gradual decline over the next month. Nerve edema and ischemic fiber degeneration consistently became more severe with reperfusion, indicating that oxidative stress impairs the blood-nerve barrier (edema) and causes ischemic fiber degeneration. Reduced reperfusion was greatest over distal sciatic nerve and mid-tibial nerve at day 7. The most ischemic segment (mid-tibial) of nonreperfused ischemic nerves (duration 3 hours) underwent both edema and ischemic fiber degeneration that were as pronounced as those of other segments after reperfusion and underwent a smaller increase in NBF with reperfusion, suggesting that ischemia alone can also cause ischemic fiber degeneration and edema. The type of fiber degeneration was that of axonal degeneration.72 Free radical–mediated lipid and protein oxidation results in carbonyl formation in nerve. Its presence in tissue can be detected histochemically. Reperfusion following 7 hours of near-complete ischemia resulted in histochemically detectable carbonyl compounds in microvessels of reperfused nerves.2 Endothelial cells were positively stained and swollen. ␣-Tocopherol pretreatment reduced staining and size of endothelial cells.2 In a subsequent study, He et al.,31 using a sensitive enzyme-linked immunosorbent assay, measured protein carbonyl content in experimental ischemia to nerve followed by reperfusion. They found that carbonyl content was unaffected by ischemia alone but increased by 55% after 12 to 18 hours of reperfusion, correlating with the onset of nerve pathology. Pretreatment with the xanthine oxidase inhibitor allopurinol prevented these abnormalities, suggesting that xanthine oxidase activity precedes oxidative damage during reperfusion injury. Nukada et al.76 differentiated the pathologic changes caused by reperfusion from those of ischemia alone. Changes caused by less severe ischemia followed by reperfusion were characterized by vascular swelling, endoneurial and intramyelinic edema, demyelination, and a paucity of axonal degeneration. Breakdown of the blood-nerve barrier was characteristic. Reperfusion results in alterations to the blood-nerve barrier, endothelial cells, and Schwann cells. Part of the response is mediated by an inflammatory response. Nukada et al.79 investigated the time course of

675

acute inflammation and its role in the development of demyelination in reperfused rat sciatic, tibial, and peroneal nerves after a 5-hour period of near-complete ischemia. Polymorphonuclear neutrophil migration was seen early in the endoneurial lesion. After 18 hours of reperfusion, there was maximal intercellular adhesion molecule-1 (ICAM-1) expression on endoneurial vessels, and polymorphonuclear neutrophil accumulation was then prominent, reaching a peak 24 hours after reperfusion. Endoneurial mononuclear macrophages increased nearly fourfold after 48 to 72 hours of reperfusion. Macrophages were observed invading Schwann cells and myelinated lamellae, with associated demyelination. Thus this study provides evidence of macrophage-associated demyelination after reperfusion similar to that seen in inflammatory neuropathies. We evaluated the mRNA expression of the pro-inflammatory cytokines tumor necrosis factor-␣ (TNF-␣) and interleukin-1␤ (IL-1␤) in rat sciatic and tibial nerves following ischemia-reperfusion injury, using competitive reverse transcriptase–polymerase chain reaction, to explore the role of cytokines in ischemia-reperfusion injury.66 The expressions of both TNF-␣ and IL-1␤ mRNA were related to severity of ischemia and occurred with reperfusion rather than ischemia alone. TNF-␣ gene expression peaked at 24 hours of reperfusion, whereas that of IL-1␤ peaked at 12 hours. These data support the notion that the proinflammatory cytokines TNF-␣ and IL-1␤ are involved in the inflammatory response of ischemia-reperfusion injury to the peripheral nervous system and may be involved in the pathophysiology of ischemic fiber degeneration. Nitric oxide synthase alterations also occur during ischemia-reperfusion. In a study evaluating mRNA and protein expressions of neuronal (nNOS), inducible (iNOS), and endothelial (eNOS) nitric oxide synthases in peripheral nerve after ischemia-reperfusion, the investigators reported an interesting pattern of isoform expression.84 They evaluated the effects of ischemia (2 hours) followed by reperfusion (0 or 3 hours). Ischemia alone resulted in a reduction in both nNOS and eNOS and their protein expressions. After ischemia-reperfusion (2 hours of ischemia followed by 3 hours of reperfusion), expression of both nNOS and eNOS mRNA and protein decreased further. In contrast, iNOS mRNA significantly increased after ischemia and was further upregulated (14-fold) after reperfusion, whereas iNOS protein was not detected. The results highlight the importance of the inflammatory response associated with reperfusion following ischemia. Recently, we explored the recovery process by extending the period of observation following reperfusion to 42 days.33 Pathologically, three phases were identifiable: phase 1 (0 to 3 hours): minimal pathologic changes and minimal edema; phase 2 (7 days, 14 days): prominent fiber degeneration and endoneurial edema; and phase 3 (28 days, 42 days): abundant small regenerating fiber clusters and minimal edema. Compound muscle action potential was the most sensitive

676

Functional Compartments and Nerve-Fiber Environment

index of neural deficits and recovery, showing progressive recovery beyond 14 days. Severe functional deficits developed immediately and persisted, with a trend to recovery at the 42-day time point. It was concluded that reperfusion by oxidative injury worsened nerve function and aggravated fiber degeneration, but in the longer time frame permitted fiber regeneration to occur. Neuroprotection Knowledge of the pathogenesis of ischemia-reperfusion injury has naturally led to studies designed to minimize the effects of this injury. Most of the studies to date have focused on hypothermia, antioxidants, and reducing the inflammatory response. Because peripheral nerve has a large ischemic safety factor, hypothermia (by reducing metabolic demands) is potentially an efficacious technique to rescue nerve from ischemic fiber degeneration. We therefore evaluated the influence of temperature on the severity of ischemic fiber degeneration in sciatic nerve resulting from embolization with microspheres (14 ␮m) into its supplying arteries. The limb was embolized at three temperatures (37°, 32°, and 28° C) and was maintained at each temperature for an additional 4 hours. End points, evaluated 7 days after embolization, for the embolized limb were (1) behavioral scores (0 to 11, in increasing limb function); (2) compound nerve action potential of sciatic-tibial nerve; (3) sciatic NBF (in mL/100 g/min); and (4) histologic grade, expressed as percentage of fibers undergoing ischemic fiber degeneration. NBF was reduced in all groups varying with temperature, and all indices of nerve structure and function were significantly improved with hypothermia. We concluded that hypothermia, easily achievable in a limb nerve, is highly efficacious in the rescue of nerve from ischemic fiber degeneration.45 We subsequently evaluated the effectiveness of hypothermia in protecting peripheral nerve from ischemia followed by reperfusion injury to rat peripheral nerve. Ischemia was caused by ligation of supplying arteries and collateral vessels, and reperfusion resulted when the ligatures were released after a predetermined period of ischemia (3 or 5 hours). The effects of temperature (28°, 32°, and 37° C) were evaluated. End points were behavioral score, nerve electrophysiology, and nerve histology. The groups treated at 36° to 37° C underwent marked fiber degeneration associated with a reduction in action potential and impairment in behavioral score. The groups treated at 28° C (for both 3 and 5 hours) showed significantly less (P ⬍ .01; analysis of variance, Bonferoni post hoc test) reperfusion injury for all indices (behavioral score, electrophysiology, and neuropathology), and the groups treated at 32° C had scores intermediate between the groups treated at 36° to 37° C and at 28° C. Our results showed that cooling the limbs dramatically protects the peripheral nerve from ischemia-reperfusion injury.67 Once hypothermia was applied in the ischemic period, the

resultant neuroprotection continued into the reperfusion period, even if nerve temperature was then raised during the reperfusion period. These results indicate that hypothermic neuroprotection is dramatically more efficacious during the intra-ischemic period than during reperfusion, when a lesser degree of neuroprotection ensued.68 As part of our focus on improving peripheral nerve salvage from ischemic fiber degeneration, we evaluated whether hyperbaric oxygen (HBO), by increasing delivery of oxygen to ischemic nerve, will rescue peripheral nerve rendered ischemic by microembolization. HBO appears to be less effective than hypothermia but is effective in improving behavioral, electrophysiologic, and morphologic indices of nerve function.43 We concluded that HBO will effectively rescue fibers from ischemic fiber degeneration, providing the ischemia is not extreme. The potent antioxidant ␣-lipoic acid was demonstrated to provide neuroprotection from ischemic fiber degeneration resulting from ischemia-reperfusion injury when the ischemic insult was not overwhelming.66 Distal sensory conduction was significantly improved in the 3-hour ischemia group treated with ␣-lipoic acid (P ⬍ .05). The antioxidant failed to show favorable effects if the duration of ischemia was longer (5 hours of ischemia). These results suggest that ␣-lipoic acid is efficacious for moderate ischemia-reperfusion, especially on distal sensory nerves. In a study of mild ischemic rat facial nerve injury induced by embolization of thrombin-coated microspheres,69 rapid and sustained labeling of facial nuclei (peaking by day 1 and lasting 8 days) of mRNA of c-Jun and growth-associated protein-43 was seen, followed by a delayed peak (day 3) for CGRP. Superoxide dismutase alleviated the behavioral impairment and decreased the CGRP mRNA expression at the third day after injury. The protocol for facial nerve paresis results in ischemia of both axons and facial neurons.69

PATHOGENESIS OF NERVE ISCHEMIA Data are gradually accumulating on the molecular pathogenesis of nerve ischemic fiber degeneration. There is a complex interplay of ischemia, eicosanoids, oxidative injury, the inflammatory response, and norepinephrine. Ischemia would result in phospholipase activation and phospholipid breakdown, liberating arachidonic acid with its cascade to produce the prostaglandins and leukotrienes. Reduced oxygen species are generated during ischemia and especially during reperfusion by mechanisms described earlier. The formation of lipid hydroperoxides would inhibit prostacyclin synthetase. The increased biosynthetic rate of thromboxane A2 coupled with the ischemia-related inhibition of prostacyclin production by endothelial cells would result in vasoconstriction, aggravating the ischemic insult. In nerve, there is a breakdown of the blood-nerve barrier after 3 hours of ischemia but not after 1 hour.

677

Nerve Blood Flow and Microenvironment

Reperfusion after 1 hour of ischemia will, however, regularly cause a breakdown of the barrier with a reduction in NBF94 as a result of endothelial cell swelling, consistently found in areas associated with demyelination.78 We used the blood-nerve barrier as a physiologic index of the oxidative injury, which is known to increase microvascular permeability.23,50 The permeability–suface area product may be increased as a result of an increase in permeability or an increase in the surface area of endoneurial capillaries. In this study, NBF remained persistently reduced with reperfusion. Therefore, the progressive increase in the product must be due to an increase in permeability. Our results suggest that reperfusion injury of nerve requires near-total ischemia of several hours’ duration, because it occurs following 3 hours but not 1 hour of occlusion. Using this paradigm, we demonstrated that reperfusion resulted in an increase in nerve lipid hydroperoxides, greatest at 3 hours, followed by a gradual decline over the next month.72 Endoneurial edema, indicative of bloodnerve barrier breakdown, consistently became more severe with reperfusion, indicating that oxidative stress impairs the barrier. Protein carbonyl content was unaffected by ischemia alone but increased by 55% after 12 to 18 hours’ reperfusion, correlated with the onset of nerve pathology, suggesting a possible role of peroxynitrite.2,31 Hypoxia/ischemia results in an increase in tissue reducing equivalents57,73 and an increase in tissue xanthine.36 Ischemia or hypoxia has been suggested to activate a calcium-dependent protease (possibly calpain), which converts cytosolic xanthine dehydrogenase to xanthine oxidase.4 This important reaction occurs in capillary endothelium.35 The above three conditions create an oxygen free radical–generating system. Furthermore, reperfusion accelerates reduced oxygen species formation by introducing oxygen into a system primed and generating reduced oxygen species.62 In nerve, pretreatment with the xanthine oxidase inhibitor allopurinol prevented these abnormalities, suggesting that xanthine oxidase activity precedes oxidative damage during reperfusion injury.31 Nukada et al.79 further explored the role of the adhesion molecule ICAM-1 and the inflammatory response. Polymorphonuclear neutrophil migration was seen early in the endoneurial lesion. After 18 hours of reperfusion, there was maximal ICAM-1 expression on endoneurial vessels, and polymorphonuclear neutrophil accumulation was then prominent, reaching a peak 24 hours after reperfusion. Endoneurial mononuclear macrophages increased nearly fourfold after 48 to 72 hours of reperfusion. Macrophages were observed invading Schwann cells and myelinated lamellae, with associated demyelination. Ischemia induces TNF-␣ mRNA in nerve,66 which in turn activates the nuclear factor (NF)-␬ pathway, causing oxidative stress.13 The expressions of both TNF-␣ and IL-1␤ mRNA were related to severity of ischemia and

occurred with reperfusion rather than ischemia alone.65 TNF-␣ gene expression peaked at 24 hours of reperfusion, whereas that of IL-1␤ peaked at 12 hours. These data support the notion that the pro-inflammatory cytokines TNF-␣ and IL-1␤ are involved in the inflammatory response of ischemia-reperfusion injury to the peripheral nervous system and may be involved in the pathophysiology of ischemic fiber degeneration. There is some indirect evidence for intracellular calcium accumulation with ischemia in peripheral nerve. Calcium ionophore will cause vesicular disruption of myelin93,96 and axonal degeneration,92 and nerve reconnection in a calciumfree medium resulted in improved functional recovery.22 The vasoconstriction and microvascular ischemia/hypoxia in disorders such as diabetes have been in part ascribed to perturbations of prostaglandins and reduced oxygen species generation. Lipid hydroperoxides are increased and inhibit prostacyclin synthetase activity, resulting in a reduced prostacyclin:thromboxane ratio and vasoconstriction and platelet aggregation.106 Three abnormalities in PGI2 have been described in diabetes. Apart from a reduced synthesis of PGI2, serum from patients with diabetes or rats with streptozotocin-induced diabetes contains less PGI2 releasestimulating factor or may have more of an inhibitory factor.59 The presence of a stimulating factor appears to be well established in normal serum.95,97 Third, platelets of diabetic patients have reduced responsiveness to PGI2,8 and dipyridamole will restore reponsiveness.39 The current pathogenetic schema of the pathogenesis of nerve ischemia-reperfusion injury may be summarized as follows. There is energy run-down,112 poly(ADP-ribose) polymerase (PARP) activation, NF-␬B activation, and activation of adhesion molecules including ICAM-1. At this time, it is known that there is immunolabeling of endothelial cells with ICAM-1, PARP, and 8-hydroxydeoxyguanosine (indicating oxidative DNA injury). A prominent inflammatory infiltration with polymorphs is seen. Next there is a switch to mononuclear cells, upregulation of iNOS, and increased expression of certain cytokines, including TNF-␣.65 These changes occur between 48 hours and 1 week. There is immunolabeling of Schwann cells with 8-hydroxydeoxyguanosine and caspase-3 and terminal deoxynucleotidyl transferase–mediated dUTP nick end-labeling (TUNEL) positivity, indicating commitment to the apoptotic pathway (see Fig. 30–4).

REFERENCES 1. Adams, W. E.: The blood supply of nerves. II. The effects of exclusion of its regional sources of supply on the sciatic nerve of the rabbit. J. Anat. 77:243, 1943. 2. Anderson, G. M., Nukada, H., and McMorran, P. D.: Carbonyl histochemistry in rat reperfusion nerve injury. Brain Res. 772:156, 1997.

678

Functional Compartments and Nerve-Fiber Environment

3. Appenzeller, O., Dhital, K. K., Cowen, T., and Burnstock, G.: The nerves to blood vessels supplying blood to nerves: the innervation of vasa nervorum. Brain Res. 304:383, 1984. 4. Battelli, M. G., Corte, E. D., and Stirpe, F.: Xanthine oxidase type d (dehydrogenase) in the intestine and other organs of the rat. Biochem. J. 126:747, 1972. 5. Baumbach, G. L., and Heistad, D. D.: Regional, segmental, and temporal heterogeneity of cerebral vascular autoregulation [review]. Ann. Biomed. Eng. 13:303, 1985. 6. Bell, M. A., and Weddell, A. G.: A morphometric study of intrafascicular vessels of mammalian sciatic nerve. Muscle Nerve 7:524, 1984. 7. Bell, M. A., and Weddel, A. G. M.: A descriptive study of the blood vessels of the sciatic nerve in the rat, man and other mammals. Brain 107:871, 1984. 8. Betteridge, D. J., El Tahir, K. E., Reckless, J. P., and Williams, K. I.: Platelets from diabetic subjects show diminished sensitivity to prostacyclin. Eur. J. Clin. Invest. 12:395, 1982. 9. Bevan, J. A.: Autonomic pharmacologist’s guide to the cerebral circulation. Trends Pharmacol. Sci. 5:234, 1984. 10. Blunt, M. J., and Stratton, K.: The immediate effects of ligature of vasa nervorum. J. Anat. 90:204, 1956. 11. Borgstrom, P., Grande, P. O., and Lindbom, L.: Responses of single arterioles in vivo in cat skeletal muscle to change in arterial pressure applied at different rates. Acta Physiol. Scand. 113:207, 1981. 12. Boveris, A., and Chance, B.: The mitochondrial generation of hydrogen peroxide: general properties and effect of hyperbaric oxygen. Biochem. J. 134:707, 1973. 13. Bowie, A., and O’Neill, L. A.: Oxidative stress and nuclear factor-kappaB activation: a reassessment of the evidence in the light of recent discoveries. Biochem. Pharmacol. 59:13, 2000. 14. Brismar, T., Sima, A. A., and Greene, D. A.: Reversible and irreversible nodal dysfunction in diabetic neuropathy. Ann. Neurol. 21:504, 1987. 15. Burnstock, G.: Innervation of vascular smooth muscle: histochemistry and electron microscopy. Clin. Exp. Pharmacol. Physiol. Suppl. 2:7, 1975. 16. Burnstock, G., Griffith, S. G., and Sneddon, P.: Autonomic nerves in the precapillary vessel wall. J. Cardiovasc. Pharmacol. 6(Suppl.):344, 1984. 17. Burton, K. P., Buja, L. M., Sen, A., et al.: Accumulation of arachidonate in triacylglycerols and unesterified fatty acids during ischemia and reflow in the isolated rat heart: correlation with the loss of contractile function and the development of calcium overload. Am. J. Pathol. 124:238, 1986. 18. Cameron, N. E., Cotter, M. A., and Hohman, T. C.: Interactions between essential fatty acid, prostanoid, polyol pathway and nitric oxide mechanisms in the neurovascular deficit of diabetic rats. Diabetologia 39:172, 1996. 19. Cameron, N. E., Eaton, S. E., Cotter, M. A., and Tesfaye, S.: Vascular factors and metabolic interactions in the pathogenesis of diabetic neuropathy. Diabetologia 44:1973, 2001. 20. Chang, K., Ido, Y., LeJeune, W., et al.: Increased sciatic nerve blood flow in diabetic rats: assessment by “molecular” vs. particulate microspheres. Am. J. Physiol. 273:E164, 1997. 21. Day, T. J., Lagerlund, T. D., and Low, P. A.: Analysis of H2 clearance curves used to measure blood flow in rat sciatic nerve. J. Physiol. (Lond.) 414:35, 1989.

22. de Medinaceli, L., Wyatt, R. J., and Freed, W. J.: Peripheral nerve reconnection: mechanical, thermal, and ionic conditions that promote the return of function. Exp. Neurol. 81:469, 1983. 23. Del Maestro, R. F., Bjork, J., and Arfors, K. E.: Increase in microvascular permeability induced by enzymatically generated free radicals. II. Role of superoxide anion radical, hydrogen peroxide, and hydroxyl radical. Microvasc. Res. 22:255, 1981. 24. Denny-Brown, D., and Brenner, C.: Paralysis of nerve induced by direct pressure and by tourniquet. Arch. Neurol. Psychiatry 51:1, 1944. 25. Dhital, K., Lincoln, J., Appenzeller, O., and Burnstock, G.: Adrenergic innervation of vasa and nervi nervorum of optic, sciatic, vagus and sympathetic nerve trunks in normal and streptozotocin-diabetic rats. Brain Res. 367:39, 1986. 26. Dintenfass, L.: Haemorheology of diabetes mellitus. J. Pharmacol. Exp. Ther. 278:1262, 1979. 27. Dyck, P. J., Hansen, S., Karnes, J., et al.: Capillary number and percentage closed in human diabetic sural nerve. Proc. Natl. Acad. Sci. U. S. A. 82:2513, 1985. 28. Fink, B. R., and Cairns, A. M.: Differential tolerance of mammalian myelinated and unmyelinated nerve fibers to oxygen lack. Reg. Anesth. 7:2, 1982. 29. Folkow, B.: Autoregulation in muscle. Circ. Res. 15:19, 1964. 30. Gilmore, N., Vane, J. R., and Wyllie, J. H.: Prostaglandins released by the spleen. Nature 218:1135, 1968. 31. He, K., Nukada, H., McMorran, P. D., and Murphy, M. P.: Protein carbonyl formation and tyrosine nitration as markers of oxidative damage during ischaemia-reperfusion injury to rat sciatic nerve. Neuroscience 94:909, 1999. 32. Hess, K., Eames, R. A., Darveniza, P., and Gilliatt, R. W.: Acute ischaemic neuropathy in the rabbit. J. Neurol. Sci. 44:19, 1979. 33. Iida, H., Schmelzer, J. D., Schmeichel, A. M., et al.: Peripheral nerve ischemia: reperfusion injury and fiber regeneration. Exp. Neurol. 184:997, 2003. 34. Irvine, R. F.: How is the level of free arachidonic acid controlled in mammalian cells? Biochem. J. 204:3, 1982. 35. Jarasch, E. D., Grund, C., Bruder, G., et al.: Localization of xanthine oxidase in mammary-gland epithelium and capillary endothelium. Cell 25:67, 1981. 36. Jennings, R. B. and Reimer, K. A.: Lethal myocardial ischemic injury. Am. J. Pathol. 102:241, 1981. 37. Johnson, P. C., Doll, S. C., and Cromey, D. W.: Pathogenesis of diabetic neuropathy. Ann. Neurol. 19:450, 1986. 38. Joyner, W. L., and Davis, M. J.: Pressure profile along the microvascular network and its control. Fed. Proc. 46:266, 1987. 39. Karpen, C. W., Pritchard, K. A. Jr., Merola, A. J., and Panganamala, R. V.: Alterations of the prostacyclinthromboxane ratio in streptozotocin induced diabetic rats. Prostaglandins Leukot. Med. 8:93, 1982. 40. Kihara, M., and Low, P. A.: Regulation of rat nerve blood flow: role of epineurial alpha- receptors. J. Physiol. (Lond.) 422:145, 1990. 41. Kihara, M., and Low, P. A.: Impaired vasoreactivity to nitric oxide in experimental diabetic neuropathy. Exp. Neurol. 132:180, 1995.

Nerve Blood Flow and Microenvironment 42. Kihara, M., and Low, P. A.: Vasoreactivity to prostaglandins in rat peripheral nerve. J. Physiol. (Lond.) 484:463, 1995. 43. Kihara, M., McManis, P. G., Schmelzer, J. D., et al.: Experimental ischemic neuropathy: salvage with hyperbaric oxygenation. Ann. Neurol. 37:89, 1995. 44. Kihara, M., Nickander, K. K., and Low, P. A.: The effect of aging on endoneurial blood flow, hyperemic response and oxygen-free radicals in rat sciatic nerve. Brain Res. 562:1, 1991. 45. Kihara, M., Schmelzer, J. D., Kihara, Y., et al.: Efficacy of limb cooling on the salvage of peripheral nerve from ischemic fiber degeneration. Muscle Nerve 19:203, 1996. 46. Kihara, M., Weerasuriya, A., and Low, P. A.: Endoneurial blood flow in rat sciatic nerve during development. J. Physiol. (Lond.) 439:351, 1991. 47. Kihara, M., Zollman, P. J., Schmelzer, J. D., and Low, P. A.: The influence of dose of microspheres on nerve blood flow, electrophysiology, and fiber degeneration of rat peripheral nerve. Muscle Nerve 16:1383, 1993. 48. Kishi, Y., Nickander, K. K., Schmelzer, J. D., and Low, P. A.: Gene expression of antioxidant enzymes in experimental diabetic neuropathy. J. Peripher. Nerv. Syst. 5:3, 2000. 49. Korthals, J. K., and Wisniewski, H. M.: Peripheral nerve ischemia. Part 1. Experimental model. J. Neurol. Sci. 24:65, 1975. 50. Korthuis, R. J., Granger, D. N., Townsley, M. I., and Taylor, A. E.: The role of oxygen-derived free radicals in ischemiainduced increases in canine skeletal muscle vascular permeability. Circ. Res. 57:599, 1985. 51. Lagerlund, T. D., and Low, P. A.: A mathematical simulation of oxygen delivery in rat peripheral nerve. Microvasc. Res. 34:211, 1987. 52. Low, P. A., Lagerlund, T. D., and McManis, P. G.: Nerve blood flow and oxygen delivery in normal, diabetic, and ischemic neuropathy. Int. Rev. Neurobiol. 31:355, 1989. 53. Low, P. A., Nukada, H., Schmelzer, J. D., et al.: Endoneurial oxygen tension and radial topography in nerve edema. Brain Res. 341:147, 1985. 54. Low, P. A., Schmelzer, J. D., and Ward, K. K.: The effect of age on energy metabolism and resistance to ischaemic conduction failure in rat peripheral nerve. J. Physiol. (Lond.) 374:263, 1986. 55. Low, P. A., Schmelzer, J. D., Ward, K. K., and Yao, J. K.: Experimental chronic hypoxic neuropathy: relevance to diabetic neuropathy. Am. J. Physiol. 250:E94, 1986. 56. Low, P. A., and Tuck, R. R.: Effects of changes of blood pressure, respiratory acidosis and hypoxia on blood flow in the sciatic nerve of the rat. J. Physiol. (Lond.) 347:513, 1984. 57. Low, P. A., Ward, K., Schmelzer, J. D., and Brimijoin, S.: Ischemic conduction failure and energy metabolism in experimental diabetic neuropathy. Am. J. Physiol. 248:E457, 1985. 58. Low, P. A., Yao, J. K., Kishi, Y., et al.: Peripheral nerve energy metabolism in experimental diabetic neuropathy. Neurosci. Res. Commun. 21:49, 1997. 59. Lubawy, W. C., and Valentovic, M.: Streptozocin-induced diabetes decreases formation of prostacyclin from arachidonic acid in intact rat lungs. Biochem. Med. 28:290, 1982. 60. Lundborg, G.: Ischemic nerve injury: experimental studies on intraneural microvascular pathophysiology and nerve

61.

62. 63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73. 74. 75.

76.

77.

78.

79.

679

function in a limb subjected to temporary circulatory arrest. Scand. J. Plast. Reconstr. Surg. 6:3, 1970. Lundborg, G.: Structure and function of the intraneural microvessels as related to trauma, edema formation, and nerve function. J. Bone Joint Surg. Am. 57:938, 1975. McCord, J. M.: Oxygen-derived free radicals in postischemic tissue injury. N. Engl. J. Med. 312:159, 1985. McManis, P. G., and Low, P. A.: Factors affecting the relative viability of centrifascicular and subperineurial axons in acute peripheral nerve ischemia. Exp. Neurol. 99:84, 1988. McManis, P. G., Low, P. A., and Yao, J. K.: Relationship between nerve blood flow and intercapillary distance in peripheral nerve edema. Am. J. Physiol. 251:E92, 1986. Mitsui, Y., Okamoto, K., Martin, D. P., et al.: The expression of proinflammatory cytokine mRNA in the sciatic-tibial nerve of ischemia-reperfusion injury. Brain Res. 844:192, 1999. Mitsui, Y., Schmelzer, J. D., Zollman, P. J., et al.: Alphalipoic acid provides neuroprotection from ischemiareperfusion injury of peripheral nerve. J. Neurol. Sci. 163:11, 1999. Mitsui, Y., Schmelzer, J. D., Zollman, P. J., et al.: Hypothermic neuroprotection of peripheral nerve of rats from ischaemiareperfusion injury. Brain 122:161, 1999. Mitsui, Y., Schmelzer, J. D., Zollman, P. J., et al.: Hypothermic neuroprotection of peripheral nerve of rats from ischemia-reperfusion injury: intraischemic vs. reperfusion hypothermia. Brain Res. 827:63, 1999. Mohri, D., Satomi, F., Kondo, E., et al.: Change in gene expression in facial nerve nuclei and the effect of superoxide dismutase in a rat model of ischemic facial paralysis. Brain Res. 893:227, 2001. Moncada, S., and Vane, J. R.: Arachidonic acid metabolites and the interactions between platelets and blood-vessel walls. N. Engl. J. Med. 300:1142, 1979. Myers, R. R., Murakami, H., and Powell, H. C.: Reduced nerve blood flow in edematous neuropathies: a biomechanical mechanism. Microvasc. Res. 32:145, 1986. Nagamatsu, M., Schmelzer, J. D., Zollman, P. J., et al.: Ischemic reperfusion causes lipid peroxidation and fiber degeneration. Muscle Nerve 19:37, 1996. Neely, J. R., and Feuvray, D.: Metabolic products and myocardial ischemia. Am. J. Pathol. 102:282, 1981. Nukada, H.: Increased susceptibility to ischemic damage in streptozocin-diabetic nerve. Diabetes 35:1058, 1986. Nukada, H.: The susceptibility of rat diabetic nerve to ischemia: increased or decreased? J. Neurol. Sci. 119:162, 1993. Nukada, H., Anderson, G. M., and McMorran, P. D.: Reperfusion nerve injury: pathology due to reflow and prolonged ischaemia. J. Peripher. Nerv. Syst. 2:60, 1997. Nukada, H., and Dyck, P. J.: Microsphere embolization of nerve capillaries and fiber degeneration. Am. J. Pathol. 115:275, 1984. Nukada, H., and McMorran, P. D.: Perivascular demyelination and intramyelinic oedema in reperfusion nerve injury. J. Anat. 185:259, 1994. Nukada, H., McMorran, P. D., and Shimizu, J.: Acute inflammatory demyelination in reperfusion nerve injury. Ann. Neurol. 47:71, 2000.

680

Functional Compartments and Nerve-Fiber Environment

80. Omawari, N., Dewhurst, M., Vo, P., et al.: Deficient nitric oxide responsible for reduced nerve blood flow in diabetic rats: effects of L-NAME, L-arginine, sodium nitroprusside and evening primrose oil. Br. J. Pharmacol. 118:186, 1996. 81. Palinski, W., Torsellini, A., and Doni, L.: Influence of platelet activation of erythrocyte deformability. Thromb. Haemost. 49:84, 1983. 82. Parry, G. J., and Brown, M. J.: Arachidonate-induced experimental nerve infarction. J. Neurol. Sci. 50:123, 1981. 83. Parry, G. J., and Brown, M. J.: Selective fiber vulnerability in acute ischemic neuropathy. Ann. Neurol. 11:147, 1982. 84. Qi, W. N., Yan, Z. Q., Whang, P. G., et al.: Gene and protein expressions of nitric oxide synthases in ischemia-reperfused peripheral nerve of the rat. Am. J. Physiol. Cell Physiol. 281:C849, 2001. 85. Rechthand, E., Hervonen, A., Sato, S., and Rapoport, S. I.: Distribution of adrenergic innervation of blood vessels in peripheral nerve. Brain Res. 374:185, 1986. 86. Rechthand, E., Smith, Q. R., Latker, C. H., and Rapoport, S. I.: Altered blood-nerve barrier permeability to small molecules in experimental diabetes mellitus. J. Neuropathol. Exp. Neurol. 46:302, 1987. 87. Ritchie, J. M.: A note on the mechanism of resistance to anoxia and ischaemia in pathophysiological mammalian myelinated nerve. J. Neurol. Neurosurg. Psychiatry 48:274, 1985. 88. Rosen, P., Senger, W., Feuerstein, J., et al.: Influence of streptozotocin diabetes on myocardial lipids and prostaglandin release by the rat heart. Biochem. Med. 30:19, 1983. 89. Roth, D. M., Reibel, D. K., and Lefer, A. M.: Vascular responsiveness and eicosanoid production in diabetic rats. Diabetologia 24:372, 1983. 90. Rundquist, I., Smith, Q. R., Michel, M. E., et al.: Sciatic nerve blood flow measured by laser doppler flowmetry and [14C]iodoantipyrine. Am. J. Physiol. 248:H311, 1985. 91. Sasaki, H., and Low, P. A.: Extreme vasoreactivity of rat epineurial arterioles to vasopressin. Am. J. Physiol. 271:H1307, 1996. 92. Schlaepfer, W. W.: Calcium-induced degeneration of axoplasm in isolated segments of rat peripheral nerve. Brain Res. 69:203, 1974. 93. Schlaepfer, W. W.: Vesicular disruption of myelin simulated by exposure of nerve to calcium ionophore. Nature 265:734, 1977. 94. Schmelzer, J. D., Zochodne, D. W., and Low, P. A.: Ischemic and reperfusion injury of rat peripheral nerve. Proc. Natl. Acad. Sci. U. S. A. 86:1639, 1989. 95. Seid, J. M., Jones, P. B., and Russell, R. G.: The presence in normal plasma, serum and platelets of factors that stimulate the production of prostacyclin (PGI2) by cultured endothelial cells. Clin. Sci. 64:387, 1983. 96. Smith, K. J., Hall, S. M., and Schauf, C. L.: Vesicular demyelination induced by raised intracellular calcium. J. Neurol. Sci. 71:19, 1985.

97. Snopko, R., Guffy, T., Rafelson, M., and Hall, E.: Serum stimulation of prostacyclin synthesis in aortically, venously and microvascularly derived endothelial cells. Clin. Physiol. Biochem. 5:70, 1987. 98. Stewart, M. A., Passonneau, J. V., and Lowry, O. H.: Substrate changes in peripheral nerve during ischaemia and wallerian degeneration. J. Neurochem. 12:719, 1965. 99. Sundqvist, T., Oberg, P. A., and Rapoport, S. I.: Blood flow in rat sciatic nerve during hypotension. Exp. Neurol. 90:139, 1985. 100. Takeuchi, M., and Low, P. A.: Dynamic peripheral nerve metabolic and vascular responses to exsanguination. Am. J. Physiol. 253:E349, 1987. 101. Timperley, W. R., Boulton, A. J., Davies-Jones, G. A., et al.: Small vessel disease in progressive diabetic neuropathy associated with good metabolic control. J. Clin. Pathol. 38:1030, 1985. 102. Tuck, R. R., Schmelzer, J. D., and Low, P. A.: Endoneurial blood flow and oxygen tension in the sciatic nerves of rats with experimental diabetic neuropathy. Brain 107:935, 1984. 103. Wachtler, J., Mayer, C., Rucker, F., and Grafe, P.: Glucose availability alters ischaemia-induced changes in intracellular pH and calcium of isolated rat spinal roots. Brain Res. 725:30, 1996. 104. Ward, K. K., Low, P. A., Schmelzer, J. D., and Zochodne, D. W.: Prostacyclin and noradrenaline in peripheral nerve of chronic experimental diabetes in rats. Brain 112:197, 1989. 105. Wong, P. Y., and Cheung, W. Y.: Calmodulin stimulates human platelet phospholipase A2. Biochem. Biophys. Res. Commun. 90:473, 1979. 106. Ziboh, V. A., Maruta, H., Lord, J., et al.: Increased biosynthesis of thromboxane A2 by diabetic platelets. Eur. J. Clin. Invest. 9:223, 1979. 107. Zochodne, D. W., Cheng, C., and Sun, H.: Diabetes increases sciatic nerve susceptibility to endothelin-induced ischemia. Diabetes 45:627, 1996. 108. Zochodne, D. W., and Ho, L. T.: Influence of perivascular peptides on endoneurial blood flow and microvascular resistance in the sciatic nerve of the rat. J. Physiol. (Lond.) 444:615, 1991. 109. Zochodne, D. W., and Ho, L. T.: Normal blood flow but lower oxygen tension in diabetes of young rats: microenvironment and the influence of sympathectomy. Can. J. Physiol. Pharmacol. 70:651, 1992. 110. Zochodne, D. W., Ho, L. T., and Gross, P. M.: Acute endoneurial ischemia induced by epineurial endothelin in the rat sciatic nerve. Am. J. Physiol. 263:H1806, 1992. 111. Zochodne, D. W., and Low, P. A.: Adrenergic control of nerve blood flow. Exp. Neurol. 109:300, 1990. 112. Zollman, P. J., Awad, O., Schmelzer, J. D., and Low, P. A.: Effect of ischemia and reperfusion in vivo on energy metabolism of rat sciatic-tibial and caudal nerves. Exp. Neurol. 114:315, 1991.

31 Pathology of Peripheral Neuron Cell Bodies MICHAEL J. GROVES AND FRANCESCO SCARAVILLI

Pathologic Alterations in Peripheral Neuron Cell Bodies following Axotomy Chromatolysis Histologic and Ultrastructural Features Factors Affecting the Extent and Duration of Chromatolysis after Axotomy Chromatolysis Produced by Insults Other Than Axotomy Summary Intracytoplasmic Vacuolation Histologic and Ultrastructural Features of Axotomy-Induced Vacuolation Vacuolation Produced by Insults Other Than Axotomy Summary Biochemical Alterations Early Changes Cytoskeletal Proteins Neuropeptides Transcription Factors Enzymes Glycoconjugates Neurotrophic Factors and Receptors Ion Channels Summary Neuron Death Mechanisms of Neuron Death Identified Biochemical Events in Neuron Death Factors Affecting the Magnitude of Neuron Death Summary

Pathologic Alterations in the Cell Bodies of Peripheral Neurons in Age and Disease Aging General Observations Lipofuscin

Melanin Neurofibrillary Tangles Hereditary Neuropathies Hereditary Motor and Sensory Neuropathy (Charcot-Marie-Tooth Disease) Hereditary Sensory and Autonomic Neuropathy Types I through V Spinal Motor Atrophy Hirschsprung’s Disease Giant Axonal Neuropathy Neuroaxonal Dystrophy Defects in DNA Repair (Ataxia-Telangiectasia, Xeroderma Pigmentosum, Cockayne’s Syndrome) Disorders Caused by Trinucleotide Repeat Instability Friedreich’s Ataxia Spinocerebellar Ataxias X-Linked Spinal and Bulbar Muscular Atrophy (Kennedy’s Disease) Neuronal Intranuclear Hyaline Inclusion Disease Lysosomal Storage Disorders Gaucher’s Disease GM1 Gangliosidosis GM2 Gangliosidosis Neuronal Ceroid Lipofuscinosis (Batten Disease) Sphingomyelin Lipidosis (Niemann-Pick Disease) Mucopolysaccharidosis Mucolipidosis Mannosidosis Fabry’s Disease Type II Glycogenosis (Pompe’s Disease) and Other Glycogen Storage Diseases Chédiak-Higashi Syndrome

Leukodystrophies Refsum’s Disease Tangier Disease (Hereditary High-Density Lipoprotein Deficiency) Neurologic Mutants Dystonia Musculorum The Sprawling Mouse and the Mutilated Foot Rat The Wobbler Mouse Mouse Motor Neuron Degeneration Inherited Dog Neuropathies Neuropathies Produced by Systemic Metabolic Disorders Porphyria Diabetes Mellitus Amyloidosis Infective and Inflammatory Neuropathies Varicella Zoster and Herpes Zoster Infection Poliovirus Cytomegalovirus Infection Human Immunodeficiency Virus Other Viruses Producing Peripheral Neuropathy Diphtheria Leprosy Sensory Ganglionitis in Sjögren’s Syndrome Lymphoma Paraneoplastic Neuropathy Toxins Hexacarbons Metals Vincristine Pyridoxine Amiodarone Acrylamide Organophosphates

683

684

Pathology of the Peripheral Nervous System

The study of the pathologic alterations to neuronal perikarya following axotomy has been an area of research since Nissl reported alterations in the staining properties and appearance of motor neurons.263 Since the last edition of Peripheral Neuropathy, many advances in technique have enabled the molecular events underlying pathologic alterations, as well as those involved in the death of neurons, to be studied. We summarize these new data from animals and humans, as well as some of the historical background, by dividing this chapter into two parts. The first largely concerns the effects of axotomy upon peripheral neurons in experimental animals, with some data from human studies. The second describes pathologic alterations seen in defined human neuropathies, with some data from animal models.

Pathologic Alterations in Peripheral Neuron Cell Bodies following Axotomy CHROMATOLYSIS Histologic and Ultrastructural Features Chromatolysis, also known as the axon reaction, encompasses a sequence of morphologic changes in neuronal cell bodies following an insult, most often after axotomy. Most reports have involved the use of histologic DNA/ RNA stains such as toluidine blue and cresyl fast violet to visualize these changes because they show the Nissl substance very clearly4,50,154,168,263,362,363,394,494 (Fig. 31–1). The morphologic features of axotomy-induced chromatolysis in adult animals appear to be relatively consistent in motor, sensory, and sympathetic138,139,263,371 neuron populations, as well as in different species.44,50,291,357,363,371 The term chromatolysis refers to the dispersion of the clumps of Nissl substance (rough endoplasmic reticulum) from the center to the periphery of the perikaryon, resulting in loss of RNA staining in the central region (Fig. 31–1). In the case of axotomy-induced chromatolysis, this can progress to, and result in, an almost complete absence of staining in the center of the neuron. The first indication of chromatolysis is often the relocation of the nucleus from the center to the periphery of the perikaryon (nuclear eccentricity); this relocation may persist far longer than the loss of central basophilia.232 Indentations on the inward side of the eccentric nucleus340 may also be a feature (Fig. 31–1D). Ultrastructurally, the perikaryal changes involve a disruption of the cytoskeletal network of filaments and microtubules, which also become shorter, leading to a dispersion of rough endoplasmic reticulum to the periphery

and alteration of neuron shape.291 The process of chromatolysis does not itself appear to affect the speed of fast axonal transport,248 which is carried out by microtubules. The disappearance of the paraphyte processes of satellite cells, which normally interdigitate with the cytoplasm of the associated DRG neuron, has also been observed265 (Fig. 31–1D). The normal function of these processes and cells is not known; “glial” cells around large-diameter dorsal root ganglion (DRG) neurons proliferate after axotomy or nerve damage in the adult rat and human, often forming “perineuronal onion bulbs,”392 the function or significance of which is not clear. Another alteration often seen in conjunction with chromatolysis is swelling or shrinkage of the perikaryon, although care should be taken when evaluating volume changes in peripheral neurons because even very small variations in fixation technique will have quite profound consequences in relation to perikaryal volume. Furthermore, if proportions of large and small neurons are being measured, a preferential loss of small or large neurons could be misinterpreted as an overall decrease or increase in mean perikaryal volume, respectively. The occurrence of swelling or shrinkage seems to depend upon the type and duration of injury as well as the type of neuron and age of the animal. For example, following permanent axotomy of the sciatic nerve, chromatolytic adult rat DRG neurons undergo a decrease in perikaryal size initially affecting all sizes of neuron; however, after 6 weeks, only the larger neurons were reduced in size.367 This was maintained for up to 26 weeks. In contrast, transient axotomy of the sciatic nerve (produced by a crush injury) produced an initial decline followed by an increase in mean perikaryal area to levels above control values.366,494 There are many other reports of volume changes in peripheral neuron perikarya following peripheral axotomy,263 but in general it seems that permanent axotomy leads to a permanent decrease in volume in some or all of the injured neurons, whereas transient axotomy followed by axonal regeneration leads to an increased perikaryal volume that is linked to certain biochemical alterations (see below). Nuclear and nucleolar enlargement also seems to be a feature of axotomy-induced chromatolysis. Adult hamster facial neurons showed an increase in size that peaked at 2 days after transection of the facial nerve, preceding the peak chromatolytic response at 4 days.221 Nuclei and nucleoli of adult rat DRGs were also reported to show phasic increases in volume after crushing of the sciatic nerve, peaking at 3 to 4 and 8 to 11 days later, respectively.494 In the latter study significant volume changes were found in neuronal perikarya, nuclei, and nucleoli in the contralateral, uninjured DRGs: Such bilateral effects of a unilateral nerve injury have been reported to occur in rat trigeminal ganglia324 and elsewhere.242

FIGURE 31–1 A, Photomicrograph of normal adult rat lumbar DRG neurons stained with cresyl fast violet (Nissl stain). The large, circular nuclei are positioned in the center of the perikaryon, and show very little chromatin. They have either a single large, centrally positioned nucleolus in large neurons (arrow), or two to four small, peripherally located nucleoli in “small dark” neurons (arrowheads). The Nissl substance (rough endoplasmic reticulum) is evenly distributed in smaller or larger clumps throughout the cytoplasm. Bar: 10 ␮m. B, Chromatolytic features in an adult rat large DRG neuron stained with cresyl fast violet, 2 weeks after transection of the sciatic nerve. The nucleus and Nissl substance are displaced to the periphery, resulting in a lack of staining in the center of the neuron. Bar: 10 ␮m. C, Chromatolytic human spinal motor neuron from an autopsy of a patient with Guillain-Barré syndrome. Displacement of the nucleus and Nissl substance to the periphery and lack of central staining are evident. (Luxol fast blue/cresyl fast violet stain.) Bar: 10 ␮m. D, Electron micrograph of a chromatolytic adult rat DRG neuron 2 weeks after sciatic transection. The nucleus (N) is indented on the side facing the center of the neuron, a feature often seen in chromatolytic neurons. The rough endoplasmic reticulum (Nissl substance) is concentrated around the periphery with the mitochondria, Golgi apparatus, and other organelles concentrated in the center. This can clearly be seen at a higher magnification (E). Bar: 1 ␮m.

686

Pathology of the Peripheral Nervous System

Factors Affecting the Extent and Duration of Chromatolysis after Axotomy The age of the animal when axotomy occurs appears to be a crucial determinant in the appearance, speed of onset, duration, and outcome of chromatolysis, and whether it is seen at all. Peripheral nerve transection in newborn (⬍24 hours old) rats, hamsters, or guinea pigs does not produce any chromatolytic DRG neurons: affected neurons become smaller and more intensely stained,34 and many die.363 However, sciatic transection in 7- to 8-day-old kittens did lead to chromatolysis of some DRG neurons.5,368 Nuclear and nucleolar swelling220 was also absent in axotomized facial motor neurons of 15-day-old hamsters, but did occur at 25 days. The authors suggested that this difference may be due to the younger neurons being at the peak growth phase, and not capable of further enlargement. This sharp delineation with regard to age in the morphologic response to axotomy indicates a critical difference in neuronal metabolism between newborn and slightly older animals: Many rat DRG neurons do not attain their mature phenotype until the week after birth.22 Kerezoudi et al.232 reported a greater proportion of DRG neurons with eccentric nuclei at 6 months after peripheral axotomy performed on young rats than at 6 months after axotomy performed on aged rats. Chromatolysis appears to be more intense and resolves less rapidly in old animals than in young ones.233 The location of the injury or insult and the type of peripheral neuron are also major determinants in the time of onset and extent of chromatolytic changes. Axotomy of the centrally directed axons of DRG neurons (rhizotomy) does not produce any observable morphologic change in the perikarya of the affected neurons, regardless of the distance of the injury from the DRG.264 Likewise, proximal section of the central axons of the trigeminal and geniculate nerves and the inferior ganglia of the vagus and glossopharyngeal nerves of the monkey did not produce any significant chromatolysis in these ganglia, in contrast to transection of their peripheral axons.44 A lesion to the peripheral axons of DRG neurons induces a chromatolytic response in perikarya 24 to 48 hours later, depending upon the distance of the injury from the DRG.152,264 Small DRG neurons seem to respond to axotomy earlier than large neurons,363 a difference that is reflected in the increased expression of growth-associated protein-43 (GAP-43) that occurs in small DRG neurons before large ones.428 Adult rat thoracic spinal cord hemisection leads to chromatolytic changes in some lumbosacral motoneurons after 21 days,315 which is an example of transsynaptic effects because the motor neurons are unlikely to have been injured directly by the lesion. Following peripheral axotomy of motor axons, any changes start to be seen within 7 days.363,371 Unlike adult rat DRG neurons, few

(if any) chromatolytic adult rat spinal motor neurons are seen 1 month after sciatic transection (M. J. Groves and F. Scaravilli, personal observations). The duration of the chromatolytic response following axotomy depends largely on whether axons are permitted to regenerate or not. Crush injuries (axonotmesis) produce a transient axotomy that, as long as the perineurium is not breached, regenerates at the same rate in adult rats regardless of the width or duration of the crush.37 Resolution of the chromatolytic appearance occurs once contact with the peripheral targets been reestablished.367,444 Permanent axotomy, in which axonal regeneration is prevented, results in a longer duration of the chromatolytic response, which gradually diminishes.163,366

Chromatolysis Produced by Insults Other Than Axotomy Diseases and toxins that disrupt axonal transport or stimulate a regenerative response may produce the alterations characteristic of chromatolysis in peripheral neurons by producing a functional axotomy, or by inducing the same pattern of gene expression that follows axotomy. Chromatolytic motor neurons have been found in autopsies of patients with a wide range of acquired and inherited diseases, such as chronic inflammatory demyelinating polyneuropathy,316 acute motor axonal neuropathy/acute motor and sensory axonal neuropathy,161 WerdnigHoffmann disease,61 poliovirus,30 amyotrophic lateral sclerosis,281 macroglobulinemia,373 and radiation injury in dogs.25,354 Chromatolysis of enteric72 and autonomic neurons174 has been reported in cases of equine grass sickness. Chromatolytic neurons have been reported in sensory ganglia in certain inherited progressive sensory neuropathies in dogs,78 mice,94 and humans (see below). Toxins will also produce chromatolysis of sensory neurons, particularly those that inhibit axonal transport. Examples include intrathecal injection of local anesthetics329 and intraneural injection of cytotoxic lectins,500,516 systemic 2,5-hexanedione,432 3-acetylpyridine,20 and capsaicin.192 Sympathectomy using an immunotoxin will lead to a large proportion of sympathetic neuron cell bodies becoming chromatolytic in the adult rat.348 Drugs that block axonal transport provide interesting clues regarding the initiation of chromatolysis. Following transection of the hypoglossal nerve of adult cats and rats, a subepineurial injection of colchicine will delay the onset of chromatolysis and the increased uptake of 2-deoxyglucose (an early indicator of injury) in the adult rat hypoglossal nucleus.419 Furthermore, subepineurial injections of extracts of transected nerve into uninjured hypoglossal nerves stimulated 2-deoxyglucose uptake in the hypoglossal nucleus.420 Local application of vinblastine

Pathology of Peripheral Neuron Cell Bodies

(which inhibits microtubule polymerization and axonal transport) to the rat sciatic nerve produces some, but not all, of the changes associated with chromatolysis.121 Systemic administration to adult rats causes swelling of large DRG neurons that appears to be due to the accumulation of neurofilaments following the inhibition of anterograde transport.464

Summary When all of the disparate insults that can produce the appearance of chromatolysis in peripheral neurons are considered, the overall impression is that a stimulus produced in nerve at the site of injury is involved in triggering the chromatolytic response. The appearance of chromatolysis may therefore be governed by mechanisms that are switched on by the specific activation of genes involved in axonal regeneration that are quite distinct from those expressed during axonal elongation in development.266 This may also explain why neonatal responses to axotomy are qualitatively (and quantitatively) different from those in the adult. When the concomitant biochemical changes that occur after axotomy are considered, chromatolysis does appear to be a process associated with a regenerative, and not a degenerative, response to an insult by neurons.

INTRACYTOPLASMIC VACUOLATION Histologic and Ultrastructural Features of Axotomy-Induced Vacuolation There are occasional reports in the literature of “atypical” neurons in ipsilateral DRGs following axotomy or peripheral nerve damage that have single or multiple large intracytoplasmic vacuoles,3,44,50,164,232,290,324 up to 50 ␮m in diameter (Fig. 31–2). They usually have a low incidence but persist for long periods of time if axonal regeneration is prevented.50,164,232,290 This form of pathology is not confined to DRG neurons: Occasional grossly distorted, vacuolated sympathetic neurons have been observed in the adult rat 1 month after sciatic transection (M. J. Groves and F. Scaravilli, personal observations). Large intracytoplasmic vacuoles have been reported in adult mouse motor neurons 2 weeks after ventral root avulsion, an injury that leads to the death of a significant number of them.261 The morphology of vacuolated adult rat DRG neurons after peripheral axotomy is relatively constant: Neurons develop large spherical spaces that increase the perikaryal volume, often reducing the neuronal cytoplasm to a thin rim around the circumference and compressing the nucleus (see Fig. 31–2B). These vacuoles usually occur singly, leading Cavanaugh50 to call the affected neurons

687

“signet ring” cells, but neurons containing multiple vacuoles are not uncommon. Authors have assumed that they are not large invaginations (or “insudation”220) of the plasma membrane. Their intracellular location is confirmed by the absence of systemically administered horseradish peroxidase.164 Ultrastructurally, the vacuoles seen after axotomy appear to be identical to those seen after acrylamide intoxication in rats.220 They are membrane bound and have occasional villus-type protrusions of neuronal cytoplasm extending into the vacuole (see Fig. 31–3G below). They are filled with an aqueous phase containing a sparse, flocculent material (seen in electron micrographs) that is probably proteinaceous in nature and are probably derived from swollen Golgi apparatus or endoplasmic reticulum. Vacuolated rat DRG neurons induced by peripheral axotomy belong to the large DRG neuron population, based on the expression of large DRG neuron markers such as stage-specific embryonic antigen-4 (SSEA-4) and phosphorylated neurofilament (see Fig. 31–2F). Vacuolated neurons actively transport a retrograde tracer (the B subunit of cholera toxin) (see Fig. 31–2D); they express the TrkC neurotrophin-3 (NT-3) receptor (M. J. Groves and F. Scaravilli, personal observations), and their incidence can be reduced by administration of NT-3 (but not nerve growth factor [NGF] or brain-derived neurotrophic factor [BDNF]) to the proximal stump.164 Axotomy-induced vacuolation did not appear to lead to the immediate death of affected neurons, because the proportion of axotomized neurons expressing large neuron markers actually increased overall. We have found only one vacuolated neuron that could be stained using the in situ end-labeling technique for single-stranded DNA,498 and other authors using the terminal deoxynucleotidyl transferase–mediated dUTP nick end-labeling (TUNEL) method for DNA strand breaks137 similarly found only occasional ones that could be stained.290 None of these stained neurons showed condensed bodies of DNA characteristic of apoptosis, indicating that other mechanisms (such as excitotoxicity) may be involved in this DNA damage that could lead to the death of some vacuolated neurons. Vacuolated neurons can be seen in uninjured DRGs from rats at an incidence of around 0.005% to 0.01%, and their numbers seem to increase with age.164 Kerezoudi et al.232 also found that they increase in incidence with age in rabbit DRGs. Following successful regeneration of axons after crushing the sciatic nerve, the incidence of vacuolation returned to control levels by 2 to 3 months; after permanent transection their incidence remained relatively high (1% to 2%), gradually declining over 6 months to levels that were still higher than in controls.164 It would appear that, like chromatolytic neurons, the rate of disappearance or resolution of large

688

Pathology of the Peripheral Nervous System

FIGURE 31–2 A, Cresyl fast violet–stained adult rat L4 DRG neuron containing a single large vacuole, 1 month after sciatic nerve transection. Bar: 10 ␮m. B, Cresyl fast violet–stained adult rat L4 DRG neuron containing three vacuoles (same animal as A). Bar: 10 ␮m. C, Hematoxylin and eosin–stained adult rat sympathetic neuron from a sacral sympathetic ganglion, 1 month after sciatic transection, containing a single large vacuole. Bar: 10 ␮m. D, Vacuolated adult rat lumbar DRG neurons stained for peroxidase, 40 hours after injection of a cholera toxin–horseradish peroxidase conjugate into the sciatic nerve proximal stump transected 2 months previously. This showed that vacuolated neurons are axotomized neurons, possess axons projecting to the lesion site 2 months after axotomy, and are capable of active transport. Bar: 10 ␮m. Figure continued on opposite page

vacuoles in DRG neurons is governed by the rate of axonal regeneration.

Vacuolation Produced by Insults Other Than Axotomy Toxins and ionizing radiation also lead to vacuolated, or “cavitated,” sensory neurons. Acrylamide220 produces vacuolation of some rat DRG neurons at the same dose as that required to trigger the appearance of chromatolysis in DRG neurons within 1 week of administration (30 mg/kg391). Intraneural pronase injection,254 neonatal exposure to capsaicin,192 3-acetylpyridine administration,20 and acute radiation injury354 also produced DRG neuron vacuolation, indicating that vacuolation may be a

general reaction to insult. Most of these insults are thought to interfere with axonal transport, and (as in chromatolysis) this may be the key underlying event that triggers vacuolation. Vacuolation has also been reported in sympathetic neurons of young rats following NT-3 antiserum treatment272 and exposure to large doses of ethanol.209,210 The cytostatic drug doxorubicin (Adriamycin), when injected into the tongue, produces intracytoplasmic vacuolation in mouse hypoglossal motor neurons 14 days later,29 as does trimethyl tin and methylmercury intoxication in rat and gerbil spinal motor neurons.325,434 Vacuolated spinal motor neurons have also been seen in pigs with an inherited lower motor neuron disease,336 and in cases of porphyria49 and tetanus60 in humans. Vacuolated human DRG and sympathetic neurons

Pathology of Peripheral Neuron Cell Bodies

689

FIGURE 31–2 Continued E, Cryosections of adult rat lumbar DRG 1 month after sciatic transection, immunostained for growth-associated protein-43 (GAP-43). Vacuolated neurons seen after peripheral axotomy always demonstrate intense GAP-43 immunoreactivity. Bar: 10 ␮m. F, Cryosections of adult rat lumbar DRG 1 month after sciatic transection immunostained for the oligosaccharide conjugate stage-specific embryonic antigen-4 (SSEA-4), which is expressed by a subset of large sensory neurons. Vacuolated neurons express large sensory neuron markers such as SSEA-4 and phosphorylated neurofilament. Bar: 10 ␮m. G, Electron micrograph of an intracytoplasmic vacuole (v) within an adult rat L4 DRG neuron 2 months after transection of the ipsilateral sciatic nerve. Vacuoles are filled with an aqueous phase, with villus-like cytoplasmic processes occasionally projecting into it (arrowhead). Abundant lipofuscin (arrows) is usually present in vacuolated neurons. (Uranyl acetate and lead citrate stain.) Bar: 1 ␮m.

have also been described in diabetic neuropathy,92,156 although one of these reports is probably describing postmortem artifact in human DRGs rather than true vacuolation.156 Vacuolated human DRG neurons are also seen in herpes zoster infection.90 There are reports in the literature of “vacuolization” involving the formation of small fluid-filled vacuoles. This pathology seems to be associated with an excitotoxic insult to neurons, such as laser ablation of spiral ganglion neurons423 or in motor neurons following ventral root avulsion, and is further discussed in the section on Neuron Death.64 “Vacuolization” has also been reported in a transgenic mouse model of motor neuron degeneration (Mnd) involving the overexpression of superoxide dismutase 1,21 and vacuolation of motor neuron perikarya was found in one patient (of eight examined) with Creutzfeldt-Jakob disease.508

Summary Although grossly distorting intracytoplasmic vacuolation is seen as an incidental finding in DRGs of normal rats (particularly aged rats), its increased incidence is a significant pathologic alteration that has been described in sensory,

sympathetic, and motor neuron perikarya after trauma or exposure to certain neurotoxins or radiation, and in some neuropathies such as diabetes. It is not known how these large vacuoles relate to the formation of smaller cytoplasmic vacuoles/vesicles (“vacuolization”), or to the chromatolysis that is normally also present in other neurons. The incidence of large vacuoles in DRG neurons appears to be dependent upon neuron type, and may be a specific reaction to injury by certain subsets of peripheral neurons. The fact that it has been largely ignored is probably due to its relatively low incidence; however, knowledge of the mechanisms involved in vacuolation may be useful in understanding wider disease processes affecting neurons in which it is seen.

BIOCHEMICAL ALTERATIONS Numerous alterations to the biochemical content and processes of peripheral neurons have been described after an insult, particularly in sensory and motor neurons following peripheral axotomy. Many of these changes seem to reflect a switch to a regenerative state; broadly, neurotransmitters, neuropeptides, and enzymes associated with

690

Pathology of the Peripheral Nervous System

nerve conduction are downregulated while proteins associated with axonal growth and neuroprotection are upregulated. It is becoming apparent that the biochemical changes described so far in the literature are part of a complex interaction between neurons, axons, and satellite and Schwann cells, as well as peripheral and central targets, seemingly mediated at least in part by cytokines and neurotrophic factors.409 The restoration of the normal biochemical content of peripheral neurons appears to be governed by considerations similar to those involved in resolution of chromatolysis, but there are exceptions to this generalization.

Early Changes The first biochemical changes described after peripheral axotomy appear to consist of upregulation of proteins that probably have a neuroprotective role, and the relationship between ionic and transcriptional changes (e.g., activating transcription factor; see below) in this early phase is not yet established. One of the earliest changes described in peripheral neurons is the increased uptake of 2-deoxyglucose in rat hypoglossal motor neurons,420 and a massive influx of calcium ions into the axoplasm of the transected axon that may not reach the perikaryon.531,532 Twenty-four hours after axotomy of the rat hypoglossal nerve, injured motor neurons upregulate the calcium-buffering protein calbindin, which may protect the neuron from the deleterious effects of free calcium ions.247 Calbindin expression after axotomy has also been reported in rat sympathetic ganglion neurons.380 Another potentially neuroprotective molecule, heat shock protein 27 (HSP27), a chaperone protein thought to protect cellular macromolecules from damage, is upregulated in DRG neurons within 48 hours after sciatic transection at mid-thigh level.71,259 The stressactivated protein kinase p38 pathway is activated in spinal motor neurons after axotomy, which leads to the induction of HSP25,313 and HSP72 expression was observed in rat superior cervical ganglion neurons 24 hours after axotomy.198 Increased activity of ornithine decarboxylase, an enzyme involved in the synthesis of polyamines, which may protect DNA from damage, was detected in DRG neurons 13 hours after sciatic injury,493 and also in rat superior cervical ganglion after axotomy.139 The urokinase plasminogen activator receptor starts to be upregulated in adult mouse DRG neurons 8 hours after a crush injury to the sciatic nerve,414 and probably facilitates axonal regeneration by dissolving the extracellular matrix.

Cytoskeletal Proteins Changes in the expression of cytoskeletal components following axotomy reflect, as with many other macromol-

ecules, a switch of synthetic activity from that required for neurotransmission to that for axonal regeneration. Neurofilament messenger RNA (mRNA) is downregulated, while tubulin and actin mRNAs are upregulated, in rat DRG neurons beginning at around 5 days after sciatic axotomy.327,504 This probably underlies decreases in perikaryal volume seen following insult, because neurofilament content is a major determinant of the size of the axon and perikaryon.127,193 Dorsal rhizotomy also produces alterations in microtubule, actin, and neurofilament synthesis, although of a smaller magnitude than those produced by peripheral axotomy.157 Peripherin, an intermediate filament protein normally expressed in small DRG neurons, is upregulated in large DRG neurons (Fig. 31–3C and D) of adult rats505 by 7 days after sciatic crush and declines to control levels as axons regenerate. GAP43, a phosphoprotein involved in axonal elongation, is upregulated within 2 days in small-diameter DRG neurons after sciatic nerve transection, whereas L-type DRG neurons upregulate GAP-43 between 4 and 14 days later428 (Fig. 31–3B). Adult rat motor neurons demonstrate alterations similar to those in cytoskeletal components after axotomy that are dependent upon the distance from the site of peripheral axotomy to the cell body.115 ␣I- and ␤II-tubulin, actin, GAP-43,148,458,459 and peripherin470 are upregulated after axotomy (see Fig. 31–3), whereas neurofilament312,470 and microtubule-associated proteins443 are downregulated. Interestingly, peripheral axotomy in adult rats and cats will induce the dye coupling of spinal motor neurons normally only seen during embryonic development, without any alterations to connexin (a gap junction protein) mRNA expression.54,55 Avulsion of the sciatic nerve leads to the death of a proportion of the affected spinal motor neurons; this is preceded by an aberrant accumulation of phosphorylated neurofilaments in perikarya.283 Abnormal phosphorylation and glycosylation of neurofilaments has also been implicated in the loss of motor neurons in Werdnig-Hoffmann disease.61

Neuropeptides Neuropeptides are peptides or polypeptides synthesized by neurons that have neuromodulatory or neurotransmitter functions. With one or two exceptions, their synthesis decreases after axotomy as neuronal synthetic activity switches to proteins required for axonal regeneration. The neuropeptide galanin, thought to have an analgesic role, is not normally expressed at all but is upregulated in DRG neurons within 24 hours of sciatic transection in adult rats,485 and within 14 days in monkeys.525 The expression of substance P starts to decline in DRG neurons around 3 days after sciatic transection.217,485 Similarly, calcitonin gene–related peptide (CGRP) is mainly expressed by smalldiameter DRG neurons but shows a more delayed decrease

Pathology of Peripheral Neuron Cell Bodies

691

FIGURE 31–3 A, Cryosection of normal adult rat lumbar DRG immunostained for growth-associated protein-43 (GAP-43). Immunoreactivity is confined to small- to medium-diameter neurons. Bar: 10 ␮m. B, Cryosection of adult rat ipsilateral lumbar DRG, 2 weeks after transection of the sciatic nerve. Following peripheral axotomy, neuronal GAP-43 immunoreactivity increases in intensity and in the proportion of immunoreactive neurons, with many larger diameter neurons becoming immunoreactive (arrow). Bar: 10 ␮m. C, Paraffin section of normal adult rat lumbar DRG immunostained for the intermediate filament protein peripherin. As with GAP-43, immunoreactivity is confined to the smaller DRG neurons. Bar: 10 ␮m. D, A 4-␮m wax section of adult rat lumbar DRG immunostained for peripherin 1 month after sciatic transection. Larger DRG neurons are starting to become immunoreactive (arrows). Bar: 10 ␮m. Figure continued on following page

in expression than substance P following axotomy165 (see Fig. 31–3E to H). However, the number of DRG neurons that express CGRP seems to increase following dorsal rhizotomy, possibly as a result of its accumulation in neurons.206 CGRP and substance P downregulation in DRG neurons after axotomy is probably due to decreased target-derived NGF,481 as shown when endogenous NGF was removed with administration of a specific antiserum.409 However, guinea pigs appear not to show decreased CGRP immunoreactivity after axotomy,376 whereas monkeys do.525 Other DRG neuron neuropeptides show altered patterns of expression beginning around 1 week after peripheral nerve injury: somatostatin and cholecystokinin levels are decreased while vasoactive peptide and neuropeptide Y are

upregulated, mainly in small DRG neurons.225,530 These alterations are largely reversed following successful axon regeneration. Unlike in DRG neurons, galanin, substance P, vasoactive peptide, somatostatin, and ␣-CGRP (see Fig. 31–3) are upregulated, and cholecystokinin downregulated, in rat spinal motor neurons following sciatic transection.15,349,525 Noradrenergic neurons of the adult rat superior cervical ganglion increase their expression of substance P, neuropeptide Y, galanin, and vasoactive intestinal peptide mRNA after axotomy.530 This upregulation is induced by leukemia inhibitory factor (LIF) produced in the lesioned nerve435 that, with NGF upregulation in macrophages, is partly triggered by interleukin-1.43 The differences

692

Pathology of the Peripheral Nervous System

FIGURE 31–3 Continued E, Cryosection of L5 DRG immunostained for calcitonin gene–related peptide (CGRP), a neuropeptide expressed in 50% to 60% of normal adult rat lumbar DRG neurons, mainly of small diameter. Bar: 10 ␮m. F, Cryosection of ipsilateral adult rat lumbar DRG immunostained for CGRP, 2 weeks after sciatic transection. CGRP expression is downregulated in DRG neurons after peripheral axotomy. Bar: 10 ␮m. G, Cryosection of normal adult rat lumbar spinal cord immunostained for CGRP, showing the faint immunoreactivity normally present in spinal motor neurons. Bar: 10 ␮m. H, Cryosections of adult rat lumbar spinal cord, immunostained for CGRP 1 month after sciatic transection. CGRP immunoreactivity in spinal motor neurons is transiently increased in intensity following peripheral axotomy. Bar: 10 ␮m.

between neuronal populations in the axotomy-induced alterations to neuropeptide expression are probably due to differences in trophic factor dependence as well as differences in the function of the neuropeptides concerned.

Transcription Factors Most, if not all, pathologic alterations to neurons are ultimately due to altered gene expression.184 These alterations are starting to be understood using molecular biologic techniques such as DNA microarrays and reverse transcription–polymerase chain reaction, but it is already evident that different neuronal populations show differing patterns of transcription factor activation and inhibition that probably underlie differences in reaction to insult.

One of the first transcription factors known to be upregulated following axotomy is activating transcription factor 3, which binds to DNA as a homodimer or as a heterodimer with c-Jun; the time course of its induction depends upon the distance of the nerve cell body from the site of injury.471 N-Shc mRNA is predominantly expressed in hypoglossal motor neurons compared with Shc mRNA, but axotomy reverses this pattern of expression while SCK and Grb2 mRNA expression is unaffected.450 The LIM-type homeobox gene islet-1 plays an important role in differentiation and axonal elongation of sensory and motor neurons during development; however, following peripheral axotomy in adult rats, islet-1 mRNA is downregulated, with maximal depletion occurring at 3 to 7 days, but returns to control levels when regeneration is complete.194 Six days after sciatic

693

Pathology of Peripheral Neuron Cell Bodies

nerve transection, when neuronal death is occurring, increased immunoreactivity for the transcription factor c-Jun and the proapoptotic protein Bax takes place in spinal cord motor neurons and small-diameter DRG neurons of young rats.144 c-Jun expression and activation is elevated in neonatal and adult rat sensory and motor neurons after sciatic axotomy, as well as in some uninjured sensory neurons, and remains elevated until axonal regeneration is complete.183 c-Jun expression does not predict neuronal death or survival according to one study,422 although another study found that the axotomy-induced upregulation of c-Jun starts within 3 hours in neonatal motor neurons, becoming maximal between 1 and 10 days, and expression of a certain form of c-Jun was associated with apoptotic motor neurons.46 c-Fos expression is elevated in neonatal and adult motor neurons after axotomy, but not altered in sensory neurons.183

Enzymes Alterations to enzyme activity following axotomy seem to reflect the change to a regenerating state: enzymes involved with neurotransmission are downregulated while those thought to have roles in axonal elongation and neuroprotection are upregulated. Motor neurons synthesize acetylcholinesterase and choline acetyltransferase, the enzymes responsible for the degradation and synthesis of acetylcholine, respectively. These enzymes are downregulated in lumbar spinal motor neurons following peripheral axotomy65,116 and thoracic spinal cord hemisection315 in the adult rat. Fluoride-resistant acid phosphatase (FRAP) is probably the same enzyme as thiamine monophosphatase, and is exclusive to the nonpeptidergic small-diameter DRG neuron population.256 This neuron subpopulation also expresses an oligosaccharide glycoconjugate recognized by the LA4 antibody and Griffonia simplicifolia isolectin B4.416 FRAP activity in rat DRGs starts to decline around 3 days after sciatic transection but recovers to control levels by 21 days,455 and does not decline at all after rhizotomy. Conversely, carbonic anhydrase activity is found in large-diameter DRG neurons and—like expression of another large-diameter DRG neuron marker, SSEA-4165—is unaffected by sciatic transection.346 Some enzyme levels are increased. Nitric oxide synthase (NOS) produces nitric oxide, an important molecule in many pathologic processes because it reacts with superoxide radicals to produce highly damaging hydroxyl and nitronium radicals.85 NOS is upregulated in adult rat,480,526 but not guinea pig or monkey,376,526 DRG neurons within 48 hours of sciatic axotomy, indicating a potentially important species difference. The levels of urokinase and tissue plasminogen activator mRNA increase in adult mouse DRG neurons 3 days after crushing the sciatic nerve; the proteins are transported to the regenerating axon tip, where they are thought to assist regeneration by dissolving the extracellular matrix.414

NOS upregulation is also induced in adult rat motor neurons following ventral root avulsion, and this can be blocked by administration of BDNF.326 The induction of NOS activity in rat facial motor neurons after peripheral axotomy seems to be age dependent, since neonatal motor neurons show little evidence of any increase in activity, and upregulation is only seen in rats over 1 week old.277 A subpopulation of adult rat sympathetic neurons (about 2%) upregulates NOS by 1 week following axotomy,239 while no increase in the number of NOS-expressing enteric neurons could be detected following a crush injury to gut or treatment with colchicine.104 Expression of superoxide dismutases, the copper-, zinc-, and manganese-containing enzymes that catalyze the breakdown of toxic superoxide radicals, is also increased in rat facial motor neurons by peripheral axotomy,523 which represents an increase in neuroprotection. Tyrosine kinase activity, as indicated by increased phosphotyrosine immunoreactivity, has been reported to increase in the hypoglossal nucleus and the dorsal motor nucleus of the vagus nerve 48 hours after axotomy of the hypoglossal and vagus nerves, respectively514; this is in keeping with the upregulation in phosphatase/kinase activity reported after axotomy.266

Glycoconjugates The function of the (known) oligosaccharide glycoconjugates in DRG neurons is unknown, although they are likely to be involved in cell-cell recognition in the dorsal horn of the spinal cord. As with carbonic anhydrase, glycoconjugates associated with large DRG neurons such as SSEA-4165 appear to be unaffected by sciatic transection. In contrast, the expression of oligosaccharide glycoconjugates by the small-diameter DRG neuron population decreases after peripheral axotomy,347 particularly the glycoconjugate recognized by the LA4 antibody and G. simplicifolia isolectin B4,165,496 which may be a sensitive indicator of peripheral axon injury.

Neurotrophic Factors and Receptors Neurotrophic factors are proteins secreted by tissues, glia, and sometimes neurons themselves. Among other functions, they maintain the phenotype of the target neuron population and may even be required for their survival in adulthood. Neurotrophic factor expression in peripheral neurons before and after insult shows wide differences between, and also within, motor, sensory, and sympathetic neuron populations. The most researched of these molecules are the neurotrophins, basic polypeptides with a characteristic and highly conserved structure that are ligands for the p75 lowaffinity neurotrophin receptor (p75NTR) and one of the three membrane-bound tyrosine kinase (Trk) receptors.456 NGF, BDNF, and NT-3 appear to be the most widely expressed

694

Pathology of the Peripheral Nervous System

and important neurotrophins for the peripheral nervous system in development, maintenance of neuronal phenotype, and survival.1,461,481 Other proteins that can stimulate or protect peripheral neurons are usually referred to as neurotrophic factors and include cytokines and growth factors.456 BDNF mRNA is normally expressed in a small number of adult rat DRG neurons that also express the high-affinity nerve growth factor receptor TrkA.12 It is upregulated in response to NGF administration,295 inflammation,59 and sciatic nerve crush and transection406,463 in rat DRG neurons that express TrkB (the high-affinity BDNF receptor) and TrkC (the high-affinity NT-3 receptor), and is transported to the spinal cord, where it may modulate somatosensory pathways. This altered pattern of expression in DRGs was apparent 2 days after sciatic transection,295 and would appear at present to be specific to sensory neurons. NT-3 and NGF mRNA and protein are normally expressed in some adult rat DRG neurons, but transection of the spinal nerve seems to produce a slight decrease in this neuronal expression 1 week later, while satellite cells surrounding large-diameter DRG neurons upregulate NT-3 and NGF mRNA and protein.527 This upregulation in non-neuronal cells is probably responsible for the reported overall increase of NGF mRNA in rat DRGs following axotomy,406 and has also been reported for glial cell line–derived neurotrophic factor (GDNF).175 Most adult rat lumbar motor neurons express mRNA for BDNF and the related neurotrophin-4, and peripheral nerve transection causes an upregulation of BDNF mRNA between 2 and 7 days later.176,240 However, adult rat nodose ganglion neurons contained no mRNA for any of the main neurotrophins (NGF, BDNF, NT-3) before or after transection of the vagus nerve.258 The neuronal expression of neurotrophin receptors in neurons also shows differing responses to peripheral axotomy. Immunoreactivity for p75NTR declines in ipsilateral DRG neurons,299 and is upregulated in the contralateral neurons, during the first week after sciatic transection. This becomes maximal at around 3 weeks and remains depressed for the duration of observation (8 weeks528); a similar effect occurs in sympathetic neurons.360 The expression of p75NTR in adult rat motor neurons is induced by axotomy, but only in those allowed to regenerate.41 Immunoreactivity and mRNA for the high-affinity NGF receptor TrkA in DRG neurons starts to be downregulated 3 days after spinal nerve ligation411 or sciatic transection.24,299 TrkA, TrkC, and p75NTR expression also declines in nodose ganglion neurons after vagotomy.257 Caution must be exercised when assessing data on TrkB and TrkC expression because of the existence of alternatively spliced isoforms that lack the catalytic kinase domain. TrkB and TrkC mRNA and protein decrease in DRG neurons 1 week after sciatic transection in young and old rats.24 However, when mRNA was extracted from the whole DRG, increased levels of mRNA for TrkB and TrkC were found 1 week after axotomy, which was largely due to an increase

in the nonkinase forms110 and which also shows that non-neuronal expression of nonkinase Trk receptors is upregulated in DRGs after axotomy. Adult motor neurons normally express the neurotrophin receptors TrkB and TrkC, but, following peripheral axotomy, mRNA for TrkC is downregulated while that of TrkB is upregulated.116,176 The more severe injury produced by ventral root avulsion leads to the downregulation of both TrkB and TrkC.176 The role of other neurotrophic factors and cytokines in neuronal responses to injury is starting to be clarified: GDNF, LIF, glial growth factor, transforming growth factor (TGF), and ciliary neurotrophic factor (CNTF) all have trophic effects on various peripheral neuron populations.456 Basic fibroblast growth factor (bFGF) is mitogenic for a wide variety of cell types of mesodermal and neuroectodermal origin; bFGF protein and mRNA begin to be upregulated in rat DRG neurons 15 hours after axotomy, before declining after 1 week.218 This period corresponds to the peak incidence of mitotic activity in glial cells of axotomized rat ganglia.126,270 LIF and GDNF receptor mRNAs are upregulated in adult rat spinal motor neurons following a crush injury to their target muscle, sciatic nerve axotomy, or ventral root avulsion.176 Intriguingly, mRNA for a critical component of signal transduction for the LIF/GDNF receptor complex, known as gp130, was downregulated after ventral root avulsion.176 CNTF receptor mRNA was downregulated and the LIF receptor upregulated in rat facial motor neurons within 24 hours of axotomy of the facial nerve, returning to control levels after successful regeneration.170 The mRNA for the GDNF receptor is also upregulated in rat DRG neurons after sciatic nerve transection, but another component of the receptor complex, c-Ret, is not.224 The mRNA for epidermal growth factor receptor, which is also the receptor for TGF-␣, is upregulated in rat DRG neurons within 24 hours of peripheral nerve injury and remains so for at least 14 days, while TGF-␣ is upregulated in surrounding satellite cells.511

Ion Channels Pathologic alterations to voltage- and ligand-gated ion channels in sensory neurons could contribute to the ectopic spontaneous discharge associated with neuropathic pain. In other neuronal populations their expression could also significantly alter the normal function of affected neurons, or make them more or less vulnerable to excitotoxic neuron death. The expression of mRNAs for subunits of voltage-gated calcium channels are downregulated in adult rat DRG neurons 7 days after axotomy or chronic constriction injury of the sciatic nerve.236 Immunoreactivity for the ␣ and ␤1 and ␤2 subunits of voltage-gated sodium channels in human DRG neurons is also decreased after DRG avulsion.77 Immunoreactivity for SK1 and IK1 (calcium-activated potassium ion channels) was decreased in avulsed human DRGs

Pathology of Peripheral Neuron Cell Bodies

from 8 hours to 12 months after surgery, and was influenced by NGF levels in cultures of neonatal rat DRG neurons.33 The P2X3 receptor is a ligand-gated cation channel expressed by the nonpeptidergic subset of small-diameter DRG neurons169; it is activated by the binding of extracellular adenosine 5⬘-triphosphate and may therefore be involved in inflammatory pain. After transection of the tibial and common peroneal nerves or the infraorbital nerve, P2X3 mRNA is downregulated in associated rat DRG or trigeminal ganglion neurons, but upregulated in adjacent uninjured neurons, 3 days after injury.472 The number of neurons expressing P2X3 immunoreactivity in human DRGs has been reported to decrease following DRG avulsion.520 The expression of P2X3 channel and acid-sensing ion channel-3 (ASIC-3) immunoreactivity seems to be upregulated in enteric neurons in inflammatory bowel disease,519,521 and immunoreactivity for ASIC-1, -2 and -3 decreased in human DRG neurons following nerve avulsion.521 Motor neurons also downregulate ion channel expression after axotomy. mRNAs encoding sodium channel types I, II, and III are all decreased,208 as are glutamate-activated N-methyl-D-aspartate (NMDA) receptors.350 This last observation could conceivably alter the degree of excitotoxic neuron death in motor neurons, for example, after ventral root avulsion (see below).

Summary In conclusion, the biochemical changes described following peripheral axotomy largely mirror chromatolysis in the time course and in the factors that govern its duration and severity. Generally, axotomy-induced biochemical changes reflect the increased synthesis of macromolecules involved in regeneration and neuroprotection, and downregulation of molecules associated with neurotransmission. To the investigator, biochemical changes may be more sensitive indicators of injury than morphologic ones, and could give clues to the nature of the injury when the underlying controlling mechanisms are known.

NEURON DEATH One possible outcome of injury to peripheral neurons is the death of all or some of the affected neurons. Neuron death is now thought to encompass several distinct morphologic appearances that reflect different biochemical processes (e.g., apoptotic, autophagic, excitotoxic, cytoplasmic, necrotic) that could be regarded as representing a spectrum of neuron death mechanisms. Necrosis would be at one end, representing death as a direct consequence of very acute and severe insults, and apoptosis at the other, representing the most controlled and deliberate form of neuron destruction that is largely a secondary effect of an insult.64,345,353 The factors that determine which mechanism

695

will operate in a particular neuronal population after insult are not well understood; readers interested in a more detailed analysis of neuronal death are referred to some excellent monographs and reviews.64,241,345

Mechanisms of Neuron Death Axotomy-Induced Neuron Loss Death of neurons after axotomy has often been inferred from evidence of neuronophagia, or from neuronal counts that demonstrate a loss when compared to control groups of neurons.4,50,107,363,440 Neuronophagia—clumps of presumed cells with phagocytic activity in ganglia (nodules of Nageotte) or the anterior horn of the spinal cord—results from the death of a neuron at that site and the subsequent infiltration of microglial cells or macrophages (Fig. 31–4). Until recently, the methods used for estimating neuron number involved potentially biased profile-based counting methods.66 Without observations of degenerating neurons, neuronal loss inferred from neuron counting can produce misleading results, particularly if obtained from single dorsal root or sympathetic ganglia, because of the large variation within (between left and right) and between animals. Sensory neuron loss following unilateral axotomy has been reported in a variety of species, including humans,3,14,16,44,50,163,290,363,366,440 with most authors finding a selective loss of small-diameter DRG neurons.3,50,165,363,367,452 Motor neurons are vulnerable to axotomy-induced death in immature animals226 but appear to be more resistant in adults.261,357 Sympathetic neurons also seem to be lost in adult rats after axotomy.181 Apoptosis in the Peripheral Nervous System The discovery of apoptosis as a particular type of death with a defined physiologic role,234 and its rediscovery during the hunt for proto-oncogenes, stimulated interest in the field of “thanatology.” Apoptosis was thought to require the de novo synthesis of macromolecules278,510 as part of a controlled destruction mechanism that does not trigger an inflammatory response. It is now recognized that other forms of cell death involve macromolecular synthesis, and that apoptosis can occur in the absence of a cell nucleus,64 so many of the original assumptions are being challenged, particularly in the context of the nervous system. Apoptosis is recognized as a vital process in the regulation of cell number during the development of the peripheral nervous system.333 Indeed, descriptions of naturally occurring cell death during development frequently appear in the historical literature, starting with Vogt in 1842.63 The term often used to describe the morphology of this “natural” death was pyknosis, which describes the condensation and/or fragmentation of the neuron and its nucleus into darkly staining, irregularly shaped bodies. Pyknotic, degenerating, or necrotic neurons are mentioned in some studies on DRG neuron death following axotomy3,50,63,363 and in mouse

696

Pathology of the Peripheral Nervous System

FIGURE 31–4 A, Neuronophagia of motor neurons (arrows) from an autopsy of a patient with poliomyelitis. The spaces formerly occupied by motor neurons are filled with phagocytic cells removing cellular debris. (Hematoxylin and eosin stain.) Bar: 10 ␮m. B, Hematoxylin and eosin–stained human lumbar DRG from an autopsy of a patient with a paraneoplastic sensory neuropathy. Neuronophagia is often seen in aged or neuropathic human DRGs, where the clusters of phagocytic cells are known as nodules of Nageotte (arrows). Bar: 20 ␮m. C, Nodules of Nageotte are seldom seen in rat DRGs unless an insult is very acute and severe, such as exposure to neurotoxic agents. The nodules of Nageotte (arrows) indicated here were present in a lumbar DRG of an adult rat exposed to large doses of triethyl tin. (Toluidine blue–stained 1-␮m resin section.) Bar: 10 ␮m. (C courtesy of J. M. Jacobs.)

models of hereditary neuropathy386 from the time period before apoptosis was characterized in depth. Apoptosis was originally, and in many respects still is, a term that described a specific morphology. Nuclear DNA condenses into “caps” or nucleosomes (spherical bodies of highly condensed DNA that stains intensely with DNA stains; Fig. 31–5), the cell and nuclear volumes decrease, and the plasma membrane frequently shows “blebbing” (small protrusions of cytoplasm510). Ultrastructurally (Fig. 31–6), the cell cytoplasm becomes electron dense and filled with numerous osmiophilic bodies and fluid-filled vesicles that are probably the remains of organelles and membranes, although mitochondria are initially preserved. The nuclear membrane also becomes involuted and indistinct while the nuclear material condenses into spherical bodies. The ultimate fate of cells undergoing apoptosis is thought to involve them being phagocytosed, and the time taken for the whole process is very short, around 1 to 3 hours,64,509 leaving no obvious trace. The exposure of extracellular phosphatidylserine residues on lipids in the plasma membrane stimulates macrophages to engulf the cell,113 which in DRGs could lead to the formation of a nodule of Nageotte. The time between cultured sympathetic neurons having a normal morphology

and being completely unrecognizable as neurons following apoptosis induced by withdrawal of NGF was found to be 2 to 3 hours.102 This short time span means that a small observed incidence of neuronal apoptosis, maintained for a period of time, would lead to a significant loss of neurons. One of the first detailed descriptions of the morphology and ultrastructure of degenerating adult peripheral neurons (which today would be called apoptotic death) was of dying trigeminal ganglion neurons after transection of the infraorbital nerve.3 These authors also reported that degenerating neurons were not seen after 30 days posttrauma and that none were seen in the control, contralateral ganglia. They found that the number of degenerating neurons was small compared to the estimated neuronal loss, and speculated that this may be due to a rapid removal process. The features of the degenerating neurons described in this work are very similar to the features of apoptotic adult rat162,163,290 or mouse105 DRG neurons after peripheral axotomy. At least some of these apoptotic DRG neurons undergoing axotomy-induced apoptosis can be stained with the in situ end-labeling (ISEL)163 or TUNEL105,290 techniques (see Fig. 31–5) (see below). The numbers of apoptotic neurons seen in vivo in these reports were low, starting to

Pathology of Peripheral Neuron Cell Bodies

697

FIGURE 31–5 A, This photomicrograph shows an adult rat lumbar DRG neuron that is probably in the early stages of apoptosis 2 weeks after transection of the sciatic nerve. The nucleus and neuron have an irregular shape, and the nucleus (arrow) contains numerous small and darkly staining bodies. (Cresyl fast violet–stained paraffin section.) Bar: 10 ␮m. B and C, Cresyl fast violet–stained paraffin sections of ipsilateral lumbar DRG, 1 month after sciatic transection. The small clumps of DNA probably coalesce into larger bodies while the perikaryon starts to be broken down. At this stage the neurons undergoing apoptosis have an irregular shape, darkly stained cytoplasm, indistinct nuclear outline, and several spherical bodies of extremely condensed DNA. Bar: 10 ␮m. D, The end result of the apoptotic process is a body with no discernible structure (a “ghost cell”), apart from remains of condensed DNA (arrow). The size of the structure and the presence of an associated satellite cell (arrowhead) indicate that it is neuronal in origin. (Cresyl fast violet–stained paraffin section.) Bar: 10 ␮m.

appear from 1 day after adult rat sciatic nerve transection and ligation at mid-thigh level.290 The peak incidence occurred at 2 weeks; interestingly, apoptotic DRG neurons were still observed 6 months later.163,290 This axotomyinduced DRG neuron loss (relative to contralateral, uninjured DRGs) can be reduced by the administration of NT-3 and NGF,162,268,337,366 and NT-3 administration to the proximal stump also reduced the incidence of apoptotic DRG neurons.162 Following axotomy of a peripheral nerve, the initial indication that a DRG neuron is going to undergo apoptosis following chromatolytic changes seems to be the formation

of small “droplets” of condensed DNA in the nucleus, which also starts to develop an indistinct nuclear membrane (see Fig. 31–5). These DNA droplets seem to become bigger and the nucleus more irregularly shaped and indistinct, while the progressive loss of basophilia in the center of the perikaryon becomes more pronounced. The Nissl substance around the periphery of the perikaryon stains more intensely, and the shape of the neuron becomes irregular and shrunken. The neuron seems to become physically harder in paraffin-embedded material, which is probably due to the cross-linking of proteins and other constituents by transglutaminase. Finally, the nucleus becomes

698

Pathology of the Peripheral Nervous System

FIGURE 31–6 A, Apoptotic DRG neurons can be identified in resin-embedded material using similar criteria as for paraffin sections, although the remains of the nucleus and DNA are more difficult to identify. The condensed DNA (arrows) and darker staining cytoplasm distinguish them from nonapoptotic neurons. (Toluidine blue–stained 1-␮m resin section.) Bar: 10 ␮m. B, Ultrastructurally, the cytoplasm of apoptotic neurons is more electron dense than that of other nonapoptotic cells and neurons, and contains numerous vesicles, lipid bodies, and lamellar structures. The distorted nucleus (N) is located at the extreme periphery and contains condensed masses of DNA (arrows). The entire structure is surrounded by an associated satellite cell (S), but no paraphyte processes interdigitate with the neuron. (Uranyl acetate and lead citrate staining.) Bar: 1 ␮m. C, At a higher magnification it is possible to see that numerous intact mitochondria are present (arrows), thought to be a hallmark of apoptosis. The nuclear membrane is visible in places (arrowheads). Bar: 1 ␮m. D, Electron micrograph of a degenerating mouse spinal motor neuron, 8 days after injection of serum from a patient with an amyotrophic lateral sclerosis–like syndrome. The key features are dilatation of endoplasmic reticulum and Golgi apparatus (arrows), “blown” mitochondria (arrowheads), and a nucleus that appears near normal. This appearance suggests that apoptosis is not occurring and that some disruption to membrane integrity/ionic homeostasis has happened. Bar: 10 ␮m. (D courtesy of A. Pullen.)

unidentifiable, all basophilia has disappeared, and what remains is a seemingly empty sack containing the condensed remnants of neuronal DNA, a so-called ghost cell.102 Adult rat or mouse spinal motor neurons do not appear to die following transection of a peripheral nerve, whereas neonatal motor neurons seem to undergo a type of death that has similarities to apoptosis and necrosis. This involves DNA fragmentation261,277,372 that is mediated by NMDA receptor–linked ion channels.158,255,372 However, spinal

motor neuron death following ventral root avulsion in neonatal or adult mice does not appear to involve DNA fragmentation, and the process is more reminiscent of necrosis.261 Sciatic nerve avulsion causes adult rat spinal motor neurons to die via a process that has the morphologic features of apoptosis283 and that involves DNA strand breaks.267 These results suggest that there is a hierarchy of nerve injuries that determines the mechanism of motor neuron death in neonates and in adults, with ventral root

Pathology of Peripheral Neuron Cell Bodies

avulsion being the most severe and acute form of injury and peripheral nerve transection the least severe. This is probably due to significant trophic support being supplied by surrounding glia and neurons in adult spinal cord332 that is disrupted by root avulsion. Reperfusion following spinal cord ischemia has also been reported to produce apoptosis of rabbit spinal motor neurons, as determined by TUNEL staining and upregulation of caspase enzymes.51,179 Apoptosis in sympathetic neurons has been studied in vitro, where it is triggered by removal of NGF from the medium containing the cultured neurons within 24 to 48 hours.101,102,123,278 Its occurrence in vivo has not been reported. Necrosis True necrosis is thought to occur when cells are overwhelmed by an insult without having time to initiate a controlled destruction, leading to cessation of synthetic functions, loss of cell volume regulation, influx of water, and cell lysis.275 It is often reported as occurring in the central nervous system (CNS) following acute ischemic injury, with cell necrosis in the center of infarcts, which are often surrounded by a penumbra of damaged cells that have had time to initiate apoptosis. There are few, if any, images of neurons undergoing necrosis in vivo, presumably because it happens too rapidly. Reports of unequivocal necrosis (and not any other form of neuron death) in peripheral neurons are rare and should be treated with caution; acute injuries that produce this pattern of injury are not commonly encountered or studied in peripheral neurons. Frank necrosis of rabbit cranial and DRG sensory neurons has been described following acute intoxication with methylmercury,213 a toxin that is thought to inactivate membrane ion transport. In this report, neurons undergoing morphologic changes typical of necrosis can be seen: an electron-lucent cytoplasm, dilatation of Golgi apparatus and endoplasmic reticulum, and mitochondrial swelling and bursting while nuclear morphology remains almost normal.64 Intraneural injection of the cytotoxic lectins ricin, abrin, modeccin, or volkensin has been reported to produce necrotic rat sensory neurons and spinal motor neurons,499,500 although the presence of irregular, shrunken, dark nuclei in DRGs suggests that the process may not be true necrosis (if indeed true necrosis exists as a distinct entity). Herpes zoster infection can produce neuron death via a process that appears similar to necrosis, because a massive inflammatory cell infiltrate was involved, no structures resembling apoptotic (or pyknotic) bodies were reported despite a massive loss of neurons in the affected DRGs, and some degenerating neurons showed evidence of vacuolation.90 It is likely that the form of neuron death most closely resembling “true” necrosis would be produced by complement activation; this results in the formation of the membrane attack complex that forms a pore in the membrane to which the immunoglobulin/complement has bound.

699

The resultant influx of water and ions would, if enough pores were present in the membrane, cause neuron lysis. Excitotoxic Neuron Death Excitotoxic, or cytoplasmic, neuron death64,100,261 incorporates features of both apoptosis (nuclear DNA condensation, electron-dense cytoplasm) and “necrosis” (dilatation of endoplasmic reticulum, Golgi cisternae, and mitochondria). Its role has been investigated in research on some types of neurodegenerative disease, and experimentally in neurons using NMDA, kainic acid, and other analogues of glutamate. These agonists activate glutamate-activated ion channels, some types of which (e.g., the AMPA glutamate receptor subunit GluR2) are expressed in motor neurons. This may make them vulnerable to excitotoxic insults through the overstimulation of these receptors/ion channels and the subsequent disruption of intracellular ionic homeostasis.47 The activation and opening of ion channels would be a key event in excitotoxic mechanisms, allowing an influx of calcium and other ions that overwhelm the normal calcium buffering mechanisms, leading to the activation of calcium-activated degradative enzymes in the cytoplasm.57 It is therefore likely that this would be associated with a particular progression of morphologic changes, recognition of which could assist in the elucidation of pathologic mechanisms. Such a mechanism has been suggested to underlie some forms of motor neuron disease,478 although the process by which motor neurons die in motor neuron disease remains unclear to date.285,410 It is likely that the more severe the injury to motor neurons (ventral root avulsion ⬎ sciatic avulsion ⬎ ventral root transection ⬎ nerve transection), the more rapid and “cytoplasmic” the pattern of neuron death becomes. In sensory neurons, lidocaine has been reported to be neurotoxic via an excitotoxic mechanism involving membrane depolarization and calcium influx and release from intracellular stores151; it is interesting to compare this finding with reports of motor neuron pathology following intrathecal administration of local anesthetics.329 In a process very similar to excitotoxic cell death, laser ablation of spiral ganglion neurons423 produced features of both apoptosis (nuclear condensation) and necrosis (cell edema and “vacuolization”). An excitotoxic mechanism is suspected to underlie the sensory neurotoxicity of capsaicin and resiniferatoxin177,180,191: These molecules bind to and activate vanilloid receptor type 1, a ligand-gated divalent ion channel. Other Types of Neuron Death Reports of other forms of motor neuron death are beginning to appear in the literature. Motor neurons from transgenic mice overexpressing superoxide dismutase appeared to die via a mechanism that is not apoptosis, and instead involved mitochondrial swelling.21 Some axotomized facial motor neurons have recently been found to accumulate ␣-synuclein, and then to die slowly via a nonapoptotic process.305

700

Pathology of the Peripheral Nervous System

Identified Biochemical Events in Neuron Death One characteristic of apoptosis is activation of a calciumactivated endonuclease that digests DNA into 200–base pair fragments; these fragments form a characteristic “ladder” pattern in DNA gels.509 This degraded DNA can be visualized in situ using histochemical techniques that detect DNA strand breaks (TUNEL137) or single-stranded DNA produced by calcium-activated endonuclease (ISEL).498 However, both techniques stain both necrotic and apoptotic (i.e., dying) cells, and the point at which DNA degradation becomes detectable using ISEL or TUNEL is not known, nor is the precise point at which the “point of no return”

occurs. More recent developments have utilized the detection of specific components of the apoptotic pathway, such as activated caspases, transglutaminase, or poly(ADPribose) polymerase (Fig. 31–7). Caution should be exercised in using these techniques because it is still not clear how many other apoptotic pathways exist that do not involve these molecules. In our view, the actual type of cell/neuron death can only be characterized with any confidence by using morphologic criteria (particularly electron microscopy) on well-fixed (i.e., perfused) tissues. It is now recognized that apoptosis can be triggered by a variety of factors, among them DNA damage,284,506 loss of contact with targets and target-derived trophic factors,282 and defects in the ubiquitin degradative pathway.302 This

FIGURE 31–7 A and B, Photomicrographs of dying adult rat DRG neurons (arrows) stained with the ISEL technique, 1 month after sciatic transection. Histochemical staining methods for damaged DNA, such as TUNEL and ISEL, are not specific for apoptosis but can assist in identifying apoptotic neurons or cells in tissue sections by revealing degraded DNA. Bar: 10 ␮m. C and D, Immunohistochemical methods for apoptosis-specific proteins may be a more reliable way of identifying apoptotic cells and neurons. C, Photomicrograph of an adult rat DRG neuron, immunostained for activated caspase 3: the remains of the nucleus are just visible (arrows). Paraffin section of ipsilateral DRG, 2 weeks after sciatic nerve transection. Bar: 10 ␮m. D, The normally occurring apoptotic neuron death (arrows) in embryonic DRG can also be identified using activated caspase 3 immunohistochemistry. Paraffin section of embryonic day 14/15 rat embryo DRG. Bar: 10 ␮m.

Pathology of Peripheral Neuron Cell Bodies

initial “initiator” stage249 is thought to be followed by the activation of at least two signaling cascades (the effector stage), one involving the activation of the caspase family of proteases, which are thought to be specific for apoptosis, and at least one other caspase-independent pathway.302 Caspases 3, 6, and 7 are thought to be common to most apoptotic pathways, and immunohistochemistry for activated caspases has been used to detect apoptosis in situ in mouse models of motor neuron disease,166 in injuryinduced motor neuron changes,87,284,479 and in sensory neurons after NGF withdrawal.314 The caspase pathway can be activated by proteins released from damaged or permeabilized mitochondria.344 This might explain the combined features of apoptosis and necrosis seen in excitotoxic or “cytoplasmic” neuron death, in which mitochondria may be damaged by loss of ionic homeostasis. The final phase of apoptosis is the degradation phase, which involves the destruction of cellular organelles, cross-linking of cytoplasmic constituents, and DNA destruction. One enzyme involved at this stage is apoptosis-related transglutaminase type 2, thought to be involved in the cross-linking of cytoplasmic constituents; this has been detected in degenerating motor neurons in the Mnd transgenic mouse model of motor neuron disease.196 Apoptosis appears to be regulated by a balance between the expression of pro- and antiapoptotic signals. Differences in the expression of pro- (such as Bax) and antiapoptotic (such as Bcl-2 and Bcl-XL) proteins have been proposed as a basis of the selective vulnerability of sensory and motor neuron populations to axotomy-induced apoptosis.144 The role of p75NTR in this equilibrium is intriguing: It appears to be both pro- and antiapoptotic depending upon the type of neuron and the stage of development of the animal.17,36,45,56,124 Moreover, Bcl-2 (normally antiapoptotic) seems to promote, and be required for, the p75-mediated apoptotic death of cultured sensory neurons, while Bcl-XL (normally proapoptotic) seems to inhibit this.73 In axotomized facial motor neurons, the downregulation of Bcl-2 was greater the closer the site of axotomy was to the facial nucleus, which may be the explanation for proximal stump length–dependent survival in axotomized neurons.490,518 Excitotoxic neuron death mediated by NMDA receptors has been reported to involve the production of nitric oxide and oxygen radicals167,267 that can kill cells or neurons by reacting with, and inactivating, cellular macromolecules. Nitrotyrosine residues in proteins are indicative of the presence of nitric oxide and superoxide radicals and can be detected using immunohistochemistry. Such a mechanism has been proposed to underlie certain transgenic models of motor neuron disease,257 and nitrotyrosine residues have been detected in cases of motor neuron disease.382,474 A more important action of nitric oxide in chronic disease might be the inhibition of neuronal energy production.38 Disruption of calcium homeostasis certainly appears to produce many of the features of excitotoxic death in motor

701

neurons.2 Overexpression of the calcium-buffering protein parvalbumin in transgenic mice475 can protect motor neurons from this form of neuron death. Moreover, administration of certain calcium channel blockers can reduce facial motor neuron apoptosis and loss following transection of the facial nerve in neonatal rats.462 Theoretically, staining for soluble Ca2⫹ ions in histologic sections would identify apoptotic and excitotoxic neuron death in tissue sections, because calcium ions appear to be crucial in both proposed mechanisms. This is technically difficult because of postmortem dispersion of soluble ions. Complement activation by the Fc portions of bound antibodies leads to the formation of the membrane attack complex. This is essentially a pore or channel composed of various activated complement proteins that is inserted into the cell membrane of the target cell or neuron, leading to the lysis (as a result of loss of ionic homeostasis) of the cell/neuron being attacked. Consequently, in tissue sections this will appear similar to necrotic or excitotoxic death in morphology but in theory could be distinguished by the presence of activated complement proteins such as C3d or C9neo. C3d bound to the axolemma of motor axons has been reported in acute motor axonal neuropathy,171 which may result in the chromatolytic motor neurons reported in acute motor and sensory axonal neuropathy.161 Complement activation has also been described in the death of hypoglossal motor neurons produced by intraneural injection of ricin, in which complement C9 and the complement inhibitor clusterin immunoreactivities were found in neuronal perikarya, and C3d found in surrounding glial cells.465 Why complement should be activated by neuronal death is not clear. The stimulus for the macrophage and T-lymphocyte invasion of adult rat DRGs after transection of the sciatic nerve is also not clear202; the question of whether they are essentially protective or deleterious toward neurons is an interesting one. The mechanism of neuron death in paraneoplastic neuropathies might be expected to involve complement activation, but so far this has not been reported; neuron loss in anti-Hu–associated paraneoplastic neuropathy and encephalomyelitis did not apparently involve apoptosis or complement activation.26

Factors Affecting the Magnitude of Neuronal Death There seems little doubt that axotomy in the neonate produces a more rapid and much greater incidence of sensory and motor neuron death than in the adult.5,16,34,189,199,263,368,394,517 This may be due to the greater dependence of neonatal neurons upon target-derived trophic factors, or to a greater vulnerability of neurons with growing axons than of those without.495 There also appear to be qualitative differences in that condensed DNA caps, as well as the spherical nucleosomes seen in adults, were observed in neonatal apoptotic neurons16; these caps have not been reported or observed in adult rats.

702

Pathology of the Peripheral Nervous System

The distance between DRGs and the site of peripheral axotomy in adult rats directly affects the rate and magnitude of neuron death.413,518 Transection with ligation of the proximal and distal stumps of the adult rat spinal nerve will produce a loss of around 35% of L5 DRG neurons by 45 days later.484 An identical lesion at mid-thigh level produced a neuron loss of around 14% at 8 weeks and 37% at 32 weeks.452 The type of injury may also determine the extent of neuron loss: Transection, ligation, and “capping” of both distal and proximal stumps of the sciatic nerve at mid-thigh level appeared to produce an eventual neuron loss of around 35% in injured DRGs,290,452 whereas transection of the sciatic nerve and ligation of the proximal stump at mid-thigh level produced a loss of only 7% to 14%.163 This smaller loss could be due to neurotrophic factors produced in the distal stump being able to diffuse out of the cut end of the stump. The apparent differences in susceptibility to axotomyinduced neuron death between populations of adult peripheral neurons, particularly between motor and sensory/sympathetic neurons, may ultimately reflect biochemical differences between placode-derived and neural crest–derived neuron populations and their respective responses to cytokines and neurotrophic factors. However, it is also possible that susceptibility to death is purely related to neuron size. This question will not be answered until the mechanisms causing neuron death are established.

Summary This is a rapidly developing field, but it is already clear that the morphology and biochemistry of dying neurons can give valuable insights into the underlying mechanisms responsible. Mechanisms are likely to be different in adult and neonatal animals, but whether axotomy-induced neuron death is triggered by a loss of axoplasm, or something released from the injury site that is neurotoxic, or a lack of a survival factor transported from the periphery is still not known.

Pathologic Alterations in the Cell Bodies of Peripheral Neurons in Age and Disease AGING

Table 31–1. Incidence of Nodules of Nageotte with Age in Human Dorsal Root Ganglia Age

Sex

7 mo 13 yr 17 yr 17 yr 17 yr 35 yr 52 yr 53 yr 60 yr 69 yr 75 yr 80 yr 91 yr

F F F M M M F F F F M F F

Nageotte Nodules (%) 0 ⬍0.1 1.0 0.17 0.87 1.5 0.92 0.35 2.2 0.78 3.5 0.8 1.5

From Scaravilli, F.: Changes in neuronal structure and cell populations with ageing. In Thomas, P. K. (ed.): Peripheral Nerve Changes in the Elderly: New Issues in Neurosciences. Amsterdam, Wiley, p. 95, 1988, with permission.

of nodules have been observed in human DRGs after amputation of the upper arm.440 Age-related changes in autonomic ganglia are more obvious in humans than in other mammals, and in the superior cervical ganglion more than elsewhere.131 These agerelated changes mainly affect the cell processes. Enlarged argyrophilic neurites containing highly phosphorylated neurofilaments were observed in human paravertebral sympathetic neurons399; swollen axon terminals containing aggregates of filaments have also been reported.397,398,401 Sensory neurons in rats352 and humans108 have been reported to increase in number with age, whereas motor neurons appear to decrease in number but increase in size in rats.211 Enteric neurons have been reported to decline in number with age in experimental animals, based on neuron counts of gut neurons in young and old rats.76 Andrews9 observed that autonomic neurons increase in size postnatally, albeit not in a synchronized way: sympathetic neurons innervating the submandibular gland continue to grow into maturity from early postnatal periods, while those innervating the iris do not. Similarly, neuronal atrophy observed in aging is by no means a general phenomenon; rat DRG neurons appear to increase in size with age (M. J. Groves and F. Scaravilli, personal observations).

General Observations Nodules of Nageotte are frequently seen in DRGs from autopsies of aged patients without symptoms of neuropathy (Table 31–1). An age-related loss of sensory neurons may be behind the loss of unmyelinated axons often seen in sural nerves of aged patients, as well as the loss of deep tendon reflexes often observed in the elderly (M. J. Groves and F. Scaravilli, personal observations). Larger numbers

Lipofuscin Progressive accumulation of lipofuscin is the most obvious change in aging neurons of the peripheral, including the autonomic, nervous system. Lipofuscin is a brown pigment, fluorescent at all wavelengths, that is usually formed by the peroxidation and polymerization of unsaturated membrane lipids following attack by free radicals, and as such is often

703

Pathology of Peripheral Neuron Cell Bodies

thought of as a “wear and tear” pigment. In mice, the incidence of (all types of) neurons containing lipofuscin varies from none at birth to about 90% in 30-month-old animals. Mouse DRG and motor neurons containing lipofuscin increased in number in both types of cells from birth,320 and the increased amount of lipofuscin in motor neurons of aged rats has been suggested to be the reason for their increased size.211 Whereas in young mice this pigment is widespread in the cytoplasm of neurons, in older animals it clusters around the nucleus at one end of the neuron as well as near the axon hillock.320 In humans, the incidence of DRG and motor neurons containing lipofuscin (Table 31–2) had reached the values existing in the elderly by the end of the third decade.384 Increased lipofuscin deposits and lamellar cytoplasmic bodies have been reported in autonomic ganglia of 26- to 28-month-old Wistar rats.429 Together with a change in amount, age also brings about a modification of the physical and histochemical characteristics of lipofuscin.320 The relationship between lipofuscin and lysosomes has been studied in motor neurons of the mouse: Primary lysosome-like, autophagic vacuole–like, and mature pigment granules represent early, intermediate, and late stages, respectively.408 Aging is not the only factor involved in the process of accumulation of lipofuscin. Ubiquinated lipofuscin granules have been described in surviving motor neurons in cases of autosomal recessive spastic paraplegia.487 An increased amount of lipofuscin in DRG cells has been described in Tangier disease.396 Stress also seems to produce an increase in neuronal lipofuscin,231 whereas vitamin E298 and the drug centrophenoxine319 seem to have a preventive effect on its accumulation in CNS neurons. The role of vitamin E in protecting membrane lipids against attack by free radicals, and hence against lipofuscin accumulation, is illustrated by a new Table 31–2. Percentage of Human Dorsal Root Ganglion and Motor Neurons Containing Lipofuscin at Various Ages

Age

Sex

DRG Neurons with Lipofuscin (%)

7 mo 13 yr 17 yr 17 yr 27 yr 52 yr 69 yr 80 yr 91 yr

F F M M M F F F F

0 20 14 13 52 76 64 66 61

Motor Neurons with Lipofuscin (%) 0 ⬍0.1 ⬍3 24 64 81 65 50 63

From Scaravilli, F.: Changes in neuronal structure and cell populations with ageing. In Thomas, P. K. (ed.): Peripheral Nerve Changes in the Elderly: New Issues in Neurosciences. Amsterdam, Wiley, p. 95, 1988, with permission.

syndrome522 of vitamin E deficiency. This is produced by a missense mutation of the ␣-tocopherol transfer protein gene that, in addition to retinal atrophy and severe degeneration of the posterior columns of the dying-back type, also produces a massive accumulation of lipofuscin in DRG neurons.

Melanin A melanin-like pigment has been described in some neurons, particularly those of the human trigeminal and spinal ganglia.88 This appears early in life, and seems to precede lipofuscin and to increase in incidence with age. Scaravilli observed that staining methods for melanin gave positive results in the regions of the cytoplasm of sensory and autonomic ganglion (but not motor) cells occupied by lipofuscin.384 Ultrastructural features of melanin also are similar to those of lipofuscin19; moreover, the finding that neuromelanin in the pigmented nuclei of the brainstem resembles mature lipofuscin190 indicates that, in old sensory ganglia, lipofuscin bodies may also become melanized.265 The observation that the pigment is histochemically similar to melanin, and is present in sensory and autonomic nerve cells but not in motor neurons, may be explained by melanocytes, autonomic neurons, and sensory neurons having a common origin during development (the neural crest).

Neurofibrillary Tangles Neurofibrillary tangles (NFTs) are most often associated with the neuropathology of CNS neurons affected by Alzheimer’s disease and related neurodegenerative disorders, but have also been observed in motor neurons of patients with Alzheimer’s disease,378,404 with Parkinson dementia complex and amyotrophic lateral sclerosis from Guam,338,397,473 and with progressive supranuclear palsy (PSP) (Fig. 31–8A).227,235,486 NFTs were also observed in the nucleus of Onufrowicz of patients with PSP.390 Transgenic mice expressing the G272V mutant form of tau develop NFTs in their motor neurons.153 NFTs have been observed in dorsal root ganglion neurons in old Wistar rats,476 in the upper cervical ganglia of a 76-year-old man,229 and in cervical DRG neurons of patients with PSP.323

HEREDITARY NEUROPATHIES Hereditary Motor and Sensory Neuropathy (Charcot-Marie-Tooth Disease) The mutations that cause hereditary motor and sensory neuropathy (HMSN) types I and III (Déjérine-Sottas disease) are located within the genes encoding the myelin compaction proteins PMP22 and P0 expressed by Schwann cells,98 and

704

Pathology of the Peripheral Nervous System

FIGURE 31–8 A, Paraffin section of DRG from an autopsy of a patient with SCA7, immunostained for polyglutamine with the 1C2 antibody. Despite postmortem artifact (severe shrinkage of neuronal perikarya), the immunoreactive nuclear inclusion is clearly visible (arrow). A neighboring nodule of Nageotte (arrowhead) indicates previous loss of a neuron. Bar: 10 ␮m. (Courtesy of O. Ansorge.) B, Paraffin section of lumbar DRG from an autopsy of a patient with familial amyloidosis, stained with hematoxylin and eosin. The visible neurons (arrows) show features of chromatolysis (eccentric or displaced nuclei), and large deposits of amyloid are present throughout the DRG (*). Bar: 10 ␮m. C, Same case and DRG as B stained with Congo red and photographed under polarized light. Amyloid deposits surrounding a lone DRG neuron show up with a green birefringence in this preparation (arrows). Bar: 10 ␮m.

result in hypomyelinating/demyelinating neuropathies. There are few reports of pathologic alterations to neuronal perikarya in these diseases, although a loss of lumbar motor neurons has been reported in patients with HMSN I.449 Studies on the “trembler” mouse, a spontaneous mutant that has mutations in the P0 gene and is the most studied mouse model of HMSN I and III, failed to find any loss of sensory or motor neurons compared with normal mice.269 HMSN II appears to be a more heterogeneous genetic disease than HMSN I and III, with several loci linked to this neuropathy.99 Charcot-Marie-Tooth disease type X is an axonal neuropathy linked to mutations in the connexin 32 gene located on the X chromosome. The few published reports of autopsies of patients with HMSN II found degeneration and loss of DRG and motor neurons, as well as chromatolysis and atrophy of surviving motor neurons,23,524 and loss of large DRG and motor neurons in HMSN associated with deafness, mental retardation, and epilepsy.310

Hereditary Sensory and Autonomic Neuropathy Types I through V The genetic defects in hereditary sensory and autonomic neuropathy (HSAN) types I, II, and III have not been

identified at the time of this writing, and at least three loci have been found in genetic linkage studies of families with the HSAN I phenotype.99 An autopsy of a patient with a slowly progressive ataxic neuropathy of adult onset (which would probably be classified as HSAN I) revealed a slight decrease in the number of large DRG neurons and nodules of Nageotte in the cervical and lumbar DRGs, with no evidence of motor neuron loss.418 HSAN types IV and V (also known as congenital insensitivity to pain with anhidrosis) have been found to be the result of mutations in the TrkA NGF receptor.200,207,466 In transgenic TrkA knockout mice, the lack of functional TrkA leads to the death of non–TrkA-expressing smalldiameter DRG neurons and sympathetic neurons during embryonic development.415

Spinal Muscular Atrophy Spinal muscular atrophy (SMA) is a hereditary neurodegenerative disease affecting motor neurons mainly caused by homozygous deletions or mutations in the SMN gene, mapped to chromosome 5q12.2-q13.82,395 SMA spans a spectrum, from the severe SMA I (Werdnig-Hoffmann disease) to the mild SMA III (Kugelberg-Welander disease),

705

Pathology of Peripheral Neuron Cell Bodies

and pathologic findings from autopsies of 3- to 9-month-old infants with SMA I included chromatolysis and aberrant glycosylation/phosphorylation in spinal motor neurons.61 This appears to lead to the death of motor neurons by an apoptotic pathway, as revealed by TUNEL staining and ultrastructural examination.417

Hirschsprung’s Disease Also known as congenital megacolon, Hirschsprung’s disease is caused by a congenital absence of neural crest–derived enteric neurons in the myenteric plexus.32 The disease has a complex inheritance and is frequently associated with other congenital abnormalities. The abnormality can occur as a chromosome 2-to-13 translocation488 or as a deletion on chromosome 2.273,309

Giant Axonal Neuropathy Giant axonal neuropathy (GAN) is caused by mutations in the gigaxonin gene, localized to chromosome 16q24,122 and leads to accumulations of neurofilaments in axons that can be extremely large. The resultant neuropathy is often severe, leading to the death of affected individuals in the first or second decade. An autopsy of a 25-year-old man with GAN found accumulations of neurofilaments in spinal motor neurons that were slightly distended with central chromatolysis; coarsely clumped or a “dustlike” distribution of chromatin was also observed in motor neurons.343 Myenteric neurons in a rectal biopsy from an 8-year-old girl with GAN were reported to show neurofilament accumulation.133

Neuroaxonal Dystrophy Neuroaxonal dystrophy is characterized by swollen axons (smaller than those seen in GAN) and enlarged initial segments of axons or perikaryal projections in sensory ganglia, which may distort otherwise normal DRG neuron perikarya. Neuroaxonal dystrophy and GAN can be thought of as representing the two poles of a spectrum of neurofilamentrelated abnormalities, but whether this is due to mutations in the same genes is not known. Swellings similar to those seen in neuroaxonal dystrophy can develop as a function of aging and diabetes, and are composed of accumulations of phosphorylated neurofilaments; those seen in aged individuals are identical to those seen in diabetes.400

Defects in DNA Repair (Ataxia-Telangiectasia, Xeroderma Pigmentosum, Cockayne’s Syndrome) There are few reports of the peripheral neuron pathology of these diseases, which are usually caused by mutations in enzymes involved in DNA repair mechanisms. A progressive loss of DRG neurons was observed in an autopsy of a

49-year-old patient with xeroderma pigmentosum. The authors concluded that the disease is a primary neuronal degeneration that manifests first in the peripheral nervous system.369 NFTs have been observed in CNS neurons in isolated cases of Cockayne’s syndrome.446

Disorders Caused by Trinucleotide Repeat Instability Under this heading are included a number of genetically transmitted disorders previously classified in different groups. The first mutations to be described were X-linked SMA and bulbar muscular atrophy and fragile X syndrome, caused by expansions of trinucleotide repeats.252,482 Originally thought to be translated into polyamino acid tracts, the CGG repeat in fragile X syndrome was later identified as a nontranslated region of the FMR1 gene, showing that there are at least two types of trinucleotide repeat disorders251: ones in which the mutant triplet repeat is located in coding regions and ones in which it is located in noncoding regions of the gene.491 In the diseases in which the repeat is localized within the coding region of the gene involved, the repeat is translated into a stretch of polyglutamine, which may accumulate as neuronal intranuclear inclusions (NIIs). These may be seen with routine staining but can be specifically visualized using antibodies such as antiubiquitin (because the inclusions are always ubiquitinated), the 1C2 antibody that recognizes the expanded polyglutamine domain,469 or antibodies specific to each disorder. Indeed, the existence of NIIs in these diseases has only been recognized relatively recently, because it is difficult to see NIIs with normal histologic staining methods. These diseases can also be classified into four groups based on the nucleotide composition of the repeated triplet243: 1. Long cytosine-guanine-guanine (CGG) repeats in the two fragile X syndrome nucleotides, FRAXA and FRAXE 2. Long cytosine-thymidine-guanine (CTG) repeats in myotonic dystrophy 3. Long guanine-adenine-adenine (GAA) repeats in Friedreich’s ataxia 4. Short cytosine-adenine-guanine (CAG) repeat expansions implicated in at least 11 neurodegenerative disorders Transgenic mouse models of trinucleotide repeat disorders have been created, and NIIs have been observed in spinal motor, cranial sensory, and cranial motor neurons in some cases.18

Friedreich’s Ataxia Friedreich’s ataxia is the most prevalent inherited ataxia, with a frequency of 1 to 2/50,000 live births. It is an autosomal

706

Pathology of the Peripheral Nervous System

recessive disorder with onset in early childhood, although late-onset forms have been described. The disease is caused by the abnormal expansion of the GAA repeat in intron 1 of the FRDA gene on chromosome 9, which encodes frataxin, a protein that has been localized to mitochondria.35 This results in a severe and diffuse loss of large DRG neurons: A report of an autopsy found an increased number of Nageotte nodules with some of the remaining neurons showing chromatolytic changes.204 Motor neurons are not affected.

Spinocerebellar Ataxias Spinocerebellar ataxia (SCA) type 1 is an autosomal dominant disease with onset during midlife and is characterized by motor symptoms. It follows an expansion of the CAG repeat within a gene located in the short arm of chromosome 6; the gene product is called ataxin-1.335 A report of 11 autopsies of patients with SCA1 found an extensive loss of spinal motor neurons, with sparing of the nucleus of Onufrowicz; DRGs appeared to be normal.370 The mutation underlying SCA2 is a trinucleotide repeat (CAG) expansion within the coding sequence of the SCA2 gene, located in chromosome 12q24, which encodes a protein called ataxin-2. Autopsies of patients with SCA2 revealed a loss of motor neurons described either as mild334 or severe,112,339 and a moderate loss of DRG neurons.112 Intranuclear inclusions, described in the brain and brainstem, were not found in cranial ganglia.245 SCA3 (Machado-Joseph disease) is probably the most common dominantly inherited ataxia, and is caused by a CAG/polyglutamine repeat in a gene located to chromosome 14q that encodes a protein called ataxin-3.342 Autopsies of patients with SCA3 have revealed a loss of spinal motor507 and DRG neurons,74 loss of the latter being associated with nodules of Nageotte. NIIs in DRG and paravertebral sympathetic and celiac ganglion neurons have recently been reported in four of four autopsies of patients with the disease.515 These inclusions are similar to those seen in the CNS: dense, spherical bodies inside the nucleus as well as a diffuse, granular material present in the cytoplasm also visualized with the IC2 antibody. The genetic defect in SCA6 is located to chromosome 19p13; the CAG repeat is found within the gene for the human a1A voltage-dependent calcium channel unit that is expressed in Purkinje cells.529 This produces a purely cerebellar form of ataxia that progresses over 20 to 30 years. A loss of motor neurons in the motor nuclei of the medulla and spinal cord has been reported from autopsies of patients with SCA7,280 which is caused by a CAG repeat mapped to chromosome 3p12-13.83 The neuropathology of SCA7 is also characterized by the presence of NIIs195 that have been seen also in DRG neurons (Fig. 31–9A).

X-Linked Spinal and Bulbar Muscular Atrophy (Kennedy’s Disease) Kennedy’s disease is a motor neuronopathy of late onset and was the first disorder related to a CAG repeat expansion.252 The disease was mapped on the region of chromosome X where the gene for androgen receptor is located. Pathologic changes include loss of lower motor neurons in brainstem and spinal cord and the presence of NIIs, which are best visualized by antibodies to ubiquitin or to the abnormal protein. No obvious signs of neuronal degeneration or loss could be found in lumbar DRGs from autopsies of two patients with the disease, although a distal axonopathy was found.424

Neuronal Intranuclear Hyaline Inclusion Disease Neuronal intranuclear hyaline inclusion disease is a rare neurodegenerative disorder that has a heterogeneous clinical presentation and pathologic features and usually affects young patients, with death occurring by the third decade. The mode of inheritance is not clear, with both sporadic and familial cases having been described as well as cases in adults. In addition to nerve cell loss, the salient pathologic finding is the presence of NIIs. Because these inclusions share morphologic similarities with those observed in polyglutamine disorders, the disease is presented in this section. These inclusions are ubiquitin positive, react to a varying degree with the antibodies used to detect inclusions in the group presented above, and are ubiquitous in the nervous system. They are composed of aggregates of 8- to 9-nm diameter straight filaments that are autofluorescent, and have been reported in spinal motor,136,237,296,427,447 DRG,136,237,296,436,447 and autonomic436 neurons.

LYSOSOMAL STORAGE DISORDERS These disorders are caused by defects in single lysosomal enzymes. The genetic mutations that underlie a majority of them have been discovered, and the diseases themselves have been classified by detailed clinical and pathologic descriptions.442 This section summarizes any pathologic changes to peripheral neurons that have been described in these disorders.

Gaucher’s Disease Gaucher’s disease is an autosomal recessive disease usually caused by a deficient activity of glucosylceramidase, the gene for which is located in band q21 of chromosome 1.147 However, a deficiency of sphingolipid activator protein C also causes a variant of Gaucher’s disease. These mutations result in massive accumulation of glucocerebroside in

707

Pathology of Peripheral Neuron Cell Bodies

FIGURE 31–9 A, Paraffin section of postmortem lumbar spinal cord from a patient with progressive supranuclear palsy (PSP). Accumulations of tau protein and neurofibrillary tangles have been described in motor neurons of patients with PSP. These tau-immunoreactive deposits may coalesce into large clumps (arrow), displacing the nucleus of the motor neuron to the periphery. Bar: 10 ␮m. B, Luxol fast blue/cresyl fast violet–stained paraffin section of the ventral horn of lumbar spinal cord from an autopsy of a patient with chronic lymphocytic leukemia who developed a GuillainBarré–like syndrome. This motor neuron shows features of chromatolysis, with displacement of the nucleus (arrow) and loss of central basophilia. Bar: 10 ␮m. C, This chromatolytic motor neuron from the same case as B (nucleus indicated by arrow) also contains intracytoplasmic vacuoles (arrowheads). Bar: 10 ␮m.

reticuloendothelial cells, and only the type II and III forms of the disease are neuronopathic. An autopsy of an infant with severe type II Gaucher’s disease found an extensive loss of spinal motor neurons that the authors speculated may involve an apoptotic pathway.119 The sensory and motor neurons of transgenic mice with a targeted disruption of the glucocerebrosidase gene contained tubular Gaucher-type inclusions at the ultrastructural level.501

and have similar neurologic abnormalities and disease progression, showing neuronal accumulation of periodic acid–Schiff reagent (PAS)-positive material from 5 weeks of age in neurons of the brain and spinal cord.173,286 Suramin is an experimental chemotherapeutic agent that seems to produce an experimental GM1 gangliosidosis in cultured DRG neurons, which develop large multilamellar inclusion bodies composed of GM1 ganglioside.142

GM1 Gangliosidosis

GM2 Gangliosidosis

GM1 gangliosidosis is a progressive neurologic disorder caused by a deficiency of lysosomal acid ␤-galactosidase, the gene for which has been located to chromosome 3p21.33.448 There are two childhood forms (infantile type 1 and late infantile/juvenile type 2) and one adult form (type 3). In type 1, accumulation of ganglioside is seen in DRG neurons from 24 week-old human embryos28; subsequently the peripheral neuron pathology tends to be overshadowed by the CNS involvement. Neuronal involvement in the adult-onset form of the disease (type 3) is largely confined to the basal ganglia.441 Transgenic mice that have had the ␤-galactosidase gene deleted have been created

Formerly known as Tay-Sachs disease (a name retained for the infantile form of the disease), GM2 gangliosidosis is the result of a decreased activity of either ␤-hexosaminidase or GM2 activator protein resulting from deletion or mutation of the genes encoding the subunits of the enzyme or activator protein.68 The gene for the ␤-hexosaminidase a subunit is on chromosome 15; the genes for the ␤ subunit and GM2 activator protein are on chromosome 5. Mutations lead to a massive accumulation of GM2 ganglioside and other glycolipids in neuronal lysosomes, which appear as multilamellar structures known as membranous cytoplasmic bodies (MCBs). These have parallel or concentric

708

Pathology of the Peripheral Nervous System

lamellae when examined with the electron microscope,457 and may trigger apoptosis in the neurons affected.203 In the peripheral nervous system, accumulation of gangliosides has been found in enteric neurons in the rectal biopsies from nine adult patients ages 26 to 39 years.502 Transgenic mice with targeted mutations of ␤-hexosaminidase type B (Sandhoff variant) demonstrated extensive MCBs in neurons of the dorsal root and trigeminal ganglia, spinal cord, and myenteric plexus.381

in neurons of the brain and gastrointestinal (GI) tract. However, the stored material of Kufs’ disease is associated with normal lipofuscin, and to differentiate between the two is difficult in routine preparations. According to Lake,253 diagnostic rectal biopsy is not advisable because of the considerable lipofuscin deposition that takes place normally in GI neurons of the adult. Ultrastructurally, inclusions are not uniform and include curvilinear and fingerprint bodies and rectilinear profiles.

Neuronal Ceroid Lipofuscinosis (Batten Disease)

Sphingomyelin Lipidosis (Niemann-Pick Disease)

The term neuronal ceroid lipofuscinosis is incorrect because the stored material is neither ceroid nor lipofuscin but a protein, and also because storage takes place in all tissues and not just the nervous system. The various forms are related to lysosomal enzymes classified as CLN1 through CLN8, with at least six gene loci being implicated in their inheritance.134 In the infantile form (CLN1), the gene for the lysosomal enzyme palmitoyl protein thioesterase, located on chromosome 1p32,215 is mutated. This leads to the storage of PAS-positive material (in paraffin and frozen sections) in enteric neurons. Ultrastructural features are common to cells in all organs and consist of membrane-bound, granular, osmiophilic deposits forming globules 0.5 ␮m in diameter, or arranged in aggregates 3 ␮m in diameter. The late infantile form (CLN2) is caused by a mutation in a gene located to chromosome 11p15 encoding a pepstatin-insensitive lysosomal peptidase301 and is characterized by less severe changes. The enteric neurons of Auerbach’s plexus contain abnormal material arranged as curvilinear bodies that form the characteristic inclusions of this form of neuronal ceroid lipofuscinosis,97 and which are only weakly PAS positive. The normal function of the mutated genes responsible for CLN3, -4, -5, and -8 are unknown, although the genetic loci have largely been established301: juvenile Batten disease (CLN3) is caused by a mutation in a gene located at chromosome 16p12, and late infantile Batten disease (CLN5) is due to a mutation in a gene mapped to chromosome 13p22. It is worth noting that mutations in the CLN8 gene in mice produce the Mnd mouse.134 In this form of the disease, enteric neurons containing stored material similar to that seen in CNS neurons can be observed; this material can be stained with the PAS, Sudan black, and Luxol fast blue techniques, and shows strong autofluorescence. Ultrastructurally, these bodies are different from curvilinear bodies in having a fingerprint-like appearance; neurons may also contain MCBs. The pathologic features of CLN4 and -5 are identical to those of the juvenile form. The adult form of Batten disease (CLN4; Kufs’ disease) is also characterized by an excess of lipofuscin-like material

This group of autosomal recessive disorders is characterized by cholesterol accumulation in various cell types, the most common of which is Niemann-Pick disease type A/B resulting from deficient acid sphingomyelinase activity. The gene for this enzyme has been located to chromosome 11p15.1-p15.4.79 One mutation for Niemann-Pick disease type C1 has been located to chromosome 7p13,84 and another for the major form of types C and D to the centromeric region of chromosome 18q.159 In 1958, Croker and Farber75 proposed a subdivision into four groups (A through D). Subsequently two types, type 1 (groups A and B) and type 2 (C and D), were defined on the basis of the absence or presence of sphingomyelinase, respectively.106 Axonal neuropathy develops in advanced stages of the disease, and is similar to the pathologic alterations to axons in the CNS. “Ballooned” enteric neurons are found in the type 1 diseases, and ultrastructurally neuronal inclusions appear as loosely packed lipid lamellae enclosed within membranebound vacuoles measuring 1 to 2 ␮m in diameter. A transgenic mouse model of type 1 Niemann-Pick disease, produced by deletion of exon 3 of the sphingomyelinase gene, has histopathologic features similar to those of the human disease.250 In the group 2 diseases, ballooned neurons were reported in the anterior horn and in neurons of the myenteric and submucosal plexuses.172 The stored material in neurons from patients with group 2 diseases is weakly sudanophilic and PAS positive in frozen sections. The positive staining with ferric hematoxylin suggests the presence of phospholipids, and electron microscopy shows that the membrane-bound cytoplasmic bodies are polymorphous, are up to 3 ␮ in diameter, and contain loosely packed lamellae that are concentric in places. Mouse transgenic models have been produced by targeted deletion of the genes NPC1 or NPC2.331

Mucopolysaccharidosis The mucopolysaccharidoses form a group of disorders the majority of which are inherited as an autosomal recessive trait. An exception is represented by mucopolysaccharidosis

709

Pathology of Peripheral Neuron Cell Bodies

(MPS) type II (Hunter’s disease), which is an X-linked recessive disorder, although a few cases in females are recorded.308,503 These disorders are characterized by the accumulation of water-soluble mucopolysaccharides that are not retained in fixed material and are due to defects in the various mucopolysaccharide synthesis and storage genes. They have been classified by Lake253 on the basis of criteria that include eponyms, specific enzyme defects, identification of urinary mucopolysaccharides, genetic studies, and genetic/phenotype correlations. In cases of MPS IV, storage is seen in the cerebral hemispheres, including deep gray nuclei,145,244 but not in spinal motor neurons or in ganglia of the autonomic system. Neuronal storage of lipid (or ceroid) and/or membranous cytoplasmic bodies (“zebra bodies”) is present in motor neurons and autonomic ganglia in MPS IH, IH/S, II, III, and VII.178 However, an autopsy of a patient with MPS I found the spinal motor neurons to contain only lipid/ceroid and no lamellated inclusions.216

Mucolipidosis The mucolipidoses are classified into four categories; mucolipidosis (ML) I is now included under the neuraminidase deficiency disorders. ML II and III have the same enzyme defect (N-acetylglucosamine-1-phosphotransferase), the gene for which has been located to chromosome 16p.359 ML IV is caused by mutations in the MCOLN1 gene that encodes for mucolipin, a member of the transient receptor potential gene family,7 located on chromosome 19p.421 Neuronal storage of lipid-like material in spinal ganglia has been reported from an autopsy of a patient with ML II, which is also known as I-cell disease.318 However, autopsies of four patients with ML II found no inclusions in DRG or parasympathetic neurons, but did find basophilic stored material in motor neurons that was seen to be identical to zebra bodies under the electron microscope.279 Neuronal storage of PAS- and Sudan black–positive material in the central454 and peripheral150 nervous systems has been reported in cases of ML IV.

Mannosidosis Mannosidosis is divided into two categories according to whether it is due to a deficiency of either lysosomal ␣- or ␤-mannosidase. These enzymes degrade asparagine-linked carbohydrate cores of glycoproteins, and deficiencies in their activity results in intralysosomal accumulation of mannoserich oligosaccharides. The gene encoding ␣-mannosidase has been mapped to chromosome 19cenq12,322 and that for ␤-mannosidase to chromosome 4q22-25.6 An autopsy of a 3.5-year-old patient with mannosidosis found severely ballooned neurons (distended with large numbers of small vacuoles or vesicles) in the paravertebral

sympathetic, trigeminal, and dorsal root ganglia with occasional evidence of neuronal loss; spinal motor neurons appeared less severely affected. Ultrastructurally, the vacuoles/vesicles were often interdigitated or invaginated with each other, sometimes fusing or coalescing together and often contained small numbers of fine fibrils in stacks as well as lipid globules.439 Certain plant alkaloids, such as swainsonine (from Ipomoea carnea, Swainsonea, Astragalus, and Oxytropis species), inhibit mannosidase and produce similar effects upon cells and neurons in intoxicated animals.86,205 Intoxication produced characteristic accumulation of small vacuoles or vesicles in certain neuron populations in the CNS, and in lumbar DRG neurons with no evidence of loss.205 Only neurons unprotected by a blood-brain barrier were affected.

Fabry’s Disease Fabry’s disease is an X-linked genetic disease that results in a deficit of ␣-galactosidase activity238; the mutation responsible has been localized to chromosome Xq22.1.109 It leads to the accumulation of stored glycolipid, comprising globotriaosylceramide and galabiosylceramide,201 in DRG and sympathetic ganglion neurons, and in enteric neurons.412 Small-diameter, nociceptive neurons appear to be particularly affected,132,328 and a moderate loss of DRG neurons and a slight loss of sympathetic neurons was described in a report of an autopsy of a 50-year-old heterozygous woman.201 The stored material is Sudan black and PAS positive in frozen sections and is often birefringent in polarized light. It appears as myelinoid, lamellated, electrondense, intralysosomal, zebra-like bodies, 0.5 to 0.75 ␮m in diameter, under the electron microscope.445

Type II Glycogenosis (Pompe’s Disease) and Other Glycogen Storage Diseases Pompe’s disease was the first lysosomal enzyme deficiency to be recognized, and is the result of a mutation in the gene coding for acid maltase (acid ␣-1,4-glycosidase), located to chromosome 17q23.185,186 Spinal motor and DRG neurons and enteric neurons of the myenteric and submucosal plexuses show evidence of excessive glycogen storage in the infantile276 and childhood,287 but not adult,477 forms. The glycogen deposits appear under the light microscope as masses of PAS-positive granules in the cytoplasm of neurons, which often displace the nucleus and organelles to one end of the neuron.276 Defects in the glycogen branching enzyme cause glycogenosis type IV (Andersen’s disease), and other glycogenmetabolizing enzyme defects produce what are collectively now known as adult polyglucosan body diseases. These diseases lead to the formation of characteristic inclusions in neurons and glia of the central and peripheral nervous systems called corpora amylacea, or polyglucosan bodies

710

Pathology of the Peripheral Nervous System

(PGBs).48 These are strongly PAS positive and are usually found in axons rather than perikarya, but in at least one autopsy report of a patient in early infancy with glycogenosis type IV, PAS-positive PGBs were observed in spinal motor neurons.453 An autopsy of an infant with glycogen branching enzyme and phosphorylase deficiency found extensive storage of PAS-positive material in spinal and cranial motor neurons, DRG neurons, and enteric neurons.187 This material was present as membrane-bound granular aggregates around the Golgi apparatus and as large filamentous bodies similar to PGBs.

Chédiak-Higashi Syndrome Chédiak-Higashi syndrome is a multisystem disorder that is due to defective lysosomal trafficking regulator protein,489 the gene encoding which is located to chromosome 1q42q44.128 The disease is characterized pathologically by the occurrence of anomalous giant granules in leukocytes, as well as occasional lymphocytic infiltration of sympathetic and sensory ganglia. Pathologic changes to spinal motor neurons, such as chromatolysis and degeneration, appear to be related to the level of spinal cord examined: lumbosacral segments were the most affected, with little alteration seen at thoracic or cervical levels in two autopsies.438 Three types of inclusion body were observed in DRG and spinal motor neurons: spherical cytoplasmic bodies, “glomerate” bodies, and small discrete granules. Because the last two types of body were never seen in the same neurons as the large spherical bodies, the large spherical bodies may have been formed by fusion of the smaller ones. All types of body were strongly PAS positive and autofluorescent.

Leukodystrophies Metachromatic leukodystrophy is normally caused by a deficiency of activity of arylsulfatase, the gene for which is located in the region of chromosome 22q13.31.321 However, a deficiency of sphingolipid activator protein 1 (gene located to chromosome 10) produces a very similar syndrome. The only pathology described in the peripheral nervous system was the presence of PAS-positive MCBs in enteric neurons.393 Adrenomyeloneuropathy (also known as adrenoleukodystrophy) caused atrophy, but no significant loss, of large DRG neurons in three patients with this disease; many mitochondria contained “lipidic” inclusions.355,356 Globoid cell leukodystrophy (or Krabbe’s disease) is caused by a deficiency of galactocerebrosidase, an enzyme involved in myelin metabolism, the gene for which is located in the region of chromosome 14q31.42 This leads to myelin breakdown and a demyelinating neuropathy. Although few data are available on pathologic alterations in the human disease, membrane-bound bodies containing irregularly striated or lamellated material were described

in DRG neurons of the “twitcher” mouse, which has an identical enzyme deficiency.93

Refsum’s Disease Also known as heredopathia atactica polyneuritiformis, Refsum’s disease is a peroxisomal disorder that leads to the accumulation of phytanic acid throughout the body and subsequently ataxia, blindness, deafness, skeletal hyperostosis, and peripheral neuropathy. Around 45% of patients with Refsum’s disease were found to have mutations in the phytanoyl–coenzyme A hydroxylase gene located in the region of chromosome 10p13,497 indicating that it is a heterogeneous syndrome. There is at least one report of autopsy findings,52 and cultured rat DRG neurons exposed to phytanic acid accumulated non–membrane-bound osmiophilic inclusions at the periphery of perikarya and within mitochondria.91

Tangier Disease (Hereditary High-Density Lipoprotein Deficiency) Tangier disease is inherited as an autosomal codominant disease in which some heterozygotes have a peripheral neuropathy and premature coronary heart disease. Cholesterol ester deposition in various organs, particularly those of the reticuloendothelial system, as well as peripheral neuropathy, characterize homozygotes. It is caused by mutations in a gene that encodes an ATP-binding cassette transporter 1, localized to chromosome 9q31,39 and results in defective apolipoprotein-mediated efflux of cellular cholesterol and phospholipids and a marked deficiency of serum highdensity lipoprotein.375 In an autopsy study of a patient with a homozygous form of the disease, there was evidence of neuronal death in the L5 DRGs, with small-diameter neurons being particularly affected. Remaining neurons were found to contain large, membrane-bound lipid inclusions resembling giant lipofuscin granules. The neuronal inclusions were around 3 ␮m in diameter, were nonpigmented, had electron-dense and electron-lucent components, and displayed only weak autofluorescence.396 Spinal motor neurons also contained these inclusions, but there was no evidence of their loss at the sacral level. However, in another report of an autopsy of a patient with a homozygous form, the authors described a severe loss of motor neurons at the cervical level of the spinal cord and in the facial nucleus, but only a slight loss at lumbar levels.11

NEUROLOGIC MUTANTS These animals are the result of spontaneous mutations that produce a peripheral neuropathy, and are included where they have contributed to the understanding of equivalent human diseases.

711

Pathology of Peripheral Neuron Cell Bodies

Dystonia Musculorum The mutation is transmitted by a single autosomal recessive gene (dt), which has been mapped to mouse chromosome 6p12.40 It leads to alterations in the structure of dystonin, a cytoskeletal link protein forming bridges between F-actin and intermediate filaments.81 The high level of expression of dystonin in neurons of the cranial and spinal sensory ganglia matches those populations vulnerable to degeneration in dt mice.27 By the age of 3 to 4 weeks, the animal’s movements appear jerky and progressively less coordinated: The mice cannot now make purposeful movements, and affected animals usually die at about the time of weaning. Animals are born with fewer neurons in all spinal and cranial (5th, 7th, 9th, and 10th) ganglia examined, with the remaining neurons often appearing chromatolytic.95

The Sprawling Mouse and the Mutilated Foot Rat These two neurologic mutants have similar clinical and pathologic phenotypes, but do not totally match any known human neuropathy or transgenic mouse neuropathy. The most similar human disorders would be the ulcero-mutilating neuropathies HSAN I/HMSN IIB.99 The mouse mutation (Swl) is transmitted as an autosomal dominant trait, while the rat mutation (mf) is transmitted as an autosomal recessive trait, which has recently been found to be due to a mutation in a gene on chromosome 14 (M. J. Groves and F. Scaravilli, unpublished observations). Both the mouse and the rat manifest ataxia and loss of nociception in the limbs from birth: The ataxia progresses and the feet become ulcerated and eventually mutilated, although mutilation (not observed in the Swl mouse) can be seen occasionally at birth. There is neither muscle weakness nor wasting; animals that survive the early postnatal period usually live for a normal length of time. Pathologic findings shared by the adult forms of the two mutants consist of severe reduction in the number of both large- (with myelinated axons) and small- (with unmyelinated axons) diameter neurons of the lumbar and cervical DRGs. In the mf rat this reduction is around 50% in L4 and L5 DRGs, but some ganglia were reported as containing only 25% or less of the normal number of neurons.94,214 Motor neurons and axons are also decreased in number, although the reduction is less severe and mainly affects the smaller gamma motor neurons. Cranial and sympathetic ganglia and thoracic DRGs are reported as having normal numbers of neurons.214 The only morphologic difference between the surviving motor and sensory neurons of the mf rat and normal rats is that the remaining motor neurons are approximately 25% larger than the largest seen in normal littermates.388 A striking difference between the Swl mouse and the mf rat consists in the presence of light and electron microscopic features of

chromatolysis in some of the residual ganglion neurons in the Swl mouse. This loss of sensory neurons is due to events in development. As early as at the 11th embryonic day (E11), neurons with abnormally indented nuclei are present in the nascent DRGs of the Swl mouse. From E15 to E20–21, excessive numbers of immature DRG neurons die through apoptosis in the cervical and lumbar DRGs of both mf rats and Swl mice.58,386 This loss of sensory neurons at this time probably underlies the subsequent loss of gamma motor neurons330,388 and second-order sensory neurons in the gracile nucleus.58

The Wobbler Mouse This spontaneous mutation (gene symbol wr114) is located on chromosome 11 and is transmitted in an autosomal recessive manner.365 It leads to a progressive degeneration of motor neurons with a time course and pattern similar to those usually seen in human motor neuron disease. The earliest signs of the disease become evident during the fourth week, and progressive weakness, particularly in the forelimbs, soon becomes obvious and is linked to numbers of surviving motor neuron axons.351 Early morphologic changes in spinal and cranial motor neurons of the wobbler mouse include swelling and chromatolysis,31,96 and at later stages perikaryal vacuoles derived from dilated cisternae of the rough endoplasmic reticulum appear and may extend into the dendritic processes.8 Large motor neurons appear to be particularly affected, either by loss or lack of differentiation.31 These cellular changes seem to result in their death via an apoptotic pathway, involving the activation of protease-activated receptor 1, the incidence of which is greater at cervical than at lumbar levels of the spinal cord.118

Mouse Motor Neuron Degeneration This spontaneous mutation in C57Bl/6 mice follows an autosomal dominant inheritance pattern, and produces a disorder that begins by 6 months of age with limping of the hind limbs and progresses to spastic paralysis of the hind and, later, forelimbs.293 The mutation produces a progressively larger loss of spinal and cranial motor neurons, particularly those in the lumbosacral region, which undergo chromatolytic changes and accumulation of ubiquitinated and autofluorescent lipofuscinoid inclusions that contain various lysosomal enzymes.289,294 The apoptosis-associated enzyme transglutaminase type 2 is “superactivated” in the spinal cord at the same time that the motor neuron disease begins to manifest itself.196 It is now apparent that this mouse is a homologue196 of a human neuronal ceroid lipofuscinosis (progressive epilepsy with mental retardation) caused by a mutation in the CLN8 gene, which encodes for a membrane protein with as yet unknown functions.364

712

Pathology of the Peripheral Nervous System

Inherited Dog Neuropathies The neuropathy in the Brittany spaniel dog is an inherited motor neuron disease (hereditary canine SMA) that is transmitted as an autosomal dominant trait.377 The clinical course of the intermediate (and most common) heterozygous phenotype resembles that of the KugelbergWelander type of juvenile SMA. Furthermore, the accelerated, homozygous form of the disease is of interest in the context of both SMA and possible molecular mechanisms in motor neuron disease. Analysis of the ventral horn cholinergic neuron population showed that large motor neurons either fail to reach the correct size or undergo atrophy.69 Subsequent analysis has shown that alterations in cyclin-dependent kinase 5 expression precede altered expression of low-molecular-weight neurofilament protein155 and subsequent inhibition of neurofilament assembly or transport.311 An inherited sensory neuropathy has been reported in a litter of nine English Pointer dogs that presented as acral mutilation and analgesia78 very similar to that suffered by the mf rat (see above). Upon examination, the density of DRG neurons was reduced: Some vacuolated and chromatolytic neurons and some nodules of Nageotte were present, but the authors believed that much of the neuron loss (25% to 50% in DRGs) had probably occurred before birth and had particularly affected the large neuron population.

NEUROPATHIES PRODUCED BY SYSTEMIC METABOLIC DISORDERS Porphyria This group of conditions is due to defects in any of the enzymes and proteins involved in porphyrin metabolism, and only those forms caused by defects in hepatic heme synthesis produce a distal axonopathy. They are inherited in an autosomal dominant fashion. Autopsies have found either no evidence of loss of either spinal motor or DRG neurons49 or evidence of some loss of both. 188,460 Chromatolysis was seen in some of the larger spinal motor neurons in the cervical and lumbar enlargements as well as occasional vacuolated neurons in the lumbar region; chromatolysis was less evident in DRG neurons.49

Diabetes Mellitus This disease produces a loss of axons in sural nerve biopsies, but evidence that this is directly related to a loss of DRG neurons, at least initially in the progression of neuropathy, is by no means certain. However, pathologic alterations of spinal cord motor neurons, DRGs, and autonomic ganglia in autopsy specimens from diabetics have been described.

DRGs from a patient with a 50-year history of diabetes contained sparse infiltrates of mononuclear cells without any obvious neuron loss, as well as no obvious loss of spinal motor neurons.492 This case also demonstrated atrophy of the superior cervical ganglion with prominent neuron loss and lymphocytic infiltration, but a normal myenteric plexus in both stomach and intestine. Schmidt398 also found lymphocytic infiltration of autonomic ganglia, but found that it was not selectively increased in frequency and intensity compared to controls. An absence of significant changes in DRG and motor neurons of diabetic patients was also reported by Schmidt et al.400 However, DRG neuron vacuolation was a significant finding in three of five autopsies of diabetic patients, but only one or two of these cases showed evidence of either motor or sensory neuron loss.156 Pathologic alterations in autonomic neurons seem to be more significant: Abnormally enlarged sympathetic ganglion cells containing PAS-positive material have been observed.13,156 Infiltration of autonomic ganglia by inflammatory cells, vacuolated and enlarged sympathetic neurons, club-shaped enlargements of neuronal processes, and the formation of “neuromatous nodules” have been reported in each of five autopsies of diabetic patients with a severe autonomic neuropathy.92 Experimental work in rats has shown that hyperglycemia can induce apoptosis in DRG and Schwann cells,374 but studies of animal models of diabetes have found DRG neuron atrophy, rather than loss, a more significant feature.512,533 Further experimental work showed abnormal phosphorylation of neurofilaments of DRG cells both in spontaneous and streptozotocin-induced diabetic rats,117 as well as decreased expression of cytoskeletal proteins and their reduced incorporation into distal axons.405 A study of BB/Wor diabetic rats showed normal histologic appearances and numbers of motor neurons at spinal level L5.300

Amyloidosis The types, presentation, and treatment of amyloidoses are covered in Chapter 83. This section deals with any pathologic alterations to peripheral neuron perikarya described in amyloidosis. Amyloidosis is a disorder of protein conformation involving proteins normally present in the body fluids as soluble precursors. An alteration to the genetic coding for these proteins, or to other proteins involved in the formation of their tertiary and quaternary structure, can lead to their deposition as insoluble fibrils in tissues. The fibrils consist of self-assembled, low-molecular-weight mass peptides with beta pleated sheet staining characteristics,125 and seem to be directly toxic to neurons. Both peripheral nerves and ganglia are affected in cases of primary and familial amyloidoses. The pathologic description that follows is based on personal observations

713

Pathology of Peripheral Neuron Cell Bodies

on the autopsy of a patient, 49 years old, suffering from a familial form of type I amyloid polyneuropathy. In the spinal cord, a number of meningeal vessels showed focal deposits of amyloid; there were no obvious signs of degeneration of motor neurons, but some were chromatolytic. The dorsal columns showed striking pallor involving both the gracile and cuneate tracts. Several muscles from the upper and lower limbs were sampled; they showed a mild to moderate degree of denervation. DRGs (see Fig. 31–9B and C) were severely affected. Areas of amyloid formed irregular aggregates replacing and separating groups of neurons. The latter were severely reduced in number, as confirmed by the large number of nodules of Nageotte, and atrophy of the surviving neurons was suggested by hypertrophy of their surrounding satellite cells, which formed concentric layers around them. A number of the surviving neurons showed signs of degeneration and chromatolysis. In addition, amyloid was present infiltrating endoneurial blood vessels and was seen to continue into the roots and peripheral nerves, but there was no evidence of an inflammatory cell infiltrate. The superior cervical and celiac sympathetic ganglia showed comparable changes, with the nervous tissue being largely replaced by areas of amyloid deposition. Increased numbers of nodules of Nageotte, evidence of chromatolysis, and absence of inflammatory cells were features similar to those seen in the sensory ganglia. In striking contrast, the two gasserian ganglia of the same patient were minimally involved by the changes and showed a small number of nodules of Nageotte and no amyloid deposits. The changes described above are similar to those reported by a number of authors in patients with familial forms of amyloidosis.297,425,467 All observers agree that spinal cord motor neurons are not decreased in number, although they show evidence of chromatolysis, and that the posterior columns are pale. With regard to the severity of the changes in dorsal root and autonomic ganglia, unlike other groups, Misu et al.297 reported only moderate neuron loss and amyloid deposition in patients with familial amyloid polyneuropathy TTR Met 30.

INFECTIVE AND INFLAMMATORY NEUROPATHIES Varicella Zoster and Herpes Zoster Infection Following primary infection, varicella-zoster virus (VZV) establishes latency in DRG neurons and can reactivate years later to produce herpes zoster infection; studies in humans have confirmed that VZV DNA persists in DRG neurons and VZV proteins are expressed during the latent period,274 although there have been conflicting reports about the cellular localization of VZV within DRGs. In a study of DRGs of humans and rats, Annunziato et al.10 and

Lungu et al.271 detected VZV in both satellite and nerve cells; in contrast, Kennedy et al.230 found VZV DNA predominantly in the latter. When reactivated, the lesions are often localized to a dermatome, and pathologic changes such as hemorrhagic infarction, massive lymphocytic infiltration, and thrombosed small arteries may be found in the corresponding DRGs, roots, and spinal cord. Within these areas there is usually massive loss of DRG neurons, and those remaining show shrinkage, chromatolysis, and vacuolation.90 Motor symptoms may also complicate herpes zoster.140

Poliovirus Poliovirus is the etiologic agent that causes paralytic poliomyelitis by infecting and killing motor neurons, and that can produce additional neurologic problems (postpolio syndrome) years after the initial infection and disease. This latter effect may be due to persistence of the virus in neural and non-neural cells.67 Studies in monkeys show that infected motor neurons may increase in size303 before chromatolysis, nuclear pyknosis, and neuronophagia (see Fig. 31–4), leading to their loss.379

Cytomegalovirus Infection Cytomegalovirus (CMV) is another member of the Herpesviridae family and has been identified as a pathogen in several mammals, including humans. Cytomegalovirus involvement of the nervous system has become very frequent with the advent of human immunodeficiency virus (HIV) infection,387 and peripheral neuropathies resulting from CMV infection are becoming relatively common.129,307 Eidelberg and co-workers103 examined DRGs from patients with acquired immunodeficiency syndrome (AIDS) and CMV polyradiculopathy and found them to be unaffected, although spinal motor neurons often showed chromatolytic changes secondary to root damage. In a personally observed case, a patient was found to have florid lymphocytic ganglionitis with intranuclear inclusions in a small number of ganglion cells. A study by Nagano et al.317 found CMV antigen in the DRGs from 4 of 10 patients, but also in 1 of 5 control DRGs. However, no intranuclear or intracytoplasmic inclusions were seen by this group, and no relationship could be established between the expression of CMV antigen, density of nodules of Nageotte, and other pathologic changes.

Human Immunodeficiency Virus Peripheral neuropathies related to HIV and human T-lymphotropic virus type 1 (HTLV-1) infection are considered in Chapter 112, and generally present as distal sensory polyneuropathies.361 The main pathologic features seen in DRGs are inflammatory cell infiltration with loss of

714

Pathology of the Peripheral Nervous System

neurons and the formation of nodules of Nageotte; these are also seen in patients with antiretroviral drug toxic neuropathies.341 The inflammatory component in the ganglia consists of macrophages and a diffuse infiltration with T lymphocytes, described in seven patients with AIDS, two of whom also had evidence of CMV ganglionitis.389 In a study of sensory and sympathetic ganglia from 12 HIVpositive patients,111 many ganglia contained lymphocytes and macrophages as well as showed expression of major histocompatibility complex class II and occasional degenerating nerve cells. The gp41 HIV-1 antigen was found in larger amounts in sensory than in sympathetic ganglia.

DRGs160 showing collections of T lymphocytes around neurons. In milder cases there was degeneration of individual ganglion neurons; in an advanced case only a few ganglion cells remained. Inflammatory cell infiltrates have also been described in DRGs in Sjögren’s syndrome associated with lymphoma.197 The pathogenetic mechanism of neuron damage and loss in Sjögren’s syndrome remains unclear, but immunostaining of rat small sensory neurons by serum and cerebrospinal fluid (CSF) of a 59-year-old woman suffering from the syndrome suggests an autoimmune-related mechanism.383

Lymphoma Other Viruses Producing Peripheral Neuropathy A motor neuron disease in the mouse caused by “type C” retrovirus has been reported.135 The disorder is transmitted via the mother’s milk, begins at 6 months, and is characterized by hind limb paralysis and spongiform degeneration. It is clinically similar to that described in Mnd mice; however, the former is associated, in 10% of the animals, with lymphoma. HTLV-1 is the cause of endemic tropical spastic paraparesis (TSP) or HTLV-1–associated myelopathy. An autopsy of a TSP case found increased lipofuscin deposition in DRG neurons, but these were otherwise normal.262

Diphtheria The bacterium that causes this disease, Corynebacterium diphtheriae, produces a toxin that inactivates elongation factor 2, which is required for protein synthesis. When this toxin enters the systemic circulation, one of the clinical manifestations is a demyelinating neuropathy. When the purified toxin is injected into muscles, marked morphologic alterations to motor neurons occur, including the dilatation and fragmentation of rough endoplasmic reticulum similar to that produced by ricin.358

Leprosy There are few reports of pathologic alteration to peripheral neurons in leprosy, but experimental inoculation of nude mice with Mycobacterium leprae resulted in a progressive depletion of substance P and CGRP from sensory neurons, and a total loss of CGRP from motor neurons 12 months later.223

Sensory Ganglionitis in Sjögren’s Syndrome An ataxic sensory and autonomic neuropathy may develop in patients with Sjögren’s syndrome. The primary involvement of sensory neurons has been confirmed in biopsies of

There are occasional reports of peripheral neuropathy being a remote effect of lymphoma, as opposed to a neuropathy caused by lymphomatous infiltration of peripheral nerve. Autopsies of five patients with a subacute lower motor neuron syndrome found prominent motor neuron (see Fig. 31–8B and C) degeneration,402 and a severe depletion of DRG neurons with inflammatory changes was found in autopsies of patients with a subacute sensory neuronopathy.197

Paraneoplastic Neuropathy Paraneoplastic neuropathies are associated with systemic malignancies, and their presentation may precede the clinical appearance of the neoplasm. Their pathogenesis is not well established, although increasing evidence supports an autoimmune mechanism.385 Involvement of sensory and autonomic ganglia (ganglioradiculoneuritis) in paraneoplastic disorders may be focal or diffuse, and in the former it can be asymmetrical and localized to the cervical or lumbar region. The degree of severity of the lesions varies, and complete destruction of the affected ganglion can result. Disappearance of neurons is accompanied by proportional increase of nodules of Nageotte, whereas the remaining nerve cells may show cytoplasmic vacuolation, chromatolysis, and nuclear shrinkage. Inflammation can appear as diffuse lymphocytic infiltration or cuffing of the vessels, and may extend to the posterior roots. Henson and Urich182 questioned the existence of two separate subgroups of ganglionopathies, inflammatory and noninflammatory. They argued that the likelihood of finding inflammatory cells depends on the number of ganglia available and the duration of the illness, because the inflammatory process tends to fade and disappear with time. Disappearance of ganglion cells is followed by atrophy of the corresponding posterior roots, with the degree of posterior column pallor being proportional to the extent and severity of the neuronal loss. Involvement of the autonomic nervous system has also been described, although nerve cell loss in the sympathetic ganglia has not been reported to reach the severity of that

Pathology of Peripheral Neuron Cell Bodies

seen in sensory ganglia; this involvement may be accompanied by a lymphocyte infiltrate.219 The parasympathetic component can also be affected,426 and inflammation in the myenteric plexus has also been described.260 The existence of a pure form of motor neuron disease in paraneoplastic disorders, in addition to motor involvement as part of a poliomyelitis affecting both anterior and posterior horns, has been debated for a long time.385 Once the cases of paraneoplastic motor neuron disease associated with myelitis are excluded, a pure degenerative form of paraneoplastic motor neuron disease is very rare. However, Verma et al.483 detected anti-Hu antibodies in a 51-year-old man with small cell cancer of the lung who developed weakness and muscle wasting in the upper half of the body. Electromyography showed signs of denervation, and spinal motor neuron loss was found post mortem. The serum and CSF of a large number of patients suffering from the disorders described above contain polyclonal antibodies, called anti-Hu (also type 1 antineuronal nuclear antibodies, or ANNA-1).80 Virtually all patients belonging to this subgroup have (or have had) small cell lung carcinoma, while a small number suffer from neuroblastoma or breast or prostate carcinoma. The characteristic feature of these antibodies is that they stain the nuclei and, discretely, the cytoplasm of certain neurons of both the central and peripheral nervous systems.146

TOXINS Hexacarbons Exposure to toxic doses of hexacarbons such as n-hexane produces a distal axonopathy in humans by interfering with neurofilament structure and assembly. The neurotoxic metabolite 2,5-hexanedione is thought to interact directly with lysine residues, which are prevalent in neurofilament proteins, and n-hexane neuropathy has morphologic alterations to axons similar to those seen in GAN. Investigations into the effects of 2,5-hexanedione on cultured human DRG neurons found tangles of neurofilaments in the perikarya of neurons, as well as in their axons.306 Systemic administration of 2,5-hexanedione induces chromatolysis in rat DRG neurons over a time course of weeks, rather than days as seen following peripheral axotomy.432 Chronic dosing eventually led to the formation of nodules of Nageotte,431 and 2,5-hexanedione itself does not appear to be as acutely toxic to neurons as it is to glia.433

Metals “Heavy” metals (such as lead, mercury, arsenic, and cadmium) often have an affinity for nucleophilic moieties, particularly sulfhydryl groups, with intoxication tending to affect membrane transport processes. Organometal com-

715

pounds are generally far more toxic than inorganic salts, probably as a result of their greater lipid solubility and concomitantly more rapid and wider distribution in vivo. Jacobs et al.212,213 originally demonstrated the selective effect of methylmercury upon large-diameter rat and rabbit DRG neurons. A dose of 7.5 mg/kg/day for 1 to 4 (rabbits) or 8 (rats) days initially produced a loss of ribosomes from rough endoplasmic reticulum. This was followed by nuclear and perinuclear changes, cytoplasmic vesiculation, and necrosis of some neurons with evidence of neuronophagia.213 At a dose of 5 mg/kg/day for 10 days, largediameter rat DRG neurons demonstrated significant delays in action potential onset; vacuolation and occasional evidence of neuronophagia were also observed in DRGs.89 Methylmercury intoxication also affects motor neurons, producing vacuolation of this population in rats.434 Indeed, acute methylmercury intoxication leads to the death of rat spinal motor neurons (as indicated by evidence of neuronophagia) by the 16th day after 10 days of daily injections of 10 mg/kg, with all large motor neurons disappearing by 18 days.434 The organotin compounds triethyl tin and trimethyl tin are toxic to myelinating Schwann cells and neurotoxic, respectively: 6 mg/kg of trimethyl tin orally administered to rats produced vacuolar changes, accumulation of lysosomes, formation of myeloid bodies, and dissolution of the Nissl substance in DRG neurons (Fig. 31–10A) up to 2 weeks later.53 Vacuolation was also seen in motor neurons of gerbils and rats with trimethyl tin intoxication.325 Lead intoxication produces a differing pattern of neuropathy depending upon the species studied and dosing schedule used. One of the few reports of the histopathology of peripheral neurons in lead poisoning involved long-term intoxication in chickens, which produced degeneration of motor neurons that the authors suggested was similar to that seen in human motor neuron diseases.288 Platinum is particularly neurotoxic to sensory neurons, neuropathy often being the result of treatment with platinum-based cytotoxic drugs such as cisplatin. Cisplatin induces rat DRG neurons to undergo apoptosis in vitro and in vivo,120,143 and systemic exposure to 2 mg/kg twice weekly for 4.5 weeks produced reductions in perikaryal, nuclear, and nucleolar areas of DRG neurons in rats.468 “Necrotic” (meaning dying rather than true necrosis) neurons, nodules of Nageotte, and reduced perikaryal volumes have been reported in DRGs from autopsies of some patients treated with cisplatin and carboplatin.246 Studies using cultured DRG neurons showed that exposure to aluminum phosphate produced intraneuronal neurofibrillary spheroids,407 as well as a decreased expression of neurofilament proteins.141 All of these changes were reversible if dosing with aluminum was discontinued. Chronic administration of aluminum chloride to rabbits was reported to produce a loss of CNS and spinal motor neurons via a process of neurofibrillary degeneration.228

716

Pathology of the Peripheral Nervous System

FIGURE 31–10 A, A degenerating rat DRG neuron following acute intoxication with triethyl tin. A presumptive nucleus of a phagocytic cell (arrow) apparently within the degenerating perikaryon probably indicates the beginning of neuronophagia (see Fig. 31–4D). The apparent dissolution of the nucleus (arrowheads) and vesiculation of the cytoplasm indicates an acute, nonapoptotic form of neuron death probably involving loss of membrane integrity or homeostatic function. Bar: 10 ␮m. (Courtesy of J. M. Jacobs.) B, Photomicrograph of a toluidine blue–stained resin section of adult rat DRG shows chromatolytic large-diameter neurons (arrows) following intraperitoneal injection of pyridoxine. Excessive ingestion of pyridoxine (vitamin B6) is toxic to large sensory neurons in humans (and rats). Acutely toxic doses lead to loss of neurons while lower doses produce axonal atrophy. Bar: 10 ␮m. (Courtesy of J. M. Jacobs and M. Pires.) C and D, Amiodarone can produce a sensorimotor neuropathy of a mixed axonal and demyelinating type, which is associated with the presence of intracellular electron-dense, lamellated inclusions. C, Photomicrograph showing a toluidine blue–stained resin section of adult rat celiac ganglion following amiodarone intoxication. The osmiophilic inclusions (arrows) can be seen within the neuronal perikarya. Bar: 10 ␮m. (Courtesy of J. M. Jacobs.) D, Electron micrograph of an adult rat superior cervical ganglion neuron 2 weeks after administration of amiodarone. The electron-dense inclusions characteristic of amiodarone neuropathy can be seen in neurons and axons (arrows). Bar: 1 ␮m. (Courtesy of J. M. Jacobs.)

Cadmium is directly toxic to a wide variety of tissues, and also exerts secondary effects through its action as a very powerful vasoconstrictor. Acute intoxication has been reported by Gabbiani et al. to produce hemorrhagic lesions in dorsal root and fifth cranial nerve ganglia, but in addition chromatolytic and (apparently) necrotic neurons are clearly visible in their figures.130 Other metals, such as thallium and gold, which have been reported to produce neuropathy, have not had any

associated pathologic alterations to peripheral neurons described.

Vincristine Vincristine is a microtubule-depolymerizing drug used in the treatment of certain cancers that produces a painful sensory neuropathy in humans. When administered to rats, it induced the swelling of large-diameter DRG neurons,

Pathology of Peripheral Neuron Cell Bodies

possibly caused by the observed buildup of neurofilaments in their perikarya and proximal segments of axon.464

Pyridoxine Pyridoxine (vitamin B6), when taken in excessive amounts, is a sensory neurotoxin that interferes with sensory neuron (see Fig. 31–10B) cytoskeletal organization.304 A dose of 600 to 1200 mg/kg/day administered to rats produced necrosis of DRG neurons, particularly large-diameter neurons513; these authors also reported substantial species susceptibilities to pyridoxine neurotoxicity.

Amiodarone Amiodarone is a diiodinated benzofuran derivative used to treat cardiac arrhythmias, and has been associated with a number of cases of peripheral neuropathy. A dose of 50 mg/kg/day administered orally to rats induced accumulations of lipids within lysosomes, leading to the formation of cytoplasmic bodies in neurons from regions without bloodbrain or blood-nerve barriers. Small to medium-sized DRG and gasserian ganglion neurons, autonomic neurons, and enteric neurons subsequently showed inclusions (see Fig. 31–10C and D), with autonomic neurons the worst affected and showing evidence of degenerative changes.70

Acrylamide Chronic low-level exposure to acrylamide produces a distal axonopathy, and animals exposed to acute high doses (30 mg/kg/day) were found to have chromatolytic and vacuolated DRG neurons.220,430 Acrylamide does not appear to produce a loss of DRG neurons, just atrophy (by up to 28%) of the large-diameter subpopulation, indicating that alterations to neurofilament function may underlie its neurotoxicity.451

Organophosphates Exposure to organophosphates has been linked to a neuropathy termed organophosphate-induced delayed neuropathy, an axonopathy affecting neurons with long axons. Their neurotoxicity is thought to be due to interaction with “neuropathy target esterase,” which is located in neuronal cell bodies and axons.149 One of the effects of this interaction appears to be the hyperphosphorylation of tubulin and microtubule-associated protein-2, which in turn may cause destabilization of microtubule assemblies and axonal degeneration.62

REFERENCES 1. Acheson, A., and Lindsay, R. M.: Non-target-derived roles of the neurotrophins. Philos. Trans. R. Soc. Lond. B 351:417, 1995.

717

2. Adalbert, R., Engelhardt, J. I., and Siklos, L.: DLHomocysteic acid application disrupts calcium homeostasis and induces degeneration of spinal motor neurons in vivo. Acta Neuropathol. (Berl.) 103:428, 2002. 3. Aldskogius, H., and Arvidsson, J.: Nerve cell degeneration and death in the trigeminal ganglion of the adult rat following peripheral nerve transection. J. Neurocytol. 7:229, 1978. 4. Aldskogius, H., Arvidsson, J., and Grant, G.: Axotomyinduced changes in primary sensory neurons. In Scott, S. A. (ed.): Sensory Neurons. New York, Oxford University Press, p. 363, 1992. 5. Aldskogius, H., and Risling, M.: Effect of sciatic neurectomy on neuronal number and size distribution in the L7 ganglion of kittens. Exp. Neurol. 74:597, 1981. 6. Alkhayat, A. H., Kraemer, S. A., Leipprandt, J. R., et al.: Human beta-mannosidase cDNA characterization and first identification of a mutation associated with human betamannosidosis. Hum. Mol. Genet. 7:75, 1998. 7. Altarescu, G., Sun, M., Moore, D. F., et al.: The neurogenetics of mucolipidosis type IV. Neurology 59:306, 2002. 8. Andrews, J. M.: The fine structure of the cervical spinal cord, ventral root and brachial nerves in the wobbler (W2) mouse. J. Neuropathol. Exp. Neurol. 34:12, 1975. 9. Andrews, T. J.: Autonomic nervous system as a model of neuronal aging: the role of target tissues and neurotrophic factors. Microsc. Res. Tech. 35:2, 1996. 10. Annunziato, P. W., La Russa, P., Lee, P., et al.: Evidence of latent varicella-zoster virus in rat dorsal root ganglia. J. Infect. Dis. 178(Suppl. 1S):48, 1998. 11. Antoine, J. C., Tommasi, M., Boucheron, S., et al.: Pathology of roots, spinal cord and brainstem in syringomyelia-like syndrome of Tangier disease. J. Neurol. Sci. 106:179, 1991. 12. Apfel, S. C., Wright, D. E., Wiideman, A. M., et al.: Nerve growth factor regulates the expression of brain-derived neurotrophic factor mRNA in the peripheral nervous system. Mol. Cell. Neurosci. 7:134, 1996. 13. Appenzeller, O., and Richardson, E. P.: The sympathetic chain in patients with diabetic and alcoholic polyneuropathy. Neurology (Minneap.) 16:1205, 1966. 14. Arvidsson, J., Ygge, J., and Grant, G.: Cell loss in lumbar dorsal root ganglia and transganglionic degeneration after sciatic nerve resection in the rat. Brain Res. 373:15, 1986. 15. Arvidsson, U., Johnson, H., Piehl, F., et al.: Peripheral nerve section induces increased levels of calcitonin generelated peptide (CGRP)-like immunoreactivity in axotomized motoneurons. Exp. Brain Res. 79:212, 1990. 16. Bahadori, M. H., Al-Tiraihi, T., and Valojerdi, M. R.: Sciatic nerve transection in neonatal rats induces apoptotic neuronal death in L5 dorsal root ganglion. J. Neurocytol. 30:125, 2001. 17. Barrett, G. L., and Bartlett, P. F.: The p75 nerve growth factor receptor mediates survival or death depending on the stage of sensory neuron development. Proc. Natl. Acad. Sci. U. S. A. 91:6501, 1994. 18. Bates, G. P., Mangiarini, L., and Davies, S. W.: Transgenic mice in the study of polyglutamine repeat expansion diseases. Brain Pathol. 8:699, 1998.

718

Pathology of the Peripheral Nervous System

19. Beaver, D. L., Moses, H. L., and Ganote, C. E.: Electron microscopy of the trigeminal ganglion. II. Autopsy study of human ganglia. Arch. Pathol. 79:557, 1965. 20. Beiswanger, C. M., Roscoe-Graessle, T. L., Zerbe, N., et al.: 3-acetylpyridine-induced degeneration in the dorsal root ganglia: involvement of small diameter neurons and influence of axotomy. Neuropathol. Appl. Neurobiol. 19:164, 1993. 21. Bendotti, C., Calvaresi, N., Chiveri, L., et al.: Early vacuolization and mitochondrial damage in motor neurons of FALS mice are not associated with apoptosis or with changes in cytochrome oxidase histochemical reactivity. J. Neurol. Sci. 191:25, 2001. 22. Bennett, D. L., Averill, S., Clary, D. O., et al.: Postnatal changes in the expression of the trkA high-affinity NGF receptor in primary sensory neurons. Eur. J. Neurosci. 8:2204, 1996. 23. Berciano, J., Combarros, O., Figols, J., et al.: Hereditary motor and sensory neuropathy, type II. Clinicopathological study of a family. Brain 109:897, 1986. 24. Bergman, E., Fundin, B. T., and Ulfhake, B.: Effects of aging and axotomy on the expression of neurotrophin receptors in primary sensory neurons. J. Comp. Neurol. 410:368, 1999. 25. Berlit, P., and Schwechheimer, K.: Neuropathological findings in radiation myelopathy of the lumbosacral cord. Eur. Neurol. 27:29, 1987. 26. Bernal, F., Graus, F., Pifarre, A., et al.: Immunohistochemical analysis of anti-Hu-associated paraneoplastic encephalomyelitis. Acta Neuropathol. (Berl.) 103:509, 2002. 27. Bernier, G., Brown, A., Dalpe, G., et al.: Dystonin expression in the developing nervous system predominates in the neurons that degenerate in dystonia musculorum mutant mice. Mol. Cell. Neurosci. 6:509, 1995. 28. Bieber, F. R., Mortimer, G., Kolodny, E. H., and Driscoll, S. G.: Pathologic findings in fetal GM1 gangliosidosis. Arch. Neurol. 43:736, 1986. 29. Bigotte, L., and Olsson, Y.: Cytotoxic effects of Adriamycin on mouse hypoglossal neurons following retrograde axonal transport from the tongue. Acta Neuropathol. (Berl.) 61:161, 1983. 30. Bledsoe, A. W., Jackson, C. A., McPherson, S., and Morrow, C. D.: Cytokine production in motor neurons by poliovirus replicon vector gene delivery. Nat. Biotechnol. 18:964, 2000. 31. Blondet, B., Carpentier, G., Ait-Ikhlef, A., et al.: Motoneuron morphological alterations before and after the onset of the disease in the wobbler mouse. Brain Res. 930:53, 2002. 32. Bodian, M., Stephens, F. D., and Ward, B. L. H.: Hirschsprung’s disease and idiopathic megacolon. Lancet 1:6, 1949. 33. Boettger, M. K., Till, S., Chen, M. X., et al.: Calciumactivated potassium channel SK1- and IK1-like immunoreactivity in injured human sensory neurons and its regulation by neurotrophic factors. Brain 125:252, 2002. 34. Bondok, A. A., and Sansone, F. M.: Retrograde and transganglionic degeneration of sensory neurons after a peripheral nerve lesion at birth. Exp. Neurol. 86:322, 1984. 35. Bradley, J. L., Blake, J. C., Chamberlain, S., et al.: Clinical, biochemical and molecular genetic correlations in Friedreich’s ataxia. Hum. Mol. Gen. 9:275, 2000.

36. Bredesden, D. E., and Rabizadeh, S.: p75NTR and apoptosis: trk-dependent and trk-independent effects. TINS 20:287, 1997. 37. Bridge, P. M., Ball, D. J., Mackinnon, S. E., et al.: Nerve crush injuries—a model for axonotmesis. Exp. Neurol. 127:284, 1994. 38. Brorson, J. R., Schumacker, P. T., and Zhang, H.: Nitric oxide acutely inhibits neuronal energy production. J. Neurosci. 19:147, 1999. 39. Brousseau, M. E., Schaefer, E. J., Dupuis, J., et al.: Novel mutations in the gene encoding ATP-binding cassette 1 in four Tangier disease kindreds. J. Lipid Res. 41:433, 2000. 40. Brown, A., Lemieux, N., Rossant, J., and Kothary, R.: Human homolog of the mouse sequence from the dystonia musculorum locus is on chromosome 6p12. Mamm. Genome 5:434, 1994. 41. Bussmann, K. A., and Sofroniew, M. V.: Re-expression of p75NTR by adult motor neurons after axotomy is triggered by retrograde transport of a positive signal from axons regrowing through damaged or denervated peripheral nerve tissue. Neuroscience 91:273, 1999. 42. Cannizzaro, L. A., Chen, Y. Q., Rafi, M. A., and Wenger, D. A.: Regional mapping of the human galactocerebrosidase gene (GALC) to 14q31 by in situ hybridisation. Cytogenet. Cell Genet. 66:244, 1994. 43. Carlson, C. D., Bai, Y., Ding, M., et al.: Interleukin-1 involvement in the induction of leukemia inhibitory factor mRNA expression following axotomy of sympathetic ganglia. J. Neuroimmunol. 70:181, 1996. 44. Carmel, P. W., and Stein, B. M.: Cell changes in sensory ganglia following proximal and distal nerve section in the monkey. J. Comp. Neurol. 135:145, 1969. 45. Carter, B. D., Kaltschmidt, C., Kaltschmidt, B., et al.: Selective activation of NF-kB by nerve growth factor through the neurotrophin receptor p75. Science 272:542, 1996. 46. Casanovas, A., Ribera, J., Hager, G., et al.: C-jun regulation in rat neonatal motoneurons postaxotomy. J. Neurosci. Res. 63:469, 2001. 47. Casanovas, A., Ribera, J., Hukkanen, M., et al.: Prevention by lamotrigine, MK-801 and N-omega-nitro-L-arginine methyl ester of motoneuron cell death after neonatal axotomy. Neuroscience 71:313, 1996. 48. Cavanagh, J. B.: Corpora-amylacea and the family of polyglucosan diseases. Brain Res. Rev. 29:265, 1999. 49. Cavanagh, J. B., and Mellick, R. S.: On the nature of the peripheral nerve lesions associated with acute intermittent porphyria. J. Neurol. Neurosurg. Psychiatry 28:320, 1965. 50. Cavanaugh, M. W.: Quantitative effects of the peripheral innervation area on nerves and spinal ganglion cells. J. Comp. Neurol. 94:181, 1951. 51. Celik, M., Gokmen, N., Erbayraktar, S., et al.: Erythropoietin prevents motor neuron apoptosis and neurologic disability in experimental spinal cord ischemic injury. Proc. Natl. Acad. Sci. U. S. A. 99:2258, 2002. 52. Cervos-Navarro, J.: Heredopathia atactica polyneuritiformis (Refsum’s disease). Histol. Histopathol. 5:439, 1990. 53. Chang, L. W., and Dyer, R. S.: Trimethyltin induced pathology in sensory neurons. Neurobehav. Toxicol. Teratol. 5:673, 1983.

Pathology of Peripheral Neuron Cell Bodies 54. Chang, Q., and Balice-Gordon, R. J.: Gap junctional communication among developing and injured motor neurons. Brain Res. Brain Res. Rev. 32:242, 2000. 55. Chang, Q., Pereda, A., Pinter, M. J., and Balice-Gordon, R. J.: Nerve injury induces gap junctional coupling among axotomized adult motor neurons. J. Neurosci. 20:674, 2000. 56. Chao, M. V., and Bothwell, M.: Neurotrophins: to cleave or not to cleave. Neuron 3:9, 2002. 57. Chard, P. S., Bleakman, D., Savidge, J. R., and Miller, R. J.: Capsaicin-induced neurotoxicity in cultured dorsal root ganglion neurons: involvement of calcium-activated proteases. Neuroscience 65:1099, 1995. 58. Chimelli, L., and Scaravilli, F.: The abnormal development of the gracile nucleus in the neurological mutant rat mf. Dev. Brain Res. 29:193, 1986. 59. Cho, H. J., Kim, S. Y., Park, M. J., et al.: Expression of mRNA for brain-derived neurotrophic factor in the dorsal root ganglion following peripheral inflammation. Brain Res. 749:358, 1997. 60. Chou, S. M., and Payne, W. N.: Vacuolation and chromatolysis of lower motoneurons in tetanus. A case report and review of the literature. Cleve. Clin. Q. 49:255, 1982. 61. Chou, S. M., and Wang, H. S.: Aberrant glycosylation/phosphorylation in chromatolytic motoneurons of WerdnigHoffmann disease. J. Neurol. Sci. 152:198, 1997. 62. Choudhary, S., Joshi, K., and Gill, K. D.: Possible role of enhanced microtubule phosphorylation in dichlorvos induced delayed neurotoxicity in rat. Brain Res. 897:60, 2001. 63. Clarke, P. G., and Clarke, S.: Nineteenth century research on naturally occurring cell death and related phenomena. Anat. Embryol. 193:81, 1996. 64. Clarke, P. G. H.: Apoptosis versus necrosis: how valid a dichotomy for neurons? In Koliatsos, V. E., and Ratan, R. R. (eds.): Cell Death and Diseases of the Nervous System. Totowa, NJ, Humana Press, p. 3, 1999. 65. Clatterbuck, R. E., Price, D. L., and Koliatsos, V. E.: Further characterization of the effects of brain-derived neurotrophic factor and ciliary neurotrophic factor on axotomized neonatal and adult mammalian motor neurons. J. Comp. Neurol. 342:45, 1994. 66. Coggeshall, R. E.: A consideration of neural counting methods. TINS 15:9, 1992. 67. Colbere-Garapin, F., Duncan, G., Paviao, N., et al.: An approach to understanding the mechanisms of poliovirus persistence in infected cells of neural and non-neural origin. Clin. Diagn. Virol. 9:107, 1998. 68. Cordeiro, P., Hechtman, P., and Kaplan, F.: The GM2 gangliosidoses databases: allelic variation at the HEXA, HEXB, and GM2A gene loci. Genet. Med. 2:319, 2000. 69. Cork, L. C., Altschuler, R. J., Bruha, P. J., et al.: Changes in neuronal size and neurotransmitter marker in hereditary canine spinal muscular atrophy. Lab. Invest. 61:69, 1989. 70. Costa-Jussà, F. R., and Jacobs, J. M.: The pathology of amiodarone neurotoxicity. Brain 108:735, 1985. 71. Costigan, M., Mannion, R. J., Kendall, G., et al.: Heat shock protein 27: developmental regulation and expression after peripheral nerve injury. J. Neurosci. 18:5891, 1998. 72. Cottrell, D. F., McGorum, B. C., and Pearson, G. T.: The neurology and enterology of equine grass sickness: a review of basic mechanisms. Neurogastroenterol. Motil. 11:79, 1999.

719

73. Coulson, E. J., Reid, K., Barrett, G. L., and Bartlett, P. F.: p75 neurotrophin receptor-mediated neuronal death is promoted by Bcl-2 and prevented by Bcl-xL. J. Biol. Chem. 274:16387, 1999. 74. Coutinho, P., Guimarães, A., and Scaravilli, F.: The pathology of Machado-Joseph disease: report of a possible homozygous case. Acta Neuropathol. (Berl.) 58:48, 1982. 75. Croker, A. C., and Farber, S.: Niemann-Pick disease: a review of eighteen patients. Medicine 37:1, 1958. 76. Cowan, T., Johnson, R. J., Soubeyre, V., and Santer, R. M.: Restricted diet rescues rat enteric motor neurons from age related cell death. Gut 47:653, 2000. 77. Coward, K., Jowett, A., Plumpton, C., et al.: Sodium channel beta1 and beta2 subunits parallel SNS/PN3 alphasubunit changes in injured human sensory neurons. Neuroreport 12:483, 2001. 78. Cummings, J. F., de Lahunta, A., and Winn, S. S.: Acral mutilation and nociceptive loss in English Pointer dogs. Acta Neuropathol. (Berl.) 53:119, 1981. 79. da Veiga Pereira, L., Desnick, R. J., Adler, D. A., et al.: Regional assignment of the human acid sphingomyelinase gene (SMPD1) by PCR analysis of somatic cell hybrids and in situ hybridisation to 11p15.1-p15.4. Genomics 9:229, 1991. 80. Dalmau, J., Graus, F., Rosenblum, M. K., and Posner, J. B.: Anti-Hu-associated paraneoplastic encephalomyelitis/sensory neuropathy: a clinical study of 71 patients. Medicine 71:59, 1992. 81. Dalpe, G., Leclerc, N., Vallee, A., et al.: Dystonin is essential for maintaining neuronal cytoskeleton organization. Mol. Cell. Neurosci. 10:243, 1998. 82. Daniels, R. J., Suthers, G. K., Morrison, K. E., et al.: Prenatal prediction of spinal muscular atrophy. J. Med. Genet. 29:165, 1992. 83. David, G., Abbas, N., Stevanin, G., et al.: Cloning of the SCA7 gene reveals a highly unstable CAG repeat expansion. Nat. Genet. 17:65, 1997. 84. Davies, J. P., Levy, B., and Ioannou, Y. A.: Evidence for a Niemann-Pick C (NPC) gene family: identification and characterization of NPC1L1. Genomics 65:137, 2000. 85. Dawson, V. L.: Nitric oxide: role in neurotoxicity. Clin. Exp. Pharmacol. Physiol. 22:305, 1995. 86. De Balogh, K. K., Dimande, A. P., van der Lugt, J. J., et al.: A lysosomal storage disease induced by Ipomoea carnea in goats in Mozambique. J. Vet. Diagn. Invest. 11:266, 1999. 87. De Bilbao, F., Giannakopoulos, P., Srinivasan, A., and Dubois-Dauphin, M.: In vivo study of motoneuron death induced by nerve injury in mice deficient in the caspase 1/interleukin-1 beta-converting enzyme. Neuroscience 98:573, 2000. 88. De Castro, F.: Sensory ganglia of the cranial and spinal nerves: normal and pathological. In Penfield, W. (ed.): Cytology and Cellular Pathology of the Nervous System. New York, Hoeber, p. 32, 1932. 89. Delio, D. A., Reuhl, K. R., and Lowndes, H. E.: Ectopic impulse generation in dorsal root ganglion neurons during methylmercury intoxication: an electrophysiological and morphological study. Neurotoxicology 13:527, 1992. 90. Denny-Brown, D., Adams, R. D., and Fitzgerald, P. J.: Pathologic features of herpes zoster: a note on “geniculate herpes.” AMA Arch. Neurol. Psychiatry 51:216, 1944.

720

Pathology of the Peripheral Nervous System

91. Dubois-Dalcq, M., Menu, R., and Buyse, M.: Influence of fatty acids on fine structure of cultured neurons: an experimental approach to Refsum’s disease. J. Neuropathol. Exp. Neurol. 31:645, 1972. 92. Duchen, L. W., Anjorin, S., Watkins, P. J., and McKay, J. D.: Pathology of autonomic neuropathy in diabetes. Ann. Intern. Med. 92:301, 1980. 93. Duchen, L. W., Eicher, E. M., Jacobs, J. M., et al.: Hereditary leukodystrophy in the mouse: the new mutant twitcher. Brain 103:695, 1980. 94. Duchen, L. W., and Scaravilli, F.: Quantitative and electron microscopic studies of sensory ganglion cells of the Sprawling mouse. J. Anat. 123:763, 1977. 95. Duchen, L. W., and Strich, S. J.: Clinical and pathological studies of an hereditary neuropathy in mice (dystonia musculorum). Brain 87:367, 1964. 96. Duchen, L. W., Strich, S. J., and Falconer, D. S.: An hereditary motor neurone disease with progressive denervation of muscle in the mouse: the mutant “wobbler”. J. Neurol. Neurosurg. Psychiatry 31:535, 1968. 97. Duffy, P. E., Kornfeld, M., and Suzuki, K.: Neurovisceral storage disease with curvilinear bodies. J. Neuropathol. Exp. Neurol. 27:351, 1968. 98. D’Urso, D., Ehrhardt, P., and Muller, H. W.: Peripheral myelin protein 22 and protein zero: a novel association in peripheral nervous system myelin. J. Neurosci. 19:3396, 1999. 99. Dyck, P. J., Dyck, P. J. B., and Schaid, D. J.: Genetic heterogeneity in hereditary sensory and autonomic neuropathies: the need for improved ascertainment. Muscle Nerve 23:1453, 2000. 100. Dykens, J. A.: Free radicals and mitochondria dysfunction in excitotoxicity and neurodegenerative disease. In Koliatsos, V. E., and Ratan, R. R. (eds.): Cell Death and Diseases of the Nervous System. Totowa, NJ, Humana Press, p. 45, 1999. 101. Edwards, S. N., Buckmaster, A. E., and Tolkowsky, A. M.: The death programme in cultured sympathetic neurons can be suppressed at the posttranslational level by nerve growth factor, cyclic AMP, and depolarization. J. Neurochem. 57:2140, 1991. 102. Edwards, S. N., and Tolkovsky, A. M.: Characterization of apoptosis in cultured rat sympathetic neurons after nerve growth factor withdrawal. J. Cell Biol. 124:537, 1994. 103. Eidelberg, D., Sotrel, A., Vogel, H., et al.: Progressive polyradiculopathy in acquired immune deficiency syndrome. Neurology 36:912, 1986. 104. Ekblad, E., Mulder, H., and Sundler, F.: Vasoactive intestinal peptide expression in enteric neurons is upregulated by both colchicine and axotomy. Regul. Pept. 63:113, 1996. 105. Ekström, P.: Neurones and glial cells of the mouse sciatic nerve undergo apoptosis after injury in vivo and in vitro. Neuroreport 6:1029, 1995. 106. Elleder, M., and Jirasek, A.: Niemann-Pick disease: report on a symposium held in Hlava’s Institute of Pathology, Charles University, Prague, September 1982. Acta Univ. Carol. Med. Praha 29:259, 1983. 107. Elliott, J. L., and Snider, W. D.: Axotomy-induced motor neuron death. In Koliatsos, V. E., and Ratan, R. R. (eds.): Cell Death and Diseases of the Nervous System. Totowa, NJ, Humana Press, p. 181, 1999.

108. Emery, J. L., and Singhal, R.: Changes associated with growth in the cells of the dorsal root ganglion in children. Dev. Med. Child Neurol. 15:460, 1973. 109. Eng, C. M., and Desnick, R. J.: Molecular basis of Fabry disease: mutations and polymorphisms in the human alphagalactosidase A gene. Hum. Mutat. 3:103, 1994. 110. Ernfors, P., Rosario, C. M., Merlio, J. P., et al.: Expression of mRNAs for neurotrophin receptors in the dorsal root ganglion and spinal cord during development and following peripheral or central axotomy. Mol. Brain Res. 17:217, 1993. 111. Esiri, M. M., Morris, C. S., and Millard, P. R.: Sensory and sympathetic ganglia in HIV infection: immunocytochemical demonstration of HIV-1 viral antigens, increased MHC class II antigens expression and mild reactive inflammation. J. Neurol. Sci. 114:178, 1993. 112. Estrada, R., Galaraga, J., Orozco, G., et al.: Spinocerebellar ataxia 2 (SCA2): morphometric analyses in 11 autopsies. Acta Neuropathol. (Berl.) 97:306, 1999. 113. Fadok, V. A., Voelker, D. R., Campbell, P. A., et al.: Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J. Immunol. 148:2207, 1992. 114. Falconer, D. S.: Wobbler (rat) mouse. Mouse Newslett. 15:23, 1956. 115. Fernandes, K. J., Fan, D. P., Tsui, B. J., et al.: Influence of the axotomy to cell body distance in rat rubrospinal and spinal motoneurons: differential regulation of GAP-43, tubulins, and neurofilament-M. J. Comp. Neurol. 414:495, 1999. 116. Fernandes, K. J., Kobayashi, N. R., Jasmin, B. J., and Tetzlaff, W.: Acetylcholinesterase gene expression in axotomized rat facial motoneurons is differentially regulated by neurotrophins: correlation with trkB and trkC mRNA levels and isoforms. J. Neurosci. 18:9936, 1998. 117. Fernyhough, P., Gallagher, A., and Averill, S. A.: Aberrant neurofilament phosphorylation in sensory neurons in rats with diabetic neuropathy. Diabetes 48:881, 1999. 118. Festoff, B. W., D’Andrea, M. R., Citron, B. A., et al.: Motor neuron cell death in wobbler mutant mice follows overexpression of the G-protein-coupled, protease-activated receptor for thrombin. Mol. Med. 6:410, 2000. 119. Finn, L. S., Zhang, M., Chen, S. H., and Scott, C. R.: Severe type II Gaucher disease with ichthyosis, arthrogryposis and neuronal apoptosis: molecular and pathological analyses. Am. J. Med. Genet. 91:222, 2000. 120. Fischer, S. J., McDonald, E. S., Gross, L., and Windebank, A. J.: Alterations in cell cycle regulation underlie cisplatin induced apoptosis of dorsal root ganglion neurons in vivo. Neurobiol. Dis. 8:1027, 2001. 121. Fitzgerald, M., Woolf, C. J., Gibson, S. J., and Mallaburn, P. S.: Alterations in the structure, function, and chemistry of C fibres following local application of vinblastine to the sciatic nerve of the rat. J. Neurosci. 4:430, 1984. 122. Flanigan, K. M., Crawford, T. O., Griffin, J. W., et al.: Localization of the giant axonal neuropathy gene to chromosome 16q24. Ann. Neurol. 43:143, 1998. 123. Fletcher, G. C., Xue, L., Passingham, S. K., and Tolkovsky, A. M.: Death commitment point is advanced by axotomy in sympathetic neurons. J. Cell Biol. 150:741, 2000.

Pathology of Peripheral Neuron Cell Bodies 124. Frade, J. M., Rodriguez-Tébar, A., and Barde, Y.-A.: Induction of cell death by endogenous nerve growth factor through its p75 receptor. Nature 383:166, 1996. 125. Frangione, B., Vidal, R., Rostagno, A., and Ghiso, J.: Familial cerebral amyloid angiopathies and dementia. Alzheimer Dis. Assoc. Disord. 14 Suppl. 1S:25, 2000. 126. Friede, R. L., and Johnstone, M. A.: Responses of thymidine labeling of nuclei in gray matter and nerve following sciatic transection. Acta Neuropathol. (Berl.) 7:218, 1967. 127. Friede, R. L., and Samorajski, T.: Axon caliber related to neurofilaments and microtubules in sciatic nerve fibers of rats and mice. Anat. Rec. 167:379, 1970. 128. Fukai, K., Oh, J., Karim, M. A., et al.: Homozygosity mapping of the gene for Chediak-Higashi syndrome to chromosome 1q42-q44 in a segment of conserved synteny that includes the mouse beige locus (bg). Am. J. Hum. Genet. 59:620, 1996. 129. Fuller, G. N., and Jacobs, J. M.: Peripheral nerve and muscle disease in HIV infection. In Scaravilli, F. (ed.): The Neuropathology of HIV Infection. London, Springer, p. 215, 1993. 130. Gabbiani, G., Gregory, A., and Baic, D.: Cadmium-induced selective lesions of sensory ganglia. Exp. Neurol. 26:498, 1967. 131. Gabella, G.: Structure of the Autonomic Nervous System. London, Chapman & Hall, 1976. 132. Gadoth, N., and Sandbank, U.: Involvement of dorsal root ganglia in Fabry’s disease. J. Med. Genet. 20:309, 1983. 133. Gambarelli, D., Hassoun, J., Pellissier, J. F., et al.: Giant axonal neuropathy: involvement of peripheral nerve, myenteric plexus and extra-neuronal area. Acta Neuropathol. (Berl.) 39:261, 1977. 134. Gardiner, R. M.: The molecular genetic basis of the neuronal ceroid lipofuscinoses. Neurol. Sci. 21:S15, 2000. 135. Gardner, M. B., Rashid, S., Klement, V., et al.: Lower motor neuron disease in wild mice caused by indigenous type C virus and search for a similar etiology in human amyotrophic lateral sclerosis. In Andrews, J. M., Johnson, R. T., and Brazier, M. B. (eds.): Amyotrophic Lateral Sclerosis. New York, Academic Press, p. 217, 1976. 136. Garen, P. D., Powers, J. M., Young, G. F., and Lee, V.: Neuronal intranuclear hyaline inclusion disease in a nine year old. Acta Neuropathol. (Berl.) 70:327, 1986. 137. Gavrieli, A., Sherman, Y., and Ben-Sasson, S. A.: Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol. 119:493, 1992. 138. Gilad, G. M., and Gilad, V. H.: Increased choline kinase activity in the rat superior cervical ganglion after axonal injury. Brain Res. 220:420, 1981. 139. Gilad, G. M., and Gilad, V. H.: Polyamine biosynthesis is required for survival of sympathetic neurons after axonal injury. Brain Res. 273:191, 1983. 140. Gilbert, G. J.: Herpes zoster ophthalmicus and delayed contralateral hemiparesis. JAMA 229:302, 1974. 141. Gilbert, M. R., Harding, B. L., Hoffman, P. N., et al.: Aluminum-induced neurofilamentous changes in cultured rat dorsal root ganglia explants. J. Neurosci. 12:1763, 1992. 142. Gill, J. S., Hobday, K. L., and Windebank, A. J.: Mechanism of suramin toxicity in stable myelinating dorsal root ganglion cultures. Exp. Neurol. 133:113, 1995.

721

143. Gill, J. S., and Windebank, A. J.: Cisplatin-induced apoptosis in rat dorsal root ganglion neurons is associated with attempted entry into the cell cycle. J. Clin. Invest. 101:2842, 1998. 144. Gillardon, F., Klimaschewski, L., Wickert, H., et al.: Expression pattern of candidate cell death effector proteins Bax, Bcl-2, Bcl-X, and c-Jun in sensory and motor neurons following sciatic nerve transection in the rat. Brain Res. 739:244, 1996. 145. Gilles, G. H., and Deuel, R. K.: Neuronal cytoplasmic globules in the brain in Morquio’s syndrome. Arch. Neurol. 25:393, 1971. 146. Giometto, B., Scaravilli, T., Nicolao, P., et al.: Detection of paraneoplastic anti-neuronal autoantibodies on paraffinembedded tissues. Acta Neuropathol. (Berl.) 92:435, 1996. 147. Glenn, D., Gelbart, T., and Beutler, E.: Tight linkage of pyruvate kinase (PKLR) and glucocerebrosidase (GBA) genes. Hum. Genet. 93:635, 1994. 148. Gloster, A., Wu, W., Speelman, A., et al.: The T alpha 1 alpha-tubulin promoter specifies gene expression as a function of neuronal growth and regeneration in transgenic mice. J. Neurosci. 14:7319, 1994. 149. Glynn, P., Holton, J. L., Nolan, C. C., et al.: Neuropathy target esterase: immunolocalization to neuronal cell bodies and axons. Neuroscience 83:295, 1998. 150. Goebel, H. H., Kohlschutter, A., and Lenard, H. G.: Morphologic and chemical biopsy findings in mucolipidosis IV. Clin. Neuropathol. 1:73, 1982. 151. Gold, M. S., Reichling, D. B., Hampl, K. F., et al.: Lidocaine toxicity in primary afferent neurons from the rat. J. Pharmacol. Exp. Ther. 285:413, 1998. 152. Goldstein, M. E., Cooper, H. S., Bruce, J., et al.: Phosphorylation of neurofilament proteins and chromatolysis following transection of rat sciatic nerve. J. Neurosci. 7:1586, 1987. 153. Gotz, J., Tolnay, M., Barmettler, R., et al.: Oligodendroglial tau filament formation in transgenic mice expressing G272V tau. Eur. J. Neurosci. 13:2131, 2001. 154. Grafstein, B.: The nerve cell body response to axotomy. Exp. Neurol. 48:32, 1975. 155. Green, S. L., Vulliet, P. R., Pinter, M. J., and Cork, L. C.: Alterations in cyclin-dependent protein kinase 5 (CDK5) protein levels, activity and immunocytochemistry in canine motor neuron disease. J. Neuropathol. Exp. Neurol. 57:1070, 1998. 156. Greenbaum, D., Richardson, P. C., Salmon, M. V., and Urich, H.: Pathological observations on six cases of diabetic neuropathy. Brain 87:201, 1964. 157. Greenberg, S. G., and Lasek, R. J.: Neurofilament protein synthesis in DRG neurons decreases more after peripheral axotomy than after central axotomy. J. Neurosci. 8:1739, 1988. 158. Greensmith, L., Hasan, H. I., and Vrbova, G.: Nerve injury increases the susceptibility of motoneurons to N-methyl-Daspartate-induced neurotoxicity in the developing rat. Neuroscience 58:727, 1994. 159. Greer, W. L., Riddell, D. C., Byers, D. M., et al.: Linkage of Niemann-Pick disease type D to the same region of human chromosome 18 as Niemann-Pick disease type C. Am. J. Hum. Genet. 61:139, 1997.

722

Pathology of the Peripheral Nervous System

160. Griffin, J. W., Cornblath, D. R., Alexander, E., et al.: Ataxic sensory neuropathy and dorsal root ganglionitis associated with Sjögren’s syndrome. Ann. Neurol. 27:304, 1990. 161. Griffin, J. W., Li, C. Y., Ho, T. W., et al.: Pathology of the motor-sensory axonal Guillain-Barré syndrome. Ann. Neurol. 39:17, 1996. 162. Groves, M. J., An, S.-F., Giometto, B., and Scaravilli, F.: Inhibition of sensory neuron apoptosis and prevention of loss by NT-3 administration following axotomy. Exp. Neurol. 155:284, 1999. 163. Groves, M. J., Christopherson, T., Giometto, B., and Scaravilli, F.: Axotomy-induced apoptosis in adult rat sensory neurons. J. Neurocytol. 26:615, 1997. 164. Groves, M. J., Giometto, B. and Scaravilli, F.: Axotomyinduced vacuolation of primary sensory neurons and effect of administered neurotrophic factors: a morphometric, immunocytochemical and ultrastructural study. Prim. Sensory Neuron 2:111, 1997. 165. Groves, M. J., Ng, Y.-W., Ciardi, A., and Scaravilli, F.: Sciatic nerve injury in the adult rat: comparison of effects on oligosaccharide, CGRP, and GAP43 immunoreactivity in primary afferents following two types of trauma. J. Neurocytol. 26:219, 1996. 166. Guegan, C., Vila, M., Rosoklija, G., et al.: Recruitment of the mitochondrial-dependent apoptotic pathway in amyotrophic lateral sclerosis. J. Neurosci. 21:6569, 2001. 167. Gunasekar, P. G., Kanthasamy, A. G., Borowitz, J. L., and Isom, G. E.: NMDA receptor activation produces concurrent generation of nitric oxide and reactive oxygen species: implication for cell death. J. Neurochem. 65:2016, 1995. 168. Guntinas-Lichius, O., Schulte, E., Stennert, E., and Neiss, W. F.: The use of texture analysis to study the time course of chromatolysis. J. Neurosci. Methods 78:1, 1997. 169. Guo, A., Vulchanova, L., Wang, J., et al.: Immunocytochemical localization of the vanilloid receptor 1 (VR1): relationship to neuropeptides, the P2X3 purinoceptor and IB4 binding sites. Eur. J. Neurosci. 11:946, 1999. 170. Haas, C. A., Hofmann, H. D., and Kirsch, M.: Expression of CNTF/LIF-receptor components and activation of STAT3 signaling in axotomized facial motoneurons: evidence for a sequential postlesional function of the cytokines. J. Neurobiol. 41:559, 1999. 171. Hafer-Macko, C., Hsieh, S.-T., Li, C. Y., et al.: Acute motor axonal neuropathy: an antibody-mediated attack on axolemma. Ann. Neurol. 40:635, 1996. 172. Hagberg, B., Haltia, M., Sourander, P., et al.: Neurovisceral storage disorder simulating Niemann-Pick disease: a new form of oligosaccharidosis? Neuropädiatrie 9:59, 1978. 173. Hahn, C. N., del Pilar Martin, M., Schröder, M., et al.: Generalised CNS disease and massive GM1-ganglioside accumulation in mice defective in lysosomal acid ␤gangliosidase. Hum. Mol. Genet. 6:205, 1997. 174. Hahn, C. N., Mayhew, I. G., and de Lahunta, A.: Central neuropathology of equine grass sickness. Acta Neuropathol. (Berl.) 102:153, 2001. 175. Hammarberg, H., Piehl, F., Cullheim, S., et al.: GDNF mRNA in Schwann cells and DRG satellite cells after chronic sciatic nerve injury. Neuroreport 7:857, 1996. 176. Hammarberg, H., Piehl, F., Risling, M., and Cullheim, S.: Differential regulation of trophic factor receptor mRNAs in

177.

178.

179.

180.

181.

182. 183.

184.

185.

186.

187.

188. 189.

190.

191. 192.

spinal motoneurons after sciatic nerve transection and ventral root avulsion in the rat. J. Comp. Neurol. 426:587, 2000. Harvey, J. S., Davis, C., James, I. F., and Burgess, G. M.: Activation of protein kinase C by the capsaicin analogue resiniferatoxin in sensory neurones. J. Neurochem. 65:1309, 1995. Haust, M. D., and Gordon, B. A.: Ultrastructural and biochemical aspects of the Sanfilippo syndrome—type III genetic mucopolysaccharidosis. Connect. Tissue Res. 15:57, 1986. Hayashi, T., Sakurai, M., Abe, K., et al.: Apoptosis of motor neurons with induction of caspases in the spinal cord after ischemia. Stroke 29:1007, 1998. Helliwell, R. J., McLatchie, L. M., Clarke, M., et al.: Capsaicin sensitivity is associated with the expression of the vanilloid (capsaicin) receptor (VR1) mRNA in adult rat sensory ganglia. Neurosci. Lett. 258:77, 1998. Hendry, I. A., and Campbell, J.: Morphometric analysis of rat superior cervical ganglion after axotomy and nerve growth factor treatment. J. Neurocytol. 5:351, 1976. Henson, R. A., and Urich, H.: Cancer and the Nervous System. London, Blackwell Scientific, 1982. Herdegen, T., Fiallos-Estrada, C. E., Schmid, W., et al.: The transcription factors c-JUN, JUN D and CREB, but not FOS and KROX-24, are differentially regulated in axotomized neurons following transection of the rat sciatic nerve. Brain Res. Mol. Brain Res. 14:155, 1992. Herdegen, T., and Leah, J. D.: Inducible and constitutive transcription factors in the mammalian nervous system: control of gene expression by Jun, Fos, and Krox, and KREB/ATF proteins. Brain Res. Brain Res. Rev. 28:370, 1998. Hermans, M. M., de Graaff, E., Kroos, M. A., et al.: Identification of a point mutation in the human lysosomal alpha-glucosidase gene causing infantile glycogenosis type II. Biochem. Biophys. Res. Commun. 179:919, 1991. Hermans, M. M., Kroos, M. A., de Graaff, E., et al.: Two mutations affecting the transport and maturation of lysosomal alpha-glucosidase in an adult case of glycogen storage disease type II. Hum. Mutat. 2:268, 1993. Herrick, M. K., Twiss, J. L., Vladutiu, G. D., et al.: Concomitant branching enzyme and phosphorylase deficiencies: an unusual glycogenosis with extensive neuronal polyglucosan storage. J. Neuropathol. Exp. Neurol. 53:239, 1994. Hierons, R.: Changes in the nervous system in acute porphyria. Brain 80:176, 1957. Himes, B. T., and Tessler, A.: Death of some dorsal root ganglion neurons and plasticity of others following sciatic nerve section in adult and neonatal rats. J. Comp. Neurol. 284:215, 1989. Hirosawa, K.: Electron microscopic studies on pigment granules in the substantia nigra and locus coeruleus in Japanese monkey (Macaca fuscate yakui). Z. Zellforsch. Mikrosk. Anat. 88:187, 1968. Hiura, A.: Neuroanatomical effects of capsaicin on the primary afferent neurons. Arch. Histol. Cytol. 63:199, 2000. Hiura, A., and Ishizuka, H.: Changes in features of degenerating primary sensory neurons with time after capsaicin treatment. Acta Neuropathol. (Berl.) 78:35, 1989.

Pathology of Peripheral Neuron Cell Bodies 193. Hoffman, P. N., Cleveland, D. W., Griffin, J. W., et al.: Neurofilament gene expression: a major determinant of axonal caliber. Proc. Natl. Acad. Sci. U. S. A. 84:3472, 1987. 194. Hol, E. M., Schwaiger, F. W., Werner, A., et al.: Regulation of the LIM-type homeobox gene islet-1 during neuronal regeneration. Neuroscience 88:917, 1999. 195. Holmberg, M., Duyckaerts, C., Dürr, A., et al.: Spinocerebellar ataxia type 7 (SCA7): a neurodegenerative disorder with neuronal intranuclear inclusions. Hum. Mol. Genet. 7:913, 1998. 196. Holmes, F. E., and Haynes, L. W.: Superactivation of transglutaminase type 2 without change in enzyme level occurs during progressive neurodegeneration in the mnd mouse mutant. Neurosci. Lett. 213:185, 1996. 197. Horwich, M. S., Cho, L., Porro, R. S., and Posner, J. B.: Subacute sensory neuropathy: a remote effect of carcinoma. Ann. Neurol. 2:7, 1977. 198. Hou, X. E., Lundmark, K., and Dahlstrom, A. B.: Cellular reactions to axotomy in rat superior cervical ganglia include apoptotic cell death. J. Neurocytol. 27:441, 1998. 199. Houenou, L. J., Li, L., Lo, A. C., et al.: Naturally occurring and axotomy-induced motoneuron death and its prevention by neurotrophic agents: a comparison between chick and mouse. Prog. Brain Res. 102:217, 1994. 200. Houlden, H., King, R. H., Hashemi-Nejad, A., et al.: A novel TRK A (NTRK1) mutation associated with hereditary sensory and autonomic neuropathy type V. Ann. Neurol. 49:521, 2001. 201. Hozumi, I., Nishizawa, M., Ariga, T., et al.: Accumulation of glycosphingolipids in spinal and sympathetic ganglia of a symptomatic heterozygote of Fabry’s disease. J. Neurol. Sci. 90:273, 1989. 202. Hu, P., and McLachlan, E. M.: Macrophage and lymphocyte invasion of dorsal root ganglia after peripheral nerve lesions in the rat. Neuroscience 112:23, 2002. 203. Huang, J. Q., Trasler, J. M., Igdoura, S., et al.: Apoptotic cell death in mouse models of GM2 gangliosidosis and observations on human Tay-Sachs and Sandhoff diseases. Hum. Mol. Genet. 6:1879, 1997. 204. Hughes, J. T., Brownell, B., and Hewer, R. L.: The peripheral sensory pathway in Friedreich’s ataxia: an examination by light and electron microscopy of the posterior nerve roots, posterior root ganglia and peripheral sensory nerves in cases of Friedreich’s ataxia. Brain 91:803, 1968. 205. Huxtable, C. R., and Dorling, P. R.: Mannoside storage and axonal dystrophy in sensory neurones of swainsoninetreated rats: morphogenesis of lesions. Acta Neuropathol. (Berl.) 68:65, 1985. 206. Inaishi, Y., Kashihara, Y., Sakaguchi, M., et al.: Cooperative regulation of calcitonin gene-related peptide levels in rat sensory neurons via their central and peripheral processes. J. Neurosci. 12:518, 1992. 207. Indo, Y., Tsuruta, M., Hayashida, Y., et al.: Mutations in the TRKA/NGF receptor gene in patients with congenital insensitivity to pain with anhidrosis. Nat. Genet. 13:485, 1996. 208. Iwahashi, Y., Furuyama, T., Inagaki, S., et al.: Distinct regulation of sodium channel types I, II and III following nerve transection. Brain Res. Mol. Brain Res. 22:341, 1994. 209. Jaatinen, P., and Hervonen, A.: Reactions of rat sympathetic neurons to ethanol exposure are age-dependent. Neurobiol. Aging 15:419, 1994.

723

210. Jaatinen, P., Riihioja, P., Haapalinna, A., et al.: Prevention of ethanol-induced sympathetic overactivity and degeneration by dexmedetomidine. Alcohol 12:439, 1995. 211. Jacob, J. M.: Lumbar motor neuron size and number is affected by age in male F344 rats. Mech. Ageing Dev. 106:205, 1998. 212. Jacobs, J. M., Carmichael, N., and Cavanagh, J. B.: Ultrastructural changes in the dorsal root and trigeminal ganglia of rats poisoned with methyl mercury. Neuropathol. Appl. Neurobiol. 1:1, 1975. 213. Jacobs, J. M., Carmichael, N., and Cavanagh, J. B.: Ultrastructural changes in the nervous system of rabbits poisoned with methyl mercury. Toxicol. Appl. Pharmacol. 39:249, 1977. 214. Jacobs, J. M., Scaravilli, F., Duchen, L. W., and Mertin, J.: A new neurological rat mutant “mutilated foot.” J. Anat. 132:525, 1981. 215. Jarvela, I., Vesa, J., Santavuori, P., et al.: Molecular genetics of neuronal ceroid lipofuscinoses. Pediatr. Res. 32:645, 1992. 216. Jellinger, K., Paulus, W., Grisold, W., and Paschke, E.: New phenotype of adult alpha-L-iduronidase deficiency (mucopolysaccharidosis I) masquerading as Friedreich’s ataxia with cardiopathy. Clin. Neuropathol. 9:163, 1990. 217. Jessell, T. M., Tsunoo, A., Kanazawa, I., and Otsuka, M.: Substance P: depletion in the dorsal horn of rat spinal cord after section of the peripheral processes of primary sensory neurons. Brain Res. 168:247, 1979. 218. Ji, R.-R., Zhang, Q., Zhang, X., et al.: Prominent expression of bFGF in dorsal root ganglia after axotomy. Eur. J. Neurosci. 7:2458, 1995. 219. Johnson, P. C., Roluk, L. A., Hamilton, R. H., and Laguna, J. F.: Paraneoplastic vasculitis of nerve: a remote effect of cancer. Ann. Neurol. 5:437, 1979. 220. Jones, H. B., and Cavanagh, J. B.: Cytoplasmic vacuoles. Ultrastruct. Pathol. 6:359, 1984. 221. Jones, K. J., and Lavelle, A.: Differential effects of axotomy on immature and mature hamster facial neurons: a time course study of initial nucleolar and nuclear changes. J. Neurocytol. 15:197, 1986. 222. Kami, K., Morikawa, Y., Kawai, Y., and Senba, E.: Leukemia inhibitory factor, glial cell line-derived neurotrophic factor, and their receptor expressions following muscle crush injury. Muscle Nerve 22:1576, 1999. 223. Karanth, S. S., Springall, D. R., Kar, S., et al.: Time-related decrease of substance P and CGRP in central and peripheral projections of sensory neurons in Mycobacterium leprae infected nude mice: a model for lepromatous leprosy in man. J. Pathol. 160:335, 1990. 224. Kashiba, H., Hyon, B., and Senba, E.: Glial cell line-derived neurotrophic factor and nerve growth factor mRNAs are expressed in distinct subgroups of dorsal root ganglion neurons and are differentially regulated by peripheral axotomy in the rat. Neurosci. Lett. 252:107, 1998. 225. Kashiba, H., Senba, E., Yoshihiro, U., and Tohyama, M.: Co-localized but target-unrelated expression of vasoactive intestinal polypeptide and galanin in rat dorsal root ganglion neurons after peripheral nerve crush injury. Brain Res. 582:47, 1992. 226. Kashihara, Y., Kuno, M., and Miyata, Y.: Cell death of axotomized motoneurons in neonatal rats, and its prevention

724

227.

228.

229.

230.

231.

232.

233.

234.

235.

236.

237.

238. 239.

240.

241. 242. 243.

244.

Pathology of the Peripheral Nervous System by peripheral reinnervation. J. Physiol. (Lond.) 386:135, 1987. Kato, T., Hirano, A., Weinberg, M. N., and Jacobs, A. K.: Spinal cord lesions in progressive supranuclear palsy: some new observations. Acta Neuropathol. (Berl.) 71:11, 1986. Katsetos, C. D., Savory, J., Herman, M. M., et al.: Neuronal cytoskeletal lesions induced in the CNS by intraventricular and intravenous aluminium maltol in rabbits. Neuropathol. Appl. Neurobiol. 16:511, 1990. Kawasaki, H., Murayama, S., Tomonaga, M., et al.: Neurofibrillary tangles in human upper cervical ganglia: morphological study with immunohistochemistry and electron microscopy. Acta Neuropathol. (Berl.) 75:156, 1987. Kennedy, P. G., Grinfeld, E., and Gow, J. W.: Latent varicella-zoster virus is located predominantly in neurons in human trigeminal ganglia. Proc. Natl. Acad. Sci. U. S. A. 95:4658, 1998. Kerenyi, T., Havanhyl, L., and Huttner, I.: Investigations on experimentally produced age-pigment in the nervous system. Exp. Gerontol. 3:155, 1968. Kerezoudi, E., King, R. H. M., Muddle, J. R., et al.: Influence of age on the retrograde effects of sciatic nerve section in the rat. J. Anat. 187:27, 1995. Kerezoudi, E., and Thomas, P. K.: Influence of age on regeneration in the peripheral nervous system. Gerontology 45:301, 1999. Kerr, J. F. R., Wyllie, A. H., and Currie, A. R.: Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br. J. Cancer 26:239, 1972. Kikuchi, H., Doh-ura, K., Kira, J., and Iwaki, T.: Preferential neurodegeneration in the cervical spinal cord of progressive supranuclear palsy. Acta Neuropathol. (Berl.) 97:577, 1999. Kim, D. S., Yoon, C. H., Lee, S. J., et al.: Changes in voltage-gated calcium channel alpha(1) gene expression in rat dorsal root ganglia following peripheral nerve injury. Brain Res. Mol. Brain Res. 96:151, 2001. Kimber, T. E., Blumbergs, P. C., Rice, J. P., et al.: Familial neuronal intranuclear inclusion disease with ubiquitin positive inclusions. J. Neurol. Sci. 160:33, 1998. Kint, J. A.: Fabry’s disease: ␣-galactosidase deficiency. Science 167:1268, 1970. Klimaschewski, L., Obermuller, N., Majewski, M., et al.: Increased expression of nitric oxide synthase in a subpopulation of rat sympathetic neurons after axotomy—correlation with vasoactive intestinal peptide. Cell Tissue Res. 285:419, 1996. Kobayashi, N. R., Bedard, A. M., Hincke, M. T., and Tetzlaff, W.: Increased expression of BDNF and trkB mRNA in rat facial motoneurons after axotomy. Eur. J. Neurosci. 8:1018, 1996. Koliatsos, V. E., and Ratan, R. R.: Cell Death and Diseases of the Nervous System. Totowa, NJ, Humana Press, 1999. Koltzenburg, M., Wall, P. D., and McMahon, S. B.: Does the right side know what the left is doing? TINS 22:122, 1999. Koshy, B. T., and Zoghbi, H. Y.: The CAG/polyglutamine tract diseases: gene products and molecular pathogenesis. Brain Pathol. 7:927, 1997. Koto, A., Horwitz, A. L., Suzuki, K., and Tiffany, C. W.: The Morquio syndrome: neuropathology and biochemistry. Ann. Neurol. 4:26, 1978.

245. Koyano, S., Uchihara, T., Fujigasaki, H., et al.: Neuronal intranuclear inclusions in spinocerebellar ataxia type 2: triple labelling immunofluorescent study. Neurosci. Lett. 273:117, 1999. 246. Krarup-Hansen, A., Rietz, B., Krarup, C., et al.: Histology and platinum content of sensory ganglia and sural nerves in patients treated with cisplatin and carboplatin: an autopsy study. Neuropathol. Appl. Neurobiol. 25:29, 1999. 247. Krebs, C., Neiss, W. F., Streppel, M., et al.: Axotomy induces transient calbindin D28K immunoreactivity in hypoglossal motoneurons in vivo. Cell Calcium 22:367, 1997. 248. Kristensson, K., and Olsson, Y.: Retrograde transport of horseradish peroxidase in transected axons. 1. Time relationships between transport and induction of chromatolysis. Brain Res. 79:101, 1974. 249. Kroemer, G., Zamzami, N., and Susin, S. A.: Mitochondrial control of apoptosis. Immunol. Today 18:44, 1997. 250. Kuemmel, T. A., Schroeder, R., and Stoffel, W.: Light and electron microscopic analysis of the central and peripheral nervous systems of acid sphingomyelinase-deficient mice resulting from gene targeting. J. Neuropathol. Exp. Neurol. 56:171, 1997. 251. La Spada, A. R., Paulson, H. L., and Fishbeck, K. H.: Trinucleotide repeat expansion in neurological disease. Ann. Neurol. 36:814, 1994. 252. La Spada, A. R., Wilson, E. M., Lubahn, D. B., et al.: Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352:77, 1991. 253. Lake, B.: Lysosomal and peroxisomal disorders. In Graham, D. I., and Lantos, P. L. (eds.): Greenfield’s Neuropathology. London, Arnold, p. 657, 1997. 254. LaMotte, C. C., and Kapadia, S. E.: Deafferentationinduced alterations in the rat dorsal horn: II. Effects of selective poisoning by pronase of the central processes of a peripheral nerve. J. Comp. Neurol. 266:198, 1987. 255. Lawson, S. J., and Lowrie, M. B.: The role of apoptosis and excitotoxicity in the death of spinal motoneurons and interneurons after neonatal nerve injury. Neuroscience 87:337, 1998. 256. Lawson, S. N.: Morphological and biochemical types of sensory neurons. In Scott, S. A. (ed.): Sensory Neurons. Oxford, UK, Oxford University Press, p. 27, 1992. 257. Lee, M., Hyun, D. H., Halliwell, B., and Jenner, P.: Effect of overexpression of wild-type and mutant Cu/Zn-superoxide dismutases on oxidative stress and cell death induced by hydrogen peroxide, 4-hydroxynonenal or serum deprivation: potentiation of injury by ALS-related mutant superoxide dismutases and protection by Bcl-2. J. Neurochem. 78:209, 2001. 258. Lee, P., Zhuo, H., and Helke, C. J.: Axotomy alters neurotrophin and neurotrophin receptor mRNAs in the vagus nerve and nodose ganglion of the rat. Brain Res. Mol. Brain Res. 87:31, 2001. 259. Lewis, S. E., Mannion, R. J., White, F. A., et al.: A role for HSP27 in sensory neuron survival. J. Neurosci. 19:8945, 1999. 260. Lhermitte, F., Gray, F., Lyon-Caen, O., et al.: Paralysie du tube digestif avec lesions des plexus myentériques: nouveau syndrome paranéoplasique. Rev. Neurol. (Paris) 136:825, 1980.

Pathology of Peripheral Neuron Cell Bodies 261. Li, L., Houenou, L. J., Wu, W., et al.: Characterization of spinal motoneuron degeneration following different types of peripheral nerve injury in neonatal and adult mice. J. Comp. Neurol. 396:158, 1998. 262. Liberski, P. P., Buczynski, J., Yanagihara, R., et al.: Ultrastructural pathology of a Chilean case of tropical spastic paraparesis/human T-cell lymphotropic type 1-associated myelopathy (TSP/HAM). Ultrastruct. Pathol. 23:157, 1999. 263. Lieberman, A. R.: The axon reaction: a review of the principal features of perikaryal responses to axon injury. Int. Rev. Neurobiol. 14:49, 1971. 264. Lieberman, A. R.: Some factors affecting retrograde neuronal responses to axonal lesions. In Bellairs, R., and Gray, E. G. (eds.): Essays on the Nervous System. Oxford, UK, Clarendon Press, p. 71, 1974. 265. Lieberman, A. R.: Sensory ganglia. In Landon, D. N. (ed.): The Peripheral Nerve. London, Chapman and Hall, p. 188, 1976. 266. Liu, R. Y., and Snider, W. D.: Different signaling pathways mediate regenerative versus developmental sensory axon growth. J. Neurosci. 21:RC164, 2001. 267. Liu, Z., and Martin, L. J.: Motor neurons rapidly accumulate DNA single-strand breaks after in vitro exposure to nitric oxide and peroxynitrite and in vivo axotomy. J. Comp. Neurol. 432:35, 2001. 268. Ljungberg, C., Novikov, L., Kellerth, J. O., et al.: The neurotrophins NGF and NT-3 reduce sensory neuronal loss in adult rat after peripheral nerve lesion. Neurosci. Lett. 262:29, 1999. 269. Low, P. A.: Hereditary hypertrophic neuropathy in the trembler mouse. Part 1. Histopathological studies: light microscopy. J. Neurol. Sci. 30:327, 1976. 270. Lu, X., and Richardson, P. M.: Inflammation near the nerve cell body enhances axonal regeneration. J. Neurosci. 11:972, 1991. 271. Lungu, O., Annunziato, P. W., Gershon, A., et al.: Reactivated and latent varicella-zoster virus in human dorsal root ganglia. Proc. Natl. Acad. Sci. U. S. A. 92:10980, 1995. 272. Luo, X. G., Zhou, X. F., and Rush, R. A.: Ultrastructural changes of sympathetic neurons following neurotrophin-3 antiserum treatment in young rat. Exp. Neurol. 147:401, 1997. 273. Lurie, I. W., Supowitz, K. R., Rosenblum-Vos, L. S., and Wulfsberg, E. A.: Phenotypic variability of del (2) (q22-q23): report of a case with a review of the literature. Genet. Couns. 5:11, 1994. 274. Mahalingam, R., Wellish, M., Cohrs, R., et al.: Expression of protein encoded by varicella-zoster virus open reading frame 63 in latently infected human ganglionic neurons. Proc. Natl. Acad. Sci. U. S. A. 93:2122, 1996. 275. Majno, G., and Joris, I.: Apoptosis, oncosis, and necrosis: an overview of cell death. Am. J. Pathol. 146:3, 1995. 276. Mancall, E. L., Aponte, G. E., and Berry, R. G.: Pompe’s disease (diffuse glycogenosis) with neuronal storage. J. Neuropathol. Exp. Neurol. 24:85, 1965. 277. Mariotti, R., Peng, Z.-C., Kristensson, K., and Bentivoglio, M.: Age-dependent induction of nitric oxide synthase activity in facial motoneurons after axotomy. Exp. Neurol. 145:361, 1997.

725

278. Martin, D. P., Schmidt, R. E., DiStefano, P. S., et al.: Inhibitors of protein synthesis and RNA synthesis prevent neuronal death caused by nerve growth factor deprivation. J. Cell Biol. 106:829, 1988. 279. Martin, J. J., Leroy, J. G., van Eygen, M., and Ceuterick, C.: I-cell disease: a further report on its pathology. Acta Neuropathol. (Berl.) 64:234, 1984. 280. Martin, J. J., Van Regemorter, N., Krols, L., et al.: On an autosomal dominant form of retinal-cerebellar degeneration: an autopsy study of five patients in one family. Acta Neuropathol. (Berl.) 88:277, 1994. 281. Martin, L. J.: Neuronal death in amyotrophic lateral sclerosis is apoptosis: possible contribution of a programmed cell death mechanism. J. Neuropathol. Exp. Neurol. 58:459, 1999. 282. Martin, L. J., Al-Abdulla, N. A., Brambrink, A. M., et al.: Neurodegeneration in excitotoxicity, global cerebral ischemia, and target deprivation: a perspective on the contributions of apoptosis and necrosis. Brain Res. Bull. 46:281, 1998. 283. Martin, L. J., Kaiser, A., and Price, A. C.: Motor neuron degeneration after sciatic nerve avulsion in adult rat evolves with oxidative stress and is apoptosis. J. Neurobiol. 40:185, 1999. 284. Martin, L. J., and Liu, Z.: Injury-induced spinal motor neuron apoptosis is preceded by DNA single-strand breaks and is p53- and Bax-dependent. J. Neurobiol. 50:181, 2002. 285. Martin, L. J., Price, A. C., Kaiser, A., et al.: Mechanisms for neuronal degeneration in amyotrophic lateral sclerosis and in models of motor neuron death [review]. Int. J. Mol. Med. 5:3, 2000. 286. Matsuda, J., Suzuki, O., Oshima, A., et al.: Neurological manifestations of knockout mice with b-galactosidase deficiency. Brain Dev. 19:19, 1997. 287. Matsuishi, T., Teresawa, K., Yoshida, I., et al.: Vacuolar myopathy with type 2 fibre atrophy and type 2B fibre deficiency: a case of childhood form of acid ␣-1,4-glucosidase deficiency. Neuropediatrics 13:173, 1982. 288. Mazliah, J., Barron, S., Bental, E., et al.: The effects of longterm lead intoxication on the nervous system of the chicken. Neurosci. Lett. 101:253, 1989. 289. Mazurkiewicz, J. E., Callahan, L. M., Swash, M., et al.: Cytoplasmic inclusions in spinal neurons of the motor neuron degeneration (mnd) mouse. I. Light microscopic analysis. J. Neurol. Sci. 116:59, 1993. 290. McKay-Hart, A., Brannstrom, T., Wiberg, M., and Terenghi, G.: Primary sensory neurons and satellite cells after peripheral axotomy in the adult rat. Exp. Brain Res. 142:308, 2002. 291. Meller, K.: Chromatolysis of dorsal root ganglion cells studied by cryofixation. Cell Tissue Res. 256:283, 1989. 292. Mentis, G. Z., Greensmith, L., and Vrbova, G.: Motoneurons destined to die are rescued by blocking N-methyl-D-aspartate receptors by MK-801. Neuroscience 54:283, 1993. 293. Messer, A., and Flaherty, L.: Autosomal dominance in a lateonset motor neuron disease in the mouse. J. Neurogenet. 3:345, 1986. 294. Messer, A., and Plummer, J.: Accumulating autofluorescent material as a marker for early changes in the spinal cord of the mnd mouse. Neuromuscul. Disord. 3:129, 1993. 295. Michael, G. J., Averill, S., Shortland, P. J., et al.: Axotomy results in major changes in BDNF expression by dorsal root

726

296.

297.

298.

299.

300.

301. 302.

303.

304.

305.

306.

307.

308.

309.

310.

Pathology of the Peripheral Nervous System ganglion cells: BDNF expression in large trkB and trkC cells, in pericellular baskets, and in projections to deep dorsal horn and dorsal column nuclei. Eur. J. Neurosci. 11:3539, 1999. Michaud, J., and Gilbert, J. J.: Multiple system atrophy with neuronal intranuclear hyaline inclusions: report of a new case with light and electron microscopic studies. Acta Neuropathol. (Berl.) 54:113, 1981. Misu, K., Hattori, N., Nagamatsu, M., et al.: Late onset familial amyloid polyneuropathy type 1 (transthyretin Met 30-associated familial amyloid polyneuropathy) unrelated to endemic focus in Japan: clinicopathological and genetic features. Brain 122:1951, 1999. Miyagishi, T., Takahata, N., and Izuka, R.: Electron microscopic studies on the lipopigments in the cerebral cortex nerve cells of senile and vitamin E deficient rats. Acta Neuropathol. (Berl.) 9:7, 1967. Mohiuddin, L., Delcroix, J. D., Fernyhough, P., and Tomlinson, D. R.: Focally administered nerve growth factor suppresses molecular regenerative responses of axotomized peripheral afferents in rats. Neuroscience 91:265, 1999. Mohseni, S.: Hypoglycaemic neuropathy in diabetic BB/Wor rats treated with insulin implants affected ventral root axons but not dorsal root axons. Acta Neuropathol. (Berl.) 100: 415, 2000. Mole, S., and Gardiner, M.: Molecular genetics of the neuronal ceroid lipofuscinoses. Epilepsia 40:S329, 1999. Monney, L., Otter, I., Olivier, R., et al.: Defects in the ubiquitin pathway induce caspase-independent apoptosis blocked by Bcl-2. J. Biol. Chem. 273:6121, 1998. Montgomery, M. M., Wood, A., Stott, E. J., et al.: Changes in neuron size in cynomolgus macaques infected with various immunodeficiency viruses and poliovirus. Neuropathol. Appl. Neurobiol. 24:468, 1998. Montpetit, V. J., Clapin, D. F., Tryphonas, L., and Dancea, S.: Alteration of neuronal cytoskeletal organization in dorsal root ganglia associated with pyridoxine neurotoxicity. Acta Neuropathol. (Berl.) 76:71, 1988. Moran, L. B., Kosel, S., Spitzer, C., et al.: Expression of alpha-synuclein in non-apoptotic, slowly degenerating facial motoneurones. J. Neurocytol. 30:515, 2001. Moretto, G., Monaco, S., Passarin, M. G., et al.: Cytoskeletal changes induced by 2,5-hexanedione on developing human neurons in vitro. Arch. Toxicol. 65:409, 1991. Morgello, S., Cho, E.-S., Nielsen, S., et al.: Cytomegalovirus encephalitis in patients with acquired immunodeficiency syndrome: an autopsy study of 30 cases and a review of the literature. Hum. Pathol. 18:289, 1987. Mossman, J., Blunt, S., Stephens, R., et al.: Hunter’s disease in a girl: association with X:5 chromosomal translocation disrupting the Hunter gene. Arch. Dis. Child. 58:911, 1983. Mowat, D. R., Croaker, G. D., Cass, D. T., et al.: Hirschsprung disease, microcephaly, mental retardation and characteristic facial features: delineation of a new syndrome and identification of a locus at chromosome 2q22-q23. J. Med. Genet. 35:617, 1998. Müller, H. D., Mugler, M., Ramaekers, V. T., and Schröder, J. M.: Hereditary sensory and motor neuropathy with absence of large myelinated fibers due to absence of large neurons in dorsal root ganglia and anterior horns, clinically associated

311.

312.

313.

314.

315.

316.

317.

318.

319.

320. 321.

322.

323.

324.

325.

326.

327.

with deafness, mental retardation, and epilepsy (HMSNADM). J. Peripher. Nerv. Syst. 5:147, 2000. Muma, N. A., and Cork, L. C.: Alterations in neurofilament mRNA in hereditary canine spinal muscular atrophy. Lab. Invest. 69:436, 1993. Muma, N. A., Hoffman, P. N., Slunt, H. H., et al.: Alterations in levels of mRNAs coding for neurofilament protein subunits during regeneration. Exp. Neurol. 107:230, 1990. Murashov, A. K., Haq, I. U., Hill, C., et al.: Crosstalk between p38, HSP25 and Akt in spinal motor neurons after sciatic nerve injury. Mol. Brain Res. 93:199, 2001. Musaka, T., Urase, K., Momoi, Y. M., et al.: Specific expression of CPP32 in sensory neurons of mouse embryos and activation of CPP32 in the apoptosis induced by a withdrawal of NGF. Biochem. Biophys. Res. Commun. 231:770, 1997. Nacimiento, W., Schlozer, B., Brook, G. A., et al.: Transient decrease of acetylcholinesterase in ventral horn neurons caudal to a low thoracic spinal cord hemisection in the adult rat. Brain Res. 714:177, 1996. Nagamatsu, M., Terao, S., Misu, K., et al.: Axonal and perikaryal involvement in chronic inflammatory demyelinating polyneuropathy. J. Neurol. Neurosurg. Psychiatry 66:727, 1999. Nagano, I., Shapshak, P., Yoshioka, M., et al.: Increased NADPH-diaphorase reactivity and cytokine expression in dorsal root ganglia in acquired immunodeficiency syndrome. J. Neurol. Sci. 136:117, 1996. Nagashima, K., Sakakibara, K., Endo, H., et al.: I-cell disease (mucolipidosis II): pathological and biochemical studies of an autopsy case. Acta Pathol. Jpn. 27:251, 1977. Nandy, K.: Further effects of centrophenoxine on the lipofuscin pigment in the neurons of senile guinea pigs. J. Gerontol. 23:82, 1968. Nandy, K.: Properties of neuronal lipofuscin pigment in mice. Acta Neuropathol. (Berl.) 19:25, 1971. Narahara, K., Takahashi, Y., Murakami, M., et al.: Terminal 22q deletion associated with a partial deficiency of arylsulphatase A. J. Med. Genet. 29:432, 1992. Nebes, V. L., and Schmidt, M. C.: Human lysosomal alphamannosidase: isolation and nucleotide sequence of the full-length cDNA. Biochem. Biophys. Res. Commun. 200:239, 1994. Nishimura, M., Namba, Y., Ikeda, K., et al.: Neurofibrillary tangles in the neurons of spinal dorsal root ganglia of patients with progressive supranuclear palsy. Acta Neuropathol. (Berl.) 85:453, 1993. Nittono, K.: On bilateral effects from the unilateral section of branches of the nervous trigeminus in the albino rat. J. Comp. Neurol. 35:133, 1923. Nolan, C. C., Brown, A. W., and Cavanagh, J. B.: Regional variations in nerve cell responses to trimethyltin intoxication I. Mongolian gerbils and rats: further evidence for involvement of the Golgi apparatus. Acta Neuropathol. (Berl.) 81:204, 1990. Novikov, L., Novikova, L., and Kellerth, J. O.: Brain-derived neurotrophic factor promotes survival and blocks nitric oxide synthase expression in adult rat spinal motoneurons after ventral root avulsion. Neurosci. Lett. 200:45, 1995. Oblinger, M. M., and Lasek, R. J.: Axotomy-induced alterations in the synthesis and transport of neurofilaments and

Pathology of Peripheral Neuron Cell Bodies

328. 329.

330.

331.

332. 333. 334.

335.

336.

337.

338.

339.

340.

341.

342.

343.

344.

345.

microtubules in dorsal root ganglion cells. J. Neurosci. 8:1747, 1988. Ohnishi, A., and Dyck, P. J.: Loss of small peripheral sensory neurons in Fabry disease. Arch. Neurol. 31:120, 1974 Ohtake, K., Matsumoto, M., Wakamatsu, H., et al.: Glutamate release and neuronal injury after intrathecal injection of local anaesthetics. Neuroreport 11:1105, 2000. Okado, N., and Oppenheim, R. W.: Cell death of motoneurons in the chick embryo spinal cord. IX. The loss of motoneurons following removal of afferent inputs. J. Neurosci. 4:1639, 1984. Ong, W. Y., Kumar, U., Switzer, R. C., et al.: Neurodegeneration in Niemann-Pick type C disease mice. Exp. Brain Res. 141:218, 2001. Oorschot, D. E., and McLennan, I. S.: The trophic requirements of mature motoneurons. Brain Res. 789:315, 1998. Oppenheim, R. W.: Cell death during development of the nervous system. Annu. Rev. Neurosci. 14:453,1991. Orozco, G., Estrada, R., Perry, T., et al.: Dominantly inherited olivopontocerebellar atrophy from Eastern Cuba: clinical, neuropathological and biochemical findings. J. Neurol. Sci. 93:37, 1989. Orr, H. T., Chung, M.-Y., Banfi, S., et al.: Expansion of an unstable trinucleotide CAG repeat in spinocerebellar ataxia type 1. Nat. Genet. 4:221, 1993. O’Toole, D., Ingram, J., Welch, V., et al.: An inherited lower motor neuron disease of pigs: clinical signs in two litters and pathology of an affected pig. J. Vet. Diagn. Invest. 6:62, 1994. Otto, D., Unsicker, K., and Grothe, C.: Pharmacological effects of nerve growth factor and fibroblast growth factor applied to the transectioned sciatic nerve on neuron death in adult rat dorsal root ganglia. Neurosci. Lett. 83:156, 1987. Oyanagi, K., Makifuchi, T., Ohtoh, T., et al.: Amyotrophic lateral sclerosis of Guam: the nature of the neuropathological findings. Acta Neuropathol. (Berl.) 88:405, 1994. Pang, J., Giunti, P., Chamberlain, S., et al.: Neuronal intranuclear inclusions in SCA2: a genetic, morphological and immunohistochemical study of two cases. Brain 125:656, 2002. Pannese, E.: Investigations on the ultrastructural changes of the spinal ganglion neurons in the course of axon regeneration and cell hypertrophy. I. Changes during axon regeneration. Z. Zellforsch. Mikroskop. Anat. 60:711, 1963. Pardo, C. A., McArthur, J. C., and Griffin, J. W.: HIV neuropathy: insight in the pathology of HIV peripheral nerve disease. J. Peripher. Nerv. Syst. 6:21, 2001. Paulson, H. L., Perez, M. K., Trottier, Y., et al.: Intranuclear inclusions of expanded polyglutamine protein in spinocerebellar ataxia type 3. Neuron 19:333, 1997. Peiffer, J., Schlote, W., Bischoff, A., et al.: Generalized giant axonal neuropathy: a filament-forming disease of neuronal, endothelial, glial, and Schwann cells in a patient without kinky hair. Acta Neuropathol. (Berl.) 40:213, 1977. Petit, P. X., Goubern, M., Diolez, P., et al.: Disruption of the outer mitochondrial membrane as a result of large amplitude swelling: the impact of irreversible permeability transition. FEBS Lett. 426:111, 1998. Pettmann, B., and Henderson, C. E.: Neuronal cell death. Neuron 20:633, 1998.

727

346. Peyronnard, J.-M., Charron, L., Messier, J.-P., and Lavoie, J.: Differential effects of distal and proximal nerve lesions on carbonic anhydrase activity in rat primary sensory neurons, ventral and dorsal root axons. Exp. Brain Res. 70:550, 1988. 347. Peyronnard, J.-M., Charron, L., Messier, J.-P., et al.: Changes in lectin binding of lumbar dorsal root ganglia neurons and peripheral axons after sciatic and spinal nerve injury in the rat. Cell Tissue Res. 257:379, 1989. 348. Picklo, M. J., Wiley, R. G., Lappi, D. A., and Robertson, D.: Noradrenergic lesioning with an anti-dopamine betahydroxylase immunotoxin. Brain Res. 666:195, 1994. 349. Piehl, F., Arvidsson, U., Johnson, H., et al.: GAP43, aFGF, CCK, ␣ and ␤ CGRP in rat spinal motoneurons subjected to axotomy and/or dorsal root severance. Eur. J. Neurosci. 5:1321, 1993. 350. Piehl, F., Tabar, G., and Cullheim, S.: Expression of NMDA receptor mRNAs in rat motoneurons is down regulated after axotomy. Eur. J. Neurosci. 7:2101, 1995. 351. Pollin, M. M., McHanwell, S., and Slater, C. R.: Loss of motor neurons from the median nerve motor nucleus of the mutant mouse ‘wobbler.’ J. Neurocytol. 19:29, 1990. 352. Popken, G. J., and Farel, P. B.: Sensory neuron number in neonatal and adult rats estimated by means of stereologic and profile-based methods. J. Comp. Neurol. 386:8, 1997. 353. Portera-Cailliau, C., Price, D. L., and Martin, L. J.: NonNMDA and NMDA receptor-mediated excitotoxic neuronal deaths in adult brain are morphologically distinct: further evidence for an apoptosis-necrosis continuum. J. Comp. Neurol. 378:88, 1997. 354. Powers, B. E., Beck, E. R., Gillette, E. L., et al.: Pathology of radiation injury to the canine spinal cord. Int. J. Radiat. Oncol. Biol. Phys. 23:539, 1992. 355. Powers, J. M.: Normal and defective neuronal membranes: structure and function. Neuronal lesions in peroxisomal disorders. J. Mol. Neurosci. 16:285, 2001. 356. Powers, J. M., DeCiero, D. P., Cox, C., et al.: The dorsal root ganglia in adrenomyeloneuropathy: neuronal atrophy and abnormal mitochondria. J. Neuropathol. Exp. Neurol. 60:493, 2001. 357. Price, D. L., and Porter, K. R.: The response of ventral horn neurons to axonal transection. J. Cell Biol. 53:24, 1972. 358. Pullen, A. H.: Morphometric evidence from c-synapses for phased Nissl body response in alpha-motoneurones retrogradely intoxicated with diphtheria toxin. Brain Res. 509:8, 1990. 359. Raas-Rothschild, A., Cormier-Daire, V., Bao, M., et al.: Molecular basis of variant pseudo-Hurler polydystrophy (mucolipidosis IIIC). J. Clin. Invest. 105:673, 2000. 360. Raivich, G., Hellweg, R., and Kreutzberg, G. W.: NGF receptor-mediated reduction in axonal NGF uptake and retrograde transport following sciatic nerve injury and during regeneration. Neuron 7:151,1991. 361. Rance, N. E., McArthur, J. C., Cornblath, D. R., et al.: Gracile tract degeneration in patients with sensory neuropathy and AIDS. Neurology 38:265, 1988. 362. Ranson, S. W.: Retrograde degeneration in the spinal nerves. J. Comp. Neurol. Psychol. 16:265, 1906. 363. Ranson, S. W.: Alterations in the spinal ganglion cells following axotomy. J. Comp. Neurol. 19:125, 1909.

728

Pathology of the Peripheral Nervous System

364. Ranta, S., Savukoski, M., Santavuori, P., and Haltia, M.: Studies of homogenous populations: CLN5 and CLN8. Adv. Genet. 45:123, 2001. 365. Resch, K., Korthaus, D., Wedemeyer, N., et al.: Homology between human chromosome 2p13.3 and the wobbler critical region on mouse chromosome 11: comparative highresolution mapping of STS and EST loci on YAC/BAC contigs. Mamm. Genome 9:893, 1998. 366. Rich, K. M., Disch, S. P., and Eichler, M. E.: The influence of regeneration and nerve growth factor on the neuronal cell body reaction to injury. J. Neurocytol. 18:569, 1989. 367. Rich, K. M., Luszczynski, J. R., Osborne, P. A., and Johnson, E. M.: Nerve growth factor protects adult sensory neurons from cell death and atrophy caused by nerve injury. J. Neurocytol. 16:261, 1987. 368. Risling, M., Aldskogius, H., and Hildebrand, C.: Effects of sciatic nerve crush on the L7 spinal roots and dorsal root ganglia in kittens. Exp. Neurol. 79:176, 1983. 369. Robbins, J. H., Kraemer, K. H., Merchant, S. N., and Brumback, R. A.: Adult-onset xeroderma pigmentosum neurological disease—observations in an autopsy case. Clin. Neuropathol. 21:18, 2002. 370. Robitaille, Y., Schut, L., and Kish, S. J.: Structural and immunocytochemical features of olivopontocerebellar atrophy caused by the spinocerebellar ataxia type 1 (SCA-1) mutation define a unique phenotype. Acta Neuropathol. (Berl.) 90:572, 1995. 371. Romanes, G. J.: Motor localization and the effects of nerve injury on the ventral horn cells of the spinal cord. J. Anat. 42:117, 1946. 372. Rossiter, J. P., Riopelle, R. J., and Bisby, M. A.: Axotomyinduced apoptotic cell death of neonatal rat facial motoneurons: time course analysis and relation to NADPH-diaphorase activity. Exp. Neurol. 138:33, 1996. 373. Rowland, L. P., Defendini, R., Sherman, W., et al.: Macroglobulinemia with peripheral neuropathy simulating motor neuron disease. Ann. Neurol. 11:532, 1982. 374. Russell, J. W., Sullivan, K. A., Windebank, A. J., et al.: Neurons undergo apoptosis in animal and cell culture models of diabetes. Neurobiol. Dis. 6:347, 1999. 375. Rust, S., Rosier, M., Funke, H., et al.: Tangier disease is caused by mutations in the gene encoding ATP-binding cassette transporter 1. Nat. Genet. 22:352, 1999. 376. Rydh-Rinder, M., Holmberg, K., Elfvin, L.-G., et al.: Effects of peripheral axotomy on neuropeptides and nitric oxide synthase in dorsal root ganglia and spinal cord of the guinea pig: an immunohistochemical study. Brain Res. 707:180, 1996. 377. Sack, G. H., Cork, L. C., Morris, J. M., et al.: Autosomal dominant inheritance of hereditary canine spinal muscular dystrophy. Ann. Neurol. 15:369, 1984. 378. Saito, Y., and Murayama, S.: Expression of tau immunoreactivity in the spinal motor neurons in Alzheimer’s disease. Neurology 55:1727, 2000. 379. Samuel, B. U., Ponnuraj, E., Rajasingh, J., and John, T. J.: Experimental poliomyelitis in bonnet monkey: clinical features, virology and pathology. Dev. Biol. Stand. 78:71, 1993. 380. Sanchez-Vives, M. V., Valdeolmillos, M., Martinez, S., and Gallego, R.: Axotomy-induced changes in Ca2⫹ homeostasis in rat sympathetic ganglion cells. Eur. J. Neurosci. 6:9, 1994.

381. Sango, K., Yamanaka, S., Hoffmann, A., et al.: Mouse models of Tay-Sachs and Sandhoff diseases differ in neurologic phenotype and ganglioside metabolism. Nat. Genet. 11:170, 1995. 382. Sasaki, S., Shibata, N., Komori, T., and Iwata, M.: iNOS and nitrotyrosine immunoreactivity in amyotrophic lateral sclerosis. Neurosci. Lett. 291:44, 2000. 383. Satake, M., Yoshimura, T., Iwaki, T., et al.: Anti-dorsal root ganglion neuron antibody in a case of dorsal root ganglionitis associated with Sjögren syndrome. J. Neurol. Sci. 132:122, 1995. 384. Scaravilli, F.: Changes in neuronal structure and cell populations with ageing. In Thomas, P. K. (ed.): Peripheral Nerve Changes in the Elderly (New Issues in Neurosciences). Amsterdam, Wiley, p. 95, 1988. 385. Scaravilli, F., An, S.-F., Groves, M., and Thom, M.: The neuropathology of paraneoplastic syndromes. Brain Pathol. 9:251, 1999. 386. Scaravilli, F., and Duchen, L. W.: Electron microscopic and quantitative studies of cell necrosis in developing sensory ganglia in normal and sprawling mutant mice. J. Neurocytol. 9:373, 1980. 387. Scaravilli, F., Gray, F., Mikol, J., and Sinclair, E.: Pathology of the nervous system. In Scaravilli, F. (ed.): The Neuropathology of HIV Infection. London, Springer, p. 99, 1993. 388. Scaravilli, F., and Jacobs, J. M.: Quantitative studies of motor, spino-cerebellar and secondary sensory neurons in the mutilated foot mutant rat. Neuroscience 6:1663, 1981. 389. Scaravilli, F., Sinclair, E., Arango, J. C., et al.: The pathology of the posterior root ganglia in AIDS and its relationship to the pallor of the gracile tract. Acta Neuropathol. (Berl.) 84:163, 1992. 390. Scaravilli, T., Pramstaller, P. P., Salerno, A., et al.: Neuronal loss in Onuf’s nucleus in three patients with progressive supranuclear palsy. Ann. Neurol. 48:97, 2000. 391. Schaumberg, H. H., and Berger, A. R.: Human toxic neuropathy due to industrial agents. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1533, 1993. 392. Schinder, V., Govrin-Lippmann, R., Cohen, S., et al.: Structural basis of sympathetic-sensory coupling in rat and human dorsal root ganglia following peripheral nerve injury. J. Neurocytol. 28:743, 1999. 393. Schlote, W., Harzer, K., Christomanou, H., et al.: Sphingolipid activator protein 1 deficiency in metachromatic leukodystrophy with normal arylsulphatase A activity: a clinical, morphological, biochemical, and immunological study. Eur. J. Pediatr. 150:584, 1991. 394. Schmalbruch, H.: Loss of sensory neurons after sciatic nerve section in the rat. Anat. Rec. 219:323, 1987. 395. Schmalbruch, H., and Haase, G.: Spinal muscular atrophy: present state. Brain Pathol. 11:231, 2001. 396. Schmalbruch, H., Stender, S., and Boysen, G.: Abnormalities in spinal neurons and dorsal root ganglion cells in Tangier disease presenting with a syringomyelia-like syndrome. J. Neuropathol. Exp. Neurol. 46:533, 1987. 397. Schmidt, M. L., Zhukareva, V., Perl, D. P., et al.: Spinal cord neurofibrillary pathology in Alzheimer’s disease and Guam

Pathology of Peripheral Neuron Cell Bodies

398. 399.

400.

401.

402.

403.

404.

405.

406.

407.

408.

409.

410.

411.

412.

413.

414.

parkinsonism-dementia complex. J. Neuropathol. Exp. Neurol. 60:1075, 2001. Schmidt, R. E.: Neuropathology of human sympathetic autonomic ganglia. Microsc. Res. Tech. 35:107, 1996. Schmidt, R. E., Beaudet, L. N., Plurad, S. B., and Dorsey, D. A.: Axonal cytoskeletal pathology in aged and diabetic human sympathetic autonomic ganglia. Brain Res. 769:375, 1997. Schmidt, R. E., Dorsey, D., Parvin, C. A., et al.: Dystrophic axonal swellings develop as a function of age and diabetes in human dorsal root ganglia. J. Neuropathol. Exp. Neurol. 56:1028, 1997. Schmidt, R. E., Plurad, S. B., Parvin, C. A., and Roth, K. A.: Effect of diabetes and ageing on human sympathetic autonomic ganglia. Am. J. Pathol. 143:143, 1993. Schold, S. C., Cho, E. S., Somasundaram, M., and Posner, J. B.: Subacute motor neuronopathy: a remote effect of lymphoma. Ann. Neurol. 5:271, 1979. Schroer, J. A., Plurad, S. B., and Schmidt, R. E.: Fine structure of presynaptic axonal terminals in sympathetic autonomic ganglia of ageing and diabetic human subjects. Synapse 12:1, 1992. Schwab, C., DeMaggio, A. J., Ghoshal, N., et al.: Casein kinase 1 delta is associated with pathological accumulation of tau in several neurodegenerative diseases. Neurobiol. Aging 21:503, 2000. Scott, J. N., Clark, A. W., and Zochodne, D. W.: Neurofilament and tubulin gene expression in progressive experimental diabetes: failure of synthesis and export by sensory neurons. Brain 122:2019, 1999. Sebert, M. E., and Shooter, E. M.: Expression of mRNA for neurotrophic factors and their receptors in the rat dorsal root ganglion and sciatic nerve following nerve injury. J. Neurosci. Res. 36:357, 1993. Seil, F. J., Lampert, P. W., and Klatzo, I.: Neurofibrillary spheroids induced by aluminum phosphate in dorsal root ganglia neurons in vitro. J. Neuropathol. Exp. Neurol. 28:74, 1969. Sekhon, S. S., and Maxwell, D. S.: Ultrastructural changes in neurons of the spinal anterior horn of ageing mice with particular reference to the accumulation of lipofuscin pigment. J. Neurocytol. 3:59, 1974. Shadiack, A. M., Sun, Y., and Zigmond, R. E.: Nerve growth factor antiserum induces axotomy-like changes in neuropeptide expression in intact sympathetic and sensory neurons. J. Neurosci. 21:363, 2001. Shaw, C. E., al-Chalabi, A., and Leigh, N.: Progress in the pathogenesis of amyotrophic lateral sclerosis. Curr. Neurol. Neurosci. Rep. 1:69, 2001. Shen, H., Chung, J. M., Coggeshall, R. E., and Chung, K.: Changes in trkA expression in the dorsal root ganglion after peripheral nerve injury. Exp. Brain Res. 127:141, 1999. Sheth, K. J., Werlin, S. L., Freeman, M. E., and Hodach, A. E.: Gastrointestinal structure and function in Fabry’s disease. Am. J. Gastroenterol. 76:246, 1981. Shi, T. J., Tandrup, T., Bergman, E., et al.: Effect of peripheral nerve injury on dorsal root ganglion neurons in the C57 BL/6J mouse: marked changes both in cell numbers and neuropeptide expression. Neuroscience 105:249, 2001. Siconolfi, L. B., and Seeds, N. W.: Induction of the plasminogen activator system accompanies peripheral nerve regeneration after sciatic crush. J. Neurosci. 21:4336, 2001.

729

415. Silos-Santiago, I., Molliver, D. C., Ozaki, S., et al.: Non-trkAexpressing small DRG neurons are lost in TrkA deficient mice. J. Neurosci. 15:5929, 1995. 416. Silverman, J. D., and Kruger, L.: Selective neuronal glycoconjugate expression in sensory and autonomic ganglia: relation of lectin reactivity to peptide and enzyme markers. J. Neurocytol. 19:789, 1990. 417. Simic, G., Seso-Simic, D., Lucassen, P. J., et al.: Ultrastructural analysis and TUNEL demonstrate motor neuron apoptosis in Werdnig-Hoffmann disease. J. Neuropathol. Exp. Neurol. 59:398, 2000. 418. Simon, L. T., Ricaurte, G. A., and Forno, L. S.: Chronic idiopathic ataxic neuropathy: neuropathology of a case. Acta Neuropathol. (Berl.) 79:104, 1989. 419. Singer, P. A., Mehler, S., and Fernandez, H. L.: Blockade of retrograde axonal transport delays the onset of metabolic and morphologic changes induced by axotomy. J. Neurosci. 2:1299, 1982. 420. Singer, P. A., Mehler, S., and Fernandez, H. L.: Effect of extracts of injured nerve on initiating the regenerative response in the hypoglossal nucleus in the rat. Neurosci. Lett. 84:155, 1988. 421. Slaugenhaupt, S. A., Acierno, J. S. Jr., Helbling, L. A., et al.: Mapping of the mucolipidosis type IV gene to chromosome 19p and definition of founder haplotypes. Am. J. Hum. Genet. 65:773, 1999. 422. Soares, H. D., Chen, S. C., and Morgan, J. I.: Differential and prolonged expression of Fos-lacZ and jun-lacZ in neurons, glia, and muscle following sciatic nerve damage. Exp. Neurol. 167:1, 2001. 423. Sobkowicz, H. M., Slapnick, S. M., and August, B. K.: Apoptosis of inner hair cells caused by laser ablation of their spiral ganglion neurons in cultures of the mouse organ of Corti. J. Neurocytol. 28:939, 1999. 424. Sobue, G., Hashizume, Y., Mukai, E., et al.: X-linked recessive bulbospinal neuronopathy. Brain 112:209, 1989. 425. Sobue, G., Nakao, N., Murakami, K., et al.: Type 1 familial amyloid polyneuropathy. Brain 113:903, 1990. 426. Sodhi, N., Camilleri, M., Camoriano, J. K., et al.: Autonomic function and motility in intestinal pseudo-obstruction caused by paraneoplastic syndrome. Dig. Dis. Sci. 34:1937, 1989. 427. Soffer, D.: Neuronal intranuclear hyaline inclusion disease presenting as Friedreich’s ataxia. Acta Neuropathol. (Berl.) 65:322, 1985. 428. Sommervaille, T., Molander, C., and Woolf, C. J.: Timedependent differences in the increase in GAP-43 expression in dorsal root ganglion cells after peripheral axotomy. Neuroscience 45:213, 1991. 429. Sosunov, A. A., Krugliakov, P. P., Shvalev, V. N., et al.: Age related changes in the autonomic ganglia [in Russian]. Arkh. Pathol. 59:32, 1997. 430. Sterman, A. B.: Acrylamide induces early morphologic reorganization of the neuronal cell body. Neurology 32:1023, 1982. 431. Sterman, A. B.: Cell body remodeling during dying-back axonopathy: DRG changes during advanced disease. J. Neuropathol. Exp. Neurol. 41:400, 1982. 432. Sterman, A. B., and Delannoy, M. R.: Cell body responses to axonal injury: traumatic axotomy versus toxic neuropathy. Exp. Neurol. 89:408, 1985.

730

Pathology of the Peripheral Nervous System

433. Stoltenburg-Didinger, G., Boegner, F., Gruning, W., et al.: Specific neurotoxic effects of different organic solvents on dissociated cultures of the nervous system. Neurotoxicology 13:161, 1992. 434. Su, M., Wakabayashi, K., Kakita, A., et al.: Selective involvement of large motor neurons in the spinal cord of rats treated with methylmercury. J. Neurol. Sci. 156:12, 1998. 435. Sun, Y., and Zigmond, R. E.: Regulation of neuropeptide expression in sympathetic neurons: paracrine and retrograde influences. Ann. N. Y. Acad. Sci. 814:181, 1997. 436. Sung, J. H.: Light, fluorescence, and electron microscopic features of neuronal intranuclear hyaline inclusions associated with multisystem atrophy. Acta Neuropathol. (Berl.) 50:115, 1980. 437. Sung, J. H., and Mastri, A. R.: Spinal autonomic neurons in Werdnig-Hoffmann disease, mannosidosis, and Hurler’s syndrome: distribution of autonomic neurons in the sacral spinal cord. J. Neuropathol. Exp. Neurol. 39:441, 1980. 438. Sung, J. H., Meyers, J. P., Stadlan, E. M., et al.: Neuropathological changes in Chediak-Higashi disease. J. Neuropathol. Exp. Neurol. 28:86, 1969. 439. Sung, J. O., Hayano, M., and Desnick, R. J.: Mannosidosis: pathology of the nervous system. J. Neuropathol. Exp. Neurol. 36:807, 1977. 440. Suzuki, H., Oyanagi, K., Takahashi, H., et al.: A quantitative pathological investigation of the cervical cord, roots and ganglia after long-term amputation of the unilateral upper arm. Acta Neuropathol. (Berl.) 85:666, 1993. 441. Suzuki, K.: Neuropathology of late onset gangliosidosis: a review. Dev. Neurosci. 13:205, 1991. 442. Suzuki, K., and Suzuki, K.: Lysosomal diseases. In Graham, D. I., and Lantos, P. L. (eds.): Greenfield’s Neuropathology. London, Arnold, p. 653, 2002. 443. Svensson, M., and Aldskogius, H.: The effect of axon injury on microtubule-associated proteins MAP2, 3 and 5 in the hypoglossal nucleus of the adult rat. J. Neurocytol. 21:222, 1992. 444. Swett, J. E., Hong, C.-Z., and Miller, P. G.: Most dorsal root ganglion neurons of the adult rat survive nerve crush injury. Somatosens. Motor Res. 12:177, 1995. 445. Tabira, T., Goto, I., and Kuroiwa, Y.: Neuropathological and biochemical studies in Fabry’s disease. Acta Neuropathol. (Berl.) 30:345, 1974. 446. Takada, K., and Becker, L. E.: Cockayne’s syndrome: report of two autopsy cases associated with neurofibrillary tangles. Clin. Neuropathol. 5:64, 1986. 447. Takahashi, J., Tanaka, J., Arai, K., et al.: Recruitment of nonexpanded polyglutamine proteins to intranuclear aggregates in neuronal intranuclear hyaline inclusion disease. J. Neuropathol. Exp. Neurol. 60:369, 2001. 448. Takano, T., and Yamanouchi, Y.: Assignment of human betagalactosidase-A gene to 3p21.33 by fluorescence in situ hybridization. Hum. Genet. 92:403, 1993. 449. Takase, Y., Takahashi, K., Takada, K., et al.: Hereditary motor and sensory neuropathy type 1 (HMSN 1) associated with cranial neuropathy: an autopsy case report. Acta Neurol. Scand. 82:368, 1990. 450. Tanabe, K., Kiryu-Seo, S., Nakamura, T., et al.: Alternative expression of Shc family members in nerve-injured motoneurons. Brain Res. Mol. Brain Res. 53:291, 1998.

451. Tandrup, T., and Braendgaard, H.: Number and volume of rat dorsal root ganglion cells in acrylamide intoxication. J. Neurocytol. 23:242, 1994. 452. Tandrup, T., Woolf, C. J., and Coggeshall, R. E.: Delayed loss of small dorsal root ganglion cells after transection of the rat sciatic nerve. J. Comp. Neurol. 422:172, 2000. 453. Tang, T. T., Segura, A. D., Chen, Y. T., et al.: Neonatal hypotonia and cardiomyopathy secondary to type IV glycogenosis. Acta Neuropathol. (Berl.) 87:531, 1994. 454. Tellez Nagel, I., Rapin, I., Iwamoto, T., et al.: Mucolipidosis IV: clinical, ultrastructural, histochemical and chemical studies of a case, including a brain biopsy. Arch. Neurol. 33:828, 1976. 455. Tenser, R. B.: Sequential changes of sensory neuron (fluoride resistant) acid phosphatase in dorsal root ganglion neurons following neurectomy and rhizotomy. Brain Res. 332:386, 1985. 456. Terenghi, G.: Peripheral nerve regeneration and neurotrophic factors. J. Anat. 194:1, 1999. 457. Terry, R. D., and Weiss, M.: Studies in Tay-Sachs disease. II. Ultrastructure of the cerebrum. J. Neuropathol. Exp. Neurol. 22:18, 1963. 458. Tetzlaff, W., Alexander, S. W., Miller, F. D., and Bisby, M. A.: Response of facial and rubrospinal neurons to axotomy: changes in mRNA expression for cytoskeletal proteins and GAP-43. J. Neurosci. 11:2528, 1991. 459. Tetzlaff, W., and Bisby, M. A.: Cytochemical protein synthesis and regulation of nerve regeneration in PNS and CNS neurons of the rat. Restor. Neurol. Neurosci. 1:189, 1990. 460. Thorner, P. S., Bilbao, J. M., Sima, A. A. F., and Briggs, S.: Porphyric neuropathy: an ultrastructural and quantitative case study. Can. J. Neurol. Sci. 8:281, 1981. 461. Tolkovsky, A.: Neurotrophic factors in action—new dogs and new tricks. TINS 20:1, 1997. 462. Tong, J. X., and Rich, K. M.: Diphenylpiperazines enhance regeneration after facial nerve injury. J. Neurocytol. 26:339, 1997. 463. Tonra, J. R., Curtis, R., Wong, V., et al.: Axotomy upregulates the anterograde transport and expression of brain-derived neurotrophic factor by sensory neurons. J. Neurosci. 18:4374, 1998. 464. Topp, K. S., Tanner, K. D., and Levine, J. D.: Damage to the cytoskeleton of large diameter sensory neurons and myelinated axons in vincristine-induced painful peripheral neuropathy in the rat. J. Comp. Neurol. 424:563, 2000. 465. Tornquist, E., Liu, L., Mattsson, P., and Svensson, M.: Response of glial cells and activation of complement following motorneuron degeneration induced by toxic ricin. Neurosci. Res. 28:167, 1997. 466. Toscano, E., della Casa, R., Mardy, S., et al.: Multisystem involvement in congenital insensitivity to pain with anhidrosis (CIPA), a nerve growth factor receptor (trk A)-related disorder. Neuropediatrics 31:39, 2000. 467. Toyooka, K., Fujimura, H., Ueno, S., et al.: Familial amyloid polyneuropathy associated with transthyretin Gly42 mutation: a qualitative light and electron microscopic study of the peripheral nervous system. Acta Neuropathol. (Berl.) 90:516, 1995. 468. Tredici, G., Tredici, S., Fabbrica, D., et al.: Experimental cisplatin neuronopathy in rats and the effect of retinoic acid administration. J. Neurooncol. 36:31, 1998.

Pathology of Peripheral Neuron Cell Bodies 469. Trottier, Y., Lutz, Y., Stevanin, G., et al.: Polyglutamine expansion as a pathological epitope in Huntington’s disease and four dominant cerebellar ataxias. Nature 378:403, 1995. 470. Troy, C. M., Muma, N. A., Greene, L. A., et al.: Regulation of peripherin and neurofilament expression in regenerating rat motor neurons. Brain Res. 529:232, 1990. 471. Tsujino, H., Kondo, E., Fukuoka, T., et al.: Activating transcription factor 3 (ATF3) induction by axotomy in sensory and motoneurons: a novel neuronal marker of nerve injury. Mol. Cell Neurosci. 15:170, 2000. 472. Tsuzuki, K., Kondo, E., Fukuoka, T., et al.: Differential regulation of P2X(3) mRNA expression by peripheral nerve injury in intact and injured neurons in the rat sensory ganglia. Pain 91:351, 2001. 473. Umahara, T., Hirano, A., Kato, S., et al.: Demonstration of neurofibrillary tangles and neuropil thread-like structures in spinal cord white matter in Parkinson-dementia complex of Guam and in Guamanian amyotrophic lateral sclerosis. Acta Neuropathol. (Berl.) 88:180, 1994. 474. Urushanti, M., and Shimohama, S.: The role of nitric oxide in amyotrophic lateral sclerosis. Amyotroph. Lateral Scler. Other Motor Neuron Disord. 2:71, 2001. 475. Van Den Bosch, L., Schwaller, B., Vleminckx, V., et al.: Protective effect of parvalbumin on excitotoxic motor neuron death. Exp. Neurol. 174:150, 2002. 476. Van den Bosch de Aguilar, P. H., and Goemare-Vanneste, J.: Paired helical filaments in spinal ganglion neurons of elderly rats. Virchows Arch. B 47:217, 1984. 477. van der Walt, J. D., Swash, M., Leake, J., and Cox, E. L.: The pattern of involvement of adult-onset acid maltase deficiency at autopsy. Muscle Nerve 10:272, 1987. 478. Vandenberghe, W., Robberecht, W., and Brorson, J. R.: AMPA receptor calcium permeability, GluR2 expression, and selective motoneuron vulnerability. J. Neurosci. 20:123, 2000. 479. Vanderluit, J. L., McPhail, L. T., Fernandes, K. J., et al.: Caspase-3 is activated following axotomy of neonatal facial motoneurons and caspase-3 gene deletion delays axotomyinduced cell death in rodents. Eur. J. Neurosci. 12:3469, 2000. 480. Verge, V., Zhang, X., Xu, X.-J., et al.: Marked increase in nitric oxide synthase mRNA in rat dorsal root ganglia after peripheral axotomy: in situ hybridization and functional studies. Proc. Natl. Acad. Sci. USA 89:11617, 1992. 481. Verge, V. M. K., Richardson, P. M., Wiesenfeld-Hallin, Z., and Hökfelt, T.: Differential influence of nerve growth factor on neuropeptide expression in vivo: a novel role in peptide suppression in adult sensory neurons. J. Neurosci. 15:2081, 1995. 482. Verkerk, A. J., Pieretti, M., Sutcliffe, J. S., et al.: Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome. Cell 65:905, 1991. 483. Verma, A., Berger, J. R., Snodgrass, S., and Petito, C.: Motor neuron disease: a paraneoplastic process associated with anti-Hu antibody and small cell lung carcinoma. Ann. Neurol. 40:112, 1996. 484. Vestergaard, S., Tandrup, T., and Jakobsen, J.: Effect of permanent axotomy on number and volume of dorsal root ganglion cell bodies. J. Comp. Neurol. 388:307, 1997. 485. Villar, M. J., Cortés, R., Theodorsson, E., et al.: Neuropeptide expression in rat dorsal root ganglion cells and spinal cord

486.

487.

488.

489.

490.

491. 492.

493.

494.

495.

496.

497.

498.

499.

500.

501.

502.

731

after peripheral nerve injury with special reference to galanin. Neuroscience 33:587, 1989. Vitaliani, R., Scaravilli, T., Egarter-Vigl, E., et al.: The pathology of the spinal cord in progressive supranuclear palsy. J. Neuropathol. Exp. Neurol. 61:268, 2002. Wakabayashi, K., Kobayashi, H., Kawasaki, S., et al.: Autosomal recessive spastic paraplegia with hypoplastic corpus callosum, multisystem degeneration and ubiquinated eosinophilic granules. Acta Neuropathol. (Berl.) 101:69, 2001. Wakamatsu, N., Yamada, Y., Yamada, K., et al.: Mutations in SIP1, encoding Smad interacting protein-1, cause a form of Hirschsprung disease. Nat. Genet. 27:369, 2001. Wang, X., Herberg, F. W., Laue, M. M., et al.: Neurobeachin: a protein kinase A-anchoring, beige/Chediak-Higashi protein homolog implicated in neuronal membrane traffic. J. Neurosci. 20:8551, 2000. Wang, Z. M., Dai, C. F., Kanoh, N., et al.: Apoptosis and expression of Bcl-2 in facial motoneurons after facial nerve injury. Otol. Neurotol. 23:397, 2002. Warren, S. T.: The expanding world of trinucleotide repeats. Science 271:1374, 1996. Watkins, P. J., Gayle, C., Alsanjari, N., et al.: Severe sensoryautonomic neuropathy and endocrinopathy in insulindependent diabetes. Q. J. Med. 88:795, 1995. Wells, M. R.: Changes of ornithine decarboxylase activity in dorsal root ganglion cells after axon injury: possible relationship to alterations in neuronal chromatin. Exp. Neurol. 95:313, 1987. Wells, M. R., and Vaidya, U.: Morphological alterations in dorsal root ganglion neurons after peripheral axon injury: association with changes in metabolism. Exp. Neurol. 104:32, 1989. White, C. M., Greensmith, L., and Vrbova, G.: Repeated stimuli for axonal growth causes motoneuron death in adult rats: the effect of botulinum toxin followed by partial denervation. Neuroscience 95:1101, 2000. White, F. A., Bennett-Clarke, C. A., Macdonald, G. J., et al.: Neonatal infraorbital nerve transection in the rat: comparison of effects on substance P immunoreactive primary afferents and those recognized by the lectin Bandierea simplicifolia-I. J. Comp. Neurol. 300:249, 1990. Wierzbicki, A. S., Mitchell, J., Lambert-Hammill, M., et al.: Identification of genetic heterogeneity in Refsum’s disease. Eur. J. Hum. Genet. 8:649, 2000. Wijsman, J. H., Jonker, J. R., Keijzer, R., et al.: A new method to detect apoptosis in paraffin sections: in situ end-labeling of fragmented DNA. J. Histochem. Cytochem. 41:7, 1993. Wiley, R. G., Blessing, W. W., and Reis, D. J.: Suicide transport: destruction of neurons by retrograde transport of ricin, abrin, and modeccin. Science 216:889, 1982. Wiley, R. G., and Stirpe, F.: Neuronotoxicity of axonally transported toxic lectins, abrin, modeccin and volkensin in rat peripheral nervous system. Neuropathol. Appl. Neurobiol. 13:39, 1987. Willemsen, R., Tybulewicz, E., Sidransky, E., et al.: A biochemical and ultrastructural evaluation of the type 2 Gaucher mouse. Mol. Chem. Neuropathol. 24:179, 1995. Willner, J. P., Grabowski, G. A., Gordon, R. E., et al.: Chronic GM2 gangliosidosis masquerading as atypical Friedreich’s ataxia: clinical, morphologic and biochemical studies of nine cases. Neurology 31:787, 1981.

732

Pathology of the Peripheral Nervous System

503. Winchester, B., Young, E., Geddes, S., et al.: Female twin with Hunter disease due to non-random inactivation of X-chromosome: a consequence of twinning. Am. J. Med. Genet. 44:834, 1992. 504. Wong, J., and Oblinger, M. M.: A comparison of peripheral and central axotomy effects on neurofilament and tubulin gene expression in rat dorsal root ganglion neurons. J. Neurosci. 10:2215, 1990. 505. Wong, J., and Oblinger, M. M.: Differential regulation of peripherin and neurofilament gene expression in regenerating rat DRG neurons. J. Neurosci. Res. 27:332, 1990. 506. Wood, K. A., and Youle, R. J.: The role of free radicals and p53 in neuron apoptosis in vivo. J. Neurosci. 15:5851, 1995. 507. Woods, B., and Schaumburg, H. H.: Nigrospinodental degeneration with nuclear ophthalmoplegia: an unique and partially treatable clinicopathological entity. J. Neurol. Sci. 17:149, 1972. 508. Worrall, B. B., Rowland, L. P., Chin, S. S., and Mastrianni, J. A.: Amyotrophy in prion diseases. Arch. Neurol. 57:33, 2000. 509. Wyllie, A. H.: Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284:555, 1980. 510. Wyllie, A. H., Morris, R. G., Smith, A. L., and Dunlop, D.: Chromatin cleavage in apoptosis: association with condensed chromatin morphology and dependence on macromolecular synthesis. J. Pathol. 142:67, 1984. 511. Xian, C. J., and Zhou, X.-F.: Neuronal-glial differential expression of TGF-␣ and its receptor in the dorsal root ganglia in response to sciatic nerve lesion. Exp. Neurol. 157:317, 1999. 512. Xu, G., Pierson, C. R., Murakawa, Y., and Sima, A. A. F.: Altered tubulin and neurofilament expression and impaired axonal growth in diabetic nerve regeneration. J. Neuropathol. Exp. Neurol. 61:164, 2002. 513. Xu, Y., Sladky, J. T., and Brown, M. J.: Dose-dependent expression of neuronopathy after experimental pyridoxine intoxication. Neurology 39:1077, 1989. 514. Yamada, E., Kataoka, H., Isozumi, T., and Hazama, F.: Increased expression of phosphotyrosine after axotomy in the dorsal motor nucleus of the vagus nerve and the hypoglossal nucleus. Acta Neuropathol. (Berl.) 88:14, 1994. 515. Yamada, M., Hayashi, S., Tsuji, S., and Takahashi, H.: Involvement of the cerebral cortex and autonomic ganglia in Machado-Joseph disease. Acta Neuropathol. (Berl.) 101:140, 2001. 516. Yamamoto, T., Iwasaki, Y., Konno, H., and Kudo, H.: Primary degeneration of motor neurons by toxic lectins conveyed from the peripheral nerve. J. Neurol. Sci. 70:327, 1985. 517. Yan, Q., Elliott, J., and Snider, W. D.: Brain-derived neurotrophic factor rescues spinal motor neurons from axotomyinduced cell death. Nature 360:753, 1992. 518. Ygge, J.: Neuronal loss in lumbar dorsal root ganglia after proximal compared to distal sciatic nerve resection: a quantitative study in the rat. Brain Res. 478:193, 1989.

519. Yiangou, Y., Facer, P., Baecker, P. A., et al.: ATP-gated ion channel P2X(3) is increased in human inflammatory bowel disease. Neurogastroenterol. Motil. 13:365, 2001. 520. Yiangou, Y., Facer, P., Birch, R., et al.: P2X3 receptor in injured human sensory neurons. Neuroreport 11:993, 2000. 521. Yiangou, Y., Facer, P., Smith, J. A., et al.: Increased acidsensing ion channel ASIC-3 in inflamed human intestine. Eur. J. Gastroenterol. Hepatol. 13:891, 2001. 522. Yokota, T., Uchihara, T., Kumagai, J., et al.: Postmortem study of ataxia with retinitis pigmentosa by mutation of the alpha-tocopherol transfer protein gene. J. Neurol. Neurosurg. Psychiatry 68:521, 2000. 523. Yoneda, E., Inagaki, S., Hayashi, Y., et al.: Differential regulation of manganese and copper/zinc superoxide dismutases by the facial nerve transection. Brain Res. 582:342, 1992. 524. Yoshimura, I., Yoshimura, N., Hanazono, T., et al.: An autopsy case of neuronal type Charcot-Marie-Tooth disease (HMSN type II) with nerve deafness and psychiatric symptoms [in Japanese]. No To Shinkei 44:571, 1992. 525. Zhang, X., Ju, G., Elde, R., and Hökfelt, T.: Effect of peripheral nerve cut on neuropeptides in dorsal root ganglia and the spinal cord of monkey with special reference to galanin. J. Neurocytol. 22:342, 1993. 526. Zhang, X., Verge, V., Wiesenfeld-Hallin, Z., et al.: Nitric oxide synthase-like immunoreactivity in lumbar dorsal root ganglia and spinal cord of rat and monkey and effect of peripheral axotomy. J. Comp. Neurol. 335:563, 1993. 527. Zhou, X.-F., Deng, Y.-S., Chie, E., et al.: Satellite-cell-derived nerve growth factor and neurotrophin-3 are involved in noradrenergic sprouting in the dorsal root ganglia following peripheral nerve injury in the rat. Eur. J. Neurosci. 11:1711, 1999. 528. Zhou, X.-F., Rush, R. A., and McLachlan, E. M.: Differential expression of the p75 nerve growth factor receptor in glia and neurons of the rat dorsal root ganglia after peripheral nerve transection. J. Neurosci. 16:2901, 1996. 529. Zhuchenko, O., Bailey, J., Bonnen, P., et al.: Autosomal dominant cerebellar ataxia (SCA6) associated with small polyglutamine expansions in the ␣ 1A-voltage-dependent calcium channel. Nat. Genet. 15:62, 1997. 530. Zigmond, R. E.: LIF, NGF, and the cell body response to axotomy. Neuroscientist 3:176, 1997. 531. Ziv, N. E., and Spira, M. E.: Axotomy induces a transient and localized elevation of the free intracellular calcium concentration to the millimolar range. J. Neurophysiol. 74:2625, 1995. 532. Ziv, N. E., and Spira, M. E.: Localized and transient elevations of intracellular Ca2⫹ induce the dedifferentiation of axonal segments into growth cones. J. Neurosci. 17:3568, 1997. 533. Zochodne, D. W., Verge, V. M. K., Cheng, C., et al.: Does diabetes target ganglion neurons? Progressive sensory neuron involvement in long-term experimental diabetes. Brain 124:2319, 2001.

32 Pathologic Alterations of Nerves PETER J. DYCK, P. JAMES B. DYCK, AND JANEAN ENGELSTAD

Role of Pathologic Study of Peripheral Nerves Nerve Biopsy Indications for Nerve Biopsy Nerve Biopsy versus Other Neuropathic Evaluations Peripheral Nerves That May Be Biopsied Symptoms after Nerve Biopsy Postmortem Spinal Cord and Nerve Removal and Processing Histologic Methods For Nerve Biopsy For Postmortem Tissue Teased Fiber History Uses Histologic Technique Teasing Technique Fixation and Teasing Artifacts Sampling Evaluation of Teased Fibers Classification of Teased Fibers Comparison of the Frequency of Teased Fiber Abnormalities Electron Microscopic Assessment of Teased Fibers Electron Microscopic Correlation with Teased Fiber Abnormality Morphometry of Neuron Somas Motor Neuron Columns Spinal Ganglia

Intermediolateral Column Sympathetic Trunk Ganglion Morphometry of Peripheral Nerves Histologic Processing Section Thickness and Axon-Myelin Relationship Attribute of Shape of Sampled Sections of Internodes That Best Estimates Average Diameter Myelinated Fiber Composition: Photographic Method Myelinated Fiber Composition: Computer Imaging Method Normal Numbers of Fibers in Control Human Nerves Artifacts of Surgical Removal and of Histologic Preparation Histologic Structures with and without Pathologic Significance Renaut Corpuscles Schwann Cell Inclusions Fiber Alterations Neuronal Degeneration Wallerian Degeneration Axonal Dystrophy Axonal Atrophy, Myelin Remodeling (Secondary Demyelination) and Degeneration Segmental Demyelination and Remyelination Myelin Remodeling in Normal Nerves

ROLE OF PATHOLOGIC STUDY OF PERIPHERAL NERVES The studies possible on nerves taken at biopsy are different from those possible on nerves taken at necropsy. Both may provide important information about the kind and severity of nerve fiber damage, or lack of it, and of

Differences from Secondary Demyelination Debris Removal from the Endoneurium Nerve Regeneration Microfascicular Regeneration Interstitial Pathology of Nerves Edema and Hemorrhage Compression Ischemia Intraneural Injection of a Foreign Substance Inflammatory Demyelinating Disorders Chronic Immune Sensory Polyradiculopathy Multifocal Motor Neuropathy Leprosy Sarcoidosis Amyloidosis Tumor Perineurial Pathology Parenchymatous Pathologic Alterations Neuronal (Axonal) Pathologic Alterations Classification of Neuronal (Axonal) Pathologic Alterations Schwann Cell Pathologic Alterations

interstitial changes, but there are important advantages and limitations of each. Only on biopsied nerve is it possible to do functional, metabolic, or high-quality ultrastructural studies. Also on biopsy specimens it is possible to study a specific stage of disease. Even with biopsy, a degree of autolysis occurs, but it is much less than in necropsy specimens. Furthermore, insights from biopsy 733

734

Pathology of the Peripheral Nervous System

may be used for the diagnosis and management of the patient from whom it was obtained, obviously not possible on specimens taken at death. However, only at necropsy is it possible to simultaneously sample different rostral-to-caudal segmental levels of the peripheral nervous system, such as different levels of the spinal cord; the roots, spinal ganglia, segmental nerves, and plexuses; and different proximal-to-distal levels of limb nerves. Additionally, large samples of tissue are possible. Pathologic study of nerve focuses on alteration from normal of nerve fibers (parenchymatous change) and on interstitial change. It is possible to determine whether nerve fibers are normal in number or decreased (as a result of failure of development, axonal degeneration, or atrophy and degeneration) or increased (e.g., from regeneration). Next, it is possible to determine whether selective classes of fibers are increased or decreased. The changes that fibers may undergo may indicate underlying pathologic mechanisms of disease (e.g., compression, ischemia, metabolic derangement, storage, infection, inflammation, neoplasia, or other). Nerve biopsy is used to diagnose infections such as leprosy and certain inflammatory-immune reactions: inflammatory demyelination, granulomas, infections, necrotizing vasculitis, and tumor. Some authors think that there is less need for nerve biopsy than formerly. However, in our medical practice and with the development of better imaging of focal nerve enlargement and enhancement, we are demonstrating new needs for and uses of nerve biopsy (Figs. 32–1 and 32–2).

NERVE BIOPSY Indications for Nerve Biopsy It may be easier to rationalize when nerve biopsy should not be done than when it should be done. It should not be done (1) before adequate characterization of the neuropathy by history, neurologic examination, and other test approaches have been done and thought about; (2) when the cause of the neuropathic process is known; (3) when the cost and side effects of the procedure cannot be justified by the information that is likely to be obtained; (4) when the disease process appears to be self-limited or improving; and (5) when, in previous cases of a similar kind, nerve biopsies have repeatedly not been helpful. Nerve biopsy may be considered when the patient: (1) appears to have an interstitial process and one of the following diagnoses is suspected but other less invasive tests have been negative or inconclusive: tumor, leprosy, necrotizing vasculitis, granuloma, amyloidosis, lysosomal storage diseases, or other diseases recognizable by their tissue alterations; (2) has an undiagnosed symptomatic and progressive focal nerve lesion and a putative focal nerve enlargement or enhancement of an accessible nerve (for biopsy) has been identified; and (3) has neuropathic symptoms or impairments and determining neuropathic activity is important. Nerve biopsy can be a very helpful diagnostic procedure, but it should be reserved for cases that have been carefully characterized by other clinical approaches first,

FIGURE 32–1 T2-weighted transverse (left) and sagittal (middle) and postgadolinium sagittal (right) magnetic resonance images of the lumbosacral spine from a patient with a polyradiculoneuropathy of unknown cause. The arrows show thickened and enlarged nerve roots, which enhance with contrast. One of these roots was later biopsied subsequently and diagnosed as sarcoidosis (see Fig. 32–2).

735

Pathologic Alterations of Nerves

FIGURE 32–2 Serial transverse paraffin sections of a dorsal lumbar rootlet from a patient with a polyradiculoneuropathy of unknown cause whose magnetic resonance imaging study showed enlarged nerve roots (shown in Fig. 32–1). Top, Granuloma formations (nests of macrophages, arrow) surrounded by intense mononuclear inflammatory infiltrates (hematoxylin and eosin). Middle, The CD68 (KP-1) immunoreactivity confirms that the granuloma cells are macrophages (arrow). Bottom, The CD45 (leukocyte common antigen) immunoreactivity indicates that mononuclear cells are lymphocytes. Based on these biopsy results, the patient was diagnosed as having sarcoidosis. See Color Plate

and these have failed to provide a definite answer and informed judgment is made that nerve biopsy may be useful. It should be done by surgeons experienced in taking such tissue and be evaluated in laboratories that are able to evaluate teased fibers, cryostat sections, and paraffin and epoxy sections and have available facilities to do histochemistry, electron microscopy, and morphometry. The availability of pathologists expert in hematopathology and tumor pathology is also a necessary requirement.

100μ

Nerve Biopsy versus Other Neuropathic Evaluations The role of nerve biopsy cannot be discussed without considering other neuropathic evaluations. Evaluations useful in detecting, characterizing, and judging the severity of peripheral neuropathy are the neurologic history (including assessment of symptoms), neurologic deficits, electromyography (EMG) (nerve conduction and needle

736

Pathology of the Peripheral Nervous System

EMG), quantitative sensation testing (QST), quantitative autonomic examination (QAE), and a variety of other laboratory, roentgenologic, and imaging tests. In recent publications, we have assessed the comparative value of these six evaluations (symptoms, deficits, nerve conduction/EMG, QSE, QAE, and nerve biopsy) in the detection, characterization, and assessment of diabetic polyneuropathy.31,41,52 Is the information provided by these evaluations simply confirmatory of neuropathy or complementary, providing useful information for differential diagnosis? The answer is that the tests provide overlapping but different characterizing information useful in differential diagnosis. Symptoms must be assessed because they reflect the subjective experience of patients. Symptoms (1) are unpleasant, painful, disabling, or worrisome to patients; (2) bring patients to doctors for relief; (3) cannot be inferred adequately from other tests; and (4) may convey the patient’s reaction to primary experiences. Neurologic deficit can be assessed as disparate information that the neurologist uses in formulating an opinion as to kind and severity, as a summated score of neurologic deficit (e.g., the Neuropathy Impairment Score), or by a scale of impairments in acts of daily living. Abnormalities of nerve conduction and EMG are known to be sensitive and specific in the detection of peripheral neuropathy in diabetes and in inherited neuropathy, and they appear to correlate with severity. Electrodiagnostic studies provide considerable information about focal or generalized nerve conduction abnormalities and about degenerative and regenerative events of large-diameter fibers. Electrodiagnostic studies can be used to infer fiber degeneration and loss, fiber regeneration, generalized or focal demyelination, collateral sprouting, reorganization of motor units, and such phenomena as fasciculations and cramps. Generally electrodiagnostic studies are not reliable for diagnosing a specific variety of neuropathy, inferring symptoms, predicting global neurologic deficits, detecting and characterizing small-fiber neuropathy, characterizing specific pathologic alterations of fibers, or characterizing interstitial pathologic alteration of nerve. QST may be used to detect and characterize sensory abnormality at specific sites. It is useful in recognizing abnormality of small-diameter sensory fibers not well assessed by present electrophysiologic techniques. QST is especially useful for recognizing small fiber dysfunction and such phenomena as thermal hyperalgesia. QAE is used to detect and characterize autonomic nerve fiber dysfunction of eye, heart, vessel, gut, sphincters, and sweat glands. In a recent study of diabetics with and without neuropathy, we found that nerve conduction velocities (NCVs) were especially sensitive and objective (uninfluenced by the will of the patient) and therefore were useful as minimal criteria for the diagnosis of polyneuropathy. Vibratory detection

threshold using the 4, 2, 1 stepping algorithm with null stimuli and the CASE IV system (WR Medical Electronics, Stillwater, MN) was also a rapid, nonpainful, and accurate method of determining whether polyneuropathy was present. Assessment of symptoms, summated neurologic deficit (e.g., The Neuropathy Impairment Score), QST, and the summated compound muscle action potential of ulnar, peroneal, and tibial nerves were best for assessing staged severity of neuropathy. Nerve biopsy is useful in detecting and characterizing pathologic changes in nerve fibers and Schwann cells (myelin) as described above. Nerve biopsy may also be used to discriminate monophasic from continuing axonal degeneration, axonal atrophy from loss of large fibers, and primary from secondary demyelination. Nerve biopsy can be used to decide whether fiber loss is focal, multifocal, or diffuse. Nerve biopsy is especially useful in recognizing interstitial pathologic alterations. The neuropathy of leprosy is suspected when focal epithelioid and lymphocyte infiltration of the endoneurium is found and by visualization of Mycobacterium leprae organisms in acid-fast stains. In herpes, poliomyelitis, cytomegalovirus, and human immunodeficiency virus infections, the pattern and localization of the inflammation and cell injury may suggest the diagnosis. Inflammation in nerve may be categorized by the level of nerve affected (ventral horn columns, roots, ganglia, plexuses, or nerves), compartment of nerve involved (endoneurium, perineurium, or epineurium), relation to certain cells’ or tissues’ pathologic reaction (vasculitis, inflammatory demyelination, granuloma, or other), and identity of the inflammatory cells. Nerve biopsy is frequently done for the purpose of diagnosing amyloid deposition. Lysosomal storage diseases may be recognized by characteristic organelles in the tissue.25 Various alterations of vessels and patterns of ischemic injury may be recognized in nerve biopsy specimens, and such information is diagnostically useful and has treatment implications. Necrotizing vasculitis of several types may occur. Vasculitic disorders are classified by the vessels involved (arteries, arterioles, venules, microvessels), by the distribution of involvement, by the inflammatory reaction, and by mechanism. Nerve biopsy may show bleeding or characteristic patterns of ischemic injury of the nerve (see Interstitial Pathology of Nerves below). Nerve biopsy may be needed to define the pathologic basis of focal nerve enlargement or enhancement on magnetic resonance imaging (MRI). Nerve biopsy tissue is of limited or no value in explaining such symptoms as paresthesia, pain, cramps, and other manifestations of neural hyperactivity. Study of nerve tissue must be limited to a small sample of accessible tissue, and it can be done only on a single or at most a few occasions.

Pathologic Alterations of Nerves

Often, detailed study is possible only at death on poorly preserved tissue.

Peripheral Nerves That May Be Biopsied Nerve biopsies may be categorized into those done from sites of focal nerve enlargement or enhancement, as seen on MRI, and those done from sites thought to be affected by disease but not known to show enlargement or enhancement. Most nerve biopsies fall into the latter category. For the first group, the focal nerve enlargement and enhancement determine the site of the biopsy. For the latter category, it is important to choose a nerve that is involved with the neuropathy to be studied, contains sufficient tissue for study, is readily accessible, and the removal of which will not cause excessive side effects, considering the information obtained. Biopsy at Sites of Nerve Enlargement or Enhancement Biopsy of nerve should be considered when progressive focal or multifocal neuropathic symptoms and signs appear to be attributable to focal nerve enlargement or enhancement on MRI, the cause of which remains undiagnosed and for which information about diagnosis, prognosis, and treatment depends on a pathologic diagnosis. The putative cause may be focal infection (leprosy), inflammation (inflammatory demyelination, vasculitis, or granuloma), tumor (lymphoma, perineurioma, neurofibroma, Schwannoma, or other), or other (see Figs. 32–1 and 32–2). As for other biopsies, the physician must decide whether the potential benefits outweigh the morbidity, subsequent symptoms and impairments, and costs. This procedure should only be done at medical centers with the necessary surgical experience and clinical and pathologic facilities to perform these studies adequately. Sites for biopsy may include nerve rootlets, portions of spinal ganglia, mixed segmental nerve, and various proximal-to-distal levels of nerve. Because these fascicular biopsies are taken from mixed limb nerves, some increased weakness, sensory loss, and positive neuropathic sensory symptoms may ensue. Fortunately, these complications are usually minimal. Nevertheless, the patient needs to be adequately forewarned, permission must be obtained, and only a minimal amount (but sufficient for pathologic study) of tissue should be procured. Ideally, two to three rootlets or fascicles are taken together with intervening tissue spanning from above the point of enlargement to within the enlargement. Cutaneous Nerve Biopsied without Focal MRI Enlargement or Enhancement In cases of apparent disease but without focal enlargement or enhancement on MRI, the nerves that may be considered for biopsy include the sural, superficial peroneal,

737

saphenous, cutaneous branch of the radial, medial or lateral cutaneous nerves of the forearm, greater auricular, and other nerves. For multifocal neuropathy (e.g., leprosy, necrotizing vasculitis, amyloidosis, or other interstitial processes), we prefer using a leg sensory nerve (the sural or the superficial peroneal nerve) that is known to be affected. We have generally preferred the sural nerve at the ankle level as the nerve of choice for biopsy because it usually has a generous epineurium and usually contains several small arteries or arterioles. Others have favored the superficial peroneal nerve, often combining the procedure of nerve biopsy with muscle biopsy (of the peroneus longus or brevis) (see Chapter 106). We have had concern about biopsying a large piece of the peroneus brevis muscle because it is an important everter of the foot at the ankle. A whole or a fascicular cutaneous biopsy can be taken. We tend to take whole cutaneous biopsies because whole nerves contain much more epineurium, which contains the larger nerve vessels and therefore provides a better sample for recognizing interstitial events. The nerve that is chosen for biopsy should be shown to be affected by the disease process. This can be determined from the clinical findings, study of nerve conduction, or subjective alterations of skin sensation in the cutaneous distribution of the nerve (e.g., asleep-numbness or prickling when the examining hand is passed over the cutaneous distribution of the nerve). Sural Nerve Biopsy. Sural nerve biopsy is performed under sterile conditions in the operating room, with the extremity shaved, scrubbed, disinfected, and draped as in any general surgical procedure. The surgical team wears gowns, masks, and gloves. Surgical instruments must be delicate and in good working order. Fine suture material, 4-0 or 5-0, and atraumatic needles are used to minimize foreign body reactions. Magnifying glasses or a surgical dissecting microscope are essential for visualizing the nerve fascicles and for accurate detection of bleeding vessels. Because most patients are ambulatory outpatients, preanesthetic sedation is avoided. We prefer local infiltration anesthesia with 0.5% lidocaine (Xylocaine) without epinephrine. General anesthesia is rarely used, being reserved for the hyperactive young child. The sural nerve is most readily exposed at the ankle level. The patient lies prone on the operating table with the ankle, slightly everted, resting on a pillow so that the foot assumes a 90-degree angle with the leg. This places the sural nerve at normal tension. When the calf is compressed, the lesser saphenous vein becomes distended and may become visible or palpable in the trough between the external malleolus and the Achilles tendon. The saphenous vein is the most reliable anatomic landmark for locating the sural nerve, which lies immediately adjacent or deep to the vein at this level.

738

Pathology of the Peripheral Nervous System

Eight to 10 mL of 0.5% lidocaine is infiltrated into the skin and subcutaneous tissue, extending from just behind the external malleolus for a distance of 10 cm proximally. The incision is placed to overlie the course of the saphenous vein. As the tough, semitransparent Scarpa’s fascia is exposed, the saphenous vein can be seen beneath it. When the Scarpa’s fascia directly over the vein is divided and the vein is retracted gently, the sural nerve comes into view, most commonly on the medial and usually deep side of the vein. The bleeding points are then ligated. Occasionally, a tributary vein to the lesser saphenous at the malleolus requires division and ligation because it crosses over the sural nerve. When the lesser saphenous vein goes into spasm, it may be mistaken for the sural nerve because it becomes white like a nerve. This error can be avoided by knowing that the tributaries of the vein come off of the lesser saphenous vein at right angles like the branches of an oak tree, whereas the branches of nerve come off at acute (much smaller than 90-degree) angles, like the branches of an inverted elm tree. After exposure of the sural nerve for 10 cm, a small amount of local anesthetic is injected directly into the sural nerve just proximal to the apex of the incision (Fig. 32–3). A 5-0 silk suture with a curved needle fused to the suture is transfixed through the substance of the nerve at a proximal point of the exposed nerve but just distal to the local anesthetic injection site. The suture is not tied. The free ends of the suture are then used as a retractor. The nerve is then transected just above the suture. Using the free ends of the suture between thumb and finger and avoiding traction (or applying only very light traction), the nerve is undercut with curved scissors, excluding most of the interstitial fat and vessels, for a distance of approximately 3 cm. The bent hook of a 10-mg weight is impaled into the distal end of the mobilized nerve specimen at a point 2.5 to 3 cm distal to the proximal transection. The nerve is then transected just distal to the weight. The nerve specimen is lifted using the retracting suture and is hung directly (in the operating room and without any delay) into 2.5% glutaraldehyde in buffer (see Histologic Methods below) for 5 minutes and then is transferred to 4.0% paraformaldehyde in buffer at 10° C and pH 7.4 (see Fig. 32–3). The next 2.5- to 3-cm segment of nerve (a direct continuation of the most proximal specimen) is similarly transfixed with a suture, undercut, weighted, and hung into 2.5% glutaraldehyde in buffer, the final solution isotonic with plasma and at 10° C and pH 7.4. The final 2.5- to 3-cm piece is immersed into just melting isopentane in a metal beaker suspended on liquid nitrogen. This is stored at ⫺80° C for later studies or is used for cryostat sections. The nerve itself is never tweaked or grasped with forceps. Its edges (the epineurium) may be grasped with fine-pointed forceps. Great care is also taken to avoid stretching the nerve. Removal of a single fascicle as compared to removal of the whole nerve is best performed using a surgical dissecting

FIGURE 32–3 Surgical technique of sural nerve biopsy (see text for details). (From Dyck, P. J., and Lofgren, E. P.: Method of fascicular biopsy of human peripheral nerve for electrophysiologic and histologic study. Mayo Clin. Proc. 41:778, 1966, with permission.)

microscope. The epineurium is slit along the length of the nerve trunk, exposing individual or groups of fascicles. A fascicle (or two or three fascicles), usually one located on the posteromedial aspect of the nerve, is selected for biopsy. A 7-0 silk suture is passed through the fascicle at the upper end of the incision and held as a loop retractor. The fascicle is transected just proximal to the suture. By dissecting the fascicle free of the epineurium and occasionally dividing small bridging fascicles, several fascicular lengths of nerve are obtained. Each of these is suspended in the fixative by the silk suture at one end, and a small weight is hooked into the other end (see Fig. 32–3). After meticulous hemostasis, Scarpa’s fascia is reapproximated, with care not to pass the suture around the remaining nerve. The skin is approximated with subcuticular stitches. A dry dressing is applied and held in place with a 10-cm elastic bandage. We encourage the patient to walk away following the procedure rather than use a wheelchair. He or she is instructed not to engage in sports in the ensuing 3 to 4 days. Analgesics generally are not necessary for postoperative discomfort, but the patient is forewarned that

Pathologic Alterations of Nerves

a surge or twinge of pain may be felt if the sural nerve is stretched, by walking or stooping forward. The dressings may be changed on the day after the procedure. The elastic bandage is worn for protection for 2 weeks. Biopsy of the Superficial Peroneal Nerve and Other Nerves. The superficial peroneal nerve emerges from between the peroneus longus and brevis and passes through the deep fascia in the distal third of the lateral leg, innervating the cutaneous distribution of the lower lateral leg and dorsum of the foot. Its size is comparable to that of the sural nerve. The surgical technique of biopsy of this and other nerves is similar to that described for the sural nerve.

Symptoms after Nerve Biopsy The symptoms from sural nerve biopsy are quite variable among healthy subjects and patients with neuropathy (see Chapter 30 in the second edition of this textbook). Patients with severe pain loss in the feet and legs generally experienced no adverse symptoms of nerve biopsy. Among patients biopsied for neuropathy, 60% had no symptoms 1 year after sural nerve biopsy, 30% had intermittent mild persisting symptoms that then disappeared, and 10% had greater degrees of pain or paresthesia. It is possible that some of the discomfort experienced by patients was related to the neuropathic process and not to the biopsy. One of the authors underwent fascicular biopsy of the sural nerve at the ankle, without injection of local anesthetic directly into the nerve.49 There was no discomfort until transection of the nerve fascicle. Immediately with nerve transection, a sharp, stinging, burning discomfort occurred in a region just below the lateral malleolus and spreading to the Achilles tendon. Although severe, the pain lasted only a few seconds. This pain could have been avoided by local anesthetic injection of the proximal nerve. In the first 3 or 4 postoperative days there was soreness at the site of the incision, especially when the tissues were stretched. On stretching the nerve, as in bending, sharp electric shock–like pain was experienced in the lateral heel. Altered sensibility and a mild raw, burning discomfort were experienced over and below the lateral malleolus and over the Achilles FIGURE 32–4 Anterior approach to spinal cord removal at postmortem examination. Dotted lines indicate planes of saw cut adjusted to shapes of different levels of vertebral column. A, Cervical. B, Thoracic. C, Lumbar. (From Okazaki, H., and Campbell, R. J.: Nervous system. In Ludwig, J. [ed.]: Current Methods of Autopsy Practice, 2nd ed. Philadelphia, W. B. Saunders, p. 95, 1979, with permission.)

739

tendon. This was noticeable especially on prolonged sitting or on taking the first steps after sitting. The pain vanished completely on walking. When the sural nerve was palpated at the level of the proximal transection site, no local discomfort was experienced. In the cutaneous distribution of innervation (below the lateral malleolus and over the lower Achilles tendon), a mild prickling, unpleasant burning sensation was produced. When the affected dermatome was touched, some sensation was elicited. Although the threshold for touch and pinprick was greatly heightened, stronger stimuli produced a diffuse, unpleasant, prickling burning in the affected region. These symptoms gradually faded away, returning intermittently with less intensity over several years. At 5 years the symptoms had disappeared, although touching the affected dermatome still elicited a slightly unpleasant and altered sensation. After approximately 10 to 15 years, sensation had returned to normal.

POSTMORTEM SPINAL CORD AND NERVE REMOVAL AND PROCESSING The spinal cord, roots, spinal ganglia, and a variable extent of segmental nerves can be removed post mortem by a posterior or anterior approach. The following description is Okazaki and Campbell’s121 modification of Kernohan’s method. The anterior approach is generally used because it is preferred by morticians, is more rapid, and allows removal of the spinal cord with roots, spinal ganglia, and plexus nerves in continuity. After evisceration, the block supporting the head is removed and placed under the middle of the thoracic spine to straighten it. A transverse cut is made through the first or second thoracic vertebra using an oscillating saw. With the oscillating saw, cuts are made along the sides of the vertebra, holding the blade at the angles shown in Figure 32–4. The cut is made just through the bone to avoid injury to roots or spinal cord. The angle changes from almost vertical to almost horizontal as one passes from the cervical through the thoracic to the lumbar spine. Okazaki suggests short cuts on both sides of the vertebra, beginning in the thoracic region

740

Pathology of the Peripheral Nervous System

before proceeding to more caudal levels. With the body properly positioned and with adequate cuts, the vertebral bodies will spring forward away from the spinal cord toward the prosector. The freed spine can be grasped with the noncutting hand as the cuts are extended caudally. As one proceeds to the lumbar region, the muscles must be cut away from the spine, making certain that one does not transect the emerging nerve roots. The lumbar spine is transversely cut at the L4-L5 interspace with a curved short knife. The L5 vertebral body is then cut away using the oscillating saw, chisel, and mallet. The nerve roots at the lumbosacral interspace can be cut across or can be dissected out after the bone is chipped away with the oscillating saw and rongeur. The exposed spinal cord and cauda equina, covered by the dura mater, can be lifted out of the spinal canal. If the bony floor of the spinal ganglia has been chipped away, the spinal ganglia, segmental nerves, and lumbar and sacral plexus can be taken as one. A suture may be used to transfix one of the nerve roots (e.g., L5, identified by its point of exit) to be able to correctly identify specific roots.

HISTOLOGIC METHODS For Nerve Biopsy Sural or other nerves taken at biopsy are suspended in fixative in the operating room (described above). On returning to the laboratory, excess fat is trimmed away. The 2.5-cm length for embedding into paraffin may be initially fixed in 2.5% glutaraldehyde for 5 to 10 minutes, then immersed in 4% paraformaldehyde in buffer. This tissue is fixed overnight, taking care that there is a large excess of fluid relative to tissue (ⱖ25:1). If the fixative solution becomes bloody on immersion of the nerve specimen, it is replaced by fresh fixative solution. After fixation and washing, 3- to 4-mm-long collars of tissue are cut transversely (to the longitudinal axis of fascicles on a sheet of dental wax) using a degreased razor blade. The collars are carefully embedded into paraffin so that transverse and longitudinal sections will result (Fig. 32–5). These paraffin sections are stained with hematoxylin and eosin, trichrome, Luxol fast blue, periodic acid–Schiff, methyl violet, Congo red, and an iron stain (e.g., Turnbull blue). Other sections are histochemically reacted for CD45 (leukocyte common antigen), for CD68 (macrophages), or for other markers. Other histochemical reactions for amyloid proteins, Schwann cell proteins (e.g., S100), epithelial membrane antigen, vascular wall markers (smooth muscle actin), and others may be obtained depending on need. The 2.5-cm length to be used for preparation of teased fibers and for embedding into epoxy is fixed in 2.5% glutaraldehyde in 0.025 mol/L cacodylate buffer at pH

7.38 and 10° C. The solution is isotonic to prevent uneven shrinkage of fascicles and axons, and we fix this tissue at 10° C—a compromise to avoid the depolymerization of microtubules (MTs), which occurs at 0° to 4° C, and to minimize autolysis, which is greater at room temperature than at low temperatures. In the operating room or shortly after return to the histology laboratory, the length of nerve fixed in buffered glutaraldehyde must be divided into two approximately equal lengths. For full-thickness biopsies this transection occurs 1 to 2 hours after surgery; for fascicular biopsy specimens it occurs 20 to 45 minutes after surgery. At the time of this transection, excess fat is removed and the specimen to be used for teased fibers is split into individual fascicles using a razor blade with a single cutting edge and on dental wax (see Fig. 32–5). The duration of aldehyde fixation for the sural nerve specimens to be embedded into epoxy varies from 4 to 24 hours depending on the size of the specimen. One collar is left undivided and another is divided into individual fascicles and embedded into epoxy for transverse sections. Another collar is embedded for longitudinal sections (see Fig. 32–5). The blocks are washed six times, for 5 minutes each, in isotonic cacodylate buffer; osmicated for 1.5 to 3 hours, depending on the tissue block size; and dehydrated, infiltrated, and embedded into epoxy. In preparing tissue blocks, one avoids splitting fascicles because the architecture of the nerve is better maintained with the perineurium left intact. Special molds are used to get true transverse and longitudinal sections (see Fig. 32–5). Semithin sections (1.0 ␮m) can be viewed with a phase contrast microscope without further staining or with a light microscope after staining with methylene blue (or toluidine blue) or paraphenylenediamine. Nerve fascicles are surrounded by perineurium, which, in concert with endothelial cells, maintains a special endoneurial environment. Dyck and colleagues found in a series of reported studies that immersion fixation with hyperosmolar glutaraldehyde fixation resulted in serious distortions not only of the transverse cross-sectional shape of fascicles but also of nerve fibers.50 Such fixation resulted in flattened, crescentic, and fluted shapes of the transverse profiles of fascicles and fibers—markedly different than the round profile of these structures in cryostat sections (Fig. 32–6).

For Postmortem Tissue Ideally, nerve tissue should be obtained as soon after death as possible. Very occasionally it may be possible to do this within 1 to 4 hours of death. Usually it is obtained within 6 to 12 hours of death. It is advisable to obtain separate permission to remove limb nerves. The medical record, especially the neurologic and electromyographic examination, should be reviewed in order to take tissue from appropriate anatomic sites. Care should be taken to sample known

741

Pathologic Alterations of Nerves

Epoxy embedding Whole nerve A1T1 Fascicles Proximal Distal

For transverse sections

For transverse sections A2T1, A2T2, ... A2Tn

Rinse Osmicate Dehydrate Infiltrate

Embed into epoxy For longitudinal sections

For longitudinal sections A3L1, A3L2, ... A3Ln Teased fibers see specific section

Rinse Osmicate Glycerinate Tease

See Fig. 32.6

Paraffin embedding Whole nerve Rinse Dehydrate Infiltrate

FIGURE 32–5 Steps used in processing nerve biopsy tissue for epoxy and paraffin embedding and for preparation of teased fibers as described in text.

Stains Methylene Blue Phenylenediamine Unstained - phase contrast Uranyl acetate and lead citrate for tem

Embed into paraffin

Stains Hematoxylin + Eosin Luxol Fast Blue - Periodic Acid Schiff Methyl Violet Congo Red Turnbull Blue Other Immunohistochemistry CD68 (KP-1) CD45 (LCA) Other

Freeze a block in chilled isopentane

anatomic segments and different proximal-to-distal levels of the peripheral nervous system and their central nervous system (CNS) extensions. It may be important to sample levels for which control values are present. For biochemical study, portions of peripheral nerve, of brain, and of parenchymatous organs should be obtained quickly after death and preserved in liquid nitrogen. Even with postmortem material, tissue may be obtained for teased fiber preparations. For this purpose, the nerve selected should be one that is affected by the disorder and one for which normative morphometric measurements are available.

The many histologic methods that are used for postmortem peripheral nerve tissue can be found in such books as the one by Romeis130 and the manual of the Armed Forces Institute of Pathology.108 A few histologic procedures that are especially helpful are described here. Further processing depends on what use will be made of the tissue. Spinal cord, spinal ganglia, and autonomic ganglia may be embedded by segment or by ganglion into celloidin so that serial sections can be cut and systematically sampled ones can be used for morphometric analysis (see Morphometry of Peripheral Nerves below). Celloidin is ideal for this purpose because heat is not employed and

742

Pathology of the Peripheral Nervous System

FIGURE 32–6 Transverse sections of human sural nerve fixed by immersion in buffered glutaraldehyde to show the reproducible alteration in fascicular and myelinated fiber shape when hyperosmolar fixation is used. The nerve on the left has been fixed in isosmolar and that on the right in hyperosmolar fixative. The fascicles and fibers assume flattened, crescentic, and other shapes when fixed in hyperosmolar solutions. (From Dyck, P. J., Low, P. A., Sparks, M. F., et al.: Effect of serum hyperosmolality on morphometry of healthy human sural nerve. J. Neuropathol. Exp. Neurol. 39:285, 1980, with permission.)

shrinkage appears to be minimal. Cresyl violet, which stains nuclei and cytoplasm well, is employed for such studies. Other portions may be embedded into paraffin, methacrylate, or epoxy. For nerve trunks one may wish to fix and process serial blocks differently. The first block may be fixed in aldehyde, embedded in paraffin or methacrylate, and stained so that it can be evaluated for interstitial pathologic abnormality. The next block is cut up into smaller pieces and is additionally fixed in osmium tetroxide and embedded in epoxy to be able to evaluate semithin and thin sections. The third block may then be used for study of teased fibers. Serial blocks may be evaluated in this way along the length of the nerve.

TEASED FIBER History The use of teased single or small strands of myelinated fibers (MFs) to study anatomic features and pathologic abnormalities dates back more than 100 years. Early investigators of peripheral nerve—Remak, Ranvier, Kölliker, Renaut, Gombault, Retzius, and Ramon y Cajal—illustrated both normal and abnormal features of MFs using teased fibers.90,128,130 Even though the preparation and analysis of teased fibers are time consuming, teased fiber analysis has been reintroduced, especially for evaluation of MF

743

Pathologic Alterations of Nerves

anatomy153 and neuropathology. Introduction and use of improved fixatives,131 simpler techniques of preparing teased fibers, a descriptive classification of pathologic abnormalities,48 and systematic sampling of fibers and statistical treatment of results48 have permitted wider and more diverse use and validation of results. Analysis of teased fibers is so helpful in characterizing fiber changes that we perform it on all nerve biopsies we assess.

Uses The procedure (1) permits evaluation of consecutive internodes or segments of the same myelinated nerve fiber over long distances; (2) may be used to establish the frequency of certain neuropathologic fiber abnormalities and hence detect neuropathic abnormality generally with more sensitivity than by use of fixed sections; (3) may serve to show whether there is ongoing fiber degeneration; (4) may indicate whether fiber branching or sprouting is occurring; and (5) allows the performance of ultrastructural studies at preselected sites of focal pathologic abnormality, if fibers have been embedded in epoxy.35,146 Obviously it cannot be used to study unmyelinated fibers.

Histologic Technique Sural nerve biopsy, described in a previous section, should be performed with care to avoid crush and stretch. The sural nerve specimen, with the proximal end cut at right angles and the distal end cut obliquely, with a suture attached to one end and a light stainless steel weight attached to the other end, is suspended in freshly prepared 2.5% glutaraldehyde fixative in 0.025 mol/L cacodylate buffer, pH 7.38, at a temperature of 10° C for 20 minutes (single fascicle) or for 60 to 120 minutes (full-thickness nerve). The duration of aldehyde fixation is critical. Insufficient aldehyde fixation will later be associated with breaks or separation at nodes of Ranvier and at Schmidt-Lanterman incisures. Excessive fixation may make it difficult to tease apart single fibers or small strands of fibers and may cause breaking and splitting of fibers. After aldehyde fixation, the full-thickness nerve specimen is divided into its component single fascicles using a razor blade on dental wax. Epineurial fat is cut away. The fascicles are washed (six times, each for 5 minutes, in isotonic cacodylate buffer) and osmicated for 1.5 to 3 hours. The fascicles are carried through 45%, 66%, and 100% glycerin each for 24 hours and at 45° C.

Teasing Technique Our (Dyck, Lais, and Engelstad) method of preparing teased fibers is illustrated in Figure 32–7. On lightly glycerinated glass slides and with curved, pointed forceps and using a dissecting microscope, epineurium and perineurium are stripped away from the fascicular bundle of nerve fibers

(endoneurium). Next, small strands of nerve fibers are torn from the endoneurium. From these strands, single fibers or a bundle of a small group of fibers are pulled away from the main strand, grasping the proximal ends of each strand with curved pointed forceps. For a right-handed person, the left forceps grasps the proximal end of the main strand and remains motionless while the right one, grasping the small strand, traces an inverted U-shaped pathway. The tip of the forceps is kept in constant contact with the glass slide to prevent the fiber from curling up around it. Teasing is always performed from the proximal end of the fascicle so that branches are not torn off. A clean slide is laid adjacent to the right of the slide on which the teasing has been done. The proximal end of each separated fiber strand is grasped with forceps and slid onto the clean slide through a minute drop of glycerin; again, the tip of the forceps is kept constantly touching the glass. As the middle of the new slide is reached, the points of the forceps are allowed to spread apart. The friction of the fiber strand on the glass will straighten the fiber and cause it to pull away from the tips of the forceps. Fifty teased fiber strands are placed side by side in the center of each of two glass slides, allowing for an analysis of 100 teased fibers in most cases. If the proper technique is used, the proximal ends of the fibers will always be oriented in the same way, the fibers will be straight, and there will be a minimum of glycerin with them. The slides are air dried overnight in a warm oven. With observation under the dissecting microscope, a drop of mounting medium is applied to the coverslip, so that the apex of this drop is at the center of the fiber. The coverslip is dropped onto the slide. As the drop spreads outward, it will carry most of the glycerin at its moving front, displacing it to the edges of the coverslip. The coverslip edges may be sealed with nail polish. Using these methods on suitably prepared tissue, an experienced, diligent person can prepare 150 to 400 teased fibers per day. An elegant display of teased fibers laid side by side is shown in Figure 32–8 for a control nerve. A nerve with most fibers undergoing early axonal degeneration is shown in Figure 32–9. Figure 32–10 is from a nerve with most fibers undergoing late degeneration. The nerve in Figure 32–11 shows many large tomacula, with myelin reduplication typical of hereditary neuropathy with a tendency to pressure palsy. It is important to note that these figures represent only a quarter or so of the length of the fibers teased but, because they are isolated and lie flat on glass, each fiber can be visualized at all points along its length.

Fixation and Teasing Artifacts It is important to recognize that various artifacts, as shown in Figure 32–12, can be produced by the method described here. Excessive stretching of the nerve during biopsy or of

744

Pathology of the Peripheral Nervous System

FIGURE 32–7 From left to right and from top to bottom, consecutive steps in fiber teasing: a fascicle of nerve, fixed in glutaraldehyde and osmium tetroxide, lying in a pool of glycerin on a glass slide; proximal ends are grasped and fascicles are pulled apart; epineurium and perineurium are stripped off; strands of fibers are pulled apart; from separated strands of nerve, a single teased fiber is slid onto an adjacent slide as described in the text; and teased fibers in place under coverslip.

the fiber during teasing, when the tissue is inadequately fixed, produces separation at nodes of Ranvier (see Fig. 32–12, frame 2) and at Schmidt-Lanterman incisures. In addition, the fiber may assume a varicose appearance that mistakenly may be assumed to be due to pathologic abnormality. Separation at nodes and at Schmidt-Lanterman incisures is particularly common when this method of teasing is used on tissue fixed with osmium tetroxide alone; with correct aldehyde fixation, the artifact is not prominent. Any morphologic change observed at the ends of teased fibers

should be ignored because these may represent damage by the forceps. With overfixation with glutaraldehyde, splitting of fibers may occur (see Fig. 32–12, frames 3 through 5).

Sampling Before beginning to tease the specimen, the technician must decide how many fibers per fascicle are to be assessed and how representative sampling is to be

Pathologic Alterations of Nerves

745

FIGURE 32–10 Low- (upper) and intermediate- (lower) power light microscopic views of teased fibers showing late axonal degeneration (condition E). All fibers on this slide are undergoing degeneration. FIGURE 32–8 Low- (upper) and intermediate- (lower) power light microscopic views of about one-quarter of the length of teased fibers laid side by side to illustrate that the method of teasing allows fibers to be visualized for long distances along their length by our technique of preparation. The teased fibers are from the sural nerve of a healthy subject using fixatives and approaches outlined in this chapter. All fibers would be graded as normal (condition A). (Modified from Dyck, P. J., Dyck, P. J. B., Giannini, C., et al.: Peripheral nerves. In Graham, D. I., and Lantos, P. L. [eds.]: Greenfield’s Neuropathology, 7th ed., Vol. 2. London, Arnold Publishing, p. 551, 2002.)

FIGURE 32–9 Low- (upper) and intermediate- (lower) power light microscopic views of teased fibers from the sural nerve of a patient with necrotizing vasculitis showing severe axonal degeneration (condition E). (Modified from Dyck, P. J., Dyck, P. J. B., Giannini, C., et al.: Peripheral nerves. In Graham, D. I., and Lantos, P. L. [eds.]: Greenfield’s Neuropathology, 7th ed., Vol. 2. London, Arnold Publishing, p. 551, 2002.)

achieved. In health, a fascicle of sural nerve may contain from 500 to 2500 MFs. The sensitivity and reliability of the estimate of pathologic abnormalities and the variability of the estimate will depend on the rate and distribution of abnormality and on the sampling procedure. The error will decrease with increasing numbers of fibers sampled. For clinical and many research applications, at least 100 fibers/nerve should be sampled. This number was chosen to get representative results, sufficient reliability for detection of neuropathic abnormality, and a reasonable expenditure of time and money. For certain research

FIGURE 32–11 Low- (upper) and intermediate- (lower) power light microscopic views of teased fibers showing typical tomacula (condition G) in closely applied teased fibers from a patient with hereditary neuropathy with tendency to pressure palsy. (Modified from Dyck, P. J., Dyck, P. J. B., Giannini, C., et al.: Peripheral nerves. In Graham, D. I., and Lantos, P. L. [eds.]: Greenfield’s Neuropathology, 7th ed., Vol. 2. London, Arnold Publishing, p. 551, 2002.)

746

Pathology of the Peripheral Nervous System

FIGURE 32–12 Teased fiber preparations (⫻380). 1, Five consecutive portions along the length of a teased fiber. Nodes of Ranvier are seen in first and fifth strips. Unstained Schwann cell nucleus is seen in the third strip. 2, Three consecutive portions along the length of a teased fiber, showing excessive separation at nodes of Ranvier (top and bottom strips), separation at Schmidt-Lanterman incisures, and undulating thickness of fiber, all caused by improper fixation (osmium tetroxide alone) for this method of teasing. 3, Two adjacent teased fibers with the upper showing patchy loss of myelin, an artifact encountered during teasing when excessive hardening has occurred with glutaraldehyde fixation. 4, A more severe artifact in which the lower right half of the fiber has been split and pulled away during teasing from tissue that is excessively hard. 5, Artifact in the middle of strip of fiber caused by grasping with forceps. These artifacts are inevitable in this method and should be ignored. This fiber also shows the effect of excessive stretching; the undulating appearance comes from excessive stretching before adequate fixation.

purposes, as many as 500 fibers/nerve might be prepared. It may be preferable to increase the number of nerves evaluated as compared to increasing the number of teased fibers evaluated per nerve (beyond 100). Approaches to sample teased fibers randomly have generally not been

used. We suggest systematic sampling by approaches described here. This is accomplished by initially dividing the fascicle into 10 thick strands of approximately equal thickness. Each of these thick strands is then subdivided into five intermediate-thickness strands. From the right-hand

747

Pathologic Alterations of Nerves

side of each of these 50 strands, two small strands (without selection) containing one, several, or no (empty) MFs is teased. For certain purposes, it may be desirable to assess the frequency of abnormalities in teased fibers taken from the edges of fascicles versus those from the center (e.g., in ischemic neuropathy). To do so, the outer one sixth of the endoneurial portion is stripped from the right and then the left side of the fascicle. The fascicle is then rotated 90 degrees on its long axis, and again the outer one sixth is stripped from the left and right sides. The four strands stripped from the edges are then divided as described in the preceding paragraph so that 100 fibers are systematically sampled. Next, the center of the fascicle is subdivided in order to grade 100 fibers. A direct comparison of the frequencies of pathologic abnormalities from the edges and centers of fascicles is then possible.

Evaluation of Teased Fibers Assuming that teased fibers have been prepared as outlined in the preceding sections, four types of evaluation are possible. In the first, the percent of teased endoneurial strands containing MFs or their breakdown products can be estimated. In control nerves MFs will be present in most if not all of the teased nerve strands. This evaluation therefore provides a rough estimate of the density of MFs. In the second, the frequency distribution of descriptive pathologic alterations of MFs is estimated. In the third, internode length (IL) and caliber are assessed. In the fourth, ultrastructural alterations are assessed at predetermined sites of teased fibers. The second, third, and fourth approach are discussed below.

FIGURE 32–13 Drawings of consecutive lengths, from top to bottom, of teased myelinated fibers. A, Fiber in condition A as described in text. B, Fiber also in condition A, but note that the internode lengths are short compared with fiber shown above. This may be a regenerated fiber; however, because of the considerable variability in mean internode lengths between fibers of a given diameter even in healthy nerve, it is not possible with certainty to designate this fiber as being regenerated.

Classification of Teased Fibers Dyck and co-workers introduced the following classification of descriptive teased fiber pathologic abnormalities38 to provide objective criteria of teased fiber changes observed based solely on light microscopic observation without need for sophisticated measurement and independent of views related to mechanism of fiber pathology. The pathologic abnormalities that were codified were ones that had been encountered in human and experimental neuropathy. Because these abnormalities had also been seen, albeit at low frequencies, in nerves of healthy subjects, it is incorrect to declare a nerve “abnormal” simply by the occurrence of these abnormalities. Whether the nerve is abnormal or not depends on whether rates of teased fiber abnormalities exceed what is expected (from study of control nerve) for the nerve and the age of the patient. Condition A: Normal Fiber. This is teased fiber of normal appearance, ignoring criteria of IL and of internode diameter (ID) (Fig. 32–13). Myelin is not more irregular than that encountered in most MF internodes of control nerves. This judgment assumes that the observer is aware that myelin irregularity varies with the preparation used, the species, the nerve assessed, and age. In the young, myelin is more regular than it is in the old. The average thickness of myelin of the internode with the thinnest myelin is 50% or more of that of the internode with the thickest myelin. No paranodal or internodal segmental demyelination is seen. Condition B: Myelin Wrinkling. This is teased fiber with excessive irregularity, wrinkling, and folding of myelin but with the other features of condition A (Fig. 32–14).

748

Pathology of the Peripheral Nervous System

FIGURE 32–14 Drawing of consecutive lengths, from top to bottom, of a teased myelinated fiber in condition B as described in text.

This judgment assumes that the observer is familiar with the degree of myelin regularity considering species, nerve, site, and age. Condition C: Segmental Demyelination. This is teased fiber with a region or regions of paranodal or internodal segmental demyelination with or without myelin ovoids or balls in the cytoplasm of the associated Schwann cells* (Fig. 32–15). Thickness of myelin of the internode with the thinnest myelin is 50% or more of that of the internode with the thickest myelin. Myelin of internodes may be regular or irregular. Condition D: Segmental Demyelination and Remyelination. These are teased fibers with a region or regions of paranodal or internodal segmental demyelination with or without myelin ovoids in the cytoplasm of the associated Schwann cells and with remyelination. Thickness of myelin of the internode with the thinnest myelin is less than 50% of that of the internode with the thickest myelin (remyelination). Myelin of internodes may be regular or irregular (Fig. 32–16). Condition E: Axonal Degeneration. This is a teased strand of nerve tissue with linear rows of myelin ovoids and balls at the same stage of degeneration along the length of the teased fiber (Fig. 32–17). Fibers can be segmented into large ovoids, as seen in early axonal degeneration, or into widely separated small myelin balls, as is seen in late axonal degeneration.

*In demyelination, as judged by the high-dry objective of the light microscope, no myelin can be recognized. From our previous evaluation of teased fibers under the electron microscope, approximately four lamellae of myelin can be recognized as a thin dark line under the light microscope except when there is excessive tissue around the fiber. By paranodal demyelination it is implied that the site of the node of Ranvier is recognized and that the nodal gap is increased beyond that seen in normal fibers. In internodal demyelination the entire former internode is demyelinated.

Condition F: Remyelination. This is teased fiber without a region or regions of segmental demyelination but with excessive variability of myelin thickness among internodes. Thickness of myelin of the internodes with the thinnest myelin is less than 50% of that of the internode with the thickest myelin (Fig. 32–18). Myelin of internodes may be regular or irregular. Condition G: Tomacula. This is teased fiber with excessive variability of myelin thickness within internodes as a result of myelin reduplication to form “globules” or “sausages” (Fig. 32–19). Although the myelin thickness in the regions of the globule42 is greater, it probably merely represents a process of infolding and reduplication of myelin rather than excessive production of myelin. Frequently, some regions of the fiber have excessively thick myelin, whereas other regions have excessively thin myelin. Condition H: Regeneration after Axonal Degeneration. This is teased fiber of normal appearance as described in condition A, but in which there are myelin ovoids or balls contiguous to two or more internodes (Fig. 32–20). Condition I: Proximal Segmental Demyelination and Distal Axonal Degeneration. This is teased fiber having several proximal internodes or parts of internodes with or without paranodal or internodal segmental demyelination and, distal to these, a linear row of myelin ovoids or balls (Fig. 32–21). This condition implies axonal dystrophy proximally and distal degeneration. Empty Nerve Strands. These are strands of teased fibers without myelin or myelin breakdown products and without another classifiable fiber condition. They are the by-products of fibers after they have undergone axonal degeneration. These strands are tallied but are not included in the total of classifiable fibers.

FIGURE 32–15 Drawing of consecutive lengths, from top to bottom, of a teased myelinated fiber in condition C as described in text. Note that the myelin thickness is not as variable as it is in condition D.

FIGURE 32–16 Drawings of consecutive lengths, from top to bottom, of a teased myelinated fiber in condition D as described in text. Note that the ovoids of the internode in which the myelin has degenerated are much smaller in diameter than the large ovoids that form initially in condition E. Note also regions of segmental demyelination and the considerable variability of myelin thickness resulting from the presence of “old” and “new” internodes.

A

FIGURE 32–17 Drawings of consecutive lengths, from top to bottom, of a teased myelinated fiber with condition E as described in text. A, Segmentation into large ovoids. B, Mixture of myelin ovoids and balls of intermediate size. C, Clusters of small myelin balls occurring at widely separated points.

B

C

750

Pathology of the Peripheral Nervous System

A

B

C

Clinical Usefulness of Teased Fiber Assessment Teased fiber assessment is a valuable component of the diagnostic sural nerve biopsy assessment. The evaluation may answer these questions: • Is neuropathy present? • What type of fiber degeneration is taking place?

FIGURE 32–18 Drawings of consecutive lengths, from top to bottom, of three teased myelinated fibers in condition F as described in text. The teased fiber in A shows a single intercalated internode, a not-uncommon finding in fibers from healthy nerve.

• Is there active axonal degeneration and what is its frequency? • Is there evidence of repair? • Is there evidence of sprouting? • Is there evidence of axonal atrophy and secondary segmental demyelination, axonal swellings, tomacula, or focal interstitial effects on fibers? • Is demyelination primary or secondary?

FIGURE 32–19 Drawing of consecutive lengths, from top to bottom, of a teased myelinated fiber in condition G as described in text.

Pathologic Alterations of Nerves

751

FIGURE 32–20 Drawing of consecutive lengths, from top to bottom, of a teased myelinated fiber in condition H as described in text. This condition implies regeneration of a myelinated fiber. This condition is not to be confused with that of the teased fiber that has a single osmiophilic droplet in Schwann cell cytoplasm outside of myelin, which is discussed subsequently as the Elzholz body (␮ granule of Reich).

Comparison of the Frequency of Teased Fiber Abnormalities Determining the percent frequency of abnormality (of 100 teased fibers of a sural nerve) allows comparing teased fibers in disease specimens versus those of control specimens. For the comparisons to be valid, nerves should be taken from the same level of nerve, prepared by the same histologic and teased fiber approaches, and

A

FIGURE 32–21 A, Drawing of consecutive lengths, from top to bottom, of a teased myelinated fiber in condition I as described in text. This type of fiber change is seen typically several days after and at the site of crush. B, The same fiber as shown above after repair has occurred—this would be graded as condition F.

B

graded without the observer knowing whether each came from the disease or control specimen (masked grading). This is readily done by random and coded assignment of slides to be graded with the observer masked as to the identity of the slide. Later, the mean percentage of pathologic abnormalities from disease and control specimens can be compared using standard statistical tests.

752

Pathology of the Peripheral Nervous System

Sources of Variability in Grading Pathologic Abnormalities The degree of myelin irregularity is hard to define and grade objectively because it varies with age, species, nerve, and level of nerve. Assessment of the frequency of condition B may therefore vary considerably among observers. Another source of variability may relate to assessment of nodal length. A wide nodal gap resulting from excessive stretching of improperly fixed nerve may be spuriously judged as segmental demyelination (condition C). A careful preevaluation of myelin regularity and nodal length variability in control nerves, fixed and prepared by similar techniques, helps minimize this problem. A further source of variability relates to distinguishing condition C from condition F and from condition D. A careful reading of the description of the criteria and use of an intermediate high-dry objective should reduce this source of variability. Another source of variability relates to discriminating a late condition B from an early condition E. A common source of variability may relate to lack of detection of late condition E. Small myelin ovoids distributed at great distances from each other are readily missed by inexperienced observers. Large onion bulbs with many Schwann cell processes can also be mistaken for empty nerve strands. Frequency of Graded Pathologic Abnormalities in Control Nerves Table 32–1 provides normal values based on evaluation of teased fibers from sural nerves of paid volunteers. These persons were neurologically examined to ensure that they did not have neurologic disease.

Internode Length and Diameter Measurement of IL and ID of teased fibers is of interest because these attributes may provide an insight into altered myelin internode development and previous demyelination and remyelination. To illustrate, in diphtheritic and lead neuropathy, segmental demyelination and remyelination result in the formation of shortened internodes having more variable lengths. This can be shown graphically when IL of teased fibers is plotted against their average diameter.61 This approach, although useful for graphic display, does not allow statistical analysis. To assess IL and ID more fully, one should consider the issue of sampling, of how to estimate the average IL or ID per teased fiber or per nerve, and of how to estimate the variability of IL or ID per teased fiber or per nerve. First, for the assessment to be meaningful, it is necessary to obtain a representative sample of teased fibers from nerve. This can be done by a systematic sampling of endoneurial strands from nerve as described above (see Sampling). Nonrepresentative teasing of MFs may affect both average IL and its variability. Because internodes of large-diameter fibers are longer than those of small-diameter fibers, the mean of mean IL of teased fibers may be greater than it should be if one selectively picks large fibers, or smaller than it should be if one selectively chooses small fibers. Values indicating a short mean of mean IL may be due to selection of too many small fibers during teasing, selective loss of large fibers, or previous remyelination after segmental demyelination. Conversely, longer than normal mean of mean IL values may be due to selection of too many large fibers during teasing or loss of small fibers.

Table 32–1. Graded Pathologic Conditions of Teased Fibers (%) from Sural Nerve of Healthy Volunteers A* Normal Appearance Mean Median SD Range Mean Median SD Range Mean Median SD Range

B Myelin Irregularity

C, D, F, G Segmental De- and Remyelination

15–⬍30 yr (n ⫽ 21; 9F and 12M; mean, 22 yr; median, 22 yr; range, 20–27 yr) 95.0 0.2 3.9 94.9 0 3.1 3.1 0.5 2.7 90.2–100.0 0–2.3 0–8.6 30–⬍45 yr (n ⫽ 4; 3F and 1M; mean, 35.5 yr; median, 34.0 yr; range, 32–47 yr) 95.0 0.5 4.0 96.5 0 2.9 3.6 1.0 4.4 89.8–98.1 0–2.0 0–10.2 45–⬍60 yr (n ⫽ 4; 1F and 3M; mean, 49.3 yr; median, 49.0 yr; range, 45–54 yr) 81.5 1.8 14.5 83.0 0.6 15.9 11.0 2.8 9.3 67.3–92.6 0–5.9 2.2–23.8

F ⫽ female; M ⫽ male. *Descriptions of conditions A to H are given in the text.

E, H Axonal Degeneration and Regeneration 1.0 0 1.8 0–7.6 0 0 0 0–0 1.9 1.1 2.3 0–5.3

753

Pathologic Alterations of Nerves

The mean of mean IL may be normalized or adjusted to reflect a normal size distribution, allowing direct comparison to healthy nerve. The approach minimizes the effect of selection of small myelinated fibers (SMFs) or large myelinated fibers (LMFs) fibers by selective teasing or by disease. The approach can be illustrated for sural nerve, in which there is a bimodal size distribution and 60% of MFs generally are less than 5.5 ␮m in diameter (SMFs) and 40% are greater than or equal to 5.5 ␮m in diameter (LMFs). Let us assume, in our first example, that too many large teased fibers were selected (60% LMFs and 40% SMFs) and that the mean of mean IL of LMFs is 700 ␮m, whereas that of SMFs is 300 ␮m, to give a mean of mean IL (unadjusted) of 540 ␮m. Assuming a normal 60:40 ratio of SMFs to LMFs, the adjusted mean of mean IL is 0.6 (300) ⫹ 0.4 (700) ⫽ 460 ␮m. For our second example, let us assume that, in a nerve from a patient with Friedreich’s ataxia, 90% of MFs are small and 10% large. Let us also assume that the mean of mean IL of small fibers is 300 ␮m, whereas the 10% of large fibers have a mean of mean IL of 700. The mean of mean IL (unadjusted) is 340 ␮m, but the adjusted mean of mean IL is also 460 ␮m (normal)—evidence that remaining fibers have approximately normal ILs, considering the caliber of fiber. The variability of IL may be tested by determining the coefficient of variation (CV) of IL per individual teased fiber. The mean of mean CV of IL of 100 teased fibers (CV of IL/nerve) is then determined. Because in health the individual CV is somewhat larger for small than for large fibers, one must consider whether an abnormally high CV of IL is due to a selective loss of large fibers, selective teasing of small fibers, or previous segmental demyelination and remyelination. An abnormally low CV of IL may be due to selective loss of small fibers or a selective teasing of large fibers. Systematic sampling should preclude selective teasing by fiber size, but how can one compare the mean of mean CV of IL when there is an altered size distribution of fibers in disease? A normalized CV of IL, using the CV of IL of various size categories from disease but calculating values assuming

a normal distribution, will allow direct comparison of values from disease and control. In Table 32–2 we give values for CV of IDs and ILs of MFs obtained from sural nerves of two controls and a patient with an inherited neuropathy. In the latter disease, the mean of mean IL is much shorter and the variability of IL and ID is greater than in controls. Assessing for Clustering of Demyelination The study of the distribution of demyelination among “old internodes” of teased fibers can provide an important insight as to the mechanism of demyelination. If demyelination occurs at random among old internodes, a diffusely mediated damage to Schwann cells is suggested. A metabolic derangement of Schwann cells or an intoxication of Schwann cells mediated through the endoneurial fluid bathing them might cause such diffuse damage. This appears to be the case in lead neuropathy.53 Alternatively, if demyelination occurs repeatedly on a selected few teased fibers but not on many teased fibers, secondary demyelination should be suspected. Such secondary demyelination is found in uremic neuropathy38 and Friedreich’s ataxia,43 and is attributed to axonal atrophy. The approach to identification of “original” or old internodes is described and illustrated in Figures 32–22 and 32–23.

Electron Microscopic Assessment of Teased Fibers This is an elegant but time consuming and expensive evaluation seldom practical for the assessment of clinical biopsy material. The technique permits ultrastructural assessment at defined positions along the length of teased fibers.42,146 Uncommon focal abnormalities can then be quickly located and appropriately sectioned. Another use is to systematically examine structural features along the length of fibers.42 This approach has been used to advantage to show the variability in axonal caliber and of myelin thickness along abnormal fibers central to neuroma146 and in fibers from sural nerve in

Table 32–2. Coefficients of Variation of Diameters and Lengths of Internodes of 100 Teased Fibers from Healthy Sural Nerves and from Sural Nerve of Patient with HMSN type I Coefficients of Variation (%) Diameters Nerve

Age (yr), Sex

68–69 69–69 65–69

10, M 19, M 22, M

Lengths

Category of Nerve Donor

Mean

Range

SD

Mean

Range

SD

Accident Congenital heart disease HN-CMT

10.5 9.0 19.3

3.0–23.5 1.6–23.2 7.0–52.2

4.72 4.24 8.97

9.7 9.4 35.2

2.8–32.8 1.7–32.2 12.2–68.1

4.68 5.80 11.18

HMSN type I ⫽ hereditary motor and sensory neuropathy type I; HN-CMT ⫽ hypertrophic neuropathy of Charcot-Marie-Tooth type; SD ⫽ standard deviation. From Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in Dejerine-Sottas disease. Mayo Clin. Proc. 45:286, 1970, with permission.

754

Pathology of the Peripheral Nervous System

MORPHOMETRY OF NEURON SOMAS

FIGURE 32–22 Schematic drawings of teased myelinated fibers with segmental demyelination (upper) and remyelination (lower). The method of scoring is by the “territory of an old internode” for the determination of randomness of segmental demyelination (see text and original article for derivation). (From Dyck, P. J., O’Brien, P. C., and Ohnishi, A.: Lead neuropathy. 2. Random distribution of segmental demyelination among “old internodes” of myelinated fibers. J. Neuropathol. Exp. Neurol. 36:570, 1977, with permission.)

hereditary motor and sensory neuropathy (HMSN) type I42 (Fig. 32–24) and Friedreich’s ataxia.53

Electron Microscopic Correlation with Teased Fiber Abnormality There is only limited information on this subject. This is given in previous editions of this textbook.

Estimates of the number and size of neuron somas in nuclei, nuclear columns, or ganglia as well as of fibers in tracts or nerves have been reported for more than 100 years. Early results were usually so variable, and are so different from results now obtained using improved approaches, that they must be considered unreliable. However, some early studies of the number and size distribution of nerve fibers were well done; for example, Erlanger and Gasser’s results relating components of the compound action potential to peaks of the fiber diameter histogram are noteworthy.58 Now, better fixatives and recognition that hyperosmolar fixatives affect estimates of number, size, and shape, as well as improved embedding media, thinner sections, and improved microscopic resolution and morphometric approaches, are available that permit performing reliable morphometric analysis of fibers of nerves as summarized in this chapter. The objective of morphometric analysis is to determine the number, size, shape, and spatial distribution of various neuron cell bodies and their axons at predetermined anatomic levels. Such quantitative data are used to characterize alterations with development, with aging, with various environmental and toxic exposures, with disease, and with therapy. If the number and size distribution of cell bodies and of proximal and distal axons of peripheral neurons are estimated for a given disease and compared to controls, the altered three-dimensional structural changes associated with disease can be inferred. Reference morphometric results are now available for defined levels of certain motor neuron columns, intermediolateral columns (ILCs), spinal ganglia, sympathetic ganglia, and various levels of different nerves.

Motor Neuron Columns

FIGURE 32–23 The same schematic drawings of teased fibers as shown in Figure 32–22 but as scored by method 2 (see original article for derivation). Because evaluation for randomness using this method will result in a clustered distribution, even when the involvement of Schwann cells is random, this method cannot be used. (From Dyck, P. J., O’Brien, P. C., and Ohnishi, A.: Lead neuropathy. 2. Random distribution of segmental demyelination among “old internodes” of myelinated fibers. J. Neuropathol. Exp. Neurol. 36:570, 1977, with permission.)

Kawamura and colleagues87 systematically evaluated the number and size distribution of motor neuron column somas of the third, fourth, and fifth lumbar segments of humans. These somas are clustered into medial and lateral groups (Fig. 32–25). The medial group probably innervates truncal muscles while the lateral group innervates limb muscles. The typical morphology of one of the columns of a lateral group of motor neurons is shown in Figure 32–26. Plotting the frequency of diameters of lumbar spinal cord motor neuron somas, three peaks were found: one at a small cyton diameter (CS), one at an intermediate diameter (CI), and one at a large diameter (CL) (Fig. 32–27). The peaks and troughs were approximately the same from one spinal cord to another spinal cord. Using troughs as division points, it is possible to estimate the number of somas (or cytons15) in CL, CI, and CS peaks for L3, L4, and L5 spinal cord segments. The values for 18 control spinal cords of human patients without known disease of peripheral nerve are given in Table 32–3.

Pathologic Alterations of Nerves

755

FIGURE 32–24 Transverse sections at the same magnification (⫻79,050) from two sites on the same teased fiber from a child with hypertrophic neuropathy of the Charcot-Marie-Tooth type. In insets, both at the same low magnification (⫻5088), rectangles show area reproduced at higher magnification. These pictures illustrate that good electron microscopic preservation is possible by the method of teasing described in the text. (From Dyck, P. J., and Lais, A. C.: Electron microscopy of teased nerve fibers: method permitting examination of repeating structures of same fiber. Brain Res. 23:418, 1980, with permission.)

The usefulness of this approach is illustrated by a study of motor neuron columns in amyotrophic lateral sclerosis86 and in inherited dysautonomia.40 Figure 32–28, taken from the paper on dysautonomia, shows the selective absence of CIs in familial dysautonomia.

Spinal Ganglia The number and size distribution of somas of S1 spinal ganglia embedded in paraffin were reported for a group of control persons.119 In a later study, Kawamura and Dyck

examined L5 ganglia using celloidin embedding, greater enlargement to increase the number of counting windows, and averaging to improve peak recognition.85 Figure 32–29 shows six histograms of patients 17 to 77 years of age. Numbers of cytons in each peak are given in Table 32–4. Three overlapping peaks are present. In contrast to motor neurons, in these ganglia CS is generally the highest, followed by CI and then by CL. Using the troughs between CS and CI and between CI and CL, the numbers of small, intermediate, and large somas were calculated, and these are given in Table 32–3.

756

Pathology of the Peripheral Nervous System

FIGURE 32–25 Schematic presentation of transverse sections of L3, L4, and L5 spinal cord of human showing motor neuron columns. (From Dyck, P. J., Low, P. A., and Stevens, J. C.: Diseases of peripheral nerves. In Baker, A. B., and Baker, L. H. [eds.]: Clinical Neurology. Philadelphia, Harper Medical, p. 1, 1981, with permission.)

To illustrate the abnormality that may occur in disease, Figure 32–30 shows the severe depletion of L5 spinal ganglion somas in dysautonomia.

Intermediolateral Column The thoracic sympathetic preganglionic outflow is important in the maintenance of postural normotension in humans. Low and co-workers determined the number and size distribution of ILC somas of the T6, T7, and T8 segments of 12 spinal cords of humans.104 The mean counts for T6, T7, and T8 segments were 5002, 5004, and 4654, respectively. No sex difference was shown. Most somas ranged from 8 to 23 ␮m. The major peak was at 12 to 13 ␮m and a smaller peak was seen at 16 ␮m. There was a progressive reduction of ILC somas with age, amounting to 8% per decade. This reduction in ILC cytons with age may be a morphologic basis for postural hypotension of the aged.

Sympathetic Trunk Ganglion Dyck and colleagues have compared the number and size of T7 sympathetic ganglia neurons for a control individual and for a patient with dysautonomia (Fig. 32–31).40

MORPHOMETRY OF PERIPHERAL NERVES The approaches described here can be used to determine the number, density, diameter distribution, shape, and spatial distribution of fibers of nerves or of spinal cord tracts. In addition, the approaches used permit assessment of the

FIGURE 32–26 Representative microscopic field of one of the columns of the lateral group of motor neurons of the L5 segment of the human spinal cord showing the favorable histologic preparation for morphometric analysis, which can be obtained using aldehyde fixation, celloidin embedding, and cresyl violet staining, as described in the text. The spinal cord was fixed in buffered paraformaldehyde and glutaraldehyde solution, embedded into celloidin, sectioned at 25 ␮m, and stained with cresyl violet. (From Kawamura, Y., O’Brien, P. C., Okazaki, H., and Dyck, P. J.: Lumbar motoneurons of man. II. The number and diameter distribution of large- and intermediate-diameter cytons in “motoneuron columns” of spinal cord of man. J. Neuropathol. Exp. Neurol. 36:861, 1977, with permission.)

diameter distribution of MFs, axon areas, myelin areas, axon perimeters, and myelin perimeters. Each measured attribute can be separately assessed for each fiber or per nerve and related to one another (e.g., axon area regressed on myelin thickness). Should it become possible to reliably identify functional classes of axons in transverse sections using markers, it will be possible to determine their size class in health and disease. The density, size distribution of MFs, axon-to-myelin thickness relationships, and index of circularity (IC) appear to change during development, regeneration, aging, and degeneration. In addition, these approaches may be used to characterize neuropathologic abnormalities.

757

Pathologic Alterations of Nerves

a fiber as obtained from serial section will be larger and smaller than the average diameter. This variance in healthy nerve was close enough to allow Erlanger and Gasser to show a relationship between conduction velocities and diameter.58 Lambert and Dyck found that in demyelinating neuropathies, however, histograms from a sampled section are not a true reflection of the average diameter of fibers. Such histograms show fiber profiles that are much too large and much too small—the histogram from a single section does not accurately reflect the average diameter distribution of fibers (see Chapter 39).

Histologic Processing

FIGURE 32–27 Diameter frequency distributions of cytons of motor neuron columns of L3 (top), L4 (middle), and L5 (bottom) segments of a representative spinal cord from a 17-year-old male. Three distinct diameter peaks are seen: large diameter (CL), intermediate diameter (CI), and small diameter (CS). It is reasonable to assume that the majority of CL cytons are alpha motor neurons, that the majority of CI cytons are gamma motor neurons, and that the CS cytons are interneurons. (From Kawamura, Y., O’Brien, P. C., Okazaki, H., and Dyck, P. J.: Lumbar motoneurons of man. II. The number and diameter distribution of large- and intermediate-diameter cytons in “motoneuron columns” of spinal cord of man. J. Neuropathol. Exp. Neurol. 36:861, 1977, with permission.)

Because fibers of healthy nerves are cylinders of approximately similar caliber for long distances, it is possible to infer that the range and diameter peaks of one transverse section are representative for other levels of nerve. Extensive work relating diameter peaks to the various peaks of the compound action potential in vitro were based on this premise. Estimates of numbers of MFs will be slightly too low because sections through nodes of Ranvier will not be recognized as MFs. With disease, variability of axonal caliber, segmental demyelination and remyelination, and axonal degeneration may develop, so that the number and diameter distribution from one section may not be an accurate reflection of the number or average size of fibers. To illustrate, the largest and smallest diameter of

Immersion fixation of nerves in hyperosmolar glutaraldehyde results in reproducible fascicular and fiber shape alterations from normal (see Fig. 32–6). Whereas nerve fascicles and MFs assume roughly spherical shapes in rapidly frozen and freeze-dried transverse sections, they assume flattened, boomerang, or crenated shapes, to a variable degree, with hyperosmolar fixation. The IC (⫽ ␲D/P, where P is perimeter and D is diameter, and which is 1 for a circle) is significantly lower for nerves fixed in hyperosmolar fixative than it is for those fixed in isosmolar fixative. The perineurium and the axolemma act as semipermeable membranes. These distortions can be avoided, at least to a large degree, when nerves are fixed in isosmolar aldehyde fixative (e.g., 2.5% glutaraldehyde in 0.025 mol/L cacodylate buffer). With perfusion or in situ fixation, hyperosmolar fixatives do not have such a large adverse effect. Low temperatures (0° to 4° C) are associated with severe depolymerization of MTs, and it is for this reason that we fix tissues at 10° C. We employ a specially designed cart having an insulated (during transport from the hospital to the laboratory) metal block housing the fixative bottles and a thermoelectric unit (with controller and power supply) that maintains temperature at 10° C when connected to line current.

Section Thickness and Axon-Myelin Relationship Thick, semithin, and thin sections can be used to provide accurate estimates of numbers of MFs, but only the last two may be used to evaluate size, shape, and myelin-axon relationship. Use of thick sections or cutting the fibers slightly obliquely will result in too large a profile area, spuriously thick myelin, and thin axons. Although thin (~80-nm) sections, as used for electron microscopy, permit the most accurate measurement of fiber diameter, shape, and myelin-axon relationship, semithin sections are commonly used because systematic sampling is better on these larger sections, because the process is less time consuming, and because quite accurate results can be obtained. Although semithin sections 1.5 ␮m thick

758

Pathology of the Peripheral Nervous System

Table 32–3. Number of Large (CL), Intermediate (CI), and Small (CS) Cytons in 18 Spinal Cords of Humans L3

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Mean SD Range

L4

L5

CL

CI

CS

CL

CI

CS

CL

CI

CS

5720 6720 5731 4795 4797 5925 4481 4736 5133 5637 5035 5726 4595 5526 4910 4193 4017 5652

2595 3138 2538 1820 2220 1797 1334 2112 1250 3348 2024 1764 1195 1482 1429 1348 1600 2883

1316 924 932 1733 1075 1533 635 584 1497 479 1451 593 1247 592 710 725 1310 913

4753 5647 4138 4920 5311 4541 5275 5096 4219 4496 4453 4768 4361 3797 3470 4753 4672 3720

1489 2017 1984 2601 2536 1131 1953 2152 1450 1720 1230 1150 1401 1424 1213 1912 2131 1355

636 909 940 1154 1228 959 370 576 446 223 533 690 968 520 923 729 694 937

5108 4575 5353 5979 5454 5799 5691 4062 5101 4721 5727 4556 4643 4836 4739 4629 4372 3887

1480 1562 2315 1818 1562 1706 1188 1115 1445 1143 2077 1890 1365 2371 1485 1662 1770 1153

1306 1282 1422 857 490 638 318 502 500 1636 544 989 1360 1014 1452 932 1619 1027

5185 691.0 4017–6720

1993 666.8 1195–3348

1014 392.5 479–1733

4577 575.2 3470–5647

1714 464.0 1131–2601

746 273.8 223–1228

4957 606.4 3387–5979

1617 377.8 1115–2315

994 424.3 318–1619

From Kawamura, Y., O’Brien, P. C., Okazaki, H., and Dyck, P. J.: Lumbar motoneurons of man. II. The number and diameter distribution of large- and intermediate-diameter cytons in “motoneuron colons” of spinal cord of man. J. Neuropathol. Exp. Neurol. 36:861, 1977, with permission.

are usually employed, with adequate osmication and use of paraphenylenediamine, it is possible to reduce thickness to 0.5 to 1.0 ␮m, thus improving error caused by section thickness.

Attribute of Shape of Sampled Sections of Internodes That Best Estimates Average Diameter Karnes and co-workers84 have addressed this issue. Using computer imaging, measurements of MF profiles were made at intervals over half-internodes of 20 fibers from each of four sural nerves of healthy rats. In these serial skip transverse sections (every 10th section) of the 80 MFs followed from nodes of Ranvier to Schwann cell nucleus or vice versa, the following measurements were made: 1. Long axis (DL): longest distance across fiber 2. Short axis (DS): greatest distance across the fiber at right angles to the long axis 3. Area (A): calculated by a modified Simpson’s rule for irregular contours 4. Perimeter (P): summation of distances between adjacent border points

5. Diameter, assuming calculated area to be that of a circle: Dc ⫽ 2√(A/␲) 6. Diameter computed as mean of long and short axes (arithmetic mean): DM ⫽ (DL ⫹ DS)/2 7. Diameter computed as square root of product of long and short axes of ellipse (geometric mean): DE ⫽ √(DL ⫻ DS) 8. Diameter, assuming calculated perimeter to be that of a circle An idealized cylinder was reconstructed for each fiber over the half-internode measured using the cross-sectional areas (A), the thickness of sections (T), and the number of sections between pictures (S). The diameter of this idealized cylinder (DV) could then be calculated from its volume (V), which was computed using the following equation: N

V ⫽ T ⫻ ⌺ 1/2 (An ⫹ An⫺1)(Sn ⫺ Sn⫺1) n⫽2

The diameters, computed in six different ways as described, were compared for bias, precision, and accuracy between sections and were compared with the diameter of an idealized cylinder reconstructed for each fiber from multiple actual cross sections. In this study, DV for a single internode of a given fiber was evaluated from multiple cross sections.

759

Pathologic Alterations of Nerves

To evaluate how well an estimation method performed for a given internode, we computed the mean, standard deviation (SD), and square root of the mean square error of the estimates. These statistics measure bias, precision (consistency), and accuracy, respectively. On the basis of all three criteria, the diameter of a circle of equivalent area (DC) is clearly the method of choice. It was not surprising that this method shows least bias because DV and DC are derived from the area of the cross sections as determined by different methods, but this fact would not necessarily favor DC for consistency or accuracy between sections. DM and DE, which are similar to methods described in the literature, compared reasonably well to DC. Karnes and colleagues also found that a crenated shape was characteristic of the paranodal region, a boomerang shape of the nuclear region, and a circular shape of the area between paranodal and nuclear regions (Fig. 32–32).84

Myelinated Fiber Composition: Photographic Method FIGURE 32–28 Diameter histograms of cell bodies of L5 motor neuron columns in representative control (top) and dysautonomic patient (bottom). In the control, three peaks are found, large (CL), intermediate (CI), and small (CS). In the patient, only two peaks are found, large and small. Because CI and intermediate axons (AI) of ventral root are thought to be from gamma motor neurons, this is evidence of a preferential absence of gamma motor neurons in dysautonomia. (From Dyck, P. J., Kawamura, Y., Low, P. A., et al.: The number and sizes of reconstructed peripheral autonomic, sensory, and motor neurons in a case of dysautonomia. J. Neuropathol. Exp. Neurol. 37:741, 1978, with permission.)

FIGURE 32–29 Each of the diameter histograms of the cell bodies of neurons in L5 spinal ganglion shows three peaks at approximately the same diameter location, an indication that the methodology produces reproducible results and that there are at least three populations by size in this ganglion. (From Kawamura, Y., and Dyck, P. J.: Evidence for three populations by size in L5 spinal ganglion in man. J. Neuropathol. Exp. Neurol. 37:269, 1978, with permission.)

Earlier editions of this textbook outlined photographic morphometric methods that might be used. For some investigators the use of microprocessor digitizers, now widely available, will replace the mechanical counters of an earlier era.

Myelinated Fiber Composition: Computer Imaging Method The advantages of computer imaging for determining the number, size, shape, and distribution of myelin fibers and their components are as follows: it is speedy, automatic

760

Pathology of the Peripheral Nervous System

Table 32–4. Number of CL, CI, and CS Cytons in L5 Spinal Ganglion of Six Men 1 2 3 4 5 6 MV SD %

Age/Sex

CS

CI

CL

Total

17/M 22/M 40/M 56/M 68/M 77/M

29,654 22,642 26,039 28,160 27,431 18,410

35,537 30,229 29,366 30,950 29,954 33,943

13,897 12,155 17,713 10,145 7345 7316

78,088 65,076 73,119 69,255 64,730 59,669

25,389 4164 37

31,675 2481 46

11,429 4033 17

68,323 6597

From Kawamura, Y., and Dyck, P. J.: Evidence for three populations by size in L5 spinal ganglion in man. J. Neuropathol. Exp. Neurol. 37:269, 1978, with permission.

bordering occurs in approximately 85% of MFs, machine criteria not subject to human judgment are used in bordering, exact perimeters and areas of fibers are provided, the need for photographic processing is eliminated, and maximum flexibility in altering programs and handling

FIGURE 32–31 Diameter histogram of T7 sympathetic trunk neurons from control (top) and from patient with familial dysautonomia (bottom). (From Dyck, P. J., Kawamura, Y., Low, P. A., et al.: The number and sizes of reconstructed peripheral autonomic, sensory, and motor neurons in a case of dysautonomia. J. Neuropathol. Exp. Neurol. 37:741, 1978, with permission.)

data is possible. The system is operator interactive so that each enumerated and bordered fiber visualized on the video screen can be verified under the microscope. Programs have been developed that identify MFs and border the inner edge of myelin (Fig. 32–33). This is averaged (eight times) to increase accuracy. For each identified MF the following measurements are made: length of perimeter (from bordering the inner edge of myelin), area enclosed by perimeter, and myelin thickness. From these measurements a variety of derived attributes are calculated. Because the data for each fiber are stored individually, they can be flexibly assessed later by multiple approaches. Subroutines have been prepared for drawing histograms and relating various variables by regression analysis. FIGURE 32–30 Diameter histogram of L5 spinal ganglion neuron cell bodies in control (top) and dysautonomic patient (bottom). Note that the scale of the ordinate for the patient is approximately one tenth that of the control. Neurons of all sizes are severely reduced in number, but small ones are preferentially affected. (From Dyck, P. J., Kawamura, Y., Low, P. A., et al.: The number and sizes of reconstructed peripheral autonomic, sensory, and motor neurons in a case of dysautonomia. J. Neuropathol. Exp. Neurol. 37:741, 1978, with permission.)

Hardware Components The hardware components of the imaging system for nerve morphometry (ISNM) that we have developed were described in earlier editions of this textbook. System Approaches The approaches and feasibility studies that led to the histologic preparation used and to the design and software features employed in ISNM were given in detail in earlier editions of this textbook.

Pathologic Alterations of Nerves

761

FIGURE 32–32 A drawing of a hemi-internode of a myelinated fiber showing that characteristic shapes are seen near the internode (I1), adjacent to the nucleus (I3), and between I1 and I3. (From Dyck, P. J., and Karnes, J. L.: Morphometry of neuron columns and fiber tracts in neurobiology and pathology using computer imaging. Trends Neurosci. 4:138, 1981, with permission.)

FIGURE 32–33 Using an operator-interactive computer imaging system for nerve morphometry, the inner boundaries of myelin are bordered as described in the text. Axon area and diameter, myelin area and thickness, and number and size distribution of myelinated fibers are automatically determined as described in the text. (From Zimmerman, I. R., Karnes, J. L., O’Brien, P. C., and Dyck, P. J.: Imaging systems for nerve and fiber tract morphometry: components, approaches, performance and results. J. Neuropathol. Exp. Neurol. 39:409, 1980, with permission.)

As in the photographic method, the transverse area of fascicles is determined on low-power enlargements using digitization. Then at high-dry magnification (~2000⫻), the endoneurial area is conceptually overlaid by frames corresponding to a square frame in the eyepiece. Each area of the endoneurium, frame by frame, and parts of frames, is brought under the ocular frame by a x and y traverse of the endoneurial area. The first frame for analysis is then chosen from a random number table, and thereafter frames are assessed at regular intervals (e.g., 1 in 4 or 1 in 6). The ratio of assessed to all frames chosen is predetermined so that 400 to 1000 MFs per nerve will be assessed. This number assures that the histogram will be smooth. For each frame to be assessed, hemocytometric technique is employed. Because the analysis is interactive and because every fiber included is bordered and otherwise marked (by a white square in the axon) on the video screen, every fiber within a frame can be detected and bordered with certainty, without spurious inclusion of other histologic profiles, since the operator can accept or reject every profile by histologic criteria on the video screen or, if he or she wishes, by direct examination of the same frame of the histologic section in the microscope. After recognition and bordering of the inner edge of myelin, computer programs estimate the perimeter and areas of the bordered profiles by rules described in detail in earlier editions of this textbook. MF size, shape, and IC are obtained by the following steps. Measured and Derived MF Measurements from Transverse Sections. With the ISNM, the areas and perimeter of all fascicles are measured using a low-power objective. Using a high-power objective (63⫻), the endoneurial area, the number of MFs per frame, and the axon area, axon perimeter, and myelin thickness for each MF

762

Pathology of the Peripheral Nervous System

even in health. Therefore, each level must be compared to its own control. This approach may be used to recognize dying-back alterations or a focal pathologic abnormality. Density, average size, or range of diameters may be compared among frames within fascicles, and among fascicles to recognize multifocal pathology. P. C. O’Brien (personal communication, 1980) has suggested two approaches to assess variability. In the first approach, the variability in the total series of frames consists of two components: variability among frames within fascicles and variability among fascicles. That is, ␴ 2 ⫽ ␴A2 ⫹ ␴␻2 . In these terms, a suitable index is ␴A / √(␴A2 ⫹ ␴␻2 ). The second approach is the same as that used in evaluating clustering based on a series of frames within fascicles. The basic measurement becomes Xi, or the number of fibers observed in the ith fascicle, where i ⫽ 1, . . . , n. However, since the area (Ai) sampled will vary, it will be necessary to take this into account. (The same would have been true in our previous index, if we had taken partial frames into account.) For notation, if we define

within the rectangular frame are measured. Derived attributes that are calculated include the area of myelin, the area of combined axon and myelin, and the diameter of a circle of equivalent area (diameter of MF). An IC (⫽ ␲D/P) is calculated as a measure of noncircularity. From the measured and derived attributes for each MF, which are stored, various analyses are possible. First, the density of MFs (number/mm2 or per nerve) is determined. Second, the mean, median, range, and SD of the nerves, axons, myelin, and MFs as areas or as diameters (of a circle of equivalent area) are listed. Similar calculations can be made for IC. Third, any of these measured or derived attributes can be displayed as a frequency distribution, preferably by square millimeter of fascicular area or by nerve. For MFs we employ an exponential plot with bars of equal relative width to use the same precision in locating peaks for fibers of small or large diameter. If well-developed peaks and troughs are visualized, the number of MFs per peak can be determined. Fourth, one variable can be regressed on another. Derived values in a typical nerve recently analyzed are as shown in Table 32–5. Assessing Variability in MF Density. The density and size distribution of MFs may be used to recognize focal or multifocal abnormality along the length of nerves and between or within fascicles at one level. To do so, the location of the sampled frames and the measured attributes of MFs by frame must be known. In evaluating for alterations along the length of nerves, it is important to recognize that a direct comparison between sampled levels generally is not possible because they may be different

Di ⫽ Xi /Ai n

D ⫽ i⫽1 ⌺ Di /n S2D ⫽ ⌺(Di ⫺ D)2/(n ⫺ 1) AH ⫽ n(⌺Ai⫺1) ⫺1 (harmonic mean range)

Table 32–5. Peroneal Nerve of Cat: Number and Size Distribution of Myelinated Fibers Nerve 140–79L LT1 Whole nerve Number of frames Number of measured fibers Number of unmeasured fibers Fascicular area (mm2) % of fascicle sampled Fascicular IC

Mean SD Median Number/mm2 SD Confidence interval Upper limit: 8.72114E Lower limit: 7.80565E

50 891 61 1.05 11.00 0.96 Diam (␮m)

IC

MT (␮m)

MA (␮m2)

Per (␮m)

8.50 3.94 8.73

0.88 0.03 0.89

1.23 0.44 1.34

32.83 24.18 31.04

29.88 13.24 30.41

8262 273 8721 7806

IC ⫽ index of circularity as defined in text; MA ⫽ myelin area; MT ⫽ myelin thickness; Per ⫽ axon perimeter.

Pathologic Alterations of Nerves

then the proposed index is ID ⫽

S2D AH D

Notice that, if A1 ⫽ . . . ⫽ An, this reduces to the wellknown index of dispersion for testing a common Poisson distribution. The derivation is as follows: E ⌺(Di ⫺ D)2 ⫽ ⌺ E(D2i ) ⫺ n E(D)2 ⫽ ␭ ⌺ Ai⫺1 ⫹ n␭2 ⫺ n ⫺1 E(⌺ Di)2 E ⌺(Di)2 ⫽ ⌺ E(D2i ) ⫹ ⌺ DiDj ij

⫽ ␭ ⌺ Ai⫺1 ⫹ n2␭2 Therefore, E ⌺(Di ⫺ D)2 ⫽ ␭ ⌺ Ai⫺1 ⫹ n␭2 ⫺ ⫽

1 ␭ ⌺ Ai⫺1 ⫺ n␭2 n

n⫺1 ␭ ⌺ Ai⫺1 n

⫽ ␭(n ⫺ 1)/AH Because E(D) ⫽ ␭ E[S2D AH] E(D)

⫽1

if the fibers are distributed according to the same random distribution on all n fascicles.

NORMAL NUMBERS OF FIBERS IN CONTROL HUMAN NERVES This section provides an abbreviated account of MF density (number/mm2 of fascicular area), number/nerve, diameter distribution (plotted either per nerve or per square millimeter of fascicular area), and ranges and peaks of human nerves taken at autopsy or at biopsy (sural nerve). Patients or subjects who provided nerves were without neurologic disease or disease predisposing to peripheral neuropathy. The method of fixation and embedding was as described for postmortem and biopsy tissue in the preceding sections. Epoxy transverse sections were cut at 0.75 ␮m and stained with paraphenylenediamine and evaluated with ISNM as described above. The diameter histograms plotted per square millimeter of endoneurial area for various

763

human nerves are shown in the illustrations in this section. In each figure, short single vertical lines pointing down from the baseline indicate the position of the first percentile and the 99th percentile positions. Double lines indicate the median diameter. The position of frequency distribution peaks is sometimes also shown as long vertical lines pointing down from the baseline. To the left of the histogram we show the age and sex and the number of MFs per nerve and per square millimeter; to the right we show the MF percent of fascicular area. For subsequent histograms of the same nerve in each figure, only the baselines of the histogram are shown to be able to include many histograms in one figure. The bottom line provides the values for the composite histograms. By assessing these normative morphometric data, some inferences can be formulated. The first is that, for a given level of a specific nerve, there is considerable variability both in density and in number of MFs/nerve, but the range, diameter histogram, median diameter, and diameter position of peaks is remarkably constant. The second is that, for an unbranching nerve, the number of MFs/nerve appears to remain constant, whereas the number/mm2 decreases from proximal to distal. The diameter histogram shows no evidence for fiber attenuation over long proximal-to-distal distances—the peaks being essentially at the same diameter position in proximal as in distal nerve. This evidence suggests that there are mechanisms for maintaining axon caliber over long distances. The third inference is that there are differences in fiber composition among roots and nerves reflecting differences in composition by fiber class. The fourth is that there are striking differences in diameter histograms among species. The fifth is that morphometric values are sufficiently consistent that, when they are evaluated in patients with neuropathy, important insights may emerge regarding the nerves affected, the class of neurons (axons) affected, the level within the neuron affected, and the components (axon or myelin) affected. A comparison of the MF diameter histograms of T7 and L5 ventral spinal roots (VSRs) (Figs. 32–34 and 32–35) illustrates that the size distribution pattern may be characteristic for segmental levels and different among levels. This variability is described here in more detail. Designating small axons as AS, intermediate axons as AI, and large axons as AL, the T7 VSR has a high peak for AS, a low peak for AI, and an intermediate-height peak for AL. These peaks correspond to preganglionic splanchnic fibers, gamma fibers, and alpha fibers, respectively. The height and diameter position of the peaks in L5 VSRs are strikingly different, there being only a few AS axons and a ratio of AI to AL axons similar to that found in T7. There is another difference between T7 and L5. The AI and AL peaks of T7 are at smaller diameter positions than they are for L5. These differences reflect the greater number of preganglionic axons in T7 and the smaller caliber of gamma and alpha axons for thoracic than for lumbar roots, perhaps reflecting the shorter length of thoracic neurons.

764

Pathology of the Peripheral Nervous System

FIGURE 32–34 Histograms and morphometric data on T7 human ventral spinal roots (VSR).

FIGURE 32–35 Histograms and morphometric data on L5 human ventral spinal roots (VSR). MF ⫽ myelinated fiber. (From Dyck, P. J., Karnes, J., Sparks, M., and Low, P. A.: The morphometric composition of myelinated fibers by nerve, level, and species related to nerve microenvironment and ischaemia. Electroencephalogr. Clin. Neurophysiol. Suppl. 36:39, 1982, with permission.)

Pathologic Alterations of Nerves

There is also a difference in the diameter position of MF diameter peaks between species. The VSR axons of the cat are considerably larger than those of humans (Fig. 32–36), the AL peak of the former being at approximately 14 to 15 ␮m and the latter at approximately 13 to 14 ␮m. Other evidence for diameter distribution differences among species may be found in the work of Boyd and Davey16 and in our previous communications.39 The striking difference in fiber composition between ventral and dorsal spinal roots of humans is illustrated in Figures 32–35 and 32–37. Diameter histograms of human L5 segmental nerve, midthigh sciatic nerve, tibial nerve, peroneal nerve, and proximal and distal sural nerve are presented in Figures 32–38, 32–39, 32–40, 32–41, 32–42, and 32–43, respectively. The second inference (the lack of tapering of nerve fibers) is illustrated by a comparison of histograms of the sural nerve at midcalf and at the ankle (see Figs. 32–42 and 32–43). The positions of diameter peaks (AS and AL) are not significantly different for the two, but the density is somewhat less at ankle level—the fascicular area tends to be large distally. The fifth inference is illustrated by providing two examples, the first related to the selective vulnerability of alpha motor neurons in motor neuron disease (Fig. 32–44) and the second the selective vulnerability (considering

FIGURE 32–36 Histograms and morphometric data on L7 ventral spinal roots (VSR) of cat. MF ⫽ myelinated fiber.

765

motor neurons) of gamma motor neurons in inherited dysautonomia (Fig. 32–45).

ARTIFACTS OF SURGICAL REMOVAL AND OF HISTOLOGIC PREPARATION Before nerve tissue can be visualized under the microscope, it must be surgically removed, fixed, dehydrated, embedded, cured, sectioned, and stained. Despite this severe manipulation, structures seen under the microscope bear a strong similarity to what is seen in life, in freshly teased tissue, or in cryostat sections. The fact that compound action potentials in vitro can be reconstructed from diameter histograms of fibers provides further evidence that the structure of fibers has been approximately retained during histologic processing. Likewise, subcellular structures such as nodes of Ranvier, Schmidt-Lanterman incisures, and organelles observed in living preparations are retained in embedded tissue. However, artifactual alterations caused by removal or processing of nerve also occur and must be recognized so that they are not interpreted as evidence of pathologic alteration. At nerve biopsy the major artifacts that may be induced are hemorrhage, excessive stretch, compression, osmotic

766

Pathology of the Peripheral Nervous System

FIGURE 32–37 Histograms and morphometric data on L5 human dorsal spinal roots (DSR). MF ⫽ myelinated fiber.

FIGURE 32–38 Histograms and morphometric data on L5 segmental human nerve. MF ⫽ myelinated fiber. (From Dyck, P. J., Karnes, J., Sparks, M., and Low, P. A.: The morphometric composition of myelinated fibers by nerve, level, and species related to nerve microenvironment and ischaemia. Electroencephalogr. Clin. Neurophysiol. Suppl. 36:39, 1982, with permission.)

Pathologic Alterations of Nerves

767

FIGURE 32–39 Histograms and morphometric data on human sciatic nerve. MF ⫽ myelinated fiber.

shrinkage, and introduction of foreign material. Some degree of hemorrhage is probably inevitable because sharp dissection is used. Typically it is found at the edges of the biopsy specimen. Fresh hemorrhage may also occur in disease and is recognized as such by its localization to

vessel alterations or to inflammatory reaction. Excessive traction, compression, and manipulation induce major tissue alteration; avoidance of these artifacts is always better than trying to distinguish them from pathologic alterations. Local anesthetic should be injected several

FIGURE 32–40 Histograms and morphometric data on human tibial nerve. MF ⫽ myelinated fiber.

768

Pathology of the Peripheral Nervous System

FIGURE 32–41 Histograms and morphometric data on human peroneal nerve. MF ⫽ myelinated fiber. (From Dyck, P. J., Karnes, J., Sparks, M., and Low, P. A.: The morphometric composition of myelinated fibers by nerve, level, and species related to nerve microenvironment and ischaemia. Electroencephalogr. Clin. Neurophysiol. Suppl. 36:39, 1982, with permission.)

FIGURE 32–42 Histograms and morphometric data on human proximal sural nerve. MF ⫽ myelinated fiber. (From Dyck, P. J., Karnes, J., Sparks, M., and Low, P. A.: The morphometric composition of myelinated fibers by nerve, level, and species related to nerve microenvironment and ischaemia. Electroencephalogr. Clin. Neurophysiol. Suppl. 36:39, 1982, with permission.)

Pathologic Alterations of Nerves

769

FIGURE 32–43 Histograms and morphometric data on human distal sural nerve. MF ⫽ myelinated fiber. (From Dyck, P. J., Karnes, J., Sparks, M., and Low, P. A.: The morphometric composition of myelinated fibers by nerve, level, and species related to nerve microenvironment and ischaemia. Electroencephalogr. Clin. Neurophysiol. Suppl. 36:39, 1982, with permission.)

centimeters above the upper end of the nerve to be transected so the injected fluid is not misinterpreted as edema. The suture should be passed through the nerve and the free ends used as loop retractors to elevate the nerve from its bed. The operator must avoid tying and thus crushing the nerve. The degree of tension applied on the suture used as a retractor should be just sufficient to straighten the nerve or fascicles while being undercut. Excessive stretch induces tears, bleeding, separation at nodes, and a beaded appearance of teased fibers. Immediately after removal of the nerve specimen, it should be put into the fixative solution. The issue of when artifactual crush of nerve fibers seen in histologic sections occurs is still not finally settled. Perhaps it can occur at several times (i.e., at surgery and during histologic fixation). A severe example of crush is shown in Figure 32–46. Crush can be distinguished from axonal degeneration by the following criteria: (1) large fibers are more affected than small fibers; (2) edge fibers are more affected than central fascicular

fibers; (3) in the former, myelin is splayed apart and sometimes disrupted, whereas in the latter and especially at later stages of degeneration, myelin is in discrete ovoids or solid balls; and (4) in the latter, there is an increase in Schwann cell cytoplasm and there may be autophagic vacuoles (some of which may contain sudanophilic lipids). This artifact is perhaps more readily recognized in teased fibers by the observation that the area of crush occurs at the same proximal-to-distal level in multiple examined fibers (Fig. 32–47). Crush artifact may also be misinterpreted as segmental demyelination in teased fibers. This possibility should be considered when (1) the demyelination does not affect the paranode and (2) myelin at the edge of the apparent demyelination appears thickened, distorted, frayed, and irregular; there are not associated smooth myelin ovoids and balls in the adjacent cytoplasm; and similar alterations occur at approximately the same region of other teased fibers (see Fig. 32–47). Other artifacts related to teasing were shown in Figure 32–12. In serial sections,

770

Pathology of the Peripheral Nervous System

FIGURE 32–44 Representative histograms of L5 motor neuron cell bodies (left) and ventral root myelinated axons (right) from control subject (top), Case 1 with amyotrophic lateral sclerosis (ALS) (middle), and Case 3 with ALS (bottom). The histogram of Case 1 is similar to that of most cases of ALS in that the CL and AL peaks are missing. The histogram of Case 3 is different in that the CL and AL peaks are higher and broader than normal. (From Kawamura, Y., Dyck, P. J., Shimono, M., et al.: Morphometric comparison of the vulnerability of peripheral motor and sensory neurons in amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 40:667, 1981, with permission.)

the artifact may be shown to be restricted to a given proximal-to-distal level of nerve. Major artifacts come from freezing and from autolysis. In Figure 32–48, ice crystal formation has severely distorted the histologic structures. Autolysis, which may be due to prolonged periods before immersion into fixative or use of too weak a fixative solution, results in watery axons, distorted and ballooned mitochondria, dispersed chromatin, and other artifacts.

HISTOLOGIC STRUCTURES WITH AND WITHOUT PATHOLOGIC SIGNIFICANCE Renaut Corpuscles Histologic structures referred to as Renaut corpuscles were first described in the nerves of human, horse, dog, and ass by Renaut (Fig. 32–49).128,129 Subsequently they have been seen in nerves of persons with various disorders. They have been extensively studied and reported in the German and French literature.35,97 In the English

literature they were sometimes misinterpreted as infarcts.88 The corpuscles usually are found on the inside of the perineurium and protrude into the endoneurial space. Fibers course around them without being distorted by them. In longitudinal profile they taper at both ends and merge with the perineurium. In transverse sections they consist of concentric lamellae of filamentous strands. Oval, plump, pale nuclei—the cellules gondronnées of Renaut—are associated with filamentous strands. At the center of the corpuscle there usually are three or four plump nuclei similar to those noted in mucoid degeneration. The lamellae stain light pink with hematoxylin and eosin, pale green with trichrome, not at all with methyl violet, and strongly blue with Alcian blue at pH 2.5. The staining reaction suggests that they contain acid mucopolysaccharides. They are seen in the nerves of healthy persons, in greater frequency with increasing age, and, according to Asbury and co-workers,4 especially adjacent to limb joints. The consensus is that Renaut corpuscles are not specific for a particular disease and certainly are not infarcts of nerve.120 Occasionally, pacinian corpuscles are located in the epineurium (Fig. 32–50).

Pathologic Alterations of Nerves

FIGURE 32–45 Diameter histogram of myelinated axons of L5 ventral root of control (top) and of dysautonomic patient (bottom). In the control, two well-defined peaks are seen, large (AL) and intermediate (AI). In dysautonomia AI is well developed but AL is markedly reduced in size from normal values. (From Dyck, P. J., Kawamura, Y., Low, P. A., et al.: The number and sizes of reconstructed peripheral autonomic, sensory, and motor neurons in a case of dysautonomia. J. Neuropathol. Exp. Neurol. 37:741, 1978, with permission.)

Schwann Cell Inclusions Schwann cell inclusions occur primarily at polar regions of the nuclei of Schwann cells and in paranodal cytoplasm of Schwann cells. Found at these and other locations are mitochondria, glycogen granules, lipid droplets, lysosomal organelles, ␲ (pi) granules, ␮ (mu) granules of Reich (also called Elzholz bodies), tuffstone bodies, and less well characterized inclusions, such as needle-like inclusions and multilamellar bodies (e.g., zebra bodies) (Fig. 32–51). In paraffin or thick epoxy sections, the ␲ granules stain with thionine, toluidine blue, or methylene blue and appear as metachromatic granules. Usually they are rodlike or curvilinear in shape and approximately 1 ␮m long (Fig. 32–52). They are directed radially or circumferentially to the direction of the fiber. In electron micrographs, they undoubtedly correspond to Schwann cell cytoplasmic lamellar bodies. It may be that zebra bodies in lysosomal storage disease are closely related but similar organelles. They have a periodicity of approximately 4.8 nm, may be surrounded by a double membrane, and may occur singly or in clusters. Because they are more frequent and larger in

771

FIGURE 32–46 A nerve fascicle that has been severely crushed during surgical removal or before adequate fixation. The distinctive features are that edge fibers on one side of the fascicle and large fibers are selectively affected, there is no consequent Schwann cytoplasmic increase or reactivity, and the changes are restricted to the crushed region (the last not shown). (From Dyck, P. J., Dyck, P. J. B., Giannini, C., et al.: Peripheral nerves. In Graham, D. I., and Lantos, P. L. [eds.]: Greenfield’s Neuropathology, 7th ed., Vol. 2. London, Arnold Publishing, p. 551, 2002, with permission.)

lysosomal storage diseases, they are assumed to represent degradative products generated by lysosomal enzymes in healthy nerves. They have been described by Thomas and Slatford152 in nerve of the rabbit, by Evans and co-workers59 in nerves of patients with various neuromuscular disorders, by Gonatas and colleagues65 in nerves of a patient with a disorder resembling Refsum’s disease, and by Dyck and Lambert47 in a series of nerves from healthy persons and nerves from patients with hypothyroid neuropathy. Another Schwann cell degradative inclusion is the ␮ granule of Elzholz (Fig. 32–53). Because it stains with the Marchi method, it is assumed to be a degradative product of myelin. Rarely a lipocyte is found in the endoneurial area (Fig. 32–54).

772

Pathology of the Peripheral Nervous System

FIGURE 32–47 Teased fibers laid side by side from a nerve grasped during surgery. Notice that the point of crush is recognized in all sampled teased fibers. Although this is a severe example, finding fibers without myelin showing at approximately the same proximal-to-distal level of different teased fibers provided the clue that the nerve has been crushed. (Modified from Dyck, P. J., Dyck, P. J. B., Giannini, C., et al.: Peripheral nerves. In Graham, D. I., and Lantos, P. L. [eds.]: Greenfield’s Neuropathology, 7th ed., Vol. 2. London, Arnold Publishing, p. 551, 2002.)

FIBER ALTERATIONS Neuronal Degeneration Historically, anterior horn cell degenerations as occur in progressive spinal muscular atrophy or amyotrophic lateral sclerosis were separated from peripheral neuropathies by clinical, pathologic, and later electrophysiologic features. Simply stated, in the first group of disorders there was selective involvement of motor neurons and the changes were not length dependent. In peripheral neuropathy, other classes of neurons are involved and the process is

often length dependent. However, the distinction between motor neuron degeneration and peripheral neuropathy is not always apparent when it is appreciated that, in motor neuron degeneration, axonal spheroids may be located in peripheral nerves (e.g., ventral root segments), and that, in inherited motor neuropathies, signs may indicate some degree of length dependence. In some neuronal degeneration, the brunt of the disease process appears to be the perikaryon, in others the nucleus, and in still others the axon just distal to the soma. If severe enough, the entire neuron degenerates. Chapter 35 outlines the electrophysiologic features that separate anterior horn from peripheral nerve disorders. Chapter 31 outlines pathologic alterations and mechanisms of degeneration of somas. In acute poliomyelitis, inflammation is localized to the ventral gray columns of the spinal cord, especially damaging the perikaryon of lower motor neuron somas (Fig. 32–55). Figures 32–25 and 32–26 show lower motor neuron cell columns in health. In Fabry’s disease, an X-linked metabolic disorder caused by deficiency of a lysosomal enzyme (see Chapter 81), lipids accumulate especially in somas of primary afferent neurons but also in perineurial and endothelial cells (Fig. 32–56). Neuron somas may also be adversely affected by the action of sensitized mononuclear cells, as may be the case in inflammatory polyganglionopathy (Fig. 32–57).

Wallerian Degeneration FIGURE 32–48 A cross section of a nerve fixed in glutaraldehyde and osmium tetroxide, embedded in epoxy, and stained with methylene blue. This section shows major artifactual separation of endoneurial contents as a result of ice crystal formation. The nerve in fixative was inadvertently frozen during transport to our laboratory. See Color Plate

When axons are transected, distal axons undergo stereotyped sequential alterations resulting in disappearance of the axon and myelin, known as wallerian degeneration.107 In the rat, after nerve transection, axons continue to conduct for 36 to 48 hours, with only a minor reduction of conduction velocity of the fastest fibers. The muscle action

Pathologic Alterations of Nerves

773

FIGURE 32–49 Epoxy transverse sections of sural nerve stained with methylene blue showing prominent Renaut corpuscles—the spherical pale areas in the endoneurium. In the American medical literature, these were sometimes incorrectly identified as infarcts. See Color Plate

FIGURE 32–50 A pacinian corpuscle in epineurium of sural nerve paraffin section stained with hematoxylin and eosin. This is an infrequent finding. See Color Plate

potential also remains relatively normal for 36 hours and then drops to 0 over the next 6 to 8 hours. In transverse electron micrographs at 36 to 48 hours, some axons have a normal appearance, others show clumping of their cytoskeleton, and still others are devoid of axonal contents and show only degenerating organelles. By the third or fourth day, most recognizable axonal contents will have disappeared (Fig. 32–58). Schlaepfer and Hasler133 have suggested that axonal contents are extruded into the interstitial fluid, but immunohistologic localization of neurofilament (NF) and tubular protein identification remain to be done to confirm this observation. Using immunologic markers, Bignami and colleagues13 demonstrated that axonal debris reacting with NF antisera disappeared much earlier from sectioned sciatic nerve (10 days) than from sectioned optic nerve (up to 4 months). The sequential changes that fibers undergo in wallerian degeneration are studied to advantage in teased fibers. Beginning at 36 to 48 hours, MFs become segmented into linearly arranged, closely spaced ovoids and myelin balls. It appears that separations occur at nodes of Ranvier and at Schmidt-Lanterman incisures. Over the next 2 to 4 weeks, the ovoids become progressively smaller and are increasingly separated from each other by longer regions of the nerve tube without ovoids. Before their final disappearance, ovoids lose their osmiophilia and are widely separated from each other. In young adult rats (~300 to 500 g), segmentation of myelin is evident in essentially all fibers at 3 to 6 days,

774

Pathology of the Peripheral Nervous System

FIGURE 32–51 A, Within the cytoplasm of the Schwann cell of a myelinated fiber are clusters of ␲ granules (arrow), also called Schwann cell cytoplasmic lamellar bodies or zebra bodies (⫻8480). B, Higher magnification of ␲ granule cluster. C, Electron micrograph of mast cell from nerve (⫻11,280). The metachromatic staining properties of this cell may be confused with the metachromasia of metachromatic leukodystrophy disease granules or with deposits of amyloid. D, Portion of longitudinal section through a Schmidt-Lanterman incisure, showing deposit of glycogen (arrow) (⫻39,480). E, Perineurial cell with glycogen (arrow) (⫻20,800).

discrete separation into ovoids and balls is typical in 1 to 3 weeks, following which the ovoids and balls are smaller and separated by long distances, and thereafter only small and widely spaced osmiophilic droplets are seen. It is important to recognize that the linear orientation of ovoids and late breakdown products in nerve tubes are maintained for

weeks or months. This linear array, consisting of the basement membrane of the former nerve fiber, Schwann cells, and macrophages, appears to be important for regeneration to originally innervated target tissue. Under the electron microscope, the ovoids initially have the characteristic periodicity of myelin, but later

Pathologic Alterations of Nerves

775

FIGURE 32–52 The rod-shaped ␲ granules of Reich (discussed in text) are shown in cytoplasm adjacent to compact myelin in both A and C. The boxed areas in A and C are shown at high magnification in B and D, respectively. The lamellar periodicity of granules is approximately 4 to 5 nm. (From Dyck, P. J., and Lambert, E. H.: Polyneuropathy associated with hypothyroidism. J. Neuropathol. Exp. Neurol. 29:631, 1970, with permission.)

these degenerate into amorphous lipid, which loses its osmiophilia and becomes sudanophilic. The events of myelin degradation begin in autophagic vacuoles of Schwann cells. Later, myelin digestion appears to take place within macrophages in the nerve tubes referred to above. Macrophages of hematogenous origin play important roles in dealing with debris. Beginning at approximately 6 days, there is a great influx of these

cells into the endoneurium, and these enter the nerve tubes and aid in fiber degeneration. Simultaneously, there is a great increase also in the number of Schwann cell nuclei. The total increase in nuclei is variable, depending on the fiber composition of the nerve, but it may be as much as six fold or greater. The volume of the endoneurial area may increase considerably during this time.

FIGURE 32–53 A ␮ granule of Reich (Elzholz body). A myelin ovoid in the Schwann cell cytoplasm is indicated by a single arrow. A similar compact lamellar structure is shown by the double arrows (as discussed in text). AC ⫽ axis cylinder. (From Dyck, P. J., Johnson, W. J., Lambert, E. H., and O’Brien, P. C.: Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin. Proc. 46:400, 1971, with permission.)

776

Pathology of the Peripheral Nervous System

FIGURE 32–54 A lipocyte in the endoneurium of sural nerve (transverse section, epoxy embedding, methylene blue stain) is a most unusual finding. See Color Plate

The sequence of wallerian degeneration of unmyelinated fibers has not been studied as intensively as for MFs. As for MFs, so also for the unmyelinated fibers: the earliest evidence of fiber alteration is the dissolution and clumping of the NFs and MTs of axons (Fig. 32–59). The axons of transected distal nerve unmyelinated fiber stumps at 2 to 3 days may be recognized by their position in a trough external to the Schwann cell cytoplasmic membrane and by their having an intact axolemma but axoplasm that appears watery, with cytoskeletal elements decreased or clumped. At this stage some unmyelinated fibers in a group of those associated with one Schwann cell may show degenerative changes, and others may still be normal. At later stages the trough of former sites of unmyelinated fibers may be recognized. Troughs may be filled with lipid deposits (e.g., in Tangier disease) or by collagen (collagen packets). At later times, remnants of myelin may persist for 30 to 60 days in bands of Büngner (Fig. 32–60). Clusters or stacks of Schwann cell cytoplasmic processes surrounded by a common basement membrane may indicate the previous site of a MF or of a cluster of unmyelinated fibers. It was previously thought that wallerian degeneration is highly stereotyped and inevitable both by the nature of the histologic reactions and by temporal events. Studies of certain mutant mouse strains indicate that this is not always the case, the process being much altered in certain mutations. Perhaps this observation suggests that extrinsic factors, not yet fully understood, play key roles in the development and progression of wallerian degeneration. Wallerian degeneration occurs in many human diseases. The causes are physical transection, compression, tear,

FIGURE 32–55 1, Round-cell infiltration of the ganglion cells of the lateral group of the gray matter, in the sacral region of the cord, from a polio case two and a half days after the onset. 2, Ganglion cells from the same region in the same case almost destroyed by phagocytic neuroglia cells (neurophages). 3, Infiltration of the lymphatic space of a central vessel from the lumbar cord of the same case. 4, Infiltration of a vascular space and of the surrounding tissue in the anterior horn of the lumbar cord in the same case. (Reprinted with permission from Wickman, I.: Acute poliomyelitis. J. Nerv. Ment. Dis. 16:1, 1913, with permission.) See Color Plate

radiation, burn, or other physical injuries. External nerve compression may be caused by tumor enlargement beneath tight fascial sheaths or in foramina. Another cause is ischemic transection. Wallerian degeneration may be caused by external compression of a fiber by focal interstitial pathologic alterations, as from a nodule of amyloid. We have postulated that it may be caused by a loop of myelin indenting and transecting the axon (internal strangulation; see below).37 Wallerian degeneration may be initiated at sites of active segmental demyelination.6 According to this view, the axons could be damaged as a nonspecific consequence of a specific cell-mediated immune reaction occurring in the vicinity—that is, a consequence of the so-called bystander effect.150,163 Toxic factors secreted by activated lymphocytes or other metabolic events have been postulated.109,132

Pathologic Alterations of Nerves

777

FIGURE 32–56 Photomicrographs from patient with Fabry’s disease, a lysosomal disorder resulting in storage of glycosphingolipid. A, One cell body appears to be filled by the stored material. B–D, Electron micrographs showing the typical lamellar structure of the stored material. (From Ohnishi, A., and Dyck, P. J.: Preferential loss of small sensory neurons in Fabry’s disease: histological and morphometric evaluation of cutaneous nerves, spinal ganglia, and posterior columns. Arch. Neurol. 31:120, 1974, with permission.)

Demyelinative Internal Strangulation The idea that secondary demyelination might cause axonal degeneration came from studies of the permanent axotomy model (see Permanent Axotomy Studies below). In this model axonal atrophy leads to secondary demyelination and axonal degeneration. Our studies suggest that, as the Schwann cells retract from the node, axial myelin buckling occurs. The infolded myelin loops may, at some sites, be large enough to compress and transect the axon. We have repeatedly demonstrated such myelin infoldings in various demyelinating conditions: after application of a pneumatic cuff,32 in the permanent axotomy model, and in various metabolic and inherited diseases. That the axon can be compromised by these myelin infoldings is suggested

by the finding that the infolded loop may contain fullthickness myelin, a size sufficient to severely compromise axonal contents. Axonal Degeneration in Severe Demyelination In acquired demyelinating neuropathies it is common to also find axonal degeneration. Could this axonal degeneration be the result of severe demyelination? Teased fibers that proximally show segmental demyelination and distally show axonal degeneration (teased fiber condition I) suggest that the two processes may be linked. Others have explained the occurrence of both processes by a bystander effect. Inflammation with the release of cytokines may damage Schwann cells (myelin) and also axons.

778

Pathology of the Peripheral Nervous System

FIGURE 32–57 Paraffin section with hematoxylin and eosin staining of spinal ganglion from a patient with inflammatory polyganglionopathy as described in the text. (Reprinted with permission from Grant, I. A.: Treatment of nonmalignant sensory ganglionopathies. In Noseworthy, J. H. [ed.]: Neurological Therapeutics: Principles and Practice. Vol. 2, London, Martin Dunitz, p. 2070, 2003, with permission.) See Color Plate

Axonal Dystrophy Definition Generally, the term axonal dystrophy is used synonymously with axonal swellings or axonal spheroids, which are multifocal accumulations of cytoskeletal elements. Here we extend the definition to include focal accumulations of other organelles or other formed structures. History Seitelberger found argyrophilic axonal swellings in the brains of children who died from a disorder that he named neuroaxonal dystrophy.136 Subsequently, axonal swellings were also described in various human disorders (giant axonal neuropathy, amyotrophic lateral sclerosis, hexacarbon intoxications, and various types of HMSN) and in experimental neuropathies (those caused by acrylamide, ␤,␤⬘-iminodipropionitrile [IDPN], n-hexane, methyl butyl ketone, and carbon disulfide). With the introduction of the electron microscope, these enlargements were shown to be focal accumulations of cytoskeletal components, particularly of NFs. Impairment of transport of NF components was implicated in these accumulations. Description of Typical Axonal Spheroids Axonal spheroids may be found in proximal nerves (e.g., in amyotrophic lateral sclerosis, neuroaxonal dystrophy, and IDPN and dimethyl hexanedione neuropathies) and in distal nerves (2,5-hexanedione–induced neuropathy, giant axonal neuropathy, and a variety of HMSN).

Giant axonal neuropathy is a disorder of children, probably recessively inherited, with large axonal spheroids in both peripherally and centrally directed primary afferent neuron axons (Fig. 32–61).4,28,29 Abnormal intermediate filament inclusions are also found in Schwann cells, endothelial cells, and fibroblasts. In toxic neuropathies the distribution of axonal spheroids may be related to the dosage and potency of the toxins. Typically, spheroids enlarge the axon, forming ellipsoidal accumulations of NFs. Neurofilaments are absolutely increased in number, closely packed, and often misdirected. In transverse sections of spheroids, several patterns are recognized: accumulation of NFs only; segregated regions of NFs only and of MTs only; and segregated regions of NFs, MTs, and accumulated organelles and vesicular profiles (dense lamellar, vesicular, and paracrystalline structures). Although the axon is generally enlarged, sometimes greatly, at other times only modest enlargement is found. The overlying myelin, especially near the equator of the spheroid, may be absent or thinned. At the poles of spheroids or in small spheroids, myelin may be of normal thickness. Distal to an axonal spheroid, the axon caliber is smaller than normal (atrophied) or may undergo degeneration. Usually the term axonal spheroid is used for the pathologic alteration of fibers in the conditions listed above, for example, giant axonal neuropathy and neuropathies caused by many industrial toxins. Accumulation of axonal organelles different from those described above may occur in compression and ischemic injury. Korthals and Wisneiwski96 and Nukada and colleagues114,115 have described in detail the three-dimensional alterations for ischemic lesions. If serial sections of MFs are followed from a nonischemic region to an ischemic region, as studied by Nukada and colleagues,115 the following changes from proximal to distal are encountered: (1) no abnormality of the axon; (2) slight enlargement and increased density of the axon (with or without light cores) and increased darkness on electron microscopy as a result of accumulated mitochondria, dense bodies, and vesicular profiles; (3) severe enlargement of the axon with accumulated organelles; (4) attenuated dark axons (also with accumulated organelles); and (5) disappearance of the axons or return to a normally appearing axon (see Fig. 32–58 below). Light cores are relatively preserved axoplasm surrounded by accumulated organelles. At the lateral margin of ischemic cores, the changes are less severe. Changes 1 to 4 may be seen, but loss of axons is less severe. Myelin thickness and size may be unaffected with changes 1 or 2 above. With change 3 above, myelin is usually thinned or absent. With changes 4 and 5 above, the myelin profile may be spherical or collapsed but the myelin may be of normal thickness or thin. The changes described above were observed within days or a week or two of the ischemic event. When fibers in ischemic cases undergo degeneration, events of axonal degeneration and repair occur and

Pathologic Alterations of Nerves

779

FIGURE 32–58 Transverse electron micrographs of peroneal nerve of mouse at 36 (A), 48 (B), 72 (C), and 96 (D) hours after sciatic nerve transection (⫻8000). At 36 hours, the appearance of myelinated fibers (MFs) is not greatly disturbed. Note the relative preservation of unmyelinated fibers (UFs). There is dissolution of neurofilaments and microtubules in axons. At 48 hours, granular clumping of axonal contents of MFs and degenerative changes in UFs are evident. By 72 hours, nerve microtubules of MFs show myelin ovoids without axons and many UFs have disappeared. By 96 hours, these changes are more advanced.

regenerating axons replace the degenerating ones. At a later date only small regenerating fibers occupy the ischemic cores. A different focal enlargement of axons is represented by polyglucosan bodies (PGBs) (Fig. 32–62). In this condition the transverse area of axoplasm probably remains normal

but total axonal area is increased because of the presence of PGBs. The PGBs are focal accumulations of abnormal glycogen polymers and are found in nerve fibers in both health and disease. Assessing the number of myelin lamellae and myelin spiral length, Yoshikawa and co-workers165 found that the number of myelin lamellae overlying PGBs

780

Pathology of the Peripheral Nervous System

FIGURE 32–59 Wallerian degeneration in distal segment of cervical sympathetic trunk 34 hours after crush. Many degenerating fibers have a watery appearance; others have material that is clumped and densely opaque. (From Dyck, J., and Hopkins, A. P.: Electron microscopic observations on degeneration and regeneration of unmyelinated fibres. Brain 95:223, 1972, with permission.)

bodies was normal but the myelin spiral length was greatly increased as compared to internode segments beyond the bodies. Because movement of myelin components along membranes is unlikely to account for this focal increase of the myelin sheet, local synthesis of myelin components is postulated. Accumulation of a variety of organelles (glycogen, myelin figures, endoplasmic reticulum, degenerated mitochondria, and lysosomes) may be encountered in both experimental and human neuropathies. Although these may not cause focal enlargements (swellings), they may represent dystrophic changes.

Axonal Atrophy, Myelin Remodeling (Secondary Demyelination), and Degeneration This subject was reviewed as a separate chapter in the second edition of this textbook.51 Although early pathologists talked about axonal atrophy as a mechanism by which fibers disappeared, the histologic events had not been described. The cellular events involved were revealed by a study of sural nerves of patients with uremic neuropathy and Friedreich’s ataxia and by a study of the permanent axotomy model. The

Pathologic Alterations of Nerves

781

A

FIGURE 32–60 Electron micrographs of bands of Büngner. A, Six days after crush. B, Fifteen days after crush (AC ⫽ axis cylinder). One, or possibly two, axon sprouts are seen in B. (From Dyck, P. J.: Ultrastructural alterations in myelinated fibers. In Desmedt J. E. [ed.]: New Developments in Electromyography and Clinical Neurophysiology. Basel, S. Karger, p. 192, 1973, with permission.)

B

sequential steps involved are postulated to be as shown in Figure 32–63. Studies of Uremic Neuropathy and Friedreich’s Ataxia The mechanism underlying secondary segmental demyelination was suggested from results of study of the sural nerve from the midcalf and ankle levels from patients with uremic neuropathy.38 The findings were as follows: 1. The MF density was much lower at the ankle than at the midcalf level, indicating greater fiber loss distally. 2. The peaks of fiber diameter histograms were at smaller diameter categories at the ankle than at the midcalf level, suggesting disappearance of the largest axons from distal nerves or distal fiber atrophy. 3. Teased fibers from the midcalf level showed myelin wrinkling and paranodal demyelination predominantly,

whereas teased fibers from the ankle level showed axonal degeneration predominantly, suggesting that the lesser pathologic involvement caused segmental demyelination and the greater involvement caused axonal degeneration. 4. Teased fibers frequently showed multiple regions of demyelination or remyelination of the same fiber, indicating that myelin remodeling was clustered on certain fibers. 5. The slopes of regression lines relating axonal area to myelin thickness were lower for distal than for proximal nerve and both were lower than in normal nerve, an indication that axons were attenuated relative to the amount of myelin. 6. Serial electron micrographs of fibers with multiple segments of demyelination and/or remyelination had small axons relative to the thickness of myelin of old internodes.

FIGURE 32–61 Selected electron micrographs from a young Iranian boy with giant axonal neuropathy. Upper left, Transverse section through a typical spheroid. The axon is without myelin and is surrounded by concentric layers of Schwann cell processes. Upper right, A longitudinal view of a spheroid in a myelinated fiber. Lower left, An axon showing a central zone of neurofilaments surrounded by a circumferential zone containing microtubules and accumulated organelles. Lower right, Electron micrograph of transverse section through an axonal spheroid, showing close spacing of neurofilaments and absence of side arms—considered by the authors to be evidence of an abnormality of neurofilament cross-linking in this disorder. (From Donaghy, M., King, R. H. M., Thomas, P. K., and Workman, J. M.: Abnormalities of the axonal cytoskeleton in giant axonal neuropathy. J. Neurocytol. 17:197, 1988, with permission.)

Pathologic Alterations of Nerves

783

FIGURE 32–62 Transverse section through a myelinated fiber of a patient with chronic inflammatory demyelinating polyneuropathy who also had frequent polyglucosan bodies (PGBs) in axons, as shown here in the lower left. The PGB had caused focal axonal enlargements that had not changed the number of overlying lamellae but was associated with increased myelin spiral length (MSL). The increased MSL was thought to be due to local synthesis of myelin components rather than movement of lipid or protein components along the membrane. BM ⫽ basement membrane; IM ⫽ inner membrane; OM ⫽ outer membrane. (From Yoshikawa, H., Dyck, P. J., Poduslo, J. F., and Giannini, C.: Polyglucosan body axonal enlargement increases myelin spiral length but not lamellar number. J. Neurol. Sci. 98:107, 1990, with permission.)

The idea of a secondary form of segmental demyelination was also suggested by a study of nerves from patients with early Friedreich’s ataxia. In biopsy samples of sural nerve from such patients, Dyck and Lais43 found that 1. Diameter histograms of MFs of patients with early involvement showed striking abnormalities, with the peak for small fibers being higher and to the left of the peak of normal nerves. The height of the largediameter fiber peak was smaller than in controls, suggesting that some large fibers had become smaller and some small fibers had become even smaller. 2. Most teased fibers had a normal appearance, but others had multiple regions of demyelination and/or remyelination, indicating that the events of demyelination and remyelination were not randomly distributed among Schwann cell territories but were clustered on certain abnormal (atrophying) fibers (Fig. 32–64). 3. Electron micrographs from fibers with demyelination and remyelination showed axonal atrophy.

4. Axonal degeneration occurred occasionally but appeared to be a final event in axonal atrophy. From these studies from human neuropathies, a number of conclusions can be drawn: 1. The study of fiber degeneration in uremic neuropathy and Friedreich’s ataxia supports the idea that there is a type of segmental demyelination and remyelination that is associated with axonal atrophy. Dyck and colleagues called this type of demyelination and remyelination “secondary.”38,43 This secondary segmental demyelination and remyelination should not be confused with nonspecific loss of myelin staining resulting from loss or decreased numbers of fibers. 2. Secondary demyelination is characterized by the following features: a. Initially, there is wrinkling of myelin of old internodes. b. Then, paranodal demyelination occurs.

784

Pathology of the Peripheral Nervous System

FIGURE 32–63 The proposed cellular events of chronic neuronal (axonal) atrophy, myelin remodeling, and degeneration from a study of uremic neuropathy and Friedreich’s ataxia. The permanent axotomy model has provided evidence for steps 1, 2, 3, 4, and 5. Several cycles of demyelination and remyelination (indicated by steps 5 and 7) may occur with development of hypertrophic neuropathy. Depending on the balance of degenerative and regenerative events, fibers might recover from various stages of fiber alteration, dwindle away, or undergo acute degeneration. These cellular events are common in a variety of metabolic and inherited diseases affecting peripheral and, conceivably, central nervous system neurons. MF ⫽ myelinated fiber. (Modified from Dyck, P. J., Lais, A. C., Karnes, J. L., et al.: Permanent axotomy, a model of axonal atrophy and secondary segmental demyelination and remyelination. Ann. Neurol. 9:575, 1981.)

c. With a and b, it may be possible to demonstrate reduced caliber of the axon relative to number of myelin lamellae. d. Secondary demyelination may be recognized in some diseases and in teased fibers by certain fibers having multiple demyelinated and remyelinated old internodes, whereas adjacent fibers are without these changes. In other words, the demyelination is clustered on some fibers and not on others. 3. Axonal atrophy, myelin remodeling with or without hypertrophic neuropathy, and axonal degeneration or loss constitute a common pattern of fiber abnormality in human inherited, metabolic, and other neuropathies. 4. Myelin alterations occur, including radial and axial infolding and myelin cleavage, discussed in detail later (see Permanent Axotomy Studies below). 5. Finally, there is adaxonal sequestration and removal of cellular breakdown products (Fig. 32–65). The sequence of postulated cellular events underlying axonal atrophy is shown in diagrammatic form in Figure 32–63. This sequence might be thought of as a common pathway for injury of various types. The scheme suggests that chronic neuronal (axonal) injury may lead ultimately to axonal degeneration, atrophic disappearance, or recovery. In an adult, a previously atrophic fiber with multiple segments of demyelination might be recognized after

recovery by having multiple short (remyelinated) segments intermixed with longer normal internodes. After recovery in a child whose limbs were still growing, the previously remyelinated internodes would elongate and could not be recognized as remyelinated because they are no longer short. Retrograde Effects of Nerve Transection Gutmann and Sanders71 evaluated the number and sizes of MFs in the proximal peroneal nerve of rabbit at various times up to 300 days after crush, resuturing, and nerve grafting of the distal nerve. They found no decrease in the number of MFs but a shift in caliber toward smaller sizes by 120 days, which was no longer present after 200 days. To explain the reduced NCV of nerves proximal to such distal peripheral nerve lesions as carpal tunnel syndrome, Cragg and Thomas24 compared the NCV and the size of the five largest MFs in each of four sections from the peroneal nerves of a rabbit proximal to the following lesions: crush, ligature, transection and reconnection, and transection and distal nerve avulsion. In the first three experiments, NCV fell to 90% of normal by 25 to 30 days, and to 80% by 50 to 100 days, and then rose to normal by 200 days. In sharp contrast, in transection and distal avulsion NCV fell to between 60% and 70% of normal by 200 days and did not

Pathologic Alterations of Nerves

785

FIGURE 32–64 Consecutive lengths along one teased myelinated fiber from the sural nerve of a patient with Friedreich’s ataxia. The atrophy of this fiber is evident from the multiple regions of demyelination and remyelination of the same fiber. Many contiguous large-diameter fibers had no demyelination, an indication that the demyelination is not occurring in a random fashion between “old” territories of Schwann cells but is clustered on some axons. The demyelination is therefore secondary to an axonal influence as discussed in this chapter (see text) and in Chapter 37. (From Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerves in health, Friedreich’s ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb W. A. [ed.]: Handbook of Electroencephalography and Clinical Neurophysiology, Vol. 9. Amsterdam, Elsevier, p. 83, 1971, with permission.)

improve thereafter. Later, Aitken and Thomas2 concluded that atrophy affected both myelin and axons but that the axons were slightly more atrophied. Assuming that no segmental demyelination occurred, these findings suggest that fiber atrophy is transitory when the nerve is allowed to regrow to its target tissue, but it is more severe when reconnection is inhibited. After recording from the spinal roots of cats whose leg nerves had been cut and prevented from regrowing, Hoffer and co-workers79 reported a decline of NCV, which stabilized for ventral root fibers but was progressive for posterior root fibers.

Lubinska106 observed that, if only part of an internode was damaged by crush, the remainder (containing the Schwann cell nucleus) might survive in a truncated condition. This type of damage may be related to direct injury to the Schwann cell. Lubinska also noted that “in a certain proportion of fibers . . . the reaction to the lesion spreads back in the proximal direction and may involve several internodes.” The proximal end of the last “preserved internode” became demyelinated, and the space between the axon and neurilemma was filled with hypertrophied Schwann cell cytoplasm and many nuclei. The frequency of such demyelination was highest in the second to the last

786

Pathology of the Peripheral Nervous System

FIGURE 32–65 A, Small, myelinated fiber showing irregular configuration of axolemma with densely packed microtubules and neurofilaments in its axoplasm. (Original magnification, ⫻18,000.) B, Large myelinated fiber with abnormal round and oblong membrane-bound profiles containing dense bodies, mitochondria, and filaments between axolemma and compact myelin. (Original magnification, ⫻8500.) (From Ohnishi, A., Peterson, C. M., and Dyck, P. J.: Axonal degeneration in sodium cyanate-induced neuropathy. Arch. Neurol. 33:530, 1975, with permission. Copyright 1975, American Medical Association.)

internode and less in the third to the last internode. Lubinska observed wrinkling, folding, and invagination of the myelin of the terminal internodes. After paranodal demyelination, the demyelinated segment was replaced by short “intercalated” remyelinated segments. Lubinska likened these intercalated internodes to those found after recovery from periaxonal neuritis produced by lead, alcohol, and mechanical trauma63,64 and by other factors.8 She observed that the segmental demyelination occurred within days of crush and was quickly followed by repair. The mechanisms for the myelin remodeling were not explored further. The retrograde teased fiber effect of peroneal nerve transection and distal nerve avulsion was studied in the rat by Spencer and Thomas.146 Their interest focused on portions of teased fibers with myelin swellings or “bubbles.” The nerve above and below the bubble was reported to have a normal appearance but was composed of short internodes of smaller diameter and with thinner myelin than normal. Transverse sections through the myelin

swellings revealed that an axon ran through the swelling, partly surrounded by thin myelin and partly without myelin. The myelin had been split at the interperiod line to form large vacuoles containing myelin debris and macrophages. The authors reported that the bubbles formed on remyelinated portions of a fiber but were uncertain about the cause of these bubbles, although they noted similarities to alterations occurring in experimental tin intoxication and experimental allergic encephalitis. In a later report, Spencer144 stated that there was fiber attenuation and suggested that fiber degeneration starts distally and retrogresses to the cell body. An attempt to quantitate fiber attenuation by regressing internodal length on fiber diameter, however, failed to demonstrate a significant difference in the line for amputated and control nerves. The study of the retrograde effects of permanent axotomy in cats confirmed the correctness of the concept that axonal atrophy leads to secondary demyelination and axonal degeneration.

Pathologic Alterations of Nerves

Permanent Axotomy Studies The seventh segmental nerve and sciatic nerve stumps were studied in groups of cats 3 weeks and 4, 12, and 24 months after hind limb amputation. The sequential changes that MFs undergo were readily demonstrated in teased fibers (Figs. 32–66 and 32–67). At 3 weeks there was a statistically significant increased percentage of stump fibers showing myelin wrinkling. By 4 months a significantly increased percentage of amputated teased fibers showed wrinkling even in the proximal nerve. At this time also the percentage of teased fibers showing demyelination and remyelination was increased in the distal segment of the amputated side compared to the nonamputated side. By 12 months the frequency of demyelinated and remyelinated fibers had reached statistical significance. By 24 months significant percentages of teased fibers were undergoing axonal degeneration in both proximal and, to a greater degree, distal segments of the stump nerves.

FIGURE 32–66 Pathologic abnormalities of teased fibers derived from amputated proximal nerve stumps and comparable levels of nonamputated nerve. A, Four consecutive lengths of a teased fiber from the nonamputated side without abnormality (type A). B, Two consecutive lengths of fiber showing a slight to moderate degree of myelin wrinkling sometimes seen at 3 weeks but more characteristically seen 4 months after amputation (type B). B⬘, Three consecutive lengths of fiber showing severe myelin irregularity characteristic of the change seen at 4 and 12 months (type B). C, Severe myelin wrinkling and partial nodal lengthening at 12 months (type C). D, Three consecutive lengths of fiber showing myelin wrinkling, demyelination, and remyelination at 12 months (type D). F, Four consecutive lengths of fiber showing myelin wrinkling and segmental remyelination, located approximately at the former nodes of Ranvier, at 12 months (type F). (From Dyck, P. J., Lais, A. C., Karnes, J. L., et al.: Permanent axotomy, a model of axonal atrophy and secondary segmental demyelination and remyelination. Ann. Neurol. 9:575, 1981, with permission.)

787

Insight regarding the change in size of MFs and their components is gained from inspection of histograms of MF diameter, myelin areas, and axon areas of the L7 segmental nerve and the sciatic nerve (Fig. 32–68). The most severe alteration appears to be the decrease in axonal area. Assuming the primary event to be axonal atrophy, regression lines relating axonal area to myelin thickness in amputated as compared to nonamputated sciatic nerves are revealing (Fig. 32–69). Axons are smaller, relative to myelin thickness, in axotomized nerves than they are in control nerves. To test whether decreased NF production might underlie the atrophy after permanent axotomy, the number of NFs was regressed on axonal area, number of myelin lamellae, or myelin spiral length of proximal and distal segments of control and axotomized sciatic nerves. Regression lines relating numbers of NFs to axonal area were not significantly different between control and

788

Pathology of the Peripheral Nervous System

These results led to the following conclusions: 1. Permanent axotomy by hind limb amputation provides an informative model of chronic axonal atrophy and of myelin changes. 2. Permanent axotomy confirms conclusions, based on study of nerves of patients with uremic neuropathy and Friedreich’s ataxia, that MFs might undergo degeneration by the following sequential steps: Axonal atrophy : myelin wrinkling : segmental demyelination and remyelination : further atrophy : axonal degeneration.

FIGURE 32–67 Average percentages of teased fiber pathologic abnormalities at various intervals of time after amputation (A) compared with contralateral control (C). A, No abnormality. B, Myelin wrinkling. C, D, and F, Demyelination and remyelination as described in the text. E and H, Axonal degeneration and axonal regeneration. After axotomy, fibers undergo myelin wrinkling, then at a later time demyelination and remyelination, and finally axonal degeneration as described in the text. The changes begin in distal nerve and are more severe in distal (sciatic) than in proximal (segmental) nerve.

amputated nerves. This finding indicates that the density of NFs is unchanged even in atrophic fibers. Regression lines relating numbers of NFs to myelin spiral length were significantly lower for amputated than for nonamputated nerves at both proximal and distal levels. This result suggests that NF number is decreased in atrophic fibers. An explanation for axonal degeneration is needed. It may be that, when the axon becomes very attenuated, it no longer can sustain a distal axon and degenerates. Alternatively, an infolded loop of myelin can cause demyelinative internal strangulation (Fig. 32–70).

3. The events of axonal atrophy, myelin remodeling, and axonal degeneration begin and appear to be more severe in the distal nerve segment but later also develop in the proximal nerve segment. 4. The myelin changes can be explained by an attenuation of the volume of the axon in the presence of a mostly unchanged myelin membrane sheet, which therefore becomes too large for the axon. With a further decrease in axonal caliber, the wall collapses radially, so that flattened and crenated shapes develop in a transverse profile. With retraction of Schwann cells from the node, further axial wrinkling and infolding develop. Clefts, sometimes beginning at Schmidt-Lanterman incisures, are thus formed in compact myelin. 5. Two mechanisms are suggested by which axonal atrophy may lead to axonal degeneration. In the first, the decrease in NF content of the axon reaches a critical level below which axon integrity cannot be maintained. In the second, myelin infoldings compress the axon, resulting in internal axonal strangulation (described earlier). 6. Decreased numbers of NFs could explain the axonal attenuation induced by permanent transection. Decreased synthesis of NF components after nerve transection has been reported by Hoffman and Lasek.80

SEGMENTAL DEMYELINATION AND REMYELINATION Myelin Remodeling in Normal Nerves During development, Schwann cells come to occupy short territories along axons. Each of these Schwann cells forms an internode of myelin. The place at which adjacent internodes abut is the node of Ranvier. With lengthening of the limb during growth, internodes also lengthen. In normal nerve, the Schwann cells of internodes appear to remain relatively unchanged in form over many years—sometimes for a lifetime. This conclusion seems to follow from the observation that the mean ILs of MFs of nerves remain approximately the same for

Pathologic Alterations of Nerves

FIGURE 32–68 Composite histograms of myelinated fiber diameters (left), myelin areas (middle), and axon areas (right) of nonamputated (control) and amputated segmental (upper five rows) and sciatic (lower two rows) nerves of cats to show the altered size distribution that follows axotomy, as described in the text.

FIGURE 32–69 Common regression lines and regression line intercepts for individual nerves of ln axon area to myelin thickness (left) and of index of circularity (IC) to myelin thickness (right) for nonamputated (⫹ symbols and continuous lines) and amputated (open circles and broken lines) sciatic nerve at 24 months after amputation. Both axonal area and IC are lower for axotomy nerves than for controls affecting fibers with all myelin thicknesses.

789

790

Pathology of the Peripheral Nervous System

FIGURE 32–70 Left, Partial view of a myelinated fiber showing the compression that an infolded loop of myelin may produce on the axon. This may be sufficient, in some cases, to cause axonal interruption and distal regeneration. Right, A partial myelin cleft that appears to begin at a Schmidt-Lanterman incisure (arrow).

most of adult life and probably decrease in old age and the fact that, in health, myelin membrane protein and lipids are incorporated at a very low rate. Actually, with age an increasing percentage of paranodal regions may break down. New short intercalated internodes are formed in these demyelinated paranodes. The term segmental demyelination, as used in this chapter, refers to degeneration of the myelin of a paranode (paranodal segmental demyelination) or of an internode (internodal segmental demyelination). As described earlier, absence of paranodal or internodal myelin may be developmental, physiologic (at a low frequency even in healthy persons), primary (metabolic, immune, compression, and other causes), or secondary (to axonal enlargement or atrophy). It is noteworthy that, for most demyelinating neuropathies, remyelination occurs concurrently with demyelination. This suggests that the factors that induce demyelination are presumably not the same ones that would interfere with remyelination. In an experimental study, after nerve crush, regeneration and remyelination occurred as in control animals even when the experimental group had circulating antibodies to galactocerebrosides and while experimental allergic neuritis was being induced.149

Segmental demyelination was first described by Gombault.63,64 He emphasized the segmental nature of the process, integrity of the axon, and smaller ovoids and lipid droplets than in wallerian degeneration. Since the time of Gombault, several varieties of primary segmental demyelination (macrophage induced, pressure induced, antibody induced, and caused by faulty composition of myelin) have been recognized. In addition, a secondary type of segmental demyelination resulting from axonal enlargement or atrophy is now recognized. In human neuropathies, this secondary type of segmental demyelination is common. When confronted with nerves showing segmental demyelination, it therefore now becomes necessary to know whether it is primary (extrinsic or intrinsic events affecting Schwann cells or myelin) or secondary (to axonal enlargement or atrophy). Gombault’s criteria would not discriminate between primary and secondary demyelination.

Differences from Secondary Demyelination Primary and secondary demyelination both show: (1) paranodal or internodal demyelination, (2) remyelination, (3) small myelin breakdown products, (4) intact axons, and (5) normal ultrastructural features of axons.

Pathologic Alterations of Nerves

Three major features distinguish secondary from primary demyelination. In secondary demyelination: (1) the axons of old internodes are (or were) smaller or larger than they should be considering myelin thickness; (2) demyelination and remyelination occur especially on fibers with severe axonal attenuation; and (3) among old internodes, demyelination and remyelination are clustered and not random. A too-small axon may be recognized by infolded loops of myelin between myelin and axon, severely attenuated axons not caused by hyperosmolar fixation, and less-steep regression lines relating axonal area to number of myelin lamellae or myelin spiral length. It should be appreciated that, with myelin remodeling, many fibers may be clothed with newly developed myelin, making interpretation of regression lines difficult. For evaluation of nerve biopsy specimens, it may be too time consuming or difficult to assess for randomness of demyelination among old internode territories. A simpler approach may suffice for clinical purposes. Secondary demyelination should be considered when (1) there is no obvious pathologic alteration that is primary (e.g., inflammation or infiltration); (2) there is unequivocal fiber loss; (3) demyelinated axons (larger than unmyelinated fibers) are not seen; and (4) some teased fibers have several or many old internodes with demyelination and remyelination, whereas many other fibers are normal (clustered demyelination). Irrespective of the type of demyelination, axons do not remain bare except at nodes of Ranvier and quickly become surrounded by Schwann cell cytoplasm and its cytoplasmic membrane. Repeated demyelination and remyelination, whether primary or secondary, result in the formation of onion bulbs and increased collagen synthesis. It has been possible to create onion bulbs experimentally by intoxication with lead99 or phenanthracene162 and repeated application of a pneumatic cuff.32 Dyck postulated the following events: (1) partial or complete demyelination of an internode, (2) mitosis of a Schwann cell in response to demyelination, (3) capture of a demyelinated internode by one Schwann cell, (4) circumferential orientation (influenced by the persistence of an earlier basement membrane) of the displaced Schwann cell and its cytoplasm, and (5) successive outward displacement of layers of basement membrane and Schwann cells.32

DEBRIS REMOVAL FROM THE ENDONEURIUM As outlined in Chapters 3, 29, and 30, nerve fascicles are maintained in a protected microenvironment by the structure and properties of the perineurium, the capillary endothelium, and the anatomic structure at the proximal and distal ends of peripheral nerves. No lymphatic channels are found within the endoneurial space.

791

Our studies, using the dye patent blue (National Aniline Division of Allied Chemical Corporation) to map lymphatic channels, do not provide evidence of intraneural lymphatic channels and indicate that dye removal is delayed when the agent is injected intraneurally (unpublished results). When dye is injected into the epineurial tissue, it is generally removed in several hours and the lymphatic channels by which it is carried to regional lymph nodes can be traced. When microinjections of 2 ␮L of the dye are made endoneurially into sciatic nerves of 300-g rats, the dye spreads almost instantaneously up and down the nerve for a distance of 4 to 6 cm, stains the entire endoneurial contents, and remains unchanged for many hours. After about 6 hours the staining becomes lighter, but pale staining may persist for 2 to 3 days. No lymphatic channels leaving the nerve were identified in several dozen experiments of this kind. In addition to demonstrating the lack of obvious lymphatic channels, these studies indicate that, after injection of minute amounts of fluid into the endoneurial pool, the fluid spreads over a very large distance. Removal of this small volume of dye is greatly delayed compared with removal of dye via lymphatic channels from the epineurial space. In far-advanced lead neuropathy resulting from feeding 4% inorganic lead carbonate in the diet to 200-g rats, macrophages often line up in rows on the inside of the perineurium and (to a lesser extent) around capillaries (Fig. 32–71). These macrophages contain lipid debris. In addition, macrophages may be found between all layers of perineurium and even just inside the last lamella. The perineurial cells may also contain extensive lipid droplets. These findings raise the question of whether macrophages may unload lipid debris to perineurial cells or simply pass through the perineurium and enter the lymphatic system. Subperineurial palisading of vacuolated cells, identified as fibroblasts, was first described by Schoene and co-workers134 in cases of hereditary sensory neuropathy. Various Schwann cell inclusion bodies may be involved in degradative events. The ␲ granules of Reich are elongated, comma- or rod-shaped structures that are found at paranodal regions and at polar ends of Schwann cells.27,47,59,134,152,154 These granules stain metachromatically with basic aniline dyes. With the electron microscope they are seen to be membrane bound and have a periodicity of 4 to 5 nm. Their frequency and size are reported to increase with age and possibly in various disease states.135 We distinguish the ␲ granules of Reich127 from the ␮ granules of Elzholz,57 although the ␲ granules of Reich are commonly situated at paranodal sites in Schwann cell cytoplasm just outside compact myelin. They stain metachromatically with aniline dyes and appear as short rods or curvilinear structures surrounded by a double membrane. Elzholz bodies are osmiophilic and lie adjacent (on the outside or inside of compact myelin). They have a lamellar structure whose

792

Pathology of the Peripheral Nervous System

FIGURE 32–71 Transverse semithin sections of peroneal nerve from animal intoxicated with 4% lead carbonate for 6 months. A, Low-power view showing the subperineurial margination of macrophages. The arrow indicates a region of the perineurium that is very attenuated. B, A macrophage is situated in the perineural defect. Serial sections confirmed the presence of this defect. C, Multinucleated giant cell, presumably a fused macrophage cell. D–G, Other segments of the subperineurial and perineurial regions, showing lipid deposits in perineurial cells, perineurial separation, macrophages between perineurial leaflets, and subperineurial edema.

periodicity is about one third less than that of myelin. As shown in Figures 32–52 and 32–53, such inclusions may also be situated between the axolemma and the myelin sheath. Spencer and Thomas147 postulated the existence of a mechanism for the sequestration and phagocytosis of axoplasmic organelles by Schwann cells in hexacarbon-

induced neuropathies. They suggested that, adjacent to an organelle to be removed, the adaxonal cytoplasm forms a thin ridge that indents the axon. Next, this sheet of adaxonal cytoplasm surrounds the axon, infolds on itself, and sequesters groups of axoplasmic organelles to form an interdigitating profile when viewed in cross section. A possible example of this phenomenon is shown in Figure 32–65.

793

Pathologic Alterations of Nerves

NERVE REGENERATION Soon after focal nerve injury, regenerative events occur. The transected axons become sealed off and are transformed into growth cones that elongate along former Schwann tubes (see Fig. 32–60). Here, they lie among Schwann cell cytoplasmic processes and macrophages in former or old Schwann cell basement membrane (SCBM) tubes. The cluster of axonal sprouts, Schwann cells, and macrophages enclosed in an SCBM tube are the counterpart of what light microscopists called bands of Büngner. These bands of Büngner may be subdivided into those with neurites (myelinated, amyelinated, or unmyelinated) and those without neurites. The key to their identification would be the presence of former SCBM or remnants of SCBM. With time the old SCBM disappear, and regenerating nerve fiber clusters are recognized by their close apposition to one another. Schwann cells multiply inside SCBMs, maintaining their columnar orientation. During early phases of regeneration, proximal segments of regenerating axons become ensheathed by Schwann cells and myelination begins. Usually more than one regenerating neurite is found in one SCBM. Sometimes in abortive regeneration, many (as many as 15 to 20 or more) neurites will be found in a cluster—“regenerative nerve fiber clusters.” We assume that at a later time the number of regenerating axons may be decreased from the earlier value, suggesting that axons that do not innervate target tissue disappear. The completeness of nerve regeneration after nerve degeneration caused by a focal injury depends on the species, age, site of injury, type and degree of complexity of the lesion, type of repair, and interval between the time of the injury and the time of the repair. Our knowledge about adequacy of regeneration has come especially from study of the distal nerve stump at various times after nerve crush, transection and suture, cable grafting, and permanent transection. The striking difference between regeneration after crush (restoration to nearly normal number and size distribution) and that after transection and suture (poor restoration of number and size distribution) is probably related to events and mechanisms at the site of injury, because the distal stump mechanism should not be different between the two. With nerve transection the blood supply is also interrupted, the endoneurial environment is exposed to the epineurial environment, and the integrity of the perineurium and SCBM tubes is breached. By contrast, in crush, vessel continuity is maintained and the perineurium and SCBM tubes remain uninterrupted. Nerve regeneration is always worse after transection than it is after compression, in torn and in macerated lesions than in clean transections, in proximal than in distal lesions, in nerves sutured under tension than without tension, and probably in the old than in the young. To illustrate, in severe brachial plexus lesions virtually

no recovery below the elbow can be expected, whereas virtually complete recovery is usual from a severe compression injury whether proximal or distal. Because degenerative and regenerative events in the distal stump should be the same whether the nerve is cut or crushed, whether there is a gap or not, and so on, the difference must be related mainly to events at the site of injury. Thus factors that influence regenerative outcome probably include (1) the number of neurites growing into the distal stump, (2) whether they are directed to an appropriate or inappropriate target (which they cannot innervate, and therefore they fail to develop or undergo atrophy and degenerate), and (3) whether they are so delayed in their outgrowth to distal limbs that SCBM tubes become sufficiently disorganized that neurites frequently do not regrow to appropriate targets. A theoretical possibility is that, when regeneration is delayed, the target tissue itself can no longer be reinnervated because it has degenerated. In a study of Schwann cell tubes at various times after permanent nerve transection in mouse, changes in the tubes have been described that may help explain poor regeneration (Figs. 32–72 through 32–75). Initially during fiber degeneration, the SCBM has dilated and collapsed regions. Increasingly the SCBM is no longer in close approximation with Schwann cell cytoplasmic basement membrane, thus losing a nutritive association. Even within a few weeks the SCBM begins to fragment. Later the fragmented pieces become dispersed in the endoneurial space. It seems likely that these changes in the normal scaffold for regrowth are sufficient alteration that neurites may not successfully regenerate back to original or appropriate targets. There is suggestive evidence that a species of laminin that is found on the inner surface of SCBM might be important in guidance of regenerating axons. It has been suggested that SCBMs from patients with diabetic neuropathy are less corrugated and collapsed than are those in controls and may persist longer before degenerating.91

Microfascicular Regeneration After nerve transection microfascicles, each containing only a few fibers, may regrow outside or within the main nerve trunk. Korthals and co-workers94 have described an extreme degree of microfascicular degeneration after ischemic nerve injury (Fig. 32–76) in cat.

INTERSTITIAL PATHOLOGY OF NERVES Pathologic alterations of peripheral nerves may be classified into parenchymatous—alterations of the parenchyma (neurons [axons] or Schwann cells [myelin])—or interstitial— alterations of the interstitium. Under interstitial alterations we include inflammatory, infiltrative, mechanical, or entrapment injury and vascular and ischemic injury.

794

Pathology of the Peripheral Nervous System

FIGURE 32–72 A low-power electron micrograph of peroneal nerve of mouse to illustrate the Schwann cell basement membrane (SCBM), which is closely applied to the external surfaces of Schwann cells that invest individual myelinated fibers and unmyelinated fiber bundles. Inset, A tracing of the SCBM at low power. Because the SCBM is continuous from one internode to the next, one must understand this structure to be a sleeve that extends from one end of a peripheral nerve fiber to its other end. Containing as it does a chain of many Schwann cells along its length, after degeneration the sleeve serves to maintain the longitudinal orientation of Schwann cells and to act as a conduit for neurite regrowth. If the nerve is crushed, leaving the SCBM intact, there is a better chance for the neurite to grow back to its original target. If the nerve is cut, a neurite may grow down to an inappropriate target and fail to innervate it; as a result, the regenerating fiber does not develop or undergoes atrophy. (From Giannini, C., and Dyck, P. J.: The fate of Schwann cell basement membranes in permanent transected nerves. J. Neuropathol. Exp. Neurol. 49:550, 1990, with permission.)

Edema and Hemorrhage Although edema of the endoneurial space is characteristic of certain neuropathies, there is little evidence that it, or the associated increased endoneurial fluid pressure, directly causes fiber damage (see Chapters 29 and 30). An exception may be the sudden accumulation of endoneurial fluid resulting from accidental needle injection into nerve45 or from hemorrhage into nerve.105 In most chronic neuropathies with associated edema, there is sufficient compliance that damaging levels of pressure do not develop. Interstitial edema may be severe in experimental lead and galactose neuropathies. Because comparable levels of increased endoneurial pressure are associated with frequent segmental demyelination in lead neuropathy and infrequent axonal degeneration in galactose neuropathy, mechanisms other than edema must be postulated for the fiber degeneration. Interstitial edema is also seen in various types of inflammatory demyelinating polyradiculo-

neuropathies, in certain monoclonal protein-associated and paraneoplastic neuropathies, in myxedema, in acromegaly, and in certain inherited neuropathies (e.g., Déjérine-Sottas disease). In tissue sections, edema must be distinguished from artifactual tissue separation resulting from histologic processing. In well-fixed and processed nerve specimens with intact perineurium, edema fluid accumulates around vessels and beneath the perineurium and may diffusely separate fibers. The total transverse fascicular area of edematous nerves is greater than that of control nerves. Generally, edema fluid also becomes stained with protein stains. Hemorrhage into nerve may be artifactual, as occurs at the time of surgery, or may be due to disease. Hemorrhage at the edges of the biopsy specimen, without inflammatory reaction, is probably related to transection of vessels during biopsy. Disease bleeding into nerve may occur because

Pathologic Alterations of Nerves

FIGURE 32–73 In acute wallerian degeneration, the Schwann cell basement membrane is focally dilated (A) and dilated and collapsed (B). Myelin debris lying free in the endoneurium is seen in A. (From Giannini, C., and Dyck, P. J.: The fate of Schwann cell basement membranes in permanent transected nerves. J. Neuropathol. Exp. Neurol. 49:550, 1990, with permission.)

FIGURE 32–74 Electron micrograph of a collapsed Schwann cell basement membrane (SCBM), showing redundant basement membrane and Schwann cell processes. This is from a permanent axotomy mouse model. The processes in the SCBM are Schwann cell processes without neurites. (From Giannini, C., and Dyck, P. J.: The fate of Schwann cell basement membranes in permanent transected nerves. J. Neuropathol. Exp. Neurol. 49:550, 1990, with permission.)

795

796

Pathology of the Peripheral Nervous System

FIGURE 32–75 At progressively longer times after permanent axotomy (A to D), the Schwann cell basement membranes (SCBMs) fragment and are dispersed. Some of the basement membrane disappears over 3 months. Although the SCBM sleeve was found to remain continuous over a period of 3 months, by this time there were many discontinuities along its length and there was dispersion. Some neurites may now follow original SCBMs to original targets because of this fragmentation and dispersion. (From Giannini, C., and Dyck, P. J.: The fate of Schwann cell basement membranes in permanent transected nerves. J. Neuropathol. Exp. Neurol. 49:550, 1990, with permission.)

797

Pathologic Alterations of Nerves

FIGURE 32–76 Microfasciculation following nerve ischemic lesion at low power (A) and higher power (B). (From Korthais, J. K., Gieron, M. A., and Wisniewski, H. M.: Nerve regeneration patterns after acute ischemic injury. Neurology 39:932, 1989, with permission.)

of trauma, necrotizing angiopathy, hemorrhagic diathesis, and septic emboli. In necrotizing angiopathy the point of hemorrhage seldom is identified in the sections evaluated (Fig. 32–77). Ischemic damage of capillaries or venules may be sufficient to cause diapedesis or frank hemorrhage. Hemosiderin in macrophages provides evidence that the hemorrhage occurred antemortem or prebiopsy. Finding hemosiderin in macrophages in nerve usually implies diseases of vessels and is an important finding, and is common in necrotizing vasculitis (see Fig. 32–77).

Inflammation of microvessels with neural separation and bleeding occurs in necrotizing vasculitis of microvessels. This may occur in Sjögren’s neuropathy, nonsystemic vasculitic neuropathy, and diabetic or nondiabetic lumbosacral radiculoplexus neuropathy (Figs. 32–78 and 32–79).

Compression The subject of compression is reviewed in Chapters 55, 56, and 57. The terms acute compression and entrapment are

798

Pathology of the Peripheral Nervous System

FIGURE 32–77 Alterations in nerve trunks typical of periarteritis nodosa. A, Gross hemorrhage into nerve trunk. B, Occluded small artery surrounded by hemorrhage. C, Occlusion of lumen and segmental absence (necrosis) of media of small artery. D, Complete absence of media and perivascular infiltrate on mononuclear cells of small artery. E, Mononuclear infiltrate around small vessels. F, Hemosiderin macrophages at the site of a severely damaged artery.

not synonymous. The term compression is reserved for a monophasic application of force exerted to the nerve from the outside so that it is compressed against a rigid underlying structure. Acute compression may also occur repeatedly. Entrapment refers to the injury that ensues when a nerve passes through an opening that is too small considering the caliber of the nerve. Acute or repeated compression, stretch, and other factors may be operative. If the nerve is tethered

because of previous injury, stretch, repeated injury, scarring, or other mechanisms may be involved. The characteristic clinical features of nerve injury arising from application of a tourniquet to an arm are (1) muscle weakness overshadows other neuropathic manifestations, (2) recovery may begin in days or weeks, and (3) eventual recovery is expected. Nerve crush is a more severe form of nerve compression. In both compression and crush, the nerve

Pathologic Alterations of Nerves

799

FIGURE 32–78 Serial skip paraffin sections of a microvessel above (upper), at (middle), and below (lower) a region of microvasculitis in the sural nerve of a patient with nondiabetic lumbosacral radiculoplexus neuropathy. The sections on the left are stained with hematoxylin and eosin, the sections in the middle are reacted to anti–human smooth muscle actin (Dako), and the sections on the right are reacted to leukocytes (leukocyte common antigen [LCA]). The smooth muscle of the tunica media in the region of microvasculitis (middle) is separated by mononuclear cells, fragmented, and decreased in amount. The changes are those of a focal microvasculitis. (Reprinted with permission from Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy [LSRPN]: similarity to the diabetic variety [DLSRPN]. J. Neuropathol. Exp. Neurol. 59:525, 2000, with permission.) See Color Plate

fascicles and the nerve basement membrane tubes remain more or less in continuity. By contrast, in nerve transection the nerve fascicles and the nerve fiber tubes are severed and, on resuturing, cut ends are misaligned. In nerve crush virtually all fibers undergo axonal degeneration, but regeneration is much better than after nerve transection and reconnection or after cable grafting. Nerve regeneration after crush injury restores fiber number and size distribution almost to normal. It is thought that very little misdirection of fibers occurs. By contrast, nerve regeneration after nerve transection usually does not restore fiber number or size distribution to normal. In nerve transection, not only fibers but also the gross nerve,

blood supply, fascicles, and Schwann tubes are transected. Even with surgical reconnection there is failure to establish blood supply immediately, fascicles are not adequately aligned, a gap develops between the proximal and distal stump, and SCBM tubes are not correctly apposed to their distal stumps. The effect is that many fibers do not regrow to their original targets or to appropriate targets and so do not innervate tissues and may degenerate. Injury caused by nerve compression is less severe than that caused by nerve crush. The spectrum of alterations may vary from reversible physiologic block to paranodal or internodal demyelination to frank degeneration.

800

Pathology of the Peripheral Nervous System

FIGURE 32–79 Transverse semithin epoxy sections of sural nerve from patients with nondiabetic lumbosacral radiculoplexus neuropathy stained with methylene blue showing multifocal fiber degeneration and loss (top) and reactive repair injury neuroma (bottom), which we attribute to ischemic injury and repair. Top, The upper left fascicle shows an admixture of normal and degenerating (arrowhead) fibers. The lower left fascicle shows an almost normal density of fibers and an occasional degenerating fiber. In contrast, the lower and upper right fascicles are essentially devoid of intact fibers and only rare degenerating profiles remain. Bottom, Transverse section of part of a fascicle showing nerve regeneration. There is a crescent of microfasciculi (injury neuroma, short thick arrow) between the leaflets of thickened perineurium (edges demonstrated by long, thin arrows). The arrowhead identifies a cluster of regenerated fibers. Most of the fibers in the large fascicle are small myelinated fibers and are probably regenerating. (Reprinted with permission from Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy [LSRPN]: similarity to the diabetic variety [DLSRPN]. J. Neuropathol. Exp. Neurol. 59:525, 2000, with permission.)

The mechanism of nerve injury resulting from compression was formerly thought to involve ischemia. DennyBrown and Brenner26 compressed nerves with a spring clip and found that severity of injury was related to duration of compression. Direct compression was not thought to be the essential factor because excised segments of nerve in a

chamber could be severely compressed without their function being lost.70 However, this argument is no longer considered persuasive because shear forces are not operative when nerves are compressed in a chamber. By contrast, in pneumatic cuff compression major shear forces develop at the edge of the cuff. In a series of studies, Ochoa and co-workers116,117 emphasized the role of mechanical injury. They suggested that axons had been telescoped into adjacent internodes with actual displacement of nodes of Ranvier. The loops of myelin that overrode nodes of Ranvier later degenerated and resulted in demyelination. Further evidence that ischemia was not the responsible mechanism was that compression injury was not worsened by a preceding period (of an hour) of leg ischemia.161 Studies by Dyck and colleagues provided definitive further evidence that the major functional and structural alterations of nerve compression were due to mechanical and not ischemic injury.44 As reviewed in other parts of this chapter, histologic alterations attributable to ischemia take from 12 to 24 hours to be expressed. Any structural alterations that are already fully expressed within a few minutes of tourniquet application, therefore, cannot be attributed to ischemic injury.114 In our experiments, histologic alterations were stopped within a minute or so of cuff application by perfusion fixation and by flooding the nerve with fixative. The stereotypical alterations of compression injury were fully expressed within a few minutes; therefore they cannot be due to ischemia. Under the cuff the following sequential changes were observed: (1) the endoneurial fluid was expressed from under the region of the cuff, resulting in close apposition of fibers and of cellular elements (Fig. 32–80); (2) fluid was also expressed from the axon with compaction of cytoarchitectural elements; (3) with greater compression, cytoskeletal elements were dislocated and expressed; and (4) internodes were lengthened and shear injury of myelin occurred. At the edges of the cuff, additional changes occurred: (1) increase of endoneurial area from expression of endoneurial fluid from under the cuff (Fig. 32–81); (2) increase of axonal fluid causing distention of axons, especially at nodes of Ranvier; (3) widening and dilatation of nodes of Ranvier; and (4) cleaved myelin loops appearing to overlap nodes of Ranvier. These results provide strong evidence that stress force causes these pathologic changes. If cuff compression is maintained for long times, ischemic injury may contribute to the pathologic changes described. We noted further that the histologic alterations resulting from compression are quite different in time course, anatomic distribution, and histologic changes from those caused by ischemia.

Ischemia A schematic view of the blood supply to nerves is shown in Figure 32–82. The human sciatic nerve, at its upper end, receives an arterial branch from the inferior gluteal

Pathologic Alterations of Nerves

FIGURE 32–80 Electron micrographs of transverse section of peroneal nerve of rat showing the compaction of cytoskeleton and organelles of axons that occurred under the cuff when the peroneal nerve was compressed at 300 mm Hg for 2 minutes (top) and for 2 hours (bottom). Inset, Low-power view (⫻4800) of a transverse section of a myelinated fiber showing the rectangular area at a higher magnification (top) (⫻38,300). The ultrastructural preservation of microtubules and of cristae of mitochondria provides evidence that fixation is adequate (bottom). In the transverse section (⫻7700) of the fiber, myelin is split and crumpled and axoplasm is not seen. The pressure that induced these changes and their sites and mechanisms of the changes are described in the text. (From Dyck, P. J., Lais, A. C., Giannini, C., and Engelstad, J. K.: Structural alterations of nerve during cuff compression. Proc. Natl. Acad. Sci. U. S. A. 87:9828, 1990, with permission.)

artery. In angiograms at autopsy, this arterial branch may be visualized for 10 to 15 cm down the sciatic nerves. The distal tibial and peroneal nerves receive their blood supply from popliteal branches. Between these levels, the distal sciatic and proximal tibial and peroneal nerves receive a variable arterial supply from branches of the deep femoral artery. The epineurial arteries branch, and

801

their arteriolar and precapillary branches penetrate the perineurium to perfuse the endoneurium. Arteries and veins have three layers of cells: tunica intima, tunica media, and tunica adventitia. Elastic arteries in nerve are infrequent and of large caliber. Their tunica media is made up of smooth muscle and elastic fibers. Muscular arteries vary from 200 ␮m to several millimeters in diameter. Most, especially the large and intermediate vessels, have an internal elastic lamina. A clear boundary does not distinguish large arterioles from small muscle arteries, although in some older textbooks 100 ␮m was set as the boundary. The smallest arterioles (precapillaries) have only endothelial cells, one or two layers of smooth muscle cells, and pericyte cells. Capillaries have only a few abutting endothelial cells and associated pericytes. Vessels may also be classified by function. The large elastic and muscular arteries might be thought of as conduits (conducting and distributing vessels) from the heart to the tissue. Arterioles provide regulation of blood supply to tissue and resistance to modulate blood pressure (regulating and resistance vessels). Capillaries are exchange vessels. Muscular venules and veins are capacitance and returning vessels. In nerve essentially only the epineurial and perineurial compartments have arterioles and small arteries 50 to 400 ␮m in diameter. That is why the characteristic pathologic abnormalities of necrotizing angiopathy must be sought in the epineurium and why whole-nerve biopsy must be performed to be able to diagnose necrotizing vasculitis of these vessels. Typical changes of small artery and arteriole necrotizing vasculitis, as occur in polyarteritis nodosa, are seen in Figures 32–77 and 32–83 (also see Chapters 106 and 107). The epineurium tends also to have the larger collecting veins. The capillaries of epineurium do not have tight junctions between endothelial cells, and tracers such as horseradish peroxidase pass freely between cells. The arterioles penetrating the perineurium are small, usually less than 50 ␮m. The vessels in the endoneurial space include only an occasional small arteriole (25 to 75 ␮m in diameter). The endoneurial vessels are precapillaries, capillaries, venules, and thin-walled veins. The endothelial cells of capillaries abut against each other and form tight junctions not allowing horseradish peroxidase to pass between them, providing a morphologic basis for the blood-nerve barrier. Three vascular processes typically cause ischemia of nerve. The first is large-vessel occlusive disease or emboli. Ligation of a single large limb vessel does not cause ischemic injury—several (actually many) must be ligated to induce ischemic injury. The second process that may cause nerve ischemia is necrotizing vasculitis. In fully expressed disease, many epineurial arterioles 50 to 400 ␮m in diameter become occluded at multiple sites, which occurs in such disorders as polyarteritis nodosa, rheumatoid vasculitis, Churg-Strauss syndrome, and Wegener’s granulomatosis (see Chapter 106). The third process is microvascular disease leading to nerve ischemia.

802

Pathology of the Peripheral Nervous System

FIGURE 32–81 Three frames of teased fibers showing normal node from above the cuff (first) and obscured node from under the cuff (second) (upper), and lengthened node from the edge of the cuff (lower). (From Dyck, P. J., Lais, A. C., Giannini, C., and Engelstad, J. K.: Structural alterations of nerve during cuff compression. Proc. Natl. Acad. Sci. U. S. A. 87:9828, 1990, with permission.)

There is good autopsy and nerve biopsy evidence that peripheral vascular disease (large artery occlusive disease) may result in ischemic nerve damage, but the nerve damage is in the general region and roughly proportionate with distal limb gangrene. There is little evidence that peripheral vascular disease is the basis of or a major contributor to diffuse polyneuropathy (e.g., old age or diabetic sensorimotor polyneuropathy). Results of experimental ligation of the large arteries to lower limbs76,95,96 confirm and extend these views: (1) very extensive ligation of large arteries are needed to cause ischemic nerve damage to nerves of lower limbs because collateral circulation is extensive and

(2) the lower sciatic and upper tibial and peroneal nerves and central fascicular regions of widely separated fascicles are sites of vulnerability. Study of patients with necrotizing vasculitis and multiple mononeuropathy has provided important information about the spatial distribution of nerve ischemic damage (Fig. 32–84). In necrotizing vasculitis, neurologic deficits may develop acutely first in the territory of one nerve, then in another, and then in others. This pattern of involvement is called “mononeuritis multiplex” or, a preferable term, multiple mononeuropathy. The three-dimensional relationships between occluded arteries and regions of ischemic nerve damage have been studied.34

Pathologic Alterations of Nerves

803

FIGURE 32–82 Diagrammatic representation of the blood supply to nerve as discussed in the text. (From Dyck, P. J.: Peripheral neuropathy. Postgrad. Med. 41:279, 1967, with permission.)

In several patients, nerve sections from serial tissue blocks of median, ulnar, radial, sciatic, tibial, and peroneal nerves were fixed in buffered formaldehyde; then, removing the entire sciatic nerve and its tibial and proximal branches, serial blocks were cut from along its length from proximal to distal. The first block was embedded into paraffin and the sections stained (hematoxylin and eosin, trichrome, and other stains), and the next block was additionally fixed in Flemming’s fixative (Weigert stain) for fiber alterations. These preparations were repeated such that consecutive blocks along the length of nerve were alternatively studied for vessel alterations and for fiber alterations. Segmental occlusions of small epineurial arteries or arterioles were found to be diffusely distributed along the length of the large nerve trunks. By contrast, fiber degeneration began at mid-thigh levels only in central fascicular or wedge-shaped regions of widely distributed fascicles. In affected fascicles subperineurial fibers tended to be unaffected. At more distal levels the discrete regions of degeneration could no longer be identified, but degenerating fibers were scattered diffusely throughout most fascicles. These postmortem studies demonstrated that (1) small artery or arteriole occlusion occurring at multiple sites along the length of nerve, resulting in nerve fiber damage beginning in a patchy distribution at mid-upper arm and mid-thigh levels; (2) individual occlusion of arteries or arterioles could not be related to regions of ischemic damage, so multiple sites of multiple vessels must be occluded before multifocal fiber loss occurs and ischemic injury occurs at watershed zones of poor perfusion; and (3) axons seemed more vulnerable to ischemia than did Schwann cells.

50μ FIGURE 32–83 Oblique section of sural nerve epineurial arteriole showing florid necrotizing vasculitis. Top, The trichrome stain of a paraffin section shows the bright red fibrinoid degeneration, neural necrosis, and inflammation. Bottom, An adjacent paraffin section has been reacted for CD68 (KP-1), revealing macrophages in vessel walls and in the surrounding inflammatory infiltrates. The changes are diagnostic of necrotizing vasculitis of small arteries and arterioles. See Color Plate

These studies describe the pathologic alterations of the large nerve vessel vasculitides as occur in rheumatoid arthritis, polyarteritis nodosum, Wegener’s granulomatosis, and systemic lupus erythematosus. In contrast, there are small blood vessel vasculitides that include nonsystemic vasculitis of nerve, Sjögren’s syndrome, and the radiculoplexus neuropathies. The radiculoplexus neuropathies are asymmetrical disorders causing pain and weakness in the upper (cervical), trunk (thoracic), and lower (lumbosacral) regions. These are monophasic illnesses that happen to diabetic and nondiabetic patients. The pathologic lesion in these conditions is ischemic injury of nerve (perineurial degeneration and thickening, neovascularization, multifocal fiber loss and degeneration, and microfasciculation) (see Fig. 32–79).55,56 This ischemic injury appears to be due to a microvasculitis of small nerve vessels. Because of these

804

Pathology of the Peripheral Nervous System

FIGURE 32–84 Upper row, Low-power tracings showing the density of myelinated fibers and the fascicular pattern of fiber loss from the median nerve of a patient with rheumatoid arthritis and necrotizing angiopathic neuropathy. These were taken at 4 cm (left), 8 cm (middle) and 12 cm (right) from the axilla. The density of intact myelinated fibers in fascicles was graded as follows: A ⫽ normal numbers; B ⫽ equivocally decreased; C ⫽ moderately decreased; and D ⫽ greatly decreased. Lower row, Epon sections of fascicles as seen in the rectangles in the 8-cm tracing. Note the central fascicular loss of myelinated fibers, surrounded by a borderline zone of degenerating fibers, which in turn is surrounded by fibers of normal density. (From Dyck, P. J., Conn, D. L., and Okazaki, H.: Necrotizing angiopathic neuropathy: three-dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972, with permission.)

vessels’ small size, fibrinoid degeneration of the wall is rare. Typically, there is fragmentation and disruption of the vessel wall by mononuclear inflammatory cells that is often very focal (see Fig. 32–78). These studies take on added significance because diabetic third-nerve palsy was found to be due to central fascicular damage, surmised to result from ischemic damage.5,30,158 Which cells in nerve are especially vulnerable to ischemia? Studies by Korthals and colleagues,94 Dyck and co-workers,39 and Nukada and Dyck114 suggested that axons are more vulnerable to ischemia than Schwann cells or myelin. Korthals and colleagues94 found that axons entering an ischemic area became dilated and filled with organelles. They suggested that ischemia resulted in interference with rapid axonal flow, resulting in the accumulation of organelles.

Nukada and Dyck114 studied these changes in more detail (Fig. 32–85). Work by Johnson and co-workers82 and Benstead and colleagues11 found that endothelial cells of capillaries and perineurial cells also were vulnerable to ischemic injury. Can precapillary and capillary disease cause nerve fiber degeneration? The evidence from experimental studies is unequivocal—it can. In human neuropathy the evidence is incomplete and less certain. Precapillary and capillary occlusion can be produced by intra-arterial injection of arachidonic acid122 and by injection of 15-␮m polystyrene microspheres.114 Such studies have demonstrated that central fascicular nerve damage can be produced by capillary occlusion. When these substances are injected into vessels of supply to the sciatic nerve, the vulnerable region is the

Pathologic Alterations of Nerves

805

FIGURE 32–85 Upper, Skip serial electron micrographs of individual myelinated fibers entering an ischemic core of sciatic nerve from an area without pathologic abnormality (proximally) to well-developed ischemic changes distally. Both fibers show the following sequence of events: normal-appearing fibers (1), dark axons with light core (2), enlarged dark axon (3 and 4), attenuated dark axons (5), and missing axon (5 and 6). Lower, The results from serial transverse sections were used to draw a representative fiber entering an ischemic core. On the left, the fiber is normal; in the middle, the axon is enlarged and dark owing to accumulated organelles; and on the right, it has become attenuated and has disappeared. AX ⫽ axon; BM ⫽ basement membrane; MY ⫽ myelin; SC ⫽ Schwann cell. (Modified from Nukada, H., and Dyck, P. J.: Acute ischemia causes axonal stasis, swelling, attenuation, and secondary demyelination. Ann. Neurol. 22:311, 1987.)

central fascicular region of proximal tibial and peroneal nerves. Axons at proximal borders of ischemic areas may be swollen and filled with various organelles.

Intraneural Injection of a Foreign Substance Rarely, the sciatic nerve of humans is directly injected with a medication, sometimes resulting in a severe protracted neuropathic deficit. Injecting penicillin in oil directly into the sciatic nerve may result in a granulomatous reaction. In war, nerves may become lacerated and involved in purulent abscesses. The perineurium provides a barrier to the entry of purulent material into the fascicular contents.

Microinjection of minute amounts of lysolecithin73 and other monochain phospholipids45,102 into a fascicle of nerve causes focal demyelination (Fig. 32–86). Demyelination can be passively transferred by intraneural injection of experimental allergic neuritis serum.132 Passive transfer of demyelination resulting from microinjection of Guillain-Barré serum has been reported,60 but other studies are not convincing.102 The results of these types of studies have to be interpreted with great caution, however, because the technique of injection itself (the size of the needle, the movement of the needle or animal, and the volume and rate of fluid injected) may produce fiber degeneration.45

806

Pathology of the Peripheral Nervous System

FIGURE 32–86 Transverse sections of peroneal nerve of rat to show control (A) and the confluent segmental demyelination from microinjection of 5 nmol/5 ␮L of lysolecithin (B). (From Dyck, P. J., Lais, A. C., Hansen, S. M., et al.: Technique assessment of demyelination from endoneurial injection. Exp. Neurol. 77:359, 1982, with permission.)

Inflammatory Demyelinating Disorders Inflammatory demyelinating polyradiculoneuropathies (IDPs) are usually symmetrical polyradiculoneuropathies caused by immune mechanisms resulting in multifocal demyelination and fiber degeneration, which results in a paretic or paralytic disorder affecting proximal and distal limbs or even bulbar or truncal muscles (especially in acute IDP) and sensation; nerve conduction abnormalities characteristic of segmental demyelination; a cytoalbuminologic dissociation of the cerebrospinal fluid; and mononuclear cell exudates, edema, and segmental demyelination in nerves (Figs. 32–87 through 32–91). The IDP disorders may be subdivided into acute IDP (AIDP; see Chapter 99) and chronic IDP (CIDP; see Chapter 100). Others and we have described other inflammatory demyelinating varieties.6,33 Whether inflammatory demyelinating processes underlie plexus neuropathies, diabetic radiculoplexus neuropathy and multiple mononeuropathy, and prednisone-responsive inherited motor and sensory neuropathy has not been demonstrated, but this remains a possibility. The hallmarks of inflammatory demyelinating disorders are edema, particularly around capillaries and beneath the perineurium; endoneurial perivascular collections of mononuclear cells; macrophage contact with Schwann cells; and macrophage delamination of myelin. With chronicity, onion bulbs are formed. In AIDP, early pathologic reports emphasized the vulnerability of ventral spinal roots,75 but Asbury and co-workers6 found the involvement to be more diffuse and both proximal and distal nerve trunks to be affected. Biopsy of sural nerve may not show

FIGURE 32–87 A cytoplasmic process of a macrophage has inserted itself into the cytoplasm of a Schwann cell. Macrophages make contact with Schwann cells and are though to participate in active demyelination in experimental allergic neuritis. (From Wisniewski, H., Prineas, J., and Raine, C. S.: An ultrastructural study of experimental demyelination and remyelination. I. Acute experimental allergic encephalomyelitis in the peripheral nervous system. Lab. Invest. 21:105, 1969, with permission.)

50μ

25μ

50μ

50μ FIGURE 32–88 Serial transverse paraffin sections from a fascicular sciatic nerve biopsy in a patient with focal chronic inflammatory demyelinating polyradiculoneuropathy. The patient presented with a sciatic mononeuropathy, and on magnetic resonance imaging the sciatic nerve was enlarged and had increased T2 signal. Top, A transverse paraffin section of the biopsied nerve (hematoxylin and eosin staining) shows large epineurial perivascular inflammatory cell infiltrates. In addition, there is a small perivascular endoneurial inflammatory cell collection (arrowhead). Middle, Section immunoreacted for CD45 (leukocyte common antigen) that confirms that the mononuclear cell infiltrates in the epineurium and endoneurium (arrowhead) are lymphocytes. Bottom, Section immunoreacted with CD68 (KP-1) shows that macrophages are present but less common than lymphocytes. Teased fibers from this biopsy showed increased rates of segmental demyelination. These findings suggest that focal demyelinating inflammatory lesions occur and that some mononeuropathies are chronic inflammatory demyelinating mononeuropathies. See Color Plate

FIGURE 32–89 Paraffin and epoxy transverse sections from the greater auricular nerve from a patient with multifocal hypertrophic chronic inflammatory demyelinating polyradiculoneuropathy. She had apparent “tumors” within the brachial plexus and greater auricular nerves and had previously been incorrectly diagnosed as having neurofibromatosis. Top, Transverse paraffin section (hematoxylin and eosin staining) showing an epineurial perivascular mononuclear inflammatory cell infiltrate (right) and diffuse onion bulbs (left). Middle, Dense onion bulbs affecting all fibers with only thin myelin remaining (epoxy, methylene blue stain). Bottom, Large subperineurial mononuclear inflammatory cell infiltrate adjacent to frequent onion bulbs (epoxy, methylene blue stain). See Color Plate

prominent inflammatory change even when there is considerable evidence of segmental demyelination. On the assumption that sensitized hematogenous mononuclear cells invade nerve, Asbury and colleagues6 proposed that these initially caused segmental demyelination and, if more severe, axonal degeneration. Whatever the mechanism, axonal degeneration of distal nerves, such as the radial cutaneous nerve of forearm and sural nerve, may be severe in fulminant cases of AIDP. There is increasing evidence that humoral mechanisms are involved and perhaps are primary and initiate the T-cell

808

Pathology of the Peripheral Nervous System

30μ

FIGURE 32–90 Teased nerve fibers from the sural nerve of a patient newly diagnosed with chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). These are separately teased myelinated fibers laid side by side demonstrating multiple areas of segmental demyelination typical of active CIDP.

or macrophage response. The efficacy of plasma exchange in AIDP and CIDP is generally attributed to removal of antibodies, but other mechanisms may be involved.

30μ

Multifocal CIDP An inflammatory demyelinating multiple mononeuropathy has been described by Lewis and Summer101 (see Fig. 32–89). Dyck and Dyck have also identified mononeuropathies (often sciatic) that are due to inflammatory demyelinating lesions (see Fig. 32–88).54

Chronic Immune Sensory Polyradiculopathy Sinnreich and co-workers142 have described a sensory syndrome with ataxia with raised cerebrospinal fluid protein, normal sensory action potentials and abnormalities of somatosensory nerve potential, abnormality of dorsal spinal root, and MRI enlargement and enhancement. Figures 32–92 and 32–93 show the typical MRI and histologic features.

Multifocal Motor Neuropathy Multifocal motor neuropathy is discussed elsewhere in the text (see Chapter 102).

Leprosy The pathology of leprosy is discussed elsewhere in the text (see Chapter 91).

Sarcoidosis Sarcoidosis (see Chapter 108) (see Figs. 32–1 and 32–2), a noncaseating granuloma of unknown cause, involves cranial

6μ

FIGURE 32–91 Paraffin and epoxy transverse sections of a nerve root from a patient with a long history of relapsing/remitting chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) that initially responded to immunosuppression but later was less responsive, and he became wheelchair dependent. Magnetic resonance imaging of his lumbosacral spine showed markedly enlarged nerve roots, one of which was biopsied to exclude other causes of polyradiculopathy. Top, A paraffin section immunoreacted with CD68 (KP-1). The picture is taken at high contrast to demonstrate the frequent background onion bulbs and frequent macrophages, which are associated with them. Middle, An epoxy transverse section stained with methylene blue that shows frequent large onion bulbs surrounding fibers without myelin or with thin myelin interspersed with myelinated fibers without onion bulbs and normal myelin thickness. This “mixed pattern” of onion bulb formation is typical of acquired demyelinating neuropathies (e.g., CIDP). Bottom, An electron micrograph showing three large onion bulbs. The onion bulb in the middle has a thinly myelinated profile in its center. The onion bulb on the right is denervated (no longer has an axon) and has collagen at its center. This patient was treated with aggressive immunotherapy with a combination of intravenous immunoglobulin and plasma exchange, and he had marked improvement. See Color Plate

Pathologic Alterations of Nerves

809

FIGURE 32–92 Transverse T2-weighted (left) and sagittal postgadolinium (right) magnetic resonance images of the cauda equina from a patient with chronic immune sensory polyradiculopathy. The nerve roots are enlarged and show enhancement (arrows). One of these rootlets was later biopsied (see Fig. 32–93).

nerves, meninges, and roots and infrequently causes a polyneuropathy. In transverse sections of the nerve, the granuloma may form crescentic lesions involving the perineurium and extend into the epineurium and endoneurium. The lesions tend to be focal, affecting short lengths of certain fascicles. At these sites severe damage of certain fascicles develops by unknown mechanisms. Four possible mechanisms for fiber damage might be considered: (1) cellular proliferation causing direct compression and distortion of nerve fibers; (2) cellular proliferation leading to vascular occlusion and ischemic damage; (3) release of cellular toxins, with cytokines damaging fibers; and (4) other immunologic mechanisms. The sarcoid nodules contain epithelioid cells, giant cells, lymphocytes, plasma cells, and macrophages, and caseation does not occur. The epithelioid cells may be arranged in rows or clusters surrounded by lymphocytes, plasma cells, and macrophages. Typically they form bands of inflammation within, or just outside of, the perineurium and extend into the epineurium or endoneurium. Giant cells are not always evident. Usually there is considerable scarring associated with the inflammatory process. Lepromatous leprosy must be distinguished from sarcoidosis by a different histologic appearance and by the finding of the leprous bacilli in globoids using a petroleumprotected acid-fast stain (Fite stain) (see Chapter 91).

Amyloidosis The subject of amyloidosis is discussed in Chapter 83. Amyloid is polymerized in interstitial fluid, particularly in connective tissue sheaths. In nerve it is found in epineurium,

perineurium, and endoneurium. It may be seen as linear streaks along connective tissue sheaths, as cuffs around blood vessels, or as nodules. In sections it has a ground-glass, feltlike, acellular appearance (Figs. 32–94 through 32–96). Several stains may be used to reveal it. In sections stained with hematoxylin and eosin, it is pink; with methyl violet it stains bright pink; with alkaline Congo red stains and under a polarizing filter, it is apple green and is birefringent; and with thioflavin T, it fluoresces (see Fig. 32–94). Based on the variety of precursor proteins that can be identified immunohistochemically, at least 10 varieties of amyloidosis occur. Four varieties affect nerve: light chain (LC) causes primary amyloidosis, mutations of transthyretin (TTR) cause familial amyloid polyneuropathy, mutations of gelsolin cause Icelandic amyloidosis, and apolipoprotein A-I can also cause neuropathy. Typically, in TTR amyloidosis, monoclonal LCs are not found in the serum. Many allelic varieties of TTR amyloidosis have now been described encompassing varieties previously called PortugueseJapanese, Appalachian, German-Swiss, and other varieties. In each of the alleles, a simple amino acid substitution of the DNA that codes for TTR is responsible. Increasingly, these known mutations can be detected rapidly with molecular genetic techniques. Primary amyloidosis (LC amyloidosis) is usually associated with monoclonal proteins in the serum. Immunohistochemical characterization of amyloid in nerve now permits a laboratory diagnosis of inherited versus primary amyloidosis. The mechanism of nerve fiber injury in amyloidosis remains unsettled. Earlier authors invoked ischemia.89 The frequent occurrence of carpal tunnel syndrome suggests that entrapment plays a role in some patients. In addition,

810

Pathology of the Peripheral Nervous System

nerve fibers may be compressed by an amyloid nodule immediately adjacent to the nerve46 (see Fig. 32–94). In a patient with dominantly inherited amyloidosis, amyloid nodules had indented MFs, causing marked dilatation immediately above and below the nodule.46 At other sites, absent myelin or remyelinated segments were seen adjacent to such nodular indentations—suggestive evidence that nodules can result in demyelination. It is evident that compression is implicated in the pathologic alterations of fibers; what is less clear is how much of an impact this has on the symptomatology of amyloid neuropathy. Ischemia may play a role in nerve damage, because both epineurial and endoneurial vessels may be heavily surrounded and infiltrated with amyloid even to the point of occlusion. Central fascicular fiber loss, as found in ischemic damage, has not, however, been described. Deposition of amyloid in spinal and autonomic ganglia may cause degeneration of primary afferent neurons by unknown mechanisms. Small-diameter neurons may be preferentially affected (Figs. 32–97 and 32–98).

Tumor

10μ

2μ

FIGURE 32–93 A paraffin preparation reacted to a macrophage antigen (CD68) (top), a methylene blue–stained epoxy transverse section (middle), and an electron micrograph (bottom) from a patient with chronic immune sensory polyradiculopathy. Top, Scattered macrophages associated with onion bulbs and myelinated fibers (arrows). Middle, A normal density of myelinated fibers but a altered size distribution with fewer large myelinated fibers and more small myelinated fibers than normal. The myelin of many fibers is abnormally thin, frequent onion bulbs are present, and some fibers are demyelinated (arrowheads). Bottom, Frequent onion bulb formations associated with thinly myelinated profiles. These findings provide evidence of chronic demyelination and abortive repair, and the loss of large myelinated fibers explains the ataxia. See Color Plate

This subject is reviewed in Chapter 116. Peripheral nerves tend to be spared unless they are compressed by tumor beneath tight fascial sheaths, by infiltrated meninges, or by narrowed foramina at points of exit of cranial and spinal nerves. The lymphoproliferative and plasma proliferative tumors tend to cause radicular and retroperitoneal involvement of nerves. The factors that lead to fiber degeneration may be related to direct compression of nerves. Ischemia seems likely to be a factor. Monoclonal proteins may be produced in lymphoproliferative and plasma proliferative disorders, in amyloidosis, and in conditions unrelated to tumor (monoclonal gammopathy of unknown significance). Might these monoclonal proteins have a deleterious effect on components of Schwann cells (such as myelin) or on axons? There is evidence that monoclonal proteins and LCs are found in the nerve microenvironment, but their effect there is still speculative. The passive transfer of an abnormality of nerve conduction to mice using monoclonal protein serum is suggestive of a direct deleterious effect.12 In some cases plasmapheresis appears to have a beneficial effect on the course of neuropathy, possibly providing further evidence that these proteins are involved in the neuropathic effect. The mechanism of nerve fiber damage in such primary tumors as neurofibromatosis is not well understood, but compression probably is a factor (see Chapter 116).

PERINEURIAL PATHOLOGY The important barrier functions of perineurium have been discussed in Chapters 29 and 30.

Pathologic Alterations of Nerves

811

FIGURE 32–94 A–F, Selected teased fibers from the sural nerve of a woman with the dominantly inherited Andrade type of amyloidosis causing early bladder and bowel incontinence, postural hypotension, and a syringomyelia-like dissociated sensation loss in lumbosacral dermatomes due to a selective severe loss of unmyelinated and small afferent myelinated fibers (MFs). The position of amyloid nodules (arrows) adjacent to pathologic MF alterations is shown in selected frames. C, The nodule is shown to have markedly indented an MF axon with dilatation of the fiber above and below. E and F, Remyelinated segments are found adjacent to several nodules. G, In the sural nerve of a patient with primary amyloidosis, the marked thickening of a vessel wall is demonstrated. H, The endoneurial position of a large nodule in a paraffin section. (Modified from Dyck, P. J., and Lambert, E. H.: Dissociated sensation in amyloidosis: compound action potential, quantitative histologic and teased fiber, and electron microscopic studies of sural nerve biopsies. Arch. Neurol. 20:490, 1969. Copyright 1969, American Medical Association.)

As mentioned in the earlier sections on interstitial pathologic alterations, the perineurium may be infiltrated by inflammatory cells, as in perineuritis, and may be involved in sarcoidosis, in necrotizing angiopathy, and in tumor. In severe ischemic damage, it may become necrotic. Also in ischemic injuries it will become thickened and, when its integrity is interrupted, abortive regeneration (microfasciculation) can occur within or beyond its original boundary (see Fig. 32–79). As described later, the perineurium may contain lipid debris in severe lead neuropathy. Whether this represents disturbed metabolic events of the perineurial cells themselves or a degradative or transport function of debris across

the perineurium is not known. Because similar lipid breakdown products are found in macrophages lined up along the inside of the perineurium, it may be that macrophages unload debris to the perineurium either for degradation or for transport out of the endoneurial space. The perineurium may be thickened in various inherited neuropathies. In neurofibromatosis (Fig. 32–99), the number of perineurial leaflets may be greatly increased and they may be attenuated and disorganized. The perineurium is markedly abnormal in a curious condition of digit gigantism caused by macrodystrophia lipomatosa or macrodystrophia hamartoma. These nerve tumors are associated with overgrowth of

812

Pathology of the Peripheral Nervous System

FIGURE 32–96 The typical crisscrossing, feltlike pattern of fibril in amyloidosis.

fiber alterations. That this separation is somewhat arbitrary may be illustrated by the example of lead neuropathy. In lead neuropathy, widespread segmental demyelination without inflammation is typical. Study of the experimental model has shown that endoneurial fluid is under increased pressure and is rich in lead, which exposes Schwann cells to lead, which results in segmental demyelination. Thus changes in microvessels and endoneurial fluid are implicated in pathogenesis.

Neuronal (Axonal) Pathologic Alterations FIGURE 32–95 Photomicrographs of amyloid from sural nerve of a patient with inherited amyloidosis. A, Congophilia of amyloid. B, Transthyretin positivity using immunohistochemistry. C and D, Lack of kappa (left) and lambda (right) light chain reactivity. The transthyretin reactivity supports the clinical diagnosis of inherited amyloidosis. (Courtesy of K. Ii, M.D.)

part of the hand and fibrous and lipomatous alterations. The digital nerves are made of multiple small fascicles (microfasciculation), each surrounded by an excessive number of normal-appearing perineurial lamellae.151

PARENCHYMATOUS PATHOLOGIC ALTERATIONS The separation of the pathologic alterations of nerves into interstitial and parenchymatous is somewhat arbitrary. In parenchymatous alterations, there is no obvious circulatory, inflammatory, infiltrative, or interstitial cause for the

For most neuropathies with neuronal or axonal degenerations, the three-dimensional structural features on light and electron microscopy are known only incompletely. Generally the available information is descriptive only. The causes and mechanisms of fiber degeneration are poorly understood. For these reasons the classification provided is a descriptive one. The following characteristics are taken into account: 1. 2. 3. 4. 5. 6. 7. 8. 9.

The segmental levels of neurons affected The size or functional classes of neurons affected The level(s) within neurons affected The temporal order and profile of classes of neurons affected The pathologic alterations of the soma and axon The pathologic alterations of the Schwann cells and myelin Pathologic reactions to axon or Schwann cell alterations The primary sites of neuronal vulnerability Specific mechanisms

Selectivity by Segmental Level Different segmental levels may be affected in peripheral neuron degeneration. To illustrate, in adult-onset Tangier

Pathologic Alterations of Nerves

813

FIGURE 32–97 A, Cluster of unmyelinated fibers (UFs) from healthy human sural nerve to emphasize species, age, and fiber-to-fiber variability. Characteristically, one to two UFs lie in separate Schwann cell cytoplasmic processes. Mesaxons may be short or long. Schwann cell cytoplasmic (SCC) stacks and pseudopod-like structures are not uncommon and occur more frequently with age. B and C, The fiber variability of mitochondria (M) and endoplasmic reticulum (ER) size and the density of both microtubules (MT, NT) and neurofilaments (NF/F) are illustrated. (From Dyck, P. J., and Lambert, E. H.: Dissociated sensation in amyloidosis: compound action potential, quantitative histologic and teased fiber, and electron microscopic studies of sural nerve biopsies. Arch. Neurol. 20:490, 1969, with permission. Copyright 1969, American Medical Association.)

disease, a genetic disorder with a deficiency of high-density lipoprotein apolipoprotein and tissue deposition of cholesteryl esters, cranial, cervical, and brachial neurons are severely affected at a time when lumbosacral neurons are only mildly or not affected.36,72,92,98 This distribution, however, is not found in the childhood form of the disorder even though it involves an identical molecular abnormality. The reasons for this are not known. For many parenchymatous peripheral neuropathies, the brunt falls on lumbosacral neurons, sparing those of higher segmental levels. In still others, most segmental levels are affected.

Selectivity of Functional Classes of Neurons Different functional classes of neurons may be involved in neuromuscular diseases. This selectivity is sometimes quite striking in inherited neuropathy and may be the basis of classification. Selective corticospinal or corticobulbar involvement is typical of spastic paraplegia, motor neuron involvement is characteristic of progressive muscular atrophy (hereditary motor neuropathy); sensory and autonomic neurons are involved in spinocerebellar degeneration and hereditary sensory and autonomic neuropathies (HSANs); and motor, sensory, and autonomic neurons are involved in HMSNs. Increasingly,

814

Pathology of the Peripheral Nervous System

FIGURE 32–98 Electron micrograph of a denervated myelinated fiber cluster from the sural nerve. No unmyelinated fibers are seen; only Schwann cell cytoplasm (SCC) and collagen fibrils (CF) of the epineurium are present. This figure should be compared with Figure 32–97, an unmyelinated fiber cluster from healthy sural nerve. (From Dyck, P. J., and Lambert, E. H.: Dissociated sensation in amyloidosis: compound action potential, quantitative histologic and teased fiber, and electron microscopic studies of sural nerve biopsies. Arch. Neurol. 20:490, 1969, with permission. Copyright 1969, American Medical Association.)

each of these major groups of neural disorders is divided into defined molecular genetic disorders. Thus there are now at least 10 unique disorders that might be classified as spinocerebellar degeneration. Similarly, HMSN (Charcot-Marie-Tooth disease) is known to be associated with perhaps 10 gene mutations. Selectivity by Levels within Neurons (Axons) Different levels within neurons may be selectively vulnerable. Krucke,97 in a historic account of the neuropathology of peripheral nerve disorders, emphasized the importance of selective vulnerability of different neuronal levels for pathologic change. Sometimes the major brunt of the pathologic alteration is on the soma (neuronopathies), in the root (radiculopathies), or in different proximal-to-distal levels of the axon. The postulated pharmacologic sites of action of toxins within neurons are discussed in Chapters 18 and 113. In some disorders the sites of metabolic action and initial structural change are localized to the perikaryon. Examples of this are certain viral infections, lysosomal storage diseases, and toxins. There may be examples of selective involvement of the centrally directed axons of peripheral sensory neurons. In other intoxications, different levels (from proximal to distal) of the peripheral axon may be affected.20–23,68,69,138 In murine dystrophy a focal lesion

involving a developmental abnormality of myelination appears to be confined to the roots.17 In other disorders the distal aspects of central and peripheral axons undergo pathologic changes.125,126,145 In some diseases abnormal metabolic products may accumulate in the soma (e.g., glycosphingolipid in Fabry’s disease; see Fig. 32–56); in other diseases, they may accumulate in the axon (e.g., glycogen in myxedema neuropathy47 and experimental diabetes mellitus124); and in still other diseases, they may accumulate in Schwann cell cytoplasm (e.g., metachromatic granules in metachromatic leukodystrophy [MLD]; Fig. 32–100). Neurofilaments with mitochondria dense bodies and other vesicular profiles may accumulate in spherical swellings (spheroids) in axons. This is a prominent feature in proximal axons of lesions caused by IDPN22–25,68,138; in distal aspects of both central and peripheral axons; in lesions caused by acrylamide125; in neuropathies produced by such hexacarbons as n-hexane, methyl-n-butyl ketone, and 2,5-hexanedione145,146; and in distal nerve of some patients with inherited neuropathy.75 Temporal Selectivity of Neurons (Axons) The temporal sequence of involvement of neurons may vary in different disorders. In Friedreich’s ataxia only a minority of large primary afferent neurons appear to be

Pathologic Alterations of Nerves

FIGURE 32–99 Semithin transverse sections stained with methylene blue at various locations of cutaneous neurofibroma.

FIGURE 32–100 Transverse semithin sections of sural nerve of a child with metachromatic leukodystrophy (MLD). The arrows indicate regions in which there is a collection of MLD granules, which stain metachromatically brown with acid cresyl violet. (From Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968, with permission.)

815

816

Pathology of the Peripheral Nervous System

undergoing degeneration at any one time. Large-diameter axons undergo axonal atrophy, myelin remodeling, and eventually degeneration. In other disorders, such as HMSN type I (a dominantly inherited hypertrophic neuropathy), most of the axons appear to be abnormal, and this abnormality is now known to be secondary to a Schwann cell abnormality. Anterograde and Retrograde Effects The morphologic differences that have been illustrated in the preceding paragraphs imply an interplay of different pathophysiologic mechanisms. It is clear that the neurons affected, the site of vulnerability within neurons, and the tissue reaction for most human neuropathies are poorly understood. The problem is made more difficult by the fact that derangement of the function of the soma may lead to dysfunction and degeneration of central and peripheral axons, and vice versa. The permanent axotomy model appears to show that decreased metabolic activity of the soma can lead to axonal atrophy, myelin remodeling, and axonal degeneration most severe distally and less severe centripetally. It is not known whether perikaryal metabolic failure in disease may selectively cause axonal atrophy, myelin remodeling, and degeneration of distal peripheral axons, of central axons, or of both central and peripheral axons, although the permanent axotomy model implies that this may be the case. It is also difficult to establish sites of vulnerability because axonal damage causes retrograde effects in the soma, which then have to be distinguished from possible primary effects.

Classification of Neuronal (Axonal) Pathologic Alterations The diversity of pathologic changes, the overlap of primary and secondary events and effects, the incompleteness of knowledge of the three-dimensional pathologic and morphometric changes with disease, and the lack of knowledge regarding the underlying molecular mechanisms make classification difficult. The tentative organization provided here emphasizes descriptive differences: 1. Neuronal agenesis or maldevelopment 2. Acute neuronal degeneration 3. Chronic neuronal degeneration ⫾ axonal spheroids ⫾ axonal atrophy, myelin remodeling, and axonal degeneration 4. Wallerian degeneration 5. Acute axonal degeneration 6. Axonal spheroids ⫾ axonal atrophy ⫾ myelin remodeling ⫾ axonal degeneration 7. Axonal atrophy ⫾ myelin remodeling ⫾ axonal degeneration

Agenesis and Maldevelopment Congenital absence or failure of development in utero may account for various cranial, motor, and sensory and autonomic neuropathies. A unilateral third cranial nerve involvement is reported. Because some patients with this disorder have a jaw-winking syndrome (the Marcus Gunn phenomenon, in which the ptotic eyelid elevates with jaw opening), we assume that faulty reinnervation after nerve damage, conceivably in utero, may be involved. Other cranial nerves are reported to be congenitally absent. Congenital absence of one or both third cranial nerves is called Möbius’ syndrome. Failure of development may underlie some cases of congenital progressive muscular atrophies. Failure of development may be the basis of inherited dysautonomia (HSAN type III) and other types of hereditary sensory and autonomic neuropathies (HSAN types II, IV, and V) (see Chapter 78). In HSAN types II through V, several observations suggest that a congenital failure of small neurons to develop is responsible: (1) the deficit is already present at birth, (2) the deficit appears to remain static for many years, (3) specific size classes of neurons are absent, and (4) obvious degeneration of fibers is not apparent. An abnormality of the neural crest involving neuron multiplication, differentiation, migration, or other events may be involved. Because remaining fibers do not have a normal appearance, one needs to consider an ongoing dystrophic process. Congenital maldevelopment may be involved in such disorders as inherited neurofibromatosis (see Fig. 32–99) and multiple endocrine neoplasia type 2B (Fig. 32–101). Faulty development may be involved in redundant lumbosacral roots, gigantism of digits and other structures, and certain neural tumors. Acute Neuronal Degeneration Primary pathologic derangement of the soma may lead to neuronal degeneration of lower motor, primary sensory, and autonomic neurons. In certain viral and lysosomal storage diseases, the soma appears to be the site of primary pathologic abnormality. The axon undergoes acute axonal degeneration having the morphologic features of wallerian degeneration. The two prototypes of neuronal degeneration resulting from virus infection of the soma and axonal degeneration are poliomyelitis (in motor neurons) and herpes zoster (in sensory neurons). The poliomyelitis virus probably enters the body through oral intake. It gains entrance into the neuron through the action of surface receptors. Why the disorder tends to be asymmetrical and remains limited to certain segmental levels is poorly understood.132,133 Postmortem examination in acute poliomyelitis reveals congestion and

Pathologic Alterations of Nerves

817

FIGURE 32–101 Postmortem findings in a typical case of multiple endocrine neoplasia, type 2B. A, Drawing of the cervical cord showing aberrant hyperplastic neural tissue (containing Schwann cells and myelinated axons) posterior to the cord and sending pegs into posterior columns. B, Photomicrograph of posterior column (bottom) and of aberrant neural tissue (top). C, A semithin section of the plaque showing its Schwann cell myelinated axon content. (From Dyck, P. J., Carney, J. A., Sizemore, G. W., et al.: Multiple endocrine neoplasia, type 2: phenotype recognition, neurological features and their pathological basis. Ann. Neurol. 6:302, 1979, with permission.)

infiltration of the meninges, ventral gray horn, and ventral aspect of the posterior horn with lymphocytes, polymorphonuclear cells, and macrophages. In chronic lesions ventral gray horns and ventral roots may be atrophied because of loss of somas. Because the poliomyelitis virus does not produce exotoxins or endotoxins, its cytolytic effect must be through other mechanisms. Replication of virus to the point where virions exceed the neuron’s ability to meet metabolic requirements is said to be unlikely. Therefore, specific synthesized proteins, liberation of autolysosomes, and direct toxicity of viral proteins have

been postulated. Whatever the mechanism, the effect is to cause destruction of the cell body with acute axonal degeneration. These matters are discussed in more detail elsewhere.93 Herpes zoster injury of sensory neurons also appears to be due to primary damage of the soma. The virus for varicella and herpes zoster is the same. Whereas varicella occurs in children, herpes zoster generally occurs in the old. In exposed susceptible children, a viremia leads to the typical skin lesions of varicella. The virus replicates in skin and is then transported centripetally to sensory cranial or

818

Pathology of the Peripheral Nervous System

spinal ganglia, where it may lie dormant for years without cytopathologic effect. In old age, possibly because of decreasing immunity, viral replication in sensory ganglia may be activated. Other activating factors may include an associated lymphoma or other malignancy, irradiation, trauma, or entrapment. Because pain in the dermatome precedes cutaneous vesiculation, it is thought that the virus begins to replicate in the sensory ganglia before it is transported to the skin. Replication in epidermal cells is associated with vesicle development. In herpes zoster, the gasserian ganglia are most frequently affected, followed by thoracic spinal ganglia and then other segmental ganglia. In a low percentage of cases, motor neurons or a generalized encephalomyelitis, or both, may develop. The center of the ganglion in an acute case may be necrotic and hemorrhagic (possibly as a result of ischemia) and the margins may be heavily infiltrated by lymphocytes and to a lesser degree by polymorphonuclear cells and plasma cells. As in poliomyelitis, neuron cell bodies undergo cytolysis with acute axonal degeneration. Certain lysosomal storage diseases appear to cause acute neuronal degeneration by interference with metabolic events at the level of the soma. Such lysosomal storage diseases as Fabry’s disease are of special interest because the glycosphingolipid is preferentially stored in the soma of spinal ganglion neurons, implying that this is the primary site of neuronal damage. In other storage diseases, other cells may be preferentially affected, for example, the Schwann cell in MLD. The lysosome, a cytoplasmic organelle containing various hydrolases, proteases, and other degradative enzymes, plays an important role in removing insoluble cell debris during cell metabolism. In lysosomal storage disease there is generally a severe reduction of one lysosomal enzyme resulting from a structural gene mutation, so that removal mechanisms are interfered with, leading to accumulation of specific precursors that cannot be degraded. In lysosomal storage diseases affecting sphingolipids, multimembranous bodies accumulate in the cytoplasm of neuron cell bodies to the extent that the cytoplasm seems filled with the abnormal material. Fabry’s disease, resulting from a gene mutation on the X chromosome in males, presents with lancinating and other types of pain; angiokeratoma, especially of the lower abdomen and flanks; and premature vascular occlusion and renal disease. That the pain might be due to a peripheral neuropathy was suggested by the finding of pathologic abnormalities in biopsy specimens of sural nerve.14,93 Deposition of lipid granules in spinal ganglion perikarya in this condition was described by Steward and Hitchock148 (see Fig. 32–56). Considering peripheral neurons, a pathologic and morphometric study of two cases at death confirmed that the brunt of the process fell on small-diameter spinal ganglia neurons; this study also provided insight into the sites of primary storage within neurons.118 The main locus of pathologic abnormality was

the soma, which was lipid laden to the point that the endoplasmic reticulum appeared to be decreased in amount. The typical lysosomal deposits also occurred in perineurium and endothelial cells but not in axons or Schwann cells. Ultrastructural abnormalities of axons—accumulation of NFs, dense bodies, and mitochondria—were occasionally seen. Considering the observed three-dimensional alteration of primary afferent neurons, several conclusions may be drawn: (1) considerable deposition of glycolipid occurs in the soma of primary sensory neurons; (2) there is selective vulnerability of neurons whose axons are small; (3) the entire neuron degenerates; (4) because glycolipid deposition is abundant in the soma and not in the axon and because axon death and cell body death occur at approximately the same time, it is inferred that the primary site of neuronal damage is the perikaryon; and (5) whether axonal atrophy with myelin remodeling precedes fiber degeneration is not known. Chronic Neuronal Degeneration ⫾ Axonal Spheroids ⫾ Axonal Atrophy, Myelin Remodeling, and Axonal Degeneration Descriptively, this pathologic process is characterized by (1) loss of neuronal somas, (2) presence of axonal spheroids, (3) axonal atrophy, (4) myelin remodeling, and (5) axonal degeneration. In some disorders, the proximally directed axon of the primary sensory neuron may be most affected; in others, the distally directed axon is most affected; and in still others, both centrally and distally directed axons are affected. An adequate three-dimensional description of the pathologic alterations, especially by electron microscopy, is not available for any neuropathy. Whether the primary abnormality is in the soma, in proximal axons, or in both soma and axon is not known. The pathologic alteration in motor neuron disease (amyotrophic lateral sclerosis) is a prototype of this type of neuronal degeneration. The nature of the neuronal degeneration in motor neuron disease is still poorly understood, possibly because spinal cord tissue from patients with early disease seldom becomes available for study, fixation of tissue (both disease and control) obtained at postmortem examination is generally poor, and it is not possible to identify motor axons in mixed nerves. The neuronal populations affected in motor neuron disease have been studied. Pyramidal neurons of the cortex, motor nuclei of the brainstem, and motor neuron columns of the ventral gray horns of the spinal cord are especially vulnerable. Helpful clinical information for differential diagnosis is the fact that the involvement tends to be asymmetrical. The suggestion has been made that the process begins to appear at one segmental level and then spreads outward from it. Although motor neurons are selectively vulnerable, there is evidence from detection threshold

Pathologic Alterations of Nerves

measurement of sensation, morphometric and teased fiber studies of sural nerve, and morphometric study of spinal ganglia that large primary afferent neurons are also affected but to a lesser degree. How does the neuron degenerate? Using diameter histograms of neuron cell bodies and of ventral root myelinated axons, a selective decrease of the large-diameter classes can be demonstrated. The three-dimensional morphologic alterations during degeneration are only incompletely known. Does the neuron degenerate acutely? Does such degeneration begin with distal axonal degeneration or with axonal atrophy? Axonal spheroids, containing NFs and other organelles, are found in proximal axons. Are these spheroids early primary events or can they in some way be related to neuronal failure? For sural nerve the predominant pathologic abnormality is acute axonal degeneration with degeneration of the fiber in linear rows of myelin ovoids and balls. The reported morphologic alterations of motor neuron cell bodies are atrophy, chromatolysis, and eosinophilic cytoplasmic inclusion bodies approximately 1 to 2 ␮m in diameter.78 The possible significance of the argentophilic spheroids19,164 in proximal axons is discussed in Chapters 53 and 114. Wallerian Degeneration The experimental features of wallerian degeneration were described in an earlier section of this chapter. In humans, acute transection, compression, laceration, and stretch of nerve may cause wallerian degeneration. Ischemia or compression by accumulated material in nerve may also cause wallerian degeneration. Acute Axonal Degeneration In many neuropathies distal fibers undergo the changes typical of wallerian degeneration; examples are arsenic and thallium intoxication, diabetes, uremia, alcohol, malnutrition, and vitamin deficiency. Because the mechanism is not understood and acute transection of fibers by a disease process has not been demonstrated, it may be preferable to refer to this process as acute axonal degeneration. Axonal Spheroids ⫾ Distal Axonal Atrophy ⫾ Myelin Remodeling ⫾ Axonal Degeneration IDPN neuropathy is the model best exemplifying this reaction. With IDPN intoxication, axonal spheroids form, causing focal enlargements of axons. The focal enlargements contain closely packed NFs. Altered transport of NF components is thought to be involved in the local accumulation. Sometimes an enlarged spheroid is surrounded by a thinned layer of myelin or the axon becomes demyelinated. The mechanism by which myelin becomes thin is not fully understood. Slippage of myelin lamellae on one another seems unlikely.

819

Frank cleavage and retraction of some of the myelin is another possibility. A third possibility is degeneration and incomplete regeneration of myelin. Distal to the spheroids the axon is atrophied and demyelinated and remyelinated. If the process is severe, fiber degeneration may occur. This type of pathologic alteration may be found in certain intoxications, such as IDPN, n-hexane, and acrylamide; inherited neuropathies (giant axonal neuropathy, a variant of HMSN type II, infantile neuroaxonal dystrophy, and other neuropathies); and possibly acquired neuropathy. Axonal Atrophy ⫾ Myelin Remodeling ⫾ Axonal Degeneration This fiber alteration is like that just described, but spheroids are not encountered. Whether this is due to incomplete studies or true absence of spheroids has not yet been determined. Uremic neuropathy (see Chapter 87) and Friedreich’s ataxia appear to have the fiber changes described here.

Schwann Cell Pathologic Alterations Developmental Abnormality Failure to develop myelin is found in muscular dystrophy of mouse. Axons may remain amyelinated. There appears to be a failure of Schwann cell segregation, ensheathment, and myelination of axons. Basal lamina formation is thought to be abnormal.18 Schwann cells of Trembler mice (in which there is hypomyelination of axons) express the hypomyelinated state when transplanted into nerves of immunologically suppressed normal mice.1 Normal Schwann cells transplanted into Trembler mice express the normal myelinated state. This provides clear evidence that the major abnormality of myelination is related to the Schwann cell. Developmental Abnormality in HMSN III The nerves of patients with Déjérine-Sottas disease (HMSN type III; see Chapter 69) have many of the histologic features seen in Trembler mice. The MFs have regions with excessively thin myelin and other regions without myelin. The findings suggest a hypomyelinated state possibly caused by a metabolic abnormality of Schwann cells. Xenograft studies have been difficult to interpret, but there may be evidence favoring faulty myelination. Storage Diseases In certain lysosomal storage diseases abnormal product, not degraded because of an enzyme deficiency, accumulates in the Schwann cell. Two examples are MLD (Fig. 32–102) and Krabbe’s disease (see Chapter 79). In MLD, metachromatic granules have a typical ladder-and-rung appearance in

820

Pathology of the Peripheral Nervous System

FIGURE 32–102 Typical ultrastructural features of MLD granules. A, Granules at polar ends of Schwann cells. B, MLD granules and ovoids adjacent to compact myelin. C, High-power view of ladder-and-rung periodicity typical of this organelle. D, Low-power view of MLD granules. E, Segmentally amyelinated or demyelinated region of myelinated fiber of sural nerve. (From Dyck, P. J.: Ultrastructural alterations in myelinated fibers. In Desmedt, J. E. [ed.]: New Developments in Electromyography and Clinical Neurophysiology. Basel, S. Karger, p. 192, 1973, with permission.)

electron micrographs (see Fig. 32–102). In MLD, myelin appears to be thinner than it should be, segmental demyelination is widespread, and there appears to be excessive myelin ovoid formation in paranodal and internodal sites.159 Toxins The role of diphtheria toxin in producing a primary demyelinative neuropathy is discussed in Chapter 95. Lysolecithin and various other monochain phospholipids are potent demyelinating agents when injected into the endoneurial space; these compounds have as yet unknown actions on Schwann cells or myelin.73,102

Immune Demyelination Antigalactocerebroside and experimental allergic neuritis serum, when injected endoneurially, produced demyelinative changes. Serum from patients with Guillain-Barré syndrome has also been reported to cause demyelination,60 but other investigators were unable to confirm the observation.45 Lead Intoxication as a Model of Primary Segmental Demyelination A series of observations suggest that lead neuropathy induces primary demyelination63: (1) irrespective of whether a low dose or a high dose of lead is given,

Pathologic Alterations of Nerves

segmental demyelination is induced; (2) ultrastructural abnormalities of axons have not been observed; (3) Schwann cells contain intranuclear inclusions, hydropic degeneration, and myelin degeneration; (4) the numbers of somas of lumbar spinal ganglia are not decreased; (5) the frequency of segmental demyelination and remyelination is not different in proximal and distal nerves; (6) demyelination is randomly distributed among old internodes; and (7) the putative mechanisms described in the next section favor involvement of Schwann cells. Mechanisms of Schwann Cell Damage. Nerve fibers are bathed by abundant endoneurial fluid whose constituents are maintained within narrow limits by blood-nerve, perineurial, and fluid gradient (proximally at the spinal ganglia and distally at target tissue) barriers. The possibility that endoneurial fluid is increased in lead neuropathy was suggested by Lampert and Schochet99 and demonstrated by Ohnishi and Dyck118 (Fig. 32–103). The latter authors found that the transverse fascicular area was greatly increased in lead intoxication, the density of MFs (MF/mm2) was decreased, but the number of MFs per nerve was normal. Later, Windebank and co-workers162 reported that the water content of the nerve was greatly increased. These observations suggested that edema represented a toxic breakdown in the blood-nerve barrier with subsequent entry of water and lead into the endoneurial microenvironment. A second hypothesis was that, with entry of fluid into the endoneurial space, hydrostatic pressure was increased

FIGURE 32–103 Epoxy embedded transverse sections of the peroneal nerve at mid-thigh level from a control animal (left) and from another after 6 months on a diet containing 4% lead carbonate (right). Note the markedly increased endoneurial area of the nerve from the leadfed animal. (⫻112.) (From Ohnishi, A., Schilling, K., Brimijohn, W. S., et al.: Lead neuropathy. 1. Morphometry, nerve conduction, and choline acetyltransferase transport: new finding of endoneurial edema associated with segmental demyelination. J. Neuropathol. Exp. Neurol. 36:499, 1977, with permission.)

821

and might be sufficient to damage fibers. The question of whether hydrostatic pressure was elevated in lead neuropathy was investigated by Low and Dyck103 using polyethylene matrix capsules implanted into the sciatic nerves of rats. The pressure did indeed begin to rise after 3 weeks of lead feeding. It did not reach values considered high enough to collapse capillaries. Windebank and co-workers162 performed studies to test the hypothesis that a toxic breakdown of the bloodnerve barrier might lead to Schwann cell intoxication by lead in the endoneurial fluid. In one series of experiments, they evaluated the lead and water content of nerve and looked at the association of these changes with rate of teased fibers showing demyelination and remyelination. Lead accumulated in the endoneurium very rapidly and reached a maximum between 20 and 35 days. With time the lead content fell. The onset of demyelination occurred at about the time when the endoneurial lead level was at its highest. Edema lagged behind maximal lead levels. Horseradish peroxidase, dextran of various molecular weights, and albumin normally excluded from the endoneurial space have been shown to enter the endoneurial fluid after induction of lead neuropathy. Windebank and Dyck161 examined the kinetics of lead entry into the endoneurium. After various time periods of feeding with 4% lead carbonate, tracer doses of 210Pb were given as single intravenous pulses. The entry of this tracer into the endoneurium was then followed by autoradiography. The tracer entered the endoneurium at a low rate in both control animals and animals that had been on a lead

822

Pathology of the Peripheral Nervous System

diet for 4 weeks. However, there was a 15-fold increase in the rate of entry of the lead tracer after 10 weeks of the lead diet. Thus there is a change in rate of bulk lead entry from the blood to the endoneurium, but this does not occur until 5 to 10 weeks after the onset of chronic lead ingestion. All of these studies produced remarkably consistent results using a wide range of tracers and detection techniques. During chronic lead intoxication, there is a change in the blood-nerve barrier between 6 and 12 weeks, and this change in barrier function parallels the accumulation of endoneurial edema and rising endoneurial fluid pressure. However, the actual accumulation of lead3,77,123,162 starts within 1 week of starting lead feeding and is maximal after about 5 weeks of lead diet; in addition, two studies have shown a fall in the endoneurial or whole-nerve lead content after 6 weeks of lead feeding. Segmental demyelination begins at the time of maximal lead accumulation at 5 weeks162 and before any changes are seen in the bloodnerve barrier. Thus it appears that the breakdown of the blood-nerve barrier is a late or epiphenomenon rather than a change of primary importance in the development of experimental lead neuropathy. It is appropriate to consider the complexities of barrier function when considering heavy metals such as lead. Previously we have considered compartments solely as single units—blood, extracellular space, endoneurium, and others. This is clearly an oversimplification. In the blood, for example, about 95% of lead is bound to red cells. The remainder, in the plasma, is bound to various protein species, and presumably a very small amount exists in some free solute form. In the same way, within the endoneurial compartment it is known that lead exists in a bound intracellular form. Lead-containing intranuclear inclusion bodies (Fig. 32–104) have been described in both capillary endothelial and Schwann cells.112,119 These dense, speculated inclusion bodies are identical to those found in other tissues during lead intoxication.110 In the kidney, these bodies have been shown to have a high lead and protein content,110,137 and it is assumed that they represent an insoluble lead-protein complex in vivo. Thus it is reasonable to postulate that intranuclear, intracytoplasmic, and extracellular bound and diffusible forms exist in the endoneurial compartment. At present we have no knowledge of which of these forms is biologically active in determining cell damage and what proportion each represents of the whole endoneurial lead content previously discussed. Cellular Mechanisms. To this point the major discussion has concentrated on the circumstances of lead entry into the endoneurium and its contact with the Schwann cell. In terms of concentration at the site of action and time course, it is reasonable to propose that

FIGURE 32–104 A, Schwann cell nucleus (unstained with lead; ⫻21,000). B, Endothelial cell nucleus (stained with lead; ⫻28,000). Both contain inclusion bodies. (From Ohnishi, A., and Dyck, P. J.: Retardation of Schwann cell division and axonal regrowth following nerve crush in experimental lead neuropathy. Ann. Neurol. 10:469, 1981, with permission.)

lead has a direct effect on Schwann cell function. There are several proposed mechanisms by which lead may exert this effect. However, direct evidence concerning the effect of lead on isolated enzyme or subcellular fractions is difficult to interpret because lead has affected virtually every enzyme system in which it was tested. Thus identifying a lead effect in vitro does not necessarily indicate a biologic effect in the whole animal. It is also true that identifying a cellular site of accumulation does not imply a cellular site of major action. Intranuclear inclusion bodies are one site at which lead is thought to accumulate within the cell nucleus, but it is suggested66,67 that this may represent a cellular protective response, sequestering the lead in an insoluble, and hence biologically inactive, complex.

Pathologic Alterations of Nerves

Our present knowledge suggests that the major site of action is either at the membrane level (myelin or mitochondrial) or at the level of mitochondrial oxidative metabolism. The spectrum of possible cellular sites of action is summarized. Lead might alter the composition of myelin either by binding directly to the membrane or by altering the synthesis of the normal lipid and protein components so that myelin structure can no longer be maintained. Preliminary results from electron microscopic autoradiographic studies suggest that 210Pb binds preferentially to myelin membranes. This binding may be reflected by the observed increase in rigidity in nuclear magnetic resonance studies of myelin membranes of chronically lead-intoxicated rats.160 A widespread membrane effect is suggested by the observation that red cell mechanical fragility is increased and osmotic fragility decreased as a result of increased membrane rigidity in lead-intoxicated animals 83,100 (Y. Wedmid, personal communication, 1993). These changes in membrane fluidity may reflect changes in the lysolecithin and cholesterol content of the lipid bilayer. This membrane effect might be implicated in lead interference with mitochondrial oxidative metabolism. Thus mitochondrial conformational changes thought to be essential for the coupling of oxidative phosphorylation might be prevented. Walton and Buckley157 found early mitochondrial changes in a cell culture system (chick embryo kidney tubule cells). Cell cultures exposed to concentrations of lead as low as 10⫺7 mol/L showed accumulation of electron-dense particles in cytoplasmic vesicles and mitochondria. Walton and Buckley confirmed by microincineration that the particles contained lead (or at least were mineral rich) and postulated that these particles mechanically interfered with normal mitochondrial membrane conformational changes. Abnormal mitochondrial conformations have been noted in renal tubular cells67 and in cultured muscle cells.74 This hypothesis is strengthened by the observation of Barltrop and colleagues7 that the radioactive isotope 210Pb was rapidly concentrated in the renal mitochondria of rat. That lead might interfere with the replication, transcription, or translation of genetic information has been suggested, although there is little direct evidence from other cell systems to support this. Ohnishi and Dyck demonstrated that Schwann cell division after crush injury is retarded in lead-intoxicated animals.118 However, this may not have a bearing on the primary process of demyelination, particularly since the number of Schwann cells is actually increased in the uncrushed nerve in lead-intoxicated animals. Intranuclear inclusion bodies have been reported as characteristic of lead poisoning for many years, but the significance of these bodies remains unclear. They appear to be separate from

823

chromatin,157 and, as discussed earlier, it has been suggested that they sequester lead in an insoluble and hence metabolically inactive protein complex. Several studies have suggested an effect on chromosomes and on passage of genetic information, but none of these has concerned the nervous system.10,62,156 It has been proposed that lead interferes with calcium-dependent metabolism by competitively binding to calcium receptor sites on enzymes and membrane proteins, thus interfering with substrate binding or inducing conformational changes in these proteins. Such interactions would explain the fact that decreased dietary calcium produces increased lead retention in the body, although the physiologic basis for this observation is not clear.9,111,113,143 It has also been suggested that calcium and lead interact at the cellular level.81 Walton and Buckley presented evidence that, in chick embryo renal cell culture, lead interacts with the sodium-dependent calcium ATPase pump at the level of the plasma membrane and with a postulated bivalent cation transport mechanism in the inner mitochondrial membrane.157 In studies of the CNS, Silbergeld and co-workers139–141 have demonstrated potential interactions of lead and calcium at both the synaptosomal and capillary endothelial mitochondrial levels. It still remains possible that lead interferes with some component outside the Schwann cells, with demyelination caused not directly by action of lead but by alterations in the endoneurial environment produced by lead. However, the possibility that a factor carried specifically in the edema fluid is responsible for the cellular changes has probably been excluded. Inherited Tendency to Pressure Palsy In this dominantly inherited disorder the peripheral nerves are unusually susceptible to pressure injury. Degrees of pressure not known to affect nerves cause focal neurologic abnormality. Uncompressed nerves may have abnormally low conduction velocities. Biopsy of nerves has shown short segments along internodes with excessively reduplicated myelin. Yoshikawa has found that a striking subclinical structure alteration characterizes unaffected myelin internodes.165 There are focal regions of uncompacted myelin adjacent to axolemma (Fig. 32–105). This may reflect an abnormality of adhesion molecules related to the susceptibility of these nerve fibers. Congenital Tomaculous Neuropathy Several patients have been described whose nerve fibers were characterized by hypomyelination and myelin thickening caused by reduplicated myelin.155,165

824

Pathology of the Peripheral Nervous System

FIGURE 32–105 Longitudinal electron micrograph of myelinated fiber of sural nerve of a patient with inherited tendency to pressure palsy. The failure of myelin compaction may suggest a developmental abnormality of myelin that accounts for the inherited liability to pressure injury. (From Yoshikawa, Y., Dyck, P. J., Poduslo, J. F., and Giannini, C.: Polyglucosan body axonal enlargement increases myelin spiral length but not lamellar number. J. Neurol. Sci. 98:107, 1990, with permission.)

REFERENCES 1. Aguayo, A. J., Attiwell, M., Trecarten, J., et al.: Abnormal myelination in transplanted Trembler mouse Schwann cells. Nature 265:73, 1977. 2. Aitken, J. T., and Thomas, P. K.: Retrograde changes in fiber size following nerve section. J. Anat. 96:121, 1962. 3. Anders, E., Dietz, D. D., Bagnell, C. R. Jr., et al.: Morphological, pharmacokinetic, and hematological studies of lead-exposed pigeons. Environ. Res. 28:344, 1982. 4. Asbury, A. K.: Renaut bodies: a forgotten endoneurial structure. J. Neuropathol. Exp. Neurol. 32:334, 1973. 5. Asbury, A. K., Aldredge, H., Hershberg, R., and Fisher, C. M.: Oculomotor palsy in diabetes mellitus: a clinico-pathological study. Brain 93:555, 1970. 6. Asbury, A. K., Arnason, B. G., and Adams, R. D.: The inflammatory lesion in idiopathic polyneuritis: its role in pathogenesis. Medicine 48:173, 1969. 7. Barltrop, D., Barrett, A. J., and Dingle, J. T.: Subcellular distribution of lead in the rat. J. Lab. Clin. Med. 77:705, 1971. 8. Barnes, J. M., and Denz, F. W.: Experimental demyelination and organophosphorus compounds. J. Pathol. Bacteriol. 65:597, 1953. 9. Barton, J. C., Conrad, M. E., Harrison, L., and Nuby, S.: Effects of calcium on the absorption and retention of lead. J. Lab. Clin. Med. 91:366, 1978.

10. Beckman, S. S.: Intranuclear inclusion bodies in the kidney and liver caused by lead poisoning. Bull. Johns Hopkins Hosp. 58:384, 1936. 11. Benstead, T. J., Dyck, P. J., and Sangalang, V.: Inner perineurial cell vulnerability in ischemia. Brain Res. 489:177, 1989. 12. Besinger, U. A., Toyka, K. V., Anzil, A. P., et al.: Myeloma neuropathy: passive transfer from man to mouse. Science 213:1027, 1981. 13. Bignami, A., Dahl, D., Nguyen, B. T., and Crosby, C. J.: The fate of axonal debris in Wallerian degeneration of rat optic and sciatic nerves: electron microscopy and immunofluorescence studies with neurofilament antisera. J. Neuropathol. Exp. Neurol. 40:537, 1981. 14. Bischoff, A., Fierz, U., Regli, F., and Ulrich, J.: Peripheral neurological disorders in Fabry’s disease (angiokeratoma corporis diffusum universale): clinical and electron microscopic findings in a case [in German]. Klin. Wochenschr. 46:666, 1968. 15. Bourque, C. N., Anderson, B. A., Martin del Campo, C., and Sima, A. A.: Sensorimotor perineuritis—an autoimmune disease? Can. J. Neurol. Sci. 12:129, 1985. 16. Boyd, I. A., and Davey, M. R.: Composition of Peripheral Nerves. Edinburgh, E & S Livingston, 1968. 17. Bradley, W. G., and Jenkison, M.: Abnormalities of peripheral nerves in murine muscular dystrophy. J. Neurol. Sci. 18:227, 1973.

Pathologic Alterations of Nerves 18. Bunge, R. P., Bunge, M. B., Williams, A. K., and Wartels, L. K.: Does the dystrophic mouse nerve lesion result from an extracellular matrix abnormality? In Schotland, D. L. (ed.): Disorders of the Motor Unit. New York, Wiley, p. 23, 1982. 19. Carpenter, S.: Proximal axonal enlargement in motor neuron disease. Neurology 18:841, 1968. 20. Cavanagh, J. B.: The toxic effects of tri-ortho-cresyl phosphate on the nervous system: an experimental study in lens. J. Neurol. Neurosurg. Psychiatry 17:163, 1954. 21. Chou, S. M., and Hartmann, H. A.: Axonal lesions and waltzing syndrome after IDPN administration in rats. Acta Neuropathol. (Berl.) 3:428, 1964. 22. Chou, S. M., and Hartmann, H. A.: Electron microscopy of focal neuroaxonal lesions produced by beta-betaiminodipropionitrile (IDPN) in rats. Acta Neuropathol. (Berl.), 4:590, 1965. 23. Clark, A. W., Griffin, J. W., and Price, D. L.: The axonal pathology in chronic IDPN intoxication. J. Neuropathol. Exp. Neurol. 39:42, 1980. 24. Cragg, G. B., and Thomas, P. K.: Changes in conduction velocity and fibre size proximal to peripheral nerve lesions. J. Physiol. (Lond.) 157:315, 1961. 25. Dayan, A. D., Graveson, G. S., Robinson, P. K., and Woodhouse, M. A.: Globular neuropathy: a disorder of axons and Schwann cells. J. Neurol. Neurosurg. Psychiatry 31:552, 1968. 26. Denny-Brown, D., and Brenner, C.: Paralysis of nerve induced by direct pressure and by tourniquet. Arch. Neurol. Psychiatry 51:1, 1944. 27. Doinikow, B.: Beitrage zur histologie und histopathologie der peripheren nerven. Histol. Histopathol. Arb. 4:445, 1911. 28. Donaghy, M., Brett, E. M., Ormerod, I. E., et al.: Giant axonal neuropathy: observations on a further patient. J. Neurol. Neurosurg. Psychiatry 51:991, 1988. 29. Donaghy, M., King, R. H., Thomas, P. K., and Workman, J. M.: Abnormalities of the axonal cytoskeleton in giant axonal neuropathy. J. Neurocytol. 17:197, 1988. 30. Dreyfus, R. M., Hakim, S., and Adams, R. D.: Diabetic ophthalmoplegia: report of a case with postmortem study and comments on vascular supply of human oculomotor nerve. Arch. Neurol. Psychiatry 77:337, 1957. 31. Dyck, P. J.: Detection, characterization, and staging of polyneuropathy: assessed in diabetics. Muscle Nerve 11:21, 1988. 32. Dyck, P. J.: Experimental hypertrophic neuropathy: pathogenesis of onion-bulb formations produced by repeated tourniquet applications. Arch. Neurol. 21:73, 1969. 33. Dyck, P. J.: Is there an axonal variety of GBS? Neurology 43:1277, 1993. 34. Dyck, P. J., Conn, D. L., and Okazaki, H.: Necrotizing angiopathic neuropathy: three-dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972. 35. Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in Dejerine-Sottas disease. Mayo Clin. Proc. 45:286, 1970.

825

36. Dyck, P. J., Ellefson, R. D., Yao, J. K., and Herbert, P. N.: Adult-onset of Tangier disease: 1. Morphometric and pathologic studies suggesting delayed degradation of neutral lipids after fiber degeneration. J. Neuropathol. Exp. Neurol. 37:119, 1978. 37. Dyck, P. J., Giannini, C., and Lais, A.: Pathologic alterations of nerves. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 514, 1993. 38. Dyck, P. J., Johnson, W. J., Lambert, E. H., and O’Brien, P. C.: Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin. Proc. 46:400, 1971. 39. Dyck, P. J., Karnes, J., Sparks, M., and Low, P. A.: The morphometric composition of myelinated fibers by nerve, level, and species related to nerve microenvironment and ischemia. Electroencephalogr. Clin. Neurophysiol. Suppl. 36:39, 1982. 40. Dyck, P. J., Kawamura, Y., Low, P. A., et al.: The number and sizes of reconstructed peripheral autonomic, sensory and motor neurons in a case of dysautonomia. J. Neuropathol. Exp. Neurol. 27:741, 1978. 41. Dyck, P. J., Kratz, K. M., Lehman, K. A., et al.: The Rochester Diabetic Neuropathy Study: design, criteria for types of neuropathy, selection bias, and reproducibility of neuropathic tests. Neurology 41:799, 1991. 42. Dyck, P. J., and Lais, A. C.: Electron microscopy of teased nerve fibers: method permitting examination of repeating structures of same fiber. Brain Res. 23:418, 1970. 43. Dyck, P. J., and Lais, A. C.: Evidence for segmental demyelination secondary to axonal degeneration in Friedreich’s ataxia. In Kakulas, B. A. (ed.): Clinical Studies in Myology. Amsterdam, Excerpta Medica, p. 253, 1973. 44. Dyck, P. J., Lais, A. C., Giannini, C., and Engelstad, J. K.: Structural alterations of nerve during cuff compression. Proc. Natl. Acad. Sci. U. S. A. 87:9828, 1990. 45. Dyck, P. J., Lais, A. C., Hansen, S. M., et al.: Technique assessment of demyelination from endoneurial injection. Exp. Neurol. 77:359, 1982. 46. Dyck, P. J., and Lambert, E. H.: Dissociated sensation in amyloidosis: compound action potential, quantitative histologic and teased fiber, and electron microscopic studies of sural nerve biopsies. Arch. Neurol. 20:490, 1969. 47. Dyck, P. J., and Lambert, E. H.: Polyneuropathy associated with hypothyroidism. J. Neuropathol. Exp. Neurol. 29:631, 1970. 48. Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerves in health, Friedreich’s ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb, W. A. (ed.): Handbook of Electroencephalography and Clinical Neurophysiology. Amsterdam, Elsevier, p. 83, 1971. 49. Dyck, P. J., and Lofgren, E. P.: Nerve biopsy: choice of nerve, method, symptoms, and usefulness. Med. Clin. North Am. 52:885, 1968. 50. Dyck, P. J., Low, P. A., Sparks, M. F., et al.: Effect of serum hyperosmolality on morphometry of healthy human sural nerve. J. Neuropathol. Exp. Neurol. 39:285, 1980.

826

Pathology of the Peripheral Nervous System

51. Dyck, P. J., Nukada, H., Lais, A. C., and Karnes, J. L.: Permanent axotomy: a model of chronic neuronal degeneration preceded by axonal atrophy, myelin remodeling and degeneration. In Dyck, P. J., Thomas, P. K., Lambert, E. H. and Bunge, R. P. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 666, 1984. 52. Dyck, P. J., and O’Brien, P. C.: Meaningful degrees of prevention or improvement of nerve conduction in controlled clinical trials of diabetic neuropathy. Diabetes Care 12:649, 1989. 53. Dyck, P. J., O’Brien, P. C., and Ohnishi, A.: Lead neuropathy: II. Random distribution of segmental demyelination among “old internodes” of myelinated fibers. J. Neuropathol. Exp. Neurol. 36:570, 1977. 54. Dyck, P. J. B., and Dyck, P. J.: Chronic inflammatory demyelinating mononeuropathy (CIDM): a focal form of CIDP? Neurology 58:A303, 2002. 55. Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy (LSRPN): similarity to the diabetic variety (DLSRPN). J. Neuropathol. Exp. Neurol. 59:525, 2000. 56. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999. 57. Elzholz, A.: Zur Kenntnis der Veranderungen im centralen Strumpf ladierter gemischter Nerven. Jahrb. Psychiatr. 17:323, 1898. 58. Erlanger, J., and Gasser, H. S.: Electrical Signs of Nervous Activity. Philadelphia, University of Pennsylvania Press, 1937. 59. Evans, M. J., Finean, J. B., and Woolf, A. L.: Ultrastructural studies of human cutaneous nerve with special reference to lamellated cell inclusions and vacuole-containing cells. J. Clin. Pathol. 18:188, 1965. 60. Feasby, T. E., Hahn, A. F., and Gilbert, J. J.: Passive transfer of demyelinating activity in Guillain-Barre polyneuropathy. Neurology (N. Y.) 30:363, 1980. 61. Fullerton, P.: The relation between fibre diameter and internodal length in chronic neuropathy. J. Physiol. (Lond.) 178:26, 1965. 62. Gerber, G. B., Leonard, A., and Jacquet, P.: Toxicity, mutagenicity and teratogenicity of lead. Mutat. Res. 76:115, 1980. 63. Gombault, A.: Contribution a l’etude anatomique de la nevrite parenchymateuse subaigue et chronique: nevrite sementaire peri-axile. Arch. Neurol. (Paris) 1:11, 1880. 64. Gombault, A.: Sur les lesions de la nevrite alcoolique. C. R. Soc. Biol. (Paris) 102:439, 1886. 65. Gonatas, N. K., Evangelista, I., and Martin, J.: A generalized disorder of nervous system, skeletal muscle and heart resembling Refsum’s disease and Hurler’s syndrome. Am. J. Med. 42:169, 1967. 66. Goyer, R. A.: Lead and the kidney. Curr. Top. Pathol. 55:147, 1971. 67. Goyer, R. A., and Rhyne, B. C.: Pathological effects of lead. Int. Rev. Exp. Pathol. 12:1, 1973. 68. Griffin, J. W., Hoffman, P. N., Clark, A. W., et al.: Slow axonal transport of neurofilament proteins: impairment by ␤,␤⬘-iminodipropionitrile. Science 202:633, 1978.

69. Griffin, J. W., and Price, D. L.: Proximal axonopathies induced by toxic chemicals. In Spencer, P. S., and Schaumberg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 161, 1980. 70. Grundfest, H.: The effects of hydrostatic pressure upon excitability, the recovery, and the potential sequence of frog nerve. Cold Spring Harb. Symp. Quant. Biol. 4:179, 1936. 71. Gutmann, E., and Sanders, F. K.: Recovery of fibre numbers and diameters in the regeneration of peripheral nerves. J. Physiol. (Lond.) 101:489, 1943. 72. Haas, L. F., Austad, W. I., and Bergin, J. D.: Tangier disease. Brain 97:351, 1974. 73. Hall, S. M., and Gregson, N. A.: The in vivo and ultrastructural effects of injection of lysophosphatidylcholine into myelinated peripheral nerve fibers of the adult mouse. J. Cell Sci. 9:769, 1971. 74. Harary, I., and Berliner, J.: Electron microscopic examination of lead treated L6 skeletal muscle line cells in culture. J. Environ. Pathol. Toxicol. 4:305, 1980. 75. Haymaker, W., and Kernohan, J. W.: The Landry-GuillainBarre syndrome: clinicopathologic report of 50 fatal cases and a critique of the literature. Medicine 28:59, 1949. 76. Hess, K., Eames, R., Darveniza, P., and Gilliatt, R. W.: Acute ischemic neuropathy in the rabbit. J. Neurol. Sci. 44:19, 1979. 77. Hietanen, E., Kilpio, J., Narhi, M., et al.: Biotransformational and neurophysiological changes in rabbits exposed to lead. Arch. Environ. Contam. Toxicol. 9:337, 1980. 78. Hirano, A.: Aspects of the ultrastructure of amyotrophic lateral sclerosis. In Rowland, L. P. (ed.): Advances in Neurology: Human Motor Neuron Diseases. New York, Raven Press, p. 75, 1982. 79. Hoffer, J. A., Stein, R. B., and Gordon, T.: Differential atrophy of sensory and motor fibers following section of cat peripheral nerves. Brain Res. 178:347, 1979. 80. Hoffman, P. N., and Lasek, R. J.: Axonal transport of the cytoskeleton in regenerating motor neurons: constancy and change. Brain Res. 202:317, 1980. 81. Jacquet, P., and Gerber, G. B.: Teratogenic effects of lead in the mouse. Biomedicine 30:223, 1979. 82. Johnson, P. B., Brendel, K., and Meezan, E.: Human diabetic perineurial cell basement membrane thickening. Lab. Invest. 44:265, 1981. 83. Karai, I., Fukumoto, K., and Horiguchi, S.: Studies on osmotic fragility of red blood cells determined with a coil planet centrifuge for workers occupationally exposed to lead. Int. Arch. Occup. Environ. Health 48:273, 1981. 84. Karnes, J., Robb, R., O’Brien, P. C., et al.: Computerized image recognition for morphometry of nerve attribute of shape of sampled transverse sections of myelinated fibers which best estimates their average diameter. J. Neurol. Sci. 34:43, 1977. 85. Kawamura, Y., and Dyck, P. J.: Evidence for three populations by size in L5 spinal ganglion in man. J. Neuropathol. Exp. Neurol. 37:269, 1978. 86. Kawamura, Y., Dyck, P. J., Shimono, M., et al.: Morphometric comparison of the vulnerability of peripheral motor and sensory neurons in amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 40:667, 1981.

Pathologic Alterations of Nerves 87. Kawamura, Y., O’Brien, P., Okazaki, H., and Dyck, P. J.: Lumbar motoneurons of man: II. The number and diameter distribution of large- and intermediate-diameter cytons in “motoneuron columns” of spinal cord of man. J. Neuropathol. Exp. Neurol. 36:861, 1977. 88. Kernohan, J. W., and Woltman, H. W.: Periarteritis nodosa: a clinicopathologic study with special reference to the nervous system. Arch. Neurol. 39:655, 1938. 89. Kernohan, J. W., and Woltman, H. W.: Amyloid neuritis. Arch. Neurol. Psychiatry 47:132, 1942. 90. Key, A., and Retzius, G.: Studien in der Anatomie des Nerven-systems und des Bingegewebes, Vol. II. Stockholm, Nordstedt and Soner, 1876. 91. King, R. H., Llewelyn, J. G., Thomas, P. K., et al.: Diabetic neuropathy: abnormalities of Schwann cell and perineurial basal laminae. Implications for diabetic vasculopathy. Neuropathol. Appl. Neurobiol. 15:339, 1989. 92. Kocen, R. S., Lloyd, J. K., Lascelles, P. T., et al.: Familial alpha-lipoprotein deficiency (Tangier disease) with neurological abnormalities. Lancet 1:1341, 1967. 93. Kocen, R. S., and Thomas, P. K.: Peripheral nerve involvement in Fabry’s disease. Arch. Neurol. 22:81, 1970. 94. Korthals, J. K., Gieron, M. A., and Wisniewski, H. M.: Nerve regeneration patterns after acute ischemic injury. Neurology 39:932, 1989. 95. Korthals, J. K., Korthals, M. A., and Wisniewski, H. M.: Peripheral nerve ischemia. Part 2. Accumulation of organelles. Ann. Neurol. 4:487, 1978. 96. Korthals, J. K., and Wisniewski, H. M.: Peripheral nerve ischemia. Part 1. Experimental model. J. Neurol. Sci. 24:65, 1975. 97. Krucke, W.: Erkrankungen der peripheren Nerven. In Henke, F., Labarsch, O. and Rossle, R. (eds.): Handbuch der speziellen pathologischen Anatomie und Histologie. Berlin, Springer-Verlag, p. 105, 1955. 98. Kummer, H., Laissue, J., Spiess, H., et al.: Familial alphalipoproteinemia (Tangier disease) [in German]. Schweiz. Med. Wochenschr. 98:406, 1968. 99. Lampert, P. W., and Schochet, S. S. Jr.: Demyelination and remyelination in lead neuropathy: electron microscopic studies. J. Neuropathol. Exp. Neurol. 27:527, 1968. 100. Levander, O. A., Fisher, M., Morris, V., and Ferretti, R. J.: Morphology of erythrocytes from vitamin E-deficient leadpoisoned rats. J. Nutr. 107:1828, 1978. 101. Lewis, R. A., Sumner, A. J., Brown, M. J., and Asbury, A. K.: Multifocal demyelinating neuropathy with persistent conduction block. Neurology 32:958, 1982. 102. Low, P. A., Baumann, W. J., and Parthasarathy, S.: Structure-demyelinating-activity study of lysolecithin analogs in mammalian peripheral nerve. Neurology 32:A104, 1982. 103. Low, P. A., and Dyck, P. J.: Increased endoneurial fluid pressure in experimental lead neuropathy. Nature 269:427, 1977. 104. Low, P. A., Okazaki, H., and Dyck, P. J.: Splanchnic preganglionic neurons in man. I. Morphometry of preganglionic cytons. Acta Neuropathol. (Berl.) 40:55, 1977. 105. Low, P. A., Schmelzer, J. D., and Dyck, P. J.: Results of endoneurial injection of Guillain-Barre serum in Lewis rats. Mayo Clin. Proc. 57:360, 1982.

827

106. Lubinska, L.: Region of transection between preserved and regenerating parts of myelinated nerve fibers. J. Comp. Neurol. 113:315, 1959. 107. Lubinska, L.: Patterns of wallerian degeneration of myelinated fibres in short and long peripheral stumps and in isolated segments of rat phrenic nerve: interpretation of the role of axoplasmic flow of the trophic factor. Brain Res. 233:227, 1982. 108. Luna, L. G.: Manual of Histologic Staining Methods of the Armed Forces Institute of Pathology, 3rd ed. New York, McGraw-Hill, 1968. 109. Madrid, R. E., and Wisniewski, H. M.: Axonal degeneration in demyelinating disorders. J. Neurocytol. 6:103, 1977. 110. Moore, J. F., Goyer, R. A., and Wilson, M.: Lead-induced inclusion bodies: solubility, amino acid content, and relationship to residual acidic nuclear proteins. Lab. Invest. 29:488, 1973. 111. Moore, M. R.: Diet and lead toxicity. Proc. Nutr. Soc. 38:243, 1979. 112. Myers, R. R., Powell, H. C., Shapiro, H. M., et al.: Changes in endoneurial fluid pressure, permeability, and peripheral nerve ultrastructure in experimental lead neuropathy. Ann. Neurol. 8:392, 1980. 113. Mylroie, A. A., Moore, L., Olyai, B., and Anderson, M.: Increased susceptibility to lead toxicity in rats fed semipurified diets. Environ. Res. 15:57, 1978. 114. Nukada, H., and Dyck, P. J.: Microsphere embolization of nerve capillaries and fiber degeneration. Am. J. Pathol. 115:275, 1984. 115. Nukada, H., and Dyck, P. J.: Acute ischemia causes axonal stasis, swelling, attenuation, and secondary demyelination. Ann. Neurol. 22:311, 1987. 116. Ochoa, J., Danta, G., Fowler, T. J., and Gilliatt, R. W.: Nature of the nerve lesion caused by a pneumatic tourniquet. Nature 233:265, 1971. 117. Ochoa, J., Fowler, T. J., and Gilliatt, R. W.: Anatomical changes in peripheral nerves compressed by a pneumatic tourniquet. J. Anat. 113:433, 1972. 118. Ohnishi, A., and Dyck, P. J.: Loss of small peripheral sensory neurons in Fabry disease: histologic and morphometric evaluation of cutaneous nerves, spinal ganglia, and posterior columns. Arch. Neurol. 31:120, 1974. 119. Ohta, M., Offord, K., and Dyck, P. J.: Morphometric evaluation of first sacral ganglia of man. J. Neurol. Sci. 22:73, 1974. 120. Okada, E.: Ueber Zwiebelartige Gebilde im peripherischen Nerven (Renaut’sche Korperchen) bei einem Fall von Kakke (Beriberi). Tokyo Imperial Univ. Coll. Med. 6:93, 1903. 121. Okazaki, H., and Campbell, R. J.: Nervous system. In Ludwig, J. (ed.): Current Methods of Autopsy Practice, 2nd ed. Philadelphia, W. B. Saunders, p. 95, 1979. 122. Parry, G. J., and Brown, M. J.: Selective fiber vulnerability in acute ischemic neuropathy. Ann. Neurol. 11:147, 1982. 123. Powell, H. C., Myers, R. R., and Lampert, P. W.: Changes in Schwann cells and vessels in lead neuropathy. Am. J. Pathol. 109:193, 1982. 124. Powell, H. C., Ward, H. W., Garrett, R. S., et al.: Glycogen accumulation in the nerves and kidney of chronically diabetic rats: a quantitative electron microscopic study. J. Neuropathol. Exp. Neurol. 38:114, 1979.

828

Pathology of the Peripheral Nervous System

125. Prineas, J.: The pathogenesis of dying-back polyneuropathies: I. An ultrastructural study of experimental tri-ortho-cresyl phosphate intoxication in the cat. J. Neuropathol. Exp. Neurol. 28:571, 1969. 126. Prineas, J.: The pathogenesis of dying-back polyneuropathies: II. An ultrastructural study of experimental acrylamide intoxication in the cat. J. Neuropathol. Exp. Neurol. 28:598, 1969. 127. Reich, F.: Uber den zelligen Aufbau der Nervenfaser auf Grund mikrohistichemischer Untersuchungen. J. Psychol. Neurol. 8:244, 1907. 128. Renaut, J.: Systeme hyalin de soutenement des centres nerveux et de quelques organes des sens. Arch. Physiol. Normale Pathol. 8:854–860, 1881. 129. Renaut, M. J.: Recherches sur quelques points particuliers de l’histologie des nerfs. I. La gaine lamelleuse et le systeme hyalin intravaginal. Arch. N. Psych. 8:161–190, 1881. 130. Romeis, B.: Mikroskopische Technik, Vol. 15. Munchen, Leibniz, 1948. 131. Saida, G., Saida, K., Saida, T., and Asbury, A. K.: Axonal lesions in acute experimental demyelination: sequential teased nerve fiber study. Neurology (N. Y.) 31:413, 1981. 132. Saida, T., Saida, K., Silberberg, D. H., and Brown, M. J.: Transfer of demyelination by intraneural injection of experimental allergic neuritis serum. Nature 272:639, 1978. 133. Schlaepfer, W. W., and Hasler, M. B.: The persistence and possible externalization of axonal debris during wallerian degeneration. J. Neuropathol. Exp. Neurol. 38:242, 1979. 134. Schoene, W. C., Asbury, A. K., Astrom, K. E., and Masters, R.: Hereditary sensory neuropathy: a clinical and ultrastructural study. J. Neurol. Sci. 11:463, 1970. 135. Schuchmann, J. A.: Isolated sural neuropathy: report of two cases. Arch. Phys. Med. Rehabil. 61:329, 1980. 136. Seitelberger, F.: Neuroaxonal dystrophy: its relation to aging and neurologic diseases. In Vinken, P., Bruyn, G., and Klawans, H. (eds): Extrapyramidal Disorders. Amsterdam, Elsevier, p. 391, 1986. 137. Shelton, K. R., and Egle, P. M.: The proteins of leadinduced intranuclear inclusion bodies. J. Biol. Chem. 257:11802, 1982. 138. Shimono, M., Izumi, K., and Kuroiwa, Y.: 3, 3⬘-Iminodipropionitrile induced centrifugal segmental demyelination and onion bulb formation. J. Neuropathol. Exp. Neurol. 37:375, 1978. 139. Silbergeld, E. K., Adler, H. S., and Costa, J. L.: Subcellular localization of lead in synaptosomes. Res. Commun. Chem. Pathol. Pharmacol. 17:715, 1977. 140. Silbergeld, E. K., Hruska, R. E., Miller, L. P., and Eng, N.: Effects of lead in vivo and in vitro on GABAergic neurochemistry. J. Neurochem. 34:1712, 1980. 141. Silbergeld, E. K., Wolinsky, J. S., and Goldstein, G. W.: Electron probe microanalysis of isolated brain capillaries poisoned with lead. Brain Res. 189:369, 1980. 142. Sinnreich, M., Daube, J. R., Klein, C. J., and Dyck, P. J. B.: Inflammatory sensory polyradiculopathy. Neurology 60:A159, 2003. 143. Sobel, A. E., Yuska, H., and Kramer, B.: The biochemical behavior of lead. I. Influence of calcium, phosphorus and

144. 145.

146.

147.

148.

149.

150.

151.

152.

153. 154. 155.

156.

157.

158.

159.

160.

161.

vitamin D on lead in blood and bone. J. Biol. Chem. 132:239, 1940. Spencer, P. S.: The traumatic neuroma and proximal stump. Bull. Hosp. Joint Dis. 35:85, 1974. Spencer, P. S., and Schaumburg, H. H.: Central-peripheral distal axonopathy—the pathology of dying-back polyneuropathies. In Zimmerman, H. M. (ed.): Progress in Neuropathy. New York, Grune & Stratton, p. 252, 1976. Spencer, P. S., and Thomas, P. K.: The examination of isolated nerve fibres by light and electron microscopy, with observations on demyelination proximal to neuromas. Acta Neuropathol. (Berl.) 16:177, 1970. Spencer, P. S., and Thomas, P. K.: Ultrastructural studies of the dying-back process. II. The sequestration and removal by Schwann cells and oligodendrocytes of organelles from normal and diseased axons. J. Neurocytol. 3:763, 1974. Steward, V. W., and Hitchcock, C.: Fabry’s disease (angiokeratoma corporis diffusum): a report of five cases with pain in the extremities as the chief symptom. Pathol. Eur. 3:377, 1968. Stoll, G., Reiners, K., Schwendemann, G., et al.: Normal myelination of regenerating peripheral nerve sprouts despite circulating antibodies to galactocerebroside in rabbits. Ann. Neurol. 19:189, 1986. Stoner, G. L., Brosnan, C. F., Wisniewski, H. M., and Bloom, B. R.: Studies on demyelination by activated lymphocytes in the rabbit eye. I. Effects of a mononuclear cell infiltrate induced by products of activated lymphocytes. J. Immunol. 118:2094, 1977. Terzis, J. K., Daniel, R. K., Williams, H. B., and Spencer, P. S.: Benign fatty tumors of the peripheral nerves. Ann. Plast. Surg. 1:193, 1978. Thomas, P. K., and Slatford, J.: Lamellar bodies in the cytoplasm of Schwann cells. Proc. Anat. Soc. Great Britain Ireland 98:691, 1964. Thomas, P. K., and Young, J. Z.: Internode lengths in the nerves of fishes. J. Anat. 82:336, 1968. Tomonaga, M., and Sluga, E.: Zur Ultrastruktur der (pi)-granula. Acta Neuropathol. (Berl.) 15:56, 1970. Vallat, J. M., Gil, R., Leboutet, M. J., et al.: Congenital hypo- and hypermyelination neuropathy: two cases. Acta Neuropathol. (Berl.) 74:197, 1987. Verschaeve, L., Driesen, M., Kirsch-Volders, M., et al.: Chromosome distribution studies after inorganic lead exposure. Hum. Genet. 49:147, 1979. Walton, I., and Buckley, I. K.: The lead-poisoned cell: a fine structural study using cultured kidney cells. Exp. Mol. Pathol. 27:167, 1977. Weber, R. B., Daroff, R. B., and Mackey, E. A.: Pathology of oculomotor nerve palsy in diabetics. Neurology 20:835, 1970. Weller, R. O., and Das Gupta, T. K.: Experimental hypertrophic neuropathy: an electron microscope study. J. Neurol. Neurosurg. Psychiatry 31:34, 1968. Williams, I. R., Jefferson, D., and Gilliatt, R. W.: Acute nerve compression during limb ischaemia—an experimental study. J. Neurol. Sci. 46:199, 1980. Windebank, A. J., and Dyck, P. J.: Kinetics of 210Pb entry into the endoneurium. Brain Res. 225:67, 1981.

Pathologic Alterations of Nerves 162. Windebank, A. J., McCall, J. T., Hunder, H. G., and Dyck, P. J.: The endoneurial content of lead related to the onset and severity of segmental demyelination. J. Neuropathol. Exp. Neurol. 39:692, 1980. 163. Wisniewski, H. M., and Bloom, B. R.: Primary demyelination as a nonspecific consequence of a cell-mediated immune reaction. J. Exp. Med. 141:346, 1975.

829

164. Wohlfart, G.: Degenerative and regenerative axonal changes in the ventral horns, brain stem, and cerebral cortex in amyotrophic lateral sclerosis. Acta Univ. Lund. (Ner Ser. 2) 56:1, 1959. 165. Yoshikawa, H., Dyck, P. J., Poduslo, J. F., and Giannini, C.: Polyglucosan body axonal enlargement increases myelin spiral length but not lamellae number. J. Neurol. Sci. 98:107, 1990.

33 Diseases of the Neuromuscular Junction ANDREW G. ENGEL

The Safety Margin of Neuromuscular Transmission Autoimmune Disorders Myasthenia Gravis Clinical Features Pathogenesis Pathology Lambert-Eaton Myasthenic Syndrome Clinical Features Pathogenesis and Pathology

Congenital Myasthenic Syndromes Choline Acetyltransferase Deficiency Clinical Features End-Plate Studies Molecular Pathogenesis Paucity of Synaptic Vesicles and Reduced Quantal Release Congenital Myasthenic Syndrome Resembling LEMS

CMS with Reduced Quantal Release and Central Nervous System Symptoms End-Plate Acetylcholinesterase Deficiency Clinical Features Pathogenesis and Pathology Molecular Pathogenesis Postsynaptic CMSs The Slow-Channel CMSs Clinical Features End-Plate Studies Molecular Pathogenesis Fast-Channel Syndromes Clinical Features End-Plate Studies Molecular Pathogenesis Mutations Causing AChR Deficiency with or without Minor Kinetic Abnormality Clinical Features End-Plate Studies

In all diseases of the neuromuscular junction, the safety margin of neuromuscular transmission is compromised by one or more specific mechanisms. The clinical concomitant is abnormal weakness and fatigability on exertion. Table 33–1 shows a classification of currently recognized diseases of the neuromuscular junction. This chapter first reviews the factors that affect the safety margin of neuromuscular transmission and then summarizes the clinical, pathologic, and molecular aspects of the autoimmune and congenital disorders of the neuromuscular junction.

The Safety Margin of Neuromuscular Transmission The neuromuscular junction consists of presynaptic and postsynaptic regions separated by the synaptic space. Each presynaptic region is made up of a nerve terminal covered

Molecular Pathogenesis CMS Caused by Rapsyn Deficiency Clinical Features End-Plate Studies Molecular Pathogenesis Sodium Channel Myasthenia Clinical Features End-Plate Studies Molecular Pathogenesis Relation to Other Sodium Channel Disorders CMS Associated with Plectin Deficiency

Therapy Myasthenia Gravis Lambert-Eaton Myasthenic Syndrome The Congenital Myasthenic Syndromes

by a Schwann cell process. Each postsynaptic region is composed of junctional folds that overlie a specialized region of the muscle fiber, the junctional sarcoplasm. The synaptic space includes a single primary and multiple secondary synaptic clefts lined by basal lamina. Acetylcholine (ACh) is stored in the nerve terminal in synaptic vesicles.93,103 An average synaptic vesicle contains 6000 to 8000 ACh molecules.90,111 The contents of the synaptic vesicles are discharged into the synaptic space by exocytosis.93,94 A transmitter quantum is the amount of ACh released from a single synaptic vesicle. The end-plate (EP)–specific form of acetylcholinesterase (AChE) is positioned in the synaptic basal lamina14,83,132 at a density of about 2500 sites/␮m2.185 The acetylcholine receptor (AChR) is concentrated on the terminal expansions of the junctional folds, where its packing density is close to 104/␮m2.79,95,96,116,131 Voltage-gated sodium channels (Nav) of the Nav1.4 type are concentrated in the depths of the junctional folds.63,180 831

832

Pathology of the Peripheral Nervous System

Table 33–1. Classification of End-Plate Diseases Autoimmune Myasthenia gravis Lambert-Eaton myasthenic syndrome Congenital Presynaptic Synaptic basal lamina–associated Postsynaptic Toxic Botulism Drug induced

In the resting state, there is a steady but random release of transmitter quanta from the nerve terminal associated with exocytosis of individual synaptic vesicles.60,127 Focally released ACh molecules diffuse into the synaptic space, but most are not hydrolyzed initially by AChE because of saturation of AChE by the high concentration of ACh.62,183 The peak concentration of ACh reaching the postsynaptic membrane is close to 1 mM.200,201 Under normal conditions and with AChE fully active, ACh molecules in a single quantum exert their effect within a distance of 0.8 ␮m from their site of release.89 AChE limits the number of collisions of ACh with AChR and also the radius of spread of ACh. When AChE is inactive, the lateral spread of ACh is increased, so that each ACh molecule can sequentially bind to multiple AChRs and open multiple ion channels before leaving the synaptic space by diffusion.107 The depolarization and the concomitant current flow induced by a single quantum give rise to the miniature EP potential (MEPP) and miniature EP current (MEPC), respectively. The amplitude of the MEPP is a function of the number of ion channels opened by the quantum, the conductance change per channel opening, and the input resistance of the muscle fiber.108,126,127 The amplitude of the MEPC is a function of the number of ion channel openings and the mean current flow per opening. The number of channel openings is a function of the number of ACh molecules in the quantum, the number of available AChRs, and the geometry of the synaptic space, which allows rapid diffusion of most ACh molecules from their site of release to AChR. The duration of the MEPP depends on the duration of channel opening events,3,106 the functional state of AChE,107 and the cable properties of the muscle fiber surface membrane.60 The duration of the MEPC is independent of the cable properties of the muscle fiber surface membrane; otherwise, its duration is affected by the same factors that determine the duration of the MEPP.127 Once dissociated from AChR, ACh is rapidly hydrolyzed by AChE to choline and acetate. Choline is taken up by

the nerve terminal via a high-affinity, sodium-dependent, and hemicholinium-sensitive choline transporter.4,13,164 ACh is resynthesized by choline acetyltransferase (ChAT) and is then transported into the synaptic vesicles by the vesicular ACh transporter (VAChT)169,174 in exchange for protons delivered to the synaptic vesicle by a proton pump.133,174 Depolarization of the nerve terminal by a nerve impulse is followed by an influx of calcium ions into the terminal through voltage-gated calcium channels (Cav), and it is this influx that mediates transmitter release.27,105,122,135 The calcium ingress into the nerve terminal also results in facilitation, so that the probability of release by a subsequent impulse is increased. The effects of calcium are antagonized by magnesium.99,104 The details of the mechanism by which calcium triggers synaptic vesicle exocytosis are still not fully understood. However, it is now clear that the increased calcium concentration in the nerve terminal releases the synaptic vesicles from cytoskeletal constraints and affects the interaction of several proteins that together form a synaptic vesicle fusion complex. Nerve stimulation results in synaptic vesicle exocytosis adjacent to the active zones of the presynaptic membrane.94 In the freeze-fractured presynaptic membrane, the active zones are represented by double parallel rows of large (10- to 12-nm) intramembrane particles. The binding pattern of fluorescent ␻-conotoxin, a ligand for Cav at the frog motor nerve terminal,176,204 and the spatial disposition of calcium concentration microdomains visualized in the stimulated squid giant synapse with n-aequorin124 support this assumption. The number of quanta released by a nerve impulse (m) depends on the probability of release (p) and on at least one additional factor designated as n, according to the formula m ⫽ np.31 The factor n was originally defined as the number of quantal units capable of responding to the nerve impulse, but it more likely indicates the number of readily releasable quanta,42 the number of active release sites in the nerve terminal,212,218 or a combination of these variables. The amplitude of the end-plate potential (EPP) is affected by the same factors that affect the amplitude of the MEPP and by the number of quanta released by a nerve impulse. When the EPP exceeds the threshold for activating the Nav1.4 sodium channels in the depth of the synaptic folds, it triggers a muscle fiber action potential.127 A high concentration of AChRs on crests of the folds184 and of Nav1.4 in the depth of the folds63,180 ensures that excitation is propagated beyond the EP.128,216 The safety margin of neuromuscular transmission is a function of the difference between the depolarization caused by the EPP and the depolarization required to activate Nav1.4. All congenital or acquired defects of neuromuscular transmission identified to date have been traced to one or more factors

833

Diseases of the Neuromuscular Junction

Myasthenia gravis (MG) is caused by EP AChR deficiency. In most patients the disease stems from an autoimmune response against AChR. Consistent with this, 80% to 90% of MG patients have circulating antibodies against AChR.121,190 A recent report that MG patients without anti-AChR antibodies have a significant titer of antibodies against MuSK, a muscle-specific tyrosine kinase that plays a role in the aggregation of AChR at the EP, implies that MG could also arise from an autoimmune response against MuSK.97 The AChR deficiency diminishes the synaptic response to ACh and hence the amplitude of the MEPP; this reduces the amplitude of the EPP and hence the safety margin of neuromuscular transmission.

in approximately 15%. Close to 90% of the generalizations occur within 13 months after the onset. In approximately 25% of patients, MG is associated with another autoimmune disease. The clinical classification proposed by Osserman in 1958166 and by Osserman and Genkins in 1971167 divides the disease into ocular (group 1, 15% to 20%), mild generalized (group 2A, 30%), moderately severe generalized (group 2B, 20%), acute fulminating (group 3, 11%), and late severe (group 4, 9%) forms. This classification remains useful for defining stages of the disease, but the distinctions between groups are based on subjective criteria. Susceptibility to develop MG depends on AChR epitope presentation to particular major histocompatibility complex human leukocyte antigen (HLA) proteins whose expression on immune cells is genetically determined. This is reflected in the higher incidence of HLA-A1, -B8, and -DR3 determinants in young white MG patients, and of different HLA determinants in other age and racial groups. Other susceptibility genes in MG include those encoding cytotoxic lymphocyte–associated antigen, interleukin-1␤, interleukin-1 receptor antagonist, and tumor necrosis factor-␣ and -␤ (reviewed by Li et al.120).

Clinical Features

Pathogenesis

The clinical features of MG are well described in classic papers.24,71,78,167,196,207,214 The cardinal clinical finding is abnormal weakness and fatigability of some or all voluntary muscles. The weakness increases with repeated or sustained exertion and over the course of the day. Menses, viral or other infections, and emotional upsets can worsen the symptoms. The symptoms respond to anticholinesterase drugs. The external ocular muscles are affected initially in approximately 50% and eventually in approximately 90% of the cases. Voluntary muscles innervated by cranial nerves (facial, masticatory, lingual, pharyngeal, and laryngeal muscles) and cervical, pectoral girdle, and hip flexor muscles are also frequently affected; proximal limb muscles are usually more severely affected than distal ones. The tendon reflexes are brisk or normally active but may diminish if repeatedly elicited; reflexes can be absent from muscles that cannot be activated by voluntary effort. There are no objective sensory deficits. The symptoms can fluctuate from day to day, from week to week, or over longer periods of time. Spontaneous remissions can occur during the first 3 years of the disease but seldom thereafter; however, long and complete remissions are rare. In patients progressing from mild to more severe disease, the weakness tends to spread from ocular to facial to lower bulbar muscles, and then to truncal and limb muscles, but this sequence may vary and different muscles may be affected either together or in succession. The disease is initially ocular in approximately 40% of patients, but remains confined to the ocular muscles only

The basic event that breaks tolerance to self-AChR remains unknown. Three predisposing conditions have been recognized: treatment with penicillamine,61 treatment with ␣- or ␤-interferon,80,84 and bone marrow transplantation.12 As in other autoimmune diseases, the afferent limb of the immune response involves presentation of processed antigen (peptide fragments of AChR) by HLA class II positive antigen-presenting cells to specific autoreactive CD4⫹ T-helper cells, which, in turn, stimulate production by B cells and plasma cells of antibodies that recognize specific epitopes of AChR. The thymus gland is likely involved in autoimmunity to MG for the following reasons: it contains myoid cells that express AChR; the myasthenic thymus gland contains lymph nodes with germinal centers that contain AChR-specific B cells that secrete anti-AChR antibodies; and the gland is hyperplastic in approximately 70% of patients and harbors epithelial tumors in approximately 15% of patients.98 These findings suggest that a thymic abnormality could result in recognition of self-AChR components as nonself and thereby trigger the afferent limb of the immune response. The efferent limb of the autoimmune response is mediated by anti-AChR antibodies that reduce the number of EP AChRs by antibody-dependent complement-mediated lysis of the junctional folds,143,182 by accelerated internalization and destruction of AChRs (antigenic modulation),199 and by blocking the binding of ACh to AChR.35,102 The AChR deficiency decreases the amplitude of the MEPP and hence

that render the EPP subthreshold for activating Na v1.4, or to a defect in Nav1.4 that renders it unresponsive to EPPs of normal amplitude.

Autoimmune Disorders MYASTHENIA GRAVIS

834

Pathology of the Peripheral Nervous System

that of the EPP, which, in turn, reduces the safety margin of neuromuscular transmission. The basic event that breaks tolerance to self-MuSK is also unknown. Anti-MuSK antibodies inhibit agrin-induced clustering of extrajunctional AChR expressed by myotubes97; this suggests that they also decrease the density of AChR at the EP. Studies of MuSK antibody–positive MG, however, are still incomplete: the number of AChRs per EP has not been determined, no immune deposits have been demonstrated at the EP, EP fine structure has not been analyzed, and in vitro electrophysiologic studies of patient EPs have not been performed. Thus the pathogenesis of MuSK antibody–positive MG is less well understood than that of AChR antibody–positive MG.

Pathology When visualized at the light microscopic level by the cholinesterase reaction, the normal EP is pretzel shaped; in contrast, the MG EP consists of multiple small and often dotlike regions dispersed over an extended length of the muscle fiber. This appearance is not specific to MG; it occurs with any condition that causes recurrent cycles of degeneration and regeneration of the postsynaptic region,182 and also in congenital EP AChR deficiency, where it is not accompanied by degeneration of the junctional folds. Both conditions exist at some EPs in MG, and one or both conditions can exist at EPs in the congenital myasthenic syndromes (CMSs).137,158 At the ultrastructural level, morphometric analysis of individual EP regions shows that the nerve terminals and postsynaptic regions are smaller than normal. Many postsynaptic regions are simplified and have too few and too short junctional folds.57 Focal or diffuse degeneration of the junctional folds with concomitant widening of the synaptic space occurs at nearly 90% of the EPs56 (see Figs. 33–1B, 33–2, and 33–3). The widened synaptic spaces harbor degenerate residues of the junctional folds surrounded by loops of basal lamina that invested the preexisting folds. Small nerve sprouts are present near highly degenerate postsynaptic regions; the nerve sprouts presumably seek new sites of contact with the muscle fiber to initiate formation of a new postsynaptic region that, in turn, will degenerate when the AChR comes under immune attack. Ultrastructural localization of AChR with peroxidaselabeled ␣-bungarotoxin reveals marked attenuation of the density of AChR on the junctional folds, and the length of the postsynaptic membrane reacting for AChR normalized for the length of the primary synaptic cleft (the AChR index) is markedly reduced51 (Fig. 33–1). Immunoelectron microscopy demonstrates immunoglobulin G (IgG) (Fig. 33–2) and complement components C3 and C9 (Fig. 33–3) on the crests of the junctional folds, where AChR is normally concentrated, and on debris shed into the synaptic space.49,182 The presence of C9 on the

junctional folds implies that the membranolytic complement membrane attack complex (MAC) has been activated. Consistent with this, MAC has been immunolocalized at all EPs encountered in intercostal muscle specimens of 30 MG patients.45,143 Activation of the lytic phase of the complement reaction sequence results in formation of transmembrane ion channels,15,59 uncontrolled influx of extracellular ions, focal calcium excess,23 protease activation, and destruction of cytoskeletal elements within the junctional folds.

LAMBERT-EATON MYASTHENIC SYNDROME In the Lambert-Eaton myasthenic syndrome (LEMS), pathogenic autoantibodies deplete the P/Q type of Cav of the motor nerve terminal by antigenic modulation. The deficiency of the Cav restricts the ingress of calcium into the nerve terminal during activity. This reduces the probability of quantal release, which decreases the number of quanta released by nerve impulse, and hence the safety margin of neuromuscular transmission.

Clinical Features The essential clinical and electromyographic (EMG) features of the syndrome, including its association with small cell carcinoma of the lung, were described in six patients by Lambert, Eaton, and Rooke112 in 1956 and by Eaton and Lambert39 in 1957. Subsequent investigations established that LEMS occurs more frequently in men than women41 and that a higher proportion of males than females have an associated malignancy, but malignancy is uncommon under age 40. Close to 80% of the associated tumors are small cell carcinomas of the lung.165 The symptoms of LEMS can antedate those of the malignancy by many months. However, only about 3% of patients with small cell carcinoma of the lung develop LEMS.43,92,114 The characteristic symptoms of LEMS consist of weakness and increased fatigability of the truncal and limb muscles. The proximal lower limb muscles are selectively and severely affected. Seventy percent of the patients have mild or transient ocular symptoms. Severe respiratory muscle weakness requiring mechanical ventilation is uncommon. Some patients complain of myalgias that involve chiefly the thigh muscles, or have peripheral paresthesias.6,28,82,115,165,178,186 Eighty percent have autonomic nervous system abnormalities. Decreased salivation, producing dryness of the mouth, is the most common autonomic symptom; decreased lacrimation and sweating, orthostatism, impotence, and abnormal pupillary light reflexes can also occur. AChE inhibitors produce only slight or no improvement, but curare sensitivity is increased.115,178 Strength is reduced in rested muscles but increases over a few seconds during the beginning of maximal voluntary contraction. This is best

FIGURE 33–1 AChR localization with peroxidase-labeled ␣-bungarotoxin in external intercostal muscles of control subject (A) and patients with moderately severe generalized myasthenia gravis (MG) (B–D). At the normal neuromuscular junction, AChR is localized on the terminal expansions of the junctional folds. At the MG junctions the postsynaptic regions are simplified. Degeneration of the junctional folds and widening of the synaptic space are evident in D. In B and C, only segments of the simplified postsynaptic membrane react for AChR. No AChR can be detected in D. Presynaptic staining on Schwann cell (arrow in A) and on nerve terminal (arrowhead in A) is caused by diffusion artifact. (A–D, ⫻22,300.) (From Engel, A. G., Lindstrom, J. M., Lambert, E. H., and Lennon, V. A.: Ultrastructural localization of the acetylcholine receptor in myasthenia gravis and in its experimental autoimmune model. Neurology 27:307, 1977, with permission.)

836

Pathology of the Peripheral Nervous System

FIGURE 33–2 Ultrastructural localization of immunoglobulin G (IgG) at a human myasthenic neuromuscular junction. IgG deposits occur on short segments of some junctional folds and on degenerating material in the synaptic space (arrow). Asterisk indicates degenerate material not reacting for IgG. Reciprocal staining of presynaptic membrane represents diffusion artifact. (⫻39,000.) (From Engel, A. G., Lambert, E. H., and Howard, F. M.: Immune complexes [IgG and C3] at the motor end-plate in myasthenia gravis: ultrastructural and light microscopic localization and electrophysiologic correlations. Mayo Clin. Proc. 52:267, 1977, with permission.)

Diseases of the Neuromuscular Junction

837

FIGURE 33–3 Ultrastructural localization of the lytic C9 complement component at a myasthenic neuromuscular junction. The junctional folds are degenerating on the right, and here globular material that has been shed into the synaptic space reacts vividly for C9. The most degenerate folds are not covered by nerve terminal (asterisk). On the left, the junctional folds are well preserved and only sparse C9-reactive material appears between them. Presynaptic staining (arrow) is caused by diffusion artifact. (⫻20,100.) (From Sahashi, K., Engel, A. G., Lambert, E. H., and Howard, F. M. Jr.: Ultrastructural localization of the terminal and lytic ninth complement component [C9] at the motor end-plate in myasthenia gravis. J. Neuropathol. Exp. Neurol. 39:160, 1980, with permission.)

observed by asking the patient to squeeze the examiner’s fingers or to continue to exert force with a proximal muscle, such as the biceps or iliopsoas, against resistance, or by repeated testing of the same muscle at short intervals. On continued maximal exertion, strength again decreases. The tendon reflexes are hypoactive or absent in nearly all patients but may briefly return to normal immediately after exercise.115,178,179 Approximately one third of LEMS cases are nonneoplastic, and these present at any age. Non-neoplastic LEMS can be associated with other autoimmune disease(s), such as pernicious anemia, hypothyroidism, hyperthyroidism, Sjögren’s syndrome, vitiligo, celiac disease, juvenile-onset diabetes mellitus, and MG.25,81,117,146,177 Subacute cerebellar degeneration has been observed with neoplastic28,186 and less often with non-neoplastic LEMS.178 In some patients with a paraneoplastic cerebellar syndrome, LEMS may be clinically occult.28 In whites with non-neoplastic LEMS, there is also an increased association with the HLA-B8, -DR3, and -DRQ2 immune response genes.168,215 In clinical EMG studies, the amplitude of the first-evoked compound muscle action potential (CMAP) is abnormally low because low quantal release by nerve impulse blocks neuromuscular transmission at many EPs. Repetitive stimulation of rested muscle at 2 to 3 Hz further decreases the CMAP amplitude, but recovery sets in after the fourth to

eighth stimulus. Voluntary effort or nerve stimulation at frequencies higher than 10 Hz enhances the calcium concentration in the nerve terminal. This facilitates quantal release, decreases the number of blocked EPs, and enhances the amplitude of the CMAP. Single-fiber EMG205 shows increased jitter and frequent blocking of neuromuscular transmission even in minimally affected muscles. The abnormalities are worst on minimal activity and improve at higher rates of firing or at high rates of repetitive axonal stimulation. The opposite occurs in MG, in which blocking increases with increasing neural activity.189,205

Pathogenesis and Pathology In 1981, Lang and co-workers117 obtained crucial evidence that LEMS is an autoimmune disease by showing that IgG transferred from patients to mice decreased the quantal content of the EPP in recipient animals. Evidence that the Cav of the nerve terminal represented the target of the LEMS antibodies soon followed. The Cav correspond to the active zone particles noted in freeze-fractured presynaptic membranes. In 1982, Fukunaga and co-workers68 found that, in freeze-fractured presynaptic membranes of LEMS patients, the number of particles per active zone as well as the number of active zones per unit area were markedly reduced, and the active zone particles that remained were

838

Pathology of the Peripheral Nervous System

FIGURE 33–4 Freeze-fractured presynaptic membrane from a normal subject (A) and from a Lambert-Eaton syndrome patient (B). A, Several active zones appear along an arc (arrow). Each active zone comprises double parallel rows of large membrane particles that represent presynaptic voltage-gated calcium channels. A cluster of large particles occurs near the end of an active zone (arrowhead). B, The entire presynaptic membrane contains only a short active zone (arrow). A single cluster of large particles is also present (arrowhead). In comparison to A, the Lambert-Eaton presynaptic membrane shows a marked depletion of active zones and active zone particles. (A, ⫻98,000; B, ⫻61,000.) (From Fukunaga, H., Engel, A. G., Osame, M., and Lambert, E. H.: Paucity and disorganization of presynaptic membrane active zones in the Lambert-Eaton myasthenic syndrome. Muscle Nerve 5:686, 1982, with permission.)

aggregated into clusters (Fig. 33–4). Aggregation of the active zone particles was attributed to the particles being cross-linked by IgG, and the disappearance of the particles was attributed to antigenic modulation by the pathogenic IgG. Subsequent studies confirmed each of these postulates. That the membrane lesions in LEMS are indeed mediated by pathogenic IgG was established by showing that injection of LEMS IgG into mice reproduces the membrane lesions observed in LEMS patients.67 Stereometric analysis of freeze-fracture replicas of normal motor nerve terminals indicated that the active zone particles are sufficiently close to be cross-linked by IgG, and that the aggregation of the active zone particles precedes their depletion.46,70 Divalent IgG is required for antigenic modulation, and divalency of LEMS IgG was shown to be essential for aggregation and depletion of the active zone particles.142 Finally, immunoelectron microscopy studies of mice injected with LEMS IgG46,69 confirmed binding of LEMS IgG to the active zones.

Other morphologic features of the LEMS EPs are also noteworthy. When visualized at the light microscopic level by the cholinesterase reaction, the LEMS EP has a normal shape. Quantitative electron microscopy shows that the postsynaptic region of clefts and folds is significantly larger than normal, and that it is made up of structurally normal junctional folds. Thus the LEMS and MG EPs display opposite morphologic features.57

Congenital Myasthenic Syndromes CMSs arise from defects in presynaptic, synaptic basal lamina, and postsynaptic proteins. Table 33–2 shows a current classification of the CMSs and indicates the number of kinships with each type of CMS investigated at the Mayo Clinic to date.

839

Diseases of the Neuromuscular Junction

Table 33–2. Classification of the CMSs Based on Site of the Defect* Index Cases Presynaptic Defects (7%) ChAT deficiency† Paucity of synaptic vesicles and reduced quantal release Lambert-Eaton syndrome like Other presynaptic defects Synaptic Basal Lamina–associated Defects (14%) End-plate AChE deficiency† Postsynaptic Defects (79%) Kinetic abnormality of AChR with/without AChR deficiency† AChR deficiency with/without minor kinetic abnormality† Rapsyn deficiency† Kinetic defect in Nav1.4† Plectin deficiency Total (100%)

7 1 1 4

25 45 83 17 1 1 185

AChE ⫽ acetylcholinesterase; AChR ⫽ acetylcholine receptor; ChAT ⫽ choline acetyltransferase. *Classification based on cohort of CMS patients investigated at Mayo Clinic between 1988 and 2003. † Gene defects identified.

A generic diagnosis of a CMS is often possible on the basis of myasthenic symptoms since birth or early childhood, a typical pattern of the distribution of weakness with involvement of ocular and other cranial muscles, a high-arched palate, a history of similarly affected relatives, a decremental EMG response, and negative tests for AChR, MuSK, and calcium channel antibodies. Some CMSs, however, are sporadic or present in later life, a decremental EMG response may not be present in all muscles or at all times, and the weakness may be restricted in distribution and not involve cranial muscles. In some CMSs, a specific diagnosis can be made on the basis of simple histologic or EMG studies. In other CMSs, in vitro electrophysiologic, ultrastructural, and immunocytochemical investigations are needed for accurate diagnosis (Table 33–3).

CHOLINE ACETYLTRANSFERASE DEFICIENCY Clinical Features The distinguishing clinical feature of ChAT deficiency is sudden episodes of severe dyspnea and bulbar weakness leading to apnea precipitated by infections, fever, or

Table 33–3. Investigation of Congenital Myasthenic Syndromes* Clinical Data History, examination, response to AChE inhibitor EMG: conventional needle EMG, repetitive stimulation, SFEMG Serologic tests for AChR, MuSK, and calcium channel antibodies, and tests for botulism Morphologic Studies Routine histochemical studies Cytochemical and immunocytochemical localization of AChE, AChR, agrin, ␤2-laminin, utrophin, and rapsyn at the end plate Estimate of the size, shape, and configuration of AChE-reactive end plates or end-plate regions on teased muscle fibers Quantitative electron microscopy; electron cytochemistry Endplate-Specific Sites

125

I-Labeled ␣-Bungarotoxin Binding

In Vitro Electrophysiology Studies Conventional microelectrode studies: MEPP, MEPC, evoked quantal release (m, n, p) Single-channel patch-clamp recordings: channel types and kinetics Molecular Genetic Studies Mutation analysis (if candidate gene or protein identified) Linkage analysis (if no candidate gene or protein recognized) Expression studies (if mutation identified) AChE ⫽ acetylcholinesterase; AChR ⫽ acetylcholine receptor; EMG ⫽ electromyography; m ⫽ number of ACh quanta released by nerve impulse; MEPC ⫽ miniature end-plate current; MEPP ⫽ miniature end-plate potential; n ⫽ number of readily releasable ACh quanta; p ⫽ probability of quantal release; SFEMG ⫽ single-fiber EMG. *Not all studies need to be performed in all CMSs.

excitement. In some patients the disease presents at birth with hypotonia, and severe bulbar and respiratory weakness requiring ventilatory support that gradually improves, but is followed by apneic attacks and bulbar paralysis in later life.54 Other patients first experience the typical attacks during infancy or early childhood. Variable ptosis and fatigable weakness may persist between the attacks (Fig. 33–5). The clinical features of this disorder were recognized four decades ago,77 and later the disease was dubbed “familial infantile myasthenia,” but it was not differentiated from MG until the autoimmune origin of MG was established and until electrophysiologic and morphologic differences were demonstrated between MG and the congenital syndrome.140 Because all CMSs can be familial and because most CMSs present in infancy, the term familial infantile myasthenia has become a source of confusion and should be avoided.33 In clinical electrophysiology studies, a decremental response at 2-Hz stimulation and single-fiber EMG abnormalities are detected only when the tested muscles are weak. Weakness and a decremental response at 2-Hz stimulation can be induced in some but not all muscles either

840

Pathology of the Peripheral Nervous System

CMSs the EPP amplitude returns to baseline in less than 2 minutes.

Molecular Pathogenesis

FIGURE 33–5 A 5-year-old boy with heterozygous mutations in choline acetyltransferase attempting to look up. The boy had had numerous apneic episodes since birth and had mild to moderately severe myasthenic symptoms between these episodes. Note ptosis, ophthalmoparesis, compensatory head tilt, facial diplegia, tracheostomy, and percutaneous gastrostomy.

by exercise or by a conditioning train of 10-Hz stimuli for 5 to 10 minutes.47,85,88,140,175 This finding, however, is not specific for CMS with episodic apnea (CMS-EA); it can also occur in some CMS patients with EP AChE or EP AChR deficiency.

End-Plate Studies The number of AChRs per EP, estimated from the number of 125I-labeled ␣-bungarotoxin binding sites, and postsynaptic ultrastructure are normal, but morphometric analysis indicates that the synaptic vesicles are smaller than normal in rested muscle.140 In vitro microelectrode studies of intercostal muscle specimens reveal that the amplitude of the MEPP is normal in rested muscle but decreases abnormally after 10-Hz stimulation for 5 minutes. The amplitude of the EPP also decreases abnormally during 10-Hz stimulation and then recovers slowly over the next 10 to 15 minutes (Fig. 33–6A) but the quantal content of the EPP is essentially unaltered.22,140 An abnormal decline of the EPP during 10-Hz stimulation can also occur in other CMSs, but in these

That the MEPP and EPP amplitudes decline abnormally when neuronal impulse flow is increased and then recover slowly pointed to a defect in resynthesis or vesicular packaging of ACh and implicated four candidate genes: the presynaptic high-affinity choline transporter,4,164 ChAT,145 VAChT,58 and the vesicular proton pump.174 Mutation analysis in five CMS-EA patients uncovered no mutations in VACHT but revealed 10 recessive mutations in CHAT160 (see Fig. 33–6B). One mutation (523insCC) was a null mutation; three others (I305T, R420C, and E441K) markedly reduced ChAT expression in COS cells. Kinetic studies of nine bacterially expressed and purified missense mutants revealed that one (E441K) lacked catalytic activity, and eight (L210P, P211A, I305T, R420C, R482G, S498L, V506L, and R560H) had catalytic efficiencies for coenzyme A (CoA) (kcat/KmAcCoA) ranging from 1% to 39% of wild type. The overall catalytic efficiency for both CoA and choline, kcat/(KmAcCoA ⭈ Kmchol), ranged from less than 1% to 69% of wild type160 (see Fig. 33–6C). Because each patient carried two different mutations of different biochemical pathogenicity, no clear genotype-phenotype correlation emerged except that, in the most severely affected patient, both mutations (I305T and R420C) markedly reduced both the expression and the overall catalytic efficiency of ChAT. Four additional CHAT mutations (L210P188 and V194L, S694L, S694C, and R548X129) were recently detected in CMS patients with unexpected episodes of sudden apnea. None of these mutations has been functionally characterized. That patients harboring CHAT mutations have no central or autonomic nervous system symptoms was unexpected. The simplest explanation for this puzzling observation would be that the identified mutations occur in a neuromuscular junction–specific isoform of ChAT. This, however, is unlikely; although there are five alternative CHAT transcripts with at least three different promoters in human,40 the observed mutations localize to the coding region shared by all of the recognized ChAT isoforms. Tissue-specific vulnerability, however, could be affected by promoters of differing strength producing different levels of ChAT expression in different tissues. An alternative explanation is that ChAT is rate limiting for ACh synthesis at the neuromuscular synapse under conditions of increased neuronal impulse flow owing to presynaptic levels of ChAT, substrate availability, or rates of choline uptake. Although the exact reason for the selective neuromuscular junction vulnerability is not yet known, it is important to recognize that CHAT mutations can cause a potentially lethal

Diseases of the Neuromuscular Junction

841

FIGURE 33–6 End-plate (EP) choline acetyltransferase deficiency. A, 10-Hz stimulation for 5 minutes results in a rapid abnormal decline of the EP potential, which then recovers slowly over more than 10 minutes. 3,4-Diaminopyridine (3,4-DAP), which accelerates ACh release, enhances the defect, whereas a low-Ca2⫹, high-Mg2⫹ solution, which reduces ACh release, prevents the abnormal decline of the EP potential. B, Genomic structure of CHAT and identified mutations. Note that the gene encoding the vesicular ACh transporter VACHT is located in the first CHAT intron. C, Individually scaled kinetic landscapes of wild-type ChAT and of the L210P and R560H ChAT mutants. The L210P mutant shows no saturation over a practical range of acetyl-CoA (AcCoA) concentrations, indicating an extremely high Km for AcCoA. Similarly, the R560H mutant does not saturate with increasing concentrations of choline, indicating a very high Km for choline. (Reproduced from Engel, A. G, Ohno, K., and Sine, S. M.: Congenital myasthenic syndromes: progress over the past decade. Muscle Nerve 27:4, 2003, with permission.)

form of CMS without central or autonomic nervous system involvement, and to search for such mutations when clinically appropriate. It is also important to note that defects in the presynaptic high-affinity choline transporter,4,164 VAChT,58 or the vesicular proton pump174 may have similar phenotypic consequences, but no mutations of these proteins in humans have been detected to date.

PAUCITY OF SYNAPTIC VESICLES AND REDUCED QUANTAL RELEASE Only one patient suffering from this disorder has been reported to date. This patient’s clinical and electromyographic features were indistinguishable from those of patients with autoimmune MG but the onset was at birth, anti-AChR antibodies were absent, there was no EP AChR

842

Pathology of the Peripheral Nervous System

deficiency, and electron microscopy revealed no postsynaptic abnormality. A presynaptic defect was indicated by a decrease to approximately 20% of normal of the number of ACh quanta (m) released by a nerve impulse. The decrease in m was due to a decrease in the number of readily releasable quanta (n), which was associated with a decrease in the numerical density of synaptic vesicles to approximately 20% of normal in unstimulated nerve terminals.208 The patient’s symptoms were improved by pyridostigmine. In this disorder, the clinical consequences stem from the paucity of synaptic vesicles in the nerve terminal. Synaptic vesicle precursors associated with different sets of synaptic vesicle proteins are produced in the perikaryon of the anterior horn cell and are carried distally along motor axons to the nerve terminal by kinesinlike motors.17,109,123,163 Mature vesicles containing a full complement of vesicular proteins are assembled in the nerve terminal163 and are then packed with ACh. After ACh has been released by exocytosis, the vesicle membranes are recycled and then repacked with ACh.203 In the present syndrome, the reduction in synaptic vesicle density could arise from (1) a defect in the formation of synaptic vesicle precursors in the anterior horn cell, (2) a defect in the axonal transport of one or more species of precursor vesicles, (3) impaired assembly of the mature synaptic vesicles from their precursors, or (4) impaired recycling of the synaptic vesicles in the nerve terminal. That synaptic vesicle density is reduced even in unstimulated nerve terminals argues against a defect in vesicle recycling.

CONGENITAL MYASTHENIC SYNDROME RESEMBLING LEMS In one young child reported with this syndrome in 1987, the CMAP amplitude was abnormally small but facilitated severalfold on tetanic stimulation, and the symptoms were improved by guanidine.5 A second patient observed at the Mayo Clinic was a 6-month-old girl with severe bulbar and limb weakness, hypotonia, areflexia, and respirator dependency since birth. The EMG showed a low-amplitude CMAP that facilitated 500% on high-frequency stimulation and decremented 40% on low-frequency stimulation. Studies of an anconeus muscle specimen revealed no EP AChR deficiency. Electron microscopy of the EPs showed structurally intact pre- and postsynaptic regions and abundant synaptic vesicles in the nerve terminals. The MEPP amplitude was normal for muscle fiber size. The quantal content of the EPP (m) at 1-Hz stimulation was less than 10% of normal, and 40-Hz stimulation increased m by 300%. Thus the in vitro electrophysiologic findings were remarkably similar to those of LEMS.113 Consistent with this, the EMG abnormalities were improved by

3,4-diaminopyridine (3,4-DAP), an agent that increases the number of ACh quanta released by nerve impulse,125 but the patient remained weak and respirator dependent. The molecular basis of this CMS could reside in an abnormality of the presynaptic Cav or in a component of the synaptic vesicle release complex. Mutation analysis of CACNA1A, the gene encoding the pore-forming ␣1 subunit of the Cav2.1, or P/Q type, calcium channel expressed at the presynaptic membrane revealed no mutations (K. Ohno and A. G. Engel, unpublished data, 1999).

CMS WITH REDUCED QUANTAL RELEASE AND CENTRAL NERVOUS SYSTEM SYMPTOMS Maselli and co-workers130 reported three sporadic patients, two presenting in early infancy and one after the age of 5 years, with myasthenic symptoms that spared the external ocular muscles. All three had other neurologic symptoms that included truncal or limb ataxia and, in one case, horizontal nystagmus. Unlike in LEMS patients, none had an abnormally small first-evoked CMAP, and their decremental response on 2-Hz stimulation was not improved by stimulation at higher frequencies. None had EP AChR deficiency. Electron microscopy demonstrated nerve terminals of normal size containing a normal number of synaptic vesicles and displaying small doublemembrane-lined saccules filled with vesicles. In vitro microelectrode studies revealed marked decrease in the number of quanta (m) released by 1-Hz stimulation. In one patient, this was associated with a significantly decreased probability of quantal release (p); in another patient, there was a less than 50%, but still significant, decrease in the number of readily releasable quanta (n). A search for mutations in selected exons of CACNA1A was negative. One of the three patients responded well to combined treatment with pyridostigmine and 3,4-DAP, one showed only mild improvement in response to combined therapy with pyridostigmine and ephedrine, and one failed to respond adequately to pyridostigmine.

END-PLATE ACETYLCHOLINESTERASE DEFICIENCY Clinical Features This CMS is caused by the absence of AChE from the synaptic space basal lamina.48,100,156 In most patients the disease presents in the neonatal period and is highly disabling (Fig. 33–7), but in one kinship with partial AChE deficiency, it presented after the age of 6 years and became disabling only during the second decade of life.34 The cardinal clinical features are (1) a decremental EMG response

843

Diseases of the Neuromuscular Junction

FIGURE 33–7 Patient suffering from EP AChE deficiency at 4 months (left) and 8 years (right) of age. Left, Note hypotonia, ptosis, facial diplegia, open mouth, and nasogastric feeding tube. Right, Note Gowers sign and percutaneous gastrostomy.

(Fig. 33–8A); (2) a repetitive CMAP elicited from rested muscle by single-nerve stimuli (see Fig. 33–8A); (3) no effect of AChE inhibitors on the decremental response or the repetitive CMAP, and failure of AChE inhibitors to improve the clinical state; and (4) a slow pupillary light response in most adult cases. The diagnosis is confirmed by demonstrating absence of AChE from the EP (Fig. 33–9) or a pathogenic mutation in COLQ, the gene encoding the collagenic tail subunit of EP AChE.

Pathogenesis and Pathology AChE is the enzyme responsible for rapid hydrolysis of ACh released at cholinergic synapses. At the normal EP, AChE limits the number of collisions between ACh and AChR and, hence, the duration of the synaptic response.107 When the enzyme is inhibited or absent, each ACh molecule binds repeatedly to AChR until it leaves the synaptic space by diffusion. This prolongs the duration of the synaptic potentials and currents (see Fig. 33–8B). When the prolonged EPP exceeds the absolute refractory period of the muscle fiber surface membrane, it can evoke a second action potential detected by surface electrodes as a repetitive CMAP. The prolonged duration of the synaptic response results in cationic overloading of the postsynaptic region. Electron microscopy studies of the EP reveal abnormally small nerve terminals, frequently partially or totally isolated from the postsynaptic region by Schwann cell processes that extend into the synaptic cleft (Fig. 33–10). Smallness of the nerve terminals and their encasement by Schwann cells restrict the number of quanta that can be released by nerve impulse. This mitigates postsynaptic injury resulting from

overstimulation by unhydrolyzed ACh, but decreases the amplitude of the EPP. Despite this protective mechanism, cationic overloading of the postsynaptic region still causes an EP myopathy characterized by focal degeneration of the junctional folds with loss of AChR, and appearance of degenerating organelles and apoptotic nuclei in the junctional sarcoplasm.48,100 The safety margin of neuromuscular transmission is compromised by the decreased quantal content of the EPP, the EP myopathy, desensitization of AChR by prolonged exposure to ACh, and a depolarization block at physiologic rates of stimulation.

Molecular Pathogenesis The EP species of AChE is a heteromeric asymmetric enzyme composed of one, two, or three homotetramers of globular catalytic subunits (AChET) attached to a triplestranded collagenic tail (ColQ) (Fig. 33–11B). ColQ has an N-terminal proline-rich region attachment domain (PRAD), a collagenic central domain, and a C-terminal region enriched in charged residues and cysteines (Fig. 33–11A). Each ColQ strand can bind an AChET tetramer to its PRAD, giving rise to A4, A8, and A12 species of asymmetric AChE.16 Two groups of charged residues in the collagen domain (heparan sulfate proteoglycan binding domains)32 plus other residues in the C-terminal region110,153 assure that the asymmetric enzyme is inserted into the synaptic basal lamina. The C-terminal region is also required for initiating the triple-helical assembly of ColQ that proceeds from a C- to an N-terminal direction in a zipper-like manner.170 Expression of globular and asymmetric forms of AChE in muscle, or in COS cells

844

Pathology of the Peripheral Nervous System

FIGURE 33–8 End-plate AChE deficiency. A, Decremental electromyographic response and repetitive compound muscle action potential recorded from thenar muscle of patient during 2-Hz stimulation of the median nerve. At this rate of stimulation, the second response decrements more rapidly than the first and appears only once. B, Representative miniature EP currents (MEPCs) from a patient with EP AChE deficiency and a control subject. The best-fit exponential curve is superimposed on the decay phase of each current. Arrows indicate decay time constants. The MEPC is smaller and decays more slowly in the patient than in the control. (From Hutchinson, D. O., Walls, T. J., Nakano, S., et al.: Congenital endplate acetylcholinesterase deficiency. Brain 116:633, 1993, with permission.)

transfected with ACHET and COLQ complementary DNA (cDNA), is readily monitored by density gradient centrifugation of muscle or COS cell extracts (Fig. 33–11C and D). In 1998, human COLQ cDNA was cloned,34,151 the genomic structure of COLQ determined,151 and the molecular basis of EP AChE deficiency traced to recessive mutations in COLQ.34,151 Twenty-four COLQ mutations in 25 kinships have been identified to date34,101,150,151,153,191 (see Fig. 33–11B). The mutations are of three major types: (1) PRAD mutations preventing attachment of AChET to ColQ (Fig. 33–11E and I); (2) collagen domain mutations producing a short, single-stranded ColQ that binds a single AChET tetramer and is insertion incompetent (Fig. 33–11F and J); and (3) C-terminal mutations hindering the triple helical assembly of the collagen domain or producing an asymmetric species of AChE that is insertion incompetent, or both (Fig. 33–11H and K).

FIGURE 33–9 Electron cytochemical localization of AChE at a control EP (A) and at an AChE-deficient patient EP (B). At the control EP, heavy reaction product fills the synaptic space and extends into the adjacent regions. At the patient EP there is no reaction product for AChE in the synaptic space. (A, ⫻20,000; B, ⫻9000.)

POSTSYNAPTIC CMSs Most postsynaptic CMSs identified to date stem from a defect in the level of expression, kinetic properties, or aggregation of AChR. We therefore begin with a brief review of the structure and kinetic properties of the muscle species of AChR. Muscle AChR is a transmembrane macromolecule composed of five homologous subunits: two of ␣, one of ␤ and ␦, and one of ⑀ in adult AChR, or one of ␥ instead of ⑀ in fetal AChR. The genes coding for ␣, ␦, and ␥ subunits are at different loci on chromosome 2q, and those coding for ␤ and ⑀ subunits are at different loci on chromosome 17p. The subunits are highly homologous, have similar secondary structures, fold similarly, and are organized like barrel staves around a central cation channel. Each subunit has an Nterminal extracellular domain that comprises approximately

845

Diseases of the Neuromuscular Junction

FIGURE 33–10 End-plate AChE deficiency. A small nerve terminal completely encased by Schwann cell is applied against a small fraction of a postsynaptic region. The junctional folds are decorated with labyrinthine membranous networks. Mitochondria are absent, and a small focus of myofibrillar degeneration appears under the junctional folds (asterisk). (⫻25,900.)

50% of the primary sequence, four putative transmembrane domains (TMD1 through TMD4), and a small C-terminal extracellular domain. TMD2, which lines the ion channel, forms an ␣ helix interrupted by a short stretch of ␤ sheet. The TMDs are connected by an extracellular TMD2/TMD3 linker and by intracellular TMD1/TMD2 and TMD3/ TMD4 linkers. The TMD3/TMD4 linker forms a long cytoplasmic loop that serves as an attachment site for rapsyn, which links AChR to the subsynaptic cytoskeleton. The long cytoplasmic loops also bear phosphorylatable residues that may be important for desensitization, and a segment of the long cytoplasmic loop of the ⑀ subunit stabilizes the gating mechanism. Each AChR has two ACh binding pockets, one at the ␣/⑀ (or ␣/␥) and one at the ␣/␦ interface. Residues contributing to the binding pocket appear on three peptide loops on the ␣ subunit and on four peptide loops on the ⑀, ␦, and ␥ subunits. The recent x-ray analysis of the structure of the snail glial ACh binding protein confirmed that con-

served residues in all seven loops are present at the ligand binding sites.18 Activation of AChR by ACh can be described by a linear scheme that accounts for essential features of AChR activation in a variety of species197,217:

A+R

k+1 k–1

AR + A

k+2 k–2

A 2R

β α

A 2R*

where A is agonist, R is the resting state, R* is the openchannel state, ␤ the rate of channel opening, ␣ the rate of channel closure, and k⫹ and k⫺ the forward binding and dissociation rates of ACh. The dissociation constants of ACh from monoliganded and diliganded AChR, K1 and K2, obtained from the k⫺1/k⫹1 and k⫺2/k⫹2 ratios, are measures of the affinity of the monoliganded and diliganded receptor for ACh. The above scheme represents a

846

Pathology of the Peripheral Nervous System

FIGURE 33–11 A, Schematic diagram showing domains of a ColQ strand. B, Schematic diagram showing the A1 through A12 species of asymmetric AChE with 24 identified ColQ mutations. C–H, Density gradient profiles of AChE extracted from COS cells transfected with wild-type ACHET and different types of COLQ mutants. I–K, Schematic diagrams of the abnormal species of AChE formed after the transfections. In E and I, note that disruption of the PRAD domain produces a sedimentation profile identical to that obtained after transfection with ACHET alone. Thus ACHET fails to attach to ColQ and no asymmetric AChE is formed. In F and J, note that the asymmetric A4, A8, or A12 moieties are absent and there is a prominent 10.5-S mutant (M) peak, representing a G4 tetramer of the catalytic subunit linked to the truncated ColQ peptide. In G and the left diagram in K, note the presence of a small M peak but absence of peaks corresponding to triple-stranded asymmetric enzymes. In H and the right diagram in K, note that both an M peak and asymmetric AChE are present. HSPBD ⫽ heparan sulfate proteoglycan binding domain; PRAD ⫽ proline-rich attachment domain.

847

Diseases of the Neuromuscular Junction

Table 33–4. Kinetic Abnormalities of AChR Slow-Channel Syndromes

Fast-Channel Syndromes

End-plate currents Channel opening events Open states Closed states Mechanisms*

Slow decay Prolonged Stabilized Destabilized Increased affinity Increased ␤ Decreased ␣

Pathology Genetic background Response to therapy

Endplate myopathy from cationic overloading Dominant gain-of-function mutations Long-lived open channel blockade of AChR with quinidine or fluoxetine

Fast decay Brief Destabilized Stabilized Decreased affinity Decreased ␤ Increased ␣ Mode-switching kinetics No anatomic footprint Recessive loss-of-function mutations 3,4-DAP and AChE inhibitors

␣ ⫽ channel closing rate; ␤ ⫽ channel opening rate. *Different combinations of mechanisms operate in the individual slow- and fast-channel syndromes.

simplification because the unliganded and monoliganded receptors can open, the nonconducting blocked and desensitized states are not represented, and transitions can occur between all adjacent states. A more complete description of AChR activation is provided by the Monod-Wyman-Changeux description of allosteric protein function139:

R* + A

K1 *

AR* + A

θ1

θ0 R+A

K2 *

A2R* θ2 A2R

AR + A K1

affected patients, the cranial muscles tend to be spared. The weakness and fatigability can fluctuate, but not as rapidly as in autoimmune MG. The tendon reflexes are usually normal but can be reduced in severely affected limbs. The more severely affected muscles become atrophic. Progressive spinal deformities and respiratory embarrassment are common complications during the evolution of the illness.

K2

To date, two major kinetic abnormalities of AChR have emerged that give rise to the slow-channel and the fast-channel CMSs. The two syndromes are physiologic opposites. Table 33–4 summarizes the divergent properties of the two syndromes.

THE SLOW-CHANNEL CMSs Clinical Features The slow-channel syndromes are caused by dominant gain-of-function mutations. The clinical phenotypes vary. Some slow-channel CMSs present in early life and cause severe disability by the end of the first decade137; others present later in life and progress slowly, resulting in little disability even in the sixth or seventh decade.50,54,198 Most patients show selectively severe involvement of cervical and of wrist and finger extensor muscles (Fig. 33–12). Except for the more severely

FIGURE 33–12 Slow-channel syndrome patient is attempting to extend wrists and fingers as shown by examiner (with sleeve). Note atrophy of patient’s forearm muscles. (From Engel, A. G., Lambert, E. H., Mulder, D. M., et al.: A newly recognized congenital myasthenic syndrome attributed to a prolonged open time of the acetylcholine-induced ion channel. Ann Neurol 11:553, 1982, with permission.)

848

Pathology of the Peripheral Nervous System

FIGURE 33–13 Slow-channel syndrome. Repetitive and decrementing compound muscle action potential evoked from a limb muscle by a single nerve stimulus. The second and third responses are triggered by the prolonged EPP that outlasts the absolute refractory period of the muscle fiber. The second and third responses decrement faster than the first response.

End-Plate Studies As in EP AChE deficiency, single-nerve stimuli elicit one or more repetitive CMAPs (Fig. 33–13) but, unlike in EP AChE deficiency, AChE inhibitors increase the number of repetitive responses. Repetitive nerve stimulation reveals a

decremental response that is present at low stimulation frequency (Fig. 33–13) and increases progressively when stimulation frequency is increased. The repetitive CMAPs are of lower amplitude and decrement faster than the first CMAP. In vitro microelectrode studies demonstrate markedly prolonged EPPs and currents that decay biexponentially53,155,198 (Fig. 33–14C). Single-channel patch-clamp recordings reveal a dual population of AChR channels, one with normal and one with prolonged opening episodes, reflecting the presence of both wild-type and mutant receptors at the EP53,137,155 (Fig. 33–14B). In addition, the mutant AChR channels open even in the absence of ACh,137,155 resulting in a continuous cation leak into the postsynaptic region. The morphologic consequences stem from prolonged activation episodes of the AChR channel that cause cationic overloading of the postsynaptic region. Excessive accumulation of Ca2⫹ can be demonstrated at some EPs with alizarin red, which detects millimolar concentrations of Ca2⫹. The EP myopathy that develops is like that in EP AChE deficiency but is even more severe, sometimes causing massive destruction of the junctional folds (Fig. 33–15), nuclear

FIGURE 33–14 A, Schematic diagram of AChR subunits with slow-channel mutation. The mutations occur in the extracellular and transmembrane domains of the different subunits. B, Examples of single channel currents from wild-type and slow-channel (␣V249F) AChRs expressed in HEK cells. C, Miniature EP currents (MEPCs) recorded from EPs of a control subject and a patient harboring the ␣V249F slow-channel mutation. The slow-channel MEPC decays biexponentially as a result of expression of both wild-type and mutant AChRs at the EP, so one decay time constant is normal and the other is markedly prolonged.

Diseases of the Neuromuscular Junction

849

FIGURE 33–15 Slow-channel syndrome EP. Note degeneration of junctional folds and accumulation of degenerate material in the widened synaptic space. (⫻25,900.)

apoptosis (Fig. 33–16), and vacuolar degeneration near the EPs50,53,54,137 (Fig. 33–17).

Molecular Pathogenesis The abnormal kinetic properties of AChR predicted that the slow-channel syndrome stemmed from mutations in AChR subunits. Since 1995, 18 slow-channel mutations have been uncovered.29,53,74–76,91,137,156,162,192,198,209 The different mutations occur in different AChR subunits and in different functional domains of the subunits (see Fig. 33–14A). In the kinships observed to date, mutations in the channel domain had more severe phenotypic consequences than those at the ACh binding site, but there were also variations in phenotypic expressivity between and within kinships harboring the same mutation. Thus, although suggestive, phenotypic severity is not a fully reliable predictor of mutation site.

Patch-clamp studies at the EP, mutation analysis, and expression studies in human embryonic kidney (HEK) cells indicate that the ␣G153S mutation near the extracellular ACh binding site198 and the ␣N217K209 and ⑀L221F91 mutations in the N-terminal part of TMD1 act mainly by enhancing affinity for ACh. This slows dissociation of ACh from the binding site and results in repeated channel reopenings during the prolonged receptor occupancy. Another slow-channel mutation near the binding site region, ␣V156M, probably has similar effects, although its mechanistic consequences have not been investigated.29 Mutations in TMD2 (which lines the channel pore), such as ␤V266M, ⑀L269F, ⑀T264P, and ␣V249F, promote the open state by affecting channel opening and closing steps.53,137,155 Some TMD2 mutations also increase affinity for ACh; this is most marked in the case of ␣V249F,137 pronounced with ⑀L269F53 and ⑀T264P,155 and not apparent with

FIGURE 33–16 Slow-channel syndrome. Apoptotic nucleus appears near a degenerating postsynaptic region. (⫻15,500.)

FIGURE 33–17 Slow-channel syndrome. Autophagic vacuoles (V) containing degraded material near a postsynaptic region (X) abandoned by its nerve terminal. (⫻11,900.)

851

Diseases of the Neuromuscular Junction

␤V266M.53 These TMD mutations likely exert their effects through structures that couple the binding site to the channel gate. The safety margin of neuromuscular transmission is compromised by the altered EP geometry, the loss of AChR from degenerating junctional folds, and a depolarization block during physiologic activity owing to staircase summation of the markedly prolonged EPPs.

FAST-CHANNEL SYNDROMES Clinical Features The fast-channel mutations derange one or more of the following functions of AChR: affinity for ACh, efficiency of gating, and stabilization of channel kinetics. The clinical features resemble those of autoimmune MG, but symptoms are mild when the main effect is on gating efficiency,210 moderately severe when channel kinetics are unstable,138,211 and severe when affinity for ACh, or both affinity and gating efficiency, are impaired.161,195

End-Plate Studies The low-affinity fast-channel syndromes leave no anatomic footprint161,195: the structural integrity of the postsynaptic region is maintained, and the density and distribution of AChR on the junctional folds are normal. In the CMS that only affects gating efficiency210 and in the CMS with unstable channel kinetics,138 the number of AChRs per EP is also reduced. In these CMSs, multiple small EP regions are dispersed over an extended length of the fiber surface, some postsynaptic regions are simplified, and the expression of AChR on the junctional folds is patchy and attenuated. The common electrophysiologic features of the fast-channel CMSs are rapidly decaying EP currents (Fig. 33–18A), abnormally brief channel activation episodes (Fig. 33–18B), and a reduced quantal response owing to a decreased probability of channel opening (see Table 33–4).

Molecular Pathogenesis The fast-channel CMSs are caused by recessive, loss-offunction mutations. Several fast-channel mutations have

FIGURE 33–18 A, Miniature EP currents (MEPCs) recorded from EPs of a control subject and a patient harboring the ␣V285I fastchannel mutation. Arrows indicate decay time constants. B, Examples of single channel currents from wild-type and fast-channel (␣V285I) AChRs expressed in HEK cells. C, Schematic diagram of fast-channel mutations in the AChR ␣, ␤, and ␦ subunits. (From Engel, A. G., Ohno, K., and Sine, S. M.: Congenital myasthenic syndromes: progress over the past decade. Muscle Nerve 27:4, 2003, with permission.)

852

Pathology of the Peripheral Nervous System

been reported to date (see Fig. 33–18C). In each instance, the mutated allele causing the kinetic abnormality is accompanied by a null mutation in the second allele; therefore, the kinetic mutation dominates the clinical phenotype. Low-Affinity Fast-Channel Syndromes Four identified kinetic mutations fall in this group. Two different substitutions of residue 121 in the extracellular domain of the ⑀ subunit, ⑀P121L161 and ⑀P121T,194 both result in abnormally brief channel events, reduce the amplitude of the quantal response by decreasing the probability of channel openings, decrease the number of channel reopenings in bursts of openings, and decrease the affinity for ACh in the open channel state. The ⑀P121L mutation also reduces gating efficiency by reducing ␤, the channel opening rate. In contrast, the ⑀P121T mutation does not alter gating efficiency, but reduces closed-state affinity for ACh and thereby stabilizes the closed state. The third mutation in this group occurs in the ␣ subunit, in the most highly conserved domain of the AChR superfamily, the disulfide-bridged ␤-hairpin formed between cysteine 128 and cysteine 142, or the Cys-loop. The mutation consists of replacement of a conserved valine at position 132 by a leucine (␣V132L). The ␣V132L mutation markedly reduces closed-state affinity for ACh, and impairs gating efficiency by slowing ␤, the rate of channel opening, and speeding ␣, the rate of channel closing. The duration of channel opening events is only 15% of normal, and the amplitude of quantal response only 10% of normal.195 The fourth mutation in this group, ␦E59K, is in the extracellular domain of the ␦ subunit. It is of special interest because it causes multiple congenital joint contractures owing to fetal hypomotility in utero.20

the EP, ⑀1254ins18-mutated AChR shows abnormally brief activation episodes during steady-state ACh application. When expressed in HEK cells and exposed to desensitizing concentrations of ACh, openings of individual receptors occur in clusters separated by silent periods during which all receptors are desensitized. For the normal receptor, the kinetics of gating within a cluster is essentially uniform and highly efficient. By contrast, the kinetic behavior of the ⑀1254ins18-mutated AChR changes abruptly, so that a normal mode of activation is replaced by three abnormal and inefficient modes in which the receptor opens slowly and closes more rapidly than normal. In this disorder, the reduced gating efficiency and decreased AChR expression are partially offset by expression of fetal AChR harboring the ␥ instead of the ⑀ subunit (␥-AChR); this improves electrical activity at EP and likely rescues the phenotype.138 The second mutation that destabilizes channel kinetics is a nearby missense mutation in the ⑀ subunit, ⑀A411P. When this mutation is expressed in HEK cells, different clusters of channel openings differ widely in their activation kinetics, so that the spread in the distribution of the channel opening and closing rates is greatly expanded.211 Intracellular microelectrode and patch-clamp studies on EPs of patients harboring this mutation are still unavailable. That both ⑀1254ins18 and ⑀A411P occur in the amphipathic helix region of the long cytoplasmic loops of the ⑀ subunit implicates this region of the ⑀ subunit in stabilization of channel kinetics.

MUTATIONS CAUSING AChR DEFICIENCY WITH OR WITHOUT MINOR KINETIC ABNORMALITY Clinical Features

Fast-Channel Syndrome Caused by a Selective Abnormality of Gating Replacement of a valine by an isoleucine at residue 285 in TMD3 of the ␣ subunit (␣V285I) selectively reduces gating efficiency by depressing the channel opening rate (␤) and enhancing the channel closing rate (␣). The ␣V285I mutation also decreases AChR expression, which further impairs the safety margin of neuromuscular transmission.210 Fast-Channel Syndromes Caused by Unstable (Mode-Switching) Kinetics Two mutations causing unstable channel kinetics have been identified in expression studies, but only one of these, ⑀1254ins18, was investigated by in vitro microelectrode studies of the EP. The ⑀1254ins18 mutation, which is an in-frame duplication of codons 413 to 418 (STRDQE) in the long cytoplasmic loop of ⑀, also reduces AChR expression at the EP.138 At

The clinical phenotypes with these CMSs vary from mild to severe. Patients with recessive mutations in the ⑀ subunit are generally less affected than those with mutations in other subunits, but some harboring ⑀ subunit mutations can also be severely affected. The sickest patients have severe ocular, bulbar, and respiratory muscle weakness from birth and survive only with respiratory support and gavage feeding. They may be weaned from a respirator and begin to tolerate oral feedings during the first year of life, but have bouts of aspiration pneumonia and may need intermittent respiratory support during childhood and adult life. Motor milestones are severely delayed; they seldom learn to negotiate steps and can walk for only a short distance. Older patients close their mouth by supporting the jaw with the hand and elevate their eyelids with their fingers (Fig. 33–19). Facial deformities, prognathism, malocclusion, and scoliosis or kyphoscoliosis become noticeable during the second

853

Diseases of the Neuromuscular Junction

patchy (Fig. 33–20). The integrity of the junctional folds is preserved, but some EP regions are simplified and smaller than normal (Fig. 33–20A). The quantal response at the EP, indicated by the amplitude of MEPPs and MEPCs, is reduced, but quantal release by nerve impulse is frequently higher than normal. In patients with low-expressor or null mutations of the ⑀ subunit, single-channel patch-clamp recordings136,157,158 and immunocytochemical studies52 reveal the presence of fetal ␥-AChR at the EP.

Molecular Pathogenesis CMSs with severe EP AChR deficiency result from different types of homozygous or, more frequently, heterozygous recessive mutations in AChR subunit genes. The mutations are concentrated in the ⑀ subunit (Fig. 33–21). There are two possible reasons for this:

FIGURE 33–19 A severely affected 4-year-old girl who carries two heteroallelic mutations in the AChR ␤ subunit: a 9–base pair deletion in the long cytoplasmic loop (␤1276del9), which curtails AChR expression, and skipping of exon 8, which abolishes AChR expression. Note ptosis, exotropia, facial diplegia, and open mouth as well as tracheostomy and gastrostomy.

decade. Muscle bulk is reduced. The tendon reflexes are normal or hypoactive. The least affected patients pass their motor milestones with slight or no delay and only show mild ptosis and limited ocular ductions. They are clumsy in sports, fatigue easily, and cannot run well, climb rope, or do push-ups. In some instances, a myasthenic disorder is suspected only when the patient develops prolonged respiratory arrest on exposure to a curariform drug during a surgical procedure. Patients with intermediate clinical phenotypes experience moderate physical handicaps from early childhood. Ocular palsies and ptosis of the lids become apparent during the first year of life. These patients fatigue easily and cannot keep up with their peers in sports; they walk and negotiate stairs with difficulty, but can perform most activities of daily living.

End-Plate Studies Morphologic studies show an increased number of EP regions distributed over an increased span of the muscle fiber. AChR expression at the EP is markedly attenuated and

1. Expression of the fetal-type ␥ subunit, although at a low level, may compensate for absence of the ⑀ subunit,52,138,158 whereas patients harboring null mutations in subunits other than ⑀ might not survive for lack of a substituting subunit. 2. The gene encoding the ⑀ subunit, and especially exons coding for the long cytoplasmic loop, have a high GC content that likely predisposes to DNA rearrangements. Different types of recessive mutations causing severe EP AChR deficiency have been identified. Some mutations cause premature termination of the translational chain. These mutations are frameshifting,1,30,52,134,149,158,193,195 occur at a splice site,134,149 or produce a stop codon directly.159 An important mutation in this group is the 1369delG in the ⑀ subunit that results in loss of a C-terminal cysteine, C470, crucial to both maturation and surface expression of the adult receptor.38 Thus any mutation that truncates the ⑀ subunit upstream of C470 is predicted to inhibit ⑀ expression. A second type of recessive mutations are point mutations in the Ets binding site, or N-box, of the promoter region of the ⑀ subunit gene: ⑀-154G⬎A,2 ⑀-155G⬎A,147 and ⑀-156C⬎T.144 The N-box represents the end point of a signaling cascade driven by neuregulin through ERBB receptors. ERBB receptors phosphorylate mitogen-activated protein (MAP) kinases. Phosphorylated MAP kinases phosphorylate GABP␣ and GABP␤ (members of the Ets family of transcription factors), which bind to the N-box.36,65,187 That these mutations impair AChR expression is strong direct evidence that the neuregulin signaling pathway regulates synapse-specific transcription at the human neuromuscular junction. In addition to the above mutations, there are also missense mutations in a signal peptide region (⑀G-8R161 and ⑀V-13D134), and missense mutations involving residues essential for assembly of the pentameric receptor. Mutations of the latter type were observed in the

854

Pathology of the Peripheral Nervous System

FIGURE 33–20 Ultrastructural localization of AChR with peroxidase-labeled ␣-bungarotoxin at EP of patient homozygous for the ⑀553del7 null mutation (A) and at a control EP (B). The control EP shows heavy reaction for AChR on the terminal expansions of the junctional folds. At the patient’s EP, the junctional folds are simplified and the reaction for AChR is attenuated and patchy. (A, ⫻25,900; B, ⫻30,000.)

⑀ subunit at an N-glycosylation site (⑀S143L)161; in Cys 128 (⑀C128S), a residue that is an essential part of the C128-C142 disulfide loop in the extracellular domain138; in arginine 147 (⑀R147L), which is part of a short extracellular span of residues that contributes to subunit assembly158; in Thr 51 (⑀T51P)134; and in the long cytoplasmic loop of the ␤ subunit, causing the deletion of three codons.171 Finally, another group of missense mutations lead to the production of channels with minor kinetic effects. The kinetic consequences of these mutations are modest and overshadowed by reduced expression of the mutant gene. For example, ⑀R311W in the long cytoplasmic

loop between TMD3 and TMD4, and ⑀P245L in TMD1, decrease EP AChR below 10% of normal; ⑀R311W also reduces the open duration of channel events threefold158 and has thus mild fast-channel properties, whereas ⑀P245L increases the open duration of channel events threefold and has mild, nondominant slow-channel properties.158 Mutation of a corresponding proline in TMD1 of the ␦ subunit (␦P250Q) also markedly reduces AChR expression, but decreases rather than increases the duration of channel opening events by about twofold.193 It is also noteworthy that homozygous low-expressor ⑀ subunit mutations are endemic in Mediterranean or other

Diseases of the Neuromuscular Junction

855

FIGURE 33–21 Schematic diagram of low-expressor and null mutations reported in the ␣, ␤, ␦, and ⑀ subunits of AChR. Note that most mutations appear in the ⑀ subunit and especially in the long cytoplasmic loop between the third and fourth transmembrane domains (M). Square indicates a chromosomal microdeletion, hexagons are promoter mutations, open circles are missense mutations, closed circles are nonsense mutations, shaded circles are frameshifting mutations, and dotted circles are splicesite mutations. The most likely consequence of a splice-site mutation is skipping of a flanking exon; therefore, the splicesite mutations point to N-terminal codons of the predicted skipped exons. (From Engel, A. G., Ohno, K., and Sine, S. M.: Congenital myasthenic syndromes: progress over the past decade. Muscle Nerve 27:4, 2003, with permission.)

Near Eastern countries,134,148 and that the frameshifting ⑀1267delG mutation occurring at homozygosity is endemic in Gypsy families1,30,149 in whom it derives from a common founder.1 For a full list of AChR subunit gene mutations and the appropriate references, the reader is referred to a recently published gene table.152

CMS CAUSED BY RAPSYN DEFICIENCY In a subset of CMS patients with EP AChR deficiency but no mutation in any of the AChR subunits, the disease is traced to mutations in rapsyn. Rapsyn, under the influence of neurally supplied agrin, has a crucial role in concentrating the AChR in the postsynaptic membrane64

856

Pathology of the Peripheral Nervous System

and linking it to the subsynaptic cytoskeleton through dystroglycan.26 Rapsyn knockout mice cluster AChR poorly and fail to accumulate AChR at the EP.72 Thus rapsyn is an effector of agrin-induced clustering of AChR. In myotubes, agrin, MuSK, and possibly other myotube-specific mechanisms regulate rapsyn aggregation,213 but rapsyn expressed in heterologous systems self-aggregates and can then recruit AChRs, dystroglycan, and MuSK. The primary structure of rapsyn predicts distinct structural domains: a myristoylation signal at the N-terminus required for membrane association173; seven tetratricopeptide repeats (TPRs; codons 6 to 279) that subserve rapsyn self-association172,173; a coiled-coil domain (codons 298 to 331), the hydrophobic surface of which can bind to determinants within the long cytoplasmic loop of each AChR subunit10,172; a Cys-rich RING-H2 domain (codons 363 to 402) that binds to the cytoplasmic domains of ␤-dystroglycan11 and mediates the MuSK-induced phosphorylation of AChR118; and a serine phosphorylation site at codon 406 (Fig. 33–22A). Transcription of rapsyn in muscle is under the control of helix-loop-helix myogenic determination factors that bind to the cis-acting E-box sequence in the RAPSN promoter.159

Clinical Features Myasthenic symptoms can present in the neonatal period or later in life. Eight of 13 patients in one series21 and 1 of 4 patients in another series154 were born with arthro-

gryposis owing to hypomotility in utero. Other dysmorphic features are also common.7,21,154 Motor milestones are typically delayed, and fatigable weakness persists during life. Respiratory infections or other intercurrent illnesses precipitate increased weakness and respiratory insufficiency. A severely affected patient had episodes of respiratory compromise during infancy resulting in anoxic encephalopathy.7 Three patients were also reported in whom the disease was first noted at 13, 21, and 48 years of age. In the late-onset cases, the clinical course was mild and the disease mimicked seronegative autoimmune MG.21 Facial deformities associated with prognathism and malocclusion are pronounced in Near-Eastern Jewish patients who carry an E-box mutation (⫺38A⬎G) in the RAPSN gene159 but also can occur in other cases of rapsyn deficiency.8 The Near-Eastern Jewish patients have mild to severe weakness of the masticatory muscles, moderate to severe eyelid ptosis without ophthalmoparesis, facial weakness, and slurred or hypernasal speech. Cervical, truncal, and limb muscles are usually spared. Although these patients present at birth or soon thereafter, they show no signs of arthrogryposis. They have a stable and essentially benign course. EMG and single-fiber EMG studies may or may not reveal a defect of neuromuscular transmission. In some patients exercise or subtetanic stimulation is required to elicit a decremental EMG response.154 Similar EMG findings have been reported in the Near-Eastern Jewish patients with facial malformations.73,181

FIGURE 33–22 Schematic diagram showing structure of the RAPSN gene (A) and domains of rapsyn (B) with identified mutations. Seven tetratricopeptide repeats (TPRs) are required for rapsyn self-association; the coiled-coil domain binds to the long cytoplasmic loop of AChR subunits, and the RING-H2 domain links rapsyn to ␤-dystroglycan. Shaded areas in A indicate untranslated regions in RAPSN. E ⫽ E-box.

857

Diseases of the Neuromuscular Junction

End-Plate Studies The EPs, like those of patients with low-expressor mutation of the AChR, show a reduced expression of rapsyn and a proportionately reduced expression of AChR. In some patients, however, the number of AChRs per EP is only mildly reduced.154 Cholinesterase stains reveal multiple small EP regions dispersed over an increased span of individual muscle fibers. Ultrastructural studies show shallow postsynaptic folds and clefts, few secondary clefts, and smaller than normal nerve terminals and postsynaptic regions; the structural integrity of the pre- and postsynaptic regions is preserved.154,159 In vitro electrophysiologic studies show a higher than normal quantal release in some patients. Consistent with the EP AChR deficiency, the MEPP and MEPC amplitudes are reduced. Single-channel patch-clamp recordings show no kinetic abnormality of the AChR channel.

Molecular Pathogenesis Among the first four identified patients with rapsyn deficiency, one carried L14P in TPR1 and N88K in TPR3; two were homozygous for N88K; and one carried N88K and 553ins5 that frameshifts in TPR5.154 Several other rapsyn mutations were subsequently identified (see Fig. 33–22B). Expression of the first identified mutations in HEK cells showed that none hindered rapsyn self-association but all diminished co-clustering of AChR with rapsyn.154 That missense mutations in TPR domains decrease co-clustering of AChR with rapsyn implies that effects of these mutations propagate downstream to the coiled-coil domain, or that the mutations have an allosteric effect on the conformation of the coiled-coil or RING-H2 domains, or that the TPR mutations might compromise the cotransport of AChR with rapsyn to the cell surface. The N88K rapsyn mutation appears to be a relatively frequent cause of CMS.7,21,37,141,154,159 Haplotype analysis of 44 subjects in 10 kinships carrying the N88K mutation with four microsatellite markers revealed that only 20% of the mutant alleles shared the same haplotypes. Thus the N88K mutation either arose repeatedly and independently in different populations, or is an ancient mutation that arose even before the microsatellites diverged. A founder effect was noted in 10 independent European kinships,141 but the analysis was based on only one microsatellite.141 Other frameshift (46insC, 55ins5, C97X, Q124X and Y269X)37 and missense (Q3K, R91L, A142D, R151P, A246V, and G291D)141 mutations of rapsyn were also detected (see Fig. 33–22B). The E-box elements in the RAPSN promoter region with the consensus sequence of CANNTG are targets of myogenic determination factors and therefore par-

ticipate in modulating RAPSN expression. The two identified mutations in RAPSN E-box elements (⫺38A⬎G and ⫺27C⬎G) are the first examples of E-box mutations in humans. The ⫺38A⬎G mutation was homozygous in seven Oriental Jewish kinships with characteristic facial malformations and was traced to a common founder. The pathogenicity of both E-box mutations was confirmed in a luciferase reporter assay that showed attenuated reporter gene expression in C2C12 myotubes.159 The phenotype associated with the ⫺38A⬎G mutation appears to be homogeneous. However, there is no clear phenotype-genotype correlation in other rapsyndeficient patients. For example, among two patients homozygous for the same N88K mutation, one has severe myasthenic symptoms at the age of 2 years but the other has only mild weakness at the age of 27 years. A similar lack of genotype-phenotype correlation has been noted in other series.21

SODIUM CHANNEL MYASTHENIA Clinical Features Only one patient with this syndrome has been observed to date.206 This was the case of a 20-year-old normokalemic woman who had abrupt attacks of respiratory and bulbar paralysis since birth lasting 3 to 30 minutes and recurring one to three times per month. She survived only because since infancy she has been on an apnea monitor and received ventilatory support during apneic attacks. Her motor development was delayed, she always fatigued easily, and she had droopy eyelids. At age 20, she had eyelid ptosis and limited ocular ductions, as well as facial, truncal, and limb muscle weakness worsened by activity (Fig. 33–23); she could walk less than half a block, and could elevate her arms to the horizontal for only 20 seconds. She had a high-arched palate, adduction deformity of the knees and ankles, and increased lumbar lordosis. She was mentally retarded; brain magnetic resonance imaging revealed mild cerebral atrophy that was attributed to previous episodes of cerebral anoxia. Tests for anti-AChR antibodies were negative. The patient’s mother and sister were asymptomatic. No clinical data or DNA was available from the patient’s father. In EMG studies, a decremental response was elicited only by sustained high-frequency stimulation, or by low-frequency stimulation after a conditioning train of high-frequency stimulation. These findings were similar to those observed in the CMS caused by mutations in CHAT.

End-Plate Studies In an intercostal muscle specimen, type I fibers had a smaller mean diameter than type II fibers. The random

858

Pathology of the Peripheral Nervous System

from control muscle fibers. EPPs of the order of 40 mV depolarizing the membrane potential to ⫺40 mV or more failed to trigger action potentials and were recorded in the absence of d-tubocurarine. The MEPP amplitude and quantal content of the EPP were normal. Patch-clamp recordings from three EPs revealed AChR channels opening to a normal conductance of approximately 60 pS, and channel opening burst of normal duration. These findings established that the stimulation-dependent decrement of the CMAP was not caused by impaired resynthesis of acetylcholine and concomitant abnormal decrease of the EPP, and pointed to a defect in Nav1.4, the adult muscle fiber sodium channel. Other factors that compromise neuromuscular transmission in previously recognized CMS (e.g., decrease of evoked quantal release, diminished expression of AChE or AChR, a kinetic defect in AChR, or alteration of EP geometry) were excluded by the EP studies.

Molecular Pathogenesis

FIGURE 33–23 Patient with myasthenic syndrome caused by mutation of the Nav1.4 sodium channel. Note asymmetric ptosis, strabismus, lumbar lordosis, and adduction deformities of knees and ankles.

distribution of the histochemical fiber types was maintained. None of the intercostal muscle fibers harbored vacuoles of the type often observed in periodic paralysis muscle specimens. The configuration of the EPs on cholinesterase-reacted teased muscle fibers and the number of AChRs per EP, determined with 125I-labeled ␣-bungarotoxin, were normal. Electron microscopy showed normal nerve terminals and well-developed junctional folds at all EPs. One of the 63 observed EP regions was denuded of its nerve terminal. The density and distribution of AChR on the crests of the junctional folds, determined with peroxidase-labeled ␣-bungarotoxin, were also normal. Immunolocalization of sodium channels with an anti–pan sodium channel antibody showed similar surface membrane and similarly enhanced synaptic expression at patient and control muscle fibers. In vitro electrophysiology studies revealed that the muscle fiber membrane potential was similar to that recorded

That EPPs depolarizing the muscle fiber membrane potential to ⫺40 mV failed to activate action potentials prompted us to search for a mutation, SCN4A, that encodes Nav1.4. SCN4A harbored two heteroallelic missense mutations: S246L in the cytoplasmic link between the S4 and S5 segments of domain I, and V1442E in the extracellular link between the S3 and S4 segments of domain IV (Fig. 33–24). Neither mutation was observed in 400 normal alleles. Both S246 and V1442 are conserved across Nav1.4 channels of different species. Family analysis showed that the patient’s asymptomatic mother and sister are heterozygous for S246L. Expression studies in HEK showed that both sodium channel mutants expressed to normal levels in HEK cells. The salient finding consisted of a hyperpolarizing shift in the voltage dependence of fast inactivation; this was marked (⫺33 mV) for the V1442E mutant and milder (⫺7 mV) for the S246L mutant. The hyperpolarizing shift caused by the V1442E mutation predicts that nearly all V1442E channels at the EP are fast-inactivated, and hence inexcitable, at a normal resting membrane potential of ⫺80 mV. To determine whether use-dependent sodium channel inactivation might contribute to the abnormal decrement of the CMAP in the patient’s muscle at high-frequency stimulation, the response to a 50-Hz train of 3-ms pulses was measured in vitro. This revealed a precipitous drop of 30% of the normalized peak current amplitude during the first few pulses for the V1442E mutant, whereas wild-type and S246L channels showed only a 5% decrease. The rapid decrement can be attributed to trapping of V1442E channels in the fast-inactivated state

859

Diseases of the Neuromuscular Junction

FIGURE 33–24 Scheme of skeletal muscle sodium channel Nav1.4 encoded by SCN4A and the identified mutations in patient with sodium-channel myasthenia.

because recovery at ⫺100 mV was slowed threefold for this mutant.

Relation to Other Sodium Channel Disorders The present phenotype differs from that of periodic paralyses associated with hitherto identified mutations of SCN4A. The onset is neonatal, the disorder is normokalemic, the attacks selectively involve bulbar and respiratory muscles, physiologic rates of stimulation decrement the CMAP abnormally, and the muscle fiber membrane potential is normal when action potential generation fails. Periodic paralyses stemming from mutations in SCN4A present later in life, and these attacks typically spare cranial, bulbar, and respiratory muscles; the serum potassium level increases or declines during attacks in most cases, mild exercise for brief periods does not decrement the CMAP, and the resting membrane potential of the muscle fiber is decreased when action potential generation fails.55,119 The syndrome also differs from most other sodium channel disorders involving skeletal muscle. These are typically associated with a gain of function consisting of an excessive inward sodium current, whereas the present syndrome is associated with loss of function consisting of a markedly diminished inward sodium current. However, a subset of hypokalemic periodic paralyses caused by mutations of Arg 672 is also associated with loss of function and a hyperpolarizing shift of fast inactivation, but the shift is only ⫺10 mV and does not prevent action potential generation.202 The inheritance pattern for this CMS remains uncertain. The more severe V1442E mutation may be dominant, but this cannot be proven because this mutation derived from the patient’s father, who could not be evaluated, and because V1442E was observed only in combination with the heteroallelic S246L mutation. The S246L mutation alone was clinically silent and therefore either represents a rare polymorphism with detectable biophysical changes but no clinical or EMG phenotype, or is a recessive mutation. Despite this uncertainty, sodium channel myasthenia is a new allelic disorder to be grouped with periodic paralyses

and myotonias attributed to mutations in Nav1.4. This CMS is also the first in which the reduced safety margin of neuromuscular transmission occurs in the setting of a normal EP potential.

CMS ASSOCIATED WITH PLECTIN DEFICIENCY Plectin is a highly conserved and ubiquitously expressed intermediate filament–linking protein concentrated at sites of mechanical stress, such as the postsynaptic membrane of the EP, the sarcolemma, Z-discs in skeletal muscle, hemidesmosomes in skin, and intercalated disks in cardiac muscle. Pathogenic mutations in plectin are associated with a simplex variety of epidermolysis bullosa, a progressive myopathy, and a myasthenic syndrome.9 In a patient with epidermolysis bullosa simplex, a progressive myopathy, abnormal fatigability involving the ocular, facial, and limb muscles, as well as a decremental EMG response and no anti-AChR antibodies, revealed that plectin expression was absent in muscle and severely decreased in skin. Morphologic studies of muscle demonstrated necrotic and regenerating fibers and a wide spectrum of ultrastructural abnormalities. Many EPs had an abnormal configuration with chains of small regions over the fiber surface, and a few EPs displayed focal degeneration of the junctional folds. The EP AChR content was normal. In vitro electrophysiologic studies showed normal quantal release by nerve impulse, small MEPPs, and expression of fetal as well as adult AChR at the EPs.9

Therapy MYASTHENIA GRAVIS The mainstays of therapy are anticholinesterase medications, thymectomy, and immunosuppressive measures. A detailed discussion of modalities of therapy appropriate

860

Pathology of the Peripheral Nervous System

for the different grades of autoimmune MG are beyond the scope of this chapter. For a detailed discussion of the treatment of autoimmune MG, the reader is referred to a recent review.44

LAMBERT-EATON MYASTHENIC SYNDROME In noncarcinomatous LEMS, the mainstays of therapy are 3,4-DAP, which augments quantal release from the nerve terminal, and immunosuppressive measures. In carcinomatous LEMS, removal of the neoplasm as well as 3,4-DAP is indicated. For a detailed discussion of the treatment of LEMS, the reader is referred to a recent review.44

5.

6. 7.

8.

THE CONGENITAL MYASTHENIC SYNDROMES 9. For the purposes of therapy, the CMSs can be classified into two major categories: those that decrease and those that increase the synaptic response to ACh. The CMSs that decrease the synaptic response to ACh are treated with cholinergic agonists, namely drugs that inhibit AChE, and, in some cases, with 3,4-DAP. 3,4-DAP increases the number of quanta released by nerve impulse, and AChE inhibitors increase the number of AChRs activated by each quantum. An increased synaptic response to ACh occurs in the slow-channel syndromes and in EP AChE deficiency. The slow-channel syndromes are treated with long-lived open-channel blockers of AChR. There is no satisfactory therapy for EP AChE deficiency. The following special considerations apply: 1. End-plate ChAT deficiency is treated prophylactically with an oral AChE inhibitor (e.g., pyridostigmine) to prevent respiratory crises, and with parenteral prostigmine methylsulfate during respiratory crises. The patient’s parents must be taught how to administer prostigmine parenterally, and must own and be familiar with the use of a portable respiratory device. 2. The CMS associated with paucity of the synaptic vesicles and decreased quantal release is treated with AChE inhibitors. 3. One case of the LEMS-like CMS has responded to guanidine,5 a medication that acts like 3,4-DAP, but another patient, observed by the author, failed to respond to 3,4-DAP clinically. 4. In EP AChE deficiency, it is important to avoid cholinergic agonists. Anecdotal reports suggest that

ephedrine may mitigate the course of the disease. One respirator-dependent patient benefited from intermittent blockade of AChR by atracurium, an agent that protects AChR from overexposure to ACh, allowing for temporary withdrawal of respiratory support.19 The slow-channel syndromes are treated with quinidine sulfate66,86 or fluoxetine.87 Both medications act as longlived open-channel blockers of AChR and gradually improve the disease. The fast-channel syndromes respond well to combined therapy with pyridostigmine and 3,4-DAP. Patients with rapsyn deficiency also respond to pyridostigmine, and some derive striking further benefit from the use of 3,4-DAP.8 A patient with sodium channel myasthenia showed improved endurance on treatment with pyridostigmine; additional therapy with acetazolamide, which is known to mitigate periodic paralysis caused by Nav1.4 mutations, prevented further attacks of respiratory and bulbar weakness.206 In a patient with a CMS associated with plectin deficiency, pyridostigmine failed to improve the patient’s symptoms, but 3,4-DAP improved her strength and endurance.9

For further discussion of the treatment of the CMSs, the reader is referred to a recent review.44

REFERENCES 1. Abicht, A., Stucka, R., Karcagi, V., et al.: A common mutation (⑀1267delG) in congenital myasthenic patients of Gypsy ethnic origin. Neurology 53:1564, 1999. 2. Abicht, A., Stucka, R., Schmidt, C., et al.: A newly identified chromosomal microdeletion and an N-box mutation of the AChRe gene cause a congenital myasthenic syndrome. Brain 125:1005, 2002. 3. Anderson, C. R., and Stevens, C. F.: Voltage clamp analysis of acetylcholine produced end-plate current fluctuations at frog neuromuscular junction. J. Physiol. (Lond.) 235:655, 1973. 4. Apparsundaram, S., Ferguson, S. M., George, A. L. Jr., and Blakely, R. D.: Molecular cloning of a human, hemicholinium3-sensitive choline transporter. Biochem. Biophys. Res. Commun. 276:862, 2000. 5. Bady, B., Chauplannaz, G., and Carrier, H.: Congenital Lambert-Eaton myasthenic syndrome. J. Neurol. Neurosurg. Psychiatry 50:476, 1987. 6. Bady, B., Vial, C., and Chauplannaz, G.: Syndrome de Lambert-Eaton: étude clinique et électrophysiologique de 18 cas associes à un cancer du poumon. Rev. Neurol. (Paris) 148:513, 1992. 7. Banwell, B. L., Brengman, J. M., Ohno, K., et al.: Novel truncating mutation causing congenital myasthenic syndrome. Neurology 60(Suppl. 1):A420, 2003.

Diseases of the Neuromuscular Junction 8. Banwell, B. L., Ohno, K., Sieb, J. P., and Engel, A. G.: Novel truncating RAPSN mutation causing congenital myasthenic syndrome responsive to 3,4-diaminopyridine. Neuromuscul. Disord. 14:202, 2004. 9. Banwell, B. L, Russel, J., Fukudome, T., et al.: Myopathy, myasthenic syndrome, and epidermolysis bullosa simplex due to plectin deficiency. J. Neuropathol. Exp. Neurol. 58:832, 1999. 10. Bartoli, M., and Cohen, J. B.: Identification of the modular domains of rapsyn binding to nicotinic acetylcholine receptor (AChR) and to dystroglycan. Soc. Neurosci. Abstr. 27:904.16, 2001. 11. Bartoli, M., Ramarao, M. K., and Cohen, J. B.: Interactions of the rapsyn RING-H2 domain with dystroglycan. J. Biol. Chem. 276:24911, 2001. 12. Batocchi, A. P., Evoli, A., Servidei, S., et al.: Myasthenia gravis during interferon alpha therapy. Neurology 42:382, 1995. 13. Beech, R. L., Vaca, K., and Pilar, G.: Ionic and metabolic requirements for high-affinity choline uptake and acetylcholine synthesis in nerve terminals of a neuromuscular junction. J. Neurochem. 34:1387, 1980. 14. Betz, W., and Sakmann, B.: Effects of proteolytic enzymes on function and structure of frog neuromuscular junctions. J. Physiol. (Lond.) 230:673, 1973. 15. Bhakdi, S., and Trnum-Jensen, J.: Complement lysis: a hole is a hole. Immunol. Today 12:318, 1991. 16. Bon, S., Coussen, F., and Massoulié, J.: Quaternary associations of acetylcholinesterase. II. The polyproline attachment domain of the collagen tail. J. Biol. Chem. 272:3016, 1997. 17. Bööj, S., Larsson, P.-A., Dahllöf, A.-G., and Dahlström, A.: Axonal transport of synapsin I and cholinergic synaptic vesiclelike material: further immunohistochemical evidence for transport of axonal cholinergic transmitter vesicles in motor neurons. Acta Physiol. Scand. 128:155, 1986. 18. Brejc, K., van Dijk, W. V., Schuurmans, M., et al.: Crystal structure of ACh-binding protein reveals the ligand-binding domain of nicotinic receptors. Nature 411:269, 2001. 19. Breningstall, G. N., Kuracheck, S. C., Fugate, J. H., and Engel, A. G.: Treatment of congenital endplate acetylcholinesterase deficiency by neuromuscular blockade. J. Child Neurol. 11:345, 1996. 20. Brownlow, S., Webster, R., Croxen, R., et al.: Acetylcholine receptor ␦ subunit mutations underlie a fast-channel myasthenic syndrome and arthrogryposis multiplex congenita. J. Clin. Invest. 108:125, 2001. 21. Burke, G., Cossins, J., Maxwell, S., et al.: Rapsyn mutations in hereditary myasthenia: distinct early- and late-onset phenotypes. Neurology 61:826, 2003. 22. Byring, R. F., Pihko, H., Shen, X.-M., et al.: Congenital myasthenic syndrome associated with episodic apnea and sudden infant death. Neuromuscul. Disord. 12:548, 2002. 23. Campbell, A. K., Daw, R. A., and Luzio, J. P.: Rapid increase in intracellular free Ca2⫹ induced by antibody and complement. FEBS Lett. 107:55, 1979. 24. Campbell, H., and Bramwell, E.: Myasthenia gravis pseudoparalytica: review of 70 case reports, including nine new patients. Brain 23:277, 1900. 25. Caroni, P., and Grandes, P.: Nerve sprouting in innervated adult skeletal muscle induced by exposure to elevated levels of insulin-like growth factors. J. Cell Biol. 110:1307, 1990.

861

26. Cartaud, A., Coutant, S., Petrucci, T. C., and Cartaud, J.: Evidence for in situ and in vitro association between ␤-dystroglycan and the subsynaptic 43K rapsyn protein: consequence for acetylcholine receptor clustering at the synapse. J. Biol. Chem. 273:11321, 1998. 27. Charlton, M. P., Smith, S. J., and Zuker, R.: Role of presynaptic calcium ions and channels in synaptic facilitation and depression at the squid giant synapse. J. Physiol. (Lond.) 323:173, 1982. 28. Clouston, P. D., Saper, C. B., Arbizu, T., et al.: Paraneoplastic cerebellar degeneration. III. Cerebellar degeneration, cancer, and the Lambert-Eaton myasthenic syndrome. Neurology 42:1944, 1992. 29. Croxen, R., Newland, C., Beeson, D., et al.: Mutations in different functional domains of the human muscle acetylcholine receptor ␣ subunit in patients with the slow-channel congenital myasthenic syndrome. Hum. Mol. Genet. 6:767, 1997. 30. Croxen, R., Newland, C., Betty, M., et al.: Novel functional ⑀-subunit polypeptide generated by a single nucleotide deletion in acetylcholine receptor deficiency congenital myasthenic syndrome. Ann. Neurol. 46:639, 1999. 31. del Castillo, J., and Katz, B.: Quantal components of the end-plate potential. J. Physiol. (Lond.) 124:560, 1954. 32. Deprez, P. N., and Inestrosa, N. C.: Two heparin-binding domains are present on the collagenic tail of asymmetric acetylcholinesterase. J. Biol. Chem. 270:11043, 1995. 33. Deymeer, F., Serdaroglu, P., and Özdemir, C.: Familial infantile myasthenia: confusion in terminology. Neuromuscul. Disord. 9:129, 1999. 34. Donger, C., Krejci, E., Serradell, P., et al.: Mutation in the human acetylcholinesterase-associated gene, COLQ, is responsible for congenital myasthenic syndrome with end-plate acetylcholinesterase deficiency. Am. J. Hum. Genet. 63:967, 1998. 35. Drachman, D. B., Adams, R. N., Josifek, L. F., and Sel, S. G.: Functional activities of autoantibodies to acetylcholine receptors and the clinical severity of myasthenia gravis. N. Engl. J. Med. 307:769, 1982. 36. Duclert, A., Savatier, N., Schaeffer, L., and Changeux, J.-P.: Identification of an element crucial for the sub-synaptic expression of the acetylcholine receptor epsilon-subunit gene. J. Biol. Chem. 271:17433, 1996. 37. Dunne, V., and Maselli, R. A.: Identification of pathogenic mutations in the human rapsyn gene. Hum. Genet. 48:204, 2003. 38. Ealing, J., Webster, R., Brownlow, S., et al.: Mutations in congenital myasthenic syndromes reveal an e subunit C-terminal cysteine, C470, crucial for maturation and surface expressions of adult AChR. Hum. Mol. Genet. 11:3087, 2002. 39. Eaton, L. M., and Lambert, E. H.: Electromyography and electrical stimulation of nerves in diseases of the motor unit: observations on a myasthenic syndrome associated with malignant tumors. J. Am. Med. Assoc. 163:1117, 1957. 40. Eiden, L. E.: The cholinergic gene locus. J. Neurochem. 70:2227, 1998. 41. Elmqvist, D., and Lambert, E. H.: Detailed analysis of neuromuscular transmission in a patient with the myasthenic syndrome sometimes associated with bronchogenic carcinoma. Mayo Clin. Proc. 43:689, 1968.

862

Pathology of the Peripheral Nervous System

42. Elmqvist, D., and Quastel, D. M. J.: A quantitative study of end-plate potentials in isolated human muscle. J. Physiol. (Lond.) 178:505, 1965. 43. Elrington, G. M., Murray, N. M., Spiro, S. G., and NewsomDavis, J.: Neurological paraneoplastic syndromes in patients with small cell lung cancer: a prospective survey of 150 patients. J. Neurol. Neurosurg. Psychiatry 54:764, 1991. 44. Engel, A. G.: Myasthenia gravis and myasthenic syndromes. In Noseworthy, J. N. (ed.): Neurological Therapeutics: Principles and Practice. New York, Martin Dunitz, p. 2378, 2003. 45. Engel, A. G., and Arahata, K.: The membrane attack complex of complement at the endplate in myasthenia gravis. Ann. N. Y. Acad. Sci. 505:326, 1987. 46. Engel, A. G., Fukuoka, T., Lang, B., et al.: Lambert-Eaton myasthenic syndrome IgG: early morphologic effects and immunolocalization at the motor endplate. Ann. N. Y. Acad. Sci. 505:333, 1987. 47. Engel, A. G., and Lambert, E. H.: Congenital myasthenic syndromes. Electroencephalogr. Clin. Neurophysiol. Suppl. 39:91, 1987. 48. Engel, A. G., Lambert, E. H., and Gomez, M. R.: A new myasthenic syndrome with end-plate acetylcholinesterase deficiency, small nerve terminals, and reduced acetylcholine release. Ann. Neurol. 1:315, 1977. 49. Engel, A. G., Lambert, E. H., and Howard, F. M.: Immune complexes (IgG and C3) at the motor end-plate in myasthenia gravis: ultrastructural and light microscopic localization and electrophysiologic correlations. Mayo Clin. Proc. 52:267, 1977. 50. Engel, A. G., Lambert, E. H., Mulder, D. M., et al.: A newly recognized congenital myasthenic syndrome attributed to a prolonged open time of the acetylcholine-induced ion channel. Ann. Neurol. 11:553, 1982. 51. Engel, A. G., Lindstrom, J. M., Lambert, E. H., and Lennon, V. A.: Ultrastructural localization of the acetylcholine receptor in myasthenia gravis and in its experimental autoimmune model. Neurology 27:307, 1977. 52. Engel, A. G., Ohno, K., Bouzat, C., et al.: End-plate acetylcholine receptor deficiency due to nonsense mutations in the ⑀ subunit. Ann. Neurol. 40:810, 1996. 53. Engel, A. G., Ohno, K., Milone, M., et al.: New mutations in acetylcholine receptor subunit genes reveal heterogeneity in the slow-channel congenital myasthenic syndrome. Hum. Mol. Genet. 5:1217, 1996. 54. Engel, A. G., Ohno, K., and Sine, S. M.: Congenital myasthenic syndromes. In Engel, A. G. (ed.): Myasthenia Gravis and Myasthenic Disorders. New York, Oxford University Press, p. 251, 1999. 55. Engel, A. G., Potter, C. S., and Rosevear, J. W.: Clinical and electromyographic studies in a patient with primary hypokalemic periodic paralysis. Am. J. Med. 38:626, 1965. 56. Engel, A. G., Sahashi, K., and Fumagalli, G.: The immunopathology of acquired myasthenia gravis. Ann. N. Y. Acad. Sci. 377:158, 1981. 57. Engel, A. G., and Santa, T.: Histometric analysis of the ultrastructure of the neuromuscular junction in myasthenia gravis and in the myasthenic syndrome. Ann. N. Y. Acad. Sci. 183:46, 1971.

58. Erickson, J. D., Varoqui, H., Eiden, L. E., et al.: Functional identification of a vesicular acetylcholine transporter and its expression from a ‘cholinergic’ gene locus. J. Biol. Chem. 269:21929, 1994. 59. Esser, A. F.: Big MAC attack: complement proteins cause leaky patches. Immunol. Today 12:316, 1991. 60. Fatt, P., and Katz, B.: Spontaneous subthreshold activity at motor nerve endings. J. Physiol. (Lond.) 117:109, 1952. 61. Fawcett, P. R. W., McLachlan, S. M., Nicholson, L. V. B., et al.: D-penicillamine associated myasthenia gravis: immunological and electrophysiological studies. Muscle Nerve 5:328, 1982. 62. Fertuck, H. C., and Salpeter, M. M.: Quantitation of junctional and extrajunctional acetylcholine receptors by electron microscope autoradiography after 125I-␣-bungarotoxin binding at mouse neuromuscular junctions. J. Cell. Biol. 69:144, 1976. 63. Flucher, B. E, and Daniels, M. P.: Distribution of Na⫹ channels and ankyrin in neuromuscular junctions is complementary to that of acetylcholine receptors and the 43 kd protein. Neuron 3:163, 1989. 64. Froehner, S. C., Luetje, C. W., Scotland, P. B., and Patrick, J.: The postsynaptic 43K protein clusters muscle nicotinic acetylcholine receptors in Xenopus oocytes. Neuron 5:403, 1990. 65. Fromm, L., and Burden, S. J.: Synapse-specific and neuregulin-induced transcription require an Ets site that binds GABP␣/GAPB␤. Genes Dev. 12:3074, 1998. 66. Fukudome, T., Ohno, K., Brengman, J. M., and Engel, A. G.: Quinidine normalizes the open duration of slow-channel mutants of the acetylcholine receptor. Neuroreport 9:1907, 1998. 67. Fukunaga, H., Engel, A. G., Lang, B., and Vincent, A.: Passive transfer of Lambert-Eaton myasthenic syndrome with IgG from man to mouse depletes the presynaptic membrane active zones. Proc. Natl. Acad. Sci. U. S. A. 80:7636, 1983. 68. Fukunaga, H., Engel, A. G., Osame, M., and Lambert, E. H.: Paucity and disorganization of presynaptic membrane active zones in the Lambert-Eaton myasthenic syndrome. Muscle Nerve 5:686, 1982. 69. Fukuoka, T., Engel, A. G., Lang, B., and Vincent, A.: Lambert-Eaton myasthenic syndrome: II. Immunoelectron microscopy localization of IgG at the mouse motor endplate. Ann. Neurol. 22:200, 1987. 70. Fukuoka, T., Engel, A. G., Lang, B., et al.: Lambert-Eaton myasthenic syndrome: I. Early morphological effects of IgG on the presynaptic membrane active zones. Ann. Neurol. 22:193, 1987. 71. Fulpius, B. W., Zurn, A. D., Granato, D. A., and Leder, R. M.: Acetylcholine receptor and myasthenia gravis. Ann. N. Y. Acad. Sci. 274:116, 1976. 72. Gautam, M., Noakes, P. G., Mudd, J., et al.: Failure of postsynaptic specialization to develop at neuromuscular junctions of rapsyn-deficient mice. Nature 377:232, 1995. 73. Goldhammer, Y., Blatt, I., Sadeh, M., and Goodman, R. M.: Congenital myasthenia associated with facial malformations in Iraqi and Iranian Jews. Brain 113:1291, 1990.

Diseases of the Neuromuscular Junction 74. Gomez, C. M., Maselli, R., Gammack, J., et al.: A beta-subunit mutation in the acetylcholine receptor gate causes severe slow-channel syndrome. Ann. Neurol. 39:712, 1996. 75. Gomez, C. M., Maselli, R., Staub, J., et al.: Novel ␦ and ␤ subunit acetylcholine receptor mutations in the slow-channel syndrome demonstrate phenotypic variability. Soc. Neurosci. Abstr. 24:484, 1998. 76. Gomez, C. M., Maselli, R., Vohra, B. P. S., et al.: Novel delta subunit mutation in slow-channel syndrome causes severe weakness by novel mechanism. Ann. Neurol. 51:102, 2002. 77. Greer, M., and Schotland, M.: Myasthenia gravis in the newborn. Pediatrics 26:101, 1960. 78. Grob, D., Brunner, N. G., and Namba, T.: The natural course of myasthenia gravis and effect of therapeutic measures. Ann. N. Y. Acad. Sci. 377:652, 1981. 79. Grohovaz, F., Limbrick, A. R., and Miledi, R.: Acetylcholine receptors at the rat neuromuscular junction as revealed by deep etching. Proc. R. Soc. Lond. Ser. B 215:147, 1982. 80. Gurtubay, I. G., Morales, G., Arechaga, O., and Gallego, J.: Development of myasthenia gravis after interferon alpha therapy. Electromyogr. Clin. Neurophysiol. 39:75, 1999. 81. Gutmann, L., Crosby, T. W., Takamori, M., and Martin, J. D.: The Eaton-Lambert syndrome and autoimmune disorders. Am. J. Med. 53:354, 1972. 82. Gutmann, L., Weidman, D., and Gutierrez, A.: LambertEaton myasthenic syndrome with prominent postexercise exhaustion. Muscle Nerve 16:716, 1993. 83. Hall, Z. W., and Kelly, R. B.: Enzymatic detachment of endplate acetylcholinesterase from muscle. Nature 232:62, 1971. 84. Harada, H., Tamaoka, A., Kohno, Y., et al.: Exacerbation of myasthenia gravis in a patient after interferon-beta treatment for chronic active hepatitis C. J. Neurol. Sci. 165:182, 2001. 85. Harper, C. M.: Electrodiagnosis of endplate disease. In Engel, A. G. (ed.): Myasthenia Gravis and Myasthenic Disorders. New York, Oxford University Press, p. 65, 1999. 86. Harper, C. M., and Engel, A. G.: Quinidine sulfate therapy for the slow-channel congenital myasthenic syndrome. Ann. Neurol. 43:480, 1998. 87. Harper, C. M., Fukudome, T., and Engel, A. G.: Treatment of slow channel congenital myasthenic syndrome with fluoxetine. Neurology 60:170, 2003. 88. Hart, Z., Sahashi, K., Lambert, E. H., et al.: A congenital, familial, myasthenic syndrome caused by a presynaptic defect of transmitter resynthesis of mobilization. Neurology 29:559, 1979. 89. Hartzell, H. C., Kuffler, S. W., and Yoshikami, D.: Postsynaptic potentiation: interaction between quanta of acetylcholine at the skeletal neuromuscular synapse. J. Physiol. (Lond.) 251:427, 1975. 90. Hartzell, H. C., Kuffler, S. W., and Yoshikami, D.: The number of acetylcholine molecules in a quantum and the interaction between quanta at the subsynaptic membrane of the skeletal neuromuscular synapse. Symp. Quant. Biol. 40:175, 1976. 91. Hatton, C. J., Shelley, C., Brydson, M., et al.: Properties of the human muscle nicotinic receptor, and of the slow-channel

92.

93.

94.

95.

96.

97.

98.

99.

100. 101.

102.

103.

104.

105.

106.

107.

108.

863

myasthenic syndrome mutant ⑀L221F, inferred from maximum likelihood fits. J. Physiol. (Lond.) 547:729, 2003. Hawley, R. J., Cohen, M. H., Saini, N., et al.: The carcinomatous neuromyopathy of oat cell lung cancer. Ann. Neurol. 7:65, 1980. Heuser, J. E., and Reese, T. S.: Evidence for recycling of synaptic vesicle membrane during transmitter release at the frog neuromuscular junction. J. Cell Biol. 57:315, 1973. Heuser, J. E., Reese, T. S., Dennis, M. J., et al.: Synaptic vesicle exocytosis captured by quick freezing and correlated with quantal transmitter release. J. Cell Biol. 81:275, 1979. Heuser, J. E., and Salpeter, S. R.: Organization of acetylcholine receptors in quick-frozen, deep-etched, and rotary-replicated Torpedo postsynaptic membrane. J. Cell Biol. 82:150, 1979. Hirokawa, N., and Heuser, J. E.: Internal and external differentiations of the postsynaptic membrane at the neuromuscular junction. J. Neurocytol. 11:487, 1982. Hoch, W., McConville, J., Helms, S., et al.: Autoantibodies to the receptor tyrosine kinase MuSK in patients with myasthenia gravis without acetylcholine receptor antibodies. Nat. Med. 7:365, 2001. Hohlfeld, R., and Wekerle, H.: The immunopathogenesis of myasthenia gravis. In Engel, A. G. (ed.): Myasthenia Gravis and Myasthenic Disorders. New York, Oxford University Press, p. 87, 1999. Hubbard, J. I., Jones, S. F., and Landau, E. M.: On the mechanism by which calcium and magnesium affect the release of transmitter by nerve impulses. J. Physiol. (Lond.) 196:75, 1968. Hutchinson, D. O., Walls, T. J., Nakano, S., et al.: Congenital endplate acetylcholinesterase deficiency. Brain 116:633, 1993. Ishigaki, K., Nicolle, D., Krejci, E., et al.: Two novel mutations in the COLQ gene causing endplate acetylcholinesterase deficiency. Neuromuscul. Disord. 13:236, 2003. Jahn, K., Franke, C., and Bufler, J.: Mechanism of block of nicotinic acetylcholine receptor channels by purified IgG from seropositive patients with myasthenia gravis. Neurology 54:474, 2000. Jones, S. F., and Kwanbunbumpen, S.: The effects of nerve stimulation and hemicholinium on synaptic vesicles at the mammalian neuromuscular junction. J. Physiol. (Lond.) 207:31, 1970. Katz, B., and Miledi, R.: The release of acetylcholine from nerve endings by graded electrical pulses. Proc. R. Soc. Lond. Ser. B 167:23, 1967. Katz, B., and Miledi, R.: Tetrodotoxin-resistant electrical activity in presynaptic terminals. J. Physiol. (Lond.) 203:459, 1969. Katz, B., and Miledi, R.: The statistical nature of the acetylcholine potential and its molecular components. J. Physiol. (Lond.) 224:665, 1972. Katz, B., and Miledi, R.: The binding of acetylcholine to receptors and its removal from the synaptic cleft. J. Physiol. (Lond.) 231:549, 1973. Katz, B., and Thesleff, S.: On the factors which determine the amplitude of the miniature end-plate potential. J. Physiol. (Lond.) 137:267, 1957.

864

Pathology of the Peripheral Nervous System

109. Kiene, L.-M., and Stadler, H.: Synaptic vesicles in electromotoneurones. I. Axonal transport, site of transmitter uptake and processing of a core proteoglycan during maturation. EMBO J. 6:2209, 1987. 110. Kimbell, L. M., Ohno, K., Rotundo, R. L., and Engel, A. G.: Transplanting mutant human collagenic tailed acetylcholinesterase onto the frog neuromuscular junction: evidence for an attachment defect in a congenital myasthenic syndrome. Mol. Biol. Cell 12(Suppl.):161a, 2001. 111. Kuffler, S. W., and Yoshikami, D.: The number of transmitter molecules in the quantum: an estimate from iontophoretic application of acetylcholine at the neuromuscular synapse. J. Physiol. (Lond.) 251:465, 1975. 112. Lambert, E. H., Eaton, L. M., and Rooke, E. D.: Defect of neuromuscular transmission associated with malignant neoplasm. Am. J. Physiol. 187:612, 1956. 113. Lambert, E. H., and Elmqvist, D.: Quantal components of end-plate potentials in the myasthenic syndrome. Ann. N. Y. Acad. Sci. 183:183, 1971. 114. Lambert, E. H., and Rooke, E. D.: Myasthenic state and lung cancer. In Brain, W. R., and Norris, F. H. (eds): The Remote Effects of Cancer on the Nervous System. New York, Grune & Stratton, p. 67, 1965. 115. Lambert, E. H., Rooke, E. D., Eaton, L. M., and Hodgson, C. H.: Myasthenic syndrome occasionally associated with bronchial neoplasm: neurophysiologic studies. In Viets, H. R. (ed.): Myasthenia Gravis. Springfield, IL, Charles C Thomas, p. 362, 1961. 116. Land, B. R., Salpeter, E. E., and Salpeter, M. M.: Acetylcholine receptor site density affects the rising phase of miniature endplate currents. Proc. Natl. Acad. Sci. U.S.A. 77:3736, 1980. 117. Lang, B., Newsom-Davis, J., Wray, D. W., et al.: Autoimmune aetiology for myasthenic (Lambert-Eaton) syndrome. Lancet 2:224, 1981. 118. Lee, Y. I., Swope, S. L., and Ferns, M. J.: Rapsyn’s C-terminal domain mediates MuSK-induced phosphorylation of the AChR. Mol. Biol. Cell 13:395a, 2002. 119. Lehmann-Horn, F., and Jurkat-Rott, K.: Voltage-gated ion channels and hereditary disease. Physiol. Rev. 79:1317, 1999. 120. Li, F.-Y., Szobor, A., Croxen, R., et al.: Dominantly inherited myasthenia gravis as a separate genetic entity without involvement of defined candidate gene loci. Int. J. Mol. Med. 7:289, 2001. 121. Lindstrom, J. M., Seybold, M. E., Lennon, V. A., et al.: Antibody to acetylcholine receptor in myasthenia gravis: prevalence, clinical correlates and diagnostic value. Neurology 26:1054, 1976. 122. Llinás, R., and Nicholson, C.: Calcium in depolarization secretion coupling: an aequorin study in squid giant synapse. Proc. Natl. Acad. Sci. U.S.A. 72:187, 1975. 123. Llinás, R., Sugimori, M., Lin, J.-W., et al.: ATP-dependent directional movement of rat synaptic vesicles injected into the presynaptic terminal of squid giant synapse. Proc. Natl. Acad. Sci. U.S.A. 86:5656, 1989. 124. Llinás, R., Sugimori, M., and Siver, R. B.: Presynaptic calcium concentration microdomains and transmitter release. J. Physiol. (Paris) 86:135, 1992.

125. Lundh, H., Nilsson, O., and Rosen, I.: Treatment of LambertEaton syndrome: 3,4-diaminopyridine and pyridostigmine. Neurology 34:1324, 1984. 126. Martin, A. R.: Current concepts of pre- and post-junctional mechanisms in neuromuscular transmission. Ann. N. Y. Acad. Sci. 274:3, 1976. 127. Martin, A. R.: Junctional transmission. II. Presynaptic mechanisms. In Handbook of Neurophysiology. Bethesda, MD, American Physiological Society, p. 329, 1977. 128. Martin, A. R.: Amplification of neuromuscular transmission by postjunctional folds. Proc. R. Soc. Lond. B 258:321, 1994. 129. Maselli, R. A., Chen, D., Delores, M. O., et al.: Choline acetyltransferase mutations in myasthenic syndrome due to deficient acetylcholine resynthesis. Muscle Nerve 27:180, 2003. 130. Maselli, R. A., Kong, D. Z., Bowe, C. M., et al.: Presynaptic congenital myasthenic syndrome due to quantal release deficiency. Neurology 57:279, 2001. 131. Matthews-Bellinger, J., and Salpeter, M. M.: Distribution of acetylcholine receptors at frog neuromuscular junctions with a discussion of some physiological implications. J. Physiol. (Lond.) 279:197, 1978. 132. McMahan, U. J., Sanes, J. S., and Marshall, L. M.: Cholinesterase is associated with the basal lamina at the neuromuscular junction. Nature 271:172, 1978. 133. McMahon, H. T., and Nicholls, D. G.: The bioenergetics of neurotransmitter release. Biochim. Biophys. Acta 1059:243, 1991. 134. Middleton, L., Ohno, K., Christodoulou, K., et al.: Congenital myasthenic syndromes linked to chromosome 17p are caused by defects in acetylcholine receptor ⑀ subunit gene. Neurology 53:1076, 1999. 135. Miledi, R.: Transmitter release by injection of calcium ions into nerve terminals. Proc. R. Soc. Lond. Ser. B 183:421, 1973. 136. Milone, M., Ohno, K., Pruitt, J. N., et al.: Congenital myasthenic syndrome due to frameshifting acetylcholine receptor epsilon subunit mutation. Soc. Neurosci. Abstr. 22:1942, 1996. 137. Milone, M., Wang, H.-L., Ohno, K., et al.: Slow-channel syndrome caused by enhanced activation, desensitization, and agonist binding affinity due to mutation in the M2 domain of the acetylcholine receptor alpha subunit. J. Neurosci. 17:5651, 1997. 138. Milone, M., Wang, H.-L., Ohno, K., et al.: Mode switching kinetics produced by a naturally occurring mutation in the cytoplasmic loop of the human acetylcholine receptor ⑀ subunit. Neuron 20:575, 1998. 139. Monod, J., Wyman, J., and Changeux, J.-P.: On the nature of allosteric transitions: a plausible model. J. Mol. Biol. 12:88, 1965. 140. Mora, M., Lambert, E. H., and Engel, A. G.: Synaptic vesicle abnormality in familial infantile myasthenia. Neurology 37:206, 1987. 141. Müller, J. S., Mildner, G., Müller-Felber, W., et al.: Rapsyn N88K is a frequent cause of CMS in European patients. Neurology 60:1805, 2003. 142. Nagel, A., Engel, A. G., Lang, B., and Fukuoka, T.: LambertEaton myasthenic syndrome IgG depletes presynaptic

Diseases of the Neuromuscular Junction

143.

144.

145.

146. 147.

148.

149.

150.

151.

152. 153.

154.

155.

156.

157.

158.

membrane active zone particles by antigenic modulation. Ann. Neurol. 24:552, 1988. Nakano, S., and Engel, A. G.: Myasthenia gravis: quantitative immunocytochemical analysis of inflammatory cells and detection of complement membrane attack complex at the end-plate in 30 patients. Neurology 43:1167, 1993. Nichols, P. R., Croxen, R., Vincent, A., et al.: Mutation of the acetylcholine receptor ⑀-subunit promoter in congenital myasthenic syndrome. Ann. Neurol. 45:439, 1999. Oda, Y., Nakanishi, I., and Deguchi, T.: A complementary DNA for human choline acetyltransferase induces two forms of enzyme with different molecular weights in cultured cells. Brain Res. Mol. Brain Res. 16:287, 1992. Oh, S. J.: The Eaton-Lambert syndrome in ocular myasthenia gravis. Arch. Neurol. 31:183, 1974. Ohno, K., Anlar, B., and Engel, A. G.: Congenital myasthenic syndrome caused by a mutation in the Ets-binding site of the promoter region of the acetylcholine receptor ⑀ subunit gene. Neuromuscul. Disord. 9:131, 1999. Ohno, K., Anlar, B., Ozdemir, C., et al.: Frameshifting and splice-site mutations in acetylcholine receptor ⑀ subunit gene in 3 Turkish kinships with congenital myasthenic syndromes. Ann. N. Y. Acad. Sci. 841:189, 1998. Ohno, K., Anlar, B., Özdirim, E., et al.: Myasthenic syndromes in Turkish kinships due to mutations in the acetylcholine receptor. Ann. Neurol. 44:234, 1998. Ohno, K., Brengman, J. M., Felice, K. J., et al.: Congenital endplate acetylcholinesterase deficiency caused by a nonsense mutation and an A-to-G splice site mutation at position ⫹3 of the collagen-like tail subunit gene (COLQ): how does G at position ⫹3 result in aberrant splicing? Am. J. Hum. Genet. 65:635, 1999. Ohno, K., Brengman, J. M., Tsujino, A., and Engel, A. G.: Human endplate acetylcholinesterase deficiency caused by mutations in the collagen-like tail subunit (ColQ) of the asymmetric enzyme. Proc. Natl. Acad. Sci. U.S.A. 95:9654, 1998. Ohno, K., and Engel, A. G.: Congenital myasthenic syndromes: gene mutations. Neuromuscul. Disord. 13:283, 2003. Ohno, K., Engel, A. G., Brengman, J. M., et al.: The spectrum of mutations causing endplate acetylcholinesterase deficiency. Ann. Neurol. 47:162, 2000. Ohno, K., Engel, A. G., Shen, X.-M., et al.: Rapsyn mutations in humans cause endplate acetylcholine receptor deficiency and myasthenic syndrome. Am. J. Hum. Genet. 70:875, 2002. Ohno, K., Hutchinson, D. O., Milone, M., et al.: Congenital myasthenic syndrome caused by prolonged acetylcholine receptor channel openings due to a mutation in the M2 domain of the ⑀ subunit. Proc. Natl. Acad. Sci. U.S.A. 92:758, 1995. Ohno, K., Milone, M., Brengman, J. M., et al.: Slow-channel congenital myasthenic syndrome caused by a novel mutation in the acetylcholine receptor ⑀ subunit. Neurology 50:A432, 1998. Ohno, K., Milone, M., Shen, X.-M., and Engel, A. G.: A frameshifting mutation in CHRNE unmasks skipping of the preceding exon. Hum. Mol. Genet. 12:3055, 2003. Ohno, K., Quiram, P., Milone, M., et al.: Congenital myasthenic syndromes due to heteroallelic nonsense/missense

159.

160.

161.

162.

163.

164.

165.

166. 167.

168.

169.

170.

171.

172.

173.

174.

175.

865

mutations in the acetylcholine receptor ⑀ subunit gene: identification and functional characterization of six new mutations. Hum. Mol. Genet. 6:753, 1997. Ohno, K., Sadeh, M., Blatt, I., et al.: E-box mutations in RAPSN promoter region in eight cases with congenital myasthenic syndrome. Hum. Mol. Genet. 12:739, 2003. Ohno, K., Tsujino, A., Brengman, J. M., et al.: Choline acetyltransferase mutations cause myasthenic syndrome associated with episodic apnea in humans. Proc. Natl. Acad. Sci. U.S.A. 98:2017, 2001. Ohno, K., Wang, H.-L., Milone, M., et al.: Congenital myasthenic syndrome caused by decreased agonist binding affinity due to a mutation in the acetylcholine receptor ⑀ subunit. Neuron 17:157, 1996. Ohno, K., Wang, H.-L., Shen, X.-M., et al.: Slow-channel mutations in the center of the M1 transmembrane domain of the acetylcholine receptor ␣ subunit. Neurology 54(Suppl. 3):A183, 2000. Okada, Y., Yamazaki, H., Sekine-Aizawa, Y., and Hirokawa, N.: The neuron-specific kinesin superfamily protein KIF1A is a unique monomeric motor for anterograde axonal transport of synaptic vesicle precursors. Cell 81:769, 1995. Okuda, T., Haga, T., Kanai, Y., et al.: Identification and characterization of the high-affinity choline transporter. Nat. Neurosci. 3:120, 2000. O’Neill, J. H., Murray, N. M. F., and Newsom-Davis, J.: The Lambert-Eaton myasthenic syndrome: a review of 50 cases. Brain 3:577, 1988. Osserman, K. E.: Myasthenia Gravis. New York, Grune & Stratton, 1958. Osserman, K. E., and Genkins, G.: Studies in myasthenia gravis: review of a twenty-year experience in over 1200 patients. Mt. Sinai J. Med. 38:497, 1971. Parsons, K. T., Kwok, W. W., Gaur, L. K., and Nepom, G. T.: Increased frequency of HLA class II alleles DRB1*0301 and DQB!*0201 in Lambert-Eaton myasthenic syndrome without associated cancer. Hum. Immunol. 61:828, 2000. Parsons, S. M., Carpenter, R. S., Koenigsberger, R., and Rothlein, J. E.: Transport in the cholinergic synaptic vesicle. Fed. Proc. 41:2765, 1982. Prockop, D. J., and Kivirikko, K. I.: Collagens: molecular biology, diseases, and potentials for therapy. Annu. Rev. Biochem. 64:403, 1995. Quiram, P., Ohno, K., Milone, M., et al.: Mutation causing congenital myasthenia reveals acetylcholine receptor ␤/␦ subunit interaction essential for assembly. J. Clin. Invest. 104:1403, 1999. Ramarao, M. K., Bianchetta, M. J., Lanken, J., and Cohen, J. B.: Role of rapsyn tetratrichopeptide repeat and coiledcoil domains in self-association and nicotinic acetylcholine receptor clustering. J. Biol. Chem. 276:7475, 2001. Ramarao, M. K., and Cohen, J. B.: Mechanism of nicotinic acetylcholine receptor cluster formation by rapsyn. Proc. Natl. Acad. Sci. U.S.A. 95:4007, 1998. Reimer, R. J., Fon, A. E., and Edwards, R. H.: Vesicular neurotransmitter transport and the presynaptic regulation of quantal size. Curr. Opin. Neurobiol. 8:405, 1998. Robertson, W. C., Chun, R. W. M., and Kornguth, S. E.: Familial infantile myasthenia. Arch. Neurol. 37:117, 1980.

866

Pathology of the Peripheral Nervous System

176. Robitaille, R., Adler, E. M., and Charlton, M. P.: Strategic location of calcium channels at transmitter release sites of frog neuromuscular synapses. Neuron 5:773, 1990. 177. Rondepierre, P., Furby, A., Godefroy, O., et al.: Bloc neuromusculaire mixte pré- et post-synaptique. Rev. Neurol. (Paris) 148:193, 1992. 178. Rooke, E. D., Eaton, L. M., Lambert, E. H., and Hodgson, C. H.: Myasthenia and malignant intrathoracic tumor. Med. Clin. North Am. 44:977, 1960. 179. Rubenstein, A. E., Horowitz, S. H., and Bender, A. N.: Cholinergic dysautonomia and Lambert-Eaton syndrome. Neurology 29:720, 1979. 180. Ruff, R. L.: Sodium channel slow inactivation and the distribution of sodium channels on skeletal muscle fibres enable the performance properties of different skeletal muscle fiber types. Acta Physiol. Scand. 156:159, 1996. 181. Sadeh, M., Blatt, I., and Goldhammer, Y.: Single fiber EMG in a congenital myasthenic syndrome associated with facial malformations. Muscle Nerve 16:177, 1993. 182. Sahashi, K., Engel, A. G., Lambert, E. H., and Howard, F. M. Jr.: Ultrastructural localization of the terminal and lytic ninth complement component (C9) at the motor end-plate in myasthenia gravis. J. Neuropathol. Exp. Neurol. 39:160, 1980. 183. Salpeter, M. M.: Molecular organization of the neuromuscular synapse. In Albuquerque, E. X., and Eldefrawi, A. T. (eds): Myasthenia Gravis. New York, Chapman and Hall, p. 105, 1983. 184. Salpeter, M. M.: Vertebrate neuromuscular junctions: general morphology, molecular organization, and functional consequences. In Salpeter, M. M. (ed.): The Vertebrate Neuromuscular Junction. New York, Alan Liss, p. 1, 1987. 185. Salpeter, M. M., Rogers, A. W., Kasprzak, H., and McHenry, F. A.: Acetylcholinesterase in the fast extraocular muscle of the mouse by light and electron microscopy autoradiography. J. Cell Biol. 78:274, 1978. 186. Satoyoshi, E., Kowa, H., and Fukunaga, N.: Subacute cerebellar degeneration in Eaton-Lambert syndrome with bronchogenic carcinoma. Neurology 23:764, 1973. 187. Schaeffer, L., Duclert, N., Huchet-Dymanus, M., and Changeux, J.-P.: Implication of a multisubunit Ets-related transcription factor in synaptic expression of the nicotinic acetylcholine receptor. EMBO J. 17:3078, 1998. 188. Schmidt, C., Abicht, A., Krampfl, K., et al.: Congenital myasthenic syndrome due to a novel missense mutation in the gene encoding choline acetyltransferase. Neuromuscul. Disord. 13:245, 2003. 189. Schwartz, M. S., and Stålberg, E.: Myasthenic syndrome studied with single fiber electromyography. Arch. Neurol. 32:815, 1975. 190. Seybold, M. E.: Diagnosis of myasthenia gravis. In Engel, A. G. (ed.): Myasthenia Gravis and Myasthenic Disorders. New York, Oxford University Press, p. 167, 1999. 191. Shapira, Y. A., Sadeh, M. E., Bergtraum, M. P., et al.: Three novel COLQ mutations and variation of phenotypic expressivity due to G240X. Neurology 58:603, 2002. 192. Shen, X.-M., Ohno, K., Aams, C., and Engel, A. G.: Slowchannel congenital myasthenic syndrome caused by a novel epsilon subunit mutation in the second AChR transmembrane domain. J. Neurol. Sci. 199(Suppl. 1):S96, 2002.

193. Shen, X.-M., Ohno, K., Fukudome, T., et al.: Congenital myasthenic syndrome caused by low-expressor fast-channel AChR ␦ subunit mutation. Neurology 59:1881, 2002. 194. Shen, X.-M., Ohno, K., Milone, M., et al.: Fast-channel syndrome. Neurology 56(Suppl. 3):A60, 2001. 195. Shen, X.-M., Ohno, K., Tsujino, A., et al.: Mutation causing severe myasthenia reveals functional asymmetry of AChR signature Cys-loops in agonist binding and gating. J. Clin. Invest. 111:497, 2003. 196. Simpson, J. F., Westbery, M. R., and Magee, K. R.: Myasthenia gravis: an analysis of 295 cases. Acta Neurol. Scand. Suppl. 23:1, 1966. 197. Sine, S. M., Claudio, T., and Sigworth, F. J.: Activation of Torpedo acetylcholine receptors expressed in mouse fibroblasts: single-channel current kinetics reveal distinct agonist binding affinities. J. Gen. Physiol. 96:395, 1990. 198. Sine, S. M., Ohno, K., Bouzat, C., et al.: Mutation of the acetylcholine receptor ␣ subunit causes a slow-channel myasthenic syndrome by enhancing agonist binding affinity. Neuron 15:229, 1995. 199. Stanley, E. F., and Drachman, D. B.: Effect of myasthenic immunoglobulin in acetylcholine receptors of intact mammalian neuromuscular junctions. Science 200:1285, 1978. 200. Stiles, J. R., Kovyazina, I. V., Salpeter, E. E., and Salpeter, M. M.: The temperature sensitivity of miniature endplate currents is mostly governed by channel gating: evidence from optimized recordings and Monte Carlo simulations. Biophys. J. 77:1177, 1999. 201. Stiles, J. R., Van Helden, D., Bartol, T. M., et al.: Miniature endplate current rise times ⬍100 ␮s from improved dual recordings can be modeled with passive acetylcholine diffusion from a synaptic vesicle. Proc. Natl. Acad. Sci. U.S.A. 93:5747, 1996. 202. Struyk, A. F., Scoggan, K. A., Bulman, D. E., and Cannon, S. C.: The human muscle Na channel mutation R699H associated with hypokalemic periodic paralysis enhances slow inactivation. J. Neurosci. 20:8610, 2000. 203. Südhof, T. C., and Jahn, R.: Proteins of synaptic vesicles involved in exocytosis and membrane recycling. Neuron 6:665, 1991. 204. Tarelli, F. T., Passafaro, M., Clementi, F., and Sher, E.: Presynaptic localization of omega-conotoxin-sensitive calcium channels at the frog neuromuscular junction. Brain Res. 547:331, 1991. 205. Trontelj, J., and Stålberg, E.: Jitter measurement by axonal micro-stimulation: guidelines and technical notes. Electroencephalogr. Clin. Neurophysiol. 85:30, 1992. 206. Tsujino, A., Maertens, C., Ohno, K., et al.: Myasthenic syndrome caused by mutation of the SCN4A sodium channel. Proc. Natl. Acad. Sci. U.S.A. 100:7377, 2003. 207. Viets, H. R., and Schwab, R. S.: Thymectomy for Myasthenia Gravis. Springfield, IL, Charles C Thomas, 1960. 208. Walls, T. J., Engel, A. G., Nagel, A. S., et al.: Congenital myasthenic syndrome associated with paucity of synaptic vesicles and reduced quantal release. Ann. N. Y. Acad. Sci. 681:461, 1993. 209. Wang, H.-L., Auerbach, A., Bren, N., et al.: Mutation in the M1 domain of the acetylcholine receptor alpha subunit decreases the rate of agonist dissociation. J. Gen. Physiol. 109:757, 1997.

Diseases of the Neuromuscular Junction 210. Wang, H.-L., Milone, M., Ohno, K., et al.: Acetylcholine receptor M3 domain: stereochemical and volume contributions to channel gating. Nat. Neurosci. 2:226, 1999. 211. Wang, H.-L., Ohno, K., Milone, M., et al.: Fundamental gating mechanism of nicotinic receptor channel revealed by mutation causing a congenital myasthenic syndrome. J. Gen. Physiol. 116:449, 2000. 212. Wernig, A.: Estimates of statistical release parameters from crayfish and frog neuromuscular junctions. J. Physiol. (Lond.) 244:207, 1975. 213. Willman, R., and Fuhrer, C.: Neuromuscular synaptogenesis. Cell. Mol. Life Sci. 59:1296, 2002.

867

214. Wilson, K. S. A.: Neurology. London, Butterworth, 1955. 215. Wirtz, P. W., Roep, B. O., Schreuder, G. M., et al.: HLA class I and II in Lambert-Eaton myasthenic syndrome without associated cancer. Hum. Immunol. 62:809, 2001. 216. Wood, S. J., and Slater, C. P.: Safety factor at the neuromuscular junction. Prog. Neurobiol. 64:393, 2001. 217. Zhang, Y., Chen, J., and Auerbach, A.: Activation of recombinant mouse acetylcholine receptors by acetylcholine, carbamylcholine and tetraethylammonium. J. Physiol. (Lond.) 486:189, 1995. 218. Zucker, R. S.: Changes in the statistics of transmitter release during facilitation. J. Physiol. (Lond.) 229:787, 1973.

34 Pathology and Quantitation of Cutaneous Innervation WILLIAM R. KENNEDY, GWEN WENDELSCHAFER-CRABB, MICHAEL POLYDEFKIS, AND JUSTIN C. MCARTHUR

Overview Anatomic Features of the Skin Methods of Cutaneous Nerve Analysis Choice of Biopsy/Blister Location Biopsy Methods Staining Skin Biopsy Specimens Skin Blister Method Epidermal Sheet Preparations Polymerase Chain Reaction Techniques in Skin Microscopy Quantitation of Cutaneous Nerves Epidermal Nerves Normal Epidermal Nerve Density Other Cutaneous Nerves Morphologic Changes in Sensory Neuropathy

Correlation with Tests of Unmyelinated Nerve Function Correlation between Epidermal Nerve and Sural Nerve Utility of Skin Biopsy Contributions of Skin Biopsy to Diagnosis Painful Sensory Neuropathy Diabetic Neuropathy HIV-Associated Sensory Neuropathies Friedreich’s Ataxia Restless Legs Syndrome Familial Dysautonomia (Riley-Day Syndrome, Hereditary Sensory and Autonomic Neuropathy Type III) Congenital Insensitivity to Pain with Anhidrosis (Hereditary Sensory and Autonomic Neuropathy Type IV)

OVERVIEW The validity of skin biopsy/blister for assessing the extent of cutaneous innervation, and especially epidermal nociceptors, has now been established. These techniques provide a reliable and reproducible means of assessing the numbers of C-fiber nociceptors. Easily applicable and robust quantitation of the dermal innervation, sweat glands, hair shafts, and other potentially instructive nerve fiber populations remains to be developed. Skin biopsy/blister methods have opened new opportunities to learn about the response of unmyelinated nerve fibers to a number of disease entities and experimental conditions. The minimally invasive nature of the biopsy, blister, and epidermal sheet techniques makes it possible to study reactions of unmyelinated nerves to experimental conditions directly in human subjects. The methods are useful for diagnosis and

Psoriasis Port-Wine Stains Leprosy Fabry’s Disease Sensory Ganglionopathies Postherpetic Neuralgia CADASIL Chronic Inflammatory Demyelinating Polyneuropathy Postural Tachycardia Syndrome Pediatric Neurologic Disorders Other Conditions Research Uses of Skin Biopsy Human Models of Nerve Regeneration Animal Models Using Skin Biopsy

may find a place in the longitudinal evaluation of disease progression. An area of particular interest is the potential value of serial skin biopsies in clinical trials of agents that are intended to promote regeneration of nerve fibers. Evaluation of unmyelinated cutaneous nerves obtained by skin biopsy has emerged as a useful method to diagnose and study peripheral nerve disorders.27,46,48,52,63 Until recently, work with peripheral nerves was restricted to studies of myelinated nerve fibers because methods available for assessment of small, unmyelinated fibers were limited. With standard electrodiagnostic studies, unmyelinated fibers remain “invisible” because nerve conduction studies assess only large sensory and motor fibers. Even assessment by sural nerve biopsy is problematic; electron microscopy is required to visualize unmyelinated fibers and nerve biopsy only gives a “window” into one location along the nerve at a single time point. Finally, from a functional standpoint, 869

870

Pathology of the Peripheral Nervous System

cutaneous sensation is transduced by identifiable nerve fibers that enter their “targets” in the skin; these cannot be identified by nerve biopsies. These limitations and the observations that individuals with sensory neuropathies have spontaneous acral neuropathic pain with marked allodynia stimulated development of methods to identify and quantify unmyelinated nerves in the skin. Early studies of cutaneous nerves obtained by punch skin biopsy defined the density and distribution of Meissner’s corpuscles and their myelinated nerves in different age groups and in several hereditary disorders. Unmyelinated nerves were not visualized.7,20 The development of sensitive immunohistochemical techniques offered a new approach to identify and quantify small sensory nerves. In particular, immunostaining of skin biopsies for protein gene product 9.5 (PGP 9.5), a neuronal ubiquitin carboxy-terminal hydrolase,106 has been used by a number of groups to visualize the subpapillary plexus of small myelinated and unmyelinated nerve fibers. Epidermal nerve fibers have received the greatest scrutiny, mainly because they appear to be early indicators of neuropathy and adequate samples can easily be obtained for quantitation.18,45,63,110 These fibers include both A and C fibers that convey “slow” nociception. We developed robust normative data,61 and have demonstrated a usually distally dominant pattern of epidermal nerve fiber loss in diabetic neuropathy,47 human immunodeficiency virus (HIV)–associated sensory neuropathies,82 and idiopathic small fiber sensory neuropathies.35 In patients with severe neuropathy, decreased epidermal densities are most severe distally, with less marked reductions found at progressively more rostral levels and prominent predegenerative axonal swellings identified even at asymptomatic sites. In fact, abnormalities of cutaneous innervation are found in some individuals with normal tendon reflexes at the ankles, normal sural nerve action potential amplitudes, and normal quantitative sensation tests (QSTs). Epidermal denervation also occurs in individuals with spontaneous allodynic pain.34,78 While the precise structural correlates of allodynia remain uncertain, it can clearly occur even with marked depletion of A and C fibers in the skin. The association of neuropathic pain with epidermal nerve loss is, en face, paradoxical because nerve fiber loss is traditionally related to loss of sensation without pain. Neuropathic pain has been correlated with epidermal nerve loss in small fiber sensory neuropathy, postherpetic neuralgia, HIV-associated sensory neuropathies, and diabetic neuropathy. Explanations for this apparent paradox include possible changes in both peripheral and central nervous systems. In peripheral neuropathy caused by persistent action of an etiologic agent, continuous spontaneous pain can result from ectopic discharges in peripheral nociceptive fibers.77 Alternatively, sensitized nociceptors responding with reduced threshold to weak stimuli that normally evoke touch, warm, or cool sensations also evoke

pain (hyperalgesia). Capsaicin-induced mechanical and heat hyperalgesia is one example. Another is the neuronal excitability that results from increased voltage-gated Na currents after exposure to serotonin, prostaglandin E2, or adenosine, agents that produce tenderness or hyperalgesia.26 Likewise, products of nerve degeneration might increase nerve excitation by unmasking proteins normally hidden from immune surveillance (e.g., P0, P2). Resultant immune cell activation exposes surviving nerves to inflammatory cytokines and other substances that alter threshold to stimuli.111 There is growing evidence that injury to nociceptors leads to increased firing not only in injured axons but also in uninjured neighboring axons115 and perhaps in dorsal root ganglia neurons at an adjacent level.4 In fact, selective injury of a ventral root causes hypersensitivity of C fiber afferents in an adjacent root.116 Subsequent activation of nociceptors is believed to produce central sensitization and secondary hyperalgesia. One mechanism is by central terminals of injured A fibers that develop nerve sprouts from deep dorsal horn lamina into superficial laminae. If these synapse with nociceptive neurons, the neurons could become susceptible to activation by otherwise innocuous peripheral stimuli.114 In addition, a shift of descending modulation from inhibition to facilitation can contribute to hypersensitivity.84

ANATOMIC FEATURES OF THE SKIN Skin separates the external environment from internal tissue while providing sensory input, protection from pathogens and allergens, and a water barrier. The layered structure of skin imparts characteristics essential to these functions (Fig. 34–1). The distinct dermal and epidermal layers of skin comprise the major compartments of cutaneous structure. The epidermis is the topmost living layer of skin and consists of layers of keratinocytes that are replaced in approximately a 28-day cycle in humans as they are desquamated to form the cornified surface layer of dead cells that provides the water barrier. Intermixed among the vital keratinocytes are basal melanocytes, dendritic Langerhans cells, and sensory nerve terminals. The epidermis is separated from the dermis by the dermalepidermal basement membrane. The area of dermis immediately internal to the epidermis, the papillary dermis, contains a vascular plexus with periodic capillary loops that reside within dermal protrusions into the epidermis called “dermal papillae.” Within the matrix of the dermis, cutaneous structures include hair follicles with associated arrector pilorum muscles and sebaceous glands, blood vessels, nerve bundles, and sweat glands (Fig. 34–1). The nerves to skin arise from sensory and motor neurons residing in dorsal root ganglia and sympathetic ganglia. Bundles of nerves enter the skin deep in the dermis and course toward the skin surface, giving off axons to

871

Pathology and Quantitation of Cutaneous Innervation

Normal Skin

FIGURE 34–2 Normal human epidermal and papillary dermis innervation. Nerves are localized with antibody to PGP 9.5, and basement membrane is demarcated with antibody to type IV collagen. Vasculature is labeled with Ulex europaeus agglutinin type I. Epidermal nerve fibers appear aqua and lie within the blue epidermis (E). The subepidermal neural plexus (SNP) appears green or yellow. The dermal-epidermal junction appears red. Capillaries (C) appear magenta. Nerve fibers (green and aqua) course in bundles through the dermis and branch in the papillary dermis to form the subepidermal neural plexus. Fibers arise from this plexus and penetrate the epidermal-dermal basement membrane to enter the epidermis. Note that some non-neuronal fibroblasts appear green. Epidermal nerve fibers are abundant and uniformly distributed in normal human skin. See Color Plate

FIGURE 34–1 Human skin innervation and vasculature. Confocal image of 150-m-thick section of human skin immunostained to demonstrate nerves (PGP 9.5—green and yellow) and basement membrane (type IV collagen—red). Image is a projection of 31 5-m optical sections acquired with a 10  objective lens. Basement membrane outlines cutaneous structure, including the basement membrane separating the dermis from the epidermis (Ep), capillary (Cap) loops of the papillary dermis and vasculature throughout the dermis, the central hair follicle (HF) and associated sebaceous gland, arrector pili muscles (AP), and sweat glands (SG). Nerve bundles (NB) arise from the deep dermis to innervate cutaneous structures. Characteristic nerve networks associated with arteries (Ar), sweat glands, arrector pili muscles, the hair follicle, and the subepidermal neural plexus provide sufficient morphologic criteria for identification of these structures based on their innervation patterns. See Color Plate

innervate the associated end organs. Unmyelinated nerve fibers comprise the vast majority of cutaneous innervation to the above dermal structures. The few myelinated nerve fibers terminate at hair follicles, Meissner’s corpus-

cles, and Merkel complexes. When the vertically oriented nerve bundles enter the superficial papillary dermis, they form a horizontal subepidermal neural plexus (Fig. 34–2). Epidermal nerve fibers branch from this plexus and, while penetrating the dermal-epidermal basement membrane to enter the epidermis, they lose their Schwann cell ensheathment and collagen collar.12 Within the epidermis they extend between keratinocytes to the surface of the vital epidermis. Autonomic nerves to sweat glands enwrap the coiled sweat tubules in such a dense pattern that they virtually cover much of the surface (see Fig. 34–1, SG). The autonomic innervation of the arrector pilorum muscles is easily distinguished within the muscle by a characteristic wavy pattern. The complex innervation pattern of hair follicles is composed of both myelinated and unmyelinated fibers, with specialized nerve endings at the base and along the shaft of the hair.22,28,85 Fine unmyelinated nerve endings form a meshlike network covering larger arteries in the deep dermis, but only one or two nerves are on the more superficial smaller arterioles (see Fig. 34–1, Ar). Epidermal nerve fibers are constantly remodeling in concert with the migration and shedding of keratinocytes during the turnover of epidermal layers. In addition,

872

Pathology of the Peripheral Nervous System

epidermal nerve fibers respond to injury by regenerative regrowth and collateral sprouting. The neurotrophic requirements of epidermal nerve fibers have been studied, and at least three trophic factors have been found to affect C-fiber nociceptors: nerve growth factor (NGF), glial cell line–derived neurotrophic factor (GDNF), and insulin-like growth factor type 1. The neurons that depend on NGF, approximately half the small sensory neurons in rat dorsal root ganglia, express the high-affinity NGF receptor TrkA as well as the low-affinity receptor p75. The other half are predominantly GDNF-responsive neurons that express the receptor c-Ret65 as well as a series of other markers. These markers include the ability to bind a specific Griffonia lectin for fucosyl moieties, isolectin B4, which provides a convenient, but not completely specific, marker for the GDNF-dependent population. Finally, most nociceptors bear the receptor for the chili pepper toxin capsaicin. This receptor, vanilloid receptor type 1 (VR1), is the basis for the burning pain elicited by topical application of capsaicin. It normally responds to heat and to protons; mice genetically engineered to lack VR1 fail to respond to heat and acid.10 A related receptor, vanilloid receptor–like receptor type 1, has recently been identified, primarily on somewhat larger sensory neurons probably giving rise to A fibers. Vanilloid receptor–like receptor type 1 responds to very high temperatures (52° C).11 In the human, almost all of the epidermal axons are VR1 positive.82 Epidermal fibers containing the peptides calcitonin gene–related peptide (CGRP) and substance P (SP) are NGF-dependent axons. In some animals these commonly enter the epidermis, but in human skin they usually terminate near capillary loops in the dermal papilla. Attempts to immunostain for other peptides in human nerves, including the isolectin B4, have been unsuccessful.

reduced number of epidermal nerves or may be devoid of epidermal nerve fibers altogether. The foot has the advantage of being more distal, but its nerves are subject to trauma. Supplementary biopsies at more proximal levels (e.g., calf, thigh, or upper extremity) can yield additional information about the severity and distribution of small fiber loss (Fig. 34–3). If the plan is to follow the course of neuropathy or the response to therapy, it is best to select a biopsy location with a definite reduction of epidermal nerves, but with a sufficient nerve residual in the epidermis and subepidermal neural plexus to either provide a substrate for nerve regeneration or allow documentation of further degeneration. This can usually be accomplished by removing biopsies from the upper sensory level, if present, and from one or two more proximal locations, keeping in mind that the loss of epidermal nerves is generally more severe than would be inferred from the clinical symptoms or findings. Future biopsies for comparison should be performed horizontally adjacent to the original biopsy at least 5 to 10 mm away from scarring and within the same peripheral nerve distribution. Normal epidermal nerve fiber values have been determined for proximal and distal anterolateral thigh, proximal and distal posterolateral calf, dorsum of the foot over the extensor digitorum brevis muscle, over the first dorsum interosseus muscle, proximal volar forearm, and the paraspinal area at T4-T5. The most frequently biopsied locations selected to assess a suspected length-dependent neuropathy are thigh and calf. The flexibility of the technique allows biopsy or blister to be performed safely almost anywhere on the body. Thus a localized sensory complaint (e.g., a thoracic radiculopathy) can be examined by biopsy of the symptomatic site and a mirror-image control, with postherpetic neuralgia being a possible exception because there may be contralateral changes in the unaffected side.76

METHODS OF CUTANEOUS NERVE ANALYSIS

Biopsy Methods

Choice of Biopsy/Blister Location The minimal invasiveness and negligible scarring remove most cosmetic objections to skin biopsy and skin blister procedures. When choosing a biopsy site, the susceptibility to nerve loss with neuropathy and the availability of normative data should be considered. Areas subject to trauma or with scars should be avoided. Site selection also depends upon the purpose for which the biopsy/blister is being considered (e.g., for diagnosis or for longitudinal study). Diagnosis of small fiber pathology can be made from a distal skin location where the clinical examination shows decreased sensation, particularly decreased sensitivity to mechanical (pin) and thermal (hot) pain stimuli. In patients with neuropathy, biopsies from the dorsum of the foot, distal calf, and proximal calf frequently contain a

Skin biopsy is commonly performed with a 3-mm diameter punch instrument (Fig. 34–4) (Acupunch; Acuderm, Inc., Ft. Lauderdale, FL) with local anesthetic. The wound rarely requires cautery or suture, and the infection rate in the lower extremity is less than 1:200. A shallow biopsy is adequate if interest is limited to epidermal nerves. A deeper biopsy (3 to 4 mm) is necessary to acquire sweat glands, the full hair follicle, subcutaneous fat, and larger arterioles. The biopsy is immediately placed into 4° C fixative and stored overnight, then transferred to 20% sucrose in phosphatebuffered saline for cryoprotection and storage. The biopsy can be held in sucrose in phosphate-buffered saline at 4° C for up to 1 month or frozen for more prolonged storage. Formaldehyde-based fixatives provide the best tissue and antigen preservation. Glutaraldehyde destroys the antigens and should not be used. Zamboni (2% paraformaldehyde and picric acid), Lana (4% formaldehyde and picric acid),

Pathology and Quantitation of Cutaneous Innervation

873

FIGURE 34–3 Cutaneous innervation of normal and diabetic subjects. Confocal microscope images of sectioned superficial skin of a normal (A and B) and a diabetic (C and D) subject. Epidermal nerve fibers arise from a subepidermal neural plexus (SNP) that extends horizontally immediately below the dermal-epidermal basement membrane. Numerous nerve fibers cross the basement membrane to innervate the epidermis of a control subject (A, thigh; B, calf). Diabetic subjects display a range of epidermal nerve fiber density depending on the body location and the degree of neuropathy. Skin biopsies from a diabetic subject have normal density of innervation in the thigh (C); however, several abnormalities are observed. Swellings in the nerve fibers are common in the SNP, and some nerve fibers are longer and more branched. The bulbous nerve endings are not seen at other depths of the dermis. The calf biopsy of the same subject contains no epidermal nerve fibers (D). A single nerve fiber in the SNP branches at the dermal-epidermal junction but does not cross into the epidermis. (From Kennedy, W. R., Wendelschafer-Crabb, G., and Walk, D.: Use of skin biopsy and skin blister in neurological practice. J. Clin. Neuromuscul. Dis. 1:196, 2000, with permission.) See Color Plate

and PLP (paraformaldehyde, lysine, and periodate) fixatives all provide optimal preservation. If these fixatives are not available, 10% formalin preservation for 12 to 18 hours is an alternative, but produces a more fragmented appearance of the epidermal nerves and is suboptimal. For multicenter studies or trials, in which the biopsy will be performed at a location geographically removed from the processing laboratory, the specimen should be placed into fresh cold fixative for 12 to 24 hours, then transferred into cryoprotectant for overnight shipping on wet ice.

Staining Skin Biopsy Specimens Nerve evaluation from skin biopsy relies on immunohistochemical localization of neural antigens within skin tissue sections. This is accomplished by applying a primary antibody directed against the antigen of interest followed by a labeled secondary antibody directed to the primary antibody. The major primary antibody used for the localization

of nerves in skin biopsy recognizes the pan-neuronal marker PGP 9.5 (Ultraclone, Wellow, UK; Chemicon, Temecula, CA) that is present in all cutaneous nerves. Localization of additional antigens can aid in diagnosis but may not be necessary in all situations. Several antigens can be localized within a single tissue section by using secondary antibodies that recognize and react with the immunoglobulin G (IgG) of the primary antibodies. Typically rabbit, mouse, and goat primary antibodies are localized with donkey antiserum specifically prepared to recognize IgG from only a single species (rabbit, mouse, or goat). Complementary markers such as fluorophores or enzymes label each of the speciesspecific secondary antibodies. Each fluorophore’s emission wavelength peak must be distinct so that it can be viewed individually with appropriate light filters (Fig. 34–5). Alternatively, enzymes such as peroxidase or alkaline phosphatase are reacted with a variety of available substrates to provide reaction products of different colors. All secondary antibodies should be derived from the same species

874

Pathology of the Peripheral Nervous System

Cy5

AMCA Cy2, FITC Cy3

400

500

600

700

Wavelength (nm)

FIGURE 34–4 Punch biopsy tool. A 3-mm punch biopsy is made on clean, shaved skin that has been swabbed with povidone-iodine (Betadine) and numbed with local anesthetic. The biopsy tool is inserted with a twisting motion to approximately half the depth of the metal collar. The depth of the biopsy depends on the thickness of the skin and the sample content desired. Sweat glands lie deep in the dermis and require a full-thickness biopsy, while shallow biopsies are sufficient for analysis of epidermal nerve density.

(e.g., donkey) and must be from a species different than hosts of the primary antibodies. Secondary antibodies used for multiple staining must be preadsorbed with IgG from all species other than the primary antibody being localized. Frozen 50- to 100-m-thick sections are cut perpendicular to the epidermal surface. The thicker sections provide greater sampling and visualization of anisotropic nerve fibers as they weave through the skin. Sections are floated in staining reagents containing Triton X-100 to promote penetration of antibodies. Samples are washed and antibodies are diluted in a solution of phosphate-buffered saline, 1% normal donkey serum, and 0.3% Triton X-100. Sections are incubated in 5% normal donkey serum so that nonspecific binding of IgG will be “blocked.” Sections are incubated in primary antibodies (e.g., rabbit anti–PGP 9.5 and mouse anti–type IV collagen; Chemicon) with gentle rotation for 5 hours at room temperature and overnight at 4° C, then washed in three changes of wash buffer, 1 hour each, with gentle rotation. This is followed with overnight incubation in labeled secondary antibody to the rabbit and mouse primary antibodies (Jackson ImmunoResearch, West Grove, PA), then washed three times. The secondary antibodies are labeled with either peroxidase or a stable fluorescent marker. Peroxidase markers are reacted enzymatically to produce a precipitate at the site of the antigen. The peroxidase reaction product is visible by bright-field microscopy and by electron microscopy. With the perox-

FIGURE 34–5 Fluorophore emission wavelength profile. Fluorophores with distinct wavelengths that can be isolated visually by use of optical filters are used to label antibodies for multiantigen immunofluorescent staining applications. Nerves are usually labeled with Cy 3 (fluorescence peaking at 565 nm), while basement membrane is labeled with Cy 2, fluorescing at 520 nm. By applying a blocking filter between 540 and 550 nm, the two antigens can be viewed independently; while applying a dual-band filter, the two antigens can be viewed simultaneously and distinguished microscopically by their colors. (Courtesy of Jackson ImmunoResearch, West Grove, PA.)

idase reaction, the basement membrane is localized as being just below the basal membrane of basal keratinocytes. Fluorescent markers require no further reaction before being studied by fluorescence or confocal microscopy. A new generation of fluorescent markers, Cy dyes 2, 3, and 5 (Jackson ImmunoResearch), are very bright and remain stable for several years in permanent mountants. The markers are useful for labeling two or three antigens within the same section. Labeling nerve with Cy 3, basement membrane with Cy 2, and endothelial cells with Ulex europaeus agglutinin type I with Cy 5 provides a vivid image of skin structure, innervation, and vasculature (Fig. 34–6). Non–immune serum specificity controls are essential and should be run with each biopsy. Accurate counts of epidermal nerve fibers, interpretation, and quantitation are difficult when sections fold or curl during drying of the epidermis. For optimal viewing and quantification, the epidermis should lie flat on the slide. Premounting sections in a small drop of Nobel agar prevents drying, adheres the sections, and holds the surface of the stratum corneum perpendicular to the slide. The premounted section is dehydrated in alcohol, cleared with methyl salicylate, and mounted on a coverslip in DPX (Fluka, Buchs, Switzerland). This provides a permanent preparation and is useful for slides labeled with Cy fluorophores and some enzyme reactions. Glycerol can be used to coverslip tissue without dehydrating the specimens and is useful for short-term storage.

Pathology and Quantitation of Cutaneous Innervation

875

FIGURE 34–6 Staining nerve and basement membrane. A, Nerves are immunostained with rabbit anti–PGP 9.5 and Cy 3. B, Basement membrane is immunostained with mouse anti–type IV collagen and Cy 2. C, Epidermis and vessels are stained with the lectin Ulex europaeus agglutinin type I and Cy 5. D, All images are combined in this colorized representation of the localization of nerves, vessels, dermal-epidermal basement membrane, and epidermis. Confocal images are acquired as gray-scale image stacks for each antigen-fluorophore combination. These are projected, pseudo colored, and combined to create a color image of the entire z series. See Color Plate

Skin Blister Method Skin blister is a painless, bloodless method to acquire, quantify, and plot distribution of epidermal nerve fibers.43 The blister roof, which consists of pure epidermis and epidermal nerve fibers, separates from the dermis on a plane between the basal membrane of basal keratinocytes and the dermal-epidermal basement membrane. Blisters are made by applying suction to a blister capsule (WR Medical Electronics, Stillwater, MN) that is adhered to shaved, cleansed skin with double-sided tape (Fig. 34–7). The surface of the capsule resting on skin contains one or more round openings to accommodate the desired number and size of blisters. Tegaderm tape (3M, Maplewood, MN) disks, cut slightly smaller than the size of the desired blister (e.g., 3-mm diameter), are placed on the skin surface to conform to the openings in the blister capsule. These prevent overstretch of the blister roof. The capsule is secured to skin with an elastic bandage, leaving the transparent top of the capsule uncovered to permit the examiner and subject to observe blister formation, and the capsule is evacuated to a negative pressure of 300 mm Hg. The small volume of the capsule makes it imperative to have tight seals to prevent leakage. A reservoir placed in series with the tubing between the pump and capsule enhances blister formation by equilibrating small leaks. We use a commercially available mechanical pump to evacuate the capsule,

but an inexpensive hand-held pump is adequate when a single capsule is used. Initially small blisters form, then these coalesce to form a full blister after 20 to 40 minutes. The time is less with increasing subject age and skin temperature, but also depends upon skin location and successful maintenance of negative pressure. Warming the area

FIGURE 34–7 A blister formed by application of negative pressure in the blister capsule. The blister conforms to the 3-mm opening in the base of the capsule. (From Kennedy, W. R., Nolano, M., WendelschaferCrabb, G., et al.: A skin blister method to study epidermal nerves in peripheral nerve disease. Muscle Nerve 22:360, 1999, with permission.)

876

Pathology of the Peripheral Nervous System

with a heating pad hastens blistering. After the capsule is removed, the blister roof with tape disk intact is excised with microscissors and placed in cold Zamboni fixative overnight at 4° C. Fixed blisters are cryoprotected with 20% sucrose in 0.1 M phosphate-buffered saline until processed. Prior to processing, blister roofs are frozen on dry ice. Specimens are processed as whole mounts without sectioning; otherwise the procedure is identical to that for thick sections.

Epidermal Sheet Preparations This is a modification of the standard punch technique that allows for separation of the epidermis from the dermis before fixation.5 The technique is particularly useful for examining the epidermis in the horizontal plane without dermal vessels or nerves. In addition, the epidermal sheet can be used for protein or messenger RNA (mRNA) assays, as described below, rather than immunocytochemical staining. Briefly, after obtaining the punch biopsy, the specimen is bathed in ethylenediaminetetraacetic acid for 2 hours; then the epidermis can be easily removed from the dermis, fixed, and immunostained as above. This preparation permits the identification of the nerve fiber terminals within the epidermis without blood vessels or the dermal structures. The sheet can be stained in exactly the same way to identify PGP-immunoreactive nerve fibers.

Polymerase Chain Reaction Techniques in Skin It is feasible to quantify neurotransmitter or neurotrophin mRNA levels in either the epidermis (using the sheet preparation) or the dermis (using the standard punch biopsy). In the field of HIV/acquired immunodeficiency syndrome (AIDS), there is interest in the concept that specific antiretrovirals may affect mitochondrial DNA levels by inhibition of gamma DNA polymerase.13 Subcutaneous fat is a rich source of mitochondria. We use the subcutaneous fat, which is usually trimmed and discarded from punch skin biopsies, to measure levels of mitochondrial DNA with real-time polymerase chain reaction23 in HIV-seropositive patients receiving antiretrovirals. Specific antiretrovirals, the dideoxynucleosides, reduce mitochondrial DNA levels by inhibition of gamma DNA polymerase,60 and thus this technique may provide a convenient measure of the severity of dideoxynucleoside toxicity and the metabolic complications common in individuals treated with these drugs long term.14

MICROSCOPY Thick sections are advantageous for viewing the full meandering course of nerves from the nerve trunk in the dermis throughout their ramifications to the endings in

sweat glands, arrector pili, hair follicles, arterioles, or epidermis. Sections are first scanned under a 10  objective in the microscope to evaluate the orientation and flatness of the section and staining of nerves, blood vessels, and cutaneous organs. The course and ending of single unmyelinated nerve fibers are better visualized with a 20  (or higher) objective, as, for example, when following epidermal nerve fibers through basement membrane and the pathway between keratinocytes to their endings under stratum corneum. Images taken from thick sections are often marred by out-of-focus blur because the focal plane of a 20  objective is about 2 m. Objects in 50- to 100-m sections that are above and below this focal plane are out of focus. Some laboratories use thinner sections to obtain sharp images by conventional microscopy. This is counterbalanced by the inability to follow nerves for more than a short distance in a section and reduced sampling. Confocal microscopy of thick sections is one solution to this problem, although the added complexity may not always be necessary for simple clinical questions. The confocal microscope takes in-focus images while retaining the advantages of working with thick sections. This is accomplished by “optically sectioning” the fluorescentlabeled thick sections into a series of smaller increments that correspond to the focal length of the lens being used. A confocal z series comprising multiple images taken at incremental focal planes throughout the tissue eliminates out-of-focus blur and provides an in-focus three-dimensional view of the tissue (Fig. 34–8). The operator can select the top and bottom of the section to be imaged, the size of the z-focus increments, and the fluorophores to image before collecting the z series. A laser scanning confocal microscope requires about 30 minutes to scan double-stained sections (two antibodies) at 2-m increments through 50 m of tissue. The nonlaser spinning disk confocal system (e.g., CARV) can collect equivalent images in about 3 minutes. The z series is projected into a single in-focus image for viewing. The series can also be “paged through” to follow nerves throughout the tissue, or the series can be rendered into a three-dimensional object.

QUANTITATION OF CUTANEOUS NERVES Epidermal Nerves Standard punch biopsies removed from anatomic locations that are often sampled to diagnose neuropathy contain an adequate sample of unmyelinated epidermal nerves and subepidermal neural plexus for detecting neuropathic changes. Epidermal innervation can be readily quantified using either confocal microscopy46,78 or direct counting of chromogen-stained specimens.16,32 The diagnosis of neuropathy is dependent upon finding

Pathology and Quantitation of Cutaneous Innervation

877

FIGURE 34–8 Confocal versus nonconfocal microscopy. The confocal microscope uses focused light beams (laser or mercury vapor) to optically section thick tissue sections, much like a computed tomography scanner. A series of many images are acquired at focal increments specific to the objective lens’ depth of field. Out-of-focus blur is eliminated, and clear, in-focus images of sections up to 200 m thick can be collected and used for three-dimensional analysis of nerve morphology and quantification.

a reduced density of epidermal nerves. Abnormalities of nerve morphology provide valuable secondary help. Epidermal nerve fibers can be quantified accurately because, after they leave the nerve bundles of the subepidermal nerve plexus and before they enter the epidermis, they separate as single nerve fibers. In thick sections the entire intraepidermal segment of most epidermal nerve fibers can be followed rather than isolated nerve segments, as is necessary in thinner sections. A set of almost identical counting rules evolved simultaneously in laboratories at both the University of Minnesota and Johns Hopkins University (Fig. 34–9): A. Count each nerve as it crosses the basement membrane of the epidermis. Because epidermal nerve fibers often branch below the basement membrane as well as in the epidermis, it is necessary to establish a standard counting site. Most investigators have agreed in practice to count as one unit each nerve that penetrates the basement membrane, even if there is branching within the epidermis. In peroxidase-stained sections, the membrane and cytoplasm of basal keratinocytes stain darker than surrounding tissue. Because the basal

FIGURE 34–9 Nerve counting rules. A set of concise rules was developed to normalize counting among investigators.

membrane is only a few micrometers above the dermalepidermal basement membrane, it can be used as the site for counting penetrating epidermal nerve fibers. In sections that are double-stained for PGP 9.5 in nerve and type IV or type VII collagen for basement membrane and labeled with compounds that fluoresce at different wavelengths, it is possible to directly observe the epidermal nerve fibers as they penetrate basement membrane. If confocal images of thick sections are available, the optical section in which the epidermal nerve fiber penetrates the basement membrane is located by paging through the z series of optical sections. B. Nerves that branch after crossing the basement membrane are counted as one. Epidermal nerve endings in normal subjects are usually quite simple—often free of branches or with minimal branching as they course through the epidermis. Nerves of patients with neuropathy frequently appear to have increased fiber

878

C.

D. E.

F.

G.

Pathology of the Peripheral Nervous System

branching. To address both nerve loss and collateralization of surviving axons, the number of fibers crossing the basement membrane is used to provide counts (how many nerves) while analysis of tracing data provides information about the complexity (length and branching pattern) of the nerve (how much nerve). Nerves that split below the basement membrane are counted as two units. Single epidermal nerve fibers arise as branches from nerve bundles of the subepidermal neural plexus. Each branch is counted as it penetrates the basement membrane. Nerves that appear to branch at the basement membrane are counted as two or more units. Nerve fragments that cross the basement membrane are counted. Epidermal nerve fibers that appear short are counted. They may be short fibers or the roots of fragments that terminate in adjacent sections. Nerve fibers that approach but do not cross the basement membrane are not counted. Nerve fibers that terminate immediately proximal to the basement membrane are common in neuropathy. They frequently follow the basement membrane a short distance before ending with a terminal swelling. Epidermal nerve fragments that do not cross the basement membrane are not counted. This rule has been followed in all publications by the Minnesota group. In previous publications by the Johns Hopkins group, such epidermal nerve fragment were included in the total numbers, but they will not be included in future counts.

Tracings made of epidermal nerve fibers with a software program (Neurolucida; MicroBrightField, Williston, VT) provide the number of nerve fibers, branch points, nerve fiber length, and coordinates of basement membrane penetration (Fig. 34–10). This information is useful because in some patients nerve branching and length increase in apparent compensation for a decrease in nerve number. The data are also useful for analysis of nerve distribution.

Results of epidermal nerve fiber density are commonly expressed as number of epidermal nerve fibers per millimeter of epidermis, providing a “linear density.” Comparison of such values from different reporting laboratories requires that the thickness of the section be known because thicker sections contain more nerve fibers. Results can also be reported as epidermal nerve fibers per square millimeter area of skin surface for easier comparison.32,46 Measurement accuracy is potentially adversely influenced by compression of the section between the slide and coverslip and by tissue shrinkage during preparation. For accurate comparison, these factors must be taken into consideration. Our laboratories report epidermal nerve fiber density based on nerve counts made in the equivalent of 50-m-thick sectioned tissue, making adjustments for changes resulting from dehydration and coverslip compression. Manual counting should be performed on at least three to four nonadjacent sections taken from different places within the biopsy. After manually counting the number of individual epidermal nerve fibers, the length of the epidermis is measured using Bioquant V (R&M Biometrics, Nashville, TN). Counting rules should be established a priori for each laboratory, and preferably only one observer used for research studies involving serial specimens. We have compared manual counting with light microscopy to stereologic techniques and demonstrated that the manual counting produced comparable densities but was much quicker.101 Reliability is highest when at least four sections at each anatomic site are counted. A technician can be trained to perform the manual quantitation, with excellent inter- and intraobserver reliability (0.95).61 We have also demonstrated that different laboratories can achieve reasonable interlaboratory reliability.98 For example, mean ( standard error of the mean) interobserver reliability was 12.2%  1.1% for each biopsy site and mean intraobserver variability was 10.2%  1.5% for individual sections. The variability between laboratories was 25.5% and the correlation coefficient for both intraobserver and

FIGURE 34–10 Nerve quantification by tracing. Neurolucida software (MicroBrightField, Williston, VT) uses the confocal image series to identify and trace each nerve fiber from the point where it crosses the basement membrane to its termination within the epidermis. Nerve number, length, complexity, and density as well as epidermal length and volume can be determined from the acquired data.

Pathology and Quantitation of Cutaneous Innervation

interobserver reliability was .98. The correlation coefficient between sites was .94. There was no relationship between absolute intraepidermal nerve fiber counts and reliability.

Normal Epidermal Nerve Density The density of epidermal nerves varies depending upon the biopsy site.40 In general, a consistent gradient in intraepidermal nerve fiber density exists from proximal to distal sites in the lower extremities, with minimal effects of race or sex. Significantly higher intraepidermal nerve fiber densities were noted in a younger group, age 10 to 19 years, but over the age of 20 years there is little change up to 80 years of age (Fig. 34–11).15,78 Stereologic quantitation has been compared with linear density estimation and is significantly associated with a correlation coefficient of .79 (P  .001). With formalin fixation (which tends to underestimate by about 20% compared to PLP fixation), the normative range for intraepidermal nerve fiber density (including intraepidermal fragments) in healthy normal controls is 21.1  10.4 fibers/mm (mean and standard deviation) in the thigh, with a fifth percentile of 5.2 fibers/mm. At the distal leg the normative range is 13.8  6.7, with a fifth percentile reading of 3.8 fibers/mm. Using the cutoff values for the fifth percentile for intraepidermal nerve fiber density at the distal part of the leg, the diagnostic efficiency (percentage correctly classified) was 88%, with a specificity (percentage true negative) of 97% and a sensitivity (percentage true positive) of 45%. Positive predictive value was 92% and negative predictive value was 90%. Although there is a relatively wide biologic variation in the intraepidermal nerve fiber density, the distribution of densities at the distal part of the leg permits the delineation of lower limit of normal

879

based on the fifth percentile value for healthy controls. The high specificity of the measure increases the positive predictive value of the test, which is therefore ideal for its anticipated use, namely to verify the presence of a disease for which there may be little clinical or electrophysiologic evidence, or when the clinician must be virtually certain of a diagnosis, for example, before initiating some form of therapy. The relatively low sensitivity of the measure implies that a normal intraepidermal nerve fiber density does not rule out the presence of a sensory neuropathy. Normative values of epidermal nerve fiber density for six body locations have been determined at the Kennedy laboratory (Table 34–1). Epidermal Nerves in Skin Blisters and Epidermal Sheets Whole-mounted blister and epidermal sheet preparations have sampling advantages over sectioned skin biopsies. The “bird’s-eye” perspective of epidermal nerve fibers in these 3-mm diameter (area  7.1 mm2) preparations presents for visualization and analysis the same number of nerves contained in the combined sections of a 3-mm biopsy (Fig. 34–12). Using a 20  objective lens, we count the epidermal nerve fibers in five representative segments of the blister roof, each measuring 0.4  0.3 mm. The 0.6-mm2 skin surface area is approximately 12% of the whole blister roof. The results, expressed as counts of epidermal nerve fibers per surface area of skin, vary according to body location in the same manner as that observed with biopsies. Epidermal nerve fiber density determinations from biopsies and blisters show a high degree of correlation (r  .71; P .005).49 The blister roof also provides an advantageous oversight of the horizontal territory of individual epidermal nerve fibers and the distribution of all epidermal nerve fibers within the specimen. Changes in epidermal nerve distribution caused by reorientation of nerves in their course through the epidermis by elongation, branching, or loss of branching probably occur normally during the continual regeneration of the epidermis, perhaps similar to normal reshaping of the terminals of motor axons.57 We suspect that exaggerated redistribution of epidermal nerves may be the earliest reaction of epidermal nerves to neuropathy (L. Waller, personal observations, 1998).

Other Cutaneous Nerves FIGURE 34–11 Effects of age on intraepidermal nerve fiber density at the distal part of the leg by age decile for healthy controls. (From McArthur, J. C., Stocks, E. A., Hauer, P., et al.: Epidermal nerve fiber density: normative reference range and diagnostic efficiency. Arch. Neurol. 55:1513, 1998, with permission.)

In general, dermal nerves are more difficult to quantify than epidermal nerve fibers. The subepidermal neural plexus that occupies the papillary dermis is often less dense in subjects with neuropathy. These bundled nerve fibers cannot be accurately counted or traced. The volume of nerve in the subepidermal neural plexus has been

880

Pathology of the Peripheral Nervous System

Table 34–1. Epidermal Nerve Fiber Density* Normal

Foot

Calf

Thigh

Hand

Forearm

Back

Mean 95% Cutoff Standard Deviation Number

22.83 12.80 7.61 39

18.58 6.73 6.54 39

34.99 19.60 15.88 39

24.57 10.40 11.31 38

37.57 17.10 12.96 37

62.61 33.00 21.25 23

*Normative data have been determined at the University of Minnesota for proximal and distal anterolateral thigh, proximal and distal posterolateral calf, dorsum of the foot over the extensor digitorum brevis muscle, over the first dorsum interosseus muscle of the hand, the proximal volar forearm, and the paraspinal area of the back at T4 to T5.

estimated based on the total fluorescence,55 but accurate quantification of nerves in the subepidermal neural plexus by thresholding from background is complicated by the positive immunoreactivity of dermal fibroblasts for PGP 9.5. A semiquantitative rating system can be used to assess

the density of the plexus (Fig. 34–13): 1  hyperinnervation, 0  normal, 1  mild (50%) loss, 2  severe (75%) loss, 3  a trace, and 4  no nerve. A similar semiquantitative rating system can be used to quantify sudomotor nerves. These rating systems require that the

FIGURE 34–12 Blister analysis. A, Blister profile: confocal image of an immunostained sectioned blister. Nerves (PGP 9.5) are green or yellow, and basement membrane (type IV collagen) is red. Epidermis separated from dermis just above the dermal-epidermal basement membrane. Dermal capillaries and the subepidermal neural plexus remained intact (left box). Epidermal nerves, severed from their proximal segment, remained in the blister roof (right box). Bar: 100 m. B, Survey confocal image of a 3-mm blister roof immunostained for nerve with antibody to PGP 9.5. Rectangles indicate the areas imaged at 20  for nerve counts. Bar: 500 m. C, Higher magnification image is used for quantification of epidermal nerve fibers (from right rectangle in B). Bar: 100 m. (From Kennedy, W. R., Nolano, M., Wendelschafer-Crabb, G., et al.: A skin blister method to study epidermal nerves in peripheral nerve disease. Muscle Nerve 22:360, 1999, with permission.) See Color Plate

Pathology and Quantitation of Cutaneous Innervation

881

FIGURE 34–13 Subepidermal neural plexus (SNP) rating. A semiquantitative rating system can be used to assess the density of the plexus. SNP  0 indicates a normal innervation pattern in the papillary dermis with nerve bundles running subjacent and parallel to the epidermis. SNP  1 denotes hyperinnervation, a condition that is encountered in early stages of neuropathy when active regeneration is occurring. It is often associated with morphologic abnormalities such as swellings, clustering of nerves, and truncation of nerve fibers at the dermal-epidermal basement membrane. SNP  1 signifies a moderate (50%) loss of SNP innervation, and is usually accompanied by a reduction in epidermal nerve fiber density. SNP  2 indicates a severe (75%) loss. SNP  3 applies when only a trace of SNP remains. SNP  4 indicates that no nerve remains in the papillary dermis. See Color Plate

viewer have an excellent interpretation of “normal.” Other techniques utilizing image analysis to quantify cutaneous nerves have recently appeared107 but have not yet been widely accepted. The density of sudomotor nerves around secretory tubules of sweat glands can be quantified by calculating nerve volume per sweat gland volume in confocal images.46 Sudomotor nerve density is below normal in patients with severe neuropathy (Fig. 34–14), whereas sweat glands often appear well innervated in mild or moderate neuropathy, even though the secretion of sweat is decreased. This suggests that quantitation of sudomotor nerves may not be helpful for early diagnosis of diabetic neuropathy. A possible explanation for this apparent discrepancy is that sweat glands in nonhairy human and murine skin receive multiple sudomotor nerves. Multiple innervation facilitates collateral reinnervation after a

partial nerve lesion,44 but sweat volume may be compromised. The occasional findings of increased sweat gland innervation46 and of increased sweat secretion58 in proximal skin locations of diabetic subjects may represent localized nerve growth in response to a general stimulus for collateral reinnervation during the early denervation stage of neuropathy. We have not attempted to quantify the innervation of arrector pilorum muscles, hair follicles, or arteries. Quantitation of the nerves to arrector pilorum muscles is possible with the quantitative methods described for sweat glands. This may be useful in the clinical diagnosis of peripheral autonomic nerve disease if adequate sampling of these muscles is achieved. Hair follicles have an abundant supply of associated myelinated and unmyelinated nerve fibers, but the relatively few follicles visualized in biopsy of nonhairy skin create a sampling problem. Evaluation for

882

Pathology of the Peripheral Nervous System

clawlike terminals. In general, these morphologic features, particularly the intra-axonal swellings, are more evident at the thigh than in regions below the knee (see Fig. 34–3C). The composition of intra-axonal swellings remains uncertain. Neurofilament stains are negative and electron microscopy studies to date have been relatively inconclusive, suggesting that the swellings represent collections of organelles. Because of their location proximal to sites with epidermal denervation, we believe that swellings represent predegenerative changes. In models of epidermal regeneration, clusters of terminal nerve fiber swellings appear more commonly in neuropathic individuals, suggesting that such clusters of swellings may represent aberrant efforts at nerve sprouting, or microneuromas. The interaction of nerve with basement membrane may be altered in neuropathy as evidenced by morphologic irregularities found in patient biopsies. Some nerve fibers follow a course along the dermal side of the basement membrane but do not penetrate, frequently ending in a bulblike swelling (see Fig. 34–3D). In patient biopsies with low nerve fiber density, nerves entering the epidermis often branch at the point of penetration through the basement membrane and form a tuft of nerve endings that extend into epidermis. Similarly, clustering of epidermal nerve fibers is frequently seen in disease states, wherein multiple nerve fiber penetrations are interspersed with relatively long segments of nerve-free epidermis. FIGURE 34–14 Sweat gland nerve density. These images depict the innervation of sweat glands from normal (top) and diabetic (bottom) human skin. Samples used for quantification are immunostained with anti–PGP 9.5 and imaged confocally. Nerve volume is computer-rendered.

clinical purposes may be possible in the scalp, depending upon the number of follicles available for analysis. The wide distribution of cutaneous arteries and arterioles and the variable number of nerves per vessel complicate attempts to quantify vasomotor nerves.

Morphologic Changes in Sensory Neuropathy In healthy subjects, epidermal fibers arise from subpapillary nerve fiber bundles and run toward the stratum corneum in either branched or unbranched patterns. The fibers are usually thin, are sometimes varicose, and often end with a single clublike enlargement. In cases of sensory neuropathy, intraepidermal nerve fibers show distinct morphologic changes. Fibers frequently show more tortuous courses and more complex ramification.16,53 In addition to the increased branching complexity, there often is increased varicosity with stubby, rounded projections; intra-axonal swellings; and

Correlation with Tests of Unmyelinated Nerve Function Relatively few studies have examined the relationship of epidermal nerve fiber densities to other tests of nerve function. Heat pain has been examined in humans using a chemical (topical capsaicin) denervation model. While no correlation was found between heat pain sensitivity and epidermal nerve fiber density using a large (900-mm2) diameter probe, statistically significant correlations between fiber density and heat pain sensitivity were found when a small (7.1-mm2) probe delivered the stimulus.50 This apparent incongruity is believed to result from the spatial summation of surviving epidermal nerves plus stimulation of the proximal stumps of degenerated epidermal nerves in the dermis (some of which may be sensitized by the degeneration of the nerve ending) that occurs when the larger probe heats a deeper area. The smaller probe heats a more isolated shallow area, with fewer nerve endings activated. Comparisons of intraepidermal nerve fiber densities with just noticeable difference (JND) thresholds for cooling and vibration levels were made in HIV-associated sensory neuropathies using the CASE IV instrument (WR Company, Stillwater, MN) and were inversely correlated. Because high JND indicates decreased sensitivity to sensory stimuli, it is not unexpected that the correlation estimates were negative.

Pathology and Quantitation of Cutaneous Innervation

A correlation analysis based upon the percentile distribution rather than the JND is another statistical approach, but given that the sample size was relatively small and that many of the subjects had mild neuropathy, this analysis was based on the categorization of subjects as normal/abnormal based on their JNDs. Surprisingly, intraepidermal nerve fiber density correlated inversely more closely with vibratory threshold than with cooling threshold. Vibration is subserved by A fibers that are not present in the epidermis while cool threshold detection is mediated by A fibers, which may be included in measurements of intraepidermal nerve fiber densities. An association with C fibers, which subserve noxious heat and mechanical pain and represent the bulk of epidermal nerve fibers, was not included in the QST battery. The inverse correlation of intraepidermal nerve fiber density and vibratory threshold may be partially explained by concurrent involvement of large and small fiber nerves. Furthermore, the fact that only a minority of subjects (43%) had abnormal epidermal nerve fiber densities at the distal leg at baseline suggests that the severity of the neuropathy in these subjects was relatively mild (and perhaps that the epidermal denervation in these subjects had not progressed above the ankle).82 In another study of 32 patients with painful, burning feet there was no concordance between the results from quantitative sudomotor axon reflex testing (QSART), a QST (vibration and cooling by CASE IV) performed on the upper and lower extremities, and skin biopsy from the distal calf, in which 28 (87.5%) had a reduction of intraepidermal nerve fibers.78 Intraepidermal nerve fiber densities have also been assessed in small fiber and autonomic neuropathies, and compared with autonomic screening tests. In a study that was designed to quantify the severity of autonomic impairment in patients with painful neuropathy, excellent correlation was observed between QSART and cooling abnormalities and loss of intraepidermal nerve fiber density.73 The QSART quantitatively evaluates the postganglionic sympathetic sudomotor axon58 (see also Postural Tachycardia Syndrome below).

883

densities were highly correlated (r  .87, P .0001). Intraepidermal nerve fiber density and sural nerve small fiber measures were concordant in 73% of patients. In 23% of cases, reduced intraepidermal nerve fiber density was the only indicator of small fiber depletion. It was usually normal in acquired demyelinating neuropathies and where the clinical suspicion for neuropathy was low. The inferences are that determination of distal leg intraepidermal nerve fiber density may be more sensitive than sural nerve biopsy in identifying small fiber sensory neuropathies and that intraepidermal nerve fiber density and assessments of large nerve fibers by nerve biopsy and electrophysiology are all useful for characterizing sensory and sensorimotor neuropathies.31

Utility of Skin Biopsy It is important in clinical practice to have a good sense of the utility (and limitations) of skin biopsy for evaluation of the peripheral nervous system. The biopsy can only provide information about the most distal terminals of sensory and autonomic nerves where they meet their target organ. Interpretation can be affected by local trauma, other diseases, or biopsy artifacts. Nonetheless, biopsy of proximal and distal sites can provide inbuilt intraindividual controls. We have found the punch skin biopsy technique to be clinically useful for defining presence and severity of focal neuropathy and of distally dominant sensory neuropathy and for differentiation of neuropathy from radiculopathy. It remains to be conclusively demonstrated that skin biopsy can track changes in neuropathies over time, or with regenerative treatments, although the technique has this potential.

CONTRIBUTIONS OF SKIN BIOPSY TO DIAGNOSIS Quantitation of epidermal nerve fiber density and morphology has been helpful for evaluating several clinical disorders.

Correlation between Epidermal Nerve and Sural Nerve

Painful Sensory Neuropathy

Twenty-six patients with neuropathic complaints had sural nerve morphometry and determination of epidermal nerve fiber density at the distal part of the leg. Nonparametric correlations were used because of the nonlinearity of the values. Intraepidermal nerve fiber density correlated with the densities of total myelinated fibers within the sural nerve (r  .57, P  .0011), small myelinated fibers (r  .53, P  .029), and large myelinated fibers (r  .49, P  .0054). There was a trend toward an association between intraepidermal nerve fiber density and sural nerve unmyelinated nerve fiber densities (r  .32, P  .054). Sensory nerve action potential amplitudes and large myelinated nerve fiber

In 20 patients with painful sensory neuropathies, intraepidermal nerve fiber density was significantly reduced compared with age-matched controls, even in regions proximal to the areas of clinically identifiable sensory abnormalities. For example, patients with grade 1 neuropathy in whom sensory abnormalities were restricted to the toes and feet had significantly reduced intraepidermal nerve fiber densities at the calf. Cooling thresholds on QST (CASE IV device) did not correlate with intraepidermal nerve fiber density or the clinical grade of sensory neuropathy (Fig. 34–15). Intraepidermal nerve fiber density from calf and thigh correlated against clinical estimates of neuropathy severity.35

884

Pathology of the Peripheral Nervous System

25

Axonai Density

IENF Density

40,000

20 15 10 5

30,000

20,000

10,000 0

0

25

The clinical features, natural history, and neuropathology were described for 32 patients presenting with “burning feet,” thought to represent idiopathic small fiber sensory neuropathies. Two clinical patterns were apparent based on natural history and anatomic location of cutaneous denervation. Most patients (28 of 32) presented with neuropathic pain that was initially restricted to the feet and toes, but eventually extended more proximally to involve the legs and hands. Intraepidermal nerve fiber density was most severely reduced distally, with more normal intraepidermal nerve fiber densities at proximal sites. The minority (4 of 32) presented with abrupt onset of generalized cutaneous burning pain and hyperesthesia. In these patients intraepidermal nerve fiber densities were reduced at both proximal and distal sites. For the entire group, intraepidermal nerve fiber densities were reduced at the calf site below the fifth percentile in 81% of all patients.34 Figure 34–16 compares nerve densities in biopsies of the skin and sural nerve from one patient from each group. Skin biopsy was obtained in a separate study of 57 patients with painful, burning feet; minimal signs of neuropathy; and normal nerve conduction studies. Of these, 44 (77%) had a reduced number of intraepidermal nerves at the distal calf. Some nerves ended abruptly just below the basement membrane; others had enlarged, swollen terminal nerve endings. Reduced ankle reflexes were found in 11%, pinprick loss on toes and feet in 45%, and loss of vibration sense in 43%. Only three patients had conditions known to be associated with peripheral neuropathy.78

40,000 20

Axonai Density

IENF Density

FIGURE 34–15 Intraepidermal nerve fiber density from calf (black circles) and thigh (white squares) skin correlated against clinical estimates of neuropathy severity in patients with painful sensory neuropathy. The data shown are group means with standard error bars. (From Holland, N. R., Stocks, A., Hauer, P., et al.: Intraepidermal nerve fiber density in patients with painful sensory neuropathy. Neurology 48:708, 1997, with permission.)

15 10 5 0

30,000

20,000

10,000

0

Distal Proximal Relative Position of skin Biopsy Skin fibers

Unmyelinated

FIGURE 34–16 Intraepidermal nerve fiber (IENF) densities at various sites for chronic progressive idiopathic small fiber sensory neuropathy (SFSN) (left) and sural nerve morphometry for chronic progressive idiopathic SFSN patient 1 (top right) and patient 2 (bottom right). Left, IENF density (mm 1) at various sites for patients compared with normal controls (mean  standard error shown by shaded rectangle and 5th and 95th percentile range shown by bars). Right, Unmyelinated fiber density (axons/mm2) for patients compared to the 5th and 95th percentile range for normal controls (bars). Patient 1 (top) had normal densities of myelinated and unmyelinated fibers in the sural nerve biopsy specimen (taken from the calf) despite marked cutaneous denervation in skin biopsies. Patient 2 (bottom), with more severe progressive idiopathic SFSN, had reduced densities of smalldiameter myelinated and unmyelinated axons in the sural nerve biopsy specimen (taken from the ankle) in addition to marked length-dependent cutaneous denervation. (From Holland, N. R., Stocks, A., Hauer, P., et al.: Intraepidermal nerve fiber density in patients with painful sensory neuropathy. Neurology 48:708, 1998, with permission.)

Diabetic Neuropathy Earlier work in painful diabetic neuropathy suggested that there was a predominance of involvement of unmyelinated and small myelinated nerve fibers in the sural nerve.8 Small fibers have long been recognized to be involved in diabetic polyneuropathy, and skin biopsies have confirmed that epidermal denervation can occur relatively early, and generally

Pathology and Quantitation of Cutaneous Innervation

0.06

0.25

Nerve vol/sweat gland vol Nerve length/epidermal vol

0.20

0.05 0.04

0.15

0.03 0.10

0.02 0.05

0.00

0.01

0 (Mild)

25

50

75

Clinical Score

100

0.00 125

Length of epidermal nerve (μm) per epidermal volume (μm3)

Sweat gland nerve volume per volume of sweat gland

in a length-dependent manner. The total length of epidermal nerve per volume of epidermis correlates with the overall clinical severity (Fig. 34–17). Some of the surviving epidermal nerves have morphologic abnormalities, particularly axon swellings.47,80 The subepidermal neural plexus is often thinner than in control skin, and fibers within the subepidermal plexus may have a thickened, dystrophic appearance. In some neuropathies single nerve fibers are seen ending in a terminal enlargement just under the basement membrane.46 These fibers appear to be unsuccessfully attempting to penetrate basement membrane en route to regeneration into the epidermis, although it is also possible that they are in the process of dying back. Sweat gland innervation is sometimes abnormal or completely lost in diabetic patients (Fig. 34–17), but it appears not to be useful as an early indication of neuropathy.46 Epidermal nerve fibers also appear to be prominently affected in patients with impaired glucose tolerance (IGT). Several groups have reported an increased prevalence of IGT among patients with painful sensory neuropathy,74,97 and randomly selected patients with sensory neuropathy and IGT were found to have reduced epidermal nerve fiber densities.99 Of 97 patients with clinically diagnosed predominantly sensory neuropathy of undetermined etiology evaluated with oral glucose tolerance tests, 36% were classed

(Severe)

FIGURE 34–17 Neuropathy score correlated with epidermal nerve fiber density and sweat gland innervation density in diabetic subjects. Correlation of cutaneous nerve quantitation with clinical assessment. The amount of nerve in calf biopsies correlated with the clinical evaluation of most diabetic subjects. The data suggest that the epidermal nerve assessment is useful in evaluating mild to moderate neuropathy, and sweat gland nerve analysis is helpful in later stages of neuropathy. Reduction of epidermal nerve is seen in mild cases of neuropathy, whereas no epidermal nerve is seen in severe cases. Sweat gland innervation remains at normal levels until neuropathy is quite severe. The peripheral and autonomic systems of the diabetic subjects were evaluated from graded results of history and examination, nerve conduction, cardiorespiratory reflexes, warm and cold sensitivity, and sweat testing. (Kennedy, W. R., and Wendelschafer-Crabb, G.: Utility of skin biopsy in diabetic neuropathy. Semin. Neurol. 16:163, 1996, with permission.)

885

as having IGT and 20% as having diabetes. Both the IGT and diabetes groups had reduced distal leg intraepidermal nerve fiber densities. The neuropathy associated with IGT was less severe than the neuropathy associated with diabetes. Patients with predominantly small fiber neuropathy had a shorter duration of neuropathy symptoms, and most had IGT. The results suggest that small-caliber nerve fiber loss may be the earliest detectable sign of neuropathy in glucose dysmetabolism.103 Diabetic Truncal Neuropathy In diabetic truncal neuropathy, skin biopsies from symptomatic regions show a loss of intraepidermal nerve fibers. After clinical recovery there may be nerve fiber regeneration.53

HIV-Associated Sensory Neuropathies Sensory neuropathies of HIV/AIDS affect at least 30% of individuals with AIDS.90 The distal sensory polyneuropathy associated with HIV overlaps clinically with the toxic neuropathy provoked by specific antiretrovirals. The sensory neuropathies produce paresthesias or burning pain in the feet, often associated with hyperalgesia and lightning pains. Unlike HIV-associated dementia, the incidence rates have not declined with the advent of highly active antiretroviral therapy, probably because specific dideoxynucleoside analogues (ddI, d4T, and ddC) produce peripheral neurotoxicity.66 Punch skin biopsies were originally developed at the Johns Hopkins University laboratory to study HIV-associated sensory neuropathies that had previously been shown by sural nerve biopsy to be associated with prominent involvement of unmyelinated nerve fibers.17 The most systematic survey of skin biopsies in HIV-associated sensory neuropathy comes from a trial of recombinant human NGF. Skin biopsies were included in this trial as an outcome measure, the first reported application of skin biopsy to measure treatment effect. Sixty-two of 270 patients with sensory neuropathies who participated in the trial of recombinant human NGF were included in a substudy examining the density of intraepidermal nerve fibers.62 Fiber density was inversely correlated with neuropathic pain as measured both by patient and physician global pain assessments, but not using the Gracely Pain Scale. The reproducibility of the technique, as assessed from intrasubject correlation between baseline and the week 18 densities, was 81% in the distal part of the leg and 77% in the upper thigh. Decreased intraepidermal nerve fiber density at the distal leg was associated with lower CD4 counts and higher plasma HIV RNA levels. There was no treatment effect over the relatively short period of 18 weeks, and a few biopsies studied out to 70 weeks remained stable over time. Although no significant treatment effect was observed over the 18 weeks, the reproducibility of the technique appears to be good, suggesting that this technique could be incorporated as an outcome measure in future trials of regenerative agents for sensory neuropathies.

886

Pathology of the Peripheral Nervous System

Friedreich’s Ataxia Friedreich’s ataxia is an autosomal recessive inherited ataxia. Hyperexpression of a GAA trinucleotide on the causative gene (X25) localized to chromosome 9 results in reduction of the protein frataxin.9 Pathology of large-diameter myelinated nerve fibers in peripheral nerves is consistent with the loss of vibratory sense and deep tendon reflexes and severely reduced sensory action potentials. Small fiber involvement was postulated but unproven19 until severe loss of unmyelinated sensory nerves and of the autonomic innervation to sweat glands, arrector pili, and arterioles was uncovered in thick sections of skin biopsies.70 The small fiber loss correlated with reduced mechanical and thermal nociception.71

Restless Legs Syndrome The restless legs syndrome (RLS) is characterized by a compelling urge to move the extremities. It is often associated with paresthesias or dysesthesias, motor restlessness, worsening of symptoms with rest, relief by activity, and worsening of symptoms in the evening or night. Two forms are described: crurum dolorosum, characterized by pain, and crurum paresthetica, characterized by paresthesias.21 Numerous risk factors are reported for RLS, and there is a suggestion that early-onset (age 45 or younger) and late-onset RLS differ etiologically.1,2 Nerve biopsies performed in an unselected group of primary patients suggest that the incidence of peripheral neuropathy may be underestimated.37 Twenty-two consecutive patients with RLS were evaluated for evidence of large fiber neuropathy (LFN) and small sensory fiber loss (SSFL). Neuropathy was identified in eight (36%). Three patients had pure LFN, two had mixed LFN and SSFL, and three had isolated SSFL. Patients with SSFL had a later onset, reported pain in their feet more frequently, and tended not to have a family history of RLS. Patients without SSFL did not associate onset with age, family history, or presence of pain. The report suggested that two forms of RLS exist: one is triggered by painful dysesthesias associated with SSFL and has a later onset and a negative family history; the other does not have small sensory fiber involvement, has an earlier onset and a positive family history, and is without pain.79

Familial Dysautonomia (Riley-Day Syndrome, Hereditary Sensory and Autonomic Neuropathy Type III) Familial dysautonomia3,86 is an autosomal dominant disorder of sensory and autonomic nerves found in Ashkenazi Jews that is caused by mutation of the IKBKAP gene on chromosome 9q31.6 Clinical features include increased sweating, absent overflow lacrimation, vomiting crises, decreased sensitivity to pain, depressed

Achilles reflexes, postural hypotension, and absence of tongue fungiform papillae and of the axon flare response to intradermal histamine. Biopsy of glabrous skin shows severe loss of unmyelinated sensory and autonomic nerves at the calf and moderate loss in the paraspinal region of the upper back.33

Congenital Insensitivity to Pain with Anhidrosis (Hereditary Sensory and Autonomic Neuropathy Type IV) Congenital insensitivity to pain with anhidrosis (CIPA) is a rare condition caused by mutation of the TrkA (NTRK1) gene on the 1q 21-22 chromosome,38 characterized by mental retardation; congenital analgesia that leads to selfmutilation, multiple scars, and fractures; and anhidrosis with repeated bouts of fever.87 Nerve biopsy shows loss of unmyelinated and small myelinated fibers.25 Recently a 10-year-old girl was reported with a low IQ, pain, and thermal insensitivity. The remainder of the neurologic examination, sensory nerve conduction velocities, cardiovascular reflexes, and visual, brainstem, and somatosensory evoked potentials were normal. Sural nerve biopsy was devoid of small myelinated nerves and contained only rare unmyelinated nerves. Epidermal nerve fibers were essentially absent in skin from the thigh, back, and calf. The few sweat glands present were hypotrophic and without innervation.69

Psoriasis Psoriasis is a common skin disorder characterized histologically by epidermal hyperproliferation, dilated tortuous capillaries, and a lymphocytic infiltrate. Skin biopsy shows that the epidermal nerves are greatly elongated through a thickened epidermis to the stratum corneum.39 Several treatment modalities are available: therapy that targets epidermal proliferation, immunomodulators, or physical treatment, such as the use of lasers. The pulsed dye laser at 585-nm wavelength targets blood vessels and selectively destroys dilated vessels in the dermal papillae. Laser therapy is followed by remodeling of the epidermis and subsequent regeneration of capillaries and epidermal nerves with return to a more normal appearance.118

Port-Wine Stains It is proposed that the pathogenesis of port-wine stains might be related to a lack of innervation around the ectatic blood vessels.89 Biopsy of the involved skin shows a deficiency of innervation that is inversely proportional to the size of the vascular spaces. Moreover, nerve density may be predictive of the response of port-wine stain lesions to laser treatment.117

887

Pathology and Quantitation of Cutaneous Innervation

Leprosy

Postherpetic Neuralgia

Severe loss of immunoreactivity to PGP 9.5 and neurofilament antibodies occurs in leprosy. Loss is greatest in the tuberculoid variety of leprosy, less in lepromatous, and least in the indeterminate variety.41 Curiously, there is almost complete loss of reactivity for the neuropeptides CGRP, SP, and neuropeptide Y. This raises the possibility that neuropathy may be detected early by loss of immunoreactivity to some antigens while reactivity to other antigens remains, at least temporarily.

Shingles represents reactivation of latent varicella-zoster virus in a sensory ganglion, causing loss of a variable proportion of the sensory neurons. In some patients, neuropathic pain persists for months or years (postherpetic neuralgia). Hyperalgesia and allodynia to light touch are often prominent. Two studies have demonstrated that the density of epidermal nerves in areas of allodynia is reduced, although they came to differing conclusions. One study found that higher fiber densities correlated with pain, an observation that might suggest that intact but hyperactive nociceptors were responsible. The reduction in intraepidermal nerve fiber density correlated with the magnitude of thermal sensory deficits.88 By contrast, a second study75 found that pain correlated with lower intraepidermal nerve fiber densities, and interpreted the results as more compatible with central sensitization. Interestingly, there was also a reduction of intraepidermal nerve fiber density in the contralateral, unaffected epidermis.115

Fabry’s Disease Fabry’s disease is an X-linked recessive disorder caused by deficiency of -galactosidase A activity. Hemizygotes develop deposits of neutral glycosphingolipids, principally ceramide trihexoside, throughout the nervous system but predominantly in vascular endothelial cells. A painful small fiber neuropathy develops that is difficult to detect and quantitate by conventional methods. The neuropathy can develop in children as young as 5 years of age, with characteristic episodes of acral burning pain. Twenty Fabry’s disease patients (hemizygotes, ages 19 to 56 years) with preserved renal function were found to have normal nerve conduction studies and large fiber quantitation by sural nerve biopsy. By contrast, involvement of small cutaneous fibers in these patients was easily demonstrated and quantified by punch skin biopsy. All patients showed severe loss of intraepidermal innervation at the distal part of the leg, with a density of 0 to 2.4 fibers/mm versus a density for control subjects at this site of 4.7 to 6.5 fibers/mm. Fiber loss at the distal thigh was proportionately less severe. Fabry’s disease patients underwent biopsy at 6-month intervals for periods ranging from 1 to 3 years with no significant longitudinal changes in density except in two patients who demonstrated a rapid decrement in innervation density following an increase in spontaneous pain.92

Sensory Ganglionopathies Ganglionopathy patients tend not to show the normal gradient in intraepidermal nerve fiber density between the proximal thigh and the distal part of the leg—that is, they show a non–length-dependent process. One cohort presented with predominant gait and limb ataxia and proprioceptive sensory loss; 8 of 16 patients had positive sensory symptoms. Causes of the ganglionopathy were paraneoplastic in six and idiopathic in seven, and one had a hereditary sensory and autonomic neuropathy. In ganglionopathies, the mean intraepidermal nerve fiber density did not differ between the proximal thigh (10.37/mm) and the distal leg (10.41/mm). Compared to other sensory neuropathies and controls, the intraepidermal nerve fiber density was significantly lower at the proximal thigh site in patients with ganglionopathies.54

CADASIL Cerebral autosomal dominant arteriopathy with subcortical infarct and leukoencephalopathy (CADASIL) is one example of how skin biopsy can be used for analysis of nonneural substances in investigation of neurologic disease. CADASIL is an autosomal dominant disorder that presents as migrainous headaches, cerebral ischemia, and multi-infarct dementia. An abnormality on chromosome 19q12 produces a missense point mutation in the notch3 gene. Punch skin biopsy has been used to identify characteristic granular, electron-dense, osmiophilic material attached to vascular smooth muscle cells.109 However, a commercial laboratory test for the genetic abnormality is now available from Athena Labs (Worcester, MA), which should provide improved specificity and sensitivity.30

Chronic Inflammatory Demyelinating Polyneuropathy Patients with chronic inflammatory demyelinating polyneuropathy (CIDP) were recently shown to have about 50% reduction in intraepidermal nerve fiber density compared to age- and gender-matched controls, suggesting that smalldiameter sensory nerves as well as large-caliber nerve fibers are affected in CIDP.15

Postural Tachycardia Syndrome A collaborative study by Johns Hopkins University and the Mayo Clinic examined the relationship between the QSART and epidermal nerve fiber densities in patients with postural tachycardia syndrome (POTS).96 Inclusion criteria for POTS in the study included orthostatic heart

888

Pathology of the Peripheral Nervous System

rate increment (from 30 beats/min to an absolute heart rate of 120 beats/min within 5 minutes of head-up tilt), symptoms of orthostatic intolerance, distal abnormal QSART, and abnormalities or loss of vasoconstrictor response. In POTS, distal autonomic fibers are affected in isolation, sparing somatic fibers, which have consistently normal nerve conduction studies.91 Intraepidermal nerve fiber density correlated with peripheral postganglionic sudomotor status as assessed by the Composite Autonomic Severity Score (CASS)102 both in the category of the sudomotor subscores and in the overall total score, which is a composite index of combined sudomotor, adrenergic, and cardiovagal function. In POTS the mean intraepidermal nerve fiber density was normal, while the CASS was abnormal. Although three of the eight patients showed some minor morphologic abnormalities, the normalcy of the epidermal nerve fiber densities suggests that distal involvement of autonomic fibers occurs in isolation in POTS.

Pediatric Neurologic Disorders Skin biopsies have been less widely studied in pediatric neurologic disorders, but there are some conditions in which examination of the cutaneous innervation has characteristic changes. Two examples are provided. Giant Axonal Neuropathy Giant axonal neuropathy is an autosomal recessive neurologic disorder characterized by the development of a severe polyneuropathy, central nervous system abnormalities, and characteristic tightly curled hair. Mutations in the gigaxonin gene have been identified as the underlying genetic defect.51 Ultrastructurally, accumulations of neurofilaments and osmiophilic aggregates are found in giant axons. Intermediate filaments accumulate in Schwann cells, perineural cells, fibroblasts, and endothelial and epithelial cells in both nerve and skin biopsies.104 We have observed giant swellings within dermal nerve fibers, although the density of epidermal nerves appears to be normal. Neuroaxonal Dystrophies Neuronal intranuclear inclusion disease has no defined metabolic abnormalities, but it affects the peripheral nervous system, which could therefore be used to identify morphologic abnormalities.24 In infantile neuroaxonal dystrophy, there is also an accumulation of intermediate neurofilaments, as well as mitochondria and organelles with vesiculotubular profiles.59 There is limited experience with the use of skin biopsies in this condition; however, some children have shown dystrophic changes in the subepidermal neural plexus.

Other Conditions Depletion of epidermal nerve fibers has been demonstrated in a variety of conditions with dermatologic manifestations. Image analysis detected a significant decrease in PGP 9.5–immunoreactive nerves in the epidermis and subepidermal layers and in CGRP-immunoreactive nerves around the capillaries within dermal papillae of patients with Raynaud’s phenomenon, although the changes were not always apparent by visual screening.105 Patients with systemic sclerosis were found to have decreased CGRP- and PGP 9.5–immunoreactive nerves in all skin areas and vasoactive intestinal polypeptide–immunoreactive nerves in the sweat glands.105 An increased number of PGP 9.5–immunoreactive fibers has been reported in the dermis and dermal-epidermal junction in atopic dermatitis, a condition in which the sensation of itching is prominent.108 It has been suggested that increased dermal innervation could explain the pruritus, burning pain, and hyperalgesia in notalgia paresthetica,100 and that an increase of sensory neuropeptides SP and CGRP might be involved in the pathogenesis of nodular prurigo.64 No changes of innervation were found in patients with lichenified eczema.64

RESEARCH USES OF SKIN BIOPSY The existence of unmyelinated sensory nerves within a few microns of the skin’s surface has provided opportunities to study the reactions of peripheral nerves to mechanical, thermal, and chemical stimuli under safe, minimally invasive circumstances. Early experiments, using methylene blue staining, localized the itching sensation caused by cowphage spicules and injected proteases to the subepidermal neural plexus.93 Use of immunohistochemical staining has allowed expansion of research into the correlates of sensory testing, effects of mechanical and chemical trauma to cutaneous nerves, and regeneration of sensory nerves.

Human Models of Nerve Regeneration Skin Biopsy Variations We developed two models to study reinnervation of the epidermis.83 One model uses a circular incision that transects the subepidermal plexus, resulting in Wallerian degeneration of the nerve fibers that enter the incised cylinder, leaving a defined zone of denervated dermis and epidermis. The earliest reinnervation of epidermis occurred by collateral sprouting from the terminals of epidermal axons from just outside the incision line. These terminals extended horizontally across the incision line and through the superficial layers of the epidermis, beneath the stratum corneum. By 13 days, numerous regenerating axons appeared in the deeper dermis derived

889

Pathology and Quantitation of Cutaneous Innervation

After Capsaicin Treatment

During Capsaicin Treatment % Deviation from normal where normal = 100 (treated/control area)

from transected axons. The latter regenerating axons grew toward and ultimately into the epidermis, so that epidermal axonal density had normalized by 30 to 75 days. The invasion of these axons was associated with regression of the horizontally growing collateral sprouts. The second model utilizes an identical incision followed by removal of the incised cylinder of skin, leaving a denervated area in which Schwann cells are absent. New fibers arose by terminal elongation of the epidermal axons outside the incision line, as in the incision model, and especially by collateral branching of epidermal fibers at the incision margins. These collaterals reached the epidermal surface of the basal lamina at the dermal-epidermal junction, and then grew slowly toward the center of the denervated circle. In contrast to the first model, complete reinnervation was not achieved even after 23 months. These and other models might be used to study reinnervation of denervated skin in humans in different injury models and have relevance for exploring the stimuli for axonal growth and remodeling.

700 600 500 400 300 200 100 0

Heat (Latency) Cold (Latency)

100 80 60 40 20 0 time 0

D3 W1 W2 Heat Pain (VAS) Cold Pain (VAS) ENF Density

W3

W1

W2 W3 Touch Pain (%) Pain (VAS)

W6

Skin Blister: A Model of Mechanical Denervation Mechanical denervation of the terminal 40 to 60 m of epidermal sensory nerve fibers can be produced to study wound healing and sensory reinnervation by removing the epidermis with a skin blister.49 Regeneration of new epidermis occurs within 3 days after a 3-mm blister wound. The new epidermis is initially innervated by collateral sprouts arising from epidermal nerve fibers in the normal skin surrounding the blistered area. Slightly later, the proximal stumps of epidermal nerves at the base of the blister begin to enter the new epidermis, first near the edge of the lesion and later at the center. Eventually, regenerated nerves from the proximal stumps in the base reinnervate the epidermis in near-normal numbers and the collateral branches disappear.112 Reinnervation is accompanied by return of sensation within 30 days. Similar reinnervation occurs in patients with diabetic neuropathy, even in areas of reduced sensation, but the return of epidermal nerve fibers is only to the level of the prelesion innervation.42

FIGURE 34–18 Capsaicin-induced denervation and reinnervation. The mean ( standard error) change in sensation and number of nerve fibers expressed as percent deviation from normal during and following the course of topical capsaicin application. Note that scale for latencies (top) is different from that for other sensitivities and nerve fiber numbers (bottom). Touch was measured using Semmes Weinstein monofilaments. Intensity of pain was measured using a visual analogue scale (V.A.S.) for mechanical pain, heat pain, and cold pain. Mechanical pain (pain % and pain V.A.S.) was evaluated with a spring-mounted pin. Heat pain (V.A.S.) and cold pain (V.A.S) were tested with a 2-mm probe. Epidermal nerve fiber (ENF) number per area was calculated from analysis of nerves in blister roofs. (From Nolano, M., Simone, D. A., Wendelschafer-Crabb, G., et al.: Topical capsaicin in humans: parallel loss of epidermal nerve fibers and pain sensation. Pain 81:135, 1999, with permission.)

Capsaicin: A Model for Chemical Denervation When capsaicin, a pungent ingredient in hot chili peppers, is applied topically, it causes warm, painful sensations. These are replaced after several hours by a period of hyposensitivity. Sensation returns toward normal in 2 to 3 weeks.95 This sequence was initially believed to result from desensitization of nociceptive receptors but was later shown to be induced by the influx of calcium ions with osmotic changes as well as activation of calcium-sensitive proteases resulting in axonal degeneration.113 More recently, skin biopsy showed that capsaicin administered intradermally is soon followed by depletion of epidermal nerves and superficial dermal nerves within 24 hours (Fig. 34–18). Reinnervation of the epidermis

reached approximately 50% of normal, and heat pain sensation reached 60% of normal after 4 weeks.94 Capsaicin administered topically also causes loss of epidermal and superficial dermal nerves and decreased sensitivity to noxious heat and mechanical (pin) stimulation. Serial biopsies reveal that nerve regeneration after topical capsaicin is faster than after intradermal delivery. The return of epidermal nerves and sensation is nearly complete in 30 days (Fig. 34–19).72 These findings strengthen the hypothesis that epidermal nerves are polymodal nociceptors. The detection of a loss and subsequent recovery of sensitivity to noxious heat was possible when stimuli were delivered by a small probe (2- to 3-mm diameter) but not by the larger 30  30-mm probe of the type common on

890

Pathology of the Peripheral Nervous System

Diabetic subjects

100 μm

Healthy control subjects

150

100

% Baseline ENFD 50

0 0

25

50

Days

100

0

25

50

75

100

Days

FIGURE 34–20 ENFD regeneration expressed as a percent of baseline values in diabetic and healthy, nondiabetic control subjects. Healthy volunteers recovered their ENF density at a rate of 0.177  0.075 fibers/mm/day. The rate of nerve fiber regeneration was significantly decreased among those with diabetes (0.074  64 fibers/mm/day; P .001) and further reduced among diabetic subjects with neuropathy as compared to those without neuropathy (P  .003). (From Polydefkis, M., Hauer, P., Sheth, S., et al.: The time course of epidermal nerve fiber regeneration: Studies in normal controls and in people with diabetes, with and without neuropathy, Brain 127:1606, 2004, with permission.)

Control

Day 6

75

Kennedy & Wendelschafer- Crabb. University of Minnesota

FIGURE 34–19 Capsaicin-treated human epidermis. Nerves degenerate following topical capsaicin treatment, as can be seen by comparing control tissue at time 0 (top) with a sample taken after 1 week of capsaicin application (bottom). Epidermal nerve fibers have degenerated, while fibers in the subepidermal neural plexus are still plentiful. This provides a good model for the peripheral neuropathy seen in diabetic subjects, in whom epidermal nerve fiber loss precedes more proximal loss. We will use this model to study basement membrane during nerve regeneration. This model provides denervation without remodeling basement membrane. Bar: 100 m. See Color Plate

commercial instruments.50 Both of our groups have developed reproducible systems for applying capsaicin and examining the patterns and rate of regrowth of epidermal nerve fibers. In normal volunteers treated with topical capsaicin, epidermal nerve density returns at an average rate of 0.177 fibers/day (range, 0.023 to 0.362 fibers/day). The rate of nerve fiber regeneration is significantly decreased in diabetics (0.074 fibers/day; range, 0.002 to 0.258 fibers/day, P .001), and there was a trend for a lower rate of regeneration with increasing severity of neuropathy (Fig. 34–20).81a These models have potential as outcome measures in trials of regenerative agents and for improving understanding of trophic factor regulation after nerve injury.

Comparison of Nerve Function with Structure Comparisons have been made between the morphology of cutaneous nerve receptors and the results of electrophysiologic stimulation. Highly significant correlations were found between the density of normal Meissner’s corpuscles and proximally recorded sensory nerve action potential areas whether generated by electrical stimulation of the digital nerve or by tactile stimulation of the fingerpad. The density of abnormal Meissner’s corpuscles correlated significantly with the sensory action potential area but not with tactile potential area, whereas the density of papillary myelinated nerve fibers correlated with tactile but not with electrical stimuli. The natural stimuli appeared to be transmitted centrally more effectively by normal receptors than by atrophic receptors.71

Animal Models Using Skin Biopsy Several groups have now used skin biopsy techniques in animal models to examine denervation and regeneration after various surgical models of nerve injury.67,68,36,56,115 Early experiments showed that rat sciatic nerve transection resulted in denervation, thinning of the epidermis, and increased expression of PGP 9.5 by Langerhans cells.36 A mouse model has been devised to correlate the reinnervation time of muscles, sweat glands, and skin with the time required for return of function to these organs (Fig. 34–21). After sciatic nerve crush, sequential biopsies of foot muscles and footpads from the hind paw of the same mouse can be examined to detect the time of reappearance of alpha motor, sudomotor, and epidermal nociceptor nerves. Reappearance

891

Pathology and Quantitation of Cutaneous Innervation

of end-organ function can be tested at the same time intervals by sciatic nerve stimulation recording muscle, pilocarpine (intraperitoneal) stimulation to detect reappearance of sweat droplets, and pinprick to skin to detect return of nociception.68 The model has potential for testing the efficacy of therapeutic agents to facilitate nerve regeneration. Reinnervation of epidermal nerves in pigskin was tested after removal of a 3-mm skin biopsy and immediate replacement (autotransplantation). The first nerves that entered the newly regenerated epidermis over the wound were collateral branches of epidermal nerves in the normal uninjured epidermis surrounding the wound. These nerves grew on the epidermal surface of the basement membrane toward the center of the wound, then turned at an acute angle toward the stratum corneum. Nerve regeneration from the severed subepidermal plexus nerve trunks along the sides of the lesion began much later (weeks) and extended slowly toward the center of the lesion. Reinnervation never approached normal.29 The same experiment on one human subject gave identical results (M. Nolano, personal observation, 1997). An important experimental study demonstrated that intraepidermal nerve fiber density is significantly reduced after dorsal root ganglionectomy or sciatic nerve transection in rats, but not after dorsal and ventral radiculotomy, sympathectomy, or spinal motor neuron lesion.56

ACKNOWLEDGMENT A portion of this research is supported by grant NS44807 (J.M.).

REFERENCES FIGURE 34–21 Confocal micrographs of PGP 9.5 immunofluorescence in control (A) and reinnervating (B and C) mouse footpads following sciatic nerve crush. A, Nerves in control footpad form a dense network. Sweat glands are very heavily innervated. Large trunks extend through the central sweat gland area to form the subepidermal nerve plexus and Meissner’s corpuscles in the tufts of papillary dermis. The epidermis has many fine nerve endings. B, At day 14 postcrush, PGP 9.5–immunoreactive fibers returned in large nerve trunks from the base of the pad to the papillary dermis. C, By day 46 postcrush, the pattern of innervation resembles the control, but regenerated nerve fibers to the nerve trunks, subepidermal neural plexus, and sweat glands are less well compartmentalized, and the epidermis has shorter nerve fibers. (From Navarro, X., Verdu, E., Wendelschafer-Crabb, G., and Kennedy, W. R.: Immunohistochemical study of skin reinnervation by regenerative axons. J. Comp. Neurol. 380:164, 1997, with permission.)

1. Allen, R., LaBuda, M., Becker, P., and Earley, C.: Family history study of RLS patients from two clinical populations. Sleep Res. 26:309, 1997. 2. Allen, R. P., and Earley, C. J.: Defining the phenotype of the restless legs syndrome (RLS) using age-of-symptom-onset. Sleep Med. 1:11, 2000. 3. Axelrod, F. B., and Hilz, M. J.: Familial dysautonomia. In Appenzeller, O. (ed.): Handbook of Clinical Neurology: The Autonomic Nervous System, Part II. Dysfunctions. Amsterdam, Elsevier Science, p. 144, 2000. 4. Bajrovic, F., Kovacic, U., Pavcnik, M., and Sketelj, J.: Interneuronal signaling is involved in induction of collateral sprouting of nociceptive axons. Neuroscience 111:587, 2002. 5. Baker, K. W., and Habrowsky, J. E.: EDTA separation and ATPase Langerhans staining in the mouse epidermis. J. Invest. Dermatol. 80:625, 1983. 6. Blumenfeld, A., Slaugenhaupt, S. A., Liebert, C. B., et al.: Precise genetic mapping and haplotype analysis of the familial dysautonomia gene on human chromosome 9q31. Am. J. Hum. Genet. 64:1110, 1999.

892

Pathology of the Peripheral Nervous System

7. Bolton, C. F., Winkelmann, R. K., and Dyck, P. J.: A quantitative study of Meissner’s corpuscles in man. Neurology 16:1, 1966. 8. Brown, M. J., Martin, J. R., and Asbury, A. K.: Painful diabetic neuropathy: a morphometric study. Arch. Neurol. 33:164, 1976. 9. Campuzano, V., Montermini, L., Molto, M. D., et al.: Friedreich’s ataxia: autosomal recessive disease caused by an intronic GAA triplet repeat expansion. Science 271:1423, 1996. 10. Caterina, M. J., Leffler, A., Malmberg, A. B., et al.: Impaired nociception and pain sensation in mice lacking the capsaicin receptor. Science 288:306, 2000. 11. Caterina, M. J., Rosen, T. A., Tominaga, M., et al.: A capsaicin-receptor homologue with a high threshold for noxious heat. Nature 398:436, 1999. 12. Cauna, N.: The free penicillate nerve endings of the human hairy skin. J. Anat. 115:277, 1973. 13. Chen, C. H., Vazquez-Padua, M., and Cheng, Y. C.: Effect of anti-human immunodeficiency virus nucleoside analogs on mitochondrial DNA and its implication for delayed toxicity. Mol. Pharmacol. 39:625, 1991. 14. Cherry, C. L., Gahan, M. E., McArthur, J. C., et al.: Exposure to dideoxynucleosides is reflected in lowered mitochondrial DNA in subcutaneous fat. J. Acquir. Immune Defic. Syndr. 30:271, 2002. 15. Chiang, M. C., Lin, Y. H., Pan, C. L., et al.: Cutaneous innervation in chronic inflammatory demyelinating polyneuropathy. Neurology 59:1094, 2002. 16. Chien, H. F., Tseng, T. J., Lin, W. M., et al.: Quantitative pathology of cutaneous nerve terminal degeneration in the human skin. Acta Neuropathol. (Berl.) 102:455, 2001. 17. Cornblath, D. R., and McArthur, J. C.: Predominantly sensory neuropathy in patients with AIDS and AIDS-related complex. Neurology 38:794, 1988. 18. Dalsgaard, C. J., Rydh, M., and Haegerstrand, A.: Cutaneous innervation in man visualized with protein gene product 9.5 (PGP 9.5) antibodies. Histochemistry 92:385, 1989. 19. Dyck, P. J.: Neuronal atrophy and degeneration predominantly affecting peripheral sensory and autonomic neurons. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1065, 1993. 20. Dyck, P. J., Winkelmann, R. K., and Bolton, C. F.: Quantitation of Meissner’s corpuscles in hereditary neurologic disorders: Charcot-Marie-Tooth disease, Roussy-Levy syndrome, Dejerine-Sottas disease, hereditary sensory neuropathy, spinocerebellar degenerations, and hereditary spastic paraplegia. Neurology 16:10, 1966. 21. Ekbom, K.: Restless legs. Acta Med. Scand. Suppl. 158:1, 1945. 22. Fundin, B. T., Arvidsson, J., Aldskogius, H., et al.: Comprehensive immunofluorescence and lectin binding analysis of intervibrissal fur innervation in the mystacial pad of the rat. J. Comp. Neurol. 385:185, 1997. 23. Gahan, M. E., Miller, F., Lewin, S. R., et al.: Quantification of mitochondrial DNA in peripheral blood mononuclear cells and subcutaneous fat using real-time polymerase chain reaction. J. Clin. Virol. 22:241, 2001.

24. Goebel, H. H.: Extracerebral biopsies in neurodegenerative diseases of childhood. Brain Dev. 21:435, 1999. 25. Goebel, H. H., Veit, S., and Dyck, P. J.: Confirmation of virtual unmyelinated fiber absence in hereditary sensory neuropathy type IV. J. Neuropathol. Exp. Neurol. 39:670, 1980. 26. Gold, M. S., Reichling, D. B., Shuster, M. J., and Levine, J. D.: Hyperalgesic agents increase a tetrodotoxin-resistant Na current in nociceptors. Proc. Natl. Acad. Sci. U. S. A. 93:1108, 1996. 27. Griffin, J. W., McArthur, J. C., and Polydefkis, M.: Assessment of cutaneous innervation by skin biopsies. Curr. Opin. Neurol. 14:655, 2001. 28. Halata, Z.: Sensory innervation of the hairy skin (light- and electronmicroscopic study). J. Invest. Dermatol. 101:75S, 1993. 29. He, C. L., Wendelschafer-Crabb, G., and Kennedy, W. R.: Reinnervation of epidermal nerve fibers in pig skin. Soc. Neurosci. Abstr. 22:760, 1996. 30. Helm, T., and Helm, K. F.: CADASIL: blood test versus skin biopsy? J. Am. Acad. Dermatol. 46:798, 2002. 31. Herrmann, D. N., Griffin, J. W., Hauer, P., et al.: Epidermal nerve fiber density and sural nerve morphometry in peripheral neuropathies. Neurology 53:1634, 1999. 32. Hilliges, M., and Johansson, O.: Comparative analysis of numerical estimation methods of epithelial nerve fibers using tissue sections. J. Peripher. Nerv. Syst. 4:53, 1999. 33. Hilz, M. J., Kennedy, W. R., Stemper, B., et al.: Cutaneous innervation in familial dysautonomia. Clin. Auton. Res. 11:182, 2001. 34. Holland, N. R., Crawford, T. O., Hauer, P., et al.: Smallfiber sensory neuropathies: clinical course and neuropathology of idiopathic cases. Ann. Neurol. 44:47, 1998. 35. Holland, N. R., Stocks, A., Hauer, P., et al.: Intraepidermal nerve fiber density in patients with painful sensory neuropathy. Neurology 48:708, 1997. 36. Hsieh, S. T., Choi, S., Lin, W. M., et al.: Epidermal denervation and its effects on keratinocytes and Langerhans cells. J. Neurocytol. 25:513, 1996. 37. Iannaccone, S., Zucconi, M., Marchettini, P., et al.: Evidence of peripheral axonal neuropathy in primary restless legs syndrome. Mov. Disord. 10:2, 1995. 38. Indo, Y., Tsuruta, M., Hayashida, Y., et al.: Mutations in the TRKA/NGF receptor gene in patients with congenital insensitivity to pain with anhidrosis. Nat. Genet. 13:485, 1996. 39. Johansson, O., Han, S. W., and Enhamre, A.: Altered cutaneous innervation in psoriatic skin as revealed by PGP 9.5 immunohistochemistry. Arch. Dermatol. Res. 283:519, 1991. 40. Johansson, O., Wang, L., Hilliges, M., and Liang, Y.: Intraepidermal nerves in human skin: PGP 9.5 immunohistochemistry with special reference to the nerve density in skin from different body regions. J. Peripher. Nerv. Syst. 4:43, 1999. 41. Karanth, S. S., Springall, D. R., Lucas, S., et al.: Changes in nerves and neuropeptides in skin from 100 leprosy patients investigated by immunocytochemistry. J. Pathol. 157:15, 1989.

Pathology and Quantitation of Cutaneous Innervation 42. Kennedy, W. R., Brown, J., and Wendelschafer-Crabb, G.: Sensory reinnervation in diabetic neuropathy. J. Peripher. Nerv. Syst. 6:150, 2001. 43. Kennedy, W. R., Nolano, M., Wendelschafer-Crabb, G., et al.: A skin blister method to study epidermal nerves in peripheral nerve disease. Muscle Nerve 22:360, 1999. 44. Kennedy, W. R., and Sakuta, M.: Collateral reinnervation of sweat glands. Ann. Neurol. 15:73, 1984. 45. Kennedy, W. R., and Wendelschafer-Crabb, G.: The innervation of human epidermis. J. Neurol. Sci. 115:184, 1993. 46. Kennedy, W. R., and Wendelschafer-Crabb, G.: Utility of skin biopsy in diabetic neuropathy. Semin. Neurol. 16:163, 1996. 47. Kennedy, W.R., Wendelschafer-Crabb, G. and Johnson, T.: Quantitation of epidermal nerves in diabetic neuropathy. Neurology 47:1042, 1996. 48. Kennedy, W. R., Wendelschafer-Crabb, G., and Walk, D.: Use of skin biopsy and skin blister in neurological practice. J. Clin. Neuromuscul. Dis. 1:196, 2000. 49. Kennedy, W .R., Wendelschafer-Crabb, G., and Lindall, A. W.: Quantitative epidermal nerve fiber analysis in skin blisters and skin biopsies. Ann. Neurol. 14:573, 1999. 50. Khalili, N., Wendelschafer-Crabb, G., Kennedy, W. R., and Simone, D. A.: Influence of thermode size for detecting heat pain dysfunction in a capsaicin model of epidermal nerve fiber loss. Pain 91:241, 2001. 51. Kuhlenbaumer, G., Young, P., Oberwittler, C., et al.: Giant axonal neuropathy (GAN): case report and two novel mutations in the gigaxonin gene. Neurology 58:1273, 2002. 52. Lauria, G.: Innervation of the human epidermis: a historical review. Ital. J. Neurol. Sci. 20:63, 1999. 53. Lauria, G., McArthur, J. C., Hauer, P. E., et al.: Neuropathological alterations in diabetic truncal neuropathy: evaluation by skin biopsy. J. Neurol. Neurosurg. Psychiatry 65:762, 1998. 54. Lauria, G., Sghirlanzoni, A., Lombardi, R., and Pareyson, D.: Epidermal nerve fiber density in sensory ganglionopathies: clinical and neurophysiologic correlations. Muscle Nerve 24:1034, 2001. 55. Levy, D. M., Terenghi, G., Gu, X. H., et al.: Immunohistochemical measurements of nerves and neuropeptides in diabetic skin: relationship to tests of neurological function. Diabetologia 35:889, 1992. 56. Li, Y., Hsieh, S. T., Chien, H. F., et al.: Sensory and motor denervation influence epidermal thickness in rat foot glabrous skin. Exp. Neurol. 147:452, 1997. 57. Lichtman, J. W., Magrassi, L., and Purves, D.: Visualization of neuromuscular junctions over periods of several months in living mice. J. Neurosci. 7:1215, 1987. 58. Low, P. A., Caskey, P. E., Tuck, R. R., et al.: Quantitative sudomotor axon reflex test in normal and neuropathic subjects. Ann. Neurol. 14:573, 1983. 59. Mahadevan, A., Santosh, V., Gayatri, N., et al.: Infantile neuroaxonal dystrophy and giant axonal neuropathy—overlap diseases of neuronal cytoskeletal elements in childhood? Clin. Neuropathol. 19:221, 2000. 60. Martin, J. L., Brown, C. E., Matthews-Davis, N., and Reardon, J. E.: Effects of antiviral nucleoside analogs on

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73. 74.

75.

76.

77.

893

human DNA polymerases and mitochondrial DNA synthesis. Antimicrob. Agents Chemother. 38:2743, 1994. McArthur, J. C., Stocks, E. A., Hauer, P., et al.: Epidermal nerve fiber density: normative reference range and diagnostic efficiency. Arch. Neurol. 55:1513, 1998. McArthur, J. C., Yiannoutsos, C., Simpson, D. M., et al.: A Phase II trial of nerve growth factor for sensory neuropathy associated with HIV infection. AIDS Clinical Trials Group Team 291. Neurology 54:1080, 2000. McCarthy, B. G., Hsieh, S. T., Stocks, A., et al.: Cutaneous innervation in sensory neuropathies: evaluation by skin biopsy. Neurology 45:1848, 1995. Molina, F. A., Burrows, N. P., Jones, R. R., et al.: Increased sensory neuropeptides in nodular prurigo: a quantitative immunohistochemical analysis. Br. J. Dermatol. 127:344, 1992. Molliver, D. C., Wright, D. E., Leitner, M. L., et al.: IB4binding DRG neurons switch from NGF to GDNF dependence in early postnatal life. Neuron 19:849, 1997. Moore, R. D., Wong, W. M., Keruly, J. C., and McArthur, J. C.: Incidence of neuropathy in HIV-infected patients on monotherapy versus those on combination therapy with didanosine, stavudine and hydroxyurea. AIDS 14:273, 2000. Navarro, X., Verdu, E., Wendelschafer-Crabb, G., and Kennedy, W. R.: Innervation of cutaneous structures in the mouse hind paw: a confocal microscopy immunohistochemical study. J. Neurosci. Res. 41:111, 1995. Navarro, X., Verdu, E., Wendelschafer-Crabb, G., and Kennedy, W. R.: Immunohistochemical study of skin reinnervation by regenerative axons. J. Comp. Neurol. 380:164, 1997. Nolano, M., Crisci, C., Santoro, L., et al.: Absent innervation of skin and sweat glands in congenital insensitivity to pain with anhidrosis. Clin. Neurophysiol. 111:1596, 2000. Nolano, M., Provitera, V., Crisci, C., et al.: Small fibers involvement in Friedreich’s ataxia. Ann. Neurol. 50:17, 2001. Nolano, M., Provitera, V., Lullo, F., et al.: Tactile stimulation and mechanoreceptors in sensory neuropathies. Neurol. Sci. 22(Suppl. 1):31, 2001. Nolano, M., Simone, D. A., Wendelschafer-Crabb, G., et al.: Topical capsaicin in humans: parallel loss of epidermal nerve fibers and pain sensation. Pain 81:135, 1999. Novak, V., Freimer, M. L., Kissel, J. T., et al.: Autonomic impairment in painful neuropathy. Neurology 56:861, 2001. Novella, S. P., Inzucchi, S. E. and Goldstein, J. M.: The frequency of undiagnosed diabetes and impaired glucose tolerance in patients with idiopathic sensory neuropathy. Muscle Nerve 24:1229, 2001. Oaklander, A. L.: The density of remaining nerve endings in human skin with and without postherpetic neuralgia after shingles. Pain 92:139, 2001. Oaklander, A. L., Romans, K., Horasek, S., et al.: Unilateral postherpetic neuralgia is associated with bilateral sensory neuron damage. Ann. Neurol. 44:789, 1998. Ochoa, J. L., and Yarnitsky, D.: The triple cold syndrome: cold hyperalgesia, cold hypoaesthesia and cold skin in peripheral nerve disease. Brain 117(Pt. 1):185, 1994.

894

Pathology of the Peripheral Nervous System

78. Periquet, M. I., Novak, V., Collins, M. P., et al.: Painful sensory neuropathy: prospective evaluation using skin biopsy. Neurology 53:1641, 1999. 79. Polydefkis, M., Allen, R. P., Hauer, P., et al.: Subclinical sensory neuropathy in late-onset restless legs syndrome. Neurology 55:1115, 2000. 80. Polydefkis, M., Hauer, P., Abraham, B., et al.: Novel measures of human axonal regeneration. J. Peripher. Nerv. Syst. 6:170, 2001. 81. Polydefkis, M., Hauer, P., Griffin, J. W., and McArthur, J. C.: Skin biopsy as a tool to assess distal small fiber innervation in diabetic neuropathy. Diabetes Technol. Ther. 3:23, 2001. 81a. Polydefkis, M., Hauer, P., Sheth, S., et al.: The time course of epidermal nerve fiber regeneration: Studies in normal controls and in people with diabetes, with and without neuropathy. Brain 127:1606, 2004. 82. Polydefkis, M., Yiannoutsos, C. T., Cohen, B. A., et al.: Reduced intraepidermal nerve fiber density in HIV-associated sensory neuropathy. Neurology 58:115, 2002. 83. Rajan, B., Polydefkis, M., Hauer, P., et al.: Epidermal reinnervation after intracutaneous axotomy in man. J. Comp. Neurol. 457:24, 2003. 84. Ren, K., and Dubner, R.: Descending modulation in persistent pain: an update. Pain 100:1, 2002. 85. Rice, F. L., Fundin, B. T., Arvidsson, J., et al.: Comprehensive immunofluorescence and lectin binding analysis of vibrissal follicle sinus complex innervation in the mystacial pad of the rat. J. Comp. Neurol. 385:149, 1997. 86. Riley, C. M., Day, R. L., Greely, D., and Langford, W. S.: Central autonomic dysfunction with defective lacrimation. Pediatrics 3:468, 1949. 87. Rosenberg, S., Nagahashi Marie, S. K., and Kliemann, S.: Congenital insensitivity to pain with anhidrosis (hereditary sensory and autonomic neuropathy type IV). Pediatr. Neurol. 11:50, 1994. 88. Rowbotham, M. C., Yosipovitch, G., Connolly, M. K., et al.: Cutaneous innervation density in the allodynic form of postherpetic neuralgia. Neurobiol. Dis. 3:205, 1996. 89. Rydh, M., Malm, M., Jernbeck, J., and Dalsgaard, C. J.: Ectatic blood vessels in port-wine stains lack innervation: possible role in pathogenesis. Plast. Reconstr. Surg. 87:419, 1991. 90. Schifitto, G., McDermott, M. P., McArthur, J. C., et al.: Incidence of and risk factors for HIV-associated distal sensory polyneuropathy. Neurology 58:1764, 2002. 91. Schondorf, R., and Low, P. A.: Idiopathic postural orthostatic tachycardia syndrome: an attenuated form of acute pandysautonomia? Neurology 43:132, 1993. 92. Scott, L. J., Griffin, J. W., Luciano, C., et al.: Quantitative analysis of epidermal innervation in Fabry disease. Neurology 52:1249, 1999. 93. Shelley, W. B., and Arthur, R. P.: The neurohistology and neurophysiology of the itch sensation in man. Arch. Dermatol. Res. 76:296, 1957. 94. Simone, D. A., Nolano, M., Johnson, T., et al.: Intradermal injection of capsaicin in humans produces degeneration and subsequent reinnervation of epidermal nerve fibers: correlation with sensory function. J. Neurosci. 18:8947, 1998. 95. Simone, D. A., and Ochoa, J.: Early and late effects of prolonged topical capsaicin on cutaneous sensibility and neurogenic vasodilatation in humans. Pain 47:285, 1991.

96. Singer, W., Spies, J. M., Hauer, P., et al.: Epidermal nerve fiber density in the neuropathic type of the postural tachycardia syndrome [S35.001]. Presented at the annual meeting of the American Academy of Neurology, San Diego, 2000. 97. Singleton, J. R., Smith, A. G., and Bromberg, M. B.: Increased prevalence of impaired glucose tolerance in patients with painful sensory neuropathy. Diabetes Care 24:1448, 2001. 98. Smith, A. G., Kroll, R., Ramachandran, P., et al.: The reliability of skin biopsy with measurement of intraepidermal nerve fiber density [A500]. Neurology 58(Suppl. 3):S66, 2002. 99. Smith, A. G., Ramachandran, P., Tripp, S., and Singleton, J. R.: Epidermal nerve innervation in impaired glucose tolerance and diabetes-associated neuropathy. Neurology 57:1701, 2001. 100. Springall, D. R., Karanth, S. S., Kirkham, N., et al.: Symptoms of notalgia paresthetica may be explained by increased dermal innervation. J. Invest. Dermatol. 97:555, 1991. 101. Stocks, E. A., McArthur, J. C., Griffen, J. W., and Mouton, P. R.: An unbiased method for estimation of total epidermal nerve fibre length. J. Neurocytol. 25:637, 1996. 102. Suarez, G. A., Opfer-Gehrking, T. L., Offord, K. P., et al.: The Autonomic Symptom Profile: a new instrument to assess autonomic symptoms. Neurology 52:523, 1999. 103. Sumner, C. J., Sheth, S., Griffin, J. W., et al.: The spectrum of neuropathy in diabetes and impaired glucose tolerance. Neurology 60:108, 2003. 104. Taratuto, A. L., Sevlever, G., Saccoliti, M., et al.: Giant axonal neuropathy (GAN): an immunohistochemical and ultrastructural study report of a Latin American case. Acta Neuropathol. (Berl.) 80:680, 1990. 105. Terenghi, G., Bunker, C. B., Liu, Y. F., et al.: Image analysis quantification of peptide-immunoreactive nerves in the skin of patients with Raynaud’s phenomenon and systemic sclerosis. J. Pathol. 164:245, 1991. 106. Thompson, R. J., Doran, J. F., Jackson, P., et al.: PGP 9.5—a new marker for vertebrate neurons and neuroendocrine cells. Brain Res. 278:224, 1983. 107. Underwood, R. A., Gibran, N. S., Muffley, L. A., et al.: Color subtractive-computer-assisted image analysis for quantification of cutaneous nerves in a diabetic mouse model. J. Histochem. Cytochem. 49:1285, 2001. 108. Urashima, R., and Mihara, M.: Cutaneous nerves in atopic dermatitis: a histological, immunohistochemical and electron microscopic study. Virchows Arch. 432:363, 1998. 109. Walsh, J. S., Perniciaro, C., and Meschia, J. F.: CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy): diagnostic skin biopsy changes determined by electron microscopy. J. Am. Acad. Dermatol. 43:1125, 2000. 110. Wang, L., Hilliges, M., Jernberg, T., et al.: Protein gene product 9.5-immunoreactive nerve fibres and cells in human skin. Cell Tissue Res. 261:25, 1990. 111. Watkins, L. R., and Maier, S. F.: Beyond neurons: evidence that immune and glial cells contribute to pathological pain states. Physiol. Rev. 82:981, 2002. 112. Wendelschafer-Crabb, G., Nolano, M., and Kennedy, W. R.: Sensory reinnervation of a skin blister wound. J. Peripher. Nerv. Syst. 6:150, 2001.

Pathology and Quantitation of Cutaneous Innervation 113. Wood, J. N., Winter, J., James, I. F., et al.: Capsaicininduced ion fluxes in dorsal root ganglion cells in culture. J. Neurosci. 8:3208, 1988. 114. Woolf, C. J., Shortland, P., and Coggeshall, R. E.: Peripheral nerve injury triggers central sprouting of myelinated afferents. Nature 355:75, 1992. 115. Wu, G., Ringkamp, M., Hartke, T. V., et al.: Early onset of spontaneous activity in uninjured C-fiber nociceptors after injury to neighboring nerve fibers. J. Neurosci. 21:RC140, 2001.

895

116. Wu, G., Ringkamp, M., Murinson, B. B., et al.: Degeneration of myelinated efferent fibers induces spontaneous activity in uninjured C-fiber afferents. J. Neurosci. 22:7746, 2002. 117. Zelickson, B. D., Kilmer, S. L., Bernstein, E., et al.: Pulsed dye laser therapy for sun damaged skin. Lasers Surg. Med. 25:229, 1999. 118. Zelickson, B. D., Mehregan, D. A., Wendelschafer-Crabb, G., et al.: Clinical and histologic evaluation of psoriatic plaques treated with a flashlamp pulsed dye laser. J. Am. Acad. Dermatol. 35:64, 1996.

35 Nerve Conduction and Needle Electromyography JUN KIMURA

Nerve Conduction Studies Fundamentals of Nerve Conduction Studies Basic Principles Motor Nerve Conduction Studies Sensory Nerve Conduction Studies Nerve Conduction in the Clinical Domain Commonly Assessed Nerves Cranial Nerves Major Nerves in the Upper Limb Nerves of the Shoulder Girdle Cutaneous Nerves in the Forearm Major Nerves in the Lower Limb Nerves of the Pelvic Girdle Human Reflexes and Late Responses Blink Reflex H and T Reflexes Masseter Reflex F Waves A Wave Clinical Applications Common Sources of Error

Collision Technique Anomalous Innervations Temporal Dispersion and Phase Cancellation Practical Assessment of Conduction Block: Criteria for Conduction Block Long and Short Segmental Nerve Conduction Studies

Needle Electromyography Potentials Associated with Needle Insertion Injury Potentials End-Plate Activities Myotonic Discharge Spontaneous Activity Fibrillation Potentials Positive Sharp Waves Single-Fiber Discharge and Denervation Complex Repetitive Discharges Fasciculation Potentials and Myokymic Discharges

Nerve Conduction Studies FUNDAMENTALS OF NERVE CONDUCTION STUDIES Basic Principles Nerve conduction studies play an important role in the assessment of peripheral nerve function. They are based on the simple principle, that electrical stimulation initiates nerve impulses, which propagate along motor or sensory nerve fibers. Standardization of both stimulating and

Continuous Muscle Fiber Activity Cramp Motor Unit Action Potential Amplitude and Area Rise Time Duration Phases Abnormalities of Motor Unit Potentials Lower Motor Neuron Versus Myopathic Disorders Discharge Pattern of Motor Units Recruitment Measurements of Turns and Amplitude Lower and Upper Motor Neuron Disorders Myopathy Involuntary Movement Other Aspects of Electromyography Principles of Localization Sequence of Abnormalities Summary

recording events are needed to get accurate and reproducible results and to avoid pitfalls. The study of the motor fibers depends on the analysis of compound evoked potentials recorded from the muscle. The function of the sensory fibers can be tested by recording sensory action potentials from the nerve itself. Nerve conduction studies have become a reliable test to assess peripheral nerve dysfunction and serve as a natural extension of the clinical evaluation. If conducted in the proper clinical context, the method provides a means of confirming the clinical diagnosis, localizing and characterizing the lesion, and quantifying the degree of involvement. 899

900

Nerve Conduction and Electromyography

Stimulus Intensity and Possible Risk To fully activate a healthy nerve, sufficient stimulation must be provided. Generally, it takes surface stimulation of 0.1 ms duration and 100 to 300 V or, assuming a tissue resistance of 10 k, 10 to 30 mA. However, as much as 400 to 500 V, or 40 to 50 mA, may be necessary to study diseased nerves with decreased excitability. Electrical stimulation within this intensity range ordinarily poses no particular risk to the patient. Routine nerve conduction studies, however, may be contraindicated in certain situations. For example, the entire current may reach the cardiac tissue in patients with indwelling catheters or central venous pressure lines inserted directly into the heart. Similarly, electric stimulation near the heart poses safety hazards in patients with cardioverters and defibrillators. Placing the stimulator at least 6 inches away reduces the chance of externally triggering the sensors.272 If necessary, a cardiologist with special expertise in electrophysiology can provide advice regarding the feasibility of a nerve conduction study in these patients, and the possible need to turn off such medical instruments during the procedure. Surface and Needle Electrodes Surface electrodes, in general, are better than needle electrodes for recording a compound muscle action potential, because they register the total contributions from all discharging motor units. The onset latency of an evoked potential recorded with surface electrodes refers to the conduction time of the fastest motor fibers, and the amplitude provides a measure of the number of axons innervating the muscle. In contrast, a needle electrode registers activity from only a small portion of the muscle, but may record action potentials from atrophic muscles when surface recording fails. The use of a needle electrode also helps to selectively record an action potential solely from a target muscle when proximal stimulation excites many other muscles simultaneously. An averaging technique, though usually unnecessary for recording muscle potentials, may help to identify markedly reduced responses from small atrophic muscles.14 Most electromyographers use surface electrodes for recording sensory and mixed nerve action potentials. A pair of ring electrodes placed over the proximal and distal interphalangeal joints is used to record antidromic sensory potentials. Single stimuli give rise to sensory potentials of sufficient amplitude for studying commonly tested nerves, thereby eliminating the need for averaging techniques. Some electromyographers prefer to use needle electrodes to increase the amplitude of the recorded sensory nerve potential and to decrease the noise from surrounding tissue. The combined effect of increased amplitude and decreased noise may enhance the signal-to-noise ratio by as much as five times,312 and, when averaging, markedly reduce the number of trials required to obtain the same resolution.

Averaging Technique Very small signals fall within the expected noise level. Interposing a step-up transformer between the recording electrodes and the amplifier improves the signal-to-noise ratio. Averaging techniques also help to identify small signals that may otherwise escape detection. For this purpose, the electronic devices now in use sum consecutive samples of waveforms in each sweep digitally and divide this by the number of trials. The voltage from noise that has no temporal relationship to stimulation will average close to zero with this technique, whereas signals timelocked to a stimulus will appear at a constant latency, distinct from the background noise. In one study testing the limit of an averaging technique, a sensory nerve action potential could be recorded against the background noise up to 50 times, but not 100 times, the signal.273

Motor Nerve Conduction Studies Waveform, Amplitude, and Duration A single shock applied to the nerve activates a group of motor units that summate slightly asynchronously because individual nerve axons vary in conduction velocity and in terminal length. Thus a compound muscle action potential consists of the temporally dispersed sum of many motor unit action potentials located within the recording radius of the electrode, a distance of approximately 20 mm.15 Motor units closer to the surface of the skin contribute more to the compound action potential than distant units, although their spatial orientation to the recording electrode also plays an important role.29,252,373 In belly-tendon recording (Fig. 35–1) with an active lead (E1) placed on the belly of the muscle and an indifferent lead (E2) on the tendon, muscle action potentials show an initially negative biphasic waveform. The usual measurements taken include (1) latency, the time from the stimulus to the onset of the negative response; (2) amplitude, the distance from the baseline to the negative peak or between negative and positive peaks; and (3) duration, the time from onset to the negative or positive peak or to the final return to the baseline. The area under the waveform measured by electronic integration correlates linearly with the product of the amplitude and duration. In general, onset latency provides a measure of the fast-conducting motor fibers. The shortest axons, however, may give rise to the initial potential, although they are not necessarily the fastest fibers. Latency and Conduction Velocity The latency of an action potential consists of three components: (1) nerve activation, the time from a stimulus to the generation of an action potential; (2) nerve conduction, the time from the activation point to the nerve terminal; and (3) neuromuscular transmission, the time from the axon terminal to the motor end plate, including the time required for generation of muscle action potential. The

Nerve Conduction and Needle Electromyography

901

FIGURE 35–1 A, Motor and sensory conduction studies of the median nerve. The photo shows stimulation at the wrist, 3 cm proximal to the distal crease, and recording over the belly (E1) and tendon (E2) of the abductor pollicis brevis for motor conduction, and around the proximal (E1) and distal (E2) interphalangeal joints of the second digit for antidromic sensory conduction. The ground electrode is located in the palm. B, Alternative recording sites for sensory conduction study of the median nerve with the ring electrodes placed around the proximal (E1) and distal (E2) interphalangeal joints of the third digit, or the base (E1) and the interphalangeal joint (E2) of the first digit. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

latency difference between two stimulation points, in effect, excludes components common to both stimuli, yielding the conduction time from one point of stimulation to the next (Fig. 35–2). Dividing the distance between these two stimuli by the latency difference, the motor nerve conduction velocity (MNCV) can be calculated by MNCV(m/s) 

D (mm) Lp  Ld(ms)

where D is the distance between the two stimulus points in millimeters, and Lp and Ld are the proximal and distal latencies in milliseconds, respectively. The reliability of

nerve conduction studies depends on how accurately the latencies are measured. Additionally, the length of nerve segment used in the calculation is only an estimate based on surface distance. In segmental conduction studies, separation of successive points of stimulation by at least 10 cm improves the accuracy of surface measurement. However, for localized lesions such as a compressive neuropathy, the inclusion of longer unaffected segments dilutes the effect of focal slowing and lowers the sensitivity. In contrast, the use of incremental stimulation across multiple shorter segments in the setting of local compression may reveal abrupt changes in latency and waveform. This may help localize an abnormality that might otherwise be difficult to identify with conventional stimulation lengths.

902

Nerve Conduction and Electromyography

FIGURE 35–2 Compound muscle action potential recorded from the thenar eminence following the stimulation of the median nerve at the elbow. The nerve conduction time from the elbow to the wrist equals the latency difference between the two responses elicited by the distal and proximal sites of stimulation. The motor nerve conduction velocity (MNCV), calculated by dividing the surface distance between the stimulus points by the subtracted times, concerns the fastest fibers. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

Conduction velocity cannot be calculated for the most distal segment because the terminal latency includes the time needed for neuromuscular transmission. The use of a fixed distance or anatomic landmarks for electrode placement standardizes the recording technique for improved accuracy and reproducibility. Because of distal slowing of nerve conduction, the terminal latency (Ld) estimated from the terminal distance (D) and proximal conduction velocity (CV) as the ratio D/CV is slightly less than the measured latency (Ld). The difference between measured and estimated latencies (Ld  Ld) is referred to as the residual latency. It represents a measure of the conduction delay at the nerve terminal and at the neuromuscular junction.215 The ratio between the estimated and measured latency (Ld/Ld), the terminal latency index, also relates to distal conduction delay.329 In a study of the median nerve, for example, a patient with a measured distal latency (Ld) of 4.0 ms, a terminal distance (D) of 8 cm, and a forearm conduction velocity (CV) of 50 m/s would have a calculated terminal latency (Ld) of 1.6 ms (8 cm/50 m/s), a residual latency of 2.4 ms (4.0 ms  1.6 ms) and a terminal latency index of 0.4 (1.6 ms/4.0 ms). Types of Abnormalities Nerve conduction studies help distinguish between two types of abnormalities: axonal degeneration and demyelination. This distinction is sometimes less clear for the

assessment of a nerve as a whole, as opposed to individual nerve fibers. Depending on the type of pathology, stimulation proximal to the site of presumed lesion gives rise to three basic types of conduction abnormalities (Fig. 35–3): (1) reduced amplitude with a relatively normal latency, (2) increased latency with relatively normal amplitude, and (3) absent response. Reduced amplitude with normal latency indicates an incomplete nerve lesion, either neurapraxia or axonotmesis involving a portion of the axons. In both situations, a shock applied distal to the lesion within the first few days of injury elicits a normal response showing a larger amplitude than stimulation proximal to the lesion. Within several days of the injury, however, distal stimulation may distinguish the two (Fig. 35–4). Regarding axonotmesis, stimulation below the lesion several days after its occurrence elicits an equally small response as stimulation above the lesion; this is secondary to wallerian degeneration. Regarding neurapraxia, often in the setting of demyelinative lesions, stimulation distal to the lesion continues to give rise to a larger compound muscle action potential than does stimulation proximal to the lesion, indicating a conduction block (Fig. 35–5). Increased latency combined with relatively normal amplitude generally implies that a majority of the nerve fibers have segmental demyelination without conduction block. As shown in rabbits, incomplete proximal compressive lesions may also give rise to slowed conduction with a reduction in external fiber diameter distal to the site of constriction. In this situation, the time course of recovery

FIGURE 35–3 Three basic types of alteration in the compound muscle action potential occur after a presumed nerve injury distal to the site of stimulation: reduced amplitude with nearly normal latency (top), normal amplitude with substantially increased latency (middle), or absent response even with a shock of supramaximal intensity (bottom). (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

Nerve Conduction and Needle Electromyography

903

a loss of large anterior horn cells in myelopathies may also cause slowed motor conduction. Here, the motor conduction velocity may decrease to 70% of the mean normal value with diminution of amplitude to less than 10% of normal. Despite a loss of the amplitude, however, a conduction velocity reduced to less than 60% of the mean normal value suggests peripheral nerve disease, rather than myelopathy.220

Sensory Nerve Conduction Studies

FIGURE 35–4 Nerve excitability distal to the lesion in neurotmesis or substantial axonotmesis. Distal stimulation elicits a normal compound muscle action potential during the first few days after injury, even with a complete separation of the nerve. Unlike neurapraxia, wallerian degeneration subsequent to transection will render the distal nerve segment inexcitable in 3 or 4 days. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

suggests distal paranodal demyelination as the cause of conduction slowing along the distal nerve segment.13 Absent responses indicate a complete failure of nerve conduction across the presumed site of the lesion. This may occur after either nerve transection with total axonotmesis, or neurapraxia involving all the motor fibers. Regardless of etiology, nerve stimulation distal to the lesion during the first week after injury elicits an entirely normal muscle action potential. By the second week, however, distal stimulation can distinguish neurapraxic changes from axonal degeneration as described above: normal response in neurapraxia and absent response in neurotmesis (see Fig. 35–4). In cases of nerve fiber degeneration, serial electrophysiologic studies help delineate progressive recovery of the amplitude of evoked potentials concomitant with nerve regeneration (Fig. 35–6). Additionally, a prolonged latency or slowing of conduction may also result from axonal neuropathy with a loss of the fast-conducting fibers.103 A major reduction in amplitude to below 40% to 50% of the mean normal value usually accompanies this type of slowing. If the amplitude remains above 80% of the control value in the absence of demyelination, the conduction velocity should not fall below 80% of the lower limit of normal. With further diminution of the amplitude to less than half the mean normal value, the conduction velocity may fall to 70% of the lower limit solely on the basis of axonal loss. Similarly,

Waveform, Amplitude, and Duration During sensory conduction studies in the upper limbs, stimulation of the digital nerves elicits an orthodromic sensory potential, which can be recorded proximally from the nerve trunk. Alternatively, stimulation of the nerve trunk evokes an antidromic potential distally in the digit (see Fig. 35–1) and a mixed nerve potential proximally. For example, a shock applied to the ulnar or median nerve at the wrist gives rise to an antidromic potential in the corresponding digits and a mixed nerve potential along the nerve trunk at the elbow. For routine clinical recordings, noninvasive recording with surface electrodes provide reliable and reproducible results.5,104 The use of needle recording, however, may improve the signal-to-noise ratio, especially when assessing temporal dispersion.217,277 Here again, signal averaging may help elucidate small late components originating from demyelinated, remyelinated, or regenerated fibers.40,131 These potentials correspond to fibers approximately 4 m in diameter with an average conduction velocity of 15 m/s.332 A reduction in the minimum conduction velocity may constitute the sole abnormality in several neuropathies that otherwise remain diagnostically elusive.156 Recording from surface electrodes generally result in higher amplitudes for an antidromic potential from the digits than an orthodromic response from the nerve trunk. This is in part because the digital nerves lie nearer to the surface.43 A change in the position of the recording electrodes may alter the waveform of an orthodromic sensory nerve action potential.304,385 Placing an active electrode (E1) on the nerve and a reference electrode (E2) at a remote site generally produces an initially positive triphasic potential. A separate late phase may also appear in a temporally dispersed response recorded at a more proximal site. In addition, the placement of E2 near the nerve at a distance greater than 3 cm from E1 results in a tetraphasic potential with the addition of the final negative component.43 The amplitude of an initially negative diphasic antidromic sensory potential is measured either from the baseline to its negative peak or between the negative and positive peaks. The amplitude of the initially positive triphasic orthodromic sensory potential may also be measured either from the baseline to the negative peak or

904

Nerve Conduction and Electromyography

FIGURE 35–5 A 67-year-old man with an acute onset of footdrop following chemotherapy and radiation treatment of prostate cancer. Although epidural metastasis was suspected clinically because of backache, the nerve conduction studies revealed the evidence of a conduction block at the knee, indicating a compressive neuropathy. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

between the positive and negative peaks. Amplitudes vary substantially among subjects and to a lesser extent between the two sides of the same individual, whether recorded with surface or needle electrodes.43 Factors that influence the amplitude of a nerve action potential include the density of sensory innervation and the subject’s body mass index, the latter reflecting the depth of the nerve from the skin surface.44 Women tend to have greater sensory nerve action potentials than men,157,230 possibly because the nerves lie more superficially. Left-handed individuals often have greater median sensory potentials at the wrist on their right side, whereas right-handed individuals often have greater amplitudes on their left side.248 The duration of the orthodromic or antidromic sensory potential is usually measured from the initial deflection from the baseline to the intersection between the descending phase and the baseline. Alternative points of reference occasionally used include the initial positive or negative peak, the subsequent positive peak, and, though less definable, the point of return to the baseline. Latency and Conduction Velocity Stimulation of the nerve at a single site generally suffices for calculation of sensory conduction velocity because, unlike a motor latency, a sensory latency consists solely of

the conduction time to the recording point. The latency of activation, a fixed delay of about 0.15 ms at the stimulus site,216 produces a measured latency slightly larger and the calculated conduction velocity slightly slower than actuality. The use of a second stimulus flanking the nerve segment corrects for this error because the latency difference between the two responses excludes the nerve activation time common to both stimulations. The latency of an antidromic potential is measured to the onset of the negative peak, whereas the latency of an orthodromic sensory potential is measured to the peak of the initial positive potential. The measurement to the negative peak may also be used if technical problems make identification of the preceding smaller positive peak difficult, as may occur in diseased or damaged nerves. Types of Abnormalities In general, similar types of abnormalities are found in both motor and sensory nerves. Demyelination of sensory fibers typically causes substantial slowing of conduction. Axonal degeneration typically results in a reduction in amplitude of the compound nerve action potentials elicited by distal stimulation. The sural nerve potential serves as a sensitive measure for identifying a lengthdependent distal axonal polyneuropathy.4 In patients with

Nerve Conduction and Needle Electromyography

905

FIGURE 35–6 A, A 21⁄2-year-old boy with hypothermia-induced axonal polyneuropathy following prolonged exposure to severe freezing weather on a frigid winter night in Iowa. Compound muscle action potentials recorded over the thenar eminence after stimulation of the median nerve at the wrist (W), elbow (E), or axilla (A). The initial study on March 17, 1986, revealed no response on either side followed by progressive return in amplitude and latency, with full recovery by January 8, 1987. B, Compound muscle action potential recorded from the abductor hallucis after stimulation of the tibial nerve at the ankle (A) or knee (K). The studies on March 17 and May 7, 1986, revealed no response on either side, with full recovery by January 8, 1987. Figure continued on following page

neuropathy, the ratio of the sensory potential of the sural nerve to that of the radial nerve often falls below 0.40, with a normal ratio of 0.71.320 Sensory fibers generally degenerate only after a lesion involving the sensory ganglion or postganglionic axon (see Fig. 35–6). Thus nerve conduction studies of the distal sensory potential can often differentiate preganglionic root avulsion from plexopathy.134 Radicular lesions within the spinal canal, however, may also involve the ganglion or postganglionic portion of the root, affecting the digital nerve potential.176,234 Localization of the lesion therefore depends on the distribution of the sensory potential abnormalities. In addition, plexopathy tends to affect multiple digits, whereas radiculopathy

usually reveals a selective abnormality of specific digits related to the root involved; the first digit by C6, the second and third digits by C7, and the fourth and fifth digits by C8.105

Nerve Conduction in the Clinical Domain Physiologic Variation among Different Nerve Segments Both motor and sensory fibers conduct slower in the legs than in the arms, with differences ranging from 7 to 10 m/s.205,358 Available data also confirm a length-dependent change in the conduction velocity for peroneal and sural nerves, but

906

Nerve Conduction and Electromyography

FIGURE 35–6 Continued C, Motor conduction studies of the tibial nerve on May 17, 1986. Stimulation at the knee elicited no response in the intrinsic foot muscle on either side (top three tracings), but a small compound action potential was elicited in the gastrocnemius bilaterally (bottom tracing) as the result of early reinnervation. D, Antidromic sensory nerve action potential recorded from the second digit after stimulation of the median nerve at the wrist (W) or elbow (E). The studies on March 17 and May 7, 1986, showed no response on either side, with full recovery by January 8, 1987. (From Afifi, A. K., Kimura, J., and Bell, W. E.: Hypothermiainduced reversible polyneuropathy: electrophysiologic evidence of axonopathy. Pediatr. Neurol. 4:49, 1988, with permission.)

not for motor or sensory fibers of the median nerve.338 Factors that may account for this velocity gradient in the lower limbs include, in the absence of histologic proof, abrupt axonal tapering, progressive reduction in axonal diameter, a shorter internodal distance, and lower temperatures. Statistical analyses of conduction velocities show no difference between median and ulnar nerves or between tibial and peroneal nerves, with a high degree of symmetry between the two sides.28 The nerve impulse propagates faster in the proximal than in the distal nerve segments,135 as evidenced by the F-wave conduction velocity between the spinal cord and axilla, which clearly exceeds the conventionally derived conduction velocity between the elbow and wrist.91,181,196 Statistical analyses show no significant difference between the two proximal segments, spinal cord to axilla and axilla to elbow, probably indicating a nonlinear slowing over the most distal segment.181 Effects of Temperature The effects of lower temperatures include a slowdown of impulse propagation and, at the same time, augmentation in amplitude of nerve and muscle action potentials, as

demonstrated in the squid axon152 and in human studies.73,300 For example, distal latencies increase by 0.3 ms/degree centigrade for both median and ulnar nerves upon cooling the hand.49 These principles apply to normal as well as demyelinated fibers as a straightforward consequence of the temperature coefficients governing voltage-sensitive sodium (Na) and potassium (K) conductance. In particular, cold-induced slowing of Na channel opening accounts for the slowing of conduction, and a delay of its inactivation, for a given increase in amplitude. In demyelinated nerve fibers, conduction fails at the lesion site when the rising temperature reduces the already marginal action potential below the critical level by fast inactivation of Na channels.116,374 A temperature rise, however, also quickens activation of Na channels, leading to a faster conduction over a length of a fiber that is normal.321 Thus the latency and amplitude measures follow two completely separate effects induced by change in temperature. Studies conducted in a warm room with ambient temperature maintained between 21° and 23° C reduce this type of variability. In practice, the skin temperature is measured with a plate thermistor to maintain it above 32° C, if necessary, using an infrared heat lamp or prior

Nerve Conduction and Needle Electromyography

immersion of the limb in warm water for 30 minutes.115 Alternatively, addition of 5% of the calculated conduction velocity for each degree below 32° C theoretically normalizes the result. Such conversion factors established in a normal population, however, may not necessarily apply in patients with diseases of the peripheral nerve.11,79,275 Maturation and Aging The process of maturational myelination accompanies a rapid increase in nerve conduction velocities from roughly half the adult value in full-term infants357 to the adult range at age 3 to 5 years (Fig. 35–7). Conduction velocities of physiologically slower fibers also show a similar time course of maturation.145 One series (Table 35–1) showed a steep increase in conduction of the peroneal nerve in infancy as compared to a slower maturation of the median nerve during early childhood.125 Conduction velocities in premature infants had an even slower range, from 17 to 25 m/s for the ulnar nerve and from 14 to 28 m/s for the peroneal nerve.51 The values reported at 23 to 24 weeks of fetal life averaged roughly one-third those of newborns of normal gestational age.255,335 Studies based on the expected date of birth of premature infant showed a different time course of maturation for motor and proprioceptive conduction.23 Fetal nutrition may alter peripheral nerve function by influencing myelin formation.309,310 In later childhood and adolescence, from age 3 to 19 years, both motor and sensory conduction velocities change as a function of age and growth in length, showing a slight increase in the upper limb and slight decrease in the lower limb.223 Conduction velocities begin to decline after 30 to 40 years of age, but not more than 10 m/s by

FIGURE 35–7 Relation of age to conduction velocity of motor fibers in the ulnar nerve between the elbow and wrist. Velocities in normal young adults range from 47 to 73 m/s, with most values between 50 and 70 m/s. Ages plotted indicate month after expected birth date based on calculation from the first day of last menstruation. (From Thomas, J. E., and Lambert, E. H.: Ulnar nerve conduction velocity and H-reflex in infants and children. J. Appl. Physiol. 15:1, 1960, with permission.)

907

60 years353 or even by 80 years of age.274 The most distal branches, such as the interdigital nerves, may show agerelated degeneration earlier.229 In one study,250 a reduction in the mean conduction rate averaged 10% at 60 years of age (Table 35–2). Aging also causes a diminution in amplitude and changes in the shape of the evoked potential,101 especially when recorded across the common sites of compression.64,65 Preferential loss of the largest and fastest conducting motor units375 probably results in a gradual increase in latencies of the F wave and somatosensory evoked potentials with advancing age. Height and Other Factors Various anthropometric characteristics also influence nerve conduction measures.322,348 Sural, peroneal, and tibial nerve conduction velocities all have inverse correlation with height in normal subjects308 and in patients with diabetic neuropathy,124 although in one study366 the height-related changes of sural nerve conduction velocity fell within the experimental error of 2.3% expected from the method. Women have faster conduction velocity and greater amplitude for both motor and sensory studies than men.309 Most gender-related velocity differences resolve when adjusted by height, whereas amplitude differences persist despite such correction and may, at least in part, relate to volume conductor characteristics of the body mass. Clinical Values and Limitations Besides a broad classification, the pattern of nerve conduction abnormalities can often characterize the general nature of some specific clinical disorders. In hereditary demyelinating neuropathies, for example, typical features

908

Nerve Conduction and Electromyography

Table 35–1. Normal Motor Nerve Conduction Velocities (m/s) in Different Age Groups Age 0–1 wk 1 wk to 4 mo 4 mo to 1 yr 1–3 yr 3–8 yr 8–16 yr Adults

Ulnar

Median

Peroneal

32 (21–39) 42 (27–53) 49 (40–63) 59 (47–73) 66 (51–76) 68 (58–78) 63 (52–75)

29 (21–38) 34 (22–42) 40 (26–58) 50 (41–62) 58 (47–72) 64 (54–72) 63 (51–75)

29 (19–31) 36 (23–53) 48 (31–61) 54 (44–74) 57 (46–70) 57 (45–74) 56 (47–63)

From Gamstorp, I.: Normal conduction velocity of ulnar, median and peroneal nerves in infancy, childhood and adolescence. Acta Paediatr. Suppl. 146:68, 1963, with permission.

include diffuse abnormalities with little difference from one nerve to another in the same patient and among different members in the same family.236 The studies also show approximately equal involvement of different nerve fibers with a limited degree of temporal dispersion despite a considerably increased latency. This stands in contrast to acquired demyelination, wherein disproportionate involvement of certain segments of the nerve185,206 often gives rise to more asymmetrical abnormalities with temporal dispersion. The reverse, however, does not always hold, because acquired demyelination can also produce diffuse abnormalities reminiscent of hereditary induced patterns.

Demyelinating and axonal polyneuropathies show a characteristic pattern of distribution in sensory nerve conduction abnormalities. Thus a reduced median amplitude compared with the sural amplitude supports the diagnosis of a primary demyelination,27 whereas a reduced sural amplitude compared with the radial amplitude tends to suggest an axonal polyneuropathy.320

COMMONLY ASSESSED NERVES Cranial Nerves Facial Nerve A quantitative assessment of nerve excitability depends on the size of muscle action recorded after stimulation of the facial nerve just below the ear and anterior to the mastoid process378 or directly over the stylomastoid foramen.352 Table 35–3 summarizes the normal onset latencies in 78 subjects divided into different age groups.377 Other normal values in adults (mean  standard deviation [SD]) range from 3.4  0.8 ms to 4.0  0.5 ms.352 In assessing a proximal lesion, as in Bell’s palsy, the latency of the muscle action potential rarely reveals a clear-cut abnormality even with substantial axonal degeneration because the remaining axons conduct normally. In contrast, the amplitude of the direct response serves importantly in determining the degree of axonal loss, which in turn dictates the prognosis.

Table 35–2. Normal Sensory and Motor Nerve Conduction Velocities (m/s) in Different Age Groups* Age 10–35 Yr (30 Cases) Nerve Median nerve Digit–wrist Wrist–muscle Wrist–elbow Elbow–axilla Ulnar nerve Digit–wrist Wrist–muscle Wrist–elbow Elbow–axilla Common peroneal nerve Ankle–muscle Ankle–knee Posterior tibial nerve Ankle–muscle Ankle–knee H reflex, popliteal fossa

Sensory

Motor

67.5  4.7 67.7  4.4 70.4  4.8

Sensory

Motor

65.8  5.7 3.2  0.3† 59.3  3.5 65.9  5.0

64.7  3.9

65.8  3.1 70.4  3.4

64.8  3.8 69.1  4.3

53.0  5.9

4.3  0.9† 49.5  5.6 5.9  1.3† 45.5  3.8 71.0  4.0 27.9  2.2†

Sensory

Motor

62.8  5.4 66.2  3.6

3.5  0.2† 54.5  4.0 63.6  4.4

57.5  6.6

67.1  4.7 70.6  2.4

2.7  0.3† 57.8  2.1 63.3  2.0

50.4  1.0

4.8  0.5† 43.6  5.1

49.0  3.8

Age 51–80 Yr (18 Cases)

59.4  4.9 3.7  0.3† 55.9  2.6 65.1  4.2

66.5  3.4 2.7  0.3† 58.9  2.2 64.4  2.6

56.9  4.4

Age 36–50 Yr (16 Cases)

7.3  1.7† 42.9  4.9 64.0  2.1 28.2  1.5†

*Values are means  1 standard deviation. † Latency in milliseconds. From Mayer, R. F.: Nerve conduction studies in man. Neurology (Minneap.) 13:1021, 1963, with permission.

56.7  3.7 64.4  3.0

3.0  0.35† 53.3  3.2 59.9  0.7

46.1  4.0

4.6  0.6† 43.9  4.3

48.9  2.6

6.0  1.2† 41.8  5.1 60.4  5.0 32.0  2.1†

Nerve Conduction and Needle Electromyography

Table 35–3. Facial Nerve Latency in 78 Subjects Divided into Different Age Groups Age 0–1 mo 1–12 mo 1–2 yr 2–3 yr 3–4 yr 4–5 yr 5–7 yr 7–16 yr

Mean (ms)

909

series, normal latencies to the upper trapezius in 25 subjects, 10 to 60 years of age, varied from 1.8 to 3.0 ms.54

Range (ms)

10.1 7.0 5.1 3.9 3.7 4.1 3.9 4.0

6.4–12.0 5.0–10.0 3.5–6.3 3.8–4.5 3.4–4.0 3.5–5.0 3.2–5.0 3.0–5.0

From Waylonis, G. W., and Johnson, E. W.: Facial nerve conduction delay. Arch. Phys. Med. Rehabil. 45:539, 1964, with permission.

Accessory Nerve Surface or needle stimulation along the posterior border of the sternocleidomastoid muscle activates the accessory nerve, eliciting a compound muscle action potential of the trapezius. For recording, an active electrode (E1) is placed at the angle of the neck and shoulder, and a reference electrode (E2) over the tendon near the acromion process.297,298 In one

FIGURE 35–8 A, Motor nerve conduction study of the median nerve. The sites of stimulation include Erb’s point (A), axilla (B), elbow (C), wrist (D), and palm (E). Compound muscle action potentials are recorded with surface electrodes placed on the thenar eminence. B, Sensory nerve conduction study of the median nerve. The sites of stimulation include axilla (A), elbow (B), wrist (C), and palm (D). Antidromic sensory potentials are recorded with a pair of ring electrodes placed around the second digit. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

Major Nerves in the Upper Limb Median Nerve The usual sites of stimulation of the median nerve include Erb’s point and the axilla, elbow, wrist, and palm along the relatively superficial course of the nerve (Fig. 35–8). Its stimulation at Erb’s point, the axilla, and the palm tends to coactivate other nerves in close proximity.186,288,314 Table 35–4 summarizes normal values in our laboratory. Separate stimulation of the median and ulnar nerves at the wrist evokes corresponding muscle action potentials from the second lumbricalis and the first volar interosseous at nearly identical latency if the distance between the stimulating and recording electrodes is kept the same for both nerves.74,112,303,322,330,365 A latency difference greater than 0.4 to 0.5 ms suggests an abnormal delay over the distal segment as might be seen in carpal tunnel syndrome. Table 35–4 summarizes normal values for the digital potentials recorded with ring electrodes placed 2 cm apart

910

Nerve Conduction and Electromyography

Table 35–4. Normal Values for Median Nerve* Site of Stimulation

Amplitude†: Motor (mV), Sensory (V)

Motor Fibers Palm Wrist Elbow Axilla Sensory Fibers Digit Palm Wrist Elbow

Latency‡ to Recording Site (ms)

Difference between Right and Left (ms)

6.9  3.2 (3.5)§ 7.0  3.0 (3.5) 7.0  2.7 (3.5) 7.2  2.9 (3.5)

1.86  0.28 (2.4)储 3.49  0.34 (4.2) 7.39  0.69 (8.8) 9.81  0.89 (11.6)

0.19  0.17 (0.5)储 0.24  0.22 (0.7) 0.31  0.24 (0.8) 0.42  0.33 (1.1)

39.0  16.8 (20) 38.5  15.6 (19) 32.0  15.5 (16)

1.37  0.24 (1.9) 2.84  0.34 (3.5) 6.46  0.71 (7.9)

0.15  0.11 (0.4) 0.18  0.14 (0.5) 0.29  0.21 (0.7)

Conduction Time between Two Points (ms)

Conduction Velocity (ms)

1.65  0.25 (2.2)储 3.92  0.49 (4.9) 2.42  0.39 (3.2)

48.8  5.3 (38)¶ 57.7  4.9 (48) 63.5  6.2 (51)

1.37  0.24 (1.9) 1.48  0.18 (1.8) 3.61  0.48 (4.6)

58.8  5.8 (47) 56.2  5.8 (44) 61.9  4.2 (53)

*Mean  standard deviation (SD) in 61 patients, 11 to 74 years of age (average, 40), with no apparent disease of the peripheral nerves. † Amplitude measured from the baseline to the negative peak. ‡ Latency measured to the onset. § Lower limits of normal, based on the distribution of the normative data. 储 Upper limits of normal, calculated as the mean  2 SD. ¶ Lower limits of normal, calculated as the mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

around the proximal (E1) and distal (E2) interphalangeal joints of the second digit. Serial stimulation from the midpalm to the distal forearm in 1-cm increments normally shows a predictable latency change of 0.16 to 0.20 ms/cm (Fig. 35–9). Focal abnormalities of the median nerve usually accompany a sharply localized latency increase across a 1-cm segment (Fig. 35–10) associated with an abrupt change in waveform from abnormal temporal dispersion. The use of multichannel electrodes allows simultaneous recording of nerve potentials at several sites across the wrist after stimulation of the median nerve at the digit189,327 or at the elbow.161 This technique offers instantaneous comparison of latencies but suffers from a major drawback in the assessment of amplitudes, which vary so much depending on the depth of the nerve at the site of recording.189 Separate stimulation of the median and ulnar nerves at the wrist evokes a corresponding sensory potential of the fourth digit at nearly the identical latency for the same conduction distance. In our series (Table 35–5), normal values for the median nerve consisted of an onset latency of 2.88  0.35 ms (mean  SD) after wrist stimulation and a distal amplitude of 37.6  17.2 V after palmar stimulation. The corresponding values for the ulnar nerve were 2.86  0.37 ms and 46.1  24.3 V. The latency difference between the two nerves ranged from 0.01  0.17 ms to 0.02  0.17 ms, with an upper limit of normal of 0.4 ms (mean  2 SD). Ulnar Nerve Routine motor conduction studies consist of stimulating the ulnar nerve at multiple sites along the superficial course of the nerve (Fig. 35–11) and recording the muscle

potential from the hypothenar muscle with a pair of surface electrodes placed over the belly of the abductor digiti minimi (E1) and its tendon (E2) 3 cm distally (Tables 35–6 and 35–7).7,8 For accurate calculation of conduction velocities, the distance between sites of stimulation above and below the elbow should exceed 10 cm to minimize measurement error. Conventional studies, however, may fail to uncover focal abnormalities in tardy ulnar palsy or a cubital tunnel syndrome, which induces an insignificant delay when assessed over a 10-cm segment. In contrast, an “inching” study across the elbow with stimulation in 1- to 2-cm increments often detects an abrupt change in latency and waveform at the site of localized compression.47,194 The conduction along the deep palmar branch of the ulnar nerve can be tested by stimulating the ulnar nerve at the wrist and recording the muscle potential from the first dorsal interosseous or adductor pollicis. In one series of 373 patients, the upper limit of the normal range included 4.5 ms for the distal latency to the first dorsal interosseous, 2.0 ms for the latency difference between this muscle and the abductor digiti minimi, and 1.3 ms for the latency difference between the two sides.280 Stimulation of the deep palmar branch distal to the site of lesion also evokes muscle responses as a good measure of surviving motor axons. Lumbrical-interosseous comparison as described for median nerve study is also useful in assessing a distal ulnar nerve lesion, which typically causes a latency difference greater than 0.4 ms in the reverse direction.212,331 The common sites of stimulation for an antidromic sensory study include above and below the elbow, 3 cm proximal to the distal crease at the wrist, and 5 cm distal

Nerve Conduction and Needle Electromyography

911

FIGURE 35–9 A, Twelve sites of stimulation in 1-cm increments along the length of the median nerve. The “0” level at the distal crease of the wrist corresponds to the origin of the transverse carpal ligament. The photo shows a recording arrangement for sensory nerve potentials from the second digit and muscle action potentials from the abductor pollicis brevis. B, Sensory nerve potentials in a normal subject recorded after stimulation of the median nerve at multiple points across the wrist. The numbers on the left indicate the site of each stimulus (compare to A). The latency increased linearly with stepwise shifts of stimulus site proximally in 1-cm increments. (From Kimura, J.: The carpal tunnel syndrome: localization of conduction abnormalities within the distal segment of the median nerve. Brain 102:619, 1979, with permission.)

to the crease in the palm (see Fig. 35–11) to make the studies comparable to those of the median nerve (see Fig. 35–8). Recording from either the fourth or fifth digit tests the integrity of the C8 and T1 roots, lower trunk, and medial cord. Alternative studies consists of stimulation of the digital nerve with ring electrodes placed around the interphalangeal joints of the fifth digit, cathode proximally, and recording orthodromic sensory potentials at various sites along the course of the nerve. The dorsal ulnar cutaneous nerve, like the ulnar nerve proper, derives from the C8 to T1 roots, the lower trunk, and the medial cord. It leaves the common trunk of the ulnar nerve 5 to 8 cm proximal to the ulnar styloid between the tendon of the flexor carpi ulnaris and the ulna.22,168 Surface stimulation at this point selectively stimulates the dorsal ulnar cutaneous nerve, which subserves the sensation over the dorsum of the hand. Its abnormality helps localize the lesion in the segment proximal to the takeoff of this branch, although anatomic variations may alter cutaneous innervation.296

Radial Nerve The radial nerve becomes accessible to surface stimulation at (1) Erb’s point in the supraclavicular fossa; (2) the axilla near the spinal groove between the coracobrachialis and medial edge of the triceps, about 18 cm proximal to the medial epicondyle; (3) above the elbow between the brachioradialis and the tendon of the biceps, 6 cm proximal to the lateral epicondyle; and (4) in the forearm between the extensor carpi ulnaris and extensor digiti minimi on the dorsal aspect of the ulna, 8 to 10 cm proximal to the styloid process (Fig. 35–12). Either needle or surface electrodes may be used to record muscle action potentials from the extensor digitorum communis387 or the extensor indicis. The sensory branch gives off the posterior antebrachial cutaneous nerve, which innervates the dorsolateral aspect of the forearm at the level of the elbow before emerging near the surface about 10 cm above the lateral styloid process.52 Surface stimulation at any site along the lateral edge of the radius up to 10 to 14 cm proximal to the base of the thumb activates the sensory fibers selectively.

912

Nerve Conduction and Electromyography

FIGURE 35–10 A, Sensory nerve potentials in a patient with the carpal tunnel syndrome. Both hands showed a sharply localized slowing from 2 to 1, with the calculated segmental conduction velocity of 14 m/s on the left (top) and 9 m/s on the right (bottom). Note a distinct change in waveform of the sensory potential at the point of localized conduction delay. The double-humped appearance at 2 on the left suggests sparing of some sensory axons at this level. B, Sensory nerve potential in a patient with the carpal tunnel syndrome. Temporally dispersed responses on the right at 1 and beyond had greater negative and positive peaks in this area compared to normal, more distal responses, presumably because of loss of physiologic phase cancellation. Both hands show a sharply localized slowing from 3 to 2, with a segmental conduction velocity of 10 m/s on the left (top) and 7 m/s on the right (bottom). An abrupt change in waveform of the sensory potential also indicates the point of localized conduction delay. C, Sensory nerve potential in a patient with the carpal tunnel syndrome before (top) and after surgery (bottom). Preoperative study showed a localized slowing from 4 to 3, with a calculated segmental conduction velocity of 8 m/s, which normalized in a repeat study 6 months postoperatively. (From Kimura, J.: The carpal tunnel syndrome: localization of conduction abnormalities within the distal segment of the median nerve. Brain 102:619, 1979, with permission.)

913

Nerve Conduction and Needle Electromyography

Table 35–5. Distal Sensory Conduction Study Comparing Median and Ulnar Nerves* Measurement of Antidromic Sensory Potential Recording

Stimulation

Amplitude† (V)

Latency‡ (ms)

Median nerve, 2nd digit Median nerve, 4th digit Ulnar nerve, 4th digit Median & ulnar difference

Palm Wrist Palm Wrist Palm Wrist Palm Wrist

49.8  21.5 (25)§ 38.4  15.6 (19) 37.6  17.2 (19) 22.3  8.2 (11) 46.1  24.3 (23) 29.0  14.8 (25) 8.5  20.7 5.9  10.1

1.43  0.16 (1.7)储 2.87  0.31 (3.5) 1.45  0.20 (1.9) 2.88  0.35 (3.6) 1.48  0.26 (2.0) 2.86  0.37 (3.6) 0.02  0.17 (0.3) 0.01  0.17 (0.4)

Calculated Values for Wrist to Palm Segment Conduction Time (ms)

Conduction Velocity (m/s)

1.44  0.20 (1.9)储

57.1  8.3 (40)¶

1.43  0.22 (1.9)

57.4  8.9 (40)

1.38  0.30 (1.8)

59.1  8.3 (43)

0.04  0.20 (0.4)

*Mean  standard deviation (SD) in 31 healthy subjects, 16 to 64 years of age (average, 38), with no apparent disease of the peripheral nerve. † Amplitude measured from the baseline in the negative peak. ‡ Latency measured to the onset with a standard distance of 8 cm between the stimulus sites at the wrist and palm. § Lower limits of normal, based on the distribution of the normative data. 储 Upper limits of normal, calculated as the mean  2 SD. ¶ Lower limits of normal, calculated as the mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

FIGURE 35–11 A, Motor nerve conduction study of the ulnar nerve. The sites of stimulation include Erb’s point (A), axilla (B), above the elbow (C), elbow (D), below the elbow (E), and wrist (F). Compound muscle action potentials are recorded with surface electrodes placed on the hypothenar eminence. B, Sensory nerve conduction study of the ulnar nerve. The sites of stimulation include axilla (A), above the elbow (B), elbow (C), below the elbow (D), wrist (E), and palm (F). The tracings show antidromic sensory potentials recorded with the ring electrodes placed around the fifth digit. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

914

Nerve Conduction and Electromyography

Table 35–6. Normal Values for Ulnar Nerve* Site of Stimulation

Amplitude†: Motor (mV), Sensory (V)

Latency‡ to Recording Site (ms)

Difference between Right and Left (ms)

Conduction Time between Two Points (ms)

Conduction Velocity (m/s)

Motor Fibers Wrist Below elbow Above elbow Axilla

5.7  2.0 (2.8)§ 5.5  2.0 (2.7) 5.5  1.9 (2.7) 5.6  2.1 (2.7)

2.59  0.39 (3.4)储 6.10  0.69 (7.5) 8.04  0.76 (9.6) 9.90  0.91 (11.7)

0.28  0.27 (0.8)储 0.29  0.27 (0.8) 0.34  0.28 (0.9) 0.45  0.39 (1.2)

3.51  0.51 (4.5)储 1.94  0.37 (2.7) 1.88  0.35 (2.6)

58.7  5.1 (49)¶ 61.0  5.5 (50) 66.5  6.3 (54)

35.0  14.7 (18) 28.8  12.2 (15) 28.3  11.8 (14)

2.54  0.29 (3.1) 5.67  0.59 (6.9) 7.46  0.64 (8.7)

0.18  0.13 (0.4) 0.26  0.21 (0.5) 0.28  0.27 (0.8)

2.54  0.29 (3.1) 3.22  0.42 (4.1) 1.79  0.30 (2.4)

54.8  5.3 (44) 64.7  5.4 (53) 66.7  6.4 (54)

Sensory Fibers Digit Wrist Below elbow Above elbow

*Mean  standard deviation (SD) in 65 patients, 13 to 74 years of age (average, 39), with no apparent disease of the peripheral nerves. † Amplitude measured from baseline to the negative peak. ‡ Latency measured to the onset after stimulation of the ulnar with the cathode 3 cm above the distal crease in the wrist. § Lower limits of normal, based on the distribution of the normative data. 储 Upper limits of normal, calculated as the mean  2 SD. ¶ Lower limits of normal, calculated as the mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

Recording arrangements for an antidromic sensory potential include a pair of ring electrodes placed around the thumb, with the disc electrode (E1) over the first web space or slightly more proximally in the snuffbox, and the reference electrode (E2) near the first dorsal interosseous or between the second and third metacarpals. The radial nerve subserves the sensation of the thumb by fibers originating from the C6 and C7 roots, which traverse the upper and middle trunk and enter the posterior cord.

Nerves of the Shoulder Girdle Phrenic Nerve Optimal placement of the needle along the posterior edge of the sternocleidomastoid muscle reduces the shock intensity necessary for maximal stimulation of the phrenic nerve. The contraction of the diaphragm, inducing hiccup

or interrupting voluntarily sustained vocalization, serves as a clinical sign of effective stimulation. Excessive stimulation voids selective recording by coactivating the brachial plexus, located posteriorly behind the anterior scalene muscle. The diaphragmatic action potential results in the seventh or eighth intercostal space near the costochondral junction being strongly positive and the xiphoid process mildly negative.245 The largest amplitude is recorded between the electrodes placed at these sites as the difference of the two potentials of opposite polarity. Brachial Plexus The brachial plexus comprises the anterior rami of the spinal nerves derived from the C5 through C8 and T1 roots. Thus stimulation at Erb’s point activates the proximal muscles of the shoulder girdle as well as the distal muscles, such as those of the thenar and hypothenar eminences. Table 35–8

Table 35–7. Latency Comparison between Median and Ulnar Nerves in the Same Limb* Site of Stimulation Motor Fibers Wrist Elbow Sensory Fibers Palm Wrist

Median Nerve (ms)

Ulnar Nerve (ms)

Difference (ms)

3.34  0.32 (4.0)† 7.39  0.72 (8.8)

2.56  0.37 (3.3)† 7.06  0.79 (8.6)

0.79  0.31 (1.4)† 0.59  0.60 (1.8)

1.33  0.21 (1.8) 2.80  0.32 (3.4)

1.19  0.22 (1.6) 2.55  0.30 (3.2)

0.22  0.17 (0.6) 0.29  0.21 (0.7)

*Mean  standard deviation (SD) in 35 patients, 14 to 74 years of age (average, 37) with no apparent disease of the peripheral nerve. † Upper limits of normal, calculated as mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

915

Nerve Conduction and Needle Electromyography

FIGURE 35–12 A, Motor nerve conduction study of the radial nerve. The sites of stimulation include Erb’s point (A), axilla (B), above the elbow (C), and mid-forearm (D). Compound muscle action potentials are recorded from the extensor indicis with a pair of surface electrodes. B, Sensory nerve conduction study of the radial nerve. The sites of stimulation include elbow (A) and distal forearm (B). Antidromic sensory potentials are recorded using the ring electrodes placed around the first digit. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

summarizes the conduction time across the brachial plexus calculated by subtracting the distal latency of the ulnar nerve.242 For a unilateral lesion, the side-to-side difference, which should not exceed 0.6 ms, serves as a more sensitive index than the absolute latency. The latency criteria, however, rarely provide useful information in the evaluation of axonal degeneration because remaining motor fibers tend to show a relatively normal value.305 In contrast, the amplitude of the recorded response determines the degree of axonal loss. Despite its considerable variability between the two sides in the same individual, amplitude preservation in the

injured side above one half of the normal side suggests limited distal degeneration and a good prognosis. Musculocutaneous Nerves Optimal sites of stimulation for motor conduction studies361 include the posterior cervical triangle 3 to 6 cm above the clavicle just behind the sternocleidomastoid muscle and the axilla between the axillary artery medially and the coracobrachialis muscle laterally.305 Either surface or needle electrodes may be used to stimulate the nerve and to record the muscle action potentials from the biceps brachii.

Table 35–8. Brachial Plexus Latency with Nerve Root Stimulation Latency across Plexus (ms)

Plexus

Site of Stimulation

Recording Site

Range

Mean

SD

Brachial (upper trunk and lateral cord) Brachial (posterior cord) Brachial (lower trunk and medial cord)

C5 and C6 C6, C7, and C8 C8 and T1, ulnar nerve

Biceps brachii Triceps brachii Abductor digiti quinti

4.8–6.2 4.4–6.1 3.7–5.5

5.3 5.4 4.7

0.4 0.4 0.5

From MacLean, I. C.: Nerve root stimulation to evaluate conduction across the brachial and lumbosacral plexuses. In Third Annual Continuing Education Course, American Association of Electromyography and Electrodiagnosis, Philadelphia, September 25, 1980, with permission.

916

Nerve Conduction and Electromyography

Cutaneous Nerves in the Forearm Lateral Antebrachial Cutaneous Nerves Stimulation applied between the tendon of the biceps medially and the brachioradialis laterally activates the sensory branch of the musculocutaneous nerve that runs superficially at the level of the elbow. Studies of this sensory potential detect abnormalities of the C6 root, upper trunk, or lateral cord better than the median sensory potentials recorded from the second digit, which, more often than not, receives sensory fibers from the C7 root and middle trunk.105 Medial Antebrachial Cutaneous Nerve The medial antebrachial cutaneous nerve subserves the sensation over the medial aspect of the forearm, the area not affected by lesions of the ulnar nerve. Like the ulnar nerve, it originates from the C8 and T1 roots, traverses the lower trunk and medial cord, and pierces the deep fascia in the lower end of the arm, thus bypassing the ulnar glove and cubital tunnel.211

Major Nerves in the Lower Limb Tibial Nerve Motor conduction studies consist of stimulating the tibial nerve at the popliteal fossa and at the ankle lateral to the medial malleolus and recording the muscle response from the abductor hallucis medially or abductor digiti quinti laterally (Fig. 35–13). Reported normal values (mean  SD) include distal latencies of 4.9  0.6 ms for the medial and 6.0  0.7 ms for the lateral plantar nerves over a 12-cm segment,165 and nerve conduction times, calculated as the latency difference between proximal and distal stimulation, of 3.8  0.5 ms for the medial and 3.9  0.5 ms for the lateral plantar nerves over a 10-cm segment across the medial malleolus.119 Tables 35–9 and 35–10 summarize the normal values in our laboratory. Common and Deep Peroneal Nerve Motor conduction studies consist of stimulation of the common peroneal nerve above and below the head of the fibula, and just above the ankle, and recording of muscle action

potentials from the extensor digitorum brevis (Fig. 35–14). Increasing the distance flanked by two sites of stimulation across the knee to 10 cm or longer improves the accuracy in the determination of conduction velocity. A series of shocks applied in short increments, however, delineates a focal conduction abnormality better.174,188 The extensor digitorum longus59 or tibialis anterior77 can usefully substitute for the atrophic extensor digitorum brevis, which may give rise to a small or no response, especially when assessing an advanced neuropathy. The use of needle electrodes and an averaging technique improves resolution in recording small sensory potentials of the deep peroneal nerve over the web between the first and second toes228 or mixed nerve potentials at the fibular head133 after stimulation of the peroneal nerve at the ankle. Tables 35–10 and 35–11 summarize the normal values in our laboratory. Superficial Peroneal Nerve This mixed nerve, originating from the L5 root, begins at the common peroneal bifurcation at the fibular head and gives rise to two sensory nerves in the lower third of the leg. The medial and intermediate dorsal cutaneous nerves subserve the skin of the dorsum of the foot, and the anterior and lateral aspects of the leg. Stimulation of the intermediate dorsal cutaneous branch with the cathode placed against the anterior edge of the fibula elicits the antidromic sensory potential at the ankles just medial to the lateral malleolus. The nerve may be stimulated at two points, 12 to 14 cm from the recording electrode and 8 to 9 cm further proximally, to differentiate the distal and proximal segments. The study helps distinguish a distal lesion from an L5 radiculopathy, which usually spares sensory potentials.169 Sural Nerve The sural nerve, primarily derived from the S1 root, originates from the tibial nerve in the popliteal fossa. It receives a communicating branch from the common peroneal nerve at the junction of the middle and lower third of the leg, where it becomes superficial. Antidromic sensory studies

FIGURE 35–13 Motor nerve conduction study of the tibial nerve. The sites of stimulation include knee (A), above the medial malleolus (B), and below the medial malleolus (C). Compound muscle action potentials are recorded with surface electrodes placed over the abductor hallucis. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

917

Nerve Conduction and Needle Electromyography

Table 35–9. Normal Values for Tibial Nerve*

Site of Stimulation Ankle Knee

Amplitude† (mV)

Latency‡ to Recording Site (ms)

Difference between Two Sides (ms)

Conduction Time between Two Points (ms)

Conduction Velocity (m/s)

5.8  1.9 (2.9)§ 5.1  2.2 (2.5)

3.96  1.00 (6.0)储 12.05  1.53 (15.1)

0.66  0.57 (1.8)储 0.79  0.61 (2.0)

8.09  1.09 (10.3)储

48.5  3.6 (41)¶

*Mean  standard deviation (SD) in 59 patients, 11 to 78 years of age (average, 39), with no apparent disease of the peripheral nerve. † Amplitude measured from the baseline to the negative peak. ‡ Latency measured to the onset with a standard distance of 10 cm between the cathode and the recording electrode. § Lower limits of normal, based on the distribution of the normative data. 储 Upper limits of normal, calculated as the mean  2 SD. ¶ Lower limits of normal, calculated as the mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

Table 35–10. Latency Comparison between Peroneal and Tibial Nerves in the Same Limb* Site of Stimulation Ankle Knee

Peroneal Nerve

Tibial Nerve

3.89  0.87 (5.6) 12.46  1.38 (15.2)

4.12  1.06 (6.2) 12.13  1.48 (15.1)

†

Difference †

0.77  0.65 (2.1)† 0.88  0.71 (2.3)

*Mean  standard deviation (SD) in 52 patients, 17 to 86 years of age (average, 41), with no apparent disease of the peripheral nerve. † Upper limits of normal, calculated as the mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

consist of stimulation of the nerve in the lower third of the leg over the posterior aspect slightly lateral to the midline and recording sensory potentials either around the lateral malleolus (Fig. 35–15) or more distally from the lateral dorsal cutaneous branch.228 In one study comparing three contiguous portions of 7 cm each, the most distal segment revealed a smaller mean velocity than the middle or proximal segment.367 Sural nerve study serves as one of the most sensitive measures for detecting electrophysiologic abnormalities in various types of neuropathies.366 In equivocal cases, the sural-to-radial amplitude ratio may help document abnormalities not apparent based on the absolute values. Preganglionic lesions such as S1 or S2 radiculopathy or a cauda equina involvement spare the sensory action potentials despite clinical sensory loss. This stands in contrast to postganglionic lesions, which consistently cause reduction in amplitude of sensory action potentials. Physiologic studies of the sural nerve provide a unique opportunity for direct histologic confirmation of pathology found in biopsied specimens.88

Nerves of the Pelvic Girdle Lumbosacral Plexus The lumbosacral plexus comprises the lumbar plexus, derived from the L2, L3, and L4 roots, and the sacral plexus, arising from the L5, S1, and S2 roots jointly. Their inaccessibility

to ordinary surface electrical stimulation prevents the use of conventional conduction studies to evaluate their integrity. The recording of the F wave and H reflex serves as an indirect measure of nerve conduction across this region. An alternative method involves needle stimulation99,100,242,254 or high-voltage surface stimulation162 of the L4, L5, or S1 spinal nerve just proximal to the plexus combined with distal stimulation of the plexus. The latency difference between the two stimulus sites equals the conduction time through the plexus. Femoral Nerve Shocks delivered to the femoral nerve above or below the inguinal ligament elicit a muscle potential of the rectus femoris at various distances from the point of stimulation. The latency of the recorded response increases progressively with the distance, reflecting the vertically oriented end-plate region in this muscle.127 The femoral nerve conduction averages 70 m/s, based on the latency difference between the two responses recorded at 14 and 30 cm from the point of stimulation (Table 35–12). This calculation holds only if the nerve branches supplying proximal and distal parts of the muscle have similar and directly comparable electrophysiologic characteristics. Saphenous Nerve The saphenous nerve, derived from the L3 and L4 roots, originates from the femoral nerve as its largest and longest

918

Nerve Conduction and Electromyography

FIGURE 35–14 Motor nerve conduction study of the common peroneal nerve. The sites of stimulation include above the knee (A), below the knee (B), and ankle (C). Compound muscle action potentials are recorded with surface electrodes over the extensor digitorum brevis. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

sensory branch, lying deep along the medial border of the tibialis anterior tendon. The nerve is stimulated with the surface electrodes pressed firmly between the tibia and medial gastrocnemius muscle, usually 12 to 14 cm above the ankle. Signal averaging facilitates the recording of small antidromic sensory potentials with a pair of electrodes placed just anterior to the highest prominence of the medial malleolus. Lateral Femoral Cutaneous Nerve The lateral femoral cutaneous nerve divides into large anterior and small lateral branches about 10 to 12 cm below the anterior superior iliac spine. Studies consist of surface stimulation of the nerve at this point to record orthodromic sensory potentials with a needle electrode inserted 1 cm medial to the lateral end of the inguinal ligament.240 An alternative technique depends on stimulation with a needle or surface electrode placed at the inguinal ligament, and recording of antidromic sensory potentials from the thigh.

HUMAN REFLEXES AND LATE RESPONSES Blink Reflex The blink reflex, elicited by stimulation of the trigeminal nerve, tests the integrity of both the trigeminal and facial nerves, including the proximal segment. A single shock to the supraorbital nerve evokes two separate contractile responses of the orbicularis oculi, R1 and R2. The latency of R1 represents the conduction time along the reflex pathway, including pontine relay, whereas the latency of R2 primarily reflects the excitability of interneurons and a delay of multiple synaptic transmissions. The upper limits of normal (mean  3 SD) include 13.0 ms for electrically elicited R1 and 16.7 ms for mechanically evoked R1. These compare to the distal latency of 4.1 ms after direct stimulation of the facial nerve. The latency difference between the two sides serves as the most sensitive criterion for unilateral lesions, with upper limits of 0.6 ms for distal latency and 1.2 ms and 1.6 ms for the latencies of electrically and mechanically evoked R1, respectively.

Table 35–11. Normal Values for Common and Deep Peroneal Nerves*

Site of Stimulation

Amplitude† (mV)

Latency‡ to Recording Site (ms)

Ankle Below knee Above knee

5.1  2.3 (2.5)§ 5.1  2.0 (2.5) 5.1  1.9 (2.5)

3.77  0.86 (5.5)储 10.79  1.06 (12.9) 12.51  1.17 (14.9)

Difference between Right and Left (ms)

Conduction Time between Two Points (ms)

Velocity (m/s)

0.62  0.61 (1.8)储 0.65  0.65 (2.0) 0.65  0.60 (1.9)

7.01  0.89 (8.8)储 1.72  0.40 (2.5)

48.3  3.9 (40)¶ 52.0  6.2 (40)

*Mean  standard deviation (SD) in 60 patients, 16 to 86 years of age (average, 41), with no apparent disease of the peripheral nerves. † Amplitude measured from the baseline to the negative peak. ‡ Latency measured to the onset with a standard distance of 7 cm between the cathode and the recording electrode. § Lower limits of normal, based on the distribution of the normative data. 储 Upper limits of normal, calculated as the mean  2 SD. ¶ Lower limits of normal, calculated as the mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

919

Nerve Conduction and Needle Electromyography

FIGURE 35–15 Antidromic sensory nerve conduction study of the sural nerve. The diagram shows stimulation on the calf slightly lateral to the midline in the lower third of the leg, and recording with surface electrodes placed behind the lateral malleolus. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

Tables 35–13 and 35–14 summarize a 10-year experience with the blink reflex in our laboratory. A substantial increase in latency of R1 implies demyelination of the central reflex pathway either in the pons,155,177,180,182 the trigeminal151,203,282 or facial nerve,198,202,292 or both.90,179,199,239 Posterior fossa tumors may increase the latency of R1 by compressing the trigeminal or facial nerve either extra-axially or intra-axially.58,178,199

H and T Reflexes Compared to clinical examination of the muscle stretch reflex, electrophysiologic recordings have the advantage of quantitating the response by objectively evaluating its briskness, velocity, or symmetry. The commonly used techniques to measure motor neuron excitability in spasticity and other related conditions include a mechanical tap to the Achilles tendon and electrical stimulation of the tibial nerve. In healthy adults, H reflexes are elicited only in the soleus muscle after stimulation of the tibial nerve and, less consistently, the flexor carpi radialis after stimulation of the median nerve, whereas mechanical stretch of any muscle evokes T reflexes. For example, stretch reflexes elicited by tapping the voluntarily contracted erector spinae comprise two components: the short latency R1, considered segmental in origin, and the long latency R2, induced by a suprasegmental pathway.351 In one study,218 mechanical

stimuli to the ankle and patella elicited T reflexes consistently in healthy subjects, suggesting their clinical value. For evaluation of localized lesions such as radiculopathy, studies of a longer segment including the normal portions of the reflex pathway tend to dilute the effect of focal slowing. Magnetic or electrical stimulation of the S1 nerve root evokes an M response via direct activation of the motor axons and, a few milliseconds later, an H reflex by reflexive excitation of the motor neurons through the group Ia afferent fibers. The peak latency difference between the simultaneously recorded M response and H reflex, or H-M interval, provides a reliable measure of conduction across a short segment within the spinal canal.98,241,289,390 In one study of 100 healthy subjects (Fig. 35–16),390 the H-M interval ranged from 2.6  0.7 ms (mean  SD) with stimulation at the T12 to L1, to 4.2  0.6 ms at the L2 to L3, 5.5  0.3 ms at the L4 to L5, and 6.8  0.5 ms at the S1 spinal process. Table 35–15 summarizes the normal values for the H reflex in healthy adults in our laboratory. In evaluating a unilateral lesion, the latency difference between the two sides provides the most sensitive measure of the T or H reflex.25 Unilateral absence or a right-left latency difference greater than 2.0 ms supports the diagnosis of S1 radiculopathy in the proper clinical context but does not by itself constitute sufficient evidence of a herniated disc. Preterm neonates have slower H reflex conduction velocity

Table 35–12. Normal Values for Femoral Nerve

No.

Age

Onset Latency (ms)

14 cm from stimulus point

42

8–79

3.7  0.45

30 cm from stimulus point

42

8–79

6.0  0.60

Stimulation Point

Recording Site

Just below inguinal ligament

Modified from Gassel, M. M.: A study of femoral nerve conduction time. Arch. Neurol. 9:57, 1963.

Conduction Velocity (m/s) 70  5.5 between the two recording sites

12 4

0

9 0 2 0

14

4

62 17 86

62

0

83 (glabellar tap, 21)* 90

0

88 0 20

0

13

63

0

Abs Delay

124

27 34 150

8

11

105

166

Nl

1

0 1 1

0

7

20

0

44

105 0 17

1

13

78

0

Abs Delay

79

19 33 154

7

8

82

166

Nl

R1 Right and Left Combined

10.5  0.8 (12.5  1.4)* 15.1  5.9 16.4  6.4 10.7  0.8

17.0  3.7 10.1  0.6 11.4  1.2 12.3  2.7

4.2  2.1 5.8  2.6 2.7  0.2

6.7  2.7 2.9  0.4 3.4  0.6 2.9  0.5

R1 (ms)

2.9  0.4

Direct Response (ms)

4.3  0.9

2.8  0.9 3.6  0.6 3.4  0.5

3.9  0.4

3.1  0.5

3.9  1.3

3.6  0.5

R/D Ratio

Abs  absent response; Nl  normal. *R1 elicited bilaterally by a midline glabellar tap in another group of 21 healthy subjects. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

Guillain-Barré syndrome Chronic inflammatory polyneuropathy Fisher syndrome Hereditary sensorimotor neuropathy Type I Type II Diabetic polyneuropathy Multiple sclerosis

Normal

Category

Number of Patients

Direct Response Right and Left Combined

35.8  8.4

39.5  5.7 30.1  3.8 33.7  4.6

31.8  1.3

39.5  9.4

37.4  8.9

30.5  3.4

Ipsilateral R2 (ms)

37.7  8.0

39.3  6.4 30.1  3.7 34.8  5.3

31.4  1.9

42.0  10.3

37.7  8.4

30.5  4.4

Contralateral R2 (ms)

Table 35–13. Blink Reflex Elicited by Electrical Stimulation of Supraorbital Nerve in Normal Subjects and Patients with Bilateral Neurologic Diseases (Mean  SD)

921

Nerve Conduction and Needle Electromyography

Table 35–14. Blink Reflex Elicited by Electrical Stimulation of Supraorbital Nerve on the Affected and Normal Sides in Patients with Unilateral Neurologic Diseases (Mean  SD) Category and Side of Stimulation Trigeminal neuralgia Affected side Normal side Compressive lesion of the trigeminal nerve Affected side Normal side Bell’s palsy Affected side Normal side Acoustic neuroma Affected side Normal side Wallenberg’s syndrome Affected side Normal side

Number of Patients

Direct Response (ms)

R1 (ms)

R/D Ratio

Ipsilateral R2 (ms)

89 89

2.9  0.4 2.9  0.5

10.6  1.0 10.5  0.9

3.7  0.6 3.7  0.6

30.4  4.4 30.5  4.2

31.6  4.5 31.1  4.7

17 17

3.1  0.5 3.2  0.6

11.9  1.8 10.3  1.1

3.9  1.0 3.4  0.6

36.0  5.5 33.7  3.5

37.2  5.7 34.8  4.1

100 100

2.9  0.6 2.8  0.4

12.8  1.6 10.2  1.0

4.4  0.9 3.7  0.6

33.9  4.9 30.5  4.3

30.5  4.9 34.0  5.4

26 26

3.2  0.7 2.9  0.4

14.0  2.7 10.9  0.9

4.6  1.7 3.8  0.5

38.2  8.2 33.1  3.5

36.6  8.2 35.3  4.5

23 23

3.2  0.6 3.2  0.4

10.9  0.7 10.7  0.5

3.6  0.6 3.4  0.4

40.7  4.6 34.0  5.7

38.4  7.1 35.1  5.8

Category R2 (ms)

From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

compared to full-term babies.259 Normal values (mean  SD) for the soleus H reflex established in 83 preterm and term infants include a latency of 19.2  2.16 ms for conceptional ages of 31 to 34 weeks, 16.7  1.5 ms for 35 to 39 weeks, and 15.9  1.5 ms for 40 to 45 weeks.31

body (intra-axial mesencephalic nucleus and extra-axial craniospinal ganglia). Similarly, ganglionopathy, but not axonal sensory neuropathy, tends to spare the masseter reflex in patients with pure sensory symptoms.12

F Waves Masseter Reflex A sharp tap to the mandible stretches the muscle spindles of the masseter muscle, causing the jaw reflex, or masseteric T reflex.155 The jaw reflex shows physiologic characteristics different from those of the spinal monosynaptic reflex. For example, muscle vibration that inhibits the soleus T and H reflexes potentiates the masseteric T and H reflexes.136,137 The criteria for abnormality established in one study, with the use of a needle recording electrode,283 comprised unilateral absence of the reflex, a difference of more than 0.5 ms in latency between the two sides, or bilateral absence of the reflex up to the age of 70 years. Table 35–16 summarizes normal values in our laboratory.195 Despite technical problems in standardizing the mechanical stimulus and regulating the tonus of the masseter muscle for optimal activation, an unequivocal unilateral delay or absent response suggests a lesion of the trigeminal nerve or the brainstem.203 Patients with Friedreich’s ataxia have hypoactive or no stretch reflexes in the upper and lower limbs, but masseter reflexes are normal or paradoxically hyperactive. This discrepancy may result from a different location of the afferent nerve cell

The F wave, which results from the backfiring of antidromically activated anterior horn cells, serves as a measure of motor conduction along the entire peripheral axon, including the most proximal segment.56,91,107,117,196,197,206,284,285,311,388 Recording Procedures For clinical studies, routine procedures include stimulation of the median and ulnar nerves at the wrist and elbow and of the tibial and peroneal nerves at the ankle and knee. In general, the sum of the F latency and M latency remains the same regardless of stimulus sites. Thus Fd  Md  Fp  Mp, where Fd and Fp are F latencies and Md and Mp are M latencies with distal and proximal stimulation, respectively. When necessary, this equation provides an estimated latency of the F wave from any proximal site by subtracting the proximal M latency from the sum of the distal F and M latencies. Stimulation of the facial nerve also elicits an F wave,324 although, in the short segment under study, superimposition of the M response usually makes its recognition difficult. Furthermore, inadvertent stimulation of neighboring trigeminal afferent fibers may simultaneously activate reflex responses that may mimic the late response.

922

Nerve Conduction and Electromyography

FIGURE 35–16 A, Schematic representation of soleus H reflexes from electrical stimulation of the S1 nerve root at the S1 foramen and of the tibial nerve at the popliteal fossa. B, The response complex of the H reflex and M response in one of the subjects elicited by magnetic (upper traces) and electrical (lower traces) stimulation of the S1 nerve root at the S1 foramen. The intensity of the nerve stimulus is to the right of the traces. (From Zhu, Y., Starr, A., Haldeman, S., et al.: Soleus H-reflex to S1 nerve root stimulation. Electroencephalogr. Clin. Neurophysiol. 109:10, 1998, with permission.)

Central Latency Central latency from the stimulus point to and from the spinal cord is derived as F  M, where F and M are latencies of the F wave and the M response (Fig. 35–17). Subtracting an estimated delay of 1.0 ms for the turnaround time at the cell and dividing by 2 ([F  M  1]/2) provides the conduction time along the proximal segment

from the stimulus site to the spinal cord. Although no human studies measured the central activation time at the anterior horn cells, animal data indicate a delay of nearly 1.0 ms.306 On one hand, the absolute refractory period of the fastest human motor fibers lasts about 1.0 ms or slightly less.184 Thus the recurrent discharge cannot propagate distally across the refractory period of

Table 35–15. Normal Values for H Reflex*

Amplitude† (mV) 2.4  1.4

Difference between Right and Left (mV)

Latency‡ to Recording Site (ms)

Difference between Right and Left (ms)

1.2  1.2

29.5  2.4 (35)§

0.6  0.4 (1.4)§

*Mean  standard deviation (SD) in the same 59 patients shown in Table 35–9. † Amplitude measured from the baseline to the negative peak. ‡ Latency measured to the onset. § Upper limits of normal calculated as mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

923

Nerve Conduction and Needle Electromyography

Table 35–16. Latency and Amplitude of Masseter Reflex in 20 Normal Subjects Latency (ms) Mean right Mean left Total SD Mean  3 SD

Latency Difference (Large Value Minus Small Value)

Amplitude (mV)

0.27 0.15 0.8

0.23 0.21 0.22 0.24 Variable

7.10 7.06 7.08 0.62 9.0

Amplitude Ratio (Large Value over Small Value)

1.44 0.42 2.7

From Kimura, J.: Needle electromyography. In Bertorini, R. (ed.): Evaluation and Diagnostic Tests in Neuromuscular Disorders. Woburn, MA, Butterworth-Heinemann, p. 331, 2002, with permission.

the axon hillock for 1.0 ms after the passage of an antidromic impulse. On the other hand, the somatodendrite spike would abate after Renshaw inhibition activated by an antidromic impulse with a synaptic delay of 1.0 ms. A narrow window based on these limiting factors allows backfiring to materialize in only 1% to 5% of the motor neuron population invaded by antidromic impulses.

FIGURE 35–17 The latency difference between the F wave and the M response represents the passage of a motor impulse to and from the cord through the proximal segment. Considering an estimated minimal delay of 1.0 ms at the motor neuron pool, the proximal latency from the stimulus site to the cord equals (F  M  1)/2, where F and M are latencies of the F wave and M response, respectively. In the segment to and from the spinal cord, FWCV  (D 2)/(F  M  1), where D is the distance from the stimulus site to the cord, (F  M  1)/2 is the time required to cover the length D, and FWCV is F-wave conduction velocity. Dividing the conduction time in the proximal segment to the cord by that of the remaining distal segment to the muscle, the F ratio  (F  M  1)/2M, where (F  M  1)/2 and M are proximal and distal latencies, respectively. (From Kimura, J.: Proximal versus distal slowing of motor nerve conduction velocity in the Guillain-Barré syndrome. Ann. Neurol. 3:344, 1978, with permission.)

F-Wave Conduction Velocity F-wave latencies suffice in studying a limb of average length or in documenting sequential changes in the same subject. For unilateral lesions affecting one nerve, latency comparison between the right and left sides in the same subject or of one nerve with another in the same limb serves as a sensitive measure of abnormality (Tables 35–17 and 35–18).185 Otherwise, clinical assessment of F-wave latency calls for a surface determination of the limb length or the patient’s height with the use of a nomogram359 to adjust for individual differences.350 In the upper limbs, the surface measurement should follow from the stimulus point to the C7 spinous process via the axilla and midclavicular point to closely estimate the nerve length corresponding to the central latency.181 In the lower limb, surface measurement approximates the nerve course best if done from the stimulus site to the T12 spinous process by way of the knee and greater trochanter of the femur.196 The estimated length of a nerve segment by surface measurement correlates well with its F-wave latency. Observations in cadavers showed a good agreement between surface determinations and actual lengths of the nerves in the upper238 as well as lower limbs.196 F-wave conduction velocity (FWCV) in the proximal segment equals the estimated nerve length divided by the conduction from the stimulus point to the spinal cord. Thus FWCV  (2D)/(F  M  1) where D represents the distance from the stimulus site to the cord and (F  M  1)/2 is the central latency, or the time required to cover the length. Clinical Value and Limitations Neurologic disorders showing consistent F-wave abnormalities include hereditary motor sensory neuropathy,181,284 acute or chronic demyelinating neuropathy,197,206 diabetic neuropathy,60,205 uremic neuropathy,1,286 alcoholic neuropathy,231 and a variety of other neuropathies. F waves show less consistent changes in such diverse categories of disorders as entrapment neuropathies,91,386 amyotrophic lateral sclerosis,10 radiculopathies,110 and cervical syringomyelia.291

Wrist Elbow Axilla储 Wrist Above elbow Axilla储 Ankle Above knee Ankle Knee

Site of Stimulation

0.73  0.54 (1.8) 1.42  1.03 (3.5) 1.28  0.91 (3.1) 1.40  1.04 (3.5) 1.25  0.92 (3.1)

20.3  1.6 (24) 48.4  4.0 (56) 39.9  3.2 (46) 47.7  5.0 (58) 39.6  4.4 (48)

43.8  4.5 (53) 27.6  3.2 (34)

10.4  1.1 (13) 44.7  3.8 (52) 27.3  2.4 (32)

23.0  2.1 (27)¶ 15.4  1.4 (18) 10.6  1.5 (14) 25.0  2.1 (29) 16.0  1.2 (18)

0.95  0.67 (2.3)¶ 0.76  0.56 (1.9) 0.85  0.61 (2.1) 1.0  0.83 (2.7) 0.68  0.48 (1.6)

26.6  2.2 (31)¶ 22.8  1.9 (27) 20.4  1.9 (24) 27.6  2.2 (32) 23.1  1.7 (27)

1.52  1.02 (3.6) 1.23  0.88 (3.0)

0.76  0.52 (1.8) 1.28  0.90 (3.1) 1.18  0.89 (3.0)

0.93  0.62 (2.2)¶ 0.71  0.52 (1.8) 0.85  0.58 (2.0) 0.84  0.59 (2.0) 0.73  0.52 (1.8)

Difference between Right and Left (ms)

52.6  4.3 (44) 53.7  4.8 (44)

1.11  0.11 (0.89–1.33)

1.05  0.09 (0.87–1.23)

1.05  0.09 (0.87–1.23)

65.3  4.8 (55) 65.7  5.3 (55) 49.8  3.6 (43) 55.1  4.6 (46)

0.98  0.08 (0.82–1.14)¶**

F Ratio§ between Proximal and Distal Segments

65.3  4.7 (56)** 67.8  5.8 (56)

Conduction Velocity‡ to and from the Spinal Cord (m/s)

*Mean  standard deviation (SD) in the same patients shown in Tables 35–4, 35–6, 35–9, and 35–11. † Central latency  F  M, where F and M are latencies of the F wave and M response, respectively. ‡ Conduction velocity  2D/(F  M  1), where D is the distance from the stimulus point to C7 or T12 spinous process. § F ratio  (F  M  1)/2M with stimulation with the cathode on the volar crease at the elbow (median), 3 cm above the medial epicondyle (ulnar), just above the head of fibula (peroneal), and in the popliteal fossa (tibial). 储 F(A)  F(E)  M(E)  M(A), where F(A) and F(E) are latencies of the F wave with stimulation at the axilla and elbow, respectively, and M(A) and M(E) are latencies of the corresponding M response. ¶ Upper limits of normal calculated as mean  2 SD. **Lower limits of normal calculated as mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

120 peroneal nerves from 60 subjects 118 tibial nerves from 59 subjects

122 median nerves from 61 subjects 130 ulnar nerves from 65 subjects

Number of Nerves Tested

Central Latency† to and from the Spinal Cord (ms)

Difference between Right and Left (ms)

F-Wave Latency to Recording Site (ms)

Table 35–17. F Waves in Normal Subjects*

Wrist Elbow

Ankle Knee

70 nerves from 35 patients

104 nerves from 52 patients

Tibial Nerve 48.1  4.2 (57) 40.1  3.7 (48)

Peroneal Nerve 47.7  4.0 (55) 39.6  3.7 (47)

‡

1.68  1.21 (4.1) 1.71  1.19 (4.1)

Difference

1.00  0.68 (2.4) 0.84  0.55 (1.9)

27.2  2.5 (32) 23.0  1.7 (26)

26.6  2.3 (31) 22.9  1.8 (26)

Difference

‡

‡

Ulnar Nerve

Median Nerve

43.6  4.0 (52) 27.1  2.9 (33)

Peroneal Nerve

23.3  2.2 (28) 15.5  1.4 (18)

‡

Median Nerve

‡

44.1  3.9 (52) 28.0  2.7 (33)

Tibial Nerve

24.5  2.4 (29) 16.0  1.2 (18)

Ulnar Nerve

Difference

1.79  1.20 (4.2) 1.75  1.07 (3.9)

Difference

1.24  0.75 (2.7)‡ 0.79  0.65 (2.1)

Central Latency† to and from the Spinal Cord

*Mean  standard deviation (SD) in the same patients shown in Tables 35–7 and 35–10. † Central latency  F  M, where F and M are latencies of the F wave and M response, respectively. ‡ Upper limits of normal calculated as mean  2 SD. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

Site of Stimulation

Number of Nerves Tested

F-Wave Latency to Recording Site

Table 35–18. Comparison between Two Nerves in the Same Limb*

926

Nerve Conduction and Electromyography

Normal Values Tables 35–17 and 35–18 summarize the normal ranges and the upper and lower limits of F-wave latency and other aspects of the F-wave values, defined as 2 SD around the mean, in the same control subjects whose data were presented in the preceding section on nerve conduction studies. Neonates and infants tend to have large F waves probably because the immaturity of physiologic inhibition renders the anterior horn cells more excitable than in adults.259 The F-wave latency remains relatively constant during childhood because a rapid change in conduction velocity compensates for the increase in arm length. In one study,257 the minimal latency of F waves elicited with stimulation of the median nerve at the wrist averaged 17 ms in neonates (1 to 28 days), 15 ms in infants (1 month to 1 year), and 16 ms in children (2 to 12 years). The F-wave latency then increases, reaching its maximal value seen at 20 years of age.219 An older group of subjects has longer F-wave latencies compared to young healthy subjects.268

A Wave With a submaximal stimulation, activating one branch of an axon but not the other, the antidromic impulse propagated up to the point of division may turn around to proceed distally along the second fiber not excited by the stimulus. This orthodromic impulse induces a constant late response, called an A wave, a term preferred over the traditional designation, “axon reflex,” which erroneously implies a reflexive origin. As suggested by its alternative description, “intermediate latency response,” the A wave usually, though not always, appears between the M response and F wave.122 A decrease in A-wave latencies with more proximal stimulation indicates that the response results from an initially antidromic passage of impulse analogous to the F wave (Fig. 35–18). Pathophysiologic mechanisms possibly involving a majority of A waves include, in addition to the less common collateral sprouting described above, ephaptic and ectopic discharges generated at a hyperexcitable site along the proximal portion of the nerve.19,243 Shocks of a higher intensity, activating both branches distally, can eliminate the A waves generated by collateral sprouting because two antidromic impulses now collide at the turning point. Thus supramaximal stimuli normally abolish the collateral A wave altogether, unless the structural changes of the nerve or surrounding tissues prevent the current from activating one of the branches. In contrast, an ephaptic A wave may persist even with the use of very highintensity stimuli because the antidromic impulse of the fastconducting axon may have already passed the site of ephaptic transmission induced by a neighboring slow-conducting axon, thus precluding the collision. An increase in shock intensity also fails to inhibit an ectopic A wave generated by antidromic passage of an impulse across a hyperexcitable segment of a nerve branch. In this case, paired shocks abolish the A wave because of the collision between the ectopically

generated orthodromic impulse and the second antidromic impulse. With repetitive stimuli, even numbered shocks cause collision for the same reason, producing ectopic A waves only in response to odd-numbered stimuli. The presence of A waves signals a heterogeneous group of neurogenic abnormalities encompassing acute and chronic neuropathies widely varying in pathophysiology, from nerve regeneration to demyelination. The disease entities commonly associated with the A wave include entrapment syndromes, brachial plexus lesions, diabetic neuropathy, hereditary motor sensory neuropathy, facial neuropathy, amyotrophic lateral sclerosis, Guillain-Barré syndrome, and cervical root lesions.19,122,210,315

CLINICAL APPLICATIONS Common Sources of Error Although simple principles dictate the nerve conduction studies, pitfalls abound in practice mostly because of technical problems that usually account for unexpected findings. Often overlooked sources of error include intermittent power failure in the stimulating or recording system, excessive spread of stimulation current, anomalous innervation, temporal dispersion, and inaccuracy of surface measurement. Stationary far-field peaks may result from a moving source not only in a referential but also in a bipolar derivation, usually used for recording near-field potentials. Lack of awareness of these possibilities can cause confusion in the interpretation of the data, suggesting an incorrect diagnosis, especially if the findings mimic common manifestations of a disease. Special care in regard to technical details leads to good quality control, which in turn improves the reproducibility of the results.

Collision Technique An appropriately placed cathode can selectively activate the median or ulnar nerve at the wrist or elbow, but usually not at the axilla, where the two nerves lie in close proximity. If the current intended for the median nerve spreads to the ulnar nerve when studying carpal tunnel syndrome, the thenar eminence potentials comprise not only a slowed median but also a normal ulnar component. The measured onset latency will then indicate the normal ulnar response, which precedes the median response (Fig. 35–19). In the same case, selective stimulation of the median nerve at the elbow reveals a prolonged latency. The calculated conduction time between the ulnar response from the axilla and the median response from the elbow would suggest an erroneously fast conduction velocity. In extreme cases, the median response from the elbow has a greater latency than the ulnar component from the axilla. The reverse discrepancy can occur in a study of tardy ulnar palsy, with spread of axillary stimulation to the median nerve.

FIGURE 35–18 A, A 51-year-old man with low back pain. Stimulation of the right tibial nerve at the ankle elicited a number of A waves. A series of eight tracings displayed with stepwise vertical shift of the baseline confirm the consistency. This type of display not only facilitates the selection of the F wave with minimal latency but also allows individual assessments of all the late responses. Of the three A waves (small arrows, 1, 2, and 3) elicited by weak shocks (S(1)), stronger shocks (S(2)) eliminated only the earliest response. B, Collateral sprouting in the proximal part of the nerve. A strong shock, activating both branches, can eliminate the A wave generated by weak stimulation by collision. (From Fullerton P. M., and Gilliatt, R. W.: Axon reflexes in human motor nerve fibres. J. Neurol. Neurosurg. Psychiatry 28:1, 1965, with permission.) C, A waves after stimulation of the left tibial nerve at the ankle or knee. Proximal stimulation eliminated the A wave (arrow) that followed the F wave with distal stimulation. D, A 50-year-old man with recurrent backaches following laminectomy. Stimulation of the tibial nerve at the ankle or knee elicited the A wave (arrow). Like the F wave, the latency of the A wave decreased with proximal site of stimulation. This indicates that the impulse first propagates in the centripetal direction. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

928

Nerve Conduction and Electromyography

FIGURE 35–19 A 39-year-old man with carpal tunnel syndrome. The stimulation of the median nerve at the wrist (S1) or elbow (S2) elicited a muscle action potential with increased latency in the thenar eminence. Spread of axillary stimulation (S3) to the ulnar nerve (third tracing from top) activated ulnar-innervated thenar muscles with shorter latency. Another stimulus (S4) applied to the ulnar nerve at the wrist (bottom tracing) blocked the proximal impulses by collision. The muscle action potential elicited by S4 occurred much earlier. The diagram on the left shows collision between the orthodromic (solid arrows) and antidromic (dotted arrows) impulses. (From Kimura, J.: Collision technique— physiological block of nerve impulses in studies of motor nerve conduction velocity. Neurology [Minneap.] 26:680, 1976, with permission.)

The addition of distal stimulation achieves a physiologic nerve block through collision, allowing selective recording of the median or ulnar component despite coactivation of both nerves proximally.183 In studying the median nerve, for example, a distal stimulus is delivered to the ulnar nerve at the wrist. The antidromic impulse generated here collides with the orthodromic ulnar impulse from the axilla, leaving only the median impulse to reach the muscle. Delaying the proximal stimulus a few milliseconds accomplishes an optimal separation between the ulnar response induced by the distal stimulus and the median response under study. If this delay exceeds the conduction time between the distal and proximal points of stimulation, the antidromic impulse from the wrist passes the stimulus site at the axilla without collision. By the same principles, the use of a distal stimulus can block the median nerve in selective recording of the ulnar response after coactivation of both nerves at the axilla.

Anomalous Innervations Martin-Gruber Anastomosis Martin (in 1763)247 and Gruber (in 1870)140 demonstrated frequent communication from the median to the ulnar nerve at the level of the forearm. This anastomosis, usually

seen a few centimeters distal to the medial humeral epicondyle,369 often originates from the anterior interosseous nerve and predominantly involves motor axons. Rare sensory contribution, when present, may follow a different distribution.57,372 The number of axons taking the anomalous course varies widely, usually, though not always,232 supplying ordinarily ulnar-innervated intrinsic hand muscles, most notably the first dorsal interosseous, adductor pollicis, and abductor digiti minimi.382 The possibility of selectively stimulating the anomalous fibers at the elbow without coactivating the median nerve proper (Fig. 35–20)191 suggests a grouping of the anastomotic fibers in a separate bundle, rather than intermingling with the median nerve proper. The anomaly is reported in 15% to 32% of an unselected population, often bilaterally.9,201 A higher incidence among congenitally abnormal fetuses in general and those with trisomy 21 in particular favors its phylogenetic origin.340 Other anomalies related to Martin-Gruber anastomosis include rare communication crossing from the ulnar to the median nerve in the forearm,138,278 occasionally involving only the sensory axons,154 and innervation of the ulnar aspect of the dorsum of the hand by the superficial radial sensory nerve.251 Careful waveform analysis of the muscle action potentials can confirm the presence of a Martin-Gruber anastomosis if suspected during nerve conduction studies. Thus

Nerve Conduction and Needle Electromyography

929

FIGURE 35–20 A 46-year-old woman with carpal tunnel syndrome and the Martin-Gruber anastomosis. Stimulation at the elbow (S2) activated not only the median nerve but also communicating fibers, giving rise to a complex compound muscle action potential. With proper adjustment of electrode position and shock intensity, another stimulus at the elbow (S2) excited the median nerve selectively without activating the anastomosis. Another stimulus (S3) applied to the ulnar nerve at the wrist (bottom tracing) achieved the same effect by blocking the unwanted impulse transmitted through the communicating fibers. (From Kimura, J.: Electromyography and Nerve Stimulation Techniques: Clinical Applications [in Japanese]. Tokyo, Igaku-Shoin, 1990, with permission.)

stimulation of the median nerve at the elbow coactivates the communicating ulnar nerve fibers, producing action potentials of not only the median-innervated thenar muscle but also anomalously innervated thenar and hypothenar muscles. In contrast, stimulation of the median nerve at the wrist evokes a smaller response lacking the ulnar component. Studies of the ulnar nerve in the presence of Martin-Gruber anastomosis also show confusing results.208 Stimulation at the elbow, sparing the communicating branch still attached to the median nerve, evokes only a partial response, whereas stimulation at the wrist activates the additional anomalous fibers, giving rise to a full response, mimicking a conduction block of the ulnar nerve at the elbow.246 The collision technique183,323 allows selective blocking of unwanted impulses transmitted via the communicating fibers. If a patient with this anomaly develops carpal tunnel syndrome, stimulation of the median nerve at the elbow reveals two temporally dispersed potentials, a normal ulnar and a delayed median response. The initial ulnar component may give a mistaken notion of normalcy of the median nerve unless the anomaly is detected. In the same patient, stimulation at the wrist only elicits a delayed median component without an ulnar response.183,220 The short latency difference incorrectly calculated between the ulnar and median responses would lead to an unreasonably fast conduction velocity from the elbow to the wrist.41,181,220

The anomalously innervated ulnar muscles located at some distance from the recording site on the thenar eminence often, though not always, display an initial positive deflection.143 In most cases, a collision technique can block impulses in the anomalous fibers without affecting those transmitted along the median nerve proper to clarify the ambiguity. Severance or substantial injury of the ulnar nerve at the elbow ordinarily leads to inexcitability of the distal segment concomitant with evolving wallerian degeneration. In the presence of this anomaly, the communicating fibers bypass the lesion, maintaining a normal supply of anomalously innervated muscle fibers. Stimulation of the ulnar nerve at the wrist, therefore, evokes a small but otherwise normal muscle action potential. As an extreme example, an injury that would usually separate the ulnar nerve completely will not appreciably affect the intrinsic hand muscles if all or nearly all ulnar fibers are attached to the median nerve at the elbow. In this rare condition, called all-median hand, the intrinsic hand muscles usually supplied by the ulnar nerve receive innervation via the communicating fibers.244 Normal motor unit potentials in the ulnar-innervated muscles, despite severe damage to the ulnar nerve at the elbow, may signal the presence of this anomaly. Conversely, studies will reveal spontaneous discharges in the ulnar-innervated intrinsic hand muscles after an injury to the median nerve at the elbow.

930

Nerve Conduction and Electromyography

Accessory Deep Peroneal Nerve The extensor digitorum brevis, the muscle commonly used in peroneal nerve conduction studies, usually derives its supply from the deep peroneal nerve, a major branch of the common peroneal nerve. As the most frequent anomaly of the lower limb, seen in 20% to 28% of unselected subjects, this muscle may receive innervation from the superficial peroneal nerve via a communicating branch, called the accessory deep peroneal nerve (Fig. 35–21). The anomalous fibers descend on the lateral aspect of the leg after arising from the superficial peroneal nerve, then pass behind the lateral malleolus before proceeding anteriorly to innervate the lateral portion of the extensor digitorum brevis.163,221 The anomaly, inherited as dominant trait,63 occasionally supplies the extensor digitorum brevis

exclusively with no contribution from the common peroneal nerve.270 In the presence of this anastomosis, the compound muscle action potential evoked by stimulation of the common peroneal nerve at the knee equals the sum of responses elicited by stimulation of the deep peroneal nerve at the ankle and stimulation of the accessory deep peroneal nerve behind the lateral malleolus. Injury to the deep peroneal nerve causes weakness of the tibialis anterior, extensor digitorum longus, extensor hallucis longus, and medial but not lateral part of the extensor digitorum brevis supplied by the anastomosis.76,142 The collision technique181 may help identify isolated abnormalities and localize a lesion to various branches, for example, the accessory deep peroneal nerve.323

FIGURE 35–21 A, Compound muscle action potentials recorded from surface electrodes over the extensor digitorum brevis after a maximal stimulus to the common peroneal nerve at the knee (A), deep peroneal nerve on the dorsum of the ankle (B), accessory deep peroneal nerve posterior to the lateral malleolus (C and D), and tibial nerve posterior to the medial malleolus (E) at the ankle. Diagram of the foot indicates the site of block (X) of the accessory deep peroneal nerve with 2% lidocaine and the points of stimulation (B, C, and D) and recording (R). B, Course of the accessory deep peroneal nerve and action potentials recorded with coaxial needle electrode (R) in the lateral belly of the extensor digitorum brevis muscle following stimulation of the common peroneal nerve at the knee (A), just below the head of the fibula (B), superficial peroneal nerve (C), accessory deep peroneal nerve posterior to the lateral malleolus (D), and deep peroneal nerve on the dorsum of the ankle (E). The volume-conducted potential from the medial bellies of the extensor digitorum brevis (E) reduces the amplitude of the action potential of the lateral belly with simultaneous stimulation of the common peroneal nerve at A or B. (From Lambert, E. H.: The accessory deep peroneal nerve: a common variation in innervation of extensor digitorum brevis. Neurology [Minneap.] 19:1169, 1969, with permission.)

Nerve Conduction and Needle Electromyography

Temporal Dispersion and Phase Cancellation Physiologic Temporal Dispersion In the study of nerve function, not only calculation of the maximal velocities based on the onset latency, but also waveform analyses of the recorded response play an essential role. The latter aspect of the study provides a measure of greater importance, especially in evaluating peripheral neuropathies with segmental block, in which surviving axons conduct normally. In clinical tests of motor and sensory conduction, the size of the recorded response serves as a measure of the number of excitable nerve axons. Any discrepancy between responses to proximal and distal shocks, however, does not necessarily imply the presence of conduction block or other abnormalities. The impulses of physiologically slow- and fast-conducting nerve fibers become increasingly separated in latency over a long conduction path.43,66,220 Thus the longer the distance between stimulating and pickup electrodes, the greater the temporal dispersion among different nerve fibers. Because of this desynchronization, a response recorded at a point distant from the stimulation site becomes lower in amplitude and longer in duration and, contrary to the common belief, smaller in area under the waveform. This diminution of evoked responses associated with temporal dispersion is caused by phase cancellation, which affects nerve action potentials more than muscle responses, as described below. A slight latency difference could line up the negative and positive peaks of short-duration diphasic sensory spikes of the fast and slow fibers, causing them to cancel each other (Fig. 35–22). In contrast, the same slight shift in latency still superimposes longer duration motor unit potentials nearly in phase rather than out of phase, without much cancellation. Indeed, the same degree of temporal dispersion, if measured in percentage, gives rise to a much greater increase in duration of the sensory potential than that of the muscle response.200 As expected from the principle of duration-dependent phase cancellation, a physiologic temporal dispersion also substantially reduces the size of short-duration muscle action potentials such as those recorded from intrinsic foot muscles. A number of technical factors influence the degree of physiologic phase cancellation. Of particular importance is the interelectrode distance between a pair of recording electrodes that dictates the duration of the recorded response. Pathologic Temporal Dispersion If the latency increases for some but not other axons, as might be expected in demyelinating neuropathy, compound muscle action potentials also diminish dramatically based solely on phase cancellation between normally conducting and pathologically slow fibers. As predicted by our model,204 this type of phase cancellation reduces the amplitude of the muscle response well beyond the usual

931

physiologic limits. This finding could give rise to a false impression of motor conduction block, and probably accounts for an occasionally encountered discrepancy between diminished size of proximally evoked responses and relatively normal recruitment of the motor unit potentials with preserved strength. Model for Phase Cancellation A simple model tests the effects of desynchronized inputs204 by stimulating the median (S1) and ulnar (S2) nerve at varying interstimulus intervals at the wrist, and recording a sensory potential of the fourth digit and a muscle potential over the thenar eminence. As predicted, the sensory potential declines nearly 50%, with a 1.0- to 1.5-ms delay of S2 after S1, whereas the muscle response reaches a minimal size with a latency change of 5 to 6 ms. These findings indicate that a latency change on the order of one half the total duration of unit discharge leads to a maximal phase cancellation and consequently loss of amplitude and area. With further separation, the two units no longer overlap out of phase to cause cancellation, and in fact may paradoxically increase the size of the response because excessive desynchronization may now counter the physiologic phase cancellation. Because variables such as interelectrode distance between E1 and E2 influence the outcome substantially, the commonly held criteria based on percentage reduction are untenable except in entirely standardized studies.224 Thus comparison between distally and proximally elicited responses often fails to differentiate physiologic, as opposed to pathologic, temporal dispersion, not to mention conduction block. An alternative method relies on segmental stimulation at more than two sites to test a linear relationship between the latency and the size of the recorded responses seen in physiologic phase cancellation.193 This approach enjoys the distinct advantage of having a built-in internal control for all recording variables such as interelectrode spacing (Fig. 35–23). A nonlinear change in amplitude or waveform indicates either a conduction block or a pathologic temporal dispersion (Fig. 35–24). The distinction between the two possibilities must, in part, depend on clinical clues and needle electrode study of motor unit recruitment as described below.

Practical Assessment of Conduction Block: Criteria for Conduction Block Motor nerve conduction studies play a key role in the differential diagnosis of clinical weakness. A conduction block usually implies demyelination,102 although other conditions such as ischemia and channelopathies can cause similar reversible changes.132,153 Additional characteristics of demyelination include slowing of nerve conduction, pathologic temporal dispersion, and repetitive discharges. These abnormalities may also alter the waveform of the evoked

932

Nerve Conduction and Electromyography

FIGURE 35–22 A, Sensory action potentials. A model for phase cancellation between fast-conducting (F) and slow-conducting (S) sensory fibers. With distal stimulation, two unit discharges summate in phase to produce a sensory action potential twice as large. With proximal stimulation, a delay of the slow fiber causes phase cancellation between the negative peak of the fast fiber and positive peak of the slow fiber, resulting in a 50% reduction in size of the summated response. B, Compound muscle action potentials. Same arrangements as above to show the relationship between fast-conducting (F) and slow-conducting (S) motor fibers. With distal stimulation, two unit discharges representing motor unit potentials summate to produce a muscle action potential twice as large. With proximal stimulation, long-duration motor unit potentials still superimpose nearly in phase despite the same latency shift of the slow motor fiber as the sensory fiber. Thus a physiologic temporal dispersion alters the size of the muscle action potential only minimally, if at all. (From Kimura, J., Machida, M., Ishida, T., et al.: Relation between size of compound sensory or muscle action potentials and length of nerve segment. Neurology 36:647, 1986, with permission.)

action potential but, unlike conduction block, do not by themselves cause muscle weakness. Electromyography reveals little or no evidence of denervation after selective damage of the myelin sheath unless the patient develops secondary axonal degeneration. In partial conduction blocks, the surviving motor unit potentials, normal in amplitude and waveform, discharge at a high rate to com-

pensate for a decreased recruitment in proportion to the number of blocked nerve fibers. Axonal damage, after the first week of injury, reduces distal and proximal responses equally, maintaining a normal ratio because excitability fails at the neuromuscular junction and along the distal nerve segment. Thus the usual criteria for conduction block depend on amplitudes or area ratio of

Nerve Conduction and Needle Electromyography

933

FIGURE 35–23 Simultaneous recordings of compound muscle action potentials from the thenar eminence and sensory nerve action potentials from index and middle fingers after stimulation of the median nerve at palm, wrist, elbow, and axilla. Progressively more proximal series of stimuli elicited nearly the same muscle response but progressively smaller sensory response from the wrist to the axilla, showing a linear relationship to latency. (From Kimura, J., Machida, M., Ishida, T., et al.: Relation between size of compound sensory or muscle action potentials and length of nerve segment. Neurology 36:647, 1986, with permission.)

compound muscle action potentials elicited by proximal versus distal stimulation.8,279,349,370 For example, motor conduction block may be defined as a reduction in amplitude ratio greater than 20% to 50% with less than 15% increase in duration. These definitions, however, suffer from considerable

uncertainty because similar waveform changes may also result from pathologic temporal dispersion and phase cancellation in the absence of conduction block. The combination of clinical and electrophysiologic findings usually documents a motor conduction block better

FIGURE 35–24 A 34-year-old man with selective weakness of foot dorsiflexors and low back pain radiating to the opposite leg. The nerve conduction studies revealed a major conduction block between the two stimulation sites, b and b, at the knee. The weakness abated promptly when the patient refrained from habitual leg crossing. (From Kimura, J.: Electromyography and Nerve Stimulation Techniques: Clinical Applications [in Japanese]. Tokyo, Igaku-Shoin, 1990, with permission.)

934

Nerve Conduction and Electromyography

than the commonly used criteria based solely on waveform analysis.192 In typical situations, stimulation distal to the lesion elicits a large muscle potential associated with a vigorous twitch,190 despite clinical weakness and a paucity of voluntarily activated motor unit potentials.62 If distal stimulation elicits normal or nearly normal M response but fail to evoke F waves, this too suggests a conduction block in the proximal nerve segment.108,181 A sustained period of immobility, however, may render the motor neurons hypoexcitable, making it difficult to evoke a normal number of F waves. An unexpected absence of F wave, therefore, may be seen in some patients with a hysterical paralysis, even if the integrity of the peripheral nerve is preserved. Careful analysis of the waveform of an evoked potential improves the accuracy of electrophysiologic interpretation by elucidating technical problems that account for a majority of unusual findings. For example, a pattern of waveform changes mimicking a conduction block may result from the use of a submaximal stimulus at one point and a supramaximal stimulus at a second site. Unrecognized contribution from anomalous branches such as the MartinGruber anastomosis also may lead to a puzzling alteration in amplitude, as does unintended current spread to a neighboring nerve.183,184 Any of these circumstances gives rise to a confusing set of electrophysiologic findings, sometimes leading to an erroneous conclusion. In addition, dissimilar responses elicited by distal and proximal stimuli usually imply that the two onset latencies measured relate to fibers of different conduction characteristics. This finding, therefore, invalidates the value of calculation of conduction velocities based on the conventional formula. A proximal stimulation may fail despite the use of ordinarily adequate intensities if structural abnormalities render the nerve segment inexcitable. In some cases of multifocal motor neuropathy, for example, failure to maximally excite the involved segment sometimes gives a false impression of sustained conduction block after clinical recovery. This situation calls for the use of near-nerve needle stimulation to counter the increased threshold of the nerve fibers. Alternatively, stimulation at a more proximal, unaffected portion of the nerve segment will help: A muscle response, if elicited by a distant stimulation, proves the propagation of the nerve impulse across the lesion site despite its abnormally elevated threshold for local excitation (Fig. 35–25).173,192

Long and Short Segmental Nerve Conduction Studies Segmental Stimulation in Short Increments The inching technique, or short incremental stimulation along the affected segment,45,187,193 can isolate a lesion more precisely than ordinary conduction studies, which localize the approximate site of involvement. In studying a focal lesion such as compressive neuropathy, inclusion of

the unaffected segments lowers the sensitivity of the study, whereas short segmental stimulation helps localize an abnormality that may otherwise escape detection. With a nerve impulse conducting at a rate of 0.2 ms/cm (50 m/s) except for a 1-cm segment where it drops to 0.4 ms/cm, the conduction time over a 10-cm segment increases only 10%, from 2.0 ms to 2.2 ms, a change well within the normal range of variability. If measured over a 1-cm segment, the same 0.2-ms increase would constitute a 100% change, from 0.2 ms to 0.4 ms, signaling a clear abnormality. In this case, the large per-unit increase in latency more than compensates for the inherent measurement error associated with short incremental stimulation.46,129 Inching study helps not only to detect but also to precisely localize a focal pathology of the median nerve at the wrist,161,187,313,327 the ulnar nerve at the elbow,47,48,175 and the peroneal nerve at the knee.174 For example, stimulation of a normal median nerve with the advance of the cathode in 1-cm increments across the wrist causes latency changes ranging from 0.16 to 0.21 ms/cm from midpalm to distal forearm. In carpal tunnel syndrome, a sharply localized nonlinear latency increase averages 0.8 ms/cm at the compression site. This is nearly always accompanied by an abrupt change in waveform, which confirms a focal abnormality.187 Regardless of the nerve under consideration, waveform analysis can distinguish a pathologic latency change from apparent shifts that might have resulted from inaccurate advances of the stimulating electrodes or inadvertent spread of stimulus current. Even if technical difficulties prevent sequential stimulation near the site of lesion, successive responses above and below the affected zone can characterize local abnormality by forming two parallel lines rather than one (see Fig. 35–25). Together with abrupt waveform changes, this confirms a focal lesion within the short interval flanked by normal segments proximally and distally. Late Responses for Evaluation of Long Pathways Ordinary nerve conduction studies mainly assess the relatively short distal segments of the peripheral nerves. For a diffuse or multisegmental process, the longer the segment under study, the more evident is the conduction delay as a sum of all the segmental abnormalities. If a nerve impulse propagates at a rate of 0.2 ms/cm (50 m/s), for example, a 20% delay amounts only to 0.4 ms for a 10-cm segment. The same abnormality, if measured for a 100-cm segment, causes a 4.0-ms delay, an obvious difference. In addition, the same absolute error constitutes a smaller percentage in evaluating a longer as compared to shorter segment, improving the accuracy of latency and distance determination. Measuring the surface distance for a 10-cm segment, a 1-cm error constitutes 10% of the actual value, resulting in a calculated conduction velocity varying between 50 and 55 m/s. When measuring a 100-cm segment, the same 1-cm

Nerve Conduction and Needle Electromyography

FIGURE 35–25 A, Motor and sensory conduction studies of the left median nerve in a patient with multifocal motor neuropathy. The left diagram illustrates the consecutive magnetic resonance imaging sections in relation to the sites of stimulation at the wrist crease (A1) and at 2-cm increments more proximally. One horizontal division equals 5 ms (motor) or 2 ms (sensory), and one vertical division corresponds to the gain indicated at the end of each trace together with stimulus intensity. Note a complete and selective motor conduction block across the segment between A2 and A3, corresponding to the site of maximal nerve enlargement. (From Kaji, R., Oka, N., Tsuji, S., et al.: Pathological findings at the site of conduction block in multifocal motor neuropathy. Ann. Neurol. 33:152, 1993, with permission.) B, A repeat study in the same patient after return of strength of the median-innervated intrinsic hand muscles. High-intensity stimulation failed to excite the nerve along the affected segments (A5 through A8), mimicking a conduction block. More proximal stimulation at the elbow applied to the presumably normal nerve segments (A9 through A11), however, induced a series of temporally dispersed muscle responses associated with thumb abduction, indicating recovery of conduction. (From Kimura, J.: Facts, fallacies, and fancies of nerve conduction studies: twenty-first annual Edward H. Lambert lecture. Muscle Nerve 20:777, 1997, with permission.)

935

936

Nerve Conduction and Electromyography

error represents only 1% of the actual value, with a range of calculated conduction velocity between 50 and 50.5 m/s. The same argument holds in latency measurement, in which the same absolute error poses a smaller percent difference over a longer as compared to a shorter segment. In summary, the study of a longer path offers better sensitivity and accuracy and, as noted below, improves reproducibility in serial studies. Of a number of neurophysiologic methods that supplement assessment of longer pathways, those of general interest include the F wave and the H reflex. Reproducibility of Various Measures Nerve conduction studies offer a sensitive and objective indicator53 to document serial changes during the clinical course.205 Here the adherence to technical details plays an important role because any deviations from the standards result in inconsistencies of the results, especially in designing a multicenter clinical trial involving many participants of different training backgrounds.371 The assessment of reproducibility exploits two independent indices: the relative intertrial variation (RIV) and the intraclass correlation coefficient (ICC). Of the two, the RIV is derived as a variation of measurements expressed as a percentage of the difference between the two measures over the mean value of the two. Thus RIV (%)  100(V2  V1)/0.5(V1  V2) where V1 and V2 represent the values of the first and the second measurements of the pair, respectively. Ranges of RIV between 10% and 10% indicate an acceptably low variability. Measures having a larger among-subject difference show a greater intraindividual variability as well. Taking this into consideration, calculation of the ICC partially offsets the effects of a large interindividual variability as follows: ICC  s2/( s2  2) where s2 and 2 represent among-subject variance and experimental error. Values exceeding 0.9 indicate a reliable measure, although, as seen from the formula, this may result from a large among-subject variance rather than a small experimental error. Our own experience with a multicenter analysis dealt with a study of the intertrial variability of nerve conduction studies to determine the confidence limits of the variations in preparation for a drug trial.192,209 All measurements were repeated twice at a time interval of 1 to 4 weeks, using a standardized method. In all, 32 centers participated in the study of 132 healthy subjects (63 men) and 65 centers in the evaluation of 172 patients (99 men) with mild diabetic polyneuropathy. Motor nerve conduction studies included

the left median and tibial nerves, measuring amplitude, terminal latency, and minimal F-wave latency and calculating motor conduction velocity and F-wave conduction velocity. Sensory nerve conduction studies included the left median and sural nerves, measuring amplitude and latency of antidromic responses and calculating sensory conduction velocities over the distal segment. Figure 35–26 shows the ICC and the 5th to 95th percentiles of RIVs in both groups, and Figure 35–27 illustrates individual data from healthy subjects. Both the controls and the patient group showed the most variability in amplitude, followed by terminal latency and motor and sensory conduction velocity. The minimal F-wave latency showed the least variability, averaging only 10% for the median nerve and 11% for the tibial nerve in normals and 12% and 14%, respectively, in patients with mild diabetic polyneuropathy. The measures meeting the RIV criterion of less than 10% included F-wave latency and F-wave conduction velocity of the median and tibial nerves and sensory conduction velocity of the median nerve in both healthy subjects and patients. Similarly, the measures meeting the ICC criterion of over 0.9 included F-wave latency of the median and tibial nerves in both groups. In some amplitude measurements, a large among-subject variance concealed a large experimental error, leading to a high ICC despite a considerable variability as evidenced by a large RIV. These included median sensory nerve potential and median and tibial compound muscle action potentials. These findings suggest that a high ICC, indicating a statistical correlation between two measurements,87,384 does not necessarily imply a good reproducibility. Thus, to achieve an optimal comparison, sequential studies must exclude the use of any measurements with a wide RIV regardless of ICC values. Our data indicate that F-wave latencies of the median and tibial nerves meet the criteria for a reliable measure, showing a large ICC ( 0.9) with a small RIV ( 10%), provided, of course, that advanced stage of illness does not preclude adequate recording of F waves. When evaluating single patients against a normal range established in a group of subjects, however, F-wave conduction velocity suits better, minimizing the effect of limb length. Similarly effective is the use of a nomogram plotting the latency against height as an indirect measure of limb length. Clinical Considerations A question often posed in regard to the accuracy, sensitivity, and reliability of latency and velocity measurements relates to the length of the segment under study. Other factors being equal, should one study a shorter or longer segment for better results? The choice depends entirely on the pattern of the conduction change; short segmental approaches better suit a focal lesion than a longer distance, which tends to obscure the abnormality. In contrast, a longer segment reveals diffuse or multisegmental

Nerve Conduction and Needle Electromyography

937

FIGURE 35–26 Reproducibility of various measures in healthy volunteers (A) and patients with diabetic neuropathy (B). All studies were repeated twice at a time interval of 1 to 4 weeks to calculate relative intertrial variations as an index of comparison. (From Kimura, J.: Facts, fallacies, and fancies of nerve conduction studies: twenty-first annual Edward H. Lambert lecture. Muscle Nerve 20:777, 1997; data courtesy of Kohara et al.,209 from a multicenter reliability study sponsored by Fujisawa Pharmaceutical Co., Ltd.)

abnormalities better than a short distance, which tends to reduce sensitivity, increase measurement errors, and lower reproducibility. In summary, short distances magnify focal conduction abnormalities despite increased measurement error, and long distances, though insensitive to focal lesions, accumulate diffuse or multisegmental abnormalities for better sensitivity, accuracy, and reliability.

Needle Electromyography Electromyography tests the motor system as an extension of the physical examination, rather than a laboratory procedure.68,343 It helps to localize the site of a lesion that may cause muscle weakness into upper and lower motor neurons, neuromuscular junction, and muscle (Fig. 35–28). Each

muscle of interest is tested at rest and during degrees of voluntary contraction. The patient’s symptoms and signs guide the optimal selection of specific muscle groups.128,293 The findings in the initially tested muscles determine the subsequent course of exploration. This strategy precludes a routine approach with a predetermined list of muscles for a given condition. Denervated muscle fibers fire at rest independent of volitional neural control, giving rise to spontaneous singlefiber potentials, which constitute one of the most distinct findings in electromyography. The motor unit and its electrical counterpart, the motor unit potential, represent the smallest functional element of volitional contraction. Diseases of the nerve or muscle cause structural or functional disturbances leading to alterations in waveform and discharge patterns. Recorded signals also reflect other biologic and nonbiologic factors, such as the characteristics of field, which serve as the volume conductor. In particular,

938

Nerve Conduction and Electromyography

FIGURE 35–27 Comparison between the first and the second measures of median nerve motor conduction velocity (A) and F-wave latency (B). Individual values plotting the first study on the abscissa and second study on the ordinate show a greater reproducibility of the F-wave latency compared to the motor nerve conduction velocity (cf. Fig. 35–26). (From Kimura, J.: Facts, fallacies, and fancies of nerve conduction studies: twenty-first annual Edward H. Lambert lecture. Muscle Nerve 20:777, 1997; data courtesy of Kohara et al.,209 from a multicenter reliability study sponsored by Fujisawa Pharmaceutical Co., Ltd.)

the spatial relationship between the source and the tip of the needle electrode substantially impacts the size of an action potential. When recorded using an electrode with a small leadoff surface, the amplitude falls off sharply, often to less than 10%, at a distance of 1 mm from the generator.

A patient treated with anticoagulants should not undergo needle study unless appropriate laboratory tests exclude an intolerable bleeding tendency. Our criterion calls for a partial thromboplastin time less than 1.5 times the control value with heparin infusion, and an international normalized ratio of 2.0 or less with warfarin (Coumadin) therapy.

FIGURE 35–28 Schematic view of the motor system with seven anatomic levels. They include upper motor neuron from the cortex to the spinal cord (I and II), lower motor neuron with the anterior horn cell and nerve axon (III and IV), neuromuscular junction (V), and muscle membrane and contractile elements (VI and VII). (Adapted from Netter, F. H.: The Ciba Collection of Medical Illustrations. Vol. 1: Nervous System. Part 1: Anatomy and Physiology. Summit, NJ, Ciba Pharmaceutical Co., Medical Education Division, 1983.) The inset illustrates diagrammatically the four steps of electromyographic examination and normal findings. (Figure from Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

939

Nerve Conduction and Needle Electromyography

Patients with other coagulopathies, such as hemophilia, must have the same precautions.325 Local pressure can usually accomplish hemostasis with a platelet count above 20,000/L for thrombocytopenia.24 In questionable cases, we first test a superficial muscle, where external compression can counter bleeding adequately. This also helps determine the degree of bleeding tendency and the feasibility of further study of deeper muscles. Electromyographic needle explorations often render the muscle unsuitable for a subsequent biopsy examination. Prior needle studies also make it difficult to interpret an elevated serum level of creatine kinase, which may signal certain muscle diseases such as muscular dystrophy and polymyositis. Other conditions associated with this finding include cardiac ischemia, hypothyroidism, and sustained athletic participation. Needle examination by itself should not elevate the enzyme to a misleading level, although, when combined with diurnal variation and prolonged exercise,30,290 it may alter the level considerably even in normal muscles. Electromyographic examination comprises the following four steps: 1. Needle insertion to find the muscle and assess the magnitude of injury potential 2. Observation of spontaneous activity recorded with the needle stationary in a relaxed muscle 3. Study of individual motor unit potentials activated by mild voluntary contraction

4. Evaluation of recruitment during progressively stronger effort and interference patterns at maximum level of contraction Adequate sampling calls for frequent repositioning of the needle electrode in small steps because a needle electrode registers muscle action potentials only from a restricted area. A single puncture site allows exploration in four directions to minimize patient discomfort. In studying larger muscles, however, additional insertions should include proximal, central, and distal portions. Electromyography helps categorize motor dysfunction into upper and lower motor neuron disorders and myogenic lesions, as illustrated schematically in Figure 35–29, which emphasizes typical findings at the risk of oversimplification.

POTENTIALS ASSOCIATED WITH NEEDLE INSERTION Injury Potentials A brief burst of electrical discharges follows movement of a needle electrode inserted into the muscle, lasting on average a few hundred milliseconds. The magnitude of the response primarily depends on the extent and speed of needle movement. Semiquantitative analysis of this activity provides an important measure of muscle excitability, typically reduced in fibroses and exaggerated in denervated or

FIGURE 35–29 Typical findings in lower and upper motor neuron disorders and myogenic lesions. Myotonia shares many features common to myopathy in general, in addition to myotonic discharges triggered by insertion of the needle or with voluntary effort to contract the muscle. Polymyositis shows combined features of myopathy and neuropathy, including (1) prolonged insertional activity; (2) abundant spontaneous discharges; (3) small-amplitude, short-duration, polyphasic motor unit potentials; and (4) early recruitment leading to low-amplitude, full-interference pattern. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

940

Nerve Conduction and Electromyography

inflammatory muscles. A reduced insertional activity may also signal functional inexcitability of muscle fibers, for example, during attacks of familial periodic paralysis. An irritable muscle with instability of the muscle membrane may show an abnormally prolonged insertional activity outlasting the cessation of needle movement, as might be seen in denervation, myotonic disorders, or myositis. A sustained run of positive waves may follow insertional activity during early stages of denervation, 10 days to 2 weeks after nerve injury. These early abnormalities of denervation are seen before the appearance of spontaneous activity. A normal insertional activity may also take the form of a few repetitive positive sharp potentials, but denervated muscles often show reproducible trains lasting several seconds to minutes after cessation of the needle movement (Table 35–19).

End-Plate Activities In normal resting muscles, a needle placed at the end-plate region detects end-plate activities that consist of two components, occurring conjointly or independently: low-amplitude, undulating end-plate noise and high-amplitude intermittent spikes (Fig. 35–30). The patient typically reports a dull pain in association with end-plate activities, which dissipates with slight withdrawal of the needle. The end-plate noise represents extracellularly recorded miniature end-plate potentials, which frequently recur as irregular negative potentials, 10 to 50 V in amplitude and 1 to 2 ms in duration. Over the loudspeaker, it characteristically sounds much like a seashell held to the ear. If this depolarization reaches the threshold, single muscle fibers fire repetitively, giving rise to the end-plate

spikes.43,148 These spikes, 100 to 200 V in amplitude and 3 to 4 ms in duration, fire irregularly at 5 to 50 Hz and typically show a diphasic waveform with initial negativity because the discharges originate at the tip of the recording electrode. Fibrillation potentials, when recorded at the endplate region, also have initial negativity, although they fire more regularly at a slower rate.

Myotonic Discharge Myotonia and its electrical counterpart, myotonic discharge, represents a sustained muscle fiber discharge outlasting the external source of excitation (Table 35–20). The disorders associated with this clinical finding include myotonia congenita, myotonia dystrophica, paramyotonia congenita,379 and hyperkalemic periodic paralysis.38 Clinical myotonia usually results from myotonic discharges triggered by a minimal voluntary contraction or mechanical stimulation with a tap on the muscle belly. Electromyography, however, may uncover myotonic discharges without manifest myotonia in a variety of disorders, such as polymyositis, type II glycogen storage disease with acid maltase deficiency,159 cytoplasmic body myopathy resembling myotonic dystrophy,260 and other conditions with chronic denervation. Depending on the spatial relationship of the discharging muscle fibers to the tip of the needle electrode, myotonic discharge takes one of two forms: (1) negative spikes with a small initial positivity or (2) positive sharp waves, each followed by a low-amplitude negative component of much longer duration (Fig. 35–31). The negative and positive waveforms, like those of denervation, represent recurring

Table 35–19. Origin of Spontaneous and Triggered Discharges Muscle Fiber Insertional positive waves Myotonic discharge Fibrillation potential Positive sharp waves Complex repetitive discharge End-plate noise End-plate spikes

Briefly sustained single muscle fiber discharges triggered by needle movement Repetitive single muscle fiber discharges triggered by needle movement Spontaneous single muscle fiber discharges, negative type Spontaneous single muscle fiber discharges, positive type A group of ephaptically activated spontaneous single muscle fiber discharges Miniature end-plate potentials recorded extracellularly at motor point Single muscle fiber discharges triggered by needle movement at motor point

Lower Motor Neuron Fasciculation potential Myokymic discharge Neuromyotonic discharge Cramp discharges Hemifacial spasm Hemimasticatory spasm

Spontaneous motor unit discharges, involving a single unit, totally or fractionally Clusters of repetitive firing of the same motor unit, usually from demyelination Continuous high-frequency discharges, involving many motor units Briefly sustained high-frequency discharges, involving many motor units Intermittent unilateral contraction of facial muscles, either idiopathic or subsequent to Bell’s palsy Intermittent unilateral contraction of masseter muscle

Upper Motor Neuron Stiff-man syndrome Involuntary movement

Sustained contraction of motor units in many agonistic and antagonistic muscles Tremor, chorea, hemiballismus, athetosis, dystonia, myoclonus, epilepsia partialis continua

From Kimura, J.: Needle electromyography. In Bertorini, R. (ed.): Evaluation and Diagnostic Tests in Neuromuscular Disorders. Woburn, MA, Butterworth-Heinemann, p. 331, 2002, with permission.

Nerve Conduction and Needle Electromyography

941

FIGURE 35–30 End-plate activities recorded from the tibialis anterior in a healthy subject. Two types of potentials shown represent the initially negative, high-amplitude end-plate spikes (a, b, and c) and low-amplitude end-plate noise (g, h, and i). The spikes and end-plate noise usually, though not necessarily, appear together (d, e, and f). (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

single-fiber potentials recorded from an intact and an injured muscle membrane. Thus the negative spikes, resembling fibrillation potentials, tend to occur at the beginning of slight volitional contraction, whereas the positive sharp waves are usually initiated by needle insertion. In both forms of discharge, amplitude waxes and wanes over the range of 10 V to 1 mV, often, though not always, showing inversely varying firing rate within the range of 50 to 100 Hz. This fluctuating pattern of discharge gives the characteristic noise over the loudspeaker, reminiscent of an accelerating or decelerating motorcycle or chain saw.

SPONTANEOUS ACTIVITY Table 35–21 summarizes types of spontaneous activities commonly seen in electromyography. These include fibrillation potentials, positive sharp waves, complex repetitive

discharges, fasciculation potentials, and myokymic discharges. Complex repetitive and myokymic discharges and fasciculation potentials cause isolated or briefly sustained visible muscle twitches over a localized area, whereas fibrillation potentials and positive sharp waves do not. A spontaneously activated single muscle fiber produces fibrillation potentials and positive sharp waves.39,67,68,82,86,213,214 Rapid firing of a cluster of muscle fibers in sequence gives rise to complex repetitive discharges. These discharges are driven ephaptically at a point of lateral contact between neighboring units.93,346 One fiber serves as a pacemaker, regulating the frequency and pattern of discharge by the rate of its rhythmic depolarization and circus movements of currents among muscle fibers.167 Fasciculation potentials result from isolated spontaneous discharges of a motor unit, and myokymic discharges from repetitive firing of a motor unit, as the name grouped fasciculation potentials indicates. These discharges are associated

942

Nerve Conduction and Electromyography

Table 35–20. Disorders with Myotonic Discharges With Clinical Myotonia Myotonia dystrophica Myotonia congenita Paramyotonia congenita Hyperkalemic periodic paralysis Proximal myotonic myopathy Without Clinical Myotonia Myositis Acid maltase deficiency Cytoplasmic body myopathy Hyperthyroidism Hypothyroidism Familial granulovacuolar lobular myopathy Malignant hyperpyrexia Multicentric reticulohistiocytosis Myopathies induced by Glycyrrhizin Hypocholesterolemic agent Colchicine From Kimura, J.: Needle electromyography. In Bertorini, R. (ed.): Evaluation and Diagnostic Tests in Neuromuscular Disorders. Woburn, MA, Butterworth-Heinemann, p. 331, 2002, with permission.

with local muscle twitches, called fasciculation, and sustained segmental contractions seen in cramp syndromes, called myokymia. Similarly, continuous motor unit discharges give rise to generalized muscle spasms or the syndrome of neuromyotonia, representing peripheral nerve hyperexcitability. In contrast, involuntary muscle contractions seen in patients with the stiff-man syndrome result from spontaneous discharges originating in the central nervous system. Semiquantitating each of these spontaneous activities can provide a numeric grading as follows: 1—Rare spontaneous potentials recordable in one or two sites only after some search. This category includes insertional positive sharp waves or abnormally sustaining injury potentials elicited after moving the needle electrode. 2—Occasional spontaneous potentials registered in more than two different sites. 3—Frequent spontaneous potentials recordable regardless of the position of the needle electrode. 4—Abundant spontaneous potentials nearly filling the screen of the oscilloscope.

Fibrillation Potentials Fibrillation potentials have diphasic or triphasic waveforms with initial positivity (Fig. 35–32) ranging from 1 to 5 ms in duration and from 20 to 500 V in amplitude, when recorded with a concentric needle electrode.43 Fibrillation potentials remain the same in shape from

FIGURE 35–31 Myotonic discharges from the right anterior tibialis in a 39-year-old man with myotonic dystrophy. The tracings show two types of discharges: trains of positive sharp waves (a, b, and c) and negative spikes with initial positivity (d, e, and f). The discharges in a and d reveal a waxing and waning quality. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

the first to the last discharges of a train, and have the same shape and amplitude as voluntarily activated singlefiber potentials when recorded with single-fiber electromyography.342 All these findings support the view that fibrillation potentials represent single muscle fibers or the smallest unit recordable by the needle electrode. A crisp clicking noise that fibrillation potentials produce over the loudspeaker resembles the sound caused by wrinkling tissue paper. Spontaneous oscillations in the membrane potential probably regulate the firing pattern in the range of 1 to 30 Hz, with an average rate of 13 Hz.354 Occasional fibrillation potentials may discharge irregularly in the range of 0.1 to 25 Hz.37,256 In profound denervation, discharges from more than one fiber may give a false impression of a very irregular firing pattern.

Positive Sharp Waves Positive sharp waves also result from single-fiber activation despite their sawtooth appearance, much lower in amplitude and longer in duration compared to fibrillation potentials

Nerve Conduction and Needle Electromyography

Table 35–21. Common Types of Spontaneous Discharges Fibrillation Potentials and Positive Sharp Waves Neuropathic condition Muscular dystrophy Myositis Complex Repetitive Discharges Motor neuron disease Radiculopathy Chronic polyneuropathy Polymyositis Muscular dystrophy Myxedema Schwartz-Jampel syndrome Fasciculation Potentials Motor neuron disease Radiculopathy Entrapment neuropathy Muscular pain–fasciculation syndrome Healthy subjects Myokymic Discharges Guillain-Barré syndrome Radiation plexopathy Spinal stenosis Nerve root compression Bell’s palsy Multiple sclerosis Syringobulbia From Kimura, J.: Needle electromyography. In Bertorini, R. (ed.): Evaluation and Diagnostic Tests in Neuromuscular Disorders. Woburn, MA, Butterworth-Heinemann, p. 331, 2002, with permission.

(see Fig. 35–32). In general, the absence of a negative spike implies recording near the damaged part of the muscle fiber. If the inserted needle penetrates the membrane, sustained standing depolarization precludes the generation of a negative spike at this point. Therefore, a propagating action potential only approaches the site of injury, giving rise to a sharp positive discharge, followed by low-amplitude negative deflection in the absence of a negative spike. As discussed earlier, positive sharp waves may form part of myotonic discharges, usually triggered by insertion of the needle and less often by mild voluntary contraction. Although both represent single-fiber discharges, denervation potentials occur spontaneously, whereas myotonic discharges appear in response to mechanical stimuli or following voluntary contraction. This distinction, however, may often blur, because needle movements also enhance spontaneous discharges associated with denervation.

Single-Fiber Discharge and Denervation Spontaneous single-fiber activity, in the appropriate clinical setting, usually signals disorders of the lower motor neuron as one of the most useful signs of abnormality in

943

clinical electromyography. Its presence implies denervation, although it is otherwise nonspecific, being seen in such diverse diseases as anterior horn cell diseases, radiculopathies, plexopathies, and axonal polyneuropathies. Its absence, however, proves little during the latent period of 2 to 3 weeks after nerve injury. The distribution of spontaneous potentials helps localize lesions of the spinal cord, root, plexus, and peripheral nerve. Certain myopathic processes, such as muscular dystrophy,43 dermatomyositis, and polymyositis, also show spontaneous discharges, possibly because of muscle necrosis75 or focal degeneration,334 which separates a part of the muscle fiber from the end-plate region, thus denervating that portion. Alternatively, spontaneous activity may also result in polymyositis from increased membrane irritability or inflammation of intramuscular nerve terminals.307 Other conditions less consistently associated with fibrillation potentials include diseases of the neuromuscular junction,123 facioscapulohumeral dystrophy, limb-girdle dystrophy, oculopharyngeal dystrophy,149 myotubular (centronuclear) myopathy,339 and trichinosis.378

Complex Repetitive Discharges Complex repetitive discharges result from a group of muscle fibers that fire repetitively in the same sequence, producing the sound of a machine gun over the loudspeaker. The entire epoch, ranging from 50 V to 1 mV in amplitude and up to 50 to 100 ms in duration (Fig. 35–33), repeats itself at slow or fast rates, usually in the range of 5 to 100 Hz. These discharges typically have a sudden onset, a constant rate of firing for a short period, and an abrupt cessation. The polyphasic waveform remains identical from one burst to the next until a sudden shift to a new pattern of discharge intervenes. The unique repetitive features once prompted the use of the now discarded terms bizarre high frequency discharges and pseudomyotonia. The uniform firing pattern showing a design-like waveform makes these discharges distinct from myokymia, neuromyotonia, and cramp syndromes despite their superficial resemblance. This discharge usually indicates a wide variety of chronic denervating conditions. These include motor neuron disease, radiculopathy, chronic polyneuropathy, myxedema, and the Schwartz-Jampel syndrome sometimes associated with neurogenic muscle hypertrophy,319 as well as certain myopathies, such as muscular dystrophy and polymyositis. An overall analysis in a large series93 revealed the highest prevalence in Duchenne type muscular dystrophy, spinal muscular atrophy, and Charcot-Marie-Tooth disease. In women with urinary retention, the striated muscle of the urethral sphincter may show profuse activity of this type.113,114 Some apparently healthy subjects may have incidental findings of complex repetitive discharges, probably indicating the presence of clinically silent foci of an irritative process that tends to involve deeper muscles in general and the iliopsoas in particular.

944

Nerve Conduction and Electromyography

FIGURE 35–32 A, Spontaneous single-fiber discharges recorded from the denervated tibialis anterior in a 67-year-old man with acute onset of footdrop. Note gradual alteration of the waveform from triphasic spike with major negativity to paired positive potentials and finally to a single positive sharp wave over the time course of some 8 seconds without movement of the needle. This fortuitous recording provides direct evidence that the same single-fiber discharge can be recorded either as fibrillation potentials or positive sharp waves. Long-duration positive deflections seen in c, f, and g represent a pulse artifact. (From Kimura, J.: Electromyography and Nerve Stimulation Techniques: Clinical Applications [in Japanese]. Tokyo, Igaku-Shoin, 1990, with permission.) B, Spontaneous single-fiber activity of the anterior tibialis in a 68-year-old woman with amyotrophic lateral sclerosis. The tracings show two types of discharges: positive sharp waves (a, b, and c) and fibrillation potentials (d, e, and f). C, Spontaneous single-fiber activity of the paraspinal muscle in a 40-year-old man with radiculopathy, consisting of positive sharp waves (a, b, and c) and fibrillation potentials (d, e, and f). D, Spontaneous single-fiber activity of the deltoid (a, b, and c) and tibialis anterior (d, e, and f) in a 9-year-old boy with a 6-week history of dermatomyositis, with two types of discharges: positive sharp waves (a, b, and c) and fibrillation potentials (d, e, and f). E, Spontaneous single-fiber activity of the tibialis anterior in a 7-year-old boy with Duchenne type muscular dystrophy, showing positive sharp waves (a, b, and c) and fibrillation potentials (d, e, and f). (B–E from Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.) Figure continued on opposite page

Nerve Conduction and Needle Electromyography

945

FIGURE 35–32 Continued

Fasciculation Potentials and Myokymic Discharges Fasciculation potentials, or spontaneous discharges of a single motor unit either as a whole or possibly in part, usually, but not necessarily, induce visible twitches depending on the depth of discharge. In some instances, therefore, electromyography reveals fasciculation potentials in the absence of a clinical counterpart. Normal voluntary motor unit potentials maintain constant shape if the needle remains stationary, whereas fasciculation potentials undergo changes in amplitude and waveform from time to time. Discharge characteristics help discriminate semirhythmic voluntary units firing at a rate ranging from 5 to 10 Hz and irregular fasciculation potentials showing a much slower rate. In questionable cases, attempts to volitionally control the discharge pattern by mildly contracting agonistic or antagonistic muscles alters the firing rate of voluntary units but not fasciculation potentials. Although the neural discharge responsible for fasciculation potentials may originate in the spinal cord or anywhere along the length of the peripheral nerve,380 the existing evidence favors a very distal site of origin at or near the motor terminals.227 Fasciculation potentials are seen not only in diseases of anterior horn cells, but also in other neuropathic conditions such as radiculopathy, entrapment neuropathy, and the muscular pain–fasciculation syndrome,160 as well as in healthy subjects. Patients with cervical spondylotic myelopathy may have fasciculation potentials in the lower limbs. Possible underlying

mechanisms include loss of inhibition, vascular insufficiency, cord traction, and denervation, although none of these hypotheses has anatomic or physiologic evidence. Spontaneous discharges abate after cervical decompression.172 Fasciculation potentials abound in some metabolic derangements, such as tetany, thyrotoxicosis, and overdoses of anticholinesterase medication.67 Multiple units may discharge simultaneously in amyotrophic lateral sclerosis and progressive spinal muscular atrophy, but also in other degenerative diseases of the anterior horn cells, such as poliomyelitis and syringomyelia. Synchronous spontaneous discharges may simultaneously involve muscles supplied by different nerves or occur in homologous muscles on opposite sides. This pattern may possibly suggest an intraspinal mechanism, or a reflex origin via spindle afferent triggered by the arterial pulse.317 Healthy subjects may have either single or grouped spontaneous discharges in otherwise normal muscles,258 sometimes, but not always, associated with cramps. A number of investigators have sought to differentiate these benign fasciculations without ominous implication from those of progressive motor neuron disease. No single method has emerged based on waveform characteristics such as amplitude, duration, and number of phases.364 The frequency criteria, however, may possibly serve to separate the two categories: irregular firing at an average interval of 3.5 seconds in motor neuron disease versus 0.8 seconds in asymptomatic individuals.71,364 Also, fasciculation potentials characteristically arise proximally early and occur distally in the later stages of the disease in amyotrophic lateral

946

Nerve Conduction and Electromyography

Continuous Muscle Fiber Activity

FIGURE 35–33 Complex repetitive discharges of the left quadriceps in a 58-year-old man with herniated lumbar disc. The tracings show two types of discharges: trains of single- or double-peaked negative spikes (a, b, and c) and complex positive sharp waves (d, e, and f). In f, each sweep, triggered by a recurring motor unit potential, shows remarkable reproducibility of the waveform within a given train. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

sclerosis.71 In conclusion, fasciculation potentials by themselves cannot provide an absolute proof of abnormality, unless accompanied by either fibrillation potentials or positive sharp waves, indicating disease of the lower motor neuron. In contrast to isolated discharges, more complex bursts of repetitive discharges of one motor unit cause vermicular movements of the skin, called myokymia.61 Repetitive firing of the same motor units typically comprises 2 to 10 spikes discharging at 30 to 40 Hz in each burst (Fig. 35–34), which repeats at fairly regular intervals of 0.1 to 10 seconds. Myokymic discharges often involve facial muscles in patients with brainstem glioma or multiple sclerosis (see Table 35–21). When present in the limb muscles, they usually favor the diagnosis of certain chronic demyelinative neuropathic processes, such as multifocal motor neuropathy, GuillainBarré syndrome,249 and radiation plexopathies.3,6,70 Hyperventilation enhances myokymic bursts generated ectopically in a demyelinated segment of motor fibers because induced hypocalcemia amplifies axonal excitability.26

A heterogeneous group of central or peripheral disorders shows prolonged spontaneous motor unit activity, or continuous muscle fiber activity.89 In stiff-man syndrome, a rare, well-recognized but poorly characterized entity, sustained involuntary discharges of central origin produce a continuous interference pattern in the agonists as well as antagonists simultaneously. These motor unit potentials abate with peripheral nerve or neuromuscular block, after spinal or generalized anesthesia, or during sleep. The administration of diazepam, but not phenytoin or carbamazepine, also abolishes or attenuates the activity. Neuromyotonia implies continuous muscle fiber activity of peripheral origin126 seen in various, probably related, disease entities such as Isaacs’ syndrome, quantal squander, generalized myokymia, pseudomyotonia, and normocalcemic tetany.166,271 In these syndromes, clinical and electrophysiologic presentations vary, although they all show the feature of sustained spontaneous discharges generated at various sites of axons, from proximal segments to the intramuscular terminals.243,316,360 Involuntary motor activity continues during sleep and after general or spinal anesthesia but abates totally with neuromuscular blocking, confirming its peripheral origin. Procaine injection of the peripheral nerve will abolish the abnormal activity only if the discharges originate in the more proximal segment. Undulating movements of the overlying skin and a delay of relaxation after muscle contraction result from sustained firing of motor units characteristically discharging at frequencies as high as 300 Hz. The firing motor unit potentials produce a typical “pinging” sound over the loudspeaker. An increasing number of single muscle fibers fail to follow the high rate of repetitive firing pattern, leading to a decline in amplitude slowly or rapidly. Ischemia and electrical nerve stimulation, but not voluntary contraction, accentuate the high-frequency discharge. Phenytoin or carbamazepine effectively reduces involuntary movements in most patients with neuromyotonia.

Cramp Cramp may commonly develop in normal subjects but also occasionally as a sign of abnormality in patients with neuromuscular disorders. Following severe cramps, the pain may persist for days. This briefly sustained involuntary muscle contraction probably results from spontaneous impulses that originate in the peripheral nerve. Although the exact underlying mechanism remains unknown, mechanical excitation of motor nerve terminals during muscle shortening may serve as a trigger.226,227,276 Spinal or general anesthesia has no effects, but peripheral nerve block often abolishes the activity. The cramp discharges comprise repetitive discharges of normal motor unit potentials at a high rate, in the range of 200 to 300 Hz. They begin with single potentials or

Nerve Conduction and Needle Electromyography

947

FIGURE 35–34 A, Myokymic discharges in a 21-year-old woman with multiple sclerosis. The patient had visible undulating movement of the facial muscles on the right associated with characteristic bursts of spontaneous activity recorded from the orbicularis oris (a, b, c, and d) as well as orbicularis oculi (e, f, g, and h). In d, each sweep, triggered by a recurring spontaneous potential, shows a repetitive but not exactly time-locked pattern of the waveform. B, Myokymic discharges in a 57-year-old man with a 2-week history of Guillain-Barré syndrome and a nearly complete peripheral facial palsy. Despite the absence of visible undulating movement, rhythmically recurring spontaneous discharges appeared in the upper (a, b, and c) and lower (d, e, and f) portions of the left orbicularis oris. In c and f, each sweep triggered by a recurring spontaneous potential shows the repetitive pattern. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

doublets, and gradually spread to involve other areas of a muscle. The discharges may wax and wane for several minutes, involving several different sites that may be activated simultaneously or sequentially, and then abate spontaneously.

MOTOR UNIT ACTION POTENTIAL A group of muscle fibers innervated by a single anterior horn cell constitutes a motor unit, with anatomic and physiologic properties determined by the innervation ratio, fiber density, propagation velocity, and integrity of neuromuscular transmission. Motor unit potentials vary not only from one muscle group to another but also with age for a given muscle. According to principal components analysis, three elements determine 90% of the variance of the data

set: changes in the size of the motor units, variations in the arrival time at the recording electrode, and loss of muscle fibers within the motor unit territory.264 Other physiologic factors that influence the shape of motor unit potentials include the resistance and capacitance of the intervening tissue and intramuscular temperature.18,73 In addition, the spatial relationships between the needle and individual muscle fibers critically determine the waveform.33 The measures that define a motor unit potential include amplitude, rise time, duration, phases, stability, and territory (Fig. 35–35). A wide variety of neuromuscular disorders cause characteristic combinations of abnormalities, which help distinguish primary muscle diseases from disorders of the neuromuscular junction and lower motor neurons (Table 35–22). A random loss of individual fibers seen in myopathies reduces the spike duration and amplitude of motor unit potentials.42 In contrast, an increased

948

Nerve Conduction and Electromyography

Turn

Baselineto-peak amplitude

Phase

Late component

Peakto-peak amplitude

weakness in the diseases of nerve and muscle.93,146 This, taken together with abnormalities of insertional and spontaneous activities, and recruitment pattern, helps delineate a characteristic abnormality. In addition, serial assessment provides a good means to monitor the disease process based on the established correlation between physiologic and histologic alterations of the motor unit.341

Amplitude and Area 200μV 500μs Risetime Duration

FIGURE 35–35 Profile of motor unit potential determined by (1) peak-to-peak or baseline-to-peak amplitude; (2) rise time, or the time interval from the onset of a change to its peak; (3) turn, or point of change in direction; (4) phase, or portion of a waveform between departure from and the return to the baseline; and (5) late component, or satellite potential, separated from the main motor unit potential. (Courtesy of David Walker, M.S.E.E.; from Kimura, J.: Needle electromyography. In Bertorini, R. [ed.]: Evaluation and Diagnostic Tests in Neuromuscular Disorders. Woburn, MA, Butterworth-Heinemann, p. 331, 2002, with permission.)

fiber density with sprouting after a loss of axons results in a larger potential in neuropathies or anterior horn cell diseases. Thus changes in the size of the motor unit potential show characteristic dichotomy in the classification of

Table 35–22. Types of Motor Unit Potentials Brief Duration, Small Amplitude, Early Recruitment Muscular dystrophy Congenital myopathy Metabolic myopathy Endocrine myopathy Myositis Myasthenia gravis Myasthenic syndromes Long Duration, Large Amplitude, Late Recruitment Motor neuron disease Radiculopathies Plexopathies Polyneuropathies Entrapment syndromes From Kimura, J.: Needle electromyography. In Bertorini, R. (ed.): Evaluation and Diagnostic Tests in Neuromuscular Disorders. Woburn, MA, Butterworth-Heinemann, p. 331, 2002, with permission.

Of all the individual muscle fibers of a motor unit that fire in near-synchrony, only a limited number located within the recording radius of the electrode contribute to the amplitude of a motor unit potential. Thus the position of the needle electrode relative to the discharging unit alters the recorded amplitude greatly. With the use of a concentric needle, for example, the amplitude falls off to less than 50% at a distance of 200 to 300 m from the source and to less than 1% a few millimeters away.92,139 To avoid a distant unit, it is recommended to select a motor unit potential with a short rise time of 500 s or less, which guarantees its proximity to the recording surface (see Fig. 35–35). Histologic studies reveal that only a few single muscle fibers lie within the recording radius from the tip of the needle. Thus the high-voltage spike of the motor unit potential results from fewer than 5 to 10 muscle fibers lying within a 500-m radius of the electrode tip.342,355 Computer simulation indicates that the proximity of the closest muscle fiber to the recording electrode determines the amplitude.85,266,347 Therefore, the muscle fibers lying closer together within the recording surface of the needle give rise to a higher amplitude. These findings indicate that, in general, the measure of amplitude aids in determining the muscle fiber density, not the motor unit territory. Many different profiles of motor unit potentials appear from the same motor unit with repositioning of the recording electrode. Motor unit potentials normally vary in amplitude from several hundred microvolts to a few millivolts. Recording with a monopolar needle tends to enhance the average amplitude of motor unit potentials compared to a concentric needle, which has a smaller recording radius.207 The “thickness” of the potential or ratio between area and amplitude varies much less with changes in electrode position.262 The combination of amplitude and area/amplitude ratio further improves discrimination, detecting around 70% of neurogenic changes compared to only 15% to 30% by duration criteria alone.337

Rise Time The rise time, or time lag from the initial positive peak to the subsequent negative peak, serves as a measure of proximity of the signal to the recording tip of the electrode. A greater rise time indicates a distant unit because

Nerve Conduction and Needle Electromyography

949

the resistance and capacitance of the intervening tissue act as a high-frequency filter. A distant motor unit also has a dull sound, as an audio counterpart of a long rise time. Repositioning the electrode closer to the source shortens the rise time, adding a high-pitched, crisp sound over the loudspeaker. In general, a rise time less than 500 s ensures that the electrode tip lies within the motor unit territory,164 although some argue for less restrictive criteria.16 Most electromyographers depend on a sharp, crisp sound over the loudspeaker as an important clue for the proximity of the discharging unit to the electrode. The measurement of the rise time determines the suitability of the recorded potential for inclusion in quantitative assessment.

loss of fibers (Fig. 35–37). These polyphasic potentials are also found in a healthy muscle, constituting between 5% and 15% of the total population, when recorded with a concentric needle electrode. The percentage of polyphasic activities increases with the use of a monopolar needle, although the exact incidence is not known. Some action potentials show several “turns,” defined as directional changes, without crossing the baseline. These serrated or, less appropriately, complex or pseudopolyphasic potentials have the same implication as polyphasic units, signaling desynchronization among discharging muscle fibers. In one study,389 these irregular potentials appeared more commonly in acute processes.

Duration

Abnormalities of Motor Unit Potentials

The duration of a motor unit potential, measured from the initial takeoff to the return to the baseline (see Fig. 35–35), reflects the length, conduction velocity, and degree of synchrony of many individual muscle fibers of the discharging motor unit.83 Although only a very small number of muscle fibers near the electrode determine the spike amplitude, a greater number of muscle fibers contribute to the duration of a motor unit potential. A concentric needle has a recording surface with an uptake area extending about 2.5 mm from the core.266,344 This allows distant units, not contributing to the amplitude of the negative spike, to add to the time of the initial and terminal positivity of the motor unit, increasing its duration. Thus the motor unit duration serves as a measure of a larger part of the muscle fiber population than the amplitude, although it still falls short of covering the entire motor unit territory, which measures 1 to 2 cm. In contrast to the amplitude, the duration remains relatively unchanged with a slight shift or rotation of the needle,263 normally varying from 5 to 15 ms, depending on the age of the subject (Table 35–23).32 With the use of a wide-open amplifier bandpass combined with enhanced signal-to-noise ratio, the duration increases substantially, approaching 30 ms. Under this circumstance, either a single-fiber electrode or macroelectrode seems to register the total time of a single action potential from end-plate zone to musculotendinous junction as the overall duration of the motor unit action potential.84

The contrasting abnormalities of the motor unit potential in myopathies versus lower motor neuron disorders occur as a common feature in a number of diseases in the respective categories. Thus such abnormalities per se cannot confirm a specific diagnosis, although they provide supportive evidence for the clinical impression. A majority of motor unit potentials have two or three phases in normal muscles. The number of polyphasic units having four or more phases increases in myopathy, neuropathy, and motor neuron disease (see Fig. 35–37). In these motor units, abnormal temporal dispersion of muscle fiber potentials probably results from unequal conduction time along either the nerve terminal or muscle fibers. Satellite potentials or extrapotentials clearly separated from the main unit have the same implications as polyphasic activities,69,362 occurring five times more often in patients with neuropathy and myopathy as compared to normal subjects.106 Motor unit potentials, if recorded at all during neurapraxia or an acute stage of axonotmesis, have normal waveforms, which result from unaffected axons. Motor units activated by voluntary effort normally fire semirhythmically, showing nearly identical configuration for successive discharges. Fatigue causes irregularity and reduction in the firing rate of a motor unit, without altering its waveform. With defective neuromuscular transmission, a repetitively firing motor unit fluctuates in amplitude and waveform from intermittent blocking and increased jitter of the constituent single-fiber potentials.287 Waveform variability of the motor unit potential plays an important role in evaluating deficiency of neuromuscular transmission, especially in proximal muscles not readily accessible to repetitive nerve stimulation. Instability of a motor unit potential, termed jiggle,345 however, accompanies a large group of disorders, which include myasthenia gravis, myasthenic syndrome, botulism, motor neuron disease, poliomyelitis, and syringomyelia as well as early stages of reinnervation. Motor unit potentials in myotonia congenita typically show a progressive decline in amplitude with successive discharges, reflecting changes that

Phases The portion of a waveform between the departure from and return to the baseline is called a phase. The number of phases, determined by counting negative and positive peaks to and from the baseline, equals the number of baseline crossings plus one. In contrast to normal motor unit potentials, with four or fewer phases (Fig. 35–36), desynchronized units show polyphasic potentials with more than four phases, reflecting either fiber size variability or random

8.8 9.0 9.2 9.4 9.6 9.9 10.1 10.4 10.7 11.4 12.2 13.0 13.4 13.8 14.3 14.8 15.1 15.3 15.5 15.7

7.1 7.3 7.5 7.7 7.8 8.0 8.2 8.5 8.7 9.2 9.9 10.6 10.9 11.2 11.6 12.0 12.3 12.5 12.6 12.8

Biceps Deltoideus Brachii 8.1 8.3 8.5 8.6 8.7 9.0 9.2 9.6 9.9 10.4 11.2 12.0 12.4 12.7 13.2 13.6 13.9 14.1 14.3 14.4

Triceps Brachii 6.6 6.8 6.9 7.1 7.2 7.4 7.5 7.8 8.1 8.5 9.2 9.8 10.1 10.3 10.7 11.1 11.3 11.5 11.6 11.8

Extensor Digitorum Communis 7.9 8.1 8.3 8.5 8.6 8.9 9.1 9.4 9.7 10.2 11.0 11.7 12.1 12.5 12.9 13.3 13.6 13.9 14.0 14.2

Opponens Pollicis; Interosseus 9.2 9.5 9.7 9.9 10.0 10.3 10.5 10.9 11.2 11.9 12.8 13.6 14.1 14.5 15.0 15.5 15.8 16.1 16.3 16.5

8.0 8.2 8.4 8.6 8.7 9.0 9.2 9.5 9.8 10.3 11.1 11.8 12.2 12.5 13.0 13.4 13.7 14.0 14.1 14.3

7.1 7.3 7.5 7.7 7.8 8.0 8.2 8.5 8.7 9.2 9.9 10.6 10.9 11.2 11.6 12.0 12.3 12.5 12.6 12.8

8.9 9.2 9.4 9.6 9.7 10.0 10.2 10.5 10.8 11.5 12.3 13.2 13.6 13.9 14.4 14.9 15.2 15.5 15.7 15.9

6.5 6.7 6.8 6.9 7.0 7.2 7.4 7.6 7.8 8.3 8.9 9.5 9.8 10.1 10.5 10.8 11.0 11.2 11.4 11.5

Biceps Abductor Femoris; Tibialis Peroneus Digiti Quinti Quadriceps Gastrocnemius Anterior Longus

Leg Muscles

7.0 7.2 7.4 7.6 7.7 7.9 8.1 8.4 8.6 9.1 9.8 10.5 10.8 11.1 11.5 11.9 12.2 12.4 12.5 12.7

4.2 4.3 4.4 4.5 4.6 4.7 4.8 5.0 5.1 5.4 5.8 6.2 6.4 6.6 6.8 7.0 7.1 7.3 7.4 7.5

Orbicularis Oris Extensor Superior; Digitorum Triangularis; Brevis Frontalis

Facial Muscles

*The values given are mean values from different subjects without evidence of neuromuscular disease. The standard deviation of each value is 15% (20 potentials for each muscle). Therefore, deviations up to 20% are considered within the normal range when comparing measurements in a given muscle with the values of the table. From Buchthal, F.: An Introduction to Electromyography. Copenhagen, Scandinavian University Books, 1957, with permission.

0 3 5 8 10 13 15 18 20 25 30 35 40 45 50 55 60 65 70 75

Age in Years

Arm Muscles

Table 35–23. Mean Action Potential Duration (in Milliseconds) in Various Muscles at Different Ages (Concentric Electrodes)*

Nerve Conduction and Needle Electromyography

951

FIGURE 35–36 A, Normal motor unit potentials from minimally contracted biceps in a 40-year-old healthy man (a, b, and c) and maximally contracted tibialis anterior in a 31-year-old woman with hysterical weakness (d, e, and f). In both, low firing frequency indicates weak voluntary effort. B, Normal variations of motor unit potentials from the same motor unit in the biceps of the same healthy subject shown in a through c above. Tracings a through h represent eight slightly different sites of recording with the patient maintaining isolated discharges of a single motor unit. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

affect constituent single muscle fiber action potentials during continued contraction.

Lower Motor Neuron Versus Myopathic Disorders Various disorders of the lower motor neuron that share the same abnormalities in electromyography include motor neuron disease, poliomyelitis, syringomyelia, and diseases of the peripheral nerve.94,337 In these disorders, denervated muscle fibers undergo reinnervation, which leads to an increase in size of motor unit potentials (Fig. 35–38). The small recording radius of a monopolar or concentric needle is not suited to identify the territory of motor unit potentials, but a macro study can delineate the size of discharging units. Sprouting axon terminals usually remain within their own motor unit territory, without reaching out to the denervated muscle fibers located outside this boundary. Thus, after reinnervation, the number of muscle fibers increases with incorporation of denervated fibers, but the territory of the surviving motor unit remains the same. More specifically, a larger amplitude reflects a greater

muscle fiber density, whereas an increased duration results from a greater range in length and conduction time of regenerating axon terminals, as predicted by computer simulation.233 With abnormal synchronization at the cord level or ephaptic activation at the root level or near the axon terminal,318 two or more motor units may discharge simultaneously showing a complex waveform.55 Primary myopathic disorders, characterized by reduction in amplitude and duration of the motor unit potential (Fig. 35–39), include muscular dystrophy, congenital myopathies, periodic paralysis, myositis, and disorders of neuromuscular transmission such as myasthenia gravis and myasthenic syndromes. All of these entities have in common a random loss of functional muscle fibers from each motor unit as the result of muscle degeneration, inflammation, metabolic changes, or failure of neuromuscular activation. A lower fiber density from a decreased number of muscle fibers, in turn, causes reduction in amplitude and duration of motor unit potentials. In extreme cases, only a single muscle fiber represents a motor unit potential, showing voluntary activities indistinguishable from a fibrillation potential. Thus a high-frequency sound produced

952

Nerve Conduction and Electromyography

or increased duration associated with myopathic changes probably reflect an abnormally increased range of muscle fiber diameter.264 Additionally, a greater fiber density from muscle fiber regeneration results in much larger amplitude than might ordinarily be expected for a myopathy. In analyzing abnormalities of motor unit potentials, therefore, an oversimplified clinical dichotomy between myopathy and neuropathy does not necessarily hold.96,376 Nonetheless, it is reassuring that electromyography and histochemical findings from muscle biopsies have an overall concordance of 90% or greater.20,35,36,146

DISCHARGE PATTERN OF MOTOR UNITS Recruitment

FIGURE 35–37 Polyphasic motor unit potentials from the anterior tibialis in a 52-year-old man with amyotrophic lateral sclerosis. Temporal variability of repetitive discharges in waveform suggests intermittent blocking of some axon terminals. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

by short spikes, 1 to 2 ms in duration, resembles that of spontaneously discharging fibrillation potentials over the loudspeaker. Reversible changes not seen in inherited disorders of muscle may characterize metabolic or toxic myopathies.34 Motor unit potentials usually remain normal in mild metabolic and endocrine myopathies, with little or no alteration in duration or amplitude. Despite ordinarily contrasting changes in the waveform of motor unit potentials between myopathies and lower motor neuron disorders,35,36 some cases with equivocal distinction challenge even experienced clinicians.95 Distal neuropathies have sick axon terminals with a random loss of muscle fibers within a motor unit. Similarly, recently reinnervated immature motor units with a few muscle fibers give rise to low-amplitude and short-duration motor unit potentials. Thus, in both instances of a neuropathic process, motor units show changes classically regarded as consistent with a myopathy.261 Conversely, motor unit potentials with a long duration may erroneously suggest neuropathic changes in myopathies with regenerating muscle fibers.75,222,302 Some of these potentials may be so delayed from the main unit as to cause the appearance of a satellite or late component. Complex potentials with normal

In a healthy subject, a mild voluntary contraction initially activates one or two motor units with small, type I muscle fibers at a rate of several impulses per second (see Fig. 35–36). Additional units are then recruited in a fixed order according to the size principle.97,150,326 During mild constant contraction, motor units typically discharge semirhythmically, slowly increasing, then decreasing the interspike intervals. Greater levels of muscle contraction change the firing rate to grade the muscle force. Thus increase in muscle force induces two separate but related changes in the pattern of motor unit discharge: (1) recruitment of previously inactive units and (2) higher firing rate of already active units. Which of the two plays a greater role is not known, but both mechanisms operate simultaneously. In a normal recruitment pattern, the number of discharging motor units increases appropriately for the muscle force generated by the effort. In neuropathy, a loss of functional motor units results in late or decreased recruitment associated with rapid firing of discharging units to compensate for the reduced number (see Figs. 35–37 and 35–38). In myopathy, a random loss of muscle fibers makes each motor unit less efficient, necessitating an early or increased recruitment to compensate for a weak force generated by each unit (see Fig. 35–39). Thus recruitment is said to be decreased or increased if fewer or a greater number of units discharge than the number expected for a given force being exerted. In parkinsonism and other disorders of basal ganglia, motor units often fire in groups at or near the rates of tremor.72 Upper motor neuron lesions such as spinal cord injury also alter motor unit forces and recruitment patterns.356 Unlike lower motor neuron dysfunction, however, the discharging units fail to fire rapidly to compensate for a decreased recruitment, because of the lack of central drive. In healthy subjects, the recruitment frequency, defined as the firing frequency of the initially activated unit at the time an additional unit is recruited, averages 5 to 10 Hz, depending on the types of motor units under study.141,294,295

Nerve Conduction and Needle Electromyography

953

FIGURE 35–38 A, Motor unit potentials from the extensor digitorum communis in a 20-year-old man with partial radial nerve palsy. Minimal (a and d), moderate (b and e) and maximal (c and f) voluntary contraction recruited only a single motor unit that discharged at progressively higher rates. B, Motor unit potentials from the extensor carpi ulnaris (a, b, and c) and extensor carpi radialis longus (d, e, and f) in the same subject. Maximal voluntary contraction recruited only a single motor unit firing at a high discharge rate. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

This measure often deviates from the expected range in patients with neuromuscular disorders,121 although normal and abnormal values overlap considerably. Another practical measure of recruitment is the ratio of the average firing rate to the number of active units.67 Healthy subjects have a ratio less than 5 with, for example, three units firing less than 15 Hz each. A ratio greater than 10, with two units firing over 20 Hz, indicates a loss of motor units. At greater effort (Fig. 35–40), simultaneous activation of many different units makes recognition of individual motor unit potentials difficult; hence the term interference pattern. Its analysis by the spike density and the average amplitude of the summated response provides a simple quantitative means of evaluating the number of firing units

activated with maximal contraction. This measure reflects the complex interaction among many different biologic factors. These include descending input from the cortex, number of discharging motor neurons, firing frequency of each motor unit, waveform of individual potentials, and probability of phase cancellation.

Measurements of Turns and Amplitude Weak voluntary effort mainly activates motor units composed of low-threshold type I muscle fibers. Wider ranges of motor units are represented in quantitative assessment of the interference pattern during strong muscle contraction.333 An automated technique to assess the number of

954

Nerve Conduction and Electromyography

duration, and number of turns with age. The ratio of turns to mean amplitude is increased in myopathy, especially at 10% to 20% of maximum force, whereas it is decreased in neurogenic disorders, mainly at a force of 20% to 30%.120 Similarly, the ratio of root mean square voltage to turns is increased in chronic neuropathies.109 Quantitative measurements of recruitment patterns in infants and young children also help differentiate primary muscle disease from neurogenic lesions.336 In contrast to evaluation of individual potentials, which allows precise description of motor units and their temporal stability, analysis of recruitment demonstrates the number and discharge pattern of all the motor units to characterize an overall muscle performance.

FIGURE 35–39 Low-amplitude, short-duration motor unit potentials from the biceps (a, b, and c) and tibialis anterior (d, e, and f) in a 7-year-old boy with Duchenne type muscular dystrophy. Minimal voluntary contraction recruited an excessive number of motor units in both muscles. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

“turns” counts directional changes of a waveform that do not necessarily cross the baseline during constant levels of muscle contraction.383 This method selects potential changes exceeding a minimum excursion usually of 100 V, measuring the amplitude from one point of change in direction to the next during a fixed time epoch.147 Automated methods usually report turns frequencies,171 the maximal ratio of turns to mean amplitude or peak ratio, and the number of time intervals between turns.237 Turns and spectral analyses of interference patterns, albeit indirectly, serve to characterize the physiologic properties of the motor units.81 With greater voluntary contraction, the number of turns increases faster than does the mean amplitude change between turns at low to moderate force levels, followed by a reversed pattern at higher force levels.265 Therefore, the relationship between the turns and amplitude critically depends on the level of effort during the recordings, as indicated by the shape of the “normal cloud.” For example, an increment of voluntary contraction from 10% to 30% of the maximal significantly increased the mean firing rate, number of turns, average amplitude, and rise rate.80 Clinical studies therefore require precise control of contractile force as a major determinant of waveform and firing properties. In one study,158 both low-threshold and high-threshold motor units showed a linear increase in mean amplitude,

FIGURE 35–40 Interference patterns seen in the triceps of a 44-year-old healthy man (a), tibialis anterior of a 52-year-old man with amyotrophic lateral sclerosis (b), and quadriceps of a 20-year-old man with limb-girdle dystrophy (c). Discrete single motor unit discharge in b stands in good contrast to abundant motor unit potentials with reduced amplitude in c. (From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.)

Nerve Conduction and Needle Electromyography

Lower and Upper Motor Neuron Disorders As mentioned earlier, the number and the average force contributed by each functional motor unit determine the recruitment pattern. A limited recruitment in disorders of the motor neuron, root, or peripheral nerve results from a reduced number of excitable motor units, which fire at an inappropriately rapid rate. In an extreme example, a single motor unit potential discharges at frequencies as high as 50 Hz, producing a discrete “picket fence” appearance at maximal effort (see Figs. 35–38 and 35–40). Patients with upper motor neuron lesions may also have characteristic firing behavior in the temporal discharge pattern of single motor units. In one study of 15 stroke patients with paretic tibialis anterior,118 low-threshold motor units fired within the lower end of the normal range, whereas high-threshold motor units, if recruited at all, discharged below their normal range. In another study with hemiparetic patients,130 two characteristic patterns emerged: compression in the range of motor neuron recruitment forces and failure to discharge motor units at a higher rate despite maximal effort to contract the paretic muscles. Thus an upper motor neuron or hysterical paralysis shows a slow rate of discharge despite a late recruitment and a reduced interference pattern. Additional characteristics of hysterical weakness or poor cooperation include irregular, grouped firing of motor units, not seen in a genuine paresis unless the patient also suffers from essential or other type of tremor.

Myopathy In myopathy, many units are recruited early to functionally compensate in quantity for a smaller force per unit. In typical myopathies (see Fig. 35–39), many axons discharge almost instantaneously with a slight voluntary effort, developing a full interference pattern at less than maximal contraction (see Fig. 35–40). In diseases of neuromuscular transmission, motor units also show an early recruitment, reaching a full interference prematurely for the same reason. Loss of whole motor units rather than individual muscle fibers in advanced myogenic disorders, however, may lead to a limited recruitment and incomplete interference pattern, mimicking a neuropathic change caused by a reduced number of motor units.

Involuntary Movement Electromyography helps characterize involuntary motor symptoms, which include neuromyotonia and related disorders associated with spontaneous discharges of motor units. In tremors, irregular bursts of motor unit potentials repeat at a fairly constant rate. During each burst, which varies in amplitude, duration, and waveform, many motor units are active simultaneously, showing no fixed temporal or spatial relationships. This pattern of grouping, when seen in association with a subclinical tremor burst, may mimic a polyphasic motor unit

955

potential. Electromyographic recordings can quantitatively delineate the rate, rhythm, and distribution of different types of tremor.328 Aberrant regeneration, in the setting of hemifacial spasm, for example, gives rise to clinical synkinesis or unintended coactivation of distant muscles not voluntarily contracted. In this condition, simultaneous recording shows the presence of time-locked discharge of motor units in multiple muscles involved in synkinetic movements. This should be distinguished from ordinary associated movement, often seen in the face, without a precise time relationship among co-contracting muscle groups.

OTHER ASPECTS OF ELECTROMYOGRAPHY Principles of Localization In addition to the studies of myopathies and neuropathies, electromyography plays a critical role in delineating the location and distribution of a lesion responsible for clinical abnormalities. In particular, a unique combination of clinical and electrodiagnostic findings helps elucidate the level of a radicular and plexus lesion.253 The clinical assessment depends on the distribution of sensory deficits and of motor involvement that consist of weakness, muscle atrophy, hyporeflexia, fatigue, cramps, and fasciculations. Electromyographic studies can confirm the clinical localization of a lesion if appropriate muscle groups are selected for needle examination based on neurologic findings225 (see Tables 35–24 and 35–25). In the upper limbs, motor deficits usually provide a more reliable localizing sign than sensory impairments (see Table 35–24). The reverse seems to prevail in the lower limbs, where anatomic peculiarity makes electrophysiologic localization of radicular lesions more difficult. Evidence of denervation in the deep cervical muscles innervated by the posterior, as opposed to anterior, rami indicates an intraforaminal lesion affecting the root or spinal nerve before the division into the two rami. Thus studies of paraspinal muscles, by documenting the involvement of the posterior rami, serve to distinguish between a radicular and a plexus lesion. Other muscles of interest in this regard include the rhomboids, supplied by the dorsal scapular nerve, and the serratus anterior, subserved by the long thoracic nerves. Because of their very proximal innervation, spontaneous activity recorded therein can also confirm a radicular as opposed to a plexus lesion. In the lumbosacral region, discs tend to compress the roots slightly above the level of their respective foramina before their lateral deviation toward the exit. Needle studies help identify the damaged root and confirm the diagnosis (see Table 35–25). As in the cervical region, the evidence of denervation in the paraspinal muscles locates the lesion to the root or spinal nerve, which is proximal to the origin of the posterior ramus. The reverse does not necessarily hold because

956

Nerve Conduction and Electromyography

Table 35–24. Innervation Patterns of the Cranial, Shoulder Girdle, and Upper Limb Muscles* Nerves Anterior Primary Rami Cervical plexus Spinal accessory nerve Phrenic nerve Brachial plexus Dorsal scapular nerve Suprascapular nerve Axillary nerve Subscapular nerve Musculocutaneous nerve

Long thoracic nerve Lateral pectoral nerve Medial pectoral nerve Radial nerve

Posterior interosseous nerve

Median nerve

Anterior interosseous nerve

Muscles

C2

Sternocleidomastoid Trapezius, upper, middle, lower Diaphragm

C3

C4

C5

C6

C7

C8

T1

d

Rhomboid Supraspinatus Infraspinatus Teres minor Deltoid, anterior, middle, posterior Teres major Brachialis Biceps brachi Coracobrachialis Serratus anterior Pectoralis major (clavicular part) Pectoralis minor Brachioradialis Extensor carpi radialis Triceps, long, lateral, middle heads Anconeus Supinator Extensor carpi ulnaris Extensor digitorum Extensor pollicis brevis Extensor indicis Pronator teres Flexor carpi radialis Abductor pollicis brevis Flexor digitorum pronfundus (I & II) Pronator quadratus Flexor pollicis longus

Ulnar nerve

Flexor digitorum profundus (III & IV) Flexor carpi ulnaris Adductor pollicis Abductor digiti minimi Interossei, volar (I–III), dorsal (I–IV)

Posterior Primary Rami

Cervical erector spinae

*Solid black indicates primary innervation; shading indicates secondary innervation. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

root compression may spare the paraspinal muscles. A cauda equina lesion, though asymmetrical in distribution, otherwise resembles a conus medullaris lesion, with bilateral involvement of the level ordinarily unaffected by a herniated lumbar disc. Electromyographic studies show fibrillation potentials

and large motor unit potentials in the myotomes of several lumbosacral roots, including paraspinal muscles17 and the urethral sphincter.114 These findings resemble those of an intrinsic conus involvement except for an asymmetrical distribution of the abnormalities with spread above the sacral region.

957

Nerve Conduction and Needle Electromyography

Table 35–25. Innervation Patterns of the Hip Girdle and Lower Limb Muscles* Nerves Anterior Primary Rami Lumbosacral plexus Femoral nerve

Obturator nerve Superior gluteal nerve

Inferior gluteal nerve Sciatic nerve Tibial division Peroneal division Common peroneal nerve Deep peroneal nerve

Superficial peroneal nerve

Tibial nerve

Medial plantar nerve Lateral plantar nerve

Posterior Primary Rami

Muscles

L2

L3

L4

L5

S1

S2

Iliopsoas Sartorius Rectus femoris Vastus lateralis, medialis Gracilis Adductor longus, brevis, magnus Gluteus medius Gluteus minimus Tensor fasciae latae Gluteus maximus

Semitendinosus, semimembranosus Biceps femoris, long head Biceps femoris, short head

Tibialis anterior Extensor digitorum longus Extensor digitorum brevis Extensor hallucis longus Peroneus longus Peroneus brevis Tibialis posterior Flexor digitorum longus Flexor hallucis longus Gastrocnemius, medial head Gastrocnemius, lateral head Soleus Abductor hallucis Abductor digiti minimi Interossei Lumbosacral erector spinae

*Solid black indicates primary innervation; shading indicates secondary innervation. From Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001, with permission.

The multifidus muscles receive innervation from a single root, in contrast to the rest of the paraspinal muscles, which are subserved polysegmentally.48 Nonetheless, paraspinal abnormalities usually do not reveal the exact level of the involved segment.144 Determination of the affected myotome, therefore, depends on careful exploration of the limb muscles to delineate the pattern of involvement. Anomalous innervation among different cord segments abounds, challenging the clinician in attributing any pattern of clinical or electromyographic findings to a specific spinal level.301 A computer-aided expert system may have some place in the evaluation of brachial plexus

injuries as advocated by some,111 although its clinical value remains unknown. Some studies report a high correlation among electromyographic evidence of denervation, myelographic abnormalities, and surgical findings.235 In one series,267 however, electromyography and magnetic resonance imaging (MRI) agreed in only 60% of patients. Because one or the other evaluation showed some abnormality in the remaining 40%, these two modes of study provided complementary diagnostic information. T2-weighted and short time to inversion recovery sequences of MRI can supplement the electrophysiologic evaluations of functional deficits by detecting denervated

958

Nerve Conduction and Electromyography

skeletal muscle through increased signal intensity. One study50 reported a close correlation between this abnormality and spontaneous activities on electromyographic examination. Electromyographic evaluation can uncover functional deficits that do not necessarily produce structural abnormalities.363 In addition, such studies may also elucidate the extent and chronicity of lesions,381 and can guide patient management by substantiating clinical progression or improvement.170 Electrodiagnosis plays a particularly important role in justifying surgical exploration if radiologic and clinical findings conflict.368 For example, extraforaminal compression of the L5 root by lumbosacral ligaments, though not detectable by myelography and other imaging studies, may cause denervation.281 Conversely, asymptomatic subjects with abnormal MRI scans of the lumbar spine must undergo further study to seek a physiologic and clinical correlation.21

volitional contraction, give rise to motor unit potentials. Structural or functional disturbances of the motor unit seen in diseases of the nerve or muscle lead to alterations of waveform and discharge patterns of their electrical signals. The study of motor unit potentials provides information useful in elucidating the nature of the disease. Certain characteristics of such abnormalities may suggest a particular pathologic process. As a clinical tool, electromyography serves best only if the examiner conducts the study and interprets the results in light of the patient’s history, physical examination, and other diagnostic findings. In fact, the study constitutes an extension of the physical examination, rather than an independent laboratory test. Therefore, it is most useful if performed by a physician thoroughly familiar with the patient’s clinical findings with the aim to prove or disprove the diagnostic impression.

Sequence of Abnormalities After nerve injury, needle examination of the affected muscle initially reveals poor recruitment of motor unit potentials, which may result from either structural or functional loss of axons. Fibrillation potentials and positive sharp waves appear 2 to 3 weeks after axonal degeneration. During active regeneration of motor axons, low-amplitude, polyphasic motor unit potentials have temporal instability. Subsequent development of high-amplitude, long-duration motor unit potentials with stable configuration signals completion of reinnervation. Axonal degeneration after a nerve injury is said to follow a length-dependent delay, with muscles subserved by a shorter nerve showing the signs of denervation earlier.256 If so, needle examination should detect fibrillation potentials and positive sharp waves first in the paraspinal muscles. In one study using multivariate estimates,299 however, paraspinal spontaneous activity showed no correlation with symptom duration. In clinical practice, therefore, this time relationship may not necessarily hold.78

SUMMARY Electromyographic studies analyze spontaneous and voluntarily activated muscle action potentials extracellularly. Following brief injury potentials coincident with the insertion of the needle, a relaxed normal muscle remains electrically silent except for the end-plate activities. Several types of spontaneous discharges all signal diseases of the nerve or muscle, although they carry different clinical implications. Both fibrillation potentials and positive sharp waves represent spontaneous excitation of individual muscle fibers. The complex repetitive discharges comprise high-frequency spikes derived from multiple muscle fibers, which discharge sequentially, maintaining a fixed order. In conventional electromyography, isolated discharges of a single motor unit, the smallest functional element of

REFERENCES 1. Ackil, A. A., Shahani, B. R., and Young, R. R.: Sural nerve conduction studies and late responses in children undergoing hemodialysis. Arch. Phys. Med. Rehabil. 62:487, 1981. 2. Afifi, A. K., Kimura, J., and Bell, W. E.: Hypothermiainduced reversible polyneuropathy: electrophysiologic evidence of axonopathy. Pediatr. Neurol. 4:49, 1988. 3. Aho, K., and Sainio, K.: Late irradiation-induced lesions of the lumbosacral plexus. Neurology 33:953, 1983. 4. Albers, J. W.: Clinical neurophysiology of generalized polyneuropathy. J. Clin. Neurophysiol. 10:149, 1993. 5. Albers, J. W.: Principles of sensory nerve conduction studies. Presented at the 49th Annual Meeting of the American Academy of Neurology, Boston, April 12–19, 1997. 6. Albers, J. W., Allen, A. A., Bastron, J. A., and Daube, J. R.: Limb myokymia. Muscle Nerve 4:494, 1981. 7. American Association of Electrodiagnostic Medicine: Consensus criteria for the diagnosis of partial conduction block. Muscle Nerve 22(Suppl 8):S225, 1999. 8. American Association of Electrodiagnostic Medicine: Practice parameter for electrodiagnostic studies in ulnar neuropathy at the elbow: summary statement. Muscle Nerve 22:408, 1999. 9. Amoiridis, G.: Median-ulnar nerve communications and anomalous innervation of the intrinsic hand muscles: an electrophysiological study. Muscle Nerve 15:576, 1992. 10. Argyropoulos, C. J., Panayiotopoulos, C. P., and Scarpalezos, S.: F- and M-wave conduction velocity in amyotrophic lateral sclerosis. Muscle Nerve 1:479, 1978. 11. Ashworth, N. L., Marshall, S. C., and Satkunam, L. E.: The effect of temperature on nerve conduction parameters in carpal tunnel syndrome. Muscle Nerve 21:1089, 1998. 12. Auger, R. G.: Role of the masseter reflex in the assessment of subacute sensory neuropathy. Muscle Nerve 21:800, 1998. 13. Baba, M., Gilliatt, W., and Jacobs, J. M.: Recovery of distal changes after nerve constriction by a ligature. J. Neurol. Sci. 60:235, 1983.

Nerve Conduction and Needle Electromyography 14. Bamford, C. R., Rothrock, R. J., and Swenson, M.: Average techniques to define the low-amplitude compound motor action potentials. Arch. Neurol. 41:1307, 1984. 15. Barkhaus, P. E., and Nandedkar, S. D.: Recording characteristics of the surface EMG electrodes. Muscle Nerve 17:1317, 1994. 16. Barkhaus, P. E., and Nandedkar, S. D.: On the selection of concentric needle electromyogram motor unit action potentials: is the rise time criterion too restrictive? Muscle Nerve 19:1554, 1996. 17. Bartleson, J. D., Cohen, M. D., Harrington, T. M., et al.: Cauda equina syndrome secondary to long-standing ankylosing spondylitis. Ann. Neurol. 14:662, 1983. 18. Bertram, M. F., Nishida, T., Minieka, M. M., et al.: Effects of temperature on motor unit action potentials during isometric contraction. Muscle Nerve 18:1443, 1995. 19. Bischoff, C., Stålberg, E., Falck, B., and Puksa, L.: Significance of A-waves recorded in routine motor nerve conduction studies. Electroencephalogr. Clin. Neurophysiol. 101:528, 1996. 20. Black, J. T., Bhatt, G. P., DeJesus, P. V., et al.: Diagnostic accuracy of clinical data, quantitative electromyography and histochemistry in neuromuscular disease: a study of 105 cases. J. Neurol. Sci. 21:59, 1974. 21. Boden, S. D., Davis, D. O., Dina, T. S., et al.: Abnormal magnetic-resonance scans of the lumbar spine in asymptomatic subjects. J. Bone Joint Surg. Am. 72:403, 1990. 22. Botte, M. J., Cohen, M. S., Lavernia, C. J., et al.: The dorsal branch of the ulnar nerve: an anatomic study. J. Hand Surg. Am. 15:603, 1990. 23. Bougle, D., Denise, P., Yaseen, H., et al.: Maturation of peripheral nerves in preterm infants: motor and proprioceptive nerve conduction. Electroencephalogr. Clin. Neurophysiol. 75:118, 1990. 24. Bouisset, S.: EMG and muscle force in normal motor activities. In Desmedt, J. E. (ed.): New Developments in Electromyography and Clinical Neurophysiology, Vol. 1. Basel, Karger, p. 547, 1973. 25. Braddom, R. I., and Johnson, E. W.: Standardization of H reflex and diagnostic use in S1 radiculopathy. Arch. Phys. Med. Rehabil. 55:161, 1974. 26. Brick, J. F., Gutmann, L., and McComas, C. F.: Calcium effect on generation and amplification of myokymic discharges. Neurology 32:618, 1982. 27. Bromberg, M. B., and Albers, J. W.: Patterns of sensory nerve conduction abnormalities in demyelinating and axonal peripheral nerve disorders. Muscle Nerve 16:262, 1993. 28. Bromberg, M. B., and Jaros, L.: Symmetry of normal motor and sensory nerve conduction measurements. Muscle Nerve 21:498, 1998. 29. Bromberg, M. B., and Spiegelberg, T.: The influence of active electrode placement on CMAP amplitude. Electroencephalogr. Clin. Neurophysiol. 105:385, 1997. 30. Brooke, M. H., Carroll, J. E., Davis, J. E., and Hagberg, J. M.: The prolonged exercise test. Neurology 29:636, 1979. 31. Bryant, P. R., and Eng, G. D.: Normal values for the soleus H-reflex in newborn infants 31–45 weeks post conceptional age. Arch. Phys. Med. Rehabil. 72:28, 1991. 32. Buchthal, F.: An Introduction to Electromyography. Copenhagen, Scandinavian University Books, 1957.

959

33. Buchthal, F.: The general concept of the motor unit. In Adams, R. D., Eaton, L. M., and Shy, G. M. (eds.): Neuromuscular Disorders. Baltimore, Williams & Wilkins, p. 3, 1960. 34. Buchthal, F.: Electrophysiological abnormalities in metabolic myopathies and neuropathies. Acta. Neurol. Scand. Suppl. 43:129, 1970. 35. Buchthal, F.: Diagnostic significance of the myopathic EMG. In Rowland, L. P. (ed.): Pathogenesis of Human Muscular Dystrophies: Proceedings of the Fifth International Scientific Conference of the Muscular Dystrophy Association, Durango, CO, June 1976. Amsterdam, Excerpta Medica, p. 205, 1977. 36. Buchthal, F.: Electrophysiological signs of myopathy as related with muscle biopsy. Acta Neurol. (Napoli) 32:1, 1977. 37. Buchthal, F.: Fibrillations: clinical electrophysiology. In Culp, W. J., and Ochoa, J. (eds.): Abnormal Nerves and Muscles as Impulse Generators. Oxford, UK, Oxford University Press, p. 632, 1982. 38. Buchthal, F., Engbaek, L., and Gamstorp, I.: Paresis and hyperexcitability in adynamia episodica hereditaria. Neurology 8:347, 1958. 39. Buchthal, F., and Rosenfalck, A.: Evoked action potentials and conduction velocity in human sensory nerves. Brain Res. 3:1, 1966. 40. Buchthal, F., Rosenfalck, A., and Behse, F.: Sensory potentials of normal and diseased nerves. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy, Vol. 1. Philadelphia, W. B. Saunders, p. 442, 1975. 41. Buchthal, F., Rosenfalck, A., and Trojaborg, W.: Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J. Neurol. Neurosurg. Psychiatry 37:340, 1974. 42. Buchthal, F., and Rosenfalck, P.: Electrophysiological aspects of myopathy with particular reference to progressive muscular dystrophy. In Bourne, G. H., and Golarz, M. N. (eds.): Muscular Dystrophy in Man and Animals. New York, Hafner Publishing Company, p. 194, 1963. 43. Buchthal, F., and Rosenfalck, P.: Spontaneous electrical activity of human muscle. Electroencephalogr. Clin. Neurophysiol. 20:321, 1966. 44. Buschbacher, R. M.: Body mass index effect on common nerve conduction study measurement. Muscle Nerve 21:1398, 1998. 45. Campbell, W. W.: The value of inching techniques in the diagnosis of focal nerve lesions. Muscle Nerve 21:1554, 1998. 46. Campbell, W. W.: Ulnar neuropathy at the elbow. Muscle Nerve 23:478, 2000. 47. Campbell, W. W., Pridgeon, R. M., and Sahni, K. S.: Short segment incremental studies in the evaluation of ulnar neuropathy at the elbow. Muscle Nerve 15:1050, 1992. 48. Campbell, W. W., Vasconcelos, O., and Laine, F. J.: Focal atrophy of the multifidus muscle in lumbosacral radiculopathy. Muscle Nerve 21:1350, 1998. 49. Carpendale, M. T. F.: Conduction time in the terminal portion of the motor fibers of the ulnar, median, and peroneal nerves in healthy subjects and in patients with neuropathy. Thesis, University of Minnesota, Minneapolis, 1956.

960

Nerve Conduction and Electromyography

50. Carter, G. T., and Fritz, R. C.: Electromyographic and lower extremity short time to inversion recovery magnetic resonance imaging findings in lumbar radiculopathy [short report]. Muscle Nerve 20:1191, 1997. 51. Cerra, D., and Johnson, E. W.: Motor nerve conduction velocity in premature infants. Arch. Phys. Med. Rehabil. 43:160, 1962. 52. Chang, C. W., Cho, H. K., and Oh, S. J.: Posterior antebrachial cutaneous neuropathy: case report. Electromyogr. Clin. Neurophysiol. 29:109, 1989. 53. Chaudhry, V., Corse, A. M., Freimer, M. L., et al.: Interand intraexaminer reliability of nerve conduction measurements in patients with diabetic neuropathy. Neurology 44:1459, 1994. 54. Cherington, M.: Accessory nerve: conduction studies. Arch. Neurol. 18:708, 1968. 55. Cholachis, S. C., Pease, W. S., and Johnson, E. W.: Polyphasic motor unit action potentials in early radiculopathy: their presence and ephaptic transmission as an hypothesis. Electromyogr. Clin. Neurophysiol. 32:27, 1992. 56. Chroni, E., and Panayiotopoulos, C. P.: F-wave quantitation in neuropathy. Muscle Nerve 18:786, 1995. 57. Claussen, G. C., Ahmad B. K., Sunwood, I. N., and Oh, S. J.: Combined motor and sensory median-ulnar anastomosis: report of an electrophysiologically proven case. Muscle Nerve 19:231, 1996. 58. Clay, S. A., and Ramseyer, J. C.: The orbicularis oculi reflex: pathologic studies in childhood. Neurology 27:892, 1977. 59. Colachis, S. C. III, Klejka, J. P., Shamir, D. Y., et al.: Amplitude of M responses: side to side comparability. Am. J. Phys. Med. Rehabil. 72:19, 1993. 60. Conrad, B., Aschoff, J. C., and Fischler, M.: Der Diagnostische Wert der F-Wellen-Latenz. J. Neurol. 210:151, 1975. 61. Conrad, B., Jacobi, H. M., and Prochazka, V. J.: Unusual properties of repetitive fasciculations. Electroencephalogr. Clin. Neurophysiol. 35:173, 1973. 62. Cornblath, D. R., Sumner, A. J., Daube, J., et al.: Conduction block in clinical practice. Muscle Nerve 14:869, 1991. 63. Crutchfield, C. A., and Gutmann, L.: Hereditary aspects of accessory deep peroneal nerve. J. Neurol. Neurosurg. Psychiatry 36:989, 1973. 64. Cruz Martinez, A., Barrio, M., Perez Conde, M. C., and Ferrer, M. T.: Electrophysiological aspects of sensory conduction velocity in healthy adults. 2. Ratio between the amplitude of sensory evoked potentials at the wrist on stimulating different fingers in both hands. J. Neurol. Neurosurg. Psychiatry 41:1097, 1978. 65. Cruz Martinez, A., Barrio, M., Perez Conde, M. C., and Gutierrez, A. M.: Electrophysiological aspects of sensory conduction velocity in healthy adults. 1. Conduction velocity from digit to palm, from palm to wrist, and across the elbow, as a function of age. J. Neurol. Neurosurg. Psychiatry 41:1092, 1978. 66. Cummins, K. L., Dorfman, L. J., and Perkel, D. H.: Nerve fiber conduction-velocity distributions. II. Estimation based on two compound action potentials. Electroencephalogr. Clin. Neurophysiol. 46:647, 1979.

67. Daube, J. R.: Needle Examination in Electromyography (Minimonograph no 11). Rochester, MN, American Association of Electromyography and Electrodiagnosis, 1979. 68. Daube, J. R.: AAEM Minimonograph #11: Needle examination in clinical electromyography. Muscle Nerve 14:685, 1991. 69. Daube, J. R.: Electrodiagnosis of muscle disorders. In Engel, A. G., and Franzini-Armstrong, C. (eds.): Myology. New York, McGraw-Hill, p. 764, 1994. 70. Daube, J. R., Kelly, J. J. Jr., and Martin, R. A.: Facial myokymia with polyradiculoneuropathy. Neurology 29:662, 1979. 71. de Carvalho, M., and Swash, M.: Fasciculation potentials: a study of amyotrophic lateral sclerosis and other neurogenic disorders. Muscle Nerve 21:336, 1998. 72. Dengler, R., Wolf, W., Schubert, M., and Struppler, A.: Discharge pattern of single motor units in basal ganglia disorders. Neurology 36:1061, 1986. 73. Denys, E. H.: AAEM Minimonograph #14: The influence of temperature in clinical neurophysiology. Muscle Nerve 14:795, 1991. 74. Desjacques, P., Egloff-Baer, S., and Roth, G.: Lumbrical muscles and the carpal tunnel syndrome. Electromyogr. Clin. Neurophysiol. 20:443, 1980. 75. Desmedt, J. E., and Borenstein, S.: Regeneration in Duchenne muscular dystrophy: electromyographic evidence. Arch. Neurol. 33:642, 1976. 76. Dessi, F., Durand, G., and Hoffmann, J. J.: The accessory deep peroneal nerve: a pitfall for the electromyographer. J. Neurol. Neurosurg. Psychiatry 55:214, 1992. 77. Devi, S., Lovelace, R. E., and Duarte, N.: Proximal peroneal nerve conduction velocity: recording from anterior tibial and peroneus brevis muscles. Ann. Neurol. 2:116, 1977. 78. Dillingham, T. R., Pezzin, L. E., and Lauder, T. D.: Cervical paraspinal muscle abnormalities and symptom duration: a multivariate analysis [short report]. Muscle Nerve 21:640, 1998. 79. Dioszeghy, P., and Stålberg, E.: Changes in motor and sensory nerve conduction parameters with temperature in normal and diseased nerve. Electroencephalogr. Clin. Neurophysiol. 85:229, 1992. 80. Dorfman, L. J., Howard, J. E., and McGill, K. C.: Influence of contractile force on properties of motor unit action potentials: ADEMG analysis. J. Neurol. Sci. 86:125, 1988. 81. Dorfman, L. J., and McGill, K. C.: AAEE Minimonograph #29: Automatic quantitative electromyography. Muscle Nerve 11:804, 1988. 82. Dumitru, D.: Single muscle fiber discharges (insertional activity, end-plate potentials, positive sharp waves, and fibrillation potentials): a unifying proposal. Muscle Nerve 19:221, 1996. 83. Dumitru, D., and King, J. C.: Motor unit action potential duration and muscle length. Muscle Nerve 22:1188, 1999. 84. Dumitru, D., King, J. C., and Nandedkar, S. D.: Comparison of single-fiber and macro electrode recordings: relationship to motor unit action potential duration. Muscle Nerve 20:1381, 1997.

Nerve Conduction and Needle Electromyography 85. Dumitru, D., King, J. C., and Nandedkar, S. D.: Motor unit action potentials recorded with concentric electrodes: physiologic implications. Electroencephalogr. Clin. Neurophysiol. 105:333, 1997. 86. Dumitru, D., King J. C., Rogers, W., and Stegeman, D. F.: Positive sharp wave and fibrillation potential modeling. Muscle Nerve 22:242, 1999. 87. Dyck, P. J., Kratz, K. M., Lehman, K. A., et al.: The Rochester Diabetic Neuropathy Study: design, criteria for types of neuropathy, selection bias, and reproducibility of neuropathic tests. Neurology 41:799, 1991. 88. Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerve in health, Friedreich’s ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Remond, A. (ed.): Handbook of Electroencephalography and Clinical Neurophysiology, Vol. 9. Amsterdam, Elsevier, p. 83, 1972. 89. Eisen, A.: The physiologic basis and clinical applications of needle EMG in neuromuscular abnormalities: principles and pitfalls in the practice of EMG and NCS. Presented at the 49th Annual Meeting of the American Academy of Neurology, Boston, April 12–19, 1997. 90. Eisen, A., and Danon, J.: The orbicularis oculi reflex in acoustic neuromas: a clinical and electrodiagnostic evaluation. Neurology 24:306, 1974. 91. Eisen, A., Schomer, D., and Melmed, C.: The application of F-wave measurements in the differentiation of proximal and distal upper limb entrapments. Neurology 27:662, 1977. 92. Ekstedt, J., and Stalberg, E.: How the size of the needle electrode leading-off surface influences the shape of the single muscle fibre action potential in electromyography. Comput. Prog. Biomed. 3:204, 1973. 93. Emeryk, B., Hausmanowa-Petrusewicz, I., and Nowak, T.: Spontaneous volleys of bizarre high frequency potentials (b.h.f.p.) in neuro-muscular diseases. Part I. Occurrence of spontaneous volleys of b.h.f.p. in neuro-muscular diseases. Part II. An analysis of the morphology of spontaneous volleys of b.h.f.p. in neuromuscular diseases. Electromyogr. Clin. Neurophysiol. 14:303 (Part I); 339 (Part II), 1974. 94. Emeryk-Szajewska, B., Kopec, J., and Karwanska, A.: The reorganization of motor units in motor neuron disease. Muscle Nerve 20:306, 1997. 95. Engel, W. K.: Brief, small, abundant motor-unit action potentials: a further critique of electromyographic interpretation. Neurology 25:173, 1975. 96. Engel, W. K., and Warmolts, J. R.: The motor unit: diseases affecting it in toto or in partia. In Desmedt, J. E. (ed.): New Developments in Electromyography and Clinical Neurophysiology, Vol. 1. Basel, Karger, p. 141, 1973. 97. Erim, Z., De Luca, C. J., and Mineo, K.: Rank-ordered regulation of motor units. Muscle Nerve 19:563, 1996. 98. Ertekin, C., Mungan, B., and Uludag, B.: Sacral cord conduction time of the soleus H-reflex. J. Clin. Neurophysiol. 13:77, 1996. 99. Ertekin, C., Nejat, R. S., Sirin, H., et al.: Comparison of magnetic coil and needle-electrical stimulation in diagnosis of lumbosacral radiculopathy. Muscle Nerve 17:685, 1994.

961

100. Ertekin, C., Nejat, R. S., Sirin, H., et al.: Comparison of magnetic coil stimulation and needle electrical stimulation in diagnosis of lumbosacral radiculopathy. Clin. Neurol. Neurosurg. 96:124, 1994. 101. Falco, F. J. E., Hennessey, W. J., Braddom, R. L., and Goldberg, G.: Standardized nerve conduction studies in the upper limb of the healthy elderly. Am. J. Phys. Med. Rehabil. 71:263, 1992. 102. Feasby, T. E., Brown, W. F., Gilbert, J. J., and Hahn, A. F. D.: The pathological basis of conduction block in human neuropathies. J. Neurol. Neurosurg. Psychiatry 48:239, 1985. 103. Feinberg, D. M., Preston, D. C., Shefner, J. M., and Logigian, E. L.: Amplitude-dependent slowing of conduction in amyotrophic lateral sclerosis and polyneuropathy. Muscle Nerve 22:937, 1999. 104. Ferrante, M. A., Olney, R., and Wilbourn, A. J.: Sensory nerve conduction study workshop. Presented at the 49th Annual Meeting of the American Academy of Neurology, Boston, April 12–19, 1997. 105. Ferrante, M. A., and Wilbourn, A. J.: The utility of various sensory nerve conduction responses in assessing brachial plexopathies. Muscle Nerve 18:879, 1995. 106. Finsterer, J., and Mamoli, B.: Satellite potentials as a measure of neuromuscular disorders. Muscle Nerve 20:585, 1997. 107. Fisher, M. A.: H reflexes and F-wave physiology and clinical indications. Muscle Nerve 15:1223, 1992. 108. Fisher, M. A.: Whither F waves. In Kimura, J., and Shibasaki, H. (eds.): Recent Advances in Clinical Neurophysiology. Amsterdam, Elsevier, p. 752, 1996. 109. Fisher, M. A.: Root mean square voltage/turns in chronic neuropathies is related to increase in fiber density. Muscle Nerve 20:241, 1997. 110. Fisher, M. A., Shivde, A. J., Teixera, C., and Grainer, L. S.: Clinical and electrophysiological appraisal of the significance of radicular injury in back pain. J. Neurol. Neurosurg. Psychiatry 41:303, 1978. 111. Fisher, W. S. III: Computer-aided intelligence: application of an expert system to brachial plexus injuries. Neurosurgery 27:837, 1990. 112. Fitz, W. R., Mysiw, W. J., and Johnson, E. W.: First lumbrical latency and amplitude: control values and findings in carpal tunnel syndrome. Am. J. Phys. Med. Rehabil. 69:198, 1990. 113. Fowler, C. J., Kirby, R. S., and Harrison, M. J. G: Decelerating burst and complex repetitive discharges in the striated muscle of the urethral sphincter, associated with urinary retention in women. J. Neurol. Neurosurg. Psychiatry 48:1004, 1985. 114. Fowler, C. J., Kirby, R. S., Harrison, M. J. G., et al.: Individual motor unit analysis in the diagnosis of disorders of urethral sphincter innervation. J. Neurol. Neurosurg. Psychiatry 47:637, 1984. 115. Franssen, H., and Wieneke, G. H.: Nerve conduction and temperature: necessary warming time. Muscle Nerve 17:336, 1994. 116. Franssen, H., Wieneke, G. H., and Wokke, J. H. J.: The influence of temperature on conduction block. Muscle Nerve 22:166, 1999.

962

Nerve Conduction and Electromyography

117. Fraser, J. L., and Olney, R. K.: The relative diagnostic sensitivity of different F-wave parameters in various polyneuropathies. Muscle Nerve 15:912, 1992. 118. Frontera, W. R., Grimby, L., and Larsson, L.: Firing rate of the lower motoneuron and contractile properties of its muscle fibers after upper motoneuron lesion in man. Muscle Nerve 20:938, 1997. 119. Fu, R., DeLisa, J. A., and Kraft, G. H.: Motor nerve latencies through the tarsal tunnel in normal adult subjects: standard determinations corrected for temperature and distance. Arch. Phys. Med. Rehabil. 61:243, 1980. 120. Fuglsang-Frederiksen, A., Monaco, M., and Dahl, K.: Turns analysis (peak ratio) in EMG using the mean amplitude as a substitute of force measurement. Electroencephalogr. Clin. Neurophysiol. 60:225, 1985. 121. Fuglsang-Frederiksen, A., Smith, T., and Hogenhaven, H.: Motor unit firing intervals and other parameters of electrical activity in normal and pathological muscle. J. Neurol. Sci. 78:51, 1987. 122. Fullerton, P. M., and Gilliatt, R. W.: Axon reflexes in human motor nerve fibres. J. Neurol. Neurosurg. Psychiatry 28:1, 1965. 123. Fusfeld, R. D.: Electromyographic abnormalities in a case of botulism. Bull. Los Angeles Neurol. Soc. 35:164, 1970. 124. Gadia, M. T., Natori, N., Ramos, L. B., et al.: Influence of height on quantitative sensory, nerve conduction, and clinical indices of diabetic peripheral neuropathy. Diabetes Care 10:613, 1987. 125. Gamstorp, I.: Normal conduction velocity of ulnar, median and peroneal nerves in infancy, childhood and adolescence. Acta Paediatr. Suppl. 146:68, 1963. 126. Garcia-Meriono, A., Cabello, A., Mora, J. S., and Liano, H.: Continuous muscle fiber activity, peripheral neuropathy and thymoma. Ann. Neurol. 29:215, 1991. 127. Gassel, M. M.: A study of femoral nerve conduction time. Arch. Neurol. 9:57, 1963. 128. Geiringer, S. R.: Anatomic Localization for Needle Electromyography. Philadelphia, Hanley & Belfus, 1994. 129. Geiringer, S. R.: Inching techniques are of limited use. Muscle Nerve 21:1557, 1998. 130. Gemperline, J. J., Allen, S., Walk, D., and Rymer, W. Z.: Characteristics of motor unit discharge in subjects with hemiparesis. Muscle Nerve 18:1101, 1995. 131. Gilliatt, R. W.: Sensory conduction studies in the early recognition of nerve disorders. Muscle Nerve 1:352, 1978. 132. Gilliatt, R. W.: Acute compression block. In Sumner, A. (ed.): The Physiology of Peripheral Nerve Disease. Philadelphia, W. B. Saunders, p. 287, 1980. 133. Gilliatt, R. W., Goodman, H. V., and Willison, R. G.: The recording of lateral popliteal nerve action potentials in man. J. Neurol. Neurosurg. Psychiatry 24:305, 1961. 134. Gilliatt, R. W., Le Quesne, P. M., Logue, V., and Sumner, A. J.: Wasting of the hand associated with a cervical rib or band. J. Neurol. Neurosurg. Psychiatry 33:615, 1970. 135. Gilliatt, R. W., and Thomas, P. K.: Changes in nerve conduction with ulnar lesions at the elbow. J. Neurol. Neurosurg. Psychiatry 23:312, 1960. 136. Godaux, E., and Desmedt, J. E.: Evidence for a monosynaptic mechanism in the tonic vibration reflex of the human

137.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

150. 151.

152.

153.

154.

masseter muscle. J. Neurol. Neurosurg. Psychiatry 38:161, 1975. Godaux, E., and Desmedt, J. E.: Exteroceptive suppression and motor control of the masseter and temporalis muscles in normal man. Brain Res. 85:447, 1975. Golovchinsky, V.: Ulnar-to-median anastomosis and its role in the diagnosis of lesions of the median nerve at the elbow and the wrist. Electromyogr. Clin. Neurophysiol. 30:31, 1990. Griep, P. A. M., Gielen, F. L. H., Boom, H. B. K., et al.: Calculation and registration of the same motor unit action potential. Electroencephalogr. Clin. Neurophysiol. 53:388, 1982. Gruber, W.: Ueber die Verbindung des Nervus medianus mit dem Nervus ulnaris am Unterarme des Menschen und der Saugethiere. Arch. Anat. Physiol. Med. Leipzig, p. 501, 1870. Gunreben, G., and Schulte-Mattler, W.: Evaluation of motor unit firing rates by standard concentric needle electromyography. Electromyogr. Clin. Neurophysiol. 32:103, 1992. Gutmann, L.: Atypical deep peroneal neuropathy in presence of accessory deep peroneal nerve. J. Neurol. Neurosurg. Psychiatry 33:453, 1970. Gutmann, L.: Median-ulnar nerve communications and carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 40:982, 1977. Haig, A. J., Talley, C., Grobler, L. J., and LeBreck, D. B.: Paraspinal mapping: quantified needle electromyography in lumbar radiculopathy. Muscle Nerve 16:477, 1993. Hakamada, S., Kumagai, T., Watanabe, K., et al.: The conduction velocity of slower and the fastest fibres in infancy and childhood. J. Neurol. Neurosurg. Psychiatry 45:851, 1982. Hausmanowa-Petrusewicz, I., and Jedrzejowska, H.: Correlation between electromyographic findings and muscle biopsy in cases of neuromuscular disease [abstract]. J. Neurol. Sci. 13:85, 1971. Hayward, M.: Automatic analysis of the electromyogram in healthy subjects of different ages. J. Neurol. Sci. 33:397, 1977. Heckmann, R., and Ludin, H. P.: Differentiation of spontaneous activity from normal and denervated skeletal muscle. J. Neurol. Neurosurg. Psychiatry 45:331, 1982. Heifernan, L., Rewcastle, N. B., and Humphrey, J. G.: The spectrum of rod myopathies. Arch. Neurol. 18:529, 1968. Henneman, E.: Relation between size of neurons and their susceptibility to discharge. Science 126:1245, 1957. Hess, K., Kern, S., and Schiller, H. H.: Blink reflex in trigeminal sensory neuropathy. Electromyogr. Clin. Neurophysiol. 24:185, 1984. Hodgkin, A. L., and Katz, B.: The effect of temperature on the electrical activity of the giant axon of the squid. J. Physiol. (Lond.) 109:240, 1949. Hömberg, V., Reiners, K., and Toyka, K. V.: Reversible conduction block in human ischemic neuropathy after ergotamine abuse. Muscle Nerve 15:467, 1992. Hopf, H. C.: Forearm ulnar-to-median nerve anastomosis of sensory axons. Muscle Nerve 13:654, 1990.

Nerve Conduction and Needle Electromyography 155. Hopf, H. C.: Clinical implications of testing brainstem reflexes and corticobulbar connections in man. In Kimura, J., and Shibasaki, H. (eds.): Recent Advances in Clinical Neurophysiology. Amsterdam, Elsevier, p. 39, 1996. 156. Horowitz, S. H.: Correlation of near-nerve sural conduction and quantified sensory testing in patients with diabetic neuropathy. Muscle Nerve 18:1202, 1995. 157. Horowitz, S. H., and Krarup, C.: Conduction studies of the normal sural nerve. Muscle Nerve 15:374, 1992. 158. Howard, J. E., McGill, K. C., and Dorfman, L. J.: Age effects on properties of motor unit action potentials: ADEMG analysis. Ann. Neurol. 24:207, 1988. 159. Hudgson, P., Gardner-Medwin, D., Worsfold, M., et al.: Adult myopathy from glycogen storage disease due to acid maltase deficiency. Brain 91:435, 1968. 160. Hudson, A. J., Brown, W. F., and Gilbert, J. J.: The muscular pain-fasciculation syndrome. Neurology 28:1105, 1978. 161. Imaoka, H., Yorifuji, S., Takahashi, M., et al.: Improved inching method for the diagnosis and prognosis of carpal tunnel syndrome. Muscle Nerve 15:318, 1992. 162. Inaba, A., Yokota, T., Komori, T., and Hirose, K.: Proximal and segmental motor nerve conduction in the sciatic nerve produced by percutaneous high voltage electrical stimulation. Electroencephalogr. Clin. Neurophysiol. 101:100, 1996. 163. Infante, E., and Kennedy, W. R.: Anomalous branch of the peroneal nerve detected by electromyography. Arch. Neurol. 22:162–165, 1970. 164. International Federation of Societies for Electroencephalography and Clinical Neurophysiology: Recommendations for the Practice of Clinical Neurophysiology. Amsterdam, Elsevier, p. 143, 1983. 165. Irnai, K. D., Grabois, M., and Harvey, S. C.: Standardized technique for diagnosis of tarsal tunnel syndrome. Am. J. Phys. Med. 61:26, 1982. 166. Isaacs, H.: Continuous muscle fiber activity. J. Neurol. Neurosurg. Psychiatry 24:319, 1961. 167. Jablecki, C., and Knoll, D.: Fibrillation potentials and complex repetitive discharges [abstract]. Electroencephalogr. Clin. Neurophysiol. 50:242, 1980. 168. Jabre, J. F.: Ulnar nerve lesions at the wrist: new technique for recording from the sensory dorsal branch of the ulnar nerve. Neurology 30:873, 1980. 169. Jabre, J. F.: Surface recording of the H-reflex of the flexor carpi radialis. Muscle Nerve 4:435, 1981. 170. Johnson, E. W., and Fletcher, F. R.: Lumbosacral radiculopathy: review of l00 consecutive cases. Arch. Phys. Med. Rehabil. 61:321, 1981. 171. Junge, D.: Turns and averaging as static and dynamic measures of masseter EMG activity. Electromyogr. Clin. Neurophysiol. 33:11, 1993. 172. Kadson, D. L.: Cervical spondylotic myelopathy with reversible fasciculations in the lower extremities. Arch. Neurol. 34:774, 1977. 173. Kaji, R., Oka, N., Tsuji, S., et al.: Pathological findings at the site of conduction block in multifocal motor neuropathy. Ann. Neurol. 33:152, 1993. 174. Kanakamedala, R. V., and Hong, C.-Z.: Peroneal nerve entrapment at the knee localized by short segment stimulation. Am. J. Phys. Med. Rehabil. 68:116, 1989.

963

175. Kanakamedala, R. V., Simons, D. G., Porter, R. W., and Zucker, R. S.: Ulnar nerve entrapment at the elbow localized by short segment stimulation. Arch. Phys. Med. Rehabil. 69:959, 1988. 176. Kikuchi, S., Sato, K., Konno, S., and Hasue, M.: Anatomic and radiographic study of dorsal root ganglia. Spine 19:6, 1994. 177. Kimura, J.: Alteration of the orbicularis oculi reflex by pontine lesions: study in multiple sclerosis. Arch. Neurol. 22:156, 1970. 178. Kimura, J.: Electrodiagnostic study of brainstem strokes. Stroke 2:576, 1971. 179. Kimura, J.: An evaluation of the facial and trigeminal nerves in polyneuropathy: electrodiagnostic study in CharcotMarie-Tooth disease, Guillain-Barré syndrome, and diabetic neuropathy. Neurology 21:745, 1971. 180. Kimura, J.: The blink reflex as a test for brainstem and higher central nervous system functions. In Desmedt, J. E. (ed.): New Developments in Electromyography and Clinical Neurophysiology, Vol. 3. Basel, Karger, p. 682, 1973. 181. Kimura, J.: F-wave velocity in the central segment of the median and ulnar nerves: a study in normal subjects and in patients with Charcot-Marie-Tooth disease. Neurology (Minneap.) 24:539, 1974. 182. Kimura, J.: Electrically elicited blink reflex in diagnosis of multiple sclerosis: review of 260 patients over a seven-year period. Brain 98:413, 1975. 183. Kimura, J.: Collision technique—physiological block of nerve impulses in studies of motor nerve conduction velocity. Neurology (Minneap.) 26:680, 1976. 184. Kimura, J.: A method for estimating the refractory period of motor fibers in the human peripheral nerve. J. Neurol. Sci. 28:485, 1976. 185. Kimura, J.: Clinical value and limitations of F-wave determination: a comment [letter]. Muscle Nerve 1:250, 1978. 186. Kimura, J.: A method for determining median nerve conduction velocity across the carpal tunnel. J. Neurol. Sci. 38:1, 1978. 187. Kimura, J.: The carpal tunnel syndrome: localization of conduction abnormalities within the distal segment of the median nerve. Brain 102:619, 1979. 188. Kimura, J.: F-wave determination in nerve conduction studies. In Desmedt, J. E. (ed.): Motor Control Mechanisms in Health and Disease. New York, Raven Press, p. 961, 1983. 189. Kimura, J.: Principles and pitfalls of nerve conduction studies. Ann. Neurol. 16:415, 1984. 190. Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 2nd ed. Philadelphia, F.A. Davis, 1989. 191. Kimura, J.: Electromyography and Nerve Stimulation Techniques: Clinical Applications [in Japanese]. Tokyo, Igaku-Shoin, 1990. 192. Kimura, J.: Facts, fallacies, and fancies of nerve conduction studies: twenty-first annual Edward H. Lambert lecture. Muscle Nerve 20:777, 1997. 193. Kimura, J.: Kugelberg lecture: principles and pitfalls of nerve conduction studies. Electroencephalogr. Clin. Neurophysiol. 106:470, 1998.

964

Nerve Conduction and Electromyography

194. Kimura, J.: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, 3rd ed. New York, Oxford University Press, 2001. 195. Kimura, J.: Needle electromyography. In Bertorini, R. (ed.): Evaluation and Diagnostic Tests in Neuromuscular Disorders. Woburn, MA, Butterworth-Heinemann, p. 331, 2002. 196. Kimura, J., Bosch, P., and Lindsay, G. M.: F-wave conduction velocity in the central segment of the peroneal and tibial nerves. Arch. Phys. Med. Rehabil. 56:492, 1975. 197. Kimura, J., and Butzer, J. F.: F-wave conduction velocity in Guillain-Barré syndrome: assessment of nerve segment between axilla and spinal cord. Arch. Neurol. 32:524, 1975. 198. Kimura, J., Giron, L. T. Jr., and Young, S. M.: Electrophysiological study of Bell palsy: electrically elicited blink reflex in assessment of prognosis. Arch. Otolaryngol. 102:140, 1976. 199. Kimura, J., and Lyon, L. W.: Alteration of orbicularis oculi reflex by posterior fossa tumors. J. Neurosurg. 38:10, 1973. 200. Kimura, J., Machida, M., Ishida, T., et al.: Relation between size of compound sensory or muscle action potentials and length of nerve segment. Neurology 36:647, 1986. 201. Kimura, J., Murphy, J. M., and Varda, D. J.: Electrophysiological study of anomalous innervation of intrinsic hand muscles. Arch. Neurol. 33:842, 1976. 202. Kimura, J., Powers, J. M., and Van Allen, M. W.: Reflex response of orbicularis oculi muscle to supraorbital nerve stimulation: study in normal subjects and in peripheral facial paresis. Arch. Neurol. 21:193, 1969. 203. Kimura, J., Rodnitzky, R. L., and Van Allen, M. W.: Electrodiagnostic study of trigeminal nerve: orbicularis oculi reflex and masseter reflex in trigeminal neuralgia, paratrigeminal syndrome and other lesions of the trigeminal nerve. Neurology 20:574, 1970. 204. Kimura, J., Sakimura, Y., Machida, M., et al.: Effect of desynchronized inputs on compound sensory and muscle action potentials. Muscle Nerve 11:694, 1988. 205. Kimura, J., Yamada, T., and Stevland, N.: Distal slowing of motor nerve conduction velocity in diabetic polyneuropathy. J. Neurol. Sci. 42:291, 1979. 206. King, D., and Ashby, P.: Conduction velocity in the proximal segments of a motor nerve in the Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 39:538, 1976. 207. King, J. C., Dumitru, D., and Stegeman, D.: Monopolar needle electrode spatial recording characteristics. Muscle Nerve 19:1310, 1996. 208. Kingery, W. S., Wu, P. B. J., and Date, E. S.: An unusual presentation of a traumatic ulnar mononeuropathy with a Martin-Gruber anastomosis. Muscle Nerve 19:920, 1996. 209. Kohara, N., Kimura, J., Kaji, R., et al.: Multicenter analysis on intertrial variability of nerve conduction studies: healthy subjects and patients with diabetic polyneuropathy. In Kimura, J., and Shibasaki, H. (eds.): Recent Advances in Clinical Neurophysiology. Amsterdam, Elsevier, p. 809, 1996. 210. Kornhuber, M. E., Bischoff, C., Menthrup, H., and Conrad, B.: Multiple A waves in Guillain-Barré syndrome. Muscle Nerve 22:394, 1999. 211. Kothari, M. J., Macintosh, K., Heistand, M., and Logigian, E. L.: Medial antebrachial cutaneous sensory

212.

213.

214. 215.

216.

217.

218.

219.

220.

221.

222.

223.

224.

225.

226.

227. 228.

229.

230.

studies in the evaluation of neurogenic thoracic outlet syndrome. Muscle Nerve 21:647, 1998. Kothari, M. J., Preston, D. C., and Logigian, E. L.: Lumbrical-interossei motor studies localize ulnar neuropathy at the wrist. Muscle Nerve 19:170, 1996. Kraft, G. H.: Fibrillation potentials and positive sharp waves: are they the same? Electroencephalogr. Clin. Neurophysiol. 81:163, 1991. Kraft, G. H.: Are fibrillation potentials and positive sharp waves the same? No. Muscle Nerve 19:216, 1996. Kraft, G. H., and Halvorson, G. A.: Median nerve residual latency: normal value and use in diagnosis of carpal tunnel syndrome. Arch. Phys. Med. Rehabil. 64:221, 1983. Krarup, C., Horowitz, S. H., and Dahl, K.: The influence of the stimulus on normal sural nerve conduction velocity. Muscle Nerve 15:813, 1992. Krarup, C., and Trojaborg, W.: Sensory pathophysiology in chronic acquired demyelinating neuropathy. Brain 119:257, 1996. Kuruoglu, H. R., and Oh, S. J.: Tendon-reflex testing in chronic demyelinating polyneuropathy. Muscle Nerve 17:145, 1994. Kwast, O., Krajewska, G., and Kozlowski, K.: Analysis of F wave parameters in median and ulnar nerves in healthy infants and children: age related changes. Electromyogr. Clin. Neurophysiol. 24:439, 1984. Lambert, E. H.: Diagnostic value of electrical stimulation of motor nerves. Electroencephalogr. Clin. Neurophysiol. Suppl. 22:9, 1962. Lambert, E. H.: The accessory deep peroneal nerve: a common variation in innervation of extensor digitorum brevis. Neurology (Minneap.) 19:11696, 1969. Lang, A. H., and Partanen, V. S. J.: “Satellite potentials” and the duration of motor unit potentials in normal, neuropathic and myopathic muscles. J. Neurol. Sci. 27:513, 1976. Lang, H. A., Puusa, A., Hynninen, P., et al.: Evolution of nerve conduction velocity in later childhood and adolescence. Muscle Nerve 8:38, 1985. Lateva, Z. C., McGill, K. C., and Burgar, C. G.: Anatomical and electrophysiological determinants of the human thenar compound muscle action potential. Muscle Nerve 19:1457, 1996. Lauder, T. D., and Dillingham, T. R.: The cervical radiculopathy screen: optimizing the number of muscles studies [short report]. Muscle Nerve 19:662, 1996. Layzer, R. B.: Muscle pain, cramps, and fatigue. In Engel, A. G., and Franzini-Armstrong, C. (eds.): Myology, 2nd ed. New York, McGraw-Hill, pp. 3462, 1994. Layzer, R. B.: The origin of muscle fasciculations and cramps. Muscle Nerve 17:1243, 1994. Lee, H. J., Bach, J. R., and DeLisa, J. A.: Deep peroneal sensory nerve: standardization in nerve conduction study. Am. J. Phys. Med. Rehabil. 69:202, 1990. Lee, K. W., and Oh, S. J.: Early appearance of aging phenomenon in the interdigital nerves of the foot. Muscle Nerve 17:58, 1994. Lee Lim, C., Lal, H., and Yiannikas, C.: The effect of wrist size on the orthodromic median sensory nerve action potential. Muscle Nerve 18:117, 1995.

Nerve Conduction and Needle Electromyography 231. Lefebvre d’Amour, M., Shahani, B. T., Young, R. R., and Bird, K. T.: The importance of studying sural nerve conduction and late responses in the evaluation of alcoholic subjects. Neurology 29:1600, 1979. 232. Leibovic, S. J., and Hastings, H.: Martin-Gruber revisited. J. Hand Surg. Am. 17:47, 1992. 233. Lester, J. M., Soule, N. W., Bradley, W. G., and Brenner, J. F.: An augmented computer model of motor unit reorganization in neurogenic diseases. Muscle Nerve 16:43, 1993. 234. Levin, K. H.: L5 radiculopathy with reduced superficial peroneal sensory responses: intraspinal and extraspinal causes. Muscle Nerve 21:3, 1998. 235. Levin, K. H., Maggiano, H. J., and Wilbourn, A. J.: Cervical radiculopathies: comparison of surgical and EMG localization of single root lesions. Neurology 46:1022, 1996. 236. Lewis, R. A., and Sumner, A. J.: The electrodiagnostic distinctions between chronic familial and acquired demyelinative neuropathies. Neurology 32:592, 1982. 237. Liguori, R., Dahl, K., and Fuglsang-Frederiksen, A.: Turnsamplitude analysis of the electromyographic recruitment pattern disregarding force measurement. I. Method and reference values in healthy subjects. Muscle Nerve 15:1314, 1992. 238. Livingstone, E. F., and DeLisa, J. A.: Electrodiagnostic values through the thoracic outlet using C8 root needle studies, F-waves, and cervical somatosensory evoked potentials. Arch. Phys. Med. Rehabil. 65:726, 1984. 239. Lyon, L. W., and Van Allen, M. W.: Orbicularis oculi reflex: studies in internuclear ophthalmoplegia and pseudointernuclear ophthalmoplegia. Arch. Ophthalmol. 87:148, 1972. 240. Lysens, R., Vandendriessche, G., Van Mol, Y., and Rosselle, N.: The sensory conduction velocity in the cutaneous femoris lateralis nerve in normal adult subjects and in patients with complaints suggesting meralgia paresthetica. Electromyogr. Clin. Neurophysiol. 21:505, 1981. 241. Maccabee, P. J., Lipitz, M. E., Amassian, V. E., et al.: A new method using neuromagnetic stimulation to measure conduction time within the cauda equina. Electroencephalogr. Clin. Neurophysiol. 101:153, 1996. 242. MacLean, I. C.: Nerve root stimulation to evaluate conduction across the brachial and lumbosacral plexuses. In Third Annual Continuing Education Course, American Association of Electromyography and Electrodiagnosis, Philadelphia, September 25, 1980. 243. Magistris, M. R., and Roth, G.: Motor axon reflex and indirect double discharge: ephaptic transmission? A reappraisal. Electroencephalogr. Clin. Neurophysiol. 85:124, 1992. 244. Marinacci, A. A.: Diagnosis of “all median hand.” Bull. Los Angeles Neurol. Soc. 29:191, 1964. 245. Markand, O. N., Kincaid, J. C., Pourmand, R. A., et al.: Electrophysiologic evaluation of diaphragm by transcutaneous phrenic nerve stimulation. Neurology 34:604, 1984. 246. Marras, C., and Midroni, G. Y. L.: Proximal Martin-Gruber anastomosis mimicking ulnar neuropathy at the elbow. Muscle Nerve 22:1132, 1999. 247. Martin, R.: Tal om Nervers allmanna egenskapper i manniskans kropp. Stockholm, L. Salvius, 1763. 248. Martinez, A. C., Perez Conde, M. C., Del Campo, F., et al.: Ratio between the amplitude of sensory evoked potentials

249. 250. 251.

252.

253.

254.

255. 256. 257.

258.

259.

260.

261.

262.

263.

264.

265.

266.

267.

965

at the wrist in both hands of left-handed subjects. J. Neurol. Neurosurg. Psychiatry 43:182, 1980. Mateer, J. E., Gutmann, L., and McComas, C. F.: Myokymia in Guillain-Barré syndrome. Neurology 33:374, 1983. Mayer, R. F.: Nerve conduction studies in man. Neurology (Minneap.) 13:1021, 1963. McCluskey, L. F.: Anomalous superficial radial sensory innervation of the ulnar dorsum of the hand: a cause of “paradoxical” preservation of ulnar sensory nerve function. Muscle Nerve 19:923, 1996. McGill, K. C., and Lateva, Z. C.: The contribution of the interosseous muscles to the hypothenar compound muscle action potential. Muscle Nerve 22:6, 1999. McGonagle, I. K., Levine, S. R., Donofrio, P. D., and Albers, J. W.: Spectrum of patients with EMG features of polyradiculopathy without neuropathy. Muscle Nerve 13:63, 1990. Menkes, D. L., Hood, D. C., Ballesteros, R. A., and Williams, D. A.: Root stimulation improves the detection of acquired demyelinating polyneuropathies. Muscle Nerve 21:298, 1998. Miller, R., and Kuntz, N.: Nerve conduction studies in infants and children. J. Child Neurol. 1:19, 1986. Miller, R. G.: AAEE Minimonograph #28: Injury to peripheral motor nerves. Muscle Nerve 10:698, 1987. Misra, U. K., Tiwari, S., Shukla, N., et al.: F-response studies in neonates, infants and children. Electromyogr. Clin. Neurophysiol. 29:251, 1989. Mitsikostas, D. D., Karandreas, N., Coutsopetras, P., et al.: Fasciculation potentials in healthy people. Muscle Nerve 21:533, 1998. Mitsudome, A., Yasumoto, S., and Ogata, H.: Late responses in full-term newborn infants. In Kimura, J., and Shibasaki, H. (eds.): Recent Advances in Clinical Neurophysiology. Amsterdam, Elsevier, p. 761, 1996. Nakano, S., Engel, A. G., Waclawik, A. J., et al.: Myofibrillar myopathy with abnormal foci of desmin positivity. I. Light and electron microscopy analysis of 10 cases. J. Neuropathol. Exp. Neurol. 55:549, 1996. Nakashima, K., Tabuchi, Y., and Takahashi, K.: The diagnostic significance of large action potentials in myopathy. J. Neurol. Sci. 61:161, 1983. Nandedkar, S. D., Barkhaus, P. E., Sanders, D. B., and Stålberg, E. V.: Analysis of amplitude and area of concentric needle EMG motor unit action potentials. Electroencephalogr. Clin. Neurophysiol. 69:561, 1988. Nandedkar, S. D., Dumitru, D., and King, J. C.: Concentric needle electrode duration measurement and uptake area. Muscle Nerve 20:1225, 1997. Nandedkar, S. D., and Sanders, D. B.: Principal component analysis of the features of concentric needle EMG motor unit action potentials. Muscle Nerve 12:288, 1989. Nandedkar, S. D., Sanders, D. B., and Stålberg, E. V.: On the shape of the normal turns-amplitude cloud. Muscle Nerve 14:8, 1991. Nandedkar, S. D., Sanders, D. B., Stålberg, E. V., and Andreassen, S.: Simulation of concentric needle EMG motor unit action potential. Muscle Nerve 11:151, 1988. Nardin, R. A., Patel, M. R., Gudas, T. F., et al.: Electromyography and magnetic resonance imaging in

966

268.

269.

270. 271.

272.

273.

274.

275.

276.

277.

278.

279.

280.

281.

282.

283.

284.

Nerve Conduction and Electromyography the evaluation of radiculopathy. Muscle Nerve 22:151, 1999. Nelson, C., White, J. A., Mitchell, R. U., and Hall, C. D.: Median nerve F wave conduction in healthy subjects over age sixty-five. Electromyogr. Clin. Neurophysiol. 30:269, 1990. Netter, F. H.: The Ciba Collection of Medical Illustrations. Vol. 1: Nervous System. Part 1: Anatomy and Physiology. Summit, NJ, Ciba Pharmaceutical Co., Medical Education Division, 1983. Neundorfer, B., and Seiberth, R.: The accessory deep peroneal nerve. J. Neurol. 209:125, 1975. Newsom-Davis, J.: Immunological associations of acquired neuromyotonia (Isaacs’ syndrome): report of five cases and literature review. Brain 116:453, 1993. Nora, L. M.: American Association of Electrodiagnostic Medicine guidelines in electrodiagnostic medicine: implanted cardioverters and defibrillators. Muscle Nerve 19:1359, 1996. Normand, M. M., and Daube, J. R.: Interaction of random electromyographic activity with averaged sensory evoked potentials. Neurology 42:1605, 1992. Norris, A. H., Shock, N. W., and Wagman, I. H.: Age changes in the maximum conduction velocity of motor fibers of human ulnar nerves. J. Appl. Physiol. 5:589, 1953. Notermans, N. C., Franssen, H., Wieneke, G. H., and Wokke, J. H.: Temperature dependence of nerve conduction and EMG in neuropathy associated with gammopathy. Muscle Nerve 17:516, 1994. Obi, T., Mizoguchi, K., Matsuoka, H., et al.: Muscle cramp as the result of impaired GABA function—an electrophysiological and pharmacological observation. Muscle Nerve 16:1228, 1993. Odabasi, Z., Oh, S. J., Claussen, G. C., and Kim, D. S.: New near-nerve needle nerve conduction technique: differentiating epicondylar from cubital tunnel ulnar neuropathy. Muscle Nerve 22:718, 1999. Oh, S. J., Claussen, G. C., and Ahmad, B. K.: Double anastomosis of median-ulnar and ulnar-median nerves: report of an electrophysiologically proven case. Muscle Nerve 18:1332, 1995. Oh, S. J., Kim, D. E., and Kuruoglu, H. R.: What is the best diagnostic index of conduction block and temporal dispersion? Muscle Nerve 17:489, 1994. Olney, R. K., and Wilbourn, A. J.: Ulnar nerve conduction study of the first dorsal interosseous muscle. Arch. Phys. Med. Rehabil. 66:16, 1985. Olsewski, J. M., Simmons, E. H., Kallen, F. C., and Mendel, F. C.: Evidence from cadavers suggestive of entrapment of fifth lumbar spinal nerves by lumbosacral ligaments. Spine 16:336, 1991. Ongerboer de Visser, B. W., and Goor, C.: Electromyographic and reflex study in idiopathic and symptomatic trigeminal neuralgias: latency of the jaw and blink reflexes. J. Neurol. Neurosurg. Psychiatry 37:1225, 1974. Ongerboer de Visser, B. W., and Goor, C.: Cutaneous silent period in masseter muscles: a clinical and electrodiagnostic evaluation. J. Neurol. Neurosurg. Psychiatry 39:674, 1976. Panayiotopoulos, C. P.: F-wave conduction velocity in the deep peroneal nerve: Charcot-Marie-Tooth disease and dystrophia myotonica. Muscle Nerve 1:37, 1978.

285. Panayiotopoulos, C. P., and Chroni, E.: F-wave in clinical neurophysiology: a review, methodological issues and overall value in peripheral neuropathies. Electroencephalogr. Clin. Neurophysiol. 101:365, 1996. 286. Panayiotopoulos, C. P., and Lagos, G.: Tibial nerve H-reflex and F-wave studies in patients with uremic neuropathy. Muscle Nerve 3:423, 1980. 287. Payan, L.: The blanket principle: a technical note. Muscle Nerve 1:432, 1978. 288. Pease, W. S., Cunningham, M. L., Walsh, W. E., and Johnson, E. W.: Determining neurapraxia in carpal tunnel syndrome. Am. J. Phys. Med. Rehabil. 66:117, 1988. 289. Pease, W. S., Lagattuta, F. P., and Johnson, E. W.: Spinal nerve stimulation in S1 radiculopathy. Am. J. Phys. Med. Rehabil. 69:77, 1990. 290. Pedinoff, S., and Sandhu, R. S.: Electromyographic effect on serum creatine phosphokinase in normal individuals. Arch. Phys. Med. Rehabil. 59:27, 1978. 291. Peioglou-Harmoussi, S., Fawcett, P. R. W., Howel, D., and Barwick, D. D.: F-responses in syringomyelia. J. Neurol. Sci. 75:293, 1986. 292. Penders, C., and Boniver, R.: Exploration electrophysiologique du reflexe de clignement dans la paralysie faciale a frigore. J. Otolaryngol. 34:17, 1972. 293. Perotto, A., Delagi, E. F., Iazzetti, J., and Morrison, D.: Anatomical Guide for the Electromyographer: The Limbs and Trunk, 3rd ed. Springfield, IL, Charles C Thomas, 1994. 294. Petajan, J. H.: Antigravity posture for analysis of motor unit recruitment: the “45 degree test”. Muscle Nerve 13:355, 1990. 295. Petajan, J. H.: AAEM Minimonograph #3: Motor unit recruitment. Muscle Nerve 14: 489, 1991. 296. Peterson, A. R., Giuliani, M. J., McHugh, M., and Shipe, C. C.: Variations in dorsomedial hand innervation: electrodiagnostic implications. Arch. Neurol. 49:870, 1992. 297. Petrera, J. E., and Trojaborg, W.: Conduction studies along the accessory nerve and follow-up of patients with trapezius palsy. J. Neurol. Neurosurg. Psychiatry 47:630, 1984. 298. Petrera, J. E., and Trojaborg, W.: Conduction studies of the long thoracic nerve in serratus anterior palsy of different etiology. Neurology (Cleve.) 34:1033, 1984. 299. Pezzin, L. E., Dillingham, T. R., Lauder, T. D., et al.: Cervical radiculopathies: relationship between symptom duration and spontaneous EMG activity. Muscle Nerve 22:1412, 1999. 300. Phillips, L. H. II: Pitfalls in nerve conduction studies: temperature and distance measurement. In Kimura, J., and Shibasaki, H. (eds.): Recent Advances in Clinical Neurophysiology. Amsterdam, Elsevier, p. 674, 1996. 301. Phillips, L. H. II, and Park, T. S.: Electrophysiologic mapping of the segmental anatomy of the muscles of the lower extremity. Muscle Nerve 14:1213, 1991. 302. Pickett, J. B.: Late components of motor unit potentials in a patient with myoglobinuria. Ann. Neurol. 3:461, 1978. 303. Preston, D. C., and Logigian, E. L.: Lumbrical and interossei recording in carpal tunnel syndrome. Muscle Nerve 15:1253, 1992. 304. Raynor, E. M., Preston, D. C., and Logigian, E. L.: Influence of surface recording electrode placement on nerve action potentials. Muscle Nerve 20:361, 1997.

Nerve Conduction and Needle Electromyography 305. Redford, J. W. B.: Conduction time in motor fibers of nerves which innervate proximal muscles of the extremities in normal persons and in patients with neuromuscular diseases. Thesis, University of Minnesota, Minneapolis, 1958. 306. Renshaw, B.: Influence of discharge of motoneurons upon excitation of neighboring motoneurons. J. Neurophysiol. 4:167, 1941. 307. Richardson, A. T.: Clinical and electromyographic aspects of polymyositis. Proc. R. Soc. Med. 49:111, 1956. 308. Rivner, M. H., Swift, T. R., Crout, B. O., and Rhodes, K. P.: Toward more rational nerve conduction interpretations: the effect of height. Muscle Nerve 13:232, 1990. 309. Robinson, L. R., Rubner, D. E., Wahl, P. W., et al.: Factor analysis. Am. J. Phys. Med. Rehabil. 71:22, 1992. 310. Robinson, R. O., and Robertson, W. C. Jr.: Fetal nutrition and peripheral nerve conduction velocity. Neurology 31:327, 1981. 311. Ropper, A. H., Wijidicks, E. F. M., and Shahani, B. T.: Electrodiagnostic abnormalities in 113 consecutive patients with Guillain-Barre syndrome. Arch. Neurol. 47:884, 1990. 312. Rosenfalck, A.: Early recognition of nerve disorders by near-nerve recording of sensory action potentials. Muscle Nerve 1:360, 1978. 313. Ross, M. A., and Kimura, J.: AAEM Case Report #2: The carpal tunnel syndrome. Muscle Nerve 18:567, 1995. 314. Roth, G.: Vitesse de conduction motrice du nerf median dans le canal carpien. Ann. Med. Physique 13:117, 1970. 315. Roth, G.: Reinnervation dans la paralysie plexulaire brachiale obstetricale. J. Neurol. Sci. 58:103, 1983. 316. Roth, G.: Repetitive discharge due to self-ephaptic excitation of a motor unit. Electroencephalogr. Clin. Neurophysiol. 93:1, 1994. 317. Roth, G., and Egloff-Baer, S.: ECG-related fasciculation potential. Electroencephalogr. Clin. Neurophysiol. 105:132, 1997. 318. Roth, G., and Magistris, M. D.: Ephapse between two motor units in chronically denervated muscle. Electromyogr. Clin. Neurophysiol. 25:331, 1985. 319. Rowin, J., and Meriggioli, M. N.: Complex repetitive discharges: cause or effect of neurogenic muscle hypertrophy? Muscle Nerve 22:1603, 1999. 320. Rutkove, S. B., Kothari, M. J., Raynor, E. M., et al.: Sural/radial amplitude ratio in the diagnosis of mild axonal polyneuropathy. Muscle Nerve 20:1236, 1997. 321. Rutkove, S. B., Kothari, M. J., and Shefner, J. M.: Nerve, muscle and neuromuscular junction electrophysiology at high temperature. Muscle Nerve 20:431, 1997. 322. Salerno, D. F., Franzblau, A., Werner, R. A., et al.: Median and ulnar nerve conduction studies among workers: normative values. Muscle Nerve 21:999, 1998. 323. Sander, H. W., Quinto, C., and Chokroverty, S.: Accessory deep peroneal neuropathy: collision technique diagnosis. Muscle Nerve 21:121, 1998. 324. Sawhney, B. B., and Kayan, A.: A study of the F wave from the facial muscles [abstract]. Electroencephalogr. Clin. Neurophysiol. 30:261, 1971. 325. Scranton, P. E., Hasiba, U., and Gorenc, T. J.: Intramuscular hemorrhage in hemophiliacs with inhibitors. JAMA 241:2028, 1979.

967

326. Scutter, S. D., and Turker, K. S.: Recruitment stability in masseter motor units during isometric voluntary contractions. Muscle Nerve 21:1290, 1998. 327. Seror, P.: Orthodromic inching test in mild carpal tunnel syndrome. Muscle Nerve 21:1206, 1998. 328. Shahani, B. T., and Young, R. R.: The blink, H, and tendon vibration reflexes. In Goodgold, J., and Eberstein, A. (eds.): Electrodiagnosis of Neuromuscular Diseases, 2nd ed. Baltimore, Williams & Wilkins, p. 245, 1977. 329. Shahani, B. T., Young, R. R., Potts, F., and Maccabee, P.: Terminal latency index and late response studies in motor neuron disease, peripheral neuropathies and entrapment syndromes. Acta Neurol. Scand. Suppl. 73:118, 1979. 330. Sheean, G. L., Houser, M. K., and Murray, N. M. F.: Lumbrical-interosseous latency comparison in the diagnosis of carpal tunnel syndrome. Electroencephalogr. Clin. Neurophysiol. 97:285, 1995. 331. Sheean, G. L., Kanabar, G., and Murray, N. M. F.: Lumbrical-interosseous comparison in a distal ulnar nerve lesion. Muscle Nerve 19:673, 1996. 332. Shefner, J. M., Buchthal, F., and Krarup, C.: Slowly conducting myelinated fibers in peripheral neuropathy. Muscle Nerve 14:534, 1991. 333. Shochina, M., Vatine, J. J., Mahler, Y., et al.: Diagnostic value of computer analysis of multipeaked EMG spikes. Electromyogr. Clin. Neurophysiol. 32:113, 1992. 334. Simpson, J. A.: Handbook of Electromyography and Clinical Neurophysiology, Vol. 16, Neuromuscular Diseases. Amsterdam, Elsevier, 1973. 335. Smit, B. J., Kok, J. H., De Vries, L. S., et al.: Motor nerve conduction velocity in very preterm infants. Muscle Nerve 22:372, 1999. 336. Smyth, D. P. L., and Willison, R. G.: Quantitative electromyography in babies and young children with no evidence of neuromuscular disease. J. Neurol. Sci. 56:209, 1982. 337. Sonoo, M., and Stålberg, E.: The ability of MUP parameters to discriminate between normal and neurogenic MUPs in concentric EMG: analysis of the MUP “thickness” and the proposal of “size index”. Electroencephalogr. Clin. Neurophysiol. 89:291, 1993. 338. Soudmand, R., Ward, L. C., and Swift, T. R.: Effect of height on nerve conduction velocity. Neurology 32:407, 1982. 339. Spiro, A. J., Shy, G. M., and Gonatas, N. K.: Myotubular myopathy: persistence of fetal muscle in an adolescent boy. Arch. Neurol. 14:1, 1966. 340. Srinivasan, R., and Rhodes, J.: The median-ulnar anastomosis (Martin-Gruber) in normal and congenitally abnormal fetuses. Arch. Neurol. 38:418, 1981. 341. Stålberg, E.: Invited review: electrodiagnostic assessment and monitoring of motor unit changes in disease. Muscle Nerve 14:293, 1991. 342. Stålberg, E., and Ekstedt, J.: Single fiber EMG and microphysiology of the motor unit in normal and diseased human muscle. In Desmedt, J. E. (ed.): New Developments in Electromyography and Clinical Neurophysiology, Vol. 1. Basel, Karger, p. 113, 1973. 343. Stålberg, E., and Falck, B.: The role of electromyography in neurology. Electroencephalogr. Clin. Neurophysiol. 103:579, 1997.

968

Nerve Conduction and Electromyography

344. Stålberg, E., Nandedkar, S. D., Sanders, D. B., and Falck, B.: Quantitative motor unit potential analysis. J. Clin. Neurophysiol. 13:401, 1996. 345. Stålberg, E., and Sonoo, M.: Assessment of variability in the shape of the motor unit action potential, the “jiggle,” at consecutive discharges. Muscle Nerve 17:1135, 1994. 346. Stålberg, E., and Trontelj, J. V.: Single Fiber Electromyography. Old Woking, Surrey, UK, Mirvalle Press, 1979. 347. Stashuk, D. W.: Detecting single fiber contributions to motor unit action potentials. Muscle Nerve 22:218, 1999. 348. Stetson, D. S., Albers, J. W., Silverstein, B. A., and Wolfe, R. A.: Effects of age, sex, and anthropometric factors on nerve conduction measures. Muscle Nerve 15:1095, 1992. 349. Sumner, A. J.: Consensus criteria for the diagnosis of partial conduction block and multifocal motor neuropathy. In Kimura, J., and Kaji, R. (eds.): Physiology of ALS and Related Diseases. Amsterdam, Elsevier, p. 221, 1997. 350. Tanenbaum, D. I., and Jabre, J. F.: New method for expressing F-wave data as conduction velocity. J. Electromyogr. Kinesiol. 1:68, 1991. 351. Tani, T., Yamamoto, H., Ichimiya M., and Kimura, J.: Reflexes evoked in human erector spinae muscles by tapping during voluntary activity. Electroencephalogr. Clin. Neurophysiol. 105:194, 1997. 352. Taylor, N., Jebsen, R. H., and Tenckoff, H. A.: Facial nerve conduction latency in chronic renal insufficiency. Arch. Phys. Med. Rehabil. 51:259, 1970. 353. Taylor, P. K.: Nonlinear effects of age on nerve conduction in adults. J. Neurol. Sci. 66:223, 1984. 354. Thesleff, S.: Fibrillation in denervated mammalian skeletal muscle. In Culp, W. J., and Ochoa, J. (eds.): Abnormal Nerves and Muscle as Impulse Generators. Oxford, UK, Oxford University Press, 1982, p. 678, 1982. 355. Thiele, B., and Bohle, A.: Number of spike-components contributing to the motor unit potential. Z. EEG EMG 9:125, 1978. 356. Thomas, C. K., Broton, J. G., and Calancie, B.: Motor unit forces and recruitment patterns after cervical spinal cord injury. Muscle Nerve 20:212, 1997. 357. Thomas, J. E., and Lambert, E. H.: Ulnar nerve conduction velocity and H-reflex in infants and children. J. Appl. Physiol. 15:1, 1960. 358. Thomas, P. K., Sears, T. A., and Gilliatt, R. W.: The range of conduction velocity in normal motor nerve fibres to the small muscles of the hand and foot. J. Neurol. Neurosurg. Psychiatry 22:175, 1959. 359. Tonzola, R. F., Ackil, A. A., Shahani, B. T., and Young, R. R.: Usefulness of electrophysiological studies in the diagnosis of lumbosacral root disease. Ann. Neurol. 9:305, 1981. 360. Torbergsen, T., Stålberg, E., and Brautaset, N. J.: Generator sites for spontaneous activity in neuromyotonia: an EMG study. Electroencephalogr. Clin. Neurophysiol. 101:69, 1996. 361. Trojaborg, W.: Motor and sensory conduction in the musculocutaneous nerve. J. Neurol. Neurosurg. Psychiatry 39:890, 1976. 362. Trojaborg, W.: Quantitative electromyography in polymyositis: a reappraisal. Muscle Nerve 13:964, 1990.

363. Trojaborg, W.: Clinical, electrophysiological, and myelographic studies of 9 patients with cervical spinal root avulsions: discrepancies between EMG and x-ray findings. Muscle Nerve 17:913, 1994. 364. Trojaborg, W., and Buchthal, F.: Malignant and benign fasciculations. Acta Neurol. Scand. Suppl. 13:251, 1965. 365. Trojaborg, W., Grewal, R. P., and Sheriff, P.: Value of latency measurements to the small palm muscles compared to other conduction parameters in the carpal tunnel syndrome. Muscle Nerve 19:243, 1996. 366. Trojaborg, W. T., Moon, A., Andersen, B. B., and Trojaborg, N. S.: Sural nerve conduction parameters in normal subjects related to age, gender, temperature, and height: a reappraisal. Muscle Nerve 15:666, 1992. 367. Truong, X. T., Russo, F. I., Vagi, I., and Rippel, D. V.: Conduction velocity in the proximal sural nerve. Arch. Phys. Med. Rehabil. 60:304, 1979. 368. Tullberg, T., Svanborg, E., Isacsson, J., and Grane, P.: A preoperative and postoperative study of the accuracy and value of electrodiagnosis in patients with lumbosacral disc herniation. Spine 18:837, 1993. 369. Uchida, Y., and Sugioka, Y.: Electrodiagnosis of MartinGruber connection and its clinical importance in peripheral nerve surgery. J. Hand Surg. Am. 17:54, 1992. 370. Uncini, A., Di Muzio, A., Sabatelli, M., et al.: Sensitivity and specificity of diagnostic criteria for conduction block in chronic inflammatory demyelinating polyneuropathy. Electroencephalogr. Clin. Neurophysiol. 89:161, 1993. 371. Valensi, P., Attali, J.-R., and Gagant, S.: Reproducibility of parameters for assessment of diabetic neuropathy. Diabet. Med. 10:933, 1993. 372. Valls-Sole, J.: Martin-Gruber anastomosis and unusual sensory innervation of the fingers: report of a case. Muscle Nerve 14:1099, 1991. 373. Van Dijk, J. G., Van Benten, I., Kramer, C. G. S., and Stegeman, D. F.: CMAP amplitude cartography of muscles innervated by the median, ulnar, peroneal and tibial nerves. Muscle Nerve 22:378, 1999. 374. Wang, A. K., Raynor, E. M., Blam, A. S., and Rutkove, S. B.: Heat sensitivity of sensory fibers in carpal tunnel syndrome. Muscle Nerve 22:37, 1999. 375. Wang, F. C., de Pasqua, V., and Delwaide, P. J.: Age-related changes in fastest and slowest conducting axons of thenar motor units. Muscle Nerve 22:1022, 1999. 376. Warmolts, J. R., and Mendell, J. R.: Open-biopsy electromyography: direct correlation of a pattern of excessively recruited, pathologically small motor unit potentials with histologic evidence of neuropathy. Arch. Neurol. 36:406, 1979. 377. Waylonis, G. W., and Johnson, E. W.: Facial nerve conduction delay. Arch. Phys. Med. Rehabil. 45:539, 1964. 378. Waylonis, G. W., and Johnson, E. W.: Electromyographic findings in induced trichinosis. Arch. Phys. Med. Rehabil. 46:615, 1965. 379. Weiss, M. D., and Mayer, R. F.: Temperature-sensitive repetitive discharges in paramyotonia congenita. Muscle Nerve 20:195, 1997. 380. Wettstein, A.: The origin of fasciculations in motoneuron disease. Ann. Neurol. 5:295, 1979.

Nerve Conduction and Needle Electromyography 381. Wilbourn, A. J., and Aminoff, M. J.: AAEM Minimonograph #32: The electrodiagnostic examination in patients with radiculopathies. Muscle Nerve 21:1612, 1998. 382. Wilbourn, A. J., and Lambert, E. H.: The forearm medianto-ulnar nerve communication: electrodiagnostic aspects. Neurology (Minneap.) 26:368, 1976. 383. Willison, R. G.: Analysis of electrical activity in healthy and dystrophic muscle in man. J. Neurol. Neurosurg. Psychiatry 27:386, 1964. 384. Winer, B. J.: Statistical Principles in Experimental Design, 2nd ed. New York, McGraw-Hill, p. 283, 1972. 385. Winkler, T., Stalberg, E., and Haas, L. F.: Uni- and bipolar surface recording of human nerve responses. Muscle Nerve 14:133, 1991.

969

386. Wulff, C. H., and Gilliatt, R. W.: F waves in patients with hand wasting caused by a cervical rib and band. Muscle Nerve 2:452, 1979. 387. Young, A. W., Redmond, M. D., Hemler, D. E., and Belandres, P. V.: Radial motor nerve conduction studies. Arch. Phys. Med. Rehabil. 71:399, 1990. 388. Young, R. R., and Shahani, B. T.: Clinical value and limitations of F-wave determination [letter]. Muscle Nerve 1:248, 1978. 389. Zalewska, E., Rowinska-Marcinska, K., and HausmanowaPetrusewicz, I.: Shape irregularity of motor unit potentials in some neuromuscular disorders. Muscle Nerve 21:1181, 1998. 390. Zhu, Y., Starr, A., Haldeman, S., et al.: Soleus H-reflex to S1 nerve root stimulation. Electroencephalogr. Clin. Neurophysiol. 109:10, 1998.

36 Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores PETER J. DYCK, PETER C. O’BRIEN, JENNY L. DAVIES, CHRISTOPHER J. KLEIN, AND P. JAMES B. DYCK

The Need for Normal Values Percentiles, Normal Deviates, and Composite Scores Nerve Tests for Which Reference Values Can Be Set Selection of a Reference Cohort Cost-Effectiveness of Prospectively Assessed Normal Controls Statistical Methods for Computing Normal Values 1. A Single Dependent Variable 2. A Single Independent Variable

3. Multiple Explanatory Variables 4. Multiple Independent Variables and Multiple Dependent Variables Variables Influencing Neuropathic End Points in a Normal Subject Cohort: The RDNS-NS Composite Scores: Five Attributes of Nerve Conduction, Seven Tests, NIS(LL) ⫹ 7 Tests, and Other Composite Scores Individual Attributes Versus Composite Scores of Nerve Conduction

The advantages of expressing test results as a percentile (or as a normal deviate) and as corrected for all applicable variables, rather than as “normal” or “abnormal,” are the following: (1) it permits all known applicable variables influencing test results to be taken into account in expressing abnormality; (2) it allows the user to select different percentile levels of abnormality depending on the purposes for obtaining the test results; (3) it permits direct comparison of different attributes of nerve conduction and of different test results; and (4) it permits addition of different components of nerve tests (e.g., attributes of nerve conduction) and of different nerve tests (e.g., NIS[LL] ⫹ Seven Tests normal deviates). The disadvantages are that (1) a large, randomly chosen, healthy subject cohort must be recruited and appropriately studied, the effect of variables must be estimated, and software programs must be written to facilitate use of the data; (2) electromyographers cannot use a fixed value as the normal limit for attributes of nerve conduction; and (3) the approach appears complex.6,18 The advantages of use of percentiles clearly outweigh the disadvantages. Perhaps more important, when it is

Abnormality: Sensitivity, Reproducibility, Representativeness, and Concordance Functional Abnormality Even When Test Values Fall within the Range of Normal Values Composite Scores as Measures of Severity of Neuropathy over Time: Epidemiologic Surveys and Controlled Clinical Trials Composite Scores in Genetic Linkage Analysis

appreciated that most normal test results are influenced by different combinations of age, sex, height, and weight (or body mass index or surface area), it is no longer reasonable to use a single limit of abnormality based on age alone. Also, expense and complexity are actually not valid arguments against use of percentiles because the approach is not that complex, and in any case results can be provided by use of a software program taking age, sex, height, weight, and other measured values into account. Test results can then be provided as a percentile (or normal deviate) value. Use of percentiles (or normal deviates) allows comparisons and combinations of test results.2 Thus, for a given person, it is possible to compare abnormality of one attribute of nerve conduction against that of another or against another test result (e.g., vibration detection threshold).3,5,9,15 A series of similar or dissimilar test results can be combined into a composite score. Use of such composite scores to represent overall severity of a patient’s polyneuropathy makes sense because neuropathies such as diabetic polyneuropathy or chronic inflammatory 971

972

Nerve Conduction and Electromyography

demyelinating polyradiculoneuropathy (CIDP) are the sum total of many different symptoms, impairments, and test results. We advocate use of such composite scores for detection and follow-up of severity of neuropathy. They are especially valuable for epidemiologic studies, intervention trials, and following the course of disease in medical practice. Percentile values even within the range of normal may, in certain circumstances, be shown to indicate abnormal or improved or worsened nerve function.

THE NEED FOR NORMAL VALUES As noted by O’Brien and Dyck, “Normal values occupy an important place in medical practice.”18 Normal values play an important role in differential diagnosis, in recognizing impairment or worsening of disease, in determining need for further investigation and whether to treat, and to recognize responsiveness to treatment. “Frequently, however, the quality of the available reference values [is] far from ideal, mainly because of unrepresentative selection of healthy subjects, inadequate sample size, lack of medical evaluation of subjects, failure to also measure other variables that influence the attribute studied, and failure to rigorously apply appropriate statistical analysis.”18 Most physicians think of normal values as providing a dichotomous answer to the question of whether a test value is normal or abnormal. In fact, as we discuss below, the level of abnormality set is arbitrary and depends on the purpose for which the test is obtained. It should also be appreciated that such study of nerve conduction may provide multiple answers (as many as 12 to 30) to the question, “Are nerve conductions normal or abnormal?” Some of the values may appear to be normal, others borderline, and still others abnormal. How can these disparate values be combined into an overall meaningful assessment of normal or abnormal? While it is not unreasonable to argue that the electromyographer combines this disparate information into an overall judgment of normal or abnormal and whether it fits a diagnostic pattern, use of composite scores can provide an overall quantitative answer and be used to bolster judgments. In specialized cases, study of percentiles (or normal deviates) even when they are within the range of normal values can provide information on abnormality and on worsening or improvement.

PERCENTILES, NORMAL DEVIATES, AND COMPOSITE SCORES The term percentile refers to the ranked position of a measurement relative to a reference population. For example, a normal range that extends to the 95th percentile implies that 95% of subjects in the reference population of healthy

subjects have values at or below that value. As noted above, the reference group needs to be carefully defined and randomly chosen so as not to be biased. The “normal deviate” (or Z score) is a defined position in a “normal probability” distribution having the classic “bell-shaped” distribution with a mean of 1 and a standard deviation (SD) of 1. The 50th percentile value has a normal deviate value of 0. Normal deviate values for the 1st, 2.5th, 5th, 95th, 97.5th, and 99th percentiles are ⫺2.326, ⫺1.960, ⫺1.645, 1.645, 1.960, and 2.326. Others1,2,7,14,16,19–21 and we2–5,8–11 have introduced and given reasons for using composite scores of nerve conduction and clinical impairment. In this chapter, we discuss only scores based on use of summated normal deviate values. How are percentile values used to derive a composite score? As an example, the ⌺ 5NC, or the summated five attributes of nerve conduction for motor fibers of the peroneal nerve (compound muscle action potential [CMAP], motor nerve conduction velocity [MNCV], and motor nerve distal latency [MNDL]), the tibial nerve (MNDL) and the sural nerve (sensory nerve action potential [SNAP]), is calculated as follows: Step 1. Using a software program, express the percentile values for each attribute of nerve conduction to be included, considering the attribute, age, sex, and applicable variables. Step 2. Express abnormality in the same tail of the normal distribution (i.e., CMAP, MNCV, and SNAP values are changed from 1st to 99th, 2.5th to 97.5th, and so on). Step 3. Express all percentile values as normal deviates. Step 4. Summate all normal deviates of measurable attributes (when CMAP is 0 [a measured value], MNCV and MNDL are unmeasurable and cannot be included). Step 5. Divide by the number of attributes that were measurable and multiply by the total number of attributes in the composite score. To determine abnormality of the composite score, the composite values of persons in the normal subject cohort are regressed on age and, using the regression line, the line providing the 97.5th percentile is obtained empirically by increasing the intercept.

NERVE TESTS FOR WHICH REFERENCE VALUES CAN BE SET Reference values are needed for attributes of nerve conduction, modalities of quantitative sensation, and tests of autonomic nerve function. Similarly, normal values can be set for standard assessments of neuropathic symptoms and impairments. For many risk covariates of disease, normal values are also needed.

Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores

SELECTION OF A REFERENCE COHORT For a laboratory performing thousands of tests of a given type, it is important to have highly developed and adequate reference values for the tests offered. Tests are used to detect abnormality that may suggest a disease process (e.g., low hemoglobin, raised fasting plasma glucose, or low nerve conduction, the latter being indicative of polyneuropathy). The first requirement is that the reference population be representative of the community for which it will serve as reference values. Because selection should be unbiased, patients should be randomly selected from the community; all ages and both sexes should be included. Because patients are allowed not to participate, selection bias might occur. Therefore, if possible, the degree to which the reference population who agreed to participate is representative of the community must be determined. The second requirement is that persons who have the disease for which reference test values are being established should be excluded. For reference values of nerve conduction, quantitative sensation tests, and quantitative autonomic tests, persons with neurologic disease or diseases predisposing to neuropathy should be excluded. To do this, a neurologic examination and measures of fasting plasma glucose, and perhaps a few additional tests, need to be done to rule out participation of persons with diabetes mellitus, uremia, hypothyroidism, and so on. Additional individuals may need to be excluded based on a very abnormal test result leading to the diagnosis of a neurologic disease (e.g., Charcot-Marie-Tooth disease type 1A or 1B). The number of persons needed for an adequate healthy subject reference cohort is an important consideration. So as to be able to express abnormality to the 1st or 99th percentile, a cohort of 300 persons is recommended and 100 should be considered a minimum sample size. In the Rochester Diabetic Neuropathy Study of Normal Subjects (RDNS-NS), we found that 431 persons were needed to get a cohort of 330 subjects without neurologic disease or diseases predisposing to neuropathy. It was increasingly difficult to recruit persons without confounding disease in older age groups.

COST-EFFECTIVENESS OF PROSPECTIVELY ASSESSED NORMAL CONTROLS Because prospective study of large cohorts is expensive, it is necessary to decide whether the improved information coming from such studies is cost-effective. A large expense may be justified when the values will serve for thousands of subsequent patient examinations and for a period of years. The study of normal subjects in Rochester, Minnesota (the RDNS-NS), was used to develop normal

973

values for attributes of nerve conduction, quantitative sensation tests, and quantitative autonomic tests. The results of these studies are applicable not only to persons of northern European extraction living in Minnesota but also to millions of North Americans and Europeans having a similar ethnic and lifestyle background. Whether these reference values can also be used for southern Europeans, Native Americans, Hispanic Americans, Asians, or other geographic or ethnic cohorts is unknown.

STATISTICAL METHODS FOR COMPUTING NORMAL VALUES* This section is quoted from O’Brien and Dyck.18

1. A Single Dependent Variable In considering the statistical techniques useful in obtaining normal values, we start with the simplest situation: only one characteristic (or, in statistical language, one dependent variable) for which reference values are desired, and no explanatory variables (independent variables) such as age or gender, are to be considered. In the past, interest has centered on the 2.5th and 97.5th percentiles, commonly referred to as a normal range. However, in view of the wide availability of computers, there is no need to restrict attention to the resulting dichotomy of normal/abnormal. It is more informative to indicate the actual percentile attained by an individual patient, and computing each of the percentiles (from 1 through 99) is easily done with a computer. Notationally, we will refer to the Pth percentile for a variable Y as Y (P), for P ⫽ 1, . . . , 99. Notice that Y(P) is defined as the value such that P% of values in the reference population are less than Y(P). To estimate Y(P) from a study population, one simply selects the corresponding percentile in the set of study values. This method is simple and universally valid, requiring no esoteric mathematical assumptions. Notice, however, that large sample sizes are required to estimate very large or very small percentiles. This is not unreasonable. For example, common sense would indicate that, if we are to estimate the 99th percentile (that value below which 99% of persons fall), we would need at least 99 subjects in our study. Conversely, it should be apparent that a sample of 15 subjects, for example, would not suffice. Unfortunately, there is a common misconception that statistical wizardry can obviate the need for large sample sizes in setting normal values. This misconception is based

*Adapted from O’Brien, P. C., and Dyck, P. J.: Procedures for setting normal values. Neurology 45:17, 1995, with permission.

974

Nerve Conduction and Electromyography

on the erroneous belief that 95% of values in any population can be expected to lie within the mean ⫾2 SDs. The assertion is based on the assumption that values in a population must follow what statisticians refer to as a Gaussian, or normal, distribution. The use of the label “normal distribution” is unfortunate, since there is no basis for assuming that values in any population must follow such a distribution. The lack of any rational basis for assuming values in a population follow a Gaussian distribution was pointed out eloquently by Elveback et al.12,13 We will only add here that a Gaussian distribution implies the existence of negative values in the population, a circumstance that is often impossible. On the other hand, if one accepts that a Gaussian distribution is not a mathematical necessity but might only hold approximately, one must be concerned with the accuracy of the approximation (strictly speaking, one might argue that any statement is true “approximately”). In this case, however, one must acknowledge that judging the accuracy of a normal distribution in estimating normal percentiles in any specific instance would require knowledge of the true percentiles. Simply having a bellshaped curve is not sufficient to ensure close agreement of percentiles. The observation that the “normal law of errors” lacks both a mathematical and an empirical basis was also pointed out eloquently by Poincarè: “Everyone believes in the normal law of error, the physicists because they think that the mathematicians have proved it to be a mathematical necessity, the mathematicians because they believe that physicists have established it by laboratory demonstration.”13 To illustrate the issues raised thus far, we consider a population of serum urea values obtained from consecutive patients at the Mayo Clinic. The population consisted of 5,594 values, which were stored in a computer. We then generated a true random sample from this population (Table 36–1). Note that the 95th percentile may be estimated by the value 82, the 95th largest value. In fact, any number between 69 and 82 will exceed 95% of values in the sample, and thus any of these could be used to estimate the 95th percentile. Rather than choosing the largest value, an intermediate value is usually chosen, and various strategies are available for making the choice. Note also that the mean minus 2 SDs yields a negative number. To illustrate the need for large sample sizes in estimating percentiles, we generated nine additional samples of size 100 from this population (Table 36–2). Notice that estimates of the mean and median (P50) are quite stable, indicating that n ⫽ 100 is a large sample size for estimating a typical value. (Note also that medians are more appropriate than means when the distribution of values is skewed, as in the present instance, with the largest values in the population more spread out than the smallest values.) However, the upper

Table 36–1. Distribution of Serum Urea Values in a Sample (n ⫽ 100) Randomly Drawn from a Population (N ⫽ 5594)* Value (mg/dL)

Frequency and Percentile (P)

Value (mg/dL)

Frequency and Percentile (P)

173 103 95 88 82 68 66 52 50 46 45 44 42 41 40 39 38 37

1 1 1 1 1 (P95) 1 1 2 1 (P90) 1 2 1 5 3 5 (P75) 2 1 3

36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 20 19 18 16

2 6 2 3 9 (P50) 4 6 6 2 2 (P25) 2 4 6 3 (P10) 2 5 (P5) 1 1 1

*Mean of sample is 36.56; SD is 20.27. From O’Brien, P. C., and Sampo, M. A.: Statistics for clinicians: 10, normal values. Mayo Clin. Proc. 56:639, 1981, with permission of the Mayo Foundation.

Table 36–2. Mean and Selected Percentiles of Serum Urea Values in 10 Samples (Each n ⫽ 100) Sample 1* 2 3 4 5 6 7 8 9 10 S† Population values‡

Mean (mg/dL)

Values for Selected Percentiles P50

P90

P95

P99

36.56 33.92 34.24 33.00 33.47 36.67 35.15 38.93 32.31 36.57 2.07

32 31 31 31 31 32 30 32 30 32 0.8

50 50 50 43 46 48 52 50 48 46 2.7

82 57 62 52 60 56 61 69 56 55 8.8

173 103 123 86 220 172 123 388 93 174 89.3

35.33

31

48

60

124

*From Table 36–1. † SDs of the 10 values listed directly above. ‡ From population of 5594 values. From O’Brien, P. C., and Sampo, M. A.: Statistics for clinicians: 10, normal values. Mayo Clin. Proc. 56:639, 1981, with permission of the Mayo Foundation.

975

Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores

percentiles vary markedly from one sample to the next. In general, we recommend that a sample size of 100 should be considered a minimum for reliably estimating normal values.

2. A Single Independent Variable Next, suppose that values of Y vary with age. In this case, one would wish to estimate percentiles taking into account the ages of persons in the population. The methodology for doing so is described below. Step 1. Visually inspect a scatterplot of the data for outliers. If present, these values are to be deleted from the regression analyses which follow (steps 3 through 5) but not from the estimation of percentiles at the end (steps 6 and 7). Step 2. Visually inspect for skewness, possibly indicating the need to transform the data (taking logarithms of original values, for example). Step 3. Evaluate whether the average value of Y changes with age. This may be performed by constructing a regression equation relating Y to age, hav^ ing the form: Y ⫽ b0 ⫹ b1 age. The decision to adjust for age at this point would depend on several factors. Is the association statistically significant? Is the association with age of sufficient magnitude to warrant considering an adjustment? In assessing this, one might compare the SD of the Y values to the SD of the differences between the Y values actually observed and the corresponding estimates obtained from the regression equation. (These differences are referred to as the residuals from regression.) Alternatively, one might compare the squared values of these quantities, which statisticians refer to as the variance. The percent reduction in variance of Y is given by a statistic called “R squared.” If the association is found to depend on age, one might then wish to consider adding higher-order terms (age squared, for example) to the estimation equation. If no adjust^ ment is made for age at this point, Y is set equal to the mean. Step 4. It is equally important to consider whether the variability in Y values changes with age. This is commonly the case, since the aging process can be expected to progress at different rates, and in different forms, in individual subjects in the population. For percentiles above the 50th, one may estimate the effect of age on variability by regressing the positive residuals (R) against age, obtaining ^ an equation of the form: R ⫽ C0 ⫽ C1 age. The need to adjust for an association with age at this point is judged in a manner analogous to the method described in step 2. If no adjustment is ^ made, R is set equal to 1.

Step 5. For each subject (including the outliers excluded from the regression analyses), compute an age^ adjusted Z score: Z ⫽ (Y ⫺ Y)/R. Step 6. Compute the percentiles for Z, using the methodology described in section 1, obtaining Z(P) for P ⫽ 1, . . . , 99. (We have explicitly described the algorithm for the upper percentiles. The lower percentiles are obtained in the same manner, but using the absolute value of the negative residuals in step 3.) Step 7. To obtain percentiles for Y as a function of age, one computes Y(P) ⫽ RZ(P) ⫹ Y.

3. Multiple Explanatory Variables In practice, one may have a variety of characteristics that may influence the measurement of interest, such as age, gender, weight, height, occupation, and so forth. The methodology for incorporating such variables is easily accomplished, using the basic principles described in section 2. However, in developing the analogous regression ^ equations for Y and R (accounting for factors that influence average values and variability, respectively), we recommend using a stepwise regression approach. Thus, in ^ the regression equation for Y, one first identifies the variable that best predicts Y. After including this variable in the prediction equation, one then identifies which of the remaining variables adds the most information (provides the best prediction of Y) beyond the information already available from the first variable entered. One continues in this way until none of the remaining variables provides a statistically significant improvement in the prediction of Y. This process yields a prediction equation of the form: ^

Y ⫽ b0 ⫹ b1 age ⫹ b2 weight ⫹ ⭈ ⭈ ⭈ . ^

One can obtain an equation for R in similar fashion and compute Z scores as described previously, ultimately obtaining percentiles for Y. Although the computations become more complicated than in the case of a single explanatory variable, they are easily carried out using stepwise regression algorithms that are standard with most statistical software packages. Once the equations for obtaining percentiles have been developed, one simply enters the selected explanatory variables to obtain the corresponding percentiles. In some instances, it may be desirable to have percentiles available in tabular or graphic form, rather than (or in addition to) [a] computerized printout of percentiles. Approximate tabulations can be made by defining strata obtained by dividing the explanatory variables into intervals. One may then compute a selected percentile of interest (such as the 95th) for the midpoint of each interval, then tabling these percentiles. Graphic displays may be obtained by retaining one variable (such as age) as a

976

Nerve Conduction and Electromyography

quantitative variable. For each of the strata defined by the remaining explanatory variables, one may graph Y against the single quantitative variable (age), including both data points and the estimated percentile line. Such graphic display is helpful in verifying that the algorithm functioned appropriately.

4. Multiple Independent Variables and Multiple Dependent Variables One is sometimes confronted with a battery of test results. Recognizing that healthy patients may often have one or two unusual values, one may desire to identify patients with a pattern of values that justifies further medical attention. Building on the analysis described previously, one may evaluate the possibility of unusual patterns by focusing on the Z scores obtained in the study of each dependent variable. Specifically, for each subject in the reference population, one computes the distance between the multivariate array of Z scores [to] form the mean of the Z scores. The statistical measure of distance used is called Mahalanobis’ distance and the computations required are easily performed on most statistical software packages. Letting D represent the distance computed for each subject, one may set percentiles for these D values using the algorithm in section 1.

VARIABLES INFLUENCING NEUROPATHIC END POINTS IN A NORMAL SUBJECT COHORT: THE RDNS-NS In a study of persons mainly of northern European extraction without known neurologic disease or diseases predisposing to neuropathy (e.g., diabetes mellitus), we studied the first 15 women and 15 men per hemi-decade between 18 and 74 years of age who consented to participate in studies of neuropathic symptoms, impairments, and nerve tests (attributes of nerve conduction, quantitative sensation and autonomic tests) and measured the same metabolic and other risk covariates as studied in the RDNS. The conclusions that came from this study were 1. The percentage of patients who were suitable for inclusion into the normal subject cohort decreased with increasing age. 2. Physical variables other than age influenced neuropathic signs and nerve tests—to varying degrees, sex, height, and weight (or derivations of heights and weight) affected percentile estimates. 3. With the availability of microprocessors and because the variables may be different among neuropathic end points, it is now possible to compute percentiles for each end point for a given patient, taking applicable variables into account.

4. Knowing the percentile, the user can choose the level of abnormality to be declared “abnormal.” 5. For statistical analysis and for use in composite scores, percentiles should be converted to normal deviates. In Table 36–3 we show the percentage of the RDNS-NS cohort who had abnormalities of tendon reflexes. In Table 36–4 we show the variable affecting symptoms, impairments, and quantitative sensation test results. In Table 36–5 we show the variable and correlation coefficients that influenced attributes of motor nerve conduction. In Table 36–6 we show attributes affecting sensory nerve conduction.

COMPOSITE SCORES: FIVE ATTRIBUTES OF NERVE CONDUCTION, SEVEN TESTS, NIS(LL) ⫹ 7 TESTS, AND OTHER COMPOSITE SCORES We have used three composite scores in assessment of patients, but others are readily fashioned. The five attributes of nerve conduction (5NC) normal deviates comprise the peroneal motor nerve potential (CMAP), velocity (MNCV), and distal latency (MNDL); the tibial motor nerve distal latency (MNDL); and the sural nerve amplitude (SNAP) at point A. The 7 Tests are the 5NC normal deviates listed above plus the vibration detection threshold of the great toe as assessed with Computer-Assisted Sensory Examination System IV (CASE IV) and the reduction of heart pulse with deep breathing. In the Neuropathy Impairment Score (Lower Limbs) (NIS[LL]) ⫹ 7 Tests normal deviate, the summated score of the points obtained by adding neurologic abnormalities of the lower limbs and the score of the 7 Tests normal deviates are added. As an example, in Figure 36–1, we plot the

Table 36–3. Percentage of RDNS-NS Cohort with Decreased or Absent Muscle Stretch Reflexes Age Group 20–29 30–39 40–49 50–59 60–69 70–74

Knees

Ankles

Decreased or Absent Absent

Decreased or Absent Absent

2 0 0 0 1 0

2 0 0 0 0 0

3 2 4 10 7 21

3 0 0 2 3 8

RDNS-NS ⫽ Rochester Diabetic Neuropathy Study of Normal Subjects. Modified from Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects. Neurology 45:1115, 1995.

977

Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores

Table 36–4. Order of Physical Variables* That Are Significantly† Associated in Stepwise Multiple Linear Regression Analysis with Neuropathic Symptoms and Deficits and with Quantitative Sensation Testing Results and Their Model Coefficients

Table 36–5. Order of Physical Variables* That Are Significantly Associated in Stepwise Multiple Linear Regression Analysis with Motor Nerve Conduction Values and Their Model Coefficients End Points

End Points First

Second

Third

Neuropathic Symptoms NSP Age, 0.001 BMI, 0.003† NSP-W Age, ⬍0.001 BMI, 0.001 Sex, 0.740 NSP-S BMI, 0.107† NSP-A Age, 0.001 Age, ⫺0.067 Neuropathic Deficit‡ NIS Age, 0.007† NIS-R Age, 0.006† Quantitative Sensation Testing VDT Left foot Age, 0.063 BSA, 2.408 Left hand Age, 0.029 BSA, 0.997 Left lateral leg Age, 0.020 CDT Left foot Age, 0.057 Left hand Age, 0.026 Left lateral leg Age, 0.063 BSA, 2.698

BMI, ⫺0.046†

CDT ⫽ cooling detection threshold; NIS ⫽ Neuropathy Impairment Score (formely Neuropathy Disability Score); NIS-R ⫽ Neuropathy Impairment Score–Reflexes; NSP ⫽ Neuropathy Symptom Profile (NSP-W ⫽ NSP weakness; NSP-S ⫽ NSP sensation; and NSP-A ⫽ NSP autonomic); VDT ⫽ vibration detection threshold. *The physical variables assessed were age, gender (Sex), height, weight, body mass index (BMI), and body surface area (BSA). † All variables listed in the table are p ⬍ .001 unless they are identified by a dagger; then their p value is .001 ⬍ p ⬍ .01. ‡ No physical variables were significantly associated with cranial (NIS-Cr), weakness (NIS-W), or sensory (NIS-S) subsets of the NIS. Quadratic terms were also considered for the NSP regression equations. There were no variables significantly associated with the other NIS scores. From Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects. Neurology 45:1115, 1995, with permission.

values of the 5NC, the 7 Tests, and the NIS(LL) ⫹ 7 Tests normal deviates of a healthy subject cohort (from the RDNS-NS) on age, plot the common regression line, and fit a 97.5th line. Other composite scores that can be fashioned include composite scores of individual nerves (motor amplitudes, velocity, and distal latency of ulnar, peroneal, and tibial nerves); composite scores of F waves; composite score of the sural nerve; CMAP of the ulnar, peroneal, and tibial nerves; ⌺ MNCV of the ulnar, peroneal, and tibial nerves; ⌺ MNDL of the ulnar, peroneal, and tibial nerves; ⌺ MNCV of the ulnar, peroneal, and tibial nerves; and the sensory nerve conduction velocity of the sural nerve.

First

Second

Third

Ulnar Amp CV DL F-wave L

Age, ⫺0.041 Age, ⫺0.118 Ht, 0.011 Ht, 0.157

Sex, 3.981 Age, 0.005 Age, 0.056

Sex, ⫺1.592

Medial Amp CV DL F-wave L

Age, ⫺0.061 Age, ⫺0.109 Sex, ⫺0.179 Sex, ⫺1.437

Sex, 1.853 Age, 0.008 Ht, 0.143

BSA, 0.417 Age, 0.060

Peroneal Amp CV DL F-wave L

Age, ⫺0.042 Age, ⫺0.121 Ht, 0.020 Ht, 0.405

Tibial Amp CV DL F-wave L

Age, ⫺0.114 Ht, ⫺0.248 Age, 0.009 Ht, 0.138

Ht, ⫺0.214 Age, 0.108 BSA, ⫺4.796 Age, ⫺0.114 Ht, 0.016 Age, 0.422

Amp ⫽ amplitude; CV ⫽ conduction velocity; DL ⫽ distal latency; L ⫽ latency. *The physical variables assessed were age, gender (Sex), height (Ht), weight, body mass index, and body surface area (BSA). All variables listed in the table are p ⬍.001. Modified from Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects. Neurology 45:1115, 1995.

INDIVIDUAL ATTRIBUTES VERSUS COMPOSITE SCORES OF NERVE CONDUCTION ABNORMALITY: SENSITIVITY, REPRODUCIBILITY, REPRESENTATIVENESS, AND CONCORDANCE Questions of sensitivity and reproducibility of individual attributes of nerve conduction versus composite scores were explored for the RDNS cohort, our patients with CIDP, and our patients with multifocal motor neuropathy.5 Setting specificity at the 97.5th percentile, the sensitivity of individual attributes or composite scores can be judged by the height of the bars as shown in Figure 36–2. We found that most composite scores of individual nerves and composite scores of several nerves were more sensitive, often markedly so, than were individual attributes of nerve

978

Nerve Conduction and Electromyography

Table 36–6. Order of Physical Variables* That are Significantly† Associated in Stepwise Multiple Linear Regression Analysis with Sensory Nerve Conduction Values and Their Model Coefficients End Points First

Second

Sensory Nerve Conduction Su Amp Su CV Su DL Wrist-Index Med Amp Wrist-Index Med CV Wrist-Index Med DL Palm-Wrist Left Ul Amp‡ Palm-Wrist Left Ul DL Palm-Wrist Left Med Amp Palm-Wrist Left Med DL Palm-Wrist Right Ul Amp Palm-Wrist Right Ul DL Palm-Wrist Right Med Amp Palm-Wrist Right Med DL Wrist-Fifth Amp Wrist-Fifth CV Wrist-Fifth DL Palm-Wrist Left Ul CV Palm-Wrist Left Med CV Palm-Wrist Right Ul CV Palm-Wrist Right Med CV Summated CMAP (Ul, Per, Tib) Summated SNAP (Ul, Su)

Age, ⫺0.199 Ht, ⫺0.364 Ht, 0.016 Age, ⫺0.825 Age, ⫺0.167 Age, 0.006 BSA,⫺45.920 Age, 0.003 BSA,⫺89.616 BSA, 0.372 Age, ⫺0.657 Age, 0.004 Age, ⫺2.069 Age, 0.007 Age, ⫺0.608 Age, ⫺0.133 Age, 0.005 Age, ⫺0.133 Age, ⫺0.181 Age, ⫺0.149 Age, ⫺0.162 Age, ⫺0.195 BSA,⫺47.128

BSA,⫺10.134 Age, ⫺0.134 Age, 0.007 Sex, 14.297 Ht, ⫺0.138 BSA, 0.298 Age, ⫺0.618 Ht, 0.005 Age, ⫺1.988 Age, 0.005 Sex, 15.034 Ht, 0.005 BSA,⫺90.341 Wt, 0.005 Sex, 20.218 Sex, 3.408 Ht, 0.008 Sex, 4.463 Ht, ⫺0.117 Sex, 3.819 Sex, 3.533 Wt, ⫺0.073 Age, ⫺0.873

Autonomic Heart Period Valsalva Heart Period Deep Breathing CASS

Age, ⫺0.010 Age, ⫺0.245 Age, 0.008

Third

BSA,⫺16.063

Sex,

12.909

Sex,

33.703

Wt,

⫺0.272

Sex,

30.035

Sex,

1.970†

Sex,

7.460†

Amp ⫽ amplitude; CASS ⫽ Composite Autonomic Scoring Scale; CMAP ⫽ compound muscle action potential; CV ⫽ conduction velocity; DL ⫽ distal latency; Med ⫽ median; Per ⫽ peroneal; SNAP ⫽ sensory nerve action potential; Su ⫽ sural; Tib ⫽ tibial; Ul ⫽ ulnar. *The physical variables assessed were age, gender (Sex), height (Ht), weight (Wt), body mass index, and body surface area (BSA). † All variables listed in the table are p ⬍ .001 unless they are identified by a dagger; then their p value is .001 ⬍ p ⬍ .01. ‡ Ht was also significant by association with this attribute (coefficient of 0.521, p ⬍ .01). From Dyck, P. J., Litchy, W. F., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects. Neurology 45:1115, 1995, with permission.

conduction. Therefore, for generalized polyneuropathies, composite scores appear to be more sensitive than individual attributes of nerve conduction. For mononeuropathies, composite scores of individual nerves also were more sensitive than individual attributes of nerve conduction, but composite scores of several nerves were not. Reproducibility of composite scores, measured by intraclass correlation coefficients, also tended to be higher than were individual attributes (Fig. 36–3).

In addition, we observed that composite scores, by virtue of measuring several attributes of nerve conduction, are more representative of peripheral neuropathy than individual attributes. We also found that composite scores correlated better with neurologic impairment than did individual attributes (Fig. 36–4). We concluded that, “with the availability of microprocessors and normal databases, electromyographers may increasingly seek to express nerve conduction abnormality

Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores

979

FIGURE 36–1 The values of three composite scores described in the text (⌺ 5NC, 7 Tests, and NIS[LL] ⫹ 7 Tests) for 330 persons in the RDNS-NS (normal subject) cohort plotted on age. The 50th and 97.5th percentile lines are shown. The description of how composite scores are calculated and how they may be used in epidemiologic surveys, controlled trials, and even medical practice is outlined in the text.

also as composite scores of individual or of several nerves.”5

FUNCTIONAL ABNORMALITY EVEN WHEN TEST VALUES FALL WITHIN THE RANGE OF NORMAL VALUES Thoughtful physicians have long known that abnormality of a nerve may be present even when test values fall within normal limits. Several lines of evidence can be used to illustrate this point. Normal deviates and compos-

ite scores are useful for each of these cases. In the first case, the distribution of percentile values in the normal range of a study cohort may be abnormal—more of the test results may fall between the 50th and 97.5th percentiles than fall between the 50th and 25th percentiles. Assuming that the cohort is sufficiently large, that the result is not occurring by chance, that selection bias has not occurred, and that percentile values have been adequately assigned to this altered distribution, this provides evidence suggestive of a subtle abnormality, even when individual values fall within normal limits. In the second case, serial evaluations of nerve conduction of a given

FIGURE 36–2 The sensitivity of individual attributes (bars 1–3, 5–7, 9–11, 13–15, and 17–19), composite scores of individual nerves (bars 4, 8, 12, and 20), and composite scores of multiple nerves (bars 16 and 21–27). For each of the three disease cohorts, the bars (from left to right) represent compound muscle action potential (CMAP) amplitude, motor nerve conduction velocity (MNCV), and motor nerve distal latency (MNDL) for the ulnar (bars 1–3), peroneal (bars 5–7), and tibial (bars 9–11) nerves. Bars 4, 8, and 12 are composite scores of CMAP, MNCV, and MNDL for the ulnar, peroneal, and tibial nerves. Subsequent bars: 13 ⫽ ulnar (Ul) motor nerve F wave (MNFW) ⱖ 97.5th percentile; 14 ⫽ median (Med) MNFW ⱖ 97.5th percentile; 15 ⫽ tibial (Tib) MNFW ⱖ 97.5th percentile; 16 ⫽ MNFW composite score of bars 13, 14, and 15; 17 ⫽ sural (Sur) sensory nerve action potential (SNAP) amplitude ⱖ 2.5th percentile; 18 ⫽ Sur sensory nerve conduction velocity (SNCV) ⱖ 2.5th percentile; 19 ⫽ Sur sensory nerve distal latency (SNDL) ⱖ 97.5th percentile; 20 ⫽ Sur composite score of bars 17, 18, and 19; 21 ⫽ ⌺ CMAP of Ul, peroneal (Per), and Tib nerves ⱖ 97.5th percentile; 22 ⫽ ⌺ MNCV of Ul, Per, and Tib nerves ⱖ 97.5th percentile; 23 ⫽ ⌺ MNDL of Ul, Per, and Tib nerves ⱖ 97.5th percentile; 24 ⫽ ⌺ MNCV of Ul, Per, and Tib and SNCV of Sur nerves ⱖ 97.5th percentile; 25 ⫽ ⌺ MNCV and MNDL of Ul, Per, and Tib nerves ⱖ 97.5th percentile; 26 ⫽ ⌺ MNCV and MNDL of Ul, Per, and Tib and SNCV and SNDL of Sur nerves ⱖ 97.5th percentile; 27 ⫽ Per and Sur ⌺ NC normal deviate (CMAP, MNCV, MNDL, SNAP amplitude, and SNDL) ⱖ 97.5th percentile. *Abnormality as defined in the text. **Percent of nonmissing values where abnormality was ⱖ 97.5th percentile or ⱖ 2.5th percentile. (From Dyck, P. J., Litchy, W. J., Daube, J. R., et al.: Individual attributes versus composite scores of nerve conduction abnormality: sensitivity, reproducibility, and concordance with impairment. Muscle Nerve 27:202, 2003, with permission.)

Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores

981

FIGURE 36–3 Intraclass correlation coefficients of the individual attributes and composite scores of nerve conduction of 25 patients without and with varying severities of diabetic sensory polyneuropathy as described in the text. (From Dyck, P. J., Litchy, W. J., Daube, J. R., et al.: Individual attributes versus composite scores of nerve conduction abnormality: sensitivity, reproducibility, and concordance with impairment. Muscle Nerve 27:202, 2003, with permission.)

person may show statistically significant worsening over time, indicating a worsening of nerve function. Because this worsening could conceivably be due to age or to a change of some anthropomorphic variable (e.g., weight), use of percentiles and normal deviates corrected for these variables should prevent an interpretation of deteriorating nerve function when these factors are the actual reason for the indication of worsening. In the third case, a disease or an intervention may cause worsening (or improvement) of a group of patients as compared to a control group, even when the measured values fall within the range of normal values. Here also, use of composite scores based on normal deviates is helpful because it minimizes the likelihood of anthropomorphic factors being the explanation for the difference.

might prevent worsening, a therapeutic study would need to be conducted for a period of 3 years to achieve a difference of 2 points—the level of abnormality considered by the Peripheral Nerve Society Consensus Group to be clinically meaningful.17 In our study of the risk covariates for diabetic polyneuropathy, we assessed with multivariate analysis which risk covariates over time related to severity of diabetic polyneuropathy as assessed with the NIS(LL) ⫹ 7 Tests. We found that other microvessel complications, glycemic exposure, and type of diabetes were the important risk covariates.

COMPOSITE SCORES AS MEASURES OF SEVERITY OF NEUROPATHY OVER TIME: EPIDEMIOLOGIC SURVEYS AND CONTROLLED CLINICAL TRIALS

For the purpose of conducting linkage analysis of kindreds with inherited neuropathies, it is of major importance to correctly identify kindred members as affected or unaffected. For inherited neuropathies with a primary abnormality of Schwann cells, low nerve conduction velocities are excellent indicators of genetic involvement. For neuronal or axonal neuropathies, however, lower than normal conduction velocities are not as useful. In this situation, use of normal deviates and composite scores may be helpful. Because these include composite measures of amplitudes, conduction velocities, and distal latencies, affected persons may be recognized from unaffected persons. After a certain age, it may be possible to identify patients as affected, not affected, or questionable.15

Neurologic signs, attributes of nerve conduction, quantitative sensation tests, and quantitative autonomic tests are useful to follow the course of severity of neuropathy over time. For the RDNS, we found that the NIS(LL) ⫹ 7 Tests normal deviate was a better minimal criterion for polyneuropathy than clinical judgment alone.2 We found that, on average, diabetic patients worsened by 0.34 points/year, whereas patients with diabetic polyneuropathy worsened by 0.85 points/year. Assuming that a therapeutic agent

COMPOSITE SCORES IN GENETIC LINKAGE ANALYSIS

982

Nerve Conduction and Electromyography

FIGURE 36–4 Correlation coefficients of various individual attributes or composite scores of nerve conduction and clinical impairment. In the RDNS, clinical impairment was assessed using the NIS(LL), whereas in patients with CIDP and multifocal motor neuropathy, it was assessed with the NIS. The height of the bar represents the correlation coefficient. Statistical significance (P ⬍ .05) is indicated by an asterisk above the bar. (From Dyck, P. J., Litchy, W. J., Daube, J. R., et al.: Individual attributes versus composite scores of nerve conduction abnormality: sensitivity, reproducibility, and concordance with impairment. Muscle Nerve 27:202, 2003, with permission.)

Nerve Tests Expressed as Percentiles, Normal Deviates, and Composite Scores

REFERENCES 1. Dyck, P. J.: Diabetic polyneuropathy in controlled clinical trials: consensus report of the Peripheral Nerve Society. Ann. Neurol. 38:478, 1995. 2. Dyck, P. J., Davies, J. L., Litchy, W. J., and O’Brien, P. C.: Longitudinal assessment of diabetic polyneuropathy using a composite score in the Rochester Diabetic Neuropathy Study cohort. Neurology 49:229, 1997. 3. Dyck, P. J., Davies, J. L., Wilson, D. M., et al.: Risk factors for severity of diabetic polyneuropathy: intensive longitudinal assessment of the Rochester Diabetic Neuropathy Study cohort. Diabetes Care 22:1479, 1999. 4. Dyck, P. J., Lais, A., Karnes, J. L., et al.: Fiber loss is primary and multifocal in sural nerves in diabetic polyneuropathy. Ann. Neurol. 19:425, 1986. 5. Dyck, P. J., Litchy, W. J., Daube, J. R., et al.: Individual attributes versus composite scores of nerve conduction abnormality: sensitivity, reproducibility, and concordance with impairment. Muscle Nerve 27:202, 2003. 6. Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects (RDNS-HS). Neurology 45:1115, 1995. 7. Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Neuropathy endpoints. Neurology 46:1193, 1996. 8. Dyck, P. J., Melton, L. J., O’Brien, P. C., and Service, F. J.: Approaches to improve epidemiological studies of diabetic neuropathy: insights from the Rochester Diabetic Neuropathy Study. Diabetes 46:S5, 1997. 9. Dyck, P. J., O’Brien, P. C., Litchy, W. J., et al.: Use of percentiles and normal deviates to express nerve conduction and other test abnormalities. Muscle Nerve 24:307, 2001. 10. Dyck, P. J., O’Brien, P. C., Litchy, W. J., et al.: Expression of nerve conduction test results—reply. Muscle Nerve 25:124, 2002.

983

11. Dyck, P. J., Velosa, J. A., Pach, J. M., et al.: Increased weakness after pancreas and kidney transplantation. Transplantation 72:1403, 2001. 12. Elveback, L. R.: A discussion of some estimation problems encountered in establishing “normal values.” In Gabrieli, E. R. (ed.): Clinically Oriented Documentation of Laboratory Data. New York, Academic Press, p. 117, 1972. 13. Elveback, L. R., Guillier, C. L., and Keating, F. R. Jr.: Health, normality, and the ghost of Gauss. JAMA 211:69, 1970. 14. Feki, I., and Lefaucheur, J. P.: Correlation between nerve conduction studies and clinical scores in diabetic neuropathy. Muscle Nerve 24:555, 2001. 15. Klein, C. J., Cunningham, J. M., Atkinson, E. J., et al.: The gene for HMSN 2C maps to 12q23-24: a region of neuromuscular disorders. Neurology 60:1151, 2003. 16. Lew, H., Wang, L., and Robinson, L. R.: Test, re-test reliability of the CSI compared to single test results. Muscle Nerve 23:1261, 2000. 17. Low, P. A., Deny, J. C., Opfer-Gehrking, T. L., et al.: Effect of age and gender on sudomotor and cardiovagal function and blood pressure response to tilt in normal subjects. Muscle Nerve 20:1561, 1997. 18. O’Brien, P. C., and Dyck, P. J.: Procedures for setting normal values. Neurology 45:17, 1995. 19. Robinson, L. R., and Micklesen, P. J.: Expression of nerve conduction test results. Muscle Nerve 25:123, 2002. 20. Robinson, L. R., Micklesen, P. J., and Wang, L.: Strategies for analyzing nerve conduction data: superiority of a summary index over single tests. Muscle Nerve 21:1166, 1998. 21. Solders, G., Andersson, T., Borin, Y., et al.: Electroneurography index: a standardized neurophysiological method to assess peripheral nerve function in patients with polyneuropathy. Muscle Nerve 16:941, 1993.

37 Single-Fiber Electromyography and Other Electrophysiologic Techniques for the Study of the Motor Unit ERIK V. STÅLBERG AND JOZˇ E V. TRONTELJ

Description of Methods SFEMG during Voluntary Activation SFEMG during Intramuscular Electrical Stimulation Studies of Some Spinal Events with SFEMG Multielectrode Studies Scanning Electromyography Macro Electromyography

Findings in Nerve-Muscle Disorders SFEMG in Denervation-Reinnervation SFEMG Signs of Abnormal Transmission in Nerve Branches Signs of Abnormal Nerve Excitability SFEMG in Myasthenia Gravis SFEMG during Axonal Stimulation in MG and LEMS Amyotrophic Lateral Sclerosis Syringomyelia

The functional organization and physiology of the motor unit can be studied with different electrophysiologic methods. Even seven decades after its development, classic needle electromyography (EMG) with the conventional coaxial electrode is still the one most widely used. A number of other methods have been developed, and they have contributed important new knowledge, but their clinical use remains more or less restricted to special indications and complementary to conventional needle EMG. This chapter describes three of these methods. Each focuses on a different aspect of the motor unit: singlefiber electromyography (SFEMG) is well suited for highly selective studies of individual muscle fibers and motor end plates, as well as single motor neurons; scanning EMG provides information on internal topography of the motor unit presented as an electrophysiologic cross section through its territory; and macro EMG is designed to give the summated electrical activity of the whole motor unit. Each of the three methods is described briefly, and the findings in some selected disorders highlight their use for the study of the motor unit.

Postpolio Sequelae Guillain-Barré Syndrome Hereditary Motor and Sensory Neuropathy: Charcot-Marie-Tooth Disease Diabetes Mellitus Alcoholic Neuropathy Primary Myopathies Summary

DESCRIPTION OF METHODS SFEMG during Voluntary Activation The method of SFEMG has been described in detail.44,58 In this type of EMG recording, a small recording surface is used. The electrode consists of a 0.5-mm steel cannula with 1 to 14 platinum wires, each 25 ␮m in diameter, exposed in a side port a few millimeters behind the tip (Fig. 37–1). As a consequence of the small size of the electrode surface, the recording is very selective. Action potentials from the most closely located muscle fibers are of considerably higher amplitude than those generated by muscle fibers 200 to 300 ␮m away. An electrode with a large recording surface, such as the concentric or monopolar needle EMG electrode (Fig. 37–1C and D), distorts the action potentials from the closest fibers, recording them with a much smaller amplitude, while the amplitude difference between action potentials from these and more distant fibers is less than in the case of SFEMG, making the recording less selective. To increase the selectivity of the SFEMG recording further, action potentials from distant muscle fibers containing more low-frequency 985

986

Nerve Conduction and Electromyography

FIGURE 37–1 Different types of EMG electrodes. A, SFEMG electrode. B, SFEMG multielectrode. C, Concentric needle electrode. D, Monopolar electrode. E, Macro electrode. The recording area is black. (From Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993, with permission.)

components than adjacent muscle fibers can be further attenuated by restricting the low-frequency range of the recording system. Typical filter settings are 500 Hz for the high-pass filter and 10 kHz for the low-pass filter. As a result, the electrode usually records activity from one or two muscle fibers in a normal motor unit. If details of the action potential shape are to be studied, the filter settings should be from 2 Hz to 20 kHz. Some of the results that are obtained with SFEMG are presented briefly below. It should be noted that, as with other EMG techniques for examination of individual fibers or motor units, usually motor units with a low recruitment threshold are studied with SFEMG. Propagation Velocity By means of a multielectrode (see Fig. 37–1B), the action potential from one muscle fiber, activated voluntarily or by electrical stimulation, can be recorded at two sites along the fiber approximately 200 ␮m apart.29 The conduction time between them can be measured with high accuracy, and the propagation velocity can be calculated. The velocity differs for different muscle fibers, largely depending on the size of the muscle fiber. For the individual fiber the velocity decreases during continuous activity and recovers after a short rest. This phenomenon is seen in the majority of fibers studied. Furthermore, the velocity of the individual impulse is influenced by previous impulses in a well-defined relationship, called the velocity recovery function (VRF). If two electrical stimuli are delivered to the muscle fiber, the change in velocity for the second impulse (test impulse) can be studied for different intervals. For intervals shorter than 3 ms, the fiber is refractory (absolute refractory period). For intervals from 3 to 10 ms, the velocity of the test impulse is 80% to 100% of the first (conditioning

impulse). At longer intervals, between 5 and 500 to 1000 ms, the velocity is higher for the test than for the conditioning impulse, with a maximum increase by up to 20% for intervals between 5 and 15 ms. This phenomenon is additive if repetitive impulses occur within 500 to 1000 ms, which is best seen at the beginning of a train of stimuli of sufficiently high frequency (e.g., 20 Hz). Similarly, the amplitude of the action potential varies as an effect of previous depolarization. Propagation velocity can also be studied with a single-surface SFEMG electrode in muscle fibers stimulated directly in the same fascicle with a needle cathode some 20 to 30 mm away (see below). In this way an estimate of muscle fiber size and its variation may be obtained.50 This value may be of interest in disease, in particular in myopathies. Propagation velocity measurements can thus be used to obtain information about the fiber sizes in a muscle. The interdischarge interval–dependent changes in velocity and amplitude reflect events during depolarization and give information about membrane characteristics. This phenomenon of changing excitability and impulse velocity is also seen in nerve (except for supernormal velocity) and can be used in certain situations to study normal and abnormal membrane function. Jitter Neuromuscular transmission can be studied by means of SFEMG. An SFEMG electrode with a single recording surface (see Fig. 37–1A) is positioned in such a way that activity from two muscle fibers belonging to the same motor unit is recorded. The time interval between them, reflecting the difference in conduction time for the two action potentials from the last common branching point to the recording site, varies at consecutive discharges. This phenomenon, called “jitter,” is mainly due to random variation in transmission

SFEMG and Other Techniques for Study of the Motor Unit

time across the two motor end plates involved. It can be expressed as the mean difference of interpotential intervals on consecutive discharges, or mean consecutive difference (MCD). This measure of variability is used instead of standard deviation about the mean to reduce the effects of slow trends in the interpotential intervals, which may occur over the hundred discharges usually studied.6 Jitter was earlier quantified manually from film recordings or by means of a separate computer. Nowadays all larger EMG equipment has built-in software for jitter analysis. Jitter is normally of the order of 20 to 50 ␮s MCD, with some differences among muscles and some increase at higher ages.12 Jitter for the individual motor end plate is correlated with its functional safety margin, as shown in experiments with partial curarization.39 In some recordings jitter is less than 5 ␮s. This is rarely seen in normal muscles but more often seen in muscular dystrophies. These recordings are probably from split muscle fibers.7 In situations of slightly disturbed neuromuscular transmission, jitter is increased. With more pronounced transmission impairment, impulse blocking also occurs; that is, now and then individual single-fiber action potentials are missing. In situations of increased jitter without blocking, no weakness is present and results of other electrophysiologic tests, such as repetitive nerve stimulation (RNS), are normal. It is thus possible to detect subclinical transmission abnor-

FIGURE 37–2 Principle for fiber density measurement in SFEMG. The muscle fibers in one motor unit are indicated as small solid circles, and the uptake area of the recording electrode is represented as a half-circle. In normal muscle (1 and 2), action potentials are recorded from only one or two fibers. In reinnervation (3), many fiber action potentials are recorded owing to increased fiber density in the motor unit (time calibrations are 1 ms). (From Stålberg, E., and Trontelj, J. V.: Single Fiber Electromyography in Healthy and Diseased Muscle. New York, Raven Press, 1994, with permission.)

987

malities with SFEMG. In myasthenia gravis (MG) this test shows a higher diagnostic yield than other conventionally used electrophysiologic tests.13,25 It should be noted that abnormal jitter and blocking are seen also in other conditions of disturbed neuromuscular transmission, such as in early reinnervation, and the finding is therefore not specific for MG, but indicates disturbed impulse transmission in the most distal part of the motor unit, usually the motor end plate. Fiber Density Information about the local distribution of fibers from the same motor unit within a small area is obtained in the following way. The electrode is placed close to one active muscle fiber in voluntarily contracting muscle and the number of synchronously firing fibers (i.e., those belonging to the same motor unit) is counted (Fig. 37–2). To be accepted, the action potential must have an amplitude exceeding 200 ␮V and a rise time less than 300 ␮s. With these criteria, the average uptake volume of the electrode corresponds to a hemisphere with a radius of approximately 300 ␮m.11 The mean number of muscle fibers from one motor unit within the uptake volume is calculated from 20 different recording positions obtained after three to five insertions through the skin. Fiber density (FD) is increased in conditions with reinnervation corresponding to fiber-type grouping in the muscle biopsy and can be increased also in primary myopathies.

988

Nerve Conduction and Electromyography

SFEMG during Intramuscular Electrical Stimulation SFEMG recordings can be performed during intramuscular electrical stimulation (Fig. 37–3). They are used in patients who cannot cooperate (unconscious patients, small children, patients with tremor, those with very weak muscles) or when the effect of firing frequency or rhythm should be studied, and in animal studies.45,57 A Tefloncoated needle with a 1-mm bare tip is used as stimulating cathode, introduced into the end-plate region for axonal stimulation or more remotely for direct muscle fiber stimulation. A remote surface or needle electrode is used as the anode. The stimulus duration is usually 20 to 50 ␮s. The stimulus amplitude is adjusted to a strength, usually less than 5 mA or 10 V, that produces small twitches in a part of the muscle. The SFEMG electrode is inserted approximately 2 cm distally or proximally to the cathode in the twitching part of the muscle. The muscle fiber can be stimulated either directly or indirectly via its axon. Direct Stimulation To study muscle fiber membrane parameters, the muscle fiber is stimulated directly. The evidence of direct muscle fiber stimulation is a jitter less than 4 ␮s, generated mainly at the stimulus site. However, with threshold stimulation, the jitter may be very large and the mean latency longer.59

Even when intramuscular stimulation is performed close to the end-plate zone for a jitter study, part of the responses may be due to direct stimulation. Direct muscle fiber stimulation has been used in studies of denervated muscles55 and in myopathies.17 Patients with myotonia congenita (Thomsen and Becker types) showed a pronounced decrement of action potential amplitude and changes of shape during stimulation at 2 to 20 Hz. The findings were interpreted as signs of a disturbed depolarizationrepolarization cycle. In addition, signs of focal conduction block of action potentials were found in some fibers. This is also found in studies with surface EMG.5 Both mechanisms may contribute to the myotonic weakness. Axonal Stimulation With axonal stimulation the jitter measured between the stimulus and a single-fiber action potential exceeds 4 ␮s. With suprathreshold stimulation nearly all jitter is generated at the motor end plate. As the stimulus strength is raised from subthreshold, the responses first appear with intermittent (false) blocking and large jitter. With cessation of blocking, the jitter becomes smaller and the latency shorter. When the stimulation is sufficient (e.g., 10% above the activation threshold), a further increase in stimulation strength does not produce any further decrease in jitter or latency. Most of the increased jitter during intermittent (false)

FIGURE 37–3 Intramuscular stimulation and SFEMG. A, Stimulating and recording electrodes. The muscle fiber is activated directly or through its intramuscular nerve branch. B, Ten superimposed recordings showing a jitter of less than 5 ␮s, indicating direct muscle fiber stimulation. C, Ten superimposed recordings with a jitter of more than 5 ␮s, indicating axonal stimulation. (From Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993, with permission.)

SFEMG and Other Techniques for Study of the Motor Unit

blocking as well as the changes in mean latency are due to changing muscle fiber propagation velocity during different stages of the velocity recovery cycle (VRF). The magnitude of the VRF-related jitter and of the mean latency change depends on the length of the muscle fiber segment between the end plate and the recording SFEMG electrode; if this is quite short, the effects may be negligible. Jitter at the Ranvier node, when stimulated at threshold, was found to be of the order of 5 ␮s.60 Because only one end plate is generating the jitter measured in this way, the normal values of jitter on axonal stimulation are on average lower by a factor of 冪2 compared with those obtained during voluntary activation, where, in recordings from muscle fiber pairs, two end plates contribute. Jitter can be studied at various frequencies and patterns, even for long periods. This technique has been used to study the safety margin in the normal motor end plate53 and the neuromuscular transmission in various diseases,52 and also for detailed studies of F waves and axon reflexes.61 In some laboratories jitter measurement with axonal stimulation is used routinely as a main test for neuromuscular transmission disorders such as MG or LambertEaton myasthenic syndrome (LEMS).

Studies of Some Spinal Events with SFEMG Firing Pattern The pattern of firing during voluntary activation varies among different motor units, a fact that has been used to classify them into tonic and phasic. The pattern is changed from normal in central disorders. Some of these studies have been performed with SFEMG electrodes, others with concentric or wire electrodes.42 F Response F responses are recurrent responses of antidromically activated motor axons. In surface recordings this so-called late response cannot always be differentiated from the H reflex with certainty, but by means of SFEMG reliable distinction can be made. The F response, as recorded with SFEMG from a single muscle fiber, is always preceded by a direct response of the same muscle fiber. Furthermore, the stimulus-response latency variability is less than 50 ␮s. In individual motor neurons, the F responses occur infrequently, typically one to five times per 100 stimuli. The response frequency is usually much higher in spasticity.26 The large variability in F-wave latencies recorded with surface electrodes reflects different impulse conduction times along different axons that have become activated by successive stimuli, not a variability in each individual neuron. SFEMG has also been used to study transmission in reflex arcs.47 The stimulus-response variability depends principally on the time variability in the synapses involved. It therefore varies for different reflexes and depends on the number of synapses involved.

989

H Reflex An H reflex in one muscle fiber is only exceptionally preceded by a direct response in the same muscle fiber. The stimulus-latency jitter is approximately 150 ␮s in the relaxed limb; it may decrease with Jendrassik’s maneuver and increase with activation of antagonistic muscles.51 It may occur up to 100 times per 100 stimuli at stimulation rates below 0.3 Hz and changes with background muscle activity more than the F response. Other Reflexes Compared with the H reflex, the stimulus-response jitter is two to three times larger in the oligosynaptic R1 component of the blink reflex and even larger in polysynaptic reflexes such as the R2 component of the blink reflex, flexion withdrawal reflex, and pudendal reflexes.41 In these reflex responses it is not uncommon to see repetitive discharges of one motor neuron after a single stimulus. This has been interpreted to be due to repetitive firing of interneurons and possibly to several central pathways of different lengths. Motor Neuron Excitability Single muscle fiber action potentials have also been used as event markers in studies of motor neuron excitation. In one such study,2 SFEMG and macro EMG were used to study the effect on motor neuron firing of superimposed afferent activity (e.g., that produced by weak nerve stimulation). Socalled poststimulus histograms were constructed. In this study the findings provided evidence that presynaptic inhibition of Ia afferents occurs in humans. Cortical Stimulation SFEMG has been used to study corticospinal impulse transmission after electrical or magnetic cortical stimulation.23,62 The stimulus-response jitter is of the order of 50 to 800 ␮s. With increasing stimulus strength or with voluntary coactivation, stepwise shortening of the latency may be seen. The multiple latencies appear to be due to a multiplepeak excitatory postsynaptic potential (EPSP). This may result from direct or indirect activation of the cortical motor neurons, or alternation between mono- and oligosynaptic transmission on the spinal motor neuron. A third possibility is repetitive firing of pyramidal neurons, which is known to occur on cortical stimulation. The complex EPSP with a stepwise rising slope may reach the motor neuron firing threshold with any of its peaks, depending on the summation of the various excitatory inputs and the fluctuation in background excitability. The latency jumps seem to correspond to the intervals between the cortical discharges.

Multielectrode Studies In an attempt to map the distribution of muscle fibers within the motor unit, a SFEMG multielectrode has been moved in defined steps through its territory according to a

990

Nerve Conduction and Electromyography

technique described elsewhere.27 In this way recordings were made from 44 well-defined steps 300 ␮m apart. It was shown that, in neurogenic disorders such as amyotrophic lateral sclerosis (ALS) and hereditary motor and sensory neuropathy, the local concentration of muscle fibers was increased without an associated increase in the anatomic territory. This is in disagreement with the suggestion of enlarged territories in neurogenic conditions based on studies with another type of multielectrode.3

Scanning Electromyography Another way to map the electrical activity of the motor unit, offering a higher spatial resolution than the multielectrode method, is by means of scanning EMG.34,35 This may be performed with any kind of needle EMG electrode, depending on the question to be answered. Below is a description of the use of concentric needle electrodes for these recordings. A separate electrode, preferably an SFEMG electrode, records activity from a single motor unit during voluntary contraction and triggers the oscilloscope sweep. A concentric needle electrode (or any type of needle EMG electrode) is positioned in the muscle in such a way that it will traverse the motor unit under study as it is pulled 20 mm through the muscle. A step motor is used to withdraw the electrode by 50 ␮m at a time after each discharge. The activity recorded with this electrode is displayed on the oscilloscope sweep. When an action potential from the triggering motor unit is obtained with the concentric needle electrode, it will occur at a relatively constant position along the sweep at successive discharges. This recording gives a picture of the spatial and temporal distribution of action potentials that can be obtained from a single motor unit (Fig. 37–4). In the biceps brachii and tibialis anterior muscles, motor unit potentials (MUPs) containing spike components (derived from muscle fibers close to the recording surface) were distributed over a distance of 5 to 10 mm. There is usually no “silent area” in the cross section of a normal motor unit. The wave complex of one motor unit in the normal biceps brachii and tibialis anterior muscles has one major negative peak in approximately 50% of recordings, but may show two, three, or exceptionally four peaks separated by up to 9 ms at some points of the cross section, depending on different arrival times of the muscle action potentials at the electrode. The interpeak interval is typically approximately 3 ms. The individual peaks are probably derived from groups of muscle fibers, called a “motor unit fraction,” with separate end-plate zones, innervated by different major branches of the axon divided relatively proximally, each with a different conduction time. The motor unit fraction should not be confused with the earlier concept of the “subunit” because the muscle fibers in each fraction are still randomly scattered and intermingled with those of other

FIGURE 37–4 Recording principle for scanning EMG. An SFEMG electrode is used to trigger the oscilloscope sweep and initiate a step motor pulling a concentric needle electrode. Each step is 50 ␮m. When the concentric electrode records activity from the triggering motor unit, action potentials occur time-locked on the oscilloscope screen. The temporal and spatial distribution of activity within the recording corridor can be analyzed. (From Stålberg, E., and Antoni, L.: Computer aided EMG analysis. In Desmedt, J. [ed.]: Progress in Clinical Neurophysiology, Vol. 10: Computer Aided Analysis of EMG. Basel, Karger, p. 186, 1983, with permission.)

motor units. These motor unit fractions are usually seen in macro EMG as separate peaks. Macro MUP reflects the temporal, but not the spatial, features of the motor unit, both of which can be obtained with scanning EMG. A scanning EMG study was performed in some pathologic conditions.36 In primary myopathies the spatial dispersion in scanning EMG may be normal or diminished, and there may be silent areas in the territory. The individual MUPs obtained during the scanning procedure show the characteristic changes, either decrease in duration resulting from fiber loss or increase in duration resulting from late components, particularly in Duchenne type muscular dystrophy.

SFEMG and Other Techniques for Study of the Motor Unit

991

In reinnervation the action potentials are typically of increased amplitude but the anatomic territory is usually not increased to any significant extent. Because of volume conduction, the electrical activity from the motor unit may be recorded over a larger area than in the normal motor unit. This is in agreement with the observations in histologic studies in rat muscles,19 wherein the reinnervated motor units did not extend outside their original territory. Scanning EMG is a useful tool in studies of the anatomic arrangement of the motor unit. Compared with morphologic methods, it is better suited to investigate certain anatomic details of the motor unit. Comments about typical SFEMG, scanning EMG, and macro EMG findings in some nerve-muscle disorders are given below.

Macro Electromyography The spike component of a MUP recorded with a concentric needle electrode represents only a small portion of all muscle fibers in the motor unit. To obtain a recording from the largest possible number of muscle fibers in a motor unit, a nonselective recording technique, called macro EMG was developed.31,37 In this method a modified SFEMG electrode is used. The recording is made from a 15-mm leading segment of the cannula, the rest of it being insulated. A single-fiber action potential recorded by the built-in SFEMG surface triggers averaging of activity recorded by the bare part of the cannula to extract a global recording from the motor unit, the macro MUP (Fig. 37–5). (Further details are presented in the previous edition of this book.33) The amplitude and area of the signal reflect the electrical strength of the whole or at least the major part of the motor unit, encompassing the number and size of its fibers.21 Normal values differ with age and muscle.37 These mean values are obtained among the low-threshold motor units. The higher threshold motor units have higher amplitudes, in conformity with the size principle in recruitment of motor units.15 This cannot always be demonstrated with conventional needle EMG because that does not usually reflect the activity of the entire motor unit. Typically, macro EMG shows an increase in the size of the motor unit in neuropathies and a decrease in myopathies (Figs. 37–6 and 37–7), but exceptions are seen. Macro EMG provides quantitative information about the motor unit that cannot be obtained with other methods and may be useful, for example, to estimate the capacity for reinnervation in different disorders.32 In reinnervation FD and macro EMG usually increase in parallel, but in myopathies and in end stages of neurogenic conditions there may be a discrepancy, with high FD without an associated increase in macro MUP amplitudes. This may reflect a nonhomogeneous distribution of muscle fibers in the motor unit in the form of focal clusters.

FIGURE 37–5 Recording principle for macro EMG. A, The electrode is inserted into a motor unit. B, A SFEMG recording is made on one channel used for triggering. C, On the other channel, recording is made from the cannula of the electrode using a remote surface electrode as reference. The cannula signal is averaged after SFEMG triggering. D, The resulting potential is called a macro MUP. (From Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993, with permission.)

FINDINGS IN NERVE-MUSCLE DISORDERS SFEMG in Denervation-Reinnervation Denervation In denervation some muscle fibers generate fibrillation potentials. These are seen as spontaneous impulses generated from single muscle fibers. It has been observed in some recordings that spontaneous activity in one denervated muscle fiber triggers activity in another muscle fiber by ephaptic transmission.43,44 Similarly, when a denervated muscle fiber is stimulated electrically by means of intramuscular needle electrodes, adjacent muscle fibers may be activated ephaptically. Furthermore, using slowly increasing stimulus strengths, it can be demonstrated that discrete sites along the denervated muscle fiber are more excitable than others. Sometimes an electrically evoked

992

Nerve Conduction and Electromyography

FIGURE 37–6 Macro EMG from biceps brachii muscles in a patient with limb-girdle muscular dystrophy (A), a healthy subject (B), and a patient with amyotrophic lateral sclerosis (C). (From Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993, with permission.)

action potential in a denervated muscle fiber triggers an extra discharge in the same fiber, starting at some distance from the point of stimulation, probably at a site with locally increased excitability. The increased excitability in denervated muscle and the possibilities for ephaptic activation of adjacent muscle fibers may be responsible for so-called complex repetitive discharges seen in conditions in which denervation is present.54 It should be noted that sites of lower electrical threshold along the muscle fibers can also be demonstrated in normal muscle. Reinnervation Serial SFEMG investigations have been performed in patients after facial muscle transplantation to follow the innervation process from the time of operation.14 The implantation of a predenervated muscle has usually been performed in patients with persistent weakness after Bell’s palsy or after operation on the parotid gland. The implant attracted axonal sprouts from the surrounding normal muscle, and after 1 month voluntary EMG activity could be recorded. The FD increased during the first months as a result of innervation of groups of adjacent muscle fibers. Initially the recordings showed very abnormal jitter with

blocking as a result of uncertain conduction along the new nerve twigs and failing transmission across the motor end plates. After approximately 6 months most of the action potential complexes were stable, although the jitter was not normal in all recordings. The FD remained increased. This sequence of events has been the standard finding in patients followed during reinnervation after nerve damage. It is used as the basis for interpretation of SFEMG recordings from reinnervated muscles. The FD value is a measure of the local motor unit density, indicating the morphologic organization of the muscle fibers within the motor unit. The stability of the action potential complex is a measure of the transmission safety in terminal nerve branches, motor end plates, and muscle fibers (Fig. 37–8). It should be noted that these findings are translated to parameters in conventional EMG. Changes in FD are reflected in changed MUP amplitude and duration, whereas increased SFEMG jitter and blocking are seen as unstable MUP shape, called jiggle.40 It has been difficult to define SFEMG features that distinguish between denervation in progress and early reinnervation. Usually it is impossible to tell whether the abnormal recording is made during ongoing degeneration

SFEMG and Other Techniques for Study of the Motor Unit

FIGURE 37–7 Examples of macro EMG recordings from biceps brachii muscles in patients with uremic and diabetic neuropathy, Charcot-Marie-Tooth disease (CMT), Guillain-Barré syndrome (GBS), amyotrophic lateral sclerosis (ALS), spinal muscle atrophy (SMA), and old poliomyelitis. The amplitude of each macro MUP is shown by solid circles. Upper limits for individual normal macro MUP amplitudes are indicated. In the second patient with ALS, a number of macro MUPs were of low amplitude, probably as a result of peripheral deterioration of the motor unit. In the second patient with ALS, there was a general increase in amplitudes, indicating efficient reinnervation. (From Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993, with permission.)

FIGURE 37–8 SFEMG recording from the extensor digitorum muscle. A, Healthy subject; jitter 27 ␮s. B, Myasthenia gravis; jitter 140 ␮s (note blocking). C, Amyotrophic lateral sclerosis; jitter 210 ␮s and 340 ␮s for the second and fourth components, respectively. Note concomitant blocking of these two components (x and arrows) and independent blocking of the third component (o and arrow). D, Neuropathy; jitter 189 ␮s between triggering and succeeding component, also showing concomitant blocking (x and arrow). The jitter between the blocking components is 12 ␮s. (From Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993, with permission.)

993

994

Nerve Conduction and Electromyography

or in a phase of reinnervation. In some cases of GuillainBarré syndrome investigated within 2 weeks after the onset of symptoms, before reinnervated structures are likely to transmit impulses, increased jitter has been found, which probably indicates the uncertainty in conduction along the terminal nerve branches or in transmission across the motor end plates during a phase of demyelination or axonal degeneration.

SFEMG Signs of Abnormal Transmission in Nerve Branches During reinnervation the following phenomenon is sometimes seen. In consecutive discharges of a motor unit represented in the recording with three or more spike components, two of them may miss in a random succession, although simultaneously. This is called “concomitant blocking” (see Fig. 37–8C). The blocking components also show a common jitter in relation to the rest of the complex (i.e., concomitant jitter). This is interpreted as a conduction failure in a preterminal nerve branch.41 The degree of blocking sometimes increases during activity, particularly at high innervation rates. This type of block can recover partly after edrophonium chloride (Tensilon) is injected. Theoretically, such block should produce a decrementing compound muscle response at repetitive nerve stimulation, and the decrement should become less after Tensilon, features that simulate a neuromuscular transmission defect. This should be kept in mind as a possible explanation when a decrement occurs during repetitive nerve stimulation in cases with denervationreinnervation. In another phenomenon, bimodal jitter, one or more late spike components occur with dual latencies relative to the rest of the action potential complex. This is due to alternation between two different conduction times along a nerve branch. The cause is not fully clarified, but the phenomenon may occur when impulse conduction along the nerve alternates between the continuous and the saltatory mode during a certain stage of remyelination.47 In various neurogenic disorders the SFEMG abnormalities, particularly increased FD and jitter, are present to a variable degree and with different patterns of involvement. The individual pattern may be a clue to speed of progression and degree of involvement and so offers a tool for following spontaneous or treatment-induced changes over time.

Signs of Abnormal Nerve Excitability Signs of abnormal excitability are seen in various conditions. In some of them reinnervation has taken place. Extra discharges44 are often recorded in neurogenic disorders (e.g., motor neuron disease [ALS]), but also in disorders of muscle (e.g., myotonic dystrophy, tetany,

polymyositis). The extra discharge (usually one but sometimes two or three), occurring 4 to 10 ms after the main discharge, has a shape similar to that of the main discharge except for reduced amplitude and increased duration ascribed to relative refractoriness after the preceding impulse. When an extra discharge has occurred, the next main discharge may be missing. The site of generation of the extra discharge is uncertain. It is not in the muscle fiber because all fibers activated in the main discharge appear in the extra discharge, except when this is within the absolute refractory period for some of the fibers, in which case these fibers are blocked. The complete dropout of the next normal discharge after the extra discharge is evidence for an axonal origin. The dropout is probably due to a prolonged refractoriness in a weak point of the axon, perhaps at the site of origin of the extra discharge. The multiple discharges seen in neuromyotonia may be generated at multiple points along the axon.44,49 They may also be initiated in the terminal nerve tree, as suggested by the changing order of individual single-fiber action potentials on consecutive firing in a multispike SFEMG recording. This may be the most likely site, based on findings of abnormal potassium channels in this disorder.22 In certain circumstances ephaptic transmission may occur between axons. This has, for example, been demonstrated in patients with hemifacial spasm, in whom SFEMG has shown late responses of facial muscles with small jitter when peripheral nerve branches to other ipsilateral facial muscles were stimulated. The small jitter excluded a blink reflex, and disappearance immediately after decompressing operation excluded an axon reflex.24 The so-called A waves, seen during motor nerve conduction studies in various pathologic conditions, may be due to abnormal proximal axonal branching (e.g., at the site of a nerve lesion), producing axon reflexes. In other situations, they may be due to ephaptic transmission between axons. Most commonly, however, they seem to be due to reexcitation of an axon segment after the passage of an antidromic impulse.20 These mechanisms may be differentiated by special tests. We have seen the latter type as a transient phenomenon in the early stage of GuillainBarré syndrome, interpreted as a sign of membrane hyperexcitability, probably resulting from potassium channel block. Fasciculations can probably start at various sites in the motor unit. Some benign fasciculations may be generated in the muscle by the mechanism of ephaptic excitation between adjacent muscle fibers.44 The fasciculations related to cramp syndromes are probably generated in the spinal cord. Most of the fasciculations seen in motor neuron disease have a peripheral neural origin.4

SFEMG in Myasthenia Gravis A typical finding in patients with MG is that some motor end plates have normal jitter, others have slightly increased

SFEMG and Other Techniques for Study of the Motor Unit

jitter, and still others show intermittent impulse blocking30 (Fig. 37–9). This variation of findings may occur within one motor unit. The proportion of normal versus abnormal jitter in a muscle varies with the clinical involvement. Muscles with more pronounced fatigability show more abnormalities than do muscles without clinical involvement. The degree of blocking correlates with muscle weakness and with the reduction in amplitude of the evoked muscle response. As a rule, some recordings are abnormal in clinically normal muscles of a myasthenic patient but no blocking may be seen. With established criteria for abnormality, that is, more than two abnormal recordings out of a sample of 20 or a mean jitter of the 20 exceeding 34 ␮s, the following findings were obtained.38 In ocular MG, 85% and 59% of the patients were abnormal in frontalis and extensor digitorum communis (EDC) muscles, respectively. In mild generalized disease, 96% were abnormal. In moderate or severe generalized disease, 100% were abnormal; and among patients with clinical remission, 62% showed abnormalities in EDC muscles. If only one muscle is studied, the diagnostic yield is somewhat lower, particularly in ocular MG.25 The SFEMG abnormalities follow the clinical status with improvement after anticholinesterase therapy, corticosteroids, and plasmapheresis.18 This high sensitivity of SFEMG is due to the fact that transmission abnormality is seen as increased jitter before any blocking occurs. Thus jitter is often clearly increased in a muscle in which the RNS test is normal. However, SFEMG may detect blocking even when decrement studies are normal. Conversely, a decrementing response to RNS is only seen in muscles that also show blocking on SFEMG. The abnormal jitter increases with increase in temperature and decreases after administration of cholinesterase inhibitors. In some recordings the degree of blocking increases during continuous activity. Fiber density is, on

FIGURE 37–9 SFEMG recording in a case of myasthenia gravis. A, Normal jitter. B, Increased jitter without blocking. C, Increased jitter with blocking. (From Stålberg, E.: Clinical electrophysiology in myasthenia gravis. J. Neurol. Neurosurg. Psychiatry 43:622, 1980, with permission.)

995

average, slightly increased in MG.16 There is a difference between treated and untreated groups. The untreated group shows a slight, although significant, difference from controls. This may be due to structural changes in the motor end plate, which have been demonstrated morphologically.8 In patients treated with cholinesterase inhibitors, the FD is significantly higher than in the untreated group. This may be due to a direct neurotoxic effect of the drug, but other possibilities should be kept in mind, such as longer average duration and/or greater severity of the disease in the treated group. SFEMG is a more sensitive test for detecting the neuromuscular transmission defect in MG than repetitive nerve stimulation and is also more sensitive, although less specific, than present techniques for determining acetylcholine receptor antibody titer. It should be kept in mind that increased jitter is not pathognomonic for MG but is an indicator of the disturbed neuromuscular transmission. In LEMS the jitter is greatly increased and impulse blocking occurs frequently. In contrast to MG, the jitter tends to be more pronounced at low than at high voluntary firing or stimulation rates.56 A similar phenomenon is seen in botulinum intoxication.

SFEMG during Axonal Stimulation in MG and LEMS Axonal stimulation was performed in patients with MG and in those with LEMS to study short-term effects of various firing rates.56 In MG most motor end plates showed the lowest degree of transmission failure at the lowest rates of stimulation (0.5 and 1 Hz) and the most pronounced abnormality at an intermediate rate (5 or 10 Hz). At most, but not all, motor end plates, transmission improved at rates above 15 Hz. In contrast, a large majority

996

Nerve Conduction and Electromyography

of end plates in the patients with LEMS showed the expected, often dramatic, improvement of the abnormal jitter and blocking as the stimulation rate increased from 1 to 2 Hz to 15 or 20 Hz. Whether the myasthenic motor end plates exhibiting a pattern typical of LEMS suffer from an additional presynaptic defect is not settled. Antibodies against calcium channels have been found in addition to those against acetylcholine receptors in some patients with clinical MG.

Amyotrophic Lateral Sclerosis The typical SFEMG findings in ALS are increased FD associated with increased jitter and single spike or, less often, concomitant blocking. The jitter is always increased when lower motor neuron signs are present, even in muscles without clinical symptoms, and is always accompanied by increased FD in these muscles. This is interpreted as impaired impulse transmission in immature reinnervating structures. More recordings show increased jitter in patients with rapid progression of symptoms than in patients with slow progression. The degree of instability of the action potential complexes may therefore be useful in prognosis. As mentioned above, extra discharges are often seen.

Fasciculation potentials in ALS are irregular in shape and show increased jitter and blocking. FD is typically between 2 and 5 in patients with lower motor neuron involvement. It is generally higher in weak muscle than in strong but is abnormal also in clinically normal muscles. In the SFEMG study46 it was shown that the FD did not differ significantly among the biceps brachii (3.9), EDC (4.3), and first dorsal interosseus (3.9) muscles. The values in these muscles were higher in patients whose symptoms had started in the upper limbs. Furthermore, there was a negative correlation between force and FD. The reinnervation indicated by increased FD did not fully compensate for the loss of neurons. Macro EMG shows normal action potentials in patients with purely bulbar symptoms but usually increased amplitudes in patients with lower motor neuron involvement. The amplitude is inversely correlated with the strength of the muscle. The median amplitude may be up to 3 times the normal mean and individual motor units may be 10 times the normal mean or 3 times the largest motor units seen in normal muscle. Scanning EMG shows irregular MUPs with long duration (Fig. 37–10). The total area over which the spike components of the MUP occur has not been increased in

FIGURE 37–10 Examples of scanning EMG recordings in a healthy subject (tibialis anterior muscle) and a patient with ALS (triceps brachii muscle). The territory is only slightly increased in the latter recording compared with normal. Note the irregular shape of the motor unit potentials in ALS. (From Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993, with permission.)

SFEMG and Other Techniques for Study of the Motor Unit

the patients studied so far. Because of volume conduction, the MUP amplitude can be recorded over a larger area than in the normal muscle.

Syringomyelia An SFEMG study compared the biceps brachii (representing C5/6), EDC (C7/8), and first dorsal interosseus (C8/T1) muscles in 10 patients with syringomyelia.28 The action potential complexes were usually more stable than in ALS. The jitter values differed for the different muscles. Increased jitter was noticed in 1% of the action potentials in the biceps brachii, in 16% in the EDC, and in 21% in the first dorsal interosseus muscle. Blocking was seen in these muscles in 0.7%, 7%, and 7% of the action potentials, respectively. Late components, occurring later than 9 ms after the triggering potential, usually had greater jitter and more frequent blocking than earlier components. The FD was increased in at least one muscle in all cases, but again there were differences between the muscles. The mean value was 1.9 in the biceps brachii muscle, in which 4 of the 10 patients had normal values, 3.6 in the EDC, and 4.4 in the first dorsal interosseus muscle, with no patients having normal values for the latter two muscles. In some muscles, no recordings could be obtained as a result of severe involvement. These observations indicate a relatively constant pattern of involvement of the anterior horn cells in the brachial segments in syringomyelia. Macro EMG and scanning EMG have not been performed in these patients.

Postpolio Sequelae In patients with a history of polio, both FD and macro EMG amplitudes are increased considerably in clinically involved muscles, but not in direct correlation to weakness. Individual macro MUP amplitudes were often 10 times, and sometimes more than 20 times, the normal median value. Muscle biopsy showed an increase in mean area of type I fibers by two to four times; these are the fibers of low-threshold motor units studied with macro EMG. In a longitudinal study of 20 patients with a history of polio, investigations were made 4 and 8 years after the initial study. The macro MUP amplitude was increased in all legs in which macro EMG was performed. It was on average 16.7 ⫾ 2.1 times control (range, 3 to 42) at the 8-year examination. At the end of the 8-year period, 20 of 26 legs showed increase in macro MUP amplitudes. When the cohort was divided into smaller groups, it was seen that, in those with initially large macro MUP amplitude (more than 20 times the normal), the amplitude usually decreased, whereas it increased in those with smaller initial values. This was taken as an indication of incompletely compensated age-related neuron loss. It should be noted that biopsy studies in these patients show a general fiber hypertrophy (approximately twice the normal

997

diameter). Therefore, the decrease in macro MUP amplitude with time is not due to a decrease in fiber diameters. As yet, patients with so-called postpolio syndrome have not shown any distinct neurophysiologic pattern with these techniques. With the above-mentioned marked reduction in the motor neuron pool, it is reasonable to assume that superimposition of the age-dependent demand for further reinnervation may have dramatic effects when capacity has already been utilized excessively. However, considering the fact that new weakness often occurs before physical aging becomes evident in these muscles and is not associated with any significant changes in motor unit size, other factors such as cardiovascular and musculoskeletal (disuse atrophy, joint problems) may also play important roles. In other chronic neurogenic disorders such as ALS, spinal muscular atrophies, Guillain-Barré syndrome, and in posttraumatic spinal cord conditions, it is often impossible to determine whether a change in clinical status is due to the primary disease or to other factors (e.g., age related). Postpolio syndromes are of particular interest as a possible model to study age-related changes in motor innervation of muscles.

Guillain-Barré Syndrome The diagnosis and progress of Guillain-Barré syndrome is electrophysiologically best assessed by a set of nerve conduction parameters reflecting the events during the initial and later stages of the disease. Studies using various EMG methods that may add further information about the motor unit are exemplified here. During the first week the MUPs recorded with concentric needle EMG electrode may be normal, but the interference pattern is reduced in proportion to muscle weakness. In some patients the macro MUP amplitude was increased as early as the eighth day after the onset of symptoms without an associated increase in FD; at the same time the surface-recorded compound muscle action potential elicited by supramaximal nerve stimulus was decreased. This is interpreted as a dropout of the smaller motor units, probably as a result of a conduction failure in the smaller axons. Loss of these low-threshold motor units will result in an earlier recruitment of larger but still normal motor units. In a few cases recordings have shown increased jitter within 14 days after onset, before reinnervation has taken place. The abnormal jitter may be due to uncertain impulse transmission during the phase of demyelination. Fibrillation potentials show that axonal involvement has taken place. The ensuing reinnervation can be studied by means of conventional EMG and with other techniques. SFEMG shows increased jitter when reinnervating axons become capable of conducting and transmitting impulses. The FD increases as long as active reinnervation continues.

998

Nerve Conduction and Electromyography

In the final stage, the jitter becomes more normal but the FD remains abnormal. In this stage the macro MUP gives information about the total size of the motor unit, that is, about the amount of reinnervation that has taken place (Fig. 37–11). If fibrillation potentials do not occur, this indicates that the process is mainly demyelinating and that no more than an insignificant number of axons has been damaged. There is a pronounced slowing of nerve conduction (motor, sensory, F waves). If conduction block occurs, the amplitude and the area of the compound motor and/or sensory action potentials decrease. The motor unit potentials may be normal, but the interference pattern is reduced, reflecting the nerve conduction block. The FD should not increase during the process because no reinnervation takes place without preceding denervation. It is important that several different electrophysiologic parameters be used to characterize the lesion in GuillainBarré syndrome to reach a reliable diagnosis, predict the future course and recovery, and quantify the effect of treatment procedures (e.g., plasmapheresis). SFEMG and macro EMG give additional information beyond that obtained with conventional techniques.

Hereditary Motor and Sensory Neuropathy: Charcot-Marie-Tooth Disease During the last decade, the individual types within the classification of hereditary motor and sensory neuropathies, which is based on clinical and morphologic criteria, have been shown to contain a variety of underlying genetic defects. Nevertheless, the different genetic entities are still found to share common neurophysiologic characteristics. In hereditary motor and sensory neuropathy type 1, FD may be greatly increased in the tibialis anterior muscle and to a lesser degree in the biceps brachii muscle. Jitter is typically increased in 25% to 50% of the recordings in the tibialis anterior muscle and less in the biceps brachii muscle. Impulse blocking may be seen. Macro EMG showed increased MUP amplitudes, but in muscles with pronounced atrophy and fibrotic changes, the mean amplitude was only moderately increased, although the FD values were high. In a study of patients with Charcot-Marie-Tooth disease types 1 and 2 (CMT1 and CMT2), the former showed typical neurogenic changes. CMT2 showed a histopathologic pattern of fiber diameter variation with hypertrophic fibers with internal

FIGURE 37–11 Macro EMG in a patient with Guillain-Barré syndrome 14 days after onset, with normal macro MUPs, and 9 months after onset, with increased macro MUPs indicating reinnervation (tibialis anterior muscle). (From Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993, with permission.)

SFEMG and Other Techniques for Study of the Motor Unit

nuclei and splitting, with some appearing as “whorled fibers.” Atrophic fibers, sometimes angulated, were found in groups; some seemed to be regenerating fibers. There was an increased amount of connective tissue. These findings were interpreted as myopathy rather than neuropathy.9 Furthermore, macro MUP amplitudes were increased by 2 to 14 times in CMT1 and by less than 1 to 8 times above normal in CMT2.10 This may indicate a reduced reinnervation capacity.

Diabetes Mellitus In 30 of 36 unselected patients, neuropathy was present as indicated by motor and sensory conduction velocity studies in ulnar, median, peroneal, tibial, and sural nerves and in F responses of the motor nerves. FD was increased in the EDC or biceps brachii muscles in 15% of the 26 cases with signs of neuropathy in the arm. FD was increased in the tibialis anterior muscle in 37% of the 27 cases with signs of neuropathy in the leg. The abnormal FD values were found among those patients with the most pronounced signs of neuropathy. Macro EMG showed increased mean MUP values in 1 of 10 cases in the biceps brachii and in 3 of 10 cases in the tibialis anterior. These patients also had increased FD. These findings are in accordance with an earlier report of only slight signs of reinnervation, at least in the arm muscles, in patients with diabetic neuropathy.48 In a study of 53 patients (34 non–insulin dependent) with diabetes, macro EMG was found to be increased, more so in patients with other signs of neuropathy. The strength of ankle and knee extensors was inversely related to FD and macro MUP amplitude of the tibialis anterior and vastus lateralis muscles.1

Alcoholic Neuropathy In a study of 16 consecutive cases of alcoholism, neuropathy was found in 9 patients using the conventional neurographic criteria. FD was increased in the biceps brachii muscle in four of five with neuropathy in the arms and in three of nine with neuropathy in the legs. Macro EMG was performed in nine of the patients. It was increased in three of them, in both the biceps brachii and tibialis anterior muscles. In an earlier study48 the FD of the EDC muscle was reported to be abnormal (more than 3 standard deviations above the normal mean) in 8 of 18 patients. Scanning EMG has not been performed.

Primary Myopathies It is beyond the scope of this chapter to discuss the findings in myopathies. FD is increased in many cases, but the macro MUP is typically decreased and scanning EMG shows short-duration MUPs, normal or small territory of the motor unit, and sometimes “silent areas” within the

999

territory. The combined result of the recordings is therefore quite different from those obtained in neurogenic lesions.

SUMMARY By combining SFEMG, concentric needle EMG, scanning EMG, and macro EMG, new insight can be gained into the motor unit. This approach enables the description of the microphysiology as well as the anatomic arrangement of muscle fibers in a motor unit and the motor unit size. The electrophysiologic findings help the physician reach a diagnosis and assess the current status of the neuromuscular system. The electrophysiologic tests available today can provide a more detailed description of the physiologic condition of the motor unit than has been previously possible, allowing for better understanding of the pathophysiologic processes in various neuromuscular disorders and more adequate appreciation of the clinical status of the patient.

ACKNOWLEDGMENT This work was supported by the Swedish Medical Research Council (Grant 135) and the Slovenian Ministry of Education, Science and Sports (J. T.).

REFERENCES 1. Andersen, H., Stålberg, E., Gjerstad, M. D., and Jakobsen, J.: Association of muscle strength and electrophysiological measures of reinnervation in diabetic neuropathy. Muscle Nerve 21:1647, 1999. 2. Ashby, P., Stålberg, E., and Hilton-Brown, P.: Afferent projections to human tibialis anterior motor units active at various levels of muscle contraction. Acta Physiol. Scand. 127:523, 1986. 3. Buchthal, F., Guld, C., and Rosenfalck, P.: Multielectrode study of the territory of a motor unit. Acta Physiol. Scand. 39:83, 1957. 4. Conradi, S., Grimby, L., and Lundemo, G.: Pathophysiology of fasciculations in ALS as studied by electromyography of single motor units. Muscle Nerve 5:202, 1982. 5. Drost, G., Blok, J., Stegeman, D., et al.: Propagation disturbance of motor unit action potentials during transient paresis in generalized myotonia: a high-density surface EMG study. Brain 124:352, 2001. 6. Ekstedt, J., Nilsson, G., and Stålberg, E.: Calculation of the electromyographic jitter. J. Neurol. Neurosurg. Psychiatry 37:526, 1974. 7. Ekstedt, J., and Stålberg, E.: Abnormal connections between skeletal muscle fibers. Electroencephalogr. Clin. Neurophysiol. 27:607, 1969. 8. Engel, A. G., Tsuyihata, M., Lindström, J. M., et al: The motor end-plate in myasthenia gravis and in experimental

1000

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24. 25. 26.

Nerve Conduction and Electromyography

autoimmune myasthenia gravis: a quantitative ultrastructural study. Ann. N. Y. Acad. Sci. 274:60, 1976. Ericson, U., Ansved, T., and Borg, K.: Charcot-Marie-Tooth disease: muscle biopsy findings in relation to neurophysiology. Neuromuscul. Disord. 8:175, 1998. Ericson, U., Borg, J., and Borg, K.: Macro-EMG and muscle biopsy of paretic foot dorsiflexors in Charcot-Marie-Tooth disease. Muscle Nerve 23:217, 2002. Gath, I., and Stålberg, E.: Measurements of the uptake area of small-size electromyographic electrodes. IEEE Trans. Biomed. Eng. 26:374, 1979. Gilchrist, J., Barkhaus, P. E., Bril, V., et al: Single fiber EMG reference values: a collaborative effort. Muscle Nerve 15:151, 1991. Gilchrist, J., Massey, J., and Sanders, D. B.: Single fiber EMG and repetitive nerve stimulation of the same muscle in myasthenia gravis. Muscle Nerve 17:171, 1994. Hakelius, L., Nyström, B., and Stålberg, E.: Histochemical and neurophysiological studies of autotransplanted cat muscle. Scand. J. Plast. Reconstr. Surg. 9:15, 1975. Henneman, E., Somjen, G., and Carpenter, D. O.: Functional significance of cell size in spinal motor neurones. Ann. Plast. Surg. 28:560, 1965. Hilton-Brown, P., Stålberg, E., and Osterman, P.-O.: Signs of reinnervation in myasthenia gravis. Muscle Nerve 5:215, 1982. Hilton-Brown, P., Stålberg, E., Trontelj, J. V., and Mihelin, M.: Causes of the increased fibre density in muscular dystrophies studied with single fibre EMG during electrical stimulation. Muscle Nerve 8:383, 1985. Howard, J. F., and Sanders, D. B.: Serial single-fiber EMG studies in myastheniac patients treated with corticosteroids and plasma exchange therapy. Muscle Nerve 4:254, 1981. Kugelberg, E., Edström, L., and Abbruzzese, M.: Mapping of motor units in experimentally reinnervated rat muscle: interpretation of histochemical and atrophic fibre patterns in neurogenic lesions. J. Neurol. Neurosurg. Psychiatry 33:319, 1970. Magistris, M. R., and Roth, G.: Motor axon reflex and indirect double discharge: ephaptic transmission? A reappraisal. Clin. Neurophysiol. 85:124, 1992. Nandedkar, S. D., and Stålberg, E.: Simulation of macro EMG motor unit potentials. Electroencephalogr. Clin. Neurophysiol. 56:52, 1983. Newsom-Davies, J.: Immunological association of acquired neuromyotonia (Isaac’s syndrome): report of five cases and literature review. Brain 116:543, 1993. Rossini, P. M., Stålberg, E., Winkler, T., and Zarola, F.: Motor responses to transcranial brain stimulation: evaluation of premovement facilitation by surface, coaxial needle and single fibre recordings. In Rossini, P. M., and Marsden, C. D. (eds.): Non-invasive Stimulation of Brain and Spinal Cord: Fundamentals and Clinical Applications. New York, Alan R. Liss, p. 105, 1988. Sanders, D. B.: Ephaptic transmission in hemifacial spasm: a single-fiber EMG study. Muscle Nerve 12:690, 1989. Sanders, D. B.: Clinical impact of single-fiber electromyography. Muscle Nerve Suppl. 11:S15, 2002. Schiller, H., and Stålberg, E.: F-responses studied with single fibre EMG in normal subjects and spastic patients. J. Neurol. Neurosurg. Psychiatry 41:45, 1978.

27. Schwartz, M. S., Stålberg, E., Schiller, H. H., and Thiele, B.: The reinnervated motor unit in man: a single fibre EMG multielectrode investigation. J. Neurol. Sci. 27:303, 1976. 28. Schwartz, M. S., Stålberg, E., and Swash, M.: Pattern of segmental motor involvement in syringomyelia: a single fibre EMG study. J. Neurol. Neurosurg. Psychiatry 43:150, 1980. 29. Stålberg, E.: Propagation velocity in single human muscle fibres. Acta Physiol. Scand. Suppl. 287:1, 1966. 30. Stålberg, E.: Clinical electrophysiology in myasthenia gravis. J. Neurol. Neurosurg. Psychiatry 43:622, 1980. 31. Stålberg, E.: MACRO EMG, a new recording technique. J. Neurol. Neurosurg. Psychiatry 43:475, 1980. 32. Stålberg, E.: Use of single fibre EMG and macro EMG in study of reinnervation. Muscle Nerve 13:804, 1990. 33. Stålberg, E.: Single-fiber electromyography and some other electrophysiologic techniques for the study of the motor unit. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed., Vol. 1. Philadelphia, W. B. Saunders, p. 645, 1993. 34. Stålberg, E., and Antoni, L.: Electrophysiological cross section of the motor unit. J. Neurol. Neurosurg. Psychiatry 43:469, 1980. 35. Stålberg, E., and Antoni, L.: Computer aided EMG analysis. In Desmedt, J. (ed.): Progress in Clinical Neurophysiology, Vol. 10: Computer Aided Analysis of EMG. Basel, Karger, p. 186, 1983. 36. Stålberg, E., and Dioszeghy, P.: Scanning EMG in normal muscle and in neuromuscular disorders. Electroencephalogr. Clin. Neurophysiol. 81:403, 1991. 37. Stålberg, E., and Fawcett, P. R. W.: Macro EMG changes in healthy subjects of different ages. J. Neurol. Neurosurg. Psychiatry 45:870, 1982. 38. Stålberg, E., and Sanders, D. B.: Electrophysiological tests of neuromuscular transmission. In Stålberg, E., and Young, R. (eds.): Clinical Neurophysiology. London, Butterworths, p. 88, 1981. 39. Stålberg, E., Schiller, H., and Schwartz, M.: Safety factor in single human motor end-plates studied in vivo with single fibre electromyography. J. Neurol. Neurosurg. Psychiatry 38:799, 1975. 40. Stålberg, E., and Sonoo, M.: Assessment of variability in the shape of the motor unit action potential, the “jiggle,” at consecutive discharges. Muscle Nerve 17:1135, 1994. 41. Stålberg, E., and Thiele, B.: Transmission block in terminal nerve twigs: a single fibre electromyographic finding in man. J. Neurol. Neurosurg. Psychiatry 35:52, 1972. 42. Stålberg, E., and Thiele, B.: Discharge patterns of motor neurones in humans. In Desmedt, J. (ed.): New Developments in Electromyography and Clinical Neurophysiology. Basel, Karger, p. 234, 1973. 43. Stålberg, E., and Trontelj, J. V.: Abnormal discharges generated within the motor unit as observed with single fibre electromyography. In Ochoa, J. L., and Culp, B. (eds): Abnormal Nerves Muscles as Impulse Generators. Oxford, UK, Oxford University Press, p. 443, 1982. 44. Stålberg, E., and Trontelj, J. V.: Single Fiber Electromyography in Healthy and Diseased Muscle. New York, Raven Press, 1994. 45. Stålberg, E., Trontelj, J. V., and Mihelin, M.: Electrical microstimulation with single-fiber electromyography: a

SFEMG and Other Techniques for Study of the Motor Unit

46.

47.

48.

49.

50.

51. 52.

53.

useful method to study the physiology of the motor unit. J. Clin. Neurophysiol. 9:105, 1992. Swash, M.: Vulnerability of lower brachial myotomes in motor neurone disease: a clinical and single fibre EMG study. J. Neurol. Sci. 47:59, 1980. Thiele, B., and Stålberg, E.: The bimodal jitter: a single fibre electromyographic finding. J. Neurol. Neurosurg. Psychiatry 37:403, 1974. Thiele, B., and Stålberg, E.: Single fibre EMG findings in polyneuropathies of different aetiology. J. Neurol. Neurosurg. Psychiatry 38:881, 1975. Torbergsen, T., Stålberg, E., and Brautaset, M. D.: Generator sites for spontaneous activity in neuromyotonia: an EMG study. Electroencephalogr. Clin. Neurophysiol. 101:69, 1996. Troni, W., Cantella, R., and Rainero, I.: Conduction velocity along human muscle fibers in situ. Eur. J. Pharmacol. 33:1453, 1983. Trontelj, J. V.: H-reflex of single motoneurons in man. Nature 220:1043, 1968. Trontelj, J. V., Mihelin, M., Fernandez, J. M., and Stålberg, E.: Axonal stimulation for end-plate jitter studies. J. Neurol. Neurosurg. Psychiatry 49:677, 1986. Trontelj, J. V., Mihelin, M., and Khuraibet, A.: Safety margin at single neuromuscular junctions. Muscle Nerve 21(Suppl. 11):S21, 2002.

1001

54. Trontelj, J. V., and Stålberg, E.: Bizarre repetitive discharges recorded with single fibre EMG. J. Neurol. Neurosurg. Psychiatry 46:310, 1983. 55. Trontelj, J. V., and Stålberg, E.: Responses to electrical stimulation of denervated human muscle fibre recorded with single fibre EMG. J. Neurol. Neurosurg. Psychiatry 46:305, 1983. 56. Trontelj, J. V., and Stålberg, E.: Single motor end-plates in myasthenia gravis and LEMS at different firing rates. Muscle Nerve 14:226, 1991. 57. Trontelj, J. V., and Stålberg, E.: Jitter measurement by axonal stimulation: guidelines and technical notes. Electroencephalogr. Clin. Neurophysiol. 85:30, 1992. 58. Trontelj, J. V., and Stålberg, E.: Single fiber and macroelectromyography. In Bertorini, T. E. (ed.): Clinical Evaluation and Diagnostic Tests for Neuromuscular Disorders. Woburn, MA, Butterworth-Heineman, p. 417, 2002. 59. Trontelj, J. V., Stålberg, E., and Mihelin, M.: Jitter in the muscle fibre. J. Neurol. Neurosurg. Psychiatry 53:49, 1990. 60. Trontelj, J. V., Stålberg, E., Mihelin, M., and Khuraibet, A. J.: Jitter of the stimulated motor axon. Muscle Nerve 15:449, 1992. 61. Trontelj, J. V., Trontelj, M., and Stålberg, E.: The jitter of single human muscle fibre responses in certain reflexes. Electroencephalogr. Clin. Neurophysiol. 34:825, 1973. 62. Zidar, J., Trontelj, J. V., and Mihelin, M.: Percutaneous stimulation of human corticospinal tract: a single fibre EMG study of individual motor unit responses. Brain Res. 422:196, 1987.

38 Single-Unit Recordings of Afferent Human Peripheral Nerves by Microneurography ERIK TOREBJÖRK AND MARTIN SCHMELZ

Technique of Microneurography Intraneural Stimulation Sensory Innervation of the Skin Low-Threshold Mechanoreceptors Thermoreceptors Nociceptors

Low-Threshold C Fibers Modulation of Discharge Frequency by the Axon Disorders of Sensory Nerve Function Paresthesias and Spontaneous Activity in Myelinated Fibers

Recording synchronized discharges of myelinated fibers as done in clinical nerve conduction studies is a standard tool for the diagnosis of impaired impulse conduction in peripheral neuropathies. The transcutaneously recorded compound action potential is generated by many superimposed action potentials of myelinated axons. This method provides clinically useful information about average conduction velocity and helps to characterize and localize nerve lesions. However, no information about the behavior of single axons can be deduced from the signal. Moreover, no transcutaneous compound action potential can be recorded from unmyelinated axons because of the small currents and the extensive dispersion caused by highly variable conduction velocities. By comparison, direct recordings from unmyelinated axons can be obtained in microneurography with microelectrodes placed in close proximity to the axon. For this latter purpose, thin tungsten needle electrodes (diameter of 0.2 mm) are placed inside nerve fascicles, which enables recording not only of synchronized volleys, but also of single-unit discharges of myelinated and unmyelinated fibers. Thereby the individual firing behavior in response to various external stimuli can be assessed. Because subjects and patients are awake during the recording, microneurography provides a unique tool to correlate the discharge behavior of afferent nerve fibers with the sensation evoked by defined stimuli. Because the method is invasive and time consuming, it is not used as a routine clinical test. Instead, it has been

Neuropathic Changes in Unmyelinated Fibers Sensitization of C Fibers in Erythromelalgia Future Perspectives in Microneurography

used since the 1960s as a research tool to generate basic physiologic knowledge and insights into pathophysiologic alterations in disease. Although restricted endoneural lesions can be induced by the recording electrode, only very few complications in thousands of microneurographic recordings have been found in the past 40 years.18 In this chapter we summarize important contributions of microneurographic recordings to the understanding of the physiology and pathophysiology of sensory skin innervation.

TECHNIQUE OF MICRONEUROGRAPHY In the standard analysis, action potentials are attributed to a single fiber according to the specific shape of its action potential. Thereby, the neuronal discharge can be quantified as total number and frequency of spikes. However, this type of analysis is restricted to high-amplitude units, mainly found in recordings of myelinated afferents. Unmyelinated afferents have smaller action potentials, and their close relation to neighboring axons in a Remak bundle hampers the univocal differentiation between C fibers with similar action potential shape. As a complementary method to assess single-fiber responsiveness, the “marking” technique has been developed. This technique utilizes the pronounced activity-dependent hyperpolarization of C fibers (causing slowing of conduction velocity) to identify their activation (Fig. 38–1). It allows simultaneous recordings of several 1003

1004

Nerve Conduction and Electromyography

C-nociceptor mechanoinsensitive 550

60 C-nociceptor polymodal

525 C-nociceptor polymodal

50 35

500

C-cold receptor 30 0.125 Hz

0.25 Hz

2 Hz

0.5 Hz

Response latency, ms

Response latency, ms

575

0.125 Hz

Time FIGURE 38–1 Response latency of action potentials between electrical stimulation of the receptive endings in the skin and recordings in the peripheral nerve (n. peroneus). Left, Electrical stimulation at increasing frequency provokes increased response latency for different classes of afferent C fibers (activity-dependent slowing). The activity-dependent slowing is most pronounced in mechano-insensitive C nociceptors, which slow down in conduction velocity even at low stimulatory frequencies of 0.125 or 0.25 Hz. The slowing of traditional mechano-responsive C nociceptors (“polymodal”) is far less pronounced, clearly separating the two nociceptor classes. (Data from Weidner, C., Schmelz, M., Schmidt, R., et al.: Functional attributes discriminating mechano-insensitive and mechanoresponsive C nociceptors in human skin. J. Neurosci. 19:10184, 1999.) Right, At higher stimulation frequencies of 2 Hz, polymodal nociceptors show a pronounced activity-dependent slowing that clearly separates them from cold-sensitive C fibers (C-cold receptors), which only slightly increase their response latency when stimulated at 2 Hz. (Data from Serra, J., Campero, M., Ochoa, J., and Bostock, H.: Activity-dependent slowing of conduction differentiates functional subtypes of C fibres innervating human skin. J. Physiol. [Lond.] 515:799, 1999.)

fibers, albeit with a lower temporal resolution and generation of only semiquantitative information about unitary activity.

INTRANEURAL STIMULATION With the microelectrode placed in close relation to adjacent axons, low-intensity currents of intraneural microstimulation can be given.54,67–69,85 For liminal stimulation of myelinated fibers, different sensory perceptions are generated (tapping, vibration, pressure), which can then be correlated with the type of afferent units recorded before. Recently, the successful combination of microneurography, intraneural microstimulation, and functional magnetic resonance imaging provided information about unexpectedly strong cortical activation of primary and secondary somatosensory cortex upon stimulation of single identified myelinated afferents.49,87 Cortical activation was even observed at stimulus frequencies that did not induce any sensation. This new observation corroborates the high transmission security from primary afferents to spinal projection neurons.64 Because intraneural microstimulation skips activation of the receptor and distal axon, it can also be employed in amputees, activating afferent fibers and assessing

the induced sensations. Unexpectedly, intraneural microstimulation of nerves innervating amputated limbs elicited perceptions that did not differ in quality or location from those in controls, suggesting a low degree of cortical plasticity.51 Intraneural microstimulation has also been used to elicit unitary sensation in response to C-fiber activation. However, because of the close relation of several axons in a Remak bundle, we do not think that a single C afferent can be excited electrically. 57,94 Functional clustering of axons in the peripheral nerve may explain the fact that unitary sensations may be elicited despite activation of several axons. Anatomic segregation of axons of similar sensory characteristics and innervation territories has already been shown in myelinated fibers.22–25,34,102,103 However, the significance of functional and topographic segregation in peripheral nerves is not yet clear.

SENSORY INNERVATION OF THE SKIN An overview of sensory innervation is provided in Chapter 8. Here we only summarize microneurography results in humans.

Single-Unit Recordings of Afferent Human Peripheral Nerves by Microneurography

Low-Threshold Mechanoreceptors Low-threshold mechanoreceptors of myelinated fibers have been classified as fast-adapting (FA) and slow-adapting (SA) units according to their response to continuous skin indentation. In glabrous skin, units with small receptive fields (FA I and SA I) can be separated from those with large receptive fields with diffuse borders (FA II and SA II).21 Morphologically FA I units correspond to Meissner’s corpuscles, FA II units to pacinian corpuscles, SA I units to Merkel cell complexes, and SA II units to Ruffini’s nerve endings.90 Because Merkel cell complexes have several generator sites, irregular spike frequency is observed at low stimulus intensities; in contrast, Ruffini’s endings generate activity of regular frequency according to a supposed single generator site.28 The innervation of palmar and plantar glabrous skin is similar, but larger innervation territories for type I units and higher mechanical thresholds for SA I, FA I, and FA II units were found.86 These differences and also lower innervation density may well explain lower spatial acuity in the foot as compared to the palms. There are no Meissner’s corpuscles in nonglabrous skin, but rapidly adapting mechanosensors innervate the hair follicles (hair units).92 However, there are also FA I units in nonglabrous skin that are not linked to hairs.86 The morphologic correlate of these units is still unclear. Functional Significance Based on the basic receptive properties of the different types of mechanoreceptors, their particular roles in encoding different kinds of skin stimuli have been explored. It is obvious that the high innervation density of FA I and SA I units having small innervation territories promotes high spatial acuity, as exemplified in the fingertips.21 Similarly, the optimum stimulation frequency of FA I (8 to 64 Hz) and FA II (⬎64 Hz) units correlates to the sensation that is linked to stimulation of these unit types (flutter vs. vibration). Beyond the traditional stimulation paradigms, more complex stimuli have been investigated recently. These include moving stimuli with well-controlled velocity, force, direction, and shape.3,14,20 Moreover, the functional significance of skin mechanoreceptors beyond exteroception has been explored. Stretch receptors in skin respond to joint movements and have a major role in sense of joint position,13 which traditionally has been assumed to be a function of joint receptors. The role of low-threshold mechanoreceptors thus should not be restricted to sensing touch and vibration. Instead, the information provided by these receptors is used more generally as a feedback system that is crucially involved in motor control mechanisms for grip force30,31 and speech and jaw movements.32,33,88,89

Thermoreceptors Thermoreceptors can be separated into receptors for warmth and cold detection. According to results of

1005

differential nerve blocks and response latencies, the warmth sensation has been attributed to C fibers, whereas cold detection is a function of A␦ fibers.15,105 Microneurographic recordings from A␦ fibers are sparse, and so far only a few examples of human A␦ cold units have been published.26,27,29 Data on human warm fibers have been described in several papers.27,37,38,84 They are mechano-insensitive and have small innervation territories. They are activated by moderate warming, but may also encode increasing temperature into the noxious range. Their low number and small receptive fields result in a sparse innervation for warmth. This may explain early impairment of warmth detection in peripheral neuropathy as compared to heat-pain thresholds, which may increase at a later stage of the disease. The phenomenon of paradoxical hot sensation upon mild cooling under a differential A-fiber block has provided evidence for cold-specific C fibers.81 Also, the existence of cold-sensitive C fibers has been suggested as the explanation of a heat-pain illusion occurring on simultaneous stimulation with non-noxious warm and cold (thermal grill illusion,11,12 Thunberg effect83). Recently, recordings of C fibers responsive to mild cooling have been identified in humans.8 Their activation thresholds were about 29 °C, which is compatible with a role of this class of C fibers in the paradoxical hot sensation. Interestingly, not only did these fibers differ in their receptive properties, but their axonal characteristics also clearly distinguished them from C nociceptors. Activity-dependent hyperpolarization of axons, which leads to slower conduction velocities, was much less pronounced in C-cold fibers as compared to the nociceptors (see Fig. 38–1).

Nociceptors A␦ Nociceptors Thinly myelinated nociceptors with a conduction velocity of about 20 m/s have been recorded in human radial nerve.1 According to their receptive properties, they can be separated into a high-threshold mechanoreceptive class and a mechano–heat-sensitive class. Activation of A␦ nociceptors underlies the sharp pricking pain inflicted in noxious mechanical and heat stimulation. Mechanoinsensitive A␦ chemonociceptors have also been found in the monkey that have a chemical response pattern62 similar to that of mechano-insensitive C fibers.73 Moreover, differentiation between A␦ and C fibers in the periphery can be complicated by the fact that A␦ fibers can have surprisingly long unmyelinated distal branches in the skin of up to 5 cm.60 C-Nociceptor Function Unmyelinated nociceptors comprise the most numerous class of somatic afferent nerve fibers. They subserve a variety of different sensory functions as described below. However,

1006

Nerve Conduction and Electromyography

their physiologic role also includes efferent release of neuropeptides into the peripheral tissues, needed for homeostasis (including trophic effects and modulation of the immune system).71 Thus thin-fiber neuropathy not only will result in sensory deficits but may also include trophic changes and impaired immune responses. Polymodal C-Nociceptors. The majority of unmyelinated nociceptors are activated by mechanical, chemical, and thermal stimuli. Thus they have been termed polymodal nociceptors.61 Their activation thresholds for heat match the heat-pain threshold of humans.63 In contrast, their mechanical activation thresholds are fairly low and do not correlate to mechanical pain thresholds. Inflammatory mediators such as bradykinin or prostaglandins are known to sensitize polymodal nociceptors to heat and thus probably underlie the heat hyperalgesia observed in inflamed tissues. C–Mechano-insensitive Nociceptors. Unmyelinated nociceptors unresponsive to mechanical stimulation75 differ from conventional polymodal nociceptors in their receptive properties, their biophysical characteristics, and their function. They have higher activation thresholds for heat and are not activated even by intense mechanical stimuli.75 Their innervation territories in the leg are larger77 (6 vs. 2 cm2) and conduction velocity is lower (0.8 vs. 1 m/s) than in polymodal fibers.75 Most interestingly, their high transcutaneous electrical thresholds and their activity-dependent hyperpolarization by far exceed the values observed in polymodal nociceptors. However, axonal properties analysis of activity-dependent hyperpolarization allows classification of these two nociceptor classes97 and thus predicts sensory properties of their endings. It clearly separates C-cold, C-polymodal, and C–mechano-insensitive fibers (see Fig. 38–1). It should be pointed out that this unexpected correlation between specific axonal properties and characteristics of sensory endings has a variety of implications. Because mechanisms of activity-dependent hyperpolarization are currently being clarified on a molecular level,65 immunohistochemistry might in the future enable differential staining and functional identification of axons. Early clinical results confirm that this approach can be used to improve characterization of neuropathic axonal changes.4 The high electrical activation threshold of mechanoinsensitive nociceptors has been used to specifically investigate their role in neurogenic vasodilation. Electrical stimulation of the skin evoked a widespread axon reflex erythema only when the stimulation was intense enough to stimulate mechano-insensitive nociceptors, whereas even high-frequency stimulation at an intensity known to activate all polymodal nociceptors did not.70 Thus neurogenic inflammation in humans differs from that in rodents, in

which the neurogenic inflammation can be elicited by polymodal C nociceptors19,42 and consists of a combination of vasodilation and protein extravasation.66,96 The latter is absent in healthy volunteers,66 but may develop under pathophysiologic conditions.95 Inflammatory mediators applied to the receptive endings can acutely sensitize mechano-insensitive “sleeping” nociceptors75,76 and render them mechanoresponsive. This sensitization may underlie the mechanical hypersensitivity that develops at the site of an injury (primary mechanical hyperalgesia). This concept of mechanically sensitized “sleeping” nociceptors so far has mainly been considered in the context of inflammatory mediators provoking local acute inflammatory pain and mechanical hyperalgesia. However, the sensitization process of “sleeping” nociceptors may not be restricted to local phenomena at the receptive endings. Nerve injury leading to wallerian degeneration might affect uninjured neighboring axons by inducing an axonal inflammation and provoke massive changes in gene expression in dorsal root ganglion cells of the initially “healthy” cells (upstream effect).7,104 Changed expression of transduction proteins, however, can then in a second step also sensitize the sensory endings (downstream effect). The sensitization of “sleeping” nociceptors has already been observed in pain patients59; however, the relative contributions of inflammatory changes at the nerve terminals and at the axons still have to be established. In addition to their role in primary mechanical hyperalgesia and neurogenic vasodilation, the spinal input of mechano-insensitive C fibers has been postulated to provoke central sensitization. Electrical stimulation at the high current density needed to excite mechano-insensitive nociceptors has been shown to induce large areas of punctuate hyperalgesia and allodynia around the stimulation site.39 The two types of central hypersensitivity can also be provoked experimentally by intradermal injections of algogens such as capsaicin,40 which provokes central sensitization.80 The long-lasting responses of mechano-insensitive nociceptors to capsaicin injection73 further support their role in the induction of central sensitization. Punctuate hyperalgesia and stroke-evoked allodynia are prominent features of neuropathic pain,36 and thus activity in this nociceptor class could be of major importance for neuropathic pain patients. Itch Fibers. About 10% of the mechano-insensitive nociceptors show lasting activation by histamine that parallels the itch sensation in humans.72 Supporting the proposal of “itch-specific” neurons, only the histamine-positive fibers were also excited by intracutaneous injection of the weak pruritogen prostaglandin E2, whereas histamine-negative mechano-insensitive nociceptors and polymodal nociceptors were not activated.74 Recently, “itch-specific” spinal neurons have been identified in the cat,2 which also strengthens the concept of a labeled line for the central processing of the itch sensation.

Traditionally afferent C-fiber function has been restricted to nociception and warmth detection. Therefore, the finding of C fibers that can be activated by slightly stroking the skin and that discharge at high frequencies was surprising,52,91–93 because no obvious sensory correlate for this activity was known. This class of afferent C fibers has been hypothesized to have a role in grooming behavior. Recently, touch-evoked activation of insular cortex, but not primary and secondary somatosensory cortex, combined with a faint sensation of pleasant touch has been described in a patient without myelinated fibers.58 This result strengthens the view that these fibers have a role in emotional or limbic touch, as suggested previously.5,16 Although impaired function of these fiber classes is problematic to quantify in neuropathy patients, it might well considerably affect their social life.

Modulation of Discharge Frequency by the Axon Effect of Hyperpolarization Activity-induced hyperpolarization is generally related to neuronal inhibition. However, if C nociceptors are active and have acquired activity-dependent hyperpolarization, a supernormal period can be observed in an interval from 10 to about 300 ms following an action potential (Fig. 38–2).6,100 Action potentials being incited in this interval will be conducted faster than those incited in the preceding one, and therefore the interspike interval will gradually decrease along the axonal length. Thus instantaneous frequencies will gradually increase along the axon, eventually exceeding the original stimulation frequency in the peripheral innervation territory (Fig. 38–3). Remarkably, at a high degree of hyperpolarization, the instantaneous frequencies may massively increase along the axon such that they reach the maximum frequency a C axon can generate (“entrainment”), which is about 190 Hz. Thereby, a peripheral stimulation at a frequency of 50 Hz can provoke spike trains that increase in instantaneous frequency along the axon and may reach the spinal cord at a frequency that is three to four times higher than the original frequency generated in the periphery.99 Thus axonal hyperpolarization in C nociceptors can paradoxically enhance sensory effects. Unidirectional Block Excitation of one peripheral ending in the axonal tree will generate action potentials that are conducted centrally but also antidromically, and thereby depolarize the entire axonal tree (axon reflex). Should two endings be active simultaneously, only the action potential reaching the common branching point first is expected to reach the central nervous system, whereas the later potential will collide with the retrogradely invading action potential (Fig. 38–4). Thereby, maximum discharge frequency in the

1007

20 15 Mechano-responsive (n=21) Mechano-insensitive (n=16)

10 5

Conduction delay, ms (”speeding”)

Low-Threshold C Fibers

Conduction delay, ms (”slowing”)

Single-Unit Recordings of Afferent Human Peripheral Nerves by Microneurography

0 200 –5

600

1000

2000

Interstimulus interval, ms

–10

FIGURE 38–2 Time course of increased and decreased axonal conduction velocity following an action potential. Twin pulses with various interstimulus intervals (20, 50, 100, 250, 500, 1000, and 2000 ms) were applied in the innervation territory of mechanoresponsive and mechano-insensitive nociceptors. The responses were recorded by microneurography in the peroneal nerve at knee level. The difference in conduction delay between the first and second pulse (“conduction delay”) was assessed. For short intervals (sodium channels partly inactivated), slower conduction velocity of the second action potential was observed in mechanoresponsive units. However, in mechano-insensitive units supernormal conduction was observed during an interval of 20 to 300 ms, which subsequently changed to subnormal conduction according to the long-lasting activity-dependent hyperpolarization. (Data from Weidner, C., Schmidt, R., Schmelz, M., et al.: Time course of post-excitatory effects separates afferent human C fibre classes. J. Physiol. [Lond.] 527:185, 2000.)

parent axon would be limited by the maximum frequencies tolerated by the small branches. However, retrograde axonal conduction at the branching point can be blocked (unidirectional block). In this situation, action potentials generated in the noninvaded branches will not collide and can also reach the central nervous system (see Fig. 38–4). It is interesting to note that unidirectional block in the periphery will enhance discharge frequency in the parent axon, whereas unidirectional blocks of its central endings in the spinal cord will reduce the sensory input.98

DISORDERS OF SENSORY NERVE FUNCTION Paresthesias and Spontaneous Activity in Myelinated Fibers In healthy nerves spontaneous activity can be generated by hyperventilation and ischemia, but it is also observed upon reperfusion and following long impulse trains.50 Microneurographic recordings have verified irregular

1008

Nerve Conduction and Electromyography

Trains (50 Hz) of electrical stimulation inside the receptive field ° 50 Hz ° 50 Hz ° 50 Hz Recorded trains of actions potentials at knee level: 40 Hz 43 Hz

50 traces

24 Hz

182 Hz 171 Hz 142 Hz 540 0

1

580

620

Response latency, ms

Repetition frequency, Hz

FIGURE 38–3 Repetitive trains of four pulses (open circles) were applied to the innervation territory in the foot at 50 Hz, and responses of a C nociceptor (filled circles) were recorded by microneurography in the peroneal nerve at knee level. Responses of the nociceptor are depicted in one trace, and subsequent stimulations are shown from top to bottom (dotted lines). Interspike intervals of the first response were longer as compared to the interstimulus interval, leading to instantaneous frequencies of 24 to 43 Hz. When repetition frequency of the trains was increased stepwise (left), instantaneous frequencies of the evoked response gradually increased, although the interstimulus interval was kept constant at 20 ms (50 Hz). With increasing repetition frequency, activity-dependent hyperpolarization becomes more intense and supernormal conduction becomes more pronounced. Thus gradually higher instantaneous frequencies of the evoked responses are observed, far above the stimulation frequency in the periphery of 50 Hz. Maximum frequencies (entrainment) of more than 180 Hz were recorded in this example. (Data from Weidner, C., Schmelz, M., Schmidt, R., et al.: Neural signal processing: the underestimated contribution of peripheral human C-fibers. J. Neurosci. 22:6704, 2002.)

bursting discharges in A fibers under these conditions.56 As an underlying mechanism for the generation of spontaneous activity observed in hyperventilation, high pH–induced activation of persistent axonal sodium channels has been proposed, which might be of more significance as compared to lowering of free calcium concentration.50 Under ischemic conditions the impaired electrogenic Na⫹, K⫹ pump and accumulation of extracellular potassium facilitate spontaneous activity.50 It is yet unclear as to whether the same mechanisms leading to paresthesias in healthy subjects also play a critical role under pathophysiologic conditions. Spontaneous and abnormal evoked activity in myelinated afferents

corresponding to paresthesia has been recorded by microneurography in neuropathy patients,9 in patients recovering from Guillain-Barré syndrome,46 and also in amputees.53 In some patients the generating site of the spontaneous activity was investigated by proximal and distal local anesthesia. Because anesthesia of the nerve proximal to the recording site stopped ongoing activity in one patient,9 the origin apparently was either along the central path of the axon or the dorsal root ganglion cell. Similarly, local anesthesia of the neuroma in an amputee did not abolish spontaneous activity,53 suggesting a central origin. However, microneurography results will not provide more specific information about the ional mechanism leading to spontaneous activity. In diabetic patients low-threshold mechanoreceptors showed pronounced fatigue (SA I) and lower discharge rates (SA II), which may correspond to impaired tactile function in these patients.44 More specifically, diabetes patients had shorter refractory periods despite unchanged conduction velocity and action potential duration.45 Pronounced fatigue in low-threshold mechanoreceptors was also described in regenerating fibers following nerve transection,43,47 whereas spinal injury did not substantially change sensory characteristics of the primary afferents.82

Neuropathic Changes in Unmyelinated Fibers Microneurographic studies on sympathetic fibers of patients suffering from diabetic neuropathy have shown normal conduction velocities of postganglionic sympathetic fibers, but a decrease in the number of fibers was suggested.17 Also, afferent C fibers recorded in diabetic patients did not show any signs of altered conductive properties, but after discharge in mechanosensitive C nociceptors was observed as a potentially pathologic feature.79 However, in a recent study in erythromelalgia patients, lowered conduction velocity and more pronounced activity-dependent slowing in the afferent C fibers were found59 (Fig. 38–5). Indeed, electron microscopic investigations of sural nerve biopsies in patients with diabetic neuropathy have identified smaller axonal diameter of unmyelinated fibers,41,48 which might explain lower conduction velocities. More pronounced activity-dependent hyperpolarization in these patients might be related to increased expression of ion channels and pumps involved in the afterhyperpolarization (calcium-activated K⫹ channels, HCN channels [Ih], Na⫹, K⫹ pump). In contrast to this assumed increase, decreased expression of calcium-activated K⫹ channels (hSK1, hIK1) has recently been found in patients following peripheral nerve injury.4 Data on specific changes of axonal ion channel expression in patients are currently sparse, but microneurography might contribute to further characterize these changes on a functional level.

1009

Traces

Single-Unit Recordings of Afferent Human Peripheral Nerves by Microneurography

550

560

570

580

590

600

610

B

ms

A

Par e Axo nt n

FIGURE 38–4 Top, Responses of a C nociceptor to electrical pulses (0.25 Hz) in the cutaneous receptive field in the foot dorsum were recorded by microneurography in the peroneal nerve at knee level. The responses are depicted in successive traces from top to bottom. Although only one electrical pulse was applied, a second action potential of the unit under study (marked by open box) was recorded in some traces. The shapes of the two action potentials were virtually identical (see superimposed shapes in the inset). In addition, each double response caused activity-dependent slowing in the unit under study (filled circle). Conversely, absence of a double response was always followed by a recovery of conduction velocity (open circle). (Data from Weidner et al.85,104). Bottom, Schematic view of unidirectional block of conduction. Electrical stimulation (flash symbol) in the receptive field activates two terminal branches of a C fiber. Action potential A reaches the branching point first. It does not retrogradely invade the second branch, but is only conducted centrally (unidirectional block). Thus, action potential B can also reach the branching point without undergoing collision and is also conducted centrally. Thereby, the parent axon will conduct two action potentials following a single electrical stimulus.

Sensitization of C Fibers in Erythromelalgia Assessment of sensitization traditionally involves measurements of thresholds of single fibers before and after a sensitizing event. Because of the long time course in neuropathy, this approach cannot be used in microneuro-

graphy. Additionally, statistical comparisons of thresholds in single fibers are problematic because of the high variability and the very large number of observations required, which is hard to achieve with microneurographic recordings in patients. However, as described earlier, biophysical axonal

1010

Nerve Conduction and Electromyography

Control

Total slowing

70 ms 60 50

EM patients Control

*

40

EM patient

0.125 Hz 0.25 Hz

30

0.5 Hz

20

*

10

0.25 Hz

0

Conduction velocity

1,1 ms 1,0

*

0,9 0,8

* 530

0,7

560

590 500

530

560

Response latency, ms 0,6 0,5 Mechanoinsensitive

Mechanoresponsive

Sympathetic

Unit class FIGURE 38–5 Altered conductive properties in only afferent fibers of erythromelalgia (EM) patients. Conduction velocity (bottom) and amount of activity-dependent slowing (top) differed significantly (ANOVA/ANCOVA, P ⬍ 0.05) between healthy subjects and EM patients (means ⫾ 95% confidence intervals). These differences were restricted to the afferent fiber classes (ANOVA, post-hoc LSD, P ⬍ 0.05 marked with asterisk). Sympathetic fibers were significantly less affected. (Data from Orstavik, K., Weidner, C., Schmidt, R., et al.: Pathological C-fibres in patients with a chronic painful condition. Brain 126:567, 2003.)

properties such as degree of activity-dependent slowing and transcutaneous electrical threshold can be used to identify afferent nerve fiber classes. High transcutaneous electrical threshold and pronounced activity-dependent slowing is found only in mechano-insensitive C fibers in healthy humans (see Fig. 38–1). Thus mechanical responsiveness in fibers characterized by these biophysical properties can be regarded as mechanical sensitization of originally mechano-insensitive nociceptors.59,78 Sensitization of originally mechano-insensitive fibers to mechanical stimulation correlates well with the clinically observed mechanical hypersensitivity in these patients (Fig. 38–6).

FIGURE 38-6 Left, Simultaneous recording of three afferent C fibers in a healthy subject. The first trace depicts the original nerve signal; for all successive traces action potentials are symbolized by rhombi at their appropriate latencies. Two fibers (filled symbols) developed pronounced activity-dependent slowing of conduction velocity upon increasing stimulus frequency, which was much less for the third fiber (open symbols). Stimulation with a 750-mN von Frey’s hair (arrowhead) activated this fiber, as can be seen by the increase in response latency (“marking”). The two fibers with pronounced activity-dependent slowing did not respond to the mechanical stimulus (mechano-insensitive fibers). Right, Single C nociceptor recorded in an erythromelalgia patient. The fiber had a pronounced conduction velocity slowing as seen characteristically in mechano-insensitive fibers. However, the mechanical stimulation induced a clear and reproducible marking response, suggesting mechanical sensitization. (Data from Orstavik, K., Weidner, C., Schmidt, R., et al.: Pathological C-fibres in patients with a chronic painful condition. Brain 126:567, 2003.)

In addition, spontaneous activity in mechano-insensitive nociceptors was observed in erythromelalgia patients by microneurography.59 Spontaneous activity in this nociceptor class could contribute to the spontaneous pain of these patients. As described above, this nociceptor class has been linked to the induction of central sensitization, leading to stroke-evoked allodynia and punctuate hyperalgesia. Thus spontaneous activity in the mechano-insensitive nociceptors may also be linked to the central mechanical hyperalgesia observed in these patients. The frequency of ongoing activity might even be enhanced by a higher incidence of unidirectional blocks55 in erythromelalgia patients.

Single-Unit Recordings of Afferent Human Peripheral Nerves by Microneurography

FUTURE PERSPECTIVES IN MICRONEUROGRAPHY Although microneurography has provided unique information about neural encoding in humans, functional electrophysiologic data provided by microneurography were basically isolated from developments in cellular biology on a molecular level. Because microneurography does not allow us to record directly from the sensory endings or record the membrane potential, molecular mechanisms of transduction cannot be investigated directly. However, axonal mechanisms of membrane potential modulation, such as postexcitatory hyperpolarization, are also currently being identified on a molecular level.10 The time course of postexcitatory hyperpolarization can be assessed by microneurography, and thus there is the exciting prospect of finally bridging the gap between the cellular and systemic approaches. Moreover, the unexpected segregation of axonal and sensory properties in C fibers might help to also link the electrophysiologic data to analysis using immunohistochemistry, as has been successfully shown in cold-sensitive units.35 Despite many limitations, the microneurographic approach provides unique data from healthy human subjects and patients and has an intermediary position between clinical science and basic research.

REFERENCES 1. Adriaensen, H., Gybels, J., Handwerker, H. O., and Van, H. J.: Response properties of thin myelinated (A-delta) fibers in human skin nerves. J. Neurophysiol. 49:111, 1983. 2. Andrew, D., and Craig, A. D.: Spinothalamic lamina 1 neurons selectively sensitive to histamine: a central neural pathway for itch. Nat. Neurosci. 4:72, 2001. 3. Birznieks, I., Jenmalm, P., Goodwin, A. W., and Johansson, R. S.: Encoding of direction of fingertip forces by human tactile afferents. J. Neurosci. 21:8222, 2001. 4. Boettger, M. K., Till, S., Chen, M. X., et al.: Calcium-activated potassium channel SK1- and IK1-like immunoreactivity in injured human sensory neurones and its regulation by neurotrophic factors. Brain 125:252, 2002. 5. Bolanowski, S. J., Verrillo, R. T., and McGlone, F.: Passive, active and intra-active (self) touch. Somatosens. Mot. Res. 16:304, 1999. 6. Bostock, H., Campero, M., Serra, J., and Ochoa, J.: Velocity recovery cycles of C fibres innervating human skin. J. Physiol. (Lond.) 533(Pt. 2):649, 2003. 7. Campbell, J. N.: Nerve lesions and the generation of pain. Muscle Nerve 24:1261, 2001. 8. Campero, M., Serra, J., Bostock, H., and Ochoa, J. L.: Slowly conducting afferents activated by innocuous low temperature in human skin. J. Physiol. (Lond.) 535:855, 2001. 9. Campero, M., Serra, J., Marchettini, P., and Ochoa, J. L.: Ectopic impulse generation and autoexcitation in single myelinated afferent fibers in patients with peripheral

10.

11.

12.

13.

14.

15. 16.

17.

18.

19.

20.

21.

22.

23.

24. 25.

26.

27.

1011

neuropathy and positive sensory symptoms. Muscle Nerve 21:1661, 1998. Cordoba-Rodriguez, R., Moore, K. A., Kao, J. P., and Weinreich, D.: Calcium regulation of a slow post-spike hyperpolarization in vagal afferent neurons. Proc. Natl. Acad. Sci. U. S. A. 96:7650, 1999. Craig, A. D., and Bushnell, M. C.: The thermal grill illusion: unmasking the burn of cold pain. Science 265:252, 1994. Craig, A. D., Reiman, E. M., Evans, A., and Bushnell, M. C.: Functional imaging of an illusion of pain [see comments]. Nature 384:258, 1996. Edin, B.: Cutaneous afferents provide information about knee joint movements in humans. J. Physiol. (Lond.) 531:289, 2001. Edin, B. B., Essick, G. K., Trulsson, M., and Olsson, K. A.: Receptor encoding of moving tactile stimuli in humans. I. Temporal pattern of discharge of individual low-threshold mechanoreceptors. J. Neurosci. 15:830, 1995. Erickson, R. P., and Poulos, D. A.: On the qualitative aspect of the temperature sense. Brain Res. 61:107, 1973. Essick, G. K., James, A., and McGlone, F. P.: Psychophysical assessment of the affective components of non-painful touch. Neuroreport 10:2083, 1999. Fagius, J., and Wallin, B. G.: Sympathetic reflex latencies and conduction velocities in patients with polyneuropathy. J. Neurol. Sci. 47:449, 1980. Gandevia, S. C., and Hales, J. P.: The methodology and scope of human microneurography. J. Neurosci. Methods 74:123, 1997. Gee, M. D., Lynn, B., and Cotsell, B.: The relationship between cutaneous C fibre type and antidromic vasodilatation in the rabbit and the rat. J. Physiol. (Lond.) 503:31, 1997. Goodwin, A. W., Macefield, V. G., and Bisley, J. W.: Encoding of object curvature by tactile afferents from human fingers. J. Neurophysiol. 78:2881, 1997. Hagbarth, K. E., Torebjörk, H. E., and Wallin, B. G.: Microelectrode explorations of human peripheral nerves. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 658, 1993. Hallin, R. G.: Microneurography in relation to intraneural topography: somatotopic organisation of median nerve fascicles in humans. J. Neurol. Neurosurg. Psychiatry 53:736, 1990. Hallin, R. G., Ekedahl, R., and Frank, O.: Segregation by modality of myelinated and unmyelinated fibers in human sensory nerve fascicles. Muscle Nerve 14:157, 1991. Hallin, R. G., and Wu, G.: Fitting pieces in the peripheral nerve puzzle. Exp. Neurol. 172:482, 2001. Hallin, R. G., and Wu, G.: Novel information on peripheral tactile mechanisms in man acquired with concentric needle electrode microneurography. Behav. Brain Res. 135:11, 2002. Hensel, H: Cutaneous thermoreceptors. In Iggo, A. (ed.): Handbook of Sensory Physiology. Berlin, Springer Verlag, p. 79, 1973. Hensel, H.: Correlations of neural activity and thermal sensation in man. In Zotterman, Y. (ed.): Sensory Functions of the Skin in Primates. Oxford, UK, Pergamon Press, p. 331, 1976.

1012

Nerve Conduction and Electromyography

28. Iggo, A., and Muir, A. R.: The structure and function of a slowly adapting touch corpuscle in hairy skin. J. Physiol. (Lond.) 200:763, 1969. 29. Järvilehto, T., and Hämälainen, H.: Touch and thermal sensations: psychophysical observations and unit activity in human skin nerves. In Kenshalo, D. R. (ed.): Sensory Functions of the Skin in Humans. New York, Plenum Publishing, p. 279, 1979. 30. Jenmalm, P., and Johansson, R. S.: Visual and somatosensory information about object shape control manipulative fingertip forces. J. Neurosci. 17:4486, 1997. 31. Johansson, R. S., and Cole, K. J.: Sensory-motor coordination during grasping and manipulative actions. Curr. Opin. Neurobiol. 2:815, 1992. 32. Johansson, R. S., Trulsson, M., Olsson, K. A., and Abbs, J. H.: Mechanoreceptive afferent activity in the infraorbital nerve in man during speech and chewing movements. Exp. Brain Res. 72:209, 1988. 33. Johansson, R. S., Trulsson, M., Olsson, K. A., and Westberg, K. G.: Mechanoreceptor activity from the human face and oral mucosa. Exp. Brain Res. 72:204, 1988. 34. Jorum, E., Lundberg, L. E., and Torebjörk, H. E.: Peripheral projections of nociceptive unmyelinated axons in the human peroneal nerve. J. Physiol. (Lond.) 416:291, 1989. 35. Koltzenburg, M., and Koerber, R. H.: Diversity of nociceptors—which are important for the generation of clinically relevant persistent pains? [abstract 710]. In Abstracts of the 10th World Congress on Pain. Seattle, IASP Press, 2002. 36. Koltzenburg, M., Torebjörk, H. E., and Wahren, L. K.: Nociceptor modulated central sensitization causes mechanical hyperalgesia in acute chemogenic and chronic neuropathic pain. Brain 117:579, 1994. 37. Konietzny, F.: Peripheral neural correlates of temperature sensations in man. Hum. Neurobiol. 3:21, 1984. 38. Konietzny, F., and Hensel, H.: Letters and notes: warm fiber activity in human skin nerves. Pflugers Arch. 359:265, 1975. 39. Koppert, W., Dern, S. K., Sittl, R., et al.: A new model of electrically evoked pain and hyperalgesia in human skin: the effects of intravenous alfentanil, S(⫹)-ketamine, and lidocaine. Anesthesiology 95:395, 2001. 40. LaMotte, R. H., Lundberg, L. E. R., and Torebjörk, H. E.: Pain, hyperalgesia and activity in nociceptive-C units in humans after intradermal injection of capsaicin. J. Physiol. (Lond.) 448:749, 1992. 41. Llewelyn, J. G., Gilbey, S. G., Thomas, P. K., et al.: Sural nerve morphometry in diabetic autonomic and painful sensory neuropathy: a clinicopathological study. Brain 114:867, 1991. 42. Lynn, B., Schutterle, S., and Pierau, F. K.: The vasodilator component of neurogenic inflammation is caused by a special subclass of heat-sensitive nociceptors in the skin of the pig. J. Physiol. (Lond.) 494:587, 1996. 43. Mackel, R.: Human cutaneous mechanoreceptors during regeneration: physiology and interpretation. Ann. Neurol. 18:165, 1985. 44. Mackel, R.: Properties of cutaneous afferents in diabetic neuropathy. Brain 112(Pt. 5):1359, 1989.

45. Mackel, R., and Brink, E.: Conduction of neural impulses in diabetic neuropathy. Clin. Neurophysiol. 114:248, 2003. 46. Mackel, R., Brink, E., Jorum, E., and Aisen, M.: Properties of cutaneous afferents during recovery from Guillain-Barre syndrome. Brain 117:169, 1994. 47. Mackel, R., Brink, E. E., and Wittkowsky, G.: Properties of cutaneous mechanosensitive afferents during the early stages of regeneration in man. Brain Res. 329:49, 1985. 48. Malik, R. A., Veves, A., Walker, D., et al.: Sural nerve fibre pathology in diabetic patients with mild neuropathy: relationship to pain, quantitative sensory testing and peripheral nerve electrophysiology. Acta Neuropathol. (Berl.) 101:367, 2001. 49. McGlone, F., Kelly, E., Trulsson, M., et al.: Functional neuroimaging studies of human somatosensory cortex. Behav. Brain Res. 135:147, 2002. 50. Mogyoros, I., Bostock, H., and Burke, D.: Mechanisms of paresthesias arising from healthy axons. Muscle Nerve 23:310, 2000. 51. Moore, C. E., and Schady, W.: Investigation of the functional correlates of reorganization within the human somatosensory cortex. Brain 123:1883, 2000. 52. Nordin, M.: Low-threshold mechanoreceptive and nociceptive units with unmyelinated (C) fibres in the human supraorbital nerve [published erratum appears in J. Physiol. (Lond.) 444:777, 1991]. J. Physiol. (Lond.) 426:229, 1990. 53. Nystrom, B., and Hagbarth, K. E.: Microelectrode recordings from transected nerves in amputees with phantom limb pain. Neurosci. Lett. 27:211, 1981. 54. Ochoa, J., and Torebjörk, H. E.: Sensations evoked by intraneural microstimulation of single mechanoreceptor units innervating the human hand. J. Physiol. (Lond.) 342:633, 1983. 55. Ochoa, J. L.: Truths, errors, and lies around “reflex sympathetic dystrophy” and “complex regional pain syndrome”. J. Neurol. 246:875, 1999. 56. Ochoa, J. L., and Torebjörk, H. E.: Paraesthesiae from ectopic impulse generation in human sensory nerves. Brain 103:835, 1980. 57. Ochoa, J. L., and Torebjörk, H. E.: Sensations evoked by intraneural microstimulation of C nociceptor fibres in human skin nerves. J. Physiol. (Lond.) 415:583, 1989. 58. Olausson, H., Lamarre, Y., Backlund, H., et al.: Unmyelinated tactile afferents signal touch and project to insular cortex. Nat. Neurosci. 5:900, 2002. 59. Orstavik, K., Weidner, C., Schmidt, R., et al.: Pathological C-fibres in patients with a chronic painful condition. Brain 126:567, 2003. 60. Peng, Y. B., Ringkamp, M., Campbell, J. N., and Meyer, R. A.: Electrophysiological assessment of the cutaneous arborization of Adelta-fiber nociceptors. J. Neurophysiol. 82:1164, 1999. 61. Perl, E. R.: Cutaneous polymodal receptors: characteristics and plasticity. Prog. Brain Res. 113:21, 1996. 62. Ringkamp, M., Peng, Y. B., Campbell, J. N., and Meyer, R. A.: Intradermal capsaicin produces a vigorous discharge in mechanically-insensitive A-fiber nociceptors of the monkey. Soc. Neurosci. Abstr. 23(Pt. 2):1258, 1997. 63. Robinson, C. J., Torebjörk, H. E., and LaMotte, R. H.: Psychophysical detection and pain ratings of incremental

Single-Unit Recordings of Afferent Human Peripheral Nerves by Microneurography

64.

65.

66.

67.

68.

69.

70.

71. 72.

73.

74.

75.

76.

77.

78.

79.

80.

thermal stimuli: a comparison with nociceptor responses in humans. Brain Res. 274:87, 1983. Rowe, M. J.: Synaptic transmission between single tactile and kinaesthetic sensory nerve fibers and their central target neurones. Behav. Brain Res. 135:197, 2002. Sah, P.: Ca(2⫹)-activated K⫹ currents in neurones: types, physiological roles and modulation. Trends Neurosci. 19:150, 1996. Sauerstein, K., Klede, M., Hilliges, M., and Schmelz, M.: Electrically evoked neuropeptide release and neurogenic inflammation differ between rat and human skin. J. Physiol. (Lond.) 529:803, 2000. Schady, W., Ochoa, J. L., Torebjörk, H. E., and Chen, L. S.: Peripheral projections of fascicles in the human median nerve. Brain 106:745, 1983. Schady, W. J., and Torebjörk, H. E.: Projected and receptive fields: a comparison of projected areas of sensations evoked by intraneural stimulation of mechanoreceptive units, and their innervation territories. Acta Physiol. Scand. 119:267, 1983. Schady, W. J., Torebjörk, H. E., and Ochoa, J. L.: Cerebral localisation function from the input of single mechanoreceptive units in man. Acta Physiol. Scand. 119:277, 1983. Schmelz, M., Michael, K., Weidner, C., et al.: Which nerve fibers mediate the axon reflex flare in human skin? Neuroreport 11:645, 2000. Schmelz, M., and Petersen, L. J.: Neurogenic inflammation in human and rodent skin. News Physiol. Sci. 16:33, 2001. Schmelz, M., Schmidt, R., Bickel, A., et al.: Specific C-receptors for itch in human skin. J. Neurosci. 17:8003, 1997. Schmelz, M., Schmidt, R., Handwerker, H. O., and Torebjörk, H. E.: Encoding of burning pain from capsaicin-treated human skin in two categories of unmyelinated nerve fibres. Brain 123:560, 2000. Schmelz, M., Schmidt, R., Weidner, C., et al.: Chemical response pattern of different classes of C-nociceptors to pruritogens and algogens. J. Neurophysiol. 89:2441, 2003. Schmidt, R., Schmelz, M., Forster, C., et al.: Novel classes of responsive and unresponsive C nociceptors in human skin. J. Neurosci. 15:333, 1995. Schmidt, R., Schmelz, M., Torebjörk, H. E., and Handwerker, H. O.: Mechano-insensitive nociceptors encode pain evoked by tonic pressure to human skin. Neuroscience 98:793, 2000. Schmidt, R., Schmelz, M., Weidner, C., et al.: Innervation territories of mechano-insensitive C nociceptors in human skin. J. Neurophysiol. 88:1859, 2002. Serra, J., Bostock, H., and Ochoa, J.: Responses of afferent C units in human skin to remote intradermal capsaicin injection [abstract 458-P92]. In Abstracts of the 10th World Congress on Pain. Seattle, IASP Press, 2002. Serra, J., Campero, M., Ochoa, J., and Bostock, H.: Activity-dependent slowing of conduction differentiates functional subtypes of C fibres innervating human skin. J. Physiol. (Lond.) 515:799, 1999. Simone, D. A., Baumann, T. K., Collins, J. G., and LaMotte, R. H.: Sensitization of cat dorsal horn neurons to innocuous

81.

82.

83.

84.

85.

86. 87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

1013

mechanical stimulation after intradermal injection of capsaicin. Brain Res. 486:185, 1989. Susser, E., Sprecher, E., and Yarnitsky, D.: Paradoxical heat sensation in healthy subjects: peripherally conducted by A delta or C fibres? Brain 122:239, 1999. Thomas, C. K., and Westling, G.: Tactile unit properties after human cervical spinal cord injury. Brain 118(Pt. 6):1547, 1995. Thunberg, T.: Förnimmelserne vid till samma ställe lokaliserad samtidigt pågående köld och värmeretning. Ups. Läkfören. Förh. 1:489, 1896. Torebjörk, H. E., and Hallin, R. G.: Skin receptors supplied by unmyelinated (C) fibres in man. In Zotterman, Y. (ed.): Sensory Functions of the Skin in Primates. Oxford, UK, Pergamon Press, p. 475, 1976. Torebjörk, H. E., Vallbo, A. B., and Ochoa, J. L.: Intraneural microstimulation in man: its relation to specificity of tactile sensations. Brain 110:1509, 1987. Trulsson, M.: Mechanoreceptive afferents in the human sural nerve. Exp. Brain Res. 137:111, 2001. Trulsson, M., Francis, S. T., Kelly, E. F., et al.: Cortical responses to single mechanoreceptive afferent microstimulation revealed with fMRI. Neuroimage 13:613, 2001. Trulsson, M., and Johansson, R. S.: Encoding of tooth loads by human periodontal afferents and their role in jaw motor control. Prog. Neurobiol. 49:267, 1996. Trulsson, M., and Johansson, R. S.: Orofacial mechanoreceptors in humans: encoding characteristics and responses during natural orofacial behaviors. Behav. Brain Res. 135:27, 2002. Vallbo, A. B., Hagbarth, K. E., Torebjörk, H. E., and Wallin, B. G.: Somatosensory, proprioceptive, and sympathetic activity in human peripheral nerves. Physiol. Rev. 59:919, 1979. Vallbo, A. B., Olausson, H., and Wessberg, J.: Unmyelinated afferents constitute a second system coding tactile stimuli of the human hairy skin. J. Neurophysiol. 81:2753, 1999. Vallbo, A. B., Olausson, H., Wessberg, J., and Kakuda, N.: Receptive field characteristics of tactile units with myelinated afferents in hairy skin of human subjects. J. Physiol. (Lond.) 483:783, 1995. Vallbo, A. B., Olausson, H., Wessberg, J., and Norrsell, U.: A system of unmyelinated afferents for innocuous mechanoreception in the human skin. Brain Res. 628:301, 1993. Wall, P. D., and McMahon, S. B.: Microneuronography and its relation to perceived sensation: a critical review. Pain 21:209, 1985. Weber, M., Birklein, F., Neundorfer, B., and Schmelz, M.: Facilitated neurogenic inflammation in complex regional pain syndrome. Pain 91:251, 2001. Weidner, C., Klede, M., Rukwied, R., et al.: Acute effects of substance P and calcitonin gene-related peptide in human skin—a microdialysis study. J. Invest. Dermatol. 115:1015, 2000. Weidner, C., Schmelz, M., Schmidt, R., et al.: Functional attributes discriminating mechano-insensitive and mechanoresponsive C nociceptors in human skin. J. Neurosci. 19:10184, 1999. Weidner, C., Schmelz, M., Schmidt, R., et al.: Unidirectional conduction block at branching points of human nociceptive

1014

Nerve Conduction and Electromyography

C afferents: a peripheral mechanism for pain amplification. In Devor, M., Rowbotham, M., and Wiesenfeld-Hallin, Z. (eds.): Proceedings of the 9th World Congress on Pain (Progress in Pain Research and Management, Vol. 16). Seattle, IASP Press, p. 233, 2000. 99. Weidner, C., Schmelz, M., Schmidt, R., et al.: Neural signal processing: the underestimated contribution of peripheral human C-fibers. J. Neurosci. 22:6704, 2002. 100. Weidner, C., Schmidt, R., Schmelz, M., et al.: Time course of post-excitatory effects separates afferent human C fibre classes. J. Physiol. (Lond.) 527:185, 2000. 101. Weidner, C., Schmidt, R., Schmelz, M., et al.: Action potential conduction in the terminal arborisation of nociceptive C-fibre afferents. J. Physiol. (Lond.) 547(Pt. 3):931, 2003.

102. Wu, G., Ekedahl, R., and Hallin, R. G.: Clustering of slowly adapting type II mechanoreceptors in human peripheral nerve and skin. Brain 121(Pt. 2):265, 1998. 103. Wu, G., Ekedahl, R., Stark, B., et al.: Clustering of Pacinian corpuscle afferent fibres in the human median nerve. Exp. Brain Res. 126:399, 1999. 104. Wu, G., Ringkamp, M., Hartke, T. V., et al.: Early onset of spontaneous activity in uninjured C-fiber nociceptors after injury to neighboring nerve fibers. J. Neurosci. 21:RC140–1, 2001. 105. Yarnitsky, D., and Ochoa, J. L.: Warm and cold specific somatosensory systems: psychophysical thresholds, reaction times and peripheral conduction velocities. Brain 114:1819, 1991.

39 Compound Action Potentials of Sural Nerve in Vitro in Peripheral Neuropathy EDWARD H. LAMBERT AND PETER J. DYCK

Reconstruction of the Compound Action Potential Fiber Groups and Sensation Compound Action Potentials of Sural Nerve in Peripheral Neuropathy

Recording the Compound Action Potential Method of Study Compound Action Potential and Fiber Diameter Spectrum

Electrophysiologic evidence for the involvement of different fiber groups in neuropathies can be obtained by recording the compound action potential of the nerve. In this chapter, compound action potentials of fascicles of sural nerve obtained by biopsy in a variety of neuropathies are presented together with clinical and histologic data. When a maximal electrical stimulus is applied to a mixed nerve, the resulting compound action potential recorded several centimeters away is dispersed in time and has two or more peaks. The compound action potential is the classic demonstration that the nerve is composed of groups of fibers with different conduction velocities. The action potentials of the most rapidly conducting fibers appear first in the record and those of more slowly conducting fibers appear later. In their initial studies of frog sciatic nerve, Erlanger and Gasser used Greek letters (␣, ␤, ␥, and ␦) to identify successive peaks of the action potential in order of decreasing conduction velocity.12,13 Subsequently, two additional peaks produced by fibers of much lower conduction velocity were discovered and were called B (not found in mammalian peripheral nerves) and C. The early peaks designated by Greek letters and referred to collectively as the A potential are produced by myelinated fibers ranging from large to small diameters. In mammalian nerves the designation B is reserved for an elevation produced by preganglionic (small-diameter myelinated) fibers in autonomic nerves, although for a time it was also applied to the

Illustrative Compound Action Potentials and Fiber Diameter Histograms Comments

elevation now called A␦.2,4 The C potential is produced by unmyelinated fibers. In cutaneous nerves, these include both dorsal root and sympathetic C fibers. Erlanger and Gasser demonstrated that conduction velocity, excitability, amplitude of action potential, and certain other properties of a nerve fiber are proportional to the fiber diameter.13 Among fibers that produce the A potential, conduction velocity and external fiber diameter are related almost linearly. Hursh21,22 compared maximal conduction velocity and diameter of the largest fibers of various nerves of the cat and found that the slope of the relationship was 6 m/s/␮m diameter. It has been customary to use this factor in estimating one variable from the other. Thus a fiber of 10 ␮m diameter would have a conduction velocity of approximately 60 m/s. Similar conversion factors for alpha fibers have been reported by others, but Boyd and Davey6 found a smaller conversion factor for small- than for large-diameter myelinated fibers in the same nerve (4.5 for gamma fibers and 5.7 for alpha fibers in motor nerve of cat). In unmyelinated fibers also, the velocity of conduction varies directly with diameter of the fiber, but the conversion factor of 1.7 determined by Gasser14,15 for unmyelinated fibers is much smaller than that for myelinated fibers. Compound action potentials of nerves with different functions have characteristic configurations depending on the fiber composition of the nerve.4,6,13 The compound action potentials and fiber diameter spectra of cutaneous 1015

1016

Nerve Conduction and Electromyography

nerves such as the sural nerve are relatively simple and remarkably constant.5,10 The myelinated fibers produce an A potential with two peaks, commonly designated alpha and delta, respectively. The unmyelinated fibers produce a later peak called C. There is still some confusion of terminology in the designation of peaks in the A potential of cutaneous nerves. The initial peak, called A␣ by Gasser16 because it was the first peak in the compound action potential, was called A␤ by Erlanger12 because its conduction velocity was like that of the A␤ peak in muscle nerves. An intermediate peak present in early records, called A␥, was shown by Gasser to be an artifact that was absent when optimal monophasic recording was achieved. Gasser noted a bulge on the falling phase of the alpha peak in some nerves, but saw no reason to give it a special designation. In this chapter the terms alpha and delta are used for the two peaks of the A potential. The letter designations are used to refer not only to elevations of the compound action potential but also to the fibers themselves. In the fiber diameter spectrum, the A␣ fibers are myelinated fibers with an outside diameter generally over 6 ␮m, the A␦ fibers are myelinated fibers with diameters less than 6 ␮m, and the C fibers are unmyelinated fibers that have diameters of 0.5 to 2 ␮m. To avoid the confusion in letter designations of elevations in the compound action potential, Lloyd26 introduced a classification based on fiber diameter in which major afferent fiber groups are designated by Roman numerals: I (12–21 ␮m), II (6–12 ␮m), III (1–6 ␮m), and IV (C fibers). In this classification the A␣ fibers of cutaneous nerve are in group II, the A␦ fibers in group III, and the C fibers in group IV.

RECONSTRUCTION OF THE COMPOUND ACTION POTENTIAL To test the hypothesis that the conduction velocity of a nerve fiber is proportional to its diameter, Erlanger and Gasser attempted to predict the contour of the compound action potential of myelinated fibers from the spectrum of fiber diameters in the nerve.12,17 Their assumptions were that (1) conduction velocity of a fiber is directly proportional to its outside diameter, (2) the amplitude of the action potential of a nerve fiber is proportional to its cross-sectional area, and (3) the action potential of all fibers has the same form and duration and can be approximated by a triangle in which the rising phase is one third of the total duration.13 Reconstruction of a compound action potential was carried out by placing a triangle representing an action potential in its appropriate position along the abscissa for each myelinated fiber in the nerve. The position and amplitude of the triangle were determined by the fiber diameter.

Summation of the triangles produced a contour that was a reasonable reproduction of the recorded compound action potential. More faithful reconstruction has been a challenge for many investigators; the classic one of the action potential of myelinated fibers of the cat saphenous nerve by Gasser and Grundfest18 was accomplished with a highly complicated technique. Landau and co-workers have described a simpler arithmetic method that employs corresponding fiber diameter histograms and is applicable to myelinated central nerve tracts as well as peripheral nerves.25 Gasser was able to reconstruct the compound action potential of unmyelinated fibers from fiber diameter measurements by using a set of assumptions about the relation of conduction velocity and spike size to fiber diameter that was generally similar to that for myelinated fibers.14,15

FIBER GROUPS AND SENSATION In the early 1930s, Heinbecker and co-workers recorded the compound action potential of human nerves within 15 minutes to 1 hour after obtaining them from amputated limbs or from persons who died violently.19,20 In cutaneous nerves, they identified three potential elevations that they designated A or A␤-␥ (A␣), B (A␦), and C. (The current designation of these elevations is given in parentheses.) The strengths of electrical stimuli required to excite the different fiber groups in excised nerve were compared with strengths of stimuli that produced sensations of different quality in vivo in unanesthetized humans. In some instances observations were made on the same nerves before and after excision. Electrical stimulation of nerves in vivo caused only two sensations, touch and pain. Correlating thresholds for sensation and reflex effects with thresholds for the various components of the compound action potential and with histologic measurements of fiber diameter, these investigators concluded that stimuli that excited only the largest myelinated fibers (A␣) produced tactile sensation, and stronger stimuli, which also excited A␦ fibers, caused a pricking touch sensation. Repetitive stimulation of the latter fibers (three to seven shocks) caused pain. They failed to find a specific sensory effect from stimulation of the C fibers, but reactions to such strong stimuli were described as so violent that threshold determinations were not reliable. Similar results were obtained by Collins and associates, who stimulated the exposed sural nerve in nine conscious patients just prior to anterolateral thoracic cordotomy.8 The compound action potential recorded from the cut, crushed distal end of the nerve in situ had three major elevations designated A␤-␥ (A␣), A␦, and C in order of increasing threshold and decreasing conduction velocity. Stimuli sufficient to excite only the A␣ fibers caused a

Compound Action Potentials of Sural Nerve in Vitro

tapping or thumping sensation. Neither single nor repetitive stimuli of this strength were painful. Increasing the strength of stimulus two to three times to excite the A␦ fibers added a stinging or burning quality. Repetitive stimuli of this strength were always very painful. Further increase in the stimulus to 50 times the threshold for A␣ fibers excited C fibers. Even single stimuli that excited the C group caused unbearable pain. It was noted, however, that a stimulus of this intensity, in addition to exciting C fibers, caused repetitive firing of fibers of the A group. A more definitive analysis of the sensory modalities mediated by fibers of different groups has been obtained by studies in animals. Single nerve fibers excited by adequate stimuli have been classified on the basis of their conduction velocity measured in situ. In such studies, fibers from mechanoreceptors are found in all fiber groups, A␣ and ␦ and C.9,23,24,27,33 High-sensitivity mechanoreceptors that excite C fibers occur in hairy skin, but may be infrequent or absent in glabrous skin.1 Fibers activated by noxious stimuli, warmth, and cold are in both the A␦ and C fiber groups, but not in the A␣ group. Observations in humans of the sensory abnormalities associated with disorders that cause more or less selective loss of particular fiber groups support the view that touch-pressure, two-point recognition, joint position, and vibration sensations are mediated predominantly by fibers of the A␣ group, whereas sensations of pain, warmth, and cold are mediated by A␦ and C groups.11 There is as yet no evidence in humans that small myelinated and unmyelinated fibers transmit impulses from high-sensitivity mechanoreceptors.

COMPOUND ACTION POTENTIALS OF SURAL NERVE IN PERIPHERAL NEUROPATHY Recording the Compound Action Potential A method for obtaining an ideal monophasic record of the compound action potential has not been found.16 Several sources of distortion have been discussed in detail by various authors.3,16,18,31 Some of these are as follows: Shock strengths that are at threshold for delta fibers exceed by several times the threshold for alpha fibers. When the strength of stimulus is sufficient to excite the complete spectrum of A fibers, alpha fibers are excited farther from the stimulating cathode than delta fibers, and the alpha elevation appears too early relative to the delta elevation in the compound action potential. Strong shocks required to excite delta fibers may, particularly in fresh nerves, cause repetitive firing of alpha fibers and the appearance of an extra elevation following

1017

the alpha potential. Similarly, strong stimuli required to excite C fibers may cause prolonged repetitive firing of A fibers whose action potentials may appear in the interval between the delta and C elevations. Inadequate reduction of the diphasic artifact or lead separation effect16 may accentuate elevations or cause the appearance of artifactual elevations in the compound action potential. Repetitive stimulation of C fibers at frequencies as low as 5/s causes an increase in the negative afterpotential of the C fibers and an increase in amplitude and duration of the C potential. Conduction velocity is decreased.7,14,28 The effect is reversible, but unless stimulation is infrequent, the C potential may be artificially enhanced. A drop of fluid or fragment of tissue on the nerve between recording electrodes can cause distortion of the action potential.3,30 Changes in interelectrode resistance caused by drying of the nerve or by excess fluid on the nerve produce proportional changes in amplitude of the action potential. Drying does not change the conduction velocity until it is sufficient to damage the nerve fibers, causing a decrease in both amplitude and conduction velocity of the potential.29

Method of Study In most instances a fascicular biopsy of the sural nerve was obtained for study. The nerve was exposed from a point 1 to 2 cm above the lateral malleolus proximally for 10 to 15 cm. A bundle of two to five fascicles was dissected from the nerve; the distal 4 to 5 cm were used for histologic studies, and the proximal 4 to 6 cm were used for action potential recording. Immediately after excision, the nerve was placed in cool Tyrode solution, through which oxygen with 5% carbon dioxide had been bubbled, and was taken from the operating room to the laboratory. Blood vessels, fat, and epineurium were cut away from the fascicles under a dissecting microscope. Cut or branching fascicles were removed. The nerve was placed on a series of silver electrodes in a chamber with an atmosphere of 5% carbon dioxide in oxygen saturated with water vapor. The nerve was held under slight tension with a 0.5 to 0.9-g weight hanging from each end. The chamber was sealed and immersed in a water bath at 37° C. The nerve was stimulated at its distal end while the compound action potential was displayed on an oscilloscope and photographed. Monophasic records were made at several positions along the length of the nerve for determination of the conduction velocity of the various fiber groups, using maximal stimuli for each component of the action potential. Monophasic recordings were obtained by crushing

1018

Nerve Conduction and Electromyography

the nerve between recording electrodes and applying 0.1% procaine at the distal electrode. Optimal records for illustrating the contour of the compound action potential were obtained when the distance between recording electrodes was 5 to 10 mm.16 For the illustrations that follow, the conduction distance (stimulating cathode to first recording electrode) was 20 to 22 mm, except for Figure 39–3, for which the distance was 30 mm.

Compound Action Potential and Fiber Diameter Spectrum In normal nerves, the contour of the compound action potential is determined by the relative numbers of excitable and conducting nerve fibers with various conduction velocities. Because conduction velocity and spike size are proportional to diameter of the fiber, by using appropriate proportionality factors one can make a reasonable prediction of the fiber diameter spectrum in normal nerve from the contour of the compound action potential. In this prediction, the amplitude of a component of the compound action potential (or the area of that component if its duration is not constant) is proportional to the density rather than to the total number of fibers that contribute to it. A much simplified model utilizing Ohm’s law illustrates the basis of this relationship. The current contributed by each fiber is i ⫽ e/(R ⫹ r), in which e is the electromotive force generated by the fiber membrane, r is the internal resistance of the axis cylinder, and R is the external resistance, that is, the resistance of the tissue and fluid surrounding the fiber. Since r is usually very large relative to R, because of the small diameter of the fiber, i is not greatly affected by small changes in external resistance.18 The external resistance, however, is a particularly important factor affecting the amplitude of the action potential recorded by electrodes on the surface of the nerve. The amplitude (E) of the action potential recorded by surface electrodes is determined by the product of the action current flowing along the nerve between the two recording electrodes and the external (or interelectrode) resistance (E ⫽ iR). Assuming that each nerve fiber of a particular diameter contributes an equal amount (i) to the action current, the total current contributed by that fiber group would be proportional to the total number of such fibers in the nerve (I ⫽ ni). A large nerve with a greater number of nerve fibers would have a greater action current than a small nerve. The larger nerve with the greater number of nerve fibers may not have a larger action potential, however, because as the diameter of the nerve increases, the external electrical resistance of the nerve decreases, and the product, E ⫽ IR, may remain the same. Thus a whole nerve composed of many fascicles produces no larger an action potential than a single one of the fascicles

dissected from the nerve, provided that the composition of the fascicles is uniform. Conversely, a nerve with many fibers per square millimeter of cross-sectional area produces an action potential of higher amplitude than does a nerve with few fibers per square millimeter (ni or I is greater, but R is not proportionately lower). From these considerations, it is apparent that the amplitude of a component of the compound action potential is roughly proportional to the density in the nerve of fibers that produce that component, rather than to the total number of those fibers in the nerve. In neuropathy, changes in the contour of the compound action potential may result from changes in conduction velocity of the fibers either by demyelination or attenuation of fiber diameter, or by a disproportionate change in density of active nerve fibers in the different fiber groups by degeneration or inexcitability or block of conduction of fibers, or by both. A decrease in amplitude of all components of the compound action potential might occur from a proportional decrease in density of active fibers in all groups by degeneration, conduction block, or loss of excitability or by increase in relative amount of interstitial tissue.

Illustrative Compound Action Potentials and Fiber Diameter Histograms Normal Nerve The compound action potential has prominent A␣, A␦, and C components (Fig. 39–1A). The conduction velocity of the beginning of the alpha component was 61 m/s at 37° C. The peak of the delta component was conducted at 19 m/s, and the beginning and peak of the C potential were conducted at 1.6 and 1.1 m/s, respectively. The histograms of fiber diameters in fascicles of the same nerve are plotted in two ways: the conventional way, with increasing diameter from left to right on the abscissa as in Figure 39–1B, and the opposite way, with decreasing diameter from left to right as in Figure 39–1C. In the conventional plot, the consecutive peaks of the histogram for myelinated fibers appear to correspond closely in relative amplitude and form with the alpha and delta peaks of the compound action potential. In fact, however, the large-diameter fibers in the second peak on the right in the fiber diameter spectrum conduct impulses with the higher velocity and account for the alpha potential to the left in the compound action potential. The histogram plotted in the opposite way presents the spectrum in order of decreasing diameter and conduction velocity, just as the components of the compound action potential occur in order of decreasing conduction velocity. The largest fibers present in significant numbers are 11 to 12 ␮m in diameter. If these fibers account for the beginning of the A␣ potential conducted at 61 m/s, the ratio of

Compound Action Potentials of Sural Nerve in Vitro

FIGURE 39–1 Compound action potentials (A) and fiber diameter histograms (B and C) of fiber diameters of normal nerve. Myelinated fibers are represented on the left, unmyelinated fibers on the right. In B, fiber sizes are shown from left to right as small to large. To show how the size distribution peaks relate to the A␣, A␦, and C potentials (A), in C, the diameter distribution is plotted from left to right as large to small fibers.

1019

1020

Nerve Conduction and Electromyography

conduction velocity to fiber diameter is about 5.3:1. Because varying amounts of shrinkage occur with the use of different histologic methods, this ratio may vary. There is no discontinuity between the alpha and delta components of the A potential in cutaneous nerve. Similarly, there is no discontinuity in the fiber diameter histogram between the distribution of large and small myelinated fibers. The alpha and delta peaks result from a high concentration of fibers in regions of a continuous spectrum of fiber diameters.2,20 From an evaluation of the amplitude and of the conduction velocity of various components of the compound action potential, some inferences can be made about numbers, sizes, and abnormalities of fibers in the nerve evaluated. It is usually assumed, in relating the compound action potential to the histogram of fiber diameters, that the diameter of fibers in a section is the same along the length of the nerve. Even in healthy nerve this is only approximately correct. In myelinated fibers of normal appearance, considerable variability in diameter occurs along their length. The fiber usually has a greater diameter in paranodal regions and lesser diameters at nodes of Ranvier and at the region of the Schwann cell nucleus. This means that the histogram

from a section always includes values that are spuriously large and small. In addition to this variability in diameter along the length of all myelinated fibers, occasional fibers will show pathologic alterations. Occasional fibers will be undergoing degeneration into linear rows of myelin ovoids, and occasional fibers will show remyelinated segments after previous segmental demyelination. For more information on the number of myelinated and unmyelinated fibers per square millimeter of fascicular area, on the frequency distribution of diameters of myelinated and unmyelinated fibers, and on the frequency of various pathologic abnormalities of myelinated fibers in healthy nerves, see Chapter 32. Friedreich’s Ataxia The compound action potential of the sural nerve has as its principal characteristic a great reduction in amplitude of the A␣ potential with preservation of A␦ and C potentials (Fig. 39–2A). Conduction velocity of the A␣ potential was 61 m/s (normal, ⬎49 m/s). The action potential correlates well with the histogram of fiber diameters, in which largediameter myelinated fibers are decreased in number (Fig. 39–2B). The largest diameter fibers were 10 to 11 ␮m.

FIGURE 39–2 Friedrich’s ataxia: Compound action potentials (A) and fiber diameter histograms (B).

Compound Action Potentials of Sural Nerve in Vitro

Similar compound action potentials were obtained from sural nerves of three other patients with Friedreich’s ataxia. In these the amplitude of the A␣ potential was also small. Conduction velocity of A␣ fibers was slightly low in one (46 m/s). In two the A␦ potential was of slightly less than normal amplitude. The relatively selective loss of large-diameter fibers is associated with a loss of touch-pressure and two-point sensations, but without loss of pain, warmth, and cold sensations. In the clinical electromyogram, cutaneous nerve action potentials are very low in amplitude or not detectable in routine examinations that record the action potential of large-diameter fibers. The nature of myelinated fiber degeneration in Friedreich’s ataxia has been extensively studied. Early in the course of the disease, at a time when large fibers are already decreased in number, the peak of the small fiber group in histograms of fiber diameter may be excessively high and displaced to smaller diameter categories than in

1021

those of control nerves. This observation raised the question of whether large degenerating fibers were undergoing slow axonal atrophy and were appearing successively in smaller diameter categories. An extensive evaluation of teased fibers has tended to support this hypothesis. It was concluded that this is a neuronal disease in which axonal atrophy and degeneration with secondary segmental demyelination affect large-diameter axons first and predominantly. For a more extensive discussion of the clinical, nerve conduction and electromyographic, and pathologic characteristics of Friedreich’s ataxia, see Chapter 78. Dominantly Inherited Amyloidosis In the compound action potential of this nerve, there is no C potential and the A␦ potential is greatly reduced in size (Fig. 39–3A). In contrast, the A␣ potential is only moderately reduced in amplitude (one third to one half of normal) and has an essentially normal contour. The conduction

FIGURE 39–3 Dominantly inherited amyloidosis: Compound action potentials (A) and fiber diameter histograms (B).

1022

Nerve Conduction and Electromyography

velocity of the A␣ potential is normal (54 m/s). The configuration of the compound action potential correlates well with the histogram of fiber diameters, which shows a virtual absence of C fibers and a reduction in number of myelinated fibers, particularly the small-diameter myelinated fibers (Fig. 39–3B). The abnormalities are consistent with the selective loss of pain sensation, temperature discrimination, and autonomic function in this patient. In this condition, nodules of amyloid are widely distributed within connective tissue in the epineurium, perineurium, and endoneurium and particularly within the walls of surrounding blood vessels. Ischemia from vessel involvement and direct compression of fibers by nodules may be mechanisms of nerve damage. Indentations of myelinated fibers resulting in a saccular dilatation of the fiber above and below the adjacent amyloid nodule have been described. Remyelinated segments may be seen adjacent to an amyloid nodule. An adequate explanation for the selective loss of unmyelinated fibers has not been discovered. Tabes Dorsalis The compound action potential and the histogram of fiber diameters are normal for the age of the patient (Fig. 39–4). Although all modalities of sensation were

affected in this patient, the fiber composition of the peripheral nerve is preserved because the radicular lesion is proximal to the dorsal root ganglion and has not caused degeneration of cell bodies or peripheral axons. Hereditary Sensory Neuropathy Type I The A␣, A␦, and C potentials are all greatly reduced in amplitude and area. The greatest decrease is in the C and A␦ potentials. Conduction velocity of the three peaks was normal. The histogram demonstrates that the numbers of myelinated and unmyelinated fibers in the nerve also were greatly decreased (Fig. 39–5). Large myelinated fibers were not as markedly decreased in number as the small myelinated and the unmyelinated fibers, consistent with the slightly better preservation of the A␣ potential and lesser impairment of touch-pressure than pain and temperature sensation. The genetic, clinical, and pathologic features of this disorder are described in detail in Chapter 78. Hereditary Sensory Neuropathy Type II The compound action potential of the sural nerve had no detectable A␣ or A␦ components. It is shown below that of

FIGURE 39–4 Tabes dorsalis: Compound action potentials (top) and fiber diameter histograms (bottom).

Compound Action Potentials of Sural Nerve in Vitro

1023

FIGURE 39–5 Hereditary sensory neuropathy type I: Compound action potentials (top) and fiber diameter histograms (bottom).

a normal nerve (unaffected father) in Figure 39–6A. The C potential was very low in amplitude and its conduction velocity was on the lower border of the normal range. The nerve contained no myelinated fibers, and the number of unmyelinated fibers was about half the normal number (Fig. 39–6B). In particular, there was a reduction in number of the larger diameter unmyelinated fibers, and the peak of the histogram was at approximately 0.35 ␮m (normal, 0.5 to 0.75 ␮m) in keeping with the low velocity of the C potential. The genetic, clinical, and pathologic characteristics of this disorder are described in Chapter 78. Uremic Neuropathy All components of the compound action potential are markedly reduced in amplitude in this patient with a moderately severe neuropathy (Fig. 39–7A). Conduction velocity of the A␣ component was low (30 m/s), and the action potential of the A group is dispersed. Conduction velocity of the C potential was normal (1.6 m/s). There was a reduction in number of both myelinated and unmyelinated fibers (about one third of normal) in the

nerve. There was also an absence of large-diameter fibers (⬎7 ␮m) in the histogram (Fig. 39–7B). The reduction in number of fibers was less than expected from the reduction in the size of the compound action potential; it is possible that some fibers included in the histogram did not conduct an action potential. The histogram of fiber diameter did not reveal how severely affected myelinated fibers of this nerve were, but study of teased fibers revealed that most were abnormal. More than 50% had degenerated into linear rows of myelin ovoids and balls (condition E). Lesser numbers had excessive irregularities of the myelin sheath (condition B), segments of demyelination (condition C), and segments of demyelination and remyelination (condition D). The much higher density of myelinated fibers per square millimeter and the much lower frequency of teased fibers showing condition E in the sural nerve taken from the midcalf as compared to the ankle level indicated that the pathologic alterations were much more severe in the distal parts of the nerve. This and other evidence favored the view

1024

Nerve Conduction and Electromyography

FIGURE 39–6 Hereditary sensory neuropathy type II. A, Compound action potentials of healthy subject (top) and patient with type II hereditary sensory neuropathy (bottom). B, Fiber diameter histograms.

that axonal atrophy and degeneration were primary (see Chapter 87). Relapsing Inflammatory Polyradiculoneuropathy All components of the compound action potential probably are represented (Fig. 39–8A). The A component has a very small amplitude and is dispersed. Conduction velocity of the first wave is very slow (11 m/s). Presumably this is produced by A␣ fibers, but the rate of conduction in many of the A␣ fibers may be so low that their action potentials are mixed with those of the A␦ fibers. The C fiber potential is normal in size. Some of the more slowly conducting, segmentally demyelinated A fibers may contribute to this

potential, however, and may account for its unusual shape and irregularity. The spectrum of fiber diameters does not correspond closely with the form of the compound action potential when there is segmental demyelination (Fig. 39–8B). The mean conduction velocity over the length of a myelinated fiber is determined by the mean of the velocity of conduction from one internode to the next over many internodes. When the internodes are uniform, as they usually are in healthy nerve, the diameter of the fiber at one internode may be representative of the fiber and be related to the mean conduction velocity (Fig. 39–9). When there is segmental demyelination and remyelination, there may be a wide disparity between diameters of internodes. The

Compound Action Potentials of Sural Nerve in Vitro

1025

FIGURE 39–7 Uremic neuropathy: Compound action potentials (A) and fiber diameter histograms (B).

diameter in a single section may not be representative of the fiber and therefore is not a reliable indicator of its conduction velocity. Hereditary Motor and Sensory Neuropathy Type I (Hypertrophic Neuropathy of Charcot-Marie-Tooth Type) The A␣ and ␦ potentials are reduced in size and increased in duration, but the C potential is essentially

normal (Fig. 39–10). Conduction velocity of the A␣ and ␦ potentials is low. The velocity of the beginning of the A␣ potential is 17 m/s; that of the beginning of the C potential, 1.1 m/s, is normal. Although the transverse fascicular area of the sural nerves in this disorder is abnormally large, the total number and the density of myelinated fibers in the nerve are decreased. The largest myelinated fibers are smaller than normal, and evidence of segmental demyelination and

1026

Nerve Conduction and Electromyography

FIGURE 39–8 Relapsing inflammatory polyradiculoneuropathy: Compound action potentials (A) and fiber sample (B).

remyelination is found with a variable degree of onion bulb formation. The pathologic alterations in the myelinated fibers are sufficient to explain the abnormality of the A potential. An example of the compound action potential in the hypertrophic neuropathy of the Déjérine-Sottas type is given in Chapter 99. In both these disorders, the apparent preservation of A␣ and ␦ elevations, in some cases despite marked slowing of conduction, suggests that there may be relatively uniform involvement of these two fiber groups.

COMMENTS The conformation of the compound action potential of a nerve is determined by the relative density of functioning nerve fibers in different regions of the conduction velocity spectrum. In the evaluation of a patient with peripheral neuropathy, the compound action potential of a fascicular biopsy of sural nerve can give an indication of the populations of cutaneous nerve fibers that are affected and of the general nature of the pathologic alterations.

Acute degeneration of some myelinated nerve fibers may leave others whose conduction velocity remains essentially or nearly normal. Although the amplitude of the A potential is smaller than normal, the alpha and delta elevations can be identified. If degenerating and nonconducting fibers can be excluded from the fiber diameter histogram, a reasonable correspondence between it and the compound action potential is found. Widespread segmental demyelination and remyelination within nerve is associated with a low conduction velocity of myelinated fibers. The alpha and delta potentials may remain as discrete potentials if there is uniform involvement of all fiber groups. More often, however, there is disproportionate slowing of conduction in different fibers, and the alpha and delta potentials may be widely dispersed and merge so that alpha and delta elevations cannot be identified reliably. The potentials of alpha fibers are mixed with those of delta fibers. The fiber diameter histogram does not correlate well with the compound action potential because the wide variation in diameter of successive internodes of a nerve fiber prevents reliable prediction of its conduction velocity.

Compound Action Potentials of Sural Nerve in Vitro

FIGURE 39–9 Uniform internodes in healthy nerve; the diameter of the fiber at one internode may be representative of that fiber and be related to the mean conduction velocity.

FIGURE 39–10 Hereditary motor and sensory neuropathy type I: Compound action potentials.

1027

1028

Nerve Conduction and Electromyography

REFERENCES 1. Bessow, P., Burgess, P. R., Perl, E. R., Taylor, C. B.: Dynamic properties of mechanoreceptors with unmyelinated C fibers. J. Neurophysiol. 34:116, 1971. 2. Bishop, G. H.: My life among the axons. In Hall, V. E. (ed.): Annual Review of Physiology. Palo Alto, CA, Annual Reviews, p. 1, 1965. 3. Bishop, G. H., Erlanger, J., and Gasser, H. S.: Distortion of action potentials as recorded from the nerve surface. Am. J. Physiol. 78:592, 1926. 4. Bishop, G. H., and Heinbecker, P.: Differentiation of axon types in visceral nerves by means of the potential record. Am. J. Physiol. 94:170, 1930. 5. Boyd, I. A.: Differences in diameter and conduction velocity of motor and fusimotor fibres in nerves to different muscles in the hind limb of the cat. In Curtis, D. R., and McIntyre, A. C. (eds.): Studies in Physiology Presented to J. C. Eccles. Berlin, Springer-Verlag, p. 7, 1965. 6. Boyd, I. A., and Davey, M. R.: Composition of Peripheral Nerves. Edinburgh, Churchill Livingstone, p. 57, 1968. 7. Brown, G. I., and Holmes, O.: The effects of activity on mammalian nerve fibers of low conduction velocity. Proc. R. Soc. Lond. [Biol.] 145:1, 1956. 8. Collins, W. F., Jr., Nulsen, F. E., and Randt, C. T.: Relation of peripheral nerve fiber size and sensation in man. A.M.A. Arch. Neurol. 3:381, 1960. 9. Douglas, W. W., and Ritchie, J. M.: Non-medullated fibres in the saphenous nerve which signal touch. J. Physiol. (Lond.) 139:385, 1957. 10. Dyck, P. J., and Lambert, E. H.: Compound nerve action potentials and morphometry. Electroencephalogr. Clin. Neurophysiol. 36:561, 1974. 11. Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerve in health, Friedreich’s ataxia, hereditary sensory neuropathy and tabes dorsalis. In Remond, A. (ed.): Handbook of Electroencephalography and Clinical Neurophysiology, Vol 9. Amsterdam, Elsevier, p. 83, 1971. 12. Erlanger, J.: The interpretation of the action potentials in cutaneous and muscle nerves. Am. J. Physiol. 82:644, 1927. 13. Erlanger, J., and Gasser, H. S.: Electrical Signs of Nervous Activity. Philadelphia, University of Pennsylvania Press, 1937. 14. Gasser, H. S.: Unmedullated fibers originating in dorsal root ganglia. J. Gen. Physiol. 33:651, 1950.

15. Gasser, H. S.: Properties of dorsal root unmyelinated fibers on the two sides of the ganglion. J. Gen. Physiol. 38:709, 1955. 16. Gasser, H. S.: Effect of the method of leading on the recording of the nerve fiber spectrum. J. Gen. Physiol. 43:927, 1960. 17. Gasser, H. S., and Erlanger, J.: The role played by the sizes of the constituent fibers of a nerve trunk in determining the form of its action potential wave. Am. J. Physiol. 80:522, 1927. 18. Gasser, H. S., and Grundfest, H.: Axon diameters in relation to the spike dimensions and the conduction velocity in mammalian A fibers. Am. J. Physiol. 127:393, 1939. 19. Heinbecker, P., Bishop, G. H., and O’Leary, J.: Pain and touch fibres in peripheral nerves. Arch. Neurol. Psychiatry 29:771, 1933. 20. Heinbecker, P., Bishop, G. H., and O’Leary, J.: Functional and histologic studies of somatic and autonomic nerves of man. Arch. Neurol. Psychiatry 35:1233, 1936. 21. Hursh, J. B.: The conduction velocity and diameter of nerve fibers. Am. J. Physiol. 127:131, 1939. 22. Hursh, J. B.: The properties of growing nerve fibres. Am. J. Physiol. 127:140, 1939. 23. Iggo, A.: Cutaneous mechanoreceptors with afferent C fibres. J. Physiol. (Lond.) 152:337, 1960. 24. Iriuchijima, J., and Zotterman, Y.: The specificity of afferent cutaneous C fibres in mammals. Acta Physiol. Scand. 49:267, 1960. 25. Landau, W. M., Clare, M. H., and Bishop, G. H.: Reconstruction of myelinated nerve tract action potentials: an arithmetic method. Exp. Neurol. 22:480, 1968. 26. Lloyd, D. P. C.: Neuron patterns controlling transmission of ipsilateral hind limb reflexes in cat. J. Neurophysiol. 6:293, 1943. 27. Maruhashi, J., Mizuguchi, K., and Tasaki, I.: Action currents in single afferent nerve fibres elicited by stimulation of the skin of the toad and the cat. J. Physiol. (Lond.) 117:129, 1952. 28. Ritchie, J. M., and Straub, R. W.: The after-effect of repetitive stimulation on mammalian non-medullated fibres. J. Physiol. (Lond.) 134:698, 1956. 29. Rud, J.: Local anesthetics: an electrophysiological investigation of local anesthesia of peripheral nerves with special reference to Xylocaine. Acta Physiol. Scand. Suppl. 51:178, 1961. 30. Rushton, W. A. H.: Resistance artefacts in action potential measurements. J. Physiol. (Lond.) 104:19, 1945. 31. Rushton, W. A. H.: The site of excitation in the nerve trunk of the frog. J. Physiol. (Lond.) 109:314, 1949. 32. Schaefer, H., and Schmitz, W.: Actionsstrom und Hullenleitfahigkeit. Arch. Gesamte Physiol. 234:737, 1934. 33. Zotterman, Y.: Touch, pain and tickling: an electrophysiological investigation on cutaneous sensory nerves. J. Physiol. (Lond.) 95:1, 1939.

40 Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes PETER J. DYCK, RICHARD A. C. HUGHES, AND PETER C. O’BRIEN

Overview History of Scoring Neurologic Signs Definitions End Points to Be Scored Symptom Scores Volunteered Symptoms The Neurological Symptom Score The Neuropathy Symptoms Profile The Total Symptom Score The Positive Neuropathic Sensory Symptom Score: Severity and Constancy The Neuropathy Symptoms and Change Score Pain Scores

Autonomic Scores Scoring Muscle Weakness (and Strength) Quantitating Muscle Strength Quantitating Sensory Loss Muscle Stretch Reflexes Muscle Twitch Composite Scores of Neurologic Signs History The Neuropathy Impairment Score Nerve Conduction Tests as Impairment Measures Composite Scores Employing Varying Combinations of Symptoms, Impairments, and Test Results

OVERVIEW This chapter lists and describes some of the more important clinical instruments, scores, and approaches used to quantitate overall neuropathic symptoms, neurologic deficits, dysfunctions, and outcomes of patients with peripheral neuropathy at a given time. Quantitating overall severity of a patient’s peripheral neuropathy may be useful for medical practice (to decide whether treatment should be initiated or to monitor its effects), or for conduct of epidemiologic surveys or prospective controlled clinical trials. Other chapters in this textbook contain information closely related to this topic, for example, the motor system (Chapters 33, 35, 37, and 43), the sensory system (Chapters 8, 20, 34, 38, 39, and 42), and the autonomic system (Chapters 10 through 15 and 44). Monographs on this subject are also available.6,8,27,28,48–51 This chapter focuses on overall assessment of symptoms, dysfunctions, neurologic signs, impairments, and outcomes

Electronic Case Report Form of the Clinical Neuropathy Assessment Disability Scores General Scores Health-Related Quality of Life and Clinical Outcomes Score in Peripheral Neuropathy Other Outcomes Morphometry of Nerve or Nerve Fiber Counts in Skin Biopsy Fatigue Handicap Quality of Life

as judged by medical personnel. The emphasis here is on objective and expert assessment by an impartial observer, as compared to relying mainly on the subjective experience of the patient. The focus, therefore, is on quantitating weakness, reflex decrease, sensation dysfunction and loss, autonomic dysfunction and loss, and functional disability for work, activities of daily living, and recreational activity. Also included here is assessment of symptoms because they are the complaints of patients, bring patients to physicians, and may be treatable. However, the specific symptoms we are especially interested in here are ones directly attributable to dysfunction or loss of nerve fibers. The nature, distribution, temporal characteristics, and anatomic distribution of these specific symptoms should correlate with neurologic findings. The more general subjective experiences and reactions of patients, although also important, are discussed in Chapter 41. Because varieties of peripheral neuropathy differ in course, population of neurons or fibers affected, proximal to distal levels of involvement, type of 1031

1032

Quantitating Symptoms, Impairments, and Outcomes

pathologic alterations, and outcomes, a single assessment approach is not suitable for all varieties. The broad grouping of approaches described here includes symptom scores, composite summated scores of neurologic findings, composite scores of quantitative sensation loss, scores of dysfunction in activities of daily living, scales of disability, composite scores of nerve conduction, and quantitation of histologic measures of pathologic abnormality (morphometry of biopsied nerves or of nerve endings). Generally, the instruments described here should be judged by the criteria of objectivity, accuracy, sensitivity, specificity, representativeness, reproducibility, and concordance with clinical impairment. For some tests, intraobserver and interobserver reproducibility is also important. Monotonicity is a further desired feature of measured end points. This measure addresses whether a test shows a consistent trend over time. The use of a normative database and of percentiles, normal deviates, and composite scores, particularly as they relate to nerve conduction abnormality, is a recent important development that is described in Chapter 36. Quality-of-life scores are discussed in Chapter 41. Here we describe the clinical instruments, scales, and outcome measures used to characterize and quantitate overall severity of symptoms, signs, test results, specific functional and structural derangements, and functional impairment and disability attributable to peripheral neuropathy. Therefore, the focus is on more than assessing single neurologic symptoms, signs, or test results as markers of disease—it is a more global representation that is being sought. The assessments described here may be used to track severity simply for information, to recognize a treatment effect, or to predict outcome. Specifically, these approaches are useful for cross-sectional or longitudinal epidemiologic surveys and for assessing therapy in prospective controlled clinical trials.

HISTORY OF SCORING NEUROLOGIC SIGNS Standardized instruments for the global evaluation of symptoms, impairments, and outcomes are a recent development. Beginning in 1912, Henry Plummer, a few years later Walter Sheldon, and then Henry Woltman at the Mayo Clinic in Rochester, Minnesota, began to systematically grade muscle weakness as 0 ⫽ normal and ⫺1 ⫽ slight, ⫺2 ⫽ moderate, ⫺3 ⫽ considerable, and ⫺4 ⫽ maximal weakness. Muscle stretch reflexes were similarly graded ⫺1 to ⫺4 (decreased to absent) or as ⫹1 to ⫹4 (increased). Tactile, pinprick, joint position and motion, and vibration sensation were also graded as 0 to ⫺4 and recorded on a cartoon drawing of the body. In 1919, results were entered into standard neurologic case report forms. Standard single

(one-half page), double (full page), and triple (11/2 page) forms and expanded muscle and sensation examination forms were introduced. The forms were modified on several occasions. The use of these forms was extensively described in Clinical Examination in Neurology, first published in 1956.39 Later, grading of muscle weakness was further defined: ⫺1 was 25%, ⫺2 was 50%, and ⫺3 was 75% weak; ⫺4 was paralyzed. In 2003, the paper forms were replaced by an electronic form, completed on line and entered into a database. As the result of poliomyelitis epidemics, R. W. Lovett, a Boston orthopedist, in 1917 introduced a list of muscles to be tested and graded their strength as 1 ⫽ “Normal”; 2 ⫽ “Good, the muscle is strong enough to overcome gravity and some resistance,” but is “not quite of normal strength”; 3 ⫽ “Fair,” the muscle is “able to overcome gravity” and can “perform part of the normal movement”; 4 ⫽ “Poor,” slight movement may be accomplished “but gravity cannot be overcome”; 5 ⫽ “Trace, no movement is possible but the muscle is felt to contract”; and 6 ⫽ “Paralyzed.”36 He also outlined a functional gait scale that included “cannot walk, walks unaided, walks with brace, walks with crutches, walks with brace and crutches, and walks with brace, crutches, and corset,” similar to a later scale published by Hughes et al.31 In 1943, the Medical Research Council (MCR) (United Kingdom) published Aids to the Investigation of Peripheral Nerve Injuries (War Memorandum No. 7).38 Although the action of muscles had already been described and illustrated in earlier publications, this document provided a set of easily understood line drawings of the actions of limb muscles, which was widely distributed and used. The “Aids” appear to have adopted Lovett’s 5-part grading approach of muscle weakness but changed the order of grading. For Lovett, grade 1 was normal strength; for the MRC scale, 0 was paralyzed. This difference in the assignment of grades made sense because Lovett wanted to emphasize the degree of weakness relative to normal after poliomyelitis, whereas the MRC emphasized recovery from paralysis after nerve injury. In their emphasis on grading improvement from the “greatest loss of power as a result of nerve wounds,” the MRC followed the lead of S. Weir Mitchell,47 who had written Injuries of Nerves and Their Consequences in 1872. Mitchell emphasized sensory experiences and loss, whereas the MRC emphasized only muscle weakness. The MRC grades were 0 ⫽ no contraction, 1 ⫽ flicker or trace of contraction, 2 ⫽ active movement with gravity eliminated, 3 ⫽ active movement against gravity, 4 ⫽ active movement against gravity and resistance, and 5 ⫽ normal power. The MRC notation has found wide acceptance and is used widely in neurologic practice, but it has serious shortcomings for this purpose. It is not an ideal grading approach for neuropathy because grading is not linear; the grading is much broader for mild to severe weakness (e.g., grade 4) than for severe weakness (grades 1

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

through 3). Recognizing this difficulty, many neurologists add additional categories for grade 4 (4⫹ and 4⫺). Also, the grades are not logically applicable for assessing the integrity of muscle stretch reflexes and sensation loss.

DEFINITIONS Such terms as impairment, handicap, and disability are often used loosely and therefore need definition. Impairment The World Health Organization defines impairment as it relates to health as “any loss or abnormality of psychological, physiological, or anatomical structure or function.”58 In its most recent revision of its classification of disablement, impairment was more precisely defined as a health dimension consisting of “a loss or abnormality of a body part or body function (i.e., physiological function)” that may be detected as “a significant variation from established statistical norms.”59 The U.S. Social Security Administration/American Medical Association’s Disability Evaluation Under Social Security49 includes laboratory evaluations as acceptable and even necessary components of assessment of impairment. Consequently nerve conduction abnormalities, electromyographic abnormalities, tests of motor, sensory, or autonomic dysfunction, and even pathologic alteration of nerve endings or nerves may be evidence of impairment, even if not accompanied by symptoms or abnormal signs on clinical examination. Disability The World Health Organization originally defined disability as “any restriction or lack (resulting from an impairment) of ability to perform an activity in the manner or within the range considered normal for a human being.”58 In its most recent revision, the term disability was replaced by the converse concept of “activities,” defined as “the nature and extent of functioning at the level of the person. Activities may be limited in nature, duration and quality.”59 In this definition, activities, and conversely their limitation, depend on the actual rather than the potential performance of the individual being assessed. The authors prefer the first definition. The U.S. Social Security Administration’s definition of disability has a more restricted and somewhat different meaning, focusing on providing government income to patients who have lost their ability to work. The definition is: “The inability to engage in any substantial gainful activity by reason of any medically determinable physical or medical impairment which can be expected to result in death or has lasted, or can be expected to last, for a continuous period of time, not less than 12 months.”49

1033

Handicap The World Health Organization first defined handicap as “a disadvantage for a given individual, resulting from an impairment or a disability, that limits or prevents the fulfillment of a role that is normal (depending on age, sex, and social and cultural factors) for that individual.”58 In the 1997 revision, the term is replaced with the converse, “participation,” which is defined as “the nature and extent of a person’s involvement in life situations in relation to Impairments, Activities, Health Conditions and Contextual factors. Participation may be restricted in nature, duration and quality.”59 This dimension refers to the interaction of the individual with his or her environment, and “limitation of participation” refers to the extent to which the individual is handicapped in his or her societal role. The authors prefer the first definition. Symptoms Although the U.S. Social Security Administration does not accept symptoms alone as the basis for disability in patients whose neuropathy is expressed mainly as pain, physicians can usually find sufficient objective “anatomical, physiological, or psychological concomitants” to meet the necessary requirement of impairment.49 Obviously the intent of the Social Security Administration is to withhold disability status to persons who claim symptoms without having them, simply for financial gain or because they do not want to work. Outcome Usually the term outcome refers to the primary variable of a study. It can be an outcome event, for example, development of a plantar ulcer, disability in walking, or inability to reach a specific ordinal scale of function. Quality of Life According to Simon Day,11 quality of life refers to a broad assessment of how a subject feels about his or her life. It includes considering subjects’ physical and mental status and their degree of satisfaction relative to some ideal standard. Ordinal Scale An ordinal scale is a grading scale that defines the position of the characteristic or of staged dysfunction by a series of whole numbers or by the letters of the alphabet (a, b . . . z). Surrogate Measure A surrogate measure is a substitute variable for the most clinically meaningful measure. For Guillain-Barré syndrome (GBS), the most important primary measure may be the degree of muscle weakness or paralysis. Assuming this to be the case, a surrogate measure might be a composite score consisting of the summated normal deviates of the compound muscle action potential of ulnar, median, peroneal,

1034

Quantitating Symptoms, Impairments, and Outcomes

and tibial nerves with percentile abnormalities expressed in the upper tail of the normal distribution.

SYMPTOM SCORES Volunteered Symptoms

END POINTS TO BE SCORED It should be appreciated that what the primary outcome measure of polyneuropathy should be is not always readily apparent and may be a disputed issue. Thus, in sensorimotor polyneuropathy expressed mainly as positive neuropathic sensory symptoms, the primary measure probably might be the symptoms themselves. The end points to be scored depend on the primary reason for the scoring; the variety, pattern, and course of the neuropathy studied; the time, resources, and personnel available to do the evaluations; and the degree of assessment patients will agree to. If the purpose of the study is to assess for the effectiveness of an intervention (e.g., plasma exchange) on overall neuropathic dysfunction, it is reasonable to grade this overall dysfunction serially, using a composite score of neurologic findings or a functional scaling of degree of disability. By contrast, in trigeminal neuralgia, neither of these approaches is suitable. Instead, a pain scale that assesses the frequency and severity of the pain itself is needed. The anatomicopathologic involvement, therefore, needs to be taken into account in choosing end points. To illustrate, in diabetic lumbosacral radiculoplexus neuropathy, a composite score of neurologic impairment of the affected lower limb is a better measure than one of neurologic impairments of the entire body. For many neuropathies, a major choice will be between functional impairment scales and sum scores of selected neurologic signs. Functional impairment scales are preferred by some investigators because differences between ordinal grades represent large and important functional differences and therefore represent clinically meaningful degrees of change, and usually judgments can be made quickly and without complex and expensive instruments. The disadvantages to their use are that they may be excessively dependent on patient motivation and secondary gain, and they may not be sufficiently sensitive to small alterations. Composite scores of neurologic deficit (assuming assessment is done by physicians who are specially trained and certified) tend to be more objective and can be used to recognize smaller changes. We (P.J.D. and P.C.O.) have found that, for conduct of prospective trials using composite scores of neurologic signs, training and certification of investigators is needed to reduce variability to an acceptable level. Specific assessments of such examinations as vibration detection threshold (VDT) with CASE IV (Computer-Assisted Sensory Examination System IV, using validated software algorithms of testing and percentile expression of findings, taking applicable variables into account) correlate better with nerve conduction abnormality than does the clinical examination even of highly trained neurologists (P. J. Dyck and P. C. O’Brien, unpublished data, 2004).

Although the problem-based approach has proven to be very useful in clinical medicine, it is generally not ideal for quantitating symptoms of peripheral neuropathy because (1) the approach is not standardized, (2) not all symptoms are remembered and volunteered or asked about on functional inquiry, and (3) it is difficult to quantify results.

The Neurological Symptom Score The Neurological Symptom Score (NSS) is a checklist of 18 symptoms, including muscle weakness and negative and positive sensory symptoms (Table 40–1).28 The inquiry about symptoms should be done prospectively, the interview asks about all 18 items, and only symptoms judged to be due to the studied neuropathy are to be tallied and judged by an experienced physician. Thus such symptoms as tiredness Table 40–1. The Neurological Symptom Score Score 1 point for the presence of each symptom: I. Symptoms of muscle weakness A. Bulbar 1. Extraocular 2. Facial 3. Tongue 4. Throat B. Limbs 5. Shoulder girdle and upper arm 6. Hand 7. Hip girdle and thigh 8. Leg II. Sensory disturbance A. Negative symptoms 9. Difficulty identifying objects in mouth 10. Difficulty identifying objects in hands 11. Loss of feeling in feet or unsteadiness (touch, heat, pain) B. Positive symptoms 12. Numbness, asleep feeling, like Novocain, prickling at any site 13. Pain—burning, deep aching, tenderness at any site III. Autonomic symptoms 14. Postural fainting 15. Impotence 16. Loss of urinary control 17. Night diarrhea 18. Gastroparesis Total

Score

––––– ––––– ––––– ––––– ––––– ––––– ––––– –––––

––––– ––––– –––––

––––– –––––

––––– ––––– ––––– ––––– ––––– –––––

Modified from Dyck, P. J., Sherman, W. R., Hallcher, L. M., et al.: Human diabetic endoneurial sorbitol, fructose, and myo-inositol related to sural nerve morphometry. Ann. Neurol, 8:590, 1980.

1035

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

should not be misinterpreted as weakness, and increasing weakness or impotence in men greater than 60 years of age should not be tallied as a symptom because it occurs quite frequently in healthy men of this age. It is necessary for the evaluator to know what is normal/abnormal for age, sex, anthropomorphic factors, and physical fitness. Only symptoms related to the peripheral neuropathy being studied are entered. Also, the physician judges whether the degree of asleep-numbness, prickling, or lancinating pain is physiologic (due to age), is a physiologic reaction (e.g., compression of a nerve by assuming certain positions for a prolonged time), or is pathologic and should be tallied because it is due to the specific neuropathy being studied. The NSS has been useful for screening and characterization purposes. For some purposes, severity and constancy of symptoms should also be included (see Total Symptom Score [TSS] and Positive Neuropathic Sensory Symptom score: Severity ⫻ Constancy [P-NSS-S ⫻ C] below).

Table 40–2. Scoring Approach for the Neuropathic Symptoms Included in the Total Symptom Score (Pain, Burning, Paresthesia, Numbness) Symptom Intensity Symptom Frequency Occasional Frequent (Almost) continuous

Absent

Slight

0 0 0

1.00 1.33 1.66

Moderate 2.00 2.33 2.66

Severe 3.00 3.33 3.66

From Ziegler, D., Hanefeld, M., Ruhnau, K. J., et al.: Treatment of symptomatic diabetic peripheral neuropathy with the anti-oxidant alpha-lipoic acid: a 3-week multicenter randomized controlled trial (ALADIN Study). Diabetologia 38:1425, 1995, with permission.

by scanner or by fax transmission. Only symptoms attributed to the neuropathy studied (as judged by the examining physician) are graded. The maximum score is 45 points.

The Neuropathy Symptoms Profile The Neuropathy Symptoms Profile (NSP) is a 32-item patient-completed questionnaire producing an optically scanned and scored profile of symptoms encountered in peripheral neuropathy.16 Because patients answer queries without a medical personnel interface, it was meant to avoid the bias of physicians. Tests of reproducibility, anxiety and depression, and suggestibility were included. The score was standardized for healthy subjects and for specific neuropathies. Although the results were shown to discriminate symptom patterns in varieties of neuropathy (e.g., between motor neuron disease and amyloidosis), pattern specificity seemed to decrease with severity of neuropathy. It was also discovered that patients misinterpreted descriptions, which a physician interface would have avoided.

The Total Symptom Score The TSS, developed by Schuette, Lobisch, and Ziegler,60 evaluates four symptoms: pain, burning, prickling, and asleep-numbness. These symptoms are graded by intensity (absent, slight, moderate, or severe) and frequency (occasional, frequent, or continuous) (Table 40–2). This approach was used as the primary outcome measure in a short-term trial of ␣-lipoic acid in symptomatic neuropathy.1

The Positive Neuropathic Sensory Symptom Score: Severity and Constancy The P-NSS-S ⫻ C is a modification of the TSS to include five positive neuropathic sensory symptoms: asleep-numbness, prickling or tingling, burning or hot or cold pain, deep aching pain, and abnormal tenderness to mechanical contact (Fig. 40–1). The form is completed at the bedside or in the office, and entered into a database

The Neuropathy Symptoms and Change Score The Neuropathy Symptoms and Change score (NSC) is a 38-item questionnaire of symptoms encompassing symptoms of weakness (items 1 through 19), negative sensory symptoms (items 20 through 22), positive sensory symptoms (items 23 through 29), and autonomic symptoms (items 30 through 38). It is a component of the Clinical Neuropathy Assessment (CNA) form, which allows entry of responses by pencil in the office or at the bedside but also as an electronic record for submission of the data by fax or scanner and for entry into a database after programmed evaluation to ensure completeness, logical entry of information, conformity to a medical template, and accuracy. The CNA is described and illustrated below. As in the NSS, the evaluator (a physician) asks questions that take this general form: “Do you have [the symptoms asked about] at this time?” The interviewer makes the judgment as to whether the symptom is present at this time, and whether it is due to the neuropathy evaluated. If necessary, the interviewer should cross-examine the patient (sympathetically) and, if the symptom is present, inquire about severity and how it compares in severity to a previous time—better (slightly [⫹1], moderately [⫹2], or much [⫹3]), the same (0), or worse (⫺1 to ⫺3). The computer program tallies the number of symptoms, their severity, and the change (the patient’s judgment as to change of that symptom since an indicated previous time). If the NSC is scored on multiple occasions, change in severity from serial evaluation can also be determined.

Pain Scores In addition to Chapter 119, the reader should consult standard texts on pain assessments.

1036

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 40–1 The Positive Neuropathic Sensory Symptom (P-NSS) questionnaire. (Courtesy of P. James B. Dyck, M.D.)

Autonomic Scores For information on autonomic symptom scores, the reader is referred to Chapter 44.

SCORING MUSCLE WEAKNESS (AND STRENGTH) As described above, based on manual testing Lovett developed an ordinal grading approach to muscle weakness, ranging from 1 (normal) to 5 (flicker only) (6 presumably

would be paralyzed) that later was used by an MRC Committee but reversing the order of scaling so that 5 is normal and 0 is paralyzed. The MRC approach is employed worldwide in neurologic practice. In some respects, the broad use of this scaling approach is unfortunate because it was developed especially for nerve injuries and not for more general use. The scaling is not linear. This ordinal scoring approach provides fine scaling of very weak muscles, but there is insufficient scaling of muscles 25% to 75% weak. The Mayo Clinic neurologic examination approach, initially developed by Plummer, Sheldon, and Woltman

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

between 1910 and 1919, provides more grades of severity for milder weakness. The Neuropathy Impairment Score (NIS), developed by Dyck and co-workers, used the Mayo Clinic neurology approach but added components of the Lovett score. Muscle weakness was scored linearly (0 ⫽ normal; 1 ⫽ 25% weak; 2 ⫽ 50% weak; 3 ⫽ 75% weak; 3.25 ⫽ movement against gravity just possible; 3.5 ⫽ movement with gravity eliminated just possible; 3.75 ⫽ flicker; and 4 ⫽ paralyzed). In the Dyck et al. revision of the NIS,17,19 a further important feature is that grading is to be done for age, sex, anthropomorphic features, and physical fitness. The NIS is described and discussed below.

QUANTITATING MUSCLE STRENGTH Muscle strength can be assessed by simple devices and by more complicated strain gauge approaches. Measurement of strength of different muscle groups can be useful in longitudinal studies of individual patients and as an outcome measure in clinical trials. A variety of instruments are available. Grip strength can be measured by a springloaded device called the dynamometer. The Vigorimeter measures pressure generated by squeezing a rubber bulb. Grip strength values have been calculated for healthy control subjects of different ages. Measurements with this instrument have good intra- and inter-rater reliability and are responsive to change. Grip strength was highly significantly correlated with upper limb disability.42 Of course, the measurement has little value if upper limbs are not involved in the disease process, and it is necessary to achieve a maximum effort.

QUANTITATING SENSORY LOSS The assessment of sensation is discussed in Chapter 42. Clinical assessment of sensation of touch-pressure, vibration, joint position and motion, and pinprick of index finger and great toe for the determination of the NIS is discussed in a subsequent section. A simple bedside scale derived from the clinical examination has been developed and shown to have acceptable intra- and inter-rater reliability and responsiveness to change in inflammatory neuropathy.43 The Inflammatory Neuropathy Sensory Sum Score requires assessment of pinprick and vibration sensation in the upper and lower limbs and of two-point discrimination of the index finger (Table 40–3). Each item is scored from 0 ⫽ normal to 4 ⫽ severely abnormal. In the published version, the score of the most affected limb was rescaled so that scores range from 0 ⫽ normal to 20 ⫽ most abnormal. The changes in the sensory sum score changed as expected in cohorts of patients who worsened, remained

1037

Table 40–3. Inflammatory Neuropathy Sensory Sum Score* Pinprick Sensation A. Upper limb (most affected side) No sensory loss At fingers At wrist At elbows At shoulders B. Lower limb (most affected side) No sensory loss At toes At ankle At knees At groin

Score 0 1 2 3 4 Score 0 1 2 3 4

Vibration Sensation A. Upper limb (most affected side) No sensory loss On fingers: dorsal aspect of the index finger DIP joint At wrist: styloid process of the radius At elbows: medial epicondyle of the humerus At shoulders: acromioclavicular joint B. Lower limb (most affected side) No sensory loss At toes: dorsal aspect of the hallux DIP joint At ankle: medial malleolus At knee: patella At groin: anterior superior iliac spine

Score 0 1 2 3 4 Score 0 1 2 3 4

Two-Point Discrimination on the Index Fingertip (most affected side) 4 mm or less (normal) 5–9 mm 10–14 mm 15–19 mm 20 mm or more

Score 0 1 2 3 4

DIP ⫽ distal interphalangeal. *The sensory sum score is the sum of all five measurements and ranges from 0 ⫽ normal to 20 ⫽ maximum sensory deficit. From Merkies, I. S., Schmitz, P. I., van der Meche, F. G., and van Doorn, P. A., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Psychometric evaluation of a new sensory scale in immune-mediated polyneuropathies. Neurology 54:943, 2000, with permission.

stable, or improved and in individual patients with relapses and remissions of chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). The scale commends itself as a method for measuring change in longitudinal studies of patients with inflammatory neuropathy as well as in clinical trials. However, it does not include light touch or joint position sensation, and a scale that included these items would probably be more comprehensive and responsive. A simple semiquantitative alternative to measurement of vibration (probably better defined as flutter at the

1038

Quantitating Symptoms, Impairments, and Outcomes

frequency used) threshold with a computerized system is provided by a graduated 64-Hz tuning fork designed by Rydel and Seiffer.54 The patient is asked to state when he or she can no longer perceive the vibration, and the amplitude of vibration at this point is recorded from a graduated scale. Martina et al. have shown the expected decline in perception of vibration with age and provided values for a normal population.37 The measurements have excellent intra- and inter-rater reliability, correlate closely with those obtained with an electronic vibration threshold testing device, and are sensitive at detecting change in longitudinal studies and distinguishing patients with peripheral neuropathy from normal controls.37 Thus the instrument is acceptable for studying vibration threshold in situations in which time and resources do not permit the use of more sensitive computerized instruments. However, several potential problems should be mentioned: (1) defined algorithms of testing with null stimuli are not typically used, (2) stimulus magnitudes are not precisely known, and (3) disappearance thresholds tend not to be as accurate as appearance thresholds (see Chapter 42). To obtain accurate assessments of touch-pressure, vibration, cool, warm, and heat-pain thresholds, it is best to use systems that provide (1) stimuli having a waveform that produces the modality of sensation to be studied, whose magnitudes can be given over a broad range of quantifiable magnitudes (so that both sensitive and insensitive sites can be tested), and having appropriately fashioned null stimuli; (2) algorithms of testing so that threshold can be determined efficiently and reliably; (3) computer control so that the steps of testing and finding threshold follow predetermined rules and so that all aspects of stimulus and response and threshold are entered into a database and compared to control values; and (4) normal values that are expressed as percentiles and normal deviates. Touch-pressure sensation can be evaluated by von Frey’s hairs (a modern version being the Semmes-Weinstein monofilaments). The proponents generally assume that one stimulator and one level of correct responses fit all ages, both sexes, and different anatomic sites. Obviously use of only one monofilament and a single threshold cannot be accepted as an adequate test because thresholds are known to vary with site, age, and several physical variables. Cool discrimination using disks having different heat transfer characteristics can be used to study A␤ cool fiber receptors and fibers. Computerized systems (such as CASE IV and V) can be used to estimate vibration, cooling, and heat-pain threshold. We have recently shown that VDT measurements using CASE IV correlate more strongly than does the clinical neurologic examination with a composite score of nerve conduction function, both for a single-center epidemiologic study and for multicenter controlled trials (P. J. Dyck et al., unpublished data, 2004).

MUSCLE STRETCH REFLEXES Assessment of muscle stretch reflexes can be used as a measure of severity of neuropathy, as we do in the NIS (see below).21

MUSCLE TWITCH The force of a muscle contraction from a supramaximal nerve stimulus can be measured using special apparatus and strain gauges. The supramaximal muscle twitch of thenar muscle was found to be decreased even years after median nerve reconnection.12 We also found that the compound muscle action potential from supramaximal stimulation provided a reasonable correlate of maximal muscle force.

COMPOSITE SCORES OF NEUROLOGIC SIGNS History In 1980, we developed a standard sum-score of neurologic findings based on standard grading of muscle weakness, reflex loss, and decreased sensation so as to express overall severity of peripheral neuropathy. We originally named it the Neurologic Disability Score (NDS); later we renamed it the Neuropathy Impairment Score. By analogy to completing an insurance claim for a damaged house or automobile, all damaged items are listed, and repair or replacement costs are listed and summated. Likewise, the NIS lists all abnormal items of the neurologic examination, estimates the severity of each impairment, and summates all of the impairments and for both sides of the body. Some items of the NIS are shown in a subsequent section on the CNA.

The Neuropathy Impairment Score The unique features of the NIS are that it is a standard assessment of defined items of the neurologic examination for the assessment of peripheral neuropathy, scoring the entire body or subsets (lower limbs, legs, a single limb [or a single part of a limb] or other anatomic regions). Ordinal scoring is used, but age, sex, and anthropomorphic variables are taken into account in grading. Scoring of all items is with reference to controls considering age, sex, anthropomorphic characteristics, and physical fitness. Weakness is graded 0 ⫽ normal, 1 ⫽ 25% weak, 2 ⫽ 50% weak, and 3 ⫽ 75% weak; 3.25 ⫽ just able to move against gravity, 3.5 ⫽ just able to move with gravity eliminated, 3.75 ⫽ flicker of movement, and 4 ⫽ paralysis. Reflexes are graded as 0 ⫽ normal, 1 ⫽ decreased, and 2 ⫽ absent (reinforcement if needed is used). Sensation of the dorsum of the index fingers and great toes near the base of the nails is graded as

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

0 ⫽ normal, 1 ⫽ decreased, and 2 ⫽ absent and as tested with cotton wool (tactile), stick pins (pinprick), vibration (a standard tuning fork*), and joint position and motion. Studies at the Peripheral Neuropathy Research Laboratory at the Mayo Clinic have shown that reproducibility (as measured with intraclass correlation coefficients) is not as good for reflexes and for sensation as it is for assessment of muscle weakness.17 These investigators have also shown that even neuromuscular assessment trainees tend to overestimate sensory loss by not sufficiently taking applicable age, sex, and anthropomorphic characteristics into account.5 The NIS has also been extensively validated and used in epidemiologic and controlled clinical trials. In open and prospective controlled double-blind trials of prednisone, apheresis, or intravenous immune globulin in CIDP and in polyneuropathy associated with monoclonal gammopathy of undetermined significance, the use of the NIS was the primary end point used to demonstrate efficacy.13,18,20,21 In a preliminary open trial of methylprednisolone in diabetic lumbosacral radiculoplexus neuropathy, an improvement of the NIS and an improvement of symptoms as evaluated by the NSC occurred.25 In a study of renal and pancreatic transplantation, Dyck et al. found significant worsening, especially of weakness subscores of the NIS at 3, 6, 9, and 12 months after surgery, whereas the composite score of nerve conduction was significantly better at these times—an indication that NIS and nerve conduction composite scores are independent and perhaps different measures of severity of neuropathy.24 We assumed both observations reflected worsening and improvement accurately. In the Sydney Study (␣-lipoic acid in symptomatic diabetic sensorimotor polyneuropathy for 3 weeks), symptoms improved significantly (as measured by the TSS and also by the NSC), as did the NIS of lower limbs (NIS[LL]), but quantitative sensation and nerve conduction did not improve significantly.1 Because physicians administer the NIS, the question arises whether the evaluation can be performed reproducibly. We (three neurologists trained to do quantitative neurology evaluations) tested this question assessing 20 diabetic patients with and without diabetic sensorimotor polyneuropathy. Using intraclass correlation coefficients, average reproducibility was surprisingly good (~0.9), with small confidence intervals for the NDS (an earlier version of the NIS) and for the weakness subscore.17 In a large multicenter trial of therapy in diabetic sensory polyneuropathy, the correlation of the NIS, NIS(LL), and NIS(legs) with abnormality of nerve conduction was assessed. The NIS(LL) and

*The standard tuning fork one of us (P.J.D.) prefers is made of an aluminum alloy sheet 93/4⬙ ⫻ 11/4⬙ ⫻ 1/2⬙, has counterweights, and provides a 165-Hz stimulus when the upper edge of the weight is set to the scribed lines; without the weights, the fork will vibrate at 256 Hz (middle C). This tuning fork is manufactured by Riverbank Laboratories, Galena, Illinois 60134 (630-232-2207).

1039

NIS(legs) correlated more strongly with nerve conduction abnormality than did the NIS.57 The NIS(LL) has been and is being extensively used in epidemiology surveys (e.g., the Rochester and Mdewakanton Diabetic Neuropathy Study [NS36797]) and in therapeutic trials of nerve growth factor,2 ␣-lipoic acid (VIATRIS, Inc.), pyridorin (BioStratum, Inc.), and protein kinase C (Eli Lilly, Inc.).57 Kleyweg et al.34 constructed a muscle strength score by adding the MRC scores of defined muscle groups. Shoulder abduction, elbow flexion, wrist extension, hip flexion, knee extension, and ankle dorsiflexion are tested. The muscles are tested on both sides so that the total score is 60. Inter-rater reliability was shown to be high and its responsiveness to change potentially greater than that of the simple GBS disability scale (described below). The scale is simple to use but does not test some muscles that are particularly affected in peripheral neuropathy—for instance, the first dorsal interosseous and extensor hallucis longus—nor does it include reflexes and sensation as done in the NIS. Nevertheless, the score has been used and found to be responsive in a trial of intravenous immune globulin in CIDP.40 However, since MRC scoring is not linear (scores 1, 2, and 3 being too narrow and 4 too wide) statistical comparison of groups (e.g., treatment groups) must be interpreted with caution.

Nerve Conduction Tests as Impairment Measures Chapter 36 outlines how attributes of nerve conduction, quantitative sensation tests (QSTs), and quantitative autonomic tests (QATs, e.g., change in pulse with deep breathing [P-DB]) can be combined to develop composite scores. To summarize, a group of representative attributes of nerve conduction (for the neuropathy studied) are chosen; for example, for diabetic sensorimotor polyneuropathy, five nerve conduction attributes might include peroneal nerve motor nerve conduction velocity, amplitude, and distal latency; tibial nerve motor distal latency; and sural nerve amplitude. To include a broader array of test results, the five nerve conduction attributes listed above could be added to by also including VDT of the great toe and decrease in P-DB. To combine these disparate measures, results must first be expressed in comparable units of measurement. To do this, test results are first expressed as percentiles. Then percentiles are changed so abnormality is always expressed in the upper tail of the normal distribution. Then percentiles are expressed as normal deviates (a value from the standard normal distribution [with a value of mean ⫽ 0 and a variance ⫽ 1], also called a Z score). These normal deviate values are then summed, divided by the number of measured variables, and multiplied by the number of variables in the composite score. As discussed in Chapter 36, these approaches cannot be used without having determined percentile values (e.g., each nerve conduction variable) specific for age, sex,

1040

Quantitating Symptoms, Impairments, and Outcomes

and applicable anthropomorphic characteristics. This means that a large normative study of healthy subjects must previously have been done. In order to have an adequate sample, several hundred persons representative of a population should have been evaluated, and only persons without neurologic disease or diseases known to predispose to neuropathy should be used for estimation of normal values. Then percentiles are estimated for test, nerve, age, sex, and applicable physical characteristics. This procedure should also be followed for autonomic tests and for modalities of QSTs. Composite scores such as that described above have several major advantages over using simple dichotomous judgments of normal or abnormal. They provide a continuous measure of abnormality. They allow a combination of different attributes of nerve conduction tests, QSTs, and QATs—a combination of functions that may be abnormal in polyneuropathy. When they are corrected for changes in physical characteristics (e.g., age and weight), the measures of nerve function should remain the same over time unless nerves are improving or worsening as a result of change in disease or effect of an intervention. Composite nerve conduction scores provide one of the best monotone measures of severity of neuropathy over time (P. J. Dyck et al., unpublished data, 2004). The composite test scores we have been using are the ⌺ 5NC normal deviates (the five nerve conduction attributes listed above) and the ⌺ 7 Tests (the ⌺ 5NC normal deviates ⫹ VDT toe ⫹ P-DB normal deviates). Other combinations of tests can be fashioned. We have also combined measures of clinical signs and test scores (e.g., NIS[LL] ⫹ 7 Tests). By combining points and normal deviates, we assume that 1 point of the NIS(LL) is approximately equal to 1 normal deviate of test results. Composite scores have been used to demonstrate longitudinal worsening in a study of diabetics in Rochester, Minnesota,14 nerve conduction improvement after pancreas and renal transplantation,24 risk factors in a cross-sectional and longitudinal study of diabetic neuropathy,15 and identity of affected persons with gene mutations in a kindred with hereditary motor and sensory neuropathy type 2C.33

Composite Scores Employing Varying Combinations of Symptoms, Impairments, and Test Results Feldman et al. developed a two-step quantitative clinical and electrophysiologic assessment for the diagnosis and staging of diabetic neuropathy. Initially, patients are given a screening instrument (the MNSI; Fig. 40–2).26 Among patients who scored 2 or more points, most had neuropathy as judged by other evaluations. The two-step approach has appeal in that it provides both a screening and definitive evaluation. It should, however, be noted that patients without symptoms and neurologic findings, but with functional

abnormalities of nerve conduction, might go undetected using the 2-stage approach of Feldman et al.—a substantial number in a condition such as diabetic polyneuropathy. Cornblath et al.10 have developed a Total Neuropathy Score, a numeric measure of a combination of symptoms, neurologic signs, QSTs, and specific attributes of nerve conduction. It depends heavily on topographic abnormality (e.g., for symptoms, 0 ⫽ none, 1 ⫽ symptoms of fingers and toes . . . 4 ⫽ symptoms to include knees or elbows). It is recommended for clinical trials in neuropathy and is reported to have favorable inter- and intraobserver reproducibility. However, it was compared to the NSS and NIS in only 5 healthy subjects and 30 patients with neuropathy; more validation is needed.

Electronic Case Report Form of the Clinical Neuropathy Assessment We have developed and use an electronic case report form in aspects of our medical practice to track severity of neuropathy over time, in epidemiologic surveys (e.g., the Rochester and Mdewakanton Diabetic Neuropathy studies14,15) and in single- or multiple-center therapeutic trials.23 The CNA consists of the Lower Limb Function score (LLF), the NIS, and the NSC.23 The LLF consists of the ability (scored 0) or inability (scored 1) to walk on the toes, walk on the heels, and arise from a kneeling position, and items are scored separately for each side. Abnormality is scored only when it is judged to be due to neuropathy. The NIS and NSC are described in previous sections. The CNA is preprinted on a 12-page form and is designed for controlled clinical trials, epidemiologic surveys, and longitudinal patient evaluation. Patient and study information is given in Figure 40–3, an abbreviated portion of the NIS is shown in Figure 40–4, and an abbreviated portion of the NSC is provided in Figures 40–5 and 40–6. Because the forms have been used in several large international controlled trials, the specific questions regarding symptoms are now available in 25 languages (translated by translators and neurologists to convey the same meaning). In a study of the overall performance, the frequency of the problem identified and corrected, and time saved from use of standard paper case report forms, were evaluated in multicenter treatment trials,1,2 single-center epidemiologic surveys,14,24,25 and neurologic patients.15,17 We quote the observation and conclusion from this study: The approach provides: 1) a standard, scannable paper case report form (CNA) for pen entry of neuropathy symptoms, impairments and disability at the bedside or in the office which paper record is retained as a source document at the participating medical center but a facsimile can be transferred instantaneously, its data can be interactively evaluated (by computer programs and quality control personnel), modified, and electronically

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

1041

FIGURE 40–2 The MNSI questionnaire. (From Feldman, E. L., Stevens, M. J., Thomas, P. K., et al.: A practical two-step quantitative clinical and electrophysiological assessment for the diagnosis and staging of diabetic neuropathy. Diabetes Care 17:1281, 1994, with permission.)

stored while maintaining an audit trail; 2) it allowed efficient and accurate reading, transfer, analysis, and storage of data of more than 15,000 forms used in multicenter trials; 3) in 500 consecutive CNA evaluations, software programs identified and facilitated interactive corrections of omissions, discrepancies, and disease and study inconsistencies, introducing only a few readily identified and corrected entry errors; and 4) use of programmed, as compared to non-programmed assessment, was more accurate than double keyboard entry of data and was approximately five times faster.23

DISABILITY SCORES General Scores A number of generic scales have been developed for measuring disability. The most thorough of these is the Functional Independence Measure, which is an 18-item ordinal scale that rates the level of assistance required to perform activities of daily living. It has high intra- and inter-rater reliability, has high internal consistency for the different items, and is moderately responsive to change in

1042

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 40–3 The patient and study information page (page 2) of the CNA. As described in the text, all pages of the CNA have a unique number (in the upper right box), corner symbols for proper paper alignment, and checkerboard symbols (upper left) to identify that the CNA belongs to a specific study (or group of studies). Note that instructions are provided on how the data are to be judged and entered. The figure is shown reduced from 8.5 ⫻ 11 inches. (From Dyck, P. J., Turner, D. W., Davies, J. L., et al.: Electronic case-report forms of symptoms and impairments of peripheral neuropathy. Can. J. Neurol. Sci. 29:258, 2002, with permission.)

multiple sclerosis. It has not been used in treatment trials in peripheral nerve disease and is cumbersome, requiring reference to a 48-page instruction book for its completion.56 A much simpler alternative is the modified Rankin scale (Table 40–4), which has been used as a measure of

activities in trials and observational studies of peripheral neuropathy.3,52 However, it was originally developed for stroke and, in fact, measures handicap as well as disability. In response to the need for a simple disability scale for trials of treatment in GBS, Hughes et al. developed a simple

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

1043

FIGURE 40–4 A composite of several parts of pages 4 and 5 of the CNA (reduced in size) providing information about Neuropathy Impairment Score (NIS). These pages show the layout used and the scoring of weakness, reflexes, and sensation. Test items not shown in detail in this figure are items 1 through 5, 6 through 20, 25 through 27, and 30 through 33. (From Dyck, P. J., Turner, D. W., Davies, J. L., et al.: Electronic case-report forms of symptoms and impairments of peripheral neuropathy. Can. J. Neurol. Sci. 29:258, 2002, with permission.)

7-point GBS disability scale ranging from 0 ⫽ normal through 4 ⫽ unable to walk even with aid, and 5 ⫽ requirement for artificial ventilation to 6 ⫽ dead.31 This scale has been shown to have good intra- and inter-rater reliability and responsiveness to change34 and has been used with minor modifications in almost all subsequent trials of treatment in

GBS.30,32,53 The most recent revision of the scale is shown in Table 40–5. This GBS disability scale has the advantage of simplicity, but the steps are coarse and other aspects of peripheral neuropathy and upper limb disability are not captured. In a partial solution to the problem of identifying a more responsive scale that captures more aspects of disability in

1044

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 40–5 A composite of portions of pages 6, 7, and 8 of the CNA (reduced in size) providing instructions about eliciting neuropathic symptoms and their scoring for the Neuropathy Symptoms and Change score (NSC). Test items 1 through 19 are not provided in detail in this figure. (From Dyck, P. J., Turner, D. W., Davies, J. L., et al.: Electronic case-report forms of symptoms and impairments of peripheral neuropathy. Can. J. Neurol. Sci. 29:258, 2002, with permission.)

peripheral nerve disease, the European Inflammatory Neuropathy Cause and Treatment Group developed the inflammatory neuropathy Overall Disability Score Scale (ODSS), summing upper and lower limb disability, for use in multicenter treatment trials (Table 40–6).45 This scale was derived from an even more comprehensive disability scale designed and validated for use in multiple sclerosis trials.55

In patients with GBS and CIDP, the ODSS scale has good construct validity, high intra- and inter-rater reliability, and good responsiveness. Impairment measures accounted for a higher proportion of the variance of the ODSS compared with the GBS disability score and Rankin score (r ⫽ .64 versus .56 and .45, respectively). The scale is simple and has been successfully used as the primary outcome measure in

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

1045

FIGURE 40–6 A composite of portions of pages 9 through 12 of the CNA (reduced in size) providing a representative sample of questions from the Neuropathy Symptoms and Change score (NSC). Test items 23 through 29 and 32 through 38 are not provided in the figure. (From Dyck, P. J., Turner, D. W., Davies, J. L., et al.: Electronic case-report forms of symptoms and impairments of peripheral neuropathy. Can. J. Neurol. Sci. 29:258, 2002, with permission.)

trials of immunotherapy in CIDP and paraproteinemic demyelinating neuropathy.9,29

Health-Related Quality of Life and Clinical Outcomes Score in Peripheral Neuropathy The Health-Related Quality of Life and Clinical Outcomes Score in Peripheral Neuropathy (Dyck Score) is an elabo-

ration of the Rankin score to include nonsymptomatic abnormality and to provide more definition for grades of severity (Table 40–7). Symptoms, signs, and test results are considered to be abnormal when they can be attributed to only the disease (the peripheral neuropathy) being evaluated and are not due to age, sex, applicable physical variables or physical unfitness, or another disease. When normative data are

1046

Quantitating Symptoms, Impairments, and Outcomes

Table 40–4. Modified Rankin Scale: A Generic Disability Scale 0 1 2

3 4

5

No symptoms Minor symptoms that do not interfere with lifestyle Minor handicap; symptoms that lead to some restriction in lifestyle but do not interfere with the patient’s capacity to look after himself Moderate handicap; symptoms that restrict lifestyle significantly and prevent totally independent existence Moderately severe handicap; symptoms that restrict lifestyle significantly although not needing constant attention Severe handicap; totally dependent patient who requires constant attention day and night

Modified from Rankin, J.: Cerebral vascular accidents in patients over the age of 60. II. Prognosis. Scott. Med. J. 2:200, 1957.

available for individual tests (e.g., an attribute of nerve conduction), the level of abnormality is the 2.5th or 97.5th percentile. The level of abnormality for composite scores (e.g., the NIS or NIS[LL] ⫹ 7 Tests) is less than or equal to the 2.5th or greater than or equal to the 97.5th percentile—or other levels of abnormality depending on the purpose of the study. Symptoms may be negative (decreased or absent function, e.g., decreased ability to feel tactile, thermal, or pain stimuli) or positive (increased or abnormal function, e.g., asleep-numbness, prickling, or pain of various kinds). The physician arrives at the score after assessing a patient’s symptoms, neuropathic impairments, and test results and reviewing the patient’s ability to perform activities of daily living. Only disability related to the peripheral neuropathy being studied is graded. In deciding whether the patient has difficulty with or inability to perform certain tasks or activities of daily living, more than the patient’s report should be used for judgment; the physician should use objective criteria. Feigned (for medicolegal gain) or psychiatric dysfunction should not be scored if the purpose is to estimate the

Table 40–5. Guillain-Barré Syndrome Disability Scale 0 ⫽ healthy, no signs or symptoms of Guillain-Barré syndrome 1 ⫽ minor symptoms or signs and able to run 2 ⫽ able to walk 5 m across an open space without assistance 3 ⫽ able to walk 5 m across an open space with the help of one person and waist-level walking frame, stick, or sticks 4 ⫽ chairbound/bedbound: unable to walk as in 3 5 ⫽ requiring assisted ventilation (for at least part of day or night) 6 ⫽ dead Modified from Hughes, R. A. C., Newsom-Davis, J. M., Perkin, G. D., and Pierce, J. M.: Controlled trial of prednisolone in acute polyneuropathy. Lancet 2:750, 1978.

disability from neuropathy alone. Dysfunction should be matched by an approximately similar abnormality in the neurologic examination or in test results. The dysfunction should be reproducible and common to the activities of the involved anatomic structures (e.g., when handwriting is affected, piano playing should also be affected).

Other Outcomes In some trials, a pharmaceutical house may wish to investigate prevention or amelioration of specific neuropathy outcomes. An outcome of relevance for small fiber sensory neuropathies and some patients with diabetic polyneuropathy is the severity of positive neuropathy sensory symptoms such as asleep-numbness, prickling or tingling, burning, deep aching, lancinating pain, and thermal or mechanical hyperalgesia. This was the primary end point for a prospective, randomized, double-blind trial of intravenous ␣-lipoic acid versus saline (the Sydney Study). The treatment resulted in a large and statistically significant improvement of symptoms.1 Other studies might focus on the number or frequency of degenerating or regenerating nerve fibers of biopsied sural nerve obtained from one side at onset of the study and from the other side at the end of the study, which have been used as end points in diabetic polyneuropathy. Finally the development of tissue complications (e.g., plantar ulcers or Charcot joints) has been used as an end point for other trials. Counts of nerve endings in skin might be used as an indicator of neuropathy (see Chapter 34).

Morphometry of Nerve or Nerve Fiber Counts in Skin Biopsy Counts of sural nerve myelinated fibers per nerve or per square millimeter of fascicular area, taken from one side at onset of the study and from the other side at the end of the study, have been used as measures of neuropathy in trials of diabetic treatment. Nerve fiber counts are objective and reliable measures of severity of neuropathy. However, they do not provide a direct measure of the severity of the symptoms and impairments of the neuropathy patient. Also, counted fibers may not be functional—their terminals, or a percentage of them, may have degenerated. A consensus group did not support their use in therapeutic trials of diabetic polyneuropathy.1 Punch biopsy of the skin so as to count epidermal nerve fibers may be considered for controlled trials of therapy in some neuropathies (see Chapter 34). Before using nerve fiber counts for this purpose, more comprehensive normative values are needed (especially values extending into old age), and the influence of site, age, sex, and physical characteristics on nerve ending counts should be known better. It would also be useful to know whether these measures correlate with clinical symptomatology.

1047

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

Table 40–6. Inflammatory Neuropathy Overall Disability Sum Score (17) Arm Disability Scale—Function Checklist Dressing upper part of body (excluding buttons/zips) Washing and brushing hair Turning a key in a lock Using knife and fork (/spoon—applicable if the patient never used knife and fork) Doing/undoing buttons and zips

Not affected

Affected but not prevented

Prevented

䊊 䊊 䊊 䊊

䊊 䊊 䊊 䊊

䊊 䊊 䊊 䊊

䊊

䊊

䊊

Arm Grade 0 ⫽ Normal 1 ⫽ Minor symptoms or signs in one or both arms but not affecting any of the functions listed 2 ⫽ Moderate symptoms or signs in one or both arms affecting but not preventing any of the functions listed 3 ⫽ Severe symptoms or signs in one or both arms preventing at least one but not all functions listed 4 ⫽ Severe symptoms or signs in both arms preventing all functions listed but some purposeful movements still possible 5 ⫽ Severe symptoms and signs in both arms preventing all purposeful movements Leg Disability Scale—Function Checklist

No

Yes

Not applicable

Do you have any problem with your walking? Do you use a walking aid? How do you usually get around for about 10 meters? Without aid With one stick or crutch or holding to someone’s arm With two sticks or crutches or one stick or crutch and holding to someone’s arm With a wheelchair If you use a wheelchair, can you stand and walk a few steps with help? If you are restricted to bed most of the time, are you able to make some purposeful movements?

䊊 䊊

䊊 䊊

䊊 䊊

䊊 䊊 䊊

䊊 䊊 䊊

䊊 䊊 䊊

䊊 䊊 䊊

䊊 䊊 䊊

䊊 䊊 䊊

Leg grade 0 ⫽ Walking is not affected 1 ⫽ Walking is affected but does not look abnormal 2 ⫽ Walks independently but gait looks abnormal 3 ⫽ Usually uses unilateral support to walk 10 meters (25 feet) (stick, single crutch, one arm) 4 ⫽ Usually uses bilateral support to walk 10 meters (25 feet) (sticks, crutches, two arms) 5 ⫽ Usually uses wheelchair to travel 10 meters (25 feet) 6 ⫽ Restricted to wheelchair, unable to stand and walk few steps with help but able to make some purposeful leg movements 7 ⫽ Restricted to wheelchair or bed most of the day, preventing all purposeful movements of the legs (e.g., unable to reposition legs in bed) Overall disability sum score ⫽ arm disability scale (range 0–5) ⫹ leg disability scale (range 0–7); overall range: 0 (no signs of disability) to 12 (maximum disability). For the arm disability scale: allocate one arm grade only by completing the function checklist. Indicate whether each function is “affected,” “affected but not prevented,” or “prevented.” For the leg disability scale: Allocate one leg grade only by completing the functional questions. Reprinted from Merkies, I. S., Schmitz, P. I., van der Meche, F. G., et al., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Clinimetric evaluation of a new overall disability scale in immune mediated polyneuropathies. J. Neurol. Neurosurg. Psychiatry 72:596, 2002, with permission.

Fatigue Fatigue is a disabling symptom that makes an important contribution to the ill health of many patients with peripheral neuropathy. Eighty percent of 113 patients with inflammatory neuropathies were reported to have severe

fatigue on a generic fatigue severity scale, including even patients with normal health and sensation.41 The fatigue severity scale used in this study was a 9-item questionnaire with scores ranging from no fatigue (0) to most disabling fatigue (7).35 Alternative scales have been used in other

1048

Quantitating Symptoms, Impairments, and Outcomes

Table 40–7. Health-Related Quality of Life and Clinical Outcomes Score in Peripheral Neuropathy (The Dyck Score) 0 1

2 3

4

5

6 7 8

No neuropathy: no symptoms (NSC number [or NSS] ⬍ 1), signs (NIS ⬍ 2 points), or test (e.g., 7 Tests ⬍ 97.5th percentile) abnormalities of neuropathy. Minimal neuropathy: only one of a, b, or c are abnormal a) Tests are the only abnormality (e.g., 7 Tests ⱖ 97.5th percentile), or b) Neuropathy signs are the only abnormality (e.g., NIS ⱖ 2 points), or c) Neuropathy symptoms are the only abnormality (e.g., NSC number or NSS ⱖ 1) Minimal neuropathy: 1a ⫹ lb. Symptomatic neuropathy: 1a, 1b, and neuropathy symptoms*; 1a and neuropathy symptoms; or 1b and neuropathy symptoms. The patient is able to continue with usual† activities of daily living, work, or recreational activity, and can meet usual family and social responsibilities. Symptomatic neuropathy (as defined in 3) interfering with and limiting work, usual activities of daily living, recreational activity, or family and social obligations, but independent function is possible without the help of others. At this score, there is an unequivocal limitation‡ of usual work, activities of daily living, or recreational, family, or social obligations because of neuropathy. Symptomatic neuropathy (defined in 3) restricting activities of daily living, work, and recreational activities. The help of other caregivers§ (⬍21/2 hr/day) is needed. He or she is unable to perform specific activities of daily living. If use of a wheelchair is mandatory for activities of daily living, recreational activities, or social and family responsibilities, the patient would usually fall into this or a higher score (ⱖ5). Symptomatic neuropathy (defined in 3) requiring the help of caregivers ⱖ 21/2 hr but ⬍ 8 hr/day as described in 5. Symptomatic neuropathy (defined in 3) requiring the help of a caregiver ⱖ 8 hr/day but not continuously as in stage 8. Symptomatic neuropathy (defined in 3) requiring constant care in an intensive care unit.

*Symptoms of neuropathy: NSS ⱖ 1 or NSC number ⱖ 1; symptoms of muscle weakness, atrophy, or cramps; negative or positive neuropathic sensory symptoms (N-NSS, P-NSS); or neuropathic autonomic symptoms. † Usual activities of daily living, work, recreational, and social and family activities: despite neuropathic symptoms, the patient is able to work at his or her usual activity, maintain his or her usual home obligations and participate in recreational activities. Generally, the patient can carry on despite some minor motor, sensory or autonomic symptoms. The patient may be unable to perform extraordinary activities (e.g., competitive sports, feats of endurance, etc.). ‡ The motor, sensory, or autonomic symptoms or impairment is of a sufficient degree to limit the ability to work, perform usual activities of daily living, or recreational activities, or meet family and usual social responsibilities. The use of a cane or orthotic device probably places the patient into this (or a higher) category unless the patient is able to perform “usual” activities of daily living and recreational activities and meet social and family responsibilities (then he or she would fall into lower category). § A member of the family or visiting nurse is needed to provide activities of daily living (bathing, shaving, tooth brushing, feeding, writing, etc.), daily management of analgesics or opiates, or help with management of autonomic dysfunction that the patient is not able to perform adequately or safely on his or her own.

diseases, and the responsiveness of these scales has not been compared.7 In a study of inflammatory neuropathy, the fatigue severity scale used had good internal consistency (similar scores with each of the items) and intra- and inter-rater reliability. It correlated with the patients’ assessment of their quality of life measured with the Medical Outcome Study 36-item short form health status scale (SF-36) but not with measures of impairment of strength or sensation or even with the ODSS measure of disability.41 However, it is also important to note that fatigue is common and may be severe even in patients who do not have neuropathy.

HANDICAP Measurement of handicap has been surprisingly neglected in clinical trials of interventions in peripheral neuropathy despite its importance to patients, representing the difficulties that their disease imposes on their participation in society. Recently, Merkies et al.44 developed the Rotterdam handicap scale, which scores the extent to which patients are

impaired in the following nine domains: mobility indoors; mobility outdoors; kitchen tasks; indoor domestic tasks; outdoor domestic tasks; indoor leisure activities; outdoor leisure activities; ability to drive a car, go by bus, or bicycle; and work or study. Each domain is scored according to the following principles: 0 ⫽ not applicable; 1 ⫽ unable to undertake the activity; 2 ⫽ able to undertake the activity to a minimal extent, usually needing help; 3 ⫽ able to undertake the activity to a moderate extent, sometimes needing help; and 4 ⫽ not affected. The reader is referred to the original publication for the precise description of this simple but potentially useful scale. Merkies et al.46 showed that the scale had good face validity and inter-rater reliability in 113 stable patients with inflammatory neuropathy and high responsiveness to change in a longitudinally studied cohort of 20 patients with GBS or CIDP. By focusing on the problems encountered by patients with peripheral neuropathy and avoiding those caused by cognitive impairment, the Rotterdam handicap scale commends itself as a measure of participation over previous scales that were developed for use in other conditions. Longitudinal studies of patients with different types of

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes

peripheral neuropathy show significant persistent interference with participation in daily life. For instance, in a study of 122 patients 3 to 6 years after diagnosis of GBS, 38% were still handicapped with respect to their work, 37% in domestic activities, and 44% in leisure activities.4

QUALITY OF LIFE The subject of quality of life is covered in greater detail in Chapter 41. The generic instrument most commonly used for measuring quality of life is the SF-36. Merkies et al. have demonstrated the applicability of this scale to inflammatory neuropathy.46 They showed that the SF-36 scores of 113 stable patients with GBS, CIDP, or paraproteinemic demyelinating neuropathy were lower than those of 1742 members of the Dutch population. Those with more severe disability had even lower scores than those with less disability in the “physical functioning” and “role physical” domains. Impairment of sensation measured with the Inflammatory Neuropathy Sensory Sum Score and impairment of strength measured with the MRC sum score only partly explained the Physical Component Summary score results of the SF-36. The SF-36 showed high responsiveness to change in 20 patients being treated for GBS or CIDP. Unsurprisingly, the domains that measured physical functioning were more responsive than other domains, although the mental component summary was as responsive as the physical component summary, reflecting the impact that inflammatory neuropathy may have on mental health.

REFERENCES 1. Ametov, A. S., Barinov, A., Dyck, P. J., et al.: The sensory symptoms of diabetic polyneuropathy are improved with alpha-lipoic acid. The Sydney Trial. Diabetes Care 26:770, 2003. 2. Apfel, S. C., Schwartz, S., Adornato, B. T., et al.: Efficacy and safety of recombinant human nerve growth factor in patients with diabetic polyneuropathy: a randomized controlled trial. JAMA 284:2215, 2000. 3. Bamford, J. M., Sandercock, P. A., Warlow, C. P., and Slattery, J.: Interobserver agreement for the assessment of handicap in stroke patients. Stroke 20:828, 1989. 4. Bernsen, A. J. A. M., De Jager, A. E. J., Schmitz, P. I. M., and van der Meche, F. G. A.: Long-term impact on work and private life after Guillain-Barre syndrome. J. Neurol. Sci. 201:13, 2002. 5. Burns, T. M., Taly, A., O’Brien, P. C., and Dyck, P. J.: Clinical versus quantitative vibration assessment: improving clinical performance. J. Peripher. Nerv. Syst. 7:112, 2002. 6. Capildeo, R., and Orgogozo, J.-M.: Methods in clinical trials in neurology. In Vascular and Degenerative Brain Disease. London, Macmillan Press, p. 327, 1988.

1049

7. Chalder, T., Berelowitz, G., Pawlikowska, T., et al.: Development of a fatigue scale. J. Psychosom. Res. 37:147, 1993. 8. Cocchiarella, L., and Andersson, G. B. J.: Guides to the Evaluation of Permanent Impairment. Chicago, American Medical Association Press, p. 613, 2001. 9. Comi, G., Swan, A. V., and Hughes, R. A. C., for the INCAT Group: Randomized controlled trial of IVIg in paraproteinaemic demyelinating neuropathy. J. Neurol. Sci. 187(Suppl. 1):S120, 2001. 10. Cornblath, D. R., Chaudhry, V., Carter, K., et al.: Total neuropathy score: validation and reliability study. Neurology 53:1660, 1999. 11. Day, S.: Dictionary for Clinical Trials. Chichester, UK, John Wiley & Sons, p. 217, 1999. 12. Dyck, P. J.: Neurophysiological evaluation of restorative surgery of nerve and muscle. Invited contribution to the 6th International Congress of Electromyography in Stockholm, Sweden, Symposia 85–86, June 17–20, 1979. 13. Dyck, P. J., Daube, J., O’Brien, P., et al.: Plasma exchange in chronic inflammatory demyelinating polyradiculoneuropathy. N. Engl. J. Med. 314:461, 1986. 14. Dyck, P. J., Davies, J. L., Litchy, W. J., and O’Brien, P. C.: Longitudinal assessment of diabetic polyneuropathy using a composite score in the Rochester Diabetic Neuropathy Study cohort. Neurology 49:229, 1997. 15. Dyck, P. J., Davies, J. L., Wilson, D. M., et al.: Risk factors for severity of diabetic polyneuropathy: intensive longitudinal assessment of the Rochester Diabetic Neuropathy Study cohort. Diabetes Care 22:1479, 1999. 16. Dyck, P. J., Karnes, J., O’Brien, P. C., and Swanson, C. J.: Neuropathy symptom profile in health, motor neuron disease, diabetic neuropathy, and amyloidosis. Neurology 36:1300, 1986. 17. Dyck, P. J., Kratz, K. M., Lehman, K. A., et al.: The Rochester Diabetic Neuropathy Study: design, criteria for types of neuropathy, selection bias, and reproducibility of neuropathic tests. Neurology 41:799, 1991. 18. Dyck, P. J., Litchy, W. J., Kratz, K. M., et al.: A plasma exchange versus immune globulin infusion trial in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 36:838, 1994. 19. Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects (RDNS-HS). Neurology 45:1115, 1995. 20. Dyck, P. J., Low, P. A., Windebank, A. J., et al.: Plasma exchange in polyneuropathy associated with monoclonal gammopathy of undetermined significance. N. Engl. J. Med. 325:1482, 1991. 21. Dyck, P. J., O’Brien, P. C., Oviatt, K. F., et al.: Prednisone improves chronic inflammatory demyelinating polyradiculoneuropathy more than no treatment. Ann. Neurol. 11:136, 1982. 22. Dyck, P. J., Sherman, W. R., Hallcher, L. M., et al.: Human diabetic endoneurial sorbitol, fructose, and myo-inositol related to sural nerve morphometry. Ann. Neurol. 8:590, 1980. 23. Dyck, P. J., Turner, D. W., Davies, J. L., et al., for the rhNGF Group: Electronic case-report forms of symptoms and

1050

24.

25.

26.

27.

28. 29.

30.

31.

32.

33.

34.

35.

36. 37.

38.

39.

Quantitating Symptoms, Impairments, and Outcomes

impairments of peripheral neuropathy. Can. J. Neurol. Sci. 29:258, 2002. Dyck, P. J., Velosa, J. A., Pach, J. M., et al.: Increased weakness after pancreas and kidney transplantation. Transplantation 72:1403, 2001. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Methylprednisolone may improve lumbosacral radiculoplexus neuropathy. Can. J. Neurol. Sci. 28:224, 2001. Feldman, E. L., Stevens, M. J., Thomas, P. K., et al.: A practical two-step quantitative clinical and electrophysiological assessment for the diagnosis and staging of diabetic neuropathy. Diabetes Care 17:1281, 1994. The Guarantors of Brain: Aids to the Examination of the Peripheral Nervous System. Philadelphia, W. B. Saunders, p. 62, 2001. Herndon, R. M.: Handbook of Neurologic Rating Scales. New York, Demos Vermande, p. 266, 1997. Hughes, R., Bensa, S., Willison, H., et al., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Randomized controlled trial of intravenous immunoglobulin versus oral prednisolone in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:195, 2001. Hughes, R. A., Raphael, J. C., Swan, A. V., and van Doorn, P. A.: Intravenous immunoglobulin for Guillain-Barre syndrome (Cochrane Review). In The Cochrane Library, Issue 4. Chichester, UK, John Wiley and Sons Ltd., 2003. Hughes, R. A. C., Newsom-Davis, J. M., Perkin, G. D., and Pierce, J. M.: Controlled trial of prednisolone in acute polyneuropathy. Lancet 2:750, 1978. Hughes, R. A. C., and van der Meche, F. G. A.: Corticosteroid treatment for Guillain-Barre syndrome (Cochrane Review). In The Cochrane Library, Issue 4. Chichester, UK, John Wiley and Sons Ltd., 1999. Klein, C. J., Cunningham, J. M., Atkinson, E. J., et al.: The gene for HMSN 2C maps to 12q23–24: a region of neuromuscular disorders. Neurology 60:1151, 2003. Kleyweg, R. P., van der Meche, F. G., and Schmitz, P. I.: Interobserver agreement in the assessment of muscle strength and functional abilities in Guillain-Barre syndrome. Muscle Nerve 14:1103, 1991. Krupp, L. B., LaRocca, N. G., Muir-Nash, J., and Steinberg, A. D.: The fatigue severity scale: application to patients with multiple sclerosis and systemic lupus erythematosus. Arch. Neurol. 46:1121, 1989. Lovett, R. W.: The Treatment of Infantile Paralysis. Philadelphia: P. Blakiston’s Son & Co., 1917. Martina, I. S., van Koningsveld, R., Schmitz, P. I., et al., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Measuring vibration threshold with a graduated tuning fork in normal aging and in patients with polyneuropathy. J. Neurol. Neurosurg. Psychiatry 65:743, 1998. Medical Research Council: Aids to the Investigation of Peripheral Nerve Injuries (War Memorandum No. 7). Her Majesty’s Stationery Office. London, The White Rose Press, 1943. Members of the Sections of Neurology and Section of Physiology, Mayo Clinic and Mayo Foundation: Clinical Examinations in Neurology. Philadelphia, W. B. Saunders, p. 370, 1956.

40. Mendell, J. R., Barohn, R. J., Freimer, M. L., et al.: Randomized controlled trial of IVIg in untreated chronic inflammatory demyelinating polyradiculoneuropathy. Neurology 56:445, 2001. 41. Merkies, I. S., Schmitz, P. I., Samijn, J. P., et al., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Fatigue in immune-mediated polyneuropathies. Neurology 53:1648, 1999. 42. Merkies, I. S., Schmitz, P. I., Samijn, J. P., et al., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Assessing grip strength in healthy individuals and patients with immune-mediated polyneuropathies. Muscle Nerve 23:1393, 2000. 43. Merkies, I. S., Schmitz, P. I., van der Meche, F. G., and van Doorn, P. A., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Psychometric evaluation of a new sensory scale in immune-mediated polyneuropathies. Neurology 54:943, 2000. 44. Merkies, I. S., Schmitz, P. I., van der Meche, F. G., et al.: Psychometric evaluation of a new handicap scale in immune-mediated polyneuropathies. Muscle Nerve 25:370, 2002. 45. Merkies, I. S., Schmitz, P. I., van der Meche, F. G., et al., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Clinimetric evaluation of a new overall disability scale in immune mediated polyneuropathies. J. Neurol. Neurosurg. Psychiatry 72:596, 2002. 46. Merkies, I. S., Schmitz, P. I., van der Meche, F. G., et al., for the European Inflammatory Neuropathy Cause and Treatment (INCAT) Group: Quality of life complements traditional outcome measures in immune-mediated polyneuropathies. Neurology 59:84, 2002. 47. Mitchell, S. W.: Injuries of Nerves and Their Consequences. Philadelphia, J. B. Lippincott, p. 377, 1872. 48. Munstat, T. L.: Quantification of Neurologic Deficit. Boston, Butterworth, p. 357, 1989. 49. Office of Disability: Disability Evaluation under Social Security. Baltimore, Social Security Administration, p. 102, 1986. 50. Porter, R. J., and Schoenberg, B. S.: Controlled Clinical Trials in Neurological Disease. Boston, Kluwer Academic Publishers, p. 440, 1990. 51. Potvin, A. R., Tourtellotte, W. W., Potvin, J. H., et al.: Quantitative Examination of Neurologic Functions. Boca Raton, FL, CRC Press, p. 205, 1985. 52. Rankin, J.: Cerebral vascular accidents in patients over the age of 60. 2. Prognosis. Scott. Med. J. 2:200, 1957. 53. Raphael, J.-C., Chevret, S., Hughes, R. A. C., and Annane, D.: Plasma exchange for Guillain-Barre syndrome (Cochrane Review). In The Cochrane Library, Issue 4. Chichester, UK, John Wiley and Sons Ltd., 2001. 54. Rydel, A., and Seiffer, W.: Untersuchungen uber das vibrationsqefuhl oder die sog: “Knochensensibilitat” (Pallasthesie). Arch. Psychiatr. Nervenkrankenh. 37:488, 1903. 55. Sharrack, B., and Hughes, R. A.: The Guy’s Neurological Disability Scale (GNDS): a new disability measure for multiple sclerosis. Mult. Scler. 5:223, 1999. 56. Sharrack, B., Hughes, R. A., Soudain, S., and Dunn, G.: The psychometric properties of clinical rating scales used in multiple sclerosis. Brain 122:141, 1999.

Quantitating Overall Neuropathic Symptoms, Impairments, and Outcomes 57. Skljarevski, V., Price, K. L., Bastyr, E. J., et al.: Limited neurological examination of the legs correlates with composite measures of nerve conduction in patients with diabetic peripheral neuropathy. Eur. J. Neurol. 10(Suppl. 1):122, 2003. 58. World Health Organization: Classification of Impairments: International Classification of Impairments, Disabilities and Handicaps. Geneva, World Health Organization, p. 47, 1980.

1051

59. World Health Organization: ICIDH-2: International Classification of Impairments, Activities, and Participation. Geneva, World Health Organization, p. 1, 1997. 60. Ziegler, D., Hanefeld, M., Ruhnau, K. J., et al.: Treatment of symptomatic diabetic peripheral neuropathy with the antioxidant alpha-lipoic acid: a 3-week multicenter randomized controlled trial (ALADIN Study). Diabetologia 38:1425, 1995.

41 Health Outcomes and Quality of Life DEBORAH BUCK AND ANN JACOBY

Health Outcomes of Impairment, Disability, and Handicap The Concept of Quality of Life Conceptualization of Quality of Life Dimensions of Quality of Life Who Should Measure Quality of Life? Asking the Relatives How Should Quality of Life Be Measured?

Generic And Specific Measures Mode of Administration Requirements of Health Outcome Measures Reliability Validity Responsiveness Acceptability

This chapter is concerned with the concepts of impairment, disability, handicap, and quality of life (QOL) and with the importance of including QOL as a measure of outcome both in clinical practice and in research. A summary of the ways in which QOL has been defined and the dimensions of which it is thought to consist is offered. Also considered is the debate surrounding the issue of who should actually be assessing QOL. Following these theoretical concerns, a brief outline of more practical issues surrounding QOL measurement is provided. This covers use of generic and condition-specific instruments, choices about mode of administration, and psychometric and other prerequisites of QOL measures. A number of conditionspecific and generic instruments that have been developed or used to assess QOL in conditions associated with peripheral neuropathy (PN) are described, and some key QOL issues pertinent to these populations are outlined. It is increasingly recognized that doctors’ biomedical evaluations of ill health need to be complemented by subjective accounts of patients themselves through patientbased assessments. This is particularly so in chronic conditions for which there is no cure, such as human immunodeficiency virus (HIV) infection, diabetes, or rheumatoid arthritis (RA), and where therapeutic goals are to relieve symptoms and improve function and QOL.4,5,26,60 In such cases, the use of objective assessments of health alone is insufficient. The advantages and disadvantages of

Assessment of Quality of Life in Diseases Associated with Peripheral Neuropathy Disease-Specific Measures Generic Measures Key Quality-of-Life Issues in Peripheral Neuropathy Conclusion

interventions must be weighed to determine the net benefit from the patients’ point of view, for example, in terms of disability, handicap, and QOL. Although medical interventions may be beneficial to patients in terms of their impact on impairment, the lack of patient-based measurements to assess their wider impacts will preclude a clear evaluation of their efficacy.28 Furthermore, and of much relevance in an era of cost containment, a thorough examination of an intervention’s effect on outcomes such as QOL and overall well-being is a key component in the evaluation of both effectiveness and cost-effectiveness.45 The increasing emphasis on “humanizing” health care, and the demand for informed choice for patients, renders the evaluation of outcomes such as QOL vital regardless of whether life is prolonged or not.8,48 Failure to measure disease and treatment in terms of quality, rather than simply by quantity of survival or by medical outcome alone, has been described by Fallowfield28 as “neither good science nor good medicine” (p. 29). Apart from their use in clinical trials and determining the impact of medical interventions, QOL measurements are also important in describing the nature and extent of functional and psychosocial problems experienced by patients, determining norms for psychosocial morbidity for specific conditions, tailoring management to the needs of the patient, screening individuals for possible psychosocial interventions, and improving clinician-patient interaction.1 1053

1054

Quantitating Symptoms, Impairments, and Outcomes

Patients increasingly require information about the pros and cons of treatments; QOL data present a method of translating medical outcomes into a format that is more meaningful to them.32 This is particularly important in conditions for which there is no cure. The following sections outline a range of health outcomes, including QOL, as they relate to PN.

HEALTH OUTCOMES OF IMPAIRMENT, DISABILITY, AND HANDICAP The World Health Organization’s International Classification of Impairments, Disabilities and Handicaps (ICIDH) was published in 1980.73 It illustrates how disease can engender impairment, defined as a loss or abnormality of psychological, physiologic, or anatomic structure or function. Impairments are “objective” and include a range of states such as hemianopia, sensory loss, problems with speech and thought processes, pain sensation loss, and muscle weakness.67,72 Impairment in turn may lead to disability, defined as a restriction of or inability to carry out an activity in a way considered normal for a human in his or her particular environment. Disabilities can include problems with mobility, personal care, activities of daily living (ADLs), continence, interacting with others, and work-related skills.67,72 The implications of chronic illnesses go beyond their effects on daily activities and personal care, however. Physical disability can only partially explain the impaired QOL experienced by people with chronic conditions. Therefore, it is important to also consider the handicap that can ensue. Handicap may arise either directly as a result of impairments or as a result of disability after interaction with the physical or social environment. Handicap is defined as a disadvantage that restricts or prevents the performance of a role deemed normal for the individual. The nature of the disadvantage is shaped by particular cultural factors and personal characteristics. In general, individuals will be at a disadvantage if they are unable to orient themselves, move around, manage the body’s physical needs, engage themselves in work or leisure, or maintain social and sexual relationships or financial independence. The disabilities and handicaps that can occur as a consequence of impairments resulting from PN are readily identifiable. Persons with, for example, diabetic peripheral neuropathy (DPN) or RA are sometimes restricted in terms of mobility or everyday functional activities. Physical, psychological, environmental, or economic issues may contribute to these difficulties. As a result of these limitations, many people are restricted in their participation in society, either because of tangible impairments or because of the label attached to their condition.68 Beyond the disease-handicap continuum is said to lie the concept of QOL.67 Although the ICIDH offers an important theoretical perspective for health outcome, it

neglects wider QOL issues. The ICIDH has recently been revised (ICIDH-2) to include dimensions for “activities” (covering the complete range of activities performed by an individual) and “participation” (areas of life in which an individual is involved, has access to, and/or for which there are societal opportunities or barriers).74,75 Nonetheless, the wider implications of chronic illness may not be fully addressed by the revised ICIDH. While handicap is considered the most relevant clinical outcome for patients, and impairment the least relevant, QOL is potentially even more pertinent from the patients’ point of view.56 In particular, clinicians involved with conditions and injuries that cannot be cured, but that can be treated in ways that have an impact on patients’ QOL, need to be conscious of the value of assessing such things as QOL.56 Growing interest in this area has resulted in an ever-increasing array of conceptualizations of QOL. The next section provides a detailed overview of this elusive but crucial concept.

THE CONCEPT OF QUALITY OF LIFE Quality of life is an often used but usually ill-defined term, and for good reason. It has been the subject of attention across a range of disciplines, including medicine, psychology, sociology, philosophy, economics, and geography,12,29 leading to an array of definitions. QOL has been described as “the missing measurement in health care,”28 although in recent times it has become a fashionable term in the medical literature. However, clinicians and researchers often fail to provide clear conceptual definitions of QOL. In a critical appraisal of the quality of 75 articles purporting to examine QOL in medicine, Gill and Feinstein38 found that conceptual definitions were provided in only 15% of cases. Even where definitions are provided, there is little consistency between studies,6 and the concept itself has been described as “slippery, burning, neglected, and crucial.”49

Conceptualization of Quality of Life A useful model for defining QOL has been provided by Cramer and Spilker.21 At the upper level is an overall assessment of well-being, which can include life satisfaction and general perceptions of well-being. The middle level incorporates broad dimensions (i.e., physical, psychological, economic, social) and takes account of the widely held view that QOL comprises several key domains. The lower level consists of the individual components of each domain and allows for the variations in their content. For example, components of a social domain may be different in a measure developed to assess QOL in children with asthma compared with one developed for adults with acquired immunodeficiency syndrome (AIDS). This basic model is helpful in that it encompasses several perspectives and can be applied to a range of contexts.

1055

Health Outcomes and Quality of Life

Dimensions of Quality of Life Most researchers in the field maintain that QOL is a multidimensional concept comprising at least three broad dimensions: physical, mental, and social well-being. However, although the majority of people accept that the concept of QOL exists, and that it does so within three broad dimensions, there is no real agreement as to what it comprises beyond these three dimensions.8,28 Chung et al.20 noted that at least 800 possible dimensions have been identified! Specific issues thought to be related to QOL have varied but include psychological and emotional factors (such as anxiety, distress, depression, fear, selfesteem, self-image, happiness, and satisfaction); social well-being (such as the ability to continue usual social roles, social integration, social contact, social support, interpersonal relationships, and intimacy); spirituality46; cognitive functioning (confusion, memory, concentration, difficulty in carrying out other intellectual activities); functional status (mobility, self-care, ADLs, energy); and somatic sensations or symptoms (such as unpleasant physical feelings, pain, or discomfort). In some cases, financial aspects of QOL may be important. General health is also often cited as being one of the key dimensions of QOL.

WHO SHOULD MEASURE QUALITY OF LIFE? In attempting to assess QOL, it is important to consider who should provide the information. Some believe that clinicians are in the best position to provide details about a patient because they can do so objectively. Indeed, the first QOL measures were intended to be physician completed.45 This was because clinicians were, and in some cases remain, distrustful of patient-based assessments. These are seen as yielding “soft” data and therefore viewed as unreliable. However, there is growing acknowledgment that individuals are the best source of information about their own QOL.4,32,48 It has been suggested that subjectivity is a pivotal issue in assessing QOL and that it may be impossible for one person to assess the qualitative aspects of another’s life.8 In particular, health care professionals have been shown to consistently underestimate the health status, psychological well-being, and QOL of their patients. For example, Owens et al.57 found that physicians rated health states in conditions such as AIDS as being associated with a poorer QOL than did people who actually had the condition. Thus doctors may recommend treatment that patients do not desire. Similarly, Wilson et al.71 reported that physicians systematically underestimated overall QOL, social functioning, role functioning, and pain in people with cancer. In a study that compared disability assessments made by clinicians with those made by patients with multiple

sclerosis, Rothwell et al.61 found good agreement between them in terms of physical disability. However, patients were significantly less concerned than clinicians about physical well-being and were significantly more concerned with emotional and mental well-being. Although the importance of patients’ perceptions is becoming increasingly acknowledged, not every patient will be able to report these because of communication problems or some other incapacity.4,32 Obtaining information from people with chronic conditions will sometimes be difficult because of physical, cognitive, or linguistic problems that limit their ability to complete a questionnaire or affect concentration, understanding, or communication. Therefore, the use of “proxy” informants (asking a relative or close friend to answer questions about QOL as he or she believes the patient would) is often favored.

Asking the Relatives Completing a QOL measure on behalf of another requires proxies to put themselves in someone else’s shoes, surmise what it must be like to be them, and imagine how their health and health care affect their experiences of life.4 The use of proxies in health outcomes research may be preferable to excluding more severe cases,32 may increase the probability of a more representative sample,51 and may avoid extra stress in the target (often very sick) group.12 However, some studies have found that QOL assessments made by relatives are in poor agreement with those made, for example, by stroke patients52 or by people with cancer.65 In a study of stroke patients and their caregivers, Mathias et al.52 found good agreement on questions about sensation, mobility, emotion, self-care, ambulation, and dexterity, but only poor to moderate agreement on questions about cognition, pain, hearing, and speech. There is little evidence to support the validity of proxy assessments in cognitively impaired populations.59

HOW SHOULD QUALITY OF LIFE BE MEASURED? This section compares and contrasts the use of generic and disease-specific health outcome measures. Various modes of administration are also discussed.

Generic and Specific Measures Whether to use generic or condition-specific measures in health outcomes research has met with ongoing debate.8 An advantage of generic measures is that they can be used in people with a diverse range of illnesses or health problems, therefore enabling comparisons between groups.32,58 Generic measures are also useful when patients

1056

Quantitating Symptoms, Impairments, and Outcomes

have several concurrent conditions. However, they are often unable to focus on the specific problems of a given condition. Thus some detail may be sacrificed, and such measures may not be sensitive to important changes that occur after an intervention or over the natural course of an illness. Although condition-specific scales are not conducive to comparisons between groups with different conditions, they tend to be more sensitive to changes within and differences between individuals with the specific condition. In addition, it could be argued that because such instruments are directly relevant to respondents with the particular condition, they will be more acceptable and response rates will be maximized. To address the problem of not being able to compare between conditions using specific measures, a battery of measures may be used. In other words, a number of scales can be used in conjunction, which enables some aspects of QOL to be compared with those of other illness groups if required, but also allows pertinent aspects of a particular condition to be assessed in greater depth.

Mode of Administration Those interested in health outcomes research need to consider the mode of administration of the measures they use and whether this is feasible in their particular study or sample. Self-administered measures tend to involve few resources and they eliminate the risk of responses being influenced by the presence of an interviewer. However, physical or visual difficulties that limit ability to complete a questionnaire, or cognitive problems that affect concentration or understanding, may render self-completion an arduous if not impossible task for some individuals. Moreover, someone other than the intended respondent might complete the questionnaire. Face-to-face, interviewer-administered questionnaires require a trained interviewer and more resources. However, the presence of an interviewer means that any misunderstandings about the questions can be clarified, thus increasing the chances of more valid responses. Moreover, interviewer-administered measures tend to yield better response rates for individual items. Conversely, the presence of an interviewer can introduce another kind of bias because respondents may feel uncomfortable in answering particularly sensitive questions truthfully.

REQUIREMENTS OF HEALTH OUTCOME MEASURES This section describes the key psychometric and other prerequisites of any health outcome measures. They relate to the reliability, validity, responsiveness, and acceptability of such measures. A U.K. National Health Service Health

Technology Assessment report32 has identified these properties as being important criteria for patient-based outcome measurement.

Reliability Users of health outcome measures must have some information about the reliability of these measures to be able to interpret scores rationally.28 Evidence of reliability suggests that an instrument is measuring something in a reproducible and consistent way.66 This is particularly important for areas such as perceived health status or QOL, which may alter either naturally or as a result of treatment. Without knowledge of a measure’s reliability, changes in QOL may be falsely charged to a particular intervention when in reality that change occurred as part of the natural history of the condition. Reliability coefficients estimate the proportion of the observed variation in a score that can be considered true rather than random error. The coefficient ranges from 0 (completely unreliable) to 1 (perfect reliability). Acceptable levels of reliability vary depending on the kind of analysis required and the purpose of the measure. Reliability is not an intrinsic property of any outcome measure; it depends on the context and population involved.32 Of the several aspects of reliability, the most common and relevant in relation to patient-based health outcome measures are testretest and internal consistency. Test-retest reliability, or repeatability, is concerned with agreement between results obtained from the same group of individuals on two separate occasions when there has been no real change within the individuals. The correlation between the two scores provides an estimate of the measure’s reliability over time. The gap between the first and second assessment must be sufficient to allow respondents to forget their original responses. At the same time, too long a lapse increases the chance of real changes occurring, thus biasing the test-retest findings.66 For examination of test-retest reliability, an interval of anything from 2 to 14 days is usual. Test-retest reliability is most appropriately expressed statistically as an intraclass correlation coefficient. It is argued that coefficients of more than .90 are needed if one wants to compare the responses of an individual over time, while reliabilities between .50 and .70 are adequate if making group comparisons.32 Internal consistency reliability is concerned with the homogeneity of a scale, that is, the extent to which individual items within a particular subscale or “domain” tap different aspects of the same attribute. A common method of assessing this property is to calculate the value of Cronbach’s alpha,22 a statistical assessment of the correlation of all items within a domain. It measures the “interrelatedness” of items at a given point in time. Coefficients of .70 or more are said to be indicative of a reliable measure. If the value of alpha is too high, this suggests some

1057

Health Outcomes and Quality of Life

degree of redundancy. Therefore, it is recommended that Cronbach’s alpha should not be higher than .90. When used in everyday language, the word “reliability” is associated with words such as “trustworthy” and “good.” However, a reliable instrument is not necessarily a good one because reliability reveals nothing about what is actually being measured, only that something is being measured reliably.28 Thus the validity of a measure is equally important.

Validity Tests of validity determine whether an instrument is measuring what it was intended to measure. Validating a measure involves gathering various types of evidence and, like reliability, validity is not an inherent property of a scale but depends on the context in which it is used.12 For example, a measure may be valid for use with people living in the community but not for those in the hospital. Validating an instrument entails determining the degree of confidence one has in the inferences made about people based on their scores. Because each piece of evidence accumulated will add or take away support, there is no clear point at which a measure can be considered valid. Validity should be evaluated separately each time a measure is to be used for a specific purpose (e.g., with certain groups or in a particular setting). Although there are various kinds of validity, they all address the same issue: the level of confidence one can claim in the inferences drawn from scores. Criterion validity examines the degree to which a scale correlates with another related measure. A “gold standard” is traditionally required as a criterion. There are two components of criterion validity: concurrent (the criteria are measured at the same time) and predictive (criteria are determined later, e.g., in diagnostic tests). Criterion validity is impossible to establish when no gold standard exists for comparison, as is the case for QOL measures.32 Therefore, investigators tend to assess the validity of a measure by comparing it with others (construct validity) and by examining face and content validity. Establishing face validity involves a subjective judgment by “experts” that an instrument’s items seem to measure what was intended. Content validity also involves judgment by experts that the instrument contains sufficient items to adequately measure what was intended. Often this entails only a selection from all possible relevant items. It is important to verify content validity during the construction of new measures. Otherwise, it will be difficult to establish the more empirical types of validity. Such forms of validity should not be evaluated by professional experts alone, however. The involvement of patients in determining the content of such things as QOL instruments, that is, a patient-centered approach, is deemed the most fruitful way of ensuring validity.32 Construct validity is an empirical type of validity, concerned with the extent to which an outcome measure is

related to other scales in a way that would be expected. It requires the specification of theoretical relationships between the construct of interest and other relevant variables.8 There are two components of construct validity: convergent and discriminant. Convergent validity examines correlations with other measures that should be related. Discriminant validity is concerned with the extent of low correlations with other measures that should not be related. Because constructs cannot be unequivocally proved in a single test, construct validation is an ongoing process, as is any form of validation. As more knowledge of the construct is obtained, new predictions may be made and tested. Failure to demonstrate construct validity in a particular study could indicate that the instrument being tested is not valid. However, it could also mean that the comparator measure was inappropriate while the measure being tested is in fact valid. Thus caution is needed when selecting the scales with which to evaluate an instrument’s construct validity.

Responsiveness Depending on the particular requirements of the study, an outcome measure that is responsive may be needed. Responsiveness here is concerned with how accurately a measure can detect changes over time in individuals or groups (e.g., as a result of disease progression/remission, an intervention, or psychological status).28 Responsiveness can be demonstrated by comparing pretest and posttest scores, although associations with other relevant variables, such as clinically assessed change in status, will provide stronger evidence.12 Similarly, Fletcher et al.33 noted that responsiveness may be established in placebo-controlled trials that involve treatments of known efficacy. Others suggest that associations with patient-assessed change (global transition questions) are important indicators of responsiveness.32 Useful responsive measures will be able to detect even small, clinically meaningful differences within an individual over time. Use of unresponsive measures in intervention studies, for example, trials concerned with within-patient differences over time, would fail to detect whether such changes had occurred, thus resulting in misleading findings. It is also important that measures are responsive to changes that patients themselves view as being of significance.

Acceptability Although acceptability of health outcome measures is examined much less frequently than their psychometric properties, it is vital that patients find them acceptable. Acceptability and average completion times need to be considered for any outcome measure, but especially when it is to be used in patients with conditions prone to cognitive problems or fatigue. It is also important that respondents do

1058

Quantitating Symptoms, Impairments, and Outcomes

not experience undue distress in answering the questions. Unacceptability will result in poor response rates, which in turn introduce unnecessary bias to the findings. Previous response rates from earlier research, both to the instrument overall and to individual items, provide some indication of the acceptability of a measure. However, it is best determined by pretesting individual items with relevant patients in terms of the wording, response options, and general layout. The use of such pretesting, “cognitive interviewing,” or “qualitative pretesting” procedures in the development and testing of a range of health outcome instruments has been found to be fruitful.18,27,41 Once evidence of acceptability has been provided in the developmental stages, future evidence can be gleaned by inspecting response rates in subsequent studies.

ASSESSMENT OF QUALITY OF LIFE IN DISEASES ASSOCIATED WITH PERIPHERAL NEUROPATHY A number of condition-specific measures have been developed for use in diseases that may be associated with PN, particularly diabetes, RA, AIDS, and HIV infection. A selection of the most commonly used measures that have been used to assess QOL in these conditions are described in this section.

Disease-Specific Measures Diabetes-Specific Measures The Diabetes Quality of Life Measure (DQOL)44 was developed using a patient-centered approach. It contains 46 core items across four domains (satisfaction with treatment, impact of treatment, worries about future effects of diabetes, and worries about social and vocational issues). There are an additional 10 items for adolescents. There is evidence that the DQOL is reliable (in terms of both internal consistency and test-retest reliability) and valid. There is some evidence of responsiveness, although it has been reported that the satisfaction scale is not particularly sensitive to changes in treatment.17 It was originally developed in English, and translations have been made into Chinese, French, and Spanish. Hammond and Aoki40 developed the Diabetes Impact Measurement Scale (DIMS), consisting of 44 items relating to five domains (specific symptoms, nonspecific symptoms, well-being, diabetes-related morale, and social role fulfillment). It appears to be internally consistent and there is some evidence for construct validity. However, there is no evidence of test-retest reliability or responsiveness, and development of the DIMS did not include patient-centered methods. Developed in the English language, there are also versions in French and Italian.

The Diabetes Health Profile (DHP-1)53 was originally designed for insulin-dependent patients and consists of 32 items across three domains (psychological distress, barriers to activity, and disinhibited eating). Patientcentered methods were included in its development. The DHP-1854 is an adapted version, consisting of 18 items across the same three domains, for use in non–insulin-dependent patients. Both versions seem to be reliable in terms of internal consistency, and there is some evidence of construct validity. There is no evidence of test-retest reliability, and responsiveness to change is yet to be confirmed. The DHP was developed in the English language in the United Kingdom, though there are now versions available in over 10 other languages, including Dutch, French, German, Hebrew, Italian, Spanish, Russian, and Urdu. Development of the Diabetes-39 (D-39)13 included patient-centered methods. It contains 39 items relating to energy and mobility, anxiety and worry, social and peer burden, sexual functioning, and diabetes control. It is internally consistent and there is evidence of validity, but test-retest reliability and responsiveness of the measure have not been determined. The Diabetes-Specific Quality of Life Scale (DSQOLS)11 was developed for people with type 1 diabetes. It consists of 64 items, of which 10 focus on treatment goals, 10 on treatment satisfaction, and 44 on the perceived burden of diabetes. Its development included a patient-centered component. There is some evidence of internal consistency, but none for test-retest reliability. The measure has face validity, with some evidence of construct validity and responsiveness to change.11,37 Originally developed in German, an English language version also exists. The Audit of Diabetes-Dependent Quality of Life (ADDQoL)14 contains 18 items covering physical functioning, symptoms, psychological well-being, social well-being, role activities, and personal constructs. Development of the measure, in English, included patient-centered methods. There is evidence of internal consistency and validity, but no evidence to date of test-retest reliability or responsiveness. The ADDQoL has been translated in more than 15 languages, including French, Italian, Polish, Portuguese (for Brazil), and Spanish (for Mexico). The Diabetes Care Profile (DCP)31 is a measure of selfcare and diabetes-related QOL containing 234 items covering attitudes, beliefs, self-management behaviors, and concerns. It is reasonably reliable in terms of subscale internal consistency, and there is evidence of predictive and construct validity.7,30,31 The DCP was developed in the United States in English. Rheumatoid Arthritis–Specific Measures The 78-item Arthritis Impact Measurement Scale (AIMS2)55 assesses 12 domains within five components: physical (mobility, walking and bending, hand and finger

1059

Health Outcomes and Quality of Life

function, self-care, household tasks); symptom (arthritic pain); affect (tension, mood); social interaction (social activities, support from family and friends); and role (work or employment). It is internally consistent and has good test-retest reliability and validity in people with RA.55 There is also some evidence that it is responsive to change.64 The short-form version (AIMS2-SF)39 contains 26 items across the same five components. This version has been shown to have psychometric properties similar to those of the AIMS2. It was developed in the English language, and translations of AIMS2 have been made in over 10 other languages, including Chinese, Dutch, French, Greek, Hebrew, Italian, Japanese, and Spanish. The Health Assessment Questionnaire (HAQ) contains 20 items across eight domains concerning difficulty in performing ADLs, with evidence of internal consistency, validity, and responsiveness to change.15,16,34 It was originally designed for adults with arthritis but has been used in a wide range of research settings. It has been translated into over 20 languages, including Arabic, Danish, Dutch, French, German, Greek, Hebrew, Italian, Spanish, Swedish, Thai, and Turkish. The Rheumatoid Arthritis-Specific Quality of Life instrument (RAQoL)25 contains 30 items that are summed to produce an overall score. It is reliable in terms of both internal consistency and test-retest reliability, and there is strong evidence of face and construct validity. AIDS- or HIV-Specific Measures The HIV/AIDS-Targeted Quality of Life (HAT/QoL) instrument42 is a 42-item measure developed using patient-centered methods. Areas covered include general health, social function, life satisfaction, mastery, sexual function, financial worries, and medication concerns. There is evidence of validity and moderate to good internal consistency for some domains but not for sexual function, mastery, or medication concerns.23 There is no evidence of test-retest reliability or responsiveness to change over time. The Medical Outcomes Study–HIV (MOS-HIV)43 questionnaire contains 35 items across 11 domains: general health, pain, physical function, role function, social function, mental health, cognitive function, energy/fatigue, health distress, health transition, and QOL. It is based on the Medical Outcomes Study SF-36, a 36-item short-form health status scale, using scales that are relevant to people with HIV. Moderate to good internal consistency, validity, and responsiveness have been reported, but good testretest reliability was found only for some domains.23 The Functional Assessment of Human Immunodeficiency Virus Infection (FAHI)19 contains 55 items across six domains. It is based on the Functional Assessment of Cancer Treatment–General (FACT-G), which assesses both general and HIV disease–specific QoL for patients with chronic illnesses. The FAHI uses the general components from the

FACT-G and HIV-specific questions to assess the diseasespecific portion. Areas covered are physical function, social function, mental health, cognitive function, and functional and global well-being. There is evidence of internal consistency, validity, and responsiveness, but not of test-retest reliability.23 The HIV Overview of Problems–Evaluation System (HOPES)36 contains 106 to 163 items, depending on responses. It has 33 subscales over five dimensions: physical function, psychosocial function, sexual interest, medical interaction/communication, and marital problems/communication. It is based on the Cancer Rehabilitation Evaluation System. Similar to the FAHI, the HOPES uses components from the original questionnaire and HIVspecific questions. Most of the subscales (30 of 33) have good internal consistency, and the remaining 3 have moderate internal consistency.23 There is some evidence of validity and responsiveness24 but no evidence of test-retest reliability.23

Generic Measures A number of generic measures have also been used to assess QOL in conditions associated with PN. One of the most well-known and widely used in these populations is the SF-36,70 which has been used as a measure of QOL in diabetes, RA, and HIV infection. The SF-36 consists of 36 items across eight domains: role limitations–physical, social functioning, role limitations–emotional, physical functioning, mental health, energy/vitality, pain, and general health. These eight domains make up two component scales, physical and mental. There is evidence that the physical functioning, mental health, pain, and energy/vitality domains, and the physical component, are reliable and valid for use in people with RA, but the mental health component was not reliable.63 Ruta et al.63 also found that the pain domain was highly responsive to change in people with RA. Two domains (energy/vitality and physical functioning) were moderately responsive, while the remaining scales displayed small to moderate responsiveness. Lloyd and Orchard50 reported construct validity on most of the SF-36 domains (except emotional and mental health) in people with major diabetes complications, including distal symmetrical polyneuropathy. In an evaluation of gabapentin for painful neuropathy in diabetes, SF-36 scores were more favorable for the treatment group than for the placebo group, providing some evidence of responsiveness in this group.9 Similarly, Ruhland and Shields62 compared home exercise intervention in people with chronic PNs with that in a control group, using the SF-36 to assess impact on QOL. The exercise group improved on the role limitations scales, indicating responsiveness for these two domains in the chronic PN group. There was also evidence of construct validity for the physical functioning domain.

1060

Quantitating Symptoms, Impairments, and Outcomes

Anderson et al.7 compared the SF-36 with a diseasespecific measure (the DCP) in people with diabetes and concluded that the DCP was more suitable for assessing relationships within diabetes (such as the effect of acute complications or regimen on QOL). The authors concluded that the SF-36 was more appropriate if comparing diabetes-related QOL with that of other chronic conditions.3 In another study, the responsiveness of the SF-36 was compared with that of a RA-specific instrument (AIMS2).64 The AIMS2 scales, in particular the physical and pain components, were found to be more responsive than the SF-36 to change in articular status and perceived health. These findings highlight the need to select instruments with care. Vickrey et al.69 compared the SF-36 with a number of disease-specific instruments and concluded that the latter provided unique information not captured by the generic measure. Although a range of generic instruments have been used to assess condition-specific QOL in PN, it appears that the use of condition-specific measures is preferable wherever possible and when comparison with other disease groups or healthy populations is not required. There are now a number of condition-specific measures available for use in diabetes, RA, AIDS, and HIV populations, several of which have been developed using patient-centered approaches. If available measures do not appear to capture the particular issues relating to a specific condition, the development of a new measure may be warranted. This should be based on patient-centered methods, including the pretesting of items and response options, to ensure validity and acceptability.

KEY QUALITY-OF-LIFE ISSUES IN PERIPHERAL NEUROPATHY Very little research has been undertaken on the effects of PNs on QOL.3 Of those studies that have been conducted, most have involved people with diabetes. It has been shown, for example, that the QOL of people with DPN is considerably poorer than that of diabetics without PN or controls. Benbow et al.10 reported significant differences in NHP scores between people with DPN and those without. In particular, people with DPN fared less well than other diabetic patients in terms of pain, energy, physical mobility, and sleep. Those with DPN and others with diabetes had poorer emotional reaction scores than control subjects. Galer and Jensen35 found that people with DPN had higher pain descriptor scores as measured by the Neuropathic Pain Scale. There have been even fewer QOL studies on the effects of PN in other diseases. Abresch et al.2 measured QOL in people with Charcot-Marie-Tooth (CMT) disease. Using the SF-36, they found that these people had significant

bodily pain and that this was a much larger problem than had been reported in the literature, with physical role and energy/vitality being considerably affected. Research by Abresch et al.3 illustrates how QOL measures can assist in improving patient management in those with PNs. They found that pain significantly reduced QOL of people with CMT and those with DPN in terms of the physical functioning, role limitations–physical, vitality, social functioning, role limitations–emotional, and mental health domains of the SF-36. Studies examining the differences in QOL between people with PN in certain conditions and those without are rare. Benbow et al.10 concluded that the detrimental effects of DPN on QOL highlight the need for further research into the effective management of these patients. A similar argument can be made for people with PN generally.

CONCLUSION Although the importance of a holistic approach to the assessment of chronic conditions and their treatment is becoming increasingly recognized, the traditional medical model of disease has dominated clinical trials and other evaluations of therapies.47 By incorporating patient-based outcomes such as QOL, the effects of illness and treatment can be measured in a way that is meaningful to the patient. This applies not only to the research context but also to clinical practice. Although there is no single agreed-on definition of QOL, it is certain that patients’ perceptions are key. Fortunately for those interested in health outcomes, patient-centered methods can be used in the development and testing of structured outcome measures that capture issues of importance to specific patient groups. It has been proposed that the methodologic quality of each QOL study is of greater importance than how QOL is defined,46 although those involved in assessing QOL should be clear about the conceptualizations that they are adopting.

REFERENCES 1. Aaronson, N. K.: Quality of life assessment in clinical trials: methodological issues. Control. Clin. Trials 10:195S, 1989. 2. Abresch, R. T., Carter, G. T., Jensen, M. P., and Kilmer, D. D.: Assessment of pain and health-related quality of life in slowly progressive neuromuscular disease. Am. J. Hosp. Palliat. Care 19:39, 2002. 3. Abresch, R. T., Jensen, M. P., and Carter, G. T.: Healthrelated quality of life in peripheral neuropathy. Adv. Diagn. Manage. Peripher. Nerve Dis. 12:461, 2001. 4. Addington-Hall, J., and Kalra, L.: Who should measure quality of life? BMJ 322:1417, 2001. 5. Ainsworth, M., and Teokul, W.: Breaking the silence: setting realistic priorities for AIDS control in less-developed countries. Lancet 356:55, 2000.

Health Outcomes and Quality of Life 6. Albrecht, G. L., and Devlieger, P. J.: The disability paradox: high quality of life against all odds. Soc. Sci. Med. 48:977, 1999. 7. Anderson, R. M., Fitzgerald, J. T., Wisdom, K., et al.: A comparison of global versus disease-specific quality-of-life measures in patients with NIDDM. Diabetes Care 20:299, 1997. 8. Avis, N. E., and Smith, K. W.: Conceptual and methodological issues in selecting and developing quality of life measures. Adv. Med. Sociol. 5:255, 1994. 9. Backonja, M., Beydoun, A., Edwards, K. R., et al.: Gabapentin for the symptomatic treatment of painful neuropathy in patients with diabetes mellitus: a randomized controlled trial. JAMA 280:1831, 1998. 10. Benbow, S. J., Wallymahmed, M. E., and MacFarlane, I. A.: Diabetic peripheral neuropathy and quality of life. Q. J. Med. 91:733, 1998. 11. Bott, U., Muhlhauser, I., Overmann, H., and Berger, M.: Validation of a diabetes-specific quality-of-life scale for patients with type 1 diabetes. Diabetes Care 21:757, 1998. 12. Bowling, A.: Measuring Disease: A Review of DiseaseSpecific Quality of Life Measurement Scales. Buckingham, UK, Open University Press, 1995. 13. Boyer, J. G., and Earp, J. A.: The development of an instrument for assessing the quality of life of people with diabetes. Med. Care 35:440, 1997. 14. Bradley, C., Todd, C., Gorton, T., et al.: The development of an individualized questionnaire measure of perceived impact of diabetes on quality of life: the ADDQoL. Qual. Life Res. 8:79, 1999. 15. Brazier, J. E., Harper, R., Munro, J., et al.: Generic and condition-specific outcome measures for people with osteoarthritis of the knee. Rheumatology (Oxford) 38:870, 1999. 16. Brown, J. H., Kazis, L. E., Spitz, P. W., et al.: The dimensions of health outcomes: a cross-validated examination of health status measurement. Am. J. Public Health 74:159, 1984. 17. Bruttomesso, D., Pianta, A., Crazzolara, D., et al.: Continuous subcutaneous insulin infusion (CSII) in the Veneto region: efficacy, acceptability and quality of life. Diabet. Med. 19:628, 2002. 18. Buck, D., Jacoby, A., Massey, A., and Ford, G.: Pre-testing items in a stroke specific quality of life measure: the value of cognitive interviewing. J. Neurolinguistics 13:267, 2000. 19. Cella, D. F., McCain, N. L., Peterman, A. H., et al.: Development and validation of the Functional Assessment of Human Immunodeficiency Virus Infection (FAHI) quality of life instrument. Qual. Life Res. 5:450, 1996. 20. Chung, M. C., Killingworth, A., and Nolan, P.: A critique of the concept of quality of life. Int. J. Health Care Qual. Assurance 10:80, 1997. 21. Cramer, J. A., and Spilker, B.: Quality of Life and Pharmacoeconomics: An Introduction. Philadelphia, Lippincott–Raven, 1998. 22. Cronbach, L. J.: Coefficient alpha and the internal structure of tests. Psychometrika 16:297, 1951. 23. Davis, E. A., and Pathak, D. S.: Psychometric evaluation of four HIV disease-specific quality-of-life instruments. Ann. Pharmacother. 35:546, 2001. 24. de Boer, J. B., Sprangers, M. A., Aaronson, N. K., et al.: A study of the reliability, validity and responsiveness of the HIV Overview of Problems Evaluation System (HOPES) in

25.

26. 27.

28.

29. 30.

31.

32.

33.

34.

35.

36.

37.

38. 39.

40.

41.

42.

1061

assessing the quality of life of patients with AIDS and symptomatic HIV infection. Qual. Life Res. 5:339, 1996. de Jong, Z., van der Heijde, D., McKenna, S. P., and Whalley, D.: The reliability and construct validity of the RAQoL: a rheumatoid arthritis-specific quality of life instrument. Br. J. Rheumatol. 36:878, 1997. Efrat, S.: Preventing type 1 diabetes mellitus: the promise of gene therapy. Am. J. Pharmacogenomics 2:129, 2002. Elasy, T. A., Samuel-Hodge, C. D., DeVellis, R. F., et al.: Development of a health status measure for older AfricanAmerican women with type 2 diabetes. Diabetes Care 23:325, 2000. Fallowfield, L.: The Quality of Life: The Missing Measurement in Health Care. London, Souvenir Press, 1990. Farquhar, M.: Definitions of quality of life: a taxonomy. J. Adv. Nurs. 22:502, 1995. Fitzgerald, J. T., Anderson, R. M., Gruppen, L. D., et al.: The reliability of the Diabetes Care Profile for African Americans. Eval. Health Prof. 21:52, 1998. Fitzgerald, J. T., Davis, W. K., Connell, C. M., et al.: Development and validation of the Diabetes Care Profile. Eval. Health Prof. 19:208, 1996. Fitzpatrick, R., Davey, C., Buxton, M. J., and Jones, D. R.: Evaluating patient-based outcome measures for use in clinical trials. Health Technol. Assess. 2(14), 1998. Fletcher, A.E., Dickinson, E.J., and Philp, I.: Review. Audit measures: quality of life instruments for everyday use with elderly patients. Age Ageing 21:142, 1992. Fries, J. F., Spitz, P., Kraines, R. G., and Holman, H. R.: Measurement of patient outcome in arthritis. Arthritis Rheum. 23:137, 1980. Galer, B. S., and Jensen, M. P.: Development and preliminary validation of a pain measure specific to neuropathic pain: the Neuropathic Pain Scale. Neurology 48:332, 1997. Ganz, P. A., Coscarelli-Schag, C. A., Kahn, B., et al.: Describing the health-related quality of life impact of HIV infection: findings from a study using the HIV Overview of Problems—Evaluation System (HOPES). Qual. Life Res. 2:109, 1993. Garratt, A. M., Schmidt, L., and Fitzpatrick, R.: Patientassessed health outcome measures for diabetes: a structured review. Diabet. Med. 19:1, 2002. Gill, T. M., and Feinstein, A. R.: A critical appraisal of the quality of quality-of-life measurements. JAMA 272:619, 1994. Guillemin, F., Coste, J., Pouchot, J., et al: Quality of Life in Rheumatology Group. The AIMS2-SF: a short form of the Arthritis Impact Measurement Scales 2. Arthritis Rheum. 40:1267, 1997. Hammond, G. S., and Aoki, T. T.: Measurement of health status in diabetic patients: diabetes impact measurement scales. Diabetes Care 15:469, 1992. Harris-Kojetin, L. D., Fowler, F. J., Brown, J. A., et al: The use of cognitive testing to develop and evaluate CAHPS™ 1.0 Core Survey items. Med. Care 37(Suppl. 3):MS10, 1999. Holmes, W. C., and Shea, J. A.: Performance of a new, HIV/AIDS-Targeted Quality of Life (HAT-QoL) instrument in asymptomatic seropositive individuals. Qual. Life Res. 6:561, 1997.

1062

Quantitating Symptoms, Impairments, and Outcomes

43. Holmes, W. C., and Shea, J. A.: Two approaches to measuring quality of life in the HIV/AIDS population: HAT/QoL and MOS-HIV. Qual. Life Res. 8:515, 1999. 44. Jacobson, A., Barofsky, I., Cleary, P., and Rand, L.: Reliability and validity of a diabetes quality-of-life measure of the Diabetes Control and Complications Trial (DCCT). Diabetes Care 11:725, 1988. 45. Jacoby, A.: Assessing quality of life in patients with epilepsy. Pharmacoeconomics 9:399, 1996. 46. Jenkins, C. D.: Assessment of outcomes of health intervention. Soc. Sci. Med. 35:367, 1992. 47. Jenkinson, C., Fitzpatrick, F., and Swash, M.: Amyotrophic lateral sclerosis/motor neurone disease. In Jenkinson, C., Fitzpatrick, R., and Jenkinson, D. (eds.): Health Status Measurement in Neurological Disorders. Abingdon, UK, Radcliffe Medical Press, 2000. 48. Jenney, M. E.: Theoretical issues pertinent to measurement of quality of life. Med. Pediatr. Oncol. Suppl. 1:41, 1998. 49. LaPointe, L. L.: Quality of life with brain damage. Brain Lang. 71:135, 2000. 50. Lloyd, C. E., and Orchard, T. J.: Physical and psychological well-being in adults with type 1 diabetes. Diabetes Res. Clin. Pract. 44:9, 1999. 51. Magaziner, J., Bassett, S. S., Hebel, J. R., and GruberBaldini, A.: Use of proxies to measure health and functional status in epidemiological studies of community-dwelling women aged 65 years and older. Am. J. Epidemiol. 143:283, 1996. 52. Mathias, S. D., Bates, M. M., Pasta, D. J., et al.: Use of the Health Utilities Index with stroke patients and their caregivers. Stroke 28:1888, 1997. 53. McColl, E., Eccles, M., Shiels, C., et al.: The Diabetes Health Profile (DHP): a new instrument for assessment the psychosocial profile of insulin requiring patients—development and psychometric evaluation. Qual. Life Res. 5:242, 1996. 54. Meadows, K. A., Abrams, C., and Sandbaek, A.: Adaptation of the Diabetes Health Profile (DHP-1) for use with patients with type 2 diabetes mellitus: psychometric evaluation and cross-cultural comparison. Diabet. Med. 17:572, 2000. 55. Meenan, R. F., Mason, J. H., Anderson, J. J., et al.: AIMS2: the content and properties of a revised and expanded Arthritis Impact Measurement Scales Health Status Questionnaire. Arthritis Rheum. 35:1, 1992. 56. Meyers, A. R., Gage, H., and Hendricks, A.: Health-related quality of life in neurology. Arch. Neurol. 57:1224, 2000. 57. Owens, D. K., Cardinalli, A. B., and Nease, R. F.: Physicians’ assessments of the utility of health states associated with human immunodeficiency virus (HIV) and hepatitis B virus (HBV) infection. Qual. Life Res. 6:77, 1997. 58. Peto, V., Jenkinson, C., and Fitzpatrick, R.: Parkinson’s disease. In Jenkinson, C., Fitzpatrick, R., and Jenkinson, D. (eds.): Health Status Measurement in Neurological Disorders. Abingdon, UK, Radcliffe Medical Press, 2000.

59. Riemsma, R. P., Forbes, C. A., Glanville, J. M., et al.: General health status measures for people with cognitive impairment: learning disability and acquired brain injury. Health Technol. Assess. 5:1, 2001. 60. Romanelli, F., and George, K.: 20 years of HIV treatment: past, present, and future. J. Am. Pharm. Assoc. (Wash.) 42(5 Suppl. 1):S42, 2002. 61. Rothwell, P. M., McDowell, Z., Wong, C. K., and Dorman, P. J.: Doctors and patients don’t agree: cross sectional study of patients’ and doctors’ perceptions and assessments of disability in multiple sclerosis. BMJ 314:1580, 1997. 62. Ruhland, J. L., and Shields, R. K.: The effects of a home exercise program on impairment and health-related quality of life in persons with chronic peripheral neuropathies. Phys. Ther. 77:1026, 1997. 63. Ruta, D. A., Hurst, N. P., Kind, P., et al.: Measuring health status in British patients with rheumatoid arthritis: reliability, validity and responsiveness of the short form 36-item health survey (SF-36). Br. J. Rheumatol. 37:425, 1998. 64. Salaffi, F., Stancati, A., and Carotti, M.: Responsiveness of health status measures and utility-based methods in patients with rheumatoid arthritis. Clin. Rheumatol. 21:478, 2002. 65. Sneeuw, K. C., Albertsen, P. C., and Aaronson, N. K.: Comparison of patient and spouse assessments of health related quality of life in men with metastatic prostate cancer. J. Urol. 165:478, 2001. 66. Streiner, D. L., and Norman, G. R.: Health Measurement Scales: A Practical Guide to Their Development and Use. Oxford, UK, Oxford University Press, 1989. 67. Tennant, A., Geddes, J. M. L., Fear, J., et al.: Outcome following stroke. Disabil. Rehabil. 19:278, 1997. 68. van Brakel, W. H.: Peripheral neuropathy in leprosy and its consequences. Lepr. Rev. 71(Suppl.):S146, 2000. 69. Vickrey, B. G., Hays, R. D., Genovese, B. J., et al.: Comparison of a generic to disease-targeted health-related quality-of-life measures for multiple sclerosis. J. Clin. Epidemiol. 50:557, 1997. 70. Ware, J. E., and Sherbourne, C. D.: The MOS 36-Item Short Form Health Survey (SF-36). I: conceptual framework and item selection. Med. Care 30:473, 1992. 71. Wilson, K. A., Dowling, A. J., Abdolell, M., and Tannock, I. F.: Perception of quality of life by patients, partners and treating physicians. Qual. Life Res. 9:1041, 2000. 72. Wolfe, C. D. A.: The impact of stroke. Br. Med. Bull. 56:275, 2000. 73. World Health Organization (WHO): The International Classification of Impairments, Disabilities and Handicaps. Geneva, WHO, 1980. 74. World Health Organization (WHO): International Classification of Functioning and Disability: A New Release from WHO [press release]. Geneva, WHO, 1999. Available at: www.who.int/icidh 75. World Health Organization (WHO): International Classification of Functioning, Disability and Health. Geneva, WHO, 2001. Available at: www3.who.int/icf/icftemplate.cfm

42 Quantitative Sensation Testing PETER J. DYCK, PETER C. O’BRIEN, DAVID M. JOHNSON, CHRISTOPHER J. KLEIN, AND P. JAMES B. DYCK

Uses of Quantitative Sensation Testing Cutaneous Sensory Receptors Assessing Sensation and Vision and Hearing Clinical Assessment Versus QST General Description of QST Why Test Different Modalities of Skin Sensation? Testing for Hyposensitivity and Hypersensitivity Subjective Differences among Anatomic Sites QST Threshold Variation with Anatomic Site, Age, Sex, and Anthropomorphic Characteristics in Health and Disease Types of QST Commercially Available Devices and Instruments Criteria and Standards for L-QST Using Natural Stimuli Subject or Patient Factors

Ambient Environment Instructions Comparison of QST Results among Different Devices and Systems Baseline Conditions of the Stimulus Ideal Stimuli and Algorithms of Testing Stimulus Waveform Null Stimuli Calibration of Stimulus Magnitude Algorithm of Testing and Finding Threshold Touch-Pressure Sensation Two-Point Discrimination Vibration Sensation Computer-Assisted Sensory Examination System IV VDT as the Measure of Mechanoreceptor Function VDT Using CASE IV Quality Assurance of VDT VDT in Health

USES OF QUANTITATIVE SENSATION TESTING Quantitative sensation testing (QST) is useful for characterizing sensation 1. In health at different anatomic sites with age, sex, and anthropomorphic features; with development, maturation, and aging; and with neuroanatomic, neurophysiologic, and metabolic alterations 2. In disease by class of sensory receptors and nerve fibers for recognition of a. hyposensitivity or hypersensitivity b. anatomic patterns of abnormality c. alteration with time (e.g., to recognize onset, worsening, or improvement) d. hyperalgesia

Use of VDT Thermal Sensations Cooling Detection Threshold Using CASE IV Warm Detection Threshold Warmth or Heat-Pain Sensation Estimating Percentile Abnormality Sensitivity, Reproducibility, and Concordance of L-QST Results with Neuropathic Impairment QST in Medical Practice Developing Sensation-Testing Approaches Useful in Clinical Practice Hyposensitivity and Hypersensitivity Distribution of Sensation Abnormality Inferring the Anatomic-Pathologic Abnormalities of Fibers Inferring Improvement or Worsening Assessing Impairment QST in Research

3. In epidemiologic surveys, controlled clinical trials, or in other research 4. For education—as a standard for comparison to clinical sensation testing QST is not ideal as a stand-alone test to make diagnoses or to determine impairment, disability, or damages. This is not an indictment of QST, because we would make the same statement about other tests used alone (e.g., nerve conduction and electromyography [EMG]). As noted by Dyck et al., “QST requires the attention and cooperation of the subject—[the] patient must be old and intelligent enough to understand what is asked by the test, alert and not taking mind-altering medications and not biased to a certain test outcome. We know of no algorithm of psychophysical testing that can reliably overcome the bias toward showing abnormality inherent in patients who wish 1063

1064

Quantitating Symptoms, Impairments, and Outcomes

(consciously or unconsciously) to demonstrate more disability than they have for whatever reason.”6

CUTANEOUS SENSORY RECEPTORS Cutaneous sensation is mediated by topographically distributed receptors of several kinds, innervated by different classes of afferent nerve fibers. These receptors are distributed in a grid pattern in the epidermis and dermis (Fig. 42–1; also see Chapters 8 and 34). They are also distributed throughout the deep tissues. Large mechanoreceptor fibers ascend ipsilaterally in the posterior column and synapse with secondary relay neurons in the nucleus gracilis (lower limb) or cuneatus (upper limb). The primary sensory neurons for pain, in comparison, end in the dorsal horn, and secondary neurons cross over at the entering segmental level and ascend in the lateral spinothalamic tract. Receptors vary in density and in depth among different regions of skin. The characteristics of the mammalian cutaneous receptors are described in Chapters 8 and 34 and in earlier editions of this textbook. Selective loss of a modality of sensation, for example, mechanoreception (e.g., vibration sensation), occurs early in the course of Friedreich’s ataxia, without abnormality of cool, warm, or heat-pain sensation, and this selective sensation loss is correlated with a decrease in the amplitude of the A potential of the compound action potential of the sural nerve in vitro and with decreased numbers of large myelinated fibers in diameter histograms of transverse sections of nerve. Conversely, selective loss of cooling and heat-pain sensation occurs in some cases of transthyretin amyloidosis, and this selective sensation loss is correlated with selective decrease of A and C potentials of biopsied sural nerves in vitro and with absence of small myelinated and unmyelinated sensory fibers. Thus there is a close association of loss of a modality of sensation with selective abnormality of a peak or peaks of the compound action potential in vitro of the sural nerve and with selective abnormality of peaks of the diameter histogram of nerve fibers of the same nerve. These studies are undergirded by the studies of Erlanger and Gasser16 and by later physiologists relating peaks of the compound action potential of nerve to size classes of nerve fibers (see Chapter 39) and single sensory unit recordings (see Chapter 8). From single-unit recordings of afferent cutaneous units of experimental animals2 and of humans,18,19,27 the characteristics of low-threshold mechanoreceptors have been described. The units are classified by their receptive fields and their adaptive characteristics. For glabrous (nonhairy) and semiglabrous skin, the low-threshold rapidly adapting (RA) receptors have characteristic receptive fields. The RA receptors have small fields with sharply defined borders. Within the fields are discrete zones of varying sensitivities. These findings imply that the responsible

FIGURE 42–1 Horizontal section of 3-mm punch biopsy specimen from skin taken from the left great toe of a healthy 7-year-old girl. (Acetylthiocholinesterase, 32.) (From Bolton, C. F., Winkelman, R. K., and Dyck, P. J.: A quantitative study of Meissner’s corpuscles in man. Neurology 16:1, 1966, with permission.)

receptors are clustered on the terminal arborizations of one axon and lie superficially. The likely receptors are Meissner’s corpuscles (see Fig. 42–1). The other lowthreshold RA mechanoreceptor is the pacinian corpuscle (PC) receptor, which has a large receptive field with sloping borders and without discrete zones of differing thresholds within the receptive field. The almost certain receptor is the pacinian corpuscle, which is innervated by one fiber and lies deep in the tissue.7 Two other lowthreshold, mechanosensitive units are slowly adapting and, following the convention in animal studies,1 are designated SA I and SA II. The SA II units may have a directional sensitivity. There is some evidence that the receptors for the SA I and SA II units are the Merkel cell complex and Ruffini’s corpuscle, respectively. Although there appear to be mechanoreceptor units innervated by C fibers in animals, there is no evidence for low-threshold mechanoreceptor units innervated by C fibers in human skin. No thermal or nociceptive receptors of human skin are innervated by A or  afferent fibers. The nerve fibers thought to be involved in cool sensation are A sensory fibers. Pulses of cooling are experienced without delay as a small splash of coolness. By contrast, heat-pain, especially burning pain, is slightly delayed and is mediated by naked nerve endings and conveyed by dorsal root C fibers (drC).

ASSESSING SENSATION AND VISION AND HEARING QST is the sensation counterpart of clinical tests of sight and hearing. In all three cases (sight, hearing, and sensation) psychophysical approaches are used. In performance of

Quantitative Sensation Testing

QST, it is first necessary to specify the modality of sensation that is being tested: touch-pressure, vibration, cooling, warming, heat-pain, two-point discrimination, an electrical shock, or other. Next it is necessary to define the threshold that is to be identified: detection, just noticeable difference (JND), a defined sensation quality (warming as distinguished from heat-pain), or level of intensity (e.g., a heatpain experience rated as 5 on a scale with 0  no pain and 10  most severe pain). After an explanation of the test to be given, it must be ascertained that the subject or patient is in a condition to be tested and understands what is asked and is cooperative. Then defined and quantitated stimuli are presented and the subject is asked to identify the smallest magnitude of the stimulus that can be recognized (detection or perception threshold or another stimulus-response criterion). In tests of sensation, as with sight and hearing, threshold is reported according to the magnitude of the physical stimulus (the physical component of psychophysical testing). Perception involves mental recognition and response by the subject (the psychological component of psychophysical testing). Considerable thought needs to be given to both aspects of psychophysical testing. Ideally, the physical stimulus is describable and quantifiable in physical units. The waveform used should be the same over a broad range of stimulus magnitudes and superimposed on a standard load or condition. The psychological component needs also to be controlled, standardized, and carefully assessed to ensure that only adequately prepared and appropriate persons are tested, that adequate algorithms of testing are used, that conditions of testing are optimal, and that results are compared to appropriate controls. Vision is evaluated by a variety of tests assessing the function of retinal cones and rods and their neural connections. Similarly, hearing is evaluated by assessing a patient’s ability to recognize different pure tones and to discriminate complex sounds (e.g., speech) and other auditory stimuli. Skin sensation tests have been developed to detect various physical stimuli that elicit sensations of touch-pressure, vibration, cooling, warming, heat or mechanical stimuli inducing pain, movement or static position, electric shocks to simulate naturally occurring stimuli, or other complex sensations. The important point is that measurement of skin sensation provides information about sensation in health by modality of sensation and for different regions of the body with development, maturation, and aging, and about alterations of sensation with disease (implying loss or dysfunction of classes of receptors and their nerve fibers or tracts). QST encompasses different approaches and techniques for stimulating cutaneous receptors directly with quantitative stimuli that simulate naturally occurring events in daily life (e.g., touch-pressure, vibration, flutter, warming, cooling, compression or sticking pain, and so on). QST approaches to simulate natural stimuli might be called QST natural stimulation. Other approaches employ electrical stimulation of cutaneous nerve fibers (QST electrical

1065

stimulation). The tests available for hearing also can be divided by whether natural or electrical stimulation is used. In QST there is considerable variability in the modalities of sensation being tested, degree of control of the testing environment, baseline testing conditions, the stimulus waveform, the range of stimulus magnitudes employed, the algorithm of testing and finding threshold, and the availability of normal values. Several levels of complexity of QST may be distinguished. In this chapter, we emphasize QST natural stimulation approaches and results. The QST natural stimulation approaches may further be divided into semiquantitative bedside, quantitative bedside, and quantitative laboratory.

CLINICAL ASSESSMENT VERSUS QST Physicians assess sensation to detect abnormality, which, if present, may imply disordered function of sensory end organs, afferent neurons (axons), central tracts, nuclei, or cerebral regions. The cutaneous distribution of the sensory abnormality may aid in localizing the responsible neuropathologic lesion. Sensation may also be assessed to recognize the classes of neurons (axons) affected in disease, sometimes selectively, which provides helpful information in differential diagnosis. Serial evaluation of sensation may aid in appreciating worsening or improvement of the sensory deficit; an indication of whether the disease is static, worsening, or improving; or whether there has been a therapeutic effect. Several facts must be considered before deciding whether sensation is abnormal and before interpreting the implications of the findings. First, detection thresholds in health vary with modality, site, age, and sex; therefore, to declare whether a result is normal or abnormal, these variables must be taken into account in expressing test results. Results also must be corrected for other variables affecting threshold. For example, sensation may be affected by drowsiness, inattention, altered states of consciousness, or noncooperation, and these conditions must therefore be avoided or taken into account in interpretation of findings. Neurologic disease may affect sensory systems and result in a decrease or elevation of detection threshold or alter sensation subjectively. Elevation of detection threshold, therefore, may reflect decreased and/or distorted sensation. Such spontaneous symptoms as paresthesia and pain may interfere with assessment of sensation. It may be difficult for the patient to discriminate external stimuli from internal experiences. Because neurologic disease may selectively affect different segmental levels, different classes of neurons (or axons), and different sites, and because thresholds vary with site, age, sex, and physical characteristics, standard sites for which control test values are known should be tested. The two approaches (clinical and psychophysical) to sensory testing are each suited for different purposes. The

1066

Quantitating Symptoms, Impairments, and Outcomes

clinical approach provides a quick, nonvalidated survey of sensation over the surface of the body without use of standardized or validated tests. It may be used in diverse settings, employs inexpensive hand-held instruments and simple devices, and is conducted quickly. Using a wisp of cotton, a sharp pin, a tuning fork, and other simple devices, the examiner surveys the surface of the body for regions of decreased, absent, or altered sensation. The patient reports whether he or she can or cannot feel the stimulus. The examiner also assesses for regions of altered feeling. The physician asks the patient to compare the sensation of the same region of the two sides of the body, of proximal and distal regions from the same side, and of symptomatic and nonsymptomatic regions. Abnormality is recognized by failure to report stimuli that should have been felt (based on the examiner’s notion of what constitutes a normal response considering the site, age, and sex), and by reported differences of feeling between comparable anatomic regions. Because the clinical approach provides satisfactory information about gross sensory loss, especially when this has a sharp level or border, it is the approach most commonly employed by neurologists. Its main drawbacks are that physical characteristics of the stimuli are essentially undefined and variable, the sequential steps (the algorithm) of testing and finding thresholds are not specified, the actual tests performed vary among neurologist, and the thresholds obtained are not quantitated or validated. The results of the evaluation, therefore, may be quite variable depending on the factors mentioned and the experience of the examiner and the amount of time spent. Clinical assessment of sensation is adequate for the detection and characterization of gross abnormality. The laboratory QST (L-QST) approach to assessing sensation is suited for specific clinical and research questions. First, it is well suited to assess the anatomic variability of detection thresholds (i.e., with site, side, age, and sex in health). Second, the approach may be used to study the physiology of sensation. A third use is to characterize sensory abnormality in varieties of neurologic disease and to infer the class and kind of sensory fibers that are dysfunctional. A fourth use is to detect subclinical involvement in a population of persons (e.g., genetic kindreds, industrial workers, those with a medical condition, or geographic isolates) at risk. A fifth use is to assess various modes of therapy in controlled clinical trials. A sixth use is to relate abnormalities of sensation to measured attributes of the compound action potential in vitro, to axonal flow, to numbers of myelinated and unmyelinated fibers, and to pathologic abnormalities (see Chapters 32 and 39). For epidemiologic and controlled clinical trials, QST provides a standardized test at defined anatomic sites. Sensitivity, specificity, accuracy, and reproducibility can be assessed. Quantitative sensory examination is not performed to save time or money. The clinical evaluation always can be done more quickly. We have used QST approaches serially in patients withdrawn from a toxic medication,21 in trials of therapy in chronic renal failure,7 in trials

of prednisone versus no treatment in chronic inflammatory demyelinating polyradiculoneuropathy,13 and in trials of various levels of hyperglycemia control in diabetic neuropathy.25 QST can be used as an educational tool. We studied how well sensation examinations conducted by peripheral neuropathy fellows in specialized training correlated with L-QST using Computer-Assisted Sensory Examination System IV (CASE IV) results of vibration detection threshold (VDT) in serially evaluated patients referred to us (P.J.D.). First, we established that the correlation of L-QST with a composite score of nerve conductions was higher than the correlation with clinical sensory testing. Clinical sensory evaluation overestimated sensory abnormality more often than underestimated it. This overestimate was attributable to not taking age, height, and body surface area sufficiently into account. These biographic and anthropomorphic factors are the same ones we have shown to be the important covariates of L-QST of VDT.

GENERAL DESCRIPTION OF QST Why Test Different Modalities of Skin Sensation? Different modalities of sensation should be tested to detect altered function (hyposensitivity) or hypersensitivity of different classes of sensory receptors or fibers. Disease of large-diameter sensory fiber receptors, nerve fibers, and central tracts is associated with decreased or absent sensation (increased threshold) mediated by A sensory fibers. Decreased cooling sensation is correlated with abnormality of small myelinated (A) sensory fibers. Decreased heat-pain sensation is associated with abnormality of unmyelinated fibers (drC fibers). A lowered pain threshold, a heightened response to a small heat-pain stimulus, or a pain response to a mechanical stimulus not usually painful (allodynia) is characteristic of hyperalgesia, a state of hypersensitivity of C nociceptors, their fibers or their central centers, and /or the action of pain peptides.

Testing for Hyposensitivity and Hypersensitivity QST is used to recognize hyposensitivity, hypersensitivity, and both hyposensitivity and hypersensitivity. In hyposensitivity, the threshold at which the stimulus is felt is larger than normal—a larger stimulus must be given to elicit the defined threshold. Hyposensitivity, assuming that it has been adequately evaluated, implies loss of a modality of sensation. In peripheral neuropathy, this hyposensitivity is attributable to loss of sensation receptors or fibers as a result of axonal blockade or degeneration. Decreased sensation may also be due to dysfunction and lost function in central tracts or associative centers of the brain. In hypersensitivity, the threshold

1067

Quantitative Sensation Testing

is at smaller than normal magnitudes of stimulation, the subjective response increases excessively as increasingly stronger stimuli are given (an increased stimulus-response slope), or both lowered threshold and increased stimulus-response slopes are found. This pattern has been recognized only for mechanically induced pain (allodynia) and for heat-pain (thermal hyperalgesia) and is thought to be due to a dysfunction of damaged drC receptors or their fibers or changes in the local microenvironment in the region of the receptor or distal nerves or at the dorsal horn of the spinal cord. Thus threshold for a modality of sensation at a given anatomic site may be described as normal, hyposensitive or insensitive, hypersensitive, or perhaps both hyposensitive and hypersensitive. In CASE IV testing, typical patterns of abnormality are (1) normal sensation; (2) pan-modality sensory loss— VDT, cooling detection threshold (CDT), and heat-pain threshold (HPT) are all elevated, suggesting disease or decreased function of all classes of sensory fibers; (3) selective large fiber sensory loss; (4) selective small fiber sensory loss; (5) hyperalgesia; and (6) mixtures.

Subjective Differences among Anatomic Sites Using CASE IV with a range of increasingly stronger stimuli, from very small to large and at standard sites, it is possible to order the most to the least sensitive anatomic sites. For VDT in our normal subject cohort, the index finger was most sensitive followed by the toe and then by fleshy parts of the leg, thigh, forearm, and shoulder. For CDT, the order was the dorsal hand, volar forearm, and face, followed by the foot, leg, thigh, and shoulder. Warming is felt best in the face, hand, and forearm, but may not be felt at all on the dorsum of the foot, especially in older people. For warmth on the foot, the first sensation felt may be heatpain. Warm receptors are thought to be infrequent over the dorsal foot. For the same modality of sensation and attempting to exceed threshold by approximately the same degree, there are differences of the subjective experience of sensation among anatomic sites. Using pyramidal pulses of warming or cooling, there is also a subjective difference. Whereas the cooling pulse feels like a “splash” of cool liquid with a definite onset and disappearance, warming develops more slowly and is less distinct. The heat-pain level 5 stimulus feels approximately the same among different sites over the body. This is discussed in more detail in the next section.

QST Threshold Variation with Anatomic Site, Age, Sex, and Anthropomorphic Characteristics in Health and Disease In addition to differences among anatomic sites, sensation thresholds vary with age, sex, height, and weight. Differences in threshold presumably relate to the density of the

receptors, the depth at which the receptors lie, the thickness and nature of overlying tissues, and, perhaps, the property of the receptors themselves. The area of cerebral cortex related to face and hand sensation is also greater than the areas for sensation from other regions of the body, and this may also affect sensation threshold. The pattern of hyposensitivity and hypersensitivity also varies with anatomic site and disease. Recognition of the alterations from normal, therefore, is helpful in characterization and diagnosis of neuropathy.

Types of QST Semiquantitative QST This is the least complex approach and involves use of semiquantitative handheld devices without control of environment, baseline conditions, algorithms of testing and finding threshold, or use of refined reference values. For tactile sensation, cotton wool, Frey’s hairs, glass rods, or nylon monofilaments are used to assess touch-pressure sensation. For thermal sensation, tubes or flasks containing warm or cool liquid are used. Stickpins are used to test pain sensation. The examiner’s hand is used to move fingers or toes in testing joint motion. A tuning fork is used to assess vibration sensation. The approach is usually comparative, for example, to identify sensation loss or alteration at or distal to a segmental level or in the distribution of a spinal root or peripheral nerve. Findings are expressed as absence, decrease, or alteration of feeling between compared regions. Precise quantification of the magnitude of the stimulus is not possible using handheld instruments, and results are generally not expressed quantitatively. The affected skin region or site is described as showing decreased or lost feeling (of a modality), as being hyperesthetic, or as demonstrating allodynia (discomfort or pain from mechanical stimuli that in normal persons are not painful) or hyperalgesia (pain induced at a lower threshold than normal or a response of pain that increases excessively with stronger stimuli). Handheld devices may also be used in a controlled laboratory setting to elicit quantitative results (see later sections). Use of a monofilament nylon stylus of standard diameter and length, calibrated to bend when a force of 10 g is applied (the Semmes Weinstein monofilament), has been advocated to test for sensory loss in diabetic and leprosy patients (to recognize a degree of sensation loss, making them vulnerable to development of plantar ulcers) and for recovery of sensation after nerve injury. Threshold for the anterior foot (toes and metatarsal heads) is defined as 7 correct responses in 10 trials. The approach is inexpensive and fast and may provide some useful information. However, the force delivered is undoubtedly variable, loss of touch-pressure sensation probably is not the primary modality that correlates with plantar ulcers, and one level of response to one stimulus level probably is insufficient to grade sensation loss, which varies among modalities of

1068

Quantitating Symptoms, Impairments, and Outcomes

sensation and with age, in some cases with sex, and with anthropomorphic factors. Quantitative QST At the next more sophisticated level of QST, stimuli are more precisely quantitated or more rigorously tested. Typical instruments used are hand held; testing is done at the bedside or in the physician’s office without control of the testing environment or control of all aspects of testing. The algorithm of testing and estimating threshold is not predetermined or controlled, and normative results are not usually provided. The advocates of this approach leave all aspects of testing and comparison to controls to the examiner. Laboratory QST At the most sophisticated levels, all aspects of QST are predetermined, standardized, and controlled: the precondition of patients (patients must be alert, cooperative, and not on mind-altering drugs); the conditions and environment of testing (a separate room suitable for testing); the instructions (standard and appropriate); the baseline load or condition of the stimulus; the stimulus waveform (standard, the same over a broad range of magnitudes, and quantified); the algorithm of testing and finding threshold; the comparison of results to those of a normative population; and the expression of results. Typically, the patient is brought to a special laboratory where essentially all aspects of the testing routine are predetermined and standardized (Figs. 42–2 and 42–3). The instructions are predetermined, and standard instructions are read to the patient. Responses to patients’ questions are standard. Defined anatomic sites are tested by specific modalities of sensation using preprogrammed ramps or steps of stimulus type and magnitude and validated algorithms of testing, all controlled by computer programs. Results are expressed as a percentile and normal deviate value ( scores) considering modality, site, age, sex, and applicable anthropomorphic characteristics (e.g., height, weight, and body mass index).

Commercially Available Devices and Instruments The list of test instruments, devices, and systems provided in Table 42–1 includes the ones known to the authors and may not be complete. It provides potential users with manufacturer contact information to acquire further information.

CRITERIA AND STANDARDS FOR L-QST USING NATURAL STIMULI In order to have standardized laboratory testing of various modalities of sensation, all aspects of testing and finding threshold should be predetermined, controlled,

standardized, and quantitated, and should employ approaches that have been validated and for which normal values as percentiles and normal deviates are available. Today this degree of standardization can only be obtained using microprocessors and specially designed electromechanical devices that can reproducibly provide defined stimuli of known waveform and magnitude; that can be calibrated; and that, by employing appropriate and validated algorithms of testing and finding thresholds, can keep track of all stimuli and responses, compare results to normative values, and maintain permanent records. These records should be available for inspection to ensure that quality assurance standards have been maintained. Systems can be made sufficiently alike that results from different medical centers are comparable.

Subject or Patient Factors For the test to be valid, the patient must be cooperative, alert, and not on mind-altering drugs. He or she must be adequately instructed to know what sensation or other stimulus-response criterion is to be recognized, and the responses to be given. The test should not be done when it is known that the patient wants an unfavorable outcome (demonstrating worse sensation than is actually present), such as in preparation for medicolegal action. Conversely, a bias towards providing a favorable response (or outcome) is desirable because algorithms have been designed to deal with this bias.

Ambient Environment A designated sound- and sight-shielded examination laboratory, sufficient to exclude street and medical center sights and sounds, is needed. The subject or patient should be seated facing a neutral wall without a window and displaying the instruction panel and cueing devices. The examiner should communicate instructions by earphones or unobtrusively to allow the subject or patient to focus on the test. In our QST laboratory, the examiner sits to the side of and behind the patient so as to be able to have access to the computer terminal, the testing transducers, and the patient and to observe the part to be tested (to know that the transducer is appropriately placed and to recognize potential injury). The patient should not be able to monitor the screen of the computer. During vibration testing, earphones providing a continuous sinusoidal oscillation should be used to ensure that responses are not based on hearing the stimulus. In our laboratory, the patient sits on a dental chair, which rests on a hydraulic platform that can be raised and lowered; the chair can be moved forward or backward or rotated and reclined for convenient testing of finger, toe, forearm, thigh, shoulder, or face (see Fig. 42–2).

1069

Quantitative Sensation Testing

A

B

C

D

E

FIGURE 42–2 Assessing touch-pressure (A–C), vibratory (D), and warm or cold (E) detection thresholds using the ComputerAssisted Sensory Examination System III (CASE III). A dental chair on a hydraulic platform facilitates testing of the finger and toe. The CASE III system was described in detail in the second edition of this text. (From Dyck, P. J., Zimmerman, I. R., O’Brien, P. C., et al.: Introduction of automated systems to evaluate touch-pressure, vibration, and thermal cutaneous sensation in man. Ann. Neurol. 4:502, 1978, with permission.)

Instructions Standard written instructions, the same ones used to study controls, should be given to patients and should state the purpose of the test, what will be done, the nature of the

stimulus to be given, the duration of the test, the degree and duration of discomfort or pain (if any), what the subject must do, and the need for the subject to be alert and cooperative. The criteria for the response should be carefully explained. To keep instructions standard, instructions

1070

Quantitating Symptoms, Impairments, and Outcomes

6.

7.

8.

9.

FIGURE 42–3 The tone-arm and vibratory stimulus transducer used in the Computer-Assisted Sensory Examination System IV (CASE IV). (From Dyck, P. J., Karnes, J. L., Gillen, D. A., et al.: Comparison of algorithms of testing for use in automated evaluation of sensation. Neurology 40:1607, 1990, with permission.)

are read to the subject or patient with the patient reading along using sign cards. Questions are answered by standard neutral phrases to avoid influencing the responses in a biased direction. Patients who do not wish to be tested should be deferred until they do. Inquiry should be made about use of mind-altering drugs (i.e., narcotics, analgesics, sedatives, tranquilizers, and the like), and if necessary the test should be deferred to another time or a record should be made of what medication and what dose was taken and when. We here provide the standard printed instructions we read to patients just before they have vibration, cooling, and heat-pain testing using CASE IV. 4, 2, 1 Step Testing Algorithm with Null Stimuli (VDT) 1. This is a test of your ability to feel vibration sensation. 2. This test is not painful. 3. It usually takes 5 to 10 minutes. 4. The stimulus may feel like a vibration, buzzing, trembling, or a rumbling sensation. Some people cannot describe it, but they know a stimulus was given. 5. All you have to do is decide whether you felt a vibration stimulus during number “1”. You probably will feel the stimulator resting on your toe (or finger) at all times. I will ask you to decide whether you felt an addi-

tional vibrating or other mechanical stimulus during the number “1”. After the number “1” has disappeared, you should push “yes” if you felt the vibration or some other mechanical stimulus, or “no” if you did not feel it. When the pulse of vibration is small, it is felt best near the midpoint of the time of display of the number “1”. I will demonstrate this to you in a minute. Please get comfortable, relax your foot (or hand), and do your best. Once again, the object is to determine the smallest vibration or other mechanical stimulus pulse you can feel. Do you have any questions?

4, 2, 1 Step Testing Algorithm with Null Stimuli (CDT) 1. This is a test of your ability to feel the sensation of cooling. 2. The test is not painful. 3. It usually takes 5 to 10 minutes. 4. The stimulus may feel like “a pulse of cooling”, “a cool wetness”, or “a puff of cooling”. Some people cannot describe it but they know a stimulus was given. 5. All you have to do is decide whether you felt a cooling stimulus during the display of the number “1”. You probably will feel the stimulator resting on your foot (or hand) at all times. You may feel the cooling or warming of the thermode between stimuli as it adjusts its temperature to baseline conditions. I will ask you whether you felt a cooling stimulus during the display of the number “1”. 6. Here is a hint that may be helpful. The “pulse of cooling”, if it occurs, comes near the midpoint of the display of the number “1”. I will demonstrate this to you in a minute. 7. After the number “1” has disappeared, you should push “yes” if you felt the cooling or “no” if you did not feel it. 8. Please get comfortable, relax your foot (or hand), and do your best. Once again, the object is to determine the smallest degree of cooling you can feel. 9. Do you have any questions? Nonrepeating Ascending Algorithm with Null Stimuli (HPT) 1. In this test, small pulses of heating are given during the display of the number “1”. A yellow light precedes the display of the number “1” to alert you that a stimulus is about to be given. 2. Many of the pulses of heating are so small that they are not felt or are felt only as warm or hot, but without discomfort or pain. When you feel nothing or warm or hot but no discomfort or pain, your response should be 0. 3. If you feel some discomfort or pain in addition to warmth or heat, you must grade its severity from 1 to 10.

Quantitative Sensation Testing

1071

Table 42–1. Commercially Available Devices and Instruments for QST Hand-held Devices Touch-pressure sensation Monofilaments

Glass filaments Tuning fork 97⁄8  11⁄4  1⁄2 with counterweight 165 Hz Pinprick Semiquantitative QST “Electrical stimuli” The Neurometer

Thermal pulses Thermal Sensitivity Tester, N-E-2

Vibration Vibration II Vibrameter

Thermal and Vibration QST TSA II Neuro Sensory Analyzer

Thermal test (replaces Marstock Device)

Laboratory Automated and Quantitative QST CASE IV (vibratory, cooling, warming, and heat-pain)

Number 1 would correspond to the least discomfort or pain; number 10 would correspond to the most severe pain. 4. To repeat, following every stimulus event as indicated by the number “1”, choose 0 when you feel nothing or only warmth or heat. Choose a number between 1 and 10 if you experience any degree of discomfort or pain—number 1 is the mildest

Long-figer cotton wool, from local merchants Sensory Testing Systems 1815 Dallas Dr., Suite 11A Baton Rouge, LA 70806 FAX: 504-923-3670 Telephone: 1-888-289-9293 Not commercially available

Riverbank Laboratory, Inc. 1512 South Batavia Ave. Geneva, IL 60134 Stick pins, from local merchants

Neurotron, Inc. 1501 Sulgrove Ave., Suite 203 Baltimore, MD 21209 FAX: 410-664-0831 Telephone: 410-664-0800 Physitemp, Inc. 154 Huron Ave. Clifton, NJ 07013 Telephone: 201-779-5577 Physitemp, Inc. (as above) Somedic Products Sollentuna, SWEDEN 19205 Telephone: 46-835-6827 Medoc–Advanced Medical Systems 45 Ha’oren St., PO Box 423 Ramat Vishai, ISRAEL 30095 FAX: 972-4-983-0785 Telephone: 972-4-983-0751 Somedic Scales Box 194 Ho˝rby, SWEDEN Telephone: 46-415-16550 W. R. Medical Electronics, Inc. 123 North Second St. Stillwater, MN 55082 FAX: 651-439-9733 Telephone: 651-439-9733

degree of discomfort or pain, whereas 10 is the most severe. 5. Please note that we do not intend to hurt you, so we will be testing you only at the lower discomfort or pain levels. 6. Do you want me to repeat the instructions? Any questions? 7. Now I will demonstrate the test.

1072

Quantitating Symptoms, Impairments, and Outcomes

Comparison of QST Results among Different Devices and Systems Test results from patients should be compared to test results from a healthy subject cohort tested by the same system and testing approaches. Ideally, the healthy subject cohort should be randomly drawn from the community and free of neurologic disease or disease predisposing to neuropathy (e.g., diabetes mellitus). All ages and both sexes should be included. An approach to estimating percentiles has been described by O’Brien and Dyck.23 Although the magnitude of the threshold stimulus should be expressed as micrometers of displacement (for VDT) or change of temperature in degrees centigrade from baseline (for CDT/HPT), these measurements need transformation before they are used in statistical testing because sensation increases as an exponential function or in steps of JND.26 It is inappropriate, therefore, to express group results as mean or median (and range) values of physical units (e.g., m of displacement or ° C). If physical measurements were used, a one-step increase of sensation (e.g., from an insensitive part) in an old person would influence mean results more than a one-step change at much lower magnitudes in a young person. Two approaches might be used to deal with this problem. The first is to express results in JND units. The difficulty with using this approach is that the value of JND is not finally agreed on. The second approach, and the one we recommend, is to express results as percentiles (corrected for anatomic site, for modality of sensation, and for age, sex, and applicable anthropomorphic characteristics) and then to express percentiles as normal deviate values ( scores), which are then used in statistical analysis. Use of normal deviates also allows comparisons among different sensory testing modalities and systems (assuming that adequate percentile values have been determined) and comparison of QST results. Use of  scores also allows development of composite scores that may include QST results, chosen attributes of nerve conduction, and autonomic test results. For a condition such as peripheral neuropathy, which is the sum total of symptoms, impairments, and test abnormalities, a broadly based composite score probably reports overall severity of polyneuropathy much more adequately than any single marker.

Baseline Conditions of the Stimulus It is necessary to have standardized baseline conditions on which the stimulus is superimposed. In the case of vibratory stimuli in CASE IV testing, the tactator surface resting on the tissue has a standard shape, surface area, and composition (low thermal conductivity) and rests on the surface with a standard invariant initial load. In CASE IV, the tactator is suspended by a beam that allows up-anddown slow movements to occur but has a sufficient mass so that it will respond minimally to the rapid sinusoidal oscillations of the vibrating stimulator (see Fig. 42–3). When

testing for cooling, warming, or heat-pain using a thermode, the surface area contacting the skin must be known and constant, the thermode must be firmly strapped to the skin, the entire surface of the thermode should be in contact with skin and the degree of contact should remain unchanged during testing, and an offset constant temperature is maintained during testing of cooling (30° C) or warming/heat-pain (34° C) (see Fig. 42–2). In the case of direct electrical current testing, the size of the electrodes and their electrical contact with the skin should be standardized.

Ideal Stimuli and Algorithms of Testing 1. A sufficiently broad range of stimulus magnitudes should be provided so as to test the most sensitive anatomic sites of young people, insensitive sites of old people, and areas of sensation loss. It may not be possible to provide a stimulus sufficiently large to overcome insensitivity of a certain degree because providing such a large stimulus would be tissue damaging. 2. Inasmuch as is possible, testing should be of only a single modality of sensation (e.g., touch-pressure, vibration, cooling, or warming/heat-pain) or of one functional class of fibers. Actually it is not possible to deliver only one modality because, with vibration, cooling, or heat-pain, the tactator or thermode is in constant and static contact with the skin and deeper tissues, providing a constant tactile and thermal input. Therefore, the patient needs to be instructed that he or she must identify the presence or absence of the superimposed stimulus being tested. In the CASE IV system, we indicate an interval of time when the stimulus is presented. 3. The test should employ a sufficient number of real and null stimuli so that average values indicating sensation can be derived and to ensure that null stimuli are not taken as real stimuli (see subsequent section). 4. The test should be efficient; it should be possible to ascertain threshold of a modality of sensation at a single site in a relatively short period of time (2 to 12 minutes) to avoid patient drowsiness and boredom. 5. As part of the standard examination, the examiner should record biographic and demographic features, including date and time of day; anatomic site tested; patient height and weight; modality of sensation, stimulation, and algorithm events; and estimated threshold. The calculated percentile and normal deviate value (based on comparison to a normal database) should be automatically expressed and entered into the database. 6. The test should assess both hyposensitivity and hypersensitivity (for heat-pain). 7. As much as possible, the approach should be made simple.

Quantitative Sensation Testing

1073

Stimulus Waveform For VDT, we use a sinusoidal waveform at 125 Hz contained in an envelope without an abrupt onset or offset (Fig. 42–4). The magnitude of the stimulus envelope containing the sinusoidal oscillation is shaped so that the envelope has the characteristics of a charging and discharging capacitor. Thermal ramps or pulses can also be fashioned into linear or exponentially increasing or decreasing ramps, and diamond- or other-shaped pulses. In CASE IV testing of cooling or warming/heat-pain, we use pyramidal or trapezoid-shaped stimuli that have a rise rate of 4° C/s. It is the initial direction of temperature change from baseline that is perceived as warming/heat-pain (increasing temperature) or as cooling (decreasing temperature). The return of the temperature to baseline is not felt. The range of stimulus intensities is of considerable importance because it is necessary to be able to deliver smallmagnitude stimuli to sensitive parts of young people but also to deliver large-magnitude stimuli to insensitive parts of old people and to insensitive sites of patients with abnormal sensation. Whether ramps or steps (pulses) of stimulus intensities are used, there should be an orderly increase of the strength from small to large. Because the range of stimulus magnitudes needed may vary from as little as 0.25 m to several hundred micrometers of displacement (in the case of VDT), it is not efficient to have the stimulus increase at a slow linear rate (assuming that a slow rate of increase is needed so as not to overestimate threshold as a result of delayed reaction time). Fortunately, use of an exponential function (e.g., In of m of displacement) was found to be satisfactory and is probably close to the physiologic JND. As we discuss elsewhere, step testing (steps increasing exponentially from very small to large) is perhaps preferable to continuous linear ramp or linear step testing. The approach avoids the overestimate from long reaction times and allows interpolation of results between steps, and a broad range of stimulus magnitudes can be efficiently tested.

Null Stimuli Null stimuli play an important role in QST and should be used in algorithms of testing and estimating threshold. Informing the subject that null stimuli will be given conveys the important message that a judgment must be made as to whether an external stimulus was given, and thus a patient cannot automatically respond affirmatively to each stimulus event indicated by a light display or display of a number. The number of false-positive responses to null stimuli may be used as an indication of how well the subject is differentiating real from null stimuli. If the subject answers affirmatively to all or most null stimuli, it is evident that he or she is not making the necessary judgments for correctly identifying threshold. This failure may occur when the patient has not been ade-

FIGURE 42–4 The waveforms used in forced-choice testing (top) and linear ramps with null stimulus (bottom). Although sinusoidal waveforms at 125 Hz are used for all three algorithms, the rate and shape of the increase of intensity and the method of testing differ. In two-alternative forced-choice testing, the subject is asked to choose the interval associated with the display of the numbers 1 and then 2 that had a vibratory stimulus. The rules of testing and finding threshold are described in the text. In linear ramp testing with null stimuli (middle), the intensity of vibratory stimuli increases to the point at which the response key is depressed. After the initial stimulus, seven stimuli and four null stimuli are presented as described in the text. A(1), A(2), A(3), and A(8) indicate the first, second, third, and eighth appearance thresholds in a typical sequence. Null stimuli are shown as “null.” In Bekesy testing with null stimuli (bottom), the rate of increase of the vibratory stimulus follows an exponential function based on our previous study of just noticeable difference. A(1)D, A(2)D, A(3)D, and A(8)D refer to the first, second, third, and eighth “appearance” and “disappearance” thresholds. Null stimuli are indicated by the word “null.” (From Dyck, P. J., Karnes, J. L., Gillen, D. A., et al.: Comparison of algorithms of testing for use in automated evaluation of sensation. Neurology 40:1607, 1990, with permission.)

quately instructed or is inattentive. In this situation, further instruction and testing (or another algorithm, e.g., the forced-choice algorithm) may need to be used. It should be appreciated that in some hyperesthetic neuropathies,

1074

Quantitating Symptoms, Impairments, and Outcomes

the patient may experience sensory symptoms (e.g., paresthesia) at the time of presentation of null stimuli that can be misinterpreted as an externally applied stimulus (e.g., a vibratory stimulus). Similar miscues can come from thermal or pain experiences. The second use of null stimuli is to equalize response criteria among patients. Some subjects may be prone to respond positively to each cue that a stimulus event is to be given. Reacting in the opposite direction, other subjects may respond positively only when the external stimulus is unmistakable. Quite different thresholds would be obtained simply because different response criteria were used. A two-alternative forced-choice algorithm can be used to equalize this difference in response criteria. The subject is told that pairs of stimulus events will be given, indicated by the visual display of the number 1, then the number 2. A stimulus will be given during only one of the pairs of stimulus events—the other event will be the null stimulus. The allocation of stimuli to interval 1 or 2 is by chance. The subject must choose interval 1 or 2 as having the stimulus. If he or she is uncertain which interval had

the stimulus, the subject must still choose the interval that was most likely to have the real stimulus (a choice is forced). Note that the subject cannot use words like “guess” and is unable to provide other response criteria choices. Assuming complete lack of sensation, the correct interval would be chosen 50% of the time in forced-choice testing; correct responses would be given 100% of the time if all stimuli were felt. The 75% correct response level is the midpoint of not feeling and feeling (threshold) in forced-choice testing.

Calibration of Stimulus Magnitude To ensure that stimuli are standard and the same for all systems of a particular type, it is essential that they be fabricated so that they meet high industry standards of safety and provide standard and reproducible results. For vibratory stimuli, the displacement of the tactator, at least for the largest stimulus steps, can be directly measured using laser triangulation with the tactator resting on a standard high-purity silicone sheet of known thickness and firmness (durometer setting). Figure 42–5 shows the measurements

FIGURE 42–5 A lateral view of the vibration testing assembly and transducer is shown in the left cower corner. For calibration, the balance arm is raised or lowered so that the bubble of the level is central, ensuring that the balance arm is in a horizontal position. The apex of the tactator rests on a high-purity silicone sheet. Standard stimulus steps of testing are then given and displacement is read by the Keyence sensor from a defined point on the upright member connected to the tactator disk. The Keyence sensor measures displacement using laser triangulation. Because the apex of the tactator traces a small distance of an arc, we then calculate the chord (the linear distance between two points of an arc) of the displacement. It is possible using this approach to provide a standard array of stimulus displacements from small (step 1) to large (step 25) using this calibration approach and adjusting the voltage to the motor to get standard steps of stimulus magnitude.

Quantitative Sensation Testing

1075

FIGURE 42–6 The theoretical (light lines) and recorded (dark tracings) heating (top) and cooling (bottom) pulses used in CASE IV and CASE V QST. Tracings are shown for steps 15, 20, 22, 24, and 25. The change in temperature with time occurs from left to right. In the case of heating pulses, baseline temperature is to be controlled at 34° C; then there is a linear rise at 4° C/s to an apex and return to baseline, also at 4° C. For steps 21 to 24, temperature is maintained at a peak 48° C for 1.5, 5, and 10 seconds; for step 25, 49° C is maintained for 10 seconds. In the case of cooling stimuli, baseline temperature is controlled at 30° C and the rate of temperature change is as described for heating. Steps 22 to 24 plateau at 10° C for 1.5, 5, and 10 seconds; step 25 plateaus at 9° C for 10 seconds. The heavy line is the actual temperature tracing from the copper-constantan thermocouple in contact with the face of the thermode and overlaid by Styrofoam. The tracing of the recorded waveform is offset to the left of the theoretical so that the recorded tracing and the theoretical waveform can be visualized. Because it is the initial rise or fall of temperature that the subject senses, we are especially concerned that the rate of rise or fall is as desired, and that the maximal temperature is achieved and maintained for the correct period of time—the return to normal is not critical provided baseline conditions are achieved. Recognizing that there is inertia in a measuring system such as this as a result of mass of the measuring apparatus, it appears that we can achieve a very close approximation of the desired thermal pulses. Calibration should occur in the factory and periodically thereafter.

and calculations we use to calibrate the vibratory stimulus of CASE IV and V systems. Because the apex of the tactator disk traces an arc, it is the chord (the linear distance between two points of an arc) that we calibrate. If piezoelectric devices are used for calibration, they need to be individually calibrated by an approach such as that described in Figure 42–5. For thermal pulses, a copper-constantan wire thermocouple of small diameter is calibrated for temperatures between 0° and 50° C in a calorimeter against a highpurity platinum resister. The wire thermocouple may then be placed across the thermode, insulating it from ambient temperature using a Styrofoam block and then monitoring the thermal ramps or pulses to be used in testing. In

Figure 42–6, we show theoretical and measured thermal pulses from CASE IV.

Algorithm of Testing and Finding Threshold Various algorithms of testing and finding threshold are used in QST and designated as the method of limits, forced-choice, staircase, or other. Describing an algorithm in these terms does not in itself ensure that a particular standard or adequate algorithm is being used. Many characteristics of an algorithm (the stimulus itself, the range of stimulus magnitudes available, adequacy of calibration, whether null stimuli are used, the order of testing of real and null stimuli, the number of

1076

Quantitating Symptoms, Impairments, and Outcomes

turnarounds used from insensitivity to sensitivity and vice versa), the time between intervals of testing, how threshold is estimated and recorded, and how the test is administered determine how good an algorithm is and whether it provides meaningful results. To validate algorithms of testing and finding threshold, we suggest the following steps: 1. Specify the rules of testing in writing based on intuition and previous publications. 2. Use computer simulations of the rules of testing so as to recognize steps that give spurious answers simply by chance, allowing correction and validation of the rules. 3. Test a series of healthy and diseased patients and visually inspect the events of the stimulus responses, looking for modifications that would make the algorithm more efficient or accurate, and then modify the algorithm and again test it in groups of patients. 4. Compare the improved and shortened algorithm with the best known algorithm or with other neuropathic end points. 5. Describe and program the algorithm so that it is used without modification. 6. Develop normative results. The simplest algorithm of testing may be referred to as ramp (continuous or increasing in steps) testing to the appearance of sensation (appearance threshold). The stimulus magnitude may be increased linearly, exponentially, or by another mathematical function. The converse is to decrease the stimulus to a level at which it is not felt (disappearance threshold). Again, the decrease may be continuous or in steps linearly, exponentially, or by another function. Quite different results are obtained depending on kind and rate of increase/ decrease of stimuli. Testing with ramps to appearance, disappearance, or the mean of appearance and disappearance thresholds may not provide accurate estimates of threshold. First, because null stimuli are not included, it is difficult to ascertain that responses are from feeling the stimuli rather than from another response criterion. Second, both appearance and disappearance thresholds tend to be overestimates of threshold (as a result of slow and variable reaction times), and the overestimate is greater with rapid and exponential ramps and in thoughtful, deliberate persons. This phenomenon is illustrated for vibration in Figure 42–7 and for thermal sensation in Figure 42–8. Use of the mean appearance and disappearance thresholds does not overcome the problem because the overestimate for disappearance threshold probably is greater than that for appearance threshold. We have considerable antipathy to devices that use potentiometers, which are rotated manually to the point at which sensation is experienced. Because the lowest

FIGURE 42–7 Vibratory detection threshold (VDT) for the index finger in eight subjects from a forced-choice algorithm (triangles) and “appearance thresholds” from linear ramps (circles). The VDTs based on testing with the forced-choice algorithm were performed on separate days. The first threshold value is displayed to the left of the second value. The thresholds from linear ramps at 0.08, 0.17, 0.83, 4.15, and 16.6 m/s are joined by lines from left to right for each of the eight subjects. As described in the text, the thresholds with slow linear ramps (0.08, 0.17, and possibly even 0.83 m/s) are comparable to those obtained from forced-choice testing. Slow ramps are not speedy enough for testing insensitive sites, making their use impractical in automated systems. Use of fast ramps clearly overestimated threshold. This overestimation probably results from the subject’s inability to stop the stimulus quickly enough because of reaction time. (From Dyck, P. J., Karnes, J. L., Gillen, D. A., et al.: Comparison of algorithms of testing for use in automated evaluation of sensation. Neurology 40:1607, 1990, with permission.)

voltage and rate of increase are unspecified, variable results may ensue as a result of uncontrolled stimulus presentation. A variety of algorithms can be designed if stimuli are given in steps of increasing and decreasing magnitudes and null stimuli are interspersed. In CASE IV, we test sensation using exponential steps of magnitude, as done in Bekesy’s testing of hearing acuity. This makes good sense because sensation increases as an exponential function with increasing stimulus magnitudes. We have proposed, and for some purposes use, a specifically designed two-alternative forced-choice algorithm for CASE IV testing. Pairs of stimulus events are indicated by the visual display of the number 1 and then display of the number 2. A stimulus is presented during one of intervals 1 or 2. The choice of whether it is in interval 1 or 2 is by chance. The other interval is baseline condition without the stimulus (the null stimulus). A series of pairs of a stimulus and a null stimulus are given at different magnitudes by predetermined rules, as illustrated for our forced-choice algorithm in Figure 42–9. It is called a forced-choice algorithm because, for each pair of stimulus and null stimulus

Quantitative Sensation Testing

1077

FIGURE 42–8 In eight healthy subjects, warm (WDT) and cool (CDT) detection thresholds were obtained for the dorsum of the hand. Different ramp rates using a linear ramp with null stimuli algorithms were employed. Ramp rates of 1° C/s gave approximately similar values and were comparable to values obtained using forced-choice testing (not shown); when ramp rates exceeding 1° C/s, thresholds began to rise, probably owing to reaction time. We consider the forcedchoice algorithm to perform better than the linear or Bekesy algorithm in three respects: (1) it is faster than when slow ramps are employed in linear or Bekesy testing, (2) bias is minimized, and (3) threshold is more likely to be formed when stimulus intensities have to be large.

events, the subject must choose which interval was most likely to have the stimulus. In this approach, the response criteria are forced to be the same for the cautious person who does not want to overcall sensation and the less cautious person. The rules of stimulation and a typical stimulus level and response sequence are shown in Figure 42–9. Various other algorithms of testing have been developed using stepping and null stimuli that are efficient and produce accurate results. These include 4, 2, 1 stepping with null stimuli (useful for testing VDT and CDT) and nonrepeating ascending stepping with null stimuli (for HPT) (as described below).

TOUCH-PRESSURE SENSATION Tactile sensation provides important information to humans about the external environment. The receptors, nerve fibers, and integrative centers are involved in the

expansion of disparate sensations: an insect has landed and is crawling over the skin, a thread needs to be removed from the mouth, or there is desired (or unwanted) physical contact. These receptors and systems provide information about tactile contact, form, shape, surface features, texture, and directionality. Patients with loss of tactile sensation of their hands cannot recognize objects by feel, texture, or shape. In walking, patients tend to be clumsy. Tactile sensation appears to function in concert with joint position and motion and vibration sensation (see below). Loss of oral tactile sensation results in patients losing food and pills in their mouth. In earlier editions of this textbook, we outlined the various approaches developed to assess touch-pressure sensation. Among the available approaches to test tactile sensation using natural stimuli are long-threaded cotton wool, nylon, or glass monofilaments or rods; sandpaper of different grades of coarseness; and a graduated inclined ridge above a plane surface. Frey introduced tail hairs of

1078

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 42–9 A typical sequence of testing with the forced-choice algorithm. The abbreviated rules by which the stimulus increases or decreases and the method of calculating the threshold are given in the text. F  failure; JND  just noticeable difference; s  success. The numbers in parentheses indicate the turnarounds. (From Dyck, P. J., Karnes, J. L., Gillen, D. A., et al.: Comparison of algorithms of testing for use in automated evaluation of sensation. Neurology 40:1607, 1990, with permission.)

horses of various lengths attached to handles. Hairs of different caliber and length were chosen to provide different loads before they buckle. We tested monofilament nylon and glass filaments and rods of different caliber and length for this purpose. To prevent the variability of force during the events of bending and buckling from vertical application of the filament to the surface of the skin, we apply the monofilaments as cantilevers deforming the monofilament from a horizontal position to the surface of the skin. In such an application, buckling does not occur. The free end of the cantilever has a vertical tactator affixed to it. It is usually assumed that Frey’s hairs and nylon or glass monofilaments induce only tactile sensation. As early investigators1,20,28 demonstrated, sensations of cooling, warmth, and pain may also be induced by such tactile stimulations. It is informative to test sensitive sites (the dorsal surface of the distal index finger will do) with the commercially available 10-g Semmes Weinstein monofilament. Slight degrees of burning discomfort may be induced from application of the monofilament at certain spots. From this observation, we infer that these 10-g monofilaments can, at certain spots, stimulate more than A fibers. In the Touch-Pressure System III we developed (described in earlier editions of this textbook), touch-pressure sensation could be assessed at regularly spaced grid points over the surface of skin. A vertical stylus was lowered onto the skin and maintained at a static load. A pressure waveform of different magnitudes could then be superimposed to estimate threshold at each grid point. In healthy subjects, threshold was greater for toe than finger and for men

FIGURE 42–10 Top, Linear regression lines relating touch-pressure thresholds (pressure) of sensitive points to age for four sites—forehead, lip, finger, and toe—as described in the text. Bottom, Linear regression lines relating mean displacement (pressure converted to displacement) to age for the same four sites, as described in the text.

than women and with older age (Figs. 42–10 and 42–11). In neuropathy, it often was not possible to detect a threshold at maximal stimuli. Therefore, it was necessary to express abnormality as number of insensitive points and as threshold of sensitive points. Although this system performed well, it was never commercially developed for reasons of complexity and expense. The vibration system that was developed concurrently was less complex and provided information about dysfunction of A sensory fibers.

FIGURE 42–11 Comparison of touch-pressure thresholds of this study using Touch-Pressure System III (unbroken line) versus a study with Frey’s hair (dashed line) by Weinstein.29 The figure is based on Weinstein’s drawing, which is a modification of earlier drawings.

Quantitative Sensation Testing

The practice of declaring a foot to be sensitive (7 of 10 positive responses testing the plantar surface of each toe and metatarsal head of a foot) or insensitive ( 7 of 10 positive responses) probably provides some useful and predictive information about susceptibility for development of plantar ulcer in diabetes and leprosy and for recognizing returning sensory function following peripheral nerve surgery. Strictly speaking, however, the approach is too simple—touch-pressure is known to vary with site, age, and several anthropomorphic features. If the approach is to be used, several different strengths of stimulation and responses specific for site, age, sex, and anthropomorphic factors should be developed. It would also be prudent to use null stimuli.

TWO-POINT DISCRIMINATION The ability to recognize spatially separated points as two rather than as one is used in clinical neurology. Assuming that the patient does not have central nervous system pathology, two-point discrimination conceivably could be a useful test of altered touch-pressure sensation from disorders of peripheral nerves. Apart from handheld devices, systems to do this are not available, and normative values would need to be generated. It should also be appreciated that, with stiff epidermis (as occurs in diabetes mellitus and with callus), points close to each other cannot be tested because they may move as one.

VIBRATION SENSATION Decreased ability to recognize rapid oscillations of a tuning fork (vibration) placed on a bony prominence (at the knee or the lateral malleolus at the ankle) was found to be an early sign of such neurologic diseases as tabes dorsalis and diabetic polyneuropathy. Today, the sensation produced by low-frequency mechanical oscillations is referred to as flutter for frequencies of approximately 50 to 100 Hz and as vibration or pallesthesia at frequencies greater than 100 Hz. The receptor mediating pallesthesia is thought to be the pacinian corpuscle, information from which is mediated by A sensory fibers, large spinal ganglion afferent neuron somas, and their extensions into the posterior columns of the spinal cord. As outlined in previous editions and in other chapters of this textbook, assessing vibration sensation is useful for detection, characterization, and following the course of polyneuropathy. Whereas Meissner’s corpuscles provide information about punctate mechanosensation, the role of vibration sensation in physiologic function is not as readily conceptualized. Conceivably, in predatory mammals, it is useful for recognition of predators or of prey. In humans, in addition to touch-pressure, information about flutter or vibration provides clues about

1079

mechanical impact, directionality, and quality (e.g., the pianist percussing a key) to a varying degree, or recognition of the type of oscillations set up when a foot strikes an object in walking. We assume that flutter and vibration sensations are not vestigial. It is instructive to test vibration sensation of the toe of a person with an artificial limb. Vibratory stimuli, transmitted through the artificial limb, are felt in the thigh stump. The tuning fork is the device most commonly used to assess vibration sensation. As conventionally used, it is firmly applied to the great toe, lateral malleolus, or index finger. It is known that long bones transmit physical stimuli over long distances. The tuning fork we prefer is 97⁄8 inches long  11⁄4 inches 1⁄2 inches with counterweights and has a frequency of 165 Hz (Riverbank Laboratories, Inc., Geneva, IL). Semiquantitative tuning forks have been developed to provide somewhat standardized degrees of stimulus intensity. It should be appreciated, however, that they maintain the desired magnitude of amplitude only momentarily, making it difficult to develop adequate algorithms of testing. The bioesthesiometer is a modification of a scalp vibrator to assess vibration sensation. A steel cantilever rigidly fixed at one end with the free end suspended over the pole of an electromagnet provides oscillations, which can be increased or decreased using a potentiometer. A tactator applied to the toe, finger, lateral malleolus, or other anatomic part has been affixed to the free end of the cantilever. The bioesthesiometer has three major shortcomings that, at least in part, can be remedied. The stimulus is subject to damping; therefore, a constant but appropriate load should always be used. Because a potentiometer is used to increase the voltage, the voltage increase should be set at a low rate and should be specified and used. Third, an algorithm of testing should be specified and used. Stating that the method of limits was used is saying almost nothing (see previous section). Defining a single voltage as abnormal for all sites, ages, and heights is simply incorrect. Vibration thresholds vary with site, age, height, and perhaps other characteristics. The device called the vibrometer is a handheld device that has a readout of the load exerted during testing (so as to normalize this) and provides quantified stimuli. The algorithm of testing and the expression of abnormality are left to the user. To assess vibration sensation, we developed CASE IV, which is described in more detail in the next section.15

Computer-Assisted Sensory Examination System IV The CASE IV system consists of (1) a personal computer, hard and floppy disks, and a keyboard for entry of biographical data, selection of testing format, setting test conditions, administration of the algorithm of testing,

1080

Quantitating Symptoms, Impairments, and Outcomes

estimation of threshold, comparison of threshold to that of controls, and generation of test results; (2) a video screen for display of program instructions, testing options, error messages, and results; (3) an electronic controller (containing power supplies, fail-safe devices and microprocessors, and circuitry) to provide discrete levels of stimulus intensity for forced-choice testing and for linear and exponential increases of stimulus magnitude for the other two algorithms of testing; (4) a printer for biographic information and results; (5) a visual cueing device to display the “get ready” yellow light and “test in progress” green light, and for display of the numbers 1 and 2 for alternative forced-choice testing; (6) a patient response key; and (7) a vibratory transducer and a thermode. The software for operation of the system was written in the C language.

VDT as the Measure of Mechanoreceptor Function As outlined in an earlier section, touch-pressure or tactile sensation is mediated by different receptors (RA and SA I or SA II) than is vibration sensation (PC). Knowing that RA receptors have small fields and sharp edges (not graded slopes as for PC receptors), it makes physiologic sense to assess touch-pressure using punctate stimuli at grid points. Such grid point testing is not needed for assessing VDT because receptors are more deeply situated and have large fields and the edges have graded slopes. Because of these physiologic properties of the pacinian corpuscles, it should not be necessary to assess VDT at grid points; rather, a single probe (disk) should suffice. Thus it is simpler to test vibration sensation than touchpressure sensation. For purposes of studying patients with neuropathy, does assessment of VDT alone provide a sufficiently reliable assessment of mechanoreceptor function so that touchpressure need not be assessed? To answer this question, we assessed (using linear regressions) the association of the normal deviate of touch-pressure detection threshold (using CASE III) and the normal deviate of VDT (using CASE IV) in 40 patients with and without neuropathy who had had both tests performed on the great toe. The results of the two tests were positively associated (r  .65; P .001). It appears that the association is close enough that VDT would serve as an indicator of large-diameter sensory nerve dysfunction. If tactile sensation is to be discriminated from vibratory sensation, both would need to be assessed.

VDT Using CASE IV Vibratory Transducer The stimulator making contact with the skin is a flat disk 9 mm in diameter with beveled edges and is coated with

Teflon (so as not to induce a thermal heat-loss stimulus) (see Fig. 42–3). The disk is mounted by shaft to an optical scanning motor (G 302 series; General Scanning, Watertown, MA) with vertical displacement proportional to current. The motor and probe are affixed to the front end of a balance arm with sufficient mass that it does not follow the rapid sinusoidal 125-Hz oscillations but will move up or down with the slow motion of the anatomic part being tested. The stimulating disk rests on the finger, toe, or other part to be tested, with a load of 30 g. Calibration of Stimuli In an earlier edition of this textbook, we provided the measured displacement of the tactator at stimulus steps 1 through 25. A ruby-red laser beam was reflected from a mirror attached to the shaft of the tactator onto an opposite wall and the amount of vertical displacement at the apex of the tactator was calculated using triangulation. Without changing any of the stimulus steps 1 through 25, we have now recalibrated stimulus magnitudes using improved methodologies: (1) a standard artificial surface on which the tactator rests (a silicone sheet of known hardness); (2) new instrumentation (Keyence, Inc.) for laser triangulation; and (3) measurement of the chord (the linear distance between two points of an arc) at the tip of the tactator. The details of these calculations are outlined in Figure 42–5. Forced-Choice Algorithm of Testing In a previous section we described the general characteristics of the forced-choice algorithm; here we describe the specific rules (see Fig. 42–9). Twenty-five discrete levels of vibratory stimuli are used. The magnitude of the 25 levels of stimulus intensity is distributed from smallest to largest intensity by using an exponential function based on our previous study of JND. Each level of stimulation therefore approximates one JND step. To prevent an abrupt onset and offset, we shape the stimulus envelope. For a given JND level, the rise and fall of the vibratory stimulus at 125 Hz increases from baseline to the specified level with a waveform that has the characteristics of a charging and discharging capacitor. The rules of testing are based on our previously published algorithm and on recent modifications.24 In brief, a vibratory stimulus at a given level of intensity is presented during only one of a pair of stimulus events indicated to the patient by serial display of the numbers 1 and then 2. Whether the stimulus is given during display of 1 or 2 is by chance. By depressing a response key, the subject indicates the interval that contains the vibratory stimulus. Success or failure at this level results in subsequent stimuli being given at lesser or greater stimulus intensities, respectively. Success is defined as correct identification of the stimulus interval in a pair of stimulus events six of seven times, or five of the first six times, with the seventh time incorrect and the

Quantitative Sensation Testing

eighth and ninth times correct. Testing is begun at level 13, with subsequent stepping to larger or smaller levels in increments of 4, 4, 2, 2, 1, 1, 1, and 1 preceding turnarounds 1 through 8, respectively. This sequence was altered so that single stepping to smaller levels occurred after the first failure. The threshold was the median of the last six turnaround values. A typical sequence is shown in Figure 42–9. 4, 2, 1 Stepping Algorithm with Null Stimuli This algorithm was used for systems employing step testing and for evaluation of vibration, cooling, and warming sensations. We consider it unsuitable for HPT testing for reasons outlined under Warmth or Heat-Pain Sensation below. This algorithm was developed to reduce the time of testing typical in a forced-choice testing. Thresholds from using this algorithm were not significantly different from these based on forced-choice testing and could be obtained in a much shorter period of time. Twenty-five steps, from small (step 1  less than 1 m of displacement) to large (in CASE V, step 25  approximately 315 m of displacement), and null stimuli (no sinusoidal oscillations) are available for testing. The rules for testing were intuitive and based on previous studies using ramp and step testing, computer simulations, and trials in healthy subjects and patients with neuropathy. Testing begins at an intermediate step level (step 13). A null stimulus is given in either interval 1 or 2 (by random

FIGURE 42–12 A typical and satisfactory stimulusresponse pattern of vibration threshold of the great toe in a 60-year-old woman using CASE IV and the 4, 2, 1 stepping algorithm with null stimuli. Note that the subject did not report having felt any of the five null stimuli and that the algorithm advanced efficiently from not feeling (steps 13 and 17) to feeling (step 21). The mean of the turnaround (indicated by the broken line) was at step 18.33. For this woman, this value corresponds to a threshold at the 95th percentile. Note that, in estimating threshold, this person’s age, body surface area, and body mass index were taken into account.

1081

choice). If a positive response is given to the first challenge, step 13 is retested. Thereafter we step up (when the stimulus was not felt at step 13 and thereafter) or down (when the stimulus was felt) by increments of 4, 2, and 1 steps. Thus, assuming that step 13 was felt, the next step tested would be step 9. If the stimulus is not felt at 9, level 9 is declared a turnaround level and the next level tested is level 7. If step 13 had not been felt, the next step tested would be step 17. If step 17 is not felt, step 21 is tested. Assuming step 21 is felt, step 21 is declared the turnaround step and the next step tested is 19. To summarize, stepping begins in steps of 4 from level 13. After the first turnaround, stepping is in steps of 2; after the second turnaround, it is in steps of 1. Of the 20 stimulus events, five are randomly distributed null stimuli. The interval of stepping of 4 steps is altered to 2 steps for steps 21 (or larger) or step 5 (or smaller). Three consecutive failures at step 25 results in termination of the test, declaring insensitivity. Three consecutive successes at step 1 results in termination and the patient is declared supersensitive or hypersensitive. If two or more of the null stimuli are identified as being felt, the test is declared invalid. The patient is reinstructed and retested. If the test is invalid on three successive trials, forced-choice testing should be done. Threshold is the mean of the turnarounds based on single stepping; for example, in the sequence shown in Figure 42–12, the last six turnarounds are at 19, 18, 19, 18, 19, and 17, with a mean value of

1082

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 42–13 This stimulus-response curve is typical of a person who is insensitive for vibration sensation in the foot. The patient did not report two null stimuli as sensitive. Additionally, seven stimuli of different magnitudes (including three at step 25) were not felt. His threshold is greater than the 25th step, which is above the 99th percentile considering site, age, body surface area, and body mass index.

18.333. Figure 42–13 illustrates a person who is insensitive to step 25 testing. In a comparative study, 25 healthy subjects and 25 patients with mild diabetic neuropathy were tested using the forced-choice format and the 4, 2, 1 stepping algorithm with null stimuli. The time required to do the test (after instructions were given) was much shorter for the stepping algorithm (2.2  0.2 min [mean  SD]) than for forcedchoice testing (sometimes beyond 30 minutes). However, the results on average were not significantly different from those with forced-choice testing.

Quality Assurance of VDT In order to obtain valid and reproducible results, it is necessary that the patient be cooperative, not on mind-altering drugs, not distracted, and properly instructed and that the test results be adequately quality controlled. With regard to the latter, we insist that the practice test results be within two steps of the final test results, that two or more of the null stimuli were not identified as felt, and that the stimulus-response characteristics are without problem patterns. Problem patterns are illustrated in Figure 42–14. When problem patterns occur, it is best to repeat the test after reinstruction.

VDT in Health Vibratory threshold in health varies with site, age, and perhaps weight. Figure 42–15 plots threshold as step

levels by age for the index finger and great toe. Figure 42–16 provides the same data but showing threshold as normal deviates. Note that in old age and for the great toe the stimulus may not be of sufficient magnitude to test to a given abnormal percentile value (e.g., the 95th or 97.5th).

Use of VDT VDT is a clinically meaningful indicator of A sensory dysfunction. In disease, only raised threshold (hyposensitivity), and not lowered threshold (hypersensitivity), has been shown to be abnormal. As we noted in the first section of this chapter, hyposensitivity of VDT can represent selective abnormality of A receptors and fibers or, in concert with abnormality of other modalities, pan-modality sensory abnormality. Our studies have found VDT to be almost as sensitive in detecting abnormality of large afferent fibers as nerve conduction. It is not as objective a test as nerve conduction because responses can be willed; therefore, the test is not useful in situations in which the patient wishes to demonstrate an adverse outcome for whatever reason (e.g., medicolegal action following injury). However, the test results are perhaps more meaningful than attributes of nerve conduction because sensation loss itself is being tested. The vibration test has found a special place in epidemiologic surveys and controlled clinical trials. In conditions in which the long and large sensory neurons are especially vulnerable (diabetic polyneuropathy, uremic neuropathy,

Quantitative Sensation Testing

1083

FIGURE 42–14 The four stimulus-response patterns shown here are problematic, and these QSTs should be repeated. In the upper frame, a threshold level appears to have been found at about 16.8 steps. However, five subsequent increasingly stronger stimuli were not felt. The record should not be accepted—the person may have become drowsy. In the second from the top stimulus response, a reliable threshold was not established, and the test should be repeated. In the lower two frames, threshold also has not been achieved. None of these records should be accepted as satisfactory studies; all should be repeated.

and such cancer therapy as vinca alkaloids, paclitaxel, and platinum), it might be used alone or preferably with other test results to develop composite scores that can then be used to set criteria for assessing the occurrence and severity of neuropathy. Each of the test results is expressed as a percentile and as a normal deviate () score, which can be added to give a composite score. With the availability of values from a normal subject population, it is then possible to set an abnormal limit (e.g., the 95th or 97.5th composite score) to determine abnormality. We have found that the distribution of VDT within the range of normal values can provide evidence of a

functional abnormality even before values exceed a defined level of abnormality. In our diabetic cohort, we found that there was a considerable excess of VDT values between the 50th and 97.5th percentile values as compared to the numbers between the 2.5th and 50th percentile values. By chi-square analysis, diabetics were significantly different from healthy subjects who had essentially the same numbers above and below the 50th percentile. An essentially similar pattern was also found for nerve conductions. This result clearly implies that low-grade abnormalities occur in a higher number of diabetics than obtained from standard tests.

1084

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 42–15 Vibratory detection threshold of the dorsum of the index finder (top) and of the dorsum of the great toe (bottom) in healthy persons 18 to 74 years old in the Rochester Diabetic Neuropathy Study of Normal Subjects (RDNS-NS) cohort. The values are expressed as JND steps of displacement. Quite a number of patients had supersensitive values (shown as values at 0), especially for the finger. For both the finger and toe, thresholds increase with age.

THERMAL SENSATION Unique receptors and fibers mediate cooling, warming, and heat-pain sensations. Whereas cool and pain receptors are densely and broadly distributed over the surface of the skin, warm receptors appear to be less densely and evenly distributed over the surface of the body. A pulse of cooling is experienced as a splash of coolness with a definite onset and offset. Receptors are innervated by A fibers with a conduction velocity of approximately 8 to 15 m/s. By contrast, a pulse of warming has a less definite onset and offset, and there are regions of skin (especially the foot and leg of old people) where the first experience from a heat pulse is a small burning sensation (i.e., heat-pain). Warming and heat-pain sensations appear to be mediated by drC unmyelinated fibers.

FIGURE 42–16 The same threshold values as shown in Figure 42-15 for the RDNS-NS with results plotted as normal deviates relative to age.

It is generally accepted that pain is experienced when skin temperature exceeds 44.6° C, the temperature that, if maintained for a long time, induces tissue injury. Thus, to test only warm sensation, it is important not to give stimuli in the nociceptive range of temperature. The regions of lowest thresholds for warm sensation are the dorsal hand, volar forearm, and face. Heat-pain sensation, in contrast to warmth sensation, is well represented over the surface of the body. In clinics and hospitals, cool, warm, and heat-pain sensation is seldom adequately tested because testing systems are still not broadly available. Simple devices are flasks or metal thermodes into which fluids at given temperatures are poured. Because the temperature is not controlled and maintained, stimuli of known magnitude cannot reproducibly be given. We introduced Minnesota Thermal Disks, which depend on the different heat transfer properties of copper, steel, glass, and polyvinyl chloride—all at ambient temperature.5 Between stimuli, the thermode rests on a copper block to maintain its temperature at room temperature. With the availability of thermoelectric units (thermocouples in parallel), it is possible to give defined cooling or warming pulses.

Quantitative Sensation Testing

1085

Cooling Detection Threshold Using CASE IV In experiments on healthy subjects, we found that pyramidal stimuli from baseline temperatures (30° C), lowering temperature at 4° C/s to various levels, induced a sensation of a drop or splash of cool water. The return of the thermode temperature to baseline was not felt as warming. The short delay and clearly defined onset and offset are in keeping with the known innervation of cool receptors by A sensory fibers. CDT is readily evaluated by our 10-cm2 thermode in CASE IV testing (Figs. 42–17 and 42–18; see also Fig. 42–6). As for VDT, as described earlier in this chapter, 25 steps of stimulus intensity testing are available. Baseline temperature is 30° C. The leading ramp of the cooling stimulus decreases at a rate of 4° C/s in exponential increasing JND steps of magnitude from baseline. At step 21, a temperature of 10° C is achieved. At step 22, the peak at 10° C is maintained for 1.5 seconds, at step 23 it is maintained for 5 seconds, and at step 24 for 10 seconds. At step 25, 9° C is achieved and this is maintained for 10 seconds. We compared five waveforms and algorithms of testing of CDT on the forearm of eight healthy subjects. Compared were 1. Linear ramps of cooling from baseline to appearance threshold interspersed with null stimuli at rates of 0.025°, 0.05°, 0.10°, 0.5°, 1°, and 3° C/s 2. Exponential ramps of cooling from baseline, interspersed with null stimuli to appearance threshold at rates of 0.5, 1, or 2 JND/s 3. Exponential ramps of cooling from baseline: mean of appearance and disappearance thresholds 4. A 4, 2, 1 stepping algorithm with null stimuli as described for VDT 5. Two-alternative forced-choice testing as described for VDT

FIGURE 42–17 The thermode used in CASE IV and CASE V testing of cool, warm, and heat-pain thresholds. The design of the thermode has been previously described in detail.15

FIGURE 42–18 The warm (top) and cool (bottom) stimuli employed in forcedchoice testing using CASE IV with ramps of 4° C/s. For testing warm detection threshold (WDT) (top), the stimulus is shown for level 11 and in period 2; for cool detection threshold (CDT) (bottom), the stimulus is shown at level 23 and in period 1.

These studies showed that thresholds were spuriously high when linear or exponential ramps and rapid increases of stimulus amplitudes were used—especially with ramps of 1° C/s or greater. By contrast, low rates of thermal charge resulted in values comparable to those obtained with twoalternative forced-choice testing (see Fig. 42–8). The mean of appearance and disappearance threshold also did not perform well; some subjects simply were unable to identify the point of disappearance so that a level was not identified. Both the 4, 2, 1 stepping with null stimuli and the twoalternative forced-choice algorithms performed well, providing comparable values. Using steps of stimulus intensity allowed the subject to take the needed time to decide whether a stimulus of cooling had, or had not, been given. Reaction times among subjects could vary without a deleterious effect on threshold. Thermal threshold tended to be lower with large than with small thermode areas.14 Figure 42–19 plots the threshold of the hand in steps of JND by age. Figure 42–20 provides the same data for the foot. In our normative study, age was the main determinant, although for one site (lateral leg) body surface area was also a determinant. CDT may be used to recognize hyposensitivity dysfunction of A fibers. Only hyposensitivity (not hypersensitivity) is recognized as an abnormality of cool sensation. Assessment of cool sensation has been reported to be useful in demonstrating abnormality in small fiber neuropathy.10,11,22,30 Because small sensory fibers are frequently affected in metabolic, toxic, deficiency, and inflammatory immune

1086

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 42–19 Plots of cool detection threshold (as JND steps of stimulus magnitude) of the dorsum of the hand using CASE IV in subjects from our normative study (RDNS-NS). The only variable influencing threshold was age.

conditions, it seems likely that assessment of systematic epidemiologic cohorts using CDT will be found to be useful for detection and characterization. It is assumed that such assessments will therefore also be useful in therapeutic trials for these conditions.

Warm Detection Threshold To reiterate, warm should be tested separately from cool because receptors and fiber classes are different. Cooling is mediated by A (small sensory myelinated fibers), whereas warmth is mediated by unmyelinated C (drC) fibers, and their density and distribution are different

FIGURE 42–20 Plots of cooling detection threshold (as JNS steps of stimulus magnitude) of the dorsal foot of healthy subjects from our normative study (RDNS-NS).

(cooling receptors appear to be more evenly and densely distributed). Also, warmth and heat-pain, although both conveyed by unmyelinated fibers, should also be tested separately. To ensure that heat-pain is not being tested, the temperature at the level of the receptor should not reach 44.6° C; to be safe, it probably should not exceed 40° C. For CASE IV testing, we do not have normative values for warmth threshold at different sites. We have found that the first subjective sensation experienced from pulses of heat on the foot and leg of healthy subjects of middle or old age may be that of a feeling of heat and burning (i.e., heat-pain). The most sensitive regions of warm sensation in humans are the dorsal hand, volar forearm, and face. It is therefore unclear whether assessing warm sensation is useful in recognition or characterization of disease. If it is to be used, it should be appreciated that warmth may not be detectable in 10-cm2 areas of lower limbs in healthy persons.

Warmth or Heat-Pain Sensation Several investigators have utilized pulses of thermal heating to study “thermal sensation, to the point of the experience of warmth or heat-pain.” For CASE IV testing, we have not developed normal values for such warmth or heat thresholds. Several approaches have been developed for stimulating and judging severity of pain from heat stimuli. Extensively studied was the technique developed by Hardy and colleagues using a radiant heat lamp source focused by a lens and shuttered so that various durations of exposure could be given.17 The radiant energy was expressed in millicalories per square centimeter (mcal/cm2). In order to develop consistency, the authors defined a response criterion—a “pricking pain” threshold. The approach was extensively studied but never found wide clinical applicability. Today some investigators use laser beams as minute thermal pulses. In the CASE IV system, we employ a thermoelectric unit to create heat pulses. The algorithm used is the ascending nonrepeating heat-pain algorithm with null stimuli. The heating pulses have been calibrated so that steps 18 or 19 through 25 induce a short-lived pulse of heat-pain that is transitory. With the highest steps of testing, the pulse induces erythema but not blistering. The algorithm that we developed attempts to minimize sensitization and adaptation. After receiving a nociceptive pulse above pain threshold, patients report an aftersensation remaining for a varying period of time, akin to the experience from sunburn. This is assumed to be a sensitization phenomenon. It is also appreciated that, when several large-magnitude stimuli in the nociceptive range are given, the response of the receptors might change (the next nociceptive pulse of equal magnitude may be experienced as less strong [adaptation]). In the revised CASE V ascending nonrepeating heat-pain algorithm, we have programmed in delays to minimize sensitization and adaptation.

Quantitative Sensation Testing

In the standard heat-pain test we developed using CASE IV, we use thermal pulses in steps 1 through 25. The baseline temperature is 34° C. The rise time is 4° C/s, followed by return to baseline. The steps increase exponentially, achieving a maximal pyramidal peak at 48° C at step 21. For steps 22, 23, and 24, the temperature of the thermode rises to 48° C and is maintained for 1.5, 5, and 10 seconds, respectively. At step 25, the temperature rises to 49° C and is maintained for 10 seconds. The thermal rise rate is 4° C/s. In conduct of the test, “no sensation,” “nothing,” “warmth,” and extremely “hot” are all graded as 0. With any discomfort or pain (in addition to warmth or heat), the subject grades discomfort or pain from 1 (the least) to 10 (the most discomfort or pain). Testing is begun at a level unequivocally below the HPT (i.e., step 13) and then is increased by steps of 2 until step 21 or a 1 or higher pain response is given. Then it is increased by steps of 1 until a response of 5 or higher is attained or step 25 has been tested. If a response of 1 or higher is reported for any of steps 15, 17, 19, 21, or 23, the step immediately smaller, if not previously tested, is then tested and then single stepping to successively higher steps is used thereafter. In each of three successive stimuli, one is randomly selected as a null stimulus. If the patient gives an answer between 1 and 10 for a null stimulus, the test is stopped and the patient is reinstructed. The stimulus-response pattern can be graphed into two components, no pain response at baseline (0) and the pain response (Fig. 42–21). We fitted a quadratic regression equation to the heat-pain response and on this regression line identified the HP 5, HP 0.5, and HP 5-0.5 slope. HP 0.5 is the HPT, HP 5 is an intermediate severity of heatpain response, and HP 5-0.5 is the stimulus-response slope

FIGURE 42–21 A typical stimulus-response pattern during performance of the heatpain test. The stimulus-response points can be divided into two groups, steps of testing without a pain response (steps 13, 15, 17, 19, and 22) and steps at which pain was experienced (steps 20, 21, 23, and 24). A quadratic regression is fitted to the pain responses and then HP 5, HP 0.5, and HP 5-0.5 are determined (see text).

1087

between HP 0.5 and HP 5. The details of the graphing approach used are outlined elsewhere.4 Three types of abnormality of the heat-pain response are possible in neuropathic disease. Hyposensitivity is indicated by a raised step levels of HP 0.5 and HP 5.0 and by a decreased slope of HP 5–0.5. A lowered step level or an increased slope might indicate hypersensitivity (hyperalgesia), shown graphically in Figure 42–22. The heat-pain test is informative in several respects. First, it provides information about nociception loss not evaluated by nerve conduction tests or other large-diameter fiber tests. Loss of heat-pain sensation is of value diagnostically (implying involvement of drC fibers) but also is of importance clinically because abnormality implies loss of protective sensation so that the anatomic site may be excessively injured by use, exposure to heat, and other factors. Along with decreased function of large sensory fibers, loss of heatpain sensation might indicate pan-modality sensory loss. Perhaps the most common pattern of abnormality that the heat-pain test identifies is hyperalgesia. Hyperalgesia is recognized by lower than normal HP 0.5 or HP 5 levels, by a steeper than normal HP 5-0.5 slope, or by combinations of lower thresholds and steeper slopes. The heightened heat-pain response is all the more impressive when it is coupled with a high VDT or CDT. Hyperalgesia occurs in a variety of neuropathies.

Estimating Percentile Abnormality Test results are often expressed simply as normal or abnormal. For several reasons, it is preferable to express results as percentiles and normal deviates. The latter approach allows

1088

Quantitating Symptoms, Impairments, and Outcomes

42–16, 42–19, and 42–20 plot QST values for the finger and toe from consenting, healthy subjects randomly selected from the Olmsted County, Minnesota, population. It should be recognized that factors other than age are used in setting percentile values. These additional factors cannot be taken into account in plots of the kind shown here.

SENSITIVITY, REPRODUCIBILITY, AND CONCORDANCE OF L-QST RESULTS WITH NEUROPATHIC IMPAIRMENT

FIGURE 42–22 A superimposition of the plotted values of three diabetic patients illustrating a normal response, one with hyperalgesia, and one with hypoalgesia. Note that it is possible to estimate the HP 0.5 and the HP 5.0 for the first two but only the HP 0.5 for the third. In the third case, it is possible (and useful for percentile estimation) to estimate the HP 2 (the patient has reported a pain severity of 2 at JND step 25). (From Dyck, P. J., Zimmerman, I. R., Johnson, D. M., et al.: A standard test of heat-pain responses using CASE IV. J. Neurol. Sci. 136:54, 1996, with permission.)

a better scaling of abnormality, permits the user to select a particular level of abnormality, and allows combination of different test results (e.g., in development of composite scores).12 This topic is discussed in detail in Chapter 36. Table 42-2 identifies the various biographic and anthropomorphic factors that influence L-QST. Figures 42–15, Table 42–2. Order of Physical Variables* Significantly† Associated in Stepwise Multiple Linear Regression Analysis with QST Results and Their Model Coefficients End Points

VDT Left foot Left hand Left lateral leg CDT Left foot Left hand Left lateral leg

First

Second

Third

Age, 0.063 Age, 0.029 Age, 0.020

BSA, 2.408 BSA, 0.997

BMI,0.046†

Age, 0.057 Age, 0.026 Age, 0.063

BSA, 2.698

CDT  cooling detecting threshold; VDT  vibratory detection threshold. *The physical variables assessed were age, gender, height, weight, body mass index (BMI), and body surface area (BSA). † All variables listed in the table are p 0.001 unless they are identified by a dagger; then their p value is 0.001 p 0.01. Modified from Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects. Neurology 45:1115, 1995.

Laboratory and clinical test results may be evaluated by the criteria of sensitivity, specificity, accuracy, reproducibility, representativeness, and concordance with other measures of neuropathy (symptoms, impairments, and quality-of-life outcomes), and perhaps by other measures (e.g., applicability, simplicity, and cost). Information of this kind is available for some QST results obtained with CASE IV in cohorts of diabetic patients with varying degrees of diabetic sensory polyneuropathy. Test-retest reproducibility as measured with the intraclass correlation coefficient for VDTs and CDTs was greater than 0.9—a value as good as or better than those available for most attributes of nerve conduction.9 The intraclass correlation coefficient for warm detection threshold was approximately 0.8. In diabetic cohorts, the sensitivity of nerve conduction abnormality has been somewhat higher than it is for QST abnormality, but the latter abnormalities are thought to be more meaningful.8

QST IN MEDICAL PRACTICE QST is useful for (1) developing sensation testing approaches that neurologists might then use in their medical practice, (2) establishing whether there is hyposensitivity (sensation loss) or hypersensitivity or both hyposensitivity and hypersensitivity by test modality in patients, (3) determining the anatomic distribution of sensation abnormality, (4) inferring the anatomic-pathologic alterations of fibers, (5) inferring from serial examination whether worsening or improvement is occurring, and (6) assessing degree of impairment. We take the position that QST results are not standalone tests but are especially helpful when they are done in association with clinical, nerve conduction, needle EMG, and other laboratory and pathology tests. Thus it is useful to know whether a neurologic disorder is or is not associated with sensory loss or hyperalgesia, and the pattern and distribution of abnormalities. QST offers a unique approach to quantitate a person’s ability to recognize quantified and graded sensory stimuli

Quantitative Sensation Testing

of vibration, cooling, warming, and heat-pain at defined sites of the body as compared to a reference cohort of healthy subjects. It is also useful to detect hypersensitivity and enhanced response to heat-pain stimuli. Recognition and characterization of thermal and mechanical hyperalgesia is an important role of QST because these phenomena are common and disabling. The main symptom of many neuropathies is burning feet, a symptom of small fiber sensory neuropathy commonly accompanied by mechanical and thermal hyperalgesia. QST is important in recognizing a level of neuropathy that predisposes patients to potential foot injury (plantar ulcers and neurogenic arthropathy). In foot and diabetes clinics, simple strategies are being pursued to recognize sensation loss and acral mutilation, which might lead to initiation of a surveillance and preventative program. QST may help following regeneration after nerve injury. In addition, QST may be helpful to neurologists (e.g., better normative values, use of improved algorithms including use of null stimuli) to improve their testing approaches and results.

1089

sia). Sometimes hyperalgesia is the only objective indicator of sensory neuropathy.

Distribution of Sensation Abnormality The distribution of altered sensation (by symptoms or examination) provides the neurologist with clues regarding the underlying cause of a mononeuropathy, plexopathy, radiculopathy, polyneuropathy, myelopathy, or other neurologic disorder. Generally, this determination is made by elicitation of symptoms and by clinical examination, but for some purposes anatomic localization can be verified by QST. Examples are given in Figure 42–23 (Friedreich’s ataxia), Figure 42–24 (Tangier disease), and Figure 42–25 (following restorative surgery).

Inferring the Anatomic-Pathologic Abnormalities of Fibers In Chapter 39 of this textbook, we correlate selective loss of modalities of sensation caused by disease with

Developing Sensation-Testing Approaches Useful in Clinical Practice Clinicians can learn from studies used to improve QST how to improve their clinical sensation testing skills. Here we list a few simple insights. Simple instruments may give quantifiable results (i.e., graded nylon monofilaments; thermal disks made of copper, steel, glass, and polyvinyl chloride). Sensation thresholds are quite different for anatomic site, age, sex (in some cases), height, and body surface area or body mass index. Null stimuli can readily be introduced so as to recognize responses not elicited by externally applied stimuli. QST can be used to study the factors that influence physicians’ lack of accuracy in sensation testing. Burns et al. found that the inaccuracy of fellows in a neuromuscular practice was due to failing to take age, height, and body surface area—the same factors known to influence thresholds in health—sufficiently into account.3

Hyposensitivity and Hypersensitivity It is insufficiently appreciated that both hyposensitivity and hypersensitivity are manifestations of neurologic disease of sensory neurons. Recognizing hyposensitivity is useful for recognition of neurologic disease. Hypersensitivity, alone or in combination with hyposensitivity, is also quite common in acute neurologic disease. Hyperalgesia may be recognized by the nature of the symptoms (excessive tenderness to touch, draft of air, contact with clothing, etc.) and by mechanical hypersensitivity (allodynia) or thermal hypersensitivity (thermal hyperalge-

FIGURE 42–23 The thresholds of sensation at the sites shown in a 15-year-old patient with Friedreich’s ataxia. On the face, thresholds fall within normal values ( P95). For the hand and foot, touchpressure and vibration sensations are abnormal but thermal cooling sensation is normal.

1090

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 42–25 Sensation thresholds after restorative surgery for a degloved thumb injury.

FIGURE 42–24 The distinctive pattern of sensation abnormality encountered in adult-onset Tangier disease.

decreased or absent potentials of the compound nerve action potential of excised sural nerve in vitro with selective loss of a diameter class of fibers. To illustrate, occasional patients with transthyretin amyloidosis have a selective deficit of pain and thermal sensation in a lumbosacral distribution. As compared to the A potential (which is reduced in amplitude from normal), A and C potentials are essentially absent. On morphometric study of nerve, small myelinated and unmyelinated fibers have degenerated, leaving denervated Schwann cell tubes and clusters. The converse to the pattern described for amyloidosis is found in spinocerebellar degeneration. These patients develop mechanoreceptor loss in their feet and legs but thermal and pain sensation is preserved. The A potential of the excised sural nerve is reduced in amplitude, whereas A and C fiber potentials are not reduced in size—in fact, the A potential may actually be larger than normal. In the diameter histogram, the large myelinated fiber peak is much smaller than normal, the small myelinated fiber peak is displaced to a smaller diameter category but is larger than normal, and the C fiber peak is normal. In the second and third editions of this text, we provided examples of the characteristic QST abnormalities found in various neuromuscular diseases.

Inferring Improvement or Worsening Serial evaluation of QST at standard sites may be used to infer whether the disease process is worsening or improving. To do this, the same anatomic site can be tested sequentially. Alternatively, several sites (one with borderline abnormality and another without abnormality) might also be tested.

Assessing Impairment QST results, especially when compared to results in controls, may be used to provide quantitative data on degree of sensation impairment.

QST IN RESEARCH QST has been used for surveys of populations at risk because of medicinal (cancer therapy) or industrial toxic exposure. QST also may be used for screening of populations at risk for developing sensation loss as a result of disease (e.g., screening for neuropathy among patients with diabetes mellitus); an infection (e.g., human immunodeficiency virus, leprosy, Lyme borreliosis); or occupational activity. Investigators may wish to use a very sensitive and rapid test, but not necessarily a very specific test, leaving the eventual determination for a later more specific test.

Quantitative Sensation Testing

QST is being extensively employed in controlled trials of therapy to prevent or ameliorate diabetic polyneuropathy. CASE IV results have been used in trials of aldose reductase inhibitors, nerve growth factors, -lipoic acid, protein kinase C  inhibitors, and other interventions. In these trials, these measures have been combined with attributes of nerve conduction and with clinical measures to provide a more representative score of neuropathic impairment than can be achieved with one single measure alone. We are using QST to prospectively survey populationbased glucose-intolerant or diabetes mellitus patients. We are assessing QST using CASE IV in examination of the feet in population-based cohort studies of diabetes mellitus in persons of northern European extraction, Mdewakanton Dakota Native Americans, Hispanic Americans, and Somalian Americans. In normal subject cohorts from these groups, we develop normative values for sensation modality, anatomic site, age, sex, and applicable anthropomorphic characteristics. We express results as percentiles and normal deviates. Results from patient cohorts can be expressed as a percentile or normal deviate score for a test in a patient, or normal deviates of several tests can be combined into composite scores. These approaches allow us to describe what number (percentage) of the cohort is hyposensitive for vibration, cooling, or heat-pain sensation or develop a composite score. The course of neuropathy can be followed. The association of

FIGURE 42–26 Test-retest intraclass correlation coefficient and confidence intervals (95%) for compound muscle action potential (CMAP), motor nerve conduction velocity (MNCV), motor nerve distal latency (MNDL), motor nerve F waves (MNFW), sensory nerve action potential (SNAP), sensory nerve conduction velocity (SNCV), and sensory nerve distal latency (SNDL) for nerves 1 through 5, as identified on the graph. By the criteria of a high correlation coefficient and small confidence intervals among attributes of nerve conduction, CMAP and SNAP performed best. (From Dyck, P. J., Kratz, K.M., Lehman, K. A., et al.: The Rochester Diabetic Neuropathy Study: design, criteria for types of neuropathy, selection bias, and reproducibility of neuropathic tests. Neurology 41:799, 1991, with permission.)

1091

these scores with clinical symptoms and staged severity can be evaluated. Assuming that risk covariates have also been assessed, multivariate risk factors for complications can be determined. Association can be assessed based on cross-sectional and longitudinal data. How sensitive, reproducible, representative, and meaningful is QST for these purposes? In the Rochester Diabetic Neuropathy Study, we evaluated a Neuropathy Impairment Score (nerve conductions, quantitative sensation tests, and autonomic tests) in stepwise discriminate analysis in predicting polyneuropathy. In the four models tested, VDT came in second one time, third two times, and fourth one time.8 Reproducibility of test (QST and other tests) results was assessed using the intraclass correlation coefficients in 20 diabetic patients without and with varying severities of neuropathy. Figure 42–26 shows the test-retest intraclass correlation coefficients and confidence intervals (95%) for compound muscle action potentials (CMAPs), motor nerve conduction velocity (MNCV), motor nerve distal latency, and motor nerve F wave. Figure 42–27 shows test-retest intraclass correlation coefficients and confidence intervals (95%) for VDT, CDT, and warming detection threshold. Both measures of VDT performed comparably to the MNCV and better than CMAP measurement. Among modalities of sensation tested, warm detection threshold performed well also, but not as well as VDT or CDT.

1092

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 42–27 Test-retest intraclass correlation coefficients and confidence intervals (95%) for vibratory (VDT1, just noticeable difference [JND] levels of testing; and VDT2, normal deviate), cooling (CDT; JND levels of testing), and warming (WDT) detection thresholds. (From Dyck, P. J., Kratz, K. M., Lehman, K. A., et al.: The Rochester Diabetic Neuropathy Study: design, criteria for types of neuropathy, selection bias, and reproducibility of neuropathic tests. Neurology 41:799, 1991, with permission.)

REFERENCES 1. Blix, M.: Experimenteia bidrag till losning of fragan om hudnervernas specifiko energi. Ups. Lakarefor. Forhandlingar 43:427, 1882. 2. Burgess, P. R., and Perl, E. R.: Cutaneous mechanoreceptors and nociceptors. In Iggo, A. (ed.): Handbook of Sensory Physiology: Somatosensory System. Berlin, Springer-Verlag, p. 29, 1973. 3. Burns, T. M., Taly, A., O’Brien, P. C., and Dyck, P. J.: Clinical versus quantitative vibration assessment: improving clinical performance. J. Peripher. Nerv. Syst. 7:112, 2002. 4. Dyck, P., Zimmerman, I. R., Johnson, D. M., et al.: A standard test of heat-pain responses using CASE IV. J. Neurol. Sci. 136:54, 1996. 5. Dyck, P. J., Curtis, D. J., Bushek, W., and Offord, K.: Description of “Minnesota Thermal Disks” and normal values of cutaneous thermal discrimination in man. Neurology 24:325, 1974. 6. Dyck, P. J., Dyck, P. J. B., Kennedy, W. R., et al.: Limitations of quantitative sensory testing when patients are biased toward a bad outcome. Neurology 50:1213, 1998. 7. Dyck, P. J., Johnson, W. J., Lambert, E. H., et al.: Comparison of symptoms, chemistry, and nerve function to assess adequacy of hemodialysis. Neurology 29:1361, 1979. 8. Dyck, P. J., Karnes, J. L., O’Brien, P. C., et al.: Reassessment of tests and criteria for diagnosis and staged severity. Neurology 42:1164, 1992. 9. Dyck, P. J., Kratz, K. M., Lehman, K. A., et al.: The Rochester Diabetic Neuropathy Study: design, criteria for types of neuropathy, selection bias, and reproducibility of neuropathic tests. Neurology 41:799, 1991.

10. Dyck, P. J., and Lambert, E. H.: Dissociated sensation in amyloidosis: compound action potential, quantitative histologic and teased fiber, and electron microscopic studies of sural nerve biopsies. Arch. Neurol. 20:490, 1969. 11. Dyck, P. J., and O’Brien, P. C.: Quantitative sensation testing in small-diameter sensory fiber neuropathy. Muscle Nerve 26:595, 2002. 12. Dyck, P. J., O’Brien, P. C., Litchy, W. J., et al.: Use of percentiles and normal deviates to express nerve conduction and other test abnormalities. Muscle Nerve 24:307, 2001. 13. Dyck, P. J., O’Brien, P. C., Oviatt, K. F., et al.: Prednisone improves chronic inflammatory demyelinating polyradiculoneuropathy more than no treatment. Ann. Neurol. 11:136, 1982. 14. Dyck, P. J., Zimmerman, I., Gillen, D. A., et al.: Cool, warm, and heat-pain detection of receptors: testing methods and inferences about anatomic distribution of receptors. Neurology 43:1500, 1993. 15. Dyck, P. J., Zimmerman, I. R., O’Brien, P. C., et al.: Introduction of automated systems to evaluate touch-pressure, vibration, and thermal cutaneous sensation in man. Ann. Neurol. 4:502, 1978. 16. Erlanger, J., and Gasser, H. S.: Electrical Signs of Nervous Activity. Philadelphia, University of Pennsylvania Press, 1937. 17. Hardy, J. D., Wolff, H. G., and Goodell, H.: Pain Sensations and Reactions. New York, Hafner Publishing Company, 1967. 18. Johansson, R. S.: Tactile sensibility in the human hand: receptive field characteristics of mechanoreceptive units in the glabrous skin area. J. Physiol. (Lond.) 281:101, 1978. 19. Knibestaeol, M.: Stimulus-response functions of rapidly adapting mechanoreceptors in the human glabrous skin area. J. Physiol. (Lond.) 232:427, 1973. 20. Meissner, G.: Beitrage sur Kenntnis der Anatomie and Physiologie der Haut. Leipzig, Leopold Voss, p. 47, 1853; and Untersuchungen uber den Tostsuin. Z. Rat. Med. 7:92, 1859. (Cited in Winkelmann R. K.: Nerve Endings in Normal and Pathologic Skin. Springfield, IL, Charles C Thomas, 1960.) 21. Mokri, B., Ohnishi, A., and Dyck, P. J.: Disulfiram neuropathy. Neurology 31:730, 1981. 22. Novak, V., Freimer, M. L., Kissel, J. T., et al.: Autonomic impairment in painful neuropathy. Neurology 56:861, 2001. 23. O’Brien, P. C., and Dyck, P. J.: Procedures for setting normal values. Neurology 45:17, 1995. 24. O’Brien, P. C., Dyck, P. J., and Kosanke, J. L.: A computer evaluation of quantitative algorithms for measuring detection thresholds of cutaneous sensation. In Munsat, T. L. (ed.): Quantification of Neurologic Deficit. Boston, Butterworths, p. 197, 1989. 25. Service, F. J., Daube, J. R., O’Brien, P. C., and Dyck, P. J.: Effect of artificial pancreas treatment on peripheral nerve function in diabetes. Neurology 31:1375, 1981. 26. Stevens, S. S.: Discrimination of quantitative differences in stimuli in man. In De Reuck, A. V. S., and Knight, J. (eds.): Ciba Foundation Symposium on Touch, Heat and Pain. Boston, Little, Brown, p. 3, 1966. 27. Valbo, A. B., Hagbarth, K. E., Torebjork, H. E., and Wallin, B. G.: Somatosensory, proprioceptive and sympathetic activity in human peripheral nerves. Physiol. Rev. 59:919, 1979.

Quantitative Sensation Testing 28. von Frey, M.: Untersuchungen uber die Sinnesfunktionen der menschlichen Haut. Abh. Sachs. Ges. (Akad.) Wiss. 40:175, 1896. 29. Weinstein, S.: Intensive and extensive aspects of tactile sensitivity as a function of body part, sex and laterality.

1093

In Kenshialo, D (ed.): The Skin Senses (Proceedings of the First International Symposium on the Skin Senses). Springfield, IL, Charles C Thomas, p. 195, 1968. 30. Zimmermann, M.: Pathobiology of neuropathic pain. Eur. J. Pharmacol. 429:23, 2001.

43 Quantitative Muscle Strength Assessment KENTON R. KAUFMAN

Clinical Examination Noninstrumented Measurement of Muscle Strength

Instrumented Strength Testing Electromyography Direct Muscle Force Measurement

Our functional independence depends on our ability to move about in the constantly changing conditions of our environment. However, bones and joints themselves cannot produce movement. The muscular system is needed for this function. Accordingly, muscle strength tests represent one of the most often applied approaches to evaluate a person’s ability for functional movement. The attractiveness of muscle strength tests has been based on their obvious validity and relative simplicity. A number of different methods can be used for assessing muscle strength.

CLINICAL EXAMINATION Manual muscle testing is the most common noninstrumented method for measuring muscle strength. The examiner assesses the strength of individual muscle groups by applying a counterforce to the limb to overcome the patient’s voluntary muscle contraction. This method was introduced by H. Plummer and W. Sheldon at the Mayo Clinic (personal communication, PJ Dyck) and by Lovett.61 The most often used clinic system for manual muscle testing is the six-level scale (Table 43–1), introduced by the British Medical Research Council during the Second World War to scale recovery of nerve function after injury. This scoring system ranges from 5 (normal strength) to 0 (no evidence of contractility). (The scaling is the same as used by Lovett,61 but the number

Intramuscular Pressure Measurement Summary

of the ordinal scales is reversed.) Muscle grade is determined by the ability of the subject to move voluntarily against gravity and to resist a force applied by an examiner. 11,96 Use of either of these techniques allows an experienced examiner to rapidly determine the distribution and severity of weakness over a large number of muscle groups. This method for grading strength has stood the test of time as being robust and clinically useful. However, the scale uses gravity as a benchmark against which strength is tested. Clearly, the weight of the limb and the length of the lever arm over which muscles operate differ across various joints. The proportion of maximum strength required to overcome gravity is markedly different between muscle groups. Therefore, manual muscle testing has limited value, particularly in the lower extremity (Fig. 43–1). A normal muscle (grade 5) ranges between 53% of nonparalytic normal strength for the knee and 80% for the ankle. The earliest display of weakness (grade 4) represents 40% of normal strength.88 A grade 3 muscle is approximately 15% of normal strength. During normal walking, muscles function at a 3 ⫹ level.77 This effort, averaging approximately 25% of normal strength, allows adequate reserves so that fatigue is avoided.76 However, people whose muscles only have fair strength (grade 3) will have no endurance or reserve because they must function at 100% effort level. Thus manual muscle strength grading can be misleading. Moreover, this scale is nonlinear and ordinal. This means that a change of scale from, say, 4 to 3 does not imply the same loss of strength as a change from 1095

1096

Quantitating Symptoms, Impairments, and Outcomes

Table 43–1. Scoring System for Manual Muscle Testing Muscle Grades

Description

5–Normal

Complete range of motion against gravity with full resistance Complete range of motion against gravity with some resistance Complete range of motion against gravity Complete range of motion with gravity eliminated Evidence of slight contractility; no joint motion No evidence of contractility

4–Good 3–Fair 2–Poor 1–Trace 0–Zero

These tests are thought to test lower extremity muscle strength in a functional capacity because the joint moment required to perform activities such as sit-tostand, stair ascent, and stair descent is considerable.4,47,101 However, noninstrumented testing fails to predict the ability to walk because patients with muscle weakness can modify muscle action timing to avoid threatening postures and obtain protective alignment during stance.76 Also, they find subtle ways to advance the limb in swing by postural substitution.76 Hence, although strength testing is critical, noninstrumented muscle testing has serious limitations.

INSTRUMENTED STRENGTH TESTING Strength (% of normal)

100 80 Ankle Hip Knee

60 40 20 0 1

2

3-

3

3+

4

5

Muscle grade

FIGURE 43–1 Relative value of a manual muscle test for the lower limb extensor muscles (percent of true normal). (From Beasley, W. C.: Quantitative muscle testing: principles and applications to research and clinical services. Arch. Phys. Med. Rehabil. 42:398, 1961, with permission.)

3 to 2.7,68 A different grading approach developed at the Mayo Clinic and utilized in the Neuropathy Impairment Score is described in Chapter 40.

NONINSTRUMENTED MEASUREMENT OF MUSCLE STRENGTH Several other noninstrumented methods have also been used for quantifying the strength of individuals with muscular weakness. These other noninstrumented methods do not actually measure muscle strength, but rather provide a nonspecific indication of muscle capacity to perform an activity of daily living. The most commonly selected tasks are sit-to-stand19,25,36 and stair climbing.2,72

Instrumented strength testing is a more accurate indication of actual strength status and is needed to define true functional capability. The use of an instrumented device to measure muscle force yields a quantitative value for a subject’s strength. This quantitative assessment has the advantage over manual muscle testing in that it is a continuous scale for which normal values can be determined.3,6,11 This scale also is equally accurate and sensitive over its entire range. A wide range of devices have been used. Instrumented methods include handheld dynamometers, 17,18,97 load cells, 31 handgrip dynamometers, 91 and isokinetic dynamometers.5,94 Instrumented strength testing is a more accurate indication of actual strength status and is needed to define true capability. Instrumented measurement has been shown to be more sensitive to differences in muscle strength than noninstrumented manual measurement.1,10,16,85 Handheld dynamometers are held by the examiner, who then applies force to the patient being tested. The examiner must be strong enough to withstand the forces generated by the patient, otherwise the measurement tests the examiner’s strength rather than that of the patient. Load cells are fixed in position and only measure isometric strength. Handgrip dynamometers only measure handgrip strength and, thus, are limited in terms of applications. The most common strength-testing device is an isokinetic dynamometer, whose popularity may be attributed to the ease with which it provides information. Isokinetic testing, however, can only provide generalizations about muscle function. It cannot be used to comment on the function of an individual muscle. Also, the location of peak joint moment does not necessarily correspond to the position of maximum muscular output for a muscle.58 In general, instrumented methods are better than noninstrumented methods for providing information

Quantitative Muscle Strength Assessment

1097

about muscular strength. However, these methods fail to provide detailed information about individual muscles. Electromyography is commonly used to provide quantification of individual muscle function.

to changes in muscle force.60,75 Because of these factors, the relationship between muscle force and EMG is not known for dynamic conditions.

ELECTROMYOGRAPHY

DIRECT MUSCLE FORCE MEASUREMENT

Electromyography has been used to assess muscle function. Considerable progress has been made in quantifying electromyographic activity.14,15 However, the electromyogram (EMG) provides a quantitative measurement of muscle tension only under isometric conditions.13,27,28,59,63,66,67,70,100,102 Inman et al.45 first observed changes in myoelectric signal amplitudes that corresponded to variations in muscle load. Since then, much work has been done to determine the degree to which EMG signals and muscle loads are related.12,20,40,43,51,69 Nonetheless, the EMG signal is a measure of the bioelectric events that occur in conjunction with contraction of muscle fibers. Thus it is a phenomenon related to the initiation of muscle contraction rather than an effect of the muscle’s mechanical action. Problems occur in dynamic situations when using electromyographic activity as a measure of a muscle’s functional capability. The dynamic force produced by a muscle is not proportional to the degree of muscular activity. Other factors may affect muscle force, such as a change of muscle length, change of the muscle contraction velocity, the rate and type of muscle contraction, joint position, and muscle fatigue. Several authors agree that the integrated EMG value is not affected by a length change of the muscle at maximal effort,53,87 whereas others report a contrary behavior.62,81 Barnes8 reported a decreasing value for the integrated EMG or maximal effort with increasing contraction velocity. In contrast, Rosentswieg and Hinson81 stated that isokinetic contractions elicited significantly greater integrated EMG values than either isotonic or isometric contractions. Others reported that the integrated EMG value at maximal effort is not affected by a change of velocity.42,54,82 Muscle force varies with the type of muscle action. The relative strength of eccentric contractions was reported to be equal to that of isometric contractions89 or greater by 10% to 20%.95 Concentric muscle contraction results in approximately 20% less force that isometric contraction.73 Strength varies nonlinearly with the speed of effort.73 During maximum effort at any speed, the EMG is relatively constant even though the force differs.76 As a joint moves, both muscle fiber length and the muscle moment arm change. Whereas the intensity of the EMG remains constant, the strength of the quadriceps muscle declines by 50% when the joint position is changed from 50 to 10 degrees of flexion.37 Estimation of muscle force from the myoelectric signal changes with fatigue, which is unrelated

In vivo internal muscle forces are difficult to measure directly. A tendon buckle transducer was proposed by Salmons83 to measure force from a single muscle or from a muscle group with a common tendon. This technique is limited to muscles with relatively long tendons that can be exposed surgically. Data from direct muscle force measurements in humans are very limited because of the invasive nature of the experiment. Direct muscle force measurements have been obtained from the human Achilles tendon30,32,34,35,52,55 and the patellar tendon.29 The results of these studies must be regarded cautiously, because the measurements obtained were from a group of muscles rather than individual muscles. Buckle transducers have also been implanted in the wrist flexors during carpal tunnel surgery.22 However, the usefulness of this technique for measuring muscle force during movement is severely limited.

INTRAMUSCULAR PRESSURE MEASUREMENT It is desirable to find an alternative mechanical method to measure muscle force that would be less invasive. The electromyographic signal does not furnish adequate information to assess the tension produced by a muscle, because tension reflects the sum of both active contraction and passive stretch. Measurement of intramuscular pressure (IMP) is a conceivable solution. Since the first reported measurement of interstitial fluid pressure by Landerer in 1884,57 many different measurement techniques have been used. Pressure measurement systems can be classified either as fluid filled or solid state. Fluidfilled methods consist of the needle manometer, wick, and micropipette techniques. The needle manometer21,57 is a fluid-filled system that uses a slow, continuous infusion to keep the needle patent. The wick technique for pressure measurement consists of a cotton wick inserted into a 1.5-mm Teflon tube and introduced into the tissue through a needle.84 Hargens et al.38 later modified the technique to become the wick catheter technique. A third type of system uses a glass micropipette with a 5-␮m tip diameter, making this the least traumatic method and also resulting in the least tissue distortion.98

1098

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 43–2 Microscopic view of a pressure microsensor used for measuring intramuscular pressure.

lengths greater than L0 (Fig. 43–4). A positive linear relationship was found between IMP and muscle stress for both the ascending and descending limbs (Table 43–2). These data indicate that IMP measurement provides an accurate index of muscle tension under both active and passive conditions.

Stress (kPa)

AVERAGE LENGTH-STRESS Active stress Passive stress

300 250 200 150 100 50 0 -50 0.70

0.80

0.90

1.00

1.10

1.20

1.30

L/Lo

FIGURE 43–3 Muscle length-tension relationship. (From Davis, J., Kaufman, K. R., and Lieber R. L.: Correlation between active and passive isometric stress and intramuscular pressure in the isolated rabbit tibialis anterior muscle. J. Biomech. 36:505, 2003, with permission.)

AVERAGE LENGTH-PRESSURE Active pressure Passive pressure

45 Pressure (mmHg)

However, micropipette measurements require tissue immobilization. Fluid-filled pressure recording systems are sensitive to hydrostatic artifacts.64,80,90 Thus, fluidfilled systems may be used only with limited movements that do not involve position changes relative to the horizontal plane, rendering this technology inappropriate for measurements during dynamic activities. In contrast, a fiberoptic transducer is not sensitive to hydrostatic artifact24 and is effective for measuring intramuscular pressure during exercise.23 Recently, Willy et al.99 introduced an accurate self-calibrating, battery-powered electronic transducer-tipped catheter. However, the catheter tip is quite large (0.99 mm). Commercially available fiberoptic pressure transducers are also too large for optimum comfort (0.55 mm) and may themselves induce pressure artifacts based on their large size. A new microsensor has been developed for measuring interstitial fluid pressure. The pressure microsensor has a 360-␮m diameter (Fig. 43–2). The microsensor has an accuracy, repeatability, and linearity better than 2% full-scale output (FSO) and a hysteresis of 4.5% FSO.49 Hill41 noted that mechanical pressure is developed inside a muscle when it contracts. The increase in pressure is caused by the contracting muscle fibers applying pressure on the interstitial fluid volume. This observation led to the measurement of IMP as a direct measure of muscle force. Sylvest and Hvid93 used a needle technique to measure joint torque and IMP of human striated muscles under isometric conditions. They found a linear relationship between IMP and torque. Through animal41,50,65,92 and human studies,39,44,46,56,71,74,75,78,86,93 investigators have shown that an approximately linear relationship exists between IMP and muscle force during isometric contraction. Baumann et al.9 suggested that IMP is related to the active and passive components of muscle tension. The relationship between IMP and active and passive muscle tension has been quantified.26 The fiber length–isometric tension curve was characterized by an “ascending limb” at a length less than muscle optimum length (L0) and a “descending limb” at a length greater than L0 (Fig. 43–3). The shape of this curve presumably represents a scaled and distorted version of a sarcomere length-tension curve previously published.33,79 Passive muscle tension increased in a fairly exponential fashion at lengths greater than L0. The length-pressure relationship generally mimicked the shape of the length-tension curve, with an ascending limb at lengths less than L0 and descending limb at

30 15 0 -15 0.70

0.80

0.90

1.00

1.10

1.20

1.30

L/Lo

FIGURE 43–4 Intramuscular pressure–length relationship. (From Davis, J., Kaufman, K. R., and Lieber R. L.: Correlation between active and passive isometric stress and intramuscular pressure in the isolated rabbit tibialis anterior muscle. J. Biomech. 36:505, 2003, with permission.)

1099

Quantitative Muscle Strength Assessment

Table 43–2. Summary of Isometric Stress–Pressure Correlations* Pressure Coefficients of Determination (r2)

Limb of the LengthTension Curve

Activation State

RMS Error (kPa)

Mean ⴞ SEM

Range (min–max)

Active

Ascending Limb Descending Limb

175.2 ⫾ 52.98 147.19 ⫾ 47.46

.60 ⫾ .25 .77 ⫾ .16

.138–.963 .343–.947

Passive

Ascending Limb Descending Limb

10.18 ⫾ 4.21 20.19 ⫾ 13.83

.53 ⫾ .36 .86 ⫾ .11

.045–.842 .672–.982

RMS ⫽ root mean square. *Values represent mean ⫾ standard error for 10 animal subjects. From Davis, J., Kaufman, K. R., and Lieber, R. L.: Correlation between active and passive isometric stress and intramuscular pressure in the isolated rabbit tibialis anterior muscle. J. Biomech. 36:505, 2003, with permission.

Muscle function during gait has been quantified using IMP.48 IMP increased at the beginning of single-limb stance (opposite toe-off) (Fig. 43–5). The increase of IMP corresponded with the increase in electromyographic activity of the gastrocnemius. The greatest muscle activity in the plantar flexors was required near the end of single-limb stance to meet the high intrinsic plantar flexion moment occurring at the ankle joint and to reverse the direction of ankle movement (Fig. 43–6). It should also be noted that the IMP reading continued briefly after cessation of the EMG. Furthermore, it can be noted that there is a lower-level IMP recording during the swing phase of gait. During the stance phase of gait,

FS

TO FS

OTO OFS

TO

OTO OFS

FS

TO FS

OTO OFS

(volts)

EMG right gastrocnemius +0.5 0

(mmHG)

+0.5 IMP right gastrocnemius 100 50 0

FIGURE 43–5 Raw data for a single subject during gait. Both electromyogram (EMG) activity and intramuscular pressure are being recorded from the gastrocnemius muscle. The stance phase of gait occurs from foot strike (FS) to toe-off (TO). The swing phase of gait occurs from TO to FS. Single-limb stance occurs from opposite toe-off (OTO) to opposite foot strike (OFS). (From Kaufman, K. R., and Sutherland, D. H.: Dynamic intramuscular pressure measurement during gait. Oper. Tech. Sports Med. 3:253, 1995, with permission.)

the peak IMP recording corresponds to the time when the ankle moment is at a maximum (opposite foot strike) (Fig. 43–6). Furthermore, during the swing phase of gait, the peak of IMP corresponds to the point of time when the ankle is at peak dorsiflexion (Fig. 43–6). Thus the peaks in IMP during gait can be correlated with the peaks of active contraction and passive stretch of the gastrocnemius.

SUMMARY Human mobility depends on the ability of the muscles to generate sufficient force to overcome resistance caused by gravitational forces (i.e., body posture) and inertial forces (i.e., rapid movements). Patients with neurologic disorders frequently have impairments in skeletal muscle strength. Their mobility declines as a result of pathologic processes and improves with rehabilitation. Our capability to quantify a patient’s muscle strength objectively gives us the ability to define a patient’s functional status and document changes over time. These measurements can be obtained using noninstrumented or instrumented methods. The time-related changes may also have prognostic implications. Most measurement techniques quantify strength of muscle groups rather than individual muscles. Technology is emerging that will offer the potential for quantifying in vivo muscle force. Muscle strength measurements are an essential component of the neuromuscular evaluation and provide substantial information to guide patient treatment.

ACKNOWLEDGMENTS The work in this chapter was partly supported by National Institutes of Health grant R01 HD31476. The author thanks Barb Iverson-Literski for her careful assistance with manuscript preparation.

1100

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 43–6 Ankle motion (A) and moment (B) during gait. The gait cycle is defined as the events that occur between successive footsteps of the same foot. The gait cycle begins with foot strike, continues through stance and swing phases, and ends with foot strike of the same foot.

REFERENCES 1. Aitkens, S. S., Lord, J., Bernauer, E., et al.: Relationship of manual muscle testing to objective strength assessment. Muscle Nerve 12:173, 1989. 2. Amundsen, L. R., and Graves, J. M.: Testing knee extensor muscles of survivors of poliomyelitis. J. Hum. Muscle Performance 1:25, 1991. 3. An, K.-N., Chao, E. Y., Cooney, W. P., and Linscheid, R. L.: Forces in a normal and abnormal hand. J. Orthop. Res. 3:202, 1985. 4. Andriacchi, T. P., Andersson, G. B. J., Fermier, R. W., et al.: A study of lower limb mechanics during stair climbing. J. Bone Joint Surg. Am. 62:749, 1980. 5. Armstrong, L. E., Winant, D. M., Swasey, P. R., et al.: Using isokinetic dynamometry to test ambulatory patients with multiple sclerosis. Phys. Ther. 63:1274, 1983. 6. Askew, L. J., An, K.-N., Morrey, B. F., and Chao, E. Y.: Isometric elbow strength in normal individuals. Clin. Orthop. 222:261, 1987. 7. Atkens, S., Lord, J., Bernauer, E., et al.: Relationship of manual muscles testing to objective strength measurements. Muscle Nerve 12:173, 1989. 8. Barnes, W. S.: The relationship of motor-unit activation to isokinetic muscular contraction at different contractile velocities. Phys. Ther. 16:1152, 1980. 9. Baumann, J. U., Sutherland, D. H., and Hanggi, A.: Intramuscular pressure during walking: an experimental study using the wick catheter technique. Clin. Orthop. 145:292, 1979. 10. Beasley, W. C.: Influence of method on estimates of normal knee extensor force among normal and post-polio children. Phys. Ther. Rev. 36:21, 1956. 11. Beasley, W. C.: Quantitative muscle testing: principles and applications to research and clinical services. Arch. Phys. Med. Rehabil. 42:398, 1961. 12. Bigland, B., and Lippold, O. C. J.: The relationship between force, velocity, and integrated electrical activity in human muscles. J. Physiol. (Lond.) 123:214, 1954.

13. Bigland-Ritchie, B., Kukulka, C. J., and Woods, J. J.: Surface EMG-force relationships in human muscles of different fibre composition. J. Physiol. (Lond.) 308:103P, 1980. 14. Bogey, R. A., Barnes, L. A., and Perry, J.: Computer algorithms to characterize individual subject EMG profiles during gait. Arch. Phys. Med. Rehabil. 73:835, 1992. 15. Bogey, R. A., Barnes, L. A., and Perry, J.: A computer algorithm for defining the group electromyographic profile from individual gait profiles. Arch. Phys. Med. Rehabil. 74:286, 1993. 16. Bohannon, R. W.: Manual muscle test scores and dynamometer test scores of knee extension strength. Arch. Phys. Med. Rehabil. 67:390, 1986. 17. Bohannon, R. W.: Make versus break tests of elbow flexion force using a hand-held dynamometer. Phys. Ther. 68:93, 1988. 18. Bohannon, R. W.: Muscle Strength Testing with Hand-held Dynamometers. New York, Churchill Livingstone, 1990. 19. Bohannon, R. W., Hull, D., and Palmeri, D.: Muscle strength impairments in gait performance deficits in kidney transplantation candidates. Am. J. Kidney Dis. 24:480, 1994. 20. Bouisset, S.: EMG and muscle force in normal muscle activities. In Desmedt, J. E. (ed.): New Developments in EMG and Clinical Physiology. Basel, Karger, p. 547, 1973. 21. Burch, G. E., and Sodeman, W. A.: The estimation of the subcutaneous tissue pressure by a direct method. J. Clin. Invest. 16:845, 1937. 22. Cooney, W. P., An, K.-N., and Chao, E. Y. S.: Direct measurement of tendon forces in the hand. In Transactions of the 32nd Annual Meeting of the Orthopaedic Research Society, February 17–20, p. 53, 1986. 23. Crenshaw, A. G., Styf, J. R., and Hargens, A. R.: Intramuscular pressures during exercise: an evaluation of a fiberoptic transducer-tipped catheter system. Eur. J. Appl. Physiol. 65:178, 1992. 24. Crenshaw, A. G., Styf, J. R., Mubarak, S. J., and Hargens, A. R.: A new “transducer tipped” fiber optic catheter for measuring intramuscular pressures. J. Orthop. Res. 8:464, 1990. 25. Csuka, M., and McCarty, D. J.: Simple method for measurement of lower extremity muscle strength. Am. J. Med. 78:77, 1985.

Quantitative Muscle Strength Assessment 26. Davis, J., Kaufman, K. R., and Lieber, R. L.: Correlation between active and passive isometric stress and intramuscular pressure in the isolated rabbit tibialis anterior muscle. J. Biomech. 36:505, 2003. 27. Dempster, W. T., and Finerty, J. C.: Relative activity of wrist moving muscles in static support of the wrist joint: an electromyographic study. Am J. Physiol. 150:596, 1947. 28. deVries, H.: Efficiency of electrical activity as a physical measure of the functional state of muscle tissue. Am. J. Phys. Med. 47:10, 1968. 29. Finni, T., Komi, P. V., and Lepola, V.: In-vivo human triceps surae and quadriceps femoris muscle function in a squat jump and counter movement jump. Eur. J. Appl. Physiol. 83:416, 2000. 30. Finni, T., Komi, P. V., and Lukkariniemi, J.: Achilles tendon loading during walking: application of a novel optic fiber technique. Eur. J. Appl. Physiol. 77:289, 1998. 31. Fowler, W. M., and Gardner, G. W.: Quantitative strength measurements in muscular dystrophy. Arch. Phys. Med. Rehabil. 48:629, 1967. 32. Fukashiro, S., Komi, P. V., Jarvinen, M., and Miyahita, M.: In-vivo Achilles tendon loading during jumping in humans. Eur. J. Appl. Physiol. 71:453, 1995. 33. Gordon, A. M., Huxley, A. F., and Julian, F. J.: The variation in isometric tension with sarcomere length in vertebrate muscle fibres. J. Physiol. (Lond.) 184:170, 1966. 34. Gregor, R. J., Komi, P. U., Browning, R.C., and Jarvinen, M.: A comparison of the triceps surae and residual muscle moments at the ankle during cycling. J. Biomech. 24:287, 1991. 35. Gregor, R. J., Komi, P. U., and Jarvinen, M.: Achilles tendon forces during cycling. Int. J. Sports Med. 8(Suppl. 1):9, 1987. 36. Guralnik, J. M., Simonsick, E. M., Ferrucci, L., et al.: A short physical performance battery assessing lower extremity function: association with self reported disability and prediction of mortality and nursing home admission. J. Gerontol. 49M:85, 1994. 37. Haffajee, D., Maritz U., and Suantesson, G.: Isometric knee extension strength as a function of joint angle, muscle length and motor unit activity. Acta Orthop. Scand. 43:138, 1972. 38. Hargens, A. R., Mubarak, S. J., Owen, C. A., et al.: Interstitial fluid pressure in muscle and compartment syndromes in man. Microvasc. Res. 14:1, 1977. 39. Hargens, A. R., Sejersted, O. M., Kardel, K. R., et al.: Intramuscular fluid pressure: a function of contraction force and tissue depth. Trans. Orthop. Res. Soc. 7:371, 1982. 40. Hatze, H.: A general myocybernetic control model of skeletal muscle. Biol. Cybern. 28:143, 1978. 41. Hill, A. V.: The pressure developed in muscle during contraction. J. Physiol. (Lond.) 107:518, 1948. 42. Hinson, M. N., and Rosentswieg, J.: Comparative electromyographic values of isometric, isotonic, and isokinetic contraction. Res. Q. Exerc. Sport 44:71, 1973. 43. Hof, A. L., and Vandenberg, J. W.: Linearity between the weighted sum of the EMGs of the human triceps surae and the total torque. J. Biomech. 10:529, 1977. 44. Hussain, S. N. A, and Magder, S.: Diaphragmatic intramuscular pressure in relation to tension, shortening, and blood flow. J. Appl. Physiol. 71:159, 1991. 45. Inman, V. T., Ralston, H. J., Saunders, J. B., et al.: Relationship of human electromyogram to muscular tension. Electroencephalogr. Clin. Neurophysiol. 4:187, 1952.

1101

46. Järvholm, U., Palmerud, G., Herberts, P., et al.: Intramuscular pressure and electromyography in the supraspinatus muscle at shoulder abduction. Clin. Orthop. 245:102, 1989. 47. Kaufman, K., Hughes, C., Morrey, B., et al.: Gait characteristics of adults with knee osteoarthritis. J. Biomech. 34:907, 2001. 48. Kaufman, K. R, and Sutherland, D. H.: Dynamic intramuscular pressure measurement during gait. Oper. Tech. Sports Med. 3:250, 1995. 49. Kaufman, K. R., Wavering, T., Morrow, D., et al.: Performance characteristics of a pressure microsensor. J. Biomech. 36:283, 2002. 50. Kirkebö, A., and Wisnes, A.: Regional tissue fluid pressure in rat calf muscle during sustained contraction on stretch. Acta Physiol. Scand. 11:551, 1982. 51. Komi, P. V.: Relationship between Muscle Tension, EMG, and Velocity of Contraction under Concentric and Eccentric Work. Basel, Karger, 1973. 52. Komi, P. V.: Relevance of in vivo force measurements to human biomechanics. J. Biomech. 21:23, 1990. 53. Komi, P. V., and Buskirk, E. R.: Effect of eccentric and concentric muscle conditioning on tension and electrical activity of human muscle. Ergonomics 15:417, 1972. 54. Komi, P. V., and Rusko, H.: Quantitative evaluation of mechanical and electrical changes during fatigue loadings of eccentric and concentric work. Scand. J. Rehabil. Med. 3(Suppl.):121, 1974. 55. Komi, P. V., Salonen, M., Jarvinen, M., and Kokko, O.: In vivo registration of Achilles tendon forces in man: II. Methodological development. Int. J. Sports Med. 8:3, 1987. 56. Körner, L., Parker, P., Almström, C., et al.: Relationship of intramuscular pressure to the force output and myoelectric signal of skeletal muscle. J. Orthop. Res. 2:289, 1984. 57. Landerer, A. S.: Die Gewebesspannung in ihrem Einfluss auf die ortlich Blut-und-Lymphbewegung [Tissue contraction and its influence on local blood and lymph flow]. Leipzig, Vogel, 1884. 58. Lieber, R. L., and Boakes, J. L.: Sarcomere length and joint kinematics during torque production in the frog hindlimb. Am J. Physiol. 254:C759, 1988. 59. Lippold, O. C. J.: The relation between integrated action potentials in a human muscle and its isometric tension. J. Physiol. (Lond.) 117:492, 1952. 60. Lippold, O. C. J., Redfearn, J., and Vuco, J.: Electromyography of fatigue. Ergonomics 3:121, 1970. 61. Lovett, R. W.: The Treatment of Infantile Paralysis, 2nd ed. Philadelphia, Blakiston’s Son & Co., 1917. 62. Lunnen, J. D, Yack, J., and Le Veau, B. F.: Relationship between muscle length, muscle activity, torque of the hamstring muscles. Phys. Ther. 61:190, 1981. 63. Maton, B., and Bouisset, S.: The distribution of activity among the muscles of a single group during isometric contraction. Eur. J. Appl. Physiol. 37:101, 1977. 64. Matsen, F. A., Mayo, K. A., Sheridan, G. W., and Krugmire, R. B. Jr.: Monitoring of intramuscular pressure. Surgery 79:702, 1976. 65. Mazella, H.: On the pressure developed by the contraction of striated muscle and its influence on muscular circulation. Arch. Int. Physiol. 62:334, 1954. 66. Messier, R. H., Duffy, J., Litchman, H. M., et al.: The electromyogram as a measure of tension in the human biceps and triceps muscles. Int. J. Mechan. Sci. 13:585, 1971.

1102

Quantitating Symptoms, Impairments, and Outcomes

67. Metral, S., and Cassar, G.: Relationship between force and integrated EMG activity during voluntary isometric anisotonic contraction. Eur. J. Appl. Physiol. 46:185, 1981. 68. Mills, K.: Wasting, weakness and the MRC Scale in the first dorsal interosseous muscle. J. Neurol. Neurosurg. Psychiatry 62:541, 1997. 69. Milner-Brown, H. S., and Stein, R. B.: The relation between surface electromyogram and muscle force. J. Physiol. (Lond.) 246:549, 1975. 70. Moritani, T., and de Vries, H. A.: Reexamination of the relationship between the surface integrated electromyogram (IEMG) and force of isometric contraction. Am. J. Phys. Med. 57:263, 1978. 71. Mubarak, S. J., Hargens, A. R., Owen, C. A., et al.: The wick catheter technique for measurement of intramuscular pressure: a new research and clinical tool. J. Bone Joint Surg. Am. 58:1016, 1976. 72. Olgiati, R., Burgunder, J. M., and Mumenthaler, M.: Increased energy cost of locking in multiple sclerosis: effect of spasticity, ataxia, and weakness. Arch. Phys. Med. Rehabil. 69:846, 1988. 73. Osternig, L. R., Hamill, J., Corros, D. M., and Lander, J.: Electromyographic patterns accompanying isokinetic exercise under varying speed and sequencing conditions. Am. J. Phys. Med. 63:289, 1984. 74. Owen, C. A., Garetto, L. P., Hargens, A. R., et al.: Relationship of intramuscular pressure to strength and muscular contraction. Trans. Orthop. Res. Soc. 2:246, 1977. 75. Parker, P. A., Korner, L., and Kadefors, R.: Estimation of muscle force from intramuscular total pressure. Med. Biol. Eng. Comput. 22:453, 1984. 76. Perry, J.: Gait Analysis: Normal and Pathological Function. Thorofare, NJ, Slack, 1992. 77. Perry, J., Ireland, M. L., Gronley, J., and Hoffer, M. M.: Predictive value of manual muscle testing and gait analysis in normal ankles by dynamic electromyography. Foot Ankle Int. 6:254, 1986. 78. Petrofsky, J. S., and Hendershot, D. M.: The interrelationship between blood pressure, intramuscular pressure, and isometric endurance in fast and slow twitch skeletal muscle in the cat. Eur. J. Appl. Physiol. Occup. Physiol. 53:106, 1984. 79. Rack, P. M. H., and Westbury, D. R.: The effects of length and stimulus rate on tension in the isometric cat soleus muscle. J. Physiol. (Lond.) 204:443, 1969. 80. Rorabeck, C. H., Castle, G. S. P., Hardie, R., and Logan, J.: Compartmental pressure measurements: an experimental investigation using the slit catheter. J. Trauma 21:446, 1981. 81. Rosentswieg, J., and Hinson, M. M.: Comparison of isometric, isotonic, and isokinetic exercises by electromyography. Arch. Phys. Med. Rehabil. 53:249, 1972. 82. Rothstein, J. M., DeLitto, A., Sinacore, D. R., and Rose, S. J.: Electromyographic, peak torque, and power relationships during isokinetic movement. Phys. Ther. 63:926, 1983. 83. Salmons, S.: Report on the 8th International Conference on Medical and Biological Engineering. Biomed. Eng. 4:467, 1969. 84. Scholander, P. F., Hargens, A. R., and Miller, S. L.: Negative pressure in the interstitial fluid of animals. Science 161:321, 1968.

85. Schwartz, S., Cohen, M. E., Herbison, G. J., and Shah, A.: Relationship between two measures of upper extremity strength: manual muscle test. Arch. Phys. Med. Rehabil. 73:1063, 1992. 86. Sejersted, O. M., Hargens, A. R., Kardel, K. R., et al.: Intramuscular fluid pressure during isometric contraction of human skeletal muscle. J. Appl. Physiol. Respir. Environ. Exerc. Physiol. 56:287, 1984. 87. Seliger, V., Dolejs, L., and Karas, V.: A dynamometric comparison of maximum eccentric, concentric, and isometric conditions using EMG and energy expenditure measurements. Eur. J. Appl. Physiol. 45:235, 1980. 88. Sharrad, W. J. W.: Correlations between the changes in the spinal cord and muscular paralysis in poliomyelitis. Proc. R. Soc. Lond. 40:346, 1953. 89. Smidt, G., and Wadsworth, J. B.: Floor reaction forces during gait: comparison of patients with hip disease and normal subjects. Phys. Ther. 53:1056, 1973. 90. Styf, J. R.: Evaluation of injection techniques in recording of intramuscular pressure. J. Orthop. Res. 7:812, 1989. 91. Sunderland, A., Tinson, D., and Bradley, L.: Arm function after stroke: an evaluation of grip strength as a measure of recovery and a prognostic indicator. J. Neurol. Neurosurg. Psychiatry 52:1267, 1989. 92. Sutherland, D. H., Woo, S.-Y., Schoon, J., et al.: The potential application of a small solid state pressure transducer to measure muscle activity during gait. Trans. Orthop. Res. Soc. 2:289, 1977. 93. Sylvest, O., and Hvid, N.: Pressure measurements in human striated muscles during contraction. Acta Rheumatol. Scand. 5:216, 1959. 94. Tripp, E. J., and Harris, S. R.: Test-retest reliability of isokinetic knee extension and flexion torque measurements in persons with spastic hemiparesis. Phys. Ther. 71:390, 1991. 95. Vandervoort, A. A., Kramer, J. F., and Wharram, E. R.: Eccentric knee strength of elderly females. J. Gerontol. 45:125, 1990. 96. Wakim, K. G., Gersten, J. W., Elkins, E. C., and Martin, G. M.: Objective recording of muscle strength. Arch. Phys. Med. Rehabil. 31:90, 1950. 97. Wickholm, J. B., and Bohannon, R. W.: Hand-held dynamometer measurements: testor strength makes a difference. J. Orthop. Sports Phys. Ther. 13:191, 1991. 98. Wiederhielm, C. A., Woodbury, J. W., Kirk, S., and Rushmer, R. F.: Positive pressures in microcirculation of frogs mesentery. Am J. Physiol. 207:173, 1964. 99. Willy, C., Gerngross, H., and Sterk, J. C.: Measurement of intracompartmental pressure with use of a new electronic transducer-tipped catheter system. J. Bone Joint Surg. Am. 81:158, 1999. 100. Woods, J. J., and Bigland-Ritchie, B.: Linear and nonlinear surface EMG/force relationships in human muscles. Am. J. Phys. Med. 62:287, 1983. 101. Wrentenberg, P., Lindberg, F., and Arborelius, U. P.: Effects of arm rests and different ways of using them on hip and knee during rising. Clin. Biomech. 8:95, 1993. 102. Zuniga, E. N., and Simons, D. G.: Nonlinear relationship between averaged electromyogram potential and muscle tension in normal subjects. Arch. Phys. Med. Rehabil. 50:613, 1969.

44 Quantitation of Autonomic Impairment PHILLIP A. LOW AND CHRISTOPHER J. MATHIAS

Overview Indications for Autonomic Function Tests Aims of Laboratory Evaluation of Autonomic Function The Autonomic Reflex Laboratory Patient Preparation Routine Tests of Autonomic Function Quantitative Sudomotor Axon Reflex Test Evaluation of Cardiovagal Function Using Heart Rate Recordings Beat-to-Beat Blood Pressure Recordings

Composite Autonomic Severity Score Salivation Test Biochemical and Hormonal Evaluation of Autonomic Activity Nonroutine Studies Imaging of Sympathetic Innervation to Organs Electrophysiologic Assessment of Sympathetic Neural Activity Skin Vasomotor Reflexes Orthostatic Stress Tests

OVERVIEW Motor, sensory, and autonomic fibers can be affected separately or together in the peripheral neuropathies. Neurophysiologic studies of large and small nerve fibers have provided important insights into the pathophysiology of neuropathies.44 The study of large nerve fibers has been universally applied in the clinical neurophysiology laboratory in a relatively standardized fashion. However, whereas studies of small nerve fibers continue at a rapid pace in the research laboratory, the clinical application of tests of autonomic function has progressed at a more uneven pace. There is usually significant distal autonomic involvement in the common “garden variety” sensorimotor neuropathies, especially when significant active axonal degeneration is present.195 In the past few years, autonomic function tests have become more available. It has become clear to the peripheral neurologist that the original suggested battery of five tests50 is quite inadequate, with its excessive reliance on cardiovascular heart rate tests. With this in mind, the American Academy of Neurology noted in a position paper the need to provide a more comprehensive evaluation of cardiovagal, postganglionic sudomotor, and adrenergic function. These tests are summarized in Table 44–1.

Blood Pressure and Heart Rate Response to Prolonged Tilt Other Cardiovascular Heart Rate Tests Determination of Baroreflex Sensitivity Determination of Denervation Supersensitivity Isoproterenol Infusion Test Evaluation of Splanchnic-Mesenteric Vasculature Evaluation of Central Autonomic Pathways

In this chapter we focus on the well-standardized tests of autonomic function and provide normative data. This is followed by additional sections that focus on autonomic systems that are not routinely studied.

INDICATIONS FOR AUTONOMIC FUNCTION TESTS The role of the laboratory is to extend the clinician’s capability to diagnose, quantitate, and evaluate autonomic dysfunction. Its value is enhanced when the needs of the referring physician and the capabilities of the laboratory are well matched. The list of indications given in Table 44–2 comprises those for which such a match appears to occur. To Detect the Presence of Generalized Autonomic Failure This is one of the main reasons for referring patients for autonomic evaluation. It should be possible to delineate the distribution by autonomic system (cardiovagal, adrenergic, sudomotor), as well as the severity and level (pre- versus postganglionic) of autonomic failure. The results of autonomic studies can help the physician make a diagnosis. For instance, if a patient with parkinsonism has only minor 1103

1104

Quantitating Symptoms, Impairments, and Outcomes

Table 44–1. Autonomic Reflex Screen 1. 2. 3. 4.

Quantitative sudomotor axon reflex test distribution Heart rate response to deep breathing Valsalva ratio Beat-to-beat blood pressure and heart rate responses to the Valsalva maneuver 5. Beat-to-beat blood pressure responses to tilt

autonomic impairment, then multiple system atrophy (MSA) is unlikely. If the patient clinically has Parkinson’s disease (PD) but has orthostatic hypotension (OH), the differential diagnosis between MSA and PD is problematic. If the laboratory evaluation shows OH and diffuse autonomic failure, then MSA is more likely than PD with autonomic failure.97,145 Another reason for autonomic testing is that the diagnosis has a serious prognostic impact in disorders such as MSA or pure autonomic failure (PAF). In diabetes and amyloidosis, the development of generalized autonomic failure significantly worsens the prognosis.51,139,144 The most common causes of generalized autonomic failure are the autonomic neuropathies, MSA, and PAF. The most common autonomic neuropathies are diabetic autonomic neuropathy, amyloid neuropathy, Sjögren’s syndrome, and autoimmune neuropathies, either idiopathic or paraneoplastic autonomic neuropathy. To Detect the Presence of a Limited Autonomic Failure Although the initial surge of referrals to the autonomic laboratory was focused on finding generalized autonomic failure, it soon became obvious that limited autonomic failure was much more common. Of greater practical importance is that many common syndromes such as apparent vasovagal syncope, postural tachycardia syndrome (POTS), and heat intolerance may have underlying, limited, or selective Table 44–2. Indications for Autonomic Function Tests 1. When generalized autonomic failure is suspected: a. Autonomic neuropathies b. Multiple system atrophy c. Pure autonomic failure 2. To diagnose limited autonomic neuropathy 3. When distal small fiber neuropathy is suspected 4. To diagnose orthostatic intolerance 5. To detect neuropathic basis in neurocardiogenic (vasovagal) syncope 6. To monitor the course of a neuropathy 7. To evaluate response to therapy 8. To evaluate autonomic involvement in the peripheral neuropathies 9. To detect sympathetic dysfunction in sympathetically maintained pain 10. In clinical treatment trials

autonomic neuropathy, an observation that has important treatment implications. For instance, if the patient has a peripheral adrenergic neuropathy (with impaired vasomotor tone), but with most autonomic reflexes being intact, the excessive venous pooling and reduced preload can trigger an active syncopal reflex. These patients respond best to measures that selectively improve vasomotor tone, in contrast to standard syncopal therapies. A limited autonomic neuropathy may underlie such syndromes as chronic idiopathic anhidrosis.102 Chronic idiopathic anhidrosis cannot be diagnosed without the demonstration of normal adrenergic and cardiovagal functions. Patients with POTS may have the disorder sui generis or as a result of an autonomic neuropathy.152 An autonomic screen is necessary to clarify this differential diagnosis. To Diagnose Distal Small Fiber Neuropathy Distal small fiber neuropathy is common, often distressing, and very difficult to diagnose. The patient complains of burning feet, excessive coldness or sweating, and vasomotor alterations. Routine nerve conduction studies and electromyography are usually normal because the brunt of the disorder is on unmyelinated fibers. Epidermal nerve fibers are somatic C fibers, whose attrition can be demonstrated with skin biopsy; this can be demonstrated in most patients125,130 with autonomic tests and, in about one third of cases, is the best approach. Although the evaluation of somatic C fibers is logical in these patients with burning feet, concomitant vasomotor and sudomotor changes are usually present, so that autonomic function tests are appropriate if done distally. Peripheral autonomic surface potentials will detect a small minority of cases.49 The quantitative sudomotor axon reflex test (QSART) and the thermoregulatory sweat test (TST) are abnormal in about 80% of cases.166 To Detect the Presence and Evaluate the Severity of Orthostatic Intolerance Orthostatic intolerance is present when the patient develops symptoms on standing up that clear on lying or sitting down. When there is a fall in blood pressure (BP) of sufficient degree, then OH is present. The majority of patients do not have OH. Some of these patients have excessive heart rate response (POTS) and some do not but instead faint (syncope). When orthostatic intolerance is of mild degree, no evaluation may be needed. A subset of patients with highly symptomatic orthostatic intolerance may be severely incapacitated. The role of autonomic testing is to diagnose its presence and severity and to determine if there is underlying autonomic failure. To Detect the Presence of a Neuropathic Basis in Neurocardiogenic Syncope The majority of patients with syncope (vasovagal or neurocardiogenic) do not have underlying autonomic failure. However, some patients with a limited autonomic neuropathy

1105

Quantitation of Autonomic Impairment

can develop secondary syncope. Many patients with limited autonomic neuropathy, including common entities such as diabetic neuropathy71,163 and POTS, can develop a hyperadrenergic state and be particularly prone to syncope. When the underlying autonomic deficit is determined, treatment focused on the abnormality can be effective.

role is to use the techniques to gain basic information in human subjects.

To Monitor the Course of the Autonomic Disorder The twin attributes of quantitation and noninvasiveness render autonomic laboratory evaluation ideally suited to monitor the alterations of autonomic function over time. The patient’s autonomic deficits may change in type, distribution, or severity. These characteristics have hitherto not been readily quantifiable.

The availability of noninvasive, well-validated tests of autonomic function has resulted in the application of these tests to the evaluation of autonomic function in the neuropathies. The goals of clinical autonomic testing may be summarized as follows:

To Evaluate the Response to Therapy The autonomic deficits may lessen in response to treatment. With various therapies being available for the neuropathies (e.g., antioxidants, aldose reductase inhibitors, tight glucose control, gangliosides, fish oil, etc. for diabetic neuropathy; 3,4-diaminopyridine for the Lambert-Eaton myasthenic syndrome; and immunotherapy for the immune-mediated neuropathies), quantitative methods are needed to evaluate the response to therapy. To Evaluate Autonomic Involvement in the Peripheral Neuropathies In the peripheral neuropathies, there is a typical pattern of autonomic involvement with a length-dependent distribution of sympathetic deficits (maximal distally). However, some neuropathies have an early impairment of cardiovagal function (e.g., diabetes; Chagas’ neuropathy), and the distribution of the sudomotor deficit may be multifocal (e.g., leprosy). A combination of routine tests (QSART, cardiovagal tests, beat-to-beat BP [BPBB] responses to head-up tilt [HUT] and to Valsalva maneuver, and TST) permits such an analysis. To Detect Sympathetic Dysfunction in Sympathetically Maintained Pain Patients with unilateral limb pain suspected of having complex regional pain syndrome type I (CRPS I), also known as sympathetically maintained pain and reflex sympathetic dystrophy or causalgia, will have sympathetic overreaction. Although the pathophysiology of CRPS I is unresolved as to whether sympathetic dysfunction is primary or incidental, sympathetic asymmetry, especially sudomotor asymmetry, is a reliable feature that helps in its recognition.26,107 This topic is not discussed further in this chapter. To Answer Research Questions The autonomic laboratory can be used to evaluate changes in autonomic function in response to therapy in clinical treatment trials and neuroepidemiologic studies. Another

AIMS OF LABORATORY EVALUATION OF AUTONOMIC FUNCTION

1. To detect the presence of autonomic involvement of the autonomic component of peripheral nerve 2. To quantitate the severity and apportion the type (sudomotor, adrenergic, cardiovagal) of deficits 3. To determine the distribution of autonomic failure 4. To determine the site of the autonomic lesion

THE AUTONOMIC REFLEX LABORATORY All studies of autonomic function are performed in the autonomic reflex laboratory with the subject supine and rested. The room temperature is optimally maintained at 23° C, and ideally the room is humidity controlled. This temperature is somewhat uncomfortable for the operator. It is acceptable to work with an ambient temperature of 20° C, preferred by most technologists, but only if careful warming of the limbs is done. Tests of autonomic function are broadly grouped into two categories. The first is the autonomic reflex screen, comprising QSART recordings from the forearm and three lower extremity sites, heart rate response to deep breathing (HRDB), Valsalva ratio (VR), and BPBB response to HUT and Valsalva maneuver. The second category comprises additional or specialized studies that are nonroutine. These include indices that can be measured using bioelectric impedance analysis or the “contour” method of determining stroke volume. There are also specialized studies such as prolonged tilt, infusions of phenylephrine or isoproterenol, and research protocols that measure other circulations, such as the cerebral or mesenteric beds, or utilize specialized techniques such as microneurography.

Patient Preparation No food, coffee, or nicotine is permitted for 3 hours before the study. Medications are stopped, if feasible, for 5 half-lives. Translated to practical terms, the following guidelines are reasonable. Anticholinergic (including antidepressants, antihistamines, over-the-counter cough and cold medications, diuretics, and sympathomimetics [ and  agonists]) and parasympathomimetic agents are

1106

Quantitating Symptoms, Impairments, and Outcomes

forbidden for 48 hours. Short-acting  and  antagonists are discontinued for 24 hours and long-acting ones for 48 hours, at the discretion of the referring physician. Analgesics, including opioids, are avoided the day of the test. 9--Fludrocortisone (Florinef) is continued. Compressive clothing, including Jobst stockings and corsets, is not worn the morning of the test. The patient should not have had an acute illness in the previous 48 hours. Vigorous unaccustomed exercise should be avoided in the previous 24 hours.

ROUTINE TESTS OF AUTONOMIC FUNCTION A number of tests have been in relatively common use for a sufficient duration to be considered routine. These tests are considered to be sufficiently useful that a complete autonomic laboratory should probably have them available. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

QSART distribution Heart rate response to deep breathing The Valsalva ratio Heart rate response to standing Beat-to-beat blood pressure response to the Valsalva maneuver Beat-to-beat blood pressure response and heart rate response to tilt Salivation test Thermoregulatory sweat test Sustained handgrip Plasma catecholamines Frequency analysis

Not all the tests listed above are routinely done in the autonomic reflex screen. The component tests of the autonomic reflex screen are shown in Table 44–1. The tests should follow a specific sequence. A convenient approach is to begin the study with QSART followed by HRDB, then BPBB and heart rate in response to the Valsalva maneuver, and finally BPBB and heart rate in response to HUT.

Quantitative Sudomotor Axon Reflex Test Physiologic Basis Innervation of human eccrine sweat glands is mainly by sympathetic postganglionic cholinergic fibers.105 Sato149 made additional key observations on isolated human eccrine sweat gland regulation. The regulation of sweating is cholinergic and muscarinic because it is completely inhibited by atropine. The muscarinic receptor subtype is M3.174 Histochemical labeling studies show prominent innervation with vasoactive intestinal polypeptide and calcitonin gene–related peptide fibers and presence of substance P and tyrosine hydroxylase fibers.103 In vitro studies suggest the following rank order of

sudorific effect: acetylcholine  epinephrine (  ) 艐 isoproterenol ()  phenylephrine (). There is, however, additional innervation, with adrenergic fibers present as a loose network of catecholaminecontaining nerves around sweat glands103 in both humans and rodents. In rodents, innervation is initially completely adrenergic followed by an adrenergic-to-cholinergic switch during development.93 It is not known if such a switch occurs in humans during development or with neuropathy and fiber regeneration. A number of observations suggest that adrenergic innervation can increase with disease in humans. Human sweat glands respond to intradermal and intra-arterial adrenaline (10% of cholinergic).148 There is evidence of both -adrenergic (blocked by dibenaline) and -adrenergic (blocked by propranolol but not phentolamine) mechanisms. Prominent hyperhidrosis and adrenergic sensitivity occur in neuropathies characterized by distal burning pain166 and CRPS I,147 suggesting increased adrenergic sudomotor innervation. Some recent support for this notion derives from the demonstration of enhanced adrenergic sweating in CRPS I.27 It is currently suspected that distal small fiber neuropathy might be associated with adrenergic sweating, although the hypothesis (that fiber regeneration recapitulates development, resulting in an upregulation of -adrenergic receptors and responsiveness) is currently unproven. The principle of the test can be surmised from Figure 44–1. The neural pathway consists of an axon “reflex” mediated by the postganglionic sympathetic sudomotor axon. The axon terminal is activated by acetylcholine. The receptor is presumably nicotinic because it is activated by nicotine and acetylcholine but not pilocarpine.105 Table 44–3 shows that iontophoresis followed by drying of the skin and then recording from the identical areas with acetylcholine, nicotine, or pilocarpine results in a large sweat response in the same (C) compartment of the sudorometer (direct response). Pilocarpine directly stimulates muscarinic receptors and fails to generate a response in compartment A (axon reflex or indirect response). The impulse travels antidromically, reaches a branch point, then travels orthodromically to release acetylcholine from the nerve terminal. Acetylcholine traverses the neuroglandular junction and binds to M3 muscarinic receptors on eccrine sweat glands99,100 to evoke the sweat response. Electrical stimulation alone can generate an axon reflex response, but requires a painful stimulus.94 Acetylcholinesterase in subcutaneous tissue cleaves acetylcholine to acetate and choline, resulting in its inactivation and cessation of the sweat response. When carbachol, an unsubstituted carbamyl ester and hence not cleaved by cholinesterase, is substituted for acetylcholine, a persistent sweat response ensues.95 A key component of QSART is the multicompartmental sweat cell (Fig. 44–2). Compartment C is the stimulus compartment and can be loaded with acetylcholine via

1107

Quantitation of Autonomic Impairment

FIGURE 44–1 Left, The neural substrate for the axon reflex sweat response (see text). Right, A representative axon reflex sweat response. (From Low, P. A., Opfer-Gehrking, T. L., and Kihara, M.: In vivo studies on receptor pharmacology of the human eccrine sweat gland. Clin. Auton. Res. 2:29, 1992, with permission.)

Table 44–3. Sweat Volumes Recorded from the Same (Direct) and Distant (Axon Reflex) Compartments in QSART Agent

Receptor

N

Direct (␮L/cm2)

Axon Reflex (␮L/cm2)

Pilocarpine Acetylcholine Nicotine

Muscarinic Nicotinic, muscarinic Nicotinic

6 6 6

3.94  0.58 3.02  0.36 3.37  0.28

0 2.61  0.31 4.30  0.5

FIGURE 44–2 The sudorometer and attachments (see text for details). (From Low, P. A.: Quantitation of autonomic function. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 729, 1993, with permission.)

1108

Quantitating Symptoms, Impairments, and Outcomes

cannula E and connected to the anode of a constant current generator via post F. Compartment A is separated from C by compartment B and two narrow ridges, which block diffusion but not the axon “reflex.” The infiltration of local anesthetic into this air gap will block the axon-reflex response.100 The sweat output in A is evaporated by low constant humidity, air, or gas. The gas flow of altered thermal mass and humidity returns to the sudorometer, where the altered humidity (measured) and flow rate permit translation to sweat volume. Post G permits the attachment of the sudorometer to the limb by straps. The stimulus is a constant current of 2 mA applied for 5 minutes, and the sweat response is recorded during the stimulus and for 5 subsequent minutes. The solution consists of 10% acetylcholine loaded into the outer compartment (C). This is a concentration that was demonstrated, in a dose-response study, to be a supramaximal concentration.105 Reproducibility The tests are sensitive and reproducible in controls and in patients with neuropathy.99,100 Tests repeated on two different days regress with a high coefficient of regression (0.95). Tests repeated daily to the identical site may evoke local skin alterations, possibly to the sweat duct after about the third or fourth repetition, but this “tolerance” is highly variable. The coefficient of variation is 14.7%. Regression of volumes done on day 1 versus day 2 demonstrates a correlation of 0.9.99,100 Commercial Equipment A unit called QSWEAT, manufactured by WR Medical Company in St. Paul, Minnesota (no relationship to Mayo Foundation), has become available. It utilizes an identical setup with the following changes: (1) it uses dehumidified room air instead of nitrogen gas; (2) commercial stimulators are used; and (3) the recommended stimulus for the QSWEAT unit is acetylcholine gel, although it will accept the solution. Recording Sites The recordings are symmetrical so that, in normal individuals, the left side is not significantly different from the right (see Normative Data later). We routinely record from the left but will study the right side when clinically warranted (e.g., following left sural nerve biopsy or with unilateral symptoms). Standard recording sites are the medial forearm (75% of the distance from the ulnar epicondyle to the pisiform bone), the proximal leg (lateral aspect, 5 cm distal to the fibular head), the distal leg (medial aspect, 5 cm proximal to the medial malleolus), and the proximal foot (on a flat surface over the extensor digitorum brevis muscle). The innervation of the forearm, proximal leg, distal leg, and proximal foot are by ulnar, peroneal, saphenous, and sural (mainly) nerves, respectively.

Interpretation A normal test indicates integrity of the postganglionic sympathetic sudomotor axon. An absent response indicates a lesion of the axon, providing iontophoresis is successful and eccrine sweat glands are present. Because the axonal segment mediating the axon is likely to be short, the test likely evaluates relatively distal function.105 A reduced or absent sweat response indicates postganglionic sympathetic sudomotor failure.99,100 In the peripheral neuropathies, the distribution of sweat responses is particularly important. In a distal small fiber neuropathy, the most distal site or two alone may be reduced or absent.165 As the disease progresses, sudomotor failure advances to more proximal sites. In preganglionic or central disorders, QSART is unimpaired. However, with increasing duration of the preganglionic lesion, QSART may become impaired, suggestive of a transsynaptic defect or sweat gland atrophy.149 Used in conjunction with the TST, QSART can differentiate a prefrom a postganglionic lesion (Table 44–4). The presence of resting sweat activity (RSA) indicates sweat gland activity at a skin temperature that is normally subthreshold for the sweat response at room temperature. Because there is little circulating acetylcholine, the mechanism is likely that of spontaneously active sympathetic sudomotor axons. Patients with RSA often have an early or painful neuropathy with burning dysesthesias, vasomotor changes such as cyanosis or pallor, and excessive sweating. However, it can also occur in anxious individuals or those with essential hyperhidrosis. The presence of persistent sweat activity (PSA), defined as sweat volume that does not decrement when the stimulus is turned off, likely indicates pathologic firing of sympathetic axons, because it is unlikely that continuous activity occurs at the level of the sweat gland. It is known that the damaged axon may fire spontaneously and on stimulation and may have more persistent activity than normal axons. The most common clinical situations associated with PSA are mild or painful neuropathies. The latency is sometimes markedly reduced in the context of painful neuropathies with RSA and PSA. In QSART, transmission time along the axon takes milliseconds (axonal conduction velocity is about 1 m/s), and transmission delay at the neuroglandular junction is also measured in milliseconds. Latency from receptor occupancy to sweat secretion takes seconds. Hence the major mechanism of the long latency of 0.5 to 2 minutes is the time needed to iontophorese acetylcholine. The mechanism of the very short latency (sometimes 10 to 30 seconds) is Table 44–4. Defining Site of Sympathetic Lesion QSART

TST

Site of the Lesion

Abnormal Normal

Abnormal Abnormal

Postganglionic Preganglionic

Quantitation of Autonomic Impairment

unknown. These patients often have allodynia, so that the likely mechanism is that of an augmented somatosympathetic reflex.

Evaluation of Cardiovagal Function Using Heart Rate Recordings For heart rate recordings, the patient lies supine on the tilt table for 20 minutes before the commencement of HUT. The optimal location for placement of electrodes is at sites where movement artifacts are minimal during the autonomic maneuvers. In our experience, the simplest and best electrode sites are the interscapular areas just medial to the tips of the scapula. An alternative placement is at the subclavicular areas, at the hollows at the lateral ends of the clavicles. The reference electrode site is not critical. We use the left mid-axillary line at nipple level. The R-R interval is converted to heart rate, which is displayed continuously on our computer console. Heart Rate Response to Deep Breathing HRDB is affected by a number of confounding variables, as summarized in Table 44–5. It is unaffected by the time of day, the sex of the subject, and the amount of preceding rest in relaxed subjects. The two major factors that affect HRDB are age and rate of breathing. The effects of age on HRDB are shown in Figure 44–3. The rate of breathing has a profound influence on the R-R variation. The variation is maximal at a breathing rate of 5 to 6 breaths/min.5,131 The optimal procedure is to have the subject breathe maximally at a rate of 6 breaths/min (inspiratory and expiratory cycles of 5 seconds each). The subject is taught to establish inspiratory and expiratory rhythms so that breathing is smooth and maximal. The subject is instructed to follow an oscillating bar with a period of 10 seconds. Eight cycles are recorded.108 Following 5 minutes of rest, the procedure is repeated. The five largest consecutive responses are read from the computer using a cursor, these responses are averaged, and the heart rate range (maximum  minimum) is derived. Hyperventilation can result in a progressive reduction in heart rate range, presumably as a result of hypocapnia. Sympathetic activation will attenuate heart rate variability.

Table 44–5. Some Factors That Affect the Heart Rate Response to Deep Breathing 1. 2. 3. 4. 5. 6. 7. 8.

Age Rate of breathing Hypocapnia Influence of sympathetic activity Position of the subject Medications Depth of breathing Obesity

1109

The position of the subject and depth of breathing have only modest effects on HRDB. Respiratory volumes are not routinely recorded because, above readily achievable tidal volumes, depth of respiration does not significantly influence heart rate variability.57 There is some variation among laboratories regarding the number of responses to average, whether to include the first response, and which responses to include. Bennett et al.8 considered the first response, which is often larger, preferable to an averaged value. With continued hyperventilation, there is hypocarbia11 with inhibition of sinus arrhythmia. Weight gain of 10% has been reported to reduce, and weight loss to increase, parasympathetic function.70 There is also some variability among individual responses, and some responses are marred by artifact or ectopic beats. Based on these considerations, we have standardized our testing by routinely recording eight cycles and determining the mean of the five consecutive largest responses. We find this approach preferable to the mean of all or the last few responses. We reject the first response if it is more than double the subsequent responses. Control values should be established for the particular laboratory. This approach is better than the adoption of some generic control range. The control values currently used in the Mayo Autonomic Reflex Laboratory are described under Normative Data later. Valsalva Ratio For determination of the VR, the subject, rested and recumbent, is asked to maintain a column of mercury at 40 mm Hg (not exceeding 50 mm Hg) for 15 seconds via a bugle with an air leak (to ensure an open glottis). The responses should be repeated until two responses of similar BPBB and heart rate responses are obtained. We will perform up to four maneuvers. The factors that affect the VR are shown in Table 44–6. The VR is derived from the maximum heart rate generated by the Valsalva maneuver divided by the lowest heart rate occurring within 30 seconds of peak heart rate.106,108 The choice of 40 mm Hg is made because a pressure of 40 mm Hg seems to yield reproducible results, whereas a VR below 20 mm is inadequate and above 60 mm results in less reproducibility.7 Our studies also suggest 40 mm Hg as a practical optimal pressure. It is important that BPBB recordings also be made during the maneuver, because the heart rate responses are baroreflex responses to changes in BP. The increase in heart rate occurs in response to the fall in BP, and the baroreflex response to the overshoot is responsible for the transient bradycardia. In dysautonomic patients, there typically is a loss of both the BP overshoot and the reflex bradycardia. Our criterion for an adequate Valsalva maneuver is the generation of two reproducible BP curves. It should be noted that identical expiratory pressures may result in very different BP curves. If the BP excursions are minimal, or simply follow the configuration of the expiratory pressure (“flat-top” response), then a

1110

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 44–3 Heart rate response to deep breathing. There is a significant effect of age. (From Low, P. A., Denq, J. C., Opfer-Gehrking, T. L., et al.: Effect of age and gender on sudomotor and cardiovagal function and blood pressure response to tilt in normal subjects. Muscle Nerve 20:1561, 1997, with permission.)

reduced VR is not meaningful. If a fall in BP (early phase II) is not generated (as might occur in patients with a flat-top response), then the VR may also be spuriously low. A simple approach to handling VR and the flat-top response is to repeat the study at 30 degrees of HUT. This angle reduces thoracic venous volume, thereby nulling excessive venous volume, the mechanism of the flat-top response. Additionally, a normal VR may occur in patients with impaired cardiovagal function, if sympathetically mediated heart rate alterations are sufficiently large to result in increases and reductions in heart rate.7,179 Response to Standing A detailed evaluation of the phases and mechanisms of the response to standing has been reported.12,52,158 The cardiovascular responses to standing have been analyzed in detail by Smith et al.158 These workers divided the response into two phases, an immediate phase (0 to 30 seconds) and the stable period (30 seconds to 20 minutes). The immediate response in healthy young adults is characterized by a sharp decrease in BP and total systemic resistance at 5 to 10 seconds followed by a rapid rebound and overshoot. There is a corresponding heart rate increment, which

Table 44–6. Some Factors That Affect the Valsalva Ratio 1. 2. 3. 4. 5. 6. 7. 8.

Age Gender Position of the subject Expiratory pressure Duration of effort Inspiratory volume Volume status Medications

follows in 3 to 5 seconds and then tapers. Over the first 30 seconds, there is a steady parallel decline of thoracic blood volume and stroke volume; there is also an initial surge of cardiac output followed by a steady decrease. During the stabilized response (30 seconds to 20 minutes), the hemodynamic variables are relatively steady, showing average increases in heart rate of about 15% to 30%, in diastolic blood pressure (DBP) of 10% to 15%, and in total vascular resistance of 20% to 40%. During the 0.5 to 20 minutes, there are also decreases in thoracic blood volume averaging about 25% to 30%, in cardiac output averaging 15% to 30%, and in pulse pressure averaging about 5% to 10%. It is evident that, in normal human subjects, assumption of the upright posture results in profound hemodynamic changes, most of them occurring during the first 30 seconds. The initial heart rate responses to standing consist of a tachycardia at 3 and then 12 seconds followed by a bradycardia at 20 seconds. The initial cardioacceleration is an exercise reflex and the subsequent tachycardia and bradycardia are baroreflex mediated. The 30:15 ratio (R-R interval at beat 30)/(R-R interval at beat 15) has been recommended as an index of cardiovagal function. However, the test is inferior to HRDB because of its more complex physiology and because the confounding variables are less well clarified. Comparison of Different Tests of Cardiovagal Function There has been much enthusiasm in evaluating the sensitivity of various tests of cardiovagal function. Ziegler et al.198 evaluated cardiovagal function in 261 patients with diabetes of different severity. They used spectral analysis, vector analysis, and standard tests of heart rate variation and concluded that the most frequently abnormal indices were the coefficient of variation, mid-frequency spectral power at

Quantitation of Autonomic Impairment

rest, mean circular resultant, and max/min 30:15 ratio and VR. Much of this emphasis on sensitivity is misplaced. Tests of cardiovagal function are almost all sensitive, with little to choose among them. The HRDB is probably preferable to most others: both afferent and efferent pathways are vagal, most patients are able to cooperate with the procedure, and the confounding variables are well studied and easy to detect. Other procedures, such as the Valsalva maneuver, standing up, and squat, have complex physiology involving sympathetic and central mechanisms as well.

Beat-to-Beat Blood Pressure Recordings The Finapres technique (and its successor, the Finometer) is a noninvasive method of measuring BPBB. It generates an arterial waveform that is indistinguishable from that of a peripheral arterial waveform. The principle of the technique is based on servoplethysmometry using the volume clamp technique devised by Penaz129 and further developed by Wesseling et al.190,191 An infrared sensor that records finger volume (plethysmograph) is mounted inside a finger cuff. The blood volume seen by the plethysmograph is clamped to a set point value by cuff pressure by means of an electropneumatic wide-band servosystem. This computerized servo or feedback system continuously counterbalances intra-arterial pressure to keep the pressure difference across the arterial wall (the transmural pressure) at zero. At zero transmural pressure, cuff pressure equals intramural arterial pressure. The Finapres-recorded pressures have been calibrated against intra-arterial pressures and accurately reflect intrabrachial73,177,178 or radial128 BP. The calibration appears to hold during autonomic maneuvers. There are some circumstances (intense vasoconstriction or marked vasodilation) when the concordance is less secure. When digital vessels are vasodilated, as in exercise when arteriovenous shunts open up, Finapres may under-read.124 Finapres recordings as described above are reliable under optimal conditions. Our recordings are done from the middle digit of the index finger, unless otherwise stated. Criteria we use are agreement (within 15%) with manual recordings of systolic blood pressure (SBP) and DBP and morphology. An adequate recording should have sharp contours and show a dicrotic notch. It should be recognized that there are a number of confounding variables in the noninvasive recording of BPBB from digital arteries. An alternative approach is that of radial artery tonometry using devices similar to the Colin 7000 and Colin Pilot (Colin Medical Instruments; these devices have been discontinued). Radial artery tonometry provides continuous measurement of noninvasive arterial pressure by a sensor positioned above the radial artery. An inflatable upper-arm cuff enables intermittent oscillometric calibration. The Colin has a long track record of noninvasive BP recordings,78 and there is good general agreement between intraarterial and tonometric pressures, even during controlled

1111

hypotension. As a method for tracking BPBB changes, the device is quite satisfactory.189 Recording from the larger radial artery is an advantage over recordings from a digital artery. The disadvantage of radial artery tonometry is the more time-consuming nature of the application and the ease of losing the signal during HUT. The ability of the device to track BPBB during autonomic maneuvers is less well documented than with Finapres. Beat-to-Beat Blood Pressure Response to the Valsalva Maneuver BPBB responses to the Valsalva maneuver and HUT are recorded simultaneously with the heart rate recordings. The dynamic alterations during tilt and the Valsalva maneuver are particularly important in detecting adrenergic failure. There are four main phases in the Valsalva maneuver. In phase I, there is a transient rise in BP resulting from increased intrathoracic and intra-abdominal pressure causing mechanical compression of the aorta.30 In early phase II (phase IIE), the reduced preload (venous return)18 and reduced stroke volume14 lead to a fall in cardiac output in spite of tachycardia caused by a withdrawal of cardiovagal influence.18 Total peripheral resistance (TPR) increases41,90 as a result of efferent sympathetic discharge to muscle41 and an increase in plasma norepinephrine concentration.143 Within 4 seconds after the increase in sympathetic discharge, the fall in BP is arrested.41 This is late phase II (IIL). In normal subjects phase IIL is so efficient that, by the beginning of phase III, mean arterial pressure (MAP) is at the resting MAP level or above in most normal subjects. Phase III, like phase I, is mechanical, lasting 1 to 2 seconds, during which BP falls. The major mechanism is the sudden fall in intrathoracic pressure. Additional factors may be an increase in left ventricular afterload17 and sudden expansion of intrathoracic vessels.46 There is a further burst of sympathetic activity during this phase.184 In phase IV, venous return192 and cardiac output14 have returned to normal while the arteriolar bed remains vasoconstricted, hence the overshoot of BP above baseline values. In the clinical autonomic laboratory setting, with studies done on patients lying supine, phase IV may be more dependent on cardiac adrenergic tone than on systemic peripheral resistance.146 Intravenous phentolamine (10 mg) resulted in the expected elimination of phase IIL, but augmented rather than blocked phase IV. In contrast, 10 mg intravenous propranolol completely blocked phase IV (Fig. 44–4). Valsalva-type maneuvers are performed during activities of daily living. These include straining at stool, abdominal compression, and micturition. In patients with generalized autonomic failure, these maneuvers can result in a progressive fall in BP, and syncope may occur because the baroreflexes are in abeyance. Indeed, the hypotension may persist for over a minute beyond the maneuver. In Figure 44–5, examples of sudomotor sympathetic failure, borderline OH,

1112

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 44–4 Blood pressure response to phentolamine and propranolol. Phentolamine augments and propranolol blocks phase IV of the Valsalva maneuver. (From Sandroni, P., Benarroch, E. E., and Low, P. A.: Pharmacologic dissection of components of the Valsalva maneuver in adrenergic failure. J. Appl. Physiol. 71:1563, 1991, with permission.)

FIGURE 44–5 Blood pressure and heart rate response to the Valsalva maneuver in patients with increasing severity of autonomic failure. BOH borderline orthostatic hypotension; OH orthostatic hypotension; SSF sudomotor sympathetic failure. (From Sandroni, P., Benarroch, E. E., and Low, P. A.: Pharmacologic dissection of components of the Valsalva maneuver in adrenergic failure. J. Appl. Physiol. 71:1563, 1991, with permission.)

and OH show a progressive increase in abnormalities in phases II and IV. The use of the phases of the Valsalva maneuver to evaluate adrenergic function has been validated in two ways. First, we undertook pharmacologic dissection of the maneuver.146 Phase IIL is primarily under peripheral (-) adrenergic control, being selectively blocked by phentolamine, whereas phase IV is completely blocked by propranolol, indicating  adrenoreceptor dependence. Next, we evaluated the maneuver in a control group and three ageand sex-matched patient groups with graded adrenergic failure.146 One group had generalized autonomic failure with an orthostatic fall in SBP during tilt of greater than 30 mm Hg. A second group had a lesser orthostatic fall in BP (10 to 30 mm Hg), and a third group had well-documented peripheral (distal) autonomic failure (absent QSART responses) but did not have OH. In contrast to controls, all the patient groups including the third group exhibited a significant reduction in phase IIL. An excessive BP fall in phase II and an absent phase IV overshoot were observed in the group with florid OH. Intermediate changes were seen in the borderline OH group. The BPBB changes during the Valsalva maneuver, when coupled with BP responses to tilt, provide a significantly improved evaluation of adrenergic failure when compared with bedside BP recordings. The patient with peripheral adrenergic failure, as will the neuropathic patient with involvement of autonomic C fibers, will have an absent phase IIL and some increase in phase IIE. The patient with more severe peripheral autonomic failure, involving more widespread limb and splanchnic adrenergic fibers, will have an excessive phase IIE, absent phase IIL, but intact phase IV. When phase IV is absent, cardiac adrenergic innervation has failed. One caveat is that a small or, less often, absent phase IV can occur in normal individuals if phase II is modest or absent. Another caveat is the fact that these phases reflect end organ function. Reduced or absent phase IIL indicates peripheral adrenergic denervation or poor vasomotor tone. Beat-to-Beat Blood Pressure Response to Head-up Tilt Orthostatic BPBB recordings in response to HUT are recorded using a sphygmomanometer cuff with the patient supine and following tilt to 70 degrees. A 70-degree angle provides the best compromise between maximizing orthostatic stress and minimizing the effects of muscle contraction (the latter is activated by standing up). An angle below 60 degrees generates insufficient orthostatic stress.162 We currently use an automated tilt table. The current emphasis is a slow tilt (over 10 to 20 seconds); however, the rate of tilt does not apparently affect the responses.162 Cuff recordings are obtained at 1 and 5 minutes after tilt-up. It is important to perform the upright tilt procedure at a standard time after the patient lies down because the orthostatic reduction in BP is greater following 20 minutes of preceding rest than 1 minute, increasing from a mean systolic/diastolic decrement of 8/9 to 17/19 mm Hg.171 We routinely perform tilting

Quantitation of Autonomic Impairment

at the end of the study, after approximately 45 minutes in the supine position. BPBB recordings of SBP, mean BP, and DBP (acquired by Finapres) are continuously displayed on the computer console, as is the heart rate, derived from electrocardiographic (ECG) leads and an ECG monitor. It is also important to ensure that the arm is at heart level because arm position influences measurements of BP.185 Several methods have been used to maintain proper arm position. The Finapres-containing digit can be held at a fixed position at heart level at the anterior axillary line.171 A second method is to perform the supine recordings with the arm at heart level and then, following upright tilt, to extend the arm to rest it on an armrest, again at heart level. We prefer a third method in which the arm is abducted onto an armrest and remains at heart level at all angles of tilt. It appears that, in some patients, abduction causes compression of major vessels. If the pulse contour becomes smaller with abduction to heart level, it is preferable and permissible to lower the arm to 2 inches below heart level. During HUT, normal individuals undergo a transient reduction in SBP, mean BP, and DBP followed by recovery within 1 minute.74 The decrement is typically modest (10 mm Hg mean BP) but can be larger in patients who are deconditioned or have poor vasomotor tone. Patients with adrenergic failure have a marked and progressive reduction in BP and pulse pressure (see Fig. 44–5). Some hypovolemic patients or patients with POTS will undergo a large fall in BP with recovery in the first half-minute. The heart rate response is typically attenuated, but in patients whose cardiac adrenergic innervation is spared, heart rate response is intact and may be increased. A number of indices of mild adrenergic impairment have become recognized. These include excessive oscillations in BP, an excessive reduction ( 50%) in pulse pressure, a transient (first minute) reduction in SBP of 30 mm Hg or greater, an excessive increment in heart rate ( 30 beats/min), and a failure of total systemic resistance to increment. Premonitory signs of syncope are a progressive reduction in BP (especially DBP), TPR, and pulse pressure and the loss of BP (and heart rate) variability. Some of these indices are expected abnormalities in a failure of arteriolar vasoconstriction (TPR, DBP). Some are signs of increased vascular capacitance (reduction in pulse pressure, excessive heart rate increment). The increased oscillations are indicative of intact compensatory mechanisms (but are abnormal because they indicate a system under stress), whereas the gradual loss of variability indicates the failure of compensation.

Composite Autonomic Severity Score A very large autonomic database is now available (see Normative Data later) in which it is clear that the normal response to autonomic testing varies by age and gender. To facilitate comparison, we have developed a 10-point Composite Autonomic Severity Score (CASS; Table 44–7) of

1113

Table 44–7. Composite Autonomic Severity Score (CASS) Sudomotor Subscore 0 Normal 1 Any of the following alterations: a. Single QSART site reduced but 50% of lower limit or b. Distal sweat volume  1/3 of forearm or proximal leg values or c. TST anhidrosis present but  25% 2 Any of the following alterations: a. Single QSART site  50% of lower limit b. 2 QSART sites reduced c. TST anhidrosis 25%–50% 3 a. 2 or more QSART sites  50% of lower limit b. TST % anhidrosis 50% Cardiovagal Subscore 0 Normal 1 HRDB or VR reduced but  50% of the lower limit of normal 2 HRDB or VR reduced to  50% of the lower limit of normal 3 Both HRDB and VR reduced to  50% of the lower limit of normal Adrenegic Subscore The adrenergic subscore is based on alterations in blood pressure in response to the Valsalva maneuver and HUT 0 Normal 1 The presence of any of the following changes on the Valsalva maneuver: a. Phase IIE reduction 30 mm Hg MBP or b. Phase IIL does not return to baseline or c. Pulse pressure reduction to  50% of baseline The presence of any of the following changes occur on HUT: d. Excessive oscillations in MBP ( 20 mm Hg occupying at least 50% of HUT) e. Fall in pulse pressure 50% f. Transient fall in SBP 40 mm Hg with recovery (within 1 min) g. SBP reduction 20 mm Hg beyond 1 min h. DBP reduction 10 mm Hg beyond 1 min 2 If a score of 1 is determined from the Valsalva maneuver, it can be increased to 2 if the following changes occur on HUT: a. Transient fall in SBP  30 mm Hg with recovery within 2 min b. SBP reduction 20 mm Hg beyond 1 min c. DBP reduction 10 mm Hg beyond 1 min 3 A score of 3 is assigned if the following changes occur on the Valsalva maneuver: a. Phase IIE reduction  40 mm Hg MBP  absent phases IIL and IV 4 An additional point is assigned if a reduction in manual SBP 30 mm Hg occurs beyond 2 min and is sustained for at least 2 min DBP diastolic blood pressure; HRDB heart rate response to deep breathing; HUT head-up tilt; MBP mean blood pressure; QSART quantitative sudomotor axon reflex test; SBP systolic blood pressure; TST thermoregulatory sweat test; VR Valsalva ratio.

1114

Quantitating Symptoms, Impairments, and Outcomes

autonomic function. The scheme allots 4 points for adrenergic and 3 points each for sudomotor and cardiovagal failure.97 Each score is normalized for the confounding effects of age and sex. Patients with a score of 3 or less on the CASS have only mild autonomic failure, those with scores of 7 to 10 have severe failure, and those with scores between these two ranges have moderate autonomic failure. The sensitivity and specificity of the method were assessed by evaluating CASS in four groups of patients with known degrees of autonomic failure: 18 with MSA, 20 with autonomic neuropathy, 20 with PD, and 20 with peripheral neuropathy but no autonomic symptoms. The composite scores (means  standard deviation) for these four groups, respectively, were as follows: 8.5  1.3, 8.6  1.2, 1.5  1.1, and 1.7  1.3. Patients with symptomatic autonomic failure had scores of 5 or more, those without symptomatic autonomic failure had scores of 4 or less, and no overlap existed in these groups.97

and gender is shown in Table 44–8. Our normative database by test is

Normative Data A large normative database is available in the Mayo Autonomic Reflex Laboratory. We recently reviewed our normative data in 557 normal subjects. Distribution by age

HRDB. Normative data who were evaluated distribution by gender numbers of subjects for

QSART: 357 subjects QSART (left vs. right sides): 39 subjects Deep breathing: 376 subjects Resting sweat output: 43 subjects Valsalva ratio: 425 subjects Orthostatic BP/heart rate: 270 subjects QSART. Normative data were available on 357 subjects. A consistent gender difference was found, with females having approximately one half the sweat volumes of males. A series of regressions are provided in Table 44–9 that describe the normative data by age and gender. The mean, 5th, and 95th percentile values are given in Table 44–10 for males and in Table 44–11 for females. were evaluated in 376 subjects for HRDB. Breakdown and were approximately equal. The ages 29 or fewer, 30 to 39, 40 to

Table 44–8. Age and Gender of Mayo Autonomic Reflex Laboratory Database Age (yr) ⱕ20

21–30

31–40

41–50

51–60

61–70

ⱖ70

Total

22 24 46

54 57 111

49 56 105

46 58 104

46 45 91

39 28 67

14 19 33

270 287 557

Male Female TOTAL

Table 44–9. Regression of QSART Volumes with Age for Males and Females for Standard Sites in Human Subjects* Intercept Variable Forearm Proximal leg Distal leg Proximal foot

Gender

Age

b0

P

b1

P

b2

P

Gender: b3

4.5291 4.0209 5.5929 4.5174

0.0001 0.0001 0.0001 0.0001

1.6188 1.0076 1.5921 1.5171

0.0001 0.0026 0.0001 0.0001

0.0072 0.0163 0.0437 0.0310

0.5752 0.1410 0.0002 0.0066

0.002 0.1410 0.0075 0.0101

Age: P 0.7989 0.0011 0.3049 0.1548

*Based on a normative database of 357 normal subjects evenly distributed by age and gender.

Table 44–10. Male QSART Responses: Mean, 5th, and 95th Percentile Values 10–29 yr Sites Forearm Proximal leg Distal leg Proximal foot

⬎50 yr

30–49 yr

Mean

5th

95th

Mean

5th

95th

Mean

5th

95th

2.67 2.67 3.28 2.58

0.76 1.27 1.37 0.87

5.06 4.54 5.27 4.48

2.67 2.32 2.55 2.17

0.76 0.93 0.98 0.78

5.06 4.19 4.55 4.07

2.67 1.97 1.83 1.75

0.76 0.58 0.59 0.68

5.06 3.84 3.82 3.65

1115

Quantitation of Autonomic Impairment

Table 44–11. Female QSART Responses: Mean, 5th, and 95th Percentile Values 10–29 yr Sites Forearm Proximal leg Distal leg Proximal foot

⬎50 yr

30–49 yr

Mean

5th

95th

Mean

5th

95th

Mean

5th

95th

1.15 1.48 1.83 1.27

0.20 0.36 0.61 0.23

2.78 3.17 2.85 3.07

1.15 1.48 1.26 1.05

0.20 0.36 0.39 0.18

2.78 3.17 2.28 2.85

1.15 1.48 0.68 0.84

0.20 0.36 0.18 0.12

2.78 3.17 1.70 2.64

Table 44–12. Heart Rate Response to Deep Breathing: 2.5th, 5th, 95th, and 97.5th Percentile Values by Age

49, 50 to 59, 60 to 69, and 70 or more years were 91, 81, 67, 60, 48, and 29, respectively. Ages ranged from 10 to 83 years and no significant differences were found between the sexes. However, a significant regression with age was found (y 37.5448* log10 0.9832x; P  .001, where y heart rate range in beats/min and x age in years). The values for the 2.5th, 5th, 95th, and 97.5th percentiles for ages 11 to 29, 30 to 49, 50 to 69, and 70 or greater are shown in Table 44–12.

60 to 69, and 70 or more years were 110, 85, 80, 67, 53, and 30, respectively. The effect of age on VR was calculated for males (N 205) as y 2.15982  0.00755x (P  .001) and for females (N 220) as y 2.00273  0.00868x  0.00021x2, where y VR and x age in years. The less consistent effect of age on VR than on respiratory sinus arrhythmia likely relates to the smaller change and greater complexity of the maneuver. Whereas respiratory sinus arrhythmia is a relatively pure test of cardiovagal function, many factors, including blood volume, antecedent period of rest,161 cardiac sympathetic functions, peripheral sympathetic functions, and norepinephrine response affect the maneuver. Age may affect different components of the Valsalva maneuver in different directions. The values for the 2.5th, 5.5th, 95th, and 97.5th percentiles for ages 11 to 29, 30 to 49, and 50 or greater are shown in Table 44–13. The values up to the age of 50 years are robust. Beyond age 50, the points are too few and a curve was not fitted.

The Valsalva Ratio. There is a progressive reduction of the VR with increasing age.75,98,101 Ingall et al.75 reported a slope of 0.01 per year, which is very similar to ours. VR, studied in 425 subjects ages 10 to 83 years, showed a gender difference. Hence the data for males and females are described separately (Figs. 44–6 and 44–7). Distribution by gender was even (male 205; female 220). The numbers of subjects for ages 29 or fewer, 30 to 39, 40 to 49, 50 to 59,

Orthostatic BP. A total of 270 subjects (male 129; female 141) were studied. Distribution by gender was even. The numbers of subjects for ages 29 or fewer, 30 to 39, 40 to 49, 50 to 59, 60 to 69, and 70 or more years were 54, 64, 52, 49, 32, and 19, respectively. BP and heart rate were recorded beat-to-beat for 5 minutes. Normative data for heart rate increment and BP decrement in response to tilt-up are shown in Tables 44–14 and 44–15, respectively.

Percentile

11–29 yr

30–49 yr

50–69 yr

ⱖ70 yr

2.5; 5.0 95; 97.5

13; 14 41; 43

9; 10 33; 36

7; 7 27; 29

7; 7 27; 29

FIGURE 44–6 Valsalva ratio: effect of age in males. (From Low, P. A., Denq, J. C., Opfer-Gehrking, T. L., et al.: Effect of age and gender on sudomotor and cardiovagal function and blood pressure response to tilt in normal subjects. Muscle Nerve 20:1561, 1997, with permission.)

1116

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 44–7 Valsalva ratio: effect of age in females. (From Low, P. A., Denq, J. C., Opfer-Gehrking, T. L., et al.: Effect of age and gender on sudomotor and cardiovagal function and blood pressure response to tilt in normal subjects. Muscle Nerve 20:1561, 1997, with permission.)

Table 44–13. Valsalva Ratio: 2.5th, 5th, 95th, and 97.5th Percentile Values by Age and Sex 11–29 yr Percentile 2.5; 95;

Males

5.5 97.5

1.50; 2.87;

Females

1.59 2.97

1.18; 2.64;

ⱖ50 yr

30–49 yr Males

1.24 2.88

1.51; 2.87;

Females

1.59 2.97

1.18; 2.64;

Males

1.24 2.88

1.51; 2.87;

Females

1.59 2.97

1.36; 2.41;

1.39 2.65

Table 44–14. Heart Rate Increment after 1 Minute of HUT: 95th and 97.5th Percentile Values for Males and Females by Age 11–29 yr Percentile Male Female

30–49 yr

95th

97.5th

95th

97.5th

95th

97.5th

95th

97.5th

34 31

40 34

30 27

36 30

26 23

32 26

21 19

27 22

Table 44–15. Blood Pressure Decrement After 1 Minute of HUT: 95th Percentile by Age Male, female

ⱖ70 yr

50–69 yr

tracted. Salivation less than 7.5 mL/5 min is considered reduced for adults.

11–29 yr

30–49 yr

50–69 yr

ⱖ70 yr

17, 21

20, 24

23, 26

26, 29

Orthostatic BP decrement increased by age but was not different by gender.

Salivation Test Five gauze pads are deposited within a container and weighed. The subject’s sublingual gutter is wiped with a different gauze pad, which is discarded. The subject then chews on one preweighed gauze pad each minute for 5 minutes. At the end of the 5 minutes, the pads, which have been deposited in the original container, are weighed together with the container and the original weight is sub-

Biochemical and Hormonal Evaluation of Autonomic Activity The major neurochemicals involved in peripheral autonomic activity are acetylcholine, noradrenaline, and adrenaline. Acetylcholine is unstable, unlike noradrenaline and adrenaline. Noradrenaline is the major neurotransmitter of sympathetic nerve endings, and both noradrenaline and adrenaline are released from the adrenal glands. Measurement of catecholamines and their metabolites can be made either in plasma or in urine samples. Measurement of Plasma Catecholamines Plasma noradrenaline, adrenaline, and dopamine can be measured in plasma using sensitive and specific techniques that range from radioenzymatic assays to high-performance

Quantitation of Autonomic Impairment

liquid chromatography. Measurements are made in the basal state and also in response to stimuli that either increase or decrease sympathetic activity. Thus stimuli such as gravitational stress with HUT increase sympathoneural activity and raise plasma noradrenaline levels, and hypoglycemia, a potent adrenomedullary stimulant, raises levels of both plasma adrenaline and noradrenaline.116,136 Measurement usually is made in venous blood samples, although arterial blood provides a more “global” measurement. Measurement in arterialized blood, obtained by warming the hand, provides values comparable to those obtained from arterial measurements. The concentration of noradrenaline in plasma is dependent on a number of processes at nerve terminals that involve formation (Fig. 44–8), secretion, neuronal uptake, intra- and extraneuronal metabolism, and clearance of noradrenaline (Fig. 44–9). This includes uptake into non-neural tissue. Basal supine plasma levels of catecholamines ideally are obtained with the subject supine and resting undisturbed in a quiet environment for a period of at least 5 minutes.

1117

The basal level itself may provide information on disease processes. Thus in PAF, there often are extremely low basal levels, consistent with the postganglionic lesion.64,115 This differs from MSA, with predominantly central preganglionic lesions, in which the basal levels often are within the normal range (Fig. 44–10). However, in tetraplegics with cervical cord transection who also have a preganglionic lesion, the basal levels are low, in the region of 35% of normal.117 Extremely low, virtually undetectable levels of plasma noradrenaline and adrenaline occur when there is deficiency of the enzyme dopamine -hydroxylase.112,115,142 In such patients, a further clue is the elevated plasma level of dopamine, because of the enzymatic block. Sympathoneural activation during standing or HUT is impaired in autonomic failure resulting from either preganglionic (MSA) or postganglionic (PAF) lesions and is reflected in the diminished response in plasma noradrenaline levels (see Fig. 44–10). The rise in plasma noradrenaline levels caused by exercise does not occur in autonomic failure.3,157 Edrophonium, an acetylcholinesterase inhibitor, increases ganglionic transmission

FIGURE 44–8 Biosynthetic pathway for noradrenaline synthesis and the structure of D-threo-3,4-dihydroxyphenylserine (DOPS). The enzyme dopa decarboxylase, which is present both intra- and extraneuronally, converts DOPS to noradrenaline, thus bypassing the hydroxylation step, which depends on dopamine -hydroxylase. (From Mathias, C. J., Bannister, R. B., Cortelli, P., et al.: Clinical autonomic and therapeutic observations in two siblings with postural hypotension and sympathetic failure due to an inability to synthesize noradrenaline from dopamine because of a deficiency of dopamine beta hydroxylase. Q. J. Med. 75:617, 1990, with permission.)

1118

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 44–9 Steps involved in the formation of noradrenaline (NA) from tyrosine within a sympathetic nerve terminal. DA dopamine; DH dopamine -hydroxylase; DDC dopa decarboxylase; DOPA dihydroxyphenylalanine; TH tyrosine hydroxylase. NA in granules is released by a process of exocytosis into the synaptic cleft, following which it acts on various  or  adrenoreceptors, either pre- or postsynaptically. NA is subject to various processes that involve uptake into the nerve terminal (uptake 1), following which it is either incorporated into granules, exerts negative feedback on TH, or is metabolized by monoamine oxidase (MAO). Some is taken up into non-neuronal tissues (uptake 2), some is metabolized by catechol-o-methyltransferase (COMT), and the rest spills over into the circulation. (Modified from Mathias, C. J.: Disorders of the autonomic nervous system. In Bradley, W. G., Daroff, R. B., Fenichel, J., and Jankovic, J. [eds.]: Neurology in Clinical Practice. Boston, Butterworth-Heinemann, p. 2403, 2004.)

and raises levels of plasma noradrenaline in MSA but not in PAF.59 The reverse, inhibition of sympathetic neural activity and suppression of plasma noradrenaline, occurs after administration of the 2 adrenoceptor agonist clonidine.172 A number of factors influence the basal levels of plasma noradrenaline, including age; levels are higher in the elderly. The basal levels of plasma adrenaline often are around the detection limit of most assays. Stimuli such as hypoglycemia raise adrenaline levels in normal subjects; this does not occur in tetraplegics with cervical cord transaction and in autonomic failure.118,136 An elevated basal level of plasma noradrenaline and adrenaline may be indicative of diseases that include pheochromocytoma, with a catecholamine-producing tumor either in adrenal or extraadrenal sites. These levels are not suppressed by clonidine administration, providing biochemical confirmation of autonomous catecholamine secretion (Fig. 44–11). Drugs that impair the neuronal uptake of noradrenaline, such as desipramine and venlafaxine, may result in a falsely positive elevation. Plasma dopamine levels usually are low in the basal state. Levels are elevated in dopamine -hydroxylase deficiency and when L-dopa is used to treat neurologic conditions such as PD, dystonia, and the restless legs syndrome. Urinary Catecholamines and Their Metabolites Measurement of urinary catecholamines and their metabolites provides a measure of secretion over a longer period.132 This may be of value especially in disorders with intermittent secretion, such as a pheochromocytoma that

may be missed with spot plasma catecholamine measurements. The disadvantage is the dependence on the many processes involved in the secretion, metabolism, and excretion of catecholamines. In dopamine -hydroxylase deficiency, urinary metabolites of noradrenaline and adrenaline are virtually undetectable, and dopamine metabolites (homovanillic acid and 3-methoxytyramine) are normal or elevated.115 These patients respond to treatment with the drug D-threodihydroxyphenylserine, which bypasses the enzymatic defect as it is converted by dopa decarboxylase into noradrenaline (see Fig. 44–8); this causes an increase in the urinary metabolites of noradrenaline and adrenaline. In disorders such as PAF, there is an overall reduction of noradrenaline metabolites in the urine, including normetanephrine, formed by the action of extraneuronal catechol-o-methyltransferase on released noradrenaline; this is unlike MSA.88 The metabolite dihydroxyphenylglycol is the result of intraneuronal noradrenaline metabolism; thus levels are low in PAF and in the normal range in MSA. This is in contrast to pheochromocytoma, in which a variety of catecholamine metabolites, including vanillylmandelic acid and 3-methoxy-4-hydroxyphenylglycol, may be markedly elevated. Measurement of Total Body and Regional Noradrenaline Spillover These studies provide quantitative measures of sympathetic neural activity in the whole body or in specific areas such as the splanchnic region, heart, kidneys, and brain.48,92 They are invasive, involving cannulation of arteries and

Quantitation of Autonomic Impairment

1119

FIGURE 44–10 Plasma noradrenaline, adrenaline, and dopamine levels (measured by high-performance liquid chromatography) in normal subjects (controls), patients with multiple system atrophy (MSA), and patients with pure autonomic failure (PAF) while supine and after head-up tilt to 45 degrees for 10 minutes. The asterisk indicates levels below the detection limits for the assay, which are less than 5 pg/mL for noradrenaline and adrenaline and less than 20 pg/mL for dopamine. Bars indicate  standard error of the mean. (Modified from Mathias, C. J., Bannister, R. B., Cortelli, P., et al.: Clinical autonomic and therapeutic observations in two siblings with postural hypotension and sympathetic failure due to an inability to synthesize noradrenaline from dopamine because of a deficiency of dopamine beta hydroxylase. Q. J. Med. 75:617, 1990.)

major veins, with the use of radioactive substances. They can be of particular value in determining whether specific stimuli result in differential regional changes. Thus with food ingestion, there is an expected increase in the splanchnic region but also an increase in sympathetic activity to the renal and skeletal muscle vasculature.180 Their use is restricted to research investigations rather than routine clinical practice.

NONROUTINE STUDIES Imaging of Sympathetic Innervation to Organs The radiolabeled substance 123I-meta-iodobenzylguanadine (MIBG) is taken up into sympathetic nerve endings

and can be measured by single-photon emission computed tomography (SPECT). This technique has been used to image adrenal and also extra-adrenal catecholamine-secreting tumors; it has also been used as a measure of cardiac sympathetic innervation. Cardiac uptake is diminished when there is postganglionic sympathetic damage, as in diabetic cardiac autonomic neuropathy, PAF, and PD.28,29,176 In PD, diminished uptake may occur even when cardiovascular autonomic failure is not present.31,126 These findings differ from MSA, in which uptake is essentially normal, indicating preserved cardiac sympathetic innervation (Fig. 44–12). A similar approach utilizes positron-emission tomography (PET) scanning with 6-[18F]fluorodopamine.61,63 A body of evidence on the utility of cardiac innervation studies has evolved and is described in greater detail in the next section.

1120

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 44–11 Plasma noradrenaline levels in a patient with a pheochromocytoma and in a group of patients with essential hypertension before and after intravenous clonidine (2 mg/kg over 10 minutes), administration of which is indicated by an arrow. Plasma noradrenaline levels fell rapidly in the essential hypertensive patient after clonidine and remained low over the period of observation. The stippled area indicates  standard error of the mean. Plasma noradrenaline levels are considerably higher in the pheochromocytoma patient and are not affected by clonidine. (From Mathias, C. J., and Bannister, R.: Investigation of autonomic disorders. In Mathias, C. J., and Bannister, R. [eds.]: Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System. Oxford, Oxford University Press, p. 169, 1999, with permission.)

MIBG SPECT Study The quantitative uptake of the radiopharmaceutical MIBG, a norepinephrine analogue, as measured by SPECT is an index of the functional integrity of presynaptic sympathetic terminals in the heart. Because ischemia per se may reduce MIBG uptake, the influence of perfusion defects is evaluated by technetium flow studies. Studies in the rat heart demonstrate that MIBG uptake is mainly dependent on uptake-1 and uptake-2 activity40 and a linear relationship between MIBG and norepinephrine content.68 In a single human autopsy study, the severity of nerve fiber loss correlated with the reduction in MIBG uptake.168 Dynamic studies demonstrated6 at least biexponential clearance patterns of MIBG from the heart. These ratios, calculated from dynamic and static studies, are helpful in elucidating the uptake at nonvesicular sites, which reflect the severity of sympathetic nervous system abnormalities in the heart. Normative data on small groups of patients are available. For instance, Tsuchimochi et al.176 evaluated age and gender differences in normal myocardial adrenergic neuronal

FIGURE 44–12 Comparison of early-phase heart:mediastinum (H/M) ratios using [131I]meta-iodobenzylguanidine scintigraphy scanning in different neurologic disorders and in controls. Open circles show normal and filled circles abnormal H/M ratios. C healthy controls; DC disease controls; ET essential tremor; PD Parkinson’s disease; MSA multiple system atrophy; VP vascular parkinsonism. (From Orima, S., Ozawa, E., Nakade, S., et al.: 123I-metaiodobenzylguanidine myocardial scintigraphy in Parkinson’s disease. J. Neurol. Neurosurg. Psychiatry 67:189, 1999, with permission.)

function in 29 subjects (18 men, 11 women; age range, 21 to 79 years; mean age 42  17 years) with no cardiac disorders. Early (15 minutes) and late (3 to 4 hours) planar images and SPECT images at 3 to 4 hours after MIBG injection (111 Mbq) were evaluated. There were no significant differences in the global heart values based on age or gender, but there was a significant inverse correlation between the inferior-toanterior wall count ratio and age, affecting especially inferior wall uptake. Studies have been done on patients with autonomic failure, mainly in diabetic subjects. There is a progressive increase in impairment of MIBG uptake from patients without cardiovagal impairment to those with severe autonomic failure.28,96,113,120,151 Groups with sympathetic failure, on tilt studies or on reduced R-R interval of low-frequency power, tended to have more impaired MIBG uptake.120 A correlation between global or regional myocardial MIBG uptake, however, and duration of diabetes, hemoglobin A1c, body mass index, or Q-T interval length was not observed.151 Patients with a long Q-T interval had more abnormal MIBG uptake,122,188 but a linear correlation was not found.188 Claus et al.28 studied 25 diabetic patients and 19 healthy subjects. MIBG scans in all normal subjects showed homogeneous uptake of activity. In 12 of 25 patients, at least two heart rate variation tests were abnormal, whereas 17 of 25 MIBG SPECT studies were abnormal. No significant correlation was found between MIBG

Quantitation of Autonomic Impairment

1121

SPECT results and spectral analysis of heart rate variation. Schnell et al.151 evaluated 20 diabetic patients with and 22 diabetic patients without cardiovagal neuropathy, and 9 control subjects. Normal MIBG studies were found in only six diabetic patients without (27%), and one diabetic patient with (5%) cardiovagal neuropathy. Severe impairment is usually found in patients with OH. Abnormalities have been reported in PAF, MSA, amyloid neuropathy, and myotonic dystrophy.66,69,110,123 Recent studies have indicated that cardiac sympathetic innervation may be quite different in PD than in MSA.13,61,62,76,126,150,169,196 In the original report,63 Goldstein used 6-[18F]fluorodopamine and PET scanning. Patients with parkinsonism (with OH) had no myocardial 6-[18F]fluorodopamine–derived radioactivity, indicating loss of myocardial sympathetic nerve terminals; in contrast, patients with MSA had increased levels of 6-[18F]fluorodopamine–derived radioactivity, indicating intact sympathetic terminals. Identical information can be derived using cardiac uptake of MIBG, an analogue of norepinephrine that traces the functioning of postganglionic sympathetic adrenergic neurons.13,76,126,150,169,196 The main limitation of the test is its low sensitivity in PD without autonomic failure. If patients with parkinsonism with OH are studied, they will segregate into two groups (low and normal or high MIBG), but it is not clear which of these patients have MSA and which have PD with autonomic failure, because pathologic confirmation is lacking in these studies.

Electrophysiologic Assessment of Sympathetic Neural Activity Sympathetic activity to both skin and muscle can be recorded using fine tungsten electrodes inserted into a cutaneous nerve in the arm or leg. Cutaneous skin sympathetic activity is affected by cold and heat (thermal stimuli), in contrast to muscle sympathetic nerve activity, which is responsive to various maneuvers (such as the Valsalva maneuver) that activate baroreceptors and thus is timelocked to BP changes through the baroreceptor reflex.183 There is an increase in sympathetic neural discharge in the Guillain-Barré syndrome associated with hypertension and tachycardia.55 In high spinal cord lesions during autonomic dysreflexia, when the BP rises rapidly, there is only a relative rise in sympathetic nerve activity, suggesting that other mechanisms may be operative.167 These microneurographic techniques have been of value, especially in the research setting. The electrical potential from electrodes on the skin (the sympathetic skin response; SSR), is dependent on an increase in sudomotor activity and provides a measure of sympathetic cholinergic activity.181 Stimuli that induce an SSR include inspiratory gasp, auditory stimuli, or electrical stimulation of a peripheral nerve (median or peroneal) (Fig. 44–13). The SSR is absent in PAF and in the rare

FIGURE 44–13 The sympathetic skin response (in microvolts) from the right hand and right foot of a normal subject (control) and a patient with dopamine -hydroxylase (DH) deficiency. (From Magnifico, F., Misra, V. P., Murray, N. M., and Mathias, C. J.: The sympathetic skin response in peripheral autonomic failure— evaluation in pure failure, pure cholinergic dysautonomia and dopamine-beta-hydroxylase deficiency. Clin. Auton. Res. 8:133, 1998, with permission.)

disorder pure cholinergic dysautonomia, in which the sudomotor pathways are impaired (Fig. 44–14). The SSR is present in autonomic failure when only the sympathetic adrenergic system is involved; the sympathetic cholinergic system is spared, as in dopamine -hydroxylase deficiency.111 The SSR is dependent on long descending pathways through the spinal cord. In thoracic spinal cord injury, the SSR is present in the palmar region, but in complete lesions is absent in the plantar region (Fig. 44–15).21 The technique thus helps to determine whether descending sympathetic sudomotor pathways in the spinal cord are disrupted and may provide a noninvasive and reproducible autonomic measure of completeness of the lesion affecting such autonomic pathways in spinal cord injury. Its

1122

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 44–14 The sympathetic skin response could not be recorded in two patients, one with pure autonomic failure (PAF) and the other with pure cholinergic dysautonomia. (From Magnifico, F., Misra, V. P., Murray, N. M., and Mathias, C. J.: The sympathetic skin response in peripheral autonomic failure—evaluation in pure failure, pure cholinergic dysautonomia and dopaminebeta-hydroxylase deficiency. Clin. Auton. Res. 8:133, 1998, with permission.)

measurement may refine classification of spinal damage in spinal cord injury patients, in whom classification criteria now include only motor and sensory function.

Skin Vasomotor Reflexes Skin blood flow is measured by a laser Doppler flowmeter or by plethysmography, and the vasoconstrictor response

to an autonomic maneuver is determined. Studies are usually done on the toe or finger pads because the sympathetic innervation to these sites is purely vasoconstrictor, and vasoconstriction can be induced by maneuvers such as inspiratory gasp,42 response to standing (for the finger), contralateral cold stimulus, and the Valsalva maneuver.104 The pathways of these reflexes are complex. For instance, the response to standing is mediated by the

FIGURE 44–15 Superimposed sympathetic skin responses to four right median nerve stimuli in a complete low thoracic spinal cord–injured subject. The stimulus was applied at time 0 (artifact in right-hand trace). Repeatable responses are seen on both sides. The plantar response could not be elicited in either foot. (From Cariga, P., Catley, M., Mathias, C. J., et al.: Organisation of the sympathetic skin response in spinal cord injury. J. Neurol. Neurosurg. Psychiatry 72:356, 2002, with permission.)

Quantitation of Autonomic Impairment

venoarteriolar reflex, by low- and high-pressure baroreceptors, and to a lesser extent by increases in epinephrine, norepinephrine, and renin. The test can be used to detect the presence of sympathetic denervation in the peripheral neuropathies, as in diabetic or amyloid neuropathy.104,140,155 A shortcoming of the test is the marked sensitivity of skin sympathetic fibers to emotional and temperature changes, so that there is much ambient fluctuation.104 More optimistic reports are also available.1,53 Faes et al.53 evaluated reflex cooling and, using their paradigm, found the test to be reproducible. Changes in skin temperature (P  .001) and skin blood flow (P  .005) in response to cooling were significantly larger in the control group than in the group with spinal analgesia. Repeated skin temperature measurements on 42 occasions (testretest period of 4 weeks) in 8 healthy and 34 diabetic subjects indicated a reliability coefficient of 80%. Abbot et al.1 studied fingertip skin blood flow as measured by laser Doppler flowmetry (as LDflux) under environmental conditions promoting vasodilation in Scottish patients with diabetes mellitus and Indian patients with leprosy and reported that the test separated out normal from neuropathic patients satisfactorily. C-fiber innervation of the skin can be evaluated by measuring the change in blood flow in response to iontophoresed vasodilators. For instance, vasodilation in the dorsal foot in response to heating and iontophoresis of acetylcholine (endothelium dependent) and sodium nitroprusside (endothelium independent) was measured using single-point laser Doppler and laser Doppler imaging in diabetic patients.182 Similar studies have been done on the forearm.80 There appears to be a good dose-response relationship to iontophoresed agents, provided the stimulus is not strong enough to cause vasoconstriction.

Orthostatic Stress Tests Some patients with adrenergic failure do not sustain an orthostatic drop in BP on routine upright tilt. However, they may do so after a meal or a warm bath. To elicit orthostatism in these patients, orthostatic stress tests can be undertaken. Four stresses that we have employed are sublingual trinitroglycerin, postexercise tilt, lower body negative pressure, and prolonged tilt. The principle of all four stress tests is the same: The patient is subjected to a vasodilatory stimulus. In the trinitroglycerin stress test,72,106 after routine upright tilt, 0.6 g trinitroglycerin is administered sublingually to the subject, who remains supine. Five minutes later, the tilt is repeated. An alternative approach is to repeat the study after the subject has done 12 squats. Especially with young subjects, extension of the period of upright tilt to 20 minutes or the application of lower body negative pressure164 may induce presyncope or orthostatism that is not found during 5 minutes of tilt.

1123

Blood Pressure and Heart Rate Response to Prolonged Tilt For patients with the “benign” disorders of reduced orthostatic tolerance, postural orthostatic tachycardia syndrome, or recurrent vasodepressor syncope, a prolonged tilt of up to 60 minutes is recommended. The extended time is needed to examine whether orthostatic tachycardia and excessive oscillations in heart rate and BP develop. Vasodepressor and vasovagal presyncope is sought. There is a sudden reduction in heart rate and BP associated with presyncopal symptoms. Fitzpatrick et al.56 studied patients using prolonged tilt up to 60 minutes and suggested a tilt duration of 40 minutes. Patients who develop vasodepressor syncope will do so within 40 minutes.

Other Cardiovascular Heart Rate Tests There are numerous potential tests of cardiovagal function. These include resting heart rate variation and heart rate response to coughing and to facial immersion. Coughing results in inspiration, then an expiratory effort against a closed glottis followed by an explosive expiration as the glottis suddenly opens.193 The heart rate response consists of a cardioacceleration that is maximal about 2 to 3 seconds after the last cough and a return to resting values in about 12 to 14 seconds.186,187 The mechanism is thought to be cholinergic, resulting initially from muscular contraction followed by baroreflex response to a fall in BP.20 The diving response has also been adapted as a test of cardiovagal function.54,79 The application of cold stimulus to both first divisions of the trigeminal nerve resulted in reflex bradycardia. Twenty-four hour monitoring provides an alternative approach to evaluating cardiovagal function.9 In patients with cardiac arrhythmia, pupillary responses of parasympathetic function are useful.2,114

Determination of Baroreflex Sensitivity The baroreflex is one of the most important regulators of moment-to-moment BP control. It is possible to quantitate baroreflex sensitivity (the change in heart period resulting from a change in BP) in human subjects. Baroreflex sensitivity can be determined under dynamic65,109,160 and steadystate89,91,106 conditions using pharmacologic approaches or carotid sinus stimulation.47,67 In both methods the BP is changed by a directly acting  agonist, usually phenylephrine, and the heart period response to mean BP is determined. In the dynamic method, analysis involves plotting each heart interval against the preceding SBP value. A simple linear regression is fitted, and the slope of the regression of SBP on heart period is a measure of reflex sensitivity in milliseconds per millimeter of mercury. There is an unknown delay (reflex latency) after the pressure change before the change in heart period occurs. The practical way this latency

1124

Quantitating Symptoms, Impairments, and Outcomes

is handled is to calculate two regression slopes, the first by relating SBP to its corresponding heart period and the second by relating SBP to the subsequent pulse interval. The slope with the highest correlation coefficient is accepted. A new computer-assisted method of analysis of baroreflex sensitivity, using the same raw data, has been developed and termed time-related analysis.159 The results are essentially identical to the “Oxford” method but do provide a direct estimation of reflex latency. The steady-state method relates the change in heart period to change in mean BP. By using a suitable range of phenylephrine doses (e.g., 25 to 300 g) on the pressor end and by obtaining a dose-response relationship on the hypotensive end (tilt, lower body negative pressure, or trinitroglycerin or nitroprusside), in normal subjects a sigmoidal relationship obtains.89,91 Typical doses of nitroprusside are 50 to 200 g, resulting in a 10- to 20-mm Hg fall in mean BP. The curve is computer fitted using probit transformation followed by linear regression analysis or directly by using a program to fit sigmoid curves. The indices of mean gain and heart period range are obtained.89,91,106 Baroreflex gain can also be estimated by relating the heart period to the SBP value of the preceding beat.175 These workers regressed heart period and SBP beginning with the end of forced expiration and proceeding to the fifth second of the overshoot. The individual values obtained during phase IV of the Valsalva maneuver correlate well (r 0.9) with those after phenylephrine injection,175 with the former generating higher slopes. An alternative method that obviates the use of pharmacologic agents uses instead a neck suction device, which generates a regulated negative pressure to the carotid sinus. This method permits quantitative assessment of the carotid sinus–to–sinoatrial node baroreflex.47 The neck suction–derived baroslopes yielded results that were qualitatively and quantitatively similar to those of the standard phenylephrine method.45

Determination of Denervation Supersensitivity The denervated end organ is supersensitive to its neurotransmitter.19 This principle is the basis of several tests of autonomic function, including the pupillary response to dilute pilocarpine (see Chapter 21) and the excessive pressor response to directly acting  agonists. The BP and heart rate response to intravenous phenylephrine should include low doses (25 g) to detect denervation supersensitivity. At this dose, normal individuals lack a pressor response. Patients with postganglionic adrenergic denervation have an upregulation of  receptors and respond with a large BP increment.106,121,156 Blood Pressure and Heart Rate Response to Intravenous Tyramine Tyramine is an indirectly acting  agonist that releases norepinephrine presynaptically. It is often given is doses of 10,

20, and 40 g/kg.43 It is often used in conjunction with phenylephrine to provide a full analysis. If a patient had a postganglionic lesion, the pressor response to tyramine would be reduced (because there are fewer functional postganglionic terminals) and there may be a supersensitive response to phenylephrine (because of denervation supersensitivity).

Isoproterenol Infusion Test The number of subjects who develop presyncope and syncope during upright tilt is increased by intravenous isoproterenol.4 In many of these individuals, shortduration HUT was normal.4 The inotropic stimulation and decreased ventricular volume provoked by the combined use of isoproterenol and tilt has been thought to intensely activate unmyelinated ventricular mechanoreceptor afferents and hence a vagally mediated cardioinhibitory reflex.173 We use two infusions of isoproterenol at doses of 0.01 and 0.02 g/kg/min. A minimum recovery period of 10 minutes between infusions is required. A heart rate increase of greater than 50 beats/min or subjective sensation of anxiety, dizziness, or palpitations should promptly terminate the infusion. The -adrenergic antagonist phentolamine and the adrenergic antagonist propranolol should be available in order to rapidly counteract any untoward side effects of adrenergic stimulation. In our hands such intervention has never been required.

Evaluation of Splanchnic-Mesenteric Vasculature The gastrointestinal tract has an extensive blood supply. At times it receives 25% of the cardiac output and contains a third of the blood volume. Thus it has a major influence on cardiovascular homeostasis and the maintenance of BP. It has a rich autonomic innervation and, with the third (enteric) branch of the autonomic nervous system, provides a variety of neurotransmitters, peptides, and autocoids that themselves can influence the autonomic nervous system and vasculature. Measurement of splanchnic blood flow previously has been difficult because of the relative inaccessibility of this vascular region. Invasive techniques have been dependent on cannulation of major blood vessels and the use of contrast medium or inert gases that are not ideal, especially when multiple measurements are needed. Noninvasive indirect approaches include the dye dilution technique with indocyanine green, which provide a measure of hepatic clearance rather than splanchnic blood flow. The development of Doppler ultrasound techniques now provides a direct, noninvasive, and reproducible measure of blood flow.23,25 This enables repeated measurements

Quantitation of Autonomic Impairment

during stimuli such as HUT, food and alcohol ingestion, and exercise, which increase autonomic activity.24,86,119,138 The anatomy of the celiac artery and superior mesenteric artery (SMA) lends itself to such visualization and, when combined with blood vessel diameter and velocity, enables calculation of blood flow. However, this provides a measure of blood flow in a major vessel rather than the entire region. The technique is dependent on a trained operator, and adequate vascular visualization of the major vessel is a necessity; gut motility and gas formation, as may occur after food ingestion, may interfere with the signal. The innervation and vascular responses of the splanchnic region can be assessed using physiologic and pharmacologic stimuli. Physiologic stimuli such as mental arithmetic, isometric exercise, cutaneous cold, dynamic exercise, and HUT result in constriction of the SMA by sympathoneural activation. In autonomic failure, there are diminished responses following such stimulation.24,58,137 Thus these techniques allow evaluation of sympathetic vasoconstrictor ability in this large region. In POTS there is a higher SMA velocity and diminished SMA response at rest than in controls, and the responses to hyperventilation are impaired.170 A major physiologic stimulus to the splanchnic vasculature is food ingestion (Fig. 44–16). In normal subjects, a liquid meal of mixed composition markedly increases SMA outflow, from 500 to about 1000 mL/min within 15 minutes of food ingestion86; there are similar changes with alcohol.119 The vasodilation lasts for a substantial period of time, with the responses dependent on the volume, caloric load, and composition of the meal; these determine the release of pancreatic and gastrointestinal peptides and hormones, some of which are vasoactive or modulate enteric neural transmission. In primary autonomic failure (MSA and PAF), there are similar responses to food ingestion,87 whereas in secondary autonomic failure, normal to reduced responses are observed.10,58 In POTS, the responses are similar to those of controls.170 Alcohol causes splanchnic vasodilation in autonomic failure and contributes to hypotension.22 The splanchnic responses to various drugs may provide further information on mechanisms and thus may aid treatment. The somatostatin analogue octreotide inhibits the release of a variety of gut and pancreatic peptides and prevents SMA dilatation when food is ingested (see Fig. 44–16).86,87 Octreotide has been of value in the management of postprandial hypotension in autonomic failure.

EVALUATION OF CENTRAL AUTONOMIC PATHWAYS The brain plays a key role in control of autonomic activity, and evaluation of cerebral centers and pathways is of importance in various disorders. The two major approaches used involve neuroimaging and neuroendocrine measurement.

1125

FIGURE 44–16 Superior mesenteric artery blood flow in normal subjects before and after a balanced liquid meal, when given either saline placebo (filled circles, continuous line) or 50 g octreotide (open circles, dotted line), both subcutaneously. (From Kooner, J. S., Peart, W. S., and Mathias, C. J.: The peptide release inhibitor, octreotide [SMS 201–995], prevents the haemodynamic changes following food ingestion in normal human subjects. Q. J. Exp. Physiol. 74:569, 1989, with permission.)

Advances in neuroimaging now provide means to evaluate the functional anatomy of different cerebral regions. High-resolution magnetic resonance imaging may define lesions in the brainstem in MSA153,154; this does not occur in PAF, in which the primary lesions are in the periphery.33 PET provides further information on the neurotransmitters involved.15,16,141 This enables evaluation of cerebral glucose metabolism, with hypometabolism found in specific brain regions in MSA.60,127 The dopamine precursor fluorodopa is inadequately taken up in MSA, unlike in PD and PAF.15 Allied approaches include proton magnetic resonance spectroscopy, which evaluates brain metabolism; and the N-acetylaspartate:creatine ratio, in which there are differences within the lentiform nucleus in MSA, but not in other parkinsonian syndromes.39 Although these techniques help separate the autonomic failure syndromes from other disorders, they are predominantly designed to determine differences in nonautonomic cerebral centers. Recently, PET scanning and functional magnetic resonance imaging have provided further information on cerebral autonomic centers and their activation in response to stimuli that selectively influence autonomic activity (Fig. 44–17). With sympathoneural stimuli such

1126

Quantitating Symptoms, Impairments, and Outcomes

FIGURE 44–17 Changes in cardiovascular autonomic measurements (blood pressure and heart rate) during control tasks and effortful isometric exercise and mental arithmetic in normal subjects (top) and in patients with pure autonomic failure (PAF) (bottom). Alongside are changes in regional cerebral (mainly right anterior cingulate) activity showing positive covariance with mean arterial pressure during exercise and mental arithmetic in normal subjects. There is an increase in activity in the right anterior cingulate area, and also the brainstem (pontine) areas in PAF, despite no change in blood pressure. The central autonomic centers in PAF patients are preserved, and indicate attempts to increase peripheral autonomic activity, which is not possible because of their postganglionic sympathetic lesion. (Modified from Mathias, C. J.: To stand on ones’ own legs. Clin. Med. 2:237, 2002.) See Color Plate

as mental arithmetic and isometric exercise, there is increased activity in the right anterior cingulate gyrus, among other areas.32 In PAF there is an increase mainly in the anterior executive and not the posterior evaluative segment of the anterior cingulate gyrus, with increased activity also in the pons.35 With different physiologic stimuli, activation in specific brain areas such as the insula and amygdala correlate with cardiovascular and sudomotor function34,36; specific changes also can be shown with relaxation and biofeedback techniques.37 A variety of neurochemical and neuroendocrine approaches have been used to determine the activity of autonomic centers and pathways.132 Measurement of neu-

rotransmitters and their metabolites in cerebrospinal fluid provides indirect information.133,135 Neuroendocrine approaches involve either a physiologic or pharmacologic stimulus that causes specific hormonal changes. They include the measurement of vasopressin in response to HUT and osmotic stimuli,77,83,194,197 of adrenocorticotrophic hormone in response to arecholine,134 and of growth hormone in response to clonidine.38,172 These tests evaluate specific responses of hypothalamic nuclei and neuroendocrine function. The clonidine–growth hormone stimulation test, which is used in clinical practice, is dependent upon the ability of clonidine, an 2 adrenoceptor agonist, to stimulate the hypothalamus and elevate human growth hormone–

1127

Quantitation of Autonomic Impairment

releasing hormone, which then acts upon the acidophil cells in the anterior pituitary to release growth hormone into the circulation (Fig. 44–18).82 In PAF with a peripheral sympathetic lesion, there is a rise in growth hormone levels in response to clonidine that is similar to that in normal subjects; however, in MSA with central lesions, there is no rise with clonidine (see Fig. 44–18).85,172 Anterior pituitary deficiency does not contribute to the impaired response because the secretagogue L-dopa raises growth hormone levels in patients with MSA and in normal subjects.84 Thus the clonidine–growth hormone test readily separates patients with distal autonomic lesions such as PAF from those with central lesions such as MSA. There is a growth hormone response in progressive supranuclear palsy that is different from the impaired response in MSA.81 In idiopathic PD, the growth hormone response is preserved, unlike in the parkinsonian and cerebellar forms of MSA (Fig. 44–19).85 Whether this test will differentiate these disorders in the early stages of parkinsonism remains to be determined.

FIGURE 44–19 Lack of serum growth hormone (GH) response to clonidine in two forms of multiple system atrophy (MSA)—the cerebellar form (MSA-C) and the parkinsonian form (MSA-P)—in contrast to patients with idiopathic Parkinson’s disease with no autonomic deficit, in whom there is a significant rise in GH levels. MSA-M mixed form of MSA. (Modified from Kimber, J. R., Watson, L., and Mathias, C. J.: Distinction of idiopathic Parkinson’s disease from multiple-system atrophy by stimulation of growth-hormone release with clonidine. Lancet 349:1877, 1997.)

REFERENCES

FIGURE 44–18 Serum growth hormone (GH) concentrations before (0) and at 15-minute intervals for 60 minutes after clonidine (2 g/kg/min) in normal subjects (controls) and in patients with pure autonomic failure (PAF) and multiple system atrophy (MSA). GH concentrations rise in controls and in patients with PAF, with a peripheral lesion; there is no rise in patients with MSA, with a central lesion. (Modified from Kimber, J. R., Watson, L., and Mathias, C. J.: Distinction of idiopathic Parkinson’s disease from multiple-system atrophy by stimulation of growth-hormone release with clonidine. Lancet 349:1877, 1997.)

1. Abbot, N. C., Beck, J. S., Wilson, S. B., and Khan, F.: Vasomotor reflexes in the fingertip skin of patients with type 1 diabetes mellitus and leprosy. Clin. Auton. Res. 3:189, 1993. 2. Adler, F. H.: Physiology of the Eye: Clinical Applications. St. Louis, C. V. Mosby, 1950. 3. Akinola, A. B., Smith, G. D., Mathias, C. J., et al.: The metabolic, catecholamine and cardiovascular effects of exercise in human sympathetic denervation. Clin. Auton. Res. 11:251, 2001. 4. Almquist, A., Goldenberg, I. F., Milstein, S., et al.: Provocation of bradycardia and hypotension by isoproterenol and upright posture in patients with unexplained syncope. N. Engl. J. Med. 320:346, 1989. 5. Angelone, A., and Coulter, N. A.: Respiratory sinus arrhythmia: a frequency dependent phenomenon. J. Appl. Physiol. 19:479, 1964. 6. Arbab, A. S., Koizumi, K., Takano, H., et al.: Parameters of dynamic and static iodine-123-MIBG cardiac imaging. J. Nucl. Med. 36:962, 1995. 7. Benarroch, E. E., Opfer-Gehrking, T. L., and Low, P. A.: Use of the photoplethysmographic technique to analyze the Valsalva maneuver in normal man. Muscle Nerve 14:1165, 1991.

1128

Quantitating Symptoms, Impairments, and Outcomes

8. Bennett, T., Farquhar, I. K., Hosking, D. J., and Hampton, J. R.: Assessment of methods for estimating autonomic nervous control of the heart in patients with diabetes mellitus. Diabetes 27:1167, 1978. 9. Bennett, T., Riggott, P. A., Hosking, D. J., and Hampton, J. R.: Twenty-four hour monitoring of heart rate and activity in patients with diabetes mellitus: a comparison with clinic investigations. Br. Med. J. 1:1250, 1976. 10. Best, I. M., Pitzele, A., Green, A., et al.: Mesenteric blood flow in patients with diabetic neuropathy. J. Vasc. Surg. 13:84, 1991. 11. Borgdorff, P.: Respiratory fluctuations in pupil size. Am. J. Physiol. 228:1094, 1975. 12. Borst, C., Wieling, W., van Brederode, J. F., et al.: Mechanisms of initial heart rate response to postural change. Am. J. Physiol. 243:H676, 1982. 13. Braune, S., Reinhardt, M., Schnitzer, R., et al.: Cardiac uptake of [123I]MIBG separates Parkinson’s disease from multiple system atrophy. Neurology 53:1020, 1999. 14. Brooker, J. Z., Alderman, E. L., and Harrison, D. C.: Alterations in left ventricular volumes induced by Valsalva manoeuvre. Br. Heart J. 36:713, 1974. 15. Brooks, D. J., Ibanez, V., Sawle, G. V., et al.: Differing patterns of striatal 18F-dopa uptake in Parkinson’s disease, multiple system atrophy, and progressive supranuclear palsy. Ann. Neurol. 28:547, 1990. 16. Brooks, D. J., Ibanez, V., Sawle, G. V., et al.: Striatal D2 receptor status in patients with Parkinson’s disease, striatonigral degeneration, and progressive supranuclear palsy, measured with 11C-raclopride and positron emission tomography. Ann. Neurol. 31:184, 1992. 17. Buda, A. J., Pinsky, M. R., Ingels, N. B. Jr., et al.: Effect of intrathoracic pressure on left ventricular performance. N. Engl. J. Med. 301:453, 1979. 18. Candel, S., and Ehrlich, D. E.: Venous blood flow during the Valsalva experiment including some clinical applications. Am. J. Med. 15:307, 1953. 19. Cannon, W. B., and Rosenblueth, A.: The Supersensitivity of Denervated Structures: A Law of Denervation. New York, Macmillan, 1949. 20. Cardone, C., Bellavere, F., Ferri, M., and Fedele, D.: Autonomic mechanisms in the heart rate response to coughing. Clin. Sci. 72:55, 1987. 21. Cariga, P., Catley, M., Mathias, C. J., et al.: Organisation of the sympathetic skin response in spinal cord injury. J. Neurol. Neurosurg. Psychiatry 72:356, 2002. 22. Chaudhuri, K. R., Maule, S., Thomaides, T., et al.: Alcohol ingestion lowers supine blood pressure, causes splanchnic vasodilatation and worsens postural hypotension in primary autonomic failure. J. Neurol. 241:145, 1994. 23. Chaudhuri, K. R., Thomaides, T., Hernandez, P., et al.: Noninvasive quantification of superior mesenteric artery blood flow during sympathoneural activation in normal subjects. Clin. Auton. Res. 1:37, 1991. 24. Chaudhuri, K. R., Thomaides, T., and Mathias, C. J.: Abnormality of superior mesenteric artery blood flow responses in human sympathetic failure. J. Physiol. (Lond.) 457:477, 1992. 25. Chaudhuri, K. R., Thomaides, T., and Mathias, C. J.: Variability in pressor and splanchnic responses to mental arithmetic but not other stimuli causing sympatho-neural

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

activation in normal subjects of different ages. High Blood Pressure 1:223, 1992. Chelimsky, T. C., Low, P. A., Naessens, J. M., et al.: Value of autonomic testing in reflex sympathetic dystrophy. Mayo Clin. Proc. 70:1029, 1995. Chemali, K. R., Gorodeski, R., and Chelimsky, T. C.: Alphaadrenergic supersensitivity of the sudomotor nerve in complex regional pain syndrome. Ann. Neurol. 49:453, 2001. Claus, D., Feistel, H., Brunholzl, C., et al.: Investigation of parasympathetic and sympathetic cardiac innervation in diabetic neuropathy: heart rate variation versus meta-iodobenzylguanidine measured by single photon emission computed tomography. Clin. Auton. Res. 4:117, 1994. Claus, D., Meudt, O., Rozeik, C., et al.: Prospective investigation of autonomic cardiac neuropathy in diabetes mellitus. Clin. Auton. Res. 12:373, 2002. Corbett, J. L.: Some Aspects of the Autonomic Nervous System in Normal and Abnormal Man. D. Phil. thesis, Oxford University, 1969. Courbon, F., Brefel-Courbon, C., Thalamas, C., et al.: Cardiac MIBG scintigraphy is a sensitive tool for detecting cardiac sympathetic denervation in Parkinson’s disease. Mov. Disord. 18:890, 2003. Critchley, H. D., Corfield, D. R., Chandler, M. P., et al.: Cerebral correlates of autonomic cardiovascular arousal: a functional neuroimaging investigation in humans. J. Physiol. (Lond.) 523:259, 2000. Critchley, H. D., Good, C. D., Ashburner, J., et al.: Changes in cerebral morphology consequent to peripheral autonomic denervation. Neuroimage 18:908, 2003. Critchley, H.D., Mathias, C.J., and Dolan, R.J.: Neural activity in the human brain relating to uncertainty and arousal during anticipation. Neuron 29:537, 2001. Critchley, H.D., Mathias, C.J., and Dolan, R.J.: Neuroanatomical basis for first- and second-order representations of bodily states. Nat. Neurosci. 4:207, 2001. Critchley, H. D., Mathias, C. J., and Dolan, R. J.: Fear conditioning in humans: the influence of awareness and autonomic arousal on functional neuroanatomy. Neuron 33:653, 2002. Critchley, H.D., Melmed, R.N., Featherstone, E., Mathias, C.J., and Dolan, R.J.: Brain activity during biofeedback relaxation: a functional neuroimaging investigation. Brain 124:1003, 2001. da Costa, D.F., Bannister, R., Landon, J., and Mathias, C.J.: Growth hormone response to clonidine is impaired in patients with central sympathetic degeneration. Clin. Exp. Hypertens. A Theory Pract. 6:1843, 1984. Davie, C. A., Wenning, G. K., Barker, G. J., et al.: Differentiation of multiple system atrophy from idiopathic Parkinson’s disease using proton magnetic resonance spectroscopy. Ann. Neurol. 37:204, 1995. Degrado, T. R., Zalutsky, M. R., and Vaidyanathan, G.: Uptake mechanisms of meta-[123I]iodobenzylguanidine in isolated rat heart. Nucl. Med. Biol. 22:1, 1995. Delius, W., Hagbarth, K. E., Hongell, A., and Wallin, B. G.: Manoeuvres affecting sympathetic outflow in human muscle nerves. Acta Physiol. Scand. 84:82, 1972. Delius, W., and Kellerova, E.: Reactions of arterial and venous vessels in the human forearm and hand to deep breath or mental strain. Clin. Sci. 40:271, 1971.

Quantitation of Autonomic Impairment 43. Demanet, J. C.: Usefulness of noradrenaline and tyramine infusion tests in the diagnosis of orthostatic hypotension. Cardiology 61(Suppl. 1):213, 1976. 44. Dyck, P. J., Oviatt, K. F., and Lambert, E. H.: Intensive evaluation of referred unclassified neuropathies yields improved diagnosis. Ann. Neurol. 10:222, 1981. 45. Ebert, T. J., Hayes, J. J., Ceschi, J., et al.: Repetitive ramped neck suction: a quantitative test of human baroreceptor function. Am. J. Physiol. 247:H1013, 1984. 46. Eckberg, D.: Parasympathetic cardiovascular control in human disease: a critical review of methods and results. Am. J. Physiol. 239:H581, 1980. 47. Eckberg, D. L., Cavanaugh, M. S., Mark, A. L., and Abboud, F. M.: A simplified neck suction device for activation of carotid baroreceptors. J. Lab. Clin. Med. 85:167, 1975. 48. Esler, M.: Clinical application of noradrenaline spillover methodology: delineation of regional human sympathetic nervous responses. Pharmacol. Toxicol. 73:243, 1993. 49. Evans, B. A., Lussky, D., and Knezevic, W.: The peripheral autonomic surface potential in suspected small fiber peripheral neuropathy. Muscle Nerve 11:982, 1988. 50. Ewing, D. J.: Cardiovascular reflexes and autonomic neuropathy. Clin. Sci. Mol. Med. 55:321, 1978. 51. Ewing, D. J., Campbell, I. W., and Clarke, B. F.: Assessment of cardiovascular effects in diabetic autonomic neuropathy and prognostic implications. Ann. Intern. Med. 92:308, 1980. 52. Ewing, D. J., Campbell, I. W., Murray, A., et al.: Immediate heart rate response to standing: simple test of autonomic neuropathy in diabetes. Br. Med. J. 1:145, 1978. 53. Faes, T. J., Wagemans, M. F., Cillekens, J. M., et al.: The validity and reproducibility of the skin vasomotor test— studies in normal subjects, after spinal anaesthesia, and in diabetes mellitus. Clin. Auton. Res. 3:319, 1993. 54. Fagius, J., and Sundlof, G.: The diving response in man: effects on sympathetic activity in muscle and skin nerve fascicles. J. Physiol. (Lond.) 377:429, 1986. 55. Fagius, J., and Wallin, B. G.: Microneurographic evidence of excessive sympathetic outflow in the Guillain-Barre syndrome. Brain 106:589, 1983. 56. Fitzpatrick, A. P., Theodorakis, G., Vardas, P., and Sutton, R.: Methodology of head-up tilt testing in patients with unexplained syncope. J. Am. Coll. Cardiol. 17:125, 1991. 57. Freyschuss, U., and Melcher, A.: Sinus arrhythmia in man: influence of tidal volume and oesophageal pressure. Scand. J. Clin. Lab. Invest. 35:487, 1975. 58. Fujimura, J., Camilleri, M., Low, P. A., et al.: Effect of perturbations and a meal on superior mesenteric artery flow in patients with orthostatic hypotension. J. Auton. Nerv. Syst. 67:15, 1997. 59. Gemmill, J. D., Venables, G. S., and Ewing, D. J.: Noradrenaline response to edrophonium in primary autonomic failure: distinction between central and peripheral damage. Lancet 1:1018, 1988. 60. Gilman, S., Koeppe, R. A., Junck, L., et al.: Patterns of cerebral glucose metabolism detected with positron emission tomography differ in multiple system atrophy and olivopontocerebellar atrophy. Ann. Neurol. 36:166, 1994. 61. Goldstein, D. S., Holmes, C., Eisenhofer, G., and Kopin, I. J.: Sympathetic cardioneuropathy in dysautonomias. N. Engl. J. Med. 336:696, 1997.

1129

62. Goldstein, D. S., Holmes, C., Li, S. T., et al.: Cardiac sympathetic denervation in Parkinson disease. Ann. Intern. Med. 133:338, 2000. 63. Goldstein, D. S., Holmes, C., Stuhlmuller, J. E., et al.: 6-[18F]fluorodopamine positron emission tomographic scanning in the assessment of cardiac sympathoneural function—studies in normal humans. Clin. Auton. Res. 7:17, 1997. 64. Goldstein, D. S., Polinsky, R. J., Garty, M., et al.: Patterns of plasma levels of catechols in neurogenic orthostatic hypotension. Ann. Neurol. 26:558, 1989. 65. Gribbin, B., Pickering, T. G., Sleight, P., and Peto, R.: Effect of age and high blood pressure on baroreflex sensitivity in man. Circ. Res. 29:424, 1971. 66. Hakusui, S., Yasuda, T., Yanagi, T., et al.: A radiological analysis of heart sympathetic functions with meta[123I]iodobenzylguanidine in neurological patients with autonomic failure. J. Auton. Nerv. Syst. 49:81, 1994. 67. Heidorn, G. H., and McNamara, A. P.: Effect of carotid sinus stimulation on the electrocardiograms of clinically normal individuals. Circulation 14:1104, 1956. 68. Herman, L. M., Dai, S., Hartman, N. G., and McNeill, J. H.: Meta-iodobenzylguanidine uptake in the hypertensivediabetic rat heart: a marker for myocardial dysfunction? Can. J. Physiol. Pharmacol. 72:1162, 1994. 69. Hirayama, M., Hakusui, S., Koike, Y., et al.: A scintigraphical qualitative analysis of peripheral vascular sympathetic function with meta-[123I]iodobenzylguanidine in neurological patients with autonomic failure. J. Auton. Nerv. Syst. 53:230, 1995. 70. Hirsch, J., Leibel, R. L., Mackintosh, R., and Aguirre, A.: Heart rate variability as a measure of autonomic function during weight change in humans. Am. J. Physiol. 261:R1418, 1991. 71. Hoeldtke, R. D., Dworkin, G. E., Gaspar, S. R., and Israel, B. C.: Sympathotonic orthostatic hypotension: a report of four cases. Neurology 39:34, 1989. 72. Hume, L., Ewing, D. J., Campbell, I. W., et al.: Provocation of postural hypotension by nitroglycerin in diabetic autonomic neuropathy. Diabetes Care 3:27, 1980. 73. Imholz, B. P., van Montfrans, G. A., Settels, J. J., et al.: Continuous non-invasive blood pressure monitoring: reliability of Finapres device during the Valsalva manoeuvre. Cardiovasc. Res. 22:390, 1988. 74. Imholz, B. P. M., Settels, J. J., van der Meiracker, A. H., and Wieling, W.: Orthostatic circulatory control in the elderly evaluated by noninvasive continuous blood pressure measurement. Clin. Sci. 79:73, 1990. 75. Ingall, T. J., McLeod, J. G., and O’Brien, P. C.: The effect of ageing on autonomic nervous system function. Aust. N. Z. J. Med. 20:570, 1990. 76. Iwasa, K., Nakajima, K., Yoshikawa, H., et al.: Decreased myocardial 123I-MIBG uptake in Parkinson’s disease. Acta Neurol. Scand. 97:303, 1998. 77. Kaufmann, H., Oribe, E., Miller, M., et al.: Hypotensioninduced vasopressin release distinguishes between pure autonomic failure and multiple system atrophy with autonomic failure. Neurology 42:590, 1992. 78. Kemmotsu, O., Ueda, M., Otsuka, H., et al.: Arterial tonometry for noninvasive, continuous blood pressure monitoring during anesthesia. Anesthesiology 75:333, 1991.

1130

Quantitating Symptoms, Impairments, and Outcomes

79. Khurana, R. K., Watabiki, S., Hebel, J. R., et al.: Cold face test in the assessment of trigeminal-brainstem-vagal function in humans. Ann. Neurol. 7:144, 1980. 80. Kilo, S., Berghoff, M., Hilz, M., and Freeman, R.: Neural and endothelial control of the microcirculation in diabetic peripheral neuropathy. Neurology 54:1246, 2000. 81. Kimber, J., Mathias, C. J., Lees, A. J., et al.: Physiological, pharmacological and neurohormonal assessment of autonomic function in progressive supranuclear palsy. Brain 123:1422, 2000. 82. Kimber, J., Sivenandan, M., Watson, L., and Mathias, C. J.: Age- and gender-related growth hormone responses to intravenous clonidine in healthy adults. Growth Horm. IGF Res. 11:128, 2001. 83. Kimber, J., Watson, L., and Mathias, C. J.: Abnormal suppression of arginine-vasopressin by clonidine in multiple system atrophy. Clin. Auton. Res. 9:271, 1999. 84. Kimber, J., Watson, L., and Mathias, C. J.: Neuroendocrine responses to levodopa in multiple system atrophy (MSA). Mov. Disord. 14:981, 1999. 85. Kimber, J. R., Watson, L., and Mathias, C. J.: Distinction of idiopathic Parkinson’s disease from multiple-system atrophy by stimulation of growth-hormone release with clonidine. Lancet 349:1877, 1997. 86. Kooner, J. S., Peart, W. S., and Mathias, C. J.: The peptide release inhibitor, octreotide (SMS 201–995), prevents the haemodynamic changes following food ingestion in normal human subjects. Q. J. Exp. Physiol. 74:569, 1989. 87. Kooner, J. S., Raimbach, S., Watson, L., et al.: Relationship between splanchnic vasodilation and postprandial hypotension in patients with primary autonomic failure. J. Hypertens. Suppl. 7:S40, 1989. 88. Kopin, I. J., Polinsky, R. J., Oliver, J. A., et al.: Urinary catecholamine metabolites distinguish different types of sympathetic neuronal dysfunction in patients with orthostatic hypotension. J. Clin. Endocrinol. Metab. 57:632, 1983. 89. Korner, P. I., Shaw, J., West, M. J., and Oliver, J. R.: Central nervous system control of baroreceptor reflexes in the rabbit. Circ. Res. 31:637, 1972. 90. Korner, P. I., Tonkin, A. M., and Uther, J. B.: Reflex and mechanical circulatory effects of graded Valsalva maneuvers in normal man. J. Appl. Physiol. 40:434, 1976. 91. Korner, P. I., West, M. J., Shaw, J., and Uther, J. B.: “Steadystate” properties of the baroreceptor-heart rate reflex in essential hypertension in man. Clin. Exp. Pharmacol. Physiol. 1:65, 1974. 92. Lambert, G. W., Thompson, J. M., Turner, A. G., et al.: Cerebral noradrenaline spillover and its relation to muscle sympathetic nervous activity in healthy human subjects. J. Auton. Nerv. Syst. 64:57, 1997. 93. Landis, S. C.: Neurotransmitter plasticity in sympathetic neurons. In Handbook of Chemical Neuroanatomy: The Peripheral Nervous System. Amsterdam, Elsevier, p. 65, 1988. 94. Lang, E., Foerster, A., Pfannmuller, D., and Handwerker, H. O.: Quantitative assessment of sudomotor activity by capacitance hygrometry. Clin. Auton. Res. 3:107, 1993. 95. Lang, E., Spitzer, A., Claus, D., et al.: Stimulation of sudomotor axon reflex mechanism by carbachol in healthy subjects and patients suffering from diabetic polyneuropathy. Acta Neurol. Scand. 91:251, 1995.

96. Langer, A., Freeman, M. R., Josse, R. G., and Armstrong, P. W.: Metaiodobenzylguanidine imaging in diabetes mellitus: assessment of cardiac sympathetic denervation and its relation to autonomic dysfunction and silent myocardial ischemia. J. Am. Coll. Cardiol. 25:610, 1995. 97. Low, P. A.: Composite autonomic scoring scale for laboratory quantification of generalized autonomic failure. Mayo Clin. Proc. 68:748, 1993. 98. Low, P. A.: The effect of aging on the autonomic nervous system. In Low, P. A. (ed.) Clinical Autonomic Disorders: Evaluation and Management. Philadelphia, Lippincott– Raven, p. 161, 1997. 99. Low, P. A.: Laboratory evaluation of autonomic function. In Low, P. A. (ed.) Clinical Autonomic Disorders: Evaluation and Management. Philadelphia, Lippincott–Raven, p. 179, 1997. 100. Low, P. A., Caskey, P. E., Tuck, R. R., et al.: Quantitative sudomotor axon reflex test in normal and neuropathic subjects. Ann. Neurol. 14:573, 1983. 101. Low, P. A., Denq, J. C., Opfer-Gehrking, T. L., et al.: Effect of age and gender on sudomotor and cardiovagal function and blood pressure response to tilt in normal subjects. Muscle Nerve 20:1561, 1997. 102. Low, P. A., Fealey, R. D., Sheps, S. G., et al.: Chronic idiopathic anhidrosis. Ann. Neurol. 18:344, 1985. 103. Low, P. A., and Kennedy, W. R.: Cutaneous effectors as indicators of abnormal sympathetic function. In Morris, J. L., and Gibbins, I. L. (eds.): Autonomic Innervation of the Skin. Amsterdam, Harwood Academic Publishers, p. 165, 1997. 104. Low, P. A., Neumann, C., Dyck, P. J., et al.: Evaluation of skin vasomotor reflexes by using laser Doppler velocimetry. Mayo Clin. Proc. 58:583, 1983. 105. Low, P. A., Opfer-Gehrking, T. L., and Kihara, M.: In vivo studies on receptor pharmacology of the human eccrine sweat gland. Clin. Auton. Res. 2:29, 1992. 106. Low, P. A., Walsh, J. C., Huang, C. Y., and McLeod, J. G.: The sympathetic nervous system in diabetic neuropathy: a clinical and pathological study. Brain 98:341, 1975. 107. Low, P. A., Wilson, P. R., Sandroni, P., et al.: Clinical characteristics of patients with reflex sympathetic dystrophy (sympathetically maintained pain) in the USA. In Jänig, W., and Stanton-Hicks, M. (eds.): Progress in Pain Research Management. Seattle, IASP Press, p. 49, 1996. 108. Low, P. A., Zimmerman, B. R., and Dyck, P. J.: Comparison of distal sympathetic with vagal function in diabetic neuropathy. Muscle Nerve 9:592, 1986. 109. Ludbrook, J., Mancia, G., Ferrari, A., and Zanchetti, A.: The variable-pressure neck-chamber method for studying the carotid baroreflex in man. Clin. Sci. Mol. Med. 53:165, 1977. 110. Machida, K., Honda, N., Mamiya, T., et al.: Abnormal sympathetic innervation of the heart in a patient with myotonic dystrophy detected with I-123 MIBG cardiac SPECT. Clin. Nucl. Med. 19:968, 1994. 111. Magnifico, F., Misra, V. P., Murray, N. M., and Mathias, C. J.: The sympathetic skin response in peripheral autonomic failure—evaluation in pure failure, pure cholinergic dysautonomia and dopamine-beta-hydroxylase deficiency. Clin. Auton. Res. 8:133, 1998. 112. Man in ‘t Veld, A. J., Boomsma, F., Moleman, P., and Schalekamp, M. A.: Congenital dopamine-beta-hydroxylase

Quantitation of Autonomic Impairment

113.

114.

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125. 126.

127.

128.

129.

deficiency: a novel orthostatic syndrome. Lancet 1:183, 1987. Mantysaari, M., Kuikka, J., Mustonen, J., et al.: Noninvasive detection of cardiac sympathetic nervous dysfunction in diabetic patients using [123I]metaiodobenzylguanidine. Diabetes 41:1069, 1992. Martyn, C. N., and Ewing, D. J.: Pupil cycle time—a simple way of measuring an autonomic reflex. J. Neurol. Neurosurg. Psychiatry 49:771, 1986. Mathias, C. J., Bannister, R. B., Cortelli, P., et al.: Clinical autonomic and therapeutic observations in two siblings with postural hypotension and sympathetic failure due to an inability to synthesize noradrenaline from dopamine because of a deficiency of dopamine beta hydroxylase. Q. J. Med. 75:617, 1990. Mathias, C. J., Christensen, N. J., Corbett, J. L., et al.: Plasma catecholamines, plasma renin activity and plasma aldosterone in tetraplegic man, horizontal and tilted. Clin. Sci. Mol. Med. 49:291, 1975. Mathias, C. J., Christensen, N. J., Corbett, J. L., et al.: Plasma catecholamines during paroxysmal neurogenic hypertension in quadriplegic man. Circ. Res. 39:204, 1976. Mathias, C. J., Frankel, H. L., Turner, R. C., and Christensen, N. J.: Physiological responses to insulin hypoglycaemia in spinal man. Paraplegia 17:319, 1979. Maule, S., Chaudhuri, K. R., Thomaides, T., et al.: Effects of oral alcohol on superior mesenteric artery blood flow in normal man, horizontal and tilted. Clin. Sci. 84:419, 1993. Miyanaga, H., Yoneyama, S., Kamitani, T., et al.: Clinical usefulness of 123I-metaiodobenzylguanidine myocardial scintigraphy in diabetic patients with cardiac sympathetic nerve dysfunction. Jpn. Circ. J. 59:599, 1995. Moorhouse, J. A., Carter, S. A., and Doupe, J.: Vascular responses in diabetic peripheral neuropathy. Br. Med. J. 1:883, 1966. Muller, K. D., Jakob, H., Neuzner, J., et al.: 123Imetaiodobenzylguanidine scintigraphy in the detection of irregular regional sympathetic innervation in long QT syndrome. Eur. Heart J. 14:316, 1993. Nakata, T., Shimamoto, K., Yonekura, S., et al.: Cardiac sympathetic denervation in transthyretin-related familial amyloidotic polyneuropathy: detection with iodine-123-MIBG. J. Nucl. Med. 36:1040, 1995. Nijboer, J. A., Dorlas, J. C., and Lubbers, J.: The difference in blood pressure between upper arm and finger during physical exercise. Clin. Physiol. 8:501, 1988. Novak, V., Freimer, M. L., Kissel, J. T., et al.: Autonomic impairment in painful neuropathy. Neurology 56:861, 2001. Orimo, S., Ozawa, E., Nakade, S., et al.: (123)I-metaiodobenzylguanidine myocardial scintigraphy in Parkinson’s disease. J. Neurol. Neurosurg. Psychiatry 67:189, 1999. Otsuka, M., Ichiya, Y., Kuwabara, Y., et al.: Glucose metabolism in the cortical and subcortical brain structures in multiple system atrophy and Parkinson’s disease: a positron emission tomographic study. J. Neurol. Sci. 144:77, 1996. Parati, G., Casadei, R., Gropelli, A., et al.: Comparison of finger and intra-arterial blood pressure monitoring at rest and during laboratory testing. Hypertension 13:647, 1989. Penaz, J.: Photoelectric measurement of blood pressure, volume and flow in the finger. In Albert, R., Vogt, W. S., and Helbig, W. (eds.): Digest of the International Conference on

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145.

1131

Medicine and Biological Engineering. Dresden: Conference Committee on the Xth International Conference on Medicine and Biological Engineering, p. 104, 1973. Periquet, M. I., Novak, V., Collins, M. P., et al.: Painful sensory neuropathy: prospective evaluation using skin biopsy. Neurology 53:1641, 1999. Pfeifer, M. A., Cook, D., Brodsky, J., et al.: Quantitative evaluation of cardiac parasympathetic activity in normal and diabetic man. Diabetes 31:339, 1982. Polinsky, R. J.: Neuropharmacological investigation of autonomic failure. In Mathias, C. J., and Bannister, R. (eds.): Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System. Oxford, UK, Oxford University Press, p. 232, 1999. Polinsky, R. J., Brown, R. T., Burns, R. S., et al.: Low lumbar CSF levels in homovanillic acid and 5-hydroxyindoleacetic acid in multiple system atrophy with autonomic failure. J. Neurol. Neurosurg. Psychiatry 51:914, 1988. Polinsky, R. J., Brown, R. T., Curras, M. T., et al.: Central and peripheral effects of arecoline in patients with autonomic failure. J. Neurol. Neurosurg. Psychiatry 54:807, 1991. Polinsky, R. J., Holmes, K. V., Brown, R. T., and Weise, V.: CSF acetylcholinesterase levels are reduced in multiple system atrophy with autonomic failure. Neurology 39:40, 1989. Polinsky, R. J., Kopin, I. J., Ebert, M. H., and Weise, V.: The adrenal medullary response to hypoglycemia in patients with orthostatic hypotension. J. Clin. Endocrinol. Metab. 51:1401, 1980. Puvi-Rajasingham, S., Smith, G. D., Akinola, A., and Mathias, C. J.: Abnormal regional blood flow responses during and after exercise in human sympathetic denervation. J. Physiol. (Lond.) 505:841, 1997. Puvi-Rajasingham, S., Wijeyekoon, B., Natarajan, P., et al.: Systemic and regional (including superior mesenteric) haemodynamic responses during supine exercise while fasted and fed in normal man. Clin. Auton. Res. 7:149, 1997. Rathmann, W., Ziegler, D., Jahnke, M., et al.: Mortality in diabetic patients with cardiovascular autonomic neuropathy. Diabet. Med. 10:820, 1993. Rayman, G., Hassan, A., and Tooke, J. E.: Blood flow in the skin of the foot related to posture in diabetes mellitus. Br. Med. J. 292:87, 1986. Rinne, J. O., Burn, D. J., Mathias, C. J., et al.: Positron emission tomography studies on the dopaminergic system and striatal opioid binding in the olivopontocerebellar atrophy variant of multiple system atrophy. Ann. Neurol. 37:568, 1995. Robertson, D., Goldberg, M. R., Onrot, J., et al.: Isolated failure of autonomic noradrenergic neurotransmission: evidence for impaired beta-hydroxylation of dopamine. N. Engl. J. Med. 314:1494, 1986. Robertson, D., Johnson, G. A., Robertson, R. M., et al.: Comparative assessment of stimuli that release neuronal and adrenomedullary catecholamines in man. Circulation 59:637, 1979. Sampson, M. J., Wilson, S., Karagiannis, P., et al.: Progression of diabetic autonomic neuropathy over a decade in insulindependent diabetics. Q. J. Med. 75:635, 1990. Sandroni, P., Ahlskog, J. E., Fealey, R. D., and Low, P. A.: Autonomic involvement in extrapyramidal and cerebellar disorders. Clin. Auton. Res. 1:147, 1991.

1132

Quantitating Symptoms, Impairments, and Outcomes

146. Sandroni, P., Benarroch, E. E., and Low, P. A.: Pharmacological dissection of components of the Valsalva maneuver in adrenergic failure. J. Appl. Physiol. 71:1563, 1991. 147. Sandroni, P., Low, P. A., Ferrer, T., et al.: Complex regional pain syndrome I (CRPS I): prospective study and laboratory evaluation. Clin. J. Pain 14:282, 1998. 148. Sato, K.: The physiology, pharmacology, and biochemistry of the eccrine sweat gland. Rev. Physiol. Biochem. Pharmacol. 79:51, 1977. 149. Sato, K.: Normal and abnormal sweat gland function. In Low, P. A. (ed.) Clinical Autonomic Disorders: Evaluation and Management. Philadelphia, Lippincott–Raven, p. 97, 1997. 150. Satoh, A., Serita, T., Seto, M., et al.: Loss of 123I-MIBG uptake by the heart in Parkinson’s disease: assessment of cardiac sympathetic denervation and diagnostic value. J. Nucl. Med. 40:371, 1999. 151. Schnell, O., Kirsch, C. M., Stemplinger, J., et al.: Scintigraphic evidence for cardiac sympathetic dysinnervation in long-term IDDM patients with and without ECGbased autonomic neuropathy. Diabetologia 38:1345, 1995. 152. Schondorf, R., and Low, P. A.: Idiopathic postural orthostatic tachycardia syndrome: an attenuated form of acute pandysautonomia? Neurology 43:132, 1993. 153. Schrag, A., Good, C. D., Miszkiel, K., et al.: Differentiation of atypical parkinsonian syndromes with routine MRI. Neurology 54:697, 2000. 154. Schrag, A., Kingsley, D., Phatouros, C., et al.: Clinical usefulness of magnetic resonance imaging in multiple system atrophy. J. Neurol. Neurosurg. Psychiatry 65:65, 1998. 155. Shore, A. C., Price, K. J., Sandeman, D. D., et al.: Posturally induced vasoconstriction in diabetes mellitus. Arch. Dis. Child. 70:22, 1994. 156. Smith, A. A., and Dancis, J.: Exaggerated response to infused norepinephrine in familial dysautonomia. N. Engl. J. Med. 270:704, 1964. 157. Smith, G. D., Watson, L. P., Pavitt, D. V., and Mathias, C. J.: Abnormal cardiovascular and catecholamine responses to supine exercise in human subjects with sympathetic dysfunction. J. Physiol. (Lond.) 484:255, 1995. 158. Smith, J. J., Porth, C. M., and Erickson, M.: Hemodynamic response to the upright posture. J. Clin. Pharmacol. 34:375, 1994. 159. Smith, S. A., Stallard, T. J., and Littler, W. A.: Estimation of sinoaortic baroreceptor heart rate reflex sensitivity and latency in man: a new microcomputer assisted method of analysis. Cardiovasc. Res. 20:877, 1986. 160. Smyth, H. S., Sleight, P., and Pickering, G. W.: Reflex regulation of arterial pressure during sleep in man: a quantitative method of assessing baroreflex sensitivity. Circ. Res. 24:109, 1969. 161. So, Y. T., Aminoff, M. J., and Olney, R. K.: The role of thermography in the evaluation of lumbosacral radiculopathy. Neurology 39:1154, 1989. 162. Sprangers, R. L., Veerman, D. P., Karemaker, J. M., and Wieling, W.: Initial circulatory responses to changes in posture: influence of the angle and speed of tilt. Clin. Physiol. 11:211, 1991. 163. Stevens, M. J., Dayanikli, F., Raffel, D. M., et al.: Scintigraphic assessment of regionalized defects in myocar-

164.

165.

166.

167.

168.

169.

170.

171.

172.

173. 174.

175.

176.

177.

178.

179.

180.

dial sympathetic innervation and blood flow regulation in diabetic patients with autonomic neuropathy. J. Am. Coll. Cardiol. 31:1575, 1998. Stevens, P. M., and Lamb, L. E.: Effects of lower body negative pressure on the cardiovascular system. Am. J. Cardiol. 16:506, 1965. Stewart, J. D., Low, P. A., and Fealey, R. D.: Small-fiber peripheral neuropathy: diagnostic value of sweat tests and autonomic cardiovascular reflexes. Ann. Neurol. 26:145A, 1989. Stewart, J. D., Low, P. A., and Fealey, R. D.: Distal small fiber neuropathy: results of tests of sweating and autonomic cardiovascular reflexes. Muscle Nerve 15:661, 1992. Stjernberg, L., Blumberg, H., and Wallin, B. G.: Sympathetic activity in man after spinal cord injury: outflow to muscle below the lesion. Brain 109:695, 1986. Takano, H., Ozawa, H., Kobayashi, I., et al.: Atrophic nerve fibers in regions of reduced MIBG uptake in doxorubicin cardiomyopathy. J. Nucl. Med. 36:2060, 1995. Takatsu, H., Nagashima, K., Murase, M., et al.: Differentiating Parkinson disease from multiple-system atrophy by measuring cardiac iodine-123 metaiodobenzylguanidine accumulation. JAMA 284:44, 2000. Tani, H., Singer, W., McPhee, B. R., et al.: Splanchnicmesenteric capacitance bed in the postural tachycardia syndrome (POTS). Auton. Neurosci. 86:107, 2000. ten Harkel, A. D. J., van Lieshout, J. J., van Lieshout, E. J., and Wieling, W.: Assessment of cardiovascular reflexes: influence of posture and period of preceding rest. J. Appl. Physiol. 68:147, 1990. Thomaides, T. N., Chaudhuri, K. R., Maule, S., et al.: Growth hormone response to clonidine in central and peripheral primary autonomic failure. Lancet 340:263, 1992. Thoren, P.: Role of cardiac vagal C-fibers in cardiovascular control. Rev. Physiol. Biochem. Pharmacol. 86:1, 1979. Torres, N. E., Zollman, P. J., and Low, P. A.: Characterization of muscarinic receptor subtype of rat eccrine sweat gland by autoradiography. Brain Res. 550:129, 1991. Trimarco, B., Volpe, M., Ricciardelli, B., et al.: Valsalva maneuver in the assessment of baroreflex responsiveness in borderline hypertensives. Cardiology 70:6, 1983. Tsuchimochi, S., Tamaki, N., Tadamura, E., et al.: Age and gender differences in normal myocardial adrenergic neuronal function evaluated by iodine-123-MIBG imaging. J. Nucl. Med. 36:969, 1995. van Lieshout, E. J., van Lieshout, J. J., ten Harkel, A. D., and Wieling, W.: Cardiovascular response to coughing: its value in the assessment of autonomic nervous control. Clin. Sci. 77:305, 1989. van Lieshout, E. J., van Lieshout, J. J., Wieling, W., et al.: Monitoring of finger blood pressure. J. Neurol. Neurosurg. Psychiatry 50:503, 1987. van Lieshout, J. J., Wieling, W., Wesseling, K. H., and Karemaker, J. M.: Pitfalls in the assessment of cardiovascular reflexes in patients with sympathetic failure but intact vagal control. Clin. Sci. 76:523, 1989. Vaz, M., Cox, H. S., Kaye, D. M., et al.: Fallibility of plasma noradrenaline measurements in studying postprandial sympathetic nervous responses. J. Auton. Nerv. Syst. 56:97, 1995.

Quantitation of Autonomic Impairment 181. Vetrugno, R., Liguori, R., Cortelli, P., and Montagna, P.: Sympathetic skin response: basic mechanisms and clinical applications. Clin. Auton. Res. 13:256, 2003. 182. Veves, A., Akbari, C. M., Primavera, J., et al.: Endothelial dysfunction and the expression of endothelial nitric oxide synthetase in diabetic neuropathy, vascular disease, and foot ulceration. Diabetes 47:457, 1998. 183. Wallin, B. G.: Intraneural recordings of normal and abnormal sympathetic activity in humans. In Mathias, C. J., and Bannister, R. (eds.): Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System. Oxford, UK, Oxford University Press, p. 224, 2002. 184. Wallin, B. G., and Eckberg, D. L.: Sympathetic transients caused by abrupt alterations of carotid baroreceptor activity in humans. Am. J. Physiol. 242:H185, 1982. 185. Webster, J., Newnham, D., Petrie, J. C., and Lovell, H. G.: Influence of arm position on measurement of blood pressure. Br. Med. J. (Clin. Res. Ed.) 288:1574, 1984. 186. Wei, J. Y., and Harris, W. S.: Heart rate response to cough. J. Appl. Physiol. 53:1039, 1982. 187. Wei, J. Y., Rowe, J. W., Kestenbaum, A. D., and Ben-Haim, S.: Post-cough heart rate response: influence of age, sex and basal blood pressure. Am. J. Physiol. 245:R18, 1983. 188. Wei, K., Dorian, P., Newman, D., and Langer, A.: Association between QT dispersion and autonomic dysfunction in patients with diabetes mellitus. J. Am. Coll. Cardiol. 26:859, 1995. 189. Weiss, B. M., Spahn, D. R., Rahmig, H., et al.: Radial artery tonometry: moderately accurate but unpredictable technique of continuous non-invasive arterial pressure measurement. Br. J. Anaesth. 76:405, 1996.

1133

190. Wesseling, K. H., de Witt, B., Settels, J. J., et al.: On the indirect registration of finger blood pressure after Penaz. Funkt. Biol. Med. 1:245, 1982. 191. Wesseling, K. H., Settels, J. J., van der Hoeven, G. M., et al.: Effects of peripheral vasoconstriction on the measurement of blood pressure in a finger. Cardiovasc. Res. 19:139, 1985. 192. Wexler, L., Bergel, D. H., Gabe, I. T., et al.: Velocity of blood flow in normal human venae cavae. Circ. Res. 23:349, 1968. 193. Widdicombe, J. G.: Mechanism of cough and its regulation. Eur. J. Respir. Dis. Suppl. 110:11, 1980. 194. Williams, T. D., Lightman, S. L., and Bannister, R.: Vasopressin secretion in progressive autonomic failure: evidence for defective afferent cardiovascular pathways. J. Neurol. Neurosurg. Psychiatry 48:225, 1985. 195. Wu, C. L., Denq, J. C., Harper, C. M., et al.: Clinical and electrophysiologic attributes as predictors of results of autonomic function tests. Clin. Auton. Res. 8:347, 1998. 196. Yoshita, M.: Differentiation of idiopathic Parkinson’s disease from striatonigral degeneration and progressive supranuclear palsy using iodine-123 meta-iodobenzylguanidine myocardial scintigraphy. J. Neurol. Sci. 155:60, 1998. 197. Zerbe, R. L., Henry, D. P., and Robertson, G. L.: Vasopressin response to orthostatic hypotension: etiologic and clinical implications. Am. J. Med. 74:265, 1983. 198. Ziegler, D., Laux, G., Dannehl, K., et al.: Assessment of cardiovascular autonomic function: age-related normal ranges and reproducibility of spectral analysis, vector analysis, and standard tests of heart rate variation and blood pressure responses. Diabet. Med. 9:166, 1992.

A. Differential Diagnosis and Epidemiology

45 Clinical Patterns of Peripheral Neuropathy PHILIP D. THOMPSON AND P. K. THOMAS

Distribution and Patterns of Involvement Distal Symmetrical Polyneuropathy Proximal Symmetrical and Upper Limb–Predominant Polyneuropathy Complex Distributions Cranial Nerve Involvement Focal and Multifocal Neuropathies Functional Selectivity Motor Neuropathy Sensory Neuropathy Autonomic Neuropathy

Clinical Course Intermittent Symptoms Symptoms and Signs Negative Motor Symptoms and Signs Positive Motor Symptoms and Signs Sensory Symptoms Negative Sensory Symptoms and Signs Positive Sensory Symptoms Autonomic Involvement Skeletal Deformity

The term peripheral neuropathy encompasses a variety of topographic, functional, and pathologic disturbances. When the pathologic process involves the peripheral nervous system, in either a focal or generalized manner, the concept and definition of “peripheral” neuropathy is straightforward. However, sensory “neuropathies” can also be produced by involvement within the central nervous system (CNS) of the central processes of dorsal root ganglion cells in a central-peripheral distal axonopathy,209,226 or the dorsal root ganglion cells themselves. The clinical picture of a motor “neuropathy” may be produced by primary degeneration of the cell bodies (anterior horn cells) and proximal axons of lower motor neurons within the spinal cord. Neuropathies in which axonal degeneration is accompanied by preservation of cell bodies are referred to as “axonopathies,” while loss of the whole neuron, including cell body and axon, is referred to as a “neuronopathy.” In functional terms, use of the term neuropathy is to some extent arbitrary. Asymptomatic patients with diabetes or chronic renal failure may exhibit signs of a neuropathy on clinical examination, and

Trophic Changes Nerve Thickening Differential Diagnosis Disorders Simulating Peripheral Neuropathy Neuropathies Simulating Other Conditions Diagnosis of Neuropathy Confirmation of Neuropathy Establishing Etiology

disturbances of nerve conduction may be demonstrable in the absence of either neuropathic symptoms or signs in diabetes and uremia.163

DISTRIBUTION AND PATTERNS OF INVOLVEMENT The pattern and distribution of clinical involvement often provide important clues to the cause of the disease process responsible for a neuropathy. Two main patterns are recognized. The most common pattern is a bilateral and symmetrical disturbance of function, which is referred to as a “polyneuropathy.” The terms polyradiculopathy and polyradiculoneuropathy are also used if there is significant involvement of spinal nerve roots in addition to the peripheral nerve trunks. Polyneuropathies tend to be associated with diffuse afflictions of the peripheral nervous system caused by toxins, nutritional deficiencies, systemic metabolic disorders, or immunologically mediated illnesses. 1137

1138

Diseases of the Peripheral Nervous System

A second pattern of peripheral nerve involvement comprises focal lesions of single nerves (mononeuropathy) or multifocal involvement of isolated peripheral nerves (mononeuritis multiplex). Focal or multifocal nerve damage occurs in nerve entrapment, mechanical injury or other trauma, vascular lesions, and granulomatous, infiltrative, and neoplastic processes.

Distal Symmetrical Polyneuropathy The most common pattern of a generalized polyneuropathy is a symmetrical distal motor and sensory deficit. Symptoms usually begin in the legs, followed by the upper limbs. Weakness and wasting begin in the distal limbs and spread proximally (Fig. 45–1). Sensory involvement presents with a distal glove-and-stocking sensory loss (Fig. 45–2). With progression of the neuropathy, the sensory loss in the limbs spreads proximally. Involvement of the longest fibers in the intercostal nerves may lead to sensory loss on the anterior abdominal wall. As the neuropathy advances, sensory loss progresses from the anterior abdominal wall laterally around the trunk, so that, unless the back of the trunk where sensation is still intact is examined, a myelopathy with a sensory level may be suspected. The vertex of the head and central face may be affected (see Fig. 45–2). In a severe sensory neuropathy, sensory loss may be extensive, sparing only a midline strip over the posterior trunk and neck, and the periphery of the face (Fig. 45–3). Depression and loss of tendon reflexes follow the same pattern, affecting the ankle jerks before the knee jerks and the lower limbs before the upper. A variety of factors may contribute to the distal lengthrelated distribution of clinical signs in neuropathy. The longest axons are the most vulnerable in dying-back axonopathies, while in demyelinating neuropathies and multifocal neuropathies, the longest fibers have the greatest chance of being affected. A strict length-related explanation for distaldominant involvement is not always easily applicable. In the lower limbs, the anterior tibial and peroneal muscles are usually affected before the calf muscles. This has been attributed to the greater length of the nerves supplying the anterolateral compartment muscles as compared with those innervating the posterior compartment. Such a pattern of weakness (referred to as “peroneal muscular atrophy”) characterizes type I hereditary motor and sensory neuropathy (HMSN).96 Yet in type II HMSN, the posterior compartment muscles are often affected simultaneously and with a severity of involvement similar to that of the anterolateral muscles.96

Proximal Symmetrical and Upper Limb–Predominant Polyneuropathy A notable exception to the distal emphasis of symptoms in neuropathy is seen occasionally in dysimmune neuropathies such as the Guillain-Barré syndrome and, rarely,

FIGURE 45–1 Patient with type 1 hereditary and motor sensory neuropathy showing distal distribution of muscle wasting particularly affecting the lower limbs. Note also the foot deformity.

chronic progressive inflammatory demyelinating polyneuropathy, in which a proximal distribution of muscle weakness may be observed and sensory symptoms may begin in the upper limbs. A proximal distribution of muscle weakness also may be observed in acute intermittent porphyria. Proximal muscle wasting and weakness in the upper limbs is characteristic of inherited motor neuronopathy (spinal muscular atrophy).145 The neuropathy of vitamin B12 deficiency occasionally begins in the upper limbs with sensory symptoms and clumsiness resulting from deafferentation of the hands. In the type II (Indiana) form of hereditary amyloid neuropathy, sensory loss often begins in the upper limbs with bilateral median nerve lesions resulting from

Clinical Patterns of Peripheral Neuropathy

1139

FIGURE 45–2 Diabetic sensory polyneuropathy. A, At an earlier stage, the sensory loss (shaded areas) shows a distally accentuated distribution in the limbs together with involvement of a strip over the anterior abdominal wall. B, At a later stage, the sensory loss has extended more proximally in the limbs and now affects an extensive area over the anterior trunk; the vertex of the head is also involved. (From Waxman, S. G., and Sabin, T. D.: Diabetic truncal neuropathy. Arch. Neurol. 38:36, 1981, with permission.)

compression by amyloid deposits under the flexor retinacula at the wrists.134 Proximal sensory loss may be observed in acute intermittent porphyria (“bathing trunks” distribution) and in Tangier disease, in which the whole of the trunk may be affected, with relative sparing of the distal extremities116,118 (Fig. 45–4). The explanation for selective proximal involvement is unknown. A possible answer may lie in the fact that the proximal portions of the limbs are supplied by nerve fibers of larger diameter than those innervating the more distal regions and the trunk,181 and such fibers may have differing susceptibility to metabolic change and immunologic insults. Symmetrical or asymmetrical proximal sensory loss may begin in the upper limbs in a sensory ganglionopathy. Widespread sensory loss involving proximal and distal limbs is more commonly a clinical feature of a ganglionopathy than a polyneuropathy.

Complex Distributions Some motor neuronopathies exhibit distinctive patterns of involvement. The scapuloperoneal form of hereditary

bulbospinal muscular atrophy (Davidenkow’s syndrome) presents with proximal upper limb and distal lower limb weakness.197 X-linked bulbospinal neuronopathy (Kennedy’s syndrome) presents with facial and bulbar weakness in combination with proximal limb weakness.98 In lepromatous leprosy, a distal sensory loss may be accompanied by a distinctive pattern of sensory loss affecting the helices of the pinnae, the tip of the nose, and the malar regions of the cheeks. These regions are cooler, at which temperatures leprosy bacilli proliferate more rapidly.188,201

Cranial Nerve Involvement Focal and multifocal cranial nerve lesions can occur in isolation or in combination with a generalized polyneuropathy. Facial and oculomotor nerve involvement may occur in Guillain-Barré syndrome and chronic inflammatory demyelinating neuropathies. Cranial nerves tend to be involved later in the evolution of neuropathy (after the limbs), but occasionally the reverse sequence is observed.55,233 An external ophthalmoplegia can be the presenting feature of the Miller Fisher syndrome. Isolated or multiple cranial

1140

Diseases of the Peripheral Nervous System

FIGURE 45–3 Hereditary sensory neuropathy with extensive distal sensory loss (shaded areas). The only areas spared are a narrow midline strip over the posterior trunk, the back of the neck, and the peripheral parts of the face. (From Sabin, T. D., Geschwind, N., and Waxman, S. G.: Patterns of clinical deficit in peripheral nerve disease. In Waxman, S. G. [ed.]: Physiology and Pathobiology of Axons. New York, Raven Press, p. 431, 1978, with permission.)

nerve lesions are seen in sarcoidosis, diabetes mellitus, borreliosis, and neoplastic invasion of the meninges or skull base.

Focal and Multifocal Neuropathies Focal neuropathies or mononeuropathies are commonly caused by compression of peripheral nerves at sites where they are exposed, particularly those passing through fibrous arches or adjacent to a bone (Table 45–1). Symptoms consist of weakness, paresthesias, numbness, and pain in the distribution of the affected peripheral nerve. Diffuse polyneuropathies may be complicated by superimposed isolated compressive mononeuropathies as a consequence of susceptibility to pressure palsies. This occurs in diabetic neuropathy,81,158,199 uremic neuropathy,163 and hereditary neuropathy with liability to pressure palsies.147 Symptoms in multiple peripheral nerve territories indicate a multifocal neuropathy or mononeuritis multiplex, as may occur in vasculitis, diabetes, and hereditary liability to pressure palsies. In widespread multifocal neuropathies, the lesions may summate to mimic a polyneuropathy. For example, multifocal sensory neuropathies may lead to an apparent glove-and-stocking distribution of sensory loss. In such

FIGURE 45–4 Tangier disease. The shaded regions indicate the distribution of sensory impairment for pain and temperature. There is sparing distally in the limbs. (From Kocen, R. S., Lloyd, J. K., Lascelles, P. T., et al.: Familial ␣-lipoprotein deficiency [Tangier disease with neurological abnormalities]. Lancet 1:1341, 1967, with permission.)

cases, the history of the evolution of symptoms and careful examination may reveal changes referable to sequential involvement of particular nerves. In some multifocal neuropathies, such as diabetic amyotrophy79 and neuralgic amyotrophy (idiopathic brachial plexus neuropathy), there is a proximal distribution as a result of involvement of limb-girdle plexuses, and the deficits do not conform to peripheral nerve territories.

Table 45–1. Common Compressive or Entrapment Mononeuropathies Ulnar nerve (elbow) Median nerve (wrist—carpal tunnel syndrome) Radial nerve (upper arm—“Saturday night” palsy) Posterior cutaneous thoracic rami (notalgia paresthetica) Thoracic spinal nerves (intercostal neuralgia) Lateral cutaneous nerve of the thigh (meralgia paresthetica) Peroneal nerve (fibula, knee) Saphenous nerve Infrapatellar branch of saphenous nerve (gonalgia paresthetica) Tibial nerve (ankle—tarsal tunnel syndrome) Plantar interdigital nerves (foot—Morton’s metatarsalgia)

1141

Clinical Patterns of Peripheral Neuropathy

FUNCTIONAL SELECTIVITY Motor Neuropathy Predominant motor involvement can be observed in the Guillain-Barré syndrome, chronic demyelinating neuropathy,141 notably, multifocal motor neuropathy, porphyria, lead intoxication,149 and diphtheria. It may also be seen in HMSN60,61 and the hereditary motor neuropathies and spinal muscular atrophies.91

Sensory Neuropathy Sensory changes are more easily overlooked than motor involvement, and their assessment is more demanding. Predominant sensory involvement is observed in diabetic polyneuropathy, vitamin B12 deficiency, leprosy, uremic neuropathy,230 and hereditary sensory neuropathy. Some selective sensory neuropathies, including hereditary sensory neuropathy type I,47 the neuropathy of Friedreich’s ataxia, carcinomatous or paraneoplastic sensory neuropathy,46 and the sensory (ataxic) neuropathy accompanying Sjögren’s syndrome,90,115 are due to degeneration of dorsal root ganglion cells.

Autonomic Neuropathy Autonomic changes are characteristic in acute pandysautonomia and the Riley-Day syndrome,3,44 in which they are accompanied by congenital absence of pain sensibility. Autonomic neuropathies are an early and troublesome feature of amyloid neuropathy73 and may be prominent in diabetic neuropathy and a variety of other neuropathies.143

CLINICAL COURSE An abrupt onset is typical of traumatic nerve lesions resulting from direct external compression, penetrating wounds, and thermal injury. A rapid onset is also observed in focal or multifocal ischemic neuropathies in polyarteritis nodosa, rheumatoid arthritis, diabetic cranial nerve palsies,6 and diabetic amyotrophy.79 The rapid onset of a focal neuropathy also occurs in nerve compression caused by tissue swelling or hemorrhage within a restricted anatomic compartment. Neuropathy evolving over days may occur in intoxication with thallium or tri-o-cresyl phosphate and in metabolic neuropathies, such as acute intermittent porphyria. Diabetic neuropathy may evolve rapidly during initiation of treatment, as may uremic neuropathy during hemodialysis.222,230 A “critical illness polyneuropathy” may also appear over a similar time course in the setting of systemic illness (see Chapter 87). Immune neuropathies such as the Guillain-Barré syndrome, neuralgic amyotrophy, and diphtheritic neuropathy typically evolve over several days following a characteristic delay between the initiating process and the development of neuropathy.

A subacute onset of neuropathy over weeks or months in seen in many toxic, nutritional, and metabolic neuropathies. Paraneoplastic neuropathies and chronic inflammatory demyelinating neuropathies may also pursue a similar time course. A chronic course with an insidious onset and slow progression is typical of the inherited neuropathies, such as the HMSNs, inherited sensory neuropathies, and Refsum’s disease. It is also characteristic of chronic inflammatory demyelinating polyneuropathies.54,141,234 The “migrant sensory neuritis of Wartenberg” is a benign relapsing and remitting condition in which stretching of cutaneous nerves leads to focal loss of cutaneous sensation in the distribution of the affected nerve.140,249 The onset usually is sudden, precipitated by movement and associated with pain at the site of nerve involvement. Symptoms recover over the following weeks, only to recur at the same or another site. A relapsing and remitting course over weeks and months is evident in chronic relapsing inflammatory demyelinating neuropathy. The clinical course of a neuropathy reflects the underlying pathologic changes. A rapid and complete recovery may be seen in dysimmune neuropathies, in which the clinical features are due to conduction block related to demyelination. Similar mechanisms probably underlie recovery from Bell’s palsy and some compressive neuropathies, such as “Saturday night” paralysis of the radial nerve. In contrast, recovery from neuropathies associated with extensive axonal destruction, provided the causal process is reversible, requires axonal regeneration and collateral sprouting from surviving axons and therefore is slow and often incomplete.

Intermittent Symptoms In general, symptoms in polyneuropathies are persistent, but transient symptoms may occur and are of considerable practical and theoretical interest. Their occurrence may be of assistance diagnostically, and may provide an opportunity for the investigation of pathophysiologic mechanisms. Examples are nocturnal attacks of intermittent paresthesias and numbness in the carpal tunnel syndrome, and the burning paresthesias of meralgia paresthetica that occur on standing and may be relieved by walking. Lewis and co-workers demonstrated that the application of a tourniquet at greater than arterial pressure around the upper arm in normal subjects, after first giving rise to ischemic paresthesias, leads to sensory loss and paralysis after an interval of about 30 minutes.127 This is rapidly reversible when the cuff is released. Recovery is associated with postischemic paresthesias. That these effects are due to ischemia of the nerve distal to the cuff, and not the direct pressure effect by the tourniquet, was shown by inflating a second cuff below the first after sensory loss had developed. On release of the upper cuff, the sensory loss persisted. The transient symptoms experienced by most

1142

Diseases of the Peripheral Nervous System

people as a result of localized nerve compression are likely to be of this nature. Persisting symptoms may be related to mechanical effects of the type shown to occur in experimental tourniquet paralysis.172 The attacks of paresthesia that constitute an important aspect of the symptoms of the carpal tunnel syndrome characteristically occur at night and have been the subject of a number of investigations.74,82,83 This work established that, in such cases, the median nerve is abnormally susceptible to ischemia produced by inflation of a pneumatic cuff around the arm.82,83 Gilliatt and Wilson82,83 found that median nerve sensory loss appeared with abnormal rapidity, and Fullerton74 showed that conduction in motor fibers failed with abnormal rapidity because of a conduction block in fibers distal to the wrist. Gilliatt and Wilson82 also noted that patients reported a similarity between the symptoms produced by ischemia and those of their spontaneous attacks. All these observations suggest that the attacks of paresthesias are related to episodes of ischemia. Further evidence that ischemia may precipitate attacks of acroparesthesias has been provided by the occurrence of such attacks during dialysis in patients with carpal tunnel syndrome under treatment for renal failure by chronic hemodialysis using an antebrachial Cimino-Brescia fistula.94 A vascular steal effect was postulated for the precipitation of attacks and was supported by their cessation after distal ligature of the radial artery in one instance. A vascular steal effect was also postulated for the provocation of attacks in a case of congenital arteriovenous fistula affecting the hand, associated with the carpal tunnel syndrome.34

SYMPTOMS AND SIGNS Negative Motor Symptoms and Signs Paralysis of voluntary movement or muscle weakness in neuropathy may result from conduction block in motor fibers, or from interruption of motor axons. Apart from temporary conduction block caused by local anesthesia, cooling, or anoxia, persisting conduction block is due to selective demyelination with preservation of axonal continuity. This has been demonstrated for experimental demyelinating neuropathies,38,142 in experimental tourniquet paralysis,48,172 and after the intraneural inoculation of antigalactocerebroside serum and serum from patients with the Guillain-Barré syndrome.192,216 Characteristic examples of muscle weakness resulting from localized conduction block include “Saturday night” paralysis caused by compression of the radial nerve; footdrop caused by compression of the common peroneal nerve, such as the “strawberry picker’s palsy”119; and facial weakness in Bell’s palsy. The presence of muscle contraction after stimulation of the affected nerve below the site of compression indicates axonal continuity and preservation

of conduction distal to the lesion. Some disuse wasting of the muscle may develop, but denervation atrophy does not occur when axonal continuity is maintained. Paralysis in demyelinating neuropathies, such as the Guillain-Barré syndrome and multifocal motor neuropathy, may be due to conduction block caused by demyelination or axonal hyperpolarization.220 As is true for localized demyelinating lesions, muscle bulk is maintained, electromyographic signs of denervation do not appear unless there is axonal interruption, and recovery is characteristically rapid and complete. When a localized nerve lesion or generalized polyneuropathy interrupts motor axons, weakness is accompanied by denervation atrophy of muscle. Atrophy becomes clinically evident within a few weeks and is progressive. Permanent muscle fiber loss in denervated muscle begins after 6 to 9 months, and few fibers remain after 3 years.25 Restoration of function is usually possible if reinnervation is established within 12 months.217

Positive Motor Symptoms and Signs Fasciculation Fasciculations are visible spontaneous twitches of muscle caused by sporadic discharges of motor units. The motor units may have a normal or abnormal morphology on electromyographic recordings, depending on the degree of denervation and reinnervation. Fasciculations are typically generated in terminal motor axons but can arise from a variety of locations on the motor axon.184 Fasciculations indicate denervation of muscle caused by a lesion of the motor neuron at any site from the anterior horn cell to the terminal motor axon. Fasciculation is most striking in anterior horn cell disease. Fasciculations may occur as a benign phenomenon without accompanying denervation in thyrotoxicosis, and are seen in normal subjects during muscle fatigue. Fibrillation Fibrillation refers to the spontaneous discharge of single muscle fibers, recorded electromyographically. The resulting muscle twitches are too small to be visible in skeletal muscles but may be seen in the tongue. Fibrillations reflect denervation of muscle fibers as a result of lower motor neuron damage and axonal degeneration, but may also be found in inflammatory myopathies as the result of intramuscular nerve damage. Myokymia Myokymia is defined clinically as an undulating, wavelike, vermicular or wormlike rippling of muscle.49 In electromyography, “myokymia” refers to regular or semirhythmic groups of motor units discharging in doublets, triplets, or multiplets (see Fig. 45–5). Myokymia occurs in diseases of motor axons. Generalized myokymia may occur in Guillain-Barré

Clinical Patterns of Peripheral Neuropathy

syndrome and HMSNs and is a central feature of the peripheral nerve hyperexcitability syndromes associated with neuromyotonia and delayed muscle relaxation. Focal myokymia is seen in radiation plexopathies,4 chronic radiculopathies, and focal compressive neuropathies.146 Muscle Cramps Muscle cramps consist of painful contractions of part of a muscle or whole muscles related to active involuntary firing of motor units. Healthy people commonly experience painful nocturnal muscle cramps without apparent predisposing factors. Benign, physiologic cramps occur abruptly while at rest or during sleep or may interrupt exercise. A distinguishing feature of benign physiologic cramps is their relief by forcibly stretching the affected muscle. This effect may be mediated through activation of Golgi tendon organ afferents. Cramps commonly occur in association with a number of conditions, such as dehydration, salt depletion, other electrolyte disturbances, and pregnancy. They are common in partially denervated muscle and may antedate the onset of muscle weakness.98 Cramps are thought to originate from spontaneous activity in the terminal motor axons.120 Electromyography during cramp discloses variable high-frequency discharge of normal motor units at several sites, recruited either simultaneously or sequentially.167 These features distinguish cramps from other syndromes of spontaneous muscle activity of peripheral nerve origin, myotonia, and the electrically silent muscle contracture. Peripheral Nerve Hyperexcitability Syndromes: Neuromyotonia and Continuous Motor Unit Activity The term neuromyotonia was introduced to describe a clinical syndrome of delayed muscle relaxation after voluntary contraction of peripheral nerve origin and to distinguish this from myotonia.153 The syndrome is caused by continuous motor unit and muscle fiber activity resulting from peripheral nerve hyperexcitability,99 and is clinically and electrophysiologically distinct from myotonia related to hyperexcitability of the muscle fiber membrane. The term pseudomyotonia has also been used in this sense.198 Both “neuromyotonia” and “pseudomyotonia” also describe the electromyographic phenomenon of complex and bizarre high-frequency discharges of muscle fibers. The designation “peripheral nerve hyperexcitability syndromes” emphasizes the peripheral nerve origin (specifically the motor axons) of the abnormal muscle discharges and the resulting clinical syndromes99 (Table 45–2). These syndromes are characterized by continuous muscle activity, myokymia, and fasciculation. Symptoms include muscle twitching, rippling, cramps, delayed muscle relaxation, muscle stiffness, and, in some cases, excessive sweating.76,78,89,108 Muscle stiffness typically follows voluntary muscle contraction but is also evident at rest and usually persists during sleep.132 Stiffness is most evident in distal

1143

Table 45–2. Syndromes of Peripheral Nerve Hyperexcitability, Defined as Continuous Motor Unit Activity, Neuromyotonia, and Delayed Muscle Relaxation Inherited Isolated (without neuropathy) Associated with hereditary neuropathy Associated with spinal muscular atrophy Schwarz-Jampel syndrome Antibody Mediated (Voltage-Gated K Channel Antibodies) Isolated Associated with central nervous system features (Morvan’s syndrome) Paraneoplastic Thymoma (⫾ myasthenia gravis) Small cell lung carcinoma (⫾ neuropathy) Adenocarcinoma Associated with “idiopathic” peripheral neuropathy Associated with autoimmune disorders Myasthenia gravis (without thymoma) Systemic lupus erythematosus, scleroderma Guillain-Barré syndrome Mutations of Voltage-Gated K Channel (KCNA1) Familial episodic ataxia type 1 Acquired Associated with neuropathy Idiopathic Diabetes mellitus Chronic inflammatory demyelinating neuropathy Celiac disease Paraneoplastic Paraproteinemic Thyroid disease Drugs Penicillamine Gold Radiation Toxins Data from Hart, I. K., Maddison, P., Newson-Davis, J., et al.: Phenotypic variants of autoimmune peripheral nerve hyperexcitability. Brain 125: 1887, 2002.

muscles of the limbs, but both proximal and distal muscles are affected. Delayed relaxation and stiffness lead to sustained abnormal postures, resembling tetany in the hand, and an awkward gait. Periocular159 and laryngeal109 muscles may be affected. Tendon reflexes are often absent. Areflexia may be due to the presence of an underlying neuropathy or occlusion and inhibition of the spinal monosynaptic reflexes by the continuous muscle activity (since reflexes may return with remission or effective treatment of the continuous motor activity78,107,131,132,246). Muscle hypertrophy159,203,257 and fatigability of voluntary muscle contraction

1144

Diseases of the Peripheral Nervous System

may be seen. Distal paresthesias are reported in some cases, indicating that sensory axons may also be involved.99 Electromyography reveals a variety of motor unit discharges, including fasciculations, single motor units and grouped discharges of motor units in doublets or triplets, and high-frequency discharges of muscle fibers238 (Fig. 45–5). A characteristic finding is the presence of afterdischarges following voluntary contraction, peripheral nerve stimulation, and even gentle percussion of peripheral nerves (see Fig. 45–5). These afterdischarges are the physiologic basis for the delayed muscle relaxation after voluntary contraction. The continuous motor unit activity persists during sleep21,132,159 and following spinal or general anesthesia, but is abolished by curarization.21,131,246 Peripheral nerve anesthesia at various sites along the length of the motor axon may diminish the activity.10,107,243 These findings, and the absence of jitter between discharges,238 indicate that the motor activity is

A: Peroneal nerve stimulation.

B: Voluntary contraction.

After discharge

After discharge

0.25 mV 250 ms

250 ms

C: SPONTANEOUS ACTIVITY.

400 uV 100 ms

FIGURE 45–5 Muscle activity in neuromyotonia. A, Surface electromyographic recordings (five trials superimposed) from tibialis anterior muscle following stimulation of the peroneal nerve at the head of the fibula. B, Concentric needle electromyographic recordings of brief voluntary activation of the tibialis anterior muscle (eight trials superimposed) in a patient with oat cell carcinoma of the lung who presented with symptoms of muscle stiffness and exhibited delayed muscle relaxation and myokymia on examination. Stimulation of the peroneal nerve elicits an M wave in the tibialis anterior, followed by an afterdischarge that begins after the F wave. The subject was asked to make brief voluntary contractions of the tibialis anterior by dorsiflexing the foot. Each voluntary contraction is followed by an afterdischarge. Peripheral nerve conduction velocities were normal. C, Concentric needle electromyography revealed continuous muscle activity at rest comprising motor unit discharges in doublets and triplets (myokymia) and smaller muscle fiber discharges. Treatment with carbamazepine improved the symptoms of muscle stiffness and reduced the continuous muscle activity.

generated throughout the peripheral motor axons,91,123,246 particularly distally in the terminal motor arborization.108,213 A wide variety of labels have been applied to this clinical syndrome. It is often referred to as Isaacs’ syndrome,108 although there were earlier descriptions.49,76 The syndrome has been encountered in both axonopathies and demyelinating neuropathies, including HMSNs,76,123,242 inherited motor neuropathies,92 chronic demyelinating neuropathy,239 toxic neuropathy,155,246 paraneoplastic neuropathy,179,247,251 and neuropathies of unknown cause.76,89,131,132,243 In many cases, peripheral nerve hyperexcitability occurs without evidence of an associated or underlying neuropathy. The disorder may be inherited,10,11 and can occur in association with autoimmune diseases,99 autoantibody-mediated paraneoplastic syndromes,77,93,99 and mutations in the voltage-gated potassium channel gene in episodic ataxia type 1.27,241 Newsom-Davis and colleagues have demonstrated an autoimmune etiology involving antibodies to voltage-gated potassium channels in more than 30% of cases.99,162 This observation has led to a new classification of the peripheral nerve hyperexcitability syndromes based on etiology and recognizing these various associations (see Table 45–2). Neuromyotonia is abolished by phenytoin,91,108,132,246,257 carbamazepine,91,123,159 and tocainide,91 agents that reduce conductance in sodium channels. The distribution of symptoms, characteristic electromyographic findings, and response to treatment help differentiate peripheral nerve hyperexcitability syndromes from other causes of muscle stiffness and cramp, such as the stiff-man syndrome148 and spinal rigidity caused by spinal interneuron loss.187,219 The differential diagnosis also includes the Schwartz-Jampel syndrome,198 in which continuous motor unit activity with similar high-frequency discharges occur,30,221 particularly in the axial and facial muscles, producing skeletal deformity and a characteristic facial appearance. Neuropathic Tremor A range of tremors may accompany neuropathy. Acquired and inherited axonal peripheral neuropathies, with chronic denervation and reinnervation, are often accompanied by a small-amplitude, rapid, postural upper limb tremor.190 This tremor is caused by co-contracting bursts of agonist and antagonist muscle activity and is attributed to weakness and muscle fatigue leading to synchronous motor unit discharge. The tremor has the characteristics of enhanced physiologic tremor. The voluntary recruitment of enlarged motor units following denervation and reinnervation may create a similar rapid postural or action tremor by amplifying the normal physiologic tremor, particularly during muscle fatigue. This phenomenon has been referred to as “contraction fasciculation”50 and “contraction pseudotremor of chronic denervation.”183 In chronic motor neuronopathies, such as spinal muscular atrophy, voluntary recruitment of “giant” reinnervated motor units gives a jerky or myoclonic quality to finger movement, described as minipolymyoclonus.212

Clinical Patterns of Peripheral Neuropathy

A conspicuous tremor is encountered in the setting of chronic demyelinating neuropathies. Tremor may be the presenting symptom of an immunoglobulin M (IgM) paraproteinemic neuropathy 207 or chronic relapsing demyelinating neuropathy,43 and is a prominent feature of the Roussy-Lévy variant of HMSN type I.200 This form of neuropathic tremor appears similar to severe, coarse essential tremor, with postural and kinetic components, alternating bursts of muscle activity in antagonist muscle pairs, and a frequency of 3 to 6 Hz. Tremor typically involves the upper limbs.13 Lower limbs are rarely affected, but may give the appearance of orthostatic tremor when standing.75 Peripheral sensation, in particular joint position sense, is usually preserved in the tremulous limbs (in contrast to deafferentation in sensory ataxia). This may explain the upper limb predominance of neuropathic tremor in IgM paraproteinemic neuropathy tremor, because joint position sense is frequently impaired in the lower limbs but preserved in the upper limbs.13,206 The tremor is unrelated to muscle weakness or fatigue. The frequency of neuropathic tremor may correlate with nerve conduction velocity in IgM paraproteinemic neuropathy,206 yet delay in the peripheral servo loop cannot be the sole explanation because this correlation is not evident in all cases.13 The explanation of neuropathic tremor remains uncertain. Despite clinical features that might suggest a central origin for the tremor in some cases, no CNS association has been uncovered apart from the ongoing discussion as to the possible existence of brainstem pathology in the Miller Fisher syndrome,51,125 A peripheral mechanism related to distortion of (peripheral) sensory input to the CNS, perhaps involving a mismatch of spindle and tendon organ afferents, has been postulated,2 leading to errors in the timing of muscle activity generated within the CNS.13,112 Myoclonus Myoclonus is a brisk involuntary movement or jerk caused by brief muscle contraction and is generally regarded as arising in the CNS. A number of examples of semirhythmic segmental myoclonus have been described occurring in the setting of focal peripheral nerve lesions. These include quadriceps myoclonus associated with sarcomatous infiltration of the femoral nerve,189 brachial myoclonus associated with a lesion of the posterior cord of the brachial plexus,17 myoclonus of the mid-thoracic paraspinal muscles associated with a thoracic radicular tumor,208 leg and buttock myoclonus in lumbosacral radiculopathy,111 leg myoclonus after meralgia paresthetica,223 shoulder myoclonus after accessory nerve section,84 forequarter myoclonus after thoracodorsal neuropathy,33 and myoclonus of one hand after a digital nerve lesion.138 In the latter, there was evidence of a peripheral generator for the myoclonus. In the remainder, the precise site of origin of the myoclonic discharges was not clear. As is discussed in the following section on painful legs and moving toes, the electromyographic characteristics of the myoclonus in most of these cases suggest additional involvement of the CNS.

1145

Painful Legs and Moving Toes and Variants This syndrome consists of slow, rhythmic sinuous toe movements, typically abduction-adduction and flexion-extension movements, and leg pain.210 The movements cannot be imitated but may be suppressed by voluntary effort or interrupted by voluntary movement of the affected limb.58,124,210 The most common antecedents are spinal nerve root injury,58,161 peripheral neuropathy,80,124,156,195 and peripheral limb trauma.58,161 The painful arm and moving fingers syndrome, following a brachial plexus radiation injury,245 and involuntary spasms of amputation stumps121,136 represent variations on the theme of involuntary movements following peripheral injury.196 Electromyographic analysis of the muscle activity accompanying the moving toes reveals long rhythmic bursts of normal motor units with normal recruitment patterns, suggesting a CNS origin.58 This is supported by the spread of movements and pain from one limb to another.58,210 Speculation about the origin of these movements has included reorganization within the CNS in response to altered sensory feedback after peripheral injury.58,124,156,161,195 However, in other cases, evidence has suggested the discharges arose within an abnormal segment of peripheral nerve.124,154,156,195 Restless Legs The restless legs syndrome is characterized by unpleasant “creeping, crawling” sensations felt deep within the legs, particularly between the knee and ankle, accompanied by an intense desire to move the legs and feet.66 Some patients describe additional pain or aching. The sensations typically begin a short time after lying in bed. The need to move the legs becomes overwhelming and irresistible, prompting the patient to rise from bed and pace around in an effort to obtain relief. Only movement relieves the symptoms. The continual need to move the legs interferes with sleep. Ekbom emphasized that these sensations differed from the cutaneous sensory disturbances of paresthesia, dysesthesia, hyperpathia, and formication that occur in sensory neuropathies.66 However, restless legs may be a feature of neuropathy,193 particularly uremic neuropathy.31 Many patients with restless legs have periodic movements of sleep and nocturnal myoclonus.37 This syndrome may be dominantly inherited.248 Treatment with dopamine agonists, clonazepam, and in some cases opiates affords relief and improves sleep.248 Hemifacial and Hemimasticatory Spasm Hemifacial spasm is idiopathic in most cases but occasionally is due to ectatic arteries or tumors of the facial nerve in the posterior fossa, and a small number of cases follow Bell’s palsy. The brief, repetitive unilateral facial movements are caused by high-frequency motor unit discharges that begin in the orbicularis oculi and spread to the orbicularis oris, frontalis, nasalis, and platysma. This

1146

Diseases of the Peripheral Nervous System

activity is derived from ectopic excitation of the facial nerve near the root entry zone (where the nerve is most excitable). Ephaptic cross-activation between adjacent nerve fibers at the site of injury leads to synkinetic activation of facial muscles.164,194 Synkinesis is evident on clinical examination and can be demonstrated after electrical stimulation of the facial nerve.164 Hemimasticatory spasm may occur in isolation235 or in association with facial hemiatrophy, a localized form of scleroderma.39,113 Ectopic activation of the trigeminal motor fibers produces brief repetitive spasms of the masticatory muscles, predominantly the masseter and temporalis. There is usually electrophysiologic evidence of a peripheral trigeminal nerve lesion, and the pathophysiology is thought to be similar to that of hemifacial spasm. Tendon Areflexia Loss of tendon reflexes is common in neuropathy. The tendon or muscle stretch reflex pathway involves activation of muscle spindles by muscle stretch, conduction of the ensuing spindle discharge in large-diameter Ia afferent fibers to the spinal cord, and an efferent limb from the anterior horn cell to muscle. The timing of reflex loss reflects the nature of the pathologic process. Reflex loss is an early feature in large fiber neuropathies, neuronopathies, and sensory ganglionopathies affecting spindle afferent fibers and appears late in the course of small fiber neuropathies such as amyloidosis. Accordingly, tendon areflexia most often reflects involvement of large-diameter sensory fibers. Other mechanisms include denervation of the intrafusal muscle fibers in motor neuropathies, temporal dispersion of conduction that degrades the synchronous volley of impulses in both the afferent and efferent limbs of the reflex arc, and the presence of continuous afferent impulses accompanying the motor unit activity in some cases of neuromyotonia.

Sensory Symptoms Normal somatosensory awareness of a body part requires afferent input from the natural stimulation of sensory receptors, consciousness, and attention. A loss of afferent input is recognized as sensory loss when adequate stimuli fail to evoke a subjective sensory response and the accompanying negative sensory symptoms of “numbness,” hypoesthesia, or hypoalgesia. In contrast, positive sensory symptoms or paresthesias (an abnormal sensation45) occur in the absence of receptor stimulation. The spectrum of positive sensory phenomena104 includes pain, allodynia (the experience of pain in response to a normally innocuous, usually tactile stimulus), dysesthesia (unpleasant sensations), hyperalgesia (exaggerated responses to painful stimuli), hyperesthesia, and hyperpathia (prolonged painful responses to repetitive stimuli, especially pinprick). Positive sensory phenomena therefore also represent exaggerated responses to sensory stimulation. For example,

sensations following non-noxious, natural mechanical or tactile stimulation may be perceived as painful, or the sensation of pain may be exaggerated. Non-noxious temperature elevation that normally evokes a sensation of warmth may induce pain in heat hyperalgesia,244 and the converse, cold hyperalgesia,128 may also occur. These changes in pain and temperature perception may be independent of any alteration in sensibility to touch.86 Positive and negative sensory symptoms may coexist. The earliest symptoms of a polyneuropathy are often sensory, affecting the distal feet and toes, with motor symptoms developing later. Sensory symptoms are encountered more commonly in acquired than inherited polyneuropathies. Initial sensory symptoms may comprise negative sensory symptoms of numbness and loss of sensation, or distal positive sensory phenomena such as paresthesia and dysesthesia.

Negative Sensory Symptoms and Signs Sensory loss in peripheral neuropathies results from conduction block or axonal loss. Loss of sensation may involve all sensory modalities, or the impairment may be restricted to particular forms of sensation. Two broad patterns of selective fiber loss have been recognized, small fiber and large fiber neuropathies. Small Fiber Sensory Loss Small fiber neuropathies lead to pain and temperature sensory loss, painless ulceration, injury, and neuropathic joint degeneration (Charcot joints). The selective dissociated loss of pain and temperature sensation is referred to as a “pseudosyringomyelic” pattern. Motor function and tendon reflexes are preserved. Cases of dissociated pain and temperature loss in the legs, often on a familial basis, have been recognized since the latter part of the 19th century. They were initially attributed to lumbosacral syringomyelia and other unconfirmed pathologies, but this interpretation was questioned by Thévenard.224 It is now clear that such cases are likely to have been examples of type I (dominant) hereditary sensory neuropathy.47 In 1937, Denny-Brown performed an autopsy on a member of a family originally reported by Hicks101 as having hereditary perforating ulcers of the feet. The autopsy revealed dorsal root ganglion cell degeneration. This was considered to be the primary pathologic lesion, and amyloid deposits that were also noted in the dorsal root ganglia were believed to be a secondary phenomenon. In a case of hereditary sensory neuropathy in which all modalities of sensation were affected, Dyck and co-workers showed that there was a diffuse loss of myelinated axons, particularly those of small diameter, and also a substantial loss of unmyelinated axons.63 Sluga, in an early case in which the sensory loss was limited to pain and temperature appreciation, found that the myelinated fiber depletion was limited to small-caliber fibers.204

Clinical Patterns of Peripheral Neuropathy

A further recognition of a neuropathy that produces dissociated pain and temperature loss beginning in the legs was provided when Andrade noted the existence of hereditary amyloid neuropathy.5 The pattern of fiber loss in this disorder also has been shown to involve a selective reduction in the population of small myelinated and unmyelinated axons.62,231 Another similar instance was described in two cases of Tangier disease in which selective loss of pain and temperature sensation characterized the earlier stages of the disease.118 A predominant loss of the smaller myelinated fibers and a gross loss of unmyelinated axons were demonstrated. Both cases had originally been diagnosed as syringomyelia. A pseudosyringomyelic pattern of temperature and pinprick sensory loss (over the limbs and anterior trunk) may be observed in some cases of diabetic neuropathy and has been related to a predominant loss of small myelinated and unmyelinated axons.26,191 Selective small fiber loss also occurs in Fabry’s disease.117 In recent years an autosomal recessive disorder has been recognized in which the affected individuals display congenital insensitivity to pain.56,64,130 Distal acrodystrophic changes develop. Nerve biopsies have shown preservation of unmyelinated fibers and large fiber densities with a preferential loss of small myelinated fibers. The loss of pain sensation in the setting of preserved unmyelinated axons is puzzling but might be explained if there were selective loss of the subclass of C-nociceptor axons or a neurotransmitter defect. The most notable instance of selective small fiber sensory loss is leprosy. The bacilli preferentially colonize Schwann cells associated with unmyelinated and small myelinated axons (see Chapter 91). The resultant loss of pain and temperature sensation that particularly affects the distal parts of limbs in lepromatous leprosy, in combination with preservation of motor function, often leads to mutilation, chronic ulceration, and loss of digits from repeated painless injuries. Large Fiber Sensory Loss In Friedreich’s ataxia, which involves a degeneration of the primary afferent neurons, joint position sense and vibration sense are lost and touch-pressure sensibility is impaired, while pain sensation and temperature sensation are preserved. Dyck and co-workers found that this was related to a selective disappearance of the larger myelinated fibers.63 Uremic neuropathy also is characterized by a predominant loss of larger myelinated fibers and a similar pattern of sensory loss.63,225,230 The loss of joint position sense that dominates the clinical picture of large fiber neuropathies typically manifests as a sensory ataxia (also referred to as an “ataxic neuropathy”) resulting from proprioceptive deafferentation.42 When the deafferentation affects the upper limbs and muscle strength is preserved, pseudoathetoid movements of the fingers may be evident when the hands are held outstretched and with the eyes closed. Tendon reflexes are absent. Severe deafferentation causes a significant deteri-

1147

oration in motor function, and the clinical presentation of large fiber neuropathies may be due to motor rather than sensory symptoms. This was illustrated in a patient with sensory neuropathy whose clinical presentation followed a decline in the performance of delicate skilled motor tasks such as the ability to do up shirt buttons and pick up small objects with the eyes closed.185 This patient was able to perform a wide range of preprogrammed movements with the distal upper limbs, yet his hands were relatively useless for everyday tasks. These difficulties were partly due to the absence of any automatic reflex correction of voluntary movements and also to an inability to sustain constant levels of muscle contraction without visual feedback for longer than 1 to 2 seconds (Fig. 45–6). Rapid movements were relatively unaffected. Similar phenomena may occur in the lower limbs. A sensory ataxia of gait is characterized by unsteadiness, an irregular stepping pattern, and increased

FIGURE 45–6 The effect of deafferentation (caused by sensory neuropathy) on motor performance. Five superimposed trials for a patient with a severe sensory neuropathy (right) and a normal subject (left) attempting to maintain a constant thumb position (top) or a constant thumb force (bottom). Initially the target trace and thumb trace were displayed on an oscilloscope. At the arrow, the target is jumped to a new position, shown by the horizontal bars in the center. Two seconds later the target and thumb traces were removed from view. The normal subject had little difficulty in maintaining a constant position or force, whereas the patient’s thumb starts to move randomly shortly after visual feedback is removed. (From Rothwell, J. C., Traub, M. M., Day, B. L., et al.: Manual performance in a deafferented man. Brain 105:515, 1982, with permission.)

1148

Diseases of the Peripheral Nervous System

body sway. Visual guidance is especially important during stepping. Upon eye closure, or when attempting to stand or walk in the dark, the body sway increases dramatically, causing the patient to fall. This is the basis of Romberg’s sign, distinguishing sensory from cerebellar gait ataxia. Sensory ataxia of gait also must be distinguished from the unsteadiness when standing in a patient with neuropathy caused by weakness of both the anterolateral lower leg and calf muscles. This pattern of weakness leads to difficulty in stabilizing the ankles. The unsteadiness is most noticeable when the affected individual attempts to stand still in one spot. It is less obtrusive when walking and, in contradistinction to sensory gait ataxia, is not worse in the dark or on eye closure. Sensory Ganglionopathy The clinical syndromes of a sensory ganglionopathy include sensory ataxia of upper and lower limbs caused by selective loss of proprioception, mimicking a large fiber neuropathy; a painful dysesthetic syndrome with burning paresthesias and lancinating pains as well as numbness and sensory loss; and widespread impairment of cutaneous (touch) sensation. Tendon reflexes are depressed or absent. Autonomic involvement may be evident. The hands or feet may be affected initially, followed by proximal spread of cutaneous sensory loss to involve the entire limb. Cutaneous sensory impairment may be patchy or extensive, involving whole limbs or parts of the trunk and face. The sensory ataxia is often dramatic and disabling. Paraneoplastic ganglionitis, Sjögren’s syndrome, drugs, and idiopathic causes are recognized.90,114,214,255

Positive Sensory Symptoms Paresthesias Paresthesias and other positive sensory symptoms are often uncomfortable and unpleasant. They are described in a variety of colorful ways. Sensations of creeping, crawling, prickling, tingling, or itching may be localized beneath or on the skin. The feet may feel abnormally warm, hot, or cold. Positive sensory symptoms may be described in terms of discomfort when walking. Sensations of roughness, sand, wool, eggshells, rolls of skin, crumpled socks, or pebbles are felt under the soles of the feet when walking. Wearing shoes may be uncomfortable, and the sensations may be relieved by wearing slippers or extra pairs of socks or avoiding shoes altogether. In large fiber neuropathies, a number of positive sensory symptoms may accompany paresthesias. The feet may feel as if they are wrapped in material or covered in water, the skin may feel tight, and there may be bandlike clamping sensations around the distal limbs. These symptoms may also be encountered in descriptions of the sensory phenomena accompanying myelopathies affecting the dorsal columns. The finding of tendon areflexia is an important clue to the neuropathic origin of these sensory complaints.

Paresthesias in small fiber neuropathies are typically accompanied by cutaneous hyperesthesia, dysesthesia, hyperpathia, and allodynia. Common descriptions of cutaneous dysesthesia include burning sensations as if the skin was on fire or sunburned. Dysesthesias are often worse at night. Such symptoms are typically encountered in acute painful diabetic neuropathy. Pain, both spontaneous and stimulus induced, may have a lancinating or shooting quality. While alterations in pain and temperature sensitivity may be found on examination, recognition of the above symptoms is important because they may appear in advance of any objective sensory deficit. Neuropathic Pain Pains evoked by afferent activity in peripheral nerves may have various subjective qualities. Brief lancinating pains in the distribution of a peripheral nerve are a common manifestation of neuropathic pain and may coexist with continuous pain. Microneurography has established that peripheral nerve fascicles are segregated into those conveying purely skin or muscle, or deep tissue afferents.240 Intraneural microstimulation of skin nerve fascicles elicits either dull burning (C-nociceptor induced) or sharp pricking (A␦-nociceptor induced) pain that is accurately projected onto the receptor site.176,236 In contrast, stimulation of muscle nerve fascicles evokes a cramplike pain that is projected diffusely over the muscle belly and sometimes to remote locations.135,170,173,237 Causalgia has been defined as a constant burning pain, accompanied by allodynia and hyperpathia, following partial nerve injury.152 The pain is experienced in the region innervated by the damaged nerve and may be exacerbated by touch, alterations in temperature, movement of the affected limb, and other stimuli, such as stress. The skin may be cool, red, and sweaty, and there may be atrophic changes affecting the skin, bones, joints, and muscle. These are largely the result of immobility. Mechanisms of Positive Sensory Phenomena and Painful Sensations Positive sensory phenomena such as tingling and pins and needles are the consequence of spontaneous ectopic discharges generated within primary sensory units. Paresthesias can be provoked experimentally by various means. Merrington and Nathan151 attributed the typical sequence of “buzzing, tingling, prickling and pseudocramp” in postischemic paresthesias to spontaneous nerve impulse generation in axons. In volunteers experiencing postischemic paresthesias, microneurographic recordings documented ectopic discharges in peripheral axons and confirmed their origin in nerve trunks rather than in receptors.175 In the postischemic period, natural stimulation of paresthetic skin may reinforce or revive symptoms if the evoked afferent impulses trigger further discharges when they propagate along hyperexcitable axons. A comparable phenomenon also

Clinical Patterns of Peripheral Neuropathy

occurs in neuropathy. Similarly, hyperventilation alters extracellular levels of calcium ions, increasing membrane excitability and enhancing paresthesias or even causing them in anxious individuals. In animal experiments, structures within the dorsal root ganglia are particularly prone to generate ectopic discharges, especially in response to mechanical pressure105 or to afferent impulse traffic in neighboring neurons.53 Electrical nerve impulses generated at sites other than peripheral receptor organs propagate bidirectionally, and antidromic impulses from relatively proximal sources can be detected in microneurographic recordings from peripheral nerve fascicles. This technique can record the ectopic impulses underlying paresthesias provoked by physical maneuvers impinging on hyperexcitable sensory nerves (as in the signs of Tinel, Lasègue, and Spurling)166,171 (Fig. 45–7). The biophysical basis for ectopic impulse generation is not completely understood. In general terms, it is believed to involve latent depolarization related to (1) facilitation or activation of sodium channels, (2) decreased inactivation of sodium channels, or (3) appropriate deviations in function of rectifying potassium channels.14,15,54 The unnatural subjective attributes of paresthesias may be explained by anomalous patterns of sensory afferent discharge, which create novel and unusual combinations of elementary sensations, in contrast to the predictable and stereotyped patterns of normal sensation. These unfamiliar combinations of basic sensations are perceived as abnormal sensations or paresthesias.177

1149

The positive sensory symptom of pain may be spontaneous or induced by natural stimulation of receptors. The afferent messages mediating neuropathic pain also may originate from hyperexcitable components of primary sensory units, as, for example, in the painful paresthetic signs of Tinel and Spurling. Pain as a symptom of peripheral neuropathy may also be due to pathologic excitation of nociceptor afferent supply to the nerve trunks themselves (nervi nervorum).7,226 Such nerve trunk pain is relatively uncommon compared with pain from diseased nerve trunks. Primary injury to peripheral nerve fibers may be complicated by secondary changes in the CNS that may be capable of causing pain.52 Pain is evoked by activity within A- or C-primary nociceptor afferent units, so it is likely that neuropathic pain is also related to spontaneous nociceptor discharge, as reported in animal models of traumatic and metabolic neuropathy,22,29 but this remains to be documented in humans. While this theory is attractive, it is also possible that secondary changes within the CNS may give rise to pain following lesions of primary sensory units. The following mechanisms have been proposed to explain the exaggerated perception of a natural stimulus or the transformation of natural stimulation into a painful sensation: 1. Sensitization of peripheral nerve endings causing an abnormally low receptor threshold has been documented as a basis for chronic spontaneous pain and mechanical and thermal hyperalgesia in human

FIGURE 45–7 Microneurographic (antidromic) recording from the left ulnar nerve at wrist level during elicitation of Spurling’s sign by positioning the neck to pinch the entrapped nerve root in a patient with a C8 radiculopathy. A, Integrated neurogram. B, Simultaneous discriminated neurogram. Symbols: a ⫽ stimulation of receptive field in little finger; solid arrow ⫽ head flexed to left and extended; b ⫽ onset of abnormal unitary bursting, temporally correlated with the verbal report of paresthesias reaching the hand; open arrow ⫽ head returned to neutral position; “emg” ⫽ voluntary muscle artifact; a ⫽ original receptive field confirmed by natural stimulation of skin. (From Ochoa, J., Cline, M., Dotson, R., and Marchettini, P.: Pain and paresthesias provoked mechanically in human cervical root entrapment [sign of Spurling]: single sensory unit antidromic recording of ectopic, bursting, propagated nerve impulse activity. In Pubols, L., and Sessle, B. [eds.]: Effects of Injury on Trigeminal and Spinal Somatosensory Systems. New York, Alan R Liss, p. 389, 1987, with permission.)

1150

Diseases of the Peripheral Nervous System

neuropathy.35,36 Sensitization of C nociceptors probably underlies erythromelalgia or the ABC syndrome,169 which may accompany mononeuropathy, radiculopathy, or polyneuropathy. It is characterized by red, warm skin; spontaneous burning pain; and heat hyperalgesia that is exacerbated by heat and relieved by cooling. These phenomena are identical to the effects of the acute action of capsaicin, a C-nociceptor neurotoxin.40,110,133 2. Ephaptic transmission, or electrical cross-talk between axons, may explain hyperalgesia and has been a fashionable explanation for causalgia,57,87,160,164,180 although it remains unproven. 3. Plasticity within CNS sensory neurons following primary peripheral nervous system injury was proposed to explain the failure of peripheral ablative therapies to relieve pain caused by nerve lesions.32,129,165,168 This is supported, though not proven, by changes in the neurophysiologic behavior of central neurons following peripheral injury.59,254 4. Disinhibition of central pain mechanisms, by an imbalance in sensory inputs,100 or interference with the modulation, or gating, of primary afferent input within the spinal cord, or descending brainstem inhibition, has been postulated.150 However, there is no resting impulse activity in human nociceptors, so central disinhibition is unlikely to release spontaneous pain. In addition to these mechanisms, it is also possible that changes within the CNS, following lesions of the primary sensory units, may give rise to pain.

Autonomic Involvement The symptoms arising from disturbances of the parasympathetic components of the cranial nerves (III, VII, IX, X, and XI) are discussed in the relevant chapters in this text. This section is limited to the effects of disturbances of the sympathetic system and the sacral parasympathetic outflow that may occur in peripheral nerve disorders. Horner’s syndrome may develop from localized lesions affecting the first thoracic root or the cervical sympathetic chain, but ocular innervation is not often involved in generalized neuropathies apart from diabetic and amyloid polyneuropathy. Disturbances of pupillary function have been most extensively evaluated in diabetic neuropathy. The salient changes are a reduction of pupil size in darkness, a reduction in the ability of the pupil to maintain miosis in continuous light, and a reduction in pupillary unrest (hippus) in both darkness and light.106 Denervated pupils are also a feature of the trigeminal and ataxic neuropathy that may be associated with the sicca syndrome.115 Cardiovascular symptoms have been studied most closely in diabetic neuropathy, but orthostatic hypotension can be the presenting symptom in amyloid neuropathy and may be of devastating severity. Estimates as to its frequency

in diabetic subjects have varied considerably. In a group of 73 diabetic persons with symptoms suggestive of autonomic neuropathy, 25 had symptoms of orthostatic hypotension, among whom 23 had falls in systolic blood pressure of 30 mm Hg or more.71 The main explanation is sympathetic splanchnic denervation.102 Both sympathetic and parasympathetic cardiac denervation can be demonstrated in diabetic autonomic neuropathy, but neither is important symptomatically apart possibly from a cardiac explanation for the sudden unexpected death that may occur in such individuals. Evidence is accumulating that such deaths may be related to the development of a prolonged QT syndrome as a result of an imbalance between the right and left cardiac sympathetic supply.70 Defects of vascular function have also been demonstrated in porphyria; alcoholic neuropathy; the GuillainBarré syndrome and acute pandysautonomia; neuropathy related to carcinoma, lymphoma, and myeloma; neuropathy resulting from thallium, vincristine, or perhexiline toxicity; and the Riley-Day syndrome.44,143 Orthostatic hypotension commonly occurs, but hypertension occasionally develops in association with porphyria, thallium intoxication,16,182 and the Guillain-Barré syndrome, when it may be the result of carotid sinus and aortic arch baroreceptor denervation. Anhidrosis related to deficient sudomotor innervation may occur in the cutaneous territory of a peripheral nerve in mononeuropathies. In polyneuropathies that affect the autonomic system, it displays a symmetrical distribution and initially affects the lower legs. It is encountered particularly in diabetic subjects and, when extensive, may lead to heat intolerance and excessive sweating over the upper parts of the body.20,85,137,186 It is also seen in amyloid neuropathy and is a feature of some hereditary sensory and autonomic neuropathies. The anhidrosis is the result of a postganglionic lesion of the sympathetic sudomotor nerve fibers and is often accompanied by defective piloerection.18 Attacks of excessive sweating occur in the Riley-Day syndrome, and troublesome gustatory sweating over the face and head may be a feature in diabetic autonomic neuropathy.250 Localized facial gustatory sweating may also occur following injury to the auriculotemporal nerve. Familial gustatory sweating without other signs of a neuropathy is also recognized. Lack of awareness of hypoglycemia in cases of diabetic autonomic neuropathy, related to failure of catecholamine release, is considered in Chapter 120. Disturbances of genitourinary function are most often encountered in diabetic and amyloid neuropathy. With respect to visceral function, a dilated atonic bladder develops. Initially the patient may notice that the intervals between voiding increase; that voiding is difficult, with an intermittent stream; and that postmicturition dribbling occurs. Retention with overflow incontinence follows, usually first noticed at night. Cystometry demonstrates a large atonic bladder

1151

Clinical Patterns of Peripheral Neuropathy

with a flat pressure curve on filling until the bladder capacity is achieved. There is a lack of awareness of bladder filling, and these changes result primarily from sensory denervation of the bladder wall. When impotence develops, failure of erection resulting from interference with the parasympathetic innervation usually precedes failure of ejaculation. An interesting symptom sometimes encountered in diabetic subjects is retrograde ejaculation.68,88 Despite normal erection and orgasm, seminal emission does not occur, and the semen is subsequently found in the urine. This is assumed to be due to a failure of the normal closure of the internal sphincter of the bladder neck during ejaculation, although both this sphincter and the seminal vesicles are innervated by sympathetic fibers. Symptoms related to disturbances of the innervation of the upper alimentary tract are considered in Chapter 122. Disturbances of large bowel function have been described in cases of autonomic neuropathy (e.g., the occurrence of colonic dilatation in diabetic autonomic neuropathy).19,178 Hirschsprung’s disease (congenital megacolon) is related to abnormalities of the myenteric plexus. The contracted segment contains a network of abnormal unmyelinated nerve fibers but is devoid of ganglion cells, and those of the dilated segment are morphologically abnormal.205 Megacolon is also a feature of infection with Trypanosoma cruzi (Chagas’ disease) and results from destruction of the ganglion cells in the colonic myenteric plexus.205

Skeletal Deformity Chronic neuropathies developing before cessation of the growth period frequently give rise to foot, hand, and spinal deformity. A claw hand and clawing of the toes may develop in a chronic neuropathy that begins at any age, but the presence of pes cavus or a spinal deformity indicates an onset in childhood, a point that is sometimes of diagnostic importance. In the feet, a cavus deformity with clawing of the toes is the most frequent abnormality, and in more severe instances is associated with a fixed equinovarus position. More rarely, pes planus results. Kyphoscoliosis is the usual spinal deformity. The foot deformity can in most cases be attributed to muscle weakness. Foot deformity may be prominent in early-onset distal spinal muscular atrophy97 and type I HMSN.96 In the latter, muscle weakness in the legs is greatest in the anterolateral muscle group on the lower leg and the intrinsic foot muscles. The equinovarus position of the foot, the cavus deformity, and the clawing of the toes seen in these conditions can be attributed to preservation of activity in the long flexors of the foot and the ankle invertors.

Trophic Changes The term trophic changes is in many ways unsatisfactory, but it has been hallowed by long usage. It refers to

alterations that follow denervation, but it is evident that the causes are often complex and multiple. The concept of specific trophic nerve fibers involved in the maintenance of the structural integrity of tissues and organs cannot be upheld. Yet it is becoming increasingly evident that, in certain instances, the normal innervation has a crucial part to play in tissue differentiation and in the subsequent maintenance of normal structure and activity. The nerve fibers innervating skeletal muscle have been shown to be involved in determining many of the structural and functional properties of muscle fibers.28 Moreover, loss of muscle innervation leads to denervation atrophy of the fibers, although the effects of trauma may be important in causing the degeneration and ultimate disappearance of denervated muscle fibers.25 Taste buds are known to undergo atrophic changes following loss of their sensory innervation.256 Lewis and Pickering held that alterations in the skin, nails, and subcutaneous tissues that follow denervation, instead of resulting from the loss of a specific trophic function by nerve fibers, were the effects of disuse, alterations in blood supply, and sensory loss.126 No convincing evidence has since been adduced to alter this view. The effects of denervation following peripheral nerve injury have been documented and critically appraised by Sunderland.218 Apart from their occurrence in diabetic neuropathy, trophic changes consequent upon peripheral neuropathy are most often encountered in leprosy, hereditary sensory neuropathy,47 HMSN with prominent sensory loss,69,228 and hereditary amyloid neuropathy.5 The changes are usually most evident at the periphery, which led Spillane and Wells211 to employ the term acrodystrophic neuropathy to describe cases of sensory and sensorimotor neuropathy of mixed etiology in which there was distal trophic ulceration. Bony changes also tend to begin distally. The radiographic appearances include generalized or focal loss of bone density, thinning of the phalanges, pathologic fractures, and the development of neuropathic arthropathy. Osteomyelitis may be superimposed. It is generally accepted that the factor responsible for the appearance of trophic ulcers and neuropathic joint degeneration is loss of pain sensation, which exposes the tissues to the risk of repeated trauma. It has been suggested that loss of other joint and muscle afferent impulses may be important, but this is as yet unsubstantiated.211 In diabetic neuropathy, ischemia and the liability of diabetic tissues to infection also may be significant.

Nerve Thickening Palpable or visible enlargement of nerve trunks is encountered in a variety of peripheral neuropathies, and its detection may be helpful in diagnosis. Yet assessment of nerve thickening is often difficult. Nerves of normal size may be readily visible in thin subjects, particularly when the overlying skin is stretched, and they may then mistakenly

1152

Diseases of the Peripheral Nervous System

be considered enlarged. This is especially likely in the case of the greater auricular nerve in the neck and the superficial peroneal nerve in front of the ankle or on the dorsum of the foot. In the lepromatous, tuberculoid, and borderline forms of leprous neuropathy, enlargement of nerves is often a conspicuous feature. In lepromatous neuropathy, in which peripheral nerve symptoms occur late in relation to the appearance of cutaneous lesions, there tends to be a firm, smooth enlargement limited to the most superficial nerve trunks. This is most easily detectable in the greater auricular nerve in the neck, the supraclavicular nerves as they cross the clavicles, the ulnar nerve just above the elbow, the radial nerve at the wrist, the common peroneal nerve at the neck of the fibula, and the superficial peroneal nerve over the anterior aspects of the ankles and the dorsa of the feet. The infrapatellar branch of the saphenous nerve on the medial aspect of the knee is sometimes noticeably enlarged before there is detectable thickening elsewhere. In tuberculoid leprosy, the enlargement may be uniform or nodular. Rarely, a localized cystic swelling develops along the nerve, at times reaching substantial dimensions, related to caseous abscess formation. Nerves may be locally enlarged in the vicinity of cutaneous lesions in tuberculoid leprosy. In patients who present with localized peripheral nerve lesions, such as ulnar palsy, enlargement of the affected nerve may be evident or thickening of other nerves may be demonstrable. A diagnostic point that may be helpful in the differentiation from a simple entrapment neuropathy of the ulnar nerve at the elbow is that, in leprosy, the enlargement usually extends for a greater distance up the arm or may be maximal some distance proximal to the elbow. In a simple entrapment neuropathy, the enlargement is usually restricted to a few centimeters in and above the ulnar groove. In borderline leprosy, widespread enlargement of nerves is also present before any neurologic symptoms or cutaneous lesions appear. The term hypertrophic neuropathy is used to describe nerve thickening related to concentric Schwann cell proliferation in chronic demyelinating neuropathies. The onion bulb formations that characterize this process are among the most striking types of change observed in nerve biopsies. Among the inherited disorders, diffuse nerve thickening related to this type of pathology is observed in types I and II HMSN and Refsum’s disease. In acquired demyelinating neuropathies, it may be a feature of chronic relapsing and chronic progressive inflammatory polyneuropathy.12 Sometimes it is extraordinarily localized, as in the focal hypertrophic neuritis that may affect the brachial plexus41,252 and limb nerves.154 Multifocal cases are on record.1,9 Although peripheral nerve enlargement in acromegaly has misleadingly been referred to as hypertrophic neuropathy,215 it is important to realize that it is the result of connective tissue overgrowth. Concentric Schwann cell proliferation, which is the hallmark of both

the inherited and acquired cases of hypertrophic neuropathy, does not occur to any significant degree. Nerve thickening in amyloid neuropathy is related to the presence of amyloid deposits within the nerve. In neurofibromatosis, apart from the more usual localized swelling subcutaneously or in relation to nerve trunks, diffuse peripheral nerve enlargement may occur.232 Localized nerve enlargement may also indicate the presence of a schwannoma or other solitary peripheral nerve tumor. Distinguishing intraneural tumor, localized hypertrophic neuropathy, and pseudoneuroma caused by entrapment may not be easy.174

DIFFERENTIAL DIAGNOSIS Disorders Simulating Peripheral Neuropathy Primary diseases of muscle do not commonly give rise to difficulty. In cases of doubt, electromyography is usually decisive in their differentiation. When the onset is gradual, the proximal distribution of muscle involvement is symmetrical and in general uniform, unlike that in most neuropathies, but may closely mimic a proximal spinal muscular atrophy. Acute polymyositis, if there is no muscle tenderness or involvement of other systems, could be mistaken for an acute Guillain-Barré syndrome with proximal weakness. Inclusion body myositis and rare cases of polymyositis display a distal distribution of weakness and wasting in the limbs and clinically may resemble a chronic motor polyneuropathy or chronic distal spinal muscular atrophy.103 Of the muscular dystrophies, the rare distal myopathy of Welander253 gives rise to a somewhat similar clinical picture but, in contradistinction to most neuropathies, usually commences in the upper limbs. Myotonic dystrophy also affects the distal musculature in the limbs, and detectable myotonia may be absent. However, selective atrophy of the sternocleidomastoid is likely to coexist, and other features of the disorder may be evident. Myasthenia gravis is not usually confused with peripheral neuropathy, except for occasional cases with acute onset in childhood and accompanied by proximal weakness in the limbs and little bulbar involvement. These may superficially resemble the Guillain-Barré syndrome. The Lambert-Eaton syndrome is more likely to cause diagnostic difficulty, especially since distal paresthesias may occur.122 The weakness is usually marked in the limbs, and ocular, facial, or bulbar weakness may not be evident. There may be generalized depression of the tendon reflexes. It is of importance that tendon reflexes often can be enhanced following vigorous contraction of the muscle under examination. Confirmation of diagnosis may be obtained by electrophysiologic demonstration of the postcontraction facilitation that underlies this phenomenon.

Clinical Patterns of Peripheral Neuropathy

Motor neuronopathies (spinal muscular atrophies) are difficult to differentiate from motor axonopathies by clinical or electrophysiologic means. Muscle fasciculation is more prominent in anterior horn cell degenerations, particularly in cases of amyotrophic lateral sclerosis. The diagnostic problem arises in occasional cases of amyotrophic lateral sclerosis of progressive muscular atrophy type, particularly those that begin with focal muscular atrophy or bilateral footdrop and depression of the ankle jerks. Similar difficulty arises in cases of hereditary distal spinal muscular atrophy.97 Acute anterior poliomyelitis gave rise to occasional diagnostic confusion at the time of epidemics and still may do so in unimmunized individuals who have returned from visiting endemic areas. The acute onset, total absence of sensory loss, constitutional upset with fever, and cerebrospinal fluid pleocytosis should indicate the correct diagnosis. Confirmation may be sought by virologic studies. Tabes dorsalis may present clinical resemblances to a sensory neuropathy, in particular that related to diabetes. In both, the tendon reflexes are depressed or absent, and there may be sensory ataxia of gait. Loss of pain sensibility may result in perforating ulcers in the feet and neuropathic arthropathy, although the distribution of joint involvement differs in the two disorders. Aching or lancinating pains in the legs occur in both, as may autonomic disturbances, including bladder atony, impotence, and pupillary disorders. Differentiation between tabes dorsalis and a sensory neuropathy is not difficult. Impairment of tactile sensation is not a feature of tabes, and the pattern of sensory loss is often characteristic. Only minor abnormalities of the sensory nerve action potentials are detectable, whereas in a sensory neuropathy of equivalent clinical severity, marked changes would be expected. Confirmation may be obtained by serologic test for syphilis. Spinal cord disorders sometimes mimic a sensory neuropathy by giving rise to distal paresthesias in the legs or in all four extremities. This situation may arise in multiple sclerosis, in cervical myelopathy from spondylosis, or occasionally from extramedullary tumors of the cervical region. If such symptoms develop before signs of long tract damage appear, diagnostic difficulties may arise. Studies of sensory nerve conduction usually resolve matters. Inquiry should be made for Lhermitte’s sign, which, if present, indicates cervical cord pathology. It may occur in subacute combined degeneration as a result of B12 deficiency or following megadosage of pyridoxine. It has been established that, in many dying-back neuropathies, there is combined CNS and peripheral nervous system involvement. Thus there is a distal degeneration of both the centrally and the peripherally directed axons of the dorsal root ganglion cells, constituting a centralperipheral distal axonopathy.209 Disorders with a selective distal degeneration of the centrally directed axons of the primary sensory neurons (together with degeneration of CNS long tracts) have been recognized.227 The sensory

1153

manifestations mimic a peripheral neuropathy. This pattern of involvement characterizes clioquinol neurotoxicity202 and is encountered in the early stages of the neurologic syndromes related to vitamin B1272 and vitamin E deficiencies.95 Cerebral lesions rarely present diagnostic difficulty. As originally pointed out by Foerster, it is of interest that cerebral tumors occasionally manifest with sensory or motor symptoms of a radicular or peripheral nerve distribution. The explanation is possibly to be found in the summation of a preexisting subclinical nerve or radicular lesion with the developing cerebral lesion.229 Hysterical symptoms should be readily identified as such. The inconstant and fluctuating character of hysterical weakness, its improvement by encouragement, and the lack of reflex change are important clues. A glove-and-stocking distribution is a favored pattern of hysterical sensory loss, and this may simulate a peripheral neuropathy, but inconsistencies revealed by testing and physiologically inexplicable responses to some sensory modalities, combined with the preservation of sensory nerve action potentials, will indicate the nature of the disturbance. The occurrence of tingling paresthesias in the extremities is a not infrequent reason for patients with anxiety and a suspected diagnosis of a sensory neuropathy to be referred to a neurologist. Although hyperventilation is not always noticed or admitted by the patient, the evanescent nature of the symptoms and the preservation of sensory nerve action potentials will help exclude structural changes in the nerves. A cautionary note has to be inserted that routine nerve conduction studies examine only the larger myelinated nerve fibers and may not detect a selective small fiber neuropathy. Nerve biopsy may be required in such circumstances.

Neuropathies Simulating Other Conditions In the preceding section, the conditions most commonly mistaken for peripheral neuropathies have been considered. Conversely, certain neuropathies are likely to masquerade as other disorders. The instance of diabetic pseudotabes has already been cited, as has that of acute neuropathies with proximal muscle weakness that can be mistaken for acute polymyositis. The coexistence of peripheral neuropathy and polymyositis can lead to a diagnostic challenge.157 A proximal distribution of involvement may occur in the Guillain-Barré syndrome and porphyric neuropathy, in either of which sensory changes may be unobtrusive or absent. In neuropathies that present with cranial nerve involvement, particularly if the onset is insidious, a diagnosis of myasthenia gravis may be entertained.233 Interesting but rare examples of a neuropathy that may mimic spinal cord disease are those cases of Tangier disease that present with muscle wasting in the upper limbs, loss of tendon reflexes, and dissociated pain and temperature sensory loss with a central distribution.

1154

Diseases of the Peripheral Nervous System

The ascription of cases of hereditary sensory neuropathy to lumbosacral syringomyelia, which was alluded to earlier, has now been abandoned. Finally, the monotonous regularity with which the weakness of porphyric neuropathy is attributed to hysteria is emphasized in Chapter 80.

DIAGNOSIS OF NEUROPATHY

In patients in whom there may be difficulty in distinguishing between a myopathic process and a motor neuropathy, electromyography is usually decisive. An elevated serum creatine kinase level is more likely to indicate a myopathy, although slight or moderate elevation may occur in denervation atrophy. Increase of the cerebrospinal fluid protein content also may point to a neuropathic condition, and the degree of increase can sometimes be of assistance in defining the nature of the disorder.

Confirmation of Neuropathy Confirmation of the existence of peripheral neuropathy when this has been suspected on clinical grounds is usually best achieved by nerve conduction studies. These are useful both in detecting localized peripheral nerve lesions and in establishing the presence of generalized polyneuropathies. In broad terms, generalized polyneuropathies can be subdivided into those in which the primary disturbance is axonal degeneration and those in which it involves widespread segmental demyelination. The group in which demyelination is prominent is characterized by a substantial reduction in nerve conduction velocity, whereas in those in which the changes primarily affect axons, conduction velocity tends to be within normal limits or only slightly reduced. In the latter, the examination of motor nerve conduction velocity may not be helpful in distinguishing axonopathies from anterior horn cell disease, in which nerve conduction velocity is similarly affected. In most neuropathies, however, sensory nerve conduction, even if sensation is clinically normal, is likely to be abnormal. This statement requires the qualification that, in lesions affecting the sensory roots proximal to the dorsal root ganglion cells, conduction in sensory fibers in the limbs is preserved even if there is total sensory loss. In brachial plexus injuries, this may be useful in distinguishing between avulsion of roots from the cord and lesions in the plexus distal to the dorsal root ganglia.24 In such cases, axon responses, demonstrated by the intradermal injection of histamine, for example, will be preserved.23 In cases of Guillain-Barré polyradiculopathy in which the demyelinating process is confined to the roots and in which axonal destruction has not occurred, sensory (and motor) conduction will be normal in the limbs. The measurement of F-wave latencies and estimation of proximal motor conduction from magnetic or high-voltage electrical stimulation over the spinal column may be informative. The quantitative assessment of the structural integrity of peripheral nerves is possible by nerve biopsy, but the use of a biopsy to substantiate a diagnosis of neuropathy is rarely necessary, except in rare examples of small fiber neuropathies in which routine nerve conduction studies are normal. The procedure may be helpful in some instances in defining the nature of the disorder, particularly if vasculitis, connective tissue infiltration, leprosy, tomaculous neuropathy, or storage disease is suspected.

Establishing Etiology The clinical features of the different types of peripheral neuropathy are detailed in succeeding chapters. At this juncture, it may be appropriate to outline a few general aspects related to establishing the cause of a peripheral neuropathy. The analysis of the clinical history and of the abnormalities revealed by physical examination may either suggest the diagnosis or narrow down the diagnostic possibilities, thus facilitating further investigation. Nerve conduction studies can provide crucial information and are best performed at the initial examination. Alternatively, if the patient is referred for electrophysiologic investigations, the clinical problem should be formulated precisely so that the appropriate electrical studies are performed. In general, the first broad determination that is made is whether the patient has a symmetrical polyneuropathy or a focal or multifocal neuropathy. As discussed earlier, the range of diagnostic possibilities differs for symmetrical polyneuropathy and multifocal neuropathy. Asbury and Gilliatt8 proposed a diagnostic algorithm that splits cases into these two categories and then subsequently into axonopathies and demyelinating neuropathies on the basis of nerve conduction studies. Unfortunately, this distinction is not always clear-cut, and the number of cases that will be separated off with a demyelinating neuropathy is relatively small. For those patients with a symmetrical polyneuropathy, the clinical pattern (i.e., whether it is motor, sensory, or mixed, or whether it has a proximal or distal distribution) can provide useful information, as may the rapidity of onset and, in established cases, the previous clinical course. The age of the patient is important. A neuropathy of insidious onset in childhood is likely to be of genetic origin, whereas in later life a paraneoplastic or paraproteinemic basis figures high in the list of diagnostic possibilities. The patient’s country of origin may provide useful information. The presence of an isolated ulnar neuropathy in an individual who was born in an area where leprosy is endemic should alert the clinician to this possibility and lead to a careful search for thickening of other nerves and for cutaneous changes. It needs no emphasis that the presence of abnormalities in other systems is often of crucial diagnostic value. Estimates of the frequency with which an etiologic diagnosis was achieved in cases of peripheral neuropathy

Clinical Patterns of Peripheral Neuropathy

that were investigated 40 years ago yielded the gloomy conclusion that in a high proportion the explanation remained uncertain.67,139 Since then a substantial amount of activity has been directed toward the investigation of peripheral nerve disorders. Yet it remains embarrassingly true, particularly for those cases with a chronic progressive course, that no satisfactory explanation for their occurrence can be given. However, Dyck and associates, in a study published in 1981, showed that intensive investigation of such cases is rewarding.65 In a series of undiagnosed cases referred to the Mayo Clinic, it was possible to provide an etiologic classification in 76%. An important aspect was the evaluation of the patient’s relatives, because this led to a diagnosis of HMSN type II in a significant number of individuals who might otherwise have remained with a diagnosis of cryptogenic axonal neuropathy of late onset. In a subsequent paper, McLeod and co-workers showed that, with intensive investigation and follow-up, the proportion of neuropathies (seen at tertiary referral centers) that elude diagnosis can be reduced to as little as 13%.144

REFERENCES 1. Adams, R. D., Asbury, A. K., and Michelsen, J. J.: Multifocal pseudohypertrophic neuropathy. Trans. Am. Neurol. Assoc. 90:30, 1965. 2. Adams, R. D., Shahani, B. T., and Young, R. R.: Tremor in association with polyneuropathy. Trans. Am. Neurol. Assoc. 97:44, 1972. 3. Aguayo, A. J., Nair, C. P. V., and Bray, G. M.: Peripheral nerve abnormalities in the Riley-Day syndrome. Arch. Neurol. 24:106, 1971. 4. Albers, J. W., Allan, A. A., and Bastron, J. A.: Limb myokymia. Muscle Nerve 4:494, 1981. 5. Andrade, C.: A peculiar form of peripheral neuropathy: familial atypical generalized amyloidosis with special involvement of peripheral nerves. Brain 75:408, 1952. 6. Asbury, A. K., Aldredge, H., Hershberg, R., and Fisher, C. M.: Oculomotor palsy in diabetes mellitus: a clinico-pathological study. Brain 93:555, 1970. 7. Asbury, A. K., and Fields, H. L.: Pain due to peripheral nerve damage: an hypothesis. Neurology 34:1587, 1984. 8. Asbury, A. K., and Gilliatt, R. W.: The clinical approach to neuropathy. In Asbury, A. K., and Gilliatt, R. W. (eds): Peripheral Nerve Disorders. London, Butterworths, p. 1, 1984. 9. Asbury, A. K., and Johnson, P. C.: Pathology of Peripheral Nerve. Philadelphia, WB Saunders, 1978. 10. Ashizawa, T., Butler, I. J., Harati, Y., and Roongta, S. M.: A dominantly inherited syndrome with continuous motor neurone discharges. Ann. Neurol. 13:285, 1983. 11. Auger, R. G., Daube, J. R., Gomez, M. R., and Lambert, E. H.: Hereditary form of sustained muscle activity of peripheral nerve origin causing generalized myokymia and muscle stiffness. Ann. Neurol. 15:13, 1984. 12. Austin, J. H.: Recurrent polyneuropathies and their corticosteroid treatment. Brain 81:157, 1958.

1155

13. Bain, P. G., Britton, T. C., Jenkins, I. H., et al.: Tremor associated with benign IgM paraproteinaemic neuropathy. Brain 119:789, 1996. 14. Baker, M., and Bostock, H.: Depolarization changes the mechanism of accommodation in rat and human motor axons. J. Physiol. (Lond.) 411:545, 1989. 15. Baker, M., Bostock, H., Grafe, P., and Martius, P.: Function and distribution of three types of rectifying channel in rat spinal root myelinated axons. J. Physiol. (Lond.) 383:45, 1987. 16. Bank, W. J., Pleasure, D. E., Suzuki, K., et al.: Thallium poisoning. Arch. Neurol. 26:256, 1972. 17. Banks, G., Nielsen, V. K., Short, M. P., and Kowal, C. D.: Brachial plexus myoclonus. J. Neurol. Neurosurg. Psychiatry 48:582, 1985. 18. Bárány, F. R., and Cooper, E. H.: Pilomotor and sudomotor innervation in diabetes. Clin. Sci. 15:533, 1956. 19. Berenyi, M. R., and Schwarz, G. S.: Megasigmoid syndrome in diabetes and neurologic disease. Am. J. Gastroenterol. 47:311, 1967. 20. Berge, K. G., Wollaeger, E. E., Scholz, D. A., et al.: Steatorrhoea complicating diabetes mellitus with neuropathy: report of two cases without apparent external pancreatic insufficiency. Diabetes 5:25, 1956. 21. Black, J. T., Garcia-Mullin, R., Good, E., and Brown, S.: Muscle rigidity in a newborn due to continuous peripheral nerve hyperactivity. Arch. Neurol. 27:413, 1972. 22. Blumberg, H., and Jänig, W.: Discharge pattern of afferent fibers from a neuroma. Pain 20:335, 1984. 23. Bonney, G.: Value of axon responses in determining site of lesion in traction injuries of brachial plexus. Brain 77:588, 1954. 24. Bonney, G., and Gilliatt, R. W.: Sensory nerve conduction after traction lesions of the brachial plexus. Proc. R. Soc. Med. 51:365, 1958. 25. Bowden, R. E. M., and Gutmann, E.: Denervation and reinnervation of human voluntary muscle. Brain 67:273, 1944. 26. Brown, M., Martin, J. R., and Asbury, A. K.: Painful diabetic neuropathy: a morphometric study. Arch. Neurol. 33:164, 1976. 27. Brunt, E. R. P., and van Weerden, T. W.: Familial paroxysmal kinesigenic ataxia and continuous myokymia. Brain 113:1361, 1990. 28. Buller, A. J., Eccles, J. C., and Eccles, R. M.: Interactions between motoneurones and muscles in respect of the characteristic speeds of their responses. J. Physiol. (Lond.) 150:417, 1960. 29. Burchiel, K. J., Russell, L. C., Lee, R. P., and Sima, A. A. F.: Spontaneous activity of primary afferent neurons in diabetic BB/Wistar rats. Diabetes 34:1210, 1985. 30. Cadilhac, J., Baldet, D., Greze, J. A., and Duday, H.: EMG studies of two family cases of the Schwartz and Jampel syndrome (osteo-chondro-muscular dystrophy with myotonia). Electromyogr. Clin. Neurophysiol. 15:5, 1975. 31. Callaghan, N.: Restless legs syndrome in uremic neuropathy. Neurology 16:359, 1966. 32. Campbell, J. N., Raja, S. N., Meyer, R. A., and MacKinnon, S.: Myelinated afferents signal the hyperalgesia associated with nerve injury. Pain 32:89, 1988.

1156

Diseases of the Peripheral Nervous System

33. Carnero-Pardo, C., Sanchez-Alvarez, J. C., GomezCamello, A., et al.: Myoclonus related to distal thoracodorsal neuropathy. Mov. Disord. 13:971, 1998. 34. Chopra, J. S., Khanna, S. K., and Murthy, J. M. K.: Congenital arteriovenous fistula producing carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 42:815, 1979. 35. Cline, M., and Ochoa, J.: Chronically sensitized C nociceptors in skin: patient with hyperalgesia, hyperpathia, and spontaneous pain. Soc. Neurosci. Abstr. 12:331, 1986. 36. Cline, M., Ochoa, J., and Torebjörk, H.-E.: Chronic hyperalgesia and skin warming caused by sensitized C nociceptors. Brain 112:621, 1989. 37. Coleman, R. M., Pollak, C. P., and Weitzmann, E. D.: Periodic movements in sleep (nocturnal myoclonus): relation to sleep disorders. Ann. Neurol. 8:416, 1980. 38. Cragg, B. G., and Thomas, P. K.: Changes in nerve conduction in experimental allergic neuritis. J. Neurol. Neurosurg. Psychiatry 27:106, 1964. 39. Cruccu, G., Inghilleri, M., Berardelli, A., et al.: Pathophysiology of hemimasticatory spasm. J. Neurol. Neurosurg. Psychiatry 57:43, 1994. 40. Culp, W. J., Ochoa, J., Cline, M., and Dotson, R.: Heat and mechanical hyperalgesia induced by capsaicin: cross modality threshold modulation in human C nociceptors. Brain 112:1317, 1989. 41. Cusimano, M. D., Bilbao, J. M., and Cohen, S. M.: Hypertrophic brachial plexus neuritis: a pathological study of two cases. Ann. Neurol. 24:615, 1988. 42. Dalakas, M. C.: Chronic idiopathic ataxic neuropathy. Ann. Neurol. 19:545, 1986. 43. Dalakas, M. C., Teravainen, H., and Engel, W. K.: Tremor as a feature of chronic relapsing dysgammaglobulinaemic polyneuropathies. Arch. Neurol. 41:711, 1984. 44. Dancis, J., and Smith, A. A.: Familial dysautonomia. N. Engl. J. Med. 274:207, 1966. 45. Dejerine, P. J.: Sémiologie du système nerveux. In: Traité de Pathologie Générale. Paris, Masson et Cie, 1901. 46. Denny-Brown, D.: Primary sensory neuropathy with muscular changes associated with carcinoma. J. Neurol. Neurosurg. Psychiatry 11:73, 1948. 47. Denny-Brown, D.: Hereditary sensory radicular neuropathy. J. Neurol. Neurosurg. Psychiatry 14:237, 1951. 48. Denny-Brown, D., and Brenner, C.: Paralysis of nerve induced by direct pressure and by tourniquet. Arch. Neurol. Psychiatry 51:1, 1944. 49. Denny-Brown, D., and Foley, D. M.: Myokymia and the benign fasciculation of muscular cramps. Trans. Am. Nerves Assoc. 61:88, 1948. 50. Denny-Brown, D., and Pennybacker, J. B.: Fibrillation and fasciculation in voluntary muscle. Brain 61:311, 1938. 51. Derakshan, I., Loft, J., and Kaufman, B.: Ophthalmoplegia, ataxia and hyporeflexia (Fisher’s syndrome) with a midbrain lesion demonstrated by CT scanning. Eur. Neurol. 18:361, 1979. 52. Devor, M.: Central changes mediating neuropathic pain. In Dubner, R., Gebhart, G. F., and Bond, M. R. (eds.): Pain Research and Clinical Management. Proceedings of the Fifth World Congress on Pain, Vol. 3. Amsterdam, Elsevier, p. 114, 1988.

53. Devor, M., and Wall, P. D.: Cross-excitation in dorsal root ganglia of nerve injured and intact rats. J. Neurophysiol. 64:1733, 1990. 54. Diamond, J., Ochoa, J., and Culp, W. J.: An introduction to abnormal nerves and muscles as impulse generators. In Culp, W. J., and Ochoa, J. (eds.): Abnormal Nerves and Muscles as Impulse Generators. Oxford, UK, Oxford University Press, p. 3, 1982. 55. Donaghy, M., and Earl, C. J.: Ocular palsy preceding chronic relapsing polyneuropathy by several weeks. Ann. Neurol. 17:49, 1985. 56. Donaghy, M., Hakin, R. N., Bamford, J. M., et al.: Hereditary sensory neuropathy with neurotrophic keratitis: description of an autosomal recessive disorder with a selective reduction of small myelinated fibres and a discussion of the classification of the hereditary sensory neuropathies. Brain 110:563, 1987. 57. Doupe, J., Cullen, C. H., and Chance, G. Q.: Post-traumatic pain and the causalgic syndrome. J. Neurol. Neurosurg. Psychiatry 7:33, 1944. 58. Dressler, D., Thompson, P. D., Gledhill, R. F., and Marsden, C. D.: The syndrome of painful legs and moving toes. Mov. Disord. 9:13, 1994. 59. Dubner, R.: Neuronal plasticity and pain following peripheral tissue inflammation or nerve injury. In Bond, M. R., Charlton, J. E., and Woolf, C. J. (eds.): Proceedings of the VIth World Congress on Pain. Amsterdam, Elsevier, p. 263, 1991. 60. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy: I. Neurologic, genetic, and electrophysiologic findings in hereditary polyneuropathies. Arch. Neurol. 18:603, 1968. 61. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy: II. Neurologic, genetic, and electrophysiological findings in various neuronal degenerations. Arch. Neurol. 18:619, 1968. 62. Dyck, P. J., and Lambert, E. H.: Dissociated sensation in amyloidosis. Arch. Neurol. 20:490, 1969. 63. Dyck, P. J., Lambert, E. H., and Nicols, P. C.: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerve in health, Friedreich’s ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb, W. A. (ed.): Handbook of Electroencephalography and Clinical Neurophysiology, Vol. 9. Amsterdam, Elsevier, p. 83, 1972. 64. Dyck, P. J., Mellinger, J. F., Reagan, T. J., et al.: Not “indifference to pain” but varieties of hereditary sensory and autonomic neuropathy. Brain 106:373, 1983. 65. Dyck, P. J., Oviatt, K. F., and Lambert, E. H.: Intensive evaluation of referred undiagnosed neuropathies yields improved diagnosis. Ann. Neurol. 10:222, 1981. 66. Ekbom, K.: Restless legs: a clinical study. Acta Med. Scand. 158:1, 1945. 67. Elkington, J. St. C.: Recent work on the peripheral neuropathies. Proc. R. Soc. Med. 45:661, 1952. 68. Ellenberg, M., and Weber, H.: Retrograde ejaculation in diabetic neuropathy. Ann. Intern. Med. 65:1237, 1966. 69. England, A. C., and Denny-Brown, D.: Severe sensory changes and trophic disorder in peroneal muscular atrophy

Clinical Patterns of Peripheral Neuropathy

70. 71.

72.

73.

74.

75.

76.

77.

78. 79. 80. 81.

82. 83.

84.

85. 86.

87.

88.

89.

90.

(Charcot-Marie-Tooth type). Arch. Neurol. Psychiatry 67:1, 1952. Ewing, D. J.: Cardiovascular syndromes. New Issues Neurosci. 1:317, 1988. Ewing, D. J., Campbell, I. W., and Clarke, B. F.: The natural history of diabetic autonomic neuropathy. Q. J. Med. 49:95, 1980. Fine, E. J., and Hallett, M.: Neurophysiological study of sub-acute combined degeneration. J. Neurol. Sci. 45:331, 1980. French, J. M., Hall, G., Parish, D. J., and Smith, W. T.: Peripheral and autonomic nerve involvement in primary amyloidosis associated with uncontrollable diarrhoea and steatorrhoea. Am. J. Med. 39:277, 1965. Fullerton, P. M.: The effect of ischaemia on nerve conduction in the carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 26:385, 1963. Gabellini, A. S., Martinelli, P., Gulli, M. R., et al.: Orthostatic tremor: essential and symptomatic cases. Acta Neurol. Scand. 81, 113, 1990. Gamstorp, I., and Wohlfart, G.: A syndrome characterized by myokymia, muscular wasting and increased perspiration. Acta Psychiatr. Scand. 34:181, 1959. Garcia-Merino, A., Cabello, A., Mora, J. S., and Liano, H.: Continuous muscle fibre activity, peripheral neuropathy and thymoma. Ann. Neurol. 29: 215, 1991. Gardner-Medwin, D., and Walton, J. N.: Myokymia with impaired muscular relaxation. Lancet 1:127, 1969. Garland, H.: Diabetic amyotrophy. Br. Med. J. 2:1287, 1955. Gastaut, J. L.: Jambes douloureuses et orteils instabiles. Rev. Neurol. 142:641, 1986. Gilliatt, R. W., and Willison, R. G.: Peripheral nerve conduction in diabetic neuropathy. J. Neurol. Neurosurg. Psychiatry 25:11, 1962. Gilliatt, R. W., and Wilson, T. G.: A pneumatic-tourniquet test in the carpal-tunnel syndrome. Lancet 2:595, 1953. Gilliatt, R. W., and Wilson, T. G.: Ischaemic sensory loss in patients with peripheral nerve lesions. J. Neurol. Neurosurg. Psychiatry 17:104, 1954. Glocker, F. X., Deuschl, G., Volk, B., et al.: Bilateral myoclonus of the trapezius muscles after distal lesion of an accessory nerve. Mov. Disord. 11:571, 1996. Goodman, J. I.: Diabetic anhidrosis. Am. J. Med. 41:831, 1966. Gowers, W. R.: A Manual of Diseases of the Nervous System. Vol. I. Diseases of the Spinal Cord and Nerves. London, J. A. Churchill, 1886. Granit, R., Leksell, L., and Skoglund, C. R.: Fibre interaction in injured or compressed region of nerve. Brain 67:125, 1944. Greene, L. F., Kelalis, P. P., and Weeks, R. E.: Retrograde ejaculation of semen due to diabetic neuropathy: report of 4 cases. Fertil. Steril. 14:617, 1963. Greenhouse, A. H., Bickell, J. M., Pesch, R. N., and Seelinger, D. F.: Myotonia, myokymia, hyperhidrosis, and wasting of muscle. Neurology 17:263, 1967. Griffin, J. W., Cornblath, D. R., Alexander, E., et al.: Ataxic sensory neuropathy and dorsal root ganglionitis associated with Sjögren’s syndrome. Ann. Neurol. 27:304, 1990.

1157

91. Hahn, A. F., Brown, W. F., Koopman, W. J., and Feasby, T. E.: X-linked dominant hereditary motor and sensory neuropathy. Brain 113:1551, 1990. 92. Hahn, A. F., Parkes, A. W., Bolton, C. F., and Stewart, S. A.: Neuromyotonia in hereditary motor neuropathy. J. Neurol. Neurosurg. Psychiatry 54:230, 1991. 93. Halbach, M., Homberg, V., and Freund, H.-J.: Neuromuscular, autonomic and central cholinergic hyperactivity associated with thymoma and acetylcholine receptorbinding antibody. J. Neurol. 234:433, 1987. 94. Harding, A. E., and Le Fanu, J.: Carpal tunnel syndrome related to antebrachial Cimino-Brescia fistula. J. Neurol. Neurosurg. Psychiatry 40:511, 1977. 95. Harding, A. E., Muller, D. P. R., Thomas, P. K., and Willison, H. J.: Spinocerebellar degeneration secondary to chronic intestinal malabsorption: a vitamin E deficiency syndrome. Ann. Neurol. 12:419, 1982. 96. Harding, A. E., and Thomas, P. K.: The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103:259, 1980. 97. Harding, A. E., and Thomas, P. K.: Hereditary distal spinal muscular atrophy: a report on 34 cases and a review of the literature. J. Neurol. Sci. 45:337, 1980. 98. Harding, A. E., Thomas, P. K., Baraitser, M., et al.: X-linked recessive bulbospinal neuronopathy: a report of 10 cases. J. Neurol. Neurosurg. Psychiatry 45:1012, 1982. 99. Hart, I. K., Maddison, P., Newsom-Davis, J. N., et al.: Phenotypic variants of autoimmune peripheral nerve hyperexcitability. Brain 125:1887, 2002. 100. Head, H.: Sensation and the cerebral cortex. In: Studies in Neurology, Vol. II. London, Oxford Medical Publications, p. 741, 1920. 101. Hicks, E. P.: Hereditary perforating ulcer of the foot. Lancet 1:319, 1922. 102. Hilsted, J., Parving, H. H., Christensen, N. J., et al.: Hemodynamics in diabetic orthostatic hypotension. J. Clin. Invest. 68:1427, 1981. 103. Hollinrake, K.: Polymyositis presenting as distal muscle weakness: a case report. J. Neurol. Sci. 8:479, 1969. 104. Holmes, G.: Introduction to Clinical Neurology. Edinburgh, E. & S. Livingstone, 1946. 105. Howe, J. F., Loeser, J. D., and Calvin W. H.: Mechanosensitivity of dorsal root ganglia and chronically injured axons: a physiological basis for the radicular pain of nerve root compression. Pain 3:25, 1977. 106. Hreidarsson, A. B., and Gundersen, H. J. G.: Evaluation of pupillary function in diabetes. New Issues Neurosci. 1:333, 1988. 107. Irani, P. F., Purohit, A. V., and Wadia, N. H.: The syndrome of continuous muscle fibre activity. Acta Neurol. Scand. 55:273, 1977. 108. Isaacs, H.: Syndrome of continuous muscle fibre activity. J. Neurol. Neurosurg. Psychiatry 24:319, 1961. 109. Jackson, D. L., Satya-Murti, S., Davis, L., and Drachman, B.: Isaacs syndrome with laryngeal involvement: an unusual presentation of myokymia. Neurology 29:1612, 1979. 110. Jancsó, G., Kiraly, E., and Jancsó-Gabor, A.: Direct evidence for an axonal site of action of capsaicin. Arch. Pharmacol. 313:91, 1980.

1158

Diseases of the Peripheral Nervous System

111. Jankovic, J., and Pardo, R.: Segmental myoclonus. Arch. Neurol. 43:1025, 1986. 112. Jenkins, I. H., Bain, P., Colebatch, J. G., et al.: Neuropathic and essential tremor are both associated with abnormally increased activity in cerebellar pathways. J. Neurol. Neurosurg. Psychiatry 55:1216, 1992. 113. Kaufman, M. D.: Masticatory spasm in facial hemiatrophy. Ann. Neurol. 7:585, 1980. 114. Kaufman, M. D., Hopkins, L. C., and Hurwitz, B. J.: Progressive sensory neuropathy in patients with carcinoma: a disorder with distinctive clinical and electrophysiological findings. Ann. Neurol. 9:237, 1981. 115. Kennett, R. P., and Harding, A. E.: Peripheral neuropathy associated with the sicca syndrome. J. Neurol. Neurosurg. Psychiatry 49:90, 1986. 116. Kocen, R. S., Lloyd, J. K., Lascelles, P. T., et al.: Familial ␣-lipoprotein deficiency (Tangier disease) with neurological abnormalities. Lancet 1:1341, 1967. 117. Kocen, R. S., and Thomas, P. K.: Peripheral nerve involvement in Fabry’s disease. Arch. Neurol. 22:81, 1970. 118. Kocen, R. S., Thomas, P. K., King, R. H. M., and Haas, L. F.: Nerve biopsy findings in two cases of Tangier disease. Acta Neuropathol. 26:317, 1973. 119. Koller, R. L., and Blank, K.: Strawberry picker’s palsy. Arch. Neurol. 37:320, 1980. 120. Kugelberg, E.: Neurologic mechanisms for certain phenomena in tetany. Arch. Neurol. Psychiatry 56:507, 1958. 121. Kulisevsky, J., Marti-Fabregas, J., and Grau, J. M.: Spasms of amputation stumps. J. Neurol. Neurosurg. Psychiatry 55:626, 1992. 122. Lambert, E. H., and Rooke, E. D.: Myasthenic state and lung cancer. In Brain, R. L., and Norris, F. H. Jr. (eds).: The Remote Effects of Cancer on the Nervous System. New York, Grune & Stratton, p. 67, 1965. 123. Lance, J. W., Burke, D., and Pollard, J.: Hyperexcitability of motor and sensory neurons in neuromyotonia. Ann. Neurol. 5:523, 1979. 124. Leger, J. M., Lubetski, C., Bouche, P., et al.: Jambes douloureuses et orteils instabiles. Rev. Neurol. 141:296, 1985. 125. Léger, J. M., Younes-Chennoufi, A. B., Zuber, M., et al.: Frequency of central lesions in polyneuropathy associated with IgM monoclonal gammopathy: an MRI, neurophysiological and immunochemical study. J. Neurol. Neurosurg. Psychiatry 55:112, 1992. 126. Lewis, T., and Pickering, G. W.: Circulatory changes in the fingers in some diseases of the nervous system with special reference to the digital atrophy of peripheral nerve lesions. Clin. Sci. 2:149, 1936. 127. Lewis, T., Pickering, G. W., and Rothschild, P.: Centripetal paralysis arising out of arrested blood flow to the limb, including notes on a form of tingling. Heart 16:1, 1931. 128. Lindblom, U., and Ochoa, J.: Somatosensory function and dysfunction. In Asbury, A. K., McKhann, G. M., and McDonald, W. I. (eds.): Diseases of the Nervous System, 2nd ed., Vol. I. Philadelphia, W. B. Saunders, p. 213, 1992. 129. Lindblom, U., and Verillo, R. T.: Sensory functions in chronic neuralgia. J. Neurol. Neurosurg. Psychiatry 42:422, 1979. 130. Low, P. A., Burke, W. J., and McLeod, J. G.: Congenital sensory neuropathy with selective loss of small myelinated nerve fibres. Ann. Neurol. 3:179, 1978.

131. Lublin, F. D., Tsairis, P., Streletz, L. J., et al.: Myokymia and impaired muscle relaxation with continuous motor unit activity. J. Neurol. Neurosurg. Psychiatry 42:557, 1979. 132. Lütschg, J., Jerusalem, F., Ludin, H. P., et al.: The syndrome of continuous muscle fiber activity. Arch. Neurol. 35:198, 1978. 133. Lynn, B.: Capsaicin: actions on nociceptive C-fibres and therapeutic potential. Pain 41:61, 1990. 134. Mahloudji, M., Teasdall, R. D., Adamkiewicz, J. J., et al.: The genetic amyloidoses, with particular reference to hereditary neuropathic amyloidosis, type II (Indiana and Rukavina type). Medicine (Baltimore) 48:1, 1969. 135. Marchettini, P., Cline, M., and Ochoa, J. L.: Innervation territories for touch and pain afferents of single fascicles of the human ulnar nerve: mapping through intraneural microrecording and microstimulation. Brain 113:1491, 1990. 136. Marion, M. H., Gledhill, R. F., and Thompson, P. D.: Spasms of amputation stumps: a report of 2 cases. Mov. Disord. 4:354, 1989. 137. Martin, M. M.: Involvement of autonomic fibres in diabetic neuropathy. Lancet 1:560, 1953. 138. Martinez, M. S., Fontoira, M., Celesta, G., et al.: Myoclonus of peripheral origin: case secondary to a digital nerve lesion. Mov. Disord. 16:970, 2001. 139. Matthews, W. B.: Cryptogenic polyneuritis. Proc. R. Soc. Med. 45:667, 1952. 140. Matthews, W. B., and Esiri, M.: The migrant sensory neuritis of Wartenberg. J. Neurol. Neurosurg. Psychiatry 46:1, 1983. 141. McCombe, P. A., Pollard, J. D., and McLeod, J. G.: Chronic inflammatory demyelinating polyradiculoneuropathy. Brain 110:1617, 1987. 142. McDonald, W. I.: The effects of experimental demyelination on conduction in peripheral nerve: a histological and electrophysiological study. II. Electrophysiological observations. Brain 86:501, 1963. 143. McLeod, J. G.: Autonomic nervous system. In Sumner, A. (ed.): The Physiology of Peripheral Nerve Disease. Philadelphia, W. B. Saunders, 1980. 144. McLeod, J. G., Tuck, R. R., Pollard, J. D., et al.: Chronic polyneuropathy of undetermined cause. J. Neurol. Neurosurg. Psychiatry 47:530, 1984. 145. Meadows, J. C., Marsden, C. D., and Harriman, D. G. F.: Chronic spinal muscular atrophy in adults. J. Neurol. Sci. 9:527, 1969. 146. Medina, J. L., Chokroverty, S., and Reyes, M.: Localized myokymia caused by peripheral nerve injury. Arch. Neurol. 33:587, 1976. 147. Meier, C., and Moll, C.: Hereditary neuropathy with liability to pressure palsies: report of two families and review of the literature. J. Neurol. 228:73, 1982. 148. Meinck, H.-M., and Thompson, P. D.: The stiff man syndrome and related conditions. Mov. Disord. 17:853, 2002. 149. Mellon, R. R.: The relation of fatigue to paralysis localization in plumbism. Arch. Intern. Med. 12:399, 1913. 150. Melzack, R., and Wall, P. D.: Pain mechanisms: a new theory. Science 150:971, 1965. 151. Merrington, W. R., and Nathan, P. W.: A study of postischaemic paraesthesiae. J. Neurol. Neurosurg. Psychiatry 12:1, 1949.

Clinical Patterns of Peripheral Neuropathy 152. Merskey, D. M., Bond, M. R., Bonica, J. J., et al.: Classification of pain, descriptions of chronic pain syndromes and definitions of pain terms. Pain (Suppl.) 3:1, 1986. 153. Mertens, H. G., and Zschocke, S.: Neuromyotonie. Klin. Wochenschr. 43:91, 1965. 154. Mitsumoto, H., Levin, K. H., Wilbourn, A. J., and Chou, S. M.: Hypertrophic mononeuritis clinically presenting with painful legs and moving toes. Muscle Nerve 13:215, 1990. 155. Mitsumoto, H., Wilbourn, A. J., and Subramony, S. H.: Generalised myokymia and gold therapy. Arch. Neurol. 39:449, 1982. 156. Montagna, P., Cirignotta, F., Sacquegna, T., et al.: “Painful legs and moving toes” associated with polyneuropathy. J. Neurol. Neurosurg. Psychiatry 46:399, 1983. 157. Morgan, G. J., McGuire, J. L., and Ochoa, J.: Penicillamineinduced myositis in rheumatoid arthritis. Muscle Nerve 4:137, 1981 158. Mulder, D. W., Lambert, E. H., Bastron, J. A., and Sprague, R. G.: The neuropathies associated with diabetes mellitus: a clinical and electromyographic study of 103 unselected diabetic patients. Neurology 11:275, 1961. 159. Nakanishi, T., Sugita, H., Shimada, Y., and Toyokura, Y.: Neuromyotonia: a mild case. J. Neurol. Sci. 26:599, 1975. 160. Nathan, P. W.: On the pathogenesis of causalgia in peripheral nerve injuries. Brain 70:145, 1947. 161. Nathan, P. W.: Painful legs and moving toes: evidence on the site of the lesion. J. Neurol. Neurosurg. Psychiatry 41:934, 1978. 162. Newsom-Davis, J., and Mills, K. R.: Immunological associations of acquired neuromyotonia (Isaac’s syndrome). Brain 116:453, 1993. 163. Nielsen, V. K.: The peripheral nerve function in chronic renal failure: a survey. Acta Med. Scand. 573(Suppl.): 1, 1974. 164. Nielsen, V. K.: Pathophysiology of hemifacial spasm: I. Ephaptic transmission and ectopic excitations. Neurology 34:418, 1984. 165. Noordenbos, W., and Wall, P. D.: Implications of the failure of nerve resection and graft to cure chronic pain produced by nerve lesions. J. Neurol. Neurosurg. Psychiatry 44:1068, 1981. 166. Nordin, M., Nyström, B., Wallin, U., and Hagbarth, K.-E.: Ectopic sensory discharges and paresthesiae in patients with disorders of peripheral nerves, dorsal roots and dorsal columns. Pain 20:231, 1984. 167. Norris, F. H., Gasteiger, E. L., and Chatfield, P. O.: An electromyographic study of induced and spontaneous muscle cramps. Electroencephalogr. Clin. Neurophysiol. 9:139, 1957. 168. Ochoa, J.: Pain in local nerve lesions. In Culp, W. J., and Ochoa, J. (eds.): Abnormal Nerves and Muscles as Impulse Generators. New York, Oxford University Press, p. 568, 1982. 169. Ochoa, J.: The newly recognized painful ABC syndrome: thermographic aspects. Thermology 2:65, 1986. 170. Ochoa, J.: Neuropathic pains from within: personal experiences, experiments and reflections on mythology. In Dimitrijevic, R., Wall, P. D., and Lindblom, U. (eds.): Recent Achievements in Restorative Neurology 3: Altered Sensation and Pain. Basel, Karger, p. 100, 1990.

1159

171. Ochoa, J., Cline, M., Dotson, R., and Marchettini, P.: Pain and paresthesias provoked mechanically in human cervical root entrapment (sign of Spurling): single sensory unit antidromic recording of ectopic, bursting, propagated nerve impulse activity. In Pubols, L. M., and Sessle, B. J. (eds.): Effects of Injury on Trigeminal and Spinal Somatosensory Systems. New York, Alan R Liss, p. 389, 1987. 172. Ochoa, J., Fowler, T. J., and Gilliatt, R. W.: Anatomical changes in peripheral nerves compressed by a pneumatic tourniquet. J. Anat. 113:433, 1972. 173. Ochoa, J., Marchettini, P., and Cline, M.: Lessons from human research on the pathophysiology of neuropathic pains in limbs. In Wynn Parry, C. B. (ed.): Management of Pain in the Hand and Wrist. London, Churchill Livingstone, p. 28, 1990. 174. Ochoa, J., and Neary, D.: Localised hypertrophic neuropathy, intraneural tumour, or chronic nerve entrapment. Lancet 1:632, 1975. 175. Ochoa, J., and Torebjörk, H. E.: Paraesthesiae from ectopic impulse generation in human sensory nerves. Brain 103:835, 1980. 176. Ochoa, J., and Torebjörk, H. E.: Sensations evoked by intraneural microstimulation of identified C nociceptor fibres in human skin nerves. J. Physiol. (Lond.) 415:583, 1989. 177. Ochoa, J., Torebjörk, H. E., and Culp, W.: Determinants of subjective attributes of normal cutaneous sensation and of paresthesiae from ectopic nerve impulse generation. In von Euler, C., Franzén, O., Lindblom, U., and Ottoson, D. (eds.): Somatosensory Mechanisms. Wenner-Gren International Symposium, vol. 41. London, Macmillan, p. 379, 1984. 178. Paley, R. G., Mitchell, W., and Watkinson, G.: Terminal colonic dilatation following intractable diarrhea in a diabetic. Gastroenterology 4:401, 1961. 179. Partanen, V. S. J., Soininen, H., Saska, M., and Riekkinen, P.: Electromyographic and nerve conduction findings in a patient with neuromyotonia, normocalcaemic tetany and small-cell lung cancer. Acta Neurol. Scand. 61:216, 1980. 180. Rasminsky, M.: Ectopic excitation, ephaptic excitation and autoexcitation in peripheral nerve fibers of mutant mice. In Culp, W., and Ochoa, J. (eds.): Abnormal Nerves and Muscles as Impulse Generators. Oxford, UK, Oxford University Press, p. 344, 1982. 181. Rexed, B.: Contributions to the knowledge of the postnatal development of the peripheral nervous system in man. Acta Psychiatr. Scand. Suppl. 33, 1944. 182. Ridley, A., Hierons, R., and Cavanagh, J. B.: Tachycardia and the neuropathy of porphyria. Lancet 2:708, 1968. 183. Riggs, J. E., Gutmann, L., and Schochet, S. S.: Contraction pseudotremor of chronic denervation. Arch. Neurol. 40:518, 1983. 184. Roth, G.: The origin of fasciculations. Ann. Neurol. 12:542, 1982. 185. Rothwell, J. C., Traub, M. M., Day, B. L., et al.: Manual motor performance in a deafferented man. Brain 105:515, 1982. 186. Rundles, R. W.: Diabetic neuropathy: general review with report of 125 cases. Medicine (Baltimore) 24:111, 1945. 187. Rushworth, G., Lishman, W. A., Hughes, J. T., and Oppenheimer, D. R.: Intense rigidity of the arms due to isolation of motor neurons by a spinal tumour. J. Neurol. Neurosurg. Psychiatry 24:132, 1961.

1160

Diseases of the Peripheral Nervous System

188. Sabin, T. D.: Temperature-linked sensory loss: a unique pattern in leprosy. Arch. Neurol. 20:257, 1969. 189. Said, G., and Bathien, N.: Myoclonies rythmées du quadriceps en relation avec un envalessment sarcomateux du nerf crural. Rev. Neurol. (Paris) 133:191, 1977. 190. Said, G., Bathien, N., and Cesaro, P.: Peripheral neuropathies and tremor. Neurology 32:480, 1982. 191. Said, G., Slama, G., and Selva, J.: Progressive centripetal degeneration of axons in small fibre type diabetic polyneuropathy: a clinical and pathological study. Brain 106:791, 1983. 192. Saida, T., Saida, K., Lisak, R. P., et al.: In vivo demyelinating activity of sera from patients with Guillain-Barré syndrome. Ann. Neurol. 11:69, 1982. 193. Salvi, F., Montagna, P., Plasmati, R., et al.: Restless legs syndrome and nocturnal myoclonus: initial clinical manifestation of familial amyloid polyneuropathy. J. Neurol. Neurosurg. Psychiatry 53:522, 1990. 194. Sanders, D. B.: Ephaptic transmission in hemifacial spasm: a single-fiber EMG study. Muscle Nerve 12:690, 1989. 195. Schoenen, J., Gonce, M., and Delwaide, P. J.: Painful legs and moving toes: a syndrome with different pathophysiologic mechanisms. Neurology 34:1108, 1984. 196. Schott, G. D.: Induction of involuntary movements by peripheral trauma: an analogy with causalgia. Lancet 2:712, 1986. 197. Schwartz, M. S., and Swash, M.: Scapuloperoneal atrophy with sensory involvement: Davidenkow’s syndrome. J. Neurol. Neurosurg. Psychiatry 38:1063, 1975. 198. Schwartz, O., and Jampel, R. S.: Congenital blepharophimosis associated with a unique generalized myopathy. Arch. Ophthalmol. 69:52, 1962. 199. Shahani, B., and Spalding, J. M. K.: Diabetes mellitus presenting with bilateral foot drop. Lancet 2:930, 1969. 200. Shahani, B., Young, R. R., and Adams, R. D.: The tremor in Roussy-Levy syndrome. Neurology 23:425, 1973. 201. Shepard, C. C.: Temperature optimum of Mycobacterium leprae in mice. J. Bacteriol. 90:1271, 1965. 202. Shibasaki, H., Kakigi, R., Ohnishi, A., and Kuroiwa, Y.: Peripheral and central conduction in subacute myelooptico neuropathy. Neurology (Minneap.) 32:1186, 1982. 203. Sigwald, J., Raverdy, P., Fardeau, M., et al.: Pseudomyotonie—forme particulière d’hypertonie musculaire à prédominance distale. Rev. Neurol. 115:1003, 1966. 204. Sluga, E.: Correlations of electrophysiological and morphological methods in neuropathies. In Kunze, K., and Desmedt, J. (eds.): Studies on Neuromuscular Diseases: Proceedings of the International Symposium on Neuromuscular Diseases, April 1973, Giessen. Basel, S. Karger, 1975. 205. Smith, B.: The Neuropathology of the Alimentary Tract. London, Edward Arnold, 1972. 206. Smith, I. S.: The natural history of chronic demyelinating neuropathy associated with benign IgM paraproteinaemia: a clinical and neurophysiological study. Brain 117:949, 1994. 207. Smith, I. S., Kahn, S. N., Lacey, B. W., et al.: Chronic demyelinating neuropathy associated with benign IgM paraproteinaemia. Brain 106:169, 1983. 208. Sotaneimi, K.: Paraspinal myoclonus due to a spinal root lesion. J. Neurol. Neurosurg. Psychiatry 48:722, 1985.

209. Spencer, P. S., and Schaumburg, H. H.: Central peripheral distal axonopathy—the pathology of dying-back polyneuropathies. Prog. Neuropathol. 3:253, 1976. 210. Spillane, J. D., Nathan, P. W., Kelly, R. E., and Marsden, C. D.: Painful legs and moving toes. Brain 94:541, 1971. 211. Spillane, J. D., and Wells, C. E. C.: Acrodystrophic Neuropathy. London, Oxford Medical Publications, 1969. 212. Spiro, A. J.: Minipolymyoclonus: a neglected sign of childhood spinal muscular atrophy. Neurology 20:1124, 1970. 213. Stålberg, E., and Trontelj, J. V.: Abnormal discharges generated within the motor unit as observed with single-fiber electromyography. In Culp, W. J., and Ochoa, J. (eds.): Abnormal Nerves and Muscles as Impulse Generators. New York, Oxford University Press, p. 443, 1982. 214. Sterman, A. B., Schaumberg, H. H., and Asbury, A. K.: The acute sensory neuronopathy syndrome: a distinct clinical entity. Ann. Neurol. 7:354, 1980. 215. Stewart, B. M.: The hypertrophic neuropathy of acromegaly: a rare neuropathy associated with acromegaly. Arch. Neurol. 14:107, 1966. 216. Sumner, A. J., Saida, K., Saida, T., et al.: Acute conduction block associated with experimental antiserum mediated demyelination of peripheral nerve. Ann. Neurol. 11:469, 1982. 217. Sunderland, S.: Capacity of reinnervated muscles to function efficiently after prolonged denervation. Arch. Neurol. Psychiatry 64:755, 1950. 218. Sunderland, S.: Nerves and Nerve Injury, 2nd ed. Edinburgh, E. & S. Livingstone, 1978. 219. Tarlov, I. M.: Rigidity in man due to spinal interneuron loss. Arch. Neurol. 16:536, 1967. 220. Taylor, B. V., Dyck, P. J. B., Engelstad, J., et al.: Multifocal motor neuropathy: pathologic alterations at the site of conduction block. J. Neuropathol. Exp. Neurol. 63:129, 2004. 221. Taylor, R. G., Layzer, R. B., Davis, H. S., and Fowler, W. M. Jr.: Continuous muscle fiber activity in the Schwartz-Jampel syndrome. Electroencephalogr. Clin. Neurophysiol. 33:497, 1972. 222. Tenckhoff, H. A., Boen, S. T., Jebsen, R. H., and Spiegler, J. H.: Polyneuropathy of chronic renal insufficiency. JAMA 192:1121, 1965. 223. ter Bruggen, J. P., and Tijssen, C. C.: Crural and axial myoclonic dystonia following meralgia paraesthetica. Mov. Disord. 3:176, 1988. 224. Thévenard, A.: L’acropathie ulcéro-mutilante familiale. Rev. Neurol. (Paris) 74:193, 1942. 225. Thomas, P. K.: Screening for peripheral neuropathy in patients treated by chronic hemodialysis. Muscle Nerve 1:396, 1978. 226. Thomas, P. K.: Pain in peripheral neuropathy: clinical and morphological aspects. In Ochoa, J., and Culp, W. (eds.): Abnormal Nerves and Muscles as Impulse Generators. New York, Oxford University Press, p. 553, 1982. 227. Thomas, P. K.: The selective vulnerability of the centrifugal and centripetal axons of primary sensory neurons. Muscle Nerve 5:S117, 1982. 228. Thomas, P. K., Calne, D. B., and Stewart, G.: Hereditary motor and sensory polyneuropathy (peroneal muscular atrophy). Ann. Hum. Genet. 38:111, 1974.

Clinical Patterns of Peripheral Neuropathy 229. Thomas, P. K., and Fullerton, P. M.: Nerve fibre size in the carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 26:520, 1963. 230. Thomas, P. K., Hollinrake, K., Lascelles, R. G., et al.: The polyneuropathy of chronic renal failure. Brain 94:761, 1971. 231. Thomas, P. K., and King, R. H. M.: Peripheral nerve changes in amyloid neuropathy. Brain 97:395, 1974. 232. Thomas, P. K., King, R. H. M., Chiang, T. R., et al.: Neurofibromatous neuropathy. Muscle Nerve 13:93, 1990. 233. Thomas, P. K., and Lascelles, R. G.: Hypertrophic neuropathy. Q. J. Med. 36:223, 1967. 234. Thomas, P. K., Lascelles, R. G., Hallpike, J. F., and Hewer, R. L.: Recurrent and chronic relapsing Guillain-Barré polyneuritis. Brain 92:589, 1969. 235. Thompson, P. D., and Carroll, W. M.: Hemimasticatory spasm: a peripheral paroxysmal cranial neuropathy? J. Neurol. Neurosurg. Psychiatry 46:274, 1983. 236. Torebjörk, H. E., and Ochoa, J. L.: Specific sensations evoked by activity in single identified sensory units in man. Acta Physiol. Scand. 110:445, 1980. 237. Torebjörk, H. E., Ochoa, J. L., and Schady, W.: Referred pain from intraneural stimulation of muscle fascicles in the median nerve. Pain 18:145, 1984. 238. Trontelj, J., and Stalberg, E.: Bizarre repetitive discharges recorded with single fibre EMG. J Neurol. Neurosurg. Psychiatry 46:310, 1983. 239. Valenstein, E., Watson, R. T., and Parker, J. L.: Myokymia, muscle hypertrophy and percussion myotonia in chronic recurrent polyneuropathy. Neurology 28:1130, 1978. 240. Vallbo, A. B., Hagbarth, K.-E., Torebjörk, H. E., and Wallin, B. G.: Somatosensory proprioceptive and sympathetic activity in human peripheral nerves. Physiol. Rev. 59:919, 1979. 241. Van Dyke, D. H., Griggs, R. C., Murphy, J. M., and Goldstein, M. N.: Hereditary myokymia and periodic ataxia. J. Neurol. Sci. 25:109, 1975. 242. Vasilescu, C., Alexianu, M., and Dan, A.: Neuronal type of Charcot-Marie-Tooth disease with a syndrome of continuous motor unit activity. J. Neurol. Sci. 63:11, 1984. 243. Vasilescu, C., and Florescu, A.: Peripheral neuropathy with a syndrome of continuous motor unit activity. J. Neurol. 226:275, 1982.

1161

244. Verdugo, R., and Ochoa, J.: Quantitative somatosensory thermotest: a key method for functional evaluation of small caliber afferent channels. Brain 115:100, 1992. 245. Verhagen, W. I. M., Horstink, M. W. I. M., and Notermans, S. L. H.: Painful arm and moving fingers. J. Neurol. Neurosurg. Psychiatry 48:384, 1985. 246. Wallis, W. E., van Poznak, A., and Plum, F.: Generalized muscular stiffness, fasciculations, and myokymia of peripheral nerve origin. Arch. Neurol. 22:430, 1970. 247. Walsh, J. C.: Neuromyotonia: an unusual presentation of intrathoracic malignancy. J. Neurol. Neurosurg. Psychiatry 39:1086, 1976. 248. Walters, A., Hening, W., Cote, L., and Fahn, S.: Dominantly inherited restless legs with myoclonus and periodic movements of sleep: a syndrome related to endogenous opiates? Adv. Neurol. 43:309, 1986. 249. Wartenberg, R.: Neuritis, Sensory Neuritis, Neuralgia. New York, Oxford University Press, p. 233, 1958. 250. Watkins, P. J.: Facial sweating after food: a new sign of diabetic autonomic neuropathy. Br. Med. J. 1:583, 1973. 251. Wearness, E.: Neuromyotonia and bronchial carcinoma. Electromyogr. Clin. Neurophysiol. 14:527, 1974. 252. Webster, H. de F., Schröder, J. M., Asbury, A. K., and Adams, R. D.: The role of Schwann cells in the formation of “onion bulbs” found in chronic neuropathies. J. Neuropathol. Exp. Neurol. 26:276, 1967. 253. Welander, L.: Myopathia distalis tarda hereditaria. Acta Med. Scand. Suppl. 265, 1951. 254. Willis, W. D.: Electrophysiological evidence for a role of altered discharges of spinothalamic tract neurons in hyperalgesia. In Dimitrijevic, R., Wall, P. D., and Lindblom, U. (eds.): Recent Achievements in Restorative Neurology 3: Altered Sensation and Pain. Basel, Karger, 1990. 255. Windebank, A. J., Blexrud, M. D., Dyck, P. J., et al.: The syndrome of acute sensory neuropathy: clinical features and electrophysiologic and pathologic changes. Neurology 40:584, 1990. 256. Zelená, J.: Development, degeneration and regeneration of receptor organs. Prog. Brain Res. 13:175, 1964. 257. Zisfein, J., Sivak, M., Aron, A., and Bender, A. N.: Isaacs’ syndrome with muscle hypertrophy reversed by phenytoin therapy. Arch. Neurol. 40:241, 1983.

46 Differential Diagnosis of Polyneuropathy IAN A. GRANT AND TIMOTHY J. BENSTEAD

Diagnostic Yield in Assessment of Polyneuropathy Approaches to the Diagnosis of Neuropathy Pattern Class of Affected Fibers Pathophysiology Electrophysiology Temporal Profile

Associated Symptoms and Signs of Disease Non-neurologic Disease Neurologic Disease Family History Laboratory Investigations Screening Tests Monoclonal Gammopathy and Antiglycolipid Antibodies Nerve Biopsy

Several authors have proposed diagnostic approaches and algorithms for the investigation of polyneuropathy, and this literature continues to grow. This can be explained by several factors. Peripheral neuropathy is common, and thus many physicians will be faced with it at some point in their careers. Many of these physicians are general neurologists or non-neurologists with limited expertise in the assessment of neuropathies. The differential diagnosis is formidable, with hundreds of potential causes to be considered. The majority of patients encountered in day-to-day practice have a length-dependent, predominantly sensory polyneuropathy mainly affecting the lower limbs, a nonspecific clinical picture that tells the clinician little about the underlying cause. Because of these realities, it would be useful to have a clearly defined approach to the differential diagnosis of peripheral neuropathy. Such an approach would be helpful for several reasons: 1. 2. 3. 4.

Improved diagnostic accuracy Time saving Cost-effectiveness Avoidance of unnecessary tests (particularly invasive ones) 5. Avoidance of unnecessary treatments Recommended approaches for the diagnosis of neuropathy have been published by several experts.9,27,28,73,86 These published approaches share some common points (such as

Other Investigations Disorders Simulating Peripheral Neuropathy Painful Legs and Moving Toes Erythromelalgia Myopathy Motor Neuron Disease Myelopathy Aging

important principles of investigation) while differing in others (such as recommended laboratory tests). It is not clear that any of these algorithms is better than the others, and no single algorithm will be ideal for all clinicians and institutions. An approach appropriate for the assessment of difficult cases referred to a tertiary center such as a university neuromuscular clinic may not be useful for a general neurologist in community practice. Factors to consider in choosing an algorithm include the individual expertise of the clinician, the diagnostic resources available to him or her, and the patient population to be assessed. It is difficult to compare rival approaches, and comparative data are unavailable. Thus the “best” approach will remain somewhat subjective. The guidelines provided in this chapter reflect the views and experience of the authors and are derived in part from practice in a large referral center. Some of the points to be made here have been published previously.27,28

DIAGNOSTIC YIELD IN ASSESSMENT OF POLYNEUROPATHY Despite intensive investigation, a proportion of patients with peripheral neuropathy will remain undiagnosed. The size of this group will depend on the extent of the work-up, which will in turn vary from center to center. 1163

1164

Diseases of the Peripheral Nervous System

Dyck et al.31 described the diagnostic yield in 205 cases of neuropathy without obvious cause referred to the Mayo Clinic. All patients underwent a complete history, physical examination, and electrodiagnostic testing. Other testing was performed according to the clinical indication in each patient. Some but not all patients underwent sural nerve biopsy. In 24%, the cause could not be found. Among those in whom a diagnosis was made, the largest group (42%) had an inherited neuropathy. McLeod et al.87 described the results of investigations of 519 patients, all of whom underwent sural nerve biopsy. After initial assessment, only 13% remained undiagnosed. Follow-up data ranging from 4 months to 12 years were available in 70% of these initially undiagnosed individuals, among whom a possible cause was eventually found in 36%. A more pessimistic view was offered by Fagius,37 who established a diagnosis in only 26% of 91 patients, none of whom underwent nerve biopsy. He concluded that only limited investigations are justified in most cases of chronic peripheral neuropathy if the cause is not obvious, although the aforementioned studies suggest otherwise.

APPROACHES TO THE DIAGNOSIS OF NEUROPATHY The imposing list of potential causes in an individual patient with neuropathy can be handled in a variety of ways. Some clinicians choose to employ a “shotgun” approach. This refers to the use of a standard, often small battery of tests, with little regard for the patient’s specific clinical and electrodiagnostic characteristics. Non-neurologists commonly employ this approach. For example, many primary care physicians obtain a serum vitamin B12 level, thyroid function tests, and routine biochemistry, and in doing so may occasionally identify a relevant disorder. However, this approach permits the detection of only a small number of potential causes, and the tests often selected (e.g., serologic testing for syphilis) have a low yield. The issue of utility of specific screening tests is discussed in more detail later. A diagnosis may also be made through pattern recognition. Despite lip service to the importance of a structured diagnostic approach, there is reason to believe that many experienced clinicians actually rely on this approach. This method relies on the clinician’s expertise, and will fail if the patient’s problem happens to lie outside his or her experience. Most would agree that the best strategy is to use a systematic approach that integrates clinical, electrodiagnostic, and laboratory data. A thorough history and general and neurologic examination are the foundation of this approach, because all subsequent investigations are guided by this initial clinical assessment. Electrophysiologic testing is important in practically all patients and is done at

an early stage. During this assessment process, the goal is to establish a number of specific characteristics of the patient’s neuropathy, as described below. This will gradually narrow the differential diagnosis, limit investigations, improve diagnostic accuracy, and lead to a definitive diagnosis in the majority of patients.

Pattern A basic step in narrowing the large differential diagnosis is to establish the anatomic pattern of peripheral nerve involvement. This determination will be based largely on the neurologic examination, although the history, electromyography (EMG), and sometimes other tests also contribute useful information. At issue is the proximalto-distal level of peripheral nerve affected, that is, primary sensory neuron/anterior horn cell, nerve root, plexus, peripheral nerve trunk, distal nerve terminal, or a combination of these. Also to be determined is whether the process is focal (mononeuropathy), multifocal (multiple mononeuropathies), or diffuse (polyneuropathy). Is the process a distal, length-dependent one with preferential damage to the longest axons, in which case symptoms and deficits first appear in the feet? If so, the differential diagnosis remains extensive and much work remains to be done. In contrast, the differential diagnosis will be smaller if any other pattern of neuropathy can be demonstrated. Because the pattern may evolve over time, a careful history, review of medical records, and/or serial evaluations may be required. For example, necrotizing vasculitis may produce distinct multiple mononeuropathies early in the course of the disease, becoming more confluent as the number of lesions increases over time, and eventually resulting in a more or less symmetrical polyneuropathy. Considerations in assessment include whether the neuropathic pattern is symmetrical or asymmetrical. The severity and time of onset of upper limb involvement relative to lower limb involvement, the degree to which proximal limb muscles are involved compared to distal ones, and the involvement of the trunk or head (suggesting radicular or cranial nerve involvement, respectively) should also be considered. A classification of neuropathies with uncommon or distinctive anatomic patterns is presented in Table 46–1. Upper limb predominance is an uncommon finding in polyneuropathies, but can occur in Guillain-Barré syndrome, chronic inflammatory demyelinating polyneuropathy (CIDP),14,45,134 and porphyria.117 A syringomyelia-like presentation of Tangier disease has been described in a few patients, with muscle wasting, weakness, and loss of pain and thermal sensation mainly affecting the arms and face.43,111 Vitamin B12 deficiency can present this way, although perhaps as a result of dorsal column involvement rather than neuropathy. A family with hereditary motor neuropathy mainly involving the upper limbs has been reported,68 and an

1165

Differential Diagnosis of Polyneuropathy

Table 46–1. Less Common or Distinctive Patterns of Peripheral Neuropathy and Their Causes Multiple Mononeuropathies Individual nerves are selectively involved Vasculitis Multifocal motor neuropathy with conduction block Hereditary neuropathy with liability to pressure palsies Leprosy Human immunodeficiency virus infection Malignant infiltration Trauma Multifocal acquired demyelinating sensory and motor neuropathy (MADSAM) Polyradiculoneuropathies Polyneuropathies in which there is disproportionate involvement of nerve roots Lyme disease Sarcoidosis Guillain-Barré syndrome (more often symmetrical) Sensory Polyganglionopathies The disease process targets sensory neurons in dorsal root and/or trigeminal ganglia Sjögren’s syndrome Paraneoplastic Idiopathic Regional Neuropathies Proximal diabetic neuropathy Idiopathic lumbar plexopathy Acute brachial plexus neuropathy Ischemic monomelic neuropathy Polyneuropathy with Superimposed Focal Neuropathy (Such as Carpal Tunnel Syndrome) May occur with polyneuropathy of any cause Common in: Diabetes Hypothyroidism Amyloidosis

upper limb form of hereditary motor and sensory neuropathy was described in the 19th century.50 Interpretation of upper limb symptoms may be difficult in patients with polyneuropathy because such symptoms may result from the polyneuropathy itself or from superimposed nerve compression. Carpal tunnel syndrome (CTS) is the most common focal neuropathy and often occurs in patients with polyneuropathy, who may be predisposed to it as a result of increased susceptibility of an already compromised median nerve. This phenomenon is nonspecific, although CTS is particularly common in patients with hereditary neuropathy with liability to pressure palsies (HNPP) and amyloidosis, particularly the primary form and familial amyloid polyneuropathy type II.122,144 In HNPP, a diffuse polyneuropathy may be apparent from nerve con-

duction studies even in patients with clinical changes limited to a single nerve.4,96

Class of Affected Fibers The next step is to consider the degree to which motor, sensory, and autonomic fibers are affected. Many diseases affecting both sensory and motor fibers cause exclusively sensory complaints and deficits at an early stage, with motor involvement becoming clinically apparent only later. An example would be a patient with long-standing, poorly controlled diabetes presenting with paresthesias and pain in the feet, in whom examination discloses distal sensory deficits and ankle areflexia without muscle weakness. Although this patient might be judged to have a purely sensory process, EMG studies may reveal subclinical motor abnormalities also. While a large number of diseases produce both motor and sensory changes, the differential diagnosis of a purely sensory or motor neuropathy is limited and so this distinction should be made with care. Symptoms and clinical findings usually reflect the affected population(s). In addition to muscle weakness, patients with motor involvement may describe cramps, fasciculations, or other positive phenomena. The absence of symptomatic weakness does not rule out motor involvement because mild weakness may not be evident to the patient, particularly when the process is long standing, as in many patients with inherited neuropathy. Cramps are nonspecific, although reported by three times as many patients with inherited neuropathy as by those with acquired inflammatory demyelinating neuropathy in one series.31 Where sensory involvement is present, the disease may affect large myelinated fibers, small myelinated fibers, unmyelinated fibers, or a combination of these. Individuals with large fiber dysfunction describe positive or negative symptoms. Loss of sensation is typically described as numbness, although other terms may be used; patients may describe a “wrapped” feeling or a sensation like walking on pillows. Positive phenomena (paresthesias) may be described in a variety of ways, including tingling, pins and needles, prickling, or crawling. Some describe a sense that a foreign body (such as a pebble or wrinkled sock) is underfoot. The proprioceptive deficit resulting from loss of large myelinated afferent fibers results in sensory ataxia, manifested as gait instability worsened by loss of compensatory visual information (e.g., in the shower or in darkness). Signs of large fiber sensory neuropathy include abnormal joint position and vibration sense and a positive Romberg test. Severe deafferentation in the upper limb produces pseudoathetosis. Small fiber dysfunction may produce spontaneous pain. To some degree, the quality of pain reflects the class of fibers affected. For example, several lines of evidence suggest that loss of unmyelinated C fibers results in dull or burning pain.103 Dysfunction of thinly myelinated A␦ fibers

1166

Diseases of the Peripheral Nervous System

may more often be sharp or lancinating in quality. The distinction is not absolute and has limited clinical utility, because a combination of A␦ and C fiber involvement often coexists. Small fiber loss also results in loss of protective sensation, which may lead to ulcerations, burns, and other manifestations of unrecognized trauma. Patients may describe an inability to feel the temperature of their bath until the leg has been submerged beyond a certain point. Sensitization of nociceptors may lead to hyperalgesia, whereby the intensity of a painful stimulus is exaggerated, or allodynia, whereby nonpainful mechanical stimuli such as contact from footwear or bedsheets is perceived as painful. Sensory abnormalities involving large fibers can be demonstrated via nerve conduction studies (see Electrophysiology below). However, small myelinated and unmyelinated fibers cannot be assessed with conventional nerve conduction studies. Quantitative sensation testing is increasing in availability and provides information about all three classes of sensory fibers. Large fiber function can be tested via vibration threshold, small myelinated fiber function through cooling thresholds, and unmyelinated fiber function via heat-pain thresholds.49 Quantitative information regarding epidermal unmyelinated fibers can be obtained from a punch biopsy of the skin, but this method has not yet been widely adopted.85 Sural nerve biopsy permits quantitation of all fiber classes, although a biopsy is seldom indicated for this purpose alone. Autonomic neuropathy is suggested by symptoms such as dryness of the eyes or mouth; tonic pupils; sweating abnormalities such as distal hypohidrosis or hyperhidrosis over the head or upper body; gastrointestinal symptoms such as bloating, early satiety, diarrhea, or constipation; orthostatic hypotension; vasomotor changes in the limbs; and sexual dysfunction. Autonomic testing is available at some centers to characterize these changes. For example, heart rate (R-R interval) variation with deep breathing provides information regarding cardiac parasympathetic innervation,107 while sympathetic sudomotor fibers can be assessed through the thermoregulatory sweat test, quantitative sudomotor axon reflex test, and other tests.40,81

Table 46–2. Demyelinating Polyneuropathies Inherited

Acquired

Charcot-Marie-Tooth (CMT) disease CMT1 CMT3 (DéjérineSottas disease) CMT4 (recessive inheritance) CMT X Hereditary neuropathy with liability to pressure palsies Refsum’s disease Leukodystrophies Adrenoleukodystrophy/ adrenomyeloneuropathy Metachromatic leukodystrophy Cockayne’s syndrome Krabbe’s disease Pelizaeus-Merzbacher disease

Guillain-Barré syndrome Chronic inflammatory demyelinating polyradiculoneuropathy Monoclonal gammopathy Multifocal motor neuropathy with conduction block* Diphtheria Drugs Amiodarone Perhexiline Cytosine arabinoside

*Demyelination assumed based on electrophysiology but not pathologically proven.

in which “intermediate” conduction velocities occur, or in some cases of diabetic neuropathy in which the degree of conduction slowing exceeds that which one would expect in a purely axonal process, possibly as a result of metabolic factors. This determination is based mainly on electrodiagnostic testing, but clinical clues may be available; demyelinating neuropathy should be considered in patients with long-standing muscle weakness without atrophy and in patients with nerve hypertrophy. Nerve biopsy may provide further information but is rarely necessary to make this distinction and may be misleading, because sampling error or multifocality may result in false negatives.

Electrophysiology Pathophysiology Pathologic alterations of nerve fibers may be broadly divided into primary axonal degeneration and segmental demyelination. The vast majority of toxic, metabolic, and inherited disorders encountered result in an axonal neuropathy for which the differential diagnosis is large. Primary demyelinating neuropathies are less common, and the differential diagnosis is correspondingly smaller (Table 46–2). This distinction may be unclear, for example, in chronic demyelinating neuropathies in which secondary axon loss has occurred, in inherited neuropathies (particularly X-linked Charcot-Marie-Tooth [CMT] disease)

EMG is widely available and provides fundamental information about peripheral nerve function, making it the most important diagnostic test in the evaluation of peripheral neuropathy. The electromyogram provides information about the pattern of nerves affected, the degree of motor versus sensory involvement, the presence of axon loss and/or demyelination, and the severity and chronicity of these changes. Nerve conduction studies show evidence of motor axon loss in the form of reduced compound muscle action potential amplitudes with normal or only mildly reduced conduction velocities. Denervation resulting from axon loss is seen

1167

Differential Diagnosis of Polyneuropathy

as fibrillation potentials and positive sharp waves on needle EMG. Loss of functioning motor units as a result of axon loss or demyelinating conduction block (see below) results in reduced recruitment. Chronic incomplete axon loss permits reinnervation, evident as high-amplitude or long-duration, polyphasic motor unit potentials on needle examination. Sensory axon loss leads to reduced or absent sensory nerve action potentials and loss of H reflexes. In demyelinating neuropathies, findings include reduced motor and sensory conduction velocities, prolonged distal latencies, temporal dispersion, motor conduction block, and prolonged F-wave latencies. Diagnostic criteria for primary demyelination have been published3,10,115; these criteria appear to have similar sensitivity.12 Demyelination does not necessarily affect all nerves equally. Uniform demyelination implies a similar degree of involvement of proximal and distal segments of the same nerve, and between various nerves. When present, this results in slow conduction velocities and prolonged distal latencies without dispersion or conduction block. The distinction between uniform and nonuniform demyelination is diagnostically helpful because uniform involvement is typical of a subset of inherited disorders, while nonuniform involvement typifies acquired demyelinating neuropathies as well as some inherited ones.74,75 This rule is useful but by no means absolute and should be applied cautiously. For example, pseudoconduction blocks may occur in inherited neuropathies as a result of increased thresholds requiring very large stimuli, or as a result of phase cancellation.75 Examples of inherited disorders in which nonuniform demyelination occurs are HNPP, X-linked CMT disease, adrenomyeloneuropathy, and Refsum’s disease.75

Temporal Profile The acuity of onset and rate of progression are often helpful clues. Most causes of polyneuropathy result in a chronic, slowly progressive process that worsens over years. A smaller number of conditions result in an acute or subacute process, or in a relapsing and remitting clinical course. Acute onset with rapid progression over days to weeks occurs in Guillain-Barré syndrome and in some examples of toxic exposure (e.g., heavy metals), inflammatory sensory ganglionopathies, inflammatory brachial and lumbar plexopathies, necrotizing vasculitis, and porphyria.118 It may also occur as an infrequent presentation of diabetic neuropathy, particularly following initiation of insulin therapy. Acute neuropathy may also occur spontaneously in diabetic individuals; pain is the most prominent symptom, often accompanied by weight loss and depression.5,33 Objective clinical and electrophysiologic abnormalities may be slight, and the pain may improve with institution of tight glycemic control.

A relapsing and remitting course is common in CIDP. Other causes include porphyria, Tangier disease,43 and Refsum’s disease.17,114

ASSOCIATED SYMPTOMS AND SIGNS OF DISEASE Non-neurologic Disease Evidence of a relevant disease capable of causing neuropathy will most often emerge from a careful history and review of systems. It may be already documented by the time the neuropathy is found or only come to light when the cause of neuropathy is pursued. The physical examination and adjunctive tests may also provide useful information. Given the large number of diseases to be considered, the number and kind of tests to be performed should be guided entirely by the clinical picture. The list of potentially relevant diseases is enormous, and virtually every organ system can be involved. Constitutional symptoms such as weight loss, malaise, and fever suggest the possibility of underlying malignancy (carcinoma or lymphoma), infection, or amyloidosis. Weight loss should also raise the possibility of unrecognized diabetes. Weight gain may herald hypothyroidism. Lymphadenopathy may be found in lymphoma, metastatic carcinoma, and the POEMS syndrome (polyneuropathy, organomegaly, endocrinopathy, M protein, and skin changes) that accompanies some cases of osteosclerotic myeloma.8 Gastrointestinal symptoms such as early satiety, cramps, diarrhea, and constipation may also provide a useful clue. Gastrointestinal symptoms, particularly abdominal pain, are common in heavy metal toxicity. Attacks of acute, colicky pain are typical of acute intermittent porphyria. Symptoms of gastroparesis may occur with diabetic autonomic neuropathy and with sensory ganglionopathies, particularly of the paraneoplastic variety. Polyneuropathy may develop following gastroplasty for morbid obesity, perhaps a result of vitamin B12 or thiamine deficiency, although the mechanism remains uncertain.1,84 Individuals with inflammatory bowel disease may develop a neuropathy, the type depending on the kind of colitis; Guillain-Barré syndrome may occur in patients with ulcerative colitis,80 while patients with Crohn’s disease sometimes develop an axonal sensorimotor polyneuropathy.34 Rarely, neuropathy has been described in patients with celiac disease, in whom symptoms of malabsorption are common.62 Inspection of the oral cavity may demonstrate lead lines along the border of the gum, macroglossia in some individuals with amyloidosis, or the often sought but rarely encountered orange tonsils of Tangier disease. A long list of skin diseases may relate to neuropathy as well as other neurologic diseases, as has been recently

1168

Diseases of the Peripheral Nervous System

reviewed by Hurko and Provost.57 Palpable purpura or ulcers may occur in the setting of necrotizing vasculitis. Raynaud’s phenomenon suggests Sjögren’s syndrome or cryoglobulinemia. Arsenic toxicity produces hyperkeratosis, erythema, sloughing on the palms and soles, and hyperpigmentation or depigmentation of the skin. Hyperpigmentation also occurs in POEMS syndrome and adrenoleukodystrophy. Angiokeratoma corporis diffusum is the characteristic skin lesion of Fabry’s disease.88 These appear as multiple punctate red lesions over the buttocks, groin, and genitalia. Nodular lesions or anesthetic patches occur in leprosy. Cutaneous photosensitivity is a hallmark of systemic lupus erythematosus. Subcutaneous lipomas are a rare feature of mitochondrial diseases in which neuropathy may coexist (Ekbom syndrome). Other examples include erythema chronicum migrans in Lyme disease, and a variety of skin changes (including erythema nodosum) in sarcoidosis. Inspection of the nails may demonstrate Mees’ lines, the transverse white or gray striae60,113 that occur secondary to either arsenic or thallium poisoning. The hair loss of thallium toxicity typically occurs 2 to 3 weeks after an acute exposure. Ocular manifestations of disease may be evident to the neurologist, but formal ophthalmologic consultation with slit-lamp examination or indirect ophthalmoscopy may be needed to identify subtle or unusual changes. In diabetics, the presence of retinopathy is potentially significant; given the association between neuropathy and other microvascular complications (retinopathy and nephropathy), the probability of neuropathy being caused by diabetes is increased when retinopathy is also present.29,36,108 Pigmentary retinopathy occurs in Refsum’s disease,128 abetalipoproteinemia, some forms of inherited spinocerebellar ataxia, hereditary motor and sensory neuropathy type VI, and mitochondrial cytopathies such as the neuropathy, ataxia, and retinitis pigmentosa (NARP) syndrome.55 Lattice corneal dystrophy is a feature of familial amyloid polyneuropathy type IV caused by gelsolin mutations.2 Hearing loss has long been known to occur in some patients with nutritional deficiency and Refsum’s disease. It is also seen in some patients with mitochondrial disease,15 in one form of hereditary sensory neuropathy,147 and in several types of CMT disease, including X-linked forms,71 CMT1 families with EGR2 mutations106 and PMP22 mutations,11,65 and one family with axonal CMT caused by a P0 mutation.91 Renal dysfunction may occur in patients with Fabry’s disease, diabetes, or mitochondrial disease. Many patients with neuropathy will have musculoskeletal deformities of the feet, such as pes cavus or hammer toes (Fig. 46–1). Pes cavus probably reflects muscle weakness and imbalance beginning early in life prior to completion of bony development. It is often taken as evidence

FIGURE 46–1 Pes cavus in a 20-year-old man with Charcot-Marie-Tooth disease type 1A. The patient has high plantar arches with hammer toes and atrophy of foot muscles. (From Benstead T. J., and Grant, I. A.: Charcot-Marie-Tooth disease and related inherited peripheral neuropathies. Can. J. Neurol. Sci. 28:199, 2001, with permission.)

of an inherited neuropathy such as CMT disease. While this is often the case, pes cavus should be interpreted carefully because it may result from several other disorders, including Friedreich’s ataxia, old polio, and central nervous system diseases such as cerebral palsy, spinal dysraphism, multiple sclerosis, and spinal cord injury. In the patient with an acute neuropathy, coexisting neuropsychological changes such as delirium should suggest heavy metal toxicity or porphyria. In patients with known or newly discovered systemic disease, neuropathy may occur for more than one reason. A common example would be the occurrence of polyneuropathy in a patient with known cancer. In this situation, the cause of neuropathy may be a paraneoplastic process (particularly in the setting of small cell lung cancer), drug toxicity (as in patients receiving vincristine or cis-platinum), amyloidosis, or nutritional deficiencies in the chronically ill, cachetic patient. In the same patient, a focal neuropathy may relate to compression or infiltration of nerve by metastatic tumor in the cauda equina or brachial plexus. Table 46–3 lists other examples of diseases in which neuropathy can occur for multiple reasons. In patients with cancer, a neuropathy may arise before or after the diagnosis of the malignancy. In patients presenting with idiopathic polyneuropathy, a detailed work-up for cancer is not routinely justified. Aggressive testing should be reserved for those in whom either the neuropathic picture (such as a subacute sensory neuropathy) or associated symptoms (weight loss, cough, hemoptysis, etc.) raise the specter of cancer. Camerlingo et al.16 followed 51 patients with undiagnosed sensory neuropathies for a mean period of

Differential Diagnosis of Polyneuropathy

Table 46–3. Causes of Neuropathy in the Setting of Selected Systemic Disease Rheumatoid Arthritis Vasculitis Drug toxicity Gold Hydroxychloroquine Compression Cancer Paraneoplastic (anti-Hu antibody) Drug toxicity Vincristine cis-Platinum Cytosine arabinoside Paclitaxel Suramin Amyloidosis Nutritional deficiency Leptomeningeal metastases causing polyradiculopathy Renal Failure Uremia Diabetes Fabry’s disease Mitochondrial disease Sicca Complex Sjögren’s syndrome Sarcoidosis Autonomic neuropathy HIV Infection Direct effect of HIV Drug toxicity ddI ddC Vasculitis Inflammatory demyelination Guillain-Barré syndrome CIDP CIDP ⫽ chronic inflammatory demyelinating polyneuropathy; ddC ⫽ dideoxycitidine; ddI ⫽ didanosine; HIV ⫽ human immunodeficiency virus.

51 months. A malignancy was eventually found in 35% of patients, following a mean interval of 27 months after the diagnosis of neuropathy. Most of the patients with cancer were male, with a mean age of 64 years entering the study. The most common tumor was hepatoma (four patients); because this tumor and several others identified in the cohort are not known to be specifically associated with neuropathy, their significance is doubtful. Several potentially relevant malignancies did develop, including non-Hodgkin’s lymphoma, small cell lung cancer, and breast cancer. Although the mechanism was not explored in detail, it was

1169

noted that nine patients had a monoclonal gammopathy. Serology for paraneoplastic antibodies was performed in all patients, although only one (with small cell lung cancer) was positive. The risk of developing cancer was higher in this group of neuropathy patients than in the general population. However, the evidence is insufficient to recommend longterm cancer surveillance in all individuals with undiagnosed neuropathy. Diabetic peripheral neuropathy is very common, but a judgment still needs to be made, once neuropathy is detected in a patient with diabetes, as to whether the diabetes is the cause. We know that neuropathy is unlikely to develop in patients with type 1 diabetes until the patient has had diabetes for several years and the diagnosis is well known.35 The difficulty lies in patients with undetected type 2 diabetes. When a patient undergoes investigation for neuropathy and type 2 diabetes is detected, several questions need to be addressed: 1. Has the diabetes been present long enough to produce neuropathy? 2. Is the glucose intolerance severe enough to produce neuropathy? 3. Is an alternative cause for the neuropathy present? Given the positive relationship between the prevalence of diabetic polyneuropathy and the duration and severity of hyperglycemia, one would expect symptomatic neuropathy to be uncommon in patients in whom diabetes is not known to be present. Yet it is frequently impossible to determine the duration of type 2 diabetes, and undetected diabetes may be present long enough to produce a complication such as neuropathy. Even mild glucose intolerance not fulfilling diabetes criteria may be significant. Novella et al.102 evaluated 76 patients with distal symmetrical neuropathy. In 48 patients, the cause of neuropathy was not evident from history or blood tests. These patients underwent fasting blood glucose and 2-hour oral glucose tolerance testing using a 75-g glucose load. Using American Diabetes Association criteria,116 50% of this group had type 2 diabetes (23%) or impaired glucose intolerance (27%). Abnormal glucose metabolism was much more frequent in the subgroup of patients with painful neuropathy. Singleton et al.127 also found a higher than anticipated incidence of unexpected impaired glucose metabolism in patients with idiopathic painful neuropathy. These results suggest that relatively mild glucose intolerance may be associated with neuropathy, and thorough diabetes evaluation should be considered in patients with cryptogenic polyneuropathy, particularly if it is painful. Currently available data have not drawn a clear causal relationship between mildly impaired glucose metabolism and neuropathy, but there is a potential relationship. Fasting glucose and perhaps an oral glucose tolerance test are therefore reasonable screening studies in patients with undiagnosed neuropathy. When diabetes is known or newly

1170

Diseases of the Peripheral Nervous System

detected in patients with diffuse neuropathy, it is still important to screen for other potential causes of the neuropathy, though the search need not be as exhaustive as in the patient without an obvious cause.

Neurologic Disease Symptoms and signs of neurologic disease apart from neuropathy may also point to a diagnosis. Tremor is a nonspecific finding in many patients with neuropathy but may occur in a familial pattern in patients with RoussyLévy syndrome. Upper motor neuron findings such as spasticity, hyperreflexia, and extensor plantar responses may accompany vitamin B12 deficiency, hereditary motor and sensory neuropathy type V, complicated hereditary spastic paraplegia, and leukodystrophies. Myokymia is characteristic of gold toxicity. The term neuromyopathy

Table 46–4. Associated Neurologic Findings That May Be Diagnostically Significant in Patients with Polyneuropathy Upper Motor Neuron Signs Vitamin B12 deficiency Charcot-Marie-Tooth disease subtypes (HMSN type V, Troyer syndrome) Multiple system atrophy Hereditary spastic paraplegia (complicated forms) Leukodystrophies Adrenoleukodystrophy/adrenomyeloneuropathy Metachromatic leukodystrophy Myopathy Thyroid disease Sarcoidosis Ethanol Critical illness Connective tissue disease Vasculitis Mitochondrial disease Amyloidosis Merosin (␣2-laminin) deficiency Parkinsonism Multiple system atrophy Spinocerebellar ataxia type 2 Seizures Mitochondrial disease Tremor Roussy-Lévy syndrome Nonspecific manifestation of neuropathy Myokymia Gold toxicity HMSN ⫽ hereditary motor and sensory neuropathy.

has been applied to patients in whom nerve and muscle disease coexist. While the association may be coincidental, several diseases affect both nerve and muscle. The list includes thyroid disease, ethanol, drugs, connective tissue disease (CTD), and mitochondrial dysfunction. Other examples of clues from the neurologic examination are given in Table 46–4.

FAMILY HISTORY As mentioned earlier, among patients with “idiopathic” neuropathy in whom a cause is eventually found, inherited disorders comprise the largest group. In Dyck et al.’s series, 42% of patients with undiagnosed neuropathies ultimately proved to have an inherited condition.31 The most important factor here is probably simple failure to consider the possibility of inherited disease. Even if the possibility is entertained, the family history will often be negative on initial questioning of a new patient. Potential reasons for this include lack of knowledge of affected kin, low penetrance such that affected family members do not show signs of disease, false paternity, or attribution of neuropathic symptoms to other kinds of disease such as arthritis or old polio. Consideration of sensory symptoms is worthwhile because these differ somewhat between inherited and acquired neuropathies when considered collectively. Positive sensory symptoms such as prickling paresthesias are more commonly reported by patients with acquired neuropathies than by those with inherited ones.31 The patient should be asked to collect additional family history; subsequently, information pointing to a familial problem may emerge. The patient will sometimes be aware of pes cavus in family members without recognizing the significance of this finding. In other cases, despite the patient’s best efforts, the nature of the problem remains unknown until other family members are assessed. A simple telephone interview can be a good starting point, but examination, EMG, and genetic testing may be required for a definitive judgment. With the advent of DNA testing for some inherited neuropathies, a specific diagnosis can sometimes be made in the complete absence of a family history. The most common example would be CMT disease type 1A, caused by a chromosome 17p11.2 duplication.

LABORATORY INVESTIGATIONS Screening Tests Neuromuscular specialists generally have a set of laboratory investigations they commonly employ to evaluate patients with diffuse neuropathy, but there is no universally accepted list of investigations, and the choice undoubtedly varies from clinician to clinician. As well, neuromuscular

Differential Diagnosis of Polyneuropathy

specialists become good at pattern recognition and appropriately tailor the investigation plan to the patient’s specific situation. This is more difficult for clinicians with less expertise in the evaluation of neuropathy. Multiple factors contribute to the difficulty of knowing which investigations are necessary for a particular patient and when to stop investigating: 1. The differential diagnosis of diffuse neuropathy is extremely long. 2. The list of available tests for peripheral neuropathy is continually increasing. 3. Many tests are expensive. 4. Many tests are only relevant for patients with a very specific subtype of neuropathy, and therefore have a low yield in an undifferentiated patient population. 5. The diagnostic utility of many commonly performed tests has not been firmly established through welldesigned studies. 6. Many disorders producing neuropathy have no specific treatment, thus nihilism may cause some clinicians to underinvestigate. 7. In even the most expert hands, there is a substantial number of patients for whom a cause of neuropathy cannot be determined. With these factors in mind, the clinician must decide for each patient how far to pursue investigation. Several authors have dealt with this issue.31,47,121,132,146 Drawing on the experience of knowledgeable investigators, it is possible to compile a list of investigations that are likely to detect or exclude the vast majority of diagnostic considerations for the patient with peripheral neuropathy. For example, investigations commonly performed at a clinic recognized for expertise in polyneuropathy diagnosis include the following132: Complete blood count (CBC) Chemistry profiles, including thyroid-stimulating hormone Serum and urine immunoelectrophoresis Immunologic markers, including erythrocyte sedimentation rate (ESR), rheumatoid factor, antinuclear antibodies (ANA), extractable nuclear antibodies (ENA), antineutrophil cytoplasmic antibodies (ANCA), and cryoglobulins Paraneoplastic antibodies ANNA-1 (anti-Hu) and ANNA-2 (Purkinje cell antibody) Chest radiograph Metastatic bone survey Serology for hepatitis B and C, human immunodeficiency virus (HIV), Lyme disease, and syphilis Genetic markers Cerebrospinal fluid (CSF) analysis Electrophysiology Nerve biopsy Quantitative sensation test and autonomic reflex screen Magnetic resonance imaging (MRI)

1171

The authors found that both diagnostic accuracy and certainty rose with increasing numbers of diagnostic tests. Evaluation by this group of neuromuscular experts resulted in a final specific diagnosis in 80% of patients. The authors noted that the greatest single factor in diagnostic accuracy is the medical history taken by an expert in peripheral neuropathy. This observation highlights the importance of clinical experience in the diagnosis of neuropathy. In general, a comprehensive blanket list of evaluations should only be applied to patients without an apparent cause of neuropathy after an experienced examiner has evaluated the patient. Investigations often need only be performed to confirm a clinical suspicion or aid in subtyping a process. For instance, a patient with a predominantly demyelinating sensorimotor peripheral neuropathy who has an autosomal dominant family history of neuropathy does not need a long battery of diagnostic investigations. However, selected genetic tests to clarify the specific genotype of CMT disease would be indicated. In all patients, specific chemistry tests should include liver and renal function tests, electrolytes, and glucose levels. Selected patients will benefit from lipoprotein electrophoresis and measures of vitamin E levels, very-long-chain fatty acids, and serum lactate. Some authors have found that intensive investigation most often increases diagnostic yield in patients with inherited or immune-related neuropathies.31,47 For inherited demyelinating neuropathy, genetic markers such as PMP22 duplication or deletion, P0 point mutation, and connexin-32 point mutation can provide a specific diagnosis. In patients with the appropriate clinical features, genetic tests for Friedreich’s ataxia, other spinocerebellar ataxias, spinal muscular atrophy, and mitochondrial disorders should be considered. Not all longitudinal studies have demonstrated a diagnostic benefit from repeated testing of patients with neuropathy. One group of 40 consecutive patients evaluated for chronic sensorimotor polyneuropathy of undetermined cause was followed prospectively with repeated testing for etiology.59 These patients had disease for at least a year before enrollment and all had axonal neuropathies. Despite undergoing serial laboratory testing, none of the patients had an etiologic diagnosis made over a minimum of 4 years’ follow-up. The laboratory tests performed included CBC, ESR, serum glucose, hemoglobin A1c, kidney and liver tests, serum protein electrophoresis, thyroid function tests, vitamin B12, ANA (including anti–SS-A and anti–SS-B), rheumatoid factor, Lyme antibody, HIV antibody, hepatitis C serology, syphilis serology, and urine protein electrophoresis. Patients with a smoking history had a chest radiograph and anti-Hu serology. The authors stated that they went to great lengths to exclude from the study patients with monoclonal gammopathy, substance abuse, and any suggestion of hereditary neuropathy. Their work would suggest that carefully screened and selected patients

1172

Diseases of the Peripheral Nervous System

with a chronic axonal sensorimotor neuropathy can be evaluated once and that repeated follow-up investigations are not worthwhile. However, this contrasts with other longitudinal studies16,132 that suggest reevaluation does increase the yield of new diagnosis. The variability in results of various studies may reflect differences in patient selection and the battery of investigations performed. Further study is necessary to clearly outline the optimal course to take with patients presenting with diffuse neuropathy when a diagnosis is not apparent after initial investigation. Some frequently performed investigations deserve special comment. Diabetes screening is likely performed universally in patients with peripheral neuropathy. As noted in a previous section, relatively mild glucose intolerance may be associated with neuropathy, and thorough diabetes evaluation should be considered in patients with cryptogenic polyneuropathy, particularly if it is painful. Fasting and 2-hour postmeal glucose measurements are justified in screening for unknown type 2 diabetes. The question of whether an oral glucose tolerance test is justified requires further study. Thyroid disease is another common medical problem that can lead to neuropathy. In a prospective study of patients with thyroid disease, signs of a sensorimotor axonal neuropathy were present in 42% of patients with hypothyroidism and 19% of patients with hyperthyroidism.25 Symptoms and signs of thyroid disease should be sought when evaluating neuropathy patients, and screening tests should be performed. Because of the treatable nature of thyroid disease, it is important to exclude this diagnosis early in the investigation. Another ubiquitous test is serum vitamin B12 measurement. This is readily available and, when well above the lower limit of normal, likely excludes cobalamin deficiency as the mechanism of neuropathy. In fact, most patients with neurologic manifestations of cobalamin deficiency will have myelopathy, myelopathy in combination with diffuse neuropathy, neuropsychiatric manifestations, or optic neuropathy. Peripheral neuropathy is a rare isolated manifestation. In patients with low-normal vitamin B12 levels, it may be valuable to also measure serum methylmalonic acid (MMA) levels. Serum MMA levels are elevated in some patients with low-normal serum vitamin B12 levels79 and neurologic disease responsive to vitamin B12 treatment.19,78,110 Routine screening for heavy metals is not advisable, but an environmental and occupational history is important to assess risk for heavy metal exposure. In patients with the appropriate exposure history, a heavy metal assessment can be important diagnostically.67 However, caution should be exercised in interpreting blood lead levels because lowlevel exposure is unlikely to explain a significant clinical neuropathy.32 There are many infectious causes of peripheral neuropathy, including syphilis, Lyme disease, HIV, hepatitis, and leprosy. Routine screening for all infectious causes of

neuropathy will not be fruitful. A reasonable knowledge of the epidemiology of these infectious processes is necessary to target the proper population with the appropriate diagnostic tests. For Lyme disease and leprosy, the region of origin of the patient and the nature of the neuropathy will determine whether these diagnoses should be considered. A history of potential exposure can direct the clinician toward appropriate investigation of communicable diseases. Some tests that are regarded as routine, such as syphilis serology, have an extremely low yield and may be best reserved for high-risk patients. Results of other investigations will sometimes suggest an infectious etiology for neuropathy. For example, with elevated liver enzymes, it would be appropriate to screen for hepatitis B and C. With CSF pleocytosis, HIV should be considered. Connective tissue diseases of many types may produce peripheral neuropathy, and usually the underlying CTD is apparent from systemic manifestations other than the neuropathy. When the manifestations of a specific CTD are present, then the investigation is usually straightforward, though it is often still important to establish through nerve biopsy whether the neuropathy is vasculitic or not. CTDs can produce a variety of neuropathy types, including mononeuritis multiplex, diffuse sensorimotor neuropathy, sensory ganglionopathy, and focal neuropathies. When CTDs are sought in patients presenting with multiple mononeuropathies, about 40% are found to have a rheumatologic disorder.53 Some disorders, such as rheumatoid arthritis, are common and cause associated peripheral neuropathy in a large percentage of patients. Up to two thirds of patients with advanced rheumatoid arthritis will have mild sensorimotor peripheral neuropathy,44 while vasculitic neuropathy is much less comon106; rheumatoid factor is positive in 80%. Sjögren’s syndrome and sicca complex produce a variety of neuropathy types, including sensory ganglionopathy, small fiber neuropathy, and sensorimotor neuropathy.48 Patients with Sjögren’s syndrome will sometimes have a positive ANA, rheumatoid factor, or ENA (anti-Ro [anti–SS-A] or anti-La [anti–SS-B]), but these autoantibodies are usually absent in patients in whom neuropathy is the presenting manifestation of their Sjögren’s syndrome.48 Diagnostic confirmation often requires demonstration of xerophthalmia through the Schirmer test or rose bengal staining and the demonstration of lymphocytic infiltrates in a minor salivary gland biopsy. Diffuse neuropathy occurs much less frequently in systemic sclerosis,72 though trigeminal sensory neuropathy will occur. Trigeminal sensory neuropathy can be associated with a number of CTDs, in particular mixed connective tissue disease, Sjögren’s syndrome, systemic sclerosis, and systemic lupus erythematosus.7,39,70 Screening tests for CTDs include ANA, ENA, antiDNA, ANCA, cryoglobulins, ESR, C-reactive protein, and serum complement. ANA are sensitive but nonspecific markers of CTD and are often used as a screening test; low

1173

Differential Diagnosis of Polyneuropathy

titers (less than 1:160) often have no clinical significance.94 ANCA are sensitive and specific for Wegener’s granulomatosis when present with a cytoplasmic staining pattern (c-ANCA). As a screening test for vasculitis in patients with peripheral neuropathy, ANCA have low utility because of a high number of false positives.18

Monoclonal Gammopathy and Antiglycolipid Antibodies Serologic abnormalities in patients with neuropathy can be divided into several types: monoclonal gammopathy (M protein), identified by serum protein electrophoresis and immunofixation; antibodies against non–organ-specific antigens, such as ANA (discussed above); and antibodies against specific neural antigens, often glycolipid components of the axolemma, the myelin sheath, or both. These groups are not mutually exclusive; many patients with a “generic” M protein can be shown to have activity against one or more specific antigens if appropriate tests are performed. Some antibodies are associated with particular clinical syndromes. The spectrum of neuropathy associated with anti-nerve antibodies includes both axonal and demyelinating polyneuropathies, CIDP, Guillain-Barré syndrome and its variants (including acute motor axonal neuropathy and Miller Fisher syndrome), paraneoplastic sensory polyganglionopathies, and ataxic sensory neuropathies. Serum protein electrophoresis and immunoelectrophoresis should be performed in all patients with neuropathy of unknown cause. Both are widely available and relatively inexpensive. Although a causal relationship between M protein and neuropathy cannot automatically be assumed in all cases, the presence of an M protein should at least suggest an autoimmune process, thus reducing or eliminating the need to test for other etiologies. In most patients with monoclonal gammopathy and neuropathy, the M protein is an isolated finding unassociated with evidence of a myeloproliferative disease; these patients are characterized as having monoclonal gammopathy of undetermined significance (MGUS). However, one will occasionally discover multiple myeloma, lymphoma, or other disease requiring treatment. The majority of MGUS patients with neuropathy have an M protein of the immunoglobulin (Ig) M class. In comparison to patients with less common IgG- and IgA-related neuropathies, patients with IgM-related neuropathies tend to have more severe large fiber sensory deficits, more electrophysiologic evidence of demyelination, and less satisfactory response to immunomodulatory therapy such as intravenous immune globulin.46,101,133 About 50% of patients with IgM-associated neuropathy have circulating antibodies to myelin-associated glycoprotein (MAG)63,98 and /or other glycoconjugates. Having shown that a patient with neuropathy has a monoclonal

gammopathy, one might ask whether the additional information obtained from anti-MAG testing is useful. Anti-MAG–positive IgM MGUS patients are clinically indistinguishable from anti-MAG–negative patients.46,126,133 The presence of anti-MAG antibodies has been associated in some studies with an electrophysiologic pattern of distal accentuation of nerve conduction slowing not present in patients with IgM gammopathies without anti-MAG,82 CIDP,61,82 and CMT1A,61 but this finding has not been seen consistently.46,133 It has been argued that the presence of anti-MAG increases the likelihood that the M protein is causally associated with neuropathy.100 However, in the absence of other causes of neuropathy, most clinicians would consider IgM MGUS clinically significant in the setting of a sensorimotor or predominantly sensory neuropathy, particularly one with demyelinating features. Anti-MAG antibody has been found in the absence of a monoclonal gammopathy on immunoelectrophoresis,42 but this is probably rare. Thus the additional information provided by anti-MAG testing is of limited diagnostic value. Screening serologic testing for specific anti-nerve antibodies cannot be recommended for all patients with neuropathy. However, there is clear diagnostic utility in testing selected patients. In individuals with sensory neuropathy, the presence of anti-Hu antibody implies a paraneoplastic process and necessitates a thorough search for an underlying malignancy. The diagnosis of Miller Fisher syndrome is not always straightforward, particularly in individuals with atypical features or incomplete expression of the classic triad of ophthalmoplegia, ataxia, and areflexia, which may be incorrectly attributed to a brainstem lesion. The high sensitivity and specificity of anti-GQ1b antibodies in the situation is very useful in confirming the correct diagnosis. Other commercially available antibodies are less useful. Neuropathic syndromes associated with available antibody tests and the diagnostic utility of these tests are summarized in Table 46–5.

Nerve Biopsy The role of nerve biopsy in the investigation of neuropathy has changed with time. Advances in electrodiagnostic testing, quantitative sensory testing, and autonomic testing provide less invasive means of obtaining much of the information about the presence and type of nerve fiber pathology that could once be obtained only by biopsy. The correlation between electrophysiology and nerve fiber pathology is good, and in most cases biopsy is not needed to determine whether the predominant nerve fiber process is axonal or demyelinating. Alteration of small myelinated and unmyelinated nerve fibers can be evaluated with quantitative sensation testing and autonomic testing, although these techniques are unavailable to many clinicians. Specific serologic tests, genetic tests, and well-developed diagnostic criteria that do not require

1174

Diseases of the Peripheral Nervous System

Table 46–5. Anti-nerve Antibody Tests Antibody MAG/SGPG

Clinical Features

Demyelinating sensorimotor or sensory polyneuropathy with prominent large fiber deficits Sulfatide Predominantly sensory polyneuropathy GM-1 Multifocal motor neuropathy with conduction block Guillain-Barré syndrome GD1b Sensory neuropathy GQ1b Miller Fisher syndrome Hu (ANNA-1) Paraneoplastic sensory neuropathy associated with SCCA

Diagnostic Utility Limited

Low Limited (may be better with new sensitive assays) Limited Unknown Good Good, particularly in the absence of known cancer

ANNA ⫽ antineuronal nuclear antibody; MAG ⫽ myelin-associated glycoprotein; SCCA ⫽ small cell carcinoma; SGPG ⫽ sulfated glucuronyl paragloboside.

nerve biopsy facilitate a noninvasive diagnostic approach to many conditions. However, despite advances, biopsy remains the only way to obtain certain kinds of information about nerve fibers, blood vessels, and interstitium. Nerve fiber abnormalities that may be found include the following: 1. Fiber loss. The biopsy demonstrates the severity of fiber loss, which can be quantified using morphometric techniques. The distribution of fiber loss may be multifocal (characteristic of ischemia, as with vasculitis) or diffuse. The affected fiber classes (i.e., large myelinated, small myelinated, and unmyelinated) can also be identified and quantified. 2. Axonal degeneration. As with fiber loss, this may be multifocal or diffuse. The amount of active axonal degeneration provides information about whether the underlying disease is aggressive or indolent. 3. Axonal regeneration. 4. Axonal atrophy. 5. Demyelination (see Table 46–2) and remyelination. The presence of onion bulb formations implies a chronic demyelinating neuropathy of either inherited or acquired cause. 6. Myelin reduplication. This finding is most characteristic of HNPP. Diseases that affect interstitial structures are particularly likely to require biopsy for diagnosis. These include amyloidosis, vasculitis, leprosy, sarcoidosis, Tangier disease, and giant axonal neuropathy. In addition to vasculitis, a variety of immune-mediated disorders may show evidence of inflammation; these include inflammatory demyelinating

neuropathies, sensory ganglionopathies, and perineuritis. Neuropathy diagnoses established by nerve biopsy commonly directly affect patient management decisions.6,41 The potential for making through nerve biopsy a diagnosis that cannot be made through other means is one important consideration when planning investigations for a patient with neuropathy. Another consideration is the potential morbidity from the biopsy procedure. Numbness and biopsy-induced pain are a disincentive to performing biopsy.105 Before a biopsy is performed, the clinician should consider two questions: (1) Is a neuropathy diagnosis possible for which a nerve biopsy may be informative? and (2) Are there other available diagnostic procedures with less potential morbidity that may provide the diagnosis? An example of the latter would be biopsy of other tissues in vasculitis and sarcoidosis. If a nerve biopsy is undertaken, there are several important points to consider regarding the choice of nerve, biopsy technique, and experience of the processing laboratory. These are reviewed elsewhere in this text. Nerve biopsy may be particularly helpful when the neuropathy is markedly asymmetrical or produces a pattern of multiple mononeuropathies. Nerve vessel abnormalities and sarcoidosis may be more likely in these patients.26,123 However, multiple mononeuropathies in typical areas of nerve compression are unlikely to be the result of an acquired generalized neuropathy and should prompt genetic testing for the PMP22 deletion associated with HNPP.137,141 If vasculitic neuropathy is suspected, tissue diagnosis can sometimes be obtained through biopsy of tissue with less potential for morbidity, such as muscle,124 but vasculitis may be restricted to nerve without systemic involvement.26,64

Other Investigations Several ancillary investigations should be considered for the evaluation of a patient with neuropathy. These include MRI of the spine, chest radiography, computed tomography of the chest and abdomen, skeletal survey, CSF analysis, and nerve biopsy. MRI scan of the spine is valuable when symptoms or EMG suggest multiple radiculopathies or polyradiculopathy, which can mimic a diffuse neuropathy. Patients with a smoking history or chest symptoms should have a chest radiograph. Primary lung cancers and sarcoidosis can be detected this way. An unexplained progressive neuropathy presenting in an older patient should prompt an evaluation for underlying malignancy, which may produce neuropathy as a paraneoplastic syndrome, through direct invasion of meninges, producing carcinomatous/lymphomatous meningitis, or through direct invasion of nerve trunks. Patients with a serum monoclonal gammopathy should have urine immunoelectrophoresis looking for Bence Jones protein and a skeletal survey looking for a solitary

1175

Differential Diagnosis of Polyneuropathy

plasmacytoma. CSF analysis is most helpful for detecting processes that invade or infect the meninges. As well, elevated CSF protein with normal CSF cell count may be found in the inflammatory neuropathies.

DISORDERS SIMULATING PERIPHERAL NEUROPATHY Painful Legs and Moving Toes The syndrome of painful legs and moving toes was described by Spillane et al. in 1971.129 This rare syndrome is characterized by burning or aching pain in the feet and, in some patients, the legs. The pain may be very severe; it is chronic and often unilateral. There is a distinctive associated involuntary movement of the toes, often wriggling in nature, and in some patients there is movement of more proximal leg muscles also. These movements can be inhibited by voluntary effort, but only briefly, and are not abolished by sleep. Pain typically precedes the onset of the toe movements by days to years. Similar symptoms have been described in the upper limbs.142 The cause is unknown. While some cases are idiopathic, a number of associated diseases have been described, including polyneuropathy,95 trauma,24,95,97 herpes zoster infection,97 and HIV infection.109 Many patients have symptoms or radiologic evidence of radiculopathy.97 EMG studies show rhythmic bursts of motor unit activity in distal leg muscles.95 The syndrome has been suggested to arise from spontaneous impulse generation in afferent fibers, often in proximal nerve segments such as the dorsal root, leading to reflex activation of local motor neurons.97 A spinal generator for the movements is suggested by the occurrence of cases in which reciprocal firing of flexors and extensors occurs148 and in cases with synchronous movements of upper and lower limb muscles.95 It has been suggested that the pain is sympathetically mediated; Spillane et al.129 reported that lumbar sympathetic block produced “great diminution or abolition of the movements and pain” temporarily. However, other authors have not found sympathetic block to be effective.97 Treatment is generally unsatisfactory. Anticonvulsants, tricyclic antidepressants, baclofen, benzodiazepines, and other drugs have been tried with limited success.

Erythromelalgia Weir Mitchell first described this distinctive syndrome in 1878.92 Consisting of a triad of pain, warmth, and redness of the extremities, it is rare and chronic. Erythermalgia and erythralgia are alternative terms that have been proposed. This disorder most often affects white females.20 Symptoms are usually intermittent. The feet are affected much more often than the hands. Characteristically, symp-

toms are precipitated by warming of the limb and relieved by cooling; patients may learn to immerse their legs in cold water for relief. Some patients are responsive to aspirin. Aspirin responsiveness has been proposed as a basis for classification; Drenth and Michiels divided aspirin-responsive cases—typically associated with thrombocytosis—from non–aspirin-responsive cases, which they termed erythermalgia.23 The cause of this syndrome continues to be debated. The most fundamental issue is whether the primary process involves blood vessels (i.e., altered vascular dynamics) or nerves. Layzer69 has suggested that erythromelalgia results from sensitization of C nociceptors such that they generate pain and vasodilatation at skin temperatures below which they would normally become active. Erythromelalgia may be a primary or a secondary condition. The primary form may occur in isolation or be familial. Secondary cases have been described in association with drugs (bromocriptine, pergolide, and calcium channel blockers), systemic lupus erythematosus, Raynaud’s phenomenon, vasculitis, and pernicious anemia. An association with myeloproliferative disorders—especially thrombocytosis—is often observed.20,58,66,89 Although most patients do not have evidence of peripheral neuropathy, erythromelalgia has been described in one patient with hereditary sensory neuropathy54 and another with acute diabetic neuropathy.140

Myopathy The classic proximal pattern of weakness caused by most myopathies will seldom be confused with neuropathy, although this pattern can occur in polyradiculoneuropathies such as Guillain-Barré syndrome. Confusion is more likely in patients with distal myopathies, a large and heterogeneous group of diseases whose classification continues to evolve and for which major advancements in molecular genetics have been made in recent years. The distal myopathies proper are essentially a group of muscular dystrophies with shared clinical features. Most are inherited in an autosomal dominant fashion, but there are rare recessively inherited types. Well-defined examples of distal myopathy include Welander’s myopathy, tibial muscular dystrophy, late-onset distal myopathy, Miyoshi myopathy, Nonaka myopathy, and desmin storage disorders.83,90,93,99,139,143,145 Many other muscle diseases affect distal muscles in a less selective way, including myotonic dystrophy, inclusion body myositis, facioscapulohumeral dystrophy, and the scapuloperoneal syndromes. Differentiation from polyneuropathy is based on the presence of weakness without sensory loss, and relative sparing of deep tendon reflexes. However, a kindred with areflexia has been described, and some patients with Welander’s myopathy experience sensory symptoms; these families probably have mild neuropathy in addition to

1176

Diseases of the Peripheral Nervous System

muscle disease.125,145 Muscle involvement may be quite selective, for example, affecting the posterior compartment of the leg with sparing of the anterior tibial group, as occurs in Miyoshi myopathy.93 Difficulty may arise when, as more often occurs, the anterior group is most affected, simulating the peroneal muscular atrophy phenotype. Some individuals have pes cavus.125 Electrodiagnostic studies will usually resolve any uncertainty, showing normal sensory conduction and myopathic changes on needle examination. Caution is needed to avoid misinterpreting large motor unit potentials (common in chronic myopathies) and fibrillations (a result of myonecrosis or vacuolar change) as evidence of a neurogenic process. Muscle biopsy will demonstrate dystrophic changes and, in many families, rimmed vacuoles. Immunohistochemistry for dysferlin or desmin may be informative.138

Motor Neuron Disease Anterior horn cell disease is another consideration in patients with weakness without sensory changes. Amyotrophic lateral sclerosis is unlikely to be mistaken for polyneuropathy, although many patients are initially diagnosed with a focal neuropathy (such as an ulnar neuropathy or cervical radiculopathy in a patient presenting with hand weakness) only to later display more widespread active denervation. Difficulty is more likely in patients with inherited lower motor neuron syndromes or spinal muscular atrophy (SMA). While the classic forms of SMA cause predominantly proximal weakness, several types of distal SMA have been described, and a classification scheme was proposed by Harding.51 Type II distal SMA (also called hereditary motor neuropathy type II) is an autosomal dominant disorder that presents as weakness of toe extensors beginning in the third decade, with progressive involvement of distal more than proximal leg muscles and distal upper limb muscles.135 Hyporeflexia and loss of vibration sense occur in some patients, who may eventually become wheelchair bound. This may be confused with axonal CMT disease; the presence of completely normal sensory nerve conduction studies will help resolve matters, but uncertainty will persist if sensory changes are present as a result of superimposed nerve compression or the SMA itself. Other distal SMA variants are known, including forms with childhood onset, upper limb predominance, or vocal cord weakness. Old polio occasionally also leaves symmetrical distal weakness.

Myelopathy Spinal cord lesions sometimes produce distal sensory symptoms in the legs or in all four limbs. Distal weakness may also occur. This is most likely to occur with cervical cord lesions such as multiple sclerosis and compression caused

by spondylosis. A history of neck pain or Lhermitte’s sign is helpful in suggesting a cervical lesion. Subtle upper motor neuron findings should be carefully sought, but their absence does not exclude a myelopathy. Sensory nerve conduction studies will be normal despite sensory symptoms and deficits, but somatosensory evoked potentials may show a delay in central conduction. Neuroimaging studies will usually demonstrate the lesion, which is sometimes more rostral than the clinical findings would suggest.

Aging With normal aging, clinical and electrophysiologic changes occur that resemble those caused by neuropathy. Given that neuropathy commonly affects the elderly, distinguishing between nerve pathology and normal age-related loss of neuromuscular function can be difficult. A gradual decline in muscle strength occurs late in life, mainly after age 60.22,112 This relates to reduced muscle mass, a reduced number of muscle fibers, and fiber atrophy (mainly affecting type II fibers).76,77,136 However, this is unlikely to confound the clinical assessment if reasonable tests are used. Most healthy individuals can stand on their heels and their toes into the eighth decade.30 Non-neurologic disease such as arthritis may cause apparent weakness. Sensory thresholds also increase with age. This has been demonstrated for warm,29 vibration, and cooling30 thresholds. Altered ankle reflexes are common in the elderly. In the healthy subjects (control) cohort of the Rochester Diabetic Neuropathy Study, a reduction or absence of ankle reflexes was found in over 5% of people over age 50 and almost 30% of people over 70.30 With age, nerve conduction studies show a reduction in conduction velocities and amplitudes in both motor and sensory nerves.30,119,120,131 It is generally acknowledged that sural sensory responses may be small or absent in healthy people older than 60 years, although this has been challenged.38 Regarding the needle examination, several quantitative and macro EMG studies have shown an increase in the size (duration and amplitude) of motor unit potentials with aging.52,56,130 Motor unit numbers also decline.13,21 Muscle biopsies in healthy elderly subjects show fibertype grouping or grouped atrophy, presumably reflecting low-grade denervation and reinnervation.76,136

REFERENCES 1. Abarbanel, J. M., Berginer, V. M., Osimani, A., et al.: Neurologic complications after gastric restriction surgery for morbid obesity. Neurology 37:196, 1987. 2. Akiya, S., Nishio, Y., Ibi, K., et al.: Lattice corneal dystrophy type II associated with familial amyloid polyneuropathy type IV. Ophthalmology 103:1106, 1996.

Differential Diagnosis of Polyneuropathy 3. Albers, J. W., and Kelly, J. J. Jr.: Acquired inflammatory demyelinating polyneuropathies: clinical and electrodiagnostic features. Muscle Nerve 12:435, 1989. 4. Andersson, P. B., Yuen, E., Parko, K., and So, Y. T.: Electrodiagnostic features of hereditary neuropathy with liability to pressure palsies. Neurology 54:40, 2000. 5. Archer, A. G., Watkins, P. J., Thomas, P. K., et al.: The natural history of acute painful neuropathy in diabetes mellitus. J. Neurol. Neurosurg. Psychiatry 46:491, 1983. 6. Argov, Z., Steiner, I., and Soffer, D.: The yield of sural nerve biopsy in the evaluation of peripheral neuropathies. Acta Neurol. Scand. 79:243, 1989. 7. Ashworth, B., and Tait, G. B.: Trigeminal neuropathy in connective tissue disease. Neurology 21:609, 1971. 8. Bardwick, P. A., Zvaifler, N. J., Gill, G. N., et al.: Plasma cell dyscrasia with polyneuropathy, organomegaly, endocrinopathy, M protein, and skin changes: the POEMS syndrome. Report on two cases and a review of the literature. Medicine (Baltimore) 59:311, 1980. 9. Barohn, R. J.: Approach to peripheral neuropathy and neuronopathy. Semin. Neurol. 18:7, 1998. 10. Barohn, R. J., Kissel, J. T., Warmolts, J. R., and Mendell, J. R.: Chronic inflammatory demyelinating polyradiculoneuropathy: clinical characteristics, course, and recommendations for diagnostic criteria. Arch. Neurol. 46:878, 1989. 11. Boerkoel, C. F., Takashima, H., Garcia, C. A., et al.: Charcot-Marie-Tooth disease and related neuropathies: mutation distribution and genotype-phenotype correlation. Ann. Neurol. 51:190, 2002. 12. Bromberg, M. B.: Comparison of electrodiagnostic criteria for primary demyelination in chronic polyneuropathy. Muscle Nerve 14:968, 1991. 13. Brown, W. F., Strong, M. J., and Snow, R.: Methods for estimating numbers of motor units in biceps-brachialis muscles and losses of motor units with aging. Muscle Nerve 11:423, 1988. 14. Busby, M., and Donaghy, M.: Predominant arm weakness in acute idiopathic polyneuritis: a distinct regional variant. J. Neurol. 247:343, 2000. 15. Calabresi, P. A., Silvestri, G., DiMauro, S., and Griggs, R. C.: Ekbom’s syndrome: lipomas, ataxia, and neuropathy with MERRF. Muscle Nerve 17:943, 1994. 16. Camerlingo, M., Nemni, R., Ferraro, B., et al.: Malignancy and sensory neuropathy of unexplained cause: a prospective study of 51 patients. Arch. Neurol. 55:981, 1998. 17. Campbell, A. M., and Williams, E. R.: Natural history of Refsum’s syndrome in a Gloucestershire family. Br. Med. J. 3:777, 1967. 18. Chalk, C. H., Homburger, H. A., and Dyck, P. J.: Antineutrophil cytoplasmic antibodies in vasculitis peripheral neuropathy. Neurology 43:1826, 1993. 19. Cossoy, M. B., Benstead, T. J., and Sadler, R. M.: Measurement of serum methylmalonic acid to diagnose cobalamin deficiency in patients with posterior column dysfunction and normal serum vitamin B12 levels. Can. J. Neurol. Sci. 28(Suppl. 2):S44, 2001. 20. Davis, M. D., O’Fallon, W. M., Rogers, R. S. 3rd, and Rooke, T. W.: Natural history of erythromelalgia: presentation and outcome in 168 patients. Arch. Dermatol. 136:330, 2000.

1177

21. Doherty, T. J., and Brown, W. F.: The estimated numbers and relative sizes of thenar motor units as selected by multiple point stimulation in young and older adults. Muscle Nerve 16:355, 1993. 22. Doherty, T. J., Vandervoort, A. A., and Brown, W. F.: Effects of ageing on the motor unit: a brief review. Can. J. Appl. Physiol. 18:331, 1993. 23. Drenth, J. P., and Michiels, J. J.: Erythromelalgia and erythermalgia: diagnostic differentiation. Int. J. Dermatol. 33:393, 1994. 24. Dressler, D., Thompson, P. D., Gledhill, R. F., and Marsden, C. D.: The syndrome of painful legs and moving toes. Mov. Disord. 9:13, 1994. 25. Duyff, R. F., Van den Bosch, J., Laman, D. M., et al.: Neuromuscular findings in thyroid dysfunction: a prospective clinical and electrodiagnostic study. J. Neurol. Neurosurg. Psychiatry 68:750, 2000. 26. Dyck, P. J., Benstead, T. J., Conn, D. L., et al.: Nonsystemic vasculitic neuropathy. Brain 110(Pt. 4):843, 1987. 27. Dyck, P. J., Dyck, P. J., and Chalk, C. H.: The 10 P’s: a mnemonic helpful in characterization and differential diagnosis of peripheral neuropathy. Neurology 42:14, 1992. 28. Dyck, P. J., Dyck, P. J., Grant, I. A., and Fealey, R. D.: Ten steps in characterizing and diagnosing patients with peripheral neuropathy. Neurology 47:10, 1996. 29. Dyck, P. J., Kratz, K. M., Karnes, J. L., et al.: The prevalence by staged severity of various types of diabetic neuropathy, retinopathy, and nephropathy in a population-based cohort: the Rochester Diabetic Neuropathy Study. Neurology 43:817, 1993. 30. Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects. Neurology 45:1115, 1995. 31. Dyck, P. J., Oviatt, K. F., and Lambert, E. H.: Intensive evaluation of referred unclassified neuropathies yields improved diagnosis. Ann. Neurol. 10:222, 1981. 32. Ehle, A. L.: Lead neuropathy and electrophysiological studies in low level lead exposure: a critical review. Neurotoxicology 7:203, 1986. 33. Ellenberg, M.: Diabetic neuropathic cachexia. Diabetes 23:418, 1974. 34. Elsehety, A., and Bertorini, T. E.: Neurologic and neuropsychiatric complications of Crohn’s disease. South. Med. J. 90:606, 1997. 35. Factors in development of diabetic neuropathy: baseline analysis of neuropathy in feasibility phase of Diabetes Control and Complications Trial (DCCT). The DCCT Research Group. Diabetes 37:476, 1988. 36. Fagerberg, S.-E.: Diabetic neuropathy: a clinical and histological study on the significance of vascular affections. Acta Med. Scand. Suppl. 345:1, 1959. 37. Fagius, J.: Chronic cryptogenic polyneuropathy: the search for a cause. Acta Neurol. Scand. 67:173, 1983. 38. Falco, F. J., Hennessey, W. J., Goldberg, G., and Braddom, R. L.: Standardized nerve conduction studies in the lower limb of the healthy elderly. Am. J. Phys. Med. Rehabil. 73:168, 1994. 39. Farrell, D. A., and Medsger, T. A. Jr.: Trigeminal neuropathy in progressive systemic sclerosis. Am. J. Med. 73:57, 1982.

1178

Diseases of the Peripheral Nervous System

40. Fealey, R. D.: Thermoregulatory sweat test. In Daube, J. R. (ed.): Clinical Neurophysiology. Philadelphia, F. A. Davis, p. 396, 1996. 41. Gabriel, C. M., Howard, R., Kinsella, N., et al.: Prospective study of the usefulness of sural nerve biopsy. J. Neurol. Neurosurg. Psychiatry 69:442, 2000. 42. Gabriel, J. M., Erne, B., Bernasconi, L., et al.: Confocal microscopic localization of anti-myelin-associated glycoprotein autoantibodies in a patient with peripheral neuropathy initially lacking a detectable IgM gammopathy. Acta Neuropathol. (Berl.) 95:540, 1998. 43. Gibbels, E., Schaefer, H. E., Runne, U., et al: Severe polyneuropathy in Tangier disease mimicking syringomyelia or leprosy: clinical, biochemical, electrophysiological, and morphological evaluation, including electron microscopy of nerve, muscle, and skin biopsies. J. Neurol. 232:283, 1985. 44. Good, A. E., Christopher, R. P., Koepke, G. H., et al.: Peripheral neuropathy associated with rheumatoid arthritis: a clinical and electrodiagnostic study of 70 consecutive rheumatoid arthritis patients. Ann. Intern. Med. 63:87, 1965. 45. Gorson, K. C., Ropper, A. H., and Weinberg, D. H.: Upper limb predominant, multifocal chronic inflammatory demyelinating polyneuropathy. Muscle Nerve 22:758, 1999. 46. Gosselin, S., Kyle, R. A., and Dyck, P. J.: Neuropathy associated with monoclonal gammopathies of undetermined significance. Ann. Neurol. 30:54, 1991. 47. Grahmann, F., Winterholler, M., and Neundorfer, B.: Cryptogenetic polyneuropathies: an out-patient follow-up study. Acta Neurol. Scand. 84:221, 1991. 48. Grant, I. A., Hunder, G. G., Homburger, H. A., and Dyck, P. J.: Peripheral neuropathy associated with sicca complex. Neurology 48:855, 1997. 49. Gruener, G., and Dyck, P. J.: Quantitative sensory testing: methodology, applications, and future directions. J. Clin. Neurophysiol. 11:568, 1994. 50. Hanel, P.: Ueber eine Form von nicht beschriebener hereditarer neurotischer Muskelatrophie. In Inaugural Dissertation, Medizinischen Fakultat zu Jena, 1890. 51. Harding, A. E.: Inherited neuronal atrophy and degeneration predominantly in lower motor neurons. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1051, 1993. 52. Hayward, M.: Automatic analysis of the electromyogram in healthy subjects of different ages. J. Neurol. Sci. 33:397, 1977. 53. Hellmann, D. B., Laing, T. J., Petri, M., et al.: Mononeuritis multiplex: the yield of evaluations for occult rheumatic diseases. Medicine (Baltimore) 67:145, 1988. 54. Herskovitz, S., Loh, F., Berger, A. R., and Kucherov, M.: Erythromelalgia: association with hereditary sensory neuropathy and response to amitriptyline. Neurology 43(3 Pt. 1): 621, 1993. 55. Holt, I. J., Harding, A. E., Petty, R. K., and MorganHughes, J. A.: A new mitochondrial disease associated with mitochondrial DNA heteroplasmy. Am. J. Hum. Genet. 46:428, 1990. 56. Howard, J. E., McGill, K. C., and Dorfman, L. J.: Age effects on properties of motor unit action potentials: ADEMG analysis. Ann. Neurol. 24:207, 1988.

57. Hurko, O., and Provost, T. T.: Neurology and the skin. J. Neurol. Neurosurg. Psychiatry 66:417, 1999. 58. Itin, P. H., and Winkelmann, R. K.: Cutaneous manifestations in patients with essential thrombocythemia. J. Am. Acad. Dermatol. 24:59, 1991. 59. Jann, S., Beretta, S., Bramerio, M., and Defanti, C. A.: Prospective follow-up study of chronic polyneuropathy of undetermined cause. Muscle Nerve 24:1197, 2001. 60. Jenkins, R. B.: Reynolds, Aldrich or Mees? A consideration of transverse striate leukonychia. Am. J. Med. Sci. 246:707, 1963. 61. Kaku, D. A., England, J. D., and Sumner, A. J.: Distal accentuation of conduction slowing in polyneuropathy associated with antibodies to myelin-associated glycoprotein and sulphated glucuronyl paragloboside. Brain 117(Pt. 5): 941, 1994. 62. Kaplan, J. G., Pack, D., Horoupian, D., et al.: Distal axonopathy associated with chronic gluten enteropathy: a treatable disorder. Neurology 38:642, 1988. 63. Kelly, J. J., Adelman, L. S., Berkman, E., and Bhan, I.: Polyneuropathies associated with IgM monoclonal gammopathies. Arch. Neurol. 45:1355, 1988. 64. Kissel, J. T., Slivka, A. P., Warmolts, J. R., and Mendell, J. R.: The clinical spectrum of necrotizing angiopathy of the peripheral nervous system. Ann. Neurol. 18:251, 1985. 65. Kovach, M. J., Campbell, K. C., Herman, K., et al.: Anticipation in a unique family with Charcot-Marie-Tooth syndrome and deafness: delineation of the clinical features and review of the literature. Am. J. Med. Genet. 108:295, 2002. 66. Kurzrock, R., and Cohen, P. R.: Erythromelalgia and myeloproliferative disorders. Arch. Intern. Med. 149:105, 1989. 67. Lagerkvist, B. J., and Zetterlund, B.: Assessment of exposure to arsenic among smelter workers: a five-year follow-up. Am. J. Ind. Med. 25:477, 1994. 68. Lander, C. M., Eadie, M. J., and Tyrer, J. H.: Hereditary motor peripheral neuropathy predominantly affecting the arms. J. Neurol. Sci. 28:389, 1976. 69. Layzer, R. B.: Hot feet: erythromelalgia and related disorders. J. Child. Neurol. 16:199, 2001. 70. Lecky, B. R., Hughes, R. A., and Murray, N. M.: Trigeminal sensory neuropathy: a study of 22 cases. Brain 110(Pt. 6): 1463, 1987. 71. Lee, M. J., Nelson, I., Houlden, H., et al.: Six novel connexin32 (GJB1) mutations in X-linked Charcot-MarieTooth disease. J. Neurol. Neurosurg. Psychiatry 73:304, 2002. 72. Lee, P., Bruni, J., and Sukenik, S.: Neurological manifestations in systemic sclerosis (scleroderma). J. Rheumatol. 11:480, 1984. 73. Leger, J. M.: Diagnosis of chronic neuropathy. J. Neurol. 246:156, 1999. 74. Lewis, R. A., and Sumner, A. J.: The electrodiagnostic distinctions between chronic familial and acquired demyelinative neuropathies. Neurology 32:592, 1982. 75. Lewis, R. A., Sumner, A. J., and Shy, M. E.: Electrophysiological features of inherited demyelinating neuropathies: a reappraisal in the era of molecular diagnosis. Muscle Nerve 23:1472, 2000.

Differential Diagnosis of Polyneuropathy 76. Lexell, J., and Downham, D. Y.: The occurrence of fibretype grouping in healthy human muscle: a quantitative study of cross-sections of whole vastus lateralis from men between 15 and 83 years. Acta Neuropathol. 81:377, 1991. 77. Lexell, J., Taylor, C. C., and Sjostrom, M.: What is the cause of the ageing atrophy? Total number, size and proportion of different fiber types studied in whole vastus lateralis muscle from 15- to 83-year-old men. J. Neurol. Sci. 84:275, 1988. 78. Lindenbaum, J., Healton, E. B., Savage, D. G., et al.: Neuropsychiatric disorders caused by cobalamin deficiency in the absence of anemia or macrocytosis. N. Engl. J. Med. 318:1720, 1988. 79. Lindenbaum, J., Savage, D. G., Stabler, S. P., and Allen, R. H.: Diagnosis of cobalamin deficiency: II. Relative sensitivities of serum cobalamin, methylmalonic acid, and total homocysteine concentrations. Am. J. Hematol. 34:99, 1990. 80. Lossos, A., River, Y., Eliakim, A., and Steiner, I.: Neurologic aspects of inflammatory bowel disease. Neurology 45(3 Pt. 1): 416, 1995. 81. Low, P. A., Caskey, P. E., Tuck, R. R., et al.: Quantitative sudomotor axon reflex test in normal and neuropathic subjects. Ann. Neurol. 14:573, 1983. 82. Maisonobe, T., Chassande, B., Verin, M., et al.: Chronic dysimmune demyelinating polyneuropathy: a clinical and electrophysiological study of 93 patients. J. Neurol. Neurosurg. Psychiatry 61:36, 1996. 83. Markesbery, W. R., Griggs, R. C., Leach, R. P., and Lapham, L. W.: Late onset hereditary distal myopathy. Neurology 24:127, 1974. 84. Maryniak, O.: Severe peripheral neuropathy following gastric bypass surgery for morbid obesity. Can. Med. Assoc. J. 131:119, 1984. 85. McArthur, J. C., Stocks, E. A., Hauer, P., et al.: Epidermal nerve fiber density: normative reference range and diagnostic efficiency. Arch. Neurol. 55:1513, 1998. 86. McLeod, J. G.: Investigation of peripheral neuropathy. J. Neurol. Neurosurg. Psychiatry 58:274, 1995. 87. McLeod, J. G., Tuck, R. R., Pollard, J. D., et al.: Chronic polyneuropathy of undetermined cause. J. Neurol. Neurosurg. Psychiatry 47:530, 1984. 88. Menkes, D. L.: Images in neurology. The cutaneous stigmata of Fabry disease: an X- linked phakomatosis associated with central and peripheral nervous system dysfunction. Arch. Neurol. 56:487, 1999. 89. Michiels, J. J., and ten Kate, F. J.: Erythromelalgia in thrombocythemia of various myeloproliferative disorders. Am. J. Hematol. 39:131, 1992. 90. Milhorat, A. T., and Wolff, H. G.: Studies in diseases of muscle: XIII. Progressive muscular dystrophy of atrophic distal type: report on a family: report of autopsy. Arch. Neurol. Psychiatry 49:655, 1943. 91. Misu, K., Yoshihara, T., Shikama, Y., et al.: An axonal form of Charcot-Marie-Tooth disease showing distinctive features in association with mutations in the peripheral myelin protein zero gene (Thr124Met or Asp75Val). J. Neurol. Neurosurg. Psychiatry 69:806, 2000. 92. Mitchell, S. W.: On a rare vaso-motor neurosis of the extremities, and on the maladies with which it may be confounded. Am. J. Med. Sci. 76:12, 1878.

1179

93. Miyoshi, K., Kawai, H., Iwasa, M., et al.: Autosomal recessive distal muscular dystrophy as a new type of progressive muscular dystrophy: seventeen cases in eight families including an autopsied case. Brain 109(Pt. 1):31, 1986. 94. Moder, K. G.: Use and interpretation of rheumatologic tests: a guide for clinicians. Mayo Clin. Proc. 71:391, 1996. 95. Montagna, P., Cirignotta, F., Sacquegna, T., et al.: “Painful legs and moving toes” associated with polyneuropathy. J. Neurol. Neurosurg. Psychiatry 46:399, 1983. 96. Mouton, P., Tardieu, S., Gouider, R., et al.: Spectrum of clinical and electrophysiologic features in HNPP patients with the 17p11.2 deletion. Neurology 52:1440, 1999. 97. Nathan, P. W.: Painful legs and moving toes: evidence on the site of the lesion. J. Neurol. Neurosurg. Psychiatry 41:934, 1978. 98. Nobile-Orazio, E., Manfredini, E., Carpo, M., et al.: Frequency and clinical correlates of anti-neural IgM antibodies in neuropathy associated with IgM monoclonal gammopathy. Ann. Neurol. 36:416, 1994. 99. Nonaka, I., Sunohara, N., Ishiura, S., and Satoyoshi, E.: Familial distal myopathy with rimmed vacuole and lamellar (myeloid) body formation. J. Neurol. Sci. 51:141, 1981. 100. Notermans, N. C.: Monoclonal gammopathy and neuropathy. Curr. Opin. Neurol. 9:334, 1996. 101. Notermans, N. C., Wokke, J. H., Lokhorst, H. M., et al.: Polyneuropathy associated with monoclonal gammopathy of undetermined significance: a prospective study of the prognostic value of clinical and laboratory abnormalities. Brain 117(Pt. 6):1385, 1994. 102. Novella, S. P., Inzucchi, S. E., and Goldstein, J. M.: The frequency of undiagnosed diabetes and impaired glucose tolerance in patients with idiopathic sensory neuropathy. Muscle Nerve 24:1229, 2001. 103. Ochoa, J., and Torebjork, E.: Sensations evoked by intraneural microstimulation of C nociceptor fibres in human skin nerves. J. Physiol. (Lond.) 415:583, 1989. 104. Pareyson, D., Taroni, F., Botti, S., et al.: Cranial nerve involvement in CMT disease type 1 due to early growth response 2 gene mutation. Neurology 54:1696, 2000. 105. Perry, J. R., and Bril, V.: Complications of sural nerve biopsy in diabetic versus non-diabetic patients. Can. J. Neurol. Sci. 21:34, 1994. 106. Peyronnard, J. M., Charron, L., Beaudet, F., and Couture, F.: Vasculitic neuropathy in rheumatoid disease and Sjogren syndrome. Neurology 32:839, 1982. 107. Pfeifer, M. A., Cook, D., Brodsky, J., et al.: Quantitative evaluation of cardiac parasympathetic activity in normal and diabetic man. Diabetes 31(4 Pt. 1):339, 1982. 108. Pirart, J.: Diabetes mellitus and its degenerative complications: a prospective study of 4,400 patients observed between 1947 and 1973 [author’s transl.]. Diabetes Metab. 3:97, 1977. 109. Pitagoras de Mattos, J., Oliveira, M., and Andre, C.: Painful legs and moving toes associated with neuropathy in HIVinfected patients. Mov. Disord. 14:1053, 1999. 110. Pittock, S. J., Payne, T. A., and Harper, C. M.: Reversible myelopathy in a 34-year-old man with vitamin B12 deficiency. Mayo Clin. Proc. 77:291, 2002. 111. Pollock, M., Nukada, H., Frith, R. W., et al.: Peripheral neuropathy in Tangier disease. Brain 106(Pt. 4):911, 1983.

1180

Diseases of the Peripheral Nervous System

112. Porter, M. M., Vandervoort, A. A., and Lexell, J.: Aging of human muscle: structure, function and adaptability. Scand. J. Med. Sci. Sports 5:129, 1995. 113. Quecedo, E., Sanmartin, O., Febrer, M. I., et al.: Mees’ lines: a clue for the diagnosis of arsenic poisoning. Arch. Dermatol 132:349, 1996. 114. Quinlan, C. D., and Martin, E. A.: Refsum’s syndrome: report of three cases. J. Neurol. Neurosurg. Psychiatry 33:817, 1970. 115. Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP): report from an Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force. Neurology 41:617, 1991. 116. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 21(Suppl. 1):S5, 1998. 117. Ridley, A.: The neuropathy of acute intermittent porphyria. Q. J. Med. 38:307, 1969. 118. Ridley, A.: Porphyric neuropathy. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. P. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 1704, 1984. 119. Rivner, M. H., Swift, T. R., Crout, B. O., and Rhodes, K. P.: Toward more rational nerve conduction interpretations: the effect of height. Muscle Nerve 13:232, 1990. 120. Rivner, M. H., Swift, T. R., and Malik, K.: Influence of age and height on nerve conduction. Muscle Nerve 24:1134, 2001. 121. Rosenberg, N. R., Portegies, P., de Visser, M., and Vermeulen, M.: Diagnostic investigation of patients with chronic polyneuropathy: evaluation of a clinical guideline. J. Neurol. Neurosurg. Psychiatry 71:205, 2001. 122. Rukavina, J. G., Block, W. D., and Jackson, C. E.: Primary systemic amyloidosis: a review and an experimental genetic and clinical study of 29 cases with particular emphasis on the familial form. Medicine (Baltimore) 35:239, 1956. 123. Said, G., Lacroix, C., Plante-Bordeneuve, V., et al.: Nerve granulomas and vasculitis in sarcoid peripheral neuropathy: a clinicopathological study of 11 patients. Brain 125(Pt. 2): 264, 2002. 124. Said, G., Lacroix-Ciaudo, C., Fujimura, H., et al.: The peripheral neuropathy of necrotizing arteritis: a clinicopathological study. Ann. Neurol. 23:461, 1988. 125. Servidei, S., Capon, F., Spinazzola, A., et al.: A distinctive autosomal dominant vacuolar neuromyopathy linked to 19p13. Neurology 53:830, 1999. 126. Simovic, D., Gorson, K. C., and Ropper, A. H.: Comparison of IgM-MGUS and IgG-MGUS polyneuropathy. Acta Neurol. Scand. 97:194, 1998. 127. Singleton, J. R., Smith, A. G., and Bromberg, M. B.: Painful sensory polyneuropathy associated with impaired glucose tolerance. Muscle Nerve 24:1225, 2001. 128. Skjeldal, O. H., Stokke, O., Refsum, S., et al.: Clinical and biochemical heterogeneity in conditions with phytanic acid accumulation. J. Neurol. Sci. 77:87, 1987. 129. Spillane, J. D., Nathan, P. W., Kelly, R. E., and Marsden C. D.: Painful legs and moving toes. Brain 94:541, 1971.

130. Stalberg, E., and Fawcett, P. R.: Macro EMG in healthy subjects of different ages. J. Neurol. Neurosurg. Psychiatry 45:870, 1982. 131. Stetson, D. S., Albers, J. W., Silverstein, B. A., and Wolfe, R. A.: Effects of age, sex, and anthropometric factors on nerve conduction measures. Muscle Nerve 15:1095, 1992. 132. Suarez, G. A., Chalk, C. H., Russell, J. W., et al.: Diagnostic accuracy and certainty from sequential evaluations in peripheral neuropathy. Neurology 57:1118, 2001. 133. Suarez, G. A., and Kelly, J. J. Jr.: Polyneuropathy associated with monoclonal gammopathy of undetermined significance: further evidence that IgM-MGUS neuropathies are different than IgG-MGUS. Neurology 43:1304, 1993. 134. Thomas, P. K., Claus, D., Jaspert, A., et al.: Focal upper limb demyelinating neuropathy. Brain 119(Pt. 3):765, 1996. 135. Timmerman, V., Raeymaekers, P., Nelis, E., et al.: Linkage analysis of distal hereditary motor neuropathy type II (distal HMN II) in a single pedigree. J. Neurol. Sci. 109:41, 1992. 136. Tomlinson, B. E., Walton, J. N., and Rebeiz, J. J.: The effects of ageing and of cachexia upon skeletal muscle: a histopathological study. J. Neurol. Sci. 9:321, 1969. 137. Tyson, J., Malcolm, S., Thomas, P. K., and Harding, A. E.: Deletions of chromosome 17p11.2 in multifocal neuropathies. Ann. Neurol. 39:180, 1996. 138. Udd, B., and Griggs, R.: Distal myopathies. Curr. Opin. Neurol. 14:561, 2001. 139. Udd, B., Partanen, J., Halonen, P., et al.: Tibial muscular dystrophy: late adult-onset distal myopathy in 66 Finnish patients. Arch. Neurol. 50:604, 1993. 140. Vendrell, J., Nubiola, A., Goday, A., et al.: Erythromelalgia associated with acute diabetic neuropathy: an unusual condition. Diabetes Res. 7:149, 1988. 141. Verhagen, W. I., Gabreels-Festen, A. A., van Wensen, P. J., et al.: Hereditary neuropathy with liability to pressure palsies: a clinical, electroneurophysiological and morphological study. J. Neurol. Sci. 116:176, 1993. 142. Verhagen, W. I., Horstink, M. W., and Notermans, S. L.: Painful arm and moving fingers. J. Neurol. Neurosurg. Psychiatry 48:384, 1985. 143. Vicart, P., Caron, A., Guicheney, P., et al.: A missense mutation in the alphaB-crystallin chaperone gene causes a desmin-related myopathy. Nat. Genet. 20:92, 1998. 144. Wallace, M. R., Conneally, P. M., and Benson, M. D.: A DNA test for Indiana/Swiss hereditary amyloidosis (FAP II). Am. J. Hum. Genet. 43:182, 1988. 145. Welander, L.: Myopathia distalis tarda hereditaria. Acta Med. Scand. 141:1, 1951. 146. Wolfe, G. I., Baker, N. S., Amato, A. A, et al.: Chronic cryptogenic sensory polyneuropathy: clinical and laboratory characteristics. Arch. Neurol. 56:540, 1999. 147. Wright, A., and Dyck, P. J.: Hereditary sensory neuropathy with sensorineural deafness and early-onset dementia. Neurology 45(3 Pt. 1):560, 1995. 148. Wulff, C. H.: Painful legs and moving toes: a report of 3 cases with neurophysiological studies. Acta Neurol. Scand. 66:283, 1982.

47 Epidemiologic Approaches to Peripheral Neuropathy JAMES P. BURKE AND L. JOSEPH MELTON III

Epidemiologic Methods Descriptive Studies Analytic Studies Genetic Epidemiology Clinical Applications: Results of Epidemiologic Studies

Tri-o-cresyl Phosphate Intoxication Nutrient-Deficient Diets Charcot-Marie-Tooth Disease Postherpetic Neuralgia Guillain-Barré Syndrome

Whereas clinicians are concerned with disease in individual patients, the perspective provided by epidemiology more accurately reflects the impact of peripheral neuropathy on society as a whole. By determining the circumstances under which such diseases occur, epidemiologic investigations can characterize those segments of the population that are at particularly high or low risk. Because disease does not occur randomly, but rather in patterns that reflect the operation of underlying causes, knowledge of the factors associated with a high or low risk of peripheral neuropathy is useful for designing prevention programs and for generating etiologic hypotheses to be tested in subsequent clinical research. Knowledge of the true clinical spectrum of neuropathy in the community also provides the best information about the natural history of these disorders and, thus, may be of assistance in patient management. In this chapter, the general principles of epidemiologic methodology are reviewed and classic examples are then provided to illustrate the usefulness of epidemiologic methods in the study of peripheral neuropathy.

EPIDEMIOLOGIC METHODS Epidemiology is the study of the distribution and determinants of disease in populations or groups. The relevant methodology can be divided into segments that support two general aims: (1) the tools needed to describe the

Diabetic Neuropathy Summary

distribution of peripheral neuropathy in the community, and (2) those required to identify factors responsible for the onset or progression of neuropathy. These are covered briefly here. More complete treatments of epidemiologic methodology may be found elsewhere.10,19

Descriptive Studies To say how many people have or will develop a disease has little meaning unless one also provides information about the number of people who are at risk of having or developing that particular disease. Three new cases of GuillainBarré syndrome in a village of 100 people have quite different implications than three new cases in a city of 1 million inhabitants. Thus the frequency of disease occurrence must be expressed as a rate or ratio, in which the number of cases of the disease (numerator) is related to the population at risk of the disease (denominator). There are two general approaches to collecting this information: prospective (longitudinal) studies and cross-sectional surveys. In a longitudinal study, all cases of a particular disease that arise in a defined population over a period of time (e.g., days, years, decades) are identified; the primary purpose is to determine the incidence of new cases of the disease. By contrast, a cross-sectional survey is done at one point in time and all existing cases are enumerated; the main purpose is to establish the prevalence of the disease. Incidence and prevalence rates are defined in Table 47–1. 1181

1182

Diseases of the Peripheral Nervous System

Table 47–1. Epidemiologic Indices of Morbidity Useful in Studies of Peripheral Neuropathy Prevalence

Incidence

Measures the frequency of existing cases of a disease within a specific population Calculated for a given time and place Expressed as a prevalence ratio: the number of persons with a given disease at a specific time per 1000 (or per 100 for common condition or per 100,000 for rare ones) persons capable of having the disease of that time Measures the rapidity with which new cases of disease occur within a specific population Calculated for a given time interval and place Expressed as an incidence rate: the number of new cases of a given disease during a specified period (usually 1 year) per 100,000 (or per 1000 or 10,000) persons at risk of having the disease during that time In a stable population, prevalence is approximately equal to the incidence rate multiplied by the average duration of the disease.

Modified from Schoenberg, B. S.: Neurologic disease in the elderly: epidemiologic considerations. Semin. Neurol. 1:5, 1981.

Prevalence and incidence are related (prevalence 艐 incidence times the average duration of disease). Because prevalence rates are also influenced by survival, migration, and other factors, they are usually very unreliable estimators of incidence. Both are measures of disease morbidity. Methodologic problems in descriptive studies are related to the accuracy of the numerator and denominator. The numerator depends, first, on a precise definition of the disorder (i.e., the specific signs or symptoms that must be present in order to make the diagnosis of peripheral neuropathy). Because peripheral neuropathies are a heterogeneous group of disorders, classification into meaningful subtypes may also be important, both clinically and epidemiologically. This task is complicated by the fact that neuropathies tend to have an insidious onset and a clinical course that is often slowly progressive and chronic, occasionally following a pattern of remission and exacerbation. Therefore, one may have to observe patients over a period of time to document their status. In surveys of peripheral neuropathy among large populations, one must frequently rely on the patient’s subjective evaluation of symptoms such as paresthesias or painful dysesthesias; the accuracy and reproducibility of these subjective assessments are quite variable. Incidence and prevalence estimates may vary dramatically depending on which minimal criteria and diagnostic end points are used. In diabetic polyneuropathy, if symptoms are used, about 5% of diabetics are defined as having neuropathy.6,7 If vibration threshold is used, perhaps as many as 25% are defined as having neuropathy. In addition, the percentile

at which abnormality is set has large implications for the estimates of disease frequency. The neurologist’s assessment, especially of sensory function, is a difficult and time-consuming aspect of the clinical examination of peripheral neuropathy. Lengthy examinations carried out among large populations may place a serious strain on available personnel. Nonetheless, it is essential to remember that the results of the most sophisticated analysis are no better than the quality of the original data. This problem was noted over 50 years ago by a British statistician, Sir Josiah Stamp,41 who wrote: The government are very keen on amassing statistics— they collect them, add them, raise them to the nth power, take the cube root and prepare wonderful diagrams. But what you must never forget is that every one of those figures comes in the first instance from the . . . village watchman, who puts down what he damn pleases. For nearly all studies of neurologic disease in populations, the clinician represents the village watchman, and the validity of the results depends on the accuracy and completeness with which he or she makes diagnoses and records data. A second, more subtle, numerator problem is related to the representativeness of the individuals selected for investigation. Generally, it is preferable to identify all cases of a particular neurologic disease within the target population (all new cases for an incidence study, or all existing cases for a prevalence study). Complete enumeration is required for an accurate estimate of disease incidence or prevalence but, more importantly, helps assure that the full clinical spectrum of cases is preserved. If, for example, some portion of cases is systemically missed (the mild cases, the elderly, etc.), the resulting morbidity rates will be unduly low (ascertainment bias). However, the clinical spectrum will also be distorted (selection bias). This almost always occurs in studies restricted to patients attended by a single physician or conducted at a single hospital. The denominator for these rates should be restricted to the population actually at risk for the disease. Because this information is rarely available, incidence and prevalence rates are often calculated for the entire population, with the implied assumption that everyone is at risk. Thus the population may be defined on the basis of geography (e.g., residents of a certain town), occupation (e.g., workers in a particular industry), or some other distinguishing characteristic. If the population from which the cases arise is unknown, as is usually the situation at tertiary referral centers, neither incidence nor prevalence rates can be calculated.

Analytic Studies Enumerating those individuals with a given disorder is not usually an end in itself. Instead, the emphasis in epidemiology is on discovering the causes of a disease or

Epidemiologic Approaches to Peripheral Neuropathy

the factors that lead to progression and complications. The patterns of disease in a community often allow one to postulate etiologic mechanisms. Associations between putative risk factors and disease can then be formally tested using the techniques of analytic epidemiology. Again, there are two main approaches: case-control and cohort studies. In a cohort study, a group of individuals with a particular factor thought to be related to disease etiology (exposed) and a comparable group without the characteristic (unexposed) are observed over time for the development of disease. To establish a positive association between the exposure and the occurrence of disease, it is necessary to show that the rate of disease is higher in those exposed than in those not exposed. The relative risk is calculated as the ratio of incidence rates (the incidence among the exposed divided by the incidence among the unexposed). For some studies to be feasible, the disease outcome must be relatively frequent. Case-control studies, in comparison, can be utilized to examine relatively uncommon disorders. In a case-control study, one begins with a group of individuals who have the disease being studied (cases) and a group without the disease (controls). These two groups are then compared for the presence or absence of factors thought to be related to disease occurrence (odds ratio). The odds ratio is a good estimator of the relative risk, which is the quantity actually desired, if (1) the cases studied are representative of all neuropathy cases in the underlying population in terms of the exposure of interest; (2) the controls are representative of all individuals without neuropathy in the population in terms of the exposure; and (3) the disease (i.e., neuropathy) is rare. The latter condition is easily fulfilled in practice; however, if either the cases or the controls are unrepresentative, the odds ratio will not accurately reflect the true relative risk. Hospital or referral center cases and control subjects selected from these sources are rarely representative of the underlying general population in terms of potentially important risk factors. In addition, it is important that the exposure be assessed consistently in both cases and controls. Cohort studies are usually quite expensive and time consuming because of the lengthy follow-up that is often required. Although prospective studies are generally restricted to fairly common disorders, they can be employed to examine groups of people at high risk of disease, such as those exposed to a specific environmental toxin. Cohort studies are also needed if it is important to assess several different clinical outcomes that might be caused by a single risk factor. Case-controls studies can be carried out quickly and inexpensively and are required if multiple risk factors must be evaluated in conjunction with a single disease. Because of their ability to evaluate many factors simultaneously, case-control studies are especially good for generating hypotheses. The two study designs are compared in Figure 47–1, and their relative advantages and disadvantages in the study

1183

FIGURE 47–1. Design of analytic (case-control and cohort) epidemiologic studies. (From Schoenbert, B. S.: Epidemiology of the inherited ataxias. In Kark, P., Rosenberg, R., and Schut, L. [eds]: The Inherited Ataxias: Biochemical, Viral, and Pathological Studies. New York, Raven Press, p. 421, 1978, with permission.)

of neurologic disease are reviewed in more detail elsewhere.34 It is essential to keep in mind that these are observational studies, not experiments. In an experiment, the investigator controls whether or not specific individuals are exposed to a particular risk factor (i.e., the intervention). It is important that the study group (those exposed to the factor) and the control group (those not exposed) be as comparable as possible, and this is usually achieved by randomly assigning individuals to the two groups. Controlled clinical trials are the most common example of such experiments and are especially useful in evaluating the efficacy of proposed treatment modalities for peripheral neuropathy (e.g., the use of gabapentin for the treatment of pain in patients with diabetic polyneuropathy17). More often, however, it is necessary to study disease occurrence in people who cannot be subjected to environmental manipulations by the investigators. Indeed, the condition may not be diagnosed and medically evaluated until long after the patient’s exposure to a specific etiologic agent. As a consequence of this lack of control, relationships may be found that are actually the result of various biases.31 For example, carpal tunnel syndrome, a relatively common neurologic condition in the general population, has been reported in association with diabetes. However, case reports, no matter how numerous, cannot be taken as evidence for an etiologic association because patients with both carpal tunnel syndrome and diabetes are probably more likely to be referred to academic medical centers, and reported, than are patients with either condition alone. Other associations may be due to chance. This is especially likely when a large number of factors are examined for their

1184

Diseases of the Peripheral Nervous System

possible relationship to a particular disease. Alternatively, some exposures are implicated directly in disease pathophysiology, whereas others are only indirectly related to some biologically important risk factor. Thus associations between a given factor and a specific disorder cannot be equated with cause solely on the basis of an observational epidemiologic study. Instead, the assessment of causality depends on judgments involving the strength and biologic gradient of the association between the risk factor and the disease, the consistency of the association from one study to another, the specificity of the association, the temporal relation between the exposure and the onset of disease, the plausibility of the association based on current knowledge and analogous situations, and direct experimental evidence from human on animal studies.10

between diabetic neuropathy and a polymorphism at the 5⬘ end of the aldose reductase gene.15,20 This gene is located on chromosome 7q35 and is involved in the sorbitol or polyol pathway, which converts glucose to fructose utilizing the enzymes aldose reductase and sorbitol dehydrogenase. While the potential for these studies is profound, there are limitations. The two most common problems with association analyses include multiple testing (false-positive association occurring by chance as a result of examining many genes at once) and population stratification (difference in allele frequency resulting from differential ethnic makeup of case and control populations). In addition, a publication bias exists whereby there is a tendency to publish only positive findings rather than negative findings. Although these limitations need to be addressed, genetic epidemiology has the potential to contribute to a greater understanding of complex diseases such as peripheral neuropathy.

Genetic Epidemiology Advances in human genetics, including the Human Genome Project, have led to a greater understanding of the role of genetics in the etiology of complex diseases. Because of this explosion of knowledge, the field of genetic epidemiology is developing rapidly. Genetic epidemiology has been defined in a number of ways. Morton and Chung27 described it as “a science that deals with the etiology, distribution and control of disease in groups of relatives, and with inherited causes of disease in populations.” Ward43 claimed that the primary objective of the genetic epidemiologist was to identify the genetic contribution to the etiologic pathway, while Phillippe28 stated that the focus of genetic epidemiology studies was the interaction between genetic and environmental factors in the origin of disease. The genetic epidemiology of peripheral neuropathy, like that of most diseases, is in its infancy. Genetic linkage and association analyses are two distinct approaches to understanding the genetic etiology of a complex disease.9 Linkage studies employ family members and compare segregation patterns of genetic markers with the disease state. Because closely linked sequences are less likely to be separated by the DNA reshuffling that takes place during meiotic recombination than are distantly spaced sequences, markers that tend to co-segregate with the disease provide pointers to the approximate chromosomal location of the underlying disease gene. For example, type I hereditary sensory neuropathy has been linked to chromosome 9q22.2,3 A limitation of linkage studies is that they can only narrow down the chromosomal location so far. To test for a specific polymorphism, association studies are used. Association studies use a more traditional epidemiologic case-control approach to test whether specific polymorphisms involved in putative etiologic pathways are independently associated with disease status or severity. For example, multiple studies have examined the association

CLINICAL APPLICATIONS: RESULTS OF EPIDEMIOLOGIC STUDIES Despite the many problems inherent in the epidemiologic study of peripheral neuropathy, such investigations have been carried out with a remarkable degree of success, as delineated in the following sections. From its beginnings as a scientific discipline, epidemiology has characteristically focused on unusual patterns of disease occurrence, particularly on unexpected increases in disease frequency (epidemics). Among the most fruitful applications of epidemiology has been identification of the environmental agents responsible for outbreaks of peripheral neuropathy. An increase in disease frequency may also be secondary to increases in other diseases. An example of this is peripheral neuropathy secondary to human immunodeficiency virus (HIV) infections. To illustrate the utility of epidemiologic approaches in such situations, epidemics of neuropathy resulting from intoxication with tri-o-cresyl phosphate or a nutrient-deficient diet and secondary to HIV are discussed briefly. Clusters of peripheral neuropathy may also occur among members of the same family. It should be emphasized that multiple cases of disease within a family may be related to environmental as well as genetic influences because family members typically share both. With some familial cases, however, it is possible to document a pattern of inheritance that identifies the disorder as genetically determined. Perhaps the most common of the hereditary neuropathies is Charcot-Marie-Tooth disease, and the epidemiology of this condition is reviewed. Epidemiologic methods are also useful for investigating diseases that are constantly present in the population (endemic). Emphasis has been placed on the peripheral neuropathies that are particularly important and those that are

Epidemiologic Approaches to Peripheral Neuropathy

easiest to study because they have an acute onset and clearcut symptoms. An example of the latter group is postherpetic neuralgia, and its epidemiologic features are described. Guillain-Barré syndrome, in contrast, is a potentially lifethreatening neuropathy, and a number of epidemiologic investigations related to this condition are discussed. Peripheral neuropathies that do not have these characteristics (i.e., occurring in clusters, having an acute onset or clear-cut symptoms, or being life threatening) are much harder to study. The difficulties associated with the study of diabetic neuropathy are reviewed to illustrate the problems that may be encountered.

Tri-o-cresyl Phosphate Intoxication An example of the usefulness of descriptive epidemiologic techniques is provided by a series of epidemics related to tri-o-cresyl phosphate poisoning. In 1959, an outbreak of neurologic disease was reported from Morocco.40 The illness began with aching pain in the calves, following by paresthesias and numbness distally in the lower extremities. Usually there was some improvement in the sensory symptoms, followed by weakness, which typically began distally in the legs. In some cases, distal muscles of the upper extremities also became weak. On examination, the ankle jerks were diminished or absent but knee jerks were usually exaggerated, suggesting upper motor involvement as well. Because of the severe neuropathy, there was no response on plantar stimulation. All routine laboratory tests were negative, and examination of cerebrospinal fluid did not reveal any abnormalities. The distribution of the disease provided some interesting clues regarding etiology, however. Within about 3 weeks of the first reported case, there were approximately 200 to 300 new cases daily, and over 2000 cases were documented within the first month. All ages and both sexes were affected. The outbreak was concentrated in Meknes and nearby towns and appeared to involve primarily poor Muslim residents of the area. Jews and Europeans were not affected, nor were the Muslims of higher socioeconomic status; there were also very few cases among the most indigent Muslims. Over 250,000 religious pilgrims visited the area during the outbreak but did not develop the disease. Furthermore, of the 100 soldiers stationed in Meknes, the only two who developed the disease usually ate their meals in the town and not in the barracks. Prisoners also appeared to be spared, although some developed the disease a few days after their release from jail. Because the most indigent Muslims were spared and the large number of religious pilgrims were unaffected, an infectious etiology seemed unlikely. The disease was related, instead, to the consumption of food prepared in Meknes. The physician in charge of the Meknes dispensaries reported that several patients believed that a dark-colored cooking oil was responsible for the ill-

1185

ness. The doctor had seen samples, which looked more like a lubricant for motors that the usually light-colored cooking oil. Investigation revealed that the same wholesaler supplied oil to all of the affected areas, and the dark-colored product was first available about the same time that the disease began to appear. These observations were consistent with the pattern of the epidemic. The Jews and Europeans had their own markets and the pilgrims brought their own food with them, and none was affected. Muslims in the higher socioeconomic strata could afford a better brand of cooking oil, and the poorest could not buy any oil. When analyzed, the oil was found to contain a manufactured lubricant consisting largely of cresyl phosphates; such oils are generally used in turbojet engines. This toxic material appeared to be the responsible agent.

Nutrient-Deficient Diets During 1992 and 1993, an epidemic of optic and peripheral neuropathy affected over 50,000 Cubans.14 Although Cuban epidemiologists and physicians responded quickly, identification of possible etiologic agents proved to be elusive. Few children were affected, and the majority of cases were adults between the ages of 25 and 64 years. Males were more likely to have the optic form, while the peripheral form was more frequent in women. Initial case-control studies by the Cuban Ministry of Health indicated a significant association of disease with smoking, missing meals, weight loss, high sugar consumption, and heavy drinking.24 Many patients reported lower milk, fish, and soy “hamburger” consumption and higher rice, beans, and cabbage consumption than controls. The Cuban economy had declined rapidly, and food availability dwindled soon after loss of trade with the Eastern Block countries, compounding the negative effects caused by the continuing U.S. trade embargo.14 As a consequence, the diet of the general population changed dramatically in the years just preceding the epidemic. Milk was restricted to children and pregnant women, soy hamburger was limited to one serving per person once or twice per month, and bread was also rationed. Because of the severity of the food shortages, vitamin deficiencies were recognized as a likely factor in causation of the epidemic, but the symptoms did not present as classic vitamin deficiency. Efforts to identify toxic substances that may have interacted with vitamin deficiency, including use of cassava and other cyanidecontaining foods, were begun but were inconclusive. Other potential factors, such as enteroviruses, were also examined without success.

Charcot-Marie-Tooth Disease The prevalence of Charcot-Marie-Tooth disease and other hereditary neuropathies has been estimated for a number of distinct populations (Table 47–2). Prevalence rates are highest in western Norway,39 but it is not certain if this

1186

Diseases of the Peripheral Nervous System

Table 47–2. Prevalence of Hereditary Motor and Sensory Neuropathies by Region Prevalence per 100,000 Researcher(s)

Year

Region

Boeters Herndon Brewis et al. Chen et al. Gudmundsson Kondo et al. Skre Davis et al. Brooks and Emery Lucci et al. Combarros et al. Hagberg and Westerberg Holmberg

1938 1954 1966 1968 1969 1970 1974 1978 1982 1982 1982 1982 1993

Sleisien North Carolina Carlistee (UK) Guam Iceland Ryuku Island (Japan) Western Norway Northeast England Edinburgh (UK) Reggio Emilia, Italy Cantabria (Spain) Southwest Sweden Northern Sweden

Population

All Cases

Autosomal Dominant Cases

4,500,000 3,500,000 70,000 38,000 187,200 287,727 725,000 2,500,000 — 412,324 510,000 263,400 518,742

2 5.4 1.4 23.7 1.6 3.4 41.0 4.7 — 3.6 32.0 21.6 20.1

— — — 16.0 1.6 — 36.0 4.1 4.8 2.9 30.0 16.7 16.2

Data from Chiò, A., Tribolo, A., Brignolio, F., et al.: Hereditary motor and sensory neuropathies: a genetic and epidemiological study in the province of Turin, Italy. Ital. J. Neurol. Sci. 8:369, 1987.

is due to more complete case ascertainment, extensive inbreeding, or the predominance of different forms of the disease in different populations.35 In Norway, three genetic patterns of inheritance were distinguished39; an autosomal dominant variety was most common (36 per 100,000), followed by an X-linked recessive form (3.6 per 100,000) and an autosomal recessive form (1.4 per 100,000), although the identification of X-linked inheritance in this study has been criticized.13 Family members of affected individuals had a confusing array of other neurologic problems. These observations emphasize the point that a particular clinical syndrome may represent a number of distinct pathophysiologies and, conversely, that a specific pathophysiologic process may produce a variety of clinical manifestations. Epidemiologic characteristics may help discriminate the different entities.

Postherpetic Neuralgia Pain lasting for a month or more occurs in about 9% of patients with herpes zoster in community studies.18,32 Higher figures from specialty practices16,30 reflect the disproportionate referral of complicated cases. Accurate assessment of the risk of neuralgia, then, depends on complete ascertainment of herpes zoster. Among the residents of Rochester, Minnesota, the incidence of herpes zoster over a 15-year period was 1.3 per 1000 per year.32 Higher rates based on surveys of general practices in Cirencester, England,18 and Hawick, Scotland,25 yielded incidence rates of 3.4 and 4.8 per 1000 per year, respectively. A third survey, based on the experience of physicians in Edinburgh, Scotland,38 produced an incidence rate of 2.0 per 1000 per year. The lower rate for Rochester suggests either major geographic variation in the incidence of the disorder or an

artifact resulting from different methodologic approaches; there is a high degree of case ascertainment in Rochester, however, and it is doubtful that many patients were missed. In none of the four studies was there a marked differential in the risk of herpes zoster by sex. Postherpetic neuralgia, in contrast, occurs more often in women. The apparent discrepancy is explained by the fact that persistent pain is more frequent among older patients—the average age of patients with postherpetic neuralgia in one study was 21 years greater than that of herpes zoster patients generally.32 Because there are more elderly women than elderly men, women outnumber men at these ages even though sex-specific incidence rates may be similar. The pain associated with postherpetic neuralgia generally follows the dermatomal distribution of the zoster44; the reported predilection for ophthalmic involvement most likely represents selection bias in patients seen in dermatology and ophthalmology practices.32 Similarly, the proportion of patients with persistent pain at 1 year is much higher in referral settings than in community studies.

Guillain-Barré Syndrome During the fall of 1976, a campaign was initiated in the United States to immunize the adult population against a particular strain of influenza A virus (influenza A/New Jersey or “swine flu”). Between October 1 and December 21, 1976, 172 cases of Guillain-Barré syndrome were reported to the Centers for Disease Control (CDC), 99 of which occurred among patients who had been immunized.4 In the majority of patients for whom the time of immunization and onset of neurologic symptoms could be ascertained, the neurologic dysfunction began between 8 and 28 days following immunization. Physicians in four states were subsequently

Epidemiologic Approaches to Peripheral Neuropathy

surveyed to identify cases of Guillain-Barré syndrome with onset between October 16 and December 14, 1976. Twentytwo cases among vaccinated individuals yielded an estimated annual incidence rate of 8.8 cases per 100,000 per year compared to a rate of 1.2 cases per 100,000 per year in the nonvaccinated population. Because of this apparent increase, the CDC launched a nationwide survey of Guillain-Barré syndrome, with particular attention to the association with immunization against influenza A/New Jersey.36 The results of this effort are contrasted with other studies in Table 47–3. Although the rate among the vaccinated was not as high as the figure generated by the smaller four-state survey, the excess risk among those receiving immunization against influenza A/New Jersey was still apparent. Interestingly, no excess in Guillain-Barré syndrome was seen with previous influenza immunization programs23 or following the programs of 1978–1979 and 1979–1980,37 suggesting that the 1976 influenza A/New Jersey vaccine had a specific association with the syndrome. This finding might be explained by ascertainment bias if the vaccinated group were more health conscious and likely to seek medical treatment than the unvaccinated individuals. In evaluating rates for the unvaccinated over age 17 years in the U.S. survey (see Table 47–3) in relation to comparable rates for Olmsted County, Minnesota (with a high level of case ascertainment through a records-linkage system), there appears to be underreporting in the survey data for the unvaccinated group. Ascertainment bias could also result from more complete reporting of vaccinated cases by attending physicians.23 This potential problem was compounded by the fact that the surveillance program did not establish diagnostic criteria for Guillain-Barré syndrome or evaluate suspected cases in a standardized way. Subsequent review has established that substantial overreporting occurred but to a similar extent in vaccinated and unvaccinated patients, so that the association was still present between vaccination and the onset of Guillain-Barré syndrome within the subsequent 6 weeks.33 The most attractive hypotheses concerning pathogenesis are that Guillain-Barré syndrome represents a primary viral infection or an immune reaction resulting from an antecedent illness or immunization.12 The immunologic hypothesis has support from laboratory investigations in which animals immunized with peripheral nerve and Freund’s adjuvant develop pathologic changes resembling those of the Guillain-Barré syndrome.42 A substantial proportion of patients have had infections antecedent to the onset of Guillain-Barré syndrome,21 but a case-control study of 52 patients compared with two sets of controls failed to support the theory that the syndrome is a specific primary viral disease because the viruses identified more often in cases were of diverse types.26 There was a significant association with respiratory infections occurring within 1 month of the onset of neurologic symptoms, but influenza was an uncommon antecedent infection.

1187

Another investigation found statistically significant increased titers to cytomegalovirus among Guillain-Barré cases as compared with two groups of controls (healthy adults and people with other neurologic diseases), but other possible differences between cases and controls (e.g., age) that might account for the difference in antibody titer were not addressed.5 A final study compared cases with age- and sex-matched controls who had idiopathic Bell’s palsy.22 There was a statistically significant association with antecedent illness within 4 weeks of onset of the Guillain-Barré syndrome but no association between Guillain-Barré syndrome and prior immunization, allergic or metabolic disorders, or exposure to toxins. However, there may have been overmatching of cases and controls if, as some suggest, Bell’s palsy and Guillain-Barré syndrome have similar causes. In that event, cases and controls will resemble each other with regard to causative factors, and no differences between the two groups will be discovered. Recent research indicates that Guillain-Barré syndrome is not a unitary disorder with variations. In fact, the syndrome includes several distinctive subtypes,1 with critical differences among the subtypes with regard to pathologic and electrodiagnostic features.11

Diabetic Neuropathy There is a great deal of conflicting information in the literature concerning even the most basic aspects of the epidemiology of diabetic neuropathy. In part, this is related to problems with definitions: Depending on the group of patients studied and the criteria used, the prevalence of neuropathy among diabetic individuals varies from 10% to 100% (Table 47–4). In general, as diagnostic criteria become more sensitive, the disease becomes more common (e.g., the apparent incidence and/or prevalence increases as milder cases are identified) and, simultaneously, the disease becomes less severe on average. Thus tests that measure abnormalities of nerve conduction indicate that the majority of diabetics, and after years of diabetes practically all of them, have neuropathy, but the condition is subclinical in many.6 Conversely, evaluation based on clinical symptomatology will yield a smaller proportion of patients with neuropathy but will miss some asymptomatic individuals with a significant degree of objective impairment.6,29 Additional problems result when diagnostic capabilities or standards change over time. In the first description of the incidence of neuropathy in a population-based cohort with non–insulin-dependent diabetes mellitus (NIDDM), 995 Rochester residents who did not have peripheral neuropathy at the time of diabetes diagnosis developed distal polyneuropathy at the rate of 7.8 cases per 100 person-years of follow-up. Autonomic neuropathy was least frequent at 0.3 per 100, while the incidence of femoral neuropathy, cranial nerve dysfunction, or radiculopathy was 0.6 per 100 per year, and carpal tunnel syndrome

1996–97

117

432* 516*

112

89

18

13 3 5 40

Number of Cases

1.5

1.3* 3.4*

1.2

0.8

0.7 0.6 1.9 1.7 2.3* 1.2

Incidence Rate (Cases/100,000/year)

Diagnosis verified by neurologist

Diagnosed by a physician; objective evidence of muscle involvement

None given None given None given Acute or subacute bilateral weakness in absence of concurrent febrile illness Acute or subacute onset of polyradiculopathy in the absence of a known cause (e.g., diabetes mellitus), diffuse symmetrical lower motor neuron weakness more prominent than sensory findings, and fewer than 10 white blood cells/mm3 in cerebrospinal fluid Acute or subacute onset with evolution of signs within 2 mo, bilateral roughly symmetrical weakness of lower motor neuron type, absence of conditions known to produce polyneuropathy, and fewer than 10 white blood cells/mm3 in cerebrospinal fluid; patients with a definite sensory level were excluded None given

Criteria for Defining Cases

*Over age 17 years. Modified from Schoenberg, B. S.: Epidemiology of Guillain-Barré syndrome. In Schoenberg, B. S. (ed.): Neurological Epidemiology: Principles and Clinical Applications. New York, Raven Press, p. 249, 1978.

Cheng et al. (2000)

Oct. 1, 1976– Jan. 31, 1977

Schonberger et al. (1979)

Unvaccinated Vaccinated against influenza A/New Jersey Sweden

1957–77

Nyland (1978)

Norway: midwestern region United States

1969–72

Soffer et al. (1978)

1972–76

Hogg et al. (1978)

Israel

1954–63 1955–61 1960–66 1935–76

Gudmundsson (1969) Brewis et al. (1966) Chen et al. (1968) Kennedy et al. (1978)

Iceland Carlisle, England Guam Olmsted County, Minnesota San Joaquin County, California

Years Studied

Study

Area Surveyed

Table 47–3. Incidence of Guillain-Barré Syndrome in Well-Defined Populations

1189

Epidemiologic Approaches to Peripheral Neuropathy

Table 47–4. Reported Prevalence of Symptoms and Signs of Neuropathy in Diabetes Data Source, Year

Measurement

Cleveland, 1953 Salford, England, 1963 Brussels, 1965 Stockholm, 1950 Rochester, MN, 1961 Philadelphia, 1958 New York, 1952 London, 1960 Toronto, 1961 Cincinnati, 1951 Chicago, 1966 Aarhus, Denmark, 1968 London, 1971 Edinburgh, 1977

Subjective complaints General findings Objective findings Objective findings Electromyography, objective signs Impotence Skin vessel dilatation Abnormal Valsalva maneuver Objective signs General signs Objective signs, motor conduction velocity Motor conduction velocity Motor conduction velocity Motor conduction velocity, autonomic vascular tests

Number of Patients

Neuropathy (%)

261 100 1175 150 103 198 16 337 100 77 107 14 39

62 57 21 49 42 55 44 20 52 35 10 100 100

10

100

Data from National Institutes of Health: Diabetes Data: Compiled 1977 (DHEW Pub. No. [NIH] 78–1468). Washington, D.C.: U.S. Department of Health, Education, and Welfare, 1978.

occurred at an annual rate of 1.3 per 100. Peripheral neuropathies not necessarily related to diabetes (i.e., meralgia paresthetica, alcoholic neuropathy, Bell’s palsy, and peroneal palsy) developed at a rate of 1.1 per 100. In reality, it may be extremely difficult to classify a specific patient (e.g., a hypothetical diabetic patient with peripheral neuropathy who also has a history of alcoholism and occupational exposure to known neurotoxic agents). These figures may also represent underestimates in that patients had to meet defined clinical criteria to be included in these tabulations. Furthermore, the data were based on a retrospective review of clinical records; patients were not followed prospectively using a defined protocol, nor were they specifically requested to have standardized medical evaluations for research purposes. A second study in Rochester, the Rochester Diabetic Neuropathy Study (RDNS), examined the prevalence by staged severity of various types of diabetic neuropathy in a representative population of diabetics living in the community.7 This study used cross-sectional and prospective study designs and extensive standardized testing for diabetic neuropathy. Since it is not possible to adequately characterize and quantitate diabetic polyneuropathies using only one or two clinical or test abnormalities, results of neurologic examination and abnormalities of nerve conduction, quantitative sensation tests, and quantitative autonomic tests were combined in a composite score to estimate the severity of diabetic polyneuropathy.8 This composite score appears to provide a much more comprehensive and stable numeric score by which to diagnose and grade severity of diabetic polyneuropathy than using individual clinical or test results. The RDNS enrolled 380 individuals living in Rochester recognized to have diabetes mellitus on January 1, 1986.7 Of these, 102 had insulin-dependent diabetes mellitus (IDDM),

and 278 had NIDDM. Sixty-six percent of IDDM patients and 59% of NIDDM patients had some form of neuropathy. However, symptomatic degrees of polyneuropathy occurred in only 15% of IDDM and 13% of NIDDM patients.

SUMMARY Peripheral neuropathies are difficult diseases to study epidemiologically. Nonetheless, the investigation of clusters of cases has proved particularly rewarding with regard to diseases induced by toxins and conditions such as the GuillainBarré syndrome. Despite the many problems discussed in this chapter, however, the difficulties of disease definition, classification, and diagnosis are being addressed and should lead to significant advances in our understanding of the epidemiology of this group of disorders.

REFERENCES 1. Asbury, A. K., and McKhann, G. M.: Changing views of Guillain-Barre syndrome. Ann. Neurol. 41:287, 1997. 2. Bejaoui, K., McKenna-Yasek, D., Hosler, B. A., et al.: Confirmation of linkage of type 1 hereditary sensory neuropathy to human chromosome 9q22. Neurology 52:510, 1999. 3. Blair, I. P., Dawkins, J. L., and Nicholson, G. A.: Fine mapping of the hereditary sensory neuropathy type I locus on chromosome 9q22.1:q22.3: exclusion of GAS1 and XPA. Cytogenet. Cell Genet. 78:140, 1997. 4. Centers for Disease Control: Guillain-Barre syndrome— United States. Mor. Mortal. Wkly. Rep. CDC Surveill. Summ. 25:401, 1976.

1190

Diseases of the Peripheral Nervous System

5. Dowling, P., Menonna, J., and Cook, S.: Cytomegalovirus complement fixation antibody in Guillain-Barre syndrome. Neurology 27:1153, 1977. 6. Dyck, P. J., Karnes, J. L., Daube, J., et al.: Clinical and neuropathological criteria for the diagnosis and staging of diabetic polyneuropathy. Brain 108(Pt. 4):861, 1985. 7. Dyck, P. J., Kratz, K. M., Karnes, J. L., et al.: The prevalence by staged severity of various types of diabetic neuropathy, retinopathy, and nephropathy in a population-based cohort: the Rochester Diabetic Neuropathy Study. Neurology 43:817, 1993. 8. Dyck, P. J., Melton, L. J. 3rd, O’Brien, P. C., and Service, F. J.: Approaches to improve epidemiological studies of diabetic neuropathy: insights from the Rochester Diabetic Neuropathy Study. Diabetes 46(Suppl. 2):S5, 1997. 9. Emahazion, T., Feuk, L., Jobs, M., et al.: SNP association studies in Alzheimer’s disease highlight problems for complex disease analysis. Trends Genet. 17:407, 2001. 10. Fletcher, R. H., Fletcher, S. W., and Wagner, E. H.: Clinical Epidemiology: The Essentials, 3rd ed. Baltimore, Williams & Wilkins, 1996. 11. Govoni, V., and Granieri, E.: Epidemiology of the GuillainBarre syndrome. Curr. Opin. Neurol. 14:605, 2001. 12. Hahn, A. F.: Guillain-Barre syndrome. Lancet 352:635, 1998. 13. Harding, A. E., and Thomas P. K.: Genetic aspects of hereditary motor and sensory neuropathy (types I and II). J. Med. Genet. 17:329, 1980. 14. Hedges, T. R. 3rd, Hirano, M., Tucker, K., and Caballero, B.: Epidemic optic and peripheral neuropathy in Cuba: a unique geopolitical public health problem. Surv. Ophthalmol. 41:341, 1997. 15. Heesom, A. E., Millward, A., and Demaine, A. G.: Susceptibility to diabetic neuropathy in patients with insulin dependent diabetes mellitus is associated with a polymorphism at the 5⬘ end of the aldose reductase gene. J. Neurol. Neurosurg. Psychiatry 64:213, 1998. 16. Helgason, S., Petursson, G., Gudmundsson, S., and Sigurdsson, J. A.: Prevalence of postherpetic neuralgia after a first episode of herpes zoster: prospective study with long term follow up. BMJ 321:794, 2000. 17. Hemstreet, B., and Lapointe, M.: Evidence for the use of gabapentin in the treatment of diabetic peripheral neuropathy. Clin. Ther. 23:520, 2001. 18. Hope-Simpson, R. E.: The nature of herpes zoster: a long-term study and a new hypothesis. Proc. R. Soc. Med. 58:9, 1965. 19. Hulley, S. B., and Cummings, S. R.: Designing Clinical Research: An Epidemiologic Approach. Baltimore, Williams & Wilkins, 1988. 20. Ichikawa, F., Yamada, K., Ishiyama-Shigemoto, S., et al.: Association of an (A-C)n dinucleotide repeat polymorphic marker at the 5⬘-region of the aldose reductase gene with retinopathy but not with nephropathy or neuropathy in Japanese patients with type 2 diabetes mellitus. Diabet. Med. 16:744, 1999. 21. Kaplan, J. E., Schonberger, L. B., Hurwitz, E. S., and Katona, P.: Guillain-Barre syndrome in the United States, 1978–1981: additional observations from the national surveillance system. Neurology 33:633, 1983. 22. Kennedy, R. H., Danielson, M. A., Mulder, D. W., and Kurland, L. T.: Guillain-Barre syndrome: a 42-year epidemiologic and clinical study. Mayo Clin. Proc. 53:93, 1978.

23. Kurland, L. T., Wiederholt, W. C., Kirkpatrick, J. W., et al.: Swine influenza vaccine and Guillain-Barre syndrome: epidemic or artifact? Arch. Neurol. 42:1089, 1985. 24. Mas Bermejo, P., Rodriquez, R., et al.: Neuropatia Epidemica en Cuba: Un Analisis Epidemiologico. La Habana, Ministerio de Salud Publica, p. 1, 1993. 25. McGregor, R. M.: Herpes zoster, chicken-pox, and cancer in general practice. Br. Med. J. 1:84, 1957. 26. Melnick, S. C., and Flewett, T. H.: Role of infection in the Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 27:395, 1964. 27. Morton, N. E., and Chung, C. S.: Genetic Epidemiology. New York, Academic Press, 1978. 28. Philippe, P.: L’epidemiologie genetique, fondement d’une veritable prevention: l’approache et ses resultats. Can. J. Public Health 73:350, 1982. 29. Pirart, J.: Diabetic neuropathy: a metabolic or a vascular disease? Diabetes 14:1, 1965. 30. Portenoy, R. K., Duma, C., and Foley, K. M.: Acute herpetic and postherpetic neuralgia: clinical review and current management. Ann. Neurol. 20:651, 1986. 31. Ragozzino, M. W., and Melton, L. J. 3rd: Disease associations: need for increased scrutiny of the literature. Mayo Clin. Proc. 59:719, 1984. 32. Ragozzino, M. W., Melton, L. J. 3rd, Kurland, L. T., et al.: Population-based study of herpes zoster and its sequelae. Medicine (Baltimore) 61:310, 1982. 33. Safranek, T. J., Lawrence, D. N., Kurland, L. T., et al.: Reassessment of the association between Guillain-Barre syndrome and receipt of swine influenza vaccine in 1976–1977: results of a two-state study. Expert Neurology Group. Am. J. Epidemiol. 133:940, 1991. 34. Schoenberg, B. S.: Analytic, experimental, and theoretical epidemiology. In Schoenberg, B. S. (ed.): Neurological Epidemiology: Principles and Clinical Applications. New York, Raven Press, p. 395, 1978. 35. Schoenberg, B. S.: Epidemiology of the inherited ataxias. Adv. Neurol. 21:15, 1978. 36. Schonberger, L. B., Bregman, D. J., Sullivan-Bolyai, J. Z., et al.: Guillain-Barre syndrome following vaccination in the National Influenza Immunization Program, United States, 1976–1977. Am. J. Epidemiol. 110:105, 1979. 37. Schonberger, L. B., Hurwitz, E. S., Katona, P, et al.: GuillainBarre syndrome: its epidemiology and associations with influenza vaccination. Ann. Neurol. 9(Suppl.):31, 1981. 38. Seiler, H. E.: Study of herpes zoster particularly in its relationship to chickenpox. J. Hyg. 47:253, 1949. 39. Skre, H.: Genetic and clinical aspects of Charcot-Marie-Tooth’s disease. Clin. Genet. 6:98, 1974. 40. Smith, H. V., and Spalding, J. M. K.: Outbreak of paralysis in Morocco due to ortho-cresyl phosphate poisoning. Lancet 2:1019, 1959. 41. Stamp, J.: Some Economic Factors in Modern Life. London, P. S. King and Son, 1929. 42. Waksman, B. S., and Adams, R. D.: Allergic neuritis: experimental disease of rabbits induced by the injection of peripheral nervous tissue and adjuvants. J. Exp. Med. 102:213, 1955. 43. Ward, R. H.: Genetic epidemiology: promise or compromise. Soc. Biol. 27:87, 1979. 44. Watson, P. N., and Evans, R. J.: Postherpetic neuralgia: a review. Arch. Neurol. 43:836, 1986.

B. Diseases of Cranial Nerves

48 Diseases of the Third, Fourth, and Sixth Cranial Nerves JACQUELINE A. LEAVITT AND BRIAN R. YOUNGE

Overview Nuclei of the Third, Fourth, and Sixth Nerves Cranial Nerve III (Oculomotor) Cranial Nerve IV (Trochlear) Cranial Nerve VI (Abducens) Courses of the Third, Fourth, and Sixth Nerves

Cranial Nerve III Cranial Nerve IV Cranial Nerve VI Cavernous Sinus Blood Supply Histologic Features Clinical Features Congenital Defects

This chapter updates the previous work of James C. Trautmann and Charles R. Barnett, describing the nuclear anatomy, relationships, and intracranial course of the third, fourth, and sixth cranial nerves, their relationships within the cavernous sinus, the blood supply to each nerve, and their histologic compositions. We describe the clinical features of a palsy of each nerve, with some special mention of partial palsy of the third nerve in its anatomic relationships. A discussion of congenital defects involving the three oculomotor nerves is next, and some of the more peculiar associated syndromes, such as the Möbius syndrome and Duane’s syndrome, are included, along with substitution syndromes. Trauma plays an important role in the etiology of many cranial nerve palsies, and we have devoted some discussion to the mechanisms of injury and the relationships of bony and ligamentous anatomy to the nerves as they course by. The recovery rate after surgical trauma is usually good, and we refer to a paper showing surprisingly favorable prognosis for traumatic sixth nerve palsies. Theories of aberrant regeneration of the third nerve are included. Tumors can produce isolated palsies, but more often associated signs are present, anywhere

Effects of Trauma Aberrant Regeneration of Cranial Nerve III Neoplasms Vascular Diseases and Aneurysms Infections and Toxins

from within the brainstem to the orbit and even remotely, as in the false localizing signs of increased intracranial pressure or from the paraneoplastic syndromes. The frequency and important clinical findings of vascular diseases that affect the three ocular motor nerves emphasizes the localizing value of vascular syndromes within the brainstem. It is important to suspect an aneurysm involving the interpeduncular portion of the third nerve, which almost always causes early pupillary dilatation. Diabetes and other less common vascular diseases are also discussed in this section. The variety of infections that can cause paresis of one or more of the oculomotor nerves is legion, but the frequency is not common; similarly, toxins and drugs that produce diplopia are many and varied.

OVERVIEW Cranial nerves III, IV, and VI and their nuclei constitute the lower motor neuron pathways for the extraocular and intraocular muscles. Cranial nerve III (the oculomotor nerve) innervates the superior rectus, medial rectus, inferior 1191

1192

Diseases of the Peripheral Nervous System

rectus, and the inferior oblique muscles of the globe as well as the levator muscle of the upper eyelid. Parasympathetic neurons to the intraocular muscles of the ciliary body and iris sphincter also course through the oculomotor nerve. Cranial nerve IV (the trochlear nerve) innervates the superior oblique muscle. Cranial nerve VI (the abducens nerve) innervates the lateral rectus muscle. These three cranial nerves, and their supranuclear connections, provide the neural pathways subserving the complex mechanisms of ocular motility. The literature on acquired ocular palsies has repeatedly emphasized the existence of a relatively large group of cases for which no cause is apparent or determined despite long-term follow-up.92–94,96 Richards et al.92 combined data from several series of patients at one institution and found that almost 25% of their 4278 cases had no determinable cause. These reports emphasize not only the variety of causal factors for disturbance of function of cranial nerves III, IV, and VI, but also the limitation of our present knowledge concerning the pathophysiology of this portion of the peripheral nervous system. This chapter is not an all-inclusive treatise on the diseases of cranial nerves III, IV, and VI, but rather a review of the common disease processes that affect these nerves and of the importance of the third, fourth, and sixth cranial nerves as a part of the peripheral nervous system.

NUCLEI OF THE THIRD, FOURTH, AND SIXTH NERVES Cranial Nerve III (Oculomotor) The oculomotor nuclei are located in the mesencephalon at the level of the superior colliculi. These paired midline structures are located ventrolateral to the aqueduct of Sylvius and are bordered laterally by the median longitudinal fasciculi. Warwick119 compiled studies of the anatomy of the oculomotor nuclei and proposed a scheme for the nuclei in primates. He warned that Wolff’s diagram (Fig. 48–1) appears more precise than is actually the case and cautioned against its direct application to humans. The motor neurons of the oculomotor nucleus extend over a large area and are arranged in a rostrocaudal and dorsoventral pattern. Most of the neurons are situated in two large lateral nuclei, the dorsal group, representing the neurons of the ipsilateral inferior rectus muscle, and the ventral group, representing the neurons of the ipsilateral medial rectus muscle. The inferior oblique muscle is represented by neurons in the intermediate area of the ipsilateral lateral nucleus. Warwick118 documented contralateral innervation of the superior rectus muscles. After unilateral interruption of the oculomotor nerve, chromatolysis is evident in the ipsilateral oculomotor nucleus and also in the caudal two

FIGURE 48–1 Cranial nerve III (oculomotor) nucleus of monkey. Lines (A, B, C, and D) on the right locate sites of sections shown on the left. AN ⫽ anteromedial nucleus; EW ⫽ Edinger-Westphal nucleus; CCN ⫽ central caudal nucleus; IV ⫽ cranial nerve IV. (From Wolff, E.: Anatomy of the Eye and Orbit: Including the Central Connections, Development, and Comparative Anatomy of the Visual Apparatus, 6th ed. [revised by Last, R. J.]. London, H. K. Lewis & Company, p. 283, 1968, with permission.)

thirds of the contralateral oculomotor nucleus. These cells in the caudal two thirds of the contralateral nucleus are defined as the “motor pool” of the superior rectus muscle. This seems to be the only extraocular muscle supplied by the contralateral oculomotor nucleus. The motor neurons of the nerves to the levator muscle of the upper eyelid of each eye are located in the central caudate nucleus. This midline dorsal group of neurons is also the most caudal portion of the oculomotor complex and, being unpaired, is believed to supply the levator muscle of the upper eyelids of both eyes.

Diseases of the Third, Fourth, and Sixth Cranial Nerves

The Edinger-Westphal nucleus consists of a paired group of small neurons in the dorsorostral portion of the oculomotor complex. The Edinger-Westphal complex, together with the anteromedian nucleus, gives rise to oculomotor axons, which become the preganglionic parasympathetic fibers to the ciliary ganglion and continue to the eye via the short ciliary nerves. Ciliary ganglionectomy interrupts the ocular parasympathetic supply and also causes retrograde changes in both the Edinger-Westphal and anteromedian nuclei. Warwick121 estimated that 97% of the ocular parasympathetic axons supply the ciliary muscle and mediate accommodation, with only 3% supplying the pupil light reaction.

Cranial Nerve IV (Trochlear) The nucleus of cranial nerve IV is located caudal to the oculomotor nucleus, beneath the periaqueductal gray matter at the level of the inferior colliculus in the midbrain. This nerve consists of large motor neurons that supply the contralateral superior oblique muscle; its fibers decussate around the central gray matter and emerge from the brainstem dorsally through the midline medullary velum caudal to the inferior colliculus. Wolff125 described the trochlear nerve as the most slender of the cranial nerves and the one with the longest intracranial course (75 mm). Although the overlying tentorium may offer some protection against compressive injury to the trochlear nerve in its course from the dorsal midbrain, its attachment to the anterior medullary velum is very delicate, and the nerve may easily be detached at this point.

Cranial Nerve VI (Abducens) The sixth cranial nerve nucleus is located close to the midline of the pons, beneath the floor of the fourth ventricle. It is bordered on its medial side by the median longitudinal fasciculus and capped by the facial colliculus, an indentation in the floor of the fourth ventricle made by the facial nerve (cranial nerve VII) as it travels around cranial nerve VI and its nucleus on its circuitous route through the pons.

COURSES OF THE THIRD, FOURTH, AND SIXTH NERVES Cranial Nerve III The oculomotor nerve is composed of approximately 15,000 fibers, most of which are large myelinated motor fibers that arise from the bulk of the oculomotor nucleus. Smaller myelinated fibers from the Edinger-Westphal and anteromedian nuclei also travel with the ipsilateral oculomotor nerve and provide the parasympathetic innervation to the ciliary muscle and the iris sphincter.

1193

The oculomotor fibers congregate into fascicles within the midbrain; descend medially; pass through the median longitudinal fasciculus, the tegmentum, the red nucleus, and the medial margin of the substantia nigra; and emerge as 10 to 15 rootlets on the medial aspect of the cerebral peduncle. The rootlets soon fuse into a single nerve trunk, which becomes covered by a well-formed pial sheath. This nerve trunk passes obliquely forward, downward, and laterally at the level of the tentorium. Shortly after leaving the brainstem, it passes between the posterior cerebral and superior cerebellar arteries. As it approaches the cavernous sinus, it crosses the free edge of the reflected leaf of the tentorium and enters the cavernous sinus lateral to the posterior clinoid process (Fig. 48–2).

Cranial Nerve IV The trochlear nerve emerges from the dorsal surface of the brainstem below the inferior colliculus. Cranial nerve IV passes laterally around the superior cerebral peduncle and through the quadrigeminal, ambient, crural, and pontomesencephalic cisterns. It then passes between the superior cerebellar artery and the posterior cerebellar artery. The fourth nerve pierces the dura mater below the entry of the oculomotor nerve into the cavernous sinus.

Cranial Nerve VI The abducens nerve is composed of approximately 6000 to 7000 nerve fibers125 that pass forward through the pons and emerge on the ventral surface of the brainstem between the pons and the medulla, just lateral to the pyramidal prominence. Some of its rootlets pass through the pyramidal tract. Unlike cranial nerves III and IV, the rootlets of cranial nerve VI may remain as separate fascicles for varying distances before merging as a common nerve trunk. As the nerve passes upward along the ventral surface of the pons, the anterior inferior cerebellar and anterior auditory arteries, as well as other small lateral branches of the basilar artery, cross it. In the posterior fossa, its course is upward and slightly lateral; it pierces the dura of the clivus approximately 1 cm below the crest of the petrous portion of the temporal bone and approximately 1.5 cm from its site of exit from the pons. Beneath the dura, it is in close proximity to the inferior petrosal sinus and ascends to the sharp border of the petrous portion of the temporal bone, at which point it bends forward beneath the petrosphenoid ligament of Gruber before entering the cavernous sinus. In this region, it is often within the inferior petrosal sinus as it passes through Dorello’s canal.

Cavernous Sinus Cranial nerve III travels horizontally within the superior lateral wall of the cavernous sinus with cranial nerve IV

1194

Diseases of the Peripheral Nervous System

FIGURE 48–2 Representation of cranial nerves and related basal structures in area of midbrain, middle fossa, cavernous sinus, and orbital apex. Roof of right orbit, sphenoid wing, floor of middle fossa, and petrous ridge have been sectioned; a section of the tentorial edge is removed to demonstrate course of trochlear nerve (4). Cerebellum is retracted to show structures in posterior fossa. Cross-section of midbrain is at level of superior colliculi and red nuclei. 2, Optic nerves and chiasma. 3, Oculomotor nerves (note relationship to posterior cerebral and posterior communicating arteries). 4, Trochlear nerve in edge of tentorium. 5, Trigeminal nerve (51, ophthalmic division; 52, maxillary division; 53, mandibular division). 6, Abducens nerve (note course up clivus and passage under petroclinoid ligament into posterior aspect of cavernous sinus). 7, Facial nerve. 8, Acoustic nerve. (From Glaser, J. S.: Neuro-ophthalmology. New York, Harper & Row, p. 251, 1978, with permission.)

and the ophthalmic division of cranial nerve V (trigeminal) immediately beneath it. Cranial nerve VI passes horizontally among the interstices of the cavernous sinus, first winding around the lateral aspect of the ascending portion of the internal carotid artery (ICA) and then lying beneath and lateral to the horizontal portion of the artery. Within the anterior portion of the cavernous sinus, Asbury et al.4 demonstrated enlargement and septation of cranial nerve III by connective tissue bands. Sympathetic fibers from the carotid plexus join the nerves to the ocular muscles within the cavernous

sinus but probably do not become an integral part of the nerves. All of the nerves to the ocular muscles pass into the orbit through the superior orbital fissure: cranial nerves III and VI are within the annulus of Zinn and cranial nerve IV is outside the annulus of Zinn. The abducens nerve continues a short distance under the lateral rectus muscle before it terminates among the muscle fibers. Cranial nerve IV travels forward and medially under the orbital roof, crosses the superior rectus muscle, and then enters the dorsal border of the superior oblique muscle. The superior oblique

Diseases of the Third, Fourth, and Sixth Cranial Nerves

muscle originates at the annulus of Zinn, extending superomedially within the orbit through the trochlea before attaching to the globe posterior to the equator. The third cranial nerve separates into two divisions either within the anterior portion of the cavernous sinus or as it enters the orbit. The superior division passes medially over the optic nerve, divides, and enters the ventral surface of the superior rectus muscle and the levator muscle of the upper eyelid. The inferior division immediately divides into branches to the medial rectus, interior rectus, and inferior oblique muscles. The parasympathetic fibers travel a short distance into the orbit within the inferior branch to the inferior oblique muscle; then they exit by one or two short branches to synapse within the ciliary ganglion. The postganglionic parasympathetic fibers travel from the ciliary ganglion to the globe as the short ciliary nerves, joined in this course by sympathetic fibers. Warwick120 has shown that iridectomy results in chromatolysis of only 3% of the cells in the ciliary ganglion, whereas lesions affecting the short ciliary nerves result in chromatolysis of approximately 97% of these cells. These findings confirm the facts that almost all of the cells of the ciliary ganglion innervate intrinsic ocular musculature (for accommodation) and that only a small fraction of their axons supply the iris sphincter.

BLOOD SUPPLY The blood supply of cranial nerves III, IV, and VI in their peripheral course is derived from small adjacent vessels in the subarachnoid space. There is a rich anastomosis of ascending and descending branches of these nutrient arteries within the pial septa. Capillaries penetrate the fasciculi and form a complex longitudinal ascending and descending network within the nerves. Asbury and co-workers4 have traced in detail the entire nutrient circulation of cranial nerve III. They found the vascular supply to be from three sources: (1) small arteries from the posterior cerebral and the basilar arteries; (2) three or four branches from the artery to the inferior cavernous sinus, arising from the base of the carotid siphon; and (3) recurrent branches from the ophthalmic artery to the orbital portion of the nerve. Tekemir et al.110 found that the posterior cerebellar artery supplies the proximal segment of III, whereas branches from the lateral trunk of the intracavernous ICA supply the segment within the cavernous sinus. Tekemir et al.110 found that the sixth nerve segment within Dorello’s canal is supplied by the meningeal branch of the ICA, whereas between the petrous apex and the superior orbital fissure it is supplied by branches of the lateral trunk of the intracavernous ICA. The tentorial cerebellar artery supplies the fourth nerve after it pierces the cerebellar tentorium.

1195

HISTOLOGIC FEATURES Sunderland and Hughes109 studied myelinated fibers within the nerves to the ocular muscles in humans. Myelinated fibers range in size from 3 to 18 ␮m. They classified the fibers into three groups—small, medium, and large—and stated that fibers of all diameters could be traced into the muscles from each of the nerves. Fibers passing to the ciliary ganglion were found to be of the small variety (3 to 5 ␮m). Within the intracranial portion of cranial nerve III, these small fibers, presumed to be pupillary fibers, are concentrated in and around the superior surface of the nerve. Kerr and Hollowell68 also found that the pupillomotor fibers are located superficially along the dorsomedial and medial aspect of cranial nerve III but have a gradually descending position within the nerve as it passes anteriorly. The accommodative fibers appear to be intimately associated with the pupillomotor fibers. Figure 48–3 illustrates the number of myelinated fibers and the spectrum of fiber diameters within normal cranial nerve VI. The measurements were made on photographic enlargements of 1.5-␮m transverse sections of each nerve at its intracranial-dural junction. This method for evaluating peripheral nerves has been described by Dyck and co-workers.36 The oculomotor nerve that was examined was 2.33 mm2 in transverse area and contained 18,542 fibers; the trochlear nerve measured was 0.19 mm2 and contained 2381 fibers; and the abducens nerve measured was 0.56 mm2 and contained 4016 fibers. Among the myelinated fibers, large fibers were predominant. In cranial nerves III and VI, the mean diameter of these large fibers was approximately 14 ␮m; in cranial nerve IV, the mean diameter of the large fibers was slightly less. There is a strong suggestion that the oculomotor nerve contains a population of intermediate-sized fibers not evident in cranial nerves IV or VI. Whether these fibers represent the parasympathetic fibers of cranial nerve III remains to be proved. In the sections examined, essentially no unmyelinated fibers were detected.

CLINICAL FEATURES A complete third nerve palsy would present with complete ptosis, the eye deviated down and out, and a large, unresponsive pupil. If the pupil is not involved, this is called an external oculomotor palsy. Many cases are incomplete third nerve palsies, with each innervated extraocular muscle being affected partially and equally. Some cases may affect only the superior or inferior division of cranial nerve III. In nuclear third nerve involvement, ptosis would be bilateral, whereas the superior rectus involvement would be on the contralateral side. The primary action of the superior oblique is to intort the eye; secondary and tertiary actions are depression and

1196

Diseases of the Peripheral Nervous System

FIGURE 48–3 A, Normal cranial nerve VI (⫻500). B, Normal cranial nerve III (⫻500). C, Myelinated fiber spectra of cranial nerves III and VI (top) and myelinated fiber spectrum of cranial nerve IV (bottom). Counts were made on phase-contrast enlargements of intracranial portion of normal nerves at site of contact with dura. (Courtesy of Peter J. Dyck, MD.)

abduction. Patients with trochlear nerve palsy report binocular vertical diplopia that is worse in adduction and downgaze. Hypertropia (elevated eye) occurs on the side of the palsied nerve. The hypertropia will be worse (1) in lateral gaze to the side opposite the palsy, (2) with head tilt to the same side, and (3) in downgaze. Often the patient will unconsciously tilt the head away from the eye with the nerve palsy. The inaction of the superior oblique

causes the eye to excyclotort; therefore, the two images will not be parallel. When the patient looks at a horizontal line, the images will form a tilted V (⬍or⬎). The pointed portion of the V will point in the direction of the affected eye. In bilateral fourth nerve palsy, there is often only a small vertical deviation in primary gaze. The hypertropia “reverses” in lateral gaze (right hypertropia in left gaze

Diseases of the Third, Fourth, and Sixth Cranial Nerves

and left hypertropia in right gaze). A vertical deviation is present with head tilt (ear to shoulder) and reverses direction with head tilt in the opposite direction (i.e., bilateral positive head tilt). The patient may assume a chin-down position. Excyclotorsion of greater than 10 degrees can occur. A V-pattern phoria, exophoria in upgaze, and esophoria in downgaze often coexist. The function of the trochlear nerve in the presence of third nerve palsy can be determined by asking the patient to look down. If cranial nerve IV is still functioning, the top of the globe will rotate in toward the nose. The examiner can watch a blood vessel on the conjunctiva to see the incyclotorsional movement. Nuclear cranial nerve IV lesions are very rare. A lesion of the trochlear nucleus would result in a contralateral superior oblique palsy. Associated midbrain findings might include ipsilateral internuclear ophthalmoplegia113 from median longitudinal fascicular damage, ipsilateral Horner’s syndrome from sympathetic damage, or a contralateral afferent pupil defect from damaged pretectal fibers in the superior colliculus.37 Etiologic mechanisms for a nuclear lesion could include ischemia, inflammation, trauma, tumor, or idiopathic causes. Abducens nerve palsy will cause the affected eye to deviate inward (secondary to the unopposed medial rectus), and the deviation will be more apparent in gaze to the side of the palsy.

CONGENITAL DEFECTS Congenital abnormality of ocular movements may be due to defects of the nuclei of cranial nerves or of the peripheral nerve. The literature contains several reports of such defects,56 which include unilateral or bilateral absence of the sixth cranial nerve, in isolation or associated with abnormalities of other cranial nerves, as well as hypoplasia or aplasia of the brainstem nuclei. Of interest also are a few reports of aberrant innervation of extraocular muscles, such as innervation of the lateral rectus muscle by a branch of cranial nerve III in the absence of cranial nerve VI. Congenital trochlear nerve palsy is often excluded from large studies; however, it is the most common cause for trochlear nerve palsy in children.53 Familial congenital fourth nerve palsy has been reported.5,45 Decompensation of congenital trochlear nerve palsy can present with symptoms in late childhood or adulthood. In fact, a decompensated congenital etiology should be suspected in all adults who present with an apparently new-onset fourth nerve palsy. Several characteristics distinguish congenital trochlear nerve palsy: long-standing head tilt (evident in old photos), large vertical fusional amplitudes (⬎4 to 5 prism diopters), underaction of the ipsilateral superior oblique, and possible associated horizontal strabismus.

1197

Some congenital superior oblique palsy patients have been found on magnetic resonance imaging (MRI) to have an absent superior oblique muscle and /or tendon.25 Congenital absence of the superior oblique tendon may be associated with craniofacial dysotoses, Down syndrome, and anencephaly.116 Laxity of the superior oblique tendon on traction testing has been associated with attenuated superior oblique muscles on MRI.100 At the time of surgical correction for congenital trochlear nerve palsy, a structural abnormality in the superior oblique tendon (tendon laxity or anomalous insertion) has been found in 87% of cases.47 In 1905, Duane34 described patients with congenital deficiency of abduction. These patients have limited or absent abduction with variable limitation of adduction. Also, on adduction, the globe retracts, the lid fissure narrows, and occasionally the globe moves upward. Duane’s syndrome occurs more frequently in females than in males and may be unilateral or bilateral; the left eye is affected more often than the right, and some cases are familial. Patients are usually asymptomatic and do not have diplopia, although they may be aware of their restricted abduction. There are three subtypes of Duane’s syndrome.57 Hotchkiss et al.54 reported a bilateral Duane’s syndrome with absent abducens nerve nuclei but innervation of the lateral rectus by a branch of the oculomotor nerve. Möbius87 described a 50-year-old man who presented with hand paralysis secondary to lead poisoning who also happened to have congenital bilateral sixth and seventh nerve palsies. The patient could converge and had learned to converge to look laterally with one eye at a time. These congenital cranial nerve palsies are stationary from birth. Carr et al.21 reported on 29 cases of congenital facial nerve palsy and 186 cases from the literature. Seventeen cases had isolated facial nerve palsy and six had bilateral abducens nerve palsy, three of them with hypoglossal palsy, one with bilateral trochlear nerve palsy, and one with bilateral oculomotor palsy. Carr et al.’s literature review revealed facial nerve palsy associated most frequently with abducens palsy (68%), followed by glossopharyngeal (28%), hypoglossal (26%), and oculomotor (20%) palsy. Other abnormalities associated with Möbius’ syndrome can include multiple cranial nerve palsies (cranial nerves III, IV, XII, V, IX, and X); limb malformations (syndactyly, polydactyly, brachydactyly, absent digits, talipes); malformation of the orofacial structures (bifid uvula, micrognathia, ear deformities); epicanthal folds; anomalies of the musculoskeletal system (Klippel-Feil anomaly, absent sternal head of the pectoralis major, rib defects, brachial muscle defects); cardiac anomalies; and occasionally mental retardation. Facial diplegia with ophthalmoplegia, axonal peripheral neuropathy, and hypogonadism has been reported in six cases, four males and two females.9 All of these cases had more extensive eye movement abnormalities than were originally described by Möbius. The peripheral

1198

Diseases of the Peripheral Nervous System

neuropathy onset is late in childhood and is slowly progressive in half of the cases. Muscle biopsy has shown neurogenic atrophy. Low testosterone and a normal response to gonadotropin-releasing hormone characterized the hypogonadotropic hypogonadism in the males. The etiology of this constellation of findings is unknown. There does not appear to be any genetic cause. Several papers have proposed a posterior circulation vascular cause, with subsequent necrosis and calcification in the brainstem, in Möbius’ syndrome.30,73,91 Henderson48 noted only three reports of pathologic studies in Möbius’ syndrome. These cases strongly support the view that the basis for the cranial nerve palsies is nuclear hypoplasia. The nerve roots of the affected cranial nerves were either absent or sparsely developed, depending on the degree of ganglion cell loss. The causes for this nuclear loss are obscure; only occasionally can a hereditary factor be traced in Möbius’ syndrome or in other types of congenital palsies except congenital ptosis. Congenital horizontal gaze palsy may be seen as an isolated cranial nerve defect. It differs from the ocular palsy of Möbius’ syndrome in that usually no convergent strabismus and no facial paralysis are present. Zweifach and colleagues128 described five children with complete absence of abduction on attempted conjugate gaze to either side; four of these five children substituted convergence and cross-fixation for lateral eye movements. This substitution phenomenon has been reported in both congenital and acquired conjugate gaze paralysis.18 Because the sparing of vertical gaze associated with severe impairment of horizontal gaze is similar to that seen with acquired midline pontine lesions, a congenital defect in this region is suggested.

EFFECTS OF TRAUMA Trauma to the head frequently results in injury to cranial nerves III, IV, and VI. Among 4278 cases of acquired ocular paralysis reviewed by Richards et al.,92 686 (16%) were secondary to head trauma. The oculomotor nerve is vulnerable to traumatic injury at several sites. The study by Richards et al.92 found a 15% incidence of oculomotor nerve palsy with head trauma. The nerve is easily compressed between the rigid posterior cerebral and superior cerebellar arteries and at the free edge of the reflected leaf of the tentorium. Trauma that results in shearing of the dural attachments of the third cranial nerve as it enters the cavernous sinus, as well as damage to the nerve in the cavernous sinus itself, is possible. Head injury associated with oculomotor nerve injury is usually also associated with loss of consciousness and skull fracture.7 Hegde et al.46 reported a case after head trauma of internal carotid dissection in the petrous portion with resultant oculomotor nerve palsy.

Minor head trauma is the most common (30%92) cause determined for acquired isolated trochlear nerve palsy.17,20,55,66,69,71,127 The trochlear nerve is vulnerable to trauma because of its long course and its delicate structure, despite the protection of the overlying tentorium. Both trochlear nerves can be injured by traumatic compression by the tentorium. Recent use of the fluid attenuation inversion recovery (FLAIR) technique in MRI found a traumatic fourth nerve palsy caused by a small subarachnoid hemorrhage near the trochlear nerve.44 Another case of traumatic trochlear nerve palsy was caused by compression from an unsuspected tentorial vascular malformation.61 Transient trochlear nerve palsy can occur after temporal lobectomy, which may be because of manipulation above the tentorium.62 Other neurosurgical procedures (petroclival or cavernous sinus tumor surgery, aneurysm repair) can also produce trochlear nerve palsy.65 Trauma to the trochlea has been reported after orbital penetrating injury, iatrogenic surgical trauma (sinus surgery), and blunt nonpenetrating trauma resulting in superior oblique weakness.72 Other orbital trauma can cause damage locally to the superior oblique tendon, nerve or muscle.6,97 After cataract surgery or retrobulbar anesthesia, paresis of a vertical muscle (usually inferior rectus or superior rectus) can occur, which may mimic superior oblique palsy.38 The frequent paralysis of the abducens nerve (15%) in trauma92 is related to several anatomic factors. The long course along the basiocciput of the skull beneath the lateral branches of the basilar artery, the firm dural attachment over the clivus, and arching over the ridge of the petrous bone at approximately a 90-degree angle beneath the rigid petrosphenoid ligament all predispose the abducens nerve to contusion and tearing injury. Fractures through the temporal bone or through the posterior clinoids may injure the nerve directly or may be associated with hemorrhage sufficient to cause either unilateral or bilateral paralysis. Schneider and Johnson102 have reported bilateral abducens palsies after severe hyperextension injury to the head and fractures of the cervical spinal column. The mechanism postulated by them was an upward and posterior displacement of the brain causing avulsion of the abducens nerves under the rigid petrosphenoid ligaments. Holmes et al.,52 in a prospective review of 99 patients with acute traumatic sixth nerve palsy, found a recovery rate of 71% at 6 months after onset. Complete sixth nerve palsy (inability to abduct past the midline) or bilateral palsy has a poor prognosis and may require strabismus surgery.51

Aberrant Regeneration of Cranial Nerve III As early as 1935, Bielschowsky13 described anomalous synkinetic ocular and lid movements after injury to the oculomotor nerve. The clinical findings in aberrant regeneration of the oculomotor nerve may consist of ptosis in primary gaze with lid retraction in downgaze

Diseases of the Third, Fourth, and Sixth Cranial Nerves

(pseudo-Graefe’s sign), elevation of the lid on adduction, limited vertical gaze with globe retraction on attempts at vertical gaze, adduction on attempted vertical gaze, and pupil constriction on adduction or depression. Aberrant regeneration of the oculomotor nerve occurs in approximately 50% of cases of congenital oculomotor nerve palsy.8 Schatz et al.101 first described aberrant regeneration of the oculomotor nerve without preceding nerve palsy or history of trauma. They termed this entity primary aberrant regeneration. The entity causing the primary aberrant regeneration in their patient was an intracavernous meningioma. Since that publication, other intracavernous15 as well as extracavernous etiologies have been identified for primary aberrant regeneration, including intracranial internal carotid aneurysm,29 posterior communicating artery aneurysm,22,114 abetalipoproteinemia,26 trigeminal neuroma, extracavernous meningioma, neurilemmoma, and asymmetrical mamillary bodies.63 Rare cases of aberrant regeneration have been reported after midbrain stroke84 or from a microvascular etiology.10 Several theories have been advanced to explain aberrant regeneration. The most commonly accepted theory proposes that, after damage to the axons of the nerve, regeneration may occur with new axons developing from the severed ends of the uninjured axons. Sibony et al.104 used retrograde horseradish peroxidase transport through the superior rectus muscle and found that neurons innervating the superior rectus, after oculomotor regeneration, originated from the subnuclei of the inferior rectus, medial rectus, inferior oblique, and superior rectus. Two other theories have been suggested to explain aberrant regeneration, especially of the primary type. Ephaptic transmission is thought to occur when action potentials from demyelinated portions of the injured peripheral nerve are transmitted to intact adjacent axons. This might explain reports of transient or evolving signs of synkinesis. Some reports of temporary oculomotor synkinesis, however, have theorized that the incorrectly regenerated fibers are subsequently replaced with correctly regenerating fibers.105 Nuclear synaptic reorganization might also account for aberrant regeneration.40,75,105

NEOPLASMS In a series by Richards et al.92 of 4176 cases of acquired ocular palsies, 767 were due to neoplasm (18%). The abducens nerve was involved in 413 cases, the oculomotor nerve in 141, and the trochlear nerve in 28. The types of neoplasm included meningioma, metastases, pituitary tumors, glioma (pontine and midbrain), chordoma, nasopharyngeal tumors, other primary tumors, acoustic neuroma, multiple myeloma, squamous cell carcinoma, adenoid cystic carcinoma, and others.

1199

Neoplasms involving the midbrain affect the nuclei and the intramedullary portion of the cranial nerves directly. Examples of such tumors are pontine gliomas, medulloblastomas, hemangiomas, and metastatic tumors. Specific syndromes of the midbrain are discussed below under Vascular Diseases and Aneurysms. In 1883, Parinaud89 described lesions of the midbrain that produced paralysis of conjugate upward gaze and convergence. Later, Koerber et al.70 completed the clinical picture by describing retraction nystagmus, large pupils poorly reactive to light, and papilledema associated with a tumor in the region of the pineal gland and posterior third ventricle. Bilateral superior oblique palsy, although usually secondary to trauma, has also been described in dorsal midbrain neoplasms.11 Brainstem glioma was the most common neoplasm found causing an acquired ocular cranial nerve palsy in children in the study by Kodsi and Younge.69 Most often the glioma presented with abducens nerve palsy (58.3%). In a population-based study, Holmes et al.53 found a low incidence (8%) of tumors in pediatric ocular cranial nerve palsies. None of their patients with a neoplasm presented with the cranial nerve palsy in isolation. Tumors in the region of the cerebellar pontine angle may affect the ocular nerves. Acoustic neurinomas comprise 8% of all primary intracranial tumors and account for 40% of all tumors of the posterior fossa. Van Meter and colleagues112 studied 100 patients operated on for pathologically confirmed acoustic neurinoma and found a low incidence (3%) of ocular nerve palsy (cranial nerve VI). Neoplasms in the region of the clivus affect the abducens nerves early in their course. An example of such a neoplasm is the chordoma, a tumor derived from remnants of the notochord. More frequently found in the sacrococcygeal region, this tumor occasionally manifests as an intracranial neoplasm. Intracranial chordomas account for approximately 1% of all intracranial tumors and occur predominantly in young men. Cranial nerve VI is most often affected and is often the site of initial manifestation of the tumor. Isolated oculomotor nerve palsy, isolated trochlear nerve palsy, and multiple cranial nerve palsies have also been reported with clivus chordomas and chondrosarcomas.115 Tumors of the pituitary gland can expand laterally into the cavernous sinus and compress the cavernous sinus sufficiently to cause paresis of the cranial nerves within it. Cranial nerve III, which lies within the superolateral wall of the sinus, is particularly likely to be compressed against the rigid interclinoid ligament. It is the most frequently affected cranial nerve, although cranial nerves IV and VI as well as the ophthalmic branch of cranial nerve V may also be involved. In a study by Hollenhorst and Younge50 of 1000 cases of pituitary tumor, 46 cases presented with ocular cranial nerve palsy: 24 with cranial nerve III palsy, 6 with cranial nerve VI palsy, 12 with multiple palsies of

1200

Diseases of the Peripheral Nervous System

cranial nerves III and IV or III and VI, and 2 with bilateral cranial nerve III palsy. A sudden hemorrhage within a pituitary tumor may result in precipitous loss of vision and ophthalmoplegia unilaterally or bilaterally (usually cranial nerve III, then VI).19 These symptoms are usually accompanied by severe headache in the classic “pituitary apoplexy.” In such cases, the prognosis for recovery of ocular muscle function is good after immediate surgical evacuation of the hemorrhagic tumor. Cavernous sinus tumors usually present with multiple cranial nerve palsies. The types of tumor involved are many: pituitary tumor, craniopharyngioma, meningioma, metastatic lesions, nasopharyngeal tumors, and lymphoma, as well as aneurysms and infectious/inflammatory processes. Differentiating cavernous sinus involvement from superior orbital fissure involvement is difficult, because clinically they are indistinguishable. The nasopharynx is a common site for neoplasms that extend intracranially. Nasopharyngeal tumors may extend into the cavernous sinus and affect the cranial nerve to the extraocular muscles. The patients usually present with nasal fullness/stuffiness, and nonspecific findings related to the sinuses are found on scanning. Intrinsic tumors of the ocular motor nerves, such as neurinoma,1,24 schwannoma,39a,59a,76 cavernous angioma,81,82

A

are rare. A recent case of presumed neuroma of the fourth nerve is illustrated in Figure 48–4. Perineural spread of malignant tumors is a complication from cutaneous squamous cell carcinoma and rarely from basal cell carcinoma and melanoma. McNab and colleagues83 and Wilcsek and colleagues124 described perineural spread of squamous cell carcinoma through the orbit with 67% of the patients presenting with ophthalmoplegia, usually partial oculomotor nerve paresis at the onset. Neurologic syndromes associated with carcinoma but not resulting from metastatic lesions or direct involvement of the central nervous system can assume several forms and involve any level of the nervous system. The LambertEaton syndrome (LES) may occur and result in weakness and fatigability of proximal limb muscles associated with loss of tendon reflexes. Paralysis of the extraocular muscles may be seen as a part of this disorder. LES is associated with malignancy in approximately 70% of cases, most commonly with small cell lung carcinoma (SCLC). A presynaptic calcium channel antibody has been identified in many of these cases and is a marker for paraneoplastic neurologic immunity.74 Paraneoplastic encephalomyelitis can primarily affect the brainstem, leading to ophthalmoplegia. Most of these cases are also associated with SCLC. The antibody found in these cases is antineuronal nuclear autoantibody, type 1 or anti-Hu.79

B FIGURE 48–4 Fourth nerve neuroma in a patient with long-standing fourth nerve palsy. No biopsy was obtained, and there were no other symptoms. A, Magnetic resonance imaging (MRI) with gadolinium enhancement showing lesion along the course of the fourth nerve. B, Coronal section MRI through the midbrain showing enhancing lesion along the course of the fourth nerve (arrows).

Diseases of the Third, Fourth, and Sixth Cranial Nerves

Pseudotumor cerebri (idiopathic intracranial hypertension) is an entity with elevated intracranial pressure without any intracranial mass lesion. Headache is the predominant symptom, with transient visual obscurations, disc edema, and diplopia occurring frequently. There are many cases of nonlocalizing abducens palsy associated with pseudotumor cerebri. Trochlear nerve palsy has also been reported.90,108

VASCULAR DISEASES AND ANEURYSMS Vascular disease of the central or peripheral nervous system is a frequent cause of acquired ocular palsy. Small arteries arising from the posterior cerebral or basilar artery perfuse the brainstem nuclei of cranial nerves III, IV, and VI. These arteries provide a rich vascular network to the brainstem, but this area lacks the great availability of collateral circulation that is predominant in the anterior circulation. Therefore, a vascular insult, ischemic or hemorrhagic, in this area often results in major clinical symptoms of the oculomotor system. Atherosclerosis is an important factor in the cause of ophthalmoplegia. Weber122 stated that 50% of third cranial nerve palsies occur in patients with hypertension and atherosclerosis. Vascular lesions may also be responsible for acute ocular palsy that is related to infectious or toxic disease. Several clinical syndromes can pinpoint the exact anatomic location of the lesion in the midbrain. Paramedian infarcts of the middle midbrain lead to nuclear third nerve palsy of a severe or complete nature.16 Weber’s syndrome (ipsilateral oculomotor palsy with contralateral hemiparesis) is the most common midbrain syndrome involving cranial nerve III, localizing laterally to the fascicle of the oculomotor nerve within the cerebral peduncle. Benedikt’s syndrome (ipsilateral oculomotor palsy with contralateral tremor) localizes to the fascicle of the oculomotor nerve within the red nucleus and substantia nigra.78 Claude’s syndrome includes ipsilateral oculomotor palsy with contralateral ataxia, dysmetria, and dysdiadochokinesia from a lesion of the oculomotor nerve fascicle, red nucleus, and brachium conjunctivum.78 Nothnagel described tumors of the quadrigeminal bodies causing unilateral or bilateral oculomotor palsy and ipsilateral ataxia from involvement of the third nerve fascicles, the brachium conjunctivum, and the superior and inferior colliculi.78 Ocular palsies resulting from vertebrobasilar disease are associated with other signs and symptoms of brainstem and cerebellar involvement; rarely a partial or complete ophthalmoplegia may be the only evidence of vertebrobasilar disease. Minor and associates85 reviewed 183 cases of vertebrobasilar disease and noted varying degrees of paralysis of cranial nerves III, IV, and VI in 27% of patients. Ferbert et al.39 described 85 cases of angiographically proven

1201

basilar artery occlusion. Excluding the 26 patients who were comatose, they found 30 with cranial nerve IV, VI, and VII involvement, 22 with horizontal gaze paresis, 13 with oculomotor palsies, and 4 with vertical gaze palsy. Diplopia was the most common visual symptom in 51 episodes of vertebral artery dissection described by Hicks et al.49 Abducens nerve palsy was the most common abnormality, followed by skew deviation, cranial nerve III palsy, and cranial nerve IV palsy. The most common cause of sudden paralysis of cranial nerve III in an adult with headache and a dilated fixed pupil is an aneurysm at the junction of the posterior communicating artery and the internal carotid artery. This is especially true in women older than 40 years of age. The clinical picture of oculomotor palsy from an aneurysm is similar to that of one associated with diabetes except that the pupil is only rarely normal when an aneurysm is the underlying cause. Rucker94 observed only 4 patients with a normal pupil in 114 cases of cranial nerve III palsy resulting from aneurysms. An unruptured intracranial aneurysm may present early in the course with an incomplete oculomotor paresis.12 Aneurysms of the top of the basilar artery and the junction between the basilar artery and the superior cerebellar artery can also cause oculomotor nerve palsy.14,80 Occasionally oculomotor nerve palsy associated with minor head trauma can be the first sign of the intracranial aneurysm.117 Trochlear nerve palsy caused by an intracerebral aneurysm is rare,92 but there have been cases reported with superior cerebellar artery,2,27 posterior communicating artery, posterior cerebral artery, basilar artery, and internal carotid artery3 aneurysms. Most of these cases had other associated findings: headache, nausea, or longstanding trochlear nerve palsy without evidence for a congenital palsy. Carotid-cavernous sinus fistula presents with engorged arterialized blood vessels on the conjunctiva, retinal venous engorgement, ocular bruit, and ophthalmoplegia (frequently abducens paresis).98,103 Posterior drainage of a carotid-cavernous fistula or a dural-sinus fistula may cause oculomotor damage.86 The association of ocular palsies and diabetes mellitus is well recognized.95 Rucker93,94 found cranial nerves III and VI to be about equally affected and cranial nerve IV less frequently affected. Sudden onset of ptosis and paralysis of the extraocular muscles supplied by cranial nerve III characterize the acute oculomotor palsy associated with diabetes. Initially there may be severe pain within or behind the orbit in the distribution of the ophthalmic division of cranial nerve V. The pupillomotor fibers within cranial nerve III are frequently spared, although partial pupil involvement can occur in up to one third of cases.58,59 Usually, recovery begins in a few weeks and is complete within 6 to 8 weeks. In a study by Keane and Ahmadi,67

1202

Diseases of the Peripheral Nervous System

among 49 patients with diabetes with isolated third nerve palsies, only 2% had brainstem localization. In oculomotor nerves of diabetics who had not had a previous palsy, Smith and Dyck107 found microfasciculations and a change in size distribution of myelinated fibers. Demyelination with some loss of axons within the central portion of the peripheral part of cranial nerve III has been the essential finding in three reports describing the clinicopathologic features of oculomotor palsy in diabetes mellitus. Dreyfus and co-workers33 noted a focal zone of destruction of myelin sheaths and axons within the intracavernous portion of the nerve. They were unable to find any occluded vessels in the many sections they studied, but concluded that the lesion was ischemic in origin and postulated that a nutrient vessel remote from the site of fiber degeneration was occluded. The exact pathogenesis of the pain often associated with this disorder has yet to be documented; ischemia within a sensory nerve in the area of the cavernous sinus or dura may be a contributing factor. Asbury and associates found a circumscribed patch of demyelination within the intracavernous portion of the third cranial nerve.4 Its borders were sharply demarcated on the proximal edge and less distinct on the distal edge. A paucity of wallerian degeneration distally and an absence of central chromatolysis in the cell bodies of the oculomotor nerve nucleus were notable histologic features. Both findings are consistent with a demyelinating process and compatible with the rapid and complete recovery that ensues in most oculomotor palsies with diabetes. Weber and colleagues123 reported a similar demyelination with some loss of central axons within the subarachnoid portion of cranial nerve III. Hyalinization of the intraneural arterioles was evident, but again no occluded vessels were found. All three studies of the pathologic changes in cranial nerve III noted a relative sparing of the peripheral fibers.4,33,123 Dreyfus and co-workers33 hypothesized that the superficial position of the fine “parasympathetic” fibers protected them from the effects of ischemia. The same argument is used for their susceptibility to injury from external pressure. Also relevant to the ischemic theory for diabetic neuropathy is the study by Dyck and collaborators35 of the peripheral nerves of patients with necrotizing angiopathy and neuropathy. Similar central fascicular fiber degeneration was noted, with only an occasional occluded vessel in the vicinity of the fiber degeneration. They postulated that areas within the peripheral nerves examined (the mid-upper arm and midthigh levels) represented watershed zones of poor perfusion. Under these conditions, the ischemic effects would be greatest within the most remote portion of a region supplied by a specific epineurial vessel or vessels, and the typical picture of an infarct with softening and necrosis would not be seen even though the underlying mechanism may be occlusive vascular disease.

INFECTIONS AND TOXINS The literature contains many case reports of various infections, toxins, and inflammatory conditions that can produce oculomotor palsies, and the list includes mucormycosis, sarcoidosis, cysticercosis, botulism, tetany, and diphtheria. Paranasal sinus infections with spread to the orbit or cavernous sinus may produce oculomotor palsies, and though uncommon, are often associated with an immunocompromised state, diabetes, or even chronic drug abuse, as in cocaine sniffing.42 Though less common since the advent of effective antibiotics, Gradenigo’s syndrome of a petrous abscess from chronic mastoid infection with spread to the posterior fossa can cause a sixth nerve palsy. The occurrence of isolated trochlear nerve palsy can occur after cutaneous ophthalmic herpes zoster infection43 or herpes zoster oticus.64 Meningitic processes, including bacterial, tuberculous, herpes zoster, syphilis, lymphoma, cryptococcosis,99 neuroborreliosis,88 Ehrlichia,23 and carcinomatous meningitis, have all been associated with trochlear nerve palsies. Meningismus, headache, other cranial nerve palsies, and abnormal findings on examination of the cerebrospinal fluid should help determine these entities. Drug therapy, particularly for neoplasms, can induce ophthalmoparesis, and the list of such drugs includes vincristine, phenytoin, pancuronium,106 nemaline,126 and doxepin overdosage.77 Botulinum poisoning can produce ophthalmoplegia, particularly internal, with loss of accommodation and mydriasis. Years ago diphtheria was noted to produce a selective internal ophthalmoplegia. Ocular palsies are more frequent with bacterial infections, particularly when the meninges are involved. Impaired ocular movement was reported to be the most frequent sign of cranial nerve dysfunction in 147 cases of acute bacterial meningitis.31,32 Among that group were 36 cases of ocular palsy, equally divided between cranial nerves III and VI. The ocular palsies tended to be transient and to disappear shortly after recovery from the meningitis. This result contrasts with the damage often sustained by cranial nerve VIII; permanent hearing loss is an important sequela of acute bacterial meningitis. Whether the conditions discussed next belong in this section on Infections and Toxins remains to be seen. The Tolosa-Hunt syndrome, in which a granulomatous inflammatory infiltration of the cavernous sinus region can produce a painful ophthalmoplegia, is really very rare. There are literally over 100 causes of painful ophthalmoplegia, and the Tolosa-Hunt syndrome needs to be a diagnosis of exclusion for that reason.111 The Miller Fisher syndrome, characterized by progressive ophthalmoplegia, ataxia, and areflexia, usually follows an upper respiratory syndrome, and has a benign course with complete recovery in 3 to 6 weeks.41 In some variants of Guillain-Barré syndrome, diplopia may ensue as a result

Diseases of the Third, Fourth, and Sixth Cranial Nerves

of an ocular palsy, but more often the facial nerve is involved. Multiple sclerosis patients frequently have diplopia, more often as a result of internuclear ophthalmoplegia, but approximately 5% have ocular palsies.60 Sarcoidosis affects the facial nerve much more commonly than other cranial nerves. In 118 cases reviewed by Colover,28 the facial nerve was involved in 59 cases; there were only 6 with ptosis, 3 with oculomotor paralysis, and 1 with trochlear involvement. Recovery is usually complete.

REFERENCES 1. Abe, T., Iwata, T., Shimazu, M., and Matsumoto, K.: Trochlear nerve neurinoma associated with a giant thrombosed dissecting aneurysm of the contralateral vertebral artery. Surg. Neurol. 42:438, 1994. 2. Agostinis, C., Caverni, L., Moschini, L., et al.: Paralysis of fourth cranial nerve due to superior cerebellar artery aneurysm. Neurology 42:457, 1992. 3. Arruga, J., DeRivas, P., Espinet, H. L., and Conesa, G.: Chronic isolated trochlear nerve palsy produced by intracavernous internal carotid artery aneurysm. J. Clin. Neuroophthalmol. 11:104, 1991. 4. Asbury, A. K., Aldredge, H., Hershberg, R., and Fisher, C. M.: Oculomotor palsy in diabetes mellitus. Brain 93:555, 1970. 5. Astle, W. F., and Rosenbaum, A. L.: Familial congenital fourth cranial nerve palsy. Arch. Ophthalmol. 103:532, 1985. 6. Bachynski, B. N., and Flynn, J. T.: Direct trauma to the superior oblique tendon following penetrating injuries of the upper eyelid. Arch. Ophthalmol. 103:1510, 1985. 7. Balcer, L. J., Galetta, S. L., Bagley, L. J., et al.: Localization of traumatic oculomotor nerve palsy to the midbrain exit site by MRI. Am. J. Ophthalmol. 22:437, 1996. 8. Balkan, R., and Hoyt, C. S.: Associated neurologic abnormalities in congenital third nerve palsies. Am. J. Ophthalmol. 97:315, 1984. 9. Baraitser, M., and Rudge, P.: Moebius syndrome, an axonal neuropathy and hypogonadism. Clin. Dysmorphol. 5:351, 1996. 10. Barr, D., Kupersmith, M., Turbin, R., et al.: Synkinesis following diabetic third nerve palsy. Arch. Ophthalmol. 118:132, 2000. 11. Barr, D. B., McFadzean, R. M., Hadley, D., et al.: Acquired bilateral superior oblique palsy: a localizing sign in the dorsal midbrain. Eur. J. Ophthalmol. 7:271, 1997. 12. Bartleson, J. D., Trautmann, J. C., and Sundt, T. M. Jr.: Minimal oculomotor nerve paresis secondary to unruptured intracranial aneurysm. Arch. Neurol. 43:1015, 1986. 13. Bielschowsky, A.: Lectures on motor anomalies of the eyes. II. Paralysis of individual eye muscles. Arch. Ophthalmol. 13:33, 1935. 14. Boccardo, M., Ruelle, A., and Banchero, M. A.: Isolated oculomotor palsy caused by aneurysm of the basilar artery bifurcation. J. Neurol. 233:61, 1986. 15. Boghen, D., Chartrand, J.-P., Laflamme, P., et al.: Primary aberrant third nerve regeneration. Ann. Neurol. 6:415, 1979.

1203

16. Bogousslavsky, J., Maeder, P., Regli, F., and Meuli, R.: Pure midbrain infarction: clinical syndromes, MRI, and etiologic patterns. Neurology 44:2032, 1994. 17. Burgerman, R. S., Wolf, A. L., Kelman, S. E., et al.: Traumatic trochlear nerve palsy diagnosed by magnetic resonance imaging: case report and review of the literature. Neurosurgery 25:978, 1989. 18. Burian, H. M., VanAllen, M. W., Sexton, R. R., and Baller, R. S.: Substitution phenomena in congenital and acquired supranuclear disorders of eye movement. Trans. Am. Acad. Ophthalmol. 69:1105, 1965. 19. Cardoso, E. R., and Peterson, E. W.: Pituitary apoplexy: a review. Neurosurgery 14:363, 1984. 20. Carlow, T. J.: Paresis of cranial nerves III, IV, and VI: clinical manifestations and differential diagnosis. Bull. Soc. Belge Ophtalmol. 237:285, 1989. 21. Carr, M., Ross, D., and Zuker, R.: Cranial nerve defects in congenital facial palsy. J. Otolaryngol. 26:80, 1997. 22. Carrasco, J. R., Savino, P. J., and Bilyk, J. R.: Primary aberrant oculomotor nerve regeneration from a posterior communicating artery aneurysm. Arch. Ophthalmol. 120:663, 2002. 23. Carter, N., and Miller, N. R.: Fourth nerve palsy caused by Ehrlichia chaffeensis. J. Neuroophthalmol. 17:47, 1997. 24. Celli, P., Ferrante, L., Acqui, M., et al.: Neurinoma of the third, fourth and sixth cranial nerves: a survey and report of a new fourth nerve case. Surg. Neurol. 38:216, 1992. 25. Chan, T. K., and Demer, J. L.: Clinical features of congenital absence of the superior oblique muscle as demonstrated by orbital imaging. J. AAPOS 3:143, 1999. 26. Cohen, D. A., Bosley, T. M., Savino, P. J., et al.: Primary aberrant regeneration of the oculomotor nerve: occurrence in a patient with abetalipoproteinemia. Arch. Neurol. 42:821, 1985. 27. Collins, T. E., Mehalic, T. F., White, T. K., and Pezzuti, R. T.: Trochlear nerve palsy as the sole initial sign of an aneurysm of the superior cerebellar artery. Neurosurgery 30:258, 1992. 28. Colover, J.: Sarcoidosis with involvement of nervous system. Brain 71:451, 1948. 29. Cox, T. A., Wurster, J. B., and Godfrey, W. A.: Primary aberrant oculomotor regeneration due to intracranial aneurysm. Arch. Neurol. 36:570, 1979. 30. D’Cruz, O. F., Swisher, C. N., Jaradeh, S., et al.: Möbius syndrome: evidence for a vascular etiology. J. Child Neurol. 8:260, 1993. 31. Dodge, P. R., and Swartz, M. N.: Bacterial meningitis: a review of selected aspects. I. General clinical features, special problems and unusual meningeal reactions mimicking bacterial meningitis. N. Engl. J. Med. 272:898, 1965. 32. Dodge, P. R., and Swartz, M. N.: Bacterial meningitis: a review of selected aspects. II. Special neurologic problems, postmeningitis complication and clinicopathological correlations. N. Engl. J. Med. 272:954, 1003, 1965. 33. Dreyfus, P. M., Hakim, S., and Adams, R. D.: Diabetic ophthalmoplegia: report of case, with postmortem study and comments on vascular supply of human oculomotor nerve. Arch. Neurol. Psychiatry 77:337, 1957. 34. Duane, A.: Congenital deficiency of abduction, associated with impairment of adduction, retraction movements, contraction of the palpebral fissure and oblique movements of the eye. Arch. Ophthalmol. 34:133, 1905.

1204

Diseases of the Peripheral Nervous System

35. Dyck, P. J., Conn, D. L., and Okazaki, H.: Necrotizing angiopathic neuropathy: three-dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972. 36. Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased-fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968. 37. Eliot, D., Cunningham, E. T. Jr., and Miller, N. R.: Fourth nerve paresis and ipsilateral relative afferent pupillary defect without visual sensory disturbance: a sign of contralateral dorsal midbrain disease. J. Clin. Neuroophthalmol. 11:169, 1991. 38. Esswein, M. B., and von Noorden, G. K.: Paresis of a vertical rectus muscle after cataract extraction. Am. J. Ophthalmol. 116:424, 1993 39. Ferbert, A., Bruckmann, H., and Drummen, R.: Clinical features of proven basilar artery occlusion. Stroke 21:1135, 1990. 39a. Feinberg, A. S., and Newman, N. J.: Schwannoma in patients with isolated trochlear nerve palsy. Am. J. Ophthalmol. 127:183, 1999. 40. Fernandez, E., Pallini, R., Lauretti, L., et al.: Motonuclear changes after cranial nerve injury and regeneration. Arch. Ital. Biol. 135:343, 1997. 41. Fisher, M.: An unusual variant of acute idiopathic polyneuritis (syndrome of ophthalmoplegia, ataxia, and areflexia). N. Engl. J. Med. 255:57, 1956. 42. Goldberg, R. A., Weisman, J. S., McFarland, J. E., et al.: Orbital inflammation and optic neuropathies associated with chronic sinusitis from intranasal cocaine abuse: possible role of contiguous inflammation. Arch. Ophthalmol. 107:831, 1989. 43. Grimson, B. G., and Glaser, J. S.: Isolated trochlear nerve palsies in herpes zoster ophthalmicus. Arch. Ophthalmol. 96:1233, 1978. 44. Hara, N., Kan, S., and Simiziu, K.: Localization of posttraumatic trochlear nerve palsy associated with hemorrhage at the subarachnoid space by magnetic resonance imaging. Am. J. Ophthalmol. 132:443, 2001. 45. Harris, D. J., Memmen, J. E., Katz, N. N. K., and Parks, M. M.: Familial congenital superior oblique palsy. Ophthalmology 93:88, 1986. 46. Hegde V., Coutinho, C. M. A., and Mitchell, J. D.: Dissection of the intracranial internal carotid artery producing isolated oculomotor nerve palsy with sparing of pupil. Acta Neurol. Scand. 105:330, 2002. 47. Helveston, E. M., Krach, D., Plager, D. A., and Ellis, F. D.: A new classification of superior oblique palsy based on congenital variations in the tendon. Ophthalmology 99:1609, 1992. 48. Henderson, J. L.: The congenital facial diplegia syndrome: clinical features, pathology and aetiology; a review of sixtyone cases. Brain 62:381, 1939. 49. Hicks, P. A., Leavitt, J. A., and Mokri, B.: Ophthalmic manifestations of vertebral artery dissection. Ophthalmology 101:1786–92, 1994. 50. Hollenhorst, R. W., and Younge, B. R.: Ocular manifestations produced by adenomas of the pituitary gland: analysis of 1000 cases. In Laws, E. R. Jr., Randall, R. V., Kern, E. B., and Abboud, C. F. (eds.): Management of Pituitary Adenomas and Related Lesions with Emphasis on

51.

52.

53.

54.

55.

56.

57. 58.

59.

59a.

60.

61.

62.

63.

64. 65. 66. 67. 68.

69.

Transsphenoidal Microsurgery. New York, AppletonCentury-Crofts, p. 111, 1982. Holmes, J. M., Beck, R. W., Kip, K. E., et al.: Predictors of nonrecovery in acute traumatic sixth nerve palsy and paresis. Ophthalmology 108:1457, 2001. Holmes, J. M., Droste, P. J., and Beck, R. W.: The natural history of acute traumatic sixth nerve palsy or paresis. J. AAPOS 2:265, 1998. Holmes, J. M., Mutyala, S., Maus, T., et al.: Pediatric third, fourth and sixth nerve palsies: a population-based study. Am. J. Ophthalmol. 127:388, 1999. Hotchkiss, M. G., Miller, N. R., Clark, A. W., and Green, W. R.: Bilateral Duane’s retraction syndrome. a clinical-pathologic case report. Arch. Ophthalmol. 98:870, 1980. Hoya, K., and Kirino, T.: Traumatic trochlear nerve palsy following minor occipital impact. Neurol. Med. Chir. 40:358, 2000. Hoyt, W. F., and Nachtigäller, H.: Anomalies of ocular motor nerves: neuroanatomic correlates of paradoxical innervation of Duane’s syndrome and related congenital ocular motor disorders. Am. J. Ophthalmol. 60:443, 1965. Huber, A.: Electrophysiology of the retraction syndromes. Br. J. Ophthalmol. 58:293,1974. Ing, E. B., Leavitt, J. A., and Younge, B. R.: Incidence of pupillary involvement in ischemic oculomotor nerve palsies. Ann. Ophthalmol. 32:90, 2000. Jacobson, D. M.: Pupil involvement in patents with diabetes-associated oculomotor nerve palsy. Arch. Ophthalmol. 116:723, 1998. Jackowski, A., Weiner, G., and O’Reilly, G.: Trochlear nerve schwannomas: a case report and literature review. Br. J. Neurosurg. 8:219, 1994. Jacobson, D. M., Moster, M. L., Eggenberger, E. R., et al.: Isolated trochlear nerve palsy in patients with multiple sclerosis. Neurology 53:877, 1999. Jacobson, D. M., Warner, J. J., Choucair, A. K., and Ptacek, L. J.: Trochlear nerve palsy following minor head trauma: a sign of structural disorder. J. Clin. Neuroophthalmol. 8:263, 1988. Jacobson, D. M., Warner, J. J., and Ruggles, K. H.: Transient trochlear nerve palsy following anterior temporal lobectomy for epilepsy. Neurology 45:1465, 1995. Johnson, L. N., Kamper, C. A., Hepler, R. S., et al.: Primary aberrant regeneration of the oculomotor nerve from presumed extracavernous neurilemoma, meningioma, and asymmetric mammillary body. Neuroophthalmology 9:227, 1989. Keane, J. R.: Delayed trochlear nerve palsy in a case of zoster oticus. Arch. Ophthalmol. 93:382, 1975. Keane, J. R.: Trochlear nerve pareses with brainstem lesions. J. Clin. Neuroophthalmol. 6:242, 1986. Keane, J. R.: Fourth nerve palsy: historical review and study of 215 patients. Neurology 43:2439, 1993. Keane, J. R., and Ahmadi, J.: Most diabetic third nerve palsies are peripheral. Neurology 51:1510, 1998. Kerr, F. W. L., and Hollowell, O. W.: Location of pupillomotor and accommodation fibres in the oculomotor nerve: experimental observations of paralytic mydriasis. J. Neurol. Neurosurg. Psychiatry 27:473, 1964. Kodsi, S. R., and Younge, B. R.: Acquired oculomotor, trochlear and abducent cranial nerve palsies in pediatric patients. Am. J. Ophthalmol. 114:568, 1992.

Diseases of the Third, Fourth, and Sixth Cranial Nerves 70. Koerber, H., Salus, R., and Elschnig (Cited in Wold, J. K.: The Classical Brain Stem Syndromes. Springfield, IL, Charles C Thomas, p. 157, 1971.) 71. Lavin, P. J. M., and Troost, B. T.: Traumatic fourth nerve palsy. Arch. Neurol. 41:679, 1984. 72. Legge, R. H., Hedges, T. R. III, Anderson, M., and Reese, P. D.: Hypertropia following trochlear trauma. J. Pediatr. Ophthalmol. Strabismus 29:163, 1992. 73. Lengyel, D., Zaunbauer, W., Keller, E., and Gottlob, I.: Möbius syndrome: MRI findings in three cases. J. Pediatr. Ophthalmol. Strabismus 37:305, 2000. 74. Lennon, V. A., Kryzer, T. J., Griesmann, G. E., et al.: Calcium-channel antibodies in the Lambert-Eaton syndrome and other paraneoplastic syndromes. N. Engl. J. Med. 332:1467, 1995. 75. Lepore, F. E., and Glaser, J. S.: Misdirection revisited: a critical appraisal of acquired oculomotor nerve synkinesis. Arch. Ophthalmol. 98:2206, 1980. 76. Leunda, G., Vaquero, J., Cabezudo, J., et al.: Schwannoma of the oculomotor nerves. J. Neurosurg. 57:563, 1982. 77. LeWitt, P. A.: Transient ophthalmoparesis with doxepin overdosage. Ann. Neurol. 9:618, 1981. 78. Liu, G. T., Crenner, C. W., Logigian, E. L., et al.: Midbrain syndromes of Benedikt, Claude and Nothnagel. Neurology 42:1820, 1992. 79. Lucchinetti, C. F., Kimmel, D. W., and Lennon, V. A.: Clinical, oncological and serological profiles of patients seropositive for type 1 anti-neuronal nuclear antibody (ANNA-1, a.k.a. “antiHu”). Neurology 44(Suppl. 2):A156, 1994. 80. Lustbader, J. M., and Miller, N. R.: Painless, pupil-sparing but otherwise complete oculomotor nerve paresis caused by basilar artery aneurysm. Arch. Ophthalmol. 106:583, 1988. 81. Matias-Guiu, X., Alejo, M., Sole, T., et al.: Cavernous angiomas of the cranial nerves. J. Neurosurg. 73:620, 1990. 82. Maruoka, N., Yamakawa, Y., and Shimauchi, M.: Cavernous hemangioma of the optic nerve. J. Neurosurg. 69:92, 1988. 83. McNab, A. A., Francis, I. C., Benger, R., and Crompton, J. L.: Perineural spread of cutaneous squamous cell carcinoma via the orbit. Ophthalmology 104:1457, 1997. 84. Messé, S. R., Shin, R. K., Liu, G. T., et al.: Oculomotor synkinesis following a midbrain stroke. Neurology 57:1106, 2001. 85. Minor, R. H., Kearns, T. P., Millikan, C. H., et al.: Ocular manifestations of occlusive disease of the vertebral-basilar arterial system. Arch. Ophthalmol. 62:84, 1959. 86. Miyachi, S., Negoro, M., Handa, T., et al.: Dural carotid cavernous sinus fistula presenting as isolated oculomotor nerve palsy. Surg. Neurol. 39:105, 1993. 87. Möbius, P. J.: Ueber angeborene doppelseitige AbducensFacialis-Lähmung. Munch. Med. Wochenschr. 35:91, 1888. 88. Müller, D., Neubauer, B. A., Waltz, S., and Stephani, U.: Neuroborreliosis and isolated trochlear palsy. Eur. J. Paediatr. Neurol. 2:275, 1998. 89. Parinaud, M. H.: Paralysie des mouvements associes des yeux. Arch. Neurol. (Paris) 5:145, 1883. 90. Patton, N., Beatty, S., and Lloyd, I. C.: Bilateral sixth and fourth cranial nerve palsies in idiopathic intracranial hypertension. J. R. Soc. Med. 93:80, 2000. 91. Pedraza, S., Gamez, J., Rovira, A., et al.: MRI findings in Möbius syndrome: correlation with clinical features. Neurology 55:1058, 2000.

1205

92. Richards, B. W., Jones, F. R. Jr., and Younge, B. R.: Causes and prognosis in 4,278 cases of paralysis of the oculomotor, trochlear, and abducens cranial nerves. Am. J. Ophthalmol. 113:489, 1992. 93. Rucker, C. W.: Paralysis of the third, fourth and sixth cranial nerves. Am. J. Ophthalmol. 46:787, 1958. 94. Rucker, C. W.: The causes of paralysis of the third, fourth and sixth cranial nerves. Am. J. Ophthalmol. 61:1293, 1966. 95. Rush, J. A.: Extraocular muscle palsies in diabetes mellitus. Int. Ophthalmol. Clin. 24:155, 1984. 96. Rush, J. A., and Younge, B. R.: Paralysis of cranial nerves III, IV, and VI: cause and prognosis in 1,000 cases. Arch. Ophthalmol. 99:76, 1981. 97. Ruttum, M. S., and Harris, G. J.: Orbital blowout fracture with ipsilateral fourth nerve palsy. Am. J. Ophthalmol. 100:343, 1985. 98. Sabates, F. N. Jr., Tsai, F., Sabates, N. R., and Blitsein, B.: Transient cranial nerve palsies after cavernous sinus fistula embolization. Am. J. Ophthalmol. 111:771, 1991. 99. Sadun, F., DeNegri, A. M., Santopadre, P., and Pezzi, P. P.: Bilateral trochlear nerve palsy associated with cryptococcal meningitis in human immunodeficiency virus infection. J. Neuroophthalmol. 19:118, 1999. 100. Sato, M.: Magnetic resonance imaging and tendon anomaly associated with congenital superior oblique palsy. Am. J. Ophthalmol. 127:379, 1999. 101. Schatz, N. J., Savino, P. J., and Corbett, J. C.: Primary aberrant oculomotor regeneration: a sign of intracavernous meningioma. Arch. Neurol. 34:29, 1977. 102. Schneider, R. C., and Johnson, F. D.: Bilateral traumatic abducens palsy. J. Neurosurg. 34:33, 1971. 103. Selky, A. K., and Purvin, V. A.: Isolated trochlear nerve palsy secondary to dural carotid-cavernous sinus fistula. J. Neuroophthalmol. 14:52, 1994. 104. Sibony, P. A., Evinger, C., and Lessell, S.: Retrograde horseradish peroxidase transport after oculomotor nerve injury. Invest. Ophthalmol. Vis. Sci. 27:975, 1986. 105. Sibony, P. A., and Lessell, S.: Transient oculomotor synkinesis in temporal arteritis. Arch. Neurol. 41:87, 1984. 106. Sitwell, L. D., Weinshenker, B. G., Monpetit, V., and Reid, D.: Complete ophthalmoplegia as a complication of acute corticosteroid- and pancuronium-associated myopathy. Neurology 42:939, 1992. 107. Smith, B. E., and Dyck, P. J.: Subclinical histopathological changes in the oculomotor nerve in diabetes mellitus. Ann. Neurol. 32:376, 1992. 108. Speer, C., Pearlman, J., Phillips, P. H., et al.: Fourth cranial nerve palsy in pediatric patients with pseudotumor cerebri. Am. J. Ophthalmol. 127:236, 1999. 109. Sunderland, S., and Hughes, E. S. R.: The pupilloconstrictor pathway and the nerves to the ocular muscles in man. Brain 69:301, 1946. 110. Tekdemir, I., Tuccar, E., Cubuk, H. E., et al.: Branches of the intracavernous internal carotid artery and the blood supply of the intracavernous cranial nerves. Anat. Anz. 180:343, 1998. 111. Thomas, J. E., and Yoss, R. E.: The parasellar syndrome: problems in determining etiology. Mayo Clin. Proc. 45:617, 1970. 112. Van Meter, W. S., Younge, B. R., and Harner, S. G.: Ophthalmic manifestations of acoustic neurinoma. Ophthalmology 90:917, 1983.

1206

Diseases of the Peripheral Nervous System

113. Vanooteghem, P., Dehaene, I., Van Zandycke, M., and Casselman, J.: Combined trochlear nerve palsy and internuclear ophthalmoplegia. Arch. Neurol. 49:108, 1992. 114. Varma, R., and Miller, N. R.: Primary oculomotor nerve synkinesis caused by an extracavernous intradural aneurysm. Am. J. Ophthalmol. 118:83, 1994. 115. Volpe, N. J., Liebsch, N. J., Munzenrider, J. E., and Lessell, S.: Neuro-ophthalmologic findings in chordoma and chondrosarcoma of the skull base. Am. J. Ophthalmol. 115:97, 1993. 116. Wallace, D. K., and von Noorden, G. K.: Clinical characteristics and surgical management of congenital absence of the superior oblique tendon. Am. J. Ophthalmol. 118:63, 1994. 117. Walter, K. A., Newman, N. J., and Lessell, S.: Oculomotor palsy from minor head trauma: initial sign of intracranial aneurysm. Neurology 44:148, 1994. 118. Warwick, R.: A study of retrograde degeneration in the oculomotor nucleus of the rhesus monkey, with a note on a method of recording its distribution. Brain 73:532, 1950. 119. Warwick, R.: Representation of the extra-ocular muscles in the oculomotor nuclei on the monkey. J. Comp. Neurol. 98:449, 1953. 120. Warwick, R.: The ocular parasympathetic nerve supply and its mesencephalic sources. J. Anat. 88:71, 1954.

121. Warwick, R.: Oculomotor organisation. Ann. R. Coll. Surg. Engl. 19:36, 1956. 122. Weber, R. B.: Cranial neuropathies and diabetes [letter]. Lancet 1:645, 1971. 123. Weber, R. B., Daroff, R. B., and Mackey, E. A.: Pathology of oculomotor nerve palsy in diabetics. Neurology 20:835, 1970. 124. Wilcsek, G. A., Francis, I. C., Egan, C. A., et al.: Superior oblique palsy in a patient with a history of perineural spread from a periorbital squamous cell carcinoma. J. Neuroophthalmol. 20:240, 2000. 125. Wolff, E.: Anatomy of the Eye and Orbit: Including the Central Connections, Development, and Comparative Anatomy of the Visual Apparatus, 6th ed. (revised by Last, R. J.). London, H. K. Lewis & Company, pp. 283, 286, 1968. 126. Wright, R. A., Plant, G. T., Landon, D. N., MorganHughes, J. A.: Nemaline myopathy: an unusual cause of ophthalmoparesis. J. Neuroophthalmol. 17:39, 1997. 127. Younge, B. R., and Sutula, F.: Analysis of trochlear nerve palsies: diagnosis, etiology and treatment. Mayo Clin. Proc. 52:11, 1977. 128. Zweifach, P. H., Walton, D. S., and Brown, R. H.: Isolated congenital horizontal gaze paralysis: occurrence of the near reflex and ocular retraction on attempted lateral gaze. Arch. Ophthalmol. 81:345, 1969.

49 Diseases of the Fifth Cranial Nerve RICHARD A. C. HUGHES

Anatomy and Physiology Trigeminal Nerve Divisions Trigeminal Ganglion Trigeminal Sensory Root Trigeminal Nuclei Trigeminal Motor Root Trigeminal Nerve Lesions Trauma Perineurial Spread of Tumors

Trigeminal Schwannomas Other Trigeminal Nerve Tumors Toxins Congenital Trigeminal Anesthesia Idiopathic Trigeminal Neuropathy Investigation of Trigeminal Nerve Lesions Trigeminal Neuralgia Clinical Features

ANATOMY AND PHYSIOLOGY The fifth cranial nerve provides sensation to the skin of much of the head, face, mouth, and nasal cavity and motor and proprioceptive innervation of the muscles involved in chewing. Numbness and pain in its territory are often more distressing to patients than similar symptoms in spinal nerve dermatomes. The ganglion of the fifth cranial nerve lies in the floor of the middle cranial fossa. The sensory root passes obliquely backward and medially to enter the lateral aspect of the pons. The motor root is much smaller, lies on the medial side of the sensory root and ganglion, and joins the mandibular division.

Trigeminal Nerve Divisions The three main divisions of the trigeminal nerve are the ophthalmic, maxillary, and mandibular nerves (Fig. 49–1). They supply the skin of the face, sparing the angle of the jaw, the scalp back to the bregma, the anterior part of the tragus and upper anterior part of the pinna of the ear, the anterior wall of the external auditory meatus, and the anterior part of the tympanic membrane (Fig. 49–2). In addition, they supply somatic sensation to the mouth, nasal sinuses, and the dura of the anterior and middle cranial fossae. Trigeminal ganglion or sensory root lesions

Triggers Associated Clinical Syndromes Differential Diagnosis Causes Pathology Medical Treatment

cause variable impairment of taste sensation on the ipsilateral tongue and palate. This may be because both common sensation, subserved by the trigeminal nerve, and special sensation, subserved by the facial and glossopharyngeal nerves, are necessary for the full appreciation of taste.4,19,54

Trigeminal Ganglion The trigeminal ganglion is shaped like a crescent. The three divisions enter on its lateral convex surface and the sensory root leaves from the concave medial surface. The upper and lower borders are folded inward to enclose a sinus lined by arachnoid and connecting with the subarachnoid space of the middle cranial fossa. The bundles of the sensory root form a triangle and are bathed in cerebrospinal fluid as they pass posteriorly through this sinus and join into a compact single root, which passes backward to reach the lateral pontine cistern and then the pons. The ganglion is surrounded by the dura of the middle cranial fossa. A layer of dura invests the triangular portion of the sensory root and is invaginated by a pouch of dura from the posterior cranial fossa to form the trigeminal pore, through which the sensory root reaches the posterior fossa. The microscopic anatomy of the trigeminal ganglion resembles that of the posterior spinal root ganglia (see Chapter 8).28 1207

1208

Diseases of the Peripheral Nervous System

FIGURE 49–1 The divisions of the trigeminal nerve and their branches. (From Warwick, R., Williams, P. L., Dyson, M., and Bannister, L. H.: Gray’s Anatomy, 37th ed. Edinburgh, Longman, 1989, with permission.)

Trigeminal Sensory Root The sensory rootlets emerge in a wedge shape from the concave medial surface of the ganglion and pass backward and medially through the trigeminal pore to reach the lateral aspect of the pons. The fibers from the mandibular division maintain a posterolateral position, those from the maxillary are intermediate, and those from the ophthalmic are situated dorsomedially throughout the course of the root. Accessory rootlets may end separately in the pons and mostly originate from the group of fibers from the ophthalmic division. There are usually anastomoses, of uncertain function, between the sensory and motor roots. These and the accessory rootlets might account for the preservation of some somatic sensory function following surgical division of the main sensory root in the posterior fossa.4

The nerve fibers in the sensory roots consist of myelinated and unmyelinated fibers in proportions similar to those found in posterior spinal roots.28 The myelin sheaths are formed by Schwann cells up to a point 1.5 mm outside the pons, where there is an abrupt transition to oligodendrocyte myelination.53

Trigeminal Nuclei The sensory fibers enter the pons and proceed to the main sensory nucleus, the nucleus of the spinal tract, or the mesencephalic nucleus. Half the fibers divide into an ascending branch that ends in the main sensory nucleus and a descending branch that ends in the nucleus of the spinal tract. The nucleus of the spinal tract is subdivided into

1209

Diseases of the Fifth Cranial Nerve

Trigeminal Motor Root The motor root emerges from the lateral border of the pons medial to the sensory root and passes forward beneath the sensory root and trigeminal ganglion to join the mandibular division. The first 0.3-mm portion resembles a central nervous system tract and is myelinated by oligodendrocytes, whereas the distal portion is myelinated by Schwann cells.53 The motor fibers innervate the muscles of mastication and the tensor tympani and tensor veli palatini.4

TRIGEMINAL NERVE LESIONS

FIGURE 49–2 The territory supplied by the trigeminal nerve. (From Brodal, A.: Neurological Anatomy in Relation to Clinical Medicine, 3rd ed. New York, Oxford University Press, 1991. Copyright 1969, 1981 by Oxford University Press, Inc. Used by permission of Oxford University Press, Inc.)

the nucleus caudalis, nucleus interpolaris, and nucleus oralis. Some mandibular division fibers leave the spinal tract and enter the tractus solitarius, possibly conveying visceral afferent information from the submandibular gland or blood vessels or possibly subserving taste. Some fibers from all three trigeminal divisions end in the medial cuneate nucleus.4,10 Based on general clinical experience and especially the results of medullary trigeminal tractotomy, the main sensory nucleus is considered to subserve light touch. The spinal tract and nucleus subserve pain and temperature sensation. The mesencephalic nucleus of the trigeminal nerve extends approximately from the rostral tip of the main sensory nucleus in the pons to the superior colliculus. The neurons of the mesencephalic nucleus are mostly unipolar and resemble those of the dorsal root ganglia. The cell bodies belong to the afferent neurons of the muscle spindles of the jaw-closing muscles and a part of the population of lower threshold mechanoreceptors in the periodontal membrane. The nucleus of the spinal tract receives input from the descending branch of most large myelinated fibers of the trigeminal nerve. The nucleus caudalis of the spinal tract of the trigeminal nerve subserves pain sensation from the whole of the trigeminal nerve territory, although pain sensation from the oral cavity and teeth is probably also subserved by more rostral parts of the nucleus, especially the nucleus oralis.4

Facial numbness may be produced by lesions of the central sensory pathways, the trigeminal nerve nuclei, the sensory root, the trigeminal ganglion, the trigeminal divisions, or their branches. Accompanying symptoms and signs may provide clues to the site of the lesion. In patients with isolated facial numbness, it may be necessary to consider and investigate the possibility of a wide differential diagnosis (Table 49–1). The frequency of the different causes of trigeminal nerve lesions is difficult to estimate in developed countries. Herpes zoster and dental and accidental skull base or facial trauma are the most common causes, with cerebellopontine angle and neck tumors in second and idiopathic trigeminal neuropathy in third place.12,22 Worldwide, leprosy is an important cause of impaired facial sensation. Pain is prominent when trigeminal sensory loss is due to local spread of nasopharyngeal carcinoma or infiltration by metastases.49 The other causes listed in Table 49–1 are rare.

Trauma The branches, divisions, and roots of the trigeminal nerves are at risk from accidental and surgical trauma. The variety of traumatic lesions is immense. The trigeminal nerve divisions or their branches are sometimes damaged in skull base or facial bone fractures. In one curious case after a skull base fracture causing right third, fifth, sixth, and seventh cranial nerve palsies, the patient developed involuntary abduction of the right eye.35 This unique trigeminoabducent synkinesis must have been due to regrowth of fibers originally supplying muscles of mastication into the sixth nerve. The anastomosis was probably at the apex of the petrous temporal bone. The inferior alveolar nerve is at risk during removal of impacted third molar teeth. Damage causes numbness of the molar and premolar teeth followed by pain, which may persist.12 The numb chin syndrome occurs with mandibular fracture, after dental extraction, with neoplastic infiltration,5 and in elderly people because of resorption of bone from the mandible, exposing the mental nerve to pressure.11

1210

Diseases of the Peripheral Nervous System

Table 49–1. Differential Diagnosis of Trigeminal Nerve Lesions Intrinsic brainstem lesions Infarction Hemorrhage Multiple sclerosis Syringobulbia Intrinsic tumors Congenital trigeminal anesthesia (some cases) Extrinsic posterior fossa lesions Trigeminal neurofibroma or schwannoma Meningioma Aneurysms: anterior and posterior cerebellar, inferior basilar Granulomas Foramen magnum anomalies Middle cranial fossa lesions Herpes zoster Trigeminal nerve schwannoma Meningioma Metastasis Lymphoma Carotid aneurysm Skull base lesions Nasopharyngeal carcinoma Metastasis Lymphoma Neural spread of basal or squamous cell carcinoma Facial bone fractures Dental trauma Drug intoxication Hydroxystilbamidine Trichloroethylene Oxaliplatin Basal meningitis Sarcoidosis Syphilis Tuberculosis Amyloidosis Chronic inflammatory demyelinating polyradiculoneuropathy14 Vasculitis Benign trigeminal sensory neuropathy Modified from Horowitz, S. H.: Isolated facial numbness: clinical significance and relation to trigeminal neuropathy. Ann. Intern. Med. 80:49, 1974.

Perineurial Spread of Tumors Perineurial spread of cancer is an uncommon but recognized feature of some head and neck cancers, especially adenoid cystic carcinomas of the nasopharynx, salivary glands, and paranasal cavities and skin tumors, including melanoma and squamous cell carcinoma of the skin.3,51

Trigeminal Schwannomas Trigeminal schwannomas account for approximately 0.2% of all intracranial tumors.45 They occur in adults, usually in middle age. They commonly arise in the region of the trigeminal ganglion and not from the motor root. Half arise in the middle cranial fossa and one fourth in the posterior fossa, and one fourth are dumbbell shaped, arising in both. Trigeminal schwannomas grow slowly as well-demarcated tumors with a thin capsule. They often reach a diameter of at least 2.5 cm before they are diagnosed. They usually have a smooth surface readily separated from the brain but often adherent to the dura or the wall of the cavernous sinus. Their texture varies from firm and rubbery to soft and brittle according to the extent of fatty degeneration, cyst formation, and calcification. Trigeminal schwannomas commonly erode the medial part of the floor of the middle cranial fossa and the apex of the petrous temporal bone. They can often be completely removed by skull base surgery with a lower morbidity than with older approaches.59 Most surgical series report isolated trigeminal schwannomas. However, trigeminal schwannomas are not uncommon in neurofibromatosis type 2, in which they may be only one of several tumors so that surgical resection may not be appropriate. Malignant schwannomas also arise from the trigeminal ganglion or its divisions. Malignant change cannot be determined reliably from the clinical or radiographic features, and its diagnosis is dependent on histologic examination.59 The clinical presentation is very variable. Numbness or paresthesias in the distribution of the trigeminal nerve are the most frequent presentation, occurring in approximately two thirds of patients. Intermittent painful burning or crawling sensations and trigeminal neuralgia are less common. A tumor growing into the lateral wall of the cavernous sinus produces in turn sixth, fourth, and third cranial nerve palsies, proptosis, and visual failure. A posterior fossa tumor causes deafness, tinnitus, ipsilateral cerebellar ataxia, and rarely facial weakness or hemispasm and lower cranial nerve palsies. Hydrocephalus and papilledema are late manifestations.

Other Trigeminal Nerve Tumors Other tumors that occur in the region of the trigeminal ganglion are meningiomas, gangliocytomas (rare primary malignant tumors of the ganglion), epidermoids, chondromas, chondromyxomas, chordomas, sarcomas, lymphomas, fibrous xanthomas (also called fibrous histiocytomas), amyloidomas,40 and hemangioblastomas. These are all rare except for meningiomas, which are more common tumors of Meckel’s cave than trigeminal schwannomas. Meningiomas spread in a plaque along the floor of the middle cranial fossa and along the trigeminal root into the

Diseases of the Fifth Cranial Nerve

posterior fossa. Meningiomas affect the third nerve more commonly than trigeminal schwannomas. Sometimes an internal carotid aneurysm may compress the trigeminal ganglion, usually affecting the third cranial nerve at the same time. Rarely, the ganglion may be affected by a syphilitic granuloma or an actinomycotic abscess. The trigeminal nerve may be markedly hypertrophied in chronic inflammatory demyelinating polyradiculoneuropathy.14 Trigeminal tumors displace branches of the internal carotid and basilar arteries. Magnetic resonance imaging is the technique of choice for displaying tumors of the trigeminal ganglion.58

Toxins Stilbamidine treatment may be complicated by pruritus and paresthesias of the face as a result of an unknown mechanism.6,47 Intoxication with trichloroethylene, either when used in general anesthesia or as a result of industrial accidents, has produced bilateral trigeminal nerve lesions, sometimes including the motor fibers.36 After general anesthesia, numbness and paresthesias begin around the mouth within 24 hours and spread to involve the whole trigeminal nerve territory over the next few days. Other cranial nerves may be affected, and recovery may be slow and incomplete. In a case of industrial intoxication, there was striking loss of nerve fibers in the trigeminal sensory roots and descending tracts.39 The new cytotoxic drug oxaliplatin produces acute perioral paresthesias and neuromyotonia and later sensory neuropathy, which limit the dose that can be given.56

Congenital Trigeminal Anesthesia Congenital trigeminal nerve lesions are rare. The presenting symptom and danger is corneal ulceration in infants and young children, which requires treatment by patching or suturing the eyelids together. As the children grow older, corneal ulceration becomes less of a problem. There are three subgroups, isolated trigeminal nerve palsy, trigeminal nerve palsy associated with mesenchymal abnormalities, and trigeminal nerve palsy associated with other cranial nerve palsies.44 In the first subgroup, congenital trigeminal anesthesia is an isolated finding and is usually bilateral. The children present with corneal ulceration at 8 to 12 months, an age when they are sleeping less and exposing the cornea for longer periods during the day.57 The condition may be inherited as an autosomal dominant trait.29 It is presumably due to primary hypoplasia of the trigeminal nerve nuclei, but no postmortem studies have been undertaken to confirm this hypothesis. In the second subgroup, congenital trigeminal anesthesia is associated with Goldenhar’s or Möbius’ syndromes. Goldenhar’s syndrome, or oculoauriculovertebral dysplasia,

1211

consists of abnormalities affecting the first and second branchial arches. It causes hypoplasia of the face, zygoma, and chin with a prominent forehead, epibulbar epidermoid, and preauricular tags. It affects 1 in 3000 to 5000, usually male, live births. It is usually sporadic but some cases are autosomal dominant.50 In a case of congenital trigeminal anesthesia with Goldenhar’s syndrome, the fifth and seventh nerve nuclei were hypoplastic, which would suggest a primary neural abnormality.1 Because the target organ provides a trophic function for the innervating neurons, it remains difficult to decide which is the primary pathologic event. Möbius’ syndrome consists of congenital facial palsy and a horizontal gaze disorder. In one case, the cranial nerve nuclei were normal but the muscles were abnormal at postmortem examination, supporting the idea that the pathology is a primary mesenchymal dysplasia.42 The third subgroup is very rare and consists of trigeminal anesthesia with associated neurologic abnormalities indicating a focal intrinsic pontine lesion.44

Idiopathic Trigeminal Neuropathy The diagnosis of idiopathic trigeminal neuropathy can be made only after consideration and exclusion of the many underlying causes of impaired sensation in the trigeminal nerve territory (see Table 49–1). Motor function is rarely affected in idiopathic trigeminal neuropathy. Even after extensive investigation, approximately 10% of cases of trigeminal nerve lesions have no evident cause, as in 7 of 61 cases in a series from the Mayo Clinic.12 In some idiopathic cases loss of facial sensation is transient. In 5 of 10 cases, facial numbness developed suddenly, usually in the second or third division territories, spread to adjacent divisions over a few days or weeks, and then gradually recovered over weeks or months.2 Some patients reported pins and needles but none reported pain. The cerebrospinal fluid was normal in the two cases in which it was examined. The authors stressed the need to exclude the possibilities of syphilis or malignant disease in the nasopharynx and compared these transient trigeminal nerve lesions to Bell’s palsy and transient benign sixth cranial nerve palsy. In some patients the development of facial numbness is associated with herpes labialis in the same territory.13 In two patients with transient trigeminal sensory neuropathy, magnetic resonance imaging showed enlargement and marked gadolinium enhancement of the main sensory root and trigeminal ganglion.43 In many cases trigeminal sensory neuropathy is chronic and unremitting.49 Lecky and colleagues reported 22 patients with this chronic form of idiopathic trigeminal neuropathy.31 Symptoms usually began insidiously with numbness around the mouth, often accompanied by paresthesias with a painful quality. Severe pain did not occur and the pain did not have the paroxysmal features of trigeminal neuralgia. The area of sensory impairment

1212

Diseases of the Peripheral Nervous System

enlarged gradually over months, remaining confined to the trigeminal nerve territory but extending sequentially to the other side of the face in 13 of 22 cases. In two thirds of the affected trigeminal nerve territories, all three divisions had become affected. In the trigeminal nerve lesions with partial involvement, the ophthalmic division was usually spared. Only one case had involvement of the motor root. Taste was impaired on the anterior two thirds of the tongue in 15 cases. Electrophysiologic recording of the blink reflex evoked in the orbicularis oculi by stimulation of the supraorbital nerve revealed delayed responses consistent with an afferent lesion in 6 of the 17 cases in which the recordings were undertaken. The delays were mild, consistent with a short demyelinating lesion or dropout of the fastest conducting fibers. The other recordings either were normal or elicited varied abnormalities. Trigeminal somatosensory responses evoked in the opposite somatosensory cortex were either normal or degraded or delayed, sometimes correlating with abnormality in the blink reflex prolongation. In a patient discussed by Spillane and Wells, exploration 4 months after the onset showed the ganglion, sensory root, and first and second divisions to be “grey and pink in color and not the normal white appearance.”49 In a patient explored after 3 years, there was no evidence of swelling or inflammation. In each of three patients with chronic isolated sensory trigeminal neuropathy operated on after 1 to 12 years, the sensory root was reduced to a few wisps of nerve fiber.23 Sections of the trigeminal ganglion showed cell loss, lymphocyte infiltration fibrosis, and hyaline degeneration. Examination of a biopsy specimen of a supraorbital nerve from a patient who had had unilateral sensory involvement affecting all three divisions for 1 year showed very severe loss of myelinated nerve fibers.31 Electron microscopy showed many unmyelinated axonal and Schwann cell processes. There were abundant stacks of redundant basal lamina. Schwann cell processes were pale. The perineurium showed nonspecific changes, including necrotic cells and irregular thickening of the basal lamina. The findings suggested that the primary event was a destructive lesion in the trigeminal ganglion or proximal part of the nerve. In two cases there was gadolinium enhancement of the cisternal segment of the trigeminal nerve; no comment was made about whether the enhancement extended into the ganglion.46 The available evidence implicates an inflammatory process involving the trigeminal ganglion itself, eventually causing degeneration of both the divisions and their branches and the main sensory root but usually sparing the motor root. The cause of the inflammation in the trigeminal ganglion remains obscure, but the association between trigeminal neuropathy and connective tissue disorders offers a clue. Hagen et al.15 reported a retrospective series of 81 patients with trigeminal sensory neuropathy associated with connective tissue disease. The facial numbness and connective

tissue disease were diagnosed simultaneously in half the cases, but the facial numbness preceded the connective tissue disease in 7%. The associated disorders were undifferentiated connective tissue disease in 47%, mixed connective tissue disease in 26%, and scleroderma in 19%. The numbness sometimes improved but always persisted to some extent. No further clues concerning etiology emerged, but the authors proposed that the pattern of sensory loss was more likely consistent with a lesion in the trigeminal ganglion than elsewhere. Vasculitic damage to the trigeminal ganglion may explain trigeminal neuropathy occurring in the absence of overt autoimmune disease. Alternatively, autoantibodies reactive with peripheral nerve components may have easier access to neural antigens in the trigeminal ganglion than to antigens elsewhere in the nervous system. The blood vessels of the trigeminal ganglion, like those of the dorsal root ganglia, are readily permeable to large molecules such as horseradish peroxidase.24 Four of the cases in the series of Lecky et al.31 had had dental sepsis or dental treatment before the onset of their symptoms. One of their cases and 4 of the 16 cases of Spillane and Wells49 had had sinus disease, raising the possibility that local inflammation upregulated the expression of major histocompatibility complex class II antigen in the trigeminal ganglion, where an increased number of cells express this antigen in any case. This would increase the likelihood of antigen being presented to circulating T cells and triggering an inflammatory reaction. Such suggestions are speculative, and there is a need for information from postmortem examinations. Although corticosteroids have been tried, usually rather halfheartedly, no effective treatment has been found. Fortunately, most patients tolerate the symptom well once they have been reassured that it does not have a sinister cause and that the numbness will not spread outside the trigeminal nerve territory. A few patients remain distressed by the distortion of sensation. The trigeminal trophic syndrome, ulceration of the nose and face, is rare.16 Difficulties with chewing are also uncommon despite almost complete loss of cutaneous sensation, presumably because the motor root remains intact. All patients with impaired corneal sensation should be counseled to protect their eyes by wearing glasses, preferably with protective sidepieces, and to inspect the conjunctiva each evening for signs of inflammation. The price of not doing this can be perforating keratitis.

INVESTIGATION OF TRIGEMINAL NERVE LESIONS Investigation of trigeminal nerve lesions should be guided by the history to identify or eliminate the possible causes shown in Table 49–1. The presence of pain makes malignancy more likely. Full neurologic and ear, nose, and throat

1213

Diseases of the Fifth Cranial Nerve

examinations, including the postnasal space, are essential. Blood tests should usually include a complete blood count, erythrocyte sedimentation rate, serologic tests for syphilis, and antibodies to DNA and extractable nuclear antigens. Magnetic resonance imaging of the skull base with gadolinium enhancement is the favored imaging technique.58

vascular loop. Paroxysms of pain occur in episodes. In a population-based study, episodes lasted an average of 49 days but with a very wide range: 29% had only one episode, 19% had two, and 52% had more episodes; 65% had a second episode within 5 years and 77% within 10 years.27

Triggers TRIGEMINAL NEURALGIA Trigeminal neuralgia is a characteristic syndrome of short paroxysms of severe facial pain in the territory of one or more divisions of the trigeminal nerve. The annual incidence rises with age from 8 to 9 per 100,000 in the fifth, to 17.5 per 100,000 in the sixth, and 25 per 100,000 in the seventh decade. Women are affected 1.7 times more often than men.27 The condition is usually sporadic, but rare familial examples of trigeminal neuralgia have been reported. In some instances trigeminal neuralgia has occurred in Charcot-Marie-Tooth disease and been cured by microvascular decompression, as described and reviewed by de Matas et al.9 The authors did not specify the type of Charcot-Marie-Tooth disease.

Clinical Features Trigeminal neuralgia is an easily recognizable syndrome of excruciating paroxysms of pain in the face. The pain is almost always unilateral and usually affects the mandibular or maxillary territories. The pain is usually confined to one or more of these territories from the onset but may rarely spread to adjacent territories. A touch or movement in the territory of the same division characteristically triggers it. In 60% of patients the pain begins in the lower gum in the canine or premolar region, or in the upper gum in a similar area, and then radiates toward the ear or from one gum to the other. In a few patients pain is felt deeply inside the ear, or in front of the tragus, and may remain there for up to 6 years before spreading to the gum or tongue. In approximately one third of patients pain originates in the upper gum and radiates upward toward the junction of the nose and cheek, or the inner or outer canthus. In a few instances it is localized behind the orbit or spreads from there into the upper gum. The pain usually remains confined to one zone, even at follow-up more than 20 years after the onset of the disease.20 Each paroxysm of pain is brief, usually lasting only a few seconds or at the most 1 or 2 minutes, and extremely severe, likened to an electric shock, lightning, or a red-hot wire. Patients live in dread of the next paroxysm. The pain makes them screw up their faces and avoid triggering factors. Between paroxysms of pain most patients are free of symptoms, except for anxiety about the next attack. Some patients with very frequent paroxysms have a persistent constant ache in the affected part of the face between attacks, which may indicate a structural lesion other than a

Characteristically, the spasms of trigeminal neuralgia are triggered by eating, talking, yawning, blowing the nose, cleaning the teeth, or shaving, washing, or lightly rubbing the affected side of the face. Movements of the tongue may bring on mandibular neuralgia, and combing the hair may provoke tic in the ophthalmic division. Proprioceptive or tactile stimuli are more effective triggers than deep pressure or pain. To avoid the triggers, patients may stop eating and swallowing and become emaciated and dehydrated. They may stop washing, shaving, or cleaning their teeth so that oral hygiene is compromised. Sometimes precipitating factors can be revealed only by direct questioning. Rarely, the pain is provoked by moving the eyes, by a sudden loud noise, by an unexpected jolt such as a missed step, or by intense emotional upset. Most patients have freedom from pain during sleep, probably because of the absence of triggering mechanisms. Approximately one third of patients have trigger spots, most frequently found on the upper or lower lip, outer ala nasae, nasolabial fold, upper eyelids, gum, or tongue. Peet and Schneider41 recorded trigger zones in 218 of their 689 patients (31.6%). Pain elicited by touching these trigger spots usually occurs in the same trigeminal division, but in exceptional cases the trigger area may lie within another, unaffected dermatome. Very rarely a patient can adopt maneuvers that diminish the intensity of the pain or abort an attack. In “idiopathic” trigeminal neuralgia no deficit can be discovered on conventional neurologic examination. Rarely, slight alterations in the quality of sensation in the territory affected by the pain are reported in otherwise typical cases during or soon after bouts of severe pain. These disappear when the patient has been free of pain for a period. The presence of frank sensory deficit should lead to a careful search for a structural cause. Even in the absence of sensory deficit, records of laser-evoked potentials indicate dysfunction of the A␦ thinly myelinated fibers in the affected trigeminal nerve in approximately half of cases of idiopathic trigeminal neuralgia.7

Associated Clinical Syndromes Trigeminal neuralgia is sometimes associated with hemifacial spasm on the same side, called “tic convulsif” by Cushing.8 The paroxysms of trigeminal neuralgia occur independently of the hemifacial spasm. Either symptom may precede the other, and the interval may be only a few months or as much as a quarter of a century. The association of the two

1214

Diseases of the Peripheral Nervous System

syndromes usually indicates identifiable pathology, including vascular compression and the range of tumors listed in Table 49–2. There are also rare reports of the coincidence of trigeminal and glossopharyngeal neuralgia, sometimes resulting from multiple sclerosis or vestibular schwannoma.

Differential Diagnosis Trigeminal neuralgia is so characteristic that it does not really have a differential diagnosis. Glossopharyngeal neuralgia may at times be difficult to differentiate. Coughing and swallowing, which are the main triggers of glossopharyngeal neuralgia, only rarely precipitate trigeminal tic. Radiation of pain to the tongue, palate, and eardrum may occur in both conditions, but the tonsillar area is a common site of pain in glossopharyngeal but not trigeminal neuralgia. Glossopharyngeal neuralgia may cause syncope if paroxysms are frequent. A wide variety of dental conditions give rise to pain in the upper or lower jaw. Pain from dental and alveolar disease is common but lacks the characteristic features of trigeminal neuralgia. Temporomandibular joint dysfunction and malocclusion can cause facial pain that is characteristically worse on eating and chewing. Migrainous neuralgia causes episodes of pain in and around the eye occur in clusters usually lasting several weeks, with each bout of pain lasting between 30 minutes and 2 hours. An ipsilateral watering eye and blocked nostril usually accompany the pain. Horner’s syndrome sometimes occurs. Another rare cause of brief headache, short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing (SUNCT syndrome), causes brief stabs of pain in the periorbital region accompained by conjunctival injection and lacrimation, which are absent in trigeminal neuralgia.34 Table 49–2. Causes of Trigeminal Neuralgia Vascular compression Ectopic loop of superior cerebellar artery Ectopic loop of inferior cerebellar artery Sacccular aneurysm Arteriovenous malformation Tortuous basilar artery Primitive trigeminal artery 37 Tumors Vestibular schwannoma Trigeminal schwannoma Meningioma Epidermoid cyst Bony deformity from osteogenesis imperfecta Amyloidoma40 Multiple sclerosis Charcot-Marie-Tooth disease See Love and Coakham32 for references.

Causes Trigeminal neuralgia is usually caused by compression of the trigeminal sensory root by arterial loops. The artery usually responsible for compressing the trigeminal nerve root is the superior cerebellar artery. In 100 operations for trigeminal neuralgia, compression of the superomedial side of the nerve by the superior cerebellar artery correlated with pain in the second and third divisions, whereas in four patients with first division neuralgia, the anterior inferior cerebellar artery indented the inferior surface of the trigeminal root.25 Other rare compressive causes are listed in Table 49–2. Multiple sclerosis is an important consideration in the differential diagnosis. Approximately 2% to 4% of patients with trigeminal neuralgia have multiple sclerosis, and 1% to 5% of patients with multiple sclerosis have trigeminal neuralgia.21,38 In 80% to 90% of cases, symptoms of multiple sclerosis precede the onset of trigeminal neuralgia by intervals varying from 1 to 29 years; in the remainder, tic is the first symptom, and the signs of multiple sclerosis appear from 1 month to 11 years later. Trigeminal neuralgia occurs at a younger age, has a trigger less often, and is more commonly bilateral when associated with multiple sclerosis than otherwise.17,18 The response to drugs, different procedures to damage the ganglion, and section of the main sensory root seem to be as good in cases with multiple sclerosis as without.21

Pathology In so-called idiopathic trigeminal neuralgia resulting from arterial loop compression, the proximal central nervous system portion of the trigeminal nerve root shows a small (2-mm diameter) focus of myelin loss where demyelinated axons are closely apposed.32 Sometimes multiple axons are embraced by a single myelin sheath. Astrocytic processes are confined to the periphery, and inflammatory cells are absent. Neighboring axons are often thinly myelinated, suggesting that they have undergone demyelination followed by remyelination. Demyelination is a known consequence of compression in chronic experimental lesions. In those specimens in which no lesion has been found, the small size of the lesion may have led to it being excluded from the operative sample. In cases of multiple sclerosis, demyelinating plaques have been described extending into the proximal central nervous system portion of the trigeminal root. The demyelinated focus is more extensive and astrocytic processes are more abundant than in idiopathic trigeminal neuralgia. In their excellent review, Love and Coakham32 developed the hypothesis that the demyelinated axons develop ectopic bursts of impulses, which are interpreted as pain.

1215

Diseases of the Fifth Cranial Nerve

Medical Treatment Many cases of trigeminal neuralgia can be adequately controlled with carbamazepine, an iminostilbene derivative chemically related to imipramine hydrochloride. Three randomized trials involving 539 participants have demonstrated that the rate ratio for benefit is 2.9 (95% confidence intervals, 2.2 to 3.9) compared with placebo.55 It is effective in approximately 70% of cases. Carbamazepine usually controls pain within 12 to 48 hours, and cessation of treatment produces an equally rapid recurrence. Once control has been achieved, the patient must continue taking the drug at the same dose for at least 2 weeks. Only after such a pain-free period should cautious reduction of the dosage be tried to find the minimum dose that will prevent the paroxysms. Unfortunately, the drug may lose its effect and has a number of side effects, including a morbilliform rash; blurred vision, unsteadiness, or drowsiness; bone marrow depression; and hyponatremia. Minor side effects can be minimized by gradual increases in dosage. Possible alternatives include lamotrigine,33 oxcarbazepine, gabapentin,30,48 clonazepam, phenytoin, baclofen, tizanidine, pimozide, valproate, and topiramate.60 Procedures to Anesthetize the Trigeminal Nerve Several techniques have been used to anesthetize the trigeminal nerve territory for the relief of trigeminal neuralgia.38 The simplest is neurectomy of the infraorbital or supraorbital nerves, but this is often followed by early recurrence. Cryotherapy exposes and then freezes the affected branch of the trigeminal nerve but uncommonly produces complete relief. An old treatment, alcohol injection into the nerve or ganglion, is now rarely used. It was painful, caused persistent dysesthesias, and had a high recurrence rate. Glycerol injection into the ganglion remains popular in many centers. The injection is made under fluoroscopic control with a combination of local and general anesthesia. The complication rate is low and a high proportion of patients achieve complete pain relief, but recurrence rates are 10% to 53% after 1 year and 34% to 83% after 5 years.38 Radiofrequency gangliolysis was introduced by Sweet and Wepsic52 and has remained popular. It is carried out in a way similar to glycerol injection. The thermocoagulation needle is inserted and its position checked by asking the patient to report the site of tingling induced by stimulation with a needle electrode; then the patient is anesthetized to receive cycles of heat at 60° to 90° C. Nurmikko and Eldridge38 summarized the three largest prospective series, totaling 246 patients, which showed immediate pain relief in nearly all, persistent paresthesias in 5% to 10%, and serious complications of the procedure in 1%. Recurrence occurred eventually in two thirds of patients. Balloon compression of the ganglion

with a Fogarty catheter balloon has been used with good immediate relief of pain, painful dysesthesias in 6% to 7%, and recurrence rates of 20% after 5 years. Finally, stereotactic radiosurgery directing 70 to 90 Gy at the trigeminal ganglion has produced excellent relief in 60% to 70% after approximately 1 year, but no long-term follow-up series are yet available.38 Microvascular Decompression Since the realization that microvascular compression of the trigeminal nerve main sensory root is the explanation of most cases of so-called idiopathic trigeminal neuralgia, operative decompression has become increasingly popular.26 The surgeon performs a suboccipital craniotomy, retracts the superolateral margin of the cerebellum, and reaches the origin of the sensory root. The offending arterial loop is dissected away and separated from the root with Teflon. The proportion of patients reporting initial relief is 87% to 98%, dwindling to 75% to 80% after 1 or 2 years, and 58% to 64% after 8 to 10 years. In early series patients sometimes died from the procedure, but more recently large series have been reported without loss of life. Complications, including cerebellar damage, deafness, and cerebrospinal fluid leak, each occur in 2% or fewer cases in the best series.38 Choice of Treatment There is a lack of comparative randomized trials and systematic reviews to guide the physician and patient in the choice of treatments. Because trigeminal neuralgia often remits, drug treatment is usually a sensible first option, and carbamazepine remains the traditional drug of choice. For those whose symptoms are not adequately controlled by medication, the choice lies between microvascular decompression and one of the procedures used to make a lesion in the trigeminal ganglion. For those who are fit enough, posterior fossa exploration may be the procedure most likely to achieve long-lasting remission without persistent dysesthesias. For less fit patients and those not willing to accept the slight risks of posterior fossa exploration, whichever of the procedures to numb the affected trigeminal nerve division is available at the local neurology center will be preferred. Trigeminal neuralgia can be a difficult and depressing condition; mutual support from a patient support organization is available at www.tna-support.org.

ACKNOWLEDGMENTS The author acknowledges his indebtedness to Dr. George Selby, who contributed this chapter in the first two editions, and thanks Mrs. Lynette Clover-Simpson for secretarial help.

1216

Diseases of the Peripheral Nervous System

REFERENCES 1. Aleksic, S., Budzilkovich, G., Reuben, R., et al.: Congenital trigeminal neuropathy in oculoauriculovertebral dysplasiahemifacial microsomia (Goldenhar-Gorlin syndrome). J. Neurol. Neurosurg. Psychiatry 38:1033, 1975. 2. Blau, J. N., Harris, M., and Kennett, S.: Trigeminal sensory neuropathy. N. Engl. J. Med. 281:873, 1969. 3. Boerman, R. H., Maassen, E. M., Joosten, J., et al.: Trigeminal neuropathy secondary to perineural invasion of head and neck carcinomas. Neurology 53:213, 1999. 4. Brodal, A.: Neurological Anatomy, 3rd ed. New York, Oxford University Press, 1981. 5. Burt, R. K., Sharfman, W. H., Karp, B. I., and Wilson, W. H.: Mental neuropathy (numb chin syndrome): a harbinger of tumor progression or relapse. Cancer 70:877, 1992. 6. Collard, P. J., and Hargreaves, W. H.: Neuropathy after stilbamidine treatment of kala-azar. Lancet 2:686, 1947. 7. Cruccu, G., Leandri, M., Iannetti, G. D., et al.: Small-fiber dysfunction in trigeminal neuralgia: carbamazepine effect on laser-evoked potentials. Neurology 56:1722, 2001. 8. Cushing, H. W.: The major trigeminal neuralgias and their surgical treatment based on experiences with 332 Gasserian operations: varieties of facial neuralgia. Am. J. Med. Sci. 160:157, 1920. 9. de Matas, M., Francis, P., and Miles, J. B.: Microvascular decompression for trigeminal neuralgia in Charcot-MarieTooth disease. J. Neurosurg. 92:715, 2000. 10. Fitzgerald, M.: The course and termination of primary afferent fibres. In Wall, P. D., and Melzack, R. (eds.): Textbook of Pain. New York, Churchill Livingstone, p. 46, 1989. 11. Furukawa, T.: Numb chin syndrome in the elderly. J. Neurol. Neurosurg. Psychiatry 53:173, 1990. 12. Goldstein, N. P., Gibilisco, J. A., and Rushton, J. G.: Trigeminal neuropathy and neuritis. JAMA 184:120, 1963. 13. Gominak, S., Cros, D., and Paydarfar, D.: Herpes simplex labialis and trigeminal neuropathy. Neurology 40:151, 1990. 14. Guibord, N., Chalk, C., Wein, F., et al.: Trigeminal nerve hypertrophy in chronic inflammatory demyelinating polyradiculoneuropathy. Neurology 51:1459, 1998. 15. Hagen, N. A., Stevens, J. C., and Michet, C. J. Jr.: Trigeminal sensory neuropathy associated with connective tissue diseases. Neurology 40:891, 1990. 16. Harper, N.: Trigeminal trophic syndrome. Postgrad. Med. J. 61:449, 1985. 17. Harris, W.: Bilateral trigeminal tic: its association with heredity and disseminated sclerosis. Ann. Neurosurg. 103:161, 1936. 18. Harris, W.: An analysis of 1433 cases of paroxysmal trigeminal neuralgia (trigeminal-tic) and the end results of Gasserian alcohol injection. Brain 63:209, 1940. 19. Harris, W.: The fifth and seventh cranial nerves in relation to the nervous mechanism of taste sensation: a new approach. Br. Med. J. 1:831, 1952. 20. Henderson, W. R.: Trigeminal neuralgia: the pain and its treatment. Br. Med. J. 1:7, 1967. 21. Hooge, J. P., and Redekop, W. K.: Trigeminal neuralgia in multiple sclerosis. Neurology 45:1294, 1995.

22. Horowitz, S. H.: Isolated facial numbness: clinical significance and relation to trigeminal neuropathy. Ann. Intern. Med. 80:49, 1974. 23. Hughes, B.: Chronic benign trigeminal paresis. Proc. R. Soc. Med. 51:529, 1958. 24. Jacobs, J. M., MacFarlane, R. M., and Cavanagh, J. B.: Vascular leakage in the dorsal root ganglia of the rat, studied with horseradish peroxidase. J. Neurol. Sci. 29:95, 1976. 25. Jannetta, P. J.: Microsurgical approach to the trigeminal nerve for tic douloureux. In Krayenbuhl, H., Maspes, P. E., and Sweet, W. H. (eds.): Progress in Neurological Surgery. Basel, S. Karger, p. 180, 1976. 26. Jannetta, P. J., and Rand, R. W.: Arterial compression of the trigeminal nerve at the pons in patients with trigeminal neuralgia. J. Neurosurg. 26(Suppl.):159, 1967. 27. Katusic, S., Beard, C. M., Bergstralh, E., and Kurland, L. T.: Incidence and clinical features of trigeminal neuralgia, Rochester, Minnesota, 1945–1984. Ann. Neurol. 27:89, 1990. 28. Kerr, F. W. L.: Correlated light and electron microscopic observations on the normal trigeminal ganglion and sensory root in man. J. Neurosurg. 26(Suppl.):132, 1967. 29. Keys, C. L., Sugar, J., and Mafee, M. F.: Familial trigeminal anesthesia. Arch. Ophthalmol. 108:1720, 1990. 30. Khan, O. A.: Gabapentin relieves trigeminal neuralgia in multiple sclerosis patients. Neurology 51:611, 1998. 31. Lecky, B. R. F., Hughes, R. A. C., and Murray, N. M. F.: Trigeminal sensory neuropathy: a study of 22 cases. Brain 110:1463, 1987. 32. Love, S., and Coakham, H. B.: Trigeminal neuralgia: pathology and pathogenesis. Brain 124:2347, 2001. 33. Lunardi, G., Leandri, M., Albano, C., et al.: Clinical effectiveness of lamotrigine and plasma levels in essential and symptomatic trigeminal neuralgia. Neurology 48:1714, 1997. 34. Matharu, M. S., and Goadsby, P. J.: Trigeminal autonomic cephalgias. J. Neurol. Neurosurg. Psychiatry 72(Suppl. 2):19, 2002. 35. McGovern, S. T., Crompton, J. L., and Ingham, P. N.: Trigemino-adbucens synkinesis: an unusual case of aberrant regeneration. Aust. N. Z. J. Ophthalmol. 14:275, 1986. 36. Mitchell, A. B. S., and Parsons-Smith, B. G.: Trichloroethylene neuropathy. Br. Med. J. 1:422, 1969. 37. Morita, A., Fukushima, T., Miyazaki, S., et al.: Tic douloureux caused by primitive trigeminal artery or its variant. J. Neurosurg. 70:P415, 1989. 38. Nurmikko, T. J., and Eldridge, P. R.: Trigeminal neuralgia— pathophysiology, diagnosis and current treatment. Br. J. Anaesth. 87:117, 2001. 39. Nuxton, P. H., and Hayward, M.: Polyneuritis cranialis associated with industrial trichloroethylene poisoning. J. Neurol. Neurosurg. Psychiatry 30:511, 1967. 40. O’Brien, T. J., McKelvie, P. A., and Vrodos, N.: Bilateral trigeminal amyloidoma: an unusual case of trigeminal neuropathy with a review of the literature. Case report. J. Neurosurg. 81:780, 1994. 41. Peet, M. M., and Schneider, R. C.: Trigeminal neuralgia: a review of six hundred and eighty-nine cases with a follow-up study on sixty five per cent of the group. J. Neurosurg. 9:367, 1952.

Diseases of the Fifth Cranial Nerve 42. Pitner, S. E., Edwards, J. E., and McCormick, W. F.: Observations on the pathology of the Moebius syndrome. J. Neurol. Neurosurg. Psychiatry 28:362, 1965. 43. Rorick, M. B., Chandar, K., and Colombi, B. J.: Inflammatory trigeminal sensory neuropathy mimicking trigeminal neurinoma. Neurology 46:1455, 1996. 44. Rosenberg, M. L.: Congenital trigeminal anaesthesia: a review and classification. Brain 107:1073, 1984. 45. Samii, M., Migliori, M. M., Tatagiba, M., and Babu, R.: Surgical treatment of trigeminal schwannomas. J. Neurosurg. 82:711, 1995. 46. Seidel, E., Hansen, C., Urban, P. P., et al.: Idiopathic trigeminal sensory neuropathy with gadolinium enhancement in the cisternal segment. Neurology 54:1191, 2000. 47. Smith, G. W., and Miller, J. M.: Treatment of tic douloureux with stilbamidine. Bull. Johns Hopkins Hosp. 96:146, 1955. 48. Solaro, C., Lunardi, G. L., Capello, E., et al.: An open-label trial of gabapentin treatment of paroxysmal symptoms in multiple sclerosis patients. Neurology 51:609, 1998. 49. Spillane, J. D., and Wells, C. E. C.: Isolated trigeminal neuropathy: a report of 16 cases. Brain 82:391, 1959. 50. Stoll, C., Viville, B., Treisser, A., and Gasser, B.: A family with dominant oculoauriculovertebral spectrum. Am. J. Med. Genet. 78:345, 1998. 51. Su, C. Y., and Lui, C. C.: Perineural invasion of the trigeminal nerve in patients with nasopharyngeal carcinoma: imaging and clinical correlations. Cancer 78:2063, 1996.

1217

52. Sweet, W. H., and Wepsic, J. C.: Controlled thermocoagulation of trigeminal ganglion and rootlets for differential destruction of pain fibers. Part 1: trigeminal neuralgia. J. Neurosurg. 40:143, 1974. 53. Tarlov, I. M.: Structure of the nerve root: I. Nature of the junction between the central and the peripheral nervous system. Arch. Neurol. 37:555, 1937. 54. Warwick, R., Williams, P. L., Dyson, M., and Bannister, L. H.: Gray’s Anatomy, 37th ed. Edinburgh, Longman, 1989. 55. Wiffen, P., McQuay, H., Carroll, D., et al.: Anticonvulsant drugs for acute and chronic pain. Cochrane Database Syst. Rev. 2:CD001133, 2000. 56. Wilson, R. H., Lehky, T., Thomas, R. R., et al.: Acute oxalipalatin-induced peripheral nerve hyperexcitability. J. Clin. Oncol. 20:1767, 2002. 57. Wong, V. A., Cline, R. A., Dubord, P. J., and Rees, M.: Congenital trigeminal anesthesia in two siblings and their long-term follow-up. Am. J. Ophthalmol. 129:96, 2000. 58. Woolfall, P., and Coulthard, A.: Pictorial review. Trigeminal nerve: anatomy and pathology. Br. J. Radiol. 74:458, 2001. 59. Yoshida, K., and Kawase, T.: Trigeminal neurinomas extending into multiple fossae: surgical methods and review of the literature. J. Neurosurg. 91:202, 1999. 60. Zakrewska, J. M.: Facial pain: neurological and nonneurological. J. Neurol. Neurosurg. Psychiatry 72(Suppl. 2):27, 2002.

50 Diseases of the Seventh Cranial Nerve CAROLINE M. KLEIN

Overview Gross Anatomy Vascular Anatomy Neuroimaging Functional Anatomy Peripheral Facial Nerve Paralysis Idiopathic Facial Nerve Paralysis (Bell’s Palsy)

Ramsay Hunt Syndrome Acute Facial Paralysis: Other Causes Inherited/Familial Facial Palsy Congenital/Pediatric Facial Nerve Paralysis Hemifacial Spasm Clinical Features

OVERVIEW The facial nerve, because of its complicated functional anatomy and intracranial course, is susceptible to much pathology. This includes infection, trauma, systemic disease, and vascular compression. The means to identify and characterize facial nerve involvement in these various conditions is evolving with advances in electrophysiologic and neuroimaging techniques. Likewise, use of these techniques provides a way to test hypotheses about the underlying pathophysiology leading to facial nerve dysfunction, such as electrophysiologic testing in hemifacial spasm or neuroimaging of the facial nerve intracranially. Two primary diseases of the facial nerve remain controversial in terms of diagnosis and treatment: acute facial paralysis and hemifacial spasm. The various etiologies of these conditions and the current treatment strategies are presented in this chapter, along with the differential diagnosis for potential underlying etiologies of these diseases of the facial nerve.

GROSS ANATOMY The seventh cranial nerve, or facial nerve, is derived embryologically from the second brachial arch, and it innervates anatomic structures also derived from this embryonic structure. The sensory and motor roots that comprise the facial nerve emerge at the caudal border of the pons and travel

Differential Diagnosis Etiologies Electrodiagnostic Testing Pathology Treatment Geniculate Neuralgia

through the cerebellopontine cistern toward the internal acoustic canal in the temporal bone.32 At the exit from the brainstem, the motor root is separate from the sensory root, also called the nervus intermedius or nerve of Wrisberg, with the motor root lying medial to the sensory root and with the vestibulocochlear nerve traveling most laterally along with the facial nerve as it passes through the cisternal space.32,184,242 The two roots only become fused grossly at the level of the internal acoustic meatus.242 The facial nerve travels in an anterolateral direction, accompanied by the vestibulocochlear nerve to the internal acoustic meatus, and continues into the first portion of the facial canal. The sensory root is believed by some authors to be the most vulnerable to pathologic processes within the cisternal segment.179 The facial nerve is covered by pia mater, because it does not have an epineurial connective tissue layer at this point along its course.179 Dura and arachnoid membranes surround the nerves at this level and invaginate into the internal acoustic meatus, forming a lateral extension of the cerebellopontine cistern, also called the acousticofacial cistern.132 The contents of the internal acoustic meatus are enveloped in the arachnoid and dural membranes and cerebrospinal fluid. The entrance into the facial canal, the porus acusticus or meatal foramen, is located medially and is divided into compartments by bony ridges in the passageways. This point is the narrowest part of the facial canal and may therefore be susceptible to damage from pathologic edema or inflammation of the nerve.106,107,235 1219

1220

Diseases of the Peripheral Nervous System

Once in the facial canal, the nerve is covered by epineurium within the periosteum of the bony canal.107 The facial nerve is located in the ventral aspect of the superior compartment of the foramen.34,132,178 As the facial nerve enters this part of the canal, the sensory root becomes entwined around the motor root and takes a superior position relative to the motor root at this point as it fuses with the motor root to form the completed facial nerve.132 Based on microanatomic dissection of human cadaveric temporal bones, geniculate ganglion cells have been noted along the course of the facial nerve at this point.132 The initial portion of the bony facial canal is termed the labyrinthine segment, and usually extends approximately 3 to 5 mm in length.178,179 This segment is located in close proximity to the superior semicircular canal and follows a course that is lateral and anterior through the temporal bone, creating an angle of 130 degrees within the internal acoustic canal.179 It ends at a widening or fossa in the bony facial canal where the geniculate ganglion is located, which contains the primary afferent (general somatic and special visceral afferent) cell bodies of the facial nerve.32 This fossa is approximately 2 mm in diameter.178 Beyond the geniculate fossa, the facial canal, containing the main trunk of the facial nerve, forms the first external bend or genu at the junction of the labyrinthine and tympanic segments.34,107,179 The tympanic segment is posterior and lateral to the geniculate ganglion, and is approximately 10 mm in length.178,179 The canal passes anteriorly in close proximity to the horizontal semicircular canal. It is here that the facial canal may become continuous with the tympanic cavity if there is bony dehiscence of the wall separating the two cavities.179 Absence of bony covering over the middle two thirds of the facial canal, including the geniculate fossa, may occur as a normal anatomic variant.107,178 This occurs with incomplete fusion during development of the otic capsule and Reichert’s cartilage from the second brachial arch.164 The tympanic segment terminates in the second external genu as it makes a 90- to 125-degree turn and makes a vertical descent through the temporal bone as the mastoid or vertical segment.34,179,235 The mastoid segment extends approximately 12 mm and is located just anterior to the mastoid process and medial to the jugular bulb, ending at the stylomastoid foramen of the skull, where the facial nerve becomes extracranial.107,178,179,235 Extracranially the facial nerve divides into its terminal branches, including the posterior auricular nerve branch, muscle branches to the posterior belly of the digastric and stylohyoid muscles, and temporofacial, zygomatic, and cervicofacial branches, with final termination in a plexus of branches deep to the substance of the parotid gland.107,179,242 Specific branches of the facial nerve along its course are as follows. The nerve to the stapedius has its origin at the posterior wall of the tympanic cavity and travels anteriorly. The chorda tympani arises along the mastoid segment proximal to

the stylomastoid foramen, then it perforates the posterior wall of the tympanic cavity and travels through the cavity, crossing medial to the handle of the malleus bone and eventually joining the lingual nerve along its course. The chorda tympani transmits preganglionic parasympathetic secretomotor fibers and taste afferents from the tongue as described under Functional Anatomy later.242 The motor branches to the posterior belly of the digastric and to the stylohyoid muscles arise from the main facial nerve trunk near the stylomastoid foramen. The posterior auricular nerve arises distal to the origin of the chorda tympani to innervate the skin of the auricle and pinna and provides motor branches to intrinsic auricular muscles and the occipital belly of the occipitofrontalis.242 Eshraghi et al.60 performed cadaveric dissections of the temporal bone to study the posterior auricular branch and its anatomic course and found that it typically branched off of the main facial nerve trunk proximal to the origin of the chorda tympani along the mastoid segment and then coursed superiorly and laterally and then inferiorly along the posterior and inferior walls of the external auditory canal and pinna. A patient who experienced iatrogenic injury to this nerve branch reported reduced cutaneous sensation along the posterior external auditory canal.60 Besides an anastomotic relationship with the vagal nerve,60,242 studies in cats have demonstrated that centrally facial sensory fibers project to the trigeminal nucleus, forming a feedback loop to the facial motor nucleus.105 Topographic organization studies of peripheral facial motor fibers have revealed that there is not a strict somatotopic organization to the fibers along the nerve course. Application of horseradish peroxidase to cut facial nerve branches in animal studies demonstrated labeling of axons diffusely throughout the nerve trunk proximally.38,227 The same technique, applied to the chorda tympani, revealed labeled cell bodies in the geniculate ganglion.227 Intraoperative electrical stimulation of the distal portion of the facial nerve in humans has shown three main fascicles traveling to the orbicularis oculi, the orbicularis oris, and the remaining facial muscles. These fascicles rotate in terms of their relative orientation as the main facial nerve trunk passes through the stylomastoid foramen.119

VASCULAR ANATOMY The vascular supply to the facial nerve within the intracranial cistern is the anterior inferior cerebellar artery, a branch of the basilar artery system.28,242 Once in the facial canal, however, the nerve is supplied by branches of the middle meningeal artery, as well as the superficial petrosal branch and the stylomastoid branch of the posterior auricular or occipital arteries.28,242 This is a point of potential vascular anastomosis between the basilar artery and middle meningeal artery systems.106,132 The petrosal artery supplies branches to the geniculate ganglion and continues within

Diseases of the Seventh Cranial Nerve

the facial canal and anastomoses with the stylomastoid artery. Along the distal facial canal, an arterial arcade is formed at the point of this arterial anastomosis.28 Beyond the stylomastoid foramen, direct branches of the stylomastoid artery provide the nerve with additional blood supply from the superficial temporal and transverse facial arteries developing as the nerve passes into the parotid gland.28,242

is thought to correlate with clinical features by some authors, and not to correlate by others.107 There may be mild enhancement of the nerve in its normal state as a result of the perineurial venous plexus,107 and the mucosa of the tympanic cavity or small intrapetrous fat deposits may also enhance without pathologic implications.178 The nerve may normally enhance in its infratemporal and vertical segments, and such enhancement is more likely to correlate with a pathologic process if the cisternal or proximal segments are found to be abnormal.34 Inflammation from any cause (postsurgical, postradiation) may result in variable degrees of enhancement of the nerve on MRI178 and is therefore relatively nonspecific. MRI may also help to determine the presence of tumors or other mass lesions along the course of the facial nerve, such as facial neuroma, chronic otitis media or secondary cholesteatoma involving the tympanic or mastoid segments, hemangioma or meningioma of the geniculate ganglion, and primary congenital cholesteatoma or facial nerve tumors of the labyrinthine segment, in addition to tumors involving the internal auditory canal.235 Primary facial nerve schwannoma may affect the cisternal segment of the nerve or the geniculate ganglion and may have heterogeneous enhancement, cystic changes, and increased T2-weighted signal intensity (Fig. 50–1).107,178,179 Vascular abnormalities, such as aneurysmal lesions of the anterior or posterior inferior cerebellar arteries or arteriovenous malformation in the cerebellopontine cistern causing focal facial nerve compression, may be discerned with neuroimaging as well.178,179 Hemangioma, a benign extraneural vascular tumor, may be seen as a hyperintense mass on T2-weighted images (Fig. 50–2A), isointense with an

NEUROIMAGING Current neuroimaging techniques allow visualization of the bony facial canal via computerized tomography (CT) of the temporal bone,107 and of the facial nerve itself with magnetic resonance imaging (MRI) with the aid of gadolinium contrast enhancement.107,179 The nervus intermedius typically is too small in diameter to visualize,107 and the facial nerve within the parotid gland may also be difficult normally to discriminate from surrounding tissue.179 Otherwise, the facial nerve may be imaged from the root entry zone at the lateral aspect of the pons, through the internal auditory canal, and to its exit from the skull, especially with special MRI techniques.34 More recently, development of three-dimensional magnetic resonance techniques through post-MRI image processing has led to visualization of the nerve by virtual cisternoscopy.88 Visualization of the nerve throughout its course may provide specific information in certain clinical situations. For example, with idiopathic facial nerve paralysis there may be linear enhancement of the nerve along its course following contrast administration on T1-weighted images, without focal enlargement of the nerve trunk itself, which

A

1221

B FIGURE 50–1 A, Axial T1-weighted brain magnetic resonance image, after gadolinium contrast administration, showing enhancement of the horizontal portion of the left facial nerve caused by schwannoma (arrowhead). B, Axial T1-weighted brain magnetic resonance image showing contrast enhancement of the schwannoma involving the descending portion of the left facial nerve (arrowhead). (Courtesy of Mauricio Castillo, M.D., Department of Radiology, The University of North Carolina, Chapel Hill).

1222

Diseases of the Peripheral Nervous System

A

B

C

FIGURE 50–2 A, Axial T2-weighted brain magnetic resonance image showing a hyperintense mass in the region of the first genu of the left facial nerve (arrowhead), proven to be a hemangioma. B, Axial T1-weighted brain magnetic resonance image demonstrating a hemangioma (arrowhead) of the left facial nerve, which appears slightly hyperintense compared to brain parenchyma—a typical feature of these tumors. C, Axial T1-weighted brain magnetic resonance image, after gadolinium contrast administration, showing homogeneous enhancement of a left facial nerve hemangioma (arrowhead). (Courtesy of Mauricio Castillo, M.D., Department of Radiology, The University of North Carolina, Chapel Hill).

irregular margin on T1-weighted images (Fig. 50–2B), and contrast enhancing (Fig. 50–2C) with MRI and with bony spiculation at the margins on high-resolution CT.178

FUNCTIONAL ANATOMY The functional anatomy of the facial nerve is determined by its components, including the special somatic efferent fibers innervating muscles derived from the second brachial arch, namely the muscles of facial expression and the stapedius, stylohyoid, posterior belly of the digastric, buccinator, and platysma.32,69,242 General visceral efferent fibers include preganglionic parasympathetic secretomotor fibers, which innervate lacrimal and seromucous glands in the nasal cavity and palate via the greater superficial petrosal nerve and sublingual and submandibular glands via the chorda tympani nerve. Preganglionic fibers in the greater superficial petrosal nerve travel to the sphenopalatine and pterygopalatine ganglia to synapse with the postganglionic

parasympathetic neurons, and the preganglionic fibers in the chorda tympani travel with the lingual nerve to the submandibular ganglion to connect with the postganglionic parasympathetic neurons.32,69,106,107,242 Taste from the anterior two thirds of the tongue is transmitted via special somatic afferent nerve fibers through the chorda tympani nerve branch of the facial nerve. Taste sensation from the soft palate and nasopharyngeal mucosa is also carried via special somatic afferent nerve fibers through the greater superficial petrosal nerve with cell bodies in the meatal ganglion.69,106,242 General somatic afferent fibers provide cutaneous sensation from the external auditory canal and concha and course through the posterior auricular branch of the facial nerve,32,69,106,242 with interconnections with the vagus, trigeminal, and upper cervical cutaneous nerve fibers to the same region.60,242 The reader is referred to an available review of the anatomy of the sensory components of the facial nerve by Boudreau et al.29 (Fig. 50–3) and the functional anatomy as reviewed by Sweeney and Gilden221 (Fig. 50–4).

Diseases of the Seventh Cranial Nerve

1223

FIGURE 50–3 Summary of sensory systems of the geniculate ganglion and their peripheral distribution. All sensory fibers indicated originate from sensory cells in the geniculate ganglion. (From Boudreau, J. C., Oravec, J., White, T. D., et al.: Geniculate neuralgia and facial nerve sensory systems. Arch. Otolaryngol. 103:473, 1977, with permission.)

PERIPHERAL FACIAL NERVE PARALYSIS Idiopathic Facial Nerve Paralysis (Bell’s Palsy) Peripheral facial nerve paresis or paralysis is the most common manifestation of a seventh cranial nerve lesion because of its innervation of the muscles of facial expression. Although there are many etiologies for facial nerve damage (see later), idiopathic facial paralysis is the most common.23,50,94,124,201 Hauser et al.87 examined the incidence of Bell’s palsy in the population of Rochester, Minnesota, from 1955 to 1967 and reported it to be 22.8 per 100,000 per year. These authors did not find any evidence of clustering of cases in this population or seasonal variation in terms of its occurrence.87 However, other

authors have noted a seasonal variation in regard to incidence.50 The incidence of Bell’s palsy in this Minnesota population remained stable, as follow-up studies showed the incidence to be 25.2 per 100,000 from 1968 to 1982.116 Similar incidences have been found in other populations, for example, an incidence of 24.1 per 100,000 in Madrid, Spain,46 and 15.5 per 100,000 in a recent study from Toronto, Canada.90 There does not appear to be a gender predominance,46,50,177,181 although a few studies have noted an increased incidence in female patients between ages 10 and 19.5,87,181 In regard to age distribution, the mean age of patients is 40,46,50,116 with the highest age-specific rate occurring in patients over age 70 in two studies.5,116 Another study, done retrospectively in the Netherlands, found that the peak age for incidence in men was between ages 30 and 40, whereas female patients showed a bimodal

1224

Diseases of the Peripheral Nervous System

Geniculate ganglion Internal acoustic meatus Carotid plexus (on internal carotid artery)

Nervus intermedius Motor nucleus of VII Solitary tract

Pterygopalatine Greater petrosal ganglion nerve

Superior salivatory nucleus

Motor VII

Spinal nucleus of V Nerve to stapedius Stylomastoid foramen

Chorda tympani

Tympanic membrane Special visceral efferent Special visceral efferent Sympathetic Special visceral efferent Parasympathetic General somatic efferent FIGURE 50–4 Functional anatomy of the facial nerve. Proximally, the four cranial nerve nuclei involved in facial nerve functions are shown in the pontomedullary junction: the motor nucleus of VII, the nucleus of the solitary tract, the superior salivary nucleus, and the spinal nucleus of V. Special visceral efferent motor fibers from the motor nucleus of cranial nerve VII (solid red line) exit the brainstem and travel through the internal acoustic meatus to enter the bony facial canal and exit through the stylomastoid foramen to supply facial muscles. The solitary tract receives special visceral afferent taste fibers (solid blue line) emanating from the anterior two thirds of the tongue. These fibers travel with the chorda tympani through the petrotympanic fissure (not shown). The cell bodies of these special visceral afferent fibers are in the geniculate ganglion. Special visceral efferent parasympathetic fibers (thin dotted red line) to the lacrimal and salivary glands emanate from the superior solitary nucleus, travel in the nervus intermedius, and branch at the geniculate ganglion into the greater petrosal and chorda tympani nerves. Special visceral efferent sympathetic fibers (thick dotted red line) emanate from the carotid plexus on the internal carotid artery and join the greater petrosal nerve as these structures pass through the foramen lacerum (not shown). The sympathetic fibers parallel the parasympathetic fibers because they supply the same areas. The spinal nucleus of cranial nerve V receives general somatic afferent fibers from the geniculate zone of the ear via the chorda tympani. Cell bodies of these neurons are located in the geniculate ganglion. (Modified from Sweeney, C. J., and Gilden, D. H.: Ramsay Hunt syndrome. J. Neurol. Neurosurg. Psychiatry 71:149, 2001).

age distribution in terms of peak incidence, between ages 20 and 29 and also between ages 50 and 59.50 Neither side of the face is more typically involved.5,50,116,177 Clinical Features Bell’s palsy is typically of acute onset, with occasional associated symptoms of abnormal ipsilateral tearing or taste, and ear pain or discomfort.87,147,177 The ear pain is usually poorly localized but usually retroauricular7,192 and present

only at the onset or just prior to the onset of the facial weakness.87,147,177,192 Patients have reported other sensory symptoms, such as ipsilateral facial tingling or numbness, but usually without an objective deficit on neurologic examination.87,147 The natural history of Bell’s palsy was examined by Peitersen177 in a group of 1011 patients in Denmark who were diagnosed but did not receive specific medical treatment. The majority (69%) of these patients had complete

Diseases of the Seventh Cranial Nerve

unilateral facial paralysis at the initial examination. All of these patients had some recovery, 54% within the first month and others recovering over 3 to 6 months, with 71% returning to normal facial motor function.177 Knox124 stated that 85% of patients with Bell’s palsy regain some facial movement within the first 3 weeks, and the remaining 15% usually improve within the first 3 to 6 months. The timing of the recovery seemed to have a correlation with the likelihood of sequelae such as synkinesis, facial muscle contracture, or gustatory tearing in that these phenomena were less likely to occur the earlier in the course of the illness that the recovery began. Patients who had delayed onset of recovery, after 3 months since onset of weakness, were the most likely to experience sequelae.177 Overall, approximately 10% to 15% of patients with Bell’s palsy will have residual ipsilateral facial muscle weakness and experience sequelae of synkinesis, muscle contracture or spasm, and abnormal tearing.124 Adour and Wingerd7 reported that reduced ipsilateral tearing and hyperacusis were more likely associated with more complete facial paralysis and clinical sequelae with recovery. Prognostic clinical factors in regard to recovery of function include the timing from the onset of weakness to the initial recovery, and the age of the patient (the younger the patient, the better the chance for recovery of function).7,26,50,87,116,120,177 Other factors related to development of sequelae include abnormal stapedial reflex, abnormal tearing,5,50,116 and presence of posterior auricular pain.177 Incomplete paralysis and early signs of return of facial movement are the best clinical indicators for good recovery, whereas earlier progression to complete paralysis usually suggests a poorer outcome.5,26,87,116 Regression analysis of clinical features of 206 patients with Bell’s palsy studied retrospectively showed that complete facial palsy, cranial pain other than auricular, and coexistent hypertension were risk factors for incomplete recovery of facial motor function.116 Of the patients in this study without these risk factors, 96% had complete recovery.116 Persistent facial nerve palsy beyond 1 year since onset usually necessitates an evaluation to examine for other underlying causes.147 Also, if ipsilateral hearing loss or acoustic reflex decay is determined in the setting of acute facial paralysis, additional testing to exclude retrocochlear pathology is usually needed.124 Knox124 recommended neuroimaging if otorrhea, vestibular symptoms, ipsilateral hearing loss, or facial paralysis beyond 6 months are present. Jackson and von Doersten106 suggested that, if the facial paralysis is slowly progressive over more than 1 month, there is no return of function within 6 months of onset, there is clinical involvement of other cranial nerves, and cranial pain is persistent beyond onset, then evaluation for neoplasm should be done. A small percentage of patients with a history of Bell’s palsy experience recurrent symptoms or have a history of prior idiopathic facial paralysis.5,50,87,99,116,147,177,234 This may

1225

occur with idiopathic facial nerve paralysis, but may also indicate an underlying tumor, especially if the recurrence is always on the same side of the face.106,147 Adour and Wingerd5 noted that 39% of patients with recurrent Bell’s palsy in their series of 1048 patients with Bell’s palsy from 1966 to 1976 had diabetes mellitus. A later combined retrospective and prospective study from the same authors specifically looking at recurrence rates showed that overall there was a 7.1% recurrence rate for Bell’s palsy.181 The frequency of ipsilateral versus contralateral recurrence was the same in this group of 140 patients identified from 1969 to 1986. The mean time interval to the first recurrence was 9.8 years after the initial episode. Mean age and gender predominance for these patients was not different than that described for patients with a single episode of Bell’s palsy. Fifteen percent of these patients had a second recurrence and a minority experienced up to four recurrences after the initial palsy. They concluded that patients with recurrent Bell’s palsy were 2.5 times more likely to have a positive family history of Bell’s palsy than those patients who did not have recurrent attacks.181 Other authors have noted that history of sarcoidosis, diabetes mellitus, leukemia, infectious mononucleosis, and neoplasm increase the chances of recurrent facial palsy.23 Bilateral seventh cranial nerve palsy occurs less frequently than unilateral involvement. The reader is referred to Keane’s review article118 for a history of bilateral facial nerve palsy and analysis of 43 patients identified over a 23-year period. In 22 of 43 patients, the cause of bilateral palsies was determined to be “benign,” with various causes including idiopathic facial paralysis, Guillain-Barré syndrome, Miller Fisher syndrome, brainstem encephalitis, benign increased intracranial hypertension, and multiple cranial neuritides. Nine of the 43 patients were found to have tumor and 5 cases were infectious (syphilis, leprosy, cryptococcal meningitis, tuberculous meningitis) in origin. The remaining patients had “other” potential etiologies including diabetes mellitus, sarcoidosis, history of head trauma, pontine hemorrhage, systemic lupus erythematosus, bulbospinal neuronopathy, and history of Möbius’ syndrome.118 A comprehensive review of the differential diagnoses in these cases is provided by Rontal and Sigal.191 Simultaneous bilateral facial palsies (occurring within 30 days from initial onset) may be seen with Möbius’ syndrome, bilateral temporal bone fractures, Bell’s palsy, Guillain-Barré syndrome, acute leukemia, bilateral otitis media, sarcoidosis, Lyme disease (Fig. 50–5), bacterial meningitis, syphilis, leprosy and Epstein-Barr viral infection.148,183 Bilateral Bell’s palsy is reported to have a good prognosis, although typically there is a delay in recovery between the affected sides.183 Recurrent bilateral facial palsy is rare. Stahl and Ferit217 reported a case of a 16-year-old girl with recurrent bilateral facial palsies that occurred with a 4-year interval in association with other cranial nerve symptoms suggestive of an acute cranial

1226

Diseases of the Peripheral Nervous System

FIGURE 50–5 Axial T1-weighted brain magnetic resonance image, after gadolinium contrast administration, showing diffuse contrast enhancement of the facial nerves bilaterally (arrowheads) in the deep intracanalicular portion of the nerves and the geniculate ganglia in a patient with Lyme disease. (Courtesy of Mauricio Castillo, M.D., Department of Radiology, The University of North Carolina, Chapel Hill).

polyneuritis, but with normal cerebrospinal fluid and no evidence of acute viral infection, who was treated successfully with oral steroids. Electrophysiologic Testing and Neuroimaging A variety of electrophysiologic tests have been used to assess degrees of nerve degeneration in patients with facial paralysis in order to provide prognostic information and to investigate the electrodiagnostic features of sequelae of paralysis, such as synkinesis. Because the maximal nerve injury usually occurs within the first 14 days of the onset of acute facial paralysis,98 it is within this time frame that initial electrodiagnostic testing is appropriate. The reader is referred to several review articles for details regarding the history and relative value of some of these testing procedures.98 The nerve excitability test determines the excitation threshold by recording the minimum electrical stimulus of the facial nerve required to produce visible muscle contraction on the affected and unaffected sides of the face, with greater than 3.5-mA difference between the sides considered to be significant.98,120,205 Kerbavaz et al.120 retrospectively reviewed 839 patients with Bell’s palsy and found, using multivariate analysis, that complete facial paralysis and a nerve excitability test difference greater than 3.5 mA had a negative association in terms of good outcome and a greater likelihood of development of sequelae such as muscle contracture and synkinesis. The maximal stimulation test is considered more reliable because it determines the electrical stimulus intensity required to elicit a maximal motor response on the unaffected side and, applying it to the affected side, assesses

the ipsilateral facial muscle contraction in response to this level of stimulation.98 Electroneurography compares the facial compound muscle action potential amplitude on the affected side as a percentage of the unaffected side.38,98 This value is thought to represent the percentage of intact motor neurons in the facial motor nucleus on the affected side38 or the approximate axonal degeneration or conduction block.41 Various electroneurographic techniques have been compared in terms of their reliability and validity by Coker.38 Coker also reported animal studies in which electroneurography was used to assess facial nerve degeneration after partial facial nerve transection and correlated these findings with quantitative assessment of surviving motor neurons in these same animals following application of horseradish peroxidase to the injured nerve trunk. In two of three animals studied, the amount of neuronal loss appeared to correlate with the electroneurographic results.38 Standard electromyography may also be used to assess denervation and reinnervation, although fibrillation potentials may not be present on needle examination for up to 21 days after onset of symptoms,90 making it less useful as an early prognostic test.41 However, there are cases of patients with clinically complete facial nerve paralysis who were able to generate voluntary motor unit potentials on needle examination of ipsilateral facial muscles, indicating early resolution of conduction block or subclinical evidence of early regeneration and therefore a better prognosis for recovery compared to patients who were not able to generate motor unit potentials.41,71 Rarely, needle examination of facial muscles can be abnormal in the setting of normal nerve conduction studies.90 Single-fiber electromyography has been correlated with electroneurography results in patients with peripheral nerve palsy.223 Takeda et al.223 reported that an electroneurographic result of 40% or more was associated with normal muscle fiber density, suggesting absent or only mild wallerian degeneration. Conversely, if the electroneurography result was less than 40%, there was increased muscle fiber density in affected facial muscles, suggesting that there was an increase in the number of muscle fibers per motor unit secondary to collateral nerve sprouting resulting from axonal degeneration or the presence of wallerian degeneration.223 In studying 36 patients with facial nerve paralysis using serial electroneurography and single-fiber electromyography within the first 2 weeks after the onset of facial paralysis, the minimum electroneurographic result occurred between 7 and 10 days after onset, indicating a time point when this testing may be most useful to assess axonal degeneration.223 Trigeminal blink reflex testing, comparison of facial compound muscle action potential latencies with nerve conduction testing, stapedial reflex testing, salivary flow studies, Schirmer’s test for tear production, and electrogustometry have also been used in the setting of acute

Diseases of the Seventh Cranial Nerve

facial paralysis.41,98 Some authors believe these tests are technically difficult and thus less reliable for prognostic assessment.41 Hill et al.90 found that blink reflex testing was most useful, with a diagnostic sensitivity of 81% and specificity of 94%. Cramer and Kartush41 suggested that blink reflex testing can measure conduction block but cannot determine axonal degeneration. In regard to prognostic testing, various tests may be useful. The visual assessment of muscle twitch used in the nerve excitability test is subjective and therefore less reliable.41 However, Devriese et al.50 used this testing to study 1176 patients with Bell’s palsy and reported that 17% of patients had diminished response on the affected side and 13% had an absent response. If the nerve excitability test result was normal, 80% of patients recovered with normal facial function and 20% had slight paresis. If the nerve excitability test showed a diminished ipsilateral response, 67% had paresis and 25% had complete recovery. With absent responses on the affected side, 50% of patients recovered with mild paresis and the remaining 50% had moderate to severe weakness.50 Alternatively, the maximum stimulation test is grossly useful in that, if there is no response on the affected side to application of an electrical stimulus used to elicit a maximal response on the unaffected side, then recovery will be incomplete. If the response is symmetrical or equal to the unaffected side, then there is a strong likelihood for recovery of greater than 90% of facial function.41 One of the better or more accurate tests for prognosis may be electroneurography, in that the rate of reduction of the compound muscle action potential amplitude on the affected side correlates with prognosis (i.e., the slower the rate of reduction, the better the prognosis for recovery).41 Sinha et al.212 performed a retrospective study on 23 patients who had electrodiagnostic testing at the University of Cincinnati within 14 days of onset of facial paralysis. By means of patient questionnaire for follow-up assessment of functional recovery after 6 months, of 15 patients with greater than 90% drop in facial compound muscle action potential amplitude on the affected versus the unaffected side, 7 reported near-normal to normal recovery, 5 reported moderate dysfunction, and 3 reported severe dysfunction on the ipsilateral face.212 All but one patient had some form of recovery of facial function by 6 months. These findings suggest that, at least with nerve conduction studies, the patient’s reported recovery of function may be better than predicted by the testing.212 Tojima et al.228 studied 551 patients with serial electroneurography who had Bell’s palsy and found that the electroneurographic result usually fell to a minimum up to 7 days after onset of symptoms, but then became stable. The sharpest decline occurred in patients who had the worst prognosis. The value of this percentage corresponded to the timing of functional recovery, in that patients who had a value greater than 40% recovered completely within

1227

1 month, those with a 20% to 40% value recovered within 2 months, and those with a 10% to 20% electroneurographic value recovered in 3 months. Patients with electroneurographic percentages of 1% to 10% on the affected side had a 50% chance for complete recovery, and those with 0%, no chance for functional recovery.228 Qiu et al.187 found that, in patients with Bell’s palsy, clinical weakness grade was typically more severe than electroneurographic testing suggested early in the clinical course, but by 10 to 14 days after onset of weakness these measures were more likely to correlate, and likewise, clinical recovery of muscle strength preceded improvement in electroneurographic results. The lower the minimal electroneurographic percentage on the affected side, the slower and more incomplete the recovery.187 The prognostic value of electromyography has been studied retrospectively in German patients with idiopathic facial nerve palsy.213 In this study, an initial needle examination was done in 271 patients within the first 7 days of symptom onset to assess for complete (no voluntary motor unit potentials activated) or incomplete (any voluntary motor unit potentials activated) paralysis electrodiagnostically. A subset of 240 patients underwent repeat needle examination more than 14 days after onset of paralysis. Of the 118 patients with residual impairment, 92 were classified as having neurapraxia (no fibrillation potentials) and 26 patients had axonotmesis (fibrillation potentials). The authors concluded that the presence of fibrillation potentials was predictive of a poor functional outcome.213 Following recovery of facial nerve function after acute facial paralysis, electrophysiologic testing may be utilized to detect various postparalysis sequelae. Kimura et al.122 used blink reflex testing to document synkinesis in 29 patients following facial nerve paralysis from a variety of causes, though predominantly Bell’s palsy. They used simultaneous recording from the orbicularis oculi and orbicularis oris or platysma with electrical stimulation of the supraorbital nerve. Synkinesis occurs with electrical response in the orbicularis oris or platysma with ipsilateral blink testing. Postparalysis synkinesis is thought to be due to aberrant regeneration in the facial nerve following a degenerative lesion.122 Another study demonstrated synkinesis in patients years after unilateral Bell’s palsy, as well as myokymic discharges at rest, ipsilateral to the paralysis in most but occurring in bilateral orbicularis oris muscles in a subset of patients studied.236 Bilateral abnormalities suggest an additional neuronal component to the hyperexcitable muscle activity noted.236 Maeyama et al.136 correlated findings of synkinesis on blink reflex testing after recovery with severity of findings on electromyography performed within the first 2 weeks. In that study, the amount of synkinesis was inversely proportional to the percentage loss of motor function determined by electromyography such that, if greater than 40% function was present initially, synkinesis did not occur. With recovery of

1228

Diseases of the Peripheral Nervous System

facial function, the ipsilateral R1 latency on blink reflex testing became less prolonged and the amount of initial prolongation of this response initially correlated with greater loss of facial motor function.136 Use of neuroimaging in the setting of acute facial nerve paralysis has been studied (Fig. 50–6). Kohsyu et al.125 reported that, in patients with Bell’s palsy, MRI with gadolinium contrast revealed enhancement of the internal acoustic meatal segment of the facial nerve only on the affected side. More distal segments of the facial nerve also enhanced, but not exclusively on the affected side. They concluded that enhancement of the meatal segment was

FIGURE 50–6 A, A 37-year-old man with acute onset of facial nerve palsy. Axial gadolinium-enhanced T1-weighted magnetic resonance image with fat suppression shows enhancement of geniculate ganglion (arrow) and perigeniculate segment of cranial nerve VII. B, Coronal magnetic resonance image confirms abnormal enhancement of geniculate ganglion (arrow). This was Bell’s palsy. (From Phillips, C. D., and Bubash, L. A.: The facial nerve: anatomy and common pathology. Semin. Ultrasound CT MR 23:202, 2002, with permission.)

the most characteristic finding in patients with Bell’s palsy.125 Another study examining MRI in facial nerve paralysis showed that ipsilateral facial nerve enhancement with contrast occurred in patients with moderate to severe muscle weakness, with the greatest facial nerve impairment associated with mastoid segment enhancement with contrast in particular. However, one patient had normal facial function by the time the MRI was performed, but diffuse enhancement of the ipsilateral facial nerve.162 A study correlating serial MRI examinations in patients with acute facial nerve palsy with clinical and electroneurographic findings found that the intensity, extent, and duration of facial nerve enhancement did not correlate with functional outcome or with severity of axonal degeneration.30 Some authors contend that any facial nerve enhancement on MRI in the setting of acute paralysis is nonspecific,34,192,205 and its clinical significance remains unclear. Some authors recommend use of MRI to look for other possible etiologies for acute facial paralysis, not for prognostic information, especially if there is little or no recovery of function.205 Pathology Herpes Infection. A body of literature has been directed toward elucidation of the potential role of reactivation of herpes simplex viral infection as a cause for acute facial paralysis. Adour et al.2 tested serial serum antibody titers for herpes simplex virus (HSV) in 41 patients with Bell’s palsy and compared these results with matched control patients. The prevalence of positive antibody titers in the patients was 100%, with 85% of the control patients also being positive. Viral antibody titers in the patient group remained persistently present up to 1 year later, without an increase in convalescent titers. The authors concluded that these findings suggest against a primary herpetic infection in Bell’s palsy patients but instead reactivation of a latent infection (presumably within the geniculate ganglion) in patients. They also suggest that Bell’s palsy in patients with reactivated HSV type 1 infection is only a component of a cranial polyneuritis, noting involvement of the trigeminal, glossopharyngeal, and vestibulocochlear nerves in a subset of Bell’s palsy patients.2,5 A direct causeand-effect relationship is difficult to prove, especially since evidence of reactivation of latent infection is indirect.201 A Japanese study comparing analysis of saliva samples from patients with Bell’s palsy, Ramsey Hunt syndrome, and healthy HSV-seropositive controls showed that, within 7 days of onset of facial weakness, shed herpesvirus in saliva was more prevalent in Bell’s palsy patients (40%) compared to those with Ramsey Hunt syndrome (7%), which is due presumably to reactivation of herpes zoster virus.68 Presumably the shed virus detected in the saliva from the Bell’s palsy patients originated from the chorda tympani. Serum antibodies were found in both groups of

Diseases of the Seventh Cranial Nerve

patients, although more often in those with Bell’s palsy, but a significant change in anti-HSV immunoglobulin (Ig) G or IgM titers was found in only 3 of 47 patients, indicating that reactivation of the viral infection is not necessarily accompanied by a change in serum antibody titer.68 Polymerase chain reaction studies of human geniculate ganglia obtained from autopsy cases in Japan revealed that 15 of 17, or 88%, were positive for latent HSV type 1, without a history of facial nerve paralysis or symptoms of active herpes simplex infection at the time of death. Latencyassociated transcripts, or latency-specific RNA sequences, were also identified in geniculate ganglia of these patients by reverse transcriptase–polymerase chain reaction.222 Inoculation of HSV into the tongue or auricle of mice has provided an animal model of acute and transient facial nerve paralysis, with approximately 56% of animals developing clinical symptoms between 6 and 9 days later.220 The facial weakness resolved spontaneously, with a correlation between the severity and duration of weakness and the timing of the onset after inoculation. Histopathologic analysis of the affected facial nerves showed severe edema, inflammatory cell infiltration, and vacuolar degeneration involving the ipsilateral facial motor nucleus, geniculate ganglion, and facial nerve. These changes were not found in animals who were inoculated but did not develop facial weakness or on the contralateral side of affected animals. HSV antigen was detected in these same tissues only in a subset of affected mice.220 This study demonstrates how HSV may cause acute facial paralysis but does not confirm that latent HSV activation occurs in human patients with Bell’s palsy. The reader is referred to Schirm and Mulkens201 for review of additional animal inoculation studies attempting to address this relationship between herpesvirus and facial palsy. Murakami et al.161 examined endoneurial fluid from human facial nerve and in posterior auricular muscle tissue by polymerase chain reaction for multiple viruses, including HSV. These specimens were obtained during decompressive surgery performed between 12 and 87 days after onset of facial nerve paralysis in patients with Bell’s palsy and Ramsey Hunt syndrome as well as control patients without facial weakness. Ten of 13 fluid specimens and 8 of 14 muscle specimens were positive for HSV type 1 in Bell’s palsy patients, suggesting reactivation of herpes viral infection in these cases. Interestingly, a higher percentage of Bell’s palsy patients had positive HSV serum antibodies (92%) compared to Ramsey Hunt syndrome patients (44%) and controls (56%).161 Other investigators have found a higher prevalence of HSV serum antibodies in patients with Bell’s palsy compared to controls.234 Viral culture of epineurial biopsy tissue obtained from two patients with Bell’s palsy during surgical decompression of the facial nerve was positive for HSV type 1 in one patient, who did not have any detectable change in serum acute or convalescent HSV antibody titers.160

1229

Marra141 provided a recent review of the evidence supportive of HSV type 1 infection being responsible for at least a subset of patients diagnosed with Bell’s palsy based primarily on an increased frequency of findings of HSV type 1 infection in Bell’s palsy patients compared to controls, yet conceding that reactivation of latent viral infection may occur even in asymptomatic individuals. It is postulated that symptoms of acute facial paralysis may only occur when the local immune response to viral replication is inadequate.141 Borrelia burgdorferi Infection (Lyme Disease). Neuroborreliosis as an occult cause of acute facial paralysis has been studied extensively. Either unilateral or bilateral facial nerve palsy is the most common cranial neuropathy associated with Lyme disease.108 Halperin et al.82 examined cases of Bell’s palsy with evidence of Lyme disease in an endemic area and found that testing for antibodies to Borrelia burgdorferi in all patients diagnosed with Bell’s palsy during a 3-month period showed that 7 of 32 had positive serology at the time of diagnosis, with an additional patient with seroconversion 6 weeks later. Cerebrospinal fluid examinations from five seropositive patients were negative for intrathecal antibodies. The authors estimate that, in an endemic area, Lyme disease may be associated with Bell’s palsy in 10% to 25% of patients.82 A retrospective study of nearly 1000 patients with Lyme disease showed that approximately 10% had acute facial paralysis, which may be the initial symptom of Lyme disease82 and rarely may precede the development of erythema chronica migrans.36 In these patients there may be an increased risk for bilateral facial paralysis.16,36,129,175 Many patients have associated symptoms of rash, fever, headache, and malaise, and may also manifest other neurologic involvement such as meningoencephalitis and/or radiculoneuritis.16,36,175 Alternatively, the facial weakness may be the only neurologic manifestation of initial Lyme disease.36,108,175 Kuiper et al.129 examined Lyme serology in patients with Lyme-associated facial nerve palsy and serum from controls and from patients with Bell’s palsy who did not have other symptoms of Lyme disease and found a low percentage of positive antibodies in the latter two groups. The authors concluded that these percentages were within the expected false-positive rate for a test with 95% specificity.129 Whereas some authors report a better outcome and earlier recovery in patients with Lyme-associated facial nerve palsy compared to patients with idiopathic facial paralysis,16,36,81,175 others believe that the outcome is typically worse.99 In regard to treatment, oral antibiotics are recommended unless there is abnormal cerebrospinal fluid (pleocytosis, elevated protein, increased IgG index),81,159 in which case intravenous antibiotics are indicated.81 Other authors suggest addition of oral prednisone to the medication regimen if the patient is evaluated within the first

1230

Diseases of the Peripheral Nervous System

24 hours after onset of facial weakness,175 although some investigators have not seen a benefit to adding steroids in terms of shortening the time to recovery of facial muscle function.36 Angerer et al.16 suggested early electrodiagnostic testing if clinically complete facial palsy develops or if there is incomplete recovery. Testing for Lyme disease is recommended in endemic areas, in children, and in patients with history of tick bite, rash, or other systemic or possible neurologic involvement, including bilateral facial nerve palsy.129,159 There have been rare cases of patients with acute facial paralysis, negative Lyme serology, and elevated cerebrospinal fluid antibody titers to B. burgdorferi,99 which implies that the clinician’s index of suspicion for this association in the setting of acute facial paralysis is critical. Neuropathology Limited neuropathologic studies have been done in patients with Bell’s palsy. Matsumoto et al.144 examined facial nerve fascicular biopsies obtained during surgical decompression in four patients with complete unilateral facial paralysis. These biopsies were done from 10 to 50 days after onset of facial weakness, with three of four biopsies done at 25 days or longer. The authors reported gross hyperemia and edema of the nerve at the time of surgery per visual nerve inspection. Histologically, the nerves showed wallerian degeneration with secondary demyelination, with relative preservation of unmyelinated nerve fibers. Only one case demonstrated lymphocytic infiltrates.144 An autopsy case from a patient who died 7 days after onset of Bell’s palsy (complete facial paralysis for 4 days ante mortem) reportedly showed thickened perineurium, small round inflammatory cell infiltrates in the endoneurium and perivascularly, and demyelinating more than axonal injury.133 There were also inflammatory changes noted in the geniculate ganglion, as the pathology appeared to extend from the internal auditory canal to the stylomastoid foramen. These findings were thought to be consistent with a viral neuritis, although the exact etiology was not determined.133 Harada et al.84 examined facial nerve biopsies from two patients with Bell’s palsy obtained at the time of surgical nerve decompression. In one case, no nerve fibers or Schwann cells were identified in the tissue, only collagen fibers and lipid droplets. In the other case, the authors stated that the myelin sheaths appeared “loosened.”84 Both of these patients had clinically complete unilateral facial paralysis. Treatment Adour and co-workers8 published one of the earliest studies examining treatment for Bell’s palsy in 1972. This was a retrospective, open-label trial of 194 patients treated with oral prednisone (40 mg daily for 4 days followed by a rapid taper) and 110 patients who did not receive prednisone. These authors reported that, with early treatment with

prednisone (started less than 7 days from onset of facial weakness), the functional outcome was improved compared to no treatment, including a subset of diabetic patients. The mechanism of action was postulated to be reduction of edema of the facial nerve within its bony canal, thus limiting damage from vascular compression. Incidental note was made of a marked reduction of posterior auricular pain in those patients receiving prednisone.8 Follow-up studies included a randomized, double-blind, placebo-controlled trial done at the University of California at Los Angeles 20 years later comparing placebo versus prednisone in a small number of patients with Bell’s palsy.21 Thirty-five patients received oral prednisone and 41 received placebo. All patients had paralysis for less than or equal to 5 days before entering the trial. There was a significant difference in functional grade at recovery, with a higher proportion of placebo patients with poor recovery, but no difference in the mean time to minimal recovery between the two groups.21 Although not specifically tested for in this study, the authors stated that patients with mild to moderate weakness at the onset did not receive significant benefit from steroid therapy; however, patients with high risk for denervation received a treatment benefit.21 Saito et al.195 found retrospectively that, in 288 patients with complete Bell’s palsy, the cure rate for patients receiving high-dose steroids was 97.2% compared to a 58.3% cure rate for those who had not been treated with steroids. As greater concern developed regarding the possible contribution of reactivation of HSV infection in the development of acute facial paralysis, additional studies examined the benefits of adding an antiviral medication to the regimen. De Diego et al.45 performed a randomized, prospective trial (no placebo group) comparing oral prednisone and acyclovir in 101 patients diagnosed with Bell’s palsy. The patients were evaluated for the study within 96 hours of the onset of their facial weakness. At 3 months’ follow-up, a higher percentage of the patients who received prednisone had recovered compared to those receiving acyclovir, based on facial function grading, maximal stimulation testing, and presence of sequelae.45 Another trial compared treatment with a combination of acyclovir and prednisone versus prednisone combined with placebo in patients with Bell’s palsy and found that the addition of acyclovir to prednisone was beneficial in terms of functional outcome and prevention of nerve degeneration by the maximal stimulation test.6 There was no difference, however, in the outcome at 4 months in patients who had progressed to complete paralysis by 2 weeks after onset.6 Given the natural history of prolonged time to recovery in patients with more severe paralysis, the time frame for follow-up may have been too short. As questions about the utility of medical therapy were raised, meta-analysis studies were done to review the available literature. Williamson and Whelan243 reviewed published trials from 1970 to 1995 and found that there

1231

Diseases of the Seventh Cranial Nerve

were insufficient numbers of patients in adequately controlled, prospective trials, with insufficient power to answer these questions. It was clear that patients with incomplete Bell’s palsy had a good prognosis, with initial recovery usually within the first 3 weeks; however, those patients who progress to complete paralysis over the first 2 weeks or patients without initial recovery within the first month may have already passed the point when medical intervention might be of benefit if treatment is withheld until their status is determined. The authors conceded that there was no conclusive evidence that steroid therapy is effective in terms of recovery of facial motor function; however, because of the trend for improvement in combined randomized controlled trials, they recommended a course of steroids in those patients with clinically complete paralysis, starting therapy as early as possible, preferably within the first 24 hours.243 A later meta-analysis that was limited to prospective studies with controls, treatment initiated within 7 days of onset of facial palsy, total prednisone dose of at least 400 mg, and follow-up of at least 4 months or to complete recovery found only two studies that could be included.188 However, the authors concluded that patients with complete facial paralysis treated with steroids had 17% more complete recovery, on average, than patients treated with placebo or no treatment, which was statistically significant. They recommended treating all patients initially with prednisone and antiviral medications for 5 days and, if stable (no worsening weakness), tapering the medication quickly or, if progressive at 5 days, instituting a longer taper.188 Other authors have suggested similar approaches,1 recommending that, if facial paralysis is incomplete after 5 days on the medications, the prednisone be tapered over 5 days and the acyclovir be discontinued.1 Patients treated earlier in their course appeared to have better outcomes. A recent practice parameter also examined the question of treatment for patients with Bell’s palsy and also found the available studies to have insufficient power to answer the question as to whether steroids are helpful, although the trend was toward a steroid benefit.77 Based on available studies, the practice parameter concluded that steroids are probably effective in improving functional facial nerve outcome in patients with Bell’s palsy and that the combination with acyclovir is possibly effective.77 Many authors5,50,116,120,124 have supported the use of prednisone with or without acyclovir in severe acute idiopathic facial paralysis. Gantz et al.71 reported the results of a multicenter prospective trial of surgical decompression of the facial nerve in patients with Bell’s palsy who demonstrated greater than 90% nerve degeneration on electroneurography and absence of voluntary motor unit potentials on needle examination of affected facial muscles within 14 days of onset of total facial paralysis.71 Via intraoperative recording they determined that there was conduction block medial

to the geniculate ganglion in 17 of 19 patients studied. Looking at patients who met the inclusion criteria but elected not to undergo surgery, there was a 58% chance of poor outcome at 7 months compared to no or only mild facial weakness in 91% of the patients who had the surgical procedure.71 Interestingly, a subset of patients who had the decompressive procedure done more than 14 days after onset of complete paralysis did not receive the same benefit. The practice parameter paper, published 2 years after Gantz’s study, commented that there was insufficient evidence regarding the role of facial nerve surgical decompression in patients with Bell’s palsy.77 Other investigators have either supported1,41 or questioned5 the role of surgical decompression for Bell’s palsy.

Ramsay Hunt Syndrome Ramsay Hunt syndrome is acute herpes zoster oticus associated with acute facial nerve palsy. In one series, Ramsay Hunt syndrome represented approximately 15% of 2465 patients seen from 1976 to 1997 with facial nerve paralysis.94 Another retrospective study of 102 patients with Ramsay Hunt noted that it occurred more frequently in males, in a bimodal distribution between ages 10 to 29 and ages 50 to 59.49 The reader is referred to available reviews of the history, natural history, and epidemiology of this disorder.11,49,221 Facial nerve involvement is of rapid onset and occurs typically with severe otalgia localized to the ear canal, followed by development of herpetic vesicles in the external auditory canal11,49,51,205,233 or oral cavity.49 Additional common symptoms include tinnitus, hearing loss, nausea, vomiting, vertigo, and nystagmus, indicating vestibulocochlear nerve damage.205,221 Other cranial nerves (V, IX, X, XI, and XII) may be affected less commonly.233 The facial nerve weakness is most severe at onset, which is in contrast to Bell’s palsy, in which facial weakness progresses during the initial few days.49,221 As with Bell’s palsy, the severity is worse with increasing patient age.49,51 Recovery outcomes are poorer than with Bell’s or idiopathic facial palsy.23,26,49,51,205,221 Recurrences are rare, however.26 Cerebrospinal fluid and/or serologic examinations may demonstrate elevated antibody titers to varicella zoster.11,221,233 MRI examination may demonstrate postgadolinium enhancement of the facial nerve, the extent of which seems to correlate with the severity of facial paralysis; that is, in patients with severe or complete palsy, the entire nerve enhanced after contrast administration, whereas there was absence of nerve enhancement in patients with mild facial weakness.27 Pathophysiologically, this cranial polyneuritis is due to reactivation of varicella-zoster virus in the geniculate ganglion and spreading of viral inflammation and edema, causing acute nerve compression and dysfunction.11,94 It is assumed that viral infection of the nerve leads to increased

1232

Diseases of the Peripheral Nervous System

permeability of the endoneurium, resulting in edema.205 Aleksic et al.11 described the pathologic findings in two cases, one with autopsy findings. They noted moderate fiber loss in the facial nerve, cell loss in the geniculate ganglion, small collections of inflammatory cells within the ganglion stroma, and neurons with central chromatolysis.11 Electron microscopic examination of a facial nerve biopsy during surgical decompression showed clusters of Schwann cell cytoplasmic processes and absence of axons or myelin, suggesting axonal retraction.84 Honda et al.94 examined the facial nerve in situ in 74 patients undergoing surgical decompression of the nerve for severe Ramsay Hunt syndrome (complete paralysis, denervation by electroneurography greater than 95%, absence of nerve excitability with electrical stimulus ⱖ 10 mA). They noted swelling of the nerve in the horizontal and vertical segments of the facial canal as well as the geniculate ganglion; examination of patients at different time points after the onset of facial palsy revealed that the severe swelling decreased with time but was still present in 50% of patients who had surgery more than 9 weeks after onset of facial paralysis.94 Presumably the pathophysiology of cranial polyneuritis is due to transaxonal spread of varicellazoster virus between the geniculate ganglionic afferent fibers to the vasa vasorum of adjacent cranial nerves.221 Treatment of patients with Ramsay Hunt syndrome includes early administration of antiviral medication and prednisone.221 Dickins et al.51 reviewed seven patients with Ramsay Hunt syndrome who were treated with acyclovir within the first 5 days after onset of facial paralysis, five of whom had complete facial palsy, and found a recovery rate of 57%; they recommended initiation of antiviral intravenous therapy within the first 72 hours after onset. Uri et al.233 reviewed their experience with five patients with Ramsay Hunt syndrome treated with acyclovir intravenously between 2 and 6 days after onset of facial weakness, followed by oral dosing in three of the five patients, and found 70% recovery after 1 month and up to 100% recovery by 24 months. These reports have been limited to small numbers of patients, so the true efficacy of treatment is not clearly known.

Acute Facial Paralysis: Other Causes Granulomatous Disease Neurosarcoidosis is a granulomatous disease that may frequently present with facial nerve palsy,39,108,109,219,241 with a high incidence of bilateral involvement at onset.39,109,219,241 James108 reported that facial nerve lesions affected both sides of the face with equal frequency and that bilateral facial nerve palsy is most likely due to sarcoidosis in young adults. Facial nerve palsy caused by sarcoidosis may be associated with involvement of other cranial nerves,108,109,219,241 and with papilledema, peripheral neuropathy, meningitis, and parotid gland enlargement.108 James and Sharma in

1967 reviewed 38 patients with neurosarcoidosis (out of 537 patients with confirmed sarcoidosis) and found that 25 of 38 had facial nerve palsy, bilaterally in 7.109 They found that facial nerve palsy in sarcoidosis had a good prognosis for recovery108,109 with only rare recurrence,108 which is confirmed by others.39,219 Facial nerve palsy may also rarely be the presenting symptom of other granulomatous diseases, such as Wegener’s granulomatosis.138 Amyloidosis Facial neuropathy may also be a manifestation of amyloidosis,143,231 including as an initial neurologic sign.231 Traynor et al.231 reviewed the cases of six patients with primary amyloidosis and found three with multiple cranial neuropathies, two with isolated facial neuropathy, and five of six with facial nerve involvement. Suspicion for this diagnosis, if facial neuropathy is the initial symptom, may be raised if there is proteinuria on urinalysis and/or monoclonal protein found by immunofixation of serum or urine.231 These patients had poor clinical response to treatment of their underlying disease with prednisone and melphalan.231 Massey and Massey143 published a case report of an older man who presented with congestive heart failure and facial diplegia resulting from amyloidosis. Electromyography showed bilateral low-amplitude facial compound muscle action potentials with prolonged distal latencies and a generalized mixed, primarily axonal, peripheral neuropathy.143 Trauma Blunt or closed head trauma may also lead to facial nerve paralysis. Facial nerve injury secondary to head trauma is second only to olfactory nerve injuries in terms of cranial nerve involvement from trauma.118 Guerrissi78 reviewed 30 cases of facial paralysis seen between 1991 and 1996 and found that 17% were due to blunt head trauma. There was poorer functional recovery if electroneurography demonstrated greater than 85% denervation on the affected side, if there was facial pain, and if there was complete facial paralysis clinically.78 Overall, recovery is seen in 80% to 90% of patients who are of young age, have incomplete weakness, and have early signs of recovery of function.78 Three of five patients in this series recovered completely and spontaneously after trauma, two with peripheral facial nerve branch trauma and one with intratemporal trauma but without fracture.78 Chang and Cass35 performed a meta-analysis study and concluded that patients who had normal facial nerve function immediately after nonpenetrating temporal bone trauma, whether they progressed to have incomplete or complete paralysis, usually had eventual functional recovery to normal or near normal. If patients maintained less than 90% axonal degeneration for the first 6 days by electroneurography or did not progress to greater than 95% degeneration within the first 14 days after injury, they had good prognosis for recovery.35

Diseases of the Seventh Cranial Nerve

Progression from incomplete to complete paralysis after trauma may have a good outcome, especially with delayed progression.35 Adour et al.4 found that 11 of 15 patients with incomplete facial paralysis after closed head injury had complete recovery without complications within 3 to 6 weeks after injury. The majority of patients with facial nerve paralysis (80%, 24 of 30 patients) after closed head injury had a temporal bone fracture.4 Penetrating trauma with facial nerve involvement usually presents initially with immediateonset paralysis and implies more severe nerve injury such as partial or complete nerve transection, usually requiring early surgical exploration and repair.35,205 Immediate as opposed to delayed-onset facial paralysis is thought to indicate direct nerve damage, such as transection, leading to complete denervation and is treated surgically by some authors, although this approach has not been adequately tested. Adour et al.4 found that patients treated with decompressive surgery within 8 weeks of injury because of complete and immediate facial paralysis had a poorer outcome than those patients treated with exploratory surgery or no surgery at all; however, the patients who had the surgery were clinically more severely affected at the onset.4 These authors concluded that decompressive surgery is probably only helpful if it is performed immediately after injury to prevent denervation, although this approach is usually impractical in patients who generally have multiple injuries in the setting of trauma.4 Some authors advocate treatment with steroids during the first week after traumatic injury, whether the paralysis is complete or not, and no surgical decompression for extratemporal or peripheral facial nerve branch lesions.78 Other authors advocate facial nerve surgical decompression if traumatic facial paralysis is immediate and complete23,26,205 and there is greater than 95% axonal degeneration by electrodiagnostic testing.23,26 The worst prognosis for functional recovery is thought to occur if denervation occurs within the first week of injury and is complete in severity.23 Chang and Cass35 recommended surgical exploration/decompression after traumatic facial nerve injury if there is greater than 95% axonal degeneration by electroneurography between 6 and 14 days after injury. Yanagihara et al.244 reviewed their series of 247 patients with facial nerve palsy after closed head injury seen from 1966 to 1994, of which 81 patients had surgical exploration of facial nerve with the majority (82.9%) having an associated unilateral temporal bone fracture. The most common fracture line extended across the mastoid process to the external auditory canal.244 Other authors have found that, although longitudinal temporal bone fractures are the most common after closed head injury, transversely oriented temporal bone fractures are more commonly associated with facial nerve paralysis (Fig. 50–7).26,107,179,205 Facial nerve injury secondary to temporal bone fracture may occur with compression by hematoma or stretch injury by separation

1233

FIGURE 50–7 Axial computed tomography scan showing transverse fracture (arrowhead) of the left temporal bone extending toward the vicinity of the horizontal portion of the left facial nerve. (Courtesy of Mauricio Castillo, M.D., Department of Radiology, The University of North Carolina, Chapel Hill).

of the temporal bone fragments across the fracture line,107,179,205 impingement by bony spicules, neural contusion, or even nerve transection.205 However, Yanagihara et al.244 found that the continuity of the facial nerve was usually maintained across the fracture line at the time of surgery and concluded that the paralysis was most likely due to external or internal compression from vasogenic edema. These authors recommended surgical decompression of the facial nerve from the site of the lesion to the stylomastoid foramen if the facial nerve palsy is complete and occurs immediately after the traumatic injury. If the facial weakness is progressive or delayed in onset, they recommended use of electrodiagnostic testing to assess the severity of denervation and proceeding to surgery if total or near-total denervation has occurred.244 Although the traumatized facial nerve may enhance with gadolinium on MRI images, this finding reportedly does not correlate with the severity of any electrophysiologic abnormalities seen, and presence or absence of nerve enhancement should not be used to predict prognosis after traumatic facial nerve paralysis.107 Holla et al.93 reported three cases of bilateral traumatic facial nerve paralysis that was delayed from 8 to 15 days after injury. Two of the three patients were found to have petrous bone fractures and, in terms of outcome, two had complete functional recovery by 6 weeks and the other patient experienced fair recovery of function by 3 months. The authors concluded that delayed onset of facial nerve paralysis led to a better prognosis than paralysis that was noted immediately after injury, presumably because delayed paralysis is less likely to be due to direct nerve injury.93

1234

Diseases of the Peripheral Nervous System

Connective Tissue Diseases Connective tissue diseases, such as Sjögren’s syndrome, have been associated with acute facial nerve paralysis. Uchihara et al.232 reported a case of a patient who presented with bilateral parotid gland swelling and subsequent bilateral facial palsy, without other cranial nerve deficits. The patient had a history of sicca symptoms and, upon evaluation, had positive antinuclear antibodies, antiSjögren’s syndrome antigen A and rheumatoid factor, and inflammation on minor salivary gland biopsy as well as increased cerebrospinal fluid protein. The patient had gradual improvement in facial function following treatment with oral prednisone.232 A more recent case report79 described a patient with peripheral facial nerve palsy, papillary abnormalities, and altered facial sensation as presenting symptoms of Sjögren’s syndrome. Rarely has facial nerve lesion been associated with systemic vasculitis. Dudley and Goodman53 described a patient with a history of bilateral chronic otitis media and bilateral facial paralysis who subsequently was diagnosed with periarteritis nodosa. At autopsy, there was evidence of granulation tissue and inflammation in the facial nerve and arteritis with neurotmesis of the sciatic nerve.53

neuropathies in the setting of acute aseptic meningitis in early stages of infection.24,148,159,215 The facial paralysis may be bilateral.24,215,229,240 The outcome for isolated facial nerve palsy associated with HIV infection is good.24,148,229,240 More rarely, facial weakness may be due to a brainstem mass lesion in patients with HIV, usually associated with other cranial nerve deficits and long tract signs.148 Belec et al.24 recommend that high-risk patients or those in an endemic HIV area who present with acute facial nerve palsy be checked upon initial presentation for HIV but also rechecked at a later time if negative because of the occurrence of facial paralysis early in the course of HIV infection. HIV infection may be related to acute facial paralysis as a result of a higher risk of herpes or other viral opportunistic infections as well, presenting with Ramsey Hunt syndrome,215 Burkitt’s lymphoma,163,215 or disseminated herpes.163,229 Marra141 suggested that acute facial paralysis in patients with HIV is more likely due to immune dysfunction allowing increased latent HSV type 1 replication than to direct HIV infection of the nerve. Other authors have suggested that HIV infection may result in bilateral parotid gland enlargement with resulting facial nerve paralysis.108

Diabetes Mellitus Acute facial nerve paralysis has been associated with newly diagnosed diabetes mellitus. Hattori and Schlagenhauff86 reported a 75-year-old man with acute bilateral facial paralysis and normal cerebrospinal fluid examination who was diagnosed with diabetes at presentation. The association of diabetes with acute facial nerve paralysis was examined by Adour et al.3 who reviewed 395 patients with a single episode of Bell’s palsy and found that 94 patients had a positive family history of diabetes mellitus, with 25 (6.4%) of the patients being diabetic. They also evaluated 232 patients with idiopathic facial paralysis with a 2-hour glucose tolerance test and found that the results were abnormal in 23%. The authors concluded that the prevalence rate for diabetes in their study population (11%) was greater than that in the general population (4%). Using nerve excitability testing thresholds, they found a higher frequency of complete facial paralysis in diabetic patients with idiopathic facial paralysis and concluded that diabetes was an adverse prognostic factor for Bell’s palsy.3 Another study found that 58 of 288 patients with Bell’s palsy had an abnormal hemoglobin A1C. The recovery rate for diabetic patients who received highdose oral steroids was the same as for patients without diabetes, and, with careful management of their blood glucose, these patients were able to tolerate steroid therapy without significant loss of glycemic control.195

Pregnancy The association between pregnancy and acute idiopathic facial paralysis remains controversial. Hilsinger et al.91 reported the overall incidence of Bell’s palsy in women to be 16.9 per 100,000 per year, but that of women of childbearing age to be slightly higher, 17.4 per 100,000 per year. The frequency of Bell’s palsy in pregnant women is 45.1 per 100,000 births.1,91 Therefore, per year of exposure, the risk to pregnant women is more than three times higher than that to nonpregnant women of the same age.1,5,91 The incidence during late pregnancy and early postpartum periods is the highest, reportedly 118.2 per 100,000 per year of exposure. Other studies have demonstrated an increased frequency of Bell’s palsy during the third trimester or early postpartum periods.1,5,75,207 However, other authors reported no increased incidence of Bell’s palsy in pregnancy in their study populations.87 Upon retrospective review of 42 cases of Bell’s palsy in pregnancy or early postpartum from 1966 to 1974, Hilsinger et al.91 found a low association with toxemia; however, Shapiro et al.207 reviewed the available literature and found a 22% pooled event rate of Bell’s palsy with patients with preeclampsia and concluded that facial paralysis may be a possible prodromal sign of preeclampsia. Of the 42 patients evaluated retrospectively, 26 had received prednisone therapy without adverse side effects.91 Gillman et al.75 recently presented data from a retrospective review of outcomes of Bell’s palsy in pregnancy and found that 52% of pregnant patients with complete facial paralysis within 10 days of onset recovered completely or had only mild facial weakness, compared to 77% to 88% with good

Human Immunodeficiency Virus Acute facial paralysis may be associated with human immunodeficiency virus (HIV) infection, either at the time of seroconversion24,148,163,240 or as one of multiple cranial

Diseases of the Seventh Cranial Nerve

recovery in nonpregnant women or men with complete facial paralysis. In this study, of the pregnant patients who were treated with steroids, 45% had a good functional outcome, compared to 55% with good outcome who were not treated with steroids. The patient group numbers were low and this was not a randomized, placebo-controlled trial. All of the pregnant patients in this study who had incomplete facial paralysis recovered completely or with mild deficits. In contrast, Steiner and Cohen218 reported four cases of Bell’s palsy in third-trimester patients, all of whom had an incomplete or no recovery of function. Interestingly, the majority of these patients were older mothers and it occurred during their first term pregnancy. Many authors believe that there is an increased risk for Bell’s palsy during the first 14 days of the menstrual cycle among women of childbearing age.91 Benign Intracranial Hypertension There have been case reports of acute facial diplegia associated with benign intracranial hypertension. Selky et al.206 described a case of a 17-year-old woman with pseudotumor cerebri with bilateral facial nerve paralysis, decreased tearing, and loss of taste on the anterior two thirds of the tongue. Medical therapy and serial lumbar punctures did not lead to any improvement of facial function. The authors concluded that the nerve was stretched and compressed in the proximal facial canal as a result of brainstem shifts during periods of increased intracranial pressure.206 Another patient had resolution of facial diplegia with placement of a lumboperitoneal shunt.123 Tumor Mass lesions involving the facial nerve may be an underlying cause of persistent facial paralysis, especially if there is electrophysiologic evidence for ongoing denervation or absence of regeneration more than 6 months after the onset of facial paralysis.114 Initial neuroimaging may not be positive and repeat imaging may be necessary. Jungehuelsing et al.114 reviewed 27 patients with malignant tumors affecting the facial nerve of 486 patients with facial nerve paralysis, including 8 patients with initially negative MRI studies. Follow-up imaging revealed parotid gland tumors that were found to be malignant at the time of surgery, with perineural, intraneural, and perivascular extension along the facial nerve pathologically.114 Neely165 stated that the superficial lobe of the parotid gland is the most common location for mass lesions and that development of facial paralysis as a result of a parotid gland mass usually is associated with malignancy, with extension of the tumor along the nerve. Adenoid cystic carcinoma is the most common malignant tumor type.165 Primary malignant tumors of the temporal bone typically affect the facial nerve in the tympanic or mastoid segments, whereas metastatic tumors invade the facial canal usually in the region of the geniculate ganglion.165

1235

Marzo et al.142 published a recent retrospective analysis of 11 of 320 patients with facial weakness with temporal bone neoplasms, parotid gland tumors, or a combination of these as a cause for facial paralysis. A variety of malignant tumor types were diagnosed in these patients, including adenocarcinoma, squamous cell carcinoma, adenoid cystic carcinoma, and osteogenic sarcoma. They recommended, in patients with painful facial nerve palsy that is unresolved after 3 months, obtaining a contrast-enhanced MRI of the head, internal auditory canal, and parotid gland as well as a temporal bone CT scan. They also recommended total surgical exploration of the facial nerve in patients with normal imaging results in order to exclude adequately the presence of a mass lesion.142 The facial nerve may be compressed by meningioma located in the internal auditory canal, jugular bulb, or geniculate ganglion.165 Facial nerve paralysis secondary to an acoustic neuroma is rare.165 However, a facial neuroma may occur along the course of the nerve in the facial canal, most commonly in the tympanic segment, with associated conductive hearing loss.165 Sherman et al.209 recently reviewed retrospectively 10 cases of facial nerve neuromas and reviewed the literature. The majority of patients (8) had right facial nerve involvement. The most common symptoms were facial weakness (progressive or recurrent) and hearing loss, but other symptoms included vertigo, tinnitus, otalgia, and dysequilibrium.209 The majority of the facial nerve neuromas were located in the cerebellopontine angle or the internal auditory canal.209 Hemifacial spasm was more likely to occur in patients with facial nerve neuroma than in patients with acoustic neuroma.209 Kania et al.115 recently published a case report of a patient with a 9-month history of hypoacusis and progressive unilateral facial palsy who underwent surgical exploration. Grossly, the facial nerve was found to be enlarged from the labyrinthine segment to the stylomastoid foramen. Biopsy showed fusiform enlargement of the nerve with onion bulb formation and negative histology for perineuroma. These authors labeled this case as a localized hypertrophic neuropathy affecting the facial nerve.115 Malignant tumors of the temporal bone may present with otalgia, facial paralysis, and otic drainage with primary tumors of the external auditory canal such as squamous cell carcinoma.165 Metastatic tumors are rare, but may be associated with nasopharyngeal carcinoma.165 The facial nerve may also be affected by large tumors of the glomus jugulare165 or, rarely, primary facial canal paragangliomas.115 The latter must be differentiated from hemangiomas, which may occur in the area of the geniculate ganglion and cause erosion of the facial canal.115 Miscellaneous Causes Other infectious causes for facial nerve paralysis include osteomyelitis of the skull base, which may be due to Pseudomonas aeruginosa infection and present with ear

1236

Diseases of the Peripheral Nervous System

symptoms, headache, and cranial nerve deficits.139 Diagnosis may be facilitated with routine neuroimaging (CT, MRI) combined with bone and gallium scanning.139 More unusual infectious causes include trichinosis,135 Mycoplasma pneumoniae,59 and herpes zoster ophthalmicus,210 which have been described in isolated case reports to result in facial diplegia by unknown pathophysiologic mechanisms. Otogenic facial nerve palsy occurs secondary to middle ear infections such as acute, subacute, or chronic suppurative otitis media causing inflammation, fibrosis, and ingrowth of granulation tissue affecting nearby facial nerve fibers.23 James108 reported an increased prevalence of facial nerve palsy in patients from areas endemic for human T-lymphotrophic virus type 1 (HTLV-1) infection. Affected patients were positive for HTLV-1 antibodies and presented with unilateral facial nerve paralysis with good recovery.108 More rare etiologic associations with acute facial nerve paralysis include Stevens-Johnson syndrome,67 ethylene glycol ingestion,138 and Wernicke-Korsakoff syndrome.189 Facial diplegia has been reported as an initial symptom of adult T-cell leukemia199 and as a rare complication of arterial embolization for epistaxis in a patient with OslerWeber-Rendu disease.149

Inherited/Familial Facial Palsy The possibility that a family history of Bell’s palsy may be a risk factor for this condition has been examined. One of the earliest studies was reported by Alter,14 in which 105 patients with Bell’s palsy were interviewed retrospectively regarding the occurrence of the condition in other family members and 29% of patients were found to have a positive family history, which was increased compared to a matched control group. One set of identical twins in the study was discordant for Bell’s palsy.14 Other studies have found lower incidences of positive family history.1,116,177 Yanagihara et al.245 reviewed 625 patients diagnosed with Bell’s palsy from 1976 to 1985 and found that 26 (4%) had a positive family history, with possible autosomal dominant inheritance with low penetrance. In these patients, there appeared to be a trend toward more frequent recurrence of Bell’s palsy; however, there were too few patients to conclude this.245 In this series, although there was a trend for more severe facial weakness in those patients with positive family history, their outcomes were good. Yanagihara et al.245 noted a slightly increased frequency of diabetes mellitus in patients with positive family history for Bell’s palsy, and suggested that diabetes may have been an inherited risk factor. Hageman et al.80 examined six cases in a Dutch family and determined that half of the affected patients had bony abnormalities of the mastoid and petrous bones, which may be a predisposing factor. Familial recurrent facial palsy has also been reported in the literature. Poloni et al.182 examined one family with peripheral myelin protein 22 deletion and found that the

proband had a history of recurrent facial palsy. Nerve conduction studies of the extremities showed diffusely slowed conduction velocities and prolonged distal latencies and prolonged R1 latencies on blink reflex testing. Aldrich et al.10 reported a family in which the father and 7 of his 10 offspring had histories of multiple bilateral recurrent Bell’s palsies, the initial episode occurring at as young an age as 13 and as late an age as 73, with the majority of family members who were affected having Bell’s palsy after age 40. These patients had associated oculomotor palsies as well.10

Congenital/Pediatric Facial Nerve Paralysis Facial nerve paralysis in children may be congenital (Möbius’ syndrome, hemifacial microsomia, congenital unilateral lower lip paralysis) or acquired prenatally (compression injury in utero via maternal sacrum, teratogenetic from infection or drugs) or postnatally (Bell’s palsy, Lyme disease, otitis media, Kawasaki disease).106,174,205 Hauser and co-workers87 found the birth incidence of Bell’s palsy to be 0.9 per 1000 live births in Rochester, Minnesota. Other conditions that may lead to facial nerve paralysis include Albers-Schönberg disease or osteopetrosis, infantile hypercalcemia, osteomyelitis of the skull base, and various intracranial (astrocytoma, medulloblastoma, pontine glioma, posterior fossa mass lesions) and extracranial (neurilemoma) tumors.174 Congenital cholesteatoma may present as a gradual loss of facial nerve function ipsilaterally.165 Inamura et al.101 published a retrospective review of facial nerve palsy in Japan and examined 82 patients who were 6 years old or younger. The majority (58.5%) of these patients were diagnosed with Bell’s palsy, with the remaining patients having congenital (15.9%) or traumatic (6.1%) etiologies; two patients had leukemia. Of 58 patients evaluated within the first 2 weeks of onset of facial nerve paralysis, 96.5% had complete recovery with no apparent effect of treatment with vitamins or steroids.101 Recurrent facial nerve palsy occurs rarely in pediatric patients.58 The reader is referred to available reviews in the literature of causes of facial paresis in children.128,205,214 Bilateral facial nerve paralysis in pediatric patients is rare and may be due to a variety of causes. Smith and Traquina214 reported four patients with bilateral facial nerve paralysis and found the underlying causes to be bilateral serous otitis media, Lyme disease, relapse of acute lymphocytic leukemia, and postinfectious acute demyelinating encephalomyelitis. Infectious Causes Facial nerve paralysis is the most common neurologic sign of Lyme disease in children.25 Belman et al.25 reviewed the cases of 96 children in Suffolk County, New York, an endemic Lyme area, and found that 13 patients (14%) had facial nerve palsy, with bilateral paralysis in 4 children, and in 7 patients the facial weakness was the initial manifestation

Diseases of the Seventh Cranial Nerve

of B. burgdorferi infection. All of the patients in this series developed facial nerve palsy within 12 weeks of their initial infection, but less than half had other systemic symptoms of erythema chronicum migrans, polyarthritis, or flulike symptoms. An additional child developed bilateral facial nerve paralysis associated with distal extremity weakness and sensory changes, with elevated cerebrospinal fluid protein, increased white blood cell count, and positive antibodies to B. burgdorferi, suggestive of Guillain-Barré syndrome.25 Another retrospective study of facial nerve paralysis in children from an endemic Lyme disease area found that, of 50 children with transient facial nerve palsy, 50% were diagnosed with Lyme disease and 26% with idiopathic facial nerve paralysis or Bell’s palsy; the remainder had other infectious causes such as acute otitis media (12%), varicellazoster (6%), herpes zoster (4%), and coxsackievirus (2%).40 None of the patients in this series with Lyme disease and facial nerve palsy had bilateral symptoms,40 and the majority did have a history of systemic symptoms.40 The prognosis for recovery in these patients with Lyme disease is good with antibiotic therapy.40 Ancillary testing for patients with suspected Lyme disease as a cause for acute facial nerve paralysis includes MRI and cerebrospinal fluid examination. Vanzieleghem et al.238 reported a case of a 7-year-old child with bilateral facial nerve paralysis resulting from Lyme disease whose MRI showed enhancement after contrast administration of bilateral facial nerves in the internal acoustic canal, and increased enhancement of the nerve ipsilateral to the side of the face with slower recovery of function. A study from Switzerland reviewed cerebrospinal fluid and serologic testing in 21 consecutive pediatric patients with facial nerve palsy and suspected Lyme disease and found that 20 of 21 patients had antibodies to B. burgdorferi in acutephase serum; 16 patients had lymphocytic pleocytosis in cerebrospinal fluid (associated with increased cerebrospinal fluid protein in 10) and production of intrathecal antibody to B. burgdorferi was noted in 5 of 20 seropositive patients.9 A definitive diagnosis of neuroborreliosis was made in 13 of 21 (63%). Other Causes Trauma. Traumatic injury to the facial nerve in childhood may occur more easily from external compression because of the ongoing development of the mastoid process during the first 2 years of life.174 Infants delivered by forceps with penetrating injuries of the facial nerve at the mastoid process have a good prognosis for spontaneous recovery.174 Leukemia. Facial nerve palsy may be a rare presenting symptom of childhood leukemia.128 It may also be a rare neurologic complication in Kawasaki disease,31 a systemic vasculitic disease. If present in these patients, facial nerve palsy is associated with an increased risk for coronary artery aneurysm, and thus mortality.31

1237

Cardiofacial Syndrome. Cardiofacial syndrome, or asymmetrical crying facies syndrome, is an association of facial nerve paralysis with other congenital malformations (usually cardiac, but also bony, genitourinary, and respiratory) and other cranial nerve or central nervous system abnormalities.186,198 Punal et al.186 found that in three cases this condition was associated with de novo deletions of paternal chromosome 22q11. The facial paralysis is typically limited to the lower facial muscles in these patients and is thought to be due to congenital hypoplasia of the depressor anguli oris muscle more so than a facial nerve lesion per se. Melkersson-Rosenthal Syndrome. Melkersson-Rosenthal syndrome is a condition with recurrent orofacial edema, recurrent alternating facial palsy, and tongue abnormalities (lingua plicata).190,248 The exact etiology of this syndrome is unknown,13,190 although it is postulated to have an inherited or viral cause.13,15 The peripheral facial nerve palsy is clinically similar to idiopathic facial nerve palsy and may be unilateral or bilateral, incomplete or complete, although the recurrent episodes of facial palsies tend to be associated with ipsilateral facial edema.13,43,76,173,190,248 The incidence of facial nerve palsy is estimated to be between 20% and 47% of patients with this condition throughout the course of the disease.43,76 The reader is referred to available literature for review of the history of this clinical syndrome.13,15 Although there is usually good recovery of function from the facial paralysis, some residual weakness may develop after recurrent episodes.13,173 Of 42 patients reviewed at the Mayo Clinic who were diagnosed between 1965 and 1990, 33% had facial palsy, 19% with facial palsy as the initial symptom.248 There may be other associated neurologic symptoms, including migraine headache,13,76,190 hyperhidrosis, hypogeusia, glossodynia, acroparesthesias, hyperacusis, blepharospasm, unilateral rhinorrhea, altered salivation, and facial dysesthesias.43,190 Biopsy of affected tissue may reveal noncaseating granuloma,108 but this finding is not required for the diagnosis.15,43,76,190,248 Other clinical features of sarcoidosis are absent.108 Orlando and Atkins173 reported two cases with elevated angiotensinconverting enzyme serum levels during episodes of facial paralysis, which returned to normal with recovery of facial function. Melkersson-Rosenthal syndrome may be rarely diagnosed in childhood,174,247 and should be suspected in a child with recurrent facial paralysis. Because facial nerve palsy may often be the only presenting symptom43,174 and may precede the orofacial edema by years, an index of suspicion and evaluation for other associated features may aid in the diagnosis. Given the intermittent appearance of symptoms over a period of time, Melkersson-Rosenthal syndrome is usually diagnosed in the second decade of life.76 Treatment includes oral steroids and possibly facial

1238

Diseases of the Peripheral Nervous System

nerve decompression in patients with prolonged facial nerve paralysis.23,55,76,190,248 Dutt et al.55 recommended total surgical decompression of the facial nerve in patients with Melkersson-Rosenthal syndrome who have increased frequency, duration, and severity of facial nerve paralysis and persistent residual paresis and synkinesis. Facial nerve decompression is thought to relieve recurrent nerve edema in the bony facial canal.55

HEMIFACIAL SPASM Clinical Features Hemifacial spasm consists of spontaneous, clonic, repetitive, involuntary twitching of facial musculature and at times sustained tonic contractions or contracture.52,62 It generally begins in the inferior portion of the orbicularis oculi and may spread over time to involve other lower more than upper facial nerve–innervated muscles.62,225 The typical patient is middle aged and female.57,72,202,239 Hemifacial spasm has rarely been described in children.131,208 Facial muscle spasm may occur subsequent to idiopathic facial paralysis, but usually there is no history of facial weakness. Wang and Jankovic239 reviewed retrospectively 158 patients with hemifacial spasm and compared their clinical features to those of 95 patients with isolated blepharospasm or cranial dystonia. They found that 90% of the patients reported periocular muscle involvement at the onset, worsening spasm with emotional stress or anxiety or certain voluntary facial movements, and relief of spasm with relaxation (although the movements persisted during sleep in 80% of the patients in the series), alcohol, and ipsilateral sensory input such as touch.239 In chronic hemifacial spasm, mild residual facial weakness may occur.62 Of the 158 patients, 8 had ipsilateral trigeminal neuralgia, 9 had a history of ipsilateral Bell’s palsy, and 3 had other family members with similar symptoms.239 One of the earliest published series of patients was that of Ehni and Woltman57 in 1945, which described 106 patients diagnosed at the Mayo Clinic with hemifacial spasm. They found that nine patients had experienced one or more periods of spontaneous remission that lasted for weeks to years before return of the spasm. In this series, 16 patients had facial weakness and 14 had ipsilateral hearing impairment. These authors postulated that the lesion causing the spasm occurred in the facial motor nucleus or in proximal facial nerve.57 Although not diagnosed in children in this series,57 other authors have acknowledged the possibility of pediatric hemifacial spasm, although pediatric cases are believed by some authors to be less likely to be idiopathic and more likely to be caused by tumor or other brainstem pathology.52,131,208 A follow-up retrospective study of an additional 20 patients diagnosed with hemifacial spasm at the Mayo Clinic from 1960 to 1984 revealed female

gender and right hemifacial predominance, with 19 of 20 with onset of symptoms in the orbicularis oculi and no history of Bell’s palsy.20 The average annual incidence was reported to be 0.78 per 100,000 (both genders), with the highest incidence occurring between ages 40 and 79. The average prevalence rate was 14.5 per 100,000 for women and 7.4 per 100,000 for men.20 Montagna et al.158 examined the changes in idiopathic versus postparalytic hemifacial spasm during sleep and found that, although hemifacial spasm may persist during sleep, it was reduced significantly in both types of patients during deeper stages of sleep. The authors proposed that this partial diminution of hemifacial spasm during sleep stages is suggestive of central inhibition of a central facial nuclear component involved in the generation of the spasm.158 Bilateral hemifacial spasm occurs rarely.57,62,92,193,225,237,239 In the early Mayo Clinic series, bilateral cases developed over a period of time, ranging from less than 1 year to 15 years between sides, with the initial side noted to be consistently more severely affected.57 Clinically, bilateral hemifacial spasm may be distinguished from other facial movement disorders by the nonsimultaneous onset between sides of the face and the asynchronous nature of the spasm occurring on each side.225,239 Tan and Jankovic225 reviewed the available literature and found that, in the majority of cases, bilateral hemifacial spasm had its onset on the left side of the face and was secondary to vascular pathology, with a variable latency of onset between sides and female gender predominance. Overall, the frequency of occurrence for bilateral hemifacial spasm is between 0.6% and 5% of the total number of patients seen with hemifacial spasm, according to a recent review of the literature.225 Differential diagnosis for bilateral hemifacial spasm includes blepharospasm, oromandibular dystonia, facial tics, facial myokymia, and facial synkinesis after Bell’s palsy.92,225 There is a single case report of a patient with alternating bilateral hemifacial spasm following unilateral facial palsy who was later diagnosed with multiple sclerosis by laboratory criteria,237 although another case of alternating bilateral hemifacial spasm appeared to be due to multiple sites of vascular compression requiring repeated surgical decompression to alleviate.193 Familial Hemifacial Spasm Hemifacial spasm occurring in several members of the same family has been reported occasionally in the literature. Miwa et al.150 published a recent review of 13 patients described in the literature as well as 10 patients from five families with familial hemifacial spasm. Mode of inheritance was unclear, but may be autosomal dominant with low penetrance.150 There was a left-sided facial predominance, with the typical patient being middle aged and female, although others have reported earlier age of onset in affected male family members with familial hemifacial

Diseases of the Seventh Cranial Nerve

1239

spasm.33 The basis of inherited hemifacial spasm may be a genetic susceptibility to facial nerve injury based on anatomic features37,150 or an inherently lower threshold for excitability in the facial nerve.150 Friedman et al.66 reported a family from Poland with familial hemifacial spasm. Electrophysiologically, these affected family members had no evidence of facial neuropathy or denervation, but demonstrated bursts of spontaneous, normal-appearing motor unit potentials with increased frequency of occurrence after voluntary facial muscle movement.66 Duff et al.54 reported a rare patient with familial trigeminal neuralgia associated with contralateral hemifacial spasm. In this patient, the hemifacial spasm occurred before the trigeminal neuralgia, although eight other family members across three generations had trigeminal neuralgia only. This family history suggested an inherited susceptibility to multiple cranial rhizopathies or central nervous system hyperexcitability related to cranial nerve nuclei.54

Differential Diagnosis Hemifacial spasm must be differentiated clinically from other facial movement disorders such as blepharospasm (which has bilateral, synchronous involuntary contractions of the orbicularis oculi), facial myokymia, oromandibular dystonia, facial focal motor seizures, craniofacial myoclonus, hemimasticatory spasm,70 cephalic tetanus, and craniofacial tics or tremor.52,62,70 The reader is referred to reviews of the differential diagnosis of hemifacial spasm available in the literature.52,62

Etiologies Hemifacial spasm may develop in association with a variety of etiologies, including vascular compression of the facial nerve at the root entry zone (Fig. 50–8; see later), or extraaxial mass lesions such as cholesteatoma with extension into the cerebellopontine angle52,202 or into the tympanic portion of the temporal bone from the mastoid antrum,44 facial neuroma within the internal auditory canal, meningioma, parotid tumor,52,70,170,202,216 acoustic neuroma, or angioma in the posterior cranial fossa.70,202,216 Pierry and Cameron180 reported a case of hemifacial spasm caused by posterior fossa arteriovenous malformation in which the facial muscle twitching was synchronous with the patient’s pulse, and the spasm completely resolved after resection of the vessels compressing the facial nerve. With posterior fossa tumors that cause hemifacial spasm, there are usually other localizing symptoms such as tinnitus, hearing loss, unsteady gait, and nystagmus.70 Rarely, it may be caused by brainstem ischemia,200 pontine tumor,127 or demyelinating plaques of multiple sclerosis.62 Sandberg and Souweidane196 described a pediatric case of hemifacial spasm as the only symptom of a pilocytic astrocytoma of the fourth ventricle in a child with neurofibromatosis

FIGURE 50–8 An elderly man with right hemifacial spasm. A, Source image from three-dimensional time-of-flight magnetic resonance angiogram acquisition shows deformity of cisternal segment of cranial nerve VII by ectatic and tortuous basilar artery (arrows). B, Axial T2-weighted magnetic resonance image is not convincing in showing offending vascular loop (arrow). (From Phillips, C. D., and Bubash, L. A.: The facial nerve: anatomy and common pathology. Semin. Ultrasound CT MR 23:202, 2002, with permission.)

type I. Sprik and Wirtschafter216 surveyed clinicians treating patients with hemifacial spasm and determined that, of 1676 patients evaluated, 0.54% had an intracranial tumor as the cause, although only about half of the patients had undergone neuroimaging studies. Hemifacial spasm may rarely be due to a contralateral mass in the posterior cranial fossa, presumably secondary to distortion or

1240

Diseases of the Peripheral Nervous System

angulation of the brainstem by the mass.72,145,169 Maroon and co-workers140 reported a case of a middle-aged woman who developed hemifacial spasm as an initial manifestation of a berry aneurysm at the bifurcation of the vertebral artery and the posterior inferior cerebellar artery, resulting in compression and stretching of the ipsilateral facial nerve. Clipping and retraction of the dome of the aneurysm by a sponge resulted in immediate relief of the hemifacial spasm.140 Other reports confirm aneurysmal compression as a rare cause of hemifacial spasm.72,134 An association between systemic hypertension and hemifacial spasm has also been proposed. Oliveira et al.172 retrospectively reviewed 48 cases of patients diagnosed with hemifacial spasm and found that 66.7% also had hypertension. Sixteen of the 42 patients with hemifacial spasm in this series had evidence of vascular tortuosity on neuroimaging studies. The authors suggested that hypertension may be a risk factor for development of hemifacial spasm in that longstanding and poorly controlled hypertension may lead to dilatation and elongation of blood vessels, which may subsequently cause compression of the facial nerve as it exits the brainstem.172 Ehni and Woltman57 suggested that there was a trend for younger patients (less than age 50) with hemifacial spasm to have systemic hypertension. Defazio et al.48 looked at hypertension in 115 consecutive patients with hemifacial spasm, and found a significant correlation between the two conditions, compared to age- and sexmatched control patients, with up to 50% of patients over age 65 having hypertension (compared to 37% of agematched control patients). Twenty-three of 63 patients in this series who underwent neuroimaging had findings suggestive of neurovascular compression of the facial nerve.48 Tan and Jankovic226 reviewed 158 consecutive cases of patients with hemifacial spasm and found that 37 also had systemic hypertension, 22 of whom had tortuous, ectatic blood vessels on neuroimaging. Fewer patients with hemifacial spasm without hypertension had abnormal neuroimaging for tortuous blood vessels, but the difference was not statistically significant.226 However, the most recent study published224 concluded that hypertension was not a risk factor for hemifacial spasm, at least in an Asian population of patients. This study evaluated 117 consecutive patients with hemifacial spasm and compared them to 245 control patients. No significant difference was found between the prevalence of hypertension in the two groups of patients.224 This is despite the fact that 94% of the patients with hemifacial spasm had evidence of neurovascular compression by neuroimaging studies, whether they had hypertension or not.224

Electrodiagnostic Testing Auger17 reported electrophysiologic findings from 23 patients diagnosed with hemifacial spasm who did not have prior histories of Bell’s palsy or facial nerve trauma.

He found that all patients with hemifacial spasm had synkinesis on blink reflex testing and recording from the orbicularis oculi and other facial nerve–innervated muscles ipsilaterally. He noted variability in the early and late responses of the orbicularis oris with stimulation of the ipsilateral supraorbital nerve. The early response from the orbicularis oris seemed to coincide with the ipsilateral R1 response to blink testing, with the late response correlating with the ipsilateral R2 response in the majority of patients. Needle examination of facial muscles in these patients was normal except for muscle spasm.17 Horowitz96 compared concentric needle examinations of facial muscles in patients with hemifacial spasm and myokymia and determined that patients with myokymia had synchronous, semiregular bursts of 2 to 7 motor unit potentials with silent periods of 100 to 200 ms, whereas in hemifacial spasm there were bursts of 1 to 10 motor unit potentials at 60 to 70 Hertz with a silent period of 7 to 8 seconds. With increased facial spasm the silent period in hemifacial spasm decreased to 80 to 150 ms.96 Facial nerve stimulation resulted in synkinesis of ipsilateral and contralateral facial muscles in patients with hemifacial spasm.96 Esteban and Molina-Negro61 conducted electrophysiology studies in 53 patients with hemifacial spasm and 20 control patients. By recording from the orbicularis oculi and oris muscles simultaneously, they determined that, in the majority of patients studied with hemifacial spasm, there was synkinesis with supraorbital and facial nerve electrical stimulation. With blink reflex testing, the R2 latency was shortened and duration of the response was increased on the side of the spasm with contralateral supraorbital nerve stimulation. They concluded that, in patients with hemifacial spasm, there is a lowering of the facial nucleus excitability threshold via either trigeminal or facial nerve afferent fibers.61 In contrast, Nielsen167 examined blink reflex testing in 62 patients with hemifacial spasm, recording from the orbicularis oculi and mentalis, and found increased R1 and R2 latencies on the ipsilateral side compared to control and the contralateral side of the face, respectively, associated with increased response amplitudes (Fig. 50–9). Nielsen concluded that the increased latencies and synkinetic responses were due to conduction slowing over a short segment of the facial nerve, such as the root entry zone, as a result of local microvascular compression and demyelination.167 This ectopic nerve activity is generated in an area of focal demyelination by mechanical irritation, decreased calcium in the perinodal environment, or flow of extracellular current between adjacent nerve fibers.166 This is possible because of the microscopic anatomy of the facial nerve within the cerebellopontine angle, with lack of epineurium, no interfascicular connective tissue, and lack of topographic organization of the facial nerve fibers within the main nerve trunk because the central portion of the nerve represents a transition zone between central and peripheral myelin.168

Diseases of the Seventh Cranial Nerve

1241

FIGURE 50–9 Hemifacial spasm. Blink reflex and synkinesis after stimulation of the supraorbital nerve before (A and C), and after (B, D, and E) microsurgical decompression of the facial nerve. Each pair of traces shows simultaneous recordings from the orbicularis oris (OO) and mental (Me) muscles. The amplification (⫽ distance between horizontal lines) is given in millivolts. Sweep length ⫽ 100 ms. A and B, Traces in a woman age 70 years show disappearance of a synkinetic response in the mental muscle 2 days after surgery (B). C, D, and E, Traces in a woman age 38 years show that the synkinetic response in the mental muscle was still present 3 days after surgery (D) but had disappeared 6 months later (E). Note also the decrease in amplitude of the orbicularis oculi reflex response. (From Nielsen, V. K., and Jannetta, P. J.: Pathophysiology of hemifacial spasm: III. Effects of facial nerve decompression. Neurology 34:891, 1984, with permission.)

Electrophysiologic techniques have been used by various investigators to try to support either of the two main theories as to the etiology of hemifacial spasm: ephaptic transmission between adjacent facial nerve fibers at a site of localized compression, or alteration of the excitability

threshold of the facial motor nucleus secondary to peripheral injury of the nerve leading to increased firing of the facial motor nerve fibers. Sanders197 performed singlefiber electromyographic recordings to measure jitter in the lateral spread responses in two patients with hemifacial

1242

Diseases of the Peripheral Nervous System

spasm and found that, in the orbicularis oculi and mentalis muscles, there was low jitter (less than 60 ␮s), suggesting these responses to be due to electrical transmission between facial nerve branches and not an additional central synaptic connection. However, Ishikawa et al.103 compared F waves in 10 patients with hemifacial spasm and found enhancement of F waves ipsilateral to the hemifacial spasm, with a linear correlation between the duration of the F wave and afterdischarge following the lateral spread response, and disappearance of the lateral spread response initially followed by reduction of the F wave after microvascular decompressive surgery. These authors concluded that the lateral spread response is an exaggerated F wave with the same nuclear origin, and that both the enhanced F wave and lateral spread response are indications of increased excitability of the facial motor nucleus.103 Repetitive stimulation of the zygomatic branch of the facial nerve in patients with hemifacial spasm and recording at the mentalis muscle for lateral spread responses reveals an increase in amplitude of the lateral spread response with decreased latency and enhanced F waves, suggesting facilitation of the late responses by repetitive stimulation resulting from increased excitability of the facial motor nucleus.102 Friedman and co-workers66 concluded that alteration in excitability in the motor neurons of the facial motor nucleus was responsible for R1 responses ipsilateral to hemifacial spasm that occurred with contralateral blink reflex testing in two patients they examined with possible familial hemifacial spasm. Clinical and electrophysiologic evidence of synkinesis occurring both sporadically or after ipsilateral facial nerve paralysis makes it less likely that it is due to aberrant regeneration of facial nerve fibers.122 The loss of contralateral synkinesis after unilateral facial nerve microvascular decompressive surgery in a patient with bilateral hemifacial spasm also suggests a role of the facial motor nucleus, with loss of facilitatory effect from the side ipsilateral to the surgical procedure203 (Fig. 50–10). Moller151 stimulated electrically the temporal branch of the facial nerve and the facial nerve trunk intracranially while recording from the mentalis and orbicularis oculi muscles in patients with hemifacial spasm undergoing microvascular decompressive surgery and found a latency difference between the lateral spread response and the sum of the conduction times in the segments involved if ephaptic transmission occurred at the site of compression, and concluded that this showed central synaptic transmission. The phenomenon of synkinesis has also been argued to be due to a peripheral facial nerve lesion. Comparing patients with hemifacial spasm who did or did not have percutaneous facial nerve block prior to microvascular decompressive surgery, Kim and Fukushima121 found that in both groups hemifacial spasm resolved clinically postoperatively, but that synkinesis persisted electrophysiologically postoperatively more often in patients who had a

prior peripheral facial nerve block, suggesting that synkinesis in the latter patients is due to a permanent and more peripheral nerve injury caused by the block rather than the more proximal peripheral nerve injury associated with microvascular compression in the patients without block, in whom the synkinesis resolved. Moller and Jannetta152 proposed that synkinesis was generated by a different process than hemifacial spasm, most likely by aberrant reverberation or a kindling phenomenon in the facial motor nucleus, with antidromic nerve impulses arising from a peripheral trigger zone such as the root entry zone as a result of vascular compression and spreading to the facial motor nucleus, leading to orthodromic stimulation of motor nerve branches and therefore synkinesis. Ferguson64 proposed that hemifacial spasm occurred as a result of reorganization of the facial motor nucleus after peripheral injury or compression, thereby unmasking normal and new reflexive and interconnecting pathways leading to multiple combinations of facial motor movements during hemifacial spasm, and that ephaptic transmission is too limited in its potential scope to explain the generation of stereotyped facial spasm. Because the facial motor nucleus has multiple unique afferent inputs (spinal cord, red nucleus, superior colliculus), such a central reorganization is possible.64

Pathology Ruby and Jannetta194 reported on the pathology from fascicular biopsies of the facial nerve obtained at the time of microvascular decompressive surgery in three patients with hemifacial spasm. They described some portions of the nerve appearing normal and others with “hypermyelination” of medium to large myelinated fibers with myelin ovoids that the authors thought might represent early microneuroma formation in the nerve.194 Kumagami130 described pathologic findings in two patients with hemifacial spasm, one with symptoms for 10 months and the other for 2 years prior to surgical nerve resection at the stylomastoid foramen, that consisted of nerve fibers with hypertrophied myelin sheaths and segmental demyelination intermixed with normal-appearing fibers.

Treatment Surgical Microvascular Decompression Microvascular decompression of the ipsilateral facial nerve as a definitive treatment for hemifacial spasm has been employed extensively over the last 25 years. One of the first series was published in 1977 by Jannetta et al.112 as a review of the authors’ experience with 45 patients with hemifacial spasm of variable duration (1 to 14 years) who underwent this procedure. Cross-compression of the facial nerve by the posterior inferior cerebellar artery at the level of the facial nerve root entry zone at the lateral aspect of the brainstem was the most common abnormality visualized

Diseases of the Seventh Cranial Nerve

1243

FIGURE 50–10 Four-channel surface electromyographic (EMG) recording from orbicularis oculi and oris muscles in a patient with bilateral hemifacial spasm. A, Spontaneously occurring spasms of facial muscles. EMG bursts are synchronous on each side (representing synkinesis) but asynchronous with regard to the contralateral side. Before surgery (top), bursts are more frequent and of higher amplitude and duration on the clinically symptomatic left side. After surgery (bottom), spasms occurred only on the right side at increased frequency. B, Blink reflex with electrical stimulation of the right supraorbital nerve. Before surgery (top), R1 components appear in both muscles on the stimulated side and R2 components are seen synchronously on both sides, thus also involving the contralateral orbicularis oris muscles (same phenomenon also occurs with stimulation on the left side). After surgery (bottom), synkinesis of the orbicularis oris muscles has disappeared. Arrow ⫽ time of stimulation. (From Schulze-Bonhage, A., and Ferbert, A.: Electrophysiological recordings in bilateral hemifacial spasm. J. Neurol. Neurosurg. Psychiatry 65:408, 1998, with permission.)

surgically, with compression by the anterior inferior cerebellar artery or vertebral artery occurring less commonly.112 At the time of follow-up 18 months after surgery, the majority of patients (40 of 45) had a good or excellent outcome in terms of resolution of spasm or facial weakness.112

A follow-up study by Jannetta111 in 1980 reviewed experience with a larger series of 229 patients with hemifacial spasm who had microvascular decompression, of whom 210 had arterial compression of the facial nerve noted at surgery, with a success rate of 88% after the initial surgery

1244

Diseases of the Peripheral Nervous System

(12 patients requiring a second operation and only 5 failures).111 The most recent review of the 20-year experience of these investigators with 782 patients with hemifacial spasm reported that 68.2% had compression of the facial nerve by the posterior inferior cerebellar artery. Of 612 patients followed for more than 1 year, 86% had an excellent outcome after a single surgical procedure at 1 month and 79% at 10 years.22 Auger et al.18 published the experience with microvascular decompression for hemifacial spasm in a series of 367 patients at the Mayo Clinic from 1975 to 1984, 54 of whom had the surgery, with the anterior inferior cerebellar artery most commonly compressing the facial nerve. At a mean follow-up of approximately 4 years, 91% had complete resolution of hemifacial spasm after the initial surgical procedure, either immediately or delayed up to 6 weeks later.18 Huang et al.97 reviewed treatment in Taiwan of 310 patients, with vascular compression seen in only 70% of patients (usually by the anterior inferior cerebellar artery), finding 88% with complete resolution after surgery and 12% with no improvement. The anterior inferior cerebellar artery has been reported as the most common blood vessel compressing the facial nerve at the root entry zone in additional series,19,63,72,176 yet others have reported that the posterior inferior cerebellar artery is most commonly compressing the nerve.100,134 Complications after microvascular decompression of the facial nerve include transient22,97,100,112,176 or mild residual18,19,22,100,111,134,176,239 facial weakness, dizziness,18,19,100,134 ataxia,176 brainstem ischemia,22 vocal cord palsy and dysphagia,239 and permanent hearing loss.18,22,97,100,111,112,134,176,239 Review of their long-term experience by Barker et al.22 showed that severe postoperative facial weakness or hearing loss was more common before 1980, after which intraoperative monitoring has been used. Hanakita and Kondo83 reported their experience with microvascular decompression for hemifacial spasm with intraoperative monitoring in 239 patients, noting complications of acute subdural hematoma, subarachnoid hemorrhage and intracranial hematoma, status epilepticus, and brainstem infarction, with an overall complication rate of 3.2%.83 Kondo126 reported a difference in postoperative hearing loss between patients who had surgery between 1976 and 1986 (patient-reported hearing loss rate of 8.9%) and those who had the procedure later, between 1987 and 1991 (patient-reported hearing loss of only 3.7%), a statistically significant difference. Kondo attributed this difference in complication rates to reduced cerebellar retraction during surgery and use of intraoperative monitoring. Moller and Moller157 reported postoperative hearing loss in 4 of 143 consecutive patients after microvascular decompression for hemifacial spasm (39 with auditory monitoring during surgery), 1 of whom had intraoperative monitoring; 24 of 137 patients had reduced hearing postoperatively. Early and abrupt loss of intraoperative brainstem auditory evoked potential responses associated with cerebellar retraction, without return by the end of the procedure, appears to

correlate with postoperative hearing loss following microvascular decompression of the facial nerve.211 Occasionally repeat microvascular decompression is performed because of lack of effect of the initial surgery or recurrence of hemifacial spasm, with mixed success.18,22 In the large series of 782 patients studied by Jannetta and colleagues, 49 had a second operation more than 30 days after the initial procedure for recurrence, with 50% with excellent results at 5-year follow-up and 22% with partial success.22 There are rare reports of recurrence more than 10 years after the initial microvascular decompression.22 Recurrence within a short period of time of the procedure has been shown on repeat operation to be due to reestablishment of the original arterial compression of the facial nerve as a result of shift of the prosthesis placed between the arterial loop and the nerve at the root entry zone.97,112 Long-term recurrence rates have been assessed. Paynor and Tew176 reported a retrospective review of 34 patients with hemifacial spasm and microvascular decompression surgery treated between 1976 and 1989 as well as a review of the literature and found the overall reported incidence of recurrence of hemifacial spasm to be 7%. In their own series, Paynor and Tew176 found that 29 of 34 patients, or 85%, had complete resolution of hemifacial spasm after surgery, 25 of 29 immediately following surgery and the remaining 4 between 3 months and 3 years later, with a small (1%) chance of recurrence based on their experience. Kondo126 sent a questionnaire to over 700 patients who had undergone microvascular decompression for hemifacial spasm to assess the patients’ report of cure rate and recurrence and found that the recurrence rate (more than 1 year after surgery) as reported by patients who underwent surgery between 1976 and 1986 was 8.9% compared to a reported recurrence rate of 6.9% among those who had surgery between 1987 and 1991. Lower rates of recurrence have been reported in other series, including 3.8% (3 of 78 patients followed for a mean of 8 years) in a prospective long-term follow-up study by Illingworth and co-workers in the United Kingdom in 1996,100 with recurrence of hemifacial spasm developing between 6 and 24 months postoperatively. Loeser and Chen134 reported a recurrence rate of 25% in their series (5 of 20 patients) and a recurrence rate of 4% for 450 cases from the available literature in 1983, although they noted that, in their patients with recurrence, the hemifacial spasm was mild and the symptoms were of a lesser severity than those preoperatively. Iwakuma et al.104 had a low rate of recurrence of hemifacial spasm after microvascular decompression (1 of 74 patients), but the length of followup was relatively short (1 month to 3 years). Electrophysiologic testing performed during and after microvascular decompression surgery in patients with hemifacial spasm has provided additional information about potential underlying pathophysiology. Auger and co-workers19 found that synkinesis on blink reflex testing,

Diseases of the Seventh Cranial Nerve

present preoperatively, disappeared postoperatively coincident with clinical recovery of hemifacial spasm. The lateral spread response or synkinesis persists with the patient under general anesthesia and is typically abolished intraoperatively during microvascular decompression. If the hemifacial spasm is mild, this electrophysiologic phenomenon may disappear intraoperatively with simply opening the dura mater or draining cerebrospinal fluid at the beginning of the surgical procedure.85 Moller and Jannetta156 also found that intraoperatively the ipsilateral blink reflex and synkinetic or lateral spread response disappear immediately with decompression of the facial nerve, and also noted in another study154 that allowing the offending artery to fall back onto the facial nerve caused the return of the abnormal lateral spread responses to electrical stimulation of peripheral facial nerve branches. Earlier these same investigators recorded action potentials from the orbicularis oculi and mentalis muscles and intracranial facial nerve intraoperatively during microvascular decompression, and stimulated the facial nerve at the root entry zone and peripheral facial branches as well as the supraorbital nerve.153 Measurement of the conduction times from the zygomatic branch to the root entry zone and orthodromically from the root entry zone to the mentalis muscle demonstrated that the mean latency of the mentalis response to stimulation of the zygomatic branch was 1.95 ms longer than the sum of the antidromic and orthodromic latencies from the mentalis muscle. They found that there was a late response recorded from the intracranial facial nerve associated with the lateral spread response, both of which disappeared with decompression of the nerve. These findings suggested that hemifacial spasm resulted from amplification of spontaneous activity from the facial nucleus at the site of compression or chronic abnormal antidromic electrical input from the lesion to the facial nucleus, leading to generation of increased spontaneous activity from the nucleus.153 Such kindling of the facial motor nucleus by antidromic neural activity from the site of nerve compression is one proposed theory to explain these electrophysiologic observations.155,156 Other investigators believe electrophysiologic studies in patients undergoing surgery support the theory of ephaptic transmission because of observed changes in direction of antidromic impulses postoperatively, as measured by collision and blocking electrical testing with recording of late responses.168 Other Treatments Injection of botulinum toxin into affected facial muscles in patients with hemifacial spasm has emerged as an alternative treatment to microvascular decompressive surgery. Jost and Kohl113 recently published a review of the available cases in the literature and found over 2000 patients in openlabeled studies, with less than 100 patients reported in double-blind, placebo-controlled trials. Based on these

1245

cases, good to excellent improvement in hemifacial spasm has been reported in 76% to 100% of patients, with a mean duration of action of the toxin for effect of between 2.6 and 4 months on average. Side effects reported included transient symptoms of dry eyes (7% to 18.1%), ipsilateral ptosis (2.8% to 23.3%), facial weakness (17.6% to 97%), tearing (5.5%), and diplopia (⬍1% to 6%).113 Duration of beneficial medication effect does not seem to change with repeated injections.47,56,65 Defazio and co-workers47 retrospectively reviewed 65 patients who had received botulinum toxin injections of the ipsilateral orbicularis oculi for hemifacial spasm over a 10-year period and found that response rate and duration of effect did not change for these patients over that time period, with the cumulative yearly dose per patient remaining the same and an overall response rate of 96% to the therapy (based on patient self-evaluation scale, not an observer rating). Electrophysiologic testing in patients receiving botulinum toxin shows continued spontaneously bursting motor unit potentials in affected facial muscles, but of lower amplitudes, despite absence of clinically observable spasm as well as synkinesis with blink reflex testing.74 This finding is not surprising in light of the action of the toxin at the neuromuscular junction, without potential effect on proximal nerve excitability.74 The reader is referred to an available review of the history and use of botulinum toxin for treatment of hemifacial spasm by Jankovic.110 Prior to the development of microvascular decompression and botulinum toxin injections, other surgical procedures were used to try to relieve hemifacial spasm, including transtympanic facial nerve needling,171 microvascular decompression with facial nerve wrapping,63,117 partial sectioning of the facial nerve trunk distal to the stylomastoid foramen,104,204 selective neurectomy of facial peripheral facial nerve branches,23,104,202 percutaneous radiofrequency coagulation of the facial nerve at the stylomastoid foramen,95 unilateral periorbital myectomy,73 and selective facial nerve block with phenol injection,230 each with variable results and complications. Others injected the facial nerve at the stylomastoid foramen with local anesthetic or alcohol or injected the infraorbital nerve to reduce any sensory trigger for hemifacial spasm.202 The reader is referred to an available review of surgical treatment of hemifacial spasm other than microvascular decompression.52 Medical treatment of hemifacial spasm has had limited success. Carbamazepine,12,62,146,200,208,239 gabapentin,42 clonazepam,62,89,146 baclofen,23 and diazepam104,146 have all been reported to provide some benefit, although side effects limited their utility in most cases.89,104,146 Wang and Jankovic239 recommended using medical therapy only for mild cases of hemifacial spasm, because of the modest potential benefit. Combination of botulinum toxin injections and medical therapy was reported to have good benefit in a retrospective study of 119 patients, although many patients selected botulinum toxin injections as their primary therapy.146

1246

Diseases of the Peripheral Nervous System

GENICULATE NEURALGIA Geniculate neuralgia, as a manifestation of facial nerve sensory dysfunction, is rare. Patients complain of severe paroxysmal pain localized to the ear185,246 with radiation into the cheek and ipsilateral occipital region.246 Pulec185 recently published a retrospective analysis of 64 patients diagnosed with geniculate neuralgia between 1966 and 1996. These patients were treated with surgical excision of the nervus intermedius, geniculate ganglion, and greater superficial petrosal nerve with excellent reported outcomes in terms of reduction of ear pain and lack of permanent ipsilateral hearing deficit.185 Operative complications included partial, transient ipsilateral facial paralysis and dry eye.185 Geniculate neuralgia associated with hemifacial spasm has been described in a single case report in the literature, with symptomatic relief obtained by surgical microvascular decompression of the facial nerve, suggesting vascular compression of both the sensory and motor trunks of the facial nerve at the brainstem root entry zone.246

REFERENCES 1. Adour, K. K.: Medical management of idiopathic (Bell’s) palsy. Otolaryngol. Clin. N. Am. 24:663, 1991. 2. Adour, K. K., Bell, D. N., and Hilsinger, R. L. Jr.: Herpes simplex virus in idiopathic facial paralysis (Bell palsy). JAMA 233:527, 1975. 3. Adour, K. K., Bell, D. N., and Wingerd, J.: Bell palsy: Dilemma of diabetes mellitus. Arch. Otolaryngol. 99:114, 1974. 4. Adour, K. K., Boyajian, J. A., Kahn, Z. M., and Schneider, G. S.: Surgical and nonsurgical management of facial paralysis following closed head injury. Laryngoscope 87:380, 1977. 5. Adour, K. K., Byl, F. M., Hilsinger, R. L. Jr., and Kahn, Z. M.: The true nature of Bell’s palsy: analysis of 1000 consecutive patients. Laryngoscope 88:787, 1978. 6. Adour, K. K., Ruboyianes, J. M., von Doersten, P. G., et al.: Bell’s palsy treatment with acyclovir and prednisone compared with prednisone alone: a double-blind, randomized, controlled trial. Ann. Otol. Rhinol. Laryngol. 105:371, 1996. 7. Adour, K. K., and Wingerd, J.: Idiopathic facial paralysis (Bell’s palsy): factors affecting severity and outcome in 446 patients. Neurology 24:1112, 1974. 8. Adour, K. K., Wingerd, J., Bell, D. N., et al.: Prednisone treatment for idiopathic facial paralysis (Bell’s palsy). N. Engl. J. Med. 287:1268, 1972. 9. Albisetti, M., Schaer, G., Good, M., et al.: Diagnostic value of cerebrospinal fluid examination in children with peripheral facial palsy and suspected Lyme borreliosis. Neurology 49:817, 1997. 10. Aldrich, M. S., Beck, R. W., and Albers, J. W.: Familial recurrent Bell’s palsy with oculomotor palsies. Neurology 37:1369, 1987.

11. Aleksic, S. N., Budzilovich, G. N., and Liberman, A. N.: Herpes zoster oticus and facial paralysis (Ramsey Hunt syndrome): clinico-pathologic study and review of literature. J. Neurol. Sci. 20:149, 1973. 12. Alexander, G. E., and Moses, H. III: Carbamazepine for hemifacial spasm. Neurology 32:286, 1982. 13. Alexander, R. W., and James, R. B.: Melkersson-Rosenthal syndrome: review of literature and report of case. J. Oral Surg. 30:599, 1972. 14. Alter, M.: Familial aggregation of Bell’s palsy. Arch. Neurol. 8:107, 1963. 15. Ang, K. L., and Jones, N. S.: Melkersson-Rosenthal syndrome. J. Laryngol. Otol. 116:386, 2002. 16. Angerer, M., Pfadenhauer, K., and Stohr, M.: Prognosis of facial palsy in Borrelia burgdorferi meningopolyradiculoneuritis. J. Neurol. 240:319, 1993. 17. Auger, R. G.: Hemifacial spasm: clinical and electrophysiologic observations. Neurology 29:1261, 1979. 18. Auger, R. G., Piepgras, D. G., and Laws, E. R. Jr.: Hemifacial spasm: results of microvascular decompression of the facial nerve in 54 patients. Mayo Clin. Proc. 61:640, 1986. 19. Auger, R. G., Piepgras, D. G., Laws, E. R. Jr., and Miller, R. H.: Microvascular decompression of the facial nerve for hemifacial spasm: clinical and electrophysiologic observations. Neurology 31:346, 1981. 20. Auger, R. G., and Whisnant, J. P.: Hemifacial spasm in Rochester and Olmsted County, Minnesota 1960 to 1984. Arch. Neurol. 47:1233, 1990. 21. Austin, J. R., Peskind, S. P., Austin, S. G., and Rice, D. H.: Idiopathic facial nerve paralysis: a randomized, double blind controlled trial of placebo versus prednisone. Laryngoscope 103:1326, 1993. 22. Barker, F. G. II, Jannetta, P. J., Bissonette, D. J., et al.: Microvascular decompression for hemifacial spasm. J. Neurosurg. 82:201, 1995. 23. Bauer, C. A., and Coker, N. J.: Update on facial nerve disorders. Otolaryngol. Clin. N. Am. 29:445, 1996. 24. Belec, L., Gherardi, R., Georges, A. J., et al.: Peripheral facial paralysis and HIV infection: report of 4 African cases and review of the literature. J. Neurol. 236:410, 1989. 25. Belman, A. L., Iyer, M., Coyle, P. K., and Dattwyler, R.: Neurologic manifestations in children with North American Lyme disease. Neurology 43:2609, 1993. 26. Benecke, J. E. Jr.: Facial paralysis. Otolaryngol. Clin. N. Am. 35:357, 2002. 27. Berrettini, S., Bianchi, M. C., Segnini, G., et al.: Herpes zoster oticus: correlations between clinical and MRI findings. Eur. Neurol. 39:26, 1998. 28. Blunt, M. J.: The blood supply of the facial nerve. J. Anat. 88:520. 29. Boudreau, J. C., Oravec, J., White, T. D., et al.: Geniculate neuralgia and facial nerve sensory systems. Arch. Otolaryngol. 103:473, 1977. 30. Brandle, P., Satoretti-Schefer, S., Bohner, A., et al.: Correlation of MRI, clinical and electroneuronographic findings in acute facial nerve palsy. Am. J. Otol. 17:154, 1996. 31. Bushara, K., Wilson, A., and Rust, R. S.: Facial palsy in Kawasaki syndrome. Pediatr. Neurol. 17:362, 1997.

Diseases of the Seventh Cranial Nerve 32. Carpenter, M. B., and Sutin, J.: Human Neuroanatomy, 8th ed. Baltimore, Williams & Wilkins, p. 385, 1983. 33. Carter, J. B., Patrinely, J. R., Jankovic, J., et al.: Familial hemifacial spasm. Arch. Ophthalmol. 108:249, 1990. 34. Castillo, M.: Imaging of the upper cranial nerves I, III-VIII, and the cavernous sinuses. Magn. Reson. Imaging Clin. N. Am. 10:415, 2002. 35. Chang, C. Y. J., and Cass, S. P.: Management of facial nerve injury due to temporal bone trauma. Am. J. Otol. 20:96, 1999. 36. Clark, J. R., Carlson, R. D., Pachner, A. R., et al.: Facial paralysis in Lyme disease. Laryngoscope 95:1341, 1985. 37. Coad, J. E., Wirtschafter, J. D., Haines, S. J., et al.: Familial hemifacial spasm associated with arterial compression of the facial nerve. J. Neurosurg. 74:290, 1991. 38. Coker, N. J.: Facial electroneurography: analysis of techniques and correlation with degenerating motoneurons. Laryngoscope 102:747, 1992. 39. Colover, J.: Sarcoidosis with involvement of the nervous system. Brain 71:451, 1948. 40. Cook, S. P., Macartney, K. K., Rose, C. D., et al.: Lyme disease and seventh nerve paralysis in children. Am. J. Otolaryngol. 18:320, 1997. 41. Cramer, H. B., and Kartush, J. M.: Testing facial nerve function. Otolarygol. Clin. N. Am. 24:555, 1991. 42. Daniele, O., Caravaglios, G., Marchini, C., et al.: Gabapentin in the treatment of hemifacial spasm. Acta Neurol. Scand. 104:110, 2001. 43. Daoud, M. S., and Rogers, R. S. III: Melkersson-Rosenthal syndrome. Semin. Dermatol. 14:135, 1995. 44. Davis, W. E., Luterman, B. F., Pulliam, M. W., and Templer, J. W.: Hemifacial spasm caused by cholesteatoma. Am. J. Otol. 2:272, 1981. 45. De Diego, J. I., Prim, M. P., De Sarria, M. J., et al.: Idiopathic facial paralysis: a randomized, prospective and controlled study using single-dose prednisone versus acyclovir three times daily. Laryngoscope 108:573, 1998. 46. De Diego, J. I., Prim, M. P., Madero, R., and Gavilan, J.: Seasonal patterns of idiopathic facial paralysis: a 16-year study. Otolaryngol. Head Neck Surg. 120:269, 1999. 47. Defazio, G., Abbruzzese, G., Girlanda, P., et al.: Botulinum toxin A treatment for primary hemifacial spasm. Arch. Neurol. 59:418, 2002. 48. Defazio, G., Berardelli, A., Abbruzzese, G., et al.: Primary hemifacial spasm and arterial hypertension: a multicenter case-control study. Neurology 54:1198, 2000. 49. Devriese, P. P., and Moesker, W. H.: The natural history of facial paralysis in herpes zoster. Clin. Otolaryngol. 13:289, 1988. 50. Devriese, P. P., Schumacher, T., Scheide, A., et al.: Incidence, prognosis and recovery of Bell’s palsy: a survey of about 1000 patients (1974–1983). Clin. Otolaryngol. 15:15, 1990. 51. Dickins, J. R. E., Smith, J. T., and Graham, S. S.: Herpes zoster oticus: treatment with intravenous acyclovir. Laryngoscope 98:776, 1988. 52. Digre, K., and Corbett, J. J.: Hemifacial spasm: differential diagnosis, mechanism and treatment. Adv. Neurol. 49:151, 1988. 53. Dudley, J. P., and Goodman, M.: Periarteritis nodosa and bilateral facial paralysis. Arch. Otolaryngol. 90:47, 1969.

1247

54. Duff, J. M., Spinner, R. J., Lindor, N. M., et al.: Familial trigeminal neuralgia and contralateral hemifacial spasm. Neurology 53:216, 1999. 55. Dutt, S. N., Mirza, S., Irving, R. M., and Donalson, I.: Total decompression of facial nerve for Melkersson-Rosenthal syndrome. J. Laryngol. Otol. 114:870, 2000. 56. Dutton, J. J., and Buckley, E. G.: Long-term results and complications of botulinum A toxin in the treatment of blepharospasm. J. Ophthalmol. 95:1529, 1988. 57. Ehni, G., and Woltman, H. W.: Hemifacial spasm: review of 106 cases. Arch. Neurol. Psychiatry 53:205, 1945. 58. Eidlitz-Markus, T., Gilai, A., Mimouni, M., and Shuper, A.: Recurrent facial nerve palsy in paediatric patients. Eur. J. Pediatr. 160:659, 2001. 59. Ernster, J. A.: Bilateral facial nerve paralysis associated with Mycoplasma pneumoniae infection. Ear Nose Throat J. 63:585, 1984. 60. Eshraghi, A. A., Buchman, C. A., and Telischi, F. F.: Sensory auricular branch of the facial nerve. Otol. Neurotol. 23:393, 2002. 61. Esteban, A., and Molina-Negro, P.: Primary hemifacial spasm: a neurophysiological study. J. Neurol. Neurosurg. Psychiatry 49:58, 1986. 62. Evidente, V. G. H., and Adler, C. H.: Hemifacial spasm and other craniofacial movement disorders. Mayo Clin. Proc. 73:67, 1998. 63. Fabinyi, G. C. A., and Adams, C. B. T.: Hemifacial spasm: treatment by posterior fossa surgery. J. Neurol. Neurosurg. Psychiatry 41:829, 1978. 64. Ferguson, J. H.: Hemifacial spasm and the facial nucleus. Ann. Neurol. 4:97, 1978. 65. Flanders, M., Chin, D., and Boghen, D.: Botulinum toxin: preferred treatment for hemifacial spasm. Eur. Neurol. 33:316, 1993. 66. Friedman, A., Jamrozik, Z., and Bojakowski, J.: Familial hemifacial spasm. Mov. Disord. 4:213, 1989. 67. Friedman, J. H., Challenor, Y., and Brust, J. C. M.: Bilateral facial palsy associated with Stevens-Johnson syndrome. Neurology 29:1304, 1979. 68. Furata, Y., Fukuda, S., Chida, E., et al.: Reactivation of herpes simplex virus type 1 in patients with Bell’s palsy. J. Med. Virol. 54:162, 1998. 69. Gacek, R. R.: Neuroanatomy of the nerve in the temporal bone. Adv. Otorhinolaryngol. 60:12, 2002. 70. Galvez-Jimenez, N., Hanson, M. R., and Desai, M.: Unusual causes of hemifacial spasm. Semin. Neurol. 21:75, 2001. 71. Gantz, B. J., Rubinstein, J. T., Gidley, P., and Woodworth, G. G.: Surgical management of Bell’s palsy. Laryngoscope 109:1177, 1999. 72. Gardner, W. J., and Sava, G. A.: Hemifacial spasm—a reversible pathophysiologic state. J. Neurosurg. 19:240, 1962. 73. Garland, P. E., Patrinely, J. R., and Anderson, R. L.: Hemifacial spasm: results of unilateral myectomy. J. Ophthalmol. 94:288, 1987. 74. Geller, B. D., Hallett, M., and Ravits J.: Botulinum toxin therapy in hemifacial spasm: clinical and electrophysiologic studies. Muscle Nerve 12:716, 1989. 75. Gillman, G. S., Schaitkin, B. M., May, M., and Klein, S. R.: Bell’s palsy in pregnancy: a study of recovery outcomes. Otolaryngol. Head Neck Surg. 126:26, 2002.

1248

Diseases of the Peripheral Nervous System

76. Greene, R. M., and Rogers, R. S. III: Melkersson-Rosenthal syndrome: a review of 36 patients. J. Am. Acad. Dermatol. 21:1263, 1989. 77. Grogan, P. M., and Gronseth, G. S.: Practice parameter: steroids, acyclovir, and surgery for Bell’s palsy (an evidencebased review). Neurology 56:830, 2001. 78. Guerrissi, J. O.: Facial nerve paralysis after intratemporal and extratemporal blunt trauma. J. Craniofacial Surg. 8:431, 1997. 79. Hadithi, M., Stam, F., Donker, A. J. M., and Dijkmans, B. A. C.: Sjogren’s syndrome: an unusual cause of Bell’s palsy. Ann. Rheum. Dis. 60:724, 2001. 80. Hageman, G., Ippel, P. F., Jansen, E. N. H., and Rozeboom, A. R.: Familial, alternating Bell’s palsy with dominant inheritance. Eur. Neurol. 30:310, 1990. 81. Halperin, J. J.: Facial nerve palsy associated with Lyme disease. Muscle Nerve 28:516, 2003. 82. Halperin, J. J., and Golightly, M., for the Long Island Neuroborreliosis Collaborative Study Group: Lyme borreliosis in Bell’s palsy. Neurology 42:1268, 1992. 83. Hanakita, J., and Kondo, A.: Serious complications of microvascular decompression operations for trigeminal neuralgia and hemifacial spasm. Neurosurgery 22:348, 1988. 84. Harada, Y., Kawamata, S., Suzuki, M., et al.: Electron microscopic observation of the human facial nerve. Acta Otolaryngol. Suppl. (Stock.) 446:145, 1988. 85. Harper, C. M., and Daube, J. R.: Facial nerve electromyography and other cranial nerve monitoring. J. Clin. Neurophysiol. 15:206, 1998. 86. Hattori, T., and Schlagenhauff, R. E.: Bilateral facial palsy: occurrence with diabetes mellitus. N. Y. State J. Med. 77:1492, 1977. 87. Hauser, W. A., Karnes, W. E., Annis, J., and Kurland, L. T.: Incidence and prognosis of Bell’s palsy in the population of Rochester, Minnesota. Mayo Clin. Proc. 46:258, 1971. 88. Heine, C., Klingebiel, R., and Lehmann, R.: Threedimensional MR visualization of the intracisternal course of the cranial nerves V-VIII by virtual cisternoscopy. Acta Radiol. 43:242, 2002. 89. Herzberg, L.: Management of hemifacial spasm with clonazepam. Neurology 35:1676, 1985. 90. Hill, M. D., Midroni, G, Goldstein, W. C, et al.: The spectrum of electrophysiological abnormalities in Bell’s palsy. Can. J. Neurol. Sci. 28:130, 2001. 91. Hilsinger, R. L., Adour, K. K., and Doty, H. E.: Idiopathic facial paralysis, pregnancy and the menstrual cycle. Ann. Otol. 84:433, 1975. 92. Holds, J. B., Anderson, R. L., Jordan, D. R., and Patrinely, J. R.: Bilateral hemifacial spasm. J. Clin. Neuroophthal. 10:153, 1990. 93. Holla, S. P., Smith, R. R., and Sanford, R. A.: Bilateral traumatic facial paralysis. Neurosurgery 6:290, 1980. 94. Honda, N., Yangihara, N., Hato, N., et al.: Swelling of intratemporal facial nerve in Ramsey Hunt syndrome. Acta Otolaryngol. (Stock.) 122:348, 2002. 95. Hori, T., Fukushima, T., Terao, H., et al.: Percutaneous radiofrequency facial nerve coagulation in the management of facial spasm. J. Neurosurg. 54:655, 1981. 96. Horowitz, S. H.: Hemifacial spasm and facial myokymia: electrophysiological findings. Muscle Nerve 10:422, 1987.

97. Huang, C.-I., Chen, I.-H., and Lee, L.-S.: Microvascular decompression for hemifacial spasm: analysis of operative findings and results in 310 patients. Neurosurgery 30:53, 1992. 98. Hughes, G. B.: Prognostic tests in acute facial palsy. Am. J. Otol. 10:304, 1989. 99. Hyden, D., Roberg, M., Forsberg, P., et al.: Acute “idiopathic” peripheral facial palsy: clinical, serological, and cerebrospinal fluid findings and effects of corticosteroids. Am. J. Otolaryngol. 14:179, 1993. 100. Illingworth, R. D., Porter, D. G., and Jakubowski, J.: Hemifacial spasm: a prospective, long-term follow-up of 83 cases treated by microvascular decompression at two neurosurgical centres in the United Kingdom. J. Neurol. Neurosurg. Psychiatry 60:72, 1996. 101. Inamura, H., Aoyagi, M., Tojima, H., et al.: Facial nerve palsy in children: clinical aspects of diagnosis and treatment. Acta Otolarygol. Suppl. (Stock.) 511:150, 1994. 102. Ishikawa, M., Namiki, J., Takase, M., et al.: Effect of repetitive stimulation on lateral spreads and F waves in hemifacial spasm. J. Neurol. Sci. 142:99, 1996. 103. Ishikawa, M., Ohira, T., Namiki, J., et al.: Abnormal muscle response (lateral spread) and F-wave in patients with hemifacial spasm. J. Neurol. Sci. 137:109, 1996. 104. Iwakuma, T., Matsumoto, A., and Nakamura, N.: Hemifacial spasm: comparison of three different operative procedures in 110 patients. J. Neurosurg. 57:753, 1982. 105. Iwata, N., Kitai, S. T., and Olson, S.: Afferent component of the facial nerve: its relation to the spinal trigeminal and facial nucleus. Brain Res. 43:662, 1972. 106. Jackson, C. G., and von Doersten, P. G.: The facial nerve: current trends in diagnosis, treatment and rehabilitation. Med. Clin. N. Am. 83:179, 1999. 107. Jager, L., and Reiser, M.: CT and MR imaging of the normal and pathologic conditions of the facial nerve. Eur. J. Radiol. 40:133, 2001. 108. James, D. G.: All that palsies is not Bell’s. J. Royal Soc. Med. 89:184, 1996. 109. James, D. G., and Sharma, O. P.: Neurological complications of sarcoidosis. Proc. R. Soc. Med. 60:1169, 1967. 110. Jankovic, J.: Botulinum toxin in movement disorders. Curr. Opin. Neurol. 7:358, 1994. 111. Jannetta, P. J.: Neurovascular compression in cranial nerve and systemic disease. Ann. Surg. 192:518, 1980. 112. Jannetta, P. J., Abbasy, M., Maroon, J. C., et al.: Etiology and definitive microsurgical treatment of hemifacial spasm. J. Neurosurgery 47:321, 1977. 113. Jost, W. H., and Kohl, A.: Botulinum toxin: evidence-based medicine criteria in blepharospasm and hemifacial spasm. J. Neurol. 248(Suppl. 1):21, 2001. 114. Jungehuelsing, M., Sittel, C., Fischbach, R., et al.: Limitations of magnetic resonance imaging in the evaluation of perineural tumor spread causing facial nerve paralysis. Arch. Otolaryngol. Head Neck Surg. 126:506, 2000. 115. Kania, R. E., Cazals-Hatem, D., Bouccara, D., et al.: Hypertrophic neuropathy of the facial nerve. Ann. Otol. Rhinol. Laryngol. 110:257, 2001.

Diseases of the Seventh Cranial Nerve 116. Katusic, S. K., Beard, C. M., Wiederholt, W. C., et al.: Incidence, clinical features and prognosis in Bell’s palsy, Rochester, Minnesota, 1968–1982. Ann. Neurol. 20:622, 1986. 117. Kaye, A. H., and Adams, C. B. T.: Hemifacial spasm: a long term follow-up of patients treated by posterior fossa surgery and facial nerve wrapping. J. Neurol. Neurosurg. Psychiatry 44:1100, 1981. 118. Keane, J. R.: Bilateral seventh nerve palsy: analysis of 43 cases and review of the literature. Neurology 44:1198, 1994. 119. Kempe, L. G.: Topical organization of the distal portion of the facial nerve. J. Neurosurg. 52:671, 1980. 120. Kerbavaz, R. J., Hilsinger, R. L. Jr., and Adour, K. K.: The facial paralysis prognostic index. Otolaryngol. Head Neck Surg. 91:284, 1983. 121. Kim, P., and Fukushima, T.: Observations on synkinesis in patients with hemifacial spasm. J. Neurosurg. 60:821, 1984. 122. Kimura, J., Rodnitzky, R. L., and Okawara, S.-H.: Electrophysiologic analysis of aberrant regeneration after facial nerve paralysis. Neurology 25:989, 1975. 123. Kiwak, K. J., and Levine, S. E.: Benign intracranial hypertension and facial diplegia. Arch. Neurol. 41:787, 1984. 124. Knox, G. W.: Treatment controversies in Bell palsy. Arch. Otolaryngol. Head Neck Surg. 124:821, 1998. 125. Kohsyu, H., Aoyagi, M., Tojima, H., et al.: Facial nerve enhancement in Gd-MRI in patients with Bell’s palsy. Acta Otolaryngol. Suppl. (Stock.) 511:165, 1994. 126. Kondo, A.: Follow-up results of microvascular decompression in trigeminal neuralgia and hemifacial spasm. Neurosurgery 40:46, 1997. 127. Kosary, I. Z., Shaeked, I. J., Ouakrine, G., and Braham, J.: Continuous facial spasm with tumor of the pons. J. Neurosurg. 33:212, 1970. 128. Krishnamurthy, S. N., Weinstock, A. L., Smith, S. H., and Duffner, P. K.: Facial palsy, an unusual presenting feature of childhood leukemia. Pediatr. Neurol. 27:68, 2002. 129. Kuiper, H., Devriese, P. P., de Jongh, B. M., et al.: Absence of Lyme borreliosis among patients with presumed Bell’s palsy. Arch. Neurol. 49:940, 1992. 130. Kumagami, H.: Neuropathological findings of hemifacial spasm and trigeminal neuralgia. Arch. Otolaryngol. 99:160, 1974. 131. Langston, J. W., and Tharp, B. R.: Infantile hemifacial spasm. Arch. Neurol. 33:302, 1976. 132. Lescanne, E., Velut, S., Lefranco, T., and Destrieux, C.: The internal acoustic meatus and its meningeal layers: a microanatomical study. J. Neurosurg. 97:1197, 2002. 133. Liston, S. L., and Kleid, M. S.: Histopathology of Bell’s palsy. Laryngoscope 99:23, 1989. 134. Loeser, J. D., and Chen, J.: Hemifacial spasm: treatment by microsurgical facial nerve decompression. Neurosurgery 13:141, 1983. 135. Lopez-Lozano, J. J., Garcia-Merino, J. A., and Liano, H.: Bilateral facial paralysis secondary to trichinosis. Acta Neurol. Scand. 78:194, 1988. 136. Maeyama, H., Aoyagi, M., Tojima, H., et al.: Electrophysiological study on the pathology of synkinesis after facial nerve paralysis. Acta Otolaryngol. Suppl. (Stock.) 511:161, 1994.

1249

137. Magliulo, G., Varacalli, S., and Sepe, C.: Wegener’s granulomatosis presenting as facial palsy. Am. J. Otolaryngol. 20:43, 1999. 138. Mallya, K. B., Mendis, T., and Guberman, A.: Bilateral facial paralysis following ethylene glycol ingestion. Can. J. Neurol. Sci. 13:340, 1986. 139. Malone, D. G., O’Boynick, P. L., Ziegler, D. K, et al.: Osteomyelitis of the skull bone. Neurosurgery 30:426, 1992. 140. Maroon, J. C., Lunsford, D., and Deeb, Z. L.: Hemifacial spasm due to aneurysmal compression of the facial nerve. Arch. Neurol. 35:545, 1978. 141. Marra, C. M.: Bell’s palsy and HSV-1 infection. Muscle Nerve 22:1476, 1999. 142. Marzo, S. J., Leonetti, J. P., and Petruzzelli, G.: Facial paralysis caused by malignant skull base neoplasms. Ear Nose Throat J. 81:845, 2002. 143. Massey, E. W., and Massey, J. M.: Facial diplegia due to amyloidosis. South. Med. J. 79:1458, 1986. 144. Matsumoto, Y., Pulec, J. L., Patterson, M. J., and Yanagihara, N.: Facial nerve biopsy for etiologic clarification of Bell’s palsy. Ann. Otol. Rhinol. Laryngol. 97:22, 1988. 145. Matsuura, N., and Kondo, A.: Trigeminal neuralgia and hemifacial spasm as false localizing signs in patients with a contralateral mass of the posterior cranial fossa. J. Neurosurg. 84:1067, 1996. 146. Mauriello, J. A. Jr., Leone, T., Dhillon, S., et al.: Treatment choices of 119 patients with hemifacial spasm over 11 years. Clin. Neurol. Neurosurg. 98:213, 1996. 147. May, M., and Klein, S. R.: Differential diagnosis of facial nerve palsy. Otolaryngol. Clin. N. Am. 24:613, 1991. 148. McArthur, J. C.: Neurologic manifestations of AIDS. Medicine 66:407, 1987. 149. Metson, R., and Hanson, D. G.: Bilateral facial nerve paralysis following arterial embolization for epistaxis. J. Otolaryngol. Head Neck Surg. 91:299, 1983. 150. Miwa, H., Mizuno, Y., and Kondo, T.: Familial hemifacial spasm: report of cases and review of literature. J. Neurol. Sci. 193:97, 2002. 151. Moller, A. R.: Hemifacial spasm: ephaptic transmission or hyperexcitability of the facial motor nucleus? Exp. Neurol. 98:110, 1987. 152. Moller, A. R., and Jannetta, P. J.: On the origin of synkinesis in hemifacial spasm: results of intracranial recordings. J. Neurosurg. 61:569, 1984. 153. Moller, A. R., and Jannetta, P. J.: Hemifacial spasm: results of electrophysiologic recording during microvascular decompression operations. Neurology 35:969, 1985. 154. Moller, A. R., and Jannetta, P. J.: Microvascular decompression in hemifacial spasm: intraoperative electrophysiological observations. Neurosurgery 16:612, 1985. 155. Moller, A. R., and Jannetta, P. J.: Blink reflex in patients with hemifacial spasm: observations during microvascular decompression operations. J. Neurol. Sci. 72:171, 1986. 156. Moller, A. R., and Jannetta, P. J.: Physiological abnormalities in hemifacial spasm studied during microvascular decompression operations. Exp. Neurol. 93:584, 1986. 157. Moller, M. B., and Moller, A. R.: Loss of auditory function in microvascular decompression for hemifacial spasm. J. Neurosurg. 63:17, 1985.

1250

Diseases of the Peripheral Nervous System

158. Montagna, P., Imbriaco, A., Zucconi, M., et al.: Hemifacial spasm in sleep. Neurology 36:270, 1986. 159. Morgan, M., and Nathwani, D.: Facial palsy and infection: the unfolding story. Clin. Infect. Dis. 14:263, 1992. 160. Mulkens, P. S. J. Z., Bleeker, J. D., and Schroder, F. P.: Acute facial paralysis: a virological study. Clin. Otolaryngol. 5:303, 1980. 161. Murakami, S., Mizobuchi, M., Nakashiro, Y., et al.: Bell palsy and herpes simplex virus: identification of viral DNA in endoneurial fluid and muscle. Ann. Intern. Med. 124:27, 1996. 162. Murphy, T. P.: MRI of the facial nerve during paralysis. Otolaryngol. Head Neck Surg. 104:47, 1991. 163. Murr, A. H., and Benecke, J. E. Jr.: Association of facial paralysis with HIV positivity. Am. J. Otol. 12:450, 1991. 164. Nager, G. T., and Proctor, B.: Anatomic variations and anomalies involving the facial canal. Otolaryngol. Clin. N. Am. 24:531, 1991. 165. Neely, J. G.: Neoplastic involvement of the facial nerve. Otolaryngol. Clin. N. Am. 7:385, 1974. 166. Nielsen, V. K.: Pathophysiology of hemifacial spasm: I. Ephaptic transmission and ectopic excitation. Neurology 34:418, 1984. 167. Nielsen, V. K.: Pathophysiology of hemifacial spasm: II. Lateral spread of the supraorbital nerve reflex. Neurology 34:427, 1984. 168. Nielsen, V. K.: Electrophysiology of the facial nerve in hemifacial spasm: ectopic/ephaptic excitation. Muscle Nerve 8:545, 1985. 169. Nishi, T., Matsukado, Y., Nagahiro, S., et al.: Hemifacial spasm due to contralateral acoustic neuroma: case report. Neurology 37:339, 1987. 170. Nussbaum, M.: Hemifacial spasm associated with benign parotid tumor. Ann. Otol. 86:73, 1977. 171. Ogale, S. B., Chopra, S., and Thakur, S.: Transtympanic facial nerve needling: a curative surgical option for hemifacial spasms. Acta Otolaryngol. (Stock.) 115:405, 1995. 172. Oliveira, L. D., Cardoso, F., and Vargas, A. P.: Hemifacial spasm and arterial hypertension. Mov. Disord. 14:832, 1999. 173. Orlando, M. R., and Atkins, J. S. Jr.: Melkersson-Rosenthal syndrome. Arch. Otolaryngol. Head Neck Surg. 116:728, 1990. 174. Orobello, P.: Congenital and acquired facial nerve paralysis in children. Otolaryngol. Clin. N. Am. 24:647, 1991. 175. Pachner, A. R., and Steere, A. C.: The triad of neurologic manifestations of Lyme disease: meningitis, cranial neuritis and radiculoneuritis. Neurology 35:47, 1985. 176. Paynor, T. D., and Tew, J. M. Jr.: Recurrence of hemifacial spasm after microvascular decompression. Neurosurgery 38:686, 1996. 177. Peitersen, E.: The natural history of Bell’s palsy. Am. J. Otology 4:107, 1982. 178. Perry, J. R., and Hasso, A. N.: Magnetic resonance imaging of cranial nerve VII. Top. Magn. Reson. Imaging 8:155, 1996. 179. Phillips, C. D., and Bubash, L. A.: The facial nerve: anatomy and common pathology. Semin. Ultrasound CT MR 23:202, 2002. 180. Pierry, A., and Cameron, M.: Clonic hemifacial spasm from posterior fossa arteriovenous malformation. J. Neurol. Neurosurg. Psychiatry 42:670, 1979.

181. Pitts, D. B., Adour, K. K., and Hilsinger, R. L. Jr.: Recurrent Bell’s palsy: analysis of 140 patients. Laryngoscope 98:535, 1988. 182. Poloni, T. E., Merlo, I. M., Alfonsi, E., et al.: Facial nerve is liable to pressure palsy. Neurology 51:320, 1998. 183. Price, T., and Fife, D. G.: Bilateral simultaneous facial nerve palsy. J. Laryngol. Otol. 116:46, 2002. 184. Proctor, B.: The anatomy of the facial nerve. Otolaryngol. Clin. N. Am. 24:479, 1991. 185. Pulec, J. L.: Geniculate neuralgia: long-term results of surgical treatment. Ear Nose Throat J. 81:30, 2002. 186. Punal, J. E., Angueira, F. B., Lorenzo, A. V., and CastroGago, M.: Three new patients with congenital unilateral facial nerve palsy due to chromosome 22q11 deletion. J. Child Neurol. 16:450, 2001. 187. Qiu, W. W., Yin, S. S., Stucker, F. J., et al.: Time course of Bell palsy. Arch. Otolaryngol. Head Neck Surg. 122:967, 1996. 188. Ramsey, M. J., DerSimonian, R., Holtel, M. R., and Burgess, P. A.: Corticosteroid treatment for idiopathic facial nerve paralysis: a meta-analysis. Laryngoscope 110:335, 2000. 189. Rice, J. P., Horowitz, M., and Chin, D.: Wernicke-Korsakoff syndrome with bilateral facial nerve palsies. J. Neurol. Neurosurg. Psychiatry 47:1356, 1984. 190. Rogers, R. S. III: Melkersson-Rosenthal syndrome and orofacial granulomatosis. Dermatol. Clin. 14:371, 1996. 191. Rontal, E., and Sigal, M. E.: Bilateral facial paralysis. Laryngoscope 82:607, 1972. 192. Roob, G., Fazekas, F., and Hartung, H.-P.: Peripheral facial palsy: etiology, diagnosis and treatment. Eur. Neurol. 41:3, 1999. 193. Rosso, A. L. Z., Mattos, J. P., Fogel, L. M., and Novis, S. A. P.: Bilateral hemifacial spasm. Mov. Disord. 9:236, 1994. 194. Ruby, J. R., and Jannetta, P. J.: Hemifacial spasm: ultrastructural changes in the facial nerve induced by neurovascular compression. Surg. Neurol. 4:369, 1975. 195. Saito, O., Aoyagi, M., Tojima, H., and Koike, Y.: Diagnosis and treatment for Bell’s palsy associated with diabetes mellitus. Acta Otolaryngol. Suppl. (Stock.) 511:153, 1994. 196. Sandberg, D. I., and Souweidane, M. M.: Hemifacial spasm caused by a pilocytic astrocytoma of the fourth ventricle. Pediatr. Neurol. 21:754, 1999. 197. Sanders, D. B.: Ephaptic transmission in hemifacial spasm: a single fiber EMG study. Muscle Nerve 12:690, 1989. 198. Sanklecha, M., Kher, A., and Bharucha, B. A.: Asymmetric crying facies: the cardiofacial syndrome. J. Postgrad. Med. 38:147, 1992. 199. Sawada, H., Matsui, M., Udaka, F., et al.: Adult T cell leukemia initially manifesting as facial diplegia. Am. J. Hematol. 32:61, 1989. 200. Schiess, R. J., Biller, J., and Toole, J. F.: Position-dependent hemifacial spasm. Surg. Neurol. 17:423, 1982. 201. Schirm, J. and Mulkens, P. S. J. Z.: Bell’s palsy and herpes simplex virus. APMIS 105:815, 1997. 202. Schloss, M. D., and Bebear, J. P.: Hemifacial spasm: importance of a complete investigation. J. Otolaryngol. 5:319, 1976. 203. Schulze-Bonhage, A., and Ferbert, A.: Electrophysiological recordings in bilateral hemifacial spasm. J. Neurol. Neurosurg. Psychiatry 65:408, 1998.

Diseases of the Seventh Cranial Nerve 204. Scoville, W. B.: Partial extracranial section of seventh nerve for hemi-facial spasm. J. Neurosurg. 31:106, 1969. 205. Selesnick, S. H., and Patwardhan, A.: Acute facial paralysis: evaluation and early management. Am. J. Otolaryngol. 15:387, 1994. 206. Selky, A. K., Dobyns, W. B., and Yee, R. D.: Idiopathic intracranial hypertension and facial diplegia. Neurology 44:357, 1994. 207. Shapiro, J. L., Yudin, M. H., and Ray, J. G.: Bell’s palsy and tinnitus during pregnancy: predictors of pre-eclampsia? Acta Otolaryngol. (Stock.) 119:647, 1999. 208. Shaywitz, B. A.: Hemifacial spasm in childhood treated with carbamazepine. Arch. Neurol. 31:63, 1974. 209. Sherman, J. D., Dagnew, E., Pensak, M. L., et al.: Facial nerve neuromas: report of 10 cases and review of the literature. Neurosurgery 50:450, 2002. 210. Shoji, H., Hirose, K., Uono, M., and Koya, M.: A case of facial diplegia following herpes zoster ophthalmicus. Eur. Neurol. 19:327, 1980. 211. Sindow, M., Fobe, J.-L., Ciriano, D., and Fischer, C.: Hearing prognosis and intraoperative guidance of brainstem auditory evoked potential in microvascular decompression. Laryngoscope 102:678, 1992. 212. Sinha, P. K., Keith, R. W., and Pensak, M. L.: Predictability of recovery from Bell’s palsy using evoked electromyography. Am. J. Otol. 15:769, 1994. 213. Sittel, C., and Stennert, E.: Prognostic value of electromyography in acute peripheral facial nerve palsy. Otol. Neurotol. 22:100, 2001. 214. Smith, V., and Traquina, D. N.: Pediatric bilateral facial paralysis. Laryngoscope 108:519, 1998. 215. Snider, W. D., Simpson, D. M., Nielsen, S., et al.: Neurological complications of acquired immune deficiency syndrome—analysis of 50 patients. Ann. Neurol. 14:403, 1983. 216. Sprik, C., and Wirtschafter, J. D.: Hemifacial spasm due to intracranial tumor. Ophthalmology 95:1042, 1988. 217. Stahl, N., and Ferit, T.: Recurrent bilateral peripheral facial palsy. J. Laryngol. Otol. 103:117, 1989. 218. Steiner, I., and Cohen, O.: Peripartum Bell’s palsy. Lancet 347:1121, 1996. 219. Stern, B. J., Krumholz, A., Johns, C., et al.: Sarcoidosis and its neurological manifestations. Arch. Neurol. 42:909, 1985. 220. Sugita, T., Murakami, S., Yanagihara, N., et al.: Facial nerve paralysis induced by herpes simplex virus in mice: an animal model of acute and transient facial paralysis. Ann. Otol. Rhinol. Laryngol. 104:574, 1995. 221. Sweeney, C. J., and Gilden, D. H.: Ramsey Hunt syndrome. J. Neurol. Neurosurg. Psychiatry 71:149, 2001. 222. Takasu, T., Furata, Y., Sato, K. C., et al.: Detection of latent herpes simplex virus DNA and RNA in human geniculate ganglia by the polymerase chain reaction. Acta Otolaryngol. (Stock.) 112:1004, 1992. 223. Takeda, K., Tojima, H., Aoyagi, M., and Koike, Y.: Recording of single fiber electromyography in patients with peripheral facial palsy. Acta Otolaryngol. Suppl. (Stock.) 511:156, 1994. 224. Tan, E. K., Chan, L. L., Lum, S. Y., et al.: Is hypertension associated with hemifacial spasm? Neurology 60:343, 2003.

1251

225. Tan, E. K., and Jankovic, J.: Bilateral hemifacial spasm: a report of five cases and a literature review. Mov. Disord. 14:345, 1999. 226. Tan, E. K., and Jankovic, J.: Hemifacial spasm and hypertension: how strong is the association? Mov. Disord. 15:363, 2000. 227. Thomander, L., Aldskogius, H., and Grant, G.: Motor fibre organization in the intratemporal portion of cat and rat facial nerve studied with the horseradish peroxidase technique. Acta Otolaryngol. (Stock.) 93:397, 1982. 228. Tojima, H., Aoyagi, M., Inamura, H., and Koike, Y.: Clinical advantages of electroneurography in patients with Bell’s palsy within two weeks after onset. Acta Otolaryngol. Suppl. (Stock.) 511:147, 1994. 229. Toma, E., Poisson, M., Claessens, M.-R., et al.: Herpes simplex type 2 pericarditis and bilateral facial palsy in a patient with AIDS. J. Infect. Dis. 160:553, 1989. 230. Toremalm, N. G., Elmqvist, D., Elner, A., and Mercke, U.: Hemifacial spasm: nerve block with phenol under electromyographic control. Acta Otolaryngol. (Stock.) 83:341, 1977. 231. Traynor, A. E., Gertz, M. A., and Kyle, R. A.: Cranial neuropathy associated with primary amyloidosis. Ann. Neurol. 29:451, 1991. 232. Uchihara, T., Yoshida, S., and Tsukagoshi H.: Bilateral facial paresis with Sjogren’s syndrome. J. Neurol. 236:186, 1989. 233. Uri, N., Greenberg, E., Meyer, W., and Kitzes-Cohen, R.: Herpes zoster oticus: treatment with acyclovir. Ann. Otol. Rhinol. Laryngol. 101:161, 1992. 234. Vahlne, A., Edstrom, S., Arstila, P., et al.: Bell’s palsy and herpes simplex virus. Arch. Otolaryngol. 107:79, 1981. 235. Valavanis, A., Kubik, S., and Oguz, M.: Exploration of the facial nerve canal by high-resolution computed tomography: anatomy and pathology. Neuroradiology 24:139, 1983. 236. Valls-Sole, J., Tolosa, E. S., and Pujol, M.: Myokymic discharges and enhanced facial nerve reflex responses after recovery from idiopathic facial palsy. Muscle Nerve 15:37, 1992. 237. Van de Biezenbos, J. B. M., Horstink, M. W. I. M., van de Vlasakker, C. J. W., et al.: A case of bilateral alternating hemifacial spasms. Mov. Disord. 7:68, 1992. 238. Vanzieleghem, R., Lemmerling, M., Carton, D., et al.: Lyme disease in a child presenting with bilateral facial nerve palsy: MRI findings and review of the literature. Neuroradiology 40:739, 1998. 239. Wang, A., and Jankovic, J.: Hemifacial spasm: clinical findings and treatment. Muscle Nerve 21:1740, 1998. 240. Wechsler, A. F., and Ho, D. D.: Bilateral Bell’s palsy at the time of HIV seroconversion. Neurology 39:747, 1989. 241. Wiederholt, W. C., and Siekert, R. G.: Neurological manifestations of sarcoidosis. Neurology 15:1147, 1965. 242. Williams, P. L., Warwick, R., Dyson, M., and Bannister, L. H. (eds.): Gray’s Anatomy, 37th Am. ed. New York, Churchill Livingstone, p. 1107, 1989. 243. Williamson, I. G., and Whelan, T. R.: The clinical problem of Bell’s palsy: is treatment with steroids effective? Br. J. Gen. Pract. 46:743, 1996. 244. Yanagihara, N., Murakami, S., and Nishihara, S.: Temporal bone fractures inducing facial nerve paralysis: a new classification and its clinical significance. Ear Nose Throat J. 76:79, 1997.

1252

Diseases of the Peripheral Nervous System

245. Yanagihara, N., Yumoto, E., and Shibahara, T.: Familial Bell’s palsy: analysis of 25 families. Ann. Otol. Rhinol. Laryngol. 97:8, 1988. 246. Yeh, H., and Tew, J. M. Jr.: Tic convulsive, the combination of geniculate neuralgia and hemifacial spasm relieved by vascular decompression. Neurology 34:682, 1984.

247. Ziem, P. E, Pfrommer, C., Goerdt, S., et al.: MelkerssonRosenthal syndrome in childhood: a challenge in differential diagnosis and treatment. Br. J. Dermatol. 143:860, 2000. 248. Zimmer, W. M., Rogers, R. S. III, Reeve, C. M., and Sheridan, P. J.: Orofacial manifestations of MelkerssonRosenthal syndrome. Oral Surg. Oral Med. Oral Pathol. 74:610, 1992.

51 Diseases of the Eighth Cranial Nerve DORIS-EVA BAMIOU AND LINDA M. LUXON

Anatomy and Physiology Age-Related Changes of the Eighth Nerve Diagnostic Approach Audiometric Procedures Vestibular Tests

Pathologies of the Eighth Nerve Congenital and Hereditary Disorders Acquired Disorders Other Insults to the Eighth Nerve: Drugs, Trauma, Noise, Autoimmune Disorders

ANATOMY AND PHYSIOLOGY The eighth cranial nerve subserves two distinct senses, hearing and balance. The mechanical energy of sound and acceleration is transduced into electrical activity by the hair cells in the organ of Corti and in the maculae and cristae of the vestibular organ and transmitted to the brain via the eighth nerve. The cochlear receptor (the organ of Corti) includes motile outer hair cells (OHCs), which amplify the incoming sound, and inner hair cells (IHCs), which generate current in response to sound and receive about 95% of the afferent neurons (Fig. 51–1). There are approximately 30,000 cochlear nerve fibers. Each IHC receives 10 to 30 afferent endings of type I neurons, while each OHC receives 4 to 10 afferent endings of type II neurons.98 The cell bodies of these neurons form the spiral ganglion that lies in Rosenthal’s canal. IHCs are divergent, in that many type I neurons receive input from a single IHC, while each neuron makes synapses with many cells in the cochlear nuclei. Fibers from the low-frequency encoding apex of the cochlea occupy the core of the nerve trunk and end in the ventral portion of the cochlear nucleus, while fibers from the highfrequency base are at the periphery of the nerve trunk and arborize in more dorsal parts of the cochlear nucleus.98 Vestibular sensory hair cells consist of type I, which form a calyx-like contact with the afferent fibers, and type II, with bouton-type synapses with afferent fiber (Fig. 51–2). Type II hair cells release transmitter in response to movement, while type I cells appear to play a complemen-

Secondary Degeneration of the Eighth Nerve Conclusion

tary role.21 The cell bodies of vestibular neurons form Scarpa’s ganglion. The vestibular nerve contains approximately 15,000 fibers, with a superior division that innervates the anterior and horizontal canals, the utricle, and part of the saccule, and an inferior division that innervates the posterior canal and the main part of the saccule, and with afferents that terminate within the vestibular nuclei. The ventral component of the inferior division of the vestibular nerve consists of the olivocochlear bundle, which carries efferent innervation to the cochlea, mostly to the OHCs of the contralateral cochlea by the medial bundle, and to radial afferent fibers in the ipsilateral cochlea by the lateral bundle. The eighth nerve lies in the internal auditory canal (IAC), where the cochlear branch occupies the anteroinferior part, the vestibular branch occupies the posterior half, and the facial nerve occupies the anterosuperior quadrant of the IAC. The eighth nerve enters the brainstem in the cerebellopontine angle at the pontomedullary junction (Figs. 51–3 and 51–4).

AGE-RELATED CHANGES OF THE EIGHTH NERVE The eighth nerve structure and function change with age. Myelin sheaths are not present in the auditory nerve central to the glial junction until the 26th fetal week or later.72 While cochlear function is mature by the 35th fetal week, maturation of the hair cell–auditory nerve for center 1253

1254

Diseases of the Peripheral Nervous System

Tectorial membrane Outer hair cells

Inner hair cells

FIGURE 51–1 The organ of Corti.

Stereocilia Cuticular plate Type II hair cell

Type I hair cell

FIGURE 51–2 Vestibular hair cells type I and II.

frequencies above 6 kHz is slightly more delayed.34 The III interwave interval of the auditory brainstem responses, which represents conduction in eighth nerve axons, reaches adult values before term, while the II-III interval, which is thought to contain a synapse in the cochlear nuclei, reaches adult values between 1 and 2 years after birth.91 Conversely, in neural presbyacusis, the predominantly high-frequency hearing loss with speech discrimination difficulties90 caused by degenerative loss of cochlear neurons (i.e., age related), cochlear neurons may be reduced by 50% or more compared to the neonatal

norm.105 Similarly, while unmyelinated axons remain unchanged, the transverse areas of myelinated axons of the vestibular nerve decrease with age,39,79 and these anatomic changes correspond to changes in vestibular reflexes.14

DIAGNOSTIC APPROACH While many cochlear and vestibular end-organ disorders also affect the eighth nerve to varying degrees, some will affect the eighth nerve predominantly or even in isolation.

1255

Diseases of the Eighth Cranial Nerve

FIGURE 51–3 A, High-resolution, fast-spin echo transverse brain section through the pontomedullary junction with T2-weighted contrast (3-mm thickness). Arrows indicate (clockwise from the top) the cochlea, auditory nerve, vestibular nerve, and horizontal semicircular canal. B, High-resolution, fast-spin echo transverse brain section, at slightly lower level than the previous. Double arrow indicates inferior cerebellar peduncle. White arrows indicate (from top to bottom) the pyramid, anterior inferior cerebellar artery, and flocculus. (Courtesy of Dr. John Stevens.)

In either case, identification of eighth nerve involvement is crucial for diagnostic as well as therapeutic purposes. However, no single diagnostic procedure will identify the cochlear or vestibular nerve involvement, because current audiometric procedures do not allow differentiation between lesions affecting the IHCs, OHC–nerve synapse lesions, or purely neural lesions. A multiple test battery approach is required, including both behavioral and electrophysiologic tests, together with a sound understanding of the relevant physiology, in order to attribute deficits to end-organ versus neural pathology. Currently used audiometric and vestibular procedures are discussed subsequently.

Audiometric Procedures

FIGURE 51–4 Internal auditory canal. an ⫽ auditory nerve; fn ⫽ facial nerve; ivn ⫽ inferior vestibular nerve; svn ⫽ superior vestibular nerve.

Behavioral Procedures Pure Tone Audiogram. This assesses hearing sensitivity and will identify hearing loss resulting from external/middle ear pathology (conductive) versus cochlear, neural, or more central pathology (sensorineural hearing loss [SNHL]), but cannot separate a purely sensory from a neural hearing loss. Tests that seek to make this differentiation, such as tone decay, which evaluates auditory adaptation, a hallmark of a neural lesion, and the

1256

Diseases of the Peripheral Nervous System

alternate binaural loudness balance test, which may identify recruitment characteristic of a cochlear hearing loss, have poorer sensitivity to eighth nerve lesions than electrophysiologic procedures such as auditory brainstem evoked responses (ABRs), and their use is thus limited.16 Speech Audiometry. This assesses the ability to recognize words presented at different intensities. Word recognition is least affected by conductive hearing loss and severely affected by a neural hearing loss, but there is a large variability within the same diagnostic categories.14

Additional Tests Pathology of the auditory nerve will result in deficits in temporal processing and frequency discrimination139; however, these deficits are not specific for neural pathology. Electrophysiologic Procedures. Tests of OHC/IHC function include otoacoustic emissions (OAEs) and electrocochleography. OAEs57 are sounds that originate in the cochlea as by-products of the preneural cochlear amplifier. They reflect OHC function and can be recorded even when the auditory nerve is severed, but they require good middle ear function. Electrocochleography 35 is the technique that records the electrical responses of the cochlea and nerve potential. Its components are as follows: 1. The cochlear microphonic, an alternating current that mirrors stimulus characteristics and reflects the basilar membrane displacement/hair cell (mostly OHC) function. Its presence does not exclude OHC pathology,135 and it can be present in the absence of OAEs in cases of auditory neuropathy (AN).30 2. The summating potential is a direct current response that reflects hair cell function and manifests as a shift in the cochlear microphonic baseline. 3. The compound action potential represents the summed response of numerous auditory nerve fibers that fire synchronously. Tests of the Auditory Pathway up to the Lower Brainstem. ABRs, acoustic reflexes, and OAE suppression by noise are objective tests that involve overlapping but not identical pathways. The ABR is the most widely used, with peaks I and II corresponding to the eighth nerve.69 The ABR can be absent or severely disrupted in eighth nerve pathology. An early response on the ABR that reverses polarity when the stimulus polarity is reversed (i.e., from rarefaction to condensation clicks) reflects a cochlear microphonic, which mirrors stimulus polarity, rather than a neural response that is not affected by polarity of stimulus.8 In some cases with AN, the ABR is very sensitive to stimulus rates; that is, it may be elicited by a low stimulus rate but may disappear at higher rates, reflecting the

presence of a synaptic disorder or demyelinating neuropathy.122 The acoustic reflex involves sound-elicited middle ear muscle contraction via a neural chain comprising the eighth nerve, cochlear nucleus, superior olivary complex, and ipsiand contralateral medial facial nerve motoneurons.45 Acoustic reflexes are recorded using a tympanometer. These reflexes can be absent or elevated (i.e., they require louder stimuli than normal) in eighth nerve lesions.22 Acoustic stimulation reduces the amplitude of OAEs, via afferent fibers from the cochlea to the superior olivary complex and efferent fibers from the superior olivary complex, mostly via the medial olivocochlear bundle to the ipsi- and contralateral cochlea.96 OEA suppression by noise tests the reduction in the OAE caused by noise stimulation of the ipsior contralateral ear (i.e., the integrity of the efferent olivocochlear system).58 AN patients show absent suppression to ipsi-, contra-, or bilateral stimulation by noise (Fig. 51–5).9

Vestibular Tests Standard procedures such as the caloric test and eye movement recording by electronystagmography may help differentiate between an end-organ/neural and a more central vestibular lesion, but not between an end-organ and a neural lesion. Galvanic stimulation of the vestibular system may allow for this definite differentiation; this response is absent in pathology of the eighth nerve,73 but also activates cortical areas of the auditory and nociceptive system.12 Video-oculography (VOG) and vestibular evoked myogenic potentials (VEMPs) do not differentiate between an end-organ and a neural lesion, but they offer a better characterization of the nystagmus than do conventional tests, and this information together with audiometric procedures can help define the site of pathology. VOG is a video-format, noninvasive technique that records ocular motions and positions in all degrees of freedom, thereby providing information about the origin of the nystagmus, because spatial direction of nystagmus depends on the planes of the affected organs.125 VEMPs are averaged muscular responses of the sternocleidomastoid muscle, which is activated by sound via the saccule, inferior vestibular nerve, and vestibulospinal tract,23 with complete disappearance of potentials in cases of tumors of the inferior vestibular nerve.132

PATHOLOGIES OF THE EIGHTH NERVE Congenital and Hereditary Disorders Aplasia/Hypoplasia of the Eighth Nerve Diagnosis. Aplasia/hypoplasia of the eighth nerve is an infrequent cause of deafness that may exist in 4% of profoundly deaf children, who receive no benefit from hearing aids and who are referred for a cochlear implant.6

26 year-old with HSMN

PURE TONE AUDIOGRAM

–20 –10 0 10 20 30 40 50 60 70 80 90 100 110 120

Left

Hearing level (dB ISO)

Hearing level (dB ISO)

Right

125 250 500 1000 2000 4000 8000 Frequency (Hz)

–20 –10 0 10 20 30 40 50 60 70 80 90 100 110 120 125 250 500 1000 2000 4000 8000 Frequency (Hz)

A

FIGURE 51–5 Audiometric results of a 26-year-old patient with hereditary sensorimotor neuropathy. A, Pure tone audiogram was normal. B, Otoacoustic emissions were normal in both ears. C, ABR showed poor morphology with delayed waveforms in both ears. Caloric responses (not shown) were absent, and otoacoustic emission suppression by noise (not shown) was impaired. Figure continued on following page

Sensitivity and sweep time per division 2 0.62 uV 1.0 msec

4 0.62 uV 1.0 msec

5 0.62 uV 1.0 msec

Neurological ABR ILI-a IIILI-a VLI-a ILI-b IIILI-b VLI-b IIIRI-a VRI-a IIRI-b VRI-b

C FIGURE 51–5 Continued

1.48ms 4.04ms 6.12ms 1.48ms 4.04ms 6.12ms

I-IIILI-a III-VLI-a I-VLI-a I-IIILI-b III-VLI-b I-VLI-b III-VRI-a III-VRI-b IADV-2

2.56ms 2.08ms 4.64ms 2.56ms 2.08ms 4.64ms

7 0.62 uV 1.0 msec

1259

Diseases of the Eighth Cranial Nerve

Table 51–1. Classification of Aplasia/Hypoplasia of the Eighth Nerve Type 1 Type 2A Type 2B ?Type 3*

Aplasia of VCN ⫹ stenosis of IAC Common VCN ⫹ aplasia/hypoplasia of cochlear branch ⫹ labyrinthine malformation Common VCN ⫹ aplasia/hypoplasia of cochlear branch ⫹ normal labyrinth Common VCN ⫹ aplasia/hypoplasia of vestibular branch

IAC ⫽ internal auditory canal; VCN ⫽ vestibulocochlear nerve. *Type 3 has not yet been detected on magnetic resonance imaging, and its existence is debated. Data from Casselman, J. W., Offeciers, F. E., Govaerts, P. J., et al.: Aplasia and hypoplasia of the vestibulocochlear nerve: diagnosis with MR imaging. Radiology 202:773, 1997.

The eighth nerve may be absent even in the presence of a normal IAC and other vestibulocochlear structures.6,18 If the vestibular nerve is also involved, there may be delayed motor milestones, and frequently a syndromic diagnosis may be made (Table 51–1).6 The eighth nerve can be reliably visualized on high-resolution magnetic resonance imaging (MRI),18 but the MRI underestimates the size of the nerve (T. Owa, personal communication, 2001), and a rudimentary cochlear branch may be missed as a result of partial volume effects. Thus electrophysiologic studies need to be conducted in selected cases. Treatment. While cochlear implantation can be successful in cases of a hypoplastic eighth nerve, for the child with bilateral aplasia this technique is inappropriate. In the presence of a normal brainstem, stimulation of the brainstem via a brainstem implant may be feasible.24 Inherited Eighth Nerve Disorders There are 28 cloned genes for deafness that encode proteins expressed in the organ of Corti and vestibule, as well as the eighth nerve (Fig. 51–6).133 Non-syndromic Disorders. AN may rarely be an isolated finding in some families.95 Otoferlin encodes a protein at the base of the inner hair cells that is thought to be involved in synaptic vesicle recycling (Fig. 51–7). Linkage analysis of four nonsyndromic recessive families with AN identified a critical region on 2p23, which includes the otoferlin locus, and pathologic otoferlin mutations were identified in three families,95 although AN may also be a double-recessive trait that requires mutations in two different genes to cause the condition. Syndromic Presentations. Mutations in genes that encode gap junction channels (i.e., channels between neighboring cells that permit the rapid exchange of certain molecules) may be responsible for peripheral neuropathy as well as AN resulting in hearing loss. These include

connexin 3161 and connexin 32, which are responsible for the X-linked form of Charcot-Marie-Tooth disease.112 Abnormalities of peripheral protein 22 on chromosome 17p11.2 are responsible for other forms of Charcot-MarieTooth disease137 in which the primary histopathologic lesion is loss of cochlear spiral ganglion cells and hypertrophic changes in cranial nerves VII and VIII.76 The audiologic evaluation reveals both AN and cochlear involvement in affected individuals, with increasing clinical severity and younger age of onset with each progressive generation.59 Hereditary motor and sensory neuropathy, Lom type,55 is an autosomal recessive peripheral neuropathy with AN hearing loss and absent vestibular caloric responses.17 There are reduced numbers and size of myelinated fibers, and the responsible gene, NDRG1, may have a role in the peripheral nervous system, possibly in the Schwann cell signaling necessary for axonal survival.54 AN and vestibular neuropathy (VN) may be associated with peripheral neuropathy in olivopontocerebellar and spinocerebellar degeneration121 and in Friedreich’s ataxia, which is characterized by severe loss of cochlear and to a lesser extent vestibular neurons, but with intact sensory epithelia.119 Cerebro-oculofacioskeletal syndrome is a rare autosomal recessive disorder with dysmorphic features, hypotonia, osteoporosis, and neural degeneration beyond the central nervous system that shows accelerated cochlear and to a lesser degree vestibular nerve degeneration resembling the histopathology findings in Friedreich’s ataxia.37 Apart from cochlear abnormalities, primary eighth nerve degeneration is also present in Usher’s syndrome, or recessively inherited deafness and retinitis pigmentosa.115 There is also some evidence to suggest AN in some cases of mitochondrial disorders.140

Acquired Disorders Auditory Neuropathy Background. In 1996, Starr et al.123 described 10 subjects with hearing loss, with normal OHC function and abnormal ABR and other reflexes that indicated an auditory nerve lesion. The term auditory neuropathy was coined to describe this symptom complex. Diagnosis. Diagnosis of AN requires exclusion of other potential diagnoses by appropriate investigations and all three of the following criteria116: 1. Evidence of poor hearing in the presence of normal or abnormal audiometric thresholds. AN can be associated with any degree of hearing loss or audiometric configuration and with marked hearing fluctuation,116 with characteristic temporal and speech processing deficits139 that become more evident when the audiogram is normal.

1260

Diseases of the Peripheral Nervous System

Mesenchymal cells

Basal cells Marginal cells Intermediate cells Spiral ligament

Tectorial membrane

Interdental cells

Spiral prominence epithelium Spiral limbus

Outer hair cell

Inner hair cell Spiral ganglion

COCH KVLQT1 and KCNE1 TECTA

MY07A and POU4F3 FGFR3 ATP6B1

PDS MITF and EDN3

Stria vascularis

Spiral ligament

Inner sulcus cells

Spiral limbus Root cells Deiter’s cells

Pillar cells

GJB2 and GJB3 NDP

KCNQ4 OTOF

Spiral ganglion

FIGURE 51–6 Gene expression in the cochlear duct. A section through the cochlear duct is shown, with different shades (given in the key) representing the expression patterns of different genes. The distribution of protein in the adult is given where known; otherwise, data on messenger RNA distribution are from in situ hybridization, again in the adult if this information is available. The distribution of COCH is inferred from the distribution in the bird cochlea. (From Steel, K., and Bussoli, T. J.: Deafness genes—expression of surprise. Trends Genet. 15:209, 1999, with permission.)

2. Evidence of poor auditory neural function, with elevated or absent auditory brainstem reflexes, such as middle ear muscle reflexes, and OAE suppression by noise and abnormal ABR. 3. Evidence of normal hair cell function, such as normal OAE or cochlear microphonics. OAEs can disappear over time, while cochlear microphonics may remain present.31

ence of this response correlates with good speech perception ability, possibly indicating that some synchrony of the auditory signal is preserved at the cortex.94

Approximately 50% of children with AN will have eventrelated potentials in response to tone and speech stimuli, regardless of the degree of the hearing loss, and the pres-

1. The auditory nerve in isolation 2. The auditory and vestibular nerve and in some cases the cochlea40 (Table 51–3)

Etiology. AN may present in 0.23% of infants at risk for hearing loss.93 Causes include genetic, autoimmune, infectious, toxic, neonatal illness, or idiopathic conditions (Table 51–2) that affect the following:

1261

Diseases of the Eighth Cranial Nerve

Table 51–3. Disorders That May Damage Both the Vestibular and Cochlear Nerves

K+ Ca2+

Ca2+ (Pmca2)

(Myo7a) (Myo6)

Disorder

Symptoms

Auditory neuropathy ⫾ peripheral neuropathy*

Minimal vestibular symptoms, as a result of vestibular compensation, if the cause is progressive or congenital40 VIIth nerve, peripheral neuropathies136 Vertigo, history of tick bite/erythema chronicum migrans52 Bilateral symptoms, acute SNHL after febrile illness47 Bilateral symptoms, progressive hearing loss118 Vesicular eruptions in the outer ear, followed by facial paralysis and audiovestibular symptoms1 Tinnitus, vertigo, hearing loss, and ear fullness/pressure44,77

Alcohol Neuroborreliosis Sensory hair cell

HIV Neurosyphilis

(OTOF)

Ca2

K+ (KCNQ4) Ca2+

FIGURE 51–7 Otoferlin (encoded by Otof) may help control neurotransmitter release at the synapse. (From Steel, K.: New interventions in hearing impairment. BMJ 320:622, 2000, with permission.)

Table 51–2. Causes of Auditory Neuropathy Genetic Nonsyndromic Syndromic

Autoimmune Infectious

Neonatal illness Toxic/metabolic

Idiopathic121 Other

Otoferlin95 Charcot-Marie-Tooth disease61, 112, 137 Hereditary motor and sensory neuropathy, Lom type55 Olivopontocerebellar and spinocerebellar degeneration121 Friedreich’s ataxia119 Cerebro-oculofacioskeletal syndrome37 Usher’s syndrome115 Cogan’s syndrome49 Neurosyphilis118 Human immunodeficiency virus (HIV)47 Cytomegalovirus in HIV-positive patients66 Typhus38 Hyperbirubinemia or prematurity31 Facial-auditory nerve oxalosis2 Potentially alcohol, organic mercury, and uremia76 Transient auditory neuropathy caused by high temperature124

Herpes zoster oticus (Ramsay Hunt syndrome) Ménière’s disease

HIV ⫽ human immunodeficiency virus; SNHL ⫽ sensorineural hearing loss. *Peripheral neuropathy was significantly more prevalent in patients with auditory neuropathy when vestibular paresis was also present.40

3. The auditory nerve and other peripheral nerves as a peripheral neuropathy syndrome, in which case vestibular nerve involvement is more common40 Histopathology and Pathophysiology Nerve biopsies (Fig. 51–8) in AN accompanied by peripheral neuropathy show two major changes of the auditory nerve: 1. Demyelination, leading to disruption of temporal synchrony of the auditory input 2. Axonal loss and reduced numbers of auditory fibers, which conduct the auditory input to the cortex122 AN could also be due to a disorder of the synapse between IHCs and the auditory nerve, as indicated by sensitivity of ABR to fast stimulus rates,122 or to selective IHC loss, as demonstrated by animal models of AN (Fig. 51–9).99,117 Thus AN is likely to be a group of diverse disorders that result in the same clinical presentation. Management. Management of AN will include appropriate medical management of the causative disease. Sophisticated hearing aids may benefit some cases,26 although there is no correlation between aided speech perception ability and the results of clinical tests.94 Cochlear implants have also been employed, with reports of improved listening and communication skills and restoration of the desynchronous ABR following implantation in some cases.113 Finally, auditory training and other strategies are necessary for management of the functional deficit.26

1262

Diseases of the Peripheral Nervous System

“umbrella” diagnosis for diverse vestibular disorders. Recent work, however, has defined a specific entity.12 Diagnosis. Key signs and symptoms of VN include12 1. Rotatory vertigo, nausea, and vomiting 2. Pathologic subjective vertical toward the side of lesion 3. Postural imbalance, Romberg fall, and past pointing toward the side of lesion 4. Horizontal and rotatory spontaneous nystagmus opposite to the side of lesion, with enhancement by the removal of fixation Smooth pursuit, saccades, optokinetic nystagmus, and neurologic examination are normal, and there are no auditory signs. The caloric test shows canal paresis on the side of the lesion, which on subsequent retesting may recover in 42%87 to 70% of patients.85 Ultimately, VN is a diagnosis of exclusion, and causes such as Ménière’s disease, migraine, vestibular schwannoma, and multiple sclerosis must be excluded. FIGURE 51–8 A and B, Auditory nerve adjacent to brainstem from subject with auditory neuropathy (A) and age-matched control (B) (osmium tetroxide; ⫻400). C and D, High-power (⫻1500) micrograph of auditory nerve (C) and sural nerve (D) from the same subject. The arrows point to thinly myelinated fibers reflecting incomplete remyelination. (From Starr, A., Michalewski, H. J., Zeng, F. G., et al.: Pathology and physiology of auditory neuropathy with a novel mutation in the MPZ gene [Tyr145-Ser]. Brain 126:1604, 2003, with permission.)

Vestibular Neur(on)itis Background. VN was first described as an isolated vestibular syndrome with no neurologic or auditory abnormality,32 but it then became recognized as an

Etiology, Histopathology, and Pathophysiology. The principal histopathologic finding is degeneration of the vestibular nerve as a result of infective, vascular, or immunemediated disorders.74 A viral theory13,106 has gained more momentum than the vascular theory60 as a result of temporal bone pathology findings that are similar to those of herpes zoster oticus,106 the epidemic occurrence of VN,109 and the presence of herpes simplex virus (HSV) in human vestibular ganglia.42 However, a common etiology and pathomechanism for VN cannot be established (Fig. 51–10). HSV may enter the nerve through the skin and replicates within the nerve cell nucleus, to be released following an insult affecting Schwann cells of the nerve or epithelial cells.63 HSV DNA can be detected in 60% of vestibular ganglia post mortem,42 while the latency-associated

FIGURE 51–9 A, Complete loss of inner hair cells in a preterm infant. B, Normal hair cells for comparison. (From Slack, R. W., Wright, A., Michaels, L., and Frohlich, S. A.: Inner hair cell loss and intracochlear clot in the preterm infant. Clin. Otolaryngol. 11:443, 1986, with permission. Courtesy of Professor Michaels.)

Diseases of the Eighth Cranial Nerve

1263

FIGURE 51–10 Light micrograph of a horizontal section through the superior division of the vestibular nerve of a patient with left vestibular neuronitis. A, Severe degeneration of the myelinated dendritic processes of the superior division of the vestibular nerve (SVN) with normal neuroepithelium of the crista ampullaris of the lateral semicircular canal (LSC) in the left ear. B, Normal SVN and LSC in the right ear. (Hematoxylin and eosin staining.) (From Nadol, J. B.: Vestibular neuritis. Otolaryngol. Head Neck Surg. 112:162, 1995, with permission. Courtesy of Dr. Nadol.)

transcript of HSV-1 (i.e., the DNA piece that encodes the protein that blocks the replication cycle of the virus and thus protects the infected cells from damage and enables a latent state) has also been identified in 63% of vestibular ganglia examined.128 Other viruses associated with VN include mumps, cytomegalovirus, and rubella.28 VN affects the superior and spares the inferior vestibular nerve, as indicated by VOG and VEMP studies,20,36 possibly because the inferior vestibular nerve consists of two branches in separate canals, because of the better accessibility of the superior vestibular nerve to viral migration along the faciovestibular anastomosis,4 or because of increased susceptibility of the superior vestibular nerve to compression and ischemia as a result of its narrow and long bony canal.46 Management. Most patients will recover within weeks to months by a process characteristic of central nervous system plasticity, known as vestibular compensation.

However, symptoms and signs induced by dynamic highfrequency head movements will persist in the majority of patients.12,87 Vestibular rehabilitation is the treatment of choice, and there is no evidence for the effectiveness of antiviral agents, steroids, or surgical intervention.5,56,127 Lesions That Compress or Infiltrate the Vestibulocochlear Nerve Neurovascular Cross-Compression of the Eighth Nerve BACKGROUND. This entity remains controversial, and its definition differs12,71 or remains unspecified in some reports. The finding of a blood vessel in close relationship with the eighth nerve on brain MRI does not prove the existence of a causative relationship, even in the presence of vertigo and hearing symptoms, because contact between a blood vessel and the eighth nerve may occur in 12.5% of otherwise normal MRI scans.89 The first reported cases had hemifacial spasm or trigeminal neuralgia in

1264

Diseases of the Peripheral Nervous System

association with vestibular symptoms,11 while subsequent patients with constant positional vertigo and signs compatible with compression of the vestibular nerve were reported to have relief of symptoms by eighth nerve microvascular decompression. Thus the terms disabling positional vertigo,53 and more recently vestibular paroxysmia,13 were coined.

Vestibular Schwannoma BACKGROUND. Sandifort of Leiden first documented in 1777 a case of vestibular schwannoma (VS), while Sir Charles Bell in 1830 made an early clinicopathologic correlation.92 VS represent 6% of intracranial tumors, with 13 newly diagnosed cases per 1 million population per year,15 but this may underestimate the true incidence.3

DIAGNOSIS. Vestibular paroxysmia has been attributed to paroxysmal hyperactivity of the eighth nerve with a functional deficit in the attack-free interval. Diagnostic features may include12,13

DIAGNOSIS. Gradually developing hearing loss (85%), tinnitus (70%), and vertigo (20%) are presenting symptoms.15,111 Atypical presentations, occurring in about 20% of cases, may include tinnitus in the presence of normal hearing, unilateral hearing complaints with normal thresholds, sudden or fluctuating SNHL,102 neurologic symptoms including fifth or seventh cranial nerve deficits, long-standing SNHL, or recent onset of vestibular symptoms.10,84 The clinical presentation may be considered in five stages92 (Table 51–4). Audiologic investigations including audiometry and ABR are necessary; however, gadolinium-enhanced MRI is the gold standard investigation.138 T2-weighted imaging of high quality is both an accurate and a cost-effective screening test.27

1. Rotational vertigo lasting seconds to minutes 2. Attacks provoked and modified by head positions 3. Hyperacusis and/or tinnitus permanently or during the attack; low-pitched, pulsatile tinnitus or tinnitus with hemifacial spasm is reported as characteristic97 4. Auditory or vestibular deficits on neurophysiologic testing107 5. Potential control of these attacks by carbamazepine12 6. Exclusion of other causes PATHOPHYSIOLOGY. Microscopic examination of the vestibular nerve in eighth nerve compression108 identified axonal loss or swelling with no myelin loss and significant endoneurial fibrosis, and it has been postulated that acute inflammation of the nerve caused arachnoidal adhesion and contact with a vascular loop. The compound cochlear nerve action potentials recorded during neurovascular decompression suggest that the presence of vertigo and/or tinnitus depends on the part of the nerve that is compressed by the blood vessels.97 MANAGEMENT. Carbamazepine may be of benefit, with relapse of symptoms after washout of the drug.12 In medically intractable cases, retromastoid craniotomy and microvascular decompression are reported as effective.70,71,107 However, hearing loss can be an immediate or delayed complication in some patients.43 It is not clear whether microvascular decompression is effective because it stops the triggering effect of the vessel, or by traumatizing and thus denervating the nerve, or whether it has a placebo effect.12,107 Because there is no pathognomonic sign for this syndrome, it is a diagnosis of exclusion. Medical management and vestibular and tinnitus rehabilitation should be aggressive, while surgical intervention should be considered as a last resort. Osseous Lesions of the IAC. A progressive osseous lesion of the IAC may compress the eighth nerve and give symptoms/signs from both its branches. Osteopetrosis,114 craniodiaphysial dysplasia,51 and other similar conditions fall within this category; however, the case is less clear for osteomas29 or stenosis of the IAC,81 which may well be incidental radiologic findings.

GENETICS. Neurofibromatosis 1 (von Recklinghausen’s disease; NF1) and neurofibromatosis 2 (NF2) are clinically and genetically distinct disorders.48,82,130 NF2 is an autosomal dominant disorder caused by the loss of function of a tumor suppressor gene on 22 q12, while the NF1 gene is located at chromosome 17 q11.2. VSs are rarely seen in NF1 patients (Table 51–5 and Fig. 51–11). HISTOPATHOLOGY. VSs are benign, encapsulated tumors with a nodular surface. Neoplastic cells display Antoni type A (fasciculated) and Antoni type B (reticular) neurilemoma patterns. VSs arise from the Schwann cells of the vestibular nerve,80 between the glial neurolemmal junction and the cribrose area within the IAC,103 and initially tend to compress rather than invade the nerve fibers.92 VSs grow on average 3 mm/yr,19 and, because of irregular patterns of growth, need to be monitored by MRI at least once a year.67 MANAGEMENT. Management options include interval scanning, surgical removal, and stereotactic radiosurgery/ radiotherapy. There is no study that compares the relative Table 51–4. Clinical Evolution of Vestibular Schwannoma Stage 1 Stage 2 Stage 3 Stage 4 Stage 5

Otologic stage with symptoms from eighth ⫾ seventh nerve Trigeminal nerve involvement (diameter ⱖ 2 cm) Brainstem and cerebellar compression Increased intracranial pressure Terminal stage with disturbance of brainstem centers and tonsillar herniation

Data from Ramsden, R. T.: Vestibular schwannoma. In Booth, J. B. (ed.): Otology, 6th ed., Vol. 3. London, Scott & Brown Otolaryngology/Read Educational & Professional Publishing, p. 1, 1997.

Diseases of the Eighth Cranial Nerve

Table 51–5. Diagnostic Criteria for Neurofibromatosis 1 (NF1) and Neurofibromatosis 2 (NF2) NF2

NF1

A. Bilateral eighth nerve masses seen on MRI or B. Parent, sibling, or child with NF2

ⱖ2 of the following criteria:

and either unilateral eighth nerve mass or one of the following: Neurofibroma Meningioma Glioma Schwannoma Posterior capsular cataract/opacity Young age

1. ⱖ6 café au lait macules of ⬎5 mm in prepubertal individuals and ⬎15 mm after puberty

2. ⱖ2 neurofibromas of any type or one plexiform neurofibroma 3. axillary or inguinal freckling 4. optic tumors 5. ⱖ2 Lisch nodules 6. A distinctive osseous lesion 7. A first-degree relative with NF1

FIGURE 51–11 Magnetic resonance imaging (MRI) in a patient with neurofibromatosis 2. A, Fine axial T1-weighted, unenhanced MRI at the level of the internal auditory canal. B, Fine axial T1-weighted postgadolinium MRI (at same level as A) demonstrates a 4-mm transverse-diameter enhancing intracanalicular lesion on the left, and the cisternal part of a vestibular schwannoma on the right. In addition, there is enhancement within the middle turn of the cochlea on the right, which is separate from the intracanalicular component on the coronal images. C, Fine coronal T1-weighted postgadolinium MRI demonstrates the cisternal and intracanalicular component of the right vestibular schwannoma.

1265

benefits of each treatment option, and the decision regarding management will depend on age, VS size, health status, patient preferences, and other considerations.15 Following surgical treatment, vestibular exercises improve postural stability and perception of dysequilibrium,50 but vestibular symptoms persist in about one third of operated patients.33 Prognostic indicators for preservation of hearing after surgery vary considerably in different reports and may include the preoperative hearing status, audiometric test results, ABR abnormalities, and tumor origin and size.15 DIFFERENTIAL DIAGNOSIS. Differential diagnosis must include causes of AN and/or VN as well as other compressive or infiltrating lesions. While the vast majority of cerebellopontine angle tumors are VSs, one in five is another pathologic lesion.68 The meningiomas dominate by far this group of tumors, with less uniform symptomatology and less involvement of the eighth nerve and more frequent trigeminal neuralgia than in VS.131 Other lesions may include lipomas, hemangiomas, granulomas, choristomas, and hamartomas80,83,92 (Fig. 51–12).

1266

Diseases of the Peripheral Nervous System

17 year-old male with HSMN

PURE TONE AUDIOGRAM Left

Hearing level (dB ISO)

Hearing level (dB ISO)

Right –20 –10 0 10 20 30 40 50 60 70 80 90 100 110 120 125 250 500 1000 2000 4000 8000 Frequency (Hz)

–20 –10 0 10 20 30 40 50 60 70 80 90 100 110 120 125 250 500 1000 2000 4000 8000 Frequency (Hz)

A

FIGURE 51–12 Audiometric results of a 17-year-old patient with bilateral eosinophil granulomas of the petrous bones. A, Pure tone audiogram with profound hearing loss on the right and normal hearing on the left. B, Otoacoustic emissions were normal in both ears. C, ABR waveforms were absent on the right and normal on the left. Caloric responses (not shown) and otoacoustic emission suppression by noise (not shown) were absent on the right and normal on the left. Figure continued on opposite page

Sensitivity and sweep time per division 1 0.62 uV 1.0 msec

2 0.62 uV 1.0 msec

3 0.62 uV 1.0 msec

4 0.62 uV 1.0 msec

Neurological ABR IRr-a IIIRr-a VRr-a IRr-b IIIRr-b VRr-b IIILr-a VLr-a IIILr-b VLr-b

C FIGURE 51–12 Continued

6.44ms 6.44ms

I-IIIRr-a III-VRr-a I-VRr-a I-IIIRr-b III-VRr-b I-VLI-b III-VLr-a III-VLr-b IADV-2

0.12ms

ILI-a IIILI-a VLI-a ILI-b IIILI-b VLI-b IIIRI-a VRI-a IIIRI-b VRI-b

6.56ms 6.56ms

I-IIILI-a III-VLI-a I-VLI-a I-IIILI-b III-VLI-b I-VLI-b III-VRI-a III-VRI-b IADV-2

0.12ms

1268

Diseases of the Peripheral Nervous System

Similarly, metastatic carcinoma of the temporal bone is uncommon, but the IAC is the second most common site of temporal bone metastasis and will be bilaterally affected in 54.5% of cases of IAC involvement.126 Primary sites of lesion include breast, lung, kidney, stomach, larynx, prostate, and thyroid gland, and metastasis occurs via hematogenous spread or cerebrospinal fluid dissemination.104,126 Symptoms of sudden or progressive hearing loss, tinnitus, vertigo, and other cranial neuropathies86,126,129 may be overshadowed by other metastases. Multiple sclerosis may affect the distal portion or the root entry zone of the eighth nerve to the brainstem, presenting with sudden hearing loss7,41 or vertigo.25 The eighth nerve is the cranial nerve fourth most commonly affected by sarcoidosis, with bilateral audiovestibular symptoms that may respond to steroid and azathioprine treatment.88

Other Insults to the Eighth Nerve: Drugs, Trauma, Noise, Autoimmune Disorders The eighth nerve can be affected by extraneous factors such as drugs (e.g., carboplatin primarily destroys IHCs62), trauma, or noise. Transverse fractures of the temporal bone may damage the seventh and eighth nerve in about 50% of cases,105 with enhancement of the end organ and nerve on MRI.100 Noise primarily damages OHCs and, to a lesser degree, IHCs, but IHC damage may be accompanied by damage to the synaptic region,101 leading to spiral ganglion degeneration. Vasculitis changes in autoimmune disorders such as lupus erythematosus and rheumatoid arthritis may also affect the eighth nerve, with a clinical picture that varies from acute audiovestibular failure65 to delayed ABR.64

Secondary Degeneration of the Eighth Nerve Damage of the peripheral afferent dendrites120 and atrophic changes in supporting cells103 may cause longterm degeneration of the eighth nerve. Neuronal loss is more pronounced in older persons and with longer duration of deafness,75,78 and is greater after measles and syphilis, lesser with otosclerosis, and least with ototoxicity or temporal bone fracture.103 Degeneration is more pronounced in the basal than in the apical half of the cochlea.75 Cochlear neuronal degeneration caused by central (brainstem side) pathologies is less well studied because these pathologies also tend to affect the vascular supply of the end organ.110 In cases of eighth nerve damage, growth factors, including brain-derived neurotrophic factor, neurotrophin 3, insulin, and transforming growth factor-␤1, have been shown to affect the survival and/or the regeneration and repair process of vestibular and auditory neurons and seem to hold promise for therapeutic application.134

CONCLUSION Identification of eighth nerve pathology can be challenging, and may require a multiple test battery and a high index of clinical suspicion. However, the finding of deficient eighth nerve function may be of value because it may help in the diagnosis of the underlying disorder, clarify the genetics of the disease, and enable the most appropriate therapeutic intervention. It also hoped that, in the near future, treatment with factors that will facilitate nerve regeneration will also be possible.

REFERENCES 1. Aleksic, S. N., Budzilovich, G. N., and Lieberman, A. N.: Herpes zoster oticus and facial paralysis (Ramsay Hunt syndrome). J. Neurol. Sci. 20:149, 1973. 2. Anderson, B., and Adams, Q. M.: Facial-auditory nerve oxalosis. Am. J. Med. 88:87, 1990. 3. Anderson, T. D., Loevner, L. A., Bigelow, D. C., and Mirza, N.: Prevalence of unsuspected acoustic neuroma found by magnetic resonance imaging. Otolaryngol. Head Neck Surg. 122:643, 2000. 4. Arbusow, V., Schulz, P., Strupp, M., et al.: Distribution of herpes simplex virus type 1 in human geniculate and vestibular ganglia: implications for vestibular neuritis. Ann. Neurol. 46:416, 1999. 5. Ariyasu, L., Byl, F. M., Sprague, M. S., and Adour, K. K.: The beneficial effect of methylprednisolone in acute vestibular vertigo. Arch. Otolaryngol. Head Neck Surg. 116:700, 1990. 6. Bamiou, D-E., Worth, S., Phelps, P., et al.: Eighth nerve aplasia and hypoplasia in cochlear implant candidates: the clinical perspective. Otol. Neurotol. 22:492, 2001. 7. Bergamaschi, R., Romani, A., Zappoli, F., et al.: MRI and brainstem auditory evoked potential evidence of eighth cranial nerve involvement in multiple sclerosis. Neurology 48:270, 1997. 8. Berlin, C., Borderlon, J., St. John, P., et al.: Reversing click polarity may uncover auditory neuropathy in infants. Ear Hear. 19:37, 1998. 9. Berlin, C., and Hood, L.: Auditory neuropathy (auditory desynchrony) disables efferent suppression of otoacoustic emissions. In Sininger, Y., and Starr, A. (eds.): Auditory Neuropathy. San Diego, Singular, p. 183, 2001. 10. Berrettini, S., Ravecca, F., Russo, F., et al.: Some uncharacteristic clinical signs and symptoms of acoustic neuroma. J. Otolaryngol. 26:97, 1997. 11. Bertrand, R. A., Molina, P., and Hardy, J.: Vestibular syndrome and vascular anomaly in the cerebello-pontine angle. Acta Otolaryngol. 83:187, 1977. 12. Brandt, T.: Vertigo. London, Springer-Verlag, 1999. 13. Brandt, T., and Dieterich, M.: Eighth nerve vascular compression syndrome: vestibular paroxysmia. Baillieres Clin. Neurol. 3:565, 1994. 14. Brandy, W. T.: Speech audiometry. In Katz, J. (ed.): Handbook of Clinical Audiology, 5th ed. New York, Lippincott Williams & Wilkins, p. 96, 2002.

Diseases of the Eighth Cranial Nerve 15. British Association of Otorhinolaryngologists Head and Neck Surgeons (BAOHNS): Acoustic Neuroma (Vestibular Schwannoma): Clinical Effectiveness Guidelines. London, BAOHNS, 2002. 16. Brunt, M. A.: Cochlear and retrocochlear behavioural tests. In Katz, J. (ed.): Handbook of Clinical Audiology, 5th ed. New York, Lippincott Williams & Wilkins, p. 111, 2002. 17. Butinar, D., Zidar, J., Leonardis, L., et al.: Hereditary auditory vestibular, motor, and sensory neuropathy in a Slovenian Roma (Gypsy) kindred. Ann. Neurol. 46:36, 1999. 18. Casselman, J. W., Offeciers, F. E., Govaerts, P. J., et al.: Aplasia and hypoplasia of the vestibulocochlear nerve: diagnosis with MR imaging. Radiology 202:773, 1997. 19. Charabi, S., Thomsen, J., Mantoni, M., et al.: Acoustic neuroma (vestibular schwannoma): growth and surgical and nonsurgical consequences of the wait-and-see policy. Otolaryngol. Head Neck Surg. 113:5, 1995. 20. Chen, C. W., Young, Y. H., and Wu, C. H.: Vestibular neuritis: three dimensional video-nystagmography and vestibular evoked myogenic potential results. Acta Otolaryngol. 120:845, 2000. 21. Cochran, S. L.: Vestibular hair-cell afferent transmission. In Beitz, A. J., and Anderson, J. H. (eds.): Neurochemistry of the Vestibular System. Boca Raton, FL, CRC Press, p. 27, 2000. 22. Cohen, M., and Prasher, D.: ABR and ART in neurootological diagnosis. Scand. Audiol. 17:153, 1988. 23. Colebatch, J. G.: Vestibular evoked potentials. Curr. Opin. Neurol. 14:21, 2001. 24. Colletti, V., Fiorino, F., Sacchetto, L., et al.: Hearing habilitation with auditory brainstem implantation in two children with cochlear nerve aplasia. Int. J. Pediatr. Otorhinolaryngol. 60:99, 2001. 25. Commins, D. J., and Chen, J. M.: Multiple sclerosis: a consideration in acute cranial nerve palsies. Am. J. Otol. 18:590, 1997. 26. Conne-Wesson, B., Rance, G., and Sininger, Y.: Amplification and rehabilitation strategies for patients with auditory neuropathy. In Sininger, Y., and Starr, A. (eds.): Auditory Neuropathy. San Diego, Singular, p. 233, 2001. 27. Daniels, R. L., Swallow, C., Shelton, C., et al.: Causes of unilateral sensorineural hearing loss screened by high-resolution fast spin echo magnetic resonance imaging: review of 1070 consecutive cases. Am. J. Otol. 21:173, 2000. 28. Davis, L. E.: Comparative experimental viral neurolabyrinthitis. Am. J. Otol. 11:382, 1990. 29. Davis, T. C., Thendinger, B. A., and Greene, G. M.: Osteomas of the internal auditory canal: a report of two cases. Am. J. Otol. 21:852, 2000. 30. Deltenre, P., Mansbach, A. L., Bozet, C., et al.: Auditory neuropathy: a report on three cases with early onsets and major neonatal illness. Electroencephalogr. Clin. Neurophysiol. 104:17, 1997. 31. Deltenre, P., Mansbach, A. L., Bozet, C., et al.: Auditory neuropathy with preserved cochlear microphonics and secondary loss of otoacoustic emissions. Audiology 38:187, 1999. 32. Dix, M. R., and Hallpike, C. S.: The pathology, symptomatology and diagnosis of certain common disorders of the vestibular system. Ann. Otol. Rhinol. Laryngol. 61:987, 1952.

1269

33. Driscoll, C. L., Lynn, S. G., Harner, S. G., et al.: Preoperative identification of patients at risk of developing persistent dysequilibrium after acoustic neuroma removal. Am. J. Otol. 19:491, 1998. 34. Eggermont, J. J., Brown, D. K., Ponton, C. W., and Kimberley, B. P.: Comparison of distortion product otoacoustic emission (DPOAE) and auditory brain stem response (ABR) traveling wave delay measurements suggests frequency-specific synapse maturation. Ear Hear. 17:386, 1996. 35. Ferraro, J. A., and Durrant, J. D.: Electrocochleography. In Katz, J. (ed.): Handbook of Clinical Audiology, 5th ed. Baltimore, Lippincott Williams & Wilkins, p. 249, 2002. 36. Fetter, M., and Dichgans, J.: Vestibular neuritis spares the inferior division of the vestibular nerve. Brain 119:755, 1996. 37. Fish, J. H., Scholz, A. W., Hussl, B., et al.: Cerebro-oculofacio-skeletal syndrome as a human example for accelerated cochlear degeneration. Otol. Neurotol. 22:170, 2001. 38. Friedman, I., Frohlich, A., and Wright, A.: Epidemic typhus fever and hearing loss: a histological study. J. Laryngol. Otol. 107:275, 1993. 39. Fujii, M., Goto, N., and Kikuchi, K.: Nerve fiber analysis and the aging process of the vestibulocochlear nerve. Ann. Otol. Rhinol. Laryngol. 99:863, 1990. 40. Fujikawa, S., and Starr, A.: Vestibular neuropathy accompanying auditory and peripheral neuropathies. Arch. Otolaryngol. Head Neck Surg. 126:1453, 2000. 41. Furman, J. M., Durrant, J. D., and Hirsch, W. L.: Eighth nerve signs in a case of multiple sclerosis. Am. J. Otolaryngol. 10:376, 1989. 42. Furuta, Y., Takasu, T., Fukuda, S., et al.: Latent herpes simplex virus type 1 in human vestibular ganglia. Acta Otolaryngol. Suppl. 503:85, 1993. 43. Fuse, T., and Moller, M. B.: Delayed and progressive hearing loss after microvascular decompression of cranial nerves. Ann. Otol. Rhinol. Laryngol. 105:158, 1996. 44. Gacec, R. R., and Gacec, R. M.: Menière’s disease as a manifestation of vestibular ganglionitis. Am. J. Otolaryngol. 22:241, 2001. 45. Gelfand, S. A.: The acoustic reflex. In Katz, J. (ed.): Handbook of Clinical Audiology, 5th ed. Baltimore, Lippincott Williams & Wilkins, p. 205, 2002. 46. Goebel, J. A., O’Mara, W., and Gianoli, G.: Anatomic considerations in vestibular neuritis. Otol. Neurotol. 22:512, 2001. 47. Grimaldi, L. M. E., Luzi, L., Martino, G. V., et al.: Bilateral eighth nerve neuropathy in human immunodeficiency virus infection. J. Neurol. 240:363, 1993. 48. Gutmann, D. H., Aylsworth, A., Carey, J. C., et al.: The diagnostic evaluation and multidisciplinary management of neurofibromatosis 1 and neurofibromatosis 2. JAMA 278:51, 1997. 49. Helmchen, C., Arbusow, V., Jager, L., et al.: Cogan’s syndrome: clinical significance of antibodies against the inner ear and cornea. Acta Otolaryngol. 119:528, 1999. 50. Herdman, S. J., Clendaniel, R. A., Mattox, D. E., et al.: Vestibular adaptation exercises and recovery: acute stage after acoustic neuroma resection. Otolaryngol. Head Neck Surg. 113:77, 1995. 51. Himi, T., Igarashi, M., Kataura, A., and Alford, B. R.: Temporal bone findings in craniodiaphyseal dysplasia. Auris Nasus Larynx 20:255, 1993.

1270

Diseases of the Peripheral Nervous System

52. Ishizaki, H., Pyykko, I., and Nozue, M.: Neuroborreliosis in the etiology of vestibular neuronitis. Acta Otolaryngol. Suppl. 503:67, 1993. 53. Jannetta, P. J., Moller, M. B., and Moller, A. R.: Disabling positional vertigo. N. Engl. J. Med. 310:1700, 1984. 54. Kalaydjieva, L., Gresham, D., Gooding, R., et al.: N-myc downstream-regulated gene 1 is mutated in hereditary motor and sensory neuropathy–Lom. Am. J. Hum. Genet. 67:47, 2000. 55. Kalaydjieva, L., Hallmayer, J., Chandler, D., et al.: Gene mapping in Gypsies identifies a novel demyelinating neuropathy on chromosome 8q24. Nat. Genet. 14:214, 1996. 56. Kemink, J. L., Telian, S. A., el Kashlan, H., and Langman, A. W.: Retrolabyrinthine vestibular nerve section: efficacy in disorders other than Menière’s disease. Laryngoscope 101:523, 1991. 57. Kemp, D. T.: Stimulated acoustic emissions from within the human auditory system. J. Acoust. Soc. Am. 64:1386, 1978. 58. Klinke, R., and Galley, N.: Efferent innervation of vestibular and auditory receptors. Physiol. Rev. 54:316, 1984. 59. Kovach, M. J., Campbell, K. C., Herman, K., et al.: Anticipation in a unique family with Charcot-Marie-Tooth syndrome and deafness: delineation of the clinical features and review of the literature. Am. J. Med. Genet. 108:295, 2002. 60. Lindsay, J. R., and Hemenway, W. G.: Postural vertigo due to sudden loss of vestibular function. Ann Otolaryngol. 65:692, 1956. 61. Lòpez-Bigas, N., Olivé, M., Rabionet, R., et al.: Connexin 31 (GJB3) is expressed in the peripheral and auditory nerves and causes neuropathy and hearing impairment. Hum. Mol. Genet. 10:947, 2001. 62. Macdonald, M. R., Harrison, R. V., Wake, M., et al.: Ototoxicity of carboplatin: comparing animal and clinical models at the Hospital for Sick Children. J. Otolaryngol. 23:151, 1994. 63. McCormick, D. P.: Herpes simplex virus as a cause of Bell’s palsy. Lancet 11:937, 1972. 64. McNicholl, J. M., Glynn, D., Mongey, A. B., et al.: Prospective study of neurophysiologic, neurologic and immunologic abnormalities in systemic lupus erythematosus. J. Rheumatol. 21:1061, 1994. 65. Merrin, P. K., and Macfarlane, D. G.: Vestibulocochlear dysfunction in a patient with rheumatoid disease and vasculitis. Ann. Rheum. Dis. 50:393, 1991. 66. Meynard, J. L., el Amrani, M., Meyohas, M. C., et al.: Two cases of cytomegalovirus infection revealed by hearing loss in HIV-infected patients. Biomed. Pharmacother. 51:461, 1997. 67. Modugno, G. C., Pirodda, A., Ferri, G. G., et al.: Small acoustic neuromas: monitoring the growth rate by MRI. Acta Neurochir. 141:1063, 1999. 68. Moffat, D. A., Saunders, J. E., McElveen, J. Jr., et al.: Unusual cerebello-pontine angle tumours. J. Laryngol. Otol. 107:1087, 1993. 69. Möller, A. R.: Neural generators of auditory evoked potentials. Semin. Hear. 19:11, 1994.

70. Moller, M. B.: Results of microvascular decompression of the eighth nerve as treatment for disabling positional vertigo. Ann. Otol. Rhinol. Laryngol. 99:724, 1990. 71. Moller, M. B., Moller, A. R., Jannetta, P. J., et al.: Microvascular decompression of the eighth nerve in patients with disabling positional vertigo: selection criteria and operative results in 207 patients. Acta Neurochir. 125:75, 1993. 72. Moore, J. K., and Linthicum, F. H. Jr.: Myelination of the human auditory nerve: different time courses for Schwann cell and glial myelin. Ann. Otol. Rhinol. Laryngol. 110:655, 2001. 73. Murofushi, T., Takegoshi, H., Ohki, M., and Ozeki, H.: Galvanic-evoked myogenic responses in patients with an absence of click-evoked vestibulo-collic reflexes. Clin. Neurophysiol. 113:305, 2002. 74. Nadol, J. B.: Vestibular neuritis. Otolaryngol. Head Neck Surg. 112:162, 1995. 75. Nadol, J. B.: Patterns of neural degeneration in the human cochlea and auditory nerve: implications for cochlear implantation. Otolaryngol. Head Neck Surg. 117:220, 1997. 76. Nadol, J. B.: Primary cochlear degeneration. In Sininger, Y., and Starr, A. (eds.): Auditory Neuropathy. San Diego, Singular, p. 99, 2001. 77. Nadol, J. B., and Thornton, A. R.: Ultrastructural findings in a case of Menière’s disease. Acta Otolaryngol. Suppl. 519:47, 1987. 78. Nadol, J. B., Young, Y. S., and Glynn, R. J.: Survival of spiral ganglion cells in profound sensorineural hearing loss: implications for cochlear implantation. Ann. Otol. Rhinol. Laryngol. 98:411, 1989. 79. Nagai, Y., Goto, N., Goto, J., et al.: Morphometric nerve fiber analysis and aging process of the human vestibular nerve. Okajimas Folia Anat. Jpn. 76:95, 1999. 80. Nager, G. T.: Acoustic neurinomas: pathology and differential diagnosis. Arch. Otolaryngol. 89:252, 1969. 81. Nakamura, K., Koda, J., and Koike, Y.: Stenosis of the internal auditory canal with VIIth and VIIIth cranial nerve dysfunctions. ORL J. Otorhinolaryngol. Relat. Spec. 61:16, 1999. 82. National Institutes of Health Consensus Development Conference. Neurofibromatosis: conference statement. Arch. Neurol. 45:575, 1988. 83. Nishizawa, S., Yokoyama, T., Ohta, S., et al.: Lipoma in the cerebellopontine angle—case report. Neurol. Med. Chir. 30:137, 1990. 84. Ogawa, K., Kanzaki, J., Ogawa, S., et al.: Acoustic neuromas presenting as sudden hearing loss. Acta Otolaryngol. Suppl. 487:138, 1991. 85. Ohbayashi, S., Oda, M., Yamamoto, M., et al.: Recovery of the vestibular function after vestibular neuronitis. Acta Otolaryngol. Suppl. 503:31, 1993. 86. Ohira, Y., Kaga, K., and Kodama, A.: A case of bilateral sudden hearing loss and vertigo caused by bilateral temporal bone metastasis from pancreatic carcinoma—comparison of clinical findings and temporal bone pathological findings. Nippon Jibiinkoka Gakkai Kaiho 94:9, 1991. 87. Okinaka, Y., Sekitani, T., Okazaki, H., et al.: Progress of caloric response of vestibular neuronitis. Acta Otolaryngol. Suppl. 503:18, 1993.

Diseases of the Eighth Cranial Nerve 88. O’Reilly, B. J., and Burrows, E. H.: Eighth cranial nerve involvement in sarcoidosis. J. Laryngol. Otol. 109:1089, 1995. 89. Parnes, L. S., Shimotakahara, S. G., Pelz, D., et al.: Vascular relationships of the vestibulocochlear nerve on magnetic resonance imaging. Am. J. Otol. 11:278, 1990. 90. Pauler, M., Schuknecht, H. F., and Thornton, A. R.: Correlative studies of cochlear neuronal loss with speech discrimination and pure-tone thresholds. Arch. Otolaryngol. 243:200, 1986. 91. Ponton, C. W., Moore, J. K., and Eggermont, J. J.: Auditory brain stem response generation by parallel pathways: differential maturation of axonal conduction time and synaptic transmission. Ear Hear. 17:402, 1996. 92. Ramsden, R. T.: Vestibular schwannoma. In Booth, J. B. (ed.): Otology, 6th ed., Vol. 3. London, Scott & Brown Otolaryngology/Read Educational & Professional Publishing, p. 1, 1997. 93. Rance, G., Beer, D. E., Cone-Wesson, B., et al.: Clinical findings for a group of infants and young children with auditory neuropathy. Ear Hear. 20:238, 1999. 94. Rance, G., Cone-Wesson, B., Wunderlich, J., and Dowell, R.: Speech perception and cortical event related potentials in children with auditory neuropathy. Ear Hear. 23:239, 2002. 95. Rogers, R., Kelley, P. M., Keats, B. J., et al.: Non-syndromic recessive auditory neuropathy due to mutations in otoferlin. Presented at the 25th annual mid-winter meeting of the Association for Research in Otolaryngology, Florida, 2002. 96. Ryan, S., Kemp, D. T., and Hinchcliffe, R.: The influence of contralateral acoustic stimulation on click evoked otoacoustic emissions in humans. Br. J. Audiol. 25:391, 1991. 97. Ryu, H., Yamamoto, S., Sugiyama, K., et al.: Neurovascular compression syndrome of the eighth cranial nerve: can the site of compression explain the symptoms? Acta Neurochir. 141:495, 1999. 98. Ryugo, D.: The auditory nerve. In Webster, D. B., Popper, A. N., and Fay, R. R. (eds.): The Mammalian Auditory Pathway: Neuroanatomy. New York, Springer-Verlag, p. 23, 1992. 99. Salvi, R. J., Wang, J., and Ding, D.: Auditory deprivation of the central auditory system resulting from selective inner hair cell loss: animal model of auditory neuropathy. Acta Audiol. Suppl. 51:1, 1999. 100. Sartoretti-Schefer, S., Scherler, M., Wichmann, W., and Valavanis, A.: Contrast-enhanced MR of the facial nerve in patients with posttraumatic peripheral facial nerve palsy. Am. J. Neuroradiol. 18:1115, 1997. 101. Saunders, J. C., Dear, S. P., and Schneider, M. E.: The anatomical consequences of acoustic injury: a review and tutorial. J. Acoust. Soc. Am. 78:833, 1985. 102. Saunders, J. E., Luxford, W. M., Devgan, K. K., and Fetterman, B. L.: Sudden hearing loss in acoustic neuroma patients. Otolaryngol. Head Neck Surg. 113:23, 1995. 103. Schuknecht, H. F.: Pathology of the Ear. Cambridge, MA, Harvard University Press, 1974. 104. Schuknecht, H. F., Allam, A. F., and Murakami, Y.: Pathology of secondary malignant tumours of the temporal bone. Ann. Otol. Rhinol. Laryngol. 77:4, 1968. 105. Schuknecht, H. F., and Gacec, M. R.: Cochlear pathology in presbyacusis. Ann. Otol. Rhinol. Laryngol. Suppl. 102:1, 1993.

1271

106. Schuknecht, H. F., and Kitamura, K.: Vestibular neuritis. Ann. Otol. Rhinol. Laryngol. Suppl. 78:1, 1981. 107. Schwaber, M. K., and Hall, J. W.: Cochleovestibular nerve compression syndrome. I. Clinical features and audiovestibular findings. Laryngoscope 102:1020, 1992. 108. Schwaber, M. K., and Whetsell, W. O.: Cochleovestibular nerve compression syndrome. II. Vestibular nerve histopathology and theory of pathophysiology. Laryngoscope 102:1030, 1992. 109. Sekitani, T., Imate, Y., Noguchi, T., and Inokuma, T.: Vestibular neuronitis: epidemiological survey by questionnaire in Japan. Acta Otolaryngol. Suppl. 503:9, 1993. 110. Sekiya, T., Hatayama, T., Shimamura, N., and Suzuki, S.: An in vivo quantifiable model of cochlear neuronal degeneration induced by central process injury. Exp. Neurol. 161:490, 2000. 111. Selesnick, S. H., Jackler, R. K., and Pitts, L. W.: The changing clinical presentation of acoustic tumors in the MRI era. Laryngoscope 103:431, 1993. 112. Senderek, J., Hermanns, B., Bergmann, C., et al.: X-linked dominant Charcot-Marie-Tooth neuropathy: clinical, electrophysiological, and morphological phenotype in four families with different connexin 32 mutations(1). J. Neurol. Sci. 167:90, 1999. 113. Shallop, J. K., Peterson, A., Facer, G. W., et al.: Cochlear implants in 5 cases of auditory neuropathy: postoperative findings and progress. Laryngoscope 111:555, 2001. 114. Shapiro, F.: Osteopetrosis: current clinical considerations. Clin. Orthop. 294:34, 1993. 115. Shinkawa, H., and Nadol, J. B.: Histopathology of the inner ear in Usher’s syndrome as observed by light and electron microscopy. Ann. Otol. Rhinol. Laryngol. 95:313, 1986. 116. Sininger, Y., and Oba, S.: Patients with auditory neuropathy: who are they and what can they hear? In Sininger, Y., and Starr, A. (eds.): Auditory Neuropathy. San Diego, Singular, p. 15, 2001. 117. Slack, R. W., Wright, A., Michaels, L., and Frohlich, S. A.: Inner hair cell loss and intracochlear clot in the preterm infant. Clin. Otolaryngol. 11:443, 1986. 118. Smith, M. M., and Anderson, J. C.: Neurosyphilis as a cause of facial and vestibulocochlear nerve dysfunction: MR imaging features. Am. J. Neuroradiol. 21:1673, 2000. 119. Spoendlin, H.: Optic and cochleovestibular degenerations in hereditary ataxias. Brain 97:41, 1974. 120. Spoendlin, H.: Factors inducing retrograde degeneration of the cochlear nerve. Ann. Otol. Rhinol. Laryngol. Suppl. 112:76, 1984. 121. Starr, A.: The neurology of auditory neuropathy. In Sininger, Y., and Starr, A. (eds.): Auditory Neuropathy. San Diego, Singular, p. 37, 2001. 122. Starr, A., Picton, T. W., and Kim, R.: Pathophysiology of auditory neuropathy. In Sininger, Y., and Starr, A. (eds.): Auditory Neuropathy. San Diego, Singular, p. 67, 2001. 123. Starr, A., Picton, T. W., Sininger, Y., et al.: Auditory neuropathy. Brain 119:741, 1996. 124. Starr, A., Sininger, Y., Winter, M., et al.: Transient deafness due to temperature-sensitive auditory neuropathy. Ear Hear. 19:169, 1998. 125. Straumann, D., and Zee, D. S.: Three dimensional aspects of eye movements. Curr. Opin. Neurol. 8:69, 1995.

1272

Diseases of the Peripheral Nervous System

126. Streitmann, M. J., and Sismanis, A.: Metastatic carcinoma of the temporal bone. Am. J. Otol. 17:780, 1996. 127. Strupp, M., Jägger, L., Maag, K. P., et al.: Vestibular exercises improve vestibulospinal compensation after vestibular neuritis. Neurology 51:838, 1998. 128. Suzuki, S.: Detection of latent herpes simplex virus in human vestibular ganglia. Hokkaido Igaku Zasshi 71:561, 1996. 129. Suzuki, Y., Kaga, K., Sugiuchi, Y., et al.: Sudden bilateral hearing loss due to gastric carcinoma and its histological evidence. J. Laryngol. Otol. 111:1142, 1997. 130. The Consensus Development Panel: National Institutes of Health Consensus Development Conference statement on acoustic neuroma, December 1991. Arch. Neurol. 51:201, 1994. 131. Thomsen, J.: Cerebellopontine angle tumours, other than acoustic neuromas: a report on 34 cases; a presentation of 7 bilateral acoustic neuromas. Acta Otolaryngol. 82:106, 1976. 132. Tsutsumi, T., Tsunoda, A., Noguchi, Y., and Komatsuzaki, A.: Prediction of the nerves of origin of vestibular schwannomas with vestibular evoked myogenic potentials. Am. J. Otol. 21:712, 2000. 133. Van Camp, G., and Smith, R. J. H.: Hereditary Hearing Loss Homepage. Available at www.uia.ac.be/dnalab/hhh/

134. Van de Water, T., Staecker, H., Apfel, S. C., and Lefebvre, P. P.: Regeneration of the auditory nerve: the role of neurotrophic factors. In Van de Water, T., Popper, A. N., and Fay, R. R. (eds.): Clinical Aspects of Hearing. New York, Springer-Verlag, p. 41, 1996. 135. Withnell, R. H.: Brief report: the cochlear microphonic as an indication of outer hair cell function. Ear Hear. 22:75, 2001. 136. Ylikoski, J. S., House, J. W., and Hernandez, I.: Eighth nerve alcoholic neuropathy: a case report with light and electron microscopic findings. J. Laryngol. Otol. 95:631, 1981. 137. Yoshikawa, H., Nishimura, T., Nakatsuji, Y., et al.: Elevated expression of messenger RNA for peripheral myelin protein 22 in biopsied peripheral nerves of patients with CharcotMarie-Tooth disease type 1A. Ann. Neurol. 35:445, 1994. 138. Zealley, I. A., Cooper, R. C., Clifford, K. M., et al.: MRI screening for acoustic neuroma: a comparison of fast spin echo and contrast enhanced imaging in 1233 patients. Br. J. Radiol. 73:242, 2000. 139. Zeng, F. G., Oba, S., and Garde, S.: Psychoacoustics and speech perception in auditory neuropathy. In Sininger, Y., and Starr, A. (eds.): Auditory Neuropathy. San Diego, Singular, p. 141, 2001. 140. Zwirner, P., and Wilichowski, E.: Progressive sensorineural hearing loss in children with mitochondrial encephalomyopathies. Laryngoscope 111:515, 2001.

52 Diseases of the Ninth, Tenth, Eleventh, and Twelfth Cranial Nerves P. K. THOMAS AND CHRISTOPHER J. MATHIAS

The Glossopharyngeal Nerve Anatomy Clinical Features of Glossopharyngeal Nerve Lesions Tests of Function Lesions Affecting the Glossopharyngeal Nerve The Vagus and Cranial Accessory Nerves Anatomy

Clinical Features of Vagus Nerve Lesions Tests of Vagal Autonomic Function Disorders of Vagal Autonomic Function Lesions Affecting the Vagus Nerve The Spinal Accessory Nerve Anatomy

THE GLOSSOPHARYNGEAL NERVE Anatomy The glossopharyngeal nerve has five or six slender fila or rootlets as it emerges from the medulla oblongata in a sulcus at the dorsal border of the inferior olive. After the fila have united, the nerve is separated from the Xth nerve by a dural isthmus as it passes through the jugular foramen. It possesses superior and inferior (petrous) ganglia. The larger inferior ganglion is located in a notch in the lower border of the petrous portion of the temporal bone; the superior ganglion lies on the nerve as it passes through the jugular foramen. The tympanic branch arises from the inferior ganglion, ascends to the middle ear, and becomes the lesser superficial petrosal nerve, which joins the otic ganglion. In the upper neck, the nerve lies in front of the Xth nerve and passes between the internal carotid artery and internal jugular vein before passing superficial to the internal carotid artery behind the styloid process. Thereafter it follows the posteroinferior part of the stylopharyngeus muscle, traveling across the lateral surface of this muscle and between the constrictors of the pharynx. Finally, before it divides, its course lies deep to the hyoglossus muscle. The nerve contains motor, parasympathetic, and sensory fibers.

Clinical Features of Spinal Accessory Nerve Lesions Lesions of the Spinal Accessory Nerve The Hypoglossal Nerve Anatomy Clinical Features of Hypoglossal Nerve Lesions Lesions of the Hypoglossal Nerve

Motor Fibers Motor fibers innervate the stylopharyngeus muscle, which has some part in the elevation of the pharynx. Their cells of origin form part of the nucleus ambiguus; studies in the cat show that these neurons form a compact mass of cells lying level with, but ventrolateral to, the rostral tip of the main column of the nucleus.132 Occasionally, the superior constrictor of the pharynx is supplied by this nerve rather than by the vagus. Parasympathetic Fibers The parasympathetic fibers mediate secretion of the parotid gland and have their cells of origin in the inferior salivary nucleus. They leave the main body of the nerve in the tympanic nerve, passing through the tympanic plexus and reaching the otic ganglion in the lesser superficial petrosal nerve. Here the fibers synapse with the postganglionic fibers to the parotid gland. Occasionally, however, the gland may be innervated by a nerve other than the glossopharyngeal.55 Sensory Fibers Visceral afferent fibers convey general sensation from the posterior third of the tongue, fauces, tonsils, nasopharynx, inferior surface of the soft palate, uvula, eustachian tube, and tympanic cavity. In addition, the nerve innervates a 1273

1274

Diseases of the Peripheral Nervous System

small cutaneous area in front of the tragus together with a little of the adjacent anterior wall of the outer auditory meatus.37 Taste afferent fibers in the glossopharyngeal nerve supply the posterior third of the tongue and taste buds in the pharynx. The cell bodies of these afferent fibers lie in the petrous ganglion. Studies in the cat and monkey have shown that these fibers terminate in the nucleus of the tractus solitarius, mainly in its rostral part but with a considerable caudal extension and with a connection to the contralateral commissural nucleus.124,178 The fibers terminating rostrally probably serve gustatory function, whereas those serving visceral function terminate more caudally. Carotid Sinus Nerve The carotid sinus nerve, another branch of the glossopharyngeal nerve, contains the primary afferent fibers of chemoreceptors in the carotid body and of baroreceptors lying in the carotid sinus wall. These fibers project to the middle third of the nucleus of the tractus solitarius.52 Some fibers also terminate on large neurons of the paramedian reticular formation of the medulla,153 but other studies have failed to reveal monosympathetic activation in this region.196 Evidence suggesting the presence of an efferent supply to the carotid body carried in the carotid sinus nerve22 has not been confirmed by retrograde transport studies.121 The baroreceptors are concerned in the regulation of blood pressure; the chemoreceptors are responsible predominantly for the ventilatory response to hypoxia but also to a lesser extent for the response to hypercapnia.

Clinical Features of Glossopharyngeal Nerve Lesions It is extremely rare for this nerve to be affected in isolation, the vagus almost always being involved as well. The symptoms related to the latter nerve usually dominate the clinical picture. The patient may be aware of difficulty in swallowing and disturbance of taste. On examination, sensation is found to be impaired on the affected side over the soft palate, pharynx, fauces, and posterior third of the tongue, and the gag reflex is reduced or absent. Taste is lost over the ipsilateral posterior third of the tongue. Salivary secretion from the parotid gland is usually reduced or absent, although in some cases it can be increased. In normal subjects bilateral local anesthesia of the glossopharyngeal and vagus nerves in the neck blocks both baroreceptor and chemoreceptor afferents, resulting in temporary hypertension, tachycardia,99 and an increase in muscle and skin sympathetic activity as measured by microneurography (Fig. 52–1).78 Transient hypertension occurs after bilateral surgical section of the glossopharyngeal nerves.223 Sustained hypertension has also been reported after unilateral section of the glossopharyngeal nerve,179 which was presumably the dominant baroreceptor afferent nerve.

In glossopharyngeal neuralgia, pain is experienced behind the angle of the jaw, deep within the ear, and in the side of the throat, as described later.

Tests of Function The integrity of taste perception on the posterior third of the tongue is most readily tested by galvanic stimulation. The parasympathetic secretomotor function can be assessed by measuring the secretion from Stensen’s duct over a 5-minute period after intravenous injection of pilocarpine. In healthy subjects, the secretion is equal on the two sides but the quantity varies from 3 to 15 mL. Asymmetrical amounts may indicate impairment of the secretory fibers. Chemoreceptor function can be examined by assessing the ventilatory response to mild hypoxia or to a few breaths of pure oxygen.60 Baroreceptor function can be tested by observing the blood pressure and heart rate changes in response to tilting, giving a pressor agent (norepinephrine, angiotensin, or phenylephrine), Valsalva maneuver,19 or external stimulation of the carotid sinus.66,67

Lesions Affecting the Glossopharyngeal Nerve Intramedullary Lesions Lesions affecting the nuclei of the glossopharyngeal nerve commonly also involve the Xth and VIIIth nerves and may cause vertigo, dysphagia, dysarthria, and contralateral hemiplegia as well as the features of IXth nerve dysfunction (Bonnier syndrome). Causes include vascular, neoplastic, and inflammatory lesions and, rarely, syringobulbia. Extramedullary Lesions The ninth nerve can be involved by lesions in the cerebellopontine angle, in which the eighth nerve deficit usually dominates the clinical picture. A neurinoma here may mimic an acoustic tumor.82,156 More peripherally, the nerve can be affected by lesions in the jugular foramen, where the Xth and XIth nerves will also be involved, giving rise to the features of the jugular foramen syndrome (Vernet’s syndrome). Glomus tumors are probably the most common cause, but a ninth nerve neurinoma may arise at this site.89 Other tumors involving the nerve here include metastases and chordomas. The nerve may be damaged with basal fractures of the skull. Paralysis may occur in diphtheria. Neoplastic or inflammatory involvement of the meninges may affect the ninth nerve as part of a polyneuritis cranialis, and the nerve is occasionally affected in acute inflammatory polyneuritis (Guillain-Barré syndrome). In tabes dorsalis and diabetes mellitus, the baroreceptor reflex may be disrupted because of involvement of carotid and aortic baroreceptor afferent fibers.192 In tabes dorsalis there is an impaired ventilatory response to hypoxia, implying involvement of chemoreceptor afferent fibers.75

Diseases of the Ninth through Twelfth Cranial Nerves

1275

FIGURE 52–1 Excerpts of records from a normal subject who underwent unilateral and then bilateral glossopharyngeal and vagus nerve blockade with a local anesthetic. These show, from 1 to 6, the electrocardiogram (ECG), muscle sympathetic activity (MSA), skin sympathetic activity (SSA), and intra-arterial blood pressure (BP). The time scale in each panel is identical. MSA and SSA are expressed as mean voltage neurograms. In 4 and 5 the arrow indicates the time after bilateral injection of anesthetic. In 2, atropine causes tachycardia and a slight increase in blood pressure, with a decrease in MSA. After unilateral nerve block there is an increase in MSA and SSA. Bilateral block causes a marked increase in MSA, along with tachycardia and a considerable rise in blood pressure. (From Fagius, J., Wallin, G. B., Sundlof, G., et al.: Sympathetic outflow in man after anaesthesia of the glossopharyngeal and vagus nerves. Brain 108:423, 1985, with permission.)

In the Guillain-Barré syndrome, hypertension and tachycardia may occur, and microneurographic recordings of peripheral muscle sympathetic activity indicate a pronounced increase during the illness with return to normal during recovery.77 These changes are similar to those observed after bilateral glossopharyngeal and vagal block-

ade in normal humans78 and suggest that the lesion is likely to involve the baroreceptor afferents. In the syndrome of crocodile tears (paroxysmal lacrimation, gustolacrimal reflex, Bogorad syndrome27), abnormal lacrimation is associated with the normal production of saliva. It occurs after injury or operations on the ear that

1276

Diseases of the Peripheral Nervous System

damage the lesser superficial petrosal nerve supplying the parotid gland.12 These are postganglionic parasympathetic fibers from the otic ganglion, which receives preganglionic fibers from the glossopharyngeal nerve via the tympanic plexus. Sprouting fibers in regenerating nerves may crossinnervate the greater superficial petrosal nerve to the lacrimal gland, because it is in close proximity. The syndrome also complicates Bell’s palsy with lesions proximal to the geniculate ganglion of the seventh nerve, where the preganglionic parasympathetic fibers to the lacrimal, nasal, palatine, pharyngeal, submandibular, and sublingual glands travel together before separating.32 Damage may cause cross-innervation of preganglionic fibers supplying the salivary and lacrimal glands. If anesthetic blockade of the glossopharyngeal nerve at the jugular foramen is successful, surgical section of its tympanic branch through the eardrum may be curative.92 If not, the sphenopalatine ganglion (from which the postganglionic parasympathetic supply of the lacrimal glands emerges) can be injected with a local anesthetic; if this is successful, ablation can be performed with absolute alcohol. Glossopharyngeal Neuralgia This condition may be idiopathic or secondary to a local structural lesion. Weisenburg first described the tic in 1910 in a patient with a cerebellopontine angle tumor.220 The idiopathic form was termed glossopharyngeal neuralgia by Harris in 1921.103 The clinical features have been comprehensively reviewed by Rushton and co-workers183 and Bruyn.36 There is an equal sex ratio with a peak age of onset between 40 and 60 years. It may rarely occur in chil-

dren.122 The pain is stabbing in quality and almost always unilateral. It is usually situated at the base of the tongue and faucial region on one side, but it may radiate into the vagal territory. Thus the pain may be felt in or behind the external auditory meatus and beneath the angle of the jaw. For this reason, some authorities have preferred the term vasoglossopharyngeal neuralgia.183,224 Trigeminal and glossopharyngeal neuralgia may occur in combination.237 A hacking cough may accompany the paroxysms, perhaps as an attempt to relieve the pain, and pain may be precipitated by swallowing, talking, pressure on the tragus or the auricle, sneezing, coughing, and moving the head. Salivation, flushing, sweating, tinnitus, lacrimation, tachycardia, hypertension, and even vertigo may be associated. Cardiac asystole and seizures have also been reported.6,88,115,116 Bradycardia and syncope presumably result from abnormal afferent impulses, which via the brainstem vasomotor centers cause excessive stimulation of the vagus. A cardiac pacemaker may prevent bradycardia but not the fall in blood pressure and syncope (Fig. 52–2),6,215 indicating that withdrawal of sympathetic activity also contributes, as in the vasovagal syndrome.214 Bouts of pain occur on average two or three times a year and last from a few minutes up to several days. With the passage of time, the bouts tend to increase in frequency. The etiology of the idiopathic form of glossopharyngeal neuralgia is obscure. The possible significance of “cross talk” resulting from ephaptic excitation is discussed in a theoretical paper by Gardner.87 The symptomatic or secondary type is commonly due to tumor. Dandy estimated that tumors in the peripheral distribution of the nerve or in

FIGURE 52–2 The electrocardiogram (upper trace), muscle sympathetic activity (as mean voltage neurogram; middle trace) and intra-arterial blood pressure (lower trace) before, during, and after an attack of glossopharyngeal neuralgia associated with syncope. Attacks were elicited when the patient coughed or yawned but at times were observed without a triggering event. The horizontal bar indicates when the cardiac pacemaker was activated; despite this, the blood pressure fell and sympathetic bursts were considerably reduced. Termination of syncope occurred spontaneously, with an increase in sympathetic activity, restoration of endogenous heart rate, and a gradual rise in blood pressure to original levels. (From Wallin, B. G., Westerberg, C. E., and Sundlof, G.: Syncope induced by glossopharyngeal neuralgia: sympathetic outflow to muscle. Neurology 34:522, 1984, permission.)

1277

Diseases of the Ninth through Twelfth Cranial Nerves

the posterior fossa accounted for a quarter of the cases.54 Other reported causes include local infection, trauma to the neck, elongation of the styloid process, occlusion of the internal carotid artery,6 and tortuosity of the vertebral arteries or posterior inferior cerebellar artery.130 The disorder is usually cured by nerve section. Medullary tractotomy has also been employed.128 Carbamazepine has been effective72,186 and should be the initial line of treatment.

Swallow syncope is a related disorder that is largely dependent on vagal reflex pathways but is sometimes associated with pathologic change in the glossopharyngeal nerve. The literature has been reviewed by Levin and Posner, who described a case of interest in the context of this discussion.134 Postmortem examination of their patient showed neoplastic infiltration of the glossopharyngeal nerve as well as degenerative changes in the vagus nerve.

Carotid Sinus Hypersensitivity The first observations on the carotid sinus reflex were by Parry,167 who noted dizziness and slowing of the heart following pressure on the bifurcation of the common carotid artery. Mechanical stimulation of the carotid body by massage normally causes only a small fall in heart rate and in blood pressure. In carotid sinus hypersensitivity, syncope is often associated with movement of the head and neck, wearing a tight collar, or stretching the skin of the neck during shaving.138,204,221 A suggestive history needs to be confirmed by objective findings, and hypersensitivity is defined as the response to carotid massage that lowers heart rate by 50%, with asystole exceeding 3 seconds in duration and/or a fall in blood pressure of over 50 mm Hg. In some the heart rate falls (the cardioinhibitory form), in others there is also a reduction in blood pressure that remains despite maintaining heart rate (the mixed variety), and rarely there is hypotension alone (the vasodepressor variety).216,239 Carotid sinus hypersensitivity has been associated with elderly patients; with cardiovascular disease; with drugs such as digoxin, ␣-methyldopa, and ␤-adrenergic blockers; and with Takayasu’s disease, biliary disease, carotid body tumor, local tumor infiltration around the sinus, and vagal lesions.17,138,162,176 In the postoperative period after carotid endarterectomy, hypotension may be related to changes in the mechanical properties of the arterial wall that stimulate the carotid sinus nerve; this can be prevented by a local anesthetic.9 It should be noted that syncope is experienced only by a minority with hyperactive carotid sinus reflexes. There are, however, potential dangers in eliciting the reflex,4,34 including ventricular fibrillation in addition to asystole. The procedure should be performed in a laboratory with adequate resuscitation facilities. Various treatments have been used.138,204 Muscarinic cholinergic blockade with atropine or propantheline may be effective, especially in the predominantly cardioinhibitory form.166 Demand cardiac pacemakers are of value,147 especially with associated sinus or atrioventricular node disease. Cervical irradiation has been used, and surgical denervation of the carotid sinus may be necessary.209 If bilateral, this may result in a moderate elevation in blood pressure. Fludrocortisone and ephedrine may be of some help in the vasodepressor or mixed variety,176,204 if carotid sinus denervation is not a practical option or has been refused.

THE VAGUS AND CRANIAL ACCESSORY NERVES Anatomy The cranial accessory nerve can be considered to constitute an aberrant bundle of the vagus nerve, and the two are therefore discussed together. The rootlets of the vagus nerve emerge from the medulla at the sulcus immediately dorsal to the prominence of the inferior olive and unite to form the nerve trunk. The rootlets of the cranial accessory nerve emerge slightly more caudally and fuse with the spinal accessory nerve, which has ascended through the foramen magnum, although they remain as a discrete fascicle. The vagus nerve emerges from the skull through the jugular foramen within the same dural sleeve as the accessory nerve, at which site is situated the jugular (sensory) ganglion. A meningeal branch arises from the jugular ganglion and is distributed to the posterior cranial fossa. The auricular branch (Arnold’s nerve) of the vagus leaves distal to the ganglion; it traverses the mastoid process and finally supplies cutaneous sensory fibers to the conch of the external ear. The larger nodose ganglion, also sensory, lies on the nerve just below the jugular foramen. At this site, a meningeal branch leaves, ascends through the hypoglossal canal, and supplies the dura in the region of the foramen magnum. At the inferior ganglion, the vagus is joined by the internal ramus of the accessory nerve, which constitutes the latter’s cranial component. At this point also, the vagus nerve gives rise to its pharyngeal branch, which joins the pharyngeal plexus and is distributed to the levator veli palatini muscle and the three paired constrictors of the pharynx. In the neck, the nerve lies within a sheath that also encloses the internal carotid artery and the internal jugular vein. The cardiac fibers are given off at about the level of the thoracic inlet. They join the sympathetic postganglionic fibers from the three cervical sympathetic ganglia, and these autonomic nerves, with the corresponding nerves from the other side, form the cardiac plexus on the anterior aspect of the bifurcation of the trachea from which the heart receives its nerve supply. On the right side the rest of the vagus nerve passes behind the innominate vein and then behind the superior vena cava to lie on the right side of the trachea. On the posterior aspect of the root of the right lung, it splits into a

1278

Diseases of the Peripheral Nervous System

posterior and smaller anterior part that together form an important part of the pulmonary plexus that provides the lung with its vagal innervation. At the lower edge of the lung root, the nerve reforms, passes down close to the esophagus, and at the level of the diaphragm is almost on the posterior aspect. On the left side the nerve, after entering the chest, crosses the lateral aspect of the aortic arch and continues beside the trachea to take part in the pulmonary plexus on the right side. After reforming at the lower edge of the lung root, it passes down close to the esophagus and at the level of the diaphragm is largely on the anterior aspect. There is, however, substantial interchange between the vagi as they accompany the lower esophagus. A morphometric study has been performed on the abdominal vagus in humans. The nerve is composed almost entirely of unmyelinated fibers and, interestingly, shows little alteration with aging, in contradistinction to limb nerves.191 The upper part of the esophagus has striated muscle in its wall. The lower part has smooth muscle that is innervated from the vagus as it passes down beside it, the preganglionic fibers synapsing with the ganglion cells of the esophageal plexus. In the abdomen the vagus supplies the stomach, liver, gallbladder, bile ducts, small intestine, pancreas, and colon almost to the splenic flexure. In the intestines the preganglionic fibers synapse in the enteric plexuses. The superior laryngeal nerve arises from the vagus in the region of the nodose ganglion and subdivides into the external laryngeal branch, which supplies the cricothyroid muscle, and the internal laryngeal branch, which pierces the thyrohyoid membrane and delivers sensory fibers to the larynx. The recurrent laryngeal nerve follows a different course on the two sides. On the right, it arises in the root of the neck and passes deep to the subclavian artery; on the left it arises in the upper thorax and passes deep to the aortic arch. On both sides it gains the groove between the trachea and the esophagus and ascends the neck to reach the larynx, where it supplies all the laryngeal muscles with the exception of the cricothyroid. It divides outside the larynx into two to six motor (anterolateral) and sensory (posteromedial) branches.229 There appears to be no constant intraneural topography within the recurrent laryngeal nerve,198 separate nerve bundles to different muscles not having been traced.28 In the horse, this appears to occur at about 10 to 15 cm from the larynx.65 Humphreys, in a study of the recurrent laryngeal nerve during thyroidectomy in 1615 patients, found that this nerve arose from the vagus at the level of the larynx and pursued a nonrecurrent course in 11 cases.111 The vagus nerve is structurally complex, with somatic sensory and visceral afferent and efferent components. Somatic Sensory Fibers Somatic sensory fibers are distributed to the skin of the concha of the external ear and travel in the auricular

branch. This nerve in the cat contains a high proportion of myelinated fibers of large diameter whose cells of origin appear to be in the jugular ganglion.63 On entering the brainstem, they reach the trigeminal sensory nucleus. Other somatic sensory fibers supply the meninges of the posterior fossa and foramen magnum. Visceral Afferent Fibers Visceral afferent fibers carry impulses from the pharynx, larynx, trachea, esophagus, and thoracic and abdominal viscera, including a small gustatory contribution from the epiglottis and vallecula. Those carrying pain from the larynx probably terminate in the spinal nucleus of the trigeminal nerve,33 the remainder mainly being distributed to the nucleus of the tractus solitarius. Their cell bodies lie in the nodose ganglion. The sensory innervation of the larynx has given rise to discussion, but it is now evident that spindles are present in all the laryngeal muscles of humans139 and that tendon organs and joint endings are also present.29 Their presence had been doubted because of the unimodal distribution of myelinated fiber diameter in the recurrent laryngeal nerve. However, Scheuer found fibers greater than 12 ␮m in diameter in both the internal and recurrent nerves in humans and obtained a bimodal distribution if the values for the two nerves were combined.187 She considered that muscle afferents might travel centrally in both these nerves in humans. Their central connections are uncertain. In the cat and rabbit their cell bodies are situated mainly in the nodose ganglion.29 Visceral Efferent Fibers Motor fibers take origin in the nucleus ambiguus from fibers that are often referred to as “special visceral efferents” because they supply striated musculature derived from the brachial arches. Some of the fibers leave the medulla in the cranial portion of the XIth nerve and reach the Xth nerve through the internal ramus of the XIth nerve. They innervate striated muscle in the palate, pharynx, larynx, and upper esophagus. The laryngeal muscles are supplied by cells from the caudal part of the nucleus ambiguus, and there is evidence to indicate cell groupings related to individual laryngeal muscles.132,201 Some fibers cross the midline to emerge with the nerves of the opposite side.152 The preganglionic parasympathetic fibers of this nerve that are distributed to the heart arise in cells close to the nucleus ambiguus. The remaining preganglionic autonomic fibers arise in the dorsal motor nucleus of the vagus in the floor of the fourth ventricle.151,152 This nucleus is especially concerned with the alimentary tract and is immediately inferior to the inferior salivary nucleus from which the parotid gland is innervated through the ninth cranial nerve. Superior to the inferior salivary nucleus is the superior salivary nucleus, concerned in innervation of the submandibular and sublingual salivary glands through the seventh cranial nerve. The cardiac fibers pass backward at first from the vicinity of the nucleus ambiguus toward the floor of the fourth

Diseases of the Ninth through Twelfth Cranial Nerves

ventricle. They then pass laterally with the noncardiac fibers, and all emerge from the medulla in the rootlets of the vagus nerve.

Clinical Features of Vagus Nerve Lesions Palatal and Pharyngeal Paralysis During normal swallowing, the nasopharynx is occluded by elevation of the palate and, at the same time, the epiglottis is raised and the inlet to the larynx is closed. Food projected from the mouth to the oropharynx is then transferred to the esophagus by the peristaltic action of the pharyngeal constrictors. Damage to the vagus nerve central to its ganglia or lesions of the pharyngeal branches result in paralysis of the levator veli palatini muscles and of the pharyngeal constrictors. Unilateral palatal paralysis gives rise to defective elevation of the palate on phonation and movement of the uvula toward the opposite side. It is usually asymptomatic but may lead to snoring at night and, in singers, to slight alterations in phonation. With bilateral palatal palsy, the nasopharynx is not closed during swallowing or phonation, leading to nasality of speech, nasal regurgitation during swallowing, and snoring when asleep. Unilateral pharyngeal paralysis produces drooping of the pharyngeal wall of the affected side and a “curtain” movement of the pharynx to the opposite side on phonation. There may be slight difficulty in clearing secretions from the throat and some interference with swallowing.208 Bilateral lesions give rise to severe dysphagia. Laryngeal Paralysis The larynx opens to allow breathing and closes during swallowing. Abduction of the vocal cords takes place during inspiration; the cords are adducted during phonation and coughing. The normal separation of the cords during respiration is about 13.5 mm, but in the fully adducted paramedian position there is a narrow separation only. The position assumed by the completely paralyzed cords (cadaveric position) is almost halfway between abduction and adduction, leaving a gap of approximately 7 mm. Unilateral Paralysis. Unilateral lesions of the recurrent laryngeal nerve result in paralysis of all the laryngeal muscles with the exception of the cricothyroid. They may be entirely asymptomatic or give rise to transient or sometimes persistent hoarseness of the voice. The occurrence of hoarseness depends on the ability of the opposite cord to compensate if the paralyzed cord is not in a median position. The affected cord usually adopts a median or paramedian position and, because the superior laryngeal nerve is intact, the cricothyroid muscle is able to exert action as a tensor of the cord and also, to some extent, as an adductor. With partial lesions of the recurrent laryngeal nerve, movements of abduction may be abolished before those of

1279

adduction (Semon’s law). Suggestions that the fibers innervating the abductor muscles occupy a particular position in the nerve that may be more vulnerable have not been confirmed by anatomic studies.198 Moreover, the electromyographic studies of Dedo and Hall have failed to substantiate the concept of “abductor” and “adductor” paralysis.58,59 Ellis has argued that whether the cord adopts a paramedian position or not depends on the activity of the cricothyroid muscle and on individual variations in the anatomy of the larynx.74 The relative bulk of the abductors and the adductors may be important, the weight of the abductors being only a quarter that of the adductors.30 When there is persisting paralysis of the cord, because of fibrotic contracture, the cord almost invariably comes ultimately to lie near the midline and at a lower level than the unparalyzed cord. If the superior laryngeal nerve is involved in addition, total paralysis of the cord on the affected side results, with the cord lying in the cadaveric position. Under these circumstances, there is usually associated palatal and pharyngeal paralysis because the lesion is likely to be central to the nodose ganglion. The voice is weak and husky and tires easily. An isolated lesion of the superior laryngeal nerve must be a very rare eventuality, but it has been described as a consequence of surgical or accidental trauma. Bilateral Paralysis. The consequences of bilateral lesions of the recurrent laryngeal nerves depend on the degree of approximation of the cords. Close approximation of the cords tends to result from acute lesions, such as those that follow thyroidectomy, whereas lesions of insidious onset tend to give rise to a less close apposition. The nearer the cords are to the midline, the better is the voice but the greater is the limitation of the airway. Acute lesions may necessitate tracheostomy, whereas insidious lesions may merely result in dyspnea and stridor on exertion, although the voice will be more severely affected. Nevertheless, paralysis with approximation of the cords necessitating tracheostomy is occasionally encountered in chronic polyneuropathies, and it has been recorded in the absence of widespread neuropathy in an alcoholic individual.210 Bilateral lesions affecting the superior laryngeal nerves usually imply damage above the nodose ganglia, and there is associated palatal and pharyngeal paralysis. The cords lie in the cadaveric position. Phonation is almost absent and is associated with much air wastage, and vocal pitch cannot be altered. There is an adequate airway for normal activities. Coughing is inefficient.

Tests of Vagal Autonomic Function The vagus nerve supplies a number of organs in the chest and abdomen, and various tests are used to assess cardiac innervation and gastrointestinal function. The former are described in Chapter 44. The latter are described here.

1280

Diseases of the Peripheral Nervous System

Esophageal, Gastric, and Intestinal Motility Esophageal movement can be observed using either radiologic techniques or manometry.127,165 In esophageal denervation there is supersensitivity to cholinomimetic agents,38 as observed after the injection of methacholine, carbachol, or bethanechol, which act directly on muscarinic receptors, especially in the lower third of the esophagus, which contains smooth muscle. In achalasia of the cardia these drugs cause strong but prolonged uncoordinated and occasionally painful contractions of the esophagus; in normal subjects or in esophageal dysfunction from other causes there is no motility change.127,165 Gastroparesis may occur in diabetes mellitus because of vagal impairment.97 The reverse may occur in patients who have undergone vagotomy and drainage operations, such as gastrojejunostomy and pyloroplasty for peptic ulcer disease.133 Gastric emptying may be assessed using either a standard barium meal or techniques that incorporate radiolabeled (gamma-emitting) solid or liquid meals with progress through the gut followed on a scintiscan gamma camera. Ultrasound and epigastric impedance techniques90,199 are being increasingly used. In gastroparesis, for instance, these noninvasive procedures can be repeated after administration of drugs such as metoclopramide, cisapride, and erythromycin to determine their therapeutic value. Increased gastric emptying (or gastric incontinence), because of altered gastric drainage, may result in the dumping syndrome,133,141,170,180 which is often associated with meals with a high sugar content. The early dumping syndrome is characterized by postprandial weakness, lightheadedness, flushing, sweating, palpitations, and tachycardia resulting from the rapid entry of hyperosmolar contents into the jejunum; this reduces blood volume by its osmotic effects, as indicated by a rise in the hematocrit. In the late dumping syndrome (2 to 4 hours later), the entry of large amounts of carbohydrate and subsequent absorption of glucose cause a later surge of insulin, resulting in reactive hypoglycemia. Diarrhea may complicate the dumping syndrome and is usually associated with truncal vagotomy, which is more likely to cause abnormalities of small bowel motility. Intestinal motility can be assessed by a range of methods, including duodenojejunal manometry and use of radiocapsules with telemetry to observe myoenteric migrating complexes (MMCs).206 Absence of MMCs and the response to food may be due either to a neuropathy or to intestinal muscle impairment. Radionuclides are used to determine small bowel transit. Hydrogen can be readily measured in breath and assesses small bowel transit time because it is related to bacterial formation of hydrogen when indigestible solids reach the colon. Gastrointestinal electromyography is of value in separating neural lesions from the rare cases of gastrointestinal myopathy.

Gastric Acid Secretion The vagus is concerned with secretion of gastric acid, and a potent stimulus is hypoglycemia, which can be induced by insulin (0.15 unit/kg body weight intravenously) within a half-hour. In vagal failure, gastric acidity does not increase although blood sugar falls below 50 mg/dL.108 To exclude achlorhydria caused by parietal cell abnormalities and atrophic gastritis, direct stimulation of acid secretion can be assessed after administration of pentagastrin.109 Enteroinsular Peptide Release Pancreatic polypeptide release is dependent on vagal activity.190 Pancreatic polypeptide levels in plasma rise with insulin-induced hypoglycemia, which increases vagally mediated cholinergic activity,190 and this is prevented by anticholinergic drugs and truncal vagotomy. Its measurement in plasma is thus a useful marker of abdominal vagal activity. In autonomic failure with vagal gut denervation there are diminished responses to hypoglycemia.173 In emotionally induced vasovagal syncope, levels rise markedly.184

Disorders of Vagal Autonomic Function Vagal Efferents Underactivity. The neural control of heart rate is predominantly exerted through the vagus. In the absence of vagal cardiac tone, heart rate rises to 100 to 110 beats/min, as in normal subjects given atropine. Diminished cardiac vagal activity is often seen in the two major groups with chronic primary autonomic failure,15 in multiple system atrophy (Shy-Drager syndrome), with lesions in the brainstem vagal nuclei, and in pure autonomic failure, where they may be predominantly peripheral. The resting heart rate, however, is not usually elevated, probably as a result of concomitant sympathetic impairment. In longitudinal studies of diabetic patients, heart rate is initially high218 but later falls,76 presumably because the cardiac vagi are affected before cardiac sympathetic nerves. In diabetes mellitus and chronic autonomic failure, there may be subnormal or absent heart rate responses to deep breathing and hyperventilation, which are tests designed to assess cardiac vagal efferent activity. Similar abnormalities may occur in polyneuropathy from other causes, including alcoholism.64 Vagal impairment occurs in a large proportion of patients with both alcoholic and nonalcoholic liver disease.207 In these disorders there appears to be a greater and earlier impairment of vagal as compared to sympathetic function; the reasons for this remain to be elucidated. Cardiac parasympathetic function tests are abnormal in a number of disorders, ranging from obesity181 to ocular hypertension,45 but their relationship to pathophysiologic processes and clinical outcome needs to be defined further. In cardiac transplantees there is both vagal and sympathetic cardiac denervation238; this may

Diseases of the Ninth through Twelfth Cranial Nerves

increase the possibility of cardiac dysrhythmias because of supersensitivity to circulating sympathetic agonists in the absence of cardiac vagal control. Impairment of vagal innervation of the gastrointestinal tract may result in a variety of functional abnormalities. In chronic primary autonomic failure, this may contribute to increased gastric emptying (Fig. 52–3)146 and an impaired pancreatic polypeptide response to hypoglycemia.173 In diabetes mellitus the vagal lesions may be more peripheral, as shown by electron microscopy; gastric dilatation with reduced peristalsis and delayed emptying (gastroparesis diabeticorum) may occur.97,110 A watery diarrhea usually without malabsorption may complicate diabetic autonomic neuropathy; it is often nocturnal and accompanied by borborygmi and abdominal discomfort. The precise cause is often unclear and may include intestinal stasis with bacterial overgrowth, which may explain the success

FIGURE 52–3 Gastric emptying curves, using a technetium-labeled meal with scintiscanning of gamma emission, in a normal subject (top) and in a patient with chronic autonomic failure (bottom). Integrated counts are indicated on the vertical axis and time in seconds on the horizontal axis. Computer exponential (Exp) fits are provided. In autonomic failure there is rapid emptying initially (Exp. fit I) with a later slower phase (Exp. fit II). (From Mathias, C. J., Da Costa, D. F., and Bannister, R.: Postcibal hypotension in autonomic disorders. In Bannister, R. [ed.]: Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 2nd ed. Oxford, Oxford University Press, p. 367, 1988, with permission.)

1281

of antibiotics such as tetracycline and metronidazole in some patients. The mechanism for the improvement with erythromycin may differ (see Chapter 85).117 Widespread gastrointestinal atony may occur in severe botulism, along with skeletal muscle paralysis and cardiovascular autonomic defects.13,212 A variant of botulinum toxicity, with only cholinergic autonomic involvement, has been reported.118 Botulinum toxin binds to cholinergic sites and through endocytosis enters the nerve terminal, resulting in damage to the presynaptic calcium-associated quantal release of acetylcholine203; this may also include the inactivation of choline acetyltransferase.40 The damage is persistent and recovery may take several weeks because it is dependent on sprouting of axon terminals with the formation of new neuromuscular terminals. Parenteral feeding is needed if there is paralytic ileus, and in the early stages of recovery the use of cholinesterase inhibitors, such as neostigmine or pyridostigmine, may aid return of gastrointestinal tone and function. Similar problems may occur in the rare syndromes of acute pure cholinergic dysautonomia.7,114,205 The lesion appears to be presynaptic, because the responses to cholinomimetic agents such as carbachol and bethanechol are preserved and even excessive, indicating functional or supersensitive postsynaptic receptors. The cause is not known. In one case, immunoglobulin G (IgG) antibodies in a skin biopsy specimen were thought to be directed against sudomotor postganglionic cholinergic fibers surrounding sweat glands,102 whereas in another there were immunologic abnormalities (low complement levels, diffuse pattern antinuclear antibodies, and positive anti– SS-A antibodies), favoring an immune disorder.202 Various toxins and drugs can impair or paralyze the vagus, but they usually involve other cholinergic systems.149 In certain disorders the intrinsic plexuses in the gut may be involved. In achalasia of the cardia, the myenteric plexus in the lower esophageal sphincter degenerates and fails to relax, with a supersensitive response to methacholine. In Chagas’ disease resulting from Trypanosoma cruzi infection, the esophagus and other parts of the alimentary tract, including the bile passages, may be affected.81,125,126 It was originally thought that a neurotoxin from dead parasites was responsible. The delay between the acute infective phase and later complications is now recognized as due to a cellular-type immune response, with the production of an anti–parasympathetic neuron IgG.143 This results in destruction of the intrinsic plexuses associated with lymphocytic infiltration. Overactivity. Overactivity of the cardiac vagal nerves may result in bradycardia and cardiac arrest, and this is of importance in the genesis of various forms of syncope ranging from emotional or common faints to carotid sinus syncope. It is often reflexly mediated, and the afferent “stimulus” could arise from a number of sites that normally cause only modest increases in vagal tone, such as pressure

1282

Diseases of the Peripheral Nervous System

over the eyeballs (oculocardiac reflex). The bradycardia may occasionally be severe enough to cause loss of consciousness or death.157 A number of trigeminal vagal reflexes that result in bradycardia and hypotension and are influenced by anesthetic agents and drugs have been described.234 Glossopharyngeal neuralgia, swallow syncope, and carotid sinus syncope are other examples. In emotional syncope the precipitant that induces mental stress varies considerably, for example, from venipuncture itself to the sight or thought of a needle and syringe. Vagal efferent blockade, with the muscarinic blockers atropine or propantheline, may prevent bradycardia and syncope. In carotid sinus supersensitivity, a demand pacemaker may help. Reflex withdrawal of sympathetic tone may additionally be involved, as has been demonstrated with microneurography in emotional syncope.214 In patients with human immunodeficiency virus infection and acquired immunodeficiency syndrome, syncope during pleural aspiration may occur; whether this is related to increased vagal efferent activity in the absence of normal sympathetic activity is unclear.53 Severe bradycardia and cardiac arrest, resulting from increased cardiac vagal activity, may complicate the management of patients with high cervical cord lesions who are receiving artificial ventilation. In such patients disconnection of the respirator and tracheal suction may precipitate a vagovagal reflex (Fig. 52–4).84,144 The increased efferent vagal activity can be prevented by atropine. Sensitivity of the vagus may be increased by hypoxia, pulmonary emboli, and acidemia, which may complicate management. Lack of cardiac sympathetic activity (because of the cord lesion) and paralysis of respiration (which normally activates the pulmonary stretch reflex to inhibit vagal activity and thus increase heart rate) are important factors contributing to the problem. Similar pathophysiologic processes probably account for bradycardia and cardiac arrest in chronically injured tetraplegics undergoing general anesthesia and tracheal intubation (Fig. 52–5).222 In children, cardiac vagal tone is greater (vagotonia) and may contribute directly to their death in certain situations. This may occur when a child falls into cold water, however shallow, because this may trigger an exaggerated vagal response, especially if it also induces apnea. Increased vagal activity may contribute to the sudden infant death syndrome, because in some near deaths there is an exaggerated cardiac response to ocular compression.119 Volatile substance abuse can enhance vagal activity and may increase the propensity to cardiac arrhythmias and sudden death.193 Abnormal Regeneration. Regeneration may occur after damage to the vagus and glossopharyngeal nerves and result in aberrant innervation, an example being in relation to sweat glands. The autonomic fibers of the vagus and glossopharyngeal nerves are cholinergic in nature. They cannot therefore innervate sympathetic postganglionic

FIGURE 52–4 Blood pressure (B. P.) and heart rate (H. R.) in a recently injured tetraplegic with a high cervical lesion needing assisted ventilation. Top, Disconnecting the respirator for aspirating the airways resulted in a fall in blood pressure and heart rate, when performed 6 hours after the last dose of intravenous atropine. Sinus bradycardia and cardiac arrest were reversed by atropine and external cardiac massage. Bottom, Disconnecting the respirator and tracheal suction caused virtually no change in blood pressure and heart rate. (Modified from Frankel, H. L., Mathias, C. J., and Spalding, J. M. K.: Mechanisms of reflex cardiac arrest in tetraplegic patients. Lancet 1:1183, 1975; and Mathias, C. J.: Bradycardia and cardiac arrest during tracheal suction—mechanisms in tetraplegic patients. Eur. J. Intens. Care Med. 2:147, 1976.)

adrenergic pathways, but they can innervate sympathetic preganglionic nerves or sudomotor postganglionic nerves, which are both cholinergic. This may cause gustatory sweating, which is inappropriate sweating in circumstances that normally cause salivary and gastric secretion. This may be unilateral or bilateral, with involvement of the head, face, neck, and shoulder depending on the nerves involved (Fig. 52–6).145 Excessive sweating may be induced while eating or even while contemplating food ingestion. This can be socially embarrassing and can be severe enough to warrant a change of clothes. It may result from damage to the glossopharyngeal nerve supplying the parotid gland after surgery, with regeneration of postganglionic cholinergic

Diseases of the Ninth through Twelfth Cranial Nerves

FIGURE 52–5 Blood pressure (BP) and heart rate (HR) in a chronically injured tetraplegic before, during, and after endotracheal intubation prior to general anesthesia. After intubation, cardiac arrest occurred, during which oxygen was administered followed by external cardiac massage. The dose of atropine that the patient had received had not adequately blocked the vagus. (From Welply, N., Mathias, C. J., and Frankel, H. L.: Circulatory reflexes in tetraplegics during artificial ventilation and during anaesthesia. Paraplegia 13:172, 1975, with permission.)

1283

sympathetic pathways in the auriculotemporal or greater auricular nerve supplying the sweat glands in that region (auriculotemporal syndrome or Frey’s syndrome85). Gustatory sweating may follow vagal and/or sympathetic damage after operations in the root of the neck and after thoracotomy or upper limb sympathectomy.106 Gustatory piloerection (which is adrenergically mediated) may accompany sweating, suggesting aberrant innervation of the cervical ganglia. Gustatory sweating in diabetics is often in the territory of the superior cervical ganglia.1,217 If treatment is needed, anticholinergic agents such as poldine methylsulfate or propantheline bromide before meals may be beneficial, but these drugs have undesirable side effects, which include a dry mouth. Local application of anticholinergic agents129 and irradiation of the skin91 may help, and the local application of aluminum chloride hexahydrate is beneficial.23 Surgical section of the sympathetic pathway, the auriculotemporal nerve, or the tympanic branch of the glossopharyngeal nerve168,195 may be needed in extreme cases. Vagal Afferents Laryngeal Receptors. The vagus supplies sensation to the mucous membrane of the larynx. This, with the corresponding sensory innervation of the tracheobronchial tree, provides the afferent pathway for the cough reflex.

FIGURE 52–6 Sequential photographs showing the effects of quinazarin red (pale powder when dry) before (A) and after (B and C) gustatory sweating induced by a cheese sandwich. This caused marked sweating, especially over the scalp. (From Mathias, C. J.: Disorders of the autonomic nervous system. In Bradley, W. G., Daroff, R. B., Fenichel, G. M., and Marsden, C. D. [eds.]: Neurology in Clinical Practice. Stoneham, MA, Butterworths, p. 1661, 1991, with permission.)

1284

Diseases of the Peripheral Nervous System

Unilateral anesthesia is generally part of a more widespread neurologic deficit. With bilateral lesions, the laryngeal and tracheobronchial reflexes are absent and inhalation of secretions and food materials is likely to occur. Stimulation of the larynx and trachea, as during tracheal toilet in patients on respirators, can result in a variety of autonomic responses. In those with an intact autonomic nervous system there is a modest increase in both blood pressure and heart rate.49 Severe hypertension and tachycardia may occur in patients with tetanus because of sympathetic overactivity,50 whereas the reverse—bradycardia, hypotension, and even cardiac arrest—may occur in tetraplegics.144 “Superior laryngeal neuralgia”11 consists of paroxysmal attacks of lancinating pain that radiates from the side of the thyroid cartilage to the angle of the jaw and sometimes the ear. The pain may be provoked by pressure on the skin over the thyrohyoid membrane. It may respond to treatment with carbamazepine.35,188 Pulmonary Receptors. Pulmonary stretch activity in the vagus is present at the resting expiratory position and increases with inspiration. In humans, however, during normal resting breathing, vagal blockade does not affect the breathing (though it does so in some animals).98,227 In monitoring tidal volume in humans at rest, the major role appears to lie with receptors in the chest wall concerned with stretch and the load on inspiratory muscles, rather than with vagal pulmonary stretch receptors. In nonresting breathing, however, vagal blockade reduces the ventilatory response to carbon dioxide and makes respiration slower and deeper. This suggests that vagal stretch receptors may be concerned in the control of the pattern of breathing when the breathing is vigorous.100 Deflation of the lung may cause an increase in respiratory rate and force, but in humans the deflation has to be of the severity that occurs in pneumothorax. The effect is abolished in animals by vagotomy, and it is likely that it is also mediated by the vagus in humans. Animal experiments indicate that, in addition to pulmonary stretch receptors, there are vagal lung irritant receptors.150 They respond to many noxious stimuli such as toxic substances (ammonia, cigarette smoke) in the airways, overinflation or marked deflation of the lungs, or pulmonary congestion or microembolism. Unlike pulmonary stretch activity, activity in these nerves does not adapt quickly if stimulation is continued, and they are probably concerned in unpleasant respiratory sensations in humans. In diabetes mellitus with autonomic neuropathy there may be diminished bronchial reactivity105 that is due not to an abnormality of bronchial smooth muscle (as tested with inhaled histamine) but to impairment of vagal afferent receptors or pathways. This also accounts for the high cough threshold of these patients when tested by inhalation of citric acid (Fig. 52–7).69

FIGURE 52–7 Cough responses to inhaled citric acid in normal subjects (open histograms; controls) and in diabetic autonomic neuropathy (hatched histograms). The responses are impaired or absent in the patients with autonomic neuropathy. (From Edmonds, M. E., and Watkins, P. J.: Clinical presentation of diabetic autonomic failure. In Bannister, R. [ed.]: Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 2nd ed. Oxford, UK, Oxford University Press, p. 632, 1988, with permission.)

Cardiac Receptors. Activation of cardiac vagal afferents in the ventricles may result in bradycardia, hypotension, and syncope. These appear to be mechanoreceptors stimulated by cardiac distention or by forceful contraction.2 This may account for the bradycardia during hemorrhage,185 the syncope of aortic stenosis,142 and the bradycardia and hypotension during coronary arteriography68 and isoprenaline infusion in the head-up tilt position.5 The precise mechanisms involved in “ventricular syncope” include a variety of factors influencing sensitivity of these receptors together with the interactive role of cerebral centers.

Lesions Affecting the Vagus Nerve Intramedullary Lesions As already stated, lesions within the brainstem commonly affect other lower cranial nerves and long tracts simultaneously and may be due to vascular (as in the lateral medullary syndrome), neoplastic, or inflammatory causes or sometimes syringobulbia.148 Lesions at this site usually produce palatal, pharyngeal, and laryngeal paralysis, but if the upper part of the nucleus ambiguus is selectively involved, there may be sparing of laryngeal function (palatopharyngeal syndrome of Avellis). The cells of the nucleus may be involved in progressive bulbar palsy (motor neuron disease) or poliomyelitis. Similar changes may occur in multiple system atrophy (Shy-Drager syndrome), resulting in laryngeal palsy affecting

Diseases of the Ninth through Twelfth Cranial Nerves

mainly the cricoarytenoid muscles in the laryngeal abductors96,228 and/or causing pharyngeal involvement with abnormalities of deglutition. Upper airway dysfunction along with central respiratory disorders may contribute to sudden death during sleep in such patients.155 Extramedullary Lesions Damage to the Xth nerve intracranially after its emergence from the medulla may occur within the posterior fossa because of tumors (see Chapter 115) or in meningovascular syphilis, in which case the lower cranial nerves are often additionally involved. It may also occur after the nerve leaves the skull because of the same types of disease as enumerated for the ninth nerve, including neurinomas94,171 and chemodectomas.80,120 The three nerves that emerge through the jugular foramen (IXth, Xth, and XIth) may simultaneously be involved, usually by a neoplasm, in Vernet’s syndrome, or with more extensive lesions the XIIth nerve may also be affected (Collet-Sicard syndrome). Other combinations have been defined, including the syndromes of Schmidt (involving the Xth and eleventh cranial nerves), Jackson (involving nerves X, XI, and XII), and Tapia (unilateral paralysis of the tongue and vocal cord).174 The vagus nerve may be affected as part of a polyneuritis cranialis, particularly in sarcoidosis (see Chapter 107), in Lyme borreliosis (see Chapter 92), and in occasional examples of the Guillain-Barré syndrome (see Chapter 98). Vagal neuropathy causing hoarseness and weakness of the voice and dysphagia may occur in neuritic beriberi and alcoholic neuropathy159 and in vincristine neuropathy.61 Bilateral vocal cord paralysis has been reported in druginduced systemic lupus erythematosus,44 and recurrent laryngeal nerve palsy may occur as a rare manifestation of idiopathic brachial plexus neuropathy (neuralgic amyotrophy). Palatal palsy may complicate pharyngeal diphtheria (see Chapter 95). Vagal palsy caused by a dissecting aneurysm of the carotid artery is on record.56 The recurrent laryngeal nerves are liable to damage by a variety of intrathoracic lesions, particularly on the left because of the longer course of the nerve on that side. Neoplastic involvement of mediastinal lymph glands from carcinoma of the bronchus or esophagus is the most common explanation, although in the past aortic aneurysm and enlargement of the left atrium in mitral stenosis figured as possible causes. Distention of the esophagus as a result of achalasia of the cardia can cause reversible bilateral recurrent laryngeal nerve palsy.226 The recurrent laryngeal nerves may also be involved in the neck because of carcinoma of the esophagus or thyroid gland or neoplastic invasion of the cervical lymph glands. Operative damage to the nerve is now considerably less frequent than formerly; a delayed postoperative laryngeal palsy may also occur. Williams analyzed his findings for 100 benign goiters.230 Preoperative inspection of the vocal

1285

cords revealed no paralysis, nor was any instance of immediate postoperative paralysis detected. Unilateral cord paralysis developed after 4 or 5 days in seven cases, however, and in two it was permanent. Because it was known that the nerves had not been transected or crushed during the operation, postoperative edema was considered to be the probable cause. Hawe and Lothian exposed the recurrent laryngeal nerves in 1011 operations.104 No paralysis was present preoperatively, but paralysis occurred following operation in 28 cases, in only 3 of whom it was permanent. If laryngoscopy at the termination of the operation revealed paralysis, the nerve on the affected side was immediately explored. In two cases the nerve had been included in a ligature; subsequent recovery occurred in both. Blackburn and Salmon reported their findings in 250 thyroidectomies in which preoperative and postoperative inspection of the larynx had been made.24 There were four instances of temporary and five of permanent unilateral recurrent laryngeal nerve paralysis. Unilateral paralysis may rarely complicate insertion of a nasogastric or endotracheal tube and is thought to be due to damage to the posterior branch of the recurrent laryngeal nerve innervating the posterior cricoarytenoid and interarytenoid muscles, as it passes behind the thyroid cartilage.86 Bilateral recurrent laryngeal damage has usually followed thyroidectomy. In a series of 22 cases reported by Williams, this occurred in 18.230 In two, it was part of a polyneuropathy, and in two others it was the result of carcinoma of the thyroid gland and of the esophagus. In about one quarter to one third of all cases of isolated recurrent laryngeal palsy there is no discoverable cause,14,112,232 although other writers have given a lower incidence rate for “idiopathic cases.”123 Males are affected twice as often as females, the left side is involved more frequently than the right, and a peak incidence occurs in the third decade.112 A series of 21 such cases was reported by Blau and Kapadia.25 Five patients recovered completely and another five improved within a few months of onset. The remainder developed no major disease over a followup period of 1 to 8 years. Berry and Blair have reported a series of 25 patents with isolated vagal mononeuropathy.21 From electromyographic studies it was possible to distinguish between recurrent laryngeal nerve palsy and recurrent plus superior laryngeal nerve palsy. Unilateral recurrent laryngeal nerve palsy was nearly twice as common as the combined lesion and appeared to have a better prognosis. The majority of cases were idiopathic, but two were associated with diabetes mellitus. Three patients had bilateral palsies. These were possibly a separate entity, occurring in younger patients and associated with upper respiratory infections. Partial or complete recovery took place in about half of this series.

1286

Diseases of the Peripheral Nervous System

THE SPINAL ACCESSORY NERVE Anatomy This nerve is largely efferent. It emerges as a linear series of rootlets extending from the last of the rootlets of the cranial accessory nerve as far as the fifth or sixth segment of the spinal cord. It is derived from a column of cells in the posterolateral part of the anterior horn, which in their most cranial portion are relatively isolated and approach the caudal termination of the nucleus ambiguus.169 The rootlets emerge from the lateral column of the cord posterior to the ligamenta denticulata and unite to form a bundle that passes through the foramen magnum. It is then joined by the cranial portion of the accessory nerve and leaves the skull through the jugular foramen. The cranial portion of the accessory nerve separates as the internal ramus and joins the vagus nerve. Having left the skull, the spinal accessory nerve descends between the internal carotid artery and the internal jugular vein and penetrates the sternomastoid muscle, which it supplies. It emerges just above the middle of the posterior border of this muscle and crosses the posterior triangle of the neck in its fascial roof to reach the trapezius muscle, which it also supplies. During its course, it receives communications from the third and fourth cranial nerves through the cervical plexus. The spinal accessory nerve also contains some afferent fibers whose cell bodies lie within the nerve trunk in its intracranial portion.169,233 These fibers are probably muscle afferents.236

Clinical Features of Spinal Accessory Nerve Lesions Damage to the spinal accessory nerve deep to the sternocleidomastoid muscle or intracranially leads to weakness of this muscle and also of the upper part of the trapezius muscle. Lesions in the posterior triangle result only in weakness of this portion of the trapezius. The lower fibers of the trapezius are supplied from the third and fourth cervical roots through the cervical plexus, although there is some variability; occasionally the whole trapezius is supplied by these roots.46 Following an accessory nerve lesion, there is weakness of rotation of the head of the opposite side because of paresis of the sternocleidomastoid muscle. Involvement of the trapezius gives rise to weakness of elevation of the shoulder and inability to lift the arm above the horizontal. It also leads to drooping of the shoulder, together with moderate winging of the scapula when the arms are hanging at the side. The winging is accentuated when the patient attempts to elevate the arm laterally. This is in contradistinction to the winging seen with serratus anterior weakness, which is minimal at rest but becomes pronounced on forward elevation of the arm.

Lesions of the Spinal Accessory Nerve The cells of origin in the spinal cord may be involved in motor neuron disease, poliomyelitis, syringomyelia, and spinal tumor. In the intraspinal and intracranial portions and in the region of the skull base, the nerve may be affected by the same disorders that have been described in the IXth and Xth nerves, including posterior fossa meningiomas42 and neurinomas (see Chapter 115). Its superficial position in the neck renders it liable to damage during surgical procedures and from external compression and injury.16,18,158,189 Stretch injury has been described.137 Minor procedures such as posterior triangle lymph node biopsy,160 cannulation of the internal jugular vein,107 and carotid endarterectomy200 may be responsible. Surgical exploration and nerve suture may be effective in restoring function in the cases in which nerve transection is a possibility.42 Delayed palsies after surgical procedures suggest entrapment in scar tissue.93 More bizarre causes of damage from external compression include love bites during sex play164 and unsuccessful attempts at hanging.194 The nerve may be affected by radiation injury, either in isolation or in combination with the vagus and hypoglossal nerves following treatment of head and neck tumors.20 Isolated unexplained lesions of the accessory nerve occasionally occur from which spontaneous recovery may take place.71 They may be heralded by pain felt along the posterior border of the sternocleidomastoid muscle and in the suboccipital region.43 This pain possibly arises locally in the nerve itself (see Chapter 46). After a few days, the severe pain subsides, to be replaced by a dull ache in the region of the shoulder, at which time the weakness is noticed. Such lesions are possibly instances of benign mononeuropathies of the type that may affect various other cranial nerves, of which Bell’s palsy is the most frequent example.83 They can be recurrent.39 Electromyography and measurement of the latency of the muscle response on nerve stimulation may be useful for confirmation of the diagnosis.79,95,172

THE HYPOGLOSSAL NERVE Anatomy The hypoglossal nerve carries motor fibers to the intrinsic tongue muscles and to the thyrohyoid, hyoglossus, styloglossus, genioglossus, and geniohyoid muscles. These fibers originate from the hypoglossal nucleus, which consists of a longitudinal arrangement of neurons lying beneath the floor of the fourth ventricle and extending caudally to the lower limit of the medulla. In their intramedullary course, the fibers traverse the reticular formation and the medial part of the inferior olive before leaving the medulla in the lateral ventral sulcus. The nerve

Diseases of the Ninth through Twelfth Cranial Nerves

emerges as several filaments, uniting into two bundles that pass separately through the dura and the anterior condyloid foramen (hypoglossal canal). Outside the skull, the nerve passes vertically downward to the level of the angle of the jaw, where it turns forward to innervate the thyrohyoid muscle and the extrinsic and intrinsic muscles of the ipsilateral half of the tongue. Sympathetic fibers travel in association with the nerve. Dural branches are given off in the hypoglossal canal. In its descending portion after it has emerged from the skull, the nerve has connections with the glossopharyngeal, vagus, and accessory nerves. Fibers from the first and second cervical nerves join the hypoglossal nerve close to its exit from the skull and leave the nerve shortly after this as its descending branch as it turns around the occipital artery. They join the ansa cervicalis. There is no doubt that the hypoglossal nerve contains afferent fibers, a view supported by both anatomic and physiologic evidence. The earlier literature is reviewed by Blom.26 Muscle spindles have been identified in the human tongue by Cooper47 and by Walker and Rajagopal,213 as well as in the tongue of other primates. The fiber caliber spectrum of the hypoglossal nerve in humans shows a peak at 7 to 9 ␮m, the largest fibers being 11 to 13 ␮m.177 In the rhesus monkey, the caliber spectrum of the distal portion of the nerve is consistent with a significant component being from spindle afferents, although this is not the case more proximally in the nerve, suggesting that spindle afferents reach the central nervous system by a different route.70 Physiologic studies in the same species indicate that this route involves the second and third cervical nerve roots.31 This finding accords with the earlier work of Corbin and Harrison, who demonstrated that the cells of origin for proprioceptive fibers from the tongue in the rhesus monkey lie in the dorsal root ganglion of C2.51 This may explain the “neck-tongue syndrome.”131 It has been suggested that afferents from the tongue in the lingual nerve join the hypoglossal nerve to reach the C2 root. Pseudoathetosis of the tongue may be a feature of this syndrome. A small ganglion is found along the course of the hypoglossal nerve in some species, and nerve cells have been identified in the extracranial portion of the hypoglossal nerve in the human fetus by Wózniak and Young, who suggested that these cells, because of their morphologic similarities, must have migrated from those of the inferior ganglion of the vagus nerve.235 Electrical stimulation of the hypoglossal nerve in the cat was shown by Downman to cause pupillary dilatation and a rise in blood pressure,62 and in the dog such stimulation evokes efferent activity in thoracic and renal sympathetic nerves.225 Stimulation at an adequate intensity to excite group III afferent fibers evokes in the cat responses in muscle innervated by the facial nerve.136 Hanson and Widén have demonstrated a reflex response in the styloglossus muscle to stimulation of the central cut end of a peripheral branch of

1287

the nerve.101 Afferent impulses have also been recorded in the hypoglossal nerve in response to stretch.26,48,101,154

Clinical Features of Hypoglossal Nerve Lesions A unilateral lesion of the nerve does not interfere with speech, but bilateral involvement causes dysarthia and difficulty in swallowing. The signs of a lesion of the hypoglossal nerve are wasting of the ipsilateral half of the tongue, which may show excessive furring. Deviation toward the side of the paralysis occurs when the tongue is protruded, although there are occasional exceptions to this.219 With bilateral lesions, there is difficulty in tongue protrusion. Lesions involving the hypoglossal nucleus usually cause fasciculation of the tongue.

Lesions of the Hypoglossal Nerve Congenital atrophy of the tongue is usually due to deficiency of innervation. The neurons constituting the hypoglossal nucleus may be affected by the same diseases that attack anterior horn cells in the spinal cord, notably motor neuron disease and poliomyelitis. Their involvement in inherited neuronal degenerations is discussed in Chapter 67. The nucleus and the intramedullary portion of the nerves may also be involved by structural lesions in the lower brainstem such as infarction, neoplasm, and syringobulbia. If the lesion is confined to one half of the medulla, ipsilateral hypoglossal paralysis may be associated with crossed hemiplegia. Extrinsic lesions of the nerve in its intracranial portion may in rare instances be due to a primary tumor of the nerve.8,10,113,231 Selective involvement of one nerve has been described as the first sign of intracranial metastasis in patients with bronchial carcinoma, lymphoma, or leukemia182 and may occur as a consequence of radiation therapy.20,41 Basal meningitis and vertebral artery aneurysms occasionally interfere with XIIth nerve function, but other cranial nerves are usually also implicated. Hypoglossal nerve palsy may occur in association with infectious mononucleosis.57 At the level of the hypoglossal canal, the nerve can be involved by glomus jugulare tumors, meningiomas, cordomas, and cholesteatomas, sometimes in combination with lesions of other lower cranial nerves. Paralysis caused by vascular entrapment161,211 and spontaneous dissection of the internal carotid artery135 has also been described. Occasionally the nerve may be damaged as a consequence of head injury163 or subluxation of the odontoid process in rheumatoid arthritis140 or, in the neck or mouth, as a result of a penetrating head wound or dental extraction.197 Isolated reversible unexplained hypoglossal nerve lesions are also encountered.3

1288

Diseases of the Peripheral Nervous System

Glossodynia Glossodynia describes the syndrome of burning pain experienced in the tongue and often also throughout the oral mucosa. It occurs usually in the middle aged and elderly and more frequently in females than in males. Deficiency of the B vitamins, including B12, has been named as a cause by Elfenbaum,73 but in the majority of patients the symptom seems to be psychogenic. In a series of 54 patients described by Quinn, for example, 19 had cancerophobia.175

REFERENCES 1. Aagenaes, Ó.: Neurovascular Examinations on the Lower Extremities in Young Diabetics, with Special Reference to the Autonomic Neuropathy. Copenhagen, C. Hamburgers Bogtrykkeri, 1962. 2. Abboud, F. M.: Ventricular syncope: is the heart a sensory organ? N. Engl. J. Med. 320:390, 1989. 3. Afifi, A. K., Rifai, Z. H., and Faris, K. B.: Isolated reversible hypoglossal nerve palsy. Arch. Neurol. 41:1218, 1984. 4. Alexander, S., and Ping, W. C.: Fatal ventricular fibrillation during carotid sinus stimulation. Am. J. Cardiol. 18:289, 1966. 5. Almquist, S. T., Goldenberg, I. F., Milstein, S., et al.: Provocation of bradycardia and hypotension by isoproterenol and upright posture in patients with unexplained syncope. N. Engl. J. Med. 320:346, 1989. 6. Alpert, J. N., Armbrust, C. A., Akhari, M., et al.: Glossopharyngeal neuralgia, asystole and seizures, Arch. Neurol. 34:233, 1977. 7. Andersen, O., Lindberg, J., Modigh, K., and ReskeNielsen, E.: Subacute dysautonomia with incomplete recovery. Acta Neurol. Scand. 48:510, 1972. 8. Andrioli, G., Rigobello, L., Mingrino, S., and Toso, V.: Tapia’s syndrome caused by a neurofibroma of the hypoglossal and vagus nerves: case report. J. Neurosurg. 52:730, 1980. 9. Angell-James, J. E., and Lumley, J. S. P.: The effects of carotid endarterectomy on the mechanical properties of the carotid sinus and carotid sinus nerve activity in atherosclerotic patients. Br. J. Surg. 61:805, 1974. 10. Arumugasamay, N., Sarvananthan, K., Rudralingan, V., and Pillay, D. R. P.: Intracranial hypoglossal neurinomas—a report of two cases. Med. J. Malaya 26:168, 1972. 11. Avellis, G.: Typische Form von Kehlkopfneuralgie. Münch. Med. Wochenschr. 47:1592, 1900. 12. Axelsson, A., and Laage-Hellman, J. E.: The gusto-lacrimal reflex. Acta Otolaryngol. 54:239, 1962. 13. Ball, A. P., Hopkinson, R. B., Farrell, I. D., et al.: Human botulism caused by Clostridium botulinum type E: the Birmingham outbreak. Q. J. Med. 191:473, 1979. 14. Ballantyne, J.: Neurological affections of the larynx. In Ballantyne, J., and Groves, J. (eds.): Diseases of the Ear, Nose and Throat, 3rd ed. London, Butterworths, 1971. 15. Mathias, C. J., and Bannister, R.: Investigation of autonomic disorders. In Mathias, C. J., and Bannister, R. (eds.): Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 4th ed. Oxford, UK, Oxford University Press, p. 169, 2002.

16. Bateman, J. E.: Nerve injuries about the shoulder in sports. J. Bone Joint Surg. Am. 49:785, 1967. 17. Bauernfeind, R., Hall, C., Denes, P., and Rosen, K. M.: Carotid sinus hypersensitivity with alpha methyldopa. Ann. Intern Med. 88:214, 1987. 18. Bell, D. S.: Pressure palsy of the accessory nerve. Br. Med. J. 1:1483, 1964. 19. Bennett, T., Hosking, D. J., and Hampton, J. R.: Baroreflex sensitivity and responses to the Valsalva manoeuvre in subjects with diabetes mellitus. J. Neurol. Neurosurg. Psychiatry 39:178, 1976. 20. Berger, P. S., and Bataini, J. P.: Radiation-induced cranial nerve palsy. Cancer 40:152, 1977. 21. Berry, H., and Blair, R. L.: Isolated vagus nerve palsy and vagal mononeuritis. Arch. Otolaryngol. 106:333, 1980. 22. Biscoe, T. J.: Carotid body: structure and function. Physiol. Rev. 51:437, 1971. 23. Black, M. J. M., and Gunn, A.: The management of Frey’s syndrome with aluminium chloride hexahydrate antiperspirant. Ann. R. Coll. Surg. 2:49, 1990. 24. Blackburn, G., and Salmon, L. F.: Cord movements after thyroidectomy. Br. J. Surg. 48:371, 1961. 25. Blau, J. N., and Kapadia, R.: Idiopathic palsy of the recurrent laryngeal nerve: a transient cranial mononeuropathy. Br. Med. J. 4:259, 1972. 26. Blom, S.: Afferent influences on tongue activity. Acta Physiol. Scand. Suppl. 49:170, 1960. 27. Bogorad, F. A.: Symptoms of crocodile tears. Vrach. Delo. 2:1328, 1928. 28. Bowden, R. E. M.: The surgery of the recurrent laryngeal nerve. Proc. R. Soc. Med. 48:437, 1955. 29. Bowden, R. E. M.: Innervation of intrinsic laryngeal muscles. In Wyke, B. (ed.): Ventilatory and Phonatory Control Systems. London, Oxford University Press, 1974. 30. Bowden, R. E. M., and Scheuer, J. L.: Weights of abductor and adductor muscles of the human larynx. J. Laryngol. Otol. 74:971, 1960. 31. Bowman, J. P., and Combs, C. M.: The cerebrocortical projection of hypoglossal afferents. Exp. Neurol. 23:291, 1969. 32. Boyer, F. C., and Gardner, W. J.: Paroxysmal lacrimation (syndrome of crocodile tears) and its surgical treatment: relation to auriculotemporal syndrome. Arch. Neurol. Psychiatry 61:56, 1949. 33. Brodal, A.: Central course of afferent fibers for pain in facial, glossopharyngeal and vagus nerves. Arch. Neurol. Psychiatry 57:292, 1947. 34. Brodie, R. E., and Dow, R. S.: Studies in carotid compression and carotid sinus sensitivity. Neurology (Minneap.) 18:1047, 1968. 35. Brownstone, P. K., Ballenger, J. J., and Vick, N. A.: Bilateral superior laryngeal neuralgia: its successful treatment with carbamazepine. Arch. Neurol. 37:525, 1980. 36. Bruyn, G. W.: Glossopharyngeal neuralgia. Cephalalgia 3:143, 1983. 37. Bues, E.: The anterior zone of the ear in the diagnosis and treatment of glossopharyngeal neuralgia. Dtsch. Z. Nervenheilkd. 185:471, 1963. 38. Cannon, W. B., and Rosenblueth, A.: The Supersensitivity of Denervated Structures: A Law of Denervation. New York, Macmillan, 1949.

Diseases of the Ninth through Twelfth Cranial Nerves 39. Chalk, C., and Isaacs, H.: Recurrent spontaneous accessory neuropathy. J. Neurol. Neurosurg. Psychiatry 53:61, 1990. 40. Chao, L. P.: The mechanism of botulism. Med. Hypoth. 9:83, 1986. 41. Cheng, V. S. T., and Schulz, M. D.: Unilateral hypoglossal nerve atrophy as a later complication of radiation therapy of head and neck carcinoma: a report of 4 cases and a review of the literature on peripheral and cranial nerve damage after radiation therapy. Cancer 39:1537, 1975. 42. Cherington, M.: Accessory nerve: conduction studies. Arch. Neurol. 18:708, 1968. 43. Cherington, M., Hendee, R., and Roland, R.: Accessory nerve palsy—a painful cranial neuropathy. Headache 18:274, 1977. 44. Chong, W. K., Dewhurst, A. G., and Dathan, J. R. E.: Acute laryngeal stridor with respiratory arrest in drug induced systemic lupus erythematosus. BMJ 297:660, 1988. 45. Clark, C. V., and Mapstone, R.: Autonomic neuropathy in ocular hypertension. Lancet 2:185, 1985. 46. Coleman, C. C., and Walker, J. C.: Technic of anastomosis of the branches of the facial nerve with the spinal accessory for facial paralysis. Ann. Surg. 131:960, 1950. 47. Cooper, S.: Muscle spindles in the intrinsic muscles of the human tongue. J. Physiol. (Lond.) 122:193, 1953. 48. Cooper, S.: Afferent impulses in the hypoglossal nerve on stretching the cat’s tongue. J. Physiol. (Lond.) 126:32P, 1954. 49. Corbett, J. L., Kerr, J. H., and Prys-Roberts, C.: Cardiovascular responses to aspiration of secretions from the respiratory tract. J. Physiol. (Lond.) 205:51P, 1969. 50. Corbett, J. L., Kerr, J. H., Prys-Roberts, C., et al.: Cardiovascular disturbances in severe tetanus due to overactivity of the sympathetic nervous system. Anaesthesia 24:198, 1969. 51. Corbin, K. B., and Harrison, F.: The sensory innervation of the spinal accessory and tongue musculature in the rhesus monkey. Brain 62:191, 1939. 52. Cottle, M. K.: Degeneration studies of primary afferents of IXth and Xth cranial nerves in the cat. J. Comp. Neurol. 122:329, 1964. 53. Craddock, C., Pasvol, G., Bull, R., et al.: Cardiorespiratory arrest and autonomic neuropathy in AIDS. Lancet 2:16, 1987. 54. Dandy, W. E.: Glossopharyngeal neuralgia (tic douloureux): its diagnosis and treatment. Arch. Surg. 15:190, 198, 1927. 55. Davie, J. C., and Bourke, R. S.: Confirmation of atypical parasympathetic innervation of the parotid gland: atypical distribution of ninth nerve. Arch. Neurol. 16:599, 1967. 56. Davies, L.: A case of vagal palsy due to dissecting aneurysm of the carotid artery. Med. J. Aust. 147:352, 1987. 57. De Simone, P. A., and Snyder, D.: Hypoglossal nerve palsy in infectious mononucleosis. Neurology (Minneap.) 28:844, 1978. 58. Dedo, H. H.: The paralysed larynx: an electromyographic study in dogs and humans. Laryngoscope 80:1455, 1970. 59. Dedo, H. H., and Hall, W.: Electrodes in laryngeal electromyography. Ann. Otol. Rhinol. Laryngol. 78:172, 1969. 60. Dejours, P.: Chemoreflexes in breathing. Physiol. Rev. 42:335, 1962.

1289

61. Delaney, P.: Vincristine-induced laryngeal nerve paralysis. Neurology (N.Y.), 32:1285, 1982. 62. Downman, C. B. B.: Afferent fibres of the hypoglossal nerve. J. Anat. 73:387, 1939. 63. DuBois, F. S., and Foley, J. O.: Quantitative studies of the vagus nerve in the cat. II. The ratio of jugular to nodose fibers. J. Comp. Neurol. 67:69, 1937. 64. Duncan, G., Johnson, R. H., Lambie, D. G., and Whiteside, E. A.: Evidence of vagal neuropathy in chronic alcoholics. Lancet 2:1053, 1980. 65. Dyer, K. R., and Duncan, I. D.: The intraneural distribution of myelinated fibres in the equine recurrent laryngeal nerve. Brain 110:1531, 1987. 66. Eckburg, D. L.: Adaptation of the human baroreceptorcardiac reflex. J. Physiol. (Lond.) 269:579, 1977. 67. Eckburg, D. L.: Baroreflex inhibition of the human sinus node: importance of stimulus intensity, duration and rate of pressure change. J. Physiol. (Lond.) 269:561, 1977. 68. Eckberg, D. L., White, C. W., Kioschos, J. M., and Abboud, F. M.: Mechanisms mediating bradycardia during coronary arteriography. J. Clin. Invest. 54:1445, 1974. 69. Edmonds, M. E., and Watkins, P. J.: Clinical presentation of diabetic autonomic failure. In Bannister, R. (ed.): Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 2nd ed. London, Oxford University Press, p. 632, 1988. 70. Egel, R. T., Bowman, J. P., and Combs, C. M.: Caliber spectra of the lingual and hypoglossal nerves of the rhesus monkey. J. Comp. Neurol. 134:163, 1968. 71. Eisen, A., and Bertrand, G.: Isolated accessory nerve palsy of spontaneous origin: a clinical and electromyographic study. Arch. Neurol. 27:496, 1972. 72. Ekbom, K. A., and Westerberg, C. E.: Carbamazepine in glossopharyngeal neuralgia. Arch. Neurol. 14:595, 1966. 73. Elfenbaum, A.: Burning tongue in older patients. J. Can. Dent. Assoc. 35:533, 1969. 74. Ellis, M.: Laryngeal paralysis. In Ellis, M. (ed.): Modern Trends in Diseases of the Ear, Nose and Throat. London, Butterworths, p. 327, 1954. 75. Evans, R. J. C., Benson, M. K., and Hughes, D. T. D.: Abnormal chemoreceptor response to hypoxia in patients with tabes dorsalis. Br. Med. J. 1:530, 1971. 76. Ewing, D. J.: Recent advances in the non-invasive investigation of diabetic autonomic neuropathy. In Bannister, R. (ed.): Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 2nd ed. London, Oxford University Press, p. 667, 1988. 77. Fagius, J., and Wallin, G. B. Microneurographic evidence of excessive sympathetic outflow in the Guillain-Barré syndrome. Brain 106:589, 1983. 78. Fagius, J., Wallin, G. B., Sundlof, G., et al.: Sympathetic outflow in man after anaesthesia of the glossopharygeal and vagus nerves. Brain 108:423, 1985. 79. Fahrer, H., Ludin, H. P., Mumenthaler, M., and Neiger, M.: The innervation of the trapezius muscle: an electrophysiological study. J. Neurol. 207:183, 1974. 80. Fernandez, B. B., Hernandez, F. J., and Staley, C. J.: Chemodectoma of the vagus nerve: report of a case with ultrastructural study. Cancer 35:263, 1975.

1290

Diseases of the Peripheral Nervous System

81. Ferreira-Santos, R.: Megacolon and megarectum in Chagas’ disease. Proc. R. Soc. Med. 54:1047, 1961. 82. Fink, L. H., Early, C. B., and Bryan, R. V.: Glossopharyngeal schwannomas. Surg. Neurol. 9:239, 1977. 83. Foley, J. M.: The cranial mononeuropathies. N. Engl. J. Med. 281:905, 1969. 84. Frankel, H. L., Mathias, C. J., and Spalding, J. M. K.: Mechanisms of reflex cardiac arrest in tetraplegic patients. Lancet 1:1183, 1975. 85. Frey, L.: Syndrome of auriculotemporal nerve. Rev. Neurol. 30:97, 1923. 86. Friedman, M., and Toriumi, D. M. Esophageal stethoscope. Arch. Otolaryngol. Head Neck Surg. 115:95, 1989. 87. Gardner, W. J.: Crosstalk—the paradoxical transmission of a nerve impulse. Arch. Neurol. 14:149, 1966. 88. Garretson, H. D., and Elvidge, A. R.: Glossopharyngeal neuralgia with asystole and seizures. Arch. Neurol. 8:26, 1963. 89. Gejrot, T.: Jugular syndrome. Acta Otolaryngol. (Stockh.) 57:450, 1964. 90. Gilbey, S. G., and Watkins, P. J.: Measurement of epigastric impedance of gastric emptying in diabetic autonomic neuropathy. Diabet. Med. 4:122, 1987. 91. Glaister, D. H., Hearnshaw, J. R., Heffron, P. F., and Peck, A. W.: The mechanisms of postparotidectomy gustatory sweating. Br. Med. J. 2:942, 1958. 92. Golding-Wood, P.: Crocodile tears. Br. Med. J. 1:1518, 1963. 93. Gordon, S. L., Graham, W. P., Black, J. T., and Miller, S. H.: Accessory nerve function after surgical procedures in the posterior triangle. Arch. Surg. 112:264, 1977. 94. Green, J. D. Jr.: Neoplasms of the vagus nerve. Laryngoscope 98:648, 1988. 95. Green, R. F., and Brien, M.: Accessory nerve latency to the middle and lower trapezius. Arch. Phys. Med. Rehabil. 66:23, 1985. 96. Guidi, G. M., Bannister, R., and Oppenheimer, D. R.: Laryngeal electromyography in multiple system atrophy with autonomic failure. J. Neurol. Neurosurg. Psychiatry 44:49, 1981. 97. Guy, R. J. C., Dawson, J. L., Garret, J. R., et al.: Diabetic gastroparesis from autonomic neuropathy: surgical considerations and changes in vagus nerve morphology. J. Neurol. Neurosurg. Psychiatry 47:686, 1984. 98. Guz, A., Noble, M. I. M., Eisele, J. H., and Trenchard, D.: The role of vagal inflation reflexes in man and other animals. In Porter, R. (ed.): Breathing: Hering-Breuer Centenary Symposium. A Ciba Foundation Symposium. London, J. & A. Churchill, p. 1, 1970. 99. Guz, A., Noble, M. I. M., Widdicombe, J. G., et al.: The role of vagal and glossopharyngeal afferent nerves in respiratory sensation, control of breathing and arterial pressure regulation in conscious man. Clin. Sci. 30:161, 1966. 100. Guz, A., and Widdicombe, J. G.: Pattern of breathing during hypercapnia before and after vagal blockade in man. In Porter, R. (ed.): Breathing: Hering-Breuer Centenary Symposium. A Ciba Foundation Symposium. London, J. & A. Churchill, p. 41, 1970. 101. Hanson, J., and Widén, L.: Afferent fibers in the hypoglossal nerve of cat. Acta Physiol. Scand. 79:24, 1970.

102. Harik, S. I., Ghandour, M. H., Farah, F. S., and Afifi, A. K.: Postganglionic cholinergic dysautonomia. Ann. Neurol. 1:393, 1977. 103. Harris, W.: Persistent pain in lesions of the peripheral and central nervous system. Br. Med. J. 2:896, 1921. 104. Hawe, P., and Lothian, K. R.: Recurrent laryngeal nerve injury during thyroidectomy. Surg. Gynecol. Obstet. 110:488, 1960. 105. Heaton, R. W., Guy, R. J., Gray, B. J., et al.: Diminished bronchial reactivity in cold air in diabetic patients with autonomic neuropathy. Br. Med. J. (Clin. Res. Ed.) 289:149, 1984. 106. Herxheimer, A.: Gustatory sweating and pilomotion. Br. Med. J. 1:688, 1958. 107. Hoffman, J. C.: Permanent paralysis of the accessory nerve after cannulation of the internal jugular vein. Anesthesiology 58:583, 1984. 108. Hollander, F.: Insulin test for presence of intact nerve fibers after vagal operation for peptic ulcer. Gastroenterology 7:607, 1946. 109. Holton, P., and Wood, J. B.: The effects of bilateral removal of the carotid bodies and denervation of the carotid sinuses in two human subjects. J. Physiol. (Lond.) 181:365, 1965. 110. Hosking, D. J., Moody, F., Stewart, I. M., and Atkinson, M.: Vagal impairment of gastric secretion in diabetic autonomic neuropathy. Br. Med. J. 2:588, 1975. 111. Humphreys, J.: A hazard of thyroidectomy. Proc. R. Soc. Med. 65:169, 1972. 112. Huppler, E. G., Schmidt, H. W., Devine, K. D., and Gage, R. P.: Ultimate outcome of patients with vocal-cord paralysis of undetermined cause. Am. Rev. Tuberc. Pulm. Dis. 73:52, 1956. 113. Ignelzi, R. J., and Bucy, P. C.: Intracranial hypoglossal neurofibroma—case report. J. Neurosurg. 26:352, 1967. 114. Inamdar, S., Easton, L. B., and Lester, G.: Acquired postganglionic cholinergic dysautonomia: case report and review of the literature. Pediatrics 70:976, 1982. 115. Jacobson, R. R., and Ross Russell, R. R.: Glossopharyngeal neuralgia: a rare but treatable cause of syncope. Br. Med. J. 1:379, 1979. 116. Jamshidi, A., and Masroor, M.: Glossopharyngeal neuralgia with cardiac syncope: treatment with a permanent cardiac pacemaker and carbamazepine. Arch. Intern. Med. 136:843, 1976. 117. Janssens, J., Peeters, T. L., Vantrappen, G., et al.: Improvement of gastric emptying in diabetic gastroparesis by erythromycin. N. Engl. J. Med. 322:1028, 1990. 118. Jenzer, G., Mumenthaler, M., Ludin, H. P., and Robert, F.: Autonomic dysfunction in botulism: a clinical report. Neurology (Minneap.) 25:150, 1975. 119. Kahn, A. A., Riazi, J., and Bloom, D.: Oculocardiac reflex in near miss for sudden infant death syndrome infants. Paediatrics 71:49, 1983. 120. Kahn, L. B.: Vagal body tumor (nonchromaffin paraganglioma, chemodectoma and carotid body–like tumor) with cervical node metastasis and familial association: ultrastructural study and review. Cancer 38:2367, 1976. 121. Kalia, M., and Davies, R. O.: A neuroanatomical search for glossopharyngeal afferents to the carotid body using horseradish peroxidase. Brain Res. 149:477, 1978.

Diseases of the Ninth through Twelfth Cranial Nerves 122. Kandt, R. S., and Daniel, F. L.: Glossopharyngeal neuralgia in a child: a diagnostic and therapeutic dilemma. Arch. Neurol. 43:301, 1986. 123. Kecht, B.: “Idiopathische” Stimmbandlämungen. Wien. Med. Wochenschr. 115:511, 1965. 124. Kerr, F. W. L.: Facial, vagal and glossopharyngeal nerves in the cat. Arch. Neurol. 6:264, 1962. 125. Koberle, F.: Zur Frage der Entstehung sogennanter “Idiopathischer Dilatation” muskulärer Hohlorgane. Virchows Arch. Pathol. Anat. Physiol. 329:337, 1956. 126. Koberle, F.: Megaesophagus. Gastroenterology 34:460, 1958. 127. Kramer, P., and Ingelfinger, F. J.: Esophageal sensitivity of Mecholyl in cardiospasm. Gastroenterology 19:242, 1951. 128. Kun`c, Z.: Treatment of essential neuralgia of the ninth nerve by selective tractotomy. J. Neurosurg. 23:494, 1965. 129. Laage-Hellman, J. E.: Gustatory sweating and flushing after conservative parotidectomy. Acta Otolaryngol. (Stockh.) 48:234, 1957. 130. Laha, R. K., and Janetta, P. J.: Glossopharyngeal neuralgia. J. Neurosurg. 47:316, 1977. 131. Lance, J. W., and Anthony, M.: Neck-tongue syndrome on sudden turning of the head. J. Neurol. Neurosurg. Psychiatry 43:97, 1980. 132. Lawn, A. M.: The localization, in the nucleus ambiguus of the rabbit, of the cells of origin of motor nerve fibers in the glossopharyngeal nerve and various branches of the vagus nerve by means of retrograde degeneration. J. Comp. Neurol. 127:293, 1966. 133. Le Quesne, L. P., Hobsley, M., and Hand, B. A.: The dumping syndrome. Br. Med. J. 1:141, 1960. 134. Levin, B., and Posner, J. B.: Swallow syncope. Neurology (Minneap.) 22:1086, 1972. 135. Lieschke, G. J.: Spontaneous internal carotid artery dissection presenting as hypoglossal nerve palsy. Stroke 19:1151, 1988. 136. Lindquist, C., and Mårtensson, A.: Reflex responses induced by stimulation of hypoglossal afferents. Acta Physiol. Scand. 77:234, 1969. 137. Logigian, E. L., McInnes, J. M., Berger, A. R., et al.: Stretch-induced spinal accessory nerve palsy. Muscle Nerve 11:146, 1988. 138. Lown, B., and Levine, S. A.: The carotid sinus: clinical value of its stimulation. Circulation 23:766, 1961. 139. Lucas Keene, M. F.: Muscle spindles in human laryngeal muscles. J. Anat. 95:25, 1961. 140. Macedo, T. F.: Bilateral hypoglossal palsy due to vertical sub-luxation of the odontoid process in rheumatoid arthritis. Br. J. Rheumatol. 27:317, 1988. 141. Machella, T. E.: Mechanism of the post gastrectomy dumping syndrome. Gastoenterology 14:237, 1950. 142. Mark, A. L., Kioschos, J. M., Abboud, F. M., et al.: Abnormal vascular responses to exercise in patients with aortic stenosis. J. Clin. Invest. 52:1138, 1973. 143. Marsden, P. D.: American trypanosomiasis: Chagas’ disease. In Weatherall, D. J., Ledingham, J. G. G., and Warrell, D. A. (eds.): Oxford Textbook of Medicine, 2nd ed. London, Oxford University Press, p. 520, 1987. 144. Mathias, C. J.: Bradycardia and cardiac arrest during tracheal suction—mechanisms in tetraplegic patients. Eur. J. Intens. Care Med. 2:147, 1976.

1291

145. Mathias, C. J.: Disorders of the autonomic nervous system. In Bradley, W. G., Daroff, R. B., Fenichel, G. M., and Jancovich, J. (eds.): Neurology in Clinical Practice. Boston, Butterworth-Heinemann, p. 2403, 2004. 146. Mathias, C. J., and Bannister, R.: Postprandial hypotension in autonomic disorders. In Mathias, C. J., and Bannister, R. (eds.): Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 4th ed. Oxford, UK, Oxford University Press, p. 283, 2002. 147. Maur, K., Nelson, E. W., Holsinger, J. W., and Eliot, R. S.: Hypersensitive carotid sinus syncope treated by implantable demand cardiac pacemaker. Am. J. Cardiol. 29:109, 1972. 148. Melerovich, A. E., and Liande, V. S.: Porazheniia gortani pri siringomielobul’bii. Vestn. Otorinolaryngol. 19:89, 1957. 149. Mikolich, J. R., Paulson, G. W., and Cross, C. J.: Acute anticholinergic syndrome due to jimson seed ingestion. Ann. Intern. Med. 83:321, 1975. 150. Mills, J. E., Sellick, H., and Widdicombe, J. G.: Epithelial irritant receptors in the lungs. In Porter, R. (ed.): Breathing: Hering-Breuer Centenary Symposium. A Ciba Foundation Symposium. London, J. & A. Churchill, p. 77, 1970. 151. Mitchell, G. A. G.: Anatomy of the Autonomic Nervous System. London, E. & S. Livingstone, 1953. 152. Mitchell, G. A. G., and Warwick, R.: The dorsal vagal nucleus. Acta Anat. 25:371, 1955. 153. Miura, M., and Reis, D. J.: Termination and secondary projections of carotid sinus nerve in the cat brain stem. Am. J. Physiol. 217:142, 1969. 154. Morimoto, T., and Kawamura, Y.: Discharge patterns of hypoglossal afferents in a cat. Brain Res. 35:539, 1971. 155. Munschauer, F. E., Loh, L., Bannister, R., and NewsomDavis, J.: Abnormal respiration and sudden death during sleep in multiple system atrophy with autonomic failure. Neurology 140:677, 1990. 156. Naunton, R. F., Proctor, L., and Elpern, B. S.: The audiologic signs of ninth nerve neurinoma. Arch Otolaryngol. 87:222, 1968. 157. Nelson, D. A., and Mahru, M. M.: Death following digital carotid artery occlusion. Arch. Neurol. 8:640, 1963. 158. Norden, A.: Peripheral injuries to the accessory nerve. Acta Chir. Scand. 94:515, 1956. 159. Novak, D. J., and Victor, M.: The vagus and sympathetic nerves in alcoholic polyneuropathy. Arch. Neurol. 30:273, 1974. 160. Olarte, M., and Adams, D.: Accessory nerve palsy. J. Neurol. Neurosurg. Psychiatry 40:1113, 1977. 161. Olivier, A., Scotti, G., and Melançon, D.: Vascular entrapment of the hypoglossal nerve in the neck. J. Neurosurg. 47:472, 1977. 162. Onrot, J., Wiley, R. G., Fogo, A., et al.: Neck tumour with syncope due to paroxysmal sympathetic withdrawal. J. Neurol. Neurosurg. Psychiatry 50:1063, 1987. 163. Orbay, T.: Late hypoglossal nerve palsy following fracture of the occipital condyle. Surg. Neurol. 31:402, 1989. 164. Paljärvi, L., and Partanen, J.: Biting palsy of the accessory nerve. J. Neurol. Neurosurg. Psychiatry 43:744, 1980. 165. Palmer, E. D.: Achalasia: response to treatment in longterm effect. Am. Practnr. Dig. Treat. 8:1595, 1957. 166. Palmer, R. A.: The treatment of carotid sinus syncope with propantheline. Can. Med. Assoc. J. 104:923, 1971.

1292

Diseases of the Peripheral Nervous System

167. Parry, C. H.: An Inquiry into Symptoms and Causes of Syncope Angiosa, Commonly Called Angina Pectoris. Bath, England, Cruttwell, 1799. 168. Patterson-Ross, J.: Surgery of the Sympathetic Nervous System, 3rd ed. London, Baillière, Tindall and Cox, 1958. 169. Pearson, A. A.: The spinal accessory nerve in human embryos. J. Comp. Neurol. 68:243, 1938. 170. Peddie, G. H., Jordan, G. I., and De Bakey, M. E.: Further studies on the pathogenesis of the dumping syndrome. Ann. Surg. 146:892, 1957. 171. Pesavento, G., Ferlito, A., and Recher, G.: Benign solitary schwannoma of the vagus nerve: a case report with a review of the literature. J. Laryngol. Otol. 93:307, 1979. 172. Petrera, J. E., and Trojaborg, W.: Conduction studies along the accessory nerve and follow-up of patients with trapezius palsy. J. Neurol. Neurosurg. Psychiatry 47:630, 1984. 173. Polinsky, R. J., Taylor, I. L., Chew, P., et al.: Pancreatic polypeptide responses to hypoglycemia in chronic autonomic failure. J. Clin. Endocrinol Metab. 54:482, 1982. 174. Quattrocolo, G.: Tapia’s syndrome caused by a neurilemmoma of the vagus and hypoglossal nerves in the neck. Acta Neurol. (Napoli) 8:535, 1986. 175. Quinn, J. H.: Glossodynia. J. Am. Dent. Assoc. 70:1418, 1965. 176. Rentmeester, T., Van Zile, J., Van Hal, M., and Levine, D.: Vasodepressor carotid sinus syncope. Br. Med. J. (Clin. Res. Ed.) 289:720, 1984. 177. Rexed, B.: Contributions to the knowledge of the postnatal development of the peripheral nervous system in man. Acta Psychiatr. Neurol. Scand. Suppl. 33:1, 1944. 178. Rhoton, A. L., O’Leary, J. L., and Ferguson, J. P.: The trigeminal, facial, vagal and glossopharyngeal nerves in the monkey. Arch. Neurol. 14:530, 1966. 179. Ripley, R. C., Hollifield, J. W., and Nies, A. S.: Sustained hypertension after section of the glossopharyngeal nerve. Am. J. Med. 62:297, 1977. 180. Roberts, K. E., Randall, H. T., and Farr, H. W.: Cardiovascular and blood volume alterations resulting from intrajejunal administration of hypertonic solutions to gastrectomised patients: the relationship of these changes to the dumping syndrome. Ann. Surg. 140:631, 1954. 181. Rossi, M., Marti, G., Ricordi, L., et al.: Cardiac autonomic dysfunction in obese subjects. Clin. Sci. 76:567, 1989. 182. Rubinstein, M. K.: Cranial mononeuropathy as the first sign of intracranial metastases. Ann. Intern. Med. 70:49, 1969. 183. Rushton, J. G., Stevens, C., and Miller, R. H.: Glossopharyngeal (vagoglossopharyngeal) neuralgia. Arch. Neurol. 38:201, 1981. 184. Sander-Jenson, K., Garne, S., and Schwartz, T. W.: Pancreatic polypeptide release during emotionally induced vasovagal syncope. Lancet 2:1132, 1985. 185. Sander-Jenson, K., Secher, N. H., Bie, P., et al.: Vagal slowing of the heart during haemorrhage: observations from 20 consecutive hypotensive patients. Br. Med. J. (Clin. Res. Ed.) 292:364, 1986. 186. Saviolo, R.: Treatment of glossopharyngeal neuralgia by carbamazepine. Br. Heart J. 58:291, 1987. 187. Scheuer, J. L.: Fibre size frequency distribution in human laryngeal nerves. J. Anat. 98:99, 1964.

188. Schmidt, D., and Strutz, L.: Superior laryngeal neuralgia. J. Neurol. 225:223, 1981. 189. Schneck, S. A.: Peripheral and cranial nerve injuries resulting from general surgical procedures. A. M. A. Arch. Surg. 81:855, 1960. 190. Schwartz, T. W.: Pancreatic polypeptide: a hormone under vagal control. Gastroenterology 85:1411, 1983. 191. Sharma, A. K., and Thomas, P. K.: Quantitative studies on age changes in unmyelinated nerve fibres in the vagus nerve of man. In Kunze, K., and Desmedt, J. E. (eds.): Studies in Neuromuscular Disease. Basel, Karger, p. 211, 1975. 192. Sharpey-Schafer, E. P., and Taylor, P. J.: Absent circulatory reflexes in diabetic neuritis. Lancet 1:559, 1960. 193. Shepherd, R. T.: Mechanism of sudden death associated with volatile substance abuse. Hum. Toxicol. 8:287, 1989. 194. Singh, S., and Schlagenhauff, R. E.: Pressure palsy of accessory nerve. Neurol. India 18:122, 1971. 195. Smith, R. O., Hemenway, W. G., Stevens, K. M., et al.: Jacobson’s neurectomy in Frey’s syndrome. Am. J. Surg. 20:478, 1970. 196. Spyer, K. M.: Central nervous control of the cardiovascular system. In Mathias, C. J., and Bannister, R. (eds.): Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 4th ed. Oxford, UK, Oxford University Press, p. 45, 2002. 197. Stankiewicz, J. A.: Hypoglossal nerve palsy after tooth extraction. J. Oral Maxillofac. Surg. 46:148, 1988. 198. Sunderland, S., and Swaney, W. E.: The intraneural topography of the recurrent laryngeal nerve in man. Anat. Rec. 114:411, 1952. 199. Sutton, J. A., Thompson, S., and Sobnack, R.: Measurement of gastric emptying rates by radioactive isotope scanning and epigastric impedance. Lancet 1:898, 1985. 200. Swann, K. W.: Accessory palsy following carotid endarterectomy: report of two cases. J. Neurosurg. 63:630, 1985. 201. Szentágothai, J.: Die Lokalization der Kehlkopfmuskulatur in den Vaguskernen. Z. Anat. Entwicklungsgesch. 112:704, 1943. 202. Takayama, H., Kazahaya, Y., Kashihara, N., et al.: A case of postganglionic cholinergic dysautonomia. J. Neurol. Neurosurg. Psychiatry 50:915, 1987. 203. Thesleff, S.: Neurophysiological aspects of botulism poisoning. In Lewis, G. E. (ed.): Biomedical Aspects of Botulism. New York, Academic Press, p. 65, 1981. 204. Thomas, J. E.: Hyperactive carotid sinus reflex and carotid sinus syncope. Mayo Clin. Proc. 44:127, 1969. 205. Thomashefsky, A. J., Horwitz, S. J., and Feingold, M. H.: Acute autonomic neuropathy. Neurology (Minneap.) 22:251, 1972. 206. Thompson, G. D., Wingate, D. L., Archer, L., et al.: Normal patterns of human upper small bowel motor activity recorded by prolonged radiotelemetry. Gut 23:517, 1982. 207. Thuluvath, P. J., and Triger, D. R.: Autonomic neuropathy in chronic liver disease. J. Med. 72:737, 1989. 208. Tresure, T.: Traumatic division of pharyngeal branch of vagus causing dysphagia. Br. Med. J. 1:808, 1976. 209. Trout, H., Brown, L. L., and Thompson, J. E.: Carotid sinus syndrome: treatment by carotid sinus denervation. Ann. Surg. 189:575, 1979.

Diseases of the Ninth through Twelfth Cranial Nerves 210. Turnbull, D. M., Venables, G. S., and Cartlidge, N. E. F.: Vagal neuropathy in the alcoholic. Lancet 2:1381, 1980. 211. Ueda, S.: Peripheral hypoglossal nerve palsy caused by lateral position of the external carotid artery and an abnormally high position of bifurcation of the external and internal carotid arteries—a case report. Stroke 15:736, 1984. 212. Vita, G., Girlanda, P., Puglisi, R. M., et al.: Cardiovascularreflex testing and single-fiber electromyography in botulism. Arch. Neurol. 44:202, 1987. 213. Walker, L. B., and Rajagopal, M. D.: Neuromuscular spindles in the human tongue. Anat. Rec. 133:438, 1959. 214. Wallin, B. G., and Sunlof, G.: Sympathetic outflow to muscles during vasovagal syncope. J. Auton. Nerv. Syst. 6:287, 1982. 215. Wallin, B. G., Westerberg, C. E., and Sundlof, G.: Syncope induced by glossopharyngeal neuralgia: sympathetic outflow to muscle. Neurology (Cleveland) 34:522, 1984. 216. Walter, P. F., Crawley, I. S., and Dorney, E. R.: Carotid sinus hypersensitivity and syncope. Am. J. Cardiol. 42:396, 1978. 217. Watkins, P. J.: Facial sweating after food: a new sign of diabetic autonomic neuropathy. Br. Med. J. 1:583, 1972. 218. Watkins, P. J., and Mackay, J. D.: Cardiac denervation in diabetic neuropathy. Ann. Intern. Med. 92:304, 1980. 219. Weber, E. L., and Odom, G. L.: Paradoxical response to peripheral hypoglossal nerve section. J. Neurosurg. 30:186, 1969. 220. Weisenburg, T. H.: Cerebellopontile tumor diagnosed for six years as tic douloureux: the symptoms of irritation of the ninth and twelfth cranial nerves. J. Am. Med. Assoc. 54:1600, 1910. 221. Weiss, S., and Baker, J. P.: The carotid sinus reflex in health and disease. Medicine (Baltimore) 12:297, 1933. 222. Welply, N., Mathias, C. J., and Frankel, H. L.: Circulatory reflexes in tetraplegics during artificial ventilation and during anaesthesia. Paraplegia 13:172, 1975. 223. Whisler, W. W., and Voris, H. C.: Effect of bilateral glossopharyngeal nerve section on blood pressure. J. Neurosurg. 23:79, 1965. 224. White, J. C., and Sweet, W. H.: Pain and the Neurosurgeon. Springfield, IL, Charles C Thomas, p. 265, 1969.

1293

225. Whitwam, J. G., Kidd, C., and Fussey, I. F.: Activity in efferent sympathetic nerves evoked by stimulation of the hypoglossal nerve. Brain Res. 14:756, 1969. 226. Wickramasinghe, L. S. P., Chowdrury, C. R., Pillai, S. S., and Ghosh, A.: Distended oesophagus as a cause of bilateral recurrent laryngeal nerve palsy. Postgrad. Med. J. 64:958, 1988. 227. Widdicombe, J. G.: Regulation of tracheobronchial smooth muscle. Physiol. Rev. 43:1, 1963. 228. Williams, A., Hanson, D., and Calne, D. B.: Vocal cord paralysis in the Shy-Drager syndrome. J. Neurol. Neurosurg. Psychiatry 42:151, 1979. 229. Williams, A. F.: Recurrent laryngeal nerve and thyroid gland. J. Laryngol. Otol. 68:719, 1954. 230. Williams, A. F.: Recurrent laryngeal nerve lesions during thyroidectomy. Surgery 43:435, 1958. 231. Williams, J. M., and Fox, J. L.: Neurinoma of the intracranial portion of the hypoglossal nerve: review and case report. J. Neurosurg. 19:248, 1962. 232. Williams, R. G.: Idiopathic recurrent laryngeal paralysis. J. Laryngol. Otol. 73:161, 1959. 233. Windle, W. F.: The sensory component of the spinal accessory nerve. J. Comp. Neurol. 53:115, 1931. 234. Wood, P. R., Foley, M. A., and Lawler, P. G.: Vagal reflexes and anaesthesia. Br. J. Hosp. Med. 44:137, 1990. 235. Wòzniak, W., and Young, P. A.: Nerve cells in the extracranial portion of the hypoglossal nerve in human fetuses. Anat. Rec. 162:517, 1968. 236. Yee, J., Harrison, F., and Corbin, K. B.: The sensory innervation of the spinal accessory and tongue musculature in the rabbit. J. Comp. Neurol. 70:305, 1939. 237. Yoshioka, J.: Combined trigeminal and glossopharyngeal neuralgia. Surg. Neurol. 24:416, 1985. 238. Yusuf, S., Theodropoulos, S., Mathias, C. J., et al.: Increased sensitivity of the denervated transplanted human heart to isoprenaline both before and after beta-adrenergic blockade. Circulation 75:696, 1987. 239. Zee-Cheng, C. S., and Gibbs, H. R.: Pure vasopressor carotid sinus hypersensitivity: unusual cause of recurrent syncope. Am. J. Med. 81:1085, 1986.

C. Diseases of the Spinal Cord, Spinal Roots, and Limb Girdle Plexus

53 The Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease CAROLINE M. KLEIN AND ANNABEL K. WANG

Overview Anatomic Considerations Motor Syndromes Classification Lower Motor Neuron Disease Combined Upper and Lower Motor Neuron Disease: ALS

Sensory Syndromes Cobalamin Deficiency Sjögren’s Syndrome Combined Motor and Sensory Syndromes Cervical Spondylotic Myelopathy

Clinical assessment of peripheral neuropathy usually involves evaluation of motor, sensory, or autonomic symptoms and their distribution, typically as these symptoms affect the distal extremities. Clinicians are faced with the task of determining whether or not the patient’s symptoms are due to disease of peripheral nerve as well as the type and anatomic distribution of the nerve fibers involved in the disease process. As part of this evaluation, it is important for the clinician to realize that peripheral neuropathy may represent disease of any portion of the peripheral nerve, including not only the peripheral process innervating target tissue (muscle, sensory receptors, visceral organs, etc.) but also dorsal and ventral nerve roots and dorsal root and autonomic ganglia, as the most proximal anatomic components of peripheral nerve. All of these structures are part of peripheral nerve, which, by definition, includes all proximal and distal portions of neurons that are located outside of the central nervous system (CNS) and are associated with Schwann cells and not oligodendrocytes. Dorsal root ganglion neurons are associated

AIDS-Related Vacuolar Myelopathy Neurosyphilis Bulbospinal Muscular Atrophy (Kennedy’s Disease) Meningeal Disease

with satellite cells, but their distal processes innervating target organs and their centrally directed axons to the point of entry into the dorsal horn reside in the peripheral nervous system (PNS), and therefore are susceptible to disease processes affecting peripheral nervous tissue. Likewise, although the cell bodies of lower motor neurons reside within the spinal cord or brainstem, once their axons exit from the CNS, they are associated with Schwann cells within the ventral root and are therefore peripheral nervous tissue. There are disease processes, such as amyotrophic lateral sclerosis (ALS), that lead primarily to degeneration of the entire motor neuron (partly in the CNS and partly in the PNS), but an early change may be an accumulation of neurofilaments in axons (a change typical of some neuropathies). Proximal involvement may be expressed as distal clinical manifestations. Preganglionic autonomic nerve fibers arising from neurons within the CNS have axons lying within both the CNS and PNS. They exit the spinal cord to synapse on the postganglionic component, which is entirely PNS. 1295

1296

Diseases of the Peripheral Nervous System

In considering the anatomic-pathologic locus of disorders with peripheral nerve manifestations, it is important to consider not only distal but also proximal levels of the neurons involved, that is, dorsal and ventral roots, dorsal root ganglion neurons, autonomic neurons, spinal nerves, plexuses, and finally the individual nerve trunks traveling to innervate target tissues in the limbs. Broadening the definition of peripheral neuropathy yields greater diagnostic possibilities in terms of consideration of many anatomicopathologic patterns such as neuronopathy, ganglionopathy, radiculoneuropathy, plexopathy, and distal polyneuropathy, among others. Ultimately, as clinicians evaluate patients presenting with distal muscle weakness or sensory or autonomic symptoms, recognition that these symptoms may be due to disease of either proximal or distal (or both) components of the neurons involved will result in better characterization of the underlying pathologic process as well as clinically better defined disease. Once the patient’s symptoms are defined anatomically in terms of level of involvement, correlation with known disease processes that affect certain levels of the neuron will result in greater diagnostic certainty and may result in more specific therapies. The objective of this chapter is to provide clinicians with an overview of pathologic processes affecting proximal portions of peripheral nerve, namely the neuronal cell bodies, roots, and proximal nerve fibers. Descriptions of these diseases are meant to provide clinicians with examples of situations in which they may need to consider what portion of the peripheral nerve is primarily involved and how to confirm that level of involvement with specific approaches or testing techniques, such as magnetic resonance imaging (MRI), cerebrospinal fluid (CSF) examination, somatosensory evoked potentials, and so on. The diseases selected for presentation in this chapter are meant to be representative but not inclusive of such considerations. Several diseases are discussed in greater detail in other chapters, to which the reader is referred for additional specific information as needed.

OVERVIEW Pathologic processes affecting peripheral nerves may be restricted to involvement of motor or sensory nerve cell bodies or their proximal or central axonal processes. Peripheral motor axons arise from the anterior horn cell located within the spinal cord or the brainstem motor nuclei, with the proximal axon found in the ventral root. Likewise, dorsal root ganglion cells, or primary sensory neurons, give rise to centrally directed and peripherally directed axonal processes. These anatomic sites of origin for peripheral nerve axons are important to consider in the pathogenesis of peripheral nerve disease, because these sites may be the location of the primary pathology

for a given condition or disease process. Determination of the site of the primary pathologic process producing neurologic symptoms in any given patient may have important clinical implications. For example, being able as a clinician to distinguish between ALS and pure motor neuropathy as a cause for progressive motor weakness in a patient is of vital importance in terms of prognosis and potentially available therapies. Likewise, primary neuronopathies may have pathologic implications for distal central or peripherally directed axons, which also need to be considered in terms of differential diagnosis and in regard to understanding the overall disease process and how it may manifest clinically and pathologically. Determination of which portions of the neuron are affected by the disease process and in what way is important to defining the stage and extent of the disease. With peripherally located, primarily distal disease of nerve processes, the chances for regeneration and repair are better than with those that are more centrally located, either in the soma, the roots, or centrally directed processes within the spinal cord. Many diseases damage the neuronal cell body as well as its central and peripheral processes in a global manner. In such situations, the clinician may not be able to limit consideration to the effects of the disease on peripherally located processes involved in isolation.188 Most disease processes that cause loss of the primary neuron, either motor or sensory, ultimately lead to secondary degeneration of the processes arising from that cell as a consequence. Whether that evolves in a proximalto-distal or distal-to-proximal sequence may depend on the underlying pathophysiology involved.179 Studies in neurotoxicology have provided useful methods to examine lesions affecting proximal versus distal neuronal processes and their implications. The reader is referred to the text by Spencer and Schaumburg180 for a detailed review of this subject. Methyl mercury and doxorubicin have selective toxicity involving dorsal root ganglion cells, but central and peripheral axons from these cells undergo degeneration as the cell bodies are lost. Distal axonopathies may involve peripherally or centrally directed distal axons. Isoniazid causes a “dying-back” toxic effect on distal peripheral sensory fibers, whereas clioquinol selectively damages the central afferent and efferent axons more than peripheral axons and results in deterioration of dorsal column and corticospinal tracts within the spinal cord preferentially. ␤,␤⬘-Iminodipropionitrile (IDPN) toxicity in animal models causes a proximal axonopathy affecting the proximal axons of anterior horn cells, with accumulation of neurofilaments and swelling thought to be due to a defect in slow axonal transport.68,188 In addition, with IDPN there is segmental demyelination and formation of onion bulbs in the distal peripheral nerve endings.74 There is progressive atrophy of the distal axon as a result of globally impaired axonal transport and secondary

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

demyelination (passive initially secondary to local edema, and later active demyelination) involving the proximal axon within the spinal cord ventral horn.68 In general, with dying-back neuropathies, whether central or peripheral, the length of the nerve fiber and its relative diameter are important factors in regard to selective vulnerability of particular parts of neurons in response to disease.188 The potential for central regenerative capacity in primary afferent nerve fibers has been studied in animal models of peripheral nerve injury. Woolf et al.199 demonstrated in a rat model of sural nerve transection in the leg that, within 1 week after injury, there was sprouting of retrogradely labeled afferent nerve fibers into lamina II of the dorsal horn of the spinal cord. Follow-up studies comparing transection versus crush injury of the sciatic nerve in the thigh, with prevention of peripheral nerve regeneration in the former and possible regeneration after injury in the latter, showed that there was sprouting of large myelinated axons into lamina II of the dorsal horn, which persisted for 6 months after injury and then disappeared by 9 months.200 These experiments suggested that, in response to peripheral nerve injury, there is A-fiber terminal reorganization in the dorsal horn by a mechanism of collateral sprouting conditioned by the injury to the peripheral sensory axon, because the sprouting centrally directed axon was not injured directly.199,200 Lamina II is normally where centrally directed unmyelinated or small myelinated afferent fibers terminate,34 so the sprouting of large myelinated fibers provides a novel input to this lamina. In these studies, the numbers of unmyelinated central afferent fibers did not differ from those of unoperated animals.200 Investigation of the differential response of centrally directed afferent fibers to different types of nerve injury has been accomplished with animal models of peripheral nerve injury. Chong et al.35 showed that sciatic nerve crush in rats resulted in an increase in growth-associated protein-43 (GAP-43) messenger RNA (mRNA) in dorsal root ganglion cells that occurred within 24 hours after injury and was maintained for 30 days postinjury. Ten days after sciatic nerve crush there was an increase in GAP-43 in the ventral horn motor neurons. However, dorsal rhizotomy produced no change in levels of GAP-43 in the dorsal root ganglion neurons or the dorsal horn of the spinal cord, even if the lesion was primed by peripheral sciatic nerve transection. Comparing dorsal rhizotomy in close proximity to the dorsal root ganglion versus a lesion made closer to the dorsal horn revealed a mild transient increase in GAP-43 in the dorsal root ganglion. GAP-43 is a marker for axonal growth in response to nerve injury.35 Thus the location of the axonal lesion in relation to the cell bodies of origin is crucial in regard to the response within the system to the injury. Coggeshall et al.39 compared sciatic nerve crush versus transection in the thigh of rats and quantified numbers of myelinated and unmyelinated axons in the L4 and L5 dorsal roots.

1297

These investigators found that the number of unmyelinated axons was reduced by 50% in the dorsal roots by 4 to 8 months postoperatively, without a concomitant change in the numbers of myelinated axons, suggesting that, following peripheral axotomy, the amount of unmyelinated fiber input to the dorsal horn is reduced with a shift toward greater large fiber afferent input. Interestingly, the type of peripheral nerve lesion did not affect these results.39 By comparison, ligation of dorsal roots distal to the dorsal root ganglion in the rat causes a decrease in the number of dorsal root unmyelinated axons and loss of dorsal root ganglion cells within the first few weeks, with an associated increase in the number of small and thinly myelinated fibers. By 32 weeks after ligation, the total number of dorsal root axons has recovered, but the number of dorsal root ganglion cells is less than 50% of that seen in control animals without ligation. These findings imply that there is axonal sprouting in the dorsal root, with loss of dorsal root ganglion cells secondary to peripheral nerve lesion.111 Transection of the sciatic nerve in rats compared to dorsal rhizotomy leads to an upregulation of heat shock protein 27 mRNA, a known promoter of cell survival and axonal sprouting, within the dorsal horn and dorsal columns of the spinal cord.40 This upregulation persists in the spinal cord for several months after peripheral nerve injury.40

ANATOMIC CONSIDERATIONS An understanding of the basic anatomy of the spinal cord and the relationships between motor and sensory cell bodies of origin and their central and peripheral processes is crucial in regard to an appreciation of how they may influence various pathologic processes. Motor neurons or anterior horn cells are found in lamina IX of the ventral horn of the spinal cord. The initial component of the motor axon lies within the CNS, myelinated by oligodendrocytes rather than Schwann cells, and surrounded by glia. The neurons that reside within the ventral horn consist of “root cells,” which give rise to axons that form the ventral or motor root that forms the motor component of the spinal nerve peripherally, and the “column cells,” whose peripheral processes remain within the CNS. In addition, there are internuncial neurons that form propriospinal circuits.28 There is somatotopic organization of groups of cells within the anterior horn, in that the medial nuclear group (posteromedial and anteromedial) motor neurons innervate axial musculature and the lateral nuclear group, which is subdivided into smaller subgroups as the cervical and lumbar enlargements, innervate distal extremity muscles (lateral subgroups) and proximal extremity muscles (medial subgroups). Ventral roots are composed of myelinated axons from motor neurons, including large-diameter myelinated axons from alpha

1298

Diseases of the Peripheral Nervous System

motor neurons destined to innervate extrafusal muscle fibers and small-diameter myelinated axons arising from gamma motor neurons to innervate intrafusal fibers within muscle spindles. There are autonomic preganglionic axons also within ventral roots, originating from specific segmental levels (T1-L2 sympathetic and S2-S4 parasympathetic general visceral efferent fibers), as well as a small number of unmyelinated primary afferent axons.28 Ventral roots combine with the dorsal root from the same segmental level to form the spinal nerve distal to the dorsal root ganglion within the intervertebral foramen. Schwann cells myelinate the axons within the dorsal and ventral roots as the proximal portions of the peripheral nerve. Dorsal root ganglia are the location of the primary sensory neurons, which are unique in their morphology as pseudounipolar neurons, with a peripherally directed axon that forms the spinal nerve in combination with the ventral root fibers and a centrally directed axon that terminates in the dorsal horn of the spinal cord. The initial process from the cell body is pseudounipolar in that the initial axon is unmyelinated and then bifurcates into the central and peripheral axons.68 The dorsal sensory roots contain both myelinated and unmyelinated axons that enter the spinal cord parenchyma, with large myelinated axons entering the spinal cord and traveling within the dorsal column white matter tracts or into deeper laminae of the dorsal horn. Small myelinated and unmyelinated axons from the dorsal root terminate within laminae I and II of the dorsal horn.28 The dorsal nucleus of Clarke is located within the spinal gray matter of the thoracic and upper lumbar segments and receives input as well from dorsal root afferent fibers, and is the origin of the posterior or dorsal spinocerebellar tracts.28 Because of the anatomic discrepancy between the length of the spinal cord and the surrounding bony spinal column, there is a longer distance between sensory ganglia and spinal cord in the lumbosacral region compared to the cervical or thoracic region.172 The blood-nerve barrier to the dorsal root ganglia is relatively deficient compared to that to the anterior horn cells, making the sensory neurons and axons more vulnerable to toxic and immune attacks.188 The lateral division of the dorsal root ganglion contains small-diameter neurons, which mediate nociceptive and temperature sensations, and the medial division contains large-diameter neurons that mediate proprioception, light touch, and mechanoreception.172 The primary sensory or motor neurons may respond morphologically to injury of the peripheral axon by central chromatolysis, with rounding of the soma, decreased staining of the Nissl substance with basophilic dyes resulting in a paler staining cytoplasm, and eccentric positioning of the nucleus.82,156 The nucleolus becomes enlarged with hypertrophy of the Golgi apparatus on a subcellular level.156 These changes typically precede pyknosis as the cell body degenerates and its nuclear material becomes condensed. A primary lesion of the cell body usually leads to secondary axonal and dendritic degeneration, which would evolve in a proximal-to-

distal gradient,179 as opposed to a dying-back process of distal axonopathy. The proximal axon, including the axon hillock and initial segment as well as the first few internodes of myelinated axons, is vulnerable to proximal neuropathic lesions and may also demonstrate early morphologic changes in response to injury.68 For motor neurons, the proximal axon lies at the transition zone between the CNS and PNS, and therefore may be a critically vulnerable location for processes affecting either peripheral or central nervous tissues.

MOTOR SYNDROMES Classification In patients who present with muscle weakness and in whom a diagnosis of peripheral neuropathy is being considered, careful investigation in order to optimally localize the portions of the motor neuron involved, either proximal or distal or both, is required for accurate diagnosis. Without associated sensory symptoms or findings, one cannot exclude a primary neurogenic process affecting the anterior horn cell or its proximal axon because such a process may appear clinically indistinct from a more distal motor axonopathy. Diseases that affect the lower motor neuron or its peripheral process range from inherited diseases such as spinal muscular atrophy to ALS, which also includes involvement of the upper motor neuron, to diseases that affect primarily the ventral root (radiculopathies) or selectively motor axons within peripheral limb nerves, such as multifocal motor neuropathy with conduction block (MMN-CB). Diagnostic confusion may arise from situations in which patients present with muscle weakness in a single extremity or nerve distribution, without or with minimal sensory symptoms, so that any of these diagnoses may be possible. Certain clinical features such as a positive family history or upper motor neuron findings on examination may help to narrow the diagnostic considerations, but in many cases either the progression of the disease process or specific tests are required in order to determine one of these diagnoses in an individual patient. Even in the situation in which upper motor neuron findings are present on examination, there is a need to differentiate between ALS and cervical spondylotic myelopathy, for example, which may present very similarly. In this section of the chapter, we discuss diseases of the anterior horn cell, the ventral root, and the peripheral motor nerve that may have similar clinical manifestations and be difficult to distinguish with certainty.

Lower Motor Neuron Disease In terms of diseases that affect primarily the lower motor neuron, the primary site of pathology may be the anterior horn cell in the spinal cord or motor nuclei of cranial nerves within the brainstem. The axon of the lower motor neuron may be affected in ventral root disease or pure motor

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

neuropathies. The reader is referred to an available review of lower motor neuron syndromes for details.192 Briefly, inherited disease of the lower motor neuron includes spinal muscular atrophy and spinobulbar muscular atrophy (Kennedy’s disease), which involve genetic defects leading to premature death of lower motor neurons. There are subtypes of familial ALS that have predominantly lower motor neuron findings, specifically the A4V mutation of the superoxide dismutase 1 gene.192 Other diseases, which have unclear genetic links, include progressive spinal muscular atrophy, distal spinal muscular atrophy, segmental spinal muscular atrophy (distal and proximal forms), and monomelic amyotrophy of the lower limb.192 Acquired lower motor neuron syndromes may include poliomyelitis and postpolio syndrome, postradiation lower motor neuron syndrome, paraneoplastic motor neuron syndromes, human immunodeficiency virus (HIV)–associated motor neuron disease, West Nile virus (WNV)–associated paralysis, spondylotic myelopathy, and MMN-CB. Some of these diseases variably involve upper motor neurons, which may lead to confusion in differentiating them from ALS, for example, or from each other. Distal spinal muscular atrophy, which may be sporadic or inherited, typically affects the lower extremities with muscle wasting, weakness, and fasciculations. However, demonstration on electrodiagnostic testing of normal motor nerve conduction velocities and normal sensory nerve conduction studies may help to differentiate this disease from hereditary motor and sensory neuropathy (HMSN) type I.46 Monomelic, focal or segmental spinal muscular atrophies are sporadic, typically involve a single upper extremity clinically in terms of distal muscle wasting and weakness, and predominantly affect young male patients.46 Van den Berg-Vos et al.193 recently published a cross-sectional study of 49 patients with sporadic, adultonset lower motor neuron disease with duration of more than 4 years and found that the typical patient was male, with onset of weakness in a single, distal upper extremity. Investigation revealed that four patients had focal spinal cord atrophy on MRI, two had elevated serum titers of anti-GM1 antibodies, and three patients had serum monoclonal proteins, suggesting more specific diagnoses or associated conditions. Based on the distribution of muscle weakness, these patients could be divided into groups including slowly progressive spinal muscular atrophy affecting the lower more than the upper extremities, symmetrical distal upper and lower extremity spinal muscular atrophy, and segmental (asymmetrical) distal or proximal spinal muscular atrophy.193 Overall, these patients did not develop upper motor neuron signs after a long period of observation and had a fairly good prognosis. Distal and segmental spinal muscular atrophy involving the upper extremities in young male patients was described initially by Hirayama and co-workers in 1963.75 Their patients presented with insidious onset of distal, asymmetrical muscle weakness in the upper extremities,

1299

with normal or reduced reflexes and little or no sensory symptoms and no sensory findings on examination. There was chronic denervation on needle electromyography, including muscles examined in the contralateral and asymptomatic upper extremity in a subset of patients.75 Sobue et al.178 later reviewed 71 similar cases and included a potential inherited case involving an affected father and son. In this series, 47 of 71 patients had a single upper extremity affected and 24 of 71 patients had both upper extremities involved. These patients had normal nerve conduction studies and CSF protein levels, with neurogenic changes on electromyography, including neurogenic changes in the contralateral limb in a majority of patients64 and mild neurogenic changes in the lower extremities in a “small percentage.”178 These patients had more rapid progression of their lower motor neuron disease within the first 2 years after onset; then their clinical course became slowly progressive or static.178 A subset of these patients has been found to have cervical spinal cord focal atrophy or abnormal signal intensities on MRI, which has led to some controversy as to whether the pathology is based on flexion-induced cervical myelopathy with secondary ischemia.64 This disease may be differentiated from ALS by the typical young age of onset, the characteristic progression, and absence of deep tendon reflexes.64 Another group of patients who present with upper extremity muscle weakness and atrophy are those with brachial amyotrophic diplegia and those with “flail arm” syndrome. Hu et al.79 described a subset of patients diagnosed with ALS who had symmetrical, severe weakness of the upper extremities bilaterally, proximally and distally, with a male predominance. Compared to the larger group of ALS patients, these patients showed a trend toward longer survival, but most eventually did develop bulbar signs and lower extremity spasticity, suggesting that they do represent a variant of ALS.79 A more benign form of lower motor neuron disease affecting the upper extremities, without significant progression after more than 18 months of follow-up since onset, is brachial amyotrophic diplegia as described by Katz et al. in 1999.93 These patients were a small percentage of patients seen for motor neuron disease. They had normal or mildly reduced compound muscle action potential amplitudes in the upper extremities, with preserved motor conduction velocities, F-wave latencies, and no evidence of conduction block with acute or chronic denervation with needle electromyography in the upper extremities. Their weakness was proximally more than distally distributed, and 9 of 10 patients had no upper motor neuron signs, implying limited lesion of the cervical anterior horn cells.93 In comparing patients with brachial amyotrophic diplegia to those classified as having “flail arm” syndrome by Hu et al., Katz et al.93 noted that the latter group probably contained a mixture of patients with classic ALS (with upper motor neuron signs and muscle weakness in other segments in addition to the arms as their disease progressed) and patients with brachial amyotrophic diplegia.

1300

Diseases of the Peripheral Nervous System

Similar syndromes of lower motor neuron–type muscle weakness and atrophy isolated to the lower extremities have also been described in the literature. Prabhakar et al.155 reported on 40 patients in India with unilateral, nonprogressive, mild leg weakness and atrophy without a history of polio or affected family members. Electromyography of the affected limb was abnormal, with normal nerve conduction studies and neurogenic changes found on muscle biopsy, all of which implies isolated involvement of the lower motor neuron within the spinal cord.155 Rosenfeld et al.161 noted seven patients with onset of unilateral leg weakness that eventually involved the contralateral leg who were followed clinically for up to 8 years and did not develop upper extremity, bulbar, or respiratory weakness. Whether this also represents a limited form of progressive muscular atrophy or a limited lower motor neuron disease syndrome is speculative. Distal lower extremity weakness may also be due to inherited pathologic processes involving the lower motor neuron, such as distal hereditary spinal muscular atrophy, which may appear clinically similar to HMSN types I and II, which involve the peripheral nerve. Distal spinal muscular atrophy has relative preservation of most deep tendon reflexes, limited upper extremity involvement, normal clinical sensory examination, and normal sensory nerve action potentials and motor nerve conduction velocities, all of which are in contrast to the typical findings in HMSN type I or II. However, as in HMSN, these patients may manifest foot deformities and onset in young adulthood.69 The fact that the motor nerve conduction velocities are usually normal indicates that this is a process of primary neuronal involvement, compared to a distal axonopathy, which would result in slowed conduction velocities as a result of secondary demyelination or a primary demyelinative process.69 Poliomyelitis Poliomyelitis represents an infectious etiology for selective damage to anterior horn cells, which presents as acute muscle weakness in the setting of a febrile illness.133 Poliovirus is one member of the Picornaviridae family of RNA viruses, genus Enterovirus, which causes acute myelitis and destruction of anterior horn cells.118,125 Poliovirus type 1 is the most frequent cause of epidemic paralytic illness, which is transmitted person to person. The incubation period has been estimated to be from 4 to 14 days.64,78,125 Various estimates have been published, but between 1 in 1000 and 1 in 100 patients infected with the virus develop paralytic or symptomatic disease,125 representing approximately 0.02% of those infected according to recent literature.64 Children present with diphasic illness, with an episode of nonspecific viral illness followed by acute onset of paralysis, whereas adult patients typically have a more indolent onset of paralysis in the setting of severe myalgias.64,78,125,133 Early in the course of the paralytic illness, there is nuchal rigidity and

spastic painful muscles, particularly of the paraspinal and hamstring muscle groups.23,64,118,133 Patients may have brisk deep tendon reflexes initially, with some patients having Babinski signs, but as the paralysis proceeds, deep tendon reflexes are lost.23,64,133 The early spasticity may be due to early loss of internuncial neurons affecting intraspinal circuitry before the effects of anterior horn cell loss become apparent.23 Proximal more than distal limb muscles become weakened in an asymmetrical or segmental distribution, with sparing of sensory function clinically.125 Buchthal23 proposed that the reason for proximal muscle involvement was the larger diameter motor axons innervating these muscles compared to those innervating smaller distal limb muscles. Deep tendon reflexes are lost within 12 to 24 hours after the onset of muscle weakness.78 Mortality rates have been estimated to be between 5% and 8%78 in early literature, but more recently mortality rates are 2% to 5% for children and between 15% and 30% for adult patients.64 Early in the course of the illness, CSF demonstrates increased polymorphonuclear lymphocytes and later lymphocytes and monocytes, which clear from the CSF after the first week of the illness.78,125,133 CSF protein gradually increases during the first 3 weeks after the onset of weakness.125,133 Therefore, at a certain point during the illness the CSF cell count may be relatively low in the setting of rising protein levels, leading to potential diagnostic confusion with acute or chronic inflammatory demyelinating polyradiculoneuropathies.78 Some discriminating clinical features between acute poliomyelitis and Guillain-Barré syndrome (GBS) include symmetrical muscle weakness, sensory loss, and less prominent muscle pain and stiffness in GBS.78 Electrophysiologically, sensory nerve conduction studies are normal in poliomyelitis, with reduced compound muscle action potential amplitudes and relative preservation of conduction velocities.64 The reader is referred to an available review of the electromyographic findings in patients with poliomyelitis for additional details.64 Pathologically, poliovirus enters the human body via an oral route, with shedding of virus in lymphatic tissue associated with the alimentary tract, leading to viremia with entrance of the virus across the blood-brain barrier into the CNS with subsequent infection and replication in lower motor neurons.118,140 The neurotropism of the virus is based on binding to specific neuronal receptors. Once inside the neuron, the virus replicates and releases its RNA into the cytoplasm, which leads to neuronal degeneration.118 Ohka and Nomoto,140 using a transgenic mouse animal model, provided evidence for an additional mode of entrance of the virus via fast retrograde axonal transport from muscle tissue via the human poliovirus receptor. In 1949, Bodian18 published a review of findings in 24 human autopsy cases of poliomyelitis as well as results of an experimental monkey model for the infection. In the infected

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

monkeys, the earliest changes in the ventral horn after infection with polio are central chromatolysis and localized inflammatory reaction, which later becomes more diffuse. The inflammatory response only develops where neurons are still present, and the severity of the inflammation correlates with the amount of anterior horn cell degeneration.18 Subclinical involvement of the motor precentral cortex and subcortical motor areas has been noted.18,23,118 Because in most patients there is some degree of return of motor function subacutely, it is postulated that there is reversible injury in a subset of anterior horn cells, which may recover morphologically within 1 month after infection.18 In limbs with complete paralysis and no or little recovery, less than 10% of the anterior horn cells survived in the corresponding spinal cord segment.18 Bodian18 also noted that in the autopsy specimens there were some pathologic abnormalities in the dorsal root ganglia and posterior columns at levels of the spinal cord, with the most severe damage to the ventral horn, which may explain why patients complain of localized limb pain prior to the onset of paralysis. Other investigators have noted mild inflammatory changes in the dorsal root ganglia in patients with polio.23,82 Chronically, besides loss of anterior horn cells and fibrosis of the ventral roots, there may be mild perivascular lymphocytic infiltrates and gliosis of the anterior horn.89,118 Pathology studies of spinal cords from patients with a history of poliomyelitis have shown differences between patients with stable neuromuscular symptoms and those with new, slowly progressive muscle weakness, diagnosed with postpolio progressive muscular atrophy, or “postpolio syndrome.” In both groups of patients, there was loss or atrophy of lower motor neurons in the cervical and lumbar cord enlargements in anteromedial and anterolateral cell groups, with severe reactive gliosis, but in the postpolio syndrome patients there was an additional finding of axonal spheroids and occasional chromatolytic changes,130,153 consistent with more recent neuronal deterioration, suggesting that in these patients some of the surviving motor neurons become unable to maintain cellular function over time.153 The scattered axonal spheroids are immunoreactive for neurofilament protein, as noted in a recent case report of a patient who developed newly progressive muscle weakness 30 years after his initial attack of polio.130 The surviving lower motor neurons demonstrated atrophy, lipofuscin accumulation, and decreased amounts of Nissl substance.153 The amount of chronic perivascular and parenchymal inflammation was similar in both groups.130,153 By some estimates, approximately 25% to 30% of patients will develop symptoms consistent with postpolio syndrome 20 to 30 years after their initial paralytic illness.64 West Nile Virus The first North America outbreak of WNV was recognized in 1999, when two cases of encephalitis associated

1301

with muscle weakness were reported to the New York City Department of Health. Ten percent of the patients in the New York outbreak had complete flaccid paralysis, and half of the hospitalized patients were reported to have severe muscle weakness.135 Initially, the patients were thought to have GBS, or acute inflammatory demyelinating polyradiculoneuropathy. Findings of lymphocytic pleocytosis (30 to 100 cells/␮L; range, 0 to 1800 cells/␮L) with normal glucose and high protein (80 to 105 mg/dL, up to 1900 mg) in the CSF and predominant axonal rather than demyelinative changes found on electrophysiologic studies made the early diagnoses of GBS unlikely.5,11,25,135 Later cases were recognized to have electrophysiologic studies consistent with a disorder of the anterior horn cells and/or their axons (i.e., a poliomyelitis-like disorder): decreased compound muscle action potential amplitudes without evidence of demyelination (normal distal latencies and conduction velocities), normal sensory nerve action potential amplitudes, and needle electromyographic findings of denervation without evidence of myopathy. Weakness was asymmetrical, often proximal, with areflexia and no or minimal sensory abnormalities.29,62,110,113,142 The pattern of weakness was often severe, leading to respiratory failure requiring mechanical ventilation, but transient weakness and severe fatigue were also reported in some cases.25,109 WNV, first identified in Uganda, is indigenous to Africa, Asia, Australia, and Europe.25 An early report of a “polio syndrome” or anterior myelitis suggested the selective involvement of motor neurons in the brainstem and spinal cord by WNV. A 22-year-old man developed a high fever with pain in his neck, back, and left leg. He developed mild left facial weakness with fasciculations and flaccid paralysis of the left lower extremity, but had preserved sensation. WNV was diagnosed by the progressive increase of complement fixation antibodies. Eighteen months later, the patient still had diminished strength and reflexes of the left leg with mild atrophy.60 Infection by WNV, a single-stranded RNA virus of the Flaviviridae family, genus Flavivirus, has been reviewed extensively.25,76,149,150,160,163 The virus is primarily transmitted to humans by mosquitoes (Culex species) from the blood of infected birds, but can also be transmitted by blood transfusion, organ donation, breast-feeding, and percutaneous inoculation in the laboratory.8,30–33,59,84,146 The incubation period ranges from 2 to 14 days. Approximately 1 in 5 infected persons will develop a mild febrile illness after infection, whereas 1 in 150, more commonly older patients, will develop neurologic symptoms including meningitis, encephalitis, meningoencephalitis, or flaccid paralysis.149 Rare cases of associated brachial plexopathy,7 rhomboencephalitis,136 areflexic quadriplegia with absent brainstem reflexes,48 rhabdomyolosis,48,85 and movement disorders (parkinsonism, tremor, and myoclonus)173 have also been reported.

1302

Diseases of the Peripheral Nervous System

The diagnosis of WNV in patients with neurologic symptoms requires one of the following160: 1. WNV immunoglobulin M (IgM) antibody in CSF detected by enzyme-linked immunosorbent assay 2. WNV RNA in CSF 3. Fourfold increase in immunoglobulin G (IgG) antibodies between acute sera and convalescent sera obtained 4 weeks later 4. Isolation of the virus from brain or spinal cord Positive IgM and IgG antibody titers by enzyme-linked immunosorbent assay should be confirmed with more specific plaque-reduction neutralization assay in cell culture because false-positive serologic results occur as a result of the cross-reactivity between other flaviviruses, such as St. Louis encephalitis virus.25 Results may not become positive for 8 days after symptom onset but can persist for up to 500 days.149,159 IgM does not cross the blood-brain barrier, so a positive CSF IgM is diagnostic for WNV.25 A positive test result has high predictive value in the months of July to December, particularly if there is additional evidence of WNV activity in the area. A large percentage of subclinical infections, however, do occur.81,159 MRI revealed nonacute abnormalities and bilateral focal lesions in the basal ganglia, thalamus, and pons on T2- and diffusion-weighted images.173 Abnormal signal in the anterior horns has been demonstrated in the lumbar spine on MRI.113 Leptomeningeal, periventricular, isolated nerve root, and cauda equina enhancement have also been reported.96,135,173 Treatment is mainly supportive. One patient in the New York outbreak received plasmapheresis, but it was unclear if recovery was related to treatment. Administration of intravenous immunoglobulin (IVIG) for cases of presumed WNV-associated GBS has not been found to be effective to date.11,57,113,173 A multicenter study has been initiated to study the effects of IVIG in patients with suspected WNVrelated meningoencephalitis and/or weakness.160 The risk of developing severe neurologic disease, longterm morbidity, and death is higher in patients over the age of 50.100,132,135,149 Case-fatality ratios range from 4% to 14%.25,113,135 Recovery of muscle weakness is generally poor and incomplete as a result of the amount of axonal damage and is dependent on collateral sprouting. Autopsy findings in patients with meningoencephalitis revealed microglial nodules, mononuclear infiltrates, perivascular inflammation, and leptomeningitis, as well as focal mononuclear inflammation in cranial nerve roots. Inflammatory destruction of brainstem motor nuclei and anterior horn of the spinal cord along with patchy gliosis have also been reported.4,48,57,62,85,96,97,110,113,165 Inflammatory infiltrates revealed CD3⫹ lymphocytes (slight CD8 over CD4 predominance), CD68⫹ macrophages, and rare CD20⫹ B lymphocytes.97 The pathologic changes in the CNS are due to viral proliferation, cytotoxic immune

response, diffuse perivascular inflammation, and microglial nodule formation.25 Infiltrates in the dorsal roots have not been reported.48 The pathologic changes are focal and mild compared to eastern and western equine encephalitides. Vasculitis has not been demonstrated pathologically. The brainstem is frequently involved.164 Limited muscle pathology in one case of rhabdomyolysis revealed only mild focal inflammation. There was no evidence of WNV or necrosis in the muscle.48 The pathogenesis of WNV-associated weakness is not yet understood. Intracerebral inoculation of WNV in monkeys led to the development of excitement, slowness of movements, clumsiness, and ataxia. Paralysis was only seen in 5 of 15 monkeys. Lesions consisting of degenerative neuronal changes, perivascular infiltrates, changes of glia, and cell nodules were found predominantly in the cerebellum, spinal cord, and medulla oblongata but were seen scattered also throughout the brain. Neuronal degeneration consisted of either dissolution of the Nissl substance, pyknosis of the nuclei, or chromatolysis; swelling of the cytoplasm; and peripheral apposition of the pyknotic nuclei. Focal and severe meningitis was seen over the cerebellum. No characteristic changes were seen in the peripheral sensory ganglia.123 A new animal model, using the golden hamster (Mesocricetus auratus),202 has been developed. Signs consistent with meningoencephalitis (lethargy and paralysis) were reported 7 days after intraperitoneal inoculation with WNV. Pathologic findings occurred in the following sequence: infection of neurons, then inflammatory infiltrate (perivascular inflammation and microgliosis), followed by the presence of well-formed microglial nodules with degenerating neurons, consistent with apoptosis. The findings suggest that the main mechanism of neuronal damage is direct infection with WNV. The viral load determines the severity of the infection, that is, whether cells undergo necrosis or apoptosis.37 Two other studies also suggest that WNV induces cell death by apoptosis. WNV-induced apoptosis was shown to be associated with upregulation of bax gene expression.143 The WNV capsid, a pathogenic protein, was also demonstrated in vitro to induce apoptosis via the mitochondrial/caspase-9 pathway. Capsid gene injected into the striatum of mouse brain and muscle led to inflammation and cell death in tissue culture.204 Paraneoplastic Lower Motor Neuron Syndrome Lower motor neuron syndromes may manifest in patients with occult malignancy, serum monoclonal proteins, and other lymphoproliferative diseases. Presumably, in the case in which a known paraneoplastic antineuronal antibody is present, the lower motor neuron dysfunction is related to selective vulnerability of these neurons to it. Verma and co-workers194 reported a case with autopsy findings of a 51-year-old man with progressive upper more

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

than lower extremity weakness, small cell lung cancer, and high titers of anti-Hu antibodies. Nerve conduction studies were normal, but there were denervation abnormalities on needle examination,194 which implies disease of the motor neuron with relatively sparing of the peripheral motor axons.15 The patient’s CSF protein was mildly increased but cytology was negative for malignant cells. Autopsy findings included decreased numbers of cerebellar Purkinje cells, loss of large anterior horn cells, and lack of inflammation. Corticospinal and corticobulbar tracts were normal pathologically.194 Additional case reports, in patients with breast cancer, have demonstrated reversible lower motor neuron syndromes. Ferracci et al.56 reported a patient who had weakness progressing to areflexic quadriparesis in 2 months’ time, with preserved sensation and absent upper motor neuron findings. Cervical spine MRI reportedly showed T2-weighted signal hyperintensities in the gray matter. Initial electrodiagnostic testing showed normal motor and sensory nerve conduction studies, with neurogenic changes on electromyography. The patient had mild improvement after cancer treatment but no improvement with additional immunosuppressive therapy.56 The authors postulated that a partially reversible lesion of the axon initial segment or nodes of Ranvier secondary to autoimmune attack by an antineuronal antibody was present.56 A more recent case of a patient with breast cancer and lower motor neuron syndrome confirmed the presence of serum autoantibodies to isoforms of beta-IV spectrin and an unspecified surface antigen localized to axonal initial segments and nodes of Ranvier.14 In this particular patient, clinical improvement in muscle strength occurred following treatment of breast cancer with reduction of autoantibody titers.14 A similar association between lymphoma and motor neuron diseases (including MMNCB, lower motor neuron syndrome, and ALS) has been proposed by some.117,205 Parry and co-workers145 reported a patient who developed flaccid quadriplegia over a period of 4 months. After initial acral paresthesias, the patient had no significant sensory symptoms or signs. Electrophysiologically the motor nerve conduction studies showed low amplitudes, but were otherwise normal. However, acute and chronic neurogenic changes were present in all limbs on electromyography. The patient had an IgM gammopathy, with elevated CSF protein. Although he improved clinically with immunosuppressive therapy, he later died as a result of a pulmonary embolus. Postmortem examination revealed central chromatolysis in anterior horn cells with occasional axonal swellings, loss of axons in the ventral roots with focal swellings, and endoneurial lymphocytic perivascular infiltrates. Dorsal horn and roots were spared.145 This case demonstrates proximal motor axonopathy associated with monoclonal gammopathy that mimicked a lower motor neuron disease. A similar case of a patient with a progressive lower motor neuron syndrome and serum monoclonal IgM

1303

protein was also reported with autopsy findings demonstrating normal numbers of anterior horn cells in the lumbar cord, although some central chromatolysis was noted. The ventral roots were severely affected, with loss of nerve fibers and evidence of demyelination, which extended to a lesser degree into the proximal peripheral nerves. A primary motor radiculopathy with retrograde changes in the ventral horn was postulated in this patient.162 Postirradiation Lower Motor Neuron Syndrome A lower motor syndrome that may mimic motor neuron disease may be seen in patients who have received radiation therapy for pelvic cancer, typically testicular cancer.20,46,64 The syndrome presents as slowly progressive weakness and atrophy in distal leg muscles with reduced to absent reflexes and minimal, if any, sensory symptoms or findings, which, if present, are very distally distributed.20,64 In some patients, there may be sphincter dysfunction.20 These symptoms may develop up to 25 years after the radiation is administered.20,64 Bowen et al.20 described six patients with this presentation, and they noted that, in two of three patients who were examined with MRI of the lumbosacral spine, there was linear and focal enhancement of the cauda equina. Electromyography demonstrated denervation in L5-S1 myotomes, with normal sural sensory conduction studies. At autopsy of one of these patients, the cauda equina roots were found to be irregularly thickened with focal areas of hemorrhage, fibrosis associated with loss of axons, and changes consistent with a radiation-induced vasculopathy affecting selectively and primarily the lumbosacral nerve roots.20 HIV Infection and Motor Neuron Disease There have been isolated case reports of patients who are infected with HIV who develop motor neuron disease. Many of these cases presented with lower motor neuron syndromes, with progressive bulbar and limb weakness, preserved sensation, normal or minimally abnormal nerve conduction studies, and neurogenic changes on electromyography. 61,80,137,147,169 In some cases, treatment with antiviral medication resulted in improved motor function,137 but no improvement with treatment was reported in others.80 Description of limited postmortem examination of one of these cases demonstrated neuronal atrophy and occasional intracytoplasmic lipofuscin and central chromatolysis in brainstem motor nuclei.61 Verma et al.195 described a case of myeloradiculoneuropathy and myopathy as an initial clinical manifestation of HIV infection. Postmortem examination in that case showed preserved upper motor neurons, normal number and morphology of anterior horn cells, but capillary proliferation and hemorrhages in the anterior horn and ventral roots with depletion of ventral root axons with fibrosis.195

1304

Diseases of the Peripheral Nervous System

Other Causes of Lower Motor Neuron Disease Amyloid neuropathy has rarely been reported to present with progressive symmetrical weakness with reduced reflexes, initially intact sensation, and normal nerve conduction studies and neurogenic changes on electromyography, all of which are suggestive of lower motor neuronopathy.158 Amyloidosis was diagnosed after fascicular nerve and muscle biopsies were performed. This patient only later developed the more common symptoms and findings of axonal sensorimotor peripheral and autonomic neuropathies.158 Lower motor neuron involvement may also appear as a prominent feature of prion disease.82,201 Worrall et al.201 reviewed 50 reported cases with various prion diseases, all of whom had amyotrophy and pathologic evidence of anterior horn cell degeneration at postmortem examination.

Combined Upper and Lower Motor Neuron Disease: ALS Patients with progressive muscular atrophy, which denotes lower motor neuron degeneration without upper motor neuron involvement, overlap clinically with patients who have ALS but who, in the early stages before upper motor neuron symptoms and signs develop, may have muscle weakness, atrophy, and fasciculations limited to a single extremity. One of the earliest pathologic studies of patients with ALS and progressive spinal muscular atrophy, by Lawyer and Netsky in 1953,107 confirmed that in the latter group of patients there was loss of anterior horn cells without degeneration of the corticospinal or pyramidal tracts within the spinal cord. Interestingly, these authors reported a single patient in their series who had a 35-year history of weakness and atrophy without upper motor neuron findings but did develop them prior to his death, with consequent pathologic evidence of demyelination of the corticospinal tracts, suggesting a form of progressive spinal muscular atrophy that may transition into ALS very late in its course, clinically and pathologically.107 In the spinal cord preparations from their patients, they also noted small perivascular lymphocytic infiltrates “in some cases” localized primarily to the gray matter of the cord, which they ascribed to a reaction to the primary change in the motor neurons in the spinal cord, which showed decreased amounts of chromatin, pyknosis, shrinkage, and axonal changes.107 Other investigators have noted lymphocytic infiltrates in the ventral horns and pyramidal tracts of the spinal cords in a percentage of patients with ALS.189 More recently, Ince et al.83 reviewed autopsies from patients with premorbid diagnoses of progressive muscular atrophy (10 patients) and ALS (63 patients), and 6 patients who seemed to have evolved clinically from the former to the latter, in addition to control patients without a history of neurologic disease. Of note, the progressive muscular atrophy patients had rapid clinical progression to death in less

than 2 years, presenting with upper extremity or bulbar weakness eventually also involving the lower extremities. Approximately 50% of the progressive muscular atrophy patients demonstrated pathologic changes in the corticospinal tracts, with more than a third with moderate or severe pathologic changes.83 More than 95% of patients in the study had ubiquitinated cytoplasmic inclusions within the lower motor neurons.83 Almost 90% of the patients clinically diagnosed with ALS had moderate to severe pathologic changes in the corticospinal tracts. With progressive upper extremity weakness without upper motor neuron signs, these patients may look very similar to patients with multifocal motor neuropathy, when, in fact, they are really a subgroup of progressive muscular atrophy that overlaps with ALS.83 Interestingly, in terms of the differences reported regarding pathologic correlates in the transitional form of progressive muscular atrophy, one finding was that 30% of those patients had pathologic abnormalities in the dorsal columns of the spinal cord (in addition to the ventral horn pathology), whereas only 8% of the progressive muscular atrophy and almost 50% of the ALS patients had abnormalities in the dorsal columns.83 Troost et al.189 examined the spinal cords of 48 ALS patients and reported that there were pathologic abnormalities in 85%. Brownell et al.22 examined pathologically the spinal cords of 36 classic ALS patients and 8 patients with possible progressive muscular atrophy and found that there was loss of lower motor neurons in the anterior horns and cranial motor nuclei without changes in the posterior columns or spinocerebellar tracts, with sparing of Clarke’s column. In the ALS patients, the degeneration of the corticospinal tracts was more severe at lower cord levels and was difficult to detect proximal to the level of the brainstem unless there was severe motor neuronal degeneration in the spinal cord.22 More extensive pathologic involvement of the central motor pathways, such as the thalamus, mammillary bodies, basal ganglia, substantia nigra, and subthalamic nucleus, was noted in a patient with ALS, although this patient also had supranuclear ophthalmoplegia as a result of astrocytosis in the superior colliculi and periaqueductal gray matter, raising the question of an alternative disease process.186 Even within the diagnostic realm of ALS, there are groups of patients who have primarily lower motor neuron involvement both clinically and pathologically. Familial ALS patients with the A4V mutation in superoxide dismutase 1, the most common mutation, have short disease durations but predominantly lower motor neuron findings.41 Cudkowicz et al.41 performed postmortem examinations of patients with this condition and compared their results to those in patients with sporadic ALS. Both groups of patients demonstrated severe loss of anterior horn cells with gliosis and intracytoplasmic inclusions in remaining anterior horn cells. However, the sporadic cases had severe abnormalities in the corticospinal tracts, in contrast to the familial cases. In this study, demyelination and gliosis in

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

the posterior columns, Clarke’s column, and the dorsal spinocerebellar tracts was more common in familial ALS but was noted in patients from both groups.41 Details of the typical pathologic findings in ALS and related motor neuron syndromes may provide important clues as to the features of the neurons that provide a substrate for these various disease processes. Hirano73 provided an early review of these features, which include spheroids, or focal accumulation of filaments with resulting focal enlargement of the neuronal process, and Bunina bodies, which are small, eosinophilic inclusions of dense granular material of unknown significance (Fig. 53–1). In comparing findings of ALS in human spinal cord with those of animals exposed to IDPN, they both have accumulations of filaments within axons, but the anterior horn cells in the animals do not demonstrate chromatolysis, nor do the animals have clinical weakness.73 The central chromatolysis that is seen in postmortem tissue in ALS patients resembles postaxotomy changes but differs in the presence of spheroids, argentophilic soma, and accumulations of filaments within the cytoplasm of the cell body, suggesting significant disruption of axonal flow within the neuron.73,122 Based on this comparison with the animal model, there is a pathologic process beyond axonal transport dysfunction that leads to the characteristic changes within the neurons in ALS. Extensive pathologic examinations of CNS tissue affected by ALS have shown that abnormalities in the process of phosphorylation of neurofilaments may lead to greater accumulation of the filaments within the cell bodies and proximal axons of lower motor neurons in ALS patients compared to age-matched controls.122 Spheroids have been found to be immunoreactive for monoclonal antibodies for phosphorylated neurofilaments.122,176 In ALS patients there is an absence of nonphosphorylated neurofilaments in the ventral roots as compared to control

FIGURE 53–1 Chromatolytic anterior horn cells and axonal spheroids in a case of amyotrophic lateral sclerosis. (Bielschowsky stain; ⫻550.) (From Hirano, A.: Aspects of the ultrastructure of amyotrophic lateral sclerosis. Adv. Neurol. 36:75, 1982, with permission.)

1305

roots, with an equivalent amount of phosphorylated neurofilaments in both groups.122 More recently, investigators have co-localized superoxide dismutase-1 and nitric oxide synthase to the accumulated neurofilaments in surviving upper and lower motor neurons in patients with ALS.36 It may be that peroxynitrite formation at the neurofilament light chain may lead to inhibition of phosphorylation of neurofilaments and eventually alter axonal transport, resulting in proximal accumulation of neurofilaments and ultimately motor neuron cell death.36 Additional pathologic findings in the anterior horn cells of patients with motor neuron disease include descriptions of Lewy body–like, eosinophilic cytoplasmic inclusions reported in patients with lower motor neuron syndromes,91,92 which were found in addition to spheroids, Bunina bodies, chromatolysis, and gliosis in the anterior horns. These Lewy body–like inclusions are immunoreactive for ubiquitin but do not contain neurofilaments.92 There is one report of similar findings in a patient with ALS.103 Other investigators have examined the synaptic alterations that occur in ALS. Sasaki and Maruyama168 found no difference in the amount and distribution of immunostaining for synaptophysin, a presynaptic vesicle membrane protein, in the ventral horns of patients with ALS compared to those with lower motor neuron disease only. With mild loss of anterior horn cells in either disease, there was decreased staining in the lateral or ventrolateral anterior horn neuropil. Synaptophysin immunostaining was preserved in the proximal dendrites and around the cell bodies of normalappearing anterior horn cells and reduced around the cell bodies and distal portion of proximal dendrites of degenerating or atrophic neurons.168 This decreased staining, presumably reflecting loss of synaptic inputs, was similar in both diseases, indicating that those inputs are not originating from upper motor neurons, which are only involved by definition in ALS.168 Matsumoto et al.126 confirmed these findings and also noted that, in the ventral horns of four patients with pyramidal tract transection resulting from focal spinal cord lesions, the neuropil synaptophysin immunoreactivity was not different from controls, also indicating that the lack of upper motor neuron input is not the cause for the altered staining in motor neuron disease patients. A subsequent ultrastructural analysis of the axon hillocks and initial segments of axons of normal-appearing anterior horn cells in patients with ALS, patients with lower motor neuron disease, and age-matched controls showed that in both disease states the mean number of synapses and mean length of synaptic contacts at the axon hillock are reduced. The mean cell body area was also reduced in ALS and lower motor neuron disease compared to controls.167 Morphometric analysis of the distribution of cell loss in the anterior horn in ALS has revealed that the amount of cell loss at a particular segment of the spinal cord is variable, with the anterolateral quadrants of the ventral horn being most affected. Variation in severity and distribution of cell

1306

Diseases of the Peripheral Nervous System

loss may occur between the two sides of the spinal cord at the same level.184 Based on serial section analysis of the lower cervical cord in patients with ALS, there appear to be zones of focal neuronal loss within the spinal cord, including subgroups of the ventral horn and Clarke’s column.184 Various hypotheses regarding the possible pathogenesis of ALS have been proposed based on the available pathologic data. Karpati et al.90 proposed that the vulnerable neurons contained bundles of neurofilaments in their dendrites that become progressively depleted, with associated dendritic atrophy and loss of afferent input into the soma leading to atrophy. Based on this theory, the neurofilaments accumulate in the soma and axon as a result of impaired transport of those neurofilaments into the dendrites. These authors proposed further that the early clinical manifestations of fasciculations and cramping were secondary to dendritic atrophy, which altered the balance of synaptic inputs to the motor neuron and motor unit.90 Others proposed that in ALS the earliest lesion was not in the lower motor neuron but resulted from loss of the upper motor neurons located in the precentral gyrus, which leads to secondary degeneration of the anterior horn cell in an anterograde fashion.53 This theory would eliminate an etiologic connection between ALS and progressive spinal muscular atrophy, despite their close similarities pathologically. The reader is referred to a recent review of potential mechanisms for cell death in motor neuron disease by Martin et al.124 Involvement of components of the nervous system beyond the upper and lower motor neurons in ALS has also been described in the literature. Wakabayashi et al.198 published a case report of a 57-year-old Japanese man who initially complained of patchy sensory symptoms but who progressed to have motor weakness and elevated CSF protein, who did not improve despite immunosuppressive therapies. Autopsy findings revealed loss of Betz cells in the motor cortex, degeneration of corticospinal and spinocerebellar tracts, severe loss of anterior horn cells with gliosis with surviving neurons containing Bunina bodies and bundles of ubiquitin-positive filaments, loss of neurons in Clarke’s column, and severe degeneration of posterior columns and dorsal root ganglion cells.198 Morphometric analysis of motor neurons, ventral root axons, dorsal root ganglion cells, and dorsal root axons in ALS patients compared to controls demonstrated loss of large motor and sensory neurons and their axons.94 Serial sensory studies have shown mild but progressive abnormalities in ALS patients compared to control patients, suggesting a more generalized neurodegenerative process.66 This may also include autonomic fibers, demonstrated by abnormalities in sympathetic skin responses in ALS patients44 and pathologic cell loss in the intermediolateral cell column.198 These studies raise the question of more widespread pathologic involvement in some patients with ALS, although the abnormalities may

be subclinical, and the possibility that the underlying pathophysiology is one of multiple system neurodegeneration within the CNS. Peripheral Motor Neuron Disease: MMN-CB Pure motor neuropathy, presenting with muscle atrophy, weakness, fasciculations, and reduced or absent deep tendon reflexes, is an important diagnosis to differentiate clinically from an inherited or sporadic segmental spinal muscular atrophy or early lower motor neuron variants of ALS or progressive muscular atrophy.46 Without sensory abnormalities, however, it can be difficult to determine if the pathology resides at the level of the lower motor neuron, the ventral root, or the peripheral motor nerve.205 MMN-CB is one type of peripheral motor neuropathy that is immune mediated and may respond to certain immunemodulating therapies.13,102,148 One study described clinical evaluations of 74 patients who presented with asymmetrical weakness, without sensory, upper motor neuron, or bulbar symptoms or findings. Twenty-five of these patients had multifocal conduction block on nerve conduction studies; the remaining patients had primarily diffuse axonal changes, with only a small number with slowed conduction velocities only.148 Of the patients with conduction block, those with distal more than proximal muscle weakness were more likely to have high serum titers of anti-GM1 antibodies, compared to those patients with proximal-predominant weakness, who were more likely to have serum antibodies to nonsialated carbohydrate epitopes of the neuronal glycolipids.148 Patients with MMN-CB are typically adult males with asymmetrical distal upper extremity weakness with minimal fasciculations clinically.46 Electrodiagnostic studies in patients with MMN-CB typically demonstrate proximal conduction block of compound muscle action potentials with associated slowing of the segmental conduction velocity across the region of block, and sparing of the sensory nerve action potential along this segment of the nerve.19,87,102,144,187 Conduction block should be demonstrated outside of sites of common compression if possible.19,87 Prolonged distal motor and F-wave latencies may also be present.46,102,191 One clinical clue to the possible presence of conduction block is weakness in a muscle with normal bulk.102 Electrodiagnostic abnormalities are more often found in nerves innervating weak muscles, but in one study, a third of the abnormal findings involved nerves innervating clinically normal muscles.191 Electromyography may demonstrate minimal fibrillation potentials but high-amplitude, polyphasic motor unit potentials with reduced recruitment in affected muscles.19,144 Compared to more centralized lower motor neuron lesions, this disorder is a peripheral nerve disease restricted to the peripheral motor axons and their myelin in particular. In evaluation of patients with lower motor neuron–type weakness, those with electrophysiologic abnormalities consistent

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

with motor axonal changes are more difficult to distinguish from those with ALS as pure motor axonal neuropathies, because both diseases may appear very similar in terms of the lower motor neuron involvement. In such patients, especially those without bulbar or signs of upper motor neuron involvement clinically, motor nerve fascicular biopsy or a therapeutic trial of immune-modulating therapy may be indicated to provide improved diagnostic accuracy.102 Early in ALS, motor nerve conduction studies may be relatively normal, although commonly there are abnormalities on needle examination in clinically normal muscles.19 Kornberg and Pestronk102 suggested that anti-GM1 antibody titers (IgM) greater than 1:6000 have the greatest specificity for MMN-CB, although such titers only occur in a third or less of patients with this disease.46 Anti-GM1 antibody titers (IgG) greater than 1:1000 are more likely seen in chronic, asymmetrical distal axonal motor neuropathies.102 Within the diagnostic possibilities in patients with MMN-CB is chronic inflammatory demyelinating polyradiculoneuropathy, which may also demonstrate conduction block and slowed conduction velocities; however, abnormal sensory nerve conduction studies, elevated CSF protein, and typically symmetrical motor weakness should aid in distinguishing these diseases in any given patient.102,144 Pathophysiologically, conduction block is believed to be due to focal demyelination or blockage of sodium channels in the axonal membrane. Persistent conduction block, as may occur in MMN-CB, has been postulated to be due to inhibition of remyelination of the affected nerve segment or of redistribution of sodium channels, thereby preventing restoration of the conduction of action potentials across the involved segment.86 Theoretically this is in part due to the presence of antiglycolipid antibodies, such as anti-GM1. There is also some evidence based on MRI evaluation of affected peripheral nerves that shows focal nerve enlargement and contrast dye enhancement at the site of the conduction block, implying a disruption of the bloodnerve barrier at this site, which may allow for exposure of the nerve to these serum antibodies.86 Neuropathologic examination of peripheral nerve at sites of conduction block demonstrated demyelination, poor remyelination, and onion bulb formation in the study by Kaji et al.,87 which examined a single nerve biopsy done distal to an area of focal enlargement from a single patient with MMN-CB. In contrast, a more recent study by Taylor et al.187 of fascicular biopsies done at the site of intraoperatively determined conduction block in seven patients with chronic disease showed multifocal loss of myelinated fibers, degenerating axons, altered size distribution, and rarely small perivascular lymphocytic infiltrates without evidence of significant demyelination or onion bulb formation. These findings imply that there is some degree of axonal involvement, possibly resulting from an immunemediated attack at the axolemma at the node of Ranvier,

1307

especially in areas of persistent conduction block.187 The fact that sensory nerve fibers are relatively spared from conduction block in these patients may be related to differences in ceramide composition between sensory and motor nerve fibers.138

SENSORY SYNDROMES Sensory loss, paresthesias, or pain in the hands and feet, along with symptoms of ataxia, are all suggestive for peripheral neuropathy but can also be seen in posterior column dysfunction associated with cobalamin deficiency and Sjögren’s syndrome.

Cobalamin Deficiency The most common neurologic symptoms are tingling and numbness.58,71 Other symptoms include generalized weakness, stiffness, deadness, tightness, feelings of cold or heat, formication, and shooting pains. Symptoms generally are symmetrical and begin in the lower extremities and progress proximally in all four limbs. Less common symptoms include vague aches and pains, girdle sensations, calf cramping, bowel or bladder disturbances, orthostatic hypotension, and changes in smell, vision, or taste or a roaring sensation in the ears. Depression, psychosis, and dementia can also occur.16,71,128,196 Symptoms progress to development of unsteady gait and awkwardness of limbs, leading to ataxic paraplegia with spasticity and contracture. Severe pseudoathetosis of the limbs has also been reported as a presenting symptom.17 Examination may reveal distal symmetrical sensory loss and absent tendon reflexes, consistent with peripheral dysfunction, or may reveal spasticity, extensor plantar responses, hyperreflexia, and segmental sensory loss, consistent with central dysfunction or myelopathy.71,82,196 Peripheral neuropathy and myelopathy can occur alone or concomitantly. Isolated hand involvement is suggestive of myelopathy.71,166,196 Examination can be normal or suggestive of small fiber dysfunction.71,166 Centrocecal scotomas have also been demonstrated in some patients.196 The peripheral neuropathy associated with cobalamin, or vitamin B12, is extensively reviewed in Chapter 89. Cobalamin deficiency is most commonly caused by pernicious anemia but can be caused by gastrointestinal malabsorption, nitrous oxide exposure, or inborn errors of metabolism.58,65,129,174,183 The lack of cobalamin leads to the accumulation of homocysteine and methylmalonic acid, with the development of megaloblastic anemia as a result of the decreased production of purines and pyrimidine and/or abnormal myelination of nerve fibers of the CNS and PNS. Neurologic symptoms can occur in the absence of any hematologic changes.26,27,82 The inability to absorb cobalamin leads to a subacute degeneration of the posterior

1308

Diseases of the Peripheral Nervous System

columns followed by the lateral columns of the spinal cord, hence the name “subacute combined degeneration.”2,82 It is estimated that 80% of patients with pernicious anemia have neurologic symptoms,196 and 8% of patients with polyneuropathy are estimated to have cobalamin deficiency.166 Subclinical cobalamin deficiency is thought to be 10 times more common than clinically evident cobalamin deficiency.27 Diagnosis of cobalamin deficiency as the cause of polyneuropathy can be difficult because, despite normal cobalamin levels, patients can still present with peripheral neuropathy.128 In one study, 12 of 27 patients (44%) with polyneuropathy had normal cobalamin but elevated serum metabolic levels.166 Asymptomatic patients have been found to have electrophysiologic abnormalities; a small proportion of these patients eventually will develop overt clinical manifestations.27 Nerve conduction studies have been reported to be mixed (axonal and demyelinating)71 or primarily axonal.128,166 Somatosensory evoked potentials can be abnormal, consistent with lesions in the spinal cord.166 In peripheral nerves, myelin loss without axonal loss has been described.196 Sural nerve biopsies have revealed axonal degeneration, with reduction of both large and small myelinated as well as unmyelinated fibers. No intrinsic abnormalities of myelin or Schwann cells have been reported.128 Focal spinal cord lesions have been found in cobalamin deficiency, although normal cervical and thoracic MRIs have also been described.12,58,116,157 Hyperintense signal seen in the posterior columns of the spinal cord in T2weighted MRI sequences are thought to be secondary to demyelination, wallerian degeneration, and gliosis. If resolution of these hyperintense signals occurs after treatment, demyelination is thought to be the underlying process. Abnormalities have also been seen with normal cobalamin levels.116 The significance of enhancement with gadolinium is unclear but possibly related to differences in vascularity or an increase in the permeability of the blood–spinal cord barrier.12 Mild spinal cord swellings on fluid-attenuated inversion recovery MRI sequences have also been described.157 Atrophy is a late change. Spinal cord pathology has revealed spongy degeneration primarily in the dorsal and lateral columns of the cervical and thoracic spinal cord. Swelling and ballooning of myelin sheaths, formation of intramyelinic vacuoles, and separation of myelin lamellae have been reported with degenerative changes of axis cylinder and myelin sheaths. Phagocytosis by macrophages, reactive astrocytosis, and fibrosis are also reported. Spinothalamic tracts and corticospinal tracts may be involved.12,65,196 Animal models do not replicate human disease, but lesions indistinguishable from subacute combined degeneration have been noted in the absence of hematologic changes in the monkey, fruit bat, and pig.26,65,129,174 Nutritional cobalamin deprivation of monkeys leads to spastic paralysis of hind limbs with separation and vacuolation

of myelin lamellae, with complete destruction of myelin sheaths in 33 to 45 months. Repeated exposure to nitrous oxide leads to neuropathy by 9 to 12 weeks and ataxia by 15 weeks. Spinal cord pathology reveals degeneration of myelin sheaths and axis cylinders in posterior columns with accumulation of lipid-laden macrophages. Nine months of cobalamin deprivation in fruit bats leads to the inability to fly, ataxia, spastic paresis, and death. With additional exposure to nitrous oxide, neuropathy occurs within 9 weeks. In the pig, repeated exposure to nitrous oxide leads to neuropathy in 9 weeks. Studies also suggest that a defect in the methylcobalamin-dependent methionine synthetase reaction is important in neurologic complications because administration of methionine can delay the onset of neuropathy in experimental cobalamin deficiency. Neuropathy has not been induced in the rabbit, mouse, or rat.

Sjögren’s Syndrome Inflammatory sensory polyganglionopathy67,105,106,121,131,177 (also reviewed in Chapter 103), such as may occur with Sjögren’s syndrome, can be mistaken for sensory axonal polyneuropathy when the initial symptoms are distal sensory loss.105 Thorough neurologic examination, however, should reveal asymmetrical or multifocal sensory loss, with or without loss of sensation over the trunk or face, which would suggest proximal sensory impairment, as with a sensory ganglionopathy.105,106 Symptoms such as light-headedness and Raynaud’s phenomenon can also be elicited. Abnormal pupillary responses, pseudoathetosis, ataxic hands and gait, loss of vibration and joint position sense, and positive Romberg’s sign may also be demonstrated on examination. Early ataxia is thought to be due to loss of afferent signals from proximal muscle spindles and joints.106 Kinesthesia and proprioceptive sensation seem to be relatively more affected than pain and temperature sensations.67 Reflexes are absent or diminished but strength is usually normal.106 A small percentage of patients have been reported to present with pain, paresthesias, and clumsiness of a single hand,106 which would seem to represent a mononeuropathy or radiculopathy clinically. Both the peripheral and central afferent axons may be damaged as a result of primary degeneration of dorsal root ganglion neurons,105 which is reflected by absent or reduced sensory responses with preserved sensory nerve conduction velocities in a non–length-dependent pattern of peripheral axonal degeneration on electrophysiologic testing,106 and by absent or abnormal somatosensory evoked potentials, reflecting pathology of the centrally directed afferent processes.105 Griffin et al.67 reported one patient who initially had normal sensory nerve action potentials 3 days after the abrupt onset of ataxia, but 6 weeks later with repeat examination all of his sensory responses were unobtainable.

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

A percentage of patients with sensory ganglionopathy caused by Sjögren’s syndrome have autonomic dysfunction.67,105,106 Cardiovascular autonomic failure, as well as abnormalities in heart rate and Valsalva maneuver, have been demonstrated with laboratory testing.67 Neurologic symptoms in Sjögren’s syndrome are typically thought to be due to sensory and autonomic neuronopathies. There is a range in the severity of large fiber myelinated fiber loss seen on nerve biopsies, although a relative preservation of unmyelinated fibers is seen. Collections of perivascular inflammatory infiltrates can be present.67,177 Involvement of short axons in the face and trunk, along with generally poor recovery, were all considered to be consistent with involvement of the sensory ganglia.177 Lymphocytic infiltration (T cell) of the dorsal root ganglia, with neuronal degeneration and loss of fibers in the dorsal roots, has previously been found on pathologic examination67,121 (Fig. 53–2). These findings are consistent with a

1309

length-independent sensory neuropathy with pathologic involvement of the primary sensory neuron. Satake and co-workers170 noted that incubation of the serum and CSF from a patient with Sjögren’s syndrome and sensory ganglionopathy with rat dorsal root ganglia revealed positive immunoreactivity in small neuronal cell bodies, which seemed to be specific because peripheral nerve and cerebellum did not react. These findings are suggestive of an autoimmune ganglionitis in Sjögren’s syndrome with an “anti–dorsal root ganglion neuron antibody.”170 Dorsal column abnormalities seen on cervical spine MRIs correlated with sensory symptoms in 14 patients with primary Sjögren’s syndrome and sensory neuronopathy.131 Twelve presented with paresthesias of the hands (nine patients) and feet (three patients), initially unilateral, before the diagnosis of Sjögren’s syndrome was made. Two presented with numbness in the face. Sensory nerve conduction studies revealed axonal loss, and abnormal median

A

B

C

D

FIGURE 53–2 Thoracic dorsal root ganglia in Sjögren’s syndrome. A, Loss of neurons and clusters of mononuclear cells. (Cryostat section stained with hematoxylin-eosin; ⫻330 before 2% reduction.) B, Subsequent section from the same biopsy as A, immunostained for T cells and photographed in dark field; immunostained cells appear white, often with a dark spot in the center representing the cell nucleus (examples identified by arrows). Two neurons are labeled N1 and N2. Note the cluster of T cells and scattered individual T cells (arrowheads). Note also the degenerating neuronal cell body on the left (N1). (⫻330 before 2% reduction.) C, Numerous clusters of mononuclear cells and reduced number of neurons. (Paraffin section stained with hematoxylin-eosin; ⫻110 before 2% reduction.) D, Neuron (N) surrounded by intense mononuclear infiltrates. (Paraffin section stained with hematoxylin-eosin; ⫻210 before 2% reduction.) (From Griffin, J. W., Cornblath, D. R., Alexander, E., et al.: Ataxic sensory neuropathy and dorsal root ganglionitis associated with Sjogren’s syndrome. Ann. Neurol. 27:304, 1990, with permission.)

1310

Diseases of the Peripheral Nervous System

somatosensory potentials suggested central pathology. High-intensity areas in T2-weighted images seen in the posterior columns of cervical MRIs, which represented degeneration of central sensory projections secondary to dorsal root ganglion damage, have been reported to occur in these patients105,106,131 (Fig. 53–3). Involvement of both cuneatus and gracilis fasciculi was seen in patients with extensive sensory deficits in the face, trunk, and upper and lower extremities. Sensory ataxia was severe, and a high frequency of autonomic abnormalities was present: pupillary dysfunction, vomiting, orthostatic hypotension, and

decreased uptake on 123I-meta-iodobenzylguanidine cardiac accumulation. High signal seen only in the gracilis fasciculus correlated with only partial sensory deficits in upper and lower limbs and mild to moderate sensory ataxia. Mild sensory deficits only were present in patients who did not have abnormal signal on their MRIs.

COMBINED MOTOR AND SENSORY SYNDROMES Cervical spondylotic myelopathy and acquired immunodeficiency syndrome (AIDS)–related vacuolar myelopathy can affect both motor and sensory pathways; however, the symptoms can be mistaken for peripheral neuropathy, and the findings of upper motor neuron dysfunction can be masked by superimposed disease such as peripheral neuropathy. Syphilis, “the great imitator,” can present with purely motor, sensory, or mixed features. Bulbospinal muscular atrophy, although primarily considered a lower motor neuron syndrome, is characterized by significant pathologic involvement of the sensory nerves.

Cervical Spondylotic Myelopathy

FIGURE 53–3 Axial T2-weighted gradient echo magnetic resonance imaging scans (repetition time, 500 ms; echo time, 9 ms; flip angle, 20 degrees) of cervical spinal cord (C4 level) in Sjögren’s syndrome. A, Normal appearance of posterior column white matter and anterior and posterior horns. B, Arrowhead shows diffuse highintensity signal in the posterior column in a patient with idiopathic sensory ganglionopathy. (From Lauria, G., Pareyson, D., Grisoli, M., and Sghirlanzoni, A.: Clinical and magnetic resonance imaging findings in chronic sensory ganglionopathies. Ann. Neurol. 47:104, 2000, with permission.)

Cervical spondylosis is often present but asymptomatic by the age of 60. If nerve roots or spinal cord are affected, symptoms of radiculopathy, cervical myelopathy, or a combination of the two may develop.21 Cervical spondylotic myelopathy refers to myelopathy associated with degenerative spine disease (e.g., cervical spondylosis), and the term often includes ossification of the posterior longitudinal ligament (OPLL). The progression of spondylosis to myelopathy is thought be uncommon and occurs in a slow and stepwise fashion.108,171 OPLL is rare among whites but is a significant cause of myelopathy in older Japanese adults. The onset of OPLL is usually insidious, beginning in the fifth decade. Patients are asymptomatic or develop symptoms similar to cervical spondylotic myelopathy.88 One of the earliest symptoms reported is difficulty walking in the dark, which occurs because compensatory visual cues are lost. Symptoms of paresthesia, dysesthesia, and sometimes pain in the toes, which can be mistaken for peripheral neuropathy, are predominant.50 Symptoms can evolve slowly with ascending sensory loss, followed by the development of a spastic paraparesis. The classic picture of cervical spondylotic myelopathy is that of a spastic tetraparesis, walking disturbance, loss of sensation, tingling in the hands, and difficulty performing fine manual tasks.82,139 Atrophy of intrinsic hand muscles may be present. Bladder, bowel, and sexual dysfunctions are late findings. Neck pain is usually not prominent unless superimposed radiculopathy is present. Other vague symptoms include headache, dizziness, and disturbances of equilibrium. Lhermitte’s sign has also been reported.50

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

Patients with atrophic upper extremity muscles and spastic and weak lower limbs have been mistakenly diagnosed with motor neuron disease.112 The increased reflexes and spasticity, however, can be masked by secondary disorders such as peripheral neuropathy, radiculopathy, and anterior horn cell disease. The degree of deep tendon reflex abnormalities depends on degree of anterior horn cell involvement. Other disorders that can be mistaken for cervical spondylotic myelopathy include multiple sclerosis (MS), ALS, polyradiculitis (GBS), Lyme disease (borreliosis), and syringomyelia.38,50 It has been suggested that up to 17% of cases attributed to cervical spondylotic myelopathy actually had another disease process such as ALS or MS.38 Spondylotic radiculopathy may be confused clinically with ALS because both may have prominent lower motor neuron signs, especially early in the course of ALS, before upper motor neuron or bulbar signs develop. Rarely, spondylotic radiculopathy may present with weakness, fasciculations, reduced reflexes, and absence of sensory changes, with improvement following decompressive cervical spine surgery.49 A subset of patients is thought to have a slightly different syndrome called “amyotrophic type of myelopathy hand.”52 They present with insidious wasting and weakness of intrinsic hand muscles, which can be unilateral or asymmetrical (Fig. 53–4). There is mild or no sensory disturbance, which is disproportionate to motor dysfunction. Posterior cervical, shoulder, or radicular pain is generally not reported by patients. Deep tendon reflexes may be normal or increased in the lower extremities, with down-going toes with plantar stimulation. There is typically no associated gait disturbance.52 These patients can be misdiagnosed with median

FIGURE 53–4 Typical amyotrophic type of myelopathy hand in cervical spondylotic myelopathy (CSM). (From Ebara, S., Yonenobu, K., Fujiwara, K., et al.: Myelopathy hand characterized by muscle wasting: a different type of myelopathy hand in patients with cervical spondylosis. Spine 13:785, 1988, with permission.)

1311

and ulnar entrapment neuropathies, ALS, or polyneuropathy. They generally have multisegmental spondylosis involving the C5-C6 and C6-C7 spinal segments, with narrowing of the anteroposterior diameter of the lower cervical spine to less than 13 mm, with improvement in their symptoms following cervical spine surgery.52 These patients can be differentiated from patients with ALS by absence of bulbar signs, no involvement of clinically normal muscles on electromyography, and correlation of spinal structural changes with their arm weakness.52 Cervical spondylosis may also simulate lower motor neuron disease or progressive spinal muscular atrophy with painless, progressive muscle weakness and fasciculations, normal to slightly brisk deep tendon reflexes, absence of sensory findings, and down-going toes.115 Dissociated motor and sensory symptoms in patients with cervical spondylosis may occur with selective compression of the ventral more than the dorsal nerve roots within the spinal canal.95 Chronic cervical myelopathy with age-related degenerative changes has previously been reviewed.21,38,50,112,171 By the fourth decade, 30% of asymptomatic individuals can have degenerative changes in the intervertebral discs. By the seventh decade, 90% have degenerative changes. The spinal cord diameter ranges between 8.5 and 11.5 mm, and the cervical spine canal ranges in diameter from 16 to 18 mm. The mechanism of cervical myelopathy is thought to be secondary to a slow progressive compression with minor but repetitive trauma and can be multifactorial in origin: disc degeneration with annulus bulging leads to osteophyte formation, which can lead to spinal cord and nerve root compression; and spondylotic transverse bars, hypertrophic facets, and hypertrophic ligamentum flavum can lead to spinal canal diameter narrowing (with and without movement), and differences in cord sensitivity to hypoxia. OPLL originates in the rostral cervical canal and extends caudally.50,171 A spinal canal anteroposterior diameter of less than 12 mm is associated with myelopathy, particularly in patients with congenital canal stenosis.139 Patients are symptomatic when the cross-sectional area of the cord is reduced by one third. Alteration of blood supply to the spinal cord may also be a factor. Intramedullary vessels may be more vulnerable than the larger extramedullary vessels to intermittent ischemia as a result of distortion and compression of the anterior spinal artery.82,115 Ischemic compression of the anterior spinal artery at upper and midcervical spinal levels with resulting ischemic changes in the lower cervical cord has been found in patients with midcervical spine subluxations associated with rheumatoid arthritis.127 Acute trauma exacerbates any of the above-mentioned conditions.21,139 Long-standing cord compression resulting from spondylosis has been found to result in wallerian degeneration of the posterior columns and lateral corticospinal tracts.139 Plain radiographs are helpful in revealing cervical stenosis. T2-weighted images reveal spinal cord abnormalities on MRI171,197 (Fig. 53–5). Electromyography may reveal

A

B C D E

B

D

C

E

FIGURE 53–5 T2-weighted magnetic resonance imaging in cervical spondylotic myelopathy. A, Sagittal image showing cord compression at the midvertebral body level as well as at the vertebral end plate level from C4 to C6. B–E, Transverse or axial images showing an isointense-signal hypertrophy of the posterior longitudinal ligament (wide arrow) as well as a low- to isointense-signal herniated intervertebral disc (small arrow) and osteophyte (large arrow) compressing the spinal cord, with a high-intensity signal in the anterior horns with left-side predominance (asterisk). (From Mizuno, J., Nakagawa, H., and Hashizume, Y.: Cervical amyotrophy caused by hypertrophy of the posterior longitudinal ligament. Spinal Cord 40:484, 2002, with permission.)

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

denervation and confirm the presence of radiculopathy but does not help in prognosis. Somatosensory evoked potentials can help differentiation between cervical myelopathy and anterior horn cell disease.50 In the early stages of myelopathy, changes of demyelination, rather than axonal damage and gray matter infarction, are seen.171 Spinal cords of patients with cervical myelopathy may be flattened and distorted with demyelination in the lateral columns and corticospinal tracts. In more severe cases, myelin destruction, axonal loss in white matter, and mild chromatolysis and neuronal loss in gray matter occur.203 Gliosis has been reported in areas of injury with damage in the posterior root entry zones.171 Sparing of anterior columns can occur because the proximal branches of the anterior spinal artery are shorter and subjected to less obliterative stress. Distal ramifications of the anterior spinal artery are thought to be most affected, and ischemia occurs as a result of the slow metabolic exchange.120 The underlying pathology of the amyotrophic type of myelopathy hand is attributed to segmental necrosis of the anterior horn cells resulting from the reduced transectional area of C7, C8, or T1.50,52,119 Autopsy cases revealed two patterns in OPLL: boomerang and triangular.88 The boomerang pattern (convex lateral surfaces and concave anterior surfaces) was restricted to gray matter, with relative preservation of white matter even with severe compression. In the triangular pattern (angular lateral surfaces and flat anterior surface), pathologic changes were seen in both white and gray matter but sparing the anterior columns. Pathologic changes were seen over more than one segment and in both descending degeneration of the lateral pyramidal tracts and ascending degeneration of the posterior columns, including the fasciculus gracilis. A triangleshaped spinal cord with transverse area of less than 60% of normal in more than one segment appears to be associated with severe and irreversible pathologic changes. The mechanisms of chronic mechanical compression have been described in an animal model of chronic cervical cord compression.203 The tiptoe-walking Yoshimura mouse develops spontaneous calcified deposits posterior to the C1-C2 vertebrae leading to compression and paresis by 4 to 8 months of age. Evaluation of 6-month-old mice revealed reduction of neurons in the gray matter, and degeneration of anterior, lateral, and posterior columns was observed; this was most severe at the level of compression. Apoptosis of both neurons and oligodendrocytes was identified using terminal deoxynucleotidyl transferase–mediated dUTP nick end-labeling (TUNEL) staining and immunostaining with caspase-3. Similar changes were identified in postmortem cervical cord from a patient with cervical myelopathy caused by OPLL. Apoptosis has been recognized to play an important role in the morbidity of cervical spondylotic myelopathy, and studies are underway to understand the molecular pathways governing apoptosis.99

1313

Diseases of the lumbar spine may also present with symptoms and signs of myelopathy. Kleopa et al.101 recently described four patients with progressive asymmetrical leg weakness with fasciculations, muscle cramps, and variable deep tendon reflexes, with only one of the four showing extensor plantar responses. Later in the course, these patients reported distal sensory symptoms of asymmetrical burning and tingling paresthesias in their feet. MRI of the spine revealed stenosis of the lower thoracic spine with compressive lumbar myelopathy with symptoms and findings restricted to the anterior horn cells, presumably resulting from spinal artery compression and anterior cord ischemia.101 There have also been rare reports of progressive lower motor neuron syndromes developing in the lower extremities years after an episode of lumbosacral transverse myelitis.134

AIDS-Related Vacuolar Myelopathy The prevalence of peripheral neuropathies in HIV infection (reviewed in Chapter 94) is estimated to be as high as 36%, despite the introduction of highly active antiretroviral therapy (HAART). In the pre-HAART era, autopsy studies estimated the prevalence of AIDS-associated vacuolar myelopathy to be between 22% and 55%.10,42,151 Myelopathic signs were not obvious in patients with pathologically mild disease or in patients with concurrent peripheral neuropathy, so the diagnosis in early reports was made post mortem. An early symptom may be erectile dysfunction, followed by a slowly progressive spastic paraparesis with loss of vibration and joint position sensations with urinary frequency and incontinence.45 AIDS-related vacuolar myelopathy has been reported to improve on a HAART regimen, especially when diagnosed early in the disease course.181 AIDS-related vacuolar myelopathy was first identified in an autopsy series in which 20 of 89 consecutive patients were found to have spongiform degeneration,151 usually of the middle and lower thoracic regions, with patchy or confluent vacuoles and intravacuolar macrophages. Cervical spinal cords were not evaluated in this initial study. These changes were later demonstrated in cervical regions but less frequently in the lumbar spine.10 HIV type 1 was isolated from macrophages in affected patients.10,151 Clinical symptoms and signs of myelopathy were noted in patients with marked pathologic changes, whereas patients with mild changes often did not have upper motor neuron signs or weakness. Myelopathy was graded in severity on a scale from I to III: grade I had infrequent myelopathic signs, grade II had symptoms similar to grade III but less frequently, and grade III was found to be associated with a steadily progressive spastic-ataxic paraparesis with urinary incontinence. Other diseases and coexisting infections that potentially could confuse the clinical picture include syphilis, low folic acid, megaloblasts in bone marrow,

1314

Diseases of the Peripheral Nervous System

Karposi’s sarcoma, and infection with Pneumocystis carinii, Candida albicans, Mycobacterium avium-intracellulare, herpes simplex virus, cytomegalovirus, Giardia lamblia, Cryptosporidium, Salmonella, and Shigella. The diagnosis of myelopathy in HIV is difficult because of the high prevalence of both peripheral neuropathy and dementia in the late stages of HIV disease.42,151 In a retrospective case-control study of AIDS patients,42 100 of 215 autopsies (47%) in which the spinal cords were evaluated were found to have vacuolar myelopathy. Cases were defined as having autopsy evidence of vacuolar myelopathy, whereas controls did not. Myelopathy was defined as upper motor neuron signs (spasticity, hyperreflexia, pyramidal weakness, and upgoing toes) in the lower extremities out of proportion to findings in the upper extremities, and sensory neuropathy was defined as distal sensory symptoms (numbness, tingling, or pain) with distal loss of pain, temperature, or vibration sensation with reduced or absent ankle reflexes. Fifty-six of 100 patients with vacuolar myelopathy had been examined 3 months prior to death, and only 15 (27%) were found to have symptoms and signs of myelopathy, most commonly spastic paraparesis with a coexistent distal peripheral neuropathy. The most severe symptoms were spasticity and sensory ataxia because severe weakness and incontinence were late findings. Only two patients were noted to have a sensory level and only one had bowel and bladder incontinence. Symptoms were present from 3 to 16 weeks before diagnosis. Only four patients had thoracic MRIs, and all were normal. Patients with vacuolar myelopathy also had a higher frequency of distal sensory neuropathy (14 of 56) compared to controls (3 of 48). Patients with both myelopathy and neuropathy were more symptomatic because of combination of the two diseases. Only two patients developed symptomatic myelopathy before developing an AIDSdefining illness. Patients with vacuolar myelopathy were older and had M. avium-intracellulare and P. carinii pneumonia, likely markers of immunosuppression, because no direct evidence of either was found in spinal cords,42 suggesting that myelopathy occurs with more advanced HIV disease (i.e., after onset of AIDS). Correlation with CD4⫹ cell counts was not available. Based on this study, the diagnosis of myelopathy appeared to confer a poorer prognosis, because 13 of the 15 cases had survival times of less than 9 months. As discussed earlier, clinical diagnosis of AIDS-related vacuolar myelopathy is difficult. MRI findings are nonspecific: atrophy and diffusely increased intrinsic cord signal, which do not correlate with clinical deficits.34 A study of 69 patients, 35 with clinical evidence for myelopathy (leg stiffness, heaviness, cramps, numbness, subjective bladder dysfunction, and objective lower extremity weakness, upper motor neuron findings, or urinary incontinence) and 34 without myelopathy, who had

somatosensory evoked potential studies was undertaken to determine if evoked potentials could be useful for early diagnosis.185 Eighty-three percent of patients were found to have peripheral neuropathy (lower extremity paresthesias and pain, muscle cramps, and weakness, with distal muscle atrophy, sensory loss, or reflex loss), and 33% had both neuropathy and myelopathy. Concomitant disorders such as syphilis, toxoplasmosis, human T-cell lymphoma, and cobalamin deficiency had been excluded in these patients. Abnormalities of tibial, and not median, central conduction times were shown to correlate with the clinical diagnosis of myelopathy. Similar evoked potential findings had previously been reported in 23 patients with AIDS.72 There was no significant correlation with CD4 counts. The lack of abnormalities in the median central conduction times suggested a predilection for the thoracolumbar spinal cord in these patients. The derived spinal conduction time was found to be a more sensitive marker of spinal cord dysfunction in patients with normal tibial central conduction times and who also did not have objective signs of myelopathy. Myelopathy was defined pathologically as vacuolation in spinal white matter with lipid-laden macrophages.151 Vacuoles were most extensive in the middle to lower thoracic levels, more in the lateral than the posterior columns. The findings were often symmetrical in mild cases but asymmetrical in moderate to severe cases. Thin myelin sheaths surrounded the vacuoles, occasionally in continuity with the sheaths of a normally myelinated axon. Axons were disrupted only in areas of severe vacuolation. Wallerian degeneration was seen in some of the lateral corticospinal tracts. Reactive astrocytosis was rare, and there was no inflammation, organisms, or intranuclear viral inclusions. Few patients had central chromatolysis of anterior horn motor cells. The pathology of vacuolar myelopathy is not dying-back or wallerian-type degeneration but more similar to subacute combined degeneration caused by vitamin B12 or folate deficiency. Spinal roots and ganglia were noted to be unremarkable.10,63,151 A small number of cases have stained positive for anti–p24 HIV core protein.10 An association between vacuolar myelopathy and HIV encephalitis152 suggests that the presence of encephalitis may have an indirect effect on the spinal cord. A correlation between HIV DNA in the spinal cord and vacuolar myelopathy has not been demonstrated.

Neurosyphilis Neurosyphilis is a spectrum of disorders that can be divided into early and late stages.54,77,141,175 Early neurosyphilis includes asymptomatic neurosyphilis and meningovascular syphilis. Asymptomatic neurosyphilis refers to CSF pleocytosis or positive serology in the absence of symptoms or signs. Meningovascular syphilis refers to progressive vascular insufficiency leading to

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease

infarction years after infection. Late or parenchymatous neurosyphilis includes general paresis; the progressive, chronic and dementing process; and tabes dorsalis. Late neurosyphilis occurs after decades of infection. Historically, 4% to 9% of patients with untreated syphilis developed symptomatic neurosyphilis.77 Neurosyphilis mimicking peripheral nerve disorders can be better understood by reviewing spinal cord pathology. Pure spinal involvement in syphilis is rare. In a series of 2231 syphilitic cases recorded at Boston City Hospital,1 only 31 cases (i.e., ⬍1% of all cases) were reported to have pure spinal involvement. A historical classification of spinal syphilis based on spinal cord pathology1 is discussed later: meningomyelitis, syphilitic spinal thrombosis, syphilitic amyotrophy, and syphilitic spinal pachymeningitis (tabes dorsalis). Mental status changes and cranial nerve involvement can occur during any stage of syphilitic infection. Recent use of electrophysiology and neuroimaging in the diagnosis of neurosyphilis is also discussed. The onset of syphilitic meningomyelitis is insidious.1,77,141 The presenting symptoms are weakness or paresthesias, with sphincter disturbances occurring less frequently. Sensory symptoms include numbness, coldness, and tingling. Vibration and joint position senses are impaired. Tone and reflexes are increased. In a case report of subacute meningomyelitis,98 the cervical cord was shown to be swollen, with diffuse high signal on T2-weighted images with gadolinium enhancement on MRI. The enhancing parenchymal lesions, which were relatively reduced on T1-weighted images, were described as having a “candle guttering appearance with flip-flop sign.” The abnormal enhancement in the superficial regions of the spinal cord under the pia mater suggested that neurosyphilis invaded the spinal cord from its surface. Pathologic changes of meningomyelitis are the result of infiltration of leptomeninges by lymphocytes and plasma cells,1 leading potentially to occlusive arteritis with secondary spinal cord ischemia or infarction.82 Persistence of inflammation can lead to damage of ventral roots. Polyradiculoneuropathy can also be a rare manifestation of syphilitic meningomyelitis, presenting with hypotonic lower extremity weakness with areflexia, without24,43 and with associated HIV infection.104 Good response to penicillin treatment was reported in all cases. Abnormal enhancement of the nerve roots and nodular enhancement of the cauda equina were demonstrated with MRI.24 Syphilitic spinal thrombosis1 can present as transverse myelitis or anterior spinal artery syndrome. Spinal arteries are suddenly thrombosed as a result of arteritis, leading to sudden flaccid paraplegia, caused by spinal shock, with urinary retention and anesthesia. Later, atrophic paralysis occurs as a result of anterior horn cell destruction with distal spastic paralysis caused by corticospinal tract destruction. Hyperplastic endarteritis, referred to as Heubner’s

1315

arteritis (with infiltration of vessel wall adventitia and media by lymphocytes and plasma cells and narrowing of vessels by subintimal fibroblasts), and myelomalacia have been described pathologically. Syphilitic amyotrophy in isolation is controversial but has been described in combination with syphilitic meningomyelitis, spinal vascular neurosyphilis, or tabetic neurosyphilis.1 It is characterized by painless and subacute weakness with lower motor neuron signs, often with a lack of progression and response to antibiotic treatment.55 Therefore, it may mimic motor neuron disease such as progressive muscular atrophy or ALS.82 Persistent meningeal inflammation causing destruction of the anterior horn cells or ventral roots can lead to atrophy, areflexia, and fasciculations. Syphilitic arteritis can cause infarction of the ventral horn without long tract involvement, particularly in the cervical spinal cord. Five percent to 10% of patients with tabes dorsalis are estimated to have muscular weakness and atrophy. Shrinkage and hyperchromatism of anterior horn cells with astrocytosis throughout the ventral horns has been described.1 The classic symptoms of tabes dorsalis or syphilitic spinal pachymeningitis are lightning-type pains, ataxia, and urinary incontinence with absent reflexes and impaired vibration and position sense in the lower extremities.1,3 Patients are ataxic, pupils are abnormal (Argyll-Robertson), and nociceptive loss is severe, leading to trophic lesions, ulcers, and Charcot joints. Strength is often normal. CSF examination reveals absent to mild pleocytosis, mildly elevated protein, and a positive Venereal Disease Research Laboratory (VDRL) test, although in some cases the CSF findings are minimal. Persistent inflammation of the meninges is postulated to lead to inflammation and eventually fibrosis of the posterior roots with degeneration of posterior columns.82 Dorsal root ganglia and peripheral nerves are less affected.3,82 Gummas, or focal nodular lesions in or around dura mater, may also develop. The clinical picture, however, is that of a rapidly growing tumor with pain and paresthesias, spastic paraplegia, and incontinence of bowel and bladder. Microscopically, there is granuloma formation with or without necrosis with lymphocytes, plasma cells, fibroblasts, and histocytes.1 Electrophysiologic studies in tabes dorsalis can be normal except for absent tibial H reflexes. Median nerve somatosensory evoked potentials were found to be normal, whereas tibial nerve somatosensory evoked potentials were abnormal—findings consistent with a disorder of the posterior columns.47,190 In vitro sural nerve studies in a patient with impaired pain and temperature sensation revealed normal A␣, A␦, and C fibers with normal number and sizes of myelinated and unmyelinated nerve fibers per unit of fascicular area, findings consistent with preservation of the dorsal root ganglia and their peripheral axons.51,70,182

1316

Diseases of the Peripheral Nervous System

Diagnosis of neurosyphilis, caused by the spirochete Treponema pallidum, requires serologic testing77,175 using rapid plasma reagent (RPR) and VDRL tests. A positive RPR should be confirmed with the fluorescent treponemal antibody absorption (FTA-ABS) test or the microhemagglutination–T. pallidum (MHA-TP) assay. The FTA-ABS is very specific; thus a negative FTA-ABS essentially excludes syphilis as a diagnosis. However, the FTA-ABS may remain positive indefinitely and cannot be used as an indicator of active disease. Although the sensitivity of the CSF VDRL test is 30% to 70%, the specificity is very high. Treatment for neurosyphilis77,141 consists of penicillin G sodium, 12 to 24 million units per day intravenously (2 to 4 million units every 4 hours) for 10 to 14 days, or penicillin G procaine, 2 to 4 million units per day intramuscularly with probenecid 500 mg orally four times a day for 10 to 14 days. Tetracycline (500 mg orally four times a day for 30 days) can be given in case of penicillin allergy. Treatment will halt the progression of the disorder, and CSF pleocytosis and elevated protein will improve by 6 months. CSF serologic tests can be reactive for a year after therapy. CSF should normalize by 1 to 2 years.

Bulbospinal Muscular Atrophy (Kennedy’s Disease) Bulbospinal muscular atrophy is an X-linked recessive inherited lower motor neuron disease that presents with primarily motor symptoms, although there is subclinical sensory neuron involvement as well. Polo et al.154 examined sensory electrophysiologic abnormalities in patients with bulbospinal muscular atrophy and found abnormalities in central afferent pathways, represented by abnormal somatosensory evoked potentials and increased wave I latency on brainstem auditory evoked potentials. Peripheral sensory nerve action potentials were variably affected, suggesting predominant involvement of central afferent pathways in these patients versus peripheral sensory axonopathy.154 Studies using laser-evoked potentials, trigeminal blink reflex testing, masseter inhibitory reflex, and jaw jerk testing revealed findings consistent with impairment of large-diameter afferent fibers and central sensory neurons.9 Sural nerve biopsies and postmortem examinations of patients with bulbospinal muscular atrophy demonstrate loss of large myelinated fibers9,114 with secondary segmental demyelination and remyelination in the sural nerve and loss of myelinated fibers in the dorsal columns.114 There was a shift in size distribution of dorsal root ganglion neurons, with reduced numbers of largediameter cells and an increase in small-diameter neurons. There was evidence of androgen receptor mRNA in dorsal root ganglia, sural nerve, and spinal cord,114 which may explain why the afferent system is also involved in this disease, because Kennedy’s disease is caused by a genetic

defect in the androgen receptor gene. These pathologic findings would imply that, in bulbospinal muscular atrophy, the sensory involvement is one of neuronal loss with central and peripheral axonal degeneration.114

MENINGEAL DISEASE Allanore et al.6 reported a case of a patient with a history of breast cancer who presented with symptoms of sciatica that progressed to also include urinary incontinence and lower extremity weakness. Serial MRI studies of this patient’s lumbosacral spine showed a focal dural mass at L2 that was a leptomeningeal metastasis.6 Such a case illustrates the importance of considering meningeal disease even with initial peripheral clinical manifestations. Pain and weakness in this case was due to an intradural mass lesion, presumably compressing dorsal and ventral roots and eventually the spinal cord.

REFERENCES 1. Adams, R. D., and Merritt, H. H.: Meningeal and vascular syphilis of the spinal cord. Medicine 48:153, 1944. 2. Adams, R. D., and Salam-Adams, M.: Chronic nontraumatic diseases of the spinal cord. In Young, R. R., and Woolsey, R. M. (eds.): Diagnosis and Management of Disorders of the Spinal Cord. Philadelphia, W. B. Saunders, p. 84, 1995. 3. Adams, R. D., and Victor, M.: Principles of Neurology. New York, McGraw-Hill, 1993. 4. Agamanolis, D. P., Leslie, M. J., Caveny, E. A., et al.: Neuropathological findings in West Nile virus encephalitis: a case report. Ann. Neurol. 54:547, 2003. 5. Ahmed, S., Libman, R., Wesson, K., et al.: Guillain-Barre syndrome: an unusual presentation of West Nile virus infection. Neurology 55:144, 2000. 6. Allanore, Y., Hilliquin, P., Zuber, M., et al.: A leptomeningeal metastasis revealed by sciatica. Rev. Rhum. 66:232, 1999. 7. Almhanna, K., Palanichamy, N., Sharma, M., et al.: Unilateral brachial plexopathy associated with West Nile virus meningoencephalitis. Clin. Infect. Dis. 36:1629, 2003. 8. Alpert, S. G., Fergerson, J., and Noel, L. P.: Intrauterine West Nile virus: ocular and systemic findings. Am. J. Ophthalmol. 136:733, 2003. 9. Antonini, G., Gragnani, F., Romaniello, A., et al.: Sensory involvement in spinal-bulbar muscular atrophy (Kennedy’s disease). Muscle Nerve 23:252, 2000. 10. Artigas, J., Grosse, G., and Niedobitek, F.: Vacuolar myelopathy in AIDS: a morphological analysis. Pathol. Res. Pract. 186:228, 1990. 11. Asnis, D. S., Conetta, R., Teixeira, A. A., et al.: The West Nile Virus outbreak of 1999 in New York: the Flushing Hospital experience. Clin. Infect. Dis. 30:413, 2000. 12. Bassi, S. S., Bulundwe, K. K., Greeff, G. P., et al.: MRI of the spinal cord in myelopathy complicating vitamin B12 deficiency: two additional cases and a review of the literature. Neuroradiology 41:271, 1999.

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease 13. Bentes, C., de Carvalho, M., Evangelista, T., and Sales-Luis, M. L.: Multifocal motor neuropathy mimicking motor neuron disease: nine cases. J. Neurol. Sci. 169:76, 1999. 14. Bergh, S. S., Feracci, F., Maksimova, E., et al.: Autoimmunity to beta IV spectrin in paraneoplastic lower motor neuron syndrome. Proc. Natl. Acad. Sci. U.S.A. 98:6945, 2001. 15. Bir, L. S., Keskin, A., Yaren, A., et al.: Lower motor neuron disease associated with myelofibrosis. Clin. Neurol. Neurosurg. 102:109, 2000. 16. Blumenthal, D. T., and Glenn, M. J.: Neurologic manifestations of hematologic disorders. Neurol. Clin. 20:265, 2002. 17. Blunt, S. B., Silva, M., Kennard, C., and Wise, R.: Vitamin B12 deficiency presenting with severe pseudoathetosis of upper limbs. Lancet 343:550, 1994. 18. Bodian, D.: Histopathologic basis of clinical findings in poliomyelitis. Am. J. Med. 6:563, 1949. 19. Bouche, P., LeForestier, N., Maisonobe, T., et al.: Electrophysiological diagnosis of motor neuron disease and pure motor neuropathy. J. Neurol. 246:520, 1999. 20. Bowen, J., Gregory, R., Squier, M., and Donaghy, M.: The post-irradiation lower motor neuron syndrome: neuronopathy or radiculopathy? Brain 119:1429, 1996. 21. Braakman, R.: Management of cervical spondylotic myelopathy and radiculopathy. J. Neurol. Neurosurg. Psychiatry 57:257, 1994. 22. Brownell, B., Oppenheimer, D. R., and Hughes, J. T.: The central nervous system in motor neurone disease. J. Neurol. Neurosurg. Psychiatry 33:338, 1970. 23. Buchthal, F.: Problems of the pathologic physiology of poliomyelitis. Am. J. Med. 6:579, 1949. 24. Byrne, T. N., Bose, A., Sze, G., and Waxman, S. G.: Syphilitic meningitis causing paraparesis in an HIV-negative woman. J. Neurol. Sci. 103:48, 1991. 25. Campbell, G. L., Marfin, A. A., Lanciotti, R. S., and Gubler, D. J.: West Nile virus. Lancet Infect. Dis. 2:519, 2002. 26. Carmel, R., Green, R., Rosenblatt, D. S., and Watkins, D.: Update on cobalamin, folate, and homocysteine. Hematology (Am. Soc. Hematol. Educ. Progr.) 2003:62, 2003. 27. Carmel, R., Melnyk, S., and James, S. J.: Cobalamin deficiency with and without neurologic abnormalities: differences in homocysteine and methionine metabolism. Blood 101:3302, 2003. 28. Carpenter, M. B., and Sutin, J.: Human Neuroanatomy, 8th ed. Baltimore, Williams & Wilkins, 1983. 29. Centers for Disease Control and Prevention: Acute flaccid paralysis syndrome associated with West Nile virus infection— Mississippi and Louisiana, July–August 2002. MMWR Morb. Mortal. Wkly. Rep. 51:825, 2002. 30. Centers for Disease Control and Prevention: Investigations of West Nile virus infections in recipients of blood transfusions. MMWR Morb. Mortal. Wkly. Rep. 51:973, 2002. 31. Centers for Disease Control and Prevention: Laboratoryacquired West Nile virus infections—United States, 2002. MMWR Morb. Mortal. Wkly. Rep. 51:1133, 2002. 32. Centers for Disease Control and Prevention: Update: investigations of West Nile virus infections in recipients of organ transplantation and blood transfusion—Michigan, 2002. MMWR Morb. Mortal. Wkly. Rep. 51:879, 2002. 33. Centers for Disease Control and Prevention: West Nile virus activity—United States, October 10–16, 2002: an

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45. 46.

47.

48.

49.

50.

1317

update on West Nile virus infections in recipients of blood transfusions. MMWR Morb. Mortal. Wkly. Rep. 51:929, 2002. Chong, J., Di Rocco, A., Tagliati, M., et al.: MR findings in AIDS-associated myelopathy. AJNR Am. J. Neuroradiol. 20:1412, 1999. Chong, M. S., Reynolds, M. L., Irwin, N., et al.: GAP-43 expression in primary sensory neurons following central axotomy. J. Neurosci. 14: 4375, 1994. Chou, S. M., Wang, H. S., and Komai, K.: Colocalization of NOS and SOD1 in neurofilament accumulation within motor neurons of amyotrophic lateral sclerosis: an immunohistochemical study. J. Chem. Neuroanat. 10:249, 1996. Chu, J. J., and Ng, M. L.: The mechanism of cell death during West Nile virus infection is dependent on initial infectious dose. J. Gen. Virol. 84:3305, 2003. Clifton, A. G., Stevens, J. M., Whitear, P., and Kendall, B. E.: Identifiable causes for poor outcome in surgery for cervical spondylosis: post-operative computed myelography and MR imaging. Neuroradiology 32:450, 1990. Coggeshall, R. E., Lekan, H. A., Doubell, T. P., et al.: Central changes in primary afferent fibers following peripheral nerve lesions. Neuroscience 77:1115, 1997. Costigan, M., Mannion, R. J., Kendall, G., et al.: Heat shock protein 27: developmental regulation and expression after peripheral nerve injury. J. Neurosci. 18:5891, 1998. Cudkowicz, M. E., McKenna-Yasek, D., Chen, C., et al.: Limited corticospinal tract involvement in amyotrophic lateral sclerosis subjects with the A4V mutation in the copper/zinc superoxide dismutase gene. Ann. Neurol. 43:703, 1998. Dal Pan, G. J., Glass, J. D., and McArthur, J. C.: Clinicopathologic correlations of HIV-1-associated vacuolar myelopathy: an autopsy-based case-control study. Neurology 44:2159, 1994. Darras-Joly, C., Belec, L., and Lortholary, O.: Subacute polyradiculoneuritis revealing syphilis in an HIV-negative man. Clin. Microbiol. Infect. 6:284, 2000. Dettmers, C., Fatepour, D., Faust, H., and Jerusalem, F.: Sympathetic skin response abnormalities in amyotrophic lateral sclerosis. Muscle Nerve 16:930, 1993. Di Rocco, A.: Diseases of the spinal cord in human immunodeficiency virus infection. Semin. Neurol. 19:151, 1999. Donaghy, M.: Classification and clinical features of motor neurone diseases and motor neuropathies in adults. J. Neurol. 246:331, 1999. Donofrio, P. D., and Walker, F. O.: Tabes dorsalis: electrodiagnostic features. J. Neurol. Neurosurg. Psychiatry 51:1097, 1988. Doron, S. I., Dashe, J. F., Adelman, L. S., et al.: Histopathologically proven poliomyelitis with quadriplegia and loss of brainstem function due to West Nile virus infection. Clin. Infect. Dis. 37:74, 2003. Dorsen, M., and Ehni, G.: Cervical spondylotic radiculopathy producing motor manifestations mimicking primary muscular atrophy. Neurosurgery 5:427, 1999. Dvorak, J., Sutter, M., and Herdmann, J.: Cervical myelopathy: clinical and neurophysiological evaluation. Eur. Spine J. 12(Suppl. 2):S181, 2003.

1318

Diseases of the Peripheral Nervous System

51. Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerve in health, Friedreich’s ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb, W. A. (ed.): Handbook of Electroencephalography and Clinical Neurophysiology. Amsterdam, Elsevier, p. 83, 1971. 52. Ebara, S., Yonenobu, K., Fujiwara, K., et al.: Myelopathy hand characterized by muscle wasting: a different type of myelopathy hand in patients with cervical spondylosis. Spine 13:785, 1988. 53. Eisen, A., Kim, S., and Pant, B.: Amyotrophic lateral sclerosis (ALS): a phylogenetic disease of the corticomotoneuron? Muscle Nerve 15:219, 1992. 54. Estanislao, L. B., and Pachner, A. R.: Spirochetal infection of the nervous system. Neurol. Clin. 17:783, 1999. 55. Etgen, T., Bischoff, C., Resch, M., et al.: Obstacles in the diagnosis and treatment of syphilitic amyotrophy. Neurology 60:509, 2003. 56. Ferracci, F., Fassetta, G., Butler, M. H., et al.: A novel antineuronal antibody in a motor neuron syndrome associated with breast cancer. Neurology 53:852, 1999. 57. Flaherty, M. L., Wijdicks, E. F., Stevens, J. C., et al.: Clinical and electrophysiologic patterns of flaccid paralysis due to West Nile virus. Mayo Clin. Proc. 78:1245, 2003. 58. Fritschi, J., and Sturzenegger, M.: Spinal MRI supporting myelopathic origin of early symptoms in unsuspected cobalamin deficiency. Eur. Neurol. 49:146, 2003. 59. From the Centers for Disease Control and Prevention: investigations of West Nile virus infections in recipients of blood transfusions. JAMA 288:2535, 2002. 60. Gadoth, N., Weitzman, S., and Lehmann, E. E.: Acute anterior myelitis complicating West Nile fever. Arch. Neurol. 36:172, 1979. 61. Galassi, G., Gentilini, M., Ferrari, S., et al.: Motor neuron disease and HIV-1 infection in a 30-year-old HIV-positive heroin abuser: a causal relationship? Clin. Neuropathol. 17:131, 1998. 62. Glass, J. D., Samuels, O., and Rich, M. M.: Poliomyelitis due to West Nile virus. N. Engl. J. Med. 347:1280, 2002. 63. Goldstick, L., Mandybur, T. I., and Bode, R.: Spinal cord degeneration in AIDS. Neurology 35:103, 1985. 64. Gooch, C. L., and Mitsumoto, H.: Infectious, syndromic and systemic disorders. In Brown, W. F., Boulton, C. F., and Aminoff, M. J. (eds.): Neuromuscular Function and Disease. Philadelphia, W. B. Saunders, p. 1281, 2002. 65. Green, R., and Kinsella, L. J.: Current concepts in the diagnosis of cobalamin deficiency. Neurology 45:1435, 1995. 66. Gregory, R., Mills, K., and Donaghy, M.: Progressive sensory nerve dysfunction in amyotrophic lateral sclerosis: a prospective clinical and neurophysiological study. J. Neurol. 240:309, 1993. 67. Griffin, J. W., Cornblath, D. R., Alexander, E., et al.: Ataxic sensory neuropathy and dorsal root ganglionitis associated with Sjogren’s syndrome. Ann. Neurol. 27:304, 1990. 68. Griffin, J. W., and Price, D. L.: Proximal axonopathies induced by toxic chemicals. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical

69. 70. 71.

72.

73. 74.

75.

76.

77. 78. 79.

80.

81. 82. 83.

84.

85.

86.

87.

Neurotoxicology. Baltimore, Williams & Wilkins, p. 161, 1980. Harding, A. E., and Thomas, P. K.: Hereditary distal spinal muscular atrophy. J. Neurol. Sci. 45:337, 1980. Hassin, G. B.: Tabes dorsalis, pathology and pathogenesis: a preliminary report. Arch. Neurol. Psychiatry 21:311, 1929. Healton, E. B., Savage, D. G., Brust, J. C., et al.: Neurologic aspects of cobalamin deficiency. Medicine (Baltimore) 70:229, 1991. Helweg-Larsen, S., Jakobsen, J., Boesen, F., et al.: Myelopathy in AIDS: a clinical and electrophysiological study of 23 Danish patients. Acta Neurol. Scand. 77:64, 1988. Hirano, A.: Aspects of the ultrastructure of amyotrophic lateral sclerosis. Adv. Neurol. 36:75, 1982. Hirano, A., and Llena, J. F.: The central nervous system as a target in toxic-metabolic states. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 24, 1980. Hirayama, K., Tsubaki, T., Toyokura, Y., and Okinaka, S.: Juvenile muscular atrophy of unilateral upper extremity. Neurology 13:373, 1963. Hirsch, M. S., and Werner, B.: Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 17–2003: a 38-year-old woman with fever, headache, and confusion. N. Engl. J. Med. 348:2239, 2003. Hook, E. W. III, and Marra, C. M.: Acquired syphilis in adults. N. Engl. J. Med. 326:1060, 1992. Horstmann, D. M.: Clinical aspects of acute poliomyelitis. Am. J. Med. 6:592, 1949. Hu, M. T. M., Ellis, C. M., Al-Chalabi, A., et al.: Flail arm syndrome: a distinctive variant of amyotrophic lateral sclerosis. J. Neurol. Neurosurg. Psychiatry 65:950, 1998. Huang, P. P., Chin, R., Song, S., and Lasoff, S.: Lower motor neuron dysfunction associated with human immunodeficiency virus infection. Arch. Neurol. 50:1328, 1993. Hubalek, Z.: Comparative symptomatology of West Nile fever. Lancet 358:254, 2001. Hughes, J. T.: Pathology of the Spinal Cord, 2nd ed. Philadelphia, W. B. Saunders, 1978. Ince, P. G., Evans, J., Knopp, M., et al.: Corticospinal tract degeneration in the progressive muscular atrophy variant of ALS. Neurology 60:1252, 2003. Iwamoto, M., Jernigan, D. B., Guasch, A., et al. for the West Nile Virus in Transplant Recipients Investigation Team: Transmission of West Nile virus from an organ donor to four transplant recipients. N. Engl. J. Med. 348:2196, 2003. Jeha, L. E., Sila, C. A., Lederman, R. J., et al.: West Nile virus infection: a new acute paralytic illness. Neurology 61:55, 2003. Kaji, R., Hirota, N., Oka, N., et al.: Anti-GM1 antibodies and impaired blood-nerve barrier may interfere with remyelination in multifocal motor neuropathy. Muscle Nerve 17:108, 1994. Kaji, R., Oka, N., Tsuji, T., et al.: Pathological findings at the site of conduction block in multifocal motor neuropathy. Ann. Neurol. 33:152, 1993.

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease 88. Kameyama, T., Hashizume, Y., Ando, T., et al.: Spinal cord morphology and pathology in ossification of the posterior longitudinal ligament. Brain 118(Pt. 1):263, 1995. 89. Kaminski, H. J., Tresser, N., Hogan, R. E., and Martin, E.: Spinal cord histopathology in long-term survivors of poliomyelitis. Muscle Nerve 18:1208, 1995. 90. Karpati, G., Carpenter, S., and Durham, H.: A hypothesis for the pathogenesis of amyotrophic lateral sclerosis. Rev. Neurol. 144:672, 1988. 91. Kato, T., Katagiri, T., Hirano, A., et al.: Sporadic lower motor neuron disease with Lewy body-like inclusions: a new subgroup? Acta Neuropathol. (Berl.) 76:208, 1988. 92. Kato, T., Katagiri, T., Hirano, A., et al.: Lewy body-like hyaline inclusions in sporadic motor neuron disease are ubiquitinated. Acta Neuropathol. (Berl.) 77:391, 1989. 93. Katz, J. S., Wolfe, G. I., Andersson, P. B., et al.: Brachial amyotrophic diplegia: a slowly progressive motor neuron disorder. Neurology 53:1071, 1999. 94. Kawamura, Y., Dyck, P. J., Shimono, M., et al.: Morphometric comparison of the vulnerability of peripheral motor and sensory neurons in amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 40:667, 1981. 95. Keegan, J. J.: The cause of dissociated motor loss in the upper extremity with cervical spondylosis. J. Neurosurg. 23:528, 1965. 96. Kelley, T. W., Prayson, R. A., and Isada, C. M.: Spinal cord disease in West Nile virus infection. N. Engl. J. Med. 348:564, 2003. 97. Kelley, T. W., Prayson, R. A., Ruiz, A. I., et al.: The neuropathology of West Nile virus meningoencephalitis: a report of two cases and review of the literature. Am. J. Clin. Pathol. 119:749, 2003. 98. Kikuchi, S., Shinpo, K., Niino, M., and Tashiro, K.: Subacute syphilitic meningomyelitis with characteristic spinal MRI findings. J. Neurol. 250:106, 2003. 99. Kim, D. H., Vaccaro, A. R., Henderson, F. C., and Benzel, E. C.: Molecular biology of cervical myelopathy and spinal cord injury: role of oligodendrocyte apoptosis. Spine J. 3:510, 2003. 100. Klein, C., Kimiagar, I., Pollak, L., et al.: Neurological features of West Nile virus infection during the 2000 outbreak in a regional hospital in Israel. J. Neurol. Sci. 200:63, 2002. 101. Kleopa, K. A., Zamba-Papanicolaou, E., and Kyriakides, T.: Compressive lumbar myelopathy presenting as segmental motor neuron disease. Muscle Nerve 28:69, 2003. 102. Kornberg, A. J., and Pestronk, A.: Chronic motor neuropathies: diagnosis, therapy and pathogenesis. Ann. Neurol. 37(Suppl. 1):S43, 1995. 103. Kusaka, H., Matsumoto, S., and Imai, T.: Granulofilamentous profiles in lower motor neurons: a sporadic case of amyotrophic lateral sclerosis with many Lewy body-like inclusions. Clin. Neuropathol. 11:20, 1992. 104. Lanska, M. J., Lanska, D. J., and Schmidley, J. W.: Syphilitic polyradiculopathy in an HIV-positive man. Neurology 38:1297, 1988. 105. Lauria, G., Pareyson, D., Grisoli, M., and Sghirlanzoni, A.: Clinical and magnetic resonance imaging findings in chronic sensory ganglionopathies. Ann. Neurol. 47:104, 2000. 106. Lauria, G., Pareyson, D., and Sghirlanzoni, A.: Neurophysiological diagnosis of acquired sensory ganglionopathies. Eur. Neurol. 50:146, 2003.

1319

107. Lawyer, T., and Netsky, M. G.: Amyotrophic lateral sclerosis: a clinicoanatomic study of fifty-three cases. Arch. Neurol. Psychiatry 69:171, 1953. 108. Lees, F., and Aldren-Turner, J. W.: Natural history and prognosis of cervical spondylosis. Br. Med. J. 2:1607, 1963. 109. Leis, A. A., Fratkin, J., Stokic, D. S., et al.: West Nile poliomyelitis. Lancet Infect. Dis. 3:9, 2003. 110. Leis, A. A., Stokic, D. S., Polk, J. L., et al.: A poliomyelitis-like syndrome from West Nile virus infection. N. Engl. J. Med. 347:1279, 2002. 111. Lekan, H. A., Chung, K., Yoon, Y. W., et al.: Loss of dorsal root ganglion cells concomitant with dorsal root axon sprouting following segmental nerve lesion. Neuroscience 81:527, 1997. 112. Lestini, W. F., and Wiesel, S. W.: The pathogenesis of cervical spondylosis. Clin. Orthop. 239:69, 1989. 113. Li, J., Loeb, J. A., Shy, M. E., et al.: Asymmetric flaccid paralysis: a neuromuscular presentation of West Nile virus infection. Ann. Neurol. 53:703, 2003. 114. Li, M., Sobue, G., Doyu, M., et al.: Primary sensory neurons in X-linked recessive bulbospinal neuronopathy: histopathology and androgen receptor gene expression. Muscle Nerve 18:301, 1995. 115. Liversedge, L. A., Hutchinson, E. C., and Lyons, J. B.: Cervical spondylosis simulating motor-neurone disease. Lancet 2:652, 1953. 116. Lorenzl, S., Vogeser, M., Muller-Schunk, S., and Pfister, H. W.: Clinically and MRI documented funicular myelosis in a patient with metabolical vitamin B12 deficiency but normal vitamin B12 serum level. J. Neurol. 250:1010, 2003. 117. Louis, E. D., Hanley, A. E., Brannagigan, T. H., et al.: Motor neuron disease, lymphoproliferative disease and bone marrow biopsy. Muscle Nerve 19:1334, 1996. 118. Love, S., and Wiley, C. A.: Viral diseases. In Graham, D. I., and Lantos, P. L. (eds.): Greenfield’s Neuropathology, 7th ed. London, Arnold, p. 1, 2002. 119. Mahieux, F., Gray, F., Fenelon, G., et al.: Acute myeloradiculitis due to cytomegalovirus as the initial manifestation of AIDS. J. Neurol. Neurosurg. Psychiatry 52:270, 1989. 120. Mair, W. G. P., and Druckman, R.: The pathology of spinal cord lesions and their relation to the clinical features in protrusion of cervical intervertebral discs. Brain 76:70, 1953. 121. Malinow, K., Yannakakis, G. D., Glusman, S. M., et al.: Subacute sensory neuronopathy secondary to dorsal root ganglionitis in primary Sjogren’s syndrome. Ann. Neurol. 20:535, 1986. 122. Manetto, V., Sternberger, N. H., Perry, G., et al.: Phosphorylation of neurofilaments is altered in amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 47:642, 1988. 123. Manuelides, E.: Neuropathology of experimental West Nile virus infection in monkeys. J. Neuropathol. Exp. Neurol. 15:448, 1956. 124. Martin, L. J., Price, A. C., Kaiser, A., et al.: Mechanisms for neuronal degeneration in amyotrophic lateral sclerosis and in models of motor neuron death [review]. Int. J. Molec. Med. 5:3, 2000. 125. Marx, A., Glass, J. D., and Sutter, R. W.: Differential diagnosis of acute flaccid paralysis and its role in poliomyelitis surveillance. Epidemiol. Rev. 22:298, 2000.

1320

Diseases of the Peripheral Nervous System

126. Matsumoto, S., Goto, S., Kusaka, H., et al.: Synaptic pathology of spinal anterior horn cells in amyotrophic lateral sclerosis: an immunohistochemical study. J. Neurol. Sci. 125:180, 1994. 127. Matthews, J. A.: Wasting of the small hand muscles in upper and mid-cervical cord lesions. Q. J. Med. 91:693, 1998. 128. McCombe, P. A., and McLeod, J. G.: The peripheral neuropathy of vitamin B12 deficiency. J. Neurol. Sci. 66:117, 1984. 129. Metz, J.: Cobalamin deficiency and the pathogenesis of nervous system disease. Annu. Rev. Nutr. 12:59, 1992. 130. Miller, D. C.: Post-polio syndrome spinal cord pathology. Ann. N. Y. Acad. Sci. 753:186, 1995. 131. Mori, K, Koike, H., Misu, K., et al.: Spinal cord magnetic resonance imaging demonstrates sensory neuronal involvement and clinical severity in neuronopathy associated with Sjogren’s syndrome. J. Neurol. Neurosurg. Psychiatry 71:488, 2001. 132. Mostashari, F., Bunning, M. L., Kitsutani, P. T., et al.: Epidemic West Nile encephalitis, New York, 1999: results of a household-based seroepidemiological survey. Lancet 358:261, 2001. 133. Mulder, D. W.: Clinical observations on acute poliomyelitis. Ann. N. Y. Acad. Sci. 753:1, 1995. 134. Narayanaswami, P., and Bertorini, T. E.: Progressive amyotrophy as a late complication of myelopathy. J. Neurol. Sci. 184:11, 2001. 135. Nash, D., Mostashari, F., Fine, A., et al.: The outbreak of West Nile virus infection in the New York City area in 1999. N. Engl. J. Med. 344:1807, 2001. 136. Nichter, C. A., Pavlakis, S. G., Shaikh, U., et al.: Rhombencephalitis caused by West Nile fever virus. Neurology 55:153, 2000. 137. Nishio, M., Koizumi, K., Moriwaka, F., et al.: Reversal of HIV-associated motor neuron syndrome after highly active antiretroviral therapy. J. Neurol. 248:233, 2001. 138. Ogawa-Goto, K., Funamoto, N., Abe, T., and Nagushima, K.: Different ceramide compositions of gangliosides between human motor and sensory nerves. J. Neurochem. 55:1486, 1990. 139. Ogino, H., Tada, K., Okada, K., et al.: Canal diameter, anteroposterior compression ratio, and spondylotic myelopathy of the cervical spine. Spine 8:1, 1983. 140. Ohka, S., and Nomoto, A.: Recent insights into poliovirus pathogenesis. Trends Microbiol. 9:501, 2001. 141. Pachner, A. R.: Spirochetal diseases of the CNS. Neurol. Clin. 4:207, 1986. 142. Park, M., Hui, J. S., and Bartt, R. E.: Acute anterior radiculitis associated with West Nile virus infection. J. Neurol. Neurosurg. Psychiatry 74:823, 2003. 143. Parquet, M. C., Kumatori, A., Hasebe, F., et al.: West Nile virus-induced bax-dependent apoptosis. FEBS Lett. 500:17, 2001. 144. Parry, G. J., and Clarke, S.: Multifocal acquired demyelinating neuropathy masquerading as motor neuron disease. Muscle Nerve 11:103, 1988. 145. Parry, G. J., Holtz, S. J., Ben-Zeev, D., and Drori, J. B.: Gammopathy with proximal motor axonopathy simulating motor neuron disease. Neurology 36:273, 1986.

146. Pealer, L. N., Marfin, A. A., Petersen, L. R., et al.: Transmission of West Nile virus through blood transfusion in the United States in 2002. N. Engl. J. Med. 349:1236, 2003. 147. Pearl, D., Noursadeghi, M., Manji, H., et al.: Lower motor neuron syndrome and HIV infection. Sex. Transm. Infect. 79:351, 2003. 148. Pestronk, A., Chaudhry, V., Feldman, E. L., et al.: Lower motor neuron syndromes defined by patterns of weakness, nerve conduction abnormalities, and high titers of antiglycolipid antibodies. Ann. Neurol. 27:316, 1990. 149. Petersen, L. R., and Marfin, A. A.: West Nile virus: a primer for the clinician. Ann. Intern. Med. 137:173, 2002. 150. Petersen, L. R., Roehrig, J. T., and Hughes, J. M.: West Nile virus encephalitis. N. Engl. J. Med. 347:1225, 2002. 151. Petito, C. K., Navia, B. A., Cho, E. S., et al.: Vacuolar myelopathy pathologically resembling subacute combined degeneration in patients with the acquired immunodeficiency syndrome. N. Engl. J. Med. 312:874, 1985. 152. Petito, C. K., Vecchio, D., and Chen, Y. T.: HIV antigen and DNA in AIDS spinal cords correlate with macrophage infiltration but not with vacuolar myelopathy. J. Neuropathol. Exp. Neurol. 53:86, 1994. 153. Pezeshkpour, G. H., and Dalakas, M. C.: Long-term changes in the spinal cords of patients with old poliomyelitis: signs of continuous disease activity. Arch. Neurol. 45:505, 1988. 154. Polo, A., Teatini, F., D’Anna, S., et al.: Sensory involvement in X-linked spino-bulbar muscular atrophy (Kennedy’s syndrome): an electrophysiological study. J. Neurol. 243:388, 1996. 155. Prabhakar, S., Chopra, J. S., Banerjee, A. K., and Rana, P. V. S.: Wasted leg syndrome: a clinical, electrophysiological and histopathological study. Clin. Neurol. Neurosurg. 83:19, 1981. 156. Price, D. L., and Griffin, J. W.: Neurons and ensheathing cells as targets of disease processes. In Spencer, P. S. and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 2, 1980. 157. Qaiyum, M., and Sandrasegaran, K.: Post-operative paraesthesia. Br. J. Radiol. 73:791, 2000. 158. Quattrini, A., Nemni, R., Sferrazza, B., et al.: Amyloid neuropathy simulating lower motor neuron disease. Neurology 51:600, 1998. 159. Roehrig, J. T., Nash, D., Maldin, B., et al.: Persistence of virus-reactive serum immunoglobulin M antibody in confirmed West Nile virus encephalitis cases. Emerg. Infect. Dis. 9:376, 2003. 160. Roos, K. L.: Fever and asymmetrical weakness in the summer: evidence of a West Nile virus-associated poliomyelitis-like illness. Mayo Clin. Proc. 78:1205, 2003. 161. Rosenfeld, J., Chang, S. W., Jackson, C. E., et al.: Lower extremity amyotrophic diplegia (LAD): a new clinical entity in the spectrum of motor neuron disease. Neurology 58(Suppl. 3):A411, 2002. 162. Rowland, L. P., Defenini, R., Sherman, W., et al.: Macroglobulinemia with peripheral neuropathy simulating motor neuron disease. Ann. Neurol. 11:532, 1982.

Peripheral Nerve Involvement of Spinal Cord, Spinal Roots, and Meningeal Disease 163. Sampathkumar, P.: West Nile virus: epidemiology, clinical presentation, diagnosis, and prevention. Mayo Clin. Proc. 78:1137, 2003. 164. Sampson, B. A., and Armbrustmacher, V.: West Nile encephalitis: the neuropathology of four fatalities. Ann. N. Y. Acad. Sci. 951:172, 2001. 165. Sampson, B. A., Nields, H., Armbrustmacher, V., and Asnis, D. S.: Muscle weakness in West Nile encephalitis is due to destruction of motor neurons. Hum. Pathol. 34:628, 2003. 166. Saperstein, D. S., Wolfe, G. I., Gronseth, G. S., et al.: Challenges in the identification of cobalamin-deficiency polyneuropathy. Arch. Neurol. 60:1296, 2003. 167. Sasaki, S., and Iwata, M.: Synaptic loss in the proximal axon of anterior horn neurons in motor neuron disease. Acta Neuropathol. (Berl.) 90:170, 1995. 168. Sasaki, S., and Maruyama, S.: Decreased synaptophysin immunoreactivity of the anterior horns in motor neuron disease. Acta Neuropathol. (Berl.) 87:125, 1994. 169. Sastre-Garriga, J., Tintore, M., Raguer, N., et al.: Lower motor neuron disease in HIV-2 infected woman. J. Neurol. 247:718, 2000. 170. Satake, M., Yoshimura, T., Iwaki, T., et al.: Anti-dorsal root ganglion neuron antibody in a case of dorsal root ganglionitis with Sjogren’s syndrome. J. Neurol. Sci. 12:122, 1995. 171. Schmidt, M. H., Quinones-Hinojosa, A., and Rosenberg, W. S.: Cervical myelopathy associated with degenerative spine disease and ossification of the posterior longitudinal ligament. Semin. Neurol. 22:143, 2002. 172. Schoenen, J.: Clinical anatomy of the spinal cord. In Young, R. R., and Woolsey, R. M. (eds.): Diagnosis and Management of Disorders of the Spinal Cord. Philadelphia, W. B. Saunders, p. 1, 1995. 173. Sejvar, J. J., Haddad, M. B., Tierney, B. C., et al.: Neurologic manifestations and outcome of West Nile virus infection. JAMA 290:511, 2003. 174. Shevell, M. I., and Rosenblatt, D. S.: The neurology of cobalamin. Can. J. Neurol. Sci. 19:472, 1992. 175. Simon, R. P.: Neurosyphilis. Arch. Neurol. 42:606, 1985. 176. Sobue, G., Hashizume, Y., Yasuda, T., et al.: Phosphorylated high molecular weight neurofilament protein in lower motor neurons in amyotrophic lateral sclerosis and other neurodegenerative diseases involving ventral horn cells. Acta Neuropathol. (Berl.) 79:402, 1990. 177. Sobue, G., Yasuda, T., Kachi, T., et al.: Chronic progressive sensory ataxic neuropathy: clinicopathological features of idiopathic and Sjogren’s syndrome-associated cases. J. Neurol. 240:1, 1993. 178. Sobue, I., Saito, N., Iida, M., and Ando, K.: Juvenile type of distal and segmental muscular atrophy of upper extremities. Ann. Neurol. 3:429, 1978. 179. Spencer, P. S., and Schaumburg, H. H.: Classification of neurotoxic disease: a morphological approach. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 92, 1980. 180. Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, 1980. 181. Staudinger, R., and Henry, K.: Remission of HIV myelopathy after highly active antiretroviral therapy. Neurology 54:267, 2000.

1321

182. Stern, R. O.: A study of the histopathology of tabes dorsalis with special reference to Richter’s theory of its pathogenesis. Brain 52:295, 1929. 183. Surtees, R.: Biochemical pathogenesis of subacute combined degeneration of the spinal cord and brain. J. Inherit. Metab. Dis. 16:762, 1993. 184. Swash, M., Leader, M., Brown, A., and Swettenham, K. W.: Focal loss of anterior horn cells in the cervical cord in motor neuron disease. Brain 109:939, 1986. 185. Tagliati, M., Di Rocco, A., Danisi, F., and Simpson, D. M.: The role of somatosensory evoked potentials in the diagnosis of AIDS-associated myelopathy. Neurology 54:1477, 2000. 186. Takahashi, T., Yagishita, S., Amano, N., et al.: Amyotrophic lateral sclerosis with numerous axonal spheroids in the corticospinal tract and massive degeneration of the cortex. Acta Neuropathol. (Berl.) 94:294, 1997. 187. Taylor, B. V., Dyck, P. J. B., Engelstad, J., et al.: Multifocal motor neuropathy: pathologic alterations at the site of conduction block. J. Neuropathol. Exp. Neurol. 63:129, 2004. 188. Thomas, P. K.: The peripheral nervous system as a target for toxic substances. In Spencer, P. S. and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 35, 1980. 189. Troost, D., van den Oord, J. J., de Jong, J. M. B. V., and Swaab, D. F.: Lymphocytic infiltration in the spinal cord of patients with amyotrophic lateral sclerosis. Clin. Neuropathol. 8:289, 1989. 190. Ugawa, Y., Kohara, N., Shimpo, T., and Mannen, T.: Central motor and sensory conduction in adrenoleukomyeloneuropathy, cerebrotendinous xanthomatosis, HTLV-1-associated myelopathy and tabes dorsalis. J. Neurol. Neurosurg. Psychiatry 51:1069, 1988. 191. Van Asseldonk, J. T. H., Van den Berg, L. H., Van den BergVos, R. M., et al.: Demyelination and axonal loss in multifocal motor neuropathy: distribution and relation to weakness. Brain 126:186, 2003. 192. Van den Berg-Vos, R. M., Van den Berg, L. H., Visser, J., et al.: The spectrum of lower motor neuron syndromes. J. Neurol. 250:1279, 2003. 193. Van den Berg-Vos, R. M., Visser, J., Franssen, H., et al.: Sporadic lower motor neuron disease with adult onset: classification of subtypes. Brain 126:1036, 2003. 194. Verma, A., Berger, J. R., Snodgrass, S., and Petito, C.: Motor neuron disease: a paraneoplastic process associated with anti-Hu antibody and small cell lung carcinoma. Ann. Neurol. 40:112, 1996. 195. Verma, R. K., Ziegler, D. K., and Kepes, J. J.: HIV-related neuromuscular syndrome simulating motor neuron disease. Neurology 40:544, 1990. 196. Victor, M., and Lear, A. A.: Subacute combined degeneration of the spinal cord. Am. J. Med. 20:896, 1956. 197. Wada, E., Ohmura, M., and Yonenobu, K.: Intramedullary changes of the spinal cord in cervical spondylotic myelopathy. Spine 20:2226, 1995. 198. Wakabayashi, K., Horikawa, Y., Oyake, M., et al.: Sporadic motor neuron disease with severe sensory neuronopathy. Acta Neuropathol. (Berl.) 95:426, 1998.

1322

Diseases of the Peripheral Nervous System

199. Woolf, C. J., Shortland, P., and Coggeshall, R. E.: Peripheral nerve injury triggers central sprouting of myelinated afferents. Nature 355:75, 1992. 200. Woolf, C. J., Shortland, P., Reynolds, M., et al.: Reorganization of central terminals of myelinated primary afferents in the rat dorsal horn following peripheral axotomy. J. Comp. Neurol. 360:121, 1995. 201. Worrall, B. B., Rowland, L. P., Chin, S. S-M., and Mastrianni, J. A.: Amyotrophy in prion diseases. Arch. Neurol. 57:33, 2000. 202. Xiao, S. Y., Guzman, H., Zhang, H., et al.: West Nile virus infection in the golden hamster (Mesocricetus auratus): a

model for West Nile encephalitis. Emerg. Infect. Dis. 7:714, 2001. 203. Yamaura, I., Yone, K., Nakahara, S., et al.: Mechanism of destructive pathologic changes in the spinal cord under chronic mechanical compression. Spine 27:21, 2002. 204. Yang, J. S., Ramanathan, M. P., Muthumani, K., et al.: Induction of inflammation by West Nile virus capsid through the caspase-9 apoptotic pathway. Emerg. Infect. Dis. 8:1379, 2002. 205. Younger, D. S., Rowland, L. P., Latov, N., et al.: Lymphoma, motor neuron diseases and amyotrophic lateral sclerosis. Ann. Neurol. 29:78, 1991.

54 Diseases of Spinal Roots COLIN CHALK

Gross Anatomy Microscopic Anatomy The Diagnosis of Radiculopathy Disorders Affecting Single Nerve Roots Acute Intervertebral Disc Herniation Spinal Root Trauma

Miscellaneous Causes of Single Root Lesions Polyradiculopathy Spondylotic Disease of the Spine Neoplastic Disease of Spinal Roots Infectious Causes of Polyradiculopathy Diabetic Polyradiculopathy

GROSS ANATOMY There are 31 pairs of spinal nerve roots: 8 cervical, 12 thoracic, 5 lumbar, 5 sacral, and 1 coccygeal. Each nerve root is formed by the confluence of ventral (anterior) and dorsal (posterior) rootlets arising from the corresponding spinal cord segment. An exception is the C1 root, which lacks posterior rootlets. The dorsal and ventral roots remain separate within the subarachnoid space, pierce the dura mater independently, and then coalesce as they enter their intervertebral foramen. In the cervical spine, nerve roots exit superior to the synonymous vertebrae. Because there are only seven cervical vertebrae, the C8 root lies superior to the T1 vertebra, and from the T1 vertebra caudally, the relationship between roots and vertebrae changes, with roots now exiting inferior to the synonymous vertebrae. Owing to the differential growth of spinal cord and vertebral column, the distance between the origin of the nerve roots and their intervertebral foramina increases progressively from the cervical to sacral levels. As a consequence, whereas the cervical nerve roots are almost perpendicular to the spinal cord, the angle between roots and cord becomes increasingly oblique as one proceeds caudally. The spinal cord ends at the L1 vertebral level in adults, so that the subarachnoid space below L1 contains only dorsal and ventral roots (the cauda equina). An

Radiation Radiculopathy Chronic Arachnoiditis Toxic Meningoradiculopathy Congenital Anomalies Miscellaneous Causes of Polyradiculopathy or Cauda Equina Syndrome

important clinical corollary is that a space-occupying lesion in the spinal canal below L1 will not affect the spinal cord, but it may compromise nerve roots from several segmental levels, particularly if the lesion is located medially. On each dorsal root is a small swelling, the dorsal root ganglion (DRG), within which are found the cell bodies of the dorsal root axons. The coalescence of dorsal and ventral roots occurs just distal to the DRG, and the DRG is generally assumed to lie extradurally, within the intervertebral foramen (Fig. 54–1). Pathologic processes that affect only structures lying within the dural sac are thus expected to spare the DRGs. Electromyographers rely on this anatomic point when localizing lesions involving the peripheral sensory pathways: Lesions distal to or involving the DRG will decrease or abolish sensory nerve action potentials (SNAPs) recorded in the distal nerve, while lesions proximal to the DRG spare the SNAPs. By extension, a process that affects sensation but does not alter sensory nerve conduction studies is usually assumed to be intraspinal. However, some recent studies suggest that this criterion may be less robust than is generally assumed. A cadaver and radiographic study found that almost half of L5 and up to three fourths of S1 DRGs were at least partly intraspinal,39 and thus potentially vulnerable to intraspinal processes. A small series of patients with L5 radiculopathy caused by clear, active intraspinal pathology in which the peroneal nerve SNAP was abnormal has been reported.44 1323

1324

Diseases of the Peripheral Nervous System

FIGURE 54–1 Cross section of the cervical spine. Note the dorsal and ventral roots lying within the dural sac, relatively vulnerable to intervertebral disc herniation, whereas the cervical dorsal root ganglia lie within the intervertebral foramina and are well protected from disc protrusions. (From Levin, K. H.: Electrodiagnostic approach to the patient with radiculopathy. Neurol. Clin. 20:397, 2002, with permission.)

MICROSCOPIC ANATOMY Spinal roots are much less familiar than peripheral nerve to histologists and pathologists, because removal of specimens for microscopic study is difficult, even in cadavers, and spinal root biopsy is only rarely performed in living subjects. The microscopic anatomy of spinal roots is discussed in some detail in Chapter 3, but some features that differ between roots and peripheral nerve are worthy of emphasis because they may at least partly explain different responses to pathologic processes. As each rootlet emerges from the spinal cord, the bundle of axons is ensheathed by an extension of the pia mater. This root sheath is analogous to the perineurium of peripheral nerve, except that the latter is thicker and stronger and contains more collagen fibers. The root sheath follows the root through its course in the subarachnoid space, and becomes continuous with the perineurium when the root exits from the dural sac. As the rootlets converge in the subarachnoid space, the more superficial fibers of the root sheath intermingle to encircle and combine adjacent rootlets until they are eventually wrapped in a loose cylindrical bundle by a thin layer of loose connective tissue. Thus, access to the endoneurial contents of spinal roots from the cerebrospinal fluid (CSF) is relatively easy, in contrast to the endoneurial space of peripheral nerves, which is ensheathed by a thicker and well-organized perineurium. The blood supply of nerve roots also differs from that of peripheral nerve. A contemporary cadaver study using light and scanning electron microscopy showed that the blood supply of spinal roots consists of intrinsic radicular vessels, which are connected by coiled anastomoses.63 Unlike peripheral nerve, in which the intrinsic endoneurial blood vessel network is regularly reinforced by extrinsic epineurial vessels that are branches of major limb arteries,

in spinal roots the source of blood appears to be limited to either end of the root. This arrangement would appear to leave the longer lumbosacral roots relatively vulnerable to ischemia, although it may be that the CSF serves as an alternate source of nutrients for the spinal roots.

THE DIAGNOSIS OF RADICULOPATHY The most common cause of single root lesions, herniation of the intervertebral nucleus pulposus, produces a characteristic picture of limb and spine pain, paresthesia in the limb, and often weakness that is familiar to all clinicians. Difficulty may arise if pain is minimal or absent, in which case making the distinction between radiculopathy and mononeuropathy depends on analysis of the motor, sensory, and reflex findings. The most frequently encountered example is the patient with footdrop, in whom weakness of ankle inversion and hip abduction in addition to footdrop points to an L5 root lesion, whereas sparing of ankle inversion and hip abduction are expected with a peroneal neuropathy. The electrophysiologic hallmarks of radiculopathy are (1) normal SNAPs despite sensory symptoms and signs; (2) needle electromyography (EMG) signs of denervation, reinnervation, or both, in proximal and distal muscles within the same myotome; and (3) EMG signs of paraspinal muscle denervation or reinnervation. The preservation of the SNAP implies that DRG cells and their peripheral processes are intact, and that the lesion disturbing sensation is proximal to the DRG. For the most part this is a reliable way to distinguish radiculopathy from neuropathy or ganglionopathy, although there are some well-described exceptions.44 In fact, given that autopsy studies suggest that an intraspinal location for L5 and S1

1325

Diseases of Spinal Roots

DRGs is common,39 it is perhaps surprising that exceptions are not more frequent. Needle EMG relies on a myotomal pattern of muscle involvement to distinguish lesions of root from nerve, and to localize the involved segmental level. Although electromyographers generally acknowledge that the exact segmental innervation of muscles varies from patient to patient, the degree of variability is debated. In one study, by directly stimulating ventral roots in series of patients undergoing dorsal rhizotomy, it was suggested that the largest contributor to the innervation of a given leg muscle might come from any one of three levels; for example, for the tibialis anterior, the largest innervation was from L4 in 55% of patients, from L5 in 41%, and from L3 in 3%.67 Other studies correlating electrodiagnostic and surgical findings in patients with lumbosacral radiculopathy have found less variability in segmental innervation of leg muscles, with little overlap among L2-L4, L5, and S1 radiculopathies.86 The presence of fibrillation potentials or positive waves in paraspinal muscles implies that the lesion is proximal to the origin of the posterior primary ramus, which leaves the spinal nerve immediately after the latter exits from the intervertebral foramen. Paraspinal denervation is a relatively insensitive indicator of spinal root lesions, however. For example, paraspinal fibrillation was found in only 5% of patients with lumbosacral radiculopathy of duration of 8 months or longer in one study.97 In acute disc herniation radiculopathies, paraspinal fibrillation is more often found, but still only in about half the patients.45,86 Paraspinal fibrillation potentials are also occasionally found in subjects without any symptoms or signs of radiculopathy.20 Standard nerve conduction studies and EMG have a significant false-negative rate for the diagnosis of radiculopathy resulting from disc herniation. In the hope of improving diagnostic sensitivity, a number of other electrophysiologic techniques have been investigated, including F waves, H reflexes, somatosensory evoked potentials, and electrical and magnetic root stimulation.28,46,93 There is no consensus at present about the role for any of these complementary techniques. A recent comprehensive review concluded that H reflexes and F waves (particularly using parameters such as F wave chronodispersion) might have a meaningful role, complementary to the needle examination, and that magnetic root stimulation deserves further study in more patients.28 Magnetic resonance imaging (MRI) of the spine has become widely available in the last decade, and in centers with MRI, myelography is now employed infrequently. MRI offers several advantages over computed tomography (CT)–myelography: MRI is noninvasive, generally offers superior anatomic resolution, and allows direct imaging of the spinal cord. Whether MRI also renders electrophysiologic testing obsolete in the investigation of a patient with suspected radiculopathy resulting from disc herniation or spondylotic disease has been debated. Some argue that

EMG should be reserved only for occasional patients in whom either the clinical or imaging data are unclear.7 Others point out that MRI is very sensitive, but is less specific than EMG. One retrospective study found that EMG and MRI came to similar conclusions in two thirds of patients, but in the remaining one third only one of the studies was abnormal, leading the authors to conclude that the two techniques remain complementary.56 Given the considerable health system costs generated by the investigation of this common clinical problem, a careful prospective study to determine the most cost-effective use of these techniques is needed. CSF examination is essential in patients with polyradiculopathy, and may be indicated in single root lesions if imaging is unrevealing. In addition to routine studies (cell count, glucose, and total protein), cytologic examination for malignant cells is often needed, so that a sufficient volume of fresh CSF (at least 5 mL) must be sent to the cytology laboratory. Depending on the clinical setting, culture, immunologic testing, or polymerase chain reaction (PCR) assays for a variety of microorganisms (e.g., Borrelia burgdorferi, human immunodeficiency virus [HIV], herpes simplex or zoster, Treponema pallidum) may be needed.

DISORDERS AFFECTING SINGLE SPINAL ROOTS Acute Intervertebral Disc Herniation Acute disc herniation resulting in spinal root compression is the most common disorder of spinal nerves encountered in clinical practice. Radiculopathy resulting from disc herniation may occur at any age in adult life, and is occasionally seen in adolescents. With advancing age, as osteoarthritis of the spine becomes almost universal, compressive radiculopathies are often due to an acute disc herniation superimposed upon preexisting intervertebral foramen narrowing. The typical clinical history is highly characteristic. The first symptom is usually back or neck pain. This may be precipitated by trauma to the neck or back, or may follow some unusual physical exertion, but equally often patients simply waken one morning with the symptom. After a short time (hours to days), pain will begin to radiate from the spine into the shoulder or hip, and then into the limb. Pain is usually unremitting and severe, and is aggravated by movements of the spine or limb. Aggravation of the limb pain during the Valsalva maneuver (e.g., coughing, sneezing, straining) is highly suggestive of intraspinal pathology. It is important to question the patient about this symptom; although present only in a minority of patients with disc herniation radiculopathy, it is highly suggestive of the diagnosis. At about the same time as pain radiation into the limb begins, patients may develop sensory disturbance, such as prickling, pins and

1326

Diseases of the Peripheral Nervous System

needles, or local anesthetic-like numbness. Many patients will report this to be well localized (e.g., the tip of the index finger, the lateral side of the foot), but in some patients numbness is felt rather diffusely throughout the limb. Weakness may be reported concomitant with the sensory symptoms, but it can be difficult to sort out whether this is true muscle weakness or function is limited because limb movement and effort worsen the pain. Alternatively, patients may be so distracted by pain that they do not realize they have lost muscle power. This seems to be frequent with C7 radiculopathies, perhaps because weakness of elbow extension is less noticeable to patients than are footdrop (L5), intrinsic hand muscle weakness (C8), or trouble abducting the arm at the shoulder (C5). Finding reflex loss or weakness in a segmental distribution on examination provides strong confirmatory evidence for the diagnosis, but the physical examination is often normal. One special maneuver for cervical radiculopathy, the Spurling test, in which the examiner attempts to precipitate symptoms in the arm by applying axial pressure to the spine while the patient’s neck is flexed laterally, was found to be insensitive but quite specific for diagnosing radiculopathy in one large series of patients referred for evaluation of arm pain.85 Aggravation or precipitation of leg pain by straight leg raising (Lasègue’s test) has long been used to support a diagnosis of lumbosacral radiculopathy, but there are conflicting opinions about the maneuver’s reliability. A recent large, prospective study examined the value of a range of historical and physical examination findings in patients with acute sciatica, using MRI evidence of root compression as the gold standard for diagnosis of radiculopathy.90 Although positive straight leg raising was predictive of root compression on MRI, it was not an independent predictor following multivariate regression analysis (inability to bend over and touch the floor, a related phenomenon, was an independent predictor). The other variables that were independent predictors of root compression in this study were patient age over 50, pain worse in the leg than the back, pain in a dermatomal distribution, pain worsened by Valsalva maneuvers, and weakness. The main investigations to consider in a patient with radiculopathy resulting from disc herniation are EMG/nerve conduction studies and imaging of the spine, which are discussed above. It is worth re-emphasizing that there is neither consensus nor good evidence about the relative merits of these two investigative tools, and most patients with persistent radicular symptoms and signs will likely end up having both. The natural history of root compression resulting from disc herniation is generally favorable, with symptoms and even neurologic deficits resolving with time in most patients. In a large cohort of patients with cervical radiculopathy followed up to 5 years in Olmsted County, Minnesota, 90% were asymptomatic or only mildly incapacitated by symptoms at last follow-up, although 26%

underwent surgery.70 In a prospective study that randomized patients with lumbar radiculopathy to conservative or surgical treatment, 60% of the nonoperated patients had satisfactory outcomes for up to 10 years.91 Treatment of acute radicular pain is often difficult, and even narcotic analgesics may not provide complete relief. Nonsteroidal anti-inflammatory drugs are often tried, but are not often helpful. Some patients obtain relief with agents used for neuropathic pain, such as amitriptyline or gabapentin. Patients usually find that immobilization of the neck or back helps to some degree, although the value of the traditional prescription of bed rest for lumbosacral radiculopathy has been in question for some time. A recent large prospective study from the Netherlands found that 14 days of bed rest was no more effective than watchful waiting in patients with acute lumbosacral radiculopathy.89 Intraspinal injections of medications such as local anesthetics or corticosteroids have been used to treat radicular pain for almost a century. Evaluating these techniques is difficult because of the scarcity of controlled, blinded trials. In addition, concerns have been raised about the accuracy of injection placement, complications related to injections, and neurotoxic effects of preservatives used in current topical steroid preparations.57 As with other invasive treatments, techniques and operator skill are likely to be important variables that influence outcomes. One recent blinded study randomized patients with lumbar radicular pain to receive fluoroscopically guided nerve root sleeve injections of betamethasone plus bupivacaine or bupivacaine alone. At about 1 year of follow-up, 20 of the 28 betamethasone patients but only 9 of the 27 controls decided against surgery.72 The usual indications for referring patients with radiculopathy resulting from disc herniation for surgery are intractable pain or significant or worsening muscle weakness. In well-selected patients, there can be impressively prompt relief of pain and improvement in motor function following surgery. However, it is difficult to provide an overall assessment of the efficacy of surgical treatment of radiculopathy resulting from disc herniation, given the condition’s generally favorable natural history and the paucity of welldesigned controlled studies of surgery. The only published randomized, controlled trial of surgery in patients with cervical radiculopathy found that surgery resulted in more rapid pain relief than did conservative treatment, but that functional outcomes at 1 year were similar.65 A prospective, randomized trial in patients with lumbar radiculopathy showed faster pain relief with surgery, but long-term outcomes were difficult to assess because many of the conservatively treated patients eventually underwent surgery.91

Spinal Root Trauma Traumatic nerve root injury generally is seen only in patients who have suffered major trauma, such as a high-velocity

1327

Diseases of Spinal Roots

motor vehicle accident. Cervical nerve root injury is the most common, presumably because the cervical spine is the most mobile spinal segment. Trauma producing root injury is almost invariably associated with vertebral, pelvic, or long bone fractures, and the initial neurologic picture is often dominated by concurrent traumatic brain injury. Root injury often coexists with plexus injury. This is particularly likely with trauma to the neck and upper body, and at the bedside it can be very difficult to distinguish between brachial plexopathy and cervical radiculopathy. Electrophysiologic testing is usually the key: Diminished or absent SNAPs indicate a lesion at or distal to the DRG, while preserved SNAP amplitudes point to a root lesion. The combination of dense dermatomal anesthesia with normal sensory nerve conduction studies is highly characteristic of nerve root avulsion. Sensory nerve conduction studies can be definitive in distinguishing between C6-C7-C8 root and middle or lower brachial plexus lesions, but there is no reliable sensory nerve conduction technique for differentiating C5C6 root from upper brachial plexus lesions. The pattern of electromyographic findings may also help with differentiating a root from a plexus lesion, although multifocal plexus or root lesions are common, and may confuse the situation. Distinguishing between root and plexus injury is primarily of prognostic importance because the prospects of recovery of function are better with a plexus lesion. The most severe variety of spinal root injury is complete avulsion of the root from the spinal cord. The diagnosis of root avulsion may be supported by imaging studies showing absence of the intrathecal portion of the spinal root and formation of a pseudomeningocele at the site of the avulsed root sleeve. Myelography and CT-myelography are still the preferred imaging modalities in this situation in most centers, although experience with modified MRI techniques is accumulating.25 Clinical and electrophysiologic abnormalities similar to those of root avulsion may occur in severe stretch injuries, in which axonal continuity is interrupted but the connective tissue elements of the root remain attached to the spinal cord. Despite the physical continuity, prognosis for recovery is poor: Any sensory axons regenerating from the DRGs will be unable to extend into the inhospitable central nervous system (CNS) extracellular milieu, and axonal transection so proximate to the anterior horn cell body generally leads to chromatolytic degeneration of the cell body. The outlook is obviously even more pessimistic with root avulsion. Despite this, there have been some attempts to reimplant avulsed roots. One group reported some recovery of motor function in 3 of 10 patients operated on within a few weeks of root avulsion.15

Miscellaneous Causes of Single Root Lesions Mimics of acute disc herniation are encountered from time to time. Probably most important is vertebral metastasis

(discussed further below), although more often this affects multiple roots. Other reported causes of single root lesions include Paget’s disease of the spine,24 gouty tophi,62 perineurial (Tarlov) cysts,84 and inadvertent extrusion of methylmethacrylate during vertebroplasty.71

POLYRADICULOPATHY Polyradiculopathy arises as a diagnostic possibility in patients with sensory disturbance and lower motor neuron weakness affecting more than a single root, often involving two or more limbs. The initial differential diagnosis may be wide, and includes polyneuropathy, multiple mononeuropathies, plexopathy, or even myopathy and neuromuscular junction disorders. Polyradiculopathy may be suggested by spine pain, or by a pattern of weakness that affects both proximal and distal muscles. In the latter circumstance, the presence of sensory symptoms and signs helps to distinguish polyradiculopathy from muscle and neuromuscular junction disease, but in some patients with polyradiculopathy, sensory features are minimal. Making the distinction between a brachial or lumbosacral plexopathy and a process involving the corresponding spinal roots may be impossible on clinical grounds alone. Perhaps the key point to bear in mind is that polyradiculopathy enters the differential diagnosis of many peripheral nervous system disorders, and the possibility of polyradiculopathy is easy to overlook. Nerve conduction studies and EMG are key to identifying that a patient’s symptoms and signs are due to polyradiculopathy. Normal sensory nerve conduction studies in territories with sensory abnormalities are a hallmark of preganglionic (i.e., root) disorders, and exclude plexopathy, multiple mononeuropathies, and polyneuropathy. The low-amplitude, short-duration motor unit potentials of myopathy are the converse of the EMG findings in polyradiculopathy, and the involvement of proximal (including paraspinal) and distal muscles is a further means to separate polyradiculopathy from polyneuropathy and plexopathy. Although it is difficult to distinguish between polyradiculopathy and anterior horn cell disease by electrophysiologic criteria alone, in the clinical setting this distinction is usually not problematic. Diseases affecting the gray matter of the spinal cord may also result in an electrophysiologic picture identical to that of polyradiculopathy, but can be distinguished by the presence of upper motor neuron signs, a sensory level, and other laboratory and imaging evidence of myelopathy. The most common etiologies of polyradiculopathy are multilevel spondylotic disease of the spine and metastatic involvement of roots and vertebrae, although there are numerous less common causes. In one study of 118 patients who had the electrophysiologic picture of polyradiculopathy without polyneuropathy, 41 had spondylotic disease, 22

1328

Diseases of the Peripheral Nervous System

had neoplastic disease, 26 were undiagnosed, 14 actually had myelopathies, and the rest were single cases of various diagnoses.53 It is also well to recall that acute inflammatory demyelinating polyradiculopathy and chronic inflammatory demyelinating polyradiculopathy (which are considered in detail in Chapters 99 and 100, respectively) are both polyradiculoneuropathies and that sometimes the root involvement predominates.

Spondylotic Disease of the Spine Spondylosis compromises spinal roots by compromising the intervertebral foramina, and typical radicular symptoms and signs can result once foraminal stenosis is sufficiently severe. Spondylosis often produces narrowing of the central spinal canal in addition to compromising the intervertebral foramina. The consequences of central canal stenosis are quite different in the cervical and lumbar spine. Cervical spinal stenosis, if sufficiently severe, compromises the cervical spinal cord, resulting in a cervical myelopathy with corticospinal tract and sensory signs caudal to the stenotic level, and sometimes bladder dysfunction. Typically the cervical central canal stenosis is combined with foraminal stenosis at one or more levels, producing a radiculomyelopathy. The myelopathic signs may develop insidiously and painlessly, and be first discovered when the patient presents with radicular pain. Management of the radiculopathy follows principles similar to those in patients without myelopathy, but the appropriate treatment for the myelopathy adds a new dimension to the patient’s problem. The natural history of cervical spondylotic myelopathy is not well understood, and it is unclear if surgical treatment is superior to conservative management.74 A recent Cochrane review of surgery for cervical spondylotic radiculomyelopathy29 found only one randomized controlled trial; in this study, surgical treatment failed to show any advantage over conservative management with respect to functional outcome.9 Central canal stenosis in the lumbar spine is associated with a quite different syndrome, neurogenic claudication or pseudoclaudication.32 Neurogenic claudication is characterized by intermittent pain, numbness, or weakness in the legs brought on by prolonged standing or walking. Unlike vascular (“true”) claudication, in which symptoms reflect arterial insufficiency, in neurogenic claudication symptoms are related to posture, most often developing when the spine is extended for some length of time. Thus a patient with neurogenic claudication who becomes symptomatic after walking several blocks can usually ride a bicycle (i.e., with the spine flexed) indefinitely. Both activities are equally limited in vascular claudication. Likewise, prolonged standing often precipitates leg symptoms in lumbar spinal stenosis, whereas standing generally brings relief in arterial insufficiency.

Most patients with lumbar spinal stenosis and neurogenic claudication also have significant back pain, but occasionally back pain is minimal or absent. With increasing degrees of injury to the cauda equina, there may be fixed neurologic symptoms and signs, which worsen when the spine is extended. When the L5 and S1 roots are affected bilaterally, the patient’s main complaint is usually distal numbness and weakness, and a misdiagnosis of polyneuropathy may result. In older patients with “polyneuropathy of unknown cause,” lumbosacral polyradiculopathy resulting from spinal stenosis should be considered if there is sparing of the upper limbs, inability to walk on the toes as well as the heels,13 decreased compound muscle action potential (CMAP) amplitudes but relatively preserved SNAPs, or worsening of numbness and weakness with walking or standing.32 A 72-year-old man reported a 2-year history of difficulty walking. He stated this was mostly due to weakness of the ankles, and that both feet were numb. He had had persistent right hip pain since a hip arthroplasty 4 years before, but no other back or limb pain. Examination showed distal weakness in the legs, most severe in the plantar and toe flexors. Vibratory sensation was decreased in the feet, and the knee and ankle reflexes were absent bilaterally. He could walk on the heels but not on the toes. The upper limbs were normal. Nerve conduction studies revealed normal conduction velocities, severely decreased tibial and peroneal CMAP amplitudes, and borderline sural SNAP amplitudes. Biopsy of the gastrocnemius showed signs of partial denervation and reinnervation, but sural nerve biopsy showed essentially normal myelinated fiber density. It was concluded that he had an axonal polyneuropathy, cause unclear. The patient sought a second opinion about the diagnosis. His preserved ability to walk on the heels but not the toes, the disparity between the CMAP and SNAP amplitudes, and the minimal findings on his sural nerve biopsy suggested the possibility of lumbosacral polyradiculopathy. A CT scan showed severe central canal stenosis at L3-L4 and L4-L5 caused by osteoarthritis and ligamentum flavum hypertrophy. Surgery was advised, but the patient declined. Over the next 5 years, he developed increased distal weakness and intermittent pain in the right leg when walking, but continued to refuse surgery.

An uncommon but important situation to recognize is a large midline disc herniation in the lumbar spine below L1 presenting as a catastrophic cauda equina syndrome. Patients present with severe acute back and leg pain, and compromise of sacral nerve root function, particularly the S2-S4 roots subserving bladder and bowel function. Most patients have a history of chronic nonspecific musculoskeletal back pain, but a substantial minority has no history of back pain prior to the acute cauda equina syndrome.76 Typically there is a saddle distribution of sensory disturbance, decreased anal sphincter tone, and a

1329

Diseases of Spinal Roots

reduced or absent anal reflex, although these findings may be difficult to appreciate because of the severity of the pain. Motor weakness and reflex changes in the legs are often minor or absent. A 20-year-old man had a 1-year history of mild low back pain. One morning, while in the shower, his back pain became acutely worse. He fell to the ground and crawled out of the shower for help. He was taken by ambulance to an emergency room, where he was observed for most of the day and was given analgesics. He reported being unable to void; in-andout catheterization was performed, and he was sent home. At home he continued to have back pain and difficulty urinating, and he returned to the emergency room the next morning. Urgent MRI of the lumbosacral spine was performed, revealing a huge midline disc protrusion at L4-L5, with compression of the cauda equina (Fig. 54–2). He was taken immediately to the operating room for laminectomy and discectomy. Six months later he still required intermittent catheterization to empty his bladder, and enemas and manual disimpaction for his bowels. He had erectile impotence and dense sensory loss

in the perineum, the posterior aspect of the thighs, and the soles of the feet. He could walk, but not on his toes. There was some weakness of toe flexors, but strength was otherwise normal. The ankle reflexes were absent. Nerve conduction studies showed decreased tibial motor amplitudes but normal sural and peroneal sensory responses. There was profuse fibrillation in the gastrocnemii, biceps femoris, and gluteus maximus; the tibialis anterior was normal.

Suspicion of this diagnosis demands urgent imaging with MRI or CT and, if the diagnosis is confirmed, immediate surgical consultation. The cauda equina syndrome is perhaps the only absolute indication for surgery in a patient with a lumbar disc herniation, and the need for urgent surgical intervention is acknowledged by most authorities. Although there are no prospective or controlled trials, a meta-analysis of reported cases and series found that all outcome measures were better when surgery was performed within 48 hours of onset of symptoms.3

Neoplastic Disease of Spinal Roots

FIGURE 54–2 T2-weighted magnetic resonance images of the lumbosacral spine. Right, Sagittal image showing a large midline protrusion of the L4-L5 disc producing severe indentation of the thecal sac and posterior displacement of the cauda equina. Left, Axial images at mid-L4 (top), at the L4-L5 disc (middle), and at mid-L5 (bottom). Note that the thecal sac is almost completely obliterated by the L4-L5 disc herniation, whereas the thecal sac has a normal appearance immediately above and below this level.

Neoplastic involvement of spinal roots is second only to spondylotic disease of the spine as a cause of polyradiculopathy. Neoplastic disease may affect spinal roots in several ways, most commonly by leptomeningeal metastasis or by local extension of vertebral metastases or tumor deposits in the paravertebral region. A variety of histologically benign tumors arising intraspinally may compress one or more spinal roots, the cauda equina, or the spinal cord. Examples of such lesions include schwannomas, neurofibromas, meningiomas, and ependymomas; these are discussed in detail in Chapter 116. Rarely, intravascular dissemination of neoplastic cells can present as polyradiculopathy.88 Although leptomeningeal metastasis and local extension from vertebral or paravertebral disease are described separately below, it should be borne in mind that the two often coexist. Leptomeningeal metastasis occurs in about 5% of all cancer patients, the most common primary tumors being breast, lung, melanoma, gastrointestinal, lymphoma, and leukemia.16,22 In children with primary CNS neoplasia, leptomeningeal dissemination is common, particularly with primitive neuroectodermal tumors, anaplastic gliomas, and ependymomas.60 By contrast, in adults meningeal metastasis from primary brain tumors is rare.4 The effects of leptomeningeal metastasis may be manifest in three domains: the cerebral hemispheres, cranial nerves, and spinal roots. Common presenting symptoms include headache, altered mental status, seizures, and gait difficulties (resulting from cerebral involvement); diplopia, facial weakness, and hearing or visual loss (resulting from cranial nerve compromise); or limb weakness or numbness plus back or neck pain. Most patients with leptomeningeal

1330

Diseases of the Peripheral Nervous System

metastasis present with some combination of the above features, but patients with a relatively pure polyradicular involvement are not uncommon.37 Usually patients are known to have a primary neoplasm, but polyradiculopathy can be the neoplasm’s presenting feature, particularly with lymphoma. Radicular pain plus accompanying weakness, reflex loss, and sensory disturbance are found. L5-S1 involvement is most common, although any segmental level can be affected. The single most important test for the diagnosis of leptomeningeal metastasis is CSF cytology. Positive cytology is found with one lumbar puncture in 30% to 60% of patients. The yield rises with a second lumbar puncture, but generally does not increase further with three or more.16 Up to 40% of patients proven to have leptomeningeal metastasis at autopsy have persistently negative CSF cytology ante mortem,30 although other CSF parameters (protein, glucose, and cell count) are almost always abnormal. More specialized techniques such as flow cytometry or cell surface markers can be valuable in certain situations, especially when trying to distinguish between reactive lymphocytes and lymphoma or leukemia cells. Biochemical markers of neoplasia such as carcinoembryonic antigen, ␤-glucuronidase, or ␤-human chorionic gonadotropin are generally regarded as useful adjuncts, and may be helpful when trying to assess response to treatment. Leptomeningeal metastasis may be detected by imaging techniques, of which gadolinium-enhanced MRI is the most sensitive. The imaging signs include enlargement of nerve roots, especially if asymmetrical or nodular, and gadolinium enhancement of the roots and associated meninges. These signs are most likely to be found in the cauda equina. MRI has a significant false-negative rate, but is particularly useful for demonstrating bulky metastatic disease, which is important in the planning of radiation therapy. Treatment of leptomeningeal metastasis is palliative, not curative. Radiation therapy is generally directed at symptomatic regions of the neuraxis, or at bulky disease. Total craniospinal irradiation, which is used prophylactically in childhood leukemia, is associated with major systemic complications in adults with leptomeningeal metastasis, and appears to offer no advantage over focused therapy in control of disease.16 Chemotherapy is usually given intrathecally, via an Ommaya reservoir or by lumbar puncture. Depending on the neoplasm, methotrexate, cytarabine, or thiotepa is used. The overall median survival of patients with leptomeningeal metastasis is about 4 months. Patients with breast cancer survive slightly longer (6 months), and up to a quarter may survive 1 year.22 Complete and sustained remission (over 2 years) may follow treatment in a substantial percentage of lymphoma patients.77 A 78-year-old woman developed urinary symptoms and pelvic discomfort. A mass posterior to the bladder was

found; biopsy revealed noncleaved, diffuse, large cell nonHodgkin’s lymphoma. Treatment with CHOP chemotherapy (three cycles) and pelvic irradiation resulted in complete resolution of the mass and her symptoms. She remained well until 1 year later, when she developed numbness in the feet. Over the next few weeks, the numbness ascended to the knees and posterior thighs, and she had progressive difficulty walking, requiring a cane and then a walker. Concurrent with these symptoms, she had low back pain when standing. There was moderate proximal and distal leg weakness, areflexia in the legs, and sensory loss distal to the knees. The cranial nerves and arms were normal. Motor and sensory nerve conduction studies were normal except for prolonged tibial and peroneal F wave latencies. The patient refused needle EMG. CSF protein was markedly increased, with normal glucose and 8 lymphocytes/␮L. Cytology for malignant cells was negative, and there were insufficient cells for cell surface markers. However, MRI showed gadolinium enhancement of cauda equina roots. The patient received two intrathecal doses of methotrexate, followed by radiation therapy to the lumbar spine. Over the next 2 months there was steady improvement in her legs, and she was able to walk independently again. Examination 1 year later showed only mild proximal leg weakness. Shortly afterward she presented with multiple cranial neuropathies. Two more doses of intrathecal methotrexate were given, but she died a month later.

Spinal metastasis is common in cancer patients, both as a site of spread from a known primary, or as the locus at which the tumor first announces its presence. In men, the most common primary is lung cancer; in women, cancer of the breast. Other tumors with a particular predilection for metastasizing to the spine are multiple myeloma, prostate cancer, and renal cell carcinoma. Spinal roots are involved either by direct invasion of cancer cells or by bony compression consequent to a pathologic fracture. The cardinal symptom of vertebral metastasis is local pain, which may be nonspecific initially but with time acquires a typical deep, unrelenting quality, often most troublesome at night. Nerve root involvement is signaled by radiation of pain into a limb or along thoracic dermatomes, sensory symptoms, and weakness. Rarely, metastatic involvement of spinal nerves is painless. A diagnosis of metastatic radiculopathy can usually be made on clinical grounds alone. In a patient known to have cancer, virtually any symptom in the spine or limb must be regarded with suspicion, with a low threshold for investigation. Plain radiographs of the spine, being simple and readily obtained, are an efficient way to begin investigation (in fact, the only real reason to obtain spine radiographs in any patient with radiculopathy is to look for evidence of vertebral metastasis). Any suspicious areas on the plain films should be imaged in greater detail, usually with MRI. Negative plain films certainly do not exclude

Diseases of Spinal Roots

neoplastic disease; in particular, paravertebral tumor will rarely be evident on plain radiographs. With vertebral metastasis rostral to the L1 vertebral level, the more pressing concern is usually whether or not there is concomitant epidural metastasis with spinal cord compression. Examination may show long tract signs in addition to radicular signs, but examination may be difficult in the face of severe pain. In addition, patients often harbor vertebral metastases at levels other than the most symptomatic. As a consequence, the most expeditious approach is to obtain urgent MRI of the whole spine, with more detailed views of involved levels. The treatment approach to metastatic radiculopathy depends upon whether or not there is also epidural spread with spinal cord compromise. The latter demands urgent treatment with corticosteroids and radiation therapy; surgery for relief of spinal cord compression is generally reserved for specific situations, such as known radiotherapyresistant tumors, or for diagnosis if the primary tumor is not known. Radiation therapy is the main treatment for metastatic root involvement. The presence of bony metastases implies incurable neoplastic disease. The primary goal of radiation for metastatic radiculopathy is pain relief, for which radiation is generally effective. A favorable response to radiation will also include arrest of neurologic worsening, and in some patients neurologic deficits may improve.

Infectious Causes of Polyradiculopathy Polyradiculopathy is a well-established manifestation of infection by Borrelia burgdorferi (the spirochete causing Lyme disease, also known in Europe as Bannwarth’s syndrome) and cytomegalovirus (CMV) (particularly in HIVinfected patients). Epstein-Barr and herpes simplex viruses also may infect spinal nerves, but generally as a component of a more widespread infection of the meninges, spinal cord, or brain. (The term Elsberg syndrome is sometimes used to refer to cauda equina syndrome or polyradiculopathy associated with herpes simplex virus infection, but the term is used inconsistently and is probably best avoided.94) There are also single case reports of polyradiculopathy or polyradiculoneuropathy associated with recent infection by agents such as Brucella,59 Listeria monocytogenes,61 and Chlamydia pneumoniae.54 Whether these single cases are true infections of nervous tissue or are best regarded as parainfectious immune-mediated disorders, analogous to the relationship between Campylobacter jejuni and GuillainBarré syndrome, is uncertain. With infectious polyradiculopathy there is typically increased CSF protein and pleocytosis. Culturing the infectious agent from the CSF is not usually possible, and making a specific diagnosis rests on serologic tests. PCR to detect microbial DNA in CSF is being increasingly used.

1331

Reported cases of Lyme polyradiculopathy in North America resemble Guillain-Barré syndrome, with limb weakness, pain, and sensory symptoms evolving over several days.75 The presence of CSF pleocytosis is a clue to the diagnosis, which is confirmed by serology. The polyradicular presentation of B. burgdorferi infection seems to be more common in Europe, and concomitant cranial neuropathies are often present. The syndrome appears to have a relatively benign natural history, with most patients recovering fully regardless of whether antibiotics are given.41 Spirochetal infection with Treponema pallidum can present as an isolated polyradiculopathy in both HIV-positive and non–HIV-infected patients.19 Involvement of dorsal roots is a key element in one of the classical manifestations of tertiary syphilis, tabes dorsalis.78 CMV polyradiculopathy is usually seen in patients with long-standing HIV infection, but occasionally can be the initial manifestation of acquired immunodeficiency syndrome.6 Rapidly evolving leg weakness and incontinence are the main findings. In addition to the presence of HIV, the diagnosis of CMV acute lumbosacral polyradiculopathy may also be suggested by the CSF, in which the pleocytosis often consists predominantly of neutrophils.55 Definitive diagnosis is by culture or PCR of CSF. Treatment with ganciclovir or foscarnet results in longer survival in HIV-positive patients than does no treatment, but neurologic recovery is generally poor.6 Other infectious causes of acute polyradiculopathy in HIV-positive patients include herpes simplex virus type 2, Toxoplasma gondii, syphilis, or mycobacteria. The differential diagnosis of acute polyradiculopathy in HIV patients also includes Guillain-Barré syndrome and neoplastic involvement of roots (particularly lymphoma).80 Difficult to classify were four previously healthy patients with acute lower motor neuron weakness confined to the legs, and no sensory or sphincter involvement, who were found to be HIV positive, and who recovered spontaneously.10 All other infectious serologies were negative, leading the authors to conclude that the patients had either a new Guillain-Barré syndrome variant or a new manifestation of HIV infection. Infectious processes affecting nerve roots often implicate neighboring tissues such as the spinal cord or meninges. For example, herpes simplex or Epstein-Barr virus infection usually results in a radiculomyelitis or encephalomyeloradiculitis; root involvement is readily demonstrated, but the major clinical manifestations stem from spinal cord or brain dysfunction.42,49,95 Strictly speaking, most examples of infectious polyradiculopathy are properly described as meningoradiculopathies. Sometimes, as with mycobacterial infection, nerve roots are affected, but the clinical picture is dominated by manifestations of intracranial meningeal involvement, such as headache and altered alertness,52 or by signs of myelopathy (tuberculous radiculomyelitis).33,96 Conversely, anterior horn cell infection

1332

Diseases of the Peripheral Nervous System

(i.e., poliomyelitis) may mimic the electrophysiologic and CSF features of polyradiculopathy. Although cases of poliovirus infection are now exceedingly rare, the poliomyelitis syndrome resulting from other viruses continues to occur from time to time (e.g., enterovirus 71 in Taiwan34; West Nile virus14).

Diabetic Polyradiculopathy In patients with diabetes mellitus, there are several peripheral nerve syndromes encountered infrequently that are distinct from the common distal symmetrical polyneuropathy. Perhaps best known is lumbosacral radiculoplexopathy (proximal diabetic neuropathy or diabetic amyotrophy), which is considered in detail in Chapter 86. Like lumbosacral radiculoplexopathy, diabetic thoracic polyradiculopathy (diabetic truncal neuropathy) is characterized by pain and has a self-limited course. The pain in thoracic polyradiculopathy is felt in a portion of the abdominal or chest wall, sometimes in an obviously dermatomal pattern. The pain usually begins abruptly; is unremitting, burning, and often severe; and may be mistaken for visceral pain.48 Patients may report reduced or heightened cutaneous sensation in the painful area, and there are often corresponding sensory signs.82 In most patients there is motor involvement as well. Fibrillation potentials in corresponding thoracic paraspinal muscles can usually be found, and asymmetrical bulging of the abdominal wall as a result of focal weakness of the abdominal musculature may be seen.38 In a diabetic, the clinical picture is so characteristic that little investigation is needed. Herpes zoster may be suspected initially, but this possibility is excluded by the absence of the cutaneous vesicular eruption. In some instances plain radiographs or other imaging of the thoracic spine may be needed to exclude a vertebral metastasis, or there may be long tract signs indicating that the thoracic pain is the segmental sign of an underlying myelopathy.38 The pathogenesis of the syndrome is unclear. One possibility is an immune-mediated inflammatory process, analogous to recent findings in diabetic lumbosacral radiculoplexopathy.26 This idea is supported by the similar profiles of the two syndromes, as well as the simultaneous existence of both in some patients.26 However, there are no published pathologic studies focused on diabetic thoracic radiculopathy. Diabetic thoracic polyradiculopathy runs a self-limited course. In one series the mean duration of pain was 6 months.38 Pain control is the main management issue, and drugs used in other types of neuropathic pain, such as amitriptyline or gabapentin, provide some relief. Topical lidocaine sprays or ointments may also be tried, but narcotic analgesics may be necessary at the height of the symptoms.

Radiation Radiculopathy The development of a progressive myelopathy months or years after radiotherapy to the spine, lumbosacral plexopathy following pelvic irradiation, or brachial plexopathy after radiation to the chest are well-recognized but fortunately uncommon complications of therapeutic radiation. Because spinal roots are included in treatment fields producing these syndromes, the possibility of radiation injury to the spinal roots, either alone or in combination with the spinal cord or plexus, has been discussed for many years. Patients with sensory loss and lower motor neuron weakness in a limb following irradiation are usually thought to have lesions of the brachial or lumbosacral plexus, based principally on electrodiagnostic findings that suggest a postganglionic localization (i.e., absent or reduced SNAPs). However, it is difficult to eliminate the possibility of coexisting spinal root injury, and some favor labeling such syndromes radiculoplexopathies rather than plexopathies.69 A related group of patients are those who develop progressive pure lower motor neuron weakness months to years after radiation, most often to the lumbar spine. Preservation of normal sensory function, on clinical and electrophysiologic grounds, argues for an anterior horn cell localization in these patients. However, the possibility that the process primarily affects ventral roots cannot be entirely excluded. In one such patient examined at autopsy, there was loss of myelin staining and axonal loss in both the ventral and dorsal roots of the cauda equina, despite the lack of any sensory disturbance clinically. Some anterior horn cell chromatolysis was also present.11 The interval between radiation therapy and onset of the neurologic symptoms in both these syndromes ranges from 3 months to 10 years or more. Weakness and sensory loss are painless and slowly progressive; in some patients the process reaches a plateau, and in others it worsens inexorably until the affected limb is useless.

Chronic Arachnoiditis There are interesting parallels between the diagnosis of arachnoiditis and changes in disease patterns and in medical practice over the last century. The term arachnoiditis literally connotes inflammation of the leptomeninges, and was first used to describe the thickened meningeal reaction seen particularly in patients with tuberculous and syphilitic meningitis. The next phase in the term’s history came with the development of myelography using oil-based contrast media such as iophendylate (Pantopaque), which were early recognized to produce an adhesive, inflammatory reaction within the thecal sac. In the last 30 years, less irritative water-soluble contrast agents (e.g., metrizamide) have essentially completely replaced oil-based agents, and examples of myelography-induced arachnoiditis are now

1333

Diseases of Spinal Roots

rarely encountered.79 In fact, extrapolation of current trends suggests that the widespread availability of MRI will lead to the virtual disappearance of myelography in clinical practice. The diagnosis of arachnoiditis is most often made today in a patient with chronic back and limb pain after lumbar spine surgery in whom imaging studies suggest the presence of meningeal thickening or clumping of roots in the cauda equina. However, controversy has developed about the relationship between the imaging and pathologic signs of arachnoiditis and the clinical symptoms attributed to them.66 In many such patients, other plausible explanations for the symptoms exist, and the situation may be complicated by unfulfilled expectations from lumbar spine surgery (the so-called failed back syndrome), or issues of litigation or workers’ compensation. Nevertheless, descriptions of patients with chronic leptomeningeal inflammation unrelated to surgery continue to appear occasionally. Recent examples include arachnoiditis caused by toxoplasmosis18 and subarachnoid hemorrhage,40 as well as those cases in which the cause is unknown.73 In some instances, the meningeal reaction may be accompanied by calcification (arachnoiditis ossificans).27 There is little information on which to base treatment, except that corticosteroids seem to be ineffective and surgery unhelpful except for diagnostic purposes. Chronic leakage of blood into the subarachnoid space, usually from a superficial hemorrhagic tumor or vascular malformation, results in superficial siderosis of the CNS, and may provoke a meningeal inflammatory reaction. The most frequent effects of widespread deposition of hemosiderin are ataxia, deafness, and dementia, but progressive flaccid paraparesis resulting from polyradiculopathy has been described.83

Toxic Meningoradiculopathy Many chemical substances evoke an inflammatory response when introduced into the thecal sac. With the gradual disappearance of myelography from clinical practice, the main situations in which intrathecal injections are now encountered are in spinal anesthesia and with intrathecal administration of antineoplastic or antimicrobial drugs. Neither antineoplastics nor antimicrobials have been implicated as causes of radiculopathy, although given the seriously ill patients who receive such treatments, it is difficult to be entirely certain that an effect on spinal roots is absent. Conversely, in the last decade there has been much discussion about the potential neurotoxic effects of local anesthetic agents used for spinal (intrathecal) anesthesia. A persistent cauda equina syndrome rarely complicates spinal anesthesia; the estimated incidence is about 1 in 10,000. With lidocaine (which offers a greater speed of onset and shorter duration of action than tetracaine or

bupivacaine), the risk of cauda equina syndrome appears to be considerably higher, up to 1 in 1300 with single-injection procedures and 1 in 200 with continuous infusion techniques.36 A related complication of lidocaine spinal anesthesia is the “syndrome of transient neurologic symptoms,” characterized by pain or dysesthesia in the buttocks and legs lasting up to a week after the procedure. This is estimated to complicate 16% to 33% of spinal anesthetics with lidocaine, but is virtually unknown with bupivicaine.36 The mechanism of neurotoxicity is unknown.

Congenital Anomalies Neural tube defects are the most common form of congenital malformation affecting spinal roots, although the routine use of folic acid supplementation during pregnancy has led to a substantial decline in their incidence. In the most severe form, encephalomyelocele, the neurologic deficits primarily involve the brain and spinal cord, and are often associated with death in utero or in infancy. Less extensive anomalies affecting caudal spinal structures produce meningomyelocele, in which a defect in the spinal meninges is accompanied by a variable combination of spasticity, lower motor neuron signs, and sensory loss in the legs, as well as impairment of bladder, bowel, and sexual function. In other words, these patients have a cauda equina syndrome, often with a superimposed thoracolumbar myelopathy. With appropriate surgical and medical management in infancy and early childhood, such patients may live full and productive lives despite their substantial neurologic deficits. The management of meningomyelocele is complex, requiring collaborative expertise drawn from neurology, neurosurgery, orthopedics, and urology. Further consideration of the management of neural tube defects is beyond the scope of this chapter. Most patients with myelomeningocele have stable neurologic deficits, although adaptation to these deficits as the child grows is an important challenge. However, in 3% to 5%, slow but definite worsening of myelopathic or lumbosacral root deficits develops over years. Most patients like this have had surgical repair of their meningeal defect shortly after birth. After some years of stable function, deterioration begins in late childhood or adolescence, and consists of some combination of worsening urologic function, increased spasticity, or increasing weakness and sensory loss. In some patients, the development of intraspinal mass lesions such as lipomas or ependymomas explains the deterioration, but in most no such lesions are found. A commonly observed anomaly in this situation is so-called tethering of the cord, meaning adherence of the conus medullaris or an abnormally short filum terminale to the caudal meninges (usually at the site of surgical closure years before). It is postulated that a tethered cord and cauda equina are subject to gradual mechanical (stretch) injury as the bony elements of the vertebral column

1334

Diseases of the Peripheral Nervous System

lengthen with growth of the child, resulting in gradual worsening of neurologic function. The usual treatment is untethering of the cord, meaning surgical lysis of the adhesions to the meninges. There has been much debate about the effects of surgical treatment of tethered cords. Part of the difficulty stems from the observation that the MRI appearance of tethered cord can also be seen in patients without any signs of worsening function. Rigorous evidence of the efficacy of untethering surgery based on controlled trials does not exist. Conviction that such procedures are helpful is mostly based on observations of arrested progression of neurologic deficits in operated patients, or the natural history of unoperated patients.68 In some instances, progressive neurologic deterioration may be due to the development of central stenosis of the lumbar spinal canal at the site of the original meningocele repair.51 A variation on the above is the adult with the imaging appearance of tethered cord in whom the diagnosis of spinal dysraphism was missed or unsuspected at birth. Such patients may present with back pain or signs of a slowly progressive cauda equina syndrome for the first time in adulthood. Untethering surgery may be performed, although a variety of abnormalities may be encountered, including lipoma of the filum or conus, dermoid cysts, or malformations of the distal spinal cord. Retrospective series of operated patients report that most patients seem to be helped by surgery.35

Miscellaneous Causes of Polyradiculopathy or Cauda Equina Syndrome Although spondylotic disease and neoplasia account for the majority of patients with polyradiculopathy, there are a number of miscellaneous causes, most of which are found in the literature as single case reports. These can be grouped under three broad headings: lesions producing root compression, inflammatory disorders, and ischemic processes. Among the reported miscellaneous causes of compressive polyradiculopathy are pannus formation in rheumatoid arthritis,92 epidural lipomatosis,47 vertebral body expansion resulting from extramedullary hematopoiesis in thalassemia,17 and epidural hematomata following an epidural blood patch23 or after disc surgery.43 In some of these reports it is difficult to be certain if the patients had polyradiculopathy, myelopathy, or both, but surgical decompression generally resulted in improvement of the symptoms and signs. Symptoms and occasionally signs attributable to multiple spinal nerves have been reported in intracranial hypertension from several causes,31,58 as well as in spontaneous intracranial hypotension.5 In these patients, it is assumed that the root dysfunction is a result of nerve root traction.

Systemic inflammatory disorders such as polyarteritis nodosa are often manifested as vasculitic polyneuropathy. Vasculitic polyradiculopathy, in contrast, seems to be almost unknown, though cases of vasculitic polyradiculoneuropathy have been described.12,81 Neurosarcoidosis occasionally presents as a polyradiculopathy or a cauda equina syndrome. Patients may have no other evidence of sarcoidosis, and the discovery of nodular, enhancing nerve roots on imaging studies may lead to a misdiagnosis of meningeal malignancy.1,50 Meningeal biopsy is usually needed to establish the diagnosis. Some patients with sarcoid polyradiculopathy respond well to immunosuppression.87 Patients with ankylosing spondylitis occasionally develop an insidiously progressive cauda equina syndrome.8 Onset is usually decades after the first diagnosis of ankylosing spondylitis, with worsening low back pain, radicular pain, and L5-S1 weakness and sensory loss, plus bowel and bladder incontinence. CSF analysis is usually unrevealing, but imaging of the spine shows striking ectasia of lumbosacral root sleeves. The cause of the syndrome is unclear. One possibility is a chronic arachnoiditis that begins with the onset of ankylosing spondylitis and slowly damages nerve roots and meninges (resulting in ectatic root sleeves). Alternatively, the mechanism may be mechanical injury to nerve roots from the expansile root sleeves. In support of the latter possibility, the only patients in whom neurologic stabilization or improvement has been reported have undergone laminectomy or lumboperitoneal shunting. Patients treated with corticosteroids or conservatively appear to slowly worsen.2 Patients with cauda equina syndrome in the setting of inferior vena cava thrombosis21 or aortic dissection64 have been described and presumably reflect instances of spinal nerve ischemic injury. Disease of the aorta is much more often associated with spinal cord ischemia, and in the acute setting, distinguishing between root and cord compromise may be difficult.

REFERENCES 1. Abrey, L. E., Rosenblum, M. K., and DeAngelis, L. M.: Sarcoidosis of the cauda equina mimicking leptomeningeal malignancy. J. Neurooncol. 39:261, 1998. 2. Ahn, N. U., Ahn, U. M., Nallamshetty, L., et al.: Cauda equina syndrome in ankylosing spondylitis (the CES-AS syndrome): Meta-analysis of outcomes after medical and surgical treatments. J. Spinal Disord. 14:427, 2001. 3. Ahn, U. M., Ahn, N. U., Buchowski, J. M., et al.: Cauda equina syndrome secondary to lumbar disc herniation: a meta-analysis of surgical outcomes. Spine 25:1515, 2000. 4. Alatakis, S., Malham, G. M., and Thien, C.: Spinal leptomeningeal metastasis from cerebral glioblastoma multiforme presenting with radicular pain. Surg. Neurol. 56:33, 2001. 5. Albayram, S., Wasserman, B. A., Yousem, D. M., and Wityk, R.: Intracranial hypotension as a cause of radiculopathy from

Diseases of Spinal Roots

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16. 17.

18.

19.

20.

21.

22. 23.

cervical epidural venous engorgement: case report. AJNR Am. J. Neuroradiol. 23:618, 2002. Anders, H.-J., and Goebel, F.-D.: Cytomegalovirus polyradiculopathy in patients with AIDS. Clin. Infect. Dis. 27:345, 1998. Ashkan, K., Johnson, P., and Moore, A. J.: A comparison of magnetic resonance imaging and neurophysiological studies in the assessment of cervical radiculopathy. Br. J. Neurosurg. 16:146, 2002. Bartleson, J. D., Cohen, M. D., Harrington, T. M., et al.: Cauda equina syndrome secondary to long-standing ankylosing spondylitis. Ann. Neurol. 14:662, 1983. Bednarik, J., Kadanka, Z., Vohanka, S., et al.: The value of somatosensory- and motor-evoked potentials in predicting and monitoring the effect of therapy in cervical spondylotic myelopathy. Spine 24:1593, 1999. Benatar, M. G., and Eastman, R. W.: Human immunodeficiency virus-associated pure motor lumbosacral polyradiculopathy. Arch. Neurol. 57:1034, 2000. Berlit, P., and Schwechheimer, K.: Neuropathological findings in radiation myelopathy of the lumbosacral cord. Eur. Neurol. 27:29, 1987. Boernke, C., Buettner, T., Heye, N., et al.: Lumbosacral radiculitis (Elsberg’s syndrome) with isolated angiitis of the central nervous system. J. Neurol. 243:662, 1996. Bourque, P. R., and Dyck, P. J.: Selective calf weakness suggests intraspinal pathology, not peripheral neuropathy. Arch. Neurol. 47:79, 1990. Leis, A. A., Stokic, D. S., Webb, R. M., et al.: Clinical spectrum of muscle weakness in human West Nile virus infection. Muscle Nerve 28:302, 2003. Carlstedt, T., Anand, P., Hallin, R., et al.: Spinal nerve root repair and reimplantation of avulsed ventral roots into the spinal cord after brachial plexus injury. J. Neurosurg. (Spine) 93:237, 2000. Chamberlain, M. C.: Leptomeningeal metastases: a review of evaluation and treatment. J. Neurooncol. 37:271, 1998. Cianciulli, P., di Toritto, T. C., Sorrentino, F., et al.: Hydroxyurea therapy in paraparesis and cauda equina syndrome due to extramedullary haematopoiesis in thalassaemia: improvement of clinical and haematological parameters. Eur. J. Haematol. 64:426, 2000. Cosan, E. T., Kabukcuglu, S., Arslantas, A., et al.: Spinal toxoplasmic arachnoiditis associated with osteoid formation: a rare presentation of toxoplasmosis. Spine 26:1726, 2001. Darras-Joly, C., Belec, L., and Lortholary, O.: Subacute polyradiculoneuritis revealing syphilis in an HIV-negative man. Clin. Microbiol. Infect. 6:284, 2000. Date, E. S., Mar, E. Y., Bugola, M. R., and Teraoka, J. K.: The prevalence of lumbar paraspinal spontaneous activity in asymptomatic subjects. Muscle Nerve 19:350, 1996. De Kruijk, J., Korten, A., Boiten, J., and Wilmink, J.: Acute cauda equina syndrome caused by thrombosis of the inferior vena cava. J. Neurol. Neurosurg. Psychiatry 67:827, 1999. DeAngelis, L. M.: Current diagnosis and treatment of leptomeningeal metastasis. J. Neurooncol. 38:245, 1998. Diaz, J. H.: Permanent paraparesis and cauda equina syndrome after epidural blood patch for postdural puncture headache. Anesthesiology 96:1515, 2002.

1335

24. Dinneen, S. F., and Buckley, T. F.: Spinal nerve root compression due to monostotic Paget’s disease of a lumbar vertebra. Spine 12:948, 1987. 25. Doi, K., Otsuka, K., Okamoto, Y., et al.: Cervical nerve root avulsion in brachial plexus injuries: magnetic resonance imaging classification and comparison with myelography and computerized tomography myelography. J. Neurosurg. (Spine) 96:277, 2002. 26. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999. 27. Faure, A., Khalfallah, M., Perrouin-Verbe, B., et al.: Arachnoiditis ossificans of the cauda equina. J. Neurosurg. (Spine) 97:239, 2002. 28. Fisher, M. A.: Electrophysiology of radiculopathies. Clin. Neurophysiol. 113:317, 2002. 29. Fouyas, I. P., Statham, P. F., and Sandercock, P. A.: Cochrane review on the role of surgery in cervical spondylotic radiculomyelopathy. Spine 27:736, 2002. 30. Glass, J., Melamed, M., Chernik, N., and Posner, J. B.: Malignant cells in cerebrospinal fluid (CSF): the meaning of a positive CSF cytology. Neurology 29:1369, 1979. 31. Groves, M. D., McCutcheon, I. E., Ginsburg, L. E., and Kyritsis, A. P.: Radicular pain can be a symptom of elevated intracranial pressure. Neurology 52:1093, 1999. 32. Hall, S., Bartleson, J. D., Onofrio, B. M., et al.: Lumbar spinal stenosis: clinical features, diagnostic procedures, and results of surgical treatment in 68 patients. Ann. Intern. Med. 103:271, 1985. 33. Hernandez-Albujar, S., Arribas, J. R., Royo, A., et al.: Tuberculous radiculomyelitis complicating tuberculous meningitis: case report and review. Clin. Infect. Dis. 30:915, 2000. 34. Huang, C.-C., Liu, C.-C., Chang, Y.-C., et al.: Neurologic complications in children with enterovirus 71 infection. N. Engl. J. Med. 341:936, 1999. 35. Iskandar, B. J., Fulmer, B. B., Hadley, M. N., and Oakes, W. J.: Congenital tethered spinal cord syndrome in adults. J. Neurosurg. 88:958, 1998. 36. Johnson, M. E.: Potential neurotoxicity of spinal anesthesia with lidocaine. Mayo Clin. Proc. 75:921, 2000. 37. Kaplan, J. G., Portenoy, R. K., Pack, D. R., and DeSouza, T.: Polyradiculopathy in leptomeningeal metastasis: the role of EMG and late response studies. J. Neurooncol. 9:219, 1990. 38. Kikta, D. G., Breuer, A. C., and Wilbourn, A. J.: Thoracic root pain in diabetes: the spectrum of clinical and electromyographic findings. Ann. Neurol. 11:80, 1982. 39. Kikuchi, S., Sato, K., Konno, S., and Hasue, M.: Anatomic and radiographic study of dorsal root ganglia. Spine 19:6, 1994. 40. Kok, A. J. M., Verhagen, W. I. M., Bartels, R. H. M. A., et al.: Spinal arachnoiditis following subarachnoid haemorrhage: report of two cases and review of the literature. Acta Neurochir. 142:795, 2000. 41. Kruger, H., Reuss, K., Pulz, M., et al.: Meningoradiculitis and encephalomyelitis due to Borrelia burgdorferi: a followup study of 72 patients over 27 years. J. Neurol. 236:322, 1989. 42. Kueker, W., Schaade, L., Ritter, K., and Nacimiento, W.: MRI follow-up of herpes simplex virus (type 1) radiculomyelitis. Neurology 52:1102, 1999.

1336

Diseases of the Peripheral Nervous System

43. Lagares, A., Gonzalez, P., Rivas, J. J., et al.: Epidural haematoma after lumbar disc surgery causing radiculopathy. Acta Neurochir. 141:1239, 1999. 44. Levin, K. H.: L5 radiculopathy with reduced superficial peroneal sensory responses: intraspinal and extraspinal causes. Muscle Nerve 21:3, 1998. 45. Levin, K. H., Maggiano, H. J., and Wilbourn, A. J.: Cervical radiculopathies: comparison of surgical and EMG localization of single-root lesions. Neurology 46:1022, 1996. 46. Linden, D., and Berlit, P.: Comparison of late responses, EMG studies, and motor evoked potentials (MEPs) in acute lumbosacral radiculopathies. Muscle Nerve 18:1205, 1995. 47. Lisai, P., Doria, C., Crissantu, L., et al.: Cauda equina syndrome secondary to idiopathic spinal epidural lipomatosis. Spine 26:307, 2001. 48. Longstreth, G. F.: Diabetic thoracic polyradiculopathy: ten patients with abdominal pain. Am. J. Gastroenterol. 92:502, 1997. 49. Majid, A., Galetta, S. L., Sweeney, C. J., et al.: Epstein-Barr virus myeloradiculitis and encephalomyeloradiculitis. Brain 125:159, 2002. 50. Martin, A. J., Connor, S. E. J., Bodi, I., and Thomas, N. W. M.: Cauda equina sarcoidosis: brief report of a special case. Acta Neurochir. 143:739, 2001. 51. Martinez-Lage, J.-F., Piqueras, C., and Poza, M. Lumbar canal stenosis: a cause of late neurological deterioration in patients with spina bifida. Surg. Neurol. 55:256, 2001. 52. Masjuan, J., Corral, I., and Fernandez-Ruiz, L. C.: Mycobacterial acute lumbosacral polyradiculopathy as the initial manifestation of AIDS. J. Acquir. Immune. Defic. Syndr. 15:175, 1997. 53. McGonagle, T. K., Levine, S. R., Donofrio, P. D., and Albers, J. W.: Spectrum of patients with EMG features of polyradiculopathy without neuropathy. Muscle Nerve 13:63, 1990. 54. Michel, D., Antoine, J. C., Pozzetto, B., et al.: Lumbosacral meningoradiculitis associated with Chlamydia pneumoniae infection. J. Neurol. Neurosurg. Psychiatry 55:511, 1992. 55. Miller, R. F., Fox, J. D., Thomas, P., et al.: Acute lumbosacral polyradiculopathy due to cytomegalovirus in advanced HIV disease: CSF findings in 17 patients. J. Neurol. Neurosurg. Psychiatry 61:456, 1996. 56. Nardin, R. A., Patel, M. R., Gudas, T. F., et al.: Electromyography and magnetic resonance imaging in the evaluation of radiculopathy. Muscle Nerve 22:151, 1999. 57. Nelson, D. A., and Landau, W. M.: Intraspinal steroids: history, efficacy, accidentality, and controversy with review of United States Food and Drug Administration reports. J. Neurol. Neurosurg. Psychiatry 70:433, 2001. 58. Obeid, T., Awada, A., Mousali, Y., et al.: Extensive radiculopathy: a manifestation of intracranial hypertension. Eur. J. Neurol. 7:549, 2000. 59. Oliveri, R., Matera, G., Foca, A., et al.: Polyradiculopathy with cerebrospinal fluid albuminocytological dissociation due to neurobrucellosis. Clin. Infect. Dis. 23:833, 1996. 60. Packer, R. J., Siegel, K. R., Sutton, L. R., et al.: Leptomeningeal dissemination of primary central nervous system tumors of childhood. Ann. Neurol. 18:217, 1985. 61. Padovan, C. S., Liebetrau, M., Danek, A., et al.: Polyradiculitis, perimyocarditis, and nephrotic syndrome: unusual manifesta-

62.

63. 64.

65.

66.

67.

68.

69. 70.

71.

72.

73.

74.

75.

76. 77.

78. 79.

80.

81.

tions of infection due to Listeria monocytogenes. Clin. Infect. Dis. 28:152, 1999. Paquette, S., Lach, B., and Guiot, B.: Lumbar radiculopathy secondary to gouty tophi in the filum terminale in a patient without systemic gout: case report. Neurosurgery 46:986, 2000. Parke, W. W., and Watanabe, R.: The intrinsic vasculature of the lumbosacral spinal nerve roots. Spine 10:508, 1985. Patel, N. M., Noel, C. R., and Weiner, B. K.: Aortic dissection presenting as an acute cauda equina syndrome: a case report. J. Bone Joint Surg. Am. 84:1430, 2002. Persson, L. C. G., Carlsson, C.-A., and Carlsson, J. Y.: Longlasting cervical radicular pain managed with surgery, physiotherapy, or a cervical collar. Spine 22:751, 1997. Petty, P. G., Hudgson, P., and Hare, W. S. C.: Symptomatic lumbar spinal arachnoiditis: fact or fallacy? J. Clin. Neurosci. 7:395, 2000. Phillips, L. H., and Park, T. S.: Electrophysiologic mapping of the segmental anatomy of the muscles of the lower extremity. Muscle Nerve 14:1213, 1991. Phuong, L., Schoeberl, K. A., and Raffel, C.: Natural history of patients with tethered cord in meningomyelocele. Neurosurgery 50:989, 2002. Pradat, P.-F., Poisson, M., and Delattre, J.-Y.: Neuropathies radiques. Rev. Neurol. (Paris) 150:664, 1994. Radhakrishnan, K., Litchy, W. J., O’Fallon, W. M., and Kurland, L. T.: Epidemiology of cervical radiculopathy: a populationbased study from Rochester, Minnesota, 1976 through 1990. Brain 117:325, 1994. Ratliff, J., Nguyen, T., and Heiss, J.: Root and spinal cord compression from methylmethacrylate vertebroplasty. Spine 26:E300, 2001. Riew, K. D., Yin, Y., Gilula, L., et al.: The effect of nerve-root injections on the need for operative treatment of lumbar radicular pain: a prospective, randomized, controlled, double-blind trial. J. Bone Joint Surg. Am. 82:1589, 2000. Rossetti, A. O., Meagher-Villemure, K., Vingerhoets, F., et al.: Eosinophilic aseptic meningitis: a neurological complication in HIV-negative drug-addicts. J. Neurol. 249:884, 2002. Rowland, L. P.: Surgical treatment of cervical spondylotic myelopathy: time for a controlled trial. Neurology 42:5, 1992. Scelsa, S. N., Herskovitz, S., and Berger, A. R.: A predominantly motor polyradiculopathy of Lyme disease. Muscle Nerve 19:780, 1996. Shapiro, S.: Medical realities of cauda equina syndrome secondary to lumbar disc herniation. Spine 25:348, 2000. Siegal, T., Lossos, A., and Pfeffer, M. R.: Leptomeningeal metastases: analysis of 31 patients with sustained off-therapy response following combined-modality therapy. Neurology 44:1463, 1994. Simon, R. P.: Neurosyphilis. Arch. Neurol. 42:606, 1985. Skalpe, I. O.: Adverse effects of water-soluble contrast media in myelography, cisternography and ventriculography. Acta Radiol. 355:359, 1977. So, Y. T., and Olney, R. K.: Acute lumbosacral polyradiculopathy in acquired immunodeficiency syndrome: experience in 23 patients. Ann. Neurol. 35:53, 1994. Stefurak, T. L., Midroni, G., and Bilbao, J. M.: Vasculitic polyradiculopathy in systemic lupus erythematosus. J. Neurol. Neurosurg. Psychiatry 66:658, 1999.

Diseases of Spinal Roots 82. Stewart, J. D.: Diabetic truncal neuropathy: topography of the sensory deficit. Ann. Neurol. 25:233, 1989. 83. Straube, A., Dudel, C., Wekerle, G., and Klopstock, T.: Polyradiculopathy in the course of superficial siderosis of the CNS. J. Neurol. 248:63, 2001. 84. Tarlov, I. M.: Spinal perineurial and meningeal cysts. J. Neurol. Neurosurg. Psychiatry 33:833, 1970. 85. Tong, H. C., Haig, A. J., and Yamakawa, K.: The Spurling test and cervical radiculopathy. Spine 27:156, 2002. 86. Tsao, B. E., Levin, K. H., and Bodner, R. A.: Comparison of surgical and electrodiagnostic findings in single root lumbosacral radiculopathies. Muscle Nerve 27:60, 2003. 87. Verma, K. K., Forman, A. D., Fuller, G. N., et al.: Cauda equina syndrome as the isolated presentation of sarcoidosis. J. Neurol. 247:573, 2000. 88. Viali, S., Hutchinson, D. O., Hawkins, T. E., et al.: Presentation of intravascular lymphomatosis as lumbosacral polyradiculopathy. Muscle Nerve 23:1295, 2000. 89. Vroomen, P. C., de Krom, M. C., Wilmink, J. T., et al.: Lack of effectiveness of bed rest for sciatica. N. Engl. J. Med. 340:418, 1999. 90. Vroomen, P. C., de Krom, M. C., Wilmink, J. T., et al.: Diagnostic value of history and physical examination in

91. 92.

93.

94.

95.

96.

97.

1337

patients suspected of lumbosacral nerve root compression. J. Neurol. Neurosurg. Psychiatry 72:630, 2002. Weber, H.: Lumbar disc herniation: a controlled, prospective study with ten years of observation. Spine 8:131, 1983. White, K. P., and Harth, M.: Lumbar pannus presenting as cauda equina syndrome in a patient with longstanding rheumatoid arthritis. J. Rheumatol. 28:627, 2001. Wilbourn, A. J., and Aminoff, M. J.: AAEM minimonograph 32: the electrodiagnostic examination in patients with radiculopathies. Muscle Nerve 21:1612, 1998. Wildhagen, D., Kuntz, R., Kermer, P., et al.: Elsberg syndrome due to infarction of the conus medullaris associated with a prothrombin mutation. J. Neurol. 246:507, 1999. Wiley, C. A., VanPatten, P. D., Carpenter, F. M., et al.: Acute ascending necrotizing myelopathy caused by herpes simplex virus type 2. Neurology 37:1791, 1987. Woolsey, R. M., Chambers, T. J., Chung, H. D., and McGarry, J. D.: Mycobacterial meningomyelitis associated with human immunodeficiency virus infection. Arch. Neurol. 45:691, 1988. Zambelis, T., Piperos, P., and Karandreas, N.: Fibrillation potentials in paraspinal muscles in chronic lumbosacral radiculopathy. Acta Neurol. Scand. 105:314, 2002.

55 Brachial Plexus Lesions ASA J. WILBOURN

Anatomy Anatomic Regions Brachial Plexopathies Anatomic Vulnerability Demographics Characteristics Clinical Features

Clinical Presentations Assessment Specific Etiologies of Brachial Plexopathy Traumatic Supraclavicular Plexopathies Nontraumatic Supraclavicular Plexopathies Iatrogenic Supraclavicular Plexopathies

The brachial plexus (BP) (literal translation: arm braid, or the interweaving of strands) innervates essentially all of the upper limb and most of the shoulder girdle. It consists of a network of 102,000 to 166,000 interlacing axons that has the form of a skewed triangle, with its apex laterally and inferiorly in the axilla, and its base oriented vertically near the midline. Compared to other peripheral nervous system (PNS) structures, the BP is one of the largest, one of the most complex in its gross anatomy, and the most intricate in its internal topography.78

ANATOMY The BP, according to anatomy texts, consists of five components: (1) five roots, (2) three trunks, (3) six divisions, (4) three cords, and (5) several terminal nerves78 (Fig. 55–1). Despite this stereotyped depiction, its first and fifth components are controversial or confusing.78,80 A description of each of these follows, beginning with the most proximal. The first components are the C5 through T1 “roots” or “roots of the brachial plexus.” Anatomists and most clinicians disagree regarding the extent of these structures. To anatomists, the “roots” consist solely of the ventral rami (VR), which are formed when the mixed spinal nerves exit the intervertebral foramina and terminate as VR and dorsal rami.78 Thus to them the BP is strictly extraforaminal. Conversely, most clinicians envision the BP encompassing not only the VR, but also the C5 through T1 ventral and

Traumatic Infraclavicular Plexopathies Nontraumatic Infraclavicular Brachial Plexopathies Iatrogenic Infraclavicular Plexopathies Thoracic Outlet Syndromes

dorsal primary roots (arising from the spinal cord) and the mixed spinal nerves (formed when the primary roots fuse). Consequently, to them the roots are more extensive, and the BP also has intraspinal canal and intervertebral foramina components. Using this construct, root avulsions are brachial plexopathies, just as are trunk and cord lesions.8,35 In any case, it is obvious that the term roots, as found in both anatomy texts and clinical reports, is anatomically inaccurate because the actual roots—the ventral and dorsal primary roots—are located solely within the intraspinal canal, where they link the spinal cord and the mixed spinal nerves. The extraforaminal portions of the “roots,” or the VR, are situated between the anterior and middle scalene muscles, deep to the sternocleidomastoid. Soon after their origin, the VR receive postganglionic sympathetic nerve fibers, via gray rami communicans, from the cervical sympathetic chain—specifically, the middle cervical and stellate ganglia. (The latter is a fusion of the inferior cervical and first thoracic ganglia.) These postganglionic fibers originate from the T3 (or T4) to T6 segments of the spinal cord, and after traversing the BP and subsequently peripheral nerves, they supply the sweat glands, piloerector muscles, and arterioles in the upper limb. In addition, the (C8), T1, and T2 VR send preganglionic sympathetic fibers, via white rami communicans, to the stellate ganglion. After entering it, these pass rostral in the sympathetic chain to the superior cervical ganglion, where they synapse with postganglionic neurons that supply the sweat glands and arteries of the head, the superior tarsal muscle 1339

1340

Diseases of the Peripheral Nervous System

FIGURE 55–1 The brachial plexus (left side; frontal view) and its relationship to nearby bony structures (the overlying clavicle is outlined). Each of the five components of the brachial plexus is encircled. Note that the “roots” are actually the extraforaminal ventral rami, not the primary roots. Certain nerves (e.g., the medial and lateral pectoral nerves) are not labeled.

of the eyelid, and the dilator muscle of the pupil.29,57,83 Motor branches arising from the VR include (1) the accessory phrenic nerve (to the diaphragm), a branch to the subclavius, and the dorsal scapular nerve (to the rhomboids), all from the C5 ventral ramus alone; (2) the long thoracic, or the nerve to the serratus anterior, from C5-C7 VR; and (3) branches to the scalenes and longus coli, from C5-C8 VR.78 The second components are the three trunks, which begin at the lateral edge of the scalene muscles. The C5 and C6 VR fuse to form the upper trunk, the C7 ventral ramus continues as the middle trunk, and the C8 and T1 VR join to form the lower trunk. The trunks are situated in the anteroinferior portion of the posterior triangle of the neck. Portions of them are relatively superficial. The lower trunk is also contiguous with the subclavian artery and the apex of the lung. From the upper trunk, the suprascapular nerve arises that innervates the spinati muscles (the infraspinatus and supraspinatus). No branches originate from the middle or lower trunks.78 The third components are the six divisions, which result when each trunk divides into an anterior and a posterior division. The divisions are located between the first thoracic rib and the clavicle (midportion). Although no branches arise from them, they are important for two reasons. First, at this level the BP undergoes a fundamental internal reorganization: The motor fibers that will supply the extensor and the flexor muscles of the shoulder and upper limb separate from one another. Second, because the clavicle overlies the divisions whenever the upper limb

is at the side, the BP can be considered as two separate regions: supraclavicular and infraclavicular. (Granted, it would be more logical to refer to these two regions as supradivisional and infradivisional.) This segmentation is clinically valuable (see later).78,80 The fourth components are the cords. The anterior divisions that derive from the upper and middle trunks unite to form the lateral cord, the one that arises from the lower trunk continues as the medial cord, and all three posterior divisions fuse to form the posterior cord. The cords, the longest components of the BP, are named for their relationship to the second segment of the axillary artery, to which they run parallel. They are in the proximal axilla, and give rise to the major limb nerves and some of the cutaneous nerves: from the lateral cord, the lateral pectoral and musculocutaneous nerves, and the lateral head of the median nerve; from the medial cord, the medial pectoral, medial brachial cutaneous, medial antebrachial cutaneous (MAC), and ulnar nerves and the medial head of the median nerve; and from the posterior cord, the subclavian, thoracodorsal, axillary, and radial nerves.78,80 The fifth and last components are the terminal nerves. Two aspects of these are confusing in anatomy texts. First, there is no consensus regarding their number. All agree that the median, ulnar, and radial nerves merit this designation; some also include the axillary, the musculocutaneous, or both. In this chapter, all five are considered “terminal nerves.” Second, their exact distal extent is never stated; that is, where do they end and the most proximal

Brachial Plexus Lesions

portions of the major nerve trunks begin? These are arbitrary boundaries because there are no anatomic dividing lines. Among both anatomists and clinicians, apparently only one surgeon has addressed this issue, and even his definition is imprecise: those portions of the peripheral nerves “at their origin and a few centimeters below” are terminal nerves (i.e., are the fifth components of the BP).78 For practical purposes, damage to any of these that occurs in the axilla probably is legitimately classified as a BP lesion. Major anomalies of BP anatomy are relatively rare. The most commonly mentioned anomaly concerns the root origins of the BP fibers. At times, these are shifted rostrally or caudally by one root, for example, a major contribution from C4 and none from T1 (pre-fixed plexus), or a major contribution from T2 with none from C5 (post-fixed plexus). At other times, the root origins appear to expand, rather than shift, in that both the C4 and T1 roots contribute. More distally, the medial and lateral cords may fuse into a single anterior cord, or the anterior division of C7 may supply the medial, as well as the lateral, cord.3,84 The importance of these various anomalies is unclear, particularly because they generally have no effect on the segmental innervation of the branches of the BP. Moreover, how common they are is questionable. Although one anatomist reported that 62% of 175 BPs he examined contained some fibers from C4 (i.e., were pre-fixed), a BP surgeon observed such root anomalies in only 2% of 655 operated patients.78 The vascular supply of the BP is provided by the subclavian artery and its branches. The supraclavicular elements are supplied as follows: (1) roots—vertebral, inferior thyroid, deep cervical, superior intercostal, and costal cervical arteries; and (2) trunks—transverse cervical, suprascapular, subclavian, and dorsal scapular arteries. The infraclavicular elements are supplied by (1) medial and lateral cords—axillary and thoracoacromial arteries; (2) posterior cord—axillary and subscapular arteries; and (3) terminal nerves—several peripheral arteries.78,84

1341

icular. First, the lesions that affect these two BP areas tend to manifest distinctly different clinical presentations, because the supraclavicular plexus largely retains a segmental arrangement, whereas the infraclavicular plexus does not. Second, many disorders have a marked preference for just one of them. Thus obstetric palsy characteristically involves solely the supraclavicular plexus. Third, traumatic lesions at these two levels differ significantly in their incidence, severity, and prognosis; for example, supraclavicular plexopathies are much more common, they more frequently require operation, and they are more likely to leave permanent residuals.78,80 The supraclavicular plexus is further subdivided, both vertically and longitudinally. Vertically, it is separated into upper, middle, and lower plexuses, based on the roots from which the axons originate: C5/6; C7; or C8, T1. An upper plexus lesion, therefore, may affect the C5 or C6 primary roots, the upper trunk, or their intervening mixed spinal nerves and VR (see Fig. 55–2). These subdivisions are quite beneficial clinically, because on initial evaluation it is often difficult to determine exactly which supraclavicular elements are damaged, particularly when the etiology was violent trauma. Moreover, these terms, although admittedly imprecise, facilitate communication. Two pertinent points can be made regarding this classification: (1) the “lower plexus” refers to the inferior elements of the supraclavicular plexus, rather than to the infraclavicular plexus, and (2) upper and lower plexus lesions sometimes are designated Erb’s palsy and Klumpke’s palsy, respectively. Lengthwise, the supraclavicular plexus is separated into preganglionic (intradural) and postganglionic (extradural; extraforaminal) segments. Even though these terms strictly apply only to sensory fibers and their relationship to their dorsal root ganglia (DRG)—which are situated at the distal end of the primary sensory roots— by convention they are used to describe lesions affecting

ANATOMIC REGIONS The BP is subdivided by clinicians into various regions. Although seldom mentioned in anatomy texts, these are extremely useful. The most widely employed is that mentioned previously, which essentially bisects the longitudinal extent of the BP using the clavicle (or, more accurately, the divisions of the BP situated deep to the clavicle) as the line of demarcation. Thus the supraclavicular plexus is composed of primary roots, mixed spinal nerves, VR, and trunks, whereas the infraclavicular plexus consists of cords and terminal nerves. The divisions themselves constitute a subclavicular or retroclavicular segment, which is mentioned infrequently because isolated lesions of it are rare (Fig. 55–2). There are several reasons for designating brachial plexopathies as either supraclavicular or infraclav-

FIGURE 55–2 Anatomic regions of the brachial plexus, including its division into supraclavicular, retroclavicular, and infraclavicular portions, and subdivisions of the supraclavicular plexus into upper, middle, and lower plexus. (Cords markedly foreshortened in this illustration.)

1342

Diseases of the Peripheral Nervous System

reconstitute the degenerated axon segments. In addition, the nerve regeneration time may be so prolonged that denervated muscles degenerate.57 The BP, unfortunately, is a vulnerable structure. It may be injured at many sites, frequently because of disorders involving neighboring structures (e.g., lung, subclavian and axillary arteries, glenohumeral joint, proximal humerus, and lymph node chains). Throughout its extent, it is near the highly mobile structures of the neck and shoulder and thus is at substantial risk of sustaining traction injuries.78

Demographics

FIGURE 55–3 The difference between a preganglionic avulsion and a postganglionic rupture of plexus fibers. The peripheral sensory (but not motor) fibers remain intact with the preganglionic lesion because they maintain continuity with their cell bodies in the dorsal root ganglia. (Intact nerve, solid line; degenerated nerve, dashed line.)

motor as well as sensory nerve fibers occurring either within the intraspinal canal (i.e., primary root lesions) or external to it (e.g., trunk lesions), respectively (Fig. 55–3).78,80 In contrast to the supraclavicular plexus, the only subdivision for the infraclavicular plexus, beyond the obvious one of cords and terminal nerves, is to view it as two planes: anterior and posterior. The anterior plane contains the anterior divisions, the medial and lateral cords, and the terminal nerves that originate from them; the posterior plane consists of the posterior divisions, the posterior cord, and the latter’s outflow. The motor, sensory, and autonomic fibers in each plane are totally independent of those in the other. Moreover, the anterior plane is not constant in its composition, principally because the anterior division of C7 (derived from the middle trunk) sometimes contributes to the medial, as well as to the lateral, cord.3

BRACHIAL PLEXOPATHIES Anatomic Vulnerability “Injuries of the brachial plexus are the worst of all peripheral nerve lesions.”8 Because they affect axons so close to their cell bodies, the latter may undergo retrograde degeneration. Moreover, the metabolic demands on the cell bodies may be excessive, because they are responsible for producing all the material that is transported distally to

Most of the published statistics regarding BP lesions are either skewed in their study population or controversial. Concerning selection bias, nearly all the reported series of unquestioned brachial plexopathies have focused on a specific subset (e.g., obstetric paralysis), or have originated from one of the relatively few civilian centers that specializes in their surgical management. However, such BP lesions are not representative of those in the general medical community, for two reasons. First, almost all are severe and persistent, because most of the patients with them are referred, usually several months after injury; those BP lesions that improved substantially or resolved in the interim were excluded.78 Second, they are heavily weighted with traumatic injuries. Yet, among patients with suspected BP lesions referred for magnetic resonance imaging (MRI), the ratio of nontraumatic to traumatic is more than 3:1.11 Regarding controversy, one professed BP lesion, disputed neurologic thoracic outlet syndrome (N-TOS), presents unique difficulties in data accumulation. Unlike every other type, its very existence is questioned by many physicians. Moreover, if it is accepted as an actual disorder then, based on its reputed high incidence, it is the most common BP lesion by far, virtually eclipsing all other types combined. Hence its inclusion or rejection significantly influences the statistics.78 Because of this, disputed N-TOS is excluded in the discussion of demographics that follows. Age Although BP lesions may affect persons of any age, there are three peaks or plateaus. First, there is a very sharp peak at birth because of obstetric paralysis. Second, there is a very high plateau, from trauma, between 18 and 30 years of age. Finally, a far more prolonged elevation is caused by cancer, radiation, and various types of trauma, particularly shoulder fractures and dislocations, beginning at age 50 or so. Patient age is of considerable importance with regard to those traumatic lesions in which surgical repair is contemplated. Because operative results tend to be so much better in the young, most surgeons will operate on children much later after lesion onset than they would in adults, and they will attempt repair of BP elements (e.g., lower trunk) that they would not consider repairing in adults.8,35,71,78

Brachial Plexus Lesions

Sex The relative incidence of men versus women having BP lesions varies with age. For obstetric paralysis there is mostly a 1:1 ratio. Conversely, in adults, the reported male:female ratio has been at least 8:1 or 9:1, because most derive from either military hospitals or specialty centers, both overwhelmingly populated by young men. At the latter, most BP lesions are attributable to closed traction, often caused by motorcycle accidents. Hence, among several large series, 88% to 97% of patients have been males.71,78 More balanced data were provided in a report that dealt with severe BP lesions caused not only by violent traction, but also by gunshot wounds, lacerations, tumors, and various iatrogenic injuries. Only a slight majority (57%) of the 171 patients were male.78 Laterality In obstetric paralysis, the right BP is affected more often than the left (56% to 61% of the time). In adults, relatively little information is available on this important point. Among four large series of adult traction injuries, the right:left involvement was essentially 1:1 in two reports, whereas the right, or at least the “dominant,” limb was affected in 64% and 79% of cases in the other two.71,78 Characteristically, these are unilateral lesions. Thus, among four large series, obstetric paralysis was bilateral in only 2.9% to 6% of cases. In adults, bilateral involvement probably is even less common; in three series containing 102 to 180 patients with primarily traction injuries, it was found in 0 to 2.3%. Nonetheless, occasionally it occurs with some low-frequency BP disorders, such as classic postoperative paralysis, post–median sternotomy lesions, “pack palsy,” and neuralgic amyotrophy.78 Incidence No information is available regarding how often brachial plexopathies occur in the general population, but there are some data concerning their incidence relative to other PNS lesions. Over a 6-year period, PNS lesions were found in 2301 outpatients seen in the neurologic and neurosurgical clinics at a major European university; of those, 232 (10%) involved the BP. Of the 1671 upper limb PNS lesions, 14% involved the BP. Far more information is available regarding the relative incidence of BP lesions sustained during wartime. In several large series reported from both World Wars I and II, they accounted for approximately 4% to 9% of all PNS lesions, and 6% to 12% of upper limb PNS lesions.78

Characteristics Mechanism of Injury Although many injurious agents can damage the BP, they act through a limited number of mechanisms. Traction (stretch) is the most common, especially when severe lesions are considered. Depending upon the rapidity, the

1343

angle, and particularly the amount of applied force, traction can be classified as low or high energy (velocity), can cause any degree of nerve fiber injury, and can affect any portion of the BP. The severe closed supraclavicular traction injuries that predominate in most large series are attributable to high-energy stretching forces applied over very brief periods. However, unless some other structure is first torn, dislocated, or broken, traction most often involves the supraclavicular, as opposed to the infraclavicular, plexus. The BP elongates over its length during movements of the cervical spine, shoulder girdle, and upper limb. It becomes vulnerable to injury whenever its longitudinal excursion exceeds about 15 mm. Proximally, its anchor points vary. The C5, C6, and, variably, the C7 mixed spinal nerves, immediately after they exit the intervertebral foramina, are firmly bound to the gutter of the transverse processes by epidural sheaths, cervical fascia, and fibrous tissue. The C8 and T1 roots lack these attachments, so their proximal anchoring point is the spinal cord itself. Distally, the supraclavicular plexus is relatively fixed at the level of the clavicle to the surrounding structures, in a fascial sheath. Even more distally, terminal nerves serve as anchoring sites for the infraclavicular plexus elements; for example, the musculocutaneous nerve is firmly attached to the coracobrachialis muscle.56,78 The principal causes for traction injuries are traumatic (e.g., vehicular accidents) and iatrogenic (e.g., birth injuries). Most traction injuries are closed in nature, but a few are open (e.g., gunshot wounds). “Contusion” or bruising is also a common mechanism of injury. It usually leaves the nerves in gross continuity, but may produce severe intraneural damage. Many injuries in continuity result from a combination of traction and contusion and are referred to as stretch/contusion lesions. Gunshot wounds frequently produce damage of this type. Although compression is a common cause of focal peripheral nerve lesions, it has a much smaller role in regard to BP disorders. Both supraclavicular (e.g., from pack straps) and infraclavicular (e.g., from crutch use) plexopathies are attributed to it. With some of these, however, traction may be a more important injurious mechanism than compression. In addition, the deleterious effects of compression may be indirect, by its compromising the blood supply to BP elements, thereby causing ischemic damage. Lacerations affect BP fibers rather infrequently; they are categorized as clean (“tidy”) or ragged, depending upon the type of injury they produce. Ischemia damages the BP mainly in the context of compartment syndromes and radiation; with the former it is caused by compromise of the nerve microcirculation by increased pressure, whereas with the latter it probably results from progressive fibrosis, with obliteration of blood vessels. Those BP injuries (usually ischemic, secondary to compartment syndromes) resulting from primary damage to major blood vessels are called neurovascular lesions.78

1344

Diseases of the Peripheral Nervous System

Degree of Nerve Fiber Injury The classifications of focal peripheral nerve injuries proposed by Seddon and Sunderland are applicable to the BP as well as to peripheral nerves. Although nerve lesions seldom are solely of one type (i.e., different axons may sustain different degrees of damage), often one type predominates.78 These are discussed under two headings: demyelinating conduction block (CB) and axon loss (AL). Demyelinating CB (Neurapraxia; First-Degree Injury). The most benign type of clinically recognizable focal nerve disorder is demyelinating CB; although axon continuity remains intact at the lesion site, the myelin damage there prevents impulses from being transmitted across it. The nerve segment distal to the lesion is unaffected, so its electrical excitability remains normal, a fact demonstrable by electrodiagnostic (EDX) examinations, specifically by nerve conduction studies (NCSs). Demyelinating CB is the underlying pathophysiology with many BP lesions, particularly those of abrupt onset caused by moderate traction or compression. Very early after injury, other types of CB can be encountered, caused by ischemia (which usually persists less than a day) and by AL (labeled an axon discontinuity CB), when the latter type is assessed with NCSs so early after injury—less than 6 to 9 days— that the degenerating distal stump fibers can still conduct impulses.81 The weakness caused by CB is indistinguishable from that produced by AL, so early clinical assessments might be quite misleading regarding the extent and severity of the lesion. Thus, immediately following injury, what ultimately will be partial brachial plexopathies may present initially as total lesions. Within a few days or weeks, however, the CB components resolve, leaving more persistent deficits in the distribution of the elements that have undergone AL. In contrast to the clinical assessment, EDX assessments, at least concerning the NCS components, are quite reliable after 7 to 10 days.81 Regarding specific etiologies, demyelinating CB frequently is found with brachial plexopathies caused by vehicular accidents, falls, childbirth, pack use, gunshot wounds (near miss), and certain operative procedures. It is the predominant mechanism in classic postoperative paralysis and in many post–median sternotomy lesions. Conversely, it rarely occurs with low-energy penetrating injuries (e.g., stab wounds) or with compartment syndromes.44,78 Demyelinating CB lesions affect only largediameter nerve fibers. Most are of abrupt onset and short lived, so they do not cause loss of pain or temperature sensation, muscle wasting, or any clinical deficits persisting more than 6 to 12 weeks. CB lesions of this type are referred to clinically as neurapraxia or first-degree injuries, and because of their temporary nature, they have been virtually excluded from every major series of BP lesions reported. Demyelinating CB, however, also can be

the predominant pathophysiology in two persistent nonneurapraxic brachial plexopathies, radiation-induced and immune-mediated (e.g., multifocal motor CB) disorders, until late in their course. In these plexopathies, unfortunately, instead of rather promptly resolving, as it does when caused by acute trauma, the demyelination usually slowly progresses to AL.78,80,81 Axon Loss Lesions (Axonotmesis; Neurotmesis; Second- to Fifth-Degree Injuries). Axon degeneration is responsible for symptom production in the majority of BP lesions. Typically, it is the sole process in those that have persisted for more than 3 months (radiation-induced and immune-mediated disorders being nearly the sole exceptions). Because the nerve segment distal to the lesion undergoes wallerian degeneration, on NCSs it ceases to conduct impulses after 6 to 9 days, at which time its pathophysiologic hallmark is conduction failure.81 With the mildest grade of AL (axonotmesis; seconddegree injury), regenerating axons encounter no impediment in traversing the lesion site because all the supporting structures of the nerve at that point, including the endoneurial tubes, are intact. Consequently, sensory recovery is excellent, and motor recovery often is equally satisfactory. The only complicating factor is the length that the motor axons must advance beyond the injury site (i.e., the regeneration distance), at approximately 1 inch per month, to reach the denervated muscle fibers, because the latter generally survive only 18 to 20 months without being reinnervated. For this reason, in adults, the time-distance factor is crucial with proximally situated nerve damage, which is characteristic of BP disorders. When AL of this mild grade involves BP elements that supply proximal muscles (e.g., the upper trunk and portions of the lateral and posterior cords), the ultimate recovery typically is excellent. Conversely, when it involves BP elements that innervate distal limb muscles (e.g., the lower trunk and medial cord) in adults, residual motor deficits in the hand are the rule.78 More severe grades of AL (neurotmesis; third- and fourth-degree injuries) typically result from stretch or stretch/contusion lesions. Even with fourth-degree injuries, the epineurium at the lesion site remains intact and, therefore, the nerve elements appear in gross continuity. Nonetheless, so much intraneural damage has occurred at that point (i.e., the endoneurium and perineurium are so disrupted and distorted) that few, if any, regenerating axons can bridge it. As a result, spontaneous recovery, even for lesions affecting nerve fibers supplying proximal muscles, generally is very poor. These more severe grades of lesions in continuity cannot be distinguished from lesser grades by either clinical or EDX examination, or even by visual inspection at operation. They can be differentiated by the amount of recovery that spontaneously occurs but, because of the time-distance factor,

Brachial Plexus Lesions

long periods of observation may be necessary. With severe third- and fourth-degree injuries, this time delay can seriously compromise the amount of recovery that can be realized if surgery is then necessary. The most reliable method for separating these various grades of AL is the use of intraoperative NCSs, as pioneered by Kline.35,78 The most severe damage (i.e., the highest grade) to nerves occurs when the continuity of all their components is lost at the lesion site, and two physically separate segments result (neurotmesis; fifth-degree injuries). Several terms are used to describe this when it affects the BP. With avulsion injuries, the primary roots are torn from the spinal cord. With ruptures and interruptions, the continuity of the nerve fibers of a specific portion of the BP (e.g., upper trunk; axillary terminal nerve) is breached. Avulsions and ruptures are common with severe closed traction injuries, and the majority occur supraclavicularly. Severe upper plexus traction lesions most often produce ruptures at the VR or trunk level (i.e., extraforaminally) because the axons of C5 and C6 (and to a lesser extent C7), unlike those of C8 and T1, are securely anchored by fascia as they exit the intervertebral foramina. Consequently, when stretched, customarily they tear distal to these anchor points and usually can be repaired surgically. In contrast, severe lower plexus traction lesions, although less common, are far more likely to be preganglionic avulsion injuries, which are irreparable. Thus, in one study of 83 patients with traction injuries, fifth-degree lesions involved the upper plexus elements 160 times in all; 38% were avulsions, whereas 62% were ruptures. Conversely, although the lower plexus elements were affected less often (111 times), 86% were avulsions and only 14% were ruptures.78 Longitudinal Extent of Damage The length of BP fibers injured varies markedly with the etiology and severity of the process. In some instances, the damage is quite discrete, whereas in others it involves an extended segment. This factor obviously has major implications for AL lesions, not only as to the amount of spontaneous improvement that occurs, but also as to whether surgical repair is feasible and, if it is, the degree of recovery likely to be achieved. Sharp transection injuries (e.g., those produced by knives or glass) characteristically are very focal; little, if any, of the axon is damaged beyond the small area that has been cut, and there is no superimposed traction. Lacerating injuries, such as those caused by dog bites and moving mechanical objects (chain saws; fan blades), cause irregular or ragged nerve fiber transections. Of these, the latter produce far more extensive injuries, both directly by obliterating long segments of nerve, and also indirectly by traction; even root avulsions can occur. Of the various AL lesions in continuity, stretch/contusion injuries vary significantly in the length of axon they affect. When severe, however, the damage characteristically extends over long segments. Consequently, those that are

1345

surgically repaired usually require relatively long nerve grafts. Similarly, radiation-induced damage, compartment syndromes, and some neoplasms often involve very extensive longitudinal segments of the plexus elements, so extensive that surgical repair frequently is essentially precluded or its effectiveness is much reduced.8,35,78 Extent of Involvement Brachial plexopathies typically are depicted as being either partial (incomplete) or total (complete). The distinction is important because the majority of unfavorable outcomes follow complete lesions that persist more than 3 or so months. Regardless of the underlying pathophysiology, the initial disability obviously is less when only some of the BP is affected. However, if AL is responsible, the degree of involvement of each plexus element, as well as the entire plexus, must also be considered, because this influences prognosis and therefore treatment. Thus an overall partial BP lesion can be complete for some of the elements involved, in which case surgery may be required if any recovery is to occur in their distribution. In contrast, when only some fibers of a particular BP element have undergone AL, then (1) the latter often is of a less severe grade (e.g., axonotmesis), and (2) another process is available for muscle reinnervation besides progressive proximal-to-distal nerve regeneration from the lesion site, namely collateral sprouting from surviving axons. As a result, satisfactory spontaneous recovery frequently occurs. When all etiologies of brachial plexopathies are considered, partial lesions are more common than total ones.35,78 Course over Time Brachial plexopathies manifest a variety of patterns in their natural history. First, the deficits may be maximal at onset and not change substantially as time passes. This presentation, fortunately uncommon, is encountered mainly when severe traction lesions involve the entire BP, or when portions of the BP, particularly the lower trunk or medial cord, sustain total high-grade AL damage. Second, the deficits may be maximal at onset, but then resolve partially or completely with the passage of time. This is by far the more common course with trauma. The tempo of improvement may be rapid or slow, depending upon whether recovery is due to remyelination or to axon regeneration. With the latter, the rate and the effectiveness also are contingent on the distance that the regenerating axons must travel, or the efficiency of collateral sprouting. More proximal muscles characteristically are reinnervated before more distal ones. Third, the process may begin with minimal symptoms in a very limited distribution and then progress. Compartment syndromes (i.e., neurovascular injuries), neoplasms, and radiation injury manifest in this manner, although their rate of progression varies significantly. Thus, once symptomatic, compartment syndromes typically evolve very rapidly, producing severe clinical deficits in just a few hours.

1346

Diseases of the Peripheral Nervous System

In contrast, brachial plexopathies caused by malignant neoplasms generally progress over one or two years before the patient succumbs to tumor extension or metastasis to vital organs, whereas those caused by radiation tend to evolve over many years. Fourth, an asymptomatic latent period of variable duration may be interposed between the time the process is initiated and when symptoms first appear. Such “delayed-onset” BP lesions are uncommon, and many are iatrogenic. One of these is radiation-induced injury, in which the latent period may be measured in years. Some orthopedic procedures also can produce brachial plexopathies after symptom-free intervals of months to several years. In addition, compartment syndromes that compromise the BP can be of delayed as well as of immediate onset; their latent periods can range from less than an hour to 2 or 3 weeks following both traumatic (e.g., gunshot wounds) and iatrogenic (e.g., axillary artery catheterizations) causes.78

Clinical Features Brachial plexopathies cause motor, sensory, and autonomic disturbances of the shoulder, the upper limb, or both, and sometimes autonomic changes in the head (i.e., Horner’s syndrome). Although motor and sensory dysfunctions usually coexist, occasionally one is very disproportionately, or even exclusively, present; for example, solely or predominantly muscle weakness may be seen with some pack palsies, whereas only sensory symptoms may occur with many neoplastic and radiation-induced BP lesions early in their course. When motor involvement is caused by AL, paretic muscles may become atrophic within a few weeks. Fasciculations rarely are seen, the major exceptions being those BP disorders caused by prolonged demyelinating CB (e.g., radiation induced). Sensory loss often is incomplete and may be inconstant within the affected territory. Pain is a frequent symptom, ranging from mild to severe and from transient to persistent. Severe, unremitting pain is especially common with malignant neoplasms and with traction injuries that avulsed multiple roots; severe but temporary pain also is characteristic of neuralgic amyotrophy. Pain occurs with lesser incidence with lesions caused by radiation, gunshot wounds (particularly when there are retained metallic fragments), and disputed N-TOS surgery. The severe pain caused by BP lesions can vary in its characteristics; for example, avulsion pain differs in many respects from causalgic pain. In contrast, pain seldom occurs with certain disorders (e.g., classic postoperative paralysis; pack palsy).35,78,85 Brachial plexopathies, especially when extensive and severe, can cause long-term complications, including muscle wasting; skin changes (e.g., skin breakdown in anesthetic areas with resulting ulceration and secondary infections); edema; joint contractures and degeneration; osteoporosis; and persistent pain. Musculoskeletal compli-

cations are especially likely to follow obstetric paralysis. Severe, permanent lesions also have significant socioeconomic implications. The affected patients often are functionally one-handed, so the type of work they can perform is relatively restricted. This limits the opportunities of those who have not yet reached working age, and often forces those already employed to change occupations. The patients probably most disabled are those who sustain multiple root avulsions during motorcycle accidents, because most of them are just beginning their careers or do semi-skilled manual work, and they are poorly equipped for other employment. In one study of 102 patients who sustained such injuries, more than one third were unemployed after 1 year.78

Clinical Presentations Proximal to the divisions (i.e., supraclavicularly), the BP fibers are arranged segmentally, excluding the relatively few named peripheral nerves (e.g., long thoracic) that arise at that level. Consequently, supraclavicular plexopathies typically produce motor and sensory deficits in the distribution of one, or usually more, limb myotomes and dermatomes. In contrast, distal to the divisions (i.e., infraclavicularly), the BP has lost its segmental character. Hence infraclavicular plexopathies present clinically as proximal peripheral nerve lesions, either singularly or in various, often complex, combinations (e.g., portions of two peripheral nerves).78,80 Supraclavicular Plexopathies These are by far the most common type of BP lesions, reportedly occurring two to seven times more often than infraclavicular ones. The primary roots, mixed spinal nerves, VR, trunks, or any combination may be affected. When traumatic and extensive, the same fibers at two or more levels (e.g., C5 root and upper trunk) may be involved, and frequently different fibers at different supraclavicular levels (e.g., rupture of the C5 VR with avulsion of the C6-T1 roots, or rupture of the upper and middle trunks with avulsion of the C8, T1 roots) are affected. Although the five roots that typically contribute to the BP can be traumatized in any combination, three patterns are seen with some frequency: C5, C6; C5-C7; and C5-T1. With more distal supraclavicular plexopathies that involve the trunks, three presentations are common: upper trunk, upper and middle trunk, and pan-trunk. Having the C7 root or middle trunk alone be either affected or spared is quite rare.8,35,78 Upper Plexus (C5, C6 Roots/Upper Trunk) Lesions. These are the most common subgroup of supraclavicular plexopathies. They frequently occur in isolation, but often are associated with middle plexus or combined middle and lower plexus lesions (the latter combination is designated a

Brachial Plexus Lesions

pan-plexus supraclavicular lesion). The majority result from trauma, particularly closed traction, which causes forcible widening of the shoulder-neck angle.15 Specific etiologies include vehicular accidents, falls onto the shoulder, and blows to the shoulder (e.g., by falling objects). Other causes are both nontraumatic (e.g., pack palsy) and iatrogenic (e.g., obstetric paralysis). The affected muscles mainly are located near the shoulder girdle. Abduction, forward flexion, and external rotation of the shoulder are impaired, as well as forearm flexion and supination. If the lesion is at the root level, the affected muscles include the serratus anterior, rhomboids, and spinati. If it is more distally situated, at the proximal upper trunk level, the serratus anterior and rhomboids are spared. When it is even more distal, at the mid or distal upper trunk level, the spinati also are unaffected. Sensory loss usually is found over the lateral deltoid, extending along the lateral aspect of the arm and forearm to the thumb. The biceps and brachioradialis deep tendon reflexes (DTRs) are diminished or lost. With severe, proximal involvement, the upper limb hangs limply beside the trunk, adducted and internally rotated, so that the palm and hand are visible from the rear. (Because this is the limb position conducive to a person surreptitiously receiving a gratuity, it is referred to as the waiter’s, bellhop’s, or policeman’s tip position.)29,35,78 Overall, upper plexus lesions have a good prognosis. Often, spontaneous recovery is rapid and complete, because demyelinating CB is a common pathophysiology. Moreover, when AL has occurred, most of the denervated muscles are proximally situated and thus the time-distance regeneration factor is favorable with axonotmesis lesions. Even many high-grade AL lesions are amenable to surgical repair, because they are extraforaminal. Thus fifth-degree neurotmesis lesions more likely are postganglionic ruptures of VR or the trunk, rather than preganglionic avulsions of primary roots.78 Middle (Intermediate) Plexus (C7 Root/Middle Trunk) Lesions. Isolated injuries of the C7 root /middle trunk fibers are quite rare. Although these supraclavicular elements are damaged with some frequency, typically this coexists with injuries of other supraclavicular elements (i.e., the upper plexus, the lower plexus, or both plexuses). In one series of 320 operated supraclavicular BP lesions, only one involved solely the middle plexus (it was a C7 root avulsion.)13 The clinical changes with middle plexus lesions include impairment of forearm pronation, radial hand flexion, and extension of the forearm, wrist, and often one or more fingers. Patchy sensory loss occurs over the dorsum of the forearm, hand, and radial fingers, and the triceps DTR is affected.29,35,78 Nearly all middle plexus lesions are caused by traction; a special mechanism is operative in which the force is directed from anterior to posterior.13 Only occasionally are they of nontraumatic origin (e.g., neoplasms). When combined with upper or

1347

lower plexus lesions, traumatic middle plexus lesions cause the clinical deficits to be substantially greater and, when surgery is necessary, the overall improvement is less satisfactory.8,35 When only or predominantly the middle plexus is involved, complete recovery never occurs.13 Lower Plexus (C8, T1 Roots/Lower Trunk) Lesions. Overall, this portion of the supraclavicular plexus is affected less often than the others. Moreover, when the etiology is trauma, it seldom is injured in isolation. Instead, typically it is one component of a combined middle and lower plexus lesion, or a pan-supraclavicular plexopathy. With these lesions wrist and finger movements are impaired, especially flexion; with complete injuries, all the intrinsic hand muscle functions are lost, and the hand and much of the forearm are wasted. Typically a sensory disturbance is noted along the medial aspect of the arm, forearm, and hand, as well as the medial two fingers. If root avulsion has occurred, Horner’s syndrome often is observed.29,35,78 Many lower plexus lesions result from trauma, particularly when there is forceful upper stretch of the plexus.15 Nontraumatic causes include neoplasms and true N-TOS, whereas iatrogenic etiologies include median sternotomy and disputed N-TOS surgery. These lesions have a very variable prognosis. With many initially extensive traumatic plexopathies, this portion of the BP is less severely injured than the others; for example, its fibers may be affected primarily by demyelinating CB, whereas the other supraclavicular fibers undergo AL. Moreover, if AL is present, it may be incomplete. Consequently, spontaneous recovery may be rapid and sometimes substantial after injury. With severe traction, however, the lower plexus damage is much more likely to be root avulsions, rather than ruptures or lesions in continuity of the VR or lower trunk. Regardless, because of the adverse time-distance factor, when very severe AL of any cause and of any grade, even the mildest (i.e., axonotmesis), affects any portion of the lower plexus, intrinsic hand function is irreparably lost (at least in adults). Surgery generally has little to offer in these instances except possible pain alleviation.78 Pan- (C5-T1 Roots/All Trunks) Supraclavicular Lesions. This type of extensive lesion is uncommon, except with closed traction injuries. With both high-energy trauma and obstetric paralysis, it occurs more often than either lower plexus, or combined lower and middle plexus, lesions. Other causes are nontraumatic (e.g., extensive malignant neoplasms) and iatrogenic (e.g., interscalene regional anesthetic blocks). With these lesions virtually all the upper limb muscles are affected, so the arm is flail; there are marked sensory deficits throughout the limb, and often there is pain. Depending upon whether the proximal (i.e., roots) or distal (i.e., trunks) portion of the supraclavicular plexus is affected, the serratus anterior,

1348

Diseases of the Peripheral Nervous System

rhomboids, and spinati muscles may or may not be weak, and a Horner’s syndrome may or may not be present. Although a few supraclavicular pan-plexus lesions recover remarkably well, most have a gloomy prognosis.78 Retroclavicular (Subclavicular) Plexopathies These lesions are discussed infrequently, because the anterior and posterior divisions seldom are damaged in isolation. Even with clavicular fractures, the stress on the BP, at least with high-energy trauma, is such that supraclavicular traction injuries are far more likely to occur than retroclavicular ones. Moreover, when BP injuries result from clavicular fractures, usually the very proximal cords, rather than the divisions, are injured (see Traumatic Thoracic Outlet Syndrome later).78 Nonetheless, occasionally the divisions are crushed between the clavicle and first rib.8 They also may be involved whenever trauma, neoplasm, or radiation produces extensive BP injuries. For this reason, Kline and Hudson sometimes refer to operated BP elements as being at the “division to cord level.”35 Infraclavicular Plexopathies These BP lesions are less common than supraclavicular ones and, overall, have a more favorable prognosis. The cords, terminal nerves, or both, may be involved. Cord lesions typically produce deficits in the distribution of two or more peripheral nerves, or portions of peripheral nerves. Those involving the lateral cord impair forearm flexion, forearm pronation, and wrist radial flexion. Sensory disturbances occur along the lateral forearm and the medial three and one-half digits, and the biceps DTR is affected. Medial cord lesions affect all intrinsic hand muscle functions, such as finger flexion, extension, and abduction, as well as ulnar wrist flexion and deviation. The motor changes caused by lesions of the medial cord are identical to those of the lower trunk except that muscles innervated by the medial pectoral nerve and the C8 root via the radial nerve (e.g., extensor pollicis brevis and extensor indicis proprius) are spared. Sensory disturbances are present along the medial arm, forearm, hand, and medial two fingers. With posterior cord lesions, many upper limb movements are impaired to varying degrees, including abduction, forward flexion, and external rotation of the arm, as well as extension of the forearm, hand, and fingers. Sensation may be compromised along the posterior and lateral deltoid, and the dorsal aspect of the arm, forearm, hand, and radial three fingers.29 Frequently, however, sensory loss is incomplete, sometimes being noticeable only in an area overlying the deltoid and, often, the base of the thumb.78 The triceps DTR is affected. With terminal nerve lesions, the clinical presentation is that of a high injury of one or more of the major upper limb nerves. This presentation is apparent with lesions of the axillary, musculocutaneous, and radial terminal nerves. However, it is not evident with injuries of the median or

ulnar terminal nerves, because neither the median nor the ulnar peripheral nerve has branches arising proximal to the elbow. Consequently, a BP lesion involving either or both of these terminal nerves alone is indistinguishable, both on clinical and EDX examinations, from damage to the median and ulnar peripheral nerves throughout the arm, down to the elbow. The main cause for infraclavicular plexopathies is trauma, both of the closed type (e.g., roadside accidents, falls) and open (e.g., gunshot wounds). The majority of traumatic infraclavicular plexopathies are sustained with, or because of, injuries to neighboring structures (e.g., humeral head dislocations; humeral, scapular, or clavicular fractures; axillary artery and/or vein tears).78 The incidence of coexisting vascular lesions is at least 10%, and it varies with etiology. Thus, in two operated series, vascular repair was necessary in 11% of 105 patients who sustained their injuries mostly in vehicular accidents, but in 33% of 90 patients with gunshot wounds.78 A major nontraumatic cause is neoplasms. Iatrogenic etiologies include radiation, medial brachial fascial compartment (MBFC) syndrome, and a variety of orthopedic procedures performed on or near the shoulder. The information available on infraclavicular plexopathies usually does not allow the incidence of involvement of individual components to be determined. Nonetheless, certain generalizations can be made. With traction injuries, diffuse involvement is noted on the initial examination in approximately half the patients. Of the three cords, the posterior is most likely to be compromised, either alone or in association with other elements, especially its outflow: the axillary and radial terminal nerves. Of the terminal peripheral nerves, the axillary most often is affected (frequently in association with the supraclavicular nerve, which is not considered a terminal nerve). The lateral cord and musculocutaneous nerve are the next most commonly injured, with the medial cord and the ulnar and median terminal nerves being the least susceptible.8,35,78 In contrast, with compartment syndromes the median terminal nerve, followed by the ulnar, is most vulnerable.74 When of the same initial severity, traumatic infraclavicular plexopathies generally recover significantly better than do supraclavicular ones, both spontaneously and following surgical repair. Thus one investigator noted that 85% of such lesions improve spontaneously, whereas only 55% of supraclavicular lesions do so.78 These better results are related principally to the mechanism of injury. The traction forces often are less with infraclavicular, compared to supraclavicular, lesions. As a result, demyelinating CB and milder grades of AL frequently are the types of nerve injury found. There are exceptions, however. Severe traction injuries of the medial cord and the terminal median and ulnar nerves do not recover well spontaneously, and even if these elements are surgically repaired, hand intrinsic muscle function very rarely is regained (because of the

Brachial Plexus Lesions

adverse time-distance factor). Nonetheless, some hand sensation may be restored after lateral cord or terminal median nerve repair.35,78 Also, infraclavicular lesions caused by compartment syndromes have very poor prognoses unless decompressed very early after onset.74 Noteworthy is that surgical repair of infraclavicular plexopathies can be particularly tedious, principally because of the proximity of the vascular structures. This is especially the case when scar tissue is prominent, when prior vascular repairs have been performed (especially if vascular grafts were used), and when the posterior cord and its outflow are involved. The threat of tearing axillary vessels during surgery is ever present, and vascular repairs, particularly of veins, can be difficult.8,35,48,70 Pan- (Cords/Terminal Nerves) Infraclavicular Lesions. These BP lesions result from extensive damage to the cords, the terminal nerves, or some combination. The main causes include trauma (e.g., shoulder dislocations, gunshot wounds) and iatrogenic (e.g., radiation). 78 Whenever a flail arm results from trauma, however, the injury is far more likely to be supraclavicular than infraclavicular. Thus, in one series of 78 such patients, only 3 (5%) had pan-infraclavicular lesions (in every case, resulting from injury of all three trunks). 5 Paninfraclavicular lesions can be difficult to distinguish from distal (i.e., trunk level) pan-supraclavicular lesions. The status of the pectoralis muscles may help. The nerves that supply them (i.e., pectoral nerves) arise from the very proximal portions of the lateral and medial cords, so usually they are not affected by pan-infraclavicular lesions, whereas they invariably are with pan-supraclavicular ones. Often the history and general physical examination are most helpful (e.g., shoulder dislocation, infraclavicular likely; motorcycle accident, supraclavicular likely). Combined Suprascapular/Infraclavicular Plexopathies Lesions of the supraclavicular and infraclavicular plexus sometimes coexist. Most are due to trauma, particularly traction. In one large series (280 patients), these two-level lesions were present in 15% of cases. Other etiologies are both nontraumatic (e.g., malignant neoplasms) and iatrogenic (e.g., radiation induced).78

Assessment Optimal assessment of patients with suspected BP lesions typically requires: (1) a detailed clinical examination that includes history taking, general physical examination, and neurologic evaluation; and (2) various laboratory diagnostic procedures, the two most pertinent being neuroimaging studies and EDX examinations.37,80 The first two questions asked are whether an organic lesion exists at all and, if so, whether it affects the BP. Although usually these

1349

are easily answered, exceptions are encountered. Occasionally, functional brachial plexopathies are diagnosed. Moreover, a number of neurologic and musculoskeletal entities can mimic BP lesions, including rotator cuff tears, motor neuron disease, radiculopathies, peripheral nerve lesions in the arm and forearm, and, especially, spinal accessory neuropathies. Whenever the etiology is not obvious, radial neuropathies frequently are misdiagnosed as diffuse BP lesions; because of the decreased ability to stabilize the thumb and wrist, spurious weakness of the median and ulnar nerves coexists with actual radial nerve weakness.78 Once it has been determined that a BP lesion exists, certain questions regarding it become paramount80,88: 1. Is it supraclavicular or infraclavicular? 2. If supraclavicular, what portion(s) (e.g., upper plexus) does it affect, and is it preganglionic or postganglionic? (With severe closed traction injuries, the latter often is the critical question.) 3. If infraclavicular, is it involving one or more cords, terminal nerves, or a combination? 4. Are there coexisting abnormalities of other, particularly nearby, structures? (This is especially pertinent with posttraumatic and possible neoplastic lesions.) In addition, the site of involvement (right or left) is important, as well as whether this is the dominant or nondominant limb. (It is appalling how frequently the latter, important fact is omitted from all the medical records.) Clinical Examination This portion of the BP assessment characteristically begins with history taking. The following information is important with abrupt-onset lesions: 1. When symptoms began; this is essential because all subsequent time measurements will derive from it. Noteworthy is that the exact time (day/month/year) should be documented, and not some soon-to-be meaningless time reference (e.g., “Sunday, a week ago.”). 2. With traumatic lesions, the circumstances under which the injury occurred (high energy versus low energy), and the position of head and limbs at the moment of occurrence. Unfortunately, many patients also sustain head injuries and have no recall for these factors. 3. With traumatic penetrating injuries, the site of entry, the injurious object (e.g., high-energy bullet versus low-energy knife), and when the symptoms began in relationship to the trauma (i.e., immediate versus delayed; the latter usually indicates that the BP damage is secondary to a vascular injury, such as an expanding hematoma). 4. The patient’s description of the motor and sensory symptoms (particularly weakness and pain) that were

1350

Diseases of the Peripheral Nervous System

present immediately after onset, the ones that are present currently, and any changes that have occurred in the interim (and when they did so). 5. If pain is or has been present, as well as its distribution, character (e.g., burning), persistence, and severity. 6. Whether any lower limb or sphincter symptoms are present (suggestive of concomitant injury to the brain or the spinal cord). With nontraumatic plexopathies, additional pertinent points include obtaining a history, recent or remote, of (1) neoplasms, particularly of the ipsilateral breast and lung; (2) radiation, surgery, or bone or joint injuries to or near the region of the affected BP; and (3) injection of various materials—for analgesic, diagnostic, or therapeutic purposes—into or around the blood vessels (particularly the subclavian and axillary arteries) near the affected BP. 37,70,80,88 The physical examination should include the following components37,70,80,88: 1. Visual inspection to detect any evidence of trauma or an ipsilateral Horner’s syndrome, and to determine the overall symmetry of the neck, shoulders, and upper limbs. Following trauma, a cervical scoliosis often is indicative of root avulsion with concomitant cervical spinal cord injury. 2. Palpation and percussion at the base of the neck and in the axilla, seeking a positive Tinel’s sign, which radiates in the distribution of the affected sensory nerve fibers. 3. Assessment of the passive range of motion of the neck, shoulder, and limb joints, and the strength of the shoulder girdle and limb muscles. With upper plexus lesions, it is especially important to evaluate muscles (e.g., serratus anterior, spinati, rhomboids) whose nerve supply arises very proximally. 4. Assessment of the limb DTRs and both the autonomic status and sensation of the limb, girdle region, and neck. Following trauma, sensory loss extending rostral to the clavicle, into the cervical plexus territory, usually indicates severe injury of the upper plexus. If any sensory deficits are detected, then the type of loss (e.g., position, vibration, and light touch alone versus all modalities), its severity, and its distribution (dermatomal versus peripheral nerve) are important. 5. Assessment of limb vascularity, with palpation of the radial and ulnar arteries. 6. Assessment of the lower limbs and trunk, seeking evidence of coexisting lesions of the spinal cord or other structures. Whenever symptoms persist, repeat clinical examinations are necessary to determine if the process is progressing, static, or regressing. This is particularly important with traumatic or iatrogenic lesions initially evaluated soon after injury. Frequently, repeat assessments performed just 1 or 2 weeks later show substantial improvement (because

demyelinating CB has resolved). Noteworthy is that recovery of sensation is a much less reliable finding than is improvement of muscle strength, because the field of sensory loss characteristically shrinks with time, as nerve fibers from adjacent territories extend into the affected area.78 Neuroimaging Studies Several neuroimaging studies are available for assessment of brachial plexopathies. Plain Radiography. This procedure shows abnormalities located both within the intraspinal canal and extraforaminally. With traumatic plexopathies, radiographs of the neck, shoulder, and upper arm are useful. Lateral tilt of the cervical spine, fractures (avulsions) of cervical transverse processes, and first rib fractures suggest root avulsions. Scapular, clavicular, and proximal humeral fractures (and bony deformities resulting from them) can be visualized, as can foreign objects (e.g., bullets, metallic fragments). An ipsilateral elevated hemidiaphragm confirms concomitant phrenic nerve injury. With nontraumatic lesions, congenital cervical ribs, elongated C7 transverse processes, and deformed first thoracic ribs may be seen, as may associated lung involvement (e.g., superior sulcus masses) with secondary malignant neoplastic plexopathies, sometimes accompanied by destruction of bony structures. With iatrogenic lesions, errant screws and radiation changes of long bones may be visualized.8,35,78 Myelography. The principal use of this procedure is with severe supraclavicular lesions, to separate preganglionic (i.e., avulsions) from postganglionic injuries. The most common finding with the former is a “traumatic meningocele”: the contrast material passes through the intervertebral foramen into an extraforaminal diverticulum, caused by dural rupture. Other findings are obliteration of the root sleeve tips and concave defects at the affected root levels, caused by scarring. Both false-negative and false-positive studies occur, and early myelography is potentially hazardous if there are coexisting head injuries.8,78 Computed Tomography. Conventional (unenhanced) computed tomography (CT) is used infrequently to assess the BP, because of its many limitations. However, when CT is performed after myelography, it can be very helpful in establishing the integrity of the primary roots. For more than two decades, many clinicians have considered CT myelography to be the primary imaging modality with proximal closed traction injuries.35,67 Magnetic Resonance Imaging. MRI has been considered the most valuable modality for assessing the extraforaminal BP for nearly a decade. It can detect neoplasms (both primary and secondary), radiation fibrosis, vascular abnormalities, and swelling of the nerve trunks in the

Brachial Plexus Lesions

1351

posterior triangle of the neck following trauma. What conventional MRI generally cannot do, however, is provide optimal visualization of the BP fibers within the intraspinal canal and the intervertebral foramen. Nonetheless, it can reveal valuable, indirect evidence of root avulsion. Unfortunately, the acquisition time for conventional MRI is long, and is further increased when studies of multiple sections and planes are necessary.45,67 Two special MRI techniques also are used. Magnetic resonance myelography evaluates the BP fibers within the cervical intraspinal canal. However, unlike CT myelography, it is not invasive. Rather, it utilizes special techniques to create a myelogram-like image. It differs from conventional MRI in that it has a relatively short acquisition time, and is comparable to CT myelography and conventional myelography in demonstrating damage to structures within the intraspinal canal. Because contrast material is not needed, it can be used in the acute phase of traumatic lesions. Its principal disadvantages are artifacts and difficulty in determining the exact level of injury, because no bony landmarks are included. In many centers it is now the primary neuroimaging study with proximal traumatic brachial plexopathies.45,67 Magnetic resonance neurography (MRN) has been used to assess the PNS, including the BP, for nearly a decade. It utilizes special techniques to enhance the signal differences between various tissues, allowing nerve fibers to be distinguished from surrounding structures, and the various structures within the perineurium itself to be distinguished from each other. Based on its stated capabilities, MRN should be the optimal neuroimaging study for assessing all portions of the BP. For uncertain reasons, however, it has not generated widespread interest, possibly because it requires custom-built components (i.e., phased-array coils).1 Nonetheless, as improvements in technology occur, this MRI technique, or one of the others described, will undoubtedly yield much more detailed information concerning BP lesions.

recording points. Consequently, with suspected BP lesions, distal stimulation and recording (e.g., stimulating the ulnar nerve at the wrist while recording from hypothenar muscles) yields normal CMAPs when the recorded muscle weakness is due to nonorganic causes or to CBs, but lowamplitude or even unelicitable responses (depending on the percentage of fibers affected) whenever AL has occurred. However, the CMAP amplitudes are reliable regarding the amount of motor AL only after wallerian degeneration has reached an advanced state along all the distal stump fibers, thereby preventing impulses from being transmitted to the recorded muscle (i.e., by 7 days after onset). The amplitudes of sensory responses, or sensory nerve action potentials (SNAPs), are useful for two reasons. First, with incomplete postganglionic AL lesions, including plexopathies, characteristically they are more affected than the corresponding CMAP amplitudes; for example, with a moderate AL lower trunk lesion, the ulnar SNAP amplitude will be low while the ulnar and median CMAP amplitudes will still be within the normal range, despite the fact that all the axons assessed traverse the lower trunk. Why this occurs is unknown but, for this reason, the SNAP amplitudes are the most sensitive NCS indicator of AL. However, they require a longer period after lesion onset to reach their nadir (e.g., 10 to 11 days). Second, unlike the CMAP amplitudes, the SNAP amplitudes are not affected by AL lesions within the intraspinal canal, because the peripheral segments being assessed are still in continuity with their cell bodies, which are located in the DRG. That the central connections of the DRG have been interrupted has no effect on the SNAPs. Consequently, with severe proximal AL lesions, whenever the same portion of the BP (e.g., lower trunk) is assessed with both sensory and motor NCSs, and a combination of normal or low-amplitude SNAPs with unelicitable CMAPs results, root avulsions are present with near certainty.81 NCSs are available for assessing almost every major element of the BP (Fig. 55–4).

Other Studies. Arteriography and venography occasionally are of benefit, mainly when neurovascular lesions are suspected and with traumatic TOS. Fluoroscopic examination may reveal paralysis of the ipsilateral diaphragm, indicating phrenic nerve injury.78,88

Needle EMG. This EDX component is useful because it can detect minimal amounts of motor AL and it permits a far more extensive investigation of the BP than do the NCSs. Thus it assesses several nerves (e.g., long thoracic, suprascapular, anterior interosseous) that cannot be adequately evaluated by NCSs. Moreover, with AL lesions it can reveal (1) minimal residual innervation; (2) early reinnervation, weeks before it becomes clinically evident; and (3) the chronicity of the lesion, via changes in the configuration of the motor unit potentials (MUPs), that is, the presence of chronic neurogenic MUP changes, which take 4 to 6 months to develop. Also, needle EMG of the cervical paraspinal muscles can assist with suspected root avulsions; if fibrillation potentials are found, the lesion very likely is preganglionic, because the paraspinal muscles derive their innervation from the dorsal rami, which separate from the VR very near the spine. Although

Electrodiagnostic Examination This diagnostic procedure consists of two major components: NCSs and needle electromyography (EMG). Nerve Conduction Studies. With NCSs, only the amplitudes of the responses, recorded with surface electrodes, are of value. The amplitudes of the motor responses, or compound muscle action potentials (CMAPs), are a semiquantitative measure of the number of viable nerve fibers innervating the recorded muscle, and capable of conducting impulses between the stimulation and

1352

Diseases of the Peripheral Nervous System

FIGURE 55–4 The components of the brachial plexus assessed by the various sensory and motor nerve conduction studies. (Nerve assessed in bold, recording site in parentheses.)

Considerations in EDX Examinations. To be most effective, the EDX examination must be extensive. Sufficient NCSs must be performed to assess the motor and sensory fibers of each of the major elements of the BP, and at least two muscles supplied by each of the motor components must be sampled on needle EMG. To accomplish the former, “nonstandard” NCSs must be performed to evaluate the upper trunk as well as the lateral and posterior cords, because the standard studies assess principally the lower trunk, medial cord and, usually, the middle trunk (see Fig. 55–4). The EDX examination has certain limitations78,80:

2. The sites to be studied (i.e., where the electrodes for NCSs are placed; where the needle EMG electrode is inserted) must be accessible (e.g., not covered by bandages, casts, or various lines). This shortcoming is especially noticeable with many traumatic plexopathies, which often are accompanied by injuries to the skin and bones of the affected limb. 3. There is no SNAP available for assessing the C5 root. 4. If the SNAPs are unelicitable, suggesting a ganglionic or postganglionic lesion, an additional, more proximal, preganglionic lesion involving the same fibers cannot be excluded. 5. Although it readily distinguishes demyelinating CB from AL loss, it cannot differentiate one grade of AL lesion from another (i.e., axonotmesis from neurotmesis; thirddegree from fifth-degree neurotmesis).

1. It is time dependent, being of limited value with AL lesions before the NCS changes are as severe as they are going to be (10 to 11 days after onset), and until fibrillation potentials have had the opportunity to develop (21 days after onset).

Somatosensory Evoked Potentials For BP assessment, somatosensory evoked potentials (SEPs) most often are obtained by percutaneously stimulating various upper limb nerves, usually the median and

fibrillation potentials are the most sensitive EDX indicator of motor AL, they require approximately 3 weeks to develop in denervated muscle.81

Brachial Plexus Lesions

ulnar, while recording over the supraclavicular region/clavicle, the posterior cervical spine, and the contralateral somatosensory cortex. However, they also can be done intraoperatively by stimulating below or above a BP lesion exposed during surgery. Percutaneous SEPs, unfortunately, suffer the same deficiencies as the sensory NCSs in assessing the BP, including the inability to evaluate the C5 root. Moreover, usually not enough SEPs are performed to assess all the pertinent DRG (C6-T1) and the sensory fibers derived from them, as often is done with sensory NCSs. Because percutaneous SEPs are more complex than sensory NCSs, and provide far less information about the status of the BP than a thorough EDX examination, they are of limited usefulness.78 (Although they have detected BP lesions in multitraumatized comatose patients,42 NCSs can do the same.) In contrast, intraoperative SEPs are valuable adjunctive procedures: they can assess C5 fibers; they can demonstrate that a “two-level” supraclavicular lesion is present, thereby preventing useless repair of the more distal injury; and they can prove that the proximal stump of a ruptured plexus element is intact and, consequently, is suitable for grafting.8,78 Intraoperative SEPs can be combined with intraoperative evoked muscle action potentials (recorded from neck muscles). In one study, this combination identified every lesion involving the posterior and anterior roots.49 Intraoperative Nerve Action Potentials This procedure, developed by Kline, is used principally to determine if regeneration is occurring with extraforaminal lesions in continuity, although it also helps in establishing that a root is avulsed. With disorders of abrupt onset, usually it is done 2 to 4 months after onset. By stimulating and recording across the lesion site of each injured element, the presence or absence of conduction through that affected region can be demonstrated. If a nerve action potential (NAP) can be recorded, a substantial number of regenerating large fibers have traversed the lesion site, which is then left intact. Conversely, if one cannot be elicited, the abnormal segment is resected, because spontaneous recovery is very unlikely to occur. Thus the surgeon can choose whether or not to resect a focal lesion using a far more reliable method than visual inspection, which is prone to error, even when the operating microscope is employed.8,78

SPECIFIC ETIOLOGIES OF BRACHIAL PLEXOPATHY In this section, BP lesions are discussed under two general categories, supraclavicular and infraclavicular, each subdivided, based on etiology, into (1) traumatic, (2) nontraumatic, and (3) iatrogenic (generally, merely a specific type of traumatic lesion). Even this simple classification is not ideal,

1353

because some overlap is inevitable and a few disorders are difficult to categorize. Regarding the latter, one of the worst is TOS, because it is not the name of a single entity but, rather, an umbrella title for at least five distinct disorders, only three of which involve, or reputedly involve, the BP.16 Consequently, only the two noncontroversial types of N-TOS are reviewed in their appropriate sections, and the entire topic of TOS is discussed at the end of the chapter. Some BP lesions, it will be noted, have a characteristic, nearpathognomonic appearance (e.g., true N-TOS), whereas others are quite variable in their presentation (e.g., closed traction injuries).

Traumatic Supraclavicular Plexopathies Closed Traction Injuries These are the most common injuries that the BP sustains. They dominate nearly every operated series derived from a general population. A considerable majority result from vehicular accidents, with motorcycle accidents contributing a disproportionate number compared to their relative incidence; for example, in a multitraumatized population, BP lesions occurred with only 0.67% of motor vehicle accidents, the most frequent cause, but with 4.2% of motorcycle accidents.42 Less often, closed traction injuries result from industrial accidents, falls (particularly from heights), objects striking the shoulder, and certain athletic activities (e.g., skiing, mountain or rock climbing, and occasionally contact sports). The patients are overwhelmingly men, and the injuries nearly always are unilateral.78 The most common presentation is solely injury of the upper plexus. Its fibers are relatively short and under mild tension when the arm is at rest. Moreover, the most common mechanism of BP stretch injury is forcible downward traction, caused by either shoulder depression, lateral head movement to the contralateral side, or both. Under these circumstances, the upper plexus is either the only supraclavicular component injured or the first, with the middle plexus and then the lower plexus damaged in sequence. Combined upper and middle plexus traction lesions are next most common, with pan-supraclavicular plexopathies being third. The causal mechanisms for these vary. They include that just described, plus forcible upward stretch on the BP (i.e., traction placed on the hyperabducted limb), which puts tension on the lower plexus initially but, if severe enough, the middle and ultimately the upper plexus also. Solely lower plexus lesions are quite uncommon, and isolated middle plexus lesions are distinctly rare.8,14,35,70 Traction produces markedly variable degrees of nerve fiber damage, ranging from low to high grade (i.e., from neurapraxia to avulsions or ruptures). With extensive injuries, frequently the degree of damage sustained by the various supraclavicular BP elements varies as well; for example, with pan-supraclavicular lesions, a common

1354

Diseases of the Peripheral Nervous System

pattern is avulsions of the C8, T1 roots coexisting with ruptures of the upper and middle trunks. When the etiology is high-energy trauma, these BP lesions may not be recognized immediately because approximately 80% of patients have associated injuries, including closed head trauma (approximately 70%); rib fractures, sometimes with flail chests, lung contusions, and hemopneumothorax (up to 40%); bone fractures (clavicle, scapula, humerus) or dislocations (more than 50%); and spinal cord damage (about 2% to 3%). Moreover, proximal vascular structures are stretched as well as BP fibers, so associated subclavian or axillary artery injuries, particularly ruptures, are common (up to 25% of cases).4,5,8,35,42,48,70 Under these circumstances, the fact that a brachial plexopathy exists may not be appreciated for several days. Even if it is, early surgical treatment usually is not feasible because (1) life-threatening injuries take precedence, (2) the patient is too ill to endure very prolonged anesthesia and surgery, or (3) a specialized surgeon is not available.48 Once a BP lesion has been identified, the critical question concerns its longitudinal location: preganglionic or extraforaminal? Certain findings are suggestive of preganglionic injury (Table 55–1). With supraclavicular closed traction injuries, the amount of recovery varies, depending on such factors as (1) underlying pathophysiology (demyelinating CB lesions always resolve without residuals, as do most low-grade AL lesions that affect the upper plexus); (2) the particular BP elements affected (upper trunk injuries recover far better than do Table 55–1. Findings Suggestive of Root Avulsion Clinical History of violent trauma Total plexopathy Early severe burning pain in anesthetic hand Horner’s syndrome Paralysis of serratus anterior, rhomboids, and spinati Ipsilateral hemidiaphragm paralysis Sensory loss proximal to clavicle (cervical plexus) Negative supraclavicular Tinel’s sign Posttraumatic cervical scoliosis Concomitant spinal cord injury Neuroimaging Fracture of transverse process Traumatic meningocele Obliteration of root pouch Absence of roots in intraspinal canal Electrodiagnostic Normal SNAPs with unelicitable CMAPs Absence of F waves Paraspinal fibrillation potentials Negative SEP (N9 to opposite cortex) CMAP ⫽ compound muscle action potential; SEP ⫽ somatosensory evoked potential; SNAP ⫽ sensory action potential.

those of the lower trunk, because the latter have an adverse time-distance factor); (3) amount of damage to each plexus element (partial injuries are much more likely to recover than are complete ones); and (4) extensiveness of lesion (e.g., at 3 to 4 months postinjury, some recovery is seen with 30% of upper plexus lesions alone, 12% of combined upper and middle plexus lesions, and just 4% of pan-plexus lesions.)34,78 Residuals of varying severity are inescapable with (1) any root avulsion; (2) extraforaminal severe AL lesions of the upper plexus, and especially the upper and middle plexus, that require graft placements; and (3) virtually all total and near-total AL lesions, regardless of grade, that affect the extraforaminal lower plexus. If severe deficits persist, operative treatment is warranted. Among two large series, this was necessary in 31% and 41% of cases.34,46 Generally, these operations are performed 3 to 4 months after injury. There is near-unanimous agreement that they should not be performed after 6 months, if at all possible, because of noticeably poor results.5,8,14,34,35,43,71 With extensive lesions, it is impossible to repair all the injured BP elements. Most surgeons consider the primary goal to be restoration of elbow flexion; otherwise, even an intact hand is essentially useless. Secondary goals include restoring wrist extension, finger flexion, shoulder abduction, and protective sensation in a median nerve distribution.5,8,35,59,70 Surgery may consist of neurolysis, placement of vascularized or nonvascularized nerve grafts, or neurotization (see later). Because the ends of torn nerves usually have retracted by the time of operation and a variable amount of damaged tissue on each end must be trimmed, direct end-to-end repair of ruptured elements generally cannot be done. Overall, about 50% of patients have good results.35 In addition to direct repair of injured BP elements, a number of secondary surgical procedures may be performed to augment specific functions. These include pedicle muscle transfers and tendon transfers, joint fusions, and a variety of osteotomies.5,8,35,70 Persistent pain is much less prominent with extraforaminal supraclavicular closed traction injuries than with avulsion injuries (10% to 20% vs. ⬎40%). Fortunately, the pain with extraforaminal damage tends to resolve when motor function is restored, either spontaneously or by operative repair.43,47,70,78 Root Avulsion Injuries. The primary roots may be torn (avulsed) from the spinal cord by high-energy traction. Although technically these are simply extremely severe radiculopathies, they are so different in their presentation, diagnosis, and treatment from the typical compressive root lesions, and so similar in many respects to total axon loss trunk injuries, that invariably they are discussed with brachial plexopathies. Root avulsions are the most serious injuries the BP can sustain, because they produce irreversible loss of function, and often severe pain. The primary roots are particularly vulnerable to traction because they are short, and they lack both protective epineurial and perineurial

Brachial Plexus Lesions

sheaths and the tensile strength of peripheral nerves.78 Although infants can sustain avulsions at birth, most occur in adults. The causes generally are the same as for the less severe supraclavicular closed traction injuries, except that the stretch characteristically is of high energy; for example, with vehicular accidents, impact velocities must exceed 10 m/s (22.4 miles/hr) to produce these injuries.78 The major source of root avulsions in adults in industrialized countries is motorcycle accidents.4,8,35,42,48,71 Motorcycles (and to a lesser extent bicycles and snowmobiles) are such potent generators because they can obtain high speeds and they lack a protective external envelope.78 Several injurious mechanisms may be operative in the same accident, that is, the victims may sustain more than one injury in rapid succession, since they “hit the oncoming vehicle, then the road, and will slide on the ground to be stopped only by the curb, a post, a tree...”48 The number of roots avulsed per patient has increased over the past two decades, principally because laws were enacted mandating that motorcyclists wear helmets. While these reduced the number of fatalities, they substantially increased the number of patients with total or near-total root avulsions. Narakas noted that in his country (Switzerland) helmet use was enforced by law beginning in mid-1981. Prior to that time, of 211 operated patients, 29 (13.7%) had avulsions of four or five roots, and 16% had coexisting arterial ruptures. In contrast, after 1982, among 108 operated cases, 33 (31%) had four or five roots avulsed, and 33% of them had coexisting arterial ruptures.48 When the latter are present, emergency vascular repairs often are required, with the treatment of the BP damage relegated to a later time. Generally, this must occur within a few weeks, because a secondary exploration many weeks or months following arterial repair usually presents the worst possible conditions. Both the repaired blood vessel and the damaged nerve trunks are encased in a massive scar, sometimes extending over 10 cm (4 inches), causing the surgery to be both extremely long and hazardous: the grafted blood vessels are at major risk of injury, as are any nerve fibers that are in continuity. Identification of the BP lesion under these circumstances often is impossible.35,48 Once avulsion injuries are diagnosed, rehabilitation usually involves two interrelated tasks, both often formidable: (1) restoring at least some useful motor function, and (2) pain control. The difficulties experienced with both are principally directly related to the number of avulsed roots.78 Concerning limb function, several different restorative procedures have been used. Unfortunately, none provides perfect results, and fewer are available as the number of avulsed roots increases. In any case, once spontaneous recovery, if any, ceases, treatment is determined by the remaining disability. Many patients sustain a combination of preganglionic and postganglionic lesions along the various supraclavicular fibers (e.g., avulsion of the C7, C8, and T1 roots, with rupture of the C5, C6 VR). With these situations, the extraforaminal lesions often can be directly

1355

repaired by grafts, and in highly selected instances even preganglionic lesions may be repaired because, as Birch and co-workers have stressed, not all are actually avulsions. The junction between the roots and the spinal cord consists of a central-peripheral transition zone, which lies outside the spinal cord itself. With some preganglionic injuries, the roots rupture at the transition zone, rather than being torn from the spinal cord. Nonetheless, there are very few reports of attempts to directly repair torn primary roots.8 The typical approach, instead, is to perform nerve transfers, or neurotization. This consists of attaching portions of the BP that have been separated by avulsion from the spinal cord to the proximal stumps of other BP elements that are still in continuity with it, or to the proximal stumps of deliberately transected extraplexal nerves (e.g., the spinal accessory nerve, intercostal nerves, and cervical plexus branches). The concept is to restore functions considered more important than those lost by the transfer. These operations are, at best, salvage procedures; generally, the goal is to recover one or, at best, a few useful functions (e.g., elbow flexion). In addition to nerve repairs and transfers, a number of reconstructive orthopedic procedures can be performed (e.g., tendon transfers). The most frustrating situation is when all the BP roots are involved, because few options remain. Formerly, the recommended treatment was shoulder arthrodesis with arm amputation and fitting of a prosthesis. This approach has been abandoned. Currently, amputation is only considered if the involved limb is infected or was massively traumatized (e.g., multiple fractures with nonunions), or if the patient prefers it (which rarely happens).8,35,78 Of all BP lesions, avulsion injuries are particularly liable to be painful. Narakas noted that significant pain was reported by 70% of patients with root avulsions, 42% with more distal supraclavicular brachial plexopathies, and just 17% with infraclavicular ones.47 The characteristics of root avulsion pain are as follows: 1. It apparently is restricted to adults, in that it does not affect children.14 2. Its incidence is related to the number of roots avulsed, that is, it is twice as likely to occur when four or five roots are avulsed as when one to three roots are. Thus, in one study it occurred in 40% of patients with C5, C6 avulsions, but in 90% with C5-T1 avulsions.46 3. It tends to appear early after injury, often immediately afterward and usually within the first 2 weeks. One report noted that pain developed within the first week in 65% of patients with root avulsions, but in just 26% of those with BP injuries that were not avulsions.47 4. It generally is experienced in the more distal portion of the limb (i.e., distal to the midportion of the arm), with the exact distribution related to the specific roots injured. Overall, it is most frequently felt, and most severe, in the hand.

1356

Diseases of the Peripheral Nervous System

5. It has a distinctive, almost unique presentation, consisting of two components: (a) a constant burning or crushing background pain, seldom varying in intensity; and (b) severe, paroxysmal, electric shock–like pains of very brief duration that abruptly superimpose on the background pain. The latter can vary markedly in frequency, occurring from many per day to a few per week, and are completely unpredictable; they are the most debilitating for the patient. 6. It increases in severity with emotional stress, intercurrent illness, and lack of intense absorption in a task. In contrast to causalgic pain, however, the only external factor that alters its intensity is temperature change (exaggerated by cold, relieved by warmth).78 7. Its duration varies. In one series, it decreased significantly during the first year in about 25% of patients, but was still severe after 3 years in 33%. Anytime it is present 3 or more years after injury, it is likely to persist indefinitely.47,78 No better description of root avulsion pain has been provided than that which appeared in a 1911 article authored by Frazier and Skillern, which dealt with a traumatized patient who had intractable left upper limb pain. They had exposed his cervical spinal cord by laminectomy, intending to perform posterior rhizotomies for pain control, only to discover that most of the roots on the affected side had been torn completely from the spinal cord. They included in their article a lengthy verbatim description of avulsion pain provided by their patient, a physician, who had sustained his C6, C7, and C8 root avulsions in a most unusual manner: while walking along the street, he was struck on the shoulder and head by a window washer who had fallen from the fourth floor of a building.22 Root avulsion pain is exceedingly difficult to control. All drugs are essentially worthless, as are stellate ganglion blocks, acupuncture, and faith healing. Alcohol, hypnotism, and especially cannabis may help, but the last-named is an illegal medication in most countries. The two most effective nonsurgical treatments are distraction (i.e., promptly returning the patient to work and engaging him or her in interesting hobbies and recreations) and transcutaneous electrical nerve stimulation. For the latter to be effective, however, electrode placements and stimulation parameters must be optimal, and accomplishing this may be quite time consuming; some physicians introduce patients to this technique only after they have been admitted to the rehabilitation ward.8,78 Many surgical procedures have been tried over the years, nearly all with short-lived success, if any. However, when only some roots are avulsed, pain often decreases or disappears following the return of any motor function resulting from BP surgery, even when nothing is changed in the distribution of the avulsed root(s).47,78 The most helpful surgical procedure for pain amelioration consists of coagulating the dor-

sal root entry zones of the avulsed roots (the DREZ, or Nashold, procedure). This therapeutic modality, introduced in 1975, has proven effective in about two thirds of the patients treated. Unlike most of the other operations used to treat avulsion pain, the DREZ procedure usually induces permanent pain relief, at least to a substantial degree. Complications are uncommon, but include ipsilateral weakness, unpleasant paresthesias, impotence, and late recurrence of pain, which may be more severe than the preoperative pain.8,51,78 Open Injuries Traumatic brachial plexopathies are produced by penetrating objects far less often than they are by closed traction. Among three large series (with obstetric palsy excluded) composed of 513, 988, and 1089 patients, open injuries were responsible for just 2%, 5%, and 17% of the lesions, respectively.14,34,46 The major causes are gunshots, clean lacerations (e.g., from knives and glass), and ragged lacerations (e.g., from moving metal structures or animal bites).8,35,78 Open injuries can affect all portions of the plexus, including the retroclavicular region. Ragged injuries appear to affect the supraclavicular more often than the infraclavicular plexus, whereas clean injuries apparently affect both with about equal incidence; for example, among a series of 45 patients with BP damage caused by knives or glass, 22 involved the supraclavicular plexus and 23 the infraclavicular plexus.8 Although gunshot wounds may affect both, the majority damage the infraclavicular plexus; nonetheless, for convenience they are discussed in this section. All open injuries of the BP share the fact that damage to nearby structures, particularly major blood vessels and lung, are common and frequently are of immediate, overriding concern. Also, in many instances the BP is not injured directly by the penetrating object but, rather, secondarily as a result of primary damage to the blood vessels. Pseudoaneurysms, arteriovenous fistulas, and expanding hematomas are responsible for these neurovascular injuries. Open injuries cause markedly variable damage to the BP, depending upon the nature of the wounding agent. Knives and glass characteristically produce low-energy injuries: the BP elements sustain either a partial or complete clean transection, without traction and, consequently, little to no nerve substance is injured beyond the actual laceration site. Gunshot wounds, in contrast, produce high-energy injuries. Nonetheless, there are marked differences in the muzzle velocities of various weapons (i.e., usually less than 60 m/s for so-called civilian handguns, 200 m/s for military pistols, and 600 to 1000 m/s for “high-powered” military rifles). High-velocity missiles produce large transient cavities in the tissues they traverse. The pressure wave generated causes variable damage to the BP fibers, ranging from CB to severe traction/contusion, without actually transecting them. Chain saws,

Brachial Plexus Lesions

propeller blades, shotguns, and fragments of shells, mortars, and grenades can cause massive high-energy injuries that literally destroy major portions of the BP, as well as surrounding structures (e.g., the clavicle). Animal bites, even though they are low energy, may produce similar BP damage.8,35,62,78 With open injuries, plain radiographs may reveal retained metallic fragments, and arteriograms are performed whenever there is a question of a vascular injury. The EDX examination can establish the full extent of the lesion and detect the presence of demyelinating CB, but it cannot determine the grade of AL present.78 Brachial plexopathies caused by stab wounds are surgically explored as soon as possible (within hours or days of the injury). In about one third of cases, acute operative intervention is required because of coexisting vascular damage. This approach offers multiple advantages for BP repair: there is no scarring and nerve length has not been lost to retraction, so usually end-to-end repair can be accomplished after minimal trimming of the stumps.35 (Generally, because of nerve retraction, end-to-end suturing is not possible with supraclavicular lesions after 1 week and with infraclavicular lesions after 2 weeks.14) Surprisingly, with clean lacerations approximately 20% of the injured plexus elements are still in continuity, having been contused rather than transected. If necessary, these can be treated with a subsequent operation performed a few months later at a time when intraoperative NAPs can be obtained. Clean lacerations are the most favorable serious BP injuries to manage; overall, they recover better than traction injuries of a similar degree. Thus more than 60% of the repaired elements recover well.8,35 With gunshot wounds, relatively acute operations frequently are required, usually for major blood vessel exploration and repair. The latter varies from solely suture to replacement by a vein or a synthetic graft. Often the status of the BP is not determined because these operations typically are not performed by BP surgeons. Even if they were, attempts at repairing any transected neural elements so soon after injury rarely produce good results, principally because not enough of the ends are trimmed to obtain uninjured tissue. Usually, the optimal time to operate on gunshot wounds is 2 to 4 months after injury. Surgery is performed to restore loss of function if it is complete in the distribution of at least one element that can be helped by operation (see later). Another reason for surgery is persistent pain. This may be caused by retained metal fragments, but frequently it results from coexisting vascular lesions.35 Occasionally, the latter are responsible for functional loss as well; regardless of the nature of the open injury, they can cause prolonged CB (presumably on a focal demyelination basis).62 At operation, only about 6% of the injured BP elements are completely separated; most of the damage is of the stretch/contusion type. Depending upon the status of the injured BP elements, with those in continuity assessed with intraoperative NAPs, neurolysis can be per-

1357

formed, grafts can be placed, and, with some terminal nerve injuries, end-to-end suturing can be done.35 Recovery following gunshot wounds is excellent for any BP elements affected by demyelinating CB (and usually some are), and generally quite good for those with incomplete AL lesions. The results also are good following surgical repair with suture or grafts of the C5 and C6 VR, the upper trunk, the lateral and posterior cords, and the outflow of the latter.35 With blunt, open injuries, particularly when massive, BP surgery almost always must be delayed. Even though patients may require emergency surgery for damage to associated structures, they are not able to undergo many hours of additional surgery to repair their BP damage. Nonetheless, operation should be within a few weeks of injury, before substantial scarring develops. Most of these, if they can be repaired at all, require long grafts, not only because of stump retraction resulting from the elapsed time since injury, but also because more of the stumps must be removed. With blunt lesions, even those of lesser severity, recovery is less than with sharp injuries, with good results approaching 50%.35 Following a penetrating injury of the neck, shoulder, or axilla, if neurologic symptoms in the ipsilateral arm and shoulder are absent immediately but then appear after a latent period (hours to weeks), it is very likely that a neurovascular lesion is present.78 These function as compartment syndromes, so even though a bruit, thrill, or palpable mass is detected over the lesion, distal peripheral pulses may be intact. Neurovascular lesions require urgent treatment: once neurologic symptoms appear, they can progress with alarming rapidity, and nerve fibers, once compromised, can progress from lower to higher grades of injury in just a few hours. Thus they are one of the few emergencies in the field of peripheral nerve surgery.78 Pack Palsy (Rucksack Paralysis) Most reports of the shoulder straps of backpacks or their equivalents (e.g., child carriers) causing neurologic symptoms in the shoulder girdle and upper limb have focused on military personnel, although individuals engaged in certain civilian avocations (e.g., scouting, hiking) are not immune. In a 1983 article concerned with 66 PNS lesions sustained by Japanese athletes, 15 (23%) were BP lesions attributed to pack use that occurred in mountain climbers.31 The salient features of pack palsy are as follows: 1. Its occurrence is related to several factors regarding the pack itself and its use, for example, design, the amount of weight carried, length of time worn, distance covered, type of terrain traversed, and use or nonuse of pack frames or waist belts. Waist belts prevent pack palsies because they shift much of the weight to the hips, thereby allowing the shoulder straps to serve principally to keep the pack in position. In one report from a U.S. Army basic training facility,

1358

2. 3.

4.

5.

Diseases of the Peripheral Nervous System

pack palsies were 7.4 times more common in recruits who did not use a pack frame and waist belt, compared with those who did. It is usually unilateral and may affect either the dominant or nondominant limb. The fully developed syndrome may be preceded by one or more transient episodes of symptoms related to pack use. Its most common presentation consists of weakness (sometimes accompanied by wasting), paresthesias, and sensory loss in principally a proximal upper trunk distribution, although combined upper and middle trunk and even pan-trunk distributions have been reported. Pain, in contrast, is rare. In about two thirds of patients its principal pathophysiology is demyelinating CB. In earlier reports a disproportionate number of patients were noted to have some underlying postural, structural, constitutional, or systemic abnormality (e.g., prior trauma to the shoulder region; recent illness), but these predisposing factors are not mentioned in more recent descriptions.

Although the BP is implicated in the majority of pack palsy cases—in one of the very few reports that mentioned EDX studies, the lesion was localized to the proximal upper trunk, based on the detection of a CB on motor NCSs— this obviously is not the location in all patients. Impairment of only the serratus anterior, trapezius, and even the sternomastoid muscles has been described; for example, one patient was reported to have bilateral involvement of the spinal accessory, dorsal scapular, and long thoracic nerves. (How the long thoracic nerves are injured by pack use remains an enigma, although it could be due to medial pressure by the straps in the axilla.) The disorder commonly cited in the differential diagnosis is neuralgic amyotrophy. However, pain typically is severe with the latter while rare with pack palsy. Neuroimaging studies are of no value, unless plain radiographs reveal a bony anomaly (typically of the clavicle). EDX examinations usually can localize the lesion, determine its severity, and establish its pathophysiology, and thereby its prognosis. Because demyelinating CB so often is the predominant pathophysiology, recovery generally occurs within 2 to 3 months of onset. However, in about one third of cases, the symptoms are due to AL, and recovery may be slow and incomplete.78 We evaluated one young man who was discharged from the U. S. Army for permanent disabilities caused by a pack palsy he sustained while testing various pack designs. Burner Syndrome Among athletes assessed in a North American EDX laboratory, burners or stingers (called such because of the predominant symptom) were the second most common

disorder encountered, responsible for 40 (18%) of 216 PNS lesions.36 This type of presumed BP lesion is almost limited to certain contact sports: rugby, wrestling, and particularly football. Its presentation is stereotyped: Triggered by sudden, forceful depression of the shoulder and backward or contralateral movement of the head, it consists of a transient episode of abrupt, intense burning dysesthesia, sometimes accompanied by weakness, involving one entire upper limb. These attacks typically resolve promptly (often within less than a minute) and subsequent neurologic examinations are normal. Several such short-lived events may occur during a single athletic contest. A few athletes, however, develop persistent weakness and sensory changes. The former usually is mild, sometimes only subjective. Curiously, these clinical abnormalities, as well as the EDX changes, almost always are restricted to an upper plexus distribution, even though the burners themselves affected the entire limb. The mechanism of injury (traction versus compression) is unclear, and the exact site of nerve fiber damage is debated: C6 root versus upper trunk of the BP. Laboratory diagnostic studies cannot resolve this localization controversy. Neuroimaging studies are normal, and on EDX examination the modest needle EMG changes characteristically seen—fibrillation potentials in the biceps, deltoid, pronator teres, and occasionally spinati— are equally consistent with a lesion at either location.78 The fact that the upper limb SNAPs that assess the upper plexus almost invariably are normal is compatible with both a cervical radiculopathy and a mild AL upper trunk BP lesion. Conversely, the fact that fibrillation potentials are never detected in the cervical paraspinal muscles (we have assessed more than 35 cases and have consistently found this to be so) is more suggestive of a BP than of a root lesion. The appropriate therapy is also debated. A few physicians (particularly those who believe it is a radiculopathy) contend that one or two episodes are ominous enough to justify withdrawal of the athlete from the sport. Most sports medicine physicians, however, are more liberal and limit participation of those with recurrent burners only if, and while, there is persistent weakness. Long-term follow-up of patients who had burners with persistent weakness has shown that about half still have occasional symptoms, and surprisingly (given the age of the patients, the mildness of AL, and the relatively short regeneration distances involved), almost all have residual needle EMG findings in the muscles that showed changes initially, consisting mainly of chronic neurogenic MUP changes.78 Unfortunately, even though burners have a characteristic presentation, some sports physicians use the term as a generic designation for any traumatic PNS lesion that occurs in the shoulder girdle region, including total AL axillary neuropathies, and the rare, severe supraclavicular closed traction injury. This creates considerable confusion.

Brachial Plexus Lesions

Electricity There are very few published reports of BP injuries caused by electricity, and most are of limited value because they provide so little detail.79 A unique case, reported in 1989, concerned a 14-year-old boy who received an electric shock when he touched a bare electrical cord with his right fifth digit. He immediately experienced some pain radiating from that finger to his axilla and gradually, over the next few hours, he developed progressive weakness and sensory loss in his right upper limb and shoulder girdle, spreading from distal to proximal. When he was examined, some 7 hours after injury, his right hand and forearm muscles were paralyzed, while his more proximal right upper limb muscles were quite weak. Pain sensation was completely absent in the right upper limb and shoulder girdle, as were DTRs. Fortunately, improvement began the day after onset, was substantial within just 2 weeks, and was essentially complete by 3 months. Presumably, the patient’s symptoms were caused by very proximal supraclavicular neurapraxia (probably resulting from electroporation rather than demyelination) involving the BP elements because, on NCSs done 27 hours after the injury, a normal ulnar CMAP was elicited on supraclavicular stimulation.64 At the opposite extreme is the patient who sustained very extensive third-degree burns of his left upper limb and axilla from contact with a high-voltage source. Five days after the injury, there was complete necrosis of the biceps muscle and partial necrosis of some of the surrounding muscles, but all nerves and blood vessels in the axilla, forearm, and hand appeared intact. By the 16th day, however, the BP had become necrotic, and on the 21st day, the entire limb was amputated.79 Miscellaneous Lesions Several closed traumatic BP lesions have been attributed to compression of various types; often the site of injury has been unclear. These include victims of earthquakes who were trapped under rubble, victims of crowd-crush disasters who also sustained traumatic asphyxia, and persons who, either accidentally or with suicidal intentions, were comatose for several hours because of excessive consumption of drugs or alcohol, or inhalation of carbon monoxide or cooking gas. In some of these, the presumed etiology was compression on the infraclavicular plexus in the axilla, resulting from positioning or a compartment syndrome, but direct pressure on the supraclavicular plexus in the neck by the edge of a hard chair was responsible for one.78,86 Several different types of supraclavicular brachial plexopathies have been associated with military personnel. These include the military brace syndrome seen in cadets who stand for long periods with the shoulders strongly retracted and depressed, the head backward, and the chin on the chest; sandbag palsy, which also occurs during training and is caused by running while carrying heavy sandbags behind the head; top cover neuropathy, the

1359

result of wearing heavy body armor (10 to 12 pounds) for protection in a mobile vehicle; and pallbearer’s palsy, sustained while carrying a casket on the shoulder, with the hands at the side, when one of the other pallbearers slips and causes a major weight shift. Surprisingly, firearm recoil palsy apparently has not been reported in military personnel but, rather, only in civilians. With this entity, the BP on the side where the stock of a rifle or shotgun rests is damaged by repeated violent retraction or rearward motion of the lateral part of the clavicle.6,27,78

Nontraumatic Supraclavicular Plexopathies True Neurologic Thoracic Outlet Syndrome (Cervical Rib and Band Syndrome) This noncontroversial type of N-TOS is rare, affects mostly young to middle-aged women, is almost invariably unilateral, and may involve either limb. It is due to anomalies at the base of the neck; specifically, a taut cervical band extending from the proximal portion of the first thoracic rib to the tip of either a rudimentary cervical rib or an elongated C7 transverse process. The distal portion of the T1 VR or the very proximal portion of the lower trunk (particularly its T1 component) is stretched and angulated around this band (Fig. 55–5B). Clinically, it manifests very predominantly as a motor syndrome, generally with one of two presentations: all the intrinsic hand muscles are weak but only the lateral thenar muscles (i.e., the median nerve–innervated muscles) are also wasted, or all the intrinsic hand muscles and the lower plexus–innervated forearm muscles are weak and wasted, but the lateral thenar muscles are the most severely affected. Typically, patients first seek medical attention soon after they abruptly notice, or have it pointed out to them, that one of their hands is wasted. Sensory symptoms, in contrast to motor, usually are mild. Most patients, on direct questioning, report experiencing intermittent pain or paresthesias along the inter aspect of the arm, forearm, and hand, sometimes exaggerated by limb position or use. Nonetheless, these seldom were severe enough to cause them to seek medical care. Clinically, the motor abnormalities are apparent, and usually a sensory loss is detectible in a patchy distribution along the medial aspect of the limb. The DTRs are normal, and the bony anomaly at the base of the neck is too small to be palpated. Typically, no evidence of concomitant arterial compression is found. This disorder usually can be strongly suspected on clinical grounds, but its definite diagnosis rests on its distinctive radiographic and EDX features. The bony anomaly is seen on plain radiographs. Often, a larger cervical rib is present on the contralateral side, but it is asymptomatic because there is no associated radiolucent cervical band. Rather unexpectedly, MRI and CT are not as helpful for demonstrating the cervical band, and usually neither visualizes the bony anomaly. The EDX examination reveals a very chronic

1360

Diseases of the Peripheral Nervous System

FIGURE 55–5 The causal mechanisms of true neurologic thoracic outlet syndrome (TOS) and post–median sternotomy lesions. A, Normal anatomy. B, Distal T1 ventral ramus/proximal lower trunk stretched and angulated around a congenital cervical band, responsible for true N-TOS. C, C8 ventral ramus injured by proximal fracture of first rib (which is located immediately inferior to it), responsible for post–median sternotomy lesion.

lower trunk BP lesion that, because the T1 fibers are disproportionately affected compared to C8, is almost pathognomonic. On NCSs, the MAC SNAP response is very low in amplitude or unelicitable, the median CMAP amplitude is very low, the ulnar SNAP amplitude is borderline low, and the ulnar CMAP amplitudes are within the normal range. (The median SNAPs, which assess an entirely different portion of the supraclavicular plexus, are normal.) On needle EMG, all the hand muscles, particularly the lateral thenar, show substantial MUP loss and prominent chronic MUP changes, but sparse fibrillation potentials. In the forearm, the flexor pollicis longus and, to a lesser extent, the ulnar nerve–innervated muscles show similar changes, but the C8/radial nerve–innervated muscles may or may not do so. The differential diagnosis includes several disorders, such as severe carpal tunnel syndrome, other lower trunk brachial plexopathies (especially neoplasms), and C8 radiculopathies. The optimal treatment is prompt sectioning of the cervical band via a supraclavicular approach. Neither the cervical rib nor the ipsilateral normal first thoracic rib needs to be removed. Postoperatively the sensory complaints, if present, are promptly relieved, and any forearm weakness or wasting may lessen or resolve over the next year or so. Although the intrinsic hand muscle abnormalities stop progressing, they do not regress, so the patient is left with a weak, wasted hand. This is to be expected, considering the severity and duration of the AL.16,38,78,80 Neoplastic Infiltration Neoplasms recognize no boundaries in the BP, being found both supraclavicularly and infraclavicularly. To avoid

redundancy, however, both are discussed in this section. Their actual incidence is unknown because data are available only for operated series, and therefore are biased against malignant neoplasms, which seldom are treated surgically. Moreover, even figures from the operated series are disparate. Thus, in one series concerned with 1068 patients, there were only 15 (1.4%) neoplasms. Conversely, in another, dealing with 1019 patients, neural sheath tumors alone were responsible for 161 cases (16%).4,46 Neoplasms of the BP can be divided into neural sheath and non–neural sheath subtypes, with each subdivided into benign and malignant. Benign neural sheath tumors are the most common primary BP neoplasms, and the majority are either neurofibromas or schwannomas. Among 168 BP neoplasms of all types seen at one institution, 103 (61%) were of one of these two types.23,40 Neurofibromas were somewhat more common than schwannomas: 59 (35%) versus 44 (26%). Although the majority of neurofibromas were solitary and fusiform, some were multiple and plexiform, a component of von Recklinghausen’s disease. Men and women were affected with about equal frequency, and there was a broad range of patient ages. Malignant neural sheath tumors were responsible for 17 of 168 neoplasms (10%). The majority were malignant schwannomas or neurogenic sarcomas. Benign non–neural sheath tumors (e.g., desmoids) were responsible for 48 of the 168 lesions (29%).23,40 Concerning malignant non–neural sheath tumors, over two thirds of them originated from the lung or breast. Various gastrointestinal (e.g., esophagus) and genitourinary (e.g., bladder) sites were responsible for most of the remainder, in addition to melanomas and lymphomas.

Brachial Plexus Lesions

When location is considered, neurofibromas, malignant neural sheath tumors, and malignant non–neural sheath tumors tend to affect the supraclavicular and infraclavicular plexus in approximately equal numbers. (Breast carcinoma is an exception, because it more often is infraclavicular.) Schwannomas, however, more commonly involve the supraclavicular plexus.23,40 Primary BP neoplasms generally present as limb pain, a palpable mass, or both. They may be tender to touch and, when manipulated, cause paresthesias or pain to radiate distally. Sensory and motor deficits may be present, particularly following biopsy. The most helpful neuroimaging procedure is MRI, which typically reveals the lesion even before clinical deficits appear. The EDX examination, in contrast, usually remains normal until there are persistent sensory or motor deficits. Because of their low incidence, these neoplasms often are mistaken for other PNS disorders, so many patients undergo needless surgery before being correctly diagnosed. Moreover, when they affect the upper plexus, they can present as a mass at the base of the neck and be biopsied or excised; this can convert a painless tumor into a painful one, and can cause unnecessary severe, permanent deficits. Consequently, once these BP tumors are suspected, they should be managed by surgeons with special expertise in their care. Neurofibromas arise within the fascicle itself, so they cannot be excised without sacrifice of the involved fascicles, although the nerve fibers within them frequently are already nonfunctional. Schwannomas typically can be removed without sacrificing nerve fascicles because they lie between the latter. Surgical treatment of plexiform neuromas generally is less satisfactory, because the segments of nerve involved are so extensive; elements may have to be sacrificed for pain control or to save the remainder of the BP, even if additional deficits are sustained. With malignant primary BP tumors, en bloc dissection is attempted and usually, after pathologic confirmation, intrascapulothoracic amputation is recommended. Nearly all these patients soon die.23,35,40 Benign non–neural sheath tumors usually are extensive and sometimes locally aggressive. Nonetheless, most can be resected.23,35,40 Malignant non–neural sheath tumors, so-called secondary neoplasms, are more common than primary BP tumors. These generally reach the BP by either direct extension or metastases; the latter mechanism is the more common. Direct extension to the BP is seen primarily with Pancoast’s syndrome (discussed later). Many of these neoplasms have already metastasized elsewhere, especially to the upper lobe of the ipsilateral lung. Metastatic BP involvement usually is the result of lymphatic spread from areas drained by the lateral group of axillary lymph nodes. Excluding those with Pancoast’s syndrome, most patients with secondary neoplasms are known to have, or have had, malignant disease. Nonetheless, the asymptomatic latent period between apparent successful

1361

treatment of the primary malignancy and the onset of BP symptoms at times is so long—latent periods of more than 30 years have been reported—that a cause-and-effect relationship is overlooked or discounted. Many patients with possible metastatic brachial plexopathies, particularly women with breast cancer as their primary lesion, received radiation therapy to the BP region months to years earlier to treat their primary neoplasm. This usually is a major confounding factor when symptoms subsequently develop.78 Differentiating a neoplastic from a radiationinduced BP lesion can be difficult. Various features of the two disorders are helpful, but, unfortunately, none has proven to be absolute, in part because the two entities are not mutually exclusive. Nonetheless, when a combination of findings all point to one of them, the diagnosis is probably accurate (Table 55–2). Secondary neoplasms characteristically present with pain in the shoulder radiating down the upper limb, often accompanied or soon followed by sensory and motor abnormalities. Frequently these initially are in a medial cord or lower plexus distribution; the latter may be accompanied by a Horner’s syndrome. However, by the time most patients are assessed, at least at tertiary care facilities, the lesions are quite diffuse. In addition to BP involvement, some patients develop cervical cord compression because of intraspinal extradural metastases. Most of these neoplasms are not palpable, either because they are too small or because they involve inaccessible neural elements. Both CT and MRI can detect neoplasms presenting as mass lesions, but both have difficulty identifying infiltrating neoplasms and distinguishing them from radiation fibrosis. When pain is the initial symptom, the EDX examination may be normal, although it is almost always abnormal as soon as persistent deficits are present. Secondary neoplasms of the BP may be difficult to diagnose initially, because their symptoms and signs are readily mistaken for those caused by other disorders; for example, lower plexus/medial cord neoplasms often are misdiagnosed as C8, T1 radiculopathies or ulnar neuropathies. Treatment is difficult, and usually just palliative. Chemotherapy and radiation therapy are used if the neoplasm is sensitive to them. With few possible exceptions, surgery has relatively little to offer beyond providing for a definite tissue diagnosis. Severe, unremitting pain frequently overshadows the progressive loss of limb function that occurs, so adequate pain control often becomes the primary goal of treatment. Most patients succumb to neoplastic involvement of vital organs within a few years.78 Pancoast’s syndrome deserves special discussion. It characteristically affects men who are heavy smokers and involves the BP by direct extension (rather than metastases) from a malignant apical lung neoplasm (also called a superior sulcus tumor, or a thoracic inlet tumor), although occasionally there are other causes. The culprit typically is a non–small cell carcinoma, most often squamous cell or

1362

Diseases of the Peripheral Nervous System

Table 55–2. Some Factors Helpful in Distinguishing Secondary Neoplastic from Radiation-Induced Lesions More Likely Neoplastic Historical Onset: shoulder pain ⬍6 months postradiation Rapidly progressive Clinical Horner’s syndrome Severe pain predominant Metastases elsewhere Neuroimaging Focal mass Neoplasms elsewhere Other Neoplasm on biopsy More Likely Radiation Induced Historical Onset: paresthesias in median nerve–innervated digits (D1, D2, or D3) Slowly progressive Symptom duration ⬎4 years Clinical Paresthesias predominant Neuroimaging Low signal on T2-weighted images Electrodiagnostic Motor conduction blocks on supraclavicular NCS stimulation Fasciculation/myokymic potentials on needle EMG EMG ⫽ electromyography; NCS ⫽ nerve conduction study.

adenocarcinoma. The classic presentation consists of pain in the shoulder, radiating down the limb in a lower plexus distribution into the hand; a Horner’s syndrome (most often resulting from direct injury of the stellate ganglia, rather than the T1 root, as occurs with closed traction injuries); bony destruction of one or more of the upper three ribs, vertebral bodies, and transverse processes; and sometimes weakness and wasting of the hand as a result of damage of the C8 root or the lower trunk. Characteristically, the BP symptoms are the first that the patient experiences, that is, he or she is not known to harbor a primary malignancy that is otherwise clinically silent at that early stage. The bony abnormalities can be detected by plain radiographs, CT scans, and MRI. The MRI findings are used to grade these lesions for optimal treatment. The EDX studies reveal substantial denervation in a lower plexus distribution, often with evidence not only of postganglionic involvement (e.g., altered ulnar SNAPs), but also of preganglionic involvement (i.e., fibrillation potentials in the paraspinal muscles), thus reflecting the spread of the disorder. For many years only palliative treatment was offered for Pancoast’s syndrome. However, with the current staging now employed, many patients can be cured

with a combination of surgery and radiation therapy. Typically, multiple surgical specialties are involved (e.g., neurosurgery, orthopedics, thoracic surgery).7,54,78 Miscellaneous Lesions Occasionally, portions of the BP are affected by multifocal motor neuropathy, and painless brachial plexopathies can result from intravenous heroin use.78,80 Although almost invariably discussed with BP lesions, neuralgic amyotrophy, in our experience, usually involves one or more proximal shoulder girdle or upper limb nerves and not the BP. All of the above are considered immune-mediated disorders and are discussed elsewhere (see Chapter 102).

Iatrogenic Supraclavicular Plexopathies Obstetric Paralysis (Obstetric Palsy, Congenital Brachial Plexus Palsy) Possibly this disorder should be discussed under traumatic, rather than iatrogenic, supraclavicular plexopathies, because it can occur when medical personnel are not in attendance, and many instances of it may be due to other than the extraction efforts of the attending clinician. Over the past decade, more publications have focused on this type of BP lesion than any other, even though it is simply a special type of closed traction injury, and one having a much lower incidence than those resulting from vehicular accidents. Obstetric paralysis was probably the first type of BP lesion reported, and for more than 125 years it has been attributed principally to excessive lateral traction applied by the clinician on the fetal head during the third stage of delivery, usually because of dystocia (i.e., difficulty in delivering the shoulder following head delivery, because of impaction of the fetuses anterior shoulder on the mother’s pubic symphysis).19 Its incidence has been estimated, at various times and in different countries, at 0.38 to 1.86 per 1000 live births.19 Predisposing factors include maternal diabetes, multiparity, and weight gain; fetal macrosomia, high birth weight, and breech presentation; shoulder dystocia; assisted deliveries (forceps or vacuum); prolonged labor; and previous birth(s) complicated by BP injury.19,69,75 Males and females are affected in approximately equal numbers, and most lesions are unilateral. The right side is affected somewhat more often than the left, presumably because the left occiput anterior presentation is more common, which causes the right shoulder to be impinged under the maternal pubic arch.69 Vertex presentations are responsible for 93% to 97% of the cases; breech presentation, about 2% to 6%; and cesarian deliveries, less than 1%.25,69 Breech presentations form a special subset because (1) the infants typically are of less than, rather than greater than, normal birth weight; (2) the incidence of bilateral lesions is significantly higher; and (3) often the BP injuries are particularly severe, consisting of avulsions of the upper roots.8,25,26 For the past two decades, the

Brachial Plexus Lesions

concept that these lesions nearly always are the result of excessive traction by the clinician in attendance has been challenged for several different reasons. It is suggested that at least some cases result from prenatal factors or from maternal propulsive forces, over neither of which the medical attendant has any control.58 Obstetric paralysis can manifest any degree of nerve fiber injury. It can also affect any portion of the supraclavicular plexus, although three presentations predominate: among 436 infants treated surgically, 48% had upper plexus lesions, 29% had combined upper and middle plexus lesions, and 23% had pan-plexus injuries.25 Surprisingly, considering how often Klumpke’s paralysis (isolated involvement of the lower plexus) is linked to obstetric palsy, Gilbert,25 who is one of the world’s experts on these disorders, apparently has encountered no instances of it, and the question of it actually existing in this context has been raised.25,69,80 Most often, the upper and middle plexus lesions are extraforaminal ruptures involving the VR, trunks, or divisions. Major exceptions are those sustained during breech deliveries, which usually are isolated avulsions of the C5, C6 or the C5-C7 roots. With total paralysis, the C8 and T1 roots typically are avulsed, whereas the C5-C7 VR or the upper and middle trunks sustain either ruptures or lesions in continuity. However, with pan-plexopathies, there is always one root that is not avulsed. Thus, unlike adult closed traction injuries, complete avulsion of all five roots does not occur.25,69 Also, in contrast to adults with closed traction injuries, concomitant vascular injuries almost never occur.26 Plain radiographs typically are obtained to assess the cervical spine and the ipsilateral shoulder and clavicle. Although MRI sometimes visualizes neuromas and disruptions of the extraforaminal plexus, it is still not as sensitive as CT myelograms for detecting pseudomeningoceles with nerve root avulsions.8,69 EDX studies are difficult to perform, not only because the infant’s small size requires the use of special electrodes, but also because conscious sedation is necessary for optimal assessment. Even when these requirements are met, some sensory NCS responses cannot be elicited. Moreover, the needle EMG findings notoriously are overoptimistic in regard to the amount of reinnervation occurring.75 Nonetheless, when NCSs reveal unelicitable CMAPs in the presence of normal or low-amplitude SNAPs, which assess the same elements of the BP, the chances of root avulsions being present are extremely high. Moreover, the EDX findings can serve as a baseline.78 The differential diagnosis is unexpectedly long, and includes both neurologic and non-neurologic disorders (see Alfonso et al.2). Two commonly associated injuries are phrenic neuropathies with upper plexus injuries, and clavicular fractures.8 Excluding the uncommon situation in which the continuity of all the injured plexus elements is lost (usually as a result of a combination of ruptures or avulsions), there is

1363

always some spontaneous recovery with obstetric paralysis. How often this occurs, however, is debated extensively. There are several reasons for this disagreement, most related to the fact that there are no generally accepted standards for defining the various grades of recovery,25 and final assessments are made too early. Probably the functional end stage can be determined reliably only when patients are 4 years of age or older. Even then, standardized protocols are required. Thus many physicians grade a result as “good” if the hand is functionally intact, even though the shoulder may lack normal rotation and abduction. Considering that the motor fibers supplying the hand are spared with the majority of cases of obstetric paralysis, it is apparent how misleading this criterion of recovery might be.25 Critical long-term assessments suggest that spontaneous recovery occurs in 75% to 78% of cases.75 Consequently, about 25% of infants require surgery to achieve optimal improvement. There are no generally accepted criteria for determining if and when surgery should be performed. Some of the more widely used are as follows25,26: 1. With upper plexus lesions, the beginning of recovery of biceps contraction by 1 month, and normal contraction by 2 months, portends an excellent recovery. (This muscle is used, rather than the deltoid, because it is easier to evaluate.) Conversely, the lack of clinical recovery of the biceps by the end of the third month is an indication for surgery. 2. Pan-plexus lesions, with a flail arm persisting after 1 month, particularly if accompanied by Horner’s syndrome, require surgery, because they will recover very little, if at all, spontaneously. 3. Complete lesions of the upper plexus occurring after a breech delivery usually are due to avulsions and require surgery. Following surgery, recovery usually begins after 6 to 8 months and can progress for 2 to 4 years. Overall, longterm follow-up reveals that surgery helps appreciably in more than 50% of operated patients.25 Sequelae are commonly seen with these lesions as a result of muscle imbalance and muscle contractures, including internal rotation contractures of the shoulder, osseous deformities of the shoulder and elbow, and dislocation of both the humeral and radial heads.69,78 Many of these can be somewhat ameliorated with various operations, so-called secondary reconstruction procedures.19 “Classic” Postoperative Paralysis (Postnarcosis Paralysis; Anesthesia Paralysis) This was the first type of postoperative brachial plexopathy described. It has a fairly typical presentation and clinical course. Following a surgical procedure, most often one performed at a distance from the BP (e.g., cholecystectomy, hysterectomy), the patient awakens from general

1364

Diseases of the Peripheral Nervous System

anesthesia with weakness of one or, occasionally, both upper limbs, sometimes accompanied by paresthesias but seldom by pain (which, if present, is rarely severe). The distribution of the abnormalities varies, but the upper plexus is nearly always affected, either alone or initially accompanied by middle and lower plexus involvement.78 Infrequently, classic postoperative paralysis follows surgery performed on the shoulder itself. In one study concerned with neurologic deficits that developed in 18 limbs during total shoulder arthroplasty, as many as 11 could be considered instances of this disorder, based on the distribution of the findings and the clinical course.41 Under these circumstances, this lesion must be distinguished from direct instrumentation injury to elements of the BP or to proximal nerves arising from it. Initially, classic postoperative paralysis was attributed to some “toxic” effect of the anesthesia. However, it was soon evident that the actual cause was malpositioning of the patient on the operating table, particularly in regard to the affected upper limb. In anesthetized patients, the BP, of all PNS structures, is most susceptible to traction injury from malpositioning, for two reasons: (1) the patients, being unconscious, are insensible to postural insults that normally would not be tolerated; and (2) their muscle tone is reduced by the anesthesia, or even abolished if muscle relaxants are employed. Although it occurs in patients placed in the lateral decubitus and the prone positions, in most instances the patients are supine. Frequently one or more of the following factors are present: use of the Trendelenburg position (particularly when steep and prolonged); abduction to 90 degrees or greater of one upper limb, and particularly both upper limbs; and restraint of the upper limb on an arm board in abduction, extension, and external rotation, along with the rotation and lateral flexion of the head to the opposite side. In contrast to patient positioning, the length of the operation appears to play little role, because this BP lesion has occurred with operations of only 40 to 45 minutes’ duration.78,80 Characteristically, recovery begins early, in most cases within 2 to 3 weeks of onset, and is often complete after just a few months. This clinical course suggests that demyelinating CB is the predominant underlying pathophysiology. This CB can be demonstrated (and the lesion localized to the distal upper trunk level) if musculocutaneous motor NCSs are performed while weakness is still present.44 The diagnosis usually is evident from the clinical circumstances. Neuroimaging studies generally are unremarkable, in contrast to the EDX examination. Before the weakness resolves, motor NCSs can demonstrate the CB, whereas afterward, the sensory NCSs and needle EMG usually show evidence of a mild AL, distal upper trunk lesion—the residuals of what was very predominantly a demyelinating CB abnormality.44 Postoperative neuralgic amyotrophy is included in the differential diagnosis, but with it almost invariably a latent period exists between the

completion of surgery and the onset of symptoms, and pain customarily is prominent. The usual pattern of recovery is for any sensory deficits to resolve first, followed by resolution of weakness in the middle and lower plexus distributions (if they were affected), and finally, in an upper plexus distribution. As would be expected, the prognosis is excellent. 44,78 Unfortunately, many surgeons erroneously assume that all postoperative BP lesions are of this type. Consequently, they assure all such patients that their symptoms will be short lived. When this is not the case, legal action often results. Post–Median Sternotomy Lesion The median sternotomy approach for surgical access to the heart, in which the sternum is split lengthwise and both sides are retracted laterally, is used for cardiac revascularization, valve repair, and transplant. Occasionally, a distinct type of postoperative BP lesion results. Its causal mechanism is debated, but both the clinical and EDX findings are most consistent with the sternal retraction producing an occult fracture of the very proximal portion of the first rib, which, in turn, injures the C8 VR located immediately cephalad to it (see Fig. 55–5C). This disorder affects both men and women. Usually it is unilateral and has no side preference, although it has a higher incidence on the side from which the internal mammary artery was used for grafting, presumably because more extensive sternal retraction was required. It varies significantly in severity. In some patients there are only temporary paresthesias, and sometime pain, in the medial aspect of the hand and fingers (i.e., in an ulnar nerve distribution). More often, however, the damage is more substantial, and the sensory symptoms are accompanied by distal motor deficits caused by weakness of not only ulnar nerve–innervated muscles, but also muscles innervated by the C8 VR via the radial nerve and the median nerve (e.g., flexor pollicis longus). Neuroimaging studies usually are unhelpful. (Specifically, demonstrating the fracture of the first thoracic rib is difficult.) The EDX examination shows minimal to no changes in those patients with short-lived sensory disturbances. However, in those with more prominent symptoms, two patterns are seen. With the first (and less common), in which clinical weakness often is minimal, all NCSs are normal, whereas needle EMG shows fibrillation potentials in a C8 root distribution, suggestive of a cervical radiculopathy. With this presentation neuroimaging studies of the neck often are performed, seeking evidence of a compressive radiculopathy (presumably caused by neck extension during intubation); these typically are unrevealing. In the second, more common, pattern in which there is hand weakness, the ulnar CMAPs and especially the SNAPs usually are low in amplitude. Conversely, the MAC SNAP and median CMAPs typically are normal; even though the axons these NCSs assess travel with the ulnar motor and sensory fibers in the lower trunk, they derive principally from the T1 VR, which

Brachial Plexus Lesions

has not been injured. On needle EMG, fibrillation potentials and MUP loss are seen in a C8 distribution (i.e., not only in ulnar nerve–innervated muscles, but also in C8/radial- and median nerve–innervated muscles). Frequently, in patients assessed within 4 to 6 weeks of onset, even though the ulnar nerve–innervated intrinsic hand muscles are quite weak, the CMAPs recorded from them are either normal or only slightly low in amplitude, indicating that demyelinating CB is responsible for the bulk of the weakness. However, supraclavicular stimulation reveals no evidence of a CB, indicating that the lesion is proximal to the stimulation site, at the midportion of the lower trunk.80 Both the clinical course and EDX examination usually denote that much of the underlying pathophysiology is demyelinating CB, although occasionally AL predominates. This entity very often is confused with postoperative ulnar neuropathy at the elbow, because most of the ulnar nerve fibers proximally traverse the C8 VR. These lesions are treated conservatively because they involve lower plexus fibers, typically are incomplete, and usually initially are dominated by neurapraxia. Unless the diagnosis is accurate, patients may undergo needless ulnar nerve operations in vain attempts to relieve symptoms. Moreover, some thoracic surgeons consider this disorder to be a postoperative instance of disputed N-TOS (see the next section) and subject these patients to unnecessary transaxillary first rib resections.78 Post–Disputed TOS Surgery Since transaxillary first rib resection was introduced in 1966 to treat disputed N-TOS, many patients have sustained intraoperative BP damage. Most often these are severe stretch/contusion injuries of the lower trunk with much scarring, but sometimes the lower plexus fibers are transected. Clinically, patients typically awaken from anesthesia with weakness, numbness, and sometimes pain in a pan-supraclavicular plexus distribution, or less often, in just a lower plexus distribution (hand and medial forearm). Within 2 or so weeks, however, the upper and middle plexus symptoms (presumably caused by demyelinating CB) resolve in those patients who initially had complete limb involvement, whereas pain usually develops in the hand and medial forearm in those who had been pain-free at first. Neuroimaging studies seldom are helpful. The EDX examination confirms the presence of a lower trunk lesion, usually severe. Intrinsic hand weakness and intense pain are very common residuals. Some patients undergo BP surgery in attempts at pain relief.78,80 Miscellaneous Lesions Portions of the supraclavicular BP have been severely injured when masses at the base of the neck have been biopsied or excised, to reveal benign neural sheath neoplasms (intermixed with normal BP fibers).78 Several reports have described BP lesions caused by internal jugular vein

1365

catheterizations (either attempted or successful). In a few instances the symptoms were of immediate onset and relatively transient, suggesting direct needle injury or local anesthetic effects. However, some cases were delayed in onset, one by 2 months; nearly all of these were attributed to expanding hematomas, except one that was caused by an arteriovenous fistula. Most often the upper trunk was injured, but the C7 and C8 roots, and the lower trunk, also have been compromised. Most patients were treated conservatively, although one hematoma was removed and the arteriovenous fistula was obliterated with a balloon. Surprisingly, given that most of these appear to be examples of compartment syndromes, nearly all had relatively rapid, satisfactory recoveries, suggesting the responsible pathophysiology was predominantly demyelinating CB.65,66,68,78 There have been several reports of supraclavicular BP lesions, mainly of the lower trunk, occurring during either attempted or successful subclavian vein catheterization via the infraclavicular approach. Hematomas were responsible in some instances, but in others the cause was unknown. Most patients were treated conservatively, and some apparently had permanent motor residuals in the hand.33,53,73,78 Brachial plexopathies following both supraclavicular and interscalene anesthetic blocks have been reported, but most were probably cases of classic postoperative paralysis, based on their reported location (upper trunk or upper and middle trunk), their pathophysiology (predominantly demyelinating CB), and their subsequent course (usually a fairly rapid recovery).41 We have seen at least one patient, however, who developed a pan-supraclavicular lesion after interscalene regional block that was probably directly due to the anesthetic agent itself. Both the clinical examination and NCSs demonstrated that her plexus involvement was proximal to the mid-trunk level (more proximal than where the lesion with classic postoperative paralysis is located), but distal to the roots. Fortunately, demyelinating CB was the predominant process.

Traumatic Infraclavicular Plexopathies Falls and vehicular accidents are two of the major causes of these BP lesions. With high-energy trauma, they often are secondary to primary damage of the axillary blood vessels, particularly the artery (i.e., they are neurovascular injuries). Arterial thrombosis, ruptures, and tears, with resultant hematomas, pseudoaneurysms, and arteriovenous fistulas, have all been described. Humeral Neck Fractures and Dislocations Infraclavicular brachial plexopathies, rotator cuff tears, and axillary artery injuries accompany fractures or dislocations of the humeral neck with some frequency. In a report dealing with 226 patients who had dislocations, 25% had one or more of these complications.78 However, the incidence of each varies significantly from one publication to another.

1366

Diseases of the Peripheral Nervous System

Thus BP lesions reportedly occur in 19% to 55% of patients with shoulder dislocations.77 Patient age and coexisting lesions play major roles. Older patients with these orthopedic injuries are twice as likely as younger patients to have BP lesions. Moreover, the majority of patients under the age of 60 years are men, whereas in those over the age of 60, the sex ratio is equal, or even shifts toward women. Also, in one study the mean age of dislocations and fractures varied by more than a decade: 53 years for the former, and 69 years for the latter.77 Patients with hematomas caused by associated arterial injuries are twice as likely to have nerve injuries compared to those without them.28 The particular portion of the infraclavicular plexus damaged depends on the position and displacement of the upper limb when the trauma is sustained.77 In a study of 101 patients, 45 had terminal nerve injuries, 31 caused by fractures and 14 by dislocations.17 The terminal nerve most commonly injured is the axillary, either alone or in various combinations. Other nerves affected, in decreasing order, are the suprascapular nerve and the musculocutaneous, radial, and ulnar terminal nerves. Coexisting nerve damage is a major factor influencing recovery of these orthopedic injuries. Fortunately, in most cases demyelinating CB is the predominant pathophysiology, so neural recovery is satisfactory with conservative therapy alone. However, early surgery usually is performed when there is associated axillary artery damage and, after 2 or 3 months, if there is total AL (with no recovery) in the distribution of one or more of the infraclavicular plexus elements.8,35,77 Traumatic Thoracic Outlet Syndrome The term traumatic thoracic outlet syndrome is used to describe two distinct disorders: (1) disputed N-TOS attributed to acute trauma, typically vehicular accidents (discussed later), and (2) the uncommon condition in which primary clavicular trauma secondarily injures nerve fibers, blood vessels, or both (the subject of this discussion). Clavicular fractures are the most common fractures sustained by humans; most are midshaft, and result from falls or vehicular accidents. Even though neurovascular structures are situated between the midportion of the clavicle and the first thoracic rib, usually they are unharmed by these bony injuries. Moreover, when both clavicular fractures and BP lesions result from high-energy trauma, typically they are independent, with the latter being closed supraclavicular traction injuries. Occasionally, however, the clavicular fracture, sometimes remote in time, directly or indirectly injures the underlying neurovascular structures, thereby producing traumatic TOS. This essentially is a disorder of adults and probably affects more men than women, although statistics are lacking on this point. Either the nerve fibers or the blood vessels may be injured in isolation, both may be damaged simultaneously, or a primary vascular injury may secondarily affect the nerve fibers (neurovascular injury). Overall, the BP is injured more often than the blood vessels. The BP damage can be

caused by displacement of fracture fragments, fracture manipulation, tight-fitting figure-of-8 bandages (which are effective in children but not in adults), expanding pseudoaneurysms and hematomas, and hypertrophic callus (exuberant callus syndrome). These infraclavicular plexopathies may develop at any time from immediately following the clavicular fracture to months afterward. The proximal cords (rather than the divisions) are affected. Isolated injuries of each of the three cords have been reported, as has pan-cord involvement. However, the medial cord most often is involved, because it passes immediately behind the midportion of the clavicle.16 Sensory symptoms (paresthesias, pain) are the most common neurologic complaints. They typically radiate down the limb, especially when it is held outstretched or overhead. Less often, there is sensory loss and limb weakness; the latter particularly involves the intrinsic hand muscles. Pain and tenderness also may be present at the fracture site. When vascular lesions coexist, ecchymoses and swelling may be noted at the base of the neck and supraclavicular area, as may profuse swelling of the affected limb. The diagnosis depends on the clinical examination, generally aided by neuroimaging studies and, of less importance, the EDX examination. On clinical examination, sensory loss, weakness, and DTR changes may be detected in the distribution of one or more cords. Depending on the specific clavicular injury and the presence or absence of coexisting vascular injury, a firm tender mass may be palpated, or a bruit heard, over the mid-clavicle. The distal pulses may be diminished and, if they are present, they may disappear when the limb is elevated. Plain radiographs characteristically demonstrate a bony abnormality in the midportion of the clavicle. Both arteriography and venography may be required if vascular compromise is suspected. Vascular lesions may also be demonstrated with CT and MRI. EDX examinations reveal the distribution of the lesion and, depending upon the amplitudes of the SNAPs and CMAPs recorded (including CMAPs obtained by supraclavicular stimulation), establish whether the process is primarily demyelinating CB or AL in type.16,21 Although SEPs may be performed, they add little to the electrophysiologic assessment. At times, the diagnosis of traumatic TOS is rather obvious. However, in its earlier stages and particularly when the clavicular fracture is remote in time and only neural elements are involved, the diagnosis becomes much more difficult. The differential diagnosis includes cervical radiculopathies and various proximal peripheral nerve lesions, depending upon the particular cord(s) involved (e.g., ulnar neuropathies with medial cord injuries). When traumatic TOS manifests very soon after the clavicular injury, for proper management it must be differentiated from supraclavicular closed traction injuries occurring independently with clavicular fractures. Curiously, many of these patients are assumed to be malingerers. In one series of 16 patients, over 50% were so mislabeled.16

Brachial Plexus Lesions

The treatment varies considerably, depending upon the specific lesion(s) present. Mild symptoms, such as intermittent paresthesias experienced with certain upper limb movements, are treated conservatively, with heat, analgesics, and shoulder exercises. However, when symptoms are severe or progressive, surgery is required. Whenever both neural and vascular compromise exists, which usually is seen soon after the clavicle is fractured, vascular repair takes precedence. Often, surgery is necessary not only to treat the BP and vascular injuries, but also to repair the clavicle itself. The prognosis with traumatic TOS varies, depending upon the severity of the lesion, its pathophysiology, and the particular plexus elements involved. With severe AL lesions of the medial cord, one fourth of 13 patients in one operated series showed no improvement.16 Miscellaneous Lesions BP injuries can be sustained with fractures of the scapula. Typically, these are due to blunt trauma and multiple major injuries coexist, including rib and clavicular fractures and pulmonary contusions.72 Blunt trauma, both to the shoulder and axilla (e.g., from a hockey stick), can cause infraclavicular plexopathies. This portion of the BP also can be damaged by burns. Two patients were reported who, while anesthetized with alcohol, sustained substantial axilla burns and extensive infraclavicular plexus injuries when one upper limb was draped over a heated towel rail. Both patients were left with substantial residuals.82

Nontraumatic Infraclavicular Brachial Plexopathies Neoplasms, both primary and secondary, can involve the infraclavicular plexus. These have been discussed earlier under Nontraumatic Supraclavicular Plexopathies. The BP can be injured by nontraumatic hematomas or aneurysms.24 One patient developed sensory symptoms caused by an axillary hematoma of spontaneous origin; he was treated conservatively and did well.30 Another patient developed a severe diffuse infraclavicular BP lesion because of an axillary artery mycotic aneurysm. Despite surgery, he did poorly.76 Immune-mediated disorders (e.g., multifocal motor neuropathy, neuralgic amyotrophy) may involve various portions of the infraclavicular plexus (the latter almost always affects the terminal nerves, especially the axillary). These lesions are discussed elsewhere (see Chapter 102).

Iatrogenic Infraclavicular Plexopathies Radiation Induced That radiation can damage the BP was first reported by Mumenthaler in 1964. At least three distinct types of lesions have been described,11 but two are quite uncommon. The first is caused by ischemia resulting from radiation injury to

1367

the subclavian artery, with subsequent occlusion. Painless weakness and sensory loss develop over a few days and are permanent. The second is a reversible disorder that often manifests just sensory symptoms (particularly paresthesias), and improves spontaneously over 6 to 12 months. In one article concerned with 1624 patients who received radiation therapy to treat breast cancer, 20 patients (1.2%) reportedly developed brachial plexopathies, and of these 16 (80%) were of this transient type; this high incidence has not been noted in any subsequent publications.52 The third is a progressive disorder caused by radiation fibrosis. It is delayed in onset and produces both sensory and motor changes. This is by far the most common type of radiation-induced BP lesion encountered, and the rest of this discussion focuses solely on it. Women who were treated for breast carcinoma dominate nearly every series. Most of the remaining patients received radiation therapy for lung cancer or lymphoma. The incidence varies substantially from one report to another. The highest incidence reported (73%) appeared in a 1966 article, which also reported a much lower incidence (15%) following adjustments to the radiation dose.78 The latent period between radiation therapy and onset of symptoms is extremely broad, ranging from a few weeks to more than 30 years.50,78 The authors of one article stressed that the reported incidence of this complication at 5, or even 10, years postradiation does not represent the full spectrum of injuries. They performed a retrospective survey of 150 patients and found that 55% of late survivors (i.e., greater than 30 years after radiation) had radiation-induced BP lesions.32 Why this symptom-free period varies so greatly from one patient to another is unknown. Unilateral involvement is typical, although bilateral lesions occasionally are seen, most often in patients who received radiation to more midline structures, such as the mediastinum. Multiple interrelated factors produce these BP lesions. Analyses over the past decade have shown that four of the major factors are total dose given, simultaneous chemotherapy, field intersection (overlapping fields, producing “hot spots”), and hypofractionation (giving fewer, larger doses).32,78 Currently, only fraction sizes of 2 Gy or less are used.78 Pathologically, there is extensive loss of myelin, with obliteration of the vascular supply and marked fibrosis.78 Pathophysiologically, these lesions characteristically begin with demyelinating CB, but then, as time passes, AL progressively supervenes. The most common presenting symptoms are paresthesias. Typically, these initially affect one or more of the median nerve–innervated fingers. How often pain occurs, and its severity and onset, are debated.78,80 The most commonly reported motor symptom is weakness of the intrinsic hand muscles. Among 19 patients with radiation-induced BP lesions who were examined after median follow-up (from receiving radiation) of 50 months, 100% had paresthesias, 58% were weak, and 47% had pain.50 The particular portion of the BP affected varies among reports, which usually

1368

Diseases of the Peripheral Nervous System

have not distinguished supraclavicular from infraclavicular involvement, although the former has been implied in the majority. Nonetheless, some authors have noted that infraclavicular elements were affected, a localization more consistent with the EDX findings we have obtained on approximately 50 patients. Although it has been reported that radiation-induced BP lesions often can be distinguished from neoplastic ones because they invariably affect either the upper trunk (77% of the time) or the entire plexus (23%), these presentations have not been confirmed by other investigators.78,80 On physical examination, depending upon how advanced the process is, sensory loss, weakness, atrophy, and DTR abnormalities are found; the atrophy may be concealed by limb edema, which often is present. The most helpful neuroimaging study is MRI. The fibrosis associated with these lesions may have different imaging appearances, showing both low or high signal intensity on T2-weighted sequences.11,78 On EDX examination, the amplitudes of the CMAPs, and particularly the SNAPs, frequently are abnormal; in our experience, the earliest NCS change is a reduction of the SNAP amplitude recorded from one of the median nerve–innervated fingers. If weakness is present, CBs usually can be demonstrated on supraclavicular stimulation, indicating that radiation-induced demyelination is affecting the BP fibers distal to the stimulation site, which is at the mid-trunk level. On needle EMG, fibrillation potentials usually are present, although seldom in substantial numbers. Sometimes fasciculation potentials and myokymic discharges are seen as well; because the latter rarely occur with neoplastic plexopathies, they are helpful for identification purposes. Unfortunately, the EDX examination may be normal very early in the course of the disorder.78,80 Long-term follow-up of as much as 30 years has shown that the natural history of radiation-induced brachial plexopathy is somewhat variable. Relatively optimistic assessments have been reported by some, who noted symptom progression in only about one third of their patients. Our experience and that of many others has been much more pessimistic, in that we have found that a progressive downhill course is characteristic. Although occasionally the process plateaus for varying periods, more often it inexorably affects more of the plexus and to a greater degree. All too frequently, the ultimate result is a diffusely swollen, flail, anesthetic, often painful limb. When the process is advanced, the EDX examination reveals unelicitable SNAPs and low-amplitude or unelicitable CMAPs on NCSs, as well as fibrillation potentials, severe MUP loss, and prominent chronic neurogenic MUP changes on needle EMG. No effective treatment is available for radiationinduced BPs. A few surgical procedures have been performed (e.g., external neurolysis; wrapping the BP with a free or pedicled omental graft) in attempts to restore circulation to the plexus and to retard fibrosis, but these have

not been beneficial.78 Most surgeons will operate on these patients only to relieve intractable pain, and with full knowledge that greater deficits may be produced.35,78 The DREZ procedure has been effective in relieving pain in the few patients in whom it has been performed.87 Medial Brachial Fascial Compartment Syndrome In 1989, the MBFC was described. It extends from the axilla to the elbow and is formed by the tough medial intramuscular septum, dividing near the surface of the arm to enclose the neurovascular bundle, and its surrounding thin fascial axillary sheath. A neurovascular injury—a compartment syndrome involving the MBFC—can occur. Initially, these were reported following axillary arteriography. A hematoma or pseudoaneurysm resulting from the procedure compromises various terminal nerves, particularly the median, in the MBFC. Subsequent experience has indicated that this same mechanism is responsible for the majority of infraclavicular plexopathies that develop following axillary regional block, and possibly also for most of the neurovascular injuries that result from closed or penetrating trauma to this region (Fig. 55–6). Predisposing patient factors for the MBFC syndrome include bleeding disorders, anticoagulants, and uncontrolled hypertension. This disorder occurs unilaterally, may affect either side, and first becomes symptomatic anywhere from immediately after the procedure or injury to more than 2 weeks later.74,80 In a recent report describing 14 patients with the MBFC syndrome caused by axillary arteriography, all patients had sensory symptoms (paresthesias, pain, and sometimes deficits) and weakness.74 Typically, the sensory symptoms precede the motor. Presumably, reversible ischemic CB is responsible for the initial symptoms, but this can fairly rapidly convert to AL, which it apparently does soon after motor deficits appear. The median terminal nerve almost always is affected, either alone (in approximately 50% of instances) or accompanied by involvement of other terminal nerves, particularly the ulnar. When severe, however, all five terminal nerves may be affected, as well as the supraclavicular nerve. The clinical history is paramount for diagnosis (injury or invasive procedure involving the axilla; progressive symptoms). Clinical examination may reveal a palpable hematoma in the proximal arm or axilla, or an axillary ecchymosis. Distal pulses are normal, as they are with most compartment syndromes, because the elevated pressure, although sufficient to collapse the vasa nervorum, is far below mean arterial pressure. Ultrasound, MRI, and CT may reveal the vascular lesion but, considering the very brief time available for surgical decompression before irreversible nerve damage occurs, obtaining these is rarely justified. The EDX examination is of no value in the acute setting, that is, for diagnosing the lesion before permanent nerve deficits occur. However, it can subsequently determine the full extent

Brachial Plexus Lesions

1369

FIGURE 55–6 The medial brachial fascial compartment and the area (in oval) where compartment syndromes develop from primary vascular injury. Various infraclavicular brachial plexus elements are injured, especially the median terminal nerve.

and severity of terminal nerve damage. Generally, these lesions are optimally managed by immediate puncture site or axilla exploration as soon as the diagnosis is considered following clinical examination. One study has shown that patients who undergo early exploration (within 4 hours of onset of symptoms) are eight times more likely to recover completely than are those who are only observed or treated with delayed surgery (i.e., performed more than 4 hours after onset of symptoms). Most patients with MBFC syndrome have not been treated with early operation, and most have permanent residuals, consisting of motor and sensory deficits, and often severe pain, in at least a median terminal nerve distribution.74,80 Orthopedic Procedures The BP can be injured by a variety of orthopedic procedures, mostly operations, performed about the shoulder girdle region for both diagnostic and therapeutic purposes. Nerve injuries reportedly occur in 1% to 2% of patients undergoing rotator cuff tear repair, 1% to 8% of those having anterior instability operations, and 1% to 4% of those undergoing prosthetic arthroplasty.10 Injuries of the BP have been reported in several series of shoulder arthroscopies; characteristically one or more terminal nerves were affected. Typically these were transient, rarely persisting more than 6 weeks. One of the most severe was a median terminal nerve injury, presumably resulting from anterior portal placement, that was permanent.10 In one series of 368 patients who underwent unilateral or bilateral

total shoulder arthroscopies, 13 BP lesions occurred.41 Most of these, however, involved the supraclavicular plexus and were neurapraxic in nature, so most likely they were examples of classic postoperative paralysis. We saw a more severe injury, an AL upper trunk lesion, occur during shoulder arthroscopy, presumably as a result of positioning rather than direct instrumentation. A few series of rotator cuff tear repairs have been published that describe low rates of nerve injuries; most of the latter involved the supraclavicular nerve or one of the terminal nerves, and some were transient.10 Surgery for anterior glenohumeral instability (i.e., shoulder dislocations) probably has yielded the highest incidence of severe BP lesions. In one study concerned with 282 patients, 23 (8.2%) had a neurologic deficit postoperatively. Although several apparently had diffuse brachial plexopathies, the majority completely recovered within 4 to 6 months.10 These sharply contrast with the eight patients reported by Richards and co-workers, all of whom were referred for persistent neurologic deficits following PuttiPlatt and Bristow procedures, and all but one of whom underwent surgical exploration of the infraclavicular BP. Seven of these eight patients had, among them, six musculocutaneous neuropathies (most complete), caused by either nerve suture or laceration; two median neuropathies; two axillary neuropathies (both complete), one caused by suture and one by excess traction; and one radial neuropathy. One patient also sustained an axillary artery laceration. The eighth patient, who had awakened from his shoulder

1370

Diseases of the Peripheral Nervous System

surgery with a pulseless flail upper limb, had the following: musculocutaneous neuropathy (complete) resulting from transection; median neuropathy; sutures in all three cords, lateral, posterior and median; suture of the axillary artery; and laceration of the axillary vein. Despite these injuries, he made an excellent recovery following the operative repair of these neurovascular structures.55 A few reports have described delayed-onset infraclavicular lesions following modified Bristow procedures. In these cases, the bone screw used to attach the coracoid process to the glenoid rim came loose, pierced the axillary artery, and produced a pseudoaneurysm that, in turn, compromised the BP. In two patients, fully 3 years elapsed between the shoulder operation and the onset of this complication, which presented abruptly with upper limb pain and numbness, soon followed by weakness. Because these are neurovascular injuries, surgical intervention must be prompt if function, particularly of the hand, is to be restored.78 BP injuries have resulted from attempts to reduce shoulder dislocations since at least the time of Hippocrates. In earlier eras, the traction employed was so severe that supraclavicular root avulsion injuries occurred. Currently, however, infraclavicular elements typically are injured, particularly the axillary nerve. The pathophysiology varies from demyelinating CB to AL.9,78 Miscellaneous Lesions The combined use of vest and wrist restraints to manage agitated or combative patients has produced infraclavicular plexopathies, both unilateral and bilateral. Apparently, in most instances the medial cord has been injured. The pathophysiology has varied, with both demyelinating CB and AL being reported, the latter resulting in bilateral “claw-hand deformities” in one patient.12,61,78 Intra-arterial injection of chemotherapy (Cisplatin) via the axillary artery to treat malignant bone neoplasm caused a relatively diffuse AL brachial plexopathy that severely affected the posterior cord. Most symptoms did not appear until 3 days after the infusion, and then progressed for several days. The BP damage was attributed to direct neurotoxic effect of the drug, although it could have been secondary to progressive ischemia resulting from damage to the arterial supply of the BP.78 Axillary hematomas have produced infraclavicular plexopathies in patients who were receiving either coumadin or heparin. These have followed crutch use, falls, and simply vigorous grasping of a bed rail. Treatment has varied. At least one patient had permanent motor residuals in the hand. Most of these probably represent instances of the MBFC syndrome.18,20,78 Several persons with normal clotting mechanisms have developed infraclavicular BP lesions, unilateral or bilateral, while using crutches. These crutch palsies typically affect various terminal nerves in the axilla, especially the radial. The underlying pathophysiology usually is demyelinating CB, so recovery generally is fairly rapid.63,78 Infraclavicular

brachial plexopathies also have been described, usually as single case reports, following use of a tissue expander after modified radical mastectomy,60 after axillary arterial bypass performed to treat subclavian and innominate artery occlusive disease,39 and after eschar removal in the burned axilla.78

Thoracic Outlet Syndromes The term thoracic outlet syndromes should always be used in the plural, because it is an umbrella title for at least five separate entities (Table 55–3). True N-TOS and traumatic TOS have already been reviewed. Neither of the two types of vascular TOS directly affects the BP; although both manifest neurologic symptoms, these are secondary to limb ischemia. Consequently, the remainder of this discussion focuses on disputed N-TOS. This highly controversial disorder is considered to be nonexistent by some physicians and very common by others. Even among its proponents, almost none of its features are universally accepted. Because of this, it was named “disputed” N-TOS in 1984. Several other names have since been proposed, including “nonspecific,” “assumed,” and “symptomatic.” It became popular in 1966, when a surgical procedure was described for removing the first thoracic rib via the axilla (i.e., transaxillary first rib resection). At that time, based on a theory advanced in 1962, this disorder was attributed to compromise of the BP by compression between the normal first thoracic rib and some other structure (e.g., clavicle). Subsequently, however, its proponents have proposed other causes, including compression: (1) between the scalene muscles, (2) by a variety of congenital bands, and (3) resulting from postural abnormalities. Currently, one or more of these mechanisms are thought to be operative, usually activated by either a single episode of neck trauma (particularly whiplash injuries sustained in vehicular accidents) or minor, repetitive trauma (e.g., office or factory work). The majority of patients diagnosed with disputed N-TOS are young to middle-aged women, and both unilateral and bilateral lesions have been described. Its incidence has been estimated to be as high as 8%. The symptomatology is quite variable, but sensory complaints (paresthesias, pain, or both) radiating down the limb in a lower plexus distribution are the most common presentation reported.

Table 55–3. Thoracic Outlet Syndromes Neurologic True Disputed Vascular Arterial Venous

Neurovascular Traumatic (Disputed)

Brachial Plexus Lesions

Typically, few if any motor changes are present. The clinical examination is normal or reveals only minimal findings (e.g., “subtle” weakness of the triceps, supraclavicular fossa tenderness). The clinical maneuver most widely used to confirm the diagnosis is the elevated arm stress test (i.e., the EAST procedure), in which both upper limbs are held in the “stick ’em up” position for 3 minutes while the hands are rapidly opened and closed. A positive test consists of the limb sagging with fatigue. Various laboratory studies, particularly EDX procedures, have been proposed over the years to confirm the diagnosis, but none is universally accepted by proponents, and all have been challenged by skeptics, usually for multiple reasons. Although some patients respond to conservative therapy, many reportedly do not and, consequently, they undergo one or more operations, the two most popular being transaxillary first rib resection and anterior (and middle) scalenectomy. Often these surgical procedures provide only temporary relief, if any, so some patients undergo multiple operations. Moreover, these operations may have major complications, the most severe neurologically being an AL lower plexus injury (described earlier), and the most severe overall being death by exsanguination.16,78 One of the many problems with disputed N-TOS is that it most often is diagnosed by physicians who lack training and expertise in recognizing PNS disorders. Consequently, not only may patients undergo unnecessary (and sometimes damaging) surgery, but the actual cause of their symptoms may be overlooked.

REFERENCES 1. Aagaard, B. D., Maravilla, K. R., and Kliot, M.: MR neurography: MR imaging of peripheral nerves. Magn. Reson. Imaging Clin. N. Am. 6:179, 1998. 2. Alfonso, I., Alfonso, D. T., and Papazian, O.: Focal upper extremity neuropathy in neonates. Semin. Pediatr. Neurol. 7:4, 2000. 3. Alnot, J. Y.: Traumatic paralysis of the brachial plexus: preoperative problems and therapeutic indications. In Terzis, J. K. (ed.): Microreconstruction of Nerve Injury. Philadelphia, W. B. Saunders, p. 285, 1987. 4. Alnot, J. Y.: Traumatic brachial plexus lesions in the adult. Hand Clin. 11:623, 1995. 5. Bentolila, V., Nizard, R., Bizot, P., and Sedel, L.: Complete traumatic brachial plexus palsy. J. Bone Joint Surg. Am. 81:20, 1999. 6. Bhatt, B. M.: “Top cover neuropathy”—transient brachial plexopathy due to body armour. J. R. Army Med. Corps. 136:53, 1990. 7. Bilsky, M. H., Vitaz, T. W., Boland, P. J., et al.: Surgical treatment of superior sulcus tumors with spinal and brachial plexus involvement. J. Neurosurg. 97:301, 2002. 8. Birch, R., Bonney, G., and Wynn-Parry, C. B.: Surgical Disorders of the Peripheral Nerves. London, ChurchillLivingstone, 1998.

1371

9. Birch, R., Jessop, J., and Scott, G.: Brachial plexus palsy after manipulation of the shoulder. J. Bone Joint Surg. Br. 73:172, 1991. 10. Boardman, N. D., and Kofield, R. H.: Neurologic complications of shoulder surgery. Clin. Orthop. 368:44, 1999. 11. Bowen, B. B.: Brachial plexus. In Stark, D. D., and Bradley, W. G. (eds.): Magnetic Resonance Imaging, Vol. 3, 3rd ed. St. Louis, Mosby, p. 1821, 1999. 12. Bromberg, M. B., and Vogel, C. M.: Vest restraint palsy. Arch. Phys. Med. Rehabil. 77:1316, 1996. 13. Brunelli, G. A., and Brunelli, G. R.: A fourth type of brachial plexus injury: middle lesion (C7). Ital. J. Orthop. Traumatol. 18:389, 1992. 14. Chuang, D. C. C.: Management of traumatic brachial plexus injuries in adults. Hand Clin. 15:737, 1999. 15. Coene, L. N. J. E. M.: Mechanisms of brachial plexus lesions. Clin. Neurol. Neurosurg. 95(Suppl.):S24, 1993. 16. Dawson, D. M., Hallett, M., and Wilbourn, A. J.: Thoracic outlet syndromes. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds.): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 227, 1999. 17. de Laat, E. A. T., Visser, C. P. J., Coene, L. N. J. E. M., et al.: Nerve lesions and primary shoulder dislocations and humeral neck fractures. J. Bone Joint Surg. Br. 76:381, 1994. 18. Desai, D. C., Uribe, A., and Lachman, T.: Brachial plexus injury due to compression: an alternate mechanism of injury. Am. Surg. 63:482, 1997. 19. Dodds, S. D., and Wolfe, S. W.: Perinatal brachial plexus palsy. Curr. Opin. Pediatr. 12:40, 2000. 20. Elesber, A. A., Kent, P. D., and Jennings, C. A.: Compressive neuropathy of the brachial plexus and long thoracic nerve. Chest 120:309, 2001. 21. England, J. D., and Tiel, R. L.: AAEM Case Report 33: costoclavicular mass syndrome. Muscle Nerve 22:412,1999. 22. Frazier, C. H., and Skillern, P. G.: Supraclavicular subcutaneous lesions of the brachial plexus not associated with skeletal injuries. J. Am. Med. Assoc. 57:1957, 1911. 23. Ganju, A., Roosen, N., Kline, D. G., and Tiel, R. L.: Outcomes in a consecutive series of 111 surgically treated plexus tumors: a review of the experience at the Louisiana State University Health Sciences Center. J. Neurosurg. 95:51, 2001. 24. Gerrand, C. H.: False aneurysm and brachial plexus palsy complicating a proximal humeral extosis. J. Hand Surg. Br. 22:413, 1997. 25. Gilbert, A.: Long-term evaluation of brachial plexus surgery in obstetrical palsy. Hand Clin. 4:583, 1995. 26. Gilbert, A., and Whitaker, I.: Obstetrical brachial plexus lesions. J. Hand Surg. Br. 16:489, 1991. 27. Goga, I. E., and Bhigjee, A. I.: Brachial plexus lesions in soldiers—the “sandbag palsy.” S. Afr. Med. J. 77:643, 1990. 28. Gumina, S., and Postacchini, F.: Anterior dislocation of the shoulder in elderly patients. J. Bone Joint Surg. Br. 79:540, 1997. 29. Haymaker, R. W., and Woodhall, B.: Peripheral Nerve Injuries. Philadelphia, W. B. Saunders, 1953. 30. Heller, A. C., Kuether, T., Barnwell, S. L., et al.: Spontaneous brachial plexus hemorrhage—case report. Surg. Neurol. 53:356, 2000. 31. Hirasawa, Y., and Sakakida, K.: Sports and peripheral nerve injury. Am. J. Sports Med. 11:420, 1983.

1372

Diseases of the Peripheral Nervous System

32. Johannsen, S., Svensson, H., Larson, L.-G., and Denecamp, J.: Brachial plexopathy after postoperative radiotherapy of breast cancer patients. Acta Oncol. 39:373, 2000. 33. Karakaya, D., Baris, S., Güldogus, F., et al.: Brachial plexus injury during subclavian vein catheterization for hemodialysis. J. Clin. Anesth. 12:220, 2000. 34. Kim, D. H., Cho, Y.-J., Tiel, R. L., and Kline, D. G.: Outcomes of surgery in 1019 brachial plexus lesions treated at Louisiana State University Health Sciences Center. J. Neurosurg. 98:1005, 2003. 35. Kline, D. G., and Hudson, A. R.: Nerve Injuries. Philadelphia, W. B. Saunders, 1995. 36. Krivickas, L. S., and Wilbourn, A. J.: Peripheral nerve injuries in athletes: a case series of over 200 injuries. Semin. Neurol. 20:225,2000. 37. Leffert, R. D.: Clinical diagnosis, testing, and electromyographic study in brachial plexus traction injuries. Clin. Orthop. 237:24, 1988. 38. Levin, K., Wilbourn, A. J., and Maggiano, H. J.: Cervical rib and median sternotomy-related plexopathies: a reassessment. Neurology 50:1407, 1998. 39. Lowell, R. C., and Mills, J. L.: Critical evaluation of axilloaxillary artery bypass for surgical management of symptomatic subclavian and innominate artery occlusive disease. Cardiovasc. Surg. 1:530, 1993. 40. Lusk, M. D., Kline, D. G., and Garcia, C. A.: Tumors of the brachial plexus. Neurosurgery 21:439, 1989. 41. Lynch, N. M., Cofield, R. H., Silbert, P. L., and Hermann, R. C.: Neurologic complications after total shoulder arthroscopy. J. Shoulder Elbow Surg. 5:53, 1996. 42. Midha, R.: Epidemiology of brachial plexus injuries in a multitrauma population. Neurosurgery 40:1182, 1997. 43. Millesi, B. P.: Brachial plexus injuries: management and results. In Terzis, J. K. (ed.): Microreconstruction of Nerve Injuries. Philadelphia, W. B. Saunders, p. 347, 1987. 44. Murray, B. E., Wilbourn, A. J., and Shields, R. W.: The clinical and electrodiagnostic features of “classic” postoperative brachial plexopathy. Neurology 54(Suppl. 3):A48, 2000. 45. Nakamura, T., Yabe, Y., Horiuchi, Y., and Takayama, S.: Magnetic resonance myelography in brachial plexus injury. J. Bone Joint Surg. Br. 79:764, 1997. 46. Narakas, A. O.: The treatment of brachial plexus injuries. Int. Orthop. 9:29, 1985. 47. Narakas, A. O.: Pain syndromes in brachial plexus injuries. In Brunelli, G. A. (ed.): Textbook of Microsurgery. Milano, Masson, p. 809, 1988. 48. Narakas, A. O.: Lesions found when operating traction injuries of the brachial plexus. Clin. Neurol. Neurosurg. 95(Suppl.):556, 1993. 49. Oberle, J., Antoniadi, S. G., Kast, E., and Richter, H. P.: Evaluation of traumatic cervical nerve root injuries by intraoperative evoked potentials. Neurosurgery 51:1182, 2002. 50. Olsen, N. K., Pfeiffer, P., Johannsen, L., et al.: Radiationinduced brachial plexopathy: neurological follow-up in 161 recurrence-free breast cancer patients. Int. J. Radiat. Oncol. Biol. Phys. 26:43, 1993. 51. Ostdahl, R. H.: DREZ surgery for brachial plexus avulsion pain. In Nashold, B. S., and Pearlstein, R. D. (eds.): The

52.

53.

54.

55.

56. 57.

58. 59. 60.

61. 62.

63. 64.

65.

66.

67.

68.

69. 70.

71.

72.

DREZ Operation. Park Ridge, IL, American Association of Neurological Surgeons, p. 105, 1996. Pierce, S. M, Reict, A., and Lingos, T. I.: Long-term radiation complications following conservative surgery (CS) and radiation therapy (RT) in patients with early stage breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 23:915, 1992. Ramdial, P., Sing, B., Moodley, J., and Haffejee, A. A.: Brachial plexopathy after subclavian vein catheterization. J. Trauma 54:786, 2003. Ricci, C., Rendina, E. A., Venuta, F., et al.: Superior pulmonary sulcus tumors: radical resection and palliative treatment. Int. Surg. 74:175, 1989. Richards, R. C., Hudson, A. R., Bertoia, J. T., et al.: Injury to the brachial plexus during Putti-Platt and Bristow procedure. Am. J. Sports Med. 15:374, 1987. Rozing, P. M.: The topographical anatomy of the brachial plexus. Clin. Neurol. Neurosurg. 95 (Suppl.):S12, 1993. Samii, M.: Special considerations regarding the treatment of brachial plexus lesions. In Samii, M. (ed.): Peripheral Nerve Lesions. Berlin, Springer-Verlag, p. 372, 1990. Sandmire, H. F., and DeMott, R. K.: Erb’s palsy: concepts of causation. Obstet. Gynecol. 95:941, 2000. Sedel, L.: Repair of severe traction lesions of the brachial plexus. Clin. Orthop. 237:62, 1988. Sinow, J. D., and Cunningham, B. L.: Postmastectomy brachial plexus injury exacerbated by tissue expansion. Ann. Plast. Surg. 27:368, 1991. Skeen, M. B., Rozear, M. P., and Morganlander, J. C.: Posey palsy. Ann. Intern. Med. 117:795, 1992. Stewart, M. P. M., and Birch, R.: Penetrating missile injuries of the brachial plexus. J. Bone Joint Surg. Br. 83:517, 2001. Subramony, S. H.: Electrophysiological findings in crutch palsy. Electromyogr. Clin. Neurophysiol. 29:281, 1989. Suematsu, N., Matsuura, J., and Atsuta, Y.: Brachial plexus injury caused by electric current through the ulnar nerve. Arch. Orthop. Trauma Surg. 108:400, 1989. Sylvestred, L., Sandson, T. A., and Nachmanoff, D. B.: Transient brachial plexopathy as a complication of internal jugular vein cannulation. Neurology 41:760, 1991. Tarng, D.-C., Huang, T.-O., and Lin, K.-P.: Brachial plexus compression due to subclavian pseudoaneurysm from cannulation of jugular vein hemodialysis catheter. Am. J. Kidney Dis. 31:694, 1998. Tavakkolizadeh, A., Saifuddin, A., and Birch, R.: Imaging of adult brachial plexus traction injuries. J. Hand Surg. Br. 26:183, 2001. Tebig, J. G., Bascoulergue, J., Dumontet, C., et al.: Brachial plexus compression by an iatrogenic arteriovenous fistula. Clin. Rheumatol. 6:593, 1987. Terzis, J. K., and Papakonstantinou, K. C.: Management of obstetric brachial plexus palsy. Hand Clin. 15:717, 1999. Terzis, J. K., and Papakonstantinou, K. C.: The surgical treatment of brachial plexus injuries in adults. Plast. Reconstruct. Surg. 106:1097, 2000. Terzis, J. K., Vekris, M. D., and Soucacos, P. N.: Outcomes of brachial plexus reconstruction in 204 patients with devastating paralysis. Plast. Reconstr. Surg. 104:1221, 1999. Thompson, D. A., Flynn, T. C., Miller, P. W., and Fischer, R. P.: The significance of scapular fractures. J. Trauma 25:974, 1985.

Brachial Plexus Lesions 73. Trentman, T. L., Rome, J. D., and Messick, J. M.: Brachial plexus neuropathy following attempt at subclavian vein catheterization. Reg. Anesth. Pain Med. 21:163, 1996. 74. Tsao, B. E., and Wilbourn, A. J.: The medial brachial fascial compartment syndrome following axillary arteriography. Neurology 61:1037, 2003. 75. Van Dijk, J. G., Pondaag, W., and Malessy, M. J. A.: Obstetric lesions of the brachial plexus. Muscle Nerve 24:1451, 2001. 76. Veilleux, M., Bourgouin, P., and Morin, J.-F.: Brachial plexopathy secondary to subclavian-axillary artery aneurysm. Muscle Nerve 19:92, 1996. 77. Visser, C. P. J., Coene, L. N. J. E. M., Brand, R., and Tavy, D. L. J.: The incidence of nerve injury in anterior dislocation of the shoulder and its influence on functional recovery. J. Bone Joint Surg. Br. 81:679, 1999. 78. Wilbourn, A. J.: Brachial plexus disorders. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, Vol. 2, 3rd ed. Philadelphia, W. B. Saunders, p. 911, 1991. 79. Wilbourn, A. J.: Peripheral nerve disorders in electrical and lightening injuries. Semin. Neurol. 15:241, 1995. 80. Wilbourn, A. J.: Brachial plexopathies. In Katirji, B., Kaminsky, H. J., Preston, D. C., et al. (eds.): Neuromuscular

81.

82. 83. 84.

85. 86.

87.

88.

1373

Disorders in Clinical Practice. Boston, ButterworthHeinemann, p. 884, 2002. Wilbourn, A. J.: Nerve conduction studies: types, components, abnormalities, and value in localization. Neurol. Clin. 20:305, 2002. Wilkinson, M. C. P., and Clarke, J. A.: Burns to the brachial plexus. Injury 23:341, 1992. Wilson-Pauwels, L., Stewart, P. A., and Akesson, E. J.: Autonomic Nerves. Hamilton, BC, Decker, 1997. Wolstock, D., and Millesi, H.: Brachial plexus—applied anatomy and operative exposure. In Gelberman, R. H. (ed.): Operative Nerve Repair and Reconstruction, Vol. 2. Philadelphia, J. B. Lippincott, p. 255, 1991. Wynn-Parry, C. B.: Brachial plexus injuries. Br. J. Hosp. Med. 32:130, 1984. Yoshida, T., Tada, K., Uemura, K., and Yonenobu, K.: Peripheral nerve palsies in victims of the Hanshin-Awaji earthquake. Clin. Orthop. 326:208, 1999. Zeidman, S. M., Rossitch, E. J., and Nashold, B. S.: Dorsal root entry zone lesions in the treatment of pain related to radiationinduced brachial plexopathy. J. Spinal Disord. 6:44, 1993. Zimmerman, N. B., and Weiland, A. J.: Assessment and monitoring of brachial plexus injury in the adult. In Gelberman, R. H. (ed.): Operative Nerve Repair Reconstruction, Vol. 2. Philadelphia, J. B. Lippincott, p. 1273, 1991.

56 Lumbosacral Plexus Lesions MICHAEL DONAGHY

Anatomy Diagnosis Investigation Causes Hemorrhage Trauma

Intra-arterial Injections Ischemia Obstetric and Gynecologic Complications Surgical Conditions and Operations

The bony ring of the pelvis encases most of the lumbosacral plexus. This protects it from direct trauma unless of severity sufficient to fracture-dislocate the pelvic bones. Thus the lumbosacral plexus is less susceptible to trauma than the brachial plexus, which lies in a more exposed location. In contrast, its tumors are relatively concealed from direct palpation. The introduction of computed tomographic imaging of the retroperitoneal region has dramatically simplified the differential diagnosis of lumbosacral plexus lesions. The following discussion of lumbosacral plexus disorders includes lesions affecting the portions of the major leg nerves that lie within the pelvis. It does not include detailed discussion of lumbosacral radiculoplexus neuropathy (see Chapter 107), vasculitic neuropathy (see Chapter 106), or diabetic proximal neuropathy (see Chapter 85). However, these disorders are mentioned briefly as the differential diagnoses of other disorders affecting the lumbosacral plexus.

ANATOMY In reality the lumbosacral plexus consists of separate lumbar and sacral plexuses, located above and below the pelvic rim, respectively. The lumbar plexus sends a large component to the sacral plexus via the lumbosacral cord, which runs across the pelvic brim (Fig. 56–1).

Tumors Irradiation Infections Hereditary Liability to Pressure Palsies Inflammatory Plexopathies

The lumbar plexus is derived from the L1-L3 ventral rami. The L1 root receives a contribution from T12, and the L3 receives a large contribution from L4. The plexus lies on the posterior aspect of the psoas major muscle in front of the vertebral transverse processes, roughly in the same plane as the intervertebral foramina. Muscular branches arising directly from the plexus supply the following muscles: quadratus lumborum (T12, L1-L4), psoas minor (L1), psoas major (L2, L3[4]), and iliacus (L2, L3). The lumbar plexus gives rise to the femoral nerve (posterior rami L2-L4), which innervates the quadriceps femoris muscle and receives the medial and intermediate cutaneous nerves of the thigh and the saphenous nerve from the medial calf. The obturator nerve (anterior rami L2-L4) innervates the adductor of the thigh muscles and the skin of the medial thigh. The lumbar plexus also gives rise to other predominantly sensory nerves supplying the abdominal wall, groin, and perineum. The iliohypogastric nerve (L1) supplies the transversus abdominus and internal oblique muscles, with a lateral cutaneous branch to the posterolateral gluteal skin and an anterior cutaneous branch to the suprapubic skin. The ilioinguinal nerve (L1) supplies the internal oblique muscle and traverses the inguinal canal to supply the skin of the proximomedial thigh, the skin over the penile root and upper scrotum, or the mons pubis and adjacent labium majus. The genitofemoral nerve (L1, L2) has a genital branch that traverses the inguinal canal and supplies the cremaster and the scrotal 1375

1376

Diseases of the Peripheral Nervous System FIGURE 56–1 Lumbosacral plexus anatomy. Diagrammatic representation of the lumbosacral plexus showing characteristic locations of discrete lesions caused by mechanical factors. A, Lumbosacral cord compression at the pelvic brim by the fetal head or midpelvic forceps causing maternal obstetric paralysis. B, Compression of the femoral nerve by surgical retractor blades in the gutter between the iliacus and the psoas muscles. C, Compression of the femoral nerve by hematoma within the iliacus fascia. D, Lumbosacral trunk damage by fracture-dislocation of the sacroiliac joint. E, Angulation of the femoral nerve under the inguinal ligament during prolonged flexion and abduction of the hips in the lithotomy position.

skin, or the mons pubis and labium majus, and a femoral branch that supplies the skin overlying the upper femoral triangle. The lateral cutaneous nerve of the thigh (L2, L3) supplies the anterolateral skin of the thigh as far distally as the knee. The sacral plexus is derived from the S1-S3 ventral rami, with a contribution from the S4 ventral ramus, and receives the lumbosacral cord (L4, L5). It overlies the lateral sacrum and posterolateral pelvic wall, anterior to the piriformis muscle but posterior to the internal iliac vessels, urethra, and sigmoid colon. The main branch is the sciatic nerve (L4-L5, S1-S3), which supplies the hamstring muscles, all the muscles below the knee, and all the skin below the knee except the area of the medial calf, which is supplied by the saphenous nerve. By the level of the sciatic notch, the sciatic nerve has already divided into its two component fascicles, which subsequently give rise to the separate common peroneal and tibial nerves within the posterior thigh or popliteal fossa. Of these two fascicles, the common peroneal component lies more superficially and is therefore more vulnerable to damage in buttock injuries, including badly sited injections. Muscular branches arising directly from the sacral plexus include the superior gluteal nerve (L4, L5, S1) supplying the gluteus medius and minimus muscles, the inferior gluteal nerve (L5, S1, S2) to the gluteus maximus muscle, and the nerve to the piriformis (S1, S2). The posterior femoral cutaneous nerve (S1-S3) enters the buttock through the greater sciatic notch, lying posteromedial to the sciatic nerve and supplying the back of the thigh, the adjacent gluteal region, and the posterior perineum. The pudendal nerve (S2-S4) also leaves the pelvis through the greater sciatic notch before emerging through the lesser

sciatic foramen to innervate the lower rectum, perineum, scrotum, and labium and to provide the dorsal nerve of the penis or clitoris. The blood supply of the lumbar and sacral plexuses has been studied in the human fetus.35 There is considerable variability in the precise pattern. Much of the lumbar plexus receives a segmental pattern of blood supply. Branches of the internal iliac artery, the superior gluteal or iliolumbar arteries, supply the roots and ganglia of L5 to S3 segments, and the inferior gluteal artery supplies the S2-S4 segments. The lumbosacral plexus is vulnerable to externally applied indirect trauma only if the pelvic ring is disrupted by double fracture-dislocations or if traction injuries result from dislocation of the hip joint.78,79 The femoral nerve is vulnerable to compression as it lies in the gutter between the psoas and iliacus muscles just above the inguinal ligament. Such pressure may arise medially from surgical retractors (Fig. 56–1B)56,141 or laterally from hematoma beneath the iliacus fascia (Fig. 56–1C).116,134,153,154 The lumbosacral cord is vulnerable to compression by the fetal head or obstetric forceps as it passes across the pelvic brim (Fig. 56-1A),31 by traction in posterior pelvic ring fractures, or by aneurysms of the common iliac or other pelvic arteries in the presacral area (Fig. 56–1D).19,28 The femoral nerve may be compressed by extreme angulation under the inguinal ligament during prolonged flexionabduction of the thighs in the lithotomy position under general anesthesia (Fig. 56–1E).117 The common peroneal division of the sciatic nerve is more vulnerable to stretch injuries than the tibial division because of its angulation at the sciatic notch above and fixation at the fibula neck below.126

1377

Lumbosacral Plexus Lesions

DIAGNOSIS Clinical suspicion of a lesion localized to the lumbosacral plexus should be aroused by demonstration of a pattern of weakness, loss of reflexes, and sensory disturbance that cannot be explained by damage either to a single root or to an individual peripheral nerve. Unsurprisingly, early plexus lesions are often so anatomically restricted that they fail to satisfy these criteria. Furthermore, simultaneous involvement of plexus, roots, and peripheral nerves can occur with pathologic processes such as infiltrating tumors or diabetes. Ease in diagnosing a plexus lesion may be hindered by a history of diabetes mellitus, laminectomy for prolapsed intervertebral disc, or chemotherapy with neurotoxins such as vincristine, cisplatin, or paclitaxel. A lesion of the lumbosacral plexus should be particularly suspected when neurologic abnormalities develop in the leg in the context of an abdominal pelvic operation, childbirth, intravenous heroin abuse, genital herpes, abdominopelvic malignancy, pelvic radiotherapy, diabetes mellitus, buttock injections, fracture-dislocation of the pelvic ring, hip joint dislocation, or a hemorrhagic diathesis. In practice, the common difficulty is to distinguish lumbosacral plexus from root lesions. Although augmentation of radiating “sciatica” pain by straight leg raising or coughing is commonly regarded as diagnostic of intraspinal root lesions, this Lasègue’s sign also accompanies plexus lesions from a variety of causes, particularly tumor or aneurysm.28,146 It can be impossible to differentiate multiple root lesions from a plexus lesion purely on clinical grounds. An underrecognized pointer to a plexus lesion, as opposed to a root lesion, is a warm, dry, red foot resulting from disturbed autonomic function.59 This reflects involvement of the retroperitoneal lumbar sympathetic nerves in the region anterolateral to the vertebral bodies. Involvement of the proximal leg muscles is a useful pointer to a plexus rather than a peripheral nerve lesion.120 Gluteal muscle power will be preserved in a pure sciatic nerve lesion because the inferior and superior gluteal nerves arise directly from the plexus. If the ankle jerk is preserved, it can be difficult to differentiate between lumbosacral cord, partial sciatic nerve, and common peroneal nerve palsies as a cause of footdrop. Iliopsoas power will be preserved in obturator or femoral nerve lesions because its innervation arises directly from the lumbar plexus. Femoral neuropathy can be confused with lesions of the L4 root or the lower lumbar plexus because the L4 dermatome and the long saphenous nerve share similar cutaneous territories. Such difficulty may be resolved by demonstrating weak ankle inversion (tibialis anterior) and hip internal rotation (gluteus medius), which cannot be accounted for by a lesion of the femoral nerve.

The general medical examination may provide clues to the nature of a lumbosacral plexus lesion. Digital examination of the rectum or vagina may reveal unexpected gynecologic, prostatic, or colorectal cancers, or an aneurysm of a pelvic artery. Inspection of the buttocks, external genitalia, uterine cervix, or anus may reveal herpes zoster eruptions or herpes simplex ulceration.

INVESTIGATION In practice, most patients harboring plexus lesions are referred to the neurologist only when magnetic resonance imaging (MRI) of the spine has failed to reveal the anticipated spinal nerve root lesion. It must be appreciated that the plane and extent of imaging used to detect spinal lesions in routine clinical practice is not sufficiently extensive to detect plexus lesions. Thus, if a plexus lesion is sought, further MRI should be requested specifying the need for axial imaging of the full width of the pelvis and extending over the full longitudinal extent of the plexus, that is, from the L1 vertebra down to the level of the inguinal ligament/greater sciatic notch. MRI and computed tomography (CT) scanning are the main ways to demonstrate structural lesions affecting the lumbosacral plexus. Although most of the initial use of scanning to detect such lesions was based on CT, MRI offers improved definition of pelvic soft tissue planes and may enhance detection of tumors infiltrating the lumbosacral plexus. Pelvic CT scanning visualizes iliacus hematomas,45,99,116,134 solid tumors,85,104,121,129,140 erosion of vertebrae or their transverse processes, aneurysms, or endometriosis in the sciatic notch.47,75,113 Pelvic CT scan detects a tumor in 96% of patients presenting with neoplastic lumbosacral plexus lesions.85 CT scanning is particularly useful in differentiating between tumor plexopathy and radiation plexopathy, although it fails to detect up to 22% of tumors in this setting129; even better sensitivity can be expected with MRI. The valuable clue of bone erosion resulting from neoplastic infiltration may be undetected by the wrong CT scan “window” settings. Detailed atlases of tomographic anatomy of the lumbosacral plexus are available.37,55,140 Electrophysiologic investigation is of limited value in defining plexus lesions. In general, electromyography often simply documents patterns of muscular involvement, which can usually be delineated by careful clinical examination. In radiation plexopathy, electromyography of moderately weakened leg muscles may reveal characteristic low-frequency periodic myokymic discharges.2,4,129 Occasionally, similar myokymic discharges occur in other chronic disorders of the peripheral nerves or spinal cord.

1378

Diseases of the Peripheral Nervous System

Electromyography of the paraspinal muscles will be normal in uncomplicated plexus lesions.98,148 It may be abnormal in more diffuse neuropathic conditions such as diabetes mellitus and radiation plexopathy. Comparison of the sural nerve sensory action potentials in the two legs can be a valuable guide to unilateral postganglionic sensory nerve fiber lesions within the lower lumbosacral plexus. The sural sensory nerve action potential is not usually affected by preganglionic lesions restricted to the nerve roots.6,131,148 Generalized demyelinating peripheral neuropathy will be detected by measuring motor conduction velocities over peripheral segments of peripheral nerves.98 Other electrophysiologic investigations, such as electromyographic evidence of denervation, delay of F responses or H reflexes, somatosensory evoked potentials, or responses to magnetoelectrical stimulation, do not conclusively distinguish between radiculopathy and plexopathy.6,30 Examination of the spinal fluid is not a routine investigation for suspected lumbosacral plexus lesions. However, it may have been undertaken when the differential diagnosis included multiple root lesions resulting from possible malignant meningitis or inflammatory disorders. The spinal fluid will be normal in lesions entirely restricted to the plexus. However, the protein content may be elevated in conditions primarily involving the plexus but that include more diffuse neurologic involvement, such as diabetes mellitus or radiation plexopathy.

CAUSES Hemorrhage Two distinct syndromes of lumbosacral plexus compression by retroperitoneal hemorrhage are recognized (Fig. 56-2). First, the lumbar plexus may be compressed diffusely within the psoas muscle, leading to weakness both in the obturator and femoral nerve territories.21,45,60,89,155 In the second, the intrapelvic femoral nerve is compressed in isolation by hematoma within the adjacent iliacus muscle.3,64,107,116,127,134,144,149,153,154 The iliacus fascia is less capacious than the fascia overlying the psoas muscle.29 Hence smaller hematomas are required to produce isolated femoral neuropathy rather than diffuse lumbar plexus damage. Iliacus or psoas muscle hemorrhage most frequently accompanies anticoagulant therapy, usually with heparin rather than warfarin, and may result from trivial trauma in this context.3,45,115,116,134,149,153–155 Plexus lesions occur less frequently in other disturbances of coagulation: disseminated intravascular coagulation,60,144 hemophilia,64,127 leukemia,21 rupture of aneurysm,89,107 localized blunt trauma,57 and athletic injuries involving hip hyperextension

and avulsion of the iliacus muscle from the ilium.66 It is possible that iliacus or psoas hematoma sometimes results from bleeding into injection sites in the buttocks in patients with a bleeding diathesis.153 Femoral nerve compression by iliacus hematoma is signaled by pain in the groin or lower iliac fossa radiating to the anteromedial thigh and medial lower leg. Hip extension exacerbates the pain, and patients characteristically choose to lie with the hip flexed. Passive rotation at the hip joint is relatively painless, unlike septic arthritis or hemarthrosis of the hip joint. The quadriceps muscle is weak and the knee jerk depressed or absent; variable sensory loss may be present on the anteromedial thigh and in the saphenous nerve territory. Iliacus fascia hematoma may be palpable in the lower iliac fossa or may track down to produce a visible and palpable mass in the groin. Unlike iliacus hematoma, psoas muscle hematomas do not cause profound pain on forced hip extension, the hip is not held flexed, no hematoma is palpable or visible in the iliac fossa or groin, and there is weakness of the thigh adductors indicating involvement of the obturator nerve territory. Hemorrhagic plexus or femoral nerve lesions are only rarely bilateral.60 The principal differential diagnosis of this acutely evolving syndrome is abscess secondary to appendicitis3 or following laparotomy.134 Hematoma of the psoas or iliacus muscles can be confidently detected and differentiated from abscess by CT scan.45,116,134,154 Restoration of normal blood coagulation is the mainstay of treatment. Extensive blood loss may cause hypovolemic shock requiring transfusion, particularly with hemorrhage into the capacious psoas muscle.29 Although good recovery with medical treatment is the rule,119 prolonged neurologic disability may follow iliacus hematoma.153 The impression that surgical drainage of iliacus hematoma may improve the chances of neurologic recovery134,153 has not been confirmed by a controlled trial.

Trauma In civilian practice traumatic lumbosacral plexus lesions generally result from double vertical fracture-dislocations of the bony pelvic ring.78,79 Neurologic disturbance occurs in half of all patients with double pelvic fractures.79,145 The neurologic abnormality lies on the same side as the sacroiliac joint injury. The lumbosacral cord is the component of the plexus most often affected, leading to signs in L5- and S1-innervated muscles.79 Autopsy studies show that rupture, compression, and traction injuries generally affect the lumbosacral trunk, the obturator or superior gluteal nerves, or the L5, S1, S2, or S3 anterior rami.78 The lumbosacral cord is 30 to 32 mm

Lumbosacral Plexus Lesions

FIGURE 56–2 Hematoma adjacent to the iliacus and iliopsoas muscles following cesarean section. MRI showing the longitudinal extent of hematoma on sagittal (A), longitudinal (B) and axial (C) slices. The relationships to the femoral nerve and vessels (B, arrow) and to the psoas muscle (C, arrow) are shown. (Courtesy of Niall Moore, FRCR.)

1379

1380

Diseases of the Peripheral Nervous System

long and has little mobility, being relatively fixed to the sacral ala, thereby rendering it particularly vulnerable to injury and fracture of the posterior pelvic ring.42 Avulsion of the lumbosacral nerve root usually affects L5, S1, or L4, in that order, happens distal to the point of emergence from the spinal canal, is associated with pseudomeningocele formation, and is visualizable on MRI.96 Often the neurologic disturbance is initially overlooked because of the pain of bony injury and the coexistent disruption of the urethra. Traction injuries of the plexus may result from fracture-dislocation of the hip joint and acetabulum.122,145 Breech delivery is a rare cause of traction injury to the plexus in neonates.77 The neurologic disability is commonly permanent after traumatic lumbosacral plexus lesions. Lumbosacral plexus lesions should be recognized as an important cause of delayed recovery of limb movement following pelvic fractures. Lumbosacral plexus neuropathy, with full clinical recovery, has been reported in a 13-month-old child who had experienced an abdominal crush injury in the absence of visceral or bony pelvic injury.44

Intra-arterial Injections Injections into the buttock can injure the lumbosacral plexus by means of ischemia rather than the familiar means of direct trauma to the sciatic nerve.123 Buttock injections may inadvertently introduce vasotoxic or crystalline drugs into the inferior gluteal artery in the medial aspect of the buttock. Such unintended intravascular injection is more likely if the aspiration test is omitted before injection. Ischemia of the sciatic nerve is thought to result from vascular obstruction by crystals or toxic vasospasm. Severe painful sensorimotor disturbance develops in the leg ipsilateral to the injection after a delay ranging from minutes to a few hours. More widespread lumbar plexus lesions may follow buttock injections and are presumed to result from retrograde extension of gluteal artery spasm to other branches of the internal iliac artery.123 The buttock skin characteristically becomes painfully swollen and cyanosed, and gangrene can develop (embolia cutis medicamentosa). This results from involvement of the artery to the buttock skin. Only partial recovery of the neurologic lesion can be expected, and severe pain may persist. On suspicion of the diagnosis, immediate administration of papaverine, heparin, and sympathetic blockade is recommended,123 although these measures are of unproven value. Usually the diagnosis of ischemic injury of the plexus resulting from intra-arterial injection is overlooked in favor of the more familiar diagnosis of direct needle injury to the sciatic nerve. Immediate onset of painless lumbosacral plexopathy has followed therapeutic injection of cisplatin or fluorouracil into the internal iliac artery.104

Ischemia Neurologic deficits localizable to the lumbosacral plexus or the proximal sciatic or femoral nerves have been noted in aortoiliac vascular disease.34 The patients presented with pain, weakness, or numbness in a leg. Aneurysmal, stenotic, or postoperative vascular lesions affected the distal aorta or proximal iliac vessels. Diabetics undergoing renal transplantation seem to be vulnerable to a presumed ischemic lesion of the lumbosacral plexus when the internal iliac arteries are used for revascularization of the graft.72 These patients present postoperatively with weakness of the leg below the knee and buttock pain. Recovery may be incomplete. Intermittent claudication of buttock and leg muscles has been attributed to ischemia of the lumbosacral plexus. This rare and generally unrecognized syndrome typically involves the following exercise-induced sequence of symptoms: pain, tingling, and sensory and motor deficits.151 Pain is mostly localized to the pelvis and is usually associated with stenosis of pelvic arteries, particularly the internal iliac. A steal phenomenon is hypothesized to be provoked by activity of the more distal leg muscles. Symptoms recover after resting for a few minutes. The differential diagnoses are the more familiar conditions of intermittent claudication resulting from more distal arterial occlusive disease of the legs and spinal canal stenosis. Successful treatment with interventional radiologic endovascular therapy has been reported.152 Buttock claudication can also occur after endovascular stent-graft repairs of abdominal aortic aneurysms that involve iliac limb extension to the external iliac artery because the common iliac artery is ectatic or aneurysmal.33

Obstetric and Gynecologic Complications Postpartum footdrop is most common after protracted labor, cephalopelvic disproportion, or midpelvic forceps delivery.20,26,31,62,86 The incidence has been estimated at 1 in 2000 to 2500 deliveries in North American obstetric practice.58,147 The lumbosacral cord is directly compressed at the pelvic brim or as it overlies the sacroiliac joint. Postpartum footdrop is rarely bilateral86 and usually occurs on the side compressed by the infant’s brow in an occiput anterior presentation.38 Weakness of ankle dorsiflexion and eversion are noticed when the woman tries to walk postpartum. Examination also reveals sensory loss on the lateral lower leg and dorsum of the foot. The differential diagnosis includes common peroneal nerve palsy and an L5 root lesion. Complete spontaneous recovery is usual within 3 months, although the disability can be permanent. Less commonly, the postpartum weakness involves the quadriceps muscle. If so, it is bilateral in 25% and may be accompanied by obturator neuropathy.24,38,101 Multiple

1381

Lumbosacral Plexus Lesions

causes for puerperal femoral nerve lesions have been invoked. The lithotomy position may compress the femoral nerve focally under the inguinal ligament, particularly if vaginal delivery is performed under general anesthesia.86 Separation of the symphysis pubis during delivery can allow direct compression of the femoral nerves by the fetal head.24 Bilateral quadriceps weakness has followed epidural anesthesia, with direct involvement of the lumbar plexus being suggested by psoas muscle involvement.1 Paracervical block of the pudendal nerves may be followed by pain radiating down the posterior thigh some days later, and a pelvic hematoma may be palpated vaginally.54 During pregnancy the lumbosacral plexus may be compressed at the pelvic brim by uterine leiomyoma71 or by rapid growth of an intrapelvic schwannoma.43 The management of a subsequent delivery in a woman who has had a prior episode of puerperal pelvic nerve compression depends on whether full recovery had previously occurred.4 Elective cesarean section is advised if there is residual neurologic damage indicative of axonal degeneration. If full recovery had occurred, a trial of labor should be undertaken but the use of midpelvic forceps avoided. If signs of a lumbosacral plexus lesion develop during labor, the labor should be allowed to continue normally because cesarean section is unlikely to reverse the nerve damage; however, midpelvic forceps should again be avoided.62 A lower sacral plexopathy has been described after vaginal delivery.81 Patients had persistent sensory disturbance in the perineum and presented with urinary (stress incontinence or dysuria), anorectal (fecal incontinence or dyskinesia), or sexual (hypo-orgasmia or anorgasmia) dysfunction. The most common electrophysiologic abnormality was prolonged bulbocavernosus muscle reflex latency. Half the patients had reduced urethral closure pressure on urodynamic investigation. There had been a high incidence of multiparity and midforceps rotation during delivery, and the mechanism of nerve damage is proposed to be direct trauma. Apparent compression of pelvic nerves has occurred in the third trimester of a teenage pregnancy.133 Catamenial sciatica is a rare syndrome resulting from implantation of endometriosis in the sciatic nerve at the sciatic notch. This can be a developmental anomaly associated with outpouching of a pocket of peritoneum into the sciatic notch. Progressive sensorimotor sciatic nerve palsy in women of childbearing age is associated with perimenstrual pain in the buttock or posterior thigh pain.9,70,75,113 Pelvic CT shows a contrast-enhancing mass in the sciatic notch.47,75,113 Relief of pain can be achieved by treatment with danazol or progesterone or by induction of an artificial menopause.38 If fertility must be preserved, microsurgical removal of the endometriosis can be successful.70,113 Indeed, surgical exploration may be necessary simply to obtain a tissue diagnosis once a mass in the sciatic notch has been demonstrated by CT.

Surgical Conditions and Operations Laterally placed self-retaining retractor blades may compress the femoral nerve along its course in the gutter between the iliacus and psoas muscles in the lower pelvis during pelvic operations such as abdominal hysterectomy56,110,141,150 or renal transplantation.105,138 This complication may be avoided by ensuring that the femoral artery remains pulsatile after initial positioning of the retractors.110 A variety of pelvic mononeuropathies, sometimes multiple, have been documented after gynecologic surgery, including symptoms spanning the extent of the lumbosacral plexus after uncomplicated abdominal hysterectomy.5 Lumbosacral plexus, sciatic, and femoral neuropathies may follow vaginal hysterectomy or other operations on the perineum. 22,48,117 Excessive stretching or angulation of the nerves by hyperabduction of the thighs under anesthesia is the probable mechanism. Symptomatic sciatic, femoral, or obturator nerve palsies were noted after hip joint replacement in 0.7% to 1% of patients in large retrospective series.12,118,143 Subclinical nerve damage may be detected in up to 70% of prospectively studied recipients of hip prostheses.143 Various causes have been implicated: stretch during perioperative hip dislocation, local hemorrhage, heat and solvent toxicity from methylmethacrylate bone cement, direct trauma, trauma from retractor blades, and postoperative aneurysm formation.103,143 Aneurysm of the iliac or hypogastric arteries in the pelvis may compress the lumbosacral plexus (Fig. 56–3).19,28,142 The presentation with an abrupt onset of sciatic pain and impaired straight leg raising may suggest prolapsed intervertebral disc. However, rectal examination often reveals a firm, pulsatile mass,28 and curvilinear calcification in the aneurysm wall may be evident on plain radiology. Vascular surgical treatment by a replacement graft is effective28 but may itself create ischemic plexus lesions.135 Without surgery, there is a high chance of aneurysmal rupture, which can be fatal,19 and the resulting retroperitoneal hemorrhage can compress the femoral nerve.107

Tumors Primary nerve sheath tumors of the lumbosacral plexus are uncommon except in patients with von Recklinghausen’s disease.14,80 They are easily visualized by CT scanning or MRI (Figs. 56–4 and 56–5). Surgical extirpation is effective, but a conservative approach should be considered in asymptomatic or stable patients.14 Rare causes of plexus compression by benign tumor include an omental dermoid cyst92 and uterine leiomyoma.71,91 A wide variety of malignant carcinomas and lymphoreticular tumors can invade the lumbosacral plexus either by direct invasion from the primary site (colorectum,

1382

Diseases of the Peripheral Nervous System

FIGURE 56–3 Iliac artery aneurysm compressing the sacral plexus. Axial CT showing a thrombosed iliac artery aneurysm extending to the pelvic side wall. A narrow lumen, visible medially (arrow), is of smaller caliber than the common iliac artery on the opposite side. (Courtesy of Jeremy Perkins, FRCS.)

uterus, prostate, ovary) or by metastatic spread (breast, sarcoma, lymphoma, bronchus, testis, myeloma, thyroid, melanoma). Locally invasive colorectal carcinoma is the most common cause of neoplastic plexus lesions.7,74,85,129 Persistent severe unilateral local or radicular pain is the first symptom. A predominantly proximal motor disturbance inexorably progresses over subsequent weeks or months. The initial pain may closely resemble that of lumbar disc prolapse146 or meralgia paresthetica.49,125 A mass

may be palpable rectally. MRI or CT scanning is usually diagnostic (Fig. 56–6). Pelvic CT is abnormal at initial presentation in 78%129 to 96%85 of patients and eventually becomes positive in all patients.85 The mode of therapy is determined by the known responsiveness of the tumor type in question. Local radiotherapy varies in its effectiveness in relieving pain.7,85,104 In one series,129 86% of patients with malignant tumor plexopathy had died within 1 year of diagnosis (Fig. 56–7).

Irradiation

FIGURE 56–4 MRI showing tumor involving the lumbar plexus on the posteromedial border of the psoas muscle. Note the wasting of the psoas muscle on the affected side. (From Donaghy, M.: Brain’s Diseases of the Nervous System, 11th ed. Oxford, UK, Oxford University Press, 2001, with permission.)

Pelvic CT and MRI have enormously simplified the differential diagnosis between radiotherapy plexopathy and tumor infiltration of the lumbosacral plexus. Radiation-induced lumbosacral plexopathy should be distinguished from two much rarer complications of radiotherapy: the postirradiation lower motor neuron syndrome16 and radiation-induced nerve sheath tumors. Radiation plexopathy typically presents 5 years after external or internal cavity irradiation of the pelvis for lymphoreticular, testicular, uterine, or ovarian malignancies.2,8,129 The delay can vary from 1 to 31 years.129 Technical variations in reported irradiation schedules make it impossible to estimate the “safe” or threshold radiation dosage before plexus lesions occur.129 Slowly progressive leg weakness is the most common presenting symptom. It is generally bilateral and starts in the distal muscles, unlike the unilateral proximal weakness typically encountered with tumor infiltration. Less often numbness and paresthesias are the presenting symptoms. Unlike infiltration by tumor, radiation plexopathy rarely presents with severe pain, although mild pain ultimately develops in up to 50% of patients.129 Neurologic disability progresses gradually and can stabilize after some years, leaving

Lumbosacral Plexus Lesions

1383

FIGURE 56–5 Pelvic neurofibroma on axial MRI of the pelvis. This large neurofibroma was neurologically occult and was discovered at cesarean section when it had obstructed labor. (Courtesy of Niall Moore, FRCR.)

permanent, severe disability.129 Myelography is normal. The spinal fluid protein level may be elevated.129 Pelvic CT is normal in radiation plexopathy, although radiation osteitis of the sacrum is noted occasionally.129 Electromyography characteristically reveals low-frequency

myokymic discharges in mild to moderately weakened muscles in 50% of patients.2,4,129 Motor nerve conduction is normal or slightly reduced, and the sural sensory nerve action potentials may be reduced. No treatment is known to arrest radiation plexopathy.

FIGURE 56–6 Lymphomatous infiltration of the intrapelvic sciatic nerve. Coronal (A) and axial (B) MRI showing a B-cell nonHodgkin’s lymphoma in the region of the greater sciatic notch, and its relationship to the intrapelvic sciatic nerve and piriformis muscle. (Courtesy of Niall Moore, FRCR.)

1384

Diseases of the Peripheral Nervous System

include malignant tumors of the lumbosacral plexus.50 Patients present with a progressive neurologic deficit or an enlarging palpable mass. The former presentation is more likely for lumbosacral plexus tumors contained within the pelvis. Malignant nerve sheath tumors are generally incurable and fatal within 32 years of diagnosis.

Infections

FIGURE 56–7 MRI showing pelvic teratoma causing intermittent sciatica-like symptoms.

Distinction between radiation plexopathy and tumor plexopathy depends on the differentiating features summarized in Table 56–1. A rare postirradiation lower motor neuron syndrome affecting the legs may occur months or years after radiotherapy to the lower thoracic and lumbar spine.16,36,51,67,111 The threshold radiotherapy dosage has generally exceeded 40 Gy, with onset of the neurologic syndrome 3 to 5 years later.16 Painless wasting and fasciculation of leg muscles is accompanied by electromyographic evidence of chronic denervation. The sensory nerve action potentials and somatosensory evoked potentials are normal, excluding a lesion of the lumbosacral plexus.36,51,111 The neuronal lesion is situated in the cauda equina, and mild sensory or micturition symptoms occur in some patients.16 No treatment is available. Occasionally nerve sheath tumors have been noted 4 to 41 years after radiotherapy to peripheral nerves. They

Table 56–1. Differentiating Features Between Radiation and Tumor Lumbosacral Plexopathy

Usual first symptom Distribution Predominantly weak muscles Pelvic CT or MRI Myokymic discharges

Radiation Plexopathy

Tumor Plexopathy

Weakness Bilateral Distal

Pain Unilateral Proximal

Normal Yes (in 50%)

Tumor mass No

Lumbosacral and coccygeal plexus neuropathy is most commonly associated with anogenital herpes simplex infections or herpes zoster infections affecting lumbosacral dermatomes. Less commonly, lumbosacral neuropathy can be caused by localized bacterial infections such as a vertebral osteomyelitis, tuberculous abscess tracking into the psoas muscle, pyelonephritis, appendicitis, or iliacus muscle abscess complicating laparotomy.3,134 CT is invaluable in diagnosing localized abscess (Fig. 56–8).97,134 Recurrent lumbosacral and brachial plexopathy has been associated with Schistosoma japonicum infection but is not thought to be due to direct infection of the nerve plexus.97 Herpes zoster may cause a painful vesicular eruption within lumbar or sacral dermatomes. Segmental muscle paralysis follows approximately 7% of such eruptions, generally developing after a delay of 3 to 20 days.130 Cutaneous zoster eruptions limited to a single dermatome may be accompanied by weakness in multiple overlapping myotomes. Sacral zoster eruptions can lead to urinary retention.83,100,106,108 There is mild cerebrospinal fluid lymphocytosis and protein elevation.100,130 The precise site of the lesion affecting the motor neuron is unknown but may lie within the spinal cord rather than the plexus or roots. The condition should be distinguished from herpes zoster transverse myelitis, in which extensor plantar responses and more widespread weakness occur. Acyclovir therapy is indicated, although many patients made complete or partial recovery of motor function before the introduction of this drug. Acute anogenital infection with herpes simplex may involve the lower motor neurons of the sacrococcygeal plexus. Neuralgia, numbness, and paresthesias of the perineum, buttocks, and posterior thighs are followed by urinary retention, constipation, or erectile failure.23,61,63,73,102 Reduced anal tone, impaired anal and bulbocavernosus reflexes, and sensory loss in sacral dermatomes are usually encountered. Occasionally more diffuse abnormalities of tendon reflexes, power, and sensation in the legs occur and are attributable to myelitis69,87 or generalized demyelinating neuropathy.15,23 Urodynamic studies show a paralytic neuropathic bladder.109 The primary anogenital herpetic lesion is usually self-evident on inspection of the perineum but is occasionally detected only on formal examination of the uterine cervix.102 The usual isolate is herpes simplex virus type 2 rather than type 1. Mild meningism may be present, and spinal fluid lymphocytosis is common. Electromyography may show fibrillations in the muscles of

Lumbosacral Plexus Lesions

1385

FIGURE 56–8 Pelvic abscess. Sagittal (A) and axial (B) MRI showing staphylococcal pre-sacral abscesses tracking posteriorly to the subcutaneous tissues in a patient with bilateral leg weakness and loss of sphincter control.

the perineum and slow nerve conduction in the legs.32 Full recovery generally occurs; neurologic symptoms generally last approximately 10 days and rarely more than 21 days.102 Acyclovir therapy is now recommended, although as yet there is no proof that it influences the neurologic illness. Only 1% of women with primary genital herpes develop sacrococcygeal plexopathy.32 The highest incidence of this neurologic complication occurs in anally receptive homosexual males with herpetic proctitis.32,61,63,102 Sacrococcygeal plexopathy only rarely follows penile herpes in heterosexual males.32,84 The syndrome generally occurs in primary genital herpes, often after a change of sexual partner, and is rare in recurrent herpes.10,32 Herpes simplex virus type 2 can be cultured from the sacral dorsal root ganglia of randomly selected cadavers.10

Hereditary Liability to Pressure Palsies Lumbosacral plexus lesions are much less common than brachial plexus lesions in patients with hereditary neuralgic amyotrophy (autosomal dominantly inherited familial

brachial neuritis) or hereditary liability to pressure palsies (hereditary pressure-sensitive neuropathy).17,27,82,128 Such lumbosacral plexus lesions may present with pain in the buttock or leg, followed by weakness of the leg, reflex depression, and sensory loss. Distinction from a prolapsed intervertebral disc can be difficult without imaging. Indeed, although leg symptoms occurred in 30% of symptomatic patients in a large family study,128 some of these patients undoubtedly had a prolapsed intervertebral disc rather than primary plexus lesions. Women with hereditary neuralgic amyotrophy or predisposition to pressure palsies can develop plexus lesions puerperally.18,27,128 Apart from the family history, features suggestive of these inherited focal neuropathies include a history of vocal cord paralysis or brachial neuritis.82 Some families with hereditary neuralgic amyotrophy display a distinctive facial appearance with close-set eyes and dwarfism.82 Motor nerve conduction may be reduced in clinically unaffected nerves,18 although this may be less common in hereditary neuralgic amyotrophy than in hereditary neuropathy with liability to pressure palsies.13,41 In hereditary neuralgic amyotrophy,

1386

Diseases of the Peripheral Nervous System

teased sural nerve fibers may show sausage-shaped swellings representing aberrant myelination, the so-called tomacula65,95 that typify hereditary neuropathy with liability to pressure palsies.13 Hereditary neuralgic amyotrophy is itself genetically heterogeneous and distinct from hereditary liability to pressure palsies.27,65,90,124,137 In contrast to hereditary neuralgic amyotrophy, patients with predisposition to pressure palsies do not generally experience severe pain, usually recall the episode of nerve compression, and develop distal rather than proximal muscle weakness. Spontaneous recovery is the rule in both conditions.

Inflammatory Plexopathies In the differential diagnosis of lumbosacral plexus lesions, a number of inflammatory conditions need to be considered: diabetic proximal neuropathy (see Chapter 88), nondiabetic lumbosacral radiculoplexus neuropathy (see Chapter 106), and vasculitic neuropathy (see Chapters 109 and 110). Only those clinical features relevant to differential diagnosis are mentioned briefly here. The clinical spectrum of diabetic proximal neuropathy varies from the extreme of acute asymmetrical painful proximal leg weakness developing over a few days to the less familiar extreme of symmetrical painless proximal weakness developing over many weeks or months. It is most common in type 2 non–insulin-dependent diabetic men, generally in their sixth or seventh decade. Previously unrecognized diabetes may present with proximal neuropathy. Proximal neuropathy is seldom accompanied by diabetic retinopathy or nephropathy.11 The clinical picture of diabetic proximal neuropathy varies. Anterior thigh pain is generally prominent in the early stages.11,94 Proximal leg weakness, principally involving the quadriceps femoris, develops with a time course ranging from a few days to many months. Sensory loss is usually mild in comparison with the disabling weakness, and knee jerks are depressed or absent in most patients. Bilateral spread occurs in over half, but the clinical picture usually remains asymmetrical. Extensor plantar responses are sometimes found.25,53,94 Weight loss exceeds 4.5 kg in more than half. More profound weight loss can occur, leading to suspicion of tumor infiltration of the lumbosacral plexus, a differential diagnosis resolvable by imaging. Neurologic improvement occurs in most patients and is generally attributed to therapy of hypoglycemia with insulin or oral hypoglycemic agents.11,25,52,68 Biopsy of cutaneous branches of the femoral nerve has shown microscopic vasculitis.39,88,93,112 This raises the question of treating severe or enduring cases of diabetic proximal neuropathy with corticosteroids or other immunomodulatory drugs. Idiopathic lumbosacral plexus neuropathy is the counterpart in the leg of acute brachial neuritis or neuralgic amyotrophy affecting the arm, although it is undoubtedly less common. The true incidence is probably underesti-

mated because of poor awareness of the condition and confusion with prolapsed intervertebral disc.46 Its age of onset, clinical course, clinical features, pathology, and outcome are remarkably similar to those of diabetic proximal neuropathy.40 The abrupt onset of unilateral severe pain occurs in the anterior thigh when lumbar plexus involvement predominates or in the buttock and posterior thigh when sacral plexus involvement predominates. The pain tends to resolve as weakness develops.46,114 Involvement of the upper plexus is most common, usually associated with an absent knee jerk, tenderness of the femoral nerve, and a positive femoral stretch sign. Lower plexus involvement is underdiagnosed because the pain resembles that of sciatica associated with prolapsed intervertebral disc and the straight leg raising test is positive.136 The ankle jerk is usually absent, and the sciatic nerve may be tender in the region of the sciatic notch. Nerve biopsy shows alterations typically of ischemic injury and microvasculitis resembling those described in diabetic proximal neuropathy. A chronic progressive form of idiopathic lumbosacral plexus neuropathy has also been described, and can involve signal hyperintensity and slight enlargement of the lumbosacral plexus nerves on MRI.76 This condition may respond to high-dose intravenous immune globulin therapy.76,132,139 The neuropathy associated with systemic vasculitis usually affects more peripheral portions of limb nerves than the brachial or lumbosacral plexuses. Vasculitic lumbosacral plexus lesions usually develop acutely, are painful, and often worsen in a stepwise fashion. The condition may initially suggest neoplastic infiltration of the plexus because of associated weight loss and raised erythrocyte sedimentation rate. In such patients, normal plexus imaging is important for differential diagnosis. The diagnosis becomes easier when patients show other manifestations of systemic vasculitis: purpura, arthritis, glomerulonephritis, respiratory tract granuloma, or eosinophilia. It is those patients whose vasculitis is restricted to the nerves of the lumbosacral plexus who pose the greatest diagnostic difficulty because biopsy of muscle or sural nerve does not reliably establish the diagnosis.

REFERENCES 1. Adelman, J. U., Goldberg, G. S., and Puckett, J. D.: Postpartum bilateral femoral neuropathy. Obstet. Gynecol. 42:845, 1973. 2. Aho, K., and Sainio, K.: Late irradiation-induced lesions of the lumbosacral plexus. Neurology 33:953, 1983. 3. Aichroth, P., and Rowe-Jones, D. C.: Iliacus compartment compression syndrome. Br. J. Surg. 58:833, 1971. 4. Albers, J. W., Allen, A. A. 2nd, Bastron, J. A., and Daube, J. R.: Limb myokymia. Muscle Nerve 4:494, 1981. 5. Alsever, J. D.: Lumbosacral plexopathy after gynecologic surgery: case report and review of the literature. Am. J. Obstet. Gynecol. 174:1769, discussion 1777, 1996.

Lumbosacral Plexus Lesions 6. Aminoff, M. J., Goodin, D. S., Parry, G. J., et al.: Electrophysiologic evaluation of lumbosacral radiculopathies: electromyography, late responses, and somatosensory evoked potentials. Neurology 35:1514, 1985. 7. Ampil, F. L.: Palliative irradiation of carcinomatous lumbosacral plexus neuropathy. Int. J. Radiat. Oncol. Biol. Phys. 12:1681, 1986. 8. Ashenhurst, E. M., Quartey, G. R., and Starreveld, A.: Lumbo-sacral radiculopathy induced by radiation. Can. J. Neurol. Sci. 4:259, 1977. 9. Baker, G. S., Parsons, W. R., and Welch, J. S.: Endometriosis within the sheath of the sciatic nerve: report of two patients with progressive paralysis. J. Neurosurg. 25:652, 1966. 10. Baringer, J. R.: Recovery of herpes simplex virus from human sacral ganglions. N. Engl. J. Med. 291:828, 1974. 11. Bastron, J. A., and Thomas, J. E.: Diabetic polyradiculopathy: clinical and electromyographic findings in 105 patients. Mayo Clin. Proc. 56:725, 1981. 12. Beckenbaugh, R. D., and Ilstrup, D. M.: Total hip arthroplasty. J. Bone Joint Surg. Am. 60:306, 1978. 13. Behse, F., Buchthal, F., Carlsen, F., and Knappeis, G. G.: Hereditary neuropathy with liability to pressure palsies: electrophysiological and histopathological aspects. Brain 95:777, 1972. 14. Benzel, E. C., Morris, D. M., and Fowler, M. R.: Nerve sheath tumors of the sciatic nerve and sacral plexus. J. Surg. Oncol. 39:8, 1988. 15. Black, D., Stewart, J., and Melmed, C.: Sacral nerve dysfunction plus generalized polyneuropathy in herpes simplex genitalis. Ann. Neurol. 14:692, 1983. 16. Bowen, J., Gregory, R., Squier, M., and Donaghy, M.: The post-irradiation lower motor neuron syndrome neuronopathy or radiculopathy? Brain 119:1429, 1996. 17. Bradley, W. G., Chad, D., Verghese, J. P., et al.: Painful lumbosacral plexopathy with elevated erythrocyte sedimentation rate: a treatable inflammatory syndrome. Ann. Neurol. 15:457, 1984. 18. Bradley, W. G., Madrid, R., Thrush, D. C., and Campbell, M. J.: Recurrent brachial plexus neuropathy. Brain 98:381, 1975. 19. Brin, B. J., and Busuttil, R. W.: Isolated hypogastric artery aneurysms. Arch. Surg. 117:1329, 1982. 20. Brown, J. T., and McDougall, A.: Traumatic maternal birth palsy. J. Obstet. Gynaecol. Br. Empire 64:431, 1957. 21. Brun, A., Caviness, V., Rudnick, P., and Tyler, H. R.: Hemorrhage into peripheral nerves in association with leukemia. J. Neuropathol. Exp. Neurol. 23:719, 1964. 22. Burkhart, F. L., and Daly, J. W.: Sciatic and peroneal nerve injury: a complication of vaginal operations. Obstet. Gynecol. 28:99, 1966. 23. Caplan, L. R., Kleeman, F. J., and Berg, S.: Urinary retention probably secondary to herpes genitalis. N. Engl. J. Med. 297:920, 1977. 24. Cary, N.: Complete paralysis of the anterior sural nerve following childbirth. J. Bone Joint Surg. Am. 7:451, 1925. 25. Casey, E. B., and Harrison, M. J.: Diabetic amyotrophy: a follow-up study. Br. Med. J. 1:656, 1972. 26. Chalmers, J. A.: Traumatic neuritis of the puerperium. J. Obstet. Gynaecol. Br. Empire 56:205, 1949. 27. Chance, P. F., Lensch, M. W., Lipe, H., et al.: Hereditary neuralgic amyotrophy and hereditary neuropathy with liability

28.

29.

30.

31. 32.

33.

34.

35. 36.

37.

38. 39.

40.

41.

42.

43.

44.

45. 46. 47.

1387

to pressure palsies: two distinct genetic disorders. Neurology 44:2253, 1994. Chapman, E. M., Shaw, R. S., and Kubik, C. S.: Sciatic pain from arteriosclerotic aneurysm of pelvic arteries. N. Engl. J. Med. 271:1410, 1964. Chiu, W. S.: The syndrome of retroperitoneal hemorrhage and lumbar plexus neuropathy during anticoagulant therapy. South. Med. J. 69:595, 1976. Chokroverty, S.: Magnetic stimulation of the human peripheral nerves. Electromyogr. Clin. Neurophysiol. 29:409, 1989. Cole, J. T.: Maternal obstetric paralysis. Am. J. Obstet. Gynecol. 52:372, 1946. Corey, L., Adams, H. G., Brown, Z. A., and Holmes, K. K.: Genital herpes simplex virus infections: clinical manifestations, course, and complications. Ann. Intern. Med. 98:958, 1983. Criado, F. J., Wilson, E. P., Velazquez, O. C., et al.: Safety of coil embolization of the internal iliac artery in endovascular grafting of abdominal aortic aneurysms. J. Vasc. Surg. 32:684, 2000. D’Amour, M. L., Lebrun, L. H., Rabbat, A., et al.: Peripheral neurological complications of aortoiliac vascular disease. Can. J. Neurol. Sci. 14:127, 1987. Day, M. H.: The blood supply of the lumbar and sacral plexuses in the human foetus. J. Anat. 98:105, 1964. De Carolis, P., Montagna, P., Cipulli, M., et al.: Isolated lower motoneuron involvement following radiotherapy. J. Neurol. Neurosurg. Psychiatry 49:718, 1986. Dietemann, J. L., Sick, H., Wolfram-Gabel, R., et al.: Anatomy and computed tomography of the normal lumbosacral plexus. Neuroradiology 29:58, 1987. Donaldson, J. O.: Neurology of Pregnancy. London, W. B. Saunders, 1988. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Non-diabetic lumbosacral radiculoplexus neuropathy: natural history, outcome and comparison with the diabetic variety. Brain 124:1197, 2001. Earl, C. J., Fullerton, P. M., Wakefield, G. S., and Schutta, H. S.: Hereditary neuropathy, with liability to pressure palsies. Q. J. Med. 33:481, 1964. Ebraheim, N. A., Lu, J., Biyani, A., et al.: The relationship of lumbosacral plexus to the sacrum and the sacroiliac joint. Am. J. Orthop. 26:105, 1997. Edelman, F. L., Hopwood, H. G. Jr., and Lonergan, W. M.: Intrapelvic neurilemoma of the lumbosacral plexus in a pregnant woman. Am. J. Obstet. Gynecol. 102:904, 1968. Egel, R. T., Cueva, J. P., and Adair, R. L.: Posttraumatic childhood lumbosacral plexus neuropathy. Pediatr. Neurol. 12:62, 1995. Emery, S., and Ochoa, J.: Lumbar plexus neuropathy resulting from retroperitoneal hemorrhage. Muscle Nerve 1:330, 1978. Evans, B. A., Stevens, J. C., and Dyck, P. J.: Lumbosacral plexus neuropathy. Neurology 31:1327, 1981. Fishman, E. K., Scatarige, J. C., Saksouk, F. A., et al.: Computed tomography of endometriosis. J. Comput. Assist. Tomogr. 7:257, 1983.

1388

Diseases of the Peripheral Nervous System

48. Flanagan, W. F., Webster, G. D., Brown, M. W., and Massey, E. W.: Lumbosacral plexus stretch injury following the use of the modified lithotomy position. J. Urol. 134:567, 1985. 49. Flowers, R. S.: Meralgia paresthetica: a clue to retroperitoneal malignant tumor. Am. J. Surg. 116:89, 1968. 50. Foley, K. M., Woodruff, J. M., Ellis, F. T., and Posner, J. B.: Radiation-induced malignant and atypical peripheral nerve sheath tumors. Ann. Neurol. 7:311, 1980. 51. Gallego, J., Delgado, G., Tunon, T., and Villanueva, J. A.: Delayed postirradiation lower motor neuron syndrome. Ann. Neurol. 19:308, 1986. 52. Garland, H.: Neurological complications of diabetes mellitus: clinical aspects. Proc. R. Soc. Med. 53:137, 1960. 53. Garland, H., and Taverner, D.: Diabetic myelopathy. Br. Med. J. 1:1405, 1953. 54. Gaylord, T. G., and Pearson, J. W.: Neuropathy following paracervical block in the obstetric patient. Obstet. Gynecol. 60:521, 1982. 55. Gebarski, K. S., Gebarski, S. S., Glazer, G. M., et al.: The lumbosacral plexus: anatomic-radiologic-pathologic correlation using CT. Radiographics 6:401, 1986. 56. Georgy, F. M.: Femoral neuropathy following abdominal hysterectomy. Am. J. Obstet. Gynecol. 123:819, 1975. 57. Gertzbein, S. D., and Evans, D. C.: Femoral nerve neuropathy complicating iliopsoas haemorrhage in patients without haemophilia. J. Bone Joint Surg. Br. 54:149, 1972. 58. Gibbs, M. A., and Beydoun, S. R.: Obstetrical lumbosacral plexus injury. Muscle Nerve 16:801, 1993. 59. Gilchrist, J. M., and Moore, M.: Lumbosacral plexopathy in cancer patients. Neurology 35:1392, 1985. 60. Gilden, D. H., and Eisner, J.: Lumbar plexopathy caused by disseminated intravascular coagulation. JAMA 237:2846, 1977. 61. Goldmeier, D., Bateman, J. R., and Rodin, P.: Urinary retention and intestinal obstruction associated with ano-rectal Herpes simplex virus infection. Br. Med. J. 2:425, 1975. 62. Gonik, B., Stringer, C. A., Cotton, D. B., and Held, B.: Intrapartum maternal lumbosacral plexopathy. Obstet. Gynecol. 63:45S, 1984. 63. Goodell, S. E., Quinn, T. C., Mkrtichian, E., et al.: Herpes simplex virus proctitis in homosexual men. Clinical, sigmoidoscopic, and histopathological features. N. Engl. J. Med. 308:868, 1983. 64. Goodfellow, J., Fearn, C. B., and Matthews, J. M.: Iliacus haematoma: a common complication of haemophilia. J. Bone Joint Surg. Br. 49:748, 1967. 65. Gouider, R., LeGuern, E., Emile, J., et al.: Hereditary neuralgic amyotrophy and hereditary neuropathy with liability to pressure palsies: two distinct clinical, electrophysiologic, and genetic entities. Neurology 44:2250, 1994. 66. Green, J. P.: Proximal avulsion of the iliacus with paralysis of the femoral nerve: report of a case. J. Bone Joint Surg. Br. 54:154, 1972. 67. Greenfield, M. M., and Stark, F. M.: Post-irradiation neuropathy. Am. J. Roentgenol. Radium Ther. 60:617, 1948. 68. Hamilton, C. R. Jr., Dobson, H. L., and Marshall, J.: Diabetic amyotrophy: clinical and electronmicroscopic studies in six patients. Am. J. Med. Sci. 256:81, 1968. 69. Handler, C. E., and Perkin, G. D.: Radiculomyelopathy due to genital herpes. Lancet 2:987, 1982.

70. Head, H. B., Welch, J. S., Mussey, E., and Espinosa, R. E.: Cyclical sciatica: report of a case with introduction of a new surgical sign. JAMA 180:521, 1962. 71. Heffernan, L. P., Fraser, R. C., and Purdy, R. A.: L-5 radiculopathy secondary to a uterine leiomyoma in a primigravid patient. Am. J. Obstet. Gynecol. 138:460, 1980. 72. Hefty, T. R., Nelson, K. A., Hatch, T. R., and Barry, J. M.: Acute lumbosacral plexopathy in diabetic women after renal transplantation. J. Urol. 143:107, 1990. 73. Hemrika, D. J., Schutte, M. F., and Bleker, O. P.: Elsberg syndrome: a neurologic basis for acute urinary retention in patients with genital herpes. Obstet. Gynecol. 68:37S, 1986. 74. Henson, R. A., and Urich, H.: Cancer and the Nervous System. Oxford, UK, Blackwell, 1982. 75. Hibbard, J., and Schreiber, J. R.: Footdrop due to sciatic nerve endometriosis. Am. J. Obstet. Gynecol. 149:800, 1984. 76. Hollinger, P., and Sturzenegger, M.: Chronic progressive primary lumbosacral plexus neuritis: MRI findings and response to immunoglobulin therapy. J. Neurol. 247:143, 2000. 77. Hope, E. E., Bodensteiner, J. B., and Thong, N.: Neonatal lumbar plexus injury. Arch. Neurol. 42:94, 1985. 78. Huittinen, V. M. Lumbosacral nerve injury in fracture of the pelvis: a postmortem radiographic and patho-anatomical study. Acta Chir. Scand. Suppl. 429:3, 1972. 79. Huittinen, V. M., and Slatis, P.: Nerve injury in double vertical pelvic fractures. Acta Chir. Scand. 138:571, 1972. 80. Hunter, V. P., Burke, T. W., and Crooks, L. A.: Retroperitoneal nerve sheath tumors: an unusual cause of pelvic mass. Obstet. Gynecol. 71:1050, 1988. 81. Ismael, S. S., Amarenco, G., Bayle, B., and Kerdraon, J.: Postpartum lumbosacral plexopathy limited to autonomic and perineal manifestations: clinical and electrophysiological study of 19 patients. J. Neurol. Neurosurg. Psychiatry 68:771, 2000. 82. Jacob, J. C., Andermann, F., and Robb, J. P.: Heredofamilial neuritis with brachial predilection. Neurology 11:1025, 1960. 83. Jacobs, S. C., Herbert, L. A., Piering, W. F., and Lawson, R. K.: Acute motor paralytic bladder in renal transplant patients with anogenital herpes infection. J. Urol. 123:426, 1980. 84. Jacome, D. E., and Yanez, G. F.: Herpes genitalis and neurogenic bladder and bowel. J. Urol. 124:752, 1980. 85. Jaeckle, K. A., Young, D. F., and Foley, K. M.: The natural history of lumbosacral plexopathy in cancer. Neurology 35:8, 1985. 86. King, A. B.: Neurologic conditions occurring as complications of pregnancy. Arch. Neurol. Psychiatry 63:611, 1950. 87. Klastersky, J., Cappel, R., Snoeck, J. M., et al.: Ascending myelitis in association with herpes-simplex virus. N. Engl. J. Med. 287:182, 1972. 88. Krendel, D. A., Costigan, D. A., and Hopkins, L. C.: Successful treatment of neuropathies in patients with diabetes mellitus. Arch. Neurol. 52:1053, 1995. 89. Kubacz, G. J.: Femoral and sciatic compression neuropathy. Br. J. Surg. 58:580, 1971. 90. Kuhlenbaumer, G., Meuleman, J., De Jonghe, P., et al.: Hereditary neuralgic amyotrophy (HNA) is genetically heterogeneous. J. Neurol. 248:861, 2001.

Lumbosacral Plexus Lesions 91. Laman, D. M., Endtz, L. J., van Well-Krouwel, H. J., and Gerretsen, G.: A rare cause of lumbosacral plexus neuropathy. Clin. Neurol. Neurosurg. 87:47, 1985. 92. Leno, C., Combarros, O., and Berciano, J.: Lumbosacral plexopathy due to dermoid cyst of the greater omentum. Postgrad. Med. J. 63:45, 1987. 93. Llewelyn, J. G., Thomas, P. K., and King, R. H.: Epineurial microvasculitis in proximal diabetic neuropathy. J. Neurol. 245:159, 1998. 94. Locke, S., Lawrence, D. G., and Legg, M. A.: Diabetic amyotrophy. Am. J. Med. 34:775, 1963. 95. Madrid, R., and Bradley, W. G.: The pathology of neuropathies with focal thickening of the myelin sheath (tomaculous neuropathy). J. Neurol. Sci. 25:415, 1975. 96. Maillard, J. C., Zouaoui, A., Bencherif, B., et al.: Imaging in the exploration of lumbosacral plexus avulsion: two cases. J. Neuroradiol. 19:38, 1992. 97. Marra, T. A.: Recurrent lumbosacral and brachial plexopathy associated with schistosomiasis. Arch. Neurol. 40:586, 1983. 98. Marra, T. R.: The clinical and electrodiagnostic features of idiopathic lumbo-sacral and brachial plexus neuropathy: a review of 20 cases. Electromyogr. Clin. Neurophysiol. 27:305, 1987. 99. Massey, E. W.: CT evaluation of lumbosacral plexus disorders. Postgrad. Med. 69:116, 1981. 100. Meyer, R., Brown, H. P., and Harrison, J. H.: Herpes zoster involving the urinary bladder. N. Engl. J. Med. 260:1062, 1959. 101. Montag, T. W., and Mead, P. B.: Postpartum femoral neuropathy. J. Reprod. Med. 26:563, 1981. 102. Oates, J. K., and Greenhouse, P. R.: Retention of urine in anogenital herpetic infection. Lancet 1:691, 1978. 103. Ovrum, E., and Dahl, H. K.: Vessel and nerve injuries complicating total hip arthroplasty. Arch. Orthop. Trauma Surg. 95:267, 1979. 104. Pettigrew, L. C., Glass, J. P., Maor, M., and Zornoza, J.: Diagnosis and treatment of lumbosacral plexopathies in patients with cancer. Arch. Neurol. 41:1282, 1984. 105. Pontin, A. R., Donaldson, R. A., and Jacobson, J. E.: Femoral neuropathy after renal transplantation. S. Afr. Med. J. 53:376, 1978. 106. Ray, B., and Wise, G. J.: Urinary retention associated with herpes zoster. J. Urol. 104:422, 1970. 107. Razzuk, M. A., Linton, R. R., and Darling, R. C.: Femoral neuropathy secondary to ruptured abdominal aortic aneurysms with false aneurysms. JAMA 201:817, 1967. 108. Richmond, W.: The genito-urinary manifestations of herpes zoster: three case reports and a review of the literature. Br. J. Urol. 46:193, 1974. 109. Riehle, R. A. Jr., and Williams, J. J.: Transient neuropathic bladder following herpes simplex genitalis. J. Urol. 122:263, 1979. 110. Rosenblum, J., Schwarz, G. A., and Bendler, E.: Femoral neuropathy—a neurological complication of hysterectomy. JAMA 195:409, 1966. 111. Sadowsky, C. H., Sachs, E. Jr., and Ochoa, J.: Postradiation motor neuron syndrome. Arch. Neurol. 33:786, 1976. 112. Said, G., Goulon-Goeau, C., Lacroix, C., and Moulonguet, A.: Nerve biopsy findings in different pat-

113.

114. 115.

116.

117. 118.

119.

120.

121.

122. 123.

124.

125. 126.

127.

128. 129.

130. 131.

132.

1389

terns of proximal diabetic neuropathy. Ann. Neurol. 35:559, 1994. Salazar-Grueso, E., and Roos, R.: Sciatic endometriosis: a treatable sensorimotor mononeuropathy. Neurology 36:1360, 1986. Sander, J. E., and Sharp, F. R.: Lumbosacral plexus neuritis. Neurology 31:470, 1981. Silverstein, A.: Neurological complications of anticoagulation therapy: a neurologist’s review. Arch. Intern. Med. 139:217, 1979. Simeone, J. F., Robinson, F., Rothman, S. L., and Jaffe, C. C.: Computerized tomographic demonstration of a retroperitoneal hematoma causing femoral neuropathy: report of two cases. J. Neurosurg. 47:946, 1977. Sinclair, R. H., and Pratt, J. H.: Femoral neuropathy after pelvic operation. Am. J. Obstet. Gynecol. 112:404, 1972. Solheim, L. F., and Hagen, R.: Femoral and sciatic neuropathies after total hip arthroplasty. Acta Orthop. Scand. 51:531, 1980. Spiegel, P. G., and Meltzer, J. L.: Femoral-nerve neuropathy secondary to anticoagulation: report of a case. J. Bone Joint Surg. Am. 56:425, 1974. Stevens, J. C.: Lumbosacral plexus lesions. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 1425, 1984. Stewart, J. D., Schmidt, B., and Wee, R.: Computed tomography in the evaluation of plexopathies and proximal neuropathies. Can. J. Neurol. Sci. 10:244, 1983. Stoehr, M.: Traumatic and postoperative lesions of the lumbosacral plexus. Arch. Neurol. 35:757, 1978. Stoehr, M., Dichgans, J., and Dorstelmann, D.: Ischaemic neuropathy of the lumbosacral plexus following intragluteal injection. J. Neurol. Neurosurg. Psychiatry 43:489, 1980. Stogbauer, F., Young, P., Kuhlenbaumer, G., et al.: Hereditary recurrent focal neuropathies: clinical and molecular features. Neurology 54:546, 2000. Suber, D. A., and Massey, E. W.: Pelvic mass presenting as meralgia paresthetica. Obstet. Gynecol. 53:257, 1979. Sunderland, S.: The relative susceptibility to injury of the medial and lateral popliteal divisions of the sciatic nerve. Br. J. Surg. 41:300, 1953. Tallroth, A.: Haemophilia with spontaneous haemorrhage in the iliopsoas muscle followed by injury to the femoral nerve. Acta Chir. Scand. 82:1, 1939. Taylor, R. A.: Heredofamilial mononeuritis multiplex with brachial predilection. Brain 83:113, 1960. Thomas, J. E., Cascino, T. L., and Earle, J. D.: Differential diagnosis between radiation and tumor plexopathy of the pelvis. Neurology 35:1, 1985. Thomas, J. E., and Howard, F. M. Jr.: Segmental zoster paresis—a disease profile. Neurology 22:459, 1972. Tonzola, R. F., Ackil, A. A., Shahani, B. T., and Young, R. R.: Usefulness of electrophysiological studies in the diagnosis of lumbosacral root disease. Ann. Neurol. 9:305, 1981. Triggs, W. J., Young, M. S., Eskin, T., and Valenstein, E.: Treatment of idiopathic lumbosacral plexopathy with intravenous immunoglobulin. Muscle Nerve 20:244, 1997.

1390

Diseases of the Peripheral Nervous System

133. Turgut, F., Turgut, M., and Mentes, E.: Lumbosacral plexus compression by fetus: an unusual cause of radiculopathy during teenage pregnancy. Eur. J. Obstet. Gynecol. Reprod. Biol. 73:203, 1997. 134. Tysvaer, A. T.: Computerized tomography and surgical treatment of femoral compression neuropathy: report of two cases. J. Neurosurg. 57:137, 1982. 135. Usubiaga, J. E., Kolodny, J., and Usubiaga, L. E.: Neurological complications of prevertebral surgery under regional anesthesia. Surgery 68:304, 1970. 136. van Alfen, N., and van Engelen, B. G.: Lumbosacral plexus neuropathy: a case report and review of the literature. Clin. Neurol. Neurosurg. 99:138, 1997. 137. van Alfen, N., van Engelen, B. G., Reinders, J. W., et al.: The natural history of hereditary neuralgic amyotrophy in the Dutch population: two distinct types? Brain 123:718, 2000. 138. Vaziri, N. D., Barton, C. H., Ravikumar, G. R., et al.: Femoral neuropathy: a complication of renal transplantation. Nephron 28:30, 1981. 139. Verma, A., and Bradley, W. G.: High-dose intravenous immunoglobulin therapy in chronic progressive lumbosacral plexopathy. Neurology 44:248, 1994. 140. Vock, P., Mattle, H., Studer, M., and Mumenthaler, M.: Lumbosacral plexus lesions: correlation of clinical signs and computed tomography. J. Neurol. Neurosurg. Psychiatry 51:72, 1988. 141. Vosburgh, L. F., and Finn, W. F.: Femoral nerve impairment subsequent to hysterectomy. Am. J. Obstet. Gynecol. 82:931, 1961. 142. Waldman, I., and Braun, A. I.: Femoral neuropathy secondary to iliac artery aneurysm. South. Med. J. 70:1243, 1977. 143. Weber, E. R., Daube, J. R., and Coventry, M. B.: Peripheral neuropathies associated with total hip arthroplasty. J. Bone Joint Surg. Am. 58:66, 1976.

144. Weintraub, M. I.: Acute femoral neuropathy secondary to consumption coagulopathy. N. Y. State J. Med. 71:2670, 1971. 145. Weis, E. B. Jr.: Subtle neurological injuries in pelvic fractures. J. Trauma 24:983, 1984. 146. Wetzel, N., Arieff, A., and Tuncbay, E.: Retroperitoneal, lumbar and pelvic malignancies simulating the “disc syndrome.” Arch. Surg. 86:1069, 1963. 147. Whittaker, W. G.: Injuries of the sacral plexus in obstetrics. Can. Med. Assoc. J. 79:622, 1958. 148. Wilbourn, A. J.: Electrodiagnosis of plexopathies. Neurol. Clin. 3:511, 1985. 149. Willbanks, O. L., and Fuller, C. H.: Femoral neuropathy due to retroperitoneal bleeding: an unusual complication of heparin therapy. Arch. Intern. Med. 132:83, 1973. 150. Winkelman, N. W.: Femoral nerve complications after pelvic surgery. Am. J. Obstet. Gynecol. 75:1063, 1958. 151. Wohlgemuth, W. A., Rottach, K. G., and Stoehr, M.: Intermittent claudication due to ischaemia of the lumbosacral plexus. J. Neurol. Neurosurg. Psychiatry 67:793, 1999. 152. Wohlgemuth, W. A., and Stoehr, M.: Percutaneous arterial interventional treatment of exercise-induced neurogenic intermittent claudication due to ischaemia of the lumbosacral plexus. J. Neurol. 249:988, 2002. 153. Young, M. R., and Norris, J. W.: Femoral neuropathy during anticoagulant therapy. Neurology 26:1173, 1976. 154. Zarranz, J. J., and Salisachs, P.: Femoral neuropathy due to compression by retroperitoneal haemorrhage: a modern evaluation. J. Neurol. Sci. 43:479, 1979. 155. Zarranz, J. J., Simon, R., and Salisachs, P.: Acute anticoagulant-induced compressive lumbar plexus neuropathy: a clinico-pathological study. Eur. Neurol. 20:469, 1981.

D. Neuropathy Caused by Compression, Entrapment, or Physical Injury

57 Mechanisms of Acute and Chronic Compression Neuropathy TED M. BURNS

Overview Differential Mechanical Forces Role in Acute Nerve Compression

Role in Chronic Nerve Compression Ischemia and Increased Endoneurial Fluid Pressure

OVERVIEW Compression neuropathies usually present as focal, chronic mononeuropathies, such as of the median, ulnar, or peroneal nerve, at sites where nerves pass through tissue tunnels. Other sites are spinal nerve in vertebral foramina. Nerve compression may also develop acutely (e.g., upper or lower limb tourniquet paralysis, compression from surgical retraction, and “Saturday night” radial nerve palsy).75 Space-occupying lesions and excess accumulation of extracellular matrix or fluid may also cause nerve compression syndromes. Metabolic disease (e.g., diabetes mellitus) or mutant genes (hereditary pressure palsy from deletion of 17p11.2) may make nerves more subject to compression injury. Nerve compression syndromes are caused by mechanical sheer forces, local microvascular disturbances, or combinations of the two. This chapter reviews the evidence for sheer forces versus ischemia inducing compression neuropathy in the experimental model and in human disease. Mechanisms in human disease are also discussed.

Additional Mechanisms Impaired Axonal Transport Impaired Nerve Trunk Gliding Capacity

DIFFERENTIAL MECHANICAL FORCES Role in Acute Nerve Compression When nerve tissues are subjected to mechanical compression, some of the compressed tissues may be displaced to sites of lower pressure. This is especially the case for acute compression neuropathies such as the prototype “Saturday night palsy,”75 surgical retractor nerve palsy, and tourniquet compression neuropathies. Mechanical compression (sheer forces or stress modules) may also be involved in chronic or repetitive entrapment. A role for direct mechanical pressure injury had been argued for by many investigators since the 1920s,2,5,23,44,72 but it was not until the 1970s that a consensus emerged, based on compelling pathologic studies, that mechanical forces play an essential role.20,21,24,51,55–57,63 In 1954, for example, Moldaver relied on suggestive clinical features when he reasoned that the focality of nerve injury (i.e., conduction block confined to the site of compression) contrasted against the normal electrical response to nerves distal to the line of pressure— nerves that would have also experienced ischemia from the 1391

1392

Diseases of the Peripheral Nervous System

tourniquet—suggested a direct mechanical mechanism.44 This line of reasoning, however, was fallacious because it did not differentiate between ischemia of the entire limb and of a focal section of nerve. Nevertheless, this argument was influential at the time and was erroneously supported by the decreased incidence of tourniquet paralysis associated with the cessation of use of thick rubber tubes in favor of Esmarch bandages or inflated cuffs.44 The work of investigators at Queen Square in the 1970s provided solid evidence for a mechanical mechanism of injury in acute compression neuropathy. In the early 1970s, Ochoa and colleagues at Queen Square reported that, after acute (1 to 3 hours) compression (1000 mm Hg) by sphygmomanometer cuff (5 cm wide) around the knees of baboons causing focal conduction block and limb paralysis, nerve damage was concentrated at the edges of the cuff rather than at limb sites of lowest blood perfusion.24,55,56 Prominent injury at the edges of the cuff suggested that the pressure gradient, rather than the absolute pressure, was the critical factor. The pathologic changes observed at the edges supported this contention (Figs. 57–1 through 57–3). The earliest histopathologic changes, seen within the first few hours, were invagination of one paranodal segment into the adjacent paranode. The invagination was directed away from the cuff and toward the uncompressed tissue. Paranodal myelin, tethered to axon, was grossly distorted, resulting in invagination on one side of the node and passive stretching on the other side. Longitudinal movement of the axon relative to the Schwann cell accompanied the paranodal myelin alterations. In extreme cases, myelin lamellae were ruptured at the site of invagination. These experimental findings, reminiscent of intussusception of the bowel, offered compelling visual evidence that the

FIGURE 57–1 Electron micrograph showing nodal axoplasm and axonal membrane moving away from the point of junction of the two Schwann cells. The pressure gradient is from higher to lower (from right to left). (From Ochoa, J., Danta, G., Fowler, T. J., and Gilliatt, R. W.: Nature of the nerve lesion caused by a pneumatic tourniquet. Nature 233:265, 1971, with permission.)

FIGURE 57–2 Electron micrograph of nodal region shown in Figure 57–1. The pressure gradient is from higher to lower (from right to left). A, Terminal myelin loops of invaginating paranode. B, Terminal myelin loops of adjacent paranode. C, Myelin fold of invaginating paranode cut tangentially. D, Schwann cell cytoplasm. E, Microvilli indicating site of Schwann cell junction. Large arrows show length of ensheathed paranode. (From Ochoa, J., Danta, G., Fowler, T. J., and Gilliatt, R. W.: Nature of the nerve lesion caused by a pneumatic tourniquet. Nature 233:265, 1971, with permission.)

pressure gradient between compressed and uncompressed nerve provided driving forces that caused expression of axoplasm “like toothpaste from a tube.”58 Paranodal demyelination, similar to that observed by Denny-Brown and Brenner,15,16 was observed at the end of the first week at sites of previously invaginated myelin. The demyelination was usually restricted to within 200 ␮ of the node, although demyelination of whole internodal segments was rarely observed. Subsequent experiments on baboons, attempting to better simulate acute human pressure palsies by using weighted nylon cords rather than cuffs, demonstrated similar early structural changes of displacement of nodes of Ranvier with secondary changes to paranodal myelin. If the displacement was severe (e.g., 200 to 350 ␮), rupture of stretched paranodal myelin was observed. The degree of conduction block correlated with the magnitude and duration of compression.63 The severity of the lesion also correlated with the size of the affected zones seen at the edges, providing more circumstantial evidence that differential mechanical forces are responsible for injury in acute compression neuropathy. In severe lesions associated with conduction block, the affected zones were large and unaffected central zones were small, whereas in milder lesions the converse was seen. The nodal displacement was consistently polarized, with movement away from sites of high pressure. Demyelination followed by the second week, whereas remyelination was observed in nerves examined 2 to 7 months out (Fig. 57–4). Dyck and colleagues confirmed that the stress forces at the apex and edges of a Silastic cuff were primarily responsible for the structural pathologic alterations typical of cuff compression.18 By perfusing and in situ fixation of peroneal rat nerve within a minute of cuff application, they

1393

Mechanisms of Acute and Chronic Compression Neuropathy

FIGURE 57–3 Diagram showing the direction of displacement of the nodes of Ranvier in relation to the cuff in acute compression neuropathy. (From Ochoa, J., Fowler, T. J., and Gilliatt, R. W.: Anatomical changes in peripheral nerves compressed by a pneumatic tourniquet. J. Anat. 113:433, 1972, with permission.)

Proximal

were able to infer sequential cellular events of compression injury and show structural changes that could not be due to ischemic injury. Using different degrees of compression and stopping histologic events at different times, it was possible to show that the sequential events resulting from focal compression are (1) expression of endoneurial fluid (the fluid between fibers) from the region underlying the apex of the cuff along the nerve to and beyond the edges of the cuff, (2) expression of fluid within axons

FIGURE 57–4 Teased myelinated nerve fibers from anterior tibial nerve of baboon 12 weeks after compression showing numerous remyelinated internodes at affected zones. The regions of nerve from which the teased fibers were taken are shown by diagram on the left. Damaged zones are shaded. Normal nodes are arrowed. (Scale on left in millimeters; bar on right: 100 ␮). (From Rudge, P., Ochoa, J., and Gilliatt, R. W.: Acute peripheral nerve compression in the baboon. Anatomical and physiological findings. J. Neurol. Sci. 23:403, 1974, with permission.)

Cuff

Distal

toward and beyond edges of the cuff, (3) expression of cytoskeletal elements to and beyond edges of the cuff, and (4) distortion and disruption of myelin and Schwann cell elements. The authors proposed that the first event was the expression of endoneurial fluid, followed by expression of axoplasm fluid, and then of axoplasm contents. The observation that during compression the myelinated fiber density was increased and the fascicular area decreased under the cuff, and vice versa at the edges of the cuff, suggested expression of endoneurial fluid in the early phase of compression. A second slower phase was attributed to further expression of endoneurial fluid, expression of axonal fluid and contents, paranodal disruption and extrusion of cytoplasm of Schwann cells, and displacement of other tissue elements from beneath the cuff. Additional damage (e.g., to the cytoskeletal network) may occur at more extreme pressures or with protracted compression. The authors confirmed that nodes of Ranvier were frequently obscured or lengthened because of displaced paranodal myelin, particularly at the edges of the cuff. Lengthened nodes were increasingly common under the cuff at higher pressures and prolonged durations. The authors suggested that the lengthened nodes represented the effect of three processes: (1) compression and expression of axonal fluid and displacement of axonal contents, resulting in passive stretching of the axolemma, particularly at nodes of Ranvier where the axon is normally narrowed and therefore offers resistance to axoplasmic movement; (2) detachment of myelin at nodes; and (3) cleavage and longitudinal displacement of layers of partial- or full-thickness myelin. On electron microscopic examination, the axons under the cuff were usually compressed, with cytoskeletal components and mitochondria closely packed together (Fig. 57–5). The degree of cytoskeletal and organelle compaction was more severe at higher pressures. The myelin sheath was altered in the form of irregular-shaped profiles at low pressures of short durations. At high pressure, myelin was severely split and distorted. On teased fiber preparations, demyelination and axonal degeneration were observed. Axonal degeneration was observed distal to the cuff site, particularly with conditions of high pressure for prolonged

1394

Diseases of the Peripheral Nervous System

state of decorticate rigidity with flexion of the wrists that demonstrated the characteristic pathologic changes of nodal displacement away from the Schwann cell junctions, with the direction of displacement apparently away from the compression site.51 Additional support came from a study of baboons in which experimental compression plus ischemia did not produce recognizably more damage to nerves than did compression alone.82 In this study, there was no significant side-to-side difference in degree of conduction block, recovery of function, or histopathologic damage with bilateral acute peroneal compression neuropathies and unilateral whole-limb ischemia.82

Role in Chronic Nerve Compression

FIGURE 57–5 Electron micrographs of transverse section of peroneal nerve of rat showing compaction of cytoskeleton and organelles of axons that occurred under the cuff when nerve was compressed at 300 mm Hg for 2 minutes (top) and 300 mm Hg for 2 hours (bottom). Inset: Top, Low-power view (⫻4800) of a transverse section of a myelinated nerve fiber showing the rectangular area at high magnification (⫻38,300). Bottom, Transverse section (⫻7700) of the nerve fiber; myelin is split and axoplasm is not seen. (From Dyck, P. J., Lais, A. C., Giannini, C., and Engelstad, J. K.: Structural alterations of nerve during cuff compression. Proc. Natl. Acad. Sci. U. S. A. 87:9828, 1990, with permission.)

duration, whereas demyelination and remyelination were seen under the cuff. The nature and time course of these histologic changes are distinctively different from those in ischemic injury.52 These animal studies provide compelling evidence that differential mechanical forces contribute significantly to the injury of acute compression neuropathy. This argument was further supported by an autopsy study of the median nerves of a 30-year-old man who died in an acute

Chronic nerve entrapment of the median nerve at the wrist may have been first described by Ramsey-Hunt in the early 1900s.61 Focal loss of myelin at the site of chronic median nerve compression was described by Marie et Foix in 1913.42 Denny-Brown and Brenner also described focal demyelination following chronic (up to 8 weeks) spring clip pressure to cat sciatic nerve.15 An autopsy study of the median nerve of a woman with chronic carpal tunnel syndrome revealed the typical changes of nerve, including fewer large myelinated fibers under the retinaculum, enlargement of the nerve proximal to the border of the flexor retinaculum, and thickened perineurium with increased endoneurial connective tissue at and slightly above the retinaculum.73 The guinea pig model of chronic pressure neuropathy confirmed the pathologic features of focal demyelination at sites of repeated trauma or chronic entrapment, but did not address the underlying mechanism.3,20,21 In one study, young guinea pigs (2 to 16 weeks), kept in cages with wire-mesh flooring for up to a year, developed hindfoot plantar neuropathies because of repeated minor pressure.21 Electrophysiologic studies suggested focal demyelination, and paranodal demyelination and remyelination were seen on teased fiber analysis, occurring approximately 1 cm below the ankle at the level of the tarsometatarsal joint. Wallerian degeneration occurred in severe cases. In another study, older guinea pigs (older than 18 months)—which are known to frequently develop spontaneous median or ulnar neuropathies at the wrist as the nerves pass underneath a transverse arch20,43—were studied looking for pathologic correlates to the chronic, focal neuropathies. Selective damage to large myelinated fibers occurred first at the wrist in these older animals, with relative preservation of the small myelinated and unmyelinated fibers.43 In more severe cases, small myelinated and unmyelinated fiber degeneration occurred at the wrist, while more proximally the large myelinated fibers were selectively affected. Perineurial thickening and endoneurial swelling were seen immediately above the wrist in severe cases. Wallerian degeneration occurred at and distal to the wrist in severe

Mechanisms of Acute and Chronic Compression Neuropathy

cases.20 These studies confirmed that focal demyelination occurs early at sites of entrapment, but were unable to provide insight into the mechanism of nerve injury. Pathology studies of young guinea pigs with early, very mild median and ulnar pressure neuropathies advanced the understanding of the mechanism of nerve injury in chronic compression.3,57 The earliest changes seen were asymmetrical distortion of the internodes of large myelinated fibers, with tapering of the internodes at one end and swelling of the internodes at the other end (Figs. 57–6 and 57–7). Electron microscopy revealed that the myelin lamellae end in a series of cytoplasmic loops at the tapered end, away from their original attachment at the node (see Fig. 57–7). At the bulbous paranode, the myelin was seen to be buckled with in-turning of some of the myelin lamellae. The redundant folds of inner lamellae indented the axon and were associated with the accumulation of modified axoplasm, suggesting possible interference of axon flow. The polarity reversed on either side of the lesion. This paranodal change was followed by demyelination and subsequent remyelination. Displacement of nodes of Ranvier was not observed, which is in contrast to

B FIGURE 57–6 Teased fiber study of young guinea pigs with early, very mild median and ulnar pressure neuropathies. Fiber samples taken from above (A) and below (B) the wrist show early changes of asymmetrical distortion of the internodes of large myelinated fibers, with tapering of the internodes at one end and swelling of the internodes at the other end. P ⫽ proximal end; D ⫽ distal end. Displacement of nodes of Ranvier is not observed in chronic compression, which is in contrast to the changes seen in acute compression. (From Ochoa, J., and Marotte, L.: Nature of the nerve lesion underlying chronic entrapment. J. Neurol. Sci. 19:491, 1973, with permission.)

1395

B

C FIGURE 57–7 Electron microscopy of young guinea pig nerves taken from the wrist. The paranode on the left is tapered. A and B, Myelin lamellae ending in a series of cytoplasmic loops at the tapered end, away from their original attachment at the node. C, At the bulbous paranode, the myelin appears buckled with in-turning of some of the myelin lamellae. The redundant folds of inner lamellae indent the axon and are associated with the accumulation of modified axoplasm, suggesting possible interference of axon flow. The polarity reversed on either side of the lesion. Displacement of nodes of Ranvier is not observed in chronic nerve compression. (From Ochoa, J., and Marotte, L.: Nature of the nerve lesion underlying chronic entrapment. J. Neurol. Sci. 19:491, 1973, with permission.)

the pathologic changes of acute compression. Thus the pathologic features differed from those demonstrated in acute compression but were, nonetheless, very suggestive for a direct mechanical cause of injury. The finding of movement of groups of inner myelin lamellae relative to the axon, in a direction away from the site of entrapment, suggested to the investigators that the detachment and slippage of myelin lamellae may be caused by repeated pressure waves.57 Similar pathologic findings were subsequently described in humans.51 In an autopsy series of presumably asymptomatic ulnar and median nerves, the remarkable finding of teased nerve fibers taken from under suspected entrapment sites was paranodal myelin swelling, consisting of redundant folds of myelin at one end of an internode, with or without thinning or retraction of the myelin at the opposite end. The polarity of swellings was reversed at the opposite side of entrapment, suggesting that the myelin changes were caused by pressure gradients of entrapment. The reversal of polarity of the ulnar nerve appeared to lie

1396

Diseases of the Peripheral Nervous System

under the aponeurosis of the flexor carpi ulnaris muscle. For the median nerve, the reversal of polarity was more variable but under or within 1 cm of the retinaculum. In keeping with animal models of chronic compression, nodes of Ranvier were not displaced. The investigators suggested that the cause of this “myelin slippage” may have been recurrent subclinical trauma or pressure sufficient to cause detachment of the inner myelin lamellae. Renaut bodies and enlargement of the nerve resulting in part from increased intrafascicular area also occurred at presumed sites of compression. Aguayo and colleagues studied chronic nerve compression by applying a Silastic tube around the medial popliteal nerve of 3- to 5-week-old rabbits.1 The tube was applied to the nerve of one limb and the other limb served as a control. As the rabbits matured, the tube gradually compressed the nerve. Nerve conduction and histologic studies revealed focal demyelination and remyelination at the compression site. The authors also pointed out that the more severe nerve pathology was distributed peripherally rather than centrally within nerve fascicles. This subperineurial distribution of nerve pathology was later confirmed71 and has subsequently been emphasized by investigators positing an important role for transperineurial ischemia in compression neuropathy.

ISCHEMIA AND INCREASED ENDONEURIAL FLUID PRESSURE Since the 1920s, many investigators have concluded that ischemia was primarily responsible for the nerve injury of compression.6,7,15,16,25,33,74,83 This conclusion was largely based on clinical observations. For example, in 1931, Lewis and colleagues postulated an ischemic mechanism in compression neuropathy based on experiments with pneumatic cuffs in humans.33 In one experiment, limb paralysis and sensory disturbance persisted after the removal of a distal cuff, if, at the time of removal of the distal cuff, a second cuff was placed proximal to the first. The authors reasoned that, if local mechanical nerve compression, rather than ischemia, were at work, then symptom resolution followed only later by development of new symptoms should have occurred. The authors also pointed out that cuff pressures below the systolic blood pressure did not cause neurologic symptoms or signs. Following the work of Lewis and colleagues, the ischemic hypothesis was erroneously bolstered by the experiments of Grundfest, who reported in 1936 that frog nerves continue to conduct at extremely high pressures (up to 14,000 psi in oxygen pressure chambers), suggesting that it must be something other than the pressure that causes injury in everyday nerve compression.25 This faulty inference was later rejected by work demonstrating that the pressure gradient rather than the absolute pressure is what

is most damaging to nerve.18,20,21,24,51,55,57,63 In 1944, Denny-Brown and Brenner reported the experimental study of acute and chronic pressure applied by tourniquet and spring clip to cat sciatic nerve.15,16 They observed that reversible nerve conduction abnormalities were limited to the portion of compressed nerve and that distal nerve fibers conducted normally (i.e., focal conduction block). They also demonstrated focal demyelination with spared axons at the site of conduction block. Other findings included nerve edema, lymphocyte and macrophage infiltration, and, in more severe cases, axonal damage. The conduction block lasted less than 3 weeks and was followed by rapid and complete recovery, which correlated with remyelination of previously demyelinated segments. Denny-Brown and Brenner correctly reasoned that the focal demyelination represented the cellular complication of conduction block.15 However, they speculated, in part based on the Grundfest experiment, that the pathologic changes were a consequence of ischemia. The conclusion that ischemia primarily caused compression neuropathies was based on physiologic and clinical observations and was subsequently tempered by the pathology studies from the 1970s discussed previously18,20,21,51,55,57,63 that demonstrated structural evidence implicating differential mechanical forces. Nonetheless, focal ischemia may yet play a significant role in some compression neuropathies, particularly in combination with the direct effects of pressure. Ischemia and endoneurial edema, discussed subsequently, may significantly contribute to the pathology of nerves that sustain chronic compression of modest pressure magnitudes, as is known to occur in carpal tunnel syndrome and ulnar neuropathy at the elbow.4,36,39,40,62,65,66,72,77 Also, transient nerve block—for instance, when a limb transiently “goes to sleep” because of modest external pressures—may be primarily caused by focal ischemia, because no recognizable structural nerve pathology has convincingly been demonstrated in such cases.58 The ischemic hypothesis has most recently focused on the transperineurial vascular system, which includes an intrafascicular circulation, composed mostly of capillaries running longitudinally within the endoneurium, and an extrafascicular network in the epineurium, composed predominantly of venules and arterioles. The extrinsic vessels penetrate the relatively rigid perineurium to anastomose with the intrinsic circulation,37,40 and it is this transperineurial vessel network that may be particularly susceptible to focal compression, especially because these vessels traverse the perineurium at oblique angles (Fig. 57–8).31,38–40,49,60,62,64,65 Moreover, the transperineurial vessels, especially the venules, are vulnerable to constriction caused by endoneurial edema and elevated (intrinsic) endoneurial fluid pressure, both of which have been shown to be consequences of external nerve compression.4,18,32,40,62,65,69,70,72,79,80 Constriction of these vessels will cause venous congestion, endoneurial capillary leakage,

Mechanisms of Acute and Chronic Compression Neuropathy

Venule 1

Perineurium 2

3

Endoneurium

FIGURE 57–8 Diagram showing a venule traversing the perineurium at an oblique angle. Transperineurial vessels oriented along the circumferential axis of the perineurium may be more vulnerable to compression than those oriented along the longitudinal axis. (From Myers, R. R., Murakami, H., and Powell, H. C.: Reduced nerve blood flow in edematous neuropathies. Microvasc. Res. 32:145, 1986, with permission.)

and elevated endoneurial fluid pressures,31,38,40,46–50,64–66,72 all of which introduce metabolic disturbances to the microenvironment and damage nerve and nerve function.35,47,54 Thus external compression may induce ischemia and endoneurial edema with concomitantly elevated endoneurial fluid pressures, the latter two of which events not only impair nerve by altering the metabolic microenvironment but also contribute to nerve injury by further constricting transperineurial venules and initiating a vicious cycle of venous congestion, ischemia, and metabolic disturbances, the so-called miniature compartment syndrome.40 The experimental evidence for a vicious cycle of microcirculatory disturbance causing nerve injury has been largely based on the following findings: (1) experimental nerve compression has been shown experimentally to reduce nerve blood flow and cause endoneurial edema15,16,18,32,40,65,66,69,70,72,79; (2) endoneurial edema elevates endoneurial fluid pressure, which may further reduce nerve blood flow40,48,50,65; (3) elevated endoneurial fluid pressure has been demonstrated in humans with chronic compression neuropathies, such as median neuropathies at the wrist22,36; and (4) focal disruption of the transperineurial microcirculation has been shown to result in the same pattern of pathology (e.g., prominent subperineurial injury)

1397

as has been demonstrated by some investigators in experimental compression neuropathy.40,46,60,65,69,70 Focal nerve compression has been shown to cause reduced nerve blood flow in animal models. Compression of rabbit tibial nerves of magnitudes of 10 and 20 mm Hg resulted in a slight reduction in epineurial blood flow, especially through venules.66 A reduction in endoneurial blood flow was first seen at compression pressures of 30 mm Hg.66 Graded compression above 30 mm Hg resulted in relatively more severe impairment of endoneurial and epineurial blood flow, so that at 60 to 80 mm Hg no blood flow was observed in the nerve. Immediately following release after cuff compression for 2 hours, the nerve became edematous but blood flow was restored, albeit appearing sluggish. After compression at 400 mm Hg for 2 hours, markedly impaired blood flow persisted for at least 7 days. Focal nerve compression has also been shown to cause nerve edema.4,15,16,18,32,40,62,65,69,70,79 Endoneurial edema and elevated endoneurial fluid pressure have been demonstrated following local compression of the rat sciatic nerve.40,65 Endoneurial edema has been shown to be most prominent at the edges of nerve compression.65 A threefold increase in endoneurial fluid pressure was found after external compression by inflatable cuff at 30 or 80 mm Hg for 8 hours, and a fourfold increase in endoneurial pressure was observed after compression at 80 mm Hg for 4 hours.40 Endoneurial edema, correlating with endoneurial pressure, was most prominent after compression at 80 mm Hg for 4 hours. Endoneurial edema was found to be particularly prominent in subperineurial and perivascular spaces. Significant endoneurial edema may impair nerve blood flow. For example, both galactose and hexachlorophene neuropathy animal models have been used to demonstrate a pathologic sequence of endoneurial edema, decreased nerve blood flow, and myelinated fiber damage.48,50 In order to normally maintain adequate intrafascicular circulation, the pressure gradient must be such that the pressure is greatest in the epineurial arteries, followed by the endoneurial capillaries, nerve fascicle, epineurial veins, and finally the surrounding tissue (e.g., the carpal tunnel) (Fig. 57–9).72 This pressure gradient protects the endoneurial capillaries from being directly affected in the early stages of mild compression, but as surrounding tissue pressure increases, the venous channels will first be affected, followed by intrafascicular and endoneurial capillary pressures.66,72 If, for example, the intracarpal canal pressure rises above 30 mm Hg, the endoneurial fluid pressure will increase significantly, often three- or fourfold from 1.5 mm Hg to 4.5 mm Hg.22,36,39 This impairs nerve function, causes nerve damage, and impairs the transperineurial microcirculation. Carpal tunnel pressure and consequently endoneurial fluid pressures probably rise significantly at night in the setting of carpal tunnel syndrome because the limb venous return is impeded by limb position and reduced limb movement.72 Endoneurial

1398

Diseases of the Peripheral Nervous System

PF PC

PA PV PT

PA > PC > PF > PV > PT FIGURE 57–9 Diagram of one fascicle of the median nerve in the carpal tunnel with its related vessels. In order to maintain an adequate intrafascicular circulation, the pressure gradient across the system must be highest in epineurial arteries (PA) and lowest within the carpal tunnel (PT). PC ⫽ capillary pressure inside the fascicle; PF ⫽ intrafascicular pressure; PV ⫽ pressure in epineurial veins draining the fascicle.

edema from other causes (e.g., diabetes) may increase nerve susceptibility to compression.11,38,45 Another observation suggesting that transperineurial microcirculation disturbances may play a critical role in compression neuropathy is that experimental nerve compression and focal epineurial microcirculation disruption both cause similar patterns of nerve injury.40,46,60,65,69,70 In one study, the sciatic nerves of adult rats were compressed by a Plexiglas pressure chamber with rubber cuffs.60 At high pressure (80 mm Hg for 2 hours), intrafascicular edema, demyelination, and axon loss were all predominantly in the subperineurial space (Fig. 57–10). Fibrin deposition, cellular proliferation, inflammation, and sometimes hemorrhage were seen in the epineurium and adjacent interstitium within hours of compression. At lower pressures (30 and 10 mm Hg for 2 hours), demyelination was the prominent finding, principally in the subperineurial region. Edema was also present. Axon loss was not as prominent as it was at higher magnitudes of compression. On electron microscopy, necrosis of Schwann cells was observed, encompassing an intact axon and myelin sheath. Demyelination was typical at days 5 to 14. Remyelination was prominent, especially subperineurial remyelination, at 28 days (Fig. 57–11). The authors argued that the distribution of pathology suggests hypoperfusion of focal transperineurial vessels, especially because similar pathologic changes were demonstrated following experimental epineurial devascularization of a 2-cm segment of rat tibial nerve.46 Marked reductions in nerve blood flow and average

FIGURE 57–10 Widespread demyelination in rat sciatic nerve 7 days after nerve compression at 30 mm Hg for 2 hours. Demyelination is more prominent in the subperineurial space. (⫻750.) (From Powell, H. C., and Myers, R. R.: Pathology of experimental nerve compression. Lab. Invest. 55:91, 1986, with permission.)

FIGURE 57–11 Thinly myelinated fibers, suggestive of remyelination, are evident throughout the subperineurial space of rat sciatic nerve endoneurium 28 days after compression at 80 mm Hg for 2 hours (⫻750). (From Powell, H. C., and Myers, R. R.: Pathology of experimental nerve compression. Lab. Invest. 55:91, 1986, with permission.)

Mechanisms of Acute and Chronic Compression Neuropathy

epineurial and subperineurial oxygen tensions were observed following the epineurial stripping. In another study, the sciatic nerves of adult rats were compressed with four ligatures.69 Prominent edema, most marked in the endoneurium and especially the subperineurial space, was observed by day 1. Gaps in perineurial integrity were observed by day 3, allowing for the outgrowth of axonal sprouts into the epineurium that were first evident on day 5. Myelinated fiber damage was most marked subjacent to the perineurium. The proposed pathologic temporal sequence of adult rat painful compression neuropathy included venous stasis, edema with elevation of endoneurial pressure, acute fiber degeneration, Schwann cell and endoneurial endothelial cell proliferation, recruitment of phagocytes, injury to the perineurium, axonal sprouting, and microneuroma formation.69,70 This subperineurial distribution of nerve injury has been emphasized at times to either support or refute the ischemic hypothesis of compression injury. Some investigators postulate that this subperineurial pattern of nerve injury is characteristic both of local disruption of the microcirculation and of experimental nerve compression, and thus the similarity supports a role for ischemia in compression neuropathy.46,54 However, other investigators argue that this pattern of injury is uncharacteristic of ischemic injury, and that, instead, a centrofascicular pattern of nerve fiber degeneration should result if ischemia is a primary mechanism because such a pattern has been clearly demonstrated following microsphere embolization of capillaries or ligation of large arteries.18,27,46,52,58,68 The crux of the argument may rest on whether microsphere embolization and arterial ligation models are relevant to studies of compression neuropathy mechanisms. It may be that, as is clearly the case for central nervous system strokes, slightly different hemodynamic disturbances (e.g., small vessel occlusion, large vessel occlusion, watershed ischemia) result in different peripheral nerve pathologic characteristics, and a subperineurial pattern of injury is seen in some cases while a centrofascicular pattern is seen in others. It has been argued by some investigators that focal compression causes focal compromise of the overlying epineurial vascular network, which results in local subperineurial injury because the subperineurial region is dependent on the local vessels that enter the endoneurium from the epineurium, whereas the central fascicular tissue is less affected because of the robust, longitudinal anastomoses at the center of the fascicle.46 In an attempt to reconcile the different patterns of injury, it has also been suggested that the differences in spatial distribution of endoneurial capillaries between large fascicles and small fascicles (capillaries are more numerous in the subperineurial region in large fascicles and in the central fascicular area in small fascicles) may in part explain the different distribution of pathology.53 However, this explanation is not entirely satisfactory because most of the animal studies

1399

showing subperineurial injury in compression neuropathy were performed on proximal nerve segments, that is, at approximately the same location along the length of nerve as the arterial watershed zones. Another argument for or against an ischemic mechanism has been based on the relative vulnerability of large nerve fibers in compression neuropathy. This initially was used as evidence to support, then later to reject, the ischemic hypothesis.1,6,15–17,19,33,43,44,51,52,56,73 More recently, however, both models of compression neuropathy demonstrate and provide explanations for selective fiber vulnerability, namely the pressure gradient preferentially affecting larger caliber fibers and ischemia damaging Schwann cells. Thus this clinical and pathologic finding may no longer translate into support for one mechanism over the other.58,60

ADDITIONAL MECHANISMS Impaired Axonal Transport Moderately elevated intrafascicular and external pressures may also disturb axonal transport.6,8,10,11,13,14,67,80 Retrograde axonal transport, particularly important for communication with the cell body for initiation of regenerative responses, has been shown to be inhibited by modest pressures in studies of the rabbit vagus nerve.14 Fast and slow anterograde axonal transport may also be reversibly impaired following compression.8–10,13,67 The blocking of axonal transport with compression is a graded effect, related to the magnitude and duration of compression. Morphologic changes in nerve cell bodies have also been demonstrated in experimental nerve compression.12 An already impaired axonal transport from another cause may increase susceptibility to compression. For example, the susceptibility to entrapment in diabetic polyneuropathy may be in part due to the combination of widespread (diabetes) and focal (entrapment) impairment of axonal flow.11,45 The double-crush hypothesis may also be in part explained by this concept.38,76

Impaired Nerve Trunk Gliding Capacity Another factor of chronic compression neuropathies may be the gliding capacity of nerve, which may be particularly relevant at common sites of entrapment such as the wrist and elbow. Gliding of nerves is necessary during movement of limbs and is made possible by conjunctiva-like adventitia that allows excursion of a nerve trunk.36,62 For example, the excursion of the median and ulnar nerves is 0.5 to 1.5 cm with elbow or wrist flexion or extension.38,81 Restriction of glide may occur with extraneural and intraneural fibrosis, especially at sites of entrapment, inducing nerve stretch lesions, edema, inflammation, and further fibrosis.38,77 Stretch may contribute to nerve injury at common sites of entrapment, although it is unlikely to be the major factor in

1400

Diseases of the Peripheral Nervous System

injury and almost always will be overshadowed by the consequences of direct pressure and perhaps ischemia. Nerve trunks are resistant to stretch because of a relatively high degree of elasticity. Nonetheless, studies of experimental stretch confirm that nerve elongation can be damaging to nerve, although the threshold for injury varies markedly from study to study. Reduced nerve blood flow has been reported at 6% to 18% strain,41,59 nerve conduction failure has been demonstrated to begin at 6% strain,78 and histopathologic changes have been reported to develop at ranges of 4% to 50% strain.26,28–30,34

REFERENCES 1. Aguayo, A., Nair, C. P. V., and Midgley, R.: Experimental progressive compression neuropathy in the rabbit: histologic and electrophysiologic studies. Arch. Neurol. 24:358, 1971. 2. Allen, F. M.: Effects of ligation on nerves of the extremities. Ann. Surg. 108:1088, 1938. 3. Anderson, M. H., Fullerton, P. M., Gilliatt, R. W., and Hern, J. E. C.: Changes in the forearm associated with median nerve compression at the wrist in the guinea-pig. J. Neurol. Neurosurg. Psychiatry 33:70, 1970. 4. Beekman, R., Schoemaker, M. C., van der Plas, J. P. L., et al.: Diagnostic value of high-resolution sonography in ulnar neuropathy at the elbow. Neurology 62:767, 2004. 5. Bentley, F. H., and Schlapp, W.: The effects of pressure on the conduction in peripheral nerves. J. Physiol. (Lond.) 102:72, 1943. 6. Causey, G.: The effect of pressure on nerve fibres. J. Anat. 82:262, 1948. 7. Clark, D., Hughes, J., and Gasser, H. S.: Afferent function in the group of nerve fibers of slowest conduction velocity. Am. J. Physiol. 114:69, 1935. 8. Dahlin, L. B., Archer, D. R., and McLean, W. G.: Axonal transport and morphological changes following nerve compression: an experimental study in the rabbit vagus nerve. J. Hand Surg. [Br.] 18:106, 1993. 9. Dahlin, L. B., and Lundborg, G.: The neurone and its response to peripheral nerve compression. J. Hand Surg. [Br.] 15:5, 1990. 10. Dahlin, L. B., and McLean, W. G.: Effects of graded experimental compression on slow and fast axonal transport in rabbit vagus nerve. J. Neurol. Sci. 72:19, 1986. 11. Dahlin, L. B., Meiri, K. F., McLean, W. G., et al.: Effects of nerve compression on fast axonal transport in streptozotocininduced diabetes mellitus: an experimental study in the sciatic nerve of rats. Diabetologia 29:181, 1986. 12. Dahlin, L. B., Nordborg, C., and Lundborg, G.: Morphologic changes in nerve cell bodies induced by experimental graded nerve compression. Exp. Neurol. 95:611, 1987. 13. Dahlin, L. B., Rydevik, B., McLean, W. G., and Sjöstrand, J.: Changes in fast axonal transport during experimental nerve compression at low pressures. Exp. Neurol. 84:29, 1984. 14. Dahlin, L. B., Sjöstrand, J., and McLean, W. G.: Graded inhibition of retrograde axonal transport by compression of rabbit vagus nerve. J. Neurol. Sci. 76:221, 1986.

15. Denny-Brown, D., and Brenner, C.: Lesion in peripheral nerve from compression by spring clip. Arch. Neurol. Psychiatry 52:1, 1944. 16. Denny-Brown, D., and Brenner, C.: Paralysis of nerve induced by direct pressure and by tourniquet. Arch. Neurol. Psychiatry 51:1, 1944. 17. Dyck, P. J., Conn, D. L., and Okazaki, H.: Necrotizing angiopathic neuropathy: three dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972. 18. Dyck, P. J., Lais, A. C., Giannini, C., and Engelstad, J. K.: Structural alterations of nerve during cuff compression. Proc. Natl. Acad. Sci. U. S. A. 87:9828, 1990. 19. Fullerton, P. M.: The effect of ischaemia on nerve conduction in the carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 26:385, 1963. 20. Fullerton, P. M., and Gilliatt, R. W.: Median and ulnar neuropathy in the guinea-pig. J. Neurol. Neurosurg. Psychiatry 30:393, 1967. 21. Fullerton, P. M., and Gilliatt, R. W.: Pressure neuropathy in the hind foot of the guinea-pig. J. Neurol. Neurosurg. Psychiatry 30:18, 1967. 22. Gelberman, R. H., Hergenroeder, P. T., Hargens, A. R., et al.: The carpal tunnel syndrome: a study of carpal canal pressures. J. Bone Joint Surg. Am. 63:380, 1981. 23. Gelfan, S., and Tarlov, I. M.: Physiology of spinal cord: nerve root and peripheral nerve compression. Am. J. Physiol. 185:217, 1956. 24. Gilliatt, R. W., Fowler, T. J., and Ochoa, J.: Tourniquet paralysis in the baboon. Trans. Am. Neurol. Assoc. 97:52, 1972. 25. Grundfest, H.: Effects of hydrostatic pressures upon the excitability, the recovery, and the potential sequence of frog nerve. Cold Spring Harb. Symp. Quant. Biol. 4:179, 1936. 26. Haftek, J.: Stretch injury to peripheral nerve. J. Bone Joint Surg. Br. 52:354, 1970. 27. Hess, K., Eames, R. A., Darveniza, P., and Gilliatt, R. W.: Acute ischaemic neuropathy in the rabbit. J. Neurol. Sci. 44:19, 1979. 28. Highet, W. B., and Holmes, W.: Traction injuries to the lateral popliteal nerve and traction injuries to peripheral nerve after suture. Br. J. Surg. 30:212, 1943. 29. Highet, W. B., and Sanders, F. K.: The effects of stretching nerves after suture. Br. J. Surg. 30:355, 1943. 30. Hoen, T. I., and Brackett, C. E.: Peripheral nerve lengthening. I. Experimental. J. Neurosurg. 13:43, 1956. 31. Kalichman, M. W., and Myers, R. R.: Transperineurial vessel constriction in an edematous neuropathy. J. Neuropathol. Exp. Neurol. 50:408, 1991. 32. Lassmann, G., Lassmann, H., and Stockinger, L.: Morton’s metatarsalgia: light and electron microscopic observations and their relation to entrapment neuropathies. Virchows Arch. A Pathol. Anat. Histopathol. 370:307, 1976. 33. Lewis, T., Pickering, G. W., and Rothschild, P.: Centripetal paralysis arising out of arrested bloodflow to the limb, including notes on a form of tingling. Heart 16:1, 1933. 34. Liu, C. T., Benda, C. E., and Lewey, F. H.: Tensile strength of human nerves: experimental, physical and histologic study. Arch. Neurol. Psychiatry 59:322, 1948. 35. Low, P. A., and Dyck, P. J.: Increased endoneurial fluid pressure in experimental lead neuropathy. Nature 269:427, 1977.

Mechanisms of Acute and Chronic Compression Neuropathy 36. Luchetti, R., Schoenhuber, R., Alfarano, M., et al.: Serial overnight recordings of intracarpal canal pressure in carpal tunnel syndrome patients with and without wrist splinting. J. Hand Surg. [Br.] 19:35, 1994. 37. Lundborg, G.: Structure and function of the intraneural microvessels as related to trauma, edema formation, and nerve function. J. Bone Joint Surg. Am. 57:938, 1975. 38. Lundborg, G., and Dahlin, L. B.: Anatomy, function, and pathophysiology of peripheral nerves and nerve compression. Hand Clin. 12:185, 1996. 39. Lundborg, G., Gelbermann, R., Minera-Convera, N., et al.: Median nerve compression in the carpal tunnel: the functional response to experimentally induced controlled pressure. J. Hand Surg. 7:252, 1982. 40. Lundborg, G., Myers, R., and Powell, H.: Nerve compression injury and increased endoneurial pressure: a “miniature compartment syndrome.” J. Neurol. Neurosurg. Psychiatry 46:1119, 1983. 41. Lundborg, G., and Rydevik, B.: Effects of stretching the tibial nerve of the rabbit: a preliminary study of the intraneural circulation and the barrier function of the perineurium. J. Bone Joint Surg. Br. 55:390, 1973. 42. Marie et Foix, P.: Atrophie isolée de l’éminence thénar d’origine névritique, rôle du ligament annulaire antérieur du carpel dans la pathogénie de la lésion. Rev. Neurol. (Paris) 26:647, 1913. 43. Marotte, L. R.: An electron microscope study of chronic median nerve compression in the guinea-pig. Acta Neuropathol. (Berl.) 27:69, 1974. 44. Moldaver, I.: Tourniquet paralysis syndrome. Arch. Surg. 68:136, 1954. 45. Mulder, D. W., Lambert, L. H., Bastrom, J. A., and Sprague, R. G.: The neuropathies associated with diabetes mellitus: a clinical and electromyographic study of 103 unselected diabetic patients. Neurology 11:275, 1961. 46. Myers, R. R., Heckman, H. M., Galbraith, J. A., and Powell, H. C.: Subperineurial demyelination associated with reduced nerve blood flow and oxygen tension after epineurial vascular stripping. Lab. Invest. 65:41, 1991. 47. Myers, R. R., Heckman, H. M., and Powell, H. C.: Endoneurial fluid is hypertonic: results of microanalysis and its significance in neuropathy. J. Neuropathol. Exp. Neurol. 42:217, 1983. 48. Myers, R. R., Mizisin, A. P., Powell, H. C., and Lampert, P. W.: Reduced nerve blood flow in hexachlorophene neuropathy: relationship to elevated endoneurial fluid pressure. J. Neuropathol. Exp. Neurol. 41:391, 1982. 49. Myers, R. R., Murakami, H., and Powell, H. C.: Reduced nerve blood flow in edematous neuropathies. Microvasc. Res. 32:145, 1986. 50. Myers, R. R., and Powell, H. C.: Galactose neuropathy: impact of chronic endoneurial edema on nerve blood flow. Ann. Neurol. 16:587, 1984. 51. Neary, D., Ochoa, J., and Gilliatt, R. W.: Sub-clinical entrapment neuropathy in man. J. Neurol. Sci. 24:283, 1975. 52. Nukada, H., and Dyck, P. J.: Microsphere embolization of nerve capillaries and fiber degeneration. Am. J. Pathol. 115:275, 1984. 53. Nukada, H., Dyck, P. J., and Karnes, J. L.: Spatial distribution of capillaries in rat nerves: correlation to ischemic damage. Exp. Neurol. 87:369, 1985.

1401

54. Nukada, H., Powell, H. C., and Myers, R. R.: Spatial distribution of nerve injury after occlusion of individual major vessels in rat sciatic nerves. J. Neuropathol. Exp. Neurol. 52:452, 1993. 55. Ochoa, J., Danta, G., Fowler, T. J., and Gilliatt, R. W.: Nature of the nerve lesion caused by a pneumatic tourniquet. Nature 233:265, 1971. 56. Ochoa, J., Fowler, T. J., and Gilliatt, R. W.: Anatomical changes in peripheral nerves compressed by a pneumatic tourniquet. J. Anat. 113:433, 1972. 57. Ochoa, J., and Marotte, L.: Nature of the nerve lesion underlying chronic entrapment. J. Neurol. Sci. 19:491, 1973. 58. Ochoa, J. L.: Nerve fiber pathology in acute and chronic compression. In Omer, G. E. Jr., Spinner, M., and Van Beek, A. L. (eds.): Management of Peripheral Nerve Problems. Philadelphia, W. B. Saunders, p. 475, 1998. 59. Ogata, K., and Naito, M.: Blood flow of peripheral nerve effects of dissection, stretching and compression. J. Hand Surg. [Br.] 11:10, 1986. 60. Powell, H. C., and Myers, R. R.: Pathology of experimental nerve compression. Lab. Invest. 55:91, 1986. 61. Ramsey-Hunt, J.: The thenar and hypothenar types of neural atrophy of the hand. Am. J. Med. Sciences 141:224, 1911. 62. Rempel, D., Dahlin, L., and Lundborg, G.: Pathophysiology of nerve compression syndromes: response of peripheral nerves to loading. J. Bone Joint Surg. Am. 81:1600, 1999. 63. Rudge, P., Ochoa, J., and Gilliatt, R. W.: Acute peripheral nerve compression in the baboon: anatomical and physiological findings. J. Neurol. Sci. 23:403, 1974. 64. Rydevik, B., Brown, M. D., and Lundborg, G.: Pathoanatomy and pathophysiology of nerve compression. Spine 9:7, 1984. 65. Rydevik, B., and Lundborg, G.: Permeability of intraneural microvessels and perineurium following acute, graded experimental nerve compression. Scand. J. Plast. Reconstr. Surg. 11:179, 1977. 66. Rydevik, B., Lundborg, G., and Bagge, U.: Effects of graded compression on intraneural blood flow: an in vivo study on rabbit tibial nerve. J. Hand Surg. [Am.] 6:3, 1981. 67. Rydevik, B., McLean, W. G., Sjostrand, J., and Lundborg, G.: Blockage of axonal transport induced by acute, graded compression of the rabbit vagus nerve. J. Neurol. Neurosurg. Psychiatry 43:690, 1980. 68. Sladky, J. T., Greenberg, J. H., and Brown, M. J.: Regional perfusion in normal and ischemic rat sciatic nerves. Ann. Neurol. 17:191, 1985. 69. Sommer, C., Galbraith, J. A., Heckman, H. M., and Myers, R. R.: Pathology of experimental compression neuropathy producing hyperesthesia. J. Neuropathol. Exp. Neurol. 52:223, 1993. 70. Sommer, C., Lalonde, A., Heckman, H. M., et al.: Quantitative neuropathy of a focal nerve injury causing hyperalgesia. J. Neuropathol. Exp. Neurol. 54:635, 1995. 71. Spinner, M., and Spencer, P. S.: Nerve compression lesions of the upper extremity. Clin. Orthop. 104:46, 1974. 72. Sunderland, S.: The nerve lesion in the carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 39:615, 1976. 73. Thomas, P. K., and Fullerton, P. M.: Nerve fiber size in the carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 26:520, 1963.

1402

Diseases of the Peripheral Nervous System

74. Thompson, I. M., and Kimball, H. S.: Effect of local ischemia upon human nerve fibers in vivo. Proc. Soc. Exp. Biol. Med. 34:601, 1936. 75. Trojaborg, W.: Rate of recovery in motor and sensory fibres of the radial nerve: clinical and electrophysiologic aspects. J. Neurol. Neurosurg. Psychiatry 33:625, 1970. 76. Upton, A. R. M., and McComas, A. J.: The double-crush in nerve entrapment syndrome. Lancet 2:359, 1973. 77. Walker, F. O., Cartwright, M. S., Wiesler, E. R., and Caress, J.: Ultrasound of nerve and muscle. Clin. Neurophysiol. 115:495, 2004. 78. Wall, E. J., Massie, J. B., Kwan, M. K., et al.: Experimental stretch neuropathy: changes in nerve conduction under tension. J. Bone Joint Surg. Br. 74:126, 1992.

79. Weisl, H., and Osborne, G. V.: The pathological changes in rats’ nerves subject to moderate compression. J. Bone Joint Surg. Br. 46:297, 1964. 80. Weiss, P.: Damming of axoplasm in constricted nerve: a sign of perpetual growth in nerve fibres. Anat. Rec. 88:464, 1944. 81. Wilgis, S., and Murphy, R.: The significance of longitudinal excursion in peripheral nerves. Hand Clin. 2:761, 1986. 82. Williams, I. R., Jefferson, D., and Gilliatt, R. W.: Acute nerve compression during limb ischemia—an experimental study. J. Neurol. Sci. 46:199, 1980. 83. Wortis, H., Stein, M. H., and Jolliffe, N.: Fiber dissociation in peripheral neuropathy. Arch. Intern. Med. 69:222, 1942.

58 Mechanisms of Repair after Traumatic Injury SUSAN HALL

Overview of the Response to Injury Classification of Nerve Injuries Perineurial Damage Axonolysis within an Intact Schwann Cell Tube Transection and Separation of Nerve Ends Wallerian Degeneration: An Overview Axonal Responses Sealing Injured Axons Axonal Degeneration Role of Ca2⫹ Delayed Axonal Degeneration Acutely Denervated Schwann Cells Schwann Cell Proliferation

Phenotype of Acutely Denervated Schwann Cells Chronic Axotomy and Chronic Denervation Chronic Axotomy Chronic Denervation Schwann Cell Death and the Axon-Independent Schwann Cell Inflammation and Wallerian Degeneration The Cytokine Network Macrophages Matrix Metalloproteinases and Chondroitin Sulfate Proteoglycans Knockout Mice Neuropathic Pain

Peripheral nerve injuries are most frequently reported in the radial nerve in the arm and the sciatic nerve in the leg.182 As noted by Robinson, they “impede recovery of function and return to work and carry risk of secondary disabilities from falls, fractures, or other secondary injuries.”205 The repair of such injuries can be a protracted and frequently incomplete process, and clinical outcomes are often unsatisfactory.16 The reader should be aware that injury in the peripheral nervous system (PNS) induces changes in the sensory and motor cortices and in numerous subcortical areas, some of them within minutes after injury, that may persist for many months or perhaps permanently.42,133,135,167,168,273 This chapter explores what is currently known about the cellular and molecular responses to mechanical injury in the PNS based on the results of experimental simulations, and the ways in which these phenomena have been, or might be, manipulated therapeutically to facilitate recovery.

Spinal Nerve Ligation Model Chronic Constriction Injury Model and the Role of TNF-␣ Axonal Regrowth Axonal Regrowth across an Interstump Gap: Use of Regeneration Chambers Cellular Outgrowth into a Regeneration Chamber Precision of Reinnervation Nerve Autografts and Their Alternatives Rationale for Using Autografts Alternatives to Autografts: Conduits and Allografts Termino-lateral Neurorrhaphy The Future

OVERVIEW OF THE RESPONSE TO INJURY Peripheral neurons are large, spatially complex cells whose size and connectivity compromise their capacity to repair. Their peripheral processes—the axons that constitute the neuronal element of peripheral nerves—can be damaged physically by a cut or tear (e.g., by a sharp instrument, shrapnel, or bone fragments); burns and electrical injuries; combat-associated trauma, including mine-blast trauma180,241; excessive traction; and external or internal compression (e.g., by tumor enlargement in foramina or under tight fascial bands, by a nodule of amyloid or an endoneurial sarcoid granuloma). Individual axons may be focally compressed by myelin slippage and infolding in some inherited dysmyelinating conditions. The effects of injury may be exacerbated by prolonged disturbance of the microvascular bed of the nerve and/or its surrounding connective tissue sheaths. 1403

1404

Neuropathy Caused by Compression, Entrapment, or Physical Injury

The neuronal response to traumatic injuries involves the entire cell, which means that the functional outcome of nerve repair is determined not only by what happens locally at the injury site (although this is critically important), but also by events that occur elsewhere either within the affected neuron (which may be many centimeters distant from the lesion), or transsynaptically within other functionally related neurons. Injury also disrupts the structural and functional relationships between neurons and their associated glial cells (oligodendrocytes and astrocytes in the central nervous system [CNS], and Schwann cells in the PNS), and this induces changes in glial behavior that impact significantly on the course and outcome of repair. Indeed, Schwann cells and olfactory ensheathing cells have become major foci of experimental attempts to manipulate the repair process in both the PNS and CNS. Caveat: Regrowing axons counted in the proximal extent of a distal nerve stump may not necessarily contribute to functional recovery. Axonal regrowth (which implies extension) should not be confused with axonal regeneration (which implies axonal extension to appropriate targets with the potential for functional recovery).

CLASSIFICATION OF NERVE INJURIES Recovery from injury is influenced by many factors, including age, nutritional and health status, and site and type of injury (including the degree of associated soft tissue injury72,77; Fig. 58–1). Even when these factors are optimized, the extent to which the connective tissue sheaths of a nerve (i.e., epineurium, perineurium, and endoneurium) are disrupted by injury remains a fundamental determinant of functional outcome, and has been the basis for the two most widely applied classifications of nerve injuries.225,249 Seddon proposed three types of nerve lesions, which he termed neurapraxia, axonotmesis, and neurotmesis.225 Sunderland introduced a more complex 5-point system that recognized a subtle gradation of injuries within the group Seddon had characterized by the term axonotmesis.249 For an authoritative critique of these classifications, the reader is referred to Birch et al.16 Caveat: These are theoretical classifications that cannot be tested at the level of individual fibers. By its very nature, trauma usually produces a spectrum of pathophysiologic changes both within and along a nerve.

Ratio of repaired side/normal side for each measured variable

0.9

L_VMAX L_FLOW

0.8

L_AXON 0.7

L_FIBER

0.6

L_MYELIN

0.5 0.4 0.3 0.2 Simple

Cavity

V-graft

Injury (late repair)

FIGURE 58–1 The adverse effect of the presence a complicating arterial injury on repair of transected sheep median nerves. Mean and standard deviation of the mean are presented for the ratios of each of the measured variables in three different injury groups observed 6 months after immediate repair of a 3-cm defect in the median nerve and a 5-cm defect in the brachial artery (using a freeze-thawed muscle graft [FTMG] and a vein graft, respectively). Injury groups are: simple ⫽ nerve repaired with FTMG; cavity ⫽ reproducible cavity injury (see below); and V graft ⫽ nerve repaired using FTMG and artery repaired using vein graft (n ⫽ 5 in each group). The variables measured are: VMAX ⫽ peak conduction velocity on the nerve (m/s); FLOW ⫽ blood flow in the nerve (arbitrary flux units); AXON ⫽ axon diameter (␮m); and MYELIN ⫽ myelin sheath thickness (␮m). A cavity injury was produced in three stages and involved crushing surrounding muscle, to produce areas of bruising and devitalized tissue, then abrading exposed soft tissues, and finally painting the tissues with creosote to provoke widespread inflammation. (From Glasby, M. A., Fullerton, A. C., and Lawson, G. M.: Immediate and delayed nerve repair using freeze-thawed muscle autografts in complex nerve injuries: associated arterial injury. J. Hand Surg. Br. 23:354, 1998, with permission.)

Mechanisms of Repair after Traumatic Injury

A simpler and more practical classification defines nerve injuries as nondegenerative or degenerative.264 Nondegenerative injuries, in which there is no axonal loss, produce a focal conduction block that may be transient or prolonged for weeks or even months. Transient conduction block affects mainly large-caliber axons, and is usually caused by a brief ischemic episode perhaps produced by traction sufficient to interfere with intraneural microcirculation.134 More prolonged block is probably attributable to an underlying primary (segmental) demyelination. In contrast, axonal degeneration occurs when a nerve is partially or completely divided as a consequence of injury and there is associated disruption to the connective tissue sheaths surrounding the nerve.264 Birch et al. noted that “[it] is the difference between preservation and destruction of continuity that underlies the division of degenerative lesions into those with the potential for complete spontaneous recovery and those that will not recover unless action is taken and even then are unlikely fully to recover.”16

Perineurial Damage Damage that is limited exclusively to the perineurium is unusual. Experimentally created partial deficits, known as “perineurial windows,”184,240 induce focal demyelination and degeneration in the underlying nerve. The clinical equivalent may be a needle stick injury, and recent experimental modeling suggests that very small defects in the perineurium, such as might be produced by a needle, result in the entrapment of herniated fibers, which causes a persistent neurologic deficit.246 When focal damage to the perineurium is part of a more invasive injury to the nerve, in which axons are transected and Schwann cell basal laminae are breached, regrowing axons may “escape” into the epineurium before the perineurial barrier is reconstituted, or grow ectopically between the layers of the perineurium, where they may give rise to a painful neuroma.

Axonolysis within an Intact Schwann Cell Tube Axons may “die back” from the periphery as a result of metabolic or toxic dysfunction,37,84 or they may degenerate in response to a focal mechanical injury; in both instances degeneration usually occurs without rupture of the Schwann cell tubes. Focal injury can be modeled experimentally by crushing a nerve bundle with fine forceps, by applying several loosely constricting ligatures (see later), or by intermittent cooling and rewarming113 (Fig. 58–2). Similar degeneration may present clinically as a response to internal (i.e., endoneurial106) or external (i.e., epineurial) compression, and it is probably secondary to altered endoneurial perfusion and a chronically compromised microvascular bed.183 Sustained compression, for example,

1405

at a point of nerve entrapment, may induce endoneurial fibrosis and axonal dropout that is sometimes manifest in the form of Renaut bodies.7,100,109 If the compressive agent is removed, there is no a priori reason why axonal regrowth should not occur within the original Schwann cell tubes. The possibility of pathfinding failures under these circumstances is low, and the likelihood that axons will reinnervate their original targets is high, particularly if the injury is “near-target.” However, when the distance between target and lesion is many centimeters long (as is the case with more proximal injuries), and/or the source of compression persists, then the probability that axons will successfully arrive at their appropriate targets is less certain.

Transection and Separation of Nerve Ends When all of the connective tissue sheaths of a nerve have been severed, it may be possible to reappose the ends of the nerve without tension at the suture site.148,260 Where this is not an option, the interstump defect may be bridged either with a nerve auto- or allograft or with some type of conduit (see later). There is a high probability that regrowing axons will be misrouted as they enter the distal stump, which means that the specificity of reinnervation inevitably decreases,139 particularly if the injured nerve is a large mixed nerve. In addition, extraneural scarring may produce tethering of the repair site in the wound bed, which is likely to interfere with the normal gliding of nerves that occurs during limb or digit movement,51a and this could produce secondary degeneration as a result of ischemia and inflammation, and generate painful dysesthesias.192 If the interstump gap and the distal stump are both long, time becomes an important determinant of outcome, and satisfactory functional recovery is an unrealistic expectation.

WALLERIAN DEGENERATION: AN OVERVIEW Traumatic injury initiates a sequence of changes called wallerian degeneration (after Augustus Waller, who described the morphologic sequelae of cutting the glossopharyngeal or hypoglossal nerves of the frog).187,274 In this sequence, all damaged axons throughout the distal extent of a traumatized nerve, irrespective of their caliber or functional modality, undergo a process of axonolysis and degenerate; where appropriate, their associated myelin sheaths are degraded as well; myelomonocytic cells invade the distal endoneurium and remove axonal and myelin debris; Schwann cells proliferate, and the dedifferentiated daughter cells line up within each basal lamina tube to form bands of Büngner (which are often referred to as Schwann tubes262). Similar changes occur proximal to an injury, but are restricted to a narrow reactive zone a few internodes long immediately adjacent to the lesion site. Degeneration tends

1406

Neuropathy Caused by Compression, Entrapment, or Physical Injury

FIGURE 58–2 Transverse resin sections demonstrate that focal intermittent cooling and rewarming of rat sciatic nerves induce axonal degeneration. The nerves were first cooled through gradients of 1° to 5° C for 3 hours and then their temperature was rapidly restored to 37° C. A, “Empty” endoneurial vessels after 1 hour of rewarming. B, Endoneurial edema seen 12 hours after nerve cooling. C–F, Axonal degeneration, including dark, “empty,” and shrunken axons and intramyelinic edema, is widespread 1 day (C), 2 days (D), 3 days (E), and 5 days (F) after cooling. Closed arrows in D, E, and F indicate thrombosed endoneurial vessels with swollen and degenerating endothelium. Bar: 25 ␮m. (From Jia, J., and Pollock, M.: The pathogenesis of non-freezing cold nerve injury: observations in the rat. Brain 120 [Pt. 4]:631, 1997, with permission.)

to involve whole internodes, so that the proximal limit of this zone usually occurs at heminodes of Ranvier. Occasionally paranodal myelin of adjacent heminodes exhibits signs of remodeling, such as attenuated, thinly remyelinated paranodes and excessive infolding of contiguous normal-thickness internodal myelin, which suggests that the process is not always limited to whole internodes. Proximal to the reactive zone, the nerve fibers appear morphologically normal. Axonal degeneration is not an exclusive consequence of trauma, but is a final common pathway seen in many disorders of the nervous system with widely varying etiologies.84 Purists might argue that, wherever the cause is not mechanical, the degeneration should be termed wallerian-like.

AXONAL RESPONSES Sealing Injured Axons The ends of severed axons become sealed within hours of injury, at a rate that appears to be independent of the type of injury.136 The mechanism of sealing has been studied extensively in invertebrate giant axons, where it requires a minimum axoplasmic Ca2⫹ concentration of 100 ␮M.54

The process involves the progressive fusion of axolemmal vesicles across the open end of an axon until an ionic seal has been reestablished, so that the movement of molecules and ions across the axolemma can return to preinjury levels55 (Fig. 58–3). The sealed ends become swollen with organelles because anterograde axoplasmic transport continues in the proximal stump, and retrograde transport persists for several days in axons in the distal stump. Potent vasoactive peptides such as calcitonin gene–related peptide accumulate in the axonal end-bulbs in the proximal stump. Their presence has been correlated with the local hyperemia that occurs within 48 hours of injury, in and around the injury site, via pathways that are both NO dependent and NO independent.286 Other mechanisms, such as local mast cell degranulation and ongoing angiogenesis, probably generate the hyperemia that persists for several weeks after injury within microvessels in the epi- and endoneurium of the distal nerve stump.

Axonal Degeneration The traditional view of axonal degeneration, as a passive process in which traumatized axons “wither away,” is probably

1407

Mechanisms of Repair after Traumatic Injury

A

B

C

D

E

FIGURE 58–3 Vesicle-mediated formation of an ionic seal at the cut end of a severed crayfish medial giant axon. A, After transection, vesicles induced by elevated Ca2⫹ form by endocytosis of the axolemma and move to the partially collapsed (box) cut end. B, Vesicles accumulate and become densely packed at the disrupted portion of axolemma. C, Vesicle membranes interact with each other and the axolemma, and some form junctional complexes (black dots joining adjacent complexes) that decrease both the number and conductance of paths (contorted dark lines with arrowheads) for ion entry into or exit from the axolemmal disruption. D, An ionic seal has formed (i.e., ionic conductance through the disrupted portion of axolemma has been reduced to the level in the intact axolemma). Reestablishment of axolemmal continuity begins or has begun either directly by vesicle fusion with the boundary of the disrupted axolemma (between open arrows) and/or indirectly by fusion with the intact portion of axolemma (closed arrows). E, At a substantial time after an ionic seal is formed (⬎⬎ 60 minutes), axolemmal continuity is restored and vesicles are no longer evident. (From Eddlemann, C. S., Bittner, G. D., and Fishman, H. M.: Barrier permeability at cut axonal ends progressively decreases until an ionic seal is formed. Biophys. J. 79:1883, 2000, with permission.)

no longer tenable. Indeed, recent evidence suggests that it may be more appropriate to think of degeneration as an active, programmed response to disconnection from target and cell body. “Sealed” axons within the distal segments of transected invertebrate or fish nerve can remain intact for months, even though they are isolated from their cell bodies and sustained by ongoing axoplasmic transport and axon-glia communication.197 In marked contrast, axonal degeneration is normally instituted very rapidly in mammalian peripheral nerves. Loss of the axoplasmic cytoskeleton, which is probably the earliest morphologic sign of the dramatic changes to come, begins 1 to 2 days after injury in small laboratory rodents and within 7 days in humans; the differing latencies are “roughly directly proportional to the size of the organism.”78 Morphologically, the periaxonal space swells, axoplasmic organelles pool and disintegrate at nodal constrictions and within myelin ovoids as axoplasmic transport fails, and axoplasm is physically segmented by involuting myelin. Physiologically, disconnected axons become unable to conduct a compound action potential (CAP) on application of an external electrical stimulus,138,158 and 80% of the CAP is lost between 24 and 36 hours in both A and C fibers in transected mouse nerve.268 Long before these changes become apparent, events that have been likened to “lighting

FIGURE 58–4 Procaspase-3 and its close relatives are apparently not activated in axons of C57Bl/6 mouse nerve undergoing wallerian degeneration. Confocal immunofluorescence micrographs of longitudinal sections of uncut nerve (0 hour) or sciatic nerve explants cultured for 14, 19, and 24 hours were stained with the N-52 antineurofilament antibody and with CM1, Ab206, or Ab127 antibody. CM1 and Ab206 were both produced against peptide sequences corresponding to the carboxyl terminus of the large subunit of activated human and rodent caspase-3; Ab127 antibodies have been characterized extensively as a biochemical and immunohistochemical marker for caspase-mediated proteolysis associated with apoptosis. Note the lack of activated caspase-3 in axons undergoing wallerian degeneration, even though neurofilaments have been degraded. Arrows indicate apoptotic glial cells containing caspase-3 or its close relatives. Bar: 50 ␮m. (From Finn, J. T., Weil, M., Archer, F., et al.: Evidence that wallerian degeneration and localized axon degeneration induced by local neurotrophin deprivation do not involve caspases. J. Neurosci. 20:1333, 2000, with permission.)

a fuse”268 occur at the submicroscopic level within disconnected axons, and they trigger a local caspase-independent program of autodestruction69,198 (Fig. 58–4).

Role of Ca2ⴙ An increase in the level of intracellular Ca2⫹ via ion-specific channels is both necessary and sufficient to induce axonal degeneration, and to activate axonal calcium-dependent proteolytic enzymes, such as phospholipases and calpains, which effect subsequent cytoskeletal breakdown.75,220–222,232,268 Since wallerian degeneration of peripheral axons apparently requires calcium levels in the range of 200 to 300 ␮M,75,78,79 it was initially assumed that axoplasmic cytoskeletal degradation was mediated by activated m-calpain.35 However, m-calpain immunoreactivity falls within hours of transection, at a time when axons are still structurally intact78 (Fig. 58–5),

1408

Neuropathy Caused by Compression, Entrapment, or Physical Injury

and so it is perhaps more appropriate to regard m-calpain activation as an early “fuse-lighting” event that does not lead directly to degeneration.

Delayed Axonal Degeneration

FIGURE 58–5 m-Calpain is activated very early during wallerian degeneration. A, Tissue cross sections of C57Bl/6J mice sciatic nerves harvested at multiple time points after transection and stained with the neuronal marker protein gene product 9.5 (PGP 9.5). Axons remain intact until between 24 and 48 hours. B, Representative immunoblots comparing the transected with the nontransected nerves (left and right of each pair, respectively) using multiple antibody markers. Note that loss of neurofilament (NF) and ␤-tubulin (␤-tub) immunoreactivity occurs between 24 and 48 hours (mirroring the findings with PGP 9.5), but m-calpain immunoreactivity (N1, N2, C1; see below) is reduced within 6 hours. Calpastatin (CAST) immunoreactivity does not change in transected nerves. C, Quantitative analysis of ratio between cut and control nerves for m-calpain and neurofilament immunoreactivity. Time points as early as 2 hours after nerve transection are included in this analysis. m-Calpain immunoreactivity in cut nerves was significantly reduced at all time points when compared to controls (P ⬍ .01 for all antibodies). Note that NF immunoreactivity did not change until between 24 and 48 hours. N1, N2, and C1 are affinity-purified antibodies targeted to amino acids 10 to 23 (N1), 20 to 36 (N2), and 531 to 554 (C1) of mouse m-calpain. (From Glass, J. D., Culver, D. G., Levey, A. I., and Nash, N. R.: Very early activation of m-calpain in peripheral nerve during wallerian degeneration. J. Neurol. Sci. 196:9, 2002, with permission.)

Degeneration in mammalian axons is not always rapid. The rate of degeneration of peripheral nerves can be slowed by exposing them to lower extracellular Ca2⫹ concentrations and by cooling within 12 hours of injury,73,268 strategies that hint at potential windows of opportunity for therapeutic manipulation of acutely injured nerves. Even more dramatically, degeneration is retarded for up to 3 weeks after axotomy in a spontaneous mouse mutant called “wallerian degeneration, slow” (Wlds), previously identified as the C57BL/Ola mouse.79,137,189,190 The Wlds mouse has attracted considerable attention since its discovery, on the basis that understanding how axonal protection is achieved in these animals might provide insights into the mechanism of wallerian degeneration in normal CNS and PNS. The Wlds mouse locus lies in the distal region of mouse chromosome 4,140 which is homologous with the human chromosomal region 1p34-1p36,141 and encodes a predominantly nuclear protein (an N-terminal fragment of ubiquitination factor E4B fused to nicotinamide adenylyltransferase). Transgenic mice expressing the Wld protein exhibit the same phenotype as the Wlds mouse, and their peripheral axons are protected in a dose-dependent manner after transection, findings which prompted Mack and colleagues to suggest that axonal protection could be mediated by altered ubiquitination or pyridine nucleotide metabolism.141 This focused attention on the potential role of the ubiquitin-proteasome system (UPS), previously implicated as a mechanism for selective protein degradation,33a as a mediator of axonal breakdown. While it now seems unlikely that the UPS contributes directly to the Wlds phenotype, recent work has indicated that UPS inhibition can profoundly delay wallerian degeneration both in vivo and in vitro.55a,283a

ACUTELY DENERVATED SCHWANN CELLS Axons and their ensheathing Schwann cells have an intimate structural and functional relationship that has been extensively documented in vivo and in vitro, in both normal and pathologic nerves. In contrast, relatively little is known about the interaction(s) between Schwann cells and other non-neuronal cells within the endoneurium (neighboring Schwann cells, perineurial cells, fibroblasts, mast cells, resident macrophages, and endothelial cells), or between Schwann cells and the extracellular matrix in which they are embedded and to which they contribute. Axonally derived signals, acting via direct contact and diffusible molecules, regulate Schwann cell genes coding for cytokines, transcription and growth factors, and myelin proteins.10,18,61,62,110,112,114,123,156,204,233 Schwann cell signals may

Mechanisms of Repair after Traumatic Injury

determine the location of intrinsic membrane glycoproteins and ion channels along the nodal and paranodal axolemma.6 Mechanical injury focally disrupts these tightly regulated relationships, and induces significant changes in Schwann cell behavior throughout the reacting portions of the nerve.

Schwann Cell Proliferation Three days after injury, most, if not all, of the Schwann cells in the injured nerve start to divide. This produces a pool of dedifferentiated daughter cells aligned within tubes of basal lamina originally secreted by their parent Schwann cells. Two Schwann cell mitogens have been described, both sourced locally within debris produced during axonolysis and myelinolysis. One is neuronally derived,200,201,210–212,214,234,235 and the other is derived from phagocytosed myelin.8 Other approximately isochronic events that may influence the gliotic burst are upregulation of erythropoietin (Epo) expression32 and downregulation of a Schwann cell–derived matrix metalloproteinase (MMP-3, also known as stromelysin-1).105 Epo may potentiate Schwann cell division through an autocrine pathway involving JAK2 phosphorylation, as happens in erythropoiesis, and in vitro evidence suggests that loss of MMP-3 would release Schwann cells from the influence of an antiproliferative fragment generated from fibronectin by MMP-3.175 When Schwann cell proliferation is delayed by application of an antimitotic agent, so too is wallerian degeneration: axonolysis continues, but other major events such as macrophage recruitment, myelinolysis, and axonal outgrowth from the proximal stump await resumption of Schwann cell division.95

Phenotype of Acutely Denervated Schwann Cells Axon-deprived Schwann cells rapidly exchange their myelinating phenotype for that of a premyelinating cell. They no longer express myelin proteins or the proteins concerned with maintaining the complex structural organization of nodes, paranodes, and incisures.217 Soon after injury, they downregulate expression of genes encoding myelin proteins P0 and PMP2287,171,277 as well as 3␤-hydroxysteroid dehydrogenase,204 which catalyzes the conversion of pregnenolone to progesterone, a neurosteroid that has been implicated in myelination12,224 because it stimulates expression of P0 and PMP22 genes51 and of Krox20 gene, which is thought to be involved in the initiation of myelination.86 Schwann cells also downregulate expression of connexin 32, a gap junction protein that forms reflexive contacts within individual myelinating Schwann cells at paranodes and incisures6,39,40; caveolin-1, a putative regulator of signal transduction and/or cholesterol transport in myelinating Schwann cells170; and E-cadherin, a calcium-dependent adhesion molecule involved in forming adherens junctions at paranodes and incisures.165

1409

In vivo, acutely denervated adult Schwann cells become “cell factories” whose products facilitate axonal regrowth either directly or indirectly. (It is worth remembering that normal, that is, intact, peripheral nerve is a poor substrate for axonal growth.13,288) What follows is an inventory of genes and their protein products known to be (re)expressed after injury (which is here taken to mean crush or transection). Injury-induced changes in the expression of some Schwann cell proteins in vivo are not always consistent with the function ascribed to that protein from in vitro studies (e.g., expression of CD910), a paradox that reinforces the difficulties inherent in extrapolating from the relative simplicity of the culture well to repair processes in the whole animal. Within days of crush or transection, dedifferentiated Schwann cells upregulate expression of the p75 low-affinity neurotrophin receptor (p75NTR)99,256; nerve growth factor (NGF)154; brain-derived neurotrophic factor (BDNF)116,169; members of the kexin/subutilisin family of pro–protein convertases—PC1 (which may play a role in the processing of pro-BDNF), PC5, PC7, and furin149; the neu receptors c-ErbB2 and c-ErbB3, which are involved in the regulation of Schwann cell differentiation and proliferation via c-ErbB–neuregulin signaling45,128,129; the cell adhesion molecules L1 and neural cell adhesion molecule, N-CAM,111,151 ninjurin 1 and 2,5 and galectin-1 (a prototypical galectin that can modulate cell-cell and cell–extracellular matrix adhesion104,107); laminin118; J1/tenascin150; ␣5␤1 fibronectin122; the ␤4 integrin subunit, although only after an initial 4-day disappearance of the protein62; netrin-1, a secreted protein that influences growth cone and axon guidance, and that is upregulated after transection but not crush147; p200, a heparin-binding adhesive glycoprotein, implicated in Schwann cell–extracellular matrix interactions43; growthassociated protein-43 (GAP-43)47,193; the cytokines leukemia inhibitory factor (LIF), interleukin (IL)-6,120,245 and tumor necrosis factor (TNF)-␣ and TNF-␣–converting enzyme228; Epo,32 a multifunctional trophic factor that is upregulated in hypoxic environments; connexin 46 (Cx46), a gap junction protein that may mediate junctional coupling in dedifferentiated Schwann cells and so facilitate the rapid diffusion of signaling molecules between cells downstream of the injury site39; the acute-phase protein hemopexin31; matrix metalloproteinases MMP-2 and MMP-9 and tissue inhibitor of metalloproteinases-1 (the latter possibly protects basal laminae from dissolution during degeneration121 [see later]); transforming growth factor (TGF)-␤1,56,218 which may induce LIF expression155; and the transcript encoding LP(A1), the canonical G protein–coupled receptor for lysophosphatidic acid, an extracellular signaling phospholipid that regulates Schwann cell morphology and adhesion in vitro.275 It is widely recognized that regrowing axons penetrate nerve autografts that have been allowed to degenerate for 3 to 4 days before harvesting more rapidly than they penetrate “fresh” grafts, in which wallerian degeneration is initiated at the time of transplantation. The changes in Schwann cell phenotype listed above, and the removal of

1410

Neuropathy Caused by Compression, Entrapment, or Physical Injury

inhibitory factors such as myelin-associated glycoprotein (MAG),161,215 all contribute to the beneficial effects of graft predegeneration. Schwann cell division, upregulation of expression of c-ErbBs and p75NTR, initiation of myelinolysis, and recruitment of exogenous macrophages also occur during primary (segmental) demyelination, wherein, by definition, axonal integrity is maintained.96 It would therefore appear that focal loss of axonal contact, rather than axonal degeneration per se, is necessary and sufficient to drive some of the phenotypic changes that characterize “reactive” Schwann cells. Schwann cells that have been contacted by regrowing axons withdraw from the cell cycle and downregulate expression of Cx46, indicating that they uncouple as they redifferentiate. In vitro evidence implicates TNF-␣ in this behavior.39,40 When appropriately directed, primarily by axonal signals, and possibly facilitated by elevated levels of prosaposin, a myelinotrophic injury–related protein,101 many Schwann cells upregulate expression of myelin-related messenger RNAs (mRNAs)71,88 and myelinate the axons

they ensheathe. Those cells that have not been reinnervated also withdraw from the cell cycle and remain within bands of Büngner. Over the ensuing months, a significant proportion disappear from the endoneurium, and many of the persisting chronically denervated Schwann cells lose their axon responsiveness.

CHRONIC AXOTOMY AND CHRONIC DENERVATION Clinical experience favors minimizing the delay before surgery.30,98,205 Experimental evidence supports this view. For example, reports that early cooling slows axonal degeneration,268 and that intraoperative electrical stimulation of the repair site for 1 hour promotes onset of motoneuron regeneration after transection and synchronizes the time taken for all available axons to cross a suture site27 (Fig. 58–6), argue persuasively that early therapeutic manipulation of the injury response, either locally and/or at the level of the neuronal cell body, is likely to be beneficial.

Repair and stimulation Repair P = 0.43

180 P = 0.008

160 Labeled motoneurons

140 120

P = 0.001

100 P = 0.31

80 60

P = 0.02

40 20 0 0—4d 4d—1wk

1wk—2wks

2wks—3wks

3wks—4wks

FIGURE 58–6 Electrical stimulation temporally compresses “regeneration stagger,” that is, it recruits motor neurons to regrow across a suture line earlier than their unstimulated counterparts. Rat femoral nerves were transected and the stumps immediately sutured. In one group, each nerve was stimulated just proximal to the repair site for 1 hour with continuous 20-Hz stimulation (100 ␮s, 3 to 5 V). Retrograde Fluororuby labeling was used to identify motor neurons as soon as possible after their axons entered the distal nerve stump. Each value on the graph was obtained by subtracting the mean number of motor neurons labeled at the beginning of an interval from the mean number labeled at the end of the interval for both stimulated and nonstimulated groups. (The means at the 0- to 4-day interval are essentially the 4-day means.) It is clear that stimulation causes a rapid increase in crossing events to a peak between 1 and 2 weeks, with a more gradual decline. In contrast, the control curve is biphasic, with a more gradual rise to an initial peak at 2 weeks followed by rapid decline and then a second peak at 4 weeks. (From Brushart, T. M., Hoffman, P. N., Royall, R. M., et al.: Electrical stimulation promotes motoneuron regeneration without increasing its speed or conditioning the neuron. J. Neurosci. 22:6631, 2002, with permission.)

1411

Mechanisms of Repair after Traumatic Injury

Chronic Axotomy Motor neurons become progressively less able to regrow their axons and reinnervate motor units with time.71 Thus, 4 to 6 months after axotomy, motor neurons produce approximately 33% of the number of axons produced by their counterparts after nerve transection and immediate repair. Several reports that the phenomenon can be manipulated have provided clues to the underlying mechanism(s) and also suggest possible therapeutic strategies. Low doses of exogenous BDNF19 or a course of daily subcutaneous injections of the immunophilin ligand FK506 (tacrolimus)247 both independently increased the number of chronically axotomized motor neurons that regrew their axons into freshly denervated distal nerve stumps. Growthassociated genes, including the BDNF gene, are only transiently expressed after axotomy in motor neurons,261 and so it is reasonable to argue that application of exogenous BDNF to chronically axotomized motor neurons at a time when endogenous levels had fallen below baseline prolonged the period of BDNF-mediated neuroprotection. These findings map nicely onto studies linking electrical activity and BDNF expression both in vitro76 and in vivo,36 and more recent demonstrations of the early upregulation of mRNAs encoding BDNF and TrkB after electrical stimulation,3 and of the promotion of the onset of motor neuron regeneration after electrical stimulation.27 FK506 improved regeneration in a model of chronic axotomy, but not in a model of chronic Schwann cell denervation,247 indicating that it acts directly on neurons rather than non-neural cells in the distal nerve stump (Fig. 58–7). It increases expression of GAP-43, a growthassociated protein, in axotomized sensory neurons, and may act similarly on chronically axotomized motor neurons. FK506 exerts its potent immunosuppressant effect via

CHRONIC AXOTOMY

A

Number of regenerated TIB motor neurons

500 400 300 200 100 0 FK506

Saline

CHRONIC DENERVATION

B 500 Number of regenerated TIB motor neurons

However, in reality, “immediate nerve repairs are nothing but exceptions in clinical practice.”247 Moreover, even when surgical intervention is undertaken relatively promptly, the length of axon to be regrown may be dauntingly long. Axons extend very slowly (1 to 3 mm/day) from a proximal nerve stump after crush or transection, and it may therefore be many months before an axon reaches and reinnervates its target. As time passes after injury, detrimental changes occur both centrally and peripherally that reduce the likelihood that these end points will be achieved. Experimental attempts to speed the rate of axonal regrowth have generally failed to demonstrate any sustained significant increase. Demonstrating an increased density of large, myelinated axons distal to a lesion within days of injury may indicate an increased rate of maturation, and/or a reduced lag time before axon sprouts cross a suture site, but it is not evidence of an increased rate of axonal elongation unless corroborated by evidence from axonal transport studies.

400 300 200 100 0 FK506

Saline

FIGURE 58–7 FK506 acts directly on neurons rather than Schwann cells. A, Numbers (means ⫾ standard error [SE]) of 2-month chronically axotomized rat tibial (TIB) motoneurons that regrew axons into freshly cut distal common peroneal nerve stumps in two experimental groups. Animals received daily subcutaneous injections of either FK506 (black bars) or saline (grey bars) (5 mg/kg rat weight). Note the significant increase in the number of regenerated tibial motor neurons with FK506 treatments. B, Numbers (means ⫾ SE) of freshly axotomized tibial motoneurons that regrew axons into 2-month chronically denervated common peroneal distal nerve stumps compared under conditions as in A. FK506 did not exert any effect on the number of tibial motor neurons that regrew axons. Statistical analysis by one-way analysis of variance and Student’s t-test. (From Sulaiman, O. A., Voda, J., Gold, B. G., and Gordon, T.: FK506 increases peripheral nerve regeneration after chronic axotomy but not after chronic Schwann cell denervation. Exp. Neurol. 175:127, 2002, with permission.)

binding to immunophilin FKBP-12 and the subsequent inhibition of calcineurin activity, but appears to exert its primary effect on neurons by binding to another immunophilin, FKBP-52 (otherwise known as heat shock protein 56), a component of mature steroid receptor complexes.80 In this context, it may be worth noting that the

1412

Neuropathy Caused by Compression, Entrapment, or Physical Injury

lar, and is far less able to support regrowing axons than its acutely denervated counterpart. There is a marked decrease in blood flow through vasa nervorum within long-term denervated distal stumps (both in the endoneurial microvascular compartment and in extrinsic microvessels) and a reduction in the caliber of new microvessels, all changes likely to produce an endoneurial microenvironment that does not favor regeneration103 (Fig. 58–8).

neuroprotective effect of a peripheral benzodiazepine receptor ligand that promotes motor neuronal survival and axonal regrowth after facial nerve transection is also thought to be steroid mediated.67

Chronic Denervation The endoneurium of a chronically denervated distal nerve stump is fibrotic,209,250,251 ischemic, and relatively acelluA

DENERVATED PERIPHERAL NERVE TRUNKS

*

40

Blood flow (ml/100g/min)

35 30

*

25 20

*

15

*

10 5 0 Intact

48h

14d

1mo

3mos

6mos

6mos 6mos sham resutured

Duration of denervation

B 800

* 700 600

* RBC flux

500 400 300 200

*

*

100 0 Intact

48h

14d

1mo

3mos

6mos

Duration of denervation

6mos 6mos sham resutured

FIGURE 58–8 Injury disturbs the microvascular bed of a peripheral nerve. Blood flow measurements are shown in intact rat sciatic nerve, distal nerve stumps that had been denervated for periods from 48 hours to 6 months, and distal nerve stumps that had been reinnervated for 6 months. Note that flow through the extrinsic vascular plexus was not restored to normal values after reinnervation. Endoneurial (A) and epineurial (B) compartments were analyzed by microelectrode hydrogen clearance polarography and laser Doppler flowmetry, respectively. There is a hyperemic response in the early stages of denervation, but this is succeeded by a reduction in blood flow when denervation is prolonged. *P ⬍ .005. (From Hoke, A., Sun, H. S., Gordon, T., and Zochodne, D. W.: Do denervated peripheral nerve trunks become ischemic? The impact of chronic denervation on vasa nervorum. Exp. Neurol. 172:398, 2001, with permission.)

Mechanisms of Repair after Traumatic Injury

Schwann Cell Death and the Axon-Independent Schwann Cell Although the gliotic response is one of the most obvious aspects of the injury response in the PNS, it should be remembered that many axon-deprived Schwann cells die during repair, apparently by apoptosis.85,253 Two separate pathways have been implicated: a Bcl-2–blockable pathway initiated on loss of trophic support and a Bcl2–independent cytokine response modifier A pathway mediated via p75NTR.236 In this latter context, it is interesting that more Schwann cells may survive in the chronically denervated distal stumps of transected nerves in p75NTR knockout mice than in their wild-type counterparts.66

1413

Although many Schwann cells disappear from a denervated endoneurium with time, others persist within collapsed bands of Büngner for at least 18 months after injury20,259 (Fig. 58–9). The ability to survive independent of axonal contact appears to be a unique property of some adult Schwann cells, because cells isolated from embryonic or perinatal nerve do not survive in the absence of axonally derived growth factors.52,85,253,267 Ultrastructurally, adult axon-independent Schwann cells can be identified by their delicate branching processes of electron-dense cytoplasm enclosed by basal laminae. Immunohistochemically, they remain S100 and laminin positive, but with time they lose other phenotypic markers, and these changes may adversely affect their ability to interact with ingrowing axons. Denervated Schwann cells upregulate expression of c-ErbB tyrosine kinase receptors for several months after injury, which is consistent with the role that c-ErbB/neuregulin signaling plays in Schwann cell biology. However, this phenotype is not sustained indefinitely, and Schwann cells that have been deprived of axonal contact for 4 to 6 months in vivo downregulate expression of receptors for both c-ErbB128 and p75NTR.282 Distal nerve stumps denervated for this length of time are unable to support axonal ingrowth71,94 (Fig. 58–10), a coincidence that prompts the speculation that axon–Schwann cell signaling becomes impaired with time. If that is the case, the finding that chronically denervated Schwann cells reexpress c-ErbB receptors and respond to the appropriate ligands (the glial growth factor subfamily of axon-related neuregulins) when they are explanted in vitro129 supports the view that it will often be necessary to maintain denervated Schwann cells in an acutely denervated, axon-responsive state in vivo for many months after clinical nerve repair.

INFLAMMATION AND WALLERIAN DEGENERATION

FIGURE 58–9 Some Schwann cells persist within a chronically denervated distal nerve stump. Electron micrograph of a transverse section through the distal stump of a human median nerve, 9 months after injury. The chronically denervated endoneurium contains several Schwann cell tubes, each containing processes of dense Schwann cell cytoplasm and surrounded by a convoluted basal lamina (arrows). Other cell processes, not enclosed by a basal lamina, are probably endoneurial fibroblasts. (Original magnification, ⫻10,000.) (From Terenghi, G., Calder, J. S., Birch, R., and Hall, S. M.: A morphological study of Schwann cells and axonal regeneration in chronically transected human peripheral nerves. J. Hand Surg. Br. 23:583, 1998, with permission.)

Cellular evidence of inflammation is seen first at the lesion site, which becomes infiltrated with T cells, neutrophils, and macrophages within 2 days.34,188,265 Expression of neutrophil chemoattractants is similarly limited to the first 2 to 3 days after injury, and is largely confined to the site of trauma, which accords with the very limited neutrophil response seen after injury in the PNS. By 4 days, macrophages have invaded the entire extent of the distal nerve stump and the “reactive zone” at the distal end of the proximal stump.

The Cytokine Network The first components of what becomes a network of proand anti-inflammatory cytokines are expressed in the distal stump within hours of injury. Immunohistochemical and in situ hybridization studies have shown that acutely denervated Schwann cells, resident macrophages, mast

% Immunostaining area

A

% Immunostaining area

0

*

*

*

0 0

0

* **

**

0

5

10

15

20

25

30

35

40

0

0

**

* *

5 10 15 20 25 30

** ** *

* *

5 10 15 20 25 30

**

2 MONTHS

Days after re-innervation

5 10 15 20 25 30

*

0

5

10

15

20

25

30

35

40

45

c-erbB4

5 10 15 20 25 30

**

1 MONTH

c-erbB2

0

5

10

15

20

25

30

35

40

0

5

10

15

20

25

30

35

40

45

0

0

*

*

**

*

**

*

*

5 10 15 20 25 30

*

5 10 15 20 25 30

*

**

4 MONTHS

0

5

10

15

20

25

30

35

40

0

5

10

15

20

25

30

35

40

45

0

0

FIGURE 58–10 Chronic denervation depresses axonal penetration of distal nerve stumps. A, Quantitative analysis of c-ErbB2 and c-ErbB4 immunolabeling areas in transverse sections taken from rat sciatic nerves 1 mm (DS1) and 1 cm (DS2) distal to the site of crush or transection/coaptation. Endoneurial c-ErbB immunolabeling areas decrease as the chronicity of denervation increases. Immunolabeling areas were determined by computerized image analysis and expressed as percentage immunolabeled area/measuring frame (mean ⫾ standard error of the mean [SEM]; n ⫽ 3). Levels of immunolabeling measured in normal, unoperated control nerves are shown as dotted lines. *P ⬍ .05 and **P ⬍ .01 versus normal controls. Figure continued on opposite page

5 10 15 20 25 30

0 0

0

0

5 10 15 20 25 30

5

15

5

*

5

15 10

*

20

10

*

15

*

10

*

20

20

25

**

25

25

**

30

**

30

30

35

35

**

35

0

5

10

15

20

25

30

40

*

*

5 10 15 20 25 30

*

**

35

40

45

40

0

*

**

CRUSH

40

0

5 10 15 20 25 30

0

10

15

5

*

*

20

25

5

10

15

20

25

30

**

30

**

35

35

40

40

**

45

45

IMMEDIATE COAPTATION

*

DS1 DS2

5 10 15 20 25 30

*

DS1 DS2

5 10 15 20 25 30

**

6 MONTHS

1415

Mechanisms of Repair after Traumatic Injury

B

C 120

Coaptation Crush

% Innervated Schwann cell columns

% Innervated Schwann cell columns

120 100 80 60 40 20 0

14 days 28 days

100 80 60 ** 40 20

**

**

0 2

4

7

14

28

Survival time (days)

1

2

4

6

Months after transection

FIGURE 58–10 Continued B and C, Quantitative analysis of the extent of axonal penetration of variously treated distal nerve stumps. Counting was performed on transverse sections 1 cm distal to the suture line. Each point represents the mean percentage of the total number of S100-positive Schwann cell tubes that contained ␤-tubulin III–positive axons in double-labeled sections (mean ⫾ SEM; n ⫽ 3 or 5). B, After crush or transection/immediate coaptation; almost all Schwann cell columns have been reinnervated by day 28. C, Fourteen or 28 days after coaptation of acutely transected proximal stumps to distal nerve stumps denervated for 1, 2, 4, or 6 months prior to suture. *P ⬍ .05 and **P ⬍ .01 versus 1-month denervated distal stumps. (From Li, H., Terenghi, G., and Hall, S. M.: Effects of delayed re-innervation on the expression of c-erbB receptors by chronically denervated rat Schwann cells in vivo. Glia 20:333, 1997, with permission.)

cells, fibroblasts, and endothelial cells all contribute to elevated endoneurial levels of TNF-␣ and IL-6, and that these are subsequently amplified by cytokines and chemokines released by recruited macrophages. The literature reveals subtle differences in temporal patterns of cytokine gene induction, that are most likely methodologic (e.g., timing of assays, inclusion of neuroma tissue in test tissue, type of lesion, sensitivity of detection protocols). The broad picture that has emerged is that transcripts for IL-1␣ and -1␤, IL-6, and IL-10 and for granulocyte-macrophage colony-stimulating factor peak within 1 day and then fall during the first week. Transcripts for IL-18 (formerly interferon-␥ [IFN-␥] inducing factor), IFN-␥, and TNF-␣ also rise during day 1, but their expression is sustained into week 2; transcripts for TGF-␤1 and the p40 subunit of IL-12 increase steadily during the first 2 weeks14,121,164,203,206,226,242,257,272 (Fig. 58–11). Of all of these molecules, the pleiotropic cytokine TNF-␣ has attracted considerable attention, probably because it has been implicated in many different activities, including macrophage recruitment and the induction of neuropathic pain after peripheral nerve injury (see later).

Macrophages Activated macrophages play multiple roles in wallerian degeneration. Long recognized for their role in myelin degradation and subsequent removal, they are also an

important source of cytokines, chemokines, and neuroactive factors. Resident Endoneurial Macrophages Five percent to 9% of the total endoneurial population is resident macrophages. Until recently, almost nothing was known about the role of these cells in the injury response, because it is impossible to distinguish reliably between resident and recruited macrophages on the basis of immunophenotype. However, the use of radiation bone marrow chimeric rats, created by transplanting WT Lewis rat bone marrow into TK-tsa transgenic Lewis rats, has permitted unequivocal identification of resident endoneurial macrophages and has revealed that they respond extremely rapidly to injury (in which respect they resemble microglia in the CNS). Transgene-positive endoneurial macrophages begin myelin phagocytosis within 2 days of injury, that is, at a time that precedes bulk macrophage invasion from the blood stream.174 Macrophage Recruitment and Fate Within 3 days of crush or transection, hematogenously derived macrophages flood into the endoneurium distal to the injury, and others arrive even earlier at the lesion site. They cross the blood-nerve barrier and Schwann cell basal laminae, aided by elevated expression of MMPs 2, 7, and 9.105,230 Many of the recruited macrophages express the chemokine receptor CCR-2, the major

1416

Neuropathy Caused by Compression, Entrapment, or Physical Injury

IL-1β

IL-6

1,6

2,0

1,2

Relative mRNA levels

Relative mRNA levels

1,4

1,0 0,8 0,6 0,4

1,5

1,0

0,5

0,2 0,0

0,0 ctr

1d

2d

4d

7d

14

ctr

1d

Days after crush

2d

4d

7d

14

Days after crush

IL-10

Relative mRNA levels

1,0

0,8

0,6

0,4

0,2

0,0 ctr

1d

2d

4d

7d

14

Days after crush

FIGURE 58–11 Profiles of cytokine expression in wallerian degeneration. Semiquantitative reverse transcription–polymerase chain reaction (RT-PCR) analysis of IL-1␤, IL-6, IL-10, IL-18, IFN-␥, TNF-␣, and IL-12 p40 mRNA expression in the distal stumps of sciatic nerves from C57Bl mice at varying times up to 14 days after sciatic nerve crush at the sciatic notch (15 pooled nerve segments/time point; results expressed after normalization against GAPDH mRNA; RT-PCR performed using mousespecific primer pairs). (From Stoll, G., Jander, S., and Myers, R. R.: Degeneration and regeneration of the peripheral nervous system: from Augustus Waller’s observations to neuroinflammation. J. Peripher. Nerv. Syst. 7:13, 2002, with permission.) Figure continued on opposite page

high-affinity receptor for the chemokine monocyte chemotactic protein-1 (MCP-1). They target Schwann cells within degenerating nerve, which upregulate expression of MCP-1 with a time course that precedes and coincides with the period of macrophage influx.34,231,265 Macrophage recruitment is impaired in: CCR-2 knockout mice231; C57BL/Ola mice, which exhibit reduced production of MCP-134; intracellular adhesion molecule-1 (ICAM-1)–deficient mice270; TNF-␣⫺/⫺ mice130 (although whether TNF-␣ is directly chemotactic for

circulating monocytes or acts indirectly via upregulation of adhesion molecules such as ICAM-1 and vascular cell adhesion molecule or chemokines such as MCP-1 is yet to be established); and after blocking Schwann cell proliferation.95 Most lipid-laden macrophages ultimately disappear from the endoneurium. A study using in situ hybridization with a Y chromosome–specific DNA probe has revealed that some are lost by local apoptosis and others migrate to draining lymph nodes and the spleen.119

1417

Mechanisms of Repair after Traumatic Injury

IL-18

IFN-γ

0,6 0,5

0,8

Relative mRNA levels

Relative mRNA levels

1,0

0,6

0,4

0,2

0,3 0,2 0,1

0,0

0,0 ctr

1d

2d

4d

7d

14

ctr

1d

2d

4d

Days after crush

Days after crush

TNF-α

IL-12p40

2,5

7d

14

7d

14

2,0

2,0 Relative mRNA levels

Relative mRNA levels

0,4

1,5

1,0

0,5

0,0

1,5

1,0

0,5

0,0 ctr

1d

2d

4d

7d

14

Days after crush

ctr

1d

2d

4d

Days after crush

FIGURE 58–11 Continued

Myelin Phagocytosis The extent to which Schwann cells participate in myelin breakdown in wallerian degeneration is unclear. Although they are capable of substantial myelin degradation in the absence of macrophages in vitro,65 in vivo after whole-body irradiation,191 and in ICAM-1–deficient mice in which macrophage recruitment is impaired,270 it seems likely that myelin degradation and subsequent removal of the debris normally involves cooperation between macrophages and Schwann cells. Macrophages appear to phagocytose myelin by opsonindependent (via Fc receptor or type 3 complement receptor49) and opsonin-independent (via macrophage

scavenger receptors48,58,179) pathways, whereas Schwann cells probably phagocytose myelin by an opsonin-independent process.102 In vitro, phagocytosed myelin induces macrophages to synthesize and secrete TNF-␣ (an activity that is suppressed by anti–complement receptor-3 antibodies46), and may induce synthesis of a Schwann cell mitogen.8 Macrophage-mediated removal of MAG, a known inhibitor of neurite elongation in vitro,255 probably facilitates axonal ingrowth into a distal nerve stump. Indeed, treatment with anti-MAG antibodies enhances regeneration and facilitates selective reinnervation of motor Schwann cell tubes in mice in vivo.161

Neuropathy Caused by Compression, Entrapment, or Physical Injury

MMP-9 0.01

Fraction of β-actin

MMPs are a family of over 20 enzymes that degrade all the components of the extracellular matrix and other nonmatrix substrates, altering their functions and releasing biologically active fragments that can affect cell growth, migration, intercellular communication, and apoptosis.157 Increased expression of activated MMPs in Schwann cells, macrophages, and endothelial cells has been correlated with a number of key stages in wallerian degeneration, including the breakdown and subsequent remodeling of the blood-nerve barrier and Schwann cell basal laminae (which are both associated with macrophage invasion); Schwann cell proliferation; myelin degradation; and clipping of TNF-␣ from its transmembrane precursor (see later). Intraperitoneal injections of hydroxamate (Ro 31-9790), a nonspecific MMP inhibitor, or local application of an MMP-9–specific antibody delays wallerian degeneration.230 Expression of MMP-2, MMP-7, and MMP-9 mRNA and protein is upregulated rapidly after nerve transection,105,121,230,258 whereas expression of MMP-3 is downregulated105 (Fig. 58–12). The timing of peaks of expression coincides with times at which the blood-nerve barrier becomes incompetent,75 granulocytes accumulate at the lesion site, and macrophages invade the endoneurium, while the timing of downregulation of MMP-3 coincides with the onset of Schwann cell proliferation (see above). MMP-2 is apparently upregulated in the contralateral nontransected nerves.105 (Over the years numerous reports have mentioned the upregulation of various proteins in contralateral, ostensibly intact, nerves after axotomy; the author is unaware of any detailed or systematic analysis of the phenomenon.) There is increasing evidence from studies in both the CNS and PNS that chondroitin sulfate proteoglycans (CSPGs) can inhibit axonal regrowth, and may suppress the growth-promoting activity of extracellular matrix components.21,22,68,159,288 Paradoxically, normal nerve is refractory to axonal ingrowth,24,93 whereas regrowing axon sprouts readily penetrate Schwann cell tubes in degenerating distal stumps, even though both normal nerve and distal stump are rich in laminin. In the PNS, it appears that CSPGs normally inhibit the neurite-promoting activity of the extracellular molecule laminin.64,287 Indeed, it has been suggested that the presence of CSPGs within the endoneurium of a normal nerve might suppress spontaneous axonal sprouting, especially at nodes of Ranvier. Cryoculture assays in which neurite outgrowth from adult dorsal root ganglion neurons explanted onto unfixed frozen sections of normal nerve was compared with that on sections of nerve pretreated with chondroitinase ABC (which removes CSPG glycosaminoglycans) revealed that neurites

A

0.001

0.0001

0.00001 0

1

3

5

7

10

15

20

Days post cut

B

MMP-3 0.001

Fraction of β-actin

MATRIX METALLOPROTEINASES AND CHONDROITIN SULFATE PROTEOGLYCANS

0.0001

0.00001 0

1

3

5

7

15

20

Days post cut

C

MMP-7 0.001

Fraction of β-actin

1418

0.0001

0.00001 0

1

3

5 7 10 Days post cut

15

20

FIGURE 58–12 Semiquantitative reverse transcription–polymerase chain reaction analysis of the mRNA expression of MMPs regulated during wallerian degeneration in rat sciatic nerve. Results are expressed as a fraction of ␤-actin. (mean ⫾ standard deviation; n ⫽ 4 at each time point). (From Hughes, P. M., Wells, G. M., Perry, V. H., et al.: Comparison of matrix metalloproteinase expression during wallerian degeneration in the central and peripheral nervous systems. Neuroscience 113:273, 2002, with permission.)

Mechanisms of Repair after Traumatic Injury

A

1419

B

FIGURE 58–13 Neurite outgrowth from embryonic chick dorsal root ganglion neurons cultured on sections of normal adult rat sciatic nerves (A) and nerve pretreated with chondroitinase ABC (B). Neurites follow longitudinal paths of exposed Schwann cell basal laminae. Chick neurons are immunofluorescently labeled using a species-specific monoclonal antibody to neural cell adhesion molecule. Bar: 50 ␮m. (From Zuo, J., Hernandez, Y. J., and Muir, D.: Chondroitin sulfate proteoglycan with neurite-inhibiting activity is up-regulated following peripheral nerve injury. J. Neurobiol. 34:41, 1998, with permission.)

grew better on the latter (Fig. 58–13). It was inferred that normal nerve had been converted from a nonpermissive to a permissive state by degradation of the CSPG core protein. MMP-2 and MMP-9, expressed by Schwann cells and macrophages during wallerian degeneration, have been implicated as the natural candidates responsible for this conversion. In vivo, local application of chondroitinase ABC improved axonal regrowth across the interface of coapted sciatic nerve stumps289 and enhanced axonal regrowth into acellular nerve grafts.117 Disinhibition of the neuritepromoting activity of laminin by chondroitinase therefore seems to mimic the naturally occurring process mediated by MMPs during degeneration, and may be exploitable therapeutically as a means of expediting the initial penetration of nerve grafts or distal stumps. That this disinhibition occurs exclusively on the luminal face of Schwann cell basal lamina tubes within the distal nerve stump accords with the findings that regrowing axons only associate with the luminal aspect of basal lamina tubes when they reenter a distal nerve stump, and that regrowing axons and their associated non-neuronal cells grow exclusively along the inner surface of muscle sarcolemma in denatured acellular muscle grafts, and of Schwann cell basal lamina tubes in acellular nerve grafts.57,117

KNOCKOUT MICE The roles of some of the neuronal or Schwann cell–derived factors that are upregulated in degeneration and subsequent repair have been examined using gene knockout mice. Perhaps not surprisingly, given the enormous complexity of

the injury response, most studies to date report either that absence of individual factors has little or no significant effect74,81 or that axonal regrowth is delayed but not prevented. Results have sometimes been contradictory for the same molecule (e.g., IL-6108,284) or for different members of a family of molecules (e.g., members of the mammalian leukocyte common antigen–related subfamily of protein tyrosine phosphatases).160,280 Lack of significant effect may reflect a functional redundancy within a particular family of molecules, coupled with compensatory expression of other members of that family (e.g., TGF-␣).279 Delayed functional recovery could reflect a prolonged lag phase before axons and their accompanying cells leave the proximal stump, as in mice lacking tissue-type plasminogen activator (PA), urokinase-type PA, or plasminogen genes, in which absence of tissue-type PA presumably might compromise the ability of Schwann cells and growth cones to remodel their substrate and so retard their outgrowth.229 Knockout of either class A178 or class B58 macrophage scavenger receptors disturbs clearance of myelin debris and reutilization of lipids from degraded myelin after crush injury.

NEUROPATHIC PAIN Axotomy often produces pain, which typically presents as ectopic mechanosensitivity, hypersensitivity, and tenderness. It is not known exactly how or where this “neuropathic” pain is generated, or how painful states are maintained after injury. It is believed that events that take place in the dorsal horn, dorsal root ganglion, and injury site contribute to the phenomenon at varying times postaxotomy,33 and that

1420

Neuropathy Caused by Compression, Entrapment, or Physical Injury

release of inflammatory mediators such as TNF-␣ and NO, the latter produced by local NO synthase activity,125–127,286 may mediate the early stages of neuropathic pain. Two animal models, spinal nerve ligation115 and chronic constriction injury (CCI),15 are providing important insights into potential mechanisms of neuropathic pain. Both protocols produce behavioral signs from which it is assumed that the animal is experiencing pain, including hyperalgesia to mechanical and thermal stimuli (the latter inferred from the rapidity with which an affected paw is withdrawn from a source of radiant heat).

A L5

S2 S1

L4 R

B

E SA Pinch Electrical

SA

0

50 100 150 200 250 300s

Spinal Nerve Ligation Model The weight of current evidence supports the view that, when the L5 spinal nerve is ligated in rats, inputs from sources other than the injured spinal nerve not only initiate but also maintain pain behavior (Fig. 58–14). It has been suggested that signals derived from denervated targets and/or denervated Schwann cells somehow alter the physiologic responses of intact nociceptors belonging to neighboring, intact, spinal nerves such that they rapidly acquire abnormal spontaneous activity and a sensitivity to catechols.33,278 This proposal is based on the fascicular anatomy of the distal branches of the sciatic nerve, which contain axons from both L4 and L5 roots, and is supported by the demonstration that C fibers from more than one spinal root coexist within Remak bundles.176

Chronic Constriction Injury Model and the Role of TNF-␣ CCI involves placing three or four loosely constricting ligatures around a sciatic nerve to produce wallerian degeneration, especially of large-caliber axons. In CCI, as in wallerian degeneration, two peaks of thermal hyperalgesia have been reported during the first week after injury. A good correlation has been demonstrated between the timing of these peaks, increases in endoneurial levels of TNF-␣ and its p55 TNF receptor, and expression of MMP-2 and MMP-9.228 The first peak corresponds to a very rapid upregulation of MMP-9 activity and an increased local expression of the transmembrane 26-kDa TNF-␣ protein, which is released by resident Schwann cells, mast cells, and macrophages. The later peak corresponds to upregulation of MMP-2 activity and an increase in active soluble 17-kDa TNF-␣ protein probably released by recruited macrophages.227 Levels of TNF produced locally at the lesion site are believed to be supplemented from central sources such as dorsal root ganglion sensory neurons (reflecting injury-induced increases in expression and anterograde transport of TNF216). Evidence from another experimental protocol also implicates TNF-␣ in the production of neuropathic pain.

C Skin pinch

0.0

0.2

0.4

0.6

0.8

1.0s

D

0 2 4 6 8 10 ms

F

RF CV = 0.5 m/s 0

10

20

30

40

50

60 ms

FIGURE 58–14 Spontaneous activity in a typical C-fiber afferent recorded 2 days after an L5 spinal nerve lesion. A, Teased-fiber techniques were used to record activity from single nerve fibers of the L4 spinal nerve (R) in normal animals and in animals after ligation and transection of the L5 spinal nerve. B, The presence of spontaneous action potential activity (SA) was assessed over a 5-minute recording interval. For this example, low-level spontaneous activity (three action potentials/5 minutes) was present in one C-fiber afferent. C, The fiber with spontaneous activity also responded to pinching of the skin. D, Electrical stimulation at the sciatic nerve (S1) produced three discrete action potential waveforms at C-fiber latencies (two are superimposed at 32 ms). E, The action potential waveform at 43 ms had the same shape as the spontaneous and mechanically induced activity, providing evidence that the spontaneous activity came from this C-fiber nociceptor. F, The receptive field (RF) for this afferent was located near the knee. Thus, in this filament, one of three C fibers was spontaneously active. (From Wu, G., Ringkamp, M., Hartke, T. V., et al.: Early onset of spontaneous activity in uninjured C-fiber nociceptors after injury to neighboring nerve fibers. J. Neurosci. 21:RC140, 2001, with permission.)

In these studies, intraneural injection of TNF produced wallerian degeneration and demyelination (albeit with a time course that suggests that these pathologic changes are secondary to microvascular plugging rather than a direct response to TNF-␣202), as well as thermal hyperalgesia and

1421

Mechanisms of Repair after Traumatic Injury

mechanical allodynia.271 Preemptive treatment with thalidomide, a selective blocker of TNF production in activated monocytes,213 partially prevents the development of hyperalgesia caused by CCI in rats,237 whereas delaying until the onset of hyperalgesia fails to ameliorate painrelated behavior, presumably because the affected nerves are exposed to other algesic substances. Distal stump

Proximal stump

AXONAL REGROWTH Within hours of injury, sprouts bud from the damaged axons at their tips and/or at nodes of Ranvier at the interface between the normal and “reactive” regions of the proximal stump. They grow toward the lesion site inside the basal lamina tubes that enveloped their parent axons, in association with the dedifferentiated progeny of the original Schwann cells and a transient population of invading macrophages. When the continuity of Schwann cell basal lamina tubes has not been disturbed by injury, regrowing axons cross the lesion site with minimal deflection and elongate within the distal stump. In experimental models in which transection is immediately followed by suture of the nerve stumps, most axon sprouts ultimately enter the distal stump, although they may cross the suture line asynchronously, exhibiting “regeneration stagger.”4,27 Sprouts from individual axons typically enter several Schwann cell tubes, any or all of which might be functionally inappropriate: the apparent imprecision of this process has obvious implications for functional outcome. The stimulus for axonal sprouting is unknown. Although factors such as BDNF have been implicated in the induction of sprouting, somewhat surprisingly there is no evidence that levels of NGF, BDNF, or neurotrophin-3 are elevated in the proximal stump after transection.285 Clinically, axon sprouts may be blocked or deflected by scarring and fibrosis at the lesion site, phenomena that are less commonly encountered in experimental models, probably because the delay to suture is minimal in the laboratory but may be of the order of days or weeks in the clinic.

Axonal Regrowth across an Interstump Gap: Use of Regeneration Chambers Where injury creates an interstump gap, regrowing axon sprouts and their associated cells are confronted with nonneural territory that must be negotiated before they can access the distal stump. Much of what is currently known about their behavior in this non-neural context is based on studies in which nerve stumps are sutured into the opposite ends of a tube (hence the generic name of “entubulation” to describe the procedure). Early accounts of the outgrowth of cells from the opposing nerve stumps and into the lumen of a tube (otherwise known as a regeneration chamber or conduit) (Fig. 58–15) were largely descriptive. More recent studies, particularly preclinical

Fluid accumulation Conduit wall

Fibrin bridge and fibroblasts

Schwann cells and regrowing axons

Regrowing axons which have crossed the gap

FIGURE 58–15 Diagram of a generic regeneration chamber. This cartoon summarizes the progression of major events that occurs when a peripheral nerve is transected and the proximal and distal nerve stumps are immediately sutured into a silicone chamber. The most common interstump gap lengths used in experimental studies of axonal regrowth through conduits in rats are in the range of 0.5 to 1 cm. (From Heath, C. A., and Rutkowski, G. E.: The development of bioartificial nerve grafts for peripheral-nerve regeneration. Trends Biotechnol. 16:163, 1998, with permission.)

evaluations of materials and/or fillings for conduits, have included some element of functional testing in their experimental protocol.11,38,44,50,89–91,101,124,131,132,162,163,172 Caveat: These experimental studies do not seek to emulate clinical presentations of traumatic nerve injury. Thus the interstump gap length that is created is typically 1 cm or less; the host animals are young and healthy, and are usually rodents (in which axonal regrowth is notoriously robust); the nerve is transected with a sharp blade; the

1422

Neuropathy Caused by Compression, Entrapment, or Physical Injury

nerve stumps are sutured into the conduit immediately; and there is minimal disturbance of the wound bed.

Cellular Outgrowth into a Regeneration Chamber Axon sprouts almost always emerge from a proximal nerve stump accompanied by a co-migrating population of non-neuronal cells organized into minifascicles. These are small bundles of axons and Schwann cells, each surrounded by at least one layer of perineurial cells, usually containing fibroblasts, and sometimes penetrated by capillaries. Minifascicular organization appears to be a continuation of the process of compartmentalization that begins within the distal end of the proximal nerve stump.173 It characterizes the outgrowth across an interstump gap, and is lost when axons enter a distal nerve stump, but is usually retained when the outgrowth ends abortively within a neuroma. The mechanism underlying axonal allocation to minifascicles, which may be of importance in determining how axons are dispersed across an interstump gap, is unknown. Schwann cells, endothelial cells, fibroblasts, and fibroblast-like cells (which are often assumed on the basis of their morphology, but not immunophenotype, to be perineurial cells), grow out from the cut end of a distal nerve stump soon after transection.223,263,276 The cellular outgrowth lacks the obvious organization of its proximal counterpart, suggesting that minifascicle formation is axonally driven. Perineurial differentiation may be similarly influenced.196 Because minifasciculation does not normally occur within a distal stump or nerve graft (providing the latter possesses an intact perineurium), it may be a mechanism for controlling the osmotic milieu of regrowing axons. This view is supported by the appearance of peripheral nerves in dhh⫺/⫺ mice, which do not express the signaling protein Desert hedgehog (Dhh), implicated in Schwann cell–fibroblast signaling during perineurial development. The perineurial sheaths of dhh⫺/⫺ mice are anatomically discontinuous and functionally incompetent, and the endoneurium is compartmentalized in a way that is strikingly reminiscent of postaxotomy minifascicles.185

Precision of Reinnervation Early experiments using straight and inverted Y-shaped regeneration chambers established two points: first, that there were limits to the length of the interstump gap that axons would cross in a tube (ⱕ1 cm in adult rats, ~3 cm in primates), and second, that axons grew preferentially toward nerve when challenged with a selection of appropriate and inappropriate targets.194 Although this last point suggests that some degree of target specificity exists, the level at which axonal regrowth can be said to exhibit specificity remains controversial. There is no doubt that, providing the gap length does not exceed 1 cm, adult rat axons and their associated cells behave as though they are responding to “tropic” signals emanating from a distal stump.2 The extent

to which specificity exists at the fascicular or nerve trunk level has been difficult to establish, perhaps because it is only unmasked in particular experimental protocols, and cannot be demonstrated when axons are challenged to enter grafts rather than intact distal stumps.1,28,194 Restoration of function requires that a sufficient number of regrowing axons reach and reinnervate the correct target organs. For motor nerves, reinnervation of an inadequate number of muscle fibers, or of inappropriate antagonists, would compromise functional outcome. Motor reinnervation is predictably specific after a crush injury,23,252 whereas after transection and repair the accuracy of reinnervation is relatively poor,17,139 not least because motor neurons retain their original activation patterns.82 The use of simultaneous and sequential double-retrograde tracing techniques in the femoral nerve model (in which sensory and motor neurons may be distinguished unambiguously) has revealed that, in this model at least, specificity is not established as axons cross a suture line and enter a distal stump25,26 (Fig. 58–16). Rather, reinnervation of bands of Büngner appears to be a serendipitous process, in which regrowing axon collaterals that have crossed a suture site initially enter both motor and sensory Schwann cell tubes. Specificity is subsequently generated as a post hoc phenomenon as collaterals lying within inappropriate tubes die back, a process termed preferential motor reinnervation (PMR) for motor axons.25 A limited degree of regeneration specificity has also been demonstrated for sensory afferents.146 Very little is known about the molecular basis for preferential reinnervation. To date, the only identified pathway marker is the L2/HNK-1 carbohydrate epitope, which is selectively expressed on Schwann cells associated with motor axons,152,153 and which may be involved in generating PMR via the “molecular memory” of the progeny of the original motor Schwann cells.25,26 Clearly events that occur as axonal growth cones enter a distal stump or nerve graft are key to the final outcome of repair, and yet currently almost nothing is known about them. It may also be significant that axotomized motor neurons downregulate expression of the chemorepulsive semaphorin Sema-III/Coll-I soon after injury, and sustain this phenotype until their target muscle has been reinnervated.186 If this results in lower levels of Sema-III/Coll-I reaching the tips of regrowing motor axons, then axonal growth across the suture site and into Schwann cell tubes in the distal stump might be facilitated. It is tempting to speculate that the upregulation of netrin-1 mRNA in Schwann cells after transection, but significantly not after crush,147 contributes to the process by deflecting some regrowing axons from inappropriate Schwann tubes in a distal nerve stump. Reduction of collateral sprouting, for example, by intraoperative focal application of neutralizing antibodies to soluble neurotrophic factors,244 coupled with measures to facilitate selection of appropriate tubes, should improve the quality of reinnervation. Once an axon has entered a band of Büngner in a nerve graft and/or a distal nerve stump, then it is physically committed to that band and will either continue to grow

Mechanisms of Repair after Traumatic Injury

A

1423

B

C

FIGURE 58–16 The femoral nerve model for determining regeneration specificity. A, The motor and sensory neuron pools innervating the terminal branch of the femoral nerve to the quadriceps branch (QB) are labeled retrogradely using DiI. Two weeks later the parent femoral nerve is transected and repaired. (One of two repair procedures was used in this particular experiment: either an entubulation repair of a high transection, wherein QB axons are dispersed throughout the entire femoral nerve, or a direct suture repair of a low transection, wherein QB axons are sequestered within a segment of the femoral nerve.) B, Four weeks later, the QB is re-transected just proximal to the site of initial DiI application, and exposed to Fluorogold. Only neurons that have regrown their axons into the QB after transection and repair of the parent femoral nerve will be double-labeled with DiI and Fluorogold on cryostat sections. C, Photomicrographs of a regenerated motor neuron labeled with both tracers: 1. Rhodamine optics (DiI); 2. Fluorogold optics (FG); 3. double exposure using successive FG and rhodamine filter sets. (From Madison, R. D., Archibald, S. J., and Brushart, T. M.: Reinnervation accuracy of the rat femoral nerve by motor and sensory neurons. J. Neurosci. 16:5698, 1996, with permission.) See Color Plate

within it or will degenerate. For this reason, it was suggested that entubulation of nerve ends, and the deliberate production of a short gap between the stumps within which outgrowing axons would be bathed in neurotrophin-rich wound fluid, might have an advantage over direct coaptation of nerve stumps, particularly when the injured nerve was a mixed nerve.199,283 However, comparisons of entubulation and end-to-end suture have not shown entubulation to confer any advantage. Indeed, Brushart et al. noted that, when the gap length within the tube is greater than 2 mm, “an enclosed gap is not an acceptable substitute for nerve graft when constructing a nerve that serves multiple functions.”28

NERVE AUTOGRAFTS AND THEIR ALTERNATIVES Rationale for Using Autografts Probably the least contentious statement that could be made about nerve repair is that a peripheral nerve autograft provides the optimum environment for regrowing axons. It provides a microenvironment full of the appropriate non-

neuronal cells, of which the majority are acutely denervated, “axon-responsive,” Schwann cells aligned within a scaffold of basal laminae.

Alternatives to Autografts: Conduits and Allografts Bain stated that “[some] major peripheral nerve deficits in otherwise salvageable extremities are unreconstructable with autogenous grafts because of their limited availability.”9 Although interfascicular and group fascicular autografts remain the “gold standard” for repairing traumatic nerve injuries, they offer a less than ideal solution to the problem of bridging an interstump gap, particularly when defects are extensive. Complications associated with their use include limited tissue volume and fascicular and modality mismatches between host and donor nerves, which are all properties of the graft, and secondary donor site morbidity, neuroma formation, and numbness within the distribution of the donor nerve, which are all sequelae of donor nerve harvest. Broadly speaking, alternatives to nerve autografts fall into two groups: either some type of conduit, whether

1424

Neuropathy Caused by Compression, Entrapment, or Physical Injury

biologic, synthetic, or bio-artificial, or a suitably prepared peripheral nerve allograft in combination with temporary immunosuppression. Conduits The biologic conduits most frequently used have been veins or acellular muscle grafts.92 Research using synthetic and bio-artificial conduits is concentrated in two major areas. One concerns the nature of the materials used to form conduits, and seeks to optimize biocompatibility, rates of bioresorption, and diffusion of molecules across the walls of the conduit (e.g., by determining optimal porosity and wall thickness), and to minimize implant adhesion to the wound bed and swelling of materials during degradation (since this could reduce the internal luminal diameter of the conduit). (The reader is referred to Doolabh et al.,53 Heath and Rutkowski,97 and Sundback et al.248 for reviews of the use of conduits in nerve repair.) The other seeks to manipulate the contents of the conduit and so to reproduce, in whole or in part, the microenvironment of a peripheral nerve, for example, by preloading the lumen or walls either with cells seeded at an appropriate density, and/or growth factors, cytokines, or physical substrates, or alternatively, by the continuous application of factors after implantation via osmotic minipumps that access the lumen of the conduit. (An obvious downside of this approach is that local application of factors could induce the overproduction of axon sprouts, leading to painful neuroma formation.) Using Schwann cells and/or fibroblasts as cellular platforms for the local production of recombinant neurotrophins within conduits represents an attractive therapeutic prospect239 that would obviate the need to use minipumps to deliver factors locally into a conduit. The Long Interstump Gap The failure of regrowing axons to cross long gaps (⬎10 cm) within a conduit98 may be related in some way to the nature of the luminal microenvironment.145 To date, most strategies for encouraging axons to grow further have been based on the assumption that axons stop growing because the pool of dedifferentiated Schwann cells available to associate with them becomes exhausted. Experimental protocols to address this concern have focused on supplementing Schwann cell numbers in the lumen or walls of a conduit, using either segments of nerve or a bolus of cultured cells (e.g., between grafts of acellular muscle)29 or longitudinal polyglactin sutures,219 or within a vein autograft.11,70,254 Addition of Schwann cells to conduits has sometimes achieved modest increases, of the order of 2 to 3 cm, in the interstump gap length that regrowing axons will cross. However, long (⬎10-cm) gaps remain an as yet unbridgeable challenge. Schwann cells are often regarded as migratory cells,266 and indeed one of the assumptions made in implanting a bolus of Schwann cells into a conduit is that they will

migrate from the bolus to associate with the regrowing axons. However, in vivo few S100-positive cells grow more than 2 to 3 mm out of the end of a distal nerve stump in the absence of an advancing front of regrowing axons and their associated Schwann cells. Moreover, when an interstump gap is long, so that the distal stump becomes chronically denervated, Schwann cells typically grow across the end of the stump, forming a multilayered “cap,” rather than away from the stump and into the surrounding epineurial connective tissue. This means that, when axons ultimately reach the distal stump, many are deflected away from its lumen because they become associated with the Schwann cells lying across its tip. This behavior has implications for implanting Schwann cells into conduits, and underscores the need to ensure either that conduits are coated with a continuous lining of Schwann cells or that multiple Schwann cell depots are placed less than 2 mm apart within a conduit. A report that Schwann cell invasion of an acellular nerve graft can be stimulated independent of axonal outgrowth by local application of vascular endothelial growth factor, apparently acting via fetal liver kinase-1 receptors expressed on the Schwann cells,238 is therefore welcome, and may be exploitable therapeutically to considerable advantage. Allografts Peripheral nerve allografts are potentially very attractive to both patient and surgeon: they offer the opportunity to use functionally matched nerves, and obviate the need to denervate a second site in the recipient (for review see Evans et al.60). Unfortunately their early clinical use produced disappointing results. Allografts were rejected (they elicit intense inflammation and Schwann cell death), and the technique was therefore abandoned clinically. The principle remained attractive, but its implementation seemed impossible, given an understandable reluctance to use long-term immunosuppression to treat nerve injuries. Over the years, there have been a number of attempts to reduce allograft antigenicity to levels requiring only temporary immunosuppression. Early attempts to pretreat grafts using predegeneration, lyophilization, or freezing, either alone or in combination with radiation, but without concomitant systemic immunosuppression, had limited success.143,144,195 More recently, the introduction of the immunophilin immunosuppressive agents cyclosporin A and FK506,63,177,243 and the use of prolonged storage in physiologic solutions or freeze-thawing in order to render allografts either nonimmunogenic or of reduced immunogenicity,59 have produced much more favorable results. Interestingly, FK506,63 but not cyclosporin A,41 can rescue peripheral nerve allografts in acute rejection. Individually and in combination,83 these technical advances give cause for optimism that allografts may provide an acceptable alternative to autografts in the future. The technique has recently returned to the clinic with encouraging results.9,142

Mechanisms of Repair after Traumatic Injury

Termino-lateral Neurorrhaphy The termino-lateral neurorrhaphy technique seeks to reduce reinnervation distance, and hence time to target reinnervation, and therefore may represent a potential advance in the currently unsatisfactory management of long interstump gaps. Like entubulation, it is an old surgical protocol that fell into disuse until a recent revival of experimental interest, and has now returned to the clinical sphere.166,181,207,269,281 Although experimental studies have reported motor and sensory functional recovery, with variable effects on the donor nerve, “clinical outcomes are yet to be determined.”208 The underlying mechanism is not clear, but presumably intact axons in the donor nerves sprout in response to factors and/or cells derived from the coapted distal nerve stump, and the sprouts access the Schwann cell tubes in the latter via a damaged perineurium.

4.

5.

6. 7. 8.

9.

THE FUTURE Improvements in the clinical management of injured nerves will depend upon manipulating the injury response, rather than fine-tuning microsurgical techniques. This will require understanding the biology of repair rather better than is currently the case, in order to design protocols that will maximize axonal growth across gaps of many centimeters within a tissue-engineered conduit; control the regulation of interactions between axons, non-neural cells, and the extracellular matrix; minimize neuronal cell death and the occurrence of neuropathic pain; maintain non-neural cells in an acute axon-responsive state along the pathway (whether that be graft, conduit, or distal stump); maintain target viability; and optimize target selectivity.

ACKNOWLEDGMENTS This review was written while the author was in receipt of European Framework V grant QLK3-CT-1999-00625. I am grateful to Dr. Stephen Bunting for helpful discussions and to Dr. Thomas Brushart for very kindly allowing me sight of a submitted manuscript.27

REFERENCES 1. Abernethy, D. A., Rud, A., and Thomas, P. K.: Neurotropic influence of the distal stump of transected peripheral nerve on axonal regeneration: absence of topographic specificity in adult nerve. J. Anat. 180(Pt. 3):395, 1992. 2. Abernethy, D. A., Thomas, P. K., Rud, A., and King, R. H.: Mutual attraction between emigrant cells from transected denervated nerve. J. Anat. 184(Pt. 2):239, 1994. 3. Al-Majed, A. A., Brushart, T. M., and Gordon, T.: Electrical stimulation accelerates and increases expression of BDNF

10.

11.

12.

13.

14.

15.

16.

17.

18. 19.

20.

21.

1425

and trkB mRNA in regenerating rat femoral motoneurons. Eur. J. Neurosci. 12:4381, 2000. Al-Majed, A. A., Neumann, C. M., Brushart, T. M., and Gordon, T.: Brief electrical stimulation promotes the speed and accuracy of motor axonal regeneration. J. Neurosci. 20:2602, 2000. Araki, T., and Milbrandt, J.: Ninjurin2, a novel homophilic adhesion molecule, is expressed in mature sensory and enteric neurons and promotes neurite outgrowth. J. Neurosci. 20:187, 2000. Arroyo, E. J., and Scherer, S. S.: On the molecular architecture of myelinated fibers. Histochem. Cell. Biol. 113:1, 2000. Asbury, A. K.: Renaut bodies: a forgotten endoneurial structure. J. Neuropathol. Exp. Neurol. 32:334, 1973. Baichwal, R. R., Bigbee, J. W., and DeVries, G. H.: Macrophage-mediated myelin-related mitogenic factor for cultured Schwann cells. Proc. Natl. Acad. Sci. U. S. A. 85:1701, 1988. Bain, J. R.: Peripheral nerve and neuromuscular allotransplantation: current status. Microsurgery 20:384, 2000. Bannerjee, A., and Patterson, P. H.: Schwann cell CD9 expression is regulated by axons. Mol. Cell. Neurosci. 6:462, 1995. Battiston, B., Tos, P., Geuna, S., et al.: Nerve repair by means of vein filled with muscle grafts. II. Morphological analysis of regeneration. Microsurgery 20:37, 2000. Baulieu, E. E., Schumacher, M., Koenig, H., et al.: Progesterone as a neurosteroid: actions within the nervous system. Cell. Mol. Neurobiol. 16:143, 1996. Bedi, K. S., Winter, J., Berry, M., and Cohen, J.: Adult rat dorsal root ganglion neurons extend neurites on predegenerated but not on normal peripheral nerves in vitro. Eur. J. Neurosci. 4:193, 1992. Be’eri, H., Reichert, F., Saada, A., and Rotschenker, S.: The cytokine network of wallerian degeneration: IL-10 and GM-CSF. Eur. J. Neurosci. 10:2707, 1998. Bennett, G. J., and Xie, Y. K.: A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain 33:87, 1988. Birch, R., Bonney, G., and Wynn Parry, C. B.: Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingstone, 1998. Bodine-Fowler, S. C., Meyer, R. S., Moskovitz, A., et al.: Inaccurate projection of rat soleus motorneurons: a comparison of nerve repair techniques. Muscle Nerve 20:29, 1997. Bolin, L. M., and Shooter, E. M.: Neurons regulate Schwann cell genes by diffusible molecules. J. Cell Biol. 123:237, 1993. Boyd, J. G., and Gordon, T.: A dose-dependent facilitation and inhibition of peripheral nerve regeneration by brainderived neurotrophic factor. Eur. J. Neurosci. 15:613, 2002. Bradley, J. L., Abernethy, D. A., King, R. H., et al.: Neural architecture in transected rabbit sciatic nerve after prolonged nonreinnervation. J. Anat. 192(Pt. 4):529, 1998. Braunewell, K. H., Martini, R., LeBaron, R., et al.: Upregulation of a chondroitin sulphate epitope during regeneration of mouse sciatic nerve: evidence that the immunoreactive molecules are related to the chondroitin sulphate proteoglycans decorin and versican. Eur. J. Neurosci. 7:792, 1995.

1426

Neuropathy Caused by Compression, Entrapment, or Physical Injury

22. Braunewell, K. H., Pesheva, P., McCarthy, J. B., et al.: Functional involvement of sciatic nerve-derived versicanand decorin-like molecules and other chondroitin sulphate proteoglycans in ECM-mediated cell adhesion and neurite outgrowth. Eur. J. Neurosci. 7:805, 1995. 23. Brown, M. C., and Hardman, V. J.: A reassessment of the accuracy of reinnervation by motorneurons following crushing or freezing of the sciatic or lumbar spinal nerves of rats. Brain 110:695, 1987. 24. Brown, M. C., Perry, V. H., Hunt, S. P., and Laper, S. R.: Further studies on motor and sensory nerve regeneration in mice with delayed wallerian degeneration. Eur. J. Neurosci. 6:420, 1994. 25. Brushart, T. M.: Preferential reinnervation of motor nerves by regenerating motor axons. J. Neurosci. 8:1026, 1988. 26. Brushart, T. M., Gerber, J., Kessens, P., et al.: Contributions of pathway and neuron to preferential motor reinnervation. J. Neurosci. 18:8674, 1998. 27. Brushart, T. M., Hoffman, P. N., Royall, R. M., et al.: Electrical stimulation promotes motoneuron regeneration without increasing its speed or conditioning the neuron. J. Neurosci. 22:6631, 2002. 28. Brushart, T. M., Mathur, V., Sood R., and Koschorke, G. M.: Joseph H. Boyes Award. Dispersion of regenerating axons across enclosed neural gaps. J. Hand Surg. Am. 20:557, 1995. 29. Calder, J. S., and Green, C. J.: Nerve-muscle sandwich grafts: the importance of Schwann cells in peripheral nerve regeneration through muscle basal lamina conduits. J. Hand Surg. Br. 20:423, 1995. 30. Calder, J. S., and Norris, R. W.: Repair of mixed peripheral nerves using muscle autografts: a preliminary communication. Br. J. Plast. Surg. 46:557, 1993. 31. Camborieux, L., Bertrand, N., and Swerts, J. P.: Changes in expression and localization of hemopexin and its transcripts in injured nervous system: a comparison of central and peripheral tissues. Neuroscience 82:1039, 1998. 32. Campana, W. M., and Myers, R. R.: Erythropoietin and erythropoietin receptors in the peripheral nervous system: changes after nerve injury. FASEB J. 15:1804, 2001. 33. Campbell, J. N.: Nerve lesions and the generation of pain. Muscle Nerve 24:1261, 2001. 33a. Campbell, D. S., and Holt, C. E.: Chemotropic responses of retinal growth cones mediated by rapid local protein synthesis and degradation. Neuron 32:1013, 2001. 34. Carroll, S. L., and Frohnert, P. W.: Expression of JE (monocyte chemoattractant protein-1) is induced by sciatic axotomy in wild type rodents but not in C57BL/Wld(s) mice. J. Neuropathol. Exp. Neurol. 57:915, 1998. 35. Castejon, M. S., Culver, D. G., and Glass, J. D.: Generation of spectrin breakdown products in peripheral nerves by addition of M-calpain. Muscle Nerve 22:905, 1999. 36. Castren, E., Zafra, F., Thoenen, H., and Lindholm, D.: Light regulates expression of brain-derived neurotrophic factor mRNA in rat visual cortex. Proc. Natl. Acad. Sci. U. S. A. 89:9444, 1992. 37. Cavanagh, J. B.: The significance of the “dying back” process in experimental and human neurological disease. Int. Rev. Exp. Pathol. 3:219, 1964.

38. Ceballos, D., Navarro, X., Dubey, N., et al.: Magnetically aligned collagen gel filling a collagen nerve guide improves peripheral nerve regeneration. Exp. Neurol. 158:290, 1999. 39. Chandross, K.: Nerve injury and inflammatory cytokines modulate gap junctions in the peripheral nervous system. Glia 24:21, 1998. 40. Chandross, K. J., Kessler, J. A., Cohen R. I., et al.: Altered connexin expression after peripheral nerve injury. Mol. Cell. Neurosci. 7:501, 1996. 41. Chen, D. L., Mackinnon, S. E., Jensen, J. N., et al.: Failure of cyclosporin A to rescue peripheral nerve allografts in acute rejection. Ann. Plast. Surg. 49:660, 2002. 42. Chen, R., Cohen, L. G., and Hallett, M.: Nervous system reorganization following injury. Neuroscience 111:761, 2002. 43. Chernousov, M. A., Scherer, S. S., Stahl, R. C., and Carey, D. J.: p200, a collagen secreted by Schwann cells, is expressed in developing nerves and in adult nerves following axotomy. J. Neurosci. Res. 56:284, 1999. 44. Chiu, D. T., and Strauch, B.: A prospective clinical evaluation of autogenous vein grafts used as a nerve conduit for distal sensory nerve defects of 3 cm or less. Plast. Reconstr. Surg. 86:928, 1990. 45. Cohen, J. A., Yachnis, A. T., Arai, M., et al.: Expression of the neu proto-oncogene by Schwann cells during peripheral nerve development and wallerian degeneration. J. Neurosci. Res. 31:622, 1992. 46. Constantinescu, C. S., Goodman, D. B., Hilliard, B., et al.: Murine macrophages stimulated with central and peripheral nervous system myelin or purified myelin proteins release inflammatory products. Neurosci. Lett. 287:171, 2000. 47. Curtis, R., Stewart, H. J., Hall, S. M., et al.: GAP-43 is expressed by nonmyelin-forming Schwann cells of the peripheral nervous system. J. Cell Biol. 116:1455, 1992. 48. da Costa, C. C., van der Laan, L. J., Dijkstra, C. D., and Bruck, W.: The role of the mouse macrophage scavenger receptor in myelin phagocytosis. Eur. J. Neurosci. 9:2650, 1997. 49. Dailey, A. T., Avellino, A. M., Benthem, L., et al.: Complement depletion reduces macrophage infiltration and activation during wallerian degeneration and axonal regeneration. J. Neurosci. 18:6713, 1998. 50. Davison, S. P., McCaffrey, T. V., Porter, M. N., and Manders, E.: Improved nerve regeneration with neutralization of transforming growth factor-beta1. Laryngoscope 109:631, 1999. 51. Desarnaud, F., Do Thi, A. N., Brown, A. M., et al.: Progesterone stimulates the activity of the promoters of peripheral myelin protein-22 and protein zero genes in Schwann cells. J. Neurochem. 71:1765, 1998. 51a. Dilley, A., Lynn, B., Greening, J., and DeLeon, N.: Quantitative in vivo studies of median nerve sliding in response to wrist, elbow, shoulder and neck movements. Clin. Biomech. 18:899, 2003. 52. Dong, Z., Brennan, A., Liu, N., et al.: Neu differentiation factor is a neuron-glia signal and regulates survival, proliferation, and maturation of rat Schwann cell precursors. Neuron 15:585, 1995. 53. Doolabh, V. B., Hertl, M. C., and Mackinnon, S. E.: The role of conduits in nerve repair: a review. Rev. Neurosci. 7:47, 1996.

Mechanisms of Repair after Traumatic Injury 54. Eddleman, C. S., Ballinger, M. L., Smyers, M. E., et al.: Endocytotic formation of vesicles and other membranous structures induced by Ca2⫹ and axolemmal injury. J. Neurosci. 18:4029, 1998. 55. Eddleman, C. S., Bittner, G. D., and Fishman, H. M.: Barrier permeability at cut axonal ends progressively decreases until an ionic seal is formed. Biophys. J. 79:1883, 2000. 55a. Ehlers, M. D.: Deconstructing the axon: Wallerian degeneration and the ubiquitin-proteasome system. Trends Neurosci. 27:3, 2004. 56. Einheber, S., Hannocks, M. J., Metz, C. N., et al.: Transforming growth factor-beta 1 regulates axon/Schwann cell interactions. J. Cell Biol. 129:443, 1995. 57. Enver, M. K., and Hall, S. M.: Are Schwann cells essential for axonal regeneration into muscle autografts? Neuropathol. Appl. Neurobiol. 20:587, 1994. 58. Eto, M., Yoshikawa, H., Fujimura, H., et al.: The role of CD36 in peripheral nerve remyelination after crush injury. Eur. J. Neurosci. 17:2659, 2003. 59. Evans, P. J., Mackinnon, S. E., Levi, A. D., et al.: Cold preserved nerve allografts: changes in basement membrane, viability, immunogenicity, and regeneration. Muscle Nerve 21:1507, 1998. 60. Evans, P. J., Midha, R., and Mackinnon, S. E.: The peripheral nerve allograft: a comprehensive review of regeneration and neuroimmunology. Prog. Neurobiol. 43:187, 1994. 61. Fabrizi, C., Kelly, B. M., Gillespie, C. S., et al.: Transient expression of the neurofilament proteins NF-L and NF-M by Schwann cells is regulated by axonal contact. J. Neurosci. Res. 50:291, 1997. 62. Feltri, M. L., Scherer, S. S., Nemni, R., et al.: Beta 4 integrin expression in myelinating Schwann cells is polarized, developmentally regulated and axonally dependent. Development 120:1287, 1994. 63. Feng, F. Y., Ogden, M. A., Myckatyn, T. M., et al.: FK506 rescues peripheral nerve allografts in acute rejection. J. Neurotrauma. 18:217, 2001. 64. Ferguson, T. A., and Muir, D.: MMP-2 and MMP-9 increase the neurite-promoting potential of Schwann cell basal laminae and are upregulated in degenerated nerve. Mol. Cell. Neurosci. 16:157, 2000. 65. Fernandez-Valle, C., Bunge, R. P., and Bunge, M. B.: Schwann cells degrade myelin and proliferate in the absence of macrophages: evidence from in vitro studies of wallerian degeneration. J. Neurocytol. 24:667, 1995. 66. Ferri, C. C., and Bisby, M. A.: Improved survival of injured sciatic nerve Schwann cells in mice lacking the p75 receptor. Neurosci. Lett. 272:191, 1999. 67. Ferzaz, B., Brault, E., Bourliaud, G., et al.: SSR180575 (7-chloro-N,N,5-trimethyl-4-oxo-3-phenyl-3,5-dihydro-4Hpyridazino[4,5-b]indole-1-acetamide), a peripheral benzodiazepine receptor ligand, promotes neuronal survival and repair. J. Pharmacol. Exp. Ther. 301:1067, 2002. 68. Fidler, P. S., Schuette, K., Asher, R. A., et al.: Comparing astrocytic cell lines that are inhibitory or permissive for axon growth: the major axon-inhibitory proteoglycan is NG2. J. Neurosci. 19:8778, 1999. 69. Finn, J. T., Weil, M., Archer, F., et al.: Evidence that wallerian degeneration and localized axon degeneration

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

1427

induced by local neurotrophin deprivation do not involve caspases. J. Neurosci. 20:1333, 2000. Foidart-Dessalle, M., Dubuisson, A., Lejeune, A., et al.: Sciatic nerve regeneration through venous or nervous grafts in the rat. Exp. Neurol. 148:236, 1997. Fu, S. Y., and Gordon, T.: Contributing factors to poor functional recovery after delayed nerve repair: prolonged axotomy. J. Neurosci. 15:3876, 1995. Fullarton, A. C., Glasby, M. A., and Lawson, G. M.: Immediate and delayed nerve repair using freeze-thawed muscle allografts: associated long-bone fracture. J. Hand Surg. Br. 23:360, 1998. Gamble, H. J., and Jha, B. D.: Some effects of temperature upon the rate and progress of wallerian degeneration in mammalian nerve fibres. J. Anat. 92:171, 1958. Genden, E. M., Watanabe, O., Mackinnon, S. E., et al.: Peripheral nerve regeneration in the apolipoprotein-Edeficient mouse. J. Reconstr. Microsurg. 18:495, 2002. George, E. B., Glass, J. D., and Griffin, J. W.: Axotomyinduced axonal degeneration is mediated by calcium influx through ion-specific channels. J. Neurosci. 15:6445, 1995. Ghosh, A., Carnahan, J., and Greenberg, M. E.: Requirement for BDNF in activity-dependent survival of cortical neurons. Science 263:1618, 1994. Glasby, M. A., Fullerton, A. C., and Lawson, G. M.: Immediate and delayed nerve repair using freeze-thawed muscle autografts in complex nerve injuries: associated arterial injury. J. Hand Surg. Br. 23:354, 1998. Glass, J. D., Culver, D. G., Levey, A. I., and Nash, N. R.: Very early activation of m-calpain in peripheral nerve during wallerian degeneration. J. Neurol. Sci. 196:9, 2002. Glass, J. D., Schryer, B. L., and Griffin, J. W.: Calciummediated degeneration of the axonal cytoskeleton in the Ola mouse. J. Neurochem. 62:2472, 1994. Gold, B. G., Densmore, V., Shou, W., et al.: Immunophilin FK506-binding protein 52 (not FK506-binding protein 12) mediates the neurotrophic action of FK506. J. Pharmacol. Exp. Ther. 289:1202, 1999. Goodrum, J. F., Fowler, K. A., Hostettler, J. D., and Toews, A. D.: Peripheral nerve regeneration and cholesterol reutilization are normal in the low-density lipoprotein receptor knockout mouse. J. Neurosci. Res. 59:581, 2000. Gramsbergen, A., IJkema-Paassen, J., and Meek, M. F.: Sciatic nerve transection in the adult rat: abnormal EMG patterns during locomotion by aberrant innervation of hindleg muscles. Exp. Neurol. 161:183, 2000. Grand, A. G., Myckatyn, T. M., Mackinnon, S. E., and Hunter, D. A.: Axonal regeneration after cold preservation of nerve allografts and immunosuppression with tacrolimus in mice. J. Neurosurg. 96:924, 2002. Griffin, J. W., George, E. B., and Chaudhry, V.: Wallerian degeneration in peripheral nerve disease. Baillieres Clin. Neurol. 5:65, 1996. Grinspan, J. B., Marchionni, M. A., Reeves, M., et al.: Axonal interactions regulate Schwann cell apoptosis in developing peripheral nerve: neuregulin receptors and the role of neuregulins. J. Neurosci. 16:6107, 1996. Guennoun, R., Benmessahel, Y., Delespierre, B., et al.: Progesterone stimulates Krox-20 gene expression in Schwann cells. Brain Res. Mol. Brain Res. 90:75, 2001.

1428

Neuropathy Caused by Compression, Entrapment, or Physical Injury

87. Gupta, S. K., Poduslo, J. F., and Mezei, C.: Temporal changes in P0 and MBP gene expression after crush-injury of the adult peripheral nerve. Brain Res. 464:133, 1988. 88. Gupta, S. K., Pringle, J., Poduslo, J. F., and Mezei, C.: Induction of myelin genes during peripheral nerve remyelination requires a continuous signal from the ingrowing axon. J. Neurosci. Res. 34:14, 1993. 89. Hadlock, T., Elisseeff, J., Langer, R., et al.: A tissueengineered conduit for peripheral nerve repair. Arch. Otolaryngol. Head Neck Surg. 124:1081, 1998. 90. Hadlock, T., Sundback, C., Hunter, D., et al.: A polymer foam conduit seeded with Schwann cells promotes guided peripheral nerve regeneration. Tissue Eng. 6:119, 2000. 91. Hadlock, T. A., Sundback, C. A., Hunter, D. A., et al.: A new artificial nerve graft containing rolled Schwann cell monolayers. Microsurgery 21:96, 2001. 92. Hall, S.: Axonal regeneration through acellular muscle grafts. J. Anat. 190 (Pt. 1):57, 1997. 93. Hall, S. M.: Observations on the progress of wallerian degeneration in transected peripheral nerves of C57BL/Wld mice in the presence of recruited macrophages. J. Neurocytol. 22:480, 1993. 94. Hall, S. M.: The biology of chronically denervated Schwann cells. Ann. N. Y. Acad. Sci. 883:215, 1999. 95. Hall, S. M., and Gregson, N. A.: The effects of mitomycin C on the process of regeneration in the mammalian peripheral nervous system. Neuropathol. Appl. Neurobiol. 3:65, 1977. 96. Hall, S. M., Li, H., and Kent, A. P.: Schwann cells responding to primary demyelination in vivo express p75NTR and c-erbB receptors: a light and electron immunohistochemical study. J. Neurocytol. 26:679, 1997. 97. Heath, C. A., and Rutkowski, G. E.: The development of bioartificial nerve grafts for peripheral-nerve regeneration. Trends Biotechnol. 16:163, 1998. 98. Hems, T. E., and Glasby, M. A.: The limit of graft length in the experimental use of muscle grafts for nerve repair. J. Hand Surg. Br. 18:165, 1993. 99. Heumann, R., Korsching, S., Bandtlow, C., and Thoenen, H.: Changes of nerve growth factor synthesis in nonneuronal cells in response to sciatic nerve transection. J. Cell Biol. 104:1623, 1987. 100. Hill, S., and Hall, S.: Microscopic anatomy of the posterior interosseous and median nerves at sites of potential entrapment in the forearm. J. Hand Surg. Br. 24:170, 1999. 101. Hiraiwa, M., Campana, W. M., Mizisin, A. P., et al.: Prosaposin: a myelinotrophic protein that promotes expression of myelin constituents and is secreted after nerve injury. Glia 26:353, 1999. 102. Hirata, K., and Kawabuchi, M.: Myelin phagocytosis by macrophages and nonmacrophages during wallerian degeneration. Microsc. Res. Tech. 57:541, 2002. 103. Hoke, A., Sun, H. S., Gordon, T., and Zochodne, D. W.: Do denervated peripheral nerve trunks become ischemic? The impact of chronic denervation on vasa nervorum. Exp. Neurol. 172:398, 2001. 104. Horie, H., Inagaki, Y., Sohma, Y., et al.: Galectin-1 regulates initial axonal growth in peripheral nerves after axotomy. J. Neurosci. 19:9964, 1999. 105. Hughes, P. M., Wells, G. M., Perry, V. H., et al.: Comparison of matrix metalloproteinase expression during wallerian

106.

107. 108.

109.

110.

111.

112.

113. 114.

115.

116. 117.

118.

119.

120.

121.

122.

degeneration in the central and peripheral nervous systems. Neuroscience 113:273, 2002. Hughes, R. A., Cameron, J. S., Hall, S. M., et al.: Multiple mononeuropathy as the initial presentation of systemic lupus erythematosus—nerve biopsy and response to plasma exchange. J. Neurol. 228:239, 1982. Hughes, R. C.: Galectins as modulators of cell adhesion. Biochimie 83:667, 2001. Inserra, M. M., Yao, M., Murray, R., and Terris, D. J.: Peripheral nerve regeneration in interleukin 6-deficient mice. Arch. Otolaryngol. Head Neck Surg. 126:1112, 2000. Jefferson, D., Neary, D., and Eames, R. A.: Renaut body distribution at sites of human peripheral nerve entrapment. J. Neurol. Sci. 49:19, 1981. Jessen, K. R., Brennan, A., Morgan, L., et al.: The Schwann cell precursor and its fate: a study of cell death and differentiation during gliogenesis in rat embryonic nerves. Neuron 12:509, 1994. Jessen, K. R., Mirsky, R., and Morgan, L.: Myelinated, but not unmyelinated axons, reversibly down-regulate N-CAM in Schwann cells. J. Neurocytol. 16:681, 1987. Jessen, K. R., Morgan, L., Stewart, H. J., and Mirsky, R.: Three markers of adult non-myelin-forming Schwann cells, 217c(Ran-1), A5E3 and GFAP: development and regulation by neuron-Schwann cell interactions. Development 109:91, 1990. Jia, J., and Pollock, M.: Cold nerve injury is enhanced by intermittent cooling. Muscle Nerve 22:1644, 1999. Kidd, G., Andrews, S. B., and Trapp, B. D.: Axons regulate the distribution of Schwann cell microtubules. J. Neurosci. 16:946, 1996. Kim, S. H., and Chung, J. M.: An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 50:355, 1992. Korsching, S.: The neurotrophic factor concept: a reexamination. J. Neurosci. 13:2739, 1993. Krekoski, C. A., Neubauer, D., Zuo, J., and Muir, D.: Axonal regeneration into acellular nerve grafts is enhanced by degradation of chondroitin sulfate proteoglycan. J. Neurosci. 21:6206, 2001. Kuecherer-Ehret, A., Graeber, M. B., Edgar, D., et al.: Immunoelectron microscopic localization of laminin in normal and regenerating mouse sciatic nerve. J. Neurocytol. 19:101, 1990. Kuhlmann, T., Bitsch, A., Stadelmann, C., et al.: Macrophages are eliminated from the injured peripheral nerve via local apoptosis and circulation to regional lymph nodes and the spleen. J. Neurosci. 21:3401, 2001. Kurek, J. B., Austin, L., Cheema, S. S., et al.: Up-regulation of leukaemia inhibitory factor and interleukin-6 in transected sciatic nerve and muscle following denervation. Neuromuscul. Disord. 6:105, 1996. La Fleur, M., Underwood, J. L., Rappolee, D. A., and Werb, Z.: Basement membrane and repair of injury to peripheral nerve: defining a potential role for macrophages, matrix metalloproteinases, and tissue inhibitor of metalloproteinases-1. J. Exp. Med. 184:2311, 1996. Lefcort, F., Venstrom, K., McDonald, J. A., and Reichardt, L. F.: Regulation of expression of fibronectin and its

Mechanisms of Repair after Traumatic Injury

123.

124.

125.

126.

127.

128.

129.

130.

131.

132. 133. 134.

135. 136.

137.

138.

139.

140.

receptor, alpha 5 beta 1, during development and regeneration of peripheral nerve. Development 116:767, 1992. Lemke, G., and Chao, M.: Axons regulate Schwann cell expression of the major myelin and NGF receptor genes. Development 102:499, 1988. Lenihan, D. V., Carter, A. J., Gilchrist, T., et al.: Biodegradable controlled release glass in the repair of peripheral nerve injuries. J. Hand Surg. Br. 23:588, 1998. Levy, D., Hoke, A., and Zochodne, D. W.: Local expression of inducible nitric oxide synthase in an animal model of neuropathic pain. Neurosci. Lett. 260:207, 1999. Levy, D., Kubes, P., and Zochodne, D. W.: Delayed peripheral nerve degeneration, regeneration, and pain in mice lacking inducible nitric oxide synthase. J. Neuropathol. Exp. Neurol. 60:411, 2001. Levy, D., Tal, M., Hoke, A., and Zochodne, D. W.: Transient action of the endothelial constitutive nitric oxide synthase (ecNOS) mediates the development of thermal hypersensitivity following peripheral nerve injury. Eur. J. Neurosci. 12:2323, 2000. Li, H., Terenghi, G. and Hall, S. M.: Effects of delayed re-innervation on the expression of c-erbB receptors by chronically denervated rat Schwann cells in vivo. Glia 20:333, 1997. Li, H., Wigley, C., and Hall, S. M.: Chronically denervated rat Schwann cells respond to GGF in vitro. Glia 24:290, 1998. Liefner, M., Siebert, H., Sachse, T., et al.: The role of TNFalpha during wallerian degeneration. J. Neuroimmunol. 108:147, 2000. Ljungberg, C., Johansson-Ruden, G., Bostrom, K. J., et al.: Neuronal survival using a resorbable synthetic conduit as an alternative to primary nerve repair. Microsurgery 19:259, 1999. Lundborg, G.: Intraneural microcirculation. Orthop. Clin. North Am. 19:1, 1988. Lundborg, G.: Brain plasticity and hand surgery: an overview. J. Hand Surg. Br. 25:242, 2000. Lundborg, G.: A 25-year perspective of peripheral nerve surgery: evolving neuroscientific concepts and clinical significance. J. Hand Surg. Am. 25:391, 2000. Lundborg, G.: Nerve injury and repair—a challenge to the plastic brain. J. Peripher. Nerv. Syst. 8:209, 2003. Lunn, E. R., Brown, M. C., and Perry, V. H.: The pattern of axonal degeneration in the peripheral nervous system varies with different types of lesion. Neuroscience 35:157, 1990. Lunn, E. R., Perry, V. H., Brown, M. C., et al.: Absence of wallerian degeneration does not hinder regeneration in peripheral nerve. Eur. J. Neurosci. 1:27, 1989. Luttges, M. W., Kelly, P. T., and Gerren, R. A.: Degenerative changes in mouse sciatic nerves: electrophoretic and electrophysiologic characterizations. Exp, Neurol. 50:706, 1976. Lutz, B. S., Ma, S. F., Chuang, D. C., and Wei, F. C.: Effects of systemically applied IGF-1 on motor nerve recovery after peripheral nerve transection and repair in the rat—a functional study. Hand Surg. 4:131, 1999. Lyon, M. F., Ogunkolade, B. W., Brown, M. C., et al.: A gene affecting wallerian nerve degeneration maps distally on mouse chromosome 4. Proc. Natl. Acad. Sci. U. S. A. 90:9717, 1993.

1429

141. Mack, T. G., Reiner, M., Beirowski, B., et al.: Wallerian degeneration of injured axons and synapses is delayed by a Ube4b/Nmnat chimeric gene. Nat. Neurosci. 4:1199, 2001. 142. Mackinnon, S. E.: Nerve allotransplantation following severe tibial nerve injury: case report. J. Neurosurg. 84:671, 1996. 143. Mackinnon, S. E., Hudson, A. R., Falk, R. E., et al.: Peripheral nerve allograft: an assessment of regeneration across pretreated nerve allografts. Neurosurgery 15:690, 1984. 144. Mackinnon, S. E., Hudson, A. R., Falk, R. E., et al.: Peripheral nerve allograft: an immunological assessment of pretreatment methods. Neurosurgery 14:167, 1984. 145. Mackinnon, S. E., Hudson, A. R., and Hunter, D. A.: Histologic assessment of nerve regeneration in the rat. Plast. Reconstr. Surg. 75:384, 1985. 146. Madison, R. D., Archibald, S. J., and Brushart, T. M.: Reinnervation accuracy of the rat femoral nerve by motor and sensory neurons. J. Neurosci. 16:5698, 1996. 147. Madison, R. D., Zomorodi, A., and Robinson, G. A.: Netrin-1 and peripheral nerve regeneration in the adult rat. Exp. Neurol. 161:563, 2000. 148. Maeda, T., Hori, S., Sasaki, S., and Maruo, S.: Effects of tension at the site of coaptation on recovery of sciatic nerve function after neurorrhaphy: evaluation by walking-track measurement, electrophysiology, histomorphometry, and electron probe X-ray microanalysis. Microsurgery 19:200, 1999. 149. Marcinkiewicz, M., Savaria, D., and Marcinkiewicz, J.: The pro-protein convertase PC1 is induced in the transected sciatic nerve and is present in cultured Schwann cells: comparison with PC5, furin and PC7, implication in pro-BDNF processing. Brain Res. Mol. Brain Res. 59:229, 1998. 150. Martini, R., and Schachner, M.: Complex expression pattern of tenascin during innervation of the posterior limb buds of the developing chicken. J, Neurosci, Res. 28:261, 1991. 151. Martini, R., and Schachner, M.: Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and myelin-associated glycoprotein) in regenerating adult mouse sciatic nerve. J. Cell Biol. 106:1735, 1988. 152. Martini, R., Schachner, M., and Brushart, T. M.: The L2/HNK-1 carbohydrate is preferentially expressed by previously motor axon-associated Schwann cells in reinnervated peripheral nerves. J. Neurosci. 14:7180, 1994. 153. Martini, R., Xin, Y., Schmitz, B., and Schachner, M.: The L2/HNK-1 carbohydrate epitope is involved in the preferential outgrowth of motor neurons on ventral roots and motor nerves. Eur. J. Neurosci. 4:628, 1992. 154. Matsuoka, I., Meyer, M., and Thoenen, H.: Cell-typespecific regulation of nerve growth factor (NGF) synthesis in non-neuronal cells: comparison of Schwann cells with other cell types. J. Neurosci. 11:3165, 1991. 155. Matsuoka, I., Nakane, A., and Kurihara, K.: Induction of LIF-mRNA by TGF-beta 1 in Schwann cells. Brain Res. 776:170, 1997. 156. Maurel, P., and Salzer, J. L.: Axonal regulation of Schwann cell proliferation and survival and the initial events of myelination require PI 3-kinase activity. J. Neurosci. 20:4635, 2000.

1430

Neuropathy Caused by Compression, Entrapment, or Physical Injury

157. McCawley, L. J., and Matrisian, L. M.: Matrix metalloproteinases: they’re not just for matrix anymore! Curr. Opin. Cell Biol. 13:534, 2001. 158. McDonald, W. I.: The time course of conduction failure during degeneration of a central tract. Exp. Brain Res. 14:550, 1972. 159. McKeon, R. J., Jurynec, M. J., and Buck, C. R.: The chondroitin sulfate proteoglycans neurocan and phosphacan are expressed by reactive astrocytes in the chronic CNS glial scar. J. Neurosci. 19:10778, 1999. 160. McLean, J., Batt, J., Doering, L. C., et al.: Enhanced rate of nerve regeneration and directional errors after sciatic nerve injury in receptor protein tyrosine phosphatase sigma knock-out mice. J. Neurosci. 22:5481, 2002. 161. Mears, S., Schachner, M., and Brushart, T. M.: Antibodies to myelin-associated glycoprotein accelerate preferential motor reinnervation. J. Peripher. Nerv. Syst. 8:91, 2003. 162. Meek, M. F., and Coert, J. H.: Clinical use of nerve conduits in peripheral-nerve repair: review of the literature. J. Reconstr. Microsurg. 18:97, 2002. 163. Meek, M. F., Den Dunnen, W. F., Schakenraad, J. M., and Robinson, P. H.: Long-term evaluation of functional nerve recovery after reconstruction with a thin-walled biodegradable poly (DL-lactide-epsilon-caprolactone) nerve guide, using walking track analysis and electrostimulation tests. Microsurgery 19:247, 1999. 164. Menge, T., Jander, S., and Stoll, G.: Induction of the proinflammatory cytokine interleukin-18 by axonal injury. J. Neurosci. Res. 65:332, 2001. 165. Menichella, D. M., Barron, P. L., Scarpini, E., et al.: Expression of E-cadherin and beta-catenin in mouse sciatic nerve during development, wallerian degeneration and experimental models of hereditary demyelinating neuropathies. Neurology 247 (Suppl. 3):III/182, 2000. 166. Mennen, U.: End-to-side nerve suture—a technique to repair peripheral nerve injury. S. Afr. Med. J. 89:1188, 1999. 167. Merzenich, M. M., Kaas, J. H., Wall, J., et al.: Topographic reorganization of somatosensory cortical areas 3b and 1 in adult monkeys following restricted deafferentation. Neuroscience 8:33, 1983. 168. Merzenich, M. M., Kaas, J. H., Wall, J. T., et al.: Progression of change following median nerve section in the cortical representation of the hand in areas 3b and 1 in adult owl and squirrel monkeys. Neuroscience 10:639, 1983. 169. Meyer, M., Matsuoka, I., Wetmore, C., et al.: Enhanced synthesis of brain-derived neurotrophic factor in the lesioned peripheral nerve: different mechanisms are responsible for the regulation of BDNF and NGF mRNA. J. Cell Biol. 119:45, 1992. 170. Mikol, D. D., Scherer, S. S., Duckett, S. J., et al.: Schwann cell caveolin-1 expression increases during myelination and decreases after axotomy. Glia 38:191, 2002. 171. Mitchell, L. S., Griffiths, I. R., Morrison, S., et al.: Expression of myelin protein gene transcripts by Schwann cells of regenerating nerve. J. Neurosci. Res. 27:125, 1990. 172. Mohammad, J., Shenaq, J., Rabinovsky, E., and Shenaq, S.: Modulation of peripheral nerve regeneration: a tissueengineering approach. The role of amnion tube nerve conduit across a 1-centimeter nerve gap. Plast. Reconstr. Surg. 105:660, 2000.

173. Morris, J. H., Hudson, A. R., and Weddell, G.: A study of degeneration and regeneration in the divided rat sciatic nerve based on electron microscopy. IV. Changes in fascicular microtopography, perineurium and endoneurial fibroblasts. Z. Zellforsch. Mikrosk. Anat. 124:165, 1972. 174. Mueller, M., Wacker, K., Ringelstein, E. B., et al.: Rapid response of identified resident endoneurial macrophages to nerve injury. Am. J. Pathol. 159:2187, 2001. 175. Muir, D., and Manthorpe, M.: Stromelysin generates a fibronectin fragment that inhibits Schwann cell proliferation. J. Cell Biol. 116:177, 1992. 176. Murinson, B. B., Li, Y., Belzber, A. J., and Griffin, J. W.: The response of unmyelinated nerve fibers and Remak bundle Schwann cells in the rat hindpaw after L5 mixed spinal root transection. Neurology 54 (Suppl. 3):A134, 2000. 177. Myckatyn, T. M., Ellis, R. A., Grand, A. G., et al.: The effects of rapamycin in murine peripheral nerve isografts and allografts. Plast. Reconstr. Surg. 109:2405, 2002. 178. Naba, I., Yoshikawa, H., Sakoda, S., et al.: Onion-bulb formation after a single compression injury in the macrophage scavenger receptor knockout mice. Exp. Neurol. 166:83, 2000. 179. Naba, I., Yoshikawa, H., Sakoda, S., et al.: Successful generation of peripheral neuropathy with onion-bulb formation in the macrophage scavenger receptor class A knockout mouse treated with isoniazid. Neurosci. Lett. 290:5, 2000. 180. Nechaev, E. A.: Mine Blast Trauma : Experience from the War in Afghanistan. St Petersburg, Russian Ministry of Public Health and Medical Industry, 1995. 181. Noah, E. M., Williams, A., Fortes, W., and Terzis, J. K.: A new animal model to investigate axonal sprouting after end-to-side neurorrhaphy. J. Reconstr. Microsurg. 13:317, 1997. 182. Noble, J., Munro, C. A., Prasad, V. S., and Midha, R.: Analysis of upper and lower extremity peripheral nerve injuries in a population of patients with multiple injuries. J. Trauma 45:116, 1998. 183. Nukada, H., Anderson, G. M., and McMorran, P. D.: Reperfusion nerve injury: pathology due to reflow and prolonged ischaemia. J. Peripher. Nerv. Syst. 2:60, 1997. 184. Nukada, H., Powell, H. C., and Myers, R. R.: Perineurial window: demyelination in nonherniated endoneurium with reduced nerve blood flow. J. Neuropathol. Exp. Neurol. 51:523, 1992. 185. Parmantier, E., Lynn, B., Lawson, D., et al.: Schwann cell-derived Desert hedgehog controls the development of peripheral nerve sheaths. Neuron 23:713, 1999. 186. Pasterkamp, R. J., Giger, R. J., and Verhaagen, J.: Regulation of semaphorin III/collapsin-1 gene expression during peripheral nerve regeneration. Exp. Neurol. 153:313, 1998. 187. Pearce, J. M.: Wallerian degeneration. J. Neurol. Neurosurg. Psychiatry 69:791, 2000. 188. Perry, V. H., Brown, M. C., and Gordon, S.: The macrophage response to central and peripheral nerve injury: a possible role for macrophages in regeneration. J. Exp. Med. 165:1218, 1987. 189. Perry, V. H., Brown, M. C., Lunn, E. R., et al.: Evidence that very slow wallerian degeneration in C57BL/Ola Mice is an intrinsic property of the peripheral nerve. Eur. J. Neurosci. 2:802, 1990.

Mechanisms of Repair after Traumatic Injury 190. Perry, V. H., Lunn, E. R., Brown, M. C., et al.: Evidence that the rate of wallerian degeneration is controlled by a single autosomal dominant gene. Eur. J. Neurosci. 2:408, 1990. 191. Perry, V. H., Tsao, J. W., Fearn, S., and Brown, M. C.: Radiation-induced reductions in macrophage recruitment have only slight effects on myelin degeneration in sectioned peripheral nerves of mice. Eur. J. Neurosci. 7:271, 1995. 192. Petersen, J., Russell, L., Andrus, K., et al.: Reduction of extraneural scarring by ADCON-T/N after surgical intervention. Neurosurgery 38:976; discussion 983, 1996. 193. Plantinga, L. C., Verhaagen, J., Edwards, P. M., et al.: The expression of B-50/GAP-43 in Schwann cells is upregulated in degenerating peripheral nerve stumps following nerve injury. Brain Res. 602:69, 1993. 194. Politis, M. J.: Specificity in mammalian peripheral nerve regeneration at the level of the nerve trunk. Brain Res. 328:271, 1985. 195. Pollard, J. D., and Fitzpatrick, L.: A comparison of the effects of irradiation and immunosuppressive agents on regeneration through peripheral nerve allografts: an ultrastructural study. Acta Neuropathol. (Berl.) 23:166, 1973. 196. Popovic, M., Bresjanac, M., and Sketelj, J.: Role of axondeprived Schwann cells in perineurial regeneration in the rat sciatic nerve. Neuropathol. Appl. Neurobiol. 26:221, 2000. 197. Raabe, T. D., Nguyen, T., Archer, C., and Bittner, G. D.: Mechanisms for the maintenance and eventual degradation of neurofilament proteins in the distal segments of severed goldfish Mauthner axons. J. Neurosci. 16:1605, 1996. 198. Raff, M. C., Whitmore, A. V., and Finn, J. T.: Axonal selfdestruction and neurodegeneration. Science 296:868, 2002. 199. Rath, S., and Green, C. J.: Lack of topographical specificity in sensory nerve regeneration through muscle grafts in rats. J. Hand Surg. Br. 16:524, 1991. 200. Ratner, N., Bunge, R. P., and Glaser, L.: A neuronal cell surface heparan sulfate proteoglycan is required for dorsal root ganglion neuron stimulation of Schwann cell proliferation. J. Cell Biol. 101:744, 1985. 201. Ratner, N., Bunge, R. P., and Glaser, L.: Schwann cell proliferation in vitro: an overview. Ann. N. Y. Acad. Sci. 486:170, 1986. 202. Redford, E. J., Hall, S. M., and Smith, K. J.: Vascular changes and demyelination induced by the intraneural injection of tumour necrosis factor. Brain 118 (Pt. 4):869, 1995. 203. Reichert, F., Levitzky, R., and Rotshenker, S.: Interleukin 6 in intact and injured mouse peripheral nerves. Eur. J. Neurosci. 8:530, 1996. 204. Robert, F., Guennoun, R., Desarnaud, F., et al.: Synthesis of progesterone in Schwann cells: regulation by sensory neurons. Eur. J. Neurosci. 13:916, 2001. 205. Robinson, L. R.: Traumatic injury to peripheral nerves. Muscle Nerve 23:863, 2000. 206. Rotshenker, S.: The cytokine network of wallerian degeneration. Curr. Top. Neurochem. 1:147, 1997. 207. Rovak, J. M., Cederna, P. S., Macionis, V., et al.: Terminolateral neurorrhaphy: the functional axonal anatomy. Microsurgery 20:6, 2000. 208. Rowan, P. R., Chen, L. E., and Urbaniak, J. R.: End-to-side nerve repair: a review. Hand Clin. 16:151, 2000.

1431

209. Roytta, M., and Salonen, V.: Long-term endoneurial changes after nerve transection. Acta Neuropathol. (Berl.) 76:35, 1988. 210. Salzer, J. L., and Bunge, R. P.: Studies of Schwann cell proliferation. I. An analysis in tissue culture of proliferation during development, wallerian degeneration, and direct injury. J. Cell Biol. 84:739, 1980. 211. Salzer, J. L., Bunge, R. P., and Glaser, L.: Studies of Schwann cell proliferation. III. Evidence for the surface localization of the neurite mitogen. J. Cell Biol. 84:767, 1980. 212. Salzer, J. L., Williams, A. K., Glaser, L., and Bunge, R. P.: Studies of Schwann cell proliferation. II. Characterization of the stimulation and specificity of the response to a neurite membrane fraction. J. Cell Biol. 84:753, 1980. 213. Sampaio, E. P., Sarno, E. N., Galilly, R., et al.: Thalidomide selectively inhibits tumor necrosis factor alpha production by stimulated human monocytes. J. Exp. Med. 173:699, 1991. 214. Saunders, R. D., Brandon, Y. W., and DeVries, G. H.: Role of intracellular pH in the axolemma- and myelin-induced proliferation of Schwann cells. J. Neurochem. 52:1576, 1989. 215. Schafer, M., Fruttiger, M., Montag, D., et al.: Disruption of the gene for the myelin-associated glycoprotein improves axonal regrowth along myelin in C57BL/Wlds mice. Neuron 16:1107, 1996. 216. Schafers, M., Geis, C., Brors, D., et al.: Anterograde transport of tumor necrosis factor-alpha in the intact and injured rat sciatic nerve. J. Neurosci. 22:536, 2002. 217. Scherer, S. S., and Arroyo, E. J.: Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Syst. 7:1, 2002. 218. Scherer, S. S., Kamholz, J., and Jakowlew, S. B.: Axons modulate the expression of transforming growth factor-betas in Schwann cells. Glia 8:265, 1993. 219. Scherman, P., Lundborg, G., Kanje, M., and Kahlin, L. B.: Neural regeneration along longitudinal polyglactin sutures across short and extended defects in the rat sciatic nerve. J. Neurosurg. 95:316, 2001. 220. Schlaepfer, W. W.: Calcium-induced degeneration of axoplasm in isolated segments of rat peripheral nerve. Brain Res. 69:203, 1974. 221. Schlaepfer, W. W.: Structural alterations of peripheral nerve induced by the calcium ionophore A23187. Brain Res. 136:1, 1977. 222. Schlaepfer, W. W., and Bunge, R. P.: Effects of calcium ion concentration on the degeneration of amputated axons in tissue culture. J. Cell Biol. 59:456, 1973. 223. Schroder, J. M., May, R., and Weis, J.: Perineurial cells are the first to traverse gaps of peripheral nerves in silicone tubes. Clin. Neurol. Neurosurg. 95 (Suppl.):S78, 1993. 224. Schumacher, M., Robel, P., and Baulieu, E. E.: Development and regeneration of the nervous system: a role for neurosteroids. Dev. Neurosci. 18:6, 1996. 225. Seddon, H.: Three types of nerve injury. Brain 66:237, 1943. 226. Shamash, S., Reichert, F., and Rotshenker, S.: The cytokine network of wallerian degeneration: tumor necrosis factoralpha, interleukin-1alpha, and interleukin-1beta. J. Neurosci. 22:3052, 2002. 227. Shubayev, V. I., and Myers, R. R.: Upregulation and interaction of TNFalpha and gelatinases A and B in painful peripheral nerve injury. Brain Res. 855:83, 2000.

1432

Neuropathy Caused by Compression, Entrapment, or Physical Injury

228. Shubayev, V. I., and Myers, R. R.: Endoneurial remodeling by TNFalpha and TNFalpha-releasing proteases: a spatial and temporal co-localization study in painful neuropathy. J. Peripher. Nerv. Syst. 7:28, 2002. 229. Siconolfi, L. B., and Seeds, N. W.: Mice lacking tPA, uPA, or plasminogen genes showed delayed functional recovery after sciatic nerve crush. J. Neurosci. 21:4348, 2001. 230. Siebert, H., Dippel, N., Mader, M., et al.: Matrix metalloproteinase expression and inhibition after sciatic nerve axotomy. J. Neuropathol. Exp. Neurol. 60:85, 2001. 231. Siebert, H., Sachse, A., Kuziel, W. A., et al.: The chemokine receptor CCR2 is involved in macrophage recruitment to the injured peripheral nervous system. J. Neuroimmunol. 110:177, 2000. 232. Smith, K. J., Kapoor, R., Hall, S. M., and Davies, M.: Electrically active axons degenerate when exposed to nitric oxide. Ann. Neurol. 49:470, 2001. 233. Snipes, G. J., and Suter, U.: Signaling pathways mediating axon–Schwann cell interactions. Trends Neurosci. 17:399, 1994. 234. Sobue, G., Kreider, B., Asbury A., and Pleasure, D.: Specific and potent mitogenic effect of axolemmal fraction on Schwann cells from rat sciatic nerves in serum-containing and defined media. Brain Res. 280:263, 1983. 235. Sobue, G., and Pleasure, D.: Adhesion of axolemmal fragments to Schwann cells: a signal- and target-specific process closely linked to axolemmal induction of Schwann cell mitosis. J. Neurosci. 5:379, 1985. 236. Soilu-Hanninen, M., Ekert, P., Bucci, T., et al.: Nerve growth factor signaling through p75 induces apoptosis in Schwann cells via a Bcl-2-independent pathway. J. Neurosci. 19:4828, 1999. 237. Sommer, C., Marziniak, M., and Myers, R. R.: The effect of thalidomide treatment on vascular pathology and hyperalgesia caused by chronic constriction injury of rat nerve. Pain 74:83, 1998. 238. Sondell, M., Lundborg, G., and Kanje, M.: Vascular endothelial growth factor stimulates Schwann cell invasion and neovascularization of acellular nerve grafts. Brain Res. 846:219, 1999. 239. Sorensen, J., Haase, G., Krarup, C., et al.: Gene transfer to Schwann cells after peripheral nerve injury: a delivery system for therapeutic agents. Ann. Neurol. 43:205, 1998. 240. Spencer, P. S., Weinberg, H. J., Raine, C. S., and Prineas, J. W.: The perineurial window—a new model of focal demyelination and remyelination. Brain Res. 96:323, 1975. 241. Stahl, S., and Rosenberg, N.: Digital nerve repair by autogenous vein graft in high-velocity gunshot wounds. Mil. Med. 164:603, 1999. 242. Stoll, G., Jander, S., and Myers, R. R.: Degeneration and regeneration of the peripheral nervous system: from Augustus Waller’s observations to neuroinflammation. J. Peripher. Nerv. Syst. 7:13, 2002. 243. Strasberg, S. R., Hertl, M. C., Mackinnon, S. E., et al.: Peripheral nerve allograft preservation improves regeneration and decreases systemic cyclosporin A requirements. Exp. Neurol. 139:306, 1996. 244. Streppel, M., Azzolin, N., Dohm, S., et al.: Focal application of neutralizing antibodies to soluble neurotrophic factors

245.

246.

247.

248.

249. 250.

251.

252.

253.

254.

255.

256.

257.

258.

259.

260.

261.

reduces collateral axonal branching after peripheral nerve lesion. Eur. J. Neurosci. 15:1327, 2002. Subang, M. C., and Richardson, P. M.: Synthesis of leukemia inhibitory factor in injured peripheral nerves and their cells. Brain Res. 900:329, 2001. Sugimoto, Y., Takayama, S., Horiuchi, Y., and Toyama, Y.: An experimental study on the perineurial window. J. Peripher. Nerv. Syst. 7:104, 2002. Sulaiman, O. A., Voda, J., Gold, B. G., and Gordon, T.: FK506 increases peripheral nerve regeneration after chronic axotomy but not after chronic Schwann cell denervation. Exp. Neurol. 175:127, 2002. Sundback, C., Hadlock, T., Cheney, M., and Vacanti, J.: Manufacture of porous polymer nerve conduits by a novel low-pressure injection molding process. Biomaterials 24:819, 2003. Sunderland, S.: A classification of peripheral nerve injuries producing loss of function. Brain 74:491, 1951. Sunderland, S., and Bradley, K. C.: Denervation atrophy of the distal stump of a severed nerve. J. Comp. Neurol. 93:401, 1950. Sunderland, S., and Bradley, K. C.: Endoneurial tube shrinkage in the distal segment of a severed nerve. J. Comp. Neurol. 93:411, 1950. Swett, J. E., Hong, C. Z., and Miller, P. G.: All peroneal motoneurons of the rat survive crush injury but some fail to reinnervate their original targets. J. Comp. Neurol. 304:234, 1991. Syroid, D. E., Maycox, P. R., Burrola, P. G., et al.: Cell death in the Schwann cell lineage and its regulation by neuregulin. Proc. Natl. Acad. Sci. U. S. A. 93:9229, 1996. Tang, J. B., Gu, Y. Q., and Song, Y. S.: Repair of digital nerve defect with autogenous vein graft during flexor tendon surgery in zone 2. J. Hand Surg. Br. 18:449, 1993. Tang, S., Shen, Y. J., DeBellard, M. E., et al.: Myelinassociated glycoprotein interacts with neurons via a sialic acid binding site at ARG118 and a distinct neurite inhibition site. J. Cell Biol. 138:1355, 1997. Taniuchi, M., Clark, H. B., and Johnson, E. M. Jr.: Induction of nerve growth factor receptor in Schwann cells after axotomy. Proc. Natl. Acad. Sci. U. S. A. 83:4094, 1986. Taskinen, H. S., Olsson, T., Bucht, A., et al.: Peripheral nerve injury induces endoneurial expression of IFNgamma, IL-10 and TNF-alpha mRNA. J. Neuroimmunol. 102:17, 2000. Taskinen, H. S., and Roytta, M.: The dynamics of macrophage recruitment after nerve transection. Acta Neuropathol. (Berl.) 93:252, 1997. Terenghi, G., Calder, J. S., Birch, R., and Hall, S. M.: A morphological study of Schwann cells and axonal regeneration in chronically transected human peripheral nerves. J. Hand Surg. Br. 23:583, 1998. Terzis, J., Faibisoff, B., and Williams, B.: The nerve gap: suture under tension vs. graft. Plast. Reconstr. Surg. 56:166, 1975. Tetzlaff, W., Alexander, S. W., Miller, F. D., and Bisby, M. A.: Response of facial and rubrospinal neurons to axotomy: changes in mRNA expression for cytoskeletal proteins and GAP-43. J. Neurosci. 11:2528, 1991.

Mechanisms of Repair after Traumatic Injury 262. Thomas, P. K.: Changes in the endoneurial sheaths of peripheral myelinated nerve fibres during wallerian degeneration. J. Anat. 98:175, 1964. 263. Thomas, P. K.: The cellular response to nerve injury. 1. The cellular outgrowth from the distal stump of transected nerve. J. Anat. 100:287, 1966. 264. Thomas, P. K., and Holdorff, B.: Neuropathy due to physical agents. In Dyck, P., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 990, 1993. 265. Toews, A. D., Barrett, C., and Morell, P.: Monocyte chemoattractant protein 1 is responsible for macrophage recruitment following injury to sciatic nerve. J. Neurosci. Res. 53:260, 1998. 266. Torigoe, K., Tanaka, H. F., Takahashi, A., et al.: Basic behavior of migratory Schwann cells in peripheral nerve regeneration. Exp. Neurol. 137:301, 1996. 267. Trachtenberg, J. T., and Thompson, W. J.: Schwann cell apoptosis at developing neuromuscular junctions is regulated by glial growth factor. Nature 379:174, 1996. 268. Tsao, J. W., George, E. B., and Griffin, J. W.: Temperature modulation reveals three distinct stages of wallerian degeneration. J. Neurosci. 19:4718, 1999. 269. Viterbo, F., Palhares, A., and Franciosi, L. F.: Restoration of sensitivity after removal of the sural nerve: a new application of latero-terminal neurorrhaphy. Rev. Paul. Med. 112:658, 1994. 270. Vougioukas, V. I., Roeske, S., Michel, U., and Bruck, W.: Wallerian degeneration in ICAM-1-deficient mice. Am. J. Pathol. 152:241, 1998. 271. Wagner, R., and Myers, R. R.: Endoneurial injection of TNF-alpha produces neuropathic pain behaviors. Neuroreport 7:2897, 1996. 272. Wagner, R., and Myers, R. R.: Schwann cells produce tumor necrosis factor alpha: expression in injured and non-injured‘ nerves. Neuroscience 73:625, 1996. 273. Wall, J. T., Xu, J., and Wang, X.: Human brain plasticity: an emerging view of the multiple substrates and mechanisms that cause cortical changes and related sensory dysfunctions after injuries of sensory inputs from the body. Brain Res. Rev. 39:181, 2002. 274. Waller, A. V.: Experiments on the section of the glossopharyngeal and hypoglossal nerves of the frog, and observations of the alterations produced thereby in the structure of their primitive fibres. Philos. Trans. R. Soc. Lond. B Biol. Sci. 140:423, 1850. 275. Weiner, J. A., Fukushima, N., Contos, J. J., et al.: Regulation of Schwann cell morphology and adhesion by receptormediated lysophosphatidic acid signaling. J. Neurosci. 21:7069, 2001. 276. Weis, J., May, R., and Schroder, J. M.: Fine structural and immunohistochemical identification of perineurial cells connecting proximal and distal stumps of transected

277.

278.

279.

280.

281.

282.

283.

283a.

284.

285.

286.

287.

288.

289.

1433

peripheral nerves at early stages of regeneration in silicone tubes. Acta Neuropathol. (Berl.) 88:159, 1994. Willison, H. J., Trapp, B. D., Bacher, J. D., and Quarles, R. H.: The expression of myelin-associated glycoprotein in regenerating cat sciatic nerve. Brain Res. 444:10, 1988. Wu, G., Ringkamp, M., Hartke, T. V., et al.: Early onset of spontaneous activity in uninjured C-fiber nociceptors after injury to neighboring nerve fibers. J. Neurosci. 21:RC140, 2001. Xian, C. J., Li, L., Deng, Y. S., et al.: Lack of effects of transforming growth factor-alpha gene knockout on peripheral nerve regeneration may result from compensatory mechanisms. Exp. Neurol. 172:182, 2001. Xie, Y., Yeo, T. T., Zhang, C., et al.: The leukocyte common antigen-related protein tyrosine phosphatase receptor regulates regenerative neurite outgrowth in vivo. J. Neurosci. 21:5130, 2001. Yan, J. G., Matloub, H. S., Sanger, J. R., et al.: A modified end-to-side method for peripheral nerve repair: large epineurial window helicoid technique versus small epineurial window standard end-to-side technique. J. Hand Surg. Am. 27:484, 2002. You, S., Petrov, T., Chung, P. H., and Gordon, T.: The expression of the low affinity nerve growth factor receptor in longterm denervated Schwann cells. Glia 20:87, 1997. Zhao, Q., Dahlin, L. B., Knaje, M., and Lundborg, G.: Specificity of muscle reinnervation following repair of the transected sciatic nerve: a comparative study of different repair techniques in the rat. J. Hand Surg. Br. 17:257, 1992. Zhai, Q., Wang, J., Kim, A., et al.: Involvement of the ubiquitin-proteasome system in the early stage of wallerian degeneration. Neuron 39:217, 2003. Zhong, J., Dietzel, I. D., Wahle, P., et al.: Sensory impairments and delayed regeneration of sensory axons in interleukin-6-deficient mice. J. Neurosci. 19:4305, 1999. Zochodne, D. W., and Cheng, C.: Neurotrophins and other growth factors in the regenerative milieu of proximal nerve stump tips. J. Anat. 196 (Pt. 2):279, 2000. Zochodne, D. W., Levy, D., Zwiers, H., et al.: Evidence for nitric oxide and nitric oxide synthase activity in proximal stumps of transected peripheral nerves. Neuroscience 91:1515, 1999. Zuo, J., Ferguson, T. A., Hernandez, Y. J., et al.: Neuronal matrix metalloproteinase-2 degrades and inactivates a neurite-inhibiting chondroitin sulfate proteoglycan. J. Neurosci. 18:5203, 1998. Zuo, J., Hernandez, Y. J., and Muir, D.: Chondroitin sulfate proteoglycan with neurite-inhibiting activity is up-regulated following peripheral nerve injury. J. Neurobiol. 34:41, 1998. Zuo, J., Neubauer, D., Graham, J., et al.: Regeneration of axons after nerve transection repair is enhanced by degradation of chondroitin sulfate proteoglycan. Exp. Neurol. 176:221, 2002.

59 Median Neuropathy J. CLARKE STEVENS

Carpal Tunnel Syndrome Historical Aspects Anatomy Incidence Symptoms Physical Examination Correlation among Symptoms, Physical Examination, and Nerve Conduction Study Findings

Clinical Criteria for Diagnosis in Epidemiologic Studies CTS Staging System and Severity Scales CTS Severity by EMG Criteria Etiology Differential Diagnosis Laboratory Diagnosis Medical Treatment

CARPAL TUNNEL SYNDROME Compression of the median nerve in the carpal tunnel is the most common and well-studied entrapment that occurs in humans. Although the condition would appear to be simple and straightforward, there are still many facets of the disorder that are not completely understood. These issues include standards for clinical and electromyographic (EMG) diagnosis, the incidence in the general population, the pathophysiology and etiology of idiopathic carpal tunnel syndrome (CTS), and the relation of occupational factors to the development of symptoms. New diagnostic techniques such as ultrasound and magnetic resonance imaging (MRI) of the wrist are undergoing evaluation and appear to give new and important information in some cases. Most people experience nocturnal acroparesthesias at one time or another. Patients with CTS typically awaken from sleep with an asleep numbness, prickling dysesthesia, and pain in the hand that can be relieved by shaking and rubbing the hand and by putting the limbs in a dependent position. Similar symptoms may be present upon awakening in the morning. The symptoms are usually worse in the right hand but are frequently bilateral. The symptoms are reproduced by everyday activities such as gripping the steering wheel, talking on the phone, and reading the newspaper. Treatment of an underlying condition, wrist

Surgical Treatment Median Nerve Lesions in the Arm Median Nerve Entrapment by the Ligament of Struthers Proximal Forearm Entrapment of the Median Nerve Anterior Interosseous Nerve Syndrome

splints at night, and steroid injections are usually helpful. Surgical treatment is highly effective if needed.

Historical Aspects Marie and Foix are credited with the first description of this nerve entrapment in 1913. Their paper provided detailed autopsy findings in an 80-year-old woman with bilateral thenar atrophy.163 At the Mayo Clinic, the first case of carpal tunnel release was performed in 1927 for severe CTS resulting from dislocation of the lunate bone.7 A case of severe “median thenar neuritis” in a stenographer was described in 1938, but she was not treated surgically.155 The first series of surgically treated cases with an accurate illustration of the divided transverse carpal ligament was published in 1946.30 In 1947, Brain et al. published an important paper describing the symptoms of CTS and surgical treatment.25 The term carpal tunnel syndrome was first used by George S. Phalen and associates in several influential papers in the early 1950s in which they described the symptoms, the wrist flexion maneuver as a provocative test, and surgical treatment.198,199 Recognition of the syndrome increased rapidly in subsequent years. Median motor nerve conduction studies for the diagnosis of CTS were described in 1956.228 Recordings from the thenar muscles using a needle electrode showed prolonged 1435

1436

Diseases of the Peripheral Nervous System

distal latencies. Two years later, orthodromic sensory nerve conduction studies in CTS were described, providing the basis for many subsequent electrophysiologic studies of the disorder.81

Anatomy The median nerve is formed by the union of a branch from C6-C7 of the lateral cord of the plexus and a lower branch (C8-T1) of the medial cord of the plexus (Fig. 59–1). The nerve has no branches in the arm. In the forearm, it supplies the pronator teres, flexor carpi radialis, flexor digito-

rum sublimis, and palmaris longus. The anterior interosseous branch arises in the distal part of the antecubital fossa and supplies the flexor pollicis longus, lateral part of the flexor digitorum profundus (tendons to digits 4 and 5), and pronator quadratus. Distal to the wrist, the median nerve supplies the abductor pollicis brevis, the flexor pollicis brevis (superficial head), the opponens pollicis, and the two lateral lumbrical muscles. The ulnar nerve usually innervates the deep head of the flexor pollicis brevis. In some cases the deep and superficial heads of the flexor pollicis brevis have dual innervation. Cutaneous branches are the palmar cutaneous branch and palmar digital nerves.

Median nerve

Flexor carpi radialis

Pronator teres

Palmaris longus

Flexor digitorum sublimis

Flexor digitorum profundus

Flexor pollicis longus Flexor digitorum profundus

Anterior interosseous nerve

Pronator quadratus

Cutaneous innervation

Abductor pollicis brevis Opponens pollicis Superficial head of flexor pollicis brevis Post.

1st and 2nd lumbricals

Ant. Ant.

FIGURE 59–1 Median nerve anatomy. The nerve indicated by 1 is the palmar cutaneous nerve; 2 indicates the palmar digital nerves. (From Haymaker, W., and Woodhall, B.: Peripheral nerve injuries. In Principles of Diagnosis, 2nd ed. Philadelphia, W. B. Saunders, p. 243, 1953, with permission.)

Median Neuropathy

Anomalous median-to-ulnar nerve communications in the forearm are referred to as the Martin-Gruber anastomosis or a “crossover.”91 Using electrophysiologic techniques, the anomaly is found in approximately 30% of subjects, sometimes with a dominant mode of inheritance. Motor fibers descending in the median nerve cross in the forearm to join the ulnar nerve. Usually the small branches originate from the anterior interosseous nerve. These axons most frequently innervate the first dorsal interosseous muscle, followed by the hypothenar and thenar muscles. The anomaly has significance as a source of confusion in the electrodiagnosis of CTS because these fibers bypass the carpal tunnel and avoid the slowing of conduction that occurs in median fibers that pass through the carpal tunnel. The RicheCannieu anastomosis is a communication between the deep ulnar and median recurrent motor branch that occasionally innervates some or all of the median-innervated thenar muscles.217 This anomaly can be suspected in patients with CTS when the median motor and sensory responses are absent yet the thenar muscle bulk and needle electromyography are normal.205 Median nerve fibers that pass through the carpal tunnel supply sensation to the palmar side of the radial three and one-half digits. The dorsal aspects of the distal phalanges of these digits are also affected and are easier to test when the palmar aspect of the fingers is calloused. Autonomic fibers accompany the branches to the fingers. The base of the palm is not affected in CTS because it is supplied by the palmar cutaneous branch that arises approximately 6 cm proximal to the carpal tunnel and passes in front of the carpal ligament. The carpal tunnel is formed anteriorly by the transverse carpal ligament (flexor retinaculum), which measures approximately 2 to 3 cm in length and width (Figs. 59–2 and 59–3). It attaches medially to the pisiform and hook of the hamate and laterally to the scaphoid and medial trapezium. The anterior surfaces of the carpal bones make up the posterior aspect of the canal. The ligament is composed of interwoven bundles of fibrous connective tissue and contains intra- and extraligamentous innervation by free nerve endings and pacinian corpuscles.150 The proximal border of the ligament is at the palmar wrist crease. The narrowest point of the tunnel is at the level of the hook of the hamate. The carpal tunnel makes up approximately 29% of the total cross-sectional area of the wrist.216 The dimensions of the carpal tunnel have been studied by MRI.200 It has a cone shape with the inlet larger than the outlet. The narrowest point of the canal is at its distal third, at 8 mm from the outlet. This correlates with neurophysiologic findings that demonstrate a sharp latency increase near the distal edge of the ligament in 52% of patients with CTS. In the remainder, the latency increase is more evenly distributed across the carpal tunnel.121 The mean length of the tunnel from inlet to outlet is 36.3 mm. The mean cross-sectional area of the tunnel inlet and volume of the tunnel are larger in women older than age 45 compared with younger age groups.

1437

FIGURE 59–2 Anatomy of the carpal tunnel and distribution of the terminal motor and sensory branches of the median nerve. (Copyright 1997, Icon Learning Systems, LLC. A subsidiary of MediMedia USA, Inc. All rights reserved.)

FIGURE 59–3 Cross-sectional anatomy of the carpal tunnel. (Copyright 1997, Icon Learning Systems, LLC. A subsidiary of MediMedia USA, Inc. All rights reserved.)

1438

Diseases of the Peripheral Nervous System

The median nerve is accompanied in the canal by nine flexor tendons of the digits. The thenar motor branch arises from the central portion of the median nerve and in 74% of subjects passes distal to the transverse carpal ligament through obliquely oriented fascia that originates from the ligament. In 19% of subjects the nerve passes distal to the ligament and is surrounded only by muscle. In 7% the nerve passes through an opening in the transverse carpal ligament.125 Occasionally there is a common trunk of origin of the motor branch and sensory branches to the thumb and index finger. These anatomic variations are important to surgeons performing carpal tunnel releases and provide a possible explanation for the rare patients with CTS that affect the motor branch exclusively.

Symptoms Sensory Symptoms and Pain The symptoms of patients with CTS confirmed by abnormal median nerve conduction across the wrist have been carefully studied (Tables 59–1 and 59–2).90,239 Numbness and tingling in the fingers and hand are the primary symptoms of CTS. Many patients also experience pain, but patients complaining of pain in the wrist, thumb, or fingers who do not have sensory symptoms rarely have CTS. It is important to recognize that the sensory symptoms experiTable 59–1. Hand Symptom Questionnaire Distribution of Carpal Tunnel Syndrome Sensory Symptoms in 159 Hands Distribution of Sensory Symptoms

Incidence In a population-based study, 78% of patients with CTS were women. In other large series, women comprise 65% to 94% of cases. Fifty percent of these patients are between 40 and 60 years of age at the time of diagnosis. Age-specific incidence rates generally increase with age in men, whereas in women a peak is reached in the 45- to 54-year age group.240 The symptoms tend to affect the dominant hand more frequently and severely, but in approximately 60% of patients the symptoms are bilateral. The incidence (cases per 100,000 person-years) in a small Minnesota city from 1961 to 1980 was 99 overall; however, according to a more recent survey the incidence may be increasing.177 A general population survey of people living in Maastricht, The Netherlands, estimated the prevalence as 7.4% in women and 0.6% in men. This may be an underestimate because only those who admitted to nocturnal awakening with hand symptoms were studied. In a mail survey, 14.4% of responders from a sample of the population in southern Sweden had sensory symptoms in the median distribution.13 The prevalence of clinically certain CTS was 3.8%, whereas the clinically and electrophysiologically confirmed CTS prevalence was 2.7%. Surveys have turned up a surprisingly high incidence of asymptomatic slowing of median nerve conduction at the wrist of up to 18%.67 This finding has considerable importance when attempting to understand the incidence of CTS in a population or the workplace. A mail survey of patients in a regional health care system in Kentucky attempted to assess the frequency of CTS using a hand diagram.191 A total of 16.3% of the responders had a classic or probable distribution of symptoms. Assuming that the nonresponders did not have CTS led to a lowest possible prevalence estimate of 3.7% in the general U.S. population. This study did not have nerve conduction study verification of the symptoms and may be an overestimate of the prevalence.

Which digits are affected by paresthesia? Uncertain Thumb Index Middle Ring Little Which is the most affected digit? Thumb Index Middle Ring Little No single finger most affected or uncertain Are the symptoms mostly on the front or on the back of the hand and fingers? Mostly palmar Mostly dorsal Both Uncertain

% of Hands 2.5 78.6 92.5 96.2 83.0 56.6 11.9 12.6 27.0 3.8 2.5 42.2

69.2 10.7 10.1 10.1

Data from Stevens, J. C., Smith, B. E., Weaver, A. L., et al.: Symptoms of 100 patients with electromyographically verified carpal tunnel syndrome. Muscle Nerve 22:1448, 1999.

Table 59–2. Symptoms of 100 Patients with Carpal Tunnel Syndrome Symptom Awakened from sleep by paresthesia/pain More than once a week Less than once a week Hands asleep upon awakening in morning Paresthesia when driving Paresthesia when reading Relief of sensory symptoms by shaking the hands

No. of Patients 63 17 61 58 70 68

Data from Stevens, J. C., Smith, B. E., Weaver, A. L., et al.: Symptoms of 100 patients with electromyographically verified carpal tunnel syndrome. Muscle Nerve 22:1448, 1999.

Median Neuropathy

enced by patients frequently extend beyond the anatomic distribution of the median nerve sensory fibers. In fact, the most common pattern of paresthesias involves both the median and ulnar nerve digits in 48% of cases (Fig. 59–4). It is unclear why patients report symptoms in the ulnar nerve distribution, but they persist in this perception even after they have been asked to carefully think about where their symptoms are located. A total of 56% of patients feel the little finger is affected by paresthesia; however, it is rarely chosen as the most affected digit. Of the medianinnervated digits, the middle and index fingers are most frequently affected. Forty-two percent of patients are unable to name any single finger as most affected. Symptoms confined to fingers innervated by the median nerve were noted in 45% of hands when patients marked a hand symptom diagram (Fig. 59–5). Usually three or four fingers are affected; symptoms confined to a single finger are quite unusual in CTS. A small number of patients experience symptoms in a glovelike distribution involving the entire hand to the wrist front and back (see Fig. 59–3). More unusual distributions of symptoms occur, including paresthesias confined to the ulnar distribution. Ulnar nerve conduction studies in patients with ulnar symptoms are normal, and it is unclear why the symptoms are displaced from median to ulnar fingers.184 The sensory examination of the little finger in patients with ulnar symptoms, however, is always normal. Failure to recognize that ulnar symptoms occur in CTS may lead the surgeon to recommend unnecessary decompression of Guyon’s canal. Some

FIGURE 59–4 Hand symptom diagram showing the most common distribution of sensory symptoms in CTS, a combination of median and ulnar innervated fingers. (From Stevens, J. C., Smith, B. E., Weaver, A. L., et al.: Symptoms of 100 patients with electromyographically verified carpal tunnel syndrome. Muscle Nerve 22:1448, 1999, with permission.)

1439

FIGURE 59–5 Sensory symptom diagram in a patient with CTS. Symptoms are confined to the median nerve distribution, with the most marked symptoms affecting the middle finger.

patients have difficulty deciding if the symptoms are on the palmar or dorsal surface of the hand, or they may feel both sides are affected. Those with persisting numbness of the fingertips complain of difficulty using the hand to feel and pick up small objects or to sew. In less than 1% of cases studied electrophysiologically, patients have selective involvement of median motor fibers and normal sensory nerve conduction studies.208,236 The symptoms of these patients are not well described, but presumably they have few sensory symptoms. Possibly, the recurrent motor branch is compressed after it has left the main trunk of the median nerve, but the exact site and mechanism of injury in these cases are unknown. Proximal Radiation of Symptoms Sensory symptoms or pain may extend from the fingers up the arm as far as the neck in 36% of patients, sometimes raising concern about proximal entrapment of the median nerve or an intraspinal process such as a cervical radiculopathy (Table 59–3). It is very common for patients to report symptoms extending proximally to the palm and wrist even though the palm receives its sensory supply by the palmar cutaneous branch, which does not pass through the carpal tunnel. Symptoms at levels above the wrist become progressively less frequent. Prominent symptoms at the neck and shoulder are very uncommon and should prompt a search for something other than CTS to explain them.

1440

Diseases of the Peripheral Nervous System

Table 59–3. Hand Symptom Questionnaire Distribution of Paresthesia and Pain Proximal to Fingers in 159 Hands with Carpal Tunnel Syndrome Symptoms Proximal to the Fingers

% of hands

Paresthesia No proximal paresthesia Palm Wrist Forearm Elbow Arm Shoulder Neck

42.8 56.0 30.2 21.4 10.1 5.7 1.9 0

Pain No associated pain Hand Wrist Forearm Elbow Arm Shoulder Neck

52.8 40.3 27.0 21.4 13.8 7.5 6.3 0.6

Data from Stevens, J. C., Smith, B. E., Weaver, A. L., et al.: Symptoms of 100 patients with electromyographically verified carpal tunnel syndrome. Muscle Nerve 22:1448, 1999.

Factors That Aggravate and Relieve Symptoms The numbness and tingling of CTS are characteristically intermittent, but can become persistent when median nerve compression is severe and sensory loss begins to develop. Most patients note that the symptoms come on or are made worse by certain activities or when they lie down. Numbness and pain in the fingers and hands awakening the patient at night are particularly characteristic of CTS. This symptom is reported significantly more often in patients with CTS and abnormal nerve conduction studies than in patients with CTS and normal nerve conduction studies or patients with other upper extremity neuromuscular problems. It is likely that many patients sleep with the wrist flexed, resulting in awakening one or more times during the night. Changes in upper extremity blood volume during sleep might also contribute to nocturnal symptoms and their relief after getting up in the morning.204 Keeping the hand in a neutral position with a wrist splint usually prevents nocturnal awakening. Awakening in the morning with numb hands is also common. During the day, activities such as driving a car, reading the newspaper, talking on the telephone, needlework, knitting, and working with the arms over the head may precipitate the symptoms. When talking on the telephone or driving, patients will often have to switch hands or put the hand down when it goes to sleep. Patients sometimes experience symptoms

when typing on a computer. Patients will try to “get the circulation back” by shaking and working the hands, which is usually effective. Dangling the afflicted hand over the edge of the bed or in some other dependent position also helps. Less-specific symptoms of CTS include pain, hand weakness, clumsiness of the hands, and dropping items.171,264 The latter symptoms occur about equally in patients with CTS who have normal and abnormal nerve conduction studies and are common in those with other upper extremity spine and neuromuscular problems.

Physical Examination Sensory and Motor Examination Many patients with CTS have little in the way of physical findings, and therefore the physical examination has limited diagnostic utility.115 Sensory testing can be hard to interpret because of calloused or thick skin. In one early laboratory-based series, two thirds of the patients with abnormal sensory nerve conduction studies had a normal sensory examination. Another study of severely affected cases showed impaired vibration and two-point discrimination in 66% and reduced pain perception in 42%.231 In the latter study, the frequency of abnormal sensory findings increased within three groups with increasing clinical symptoms of CTS, but, surprisingly, abnormal nerve conduction studies were found in approximately 80% regardless of clinical severity. The physical examination is important to assist in the detection of other or superimposed neurologic processes that cause hand paresthesia. Weakness and atrophy of the thenar muscles is a late sign and indicates severe median nerve compression. Patients rarely notice the atrophy, although they may complain of weakness of grip or a tendency to drop things. Atrophy of the thenar muscles may appear to be present in patients with arthritis, especially of the carpal-metacarpal joint of the thumb, yet the thenar motor amplitude is normal. In others, a fleshy appearance of the area may make appreciation of atrophy difficult. Occasional patients are encountered with advanced CTS on physical examination and by nerve conduction studies who claim to have had no or minimal paresthesias or pain. They may simply be stoic, or perhaps they have forgotten they had symptoms over the years. They are surprised when thenar atrophy is pointed out. Most patients with similar clinical and electrophysiologic findings have prominent numbness and discomfort. Interestingly, some patients with CTS may actually have decreased symptoms when nerve compression becomes severe.188 Disturbances of small and autonomic fiber function can occasionally be noticeable in CTS. Raynaud’s phenomenon and acrocyanosis, sometimes associated with systemic disease, may develop.136 Rarely other autonomic dysfunction, including blisters, ulceration, anhydrosis, and changes in skin temperature, will be found.42,73 Abnormalities of the

1441

Median Neuropathy

sympathetic skin response have been found in CTS, and thresholds for hot and cold sensation are higher.85 Reflex sympathetic dystrophy has been reported in patients with CTS precipitated by polymyalgia rheumatica, unaccustomed exercise, and seronegative arthritis.69 Provocative Tests In some patients, it is possible to reproduce the painful paresthesias by holding the wrists in extreme flexion for a minute (Phalen’s maneuver). Extreme extension of the wrists also increases the pressure in the carpal tunnel but is less useful. Paresthesias radiating out into the median fingers may occur when the median nerve is gently percussed with a reflex hammer (Tinel’s sign). Symptoms after pressure of the examiner’s thumbs over the carpal tunnel for 60 seconds form the basis of the carpal compression test.57 The well-known tendency for CTS symptoms to occur on arm raising has led to its suggestion as a new test for CTS.2 The patient is instructed to elevate the hands over the head, and, if paresthesias or numbness occur within 2 minutes, the test is considered positive. The development of symptoms with the arms up should make the examiner think of CTS rather than the much less common thoracic outlet syndrome. The tourniquet test is performed by inflating the cuff above systolic pressure to induce paresthesias.82 This test is quite insensitive and not very specific, and is not recommended as a routine screening test in the diagnosis of CTS.78 The sensitivity and specificity of these signs for diagnosis vary widely in published reports, but many studies show poor predictive value.49 This is because they are positive in some normal subjects and in patients with a peripheral neuropathy. In addition, the prevalence of positive tests is significantly less when the median neuropathy is severe. Many authors consider these signs to have minimal or no usefulness for diagnosis.50,118,160 With this in mind, most examiners still perform one or more of the tests while realizing they are not very specific and cannot be a significant factor by themselves for the diagnosis of CTS.

Correlation among Symptoms, Physical Examination, and Nerve Conduction Study Findings A number of studies have attempted to relate the severity of symptoms to nerve conduction study findings. Several slightly different electrophysiologic severity scales correlate fairly well with the Boston CTS functional and symptom severity scores.53,188 There is also a very good relationship between one nerve conduction study severity scale and a CTS symptom score consisting of common symptoms, including the distribution of symptoms in the fingers.22 However, patients classified in one neurophysiologic grade may interpret their symptoms on the symptom questionnaire from negative to extreme, and so the correl-

ation is not perfect.53 In an Italian study, symptoms were surprisingly more severe in those with a normal electromyogram than in those with a mildly abnormal electromyogram. Median nerve conduction study findings correlate better with the major symptoms of CTS (numbness, tingling, and nocturnal symptoms) than with minor symptoms (pain, weakness, and clumsiness).264 Electrophysiologic severity also tends to increase with increasing age and duration of symptoms.188,239 The use of a hand symptom diagram is quite sensitive for evaluation of clinic patients suspected of having CTS. Unfortunately, when used as a screening tool to predict abnormal nerve conduction studies in unselected community residents, it has a poor sensitivity and is not useful.68 Similar findings were obtained in another study examining the best methods to screen workers in manufacturing, office, and computer-related jobs for CTS.101 Physical examination, symptom surveys, and hand diagrams correlated poorly with sensory nerve conduction study findings. Eighty-two percent of workers who reported symptoms potentially consistent with CTS had normal nerve conduction studies, and, conversely, 60% who had abnormal nerve conduction studies did not have symptoms consistent with CTS.

Clinical Criteria for Diagnosis in Epidemiologic Studies A variety of criteria for the clinical diagnosis of CTS in research studies have been put forward. As noted above, symptoms involving ulnar- as well as median-innervated fingers are very common. Including only patients with symptoms limited to the median-innervated fingers and excluding patients from research studies if they have involvement of all fingers or the whole hand is too restrictive and will eliminate many patients with CTS. Consensus criteria for the classification of CTS in epidemiologic studies have been published.207 We have used the following clinical classification, which is based on symptoms: Definite/Classic/Probable—Numbness and tingling with or without pain in at least two of digits 1, 2, 3, and 4. Involvement of all fingers or the whole hand in a glove distribution is acceptable. In addition, the patient must have two of the other three major symptoms of CTS (nocturnal awakening with symptoms; daytime aggravation by reading, driving, talking on the phone, or working with the arms up; and relief or reduction of symptoms by shaking and rubbing the hands or putting them in a dependent position). Minor symptoms of CTS (weakness, clumsiness, and dropping items) are much less specific and are of minimal value for diagnosis. Possible—Must have the primary symptom of CTS (i.e., numbness and tingling with or without pain in at least two of digits 1, 2, 3, and 4). Involvement of all fingers or

1442

Diseases of the Peripheral Nervous System

the whole hand in a glove distribution is acceptable. May have no or up to one of the other major symptoms of CTS. Unlikely—No symptoms in digits 1, 2, and 3. The American Association of Electrodiagnostic Medicine CTS Task Force has recently published inclusion and exclusion criteria for the clinical diagnosis of CTS for use in research studies.107 Inclusion criteria are 1. Sensory symptoms (numbness and/or tingling) in at least two of digits 1, 2, 3, and 4 for at least 1 month. The sensory symptoms may be intermittent or constant, but, if constant, there must have been a period of time during which the symptoms were intermittent. The numbness and tingling may be accompanied by pain, but pain alone is not sufficient to meet this first inclusion criterion. 2. Sensory symptoms (numbness and/or tingling) aggravated by at least one of the following: sleep, sustained hand or arm positioning, or repetitive actions of the hand. 3. Sensory symptoms (numbness and/or tingling) mitigated by at least one of the following: changes in hand posture, shaking the hand, or use of a wrist splint. 4. If pain is present, the wrist, hand, and finger pain is greater than elbow, shoulder, or neck pain if there is pain in any or all of those locations. Exclusion criteria are 1. Sensory symptoms exclusive or predominantly in the little finger (ulnar neuropathy) 2. Neck pain or shoulder pain preceding the paresthesia in the digits (cervical radiculopathy and/or brachial plexopathy) 3. Numbness and/or tingling in the feet that preceded or accompanied the sensory symptoms in the hands (polyneuropathy) 4. Findings on the problem-focused history and physical examination that indicate an explanation for the sensory symptoms that is more probable than CTS (e.g., digital neuropathy; median nerve pathology proximal to the carpal tunnel; ulnar neuropathy; radial neuropathy; brachial plexopathy; cervical radiculopathy; spinal cord, brainstem, or brain pathology; or a polyneuropathy) The combination of the classic clinical criteria and abnormal median nerve conduction studies will identify patients who very likely have CTS with few falsely classified. Unfortunately, approximately 22% of patients meeting clinically classic criteria in a laboratory-based study have normal nerve conduction studies, and 47% with only possible symptoms of CTS can have prolonged median sensory latencies (J. C. Witt et al., unpublished data, 2003). As noted in the preceding section, the number of positive nerve conduction studies is much lower when screening workers for CTS.241

CTS Staging System and Severity Scales The staging system for CTS used in the Rochester Diabetic Neuropathy Study is as follows60: CTS 0: no electrophysiologic or clinical evidence of CTS regardless of whether the patient had previous carpal ligament section CTS 1: electrophysiologic findings typical of CTS but without typical symptoms, and including patients with previous carpal ligament section CTS 2a: patients with symptoms suggestive of CTS with or without typical electrophysiologic findings CTS 2b: patients with symptoms and findings and electrophysiologic tests diagnostic of CTS Widely used reproducible symptom severity and functional scales have been described.135 A historical-objective clinical severity scale has recently been described and is also reproducible. The five stages in this scale are80,159 0 ⫽ no symptoms suggestive of CTS in the previous 2 weeks 1 ⫽ paresthesia only at night or on waking in any part or all of the median nerve territory of the hand 2 ⫽ diurnal paresthesia 3 ⫽ any degree of sensory deficit 4 ⫽ hypotrophy and/or weakness of median-innervated thenar muscles 5 ⫽ complete atrophy or plegia of thenar muscles

CTS Severity by EMG Criteria A number of scales are used to describe the severity of CTS found on electrophysiologic testing.22,53,188 The severity scale we have found useful is as follows: Negative—no electrophysiologic evidence of CTS Mild—prolonged median palmar or antidromic sensory distal latency (absolute prolongation or compared with ulnar or radial distal latencies) ⫾ reduced amplitude Moderate—prolonged median motor distal latency Severe—absent sensory or mixed nerve action potential and reduced thenar motor amplitude Very Severe—no thenar motor or median sensory response; second lumbrical response may or may not be present

Etiology Idiopathic CTS In a retrospective population survey, idiopathic CTS was found in 43% and associated conditions that were a possible cause of CTS were documented in 57%.238 The risk factors for idiopathic CTS are largely related to personal physical attributes (Table 59–4). It is much more common in women than men by a factor of 2 to 3. The reason for this difference in incidence is not well understood. The

Median Neuropathy

Table 59–4. Risk Factors for Idiopathic CTS Female gender Obesity and increased body mass index Increasing age Hand size and shape Extra use of the dominant hand ? Menopausal hormonal changes

increased incidence in perimenopausal women suggests hormonal factors may have a role, but this is unproved. Increasing age is also correlated with the development of CTS. Obese individuals are 2.5 times more likely than slender individuals to be diagnosed with CTS.256 Morbidly obese subjects are at a nearly sevenfold greater risk for CTS.171 The wrist ratio, obtained by dividing the wrist depth by wrist width, is increased in patients with CTS, suggesting that a squarer wrist is a risk factor for CTS.87,124 Short stature and small hand size appear to be risk factors for CTS.166 Palm and third-digit length are shorter and wrist width longer in women with CTS compared with control subjects. The association between hand shape and CTS might be related to biomechanical factors. It is suggested that shorter hands require greater flexion for a given motion or require increased tension of the finger flexors during gripping that produces greater pressure in the carpal tunnel.40 Athletes have the lowest risk of CTS, whereas those with light or no avocational activities are at the highest risk. CTS more often affects the right hand than the left, suggesting that increased physical activity by the dominant hand is important. Early studies using computed tomography (CT) scanning suggested canal stenosis was a risk factor for CTS.23 More recent sophisticated studies using MRI to quantify the volume of the carpal tunnel in patients compared with control subjects have usually shown no difference in the size or cross-sectional area of the carpal tunnel.158,200 Pathophysiology. The mechanisms involved in the production of the pathologic changes and symptoms are still the subject of discussion. Measurements of pressure in the carpal tunnel have been made with a pressurerecording catheter at surgery and by an Angiocath transducer inserted under local anesthesia.84,96,255 Carpal canal pressures in the resting position, during flexion and extension of the wrist, and with gripping of the fingers are significantly higher in patients with CTS than in control subjects. There appears to be a correlation between carpal canal pressures and the severity of CTS. The pressures are highest in the distal portion of the carpal tunnel, where the median sensory amplitude is most often decreased compared with stimulation proximal and distal to the carpal tunnel.140 Overnight measurements of carpal tunnel pressure have shown the highest pressure at 6:00 AM. Splinting does not prevent critical pressure levels.141

1443

Increased resting pressures can be found in patients with normal nerve conduction studies, and these patients have more rapid resolution of symptoms. After successful surgery, carpal canal pressures return to normal. It is postulated that the pathologic change results in part from intrafascicular anoxia caused by obstruction of the venous return because of increased pressure in the carpal tunnel.244 The resulting intrafascicular edema and increased intrafascicular pressure destroy the nerve fibers by impairing their blood supply. The improvement of paresthesias that occurs almost immediately after surgery in some patients may be related to resolution of the ischemic component. The importance of ischemia is also demonstrated by a tourniquet test for CTS. The cuff is inflated above systolic pressure and is positive when hand paresthesias are induced. However, the ischemic hypothesis has the problem that, even if the arterial blood were excluded for a centimeter or more, ischemic injury is unlikely to occur because the blood supply to nerves can be ligated from long distances of nerve without causing ischemic injury. Animal models of acute nerve compression by a tourniquet show myelin invagination, particularly at the margins of the cuff. This results in paranodal demyelination and remyelination with slowing of the nerve conduction velocity.178 Most cases of CTS do not occur after acute compression, and so this model does not exactly simulate the chronic and repetitive increased pressure of idiopathic carpal tunnel syndrome.257 A single pathologic study of the nerve in a patient with CTS showed demyelination, remyelination, and loss of large myelinated fibers. The nerve was swollen proximal to the ligament.248 Swelling of the nerve proximal to the tunnel has also been noted on imaging studies and may in part be related to obstruction of axoplasmic flow. It is clear that small fibers are also affected, because pain is a common symptom and disturbed autonomic function can also be demonstrated. It is not clear why female gender is a risk factor, but wrist size and poorly understood hormonal factors are issues that need further study. With increasing age, mild wear-and-tear synovial thickening may increase pressure within the carpal tunnel. It is likely that obesity also increases the risk of CTS by increasing carpal tunnel pressures, but whether it exerts its effect by increasing interstitial fluid pressure or by some other mechanism is not clear. In a series of 145 cases that had biopsy of the tendon sheaths or transverse carpal ligament, 57% showed no significant pathologic change. Nonspecific tendonitis was found in 31%, rheumatoid synovitis in 10%, and hyaline fibrosis in less than 1% of cases.45 Similar results of biopsies in idiopathic CTS were reported from Japan. A total of 73.4% of the ligament and 56% of the tenosynovial biopsies were normal. Abnormal findings included mucoid change, amyloid deposits, inflammation, chondrometaplasia, and vascular hypertrophy. The majority of the abnormal findings were mild and probably of little

1444

Diseases of the Peripheral Nervous System

clinical significance.167,195 Amyloid deposition is usually nonspecific, but occasional cases of familial amyloidosis will be discovered.169 In summary, there are no specific histologic changes in patients with idiopathic CTS that provide information on pathogenesis. Conditions Associated with CTS Many conditions have been associated with the development of CTS for reasons that are easily understood. These factors include inflammation of the tendons, wrist trauma, arthritis, infection, edema, mass lesions, and anomalies, all of which reduce the space in the carpal tunnel. Other factors, such as the effect of menopause, occupational factors, and computer use, are controversial and require more study. The major causes of symptomatic CTS are discussed below, beginning with the more frequent disorders (Table 59–5). Trauma. Trauma, usually a Colles’ fracture and sometimes a carpal bone fracture or dislocation, is a common factor associated with the development of CTS.238 CTS can develop acutely after the injury and require emergent surgery.151 Acute CTS is often associated with conduction

block that can be identified by stimulating the median nerve proximal and distal to the carpal tunnel.86 The fracture effects may be seen 20 years later, but in most cases CTS symptoms begin within weeks to several years after the injury. Blunt trauma to the wrist and operative procedures or trauma involving the hand distal to the wrist with hand swelling are less frequent causes. Another common injury is a wrist sprain usually resulting from a fall on the outstretched hand. Hyperextension of the wrist from bracing on the steering wheel or dashboard of an automobile during an accident can lead to symptoms of CTS within a few days. Conservative treatment of these cases leads to improvement in most, but surgery may be needed in 30%.92,129 Blunt trauma to the wrist, operative procedures, and trauma involving the hand distal to the carpal tunnel with swelling or infection can also lead to symptoms.26 A few cases have been reported after radial artery puncture and catheterization.147 Endocrine Disorders DIABETES MELLITUS. Diabetes mellitus is a significant risk factor for the development of CTS. The standardized

Table 59–5. Conditions Associated with the Development of CTS Trauma Colles’ fracture Wrist bone fracture-dislocation Wrist sprain Blunt trauma to wrist Hand trauma, surgery, and infection Radial artery puncture Endocrine Diabetes mellitus Hypothyroidism Acromegaly Menopause and hormone replacement therapy Rheumatoid Arthritis and other Inflammatory Diseases Pregnancy Occupational Trauma Occupational exposure at work Unaccustomed overuse Nonspecific Tenosynovitis Degenerative Arthritis of the Wrist Tumors and other Mass Lesions Ganglion cyst Neurofibroma/schwannoma Lipoma Lipofibroma Cavernous hemangioma Localized hypertrophic neuropathy Chondroma T-cell lymphoma

Amyloidosis Hereditary amyloidosis Multiple myeloma Beta2–microglobulin in dialysis patients Infection Bacterial Mycobacterial — M. tuberculosis, M. leprae, and other species Fungal Viral and parasitic infections Borrelia (Lyme disease agent) Human immunodeficiency virus Inherited Familial autosomal dominant Hereditary neuropathy with liability to pressure palsies Mucopolysaccharidosis Anomalous Carpal Tunnel Structures Muscles/tendons in the carpal tunnel Thrombosis or aneurysm of persistent median artery Anomalous distal end of the radius Miscellaneous Gouty tophi Sarcoidosis Paget’s disease Hemophilia, anticoagulant therapy

Median Neuropathy

morbidity ratio (the ratio of the observed number of CTS patients with diabetes to the expected number) is 2.5 for men and 2.2 for women.238 A hospital-based case-control study also noted an increased frequency of CTS in those with a history of diabetes.55 Electrophysiologic abnormalities consistent with CTS but without symptoms (subclinical CTS) occur in approximately one fourth of diabetic patients, while symptomatic CTS is found in only 7.7%.60 The electrodiagnostic diagnosis of CTS in patients with a diabetic neuropathy may require the use of segmental nerve conduction study techniques because routine antidromic nerve conduction studies do not appear to be very sensitive.197 The reason CTS is so common in patients with diabetes is unknown; however, CTS is more common in patients with neuropathy from a variety of causes. Aggravation of ischemia in nerves already stressed by chronic endoneurial hypoxia may be a factor.139 There may be alterations in the connective tissue sheaths of the nerve or in the transverse carpal ligament.59 HYPOTHYROIDISM/HYPERTHYROIDISM. Hypothyroidism was a common cause of CTS before there were rapid laboratory methods of diagnosis.202 Thyroid function is now routinely tested and hypothyroid states are treated early in many cases, so that CTS resulting from myxedema is seen much less frequently. In newly diagnosed hypothyroid patients, a recent series found that 29% had CTS.58 None of the hyperthyroid patients in that study had CTS, but a prospective study of patients with hyperthyroidism found CTS in 5% of these patients when hyperthyroidism was first diagnosed.212 The symptoms of CTS remitted after treatment of the hyperthyroidism. There is disagreement as to whether patients on thyroid hormone replacement therapy continue to have an increased risk of CTS.190,230

1445

of industrial workers found no relationship between CTS and hysterectomy, oophorectomy, menopause, and the use of oral contraceptives.211 Pregnancy. CTS occurs at least 2.5 times more often among pregnant women than among nonpregnant women of the same age. The incidence of symptoms increases with each succeeding trimester, probably as a result of edema and weight gain. A prospective study evaluating women in the final stages of pregnancy found clinically diagnosed CTS in 62% and abnormal nerve conduction studies in 43%. The Italian CTS Study Group also found that smoking and alcohol consumption appeared to be associated with the development of CTS.186 Approximately half of the women with CTS during pregnancy had symptoms a year later.187 A study has looked for factors that were present in patients who had eventually needed surgery either during or after pregnancy when conservative measures failed. These patients (1) had either started having CTS symptoms during the first two trimesters or had a history of CTS symptoms, and (2) had both a positive Phalen’s maneuver within less than 30 seconds and abnormal two-point discrimination at the fingertips (⬎6 mm).233

ACROMEGALY. Acromegaly is the result of chronic growth hormone excess caused by a pituitary adenoma. Approximately one third of cases develop CTS.78,180 The associated CTS usually resolves when the acromegaly is treated. Water retention with edema of the synovial tissues is thought to be a factor. An MRI study of patients with acromegaly found that the predominant pathology seemed to be increased edema of the median nerve within the carpal tunnel rather than extrinsic compression from increased volume of the carpal tunnel contents.111

Rheumatoid Arthritis and Other Inflammatory Conditions. Rheumatoid arthritis is a very common cause of CTS. It occurs with a standardized morbidity ratio (an index of the association with CTS) of 3.5 for men and 3.9 for women, a significant association. Tenosynovitis, arthritis, wrist deformity, and nodules all contribute to canal narrowing. Other connective tissue diseases, such as lupus erythematosus and Sjögren’s syndrome, are much less frequently associated with CTS.156 Temporal arteritis and polymyalgia rheumatica are other inflammatory conditions linked to CTS.1,218 These patients can develop distal extremity swelling with pitting edema, wrist joint inflammation, and synovitis. Ultrasound evidence of tenosynovitis of the flexor tendons in the carpal tunnel has been found in more than one third of patients with polymyalgia rheumatica.65 Eosinophilic fasciitis has been linked to CTS.113,131 Granuloma annulare, a noninfectious necrobiotic granulomatous reaction of unknown cause, is a rare cause of CTS.260

Menopause and Hormone Replacement Therapy. The association of CTS with ovarian function is of considerable interest because of the peak in incidence in women of perimenopausal age. One study has found a higher incidence of CTS in perimenopausal women, but the issue requires further study.51 The use of hormone replacement therapy appears to be associated with an increased rate of carpal tunnel release.230 Several reports have described an association between CTS and bilateral oophorectomy, but this association remains questionable.21,31,194 A small study

Nonspecific Tendonitis and Degenerative Arthritis. Nonspecific tenosynovitis is a common cause of CTS. The severity varies from minor wear-and-tear thickening to inflammation severe enough to cause swelling proximal to the wrist and difficulty making a fist. Interestingly, a history of other nonspecific inflammatory disorders affecting the hand is reported by 47% of patients with nonspecific tenosynovitis of the wrist and by 20% of all patients with CTS within 1 to 2 years of the development of symptoms.238 These include trigger finger, tennis elbow,

1446

Diseases of the Peripheral Nervous System

de Quervain’s tenosynovitis, Dupuytren’s contracture, tenosynovitis of the forearm or hand, and nonspecific synovitis of the finger joints. Osteoarthritis of the wrist likely contributes to the development of CTS in some people by narrowing the carpal tunnel, but the significance of minor arthritic changes commonly seen on plain films of older individuals with CTS is unclear. Occupational Factors. The development of CTS as a result of overuse the upper extremities has been documented for many years but remains a controversial area, with some studies suggesting a link and others disputing any relationship. The association of CTS with occupations that involve repetitive hand motion, high force, various biomechanical factors, and long-term exposure to vibration is accepted by most researchers.66,137,242 Workers’ compensation and financial gain can encourage claims for CTS and may delay recovery after treatment, making it difficult to assess an association with occupation.99,221 Employers have sought ways to screen prospective employees for CTS and have made workplace changes to avoid stress on the wrists because of the high costs of treatment and disability. A permanent dropout from work in more than 1 in 10 patients after surgery indicates that CTS is a substantial socioeconomic burden on society.18 Workers with recent onset of symptoms who are able to change jobs may improve and avoid surgery in many cases.86 The list of occupations thought to cause CTS is extensive (Table 59–6). They include seafood processing and canning72; heavy manual labor such as carpentry, roofing, drywalling, boat building, loading tires and packages, production dressmaking, and garment work9; carpet work117,128,201; work as a nurse anesthetist,54 dental hygienist,132 grocery checker,16,146 or meat packer/butcher63; poultry processing10; electronic and other types of assembly work66; packing food137; repetitive heavy manual labor143,214; and apprentice construction work.213 Vibrating power tools that cause symptoms include the jackhammer, air gun, pneumatic stapling gun, chain saw, and power sander. Symptoms can develop acutely when “weekend warriors” tackle home projects that involve unaccustomed manual labor, such as home building or remodeling, using a chain saw, sanding, painting, and strenuous yard work. The symptoms usually subside after a few weeks or months after the project. Paraplegia is associated with an increased risk of CTS because of the need to transfer, lift, and propel a wheelchair with the wrists in extension.77 Golfers can develop tenosynovitis from repetitive grasping and wrist motion and from compression of the nerve just distal to the carpal tunnel.103 In marked contrast to these reports, a study of median sensory nerve conduction in those working in a steel mill, meat/food packaging, and electronics and plastics industries showed no significant increase in nerve conduction

Table 59–6. Occupations Associated with Carpal Tunnel Syndrome Repetitive Motion Grocery checkers Meat packing, poultry processing, butchers Seafood processing and canning Product dressmaking, garment workers Assembly Repetitive heavy manual labor Packing various products Dental hygienists Nurse anesthetists Carpet workers Heavy manual labor Vibrating Power Tools Jackhammer Chain saw Stapling gun Power sander Acute after Unaccustomed Labor Home remodeling Painting Cleaning Sanding Chain saw Strenuous yard work

abnormalities or clinical diagnosis of CTS over a period of 11 years. This careful study did not support a link between occupational hand use and CTS. Nerve conduction latencies tended to increase with increasing age, but the prevalence of symptoms did not increase despite many years of occupational exposure.172 Another study followed workers for 70 months. Preemployment nerve conduction study screening showed that the majority of workers with a prolonged median sensory distal latency and no symptoms of CTS did not become symptomatic. In addition, changes in median nerve conduction over time did not correlate well with the development of symptoms.258 CTS IN COMPUTER USERS. It is commonly held that long hours working at a computer lead to the development of CTS, but there is little evidence to support that conclusion. In a study of computer users at a medical facility, no apparent increase in the incidence of CTS was found.241 The prevalence in the computer users was comparable to the estimates of the incidence of CTS in the general population (see discussion of the incidence of CTS above). Studies of American and Japanese workers found no relationship between computer keyboard use and CTS; in fact, avocational keyboard use seemed to have a protective effect.173,174 Computer users commonly experience a variety of other upper body complaints causing discomfort. Some of

Median Neuropathy

the disorders and clinical findings in computer users include postural misalignment, cervical radiculopathy, sympathetic dysfunction, complex regional pain syndromes, radial and cubital tunnel syndromes, shoulder impingement, medial and lateral epicondylitis, and peripheral muscle weakness.192 Malposition of the wrist with dorsiflexion and ulnar deviation can lead to pain from increased forearm muscle workload. Prolonged use of the mouse can increase workload on the fingers and increase carpal tunnel pressures as well.120,193 Amyloidosis. The development of a peripheral neuropathy and CTS may be a clue to the diagnosis of amyloidosis. The classification of amyloidosis is based on the nature of the precursor plasma proteins that form the fibril deposits. These proteins are diverse, but all produce amyloid deposits with a common ␤-fibrillar structure.64 Amyloid light chain (AL) disease occurs in primary systemic amyloidosis, multiple myeloma, and Waldenström’s macroglobulinemia. Primary systemic amyloidosis is by far the most common cause of systemic amyloidosis in the United States. A diagnostic fat aspirate for amyloidosis is found most often in cases of peripheral neuropathy with a monoclonal protein and clinical findings such as gastrointestinal disturbance, weight loss, autonomic disturbance, and CTS.8 Nephrotic syndrome, thickening of the ventricular walls on echocardiography, congestive heart failure, macroglossia, and sensorimotor and autonomic neuropathy are common findings.175 Symptoms of CTS are frequent and may precede other features of the disease by a year or more. Familial amyloidosis is inherited as an autosomal dominant disease resulting from genetic variants of physiologic proteins, including transthyretin and, much more rarely, apolipoprotein A-I, lysozyme, fibrinogen, gelsolin, amyloid-[␤], and cystatin C.104 There is considerable clinical overlap with AL amyloidosis. Patients typically present with polyneuropathy, CTS, autonomic insufficiency, and cardiomyopathy and gastrointestinal features, occasionally accompanied by vitreous opacities and renal insufficiency. When carpal tunnel surgery is performed, fat and ligament should be examined for amyloid. Beta2-microglobulin–derived amyloid deposits in carpal tunnel tissues are very common in patients on long-term hemodialysis.122 A key factor in the pathogenesis is the uremic retention of the precursor molecule, beta2-microglobulin.71 In addition, patients on long-term hemodialysis frequently have an underlying peripheral neuropathy. Compared with idiopathic CTS, these patients may not have as satisfactory postoperative improvement.98 Forearm arteriovenous fistulas for hemodialysis can be associated with distal nerve ischemia by causing vascular steal phenomenon or increased venous pressure.148

1447

Others, however, have found no precise relationship between the side of the CTS and the side of the arteriovenous fistula.4 Amyloid deposits localized to the tenosynovium and without evidence of systemic disease are sometimes discovered at the time of carpal tunnel surgery but are of uncertain pathogenic significance. Anomalous Carpal Tunnel Structures. Anomalous structures and hypertrophied muscles in the carpal tunnel can narrow the passageway. The lumbrical muscles can have a proximal insertion in the carpal tunnel and may be hypertrophied in patients whose jobs require repetitive hand motion.226 Muscle bellies of the flexor digitorum superficialis longus, profundus, and palmaris longus can extend into the canal.210,220,229 Bony anomalies of the distal end of the radius can also cause symptoms. Rarely an unusually large patent persistent median artery or thrombosis or aneurysm of the artery is encountered.251 Tumors and Other Mass Lesions. Tumors and other mass lesions are occasionally encountered as a cause of CTS. Ganglion cysts of the carpal tunnel are seen most often.43,165 Other lesions are the subject of single case reports or a small series. These lesions in the carpal tunnel include lipomas,126 cavernous hemangiomas,196 malignant nerve sheath tumor,179 perineurioma,3,154 intraneural lipofibroma,29 T-cell lymphoma,37 and chondroma.100 Infections. A variety of infectious agents are occasionally responsible for CTS, including bacteria, viruses, parasites, mycobacteria, and fungi.27,28,149 More recently added to the list of infectious agents are Borrelia (the agent of Lyme borreliosis)95 and human immunodeficiency virus.145 Miscellaneous Conditions. There are a host of other conditions that can lead to the development of CTS. Gouty tophi in the floor of the carpal tunnel, carpal bones, radiocarpal joint, and tendon sheaths of the wrist can be responsible and may be revealed by CT and MRI imaging.36,41 Pseudogout may cause acute CTS that requires surgery.38 Acute CTS can occur in anticoagulated patients and those with hemophilia, not always after trauma.20,157 CTS is more common than expected in sarcoidosis.176 Fluid retention causes the syndrome in congestive heart failure.11 Paget’s disease of the wrist may rarely cause symptoms.56 An increased incidence of median nerve entrapment occurs in Fabry’s disease.142 Familial CTS. Familial cases, often showing an autosomal dominant inheritance, have been reported.144,243 A familial predisposition to CTS may be much more important than previously suspected. In a study of 84 patients having undergone prior CTS surgery, a family history of CTS or symptoms suggestive of CTS were found in 39.3%, as opposed to only 3% of 100 patients seen for complaints of

1448

Diseases of the Peripheral Nervous System

low back or neck pain.203 Close scrutiny of possible familial cases suggests, however, that familial cases are quite rare.88 These cases also differ from sporadic CTS because of earlier onset and increased bilateral involvement. A study of monozygotic and dizygotic twins produced a heritability estimate of 0.46, suggesting that up to half of the liability to CTS in women is genetically determined. Surprisingly, the estimate was unchanged after taking into account age, body mass index, physical activities, and hormonal factors. There was some increased risk in the peri- and postmenopausal groups.94 Development of CTS along with current or prior mononeuropathies affecting other nerves and an underlying mild peripheral neuropathy suggests hereditary neuropathy with liability to pressure palsies (HNPP). Screening for HNPP in patients with isolated CTS is, however, unlikely to be rewarding. The prevalence of electrophysiologic CTS is thought to be high in individuals with Down syndrome.39 Diagnosis of CTS in children with mucopolysaccharidoses and mucolipidoses is more challenging. Symptoms are rare, but signs such as decreased sweating, pulp atrophy, thenar wasting, and manual clumsiness are much more common than in adults. Operative findings are thickening of the flexor retinaculum and a mass of shiny white tenosynovium engulfing the flexor tendons.93 Hurler’s/Hurler-Scheie and Hunter’s syndrome patients make up the largest group of affected patients. All patients over the age of 2 years will likely have CTS. CTS in Children. CTS in children is rare and, when it occurs, is usually related to an underlying condition such as a mucopolysaccharidosis. A variety of other rare disorders can be responsible (Table 59–7).6,44 The second Table 59–7. Etiology of Carpal Tunnel Syndrome in Children and Adolescents Mucopolysaccharidosis, mucolipidosis Trauma Tenosynovitis Idiopathic Familial Hereditary liability to pressure palsies Anomalous muscle bellies Tumors and other mass lesions Dejerine-Sottas neuropathy Schwartz-Jampel syndrome Weill-Marchesani syndrome Léri’s pleonosteosis Melorheostosis Hemophilia Macrodactyly Athletic activities Modified from Al-Qattan, M. M., Thomson, H. G., and Clarke, H. M.: Carpal tunnel syndrome in children and adolescents with no history of trauma. J. Hand Surg. Br. 21:108, 1996.

most common group is idiopathic CTS.133 Familial cases can present at a young age and may be the result of thickening of the transverse carpal ligament.246,252 Motor symptoms rather than paresthesias may be the presenting problem in children. Mass lesions in the carpal tunnel include lipofibromatous hamartoma, neurofibromas in patients with neurofibromatosis, and hemangiomas. Older children are reported to develop symptoms after athletic activities such as weight lifting, golf, and basketball.

Differential Diagnosis The differential diagnosis of CTS is listed in Table 59–8. It is not uncommon for patients to have a carpal tunnel release or even several wrist operations for symptoms caused by disease at a more proximal level.261 Incomplete or inadequate nerve conduction studies and needle electrode examination and misinterpretation of the data lead to an erroneous electrophysiologic diagnosis of CTS. Patients with amyotrophic lateral sclerosis with onset of weakness and atrophy in the small hand muscles may have a low amplitude and delayed median motor response. Unless median sensory nerve conduction studies are performed (showing normal responses), a diagnosis of CTS can be made. Patients with C6 and C7 radiculopathies have sensory symptoms in the median nerve distribution that can be confused with CTS. Because both problems are common, they not infrequently coexist. The high incidence of cervical radiculopathy in patients with CTS underscores the importance of the needle electrode examination in patients with CTS. There is no evidence, however, that a double crush of median nerve fibers predisposes to CTS.162,209 Patients with abnormal median motor nerve conduction studies but normal sensory or mixed nerve conduction studies usually do not have CTS. When this electrophysiologic finding is encountered, the electrodiagnostic medical

Table 59–8. Differential Diagnosis of Carpal Tunnel Syndrome Cervical radiculopathy Peripheral neuropathy Ulnar neuropathy Amyotrophic lateral sclerosis Syringomyelia Multiple sclerosis Spondylotic myelopathy Thoracic outlet syndrome Pronator syndrome Osteoarthritis of the metacarpal phalangeal joint of the thumb Mononeuritis multiplex Multifocal motor neuropathy with conduction block Wrist and hand pain of diverse causes

Median Neuropathy

consultant should look carefully for anterior horn cell disease and other spinal cord lesions. Some patients with CTS may have slowing of median motor conduction in the forearm. This tends to occur in those with more prolonged median motor distal latencies. While forearm slowing might raise the issue of a pronator syndrome, this is rarely the case. Lesions of the nerve in the distal forearm that can have symptoms mimicking CTS include transient forearm conduction block possibly caused by focal transient ischemia, anomalous forearm tendons and muscle bellies, vascular anomalies, and other lesions.254 Ulnar neuropathies are usually easy to distinguish from CTS on clinical grounds, but the symptoms of tingling in the fingers and hand weakness sometimes lead to referral to the electromyography laboratory as possible cases of CTS. Hand paresthesias are frequent in patients with peripheral neuropathy, and clinically it may be difficult to be sure if there is a superimposed CTS. Segmental median nerve conduction studies can detect slowing that is localized to the carpal tunnel, whereas ulnar nerve conduction is diffusely slowed in the palm-to-wrist and wrist-to-finger segments. Patients are not infrequently referred for electrodiagnostic study for pain in the wrist and hand without any definite sensory complaints. CTS rarely occurs without paresthesias. Usually these patients have overuse syndromes of the hand, local arthritis, or de Quervain’s tenosynovitis, or there is no clear explanation of why they are having the discomfort.

Laboratory Diagnosis Electromyography Experienced clinicians can make a diagnosis of CTS with a high degree of certainty in most cases. As noted above, the severity of the median nerve compression may be difficult to determine on clinical grounds. There are patients with mild or vaguely described hand paresthesias who surprisingly are found to have severe CTS by nerve conduction studies. Conversely, patients may have severe symptoms with only mild median nerve conduction changes. Consideration of the severity of the electrophysiologic abnormalities is useful, along with the severity of the patient’s symptoms, when making a treatment decision. Comparison with a preoperative electromyogram is helpful when the symptoms are not relieved as expected and the question of recurrent CTS arises. Electromyography is especially valuable in distinguishing cervical cord lesions, C6 and C7 radiculopathies, thoracic outlet syndrome, brachial plexus lesions, and proximal median neuropathies from CTS. Nerve conduction studies can determine if hand paresthesias in a patient with peripheral neuropathy are due largely or in part to a focal median neuropathy. Nerve conduction study abnormalities are well known, and practice parameters have been published.107,237 The median motor distal latency is prolonged, and there may

1449

be a reduction in the amplitude of the compound action potential of the thenar muscles in more advanced cases. The median antidromic sensory distal latency recording from the index or middle fingers is more often abnormal than standard motor distal latencies. Even more sensitive sensory nerve conduction techniques use segmental conduction across the carpal tunnel, such as median stimulation in the midpalm, or a comparison of the median and ulnar latencies on stimulation of the ring finger or of median sensory latency with superficial radial latency. Prolonged median latencies with palmar stimulation are found in 75% to 90% of cases.108 It is important to include ulnar nerve conduction studies to be certain that the abnormalities are confined to the median nerve and are not part of a generalized polyneuropathy. Needle electrode examination of the thenar muscles is not as useful as studies of nerve conduction but does provide a measure of the severity of axonal degeneration affecting thenar motor fibers. It is also important to look for evidence of denervation in muscles supplied by the ulnar nerve and muscles of the forearm innervated by the median nerve to avoid errors. Equivocal changes on the nerve conduction studies may suggest the need for a more extensive sampling of muscles supplied by C6, C7, and C8. Ultrasound Reports on the diagnosis of CTS by ultrasound have appeared, and the literature has recently been reviewed.17 In normal subjects, the median nerve appears hypoechoic and elliptical in cross section. It changes in shape from ovoid proximally to wedgelike distally. It has a greater cross-sectional area in the dominant hand without a correspondingly larger carpal tunnel. Patients with moderate to severe CTS have a markedly increased cross-sectional area and frequently have marked dilatation proximal to the carpal tunnel, with a notch in the nerve as it enters the carpal tunnel.134 There is, however, considerable overlap in the caliber of the median nerve between patients with mild CTS and normal subjects.262 The median nerve has somewhat of an hourglass or “head-neck-body” configuration as it passes through the carpal tunnel. The largest cross-sectional areas are at the wrist crease and the distal edge of the flexor retinaculum, whereas the smallest area, the neck, is at the level of the hook of the hamate.170 Measurement of the median nerve cross-sectional area proximal to, distal to, and in the middle of the carpal tunnel in a series of idiopathic cases gave a sensitivity of 67% and specificity of 97% compared with control subjects.168 Compared with MRI, ultrasound is more available, takes less time, and is less expensive. Advantages of ultrasound compared with electromyography include the lack of patient discomfort and the ability to detect tenosynovitis, ganglion cysts, and other mass lesions within the carpal tunnel. Ultrasound can also confirm early duplication of the median nerve in the carpal tunnel and a persistent

1450

Diseases of the Peripheral Nervous System

median artery that is sometimes associated with high division of the median nerve or a bifid nerve configuration in the carpal tunnel.76,106 The acute onset of CTS symptoms can be the result of thrombosis of a persistent median artery.74 Excision of a large unthrombosed persistent median artery may not be advisable because it may contribute significantly to the blood supply of the hand. It has been reported that the thickness of the transverse carpal ligament increases slowly in patients undergoing chronic hemodialysis and that this may be related to the development of CTS.247 Sonography may be helpful for the diagnosis of dialysis-related amyloidosis. Amyloid infiltration of the tendons causes the tendons to become thickened and hypoechoic.134 Magnetic Resonance Imaging MRI of the carpal tunnel is not widely used for the diagnosis of CTS because of the cost of the examination and because the MRI is not always abnormal. Abnormalities visualized include swelling of the nerve proximal to the carpal tunnel and flattening within the tunnel leading to an increased swelling-to-flattening ratio, similar to what is found with ultrasound.119 Bowing of the flexor retinaculum is less specific. The abnormal nerve can become hyperintense, and sometimes a fascicular pattern can be seen. The nerve is frequently most compressed at the level of the hook of the hamate, where the cross-sectional area of the tunnel is smallest.109 Patients with CTS have a median nerve cross-sectional area that is approximately 50% larger within and proximal to the carpal tunnel than that of asymptomatic subjects. A recent study compared MRI findings in CTS and control subjects with a gold standard for CTS that included a hand symptom diagram and abnormal median and ulnar palmar nerve conduction

studies.110 The reliability of MRI was high, but diagnostic accuracy was moderate compared with the gold standard for diagnosis. MRI abnormalities with the greatest sensitivity were an overall impression of abnormality at 96% and hyperintense nerve signal at 91%. The nerves were larger at the level of the distal radioulnar joint and the pisiform bone in patients with CTS as a result of proximal nerve swelling. The cross-sectional area of the carpal tunnel was borderline smaller in patients with CTS, but the tapemeasured wrist circumference has little value in predicting the presence of CTS. The length of abnormal median nerve signal correlated well with abnormal median nerve conduction studies. The use of MRI in patients with uncomplicated CTS will remain limited because routine nerve conduction study specificity is very high (especially in laboratory-based series) compared with MRI and because MRI is expensive.70 Patients with tenosynovitis can have distention and fluid within the tendon sheaths and abnormal signal intensity within the tendons. Ganglion cysts, tumors, and bony lesions are easily identified by MRI. There may be a role for MRI in certain patients with nondiagnostic nerve conduction studies and atypical clinical presentations, but this has not been well studied. After successful surgery, there is a return toward normal signal and size of the nerve proximal to the carpal tunnel, and an increase in cross-sectional area and an increase in nerve signal in the distal carpal tunnel.46 The proximal nerve swelling and change in signal intensity have been attributed to proximal spread of edema and accumulation of axoplasm because of obstructed axoplasmic flow. MRI can be helpful in those patients with symptoms unrelieved by surgery by revealing an intact distal transverse carpal ligament and, less often, scarring around the nerve (Figs. 59–6 and 59–7). A second surgery to section the intact remnant of

FIGURE 59–6 MRI of a normal carpal tunnel. Fast-spin echo T2weighted images with fat suppression. Upper arrows indicate the transverse carpal ligament. Lower arrow points to the median nerve. (Courtesy of Dr. K. K. Amrami, Department of Radiology, Mayo Clinic, Rochester, MN.)

Median Neuropathy

1451

FIGURE 59–7 MRI of a carpal tunnel after failed carpal tunnel release. Fast-spin echo T2-weighted images with fat suppression. Upper arrows indicate the intact transverse carpal ligament. Lower arrow points to the flattened median curve. (Courtesy of Dr. K. K. Amrami, Department of Radiology, Mayo Clinic, Rochester, MN.)

ligament may lead to relief of symptoms.164,235 MRI can identify iatrogenic nerve damage such as section of the radial branches after endoscopic carpal tunnel release.227

Medical Treatment Because CTS may be precipitated by various activities involving excessive use of the hands, improvement may result from stopping the activity. Treatment of an underlying systemic disease such as myxedema, rheumatoid arthritis, and acromegaly may be helpful. Nonsteroidal anti-inflammatory medications can be tried when inflammation or synovitis is present, but there is not much evidence that nonsteroidal anti-inflammatory drugs or diuretics are helpful for treatment of idiopathic CTS.35 In a majority of patients with CTS, immobilization of the wrist in a neutral position with a posterior splint is very helpful for those who are awakened by pain and numbness during the night. These can be purchased off the shelf, but usually a custom-fitted splint is more comfortable. In some cases prolonged motor and sensory distal latencies may improve.34 Instructing patients to wear the splints during the day as well results in superior distal latency improvement.253 Although some patients successfully wear splints on and off for years as necessary, most patients eventually need additional treatment. Ultrasound treatment over the carpal tunnel has been suggested because of possible effects on blood flow and inflammation. One study noted improvement in symptoms and nerve conduction study parameters. A 68% favorable response rate was noted after 6 months of followup.61 A second study found that ultrasound and placebo ultrasound had similar effects in providing symptomatic relief.185 Ultrasound therapy seems to facilitate improvement in experimental acute median neuropathy in the rabbit.189 Injection of steroids into the carpal tunnel is an effective treatment for relief of CTS symptoms. It provides temporary relief for most patients and can give long-

lasting relief.48 It is often considered prior to surgical decompression, and in some cases with vague or unusual symptoms a response is helpful for diagnosis. Steroid injection can provide relief in pregnant patients or those with myxedema, in whom the symptoms can be expected to resolve after delivery and hormone replacement. There are occasional patients who refuse surgery who might consider steroid injection. Patients with advanced CTS should have surgery rather than steroid injection. In one study, 84% of patients had complete or partial remission of their symptoms 1 year after injection. Motor and sensory nerve conduction studies were improved in 63%. Those with mild symptoms present for less than 1 year and age 40 years and younger responded best.152 Another study found 92% of nerves benefited but relief was less sustained. Approximately 50% of the nerves became worse within 6 months and 90% within 18 months. Only 8% of the nerves remained improved at 2 years’ followup.83 Steroid injection into the carpal tunnel is more effective than low-dose, short-term oral steroid therapy for the relief of CTS.263 Short-term oral steroids provide persisting improvement in 35% to 50% of patients when assessed after 1 year. Failure of splinting and local injection of corticosteroids is more likely in those with long duration of symptoms, older age, permanent paresthesia, two-point discrimination threshold above 6 mm, positive Phalen’s maneuver within 30 seconds, and long motor and sensory distal latency or absent sensory nerve action potential.15,234 Complications are few, but intraneural injection of the steroids is a risk.152 Steroid injections were performed in 492 hands of patients in a population study of CTS (J. C. Stevens, unpublished data, 1988). Seventyseven percent had one injection, 16% had two, 5% had three, and 2% had four. An operation was subsequently required in 49% of hands. In summary, nonoperative treatment can be effective in the long term for many patients. It has the advantage of costing less than surgery and avoids risks associated with the procedure.183

1452

Diseases of the Peripheral Nervous System

Surgical Treatment Patients who fail conservative measures, including modification of their physical activities and treatment of associated medical disorders, should be offered surgical treatment. Approximately 50% of patients will end up having surgery. A minority view is that patients with symptomatic idiopathic CTS and abnormal median nerve conduction studies should be offered surgery early on because conservative treatment will only delay definitive treatment.259 Good or excellent results are obtained in 85% to 95% of patients. With few exceptions, surgery provides rapid relief of nocturnal paresthesia, pain, and episodic symptoms during the day. Patients with severe CTS also obtain relief of pain and tingling after surgery, although atrophy often persists and sensation may remain impaired.12 Older age is not a barrier to surgery because outcome in severely affected patients is unrelated to age.250 At the other end of the spectrum, it has been found that patients with normal median nerve conduction studies, carefully selected clinically, can expect a good or excellent outcome in 90% of cases.123 Predictors of less-than-favorable outcomes include measures of upper extremity functional status, mental health status, health habits, and attorney involvement.116 Workers’ compensation cases in which carrier denial of authorization for surgery was followed by authorization after intervention by a plaintiff’s attorney were much less likely and much slower to return to light-duty and to return to full-duty work than were uncontested cases.182 There is a suggestion that relief of symptoms in workers exposed to chronic vibration may not be as good as relief in those with idiopathic CTS. A possible explanation for this finding is the fact that vibration exposure may cause more diffuse pathologic changes in the distal median nerve.47 Endoscopic release of the carpal tunnel is effective in uncomplicated cases, with earlier return to work than among those who underwent the traditional open operation. Open release is technically less demanding and has a lower risk of complications.79 Complications of surgery include damage to the recurrent motor branch or digital and palmar cutaneous nerves, painful or sensitive scar, limited motion of the wrist, pain and tenderness of the wrist and hand, infections, hematoma, keloid formation, reflex sympathetic dystrophy, laceration of the superficial palmar arch, tendon laceration, and incomplete ligament release.105,127 Relationship between Surgical Outcome and Preoperative Clinical and EMG Severity There is disagreement as to whether preoperative EMG findings are related to surgical outcome.33,138 Most patients do well irrespective of their electrophysiologic findings. The duration of CTS symptoms does not correlate well with surgical outcome. Even patients with severe CTS and no motor or sensory response on nerve conduction studies

report absence of pain and disappearance or reduction of paresthesia.161 At follow-up 6 months after surgery, some low-amplitude motor and sensory responses may return. Muscle atrophy usually remains unchanged. Patients with longer distal motor and sensory latencies and good general health have better outcomes after surgical treatment.52 Perhaps patients with long latencies have more segmental demyelination and less axonal degeneration than those with shorter latencies. Median Nerve Conduction Studies after Surgery Several investigators have found an improvement in nerve conduction studies within 1 hour after surgery.62,102 Most patients will experience improvement of painful symptoms within 1 or 2 days after operation, which may be the result of a reduction in spontaneous activity generated by the compressed nerve segment. Relief of ischemia could be responsible for the rapid reduction in distal latencies and paresthesia.223 Recovery from conduction block would be expected to take some time, and repair of axonal degeneration even longer. The maximum improvement in nerve conduction occurs during the first 6 weeks after operation and is slower thereafter. Prolonged latencies do not always return to normal when symptoms are relieved, even when the study is done 1 year or more later. The neurophysiologic recovery after surgery is similar in elderly and younger patients.249 Recurrent Paresthesias in Patients Who Have Had Surgery Patients sometimes experience hand paresthesias months or years after an apparently successful carpal tunnel release. This may be due to recurrent CTS, but often another process such as a cervical radiculopathy or a cerebral infarct is responsible. In these cases, normal median sensory nerve conduction studies make CTS unlikely. When the median nerve conduction studies are abnormal, availability of preoperative nerve conduction studies is helpful for comparison. If preoperative EMG studies were not performed or if the results are unavailable, it will be difficult to know if prolonged latencies are the result of recurrent or residual CTS. Severely prolonged median latencies favor recurrent median nerve compression. A repeat nerve conduction study in 3 to 6 months can be suggested if the symptoms are worsening. MRI of the wrist can be performed to see if there is incomplete division of the transverse carpal ligament, scarring, or some other process affecting the median nerve.

MEDIAN NERVE LESIONS IN THE ARM A variety of other lesions can affect the proximal median nerve, including pseudoaneurysms of the axillary or brachial artery, hemodialysis fistulas, hematoma compressing the

1453

Median Neuropathy

nerve in anticoagulated patients, malposition of the arm on the operating table, misplaced intravenous lines, and needle trauma from attempted blood draws.89 The nerve can also be compressed during use of sleep-deepening drugs and alcohol intoxication.215 Tourniquet paralysis may also involve the median nerve.24,97

MEDIAN NERVE ENTRAPMENT BY THE LIGAMENT OF STRUTHERS Median nerve entrapment above the elbow by the ligament of Struthers is very rare. It is associated with a bony spur, the supracondylar process, present on the anteromedial side of the distal humerus in 1% to 3% of the population. The spur is pyramid shaped, bony or cartilaginous, and often bilateral, and is usually an incidental finding. Struthers’ ligament extends from this small bony process to the medial epicondyle. The brachial artery and median nerve pass beneath the spur and ligament and can be compressed. The bony prominence may be palpable, and pressure over the area may aggravate or provoke the symptoms. A few cases have been reported in which compression was caused by the ligament in the absence of the spur.245 Cases can present in children as well as in adults. Symptoms include pain in the forearm, numbness of the hand, and weakness of median-innervated muscles. Weakness affecting the pronator teres muscle favors entrapment above the elbow. Usually elbow extension aggravates the pain, but sometimes it is flexion that causes symptoms. Vascular compression is less common, but sometimes the pulse attenuates with elbow movement. Fracture of the spur can occur and cause acute neurovascular compression. Treatment is surgical excision of the spur and ligament.5,14,19,222

PROXIMAL FOREARM ENTRAPMENT OF THE MEDIAN NERVE Entrapment of the median nerve in the region of the elbow occurs at several locations, and they present in a similar fashion. At the Lahey Clinic, proximal median nerve entrapment was estimated to be 100 times less common than CTS.89 The nerve passes behind the lacertus fibrosis or bicipital aponeurosis at the elbow and is usually protected beneath the flexor muscle mass. The aponeurosis can compress the nerve if it is hypertrophied or thickened. Acute compression beneath the aponeurosis can occur after heavy lifting or venipuncture. At surgery, a tight aponeurosis with hemorrhage in the flexor and pronator muscles is found.232 The median nerve passes between the ulnar and humeral heads of the pronator teres in 80% of cases or, less commonly, posterior to them, where it may be compressed by fibrous arches (pronator syndrome). Other variations in the course of the nerve in

relation to the pronator teres muscle are described. A fibrous arch develops at the site where the two heads of the pronator teres merge. Hypertrophy of the pronator teres and overuse in the form of repetitive grasping and pronation at work and with athletic activities may contribute to symptoms. The most common site of compression in the proximal forearm is the hypertrophied pronator muscle. A little distally, the aponeurotic bridge of the flexor digitorum brevis muscle produces similar compression of median nerve branches. A case of median neuropathy has resulted from a persistent median artery that split the median nerve trunk distal to the pronator teres muscle.114 Proximal and distal branches to the pronator teres muscle originate above the level of the medial epicondyle but continue with the median nerve trunk under the proximal pronator muscle. Other branches originating before passage through the pronator teres include the palmaris longus, flexor carpi radialis, and flexor digitorum superficialis muscles. These muscles may be spared by entrapment by the fibrous arches while more distally innervated muscles may have weakness.75 Symptoms ascribed to the pronator syndrome include sensory symptoms in the median nerve distribution, muscle weakness, painful resisted pronation, and median distribution pain on compression of the pronator teres. Resisted flexion of the proximal interphalangeal joint of the middle finger may increase pain in the flexor digitorum sublimis syndrome. In contrast to CTS, nocturnal awakening with paresthesias occurs infrequently. Tinel’s sign may be elicited over the median nerve in the region of the elbow. Occasionally symptoms are bilateral.130 The pronator syndrome is rarely diagnosed in electromyography laboratories because the needle electrode examination and nerve conduction studies are usually normal in suspected cases. Other more common disorders such as CTS or a cervical radiculopathy can be excluded, however. Surgical decompression is reported to give satisfactory results in some patients; however, a significant number of patients have only partial relief and some have no change in their symptoms.181,206

ANTERIOR INTEROSSEOUS NERVE SYNDROME The anterior interosseous nerve arises from the median nerve distal to the pronator teres muscle and passes down the arm on the anterior interosseous membrane in the company of the anterior interosseous artery. It supplies the flexor digitorum profundus to digits 2 and 3, the flexor pollicis longus, and the pronator quadratus. If a Martin-Gruber anastomosis is present, paralysis of small hand muscles may confuse the diagnosis. There are no sensory branches. Weakness of flexion of the distal phalanges of the thumb and index finger causes the joints to

1454

Diseases of the Peripheral Nervous System

hyperextend when attempting to oppose or pinch the fingertips. The acute onset of the syndrome with pain followed by weakness is often a manifestation of idiopathic brachial plexopathy (neuralgic amyotrophy). This condition is the most common cause of the syndrome.153 Pain may be located in the upper arm or proximal forearm. Careful clinical and EMG study may reveal involvement of additional nerves. Cases have been reported in the postpartum period.32 Nerve conduction to the pronator quadratus can reveal a prolonged distal latency and reduced compound muscle amplitude.224 Recovery occurs in 3 to 12 months but can be incomplete. A case of anterior interosseous neuropathy as the presenting manifestation of multifocal motor neuropathy has been reported.225 Although weakness and EMG findings may be confined to muscles innervated by the anterior interosseous nerve, the lesion responsible for entrapment may be proximal to the origin of this nerve (pseudo anterior interosseous nerve syndrome). Entrapment is caused by tendinous origin of the deep head of the pronator teres and flexor digitorum superficialis, supracondylar fractures of the humerus or forearm bones, compression from plaster casts, blunt trauma, bodybuilding exercises, and neurilemoma.112 Some patients are referred to surgeons with suspected rupture of the flexor pollicis tendon. Exploration of the nerve in patients with no obvious cause should be delayed for 4 to 6 months because most will have a localized neuritis. About half of the time exploration shows constriction of the nerve.219

REFERENCES 1. Ahmed, T., and Braun, A. I.: Carpal tunnel syndrome with polymyalgia rheumatica. Arthritis Rheum. 21:221, 1978. 2. Ahn, D.-S.: Hand elevation: a new test for carpal tunnel syndrome. Ann. Plast. Surg. 46:120, 2001. 3. Alfonso, D. T., Sotrel, A., and Grossman, J. A.: Carpal tunnel syndrome due to an intraneural perineurioma in a 2year-old child. J. Hand. Surg. Br. 26:168, 2001. 4. Allieu, Y., Chammas, M., Idoux, O., et al.: Carpal tunnel syndrome and amyloid tenosynovitis in patients undergoing chronic hemodialysis: evaluation and treatment apropos of 130 cases. Ann. Chir. Main. Memb. Super. 13:113, 1994. 5. al-Naib, I.: Humeral supracondylar spur and Struthers’ ligament: a rare cause of neurovascular entrapment in the upper limb. Int. Orthop. 18:393, 1994. 6. Al-Qattan, M. M., Thomson, H. G., and Clarke, H. M.: Carpal tunnel syndrome in children and adolescents with no history of trauma. J. Hand Surg. Br. 21:108, 1996. 7. Amadio, P. C.: The Mayo Clinic and carpal tunnel syndrome. Mayo Clin. Proc. 67:42, 1992. 8. Andrews, T. R., Colon-Otero, G., Calamia, K. T., et al.: Utility of subcutaneous fat aspiration for diagnosing amyloidosis in patients with isolated peripheral neuropathy. Mayo Clin. Proc. 77:1287, 2002.

9. Armstrong, T. J., and Chaffin, D. B.: Carpal tunnel syndrome and selected personal attributes. J. Occup. Med. 21:481, 1979. 10. Armstrong, T. J., Foulke, J. A., Joseph, B. S., et al.: Investigation of cumulative trauma disorders in a poultry processing plant. Am. Ind. Hyg. Assoc. J. 43:103, 1982. 11. Arnold, G.: The carpal tunnel syndrome in congestive heart failure. Postgrad. Med. J. 53:623, 1977. 12. Atroshi, I., Axelsson, G., Gummesson, C., et al.: Carpal tunnel syndrome with severe sensory deficit: endoscopic release in 18 cases. Acta Orthop. Scand. 71:484, 2000. 13. Atroshi, I., Gummesson, C., Johnsson, R., et al.: Prevalence of carpal tunnel syndrome in a general population. JAMA 282:123, 1999. 14. Aydinlioglu, A., Cirak, B., Akpinar, F., et al.: Bilateral median nerve compression at the level of Struthers’ ligament: case report. J. Neurosurg. 92:693, 2000. 15. Ayhan-Ardic, F. F., and Erdem, H. R.: Long-term clinical and electrophysiological results of local steroid injection in patients with carpal tunnel syndrome. Funct. Neurol. 15:157, 2000. 16. Barnhart, S., and Rosenstock, L.: Carpal tunnel syndrome in grocery checkers: a cluster of a work-related illness. West. J. Med. 147:37, 1987. 17. Beekman, R., and Visser, L. H.: Sonography in the diagnosis of carpal tunnel syndrome: a critical review of the literature. Muscle Nerve 27:26, 2003. 18. Bekkelund, S. I., Pierre-Jerome, C., Torbergsen, T., et al.: Impact of occupational variables in carpal tunnel syndrome. Acta Neurol. Scand. 103:193, 2001. 19. Bilbe, T., Yalaman, O., Bilge, S., et al.: Entrapment neuropathy of the median nerve at the level of the ligament of Struthers. Neurosurgery 27:787, 1990. 20. Bindiger, A., Zelnik, J., Kuschner, S., et al.: Spontaneous acute carpal tunnel syndrome in an anticoagulated patient. Bull. Hosp. Jt. Dis. 54:52, 1995. 21. Bjorkqvist, S. E., Lang, A. H., Punnonen, R., et al.: Carpal tunnel syndrome in ovariectomized women. Acta Obstet. Gynecol. Scand. 56:127, 1977. 22. Bland, J. D.: A neurophysiological grading scale for carpal tunnel syndrome. Muscle Nerve 23:1280, 2000. 23. Bleecker, M. L., Bohlman, M., Moreland, R., and Tipton, A.: Carpal tunnel syndrome: role of carpal canal size. Neurology 35:1599, 1985. 24. Bolton, C. F., and McFarlane, R. M.: Human pneumatic tourniquet paralysis. Neurology 28:787, 1978. 25. Brain, W. R., Wright, A. D., and Wilkinson, M.: Spontaneous compression of both median nerves in the carpal tunnel: six cases treated surgically. Lancet 1:277, 1941. 26. Brown, E. Z., and Snyder, C. C.: Carpal tunnel syndrome caused by hand injuries. Plast. Reconstr. Surg. 56:41, 1975. 27. Bruno, K. M., Farhoomand, L., Libman, B. S., et al.: Cryptococcal arthritis, tendinitis, tenosynovitis, and carpal tunnel syndrome: report of a case and review of the literature. Arthritis Rheum. 47:104, 2002. 28. Brutus, J. P., Baeten, Y., Chahidi, N., et al.: Atypical mycobacterial infections of the hand: report of eight cases and literature review. Chir. Main 20:280, 2001.

Median Neuropathy 29. Camilleri, I. G., and Milner, R. H.: Intraneural lipofibroma of the median nerve. J. Hand Surg. Br. 23:120, 1998. 30. Cannon, B. W., and Love, J. G.: Tardy median palsy; median neuritis; median thenar neuritis amenable to surgery. Surgery 20:210, 1946. 31. Cannon, L. J., Bernacki, E. J., and Walter, S. D.: Personal and occupational factors associated with carpal tunnel syndrome. J. Occup. Med. 23:255, 1981. 32. Carmant, L., and Veilleux, M.: Anterior interosseous neuropathy in the postpartum period. Can. J. Neurol. Sci. 20:56, 1993. 33. Carter, T., Jordan, R., and Cummins, C.: Electrodiagnostic Techniques in the Presurgical Assessment of Patients with Carpal Tunnel Syndrome (DPHE 2000 Report No. 18). Birmingham, UK, Department of Public Health and Epidemiology, West Midlands Development and Evaluation Service, University of Birmingham, 2000. 34. Celiker, R., Arslan, S., and Inanici, F.: Corticosteroid injection vs. nonsteroidal antiinflammatory drug and splinting in carpal tunnel syndrome. Am. J. Phys. Med. Rehabil. 81:182, 2002. 35. Chang, M. H., Chiang, H. T., Lee, S. S., et al.: Oral drug of choice in carpal tunnel syndrome. Neurology 51:390, 1998. 36. Chen, C. K., Chung, C. B., Yeh, L., et al.: Carpal tunnel syndrome caused by tophaceous gout: CT and MR imaging features in 20 patients. AJR Am. J. Roentgenol. 175:655, 2000. 37. Chevalier, X., Hermine, O., Authier, F. J., et al.: Carpal tunnel syndrome due to T cell lymphoma. Arthritis Rheum. 38:1707, 1995. 38. Chiu, K. Y., Ng, W. F., Wong, W. B., et al.: Acute carpal tunnel syndrome caused by pseudogout. J. Hand Surg. Am. 17:299, 1992. 39. Christensen, J. E., Peter, P. J., Nielsen, V. K., et al.: Prevalence of carpal tunnel syndrome among individuals with Down syndrome. Am. J. Ment. Retard. 102:547, 1998. 40. Chroni, E., Paschalis, C., Arvaniti, C., et al.: Carpal tunnel syndrome and hand configuration. Muscle Nerve 24:1607, 2001. 41. Chuang, H. L., and Wong, C. W.: Carpal tunnel syndrome induced by tophaceous deposits on the median nerve: case report. Neurosurgery 34:919, 1994. 42. Cox, N. H., Large, D. M., Petersen, W. D., et al.: Blisters, ulceration and autonomic neuropathy in carpal tunnel syndrome. Br. J. Dermatol. 126:611, 1992. 43. Crowley, B., Gschwind, C. R., and Storey, C.: Selective motor neuropathy of the median nerve caused by a ganglion in the carpal tunnel. J. Hand Surg. Br. 23:611, 1998. 44. Cruz Martinez, A., and Arpa, J.: Carpal tunnel syndrome in childhood: study of 6 cases. Electroencephalogr. Clin. Neurophysiol. 109:304, 1998. 45. Cseuz, K. A., Thomas, J. E., Lambert, E. H., et al.: Longterm results of operation for carpal tunnel syndrome. Mayo Clin. Proc. 41:232, 1966. 46. Cudlip, S. A., Howe, F. A., Clifton, A., et al.: Magnetic resonance neurography studies of the median nerve before and after carpal tunnel decompression. J. Neurosurg. 96:1046, 2001. 47. Dahlin, L. H., and Lundborg, G.: Vibration-induced hand problems: role of the peripheral nerves in the pathophysiology. Scand. J. Plast. Reconstr. Surg. 25:225, 2001.

1455

48. Dammers, J. W., Veering, M. M., and Vermeulen, M.: Injection with methylprednisolone proximal to the carpal tunnel: randomised double blind trial. BMJ 319:884, 1999. 49. D’Arcy, C. A., and McGee, S.: Does this patient have carpal tunnel syndrome? JAMA 283:3110, 2000. 50. De Krom, M. C., Knipschild, P. G., Kester, A. D., and Spaans, F.: Efficacy of provocative tests for diagnosis of carpal tunnel syndrome. Lancet 335:393, 1990. 51. de Krom, M. C., Knipschild, P. G., Kester, A. D., et al.: Carpal tunnel syndrome: prevalence in the general population. J. Clin. Epidemiol. 45:373, 1992. 52. Dennerlein, J. T., Soumekh, F. S., Fossel, A. H., et al.: Longer distal motor latency predicts better outcomes of carpal tunnel release. J. Occup. Environ. Med. 44:176, 2002. 53. Dhong, E. S., Han, S. K., Lee, B. I., and Kim, W. K.: Correlation of electrodiagnostic findings with subjective symptoms in carpal tunnel syndrome. Ann. Plast. Surg. 45:127, 2000. 54. Diaz, J. H.: Carpal tunnel syndrome in female nurse anesthetists versus operating room nurses: prevalence, laterality, and impact of handedness. Anesth. Analg. 93:975, 2001. 55. Dieck, G. S., and Kelsey, J. L.: An epidemiological study of carpal tunnel syndrome in an adult population. Prev. Med. 14:63, 1993. 56. Direkze, M., and Fernando, P. S.: Peripheral nerve entrapment in Paget’s disease. Acta Neurol. Scand. 54:375, 1976. 57. Durkan, J. A., and River, H.: A new diagnostic test for carpal tunnel syndrome. J. Bone Joint Surg. Am. 73:535, 1991. 58. Duyff, R. F., Van den Bosch, J., Laman, D. M., et al.: Neuromuscular findings in thyroid dysfunction: a prospective clinical and electrodiagnostic study. J. Neurol. Neurosurg. Psychiatry 68:750, 2000. 59. Dyck, P. J., and Giannini, C.: Pathologic alterations in the diabetic neuropathies of humans: a review. J. Neuropathol. Exp. Neurol. 55:1181, 1996. 60. Dyck, P. J., Kratz, K. M., Karnes, J. L., et al.: The prevalence by staged severity of various types of diabetic neuropathy, retinopathy, and nephropathy in a population-based cohort: the Rochester Diabetic Neuropathy Study. Neurology 43:817, 1993. 61. Ebenbichler, G. R., Resch, K. L., Nicolakis, P., et al.: Ultrasound treatment for treating the carpal tunnel syndrome: randomised “sham” controlled trial. BMJ 316:731, 1998. 62. Eversmann, W. W. Jr., and Ritsick, J. A.: Intraoperative changes in motor nerve conduction latency in carpal tunnel syndrome. J. Hand Surg. Am. 3:77, 1978. 63. Falck, B., and Aarnio, P.: Left-sided carpal tunnel syndrome in butchers. Scand. J. Work Environ. Health 9:291, 1983. 64. Falk, R. H., Comenzo, R. L., and Skinner, M.: The systemic amyloidoses. N. Engl. J. Med. 337:898, 1997. 65. Falsetti, P., Frediani, B., Storri, L., et al.: Evidence for synovitis in active polymyalgia rheumatica: sonographic study in a large series of patients. J. Rheumatol. 29:123, 2002. 66. Feldman, R. G., Travers, P. H., Chirico-Post, J., et al.: Risk assessment in electronic assembly workers: carpal tunnel syndrome. J. Hand Surg. Am. 12:849, 1987.

1456

Diseases of the Peripheral Nervous System

67. Ferry, S., Pritchard, T., Keenan, J., et al.: Estimating the prevalence of delayed median nerve conduction in the general population. Br. J. Rheumatol. 37:630, 1998. 68. Ferry, S., Silman, A. J., Pritchard, T., et al.: The association between different patterns of hand symptoms and objective evidence of median nerve compression: a community based survey. Arthritis Rheum. 41:720, 1998. 69. Fitzcharles, M.-A., and Esdaile, J. M.: Carpal tunnel syndrome complicated by reflex sympathetic dystrophy syndrome. Br. J. Rheumatol. 30:468, 1991. 70. Fleckenstein, J. L., and Wolfe, G. I.: MRI vs EMG: which has the upper hand in carpal tunnel syndrome? [Editorial]. Neurology 58:1583, 2002. 71. Floege, J., and Ketteler, M.: Beta2-microglobulin-derived amyloidosis: an update. Kidney Int. Suppl. 78:S164, 2001. 72. Franklin, G. M., Haug, J., Heyer, N., et al.: Occupational carpal tunnel syndrome in Washington State, 1984–1988. Am. J. Public Health 81:741, 1991. 73. Fritz, T. M., and Burg, G.: Carpal tunnel syndrome with ulcerous skin lesions. Dermatology 201:165, 2000. 74. Fumiere, E., Dugardeyn, C., Roquet, M. E., et al.: US demonstration of a thrombosed persistent median artery in carpal tunnel syndrome. JBR BTR (J. Belge Radiol.) 85:1, 2002. 75. Fuss, F. K., and Wurzl, G. H.: Median nerve entrapment. Pronator teres syndrome: surgical anatomy and correlation with symptom patterns. Surg. Radiol. Anat. 12:267, 1990. 76. Gassner, E. M., Schocke, M., Peer, S., et al.: Persistent median artery in the carpal tunnel: color Doppler ultrasonographic findings. J. Ultrasound Med. 21:455, 2002. 77. Gellman, H., Chandler, D. R., Petrasek, J., et al.: Carpal tunnel syndrome in paraplegic patients. J. Bone Joint Surg. Am. 70:517, 1988. 78. Gellman, H., Gelberman, R. H., Tan, A. M., et al.: Carpal tunnel syndrome: an evaluation of the provocative diagnostic tests. J. Bone Joint Surg. Am. 68:735, 1986. 79. Gerritsen, A. A., Uitdehaag, B. M., vanGeldere, D., et al.: Systematic review of randomized clinical trials of surgical treatment for carpal tunnel syndrome. Br. J. Surg. 88:1285, 2001. 80. Giannini, F., Cioni, R., Mondelli, M., et al.: A new clinical scale of carpal tunnel syndrome: validation of the measurement and clinical-neurophysiological assessment. Clin. Neurophysiol. 113:71, 2002. 81. Gilliatt, R. W., and Sears, T. A.: Sensory nerve action potentials in patients with peripheral nerve lesions. J. Neurol. Neurosurg. Psychiatry 21:109, 1958. 82. Gilliatt, R. W., and Wilson, T. G.: A pneumatic-tourniquet test in the carpal tunnel syndrome. Lancet 2:595, 1953. 83. Girlanda, P., Dattola, R., Venuto, C., et al.: Local steroid treatment in idiopathic carpal tunnel syndrome: short- and long-term efficacy. J. Neurol. 240:187, 1993. 84. Gleberman, R. H., Hergenroeder, P. T., Hargens, A. R., et al.: The carpal tunnel syndrome: a study of carpal canal pressures. J. Bone Joint Surg. Am. 63:380, 1981. 85. Goadsby, P. J., and Burke, D.: Deficits in the function of small and large afferent fibers in confirmed cases of carpal tunnel syndrome. Muscle Nerve 17:614, 1994.

86. Gordon, C., Bowyer, B. L., and Johnson, E. W.: Electrodiagnostic characteristics of acute carpal tunnel syndrome. Arch. Phys. Med. Rehabil. 68:545, 1987. 87. Gordon, C., Johnson, E. W., Gatens, P. F., et al.: Wrist ratio correlation with carpal tunnel syndrome in industry. Am. J. Phys. Med. Rehabil. 67:270, 1988. 88. Gossett, J. G., and Chance, P. F.: Is there a familial carpal tunnel syndrome? An evaluation and literature review. Muscle Nerve 21:1533, 1998. 89. Gross, P. T., and Tolomeo, E. A.: Proximal median neuropathies. Neurol. Clin. 17:425, 1999. 90. Gupta, S. K., and Benstead, T. J.: Symptoms experienced by patients with carpal tunnel syndrome. Can. J. Neurol. Sci. 24:338, 1997. 91. Gutmann, L.: AAEM minimonograph #2: important anomalous innervation of the extremities. Muscle Nerve 16:339, 1993. 92. Haas, D. C., Nord, S. G., and Bome, M. P.: Carpal tunnel syndrome following automobile collisions. Arch. Phys. Med. Rehabil. 62:204, 1981. 93. Haddad, F. S., Jones, D. H., Vellodi, A., et al.: Carpal tunnel syndrome in the mucopolysaccharidoses and mucolipidoses. J. Bone Joint Surg. Br. 79:576, 1997. 94. Hakim, A. J., Cherkas, L., El Sayat, S., et al.: The genetic contribution of carpal tunnel syndrome in women: a twin study. Arthritis Rheum. 47:275, 2002. 95. Halperin, J. J., Volkman, D. J., Luft, B. J., et al.: Carpal tunnel syndrome in Lyme borreliosis. Muscle Nerve 12:397, 1989. 96. Hamanaka, I., Okutsu, I., Shimizu, K., et al.: Evaluation of carpal canal pressure in carpal tunnel syndrome. J. Hand Surg. Am. 20:848, 1995. 97. Hidou, M., Huraux, C., Viry-Babel, F., et al.: Pneumatic tourniquet paralysis: a differential diagnosis after locoregional anesthesia of the upper limb. J. Chir. (Paris) 129:213, 1992. 98. Hirasawa, Y., and Ogura, T.: Carpal tunnel syndrome in patients on long-term haemodialysis. Scand. J. Plast. Reconstr. Surg. Hand Surg. 34:373, 2000. 99. Hocking, B.: Epidemiological aspects of “repetition strain injury” in Telecom Australia. Med. J. Aust. 147:218, 1987. 100. Hofmann, A. K., Wustner, M. C., and Spier, W.: Compression neuropathy of the median nerve at the wrist joint caused by chondroma. Handchir. Mikrochir. Plast. Chir. 22:96, 1990. 101. Homan, M. M., Franzblau, A., Werner, R. A., et al.: Agreement between symptom surveys, physical examination procedures and electrodiagnostic findings for the carpal tunnel syndrome. Scand. J. Work Environ. Health 25:115, 1999. 102. Hongell, A., and Mattsson, H. S.: Neurographic studies before, after, and during operation for median nerve compression in the carpal tunnel. Scand. J. Plast. Reconstr. Surg. 5:103, 1971. 103. Hsu, W.-C., Chen, W.-H., Oware, A., et al.: Unusual entrapment neuropathy in a golf player. Neurology 59:646, 2002. 104. Hund, E., Linke, R. P., Willig, F., et al.: Transthyretinassociated neuropathic amyloidosis: pathogenesis and treatment. Neurology 27:431, 2001.

Median Neuropathy 105. Hunt, T. R., and Osterman, A. L.: Complications of the treatment of carpal tunnel syndrome. Hand Clin. 10:63, 1994. 106. Iannicelli, E., Almberger, M., Chianta, G. A., et al.: Bifid median nerve in the carpal tunnel: integrated imaging. Radiol. Med. (Torino) 101:456, 2001. 107. Jablecki, C. K., Andary, M. T., Floeter, M., et al.: Second AAEM literature review of the usefulness of nerve conduction studies and needle electromyography for the evaluation of patients with carpal tunnel syndrome. Neurology 58:1589, 2002. 108. Jablecki, C. K., Andary, M. T., So, Y. T., et al.: Literature review of the usefulness of nerve conduction studies and electromyography for the evaluation of patients with carpal tunnel syndrome. AAEM Quality Assurance Committee. Muscle Nerve 16:1392, 1993. 109. Jarvki, J. G., and Yuen, E.: Diagnosis of carpal tunnel syndrome: electrodiagnostic and magnetic resonance imaging evaluation. Neurosurg. Clin. North Am. 12:241, 2001. 110. Jarvik, J. G., Yuen, E., Haynor, D. R., et al.: MR nerve imaging in a prospective cohort of patients with suspected carpal tunnel syndrome. Neurology 58:1597, 2002. 111. Jenkins, P. J., Sohaib, S. A., Akker, S., et al.: The pathology of median neuropathy in acromegaly. Ann. Intern. Med. 133:197, 2000. 112. Joist, A., Joosten, U., Wetterkamp, D., et al.: Anterior interosseous nerve compression after supracondylar fracture of the humerus: a metaanalysis. J. Neurosurg. 90:1053, 1999. 113. Jones, H. R. J., Beetham, W. P. J., Silverman, M. L., et al.: Eosinophilic fasciitis and the carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 49:324, 1986. 114. Jones, N. F., and Ming, N. L.: Persistent median artery as a cause of pronator syndrome. J. Hand Surg. Am. 13:728, 1988. 115. Katz, J. N., Larson, M. G., Sabra, A., et al.: The carpal tunnel syndrome: diagnostic utility of the history and physical examination. Ann. Intern. Med. 112:321, 1990. 116. Katz, J. N., Lsina, E., Amick, B. C., et al.: Predictors of outcomes of carpal tunnel release. Arthritis Rheum. 44:1184, 2001. 117. Katz, J. N., Punnett, L., Simmons, B. P., et al.: Workers’ compensation recipients with carpal tunnel syndrome: the validity of self-reported health measures. Am. J. Public Health 86:52, 1996. 118. Kaul, M. P., Pagel, K. J., Wheatley, M. J., et al.: Carpal compression test and pressure provocative test in veterans with median-distribution paresthesias. Muscle Nerve 24:107, 2001. 119. Keberle, M., Jenett, M., Kenn, W., et al.: Technical advances in ultrasound and MR imaging of carpal tunnel syndrome. Eur. Radiol. 10:1043, 2000. 120. Keir, J. P., Bach, J. M., and Rempel, D.: Effects of computer mouse design and task on carpal tunnel pressure. Ergonomics 42:1350, 1999. 121. Kimura, J.: The carpal tunnel syndrome: localization of conduction abnormalities within the distal segment of the median nerve. Brain 102:619, 1979.

1457

122. Kinugasa, K., Hashizume, H., Shigeyama, Y., et al.: Histopathology and clinical results of carpal tunnel syndrome in idiopathic cases and hemodialysis patients. Acta Med. Okayama 51:63, 1997. 123. Kitsis, C. K., Savvidou, O., Alam, A., et al.: Carpal tunnel syndrome despite negative neurophysiological studies. Acta Orthop. Belg. 68:135, 2002. 124. Kouyoumdjian, J. A., Zanetta, D. M. T., and Morita, M. P. A.: Evaluation of age, body mass index, and wrist index as risk factors for carpal tunnel syndrome severity. Muscle Nerve 25:93, 2002. 125. Kozin, S. H.: The anatomy of the recurrent branch of the median nerve. J. Hand Surg. Am. 23:852, 1998. 126. Kremchek, T. E., and Kremchek, E. J.: Carpal tunnel syndrome caused by flexor tendon sheath lipoma. Orthop. Rev. 17:1083, 1988. 127. Kuschner, S. H., Brien, W. W., Johnson, D., et al.: Complications associated with carpal tunnel release. Orthop. Rev. 20:346, 1991. 128. Kutluhan, S., Akhan, G., Demirci, S., et al.: Carpal tunnel syndrome in carpet workers. Int. Arch. Occup. Environ. Health 74:454, 2001. 129. Label, L. S.: Carpal tunnel syndrome resulting from steering wheel impact. Muscle Nerve 14:904, 1991. 130. Lacey, S. H., and Soldatis, J. J.: Bilateral pronator syndrome associated with anomalous heads of the pronator teres muscle: a case report. J. Hand Surg. Am. 18:349, 1993. 131. Lakhanpal, S., Ginsburg, W. W., Michet, C. J., et al.: Eosinophilic fasciitis: clinical spectrum and therapeutic response in 52 cases. Semin. Arthritis Rheum. 17:221, 1988. 132. Lalumandier, J. A.: Prevalence and risk factors of hand problems and carpal tunnel syndrome among dental hygienists. J. Dent. Hyg. 75:130, 2001. 133. Lamberti, P. M., and Light, T. R.: Carpal tunnel syndrome in children. Hand Clin. 18:331, 2002. 134. Lee, D., van Holsbeeck, M. T., Janevski, P. K., et al.: Diagnosis of carpal tunnel syndrome: ultrasound versus electromyography. Radiol. Clin. North Am. 37:859, 1999. 135. Levine, D. W., Simmons, B. P., Koris, M. J., et al.: A selfadministered questionnaire for the assessment of severity of symptoms and functional status in carpal tunnel syndrome. J. Bone Joint Surg. Am. 75:1585, 1993. 136. Linscheid, F. L., Peterson, L. F., and Juergen, J. L.: Carpal tunnel syndrome associated with vasospasm. J. Bone Joint Surg. Am. 49:1141, 1967. 137. Liss, G. M., Armstrong, C., Kusiak, R. A., et al.: Use of provincial health insurance plan billing data to estimate carpal tunnel syndrome morbidity and surgery rates. Am. J. Ind. Med. 22:395, 1992. 138. Longstaff, L., Milner, F. H., O’Sullivan, S., et al.: Carpal tunnel syndrome: the correlation between outcome, symptoms and nerve conduction study findings. J. Hand Surg. Br. 26:457, 2001. 139. Low, P. A.: Recent advances in the pathogenesis of diabetic neuropathy. Muscle Nerve 10:121, 1987. 140. Luchetti, R., Schoenhuber, R., Alfarano, M., et al.: Carpal tunnel syndrome: correlations between pressure measurement and intraoperative electrophysiological nerve study. Muscle Nerve 13:1164, 1990.

1458

Diseases of the Peripheral Nervous System

141. Luchetti, R., Schoenhuber, R., Alfarano, M., et al.: Serial overnight recordings of intracarpal canal pressure in carpal tunnel syndrome patients with and without wrist splinting. J. Hand Surg. Br. 19:35, 1994. 142. Luciano, C. A., Russell, J. W., Banerjee, T. K., et al.: Physiological characterization of neuropathy in Fabry’s disease. Muscle Nerve 26:622, 2002. 143. Mackinnon, S. E., and Novak, C. B.: Repetitive strain in the workplace. J. Hand Surg. Am. 22:2, 1997. 144. Mahjneh, I., Saarinen, A., and Siivola, J.: Familial carpal tunnel syndrome: a report of a Finnish family. Acta Neurol. Scand. 104:377, 2001. 145. Manfredi, R., Calza, L., and Chiodo, F.: Carpal tunnel syndrome in HIV-infected patients treated with highly active antiretroviral therapy: other case reports. Rheumatol. Int. 21:81, 2001. 146. Margolis, W., and Kraus, J. F.: The prevalence of carpal tunnel syndrome symptoms in female supermarket checkers. J. Occup. Med. 29:953, 1987. 147. Marshall, G., Edelstein, G., and Hirshman, C. A.: Median nerve compression following radial artery puncture. Anesth. Analg. 59:953, 1980. 148. Martinelli, P., Baruzzi, A., Montagna, P., et al.: Carpal tunnel syndrome in a patient with a Cimino-Brescia fistula. Eur. Neurol. 20:478, 1981. 149. Mascola, J. R., and Rickman, L. S.: Infectious causes of carpal tunnel syndrome: case report and review. Rev. Infect. Dis. 13:911, 1991. 150. Mashoof, A. A., Levy, H. J., Soifer, T. B., et al.: Neural anatomy of the transverse carpal ligament. Clin. Orthop. 386:218, 2001. 151. McClain, E. J., and Wissinger, H. A.: The acute carpal tunnel syndrome: nine case reports. J. Trauma 16:75, 1976. 152. McConnell, J. R., and Bush, D. C.: Intraneural steroid injection as a complication in the management of carpal tunnel syndrome: a report of three cases. Clin. Orthop. 250:181, 1990. 153. Miller-Breslow, A., Terrono, A., and Millender, L. H.: Nonoperative treatment of anterior interosseous nerve paralysis. J. Hand Surg. Am. 15:493, 1990. 154. Mitsumoto, H., Estes, M. L., Wilbourn, A. J., et al.: Perineurial cell hypertrophic mononeuropathy manifesting as carpal tunnel syndrome. Muscle Nerve 15:1364, 1992. 155. Moersch, F. P.: Median thenar neuritis. Proc. Staff Mayo Clin. 13:220, 1938. 156. Molina, R., Provost, T. T., and Alexander, E. L.: Peripheral inflammatory vascular disease in Sjögren’s syndrome: association with nervous system complications. Arthritis Rheum. 28:1341, 1985. 157. Molitor, P. J., and Wimperis, J. Z.: Acute carpal tunnel syndrome in haemophiliacs. Br. J. Clin. Pract. 44:675, 1990. 158. Monagle, K., Kai, G., Chu, A., et al.: Quantitative MR imaging of carpal tunnel syndrome. AJR Am. J. Roentgenol. 172:1581, 1999. 159. Mondelli, M., Ginanneschi, F., Rossi, S., et al.: Interobserver reproducibility and responsiveness of a clinical severity scale in surgically treated carpal tunnel syndrome. Acta Neurol. Scand. 106:263, 2002.

160. Mondelli, M., Passero, S., and Giannini, F.: Provocative tests in different stages of carpal tunnel syndrome. Clin. Neurol. Neurosurg. 103:178, 2001. 161. Mondelli, M., Reale, F., Pauda, R., et al.: Clinical and neurophysiological outcome of surgery in extreme carpal tunnel syndrome. Clin. Neurophysiol. 112:1237, 2001. 162. Morgan, G., and Wilbourn, A. J.: Cervical radiculopathy and coexisting distal entrapment neuropathies: doublecrush syndromes? Neurology 50:78, 1998. 163. Mumenthaler, M.: Carpal tunnel syndrome: first description. Neurology 34:921, 1984. 164. Murphy, R. X., Chernofsky, M. A., Osborne, M. A., et al.: Magnetic resonance imaging in the evaluation of persistent carpal tunnel syndrome. J. Hand Surg. Am. 18:113, 1993. 165. Nakamichi, K., and Tachibana, S.: Unilateral carpal tunnel syndrome and space-occupying lesions. J. Hand Surg. Br. 18:748, 1993. 166. Nakamichi, K., and Tachibana, S.: Small hand as a risk factor for idiopathic carpal tunnel syndrome. Muscle Nerve 18:664, 1995. 167. Nakamichi, K., and Tachibana, S.: Histology of the transverse carpal ligament and flexor tenosynovium in idiopathic carpal tunnel syndrome. J. Hand Surg. Am. 23:1015, 1998. 168. Nakamichi, K., and Tachibana, S.: Ultrasonographic measurement of median nerve cross-sectional area in idiopathic carpal tunnel syndrome: diagnostic accuracy. Muscle Nerve 26:798, 2002. 169. Nakamichi, K. I., and Tachibana, S.: Amyloid deposition in the synovium and ligament in idiopathic carpal tunnel syndrome. Muscle Nerve 19:1349, 1996. 170. Nakamichi, K. I., and Tachibana, S.: Enlarged median nerve in idiopathic carpal tunnel syndrome. Muscle Nerve 23:1713, 2000. 171. Nathan, P. A., and Keniston, F. C.: Carpal tunnel syndrome and its relation to general physical condition. Hand Clin. 9:253, 1993. 172. Nathan, P. A., Keniston, F. C., Myers, L. D., et al.: Natural history of median nerve sensory conduction in industry: relationship to symptoms and carpal tunnel syndrome in 558 hands over 11 years. Muscle Nerve 21:711, 1998. 173. Nathan, P. A., Meadows, K. D., and Doyle, L. S.: Occupation as a risk factor for impaired sensory conduction of the median nerve at the carpal tunnel. J. Hand Surg. Br. 13:167, 1988. 174. Nathan, P. A., Takigawa, K., Keniston, R. C., et al.: Slowing of sensory conduction of the median nerve and carpal tunnel syndrome in Japanese and American industrial workers. J. Hand Surg. Br. 19:30, 1994. 175. Nestle, F. O., and Burg, G.: Bilateral carpal tunnel syndrome as a clue for the diagnosis of systemic amyloidosis. Dermatology 202:353, 2001. 176. Niemer, G. W., Bolster, M. B., Buxbaum, L., et al.: Carpal tunnel syndrome in sarcoidosis. Sarcoidosis Vasc. Diffuse Lung Dis. 18:296, 2001. 177. Nordstrom, D. L., DeStefano, F., Vierkant, R. A., et al.: Incidence of diagnosed carpal tunnel syndrome in the general population. Epidemiology 9:342, 1998.

Median Neuropathy 178. Ochoa, J., Fowler, T. J., and Gilliatt, R. W.: Anatomical changes in peripheral nerves compressed by a pneumatic tourniquet. J. Anat. 113:433, 1972. 179. Ochsner, F., Baumann, R. P., and Kuntzer, T.: Carpal tunnel syndrome with an unusual cause: a malignant nerve sheath tumor of the median nerve. Rev. Neurol. 157:1547, 2001. 180. O’Duffy, J. D., Randall, R. V., and MacCarty, C. S.: Median neuropathy (carpal-tunnel syndrome) in acromegaly: a sign of endocrine overactivity. Ann. Intern. Med. 78:379, 1973. 181. Olehnik, W. K., Manske, P. R., and Szerzinski, J.: Median nerve compression in the proximal forearm. J. Hand Surg. Am. 19:121, 1994. 182. Olney, J. R., Quenzer, D. E., and Makowsky, M.: Contested claims in carpal tunnel surgery: outcome study of worker’s compensation factors. Iowa Orthop. J. 19:111, 1999. 183. Osterman, A. L., Whitman, M., and Porta, L. D.: Nonoperative carpal tunnel syndrome treatment. Hand Clin. 18:279, 2002. 184. Oswald, T. A., Wertsch, J. J., Vennix, M. J., et al.: Ulnar paresthesia a presenting symptom of carpal tunnel “syndrome.” Muscle Nerve 17:1082, 1994. 185. Oztas, O., Turan, B., Bora, I., et al.: Ultrasound therapy effect in carpal tunnel syndrome. Arch. Phys. Med. Rehabil. 79:1540, 1998. 186. Padua, L., Aprile, I., Caliandro, P., et al.: Symptoms and neurophysiological picture of carpal tunnel syndrome in pregnancy. Clin. Neurophysiol. 112:1946, 2001. 187. Padua, L., Aprile, I., Caliandro, P., et al.: Carpal tunnel syndrome in pregnancy: multiperspective follow-up of untreated cases. Neurology 26:1643, 2002. 188. Padua, L., Padua, R., Lo Monaco, M., et al.: Multiperspective assessment of carpal tunnel syndrome: a multicenter study. Neurology 53:1654, 1999. 189. Paik, N. J., Cho, S. H., and Han, T. R.: Ultrasound therapy facilitates the recovery of acute pressure-induced conduction block of the median nerve in rabbits. Muscle Nerve 26:356, 2002. 190. Palumbo, C. F., Szabo, R. M., and Olmsted, S. L.: The effects of hypothyroidism and thyroid replacement on the development of carpal tunnel syndrome. J. Hand Surg. Am. 25:734, 2000. 191. Papanicolaou, G. D., McCabe, S. J., and Firrell, J.: The prevalence and characteristics of nerve compression symptoms in the general population. J. Hand Surg. Am. 26:460, 2001. 192. Pascarelli, E., and Hsu, Y.-P.: Understanding work-related upper extremity disorders: clinical findings in 485 computer users, musicians, and others. J. Occup. Rehabil. 11:1, 2001. 193. Pascarelli, E. F., and Kella, J. J.: Soft-tissue injuries related to use of the computer keyboard. J. Occup. Med. 35:522, 1993. 194. Pascual, E., Giner, V., Arostegui, A., et al.: Higher incidence of carpal tunnel syndrome in oophorectomized women. Br. J. Rheumatol. 30:62, 1991. 195. Patiala, H., Rokkanen, P., Kruuna, O., et al.: Carpal tunnel syndrome: anatomical and clinical investigation. Arch. Orthop. Trauma Surg. 104:69, 1985. 196. Peled, I., Iosipovich, Z., Rousso, M., et al.: Hemangioma of the median nerve. J. Hand Surg. Am. 5:363, 1980.

1459

197. Perkins, B. A., Olaleye, D., and Bril, V.: Carpal tunnel syndrome in patients with diabetic polyneuropathy. Diabetes Care 25:565, 2002. 198. Phalen, G. S.: Spontaneous compression of the median nerve at the wrist. JAMA 164:524, 1951. 199. Phalen, G. S., Gardner, W. J., and La Londe, A. A.: Neuropathy of the median nerve due to compression beneath the transverse carpal ligament. J. Bone Joint Surg. Am. 32:109, 1950. 200. Pierre-Jerome, C., Bekkelund, S. I., and Nordstrom, R.: Quantitative MRI analysis of anatomic dimensions of the carpal tunnel in women. Surg. Radiol. Anat. 19:31, 1997. 201. Punnett, L., Robins, J. M., Wegman, D. H., et al.: Soft tissue disorders in the upper limbs of female garment workers. Scand. J. Work Environ. Health 11:417, 1985. 202. Purnell, D. C., Daly, D. D., and Lipscomb, P. R.: Carpal tunnel syndrome associated with myxedema. Arch. Intern. Med. 108:751, 1961. 203. Radecki, P.: The familial occurrence of carpal tunnel syndrome. Muscle Nerve 17:325, 1994. 204. Radecki, P.: Personal factors and blood volume movement in causation of median neuropathy at the carpal tunnel: a commentary. Am. J. Phys. Med. Rehabil. 75:235, 1996. 205. Refacian, M., King, J. C., Dumitru, D., et al.: Carpal tunnel syndrome and the Riche-Cannieu anastomosis: electrophysiological findings. Electromyogr. Clin. Neurophysiol. 41:377, 2001. 206. Rehak, D. C.: Pronator syndrome. Clin. Sports Med. 20:531, 2001. 207. Rempel, D., Evanoff, B., Amadio, P. C., et al.: Consensus criteria for the classification of carpal tunnel syndrome in epidemiological studies. Am. J. Public Health 88:1447, 1998. 208. Repaci, M., Torrieri, F., Di Blasio, F., et al.: Exclusive electrophysiological motor involvement in carpal tunnel syndrome. Clin. Neurophysiol. 110:1471, 1999. 209. Richardson, J. K., Forman, G. M., and Riley, B.: An electrophysiological exploration of the double crush hypothesis. Muscle Nerve 22:71, 1999. 210. Robertson, D., Aghasi, M., and Halperin, N.: The treatment of carpal tunnel syndrome caused by hypertrophied lumbrical muscles: case reports. Scand. J. Plast. Reconstr. Surg. Hand Surg. 23:149, 1989. 211. Roquelaure, Y., Mechali, S., Dano, C., et al.: Occupational and personal risk factors for carpal tunnel syndrome in industrial workers. Scand. J. Work Environ. Health 23:364, 1997. 212. Roquer, J., and Cano, J. F.: Carpal tunnel syndrome and hyperthyroidism: a prospective study. Acta Neurol. Scand. 88:149, 1993. 213. Rosecrance, J. C., Cook, T. M., Anton, D. C., et al.: Carpal tunnel syndrome among apprentice construction workers. Am. J. Ind. Med. 42:107, 2002. 214. Rosenbaum, R. B., Franklin, G. M., Clemmons, D. C., et al.: Carpal tunnel syndrome. In Mancall, E. L. (ed.): Continuum in Occupational Neurology. Philadelphia, Lippincott Williams & Wilkins, p. 712, 2001. 215. Roth, G., Ludy, F. P., and Egloff-Baer, S.: Isolated proximal median neuropathy. Muscle Nerve 5:247, 1982.

1460

Diseases of the Peripheral Nervous System

216. Rotman, M. B., and Donovan, J. P.: Practical anatomy of the carpal tunnel. Hand Clin. 18:219, 2002. 217. Sachs, G. M., Raynor, E. M., and Shefner, J. M.: The all ulnar motor hand without forearm anastomosis. Muscle Nerve 18:309, 1995. 218. Salvarani, C., and Hunder, G. G.: Musculoskeletal manifestations in a population-based cohort of patients with giant cell arteritis. Arthritis Rheum. 42:1259, 1999. 219. Schantz, K., and Riegels-Nielsen, P.: The anterior interosseous nerve syndrome. J. Hand Surg. Br. 17:510, 1992. 220. Schon, R., Kraus, E., Baller, O., et al.: Anomalous muscle belly of the flexor digitorum superficialis associated with carpal tunnel syndrome. Neurosurgery 31:969, 1992. 221. Schottland, J. R., Kirschberg, G. J., Fillingim, R., et al.: Median nerve latencies in poultry processing workers: an approach to resolving the role of industrial “cumulative trauma” in the development of carpal tunnel syndrome. J. Occup. Med. 33:627, 1991. 222. Schrader, P. A., and Reina, C. R.: Struthers’ ligament neuropathy in a juvenile. Orthopedics 17:723, 1994. 223. Seiler, J. G. 3rd, Milek, M. A., Carpenter, G. K., et al.: Intraoperative assessment of median nerve blood flow during carpal tunnel release with laser Doppler flowmetry. J. Hand Surg. Am. 14:986, 1989. 224. Seror, P.: Electrodiagnostic examination of the anterior interosseus nerve: normal and pathologic data (21 cases). Electromyogr. Clin. Neurophysiol. 39:183, 1999. 225. Seror, P., Leger, J. M., and Maisonobe, T.: Anterior interosseous nerve and multifocal motor neuropathy. Muscle Nerve 26:841, 2002. 226. Siegel, D. B., Kuzma, G., and Eakins, D.: Anatomic investigation of the role of the lumbrical muscles in carpal tunnel syndrome. J. Hand Surg. Am. 20:860, 1995. 227. Silbermann-Hoffman, O., Touam, C., Miroux, F., et al.: Contribution of magnetic resonance imaging for the diagnosis of median nerve lesion after endoscopic carpal tunnel release. Chir. Main 17:291, 1998. 228. Simpson, J. A.: Electrical signs in the diagnosis of carpal tunnel and related syndromes. J. Neurol. Neurosurg. Psychiatry 19:275, 1956. 229. Singer, G., and Ashworth, C. R.: Anatomic variations and carpal tunnel syndrome: 10-year clinical experience. Clin. Orthop. 392:330, 2001. 230. Solomon, D. H., Katz, J. N., Bohn, R., et al.: Nonoccupational risk factors for carpal tunnel syndrome. J. Gen. Intern. Med. 14:310, 1999. 231. Spindler, H. A., and Dellon, A. L.: Nerve conduction studies and sensibility testing in carpal tunnel syndrome. J. Hand Surg. Am. 7:260, 1982. 232. Spinner, R. J., Carmichael, S. W., and Spinner, M.: Partial median nerve entrapment in the distal arm because of an accessory bicipital aponeurosis. J. Hand Surg. Am. 16:236, 1991. 233. Stahl, S., Blumenfeld, Z., and Yarnitsky, D.: Carpal tunnel syndrome in pregnancy: indications for early surgery. J. Neurol. Sci. 136:182, 1996. 234. Stahl, S., Yarnitsky, D., Volpin, G., et al.: Conservative therapy in carpal tunnel syndrome. Harefuah 130:241, 1996.

235. Steinback, L. S., and Smith, D. K.: MRI of the wrist. J. Clin. Imag. 24:298, 2000. 236. Stevens, J. C.: AAEE minimonograph #26: the electrodiagnosis of carpal tunnel syndrome. Muscle Nerve 10:99, 1987. 237. Stevens, J. C.: AAEM minimonograph #26: the electrodiagnosis of carpal tunnel syndrome. Muscle Nerve 20:1477, 1997. 238. Stevens, J. C., Beard, C. M., O’Fallon, W. M., et al.: Conditions associated with carpal tunnel syndrome. Mayo Clin. Proc. 67:541, 1992. 239. Stevens, J. C., Smith, B. E., Weaver, A. L., et al.: Symptoms of 100 patients with electromyographically verified carpal tunnel syndrome. Muscle Nerve 22:1448, 1999. 240. Stevens, J. C., Sun, S., Beard, C. M., et al.: Carpal tunnel syndrome in Rochester, Minnesota, 1961 to 1980. Neurology 38:134, 1988. 241. Stevens, J. C., Witt, J. C., Smith, B. E., et al.: The frequency of carpal tunnel syndrome in computer users at a medical facility. Neurology 56:1568, 2001. 242. Stock, S. R.: Workplace ergonomic factors and the development of musculoskeletal disorders of the neck and upper limbs: a meta-analysis. Am. J. Ind. Med. 19:87, 1991. 243. Stoll, C., and Maitrot, D.: Autosomal dominant carpal tunnel syndrome. Clin. Genet. 54:345, 1998. 244. Sunderland, S.: The nerve lesion in the carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 39:615, 1976. 245. Suranyi, L.: Median nerve compression by Struthers ligament. J. Neurol. Neurosurg. Psychiatry 46:1047, 1983. 246. Swoboda, K. J., Engle, E. C., Scheindlin, B., et al.: Mutilating hand syndrome in an infant with familial carpal tunnel syndrome. Muscle Nerve 21:104, 1998. 247. Takahashi, T., Kato, A., Ikegaya, N., et al.: Ultrasound changes of the carpal tunnel in patients receiving long-term hemodialysis: a cross-sectional and longitudinal study. Clin. Nephrol. 57:230, 2002. 248. Thomas, P. K., and Fullerton, P. M.: Nerve fiber size in the carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 26:520, 1963. 249. Todnem, K., and Lundemo, G.: Median nerve recovery in carpal tunnel syndrome. Muscle Nerve 23:1555, 2000. 250. Topmaino, M. M., and Weiser, R. W.: Carpal tunnel release for advanced disease in patients 70 years and older: does outcome from the patient’s perspective justify surgery? J. Hand Surg. Br. 26:481, 2001. 251. Toranto, I. R.: Aneurysm of the median artery causing recurrent carpal tunnel syndrome and anatomic review. Plast. Reconstr. Surg. 84:510, 1989. 252. Vadasz, A. G., Chance, P. F., Epstein, L. G., et al.: Familial autosomal-dominant carpal tunnel syndrome presenting in a 5-year-old-case report and review of the literature. Muscle Nerve 20:376, 1997. 253. Walker, W. C., Metzler, M., Cifu, D. X., et al.: Neutral wrist splinting in carpal tunnel syndrome: a comparison of nightonly versus full-time wear instructions. Arch. Phys. Med. Rehabil. 81:424, 2000. 254. Watson, B. V., Parkes, A., and Brown, J. D.: Transient forearm conduction block in the median nerve. Muscle Nerve 25:456, 2002. 255. Werner, C.-O., Elmqvist, D., and Ohlin, P.: Pressure and nerve lesion in the carpal tunnel. Acta Orthop. Scand. 54:312, 1983.

Median Neuropathy 256. Werner, R. A., Albers, J. W., Franzblau, A., et al.: The relationship between body mass index and the diagnosis of carpal tunnel syndrome [see comments]. Muscle Nerve 17:632, 1994. 257. Werner, R. A., and Andary, M.: Carpal tunnel syndrome: pathophysiology and clinical neurophysiology. Clin. Neurophysiol. 113:1373, 2002. 258. Werner, R. A., Gell, N., Franzblau, A., et al.: Prolonged median sensory latency as a predictor of future carpal tunnel syndrome. Muscle Nerve 24:1462, 2001. 259. Wilson, J. R., and Sumner, A. J.: Immediate surgery is the treatment of choice for carpal tunnel syndrome. Muscle Nerve 18:660, 1995. 260. Winkelmann, R. K., and Stevens, J. C.: Successful treatment response of granuloma annulare and carpal tunnel

261.

262.

263.

264.

1461

syndrome to chlorambucil. Mayo Clin. Proc. 69:1136, 1994. Witt, J. C., and Stevens, J. C.: Neurologic disorders masquerading as carpal tunnel syndrome: 12 cases of failed carpal tunnel release. Mayo Clin. Proc. 75:409, 2000. Wong, S. M., Griffith, J. F., Hui, A. C. F., et al.: Discriminatory sonographic criteria for the diagnosis of carpal tunnel syndrome. Arthritis Rheum. 46:1914, 2002. Wong, S. M., Hui, A. C. F., Tang, A., et al.: Local vs systemic corticosteroids in the treatment of carpal tunnel syndrome. Neurology 56:1565, 2001. You, H., Simmons, Z., Freivalds, A., et al.: Relationships between clinical symptom severity scales and nerve conduction measures in carpal tunnel syndrome. Muscle Nerve 22:497, 1999.

60 Upper Limb Neuropathies: Long Thoracic (Nerve to the Serratus Anterior), Suprascapular, Axillary, Musculocutaneous, Radial, Ulnar, and Medial Antebrachial Cutaneous ASA J. WILBOURN AND MARK A. FERRANTE

Pathophysiology Long Thoracic Neuropathy (Lesion of Nerve to the Serratus Anterior) Anatomy Etiologies Clinical Manifestations Diagnosis Treatment and Prognosis Suprascapular Neuropathy Anatomy Etiologies Clinical Manifestations Diagnosis Treatment and Prognosis

Axillary Neuropathy Anatomy Etiologies Clinical Manifestations Diagnosis Treatment and Prognosis Musculocutaneous Neuropathy Anatomy Etiologies Clinical Manifestations Diagnosis Treatment and Prognosis Radial Neuropathy Anatomy

Mononeuropathies of the upper limb are more common than are those of the lower limb. Regardless of the type and location of institution surveyed, and the population studied— military personnel in hospitals in England in World War I or in Australia in World War II, adults in a general neurology and neurosurgery clinic in Switzerland, or athletes in an electrodiagnostic (EDX) laboratory in the United States— the ratio of peripheral nerve disorders affecting the upper limb versus the lower limb is approximately 3:1 or 4:1.2,44,48,63

PATHOPHYSIOLOGY The underlying pathophysiology of focal peripheral nerve lesions determines their clinical and electrophysiologic

Etiologies Clinical Manifestations Diagnosis Treatment and Prognosis Ulnar Neuropathy Anatomy Etiologies Clinical Manifestations Diagnosis Treatment and Prognosis Medial Antebrachial Cutaneous Neuropathy

manifestations, and their prognosis. Peripheral nerve axons have a very limited repertoire of responses to any type of localized insult. Whenever both myelinated and unmyelinated nerve fibers sustain a sufficient degree of injury, they are interrupted at the site of the lesion; their distal segments then undergo wallerian degeneration, thereby becoming incapable of transmitting nerve impulses—that is, they suffer conduction failure. Depending upon the percentage of axons involved, the clinical deficits that result may be partial or complete. Regarding the large myelinated axons, a substantial loss of motor fibers results in paresis or paralysis, soon followed by atrophy, whereas a similar loss of sensory fibers compromises appreciation of position, vibration, and light touch. With loss of unmyelinated and lightly myelinated 1463

1464

Diseases of the Peripheral Nervous System

sensory axons, pain and temperature sensations are impaired. If large myelinated fibers sustain lesser degrees of focal damage than that necessary to interrupt them, often only their myelin is damaged at the site of injury. This can result in focal conduction abnormalities limited to that injured segment, consisting of either slowing or block. With demyelinating conduction slowing, all nerve impulses traverse the damaged portion of nerve, but their speed is reduced as they do so; nonetheless, because they all ultimately reach their destinations, there are few, if any, clinical manifestations. With demyelinating conduction block, a more severe type of focal abnormality, impulse transmission across the injured area of the nerve is stopped, rather than merely slowed. As with axon loss, demyelinating conduction block can affect any percentage of the nerve fibers. Its clinical manifestations are identical to those seen with the same degree of axon loss, except that pain and temperature sensations are not altered.77

LONG THORACIC NEUROPATHY (LESIONS OF NERVE TO THE SERRATUS ANTERIOR) Anatomy The nerve fibers composing the long thoracic nerve derive from the C5, C6, and C7 ventral rami. Typically, the C5 and C6 branches pass through the scalenus medius muscle and then fuse, whereas the branch from C7 passes anterior to that muscle and joins the others at the level of the first thoracic rib. The nerve descends along the anterolateral aspect of the thoracic wall, on the surface of the serratus anterior, to the level of the eighth or ninth rib, providing muscular branches to each digitation of the serratus anterior muscle (Fig. 60–1). In general, the more superior digi-

FIGURE 60–1 The long thoracic nerve.

tations are supplied by the C5 fibers, whereas the more inferior slips are supplied by C7.17 Along with the trapezius, levator scapulae, and rhomboid muscles, the serratus anterior is a fixator or stabilizer of the scapula. It stabilizes the scapula as the arm is abducted to 90 degrees, and then helps laterally rotate the inferior angle of the scapula as the arm is further abducted to 180 degrees (straight overhead). It also approximates the scapula against the posterior chest wall whenever the arm is pushed forward against resistance.17,39

Etiologies Long thoracic neuropathies may be of traumatic, nontraumatic, and iatrogenic origin. Traumatic injuries, in turn, may be open or closed in nature. Open lesions are rare. Birch and co-workers8 described repairing stab wounds of the long thoracic nerve in five patients, while Seddon58 reported seeing a single patient with a gunshot wound of the nerve. Most closed traumatic lesions are due to either blunt trauma or traction.42 The nerve is susceptible to blunt force because of its superficial position along the chest wall, and to traction because it is anchored where it angulates to enter the axilla and begins its descent along the lateral thoracic wall.7 Abrupt, forceful upper limb exertion, with the arm suddenly being jerked forward or backward, is a common cause of injury. Excessive usage during strenuous work or sporting activities, such as shoveling dirt, chopping wood, and participation in weight lifting, golf, tennis, gymnastics, and wrestling, has produced lesions of this nerve.69,70,73 Nontraumatic insults include entrapment, compression, and neuralgic amyotrophy. Entrapment of the long thoracic nerve is quite uncommon, but it can occur as a result of the action of the scalenus medius in patients who lift heavy weights or push heavy objects.8 Another uncommon cause is compression by the shoulder straps of backpacks and during prolonged bed rest.49,69 The most common nontraumatic etiology is neuralgic amyotrophy. With this disorder, the nerve may be affected with other nerves (especially the suprascapular), but frequently it is involved in isolation. Thus, among the 136 patients reported by Parsonage and Turner, 30 had only lesions of the long thoracic nerve.51 Iatrogenic causes usually are related to surgical procedures in which the nerve sustains direct injury, such as first rib resections, scalenotomies, mastectomies, thoracotomies, and transaxillary sympathectomies.23,33,61,69,73 Unfortunately, when long thoracic neuropathies manifest in the postoperative period in situations in which direct operative nerve injury is quite unlikely, they may be attributed to patient malpositioning on the operating table, even though neuralgic amyotrophy is actually responsible. Occasionally, no cause for lesions of this nerve is evident. In these instances, when the patients are young

1465

Upper Limb Neuropathies

adults, it is tempting to assume that their neuropathies resulted from a painless bout of neuralgic amyotrophy (or that pain was present but subsequently forgotten). However, unless other nerves also are affected, this supposition cannot be proven.

Clinical Manifestations Long thoracic neuropathies cause serratus anterior muscle weakness that presents with shoulder movement limitations, alterations in the contour of the posterior thorax, and sometimes shoulder aching. The upper limb movements most impaired are forward flexion and abduction, especially above shoulder level. The posterior thorax contour changes consist of alterations in the orientation of the scapula. At rest, the scapula is translated medially and superiorly, its inferior angle is rotated medially, and its vertebral border is displaced away from the posterior chest wall (resulting in scapular “winging”). These neuropathies often are painless, although a dull aching pain in the shoulder region may be noted.33,61,70,73 When caused by neuralgic amyotrophy, however, they typically are preceded by severe pain, usually of abrupt onset and of 7 to 10 days’ duration, which most often is located either over the lateral deltoid or over the course of the nerve along the lateral aspect of the thorax, or both.

Diagnosis Long thoracic neuropathies should be considered whenever certain shoulder movements, such as arm abduction, are limited. The major clinical finding is winging of the scapula, which may be apparent when the patient is at rest but becomes far more prominent when the patient pushes forward forcibly against a vertical surface.49,54 Neuroimaging studies may aid in diagnosis. In the setting of trauma, plain radiographs may reveal fractures or foreign bodies. Unless competing diagnoses (such as radiculopathies and plexopathies) require exclusion, computed tomography (CT) scanning and magnetic resonance imaging (MRI) usually are not performed.33,54 On EDX examination, typically only needle electromyography (EMG) assessment of the serratus anterior is performed. Usually the muscle is sampled at two or more levels. Although motor nerve conduction studies (NCSs) can be performed, they are of limited value because the serratus anterior muscle consists of a series of individual digitations rather than a single common muscle mass. Consequently, the amplitude of any compound muscle action potential (CMAP) recorded represents activation of only a portion of the muscle.33,54,78 Because the serratus anterior muscle is a shoulder stabilizer, the differential diagnosis of long thoracic neuropathies includes lesions of a number of other nerves that supply shoulder girdle muscles, such as the trapezius, deltoid, and

infra- and supraspinatus, which may appear to be weak. That these muscles are actually normal can be demonstrated by clinical examination and needle EMG. Authentic pathologic conditions that can be confused with long thoracic neuropathies include spinal accessory neuropathies, C7 radiculopathies, traumatic serratus anterior muscle avulsions resulting from scapular fractures, and certain muscular dystrophies. With spinal accessory neuropathies, both clinical and EDX examinations demonstrate that the upper trapezius, rather than the serratus anterior, is affected. With C7 radiculopathies, the serratus anterior seldom is severely weak and, invariably, other muscles innervated by the C7 root, particularly the triceps, are also involved; this can be demonstrated on both clinical and needle EMG examinations. Traumatic avulsions of the serratus anterior muscle are suspected based on the clinical history and the demonstration on radiography of a fractured scapula. Muscular dystrophy rarely affects only a single muscle, a fact apparent on both clinical and needle EMG assessments; moreover, needle EMG of the serratus anterior will reveal changes much different from those seen with long thoracic neuropathies.33,34,54,61,70

Treatment and Prognosis The treatment of long thoracic nerve lesions depends on their etiology, severity, and duration. The rare open lesions, resulting from either traumatic or iatrogenic causes, usually require surgery. Conservative management, at least initially, is employed with most other etiologies. Instigating activities (such as repetitive limb movements) are avoided, and both nonsteroidal antiinflammatory and neuropathic pain medications are used, as is physical therapy. Although braces and other orthotic devices are available for abutting the scapula to the thoracic wall, typically they are cumbersome and ineffective. Whenever shoulder function remains unsatisfactory after a reasonable observation period, operative procedures to stabilize the scapula can be considered.33,61,73 These can be divided into three categories: scapulothoracic fusion, static stabilization, and dynamic muscle transfer. Of these, dynamic muscle transfer procedures are considered to provide the most satisfactory permanent outcomes. Currently, transfer of the sternal head of the pectoralis major to the inferior angle of the scapula, with autograft extension (using either fascia lata or hamstring tendon), is probably the preferred treatment for chronic serratus anterior paralysis.61,70,73 In general, recovery with traumatic neuropathies is less satisfactory than with nontraumatic ones, and, as would be expected, it often is less satisfactory with complete than with incomplete lesions. Those neuropathies caused by overactivity may clear promptly, suggesting that demyelinating conduction block is the major pathophysiologic process. Those lesions resulting from neuralgic amyotrophy have a

1466

Diseases of the Peripheral Nervous System

good prognosis, but maximal improvement may require 2 years. Patients with isolated mild to moderate serratus anterior muscle weakness, and those who are elderly and inactive, often have satisfactory outcomes without operative intervention.30,33,73

SUPRASCAPULAR NEUROPATHY Anatomy The suprascapular nerve fibers derive from the C5 and C6 ventral rami and leave the upper trunk, just after the formation of the latter, as the suprascapular nerve. The nerve passes laterally across the posterior cervical triangle (anterior to the trapezius); traverses the suprascapular notch, located on the superior border of the scapula; and enters the supraspinatus fossa, on the dorsum of the scapula. It passes inferiorly in this fossa, deep to the supraspinatus muscle, which it innervates, and then supplies deep sensory fibers to the majority of the shoulder capsule. After winding around the lateral aspect of the spine of the scapula, in a space referred to as the spinoglenoid notch, it enters the infraspinatus fossa, where it innervates the infraspinatus muscle17,18,64 (Fig. 60–2). The suprascapular and spinoglenoid notches vary in their size and shape;

when bridged by ligaments, called the supraclavicular and spinoglenoid ligaments (also known as the superior and inferior transverse ligaments, respectively), they act as fibro-osseous tunnels. The supraspinatus muscle, along with the deltoid, abducts the arm. In addition, it is a weak external rotator and flexor of the arm. For many years, the supraspinatus muscle was thought to be responsible only for initiating arm abduction64; however, more recent research has demonstrated that it contributes a slight majority of the force required for the first 60 to 90 degrees of arm abduction, with the deltoid muscle supplying the majority of force beyond that range.18 The infraspinatus muscle acts as an external rotator of the arm, contributing approximately 50% of the force required for this movement, with the teres minor muscle supplying the remainder.18,64

Etiologies Isolated suprascapular neuropathies are relatively uncommon. They are less often responsible for weakness of the spinatus muscles than are cervical radiculopathies and upper trunk brachial plexopathies.18 Nonetheless, for uncertain reasons, far more publications have focused on lesions of this proximal upper limb nerve than any other.

FIGURE 60–2 The axillary and suprascapular nerves, whose actions often compliment one another.

Upper Limb Neuropathies

Disorders of the suprascapular nerve can be classified as traumatic, nontraumatic, and iatrogenic. Traumatic suprascapular neuropathies may be open or closed in nature. Although the suprascapular nerve is superficial in the posterior cervical triangle, and thereby vulnerable to penetrating injuries, open traumatic lesions of it are uncommon. Thus two large series reported that only 2 of 1022 and 16 of 3907 battle-related neuropathies involved the suprascapular nerve.18,64 Nonetheless, occasionally it is injured by blunt trauma in this region.18 Most traumatic injuries are related to closed trauma, either from a single event (such as a motorcycle accident), or from repetitive activity. The majority of these lesions are thought to occur at the suprascapular notch.3,4,13,18,33,46,47 Thus, among 11 patients with suprascapular nerve ruptures located at a single site, the injury occurred at the suprascapular notch in 8, proximal to that notch in 2, and at the spinoglenoid notch in only 1.47 Persons performing activities that involve extreme shoulder movements (which require wide scapular excursions) or some repetitive shoulder and upper arm movements are susceptible to suprascapular nerve injuries. These include construction workers, weight lifters, volleyball players, baseball pitchers, and athletes engaged in contact sports, such as football, hockey, and lacrosse.3,18,26,33,35,46,56,64 Presumably, the suprascapular nerve is injured in these situations because certain shoulder movements (especially depression, retraction, and hyperabduction) force the nerve to strike the sharp inferior edge of the suprascapular ligament.18 Anterior shoulder dislocations are another traumatic cause of suprascapular nerve lesions; only the axillary nerve is injured more frequently by this mechanism.18,61,80 Suprascapular neuropathies may be caused by scapular fractures, particularly those occurring near the suprascapular foramen; in these situations, the nerve may be injured immediately, or later as a result of excessive callus formation.3,7,18,61,67 Nontraumatic causes of suprascapular neuropathies include entrapment, compression, and neuralgic amyotrophy. Entrapment most commonly occurs at the suprascapular foramen; a variety of mechanisms are responsible, including ligamental abnormalities, such as tightness, hypertrophy, ossification, or calcification; excess callus formation; and various congenital anomalies, such as a small notch.13,18,46,61,68 At the spinoglenoid notch, compressive injuries related to adjacent masses predominate. These include cysts (which account for the majority), neoplasms, excess callus formation, hematomas, bone spurs, and familial calcifications. Paraglenoid cysts include synovial cysts, ganglion cysts, and pseudocysts. Synovial cysts are lined by synovial cells and form from evaginations of joint capsule or para-articular bursa. Ganglion cysts arise from joint capsules, bursae, ligaments, tendons, and subchondral bone. They are commonly found with degenerative or traumatic lesions of the glenoid capsular–labrum tissue; they probably develop when labral or capsular tears permit

1467

synovial fluid to be forced into the tissues, creating a one-way valve effect. Pseudocysts represent joint fluid extrusions into adjacent soft tissue, similar to meniscal cysts. Overall, ganglion cysts are the most common soft tissue lesions affecting the suprascapular nerve; most occur in males, a fact consistent with their presumed traumatic etiology.26,31,67 Neoplasms rarely are responsible for compressive suprascapular neuropathies. Thus, among 121 neural sheath tumors treated by Kline and Hudson, none involved the suprascapular nerve.42 Nonetheless, isolated examples have been reported; curiously, most of these have been malignant, such as Ewing’s sarcoma, synovial sarcoma, chondrosarcoma, and metastatic renal cell carcinoma.3,18,31 Neuralgic amyotrophy commonly affects the suprascapular nerve, either in isolation or along with other, predominantly motor, shoulder girdle region nerves.51 Many cases of isolated suprascapular neuropathy resulting from neuralgic amyotrophy are misdiagnosed as traumatic or entrapment in nature, especially when they are triggered by minor trauma. To compound the error, the characteristic features of suprascapular neuropathy caused by neuralgic amyotrophy—sudden, often nocturnal, onset of severe shoulder pain, followed shortly thereafter by muscle weakness and wasting—have been described as being typical of suprascapular nerve entrapment.42 Iatrogenic injuries of the suprascapular nerve have been produced by radical neck dissections, distal clavicular resections, rotator cuff tear repairs, and both open and arthroscopic shoulder surgeries.18,61 With those suprascapular neuropathies in which only the infraspinatus muscle is affected, the lesions most often are at the spinoglenoid notch region. The two most common causes are ganglion cysts and neuralgic amyotrophy.18,61 The nerve also may be entrapped in this area by a hypertrophied spinoglenoid ligament. These ligaments are more common among males and their incidence varies enormously (3% to 80%), depending upon how the term ligament is defined. Nonetheless, this hypertrophy seldom has been verified surgically.18,21,22 Often the underlying etiology cannot be determined whenever the infraspinatus muscle is affected in isolation, without the characteristic clinical presentation of neuralgic amyotrophy being present, and yet a structural lesion at the spinoglenoid notch is not demonstrable on neuroimaging studies. This presentation is common among dancers, baseball players, and volleyball players. A confusing aspect in this regard is that not all suprascapular neuropathies that cause solitary involvement of the infraspinatus muscle are located near the spinoglenoid notch. Thus there are documented examples of patients with nerve lesions near the suprascapular foramen in which the motor axons to the supraspinatus, but not the infraspinatus, muscle are spared.18 Moreover, isolated infraspinatus weakness at times is asymptomatic. Two studies in which asymptomatic top-level volleyball players were assessed found that 12% (12 of 96) to 25%

1468

Diseases of the Peripheral Nervous System

(4 of 16) of them had unilateral suprascapular neuropathies, many of which were severe and causing unappreciated weakness.29,79

Clinical Manifestations Suprascapular neuropathies present with weakness, often accompanied by pain and atrophy. With supraspinatus muscle involvement, the most obvious finding is impaired arm abduction, whereas with infraspinatus muscle weakness, external rotation of the upper limb is compromised. Atrophy of the supraspinatus muscle may be difficult to detect because of the overlying trapezius, but atrophy of the infraspinatus muscle usually is readily apparent. In general, with suprascapular neuropathies at the suprascapular notch, both spinatus muscles are involved, and often there is an accompanying dull, aching pain that is poorly localized to the posterior aspect of the shoulder and exacerbated by certain shoulder movements.3,18 With suprascapular nerve lesions at the spinoglenoid notch, usually only the infraspinatus muscle is affected, and the disorder may be asymptomatic: Often there is no pain, and the compromise of this muscle is not clinically apparent.18,31 Suprascapular neuropathies may manifest other types of pain than that described above, depending upon etiology. Thus, with neuralgic amyotrophy, the patient often is awakened by very severe pain that persists for several days and gradually resolves. Typically, this is interscapular in location, thereby mimicking the distribution of the pain characteristic of cervical radiculopathies. Conversely, with ganglion cysts, the pain may be abrupt in onset but dull and persistent, sometimes first appearing during vigorous use of the limb, such as with weight lifting.18,46

Diagnosis Suprascapular neuropathies are identified by the clinical examination, supplemented by neuroimaging and EDX studies. Adduction of the extended arm across the body may provoke or increase the pain, and in some patients tenderness is present over the suprascapular notch.53 Neuroimaging studies are of particular value with suprascapular neuropathies. Although the role of plain radiographs is limited, they can demonstrate anterior shoulder dislocations, fractures of the scapula and clavicle, bony tumors, and excess callus. Moreover, suprascapular notch views may determine the size of the foramen, although they usually fail to identify ligamental calcification. CT scanning can demonstrate osseous abnormalities and ligamental calcification, but its ability to recognize soft tissue masses, such as ganglion cysts, varies with the degree of attenuation produced by the mass. The two most helpful imaging studies for identifying soft tissue masses are MRI and ultrasonography.65 MRI can also demonstrate atrophy of one or both of the spinatus muscles and rotator cuff

tears, and, when ganglion cysts are present, it can reveal the posterior labral tear from which they arise. Shoulder arthrograms can establish rotator cuff integrity when MRI studies are ambiguous.18,26,31,33,36,59,61 Finally, involvement of the suprascapular nerve is presumed whenever injection of a local anesthetic agent into the suprascapular or spinoglenoid notch region produces complete pain relief.18,26,61 On EDX examination, the suprascapular nerve can be assessed with both motor NCSs and needle EMG. The underlying pathophysiology with most lesions is conduction failure resulting from axon degeneration. With suprascapular NCSs, surface recording electrodes are placed over the infraspinatus muscle, and the nerve fibers are stimulated in the supraclavicular fossa, where they are a component of the upper trunk. (Recording from the supraspinatus muscle is more technically difficult because of the overlying trapezius muscle.) Usually, suprascapular NCSs are performed bilaterally, and the amplitudes of the CMAPs are compared; on the symptomatic side, the response generally is very low in amplitude, and sometimes unelicitable.18,33,54,78 Although measuring the latency during NCSs, seeking evidence of conduction slowing, is mentioned in some articles,18,53,56 the reason for doing so is usually obscure; slowing is due to demyelination, and demyelinating conduction slowing never is responsible for weakness or atrophy. Both the supraspinatus and infraspinatus muscles can be assessed on needle EMG. Because the nerve lesions usually are severe, the neurogenic abnormalities present typically are obvious.18,33,78 A number of disorders are mistaken for suprascapular neuropathies, including C5-C6 radiculopathies, upper trunk brachial plexopathies, axillary neuropathies, and several orthopedic conditions, including rotator cuff tears, impingement syndrome, and adhesive capsulitis. Neuroimaging and EDX studies can eliminate essentially all of these. Thus, with both C5-C6 radiculopathies and upper trunk brachial plexopathies, needle EMG abnormalities typically will be found beyond the suprascapular nerve distribution. Moreover, with axon loss upper trunk brachial plexopathies, the sensory NCSs that assess upper trunk fibers—median, recording from the first digit, and lateral antebrachial cutaneous (LAC)—will be abnormal. With axillary neuropathies, EDX changes will be found in just an axillary nerve distribution. In contrast, with the various orthopedic (i.e., non-neurologic) entities included in the differential diagnosis, the EDX assessment of the suprascapular nerve, as well as other nerves about the shoulder girdle, will be normal, whereas MRI and other neuroimaging studies may reveal the actual lesion.18,28,33,61,68

Treatment and Prognosis The specific cause of the suprascapular nerve lesion is the most helpful factor in determining its most appropriate treatment. For some etiologies, however, there is lack of

1469

Upper Limb Neuropathies

agreement regarding which treatment is optimal, and when it should be performed. Conservative therapy includes resting the limb, activity modification (such as avoiding heavy lifting and not performing those activities that instigate or exacerbate symptoms), physical therapy, and pain management. The latter is accomplished with analgesics, antiinflammatories, and neuropathic pain medications, such as tricyclic antidepressants. Suprascapular or spinoglenoid notch injections of corticosteroids, local anesthetics, or both may be helpful.3,4,9,61 Neuralgic amyotrophy is always managed without surgery. In general, with closed lesions in which neuroimaging studies show no evidence of spaceoccupying masses, the initial treatment also is conservative. However, the optimal therapeutic approach to suprascapular neuropathies attributed to nerve entrapment at either the suprascapular or spinoglenoid notch remains controversial.13,18,46,68 Some investigators recommend early surgical intervention (within 1 to 2 months of onset) whenever the pain is severe, when weakness and atrophy are present, or when the diagnosis is definite68; others perform operations only when conservative treatment fails3; and still others essentially reject surgery altogether.10,46 Many investigators have recommended nonoperative treatment for those nerve lesions located at the spinoglenoid notch region, particularly since persistent infraspinatus muscle weakness and wasting generally cause little or no clinical deficit.18,21,29,79 Moreover, Antoniou and co-workers recently reported that suprascapular neuropathies related to traction and to overuse (apparently regardless of location) respond no better to surgery than to conservative therapy, so probably the latter treatment is indicated with them.4 (In all likelihood, the traction injuries in their series were not extremely severe, such as ruptures, which could improve only after surgical repair, if at all.) In contrast to the uncertainties of approach with presumed nerve entrapment, early operative intervention usually is mandated with open injuries of the suprascapular nerve.18,42 Similarly, space-occupying lesions, with the exception of some cysts, generally are treated with various invasive procedures, including surgery.4,31 The most appropriate treatment for paraglenoid cysts is unknown, in part because their natural history is unclear. Some are asymptomatic and spontaneously decompress or rupture, whereas other persist and slowly enlarge.31 When conservative therapy fails, invasive treatment options can be employed. These include percutaneous aspiration (under CT or ultrasonographic guidance), arthroscopic decompression, and open excision. Simple aspiration procedures may fail to drain the cyst and, even when successful, may yield only temporary relief, because the cyst may recur. Consequently, arthroscopic or open surgical procedures are preferred because they can remove not only the cyst but also its source.4,26,31,35,56,67 The prognosis of suprascapular neuropathies varies, depending principally on the etiology and severity of the

lesion. Overall, traumatic injuries tend to have a worse final outcome than those resulting from other causes.4 Most cases resulting from neuralgic amyotrophy recover considerably, but occasionally, with complete lesions, there is no improvement. Similarly, when total or near-total axon loss has resulted from nerve compression by cysts, weakness and wasting of the spinatus muscles, particularly the infraspinatus, may be permanent, although pain may be alleviated.4,18,33,61

AXILLARY NEUROPATHY Anatomy The axillary nerve (circumflex nerve) derives from the C5 and C6 ventral rami. Its fibers traverse the upper trunk and exit the posterior cord, as the axillary nerve, near the shoulder joint. This nerve has a short course. It leaves the axilla by passing posterolaterally through the quadrilateral (quadrangular) space—the anatomic space bordered by the surgical neck of the humerus laterally, the long head of the triceps medially, the teres minor superiorly, and the teres major inferiorly. While traversing this region, it innervates the teres minor. At the posterior aspect of the humeral neck, the axillary nerve divides into two branches, a longer anterior branch and a shorter posterior one. The anterior (lateral) branch passes laterally and winds anteriorly around the posterior and lateral aspects of the surgical neck of the humerus; while doing so, it innervates the middle and then the anterior heads of the deltoid. The posterior (medial) branch runs medial to the posterior head of the deltoid muscle, which it innervates, and then continues as the upper lateral brachial cutaneous nerve, which supplies cutaneous sensation to the lateral aspect of the deltoid17,64 (Fig. 60–2; see also Fig. 60–3). The teres major muscle externally rotates and weakly adducts the arm. The major function of the deltoid muscle is to abduct the arm, which it does in concert with the supraspinatus muscle. The anterior and posterior heads of the deltoid also forward flex and extend the arm, respectively.17,18,33,61

Etiologies Axillary neuropathies can be divided into traumatic (most common), nontraumatic, and iatrogenic causes. The majority of axillary neuropathies are due to trauma, especially shoulder region trauma (glenohumeral dislocations, proximal humeral fractures, and blunt blows to the shoulder), and therefore most are closed traction injuries.18,52,61 Of all the peripheral nerves, the axillary is the one most often affected with both shoulder dislocations and

1470

Diseases of the Peripheral Nervous System

FIGURE 60–3 The cutaneous distributions of the shoulder and upper limb nerves.

proximal humeral fractures.18,48,52 The nerve injuries in these situations can range from mild to extremely severe. Blunt trauma to the shoulder alone, without associated fractures or dislocations, also causes axillary neuropathies. These lesions are sustained when heavy objects strike the shoulder, during contact sports, and with falls onto the shoulder.7,52,61 Open injuries of the axillary nerve occur, such as from gunshot wounds and lacerations, but their reported incidence is low.64 Nontraumatic causes include compression resulting from space-occupying lesions such as pseudoaneurysms and hematomas; paraglenoid cysts; and neoplasms, both extrinsic (such as osteochondromas) and intrinsic (such as benign neural sheath tumors). All of these are uncommon etiologies; thus, in Kline and Hudson’s series of benign neural sheath tumors, the axillary nerve was affected in only 9 patients, compared to 35 patients with median nerve involvement.42 In contrast, neuralgic amyotrophy is the most common nontraumatic cause of axillary neuropathies, although these seldom happen in isolation.33,51 How commonly the axillary nerve is entrapped is debated, principally because the reputed cause of such entrapments, the quadrilateral space syndrome, has not been firmly established as a distinct entity. Most proponents believe that it results from entrapment of the axillary nerve and posterior humeral circumflex artery in the quadrilateral space, by either fibrous bands or muscular hypertrophy.18,59 However, the lack of objective findings in most patients, the nonanatomic distribution of the reported paresthesias, and an unclear pathogenesis have caused many physicians to be skeptical. Despite this, the quadrilateral space syndrome is diagnosed with some frequency in athletes, such as baseball pitchers and volleyball and tennis players.18,61

Iatrogenic axillary nerve injuries may occur during shoulder reconstruction procedures, rotator cuff tear repair, shoulder arthroscopy, reduction of humeral head dislocations, and breast surgery. Rare causes are malpositioning on the operating table and deep injections into the posterior deltoid region.18,33,58,62 Delayed presentations may follow operations performed to correct recurrent anterior shoulder dislocations, when the bone screw attaching the coracoid process to the glenoid rim becomes loose and pierces the axillary artery, producing a pseudoaneurysm.55 Approximately 11% of patients with medial brachial fascial compartment (MBFC) syndrome have axillary nerve involvement.43,66

Clinical Manifestations Axillary neuropathies typically present with weakness of arm abduction, and often muscle wasting (most notable in the lateral deltoid). Because of overlapping functions of other muscles, there may be no evident weakness of the teres minor muscle. Sensory symptoms (such as pain, paresthesias, or a deficit) over the lateral aspect of the shoulder, as well as subluxation of the humeral head, may also be noted. Even with complete axillary neuropathies, however, sensory changes sometimes are absent.18,33,61,64

Diagnosis Both neuroimaging and EDX studies are quite helpful in the assessment of axillary neuropathies. Neuroimaging studies are of particular benefit because so many of these nerve lesions are related to bony abnormalities about the shoulder, which are readily visualized. Thus both plain films

1471

Upper Limb Neuropathies

and CT scanning of the shoulder and humerus can demonstrate scapular or humeral fractures, glenohumeral dislocations, and, in a more chronic setting, callus formation. They also may detect osteophytes and bony tumors. MRI is the modality of choice for the recognition of soft tissue lesions, including concomitant injuries, such as rotator cuff tears. Occasionally, subclavian or axillary arteriograms and duplex scanning identify pseudoaneurysms or vascular occlusions.18,33,60,61 On EDX examination, the motor fibers, but not the sensory fibers, of the axillary nerve can be assessed. During an axillary motor NCS, the nerve fibers are stimulated supraclavicularly, where they are a component of the upper trunk, while a CMAP is recorded from the lateral deltoid muscle. On needle EMG, all three heads of the deltoid, as well as the teres minor muscle, may be sampled. Unfortunately, because the nerve fibers cannot be stimulated in the axilla (i.e., distal to the site of damage) during axillary motor NCSs, when acute lesions are assessed soon after onset, it is impossible to determine whether the pathophysiology is demyelinating conduction block or axon loss because the two processes have nearly identical EDX presentations under these circumstances.78 The differential diagnosis with axillary neuropathies includes orthopedic disorders affecting the shoulder (such as rotator cuff tears), suprascapular neuropathies, brachial plexopathies involving the upper trunk or posterior cord, and C5-C6 radiculopathies. With orthopedic (non-neurologic) conditions, sensory abnormalities are absent, and any muscle atrophy present is due to disuse, rather than denervation, a fact readily demonstrated on EDX examination. With suprascapular neuropathies, although shoulder abduction may be compromised, the clinical, and especially the EDX, examinations reveal the affected nerve. With brachial plexopathies, the sensory loss is more extensive and muscles other than the deltoid and teres minor are affected. The distinction is readily made by EDX examination. On NCSs, substantial upper trunk brachial plexopathies affect the median sensory nerve action potentials (SNAPs), recording from the first digit, and the LAC SNAPs (and sometimes the radial SNAPs), whereas posterior cord lesions produce abnormalities of the radial and posterior antebrachial cutaneous SNAPs. On needle EMG, muscles beyond the territory of the axillary nerve are affected, such as the brachioradialis, biceps, and pronator teres.18,27,33 With C5-C6 radiculopathies, ipsilateral interscapular pain is, or has been, present and myotomal changes usually are found on clinical and needle EMG examinations.

Treatment and Prognosis The appropriate treatment for axillary neuropathies varies, depending sometimes upon etiology but more

often upon the type of nerve damage present. Thus, among the nontraumatic lesions, neuralgic amyotrophy is managed conservatively, whereas neoplasms are treated surgically. Penetrating injuries, particularly those resulting from lacerations that produced immediate complete deficits, usually undergo early surgery. 42 However, for the majority of closed traumatic axillary neuropathies, such as those resulting from shoulder dislocations and blunt trauma, the underlying pathophysiology can vary from demyelinating conduction block (which recovers completely and rather rapidly) to total axon loss resulting from nerve rupture (which requires operative repair if any recovery is to be realized). Consequently, with most axillary neuropathies, particularly those resulting from trauma, conservative therapy is used initially. This consists of resting the limb, activity modification, physical therapy, and pain management. Operative intervention generally is considered whenever there is no improvement after a reasonable time period, when the symptoms are progressive, or when the mechanism of injury was such that spontaneous recovery is highly unlikely. The permissible delay between the onset of symptoms and operative treatment is controversial. In general, however, the outcome from surgical repair is better when it is performed within the first 6 months after injury, especially when grafting is necessary.7,18,33,42,60 Isolated lesions of the axillary nerve may have a satisfactory functional recovery despite persistent complete paralysis of the deltoid muscle.52 Conversely, when damage of this nerve is associated with other injuries that have compromised shoulder motion, such as rotator cuff tears and suprascapular neuropathies, recovery of the axillary nerve alone usually does not result in substantial restoration of shoulder function. Moreover, the development of adhesive capsulitis can seriously compromise functional outcome. For this reason, if the nerve is surgically repaired, mobilization by physical therapy must be introduced as early in the postoperative period as possible.18,33 Overall, the majority of axillary neuropathies related to shoulder dislocations or humeral fractures usually recover satisfactory function, as do many of those resulting from blunt trauma (particularly those sustained in sporting events).18,33,52,61

MUSCULOCUTANEOUS NEUROPATHY Anatomy The musculocutaneous nerve fibers derive from the C5 and C6 ventral rami. They traverse the upper trunk and much of the lateral cord of the brachial plexus, and then separate from the latter in the axilla, as the musculocutaneous nerve. Near its origin, the nerve pierces the

1472

Diseases of the Peripheral Nervous System

coracobrachialis, which it supplies. Then it crosses to the lateral aspect of the arm while traveling distally, between the biceps brachii and brachialis, both of which it innervates. At this point it becomes a pure sensory nerve, the LAC nerve, which penetrates the deep fascia just proximal to the elbow and lateral to the biceps tendon and passes distally in the forearm. The LAC nerve supplies cutaneous sensation to the lateral aspect of the forearm and deep sensation to the wrist and some of the carpal bones17,18 (Fig. 60–4; see also Fig. 60–3). Anastomoses between the LAC nerve and the superficial radial, posterior antebrachial cutaneous, and median nerves have been described.54

FIGURE 60–4 The musculocutaneous nerve and its sensory component, the lateral antebrachial cutaneous nerve.

The coracobrachialis muscle is a weak flexor and adductor of the arm, and the brachialis is a forearm flexor, while the biceps is both a forearm flexor and a weak supinator. Together, the biceps and brachialis contribute 75% of forearm flexion power.18

Etiologies Although the musculocutaneous nerve fibers are susceptible to injury at multiple points along their course, they are affected less often than are the other major upper extremity nerves. Isolated lesions of the musculocutaneous nerve itself (as opposed to lesions involving musculocutaneous nerve fibers when they are a component of the brachial plexus) are rare. Most reports of them in the literature concern single cases, or small series. Consequently, whenever biceps and brachialis weakness is recognized, it far more often is indicative of a plexus or cervical root lesion than of a musculocutaneous neuropathy.18,61 Lesions of both the musculocutaneous and the LAC nerves can be discussed in terms of traumatic, nontraumatic, and iatrogenic causation. Traumatic insults of the musculocutaneous nerve include blunt trauma, anterior shoulder dislocations, and displaced humeral and clavicular fractures. With the last two mechanisms, usually other proximal upper limb nerves also are involved, especially the axillary. Additional traumatic lesions include stab wounds and gunshot wounds, but both are rare. Thus, among 10 large series of gunshot wounds of nerves cataloged by Sunderland, musculocutaneous neuropathies were responsible for just 0.8% to 3.4% of the lesions.18,64 Two patients have been reported who manifested musculocutaneous neuropathies (one bilaterally) soon after sustaining severe traumatic thoracic spinal cord injuries. However, in neither patient was a definite cause determined, and the trauma itself appeared unlikely to have been responsible.1,25 Nontraumatic causes include vigorous, repetitive physical use of the upper limb, such as occurs with weight lifting, rowing, and various construction projects. Pressure on the nerve by a hypertrophied coracobrachialis muscle has been postulated as the underlying cause in these instances.9,18,61 Another nontraumatic cause is an abrupt, often violent, single movement of the arm and forearm, most often consisting of elbow extension. These movements can produce severe traction injuries, even ruptures, of the nerves. Carrying a roll of heavy carpet on the shoulder, with the arm curled about it, has produced a musculocutaneous neuropathy.18,61 Bony anomalies (such as humeral exostoses) and neoplasms (such as lipomas and neural sheath tumors) can damage the nerve, but these are quite uncommon.18,37 Thus, in a series of benign neural sheath tumors (schwannomas and neurofibromas) reported by Kline and Hudson, only 6 affected the musculocutaneous nerve, whereas 35 involved the median nerve.42 Approximately 18% of patients with MBFC syndrome have musculocutaneous nerve involvement, although

1473

Upper Limb Neuropathies

never in isolation.43,66 Similarly, neuralgic amyotrophy occasionally affects the musculocutaneous nerve, but only rarely in isolation.18,51 Iatrogenic musculocutaneous neuropathies most often are due to operative procedures. These include glenohumeral stabilization surgery, shoulder arthroscopy, shoulder replacement, and operations to correct anterior shoulder dislocations. Of the five terminal nerves of the brachial plexus, the musculocutaneous is the one most often damaged when anterior shoulder dislocations are treated surgically.18,55,61 The nerve also may be injured, in isolation, as a result of malpositioning of the patient during anesthesia. Typically the patients have been supine, with the affected limb abducted and externally rotated,18,61 that is, in the same position most often responsible for the more common type of malpositioning problem: “classic postoperative paralysis,” an upper trunk brachial plexopathy.28 Occasionally, the musculocutaneous nerve is injured by tight “figure-of-8” bandages used to treat clavicular fractures.18 Traumatic injuries of the LAC nerve are uncommon. The majority are of the open variety (gunshot wounds or lacerations), but closed blunt trauma to the elbow region is a reported cause. Among nontraumatic etiologies are prolonged or repeated compression by straps, as well as strenuous repetitive arm use, especially that which involves hyperextending the elbow. Reputedly the nerve is compressed or entrapped either between the brachialis and the biceps tendon or adjacent thickened fascia, or where it becomes superficial, lateral to the bicipital tendon.18,61 Very rarely the nerve is involved in isolation by neuralgic amyotrophy.24 In our experience, however, it is the upper limb sensory nerve that is most often affected with this disorder. Iatrogenic insults sustained near the elbow are the most common cause of LAC neuropathies. These include phlebotomies, catheterizations, placement of arteriovenous shunts, and other operative procedures. With these, the symptoms may appear at the time of injury (i.e., be due to direct injury) or be delayed (typically, be secondary to hematoma formation).18,57,61,76 Rarely, the LAC nerve is injured by tourniquet paralysis, but never alone.76

Clinical Manifestations The site of the nerve injury dictates the clinical findings. Proximally located musculocutaneous neuropathies produce weakness of forearm flexion and supination, as well as altered sensation along the radial aspect of the forearm. Sensory symptoms, such as pain and paresthesias, may or may not be present. With some nontraumatic lesions, for example, the only symptom may be weakness. When lesions are severe, biceps and brachialis muscle wasting may be obvious. The biceps deep tendon reflex usually is

diminished. With isolated lesions of the LAC nerve, only sensory abnormalities are noted, although tenderness to palpation or a Tinel’s sign may be elicited in the elbow region where the LAC nerve pierces the fascia.18,33,61,64

Diagnosis Both neuroimaging and EDX studies are used in the assessment of musculocutaneous and LAC neuropathies. In general, neuroimaging studies are of limited benefit in diagnosing isolated lesions of these nerves, because most are not due to trauma or space-occupying lesions. With acute trauma, plain radiographs of the humerus and shoulder, as well as CT scanning, can help exclude humeral fractures, shoulder dislocations, and other anatomic abnormalities. Shoulder MRI and arthrograms assess rotator cuff integrity; MRI also is the best modality for demonstrating neoplastic processes and some of the other nontraumatic causes. With LAC neuropathies, plain radiographs of the elbow (seeking evidence of fractures, dislocations, bone spurs, or degenerative joint changes), and MRI of the same region (seeking nonosseous abnormalities) may be helpful. The relief of sensory symptoms by an injection of local anesthetic at the site where the LAC nerve becomes superficial (i.e., at Olson’s point) has also been used as a diagnostic test.18,33,61 On EDX examination, the motor fibers of the musculocutaneous nerve are assessed on NCSs by stimulating them percutaneously in the axilla and in the supraclavicular fossa, while recording CMAPs from the biceps and brachialis muscle group. The fact that the motor nerve fibers can be stimulated at two points along their course allows demyelinating conduction block lesions situated along the distal upper trunk and proximal lateral cord of the brachial plexus to be detected. The LAC nerve can be evaluated by stimulating it near the elbow, while recording from it in the midforearm. Typically, with both axon loss musculocutaneous neuropathies and isolated LAC neuropathies (which customarily are due to axon loss), the LAC SNAP is unelicitable or, less often, low in amplitude. Both the biceps brachii and brachialis muscles are readily assessed on needle EMG.33,78 The differential diagnosis of musculocutaneous neuropathies includes ruptures of the biceps muscle or its tendon, lateral cord or upper trunk brachial plexopathies, and C5 or C6 radiculopathies. Biceps muscle rupture usually is painful and is accompanied by swelling, whereas distal biceps tendon rupture is painless, and manifested by contraction of the biceps into a small, firm ball with attempted forearm flexion. Sensory abnormalities are not present with either disorder.18,61 Brachial plexopathies manifest sensory abnormalities that extend beyond the LAC distribution, a fact readily demonstrated by performing a median sensory NCS, recording from the first digit. Moreover, muscles not innervated by the musculocutaneous nerve, such as the deltoid, brachioradialis, and

1474

Diseases of the Peripheral Nervous System

pronator teres, are also involved; this is usually apparent on clinical examination, but almost always it is obvious on needle EMG. With C5 or C6 radiculopathies, the symptoms include radiating ipsilateral neck and intrascapular pain, sensory abnormalities in a dermatomal (rather than just a LAC) distribution, and occasional myotomal weakness, as well as brachioradialis deep tendon reflex diminution. On EDX examination, the LAC NCS is normal, despite the sensory complaints in the cutaneous area of that nerve, and needle EMG often demonstrates changes in a myotomal distribution.28,33,61

Treatment and Prognosis The treatment of musculocutaneous neuropathies varies with the severity and location of the lesion and its presumed etiology. The majority of closed lesions are treated conservatively, at least initially. Many of these, such as those resulting from positioning during anesthesia, have excellent recovery potential because they are principally demyelinating conduction blocks (i.e., they are neurapraxic in nature). Even when caused by axon loss, such as those resulting from neuralgic amyotrophy, the recovery usually is ultimately satisfactory, at least in regard to the restoration of forearm flexion. However, with some closed lesions, such as those caused by fractures, violent limb exertion, and neoplasms, operative treatment typically is mandatory, just as it is with nearly all open injuries of the nerve. When nerve grafting procedures are required to restore forearm flexion, the outcome may be excellent. This probably reflects the simple function (flexion and supination of the forearm) and architecture (small number of fibers and fascicles) of the musculocutaneous nerve.18 With closed lesions of the LAC nerve, treatment initially is conservative (such as avoidance of precipitating activity, local anesthetic or steroid injections, ultrasound, slings, and splinting). In refractory cases, or when progressive worsening is documented, surgical decompression of the nerve at the bicipital tendon often relieves symptoms.18,33,61 Those iatrogenic LAC neuropathies located near the elbow that follow invasive procedures, such as phlebotomy, usually are initially treated conservatively, unless hematoma is evident. However, they often result in persistent pain and, consequently, ultimately require surgical treatment. Even then, permanent sensory loss, sometimes accompanied by pain, can occur.18

RADIAL NEUROPATHY Anatomy The radial nerve fibers derive from the C5 through C8 ventral rami. They traverse all three trunks and posterior divisions and then converge in the posterior cord. The lat-

ter terminates in the axilla by dividing into the axillary and radial terminal nerves. The radial nerve passes distally along the lateral wall of the axilla, on the medial humerus. From this proximal nerve segment arise (1) motor branches that innervate the various heads of the triceps muscle and the anconeus muscle, and (2) the posterior brachial cutaneous nerve, which supplies a narrow vertical strip of skin on the back of the arm. The radial nerve then passes obliquely around the humerus, in the spiral groove. In this region, two sensory nerves originate from it: the lower lateral brachial cutaneous and the posterior antebrachial cutaneous, which provide sensation to the anterolateral aspect of the distal two thirds of the arm and to a narrow longitudinal strip along the back of the forearm, respectively. Distal to the spiral groove, the radial nerve pierces the lateral intramuscular septum to enter the anterior compartment of the arm. Continuing its distal course, it supplies motor branches to the brachioradialis and the extensor carpi radialis longus. Near the elbow, the radial nerve bifurcates into its two terminal divisions: the posterior interosseous nerve (PIN), a deep motor branch; and the superficial radial nerve (SRN), a cutaneous sensory branch. Thus the radial nerve overall extends from the distal axilla to the elbow. The PIN passes around the neck of the radius, innervates the extensor carpi radialis brevis, and then enters the extensor compartment of the forearm by penetrating the supinator muscle, which it supplies; the proximal opening in the supinator that it traverses is called the arcade of Frohse. The PIN then passes distally in the forearm, innervating the extensor carpi ulnaris, abductor pollicis longus, and extensor muscles of all the digits. The SRN travels distally in the forearm, becoming superficial in the distal one third. On reaching the lateral aspect of the dorsum of the wrist, it terminates by dividing into branches that supply the lateral aspect of the dorsum of the hand and the dorsal regions of the first three digits16,17,19,54 (Fig. 60–5; see also Fig. 60–3).

Etiologies The radial nerve and its two terminal branches may be injured at any point along their lengths, but certain sites are far more common than others; for the radial nerve, this is the spiral groove; for the PIN, near the elbow; and for the SRN, near the wrist. Along with the ulnar nerve, the radial nerve is the most commonly injured peripheral nerve among military personnel during wartime.64 Occasionally, radial nerve fibers in the axilla are injured where they are a component of the distal brachial plexus, such as from crutch use or from shoulder dislocations. Focal damage to the most proximal portion of the radial nerve proper, in the upper arm, is quite uncommon. Two causes include gunshot wounds and compression of the nerve against the medial aspect of the humerus by the edge of a firm surface that the limb is draped over for

Upper Limb Neuropathies

FIGURE 60–5 The radial nerve and its two terminal divisions, the posterior interosseous nerve and superficial radial nerve. (The sensory nerves arising from the radial nerve in the axilla and arm are not shown.)

prolonged periods (such as a chair), or by the head of a sleeping partner. A far more common site for occurrence of radial neuropathies is at or near the spiral groove (RN-SG). Trauma is responsible for many of the lesions at this location, and the most common traumatic cause is humeral fractures; approximately 11% to 20% of the latter produce radial neuropathies. There are several mechanisms of injury in these situations, including those occurring immediately (direct blunt trauma, traction resulting from separation of bone fragments); at the time of surgical repair (entrapment between the two ends of approximated bone, compression beneath the metal plates used to secure the bones); or still later (compression resulting from callus formation, trauma at the time of removal of hardware).19,42,62 Other causes for traumatic RN-SG are lacerations (most often caused by knives or glass), as well as gunshot wounds and contusions resulting from blunt injury; both of the latter may occur with or without associated humeral fractures.6,19,42 Nontraumatic causes for RN-SG include external compression against the lateral aspect of the arm. Frequently

1475

these are due to the patient lying on the arm for prolonged periods while asleep or in coma, with the latter induced by head injury, alcohol, drugs, or general anesthesia. Military personnel have sustained these lesions because of pressure being placed on the nerve by rifle slings and by the position assumed during prolonged rifle firing. Noteworthy is that hereditary neuropathy with a predisposition to pressure palsies very often produces RN-SG; conversely, neuralgic amyotrophy only occasionally does so. The triceps muscle is a source of internal compressive lesions, both of a chronic and of an acute nature. With the latter, nerve injuries have resulted from strenuous repetitive use of the limb (so-called exercise-induced radial neuropathies) and from a single sudden forceful or violent arm movement. Neoplasms, most often primary to the nerve sheath, are a less common source.6,19,61 Nonetheless, in Kline and Hudson’s series, approximately 10% of all benign neural sheath tumors affecting upper limb nerves involved the radial nerve.42 Iatrogenic causes for RN-SG include injection injuries (characteristically caused by misdirected intramuscular deltoid injections) and tourniquet paralysis.19 Typically, tourniquets are placed about the arm to produce a bloodless field for subsequent forearm and hand surgery. Almost invariably, tourniquet paralysis affects either the radial nerve alone, or the radial, ulnar, and median nerves together (referred to as tri-nerve paralysis).76 At times the radial nerve is injured in the distal arm or at the elbow. The basic causes are essentially the same as those responsible for RN-SG, including fractures (both of the distal humerus and the elbow), blunt contusions, gunshot wounds, lacerations from knives and glass, soft tissue masses (such as neoplasms and cysts), and misdirected drug injections. Other etiologies are elbow dislocations and Volkmann’s ischemic contracture (usually resulting from supracondylar humeral fractures and dislocations of the elbow).19,42,62 Most lesions of the PIN occur near its origin, at the elbow region. Dislocations and fractures are principle causes. Monteggia’s fractures are the most common of these: the head of the radius is dislocated posteriorly, while the proximal portion of the ulna is fractured. Less often, the proximal portion of the radius itself is fractured. With these injuries, the nerve can sustain both primary damage (blunt contusion trauma) and secondary damage, from bone fragments and during operative treatment. Moreover, the PIN can be damaged by deformed elbow joints resulting from remote fractures or osteomyelitis; nerve injuries from the latter are sometimes referred to as tardy posterior interosseous neuropathies because they can develop years after the bony lesions are sustained. Occasionally, the PIN is injured more distally, by midforearm fractures of the radius and ulna. It may be damaged at any point along its course by gunshot wounds and lacerations; the former are more common.

1476

Diseases of the Peripheral Nervous System

Space-occupying lesions affect the PIN more than the radial nerve proper. These include neoplasms (particularly lipomas), cysts, ganglia, and rheumatoid elbow synovitis. Neuralgic amyotrophy is another nontraumatic cause. In many patients, progressive abnormalities develop in the PIN distribution without evident source. Probably the majority of these are insidious in onset, and ultimately (after surgical exploration) are attributed to compression of the nerve by various structures near the elbow. The most frequently mentioned of these is the fibrous edge of the supinator muscle, at the arcade of Frohse; others include compression by the supinator muscle itself and by various fibrous bands in the region.6,19,59,61 Iatrogenic causes of PIN lesions, which usually affect the proximal portion of the nerve, include elbow arthroscopy, resections of the radial head, and insertion of plates during fracture treatment.19,42,61 Historically, the first recorded injection injuries of peripheral nerves were lesions of the PIN resulting from ether injections into the proximal extensor forearm to treat the pneumonia associated with smallpox. However, this injection site has not been used for more than a century.76 Relatively rare iatrogenic causes for PIN dysfunction include external compression resulting from the use of Canadian crutches and progressive scarring following radial head resection.61 Occasionally, no apparent etiology for PIN lesions is detected despite many diagnostic procedures being performed, including neuroimaging studies and surgical exploration.6,19,61 Disorders of the SRN branch of the radial nerve are less common than are those of the PIN. In most instances the nerve is injured distally, near the wrist. Trauma is a rather common etiology; lacerations are more frequent than blunt contusion injuries. External compression can produce SRN lesions, typically by the wrist being encircled by various structures, such as tight casts, watch straps, or restraining devices (handcuffs or ropes). Relatively rare causes for nontraumatic SRN injuries include space-occupying lesions (both neoplasms and ganglia) and entrapment of the nerve where it becomes subcutaneous. Iatrogenic causes include surgical operations, the most common apparently being de Quervain’s tenosynovectomy, and particularly needles inserted near the wrist, seeking the nearby vein.19,61 Radial Tunnel Syndrome This controversial disorder, also known as resistant tennis elbow, is essentially a pain syndrome attributed to compression of the PIN by the extensor muscles in the proximal forearm. It reportedly presents with pain and tenderness in the proximal extensor forearm muscle mass. The pain increases on active forearm supination against resistance, and on resisted active extension of the third digit. There is no weakness.59 Distinguishing this entity from lateral epicondylitis can be very difficult, and the EDX examination

usually is normal. Moreover, the results of surgical treatment have been unimpressive. Finally, the concept that compromise of essentially a “pure” motor nerve can manifest with sensory but not motor symptoms is a difficult one to accept. For these reasons, the existence of radial tunnel syndrome is questioned by many investigators.6,19,61

Clinical Manifestations Lesions of the radial nerve and its two terminal branches may present with motor or sensory abnormalities, alone or in various combinations. The most apparent of these is weakness. Radial neuropathies in the upper arm, when severe, present with weakness of all muscles innervated by the radial nerve and the PIN (i.e., with weakness of elbow extension, wristdrop, and finger drop), and sensory changes in the distribution of the three sensory nerves that arise from the radial nerve. With RN-SG, the symptoms are similar to those described above, except that elbow extension (triceps function) is preserved, and sensation may be intact in the posterior brachial cutaneous distribution. With lesions of the PIN, most of which affect the nerve near its origin, there is finger drop and lateral wrist deviation, resulting from the extensor carpi ulnaris, but not the extensor carpi radialis longus, being paralyzed. Sensation, however, is intact. With lesions of the SRN, sensory changes are present over the lateral aspect of the dorsum of the hand and fingers, but there is no weakness.16,19,54,61

Diagnosis Disorders of the radial nerve and its branches are diagnosed mainly by the history and clinical examination (especially muscle testing), supplemented by EDX and neuroimaging studies. Generally, these lesions can be localized accurately, because a number of motor branches arise at fairly regular intervals from the radial nerve and its forearm motor continuation, the PIN, and most focal lesions are situated along a segment of nerve between two such points of origin. Two deep tendon reflexes can aid in assessing radial neuropathies: the triceps and brachioradialis. For the former to be affected, the lesion usually must be in or near the axilla—that is, proximal to the motor branches supplying the triceps muscle; for the latter, it must be at or proximal to the spiral groove.19 Some caveats are necessary regarding localization. First, with lesions at the spiral groove or more proximal, muscle examination can suggest, erroneously, coexisting involvement of the median and ulnar nerves. The muscles innervated by these nerves can appear weak because of the complicated mechanics of hand and finger movements.16,45,49 Second, although severe radial neuropathies cause paralysis of extension of the wrist and of finger

Upper Limb Neuropathies

extension at the metacarpophalangeal joints, they do not compromise finger extension at the interphalangeal joints, because those movements are controlled by the median and ulnar nerves. Third, the zone of distribution of the SRN can vary considerably; often some, and occasionally all, of its usual territory is usurped by the LAC nerve or the dorsal ulnar cutaneous nerve. The sensory supply to the dorsum of the hand is one of the most variable in the body.19 Various neuroimaging studies and the EDX examination are helpful in assessing radial neuropathies. Plain radiographs are of particular help. They can demonstrate fractures of long bones (humerus, ulna, or radius) and of the elbow; excess callus formation following fractures; elbow dislocations; various metallic objects (e.g., bullet fragments), and lipomas (usually affecting the proximal PIN). Most neoplasms and cysts can be demonstrated with MRI and CT scans. Ultrasonograms are useful for detecting ganglion cysts affecting the proximal PIN.19,61 The radial nerve and its terminal branches can be thoroughly assessed by EDX examination. On NCSs, CMAPs can be elicited both from the distal forearm extensor muscles (such as the extensor pollicis brevis and extensor indices proprius), which are innervated by the PIN, and from the proximal extensor forearm muscles (principally the brachioradialis), which are supplied by the distal radial nerve. Stimulation sites for these motor NCSs include the elbow region, the arm (both distal and proximal to the spiral groove), and the axilla. Sensory NCSs that can be performed include assessment of the SRN, obtained by recording at the base of the thumb while stimulating the SRN more proximally, and, whenever necessary, of the posterior antebrachial cutaneous nerve, which is studied along the extensor forearm. On needle EMG, all the muscles innervated by the radial nerve and the PIN can be sampled. Of particular note are the following: 1. The triceps and anconeus—if they are abnormal, along with all the other muscles innervated by the radial nerve and PIN, probably the lesion is in the axilla or upper arm. 2. The brachioradialis and extensor carpi radialis longus— if they are affected, along with PIN-innervated muscles, but the triceps and anconeus are normal, the lesion is probably at or near the spiral groove. 3. All the PIN-innervated muscles, including those innervated by the proximal portion of the nerve—if only they are affected, the lesion is probably at the elbow. Because of the favorable anatomy of the radial nerve in regard to number and site of origin of motor branches, focal neuropathies producing more than modest axon loss are almost as readily localized by needle EMG as those causing demyelinating conduction block.16,19,54,78

1477

One of the major functions of the EDX examination is to provide information regarding prognosis, principally by demonstrating the pathophysiology underlying RN-SG: demyelinating conduction block versus conduction failure caused by axon loss. With the former, the radial motor NCSs demonstrate a prominent block at the spiral groove, while the SRN is normal on NCSs. With the latter, in contrast, the radial motor CMAPs are uniformly low in amplitude or unelicitable, while the SRN response usually is unelicitable. With PIN lesions, the radial motor NCSs recording from the proximal extensor forearm usually are normal, while those recording from the distal extensor forearm are low in amplitude; the SRN response is normal. With isolated lesions of the SRN, both the radial motor NCSs and the needle EMG are normal, whereas the SRN response on NCS is low in amplitude or, more often, unelicitable.6,19,54 The differential diagnosis of radial and posterior interosseous neuropathies includes several entities. Occasionally, central nervous system disorders (focal lesions in the contralateral pre–central motor cortex) cause weakness of the wrist, finger, and thumb extensors, similar to that seen with radial neuropathies; however, on clinical examination, the fact that an upper motor neuron lesion exists usually is obvious, and it can be visualized with MRI and CT of the head. Moreover, the EDX examination is normal. Radiculopathies involving the C7 root sometimes are mistaken for radial neuropathies. The confusion is particularly likely to arise in the rather uncommon situations in which the root lesion has produced substantial weakness of the triceps muscle. This motor abnormality, an impaired triceps deep tendon reflex, and sensory changes in an SRN distribution are all features common to both disorders. However, with radiculopathies, ipsilateral posterior neck and interscapular pain usually are prominent; sensory changes often are found on the volar aspect of the second through fourth digits (i.e., in a median, rather than a radial, nerve distribution); and, if triceps weakness is present, usually muscles innervated by other than the radial nerve, such as the pronator teres and flexor carpi radialis, are also weak. Finally, on EDX examination, the abnormalities seen with radiculopathies essentially are restricted to the needle EMG, and involve muscles outside the radial nerve territory. Brachial plexopathies, principally those involving the posterior cord, can be mistaken for radial neuropathies. However, they typically produce needle EMG abnormalities in the axillary nerve–innervated deltoid and teres minor muscles, as well as the latissimus dorsi. Lesions of the PIN can be confused with radial neuropathies, particularly when the latter have caused weakness along only some of the PIN-innervated muscles as a result of fascicular involvement. Although usually these can be distinguished from one another by clinical and EDX examination, occasionally they cannot. This scenario is encountered whenever only one or two muscles

1478

Diseases of the Peripheral Nervous System

innervated by the PIN, typically those situated distally in the limb, are weak, and to only a minimal to modest degree. In these situations, it may be impossible to localize the lesion accurately.6,19,61,78

Such procedures, however, probably should be delayed for a year or so after nerve injury to allow time for nerve regeneration.19,42,61

Treatment and Prognosis

ULNAR NEUROPATHY

Although RN-SG caused by humeral fracture nearly always are axon loss in type, the majority of cases recover satisfactorily with conservative treatment. However, if the fractures are complex, or if the symptoms develop after reduction of the fracture (either open or closed), then surgical exploration is indicated, as it is for most open injuries, such as those resulting from gunshot wounds or lacerations. Nearly all external compressive lesions of the radial nerve proper, regardless of the site of injury, have demyelinating conduction block as their underlying pathophysiology. Consequently, they recover spontaneously and ultimately completely. Thus conservative therapy, consisting principally of appropriate splinting, is indicated.19,61 This includes those cases of RN-SG resulting from tourniquet paralysis. A noteworthy point regarding the latter is that very often recovery requires many months (nearly up to 1 year), rather than the 6 weeks or so that is typical with neurapraxic lesions.76 With nerve injection injuries, recovery sometimes is sufficient (although never complete) with conservative therapy. However, many of these patients require surgical exploration for at least neurolysis, if not nerve resection and the placement of cable grafts.42,76 Neoplasms affecting the radial nerve or the PIN are removed if possible; in many instances, the nerve can be repaired, and some function restored. Neuralgic amyotrophy involving either the radial nerve or the PIN is treated conservatively. Lesions of the PIN resulting from fractures or elbow dislocations, as well as those caused by arthroscopic procedures, usually are observed for 2 to 3 months, and then surgically explored if no recovery occurs. Uncomplicated lacerations of the SRN can be sutured, but certain injuries of this nerve can be particularly difficult to treat, including those caused by needles and those resulting from blunt trauma or laceration with subsequent scarring. A painful neuroma present on the proximal end of the SRN can be removed, and if the nerve is encased in scar tissue, it can be resected proximal to the site of damage. Of the various upper limb cutaneous nerves, lesions affecting the SRN most often are treated surgically. Thus, in the series by Birch and co-workers, which concerned four cutaneous nerves, 65 of 120 (54%) of the operations were performed on the SRN.8 Nonetheless, at times, no therapy, surgical or nonsurgical, proves effective.19,42,61 Whenever weakness of finger or wrist extension is present, as a result of lesions of either the radial nerve or the PIN, and nerve repair either is impossible or has been unsuccessful, tendon transfers can be performed to relieve much of the motor deficit.

Anatomy The ulnar nerve fibers originate from the C8 and T1 ventral rami, principally the former, and then traverse the lower trunk and much of the medial cord of the brachial plexus. The sensory and motor fibers of the ulnar nerve, unlike those of the median and radial nerves, are contiguous as they pass through the brachial plexus. In the proximal arm, the ulnar nerve is in the flexor compartment. In the mid-arm, however, it enters the extensor compartment and continues distally, in a groove on the medial aspect of the medial head of the triceps; often, it is firmly bound down to this muscle by some dense fascia, referred to as the arcade of Struthers. At the elbow, the nerve rests in a channel, the retroepicondylar groove, formed by the medial epicondyle of the humerus and the olecranon process of the ulna. Immediately distal to the elbow, it enters the cubital tunnel. The roof of this passageway is a fibrous arcade that extends from the medial epicondyle to the olecranon and forms the common aponeurosis of the ulnar and humeral heads of the flexor carpi ulnaris muscle. In this region, motor branches originate from the nerve to supply two forearm muscles: the flexor carpi ulnaris and the medial portion of the flexor digitorum profundus. Typically, both muscles are innervated by multiple branches, with those to the flexor carpi ulnaris muscle being the first to arise. In the distal third of the forearm, the dorsal ulnar sensory (dorsal cutaneous) branch arises, which passes between the flexor carpi ulnaris and the ulna, pierces the deep fascia to reach the dorsal aspect of the forearm, and then tracts distally. Ultimately, it provides sensation to the dorsum of both the hand and the fourth (medial aspect) and fifth digits. Near the wrist, the palmar cutaneous sensory branch originates; it supplies sensation to the proximal medial palm. After traversing the wrist, the ulnar nerve enters Guyon’s canal (also known as the ulnar tunnel), which is covered by the volar carpal ligament and situated between two metacarpal bones: the pisiform and the hook of the hamate. Within Guyon’s canal, the ulnar nerve divides into its terminal superficial ulnar and deep palmar branches. The former, predominantly sensory in nature, leaves the canal to supply sensation to the hypothenar eminence, the fifth digit, and the medial aspect of the fourth digit. The deep palmar branch, as it exits Guyon’s canal, provides motor branches to the hypothenar muscles and then bends acutely around the hook of the hamate and passes laterally, following the deep palmar arch. It innervates all the interossei, the third and fourth

Upper Limb Neuropathies

lumbricals, and two of the muscles composing the thenar eminence: the adductor pollicis and the deep head of the flexor pollicis brevis17,64,74 (Fig. 60–6; see also Fig. 60–3).

Etiologies The ulnar nerve fibers can be injured at any point along their course, from the axilla—where they are continuous with the ulnar terminal nerve of the brachial plexus—to the hand. During wartime, the ulnar and radial nerves are the most commonly injured peripheral nerves. Moreover, in one large series from World War II, more than 80% of such ulnar neuropathies were due to gunshot wounds, as opposed to accidents or lacerations.64 Arm Ulnar neuropathies are uncommon in this region. Traumatic injuries include lacerations and gunshot wounds; the former are more common. Thus, among the 654 ulnar neuropathies evaluated by Kline and co-workers,

FIGURE 60–6 The ulnar nerve and its two terminal branches, the deep palmar nerve (deep ulnar nerve) and the superficial ulnar (sensory) nerve.

1479

62 were in this area, and of them, 23 were due to laceration, while 11 resulted from gunshot wounds.40 The ulnar nerve seldom is involved by humeral fractures, the major exception being supracondylar fractures in children. With the latter, the nerve not only can sustain immediate damage at the moment of the injury, but also can be damaged at a later time, such as when the fracture is reduced, when the fracture is stabilized with wires or pins (particularly percutaneous cross pinning), and from a bony deformity if one occurs as a result of suboptimal fracture treatment.14,49,61,64 Other causes include primary neoplasms (such as schwannoma1); infarcts resulting from vasculitis; entrapment at the arcade of Struthers; and, rarely, multifocal motor neuropathy. The few iatrogenic causes in this region include tourniquet paralysis, in which the ulnar nerve nearly always is affected in combination with the radial and the median,76 and MBFC syndrome, in which the nerve is affected in almost 60% of patients but always in conjunction with the median nerve.43,66 Elbow Segment The majority of focal ulnar neuropathies occur in this region, which includes both the retrocondylar groove and the cubital tunnel. Many physicians, especially surgeons, refer to all ulnar neuropathies at the elbow segment (UN-ES) as cubital tunnel syndrome, implying that lesions at the cubital tunnel are far more common than are those at the retrocondylar groove. This assumption is not supported by the available data, however. Thus, two neurophysiologic studies that focused specifically on this point found abnormalities at the retrocondylar groove in 62% to 69%, at the cubital tunnel in 23% to 28%, and at both sites in 8% to 10% of cases.15,38 Although both open and closed injuries are sustained at this region, the latter are far more common. Among open injuries, lacerations occur with a much higher frequency than do gunshot wounds. Thus, among the 169 “elbow-forearm level” lesions operated on by Kline and co-workers, lacerations were the most common etiology, closely followed by “stretch-contusion” (42 and 41 in number, respectively). In contrast, only three patients with gunshot wounds underwent surgery.40 Single episodes of blunt injury, such as can occur with blows to the elbow or falls onto it, are a source of closed traumatic injuries, both immediate and delayed, with the latter presumably resulting from subsequent scarring.61 Repeated episodes of external pressure are one of the more common etiologies of UN-ES. A major cause includes persons or individuals, while sitting, chronically resting their elbows on firm surfaces such as desks, chair arms, and the base of car window frames. Closed UN-ES usually ultimately are related to either external pressure or elbow flexion.62 The latter increases both the extraneural and intraneural pressure on the ulnar nerve at the elbow segment. The external pressure increases fourfold on full flexion, compared to extension, because of a decrease in the

1480

Diseases of the Peripheral Nervous System

cross-sectional area of the cubital tunnel, whereas the intraneural pressure increases fivefold, presumably as a result of nerve elongation caused by traction.32 Another cause of nontraumatic closed lesions is deformity of the elbow joint. This can result from various congenital anomalies, severe osteoarthritis or rheumatoid arthritis, and unsatisfactory healing of remote fractures of the elbow. In these situations, ulnar neuropathy symptoms may first appear many years after the elbow deformity is evident. All UN-ES once were referred to as tardy ulnar palsy, because remote trauma to the elbow was the first etiology to be recognized. Episodes of prolonged external pressure also are responsible for the UN-ES that occur in bedridden patients, who frequently are emaciated and sometimes comatose. Often the elbow is flexed in these situations, which probably contributes to the nerve damage. The ulnar nerve can be entrapped in this region by an anconeus epitrochlearis muscle, by various fibrous bands, and particularly by the edge of a thickened fibrotic flexor carpi ulnaris aponeurosis where the nerve is just entering the cubital tunnel; the last is considered a very common cause for ulnar neuropathy, and gave rise to the term cubital tunnel syndrome.14,61,64 Elbow flexion is considered a major contributing factor with this particular type of ulnar neuropathy.61 Other nontraumatic etiologies include space-occupying lesions, such as primary nerve tumors, lipomas, and ganglia, residing in the retrocondylar groove or the cubital tunnel. In contrast to the above known risk factors, the role of recurrent subluxation of the ulnar nerve, out of the supracondylar groove, on elbow flexion is uncertain. Although it commonly occurs in the presence of elbow joint deformities, it can also occur when these are not present.14,54,59,61 The major iatrogenic cause for UN-ES is surgical procedures. For many years these perioperative nerve lesions have been considered the responsibility of anesthesiologists, because they have been attributed to malpositioning of the upper limb on the operating table and to inadequate elbow padding. This is the reason why the most common focal nerve injury cited in malpractice litigation filed against anesthesiologists is UN-ES.61 However, in the last decade some convincing data have appeared that question the underlying concepts of these claims. Specifically, a very large retrospective study showed that the majority of patients who developed these neuropathies were males who were at the extremes of body habitus, and who were hospitalized for more than 2 weeks; patient positioning was not a factor.72 Moreover, detailed studies of patients whose ulnar neuropathies were attributed to operations have demonstrated that these lesions can develop regardless of the adequacy of elbow protection present, and frequently they do not become symptomatic until 2 to 7 days into the postoperative period.62,71,72 The latter finding strongly suggests that perioperative ulnar neuropathies actually develop in the early postsurgical convalescent period, rather than during the operation itself.

In addition to the many etiologies described above, UN-ES often develop in the absence of any apparent precipitating cause.14,61 Despite the fact that detailed histories, comprehensive clinical examinations, and various neuroimaging studied yield no clues regarding etiology, many of these idiopathic lesions are axon loss in type and progressive (a point that can be documented with serial EDX studies). Moreover, the incidence of idiopathic UN-ES is increasing, compared to traumatic UN-ES.14 Forearm Virtually all the causes of ulnar neuropathy in this region are quite uncommon. These include compartment syndromes and focal damage by fibrous bands, by a hypertrophied flexor carpi ulnaris muscle, or by the deep flexor–pronator aponeurosis.41,61 The forearm, however, is one of the more common areas for multifocal motor neuropathy to manifest along the motor fibers of the ulnar nerve; often these occur bilaterally. In the distal forearm, the ulnar nerve fibers, along with those of the median and often the radial nerves, are damaged whenever ischemic monomelic neuropathy develops following arteriovenous shunt placements in diabetics with end-stage renal disease.78 Wrist and Hand This is the second most common site for ulnar neuropathies to occur. Those that affect the nerve at the wrist and in the proximal portion of Guyon’s canal typically involve both motor and sensory fibers, whereas those situated more distally in the hand often involve just one of the ulnar nerve’s two terminal branches: the superficial sensory or the deep palmar. The latter can be injured proximal or distal to where the fibers that supply the hypothenar muscles arise from it. Both open and closed lesions occur in this region. Among civilian populations, open lesions are relatively uncommon and most often result from wrist lacerations (medial aspect) and puncture wounds in the hand, frequently caused by industrial accidents. Common causes of closed injuries include spaceoccupying lesions (such as lipomas, anomalous muscles, and particularly ganglion cysts) and external pressure applied to the base of the hand. The latter most often is repetitive in nature; persons having a particular predilection to develop these lesions include those in certain occupations (such as butchers), those using various means of locomotion (such as bicycles, walkers, or canes), and those engaged in a few sports (such as volleyball). Single episodes of blunt trauma to the hand, often with associated fractures of one or more carpal bones, occasionally produce these nerve lesions. Iatrogenic injuries affecting the distal ulnar nerve and its terminal branches are uncommon; many occur during treatment of carpal tunnel syndrome (CTS), specifically during corticosteroid injections and endoscopic carpal tunnel release.14,49,59,61

1481

Upper Limb Neuropathies

Overall, when all ulnar neuropathies in this region are considered, lesions involving the deep palmar branch alone are the most common. Of these, the majority affect the nerve proximal to the origin of the motor branches supplying the hypothenar muscles, while the remainder affect it distal to that point, but usually in the medial aspect of the hand.

Clinical Manifestations The motor and sensory abnormalities resulting from ulnar neuropathies vary considerably, depending principally upon the severity and location of the lesions. Those affecting the ulnar nerve in the arm, along the elbow segment, and in the proximal and mid-forearm frequently have nearly identical clinical presentations. When mild, they often present solely with sensory symptoms, consisting of intermittent paresthesias and sometimes pain along the fifth digit and the medial aspect of both the fourth digit and the hand. If caused by UN-ES, these symptoms may be precipitated or exaggerated by elbow flexion, and accompanied by pain or tenderness along the medial epicondyle. When more severe, the sensory disturbances typically are constant, and may coexist with weakness of the ulnar nerve–innervated intrinsic hand muscles, manifested as a loss of dexterity and impairment of grip and pinch. With severe lesions, a dense sensory loss, sometimes accompanied by pain, is present, accompanied by marked weakness and wasting of the ulnar nerve–innervated hand muscles, the wasting being most apparent in the interossei. With lesions of this severity, wasting of the ulnar nerve–innervated forearm muscles often may be evident as well.11,14,49,61 A relatively small subset of patients with ulnar neuropathies proximal to the distal forearm present solely with motor abnormalities (weakness, wasting), which are often advanced because of the insidious nature of the lesions. A preponderance of these occur in elderly men who are diabetic. Ulnar neuropathies in the distal forearm, the wrist, and the hand have diverse clinical presentations, depending upon the exact site of nerve damage. Those affecting the nerve in the distal forearm, or in the proximal portion of Guyon’s canal, typically involve both motor and sensory fibers. Consequently, the symptoms they produce are similar to those manifested by more proximal lesions, except that the sensory disturbance spares the dorsum ulnar territory, and the ulnar nerve–innervated forearm muscles are not involved. The majority of ulnar neuropathies in the hand affect solely motor fibers and, depending upon the site of the lesion along the deep palmar nerve, usually present with either weakness and wasting of all the ulnar nerve–innervated hand muscles, or of all those except the hypothenar. An uncommon presentation in this region is the pure sensory variety in which symptoms are restricted to the superficial ulnar sensory distribution. Unfortunately,

this presentation has little localizing value, because many lesions located more proximally along the nerve manifest in an identical fashion.11,20,41,54

Diagnosis Although currently quite uncommon, a cubitus valgus or varus deformity may be seen, usually the result of prior, remote injury. With relatively severe lesions, clawing of the fourth and fifth digits is observed, with the metacarpophalangeal joints hyperextended and the interphalangeal joints, both proximal and distal, flexed. A Tinel’s sign may be elicited; most often this is detected along the elbow segment, but it may be found at other points along the course of the nerve, usually as a result of nerve entrapment caused by congenital anomalies or space-occupying lesions. Rarely, a mass can be visualized or palpated along the nerve; typically, manipulation of it produces paresthesias in an ulnar sensory distribution. With UN-ES, palpation about the medial epicondyle often reveals that the nerve is enlarged and tender, and elbow flexion may produce subluxation of the nerve from the retrocondylar groove, or cause sensory symptoms in the ulnar nerve–innervated fingers and hand to appear, or to become more severe. A sensory deficit may be demonstrable in the distribution of the superficial terminal branch, most consistently on the distal portions of the fourth and fifth digits. Theoretically, with lesions located proximal to the distal forearm, similar sensory changes should be found in the distributions of both the dorsal cutaneous and palmar cutaneous nerves, but often this does not occur. Similarly, with lesions located at or proximal to the elbow, if motor abnormalities are present in the hand they should also be detected in the medial forearm, but in at least half of such patients these are not evident. Thus, the ulnar nerve–innervated forearm muscles may be normal in the presence of marked weakness and wasting of the intrinsic hand muscles, a situation that severely compromises localization. Occasionally, the clinical examination with ulnar neuropathies is completely normal, particularly when the symptoms are restricted to intermittent paresthesias in a superficial ulnar sensory distribution.11,14,41,59,61,74 With the majority of ulnar neuropathies in the hand, the sensory examination is normal, whereas the motor assessment reveals weakness and often wasting of the first dorsal interosseous (FDI); involvement of other ulnar-innervated intrinsic hand muscles may also be found, including the hypothenar, depending upon the location of the focal damage along the deep palmar nerve. Much less often, both motor and sensory deficits are present, because the lesion is located at the wrist or in the proximal portion of Guyon’s canal, or solely sensory changes are detected, because the lesion involves only the superficial terminal branch.20,61 Clinically detecting an ulnar neuropathy frequently is much easier than localizing it accurately. This is because

1482

Diseases of the Peripheral Nervous System

lesions at almost any point along the nerve, as noted, can manifest motor and sensory symptoms restricted to the hand, presumably because of selective fascicular involvement.54,61 Both the EDX examination and various neuroimaging studies aid in the assessment of suspected ulnar neuropathies. Plain radiographs of the elbow, wrist, or hand may reveal abnormalities of the joints (sometimes clinically unsuspected) resulting from remote fractures or other trauma, arthritic changes, congenital anomalies, and occasionally space-occupying lesions, such as the ganglia. Although not performed routinely, CT and MRI studies can demonstrate space-occupying lesions and sometimes various congenital anomalies impinging upon the nerve. Britz et al. reported finding MRI abnormalities (specifically, increased signal intensity and nerve enlargement) in the cubital tunnel region in a series of patients with UN-ES of uncertain cause.12 More EDX techniques have been devised for assessing the ulnar nerve than for any other nerve in the body. Thus ulnar motor NCSs can be performed while recording from the hypothenar muscles (the standard technique), the FDI, the adductor pollicis, and the flexor carpi ulnaris, whereas ulnar sensory NCSs can be obtained while stimulating or recording from the fifth digit, the fourth digit, and the dorsum of the hand. For both motor and sensory NCSs, the nerve can be stimulated at multiple points along its course, including at the wrist, below the elbow, above the elbow, at the axilla, and supraclavicularly. Moreover, virtually every forearm and hand muscle innervated by the ulnar nerve can be sampled on needle EMG. Nonetheless, localization, and sometimes even detection, of ulnar neuropathies on EDX examination frequently is unsatisfactory, particularly for UN-ES. Both pathophysiologic and anatomic factors are principally responsible for this suboptimal performance.74 Many ulnar neuropathies produce at least some focal demyelination; these can be localized on motor NCSs because they produce conduction slowing or, less often, conduction block. However, many others manifest solely conduction failure resulting from axon loss. Although these can be detected on NCSs if they cause more than a modest number of axons to degenerate, they cannot be localized well, because no conduction abnormalities are detectible across the site of the lesion. Thus a pure axon loss lesion at or proximal to the elbow may result in an unelicitable ulnar SNAP, recording from the fifth digit, and uniformly low-amplitude ulnar CMAPs, recording from both the hypothenar and FDI, with normal motor conduction velocities along the elbow segment. This combination of changes is consistent with an ulnar neuropathy located at any point along the nerve from its origin in the axilla to the proximal portion of Guyon’s canal in the hand. Theoretically, two other procedures should assist in localization in these instances. The first is the dorsum ulnar SNAP. If it is low in amplitude or unelicitable, the lesion

should be situated along the ulnar nerve at some point at or proximal to where the dorsum ulnar cutaneous nerve originates from it, in the mid-distal forearm. Similarly, needle EMG of the ulnar nerve–innervated forearm muscles should reveal whether the axon loss lesion is at or proximal to the elbow, or distal to it, depending upon whether or not these muscles show abnormalities. Unfortunately, all too often neither of these procedures yields reliable results for localization. Thus patients with proven severe axon loss UN-ES may have normal dorsum ulnar SNAPs on NCSs and normal-appearing flexor carpi ulnaris and flexor digitorum profundus (third and fourth) muscles on needle EMG. When viewed in regard to severity, the EDX examination with ulnar neuropathies proximal to the wrist is most helpful with those that are moderate in degree and causing some element of focal demyelination, because these can be localized satisfactorily by motor NCSs. Mild lesions, in contrast, such as those manifesting only intermittent paresthesias, often produce no abnormalities on EDX examination, regardless of how extensive the latter is. Conversely, severe axon loss lesions that cause profound weakness and wasting and dense sensory deficits are quite unlikely to be satisfactorily localized by EDX studies, because usually there is no focal demyelinating component present.12,14,54,74,75,78 With ulnar neuropathies in the hand, the EDX examination is of variable assistance. Typically, those lesions affecting solely the deep palmar fibers can be localized well, because they cause substantial amounts of either demyelinating conduction block or axon loss, and usually affect the axons supplying the FDI disproportionately severely compared to those that innervate the hypothenar muscles. In contrast, those that involve solely the sensory fibers, or the motor and sensory fibers in the proximal portion of Guyon’s canal, are poorly localized because the actual location of the lesions could be much more proximal (as a result of fascicular involvement).54,78 Although the EDX examination may be suboptimal for localizing many ulnar neuropathies, it can still be of major benefit in the differential diagnosis, which includes a number of entities. The most common of these are C8 radiculopathies, which may produce somewhat similar symptoms. However, with root lesions there is typically a history of ipsilateral neck and interscapular pain. Moreover, the ulnar SNAP is not affected, and muscle weakness—or at least needle EMG abnormalities—often is found in muscles innervated by the C8 root via the radial and median nerves. Lower trunk and medial cord brachial plexopathies also can be mistaken for ulnar neuropathies, but generally they manifest sensory loss along the medial aspect of the arm and forearm—in the distributions of the medial brachial cutaneous and medial antebrachial cutaneous (MAC) nerves—and they involve muscles outside the ulnar nerve territory. The former can be demonstrated on MAC NCSs, the latter on needle EMG. The brachial

Upper Limb Neuropathies

plexopathy that most closely imitates an ulnar neuropathy is that caused by open heart surgery, the result of median sternotomy. These involve almost solely the C8 anterior primary rami, which contribute to the formation of the lower trunk itself. Consequently, they typically do not cause clinical or NCS abnormalities in the distribution of the MAC nerve, nor do they affect either the median nerve–innervated thenar muscles (which are innervated principally by the T1 root) or the median motor NCSs. However, almost invariably they produce needle EMG changes in the forearm muscles innervated by the C8 root via the median and radial nerves, even though, because of the mildness of the lesion, the same muscles may not be clinically weak.28 A number of conditions cause either sensory or motor abnormalities, but not both, suggestive of an ulnar neuropathy. Thus cerebral infarcts in the thalamus or corona radiata rarely can produce sensory symptoms in the ulnar nerve territory, whereas amyotrophic lateral sclerosis and benign focal motor neuron disease may cause weakness and wasting of the intrinsic hand muscles, particularly the FDI, similar to that seen with ulnar neuropathies. However, with the former, clinical examination demonstrates that the lesion is upper, rather than lower, motor neuron in type, whereas with the latter, the clinical and EDX examinations almost always reveal that the abnormalities extend beyond the ulnar nerve territory. Moreover, with them the ulnar SNAP typically is normal, which it would not be with an ulnar nerve lesion of such magnitude, unless the latter were located in the hand and involved solely motor axons.14,54,61

Treatment and Prognosis Nearly all ulnar neuropathies in the arm are axon loss in type, and most require operative treatment. This may result in some improvement in sensation and in forearm muscle strength and bulk, but the ulnar nerve–innervated intrinsic hand muscles, except occasionally the hypothenar muscles, usually remain severely compromised. An exception is the ulnar neuropathy that is a component of tourniquet paralysis. Because the underlying pathophysiology almost invariably is demyelinating conduction block, the ultimate recovery with conservative therapy is excellent, although it may be delayed for many months.14,54,61,74,75 Those UN-ES caused by open wounds, bony anomalies, space-occupying lesions, and sometimes blunt trauma generally require operative repair. In contrast, those resulting from repetitive or prolonged external pressure, or with an unknown etiology, usually are treated conservatively, at least initially; the latter often are milder lesions, manifesting sensory symptoms alone. Nonoperative measures include minimizing the adverse effects of both pressure on the elbow and prolonged elbow flexion by modifying the work environment, padding the elbow, splinting the elbow during sleep, and anti-inflammatory medications. Conservative

1483

therapy helps in a considerable number of patients.14,41,50,61 However, if it fails, and particularly if the symptoms are progressive, then surgical treatment is indicated. Several operations are available, including simple decompression, medial epicondylectomy, and various nerve transposition procedures (subcutaneous, submuscular, intramuscular, and transmuscular). Each of these has its champions and its critics, and the views are quite divergent on this point. Some investigators contend that the operation should be tailored to a specific lesion (thus, entrapment at the cubital tunnel should be treated with decompression alone).14,61 Others claim that certain operations, such as simple decompression and submuscular transposition, have better outcomes than others, regardless of the specific lesion present.5 Still others recommend that surgeons perform the procedure with which they are most comfortable.59 Of note is that nerve transposition carries certain risks that simple decompression and medial epicondylectomy do not. These include nerve infarction caused by devascularization, subsequent scar formation with nerve compression, and primary damage to both the ulnar nerve itself and the MAC nerve, which can produce painful, persistent, and difficultto-treat paresthesias. Although operative treatment for UNES is beneficial in some patients, the percentage actually helped is debated, and almost all sources agree that the success rate is much lower than the 90% or so experienced with CTS surgery. The two most important factors regarding surgical success are the severity and duration of the lesion. Much better postoperative results are obtained when less than severe UN-ES of less than 6 months’ duration are treated surgically.11,14,41,61 Many ulnar neuropathies in the forearm, including those caused by open wounds, soft tissue masses, and compartment syndromes, require surgical treatment. Ischemic monomelic neuropathy and multifocal motor neuropathy are treated conservatively. With the majority of these lesions the ultimate recovery is suboptimal, especially in regard to the regaining of interossei strength and bulk.14,61,78 Surgery generally is required for those ulnar neuropathies in the hand caused by masses and open wounds, whereas those caused by external compression usually respond well to conservative therapy alone.20,61

MEDIAL ANTEBRACHIAL CUTANEOUS NEUROPATHY The MAC nerve fibers originate principally from the T1 ventral rami, traverse the lower trunk and proximal portion of the medial cord, and then exit from the latter, in the axilla, as the MAC nerve. Passing distally on the medial aspect of the proximal arm, the nerve pierces the deep fascia with the basilic vein in the mid-arm to become superficial, and then divides into anterior and posterior

1484

Diseases of the Peripheral Nervous System

(ulnar) divisions. The anterior division continues directly distally, to supply cutaneous sensation to the anteriomedial aspect of the forearm. The posterior division initially tracks obliquely downward as two or three branches that pass proximal to, over, or distal to the medial epicondyle before they join and proceed distally to supply the posteriomedial aspect of the forearm (see Fig. 60–3).17,48,64 MAC neuropathies may be of traumatic, nontraumatic, or iatrogenic origin. Most are traumatic or iatrogenic in nature; nontraumatic lesions are quite rare. Although these lesions may affect any portion of the nerve, most occur near the elbow. Traumatic causes include open injuries, most often lacerations, and closed injuries, such as repetitive trauma near the elbow sustained by a tennis player while serving. Nontraumatic causes include carrying heavy trays or packages on the forearm, and possible compression of the nerve by a superficial lipoma near the medial epicondyle. Iatrogenic damage to the MAC nerve has a variety of causes, including intravenous injections or venipunctures for blood drawing in the antecubital fossa, arterial graft placements in the same region, and, particularly, surgery at the elbow to treat UN-ES (in which the branches of the posterior division are injured as they pass near the medial epicondyle). The MAC nerve is also used as a donor nerve, to provide cable grafts when a portion of a major nerve must be resected and the resulting nerve gap is too extensive to allow the two ends of the nerve to be sutured directly.8,61,64 The clinical manifestations of MAC neuropathy are quite variable. Generally, with uncomplicated lacerations and surgical removal of the nerve for donor graft purposes, the only symptom is a narrow longitudinal band of hypoesthesia (less often anesthesia) along the medial aspect of the forearm. The sensory loss is much less than the distribution of the nerve would suggest, because of extensive overlap of the territories of adjacent cutaneous nerves. With those nerve injuries in which scarring occurs at the lesion site, or with lacerations in which a neuroma forms on the proximal end of the nerve, localized pain and tenderness may be prominent. When the scarring is near the elbow (as occurs with ulnar nerve surgery) the localized pain may increase with elbow movement, particularly extension, to the point that movement is restricted. Both paresthesias and pain in the distribution of the MAC nerve have been reported with nontraumatic lesions near the elbow. In at least two instances, these were not accompanied by demonstrable sensory loss. Traumatic and iatrogenic (other than nerves sacrificed for donor purposes) lesions can manifest a variety of other sensory disturbances in the distribution of the nerve, including hypoesthesia, hyperesthesia, hyperalgesia, and allodynia.8,48,61,64 Diagnosis of a MAC neuropathy most often rests on the clinical findings. This can be confounded by the fact that, with many lesions near the elbow, other nearby nerves, particularly the ulnar and median, may also be damaged. Consequently, the findings may extend beyond the territory

of the MAC nerve. A few tests are available for verifying the diagnosis. If the site of the lesion is apparent, cortisone injections or nerve blocks using local anesthesia usually confirm the nerve damage. A MAC NCS can be performed by stimulating the nerve near the elbow while recording in the midforearm. Characteristically, this NCS is abnormal with these neuropathies, even when no sensory loss is apparent clinically; typically, the SNAP is either low in amplitude or unelicitable, indicative of axon loss.48,64 The differential diagnosis with MAC neuropathies includes C8 radiculopathies, lower trunk or medial cord brachial plexopathies, and ulnar neuropathies, all of which are usually readily identified by clinical and EDX examinations. Cervical root lesions typically manifest neck pain and stiffness, and sensory symptoms are in an ulnar nerve as well as a MAC nerve distribution. On NCSs, with plexopathies the ulnar as well as the MAC SNAPs are affected, and often the median and ulnar CMAPs, recording hand muscles, are low in amplitude. On needle EMG, with plexopathies and with most radiculopathies, abnormalities are found in the appropriate limb muscles. MAC nerve lesions are mistaken for ulnar neuropathies because, for uncertain reasons, many physicians erroneously assume that the medial aspect of the forearm is supplied by the ulnar nerve, even though this is never the case. In fact, any patient with sensory loss extending along the medial portion of both the forearm and hand has either combined lesions of the ulnar and MAC nerves, or a lesion in the lower trunk or proximal medial cord of the brachial plexus, and not UN-ES alone. In the majority of instances, MAC neuropathies require no treatment. However, at least one presumed entrapment neuropathy of the nerve at the elbow responded to corticosteroid injections, and a variety of surgical procedures are available to treat those uncommon lesions in which symptoms are prominent and persistent. These include suturing if the nerve has been transected and the two ends can be approximated. (If the gap between the nerve ends is too extensive for them to be sutured, it would be illogical to sacrifice another cutaneous nerve to obtain a donor nerve.) A painful neuroma along the proximal end of a transected nerve usually is treated with neuroma removal and the implantation of the proximal nerve stump into muscle. Alternatively, the nerve can be resected proximally, in the upper arm. Overall, MAC neuropathies have a satisfactory prognosis, although a permanent sensory deficit is a rather common residual.8,48,61,64

REFERENCES 1. Abbott, K. M., and Nesathurai, S. I.: Musculocutaneous nerve palsy following traumatic spinal cord injury. Spinal Cord 36:588, 1998. 2. Alexander, C. B.: Lesions of peripheral nerves resulting from war injuries: pathology and treatment. BMJ 1:381, 1918.

Upper Limb Neuropathies 3. Antoniadis, G., Richter, H. P., and Rath, S.: Suprascapular nerve entrapment: experience with 28 cases. J. Neurosurg. 85:1020, 1996. 4. Antoniou, J., Tae, S. K., Williams, G. R., et al.: Suprascapular neuropathy: variability in the diagnosis, treatment, and outcome. Clin. Orthop. 386:131, 2001. 5. Bartels, R. H. M. A., Menovoski, T., Van Overbeeke, J. J., and Verhagen, W. I. M.: Surgical management of ulnar nerve compression at the elbow: an analysis of the literature. J. Neurosurg. 89:722, 1998. 6. Bensted, T. J., and Grant, I. A.: Radial nerve. In Brown, W. F., Bolton, C. F., and Aminoff, M. J. (eds.): Neuromuscular Function and Disease, Vol. 1. Philadelphia, W. B. Saunders, p. 917, 2002. 7. Berry, H.: Traumatic peripheral nerve lesions. In Brown, W. F., and Bolton, C. F. (eds.): Clinical Electromyography, 3rd ed. Boston, Butterworth-Heinemann, p. 323, 1993. 8. Birch, R., Bonney, G., and Wynn-Parry, C. B.: Surgical Disorders of the Peripheral Nerves. London, ChurchillLivingstone, 1998. 9. Bird, S. J., and Brown, M. J.: Acute focal neuropathy in male weight lifters. Muscle Nerve 19:897, 1996. 10. Black, K. P., and Lombardo, J. A.: Suprascapular nerve injuries with isolated paralysis of the infraspinatus. Am. J. Sports Med. 18:225, 1990. 11. Bradshaw, D. Y., and Shefner, J. M.: Ulnar neuropathy at the elbow. Neurol. Clin. 17:447, 1999. 12. Britz, G. W., Haynor, D. R., Kuntz C., et al.: Ulnar nerve enlargement at the elbow: correlation of magnetic resonance imaging, clinical, electrodiagnostic, and intraoperative findings. Neurosurgery 38:458, 1996. 13. Callahan, J. D., Scully, T. B., Shapiro, S. A., and Worth, R. M.: Suprascapular nerve entrapment: a series of 27 cases. J. Neurosurg. 74:893, 1991. 14. Campbell, W. W.: Ulnar neuropathy at the elbow. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds.): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 123, 1999. 15. Campbell, W. W., Pridgeon, R. M., and Sahmi, K. S.: Short segment incremental studies in the evaluation of ulnar neuropathy at the elbow. Muscle Nerve 15:1050, 1992. 16. Carlson, N., and Logigian, E. L.: Radial neuropathy. Neurol. Clin. 17:499, 1999. 17. Clemente, C. D.: Gray’s Anatomy, 30th American ed., Baltimore, Williams & Wilkins, 1985. 18. Dawson, D. M., Hallett, M., and Wilbourn, A. J.: Musculocutaneous, axillary and suprascapular neuropathies. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds.): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 335, 1999. 19. Dawson, D. M., Hallett, M., and Wilbourn, A. J.: Radial nerve entrapment. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds.): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 198, 1999. 20. Dawson, D. M., Hallett, M., and Wilbourn, A. J.: Ulnar nerve entrapment in the hand. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds.): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 176, 1999. 21. DeMaio, M., Drez, D., and Mullins, R. C.: The inferior transverse scapular ligament as a possible cause of entrap-

22.

23.

24. 25.

26.

27.

28.

29.

30.

31.

32.

33. 34.

35.

36.

37.

38.

39.

1485

ment neuropathy of the nerve to the infraspinatus. J. Bone Joint Surg. Am. 73:1061, 1991. Demirhan, M., Imhoff, A. B., Debski, R. E., et al.: The spinoglenoid ligament and its relationship to the suprascapular nerve. J. Shoulder Elbow Surg. 7:238, 1998. Duncan, M. A., Lotze, M. T., Gerber, L. H., and Rosenberg, S. A.: Incidence, recovery, and management of serratus anterior muscle palsy after axillary node dissection. Phys. Ther. 63:1243, 1983. England, J. D., and Sumner, A. J.: Neuralgic amyotrophy: an increasingly diverse entity. Muscle Nerve 10:60, 1987. Fattal, C., Weber J., and Beuret-Blanquart, F.: Isolated musculocutaneous nerve palsy in a spinal cord injury. Spinal Cord 38:591, 1998. Fehrman, D. A., Orwin, J. F., and Jennings, R. M.: Suprascapular nerve entrapment by ganglion cysts: a report of six cases with arthroscopic findings and review of the literature. Arthroscopy 11:727, 1995. Ferrante, M. A., and Wilbourn, A. J.: The utility of various sensory nerve conduction responses in assessing brachial plexopathies. Muscle Nerve 18:879, 1995. Ferrante, M. A., and Wilbourn, A. J.: Plexopathies. In Levin, K. H., and Luders, H. O. (eds.): Comprehensive Clinical Neurophysiology. Philadelphia, W. B. Saunders, p. 201, 2000. Ferretti, A., Cerullo, G., and Russo, G.: Suprascapular neuropathy in volleyball players. J. Bone Joint Surg. Am. 69:260, 1987. Friedenberg, S. M., Zimprich, T., and Harper, C. M.: The natural history of long thoracic and spinal accessory neuropathies. Muscle Nerve 25:535, 2002. Fritz, R. C., Helms, C. A., Steinbach, L. S., and Genant, H. K.: Suprascapular nerve entrapment: evaluation with MR imaging. Radiology 182:437, 1992. Gelberman, R., Yamaguchi, K., Hollstein, S., et al.: Changes in interstitial pressure and cross-sectional area of the cubital tunnel and ulnar nerve with flexion of the elbow. J. Bone Joint Surg. Am. 80:492, 1998. Goslin, K. L., and Krivickas, L. S.: Proximal neuropathies of the upper extremity. Neurol. Clin. 17:525, 1999. Hayes, J. M., and Zehr, D. J.: Traumatic muscle avulsion causing winging of the scapula: a case report. J. Bone Joint Surg. Am. 63:495, 1981. Iannotti, J. P., and Ramsey, M. L.: Arthroscopic decompression of ganglion cyst causing suprascapular nerve compression. Arthroscopy 12:739, 1996. Inokuchi, W., Ogawa, K., and Horiuchi, Y.: Magnetic resonance imaging of suprascapular nerve palsy. J. Shoulder Elbow Surg. 7:223, 1998. Juel, V. C., Kiely, J. M., Leone, K. V., et al.: Isolated musculocutaneous neuropathy caused by a proximal humeral exostosis. Neurology 54:494, 2000. Kanakamedala, R. V., Simons, D. G., Porter, R. W., et al.: Ulnar nerve entrapment at the elbow localized by short segment stimulation. Arch. Phys. Med. Rehabil. 69:959, 1988. Katirji, B.: Clinical assessment in neuromuscular disorders. In Katirji, B., Kaminski, H. J., Preston, D. C., et al. (eds.): Neuromuscular Disorders in Clinical Practice. Boston, Butterworth-Heinemann, p. 3, 2002.

1486

Diseases of the Peripheral Nervous System

40. Kim, D. H., Han, K., Tiel, R. L., et al.: Surgical outcomes of 654 ulnar nerve lesions. J. Neurosurg. 98:993, 2003. 41. Kincaid, J. C., and Campbell, W. W.: Ulnar nerve. In Brown, W. F., Bolton, C. F., and Aminoff, M. J. (eds.): Neuromuscular Function and Disease, Vol. 1. Philadelphia, W. B. Saunders, p. 901, 2002. 42. Kline, D. G., and Hudson, A. R.: Nerve Injuries. Philadelphia, W. B. Saunders, 1995. 43. Krivickas, L. S.: Medial brachial fascial compartment syndrome versus tourniquet paralysis. In: Course E: Distinctive Regional Neuropathies [syllabus]. Rochester, MN, American Association of Electrodiagnostic Medicine, p. 37, 2002. 44. Krivickas, L. S., and Wilbourn, A. J.: Peripheral nerve injuries in athletes: a case series of over 200 injuries. Semin. Neurol. 20:225, 2000. 45. Labosky, D. A., and Waggy, C. A.: Apparent weakness of the median and ulnar nerves in radial nerve palsy. J. Hand Surg. Am. 11:528, 1986. 46. Martin, S. D., Warren, R. F., and Martin, T. L.: Suprascapular neuropathy: results of non-operative treatment. J. Bone Joint Surg. Am. 79:1159, 1997. 47. Mikami, Y., Nagano, A., Ochiai, N., and Yamamoto, S.: Results of nerve grafting for injuries of the axillary and suprascapular nerves. J. Bone Joint Surg. Br. 79:527, 1997. 48. Mumenthaler, M.: Some clinical aspects of peripheral nerve lesions. Eur. Neurol. 2:25, 1969. 49. Mumenthaler, M., and Schliack, H. (eds.): Peripheral Nerve Lesions. New York, Thieme, 1991. 50. Padua, L., Apule, I., Caliandro, P., et al.: Natural history of ulnar neuropathy at the elbow. Clin. Neurophysiol. 113:1980, 2002. 51. Parsonage, M. J., and Turner, J. W. A.: Neuralgic amyotrophy: the shoulder-girdle syndrome. Lancet 1:973, 1948. 52. Perlmutter, G. S.: Axillary nerve injury. Clin. Orthop. 368:28, 1999. 53. Post, M., and Grinblat, E.: Suprascapular nerve entrapment: diagnosis and results of treatment. J. Shoulder Elbow Surg. 2:190, 1993. 54. Preston, D. C.: Compressive and entrapment neuropathies of the upper extremity. In Katirji, B., Kaminski, H. J., Preston, D. C., et al. (eds.): Neuromuscular Disorders in Clinical Practice. Boston, Butterworth-Heinemann, p. 744, 2002. 55. Richards, R. R., Hudson, A. R., Bertoia, J. T., et al.: Injury to the brachial plexus during Putti-Platt and Bristow procedures: a report of eight cases. Am. J. Sports Med. 15:374, 1987. 56. Romeo, A. A., Rotenberg, D., and Bach, B. R. Jr.: Suprascapular neuropathy. J. Am. Acad. Orthop. Surg. 7:358, 1999. 57. Sander, H. W., Conigliari, M., and Masdeu, J. C.: Antecubital phlebotomy complicated by lateral antebrachial cutaneous neuropathy. N. Engl. J. Med. 339:2024, 1998. 58. Seddon, H.: Surgical Disorders of Peripheral Nerves, 2nd ed. Edinburgh, Churchill Livingstone, 1975. 59. Spinner, R. J., and Amadio, P. C.: Compressive neuropathies of the upper extremity. Clin. Plast. Surg. 30:155, 2003. 60. Steinmann, S. P., and Moran, E. A.: Axillary nerve injury: diagnosis and treatment. J. Am. Acad. Orthop. Surg. 9:328, 2001. 61. Stewart, J. D.: Focal Peripheral Neuropathies, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, 2000.

62. Stewart, J. D., and Shantz, S. H.: Perioperative ulnar neuropathies: a medical legal review. Can. J. Neurol. Sci. 30:15, 2003. 63. Sunderland, S.: Traumatic injuries of peripheral nerves: II: An analysis of the incidence in 301 consecutive cases of peripheral nerves. Aust. N. Z. J. Surg. 15:25, 1945-46. 64. Sunderland, S.: Nerves and Nerve Injuries, 2nd ed. Edinburgh, Churchill Livingstone, 1978. 65. Takagishi, K., Maeda, K., Ikeda, T., et al.: Ganglion causing paralysis of the suprascapular nerve: diagnosis by MRI and ultrasonography. Acta Orthop. Scand. 62:391, 1991. 66. Tsao, B., and Wilbourn, A.: The medial brachial fascial compartment syndrome following axillary arteriography. Neurology 61:1037, 2003. 67. Tung, G. A., Entzian, D., Stern, J. B., and Green, A.: MR imaging and MR arthrography of paraglenoid labral cysts. AJR Am. J. Roentgenol. 174:1707, 2000. 68. Vastamaki, M., and Goransson, H.: Suprascapular nerve entrapment. Clin. Orthop. 297:135, 1993. 69. Vastamaki, M., and Kauppila, L. I.: Etiologic factors in isolated paralysis of the serratus anterior muscle: a report of 197 cases. J. Shoulder Elbow Surg. 2:240, 1993. 70. Warner, J. J., and Navarro, R. A.: Serratus anterior dysfunction: recognition and treatment. Clin. Orthop. 349:139, 1998. 71. Warner, M. A., Warner, D. O., Matsumoto, J. Y., et al.: Ulnar neuropathy in surgical patients. Anesthesiology 90:54, 1999. 72. Warner, M. A., Warner, M. E., and Martin, J. T.: Ulnar neuropathy: incidents, outcome, and risk factors in sedated or anesthetized patients. Anesthesiology 87:1332, 1994. 73. Wiater, J. M., and Flatow, E. L.: Long thoracic nerve injury. Clin. Orthop. 368:17, 1999. 74. Wilbourn, A. J.: Ulnar neuropathy In: 1985 Course A: Basic Electrophysiologic Testing in Mononeuropathy [syllabus]. Rochester, MN, American Association of Electromyography and Electrodiagnosis, p. 27, 1985. 75. Wilbourn, A. J.: Ulnar neuropathy at the elbow: electrodiagnostic approaches. In: 1991 AAEM Course D: Focal Peripheral Neuropathies: Selected Topics [syllabus]. Rochester, MN, American Association of Electrodiagnostic Medicine, p. 13, 1991. 76. Wilbourn, A. J.: Nerve injuries caused by injections and tourniquets. In: 1996 AAEM Plenary Session: Physical Trauma to Peripheral Nerves [syllabus]. Minneapolis, American Association of Electrodiagnostic Medicine, p. 15, 1996. 77. Wilbourn, A. J.: Nerve conduction studies: types, components, abnormalities, and value and localization. Neurol. Clin. 80:305, 2002. 78. Wilbourn, A. J., and Ferrante, M. A.: Clinical electromyography. In Joynt, R. J., and Griggs, R. C. (eds.): Clinical Neurology, Vol. 1 (loose-leaf). Philadelphia, Lippincott–Raven, p. 178, 1997. 79. Witvrouw, E., Cools, A., Lysens, R., et al.: Suprascapular neuropathy in volleyball players. Br. J. Sports Med. 34:174, 2000. 80. Zoltan, J. D.: Injury to the suprascapular nerve associated with anterior dislocation of the shoulder: case report and a review of the literature. J. Trauma 19:203, 1979.

61 Mononeuropathies of the Lower Limb BASHAR KATIRJI AND ASA J. WILBOURN

Pathophysiology Ilioinguinal, Iliohypogastric, and Genitofemoral Neuropathies Obturator Neuropathy Femoral Neuropathy Anatomy Etiology Clinical Manifestations Diagnosis Treatment and Prognosis Saphenous Neuropathy Lateral Femoral Cutaneous Neuropathy Anatomy Etiology Clinical Manifestations Diagnosis Treatment and Prognosis

Posterior Femoral Cutaneous Neuropathy Gluteal Neuropathies Sciatic Neuropathy Anatomy Etiology Clinical Manifestations Diagnosis Treatment and Prognosis Piriformis Syndrome Common Peroneal Neuropathy Anatomy Etiology Clinical Manifestations Diagnosis Treatment and Prognosis Tibial Neuropathy Anatomy

Neuropathies of the lower limb are less common than those of the upper limb by a factor of 1:3 or 1:4 (see Chapter 60). Despite their lower frequency, lower limb neuropathies may be more disabling than those of the upper limb, because they can seriously compromise weight bearing and thereby limit walking.64 Moreover, unless the cause is immediately obvious (e.g., limb trauma), they are much more likely to be initially misdiagnosed than are their upper limb counterparts. This is because their clinical manifestations are readily confused with those produced by more proximal lesions, especially radiculopathies, which are far more common.

PATHOPHYSIOLOGY For all focal peripheral nerve disorders, including those involving lower limb nerves, the underlying pathophysiology is the major determinant of their clinical and electrophysiologic manifestations and their prognosis. Peripheral nerves are very limited in their responses to localized

Etiology Clinical Manifestations Diagnosis Treatment and Prognosis Tarsal Tunnel Syndrome Anatomy Etiology Clinical Manifestations Diagnosis Treatment and Prognosis Distal Plantar, Distal Calcaneal, and Digital Neuropathies Anatomy Lesions of Individual Nerves Sural Neuropathy

insult, regardless of the nature of the latter (e.g., traction, compression, transection). Whenever they sustain a sufficient degree of injury, both unmyelinated and myelinated axons are interrupted at the site of lesion and their distal segments subsequently undergo wallerian degeneration, leading to conduction failure along them. The deficits that appear in the distribution of the affected axons may be total or partial, depending upon the percentage of fibers involved. Whenever loss of large myelinated motor axons is severe, paresis or paralysis immediately results, soon followed by muscle atrophy; with similar large sensory axon involvement, impairment of vibration, position, and light touch sensation occurs. Severe loss of the unmyelinated and lightly myelinated sensory fibers affects pain and temperature sensations. Concerning the myelinated fibers, lesser degrees of local injury than that necessary to interrupt axons can produce focal demyelination. This often results in localized conduction abnormalities, either slowing or block. With demyelinating conduction slowing, the speed of the nerve impulses is reduced as they traverse the site of injury. 1487

1488

Diseases of the Peripheral Nervous System

Conduction slowing has almost no clinical manifestations, because all the impulses ultimately reach their destinations. However, if the slowing is not uniform, then those portions of the formal neurologic examination (such as deep tendon reflex assessment and vibration testing) for which the affected fibers are responsible can be compromised. With demyelinating conduction block, nerve impulses are halted at the lesion site. The clinical manifestations are very similar to those seen with axon loss of the same severity and, in fact, are identical for motor fibers. For sensory fibers, however, only those sensory modalities that are transmitted via large myelinated axons are affected (i.e., pain and temperature remain intact).70

ILIOINGUINAL, ILIOHYPOGASTRIC, AND GENITOFEMORAL NEUROPATHIES All three of these nerves are predominantly sensory and originate from the L1 ventral ramus, with the iliohypogastric and genitofemoral nerves receiving contributions from the T12 and L2 ventral rami, respectively. The ilioinguinal nerve passes medial to the anterior superior iliac spine, and innervates the lower abdominal muscles and a narrow strip of skin along the inguinal canal. The iliohypogastric nerve innervates two lower abdominal muscles (obliquus internus and transversus abdominis) and then divides into two cutaneous terminal branches: a lateral branch that innervates a small area of the upper lateral buttock, and an anterior branch that innervates a small skin area above the pubis symphysis. The genitofemoral nerve divides into a femoral branch that innervates a small cutaneous area on the anterior aspect of the thigh and a genital branch that supplies the cremasteric muscle and the skin of the scrotum or labium majus7,21 (Fig. 61–1). Most lesions affecting these nerves are iatrogenic. Traditional hernia repair is the most common cause. Other etiologies include appendectomy, cesarean section, abdominoplasty, blunt trauma to the lower abdomen, and iliac bone grafting.25,47,59 The clinical manifestations of these neuropathies overlap significantly. Most patients complain of burning pain and numbness in the groin that may extend into the lower abdomen, upper thigh, symphysis pubis, scrotum, or labia majora. The symptoms may worsen on extension of the hip, so that the patients keep the hip flexed for comfort. Bulging of the lower abdominal wall may occur with severe ilioinguinal neuropathies, while an absent or diminished cremasteric reflex may be found with genitofemoral neuropathies.9,25 Accurately diagnosing these sensory neuropathies is difficult because the territories of these nerves significantly overlap. Often, diagnostic nerve blocks are needed for confirmation.20 These lesions may imitate high lumbar radiculopathies, such as L1 or L2 radiculopathies, which often

FIGURE 61–1 The obturator, iliohypogastric, ilioinguinal, and genitofemoral nerves. Of these four nerves, only the obturator has any major motor component: It supplies the thigh adductor muscles.

cause groin pain and numbness. The pain and abdominal bulging associated with ilioinguinal nerve lesions may be mistaken for a frank hernia. Electrodiagnosis (EDX) is of very limited value, because no sensory nerve conduction studies (NCSs) are available for assessing these nerves. Needle electromyography (EMG) of lower abdominal muscles may reveal denervation in patients with ilioinguinal neuropathies.25,32,60 Nerve blocks may offer transient relief but are more important diagnostically. Surgical exploration with neurectomy (nerve section) has had good results in a significant percentage of patients with persistent symptoms.20,59

OBTURATOR NEUROPATHY The obturator nerve derives from the anterior divisions of the lumbar plexus, which, in turn, originates from the L2, L3, and L4 ventral rami. The obturator nerve passes along the medial edge of the psoas muscle and over the sacroiliac joint to reach the obturator canal, where it divides into anterior and posterior branches to innervate the thigh adductor muscles. (The adductor magnus

1489

Mononeuropathies of the Lower Limb

receives additional innervation from the sciatic nerve.) The obturator also supplies cutaneous sensation to the proximal inner thigh7 (see Fig. 61–1). Obturator neuropathies may be manifestations of obturator hernias, endometriosis, or new or recurrent pelvic malignancies. The nerve also can be injured by pelvic trauma (particularly that producing pelvic fractures) and during hip surgery, genitourinary procedures, and abdominal operations (such as aortofemoral bypass or oophorectomy). Rarely the obturator nerve is damaged during vaginal delivery, particularly when forceps are used.9,25,60 Neuralgic pain in the medial thigh (“obturator neuralgia”), worse with exercise, is a rather common presentation, as is numbness in the medial thigh. Occasionally subjective gait impairment is noted, but patients seldom report weakness. In a few cases, these neuropathies are only detected on clinical or EDX examination.9,25 Obturator neuropathies may be mistaken for L3 or L4 radiculopathies or lumbar plexopathies and, in diabetics, diabetic amyotrophy. Radiculopathies or plexopathies are distinguished by the presence of quadriceps and iliopsoas weakness or depressed/absent patellar reflexes. Disorders of the symphysis pubis may produce referred pain in the medial thigh, mimicking that caused by obturator neuropathies.9,25,60 The neurologic findings are limited to weakness of thigh adduction, sometimes associated with sensory loss in the medial, proximal thigh. Circumduction of the leg, caused by weakened hip adduction, occasionally results in a widebased gait. Patients with obturator neuropathies and no history of pelvic or hip surgery or pelvic trauma should undergo pelvic computed tomography (CT) or magnetic resonance imaging (MRI) to exclude pelvic malignancy. There are no NCSs available for assessing the obturator nerve. Needle EMG typically reveals fibrillation potentials, decreased motor unit action potential (MUAP) recruitment, or both in the thigh adductors only (i.e., assessment of the quadriceps, iliacus, and lumbar paraspinal muscles is normal).9,25,50,60 The treatment of obturator neuropathies depends on the cause. Mass lesions are excised if possible. Surgical correction of an obturator hernia is essential because the mortality from intestinal obstruction is high. In patients with disabling pain, obturator nerve blocks or surgical excision of the obturator nerve may relieve the pain.9,60,66

compartment. The femoral nerve innervates the psoas and the iliacus muscles. It then passes beneath the inguinal ligament to enter the proximal anterior thigh, where it soon divides into several terminal motor branches, which innervate the sartorius and all four heads of the quadriceps, and three terminal sensory branches, the medial and intermediate cutaneous nerves of the thigh and the saphenous nerve (which supply sensation to the anterior thigh and the medial aspect of the leg, respectively)7 (Fig. 61–2).

Etiology Because of its short course, the main trunk of the femoral nerve generally is injured at one of two sites: the retroperitoneal pelvic space or beneath the inguinal

FEMORAL NEUROPATHY Anatomy The lumbar plexus derives from the L2, L3, and L4 ventral rami, and divides into anterior and posterior divisions that form the obturator and femoral nerves, respectively. In the pelvis, the femoral nerve (or anterior crural nerve) and the iliopsoas muscles are the main constituents of the iliacus

FIGURE 61–2 The femoral nerve and its major sensory component, the saphenous nerve. This nerve innervates the anterior thigh muscles and supplies sensation to the anterior thigh and medial leg (the sartorius muscle is not shown).

1490

Diseases of the Peripheral Nervous System

ligament. Less localized femoral neuropathies may occur with marked hip hyperextension, such as in dancers; with prolonged squatting, such as during yoga exercise; and from penetrating injuries such as gunshot and stab wounds or lacerations.25 Femoral Neuropathy in the Pelvis Most of these lesions are iatrogenic, and occur during pelvic surgical procedures that utilize retractor blades, particularly the self-retracting type; the femoral nerve is compressed between the lateral blade of the retractor and the pelvic wall. Rarely, the nerve is damaged perioperatively by ischemia, a retroperitoneal hematoma, inadvertent laceration, or being stapled or sutured. These perioperative femoral neuropathies result from a variety of gastrointestinal, urologic, gynecologic, or vascular operations, including abdominal colectomy, inguinal herniorrhaphy, radical prostatectomy, hysterectomy, and abdominal aortic repair.5,9,19,25,35 Femoral neuropathies occurring during abdominal hysterectomy have a reported incidence of 1.5% to 7.5%.35 The incidence was decreased by 93% in one study when retractors were not used.19 Acute iliacus compartment syndrome caused by hemorrhage may produce ischemia of the iliacus muscle, the femoral nerve, or both. Occasionally, the hematoma is large and extends into the retroperitoneal space, leading to a more extensive injury to the psoas muscle, the lumbar plexus, or the entire lumbosacral plexus. These hemorrhages often are iatrogenic, resulting from anticoagulant therapy, pelvic operations, or femoral vessel catheterization. They may also be spontaneous, resulting from hemophilia or other blood dyscrasias or following a ruptured abdominal aortic aneurysm.9,25,60 Pelvic mass lesions other than hematomas may also compress the femoral nerve. These include lymphadenopathy, abscesses, cysts, iliac or aortic aneurysms, or tumors such as lymphomas and malignancies of the colon, rectum, or ilium. Primary femoral nerve neoplasms, such as schwannomas or neurofibromas, are extremely rare. Pelvic trauma may cause traumatic rupture of the iliopsoas muscle with a subsequent iliacus compartment syndrome; it may also occasionally cause direct contusion of the femoral nerve.25,60 Femoral Neuropathy at the Inguinal Ligament These lesions also are usually iatrogenic, and result from nerve compression during lithotomy positioning,1 which is employed during such procedures as vaginal delivery, vaginal hysterectomy, prostatectomy, and laparoscopy. Particularly when the leg is held in extreme hip flexion and external rotation, lithotomy positioning causes kinking of the femoral nerve under the inguinal ligament. Smokers and women with thin body habitus are at higher risk.1,34,60,67 Femoral neuropathies are sustained in up to 2.3% of total hip replacements, particularly with revisions and

complicated reconstructions; they are usually due to malpositioning of the anterior acetabular retractors during the procedure.68 Less common causes of femoral neuropathies at the inguinal ligament include inguinal lymphadenopathies, femoral vessel catheterization and localized groin hematomas.25,60

Clinical Manifestations Most femoral neuropathies are unilateral, although bilateral lesions may result from lithotomy positioning and pelvic surgery. The principal symptoms are thigh weakness along with anterior thigh and medial leg numbness. When severe, the patient may be unable to stand without support; when milder, there is leg buckling on attempted weight bearing, leading to frequent falls. Pain usually is absent or mild, the main exceptions being iliacus and retroperitoneal hematomas, in which severe pain is often experienced not only in the anterior thigh, but also the back, buttock, abdomen, and groin.9,25

Diagnosis The diagnosis of femoral neuropathies depends on the clinical presentation and laboratory tests, particularly neuroimaging and EDX studies. Neurologic examination reveals weakness of knee extension with an absent or depressed patellar reflex. Thigh adduction and ankle dorsiflexion strength is normal. Hip flexion usually is weak when the lesion is intrapelvic, but spared when it is at the inguinal region. Sensory examination frequently reveals a deficit over the anterior thigh and medial leg. With iliacus or retroperitoneal hematomas, patients often keep the affected leg flexed at the hip.25 Neuroimaging studies, particularly CT and MRI of the pelvis, are useful in patients with suspected pelvic or femoral nerve neoplasms, and with iliacus or retroperitoneal hematomas. With the latter, these studies should be done promptly if surgical evacuation is to prevent severe nerve damage.25,60 The EDX examination can be helpful with femoral neuropathies by (1) confirming that only the femoral nerve is involved; (2) identifying the site of the lesion (intrapelvic vs. inguinal ligament); and (3) determining prognosis (by demonstrating the pathophysiology of the lesion). The most useful components are the needle EMG and two NCSs: 1. The saphenous sensory nerve action potential (SNAP), which assesses the L4 postganglionic sensory fibers; this SNAP is of low amplitude or absent with axon loss lumbar plexopathies and femoral neuropathies, but normal with L4 radiculopathies and demyelinating conduction block affecting femoral nerves.

Mononeuropathies of the Lower Limb

2. The femoral motor compound muscle action potential (CMAP), obtained by recording over the rectus femoris while stimulating in the groin. The amplitude provides a semiquantitative measure of the number of functioning motor axons. If the amplitude is low or absent, axon loss is the underlying pathophysiology. Conversely, if it is unaffected, demyelinating conduction block is responsible. (With inguinal lesions, stimulation of the femoral nerve in the groin usually is distal to the site of involvement; however, stimulating proximal to the lesion has proven difficult.) The femoral CMAP latency, in contrast to the amplitude, has little diagnostic use. Needle EMG reveals fibrillation potentials and decreased recruitment of MUAPs in the quadriceps muscle. Some fibrillation potentials typically are seen, even when the pathophysiology is predominantly demyelination. The iliacus muscle is innervated by a branch that originates from the femoral nerve 4 to 5 cm above the inguinal ligament. Consequently, when it is abnormal on needle EMG, the femoral neuropathy is considered intrapelvic; conversely, when it appears normal, the lesion is at/around the inguinal ligament. The thigh adductors, innervated by the L2-L4 roots via the obturator nerve, will be normal with femoral neuropathies but will often show abnormalities with upper lumbar radiculopathies or plexopathies.25 The major differential diagnosis includes upper lumbar (L2-L4) radiculopathies and lumbar plexopathies. Clinical weakness or needle EMG changes in the thigh adductors or tibialis anterior are consistent with a radiculopathy or plexopathy, but not a femoral neuropathy.1,25 “Diabetic Femoral Neuropathy” Despite earlier clinical reports suggesting that selective femoral neuropathy may be a complication of diabetes mellitus, the term diabetic femoral neuropathy is a misnomer. Although anterior thigh pain is present and the quadriceps and iliopsoas muscles are weak, mimicking an isolated femoral nerve injury, careful clinical and needle EMG examinations consistently reveal involvement of thigh adductors and sometimes also foot dorsiflexors and lumbar paraspinal muscles. Thus these patients have more extensive deficits, caused by damage to the lumbosacral plexus and spinal roots, the so-called diabetic amyotrophy or diabetic lumbosacral plexus radiculoneuropathy. These changes frequently are bilateral but often are asymmetrical. Moreover, an underlying distal axon loss polyneuropathy is also detected in two thirds of these patients (see Chapter 60).26,60

Treatment and Prognosis Patients with confirmed iliacus or retroperitoneal hematomas might require urgent drainage, but the necessity for this is controversial. If it is done, it should be per-

1491

formed as soon as the hematoma is detected and before signs of severe femoral nerve damage appear, because there is no evidence that surgical treatment improves the outcome once the nerve lesion is established. Patients with pelvic or inguinal masses or femoral nerve tumors often require surgery. Most other femoral neuropathies are treated conservatively. A knee-ankle-foot orthosis is helpful for patients with unilateral severe weakness of the quadriceps, to assist in walking and prevent falls. Anticonvulsants or tricyclic antidepressant drugs may be used for pain and uncomfortable hyperesthesias that often are present.9,25,60 In general, compressive femoral neuropathies at the inguinal ligament following childbirth or laparoscopy tend to be demyelinating in nature and recover rapidly,1 whereas those caused by iliacus hematomas give rise to axon loss and have a more protracted recovery time. Lesions of the femoral nerve tend to have a good prognosis, even when significant axon loss has occurred. In most conditions, particularly those at the inguinal ligament, the site of the injury is not far removed from the most clinically relevant muscles, the quadriceps.25 The femoral CMAP amplitude obtained on NCSs is the only independent factor influencing prognosis. With demyelinating lesions having normal CMAPs, recovery usually occurs in 2 to 3 months by remyelination. If the CMAP amplitude is reduced to more than 50% of the normal contralateral side, all patients recover within 1 year, whereas if it is reduced to less than 50% of the contralateral side, fewer than half the patients improve during the same period.34

SAPHENOUS NEUROPATHY The saphenous nerve arises from the femoral nerve in the groin and travels distally in the thigh to reach the subsartorial (Hunter’s or adductor) canal. There, it gives off the infrapatellar branch that innervates the skin over the anterior surface of the patella. Proximal and medial to the knee, the nerve pierces the fascia to become subcutaneous. It then crosses the pes anserinus bursa at the upper medial end of the tibia, and soon divides into two terminal branches to innervate the skin of the medial knee, medial leg, and medial malleolus as well as a small area of the medial arch of the foot7 (see Fig. 61–2). Saphenous neuropathies are commonly iatrogenic, resulting from vascular or orthopedic operations. They occur during stripping of long saphenous varicose veins, harvesting of long saphenous veins for coronary artery bypass, and superficial femoral thromboendarterectomies or femoropopliteal bypass graft placements.60 Surgical procedures performed on the knee, including meniscectomies and arthroscopic procedures, may injure the main trunk of the saphenous nerve and/or its infrapatellar branch. Although entrapment of the saphenous nerve is rare, it

1492

Diseases of the Peripheral Nervous System

may occur as the nerve exits the subsartorial canal or as a result of a pes anserine bursitis. The infrapatellar branch may be entrapped behind the sartorius tendon or injured with acute or repetitive medial knee trauma.25,39,63 Patients with saphenous neuropathies often complain of paresthesias or hyperesthesias of the medial leg. Knee pain also is common and may be the presenting symptom. Selective infrapatellar branch entrapment often causes knee pain with minimal sensory loss confined to a small patch below the knee. Percussion of the saphenous nerve at the subsartorial canal may elicit Tinel’s sign.25,60 Saphenous neuropathies should be differentiated from L4 radiculopathies, lumbar plexopathies, and femoral neuropathies. Muscle strength, the patellar tendon reflex, and needle EMG of the quadriceps, iliacus, and thigh adductors are normal in patients with saphenous neuropathy, while the saphenous SNAP is often absent or asymmetrically low in amplitude. Conversely, the latter is normal in L4 radiculopathies (see Femoral Neuropathy above). The pain with infrapatellar branch lesions mimics pain from various pathologies of the knee joint, which should always be excluded. The treatment of saphenous neuropathies depends on their cause. The area of sensory loss usually becomes smaller with time, although pain may persist. In entrapment at the subsartorial canal, saphenous nerve blocks may alleviate the symptoms. Occasionally, when definite nerve entrapments do not respond to conservative treatment, surgical decompression of the saphenous nerve near the canal, or the infrapatellar branch near the sartorius tendon, is indicated.9,39,60,63

LATERAL FEMORAL CUTANEOUS NEUROPATHY Anatomy Branches from the L2 and L3 ventral rami join to form the lateral femoral cutaneous nerve (LFCN), or lateral cutaneous nerve of the thigh. This sensory nerve travels within the lower abdominal muscles and crosses the iliacus muscle. It leaves the pelvis by passing underneath or within the inguinal ligament near the anterior superior iliac spine. It then changes its course (from horizontal to vertical), pierces the fascia lata, and divides into its terminal branches to innervate the skin of the lateral thigh. The sensory territory of the LFCN does not extend beyond the knee, and seldom crosses the anterior or posterior midline of the thigh.7,25

Etiology Meralgia paresthetica (from the Greek mèros [thigh] and algo [pain]) refers to the sensory syndrome caused by a

neuropathy of the LFCN. It is most often due to compression of the nerve beneath the inguinal ligament. The exact cause often is unclear, but the disorder is commonly associated with pregnancy, obesity, and diabetes mellitus. Specific etiologies include wearing large tool belts or tight clothing, or leaning against various firm objects such as tool benches or gymnastic bars. The LFCN may be injured during iliac bone grafting, by misplaced intramuscular injections, and by car seat belts during automobile accidents. These neuropathies occasionally are associated with pelvic surgery (e.g., renal transplantation, hernia repair, gastric bypass), and with pelvic or iliac crest masses.9,17,23,25,60,72

Clinical Manifestations Thigh pain (often burning in character), numbness, and sometimes hyperesthesia constitute the main symptomatology of LFCN lesions, which usually are unilateral. The symptoms may encompass the entire sensory territory of the nerve or only part of it; often they coincide with a trouser pocket area. They may be intermittent or constant and sometimes worsen with standing, walking, or turning in bed, whereas they may improve with hip flexion.25,60

Diagnosis Diagnosing meralgia paresthetica frequently is simple. Often patients can outline the exact area of sensory disturbance, which correlates with the cutaneous distribution of the LFCN or one of its branches, and the history reveals the precipitating cause (e.g., weight gain). On neurologic examination, a well-circumscribed area of sensory loss can be demonstrated over the lateral thigh in the territory of the LFCN or one of its terminal branches. Hair loss restricted to the area of sensory loss occasionally is seen. Neuroimaging studies of the pelvis usually are unnecessary except if an abdominal mass is suspected. The EDX examination is most useful in excluding other entities, such as lumbar radiculopathies. Recording the SNAP of the LFCN may be attempted, but responses often are sometimes unrecordable bilaterally in healthy subjects, particularly in obese individuals. An asymmetrically low-amplitude or absent SNAP on the symptomatic side is a diagnostic finding. Assessing somatosensory evoked potentials of the LFCN is also possible, but their sensitivity is low with meralgia paresthetica.25,26,36,60 The differential diagnosis includes upper (L2, L3) lumbar radiculopathies, lumbar plexopathies, and particularly femoral neuropathies. With L2 or L3 radiculopathies, the sensory symptoms are more vague and often there is back or groin pain. Conversely, with femoral neuropathies and lumbar plexopathies, quadriceps weakness or depressed patellar reflexes often are found, and the numbness frequently extends into the medial leg and beyond the anterior thigh midline.9,25

1493

Mononeuropathies of the Lower Limb

Treatment and Prognosis Meralgia paresthetica symptoms usually slowly dissipate when the precipitating factor is corrected, such as weight loss in obese patients and elimination of tight clothing or large belts. In pregnant women, the symptoms resolve after delivery but may recur with subsequent pregnancies. More aggressive treatment is indicated when the symptoms are considered intolerable. Drugs useful in neuropathic pain are effective, including oral tricyclic antidepressants (such as amitriptyline), oral anticonvulsants (such as carbamazepine), and local agents (such as capsaicin). Corticosteroid injections near the anterior superior iliac spine also may be helpful. Surgical decompression of the LFCN is occasionally performed (in about 10% of patients). A preceding successful steroid injection is a good prognostic indicator. Surgical procedures include neurolysis with transposition or sectioning (neurectomy) of the LFCN.9,60,72

POSTERIOR FEMORAL CUTANEOUS NEUROPATHY The posterior femoral cutaneous nerve (PFCN), or posterior cutaneous nerve of the thigh, originates from the S1 through S3 ventral rami and exits the pelvis, with the inferior gluteal nerve, through the greater sciatic notch, under the piriformis muscle. It then enters the thigh near the sciatic nerve, at the lower border of the gluteus maximus. This sensory nerve innervates the skin of the lower buttock, posterior thigh, popliteal fossa, and proximal third of the calf, as well as the perineum and scrotum in men and labium majus in women. It can be injured by injections into the buttock, lacerations, and gunshot wounds, often along with the inferior gluteal nerve, the sciatic nerve, or both. Pelvic mass lesions, such as colorectal tumors or venous malformations, may affect the PFCN. The nerve also can be compromised by prolonged pressure on the ipsilateral buttock, as may occur during gymnastic exercises or prolonged bicycle riding. Paresthesias in the lower buttock and posterior thigh are the predominant features. Burning or throbbing pain also occurs; often it radiates to the perineum, scrotum, or labium majus and is exaggerated by sitting or lying on the buttocks. The diagnosis relies on a detailed history and the typical sensory findings. A CT or MRI scan of the pelvis is indicated if there is a suspicion of a pelvic mass. A PFCN SNAP usually is of low amplitude or unelicitable. Most of these lesions are managed conservatively; avoiding excess pressure on the buttock, such as by proper padding of bicycle seats, is recommended. Masses affecting the PFCN are usually removed surgically.25,60

GLUTEAL NEUROPATHIES The superior gluteal nerve originates from the L4, L5, and S1 ventral rami, exits the pelvis through the suprapiriform foramen, and then passes through the gluteus medius and minimus muscles (both of which it innervates) before ending in the tensor fascia lata muscle. The inferior gluteal nerve originates from the L5, S1, and S2 spinal ventral rami and leaves the pelvis beneath the piriformis muscle, in close association with the sciatic and the posterior femoral cutaneous nerves; it innervates the gluteus maximus only.7 Lesions of the gluteal nerves are rare. Injuries of the superior gluteal nerve, before or after it supplies the gluteus medius and minimus muscles, may occur with intramuscular injections.30 The inferior gluteal nerve and the nearby PFCN may be compressed by intrapelvic masses, such as colorectal malignancies or large iliac artery aneurysms. Both gluteal nerves may be injured very severely with gluteal compartment syndromes, and usually subclinically during total hip replacement.25,60 Weakness of abduction of the affected hip causing a waddling gait is the main manifestation of superior gluteal neuropathies. When patients attempt to stand on one leg, there is a pelvic tilt toward the healthy side (Trendelenburg’s sign). Buttock atrophy and weakness of hip extension are the common features of inferior gluteal neuropathies.25 Gluteal neuropathies are difficult to diagnose without EDX confirmation. Close inspection and examination of the buttock are essential. Disorders of the hip joint often mimic these lesions and should always be excluded. The EDX abnormalities are limited to needle EMG. Fibrillation potentials and loss of MUAPs in the tensor fascia lata and gluteus medius are features of superior gluteal neuropathies; the gluteus medius may be spared when the lesion is distal, such as with intramuscular injections. Isolated denervation of the gluteus maximus is the only finding in inferior gluteal neuropathies, although an abnormal PFCN SNAP or evidence of sciatic nerve involvement may coexist. The EDX findings in gluteal neuropathies should be distinguished from those seen with L5 and S1 radiculopathies. Because these nerves are short, most partial lesions improve significantly without intervention.25,26

SCIATIC NEUROPATHY Anatomy The sciatic nerve is formed from convergence of the L5, S1, and S2 ventral rami, with a small contribution from the L4 ventral ramus, after the superior and inferior gluteal nerves arise.7 The nerve is composed of a lateral division (common peroneal nerve, or lateral popliteal nerve) and a medial division (tibial nerve, or medial popliteal nerve),

1494

Diseases of the Peripheral Nervous System

which are enclosed in a common sheath. These two divisions are separate from each other, however, and do not exchange any fascicles.61 The sciatic nerve leaves the pelvis via the greater sciatic notch and usually passes underneath the piriformis muscle. However, in 10% to 30% of subjects the peroneal division alone, or the entire sciatic nerve, passes through or above the piriformis muscle. The sciatic nerve innervates all four hamstring muscles and, partially, the adductor magnus. All of these muscles are innervated by the tibial component of the nerve with the exception of the short head of the biceps femoris, which is innervated by the common peroneal division. Slightly proximal to the popliteal fossa, the sciatic nerve terminates by dividing into its common peroneal and tibial components7,61 (Fig. 61–3).

FIGURE 61–3 The sciatic nerve, the major nerve of the lower limb. It extends throughout most of the thigh and innervates the hamstring muscles before terminating as the common peroneal and tibial nerves.

Etiology Total hip joint replacement is a leading cause of sciatic neuropathies. Conversely, these nerve injuries are the most common neurologic complication of this operation, with an estimated incidence of 1% to 3%. This incidence is higher in revisions, with congenital hip dislocation or dysplasia, and with limb-lengthening procedures.52,60 Subclinical evidence of nerve damage has been detected by needle EMG in as many as 70% of patients undergoing total hip replacement.68 The causes include direct intraoperative stretch injury, hemorrhage (into the gluteal compartment), prosthetic dislocation, migrating trochanteric wire, and leaking cement (methylmethacrylate).52,55 Hip fractures, hip dislocations, closed reductions of hip fractures, and femur fractures may produce sciatic nerve injuries.11,60 Sciatic neuropathies may present after prolonged external compression of the buttock or posterior thigh, both when the patient is lying supine, such as with coma (during drug overdose or with poor positioning in the intensive care unit), or when sitting (during operations, or on toilet seats or firm chairs).11,25,55,60,74 Many of these cases are associated with a gluteal compartmental syndrome. Hemorrhage within the gluteus maximus compartment sometimes may also occur spontaneously during anticoagulant therapy, in hemophiliacs, or following rupture of an iliac artery aneurysm or hip surgery. Although the sciatic nerve is not within this compartment, it can be compressed when a gluteal compartment syndrome is present.16,55 Improperly administered intramuscular gluteal injections may damage the sciatic nerve. These injuries occur mainly in thin or emaciated adults and in children; with the latter, subsequent limb growth is disturbed.60,61 Cyclic radicular pain (i.e., sciatica) or overt sciatic neuropathy may accompany endometriosis.51 Typically, the symptoms start a few days before menstruation and stop after menses end. With time, they become more constant, though often worse during menses. Other mass lesions, such as malignant or benign tumors in the buttock or thigh, persistent sciatic artery, or enlargement of the lesser trochanter (possibly from frequent sitting on hard benches), may compress the sciatic nerve.25 Ischemia of the sciatic nerve may be due to small vessel disease, as with vasculitis involving the vasa nervorum, or large vessel disease, as with iliac or femoral artery occlusion. It also can occur during intra-aortic balloon pump therapy via the femoral artery.42 Finally, idiopathic progressive sciatic neuropathy is a diagnosis of exclusion when all tests—including at times exploratory surgery—fail to reveal a cause.74,75

Clinical Manifestations Sciatic neuropathies resulting from hip replacement typically manifest symptoms in the immediate postoperative period. The symptoms may be delayed up to several years,

1495

Mononeuropathies of the Lower Limb

however, when caused by prosthetic dislocation, osseous formation, or migrating trochanteric wires.25 Severe sciatic neuropathies generally present with pain and hyperesthesias, mostly on the dorsal aspect of the foot and the sole. The foot is flail, and all muscles below the knee, in addition to the hamstrings, are weak. Sensory loss mostly is detectable below the knee, with sparing of the medial leg (saphenous nerve distribution). Posterior thigh sensation is unaffected unless there is concomitant involvement of the PFCN.11,25,60 With partial sciatic nerve lesions, the common peroneal component often is more severely affected than the tibial component. The greater vulnerability of the former is due to (1) its having fewer and larger fascicles with less supportive tissue, rendering it more susceptible to external pressure than the tibial nerve; and (2) its anatomic course, predisposing it to traction injuries because it is taut and secured at the sciatic notch and fibular neck, whereas the tibial nerve is loosely fixed.60 Clinically, partial sciatic nerve lesions typically manifest principally as footdrop. Though weakness of the tibialis anterior muscle dominates, clinical examination usually detects weakness of the hamstrings (knee flexion), gastrocnemius (plantar flexion), or tibialis posterior (ankle inversion). The sensory loss usually is predominant over the dorsum of the foot. Rarely, sciatic neuropathies in the hip are essentially pure high common peroneal lesions, manifesting similarly to common peroneal lesions at the fibular neck, except that the short head of the biceps femoris is affected.11,25,29,60

gluteal compartmental syndromes, may reveal a hematoma or significant compartmental swelling. These neuroimaging studies are also useful in detecting endometriosis affecting the sciatic nerve.11,25 The EDX findings in sciatic neuropathies parallel the clinical manifestations. With severe lesions, the NCSs reveal that the sural and superficial peroneal sensory SNAPs, as well as the peroneal and tibial CMAPs, are absent or low in amplitude. The needle EMG discloses fibrillation potentials and neurogenic MUAP changes in all muscles below the knee and in the hamstrings. With partial or mild sciatic neuropathies, the EDX examination may reveal findings mainly suggestive of an axon loss common peroneal neuropathy, because those fibers often are affected more severely than the tibial nerve fibers. This may include absent or low-amplitude superficial peroneal SNAPs and peroneal motor CMAPs. However, detailed assessments usually demonstrate involvement of the tibial nerve fibers that may not have been detected clinically. Thus an asymmetrically low-amplitude or absent sural SNAP or H reflex may be seen. The needle EMG in these situations frequently shows severe denervation in the common peroneal nerve–innervated muscles, including the short head of the biceps femoris, with lesser denervation in the muscles innervated by the tibial nerve, including the internal hamstrings, long head of the biceps femoris, and gastrocnemius. Needle EMG of the glutei, tensor fascia lata, and lumbar paraspinal muscles is normal in patients with sciatic neuropathies, thereby helping to exclude plexopathies or radiculopathies.11,18,25,75

Diagnosis Substantial sciatic neuropathies pose little difficulty in diagnosis because weakness involves the hamstrings and all muscles below the knee, and there is a diffuse sensory loss below the knee that spares only the medial leg. In contrast, those partial lesions that present principally with footdrop may mimic peroneal neuropathies, lumbosacral plexopathies, or even lumbosacral radiculopathies. Useful features that distinguish a partial sciatic neuropathy from a common peroneal neuropathy include the presence of dysesthetic foot pain, a decreased Achilles tendon reflex, and weakness of ankle inversion or toe flexion. Similarly, the lack of back pain, radiating radicular limb pain, and well-demarcated sensory loss favors a sciatic neuropathy over an L5 or S1 radiculopathy.11,25,60 Sacral plexopathies, particularly those involving the lumbosacral trunk (cord), may be due to pelvic fracture or fetal head compression during labor in women with short stature who are delivering large babies. They present principally with footdrop. Often detailed EDX studies are necessary for correct localization.27 Plain hip radiographs are useful for detecting hip dislocations or fractures. A CT scan or MRI of the pelvis and gluteal compartments, done urgently in patients with possible

Treatment and Prognosis Patients with gluteal compartment syndrome should undergo emergency surgical decompression. Surgical intervention is also indicated for many patients with total sciatic neuropathies caused by hip fractures, dislocations, or prosthetic hip dislocations. Partial or complete surgical excision should be attempted with sciatic neoplasms. Lesions resulting from hip replacements and misplaced intramuscular injections usually are managed conservatively, but occasionally neurolysis is performed for pain control. Patients with mild manifestations caused by endometriosis may respond to pharmacologic suppression of the ovarian cycle, but those with severe symptoms require surgical resection of endometrial foci and adhesions. Because sciatic neuropathies typically are painful, pain management is necessary. Tricyclic antidepressants and anticonvulsants are the drugs of choice. A cock-up foot brace or ankle-foot orthosis assists patients in ambulation.11,25,60 The prognosis of sciatic neuropathies is guarded. In general, those patients with severe lesions have significant residual weakness, while those with partial lesions may improve significantly as a result of collateral nerve sprouting. At least three fourths of lesions sustained

1496

Diseases of the Peripheral Nervous System

during total hip arthroplasty leave permanent motor residuals. Many patients are left with chronic foot pain and symptoms of reflex sympathetic dystrophy.25

Piriformis Syndrome The piriformis syndrome is a nebulous, controversial entity. Presumably, pain in the buttock (“sciatica”), with or without accompanying back pain, results from sciatic nerve compression at the pelvic outlet by the piriformis muscle. Because the neurologic examination usually is normal, the diagnosis is difficult.11,60 To reproduce the pain in the affected buttock (and sometimes thigh), several provocative procedures are used, all of which result in passive stretching or active contraction of the piriformis muscle. Unfortunately, these maneuvers may be positive in many other disorders of the spine, hip joint, and buttock.56,60 Symptom relief by corticosteroid injection into the sciatic notch region is also used for diagnosis, but this may relieve pain with a variety of lesions. Recently, MRI has identified abnormal vessels or bands in the region of the piriformis muscle as well as hypertrophy of the muscle itself. Such changes, however, are as common in control patients and in asymptomatic limbs as in symptomatic limbs (C. Petersilge, personal communication, 2000). Only rarely does the EDX examination reveal any abnormalities. Treatment is difficult. Conservative therapy includes prolonged stretching of the muscle by flexion, adduction, and internal rotation of the symptomatic hip. Corticosteroid injections into the muscle, preferably done under imaging guidance, may relieve symptoms. Surgical exploration is done in resistant cases. Abnormal bands or vessels are often removed, but sectioning of the piriformis muscle is the most popular procedure, although of unproven value.25,60 The prognosis with the piriformis syndrome is unknown; in one series, however, favorable outcome occurred in 12 of 14 patients treated surgically.73

lateral sural cutaneous nerve, which joins the sural nerve (Fig. 61–4). The common peroneal nerve leaves the popliteal fossa by winding around the fibular neck, where it passes through the “fibular tunnel,” formed between the tendinous edge of the peroneus longus muscle and the fibula. Usually, it divides into its terminal branches, the deep and superficial peroneal nerves, near the fibular neck, but in 10% of individuals this division occurs proximal to the knee joint.7,13,61 The deep peroneal (or anterior tibial) nerve is the largest of the two terminal branches. It traverses the proximal lateral compartment of the leg to enter the anterior compartment, where it tracks distally and innervates the tibialis anterior, extensor hallucis longus, peroneus tertius,

COMMON PERONEAL NEUROPATHY Anatomy Nerve fibers composing the common peroneal nerve originate from the L4 and L5 ventral rami with a minor contribution from the S1 ventral ramus. From its origin, the common peroneal nerve shares a common sheath with the tibial nerve to form the sciatic nerve. A single motor branch originates from the common peroneal division in the thigh to innervate the short head of the biceps femoris. All other hamstring muscles are innervated by the tibial division. The common peroneal and tibial nerves separate from one another proximal to the popliteal fossa. Two nerves arise from the common peroneal nerve in the popliteal fossa: the lateral cutaneous nerve of the calf, which innervates the skin over the upper third of the lateral aspect of the leg, and the

FIGURE 61–4 The common peroneal nerve. This nerve extends from the distal thigh to the foot. It innervates the muscles in the anterior and lateral compartments of the leg and supplies sensation to the anterolateral portion of the leg and the dorsum of the foot (the lateral cutaneous nerve of the calf and the lateral sural cutaneous nerve are not shown).

1497

Mononeuropathies of the Lower Limb

and extensor digitorum longus muscles. After passing under the extensor retinaculum at the ankle, it divides into two terminal branches, a lateral branch that innervates the extensor digitorum brevis and a medial branch that supplies the skin of the web space between the first and second toes. The superficial peroneal nerve proceeds distally in the lateral compartment, during which it innervates the peroneus longus and brevis muscles, and then becomes the superficial peroneal sensory nerve. This nerve becomes subcutaneous proximal to the lateral malleolus by piercing the crural fascia, and then innervates the skin of the lower two thirds of the lateral aspect of the leg and the dorsum of the foot (except for the first web space).7,61

Etiology Common Peroneal Neuropathy at the Fibular Neck Lesions of the common peroneal nerve at the fibular neck are the most common major lower limb neuropathies encountered. The superficial location of the nerve and its close relation to the fibular neck places it at high risk at this site. Compression, traction, and other types of trauma are common causes. Compressive injuries often are complications of coma, or operations performed under general anesthesia; they are due to malpositioning of legs and failure to properly pad the knees. They also occur after weight loss, because subcutaneous fat padding around the nerve is reduced, exposing it to direct pressure such as during habitual leg crossing. A history of weight loss should be sought in all patients presenting with unexplained common peroneal neuropathies at the fibular neck. In general, patients with diabetes mellitus, polyneuropathy, or both, have a higher incidence of these lesions. Prolonged squatting, such as during crop harvesting, yoga meditation, and military training maneuvers, may stretch the common peroneal nerve around the fibular neck. These neuropathies also can develop in women during childbirth, presumably as a result of nerve compromise caused by squatting, hand pressure, or leg holders used with lithotomy positioning. Similarly, surgical procedures that employ lithotomy positioning, particularly with the use of stirrups, such as hysterectomies or prostatectomies, may cause common peroneal neuropathies, as can devices that apply pressure at the fibular neck, including knee casts, ankle-foot orthoses, pneumatic splints, bandages, straps, and antithrombotic stockings.10,24,28,58,60 Traumatic common peroneal neuropathies at the fibular neck can be open, blunt, or iatrogenic. Open injuries usually are due to lacerations, gunshot wounds, or animal bites. Blunt injuries often result from ligamental ruptures of the knee joint, fibular fractures, knee dislocations, or dislocations of the superior tibiofibular joint, such as can be sustained in vehicular accidents or during athletic activities. Traction injuries of the nerve, or one of its terminal branches, may follow inversion ankle sprains. Finally,

iatrogenic lesions occur during procedures in which the common peroneal nerve is in or near the operative field, such as with conventional knee surgery, total knee replacement, knee arthroscopy, and tibial osteotomies.10,12,13,24 The common peroneal nerve may occasionally be compressed at the fibular neck by a mass or be the site of a benign or malignant nerve tumor.25 Deep Peroneal Neuropathy at the Fibular Neck In lesions of the common peroneal nerve at the fibular neck, the deep peroneal branch often is more severely affected than the superficial branch. Occasionally, however, the deep peroneal nerve alone is injured. This is caused by the direct contact of the medially placed deep peroneal nerve with the fibular bone while the exiting fascicles of the superficial peroneal nerve are more lateral. Well-known causes for these selective proximal deep peroneal neuropathies include intraneural ganglion cysts, osteochondromas, inversion injuries of the foot (ankle sprains), and arthroscopic knee surgery.12,24,25,60 High Common Peroneal Neuropathy (Sciatic Neuropathy Involving the Common Peroneal Nerve Exclusively) Partial lesions of the sciatic nerve at the hip usually affect the common peroneal component more than the tibial component. Rarely, however, the common peroneal division alone is injured. The causes are similar to those affecting the proximal sciatic nerve in general.29

Clinical Manifestations Common peroneal neuropathies are almost three times more common in men and affect all age groups. Although most are unilateral, about 10% are bilateral. Common and proximal deep peroneal neuropathies present with complete or partial footdrop that may lead to repeated falls, as well as foot or ankle sprains and fractures. Often subjective numbness (loss of sensation more often than paresthesias) is present on the dorsum of foot, extending into the lower lateral leg and, in high (proximal) peroneal lesions, to the upper lateral leg and knee. A mild deep, boring pain around the knee or in the leg may occur. The onset of symptoms most often is acute, but occasionally may be subacute (gradual over days or weeks) or difficult to determine. In general, cases caused by a single episode of compression or direct trauma present acutely, whereas those associated with weight loss and leg crossing, or prolonged hospitalization, often present insidiously.10,24,25

Diagnosis The neurologic examination reveals weakness of ankle and toe dorsiflexion while ankle inversion, toe flexion, and ankle plantar flexion are normal. Ankle eversion may be

1498

Diseases of the Peripheral Nervous System

less impaired if the superficial branch is not as damaged as the deep branch (as is frequently the case), and it is spared in selective proximal deep peroneal neuropathy. Rarely, the deep peroneal lesion is more distal in the leg (such as with traction injuries associated with ankle sprains), resulting in selective weakness of toe extension with normal foot dorsiflexion. The hamstring muscles and deep tendon reflexes are normal. With common peroneal neuropathies, impaired touch and pain sensation often is detected over the lower two thirds of the lateral leg and dorsum of foot. With high (proximal) peroneal lesions, the sensory deficit may extend into the upper lateral leg and knee. A Tinel’s sign may be elicited by percussion of the nerve at the fibular neck, with some lesions located at that point.10,24,60 The differential diagnosis of peroneal neuropathies parallels that of footdrop and can be divided into three steps (Table 61–1): 1. Distinguish footdrop from flail foot, with which it is sometimes confused. With footdrop, the sole or predominant finding is foot dorsiflexion weakness, whereas with flail foot all foot movements are severely impaired. 2. Distinguish footdrop caused by an upper motor neuron lesion from that caused by a lower motor neuron lesion. Though most central nervous system disorders (such as cortical or subcortical cerebral lesions or myelopathies) result in flail foot, occasionally a selective deficit in foot dorsiflexion may dominate. Associated hyperreflexia, Babinski’s sign, hand manifestations, or sphincteric disturbance are helpful clues. 3. Distinguish peroneal neuropathies from sciatic neuropathies (affecting the common peroneal division predominantly), sacral plexopathies (particularly those involving the lumbosacral trunk), and lumbar radiculopathies (especially those involving the L5 root).25 Imaging studies such as plain radiographs may reveal bony abnormalities around the fibular neck, including tumors and dislocations. Ultrasonography of the popliteal fossa or fibular neck region may identify certain mass lesions, such as Baker’s cysts, ganglion cysts, or aneurysms. MRI can visualize soft tissue masses and cysts, but may not detect all intraneural ganglia or intrinsic nerve tumors.10,25,60 The EDX examination is extremely helpful in the assessment of these lesions. The most useful components are the peroneal sensory SNAP and the peroneal motor CMAP, recording from the extensor digitorum brevis, and especially the tibialis anterior, while stimulating in the popliteal fossa and below the fibular neck. The EDX examination (1) distinguishes common peroneal neuropathies at the fibular neck from those in the upper thigh or those selectively affecting the deep peroneal nerve, and also from sciatic neuropathies, sacral plexopathies, and L5 radiculopathies (see Table 61–1); (2) reveals the pathophysiology of these nerve lesions (demyelinating vs. axon loss vs. mixed), and thereby

aids in predicting prognosis; and (3) when performed sequentially with severe axon loss peroneal neuropathies, discloses the presence and extent of reinnervation. The EDX examination with peroneal neuropathies has three patterns: 1. Conduction block across the fibular neck. These lesions may be complete or partial. With partial block, there is a significant (often more than 50%) drop in peroneal CMAP amplitude when stimulating proximal to, compared to distal to, the fibular neck. The superficial peroneal SNAP usually is normal. The primary pathology is demyelination affecting some or all of the peroneal fibers, the extent of which may be determined by comparing the distal and proximal CMAP amplitudes. In contrast to demyelinating conduction block, demyelinating focal slowing across the fibular neck is rarely seen. Needle EMG reveals reduced MUAP recruitment in all muscles innervated by the common peroneal nerve distal to the knee.28 2. Axonal loss–caused conduction failure. These lesions may also be complete or partial; they may involve the common peroneal nerve at the fibular neck, the common peroneal division in the thigh, or the deep peroneal nerve. The peroneal CMAPs on proximal and distal stimulations are of equally low amplitude or absent. The motor conduction velocities across the fibular neck, when obtainable, are normal or slightly reduced diffusely, without focal slowing. The superficial peroneal SNAP usually is absent except with selective deep proximal peroneal neuropathies. On needle EMG, fibrillation potentials and reduced MUAP recruitment are seen in all deep peroneal nerve–innervated muscles, as well as in the peroneus longus (with common peroneal lesions) and the short head of the biceps femoris (with high common peroneal division lesions). EDX testing cannot detect the exact site of these lesions because they are not accompanied by a localizing conduction abnormality, such as a block. Based on the needle EMG findings, these lesions can be localized as (a) being between the fibular neck and the origin of the branch to the short head of the biceps femoris in the upper thigh; (b) being above mid-thigh (i.e., above the origin of the branch to the short head of the biceps femoris); or (c) involving the deep peroneal nerve selectively. 3. Mixed (conduction block and axon loss). These lesions are characterized by reduced amplitudes of the peroneal CMAP responses obtained on distal stimulation, accompanied by further reduction (i.e., partial or complete conduction block) across the fibular neck. The superficial peroneal SNAPs typically are low in amplitude or absent. Motor conduction velocities, when obtainable, usually are normal. Needle EMG reveals fibrillation potentials and reduced MUAP recruitment in all the common peroneal nerve–innervated muscles distal to the knee.24,28,69

1499

Mononeuropathies of the Lower Limb

Table 61–1. Differential Diagnosis of Common Causes of Footdrop Peroneal Neuropathy at the Fibular Head Clinical Common causes

Ankle inversion Toe flexion Plantar flexion Ankle jerk

Compression (weight loss, perioperative), trauma Normal Normal Normal Normal

Sensory loss distribution

Peroneal distribution only

Pain

Rare, deep

Electrodiagnosis Peroneal motor study to EDB and/or Tib ant Superficial peroneal sensory study Sural sensory study Peroneal muscles† Tibial L5 muscles‡ Other L5 muscles§ Biceps femoris (short head) Paraspinal muscles fibrillations

L5 Radiculopathy Disk herniation, spinal stenosis Weak Weak Normal Normal (unless with S1) Poorly demarcated, predominantly big toe Common, radicular

Lumbar Plexopathy (Lumbosacral Trunk) Pelvic surgery, hematoma, prolonged labor Weak Weak Normal Normal (unless with S1) Well demarcated to L5 dermatome Common, can be radicular

Sciatic Neuropathy (Mainly or Exclusively Peroneal) Hip surgery, injection injury, coma Normal or mildly weak Normal or mildly weak Normal or mildly weak Normal or depressed Peroneal distribution plus lateral cutaneous of calf and sole (tibial) Can be severe

Low in amplitude or conduction block across fibular head or both Low or absent*

Usually normal but can be low in amplitude

Low in amplitude

Low in amplitude

Normal

Low or absent

Low or absent

Normal

Normal

Abnormal Normal Normal Normal

Abnormal Usually abnormal Normal or abnormal Usually normal

Normal or low amplitude Abnormal Usually abnormal Normal or abnormal Usually normal

Normal or low amplitude Abnormal Normal or abnormal Normal Abnormal

Absent

May be absent

Absent

Absent

EDB ⫽ extensor digitorum brevis; Tib ant ⫽ tibialis anterior. *Can be normal in purely demyelinating lesions or lesion of the deep peroneal nerve only. † Below knee (tibialis anterior, extensor digitorum longus, extensor digitorum brevis, extensor hallucis, ⫾ peroneus longus). ‡ Tibialis posterior and flexor digitorum longus. § Gluteus medius and tensor fascia lata. Adapted from Katirji, B.: Electromyography in Clinical Practice. St. Louis, Mosby, 1998.

Treatment and Prognosis An ankle-foot orthosis and both active and passive rangeof-motion exercises, to improve gait and prevent ankle contractures and sprains, should be prescribed for patients with peroneal neuropathies, particularly those in whom the footdrop is profound and who are expected to have a protracted course. Patients with significant weight loss should be warned about leg crossing and should wear protective knee pads, properly placed over the proximal fibula to prevent repeated external compression. Axonal loss peroneal neuropathies at or around the fibular neck should be observed for 4 to 6 months to allow for improvement by spontaneous reinnervation. Sequential needle EMG studies of the tibialis anterior, the peroneus longus, or both may reveal MUAPs that are of low amplitude, of short

duration, and polyphasic (“nascent” MUAPs), characteristically the first sign of reinnervation.10,24,26,60 Surgical intervention is indicated with certain axon loss peroneal neuropathies. These include 1. With lacerated nerves that are visibly discontinuous. Repair can be done, usually by end-to-end anastomosis, at the time of laceration suturing (primary repair) or at a later date if local infection is anticipated (secondary repair). 2. With severe lesions, when there is no clinical or needle EMG evidence of reinnervation 4 to 6 months after injury. Intraoperative NCSs of these lesions provide optimal evaluations, while neurolysis, nerve cable grafting, or end-to-end repair, done in a timely fashion, often results in a satisfactory outcome.

1500

Diseases of the Peripheral Nervous System

3. With those caused by mass lesions such as nerve tumors or ganglion cysts. Preoperative EDX studies combined with MRI may delineate the nature of the mass structure and thereby help determine the optimal surgical procedure.4,24,31 The prognosis of peroneal neuropathies depends on their pathology. With focal demyelination, the prognosis is excellent: Recovery generally occurs within 2 to 3 months provided the cause of compression is eliminated. In contrast, axon loss lesions have a more uncertain prognosis because recovery is dependent on proximodistal reinnervation or collateral sprouting. Nonetheless, peroneal neuropathies at the fibular neck can reinnervate relatively quickly because the lesion site is not far from the main target muscles. Consequently, the ultimate result often is satisfactory. The prognosis is much worse with high peroneal neuropathies, because of the long distance to the main target muscles (anterior compartment muscles). In general, partial lesions recover faster as a result of collateral sprouting, whereas complete lesions recover slower, because reinnervation occurs at a rate of 1 mm/day in a proximaldistal pattern. Recovery of the neurologic deficit in mixed lesions (demyelination and axon loss) is usually biphasic; the first phase, which is due to remyelination, is relatively rapid, over 2 to 3 months, whereas the second phase, resulting from reinnervation, is more protracted, extending into several months or a year.24,25

tibialis posterior, flexor digitorum longus, and flexor hallucis longus—all of which it innervates. At the ankle, it passes posterior and inferior to the medial malleolus to enter the tarsal tunnel. In this region, it gives off the medial calcaneal nerve and then soon divides into its two terminal branches, the medial plantar and lateral plantar nerves7,16,54,61 (Fig. 61–5).

Etiology Overall, tibial nerves are damaged much less often than are the common peroneal nerves. Thus, among Seddon’s series of 365 sciatic neuropathies, 136 (37%) involved

TIBIAL NEUROPATHY Anatomy The tibial nerve originates principally from the L4, L5, S1, and S2 ventral rami. When the sciatic nerve leaves the pelvis through the greater sciatic foramen, these axons are its medial, larger component. As the sciatic nerve passes distally in the thigh, its tibial component provides innervation to three of the four hamstring muscles (semitendinosus, long head of biceps femoris, and semimembranosus) and the adductor magnus.7 The exact order of muscle innervation varies considerably, and frequently each muscle is supplied by more than one branch61 (see Fig. 61–3). Near the apex of the popliteal fossa, the sciatic nerve divides into its two terminal branches: the common peroneal nerve and the tibial nerve. The latter continues distally in the midline; after entering the popliteal fossa, where it is relatively superficial, it gives off the sural nerve as well as the motor branches to the gastrocnemius (both medial and lateral heads) and three other muscles (the plantaris, soleus, and popliteus). The tibial nerve enters the leg when it passes beneath the fibrous arch of the soleus muscle, and is now often referred to as the posterior tibial nerve. As it continues its distal course, it is situated in the deep posterior compartment of the leg with the posterior tibial vessels, and with three muscles—the

FIGURE 61–5 The tibial nerve. This nerve extends from the distal thigh to the foot. It innervates the muscles in the posterior compartment of the leg and most of those in the foot, and it supplies sensation to the sole of the foot.

1501

Mononeuropathies of the Lower Limb

common peroneal nerve fibers solely, whereas only 20 (5.4%) affected tibial nerve fibers alone. Among the latter, 19 were due to missile injuries and one to a ganglion cyst.54 The reasons for this are discussed above. Nonetheless, any portion of the tibial nerve or the tibial division of the sciatic nerve from which it derives can be injured. In one recently reported series of 52 patients with tibial neuropathies—which excluded sciatic neuropathies in which the peroneal divisions were also affected, distal tibial neuropathies within the tarsal tunnel, and lesions caused by systemic disorders (e.g., vasculitis)—25% occurred proximal to the popliteal fossa, 48% in the popliteal fossa, and 27% distal to it. Among the last group, in which the tibial nerve was affected in the leg, common peroneal and deep peroneal neuropathies frequently coexisted. Regarding etiology, trauma was responsible for 56% of tibial neuropathies, whereas the principal nontraumatic causes were ischemia (19%) and neoplasms (17%).14 Tibial neuropathies situated proximal to the popliteal fossa are of variable etiology. Compression from casts and tourniquets, blunt trauma with or without associated femur fractures, gunshot wounds, and lacerations are major causes, along with various space-occupying lesions, including Baker’s cysts, localized hypertrophic neuropathy, and neoplasms (e.g., schwannomas). Rarely, only the tibial division of the sciatic nerve is injured in or near the gluteal crease during hip operations or by misplaced intramuscular gluteal injections.14,52,60 The tibial nerve often is injured in the popliteal fossa. This nerve segment is more vulnerable because it is superficial, near a joint, and contiguous to major blood vessels. Trauma is a frequent cause, occasionally including knee dislocations. The nerve can be injured both by closed and by penetrating injuries; with the latter, it can be damaged directly as well as indirectly, as a result of initial injury to the nearby popliteal vessels with compromise of the tibial nerve by the subsequent hemorrhage or pseudoaneurysm that forms. Iatrogenic lesions that occur at this level include those caused by various surgical procedures, such as knee arthroscopy and arthroscopic surgery, open operations on the knee joint, and vascular repairs.25,31,60 The tibial nerve is not injured as frequently during surgery performed on the knee as are several other lower limb nerves. Thus, in one study of 63 neuropathies resulting from arthroscopic knee surgery, 25 involved the peroneal nerve, 23 the saphenous, and 7 the femoral, while only 4 each affected the sciatic and tibial nerves.12 Ischemic nerve injuries can result from femoral arteriograms performed to assess the lower limb arterial blood supply. In these cases, the advancing catheter tip in the atheromatous artery causes an embolus to lodge in one of the nutrient vessels of the tibial nerve.14 Tibial neuropathies occurring distal to the popliteal fossa have several causes, a common one being tibial fractures (often accompanied by deep peroneal neuropathies).

Other etiologies include hemorrhage into Baker’s cysts, with subsequent formation of a deep posterior compartment syndrome of the leg; nerve sheath tumors; entrapment by the tendinous arch formed by the two heads of the soleus muscle; and kinking of the nerve, and often the vessels accompanying it, by one or more fibrous bands extending between the lateral and medial heads of the gastrocnemius.40,46,60 A particularly severe type of tibial neuropathy occurs in patients with ruptured calcaneal tendons in whom the goal is to harvest the long slender tendon of the vestigial plantaris muscle to serve as a tendon graft but, instead, the tendon stripper removes a very long segment of the tibial nerve.41

Clinical Manifestations Tibial neuropathies present with various combinations of motor, sensory, and autonomic disturbances, depending upon the level and severity of the lesion. Substantial injuries to the tibial division of the sciatic nerve in the proximal thigh result in (1) sensory disturbance in the distribution of the sural, medial plantar, lateral plantar, and medial calcaneal nerves (i.e., on the lateral aspect, the sole, and the heel of the foot) along with anhidrosis of the sole of the foot (Fig. 61–6); and (2) weakness, and often atrophy, involving the tibial nerve–innervated muscles, including the three hamstrings. Thus knee flexion, foot plantar flexion and inversion, and toe flexion are affected. Claw toes may be evident. With lesions more distal, in or near the popliteal fossa, the hamstrings will be spared. In addition, depending upon the exact location of the lesion, sensation in the sural nerve distribution, foot plantar flexion strength (gastrocnemius and soleus), and the Achilles deep tendon reflex may be compromised or intact. With tibial lesions in the leg, only the muscles in the deep posterior compartment may be weak, although sensory disturbance and anhidrosis on the sole of the foot persist.21,25,38 Not all focal tibial neuropathies, however, adhere to the above patterns. Baker’s cysts, for example, may compress not only the tibial nerve but also the common peroneal nerve. In addition, a ruptured Baker’s cyst can manifest as a more distal tibial neuropathy, by producing a deep posterior compartment syndrome in the leg. Moreover, certain tibial nerve lesions have characteristic symptomatology that differs from the symptoms described above. Thus entrapment in the proximal leg, caused by the tendinous arch of origin of the soleus muscle, produces severe pain and tenderness in the popliteal fossa and upper calf regions, with the former increased by weight bearing and passive dorsiflexion of the foot; although most patients also experience paresthesias on the sole or heel of the foot, only about half of them have definite weakness of foot muscles. With the pseudoentrapment syndrome, in which the nerve is kinked in the proximal leg by a transverse fibrous band, calf pain and

1502

Diseases of the Peripheral Nervous System

FIGURE 61–6 A, The tarsal tunnel and its neural contents—the terminal portion of the tibial nerve, the medial plantar nerve, the lateral plantar nerve, and the medial calcaneal nerve—are illustrated, as are the digital nerves. B, The sensory distribution of the five nerves that supply sensation to the sole of the foot is shown.

foot paresthesias occur when the patient actively flexes the knee or toe walks; occasionally swelling of the distal lower limb and foot also is seen as a result of concomitant kinking of the vessels.8,40,43,46 Overall, among 52 patients with tibial neuropathies, sensory complaints were present in 79%, and weakness of foot plantar flexion, toe flexion, and foot inversion in 62%.14 Among the major nerves of the lower limb, the tibial nerve, when injured, has two manifestations that are nearly unique to it. First, both initially and after surgical repair, axon loss lesions often result in marked hypersensitivity (hyperpathia) and pain in the sole and heel of the foot. Second, severe (total or near-total) axon loss tibial neuropathies render the sole of the foot insensitive and therefore very much at risk to develop pressure ulcers, which can be quite disabling.4,31,44,54

Diagnosis At times, the cause and location of a tibial neuropathy are obvious from the history. In other cases they are not, often

resulting in marked delays in diagnosis. The diagnosis of tibial neuropathies typically depends on the clinical manifestations and various laboratory studies. On neurologic examination, strength or weakness of various muscle groups (e.g., hamstrings, gastrocnemius) helps determine the level of the lesion, as does intact or impaired sensation in a sural distribution and a normal or reduced Achilles tendon reflex. Palpation along the course of the nerve may reveal a mass or marked tenderness, particularly when the lesion is situated in or near the popliteal fossa and is relatively superficial.38,60 A Tinel’s sign may pinpoint the site of involvement, especially with nerve sheath tumors and entrapments,60 but it was present in only 10% of the 52 patients with tibial neuropathies in one study.14 The most frustrating situation is when sensory complaints are limited to the sole and heel of the foot, with or without some accompanying changes in the tibial nerve–innervated intrinsic foot muscles. Such presentations can result from distal tibial neuropathies, but also from incomplete axon loss lesions located at any point along the tibial nerve fibers, from the sciatic notch distally.25,38,60 Patients may undergo unsuccessful tarsal tunnel release to treat their severe sole pain, only to have tibial nerve neoplasms situated at or proximal to the popliteal fossa discovered months later.25,60 The differential diagnosis of tibial neuropathies includes S1 and S2 radiculopathies, sacral plexopathies, sciatic neuropathies, and particularly tarsal tunnel syndrome (TTS), especially with partial tibial neuropathies, in which the symptoms and deficits may be quite restricted.38,60 A variety of neuroimaging studies as well as EDX examinations are valuable in assessing these patients, depending upon the lesion present. Plain radiographs are of most benefit with traumatic cases, wherein they disclose fractures and occasionally dislocations; they may also reveal some soft tissue masses. Arthrograms are helpful in determining the source of Baker’s cysts, whereas arteriograms demonstrate various vascular lesions (e.g., popliteal fossa aneurysms). Doppler ultrasound can detect space-occupying lesions and congenital bands and is less expensive than CT or MRI. Nonetheless, the latter two are the most versatile, especially MRI, which is considered the most useful neuroimaging study overall.25,38,46,60 On EDX examination, the tibial nerve can be assessed extensively. Informative NCSs include the H wave test (both the H wave and the direct M wave), the sural SNAP, tibial motor CMAPs (recording from both the abductor hallucis and abductor digiti quinti pedis), and the mixed medial and lateral plantar nerve action potentials (NAPs). With most tibial neuropathies, substantial abnormalities are seen on EDX examination.25,38 Thus, among 52 patients, low-amplitude or unelicitable responses were found in 92% of tibial motor NCSs, 65% of H wave tests, and 61% of sural NCSs, whereas needle EMG revealed abnormalities in the abductor hallucis or abductor digiti

1503

Mononeuropathies of the Lower Limb

quinti pedis muscles in 94%, in the tibialis posterior or flexor digitorum longus in 81%, and in the gastrocnemii or soleus muscles in 73%.14

Treatment and Prognosis The appropriate treatment for tibial neuropathies depends principally on their etiology and severity. Closed injuries, such as those caused by compression, blunt trauma, and tibial fractures, usually receive conservative therapy if the nerve damage is incomplete. However, total axon loss, concomitant vascular damage, or subsequent persistent severe pain may necessitate surgical neurolysis or nerve repair at some point. Many open injuries require operation, particularly lacerations, which at times can be repaired immediately. In contrast, gunshot wounds may be managed conservatively or with delayed surgery, depending on the completeness of the lesion and evidence of nerve regeneration. The treatment of ischemic injuries varies with the cause. Those sustained during femoral arteriograms are managed conservatively, whereas those resulting from compartment syndrome or large artery obstruction may be treated surgically, particularly if diagnosed quite early, before axon loss has occurred. Neoplasms are surgically removed, if possible. A variety of treatments have been employed with Baker’s cysts that are compromising the tibial nerve, including conservative therapy (such as rest, leg elevation, and anti-inflammatory agents), aspiration of the knee joint (from which the Baker’s cyst arose), and surgical extirpation of the Baker’s cyst itself. Unfortunately, high recurrence rates have been reported following operative removal. Tibial nerve entrapment in the proximal calf has been treated successfully both by surgical division of the tendinous arch and with conservative therapy, whereas kinking of the nerve by transverse bands is alleviated by operative division of the bands.4,25,31,38,40,43,46 Any axon loss tibial neuropathy so severe as to seriously impair sensation on the sole of the foot requires special attention to be directed toward shoe wear, to avoid the development of ulcers.31 In fact, any injury that has resulted in loss of such protective sensation probably merits an operative attempt, employing nerve grafting, to reestablish sensation so as to prevent ulcerations.31 When lesions at the popliteal fossa are treated in this manner, a minimal postoperative period of 2 years—and sometimes much longer—is required before the ultimate degree of sensation is restored.44 Unfortunately, in some patients, not only does some sensation return, but it is accompanied by hypersensitivity and pain.4,54 The prognosis of tibial neuropathies varies with the level and, to a certain extent, with the severity of the lesion. Concerning those at or proximal to the popliteal fossa, the gastrocnemii often can be reinnervated, but generally not the intrinsic foot muscles. If foot plantar flexion (gastrocnemius and soleus) and inversion (tibialis posterior) are

recovered, gait essentially has been restored to normal. Similarly, if protective sensation can be restored to the sole of the foot without coexisting hyperpathia, much has been achieved. One group of surgeons has reported that tibial neuropathies at the knee or leg level are among the more favorable of all nerve injuries to manage.31 However, other surgeons have reservations regarding these repairs, because of the frequent occurrence of pain and hyperpathia affecting the sole of the foot once nerve regeneration occurs. The intensity of this disability varies, but in some patients it is very disabling.4,54

TARSAL TUNNEL SYNDROME The term tarsal tunnel syndrome applies to two distinct entities: anterior TTS, which involves the distal deep peroneal nerve and is principally an incidental finding in the EDX laboratory, and medial TTS, which involves the terminal tibial nerve and branches that arise from it. Of the two, the latter is of far more clinical importance. For this reason, whenever “tarsal tunnel syndrome” is used without qualification, it is always assumed to refer to medial TTS.25 Medial TTS is defined as entrapment of the tibial nerve or the proximal portion of one of its major branches within the tarsal tunnel, on the medial side of the ankle. Presumably, TTS is the most common of the various nerve entrapment syndromes of the foot and ankle.64 Nonetheless, it is probably the most overdiagnosed of the various nerve lesions of the lower limb. Thus most patients referred to our EDX laboratories with this diagnosis are shown to have problems other than distal tibial neuropathies.

Anatomy The tarsal tunnel is the space situated at the distal, medial aspect of the lower limb, posteroinferior to the medial malleolus (see Fig. 61–6A). It has an osseous floor, while its roof is the flexor retinaculum (or laciniate ligament), which is a thin, fibrous structure extending between the medial malleolus and the medial tuberosity of the calcaneus. It contains tendons, blood vessels, and nerves. The tendons are those of the tibialis posterior, flexor digitorum longus, and flexor hallucis longus muscles, while the vessels are the tibial artery and vein. The tibial nerve terminates within the tarsal tunnel by dividing into medial and lateral plantar nerves. The medial calcaneal nerve usually arises from the more distal portion of the tibial nerve immediately before the latter terminates, but its origin is quite variable; in many instances, it originates prior to where the tibial nerve enters the tunnel. The medial and lateral plantar nerves leave the tarsal tunnel to enter the foot by passing through openings in the fascial origin of the abductor hallucis muscle, which are referred to as the abductor tunnel.3,7,37,45,49

1504

Diseases of the Peripheral Nervous System

Etiology TTS mainly affects adults, has a slight female predominance, and usually occurs unilaterally.6,60 It has a variety of causes that can be broadly classified as trauma, spaceoccupying lesions, foot deformities, and idiopathic.37 Traumatic or idiopathic causes are the most common. Among the former are displaced fractures of the distal tibia, the tarsal bones, and the calcaneus; medial ankle sprains; and trauma-induced perineurial fibrosis and flexor tenosynovitis. Space-occupying lesions can be both within and external to the tarsal tunnel, and include varicosities (the most common), neural sheath tumors, ganglia, lipomas, and bony exostoses. Foot deformities can be both fixed and dynamic (flexible); these include varus heel with pronated forefoot; valgus heel with abducted forefoot (pes planus), and foot alterations caused by shoe wear. A relatively small number of patients have systemic diseases, particularly diabetes mellitus and inflammatory arthritis.3,6,37,45,64 The etiology of TTS has been linked in a somewhat vague fashion to double-crush syndrome,37,49,53 although very serious questions can be raised regarding the conceptual, anatomic, and experimental basis of the latter.71

Clinical Manifestations The symptoms of TTS typically are insidious in onset. A combination of foot pain (often having a burning quality), paresthesias, and numbness in the territory of one (less often, more than one) of the branches of the tibial nerve is characteristic. The distribution of the sensory disturbances depends on the nerve affected: the medial two thirds of the sole of the foot with medial plantar neuropathies, the lateral one third of the sole with lateral plantar neuropathies, and the heel with medial calcaneal neuropathies. When either the medial or the lateral plantar nerve is involved (the most common presentations), the symptoms frequently are more prominent distally, in the toes and the distal portion of the foot. In contrast, isolated heel pain is rare. In some patients, pain radiates proximally to the calf or above (referred to as the Valleix phenomenon). Typically, the symptoms initially are intermittent, but often they eventually become constant. They are triggered or aggravated by prolonged weight bearing, such as standing, walking, and running, and by the use of certain shoe wear, such as high-heeled shoes; they are relieved by rest, foot elevation, and shoe removal. In some patients, sleep is disturbed.3,37,25,45,60

Diagnosis The diagnosis of TTS is based principally on the clinical history and the physical examination. Some patients report prior foot trauma. Tinel’s sign with radiation in the distri-

bution of the affected nerve is considered by many investigators to be such a constant finding that its absence places the diagnosis in doubt. A positive “nerve compression test,” in which gentle pressure applied over the tarsal tunnel for 30 seconds reproduces the symptoms, also reportedly has a high positive yield. Although weakness and atrophy of the intrinsic foot muscles are uncommon findings, some type of foot deformity can be observed in many instances.3,37,45,49 Symptoms caused by varicosities frequently can be reproduced by proximal limb compression using a pneumatic tourniquet, which causes venous engorgement within the tarsal tunnel. Placing the foot in various positions, such as maximal dorsiflexion or heel eversion or inversion, may reproduce or aggravate the symptoms. Conversely, an injection of corticosteroids into the tarsal tunnel may alleviate symptoms and thereby serve as a diagnostic test.37,49 Neuroimaging studies and EDX examinations are both helpful in diagnosis. Plain radiographs may reveal bony abnormalities such as exostoses, displaced fractures, and degenerative arthritis, while weight-bearing radiographs can disclose some foot deformities. If a blood pressure cuff test is positive, venography usually demonstrates venous varicosities within the tarsal tunnel. Overall, MRI is the best imaging assessment available. It can detect lesions that are both within the tarsal tunnel and impinging upon it, such as tenosynovitis, neoplasms and other masses, varicosities, and hypertrophic scar tissue; it is considered to be extremely useful when surgery is contemplated.3,37,45,49 Nonetheless, a few investigators have found MRI studies to be of little benefit.65 Confirming the presence of TTS by EDX examination has rested mainly on one or more of several NCSs, including tibial motor, mixed plantar, and plantar sensory NCSs. Typically, prolonged latencies, either absolute (compared to laboratory normal values) or relative (compared to the responses obtained on the contralateral, asymptomatic limb), have been sought. Low-amplitude or absent responses occur, and may be abnormal—depending upon the particular one performed and the age of the patient— but they have little localizing value. Obtaining tibial motor CMAPs by stimulating the tibial nerve proximal to the tarsal tunnel while recording from the abductor hallucis, abductor digiti quinti pedis, or both muscles was the first procedure to be used. Unfortunately, while rather easily elicitable in nearly all patients, they are quite insensitive, yielding positive results in only about half of clinically suspected cases. Mixed plantar NCSs are currently the most widely used procedures. The intermixed motor and sensory axons in the medial plantar and lateral plantar nerves are stimulated percutaneously on the sole of the foot while mixed NAPs are recorded with surface electrodes over the tibial nerve, immediately proximal to the tarsal tunnel. Although these are more sensitive than tibial motor NCSs, reportedly being abnormal in approximately two thirds of

1505

Mononeuropathies of the Lower Limb

symptomatic cases, they are plagued by many technical problems, including being unelicitable because of foot calluses, ankle edema, and foot deformities, and in patients who are over the age of 45 years. Assessing solely the sensory fibers of the medial and lateral plantar nerves is the most sensitive method for confirming the presence of TTS. The interdigital nerves of the first and fifth toes are stimulated while recording from the tibial nerve proximal to the tarsal tunnel (orthodromic technique), or the stimulating and recording sites are reversed (antidromic technique). These plantar sensory NCSs have several limitations, including (1) the elicited responses invariably are of such low amplitude that they require signal averaging, even in healthy subjects; (2) if needle recording electrodes are employed, the test becomes invasive; and (3) they almost always are time consuming, particularly when performed bilaterally, as is frequently required. For these reasons, even though they are reported to be positive in more than 90% of symptomatic limbs, plantar sensory NCSs are not widely used for diagnosis. In many patients, the needle EMG is abnormal, disclosing fibrillation potentials, MUAP loss, or both in the distribution of the affected plantar nerve. Unfortunately, such changes have little localizing value.15,25,38,45 As with MRI, while some investigators consider EDX examinations to be mandatory for diagnosis of TTS,15,45 others have found them to be of little value65; several authors have emphasized that normal EDX studies do not exclude the diagnosis whenever the clinical presentation is characteristic.3,37,49 The differential diagnosis of TTS includes many entities, only some of which involve the peripheral nervous system. In our experience, S1 radiculopathies rather frequently are misdiagnosed as TTS. Indolent partial sciatic neuropathies and main trunk tibial neuropathies (such as those caused by compression by neoplasms or congenital anomalies), as well as ischemic monomelic neuropathy and reflex sympathetic dystrophy, may present solely with sensory symptoms in the foot, similar to TTS. A pure or predominantly pure sensory polyneuropathy can be erroneously diagnosed as bilateral TTS. Lesions of the medial plantar and lateral plantar nerves distal to the tarsal tunnel are readily mistaken for selective involvement of those nerves within the tunnel itself, because the clinical and EDX manifestations at both sites may be indistinguishable. A number of non-neurologic conditions are confused with TTS, including plantar fasciitis, tendinitis, arthritis, bursitis, and stress fractures.6,25,38,45,60,64

Treatment and Prognosis The appropriate therapy for TTS depends upon the cause. Nonsurgical treatments include local corticosteroid injections, nonsteroidal anti-inflammatory medications, foot orthoses, and alterations in shoe wear. Surgical removal is the preferred initial treatment whenever space-occupying

lesions are present. However, if these are lacking, operations—consisting of release of the flexor retinaculum— usually are considered last resorts, performed only after prolonged attempts (3 months or greater) at conservative therapy have failed.3,37,45,49 The percentage of patients who ultimately require surgery is debated. Operative treatment of the idiopathic variety may be more successful if, in addition to release of the flexor retinaculum, the neurovascular bundle is decompressed along its course, and a graft of fat is inserted between the nerves and blood vessels.33 Conservative therapy alone is highly effective in many patients, such as those with foot deformities, in whom orthoses or alterations in shoe wear can relieve nerve compression, and in those with tenosynovitis, which responds to corticosteroid injections or nonsteroidal anti-inflammatory medication. Surgery has proven most effective in patients with space-occupying lesions, those who have short histories, and those who do light work. Conversely, it has been less helpful in patients with tenosynovitis (both traumatic and nontraumatic) and flexible foot deformities, those who do heavy work or have workers’ compensation claims pending, and those who require repeat operations.2,6,25,37,45,49 The fact that several authorities have reported 85% to 90% surgical success rates, and that surgery has been described as “extremely effective” for this disorder,37 is difficult to reconcile with the experience of our own EDX laboratories, in which most of the patients referred for TTS evaluations have already undergone three to five unhelpful operations. On a yearly basis, we evaluate more patients for failed TTS surgery than we do for failed carpal tunnel syndrome surgery, although the latter operation is far more common.

DISTAL PLANTAR, DISTAL CALCANEAL, AND DIGITAL NEUROPATHIES Anatomy As the medial plantar nerve, which is larger than the lateral plantar branch, enters the foot, muscular branches arise from it to innervate the abductor hallucis and flexor digitorum brevis muscles, while its plantar cutaneous branch pierces the plantar aponeurosis to supply the skin of the medial two thirds of the sole of the foot. The medial plantar nerve terminates as the proper digital plantar nerve and three common digital nerves, which innervate the flexor hallucis brevis and first lumbrical muscles, and supply sensation to the medial three toes and the medial aspect of the fourth toe7 (see Fig. 61–6). After the lateral plantar nerve enters the foot, it tracks laterally in an oblique fashion, innervates the quadratus plantae and abductor digiti minimi muscles, supplies sensation to the lateral one third of the sole of the foot, and then divides into superficial and deep branches. These two branches innervate

1506

Diseases of the Peripheral Nervous System

those intrinsic foot muscles that are not innervated by the medial plantar nerve, and supply sensation to the fifth toe and lateral aspect of the fourth toe7 (see Fig. 61–6). The medial calcaneal nerve usually arises from the tibial nerve proximal to or within the tarsal tunnel, but its origin is highly variable. It perforates the flexor retinaculum to reach the heel, where it supplies sensation to the skin7,45 (see Fig. 61–6).

Lesions of Individual Nerves Plantar Neuropathies The same disorders that affect the plantar nerves within the tarsal tunnel can affect them when they are beyond it. These include space-occupying lesions, such as neoplasms; trauma, such as foot fractures, ankle sprains, and postsurgical scarring; foot deformities (both dynamic and fixed); and various muscle anomalies, particularly of the abductor hallucis muscle. Relatively few verified instances of medial plantar neuropathy in this area have been reported, and even fewer cases of lateral plantar neuropathy. Distal plantar nerve lesions typically present clinically with pain (often burning in character), paresthesias, or both. Many of the reported patients with medial plantar neuropathies have been runners who became symptomatic while running, causing the condition to be called “jogger’s foot”; a history of prior foot injury is common in these cases. Tenderness may be noted distal to the tarsal tunnel, at the abductor tunnel area, with both types of plantar neuropathy. Weakness of the intrinsic foot muscles innervated by the affected nerve, however, seldom is reported.38,45,48 These lesions are diagnosed by their clinical presentation, the demonstration of sensory loss in the distribution of the nerve, and, especially, by the presence of a Tinel’s sign over the abductor tunnel rather than the tarsal tunnel. Muscles innervated by the involved nerve may be weak and atrophic, but demonstrating this can be difficult. Conversely, hypertrophy of the abductor hallucis muscle has been reported as one cause of these lesions. Both foot radiographs and MRI studies may disclose abnormalities. The EDX examination frequently is abnormal. Posterior tibial motor, mixed, and sensory plantar NCSs may reveal absent or low-amplitude responses or, less often, prolonged latencies. Similarly, needle EMG may reveal fibrillation potentials and possibly MUAP loss in the affected muscles. Unfortunately, the identical EDX abnormalities can be found with more proximal involvement of these same nerve fibers within the tarsal tunnel, although such fascicular involvement is more common for the medial than the lateral plantar nerve.38,45,64 The optimal treatment for these neuropathies depends on the cause. Conservative therapy may suffice, particularly when athletes are affected. Rest, temporary avoidance

or modification of the suspected offending activity, and changes in shoe wear may relieve symptoms. Local injections of lidocaine or corticosteroids, orthotic shoe inserts, and nonsteroidal anti-inflammatory medications also may help. Surgery, consisting of either excision or neurolysis, generally is required for mass lesions and sometimes for those resulting from trauma.38,45,64 No definite conclusions can be drawn regarding prognosis with treated distal, medial, or lateral plantar neuropathies because so few have been reported in detail. Medial Calcaneal Neuropathies Localized damage to the medial calcaneal nerve after it exits the tarsal tunnel may produce heel paresthesias, pain, and sensory loss. It is unclear, however, how often lesions of this nerve actually are responsible for such heel symptoms. Reviews on the subject38,64 rely heavily on a single source: a 1984 article focused on 10 athletes who, while walking and running, had heel pain that was intense each time the affected foot struck the ground. Marked tenderness was present over the medial anterior portion of the heel pad, the anterior calcaneal branch of the tibial nerve, and the abductor tunnel. Radiographs of the os calcis were normal. Multiple conservative regimens, including rest, anti-inflammatory medications, corticosteroid injections, shoe adjustments, orthotic devices, ultrasound, and massage, were all ineffective. Conversely, surgical decompression of the anterior and posterior branches of the medial calcaneal nerve relieved symptoms in 10 of the 11 feet. Although the authors postulated that their patients’ chronic heel pain was due to entrapment of the anterior branch of the medial calcaneal nerve, no evidence of an entrapment point was found at operation.22 In another report concerned with these lesions, the most common causes were local inflammatory disorders, trauma, and vascular spasm. Heel spurs were frequently present, and usually could be treated with conservative measures, such as shoe pads, special shoes, or, if necessary, local corticosteroid injections.62 Interdigital Neuropathies The interdigital nerves can be injured by a variety of mechanisms, although the most common is trauma, either blunt or penetrating, such as lacerations of the sole of the foot. Two digital neuropathies have acquired special significance: Joplin’s neuroma and Morton’s neuroma. Joplin’s Neuroma. Joplin’s neuroma is the name given to the symptomatology resulting from compromise of one of the terminal branches of the medial plantar nerve: the medial plantar proper digital nerve, which supplies sensation to the medial aspect of the great toe. It is damaged either as it crosses the first metatarsophalangeal joint or slightly more distally, along the medial side of the toe. The most common cause is compression from poorly fitting shoe

1507

Mononeuropathies of the Lower Limb

wear; other etiologies include blunt trauma or lacerations and postoperative scarring following bunion surgery. Symptoms include pain and paresthesias along the medial side of the first toe on ambulation. Sensory loss and tenderness may be present along the course of the nerve. On clinical examination, the nerve may be palpably enlarged and tender immediately proximal to the interphalangeal joint, and usually a Tinel’s sign can be elicited on the medial aspect of the metatarsophalangeal joint. The treatment for Joplin’s neuroma includes changing shoe wear and protective padding. If surgery is required, the nerve is excised rather than decompressed, even though this produces a permanent sensory deficit.38,45,60 Morton’s Neuroma. Morton’s neuroma affects mainly adults, particularly women. It results from damage, between or near the heads of adjacent metatarsal bones, to an interdigital nerve immediately prior to where it divides into two proper digital nerves. The term initially was applied only to involvement of the interdigital nerve that supplies the web space between the third and fourth toes. However, more recently, similar lesions involving other interdigital nerves also have been labeled Morton’s neuroma. Two pathogenetic mechanisms have been described: compression of the symptomatic plantar interdigital nerve between the heads of adjacent metatarsal bones and physical distortion (compression or traction) of the nerve by the anterior edge of the deep transverse metatarsal ligament.25,38,45,60,64 Surgery frequently reveals a swelling in the affected interdigital nerve near the metatarsal heads. Although referred to as a neuroma, this actually is a fibrous nodule, so the name “Morton’s neuroma” is a misnomer.60 The symptoms typically are triggered by weight bearing, both standing and ambulation, particularly when wearing shoes. They consist of (1) localized pain and tenderness on the plantar aspect of the distal foot, between two metatarsal heads; and (2) pain or paresthesias that radiate distally from between the two metatarsal heads to the adjacent sides of the two affected toes. Typically, the symptoms are relieved by non–weight bearing and by shoe removal. Morton’s neuroma is diagnosed principally by the history and physical examination. The lesion site may be tender to palpation, which may also reproduce symptoms. Sensory loss often is present in the distribution of the affected interdigital nerve, in the web space between the adjacent sides of two toes. A nerve compression test (called a web space compression test), in which the involved nerve is squeezed in its web space, both dorsally and ventrally, using the thumb and index finger, reportedly causes severe pain in the appropriate distribution. Local anesthetic blocks have been used for diagnosis, but both false-positive and false-negative studies occur, so their value is debated. Although routine radiographs typically are normal, ultrasonography, CT, and MRI usually detect the neuroma.

Identifying a Morton’s neuroma may be difficult because the differential diagnosis includes several entities, such as metatarsal stress fractures, osteochondrosis of the metatarsal head, and metatarsophalangeal joint pathology of various types. Conservative treatment is used initially, including changes in shoe wear (such as wide shoes rather than narrow, and flat shoes rather than high heels); shoe modification to limit dorsiflexion of the metatarsophalangeal joint during walking; minimizing repetitive dorsiflexion of the toes (such as reducing stair climbing and squatting); nonsteroidal anti-inflammatory medications; and injections of corticosteroids or local anesthetics. If these measures fail, which they often do, then various surgical procedures can be performed, including excision of the fibrous nodule; sectioning of the deep transverse metatarsal ligament (the intermetatarsal ligament), sometimes combined with internal neurolysis of the nerve; and interdigital neuroectomy.25,38,45,60,64 Surgery reportedly relieves symptoms in approximately 80% of patients.64

SURAL NEUROPATHY The sural nerve arises from the tibial nerve in the popliteal fossa and passes distally between the two heads of the gastrocnemius to a location slightly below the middle of the back of the leg. At that point it pierces the deep fascia and is joined by the lateral sural cutaneous branch, which arises from the common peroneal nerve in the popliteal fossa. (Some authorities view the sural nerve as beginning only when the lateral sural cutaneous nerve unites with the branch that arises from the tibial nerve; they consider the latter to be the medial sural cutaneous nerve, and not the sural nerve itself.) On reaching the foot, the sural nerve passes posteroinferiorly around the lateral malleolus to supply the skin over the lateral aspect of the ankle and border of the sole. The sural nerve supplies sensation to a narrow vertical strip of skin on the posteromedial aspect of the leg (from the proximal calf distally), to the lateral aspect of the ankle, and to the lateral border of the sole7,21,60,61 (Fig. 61–7). The sural nerve can be damaged at any point along its course. The most common site is distally, near the ankle, because it is in this region that it is biopsied. This is the single most common cause for sural neuropathies. However, even when iatrogenic injuries are excluded, trauma is a common etiology: In a series of 46 cases, 31 (67%) were of this etiology.57 Other causes include arthroscopy and Baker’s cysts, in which the nerve is injured near the knee; varicose vein surgery and external compression (such as by the superior edge of a boot or prolonged pressure from a firm object), which damage the nerve in the calf region; and arthroscopic and open surgery, fractures, lacerations, and sprains of the ankle, as well as compression by cysts,

1508

Diseases of the Peripheral Nervous System

The treatment of sural neuropathies depends upon their cause. Those resulting from external compression typically respond to avoidance of pressure alone. Painful neuromas and space-occupying lesions usually are resected; Baker’s cysts are exceptions—most often they are treated conservatively. Although the discomfort resulting from sural nerve biopsy rarely requires operative treatment, it is present in enough instances that this diagnostic procedure should be performed only when the information obtained is likely to influence treatment.25,38,43,60

REFERENCES

FIGURE 61–7 The sural nerve. This pure sensory nerve arises principally from the tibial nerve in the popliteal fossa and extends distally to the lateral sole of the foot.

ganglia, bony anomalies, and posttraumatic fibrosis occurring near the ankle.12,25,38,60,64 In the foot itself, the inferior branch of the sural nerve can be injured by fractures of the base of the fifth metatarsal bone. The symptoms of a sural neuropathy include paresthesias, a noticeable sensory deficit and, less often, pain in the distribution of the nerve. With lacerations, the pain may be localized to the site of a neuroma.25,38,60 The clinical examination frequently can localize the lesion, although in most instances this is apparent from the history. Often the site of injury is tender to pressure or percussion (Tinel’s sign). With occasional lesions, especially space-occupying ones, neuroimaging studies may be helpful, particularly MRI studies. Sural SNAPs usually are low in amplitude or absent, thereby confirming the presence of nerve damage. However, they rarely assist in localization, unless the lesion is proximal and other NCSs are affected as well.25,38

1. Al Hakim, M., and Katirji, M. B.: Femoral mononeuropathy induced by the lithotomy position: a report of 5 cases and a review of the literature. Muscle Nerve 16:891, 1993. 2. Baba, H., Wada, M., Annen, S., et al.: The tarsal tunnel syndrome: evaluation of surgical results using multivariant analysis. Int. Orthop. 21:67, 1997. 3. Bailie, D. S., and Kelikian, A. S.: Tarsal tunnel syndrome: diagnosis, surgical treatment, and functional outcome. Foot Ankle Int. 19:65, 1998. 4. Birch, R., Bonney, G., and Wynn-Parry, C. B.: Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingstone, 1998. 5. Celebrezze, J. P. Jr., Pidala, M .J., Porter, J. A., and Slezak, F. A.: Femoral neuropathy: an infrequently reported postoperative complication. Report of four cases. Dis. Colon. Rectum 43:419, 2000. 6. Cimino, W. R.: Tarsal tunnel syndrome: review of the literature. Foot Ankle Int. 11:47, 1990. 7. Clemente, C. D. (ed).: Gray’s Anatomy, 30th American ed. Baltimore, Williams & Wilkins, 1985. 8. Dash, S., Bheemreddy, S. R., and Tiku, M. L.: Posterior tibial neuropathy from ruptured Baker’s cyst. Semin. Arthritis Rheum. 27:272, 1998. 9. Dawson, D. M., Hallett, M., and Wilbourn, A. J.: Miscellaneous uncommon syndromes of the lower extremity. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 369, 1999. 10. Dawson, D. M., Hallett, M., and Wilbourn, A. J.: Peroneal nerve entrapment. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 270, 1999. 11. Dawson, D. M., Hallett, M., and Wilbourn, A. J.: Sciatic nerve entrapment. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds.): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 264, 1999. 12. DeLee, J. C.: Complications of arthroscopy and arthroscopic surgery: results of a national survey. J. Arthroscopy 1:214, 1985. 13. Deutsch, A., Wyzykowski, R. J., and Victoroff, B. N.: Evaluation of the anatomy of the common peroneal nerve: defining nerve at-risk in arthroscopically assisted lateral meniscus repair. Am. J. Sports Med. 27:10, 1999. 14. Drees, C., Wilbourn, A. J., and Stevens, G. H. L.: Main trunk tibial neuropathies. Neurology 59:1082, 2002.

Mononeuropathies of the Lower Limb 15. Galardi, G., Amadio, S., Maderna, L., et al.: Electrophysiologic studies in tarsal tunnel syndrome. Am. J. Phys. Med. Rehabil. 73:193, 1994. 16. Garfin, S. R.: Anatomy of the extremity compartments. In Mubarak, S. J., and Hargens, A. R. (eds.): Compartment Syndrome and Volkmann’s Contracture. Philadelphia, W. B. Saunders, p. 17, 1981. 17. Gateless, D., and Gilroy, J.: Tight-jeans meralgia: hot or cold? JAMA 252:42, 1984. 18. Goh, K. J., Tan, C. B., and Tjia, H. T.: Sciatic neuropathies— a retrospective review of electrodiagnostic features in 29 patients. Ann. Acad. Med. Singapore 25:566, 1996. 19. Goldman, J. A., Feldberg, D., Dicker, D., et al.: Femoral neuropathy subsequent to abdominal hysterectomy: a comparative study. Eur. J. Obstet. Gynecol. Reprod. Biol. 20:385, 1985. 20. Hahn, L.: Clinical findings and results of operative treatment in ilioinguinal nerve entrapment syndrome. Br. J. Obstet. Gynaecol. 96:1080, 1989. 21. Haymaker, W., and Woodhall, B.: Peripheral Nerve Injuries, 2nd ed. Philadelphia, W. B. Saunders, 1953. 22. Henricson, A. S., and Westlin, N. E.: Chronic calcaneal pain in athletes: entrapment of the calcaneal nerve? Am. J. Sports Med. 12:152, 1984. 23. Jablecki, C. K.: Postoperative lateral femoral cutaneous neuropathy. Muscle Nerve 22:1129, 1999. 24. Katirji, B.: Peroneal neuropathy. Neurol. Clin. 17:567, 1999. 25. Katirji, B.: Compressive and entrapment neuropathies of the lower extremity. In Katirji, B., Kaminski, H. J., Preston, D. C., et al. (eds.): Neuromuscular Disorders in Clinical Practice. Boston, Butterworth Heinemann, p. 774, 2002. 26. Katirji, B.: Electrodiagnostic approach to the patient with suspected mononeuropathy of the lower extremity. Neurol. Clin. 20:479, 2002. 27. Katirji, B., Wilbourn, A. J., Scarberry, S., and Preston, D. C.: Intrapartum lumbosacral plexopathy. Muscle Nerve. 26:340, 2002. 28. Katirji, M. B., and Wilbourn, A. J.: Common peroneal mononeuropathy: a clinical and electrophysiologic study of 116 lesions. Neurology 38:1723, 1988. 29. Katirji, M. B., and Wilbourn, A. J.: High sciatic lesions mimicking peroneal neuropathy at the fibular head. J. Neurol. Sci. 121:172, 1994. 30. Kaufman, M. D.: Isolated superior gluteal neuropathy due to intramuscular injection. N. C. Med. J. 49:85, 1988. 31. Kline, D. G., and Hudson, A. R.: Nerve Injuries. Philadelphia, W. B. Saunders, 1995. 32. Knockaert, D. C., Boonen, A. L., Bruyninckx, F. L., and Bobbaers, H. J.: Electromyographic findings in ilioinguinaliliohypogastric nerve entrapment syndrome. Acta Clin. Belg. 51:156, 1996. 33. Kohno, M., Takahashi, H., Segawa, H., and Sano, K.: Neurovascular decompression for idiopathic tarsal tunnel syndrome: technical note. J. Neurol. Neurosurg. Psychiatry 69:87, 2000. 34. Kuntzer, T., vanMelle, G., and Regli, F.: Clinical and prognostic features in unilateral femoral neuropathies. Muscle Nerve 20:205, 1997. 35. Kvist-Poulsen, H., and Borel, J.: Iatrogenic femoral neuropathy subsequent to abdominal hysterectomy: incidence and prevention. Obstet. Gynecol. 60:516, 1982.

1509

36. Lagueny, A., Deliac, M. M., Deliac, P., and Durandeau, A.: Diagnostic and prognostic value of electrophysiologic tests in meralgia paresthetica. Muscle Nerve 14:51, 1991. 37. Lau, J. T. C., and Daniels, T. R.: Tarsal tunnel syndrome: a review of the literature. Foot Ankle Int. 20:201, 1999. 38. Levin, K. H.: Tibial neuropathies. In Brown, W. F., Bolton, C. F., and Aminoff, M. J. (eds.): Neuromuscular Function and Disease, Vol. 2. Philadelphia, W. B. Saunders, p. 965, 2002. 39. Lippitt, A. B.: Neuropathy of the saphenous nerve as a cause of knee pain. Bull. Hosp. Jt. Dis. 52:31, 1993. 40. Mastaglia, F. L.: Tibial nerve entrapment in the popliteal fossa. Muscle Nerve 23:1883, 2000. 41. McGeorge, D., Sturzenegger, M., and Buchler, U.: Tibial nerve mistakenly used as the tendon graft: reported 3 cases. J. Bone Joint Surg. Br. 74:365, 1992. 42. McManis, P. G.: Sciatic nerve lesions during cardiac surgery. Neurology 44:684, 1994. 43. Nakano, K. K.: Entrapment neuropathy from Baker’s cysts. JAMA 239:135, 1978. 44. Nunley, J. A., and Gabel, G. T.: Tibial nerve grafting for restoration of plantar sensation. Foot Ankle 14:489, 1993. 45. Oh, S. J., and Meyer, R. D.: Entrapment neuropathies of the tibial (posterior tibial) nerve. Neurol. Clin. 17:593, 1999. 46. Psathakis, D. N.: A new syndrome: the pseudo-entrapment of the popliteal artery and tibial nerve. Int. Angiography 10:250, 1991. 47. Purves, J. K., and Miller, J. D.: Inguinal neuralgia: a review of 50 patients. Can. J. Surg. 29:43, 1986. 48. Rask, M. R.: Medial plantar neurapraxia (jogger’s foot). Clin. Orthop. 134:193, 1978. 49. Reade, B. M., Longo, D. C., and Keller, M. C.: Tarsal tunnel syndrome. Clin. Podiatr. Med. Surg. 18:395, 2001. 50. Rogers, L. R., Borkowski, G. P., Albers, J. W., et al.: Obturator mononeuropathy caused by pelvic cancer: six cases. Neurology 43:1489, 1993. 51. Salazar-Grueso, E., and Roos, R.: Sciatic endometriosis: a treatable sensorimotor mononeuropathy. Neurology 36:1360, 1986. 52. Schmalzried, T. P., Amstutz, H. C., and Dorey, F. J.: Nerve palsy associated with total hip replacement: risk factors and prognosis. J. Bone Joint Surg. Am. 73:1074, 1991. 53. Schon, L. C., Glennon, T. P., and Baxter, D. E.: Heel pain syndrome: electrodiagnostic support for nerve entrapment. Foot Ankle 14:129, 1993. 54. Seddon, H.: Surgical Disorders of the Peripheral Nerve, 2nd ed. Edinburgh, Churchill Livingstone, 1975. 55. Shields, R. W. Jr., Root, K. E. Jr., and Wilbourn, A. J.: Compartment syndromes and compression neuropathies in coma. Neurology 36:1370, 1986. 56. Silver, J. K., and Leadbetter, W. B.: Piriformis syndrome: assessment of current practice and literature review. Orthopedics 21:1133, 1998. 57. Smith, B. E., and Litchy, W. J.: Sural mononeuropathy: a clinical and electrophysiological study. Neurology 39(Suppl. 1):296, 1989. 58. Sotaniemi, K.: Slimmer’s paralysis: peroneal neuropathy during weight reduction. J. Neurol. Neurosurg. Psychiatry 47:564, 1984. 59. Starling, J. R., Harms, B. A., Schroeder, M. E., and Eichman, P. L. Diagnosis and treatment of genitofemoral and ilioinguinal entrapment neuralgia. Surgery 102:581, 1987.

1510

Diseases of the Peripheral Nervous System

60. Stewart, J. D.: Focal Peripheral Neuropathies, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 431, 2000. 61. Sunderland, S.: Nerve and Nerve Injuries, 2nd ed. Edinburgh, Churchill Livingstone, 1978. 62. Tanz, S. S.: Heel pain. Clin. Orthop. 28:169, 1963. 63. Tennent, T. D., Birch, N. C., Holmes, M. J., et al.: Knee pain and the infrapatellar branch of the saphenous nerve. J. R. Soc. Med. 91:573, 1998. 64. Trepman, E.: Entrapment neuropathies of the foot and ankle. In Dawson, D. M., Hallett, M., and Wilbourn, A. J. (eds.): Entrapment Neuropathies, 3rd ed. Philadelphia, Lippincott–Raven, p. 297, 1999. 65. Turan, I., Rivero-Melian, C., Guntner, P., and Rolf, C.: Tarsal tunnel syndrome: outcome of surgery in long-standing cases. Clin. Orthop. 343:151, 1997. 66. Vasilev, S. A.: Obturator nerve injury: a review of management options. Gynecol. Oncol. 53:152, 1994. 67. Warner, M. A., Warner, D. O., Harper, C. M., et al.: Lower extremity neuropathies associated with lithotomy positions. Anesthesiology 93:938, 2000.

68. Weber, E. R., Daube, J. R., and Coventry, M. B.: Peripheral neuropathies associated with total hip arthroplasty. J. Bone Joint Surg. Am. 58:66, 1976. 69. Wilbourn, A. J.: Common peroneal mononeuropathy at the fibular head. Muscle Nerve 9:825, 1986. 70. Wilbourn, A. J.: Nerve conduction studies: types, components, abnormalities and value in localization. Neurol. Clin. 80:305, 2002. 71. Wilbourn, A. J., and Gilliatt, R. W.: Double crush syndrome: a critical reappraisal. Neurology 49:21, 1997. 72. Williams, P. H., and Trzil, K. P.: Management of meralgia paresthetica. J. Neurosurg. 74:76, 1991. 73. Wyant, G. M.: Chronic pain syndromes and their treatment. III. The piriformis syndrome. Can. Anaesth. Soc. J. 26:305, 1979. 74. Yuen, E. C., Olney, R. K., and So, Y. T.: Sciatic neuropathy: clinical and prognostic features in 73 patients. Neurology 44:1669, 1994. 75. Yuen, E. C., So, Y. T., and Olney, R. K.: The electrophysiologic features of sciatic neuropathy in 100 patients. Muscle Nerve 18:414, 1995.

62 Operating on Peripheral Nerves ROLFE BIRCH

Overview Clinical Diagnosis Pain Causalgia Neurostenalgia Posttraumatic Neuralgia Central Pain in Brachial Plexus Injury Principles of Operating upon the Nerves Alternative Methods of Repair Iatropathic Nerve Injury Injection Injury Interscalene Block

Hemorrhage and Ischemia Compression by Hematomas False Aneurysm and Arteriovenous Fistula Damage in the Field of Operation Damage to Nerves during Replacement Arthroplasty of the Hip Damage to Small Nerves Injuries of the Brachial Plexus Level of Lesion Diagnosis Physiologic Investigation Radiologic Investigation

OVERVIEW While the decision to operate on a nerve is straightforward in the acute case of an open wound or where the nerve injury is associated with damage to long bones, joints, and blood vessels, in other situations it is not always as easy. The primary reasons for operating include 1. To confirm or establish diagnosis 2. To restore continuity to a severed or ruptured nerve 3. To relieve a nerve of an agent that is compressing, distorting, or occupying it The indications for operating upon nerves after an injury include 1. Severe paralysis after a wound over the course of a major nerve 2. Severe paralysis after operation in the field of a major nerve or injection close to the course of that nerve 3. Severe paralysis after a closed injury, especially highenergy injuries, with severe damage to soft tissues and skeleton 4. Severe paralysis after closed traction injury of the brachial plexus

Repair of the Brachial Plexus The Upper Lesion: Rupture or Avulsion of C5/C6 (C7) with Intact (C7), C8/T1 The Complete Lesion Restoration of Hand Function Reconnection between the Spinal Cord and Avulsed Nerves Obstetric Brachial Plexus Palsy Incidence and Risk Factors Course and Prognosis Special Investigations and Indications for Operation

5. A nerve lesion associated with arterial injury 6. A nerve lesion associated with fracture or dislocation requiring urgent open reduction and internal fixation 7. Increasing severity of the nerve lesion while under observation 8. Failure to progress toward recovery in the expected time after a closed injury thought to be axonotmetic 9. Failure to recover from conduction block within 6 weeks of injury 10. Persistent neuropathic pain The aims of operation are to preserve or restore function by preserving or restoring the innervation of the skin, muscle, and other tissues. When a nerve is severed, repair offers the only chance of doing this. That repair may be done by suture or by graft. If the proximal stump is hopelessly damaged, or if there is no continuity between the proximal stump and the spinal cord, then transfer of other nerves to the distal stump may be possible. If the distal stump is hopelessly damaged, then muscular neurotization (the implantation of nerves into the muscle) may be possible. Reconnection between the cord and the peripheral nerve may now be a realistic possibility in selected cases of injury to the brachial or lumbosacral plexus. When the 1511

1512

Diseases of the Peripheral Nervous System

neurologic injury is, by whatever means, irreparable, then palliation may be achieved by musculotendinous transfer or other “reconstruction.” The response of the axon to injury is the critical issue for the clinician. Seddon’s142 distinction between conduction block (neurapraxia) and two types of degenerative lesion (axonotmesis and neurotmesis) remains valuable. Sunderland154 introduced a more elaborate classification of nerve injury recognizing five degrees of severity ranging from simple conduction block to loss of all continuity. This system is useful in analyzing patterns of longitudinal traction injury, when the sheath of the nerve may be intact but there has been extensive intraneural disruption. Both of these systems are often misunderstood and misused. It is common to find deep degenerative lesions labelled “neuropraxia.” The distinction between axonotmesis and neurotmesis or between Sunderland’s grade 3, 4, and 5 injuries can be made only by the passage of time or by inspection of the nerve. It is more useful for the clinician to think of the nerve lesion as degenerative or nondegenerative. Thomas and Holdorff164 elaborated this concept in a most helpful manner (Table 62–1). “Axonal constriction” is the most likely lesion in “hourglass” constriction of fascicles within the anterior interosseous nerve. Recovery after interfascicular neurolysis was good.116 In a case of compression of the common peroneal division of the sciatic nerve at the greater sciatic notch of 3 years’ duration, with complete loss of muscular function and severe pain, pain was relieved and motor function was restored by 48 hours after decompression of the nerve, which was found to be flattened to a translucent band by callus.13 The explanation for such cases of prolonged conduction block is provided by Fowler et al.,53 Harrison,70 and Gilliatt.63 A wholly unfavorable degenerative lesion is seen only when a trunk nerve has been transected or ruptured. Many injuries to trunk nerves provoke lesions of populations of nerve fibers of differing severity. A particularly common example of this is sciatic or Table 62–1. Classification of Focal Mechanical Nerve Injury Focal Conduction Block Transient Ischemic Other More persistent Demyelinating Axonal constriction Axonal Degeneration With preservation of basal laminal sheaths of nerve fibers With partial section of nerve With complete transection of nerve Data from Thomas, P. K., and Holdorff, B.: Neuropathy due to physical agents. In Dyck, P. J., and Thomas, P. K. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 990, 1991.

femoral neuropathy incurred during arthroplasty of the hip. Some fibers remain intact; in others, there are degrees of conduction block. Spontaneous recovery is usual in some nerve fibers undergoing wallerian degeneration; in others there is no recovery at all. If the cause of the nerve lesion is not removed, then that lesion will progress. Inadvertent crushing of the nerve under a plate used to stabilize a fracture of a long bone or inadvertent encircling of nerve by suture material is not uncommon. The alert surgeon who removes that agent within an hour can anticipate recovery. Failure to remove that agent causes progression from conduction block to a degenerative lesion of favorable prognosis (axonotmesis) and then to one of unfavorable prognosis (neurotmesis). Nerves trapped within fractures or joints, distorted by expanding hematoma, or compressed within a swollen ischemic limb follow the same progression. Intradural injury of nerves forming the brachial plexus and lumbosacral plexus is a special case. Axons whose cells of origin lie in the posterior root ganglion remain intact and functional in such intradural injury. The separation between the central and the peripheral nervous systems is preganglionic.

CLINICAL DIAGNOSIS Examination of the patient soon after wounding is often difficult; the general condition may be affected by the loss of blood and by other injuries, the examination very often is done in unfavorable surroundings in an emergency department, and the patient may be a child or an adult, and, if an adult, may be affected by drink or drugs or both. “The clinician should, at all times, bear in mind that if there is a wound over the line of a main nerve and if there is any suggestion of loss of sensibility or impairment of motor function in the distribution of that nerve, it must be regarded as having been cut until and unless it is proved otherwise.”21 This seems to be particularly true for nerve injuries incurred during operation or after injection. In closed lesions, understanding of the extent and application of force is critical. Serious injuries cause serious nerve lesions. Widely displaced fractures of long bones will more likely rupture adjacent nerves. Rupture of an axial artery will almost inevitably relate to rupture of adjacent nerve trunks. It seems to be widely assumed that the prognosis for nerve injured by fractures or dislocations is good, but this is not generally the case. Seddon143 found that 146 of 212 nerve lesions produced by fractures or dislocations of the upper limb spontaneously recovered. Less than one half of 57 cases of nerve palsies after skeletal injury in the lower limb recovered. Seigel and Gelberman146 found that 85% of nerve palsies associated with a closed fracture recovered; 65% to 70% of nerves injured in open fractures did so. Omer129 recorded that no less than 83% of nerve

1513

Operating on Peripheral Nerves

palsies from closed upper limb fractures recovered. One common observation in these series was that those nerves that were going to go on to recover showed clear evidence of doing so at no later than 3 months from time of injury. These were not wholly degenerative lesions. The early signs of nerve injury include disturbance of sensation, weakness or paralysis of muscles, vasomotor and sudomotor paralysis in the distribution of the affected nerve or nerves, and abnormal sensitivity of that nerve at the point of injury. One very important sign evident within 24 hours of deep injury to a nerve with a cutaneous sensory component is sympathetic paralysis: The skin is warm and dry. The hypersensitivity of the nerve at the level of lesion, recognized by Tinel165 and Hoffman,73 tells us a good deal more. Tinel distinguished between the sensitivity of the nerve to compression and the pain of that nerve in “neuralgic irritation,” noting that “Formication [the sensation induced by eliciting Tinel’s sign] in the nerve is a very important sign, for it indicates the presence of young axis-cylinders in place of regeneration. This formication is quite distinct from the pain on pressure which exists in nerve irritation.”165 A strong Tinel’s sign is regularly found on the day of injury in cases of rupture or transection of a nerve. A strong Tinel’s sign is not consistent with a diagnosis of conduction block. Diagnosis of neurapraxia (conduction block) is unwise in the presence of severe neuropathic pain because that pain indicates there is a continuing and progressing lesion of the nerve.

PAIN Injury to peripheral nerves is not always accompanied or followed by pain; most cases of complete division are painless. Conversely, incomplete lesions are, from time to time, followed or accompanied by pain of varying duration and severity. Some neuropathic pain states are exceptionally severe. They are characterized by persistence, intractability, and disproportion between the extent of lesion and the severity of the pain. There is associated sensory or motor disturbance; often there is sympathetic dysfunction. The positive sensory symptoms described by patients are sometimes so clear that diagnosis of the cause and some understanding of the subsequent pathophysiologic events are possible from the history alone. The vogue for measuring pain by visual analogue scales too often gives a spurious sense of objectivity. Explanations of neuropathic pain that assume structural changes within the central nervous system (CNS) do not account for the instantaneous onset of pain in most cases of causalgia or for the central pain seen in many cases of brachial plexus lesions. Birch and Bonney10 proposed a classification of pain states that is helpful in clinical practice: 1. Causalgia. The term complex regional pain syndrome type 2 obscures diagnosis of this rare but important

2. 3.

4.

5.

6.

disorder for which the core of treatment is operation for rectification of the prime cause: the injured nerve and adjacent vessels. “Reflex sympathetic dystrophy” (complex regional pain syndrome type 1). Posttraumatic neuralgia, which can be divided into that following major injury to a major nerve trunk and that following injury to a nerve of cutaneous sensation. Neurostenalgia, which is pain from persistent compression, distortion, or ischemia of an intact nerve trunk. Central pain from preganglionic injuries to the brachial or lumbosacral plexus. This is divided into that pain following preganglionic rupture of spinal nerves peripheral to the transitional zone, causing deafferentation, and that following avulsion of spinal nerves from the cord itself. Pain maintained either deliberately or subconsciously by the patient.

Operation is contraindicated in “reflex sympathetic dystrophy” and in pain that is maintained by the wish to obtain compensation or by resentment against public bodies. Operation is the treatment of choice in causalgia and in neurostenalgia. Repair of a damaged nerve eases pain in some 50% of cases of posttraumatic neuralgia. Reinnervation of the limb eases central pain after brachial plexus lesions in an even higher percentage of cases.

Causalgia Nathan126 described causalgia as follows: “there is hyperpathia in which all cutaneous stimuli cause the same sensation, penetrating fiery pain which spreads from the point stimulated, and allodynia in which stimuli that elsewhere on the skin of the patient and in normal subjects do not cause pain, and a deep hypersensitivity, in which squeezing or pressing on the deep tissues cause pain.” The demonstration that axonotomized muscle afferents generate ongoing activity within the dorsal root ganglion (DRG) provides an explanation for the sensitivity to stimulation of deep tissues in causalgia.112 In two reviews,10,151 31 of 36 cases of causalgia were caused by penetrating missile injuries. False aneurysm or arteriovenous fistula was seen in 13. There was partial transection of the lower roots of the brachial plexus, or of the sciatic nerve and its derivative branches, in all of these cases. Onset of pain within 24 hours of injury was usual. Physical examination of the limb was impossible because of allodynia; there was disturbance of sympathetic function with excess sweating, vasomotor disturbance, and trophic changes in the skin. Pain was increased in response to emotional or other physical stimuli. Almost unique among pain states, distraction or activity worsens rather than relieves causalgia. Formal sympathectomy

1514

Diseases of the Peripheral Nervous System

was done in 4 patients, and sympathetic blockade was used in 10 others. In the remainder, pain was abolished by correction of the vascular lesion and by repair of the damaged nerve. It could be said that these patients did not have causalgia, but rather that the practice of infusing local anesthetic about the nerves proximal to the lesion may have induced an element of sympathetic blockade (Fig. 62–1).

Neurostenalgia Chronic compression, distortion, or ischemia of the nerve produces pain. In most cases the nerve trunk is intact, and the lesion is one of conduction block or at worst axonotmesis. The nerve is in some way irritated, tethered, compressed, or ischemic. Treatment of the cause not only relieves the pain but preserves or restores the function of the afflicted nerve. Neurostenalgia is the most rewarding of all neuropathic pain states in its response to operation and removal of the cause. The instantaneous relief of pain after decompression of a nerve stretched over an expanding hematoma, or after removal of a foreign body or a spike of bone distorting or impaling the nerve, is dramatic (Fig. 62–2).

Posttraumatic Neuralgia This is pain after nerve injury that is not sympathetically maintained. Treatment remains as difficult as when Seddon143 commented: “we come to an aspect of peripheral nerve injuries that is profoundly unsatisfactory: the painful syndrome lacking distinctive features of causalgia.” The only method of treatment with any hope of success is the restoration of a normal input of nonpainful impulses; this usually involves successful repair of the affected nerve.

Central Pain in Brachial Plexus Injury

FIGURE 62–1 Causalgia resulting from a handgun bullet injury to the brachial bundle, causing partial transection of the median nerve and a massive arteriovenous fistula (AVF). Pain responded to a sympathetic block and was abolished by correction of the AVF with repair of the damaged nerve.

“The pain is continuous, it does not stop a minute either day or night. It is either burning or compressing . . . In addition there is, every few minutes, a jerking sensation similar to that obtained by touching . . . a Leyden jar. It is like a zigzag made in the sky by a stroke of lightening. The upper part of the arm is mostly free of pain; the lower part from a little above the elbow to the tips of the fingers, never.”55 This expression of two patterns of pain in brachial plexus injury is characteristic. Taggart155 followed 116 patients for 12 years who had root avulsions proven at operation. In one third the whole plexus was avulsed; in the remainder there was a mixture of pre- and postganglionic injury. All patients experienced pain. In 88% this was severe at some time during their course. In 62% of patients pain began within 24 hours of injury. The mean time to onset was 13 days (range 0 to 180 days). Pain was at its most intense at, on average, 6 months from injury. Paroxysmal pain was never experienced in the absence of constant pain, and the intensity of the pain was closely related to the extent of deafferentation of the spinal cord. It was exceptionally severe in cases of avulsion of C4 through T1. No close relation between the depth of pain and the extent of associated multiple injuries was evident. A measurable decrease in pain was noted at a mean time of 6 months after operation and repair. In 34% of cases, pain remained moderate or severe at 3 years or more from that operation.

Operating on Peripheral Nerves

1515

FIGURE 62–2 Neurostenalgia caused by a coarse suture passed through the lateral cord of the brachial plexus during operation for stabilization of the shoulder. The pain, which was severe, was abolished by removal of that suture 10 weeks later. The nerve trunks recovered as axonotmesis.

Pain in brachial plexus injury is worse in cold weather, if the patient is otherwise unwell, or when the patient is stressed or depressed. Employment of the mind in work, hobbies, or conversation almost always eases pain. The nature and distribution of the pain are diagnostic. Constant crushing or burning pain on the day of injury or soon after indicates separation of roots from the cord; convulsive shooting pain radiating into an individual dermatome indicates that that nerve has been avulsed. Bonney20 suggested that pain arose from the damaged substantia gelatinosa. Damage to the medial part of Lissaeur’s tract, with gliosis within the substantia gelatinosa, has been shown after experimental avulsion in the spinal cord of cats.130 Spontaneous firing occurred in laminae 1 and 5 of the posterior horn after experimental deafferentation.5,97 This activity began a few days after lesion and reached a peak at 1 month. Spontaneous firing was expressed as a continuous barrage with superimposed bursts of highfrequency firing. There is increasing evidence that reinnervation of the limb by graft or by nerve transfer is useful in relieving this pain. Pain from postganglionic rupture was eased in 90% of cases.119 Of patients with avulsions, 17% continued with disturbing pain at 3 years or more from neurotization compared with 25% among those who had not been operated.18 In a prospective study of 324 patients following operations for supraclavicular injury of the brachial plexus,156 pain was severe on the day of injury in one half of cases, and 80% of those who went on to develop severe pain had done so by 14 days after injury. Many of these patients related relief of their pain to operations upon the nerves, and a number stated that pain relief coincided with the recovery of muscular function. A small number of patients were convinced that the relief of their pain was

related to transfer of intercostal nerves to trunk nerves within the limb even when there had been no recovery of muscular function. This observation led to a study of 19 patients in whom intercostal nerve transfer was performed more than 1 year after injury with the sole purpose of pain relief.9 All of these patients had avulsions of at least two spinal nerves. Sixteen of 19 patients reported significant relief of pain at an average of 8 months after intercostal transfer. The possible mechanisms for this surprising result were outlined as follows: that there was restoration of afferent input from the damaged limb; that pain relief was produced by the sectioning of some functioning afferent axons within the repaired nerve; and that pain relief may have been nonspecific, occurring in response to the anesthetic or to relief of postoperative pain with the use of analgesics, or based on suggestion alone. A later study8 of 14 patients found a strong correlation and temporal relation between reduction in pain and successful nerve repair. All five patients with motor recovery gained significant relief of deafferentation pain. Of the seven patients with persisting pain, none had motor recovery. There was no correlation between pain relief and the minimal recovery of sensation, and no patient had any return of sensory or sympathetic cutaneous axon reflexes. The easement of pain coincided with clinical evidence of muscular activity or, in some cases, preceded it by a matter of weeks.

PRINCIPLES OF OPERATING UPON THE NERVES Nerve repair requires the coaptation of healthy nerve stumps, without tension and with respect for longitudinal blood supply. Accurate orientation of the fascicles or

1516

Diseases of the Peripheral Nervous System

groups of fascicles is essential. Operations are aided by magnification, and by the use of fine sutures and instruments. Fibrin clot glue,122,173 now available as a commercial product, eases nerve repair and grafting. The sheath of particular surgical interest is the internal epineurium condensed around the perineurium. In early “primary” repair of clean transections, the perineurium itself may be the connective tissue sheath of choice for the placing of sutures. Venous flow was blocked when a nerve was stretched by 8% of its resting length; stretching by 15% produced ischemia.100 The experimental work of Clarke et al.37 clearly showed the ill effect of tension upon the repaired nerves. It is better to bridge a gap with the graft than to force direct suture. It is a mistake to resect too little damaged nerve in the hope of bringing about direct suture. Better results are achieved by resecting back to a healthy nerve face in which a recognizable pattern of fascicles can be seen pouting out from within the sheath. The results of operations upon nerves depend to an extent hardly matched in any other branch of surgery on the skill of the surgeon and the quality of technique. It is a matter of regret that the lessons learned from the establishment of regional centers in World War II110 have not been absorbed. Operation itself is but the first step in the process of rehabilitation. Treatment must also be directed to the relief of pain, to restoration of function by physical methods or orthosis or by musculotendinous transfer, and above all to the reintegration of the patient into work, study, and daily life. This is, at times, a lengthy process, and it is best done through a multidisciplinary specialized unit itself working within a hospital where facilities are adequate for the emergency or urgent diagnosis and treatment of associated injuries. The regular use of electrophysiologic studies during operations brings great advantages. These studies can demonstrate neural continuity across a lesion in continuity, can demonstrate the site of a conduction block, and can demonstrate which part of the nerve has suffered axonal interruption; they also can confirm that apparently intact components of the brachial plexus have intact central connections. Intraoperative electrodiagnostic assessment is particularly valuable in analysis of the lesion in continuity. The decision as to whether to resect such a lesion and bridge the gap by graft or to leave it alone can be particularly difficult. Most helpful of all is to stimulate above the lesion and to record from the nerve or from individual fascicles below it; a response of good amplitude may well indicate a good prognosis. Response in individual fascicles may allow separation of the intact part of the nerve from the damaged portion. Kline and Hudson88 analyzed compound nerve action potential (CNAP) studies across the site of lesion in nearly 1000 nerve injuries. Of 438 nerves where a CNAP could be recorded, 92% went on to useful recovery after neurolysis.

ALTERNATIVE METHODS OF REPAIR For many years surgeons have endeavored to improve the results of grafting or even to replace nerve by alternative materials. Some methods, grounded in considerable experimental work, have progressed as far as careful clinical studies. Experimental and clinical work on the use of freeze-thawed muscle grafts led to their use in the treatment of painful neuromas.56,94,163 The possibilities and limitations of axonal regeneration through acellular muscle graft were reviewed by Hall.68 Polyglycolic acid synthetic tubes were found useful in the repair of digital nerves in the hand.102 Placing nerve stumps within a silicone tube provides a chamber separating neural elements from the surrounding tissues so that there is a local accumulation of neurotrophic factors. Longitudinal orientation of fibrin matrix occurs within the tube, and regenerating axons are guided into the distal Schwann tubes across the gap.50,98 A trial of intubation versus suture of median and ulnar nerves in 18 patients showed no significant difference.99

IATROPATHIC NERVE INJURY Iatropathic injury of peripheral nerves is a persistent and probably growing problem. The incidence of claims alleging medical negligence and the cost of those claims seems to be rising inexorably within the United Kingdom. Harris69 noted that “in 1978 settled case of malpractice in the UK amounted to 1 in 1,000 and in 1988 it was 13 in 1,000.” Harris went on to point out that, “although figures are not available for the current position—mainly due to the fact that these cases have been dealt with by Health Authorities under the Crown Indemnity Scheme since 1990—the incidence is undoubtedly much higher.” Nerve injuries accounted for 92 of a total of 424 claims arising from a 10-year survey of incidents in vascular surgery.31 Overall peripheral nerve injuries probably come second to injuries of the CNS, both in cost and in number of claims for malpractice; if allegations for negligence relating to obstetric brachial plexus palsy (OBPP) are included, then both the number and the costs are higher. Two studies conducted over 7-year periods revealed a disturbing trend. In the first,11 68 major nerve injuries requiring repair were counted. In 2001,83 the same unit reported on a total of 306 nerve lesions in 291 patients. The delay between incident and diagnosis rose from 6 to 10 months. In the later report, 34 patients had been referred by attorneys. Legal proceedings were already underway for 110 others. Iatropathic transection of major nerves account for about 10% of our practice; if OBPP, postirradiation neuropathy, and nerves injured during arthroplasty of the hip and knee are included, that proportion exceeds 30%.

1517

Operating on Peripheral Nerves

Nerves are injured by doctors in many ways: By the use of drugs causing neuropathy By the use of anticoagulants, provoking intra- or extraneural hemorrhage By injection into or near a nerve By extraneural hemorrhage resulting from accidental or deliberate arterial puncture By unnoted ischemia of the limb By pressure or traction in the anesthetized patient By traction during manipulation under anesthesia or during delivery By direct damage from pressure, traction, heat, crushing, or cutting in the field of operation By ionizing radiation Operation is strongly indicated in nerve palsies caused by injection injury, hemorrhage and ischemia, and direct damage in the field of operation, and in some severe OBPP cases.

Injection Injury Severe local and radiating pain indicates that a drug has been injected into a nerve, although motor and sensory paralysis may be detectable only after some hours. It is the injected agent rather than the needle that causes the damage.75 Intrafascicular injury produces nerve lesion; extrafascicular injection causes lesser damage.57 Mackinnon and Dellon101 thought that operation should be done only when there is no evidence of recovery at the expected time. Kline and Hudson87 advised early decompression and irrigation. In later cases, their recommended policy is operation at between 2 and 4 months in the absence of recovery, with analysis of lesion by CNAP. Useful results from resection and repair of the damaged segment are confirmed. Early operation is probably indicated when the injected substance is known to be particularly damaging, when there is instant severe local and radiating pain, and when there is complete nerve palsy within 3 hours. Exposure of the nerve within 6 hours of injection secures decompression, removal of the noxious agent by irrigation, and relief of pain by infusion of local anesthetic.

Interscalene Block Review of large series of regional anesthetic blocks showed that major complications were rare.52 Eleven serious injuries to the spinal cord from the practice of interscalene block were seen in patients undergoing operations at the shoulder. In two cases, flow through the vertebral artery was compromised, causing bulbar and cranial nerve palsies that ultimately recovered. In all 11 cases there was ischemia of the anterior and lateral part of the lower cervical spinal cord, possibly from tamponade of the radicular vessels of

C7, C8, and T1. This left a permanent legacy of profound bilateral lower motor neuron lesion of segments C8 and T1 and of the ipsilateral spinothalamic tract. In three patients there was intractable central pain. The authors advised that interscalene block be abandoned in shoulder surgery.159 Similar cases have been reported.27,125

Hemorrhage and Ischemia Some of the most severe injuries to peripheral nerves and their target tissues followed delayed recognition of ischemia from interruption of flow through an axial vessel, increased pressure within a fascial compartment, intraarterial injection of a noxious agent, and cases of “tourniquet paralysis.” It is difficult to separate the effects of pure ischemia from those associated with compression, traction, and distortion. Some patterns can be distinguished. Ischemia of the Nerve Seddon143 described infarction of the median nerve in the hand in a severe case of Volkmann’s contracture affecting the forearm. Ischemia of the sciatic nerve provoking claudication has been relieved by reconstruction of the internal iliac artery.91 When flow through a main artery is interrupted by laceration or thrombosis, there is, in the absence of adequate collateral circulation, distal ischemia (Fig. 62–3). The use of a tourniquet on a limb in which a vascular prosthesis has been used to maintain flow is absolutely contraindicated.104 The prosthesis may be crushed and will thrombose. When there are inadequate collateral vessels, amputation may result. The effect of ischemia is most severe for muscle and least so for skin. Too often the presence of circulation in the skin is used to justify a complacent attitude. Capillary and venous filling are inadequate measures of blood supply to deeper tissues.66 Spasm is never a cause of obstruction of flow through a main artery in an adult, although it does occur in children.6 Ischemia of Nerves Compressed by Swollen Muscles Parkes132 described venous obstruction as being rather like strangulation of a hernia, leading to irreversible changes in the nerve at between 12 and 24 hours. Increased permeability of cell membranes from hypoxia leads to increase of pressure within compartments by transudation.6 Matsen107,108 described this as compartment syndrome. There is, first, occlusion of the low-pressure lymphatic venous system. Hypoxia of cell membranes leads to increasing fluid in the extravascular compartment. The “vicious cycle” is completed by obstruction of the perfusing arterioles when pressure within the compartment exceeds their critical closing pressure.29,44 The effect of this obstruction on the blood supply of the nerves is not altogether clear. Even in some advanced cases of ischemic contracture of forearm flexor muscles,

1518

Diseases of the Peripheral Nervous System

Compression by Hematomas Hematoma is a common complication in total hip replacement (THR) arthroplasty when anticoagulants are used to diminish deep vein thrombosis (DVT). The onset is sudden and delayed. Fourteen cases have been seen. In one case, a 53-year-old man underwent THR during which low-molecular-weight heparin was used as a prophylactic agent against DVT. Twenty-four hours after operation, there was sudden and severe burning pain in the foot followed by progressive sciatic palsy over the next 4 hours. The nerve was reexplored within 24 hours of onset of symptoms. An external hematoma with some internal intraneural bleeding, most severe in the common peroneal division, was evacuated. Pain was abolished, and the tibial division recovered fully over the next 18 months. The common peroneal nerve did not. In three other cases, evacuation of hematoma within 48 hours of onset of symptoms relieved pain and was followed by recovery of the sciatic nerve in two cases. Sedel145 thought that operation should be done within 6 hours at most. Decompression may still be helpful up to 48 hours.

False Aneurysm and Arteriovenous Fistula

FIGURE 62–3 Ischemia occurring after the brachial artery was ruptured in this supracondylar fracture of humerus in a 23-year-old woman and was not repaired. Necrosis of the flexor forearm muscles was complete. Later recovery for the ulnar nerve was good, and that for the median nerve useful. The median nerve was seen to be sustained by intrinsic circulation over a length of 26 cm.

ulnar nerve function is preserved or it recovers after decompression. Late decompression of median or tibial nerves compressed for many months by fibrosed muscle usually relieves pain and is followed by some recovery of the nerve. In one case of gas gangrene involving the muscles of the thigh, the posterior compartment was excised, leaving only the sciatic nerve devoid of all deep soft tissue connections. There was, later, useful recovery for both divisions. Holden74 emphasized that clinical suspicion is most important: absence of a pulse, pain, and tense and tender swelling of the compartment all indicate a need for intervention, which should be done before neural deficit occurs. Sympathetic block is condemned. There is a role for monitoring intracompartmental pressure in the leg. Decompression is advised when the pressure threshold, the difference between diastolic and the compartmental pressure, falls below 30 mm Hg, which is less than the critical closing pressure of muscle arterioles.109 Cases of iatropathic intracompartmental syndromes have been described.71,105

Cases of false aneurysm (expanding hematoma) and arteriovenous fistula following surgical accidents, fracture-dislocation, and penetrating missile wounds provide insight into effects upon the nerve of hemorrhage and ischemia. In 14 cases of false aneurysm of the axillary artery or of its offset in older patients caused by fracture-dislocation of the shoulder, the diagnosis of the vascular incident was made only after delayed onset of nerve palsy. The nerve palsies, usually painless, developed at between 24 hours and 3 months from injury and rapidly progressed over the next few days. Recovery of nerve function was usual save in cases in which the nerve trunk had become embedded within the wall of the sac. The course of recovery implied a strong element of conduction block, especially for those nerves remote from the expanding hematoma. Other nerves recovered as axonotmesis (Fig. 62–4). True causalgia is seen after partial section of a major nerve trunk associated with false aneurysm or arteriovenous fistula. In these cases, repair of the vascular lesion and decompression and repair of the nerve, with prolonged infusion of local anesthetic about the plexus and trunk and the stellate ganglion, bring early and lasting relief.85,151

Damage in the Field of Operation Damage to peripheral nerves during operations is the most common cause of iatropathic injuries. The rising incidence must be a cause of real concern for those responsible for the training of surgeons. Common factors include lack of awareness of anatomic relations, a reflection of the trend

Operating on Peripheral Nerves

FIGURE 62–4 Expanding hematoma following avulsion of the circumflex humeral vessels from the axillary artery in a fracture-dislocation in a 63-year-old man, resulting in massive hematoma formation over the course of 12 weeks. There was complete paralysis of all major trunks. Corrective operation was followed by recovery of those nerve trunks as axonotmesis.

toward abolition of teaching anatomy in medical schools; pressure upon clinicians to increase work rate and to diminish costs; the reduction in scope and breadth of surgical training; and the trend to increasing subspecialization. Delay in diagnosis is too often matched by ignorance of the fact that nerves can be repaired, thus mitigating or even abolishing pain and disability. The spinal accessory, brachial plexus, radial, sciatic, and common peroneal nerves are particularly at risk. Operations for biopsies of lymph nodes or tumors of nerves, for other swellings, for varicose veins, and for the internal fixation of fractures and dislocations seem to be particularly risky. The displacement and distortion of normal tissue planes and relations by fracture hematoma or displaced fragments of bone explain some of the nerve injuries seen in fracture work. Two cases exemplify what can be achieved by an alert surgeon when a mishap occurs. A 22-year-old man suffered a closed fracture of the shaft of the humerus. The operating surgeon elected to treat this with a plate and screws. Having inserted the plate, he noted that the radial nerve had become crushed between the plate and the bone, and immediately removed the implant and liberated the nerve. A later exploration of the nerve was done, at that surgeon’s request, to establish prognosis. The nerve had reconstituted to normal appearance, and recovery was complete following the course of axonotmesis. Had the initial operating surgeon failed to

1519

recognize the event or failed to act to rectify it, the outcome would have been very much less satisfactory. A 19-year-old woman sustained a common fracture of the shaft of the femur that was treated by the operating surgeon by intramedullary nail. During clearance with scissors of the soft tissues overlying the upper end of the femur, the muscles of the calf were seen to twitch. The exposure was extended, and it was found that the sciatic nerve had been cut. It was repaired immediately, with the aid of an operating microscope. One month later, that nerve was reexplored, confirming that the suture line was intact. By 2 years, there was remarkably good recovery of motor power and of sensation in the distribution of both components of the sciatic nerve. At no time was there severe pain. The spinal accessory nerve is too often divided during lymph node biopsy in the upper part of the posterior triangle of the neck. Pain and loss of function are usually immediate.170 Pain, too often ascribed to the operation itself, has three components. First and most severe is neuropathic pain from transection of afferent fibers within the nerve.26 Next is the dragging pain from traction upon the brachial plexus following loss of support of the shoulder. A later development is pain from impingement within the disorganized glenohumeral joint. Loss of function is severe, and it includes dropping of the shoulder, winging of the scapula, and loss of abduction. Narakas124 analyzed the functional consequences of injuries to different nerves passing to the shoulder girdle and glenohumeral joint in a study of exceptional importance, showing that loss of function following accessory nerve palsy was particularly severe (Fig. 62–5). Bonney,22 in an extended and relevant discussion, pointed out that many of these injuries could have been avoided by adequate knowledge of topographic anatomy, by adequate exposure within the field, and by the use of nerve stimulation.

Damage to Nerves during Replacement Arthroplasty of the Hip THR is one of the most commonly performed operations in orthopedic practice. All three trunk nerves in relation to the hip can be damaged.135 The most common cause of injury in 78 femoral nerve palsies was “various operative procedures in the vicinity of the nerve.”84 Calandruccio30 suggested that the incidence of injury to the sciatic nerve ranged from 0.7% to 3% and recommended adequate exposure of the sciatic nerve in cases in which the anatomy of the hip was distorted. Schmalzreid and Amstutz141 found the overall incidence of major nerve palsy to be 1.7%. The incidence was highest in arthroplasty done in cases of congenital dislocation of the hip (5.8%). Intraoperative monitoring is recommended in high-risk cases by several authors.82,127

1520

Diseases of the Peripheral Nervous System

FIGURE 62–5 Accessory nerve palsy after lymph node biopsy. The displacement and winging of the scapula is characteristic. There is dropping of the shoulder and restriction of shoulder elevation.

The sciatic and other trunk nerves are vulnerable to injury by cutting during exposure of the joint, by traction during dislocation or after insertion of the implant, by crushing from forceps or retractors, by the heat of diathermy or polymerizing cement, by bleeding externally or within the nerve, or by encirclement by wire or suture. Over 150 cases have been followed in our unit; operations were performed in 58. The lesion of cement burn is localized to 1 cm from the bolus of cement. Axons, myelin, and intraneural collagen are destroyed14 (Fig. 62–6). Traction is the cause of lesion in operations in which the limb is lengthened. In these cases, the nerves show attenuation of longitudinal blood supply and partial epineurial rupture. The superior gluteal nerve is vulnerable in lateral approaches to the hip.76 The consequences of these injuries are severe, and many patients will relate that pain from the nerve injury greatly exceeds the pain from the arthritis that caused them to seek the operation in the first place. The natural history is unfavorable. Oldenburg and Müller128 found good recovery in only one third of cases in their study of incidence in over 2713 arthroplasties; the overall incidence of nerve lesion of 2.24% rose to 8.5% in acetabular revisions. Electrodiagnostic evidence of an incomplete lesion may tempt the clinician into anticipating recovery. Evidence of axonal loss is inconsistent with such recovery. The good results following evacuation of external hematoma have already been related, and the operating surgeon may be well advised to give consideration to reexploration of the nerve when there is severe pain or when the lesion is complete and deep. The advantages of defining the cause and the possibility of improving the prognosis need to be balanced against the risks of a further operation in patients who may be elderly and whose general condition may not be good.

FIGURE 62–6 Thermal injury sustained during arthroplasty of hip in a 73-yearold woman. The sciatic nerve is shown stretched over extraneous cement. Partial relief of pain was obtained after excision and grafting of the lesion, which extended to about 1 cm beyond the cement.

1521

Operating on Peripheral Nerves

Damage to Small Nerves Some of the most severe consequences of iatropathic lesions are those following damage to nerves so small that few would believe that so much trouble could result. Accidental damage to the terminal branch of the radial nerve, the medial cutaneous nerve of the forearm, the digital nerves, the supraclavicular nerve, and the saphenous and sural nerves is common. In some patients, the lesion provokes the development of a chronic painful state, with spontaneous or induced pain, hyperesthesia and hyperalgesia, and even allodynia and hyperpathia. These painful states are not usually dependent on sympathetic activity and are resistant to most methods of treatment. The piling of treatment on treatment in the “pain clinic” usually makes them worse. It is best to avoid damage to cutaneous nerves by careful operating. When such a nerve is divided, the treatment with the best prospect of success is restoration of neural continuity by direct suture or by insertion of a nonneural graft such as the freeze-thawed muscle graft.

INJURIES OF THE BRACHIAL PLEXUS Closed traction injury to the supraclavicular plexus is probably the most severe of peripheral nerve injuries and is characterized by a high incidence of avulsion of spinal nerves from the cord, of associated arterial injury, and of associated injury to the chest, viscera, spinal column, and long bones. Between 450 and 500 patients suffer severe and permanent damage from closed supraclavicular lesion each year in the United Kingdom.65 The plexus was injured in between 4% and 9% of peripheral nerve lesions in World Wars I and II.169 The common element underlying closed traction lesion is distraction of the forequarter from the head, the neck,

FIGURE 62–7 Total avulsion. Note the variation in level of rupture of the dura and extent of displacement of the dorsal root ganglia. (Courtesy of Mr. George Bonney, M.S.)

and the chest. The angle between the head and shoulder is opened. Linear abrasions on the chin and the face and corresponding abrasions and bruising at the tip of the shoulder are particularly important physical signs. Road traffic accidents account for most of these injuries, but we have seen multiple avulsions in falls; from objects falling onto the tip of the shoulder; from skiing, climbing, and rugby accidents; and from traction on an upper limb caught in moving machinery. In two studies of over 100 consecutive patients, the great majority were less than 25 years old. The dominant limb was injured in over 60%, there was severe associated injury in 50%, and more than one third remained unemployed at 1 year from injury. Two thirds remained in severe pain 1 year from injury.137

Level of Lesion The rootlets and the dura are ruptured at different levels, and the DRG is variably displaced. Inspection of the cord shows stumps of rootlets or, alternatively, defects in the cord (Figs. 62–7 and 62–8). Jamieson and Bonney,77 in their first intradural repair, noted that “it was seen that the ruptures had taken place just distal to the surface of the cord: little stumps of the rootlets were visible.” This finding has a number of implications: 1. Preganglionic injury can be subdivided into lesions of rootlets peripheral to the transitional zone (TZ) and those central to the TZ (true avulsions). 2. The pain of a preganglionic lesion is more than a reflection of deafferentation. In the avulsion lesion, there is direct injury to the CNS, and such injuries might be considered as a longitudinal lesion of the spinal cord. In these patients, pain is likely to be more severe, less well localized, and more intractable.

1522

Diseases of the Peripheral Nervous System

Tinel’s sign is important in diagnosis of postganglionic rupture. The radiation of sensation produced by percussion over the ruptured nerve is a reliable indicator of the level of injury: sensory loss extending to the elbow indicates injury to C5, that to the thumb indicates C6, and that to the back of the hand indicates C7. Sensory loss extending above the clavicle (C4), phrenic palsy, and weakness of the trapezius and serratus suggest intradural injury to the upper spinal nerves.

Physiologic Investigation

FIGURE 62–8 Avulsion of C5 and C7; C6 is intact. Note the peripheral rupture of the dorsal root of C5 (above). (Courtesy of Mr. George Bonney, M.S.)

3. Partial Brown-Séquard syndrome, seen in some 10% of cases of complete lesion, may reflect ischemia from interruption of radicular vessels, but may also indicate direct damage to CNS tissue. 4. Intradural repair is a more realistic concept in rhizotomy than in avulsion lesions. Examination of rootlet biopsies from nerves exposed at operation showed that CNS tissue was evident at the tips of the roots in 2 of 10 examined specimens. Corpora amylacea mark the watershed between central and peripheral tissue.140

Bonney19 and Bonney and Gilliatt23 demonstrated persisting conduction within afferent fibers in preganglionic lesions. The time for disappearance of motor conduction varies. It is unusual to see a muscle response to stimulation of the distal stump after about 60 hours, but in one case neuromuscular conduction was seen after stimulation of the avulsed ventral root of C7 at an interval of 132 hours, and several other cases have been seen in which that interval approached 100 hours. The histamine test is useful, especially for C5/C6 and T1 injuries. Sensory action potential (SAP) recordings have to be interpreted with care within 2 weeks of injury, before wallerian degeneration is complete. Interpretation must admit variations in distribution of the plexus, as well as the possibility of separate, more distal injuries of peripheral nerve trunks and of the loss of neurons within the DRG from particularly violent displacement. Landi et al.92 developed somatosensory evoked potential (SSEP) recordings, comparing these with preoperative SAP and operative findings. This work has been greatly extended.152,153 The method is important in the analysis of the early case, but it is not quantitative; it is sensitive to scarring about the exposed spinal nerve, and it does not detect isolated intradural injury to the ventral root. Comparison of electromyography (EMG) of paravertebral muscles with preand intraoperative SSEP recordings has confirmed that ventral rootlets are more vulnerable to rupture.35

Radiologic Investigation Diagnosis Careful clinical examination permits accurate diagnosis of the extent, depth, and level of injury in most cases. The violence of injury, the line of application of force to the damaged forequarter, and the nature and distribution of pain are particularly significant. The distribution of abrasions and bruises about the neck and shoulder indicate the violence and application of force. Deep bruising within the posterior triangle relates to tearing of deep structures and possibly of subclavian vessels. Boggy swelling in the neck indicates a collection of leaking cerebrospinal fluid and/or an expanding hematoma. A Claude Bernard-Horner syndrome is strong, but not absolute, evidence of the avulsion of C8 and T1.

Plain radiographs of the neck and chest provide valuable information. Fractures of the tips of the transverse processes of the cervical vertebrae are associated with avulsion of the adjacent spinal nerves. Fracture-dislocation of the first rib indicates severe injury to the lower trunk. In severe cases, the cervical spine is tilted away from the side of injury. Myelography is useful,40,172 but contrast-enhanced computed tomography (CT) scans are more accurate.106 It is possible to demonstrate cases of selective injury to the ventral or dorsal root and also to distinguish between lesions of roots central or peripheral to the TZ. Magnetic resonance (MR) neurography has shown remarkable contrast between

1523

Operating on Peripheral Nerves

the fascicles and the internal epineurium.39,51 It is likely that further refinements in MR imaging will cause it to displace myelography.72

REPAIR OF THE BRACHIAL PLEXUS Thoburn161 described what is probably the first case of repair of a traction lesion of the brachial plexus by suture in a 16-year-old mill girl. Leffert95 and Narakas123 analyzed early work in repair of closed traction lesion in adults and children. Jamieson and Bonney77 related their experience with 76 patients, which included an account of the first intradural repair. Jamieson and Hughes79 reported 31 repairs, noting that “surgical repair of the plexus does produce reasonable results, while later repair is unrewarding.” Millesi114 and Narakas120 presented preliminary results in over 160 cases of repair. Significant publications from French surgeons1,2,144 were followed by important contributions from Japanese surgeons, who refined techniques of diagnosis and of intercostal nerve transfer to a high degree. Eighty percent of 179 patients regained functional elbow flexion following intercostal transfer by the technique of Tsuyama.117 The arguments for urgent repair are strong. Exposure within hours or days of injury permits accurate diagnosis and greatly eases the task of repair. Dissection is easier, and the retraction of ruptured nerve trunks can be overcome. Intraoperative evoked potentials may uncover a hidden intradural lesion. Nerve stimulation describes bundles within the distal stump passing to individual muscle groups and also provides evidence of distal rupture. A further advantage is the ease with which the ventral and the dorsal rootlets of an avulsed spinal nerve may be distinguished. Reinnervation of the ventral root is possible by transfer of accessory nerve or other suitable donor material. Intradural repair can be contemplated only within days of injury. Associated life-threatening injuries rule out urgent exploration in at least 20% of cases. The most important distinction lies between those cases in which some roots are intact or recovering and those in which all roots are damaged. The prognosis for recovery in the latter group is very much worse. Table 62–2 analyzes the pattern of injury in 300 consecutive operated cases of closed supraclavicular lesion in a period of 4 years. In the largest single group (117 cases), the upper nerves were damaged and the lower were intact. Table 62–3 delineates the lesions of individual spinal nerves; 826 of 1500 spinal nerves sustained intradural injury.

The Upper Lesion: Rupture or Avulsion of C5/C6 (C7) with Intact (C7) C8/T1 This is the most favorable lesion because there is residual hand function. Great advances have been made in the treatment of this group. Early repair within a few days of injury,

Table 62–2. Patterns of Injury in 300 Consecutive Operated Supraclavicular Lesions (1989–1993) Complete Lesions: Pre- and Postganglionic Injury Rupture of upper nerves C5 (C6,C7) Intradural lower nerves (C6, C7, C8) T1 Rupture of middle nerves (C6) C7 (C8) Intradural above and below Rupture of lower nerves C8, T1 Intradural upper nerves C5, C6, C7 Rupture of C5, T1 Intradural C6, C7, C8 Total Intradural C5–T1 TOTAL

83 5 1 7 52 148

Incomplete Lesions: Some Roots Intact Damage to C5, C6 (C7) Recovering or intact (C7) C8, T1 Damage to C6, C7, T1 Recovering or intact C5, T1 Damage to C7, C8, T1 Recovering or intact C5, C6 TOTAL

117 23 13 153

either by graft, by transfer, or by both, can achieve such useful results that failure indicates significant error in diagnosis or in technique, or that more distal injury to the peripheral nerves or damage to the glenohumeral joint was not fully appreciated. The aims of treatment include restoration of control of the thoracoscapular and glenohumeral joints, elbow flexion, and if necessary extension of the wrist by either nerve repair or early musculotendinous transfer. Extensive sensory loss can be mitigated by transfer of the medial cutaneous nerve of the forearm or the superficial divisions of intercostal nerves to the lateral root of the median nerve (Fig. 62–9).

The Complete Lesion The whole of the plexus is damaged in 50% of cases, and in one third of these (about one sixth of the total) all five spinal nerves have sustained intradural injuries. In those Table 62–3. Injuries To Individual Spinal Nerves in 300 Consecutive Operated Cases (1989–1993) Intact or LIC

Rupture

Intradural

C5 C6 C7 C8 T1

41 39 76 105 133

154 90 24 15 7

105 171 200 180 170

TOTAL

394

290

826

LIC ⫽ lesion in continuity.

1524

Diseases of the Peripheral Nervous System

FIGURE 62–9 A rupture of C5, C6, and C7 was repaired within 24 hours of injury. The patient, an electrician, considered function at 20 months after operation to be normal and had no pain.

cases in which only some nerves were ruptured, modest function can be expected from grafts, and it is usual to see restoration of flexion of the elbow and stability at the shoulder. Recovery of function within the hand is rare. In cases of complete avulsion, conventional nerve transfers offer only paltry mitigation. Useful results have been gained by using the double free muscle transfer technique.41 A better prospect of restoring useful function lies in reconnection between the spinal cord and the avulsed nerves. The results of repair by graft or by transfer in 367 main nerves in 153 cases operated between 1991 and 1993 are provided in Table 62–4. Fifty-five patients achieved no useful function, and 43 achieved useful function in two joints; of these, 33 were operated on within 14 days. The three patients who regained useful hand function following repair were all operated on within 14 days, proving that delay is harmful.

cord or median nerve. In all of these patients, the eighth cervical and the first thoracic nerves were irretrievably damaged. Strong elbow flexion was restored in 51 patients. Modest hand function, defined by long finger flexors to Medical Research Council (MRC) power 3 and a degree of “protective” sensation in the median territory, was regained in 14 patients. A bold idea from the Shanghai school67 uses a part of the seventh cervical nerve from the uninjured contralateral plexus as a donor. A VUNG is interposed between this and the median nerve within the injured limb.36 A striking success was seen in one personally examined case. Power in wrist and digital flexors reached MRC 4; the patient was able to localize, accurately, touch to the thumb, index, and middle fingers.

Restoration of Hand Function

The first intradural repair was done by Jamieson and Bonney at St. Mary’s Hospital in 1977.24 The cord was exposed by hemilaminectomy. All five spinal nerves had been avulsed. The dorsal roots were sutured to the little stumps of rootlets visible at levels C7 and C8. This patient was seen 20 years later, at which time there was some recovery of the pectoralis

Reconnection between the Spinal Cord and Avulsed Nerves

The vascularized ulnar nerve graft (VUNG) was used to reinnervate the median nerve in 65 patients.12 Between one and four strands of the nerve were interposed between postganglionic ruptures of the spinal nerve and the lateral

Table 62–4. Recovery and Delay in 153 Cases Operated (1991–1993)* Within 14 days

To 3 months

To 6 months

To 12 months

Partial Useful Failure

75 11

83 25

16 3

11 11

Complete Useful Failure

43 22

15 17

3 6

7 19

*Cases involved 367 “elements” (grafted roots or transfers to trunk nerves). Intercostal transfers for pain relief, hand sensation, and medial cord function are not included. Vascularized ulnar nerve grafts excluded.

1525

Operating on Peripheral Nerves

major and triceps. The patient had not experienced severe pain at any time in his course. Regeneration of axons has been shown to occur through a reattached ventral root in the dog.78 Carlstedt and his colleagues have gone very much further both experimentally and clinically. Regeneration within the ventral roots of lumbar nerves after axonotomy within the lumbar cord of the cat after cord lesion was shown.136 Half of the motor neurons survived, and axons grew into the ventral root within 1 month of injury.96 Functional recovery after reimplantation of the ventral root in rats and in monkeys has also been demonstrated.33,34 Over 20 intradural repairs have been done in humans. The posterior subscapular approach86 is used to expose the brachial plexus. The spinal cord is displayed by hemilaminectomy. Fine grafts are inserted just below the surface of the cord through small slits in the pia and are passed to the ventral roots. Carlstedt et al.32 showed elements of functional recovery in three patients. Incoordination and co-contraction of muscle activity is usual for the first 2 years after operation. Recovery of some modalities of cutaneous sensation has been measured. One possible explanation for this is through myelinated afferents within the ventral root.139 Much remains to be done to extend the principles of this intervention. It is a lengthy and potentially hazardous procedure that must be done within days of injury. It is reserved only for those patients in good general condition, in whom there has been no injury to the cervical spinal column and no suggestion of injury to the vertebral arteries. Work is proceeding to develop less invasive methods of access to the anterior face of the spinal cord by endoscopic methods.

damage during birth. Klumpke89 explained the cause and the significance of the Bernard-Horner sign in the deep complete lesion. The heavy baby or the baby born breech was at particular risk.166 Thoburn162 thought that the injury was caused by stretching or even rupture of the spinal nerve during delivery. Ubachs and Slooff,167 in an excellent review of etiology, refer to the doctoral thesis of Engelhard,47 who showed that the injury was caused by excessive stretching during delivery by excessive pull on the head against the anterior shoulder in cephalic delivery and by excessive lateral movement of the body in breech delivery. This was later confirmed.111 The early interest in operative repair was replaced by a more cautious approach when both Sever147 and Jepson80 related that the only children they saw with no recovery at all were those in whom repairs had been done. Narakas118 and Morelli et al.115 reopened the case for surgical repair in carefully selected severe cases. A particularly important contribution from Narakas121 was his classification (Table 62–5). This system is helpfully applied at about 2 weeks, when most simple conduction block lesions should have recovered. It defines the extent of injury and offers a broad guide to prognosis. As recovery proceeds, it is not unusual to see that the middle nerves, C7 and C8, have been severely damaged.28 Recovery from a Narakas group 4 lesion is often confined to the upper nerves, leaving a paralyzed hand as described by Klumpke.89 Diagnosis is usually straightforward, but fractures of the clavicle or humerus can also cause immobility. Neonatal sepsis of the glenohumeral joint has been mistaken for OBPP. Cerebral palsy arthropyosis and ischemic injuries to the cord can mimic OBPP.

OBSTETRIC BRACHIAL PLEXUS PALSY The latter part of the 19th century saw important advances in the understanding of this disorder. Duchenne42 and Erb48 described cases of injury to C5 and C6 that they attributed to

Incidence and Risk Factors Data of exceptional importance were provided by EvansJones and Kay,49 who conducted the first national survey of incidence in any country. Using the British Paediatric

Table 62–5. Clinical Classification of OBPP, Based on Examination 2–3 Weeks After Birth in Vertex Deliveries Group

Nerves Injured

Clinical Features

Likely Course

I

C5, C6

II

C5, C6, C7

Paralysis of shoulder and elbow flexors Paralysis of shoulder, elbow flexors, wrist extensors Weakness in triceps

III

C5, C6, C7, C8, T1

Complete paralysis

IV

C5, C6, C7, C8, T1 with Bernard-Horner sign

Complete paralysis

Good spontaneous recovery in 90% Good spontaneous recovery in 70% Persisting defect of shoulder, elbow in 30% Most regain useful hand function Persisting defect in shoulder and elbow usual About one half regain some function in the hand Severe defects throughout the limb usual

Data from Narakas, A. O.: Obstetrical brachial plexus injuries. In Lamb, D. W. (ed.): The Paralysed Hand. Edinburgh, Churchill Livingstone, p. 116, 1987.

1526

Diseases of the Peripheral Nervous System

Surveillance Unit (BPSU) notification system in this survey, the largest completed, 323 confirmed cases were recorded in over 776,000 live births, an incidence of 0.42%. Significant risk factors included shoulder dystocia (60% vs. 0.3% without); high birth weight, with 53% of infants weighing more than the 90th percentile; and assisted delivery. There was a considerably lower risk of OBPP in infants delivered by cesarean section. At 6 months of age, about half of these infants had fully recovered. In 2% there was no recovery. Incomplete recovery in infants with extensive lesions was more frequent than in those with less extensive lesions. These authors commented that “the incidence of CBP in the United Kingdom and the Republic of Ireland is strikingly similar to that previously reported nearly 40 years ago. Most cases are due to trauma at delivery, which is not necessarily excessive or inappropriate.” The incidence of OBPP is at least 10 times higher in breech deliveries.25,157 Lesions after breech delivery are often bilateral and frequently complicated by phrenic palsy, and there is a particularly high incidence of avulsion of the uppermost nerves of the brachial plexus from the spinal cord. The birth weight of these infants is significantly less that the mean; many are premature.17,149 The risk factors in cephalic delivery include macrosomia, multiparity, intrapartum asphyxia, and instrumental delivery.168 Risk factors in 1100 cases of OBPP58,160 were grouped into those affecting the parents, those occurring during pregnancy, those occurring during labor, and those relating to the child. These babies were heavier than the mean, and more severe lesions were associated with even heavier babies. Mothers were shorter and heavier. Shoulder dystocia was reported in over 50%. Possible intrauterine causation was suggested by a number of cases born by planned cesarian section. A few cases may relate to abnormal intrauterine position with oligohydramnios.43 In one case, EMG findings suggested intrauterine lesion.131

Course and Prognosis Many papers describe a favorable outcome for the majority of cases,7,113,121 but others take a more pessimistic view.45,46,64,148,174 Prolonged conduction block has been demonstrated.46 Gilbert and Tassin62 recorded a poor functional outcome for the fifth and sixth cervical nerves in cases in which the deltoid and biceps muscles remained paralyzed at 3 months from birth. From this study came the recommendation that exploration is indicated if, at 3 months, there is no active elbow flexion. This recommendation has been widely followed. The course of 74 babies identified in the BPSU survey and followed in one unit showed that 88% of the children achieved good or useful recovery without operative interference.15 The most important secondary deformity involved the glenohumeral joint; 20 patients (27%)

required operation to correct medial rotation contracture, posterior subluxation, or dislocation within the first 2 years of life. The outcome was very much better for Narakas group 1 than for group 4 infants. In five babies it was clear that dislocation had occurred at birth.138 In the others, the contracture was related to muscle imbalance between the powerful medial rotators and the paralyzed lateral rotators innervated by the fifth cervical nerve.

Special Investigations and Indications for Operation Most infants born with an obstetric lesion of the brachial plexus who show signs of neurologic recovery within 2 months of life subsequently regain normal upper limb function. Those who do not recover in the first 3 months of life are at risk of an unfavorable outcome. The prognosis for breech deliveries and for group 4 lesions is particularly doubtful. Clinical examination remains the most important method of assessing the severity and level of neural injury.81,93,113,158 Myelography, CT myelography, and MR imaging may be used to distinguish between avulsion and extraforaminal ruptures.54,60,149 The role of neurophysiologic investigations remains controversial. Smith150 classified lesions according to NAP and EMG findings into three categories (Table 62–6). In a study of 251 babies seen over 3 years,16 122 children showed early spontaneous recovery and no investigations were performed. In 53 children, the combination of unfavorable clinical progress with poor neurophysiologic prognosis led to operation. In 76 children, the neurophysiologic prediction was a type A or type B lesion, and these were not operated. The neurophysiologic predictions were generally confirmed by the functional outcome. A favorable initial investigation had a very high probability of predicting a good outcome: 93.8% for injury of C6 and 96% for C7. Predictions for C5 were matched in 75% of the children. This was explained by the significantly higher frequency of secondary shoulder deformity in these children and by the fact that it is not possible to record NAPs from C5. Neurophysiologic evaluation of slowly recovering OBPP can define, correctly, the nature and prognosis of the lesion. The identification of a neurapraxia (conduction block type A) is particularly important in these babies, who at presentation may be clinically indistinguishable from those with more severe lesions. The accuracy of preoperative neurophysiologic investigations by Smith’s method was confirmed by matching preoperative projections with findings at operation in 100 operated cases.90 A total of 138 spinal nerves among 145 classed as type C were found, at operation, to be either ruptured or avulsed or to have a combination of both. Premature electrodiagnostic investigation will show evidence of a complete degenerative lesion, implying that the

1527

Operating on Peripheral Nerves

Table 62–6. Classification of the Neurologic Lesion by Electrodiagnostic Investigation NAP

EMG

Lesion

Type A

Normal

Conduction block

Type B (favorable)

Normal or ⬎ 50% of uninjured side

Type B (unfavorable)

Absent or ⬍ 50% of uninjured side Absent, occasionally present

No spontaneous activity; reduced number of normal motor units, increased firing rates Relatively good motor unit recruitment; mixture of normal units and potentials suggesting collateral reinnervation Normal units few or absent; collateral reinnervation Spontaneous activity; nascent units; poor recruitment

Type C

Mild axonal (degenerative) lesion Axonotmesis Significant axonal lesion (neurotmesis) Severe axonal injury (neurotmesis or intradural)

EMG ⫽ electromyography; NAP ⫽ nerve action potential. Data from Smith, S. J. M.: The role of neurophysiological investigation in traumatic brachial plexus injuries in adults and children. J. Hand Surg. Br. 21:145, 1996.

outlook is less favorable than in fact it is; these tests seem to be most reliable at about 12 weeks of age. Ideally, however, intervention in most of the cases would be done earlier than this. Operation should be considered when there is concordance between poor clinical progress and poor neurophysiologic prediction in Narakas groups 1, 2, and 3. A case can be made for more urgent exploration in babies born by breech, most especially in the presence of phrenic palsy. Early operation should be considered in group 4 lesions in which radiologic and neurophysiologic evidence points to intradural injury. Neurophysiologic evidence of postganglionic injury is often followed by a high level of spontaneous recovery of C8 and T1 in group 4 lesions. Findings at operation are less distinct than in the adult case. The injury is a longitudinal stretch along the spinal nerves extending from the cord to the clavicle. Intradural rupture without displacement of the dorsal root ganglion is not uncommon. There is always some regeneration of fibers across postganglionic ruptures. In other cases, it is clear that there has been both postganglionic rupture and intradural injury. Some nerves appear elongated and fibrosed, with some conduction. Recovery for such lesions is generally poor. Other nerves are electrically silent, apparently intact but with neither peripheral nor central conduction. Many of these do not recover, but some do. Intraoperative neurophysiologic studies of SSEPs and of conduction across neuromas are essential in this work. Simple and reliable methods of assessment have been described.15,59,61,103,134 Recovery after repair in OBPP is remarkably slow. In some cases no evidence of recovery was seen in the hand until 3 years or more had passed, though the trophic state of the limb improved much earlier. Possibly the prolonged course reflects the slow maturation of muscle spindle, proprioceptive, and other afferent functions.3,133 In almost

every case the injured side becomes nondominant and stays so. It is exceptionally rare to encounter cases of neuropathic pain. One possible explanation is the differing expression of sodium channels related to neuropathic pain between adults and children.38,171 The largest series of repairs comes from Gilbert,60 who reported 178 cases treated by nerve reconstruction and followed for at least 3 years. Results for repairs of Narakas group 1 and 2 lesions were generally good. In 54 children with total lesions, reinnervation of the lower trunk provided useful finger flexion in 75% of cases and useful intrinsic function in 50%. Gilbert proposed reinnervation of the hand in cases of avulsion of C7, C8, and T1 by transfer of adjacent spinal nerves that were ruptured. His, too, was the idea of selective reinnervation of the ventral root in cases of avulsion: “C’est a dire provenir d’autres nerfs du plexus ou extra plexuelles et provenir de nerfs éxterieurs au plexus (spinal intercostaux etc).” These methods provide the only chance of regaining any hand function in the most severe cases of avulsions of the lower nerves, as shown in a study of 24 group 4 lesions.4 Four nonoperated patients, presenting in later life, had neither power nor sensation within the hand. In the 20 cases in which repairs were done, only 3 out of a total of 120 spinal nerves were intact. There were 47 avulsions, 58 ruptures, and 12 lesions in continuity. The aim of operation was to restore hand function, and C8 and/or T1 were repaired in all cases. Ten of these children regained moderate hand function, but in three of these there had been some spontaneous recovery through either C8 or T1. No useful motor function within the hand was seen in the other 10 children. Recovery of sensation greatly exceeded that for skeletal motor or cholinergic sympathetic function. There was recovery of sensation to normal limits in all dermatomes for at least one modality in 16 of the 20 operated cases. This study demonstrated exquisite plasticity within the

1528

Diseases of the Peripheral Nervous System

CNS. Children showed perfect localization of restored sensation in the dermatomes of avulsed spinal roots that had been reinnervated by nerves transferred from a distal spinal segment.

REFERENCES 1. Allieu, Y.: Exploration et traitement direct les lesions nerveuse dans les paralysies traumatiques par élongation du plexus brachial chez d’adulte. Rev. Chir. Orthop. Réparatrice Appar. Mot. 63:107, 1975. 2. Alnot, J. Y., Jolly, A., and Frot, B.: Traitement direct de lesions nerveuses dans les paralysies traumatique du plexus brachial chez l’adulte. Orthopade 5:151, 1981. 3. Alvares, D., and Fitzgerald, M.: Building blocks of pain: the regulation of key molecules in spinal sensory neurones during development and following peripheral axonotomy. Pain Suppl. 6:571, 1999. 4. Anand, P., and Birch, R.: Restoration of sensory function and lack of long-term chronic pain syndromes after brachial plexus injury in human neonates. Brain 125:113, 2002. 5. Anderson, S. D., Basbaum, A. I., and Fields, H. I.: Response of medullary raphe neurons to peripheral stimulation and to systemic opiates. Brain Res. 123:363, 1977. 6. Barros D’Sa, A. A. B.: Arterial injuries. In Eastcott, H. H. G. (ed.): Arterial Surgery, 3rd ed. Edinburgh, Churchill Livingstone, p. 355, 1992. 7. Bennet, C. C., and Harrold, A. J.: Prognosis and early management of birth injuries to the brachial plexus. Br. Med. J. 1:1520, 1976. 8. Berman, J. S., Birch, R., and Anand, P.: Pain following human brachial plexus injury with spinal cord root avulsion and the effect of surgery. Pain 75:199, 1998. 9. Berman, J. S., Taggart, M., Anand, P., and Chen, L.: Pain relief from preganglionic injury to the brachial plexus by late intercostal transfer. J. Bone Joint Surg. Br. 78:759, 1996. 10. Birch, R., and Bonney, G.: Classification of pain. In Birch, R., Bonney, G., and Wynn Parry, C. B. (eds.): Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingstone, p. 338, 1998. 11. Birch, R., Bonney, G., Dowell, J., and Hollingdale, J.: Iatrogenic injuries of peripheral nerves. J. Bone Joint Surg. Br. 73:280, 1991. 12. Birch, R., Dunkerton, M., Bonney, G., and Jamieson, A. M.: Experience with the free vascularised ulnar nerve graft in repair of supraclavicular lesion of the brachial plexus. Clin. Orthop. 237:96, 1988. 13. Birch, R., and St. Clair Strange, F. G.: A new type of peripheral nerve lesion. J. Bone Joint Surg. Br. 72:312, 1990. 14. Birch, R., Wilkinson, M. C. P., Vijayan, K. P., and Gschmeissner, S.: Cement burn of the sciatic nerve. J. Bone Joint Surg. Br. 74:731, 1992. 15. Bisinella, G., and Birch, R.: Obstetric brachial plexus lesion: a study of 74 children registered with the British Surveillance Unit. J. Hand Surg. Br. 28:40, 2003.

16. Bisinella, G., Birch, R., and Smith, S. J. M.: Neurophysiological predictions of outcome in obstetric lesions of the brachial plexus. J. Hand Surg. Br. 28:148, 2003. 17. Blaauw, G., Slooff, A. C. J., and Muhlig, S.: Results of surgery after breech delivery. In Gilbert, A. (ed.): Brachial Plexus Injuries. London, Martin Dunitz, p. 217, 2001. 18. Bonnard, C., and Narakas, A.: Syndromes douloureux et lesions post-traumatiques du plexus brachial. Helv. Chir. Acta 52:621, 1985. 19. Bonney, G.: The value of axon responses in determining the site of lesion in traction lesions of the brachial plexus. Brain 77:588, 1954. 20. Bonney, G.: Causalgia. Br. J. Hosp. Med. 9:593, 1973. 21. Bonney, G.: Iatrogenic injuries of nerves. J. Bone Joint Surg. Br. 68:9, 1986. 22. Bonney, G.: Iatropathic injury. In Birch, R., Bonney, G., and Wynn Parry, C. B. (eds.): Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingstone, p. 293, 1998. 23. Bonney, G., and Gilliatt, R. W.: Sensory nerve conduction after traction lesion of the brachial plexus. Proc. R. Soc. Med. 51:365, 1958. 24. Bonney, G., and Jamieson, A.: Reimplantation at C7 and C8: communication au Symposium sur le Plexus Brachial, Lausanne, 1978 [abstract]. Int. J. Microsurg. 1:103, 1979. 25. Boo, N. Y., Lye, M. S., Kanchanala, M., and Ching, C. L.: Brachial plexus injuries in Malaysian neonates: incidence and associated risk factors. J. Trop. Pediatr. 37:327, 1991. 26. Bremner-Smith, A. S., Unwin, A. J., and Williams, W. W.: Sensory pathways in the spinal accessory nerve. J. Bone Joint Surg. Br. 81:226, 1999. 27. Brouwers, P. J. A. M., Kottink, E. J. B. L., Simon, M. A. M., and Prevo, R. L.: A cervical anterior spinal artery syndrome after diagnostic blockade of the right C6 nerve root. Pain 91:397, 2001. 28. Brunelli, G. A., and Brunelli, G. R.: A fourth type of brachial plexus lesion: the intermediate (C7) palsy. J. Hand Surg. Br. 16:492, 1991. 29. Burton, A. C.: On the physical equilibrium of small blood vessels. Am. J. Physiol. 164:319, 1951. 30. Calandruccio, R. A.: Nerve injuries in hip arthroplasty. In Edmondson, A. S., and Crenshaw, A. H. (eds.): Campbell’s Operative Orthopedics, 6th ed. St. Louis, C. V. Mosby, p. 2329, 1980. 31. Campbell, W. B., France, F., and Goodwin, H. M.: Medicolegal claims in vascular surgery. Ann. R. Coll. Surg. Eng. 84:181, 2002. 32. Carlstedt, T., Anand, P., Hallin, R., et al.: Spinal nerve root repairs and replantation of avulsed ventral roots into the spinal cord of the brachial plexus in humans: the first 10 cases. J. Neurosurg. 93:237, 2000. 33. Carlstedt, T., Hallin, R. G. Hedström, K. G., and NillsonRemahl, I.: Functional recovery in primates with brachial plexus injury after spinal cord implantation of avulsed ventral roots. J. Neurol. Neurosurg. Psychiatry 56:649, 1993. 34. Carlstedt, T., Lindå, H., Cullheim, S., and Risling, M.: Reinnervation of hind limb muscles after ventral root avulsion and implantation in the lumbar spinal cord of the adult rat. Acta Physiol. Scand. 128:645, 1986. 35. Celli, L., and Rovesta, C.: Electrophysiologic intraoperative evaluations of the damaged root in traction of the brachial

Operating on Peripheral Nerves

36.

37.

38.

39.

40. 41.

42.

43. 44.

45. 46.

47.

48.

49.

50.

51.

52. 53.

plexus. In Terzis, J. K. (ed.): Microreconstruction of Nerve Injuries. Philadelphia, W. B. Saunders, p. 473, 1987. Chen, L.: C7 versus phrenic nerve, and vascularised versus non vascularised nerve grafts in reinnervation of rat biceps muscle. Presented at the Shanghai International Brachial Plexus Symposium, 1996. (Cited in Birch, R., Bonney, G., and Wynn Parry, C. B.: Traumatic lesions of the brachial plexus. In Birch, R., Bonney, G., and Wynn Parry, C. B. [eds.]: Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingston, p. 188, 1998.) Clark, W. L., Trumble, T. E., Swiontowski, M. F., and Tencer, A. F.: Nerve tension and blood flow in a model of immediate and delayed repairs. J. Hand Surg. Am. 17:677, 1992. Coward, K., Plumpton, C., Facer, P., et al.: Immunolocalisation of SNS/PN3 and NaN/SNS2 sodium channels in human pain states. Pain 85:41, 2000. Dailey, A. T., Tsuruda, J. S., Goodkin, R., et al.: Magnetic resonance neurography for cervical radiculopathy: a preliminary report. J. Neurosurg. 38:488, 1996. Davis, E. R., Sutton, D., and Bligh, A. S.: Myelography in brachial plexus injury. Br. J. Radiol. 39:362, 1966. Doi, K., Sakai, X., Kuwata, K., et al.: Double free muscle transfer to restore prehension following complete brachial plexus avulsion. J. Hand Surg. Am. 20:408, 1995. Duchenne, G. B. A.: De l’Élèctrisation Localisée et de son Application à la Pathologie et à la Thérapeutique, 2nd ed. Paris, J.-B. Balliere, p. 357, 1872. Dunn, D. W., and Engle, W. A.: Brachial plexus palsy: intrauterine onset. Pediatr. Neurol. 1:367, 1985. Eaton, R. G., and Green, W. T.: Epimysiotomy and fasciotomy in the treatment of Volkmann’s ischemic contracture. Orthop. Clin. North Am. 3:175, 1972. Eng, G.: Brachial plexus palsy in new born infants. Pediatrics 48:18, 1971. Eng, G. D., Koch, B., and Smokvina, M. D.: Brachial plexus palsy in neonates and children. Arch. Phys. Med. Rehabil. 59:458, 1978. Engelhard, J. L. B.: Verlamimingen van den plexus brachialis en n. facialis bij het pasgeboren kind. Doctoral thesis, P. Den Boer, Utrecht, 1906. Erb, W. H.: Uber eine eigenthümliche Lokalisation von Lahmungen im Plexusbrachialis. Verhandl Naturhist Med Vereins, Vol. 2. Heidelberg, Carl Winter’s Universitats Buchandlung, p. 130, 1874. Evans-Jones, G., Kay, S. P. J., Weindling, A. M., et al.: Congenital brachial palsy: incidence, causes and outcome in the United Kingdom and Republic of Ireland. Arch. Dis. Child. Fetal Neonatal Ed. 88:185, 2003. Fields, R. D., Le Beau, J. M., Longo, F. M., and Ellisman, M. H.: Nerve regeneration through artificial tubular implants. Prog. Neurobiol. 33:874, 1989. Filler, A. G., Kliot, M., Howef, A., et al.: Application of magnetic resonance neurography in the evaluation of patients with peripheral nerve pathology. J. Neurosurg. 85:299, 1996. Fischer, H. B. J.: Regional anaesthesia—before or after general anaesthesia? [editorial]. Anaesthesia 53:727, 1998. Fowler, T. J., Dunta, G., and Gilliatt, R. W.: Recovery of nerve conduction after a pneumatic tourniquet: observa-

54.

55.

56.

57.

58.

59.

60.

61.

62.

63. 64. 65.

66. 67.

68. 69.

70.

71.

1529

tions in the hind limb of a baboon. J. Neurol. Neurosurg. Psychiatry 35:638, 1972. Francel, P. C., Koby, M., Park, T. S., et al.: Fast spin-echo magnetic resonance imaging for radiological assessment of neonatal brachial plexus injury. J. Neurosurg. 83:461, 1995. Frazier, C. H., and Skillern, P. G.: Supraclavicular subcutaneous lesions of the brachial plexus not associated with skeletal injuries. J. Am. Med. Assoc. 57:1957, 1911. Gattuso, J. M., Davies, A. H., Glasby, M. A., et al.: Peripheral nerve repair using muscle autografts. J. Bone Joint Surg. Br. 70:524, 1988. Gentili, F., Hudson, A., Kline, D. G., and Hunter, D.: Peripheral nerve injection: an experimental study. Neurosurgery 4:244, 1979. Giddins, G. E. B., Birch, R., Singh, D., and Taggart, M.: Risk factors for obstetric brachial plexus palsies. J. Bone Joint Surg. Br. 76(Orthop. Proc. Suppls. II and III):156, 1994. Gilbert, A.: Obstetrical brachial plexus palsy [English trans.]. In Tubiana, R. (ed.): The Hand, Vol. 4. Philadelphia, W. B. Saunders, p. 576, 1993. Gilbert, A.: Paralysie obstetricale du plexus brachial. In Alnot, J.-Y., and Narakas, A. (eds.): Les Paralysies du Plexus Brachial, 2nd éd. (Monographie de la Société Français de Chirurgie de la Main). Paris, Expansion Scientifique Français, pp. 270, 1995. Gilbert, A., and Raimondi, P.: Evaluation of results in obstetric brachial plexus palsy: the elbow. Paper presented at the International Meeting on Obstetric Brachial Plexus Palsy, Heerlen, 1993. (Cited in Birch, R., Bonney, G., and Wynn Parry, C. B.: Birth lesions of the brachial plexus. In Birch, R., Bonney, G., and Wynn Parry, C. B. [eds.]: Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingston, p. 220, 1998.) Gilbert, A., and Tassin, J. L.: Réparation chirurgicale de plexus brachial dans la paralysie obstétricale. Chirurgie (Paris) 110:70, 1984. Gilliatt, R. W.: Physical injury to peripheral nerves. Mayo Clin. Proc. 56:361, 1981. Gjörup, L.: Obstetrical lesion of the brachial plexus. Acta Neurol. Scand. Suppl. 18:9, 1965. Goldie, B. S., and Coates, C. J.: Brachial plexus injuries—a survey of incidence and referral pattern. J. Hand Surg. Br. 17:86, 1992. Gorman, J. F.: Combat wounds of the popliteal artery. Ann. Surg. 168:974, 1968. Gu, Y. D., Zhang, G. M., Chen, D. S., et al.: Seventh cervical nerve root transfer from the contralateral healthy side for treatment of brachial plexus root avulsion. J. Hand Surg. Br. 17:518, 1992. Hall, S.: Axonal regeneration through acellular muscle grafts. J. Anat. 190:57, 1997. Harris, N. H.: Medical negligence in trauma and orthopaedics. In Powers, M. J., and Harris, N. H. (eds.): Medical Negligence, 2nd ed. London, Butterworths, p. 647, 1995. Harrison, M. J. G.: Pressure palsy of the ulnar nerve with prolonged conduction block. J. Neurol. Neurosurg. Psychiatry 39:96, 1976. Hastings, H., and Misamore, G.: Compartment syndrome resulting from intravenous regional anesthesia. J. Hand Surg. Am. 12:559, 1987.

1530

Diseases of the Peripheral Nervous System

72. Hems, T. E. J., Birch, R., and Carlstedt, T.: The role of magnetic resonance imaging in the management of traction injuries to the adult brachial plexus. J. Hand Surg. Br. 24:550, 1999. 73. Hoffman, P.: Über eine Methode, den Erfolg einer Nervennant zu beurteilen. Med. Klin. 11:359, 1915. 74. Holden, C. E. A.: Compartmental syndrome following trauma. Clin. Orthop. 113:95, 1975. 75. Hudson, A. R.: Peripheral nerve surgery. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1674, 1993. 76. Jacobs, L. G. H., and Buxton, R. A.: Course of the superior gluteal nerve in the lateral approach to the hip. J. Bone Joint Surg. Am. 71:1239, 1989. 77. Jamieson, A., and Bonney, G.: Communication au Symposium sur le Plexus Brachial, Lausanne, 1978 [abstract]. Int. J. Microsurg. 1:103, 1979. 78. Jamieson, A., and Eames, R. A.: Reimplantation of avulsed brachial plexus roots: an experimental study in dogs. Int. J. Microsurg. 2:75, 1980. 79. Jamieson, A., and Hughes, S. P. F.: The role of surgery in the management of closed injuries of the brachial plexus. Clin. Orthop. 147:210, 1980. 80. Jepson, P. N.: Obstetrical paralysis. Ann. Surg. 91:724, 1930. 81. Kay, S. P. J.: Obstetrical brachial palsy: review. Br. J. Plast. Surg. 51:43, 1998. 82. Kennedy, W. F., Byrne, T. F., Majid, H. A., and Pavlak, L. L.: Sciatic nerve monitoring during total hip arthroplasty. Clin. Orthop. 264:223, 1991. 83. Khan, R., and Birch, R.: Iatropathic nerve injuries. J. Bone Joint Surg. Br. 83:1145, 2001. 84. Kim, D. H., and Kline, D. G.: Surgical outcome for intraand extrapelvic femoral nerve lesions. J. Neurosurg. 63:783, 1995. 85. Kline, D. G.: Civilian gun shot wounds to the brachial plexus. J. Neurosurg. 70:166, 1989. 86. Kline, D. G., Donner, T. R., Happel, L., et al.: Intraforaminal repair of plexus spinal nerves by a posterior approach: an experimental study. J. Neurosurg. 76:459, 1992. 87. Kline, D. G., and Hudson, A. R.: Injection injury. In Kline, D. G., and Hudson, A. R. (eds.): Nerve Injuries. Philadelphia, W. B. Saunders, p. 46, 1995. 88. Kline, D. G., and Hudson, A. R.: Nerve action potential recordings. In Kline, D. G., and Hudson, A. R. (eds.): Nerve Injuries. Philadelphia, W. B. Saunders, p. 101, 1995. 89. Klumpke, A.: Contribution a l’étude des paralysies radiculaires du plexus brachial: paralysies radiculaires totales. Rev. Med. (Paris) 5:591, 1885. 90. Kono, H.: (Cited in Birch, R.: Invited editorial: obstetric brachial plexus palsy. J. Hand Surg. Br. 27:3, 2002.) 91. Lamerton, A. J., Bannister, R., Withrington, R., and Eastcott, H. H. G.: Claudication of the sciatic nerve. BMJ 286:1785, 1983. 92. Landi, A., Copeland, S. A., Wynn-Parry, C. B., and Jones, S. J.: The role of somatosensory evoked potentials and nerve conduction studies in the surgical management of brachial plexus injuries. J. Bone Joint Surg. Br. 62:492, 1980.

93. Laurent, J. P., Lee, R., Shena, Q. S., et al.: Neurosurgical correction of upper brachial plexus birth injuries. J. Neurosurg. 79:197, 1993. 94. Lawson, G. M., and Glasby, M. A.: A comparison of immediate and delayed nerve repair using autologous freeze-thawed muscle grafts in a large experimental animal. J. Hand Surg. Br. 20:663, 1995. 95. Leffert, R. D.: Brachial Plexus Injuries. New York, Churchill Livingstone, p. 91, 1985. 96. Lindå, H., Risling, M., Shupliakov, O., and Cullheim, S.: Changes in the synaptic input to lumbar motorneurons after intramedullary axotomy in the adult cat. Thesis, Karolinska Institutet, 1993. 97. Loeser, J. D., and Ward, A. A.: Some effects of deafferentation on neurons of the cat spinal cord. Arch. Neurol. 17:629, 1967. 98. Lundborg, G.: Neurotropism, frozen muscle grafts and other conduits. J. Hand Surg. Br. 16:473, 1991. 99. Lundborg, G., Rosén, B., Dahlin, L., et al.: Tubular versus conventional repair of median and ulnar nerves in the human forearm: early results from a prospective, randomised clinical study. J. Hand Surg. Am. 22:99, 1997. 100. Lundborg, G., and Rydevik, B.: Effects of stretching the tibial nerve of the rabbit: preliminary study of the intraneural circulation of the barrier function of the perineurium. J. Bone Joint Surg. Br. 55:390, 1973. 101. Mackinnon, S. E., and Dellon, A. L.: Injection injury. In: Surgery of the Peripheral Nerve. New York, Thieme Medical Publishers, p. 59, 1988. 102. Mackinnon, S. E., and Dellon, A. L.: Clinical nerve reconstruction with a bioabsorbable polyglycolic acid tube. Plast. Reconstr. Surg. 85:419, 1990. 103. Mallett, J.: Paralysie obstetricale du plexus brachial. Rev. Chir. Orthop. Réparatrice Appar. Mot. 58:115, 1972. 104. Mansfield, A. O.: Personal communication on hazards of tourniquet. (Cited in Birch, R., Bonney, G., and Wynn Parry C. B.: Iatropathic nerve injuries. In Birch, R., Bonney, G., and Wynn Parry, C. B. [eds.]: Surgical Disorders of the Peripheral Nerves. London, Churchill Livingstone. p. 303, 1998.) 105. Maor, P., Levy, M., Lotem, M., and Fried, A.: Iatrogenic Volkmann’s ischemia—a result of pressure transfusion. Int. Surg. 57:415, 1972. 106. Marshall, R. W., and de Silva, R. D.: Computerised tomography in traction lesions of the brachial plexus. J. Bone Joint Surg. Br. 68:734, 1986. 107. Matsen, F. A.: The compartmental syndrome: a unified concept. Clin. Orthop. 113:8, 1975. 108. Matsen, F. A.: Compartmental Syndromes. New York, Grune & Stratton, 1980. 109. McQueen, M. M., and Court-Brown, C.-M.: Compartment monitoring in tibial fractures. J. Bone Joint Surg. Br. 78:99, 1996. 110. Medical Research Council: Peripheral Nerve Injuries (Special Report Series No. 282, Seddon, H. J, ed.). London, Her Majesty’s Stationery Office, 1954. 111. Metaizeau, J. P., Gayet, C., and Periat, F.: Les lésions obstétricales du plexus brachial. Chir. Pediatr. 20:159, 1979. 112. Michaelis, M., Lui, X. G., and Jänig, W.: Axotomised and intact muscle afferents but not skin afferents develop ongoing discharges of dorsal root ganglion origin after peripheral nerve lesion. J. Neurosci. 20:2724, 2000.

Operating on Peripheral Nerves 113. Michelow, B. J., Clarke, H. M., Curtis, C. G., et al.: The natural history of obstetrical brachial plexus palsy. Plast. Reconstr. Surg. 93:675, 1994. 114. Millesi, H.: Surgical management of brachial plexus injuries. J. Hand Surg. 2:367, 1977. 115. Morelli, E., Raimondi, P. L., and Saporiti, E.: Il loro trattamento precoce. In Pipino F. (ed.): Le Paralisi Ostetriche. Bologna, Aulo Gaggi, p. 57, 1984. 116. Nagano, A., Shibata, K., Tokimura, H., et al.: Spontaneous anterior interosseous nerve palsy with hourglass-like fascicular constriction within the main trunk of the median nerve. J. Hand Surg. Am. 21:266, 1996. 117. Nagano, A., Tsuyama, N., Ochiai, N., and Hara, T.: Direct nerve crossing with the intercostal nerve to treat avulsion injuries of the brachial plexus. J. Hand Surg. Am. 14:980, 1989. 118. Narakas, A.: Brachial plexus surgery. Orthop. Clin. North Am. 12:303, 1981. 119. Narakas, A.: The effects on pain of reconstructive neurosurgery in 160 patients with traction and/or crush injury to the brachial plexus. In Siegfried, J., and Zimmerman, M. (eds.): Phantom and Stump Pain. Berlin, Springer Verlag, p. 126, 1981. 120. Narakas, A. O.: Indications et résultats du traitement chirurgicale dans les lésions par élongation du plexus brachial. Rev. Chir. Orthop. Réparatrice Appar. Mot. 63:44, 1977. 121. Narakas, A. O.: Obstetrical brachial plexus injuries. In Lamb, D. W. (ed.): The Paralysed Hand. Edinburgh, Churchill Livingstone, p. 116, 1987. 122. Narakas, A. O.: The use of fibrin glue in repair of peripheral nerves. Orthop. Clin. North Am. 19:187, 1988. 123. Narakas, A. O.: Compression and traction neuropathies about the shoulder and arm. In Gelberman, R. H. (ed.): Operative Nerve Repair and Reconstruction. Philadelphia, J. B. Lippincott, p. 1147, 1991. 124. Narakas, A. O.: Paralytic disorders of the shoulder girdle. In Tubiana, R. (ed.): The Hand, Vol. IV. Philadelphia, W. B. Saunders, p. 112, 1993. 125. Nash, T. P.: Letter to editor. Pain 96:217, 2002. 126. Nathan, P. W.: Pain and the sympathetic system. J. Auton. Nerv. Syst. 7:363, 1983. 127. Nercessian, O. A., Gonzalez, E. G., and Stinchfield, F. E.: The use of somatosensory evoked potential during revision or reoperation for total hip arthroplasty. Clin. Orthop. 243:138, 1989. 128. Oldenburg, M., and Müller, R. T.: The frequency, prognosis and significance of nerve injuries in total hip arthroplasty. Int. Orthop. 21:1, 1997. 129. Omer, G. E.: Injuries to nerves of the upper extremity. J. Bone Joint Surg. Am. 56:1615, 1974. 130. Ovelmen-Levitt, J., Johnson, B., Bedenbaugh, P., and Nashold, B. S.: Dorsal root rhizotomy and avulsion in the cat: a comparison of long term effects on dorsal horn neuronal activity. J. Neurosurg. 15:921, 1984. 131. Paradiso, G., Granana, N., and Maza, E.: Prenatal brachial plexus injury. Neurology 49:261, 1997. 132. Parkes, A. R.: Traumatic ischemia of peripheral nerves, with some observations of Volkmann’s ischemic contracture. Br. J. Surg. 32:403, 1945. 133. Payan, J.: Clinical electromyography in infancy and childhood. In Brett, E. M. (ed.): Neurology. Edinburgh, Churchill Livingston, p. 829, 1991.

1531

134. Raimondi, P.: Evaluation of the hand. Presented at the International Meeting on Obstetric Brachial Plexus Palsy, Heerlen, 1993. (Cited in Birch, R. Bonney, G., and Wynn Parry, C. B.: Birth lesions of the brachial plexus. In Birch, R., Bonney, G., and Wynn Parry, C. B. [eds.]: Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingstone, p. 220, 1998.) 135. Ratliffe, A. H. C.: Neurological complications. In Ling, R. S. M. (ed.): Complications of Total Hip Replacement. Edinburgh, Churchill Livingstone, p. 23, 1984. 136. Risling, M., Fried, K., Linda, H., et al.: Regrowth of motor axons following spinal cord lesions: distribution of laminin and collagen in the CNS scar tissue. Brain Res. Bull. 30:405, 1993. 137. Rosson, J. W.: Closed traction lesions of the brachial plexus—an epidemic among young motor cyclists. Injury 19:4, 1988. 138. Scaglietti, O.: The obstetrical shoulder trauma. Surg. Gynecol. Obstet. 66:868, 1938. 139. Schenker, M., and Birch, R.: Intact myelinated fibres in biopsies of ventral spinal roots after preganglionic traction injury to the brachial plexus: a proof that Sherrington’s “wrong way afferents” exist in man? J. Anat. 197:383, 2000. 140. Schenker, M., and Birch, R.: Diagnosis of level of intradural ruptures of the rootlets in traction lesions of the brachial plexus. J. Bone Joint Surg. Br. 83:916, 2001. 141. Schmalzried, T. P., and Amstutz, H. C.: Nerve injury and total hip arthroplasty. In Gelberman, R. H. (ed.): Operative Nerve Repair and Reconstruction. Philadelphia, J. B. Lippincott, p. 1245, 1991. 142. Seddon, H. J.: Three types of nerve injury. Brain 66:237, 1943. 143. Seddon, H. J.: Surgical Disorders of Peripheral Nerves, 2nd ed. Edinburgh, Churchill Livingstone, 1975. 144. Sedel, L.: Traitement palliatif d’une séries de 103 paralysies par élongation du plexus brachial: evolution spontanée et résultats. Rev. Chir. Orthop. Réparatrice Appar. Mot. 63:651, 1977. 145. Sedel, L.: Surgical management of lower extremity nerve lesions. In Terzis, J. K. (ed.): Microreconstruction of Nerve Injuries. Philadelphia, W. B. Saunders, p. 253, 1987. 146. Seigel, D. B., and Gelberman, R. H.: Peripheral nerve injuries associated with fractures and dislocations. In Gelberman, R. H. (ed.): Operative Nerve Repair and Reconstruction. Philadelphia, J. B. Lippincott, p. 619, 1991. 147. Sever, J. W.: Obstetrical paralysis: report of eleven hundred cases. J. Am. Med. Assoc. 85:1862, 1925. 148. Sharrard, W. J. W.: Paediatric Orthopaedics and Fractures, 2nd ed. Boston, Blackwell, p. 904, 1971. 149. Slooff, A. C. J., and Blaauw, G.: Aspects particuliers. In Alnot, J-Y., and Narakas, A. (eds.): Les Paralysies du Plexus Brachial, 2nd ed. (Monographie de la Société Français de Chirurgie de la Main). Paris, Expansion Scientifique Français, p. 282, 1995. 150. Smith, S. J. M.: The role of neurophysiological investigation in traumatic brachial plexus injuries in adults and children. J. Hand Surg. Br. 21:145, 1996. 151. Stewart, M., and Birch, R.: Penetrating missile injuries. J. Bone Joint Surg. Br. 83:517, 2001.

1532

Diseases of the Peripheral Nervous System

152. Sugioka, H.: Evoked potentials in the investigation of traumatic lesions of peripheral nerves and the brachial plexus. Clin. Orthop. 184:852, 1984. 153. Sugioka, H., and Nagano, A.: Électrodiagnostic dans l’évaluation des lésions par élongation du plexus brachial. In Alnot, J. Y., and Narakas, A. (eds.): Les Paralysies du Plexus Brachiale (Monographies du Groupe d’Étude de la Main, 15). Paris, Expansion Scientifique, p. 123, 1989. 154. Sunderland, S.: A classification of peripheral nerve injuries producing loss of function. Brain 74:491, 1951. 155. Taggart, M.: Pain after injury to the brachial plexus. (Cited in Birch, R., Bonney, G., and Wynn Parry, C. B.: Rehabilitation. In Birch, R., Bonney, G., and Wynn Parry, C. B. [eds.]: Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingstone, p. 399, 1998.) 156. Taggart, M.: Value of rehabilitation: a prospective study of 324 patients with closed traction lesions of the brachial plexus. In Birch, R., Bonney, G., and Wynn Parry, C. B. (eds.): Surgical Disorders of the Peripheral Nerves. Edinburgh, Churchill Livingstone, p. 461, 1998. 157. Tan, K. L.: Brachial palsy. J. Obstet. Gynaecol. Br. Commonw. 80:60, 1973. 158. Tassin, J. L.: Paralysies obstétricales du plexus brachial: evolution spontanée, résultats des interventions réparatrices précoces. Thèse, Université Paris VII, 1983. 159. Tavakollizadeh, A., Anand, P., and Birch, R.: Infarction of the spinal cord: an avoidable complication of the interscalene block. J. Bone Joint Surg. Br. 85B(Suppl. 2):169, 2003. 160. Tavakollizadeh, A., Taggart, M., and Birch, R.: Risk factors in obstetric brachial plexus palsy (OBPP) . J. Bone Joint Surg. Br. 85B(Suppl. 2):161, 2003. 161. Thoburn, W.: Secondary suture of the brachial plexus. Br. Med. J. 1:1073, 1900. 162. Thoburn, W.: Obstetrical paralysis. J. Obstet. Gynaecol. Br. Empire 3:454, 1903.

163. Thomas, M., Stirratt, A., Birch, R., and Glasby, M.: Freeze thawed muscle grafting for painful cutaneous neuromas. J. Bone Joint Surg. Br. 76:474, 1994. 164. Thomas, P. K., and Holdorff, B.: Neuropathy due to physical agents. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 990, 1993. 165. Tinel, J.: Nerve Wounds (revised and edited by Joll, C. A.). London, Ballière Tindall and Cox, 1917. 166. Trombetta, A.: Sullo Stiramento dei Nervi. Messina, Frat D’Angelo, 1880. 167. Ubachs, J. M. N., and Slooff, A. C. J.: Aetiology. In Gilbert, A. (ed.): Brachial Plexus Injuries. London, Martin Dunitz, p. 151, 2001. 168. Ubachs, J. M. N., Slooff, A. C. J., and Peeters, L. L. H.: Obstetric antecedents of surgically treated obstetric brachial plexus injuries. Int. J. Obstet. Gynaecol. 102:813, 1995. 169. Wilbourn, A. J.: Brachial plexus disorders. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 911, 1993. 170. Williams, W. W., Twyman, R. S., Birch, R., and Donell, S. T.: The posterior triangle and the painful shoulder: spinal accessory nerve injury. Ann. R. Coll. Surg. 78:521, 1996. 171. Yaingou, Y., Birch, R., Sangameswaran, L., et al.: Sodium channel-like immunoreactivity in human adults and neonates injured sensory nerves. FEBS Lett. 467:249, 2000. 172. Yeoman, P. M.: Cervical myelography in traction injuries of the brachial plexus. J. Bone Joint Surg. Br. 50:253, 1968. 173. Young, J. Z., and Medawar, P. B.: Fibrin suture of peripheral nerves. Lancet 2:126, 1940. 174. Zancolli, E. A., and Zancolli, E. R.: Palliative surgical procedures in sequelae of obstetrical palsy [English trans.]. In Tubiana, R. (ed.): The Hand, Vol. IV. Philadelphia, W. B. Saunders, p. 602, 1993.

E. Sporadic Motor Neuron Disease

63 Sporadic Motor Neuron Degeneration AMMAR AL-CHALABI AND P. NIGEL LEIGH

Definition and History Epidemiology Etiology Genetics Environmental Risk Factors Viruses Clinical Features, Differential Diagnosis, and Investigation

Clinical Features Differential Diagnosis Extramotor Involvement Diagnostic Certainty Delivering the Diagnosis Pathology Hypotheses of Causation Treatments

DEFINITION AND HISTORY Diseases that result in motor neuron degeneration are a heterogeneous group, perhaps best characterized by the syndrome of amyotrophic lateral sclerosis (ALS). This is by far the most common form of sporadic motor neuron degeneration and is characterized by progressive and relentless loss of both upper and lower motor neurons in the brain and spinal cord. This results in death, usually from respiratory failure caused by diaphragmatic paralysis, typically within 3 to 5 years from symptom onset. In addition to sporadic motor neuron degenerations, there are inherited motor neuron degenerations and progressive muscular atrophies (see Chapter 67). ALS was first described in the early 19th century. At this time, the functions of the spinal nerves, and more specifically the motor functions of the ventral nerves, were being discovered. The basis of progressive wasting diseases was being debated, and there were two major groups of opinion. One group believed such diseases had a neural origin,33 and the other supported a muscular origin.16,36 In 1869, the neural basis for one such disease was established when Charcot and Joffroy described a

Disease-Modifying Agents Supportive Therapies Anecdotal Treatments Clinical Trials The Future

condition with pathologic changes of degeneration of the anterior horn cells and involvement of the corticospinal tracts. They termed this amyotrophic lateral sclerosis.28 Clinically, they described a syndrome of progressive muscular atrophy with spasmodic contractions, absence of sensory involvement, and sparing of bladder function. At the start of symptoms, the burden of ALS may fall predominantly on upper motor neurons or lower motor neurons and cause predominant bulbar symptoms or limb symptoms—more focused forms of motor neuron degeneration. The syndrome of bulbar symptoms caused by upper motor neuron degeneration is known as pseudobulbar palsy, but is probably better termed corticobulbar palsy to differentiate it from the lower motor neuron equivalent, progressive bulbar palsy. Upper motor neuron degeneration causing more widespread weakness is known as primary lateral sclerosis. Lower motor neuron syndromes are more difficult to categorize with certainty. The form that can progress to ALS is known as progressive muscular atrophy. Progressive bulbar palsy and primary lateral sclerosis were recognized as related to ALS in their original descriptions.37,108 These four conditions and their 1533

1534

Diseases of the Peripheral Nervous System

ultimate end point, ALS, are grouped together in the United Kingdom under the umbrella term motor neuron disease. The course of these more focused forms of motor degeneration is usually one of progression to ALS, but this is not inevitable. Motor neuron disease can be slowly progressive, and can slow from a previously rapid rate or arrest; if this happens at a time when the disease burden is in one of the early patterns, the condition may never progress to ALS. Indeed, this is more likely to happen when the pattern of disease is focused in this way. Are at least some of these then separate diseases?

EPIDEMIOLOGY The prevalence of ALS is 4.09 per 100,000 population (range 1.0 to 13.4) as estimated from 33 reports.67 Most population studies report an incidence of about 1 to 2 per 100,000 population-years. Other than in a few very highrisk areas, the incidence is similar worldwide. On the Pacific island of Guam, on parts of the Kii peninsula of Japan, and in West New Guinea, there is a high incidence of ALS that is either atypical, associated with dementia and parkinsonism, or, in the case of New Guinea, not well documented.40 These populations are generally impoverished and endogamous, and the previously high incidence in these areas is now falling to levels more typical of the rest of the world.47,109 The lifetime risk as estimated from published population studies is between 1 in 500 and 1 in 1000 by the age of 70, being higher for men. By the age of 90, the risk is between 1 in 250 and 1 in 1000. The peak age of onset is about 56, and it is rare below the age of 30. In general, typical ALS does not occur in individuals younger than 15. At the other end of the spectrum, it is unclear whether the incidence of ALS declines with increasing old age at a faster rate than one would expect from the population demography. In other words, the risk of ALS may decline after a peak in the sixth to seventh decades. The problem is that most population studies rely heavily on neurologists for ascertainment, whereas most health care systems filter older patients to gerontologists, resulting in a bias in ascertainment. Both the Irish and Italian registers, however, show an increasing risk with age, suggesting that the decline in incidence is simply due to the lower number of individuals in these age groups.29,118 In fact, the Irish register shows barely a slowdown in the risk of ALS with increasing age, suggesting that, like Alzheimer’s disease, it may be that if one lives long enough, ALS is inevitable. This has important implications for the future incidence of ALS as the proportion of the elderly population changes in developed countries, and suggests that ALS prevalence will increase as the population ages.

ALS more frequently affects men, with a male-to-female ratio of 3:2. This sex difference is not seen in familial autosomal dominant forms of the disease and is less marked after menopause.

ETIOLOGY Genetics In about 10% of cases, ALS is autosomal dominant with age-dependent penetrance of about 96% by age 70. In the remaining cases the disease appears sporadic, but there is increasing evidence that there is a significant genetic component to disease susceptibility in such cases. For example, about 20% of the familial cases are caused by mutations in the gene for copper/zinc superoxide dismutase 1 (SOD1).101,106 Mutations in this gene are also found in about 2% of apparently sporadic cases.13,58,60 In only one case has this been conclusively shown to be a new mutation.10 Although there are more than 100 SOD1 mutations, suggesting mutation is not uncommon, there should be more than 2000 new cases of ALS worldwide per year directly attributable to SOD1 mutation. At least some SOD1 mutations show a founder effect, and it is likely that low penetrance or other disease-modifying factors influence SOD1 in sporadic ALS. Some mutations of SOD1 are known to have lower penetrance, such as the I113T mutation. The D90A mutation in single copy is not a risk factor and indeed is polymorphic in parts of Scandinavia, where it may be present in up to 2.5% of the normal population.11,12 D90A-homozygous individuals still develop ALS, but of a characteristic phenotype with predominantly spastic legs and slow progression. Occasionally D90A heterozygotes develop ALS. Founder studies have shown a single founder for all D90A cases, more recent for homozygotes. These data can be interpreted as suggesting either a linked protective factor or an accumulation of additive risk factors on the background of a low-risk allele.5,95 Pooled data from several studies totaling about 1000 cases and 1000 controls have revealed variants of the neurofilament heavy subunit gene (NEFH) containing deletions of the tail domain in about 1% of individuals with ALS and no family history.6,39,116 Similar variants were found in about 0.2% of controls (both in their 30s, below the usual age of risk). Recently, mice with deletion of the hypoxia response element of the vascular endothelial growth factor gene (VEGF) were found to develop motor neuron degeneration. In humans, single nucleotide polymorphisms (SNPs) in the promoter or leader sequence of the VEGF gene may modify risk.70 A collaborative study of about 1900 individuals showed that a particular haplotype of three SNPs was overrepresented in ALS, nearly doubling risk. Within subgroups, however, this haplotype was not associated with ALS in patients from London. Further studies are needed

1535

Sporadic Motor Neuron Degeneration

to confirm whether it is indeed a risk factor for ALS in humans. Other genetic susceptibility factors have been examined with inconsistent results. These include the apolipoprotein E gene APOE (E4 allele associated with bulbar-onset ALS, shorter survival, or younger age at onset),7,17,82,107 the ciliary neurotrophic factor gene CNTF (null allele associated with increased risk),9,43,44,57,91,111,112 and survival motor neuron (SMN) genes SMN1 (altered copy number of SMN1 increases risk)32,59,90,92 and SMN2 (homozygous deletion as a possible risk factor).81,120 Mutation of the ALS2 gene on chromosome 2 was recently found to be the cause of some types of juvenile recessive atypical ALS, with a spectrum that overlaps with hereditary spastic paraplegia and primary lateral sclerosis.38,46,125 Because of the finding of SOD1 mutations in sporadic ALS, the role of ALS2 mutations in sporadic ALS has been examined; no association was found.8,49 Small studies have examined the roles of mitochondrial SOD,35,76,100 suggesting that variants may act as susceptibility factors. Similarly, manganese SOD has been examined with contradictory results,93,113,114,119 as has APEX nuclease, a DNA repair enzyme.51,89,115 A study of three exons of the dynein heavy chain gene found no increased risk in ALS, although with such small numbers it probably did not have the power to answer this question conclusively.4 Finally, a small study has examined the role of leukemia inhibitory factor, suggesting that variants increase risk.42 The problem with all such studies has been that a candidate gene approach has until recently been the only practical way to perform association studies, but this leaves out the other 30,000 genes by definition. Also, a small effect will only be detectable with large sample sizes, and very few of the studies to date have had sufficient power (studies of SOD1, NEFH, and VEGF being notable exceptions). A genome-wide association study is the logical unbiased approach, but this currently requires highthroughput methods such as DNA pooling, and huge numbers of markers, to be feasible. What is the recurrence risk for an individual whose sibling develops disease? For those with a clear autosomal dominant history of ALS, there is a 50% chance of inheriting a causative gene and, given a high penetrance, a nearly 50% chance of developing ALS. Because about 10% of cases are familial, the overall risk could be regarded as about 5%. This calculation, however, makes two assumptions. First, it presumes that all autosomal dominant disease has high penetrance. We know this is not true, and so even a lack of family history does not mean there is not a single-gene cause of disease for an affected individual. Second, it presumes that there is no underlying familial susceptibility, perhaps caused by multiple genetic or shared environmental factors, in those with no apparent family history. In general, for a trait with a normally distributed liability, the risk to a first-degree relative is the square root of the population risk. If we use the figures for lifetime risk

of ALS given previously, it would be reasonable to propose that the risk for a first-degree relative might be 1 in 30, or 3% if liability to ALS were normally distributed. Lambda S, a genetic epidemiology measure of familiality, is the risk to a sibling compared with the population risk. In the case of ALS, we can estimate this at between 30 and 50.

Environmental Risk Factors At present, SOD1 mutations are the only proven cause of ALS. Several studies have examined nongenetic risk factors, but few with replicated or consistent conclusions. Some factors have been examined because clinicians working with ALS have the impression that they increase risk, or because of high-profile cases. Such anecdotal risk factors include exercise, thinness, overseas travel, and immunizations. Of these, probably the most robustly examined has been exercise. A recent study showed that being a team player in varsity sports increased risk about 1.7 times, and being generally thin was also a risk factor.103 Association is not causation, however, and it is not clear if such findings are a result of protection from cardiovascular disease or because of a hidden common factor itself associated with, for example, slimness and athletic achievement. A Washington State population study found an association with dietary intake of fat and with smoking, but not with alcohol consumption,86,87 implying that this is not the mechanism. There are no occupations consistently shown to be at risk, but there may be an association with work in the armed services, being a pilot, work in the printing industry, welding, exposure to magnetic fields, exposure to agricultural chemicals, and being a league soccer player.27,77,83,110 A possible mechanism for at least some of these increased risks would be through VEGF. If the response to hypoxia induced by sport is inadequate, it may increase motor neuron loss. A similar mechanism could also account for increased risk for airline pilots and smokers. The possible role of VEGF in human ALS also has implications for management of respiratory failure. Other mechanisms could represent exposure to heavy metals or toxins.

Viruses There is suggestive evidence that viruses may play a role in ALS. Human T-lymphotrophic virus type 1 (HTLV-1) is associated with a syndrome of spasticity, and another retrovirus, human immunodeficiency virus (HIV), can mimic ALS.31 Polio, an enterovirus, causes a lower motor neuron syndrome. Using a highly sensitive assay, reverse transcriptase activity was recently detected in the serum of 33 of 56 cases, but only in 3 of 58 controls.15 This was not due to any known retroviruses and may represent activated endogenous retrovirus sequence or exogenous infection. There is also evidence for enterovirus involvement.61,123

1536

Diseases of the Peripheral Nervous System

CLINICAL FEATURES, DIFFERENTIAL DIAGNOSIS, AND INVESTIGATION Clinical Features The neurodegenerative process of ALS may start long before symptoms become apparent. Evidence for this comes, for example, from the batting score of the famous Yankees baseball player, Lou Gehrig, one of the greatest of all time. His home run record declined about 2 years before he developed symptoms.64 Muscle cramps or fasciculations, both representing lower motor neuron damage, may be an early symptom, but typically an affected individual seeks medical advice for weakness. Weakness in the bulbar region usually presents first, with slurred speech, which may be of the “Donald Duck” type, followed by dysphagia for liquids and solids, and nasal regurgitation of food. As the disease progresses, examination reveals a wasted, fasciculating, spastic tongue; brisk jaw jerk and facial reflexes; mild facial weakness; and impaired palatal elevation. Because the tongue is slow, weak, and small, there is difficulty using the natural flow of saliva to coat the teeth and oral mucosa. As a result, saliva dribbles from the mouth, dental hygiene suffers, and there may be oral thrush. Weakness in the upper limbs usually manifests as weakness of pinch, grip, holding and turning keys, or opening bottle tops, and that in the lower limbs as repeated falls or tripping up. Flexor spasms may occur in the legs and represent upper motor neuron damage. Examination reveals mixed upper and lower motor neuron signs in the form of fasciculations, wasting, increased tone, and brisk reflexes. Abdominal reflexes are preserved and the plantars down-going often until quite late in the disease, even in the presence of marked spasticity and hyperreflexia. This is in contrast to, for example, multiple sclerosis, in which such changes tend to occur early. Rarely, ALS may present with respiratory failure, although diaphragmatic weakness is inevitable as the disease progresses and is the usual cause of death. This manifests as breathlessness and daytime sleepiness, often with headache and orthopnea. Respiratory symptoms tend to manifest with forced vital capacity below 70% of predicted or reduced nasal sniff pressure. Use of accessory muscles may be obvious, and there may be paradoxical movement of the abdomen. Symptoms and signs spread contiguously through anatomic segments rather than patchily, even for pure upper or lower motor neuron syndromes. Spread cranially is faster than the reverse.21 Occasionally, a predominantly proximal, symmetrical, lower motor neuron involvement of the arms may occur, causing flail arms, and a similar phenotype may occur in the legs.55 The flail arm phenotype is overwhelmingly more common in males, with a 9:1 ratio. Rarely, what starts as typical ALS may remain confined to a single limb, causing monomelic ALS.

Differential Diagnosis Mixed upper and lower motor neuron signs are the hallmark of ALS, but other conditions can also produce this pattern. Structural lesions in the spinal cord or brainstem may produce lower motor neuron signs at the site of the lesion and upper motor neuron signs below, and for clinical diagnosis of ALS, it is important for an upper motor neuron sign to be cranial to a lower motor neuron sign. Metabolic causes such as hexosaminidase A deficiency or adrenoleukodystrophies may need to be excluded. Infections such as HIV or syphilis, or systemic conditions such as vasculitis may also need exclusion. Lyme disease, although often included in the differential of ALS, is usually straightforward to distinguish. Predominantly upper motor neuron signs may be seen with adrenoleukodystrophy or multiple sclerosis, both demonstrable with magnetic resonance imaging (MRI). Conditions such as hereditary spastic paraparesis can usually be distinguished on the basis of cerebellar, sphincter, or sensory involvement. Tropical spastic paraparesis secondary to infection with HTLV-1 can be diagnosed by examination of spinal fluid. Lower motor neuron syndromes have a greater range of differential diagnoses. Multifocal motor neuropathy with conduction block is an autoimmune peripheral demyelinating condition related to chronic inflammatory demyelinating polyneuropathy. It tends to occur in younger people and can be very asymmetrical or focal. Wasting tends to be disproportionately mild compared to weakness at first, although it may become more marked as the disease progresses.26,94,97 Antibodies, if detected, are directed against ganglioside GM1. Conduction block can be difficult to demonstrate electrophysiologically because the block may be very proximal, and, if suspected, a trial of immunoglobulins is worthwhile. Kennedy’s disease68 is an X-linked spinal muscular atrophy and therefore affects men. It presents with bulbar and limb weakness. Gynecomastia and testicular atrophy may be present. The cause is a trinucleotide repeat expansion in the androgen receptor gene. Kennedy’s disease is compatible with a normal lifespan, and it is therefore essential that the diagnosis be made correctly. Other forms of spinal muscular atrophy also produce lower motor syndromes, but there is usually a family history.50,56 Genetic tests are available. Infective processes such as polio, or exposure to heavy metals or neurotoxins, may be obvious because of the history.

Extramotor Involvement Although ALS is usually thought of as a pure motor disease, it is clear that extramotor symptoms can occur. The most common of these is emotional lability. This is a distressing symptom of inappropriate laughter or crying and is particularly associated with involvement of the corticobulbar pathways. Less frequently, ALS may occur in association with

1537

Sporadic Motor Neuron Degeneration

overt frontotemporal dementia, and, interestingly, subclinical motor neuron involvement can be seen in many individuals with frontotemporal dementia. In support of a common pathologic pathway for these conditions, genetic studies have recently shown linkage to chromosome 9q21-22 in a set of families in which ALS and frontotemporal dementia occur together not only in the same pedigree but also sometimes in the same individual.54 Even in those without overt frontotemporal features, psychological studies have shown that subclinical frontal lobe executive dysfunction is not uncommon.1,3 At its most subtle this manifests as a subtle impairment of planning, but in about a third of individuals, there are high levels of distractibility in three-choice tasks such as the Wisconsin and Stroop tests. These findings are more common in those with pseudobulbar palsy, but may be found in all type of ALS. The psychological evidence is supported by imaging studies that show subcortical frontal lobe involvement in nondemented ALS patients on volumetric MRI, functional MRI, magnetic resonance spectroscopy, and positron emission tomography scanning. There may also be subclinical autonomic involvement in ALS, as suggested by alterations in salivary gland function.45 The typical ubiquitin immunoreactive inclusions of ALS may be seen outside the normal motor regions in individuals with both ALS and frontotemporal dementia, and are found in the dentate gyrus and superficial cortical neurons.73 In motor neuron inclusion body dementia, pathologically the most common form of frontotemporal dementia, inclusions identical to those found in ALS are found in the hippocampus and superficial cortical neurons.117,124 The relationship between frontotemporal dementia and ALS and the wider implications for neurodegeneration remain intriguing.

Diagnostic Certainty In practice, the diagnosis of ALS is usually straightforward clinically, but requires investigation to exclude other possibilities, including those listed previously. Unfortunately, there is no positive diagnostic test and it remains a diagnosis of exclusion. Electromyographic evidence in support of the diagnosis is mandatory. It is usually appropriate to admit patients for the investigations and to discuss in detail the implications of the diagnosis. For about a fifth of cases, there is diagnostic difficulty, most often with presentations involving pure lower motor neuron or pure upper motor neuron signs confined to one limb, or in young patients. Because of the uncertainty of clinical features at the outset, and because no diagnostic test exists for ALS, patients are classified according to research criteria for the certainty of diagnosis. These World Federation of Neurology criteria were first elaborated in El Escorial, Spain, and take their name from this place. The original El Escorial criteria were based on the distribution of upper and lower motor neuron signs and were purely

clinical.19 A subsequent modification includes neurophysiologic and other evidence.20 Signs are classified as bulbar, cervical, thoracic, or lumbar. In all cases the diagnosis requires progression and, if both are present, an upper motor neuron sign neurologically higher than a lower motor neuron sign. The presence of upper and lower motor neuron signs in three regions then constitutes definite ALS, that in two regions probable ALS, and that in one region possible ALS. Pure upper motor neuron signs are also in the possible category, and pure lower motor neuron signs in the probable category. There have been several problems with these criteria. In practice, it is possible for someone to be in the El Escorial possible category, but for there to be little doubt about the diagnosis of ALS. This is because these are primarily research criteria, designed for use in clinical trials. Along similar lines, one might expect that El Escorial definite ALS is more advanced disease than El Escorial possible ALS. Unfortunately, it has proved difficult to use the El Escorial criteria for staging, and other systems are needed. Finally, there can be confusion about whether the different regions containing signs refer to neurologically or anatomically defined regions. This is because, although it may be obvious that biceps involvement is cervical, if the lesion is upper motor neuron then it is less logical to use a neurologic reference. As a result, patients may be inadvertently classified according to anatomic distribution rather than neurologic distribution of signs.

Delivering the Diagnosis It is extremely important that the diagnosis of ALS is given in an appropriate manner and environment. In general, it should be done with family and nursing staff present if possible, no interruptions, and plenty of time available. In practice this usually means admitting someone to the ward. Although the diagnosis may be obvious clinically almost immediately, the need to exclude other conditions provides a good reason to admit someone for investigation. Studies repeatedly show that the amount of information retained after such a meeting is low and a follow-up meeting a few days later is very helpful. Support groups such as the Motor Neuron Disease Association or the ALS Association are invaluable.

PATHOLOGY There is increasing evidence that ALS involves more than just the motor system, although clearly the majority of the pathology is in the voluntary motor pathways.2,14,22 Typically, there is loss of pyramidal neurons of the corticospinal pathway with loss and degeneration of large anterior horn cells and brainstem cranial nerve nuclei. There is

1538

Diseases of the Peripheral Nervous System

often marked astroglial proliferation in affected areas.63 Skeletal muscles show denervation atrophy. Both upper and lower motor neurons show proximal and distal axonopathy with senescent changes such as deposition of lipofuscin.78 In the corticospinal tracts, the pattern of axonal damage tends to be distal to proximal, a “dying-back” process.25,41 This is seen in both sporadic and familial ALS. Although ALS affects the voluntary motor pathways, the oculomotor, abducens, and trochlear nuclei are relatively spared, as are neurons of Onuf’s nucleus, which innervates the vesicorectal sphincters.75 Onuf’s nucleus is also unaffected in acute and chronic polio,66 and this may be because the neurons are histologically more like autonomic neurons than motor neurons.65 The androgen receptor, which is absent from Onuf’s nucleus and nuclei concerned with eye movement, may play a part in motor cell death in ALS.121 It certainly has a role in the death of lower motor neurons in Kennedy’s disease, a condition caused by trinucleotide repeat expansion in the androgen receptor gene. Another possible explanation is that these neurons express more calcium-buffering proteins than more vulnerable upper or lower motor neurons.105 Lower motor neurons show the most obvious cytoplasmic pathologic changes. Focal perikaryal accumulations of neurofilaments are rare but characteristic, and phosphorylated neurofilament accumulations are a hallmark of ALS.84 These neurofilament accumulations retain a filamentous structure and are not therefore completely disorganized. Spheroids, which are also neurofilamentous accumulations, are found in myelinated axons and perikarya.52 The most characteristic cellular inclusions are ubiquitinimmunoreactive inclusions, which are almost specific to ALS.71,74 These are skeinlike or dense, rounded structures and appear to be filamentous, with a tubular cross-section and diameter of 15 to 25 nm. Ubiquitin is involved in the degradation of abnormal proteins via an ATP-dependent nonlysosomal pathway, but ubiquitinated proteins are also processed via endosomes and lysosomes. Bunina bodies are specific for ALS and are also found in the more focused forms, occurring in about two thirds of all cases.23,30 These small (2- to 7-␮m) eosinophilic inclusions occur singly or in groups or chains. They have a honeycombed tubular and vesicular structure consisting of an amorphous electron-dense material. Although smaller Bunina bodies are rarely ubiquinated, the larger reticulated ones may well be. They react with antibodies to cystatin C,88 but attempts to characterize them further have so far been unsuccessful. Lewy body–like (hyaline) inclusions are found in familial and sporadic ALS.34,53 They are 3 to 5 ␮m in diameter and are surrounded by a clear halo. They may be associated with Bunina bodies or even contain them. They also contain granule-coated microtubules, sheaves of microtubules, and vacuoles. Another type of inclusion is seen in

young-onset ALS, as a basophilic, 4- to 16-␮m, RNA-rich structure.122 Electron microscopy demonstrates these structures as fuzzy microtubules associated with rough endoplasmic reticulum. It has been proposed that there is an evolution from a basophilic inclusion to the Lewy body–like hyaline inclusion to the Bunina body in the proximal axon, as the axon undergoes atrophy, demyelination, and dying back distally as a result of axostasis.30

HYPOTHESES OF CAUSATION The exact mechanism underlying ALS remains unknown. There are three main hypotheses, all of which have strong supporting evidence. These and the many other hypotheses should not be seen in isolation, but rather as interrelated ideas about a common pathway to ALS. The first major hypothesis is the excitotoxic hypothesis. Whereas most neurotransmission is inhibitory, glutamate, the neurotransmitter of the corticospinal tracts and certain spinal interneurons, causes excitation of the postsynaptic terminal. This is mediated through various types of glutamate receptors. Splice variants combine with subunit variations to produce a large number of possible receptors, which can be classified into two main groups: metabotropic receptors, which are G protein linked, and inotropic receptors, which allow influx of sodium (non–N-methyl-D-aspartate [NMDA] receptors) or calcium (NMDA receptors). Glutamate is cleared from the synaptic cleft by glutamate transporters found mainly in astrocytes. These excitatory amino acid transporters (EAATs 1 through 4) are essential in keeping extracellular glutamate low, because excessively high concentrations lead to necrotic (and possibly apoptotic) neuronal cell death. Calcium influx activates intracellular cascades and the production of nitric oxide, peroxynitrite, and superoxide. Plasma levels of glutamate have been found to be elevated in ALS patients,24,98 although this is not a consistent finding.96 Studies of amino acids in the cerebrospinal fluid of patients with ALS have also shown an increase in glutamate compared with agematched disease and control patients.102 It is possible that these changes represent a subgroup of patients in whom there is an abnormality of glutamate transport.104 EAAT2 is a glutamate transporter expressed solely on astrocytes. Levels are reported to be reduced in the cortex and spinal cord of patients with ALS compared with controls, and this is apparently because of aberrant messenger RNA transcripts.72 However, other studies79,85 have shown that aberrant transcripts may also be found in normal controls. EAAT2 can be inactivated by oxidative reaction products of at least two mutant forms of SOD1 (A4V and I113T) but not by wild-type SOD1. Such inactivation leads to neuronal death. Riluzole, a benzothiazole derivative with neuroprotective properties, increases

1539

Sporadic Motor Neuron Degeneration

survival in ALS patients.69 This drug decreases glutamate release at high rates of stimulation, acts to decrease glutamate transmission by inactivating neuronal sodium channels, and has a postsynaptic effect that is G protein linked. The second hypothesis comes from the discovery of SOD1 mutations in ALS. Although SOD1 mutations account for only 2% of all ALS, the mechanism by which they cause disease are discussed briefly here and in more detail in Chapter 67. Superoxide is a free radical anion normally detoxified by SOD1 to produce hydrogen peroxide, which is in turn detoxified by catalase to produce water and molecular oxygen. In the presence of reduced copper, however, hydrogen peroxide is converted to the dangerous hydroxyl radical. Because the active site of SOD1 contains reduced copper and mutant forms of the molecule result in exposure of the active site, there is a potential source of such free radicals in motor neurons. Calcium influx upregulates neuronal nitric oxide synthase, leading to greater production of nitric oxide. Excess superoxide anion and nitric oxide can be catalyzed by mutant SOD1 to form peroxynitrite, which nitrosylates tyrosine residues of proteins. This and the hydroxyl radical can damage neurofilaments and other cytoskeletal proteins. The excitotoxic mechanism and free radical production are thus linked. There are several other ideas about how mutant SOD1 causes ALS, including an increased tendency to form beta pleated sheets and therefore aggregate, and an increased tendency to form pores leading to membrane damage. The third hypothesis is based on the observation that cytoskeletal disturbances occur in ALS. Phosphorylated neurofilamentous accumulations in the proximal axon and perikarya of motor neurons are a pathologic hallmark. In transgenic mice, alterations in neurofilament subunit stoichiometry or makeup can lead to neuronal cell death.62 In humans, mutations consisting of deletions of the NEFH tail have been found in about 1% of cases. More recently, missense mutations in another cytoskeletal component, the dynein heavy chain, have been found to be the cause of a mouse model of ALS, the legs-atodd-angles (loa) mouse.48 Again, in humans a motor neuron disease similar to Kennedy’s disease has been found to be due to mutation in dynactin, a component of the cytoskeleton.99

Disease-Modifying Agents Riluzole was originally, and unsuccessfully, developed as an anti-excitatory anticonvulsant. With the emergence of excitotoxicity as a possible mechanism for neuronal cell death, attention was turned to its potential as a neuroprotective agent. Riluzole is the first drug licensed for use in the treatment of ALS. It is also the first drug shown to modify the course of any neurodegenerative disease. The effect is modest but consistently demonstrated in many different studies and designs, and a Cochrane review of the evidence suggests that the benefits outweigh the costs.80 Riluzole is taken by mouth at a dose of 50 mg twice daily. The main evidence for the efficacy of riluzole comes from two large international randomized, double-blind, placebo-controlled trials.18,69 The first involved 155 patients randomized to 100 mg/day riluzole or placebo for up to 21 months, using primary outcome measures of death and tracheostomy. There was a significant difference in survival at 12 and 21 months, with a significantly slower deterioration in muscle strength for the treated group. The median survival was 449 days for those on placebo and 532 days for those on riluzole. Overall, riluzole therapy reduced mortality by 38.6% at 12 months (relative risk, 0.61; 95% confidence interval [CI], 0.39 to 0.97; P ⫽ .014) and by 19.4% at 21 months (P ⫽ .046).18 The second trial involved 959 patients in four trial arms consisting of 50, 100, or 200 mg/day riluzole or placebo. The primary outcome measure was tracheostomy-free survival. There was a trend toward significant improvement in survival that fell short of the P ⫽ .05 level at 18 months. The optimal dose was 100 mg/day. Tracheostomy-free survival was 50.4% for those on placebo compared with 56.8% for those taking 100 mg/day riluzole. The unadjusted risk was equivalent to a gain in survival of about 3 months, although the Kaplan-Meier analysis did not reach the median by 18 months, and therefore this can only be an estimate. At 12 months the difference in survival was significant, with a hazard ratio of 0.71 (95% CI, 0.54 to 0.92). Because of problems with assumptions made in the initial power calculations, and because there were clear differences in known prognostic factors among the patients in the study despite randomization, the Cox proportional hazards model was used. In this analysis the relative risk reduction of riluzole was greater, being 0.65 (95% CI, 0.50 to 0.85; P ⫽ .002) at 18 months.69

TREATMENTS ALS was previously considered “untreatable,” despite the existence of palliative and other therapies. Treatment methods can be classified into three groups: disease-modifying agents, supportive therapies (including the multidisciplinary team and symptomatic treatments), and anecdotal treatments.

Supportive Therapies Nearly all symptoms of ALS can be ameliorated to some extent, except weakness. Emotional lability responds to antidepressants, and newer therapies will be available soon. In the early stages, excess saliva is best treated with glycerine swabs and drinks of pineapple or cranberry juice.

1540

Diseases of the Peripheral Nervous System

Pharmacologically, it can be dried up with a tricyclic antidepressant, or with an antihistamine. Antihistamines may be delivered topically as drops or parenterally in the form of skin patches. Botulinum toxin injection to the parotid gland requires care and skill because of the risk of worsening dysphagia, but can be effective in some cases. Radiotherapy to one or the other parotid gland may be useful in extreme cases. Cramps may respond to quinine sulfate (300 mg/day). Spasticity can be treated by use of a standing frame and physiotherapy, or pharmacologically by baclofen or tizanidine. Dantrolene is an alternative but is not suitable for long-term use. Depression is a common problem, and although counseling services are useful, the concomitant use of antidepressants is almost always needed. The multidisciplinary team is now a vital component of the management of ALS. The physical therapist can provide help with joints, spasticity, and mobility, including the use of ankle-foot orthoses. The occupational therapist is able to find practical solutions to many of the problems caused by progressive weakness and immobility, such as large-gripped cutlery, extending grabbers, self-tying shoelaces, hoists, handles, and wheelchairs. A speech-language therapist can help with communication aids, including speech synthesizers, and is also trained to advise about swallowing. A dietitian can find practical ways to maintain nourishment as it becomes more difficult not only to feed but also to swallow. Social workers and others can advise on financial and legal help, and community nurses can provide help with the physical tasks of caring for a disabled individual. Finally, the hospice team is able to provide palliative and terminal care in an environment more suited to this than an acute hospital setting. Respiratory support is a fast-improving area of management, and there is increasing evidence that noninvasive ventilation improves quantity as well as quality of life. Formal respiratory assessment requires involvement of a pulmonologist, respiratory function testing, overnight oximetry, and arterial blood gas estimation. Management includes provision of noninvasive positive pressure ventilation and anxiolytics such as benzodiazepines and morphine-based drugs. The use of tracheostomy and invasive ventilation is not universally accepted because of the ethical implications of the “lockedin” syndrome that is the inevitable consequence. If dysphagia can no longer be managed by liquidizing or pureeing food and adding high-calorie supplements, insertion of a gastrostomy tube becomes necessary. Insertion via endoscope (percutaneous endoscopic gastrostomy) or radiologically is best carried out while the patient remains well nourished and has a level of respiratory function that does not make the procedure dangerous.

Anecdotal Treatments Treatments such as vitamins C and E, coenzyme Q10, creatine supplements, minocycline, gabapentin, antiretroviral agents, insulin, insulin-like growth factor, and tamoxifen

are taken by many patients on the basis of variously credible scientific or other claims about efficacy and safety. In general, we adopt a policy of educating patients to be aware of overpromoted or expensive therapies, and to ask us regarding evidence of safety.

CLINICAL TRIALS There have been more than 60 clinical trials in ALS to date. The search for new therapeutic agents has been given more impetus now that one agent has been shown to be beneficial. If ALS could be slowed or reversed significantly, the prevalence would be similar to that of multiple sclerosis. The increasing age of the population and the effects of multidisciplinary care, noninvasive ventilation, and riluzole on survival will all lead to an increased prevalence and an increased need for an effective agent. Current research is focused not only on traditional methods of drug delivery, but also on delivery of gene-based therapies by viral vectors, the therapeutic use of stem cells (including bone marrow transplantation), and the use of RNA interference to switch off deleterious genes.

THE FUTURE Although it seems likely that, in the early stages of therapy discovery, successful treatments will cause the disease to be slowed rather than stopped or reversed, the ultimate aim is a cure. A predictive or diagnostic test would be invaluable, and would potentially allow the development of appropriate prevention. A few years ago, ALS was regarded as a hopeless disease that could never be beaten. With increasing advances in research and therapeutics, an effective treatment is only a matter of time.

REFERENCES 1. Abrahams, S., Goldstein, L. H., Al-Chalabi, A., et al.: Relation between cognitive dysfunction and pseudobulbar palsy in amyotrophic lateral sclerosis. J. Neurol. Neurosurg. Psychiatry 62:464, 1997. 2. Abrahams, S., Goldstein, L. H., Kew, J. J. M., et al.: Frontal lobe dysfunction in amyotrophic lateral sclerosis: a PET study. Brain 119:2105, 1996. 3. Abrahams, S., Goldstein, L. H., Lloyd, C. M., et al.: Cognitive deficits in non-demented amyotrophic lateral sclerosis patients: a neuropsychological investigation. J.Neurol.Sci. 129(Suppl.):54, 1995. 4. Ahmad-Annuar, A., Shah, P., Hafezparast, M., et al.: No association with common Caucasian genotypes in exons 8, 13 and 14 of the human cytoplasmic dynein heavy chain gene (DNCHC1) and familial motor neuron disorders. Amyotroph. Lateral Scler. Other Motor Neuron Disord. 4:150, 2003.

Sporadic Motor Neuron Degeneration 5. Al-Chalabi, A., Andersen, P. M., Chioza, B., et al.: Recessive amyotrophic lateral sclerosis families with the D90A SOD1 mutation share a common founder: evidence for a linked protective factor. Hum. Mol. Genet. 7:2045, 1998. 6. Al-Chalabi, A., Andersen, P. M., Nilsson, P., et al.: Deletions of the heavy neurofilament subunit tail in amyotrophic lateral sclerosis. Hum. Mol. Genet. 8:157, 1999. 7. Al-Chalabi, A., Enayat, Z. E., Bakker, M. C., et al.: Association of apolipoprotein E epsilon 4 allele with bulbar-onset motor neuron disease. Lancet 347:159, 1996. 8. Al-Chalabi, A., Hansen, V. K., Simpson, C. L., et al.: Variants in the ALS2 gene are not associated with sporadic amyotrophic lateral sclerosis. Neurogenetics 4:221, 2003. 9. Al-Chalabi, A., Scheffler, M. D., Smith, B. N., et al.: Ciliary neurotrophic factor genotype does not influence clinical phenotype in amyotrophic lateral sclerosis. Ann. Neurol. 54:130, 2003. 10. Alexander, M. D., Traynor, B. J., Miller, N., et al.: “True” sporadic ALS associated with a novel SOD-1 mutation. Ann. Neurol. 52:680, 2002. 11. Andersen, P. M., Forsgren, L., Binzer, M., et al.: Autosomal recessive adult-onset amyotrophic lateral sclerosis associated with homozygosity for Asp90Ala CuZn-superoxide dismutase mutation: a clinical and genealogical study of 36 patients. Brain 119:1153, 1996. 12. Andersen, P. M., Nilsson, P., Ala-Hurula, V., et al.: Amyotrophic lateral sclerosis associated with homozygosity for an Asp90Ala mutation in CuZn-superoxide dismutase. Nat. Genet. 10:61, 1995. 13. Andersen, P. M., Nilsson, P., Keranen, M. L., et al.: Phenotypic heterogeneity in motor neuron disease patients with CuZn-superoxide dismutase mutations in Scandinavia. Brain 10:1723, 1997. 14. Anderson, V. E., Cairns, N. J., and Leigh, P. N.: Involvement of the amygdala, dentate and hippocampus in motor neuron disease. J. Neurol. Sci. 129(Suppl.):75, 1995. 15. Andrews, W. D., Tuke, P. W., Al Chalabi, A., et al.: Detection of reverse transcriptase activity in the serum of patients with motor neurone disease. J. Med. Virol. 61:527, 2000. 16. Aran, F. A.: Recherches sur une maladie non encore decrite du systeme musculaire (atrophie musculaire progressive). Arch. Gen. Med. 24:172, 1850. 17. Bachus, R., Bader, S., Gessner, R., and Ludolph, A. C.: Lack of association of apolipoprotein e epsilon 4 allele with bulbar-onset motor neuron disease. Ann. Neurol. 41:417, 1997. 18. Bensimon, G., Lacomblez, L., and Meininger, V.: A controlled trial of riluzole in amyotrophic lateral sclerosis. ALS/Riluzole Study Group. N. Engl. J. Med. 330:585, 1994. 19. Brooks, B. R.: El Escorial World Federation of Neurology criteria for the diagnosis of amyotrophic lateral sclerosis. Subcommittee on Motor Neuron Diseases/Amyotrophic Lateral Sclerosis of the World Federation of Neurology Research Group on Neuromuscular Diseases and the El Escorial “Clinical Limits of Amyotrophic Lateral Sclerosis” workshop contributors. J. Neurol. Sci. 124(Suppl.):96, 1994. 20. Brooks, B. R., Miller, R. G., Swash, M., and Munsat, T. L.: El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis. Amyotroph. Lateral Scler. Other Motor Neuron Disord. 1:293, 2000.

1541

21. Brooks, B. R., Mitsumoto, H., Haverkamp, L., and Miller, R. G.: Natural history of ALS: symptoms, strength, pulmonary function, and disability. Neurology 47:S71, 1996. 22. Brownell, B., Oppenheimer, D. R., and Hughes, J. T.: The central nervous system in motor neurone disease. J. Neurol. Neurosurg. Psychiatry 33:338, 1970. 23. Bunina, T. L.: On intracellular inclusions in familial amyotrophic lateral sclerosis. Korsakov J. Neuropathol. Psychiatry, p. 1293, 1962. 24. Camu, W., Billiard, M., and Baldy Moulinier, M.: Fasting plasma and CSF amino acid levels in amyotrophic lateral sclerosis: a subtype analysis [see comments]. Acta. Neurol. Scand. 88:51, 1993. 25. Cavanagh, J. B.: The problems of neurons with long axons. Lancet 1:1284, 1984. 26. Chad, D. A., Hammer, K., and Sargent, J.: Slow resolution of multifocal weakness and fasciculation: a reversible motor neuron syndrome. Neurology 36:1260, 1986. 27. Charatan, F.: US links motor neurone disease with Gulf war service. BMJ 324:65, 2002. 28. Charcot, J. M., and Joffroy, A.: Deux cas d’atrophie musculaire progressive avec lesions de la substance grise et des faisceaux antero-lateraux de la moelle epiniere. Arch. Physiol. Neurol. Pathol. 2:744, 1869. 29. Chio, A., Magnani, C., and Schiffer, D.: Amyotrophic lateral sclerosis mortality in Italy, 1958 to 1987: a cross-sectional and cohort study. Neurology 43:927, 1993. 30. Chou, S. M.: Pathology of motor system disorder. In Leigh, P. N., and Swash, M. (eds.): Motor Neuron Disease: Biology and Management. London, Springer-Verlag, p. 53, 1995. 31. Cone, L. A., Nazemi, R., and Cone, M. O.: Reversible ALSlike disorder in HIV infection: an ALS-like syndrome with new HIV infection and complete response to antiretroviral therapy. Neurology 59:474; author reply 59:474, 2002. 32. Corcia, P., Mayeux-Portas, V., Khoris, J., et al.: Abnormal SMN1 gene copy number is a susceptibility factor for amyotrophic lateral sclerosis. Ann. Neurol. 51:243, 2002. 33. Cruveilhier, J.: Sur la paralysie musculaire, progressive, atrophique. Bull. Acad. Med. (Paris) 18:490, 1852. 34. Delisle, M. B., and Carpenter, S.: Neurofibrillary axonal swellings and amyotrophic lateral sclerosis. J. Neurol. Sci. 63:241, 1984. 35. Dhaliwal, G. K., and Grewal, R. P.: Mitochondrial DNA deletion mutation levels are elevated in ALS brains. Neuroreport 11:2507, 2000. 36. Duchenne de Boulogne, G. B. A.: Recherches faites a l’orde des galvanisine sur l’etat de la contractilite et de la sensibilite electromusculaires dans les paralysies des membres superieurs. C. R. Acad. Sci. (Paris) 29:667, 1849. 37. Duchenne de Boulogne, G. B. A.: Paralysie musculaire progressive de la langue, du voile du palais et des levres: affection non encore decrite comme espece morbide distincte. Arch. Gen. Med. 16:283, 1860. 38. Eymard-Pierre, E., Lesca, G., Dollet, S., et al.: Infantile-onset ascending hereditary spastic paralysis is associated with mutations in the alsin gene. Am. J. Hum. Genet. 71:518, 2002. 39. Figlewicz, D. A., Krizus, A., Martinoli, M. G., et al.: Variants of the heavy neurofilament subunit are associated with the development of amyotrophic lateral sclerosis. Hum. Mol. Genet. 3:1757, 1994.

1542

Diseases of the Peripheral Nervous System

40. Gajdusek, D. C., and Salazar, A. M.: Amyotrophic lateral sclerosis and parkinsonian syndromes in high incidence among the Auyu and Jakai people of West New Guinea. Neurology 32:107, 1982. 41. Gibson, S. J., Polak, J. M., Katagiri, T., et al.: A comparison of the distributions of eight peptides in spinal cord from normal controls and cases of motor neurone disease with special reference to Onuf’s nucleus. Brain Res. 474:255, 1988. 42. Giess, R., Beck, M., Goetz, R., et al.: Potential role of LIF as a modifier gene in the pathogenesis of amyotrophic lateral sclerosis. Neurology 54:1003, 2000. 43. Giess, R., Braga, M., Holtmann, B., et al.: A CNTF null mutation as a potential modifier to disease onset in familial ALS with a mutation in the SOD-1 gene and transgenic SOD-1 mutant mice. J. Neurol. 248:41, 2001. 44. Giess, R., Holtmann, B., Braga, M., et al.: Early onset of severe familial amyotrophic lateral sclerosis with a SOD-1 mutation: potential impact of CNTF as a candidate modifier gene. Am. J. Hum. Genet. 70:1277, 2002. 45. Giess, R., Werner, E., Beck, M., et al.: Impaired salivary gland function reveals autonomic dysfunction in amyotrophic lateral sclerosis. J. Neurol. 249:1246, 2002. 46. Hadano, S., Hand, C. K., Osuga, H., et al.: A gene encoding a putative GTPase regulator is mutated in familial amyotrophic lateral sclerosis 2. Nat. Genet. 29:166, 2001. 47. Haddock, R. L., and Santos, J. V.: Are the endemic motor neuron diseases of Guam really disappearing? Southeast Asian J. Trop. Med. Public Health 23:278, 1992. 48. Hafezparast, M., Klocke, R., Ruhrberg, C., et al.: Mutations in dynein link motor neuron degeneration to defects in retrograde transport. Science 300:808, 2003. 49. Hand, C. K., Devon, R. S., Gros-Louis, F., et al.: Mutation screening of the ALS2 gene in sporadic and familial amyotrophic lateral sclerosis. Arch. Neurol. 60:1768, 2003. 50. Harding, A. E., Bradbury, P. G., and Murray, N. M.: Chronic asymmetrical spinal muscular atrophy. J. Neurol. Sci. 59:69, 1983. 51. Hayward, C., Colville, S., Swingler, R. J., and Brock, D. J.: Molecular genetic analysis of the APEX nuclease gene in amyotrophic lateral sclerosis. Neurology 52:1899, 1999. 52. Hirano, A., Donnenfeld, H., Sasaki, S., and Nakano, I.: Fine structural observations of neurofilamentous changes in amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 43:461, 1984. 53. Hirano, A., Kurland, L. T., and Sayre, G. P.: Familial amyotrophic lateral sclerosis: a subgroup characterized by posterior and spinocerebellar tract involvement and hyaline inclusions in the anterior horn cells. Arch. Neurol. 16:232, 1967. 54. Hosler, B. A., Siddique, T., Sapp, P. C., et al.: Linkage of familial amyotrophic lateral sclerosis with frontotemporal dementia to chromosome 9q21-q22. JAMA 284:1664, 2000. 55. Hu, M. T., Ellis, C. M., Al-Chalabi, A., et al.: Flail arm syndrome: a distinctive variant of amyotrophic lateral sclerosis [letter]. J. Neurol. Neurosurg. Psychiatry 65:950, 1998. 56. Huang, K., and Luo, Y.: Adult spinal muscular atrophy: a report of four cases. J. Neurol. Sci. 61:249, 1983. 57. Imura, T., Shimohama, S., Kawamata, J., and Kimura, J.: Genetic variation in the ciliary neurotrophic factor receptor

58.

59.

60.

61.

62.

63.

64. 65.

66. 67.

68.

69.

70.

71.

72.

73. 74.

75.

alpha gene and familial amyotrophic lateral sclerosis [letter]. Ann. Neurol. 43:275, 1998. Jackson, M., Al-Chalabi, A., Enayat, Z. E., et al.: Copper/zinc superoxide dismutase 1 and sporadic amyotrophic lateral sclerosis: analysis of 155 cases and identification of a novel insertion mutation. Ann. Neurol. 42:803, 1997. Jackson, M., Morrison, K. E., Al-Chalabi, A., et al.: Analysis of chromosome 5q13 genes in amyotrophic lateral sclerosis: homozygous NAIP deletion in a sporadic case. Ann. Neurol. 39:796, 1996. Jones, C. T., Swingler, R. J., and Brock, D. J.: Identification of a novel SOD1 mutation in an apparently sporadic amyotrophic lateral sclerosis patient and the detection of Ile113Thr in three others. Hum. Mol. Genet. 3:649, 1994. Jubelt, B.: Motor neuron diseases and viruses: poliovirus, retroviruses, and lymphomas. Curr. Opin. Neurol. Neurosurg. 5:655, 1992. Julien, J. P., and Beaulieu, J. M.: Cytoskeletal abnormalities in amyotrophic lateral sclerosis: beneficial or detrimental effects? J. Neurol. Sci. 180:7, 2000. Kamo, H., Haebara, H., Akiguchi, I., et al.: Peculiar patchy astrocytosis of the precentral cortex in amyotrophic lateral sclerosis [in Japanese]. Rinsho Shinkeigaku 23:974, 1983. Kasarskis, E. J., and Winslow, M.: When did Lou Gehrig’s personal illness begin? Neurology 39:1243, 1989. Katagiri, T., Kuzirai, T., Nihei, K., et al.: Immunocytochemical study of Onuf’s nucleus in amyotrophic lateral sclerosis. Jpn. J. Med. 27:23, 1988. Kojima, H., Furuta, Y., Fujita, M., et al.: Onuf’s motoneuron is resistant to poliovirus. J. Neurol. Sci. 93:85, 1989. Kondo, K.: Epidemiology of motor neuron disease. In Leigh, P. N., and Swash, M. (eds.): Motor Neuron Disease: Biology and Management. London, Springer-Verlag, p. 19, 1995. La Spada, A. R., Wilson, E. M., Lubahn, D. B., et al.: Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352:77, 1991. Lacomblez, L., Bensimon, G., Leigh, P. N., et al.: Doseranging study of riluzole in amyotrophic lateral sclerosis. Lancet 347:1425, 1996. Lambrechts, D., Storkebaum, E., Morimoto, M., et al.: VEGF is a modifier of amyotrophic lateral sclerosis in mice and humans and protects motoneurons against ischemic death. Nat. Genet. 34:383, 2003. Leigh, P. N., Anderton, B. H., Dodson, A., et al.: Ubiquitin deposits in anterior horn cells in motor neurone disease. Neurosci. Lett. 93:197, 1988. Lin, C. L., Bristol, L. A., Jin, L., et al.: Aberrant RNA processing in a neurodegenerative disease: the cause for absent EAAT2, a glutamate transporter, in amyotrophic lateral sclerosis. Neuron 20:589, 1998. Lowe, J.: New pathological findings in amyotrophic lateral sclerosis. J. Neurol. Sci. 124(Suppl.):38, 1994. Lowe, J., Lennox, G., Jefferson, D., et al.: A filamentous inclusion body within anterior horn neurones in motor neurone disease defined by immunocytochemical localisation of ubiquitin. Neurosci. Lett. 94:203, 1988. Mannen, T., Iwata, M., Toyokura, Y., and Nagashima, K.: Preservation of a certain motoneurone group of the sacral cord in amyotrophic lateral sclerosis: its clinical significance. J. Neurol. Neurosurg. Psychiatry 40:464, 1977.

Sporadic Motor Neuron Degeneration 76. Mawrin, C., Kirches, E., and Dietzmann, K.: Single-cell analysis of mtDNA in amyotrophic lateral sclerosis: towards the characterization of individual neurons in neurodegenerative disorders. Pathol. Res. Pract. 199:415, 2003. 77. McGuire, V., Longstreth, W. T. Jr., Nelson, L. M., et al.: Occupational exposures and amyotrophic lateral sclerosis: a population-based case-control study. Am. J. Epidemiol. 145:1076, 1997. 78. McHolm, G. B., Aguilar, M. J., and Norris, F. H.: Lipofuscin in amyotrophic lateral sclerosis. Arch. Neurol. 41:1187, 1984. 79. Meyer, T., Munch, C., Liebau, S., et al.: Splicing of the glutamate transporter EAAT2: a candidate gene of amyotrophic lateral sclerosis [letter]. J. Neurol. Neurosurg. Psychiatry 65:954, 1998. 80. Miller, R. G., Mitchell, J. D., Lyon, M., and Moore, D. H.: Riluzole for amyotrophic lateral sclerosis (ALS)/motor neuron disease (MND) (Cochrane Review). In The Cochrane Library. Chichester, UK, John Wiley, Issue 3, 2004. 81. Moulard, B., Salachas, F., Chassande, B., et al.: Association between centromeric deletions of the SMN gene and sporadic adult-onset lower motor neuron disease. Ann. Neurol. 43:640, 1998. 82. Moulard, B., Sefiani, A., Laamri, A., et al.: Apolipoprotein E genotyping in sporadic amyotrophic lateral sclerosis: evidence for a major influence on the clinical presentation and prognosis. J. Neurol. Sci. 139:34, 1996. 83. Mundt, D. J., Dell, L. D., Luippold, R. S., et al.: Cause-specific mortality among Kelly Air Force Base civilian employees, 1981–2001. J. Occup. Environ. Med. 44:989, 2002. 84. Munoz, D. G., Greene, C., Perl, D. P., and Selkoe, D. J.: Accumulation of phosphorylated neurofilaments in anterior horn motoneurons of amyotrophic lateral sclerosis patients. J. Neuropathol. Exp. Neurol. 47:9, 1988. 85. Nagai, M., Abe, K., Okamoto, K., and Itoyama, Y.: Identification of alternative splicing forms of GLT-1 mRNA in the spinal cord of amyotrophic lateral sclerosis patients. Neurosci. Lett. 244:165, 1998. 86. Nelson, L. M., Matkin, C., Longstreth, W. T. Jr., and McGuire, V.: Population-based case-control study of amyotrophic lateral sclerosis in western Washington State. II. Diet. Am. J. Epidemiol. 151:164, 2000. 87. Nelson, L. M., McGuire, V., Longstreth, W. T. Jr., and Matkin, C.: Population-based case-control study of amyotrophic lateral sclerosis in western Washington State. I. Cigarette smoking and alcohol consumption. Am. J. Epidemiol. 151:156, 2000. 88. Okamoto, K., Hirai, S., Amari, M., et al.: Bunina bodies in amyotrophic lateral sclerosis immunostained with rabbit anti-cystatin C serum. Neurosci. Lett. 162:125, 1993. 89. Olkowski, Z. L.: Mutant AP endonuclease in patients with amyotrophic lateral sclerosis. Neuroreport 9:239, 1998. 90. Orrell, R. W., Habgood, J. J., de Belleroche, J. S., and Lane, R. J. M.: The relationship of spinal muscular atrophy to motor neuron disease: investigation of SMN and NAIP gene deletions in sporadic and familial ALS. J. Neurol. Sci. 145:55, 1997. 91. Orrell, R. W., King, A. W., Lane, R. J., and de Belleroche, J. S.: Investigation of a null mutation of the CNTF gene in familial amyotrophic lateral sclerosis. J. Neurol. Sci. 132:126, 1995.

1543

92. Parboosingh, J. S., Meininger, V., McKenna-Yasek, D., et al.: Deletions causing spinal muscular atrophy do not predispose to amyotrophic lateral sclerosis. Arch. Neurol. 56:710, 1999. 93. Parboosingh, J. S., Rouleau, G. A., Meininger, V., et al.: Absence of mutations in the Mn superoxide dismutase or catalase genes in familial amyotrophic lateral sclerosis. Neuromuscul. Disord. 5:7, 1995. 94. Parry, G. J., and Clarke, S.: Multifocal acquired demyelinating neuropathy masquerading as motor neuron disease. Muscle Nerve 11:103, 1988. 95. Parton, M. J., Broom, W., Andersen, P. M., et al.: D90ASOD1 mediated amyotrophic lateral sclerosis: a single founder for all cases with evidence for a cis-acting disease modifier in the recessive haplotype. Hum. Mutat. 20:473, 2002. 96. Perry, T. L., Krieger, C., Hansen, S. and Eisen, A.: Amyotrophic lateral sclerosis: amino acid levels in plasma and cerebrospinal fluid. Ann. Neurol. 28:12, 1990. 97. Pestronk, A., Cornblath, D. R., Ilyas, A. A., et al.: A treatable multifocal motor neuropathy with antibodies to GM1 ganglioside. Ann. Neurol. 24:73, 1988. 98. Plaitakis, A., and Caroscio, J. T.: Abnormal glutamate metabolism in amyotrophic lateral sclerosis. Ann. Neurol. 22:575, 1987. 99. Puls, I., Jonnakuty, C., LaMonte, B. H., et al.: Mutant dynactin in motor neuron disease. Nat. Genet. 33:455, 2003. 100. Ro, L. S., Lai, S. L., Chen, C. M., and Chen, S. T.: Deleted 4977-bp mitochondrial DNA mutation is associated with sporadic amyotrophic lateral sclerosis: a hospital-based case-control study. Muscle Nerve 28:737, 2003. 101. Rosen, D. R., Siddique, T., Patterson, D., et al.: Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis [see comments]. Nature 362:59, 1993. [Published erratum appears in Nature 364:362, 1993.] 102. Rothstein, J. D., Tsai, G., Kuncl, R. W., et al.: Abnormal excitatory amino acid metabolism in amyotrophic lateral sclerosis [see comments]. Ann. Neurol. 28:18, 1990. 103. Scarmeas, N., Shih, T., Stern, Y., et al.: Premorbid weight, body mass, and varsity athletics in ALS. Neurology 59:773, 2002. 104. Shaw, P. J., Forrest, V., Ince, P. G., et al.: CSF and plasma amino acid levels in motor neuron disease: elevation of CSF glutamate in a subset of patients. Neurodegeneration 4:209, 1995. 105. Shaw, P. J., and Ince, P. G.: Glutamate, excitotoxicity and amyotrophic lateral sclerosis [review]. J. Neurol. 244(Suppl 2): S3, 1997. 106. Siddique, T., Figlewicz, D. A., Pericak Vance, M. A., et al.: Linkage of a gene causing familial amyotrophic lateral sclerosis to chromosome 21 and evidence of genetic-locus heterogeneity [see comments]. N. Engl. J. Med. 324:1381, 1991. [Published errata appear in N. Engl. J. Med. 325:71, 1991, and 325:524, 1991.] 107. Siddique, T., Pericak-Vance, M. A., Caliendo, J., et al.: Lack of association between apolipoprotein E genotype and sporadic amyotrophic lateral sclerosis. Neurogenetics 1:213, 1998. 108. Spiller, W. G.: Primary degeneration of the pyramidal tracts: a study of eight cases with necrophy. Univ. Philadelphia Med. Bull. 17:390, 1904.

1544

Diseases of the Peripheral Nervous System

109. Stone, R.: Guam: deadly disease dying out [news] [see comments]. Science 261:424, 1993. [Published erratum appears in Science 262:640, 1993.] 110. Strickland, D., Smith, S. A., Dolliff, G., et al.: Amyotrophic lateral sclerosis and occupational history: a pilot case-control study. Arch. Neurol. 53:730, 1996. 111. Takahashi, R.: Deficiency of human ciliary neurotrophic factor (CNTF) is not causally related to amyotrophic lateral sclerosis (ALS). Clin. Neurol. 35:1543, 1995. 112. Takahashi, R., Yokoji, H., Misawa, H., et al.: A null mutation in the human CNTF gene is not causally related to neurological diseases. Nat. Genet. 7:79, 1994. 113. Tomblyn, M., Kasarskis, E. J., Xu, Y., and St. Clair, D. K.: Distribution of MnSOD polymorphisms in sporadic ALS patients. J. Mol. Neurosci. 10:65, 1998. 114. Tomkins, J., Banner, S. J., McDermott, C. J., and Shaw, P. J.: Mutation screening of manganese superoxide dismutase in amyotrophic lateral sclerosis. Neuroreport 12:2319, 2001. 115. Tomkins, J., Dempster, S., Banner, S. J., et al.: Screening of AP endonuclease as a candidate gene for amyotrophic lateral sclerosis (ALS). Neuroreport 11:1695, 2000. 116. Tomkins, J., Usher, P., Slade, J. Y., et al.: Novel insertion in the KSP region of the neurofilament heavy gene in amyotrophic lateral sclerosis (ALS). Neuroreport 9:3967, 1998. 117. Toyoshima, Y., Piao, Y. S., Tan, C. F., et al.: Pathological involvement of the motor neuron system and hippocampal formation in motor neuron disease-inclusion dementia. Acta Neuropathol. (Berl.) 106:50, 2003.

118. Traynor, B. J., Codd, M. B., Corr, B., et al.: Incidence and prevalence of ALS in Ireland, 1995–1997: a population-based study. Neurology 52:504, 1999. 119. Van Landeghem, G. F., Tabatabaie, P., Beckman, G., et al.: Manganese-containing superoxide dismutase signal sequence polymorphism associated with sporadic motor neuron disease. Eur. J. Neurol. 6:639, 1999. 120. Veldink, J. H., van den Berg, L. H., Cobben, J. M., et al.: Homozygous deletion of the survival motor neuron 2 gene is a prognostic factor in sporadic ALS. Neurology 56:749, 2001. 121. Weiner, L. P.: Possible role of androgen receptors in amyotrophic lateral sclerosis: a hypothesis. Arch. Neurol. 37:129, 1980. 122. Wohlfart, G., and Swank, R. L.: Pathology of amyotrophic lateral sclerosis: fiber analysis of the ventral roots and pyramidal tracts of the spinal cord. Arch. Neurol. Psychiatry 46:783, 1977. 123. Woodall, C. J., Riding, M. H., Graham, D. I., and Clements, G. B.: Sequences specific for enterovirus detected in spinal cord from patients with motor neurone disease [see comments]. BMJ 308:1541, 1994. 124. Woulfe, J., Kertesz, A., and Munoz, D. G.: Frontotemporal dementia with ubiquitinated cytoplasmic and intranuclear inclusions. Acta Neuropathol. (Berl.) 102:94, 2001. 125. Yang, Y., Hentati, A., Deng, H. X., et al.: The gene encoding alsin, a protein with three guanine-nucleotide exchange factor domains, is mutated in a form of recessive amyotrophic lateral sclerosis. Nat. Genet. 29:160, 2001.

F. Basis of Inherited Disease

64 Mendelian and Mitochondrial Inheritance, Gene Identification, and Clinical Testing VIRGINIA V. MICHELS, STEPHEN N. THIBODEAU, AND ERIK C. THORLAND

Inheritance Patterns Mendelian Inheritance Mitochondrial Inheritance Imprinting

Multifactorial Etiology Genetic Diagnostic Methodologies Fluorescent in Situ Hybridization Genetic Linkage

INHERITANCE PATTERNS Mendelian Inheritance The concept of single-gene inheritance was proposed originally by Mendel in 1865. In 1953, the gene was defined biochemically by Watson and Crick as a portion of the deoxyribonucleic acid (DNA) molecule that codes for the amino acid sequence of a particular polypeptide chain.48 Mendel’s concept of a particulate, genetic determinant resulting in a specific characteristic remains the basis for what is referred to as mendelian inheritance. The nuclear genome contains 3 billion base pairs. In the year 2001, a rough draft of the human genome was published, marking another milestone in human genetics.15 This resulted in a downward revision of the number of genes from 100,000 to approximately 30,000 to 35,000. However, much remains to be learned about this complex genome. Many of the genes, and their functions, still need to be identified, along with their role in development of specific human disease. Details about how genes are regulated and expressed through the basic processes of transcription (DNA to RNA) and translation (messenger RNA [mRNA] to protein) need to be delineated. Some genes

DNA Analysis: Polymerase Chain Reaction

only code for RNA products that function without being translated into proteins, such as the XIST gene that controls inactivation of one X chromosome in females. Some genes code for different protein isoforms that are expressed in different tissues, such as the dystrophin gene with its brain and muscle isoforms. Mutations in certain regions of the dystrophin gene cause classic Duchenne or Becker type muscular dystrophy, while other mutations in the promoter region cause isolated X-linked dilated cardiomyopathy.26 The details of how the relative expression of these isoforms is regulated in different tissues remain to be determined. The study of the function of the protein products of genes and their interactions with each other at the biochemical level to efficiently run cells and whole organisms has spawned a new term, proteomics. Thus the mapping of the human genome is not the end of the quest for understanding our genetic blueprint, it is the beginning of a new era. Along with the complexities being discovered at the molecular level, clinical complexities of mendelian inheritance also have become apparent. Although the basic principles still apply, as discussed in the previous edition of this chapter,27 the following descriptions of the patterns of inheritance focus on these new complexities. 1545

1546

Diseases of the Peripheral Nervous System

Autosomal Dominant Inheritance “Autosomal” indicates that the gene determining a given trait or disease is located on one of the 44 nonsex or autosomal chromosomes. A gene is “dominant” if one copy of that gene is sufficient for the disease or trait to be present. A person having two identical alleles is called a homozygote; a person with nonidentical gene forms at a single locus is a heterozygote. The genetic constitution of an individual with respect to a particular locus or to the complete complement of genes is referred to as the genotype. In contrast, the phenotype is the observable characteristic of an individual as determined (at least in part) by the genotype. In autosomal dominant inheritance, both males and females can be affected and can pass the trait on to their children, with a 50% risk for each child. A negative family history does not exclude the possibility that the disease is autosomal dominant. The disease may arise by new mutation; that is, a change may have taken place in the genetic material as it was being replicated to be passed on in the sperm or egg. The proportion of cases of new mutation varies from one disease to another. For example, 50% of cases of neurofibromatosis type 1 arise by new mutation, whereas 80% of cases of achondroplasia arise by new mutation. Statistically, new mutations occur more frequently in the formation of sperm of older men, so that the mean paternal age of patients with some autosomal dominant diseases is greater than the mean paternal age of the general population. This does not mean, however, that new mutations cannot occur in young individuals of either sex. Other possible explanations for a negative family history in individuals with autosomal dominant diseases include incomplete penetrance and variable expression. “Incomplete penetrance” means that an individual may have a specific gene, but he or she shows no clinical sign of it. “Variable expression” refers to the range of severity that may be present in an individual with an abnormal gene. For example, a person with tuberous sclerosis may have inherited the gene for this disease from a parent who is thought to be healthy. However, if that parent is investigated carefully and is found to have typical tubers on magnetic resonance imaging of the brain, that parent would be said to have mild expression. A third possible explanation for a negative family history in a person with an autosomal dominant disease is germinal mosaicism. In this case, the parent has some cells with the abnormal allele and other cells with the normal allele for a certain gene. If this mosaicism is limited to a minority of somatic cells or to cells in the gonad, the parent will have no phenotypic effects of the mutant gene, but the recurrence risk for future children will be increased. This type of gonadal mosaicism has been implicated in some cases of certain diseases, such as autosomal dominant lethal type II osteogenesis imperfecta.3 In fact, germinal mosaicism occurred frequently enough in families with affected siblings and healthy parents that, for a time, the disease was thought to be autosomal recessive.

Molecular analysis in some of these cases proved otherwise. Because of the possibility of gonadal mosaicism, recurrence risk for lethal type II osteogenesis imperfecta is 4% to 6%. The reason why gonadal mosaicism seems to be particularly common for some new autosomal dominant diseases, such as osteogenesis imperfecta, as compared to others, such as achondroplasia, is unknown. A final mechanism to explain an apparently isolated case of an autosomal dominant disease within a family is the possibility of mistaken paternity. Regardless of the mechanism, it is clear that a negative family history does not exclude the possibility that a disease is inherited as an autosomal dominant disease, and that the family history cannot be assumed to be negative without detailed investigation that should in some cases include physical and laboratory evaluation of family members. The distinction among isolated instances of an autosomal dominant disease on the basis of new mutation versus nonpenetrance in a parent is very important, because in the former case there is little or no increased risk for another child of that parent to be affected, whereas in the latter case the risk is 50%. The risk of an affected individual’s offspring inheriting the gene defect is 50% regardless of whether the disease arose by new mutation or was inherited. If a parent of an affected child shows no signs of the disease but the grandparent is definitely affected, it can be assumed that the parent represents an instance of nonpenetrance. This phenomenon has been referred to as a “skipped generation” and emphasizes the need for taking an extended family history, including grandparents, aunts, uncles, cousins, nephews, and nieces. Some diseases are virtually never nonpenetrant if appropriate examinations are performed (e.g., neurofibromatosis type 1), whereas others, such as von Hippel-Lindau disease, may be nonpenetrant in occasional cases. Age must also be taken into account; it is typical for a young child with von Hippel-Lindau disease to have no manifestations of the disease, with increasing expression with older ages. It always had been presumed that environment and other genetic “modifiers” interact to account for variable expression within and between families, but until recently this could not be shown at the molecular level. Now, some autosomal dominant diseases with incomplete penetrance are known to be influenced by the effect of a nonhomologous gene at another chromosomal locus. For example, Hirschsprung’s disease is heterogeneous at the molecular level, with at least eight different genes reported to cause the disorder. Of these, RET mutations are the most common, and penetrance is incomplete. Neurturin is a RET ligand. In one reported family segregating for Hirschsprung’s disease with a RET mutation, three very severely affected sibs were found to also have a mutation in the neurturin gene, thus demonstrating increased expression and severity when mutations in both genes were present.1,6 This is sometimes referred to as digenic inheritance.

Inheritance, Gene Identification, and Clinical Testing

In another example of molecular complexity that impacts phenotype, some alleles will contain two mutations; that is, there is a double mutation in cis. For example, a case of autosomal dominant supravalvular aortic stenosis caused by an elastin abnormality has been shown to have mutations in both exons 18 and 26 within the same allele that resulted in expression of the disease.44 This type of phenomenon also can occur in autosomal recessive disorders (e.g., cystic fibrosis [CF]). The term codominant inheritance is applied to some traits, such as the MNS blood group, in which the gene coding for the M antigen and the allelic gene coding for the N antigen both may be expressed fully in the same person.50 This term is generally not used when referring to disease. For example, the alleles for sickle hemoglobin and for normal hemoglobin are expressed fully in carriers for sickle cell disease, so that in a certain sense the trait is codominant; however, clinically, the disorder behaves as an autosomal recessive disease. Occasionally, individuals are homozygous for an autosomal dominant gene defect that causes disease even in the heterozygous state; for example, one form of hypercholesterolemia results from deficient low-density lipoprotein receptors. Affected heterozygotes have atherosclerosis at a young age, often less than 50 years. Persons homozygous for this gene die of atherosclerotic disease in childhood or early adolescence.10 Therefore, although the disorder is considered to be dominant clinically, it is apparent that the abnormal gene does not completely override the normal allele. In contrast, patients homozygous for the Huntington’s disease gene do not appear to be more severely affected than do heterozygotes; the disease is truly dominant.30 Trinucleotide repeat expansions result in a special category of diseases that are characterized by an increased number of a particular series of nucleotides (e.g., CAG). These expansions seem particularly prevalent in neurologic diseases such as fragile X-linked mental retardation, Huntington’s disease, many of the autosomal dominant spinocerebellar ataxias, and myotonic dystrophy.39 The DNA is unstable, and the repeat size may either expand or contract as it is transmitted from one generation to the next. Depending on the specific gene, expansion may be more frequent with maternal transmission (myotonic dystrophy) or paternal transmission (Huntington’s disease). The size of the expansion variably correlates with clinical severity, and the disease may become worse over the generations, thereby providing a molecular explanation for the “anticipation” effect seen in certain diseases. The expression of some autosomal dominant diseases is affected by other factors such as anticipation or imprinting (described below). Autosomal Recessive Inheritance Autosomal recessive disorders are coded for by genes located on the nonsex chromosomes. In contrast to autosomal dominant inheritance, the heterozygote, who has

1547

one abnormal allele and one normal allele, does not differ clinically from a person homozygous for the normal gene. Rather, the person must be homozygous for the abnormal allele for the disease or trait to be expressed. In some cases, the person has two abnormal alleles of a certain gene, but each is abnormal in a different way. Such persons are referred to as compound heterozygotes. As described above for autosomal dominant disease, trinucleotide repeat expansions can also be the type of mutation causing autosomal recessive disease, such as Friedreich’s ataxia. For the disease to be present in the offspring, both parents must have one copy of an abnormal allele, and the risk of disease for each of their offspring, of either sex, is 25%. In autosomal recessive inheritance, the previous generations usually are not affected with the disease. Although the classic description of pedigrees for autosomal recessive inheritance includes two or more affected siblings, with today’s small average family size of 2.4 children, it is not unusual for the disease to appear sporadically within the family. One cannot exclude autosomal recessive disease on the basis of a negative family history. In these cases it is sometimes necessary to rely on knowledge of the usual mode of inheritance of the disease. Although some diseases, such as CF, are always inherited in an autosomal recessive pattern, other clinically defined diseases may be inherited in one of several ways. For example, retinitis pigmentosa can be inherited as an autosomal dominant, autosomal recessive, or X-linked recessive disease. Even the same gene can have different mutations that act in a dominant or recessive fashion. For example, both autosomal dominant and autosomal recessive retinitis pigmentosa can be caused by different mutations in the rhodopsin gene.7 Furthermore, different mutations in the same gene can cause different clinical disorders. For example, mutations in the peripherin/RDS gene can cause autosomal dominant retinitis pigmentosa, as well as several types of macular dystrophy.5 For autosomal recessive diseases, the risk for an affected person to have an affected child is low, unless the disease is very common or the affected person marries a blood relative or a person also afflicted with the same autosomal recessive disease. However, even among couples who meet neither of these criteria, the risk is greater than the general population risk. For example, if one assumes that the carrier frequency of the gene for phenylketonuria (PKU) is 1 in 50 in the general population, the risk for healthy parents without a positive family history is 1/50 ⫻ 1/50 ⫻ 1/4 ⫽ 1/10,000. However, if a man has PKU, the risk for his children is 1 ⫻ 1/50 ⫻ 1/2 ⫽ 1/100. The risk for the affected man’s healthy sister to have a child with PKU is 2/3 ⫻ 1/50 ⫻ 1/4 ⫽ 1/300. There are some unusual mechanisms by which autosomal disease may occur, in which only one parent is a carrier for the gene defect. New mutations may occur, as has been documented at the molecular level. For example,

1548

Diseases of the Peripheral Nervous System

a patient with spinal muscular atrophy type I was shown to be homozygous for the common deletion of exons 7 and 8 of the SMN1 gene. The mother was a carrier of the deletion, but the father was not. Nonpaternity was excluded, and it was concluded that the mutation had arisen by new mutation.52 In another type of situation, uniparental disomy for a chromosomal segment with an autosomal recessive gene defect was shown to cause “homoallelic” disease in a patient with a retinal dystrophy.40 Uniparental disomy refers to the inheritance of both of a pair of chromosomes or chromosomal segments from one parent rather than one from each parent. This can occur by various mechanisms, such as trisomic rescue, in which the zygote starts out with trisomy for a given chromosome, but the extra chromosome is lost early in subsequent cell divisions. Therefore, caution is always warranted in making presumptions about carrier status, particularly if prenatal diagnosis may be involved. X-Linked Recessive Inheritance Genes coding for X-linked diseases are located on the X chromosome. Females carrying one abnormal gene and one normal gene can be called heterozygotes, but males have only one X chromosome and are, therefore, referred to as hemizygotes. The Y chromosome does not have most of the genes corresponding to those on the X chromosome, although there are exceptions as described below. X-linked recessive diseases are transmitted by unaffected carrier females; their daughters have a 50% chance of inheriting the gene and being carriers, and their sons have a 50% chance of being affected with the disease. All the children of an affected male will be healthy, but all daughters will be carriers. Thus there can be no male-to-male transmission of X-linked diseases, with the exception of the rare male who also has Klinefelter’s syndrome, with an 47,XXY karyotype. Lyonization is the phenomenon by which, early in embryonic life, many of the loci on one of the X chromosomes of a female become inactivated in each cell. This is a random process so that, in an individual female, for example, 90% of the X chromosomes carrying the normal gene may be inactive, and the person will show signs of the disease. Alternatively, the majority of X chromosomes with the normal gene may remain active, and no clinical or biochemical abnormality may be present. The likelihood of a female being affected with an X-linked recessive disease varies with the specific disease. For example, female carriers of ornithine transcarbamylase deficiency and Fabry’s disease frequently have symptoms, albeit often milder than in males. In contrast, carriers of hemophilia A usually do not have symptoms. This partly reflects differences in the threshold for clinical deficiency based on the biochemical effects of the abnormal or deficient protein product. There are other, less common reasons why females may have an X-linked recessive disease. Women with only one X chromosome (Turner’s syndrome) are hemizygotes and could be affected. A female with an X; autosome transloca-

tion also may be affected by an X-linked recessive disease if the breakpoint is through the site of the gene, thereby damaging the gene. Chromosome abnormalities in females affected with X-linked recessive diseases have been described that result in Duchenne type muscular dystrophy and hypohidrotic ectodermal dysplasia, for example.42 Common diseases such as X-linked color blindness can occur in females because a carrier mother could mate with an affected father. In inbred families, a carrier for a relatively rare disease, such as hemophilia A, may mate with an affected male, and female offspring could be affected. X-linked diseases also can arise by new mutation, and in these cases the family history will be negative. In such instances, knowledge of the usual inheritance pattern of the disease sometimes can be relied upon. For example, Duchenne type muscular dystrophy is inherited as an X-linked recessive disease. Biochemical and molecular testing of the patient and mother can be done in an attempt to determine if the new mutation occurred in the mother or in the affected child. As described previously for autosomal dominant disease, trinucleotide repeat expansions can be the mechanism of mutation in X-linked diseases such as fragile X-linked mental retardation and Kennedy’s disease (spinobulbar muscular atrophy). X-Linked Dominant Inheritance X-linked dominant inheritance is uncommon. The presence of one abnormal gene is sufficient to cause the condition to be expressed in the heterozygote female. However, females, who also have one normal allele, are usually less severely affected than are males. This characteristic does not, in itself, prove that a disorder is X-linked dominant. In some autosomal dominant disorders—for example, in some families with amyloid neuropathy—females can be less severely affected than males.24 The reason for this variation in severity between sexes in some autosomal dominant conditions is unknown but is presumably related to hormonal or other environmental factors. The main clinical criteria for establishing that a disorder is an X-linked dominant condition is that all of the daughters and none of the sons of affected males are affected. For affected females, on average, 50% of the daughters and 50% of the sons are affected. Therefore, in large population studies, the ratio of affected males to females should be 1:3. This ratio can be distorted if males are affected so severely that they die in utero or early in life. In these cases, when the disease is lethal for males, the ratio of children in the families should be two females to one male, with half the females being affected. An interesting example of an X-linked dominant trait is incontinentia pigmenti, which generally is lethal in males. The molecular basis for the disease recently has been shown to be caused by mutations in the NEMO gene. Rare cases of affected males are due to mosaicism or a 47,XXY karyotype. Furthermore, a novel phenotype has been reported in males with a milder NEMO mutation resulting

Inheritance, Gene Identification, and Clinical Testing

in ectodermal dysplasia, immune deficiency, osteosclerosis, and lymphedema.25 The distinction between X-linked dominant and X-linked recessive disorders is not always this clear. As discussed above, in some diseases, such as Fabry’s disease and ornithine transcarbamylase deficiency, the female heterozygote is more frequently affected than in classic X-linked recessive diseases, such as hemophilia A. However, clinical expression of these diseases is not present in all heterozygote females, as is expected for X-linked dominant disorders. Y-Linked Inheritance There are genes on the Y chromosome that are important for testicular development and spermatogenesis. Defects or deletions of these genes can result in decreased fertility in affected males. With newer reproductive technology, such as intracytoplasmic sperm injection into the egg, some of these men can father children and thus transmit the trait to their sons.17 The “pseudoautosomal” regions of the X and Y chromosomes are particularly interesting. These small regions on the short arms of chromosomes X and Y pair during male meiosis, and share some common genetic material that has the opportunity to cross over during pairing. The SHOX gene on Xp causes Leri-Weill dyschondrosteosis, characterized by short stature and bowed radii. Previously, LeriWeill dyschondrosteosis was thought to be an autosomal dominant disorder. It turns out that this gene has a homologue on Yp, and thus can be transmitted from either parent to both sons and daughters.12 Genetic Counseling Genetic counseling should be offered to all families who suffer from mendelian disease. The geneticist can explain the inheritance pattern to the patient, and its implication for relatives, including future children. This information can help the family deal realistically with the risks, and in many cases take proactive measures to lessen the impact of the disease. For example, some people mistakenly believe that, if a parent has the same disorder as they do, all their children will be affected; knowing that each child has a 50% risk of being unaffected in autosomal dominant inheritance can be very encouraging to them. Knowledge about risk for a disease can also empower patients and their families in making appropriate insurance, career, or other lifestyle choices in some cases. The health consequences of some diseases (e.g., hereditary nonpolyposis colon cancer) can be minimized by early surveillance of the colon and polyp removal. In the case of medium chain acyl-coenzyme A (acyl-CoA) dehydrogenase deficiency, avoidance of fasting, a low-fat diet, and carnitine supplementation can prevent metabolic decompensation and death. There are many other such examples in which early diagnosis of affected relatives based on risk counseling can benefit

1549

patients with genetic diseases. Evaluation of patients by a medical geneticist often can be instrumental in making specific diagnoses, utilizing chromosomal and molecular diagnostic techniques as described below. Presymptomatic diagnosis of asymptomatic relatives should be undertaken cautiously, after a thorough discussion of the potential advantages and disadvantages of having this knowledge. This is particularly important for diseases not amenable to treatment. For example, some people may not want to know if they have inherited the gene for early-onset Alzheimer’s disease from an affected parent with a presenilin mutation, whereas others may cope better with the certainty of knowing, regardless of the outcome. Such testing should only be undertaken after a person understands the issues and has had time to carefully think about the options. A visit with a psychiatrist who is knowledgeable about genetic testing issues often can be helpful. Special caution should be used in the case of presymptomatic testing for hereditary diseases in children. It is generally accepted that such testing should not be offered for incurable diseases of adult onset, thereby depriving the child of the chance to decide as a mature adult whether he or she wants this information. Conversely, testing children for disorders such as von Hippel-Lindau disease, for which clinical monitoring must be initiated early in childhood, is a reasonable option. When a patient with a hereditary disease is planning additional children, prenatal diagnosis should be offered when available. Whether or not to use prenatal diagnosis should be based on the patient’s wishes after a discussion of the potential advantages and disadvantages. Prenatal diagnosis for a single-gene disorder for which an enzyme deficiency or specific DNA mutation is known, and for which a clinical test is available, is a standard procedure using amniocytes or chorionic villus cells. Preimplantation diagnosis based on a single cell taken from a blastocyst after in vitro fertilization also is becoming available in some centers.

Mitochondrial Inheritance Mitochondria are the major site of energy metabolism in the cell. Mitochondria are maternally inherited, because the egg contains mitochondria but mitochondria from the sperm are usually actively degraded after fertilization.48 Therefore, transmission is, with rare exception, only through the maternal lineage, such that all of a woman’s children potentially inherit a proportion of mitochondria that contain the mutation. Some of these individuals will be clinically affected, while others may have a normal phenotype. The daughters can transmit the disease to the third generation, but the sons usually cannot. Each mitochondrion contains DNA that replicates relatively independently of nuclear DNA and is transmitted to daughter mitochondria. Human mitochondrial DNA has been completely sequenced and, just as with nuclear DNA, contains normal variations (polymorphisms) that are

1550

Diseases of the Peripheral Nervous System

transmitted from generation to generation. The mitochondrial genome encodes two ribosomal RNAs, 22 transfer RNAs, and 13 subunits of the respiratory chain complexes: seven subunits of NADH oxidoreductase (1, 2, 3, 4, 4L, 5, and 6) of complex I, cytochrome b of complex III, cytochrome c oxidase subunits I through III of complex IV, and ATP synthase subunits 6 and 8 of complex V. It is important to note that these complexes also contain proteins encoded by nuclear DNA, and that mitochondria contain many proteins not related to oxidative phosphorylation that are encoded by nonmitochondrial DNA and that are transported into the mitochondria. Mitochondrial inheritance of disease was first documented to occur in Leber’s hereditary optic neuropathy.47 Many other mitochondrial diseases have been described, including myoclonic epilepsy with ragged red fibers (MERRF); mitochondrial encephalopathy, lactic acidosis, and stroke (MELAS); and neuropathy, ataxia, and retinitis pigmentosa (NARP). These diseases usually result from point mutations or small deletions in mitochondrial DNA. However, Kearns-Sayre syndrome, characterized by progressive external ophthalmoplegia, pigmentary retinopathy, ataxia, and cardiac conduction defects, tends to be associated with larger mitochondrial deletions, and these patients tend not to have carrier mothers.51 Although these classic mitochondrial diseases are well delineated, there is tremendous variability in the clinical phenotypes and in the mitochondrial mutations that can be associated with variable combinations of central nervous system, peripheral nervous system, ocular, hematologic, endocrine, musculoskeletal, cardiac, and renal dysfunction. Many, but not all, of these conditions are associated with lactic acidosis, which may be intermittent or only evident in cerebrospinal fluid. Ragged red fibers in muscle are strongly suggestive of mitochondrial dysfunction, but are not always present in these mitochondrial diseases and are not pathognomonic. Furthermore, there are many disorders of the respiratory chain complex that cannot be identified, either because the relevant tissue (e.g., brain) cannot be analyzed for mitochondrial mutations, or because the defect is encoded by nuclear DNA. Many of the nuclear-encoded DNA respiratory chain subunits and other gene products that affect mitochondrial function remain unidentified. The reasons for variable expression and incomplete penetrance of mitochondrial diseases in terms of severity, age of onset, and organs affected within a family remain unknown, but heteroplasmy is a partial explanation. Heteroplasmy refers to different populations of mitochondria within the same person, such that some mitochondria harbor the deleterious mutation while others do not. Because multiple mitochondria are present in each fertilized egg, it is not surprising that, as mutations occur in individual mitochondria in the human population, a person could inherit two different populations of mitochondria. The proportion of mitochondria with the genetic

defect that is inherited by the individual and the subsequent proliferation of the abnormal population of mitochondria within a given tissue could account for variable symptoms, such as optic neuropathy in some family members and optic neuropathy plus cardiac dysrhythmia in others. Environmental and other genetic factors could contribute further to the clinical variability. Because of heteroplasmy and the inability to precisely predict clinical phenotype, genetic counseling for mitochondrial inherited disorders is difficult. This ability is further compromised when the underlying genetic defect cannot be identified, resulting in inability to distinguish a mitochondrial versus a nuclear-encoded DNA defect. For women with a known mitochondrial DNA mutation, accepting the risks, adoption, and egg donation are options for having children. Prenatal diagnosis using amniocytes or chorionic villus cells is another option. However, data on whether the degree of heteroplasmy (i.e., mutation load) in these cells will predict the mutation load at birth are limited. There is some very preliminary indication from animal studies and a few human cases that it is predictive, but extreme caution must be used, and the patient must fully understand the inherent limitations of this type of prenatal diagnosis. Even if the mutation load determined at prenatal diagnosis is shown in the future to reflect the status at birth, the long-term outcome for these subjects would remain uncertain. There are some published data on risk to future children for women who carry a few of the relatively common mitochondrial mutations, such as the A3243G mutation that causes MELAS.41 It is clear that much more needs to be learned about genetic counseling and prenatal diagnosis for disorders caused by mitochondrial DNA mutations. This knowledge is likely to come by a combination of empirically derived data based on family observations and increased information on how mutant mitochondria are transmitted and replicated.

Imprinting Imprinting refers to the phenomenon of parent-of-origin differences in the expression of certain genes.4 Changes in the gene’s activity, partly based on methylation status, can occur as the gene is transmitted in an egg versus a sperm. Such imprinting plays a role in two uncommon disorders, Prader-Willi and Angelman’s syndromes. Both usually are sporadic in occurrence and sometimes are caused by uniparental disomy or by new mutation chromosome deletions at the same site, 15q11.2-q13. However, PraderWilli syndrome results from the loss of paternal alleles, whereas in Angelman’s syndrome, it is the maternal alleles that are lost.4 Other cases have no chromosome deletion but have mutations in genes within the same chromosomal region that function as an imprinting center. Single-gene disorders also may be influenced by imprinting, such as autosomal dominant familial paragangliomas caused by SDHD

Inheritance, Gene Identification, and Clinical Testing

gene mutations. Paragangliomas can develop when the gene is inherited from the father, but not from the mother.28

Multifactorial Etiology For single-gene patterns of inheritance, a specific gene is the sole or major determinant of a given disease or trait. However, even in mendelian disorders, the remainder of the genetic material, as well as internal and external environmental factors, also influences the phenotypic expression of the gene. Incomplete penetrance, variable expression, variable age of onset, and different severity between sexes, even within the same family, are clues that the gene does not act independently of the other genes or of environmental factors. The role of environmental factors in the expression of mutant genes for neuropathy is particularly evident in acute intermittent porphyria, in which an attack may be precipitated by the use of various medications; in prednisoneresponsive hereditary motor and sensory neuropathy, in which inflammatory-demyelinating mechanisms may be involved in expression; in inherited tendency to brachial palsy, not infrequently expressed at the time of the parturition; in hereditary sensory and autonomic neuropathy; and in hereditary Charcot joints—all of which are complications that relate to use, abuse, neglect, or improper therapy. The situation in which a particular gene is the major determining factor for a given condition can be contrasted with polygenic and multifactorial causation. Although the terms polygenic and multifactorial are frequently used interchangeably, there is an important distinction between them. In polygenic inheritance, the disease or trait is determined only by the cumulative effect of several genes at several loci. In humans, relatively few conditions are purely polygenic, with no environmental component. In contrast, multifactorial conditions are very common. “Multifactorial” means that the disease or trait is determined by the interaction of environmental influences and the polygenic predisposition. Many normal traits, such as height and intelligence, and many birth defects, such as neural tube defects and congenital heart disease, are of multifactorial causation. Diseases such as atherosclerosis, asthma, and some forms of cancer also are believed to be of multifactorial etiology.

GENETICS DIAGNOSTIC METHODOLOGIES Fluorescent in Situ Hybridization Fluorescent in situ hybridization has become a valuable tool for evaluation of submicroscopic chromosome abnormalities. This is a technique utilizing DNA probes homologous to the DNA area of interest within the chromosome, thereby resulting in specific binding to that chromosomal

1551

region. These DNA probes are labeled with fluorophores, to allow microscopic visualization. The probes can be designed to “paint” an entire chromosome—whole-chromosome paint probes, which are helpful in defining complex chromosome rearrangements as can be seen in some malignancies. Whole-chromosome paints are also helpful in defining structural abnormalities that cannot be defined by standard banded karyotyping.16 Alternatively, the probes can be designed to bind to discreet regions of a chromosome, as in the case of Williams syndrome (7q11.2) or velocardiofacial syndrome (22q11.2). The chromosome deletions that cause these syndromes are too small to be seen by standard karyotype, and are therefore referred to as submicroscopic deletions. Less commonly, these probes can be used to identify single-gene defects, such as hereditary neuropathy with liability to pressure palsies, caused by deletion of peripheral myelin protein 22 on 17p11.2 (Fig. 64–1), or PelizaeusMerzbacher disease, caused by duplication of proteolipid protein 1 on Xq22.

Genetic Linkage “Linkage” refers to the nonrandom co-segregation of alleles at different genetic loci based on their close proximity on the same chromosome. As gametes form, homologous chromosomes undergo chromatid exchanges, or crossovers, during meiosis I that result in reciprocal exchanges of genetic material called recombination. On average, one to two crossovers occur between each pair of homologous chromosomes per meiotic division. Linkage is a measure of the frequency of recombination between two loci. The closer together two loci are, the less chance that a recombination event will occur between them. A recombination rate of 50% indicates random segregation of two loci. However, if two loci have less than a 50% recombination rate, they are considered linked. The frequency of recombination between two loci can be used as a measure of the genetic distance separating them. This functional “distance” is measured in centimorgans (cM). A distance of 1 cM equals a 1% chance of recombination occurring between two loci as a chromosome is passed from parent to child. Centimorgans correlate to some degree with a physical distance of 1000 kilobases (kb). However, this relationship is not linear because of differences in recombination potential that occur between males and females and between different regions of the genome. Analyses made possible by the Human Genome Project have demonstrated that the rates of recombination may vary as much as 10-fold when sex-based and regional genomic differences are taken into account.45 Linkage analysis has been used as the first step toward identifying disease genes by positional cloning and for indirect genetic testing in families in which the precise mutation causing a specific disease is unknown. These analyses are made possible by sets of polymorphic DNA

1552

Diseases of the Peripheral Nervous System

HNPP –HEREDITARY NEUROPATHY WITH TENDENCIES TO PRESSURE PALSY Normal metaphase

Deletion of D17S122 (arrow)

PROBE: D17S122(gr)/MPO(red)

FIGURE 64–1 Fluorescent in situ hybridization probe used to detect deletion of peripheral myelin protein 22 in a patient with hereditary neuropathy with liability to pressure palsy. The control probe binds to both chromosomes 17q in both metaphases (red). The D17 S122 probe binds to both chromosomes 17 p in the healthy control but to only one chromosome 17p in the affected patient (green).

markers, variations of DNA sequences that differ between individuals without phenotypic effects. These polymorphisms are distributed throughout the genome, typically occur in noncoding regions of DNA, and are inherited in a mendelian fashion, making them ideal for genetic linkage studies. In the past, these polymorphic markers mainly consisted of restriction fragment length polymorphisms (RFLPs) and variable numbers of tandem repeat polymorphisms. These markers typically were typed by Southern blotting, were not highly informative, and were nonrandomly distributed throughout the genome, making linkage analysis very time consuming and labor intensive (for a more detailed discussion of RFLPs, see Michels and Thibodeau27). More recently, the discovery of microsatellite repeat markers, also known as short tandem repeats (STRs), has made linkage analysis a much more efficient process.49 STRs are tandem repeats of simple sequence that are abundant and randomly distributed throughout the genome. They comprise approximately 3% of the human genome and are found every 2 kb on average.21 STRs are classified by the size and sequence of the repeat. The most commonly used STRs are mono-, di-, tri-, and tetranucleotide repeats. These tandem repeats are thought to arise by slippage events during DNA replication. Slippage events occur often enough that there are usually

multiple alleles (in contrast to the diallelic RFLPs) at any given STR. However, the mutation rate (between 5 ⫻ 10⫺4 and 10⫺5) at these loci is low enough to allow them to be used in linkage studies. In addition to their being highly polymorphic, another major advantage of the use of STRs is that they can be typed very quickly. The polymerase chain reaction (PCR)27 typically is used to amplify a segment of DNA containing the STR of interest, followed by analysis on a polyacrylamide gel to determine the number of repeats on each allele. Recent advances in typing technology have decreased the time necessary for linkage analyses severalfold. Multiplex PCR allows the simultaneous amplification of several STRs in one PCR reaction. Multiplex PCR can be combined with fluorescent labeling of the PCR primers with dyes that fluoresce at different wavelengths when excited by a laser. This allows multiple PCR products of similar size to be run in the same lane on a gel, followed by computer analysis of the output signal to simultaneously differentiate the allele sizes for multiple STRs. The multiple attributes of STRs make them ideal markers for constructing high-resolution genetic maps for the identification of human disease genes. In addition, data provided by the Human Genome Project have allowed the identification of many STRs that are very tightly linked to known

Inheritance, Gene Identification, and Clinical Testing

human disease genes for use in indirect testing in families without known mutations. (For a discussion of the use of markers for linkage to identify disease genes and of the limitations of linkage in studying human families, see Michels and Thibodeau.27) Establishing linkage to a specific chromosomal region is usually the first step in identifying disease genes through a positional cloning approach. Once linkage has been established, many new markers from the region of interest may be used to further narrow the region of interest. Once the candidate region has been sufficiently narrowed (typically ⬍1 cM), linkage data on a genetic map must be transcribed to a physical map of the region. Traditionally, large insert genomic clones such as yeast artificial chromosomes and bacterial artificial chromosomes (BACs) have been used to construct physical maps. Once a physical map of the region defined by linkage is complete, a variety of techniques are used to identify genes in the region that can be tested for mutations in families with the disease. Many disease genes have been identified using the positional cloning approach, such as the genes for Huntington’s disease39 and CF.19,33 However, recent data from the Human Genome Project have significantly streamlined the processes of physical mapping and gene identification. A nearly complete physical map of overlapping BAC clones is now available, and many human genes have been placed on this map. Candidate genes within an interval defined by linkage analysis can now be identified quickly and screened for disease-causing mutations. This “candidate gene” approach is quickly supplanting the more traditional positional cloning. Linkage analysis is also being applied to more complex, multifactorial diseases such as heart disease, diabetes, and schizophrenia. Linkage analysis in these cases requires the study of many more families with a greater number of polymorphic markers, because the contribution of any one gene to a multifactorial disease may be subtle. Overcoming the difficulties in this type of analysis is one of the major goals of modern genetics. The contributions of the Human Genome Project will likely prove to be the catalyst for the identification of the genes influencing the expression of these complex diseases. The human genome map has created exciting new possibilities for identification of genes that predispose to common disorders. It is known that humans share 99.9% of their DNA sequence.36 Thus 0.1% variation in the genome is responsible for the portion of our diversity that is genetically determined. The most common form of this DNA variation is the single-nucleotide polymorphism (SNP), a substitution of one base for another at a specific location. These SNPs may have no effect whatsoever, or they may subtly influence gene function in a way that predisposes to disease. Knowing the location of these SNPs within the genome will provide researchers with a tremendous opportunity to look for correlations

1553

of specific SNPs with disease status. For example, angiotensin receptor variations are believed to relate to the severity of heart failure and response to angiotensinconverting enzyme inhibitors.36 New methodologies, as described below, will facilitate these exciting new opportunities for understanding the genetic predispositions for common multifactorial diseases, as well as for singlegene disorders. Although not used as frequently as in the past, genetic linkage analysis to follow the inheritance of mendelian disorders still plays an important role in predicting disease states in families in which specific disease-causing mutations have not been identified. This indirect testing requires the ascertainment of appropriate family members, generally including individuals who are affected with the disease, and is most often used to determine the carrier status of at-risk family members and for prenatal diagnosis. Indirect testing depends critically on accurate family history and on informative polymorphic markers that reside either within or very near the disease gene. Typing of the appropriate polymorphic markers is performed to determine the allele that co-segregates with the disease state. This allele can then be followed through the family and used for disease prediction (for a specific example, see Michels and Thibodeau27). Recombination events can lead to incorrect results. Thus the accuracy of testing is directly related to the distance separating the polymorphic marker locus and the disease-causing mutation. For tightly linked markers, accuracy can be greater than 99%. However, caveats of this approach include nonpaternity and genetic heterogeneity. Data from the Human Genome Project have contributed greatly in the identification and localization of highly polymorphic markers that can be utilized for linkage analysis in families with mutations in nearly every known disease gene. Furthermore, technologic advances have provided more efficient processes, reducing the overall cost of performing such analyses.

DNA Analysis: Polymerase Chain Reaction The increasing rapidity with which new genes are discovered is due to improving molecular techniques; this and the Human Genome Project have led to the identification of genes that are responsible for a wide variety of human diseases. Some of these discoveries have been incorporated into clinical assays that can be used for the diagnosis and carrier detection of many genetic disorders. These assays require the use of powerful molecular tools for detection of mutations in these genes. The application of novel DNA technologies has permitted rapid progress in the analysis and diagnosis of human disease by the characterization of the underlying gene defects. Although recombinant DNA technology has existed for many years and is still evolving, two of the primary tools still

1554

Diseases of the Peripheral Nervous System

used in DNA diagnostic laboratories are the Southern blot and PCR.27 The Southern blot provides both qualitative and quantitative information about specific regions of the genome. Probes specific for a gene or region of interest are hybridized to genomic DNA digested with restriction enzymes, separated using agarose gel electrophoresis, and transferred to a solid support such as a nylon membrane. By comparing the pattern and intensity of hybridization in patients and appropriate controls, many different types of genetic alterations can be detected. PCR is the most widely used molecular technique today. PCR is an enzymatic method for amplification of DNA with very high specificity for a region of interest. Segments up to several kilobases can be amplified more than 1 million-fold in a single reaction. Each reaction consists of multiple cycles of denaturation of the two strands of DNA at high temperature, annealing of synthetic oligonucleotide primers that provide specificity to the reaction, and synthesis of new DNA by extension from the primers using specialized thermally stable DNA polymerases.27 Once amplified, detection of target sequences and alterations contained within them is greatly simplified. PCR amplification is the starting point for many of the innovative molecular techniques that have been applied to genetic testing in a variety of human diseases (described below). PCR is so widely used in both clinical and research laboratories because it is fast (results can be available within hours), specific (almost any area of interest can be examined), sensitive (single-cell measurements are possible), and cost effective. The objective of both Southern blot and PCR-based analysis in genetic testing is to examine a specific region of DNA for the presence or absence of sequence alterations. Sequence alterations are typically classified into either pathogenic mutations or polymorphisms based on whether they cause abnormal phenotypes or are genetically neutral and/or contribute to normal variation, respectively. Many different types of mutations (point mutations, deletions, insertions, duplications, inversions, and trinucleotide repeat expansions) cause human disease. Point mutations are the most common type of mutation in many genes and may be classified as missense mutations (those that cause one amino acid to be substituted for another), nonsense mutations (those changes that lead to the insertion of a stop codon and premature protein truncation), or mutations in noncoding regions that affect transcription, mRNA splicing, or translation. Deletions and insertions may range in size from single base pairs that change the translational reading frame and cause premature protein truncation to deletions that are megabases in length and encompass entire genes. Duplications can occur via homologous unequal recombination or recombination mediated by repeated DNA sequences. Typically, the duplicated material exists in tandem with the original sequence. Inversions are fairly uncommon and often occur in conjunction with deletions. However, a hot spot of inversion, for example, in

the factor VIII gene causes approximately 40% of the severe cases of the X-linked disorder hemophilia A.20 This inversion is due to recombination events between a copy of a small intronless gene located in intron 22 of the factor VIII gene and two more copies of the gene about 500 kb upstream of factor VIII. The inversion event occurs almost exclusively in male meioses, suggesting that the absence of homologous X chromosome pairing during male meiosis predisposes to intrachromosomal recombination resulting in inversion events.34 As described earlier in the section on Mendelian Inheritance, a recently identified class of mutation involves the expansion of trinucleotide repeat sequences. These repeats are unstable and may undergo large expansions or contractions during meiosis. The expansion of these repeats in or near some genes leads to several diseases, such as myotonic dystrophy2 and fragile X-linked mental retardation syndrome.46 This class of mutation requires special types of genetic analysis because of the dynamic mutation that occurs in these patients. Detection of these many different classes of mutations requires the use of a variety of molecular techniques. The types and distribution of mutation in an individual gene dictate the molecular methods used to detect mutations in affected families. For the purposes of discussion in this chapter, the observed mutations in disease genes commonly fall into three categories, which require three different testing strategies. Analysis of Genes with a Limited Number of Mutations The first category is made up of genes in which one or a small number of mutations cause the majority of disease. Examples of these include Canavan’s disease,18 hereditary hemochromatosis,8 medium chain acyl-CoA dehydrogenase deficiency,11 and factor V Leiden.14 Several approaches can be used to identify specific mutations in the gene of interest, both efficiently and inexpensively. Restriction enzymes have been used in molecular biology for many years. These enzymes recognize and cleave at specific sequences in double-stranded DNA. Mutations that change a single base pair of DNA can create or destroy a restriction enzyme recognition site and can be detected by utilizing the appropriate restriction enzyme. Typically, the segment of DNA containing the mutation is amplified using PCR. Amplification is followed by incubation with the appropriate restriction enzyme, which will differentially cleave the PCR product depending on the presence or absence of the mutation. The digested fragments are visualized by agarose gel electrophoresis and staining with ethidium bromide. In this way, individuals who are homozygous for the normal sequence can be differentiated from individuals who are either heterozygous or homozygous for the mutation based on the migration pattern in the gel. This assay is very specific and reliable, but is relatively time consuming.

1555

Inheritance, Gene Identification, and Clinical Testing

A variation of this theme is allele-specific amplification. Allele-specific PCR is an application of PCR that preferentially amplifies one allele over the other based on single base pair changes in DNA.35 One of the oligonucleotide primers used for the PCR amplification is designed to anneal to either the normal or mutated allele. Any mismatch drastically decreases the efficiency of amplification. The benefit of allele-specific PCR is the speed at which mutation detection can be performed. One of the disadvantages is that heterozygotes cannot be distinguished from homozygotes unless allele-specific reactions are designed for both of the alleles. In addition, an extra set of primers must be included to serve as an internal control in case of failed amplification. The main problem with the examples described above is that a post-PCR analytic step is required. That is, one performs a PCR step to amplify the desired product and a second analytic step to actually detect the mutation present within the PCR product (or amplicon). This requires additional cost in labor and supplies. In addition, further handling of the PCR product increases the potential of PCR contamination within the laboratory. It is desirable to have the ability to simultaneously amplify (via the PCR) and measure the DNA change within a single reaction tube. Because of the tremendous growth in this area, such methodologies are now being realized and, if applicable, are the preferred methods to be used. Some examples of these methods are as follows. Technology for allele determination based on fluorescence resonance energy transfer (FRET) has been widely used recently. FRET is a process that shifts energy from a donor fluorophore to a neighboring acceptor or quenching fluorophore, returning the donor to its ground state without fluorescence emission. The ability of the acceptor to quench fluorescence is highly dependent of the distance between the two molecules. Generally, pairs of fluorophores are used such that the emission spectra of the donor fluorophore overlap the excitation spectra of the acceptor fluorophore. For example, in one method developed to take advantage of the properties of FRET for mutation detection, a standard PCR first is performed to amplify the region containing a specific sequence alteration. Two short oligonucleotide probes, one complementary to each allele, are included in the PCR. Amplification of either or both alleles will allow the complementary probe(s) to hybridize to the target sequence during the annealing/extension phase of the PCR. Both probes are labeled at each end with donor fluorophores, which fluoresce at different wavelengths, and a quencher. When the probes are intact, the donor and quencher are in close proximity and FRET occurs. However, during the extension phase of PCR, when the complementary probes have annealed to the singlestranded PCR products, the probes are cleaved by the 5⬘ nuclease activity of Taq polymerase, releasing the donor

fluorophore from the rest of the probe. The cleavage increases the distance from the quencher, producing an increase in donor fluorescence. After the PCR is complete, allele discrimination is performed by measuring fluorescence at the emission wavelengths of both donor fluorophores and using software to assign the genotype23 (Fig. 64–2). The benefits to FRET-based analyses include the ability to perform both PCR and mutation detection in one tube and the elimination of post-PCR manipulations. Use of these techniques, in conjunction with rapid thermocycling techniques such as capillary PCR, allows powerful, inexpensive, and rapid mutation detection.

D

Q

Q

D

A

G

Reverse

Forward

G C

A D Q A

D

Forward

Q

G C

B

D Q

Forward Taq

G C

C FIGURE 64–2 Schematic representation of FRET-based mutation detection by the 5⬘ nuclease assay. A, PCR primers and probes complementary to both the normal and mutant alleles are included in the PCR reaction. D ⫽ donor fluorophore; Q ⫽ quenching fluorophore; forward ⫽ forward PCR primer; reverse ⫽ reverse PCR primer. B, High-temperature denaturation during PCR separates the DNA strands. Subsequently the temperature is lowered to allow annealing of primers and probes (only one of the strands is depicted). Only the probe with perfect complementarity to the “C” allele anneals to the region containing the mutation. C, During the extension phase of PCR, Taq polymerase extends from the forward primer. When it reaches the annealed fluorescent probe, the 5⬘ nuclease activity of the enzyme cleaves the donor fluorophore from the probe. Separation from the quencher allows fluorescence from the donor fluorophore.

1556

Diseases of the Peripheral Nervous System

Analysis of Genes with a Large Number of Known Mutations Many genes contain a large number of defined diseasecausing mutations. A prime example is mutations in the CFTR gene that cause CF.33 CF is an autosomal recessive disorder characterized by chronic obstructive lung disease and pancreatic enzyme insufficiency. The severity of CF depends on the type and location of the mutation within CFTR.38 The incidence of CF and spectrum of mutation vary markedly among different populations, with over 700 different mutations described.37 However, selective screening of several mutations in specific ethnic populations can detect most mutations. Rapid technology to simultaneously screen for multiple mutations had been a bottleneck in genetic testing for many years. Recently, several novel assays have been developed based on array theory. Arrays for detecting mutations vary from the very simple (the Southern blot was the first array) to the highly complex, in which thousands of oligonucleotides are placed on a solid support. Arrays can be designed to detect specific mutations or all possible changes within a sequence of DNA. The nature of the mutations within a gene dictates the necessary complexity of the array. Several commercial sources provide array assays designed to detect 30 or more specific CF mutations based on a reverse dot-blot principle. Briefly, extracted patient DNA is amplified with biotinylated primers, denatured, and hybridized to oligonucleotide probes that are specific for both the wildtype and mutant sequences, and immobilized on membrane-based strips. After hybridization, streptavidin labeled with alkaline phosphate is bound to the biotinylated products, followed by incubation with a chromogenic substrate. By analyzing the pattern of hybridization, any of the mutations applied to the strip can be detected in either a heterozygous or homozygous state (Fig. 64–3). There are other variations of array-based technology; for example, another method uses a combination of microspheres and fluorescence. Perhaps the most widely known array-based technology uses a variety of chip-based arrays. Arrays allow the genotyping of large sets of known DNA variants using oligonucleotides that are specifically designed to detect individual mutations. These arrays consist of hundreds to thousands of oligonucleotides synthesized at very high density directly on glass slides using photolithography or ink-jet technologies.22 Typically, the sequence containing each mutation is represented by four complementary oligonucleotide probes that differ by a single base at the central position. Patient DNA is amplified by PCR, fluorescently labeled, and hybridized to the glass slide. After the hybridization reaction is complete, the array is inserted into a scanner, where patterns of hybridization are detected. The hybridization data are collected as light emitted from the fluorescent reporter groups,

W-1 M-1 W-2 M-2 W-3 M-3 W-4 M-4 W-5 M-5 W-6 M-6 W-7 M-7

A

B

FIGURE 64–3 Schematic of the reverse dot-blot procedure. A, Stippled regions represent regions of immobilized oligonucleotides specific for either normal or mutant alleles. W ⫽ wild type; M ⫽ mutant. B, After hybridization of amplicons containing the region of interest, the chromogenic substrate indicates the regions of hybridization. In this example, the patient is heterozygous for mutation number 4.

which are bound to the probe array. Probes that most clearly match the target generally produce stronger signals than those that have mismatches. Because the sequence and position of each probe on the array are known, the identity of the target nucleic acid applied to the probe array can be determined. The major advantages to this technology are miniaturization, fluorescentbased detection, and automation of the process. However, while homo- and hemizygous mutations are detected easily, heterozygotes can be difficult to detect because of the intermediate fluorescent signals. Furthermore, at the time of this writing, chips remain prohibitively expensive for routine clinical use.

Inheritance, Gene Identification, and Clinical Testing

Analysis of Genes with a Large Number of Unique Mutations The final category is that of disease genes that contain a large number of unique disease-causing mutations; in other words, each family typically is associated with a private mutation. Many of these genes contain mutations spread throughout the gene, requiring that the whole gene be examined to find the causative mutation. Although direct sequencing of all of the regions of functional significance can be performed for small genes, for larger genes mutation screening techniques are used to determine the local region of the mutation, followed by direct sequencing to find the specific mutation. Familial adenomatous polyposis (FAP) is an example of one of the many diseases that fall in this category. This disorder is caused by mutations in the APC (adenomatous polyposis coli) gene, which is located on chromosome 5.13 Mutations causing FAP can be found anywhere throughout the more than 8-kb coding region of the APC gene. However, greater than 98% of mutations that cause FAP are either frameshift or nonsense mutations leading to the synthesis of a truncated protein. This makes the protein truncation test (PTT) an ideal method for screening for mutations in this gene.32 The PTT method is an assay of open reading frame integrity. The technique involves a combination of reverse transcription–PCR (RT-PCR) and in vitro translation. RT-PCR is a technique used to synthesize double-stranded complementary DNA from mRNA that can be isolated from either blood or tissue samples. Two PCR primers are designed to amplify the region of interest. The upstream primer includes a bacteriophage RNA polymerase promoter sequence and translation initiation codon (ATG) engineered 5⬘ to the gene-specific primer sequence. After RT-PCR, the end product is a double-stranded DNA template suitable for an in vitro coupled transcription/ translation reaction in which the amplified products are transcribed into RNA, followed by translation of the RNA into protein. Products from the transcription/translation reaction are analyzed by sodium dodecyl sulfate– polyacrylamide gel electrophoresis (PAGE). In samples in which premature termination occurred, the protein product will be smaller than that generated from the nonmutant sequence. Assessment of the size difference between these two products allows the position of the putative mutation to be located, followed by direct sequencing to determine the precise mutation (Fig. 64–4). This technique is advantageous because it is sensitive and can be used to screen very large segments of a gene for mutations. However, the PTT has disadvantages, including the need to use RNA, which is much less stable than DNA, as the starting material in most instances. In addition, PTT will not detect missense or small in-frame insertion/deletion mutations, and mutations may be missed if the mutant allele is not amplified as a result of polymorphisms under the primers or nonsense-mediated mRNA decay. Thus the

1557

technique is ideal for the study of some disease-causing genes but does not apply to others. Several DNA-based screening techniques are also available to screen portions of genes for mutations. These include single-strand conformation polymorphism (SSCP), conformation-sensitive gel electrophoresis (CSGE), denaturing high-performance liquid chromatography (dHPLC), and others. Some examples are described here. SSCP is a technique for detecting mutations in small (200– to 250–base pair) PCR products.31 Following amplification, the double-stranded PCR product is denatured to yield single-stranded DNA. Each strand assumes a specific folded conformation based on internal base pairing, which in turn depends on its sequence and conditions such as temperature and ionic strength. These conformers are separated by PAGE under nondenaturing conditions. A single base mutation can cause a change in the folded conformation compared to the wild type and change the separation pattern, indicating a sequence alteration. To detect the maximum number of mutations, it is common to run the samples under at least two different conditions. In this way, 70% to 90% of unknown mutations can be detected relatively quickly. Samples that demonstrate a SSCP-based shift are then sequenced to determine the precise mutation. The main drawback to this technique is that samples should be analyzed on multiple PAGE gels to maximize the detection rate, and even then a substantial portion may not be detected. CSGE is another PCR- and PAGE-based technique used to detect single-base mismatches in DNA.9 Unlike SSCP, CSGE examines double-stranded DNA rather than singlestranded PCR products. The basis of CSGE is the detection of heteroduplexes that form when one strand of wild-type and one strand of mutated DNA are present. PCR-amplified products are denatured at high temperature and then allowed to slowly reanneal. If a sequence alteration is present on one allele, a portion of the renatured PCR products will form heteroduplexes. When these heteroduplexes are run on a PAGE gel with mildly denaturing solvents, the tendency of single-base mismatches to produce conformational changes such as bends in the double helix increases the differential migration of DNA heteroduplexes and homoduplexes during gel electrophoresis. Samples with altered migration patterns are directly sequenced to determine the precise sequence alteration (Fig. 64–5). Advantages to this technique are the larger size of PCR products that can be analyzed (up to 450 base pairs), the potential to run multiplexed PCR reactions in the same lane on the gel, the need for a single gel system in most instances, and a high mutation detection rate (⬎90%). Disadvantages include the reduced mutation detection rate when sequence alterations occur near the ends of the PCR products and inability to detect homozygous mutations because only homoduplexes are formed. In this case it is necessary to mix patient and normal DNA to form detectable heteroduplexes.

1558

Diseases of the Peripheral Nervous System

FIGURE 64–4 The protein truncation test. A, Schematic of the PTT. The 5⬘ overhang on the downstream PCR primer contains a bacteriophage RNA polymerase promoter (T7) and a transcription initiation sequence (TIS). B, PTT analysis of DNA from several FAP patients. The germline band is present in all samples along with truncated products of various sizes (lanes 1 through 5).

Finally, dHPLC is an ion-pair reversed-phase highperformance liquid chromatography method that also depends on PCR amplification and heteroduplex formation.43 dHPLC relies on differential denaturing profiles of heteroduplexes separated on an HPLC column containing

alkylated nonporous particles. Under conditions of partial heat denaturation within a linear acetonitrile gradient, heteroduplexes display reduced column retention time relative to homoduplexes. Samples that demonstrate altered elution profiles are followed up with direct sequencing.

FIGURE 64–5 CSGE analysis demonstrating altered bands resulting from sequence changes. Three different PCR products, each containing an exon (exons A through C), are combined and loaded together on a PAGE gel. Arrows indicate the expected point of migration. Sequence changes in exon A (lanes 4, 5, and 6), exon B (lane 7), and exon C (lanes 1, 2, 3, 4, and 6) are represented by altered bands.

Inheritance, Gene Identification, and Clinical Testing

The major advantages of this method include the use of automated instrumentation, the speed of analysis (approximately 5 minutes per sample), the size of the DNA fragment that can be analyzed (up to 1.5 kb), and the sensitivity of mutation detection, which approaches 100%. However, the technique requires extensive optimization of conditions for each PCR segment and a high initial investment for the machine. It is apparent that no one technique is suitable for all diseases because of the different types of underlying mutations. The clinician who relies on laboratory diagnostics must be aware that each technique has limitations and none is completely sensitive. The clinician must still make an accurate clinical diagnosis and use the result of the molecular analysis in concert with, not apart from, clinical findings. However, the list of genetic disorders that can be approached with linkage or DNA analysis is growing rapidly. As the Human Genome Project is completed and the technology to analyze the vast amounts of new genetic data progresses, the ability to solve previously unapproachable genetic problems will greatly increase. The ability to rapidly assay large regions of DNA for sequence alterations and the ability to understand the nature of multifactorial diseases will drive genetics in the future.

REFERENCES 1. Amiel, J., and Lyonnet, S.: Hirschsprung disease, associated syndromes, and genetics: a review. J. Med. Genet. 38:729, 2001. 2. Brook, J. D., McCurrach, M. E., Harley, H. G., et al.: Molecular basis of myotonic dystrophy: expansion of a trinucleotide (CTG) repeat at the 3⬘ end of a transcript encoding a protein kinase family member. Cell 68:799, 1992. 3. Byers, P. H.: Osteogenesis imperfecta: an update. Growth Genet. Horm. 4:1, 1988. 4. Cassidy, S., Dykens, E., and Williams, C.: Prader-Willi and Angelman syndromes: sister imprinted disorders. Am. J. Med. Genet. 97:136, 2000. 5. Cremers, F., van de Pol, D., van Driel, M., et al.: Autosomal recessive retinitis pigmentosa and cone-rod dystrophy caused by splice site mutations in the Stargardt’s disease gene ABCR. Hum. Mol. Genet. 7:355, 1998. 6. Doray, B., Salomon, R., Amiel, J., et al.: Mutation of the RET ligand, neurturin, supports multigenic inheritance in Hirschsprung disease. Hum. Mol. Genet. 7:1449, 1998. 7. Dryja, T., and Li, T.: Molecular genetics of retinitis pigmentosa. Hum. Mol. Genet. 4:1739, 1995. 8. Feder, J. N., Gnirke, A., Thomas, W., et al.: A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis. Nat. Genet. 13:399, 1996. 9. Ganguly, A., Rock, M. J., and Prockop, D. J.: Conformationsensitive gel electrophoresis for rapid detection of singlebase differences in double-stranded PCR products and DNA fragments: evidence for solvent-induced bends in DNA heteroduplexes. Proc. Natl. Acad. Sci. U. S. A. 90:10325, 1993.

1559

10. Goldstein, J. L., and Brown, M. S.: The LDL receptor locus and the genetics of familial hypercholesterolemia. Annu. Rev. Genet. 13:259, 1979. 11. Gregersen, N., Winter, V., Curtis, D., et al.: Medium-chain acyl-CoA dehydrogenase (MCAD) deficiency: the prevalent mutation G985 (K304E) is subject to a strong founder effect from Northwestern Europe. Hum. Hered. 43:242, 1993. 12. Grigelioniene, G., Schoumans, J., Neumeyer, L., et al.: Analysis of short stature homeobox-containing gene (SHOX) and auxological phenotype in dyschondrosteosis and isolated Madelung deformity. Hum. Genet. 109:551, 2001. 13. Groden, J., Thliveris, A., Samowitz, W., et al.: Identification and characterization of the familial adenomatous polyposis coli gene. Cell 66:589, 1991. 14. Grody, W., Griffin, J., Taylor, A., et al.: American College of Medical Genetics consensus statement on Factor V Leiden mutation testing. Genet. Med. 3:139, 2001. 15. International Human Genome Sequencing Consortium: Initial sequencing and analysis of the human genome. Nature 409:860, 2001. 16. Jalal, S., Law, M., and Dewald, G.: Atlas of whole chromosome paint probes: normal patterns and utility for abnormal cases [monograph]. Rochester, MN, Mayo Foundation, p. 1, 1996. 17. Kamp, C., Hirschmann, P., Voss, H., et al.: Two long homologous retroviral sequence blocks in proximal Yq11 cause AZFa microdeletions as a result of intrachromosomal recombination events. Hum. Mol. Genet. 9:2563, 2000. 18. Kaul, R., Gao, G., and Aloya, M.: Canavan disease: mutations among Jewish and non-Jewish patients. Am. J. Hum. Genet. 55:34, 1994. 19. Kerem, B., Rommens, J. M., Buchanan, J. A., et al.: Identification of the cystic fibrosis gene: genetic analysis. Science 245:1073, 1989. 20. Lakich, D., Kazazian, H. H. Jr., Antonarakis, S. E., and Gitschier, J.: Inversions disrupting the factor VIII gene are a common cause of severe haemophilia A. Nat. Genet. 5:236, 1993. 21. Lander, E. S., Linton, L. M., Birren, B., et al.: Initial sequencing and analysis of the human genome. Nature 409:860, 2001. 22. Lipshutz, R. J., Fodor, S. P., Gingeras, T. R., and Lockhart, D. J.: High density synthetic oligonucleotide arrays. Nat. Genet. 21:20, 1999. 23. Livak, K. J.: Allelic discrimination using fluorogenic probes and the 5⬘ nuclease assay. Genet. Anal. 14:143, 1999. 24. Mahloudji, M., Teasdall, R. D., Adamkiewics, J. J., et al.: The genetic amyloidosis: with particular reference to hereditary neuropathic amyloidosis, type II (Indiana or Rukavina type). Medicine (Baltimore) 48:1, 1969. 25. Mansour, S., Woffendin, H., Mitton, S., et al.: Incontinentia pigmenti in a surviving male is accompanied by hypohidrotic ectodermal dysplasia and recurrent infection. Am. J. Med. Genet. 99:172, 2001. 26. Mestroni, L., Krajinovic, M., Severini, G., et al.: Molecular genetics of dilated cardiomyopathy. Herz 19:97, 1994. 27. Michels, V. V., and Thibodeau, S. N.: Mendelian inheritance, genetic linkage, and molecular diagnosis. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1015, 1993.

1560

Diseases of the Peripheral Nervous System

28. Milunsky, J., Maher, T., Michels, V., and Milunsky, A.: Novel mutations and the emergence of a common mutation in the SDHD gene causing familial paraganglioma. Am. J. Med. Genet. 100:311, 2001. 29. Munnich, A., and Rustin, P.: Clinical spectrum and diagnosis of mitochondrial disorders. Am. J. Med. Genet. 106:4, 2001. 30. Myers, R. M., Lumelsky, N., Farrer, L. A., et al.: Homozygote for Huntington disease. Am. J. Hum. Genet. 45:615, 1989. 31. Orita, M., Iwahana, H., Kanazawa, H., et al.: Detection of polymorphisms of human DNA by gel electrophoresis as single-strand conformation polymorphisms. Proc. Natl. Acad. Sci. U. S. A. 86:2766, 1989. 32. Powell, S. M., Petersen, G. M., Krush, A. J., et al.: Molecular diagnosis of familial adenomatous polyposis. N. Engl. J. Med. 329:1982, 1993. 33. Riordan, J. R., Rommens, J. M., Kerem, B., et al.: Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science 245:1066, 1989. 34. Rossiter, J. P., Young, M., Kimberland, M. L., et al.: Factor VIII gene inversions causing severe hemophilia A originate almost exclusively in male germ cells. Hum. Mol. Genet. 3:1035, 1994. 35. Ruano, G., and Kidd, K. K.: Direct haplotyping of chromosomal segments from multiple heterozygotes via allele-specific PCR amplification. Nucleic Acids Res. 17:8392, 1989. 36. Subramanian, G., Adams, M., Venter, J., and Broder, S.: Implications of the human genome for understanding human biology and medicine. JAMA 286:2296, 2001. 37. The Cystic Fibrosis Genetic Analysis Consortium: Population variation of common cystic fibrosis mutations. Hum. Mutat. 4:167, 1994. 38. The Cystic Fibrosis Genotype-Phenotype Consortium: Correlation between genotype and phenotype in patients with cystic fibrosis. N. Engl. J. Med. 329:1308, 1993. 39. The Huntington’s Disease Collaborative Research Group: A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington’s disease chromosomes. Cell 72:971, 1993. 40. Thompson, D., McHenry, C., Li, Y., et al.: Retinal dystrophy due to paternal isodisomy for chromosome 1 or chromosome

41.

42.

43.

44.

45. 46.

47.

48. 49.

50.

51.

52.

2, with homoallelism for mutations in RPE65 or MERTK, respectively. Am. J. Hum. Genet. 70:224, 2002. Thorburn, D., and Dahl, H.: Mitochondrial disorders: genetics, counseling, prenatal diagnosis and reproductive options. Am. J. Med. Genet. 106:102, 2001. Turleau, C., Niaudet, P., Cabanis M. O., et al.: X-linked hypohidrotic ectodermal dysplasia and t9C;12 in a female. Clin. Genet. 35:462, 1989. Underhill, P. A., Jin, L., Lin, A. A., et al.: Detection of numerous Y chromosome biallelic polymorphisms by denaturing high-performance liquid chromatography. Genome Res. 7:996, 1997. Urbán, Z., Zhang, J., Davis, E., et al.: Supravalvular aortic stenosis: genetic and molecular dissection of a complex mutation in the elastin gene. Hum. Genet. 109:512, 2001. Venter, J. C., Adams, M. D., Myers, E. W., et al.: The sequence of the human genome. Science 291:1304, 2001. Verkerk, A. J., Pieretti, M., Sutcliffe, J. S., et al.: Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome. Cell 65:905, 1991. Wallace, D. C., Singh, G., Lott, M.T., et al.: Mitochondrial DNA mutation associated with Leber’s hereditary optic nerve neuropathy. Science 242:1427, 1988. Watson, J. D., and Crick, F. H. C.: The structure of DNA. Cold Spring Harbor Symp. Quant. Biol. 18:123, 1953. Weber, J. L., and May, P. E.: Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am. J. Hum. Genet. 44:388, 1989. Wiener, A. S., di Diego, N., and Sokol, S.: Studies on the heredity of the human blood groups. I. The M-N types. Acta Genet. Med. Gemellol. (Rome) 2:391, 1953. Wong, L.: Recognition of mitochondrial DNA deletion syndrome with non-neuromuscular multisystemic manifestation. Genet. Med. 3:399, 2001. Zeesman, S., Whelan, D., Carson, N., et al.: Parents of children with spinal muscular atrophy are not obligate carriers: carrier testing is important for reproductive decision-making. Am. J. Med. Genet. 107:247, 2002.

65 Transgenic Models of Inherited Neuropathy ANDREA ROBERTSON AND CLARE HUXLEY

Peripheral Myelin Protein 22 PMP22 Overexpression Inducible PMP22-Overexpressing Model The CMT Rat as a Model for Progesterone Therapy in CMT1A Pmp22 Deficiency Point Mutations Intracellular Trafficking Abnormalities in Pmp22 Disorders

Myelin Protein Zero Mpz-Overexpressing Mice Mpz-Null Mutant Mice Mpz Epitope–Tagged, Toxic Gain-of-Function Model Early Growth Response Gene 2 Egr2 (Krox20)-Null Mice Transcriptional Activation of Late Myelin Genes by Egr2

Failure to assemble or maintain peripheral nerve myelin is the defining characteristic of a heterogeneous group of inherited peripheral neuropathies. Collectively they are termed Charcot-Marie-Tooth disease (CMT), but they are also known as hereditary motor and sensory neuropathy. This chapter details the animal models that have been generated to study the proteins involved in CMT. It also illustrates what these models have added to our understanding of the disease mechanisms. Differences between the models have been useful in determining the contributions of individual proteins, whereas the phenotypic similarities serve to illustrate the complex functional interactions among the proteins.

PERIPHERAL MYELIN PROTEIN 22 Peripheral myelin protein 22 (PMP22) is produced predominantly by myelinating Schwann cells in the peripheral nervous system (PNS).111 It localizes largely in compact myelin, where it plays a structural role in maintaining myelin integrity. However, PMP22 is also expressed in several non-neuronal tissues during embryonic development and in adulthood.10,54,61,82,112,131 Its expression in these tissues demonstrates that it has a more general role in cell biology. In addition to its role as a specialized myelin

Connexin 32 Cx32: A Gap Junction Protein Cx32-Deficient Mice Cx32 Mutant Mice Periaxin Prx-Deficient Mice

component, PMP22 also functions in cell proliferation and differentiation.76

PMP22 Overexpression Charcot-Marie-Tooth disease type 1A (CMT1A) is the most common inherited peripheral neuropathy. It results from the duplication of a region of chromosome 17 that includes the PMP22 gene.59,89 The generation of several PMP22-overexpressing animal models has demonstrated a dose-dependent worsening of peripheral nerve phenotype in response to increased gene dosage.39,60,95,106 Not surprisingly, the number of copies of the PMP22 gene is not the main determinant of phenotype but rather the relative level of introduced versus endogenous messenger RNA (mRNA) (Table 65–1). For the purposes of this chapter, overexpressing models are grouped based upon expression levels. An inducible model, wherein PMP22 overexpression can be switched on and off by the administration of antibiotics, is discussed separately. Behavior Low-Level Overexpression. Mouse strains with a low number of additional copies of the PMP22 gene (C2 and C61) show no signs of gait abnormality. They cannot be visually distinguished from unaffected littermates, even at 18 months (authors’ personal observation). When subjected 1561

1562

Diseases of the Peripheral Nervous System

Table 65–1. PMP22-Overexpressing Animal Models Species Mouse39,40 C2 hom C61 het C22 het Rat106 CMT rat het hom Mouse60 TgN248 Mouse95 My41 Mouse84 Inducible

Additional Gene

Copy Number

Relative Level of Overexpression

MNCV (m/s) (Agematched Controls)

Human

2 4 7

0.8 1.0 1.6

41 25 4 (38)

Mouse

3 6

1–4

14.7 (36.4 ⫾ 2.5)

Mouse

16

2.0

2⫾1 (41 ⫾ 10)

Mouse

Not done

Not done

Not done

Mouse

Not done

~5.0

34.5 ⫾ 3.3 (44.1 ⫾ 2.9)

to more detailed behavioral profiling, young adult C61 mice (2 months) showed generally impaired motor performance (SHIRPA protocol).75 In particular, they showed impaired muscle strength and a deficit in coordination and balance (grip strength and Rotorod tests, respectively).75 In contrast, heterozygous rats, with an equally low number of additional copies of Pmp22, show obvious physical impairment. CMT rats have an unsteady gait and move clumsily with their hind feet splayed outward.106 Like the low-copy mice, they also have impaired motor performance and show reduced muscle strength.106 High-Level Overexpression. These mice/rats have multiple copies of the PMP22/Pmp22 gene and high levels of the introduced mRNA (see Table 65–1: C22, My41, and TgN248 mice and homozygous CMT rat). Gait abnormalities in the mouse models become apparent in the first 2 to 3 weeks after birth and worsen progressively with age.39,60,95 These mice generally have weak hind limbs, an unstable gait, and muscle atrophy that becomes increasingly evident over time (Fig. 65–1). Motor tests show impairments in hindfoot positioning, grip strength, and reflex reactions and a coordination and balance deficit.75 Qualitatively, the deficits are similar to those seen in low-copy mice (C61) but more severe and more variable.75 In addition to motor impairment, the C22 strain also showed some evidence of sensory deficits (toe pinching withdrawal reflex). However, the strong motor component inherent in this test makes the results difficult to interpret.75 C22 animals have a normal lifespan, and Tg248 animals are reported to generally survive to 8 months. However, My41 mice do not normally survive beyond 5 months. All

high-copy-number animals breed poorly, presumably as a result of generalized muscle weakness.39,60,95 Homozygous CMT rats (six copies) never develop a coordinated gait and are paralyzed by 4 weeks. At this point many die or have to be sacrificed.106 Occasionally spasticity and seizures are also seen.106 Histology Low-Level Overexpression. No myelin abnormalities are seen in the nerves of C2 animals at any age.40 In C61 mice, myelin initially formed normally but it was in some

FIGURE 65–1 A 6-month-old C22 mouse showing characteristic posture of the hind limbs. (From Huxley, C., Passage, E., Manson, A., et al.: Construction of a mouse model of Charcot-Marie-Tooth disease type 1A by pronuclear injection of human YAC DNA. Hum. Mol. Genet. 5:563, 1996, with permission.)

Transgenic Models of Inherited Neuropathy

way unstable and could not be maintained in the population of larger axons. Demyelinated fibers were first seen around postnatal day (P) 28.93 By 7 to 9 months there was a small population of fibers (3%) that were either demyelinated or recently remyelinated.95 Myelin loss only affected the large fiber population (Fig. 65–2A).92,95 In the heterozygous CMT rat many axons in peripheral and cranial nerves had thin or absent myelin sheaths.106 Within the same nerve branch, motor fibers were more severely affected than sensory ones.106 As found in C61 animals, hypomyelination was more marked in larger caliber fibers. Abnormalities were more frequently seen in the ventral roots compared with the dorsal roots and sciatic and tibial nerves. Neither axonal degeneration nor active demyelination were features of the pathology in the CMT rat. Neuronal cell bodies appeared normal. Schwann cells had begun to form onion bulbs in 2.5-month-old rats but they were not well developed. By 6 months onion bulb formation was extensive.106

1563

High-Level Overexpression. A completely different type of myelin deficit was seen in lines expressing high levels of PMP22 (C22, My41, and TgN248 mice and homozygous CMT rat) (Fig. 65–2B and C). Instead of progressive demyelination, the initiation of myelination was delayed or in many fibers did not occur at all.39,60,95 Very little myelin debris was detected in these models, suggesting that the absence of myelin is a result not of demyelination but of myelination failure (dysmyelination). In homozygous overexpressing rats, the most severely affected overexpressing animals, peripheral nerves are severely dysmyelinated.106 Myelination is not merely delayed but fails completely, representing a stable block of myelination or arrest of differentiation. This rat model has been used to demonstrate the uncoupling of the two processes of myelin assembly and differentiation.74 At the molecular level, genes such as myelin protein zero (Mpz) and myelin basic protein (Mbp) are expressed abundantly despite the absence of myelin. Arrested Schwann cells also

FIGURE 65–2 A–D, Cross sections through the tibial nerves of 18-month-old mice. E, Dysmyelinated axons incompletely surrounded by Schwann cell cytoplasm in the My41 mouse. (From Robertson, A., Perea, J., Tolmachova, T., et al.: Effects of mouse strain, position of integration and tetracycline analogue on the tetracycline conditional system in transgenic mice. Gene 282:65, 2002, with permission.) See Color Plate

1564

Diseases of the Peripheral Nervous System

have a novel molecular phenotype in which both late and early differentiation markers are expressed simultaneously.74 This suggests that the Schwann cells of overexpressing rats differentiate molecularly but that Pmp22 overexpression blocks the late steps of myelin assembly.74 In mice overexpressing high levels of PMP22 (C22, My41, and TgN248), a small population of fibers are able to form some myelin. However, the sheath is disproportionately thin relative to axon size. Overexpressing mouse and rat Schwann cells successfully establish a one-to-one relationship with their axons but fail to elaborate myelin.60 This indicates that PMP22 is not involved in the initial separation of axons from the axon bundles. Muscle Pathology High-Level Overexpression. Muscle pathology in 16-copy TgN248 mice (10 weeks old) consisted of extensive fiber type grouping indicative of denervation and reinnervation.60 At more advanced stages definite signs of neurogenic atrophy were seen. These included large areas containing mainly atrophic fibers intermingled with smaller groups of hypertrophic muscle fibers. Despite the high number of atrophic fibers, no muscle fiber degeneration was seen.60 Electrophysiology In overexpressing mice, motor nerve conduction velocities (MNCVs) decrease as PMP22 copy number and mRNA level increase (see Table 65–1).40,60 Reduced nerve conduction velocities are generally accompanied (where responses could be obtained) by increased distal motor and F-wave latencies.40 In some severely affected animals no response could be elicited. In these animals muscle denervation was confirmed by the presence of spontaneous fibrillation potentials and myotonic runs on needle movement.40 In CMT rats conduction velocities were decreased by approximately 50% in both motor and sensory nerves.106 There was minimal morphologic evidence of axonal degeneration, but abundant fibrillation potentials and positive sharp waves suggested partial muscle denervation.106

Inducible PMP22-Overexpressing Model The generation of an inducible model has established that the demyelination resulting from PMP22 overexpression is reversible.84 Expression of the transgenic PMP22 gene in this model occurs in the absence of tetracycline. Overexpression can then be turned off by administering tetracycline in the drinking water. Affected animals (never fed tetracycline) have a slightly abnormal gait and mildly reduced MNCVs.84 Twenty percent to 40% of fibers in the sciatic nerve have thin or absent myelin and are associated with an increased number of Schwann cells. Nerves do not show any evidence of axon loss or signs of axonal atrophy. Given the fivefold overexpression

of mRNA in their nerves, the phenotype of these animals is relatively mild. This is probably a result of uneven mRNA distribution caused by the variegated distribution of the transgene, which leaves the majority of fibers unaffected (Fig. 65–3A). When overexpression was prevented (turned off by tetracycline administration), correction of myelination began within 1 week and the phenotype was largely restored to normal by 6 weeks (Fig. 65–3B).84 This demonstrates that Schwann cells are not permanently damaged by long-term overexpression of PMP22. Instead they are capable of myelinating axons as soon as overexpression ceases. Upregulation of the PMP22 transgene caused active demyelination within 1 week (Fig. 65–3C), and a population of demyelinated fibers was seen by 8 weeks. Thus, although demyelination can largely be corrected within a few months by downregulating PMP22, any treatment would need to be continuous because mature myelin is sensitive to upregulation of PMP22.84

The CMT Rat as a Model for Progesterone Therapy in CMT1A Steroid hormones are epigenetic regulators of gene expression. Treatment with neuroactive steroids stimulates the expression of Mpz and PMP22.64 In aged wild-type rats this reduces the incidence of age-associated morphologic abnormalities.64 The phenotype of CMT rats can also be modulated by progesterone treatment.107 Over a 7-week period, progesterone treatment upregulated Pmp22 mRNA by approximately 30%.107 This exacerbated the phenotype of the CMT rats, which results from overexpression of Pmp22. Treatment resulted in an increased number of axons without myelin and thinning of the myelin sheath.107 Treatment with onapristone, a progesterone receptor agonist, resulted in a 15% decrease in Pmp22 mRNA. The number of demyelinated fibers and myelin thickness was unaltered in onapristonetreated rats, possibly because of the high level of interindividual variation. However, treatment appeared to slow the disease progression by decreasing the rate of axon loss.107 A functional improvement in motor performance was seen after 5 weeks of treatment, and this was subsequently maintained. This study suggests an antiprogesterone strategy as a possible treatment for CMT1A.

Pmp22 Deficiency Most patients with hereditary neuropathy with liability to pressure palsies (HNPP) have a 1.5-Mb deletion of chromosome 17p11.2-p12. This deletion is reciprocal to the duplication found in CMT1A patients. Both disorders result from altered dosage of the PMP22 gene (CMT1A patients having three copies and HNPP patients one copy), which is contained within the duplicated/deleted

Transgenic Models of Inherited Neuropathy

1565

FIGURE 65–3 The inducible PMP22 model. Electron micrographs of cross sections through the sciatic nerve. A, Affected: a 16-week-old mouse not fed with tetracycline (overexpressing PMP22). Arrows indicate fibers with no myelin. Asterisks indicate fibers with very thin myelin. Arrowheads indicate fibers that are unaffected because of variegation of the transgenes. Scale bar: 2 ␮m. B, Correcting, 2 weeks: mice not fed with tetracycline to 8 weeks of age (overexpressing PMP22) and then fed with tetracycline for 2 weeks (overexpression prevented). Almost all singly ensheathed axons have a myelin sheath. Arrows indicate recently remyelinated fibers with very thin myelin. Scale bar: 5 ␮m. C, Demyelinating, 1 week: mice fed with tetracycline for 8 weeks (overexpression prevented, normally myelinated) then not fed with tetracycline for 1 week (overexpressing). A macrophage containing myelin debris (asterisk) indicative of active demyelination is seen beside a recently demyelinated axon with myelin debris in the Schwann cell cytoplasm (arrow). Scale bar: 1 ␮m. (From Perea, J., Robertson, A., Tolmachova, T., et al.: Induced myelination and demyelination in a conditional mouse model of Charcot-Marie-Tooth disease type 1A. Hum. Mol. Genet. 10:1007, 2001, with permission.)

region. Clinically, HNPP is manifest as acute or recurrent transient muscle palsies and paresthesias (abnormal burning or prickling sensations). Symptoms are often due to minor injuries or pressure being applied to the nerve. The pathologic hallmark of HNPP is the presence of multiple focal myelin thickenings along single nerve fibers. These prominent structures consist of redundant myelin loops called tomacula, named after their appearance in teased fiber preparations (tomaculum ⫽ “sausage” in Latin).

Generation of Mice with No or Reduced Pmp22 Adlkofer et al. generated mice with a mutation disrupting the first translated exon of the Pmp22 gene.3 No protein was detected in the nerves of the resulting homozygous animals (PMP22(Ald)⫺/⫺).3 A second group of investigators used pronuclear injection of a Pmp22 antisense construct to create transgenic founder mice (PMP22(May)hom).63 The nerves of these

1566

Diseases of the Peripheral Nervous System

mice showed a modest reduction in mRNA level: 15% in homozygotes and 9% in heterozygotes.63 Behavior Homozygous Pmp22-null mutant mice (PMP22(Ald)⫺/⫺) have walking difficulties resulting from progressive paralysis of the hind limbs.3 Homozygous animals can be clearly identified as early as 2 weeks after birth, whereas heterozygous mice (PMP22(Ald)⫺/⫹) cannot usually be distinguished from wild-type littermates.3 However, occasional heterozygous animals showed signs of walking difficulties.3 In some homozygous animals a mild tremor was seen. Stress-induced convulsions were reported3 and have subsequently been found to be neuromyotonia of peripheral nerve origin.121 Mice with moderately reduced Pmp22 levels (PMP22 (May)hom) have a unique and bizarre gait abnormality in which the stepping motion is grossly exaggerated.63 These authors reported that homozygous mice were “reluctant to move . . . and tended to explore the environment with both forelimbs, radially from a single point.” The gait disturbance worsened progressively with age and resulted in severe distal limb weakness in older mice. As a consequence of muscle weakness, the animals’ hind limbs collapsed when they were lifted by the tail.63 Mild intention tremors were sometimes seen.63 The high-stepping and exaggerated gait has similarities to the gait seen in humans with a peripheral neuropathy that has a strong proprioceptive component. However, this aspect of the mouse model has not yet been investigated. Histology Myelin formation is delayed in homozygous null mutant mice (PMP22(Ald)⫺/⫺). At P4 they have 66% fewer myelinated fibers than controls.3 Even at this early stage, myelin sheaths are already thickened compared with those seen in wild-type animals. Thickened myelin sheaths are also seen in the nerves of HNPP patients, suggesting a role for PMP22 in the determination of myelin thickness. Tomacula are the most prominent morphologic feature in the nerves of Pmp22-deficient mice (PMP22(Ald)⫺/⫺). Young animals (24 days) have a tomaculum associated with every axon/Schwann cell unit. However, by 10 weeks tomacula have been replaced by established signs of myelin degeneration, including thinly myelinated axons and onion bulbs.3 Tomacula are seen in a subset of peripheral nerve biopsies but their significance is not clear. Studies on these animals provided the first evidence of what tomacula represent, that is, an early and transient sign of unstable myelin that will subsequently degenerate and result in a demyelination phenotype.3 In these mice, as in most murine models of myelin disorders, myelination was more severely affected in the ventral than the dorsal roots.99 In the L3 ventral roots of older mice, most large axons were dysmyelinated/demyelinated.

In contrast, the dorsal root had underdeveloped onion bulbs, hypermyelination, and tomacula (indicative of earlier stages or less severe pathology). A milder phenotype was found in heterozygous mice (PMP22(Ald)⫺/⫹) but the features were similar. Commensurate with the small reduction in Pmp22 mRNA levels, pathology in the nerves of homozygous PMP22(May)hom animals was modest.63 There were only small histologic differences between them and control animals, most easily seen in older (7- to 9-month) animals. Pathology included the presence of tomacula, myelin breakdown products, onion bulbs, and almost completely demyelinated fibers. No abnormalities were seen in heterozygous PMP22(May)het mice (9% reduction in Pmp22 mRNA). PMP22(May)hom mice illustrate the sensitivity of the myelination process to altered Pmp22 dosage. A relatively small (15%) reduction in Pmp22 mRNA results in bizarre gait abnormalities and a mild, relatively late-onset neuropathy. Myelination thus appears less sensitive to overexpression than it does to reduced Pmp22 expression (see Table 65–1). Electrophysiology PMP22(Ald)⫺/⫺ animals have electrophysiologic abnormalities typical of a severe myelination disorder. These include decreased MNCVs, polyphasia of compound action potentials, and increased motor latencies.3 In comparison, conduction velocity is normal in PMP22(Ald)⫺/⫹ mice, indicating that near-normal conduction is maintained in the presence of tomacula.3 The only alteration found in 3- to 5-month-old PMP22(May)hom mice was increased F-wave latency, representing conduction slowing and/or partial conduction block in the fastest conducting fibers.63 By 7 to 9 months, distal caudal and distal sural nerve compound latencies were also prolonged.63

Point Mutations Several mouse strains with point mutations in the Pmp22 gene are currently available (Table 65–2). Some have arisen spontaneously,21,109,113 and others have resulted from largescale mutagenesis.44,45 The severity of the resulting phenotype varies depending on the specific mutation. In the nerves of mildly affected heterozygous Trembler-J (TrJ) mice, approximately 80% of fibers have an intact myelin sheath.94 At the other end of the spectrum, the Schwann cells of Trembler-m1H (Tr-m1H) mice fail completely to form myelin.44,45 Behavior Trembler. The behavioral features of the Trembler (Tr) mouse have been described by a number of investigators.8,21,36 Affected animals develop a tremor and gait

1567

Transgenic Models of Inherited Neuropathy

Table 65–2. Animal Models with Pmp22 Point Mutations Name

Allele Symbol

Origin

Mutation

Severity of Phenotype

Trembler

Tr

Spontaneous

Heterozygous Homozygous

Trembler-J

TrJ

Spontaneous

Trembler-Ncnp

Tr-Ncnp

Spontaneous

Trembler-m1H

Tr-m1H

Mutagenesis

Trembler-m2H

Tr-m2H

Mutagenesis

Trembler-m3H

Tr-m3H

Mutagenesis

G150A114 Transmembrane domain 4 L16P114 Transmembrane domain 1 Deletion113 Region containing the whole of exon 4 H12R44 Transmembrane domain 1 Y153TER44 Premature stop codon at aa 153; last seven aa of the Pmp22 protein are missing S72T45 Transmembrane domain 2 in humans

Human Counterpart ⴙⴙ ⴙⴙ

Heterozygous Homozygous (lethal)

ⴙ ⴙⴙⴙⴙ

Heterozygous Homozygous

ⴙⴙ ⴙⴙⴙⴙ ⴙⴙⴙⴙ

G150A43 G150C41 L16P38,125

H12Q126

ⴙⴙⴙ

ⴙⴙ

S72P124 S72W124 S72L91 Possible mutation hotspot

aa ⫽ Amino acids.

abnormalities between P10 and P14. The tremor is rapid and generalized and persists throughout the animal’s life. It is most obvious in the head and neck and disappears when the animal is asleep. The gait abnormality consists of extension at the knees and ankles and circumduction of the hind limbs at the hip. From an early age, the limbs of Tr animals are held extended, making their gait awkward. Their flexed joints are mobile but their limbs are weak. Neuromyotonia is seen in young animals. The episodes (first reported as seizures) are frequent initially but decrease with age.21,121 Homozygous Tr mice cannot be behaviorally distinguished from heterozygous ones. Trembler-J. Heterozygous (TrJ/⫹) animals are less severely affected than Tr mice. They cannot be told apart from their littermates until P20 to P25, whereas Tr animals can be recognized around P10. TrJ/⫹ animals, like Tr, have a tremor but it is smaller in amplitude. The gait abnormality in TrJ/⫹ animals is mild initially, progresses with age, and is of a similar nature to that seen in Tr mice.36 Homozygous TrJ/TrJ mice are the most severely affected mice with a Pmp22 point mutation. By P10 they have severely dystonic, ineffectual movements and struggle

to right themselves.36 Homozygous TrJ/TrJ mice generally die around P21. Myotonic seizures are seen in young homozygous TrJ/TrJ mice when they are handled. However, they occur rarely in adult TrJ/⫹ mice and then only when they are under anesthesia (A. R., personal observation). Trembler-Ncnp. Affected Trembler-Ncnp (Tr-Ncnp) animals have a gait abnormality that becomes apparent between P15 and P20 and progresses slightly with age. The gait abnormality is followed by motor and sensory ataxia that remains throughout the animal’s life.113 Most affected mice survive for more than 1 year. Homozygotes cannot be behaviorally distinguished from heterozygotes.113 Trembler-m1H, -m2H, and -m3H. Tr-m1H animals are the most severely affected of the mutagenesis-induced lines. They exhibit a number of functional deficits indicating impairment of muscle and motor function.45 Behavioral phenotype is characterized by severe tremor with seizures (presumably neuromyotonic) and a gait abnormality associated with low body position and splayed legs.45 When held by the tail, they exhibit vigorous extension

1568

Diseases of the Peripheral Nervous System

and flexion of the hind limbs. Grip, truncal, and limb strengths (wire test) are reduced and motor coordination (Rotorod test) is severely impaired. No sensory deficit is evident.44 The Trembler-m2H (Tr-m2H) animal is similar in phenotype to the Tr-m1H but the deficit is generally milder with less severe tremors and no seizures.44,45 In the Trembler-m3H (Tr-m3H) animal, the deficits seen are similar in nature but less severe than those seen in Tr-m2H and Tr-m1H.45 A resting tremor is observed from 3 weeks of age. The phenotypes of these mice were all assessed using the SHIRPA protocol. This is a standardized series of tests designed to evaluate animal models with neurologic deficits (see www.mgu.har.mrc.ac.uk/mutabase/shirpa_ summary.html). Histology Trembler. On initial inspection, Tr nerves lack myelin and have shortened internodes, increased amounts of endoneurial and perineurial connective tissue, and increased Schwann cell density.8,36,55,56 The myelin deficit becomes progressively more pronounced with age, with large myelinated fibers being preferentially affected.56 Most axons in Tr nerves do not have a myelin sheath and are instead surrounded only by two to three turns of uncompacted Schwann cell membrane.8 On the few fibers where myelin has been formed, it is commonly uncompacted, usually approaching the node of Ranvier but also located randomly along the internode.8 Myelinated fibers also often have abnormal internal and external mesaxons.8 In older Tr animals, many fibers are surrounded by concentric layers of basement membranes (basal laminal onion bulbs). A few fibers are surrounded by layers of Schwann cell processes (classic onion bulbs) that are suggestive of ongoing cycles of demyelination and remyelination.8,9,55–57 The density of Schwann cells is increased in Tr animals at all ages.8,36,55,56 Although increased Schwann cell density is a common feature of demyelinated nerves, it usually remains stable or decreases from early postnatal values.93 However, in Tr nerves Schwann cell density continues to increase throughout the animal’s life,85 possibly because of the presence of myelin breakdown products (known Schwann cell mitogens).8,36,56 There is currently no evidence of axon loss in Tr sciatic nerves.5 The number of motor neurons in the ventral horn is unaltered, and both density and diameter distribution of dorsal ganglion cells appear normal.55 Initial morphologic studies failed to discern any phenotypic differences between heterozygous and homozygous mice. This is in contrast to the TrJ mutant, in which two different classes of affected animals were obvious. As a result, Tr was referred to as a dominant mutation and TrJ as semidominant. This distinction is misleading because both mutations are dominantly inherited and dose

dependent. Subsequent studies have demonstrated different levels of myelination in heterozygous and homozygous Tr nerves. Homozygous Tr nerves have virtually no myelin, but in heterozygous animals 28% of appropriately sized axons are myelinated, albeit thinly.37 In an elegant series of nerve grafting experiments, the Tr mutation was confirmed as a primary lesion of the Schwann cell.5 Regenerating axons from control mice reproduced the features of Tr nerve as they grew through the Tr graft and vice versa. Manifestations of the Tr mutation were fully expressed by the transplanted Schwann cells. The cervical sympathetic trunk (CST) contains axons that are almost entirely unmyelinated, and in Tr mice they appear phenotypically normal. When Schwann cells from the CST are transplanted into the sural nerve of normal animals, they show typical Tr morphology. These data suggest that the Tr mutation only affects Schwann cells when they are challenged to myelinate.86 However, recent findings of ultrastructural abnormality of unmyelinated fibers in PMP22-overexpressing animals and TrJ mice suggest that closer investigation may be needed.95 Trembler-J. Axon separation from fetal bundles occurs normally in both heterozygous and homozygous nerves, but myelination is delayed. As early as P4 the nerves from homozygous animals can easily be distinguished from controls by the markedly decreased number of myelinated fibers (~2% of singly ensheathed axons in TrJ/TrJ vs. ~ 50% in controls).93 The nerves of homozygous TrJ/TrJ animals are the most severely affected of the point mutation mutants. Only a few dozen myelinated fibers are present in a nerve that would normally contain thousands.36 The abnormalities in the nerves are qualitatively similar to those in other severely affected Pmp22 mutant animals but more pervasive.95 Homozygous animals die around 3 weeks of age, but the cause of their death is not clear. Their peripheral nerves, although severely affected, do not appear worse than those of My41 animals, which usually live to be 5 months old. Myelination in TrJ/TrJ mice is delayed at a characteristic stage, not seen to date in other mutants. During normal myelination, Schwann cell cytoplasm surrounds the axon and the opposing tongues contact prior to overlapping, forming a mesaxon and beginning to spiral around the axon. In a proportion of TrJ/TrJ promyelin fibers, this contact is not made and the axon remains partially surrounded by Schwann cell cytoplasm (Fig. 65–4A).93 Abnormal spiraling around the axon is also seen in the overexpressing C22 line, but it has slightly different features. In C22 nerve the initial overlap is made but myelination appears to be delayed at this stage (Fig. 65–4B).93 These observations of delayed and abnormal myelin development suggest that PMP22 has a role in the initiation of myelination at a stage after the association of Schwann cells with the axons, and specifically in the process of mesaxon formation.

Transgenic Models of Inherited Neuropathy

1569

FIGURE 65–4 A, Fibers from a homozygous TrJ mouse in which the Schwann cell cytoplasm fails to complete a full turn around the axon. Scale bar (applies to A and B): 1 ␮m. B, Fibers from a C22 mouse in which the Schwann cell cytoplasm has completed a full turn around the axon but fails to progress further. (From Robertson, A. M., Huxley, C., King, R. H. M., et al.: Development of early postnatal peripheral nerve abnormalities in Trembler-J and PMP22 transgenic mice. J. Anat. 195:331, 1999, with permission.)

Myelination is also delayed in heterozygous TrJ/⫹ animals, but it is not as obvious visually as in homozygotes (35% of singly ensheathed axons are myelinated). The nerves of adult heterozygous animals can be identified under the light microscope by the presence of many fibers with irregular contours of Schwann cell cytoplasm. The nerves also have small, thinly myelinated axons, increased amounts of extracellular matrix, and increased numbers of Schwann cells.36,78,95 Myelination is most severely affected in the ventral root, but both spinal roots are more affected than the sciatic nerve.94 Pathology in TrJ/⫹ nerves is qualitatively similar but less severe than in Tr nerves.36 Thin myelin sheaths are found on virtually all large fibers in older TrJ/⫹ animals, whereas in Tr animals large fibers are usually unmyelinated.36,95 There are virtually no classic onion bulbs, but basal laminal onion bulbs are commonly seen, particularly in the spinal roots.94 Schwann cell numbers in TrJ/⫹ nerves do not decrease with the onset of myelination as they do in control animals. Neither do they continue to increase, as seen in Tr nerves. Instead their numbers remain relatively static throughout the animal’s life.93 There was no axon loss in the sciatic nerve of 7- to 9-month-old TrJ/⫹ animals.95 However, there were reduced axon numbers in the distal tibial (30% reduction) and tail (75% reduction) nerves of aged mice (18 months).49 Despite having much less demyelination, the amount of axon loss in TrJ/⫹ nerves is the same as in transgenic mice overexpressing PMP22.49,95 This suggests that demyelination does not directly result in axon loss and that PMP22 may have a alternative role in the maintenance of axonal integrity. Unmyelinated fibers in TrJ/⫹ mice have been reported as appearing normal.36 However, recent ultrastructural examination has shown that unmyelinated axons are often surrounded by an unusual excess of Schwann cell mem-

branes (Fig. 65–5).95 This excess of membranes is also found in overexpressing mice.95 Abnormalities of unmyelinated fibers have not previously been reported in the literature, and their significance is unclear. Membranes were also seen spiraling around collagen pockets and so could indicate a role of PMP22 in Schwann cell/axon recognition.95 Several abnormalities relating to Schwann cell–axon interactions were noted in TrJ/⫹ nerve. Terminal loops of the myelin sheath were occasionally found turning outward into the extracellular space instead of terminating on the axon.94 In a small proportion of TrJ/⫹ fibers, Schwann cell nuclei, normally located at the center of the internode, were seen very close to one end of the internode, often over terminal loops.94 Cytoplasm has been seen unusually distributed along the internode; it was concentrated around the nucleus with an attenuated layer extending along either side toward the nodes of Ranvier.94 Exuberant protrusions of Schwann cell cytoplasm into the endoneurium were also a feature of TrJ/⫹ nerves.94 These protrusions have subsequently also been found in C22 and My41 mice overexpressing PMP22.93,95 Trembler-Ncnp. The Tr-Ncnp mutant allele has an inframe deletion that results in the production of a truncated Pmp22 protein (113 amino acids).113 The deletion includes exon 4, which encodes transmembrane domain 2 and part of transmembrane domain 3. In many respects pathology in Tr-Ncnp mice is similar to that seen in Tr and TrJ. All three mutations result in dominantly inherited disorders with myelin deficits. However, Tr-Ncnp had some notably different features from the other point mutation models, particularly in homozygous animals.113 Homozygous Tr-Ncnp animals completely fail to elaborate myelin, with the result that most fibers remain in the promyelin state.113 Electron microscope studies showed

1570

Diseases of the Peripheral Nervous System

FIGURE 65–5 Unmyelinated fiber bundles (Remak fibers). A, Control mouse. B, C22 mouse. Extensive layers of Schwann cell wrapping around Remak fibers. C, TrJ mouse. A few layers of Schwann cell wrapping around Remak fibers. D, Layers of Schwann cell membrane seen wrapping around collagen pockets in a TrJ mouse. (From Robertson, A., Perea, J., Tolmachova, T., et al.: Effects of mouse strain, position of integration and tetracycline analogue on the tetracycline conditional system in transgenic mice. Gene 282:65, 2002, with permission.)

many swollen vacuolar structures near the Golgi apparatus. The vacuoles were generated not from the Golgi apparatus but from swelling of the rough endoplasmic reticulum (ER) (Fig. 65–6).113 Large vacuoles were usually electron translucent but occasionally contained closely packed membrane complexes. These vacuoles were also present in heterozygous animals, although less obviously. The nerves of homozygous mice contained an increased number of apoptotic Schwann cells (7.9%).113 Fibers in the peripheral nerves of heterozygous Tr-Ncnp mice have thin or no myelin.113 In these animals the myelin deficit is variable even within a single internode. Myelin compaction appears normal. The density of collagen is increased and basal lamina is seen irregularly surrounding the axons. Pathologic alterations were not detected in the central nervous system (CNS). The abnormal protein encoded by the Tr-Ncnp allele has a toxic gain of function. The vacuoles seen in the

nerves of affected animals are likely to be the result of impaired intracellular trafficking of the mutant protein. Retention of mutant Pmp22 appears to result in overload of the ER that ultimately leads to Schwann cell death. Trembler-m1H, -m2H, and -m3H. All three mutageninduced models have hypomyelinated peripheral nerves but no evidence of abnormality in brain, spinal cord, or muscle.44 The number of myelinated fibers is decreased and the existent myelin is abnormally thin. The density of Schwann cells and the amount of endoneurial connective tissue is increased in these nerves. Tr-m1H is the most severely affected of the three lines, closely followed by Tr-m2H (both around 90% decrease in myelinated fibers), with Tr-m3H (~30% decrease) being relatively mildly affected. Myelinated fiber and axon diameters are decreased in parallel with the overall degree of severity.44

Transgenic Models of Inherited Neuropathy

1571

Pmp22 Expression Messenger RNA. The Tr mutation was found to affect the expression of all the myelin protein genes examined.12 As expected, Pmp22 was the most affected, with mRNA levels in heterozygous animals decreased by 90% compared with controls and decreased by 95% in homozygotes. Mpz and Mbp mRNAs were also expressed at lower levels than in control nerves.

FIGURE 65–6 Transverse section through a sciatic nerve fiber from a 5-month-old homozygous Tr-Ncnp mouse. Giant vacuoles (v) in the Schwann cell are thought to be generated from rough endoplasmic reticulum (ER) because they are frequently continuous with the limiting membranes of the rough ER near the Golgi apparatus and with the inner and outer membranes of the nuclear envelope. The axon (ax) has no myelin. The Schwann cell has a heterochromatin-rich nucleus with a dark cytoplasm that contains a number of free polysomes, Golgi apparatus (g), and mitochondria. Scale bar: 1.0 ␮m. (From Suh, J.-G., Ichihara, N., Saigoh, K., et al.: An in-frame deletion in peripheral myelin protein-22 gene causes hypomyelination and cell death in Schwann cells in the new TREMBLER mutant mice. Neuroscience, 79:735, 1997, with permission.)

Electrophysiology Trembler mice have decreased MNCVs, decreased evoked muscle action potential amplitudes, and increased distal motor latencies.58 In normal animals, MNCVs increase from birth until 4 months. In Tr mice MNCVs are less than 10 m/s as early as P6 and do not increase with age.58 Muscle Pathology In young Tr animals, myofibrils are mildly disrupted. Disruption progresses into complete fiber disorganization with advancing age.26 Deep, slow-twitch muscles (predominantly fiber type I) are more severely affected than the more superficial, fast-twitch ones (predominantly type II).26 However, all the peripheral nerves appear similarly affected. This disparity is likely to result from the different types of in vivo impulses received by the two muscle types. Slow-twitch fibers are normally subjected to prolonged low-frequency stimulation, whereas fast-twitch fibers are stimulated for short periods at high frequency. Muscle motor end plates form normally but subsequently develop ultraterminal sprouting, which worsens until the structure of the end plates becomes abnormal.26

Protein. The amount of Pmp22, Mpz, and Mbp protein was found to be reduced in Tr nerves.12,27,127 This would be expected because the volume of myelin in their nerves is significantly decreased. Myelin-associated glycoprotein (Mag) staining was increased in Tr nerve, which is consistent with its localization in noncompact Schwann cell membranes.42 When immunolabeling density was adjusted for myelin surface area, Mpz and Mbp were unchanged in large myelinated fibers but reduced by 65% and 80%, respectively, in thinly myelinated fibers. Pmp22, Mpz, and Mbp were undetectable in the nerves of homozygous TrJ mice and severely reduced in heterozygotes.78 In young heterozygous mice (P18), Pmp22 expression appeared to be reduced relative to Mag or Mpz.78

Intracellular Trafficking Abnormalities in PMP22 Disorders Protein misfolding and trafficking can result in pathology through different mechanisms. Disease may arise because the function usually performed by a given protein is lost (e.g., cystic fibrosis). Alternatively, a mutant protein may disrupt the function of the wild-type protein (dominant negative) or even be toxic to the cell (gain of function). Trafficking Abnormalities in Pmp22 Mutants Both Tr and TrJ mutant proteins sequester wild-type Pmp22 into abnormal intracellular locations. Trembler-associated protein remains localized in the endoplasmic reticulum,68,69 whereas TrJ-associated protein is found largely in the intermediate compartment, a collection of vesicles involved in cargo transport from the ER to the Golgi apparatus.119 Wild-type Pmp22 is known to form dimers and possibly larger oligomers in the secretory pathway.96,119 In cells expressing both normal and mutant (L16P) Pmp22, mixed protein oligomers (heterodimers) are formed.119 Consequently, a proportion of wild-type protein is retained in the intermediate compartment along with the mutant protein. This results in a reduced amount of wild-type protein being tracked to the cell surface.119 Decreased levels of cell surface protein expression have been found in seven of eight disease-associated PMP22 point mutations. These seven mutations have resulted in less than 10% of the wildtype protein values being detected at the cell surface.17,68,69 Altered protein trafficking has been suggested as a common pathologic mechanism in disease-causing PMP22

1572

Diseases of the Peripheral Nervous System

mutations.69,100,118 Many authors have suggested that disease-associated missense mutations act as loss-of-function mutants because they retain wild-type protein. However, in both humans and mice, point mutations result in a greater degree of dysfunction than a missing allele.3,63,128 In these cases, the mutant protein is not merely failing to function but must be actively resulting in disease. Crossbreeding experiments have confirmed that Tr acts as a true gain-of-function mutation.4 Mice carrying a Tr allele (Tr/⫹) have an increased incidence of myelin abnormalities compared with mice that have a null allele (Pmp22null/⫹). Further evidence for a true gain-of-function mechanism is that the myelin deficit worsens as Tr gene dosage is increased (Tr/Tr vs. Tr/null). Protein Aggregation in Pmp22 Mutant Mice Abnormally folded proteins are subject to a quality control mechanism in the ER or Golgi apparatus. When the degradation capacity of the cell is overwhelmed, by overexpression of misfolded proteins and/or inhibition of the proteasome, intracellular aggregates form. 97 These microtubule-dependent inclusions have been termed aggresomes.28,48,77 The significance of aggresomes is unclear. Because they are found associated with several disease states, some authors have suggested that they may be linked to pathogenesis.29,51 Aggresomes are found associated with cellular degeneration in many neurodegenerative diseases (e.g., Huntington’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis).122 Other authors have postulated that aggresomes may have a protective function in preventing aggregated protein from interfering with cellular function by removing it from the cytosol. Protein accumulation into aggresomes is suggested to facilitate its disposal by autophagy.29,51 Impaired intracellular trafficking of mutant Pmp22 in Tr, TrJ, and several CMT and Déjérine-Sottas syndrome (DSS) mutations leads to its aggregation in intracellular compartments.17,18,68,69,78,119 The fate of the trapped WT/TrJ heterodimers is not clear, and it is possible that impaired cellular trafficking or overloading of the ER could result in a toxic gain-of-function effect.66 The formation of aggresomes in response to wild-type Pmp22 overexpression and their presence in the nerves of TrJ animals suggest that they may have a role in pathogenesis.77,97 Mechanisms have been postulated as to how aggresome formation may lead to cell death, but none has yet been verified in vivo.29 They include the coaggregation of essential cellular components and indirect consequences of traffic inhibition by the aggregated protein. There is currently no evidence of Schwann cell death being a primary cause of pathology in CMT1A, although because of Schwann cell overproliferation in demyelinating neuropathies, this is difficult to assess.

Aggregate Formation as the Possible Source of Toxicity and Phenotype Variability Two recent studies suggest that aggresome formation, rather than the presence of aggregated protein itself, is cytotoxic. In a transgenic model of familial amyotrophic lateral sclerosis, the initial phase of protein aggregation of Cu,Zn-superoxide dismutase precedes the manifestation of pathology by 3 months, whereas the visible accumulation of aggresomes is a late event that coincides with the death of motor neurons.47 A recent study suggested that the pathogenicity of protein-deposition disease could be primarily related to the structure of the aggregates rather than to the specific protein sequences from which they arise. Using proteins not known to be associated with disease, the authors found that small granular protein aggregates (which tend to form early in the aggregation process) were more toxic to cells than the fibrillar forms found later.15 The finding that the physical characteristics of the aggregate determine the degree of pathology is supported by the work of Isaacs et al. on a series of mutagen-induced Pmp22 mutants (see Point Mutations earlier in this chapter).45 The mutant protein that resulted in the most severe phenotype (Tr-m1H) only formed small protein aggregates. The proteins from the less severely affected mutants (Tr-m2H and Tr-m3H) occurred in the largest aggregates. It is not clear from this series whether the large aggregates have a protective role or the small aggregates/oligomers have detrimental effects.45 The difference in their tendency to form aggregates may also explain the differences in phenotype between Tr and TrJ mice. In detergent extracts from cells overexpressing Tr and TrJ protein, the Tr protein shows a much greater tendency to form high-molecular-weight aggregates than does TrJ Pmp22.119 Trembler Pmp22-expressing cells also showed the greatest number of cells containing aggresomes, followed by wild-type and then TrJ Pmp22.118 It is likely that the size/type of aggregate formed by the mutant protein alters its pathogenicity.

MYELIN PROTEIN ZERO MPZ is the most abundant protein in peripheral nerve, constituting more than 50% of protein and 8% of total mRNA.33 It is an immunoglobulin-related adhesion molecule and has been shown to mediate homophilic adhesion in cultured cells.23 X-ray crystallographic analysis of its extracellular domain suggests that it may also mediate adhesion in myelin.52,108 Functionally MPZ has been implicated in the spiraling of Schwann cell processes around myelincompetent axons81 and in myelin compaction.30,88 Mpzoverexpressing and -null mutant mice are discussed in detail in Chapter 24. The material presented in this chapter constitutes only a brief overview of these two models.

1573

Transgenic Models of Inherited Neuropathy

Table 65–3. Overexpression of Mpz in Transgenic Mice Copy number Tremor Median lifespan Grip strength (male) units Grip strength (female) units Overexpression (mRNA) (transgenic/endogenous) NCV (m/s) Myelinated fibers (%)

Control

Tg80.3

Tg80.4

Tg80.2

0 ⫺ ⬎2 yr 4.7 ⫾ 0.28 5.7 ⫾ 0.3 0

2 ⫺ ⬎2 yr 5.0 ⫾ 0.06 5.8 ⫾ 0.45 0.3

14 ⫹⫹ 10 mo 4.6 ⫾ 0.2 5.0 ⫾ 0.01 0.8

31 ⫹⫹⫹⫹ 4 mo 1.4 ⫾ 0.14 1.9 ⫾ 0.15 7.3

36 ⫾ 4.3 100

nd 100

11 ⫾ 4.0 50

2 ⫾ 0.1 ⬍10

mRNA ⫽ messenger RNA; NCV ⫽ nerve conduction velocity. Data from Wrabetz, L., Feltri, M. L., Quattrini, A., et al.: P0 overexpression causes congenital hypomyelination of peripheral nerves. J. Cell Biol. 148:1021, 2000.

Mpz-Overexpressing Mice

Mpz-Null Mutant Mice

Pronuclear injection of the complete Mpz gene resulted in several lines of transgenic mice with varying levels of overexpression. The behavioral, morphologic, and electrophysiologic phenotype of the Mpz-overexpressing mice increases in severity with increasing Mpz dosage (Table 65–3).133 Lines with more than 20 copies of the Mpz gene showed a severe phenotype, including a wobbling gait, tremor, atrophy of the paraspinal and hind limb musculature, weight reduction, and decreased lifespan.133 As Mpz expression increases, Schwann cell function is progressively impaired. Nerves with low levels of overexpression exhibit only transient hypomyelination. In nerves with moderate overexpression, a thin myelin sheath is generally formed. The nerves from high-copy mice show severe dysmyelination and a population of Schwann cells that fail to separate large axons away from axon bundles (Fig. 65–7). Developmental studies demonstrate that the phenotype of Mpz overexpression (as with PMP22 overexpression) is one in which myelin is never properly formed (dysmyelination), not of myelin that is formed and then destroyed.133

Homozygous null mice exhibit a progressive behavioral phenotype consisting of clasping of the hind limbs when lifted by the tail, uncoordinated swimming performance, slight tremors, and dragging or jerky movements of the hind limbs (4 weeks).30 Some animals have myotonic seizures30 and occasionally exhibit behavior usually considered to be indicative of pain, including shrieking and self-mutilation.30,135 The nerves of homozygous (Mpz⫺/⫺) mice are characterized initially by delayed myelination. In older animals, a proportion of fibers have abnormal myelin that lacks intraperiod lines, have widened intraperiod spaces, and have undulating major dense lines.30 Further pathologic features become apparent with age, including myelin degeneration, onion bulb formation, tomacula, pathologic features indicative of denervation, and occasionally the presence of macrophages.24,30,62 Both over- and underexpression of Mpz result in uniquely altered patterns of myelin gene expression.30,65,133,134 This contrasts with the situation seen in Pmp22 knockout animals, in which myelin-specific gene products are coordinately

FIGURE 65–7 Semithin sections through the sciatic nerve of 5-day-old mice. A, Wild-type mouse (scale bar: 50 ␮m). B, Tg80.2 mouse. Arrows indicate bundles of unsorted axons. Scale bar: 50 ␮m. (From Wrabetz, L., Feltri, M. L., Quattrini, A., et al.: P0 overexpression causes congenital hypomyelination of peripheral nerves. J. Cell Biol. 148:1021, 2000, with permission.)

1574

Diseases of the Peripheral Nervous System

reduced.3 Taken together, these data confirm that Mpz plays a regulatory as well as structural role in the PNS.

Mpz Epitope–Tagged, Toxic Gain-of-Function Model In order to track the intracellular trafficking of Mpz, a myc epitope tag was inserted.88 Full-length, normal Mpz was tagged at the mature NH2 terminus in the extracellular domain (ECD), which crystallographic studies had suggested was not involved in cis or trans interactions.88,108 However, the resulting mice exhibited an unexpected phenotype. The paranodal nature of the tomacula, poorly compacted myelin, and preserved axon caliber seen in Mpz-myc mice are features similar to those reported in CharcotMarie-Tooth disease type 1B patients.25,50,70,80,87,110,116 It appears that the myc tag at the N terminus mimics some of the effects of Mpz mutations in the ECD. Interestingly, Mpz-myc mice had both myelin packing abnormalities and tomacula, whereas patients usually have one or the other but not both.25 Behavior Three of the four founder Mpz-myc mice transmitted the transgene to their offspring. Mice from all lines displayed tremors and variable signs of muscle weakness such as

clenching of the paws, dragging of the hind limbs, and a floppy tail.88 In two of three lines, marked muscular atrophy was evident. One line self-mutilated, a behavior presumably derived from sensory symptoms. All behaviors were most severe in animals with the highest copy numbers and levels of transgene expression.88 Histology Many large-caliber axons in the sciatic nerve of Mpz-myc mice (P28) were hypomyelinated, having either thin or absent myelin sheaths (Fig. 65–8A and B).88 However, occasional fibers showed thick and redundant myelin (myelin outfoldings or tomacula).88 With advancing age, hypomyelination became more pronounced and demyelination (demyelinating figures and short internodes) and features of axonal degeneration were seen.88 Onion bulbs were not seen at any age in Mpz-myc mice. Ultrastructurally, myelin lamellae were widely spaced, in particular the inner lamellae close to the axon. High magnification of poorly compacted areas revealed the absence of an intraperiod line (Fig. 65–8C and inset), with the major dense line occasionally being uncompacted.88 Created in parallel with overexpressing mice,133 Mpzmyc animals showed clinical and morphologic abnormalities that were independent of overexpression. The myelin sheaths of all three lines of Mpz-myc mice had widened

FIGURE 65–8 Transverse sections through the sciatic nerve at 28 days postnatal. A, Wild-type mice. B, Mpz-myc nerves showing diffuse hypomyelination and a few abnormally outfolded myelin sheaths. C and inset, Myelin sheath demonstrating widely spaced myelin lamellae. (From Previtali, S., Quattrini, A., Fasolini, M., et al.: Epitope-tagged P0 glycoprotein causes Charcot-Marie-Tooth-like neuropathy in transgenic mice. J. Cell Biol. 151:1035, 2000, with permission.)

Transgenic Models of Inherited Neuropathy

myelin lamellae and tomacula. These were not seen in any of the Mpz wild-type insertion lines. In addition, Mpz-myc mice produced a more severe phenotype than that seen in mice overexpressing wild-type Mpz at comparable levels.88 When Mpz-myc was crossed onto the Mpz-null background, hypomyelination was very slightly improved but widening of myelin lamellae worsened and the number of tomacula increased, confirming that the phenotype is not due to overexpression.88 Mpz-myc was detected in uncompacted lamellae, tomacula, and normal myelin and is most likely co-integrated with wild-type Mpz in most myelin sheaths, confirming a gain-of-function mechanism. Compaction abnormalities in Mpz-myc/Mpz-null crossbreeding experiments suggest that Mpz-myc is partially adhesive, and can interact to form the intraperiod line, but is not able to tightly compact myelin.88 Reverse transcriptase–polymerase chain reaction studies showed that endogenous Mpz levels are decreased out of proportion with Mpz-myc, suggesting that Mpz-myc interferes not only with Mpz adhesion but also with the expression of endogenous Mpz.88 Electrophysiology Nerve conduction velocity was significantly decreased in both the lines tested (2.0 m/s and 15 m/s).

EARLY GROWTH RESPONSE GENE 2 Early growth response gene 2 (EGR2) (also known as KROX20) is one of a family of genes with a highly conserved DNA binding domain. EGR proteins participate in regulating the transcription of target genes by binding to specific promoter sequences (reviewed by O’Donovan et al.79). Mutations in EGR2 have been found in patients with several inherited peripheral neuropathies, including congenital hypomyelinating neuropathy (CHN), DSS, and CMT. The variety of disease phenotypes largely reflects the range of severity of the EGR2 mutations.117,129 All the patient phenotypes detected to date are associated with missense mutations. Dominantly acting mutations include R359W, S382R/D383Y, and R409W, which are located in the zinc finger domain and affect DNA binding. Of those studied, the mutant S382R/D383Y, which has the highest level of DNA binding, is associated with the most severe phenotype (CHN), whereas the mutant that shows no DNA binding (R409W) is associated with the milder CMT phenotype.130 The mutation that results in the most severe phenotype, S382R/D383Y, is unable to activate myelin protein expression in cultured Schwann cells. It also has a strong dominant-negative effect on the ability of wild-type Egr2 to mediate myelin gene expression.71 This suggests that most of the dominantly acting mutations in the DNA binding domain are dominant-negative mutations. The only autosomal recessive

1575

mutation recorded (see molgen-www.uia.ac.be), I268N, is located in the inhibitory R1 domain. This mutation probably results in deregulation of EGR2.129 Because all mutations found in human patients have been missense mutations, it was not clear what phenotype would be associated with a null mutation.

Egr2 (Krox20)-Null Mice Null mutant mice were originally created to examine the role of Egr2 in the developing hindbrain.105 Egr2 is the mouse ortholog of EGR2 and is also called Krox20. Egr2/Krox20 homozygous null mice generally die in the early postnatal stages, most probably as a result of compromised segmentation of the hindbrain.105 They also have abnormalities of endochondral bone formation and masticatory muscle development.53,123 The few animals that survived until P10 to P15 were observed to have begun to tremble, suggesting a phenotypic effect of the mutation on the PNS.105,120 Histology The peripheral nerves of homozygous Egr2-deficient mice are virtually devoid of myelin. The absence of any myelin breakdown products suggested that this was due to a failure of the onset of the myelination process.120 In the early stages of myelination, Schwann cells appeared to develop a normal one-to-one relationship with axons. They had begun to initiate the myelination process by wrapping around the axon. After they had completed one and onehalf turns around the axon, the myelination process failed to proceed any further. This suggests that Egr2 is necessary for the transition between the promyelinating and myelinating stages of myelinated fiber development. No myelination abnormalities were seen in heterozygous animals, indicating that half the normal dose of Egr2 is sufficient for normal myelination.120

Transcriptional Activation of Late Myelin Genes by Egr2 In the absence of Egr2, Schwann cell differentiation was blocked at a stage where the early myelin marker Mag was present but late myelin gene products (Mpz and Mbp) were absent.120 The implication that Egr2 is involved in the activation of late myelin genes has been confirmed by infecting Schwann cells with Egr2-expressing adenovirus.71 Infected Schwann cells upregulated genes critical for myelin formation and maintenance (Mpz, Pmp22, the connexin 32 [Cx32] gene gap junction B1 [Gjb1], periaxin [Prx], Mag, and Mbp). They also upregulated genes encoding enzymes essential for the synthesis of lipids and cholesterol.71 Transfection of the same construct into Egr2-deficient Schwann cells resulted in restoration of normal levels of myelination-related genes. This confirms that Egr2-deficient Schwann cells were unable

1576

Diseases of the Peripheral Nervous System

to execute the myelination program rather than being arrested at an immature state.71 Efficient induction of Egr2 has also been shown to require direct axonal contact, suggesting that it is a key component of the transduction cascade linking axonal signaling to myelination.67

functioning of both cell types. Transmission of essential signals, relayed by second messengers, from axon to Schwann cell nucleus or vice versa is postulated to occur through the intracellular channels formed by Cx32.2,13,73,90

Cx32-Deficient Mice CONNEXIN 32 The onset of clinical symptoms in X-linked Charcot-MarieTooth disease (CMTX) varies from congenital to the third decade. Symptoms are characterized by distal muscle weakness, muscle wasting, areflexia, and foot deformities. The major pathology in CMTX nerves consists of agerelated loss of myelinated fibers and increased numbers of regenerating axon clusters (see Chapter 76). Linkage studies localized CMTX to the proximal segment of the long arm of chromosome X, and mutations were subsequently found in the Cx32 (GJB1) gene.13 Since the first reports in 1993, 250 different mutations in the Cx32 gene have been recorded (see molgenwww.uia.ac.be/CMTMutations/Datasource/MutByGene. cfm). They are predominantly missense mutations but also include frameshift, deletion, and nonsense mutations.

Cx32: A Gap Junction Protein Connexin 32 is a member of a large gene family that encodes subunit proteins of gap junctions. Gap junctions consist of complexes of connexin proteins that form intercellular channels. These channels allow the exchange of small molecules such as ions, second messengers, and metabolites. Communication between cells is important for the regulation of cell proliferation and differentiation, both during tissue development and in carcinogenesis. Communication is also important for the propagation of signals among electrically excitable cells. Connexin 32 is highly expressed in many cells, including hepatocytes, exocrine pancreatic cells, and myelinating Schwann cells. In Schwann cells it is regulated in parallel with other myelin genes and is localized in uncompacted regions of the myelin sheath, such as paranodal loops and Schmidt-Lanterman incisures.13,16,104 This localization of Cx32 within individual Schwann cells rules out the formation of conventional cell-cell channels and suggests that Cx32 may be a component of reflexive intracellular gap junctions.83 These are channels that cross the myelin sheath, creating a radial diffusion pathway within the Schwann cell between the adaxonal and perinuclear cytoplasm. In a myelin sheath that is 4 ␮m thick, the unrolled membrane may be as long as 4 mm. A radial pathway through the Schwann cell could reduce the diffusion distance by up to 1000-fold.103 Schwann cells and axons are engaged in continuous reciprocal interactions that are essential for the proper

The Cx32 gene (Gjb1) has been inactivated by homologous recombination in ES cells that were then used to generate an animal model.72 Cx32-deficient mice were generally smaller than control animals but long lived and fertile. No behavioral abnormalities were seen even in aged animals (15 months).7,103 Connexin 32 function was assessed by measuring the efficiency of second messenger molecule propagation through gap junctions. Following sympathetic nerve stimulation, the amount of glucose released from hepatic glycogen stores was reduced in Cx32-deficient mice.72 This suggests defective communication between cells. Connexin 32–deficient mice also had a reduced number of gap junctional plaques.72 Despite Cx32 being highly expressed in peripheral nerve, the original report on Cx32deficient mice found no evidence of neurologic abnormalities (3 months).72 A late-onset progressive peripheral neuropathy was subsequently described in the same mice.7 Histology Studies of Cx32-deficient mice have described a demyelinating neuropathy that predominantly affects motor fibers.7,103 Histologic signs of demyelinating neuropathy were not seen in 7-week animals but had become evident by 3 months.103 The pathology initially consisted of a few demyelinated axons, a much higher number of thinly myelinated fibers (indicating remyelination following earlier demyelination), and some active demyelination.103 The incidence of abnormality increased with age with the addition of further pathologic features, including onion bulbs and clusters of regenerating axonal sprouts. Occasionally, denervated Schwann cells were seen distally. Because Schwann cells are not usually found free in the cytoplasm, this implies previous axon degeneration.7,103 Motor fibers were more affected than sensory ones in both studies.7,103 More subtle abnormalities of the myelin sheath were noted during electron microscope examination. The most common was an increased amount of adaxonal Schwann cell cytoplasm that was filled with large bundles of filaments.103 This adaxonal material was rarely seen in young adult animals but was commonly found in 9- to 12-monthold animals.103 The filaments, which are smaller than neurofilaments, are thought to be filamentous actin. Scherer et al.103 have noted that adaxonal bundles of filaments have been found in the ventral roots of normal rats and so may not be strictly pathologic.46 Other pathologic features included the separation of the myelin sheath from the axon and splitting of the myelin sheath itself. Unmyelinated fibers showed no evidence of abnormality at any of the

Transgenic Models of Inherited Neuropathy

ages studied. No appreciable pathologic changes were seen in the optic nerve or spinal cord. Even the large myelinated fibers of the ventral funiculus appeared unaffected, ruling out axonal caliber alone as the reason for the lack of demyelination in the CNS.103 Electrophysiology Mild electrophysiologic abnormalities were found in animals that were 4 to 6 months old. These included a slight slowing of conduction velocities (32 ⫾ 9 m/s vs. 35 ⫾ 9 m/s in wild-type mice) and increased latency of the muscle response after distal stimulation of the sciatic nerve.7 This mild slowing of nerve conduction velocity is similar to the mild slowing seen in human CMTX patients (see Chapter 76). Random Inactivation of the Cx32 Gene in Female Mice In humans, heterozygous women are less affected than males. This is presumably because they have one normal Cx32 gene in addition to the mutant allele, whereas males only have the mutant allele. Like other genes of the X chromosome, one would expect the Cx32 gene to be randomly inactivated in females. This produces a mosaic of cells some of which express normal levels of Cx32 and others none at all in heterozygotes. Consistent with this, homozygous Cx32 female mice (Cx32⫺/⫺) and male mice (Cx32⫺/Y) have the same pathologic features, age of onset, and severity at any given age, whereas heterozygous Cx32⫹/⫺ female mice have fewer demyelinated/remyelinated fibers than age-matched homozygous female (Cx32⫺/⫺) and Cx32⫺/Y mice.103 The demyelination in heterozygous female mice is probably caused by lack of Cx32 in individual Schwann cells, and examples of apposed paranodes in which one was Cx32 immunopositive and the other negative have been observed.103

Cx32 Mutant Mice Because Cx32 is located on the X chromosome, it is hard to determine whether mutations cause loss of function or dominant-negative effects. To address this, transgenic mice expressing frameshift (175)1 or missense (R142W)101 mutations have been generated. In the frameshift transgenics, no Cx32 protein was detected and there was no evidence of neuropathy, even in the lines in which transgene mRNA was highly expressed, indicating a loss-of-function mutation. R142W-expressing transgenics developed a demyelinating peripheral neuropathy. The mutant protein remained in the perinuclear region and did not reach the incisures or paranodes.104 The expression of mutant Cx32 also reduced the level of wild-type Cx32, indicating that this mutation may have a dominant-negative effect on endogenous Cx32.

1577

Alterations at the Molecular Level Schwann cells cultured from Cx32 knockout mice were found to express levels of mRNA species similar to those of control cells.73 Normal expression was found for genes characteristic of myelinating (Pmp22, Krox20, and Mbp) and nonmyelinating (Krox24, Zfp57, Mag, S100, glial fibrillary acid protein [Gfap], and suppressed cAMP-inducible POU [Scip]) Schwann cells.73 Expression of a number of connexin genes was also unaltered.73 The only major alteration found, with the exception of Cx32 itself, was an increase in Gfap expression. Gfap mRNA and protein levels were upregulated both in Cx32-deficient cultured cells and in vivo. GFAP is a cytoskeletal protein that is usually synthesized only by nonmyelinating Schwann cells. Its role in the PNS, if any, remains entirely speculative. Increased GFAP message and protein is a common phenomenon in the CNS that is normally associated with astrogliosis or the localized astrocyte proliferation that occurs following injury or inflammation.20 The primary intermediate filament protein in the PNS is not GFAP but vimentin. GFAP increase may contribute to cytoskeletal instability, but the consequences of an excess amount for Schwann cell function remain uncertain.73 Presence of Functional Gap Junctions in the Nerves of Cx32-Deficient Mice Balice-Gordon et al. demonstrated the presence of functional gap junctions in the nerves of Cx32-deficient mice.11 Injection of a low-molecular-weight fluorescent tracer into the perinuclear region of Schwann cells produced a similar distribution pattern in the nerves of Cx32-deficient and normal mice. The dye diffused radially from the outer/ abaxonal space to the inner/adaxonal cytoplasm across incisures. Diffusion was prevented by the addition of a presumed gap junction channel “blocker” (AGA). This provided the first evidence that other functional gap junction proteins may participate in the reflexive pathway in murine Schwann cells. A recent study has shown that connexin 29 (Cx29) is expressed in peripheral nerve and its mRNA is also regulated by axon–Schwann cell interactions.6 Myelinating Schwann cells express both connexins, and although their intracellular localizations differ, they overlap at incisures and paranodes. This suggests that they both contribute to reflexive gap junctions.6 Although Cx29 may participate in the formation of reflexive Schwann cell gap junctions, it cannot provide a complete functional substitute for Cx32, because Cx32-deficient mice go on to develop a neuropathy. Each connexin has unique biophysical and gating properties. The nature of the intercellular communication between cells appears to be dependent on the particular connexin composition of the channels involved. CMTX may result from mutations in Cx32 affecting gating, permeability, or pH dependence in addition to loss-of-function mutations.

1578

Diseases of the Peripheral Nervous System

Evidence from an oocyte expression system demonstrates that subtle changes in channel function appear to be sufficient to cause neuropathy.2,90 This allows speculation that certain signaling molecules may be excluded from mutated channels even though the passage of ions and electrical communication are maintained.90 The idea that a particular combination of connexins may be required to form a fully functional channel is not without precedent. Connexin 46 (Cx46) and connexin 50 (Cx50) are expressed in the eye with almost identical distributions. A different type of cataract is formed in the respective knockout mutants.32,132 Gap junctions consisting entirely of one or the other protein are not sufficient to maintain functionality, implying that the unique properties of mixed Cx46/Cx50 channels are necessary for normal lens function.73 Xenograft Model Axons from a normal mouse were allowed to grow through a transplanted piece of nerve from a patient with a CMTX mutation (Glu102Gly).98 The mutant human Schwann cells then ensheathed and myelinated the normal mouse axons. Schwann cells producing mutant Cx32 strongly influenced the cytoskeletal organization within the newly myelinated axons of the graft. Axons regenerating through the proximal portion of the Cx32 graft had increased crosssectional areas associated with increased neurofilament density, decreased numbers of microtubules, fragmentation of smooth ER, and an increase in the number of membranous organelles usually carried by fast axonal transport.98 None of these features was seen in control animals, indicating a pathologic basis for this enlargement. Beyond the proximal portion of the graft, axonal areas became progressively smaller and remained smaller into the host segment at the end of the graft. Distally there was evidence of axonal atrophy, wallerian degeneration, and axonal sprouting indicative of ongoing cycles of axonal degeneration and regeneration.98 The increased neurofilament density proximally and the subsequent axonal atrophy suggest that the neurofilament transport rate within the graft must be slower than in the host proximal area, causing the axons to become enlarged locally. Sahenk and Chen also found evidence of retarded fast bidirectional axonal transport within the graft.98 Aside from a few thinly myelinated fibers, no significant alterations of myelin were reported. This study demonstrates that the absence of normal (or the presence of a specific abnormal) Cx32 adversely affects the axon. This is likely due to the loss of a Schwann cell–axon signaling pathway normally mediated by Cx32.

PERIAXIN The demyelinating phenotype seen in Prx-deficient mice prompted investigations into the possible role of PRX in human peripheral neuropathy. Periaxin mutations have been

found associated with Charcot-Marie-Tooth disease type 3 (also known as DSS), Charcot-Marie-Tooth disease type 4F (a sensorimotor demyelinating neuropathy associated with neuropathic pain and hearing loss), and an early-onset, slowly progressing CMT-like neuropathy.14,34,115 The cases found to date are autosomal recessive. Patients with PRXassociated neuropathy have pronounced sensory involvement, which is not normally seen in patients with classic DSS or CMT. The peripheral nerve pathology in PRX patients includes demyelination with some remyelination, classic or basal laminal onion bulbs, occasional tomacula, focally folded myelin, and detached myelin loops. The murine Prx gene encodes two cytoskeletal proteins, L-periaxin and S-periaxin, of 147 and 16 kDa, respectively.19 The presence of PDZ domains in both proteins suggests that periaxin may have a role in organizing multiprotein signaling complexes35 and/or in linking transmembrane proteins with the cytoskeleton.22 Periaxin is expressed exclusively by myelinating Schwann cells.102 It is first detected early in the myelination process immediately after the Schwann cell has established a one-to-one relationship with an axon.102 Periaxin is a myelin-related protein and is regulated in parallel with other myelin-related genes during development and nerve degeneration.102 The appearance of Prx immunoreactivity precedes that of the common myelin proteins Mag, Mbp, or Mpz. In developing nerve and during regeneration, Prx expression is found predominantly on the adaxonal Schwann cell membrane (apposing the axon). As myelination proceeds, its localization changes to a predominantly abaxonal one (apposing the basal lamina). This shift in location within the Schwann cell suggests that Prx has two different roles in membrane protein interactions, the first in the later stages of axon ensheathment and the second in the stabilization of the mature myelin sheath.102

Prx-Deficient Mice Inactivation of the Prx gene by deletion of exon 6 and part of exon 7 was carried out by homologous recombination in ES cells.31 Neither the L- nor S-periaxin protein was detectable in the nerves of homozygous mutant animals. Embryonic development was not compromised by inactivation of the Prx gene. Behavior In terms of gross morphology, Prx-deficient mice initially appeared normal.31 The first indication of abnormality was seen at 4 to 6 weeks. Periaxin-deficient animals clasped their hind limbs when lifted by the tail; this clasping action was accompanied by a mild tremor in some animals. The disorder was progressive, and by 6 to 9 months deficient animals had an unsteady gait and splayed limbs, often having difficulty supporting themselves. Older animals lost weight rapidly as a result of an inability to feed. Their

1579

Transgenic Models of Inherited Neuropathy

breathing became labored, and they were killed to prevent distress. Sensory Involvement In human peripheral demyelinating disease, damage to sensory nerves is linked to a heightened response to painful stimuli (hyperalgesia) and excessive sensitivity to normally innocuous stimuli (allodynia). Prx-null mice displayed evidence of both allodynia and hyperalgesia. They had low withdrawal thresholds both when the hind paws were stimulated mechanically (von Frey’s hairs) and when a nociceptive thermal stimulus was applied. Both of these responses increased with the administration of an N-methyl-D-aspartate antagonist, which suggests a central change (probably sensitization). At the same age, grip strength test indicated no overt motor deficit (6 weeks).31 This mouse mutant provides a model of late-onset neuropathy with a severe clinical phenotype and is the first in which sensory deficits have been a significant feature. The origin of tactile allodynia and neuropathic pain in human demyelinating disorders is poorly understood. It is hoped that the Prx-deficient mouse may provide a model in which these issues can be examined. Histology Prx-null mice have a slowly progressive demyelinating neuropathy. The earliest morphologic alterations observed consisted of internodal myelin infoldings and focal thickening of the myelin sheath (indicative of myelin instability). By 8 months naked or thinly myelinated axons were commonly seen, often surrounded by redundant basal lamina or classic onion bulbs (indicative of multiple attempts at remyelination). Sensory, motor, and autonomic nerves were all extensively demyelinated. This model is unique in that all nerve types are equally affected. In all the myelin protein null mutants examined to date, the motor nerves are predominantly affected with the sensory ones relatively spared.3,7,62 There was no evidence of axonal damage, and the number of dorsal root ganglion neurons was unchanged, though axon numbers themselves were not counted. Many axons were incompletely surrounded or had abnormally thin myelin sheaths. This contrasts with the situation seen in other transgenic models with altered myelin protein expression, in which delayed myelination was usually evident during early postnatal development.93 Heterozygous mice showed occasional focal thickening of the myelin sheaths but no demyelination. Electrophysiology At 6 to 8 months there was electrophysiologic evidence of both sensory and motor nerve fiber involvement, which affected myelinated fibers specifically.31 An increased level of spontaneous low-frequency discharge (thought to contribute to pain following nerve injury) was found in the Prx mutant mice, which prompted the analysis of pain responses.

REFERENCES 1. Abel, A., Bone, L. J., Messing, A., et al.: Studies in transgenic mice indicate a loss of connexin32 function in X-linked Charcot-Marie-Tooth disease. J. Neuropathol. Exp. Neurol. 58:702, 1999. 2. Abrams, C., Yi Ri, S., and Bargiello, A.: Mutations in connexin 32: the molecular and biophysical bases for the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Brain Res. Rev. 32:203, 2000. 3. Adlkofer, K., Martini, R., Aguzzi, A., et al.: Hypermyelination and demyelinating peripheral neuropathy in Pmp22-deficient mice. Nat. Genet. 11:274, 1995. 4. Adlkofer, K., Naef, R., and Suter, U.: Analysis of compound heterozygous mice reveals that the Trembler mutation can behave as a gain-of-function allele. J. Neurosci. Res. 49:671, 1997. 5. Aguayo, A. J., Attiwell, M., Trecarten, J., et al.: Abnormal myelination in transplanted Trembler mouse Schwann cells. Nature 265:73, 1977. 6. Altevogt, B. M., Kleopa, K. A., Postma, F. R., et al.: Connexin29 is uniquely distributed within myelinating glial cells of the central and peripheral nervous systems. J. Neurosci. 22:6458, 2002. 7. Anzini, P., Neuberg, D., Schachner, M., et al.: Structural abnormalities and deficient maintenance of peripheral nerve myelin in mice lacking the gap junction protein connexin 32. J. Neurosci. 17:4545, 1997. 8. Ayers, M. M., and Anderson, R. M.: Onion bulb neuropathy in the Trembler mouse: a model of hypertrophic interstitial neuropathy (Dejerine-Sottas) in man. Acta Neuropathol. (Berl.) 25:54, 1973. 9. Ayers, M. M., and Anderson, R. M.: Development of onion bulb neuropathy in the Trembler mouse: comparison with normal nerve maturation. Acta Neuropathol. (Berl.) 32:43, 1975. 10. Baechner, D., Liehr, T., Hameister, H., et al.: Widespread expression of the peripheral myelin protein-22 gene (pmp22) in neural and non-neural tissues during murine development. J. Neurosci. Res. 42:733, 1995. 11. Balice-Gordon, R., Bone, L., and Scherer, S.: Functional gap junctions in the Schwann cell myelin sheath. J. Cell Biol. 142:1095, 1998. 12. Bascles, L., Bonnet, J., and Garbay, B.: Expression of the PMP-22 gene in Trembler mutant mice: comparison with the other myelin protein genes. Dev. Neurosci. 14:336, 1992. 13. Bergoffen, J., Scherer, S. S., Wang, S., et al.: Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262:2039, 1993. 14. Boerkoel, C. F., Takashima, H., Stankiewicz, P., et al.: Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am. J. Hum. Genet. 68:325, 2001. 15. Bucciantini, M., Giannoni, E., Chiti, F., et al.: Inherent toxicity of aggregates implies a common mechanism for protein misfolding diseases. Nature 416:507, 2002. 16. Chandross, K. J., Kessler, J. A., Cohen, R. I., et al.: Altered connexin expression after peripheral nerve injury. Mol. Cell. Neurosci. 1996:501, 1996. 17. Colby, J., Nicholson, R., Dickson, K. M., et al.: PMP22 carrying the trembler of trembler-J mutations is intracellularly

1580

18.

19.

20. 21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Diseases of the Peripheral Nervous System

retained in myelinating Schwann cells. Neurobiol. Dis. 7:561, 2000. D’Urso, D., Prior, R., Greiner-Petter, R., et al.: Overloaded endoplasmic reticulum-Golgi compartments, a possible pathomechanism of peripheral neuropathies caused by mutations of the peripheral myelin protein PMP22. J. Neurosci. 18:731, 1998. Dytrych, L., Sherman, D. L., Gillespie, C. S., et al.: Two PDZ domain proteins encoded by the murine periaxin gene are the result of alternative intron retention and are differentially targeted in Schwann cells. J. Biol. Chem. 273:5794, 1998. Eng, L. F., Yu, A. C. H., and Lee, Y. L.: Astrocytic response to injury. Prog. Brain Res. 4:353, 1992. Falconer, D. S.: Two new mutants, ‘Trembler’ and ‘Reeler,’ with neurological actions in the house mouse (Mus musculus L.). J. Genet. 50:192, 1951. Fanning, A. S., and Anderson, J. M.: Protein molecules as organizers of membrane structure. Curr. Opin. Cell Biol. 11:432, 1999. Filbin, M. T., Walsh, F. S., Trapp, B. D., et al.: Role of myelin P0 protein as a homophilic adhesion molecule. Nature 344:871, 1990. Frei, R., Mötzing, S., Kinkelin, I., et al.: Loss of distal axons and sensory Merkel cells and features indicative of muscle denervation in hindlimbs of P0-deficient mice. J. Neurosci. 19:6058, 1999. Gabreëls-Festen, A. A., Hoogendijk, J. E., Meijerink, P., et al.: Two divergent types of nerve pathology in patients with different P0 mutations in Charcot-Marie-Tooth disease. Neurology 47:761, 1996. Gale, A. N., Gomez, S., and Duchen, L. W.: Changes produced by a hypomyelinating neuropathy in muscle and its innervation. Brain 105:373, 1982. Garbay, B., and Bonnet, J.: P0 protein in normal, trembler heterozygous/homozygous mice during active PNS myelination. Neuroreport 3:594, 1992. Garcia-Mata, R., Bebik, Z., Sorscher, E. J., et al.: Characterization and dynamics of aggresome formation by a cytosolic GFP chimera. J. Cell. Biol. 146:1239, 1999. Garcia-Mata, R., Gao, Y.-A., and Sztul, E.: Hassles with taking out the garbage: aggravating aggresomes. Traffic 3:388, 2002. Giese, K. P., Martini, R., Lemke, G., et al.: Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71:565, 1992. Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., et al.: Peripheral demyelination and neuropathic pain behaviour in periaxin deficient mice. Neuron 26:523, 2000. Gong, X., Li, E., Klier, G., et al.: Disruption of the a3 connexin gene leads to proteolysis and cataractogenesis in mice. Cell 91:833, 1997. Greenfield, S. S., Brostoff, E. H., Eylar, E. H., et al.: Protein composition of myelin of the peripheral nervous system. J. Neurochem. 20:1207, 1973. Guilbot, A., Williams, A., Ravisé, N., et al.: A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum. Mol. Genet. 10:415, 2001.

35. Harris, B. Z., and Lim, W. A.: Mechanism and role of PDZ domains in signaling complex assembly. J. Cell Sci. 114:3219, 2001. 36. Henry, E. W., Cowen, J. S., and Sidman, R. L.: Comparison of Trembler and Trembler-J mouse phenotypes: varying severity of peripheral hypomyelination. Exp. Neurol. 42:688, 1983. 37. Henry, E. W., and Sidman, R. L.: Long lives for homozygous Trembler mutant mice despite virtual absence of peripheral nerve myelin. Science 241:344, 1988. 38. Hoogendijk, J. E., Janssen, E. A. M., Gabreëls-Festen, A. A. W. M., et al.: Allelic heterogeneity in hereditary motor and sensory neuropathy type Ia (Charcot-Marie-Tooth disease type Ia). Neurology 3:1010, 1993. 39. Huxley, C., Passage, E., Manson, A., et al.: Construction of a mouse model of Charcot-Marie-Tooth disease type 1A by pronuclear injection of human YAC DNA. Hum. Mol. Genet. 5:563, 1996. 40. Huxley, C., Passage, E., Robertson, A. M., et al.: Correlation between varying levels of PMP22 expression and the degree of demyelination and reduction in nerve conduction velocity in transgenic mice. Hum. Mol. Genet. 7:449, 1998. 41. Ikegami, T., Ikeda, H., Aoyama, M., et al.: Novel mutations of the peripheral myelin protein 22 gene in two pedigrees with Dejerine-Sottas disease. Hum. Genet. 102:294, 1998. 42. Inuzuka, T., Quarles, R. H., Heath, J., et al.: A human lymphocyte antigen is shared with a group of glycoproteins in peripheral nerve. Neurosci. Lett. 51:105, 1985. 43. Ionasescu, V. V., Searby, C. C., Ionasescu, R., et al.: Dejerine-Sottas neuropathy in mother and son with same point mutation of PMP22 gene. Muscle Nerve 20:97, 1997. 44. Isaacs, A. M., Davies, K. E., Hunter, A. J., et al.: Identification of two new Pmp22 mouse mutants using large-scale mutagenesis and a novel rapid mapping strategy. Hum. Mol. Genet. 9:1865, 2000. 45. Isaacs, A. M., Jeans, A., Oliver, P., et al.: Identification of a new Pmp22 mouse mutant and trafficking analysis of a Pmp22 allelic series suggesting that protein aggregates may be protective in Pmp22-associated neuropathy. Mol. Cell. Neurosci. 21:114, 2002. 46. Jacobs, J. M., and Cavanaugh, J. B.: Aggregations of filaments in Schwann cells of spinal roots of the normal rat. J. Neurocytol. 1:161, 1972. 47. Johnston, J. A., Dalton, M. J., Gurney, M. E., et al.: Formation of high molecular weight complexes of mutant Cu,Zn-superoxide dismutase in a mouse model of amyotrophic lateral sclerosis. Proc. Natl. Acad. Sci. U. S. A. 97:12571, 2000. 48. Johnston, J. A., Ward, C. L. and Kopito, R. R.: Aggresomes: a cellular response to misfolded proteins. J. Cell Biol. 143:1883, 1998. 49. King, R. H. M., Muddle, J. R., Nourallah, M., et al.: An investigation into axonal loss in transgenic and spontaneous mutant mouse models of HMSN1a. Manuscript in preparation, 2003. 50. Kirschner, D. A., Szumowski, K., Gabreëls-Festen, A. A., et al.: Inherited demyelinating peripheral neuropathies: relating myelin packing abnormalities to P0 molecular defects. J. Neurosci. Res. 46:502, 1996.

Transgenic Models of Inherited Neuropathy 51. Kopito, R. R.: Aggresomes, inclusion bodies and protein aggregation. Trends Cell Biol. 10:524, 2000. 52. Lemke, G., Lamar, E., and Patterson, J.: Isolation and analysis of the gene encoding peripheral myelin protein zero. Neuron 1:73, 1988. 53. Levi, G., Topilko, P., Schneider-Maunoury, S., et al.: Defective bone formation in Krox-20 mutant mice. Development 122:113, 1996. 54. Liehr, T., and Rautenstrauss, B.: The peripheral myelin protein 22kDa (PMP22) gene is expressed in non-neural epithelial tissues of mouse, rat, cattle and human [abstract]. Am. J. Hum. Genet. 61(SS):1015, 1997. 55. Low, P. A.: Hereditary hypertrophic neuropathy in the Trembler mouse. Part 1. Histopathological studies: light microscopy. J. Neurol. Sci. 30:327, 1976. 56. Low, P. A.: Hereditary hypertrophic neuropathy in the Trembler mouse. Part 2. Histopathological studies: electron microscopy. J. Neurol. Sci. 30:343, 1976. 57. Low, P. A.: The evolution of ‘onion bulbs’ in the hereditary hypertrophic neuropathy of the Trembler mouse. Neuropathol. Appl. Neurobiol. 3:81, 1977. 58. Low, P. A., and McLeod, J. G.: Hereditary demyelinating neuropathy in the Trembler mouse. J. Neurol. Sci. 26:565, 1975. 59. Lupski, J. R., Montes de Oca-Luna, R., Slaugenhaupt, S., et al.: DNA duplication associated with Charcot-MarieTooth disease type 1A. Cell 66:219, 1991. 60. Magyar, J. P., Martini, R., Ruelicke, T., et al.: Impaired differentiation of Schwann cells in transgenic mice with increased PMP22 gene dosage. J. Neurosci. 16:5351, 1996. 61. Manfioletti, G., Ruaro, M. E., Del Sal, G., et al.: A growth arrest-specific (gas) gene codes for a membrane protein. Mol. Cell. Biol. 10:2924, 1990. 62. Martini, R., Zielasek, J., Toyka, K. V., et al.: Protein zero (P0)-deficient mice show myelin degeneration in peripheral nerves characteristic of inherited human neuropathies. Nat. Genet. 11:281, 1995. 63. Maycox, P. R., Ortuno, D., Burrola, P., et al.: A transgenic mouse model of human hereditary neuropathy with liability to pressure palsies. Mol. Cell. Neurosci. 8:405, 1997. 64. Melcangi, R. C., Azcoitia, I., Ballabio, M., et al.: Neuroactive steroids influence peripheral myelination: a promising opportunity for preventing or treating age-dependent dysfunctions of peripheral nerves. Prog. Neurobiol. 71:57, 2003. 65. Menichella, D., Xu, W., Jiang, H., et al.: The absence of myelin P0 protein produces a novel molecular phenotype in Schwann cells. Ann. N. Y. Acad. Sci. 883:281, 1999. 66. Müller, H. W.: Tetraspan myelin protein PMP22 and demyelinating peripheral neuropathies: new facts and hypotheses. Glia 29:182, 2000. 67. Murphy, P., Topilko, P., Schneider-Maunoury, S., et al.: The regulation of Krox-20 expression reveals important steps in the control of peripheral glial cell development. Development 122:2847, 1996. 68. Naef, R., Adlkofer, K., Lescher, B., et al.: Aberrant protein trafficking in Trembler suggests a disease mechanism for hereditary human peripheral neuropathies. Mol. Cell. Neurosci. 9:13, 1997. 69. Naef, R., and Suter, U.: Impaired intracellular trafficking is a common disease mechanism of PMP22 point mutations in peripheral neuropathies. Neurobiol. Dis. 6:1, 1999.

1581

70. Nagakawa, M., Suehara, M., Saito, A., et al.: A novel MPZ gene mutation in dominantly inherited neuropathy with focally folded myelin sheaths. Neurology 52:1271, 1999. 71. Nagarajan, R., Svaren, J., Le, N., et al.: EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron 30:355, 2001. 72. Nelles, E., Bützler, C., Jung, D., et al.: Defective propagation of signals generated by sympathetic nerve stimulation in the liver of connexin32-deficient mice. Proc. Natl. Acad. Sci. U. S. A. 93:9565, 1996. 73. Nicholson, S., Gòmes, D., Néchaud, B., et al.: Altered gene expression in Schwann cells of connexin32 knockout animals. J. Neurosci. Res. 66:23, 2001. 74. Niemann, S., Sereda, M., Suter, U., et al.: Uncoupling of myelin assembly and Schwann cell differentiation by transgenic overexpression of peripheral myelin protein 22. J. Neurosci. 20:4120, 2000. 75. Norrell, J. C., Jamon, M., Riviere, G., et al.: Behavioural profiling of a murine Charcot-Marie-Tooth disease type 1A model. Eur. J. Neurosci. 13:1625, 2001. 76. Notterpek, L., Roux, K. J., Amici, S. A., et al.: Peripheral myelin protein 22 is a constituent of intercellular junctions in epithelia. Proc. Natl. Acad. Sci. U. S. A. 98:14404, 2001. 77. Notterpek, L., Ryan, M. C., Tobler, A. R., et al.: PMP22 accumulation in aggresomes: implications for CMT1 pathology. Neurobiol. Dis. 6:450, 1999. 78. Notterpek, L., Shooter, E. M., and Snipes, G. J.: Upregulation of the endosomal-lysosomal pathway in the Trembler-J neuropathy. J. Neurosci. 17:4190, 1997. 79. O’Donovan, K. J., Tourtellotte, W. G., Milbrandt, J., et al.: The EGR family of transcription-regulatory factors: progress at the interface of molecular and systems neuroscience. Trends Neurosci. 22:167, 1999. 80. Ohnishi, A., Yamamoto, T., Yamamoto, S., et al.: Myelinated fibres in Charcot-Marie-Tooth disease type 1B with Arg98His mutation of P0 protein. J. Neurol. Sci. 171:97, 1999. 81. Owens, G. C., and Boyd, C. J.: Expressing antisense P0 RNA in Schwann cells perturbs myelination. Development 112:639, 1991. 82. Patel, P. I., Roa, B. B., Welcher, A. A., et al.: The gene for the peripheral myelin protein PMP-22 is a candidate for CharcotMarie-Tooth disease type 1A. Nat. Genet. 1:159, 1992. 83. Paul, D. L.: New functions for gap junctions. Curr. Opin. Cell Biol. 7:665, 1995. 84. Perea, J., Robertson, A., Tolmachova, T., et al.: Induced myelination and demyelination in a conditional mouse model of Charcot-Marie-Tooth disease type 1A. Hum. Mol. Genet. 10:1007, 2001. 85. Perkins, C. S.: Schwann cell multiplication in Trembler mice. Neuropathol. Appl. Neurobiol. 7:115, 1981. 86. Perkins, C. S., Aguayo, A. J., and Bray, G. M.: Behaviour of Schwann cells from the Trembler mouse unmyelinated fibres transplanted into myelinated nerves. Exp. Neurol. 71:515, 1981. 87. Planté-Bordeneuve, V., Guiochon-Mantel, A., Lacroix, C., et al.: The Roussy-Lévy family: from the original description to the gene. Ann. Neurol. 46:1999. 88. Previtali, S., Quattrini, A., Fasolini, M., et al.: Epitope-tagged P0 glycoprotein causes Charcot-Marie-Tooth-like neuropathy in transgenic mice. J. Cell Biol. 151:1035, 2000.

1582

Diseases of the Peripheral Nervous System

89. Raeymaekers, P., Timmerman, V., Nelis, E., et al.: Duplication in chromosome 17p11.2 in Charcot-Marie-Tooth neuropathy type 1a (CMT 1a). The HMSN Collaborative Research Group. Neuromuscul. Disord. 1:93, 1991. 90. Ressot, C., and Bruzzone, R.: Connexin channels in Schwann cells and the development of the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Brain Res. Rev. 32:192, 2000. 91. Roa, B. B., Dyck, P. J., Marks, H. G., et al.: Dejerine-Sottas syndrome associated with point mutation in the peripheral myelin protein 22 (PMP22) gene. Nat. Genet. 5:269, 1993. 92. Robertson, A., Perea, J., Tolmachova, T., et al.: Effects of mouse strain, position of integration and tetracycline analogue on the tetracycline conditional system in transgenic mice. Gene 282:65, 2002. 93. Robertson, A. M., Huxley, C., King, R. H. M., et al.: Development of early postnatal peripheral nerve abnormalities in Trembler-J and PMP22 transgenic mice. J. Anat. 195:331, 1999. 94. Robertson, A. M., King, R. H. M., Muddle, J. R., et al.: Abnormal Schwann cell/axon interactions in the Trembler-J mouse. J. Anat. 190:423, 1997. 95. Robertson, A. M., Perea, J., McGuigan, A., et al.: Comparison of a new pmp22 transgenic mouse line with other mouse models and human patients with CMT1A. J. Anat. 200:377, 2002. 96. Ryan, M. C., Notterpek, L., Tobler, A. R., et al.: Role of peripheral myelin protein 22 N-linked glycan in oligomer stability. J. Neurochem. 75:1465, 2000. 97. Ryan, M. C., Shooter, E. M., and Notterpek, L.: Aggresome formation in neuropathy models based on peripheral myelin protein 22 mutations. Neurobiol. Dis. 10:109, 2002. 98. Sahenk, Z., and Chen, L.: Abnormalities in the axonal cytoskeleton induced by a connexin32 mutation in nerve xenografts. J. Neurosci. Res. 51:174, 1988. 99. Sancho, S., Magyar, J. P., Aguzzi, A., et al.: Distal axonopathy in peripheral nerves of PMP22-mutant mice. Brain 122:1563, 1999. 100. Sanders, C. R., Ismail-Beigi, F., and McEnery, M. W.: Mutations of the peripheral myelin protein 22 result in defective trafficking through mechanisms which may be common to diseases involving tetraspan membrane proteins. Biochemistry 40:9453, 2001. 101. Scherer, S., Bone, L. J., Deschenes, S. M., et al.: The role of the gap junction protein connexin32 in the pathogenesis of X-linked Charcot-Marie-Tooth disease. In Cardew, G. (ed.): Gap Junction-Mediated Intercellular Signalling in Health and Disease (Novartis Foundation Symposium 219). New York, John Wiley and Sons, 1999. 102. Scherer, S., Xu, Y., Bannerman, P., et al.: Periaxin expression in myelinating Schwann cells: modulation by axon-glial interactions and polarized localization during development. Development 121:4265, 1995. 103. Scherer, S., Xu, Y., Nelles, E., et al.: Connexin32-null mice develop demyelinating peripheral neuropathy. Glia 24:8, 1998. 104. Scherer, S. S., Deschenes, S. M., Xu, Y.-T., et al.: Connexin32 is a myelin-related protein in the PNS and CNS. J. Neurosci. 15:8281, 1995. 105. Schneider-Maunoury, S., Topilko, P., Seitanidou, T., et al.: Disruption of Krox-20 results in alteration of rhombomeres 3 and 5 in the developing hindbrain. Cell 75:1199, 1993.

106. Sereda, M., Griffiths, I., Pühlhofer, A., et al.: A transgenic rat model of Charcot-Marie-Tooth disease. Neuron 16:1049, 1996. 107. Sereda, M., Meyer zu Hörste, G., Suter, U., et al.: Therapeutic administration of progesterone antagonist in a model of Charcot-Marie-Tooth disease (CMT-1A). Nat. Med. 9:1533, 2003. 108. Shapiro, L., Doyle, J. P., Henesley, P., et al.: Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17:435, 1996. 109. Sidman, R. L., Cowen, J. S., and Eicher, E. M.: Inherited muscle and nerve diseases in mice: a tabulation with commentary. Ann. N. Y. Acad. Sci. 317:497, 1979. 110. Sindou, P., Vallat, J.-M., Chapon, F., et al.: Ultrastructural protein zero expression in Charcot-Marie-Tooth type 1B disease. Muscle Nerve 22:99, 1999. 111. Snipes, G. J., Suter, U., Welcher, A. A., et al.: Characterization of a novel peripheral nervous system myelin protein (PMP-22/SR13). J. Cell Biol. 117:225, 1992. 112. Spreyer, P., Kuhn, G., Hanemann, C. O., et al.: Axonregulated expression of a Schwann cell transcript that is homologous to a “growth arrest-specific” gene. EMBO J. 10:3661, 1991. 113. Suh, J.-G., Ichihara, N., Saigoh, K., et al.: An in-frame deletion in peripheral myelin protein-22 gene causes hypomyelination and cell death in Schwann cells in the new TREMBLER mutant mice. Neuroscience 79:735, 1997. 114. Suter, U., Welcher, A. A., Ozcelik, T., et al.: Trembler mouse carries a point mutation in a myelin gene. Nature 356:241, 1992. 115. Takashima, H., Boerkoel, C. F., De Jonge, P., et al.: Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51:709, 2002. 116. Thomas, F. P., Lebo, R. V., Rosoklija, G., et al.: Tomaculous neuropathy in chromosome 1 Charcot-Marie-Tooth syndrome. Acta Neuropathol. (Berl.) 87:91, 1994. 117. Timmerman, V., De Jonghe, P., Ceuterick, C., et al.: Novel missense mutation in the early growth response 2 gene associated with Dejerine-Sottas syndrome phenotype. Neurology 52:1827, 1999. 118. Tobler, A. R., Liu, N., Mueller, L., et al.: Differential aggregation of the Trembler and Trembler J mutants of peripheral myelin protein 22. Proc. Natl. Acad. Sci. U. S. A. 99:483, 2002. 119. Tobler, A. R., Notterpek, L., Naef, R., et al.: Transport of Trembler-J mutant peripheral myelin protein 22 is blocked in the intermediate compartment and affects the transport of the wildtype protein by direct interaction. J. Neurosci. 19:2027, 1999. 120. Topilko, P., Schneider-Maunoury, S., Levi, G., et al.: Krox-20 controls myelination in the peripheral nervous system. Nature 371:796, 1994. 121. Toyka, K., Zielasek, J., Ricker, K., et al.: Hereditary neuromyotonia: a mouse model associated with deficiency or increased gene dosage of the PMP22 gene. J. Neurol. Neurosurg. Psychiatry 63:812, 1997. 122. Tran, P. B., and Miller, R. J.: Aggregates in neurodegenerative disease: crowds and power? Trends Neurosci. 22:194, 1999. 123. Turman, J. E., Chopiuk, N. B., and Shuler, C. F.: The Krox-20 mutation differentially affects the development of masticatory muscles. Dev. Neurosci. 23:113, 2001.

Transgenic Models of Inherited Neuropathy 124. Tyson, J., Ellis, D., Fairbrother, U., et al.: Hereditary demyelinating neuropathies of infancy: a genetically complex syndrome. Brain 120:47, 1997. 125. Valentijn, L. J., Baas, F., Wolterman, R. A., et al.: Identical point mutations of PMP-22 in trembler-J mouse and Charcot-Marie-Tooth disease type 1A. Nat. Genet. 2:288, 1992. 126. Valentijn, L. J., Ouvrier, R. A., van den Bosch, N. H., et al.: Dejerine-Sottas neuropathy is associated with a de novo PMP22 mutation. Hum. Mutat. 5:76, 1995. 127. Vallat, J. M., Sindou, P., Garbay, B., et al.: Expression of myelin proteins in the adult heterozygous Trembler mouse. Acta Neuropathol. (Berl.) 98:281, 1999. 128. Verhagen, W. I. M., Gabreëls-Festen, A. A. W. M., Van Wensen, P. J. M., et al.: Hereditary neuropathy with liability to pressure palsies: a clinical, electrophysiological and morphological study. J. Neurol. Sci. 116:176, 1993. 129. Warner, L. E., Mancias, P., Butler, I. J., et al.: Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet. 18:382, 1998.

1583

130. Warner, L. E., Svaren, J., Milbrandt, J., et al.: Functional consequences of mutations in the early growth response 2 gene (EG2) correlate with the severity of human myelinopathies. Hum. Mol. Genet. 8:1245, 1999. 131. Welcher, A. A., Suter, U., De Leon, M., et al.: A myelin protein is encoded by the homologue of a growth arrestspecific gene. Proc. Natl. Acad. Sci. U. S. A. 88:7195, 1991. 132. White, T. W., Goodenough, D. A., and Paul, D. L.: Targeted ablation of connexin50 in mice results in microphthalmia and zonular pulverulent cataracts. J. Cell Biol. 143:815, 1998. 133. Wrabetz, L., Feltri, M. L., Quattrini, A., et al.: P0 overexpression causes congenital hypomyelination of peripheral nerves. J. Cell Biol. 148:1021, 2000. 134. Xu, W., Manichella, D., Jiang, H., et al.: Absence of P0 leads to the dysregulation of myelin gene expression and myelin morphogenesis. J. Neurosci. Res. 60:714, 2000. 135. Zimmerman, M.: Pathobiology of neuropathic pain. Eur. J. Pharmacol. 29:23, 2001.

G. Inherited Motor Neuron Degeneration

66 Transgenic Animal Models of Amyotrophic Lateral Sclerosis ERIC J. SORENSON

SOD1 Transgenic Models Neurofilament Transgenic Models

Double Transgenic Models Transgenic Models and Apoptosis

Historically, animal models have contributed only modestly to our understanding of the pathophysiology and treatment of amyotrophic lateral sclerosis (ALS). This is principally because these older animal models only partially represented the clinical syndrome of ALS. With an increased understanding of the genetics of this disease and transgenic technology, new animal models that more accurately represent human ALS have been developed. This has for the first time allowed for controlled experimentation into the cellular mechanisms of motor neuron death as it occurs in ALS and screening for potential therapeutic agents for treatment of ALS on a wide-scale basis. Prior to the development of the superoxide dismutase 1 (SOD1) transgenic mouse models, the wobbler mouse was the closest animal model of motor neuron disease. The wobbler mouse was identified as a naturally occurring mutation in a mouse colony in 1968.18 These animals are identifiable clinically within the first few weeks of life. They develop muscle weakness and atrophy and tremors (hence their name, wobbler mice). Wobbler mice develop motor neuron degeneration as the result of an autosomal recessive mutation. The motor neuron loss is greatest in the cervical segments. Although the gene has not yet been isolated, it has been linked to mouse chromosome 11.30 The gene has been mapped to a locus very close to the pseudogene glutamine synthetase.16 By crossing the wobbler mouse strain with another strain

Animals Models in Screening for New Therapeutic Agents

carrying a unique pseudogene glutamine synthetase, animals in the asymptomatic stage could be identified by analyzing this pseudogene.3 Pathologically, there is aggregation of neurofilaments in the wobbler mouse, a feature in common with human ALS.48 An astrocytosis24 and an increase in activated microglia4 are found as well; although nonspecific, both are seen in human ALS. In the available research, there is no clear evidence of apoptosis in the wobbler mouse.8 Despite the similarities the wobbler mouse has to human ALS, there are important distinctions as well. The anterior horn cells undergo vacuolar degeneration in the perikaryon.41 This feature is also identified in the transgenic models described below, but is not identified in human ALS. Pathologic analysis also demonstrates neuronal loss in the thalamus and cerebellar nuclei, features that are absent in human ALS.52 With the advent of the transgenic SOD1 animals, the wobbler mouse has taken a less prominent role in the study of motor neuron degeneration.

SOD1 TRANSGENIC MODELS The current animal models of ALS were made possible by recent advances in molecular genetics. Through linkage analysis and subsequent candidate genes, mutations in the human superoxide dismutase 1 gene (Cu/Zn) (SOD1) on 1585

1586

Diseases of the Peripheral Nervous System

chromosome 21 were identified as a cause in approximately 20% of familial forms of ALS.54,57 The principal function of this enzyme is to detoxify the superoxide anion (an oxygen free radical) into hydrogen peroxide (Fig. 66–1). The hydrogen peroxide is then converted to water by one of two enzymes: catalase and/or glutathione peroxidase.59 The initial assumption was that the mutant enzyme disrupted this enzymatic pathway. This would result in an accumulation of the superoxide anion and facilitate its reaction with nitric oxide, leading to the highly reactive peroxynitrite (ONOO•⫺). However, as an increasing number of mutations were discovered, it was also established that loss of enzymatic activity was not associated with disease. Many of the mutations retained normal activity in conversion of the superoxide anion to hydrogen peroxide.14 It was subsequently proposed that the mutant enzyme was acquiring a new toxic function, deleterious to the motor neurons. This proposed toxic gain of function was confirmed by the occurrence of motor neuron degeneration in transgenic mice expressing the mutant SOD1 gene.12 The development of transgenic mouse lines expressing the mutant SOD1 enzymes that demonstrated motor neuron loss was a major advancement in the study of ALS. These models were created using standard transgenic techniques (Fig. 66–2). Female mice are injected with the ovarian-stimulating hormones pregnant mare’s serum followed by human chorionic gonadotropin. This creates a superovulated female that is mated to males from the same lineage. Once impregnated, the embryos are harvested from the sacrificed female. These embryos are then isolated and injected with multiple recombinant DNA copies created from the mutant enzyme DNA sequence. These injected embryos are then implanted into female mouse surrogates. After birth of the litter, the animals containing the mutation are identified via DNA analysis. These animals are then mated with the original (nontransgenic) animals, and their progeny are examined

e– O2

– O2•

X

Female stimulated with human chorionic gonadotropin

Standard male

Fertilized eggs harvested from sacrificed female

Injection of recombinant transgene into the pronuclei of the fertilized egg

Egg implanted into female prepared by mating with sterilized male

H+ H2O2

Transfected mice offspring identified

SOD Catalase or

H2O2

Glutathione peroxidase

–

H2O2 + O2• FIGURE 66–1 SOD enzymatic pathway.

FE3+ FE2+

H2O

OH –+ OH

•

Germline mutations (founder animals) identified by transmission to subsequent generations FIGURE 66–2 Transgenic process.

Transgenic Animal Models of Amyotrophic Lateral Sclerosis

for the same DNA copy. If present in the offspring, then the mutation is present in the germline, and these animals serve as a founder population for subsequent generations.15 The first transgenic mouse carrying the SOD1 mutation was created by Gurney and colleagues in 1994.27 This model expressed a substitution of glycine for alanine at position 93 of the SOD1 protein and has become known as the G93A strain. Concurrently, this same group tried to create a second model using the more common human mutation A4V (alanine substitution for valine at position 4). This model failed to develop motor neuron degeneration. The A4V line expressed a low copy number of the mutation, which may explain why this model was unsuccessful. In contrast, the G93A transgenic mice were developed to express a large quantity of the mutant protein. Those mice possessing the mutant protein developed signs of limb weakness by 3 or 4 months of age and needed to be sacrificed by 5 months because of progressive weakness. Clinically the syndrome closely follows that of ALS. Pathologically, these mice possessed important differences from human familial and sporadic ALS. Most importantly, large numbers of intracellular vacuoles were observed. These findings have never been described in human ALS. Upon further investigation, it has been established that these vacuoles represent dilated mitochondria and endoplasmic reticulum.12,13,27,67 Despite this, the distribution of the pathology, affecting the anterior horn cells preferentially, closely mimics the human disease, as do other pathologic features. These animals demonstrate neurofilament-containing inclusions in the axons and Lewy body–like inclusions in the perikaryon.60 Additionally, presymptomatic mice demonstrate Golgi apparatus fragmentation that is very similar to what has been identified in human ALS.44 Transgenic control animals expressing wild-type human SOD1 were created and, despite high levels of expression, failed to develop a motor neuron syndrome. Furthermore, a SOD1 knockout mouse was created that failed to express any SOD1 activity. These mice also fail to develop the progressive motor neuron syndrome.53 These results, in combination, strongly suggest that the presence of the mutant enzyme is required for the development of murine motor neuron disease and strongly support the toxic gain-of-function hypothesis. Despite this, the severe vacuolization in the transgenic mouse and its absence in humans continued to raise concern as to the relevance of this model to human disease. Because this model had an early age of onset and a very rapid progression, it was postulated that the large number of mutant copies expressed may be responsible for these differences. To address this, a second transgenic line possessing the G93A mutation but expressing a lower number of copies of the mutant gene was created. This model possessed 10 copies of the gene, in contrast to the

1587

25 copies possessed by the first model. These subsequent animals had a later onset of disease, considerably less vacuolar change, and more filamentous inclusions.12,13 These changes more closely resembled the pathology of human ALS, which has supported the accuracy of the model. Subsequently, a number of other transgenic mouse lines have been created with other SOD1 mutations and have confirmed these findings. As with the G93A mouse model, these other models also confirm the association between number of mutant copies and disease severity. This dose response further supports the hypothesis of a toxic gain of function. Although there is now considerable evidence in the transgenic animal models to support a toxic gain of function in the SOD1 mutations, exactly what this function is remains unknown. Two principal mechanisms have been proposed. First, copper is known to catalyze peroxynitrite to reactive nitronium, which results in nitration of tyrosine residues contained in intracellular proteins. This reaction is normally restricted because of the tightly folded nature of the SOD1 protein, preventing access of peroxynitrite (formed spontaneously from nitric oxide and superoxide) to the protein-bound copper that catalyzes this reaction. However, it has been proposed that mutant enzyme is less tightly folded and therefore may facilitate access of peroxynitrite to the catalytic copper, allowing this reaction to take place.1 Once formed, the nitronium ions react with tyrosine residues, resulting in increased nitrotyrosine levels that may interfere with protein function, ultimately leading to protein aggregation as seen in the pathology of the transgenic mice and human ALS.6 An attractive target for this process would be the neurofilament proteins. Neurofilament aggregation is one of the pathologic hallmarks of ALS. However, the aggregated nitration of tyrosine residues on the neurofilaments has not been demonstrated.6 The formation of peroxynitrite requires the presence of nitric oxide, and nitric oxide synthetase, the enzyme that catalyzes the formation of nitric oxide, is expressed in the stressed motor neurons.63,68 Evidence against this mechanism is that creation of a G93A transgenic mouse model without expression of nitric oxide synthetase (presumably reducing the amount of nitric oxide) did not alter the toxicity of the SOD1 G93A mutation.20 Additionally, transgenic mice created lacking the copper chaperone protein (and hence with reduced copper-containing SOD1) still developed motor neuron degeneration.58 A second possible mechanism of toxicity is increased peroxidation of hydrogen peroxide to hydroxyl radicals. These hydroxyl radicals are highly toxic to cellular membranes found on organelles. These may also target other membrane proteins such as glutamate transporters.49,66 These two possibilities are not mutually exclusive and may be important in combination. Although other possible

1588

Diseases of the Peripheral Nervous System

mechanisms exist, these two remain the most prominent and best studied. It has been well established that glutamate metabolism is altered in human sporadic and familial ALS. Whether this is involved in the pathophysiology of the disease or is merely an epiphenomenon is uncertain. In human ALS it has been established that the glial glutamate transporter EAAT2 is reduced in most ALS subjects at autopsy.5 No mutations have been identified in the gene for this protein, nor has there been any loss of messenger RNA (mRNA) for this protein.5 This suggests that the loss of this protein occurs as a process of translation or posttranslational modification. The transgenic SOD1 mice demonstrated their earliest pathologic findings within the astrocytes.7 As in human ALS, this model also demonstrated a reduction in the glial glutamate transporter GLT1 (the murine equivalent to human EAAT2). How SOD1 mutation produces the GLT1 reduction is not yet known, but this does provide a unifying connection between these familial forms and the glutamate excess that has been identified in sporadic human ALS. Further support for the role of glutamate toxicity in the pathogenesis of ALS is that the glutamate AMPA receptor antagonists prolong survival in the G93A transgenic mouse line.61 Recently, a line of transgenic rats was created using the G93A SOD1 mutation.28 These animals thus far have confirmed the pathology and clinical features of the murine models, with evidence thus far supporting the loss of glial glutamate transporter GLT1 (equivalent to human glial transporter EAAT2) and glutamate.19 The transgenic mutant SOD1 animals described above all share the feature that the mutant SOD1 is expressed in all tissues. It is not known where the toxic effects of the mutant SOD1 are occurring. It has long been assumed that the mutant SOD1 was exerting its toxic effects in the motor neurons. More recently a transgenic animal model was developed that expressed the SOD1 G37R mutation only within the neurons. These animals failed to develop any neurologic symptoms and had no demonstrable loss of motor neurons on pathology.51 These findings were confirmed in the neuron-specific expression of the G93A and the G85R mutations as well.39 In those animals accumulation of the mutant SOD1 was demonstrated in the motor neurons. Despite this accumulation no motor neuron loss occurred. In another model, specific expression of SOD1 G86R mutation only within astrocytes also failed to produce any neurologic symptoms or motor neuron loss, but did produce gliosis.22 These findings strongly suggest that interactions between motor neurons and other cell lines are critical for the development of motor neuron degeneration and that mutant SOD1’s toxic effects within the motor neurons in isolation are insufficient for the development of motor neuron disease.

NEUROFILAMENT TRANSGENIC MODELS Neurofilament aggregation within the perikaryon and proximal axons is one of the pathologic hallmarks of ALS. These aggregated neurofilaments are phosphorylated, a finding normally only found in neurofilaments present within the axon.43 This has suggested abnormal axonal transport occurring in ALS. It has long been questioned whether these findings contribute directly to the pathogenesis of the disease or they are merely secondary results of the dying neuron. Neurofilament accumulation is present in other neurodegenerative disorders as well, and it has been proposed that disruption of neurofilaments and axonal transport may be a common mechanism of neuronal death in neurodegeneration.43 Neurofilament proteins represent a heteropolymer with three principal component proteins named by their respective molecular weights: neurofilament heavy (NF-H), neurofilament intermediate or medium (NF-M), and neurofilament light (NF-L). NF-L combines with either NF-H or NF-M to form a dimer protein that forms the neurofilaments. The principal function of the neurofilaments is in maintenance and support of the cytoskeleton. Neurofilaments that are phosphorylated are transported down into the axon, where they maintain the size and caliber of the axon. Neurofilaments that are unphosphorylated remain in the cell body serving their function there. The quivering quail is an animal that is deficient in neurofilament proteins. These animals possess a nonsense mutation in the NF-L gene.46 The quivering quail demonstrates axonal atrophy with a reduced-caliber axon. However, these animals do not develop motor neuron degeneration.46,70 A number of transgenic models affecting axonal transport with pathology similar to ALS have now been developed. These include transgenic expression of neurofilament subunits and transgenic expression of other axonal transport proteins. The neurofilament models share the common feature of disrupting the normal stoichiometric balance between the three subunits. This results in an accumulation of one set of the subunits, presumably leading to their aggregation and pooling. Three principal models have been developed: NF-L overexpressor, NF-H overexpressor, and the NF-L mutant protein. Although each of these shares important pathologic and clinical features with human ALS, none fully reproduces the human disease as well as the SOD1 transgenic mice. The NF-L mutation represents the closest of the three, sharing the most similarities to human ALS. NF-L transgenic mice were created by combing the mouse sarcoma virus promoter with the NF-L gene.69 This results in an overexpression of NF-L and a two- to fourfold increase in the amount of NF-L present. These mice died within a few weeks of birth. If carefully nurtured, however, they could recover and mature. Pathologically these mice developed changes that are strikingly similar to those of ALS. They possess neurofilament accumulation and

Transgenic Animal Models of Amyotrophic Lateral Sclerosis

swelling of the perikaryon. These neurofilaments also demonstrate phosphorylation. Neurofilament spheroids are present in the proximal axon. All of these are features that occur in human ALS. There are, however, a number of important differences. Similar aggregation occurs within the dorsal root ganglia and other neurons that are spared in ALS. Furthermore, although these animals do develop some motor neuron loss, they do not develop a progressive motor neuron degeneration. Two types of NF-H transgenic mice were created. At low levels of overexpression of murine NF-H, the transgenic mice bear little resemblance to human motor neuron disease. There is accumulation of NF-H in the axon and cell body leading to an increase in the caliber of the axon. There is no motor neuron loss, no muscle weakness, and no discernible functional motor impairment.40 Overexpression of human NF-H in the mouse, however, does produce a syndrome resembling ALS. In these mice muscle weakness and functional motor disturbances do occur. Pathologically, swelling of the axon and cell body occurs secondary to neurofilament accumulation, with atrophy of the distal axons.10 In addition to these findings, there is a loss of microtubules and mitochondria in the distal axon, suggesting that there is a disruption in axonal transport.9 As with the NF-L overexpressors, these animals demonstrate a number of differences from human ALS. These changes are identified in the dorsal root ganglia, and as with the NF-L overexpressors, these animals do not develop progressive motor neuron loss nor do they develop paralysis. The neurofilament model with the closest resemblance to human ALS is the NF-L mutant transgenic mouse. This animal was created with a proline-for-leucine amino acid substitution in the ␣-helical portion of the NF-L subunit.37 Although this mutation has never been identified in human ALS, it was chosen because of the likelihood of its interfering with the formation of the neurofilament heteromer. In these mice expression of all of the neurofilament subunits (NF-H, NF-M, and NF-L) is reduced. These mice develop the pathologic features of neurofilament accumulation, with phosphorylated subunits identified within the stroma of the cell body. There are microglial nodules and an astrocytosis. All are features in common with ALS. In contrast to the other neurofilament models, however, in this model the changes are relatively selective to the ventral roots and anterior horn cells, with minimal involvement of the dorsal horn and dorsal roots. These animals do develop a progressive syndrome of weakness and atrophy with associated loss of the anterior horn cells and die prematurely; there do not appear to be any changes in the corticospinal tracts. Nonetheless, these changes that occur in this model closely mimic those of the SOD1 transgenic mice and human ALS, making this a relevant animal in the study of this disease. Finally, a transgenic mouse model that disrupts axonal transport directly has been created. This model causes overexpression of dynamitin, a protein required for

1589

microtubule-based slow and retrograde axonal transport through the dynein/dynactin complex interaction.36 In this model there is late degeneration of the motor neurons, weakness, atrophy, and premature death that is relatively selective to the motor neurons. The pathology of this model suggests many similarities to that of ALS. Most notable is the presence of neurofilament accumulations. These findings support the theory that disruption of axonal transport is sufficient to cause motor neuron degeneration and that it is not a secondary process.

DOUBLE TRANSGENIC MODELS Crossing lines of transgenic SOD1 murine lines with other transgenic lines, creating doubly transgenic animals, has been done to evaluate interactions between the SOD1 mutations and other cellular proteins. Most notable of these are the double transgenic models expressing mutant human SOD1 and various neurofilament protein subcomponents. Because the individual expression of the SOD1 mutations and that of the neurofilament proteins both produced motor neuron degeneration, it was speculated that a double transgenic model would have more severe degeneration and a more rapid course. Paradoxically, the double transgenic models all demonstrated a protective effect when the two proteins were overexpressed together. These findings were confirmed in multiple models utilizing different SOD1 mutations and neurofilament subunits. In one double transgenic model expressing the G37R SOD1 mutation and NF-H subunit, there was prolonged survival of the transgenic mice by 6 months over the G37R SOD1 strain not expressing NF-H.11 These findings were unexpected but were confirmed in subsequent double transgenic models developed independently. Two other double transgenic strains possessing the G93A SOD1 mutation and either NF-L or NF-H demonstrated the survival benefit of this dual expression.32 In both of these double transgenic models, there was a prolongation of mouse survival of over 1 month. Although the magnitude of this benefit is less than that identified in the G37R mutation, the clinical course of the G37R model evolves more slowly than the G93A model, making the relative benefit similar. What protective effect the neurofilament proteins are conferring to the SOD1 mutant mice remains indeterminate. It has been theorized that the neurofilament aggregation acts as a cellular sink for excessive oxidative free radicals to react with, thus preventing them from exerting their damaging effects on more critical proteins. Another hypothesis is that the excessive neurofilament proteins may possess excess calcium binding sites, providing a buffer for excessive intracellular calcium, again limiting the effect of calciummediated toxicity. In support of this hypothesis is the report that calbindin D28K, an intracellular calcium-binding protein, protects PC12 cells from SOD1 toxicity.21

1590

Diseases of the Peripheral Nervous System

TRANSGENIC MODELS AND APOPTOSIS It has long been proposed that motor neuron death in ALS proceeds by programmed cell death, or apoptosis. Experimental evidence in support of apoptosis in the transgenic ALS models is mixed. Apoptosis or programmed cell death is thought to occur when the balance between apoptotic pathways and cell survival pathways is shifted in favor of apoptosis. Although many activities are involved in apoptosis, the best understood pathways leading to programmed cell death are those involving the Bcl-2 and caspase families (Fig. 66–3). The transgenic mice have now been extensively studied in reference to these two families and their functions. The Bcl-2 family consists of a number of proteins that affect the survival balance of the cell. Bcl-2 and Bcl-XL act as suppressors to programmed cell death when activated, whereas Bax, Bad, Bak, and Bcl-XS act as promoters of programmed cell death when activated. In the G93A mouse model, Bax mRNA is increased in its expression among spinal cord motor neurons.45,64 In addition to the mRNA being increased, the active Bax protein has also been identified in these animals.64 Similar findings were also identified in the G86R transgenic mouse.23 These findings are supportive that programmed cell death is activated in motor neurons in ALS. Further support that the Bcl-2 family and programmed cell death are important in ALS is that overexpression of Bcl-2 (a programmed cell death suppressor) prolonged survival of G93A transgenic mice.33 It is unlikely that the SOD1 mutations directly activate the Bcl-2 family pathways. Rather, it is likely that these mutations trigger a series of intracellular reactions that lead to the activation of the pro-apoptotic protein Bax. The p53 tumor suppressor protein is one of the few factors known to activate the Bax cascade. This protein has been found to be upregulated in the G86R transgenic mouse and has been identified in the neuronal nuclei (the active site for p53).23 However, confirmatory evidence in support

Mitochondria

Casapse-1

Bax Cytochrome C

Procaspase-9

Caspase-9

Procaspase-3 FIGURE 66–3 Apoptosis pathways.

IL-1 activation

Caspase-3

Cell death

of p53-mediated programmed cell death in SOD1 transgenic mice has not yet been provided. A second family of proteins active in programmed cell death is the caspase family. While the Bcl-2 family appear to be involved in signaling apoptosis, the caspase proteins function as the effectors of apoptosis. The caspases are cysteine proteases that cleave their target proteins after aspartic acid residues. The caspases exist principally in an inactive form and require activation before being effective. The principal function of caspase-1 is to activate interleukin-1. Caspase-1 has been extensively studied in SOD1 transgenic mice. Caspase-1 activation is known to occur concurrently with the earliest findings in the SOD1 transgenic animals.47,65 Caspase-9 is a pro-apoptotic protein whose effect is mitochondrial dependent. Caspase-9 is activated via downstream effects of the release of mitochondrial cytochrome c. In G93A transgenic mice it is known that cytochrome c is released from the mitochondria and is present in the cytosol and that caspase-9 is activated in the motor neurons.25 Furthermore, minocycline (an agent that has been shown to prolong survival in SOD1 transgenic mice) has been shown to inhibit the release of mitochondrial cytochrome c into the cytosol.71 These findings suggest a pathologic role for caspase-9 in SOD1 transgenic mice.

ANIMALS MODELS IN SCREENING FOR NEW THERAPEUTIC AGENTS Accurate animal models are an important step in developing effective therapies in ALS. These animal models serve two principal purposes: the elucidation of the etiologies of ALS and the subsequent testing of new treatments. Historically, attempts at treatment trials in ALS have been undertaken with little scientific support. This was in large part because there was no controlled manner in which to study mechanisms of action. Early animal models of ALS have been used to screen for potential new treatments. For example, the wobbler mouse and the progressive motor neuronopathy (PMN) mouse were used extensively in the early 1990s to screen for neurotrophic factors that may be beneficial in promoting motor neuron survival.42,55,56 Unfortunately, although many agents appear to be effective in these models, none of the agents that subsequently entered into human clinical trials demonstrated benefit. This is not surprising because the wobbler mouse and the PMN mouse, while sharing a number of features with ALS, do not accurately reproduce ALS in mice. Screening of potential new therapies has been dramatically changed with the use of the transgenic mouse models. By placing the SOD1 mutation transgenes into mice, for the first time there is an unequivocal ALS model in mice. Although these SOD1 mutations only account for about 1% to 2% of all human ALS, there is good rationale for using them in screening compounds for treatment in human ALS.

Transgenic Animal Models of Amyotrophic Lateral Sclerosis

In humans, familial forms of ALS and sporadic forms of ALS are indistinguishable both clinically and pathologically. This implies that ALS has heterogeneous causes but that all share the same final common pathway(s) to neuronal death. It is very likely that sporadic and familial ALS share a common set of cellular and molecular pathways regardless of what the upstream triggers are. Transgenic animals are now widely available that can be used to screen for possible treatments relatively quickly and, compared to human clinical trials, relatively cheaply. Presently this model has been used to test efficacy of a large number of compounds, many of which have demonstrated modest effects in prolonging mouse survival. However, none has halted or reversed the clinical syndrome in any of the transgenic mouse models. In many instances the experience in the SOD1 transgenic mice has mirrored the experience in human ALS. Most notably, riluzole (an agent that has a number of antiglutamate effects) was shown in human clinical trials to modestly prolong survival.2,35 In G93A transgenic mice, riluzole has been shown to prolong survival at an order of magnitude that is comparable to the human clinical trial studies.26 Furthermore, in the same set of experiments, vitamin E (a potent antioxidant) was not found to have any effect on survival but did delay the onset of the disease. Gabapentin, which has had mixed results in human clinical trials, did benefit survival but with less magnitude than riluzole. As a result of this preclinical experience in transgenic SOD1 mice, these animals have become the current model of choice to screen new potential drug treatments. The current approach is to filter new compounds through a variety of biologic assays that have been developed to represent pathways that may contribute to the pathogenesis of ALS. As an example, there are now a large number of assays directed as assessing various apoptosis pathways. Agents that demonstrate the greatest effect in these assays may be taken to the animal model and tested for efficacy and animal safety data. Those compounds that appear effective in the animal may then be advanced to human Phase I to III clinical trials. A number of agents have now completed this process and have entered into human clinical trials. A large number of agents have been tested in the SOD1 mouse, and many have been shown to either delay the onset of the disease or prolong survival of the mice. Agents that have demonstrated efficacy in this animal model have varied from nutritional supplements such as creatine and ginseng root29,31 to Food and Drug Administration–approved medications such as minocycline and cyclooxygenase-2 inhibitors used off label17,34,50,62 to truly novel agents such as caspase-1 and caspase-3 inhibitors.38 The mode of delivery has varied from oral ingestion to intraventricular administration to viral vector delivery. All these agents appear to have a modest effect on the course of the disease, either disease onset or survival; however, there is no apparent unifying mechanism of action. In general these agents have a wide range of mechanisms, many of which are seemingly unrelated. Despite the

1591

fact that a large number of agents have demonstrated benefit in the SOD1 animal models, none of these agents when tested in human clinical trials has yet shown efficacy in human ALS. There are a number of possible explanations for the failure of these agents in human clinical trials. First, the effects of these drugs in the mice are quite modest, and most of the human clinical trials have been underpowered to detect such a modest effect. Second, the animal trials largely began treatment in the asymptomatic stage of the disease, an option that is not available in human clinical trial research. Furthermore, it is possible that these drugs are working on mechanisms of action that are relevant to the SOD1 mutations but not to sporadic ALS as a whole. It is not clear yet why so many of these agents that appear to be effective in the SOD1 animal models have not proven effective for human ALS. This experience raises questions as to the value of the SOD1 mutations as a screening tool for potential new therapies. Nonetheless, the SOD1 animal models remain the most valid animal models of ALS to date. It is very likely that future efforts will be directed at specific cellular pathways and treatments will be directed at more specific mechanisms of cell death, thereby increasing the likelihood of success of future human clinical trials.

REFERENCES 1. Beckman, J. S., Carson, A. M., Smith, C. D., and Koppenol, W. H.: ALS, SOD, and peroxynitrite. Nature 364:584, 1993. 2. Bensimon, G., Lacomblezs, L., and Meininger, V., for the ALS/Riluzole Study Group: A controlled trial of riluzole in amyotrophic lateral sclerosis. N. Engl. J. Med. 330:585, 1994. 3. Blondet, B., Ait-Ikhlef, A., Murawshy, M., and Rieger, F.: Transient massive DNA fragmentation in nervous system during the early course of a murine neurodegenerative disease. Neurosci. Lett. 305:202, 2001. 4. Boillee, S., Viala, L., Peschansk, M., and Dreyfus, P. A.: Differential microglial response to progressive neurodegeneration in the murine mutant Wobbler. Glia 33:277, 2001. 5. Briston, L. A., and Rothstein, J. D.: Glutamate transporter gene expression in amyotrophic lateral sclerosis motor cortex. Ann. Neurol. 39:676, 1996. 6. Bruijn, L. I., Beal, M. F., Becher, M. W., et al.: Elevated free nitrotyrosine levels, but not protein-bound nitrotyrosine or hydroxyl radicals, throughout amyotrophic lateral sclerosis (ALS)-like disease implicate tyrosine nitration as an aberrant in vivo property of one familial ALS-linked superoxide dismutase 1 mutant. Proc. Natl. Acad. Sci. U. S. A. 94:7606, 1997. 7. Bruijn, L. I., Becher, M. W., Lee, M. K., et al.: ALS-linked SOD1 mutant G85R mediates damage to astrocytes and promotes rapidly progressive disease with SOD-1 containing inclusions. Neuron 18:327, 1997. 8. Clarke, P. G. H.: Developmental cell death: morphological diversity and multiple mechanisms. Anat. Embryol. 181:195, 1990.

1592

Diseases of the Peripheral Nervous System

9. Collard, J.-F., Cote, F., and Julien, J.-P.: Defective axonal transport in a transgenic mouse model of amyotrophic lateral sclerosis. Nature 375:61, 1995. 10. Cote, F., Collard, J.-F., and Julien, J.-P.: Progressive neuronopathy in transgenic mice expressing the human neurofilament heavy gene: a mouse model of amyotrophic lateral sclerosis. Cell 73:35, 1993. 11. Couillard-Despres, S., Zhu, Q., Wong, P. C., et al.: Protective effect of neurofilament heavy gene overexpression in motor neuron disease induced by mutant superoxide dismutase. Proc. Natl. Acad. Sci. U. S. A. 95:9626, 1998. 12. Dal Canto, M. C., and Gurney, M. E.: Development of central nervous system pathology in a murine transgenic model of human amyotrophic lateral sclerosis. Am. J. Pathol. 145:1271, 1994. 13. Dal Canto, M. C., and Gurney, M. E.: Neuropathological changes in two lines of mice carrying a transgene for mutant human Cu,Zn SOD, and in mice overexpressing wild type human SOD: a model of familial amyotrophic lateral sclerosis (FALS). Brain Res. 676:25, 1995. 14. Deng, H. X., Hentati, A., Tainer, J. A., et al.: Amyotrophic lateral sclerosis and structural defects in Cu,Zn superoxide dismutase. Science 261:1047, 1993. 15. Deng, H. X., and Siddique, T.: Transgenic mouse models and human neurodegenerative disorders. Arch. Neurol. 57:1695, 2000. 16. Des Portes, V., Coulpier, M., Melki, J., and Dreyfus, P.: Early detection of mouse wobbler mutation: a model of pathological motoneuron death. Neuroreport 5:1861, 1994. 17. Duchen, L. W., and Strick, S. J.: An hereditary motor neurone disease with progressive denervation of muscle in the mouse: the mutant “Wobbler.” J. Neurol. Neurosurg. Psychiatry 31:535, 1968. 18. Dunlop, J., McIlvain, H. B., She, Y., and Howland, D. S.: Impaired spinal cord glutamate transport capacity and reduced sensitivity to riluzole in a transgenic superoxide dismutase mutant rat model of amyotrophic lateral sclerosis. J. Neurosci. 23:1688, 2003. 19. Drachman, D. B., Frank, K., Dykes-Hoberg, M., et al.: Cyclooxygenase 2 inhibition protects motor neurons and prolongs survival in a transgenic mouse model of ALS. Ann. Neurol. 52:771, 2002. 20. Facchinetti, F., Sasaki, M., Cutting, F. B., et al.: Lack of involvement of neuronal nitric oxide synthetase in the pathogenesis of a transgenic mouse model of familial amyotrophic lateral sclerosis. Neuroscience 90:1483, 1999. 21. Ghadge, G. D., Lee, J. P., Bindokas, V. P., et al.: Mutant superoxide dismutase-1-linked familial amyotrophic lateral sclerosis: molecular mechanisms of neuronal death and protection. J. Neurosci. 17:8756, 1997. 22. Gong, Y. H., Parsadanian, A. S., Andreeva, A., et al.: Restricted expression of G86R Cu/Zn superoxide dismutase in astrocytes results in astrocytosis but does not cause motoneuron degeneration. J. Neurosci. 20:660, 2000. 23. Gonzalez de Aguilar, J. L., Gordon, J. W., Rene, F., et al.: Alteration of the Bcl-x/Bax ratio in a transgenic mouse model of amyotrophic lateral sclerosis: evidence for the implication of the p53 signaling pathway. Neurobiol. Dis. 7:406, 2000. 24. Gonzalez Densielle, M. C., Gonzalez, S., and De Nicola, A. F.: Cellular basis of steroid neuroprotection in the Wobbler

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

mouse, a model for motoneuron disease. Cell. Mol. Neurobiol. 21:237, 2001. Guegan, C., Vila, M., Rosoklija, G., et al.: Recruitment of the mitochondrial-dependent apoptotic pathway in amyotrophic lateral sclerosis. J. Neurosci. 21:6569, 2001. Gurney, M. E., Pu, H., Alexander, D. D., et al.: Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science 264:1772, 1994. Gurney, M. E., Cutting, F. B., Zhai, P., et al.: Benefit of vitamin E, riluzole, and gabapentin in a transgenic model of familial amyotrophic lateral sclerosis. Ann. Neurol. 39:147, 1996. Howland, D. S., Liu, J., She, Y., et al.: Focal loss of the glutamate transporter EAAT2 in a transgenic rat model of SOD1 mutant-mediated amyotrophic lateral sclerosis (ALS). Proc. Natl. Acad. Sci. U. S. A. 99:1604, 2002. Jiang, F., DeSilva, S., and Turnbull, J.: Beneficial effect of ginseng root in SOD-1 (G93A) transgenic mice. J. Neurol. Sci. 180:52, 2000. Kaupman, K., Simon-Chazottes, D., Guener, J. L., and Jokusch, H.: Wobbler, a mutation affecting motoneuron survival and gonadal functions in the mouse, maps to proximal chromosome 11. Genomics 13:39, 1992. Klivenyi, P., Ferrante, R. J., Matthews, R. T., et al.: Neuroprotective effects of creatine in a transgenic animal model of amyotrophic lateral sclerosis. Nat. Med. 5:347, 1999. Kong, J., and Zuoshang, X.: Overexpression of neurofilament subunit NF-L and NF-H extends survival of a mouse model for amyotrophic lateral sclerosis. Neurosci. Lett. 281:72, 2000. Kostic, V., Jackson-Lewis, V., De Bilbao, F., et al.: Bcl-2: prolonging life in a transgenic mouse model of familial amyotrophic lateral sclerosis. Science 277:559, 1997. Kriz, J., Nguyen, M. D., and Julien, J.-P.: Minocycline slows disease progression in a mouse model of amyotrophic lateral sclerosis. Neurobiol. Dis. 10:268, 2002. Lacomblez, L., Bensimon, G., Leigh P. N., et al., for the ALS/Riluzole Study Group II: Dose-ranging study of riluzole in amyotrophic lateral sclerosis. Lancet 347:1425, 1996. LaMonte, B. H., Wallace, K. E., Holloway, B. A., et al.: Disruption of dynein/dynactin inhibits axonal transport in motor neurons causing late-onset progressive degeneration. Neuron 34:715, 2002. Lee, M. K., Marszalek, J. R., and Cleveland, D. W.: A mutant neurofilament subunit causes massive, selective motor neuron death: implication for the pathogenesis of human motor neuron disease. Neuron 13:975, 1994. Li, M., Ona, V. O., Guegan, C., et al.: Functional role of caspase-1 and caspase-3 in an ALS transgenic mouse model. Science 288:335, 2000. Lino, M. M., Schneider, C., and Caroni, P.: Accumulation of SOD1 mutants in postnatal motoneurons does not cause motoneuron pathology or motoneuron disease. J. Neurosci. 22:4825, 2002. Marszalek, J. R., Williamson, T. L., Lee, M. K., et al.: Neurofilament subunit NF-H modulates axonal diameter by selectively slowing neurofilament transport. J. Cell Biol. 135:711, 1996.

Transgenic Animal Models of Amyotrophic Lateral Sclerosis 41. Mitsumoto, H., and Bradley, W. G.: Murine motor neuron disease (the Wobbler mouse): degeneration and regeneration of the lower motor neuron. Brain Res. 105:811, 1982. 42. Mitsumoto, H., Ikeda, K., Klinkosz, B., et al.: Arrest of motor neuron disease in wobbler mice cotreated with CNTF and BDNF. Science 265:1107, 1994. 43. Morrison, B. M., Morrison, J. H., and Gordon, J. W.: Superoxide dismutase and neurofilament transgenic models of amyotrophic lateral sclerosis. J. Exp. Zool. 282:32, 1998. 44. Mourelatos, Z., Gonatas, N. K., Stieber, A., et al.: The Golgi apparatus of spinal cord motor neurons in transgenic mice expressing mutant Cu,Zn superoxide dismutase becomes fragmented in early, preclinical stages of the disease. Proc. Natl. Acad. Sci. U. S. A. 93:5472, 1996. 45. Mu, X., He, J., Anderson, D. W., et al.: Altered expression of bcl-2 and bax mRNA in amyotrophic lateral sclerosis spinal cord motor neurons. Ann. Neurol. 40:379, 1996. 46. Ohara, O., Gahara, Y., Miyake, T., et al.: Neurofilament deficiency in quail caused by nonsense mutation in neurofilament-L gene. J. Cell Biol. 121:387, 1993. 47. Pasinelli, P., Houseweart, M. D., Brown, R. H. Jr., and Cleveland, D. W.: Caspase-1 and -3 are sequentially activated in motor neuron death in Cu,Zn superoxide dismutase-mediated familial amyotrophic lateral sclerosis. Proc. Natl. Acad. Sci. U. S. A. 97:13901, 2000. 48. Pernas-Alonso, R., Perrone-Capano, C., Volopicelli, F., and di Porzio, U.: Regionalized neurofilament accumulation and motoneuron degeneration are linked phenotypes in wobbler neuromuscular disease. Neurobiol. Dis. 8:581, 2001. 49. Pogun, S., Dawson, V., and Kahur, M. J.: Nitric oxide inhibits 3 H-glutamate transport in synaptosomes. Synapse 18:21, 1994. 50. Pompl, P. N., Ho, L., Bianchi, M., et al.: A therapeutic role for cyclooxygenase-2 inhibitors in a transgenic mouse model of amyotrophic lateral sclerosis. FASEB J. 17:725, 2003. 51. Pramatarova, A., Laganiere, J., Roussei, J., et al.: Neuronspecific expression of mutant superoxide dismutase 1 in transgenic mice does not lead to motor impairment. J. Neurosci. 21:3369, 2001. 52. Rathke-Hartleib, S., Schmidt, V., Jockusch, H., et al.: Spatiotemporal progression of neurodegeneration and glia activation in the wobbler neuropathy of the mouse. Neuroreports 10:3411, 1999. 53. Reaume, A. G., Elliott, J. L., Hoffman, E. K., et al.: Motor neurones in Cu/Zn superoxide dismutate-deficient mice develop normally but exhibit enhanced cell death after axonal injury. Nat. Genet. 13:43, 1996. 54. Rosen, D. R., Siddique, T., Patterson, D., et al.: Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 362:59, 1993. 55. Sagot, Y., Tan, S. A., Hammang, J. P., et al.: GDNF slows loss of motor neurons but not axonal degeneration or premature death of pmn/pmn mice. J. Neurosci. 16:2335, 1996. 56. Sendtner, M., Schmalbruch, H., Stockli, K. A., et al.: Ciliary neurotrophic factor prevents degeneration of motor neurons in mouse-mutant progressive motor neuronopathy. Nature 358:502, 1992. 57. Siddique, T., Figlewicz, D. A., Pericak-Vance, M. A., et al.: Linkage of a gene causing familial amyotrophic lateral sclerosis

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

1593

to chromosome 21 and evidence of genetic heterogeneity. N. Engl. J. Med. 324:1381, 1991. Subramaniam, J. R., Lyons, W. E., Liu, J., et al.: Mutant SOD1 causes motor neuron disease independent of copper chaperone-mediated copper loading. Nat. Neurosci. 5:301, 2002. Tu, P., Raju, P., Robinson, K. A., et al.: Transgenic mice carrying a human mutant superoxide dismutase transgene develop neuronal cytoskeletal pathology resembling human amyotrophic lateral sclerosis lesions. Proc. Natl. Acad. Sci. U. S. A. 93:3155, 1996. Tu, P., Gurney, M. E., Julien, J., et al.: Oxidative stress, mutant SOD1, and neurofilament pathology in transgenic mouse models of human motor neuron disease. Lab. Invest. 76:441, 1997. Van Damme, P., Leyssen, M., Callewaert, G., et al.: The AMPA receptor antagonist NBQX prolongs survival in a transgenic mouse model of amyotrophic lateral sclerosis. Neurosci. Lett. 343:81, 2003. VanDenBosch, L., Tilkin, P., Lemmens, G., and Robberecht, W.: Minocycline delays disease onset and mortality in a transgenic model of ALS. Neuropharmacol. Neurotox. 13:1067, 2002. Verge, V. M. K., Xu, Z., Xu, X., et al.: Marked increase in nitric oxide synthase mRNA in rat dorsal root ganglia after peripheral axotomy: in situ hybridization and functional studies. Proc. Natl. Acad. Sci. U. S. A. 89:11617, 1992. Vukosavic, S., Dubois-Dauphin, M., Romero, N., and Przedborksi, S.: Bax and Bcl-2 interaction in a transgenic mouse model of familial amyotrophic lateral sclerosis. J. Neurosci. 73:2460, 1999. Vukosavic, S., Stefanis, L., Jackson-Lewis, V., et al.: Delaying caspase activation by Bcl-2: a clue to disease retardation in a transgenic mouse model of amyotrophic lateral sclerosis. J. Neurosci. 20:9119, 2000. Wiedau-Pazos, M., Goto, J. J., Rabizadeh, S., et al.: Altered reactivity of superoxide dismutase in familial amyotrophic lateral sclerosis. Science 271:515, 1996. Wong, P. C., Pardo, C. A., Borchelt, D. R., et al.: An adverse property of a familial ALS-linked SOD1 mutation causes motor neuron disease characterized by vacuolar degeneration of mitochondria. Neuron 14:1105, 1995. Wu, Y., Li, Y., Liu, H., and Wu, W.: Induction of nitric oxide synthase and motoneuron death in newborn and early postnatal rats following spinal root avulsion. Neurosci. Lett. 194:109, 1995. Xu, Z., Cork, L. C., Grifin, J. W., and Cleveland, D. W.: Increased expression of neurofilament subunit NF-L produces morphological alterations that resemble the pathology of human motor neuron disease. Cell 73:23, 1993. Yamasaki, H., Bennett, G. S., Itakura, C., and Mizutani, M.: Defective expression of neurofilament protein subunits in hereditary hypotrophic axonopathy of quail. Lab. Invest. 66:734, 1993. Zhu, S., Stavrovskaya, I. G., Drozda, M., et al.: Minocycline inhibits cytochrome c release and delays progression of amyotrophic lateral sclerosis in mice. Nature 417:74, 2002.

67 Inherited Motor Neuron Degeneration AMMAR AL-CHALABI AND P. NIGEL LEIGH

Background and History of Familial Amyotrophic Lateral Sclerosis Models of Inheritance First Descriptions Pathology, Epidemiology, Clinical Features, and Management

Genetics of Amyotrophic Lateral Sclerosis Familiality Superoxide Dismutase 1 Amyotrophic Lateral Sclerosis Type 2 SETX

BACKGROUND AND HISTORY OF FAMILIAL AMYOTROPHIC LATERAL SCLEROSIS Models of Inheritance In discussing inherited conditions, some consideration must be given to what is meant by the term inherited. For example, single-gene disorders that have a clear inheritance pattern, such as cystic fibrosis, are obviously inherited. The situation for multigene or complex traits is less straightforward. Is height inherited? There is clearly a heritable component, but it seems unlikely to be under single-gene control. Mendelian traits are phenotypes that follow Mendel’s laws of segregation. These are traits that are dominant or recessive. Other traits are the product of a liability to disease that is the result of the action of several genes and environment. This is also heritable, but does not manifest unless a threshold of liability is exceeded. This is complex genetics. In this chapter we concentrate on Mendelian motor neuron degeneration, but we also widen the discussion where appropriate to include complex genetics as manifested by familial clustering or even sporadic disease if there is a significant heritable component. In general, however, these aspects are covered in Chapter 63. The concepts of genetics were being realized at the same time as the earliest descriptions of motor neuron diseases. Mendel published his classic paper in 1865.28 The word “genetics” was not used until 1905 by Bateson in a letter

Neurofilament Heavy Subunit Dynactin X-Linked Motor Neuron Disease Loci Linked to Autosomal Dominant FALS

to the Cambridge zoologist Adam Sedgwick, and later in public in 1906, again by Bateson.9

First Descriptions Possibly the earliest recorded case of familial amyotrophic lateral sclerosis (FALS) is that reported by the French physician Aran in 1850.8 In his paper describing progressive muscular atrophy, he reported the case of a mariner presenting with upper limb cramps followed by wasting that spread to involve all limbs. This man died 2 years after symptom onset, and had a sister and two maternal uncles affected by a similar condition. This would be consistent with autosomal dominant inheritance, a concept that had not yet come into being. Amyotrophic lateral sclerosis (ALS) itself was formally described in 1869 by Charcot and Joffroy,13 although they did not believe that it was inherited. This may be because it was not obvious that progressive muscular atrophy and ALS could be part of a single disease entity, and because Charcot and Joffroy described only 20 cases, it is quite possible they were not aware of any familial ALS among their patients.

PATHOLOGY, EPIDEMIOLOGY, CLINICAL FEATURES, AND MANAGEMENT Pathology studies support the notion that sporadic and familial ALS should not be regarded as separate conditions. In general the pathology of sporadic and familial ALS is 1595

1596

Diseases of the Peripheral Nervous System

identical. Some familial cases show involvement of the posterior and Clarke’s columns as well as the spinocerebellar tracts, but there is no way to distinguish familial from sporadic ALS pathologically. Because most of the features are identical between sporadic and familial ALS, those not covered specifically here are dealt with in Chapter 63. The management of FALS is identical to that of sporadic disease except that genetic counseling is required for patients and family members. For those with a family history suggesting autosomal dominant inheritance, a copper/zinc superoxide dismutase 1 (SOD1) mutation screening may be requested (see below). For presymptomatic testing, the issue is complicated by the reduced penetrance seen for some mutations and the fact that, unless the family is known to have a SOD1 mutation, a negative result does not exclude future disease. For diagnostic testing, similar caveats apply in that, unless the family is known to carry a SOD1 mutation, a negative result does not imply a phenocopy, or that the diagnosis should be reconsidered. Even so, at least two families are known to the authors in which SOD1 mutation is seen in one member and the other affected family members are SOD1 mutation negative.

GENETICS OF AMYOTROPHIC LATERAL SCLEROSIS Familiality In about 10% of cases, ALS is inherited as an autosomal dominant trait with age-dependent penetrance, and about 20% of such families have mutations in the gene for SOD1.35,39 Familial ALS may not be obviously dominant, and it is not uncommon for cousins or more distant relations to be affected. Some pedigrees from bottleneck populations, such as those in Scandinavia, show definite familial clustering without clear Mendelian inheritance. Such families may be explained by a single gene with very low penetrance, or by more than one locus contributing to disease susceptibility. The statistical methods underlying gene-hunting in Mendelian diseases were laid down by Morton in 1955 and constitute the basis of the logarithm of odds (LOD) score method of linkage analysis.29 The technology to utilize such techniques for disease gene mapping only became widely available in the early 1990s, and ALS was one of the first diseases in which linkage was found. The last 10 years have seen an explosion in the number of genes and loci linked to ALS or related conditions (Table 67–1), and a common theme is likely to emerge in the near future. This will make finding further disease genes easier and, it is hoped, expose a pathogenic pathway that can be attacked pharmacologically. It is difficult to perform linkage studies in FALS because it is a disease of late middle life, so parents may no longer be alive and children are not at the age of risk. A three-

generation family, ideal for classic linkage, is therefore difficult to obtain.

Superoxide Dismutase 1 Mutations in the SOD1 gene remain the only proven genetic cause of ALS. To date more than 100 have been described, occurring anywhere in the five-exon gene and reported from many parts of the world.7 The most frequent SOD1 variant worldwide is D90A, followed by A4V and I113T. In the United States, A4V is the most common, and in Scandinavia D90A is the most common. It is interesting that two of the three most common variants have low penetrance, resulting in a slightly lower evolutionary selection pressure against them. The lower penetrance will also make them less likely to occur in the kind of pedigrees used for linkage analysis. If D90A and I113T had been the only mutations of SOD1 in existence, this cause of ALS might not yet have been found. The active site of SOD1 is coded for by exon 3, and this has the fewest reported mutations. Four founder studies have been performed, for D90A (international study),2,31 L84F (Italian study),11 and A4V (two studies). The D90A, L84F, and North American A4V studies have all shown a single founder for the particular mutant under study, and the North American A4V cases had a founder distinct from the small A4V families reported in Sweden and Italy. These findings suggest that SOD1 is not particularly prone to new mutation. Given that 2000 individuals per year worldwide can be expected to develop ALS as a direct result of SOD1 mutations, it could be argued that there are far fewer variants than one might expect. Typically, ALS caused by SOD1 mutations shows autosomal dominant inheritance with age-dependent penetrance. The situation is not completely straightforward, however. In about 2% of cases, individuals with sporadic ALS are found to have SOD1 mutations, often but by no means exclusively D90A or I113T. This may be because of reduced penetrance, but in at least one case is because of new mutation.4 Other than the D90A variant, no other SOD1 mutation is routinely found in a homozygous individual. Of the three reported cases from inbred families (L84F, N86S, and L126S),23,26 at least two have been associated with a particularly aggressive form of the disease, suggesting a dose effect. The D90A variant is an enigmatic mutation of the SOD1 gene.5,6,34 Some families with this mutation show a dominant pattern of inheritance of ALS as expected, often with reduced penetrance. Consistent with reduced penetrance, some individuals with this mutation in single copy develop ALS even without a family history. Such individuals and families have been found in the United Kingdom, Belgium, France, the United States, and Belorussia. They have a variable phenotype indistinguishable from sporadic ALS. More usually, however, this SOD1 variant in single copy seems to have no effect and is even polymorphic in parts of Scandinavia at a rate of up to 5%. Because of the high carrier

1597

Inherited Motor Neuron Degeneration

Table 67–1. Loci Implicated in Various Forms of Familial Amyotrophic Lateral Sclerosis (FALS) Type of FALS and Pattern of Inheritance 1. Autosomal dominant FALS

2. Autosomal dominant juvenile ALS

3. Autosomal dominant FALS with frontotemporal dementia (FTD) 4. X-linked dominant FALS 5. Autosomal recessive juvenile FALS

6. Autosomal recessive juvenile FALS 7. Autosomal dominant FALS 8. Autosomal dominant FALS

9. Autosomal dominant 10. Autosomal dominant FALS 11. Autosomal dominant juvenile ALS 12. Autosomal recessive Fazio-Londe disease (pontobulbar palsy of childhood without deafness) 13. Autosomal recessive Brown-Vialetto-Van Laere syndrome (pontobulbar palsy of childhood with sensorineural deafness) 14. Autosomal dominant familial amyotrophy with FTD and parkinsonism

Chromosomal Linkage

Gene/Protein

21q22.1-22.2 (ALS1) Over 100 mutations in SOD1 gene, mostly missense mutations, some nonsense mutations leading to truncated protein 9q34 (ALS4) Slowly progressive childhood- or adolescent-onset upper motor neuron (UMN) and lower motor neuron (LMN) syndrome 9q21-q22 (ALS-FTD) Several families with ALS, ALS plus dementia, or dementia alone; dementia is of frontotemporal type Xcentromere (ALSX) 2q33-2q35 (ALS2) Several families with slowly progressive spastic quadriparesis; some LMN signs in one family; pathology unknown

Cu/Zn superoxide dismutase 1 (SOD1)

15q15-15q22 (ALS5)

SETX

Unknown

Unknown ALS2/Alsin Alsin is a guanine nucleotide exchange factor (GEF) and is implicated in signal transduction and a variety of other cellular processes Unknown

18q21 (ALS6) 16q12 Variable penetrance; 4 families known; typical ALS, and ALS with dementia 20 (ALS8) Unknown (ALS3) (~80% of families) 20q13 Slowly progressive LMN syndrome Unknown

Unknown Unknown

Unknown Possibly same as Madras phenotype motor neuron disease

Unknown

17q21-17q22 Many phenotypes, mainly FTD, parkinsonism, and a few associated with MND

Tau gene/protein

frequency, homozygotes are not uncommon, and such individuals do tend to develop ALS, but with a characteristic and relatively benign phenotype. This mutation therefore causes autosomal dominant, autosomal recessive, and sporadic ALS. Intriguingly, families tend to breed true, so that recessive families do not contain affected heterozygous individuals. Families in which ALS is associated with homozygosity for the D90A variant and the characteristic phenotype have now also been found in Russia, France, Germany, Italy, Canada, and the United States, and there is evidence for such a family in Australia. Founder studies have shown there is a single

Unknown Unknown Unknown Unknown

founder for all cases.2,31 From among the descendants of this single founder, one has given rise to all the recessive families. This has been used as evidence that a protective factor tightly linked to the D90A variant has been passed on to the descendants of the recent founder and explains the recessive inheritance and the fact that families breed true. Another interpretation is that a single copy of the D90A variant increases liability to ALS to more than 50% of the threshold required to develop disease, but does not exceed 100% as would usually be the case for SOD1 mutations. A further copy would then be sufficient to result in ALS,

1598

Diseases of the Peripheral Nervous System

and homozygotes would therefore be affected. In affected heterozygotes, some other factor or factors (genetic or environmental) have increased the liability to beyond the threshold. A possible example comes from the case of two siblings who showed compound heterozygosity, carrying both one copy of D90A and another apparently low penetrance mutation, D96N, and thus having two doses of a liability factor for ALS.22 Both had ALS with a phenotype similar to that of homozygous D90A cases. This dose-effect model is also consistent with the evidence from dominant SOD1 homozygotes and with the liability threshold model of complex genetics. Other than a few exceptions, there is a poor correlation between genotype and phenotype.7 By far the majority of SOD1 variants are associated with a poor prognosis, and survival is often less than 3 years. The A4V mutation, common in the United States, seems to be particularly aggressive. Variants associated with slow progression, such as recessive D90A, G93D, or E100K, are possibly more frequent in exon 4. Position in the gene is not a reliable indicator of prognosis, however, and there are exon 4 variants associated with aggressive disease (e.g., D90V and I112T). There is poor correlation between mutation position and penetrance as well. Low penetrance is possibly more likely to be associated with an exon 4 variant (e.g., heterozygous D90A and I113T), but there are exon 4 variants that have complete penetrance (recessive D90A, G108V, G114A). First symptoms in the lower limbs are consistently seen in the G41D, H46R, D76V, recessive D90A, and E100K variants, often with a slowly ascending weakness. No SOD1 mutation has been reported to cause a predominantly upper motor neuron phenotype or presentation of ALS. In general, a predominantly lower motor neuron pattern is seen, although there are exceptions. Just as typical sporadic ALS shows extramotor involvement, so too can ALS caused by SOD1 mutation. Sensory symptoms (but not signs) are not unusual in SOD1-mediated ALS and may consist of paresthesias or shooting pains.33 Autonomic disturbances have also been reported, and may present with bladder disturbance or even autonomic failure.25 It may seem that any variation on SOD1 results in ALS, and if this is the case, it seems difficult to see how it could evolve without being disastrous for the organism. Every evolutionary step toward the wild type would by necessity not be quite the wild type and therefore be the equivalent of a mutated version. There are several solutions to this apparent conundrum. First, in general FALS is a disease with onset at about age 50, well after most individuals have started their families. Second, in historical terms, 50 would be considered very old age and SOD1 would be in an “evolutionary shadow.” Third, it may be that SOD1 mutations only cause neurodegeneration in a two-synapse pyramidal system. No other animal develops a condition truly like ALS (equine or canine motor neuron diseases probably have different causes). Nevertheless, with a wide

variation in age of onset seen for most SOD1 mutations, there is a negative selection pressure.

Amyotrophic Lateral Sclerosis Type 2 Recessive juvenile-onset ALS was first described in large inbred Tunisian pedigrees in 1990.10 It was designated amyotrophic lateral sclerosis type 2 (ALS2) to distinguish it from ALS type 1 (SOD1-mediated ALS) and ALS type 3 (all other causes). Three phenotypes were described in the original paper. Group 1 was characterized by upper motor neuron involvement with weakness and wasting in the upper limbs, beginning in the second decade. Group 1 families are different from the other two phenotypes described in having no pseudobulbar affect and having early hand involvement, sometimes unilateral for several years. This form has been genetically linked to chromosome 15q15-q22 and is now designated ALS type 5, but there is almost certainly genetic heterogeneity in the group 1 families. The group 2 families had a younger age of onset, in the first or second decades, of peroneal atrophy with footdrop and spastic paraplegia with absent ankle reflexes. Electromyographic studies show denervation in all limbs, with normal nerve conduction velocities. It is the group 3 families, defined by one large consanguineous pedigree with a rather uniform clinical picture of early onset (less than 10 years), spastic paraplegia, and prominent pseudobulbar findings, that were linked to chromosome 2q33 and now comprise those with mutations in the ALS2 gene, coding for alsin. Denervational atrophy was clinically minimal, although electromyography demonstrated dysfunction of lower motor neurons. There are other recessive families with a phenotype similar to the chromosome 2–linked ALS from Kuwait and from Saudi Arabia, although, other than the original Tunisian pedigree, none has amyotrophy. All these families were found to have small deletions in a 34-exon gene coding for a novel protein.18,41 The deletions result in a frameshift and premature stop codon, with a truncated protein. The ALS2 gene codes for two transcripts of a protein now called alsin, a short transcript and a long transcript. Both are ubiquitously expressed, including in the neurons of the central nervous system, although the shorter transcript is the predominant form in the liver. Interestingly, only the Tunisian deletion affects both the short and the long transcript, and this was postulated as a cause of the amyotrophy seen in the Tunisian pedigree. Since then, a further deletion also affecting the shorter transcript but not associated with amyotrophy has been reported. The ALS2 protein consists of three guanine nucleotide exchange factor (GEF) domains, a Ran GEF, a Rho GEF, and a vacuolar sorting domain motif (VPS9 GEF). GEFs are generally small proteins involved in cycling of GTP/GDP for GTPases. The protein also contains a membrane occupation and recognition nexus (MORN) motif. From this homology,

1599

Inherited Motor Neuron Degeneration

various functions have been proposed, including organization of vacuolar traffic, control of chromatin condensation, and cytoskeletal organization. It has recently been shown that alsin is a Rab5 and Rac1 GEF with the VPS9 domain showing the Rab5 activity.40 From this evidence, it appears that alsin is involved in membrane transport, possibly linking endocytosis and actin cytoskeleton remodeling. This is further evidence of a primary role for the cytoskeleton in motor neuron degeneration. Since the original reports, similar mutations have been found in families from Algeria, Italy, France, and Pakistan, bringing the number of deletions to eight.16,17 The phenotype now extends from infantile-onset ascending hereditary spastic paraplegia with bulbar involvement to juvenile primary lateral sclerosis to juvenile ALS. Recently, a nonsense mutation rather than a deletion, but also resulting in a truncated protein, has been reported in a family with infantileonset ascending hereditary spastic paraplegia with bulbar involvement.15 Full-length alsin therefore seems necessary for normal motor neuron function, but it is not clear what phenotype would result from missense mutations that do not alter transcript length. Mutations in the ALS2 gene overwhelmingly cause upper motor neuron syndromes and, except for bulbar involvement, the phenotype is perhaps closer to hereditary spastic paraplegia than to ALS. Because of the finding of SOD1 mutations in sporadic ALS, and because a recessive gene would be able to masquerade as a cause of sporadic disease, the ALS2 gene was sequenced in 12 individuals with predominantly upper motor neuron, young-onset (mean age 29, youngest 7 years), slowly progressive ALS (all except one individual were still alive, with median survival of 8 years, at the time of study). There were five single nucleotide polymorphisms in the coding region, of which all except one were conservative. This variant (1102g ⬎ a, V368M) in blade 4 of the regulation of chromatin condensation (RCC1) propeller had already been described in the original papers and does not appear to increase risk of ALS. (It should be noted that the sequence numbering in the original papers is different in that one starts at base 1 of the coding sequence and the other starts at base 1 of exon 1). Using haplotype analysis, regions of linkage disequilibrium were mapped in the gene and used to look for association with ALS in 300 cases and 300 matched controls. No difference was found, suggesting that variants in the ALS2 gene are not a common cause of typical ALS, or of predominantly upper motor neuron, young-onset, slowly progressive ALS. This has been examined in 95 unrelated familial, 95 sporadic, and 11 young-onset individuals in a Canadian study with similar results.20

other pedigrees with a similar phenotype has allowed gene mapping to proceed rapidly, and mutations have now been identified in the SETX gene.14 This has an unknown function but contains a DNA/RNA helicase domain with strong homology to RENT1 and immunoglobulin mu-binding protein 2 (IGHMBP2). These genes have a role in RNA processing (as does the survival motor neuron gene SMN1). Other mutations in the SETX gene cause a form of ataxia associated with oculomotor apraxia (AOA type 2). The three mutations so far found to cause ALS4 are point mutations scattered through the gene, whereas those causing AOA2 are homozygous null deletions except for a single compound heterozygote. Therefore, there is to some extent genotype-phenotype correlation, but whether this will hold up to further studies remains to be seen.

SETX

X-Linked Motor Neuron Disease

Juvenile ALS with very long survival (tens of years, compatible with a normal lifespan) has been linked to 9q34 in a single large pedigree (ALS type 4).12 The finding of

Kennedy’s disease is an X-linked form of motor neuron degeneration presenting as a spinal muscular atrophy with bulbar involvement. It is associated with testicular

Neurofilament Heavy Subunit Neurofilament accumulations are seen in the perikarya and proximal axons of lower motor neurons in ALS. Although it is clear that neurofilament accumulation can occur secondary to the cell death process, it now seems likely that neurofilament defects can be a primary cause of motor neuron degeneration. Investigation of more than 1000 individuals with sporadic ALS has shown about 1% to have deletions in the neurofilament heavy subunit (NEFH) tail.3 As part of studies of the NEFH tail in sporadic ALS, deletions were found in an affected individual from a family with FALS. The older brother had had focal amyotrophy of the lower limbs for more than 20 years. Consent and DNA have not been available from sufficient family members to determine if this mutation segregates with disease.

Dynactin A dominant late-childhood-onset form of predominantly lower motor neuron ALS with prominent bulbar features has recently been described. This is clinically similar to Kennedy’s disease (see later) but shows autosomal dominant inheritance. It is now known to be due to mutation in the gene for dynactin, a protein involved in axonal transport as part of the molecular motor.32 Defects in the related dynein gene have been found in two strains of mice with inherited motor neuron disease, legs at odd angles (loa) and cramping (cra).19 This is important because neurofilament defects may underlie some forms of ALS, and cytoskeletal abnormalities are a hallmark of the disease.

1600

Diseases of the Peripheral Nervous System

atrophy and gynecomastia. The feminizing features made the androgen receptor gene a good candidate for mutation screening, and in 1990, a trinucleotide repeat (CAG) expansion was found to be the cause.27 The size of the triplet expansion correlates with disease severity and younger onset. The diagnosis is important to consider because it is simple to test for and is compatible with a normal lifespan. Patients may have a tremor of the arms and facial myokymia, especially around the chin. A U.S. family with classic ALS has been linked to the chromosome X centromeric region.

Loci Linked to Autosomal Dominant FALS Recently four families, three with typical ALS and one with ALS and frontotemporal dementia, were linked to the same region of chromosome 16.1,37,38 Typical ALS has also been linked to chromosome 2038 and to chromosome 18q.21 A slowly progressive, young-onset lower motor neuron disease has been linked to chromosome 20q13.30 As noted in Chapter 63, there is intriguing overlap between pathologic and clinical features of frontotemporal dementia and ALS. ALS with frontotemporal dementia has been linked to 9q21-q22.24 There are therefore five or more loci linked to autosomal dominant FALS without a known underlying gene defect. The condition known variously as dementia-disinhibitionparkinsonism-amyotrophy complex (DDPAC) or ALS/ parkinsonism-dementia complex (ALS/PDC) is linked to chromosome 17q and is part of the spectrum of neurodegeneration caused by mutation in the tau gene.36 Similar conditions, both clinically and pathologically, occur in the Pacific island of Guam and in the Kii peninsula of Japan. Family members may have frontotemporal dementia with an ALS-like syndrome, or frontotemporal dementia with parkinsonism. The pathologic hallmark of these conditions is neurofibrillary tangles without amyloid deposition in the third layer of the cerebral cortex. The hyperphosphorylated tau in Kii peninsula ALS/PDC is very similar if not identical to that of Alzheimer’s disease. Guamanian and Kii peninsula ALS/PDC are not caused by mutation in the tau gene, and linkage to another region or environmental causes seem likely. The finding of SOD1 gene defects in both familial and sporadic cases, and the clustering of ALS seen in families without clear autosomal dominant inheritance, unmistakably argue that the distinction between sporadic and familial ALS is to some extent artificial. As linkage in more low-penetrance families is reported, and as the technology to find multiple genes simultaneously contributing to disease improves, the line between familial and sporadic ALS will blur still further.

REFERENCES 1. Abalkhail, H., Mitchell, J., Habgood, J., et al.: A new familial amyotrophic lateral sclerosis locus on chromosome 16q12.1-16q12.2. Am. J. Hum. Genet. 73:383, 2003. 2. Al-Chalabi, A., Andersen, P. M., Chioza, B., et al.: Recessive amyotrophic lateral sclerosis families with the D90A SOD1 mutation share a common founder: evidence for a linked protective factor. Hum. Mol. Genet. 7:2045, 1998. 3. Al-Chalabi, A., and Miller, C. C.: Neurofilaments and neurological disease. Bioessays 25:346, 2003. 4. Alexander, M. D., Traynor, B. J., Miller, N., et al.: “True” sporadic ALS associated with a novel SOD-1 mutation. Ann. Neurol. 52:680, 2002. 5. Andersen, P. M., Forsgren, L., Binzer, M., et al.: Autosomal recessive adult-onset amyotrophic lateral sclerosis associated with homozygosity for Asp90Ala CuZn-superoxide dismutase mutation: a clinical and genealogical study of 36 patients. Brain 119:1153, 1996. 6. Andersen, P. M., Nilsson, P., Ala-Hurula, V., et al.: Amyotrophic lateral sclerosis associated with homozygosity for an Asp90Ala mutation in CuZn-superoxide dismutase. Nat. Genet. 10:61, 1995. 7. Andersen, P. M., Sims, K. B., Xin, W. W., et al.: Sixteen novel mutations in the Cu/Zn superoxide dismutase gene in amyotrophic lateral sclerosis: a decade of discoveries, defects and disputes. Amyotroph. Lateral Scler. Other Motor Neuron Disord. 4:62, 2003. 8. Aran, F.A.: Recherches sur une maladie non encore decrite du systeme musculaire (atrophie musculaire progressive). Arch. Gen. Med. 24:172, 1850. 9. Bateson, W.: Address. In Third Conference on Hybridisation. Cambridge, UK, 1906. 10. Ben Hamida, M., Hentati, F., and Ben Hamida, C.: Hereditary motor system diseases (chronic juvenile amyotrophic lateral sclerosis): conditions combining a bilateral pyramidal syndrome with limb and bulbar amyotrophy. Brain 113:347, 1990. 11. Ceroni, M., Malaspina, A., Poloni, T. E., et al.: Clustering of ALS patients in central Italy due to the occurrence of the L84F SOD1 gene mutation. Neurology 53:1064, 1999. 12. Chance, P. F., Rabin, B. A., Ryan, S. G., et al.: Linkage of the gene for an autosomal dominant form of juvenile amyotrophic lateral sclerosis to chromosome 9q34. Am. J. Hum. Genet. 62:633, 1998. 13. Charcot, J. M., and Joffroy, A.: Deux cas d’atrophie musculaire progressive avec lesions de la substance grise et des faisceaux antero-lateraux de la moelle epiniere. Arch. Physiol. Neurol. Pathol. 2:744, 1869. 14. Chen, Y.-Z., Bennett, C. L., Huynh, H. M., et al.: DNA/RNA helicase gene mutations in a form of juvenile amyotrophic lateral sclerosis (ALS4). Am. J. Hum. Genet. 76:1128, 2004. 15. Devon, R. S., Helm, J. R., Rouleau, G. A., et al.: The first nonsense mutation in alsin results in a homogeneous phenotype of infantile-onset ascending spastic paralysis with bulbar involvement in two siblings. Clin. Genet. 64:210, 2003. 16. Eymard-Pierre, E., Lesca, G., Dollet, S., et al.: Infantile-onset ascending hereditary spastic paralysis is associated with mutations in the alsin gene. Am. J. Hum. Genet. 71:518, 2002.

Inherited Motor Neuron Degeneration 17. Gros-Louis, F., Meijer, I. A., Hand, C. K., et al.: An ALS2 gene mutation causes hereditary spastic paraplegia in a Pakistani kindred. Ann. Neurol. 53:144, 2003. 18. Hadano, S., Hand, C. K., Osuga, H., et al.: A gene encoding a putative GTPase regulator is mutated in familial amyotrophic lateral sclerosis 2. Nat. Genet. 29:166, 2001. 19. Hafezparast, M., Klocke, R., Ruhrberg, C., et al.: Mutations in dynein link motor neuron degeneration to defects in retrograde transport. Science 300:808, 2003. 20. Hand, C. K., Devon, R. S., Gros-Louis, F., et al.: Mutation screening of the ALS2 gene in sporadic and familial amyotrophic lateral sclerosis. Arch. Neurol. 60:1768, 2003. 21. Hand, C. K., Khoris, J., Salachas, F., et al.: A novel locus for familial amyotrophic lateral sclerosis, on chromosome 18q. Am. J. Hum. Genet. 70:251, 2002. 22. Hand, C. K., Mayeux-Portas, V., Khoris, J., et al.: Compound heterozygous D90A and D96N SOD1 mutations in a recessive amyotrophic lateral sclerosis family. Ann. Neurol. 49:267, 2001. 23. Hayward, C., Brock, D. J., Minns, R. A., and Swingler, R. J.: Homozygosity for Asn86Ser mutation in the CuZn-superoxide dismutase gene produces a severe clinical phenotype in a juvenile onset case of familial amyotrophic lateral sclerosis [letter]. J. Med. Genet. 35:174, 1998. 24. Hosler, B. A., Siddique, T., Sapp, P. C., et al.: Linkage of familial amyotrophic lateral sclerosis with frontotemporal dementia to chromosome 9q21-q22. JAMA 284:1664, 2000. 25. Karlsborg, M., Andersen, E. B., Wiinberg, N., et al.: Sympathetic dysfunction of central origin in patients with ALS. Eur. J. Neurol. 10:229, 2003. 26. Kato, M., Aoki, M., Ohta, M., et al.: Marked reduction of the Cu/Zn superoxide dismutase polypeptide in a case of familial amyotrophic lateral sclerosis with the homozygous mutation. Neurosci. Lett. 312:165, 2001. 27. La Spada, A. R., Wilson, E. M., Lubahn, D. B., et al.: Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352:77, 1991. 28. Mendel, G.: Versuche uber Pflanzen-Hybriden [Experiments in plant hybridization]. In Verhandlungen es naturforschenden Vereines in Brunn 4 [Natural History Society of Brunn 4]. Bohemia, Czechoslovakia, 1865. 29. Morton, N. E.: Sequential tests for the detection of linkage. Am. J. Hum. Genet. 7:277, 1955.

1601

30. Nishimura, A. L., Mitne-Neto, M., Silva, H. C., et al.: A novel locus for late onset amyotrophic lateral sclerosis/motor neurone disease variant at 20q13. J. Med. Genet. 41:315, 2004. 31. Parton, M. J., Broom, W., Andersen, P. M., et al.: D90A-SOD1 mediated amyotrophic lateral sclerosis: a single founder for all cases with evidence for a cis-acting disease modifier in the recessive haplotype. Hum. Mutat. 20:473, 2002. 32. Puls, I., Jonnakuty, C., LaMonte, B. H., et al.: Mutant dynactin in motor neuron disease. Nat. Genet. 33:455, 2003. 33. Rezania, K., Yan, J., Dellefave, L., et al.: A rare Cu/Zn superoxide dismutase mutation causing familial amyotrophic lateral sclerosis with variable age of onset, incomplete penetrance and a sensory neuropathy. Amyotroph. Lateral Scler. Other Motor Neuron Disord. 4:162, 2003. 34. Robberecht, W., Aguirre, T., Van Den Bosch, L., et al.: D90A heterozygosity in the SOD1 gene is associated with familial and apparently sporadic amyotrophic lateral sclerosis. Neurology 47:1336, 1996. 35. Rosen, D. R., Siddique, T., Patterson, D., et al.: Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis [see comments]. Nature 362:59, 1993. [Published erratum appears in Nature 364:362, 1993.] 36. Rosso, S. M., Kamphorst, W., de Graaf, B., et al.: Familial frontotemporal dementia with ubiquitin-positive inclusions is linked to chromosome 17q21-22. Brain 124:1948, 2001. 37. Ruddy, D. M., Parton, M. J., Al-Chalabi, A., et al.: Two families with familial amyotrophic lateral sclerosis are linked to a novel locus on chromosome 16q. Am. J. Hum. Genet. 73:390, 2003. 38. Sapp, P. C., Hosler, B. A., McKenna-Yasek, D., et al.: Identification of two novel loci for dominantly inherited familial amyotrophic lateral sclerosis. Am. J. Hum. Genet. 73:397, 2003. 39. Siddique, T., Figlewicz, D. A., Pericak Vance, M. A., et al.: Linkage of a gene causing familial amyotrophic lateral sclerosis to chromosome 21 and evidence of genetic-locus heterogeneity [see comments]. N. Engl. J. Med. 324:1381, 1991. [Published errata appear in N. Engl. J. Med. 325:71 and 325:524, 1991.] 40. Topp, J. D., Gray, N. W., Gerard, R. D., and Horazdovsky, B. F.: Alsin is a Rab5 and Rac1 guanine nucleotide exchange factor. J. Biol. Chem. 279:24612, 2004. 41. Yang, Y., Hentati, A., Deng, H. X., et al.: The gene encoding alsin, a protein with three guanine-nucleotide exchange factor domains, is mutated in a form of recessive amyotrophic lateral sclerosis. Nat. Genet. 29:160, 2001.

68 Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons* A. E. HARDING

Classification Diagnosis Clinical and Genetic Features Proximal Hereditary Motor Neuronopathy X-Linked Bulbospinal Hereditary Neuronopathy

Distal Hereditary Motor Neuronopathy Scapuloperoneal Hereditary Motor Neuronopathy Types I and II Facioscapulohumeral Hereditary Motor Neuronopathy Hereditary Motor Neuronopathy with Oculopharyngeal Involvement

The disorders discussed in this chapter are usually referred to as spinal muscular atrophies. Werdnig111 and Hoffmann45 independently recognized that a disease giving rise to progressive muscle wasting and weakness in infancy was due to anterior horn cell degeneration. Hoffmann first used the term spinale Muskelatrophie to distinguish this disorder from the muscular dystrophies. Werdnig-Hoffmann disease, or acute infantile spinal muscular atrophy, may be the most common form of the inherited degenerations of the lower motor neurons, but subsequent reports have described a considerable variety of such syndromes, differing in age of onset, mode of inheritance, distribution of muscle weakness, and progression. The presence of prominent bulbar involvement in some of these23,48 makes the use of the term of “spinal muscular atrophy” inaccurate. “Hereditary motor neuronopathy” (HMN) and “neuronal atrophy and degeneration of predominantly lower motor neurons” are more appropriate generic titles. *This chapter is reprinted from the third edition because the invited author was unable to meet extended deadlines for revision. A recent review of spinal muscular atrophies by Sabine Rudnik-Schöneborn and colleagues can be found in Myology, third edition, edited by A. G. Engel and C. Franzini-Armstrong (McGraw-Hill, 2004). Several tables from that chapter appear here.

Bulbar Hereditary Motor Neuronopathy Ryukyuan Hereditary Motor Neuronopathy Segmental Motor Neuronopathies Etiology and Management

CLASSIFICATION Despite some early suggestions that the hereditary motor neuronopathies were a single disorder with a wide spectrum of clinical variation, it is quite clear that these diseases are both clinically and genetically heterogeneous. Their classification is not easy, and this problem has been vigorously discussed.14,19–21,41,42,46,80,116 Most attempts have been based on the distribution of muscle weakness, age of onset, or mode of inheritance. Unfortunately, it is not always easy to establish the mode of inheritance in some cases of HMN. This is particularly true if patients have no affected relatives (sporadic, single, or singleton cases). In this situation the disorder may be of autosomal recessive inheritance, due to fresh mutation of an autosomal dominant gene or, in a male, of X-linked recessive inheritance. Occasionally a nongenetic syndrome may mimic a recognized phenotype. The classification used here (Table 68–1) is loosely based on the work of Emery19 and Pearn80 but differs from both in several respects. The number of available classifications no doubt reflects the lack of an obvious ideal one. This situation will change in the near future, particularly in relation to childhood-onset proximal HMN, because genetic linkage studies have shown that there is a locus for both acute infantile HMN and the chronic childhood form on the long arm of chromosome 5.5,33,64,65 1603

1604

Diseases of the Peripheral Nervous System

Table 68–1. Classification of Hereditary Motor Neuronopathy

Synonyms

Inheritance†

Age of Onset (years unless indicated)

AR ?XL

In utero to 6 mo

Never able to walk

7–18 mo

AR (some singletons may be new dominant mutants) AR

3 mo–15 yr

Median about 12 (from never to 5th decade)

18 mo–40 yr

15–60 (usually 20–40)

Rare 50⫹

Normal

Type IV: Juvenile onset

Werdnig-Hoffmann disease (WHD), Oppenheim disease, amyotonia congenita, SMA type I* Arrested WHD, Oppenheim disease, Kugelberg-Welander syndrome, SMA types I and II* Kugelberg-Welander syndrome, SMA type IV* SMA type III*

AD

Rare

Probably normal

Type V: Adult onset

SMA type IV*

AD

6 mo–5 yr (rarely up to 15) 25–65

? 10 yr after diagnosis

20 yr after diagnosis

X-linked spinal ⫹ bulbar muscular atrophy

XL

15–60 (usually 20–40)

50⫹

Minimally limited

Spinal form of Charcot-Marie-Tooth disease (CMT) Spinal form of CMT Spinal form of CMT Spinal form of CMT —

AD

2–20

Rare

Normal

20–40 2–10 4 mo–20 yr 5–20

Rare Rare About 30 Never

Normal Normal ? Not reduced

—

AD AR AR AD (some sporadic) AR

Infancy

⬍1

—

AD

10–20

Never able to walk Rare

Probably normal

Scapuloperoneal Type I Type II

— —

AD AR

4–70 2–5 (under 15)

? 50⫹ ?

? reduced ?

Facioscapulohumeral

—

AD

Before 20

?

?

Oculopharyngeal Bulbar Type I: With deafness

—

AD

30–40

?

?

Vialetto–van Laere syndrome Fazio-Londe disease, progressive bulbar palsy of childhood

?AR

Before 20 (? males earlier) 1–12

—

20–40 (? males earlier) 50% within 18 mo of onset

Distribution and Type Proximal Type I: Acute infantile

Type II: Chronic childhood

Type III: Adult onset

Bulbospinal

Distal Type I: Juvenile onset

Type II: Adult onset Type III: Mild juvenile Type IV: Severe juvenile Type V: Upper limb predominance Type VI: Severe infantile Type VII: With vocal cord paralysis

Type II: Without deafness

AR

Age Unable to Walk (years unless indicated)

Life Expectancy (years unless indicated)

*After Emery, A. E. H.: Review: the nosology of the spinal muscular atrophies. J. Med. Genet. 8:481, 1971. SMA ⫽ spinal muscular atrophy. † AR ⫽ autosomal recessive; AD ⫽ autosomal dominant; XL ⫽ X-linked.

Our understanding of the nosology of these disorders will advance enormously when the relevant gene(s) has been cloned. Nevertheless, a classification based on clinical and genetic features is still needed for prognostic and genetic counseling. Such a classification should preferably fulfill

the criteria outlined by Gardner-Medwin and co-workers.31 There should be minimal overlap in terms of age of onset, course, and severity. The variation in disease characteristics (e.g., distribution of muscle wasting) should be limited, and the disorder should run true within families. To these criteria

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons

Table 68–2. Classification of Spinal Muscular Atrophies 1. Proximal SMA (80%–90%) Infantile and juvenile SMA (SMA I–III) (AD, AR) Adult SMA (SMA IV) (AR, AD, mostly sporadic) 2. Nonproximal SMA 2.1 Distal SMA (AD, AR, sporadic), predominantly involves legs, hands, or both Segmental SMA or benign monomelic amyotrophy (mostly sporadic) 2.2 Scapuloperoneal SMA (AD, AR) 3. Bulbar palsy 3.1 Progressive bulbar palsy of childhood type Fazio-Londe (AR) Bulbar palsy with deafness (Brown-Vialetto–van Laere syndrome) (AR) 3.2 Adult-onset bulbar palsy (AD) 4. Spinobulbar neuronopathy type Kennedy (XL) 5. Variants of SMA 5.1 Variants of infantile SMA Diaphragmatic SMA (AR) SMA plus pontocerebellar hypoplasia (AR) SMA plus arthrogryposis and bone fractures (AR, XL) SMA with myoclonus epilepsy (AR) 5.2 Variants of adult-onset SMA Distal SMA with vocal cord paralysis (AD) (SMA with cardiopathy) AR ⫽ autosomal recessive; AD ⫽ autosomal dominant; XL ⫽ X-linked. From Rudnik-Schöneborn, S., de Visser, M., and Zerres, K.: Spinal muscular atrophies. In Engel, A. G., and Franzini-Armstrong, C. (eds.): Myology: Basic and Clinical, 3rd ed. New York, McGraw-Hill, p. 1846, 2004, with permission.

the author would add that the mode of inheritance within each subgroup must be identical. A clinical syndrome that may be of autosomal recessive or autosomal dominant inheritance must be heterogeneous. The classification proposed by Sabine Rudnik-Schöneborn and colleagues is shown in Table 68–2.

DIAGNOSIS Although the hereditary motor neuronopathies are heterogeneous, certain features are common to all. The presenting symptom is always one of loss of motor function, but the mode of presentation varies in relation to age of onset, from the presence of abnormal fetal movements in utero to deterioration of athletic performance in later life. Delay of motor milestones may be a useful index of muscle weakness in infancy and early childhood. The site of initial weakness depends on the type of HMN and may be predominantly distal, proximal, or bulbar. Weakness is usually accompanied by wasting, although this may not be obvious initially. Fasciculation is not always present. Sensory symptoms and signs are absent, and sphincter function is normal. The tendon reflexes are usually

1605

depressed or absent. The presence of skeletal deformities such as scoliosis, lordosis, pes cavus, and joint contractures is again related to age of onset, the more marked deformities being encountered in severe childhood cases. The most helpful investigations in the diagnosis of HMN are electromyography and nerve conduction studies. It was the advent of these techniques that allowed the existence of several of the subgroups of HMN to be distinguished from primary muscle diseases.47,51,112 Sensory nerve conduction studies are normal. Motor nerve conduction velocity (MNCV) is usually within the normal range, although it may be slightly reduced, particularly in the presence of severe wasting. This is presumably due to loss of large motor neurons.30 Electromyography usually demonstrates large-amplitude motor unit potentials with normal or increased duration and an increased number of polyphasic potentials. Spontaneous fibrillation and positive sharp waves may be seen, but insertion activity is not always present. Namba and colleagues,71 in a review of 382 cases of “chronic proximal spinal muscular atrophy,” noted that electrophysiologic examination established the diagnosis in 97% of patients. Spontaneous discharges occurred in 47%, including fasciculation potentials in 30%, fibrillation potentials in 29%, and positive sharp waves in 7%. Motor unit potentials of increased amplitude and long duration were found in 90%, polyphasic potentials in 35%, and an incomplete interference pattern in 79%. MNCVs were normal in 97% of patients. There is little evidence to suggest that any of the types of HMN demonstrates specific electromyographic findings, except for a feature of cases of the acute infantile type of HMN (Werdnig-Hoffmann disease) that appears to be unique.6,39 This is the presence of spontaneous rhythmic motor unit firing in the relaxed muscles of these infants, persisting for hours and even during sleep. The potentials discharge at a rate of 5 to 18/s. Muscle biopsy had been performed in 180 of the cases reviewed by Namba and co-workers.71 These showed neuropathic changes in 68%, myopathic changes in 5%, and a mixture of both in 19%. Thirteen revealed nonspecific abnormalities or were normal. The finding of myopathic changes in denervated muscle has been described and discussed at length in the past.13,70 They are more prominent in chronic denervating diseases, presumably because of the greater degree of reinnervation that takes place. With longstanding disease there is more fiber breakdown and cellular response in the muscle, which Drachman and associates referred to as the “myopathic” changes of chronic neuropathy.13 Dubowitz and Brooke commented: “To term these changes myopathic, however, is in our opinion really not justified. We prefer to look upon them as the changes seen as part of the picture of chronic neuropathies.”17 Nevertheless, the difficulties in interpretation of mixed biopsy findings have contributed to the nosologic confusion surrounding the hereditary motor neuropathies, as is discussed later.

1606

Diseases of the Peripheral Nervous System

Dubowitz and Brooke suggested that the acute infantile and chronic childhood forms of proximal HMN demonstrate slightly different histologic abnormalities in muscle.17 Reports concerning the other forms of HMN have not been numerous enough to judge whether there are any specific features. The differences may depend purely on chronicity. In HMN of infantile onset, there is both type I and type II fiber atrophy, and considerable variation of fiber size occurs. Hypertrophied fibers (up to four times normal size) may be type I or type II. Nuclear changes are limited to the occurrence of pyknotic nuclear clumps in the atrophic fibers. Degeneration and necrosis are rare. Fibrous tissue tends to be laid down between fascicles rather than around fibers. The number of muscle spindles seen in muscle biopsies from patients with infantile-onset HMN is increased. In juvenileonset HMN, fiber group atrophy is also present, but the clusters of atrophic fibers are smaller. The fibers are less rounded and may be angulated. There is often type II predominance; giant fibers are uncommon. Architectural changes within the fibers, such as central cores and target fibers, are more frequent, as are internal nuclei. The pathology of the motor neuron in HMN has been discussed in Chapters 31 and 32. Other investigations are relatively unhelpful in the diagnosis of HMN. Plasma creatine kinase (CK) level is normal in over 70% of cases with prolonged survival71 and in a higher proportion of acute infantile cases. Increases are generally small, but rarely enzyme levels may be 10 times higher than normal.68,85,108 The electrocardiogram is usually normal, although Dubowitz has described irregularity of the baseline (presumably in relation to fasciculation) in chronic childhood HMN.16 The cerebrospinal fluid is also normal,60 and no consistent biochemical abnormalities have been identified. It is important to distinguish HMN from recognizable disorders that have a prominent component of neurogenic weakness. Primary among these is hexosaminidase A deficiency. This form of GM2 gangliosidosis has a variable phenotype, varying from infantile Tay-Sachs disease to adult-onset multisystem disorders usually consisting of neurogenic muscle weakness, often with evidence of corticospinal tract involvement, supranuclear ophthalmoplegia, ataxia, dysarthria, and dementia. A pure HMN type of picture is extremely rare, although the disease may be confined to this at the time of presentation.75 Atypical features for HMN should lead to assay of hexosaminidase activity.

CLINICAL AND GENETIC FEATURES Proximal Hereditary Motor Neuronopathy (Table 68–3) The nosology of this group of disorders is particularly controversial, the main issue being how much heterogeneity there is within the chronic childhood form of HMN.

The separate identity of the acute infantile form (WerdnigHoffmann disease) is reasonably secure, and a summary of the various arguments for and against subdivision of the chronic childhood type is presented here. These problems should be resolved in the relatively near future following localization of the locus for both acute infantile and chronic childhood forms of HMN to the long arm of chromosome 5.5,33,64,65 This was initially reported in chronic childhood families, with a 2-point logarithm of odds (LOD) score of 8.3 at a recombination fraction of 0.02 (2 cM) between the disease locus and a marker called D5S6, by Brzustowicz and colleagues5; a LOD score of 9.06 at a recombination fraction of 0.04 with D5S39 was reported from France.64,65 At that stage there was a suggestion that the acute HMN locus was in the same region. Subsequently, Gilliam and co-workers provided further evidence to support this by studying four inbred families with this disorder.33 Testing for heterogeneity suggested that two kindreds, one chronic and one acute, did not show evidence of linkage to a locus on chromosome 5q. When the acute kindred was removed from the analysis of similar families, a peak LOD score of 3.2 between markers D5S112 and D5S39 was obtained, very close to the location of the LOD score peak for chronic childhood HMN families. In 25 acute HMN families from Europe, a LOD score of 2.36 at a recombination fraction of 0 was reported with D5S39.65 Type I: Acute Infantile Form (Werdnig-Hoffmann Disease) Pearn has estimated that this disorder is the second or third most common fatal recessively inherited disease of childhood.77 The disease frequency is on the order of 1 in 25,000 births in northeast England, giving a gene frequency of 1 in 160 and a heterozygote frequency of 1 in 80. Pearn and co-workers defined the disease as follows: (1) it is fatal before the third birthday; (2) it causes delayed motor milestones before the age of 6 months, and usually before 3 months; and (3) it causes the parents to seek medical attention before the child is 6 months old.82 Pearn observed that 30% of 39 mothers noted absent fetal movements in utero.76 One of the latter felt fasciculation in the last few weeks of pregnancy. About one tenth of affected infants are born with minor orthopedic deformities.87 One third of the 76 cases described by Pearn and Wilson87 were thus judged to have a prenatal onset. The presenting features in the remainder were hypotonia and weakness (in 37 cases), poor feeding (in 24), and failure to progress to normal motor milestones (in 14). The neonate is typically inactive and hypotonic (Fig. 68–1). Proximal weakness is usually profound, but movements of the hands and feet are active. Breathing is almost entirely diaphragmatic, because the intercostal muscles are severely affected. The child is never able to roll over, and head

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons

1607

Table 68–3. Classification of Proximal Spinal Muscular Atrophy Modified According to the International SMA Consortium SMA Type

Principal Synonyms

Definition

Genetics

I

Werdnig-Hoffmann disease Acute infantile SMA

Autosomal recessive (~95% SMN1 deletions)

II

Chronic childhood SMA Arrested Werdnig-Hoffmann disease

III

Kugelberg-Welander disease Juvenile SMA

IV

Adult SMA

Sitting not achieved Onset usually within the first 6 months Death usually in the first 2 years of life Unaided sitting possible, walking not achieved Onset usually in the first year of life Survival ⬎90% by 10 years Walking without aids achieved IIIa: Onset ⱕ3 years IIIb: Onset ⬎3 years Mild course, life span not markedly reduced Onset ⬎30 years Variable severity, normal life span

Autosomal recessive (~95% SMN1 deletions)

Autosomal recessive (80%–90% SMN1 deletions), rarely autosomal dominant mutations Excess of males Mostly sporadic Autosomal dominant (gene defect unknown) Autosomal recessive (extremely rare)

Data from International SMA Collaboration: Workshop report. Neuromuscul. Disord. 9:272, 1999; and Zerres, K., Rudnik-Schöneborn, S.: Natural history in proximal spinal muscular atrophy (SMA): clinical analysis of 445 patients and suggestions for a modification of existing classifications. Arch. Neurol. 52:518, 1995. From Rudnik-Schöneborn, S., de Visser, M., and Zerres, K.: Spinal muscular atrophies. In Engel, A. G., and Franzini-Armstrong, C. (eds.): Myology: Basic and Clinical, 3rd ed. New York, McGraw-Hill, p. 1847, 2004, with permission.

control is very poor. The face is usually spared, but bulbar involvement leads to feeding difficulties. Fasciculation may, often with difficulty, be observed in the tongue, and tremor can sometimes be seen in the fingers. Affected infants tend to lie in a froglike position, with flexion and external rotation of the hips and flexion and internal rotation of the elbows. The tendon reflexes are absent. This clinical picture is typical, but occasionally the neonate may appear quite normal, even to an experienced clinician, and develop abnormal physical signs some weeks after birth.16 The disease is relentlessly progressive, and only 1 of the 76 patients described by Pearn and Wilson was ever able to sit unsupported.87 Twenty were able to lift the head (even momentarily) off the bed during life. Death usually occurs from intercurrent respiratory infection, and Dubowitz observed that the use of antibiotics did not have any appreciable effect on life expectancy.14 Pearn and Wilson found that 50% of affected children were dead by the age of 7 months and 95% by 17 months.87 The differential diagnosis of acute infantile proximal HMN includes many of the causes of hypotonia in infancy. A detailed discussion of these is outside the scope of this chapter; the subject has been reviewed by Dubowitz.15 Distinction of this disorder from type II proximal HMN is very important, especially from the point of view of genetic counseling. The recurrence risk to subsequent siblings of index patients with type I proximal HMN is 1 in 4; this is

high, but some parents may be willing to accept it if recurrence implies death in the first year or so of life. The prospect of a seriously handicapped child with prolonged survival may be more daunting. Pearn and Hudgson85 gave guidelines for counseling based on the clinical features of patients with types I and II proximal HMN. It can be concluded that the siblings of a child with the disease whose onset was before the age of 6 months and who died before 18 months of age are unlikely (⬍5%) to have a chronic course if they develop the disease. Assuming that evidence is found for locus homogeneity in the majority of acute infantile proximal HMN families, prenatal diagnosis using linked genetic markers should be available soon. A further type of infantile HMN has been described, in which inheritance appears to be X-linked.35 Four male infants from three sibships in a single family presented with hypotonia, areflexia, and congenital joint contracture. Three had fractures at birth. Decreased fetal movement was noted in two pregnancies. Death occurred within the first 2 years of life in all but one child, who was ventilator dependent. Type II: Chronic Childhood Form The genetic identity of proximal HMN type I is reasonably well established, but there exist a large number of patients in whom proximal HMN develops in infancy or early childhood and follows a more benign course than that of the acute infantile form. In the mid-1950s Wohlfart and co-workers112 and Kugelberg and Welander51 identified a

1608

Diseases of the Peripheral Nervous System

FIGURE 68–1 Infant with acute infantile proximal hereditary motor neuropathy (Werdnig-Hoffman disease). The child is alert but has a hypotonic posture. There is proximal muscle wasting, particularly in the upper limbs. Recession of the intercostal muscle gives the chest a bell-shaped appearance. (Photograph supplied by the Medical Illustration Department, Hospital for Sick Children, Great Ormond Street, London.)

chronic form of proximal spinal muscular atrophy with onset between the ages of 2 and 17 years. Such patients had previously been thought to suffer from muscular dystrophy, and it seemed likely that they had a different disorder from patients with type I proximal HMN. However, subsequent authors31,70,71,88 showed that the majority of patients with chronic proximal HMN develop symptoms before the age of 2 years, and often in the first year of life. This disorder is sometimes called the “arrested form of Werdnig-Hoffmann disease,” and it has been suggested that this is a distinct disease subgroup, intermediate between cases of type I proximal HMN and Kugelberg-Welander syndrome.19 Lugaresi and co-workers considered that the intermediate form could be divided into three groups: early onset but more chronic than Werdnig-Hoffmann disease; slow evolution with later onset; and early onset with total arrest.55 It is

likely, in retrospect, that one of the patients reported by Werdnig had a more protracted course than is now considered acceptable for examples of acute infantile HMN.111 In the last two decades a number of studies of the chronic childhood form of proximal HMN have suggested that there is no firm evidence in favor of an intermediate group of patients,7,40,87 whereas other authors believe that distinct subgroups can be defined.42,114 One of the main points against genetic heterogeneity is the considerable discordance that exists between affected members of the same sibship in terms of age of onset, progression, and age of death. Although the disease may run true within families in a proportion of cases,71,114 Gardner-Medwin and associates reported two brothers with onset of symptoms at the ages of 18 months and 15 years,31 and another sibship contained one case with onset at birth and the other at 4 years. Emery and colleagues described sibships with ages of onset at 1, 13, and 15 years, at birth in two cases and at 14 and 11 years in their siblings.21 One patient was still ambulant at the age of 41 years, whereas the siblings of this patient were unable to walk at ages 6 and 11 years. In another family, two patients died at 22 and 32 months of age, and an affected sibling was still alive at the age of 8 years. Similar discordance within sibships was reported by Pearn and Wilson.88 Bundey and Lovelace7 and Pearn and co-authors81 examined intrafamilial correlation for various quantifiable disease parameters in chronic childhood HMN and found no evidence of genetic heterogeneity. Pearn and colleagues suggested that a single autosomal recessive gene accounts for over 90 per cent of cases [of chronic childhood spinal muscular atrophy], causes a clinical syndrome which manifests its first clinical signs before 5 years of age and in almost all cases before 2 years of age, but which is compatible with life into the third decade. . . . A small group of cases (a) is caused by new dominant mutation(s), or (b) is composed of phenocopies, or both. This relatively uncommon form may comprise the majority of late-presenting cases, and may account for all cases which manifest signs after 5 years of age. The spectrum of age-at-onset of this group cannot be determined at present, but the disease may be manifest before the age of 2 years; it is clinically indistinguishable from spinal muscular atrophy caused by an autosomal recessive gene.81 The essence of these conclusions, and those of Bundey and Lovelace,7 is that the arrested forms of WerdnigHoffmann disease and Kugelberg-Welander syndrome are the two ends of a single continuous disease spectrum. Some cases of childhood or juvenile onset remain difficult to classify.100,103 In contrast, Hausmanowa-Petrusewicz and colleagues,42 analyzing data from 354 cases of chronic childhood proximal HMN, proposed subdivision into three categories. The

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons

infantile chronic form was characterized by age at onset, usually between 3 and 9 months, with a milder course in females. Inheritance was autosomal recessive. The childhood isolated form had onset between 10 and 36 months, and most cases were singletons. Segregation ratios in this type were low and incompatible with autosomal recessive transmission. It was thought probable that this subgroup contained nongenetic phenocopies or cases representing fresh dominant mutation, or both. The third, mild childhood and adolescent form, had an age of onset usually after the third year of life. Inheritance was thought to be largely autosomal recessive with a marked sex influence, expressed by a smaller proportion of affected females, which was proposed to be due to reduced penetrance. Fairly similar disease categories were proposed by Zerres,116 who also identified a “childhood isolated form,” although with onset usually in the first year of life; in this group, segregation analysis was incompatible with autosomal recessive inheritance. One feature, pointed out by Hausmanowa-Petrusewicz and co-workers,41 of the patients described by Pearn and collaborators81 is that the proportion of cases studied with the relatively benign, later onset clinical picture (Kugelberg-Welander type) was extremely small. Only three patients developed symptoms over the age of 4 years, and the latest age of onset was 8 years. This is in contrast to the series of 247 patients reported by HausmanowaPetrusewicz and associates, in which 13.2% had onset after the fifth year of life and 10.2% after the eighth year.43 Patients with onset between the ages of 5 and 15 years (including sibling pairs) have also been described by Kugelberg and Welander,51 Meadows and co-workers,62 and others. Namba and co-authors, in a review of 382 cases of “proximal spinal muscular atrophy with a protracted course,” found that 48.8% developed symptoms between the ages of 3 and 18 years.71 The lack of later onset cases in the series of Pearn and colleagues81 is almost certainly due to a degree of ascertainment bias; 63% were seen at a specialist children’s hospital. Nevertheless, the proportion of later onset patients in the study of Bundey and Lovelace, in which there was no such bias in selection, was also small (4 cases in 50 over 5 years).7 Even taking later onset childhood cases into account (in an international collaborative study that included data from both Pearn and Hausmanowa-Petrusewicz), Emery and associates again failed to find evidence for genetic heterogeneity in chronic childhood proximal HMN.21 These authors observed that parental consanguinity was less frequent among singleton than familial cases, which would support the suggestion of Bundey and Lovelace7 and Pearn and co-workers81 that a proportion of single cases are not recessively inherited. The problems of classifying this group of patients should be resolved when the disease mutations are identified. As has been mentioned, there is good evidence that the acute and chronic childhood types of proximal HMN are caused by a gene or, more likely, genes at the same locus. The wide

1609

spectrum of age at onset and disease progression in childhood of proximal HMN could thus be explicable on the basis of patients representing a mixture of homozygotes for different mutations and compound heterozygotes. The occurrence of both homozygosity and compound heterozygosity in the same family could explain the well-documented intrafamilial variation, also observed in family B in the study of Gilliam and colleagues.33 It is more difficult to provide a hypothesis for possible influence of sex in disease severity, because the locus for autosomal dominant chronic childhood-onset proximal HMN has not yet been identified. Pearn found that the prevalence of chronic childhood HMN was 1.2 in 100,000 (incidence 1 in 24,100 live births) in northeast England.79 The heterozygote frequency was about 1 in 90, and the median age of onset of 12.71 months, with a range of 1 to 96 months.83 Ninety-five percent of children were abnormal by the age of 3 years. Fetal movements are nearly always normal, although exceptions occur.70 Fourteen of the 50 patients described by Bundey and Lovelace were never able to walk; in another 9, walking was delayed.7 The rest walked at the normal time, but 10 never walked normally. A small proportion of patients (about 5%) are never able to sit unsupported.83,88 In the latter report, 19% of patients were still ambulant at the age of 10 years, and 5% at 20 years.83 Thirty-three of the 50 patients reported by Bundey and Lovelace were still walking at the age of 10 years,7 and 2 affected siblings described by Meadows and co-workers could walk unaided when ages 48 and 66 years.63 These differences must again be due to different sources of ascertainment. The disease is only very slowly progressive, and stepwise deterioration may occur.88 There may be long periods of apparent arrest.14 One important feature is that the age of onset has very little bearing on the eventual disability. Pearn and colleagues noted a tendency for the disease to be less severe in females than in males,83 whereas Hausmanowa-Petrusewicz and associates found that it was apparently milder in males and that there were more males among later onset cases.43 Muscle weakness and wasting are more marked proximally in the majority of patients, although the more severely affected ones demonstrate generalized paresis.88 In such patients there is a high incidence of lordosis, scoliosis, and joint contractures (Fig. 68–2). Skeletal deformity is much less prominent in patients who remain ambulant throughout childhood. About half have fasciculation in the limbs. Some authors have found a significant incidence of calf hypertrophy in childhood HMN.40 Pearn and Hudgson suggested that there may be a separate form of HMN associated with calf hypertrophy and onset in adolescence,84 but some of these patients were later shown to have a deletion of exons from the dystrophin gene, indicating a diagnosis of Becker’s muscular dystrophy. The same applies to some other patients considered to have Kugelberg-Welander syndrome.9 It is important to consider the diagnosis of an atypical X-linked muscular dystrophy in patients with a limb girdle syndrome,

1610

Diseases of the Peripheral Nervous System

FIGURE 68–2 Seven-year-old boy with chronic childhood (type II) proximal hereditary motor neuropathy. He has generalized muscle wasting and weakness and a severe scoliosis.

particularly males with calf hypertrophy, even if electromyography suggests neurogenic atrophy. Cranial nerve involvement appears to be very variable, even within members of the same sibship. Facial weakness and wasting of the tongue occur in roughly half of patients; dysarthria is present in some, but dysphagia is relatively rare. The intercostal muscles may be severely weak and respiration entirely diaphragmatic, particularly in severe cases.88 Emery and co-authors found that 9% of patients were considered to be mentally retarded20; Spiro and colleagues described a sibship in which three boys had chronic childhood HMN, mental retardation, and microcephaly.101 It may be that this family represents a distinct, unique disorder. Differentiation of early-onset cases of type II HMN from type I may be extremely difficult at the time of diagnosis and can be made only with hindsight after prolonged survival (after the age of 2 years). If elder siblings clearly suffer from the chronic (type II) form of HMN, this problem does not arise. However, even if a 3-year-old child dies as a result of type II HMN, subsequent affected offspring may survive into the third decade, albeit severely handicapped.88 This point is obviously crucial to the parents in genetic counseling. In patients with later onset, the differential diagnosis is chiefly from the congenital myopathies and limb girdle muscular dystrophy. Fasciculation, either in the tongue or in the limbs, is an obvious feature in favor

of HMN, but the question is more satisfactorily resolved by the use of electromyography and, if necessary, muscle biopsy. As has been mentioned, it is impossible to distinguish patients with new dominant mutations from the majority of patients who have autosomal recessive proximal HMN type II. Although it may be true that most of the former develop symptoms beyond the age of 2 years,81 dominant pedigrees have been described in which affected family members were noted to be abnormal in the first year of life.115 This overlap leads to difficulties in genetic counseling, especially if only one member of a family is affected. If two siblings are affected and the parents are normal, the recurrence risk to subsequent siblings is clearly 25% and that to their children is negligible. Pearn and co-workers suggested that the recurrence risk to siblings of a singleton case with onset before the age of 3 years was 1 in 581; if onset was over the age of 3 years, the risk drops to 1 in 10. Bundey and Lovelace found that one in eight of the offspring of index cases with onset over the age of 2 years was affected with type II HMN.7 Hausmanowa-Petrusewicz and colleagues gave slightly different risk estimates for the sibs of singleton cases, as follows: age at onset 0 to 9 months—1 in 4–5; 10 to 36 months—1 in 7; 37 months to 8 years—1 in 4 to males, 1 in 6 to females; and 9 to 18 years—1 in 5 to males, 1 in 20 to females.42 Prenatal diagnosis in pregnancies following the birth of two affected sibs should soon be possible if the affected children can be shown to share the same polymorphisms at the chromosome 5 locus. However, in a singleton case, it would not be safe to assume that the child has an autosomal recessive disorder caused by a mutation mapping to this locus, because the disease could be due to a recessive gene with a locus elsewhere or to a fresh dominant mutation. These difficulties should be clarified when the disease mutations can be detected directly. Type III: Adult-Onset Proximal HMN of Autosomal Recessive Inheritance There is good evidence to suggest that proximal HMN developing beyond the age of 15 years is genetically distinct from the chronic childhood type (type II).86 There have been several reports of the disorder occurring in more than one member of a single sibship and not in the parents7,86,106 as well as a number of single cases.62,69,86 Segregation analysis indicates that the majority of the latter are of autosomal recessive inheritance,86 but it is likely that a small proportion represent fresh mutation with the autosomal dominant form of HMN that develops in adult life (see later). Nevertheless, the latter are clinically indistinguishable from the recessive cases, so it is likely that the following description does not refer to a single genetic entity. Surprisingly for a rare recessive disorder, consanguinity among the parents of patients has not been reported.

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons

The age of onset of first symptoms is between 15 and 60 years; in the series of Pearn and co-workers the median age of onset was 35 years.86 These authors estimated a disease prevalence in northeast England of 0.32 per 100,000. Muscle weakness is slowly progressive; the majority of patients do not need to use aids for walking until the sixth decade, although Bundey and Lovelace described two brothers who were wheelchair-bound in the fifth decade.7 Life expectancy is not significantly reduced. The distribution of muscle involvement is mainly proximal, but this usually extends to involve the distal extremities. Bulbar involvement is infrequent. It has mainly been observed either in single cases in males62,90 or in affected brothers.106 It is quite possible that these patients represent examples of the X-linked form of HMN described by Kennedy and associates48 and others; bulbar involvement is particularly prominent in this latter disorder (see later). The tendon reflexes are depressed or absent. Skeletal deformity, apart from lordosis, is rare. It is likely that singleton cases of adult-onset proximal HMN have been relatively underreported in the literature for reasons of both interest and difficulties in diagnosis. In the early stages the disorder can be impossible to distinguish from motor neuron disease, especially because the distribution of weakness in this type of HMN may be asymmetrical.86 One important point is that there have been no reports of definite familial cases in which pyramidal signs were conspicuous, although some authors have suggested that their presence does not exclude HMN.62 Genetic counseling of singleton cases is complex. Pearn and co-workers suggested an empirical recurrence risk of 1 in 5 to siblings of single cases, and of 1 in 20 to their children.86 This latter figure was based on an (unsubstantiated) estimate of 10% of singletons having new dominant mutations. Clearly, in the presence of affected siblings and normal parents who survive past the age of 60 years without developing symptoms, the risk to offspring of patients is negligible. Types IV and V: Autosomal Dominant Proximal HMN of Juvenile and Adult Onset Proximal HMN of autosomal dominant inheritance is relatively rare. Pearn found that it constituted 2% of cases of childhood HMN and about 30% of adult-onset cases.78 There is a considerable amount of evidence to suggest that there are two forms of the disorder, one with an onset in infancy or early childhood, and another in which symptoms usually develop in the third and fourth decades. This division is supported by complete concordance for age of onset within families. In the juvenile form of the disease (type IV), affected individuals may be noted to be hypotonic in infancy; there is often delay in motor milestones. In over 90% of the reported cases,3,19,55,56,78,109,114,115 onset has been before the age of 5 years and only once beyond the age of 10 years.109

1611

The disorder is benign and only very slowly progressive. The patients reported by Magee and de Jong denied any worsening of symptoms after adolescence.56 Four of the six patients described by Pearn were never able to run,78 but all were able to walk unaided and three were over 30 years old at the time of study. Zellweger and co-workers noted considerable variation in severity in one kindred (as is to be expected in an autosomal dominant disorder)115; one patient was still ambulant and able to do housework at the age of 76 years. Pearn suggested that muscle weakness in the dominant form of juvenile proximal HMN was rather more generalized than in the autosomal recessive type.78 This appears to be true, but there has been a definite preponderance of proximal muscle involvement among the reported cases. Bulbar involvement is relatively rare and only mild. The affected individuals in the family described by Tsukagoshi and associates experienced mild dysphagia for many years.108 The tendon reflexes are depressed or absent. Mild joint contractures and lumbar lordosis are frequent. The group of disorders most likely to be confused with dominant juvenile-onset proximal HMN are the congenital myopathies. Clinically this distinction may be difficult, because fasciculations are not seen particularly frequently in the former, and muscle wasting may be severe in both groups. Electromyography and muscle biopsy should distinguish between the two. Genetic counseling is straightforward in the presence of a clear family history. The disorder appears to be fully penetrant.78 The problems that arise in counseling patients with this disorder who may have new mutations have already been discussed. The adult-onset form of autosomal dominant proximal HMN (type V) is less common than the juvenile type. Examples of this disorder have been reported by several authors.7,62,78,90,106 The median age of onset in the three families studied by Pearn was 37 years78; it can range from 25 to 65 years. Proximal muscle weakness and wasting are more marked than in the juvenile cases and the course is more rapidly progressive. Bulbar involvement is also more frequent. None of the seven patients reported by Pearn could run within 5 years of development of symptoms,78 and the author considered that median life expectancy after diagnosis was about 20 years. If there is a clear history of dominant inheritance and evidence of a denervating disease, the only disorder likely to be confused with adult-onset proximal HMN is familial motor neuron disease. The latter is usually asymmetrical and more rapidly progressive, and signs of pyramidal tract dysfunction usually appear eventually even if they are not found at the time of presentation. Also, penetrance in proximal HMN is complete, whereas in families with familial motor neuron disease it is usually incomplete. As with the juvenile form of proximal HMN, it is likely that some single cases with onset in adulthood represent new dominant mutations. The proportion is likely to be smaller

1612

Diseases of the Peripheral Nervous System

in the adult disease because biologic fitness will be less impaired. The counseling of such patients has already been discussed.

X-Linked Bulbospinal Hereditary Neuronopathy This syndrome, which is both clinically and genetically distinct from the proximal forms of HMN, was first described by Kennedy and co-workers in 1968.48 Other families have since been reported.38,93,98,107 Onset is usually in the third or fourth decade but may be as early as 15 years and as late as 59 years. One prominent feature is the occurrence of muscle cramps, which may precede the onset of weakness by a number of years. Proximal weakness usually develops first in the lower limbs and then spreads to the shoulder girdle, face, and bulbar musculature. Fasciculation is often prominent, particularly in the lower face, where it is more pronounced after muscular contraction. It is also present in the tongue. Tremor of the outstretched hands is common. The tendon reflexes are depressed or absent. Dysphagia and dysarthria usually develop some 10 or 20 years after the onset of symptoms, at which stage weakness in the limbs has usually spread to involve the distal musculature. Aspiration pneumonia is a common complication in the terminal stages. Life expectancy is, nevertheless, not greatly reduced. One interesting finding in the cases of bulbospinal hereditary neuronopathy (HN) reported to date is a high incidence of gynecomastia among affected males (Fig. 68–3), which has been observed in over 50% of patients. Secondary sexual characteristics are normal, although the testes may be small and infertility is common. In one of the families reported by Kennedy and co-workers, three of the nine affected individuals had maturity-onset diabetes.48 There were also numerous neurologically normal relatives with diabetes mellitus in this particular kindred, but two brothers with bulbospinal neuronopathy described by Harding and associates also had maturity-onset diabetes.38 Electromyography in this form of HN usually provides unequivocal evidence of denervation. MNCV is normal (or very slightly slowed). Sensory nerve action potentials are usually reduced in amplitude or absent, and this is not found exclusively in diabetic patients. Clinical sensory loss is unusual. Studies of sural nerve biopsies from six patients suggested that a distally accentuated axonopathy was the salient pathologic process.99 Prominent “secondary myopathic” changes as well as features of denervation were seen in muscle biopsy specimens from two of the patients studied by Harding and colleagues.38 These findings may be correlated with the presence of high plasma CK levels in this disorder, often up to five times the upper limit of normal.38 In three autopsied cases, motor neurons were markedly depleted through all spinal segments and in brainstem motor

FIGURE 68–3 This 59-year-old man with X-linked recessive bulbospinal neuropathy developed proximal muscle weakness when he was 35 years old. He is now wheelchair-bound. There is proximal wasting in the upper limbs and gynecomastia.

nuclei except for the third, fourth, and sixth cranial nerves. Neuronal cell bodies in the dorsal root ganglia and fibers in dorsal spinal roots were fairly well preserved in population and morphology, but myelinated axons in the fasciculus gracilis and sciatic, tibial, and sural nerves showed axonal loss with distal emphasis.99 These observations, together with those in sural nerves mentioned earlier, indicate that a lower motor and primary sensory neuronopathy is the major neurologic manifestation of this disease. The differential diagnosis of this condition is unlikely to be difficult if there is a clear history of X-linked inheritance. In single cases, the late onset often leads to confusion with motor neuron disease. It is highly probable that this particular form of HN is commonly not recognized as such in the absence of a family history. Yet, by the laws of chance, there is likely to be a considerable proportion of singleton males in any X-linked recessive disorder, without the need for invoking the occurrence of new mutation. Probable examples of such single cases have been described86,98; Harding and co-workers reported six males with the disease, none of whom had similarly affected relatives.38 Recognition of this disorder in males without affected relatives is important, because their daughters will be obligate carriers of the disease and will therefore have a 50% chance of having an affected male child. The disease locus was mapped to the proximal long arm of the X chromosome26; more recently this disorder was shown to be associated with an insertion in the androgen receptor gene.52 This potentially allows carrier detection and prenatal diagnosis.

1613

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons

Distal Hereditary Motor Neuronopathy The distal form of HMN is one subgroup of the disorders giving rise to the peroneal muscular atrophy syndrome; it was called progressive muscular atrophy or the spinal form of Charcot-Marie-Tooth disease.18 There were 34 patients with distal HMN in the series of 262 patients with peroneal muscular atrophy described by Harding and Thomas37; Pearn and Hudgson found that distal HMN accounted for 10% of all cases of spinal muscular atrophy in a population survey in northeast England.85 Distal HMN is genetically heterogeneous; there are at least two autosomal dominant forms of the disease and probably three autosomal recessive types (Table 68–4). No X-linked pedigrees have been identified.

A severe infantile form of distal HMN has been described, and this appears to be autosomal recessive.4 Diaphragmatic paralysis was a prominent early feature, with respiratory failure necessitating artificial ventilation. Although originally distal, muscle weakness spread to involve proximal muscles and death occurred before the age of 1 year. The patients with dominant inheritance described by Davis and co-workers12 and Harding and Thomas37 all developed symptoms before the age of 20 years and usually in the first decade. Nelson and Amick72 and McLeod and Prineas59 reported dominant pedigrees in which affected members were asymptomatic until the third and fourth decades. In the series of Harding and Thomas, patients with autosomal recessive inheritance were only mildly affected,37 whereas Pearn and Hudgson have described a much more severe form of recessive distal HMN.85 It is also possible

Table 68–4. Spinal Muscular Atrophies with Particular Distribution Name

Principal Synonyms

Clinical Features

Inheritance

Gene Defect

Distal SMA

Peroneal muscular atrophy Distal hereditary motor neuronopathy (HMN)

Distal weakness (lower or upper limb predominance) with normal conduction velocities and normal sensory findings Sometimes pyramidal features Onset birth–adulthood

AR (~75%) AD (~25%)

Segmental SMA

Monomelic juvenile SMA type Hirayama, benign monomelic amyotrophy

Asymmetrical involvement confined to upper (lower) extremities Onset 2nd decade, benign course (arrest usually after 2–4 years) Male predominance Myoclonic epilepsy, ataxic movements, hypotonia, and generalized weakness; normal intelligence Onset mostly in childhood Progressive distal wasting, hoarseness of voice, no bulbar palsy Onset in adulthood Variable weakness of foot and toe extensors, shoulder girdle, and upper arms, rarely with cardiopathy; onset in youth or adulthood Proximal/distal weakness mostly in adulthood, gynecomastia, bulbar palsy, sensory disturbances

Mostly nongenetic

11q13 (a.r., also SMARD1) 12q24 (a.d. HMN type II) 7p14 (a.d. HMN type VII) 9q34 (a.d. HMN type VB) Unknown

Distal SMA with myoclonic epilepsy

Distal SMA with vocal cord paralysis

Hereditary motor neuronopathy type VII

Scapuloperoneal muscular atrophy

Scapuloperoneal atrophy type Stark-Kaeser (dominant form)

Spinobulbar neuronopathy

Kennedy disease

Few AD cases

AR AD

Unknown

AD

2q14 (a.d.)

AD AR (X-linked)

12q23-24 (a.d.)

X-linked

Xq13.1 (CAG repeat expansion of the androgen receptor gene)

AD ⫽ autosomal dominant; AR ⫽ autosomal recessive; SMA ⫽ spinal muscular atrophy. From Rudnik-Schöneborn, S., de Visser, M., and Zerres, K.: Spinal muscular atrophies. In Engel, A. G., Franzini-Armstrong, C. (eds.): Myology: Basic and Clinical, 3rd ed. New York, McGraw-Hill, p. 1855, 2004, with permission.

1614

Diseases of the Peripheral Nervous System

that cases of distal HMN in which wasting and weakness are virtually confined to the hands53 represent a distinct genetic entity. Intrafamilial concordance for age of onset, severity, and distribution of muscle involvement is high. Young and Harper reported a large kindred in which distal HMN combined with vocal cord paralysis.113 Inheritance was autosomal dominant. Onset was most commonly in the teens, with weakness and wasting of the small hand muscles, subsequently involving the distal muscles of the legs. Hoarseness was usually noted from the second decade of life; laryngoscopy showed unilateral vocal cord paralysis in some patients, but bilateral cord paralysis had caused airway obstruction in two cases. The disease usually begins with distal wasting and weakness of the legs, particularly of the anterior tibial and peroneal compartments (Fig. 68–4). Foot deformity of pes cavus type is prominent in early-onset cases, and it has been suggested that this is more severe in distal HMN

FIGURE 68–4 Six-year-old boy with distal hereditary motor neuropathy. There is wasting of the distal muscles in the legs. (From Harding, A. E., and Thomas, P. K.: Hereditary distal spinal muscular atrophy: a report of 34 cases and a review of the literature. J. Neurol. Sci. 45:337, 1980, with permission.)

than in either type I or type II hereditary motor and sensory neuronopathy (HMSN).37 Scoliosis is present in about one quarter of patients. In the more severe autosomal recessive form of the disease, wasting and weakness may extend to the proximal musculature in the legs,85 but this is unusual in the other types. Weakness of the upper limbs is also prominent in the severe cases of recessive inheritance, but it was observed in only one quarter of the milder cases, both recessive and dominant, reported by Harding and Thomas.37 Reflex loss in distal HMN is less frequent than in HMSN: The ankle jerks are absent in about one third of cases, but in over 60% the knee jerks and upper limb reflexes are normal. Although involvement of the legs is usually more prominent than that of the upper limbs in distal HMN, a few of the patients have had more severe muscle wasting and weakness in the arms.37,60 This distribution appears to run true within families, and inheritance appears to be autosomal dominant. It is possible that such families represent an intermediate group between the more common form of distal HMN predominantly involving the legs and those reported by Lander and co-workers53 and O’Sullivan and McLeod,73 in whom muscle wasting and weakness were confined to the hands and forearms. Lander and associates described a family in which wasting and weakness were virtually confined to the hands53; inheritance was dominant. Onset was in the first or second decade. In several family members, however, there was electromyographic evidence of denervation in the small foot muscles, and in two there was a suspicion of clinical weakness. O’Sullivan and McLeod subsequently reported six sporadic cases, although in four the disorder was unilateral.73 The prognosis for this group of disorders, apart from the infantile type and the severe recessive form reported by Pearn and Hudgson,85 is good. Only one of the patients reported by Harding and Thomas required a stick in order to walk, and none was wheelchair-bound.37 The disorder is very slowly progressive and in some instances it may appear to arrest. In the series of 12 patients reported by Pearn and Hudgson, only one quarter could walk unaided in adult life.85 None was able to run; three patients had never been able to run. The distinction between those with autosomal recessive cases having a relatively poor prognosis and those with benign disease would appear to be impossible at the onset of symptoms; accurate advice can be given only after a period of observation. Electrophysiologic investigations of patients with distal HMN have been well documented. MNCV is normal, although the small foot muscles may be totally denervated. Sensory conduction is similarly normal. Electromyography shows the characteristic features of denervation in the distal musculature. Plasma CK may be raised; in some of the patients described by Pearn and Hudgson it was moderately

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons

high,85 but this is unusual. Muscle biopsy confirms the presence of denervation. McLeod and Prineas undertook morphometric and ultrastructural studies on sural nerve biopsies from four patients, with entirely normal findings.59 The differential diagnosis of distal HMN is chiefly from types I and II HMSN. The distinction between the three disorders can be impossible to make clinically, because a proportion of patients with HMSN, particularly type II (60%), have no clinical sensory loss. In type I HMSN, MNCV in the upper limbs is less than 38 m/s, and sensory conduction is abnormal. In type II HMSN, although MNCV is, on average, slightly reduced, this may not be diagnostically helpful in individual cases, and the essential distinguishing feature between the two conditions is the presence or absence of normal sensory action potentials.37 The only other form of spinal muscular atrophy likely to be confused with distal HMN is that of scapuloperoneal distribution, in which there may be wasting of the small hand muscles; nevertheless, shoulder girdle involvement is usually prominent. When the abnormal physical signs are confined to the lower limbs in distal HMN, the possibility of spina bifida occulta with cauda equina involvement should be excluded. Genetic counseling of patients with distal HMN is straightforward if there is clear evidence of autosomal dominant inheritance. If only members of a sibship are affected, it is important to examine the parents to exclude dominant inheritance. If only members of a sibship are affected, it is important to examine the parents to exclude dominant inheritance, because some affected individuals are asymptomatic. Singleton cases are more difficult to counsel. If the disorder is severe, as in the patients described by Pearn and Hudgson,85 the risk of recurrence to siblings may be as high as 1 in 4. The data presented by Harding and Thomas concerning segregation analysis of milder cases37 suggest that the risk of recurrence to siblings of singleton cases is about 1 in 16, and the risk to their offspring approximately 1 in 4.

1615

There was marked wasting of the small hand muscles as well as those of the shoulder girdle. Before the use of investigations such as nerve conduction studies, electromyography, and nerve and muscle biopsy, the underlying pathogenesis of the scapuloperoneal syndrome was not clear. It has since been established that the syndrome is a heterogeneous entity and may be a result of myopathy,96 neuropathy (Davidenkow syndrome),95 and motor neuronopathy or spinal muscular atrophy (Fig. 68–5). The first description of patients reputed to have the latter was that of Kaeser.47,91 He reported a family in which the disorder affected 12 individuals in five generations. Onset was between the ages of 30 and 50 years and the course was relatively benign, lasting up to 30 years. Muscle weakness started in the lower limbs and affected chiefly the peroneal, anterior tibial, and calf muscles. The intrinsic muscles of the foot were relatively spared. Weakness later developed in the thighs and pelvic girdle. It should be pointed out that only one of the three patients examined by Kaeser had the classic scapuloperoneal syndrome.47 Paresis spread to the shoulder girdle musculature 5 years after onset in the legs and was moderately severe. There was also weakness of the bulbar, facial, and extraocular muscles. One of the other patients

Scapuloperoneal Hereditary Motor Neuronopathy Types I and II Between 1927 and 1939, Davidenkow described 13 patients whom he considered to have ‘“scapuloperoneal amyotrophy.”10,11 The disorder was usually of autosomal dominant inheritance, and muscle weakness developed between the ages of 14 and 26 years. Symptoms began either distally in the legs or in the shoulder girdle. The disease tended to run a relatively benign, slowly progressive course. Some of the patients were noted to have fasciculation, and a considerable proportion had distal sensory loss. An autosomal dominant scapuloperoneal syndrome of probable neurogenic origin was reported by Palmer.74 The age of onset was in the third decade. Fasciculation was prominent in affected individuals.

FIGURE 68–5 Scapuloperoneal hereditary motor neuropathy. There is wasting of the scapular muscles with sparing of the trapezii and deltoids. Distal atrophy is seen in the legs.

1616

Diseases of the Peripheral Nervous System

had weakness of the sternocleidomastoid muscles but normal power in the arms, and in the third the upper limbs were not involved 18 years after onset. The electrophysiologic and histologic findings in this family are somewhat confusing; it is impossible to state with certainty whether the underlying defect was one of myopathy or chronic denervation. Thus, although Kaeser’s kindred was the “prototype” of autosomal dominant scapuloperoneal HMN,47 the clinical features are not entirely in keeping with those of a scapuloperoneal syndrome, and it is quite possible that the disorder was a myopathy. The same can be said of the father and son reported by Ricker and co-workers, who were described as suffering from the neurogenic scapuloperoneal syndrome.92 Both had myopathic facies. In the patient studied, there was sparing of the intrinsic foot muscles and clear evidence of myopathy on electromyography and muscle biopsy. Two more convincing pedigrees in which affected family members had scapuloperoneal HMN were described by Serratrice and associates.96 Onset was between the ages of 4 and 70 years, with more prominent involvement of the upper limbs. Bulbar signs were rare, but mild facial weakness was noted in several patients. The tendon reflexes were usually all absent. Fasciculation was uncommon. Progression of muscle weakness to the pelvic girdle was frequent but not severe. One point that must emerge from this review is that the autosomal dominant (type I) form of scapuloperoneal muscular atrophy is extremely rare. This is a significant consideration in genetic counseling. Single cases of scapuloperoneal HMN with early onset (type II) have been reported rather more frequently than autosomal dominant cases, and there is reasonable evidence to suggest that the majority of these are of autosomal recessive inheritance. Emery and co-authors described a girl who developed bilateral footdrop at the age of 4 years.22 Two years later she had severe muscle weakness of neurogenic origin, which was most marked in a scapuloperoneal distribution, but there was also some involvement of the hands and forearms. There was considerable skeletal deformity, with talipes equinovarus and joint contractures. The child was unable to stand without assistance by the age of 9 years and at that stage was noted to have mild facial weakness. Other early-onset cases (before the age of 15 years and usually before 5 years) have been described.61,69,92,94 Affected sibling pairs were reported by Feigenbaum and Munsat,24 Emery,19 Tohgi and co-workers,105 and Mercelis and colleagues.66 This early-onset, autosomal recessive form of scapuloperoneal HMN is more severe than the dominant cases described earlier. Pes cavus is common, and involvement of the small muscles of the hands is not infrequent (although to a lesser degree than the scapuloperoneal musculature). Facial weakness is relatively rare, and bulbar involvement has not been described. The exact prognosis is not known, because there have been no descriptions of cases followed for a long period. Certainly one of Emery’s

patients was still alive at the age of 30 years,19 and Munsat suggested that the disorder was nonprogressive after early childhood.69 Matawari and Katayama have suggested that there is an X-linked recessive form of scapuloperoneal HMN.57 These authors reported five affected males from one large family; the age of onset was in the latter part of the first decade. The small foot muscles were spared, and all the patients had disturbances of cardiac conduction. There was some evidence of primary muscle disease on both electromyography and muscle biopsy, and it seems highly likely that these cases are examples of Emery-Dreifuss muscular dystrophy. The chief differential diagnosis from scapuloperoneal HMN is that of the other scapuloperoneal syndromes. Distinction from scapuloperoneal myopathy may be extremely difficult, as is illustrated by Kaeser’s cases and those of Ricker and collaborators.47,92 It has been suggested that sparing of the small foot muscles is prominent in the myopathies, whereas this is not so in the neurogenic form of the disorder. Facial involvement appears to be more marked in scapuloperoneal myopathy.96 Fasciculation is not particularly common in this type of HMN, and the tendon reflexes may be depressed or absent in both HMN and myopathy of scapuloperoneal distribution. The electromyographic and histologic changes may by no means be unequivocal; in the patients reported by Mercelis and co-workers,66 the electromyographic abnormalities were initially interpreted as those of myopathy, but they later became frankly neurogenic. Ricker and associates described neurogenic and myopathic electromyographic changes in the lower and upper limbs, respectively, in one case. The presence of cardiomyopathy in a male should suggest a diagnosis of Emery-Dreifuss muscular dystrophy. The distinction of the Davidenkow syndrome95 from scapuloperoneal HMN can be made by the clinical finding of distal sensory loss and the presence of abnormal sensory nerve conduction. However, Thomas has observed two patients who initially appeared to have scapuloperoneal HMN and later developed distal sensory loss and abnormal sensory nerve action potentials.104 In the early stages of scapuloperoneal HMN, before there is upper limb involvement, it may be impossible to differentiate the disorder from the distal form of HMN.

Facioscapulohumeral Hereditary Motor Neuronopathy Hereditary motor neuronopathy of facioscapulohumeral distribution appears to be extremely rare. It was first reported by Fenichel and co-workers in 1967.25 The age of onset was in the second decade; inheritance was autosomal dominant. Facial weakness was the first symptom, followed by involvement of the shoulder and pelvic girdle muscles. Creatine phosphokinase levels were moderately raised, and

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons

electromyography suggested the features of both denervation and myopathy. It is possible that this family had facioscapulohumeral muscular dystrophy. More convincing evidence of denervation in a similar kindred was provided by Cao and associates,8 whose patients had wasting and fibrillation of the tongue. Furukawa and Toyokura reported another dominant pedigree.28 Onset was in the second decade. All three patients had marked fasciculation in the shoulder girdle musculature, and weakness was confined to the face and upper limbs in two. Electromyography showed the features of both myopathy and denervation. The most important differential diagnosis is from facioscapulohumeral muscular dystrophy. Facial weakness in cases of autosomal dominant proximal HMN is not universal and usually is mild, whereas it was severe in the families mentioned above. The prognosis in this disorder is relatively good. All the cases described have been very slowly progressive.

Hereditary Motor Neuronopathy with Oculopharyngeal Involvement Matsunaga and co-workers reported this disorder in seven members from two generations of a Japanese family.58 Inheritance could have been either autosomal dominant or recessive, because consanguinity was present. The patients were well until the fourth decade, and expression of the gene was variable. Ophthalmoplegia, dysarthria, and dysphagia were prominent in most cases, in association with distal weakness and wasting in the limbs. However, one patient had only ophthalmoplegia, and some had solely proximal involvement of the limbs.

Bulbar Hereditary Motor Neuronopathy There are two forms of progressive bulbar palsy developing in the first or second decade of life, which appear to be genetically distinct. They have the eponymic names of the Vialetto–van Laere syndrome and Fazio-Londe disease. The former is a disorder affecting the lower six cranial nerves. It was first clearly described by Vialetto,110 and further cases were reported by van Laere109 and others. The disorder usually presents in the second decade with facial weakness, dysphagia, and dysarthria. This is preceded by bilateral deafness, of rapid or slow onset, some years earlier. This is of the sensorineural type and it is associated with inexcitable labyrinths. The bulbar palsy is generally progressive but may run an episodic course. Weakness often later spreads to affect the limbs, in a proximal, distal, or generalized distribution. Intercostal muscle and diaphragmatic paralysis lead to respiratory failure, which may be heralded by sleep apnea. Nevertheless, patients may survive the fourth decade. Gallai and colleagues reported three cases and reviewed the available literature on the disease.29 The pair of siblings they described are of particular interest, because the girl did not

1617

develop cranial nerve palsies until the second decade, whereas her brother died at the age of 20 months with acute respiratory failure following some feeding difficulties and showed delay in speech development. Autopsy of this male child revealed degeneration and atrophy of the VIth to XIIth cranial nerve nuclei. The authors suggested that expression of the disorder may be more severe in males, leading to early death before the diagnosis is made. This would explain the excess of reported affected females. A further female patient with this syndrome had maternal relatives with deafness, and it was suggested that inheritance could be autosomal dominant.44 However, it is possible that this disorder is X-linked, with occasional lethality in affected males; male-to-male transmission has not been described. Hereditary progressive bulbar palsy (without deafness) of childhood was first described by Fazio in a mother and her son.23 Londe later reported a more rapidly progressive but clinically similar disorder in two brothers.54 The entity known as Fazio-Londe disease therefore is a heterogeneous syndrome. The majority of cases have onset in childhood and appear to be of autosomal recessive inheritance. Gomez and associates34 and Alexander and collaborators2 have described single cases, both of which were autopsied, and have also reviewed previously published reports. Onset is usually in the second or third year of life but may be as late as age 12 years. Excessive drooling, repeated respiratory infections, and stridor are the usual presenting features. Dysphagia, facial weakness, and extraocular palsies follow; the latter are often asymmetrical. In some patients there may be a degree of generalized weakness, wasting, hypotonia, and hyporeflexia in the limbs.1 About half the patients reported died within 18 months of onset, usually as a result of respiratory failure. Autopsy shows degeneration and atrophy of cranial nerve motor nuclei and a variable loss of anterior horn cells in the spinal cord. The pyramidal tracts are normal. A later-onset, very rapidly progressive bulbospinal motor neuronopathy was described in three sibs. Their father was similarly affected, and it seems likely that inheritance was autosomal dominant. The patients presented in the fifth to seventh decades with prominent bulbar weakness together with facial, neck, and upper limb weakness. The disorder was fatal within 18 months of onset.68

Ryukyuan Hereditary Motor Neuronopathy This appears to be a unique autosomal recessive disorder, which has been observed only in the inhabitants of islands near the southern part of Japan. Kondo and co-workers reported 32 cases.49 Bilateral weakness of the legs was noted shortly after birth, and walking was delayed, sometimes until the age of 8 years, in affected individuals. Wasting was prominent in the fourth or fifth year, and proximal weakness of the arms developed at the end of the first decade. The facial and neck muscles were preserved and there was no bulbar involvement. Some patients had fasciculation of

1618

Diseases of the Peripheral Nervous System

affected muscles. Pes cavus, scoliosis, and joint contractures were common. The tendon reflexes were depressed or absent. The electromyographic and histologic findings suggested the features of both myopathy and denervation. It was thought that all the patients were descended from a Hokuzan king who ruled the area in the 14th century.

Segmental Motor Neuronopathies There have been a number of reports, many from Japan and other parts of Asia, of patients exhibiting relatively well localized or asymmetrical muscle wasting and weakness of neurogenic origin. Most of these cases do not seem to have a genetic basis. An exception to this is a family described by Furukawa and co-workers.27 These authors reported a single case and two siblings affected with motor neuropathy that predominantly affected the quadriceps muscle. The parents of the siblings were first cousins. The age of onset in the single case was in the sixth decade. Of the two siblings, one developed weakness of the quadriceps at the age of 14 years, which was very slowly progressive; the rest of the proximal musculature in the legs was mildly involved later. The patient’s brother, however, had more severe proximal weakness of all four limbs and resembled patients with the chronic childhood form of proximal HMN. Sobue and colleagues described a series of 71 patients with neurogenic atrophy of the hand and forearm.100 This was unilateral in 47 and bilateral in 24. In one family a father and son were affected, but all the rest were single cases. The disorder was virtually confined to males; only 12 patients were female. Nearly 90% of the patients developed symptoms between the ages of 18 and 22 years (range 2 to 30 years). Wasting and weakness of the intrinsic hand muscles and the distal half and ulnar side of the proximal half of the forearm musculature was the most conspicuous finding. The brachioradialis muscle was usually spared. In most cases progression was rapid for about 2 years and then was arrested. Some of the patients reported by O’Sullivan and McLeod had a similar disorder,73 which has also been described in India by Singh and co-workers.97 Patients with chronic asymmetrical motor neuronopathy36 can be impossible to distinguish from those with motor neuron disease at the time of first presentation; similar cases were described by Meadows and associates.63 Weakness and wasting may develop in one arm or leg and do not follow a specific root distribution. There may be total arrest of the disorder at this stage, or signs of denervation can spread to other muscle groups, either gradually or after a period of 10 years or longer. Signs of pyramidal tract dysfunction are absent. The prognosis appears to be relatively good, with only very gradual motor disability; bulbar dysfunction was not observed in the 18 cases investigated by Harding and colleagues.36 None of the patients had similarly affected relatives, but it is of interest that two

had first-degree relatives with acute infantile proximal HMN (Werdnig-Hoffmann disease). The question as to whether this chronic asymmetrical denervating disease could be a heterozygote manifestation must be purely speculative. Chronic asymmetrical motor neuronopathy should be distinguished from multifocal inflammatory motor neuropathies. In the latter disorder, there is evidence of conduction block in motor nerves; sensory involvement may be inconspicuous or absent. Such patients may respond well to immunosuppressant mediation, so the distinction is important. High titers of anti–GM1 ganglioside antibodies may be found in the serum.89

ETIOLOGY AND MANAGEMENT The basic defect or defects giving rise to the hereditary motor neuropathies are unknown. It is likely that each of the different genes giving rise to these disorders results in unique enzymatic deficiencies or abnormalities of protein synthesis, all of which result in anterior horn cell degeneration. The predilection of the neurogenic atrophy for specific muscle groups in the HMNs is also unknown. No environmental factors are known to influence the expression of these genetic disorders other than precipitation of symptoms by coincidental illness and immobility. This phenomenon is well recognized as a feature of many degenerative neurologic disorders. As a result of our ignorance concerning the pathogenesis of HMN, much of this chapter has been concerned with the description and delineation of separate genetic disease entities. Unsatisfactory as this may seem, it is of great importance in terms of accurate prognostic and genetic counseling, which forms the mainstay of management of patients with HMN and their families. Supportive care is also of great importance, particularly in chronically handicapped children. There can be little doubt that the use of orthotic appliances and, if necessary, surgery can prevent or delay progressive skeletal deformity, which is a major source of morbidity.67,88 The use of assisted ventilation in chronic HMN is controversial, but for those with symptoms of nocturnal hypoventilation, assisted ventilation at night may improve quality of life substantially.32

REFERENCES 1. Aberfeld, D. C., and Namba, T.: Progressive ophthalmoplegia in Kugelberg-Welander disease. Arch. Neurol. 20:253, 1969. 2. Alexander, M. P., Emery, E. S., and Koerner, F. C.: Progressive bulbar paresis in childhood. Arch. Neurol. 33:66, 1976. 3. Armstrong, R. M., Fogelson, M. H., and Silberberg, D. H.: Familial proximal spinal muscular atrophy. Arch. Neurol. 14:208, 1966.

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons 4. Bertini, E., Gadisseux, J. L., Palmieri, G., et al.: Distal infantile spinal muscular atrophy associated with paralysis of the diaphragm: a variant of infantile spinal muscular atrophy. Am. J. Med. Genet. 33:328, 1989. 5. Brzustowicz, L. M., Lehner, T., Castilla, L. H., et al: Genetic mapping of chronic childhood-onset spinal muscular atrophy to chromosome 5q11.2-13.3. Nature 344:540, 1990. 6. Buchthal, F., and Olsen, P. Z.: Electromyography and muscle biopsy in infantile spinal muscular atrophy. Brain 93:15, 1970. 7. Bundey, S., and Lovelace, R. E.: A clinical and genetic study of chronic proximal spinal muscular atrophy. Brain 98:455, 1975. 8. Cao, A., Cianchetti, C., Calisti, L., and Tangheroni, W.: A family of juvenile proximal spinal muscular atrophy with dominant inheritance. J. Med. Genet. 13:131, 1976. 9. Clarke, A., Davies, K. E., Gardner-Medwin, D., et al.: Xp21 DNA probe in diagnosis of muscular dystrophy and spinal muscular atrophy. Lancet 1:443, 1989. 10. Davidenkow, S.: Über die neurotische Muskelatrophie Charcot-Marie: Klinisch-genetische studien. Z. Ges. Neurol. Psychiat. 107:259, 1927. 11. Davidenkow, S.: Scapuloperoneal amyotrophy. Arch. Neurol. Psychiatry 41:694, 1939. 12. Davis, C. J. F., Bradley, W. G., and Madrid, R.: The peroneal muscular atrophy syndrome: clinical, genetic, electrophysiological and nerve biopsy studies. J. Génét. Hum. 26:311, 1978. 13. Drachman, D. B., Murphy, S. R., Nigam, M. P., and Hill, J. R.: “Myopathic” changes in chronically denervated muscle. Arch. Neurol. 16:14, 1967. 14. Dubowitz, V.: Infantile muscular atrophy: a prospective study with particular reference to a slowly progressive variety. Brain 87:707, 1964. 15. Dubowitz, V.: The floppy infant—a practical approach to classification. Dev. Med. Child Neurol. 10:706, 1968. 16. Dubowitz, V.: Muscle Disorders in Childhood. Philadelphia, W. B. Saunders, p. 149, 1978. 17. Dubowitz, V., and Brooke, M.: Muscle Biopsy: A Modern Approach. Philadelphia, W. B. Saunders, p. 105, 1973. 18. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. II. Neurologic, genetic and electrophysiologic findings in various neuronal degenerations. Arch. Neurol. 18:619, 1968. 19. Emery, A. E. H.: Review: the nosology of the spinal muscular atrophies. J. Med. Genet. 8:481, 1971. 20. Emery, A. E. H., Hausmanowa-Petrusewicz, I., Davie, A. M., et al.: International collaborative study of the spinal muscular atrophies. Part I. J. Neurol. Sci. 29:83, 1976. 21. Emery, A. E. H., Hausmanowa-Petrusewicz, I., Davie, A. M., et al.: International collaborative study of the spinal muscular atrophies. Part II. J. Neurol. Sci. 30:375, 1976. 22. Emery, E. S., Fenichel, G. M., and Engel, G.: A spinal muscular atrophy with scapuloperoneal distribution. Arch. Neurol. 18:129, 1968. 23. Fazio, M.: Ereditarietà della paralisi bulbare progressiva. Riforma Med. 8:327, 1892. 24. Feigenbaum, J. A., and Munsat, T. L.: A neuromuscular syndrome of scapuloperoneal distribution. Bull. Los Angeles Neurol. Soc. 35:47, 1970.

1619

25. Fenichel, G. M., Emery, E. S., and Hunt, P.: Neurogenic atrophy simulating facioscapulohumeral dystrophy. Arch. Neurol. 17:257, 1967. 26. Fischbeck, K. H., Ionasescu, V., Ritter, A. W., et al.: Localization of the gene for X-linked spinal muscular atrophy. Neurology 36:1595, 1986. 27. Furukawa, T., Akagami, N., and Maruyama, S.: Chronic neurogenic quadriceps amyotrophy. Ann. Neurol. 2:528, 1977. 28. Furukawa, T., and Toyokura, Y.: Chronic spinal muscular atrophy of facioscapulohumeral type. J. Med. Genet. 13:285, 1976. 29. Gallai, V., Hockaday, J. M., Hughes, J. T., et al.: Ponto-bulbar palsy with deafness (Brown-Vialetto–van Laere syndrome): a report on 3 cases. J. Neurol. Sci. 50:250, 1981. 30. Gamstorp, I.: Progressive spinal muscular atrophy with onset in infancy or early childhood. Acta Paediatr. Scand. 56:408, 1967. 31. Gardner-Medwin, D., Hudgson, P., and Walton, J. N.: Benign spinal muscular atrophy arising in childhood and adolescence. J. Neurol. Sci. 5:121, 1967. 32. Gilgoff, I. S., Kahlstrom, E., MacLaughlin, E., and Keens, T. G.: Long-term ventilatory support in spinal muscular atrophy. J. Pediatr. 115:904, 1989. 33. Gilliam, T. C., Brzustowicz, L. M., Castilla, L. H., et al.: Genetic homogeneity between acute and chronic forms of spinal muscular atrophy. Nature 345:823, 1990. 34. Gomez, M. R., Clermont, V., and Berstein, J.: Progressive bulbar paralysis in childhood (Fazio-Londé disease). Arch. Neurol. 6:317, 1962. 35. Greenberg, F., Fenolio, K. R., Hejtmancik, J. F., et al.: X-linked infantile spinal muscular atrophy. Am. J. Dis. Child. 142:217, 1988. 36. Harding, A. E., Bradbury, P. G., and Murray, N. M. F.: Chronic asymmetrical spinal muscular atrophy. J. Neurol. Sci. 59:69, 1983. 37. Harding, A. E., and Thomas, P. K.: Hereditary distal spinal muscular atrophy: a report on 34 cases and a review of the literature. J. Neurol. Sci. 45:337, 1980. 38. Harding, A. E., Thomas, P. K., Baraitser, M., et al.: X-linked recessive bulbo-spinal neuronopathy: a report of ten cases. J. Neurol. Neurosurg. Psychiatry 45:1012, 1982. 39. Hausmanowa-Petrusewicz, I.: Electrophysiological findings in childhood spinal muscular atrophies. Rev. Neurol. (Paris) 144:716, 1988. 40. Hausmanowa-Petrusewicz, I., Askanas, W., and Badurska, B.: Infantile and juvenile spinal muscular atrophy. J. Neurol. Sci. 6:269, 1968. 41. Hausmanowa-Petrusewicz, I., Zaremba, J., and Borkowska, J.: Chronic form of childhood spinal muscular atrophy: are the problems of its genetics really solved? J. Neurol. Sci. 43:313, 1979. 42. Hausmanowa-Petrusewicz, I., Zaremba, J., and Borkowska, J.: Chronic proximal spinal muscular atrophy of childhood and adolescence: problems of classification and genetic counselling. J. Med. Genet. 22:350, 1985. 43. Hausmanowa-Petrusewicz, I., Zaremba, J., Borkowska, J., and Prot, J.: Genetic investigations on chronic forms of infantile and juvenile spinal muscular atrophy. J. Neurol. 213:335, 1976.

1620

Diseases of the Peripheral Nervous System

44. Hawkins, S. A., Nevin, N. C., and Harding, A. E.: Pontobulbar palsy and neurosensory deafness (Brown-Vialetto–van Laere syndrome) with possible autosomal dominant inheritance. J. Med. Genet. 27:176, 1990. 45. Hoffmann, J.: Über chronische spinal Muskelatrophie im Kindesalter, auf familiärer Basis. Dtsch. Z. Nervenheilkd. 3:427, 1893. 46. Hurwitz, L., Lapresle, J., and Garcin, R.: Atrophie musculaire neurogene de topographie proximale et symmetrique simulant une myopathie: étude de deux observations. Rev. Neurol. (Paris) 104:97, 1961. 47. Kaeser, H. E.: Scapuloperoneal muscular atrophy. Brain 88:407, 1965. 48. Kennedy, W. R., Alter, M., and Sung, J. H.: Progressive proximal spinal and bulbar muscular atrophy of late onset: a sex-linked recessive trait. Neurology (Minneap.) 18:671, 1968. 49. Kondo, K., Tsubaki, T., and Sakamoto, F.: The Ryukyuan muscular atrophy: an obscure heritable disease found in the islands of southern Japan. J. Neurol. Sci. 11:359, 1970. 50. Jagganathan, K.: Juvenile motor neurone disease. In Spillane, J. D. (ed.): Tropical Neurology. London, Oxford University Press, p. 127, 1973. 51. Kugelberg, E., and Welander, L.: Heredofamilial juvenile muscular atrophy simulating muscular dystrophy. Arch. Neurol. Psychiatry 75:500, 1956. 52. La Spada, A. R., Wilson, E. M., Lubahn, D. B., et al.: Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352:77, 1991. 53. Lander, C. M., Eadie, M. J., and Tyrer, J. H.: Hereditary motor peripheral neuropathy predominantly affecting the arms. J. Neurol. Sci. 28:389, 1976. 54. Londe, P.: Paralysie bulbaire progressive infantile et familiale. Rev. Méd. 14:212, 1894. 55. Lugaresi, E., Gambetti, P., and Rossi, P. G.: Chronic neurogenic muscle atrophies of infancy: their nosological relationship with Werdnig-Hoffmann’s disease. J. Neurol. Sci. 3:399, 1966. 56. Magee, K. R., and de Jong, R. N.: Neurogenic muscular atrophy simulating muscular dystrophy. Arch. Neurol. 2:677, 1960. 57. Matawari, S., and Katayama, K.: Scapulo-peroneal muscular atrophy with cardiopathy. Arch. Neurol. 28:55, 1973. 58. Matsunaga, M., Inokuchi, T., Ohnishi, A., and Kuroiwa, Y.: Oculopharyngeal involvement in familial neurogenic muscular atrophy. J. Neurol. Neurosurg. Psychiatry 36:104, 1973. 59. McLeod, J. G., and Prineas, J. W.: Distal type of chronic spinal muscular atrophy—clinical, electrophysiological and pathological studies. Brain 94:703, 1971. 60. Meadows, J. C., and Marsden, C. D.: A distal form of chronic spinal muscular atrophy. Neurology (Minneap.) 19:53, 1969. 61. Meadows, J. C., and Marsden, C. D.: Scapuloperoneal amyotrophy. Arch. Neurol. 20:9, 1969. 62. Meadows, J. C., Marsden, C. D., and Harriman, D. G. F.: Chronic spinal muscular atrophy in adults. Part 1. The Kugelberg-Welander syndrome. J. Neurol. Sci. 9:527, 1969. 63. Meadows, J. C., Marsden, C. D., and Harriman, D. G. F.: Chronic spinal muscular atrophy in adults. Part 2. Other forms. J. Neurol. Sci. 9:551, 1969.

64. Melki, J., Abdelhak, S., Sheth, P., et al.: Gene for chronic proximal spinal muscular atrophies maps to chromosome 5q. Nature 344:767, 1990. 65. Melki, J., Sheth, P., Abdelhak, S., et al., and the French Spinal Muscular Atrophy Investigators: Mapping of acute (type 1) spinal muscular atrophy to chromosome 5q12-q14. Lancet 336:271, 1990. 66. Mercelis, R., Demeester, J., and Martin, J. J.: Neurogenic scapuloperoneal syndrome in childhood. J. Neurol. Neurosurg. Psychiatry 43:888, 1980. 67. Merlini, L., Granata, C., Bonfiglioli, S., et al.: Scoliosis in spinal muscular atrophy: natural history and management. Dev. Med. Child Neurol. 31:501, 1989. 68. Mikol, J., Bourdarias, H., Dellanave, J., and Perie, G.: Familial bulbospinal atrophy in adults. Rev. Neurol. (Paris) 144:748, 1988. 69. Munsat, T. L.: Infantile scapuloperoneal muscular atrophy. Neurology (Minneap.) 18:285, 1968. 70. Munsat, T. L., Woods, R., Fowler, W., and Pearson, C. M.: Neurogenic muscular atrophy of infancy with prolonged survival. Brain 92:9, 1969. 71. Namba, L., Aberfeld, D. C., and Grob, D.: Chronic proximal spinal muscular atrophy. J. Neurol. Sci. 11:401, 1970. 72. Nelson, J. W., and Amick, L. D.: Heredofamilial progressive spinal muscular atrophy: a clinical and electromyographic study of a kinship. Neurology 16:306, 1966. 73. O’Sullivan, D. G., and McLeod, J. G.: Distal chronic spinal muscular atrophy involving the hands. J. Neurol. Neurosurg. Psychiatry 41:653, 1978. 74. Palmer, H. D.: Familial scapuloperoneal amyotrophy. Arch. Neurol. Psychiatry 28:473, 1932. 75. Parnes, S., Karpati, G., Carpenter, S., et al.: Hexosaminidase A deficiency presenting as atypical juvenile-onset spinal muscular atrophy. Arch. Neurol. 42:1176, 1985. 76. Pearn, J. H.: Fetal movements and Werdnig-Hoffmann disease. J. Neurol. Sci. 18:373, 1973. 77. Pearn, J. H.: The gene frequency of acute Werdnig-Hoffmann disease (SMA type I): a total population survey in North-East England. J. Med. Genet. 10:260, 1973. 78. Pearn, J. H.: Autosomal dominant spinal muscular atrophy: a clinical and genetic study. J. Neurol. Sci. 38:263, 1978. 79. Pearn, J. H.: Incidence, prevalence, and gene frequency studies of chronic childhood spinal muscular atrophy. J. Med. Genet. 15:409, 1978. 80. Pearn, J. H.: Classification of the spinal muscular atrophies. Lancet 1:919, 1980. 81. Pearn, J. H., Bundey, S., Carter, C. O., et al.: A genetic study of subacute and chronic spinal muscular atrophy in childhood: a nosological analysis of 124 index patients. J. Neurol. Sci. 37:227, 1978. 82. Pearn, J. H., Carter, C. O., and Wilson, J.: The genetic identity of acute infantile spinal muscular atrophy. Brain 96:463, 1973. 83. Pearn, J. H., Gardner-Medwin, D., and Wilson, J.: A clinical study of chronic and childhood spinal muscular atrophy: a review of 141 cases. J. Neurol. Sci. 38:23, 1978. 84. Pearn, J. H., and Hudgson, P.: Anterior horn cell degeneration and gross calf hypertrophy with adolescent onset. Lancet 1:1059, 1978.

Inherited Neuronal Atrophy and Degeneration Predominantly of Lower Motor Neurons 85. Pearn, J. H., and Hudgson, P.: Distal spinal muscular atrophy—a clinical and genetic study of 8 kindreds. J. Neurol. Sci. 43:183, 1979. 86. Pearn, J. H., Hudgson, P., and Walton, J. N.: A clinical and genetic study of adult-onset spinal muscular atrophy: the autosomal recessive form as a discrete disease entity. Brain 101:591, 1978. 87. Pearn, J. H., and Wilson, J.: Acute Werdnig-Hoffmann disease (acute infantile spinal muscular atrophy). Arch. Dis. Child. 48:425, 1973. 88. Pearn, J. H., and Wilson, J.: Chronic generalized spinal muscular atrophy of infancy and childhood. Arch. Dis. Child. 48:768, 1973. 89. Pestronk, A., Chaudhry, V., Feldman, E. L., et al.: Lower motor neuron syndromes defined by patterns of weakness, nerve conduction abnormalities, and high titers of antiglycolipid anti-bodies. Ann. Neurol. 27:316, 1990. 90. Peters, H. A., Opitz, J. M., Goto, I., and Reese, H. H.: The benign proximal spinal progressive muscular atrophies. Acta Neurol. Scand. 44:542, 1968. 91. Pouget, J.: Amyotrophies scapulo-péronières chroniques de type stark-Kaeser (à propos de 10 observations). Rev. Neurol. (Paris) 132:823, 1976. 92. Ricker, K., Mertens, H. G., and Schimrigk, K.: The neurogenic scapulo-peroneal syndrome. Eur. Neurol. 1:257, 1968. 93. Ringel, S. P., Lava, N. S., Treihaft, M. M., et al.: Late-onset X-linked recessive spinal and bulbar muscular atrophy. Muscle Nerve 1:297, 1978. 94. Schuchmann, L.: Spinal muscular atrophy of the scapuloperoneal type. Z. Kinderheilkd. 109:118, 1970. 95. Schwartz, M. S., and Swash, M.: Scapulo-peroneal atrophy with sensory involvement: Davidenkow’s syndrome. J. Neurol. Neurosurg. Psychiatry 38:1063, 1975. 96. Serratrice, G., Roux, H., Aquaron, R., et al.: Myopathies scapulo-péronières. (A propos de 14 observations dont 8 avec atteinte faciale). Sem. Hôp. Paris 45:2678, 1969. 97. Singh, N., Sachdev, K., and Susheela, A. K.: Juvenile muscular atrophy localized to arms. Arch. Neurol. 37:297, 1980. 98. Smith, J. B., and Patel, A.: The Wohlfart-Kugelberg-Welander disease. Neurology 15:469, 1965. 99. Sobue, G., Hashizume, Y., Mukai, E., et al.: X-linked recessive bulbospinal neuronopathy: a clinico-pathological study. Brain 112:209, 1989. 100. Sobue, I., Saito, N., Iida, M., and Ando, K.: Juvenile type of distal and segmental muscular atrophy of upper extremities. Ann. Neurol. 3:429, 1978. 101. Spiro, A. J., Fogelson, M. H., and Goldberg, A. C.: Microcephaly and mental subnormality in chronic progressive spinal muscular atrophy of childhood. Dev. Med. Child Neurol. 9:594, 1967.

1621

102. Stefanis, C., Papapetropoulos, T., Scarpalezos, S., et al.: X-linked spinal and bulbar muscular atrophy of late-onset. J. Neurol. Sci. 24:493, 1975. 103. Summers, B. A., Swash, M., Schwartz, M. S., and Ingram, D. A.: Juvenile-onset bulbospinal muscular atrophy with deafness: Vialetto–van Laere syndrome or Madras-type motor neuron disease? J. Neurol. 234:440, 1987. 104. Thomas, P. K.: Discussion on the scapuloperoneal syndrome. In Serratrice, G., and Roux, H. (eds.): Peroneal Atrophies and Related Disorders. New York, Masson, p. 230, 1979. 105. Tohgi, H., Tsukagoshi, H., and Toyokura, Y.: Neurogenic scapuloperoneal syndrome with autosomal recessive inheritance. Clin. Neurol. (Tokyo) 11:215, 1971. 106. Tsukagoshi, H., Nakanishi, T., Kondo, K., and Tsubaki, T.: Hereditary proximal neurogenic muscular atrophy in adult. Arch. Neurol. 12:597, 1965. 107. Tsukagoshi, H., Shoji, H., and Furukawa, T.: Proximal neurogenic atrophy in adolescence and adulthood with X-linked recessive inheritance—Kugelberg-Welander disease and its variant of late onset in one pedigree. Neurology (Minneap.) 20:1188, 1970. 108. Tsukagoshi, H., Sugita, H., Furukawa, T., et al.: KugelbergWelander syndrome with dominant inheritance. Arch. Neurol. 14:378, 1966. 109. Van Laere, J.: Paralysie bulbo-pontine chronique progressive familiale avec surdité. Rev. Neurol. (Paris) 115:289, 1966. 110. Vialetto, E.: Contributo alla forma ereditaria della paralisi bulbare progressiva. Riv. Sper. Fren. 40:1, 1936. 111. Werdnig, G.: Zwei frühinfantile hereditäre Fälle von progressiven Muskelatrophie unter dem Bilder der Dystrophie, uber auf neurotischer Grundlage. Arch. Psychiat. 22:437, 1891. 112. Wohlfart, G., Fex, J., and Eliasson, S.: Hereditary proximal spinal muscular atrophy: a clinical entity simulating progressive muscular dystrophy. Acta Psychiatr. Neurol. Scand. 30:395, 1955. 113. Young, I. D., and Harper, P. S.: Hereditary distal spinal muscular atrophy with vocal cord paralysis. J. Neurol. Neurosurg. Psychiatry 43:413, 1980. 114. Zellweger, H., Schneider, D. R., Schuldt, D. R., and Mergner, W.: Heritable spinal muscular atrophies. Helv. Paediatr. Acta 24:92, 1969. 115. Zellweger, H., Simpson, J., McCormick, W. F., and Ionasescu, V.: Spinal muscular atrophy with autosomal dominant inheritance. Neurology (Minneap.) 22:957, 1972. 116. Zerres, K.: Genetics of proximal spinal muscular atrophies. In Merlini, L., Granata, C., and Dubowitz, V. (eds.): Current Concepts in Childhood Spinal Muscular Atrophy. Vienna, Springer, p. 69, 1989.

H. Hereditary Motor and Sensory Neuropathies

69 Hereditary Motor and Sensory Neuropathies: An Overview of Clinical, Genetic, Electrophysiologic, and Pathologic Features MICHAEL E. SHY, JAMES R. LUPSKI, PHILLIP F. CHANCE, CHRISTOPHER J. KLEIN, AND PETER J. DYCK

Historical Overview Description of Syndromes Classification of Inherited Neuropathies Gene Identification Studies Epidemiology Clinical Features Overview Muscle Weakness Sensory Dysfunction

Autonomic Dysfunction Bone and Joint Abnormalities Hypertrophic Nerves Tremor Nerve Conduction and Electromyographic Features Pathologic Features HMSN I Déjérine-Sottas Neuropathy HMSN II

The disorders discussed in this chapter are inherited neuropathies without known metabolic derangements. Many of these disorders are known as hereditary motor and sensory neuropathy (HMSN) syndromes or Charcot-MarieTooth (CMT) syndromes. The term Charcot-Marie-Tooth syndrome, or peroneal muscular atrophy, was originally believed to be specific for one disorder.8,27,66,152 It is now known that peroneal muscular atrophy occurs in several inherited neuromuscular disorders. In fact, in 1968 Dyck and Lambert48,49 suggested that there were several disorders with peroneal muscular atrophy: a hypertrophic neuropathy variety (type I), a neuronal variety (type II), a progressive muscular atrophy variety, and other varieties. Later the term hereditary motor and sensory neuropathy was introduced to encompass a broader group of inherited neuropathy syndromes than encompassed by peroneal muscular atrophy or

Schwann Cell–Axon Interactions Unusual Forms of HMSN HMSN VI with Optic Atrophy HMSN with Retinitis Pigmentosa Giant Axonal Neuropathy Course and Outcome Treatment Traditional Therapies New Therapeutic Approaches

CMT syndrome. In part this was done to set these disorders apart from other system atrophies that affected only motor neurons (cryptogenic and inherited motor neuron disease and progressive muscular atrophies) or primary sensory neurons (spinocerebellar degenerations [large-diameter fiber involvement] and hereditary sensory and autonomic neuropathies [HSANs, especially small fiber involvement]), and from inherited autonomic neuron involvement. Since the previous edition of Peripheral Neuropathy, many of the genetic causes of HMSN syndromes have been identified. These are discussed subsequently as separate chapters in this section. In addition, other inherited neuropathies, some with known metabolic derangement, are discussed as separate chapters. With respect to nomenclature, we utilize HMSN, CMT syndrome, or both where it is appropriate to do so. 1623

1624

Diseases of the Peripheral Nervous System

HISTORICAL OVERVIEW The history of the HMSNs can be divided into three phases: (1) the description of syndromes; (2) classification based on types of inheritance, natural history, electrophysiologic, and morphologic findings; and (3) gene identification.

Description of Syndromes Peroneal Muscular Atrophy or CMT Syndrome In the late 19th century, Charcot and Marie in France27 and Tooth in England152 described the features of peroneal muscular atrophy. Charcot and Marie listed the following features as characteristic: (1) progressive muscular atrophy involving the feet and legs first and, second, after many years, affecting the hands and, later still, the forearms; (2) fibrillary contractions of atrophic muscles; (3) vasomotor abnormalities; (4) lack of joint contractures; (5) normal sensation (probably an incorrect conclusion); (6) frequent cramps; (7) reaction of degeneration in atrophic muscles; (8) frequent onset in infancy (symptomatic onset occurs at a later age in many patients); and (9) occurrence of the disorder in the “same generation and in succeeding generations.” Charcot and Marie emphasized tightness of the Achilles tendons, pes cavus, atrophy of leg and thigh muscles so that the lower limbs resembled an inverted champagne bottle, hammer toes and claw hand, steppage gait, and in some persons constant shuffling of feet when standing in one place to maintain balance. In addition to footdrop and atrophy of feet and leg muscles, one of their patients manifested lancinating pains, sensory loss of the distal parts of the limbs, and abnormalities of tactile, pain, and temperature sensation. Today the latter patient would probably be classified as having HSAN type I (see Chapter 78). Tooth contrasted his peroneal type of progressive muscular atrophy to “classic” progressive muscular atrophy.152 His progressive peroneal muscular atrophy was inherited and began early in life. Atrophy was symmetrical, distal (peroneal muscles), and in the lower limbs, unlike the classic variety (now called motor neuron disease or amyotrophic lateral sclerosis) in which atrophy spread from the small hand muscles up the arms to the shoulders, to the back, and then to the lower limbs. Tooth’s description of the features of the peroneal type was similar in many respects to the features listed by Charcot and Marie, and he concluded that the disorder was due to disease of the peripheral nerves.152 Roussy-Lévy Syndrome Roussy and Lévy described an autosomal dominant disorder with many of the features of peroneal muscular atrophy (clubfeet, difficulty in walking and standing still, a tendency to atrophy of the hands, generalized areflexia, and absence of Babinski’s sign).127–129,166 Clumsiness in walking and in use of the hands was emphasized, but cerebellar ataxia was not found. The trembling of the hands that Roussy and Lévy observed is the essential feature of the syndrome; however,

the etiology of the tremor remains confusing. In part it may be the result of impaired position sense, or “pseudoathetosis.” However, it may also be in part centrally mediated, given the variability of overt position sense loss in many patients with the tremor. Some investigators, reading the original case descriptions and seeing present-day cases, believe the unsteadiness in walking can be attributed in part to muscle weakness and perhaps sensory loss. They feel that the hand trembling, frequently brought on by activity, appears to be typical of what has been called essential tremor.166 Additional kinships with a Roussy-Lévy–like syndrome were subsequently described.29,68,121,143,153 Several early kindreds with the Roussy-Lévy syndrome were found to have sensory as well as motor involvement. In addition, they were found to have slow nerve conduction velocities (NCVs).43,166 Moreover, in both clinical and morphometric evaluation of cutaneous nerves, Dyck and colleagues detected nerve enlargement and onion bulb formation.43 Taking these findings together, therefore, these authors postulated that most cases of Roussy-Lévy syndrome were similar to the hypertrophic form of CMT syndrome, subsequently characterized as HMSN I. This has proven to be the case since, recently, the original family evaluated by Roussy and Lévy has been shown to have a mutation in the myelin protein zero (MPZ) gene,120 now classified as HMSN IB (see below). However, other patients who have mutations in the peripheral myelin protein 22 (PMP22) gene, as well as the duplication of PMP22 responsible for the most frequent form of HMSN, HMSN IA, also have similar tremors.150 In addition, affected patients within the same family may or may not have tremor. The Roussy-Lévy syndrome is thus now known to be a phenotypic variant, with variable expression, of several forms of HMSN I rather than a separate disease entity. Déjérine-Sottas Neuropathy and Congenital Hypomyelinating Neuropathy A number of authors, among them Virchow158 and Friedreich,57 had described cases of inherited hypertrophic neuropathy by the time Gombault and Mallet60 and Déjérine and Sottas35 gave their descriptions of an unusual inherited hypertrophic neuropathy. A recessive inheritance pattern was reported by Davidenkow31,32 and Dyck and Gomez42 to distinguish these disorders from the dominantly inherited HMSN I. Typically cases begin in infancy, and may be accompanied by delayed motor milestones. They are generally associated with severe pathologic alterations and nerve conduction abnormalities (usually ⬍ 10 m/s).38,43 Déjérine and Sottas described two siblings with this disorder.35 The parents were unaffected. In the sister the symptomatic onset was in infancy, and in the brother it occurred at 14 years. The signs of the disorder included clubfeet, kyphoscoliosis, and generalized muscle weakness and atrophy, with the greatest severity in distal limb muscles. There were also muscle fasciculations, decreased reaction to electrical stimulation, areflexia, sensory loss, especially over distal limbs, and incoordination of the upper limbs. Finally,

1625

Hereditary Motor and Sensory Neuropathies: Overview of Features

the patients had Romberg’s sign, miosis, decreased pupillary response to light, and nystagmus. Prior to death at age 45 years, the more severely affected patient was unable to walk. At autopsy, peripheral nerves were described as enlarged, firm, and gelatinous. Although many fibers were said to be present, the myelin sheath was found to be either thin or absent. The authors described a cuff of hypertrophic connective tissue, with increased numbers of nuclei, surrounding the abnormal fibers that assumed a circumferential direction within the cuff. Peripheral nerve changes were more pronounced distally, although similar changes were also found in the roots. There was some loss of ventral horn perikarya and demyelination of posterior columns. Déjérine and Sottas concluded that the main abnormality lay in a hyperplastic endoneurium that constricted and damaged nerve fibers. As is discussed below, this conclusion has not turned out to be correct. Gombault and Mallet60 found similar histologic changes but attributed them to tabes. Studies by Déjérine and Sottas and others performed prior to the era of molecular diagnosis suggested that the inheritance might be recessive in children with DéjérineSottas neuropathy (DSN). For example, in the Mayo Clinic experience, in clinical examination and nerve conduction studies of 5 affected and 13 unaffected close relatives, as well as subsequent evaluation of 2 affected and 3 unaffected close relatives, investigators found consanguinity in two sets of parents and no evidence of neuropathic involvement in any of the five sets of parents.48 However, Harding and Thomas68 suggested that the severe, early-onset group of patients “is probably a genetically heterogeneous group,” and this has subsequently proven to be the case. It has now been shown that many presumed DSN patients have dominantly inherited mutations caused by new mutations in PMP22, MPZ, and the early growth response element 2 (EGR2) gene (see below). Whereas many NCVs in DSN patients were extremely slow (⬍10 m/s), some severely disabled children did not have slow NCVs. Moreover, whereas sural nerve biopsies in many DSN children revealed severe demyelination and hypertrophy, others revealed predominantly axonal loss. To minimize confusion, in this chapter we use the term Déjérine-Sottas neuropathy to define all patients with severe onset in infancy. Specifically, we use the following diagnostic criteria: (1) onset by 2 years of age with delayed motor milestones; and (2) severe motor, sensory, and skeletal deficits with frequent extension to proximal muscles, sensory ataxia, and scoliosis. Other supportive findings include markedly slowed motor NCVs (⬍10 m/s) and increased cerebrospinal fluid (CSF) protein levels. Congenital hypomyelination (CH) is a pathologically based term originally used to describe peripheral nerves that were so abnormal that they suggested a developmental failure of peripheral nervous system (PNS) myelination.86,87,103 Patients with CH were noted to be hypotonic within the first year of life, to have developmental delays in walking, and, in some cases, to have swallowing or respiratory difficulties.64 Furthermore, CH patients can present as “floppy” infants.

Patients classified as having either CH or DSN have been reported with the same severe pathologic changes on sural nerve biopsies. Patients with CH or DSN have had very slow NCVs (⬍10 m/s). Finally, the same patient has been diagnosed as having CH in one publication144 and DSN in another.70 Perhaps the only clinically distinguishing characteristic of CH is presentation at birth for some cases. However, taken together, these data suggest that it is difficult to distinguish between DSN and CH; we utilize the latter designation only in its morphologic context in this chapter.

Classification of Inherited Neuropathies In part because not all patients with inherited neuropathies had autosomal dominant inheritance patterns, or similar natural histories of disease, it became apparent early on that not all patients could be simply classified as having HMSN I or HMSN II. Therefore, detailed classification systems were developed. Davidenkow may have been the first to propose a classification system based on clinical and genetic features.31,32 Subsequently, however, Dyck and Lambert developed their landmark classification system in 1968, briefly touched on above.48,49 This system, shown in Table 69–1, Table 69–1. Dyck and Lambert Classification of Hereditary Motor and Sensory Neuropathies Type

Features

HMSN I A and B (dominantly inherited hypertrophic neuropathy)

Slow nerve conduction velocities Distal weakness, mild sensory loss Palpable nerves Decreased reflexes Pathology demonstrating segmental demyelination, remyelination, onion bulb formation, and axonal loss Autosomal dominant Normal nerve conduction velocities Distal weakness, mild sensory loss Nonpalpable nerves Pathology demonstrating degeneration of motor and sensory nerves Autosomal dominant Delayed motor development Severe motor-sensory loss Slow nerve conduction velocities Autosomal recessive Refsum’s disease

HMSN II (neuronal type of peroneal muscular atrophy)

HMSN III Hypertrophic neuropathy of infancy (Déjérine and Sottas) HMSN IV Hypertrophic neuropathy (Refsum) (associated with phytanic acid excess) HMSN V (associated with spastic paraplegia) HMSN VI (with optic atrophy) HMSN VII

Spastic paraplegia present

Optic atrophy present Retinitis pigmentosa present

1626

Diseases of the Peripheral Nervous System

remains the basis of the classification of the inherited neuropathies today. HMSN I patients were subdivided into (1) those with clinical phenotypes similar to those originally described by Charcot, Marie, and Tooth or by Roussy and Lévy (HMSN I), and (2) those with what were thought to be the recessive disorders Déjérine-Sottas syndrome (HMSN III) and Refsum’s disease, or phytanic acid storage disease (HMSN IV). Of those patients classified as HMSN I, approximately three fourths of the cases were inherited in an autosomal dominant fashion (Fig. 69–1). In the remaining quarter, the disorder either was sporadic or was found in siblings, but parents were not affected. In the families with affected siblings, this most likely represents recessive inheritance; however, it could represent germline mosaicism for a new dominant mutation. Dyck and Lambert also realized that the neuronal form of HMSN was a heterogeneous group that could be subclassified based on whether motor and sensory nerves were clinically affected, without nerve hypertrophy (HMSN II); whether only lower motor neurons were affected (a distal form of hereditary motor neuronopathy [HMN V]); whether spastic paraplegia was present (HMSN V); or whether other particular neurologic abnormalities such as optic atrophy (HMSN VI) or retinitis pigmentosa (HMSN VII) were present. While this classification proved invaluable to investigators who subsequently identified the genetic cause of many of

these neuropathies, the discovery of the causal genes also led to the need to modify the classification system for several reasons. First, different mutations in the same gene, such as MPZ or PMP22 (see below), can cause either typical HMSN I phenotypes or severe early-onset neuropathies. Many of these early-onset cases present sporadically (see Chapter 71 for examples), with no family history, and would have previously been diagnosed as HMSN III, or DSN. Second, X-linked HMSN (HMSN X) does not fit easily into the original Dyck and Lambert classification both because of its X-linked inheritance and because of its intermediately slowed NCV. Finally, different mutations in the same gene, such as MPZ, may cause slow NCV characteristic of HMSN I or near-normal velocities more characteristic of HMSN II. With this in mind, the classification scheme has been modified to that illustrated in Table 69–2, which is based on the genetic cause of the neuropathy and its inheritance pattern. Neuropathies that are caused by genes expressed in Schwann cells, the myelinating cell of the PNS, and are inherited as dominant disorders are classified as HMSN I, whereas dominantly inherited neuropathies caused by mutations in neuronally expressed genes are classified as HMSN II. Mutations in Schwann cell genes may pathologically primarily affect myelin or the axon, depending on the gene and mutation (see Chapters 71 and 76 for examples). Similarly, mutations in some neuronally expressed genes, such as neurofilament light, may disrupt saltatory conduction and

FIGURE 69–1 Large kinship with HMSN type I shows autosomal dominant inheritance. (From Dyck, P. J., Lambert, E. H., and Mulder, D. W.: Charcot-Marie-Tooth disease: nerve conduction and clinical studies of a large kinship. Neurology [Minneap.] 13:1, 1963, with permission.)

1627

Hereditary Motor and Sensory Neuropathies: Overview of Features

Table 69–2. Classification of Hereditary Motor and Sensory Neuropathies Disease

Inheritance Pattern*

Locus

Gene

HMSN I HMSN IA (#118220)† HMSN IB (#118200) HMSN IC (#601098) HMSN ID (#607678)

AD AD AD AD

17p12 1q21-23 16p13.1-12.3 10q21-22

PMP22 duplication and mutation MPZ mutation LITAF/SIMPLE mutation EGR2 mutation

HNPP (#162500)

AD

17p12

PMP22 deletion

HMSN II HMSN IIA (#118210) HMSN IIB (#600882) HMSN IIC (#606071) HMSN IID (#601472) HMSN IIE (#607684) HMSN IIF (#606595)

AD AD AD AD AD AD

1p36.2 3q21 12q23-24 7p15 8p21 7q11-q21

MFN2 mutation (also KIF1B␤) RAB7 mutation ? GARS mutation NEFL mutation HSP27 mutation

HMSN X HMSN X1 (#302800) HMSN X2 (#310490)

XD XD

Xq13-22 Xq24-q26.1

GJB1 mutation ?

HMSN IV HMSN IVA (#214400) HMSN IVB1 (#605588) HMSN IVB2 (#604563) HMSN IVC (#601596) HMSN IVD (#601455) HMSN IVF (#605735) HMSN Russe (#605285) GAN (#256850) CCFDN (#604168)

AR AR AR AR AR AR AR AR AR

8q13 11q23 11p15 5q23-33 8q24.3 19q13.3 10q23.2 1p35-36 18qter

GDAP1 mutation MTMR2 mutations SBF2/MTMR13 mutation KIAA1985 NDRG1 mutations Periaxin (PRX) mutation ? GAN mutations CTDP1 mutation

Distal Hereditary Motor Neuropathies and Neuronopathies dHMN I dHMN II (#158590) dHMN III dHMN V (#600794) (same as HMSN IID) dHMN VI dHMN VII dHMN Jerash (#605726) ALS4 (#602433) HMN Dynactin (#607641)

AD AD AR AD AR AD AR AD AD

? 12q24.3 1q21-23 7p15 ? ? 9p21.1-p12 9q34 2p13

? HSP22 ? GARS mutations ? ? ? SETX DCTN1 mutations

HSAN HSAN I (#162400) HSAN II (#201300) HSAN III (#223900) (Riley-Day) HSAN IV (#256800)

AD AR AR AR

9q22 ? 9q31 1q21-q22

SPTLC1 mutations ? IKBKAP mutations NTRK1 mutations

Miscellaneous HNA (#162100) HMSN and retinitis pigmentosa HMSN VI (with optic atrophy)

AD AD ?

17q24-25 ? ?

? ? ?

*AD ⫽ autosomal dominant; AR ⫽ autosomal recessive; XD ⫽ X-linked dominant; ? ⫽ unknown. † Numbers in parentheses are those of the Online Mendelian Inheritance in Man database.

cause slow NCVs as well as damage axons (see Chapter 73). HMSN III is not used in this classification. This is because DSN, previously classified as HMSN III, is now used independently of whether a gene is expressed in Schwann

cells or neurons, and independently of whether the neuropathies are dominantly or recessively inherited. HMSN IV is used to classify the recessive forms of HMSN, also independently of whether they are demyelinating or axonal

1628

Diseases of the Peripheral Nervous System

pathologically. This is preferred over the prior classification of HMSN IV as Refsum’s disease (phytanic acid excess) given the clear involvement of multiple tissues external to the nervous system in this latter disorder. The most common form of HMSN X, caused by mutations in the gap junction gene GJB1 (see below), is sometimes listed under HMSN I because it has an X-linked dominant inheritance. However, even this genetically based classification system has limitations. For example, certain mutations in the PMP22 gene (e.g., Thr118Met) may cause recessively inherited neuropathies,124 whereas most other PMP22 mutations are inherited as autosomal dominant disorders (reviewed by Shy et al.137). In the end, it may not be possible to “perfectly” classify the various types of HMSN because of the large number of different genetic causes of these neuropathies and the phenotypic variability conveyed by different mutations within the same gene.

Gene Identification Studies HMSN I Historically, the first genetic localization in HMSN I was to chromosome 1, by linkage to the Duffy blood group marker (Fig. 69–2).13–15,115 This form, now designated HMSN IB, is discussed below. Following Bird’s observation that a kindred with slow NCVs linked to chromosome 1, Dyck and colleagues found that their extensively studied kindred did not link to Duffy.54 In a telephone conversation between Bird and Dyck, it was decided that the kindred without linkage to Duffy would be called 1A, whereas the one linked to Duffy would be called 1B. Subsequent studies with chromosome 1 markers in HMSN I pedigrees suggested that the defective gene(s) did not map to chromosome 1 in the majority of kindreds.15,26,53,54,61,77,97,109 Rather, the most common form of HMSN I, HMSN IA, was mapped to the pericentromeric region of chromosome 17,122,156 eventually to 17p11.2-p12.151,155 Finally, in a pair of landmark studies in 1991, HMSN IA was shown to be caused by a duplication on chromosome 17, containing the gene encoding peripheral myelin protein 22 (PMP22).102,123 Subsequently, it was found that the original CMT1A kindred described by Dyck et al.50 had the duplication of 17p11.2. Approximately 60% to 70% of HMSN I patients have this duplication.78,113 There is convincing evidence that the duplication of PMP22 causes HMSN IA: 1. Missense mutations in PMP22 cause the Trembler (Tr) and Trembler J (TrJ) naturally occurring mouse models of HMSN I.141,142 2. The transgenic mice and rats bearing extra copies of PMP22 develop a HMSN IA–like neuropathy.104,135 3. Some patients with missense mutations in PMP22 also develop a phenotype similar to HMSN IA.67,79,112,125,154 The frequency of the duplication event probably results from a homologous repeat sequence flanking the gene (CMT1A-REP).117

FIGURE 69–2 Chromosomes 1, 17, and X and approximate HMSN IA (chromosome 17), HMSN IB (chromosome 1), and HMSN X loci.

Interestingly, a deletion of exactly the same 1.4-Mb region containing PMP22 was subsequently shown to cause most cases of an entirely different disorder, hereditary neuropathy with liability to develop pressure palsies (HNPP).25 Missense mutations in the major PNS myelin protein gene, on chromosome 1, encoding myelin protein zero (MPZ), cause HMSN IB.71 MPZ is a member of the immunoglobulin gene superfamily, has a single transmembrane domain, and is necessary for the adhesion of concentric myelin wraps in the PNS internode. Missense mutations in EGR2 (also called Krox20), on chromosome 10, cause HMSN ID.161 EGR2 is a transcription factor involved in the regulation of as yet unspecified genes in the myelinating Schwann cell. Finally, the genetic cause of HMSN IC has just been shown to be point mutations in the gene encoding the putative protein degradation gene LITAF/SIMPLE. There is limited clinical information on these patients at present. However, patients are known to develop widely accepted criteria for HMSN I, including distal muscle weakness and atrophy, sensory loss, and slow NCV, in the range of 20 to 25 m/s. Onion bulbs of various thickness are also a feature of this disorder.140 HMSN X The second most common form of HMSN, encompassing between 10% and 16% of cases, is caused by missense

Hereditary Motor and Sensory Neuropathies: Overview of Features

mutations in GJB1, located on the X-chromosome, encoding the protein connexin 32 (Cx32).11 Cx32 is localized in the uncompacted myelin of the paranodal loops and Schmidt-Lanterman incisures, where it presumably functions as a gap junction protein permitting the passage of small molecules and ions between adjacent loops of the paranode or incisures, thus enabling a radial diffusion path directly to the axon. Currently, over 250 different mutations of Cx32 have been identified (see molgen-www.uia.ac.be/ CMTMutations/). Because they have a single X chromosome, men tend to develop HMSN X more severely than their female counterparts. Women, probably because of X-inactivation of the abnormal chromosome, usually have milder disease, though most are affected to some degree (see Chapter 76). HMSN II Molecular genetic studies have proven HMSN II to also be a genetically heterogeneous disorder, like HMSN I. At least six subtypes of HMSN II (HMSN IIA, B, C, D, E, and F) have been identified by linkage analysis (Online Mendelian Inheritance in Man database; see www3.ncbi.nlm.gov). HMSN IIA patients have typical CMT clinical presentations with sensorimotor peripheral neuropathies. The locus of CMT2A is at chromosome 1p36. It has recently been reported that mutations in the nuclearly encoded mitochondrial GTPase mitofusin 2 (MFN2) cause most cases of HMSN IIA.167a Mitochondrial morphology and function is determined by a dynamic equilibrium between organelle fusion and fission. In yeast and Drosophila mitochondrial fusion is controlled by the mitochondrial transmembrane GTPase fuzzy onions (Fzo). MFN2 and MFN1 are human homologues of Fzo and are both essential for embryonic development and mitochondrial fusion. The two mitofusins can exist as both homotypic and heterotypic oligomers that either cooperate or act individually to promote mitochondrial fusion. Mfn1 and Mfn2 deficient mitochondria lose the ability to fuse. Perhaps as important for HMSN IIA, these mitochondria also lose their ability to move along microtubules or actin filaments.27a Since mitochondria are transported along microtubules as part of the fast anterograde axonal transport system, it may be that impaired mitochondrial transport deprives the distal axon of a necessary source of energy, contributing to the axonal neuropathy of HMSN IIA. With respect to anterograde axonal transport it is interesting to note that an additional family with HMSN IIA has been reported in a family with mutations in the KIF1B gene, encoding the kinesin motor for anterograde axonal transport.167 HMSN IIB is a predominantly sensory disorder, and there is debate as to whether cases should be considered under pure sensory neuropathies. The disorder has been mapped to chromosome 3q21. Recently, HMSN IIB has been found to be caused by mutations in the small GTPase late endosomal protein RAB7.157

1629

HMSN IIC is a rare disorder in which patients have paresis of diaphragmatic and vocal cords in addition to other characteristics of HMSN II. CMAP amplitude of the diaphragm on phrenic nerve stimulation is reduced or absent. Although vocal cord and diaphragmatic paresis are not unique to HMSN IIC (e.g., HMSN IVA; see below), Klein and colleagues have recently identified linkage to chromosome 12 in families with HMSN IIC,91 suggesting that it is at least a genetically distinct disorder (see also Chapter 73). The large family reported by Klein and colleagues with HMSN IIC has varying degrees of vocal cord weakness or paralysis, intercostal and diaphragm weakness, distal lower limb weakness and atrophy, and sensory loss in severe cases, but pupillary abnormalities and hearing loss were not features.91 HMSN IID is a somewhat confusing disorder because some patients appear to have sensorimotor neuropathies while others have presented with pure motor syndromes characterized as HMN V. At least one family has been described with some individuals having the pure motor syndrome and others also having sensory loss, suggesting that the two disorders are likely to be different phenotypes of the same disease.131 The HMSN IID locus is on chromosome 7p14. The genetic cause of HMSN IVD (HMN V) has recently been identified as mutations in the glycyl transfer RNA (tRNA) synthetase (GARS) gene. The GARS protein is involved in ligating glycine to its cognate tRNAs.5 An additional locus for HMSN II was detected with linkage to chromosome 8p21. Subsequent studies have identified mutations in the neurofilament light (NEFL) gene as the cause of this neuropathy, now known as HMSN IIE.108 Since the NEFL protein is an important constituent of the neurofilaments used in axonal transport systems, and neurofilament phosphorylation is known to be abnormal in demyelinating forms of HMSN, HMSN IIE may provide important clues into mechanisms of axonal damage not only in HMSN II but also in HMSN I (reviewed by Kamholz et al.85). Mutations in the ␣-crystallin domain of small heat shock protein sHSP27 have recently been shown to cause either CMT2F or a distal hereditary motor neuropathy (dHMN).55a Similarly, mutations in sHSP22 were shown to cause distal HMN type II.79a sHSPs constitute a protein superfamily in which members share an approximately 85 amino acid residues C-terminal region known as the ␣-crystallin domain. dHSP22 and dHSP27 have previously been found to interact with each other; therefore both disorders probably involve common, though as yet unknown, mechanisms.10a,140a Mutations in sHSP27 disrupt neurofilament assembly.55a Mutations in the neurofilament light chain gene (NEFL) mutations also cause CMT2 (CMT2E).81a Taken together these results suggest that the sHSPs may play critical roles in regulating or maintaining the axonal cytoskeleton and axonal transport. Upregulation of sHSP27 is necessary for the survival of injured motor or sensory neurons and that sHSP27 is overexpressed in motor neurons of the SOD1 mouse model of familial ALS. Thus

1630

Diseases of the Peripheral Nervous System

sHSP mediated pathways may ultimately prove to play important roles in other neurodegenerative disorders as well. HMSN IV HMSN IV, the autosomal recessively inherited neuropathies, are also a heterogeneous group of disorders. HMSN IV cases are rare, usually more severe than the autosomal dominantly inherited disorders, and many patients may have systemic symptoms, such as cataracts and deafness. HMSN IV is separable into multiple forms depending on the causal gene. HMSN IVA is linked to a 5-cM region of 8q13-q21.1. The disorder was first described in four highly inbred families in Tunisia. Clinical onset began in the first 2 years of life, with delayed developmental milestones such as sitting or walking. Weakness spread to proximal muscles by the end of the first decade of life, and many patients became wheelchair dependent. Sensory loss has been mild, deep tendon reflexes absent, and motor NCVs slowed to an average of 30 m/s in the upper limbs. Pathology studies from sural nerve biopsies revealed a loss of large-diameter myelinated fibers and hypomyelination, but no abnormalities of myelin folding. So-called basal lamina onion bulbs, characterized by concentric layers of basal lamina without intervening regions of Schwann cell cytoplasm, have been described in biopsies.10 Recently, mutations in a gene encoding a novel protein of unknown function, ganglioside-induced differentiation– associated protein-1 (GDAP1), have been shown to cause the recessively inherited HMSN IVA. Depending on the patients, the mutations appear to cause either demyelinating7 or axonal28 neuropathies. Patients with the “axonal” form have also presented with vocal cord paralysis in addition to other manifestations of severe sensorimotor neuropathies. Thus whether mutations in GDAP1 cause neuropathy by disrupting Schwann cell or neuronal function, or a combination of the two, is not yet known.7,28 A likely scenario is that GDAP1 mutations disrupt signaling between Schwann cells and axons. HMSN IVB1 is a recessively inherited disorder characterized clinically by a unique pathologic feature: the presence of focally folded myelin sheaths in nerve biopsy. The genetic locus is on chromosome 11q23, and encodes a gene called myotubularin-related protein-2 (MTMR2).19 How a “dual-specific” phosphatase causes a demyelinating neuropathy is not yet known. Affected patients become symptomatic early, with an average age of onset of 34 months; unlike most forms of CMT, proximal as well as distal weakness is prominent. Motor conduction velocities are severely reduced (typically 14 to 17 m/s) with temporal dispersion, CMAPs are reduced, and sensory nerve action potentials (SNAPs) are frequently absent. Segmental demyelination is also demonstrated in nerve biopsies. Four children in a Turkish inbred family were identified with neuropathies characterized by focally folded myelin, a characteristic of HMSN IVB as cited above. However,

homozygosity mapping demonstrated linkage to a distinct locus on chromosome 11p15, leading to the designation of the locus as HMSN IVB2. Recently, mutations have been identified in a large, novel gene named SET binding factor-2 (SBF2), which lies within the interval on 11p15 and segregates with the neuropathy in affected families, suggesting that the mutations are responsible for the disease. SBF-2 is a member of the pseudophosphatase branch of myotubularins, with striking homology to MTMR-2.134 HMSN IVC is a childhood-onset demyelinating form of HMSN associated with an early-onset scoliosis and a distinct Schwann cell pathology. HMSN IVC is inherited as an autosomal recessive trait and is caused by homozygous or compound heterozygous mutations in the previously uncharacterized KIAA1985 gene.133 Scoliosis is prominent early on and may precede weakness or sensory loss. Mean median motor NCV was 22.6 m/s, and nerve biopsy, when obtained, showed demyelination, onion bulb formation, and, most typically, large cytoplasmic extensions of Schwann cells. The translated protein defines a new protein family of unknown function with putative orthologues in vertebrates. Comparative sequence alignments indicate that members of this protein family contain multiple SH3 and TPR domains that are likely involved in the formation of protein complexes. Kalaydjieva and colleagues have reported a separate disorder with linkage to chromosome 8q24 in a Gypsy population with an autosomal recessive inheritance pattern that is now recognized as HMSN IVD. The neuropathy has presented with distal muscle wasting and weakness, sensory loss, both foot and hand deformities, and the loss of deep tendon reflexes. Disability usually begins in the first decade of life and becomes severe by the fifth decade. Deafness is invariant and usually develops by the third decade. Brainstem auditory evoked responses are markedly abnormal with prolonged interpeak latencies. NCVs are severely reduced in younger patients and unobtainable after 15 years of age. Pathologically, there are decreased numbers of myelinated axons with thinly myelinated large-caliber axons and onion bulbs present.84 A subsequent study has identified mutations in what is termed the “N-myc downstream-regulated gene” (NDRG1) in these patients.83 How the gene abnormality leads to the disorder is unknown. A novel, severe form of recessive CMT has been designated HMSN IVF and defined in a large Lebanese family in which mutations have been found in the periaxin (PRX) gene on chromosome 19. PRX, expressed in Schwann cells, encodes two proteins that contain “PDZ” domains that usually interact with other PDZ domain–bearing proteins in intracellular signal transduction pathways. Binding partners for PRX in Schwann cells have not yet been identified, nor have the signal transduction pathways involving PRX been delineated. NCVs in patients with PRX mutations were markedly slowed, and onion bulbs were present on sural nerve biopsies.18,145

Hereditary Motor and Sensory Neuropathies: Overview of Features

Recently, still another autosomal recessive form of demyelinating CMT was reported, hereditary motor and sensory neuropathy–Russe (HMSNR). Patients develop primarily severe sensory loss, though motor NCVs are moderately reduced (on average, 31.9 m/s for ulnar and median nerves). The locus of HMSNR is positioned to 10q23.2, a small interval telomeric to the EGR2 gene.126

EPIDEMIOLOGY Much of the epidemiologic research on the HMSNs was performed prior to the era of molecular diagnosis. For example, the prevalence of HMSN (all types) was estimated to be 36 per 100,000 in western Norway23 and 23 per 100,000 in Scandanavia139 but only 4.7 per 100,000 in Newcastle-Upon-Tyne.30 In Olmsted and Wabasha counties in southeastern Minnesota, among a population of approximately 150,000 persons, one of us (P.J.D.) has identified approximately 45 affected persons with HMSN I. In earlier studies, Dyck and Lambert identified 72 out of 104 patients with hypertrophic forms and 30 patients with neuronal varieties of inherited neuropathies.48,49 In the hypertrophic neuropathy group were 67 patients with a typical CMT peroneal muscular atrophy syndrome and 5 with early-onset DSN. In the hypertrophic CMT syndrome group the disorder was usually inherited as an autosomal dominant trait, but in approximately one fourth of kindreds the disorder was sporadic or in siblings (the parents having been examined and found to have normal nerve conductions). Syndromes with neuronal varieties included 8 patients with neuronal CMT, 4 patients with distal progressive muscular atrophy, 12 patients with Friedreich’s ataxia, and 8 patients with spastic paraplegia. In their classic review of 227 patients with HMSN, Harding and Thomas found that 173 had type I, and 54 had type II disorder. Autosomal dominant inheritance could be shown in the majority. In some, autosomal recessive inheritance seemed probable because the disorder occurred in siblings and was not found in patients who were cousins.68,69 With the advent of molecular diagnosis, increasing data are becoming available on the frequency among different genetic types of HMSN. The 17p11.2 duplication was identified in 60% to 68% of HMSN I patients in North America78,164 and 70.7% of 819 unrelated CMT1 patients in a European collaborative study.113 HMSN X accounts for approximately 10% to 20% of HMSN cases.78 MPZ mutations account for about 5% of patients with HMSN I phenotypes,113 although slightly lower percentages have also been reported.17 Point mutations in PMP22 or EGR2 are less frequent.17,113 Remarkably, point mutations have often been identified as de novo or sporadic events without clear family histories.17,138 The overall prevalence of HNPP is not known, but 84% of 115 unrelated patients with clinical evidence of HNPP were found to have the chromosome 17p11.2 deletion.113 Fourteen of 63 families with domin-

1631

antly inherited neuropathies were found to have features consistent with CMT2 in the series of Ionasescu et al.78 At present, 33 distinct genes have been identified that cause inherited neuropathies when the gene is mutated; linkage studies suggest that at least 44 genetic causes will ultimately be identified. Although the demyelinating forms have been reported as most common, ultimately the axonal forms may be even more frequent because many are probably underreported as a result of difficulties of ascertainment and other issues. Genes and loci associated with HMSNs are illustrated in Figure 69–3, as depicted on the Inherited Peripheral Neuropathies Mutation Database website (molgenwww.uia.ac.be/CMTMutations/), maintained by investigators at the University of Antwerp in Belgium. This figure will be updated on the website as additional genes are identified that cause inherited neuropathy.

CLINICAL FEATURES The clinical presentations of patients with specific forms of HMSN are discussed in the chapters dealing with these disorders. However, in this section we review some general characteristics of how these inherited neuropathies present.

Overview Many HMSN patients present with a “classic” or typical phenotype. They are slow runners in childhood, develop foot problems in their teenage years, and often require orthotics for ankle support as adults. The development of the neurologic signs is so insidious that it becomes a clinical feature, which is diagnostically useful. Thus in most families the onset of symptoms (e.g., weakness) is often vague or not recalled. Families tell the physicians that the patient has had high arches of the feet, hammer toes, or clumsiness in walking most of his or her life. The fact that they cannot walk on their heels is often not known by the patients or is explained away as, “I never could walk on my heels.” Some patients report that previous physicians never could elicit knee or ankle reflexes. The nature of the symptoms is also instructive. For the most part, affected persons in large kindreds can be shown to have sensory fiber involvement by neurologic examination or low sensory NCVs or amplitudes, but only a small percentage of patients have positive neuropathic sensory symptoms (asleep-numbness, prickling, deep aching, burning, lancinating pain, or the like). A large percentage have negative neuropathic symptoms, recognizing that their feeling may be decreased in the feet (e.g., to recognize the texture of carpet, or thermal or pain sensation). These features are quite different from what has been found in acquired neuropathies.51,55 The frequency of pain in patients belonging to a large disease support group appears to be much higher (Linda Crabtree survey, CMT International),

1632

Diseases of the Peripheral Nervous System

1 K/F1B

2

3

5

7

8

10

dHMN V

CMT2A

NEFL

CMT2E

9

CMT2D

dHMN-J CMT2F

HMSN-P dHMN VII RAB7

MPZ LMNA

NDRG1

HMSN-L

CH, CMT1, DSS

EGR2

HMSN-R DI-CMTA

DSS

CMT2B

CMT4

HSAN IV HSAN V CMT1B DSS CH “CMT2” ARCMT2A

NTRK1

GDAP1

CMT4A CMT4C2

SPTLC1

HSN I

NBKAP

HSN III ALS4/dHMN

X

22

15 12

11 SBF2/ MTMR13

17

16

CMT4B2

AS L/TAP/ SIMPLE

PMP22 CMT1C

19

18

DI-CMTB

PRX IGHMBP2

Con. distal SMA MTMR2

CMT4B1

dMHN II

G/B1

CMT1X

CMT4F/DSS

CMT2X

ARCMT2B

HNA

dHMN VI

CMT1 +PM +WHS

SOX10

CMT1A HNPP DSS

CCFDN GAN

GAN

FIGURE 69–3 Inherited peripheral neuropathies. (Courtesy of Inherited Peripheral Neuropathies Mutation Database, University of Antwerp, Antwerp, Belgium.)

but selection bias and other factors may account for this discrepancy. Variable degrees of hand weakness occur, typically lagging about 10 years behind the development of foot weakness. Sensory loss, also variable, occurs in both large (vibration and proprioception) and small (pain and temperature) modalities. While the combination of weak ankles and decreased proprioception often leads to problems with balance, the vast majority of patients remain ambulatory throughout life, which is not shortened by their disease. One may be able to palpate the enlarged nerve trunks in subcutaneous tissue. Additional features, including postural tremor (referred as Roussy-Lévy syndrome) and muscle cramps, may also occur. While this phenotype is typical for HMSN IA patients, it is not invariable. Reflexes are often diffusely absent in HMSN I patients, but a considerable number of HMSN II patients have normal patellar reflexes. In general it is difficult to distinguish between HMSN I and HMSN II patients without resorting to neurophysiology. Some examples of clinical features of HMSN I and HMSN II are shown in Tables 69–3, 69–4, and 69–5,

Table 69–3. Hereditary Motor and Sensory Neuropathy: Incidence of Weakness No of cases (%) Type I Upper limbs Total Dominant Sporadic/recessive Lower limbs Total Dominant Sporadic/recessive Severe lower limb weakness Total Dominant Sporadic/recessive

Type II

116 (67) 89 (64) n.s. 27 (79)

28 (51) P ⬍ 0.05 18 (49) n.s. 10 (55)

150 (87) 118 (85) n.s. 32 (94)

52 (94) n.s. 34 (92) n.s. 18 (100)

} }

59 (34) 40 (29) P ⬍ 0.01 19 (56)

}

}

}

18 (33) 12 (32) 6 (33)

n.s.

}

n.s.

From Harding, A. E., and Thomas, P. K.: The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103:259, 1980, with permission.

Hereditary Motor and Sensory Neuropathies: Overview of Features

Table 69–4. Hereditary Motor and Sensory Neuropathy: Incidence of Sensory Loss No. of cases (%)

Light touch Pain Joint position sense Vibration sense

Type I

Type II

89 (53) 89 (53) 78 (46) 117 (69)

20 (36) 17 (31) 14 (25) 31 (56)

P ⬍ 0.05 P ⬍ 0.05 P ⬍ 0.01 n.s.

From Harding, A. E., and Thomas, P. K.: The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103:259, 1980, with permission.

taken from the landmark studies of Harding and Thomas in 1980.68,69 Not all patients, however, present with this classic phenotype. Clinically, DSN or CH patients have delayed motor development in infancy and motor development remains abnormal. Many of the features identified by Dyck and Lambert in 1966 are characteristic of these disorders. In evaluating four probands from different kindreds, these investigators found that walking did not begin until 15, 33, 36, and 48 months of age.47 During childhood, walking was severely impaired because of generalized limb and truncal weakness, and running, skipping, and jumping were not possible. The four probands identified the age at which motor performance had been best as 8, 8, 12, and 15 years, respectively. The two oldest patients with the disorder have been confined to wheelchairs since the ages of 31 and 19 years. The weakness was not confined to lower limbs. Increasingly during late childhood and the early teens, it became difficult to perform fine manipulative acts with the fingers (e.g., turning keys in locks, unscrewing lids from jars, and grasping coins or writing instruments). Dyck and Lambert identified weakness in all limb muscles, though distal muscles were more severely affected. Atrophy, though present in all patients, was not always detectable because of the presence of subcutaneous fat. The patients were usually areflexic. Sensory loss was definite over the distal aspects of the limbs, and it affected touch-pressure, joint position, and

Table 69–5. Hereditary Motor and Sensory Neuropathy: Incidence of Tendon Areflexia No. of cases (%)

Upper limbs Knee jerk Ankle jerk

Type I

Type II

101 (58) 67 (70) 154 (89)

5 (9) 10 (18) 44 (80)

P ⬍ 0.02 P ⬍ 0.02 n.s.

From Harding, A. E., and Thomas, P. K.: The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103:259, 1980, with permission.

1633

vibration sensation without affecting pain sensation and temperature discrimination. Dyck and Lambert’s patients had sensory ataxia and some truncal ataxia on sitting. Some patients had choreiform movements of the fingers of outstretched hands. Peripheral nerves were enlarged; the greater auricular nerve, cervical plexus, and median and ulnar nerves were found to be particularly suitable for confirming nerve enlargement. The CSF protein content in Dyck and Lambert’s patients ranged from 76 to 180 mg/dL; the lowest value was found for the patient with the least involvement. Miosis, sluggish response of pupils to light, and nystagmus have been described in some but not all patients with severe, early-onset neuropathies. Occasional patients have had sensorimotor hearing loss. Many patients with DSN or CH are of small stature. Finally, patients with HMSN may present later in life as adults. Some may develop only minimal disability throughout life. Others, however, may present as adults but progress to the point that they become wheelchair bound (see Chapter 71 for examples). In general, only a modest percentage of patients with HMSN seek the help of a physician or neurologist for neuropathy or musculoskeletal symptoms, which suggests that many have relatively mild symptoms. Alternatively, they may have become accustomed to their symptoms, have decided that help is not available, or have attributed it to other diseases or injuries for which treatment is not available. In the experience at the Mayo Clinic, less than 10% of HMSN I patients have sought medical attention for symptoms related to HMSN. When patients with HMSN come to physicians, they present in four ways: (1) with symptoms of muscle weakness or atrophy; (2) with symptoms related to structural foot abnormalities (Fig. 69–4) or soft tissue complications (e.g., calluses, ulcers, cellulitis, or lymphangitis); (3) with incidental discovery or a telltale indicator of neuropathy (e.g., abnormalities in electrodiagnostic studies); and (4) with findings from evaluation of family members. Sensory or autonomic symptoms are seldom the presenting complaint. Parents may report that children have weakness of ankles, stumble, run abnormally, or do not pick up their feet. A second common reason for evaluation is abnormality of the feet, such as pes cavus, pes equinovarus, hammer toes, symptoms related to painful calluses, or neuropathic pain. Infrequently, patients may develop mutilating neuropathy or plantar ulcers (or ulcers over the convexity of hammer toes). A physician may spot characteristic signs of the disorder when seeing the patient for perhaps unrelated symptoms. Thus a patient seen for back pain, sciatica, carpal tunnel syndrome, or various other disorders may have nerve conduction studies done to confirm the focal problem, only to have the electromyographer report the more generalized nerve conduction abnormality typical of HMSN. Loss of sensation is seldom an initial

1634

Diseases of the Peripheral Nervous System

FIGURE 69–4 Feet of patients with HMSN type I. A, Feet of teenage son (left) and his mother (right) show that pes cavus is not necessarily a feature of the disease. The mother had unequivocally abnormal nerve conduction velocity values and histologic abnormalities characteristic of the disorder revealed by sural nerve biopsy. B, Typical feet of patient with slight degree of hammer toes and high instep. C, More developed form of pes cavus and hammer toes; note absence of muscle atrophy.

symptom. Prickling, a feeling of “tightly wrapped” asleepnumbness (like a hand or foot “gone asleep”), is not an infrequent symptom in HNPP, but is otherwise not a typical presentation of HMSN. When patients report these sensory symptoms, a concomitant disorder such as carpal tunnel

syndrome or an acquired rather than an inherited neuropathy should be considered. A small percentage of patients may report painful calluses or “burning” or lancinating pain. Autonomic symptoms are usually not reported by patients, although autonomic dysfunction may be present on testing.

Hereditary Motor and Sensory Neuropathies: Overview of Features

Muscle Weakness Muscle weakness is usually distal and expressed as an abnormality of gait, or clumsiness in running. Typically, there is some increased instability at the ankle, a tendency to varus deformity of the foot, and steppage gait. In steppage gait, the knees have to be raised higher than normal to lift the feet off the ground. Muscle weakness and atrophy typically begin insidiously in the foot and leg muscles, and especially affect intrinsic foot and peroneal muscles. Patients often have extreme difficulty walking on their heels as a result of dorsiflexion weakness, which manifests itself frequently by catching the toes on the curb or carpet while walking. Difficulty toe walking is also common. However, idiopathic toe walking of childhood can be an early presentation of HMSN. Later, calf and intrinsic hand muscles become affected. Occasionally, in more severe cases, proximal thigh muscles may become affected. The atrophy tends to affect the distal part of the gastrocnemius, soleus, and distal quadriceps muscles, leaving only a small mass of muscle at the proximal end. The appearance of “inverted champagne bottle” lower limbs is actually not typical of most patients with this type of HMSN (see Fig. 69–4). This appearance is more characteristic of thin patients with fully expressed disease, and also in some cases of HMSN II. In patients with considerable subcutaneous fat, the degree and distribution of muscle atrophy may not be readily appreciated. If “inverted champagne bottle” legs were the criterion for diagnosis, most patients with HMSN would not be detected. Cramps, especially after exercise, and fasciculations in affected muscle are common. Muscle stretch reflexes disappear first in the gastrocnemius and soleus, then in the quadriceps femoris, and finally in the upper limb muscles. Reflex loss is more frequent in HMSN I, as mentioned above. The Babinski toe sign is not present.

Sensory Dysfunction Sensory symptoms of HMSN can be separated into those that involve small, thinly myelinated or unmyelinated fibers subserving pain and temperature, and those affecting large myelinated fibers subserving position sense. Common symptoms of small fiber sensory neuropathy include “feeling like my feet are walking on pebbles,” “my feet feel like ice all the time,” or “difficulties telling whether bath water is hot or cold with my feet.” As mentioned previously, painful dysesthesias such as feeling like one’s feet are on fire, or on hot coals, or being stuck with pins are infrequent complaints, although they do occur. Sensory symptoms occur in the hands but less frequently, probably because most neuropathies are length dependent. A general rule is that sensory symptoms in the hands begin at about the time sensory symptoms in the lower extremities have progressed up to the knee. One exception to this rule is that patients may develop concomitant carpal tunnel syndromes, in

1635

which case dysesthesias in the hands can be prominent and even wake them from sleep. Large fiber sensory loss usually causes patients to develop difficulties with balance. When presenting in children, for example, large fiber sensory loss may cause difficulties walking on a balance beam at school, or walking across a log over a stream. In patients of any age, difficulties with balance may be worse in a crowd, at night, or in a darkened movie theater, when vision cannot overcome proprioception loss. Because loss of proprioception is also frequently length dependent, patients may be able to stabilize themselves by simply lightly touching a wall with their hand or elbow since nerves from the arm can then provide proprioceptive input to the brain. On neurologic examination, sensory loss is usually found in a stocking-glove distribution for both large and small fiber modalities. Cold, erythematous or bluish discolored feet suggest a loss of small fiber function, although other factors may also contribute to this (see below). Large fiber sensory loss, or “sensory ataxia,” in the upper extremities can often be detected by the inability of the patient to accurately locate the thumb with the opposite index finger while the eyes are closed, or by the presence of a characteristic irregular tremor (pseudoathetosis) of the fingers (see below). The sensory examination should include vibration, position, and light touch as well as pain and temperature. A determination of the degree and the extent of sensory loss as well as the pattern of the deficits (symmetrical or asymmetrical; distal or generalized; focal, multifocal, or diffuse) in both upper and lower extremities is important because some forms of HMSN are more likely than others to present asymmetrically. Quantitative sensation examination employing good testing techniques and adequate controls provides unequivocal evidence of abnormalities affecting both largeand small-diameter fibers. Because of the high arch and weakness of intrinsic foot and peroneal muscles, the foot is often foreshortened and internally deviated, and toes are curled into a clawed deformity. Calluses often form over the metatarsal heads and over the tips and convexities of the toes. Occasionally ulcers develop over these pressure points. Common sites are the head of the metatarsal bones, convexity of curled toes, or other points of repeated pressure contact. Only occasionally do Charcot joints develop. The multiple factors that underlie mutilating neuropathy are described in Chapter 78. Generally, the causes are insensitivity, excessive mechanical pressure in a region, previous injury, improper care, and indifference or neglect of injury.

Autonomic Dysfunction Symptoms or signs of autonomic dysfunction are infrequent in most cases of HMSN. Thus most patients do not develop urinary retention or incontinence, impotence, constipation alternating with diarrhea, or orthostatic symptoms as a result of HMSN. Patients with HSAN are obviously an

1636

Diseases of the Peripheral Nervous System

exception. It is unusual for patients with typical HMSN to complain of abnormalities in sweating. However, with sophisticated studies of sweating, abnormalities may be found. Although miosis and a poor light response have been reported in HMSN,65 this was not a feature of 67 patients observed in the Mayo Clinic.48 However, controlled studies using the pupillograph were not done in this series. The presence of cold feet has been discussed above with respect to small fiber sensory loss. A decrease in skin and muscle temperatures of the foot or distal leg of 5° to 10° C is not uncommon, particularly later in the course of the neuropathy. Whether adrenergic nerve dysfunction contributes to this temperature decrease is not known, though it has been suggested in at least one study.22 Alternatively, the temperature decrease may be due, at least in part, to diminished muscle bulk and the consequent decrease in the flow of blood.

Bone and Joint Abnormalities Since the neuropathy is often asymmetrical around joints, muscle imbalances develop between agonist and antagonist muscle groups (e.g., ankle dorsi- vs. plantar flexors, ankle invertors vs. evertors). In this sense, the development of pes cavus and hammer toes is probably analogous to the development of the claw hand in ulnar nerve lesions. In our experience, only approximately one fourth of affected persons in the first decade of life had high arches; in these persons in the later decades, about two thirds had high arches. These muscle imbalances in the foot lead to contractures (e.g., ankle plantar flexion) and/or additional foot deformities (forefoot and hindfoot varus or valgus, forefoot supination or pronation). When HMSN patients experience foot pain, it is often best explained by these foot and ankle changes and the consequent increased stress placed on the articular cartilage of supporting joints and soft tissue structures (i.e., ligaments, joint capsules, and tendons). The misalignment produced by foot deformities often serves to decrease the base of support provided by the feet and secondarily decrease the efficiency of ankle muscles (which are often already weakened from the neuropathy) by altering the resting length of the muscle belly and anatomic line of pull of tendons. Functionally this can lead to decreased balance during standing and walking as well as gait deviations. The gait deviations often require the individual to expend increased energy, with consequent fatigue during ambulation. Neuropathic proprioceptive sensory impairment from the foot and ankle joints can further accentuate balance and gait problems.

FIGURE 69–5 Enlarged great auricular nerve, not infrequently seen in HMSN type I. (From Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. I. Neurologic, genetic and electrophysiologic findings in hereditary polyneuropathies. Arch. Neurol. 18:603, 1968. Copyright 1968, American Medical Association. Reprinted by permission.)

cal plexus.148 Patients with CMT were reported to have hard, thick nerves, whereas patients with Déjérine-Sottas disease had thick, succulent nerves. In a prospective series from the Mayo Clinic, one fourth of 67 HMSN I patients had clinical enlargements or excessive firmness of nerve. The nerves between the shoulder and elbow were the most suitable for such an assessment; entrapment points such as the ulnar groove are not good sites for this determination because the nerve is normally thickened at this site and contains variable amounts of epineurium and perineurium. In typical cases, we have found an increased transverse fascicular area of nerves and prominent onion bulb formation.

Hypertrophic Nerves

Tremor

Enlargement of peripheral nerves, as shown in Figure 69–5, was usually observed at autopsy, although some authors observed this in living persons.22,107 Thevenard and Berdet reported a difference in the hypertrophied superficial cervi-

Tremor, especially of the hands, may accompany HMSN. In some cases the tremors can be quite severe. Roussy and Lévy,29,127–129 as well as Marie105 and Boveri,21 described a kinship of HMSN I and tremor with the features typical of

Hereditary Motor and Sensory Neuropathies: Overview of Features

essential tremor. This family has recently been found to have an MPZ mutation120 (see Chapter 71). Subsequent studies have shown that similar tremors are not specific for particular mutations, and these tremors have been reported in patients with chromosome 17p12 duplications as well as in patients with PMP22 point mutations.150

NERVE CONDUCTION AND ELECTROMYOGRAPHIC FEATURES Because NCV studies are easily performed and provide important information on whether neuropathies are primarily demyelinating or axonal processes, they are frequently used alone in classifying HMSN as demyelinating or axonal, particularly since sural nerve biopsies are at least relatively invasive techniques. Lambert and colleagues first reported low conduction velocities of peripheral nerves in some families with inherited neuropathies.16,93–95 In a subsequent evaluation of 103 family members from a large kinship with HMSN I, Dyck and colleagues found that all patients with clinical evidence of neuropathy had slow NCVs (Figs. 69–6 and 69–7).50 In two family members, the slow NCVs led to their being diagnosed.50 Similar results were reported in HMSN I families by Amick and Lemmi3 and Myrianthopoulos and colleagues.111

FIGURE 69–6 Measurement of conduction velocity of ulnar nerve (forearm) in one member of a family having normal conduction velocity (case IV.24) and in another member having low conduction velocity (case IV.22). In each, maximal electrical stimulus was applied to the ulnar nerve at the elbow and then wrist. The moments of stimulation and response of the hypothenar muscle as detected by electrodes on skin surface (R) were displayed on a cathode ray oscilloscope and photographed; the moment of stimulation is indicated by sharp break in the baseline (shock artifact), which is followed by a large action potential of the hypothenar muscles. Latency of response in case IV.24 was 0.0062 second when the stimulus was applied at the elbow and 0.0023 second when applied at the wrist; in case IV.22 comparable figures were 0.0191 and 0.0086 second. Conduction velocity between the two points in case IV.24 was 65 m/s and in case IV.22, 21 m/s. (From Dyck, P. J., Lambert, E. H., and Mulder, D. W.: Charcot-Marie-Tooth disease: nerve conduction and clinical studies of a large kinship. Neurology [Minneap.] 13:1, 1963, with permission.)

1637

In reporting 67 patients from 21 HMSN I families, Dyck and Lambert confirmed that slow NCVs were the hallmark of the disorder.48 The conduction velocities of the motor fibers of ulnar, median, and peroneal nerves of affected persons were, on the average, less than half the values in unaffected persons (see Fig. 69–7). In an equally important observation, the authors noted that the mean amplitude of the evoked CMAPs was also reduced by half in HMSN I patients. Occasionally, a patient in direct lineage with an affected patient and affected children, although without clinical signs, was shown to have abnormally slow NCV. Nerve biopsies in these patients provided unequivocal evidence that they had the neuropathy. It was hypothesized that small children from affected kinships that had low conduction velocities but no neurologic signs would develop clinical stigmata of the disorder as they got older. Sensory nerve conduction velocities were affected also in these 21 kindreds. A digital nerve action potential usually could not be detected by procedures that are uniformly successful in healthy persons. If an action potential was detected, it had a low amplitude and a long latency. Study of the compound action potential of the sural nerve in vitro has demonstrated a low conduction velocity and decreased action potential of alpha and delta fibers without comparable changes in the action potential of C fibers. Median, ulnar, and sural SNAPs were often undetectable in patients,

1638

Diseases of the Peripheral Nervous System

FIGURE 69–7 Conduction velocity of motor fibers of ulnar nerve of unaffected (solid symbols) and affected (open symbols) persons from kinships with dominantly inherited peroneal muscular atrophy with low conduction velocities of peripheral nerves. The dashed line represents lower limit of normal. (From Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. I. Neurologic, genetic and electrophysiologic findings in hereditary polyneuropathies. Arch. Neurol. 18:603, 1968. Copyright 1968, American Medical Association. Reprinted by permission.)

and when detectable, the latencies were prolonged and the amplitudes reduced. What are the electrodiagnostic features that separate HMSN I and HMSN II? In their review of peroneal muscular atrophy, Dyck and Lambert found that marked reduction of the conduction velocity of the ulnar nerve was characteristic of many kindreds, whereas in others it was at the lower limits of normal or only slightly reduced.48,49 This separation of HMSN cases into those with a slow conduction velocity and a variety with little abnormality was supported by other investigators.23,24,149 Davis and co-workers argued for the addition of an intermediate group: they classified NCVs of less than 25 m/s as the hypertrophic neuropathy group, NCVs between 25 and 45 m/s as the intermediate group, and NCVs greater than 45 m/s as the neuronal group of patients.30 This classification seemed problematic because it sometimes appeared to separate affected persons from the same kindred, having the same gene abnormality, into those with the hypertrophic neuropathy and those with the intermediate variety of HMSN—a separation without biologic meaning. However, with the advent of molecular diagnosis, the identification of intermediately slowed NCV has proved useful in identifying patients who are likely candidates for having particular genetic forms of HMSN, such as HMSN X100,114 (see below). In their landmark studies in 1980, Harding and Thomas68 found that a bimodal distribution of conduction velocities occurred for both median (peaks at approximately 18 and 55 m/s) and peroneal (peaks at approximately 12 and 47 m/s) nerves. Based on these findings, these authors set the criterion for HMSN I as a conduction velocity less than 38 m/s.

Using this criterion, most of their cases were correctly subdivided into types I and II. We like the 38 m/s criterion but would add two caveats: (1) that mean values of the motor conduction velocity of ulnar nerves of all affected persons in the kindred be used, and (2) that only values be used in which the CMAP is at least 0.5 mV. When conduction block and dispersion of the CMAP or SNAP is found, a diagnosis other than HMSN I should be considered, such as chronic inflammatory demyelinating polyradiculopathy (CIDP). In the early 1980s, Lewis and Sumner demonstrated that most cases of inherited neuropathies had uniformly slow NCVs, whereas acquired demyelinating neuropathies, such as CIDP, had asymmetrical slowing.99 Thus NCVs could be used, along with a patient’s pedigree, to distinguish between inherited and acquired neuropathies. Over the past decade, however, this approach has had to be qualified. Most HMSN I patients, particularly those with HMSN IA, do have uniformly slow NCVs of about 20 m/s, although values as high as 42 m/s have been reported and have been used as a cutoff value.82 However, asymmetrical slowing is characteristic of HNPP, and may be found in patients with missense mutations in PMP22, MPZ, EGR2, and GJ␤1 (reviewed by Lewis et al.100). Since all of these disorders may present without a clear family history of neuropathy, one must now be cautious in using NCV to distinguish acquired from inherited demyelinating neuropathies. Forms of inherited neuropathies associated with uniformly and nonuniformly slowed NCV are illustrated in Table 69–6.100 The use of NCV to distinguish between demyelinating and axonal neuropathies is also important. Virtually all forms of HMSN I have axonal loss as well as demyelination, and

Hereditary Motor and Sensory Neuropathies: Overview of Features

Table 69–6. Electrophysiologic Findings of Inherited Demyelinating Neuropathies Inherited Disorders with Uniform Conduction Slowing HMSN IA HMSN IB Metachromatic leukodystrophy Cockayne’s syndrome Krabbe’s disease Inherited Disorders with Multifocal Conduction Slowing Hereditary neuropathy with liability to pressure palsies HMSN X HMSN IB Adrenomyeloneuropathy Pelizaeus-Merzbacher disease with proteolipid protein null mutation Refsum’s disease Inherited Disorders with Electrophysiology That Has Not Been Characterized PMP22 point mutations Adult-onset leukodystrophies Merosin deficiency Early growth response element 2 (EGR2) gene mutations Myotubularin-related protein 2 mutations Neurofilament light mutations LITAF/SIMPLE mutations Ganglioside–induced differentiation–associated protein 1 From Lewis, R. A., Sumner, A. J., and Shy, M. E.: Electrophysiological features of inherited demyelinating neuropathies: a reappraisal in the era of molecular diagnosis. Muscle Nerve 23:1472, 2000, with permisssion.

it is likely that the axonal loss correlates better than the demyelination with the patient’s actual disability (reviewed in Kamholz et al.85 and Krajewski et al.92). Thus reductions in CMAP and SNAP amplitudes are found in most HMSN I patients; in one series of 43 HMSN IA patients, 34 had unobtainable peroneal CMAPs and 41 had unobtainable sural SNAPs.92 The distinction between demyelinating and axonal features of NCV is particularly confusing in HMSN X and HMSN IB. NCVs in HMSN X patients are “intermediate,” as discussed above, and therefore are faster than in most patients with HMSN I, often with prominent reductions in CMAP and SNAP amplitudes. Thus HMSN X has even been described as an “axonal” neuropathy. However, a careful analysis of the conductions will reveal the primary demyelinating features of the neuropathy. The conduction velocities in men are not normal, but usually between 30 and 40 m/s, values that would be considered an intermediate range between HMSN I and HMSN II. Moreover, distal motor latencies and F-wave latencies are usually prolonged.98,114 Some women with HMSN X, probably through inactivation of their mutant X chromosome, have normal NCVs, though many have values similar to their male counterparts. In distinguishing between the demyelinating and axonal features

1639

of HMSN X, it is important to remember that the disease is caused by mutations in the Cx32 protein, which is expressed in the myelinating Schwann cell. Similar issues occur in some patients with mutations in the MPZ gene (see Chapter 71). For example, the Thr124Met mutation (also numbered Thr95Met) often has NCV suggestive of an axonal neuropathy, and it is only through careful evaluation of the studies that evidence of the primary demyelinating features can be detected33 (also see Chapter 71).

PATHOLOGIC FEATURES Increasingly nerve biopsies are not considered in the management of the inherited neuropathies, which may be appropriate for many cases since molecular biologic testing is less invasive and may be definitive in diagnosis. However, the role of nervous system tissue in affected individuals will gain increasing importance in further understanding disease pathogenesis. Specifically, techniques are now available to identify complex molecular pathways and molecules that probably play a role in the disability caused by specific mutations. It is only by evaluating nerve tissue that these pathways and molecules are likely to be identifiable. Clearly, much of what we understand about the pathogenesis of the inherited neuropathies comes from pathologic studies over the years, in particular from sural nerve biopsies. Some of these results are reviewed below.

HMSN I Pathologic abnormalities in peripheral nerve have long been recognized in the hypertrophic or demyelinating inherited neuropathies. Friedreich57 and Virchow158 noted that peripheral nerves from patients contained fewer myelinated fibers, that intramuscular nerves were surrounded by a rich connective tissue and a hyperplastic neurilemma, and that segments of myelin were atrophic along nerve fibers. Hoffman73 noted that the degeneration of myelinated fibers in patients resembled that which Gombault59,60 had produced in guinea pigs by the use of lead, and which was later called segmental demyelination. In studies of intramuscular nerves, Friedreich57 and Marinesco106 noted that there was concentric hypertrophy of the lamellar sheaths. Marinesco’s patient,106 while one of those originally observed by Charcot and Marie,27 was somewhat atypical in that she had features suggestive of a HSAN. For example, she had lancinating pains in the thighs, knees, and legs, and appreciable loss of pain sensation. In addition to posterior column demyelination and decrease in number and atrophy of anterior horn cells, there was atrophy of Lissauer’s zone. The fibers making up this zone are small ones. Extensive morphometric studies have been made on nerves (especially sural nerves) taken at biopsy and at

1640

Diseases of the Peripheral Nervous System

autopsy.40,41,43,45,63 In comparison with healthy nerves, there has been an increase in transverse fascicular area, which was demonstrated especially for the greater auricular and sural nerves. The numbers of myelinated fibers (expressed per nerve or per square millimeter) were decreased by a variable extent, depending on the severity of the disease. The frequency distribution of the diameters of fibers usually showed a first peak and sometimes a second peak to be present. However, the second peak was usually found at smaller size diameters than in control nerves (Fig. 69–8). Onion

bulb formations were identified under phase microscopy and electron microscopy. These typically occurred around myelinated internodes, demyelinated internodes, or former sites of myelinated fibers (Figs. 69–9 and 69–10).38 The onion bulbs were made up of circumferentially directed Schwann cells and their processes. The Schwann cell nature of the onion bulbs was identified because of the presence of basement membranes. Unmyelinated fibers were also identified that were believed to be nerve sprouts. The outer layer sometimes has the appearances of a fibroblast or

FIGURE 69–8 Spectra of myelinated and unmyelinated fibers of healthy sural nerve of a normal 19-year-old man (A), of sural nerve of a 22-year-old man with HMSN type I (B), and of sural nerve of a 21-year-old woman with probable DSN (C). Note decrease in numbers and alteration in spectrum of myelinated fibers of nerve from patients. Number and spectrum of unmyelinated fibers of sural nerve of patient with HMSN I are normal. Number of unmyelinated fibers in nerve of patient with probable DSN is slightly above normal and peak is somewhat displaced to smaller size categories. (From Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in Déjérine-Sottas disease. Mayo Clin. Proc. 45:286, 1970, with permission.)

Hereditary Motor and Sensory Neuropathies: Overview of Features

1641

FIGURE 69–9 Fascicular biopsy of sural nerve of a 20-year-old man with HMSN type I (⫻2500). Transverse section fixed in 3% glutaraldehyde and 2% osmium tetroxide, embedded in Epon, sectioned at 1.5 ␮m, and stained with methylene blue. Note prominent onion bulb formations and increase in endoneurium. (From Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased-fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968, with permission.)

macrophage. The lamellae were separated by longitudinally directed collagen fibers. Nuclei were increased in number. In teased sural nerve fibers from HMSN I patients, most of the myelinated fibers could be shown to be abnormal (Figs. 69–11 and 69–12). The mean internode length, in relation to the diameter of the fiber, was much less than in normal nerves. In a typical case, there was increased variability of internode length (four times) and of internode diameter (two times) compared to control nerves (Table 69–7). Paranodal and segmental demyelination and remyelination have been demonstrated.43,54,63 Counts of unmyelinated fibers (probably including nerve sprouts) were normal. Although C potentials of excised sural nerves appeared normal in the patient studied at the Mayo Clinic, an abnormality of unmyelinated fibers is assumed from the observed abnormality of thermal sensations and nociception and abnormal sudomotor function in many cases. Pathologic alterations in the spinal cord have also been found at autopsy in HMSN I. For example, posterior column demyelination (defined by a loss of myelinated fibers) and sclerosis have been observed since the original studies by Friedreich.57 He observed that the breakdown of myelinated fibers could be determined by the swollen varicosities and the granular particles of myelin in the posterior columns. The sclerosis and reduction in number of myelinated fibers were greatest in the fasciculus gracilis in upper spinal cord

regions and were less toward the caudal region of the cord. Atrophy of ventral horn perikarya was also noted by Friedreich,57 as well as by Marinesco.106

Déjérine-Sottas Neuropathy As mentioned previously, pathologic changes in nerves from patients diagnosed as having DSN need to be interpreted with the knowledge that DSN represents a heterogeneous group of disorders, not a single entity. Also, as mentioned above, it is unclear whether patients with CH, as described by Lyon,103 Kennedy and co-workers,87,88, and Anderson,4 have neuropathies distinct from those of patients with DSS. Déjérine and Sottas stressed the changes in the peripheral nerves of their patient.35 Nerves were grossly enlarged, more so near their distal ends. Fibers were surrounded by a periaxonal mass composed of abundant connective tissue, and in it lay nuclei. These authors believed that the Schwann sheath, the myelin sheath, and the axis cylinder played a passive role, disappearing because of the compression. Few nerve fibers had a myelin sheath. Wallerian degeneration was not seen. Myelin became increasingly thin, leaving a bare axis cylinder. In a later paper, Déjérine and Thomas noted that the demyelination state of the fibers might indicate that they were regenerating or that axons had lost their myelin sheaths.36 Gombault and Mallet’s early histologic

1642

Diseases of the Peripheral Nervous System

FIGURE 69–10 Transverse sections of consecutive portions along length of singly embedded teased fiber from a 2-yearold patient with HMSN type I. A well-developed onion bulb formation surrounds region of segmental demyelination (A) and remyelination (B). Insets show absence of myelin in former and small number of lamellae in latter. (From Dyck, P. J., and Lais, A. C.: Electron microscopy of teased nerve fibers: method permitting examination of repeating structures of same fiber. Brain Res. 23:418, 1970, with permission.)

studies attributed the disorder to tabes dorsalis.60 They emphasized the involvement of nerves, their enlargement, and their greater distal involvement. They noted abnormalities of Schwann cells that resulted in lamellae that resembled layers of an onion. Connective tissue was not much increased. The disappearance of the myelin sheath was discontinuous, periaxial, and segmental. Sclerosis and loss of myelin were observed in posterior columns. Gombault and Mallet also reported atrophy of the gray substance of the spinal cord.60 In nerve biopsies from Déjérine-Sottas patients, Dyck and colleagues have shown that the transverse fascicular

area is greater than it is in healthy nerves; that in severely affected nerves the frequency distribution of myelinated fibers is markedly abnormal (diameter of the largest fibers is only 4 ␮m, whereas that of healthy nerve [Fig. 69–13] ranges from 12 to 14 ␮m); and that there is an abnormally high ratio of axis-cylinder diameter to total fiber diameter (0.85) (Fig. 69–14).38,43 By means of electron microscopy, the nature of the lamellae of onion bulbs, with interposed collagen fibrils and segmental demyelination and remyelination, was demonstrated. As in HMSN I, the measured conduction velocities were lower than anticipated from the relationship between conduction velocity and diameter

Hereditary Motor and Sensory Neuropathies: Overview of Features

1643

FIGURE 69–11 Teased fibers from sural nerve of a 22-year-old man with HMSN type I (⫻260). Top, Contiguous portions of teased myelinated fiber, showing demyelinated segment in strips B and C. Bleb formation is seen in strip D. Longitudinal profile of onion bulb formation, with large number of longitudinally directed nuclei, is seen in strips B and C. Middle, Contiguous portions of teased myelinated fiber, showing demyelinated segments and marked dilatation or bleb formation. Bottom, Contiguous portions of myelinated fiber. In strip B, myelin sheath of portion of fiber (between arrows) is undergoing fine granular disintegration. (From Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in Déjérine-Sottas disease. Mayo Clin. Proc. 45:286, 1970, with permission.)

in control sural nerves (Fig. 69–15).38 Studies of teased fibers with longitudinal sections (Fig. 69–16) have shown that all myelinated fibers from these highly abnormal nerves contain long regions devoid of myelin (Figs. 69–17 and 69–18). Probably these demyelinated regions exceed in length the regions covered with myelin.41,42,52 Increases in epineurium and in the thickness of the perineurium were reported by Yokomori.165 Counts of the number of unmyelinated fibers in DSN (probably including demyelinated regions and nerve sprouts) have been within the

range of normal values. In comparison with normal nerves, the peak was somewhat displaced to smaller size categories.

HMSN II The pathologic changes in HMSN II have not been studied as completely as have those of HMSN I. In transverse sections of sural nerve viewed under phase and electron microscopes, small onion bulbs are occasionally seen. The

1644

Diseases of the Peripheral Nervous System

FIGURE 69–12 Teased fibers from sural nerve of a 22-year-old man with HMSN type I (⫻360). Top, Contiguous portions of myelinated fiber show only a few fragmentary internodes of myelin; more than half the length of this teased fiber is demyelinated. Bottom, Contiguous portions of myelinated fiber, showing short internode of myelin. Subsequent seven strips (not shown) did not show any internodes. In this teased fiber, 94.6% of the length is demyelinated. (From Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in Déjérine-Sottas disease. Mayo Clin. Proc. 45:286, 1970, with permission.)

numbers of myelinated fibers were decreased, especially distally. The distribution of unmyelinated fibers with respect to number and size was similar to that of control nerves. The frequency distribution of the diameters of myelinated fibers of sural and saphenous nerves was abnormal in that the

height of the large fiber peak was smaller than in normal nerves at the proximal level (thigh and midcalf). At the ankle level a similar but more severe change was seen. Here the largest fibers are smaller than in control nerves (Fig. 69–19).

Table 69–7. Coefficients of Variation of Diameters and Lengths of Internodes of 100 Teased Fibers from Sural Nerve of Humans in Health and in HMSN Type I Coefficients of Variation (%) Diameters Nerve

Age (yr), Sex

68–69 69–69 65–69

10, M 19, M 22, M

Lengths

Category of Nerve Donor

Mean

Range

SD

Mean

Range

SD

Accident Congenital heart disease HMSN type I

10.5 9.0 19.3

3.0–23.5 1.6–23.2 7.0–52.2

4.72 4.24 8.97

9.7 9.4 35.2

2.8–32.8 1.7–32.2 12.2–68.1

4.68 5.80 11.18

Hereditary Motor and Sensory Neuropathies: Overview of Features

FIGURE 69–13 Transverse sections of sural nerve. Phase contrast; ⫻1000. A, From a healthy 39-year-old volunteer. B, From an 11-year-old boy with probable DSN. Note demyelinated regions of myelinated fibers (examples shown by arrows). One or two nuclei are circumferentially arranged about some myelinated fibers. (From Dyck, P. J., and Gomez, M. R.: Segmental demyelination in Déjérine-Sottas disease: light, phase-contrast, and electron microscopic studies. Mayo Clin. Proc. 43:280, 1968, with permission.)

1645

1646

Diseases of the Peripheral Nervous System

internode length, 512.8 ␮m (mean of 13 control nerves, 551.2 ␮m; standard deviation [SD], 27.4 ␮m; range 499 to 615 ␮m), and adjusted coefficient of variation of internode length, 12.04% (mean of 13 control nerves, 13.8%; SD, 2.4%; range 10.2% to 17.4%). We interpret these histologic studies to indicate neuronal atrophy and degeneration of peripheral motor and sensory neurons. The neurons are affected to the greatest degree at their extremity. Enough neurons remain normal so that conduction velocity of nerves is not appreciably affected. The population of neurons affected is not as ubiquitous as in HMSN I, being more confined to the lower limb in HMSN II. FIGURE 69–14 Regression lines for relationship of number of lamellae of myelin to circumference of axis cylinders for healthy sural nerves and for nerves of persons with probable DSN. (From Dyck, P. J., Lambert, E. H., Sanders, K., et al.: Severe hypomyelination and marked abnormality of conduction in Déjérine-Sottas hypertrophic neuropathy: myelin thickness and compound action potential of sural nerve in vitro. Mayo Clin. Proc. 46:432, 1971, with permission.)

The most striking difference between HMSN I and HMSN II was found in teased fibers. The frequencies of the different conditions of teased fiber of saphenous nerve at thigh and ankle levels taken from a typical patient with this disorder are shown in Table 69–8. Although there is a large preponderance of normal fibers at both levels, many fibers were unequivocally abnormal. A similar conclusion was reached when the sural nerve was evaluated (Table 69–9). The ankle level of sural nerve of a typical case (from a 17-year-old boy) yielded the following values: adjusted

Schwann Cell–Axon Interactions There is considerable evidence of axonal pathology in HMSN I, despite the fact that the genetic causes of the neuropathies occur in myelinating Schwann cells. The evidence comes from pathologic, morphometric, and axonal flow studies.38,41,45,48,49 One source of axonal abnormalities involved a study of saphenous nerves taken from the upper thigh and distal leg and/or sural nerves of patients compared with controls, which illustrates the following findings: 1. The diameter histograms of myelinated fibers showed that they were much more severely affected distally (Fig. 69–20), as might occur in axonal atrophy. 2. There was a distal preponderance of onion bulb formation (Table 69–10). 3. The ratio of axon area to number of myelin lamellae (or myelin spiral length) was significantly decreased in HMSN nerves (Fig. 69–21).

FIGURE 69–15 Compound action potentials of healthy sural nerve and sural nerve from a patient with probable DSN, as discussed in text. (From Dyck, P. J., Lambert, E. H., Sanders, K., et al.: Severe hypomyelination and marked abnormality of conduction in Déjérine-Sottas hypertrophic neuropathy: myelin thickness and compound action potential of sural nerve in vitro. Mayo Clin. Proc. 46:432, 1971, with permission.)

Hereditary Motor and Sensory Neuropathies: Overview of Features

1647

FIGURE 69–16 Consecutive regions (A–C) of myelinated fiber from an 11-year-old boy with probable DSN, showing continuous axon with and without myelin. Note asymmetry of length of myelin on either side of nucleus in B. Outpouching of myelin or myelin ovoid in Schwann cell cytoplasm is at point of discontinuity of myelin (arrow). Phase contrast; ⫻1800. (From Dyck, P. J., and Gomez, M. R.: Segmental demyelination in Déjérine-Sottas disease: light, phase-contrast, and electron microscopic studies. Mayo Clin. Proc. 43:280, 1968, with permission.)

In fact, although abnormalities in myelin may cause HMSN I, there is increasing evidence that many of the clinical signs and symptoms of this disease (weakness and sensory loss) are produced by axonal degeneration, not demyelination. Children with HMSN I, for example, have slow NVCs prior to the onset of symptoms, and these velocities do not change appreciably as the disease progresses, suggesting that demyelination per se is not sufficient to cause their neurologic signs and symptoms.62,90,92 In addition, Krajewski and co-workers92 have demonstrated that CMAP amplitudes, not NCVs, correlate best with weakness in patients with HMSN IA, again suggesting that axonal loss is the cause of weakness. Finally, there is anatomic evidence of progressive length-dependent axonal loss in HMSN I46,58 as well as in mice overexpressing PMP22, an animal model of HMSN I.132 Taken together, these data strongly suggest that distal axonal degeneration, not demyelination, is the major cause of clinical disability in HMSN I. What is the molecular pathogenesis of this phenomenon?

A number of studies have demonstrated that Schwann cell–axon interactions are disrupted in demyelinating peripheral neuropathies, including patients with HMSN I, causing significant changes in axonal physiology. de Waegh and Brady, for example, have demonstrated that Trembler mice, which have a dysmyelinating peripheral neuropathy caused by a point mutation in PMP22 that is similar to HMSN IA, have significant changes in both axonal structure and function, including alterations in neurofilament phosphorylation, increased neurofilament density, and decreased axonal transport,34 and similar changes have been found in patients with CMT1.163 Transplantation of a segment of Trembler2 or HMSN IA130 nerve into normal nerve also produces similar changes in axons that have regenerated through the nerve graft, but not in the surrounding nerve, demonstrating that this effect is induced by contact with the abnormal Schwann cells. Finally, neurofilament packing density is increased and axonal caliber is decreased at the node of Ranvier of normal nerve, where there is no axonal contact with Schwann cells,

1648

Diseases of the Peripheral Nervous System

FIGURE 69–17 Transverse section of sural nerve of a 21-year-old woman with probable DSN, showing the largest demyelinated fiber seen in this nerve (⫻20,070). A few myelin remnants (probably from previous demyelination) are present in Schwann cell cytoplasm. Diameter of this fiber is approximately 3 ␮m, larger than any fiber seen in electron micrographs used for determination of fiber spectrum. Its diameter does not, however, come near the diameter of the axis cylinder of largest fibers of healthy nerve (6 to 8 ␮m). (From Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in Déjérine-Sottas disease. Mayo Clin. Proc. 45:286, 1970, with permission.)

compared to the adjacent myelinated internode.75,76 These data thus convincingly demonstrate that axonal contact with myelinating Schwann cells has a significant effect on underlying axonal physiology in both normal and abnormal nerve. Changes in the nature of this interaction, such as occur in HMSN I or other demyelinating peripheral neuropathies, lead to changes in neurofilament packing density, phosphorylation, and axonal transport, and subsequently axonal degeneration. Whether the changes in neurofilament packing density, phosphorylation, and axonal transport directly cause axonal damage, however, or are merely a marker of some other more fundamental axonal abnormality, is not currently known. Although the nature of the Schwann cell–derived signal that modulates neurofilament packing density, neurofila-

ment phosphorylation, and axonal transport is not known, it must be altered or modified in dysmyelinating Schwann cells. In addition, it is likely that it is expressed as a part of the coordinated program of myelin-specific gene expression, since this signal is also produced by normal Schwann cells. An attractive location for the origin of this signaling pathway is the paranodal region of the myelinating Schwann cell, where specializations occur in both the Schwann cell and its underlying axolemma. Further defining the molecular architecture of the paranode and surrounding region of axolemma will hopefully delineate the molecular pathways by which Schwann cells and axons communicate, contributing to understanding the axonal degeneration in demyelinating neuropathies. Schwann cell–axon interactions are discussed in more detail in Chapters 16 and 20.

Hereditary Motor and Sensory Neuropathies: Overview of Features

1649

FIGURE 69–18 Transverse section of sural nerve of a 21-year-old woman with probable DSN, showing onion bulb formation with small myelinated nerve fiber at its center (⫻12,800). Note complex Schwann cell processes forming lamellae separated by longitudinally directed collagen fibrils. Occasionally very small unmyelinated fibers (arrows) are seen in lamellae (nerve sprouts?). (From Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in DéjérineSottas disease. Mayo Clin. Proc. 45:286, 1970, with permission.)

UNUSUAL FORMS OF HMSN HMSN VI with Optic Atrophy In 1879, Vizioli160 described a kinship in which a father and two sons were found to have optic atrophy in association with peroneal muscular atrophy.32,74,110 Peripheral sensory neurons were probably also affected, given the lancinating pains that characteristically affect some patients with involvement of these neurons. Ages at onset were 59 years for the father, 26 years for one son, and 6 years for the other son. Loss of visual acuity, progressing to blindness,

occurred in the father and in the 26-year-old son. Muscle weakness and atrophy began in leg muscles and later affected the hand muscles. Tendon reflexes were normal or decreased. Two of the patients had pes cavus. Davidenkow reported seeing similar cases, and included those of Vizioli, by 1927.32 Taylor147 and Head72 presented cases of siblings affected by atrophy of the legs, optic atrophy, and loss of hearing. A further report of this association was given by Ferguson and Critchley.56 In the report by Milhorat, two brothers with peroneal muscular atrophy and hypertrophic neuropathy had optic

1650

Diseases of the Peripheral Nervous System

FIGURE 69–19 Frequency distribution of diameters of myelinated fibers at levels of thigh (left) and ankle (right) from healthy saphenous nerves (A) and from the saphenous nerve of a 57-year-old man with HMSN type II (B). Note that at midthigh level there is a decrease in the number of large fibers and an increase in the number of small fibers, and that the peak of the latter occurs at a smaller size category than in healthy nerve. It is our view that these changes are explained by axonal atrophy, so that large fibers shift into the small fiber group. The number of fibers at ankle level is decreased to a greater extent than at midcalf level.

atrophy.110 A sister with multiple sclerosis also had optic atrophy. Their grandparents were first cousins. Other reports of ocular manifestations and HMSN have been added to the literature.6,89,116

HMSN with Retinitis Pigmentosa

Dyck and colleagues have seen several sibships of retinitis pigmentosa and distal limb muscle weakness and atrophy. The patients had no sensory symptoms, but involvement of peripheral sensory neurons was demonstrable from conduction velocity studies of afferent fibers of peripheral nerves. Excessive phytanic acid in the plasma was sought but not found.

Massion-Verniory and associates described two siblings with retinitis pigmentosa associated with distal limb muscle weakness and atrophy.107 Cutaneous sensation was recorded as normal. Vibration at the malleolus was decreased. The myotatic reflex of the gastrocnemius and soleus muscles could not be obtained. Their parents were healthy.

Giant axonal neuropathy (GAN) is a rare autosomal recessive disorder presenting in childhood and progressing to death by the end of the third decade. Recently the genetic cause of

Table 69–8. Condition of Teased Fibers of Saphenous Nerve in HMSN Type II

Table 69–9. Condition of Teased Fibers of Sural Nerve at Ankle Level in HMSN Type II

HMSN Type II†

Controls*

A B C D E F G H I

Giant Axonal Neuropathy

Midthigh

Ankle

Midthigh

Ankle

96 2 — 1 1 — — — —

96 1 — 1 2 — — — —

98 — — 1 1 — — — —

78 — 1 6 3 — 12 — —

*Mean values of nine healthy sural nerves. † Sixty-three–year-old man.

Controls*

A B C D E F G H I

Midcalf

Ankle

98.8 0.1 — 0.2 0.8 0.1 — — —

99.1 — — 0.2 0.6 0.1 — — —

HMSN Type II Midcalf

Ankle 89 1 2 1 4 — 3 — —

*Mean values based on evaluation of 100 fibers from each of 13 sural nerves at midcalf and ankle levels.

Hereditary Motor and Sensory Neuropathies: Overview of Features

1651

FIGURE 69–20 Frequency distribution of the diameters of myelinated fibers at levels of thigh (left) and ankle (right) from healthy saphenous nerves (A) and from two cases (B and C) of HMSN type I.

the disease has been demonstrated to be mutations in a novel cytoskeletal gene termed gigaxonin.20 The name of the disorder comes from the characteristic pathologic abnormalities that result from the general disorganization of intermediate filaments in nerve axons in both the central nervous system and PNS. Because of disorganized intermediate filaments, many, though not all, patients also have characteristic kinky hair. NCV studies reveal severe reductions in CMAP and SNAP amplitudes, and needle electromyograms reveal active denervation.146

COURSE AND OUTCOME Information on course and outcome is provided in more detail in individual chapters dealing with these various disorders, but some general comments are made here. Although the mean age at onset of symptoms has been calculated for HMSN in general, these figures are not accurate estimates because patients often cannot state when their symptoms began, because nerve conduction abnormality is known to be present since infancy, and because the natural history Table 69–10. Incidence of Onion Bulbs Associated with Myelinated Fibers at Midthigh and Ankle Levels of Saphenous Nerve in HMSN Type I Site Midthigh Ankle

Control

Case 1

Case 2

0 0

2 92

50 73

may differ among the various genetic forms of HMSN. Average age of onset of HMSN was given by Bell9 as 19 years (SD 13.6 years), by Thomas and Calne149 as 12 years (SD 13 years), and by Bird and Kraft12 as 12.2 (SD 7.3 years). Dyck and colleagues39,48–50 stated that symptomatic onset usually began in the first, second, third, or fourth decades. In two early studies, nerve conduction changes over time were evaluated in 7 patients over 5 to 25 years119 and in 10 patients over 11 to 19 years.62 Dyck and colleagues assessed the change in neuropathic deficit and nerve conduction abnormalities in longitudinal studies of 31 patients with HMSN I followed for more than 15 years.44 On average, for patients younger than 15 years old at first examination, the neuropathy disability score (NDS) worsened by 0.6 points per year, by 1.1 point in patients 15 to 38 years old at first examination, and by 0.9 points in patients 40 years old or older at first examination. One NDS point is equal to a 25% weakness in one muscle or group of muscles, a 50% decrease of a muscle stretch reflex, or a 50% decrease of a modality of sensation of an index finger or great toe. Nerve conduction changed significantly over time depending on attribute, nerve, and age. In patients evaluated serially, ulnar conduction velocity increased by a few meters per second in patients who were 5 to 14 years old or 15 to 35 years old at first examination but decreased in patients who were older at first examination. In serial measurements, peroneal nerve amplitudes decreased in all three age groups. Dyck and colleagues found a statistically significant association between conduction velocity and amplitude or NDS at first and last examination, suggesting an association

1652

Diseases of the Peripheral Nervous System

FIGURE 69–21 Plot of relationship of number of myelin lamellae to area of axis cylinders of saphenous nerve myelinated fibers at thigh (A) and ankle (B) levels in patients with HMSN type I (see text for discussion).

between severity of the conduction velocity abnormality and neuropathic deficit. The severity of the conduction velocity abnormality in the young appeared to predict later neurologic abnormality. This finding suggests that the physician can predict (approximately) the later neurologic deficit from knowing the conduction velocity in the young when little deficit is present.

TREATMENT Traditional Therapies Since no specific cures are available for the inherited neuropathies, the physician’s role is to provide clinical and genetic counseling and to suggest symptomatic and rehabilitative treatment. Advice about the nature of the inheritance and the prognosis is best given only after knowing the kinship history, which in some cases means that available relatives will need to be evaluated. Genetic counseling should be based not on phenotype but on information obtained from study of the family history. Since HMSN can be transmitted as an autosomal dominant, autosomal recessive, X-linked dominant, and X-linked recessive trait, the physician must be familiar with how to recognize the pattern of inheritance (see Chapter 64). When, after adequate evaluation of family members, only the patient is known to be affected, one assumes that the inheritance is (1) autosomal dominant (but that the disorder has not been recognized in the parent, the disorder is due to a new mutation, or one of the parents has been misidentified), (2) autosomal recessive, (3) X-linked, or (4) due to another disease.

If uncertainty regarding inheritance exists, this should be explained to the patient or the parents. Families commonly ask about the nature of genetic transmission and the prognosis. We can now provide quite specific information about the genetic causes of many inherited neuropathies, but advice about the natural history of the various neuropathies should be given carefully. It is well known that many patients, even those with prominent phenotypes, live to be very old and that only a few become confined to a wheelchair. Most affected persons are able to work in manual labor, assembly lines, or clerical work, or as professionals. Symptoms are more difficult to anticipate because they may depend also on nondisease factors. A few patients within a large kinship may complain of such symptoms as pain that appear to be out of proportion to their neurologic deficits. In such patients, the physician has to weigh whether the symptoms relate to hyperactivity in nerves and the role of such coexisting conditions as excessive anxiety or depression. In countries in which life and disability insurance is related to risk, the physician may be of help in gaining insurance coverage for the patient. Insurance companies seem to be reluctant to insure patients with inherited neuromuscular disorders or may charge high premiums. In the young patient with this disorder, training in an occupation that does not require unusual manual dexterity or strength might be advised. Retraining at an older age may be possible. Because foot deformities are characteristic of this disorder, properly fitting and well-made shoes should be recommended. The shoes should be roomy and protective. Daily inspection of feet for injury may be worthwhile.

Hereditary Motor and Sensory Neuropathies: Overview of Features

Appropriate orthotics and bracing may be helpful for many patients (see Chapter 118). However, in our experience, foot surgery is often not needed. Foot surgery to correct excessively inverted feet, for severe degrees of pes cavus, and for hammer toes may improve walking and alleviate pain over pressure points and prevent plantar ulcers. The patient must be given realistic objectives for the operation, since obviously it is not a panacea for such other manifestations of the disorder as weakness and sensory loss.

1653

A mouse model of HMSN IA has been developed in which disease severity can be modified via a tetracycline responsive element in the Pmp22 promoter.118 Viral vectors are being used to introduce genes into myelin81 and neurons1 to develop gene therapy strategies. As discussed in Chapter 65, many animal models of HMSN have now been developed that can serve as test systems for future therapeutics. Hopefully the future era of molecular therapeutics will become as exciting and fruitful as the present era of molecular diagnosis.

New Therapeutic Approaches Recently, it is becoming evident that at least a small percentage of patients with inherited neuropathy may respond to immunomodulatory therapy, such as prednisone or intravenous gamma globulin. Usually, these patients have had atypical features for HMSN I that mimic CIDP. For example, the first report of this subgroup of patients was the report by Dyck and colleagues of a series of patients who presented with subacute exacerbations superimposed upon chronic neuropathies, greater proximal limb involvement than expected, excessive elevation of CSF protein, and a favorable therapeutic response to prednisone.53 At least one patient with HMSN IA has been reported with inflammatory changes on nerve biopsy suggestive of superimposed CIDP.159 Two patients with HMSN IB who developed symptoms in adulthood and had only mildly slow nerve conduction studies that improved with prednisone treatment.37,162 However, these patients probably represent the exception rather than the rule, and it is unlikely that immunomodulatory therapy will be effective in most patients with HMSN. Also recently, it has been reported that an antagonist to progesterone improves the neuropathy in the transgenic rat model of CMT1A. In these studies, administration of the selective progesterone receptor antagonist reduced overexpression of Pmp22 and improved the CMT phenotype, without obvious side effects, in the animals.136 It is not yet known whether side effects from a progesterone antagonist will limit this therapeutic approach in humans. Similarly, it is not yet known whether this therapy will be effective once the full-blown neuropathy has developed, since the rats were treated from a young age. Nevertheless, these data are very exciting and suggest that the progesterone receptor of myelin-forming Schwann cells is a promising pharmacologic target for therapy of CMT1A. Also, it has recently been shown that high doses of ascorbic acid (vitamin C) improved myelination and reduced Pmp22 levels in a Pmp22 over–expressing mouse model of CMT1A.115a Ascorbic acid has been approved by the FDA for other clinical indications. Thus, it is likely that human clinical trials with ascorbic acid for CMT1A will be rapidly undertaken. Finally, in the current era of molecular genetics and molecular cell biology, novel treatments are being developed that take advantage of these ever-expanding technologies.

REFERENCES 1. Acsadi, G., Anguelov, R. A., Yang, H., et al.: Increased survival and function of SOD-1 mice after glial cell-derived neurotrophic factor gene therapy. Hum. Gene Ther. 13:1047, 2002. 2. Aguayo, A. J., Kasarjian, J., Skamene, E., et al.: Myelination of mouse axons by Schwann cells transplanted from normal and abnormal human nerves. Nature 268:753, 1977. 3. Amick, L. D., and Lemmi, H.: Electromyographic studies in peroneal muscular atrophy: Charcot Marie Tooth disease. Arch. Neurol. 9:273, 1963. 4. Anderson, R. M., Dennett, X., Hopkins, I. J., and Shields, L.: Hypertrophic interstitial polyneuropathy in infancy: clinical and pathologic features in two cases. J. Pediatr. 82:619, 1973. 5. Antonellis, A., Ellsworth, R. E., Sambuughin, N., et al.: Glycyl tRNA synthetase mutations in Charcot-Marie-Tooth disease type 2D and distal spinal muscular atrophy type V. Am. J. Hum. Genet. 72:1293, 2003. 6. Baker, R. S., and Upton, A. R. M.: Variation of phenotype in Charcot Marie Tooth disease. Neuropaediatrie 10:290, 1979. 7. Baxter, R. V., Ben Othmane, K., Rochelle, J. M., et al.: Ganglioside-induced differentiation-associated protein-1 is mutant in Charcot-Marie-Tooth disease type 4A/8q21. Nat. Genet. 30:21, 2001. 8. Beck, P. E.: Humangenetik: ein kurzes Handbuch. Thieme, Stuttgart, p. 425, 1978. 9. Bell, J.: On the peroneal type of progressive muscular atrophy. In Fisher, R. H. (ed.): The Treasury of Human Inheritance. London, Cambridge University Press, p. 69, 1935. 10. Ben Othmane, K., Hentate, F., Lennon, F., et al.: Linkage of a locus (CMT4A) for autosomal recessive Charcot-MarieTooth disease to chromosome 8q. Hum. Mol. Genet. 2:1625, 1993. 10a. Benndorf, R., Sun, X., Gilmont, R. R., et al.: HSP22, a new member of the small heat shock protein superfamily, interacts with mimic of phosphorylated HSP27 ((3D)HSP27). J. Biol. Chem. 276:26753, 2001. 11. Bergoffen, J., Scherer, S. S., Wang, S., et al.: Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262:2039, 1993. 12. Bird, T. D., and Kraft, G. H.: Charcot-Marie-Tooth disease: data for genetic counseling relating age to risk. Clin. Genet. 14:43, 1978.

1654

Diseases of the Peripheral Nervous System

13. Bird, T. D., Ott, J., and Giblett, E. R.: Linkage of Charcot Marie Tooth neuropathy to the Duffy locus on chromosome I. Am. J. Hum. Genet. 32:99A, 1980. 14. Bird, T. D., Ott, J., and Giblett, E. R.: Linkage of Charcot Marie Tooth neuropathy to the Duffy locus on chromosome 1. Am. J. Hum Genet. 34:388, 1982. 15. Bird, T.D., Ott, J., Giblett, E. R., et al.: Genetic linkage evidence for heterogeneity in CharcotMarie-Tooth neuropathy (HMSN type I). Ann. Neurol. 14:679, 1983. 16. Blom, S., Hagbarth, K. E., and Lumberg, P. O.: Motor conduction velocities in amyotrophic lateral sclerosis, polyradiculoneuritis, and Charcot Marie Tooth’s disease. Acta Neurol. Scand. 40:6, 1964. 17. Boerkoel, C. F., Takashima, H., Garcia, C. A., et al.: Charcot-Marie-Tooth disease and related neuropathies: mutation distribution and genotype-phenotype correlation. Ann. Neurol. 51:190, 2002. 18. Boerkoel, C. F., Takashima, H., Stankiewicz, P., et al.: Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am. J. Hum. Genet. 68:325, 2001. 19. Bolino, A., Muglia, M., Conforti, F. L., et al.: CharcotMarie-Tooth type 4B is caused by mutations in the gene encoding myotubularin-related protein-2. Nat. Genet. 25:17, 2000. 20. Bomont, P., Cavalier, L., Blondeau, F., et al.: The gene encoding gigaxonin, a new member of the cytoskeletal BTB/kelch repeat family, is mutated in giant axonal neuropathy. Nat. Genet. 26:370, 2000. 21. Boveri, P.: De la nevrite hypertrophique familiale (type Pierre Marie). Semin. Med. 30:145, 1910. 22. Brimijoin, S., Capek, P., and Dyck, P. J.: Axonal transport of dopamine-␤-hydroxylase by human sural nerves in vitro. Science 180:1295, 1973. 23. Brust, J. C., Lovelace, R. E., and Devi, S.: Clinical and electrodiagnostic features of Charcot-Marie-Tooth syndrome. Acta Neurol. Scand. Suppl. 68:1, 1978. 24. Buchthal, F., and Behse, F.: Peroneal muscular atrophy (PMA) and related disorders. I. Clinical manifestations as related to biopsy findings, nerve conduction and electromyography. Brain 100(Pt. 1):41, 1977. 25. Chance, P. F., Alderson, M. K., Leppig, K. A., et al.: DNA deletion associated with hereditary neuropathy with liability to pressure palsies. Cell 72:143, 1993. 26. Chance, P. F., Murray, J. C., Bird, T. D., and Kochin, R. S.: Genetic linkage relationships of Charcot-Marie-Tooth disease (HMSN-Ib) to chromosome 1 markers. Neurology 37:325, 1987. 27. Charcot, J., and Marie, P.: Sue une forme particulaire d’atrophie musculaire progressive souvent familial debutant par les pieds et les jamber et atteingnant plus tard les mains. Rev. Med. 6:97, 1886. 27a. Chen, H., Detmer, S. A., Ewald, A. J., et al.: Mitofusins Mfn1 and Mfn2 coordinately regulate mitochondrial fusion and are essential for embryonic development. J. Cell Biol. 160:189, 2003. 28. Cuesta, A., Pedrola, L., Sevilla, T., et al.: The gene encoding ganglioside-induced differentiation-associated protein 1 is mutated in axonal Charcot-Marie-Tooth type 4A disease. Nat. Genet. 30:22, 2001.

29. Darre, B., Mollaret, P., and Landowski: La maladie de Roussy-Levy n’est-elle qu’une forme fruste ou qu’une forme abortive de la maladie de Friedreich? Interet des examens labyrinthique and chromaxique. Rev. Neurol. (Paris) 2:782, 1933. 30. Davis, C. J. F., Bradley, W. G., and Madrid, R.: The peroneal muscular atrophy syndrome: clinical, genetic, electrophysiological and nerve biopsy findings and classification. J. Genet. Hum. 26:311, 1978. 31. Dawidenkow, S.: Uber die neurotische Muskelatrophie Charcot-Marie: klinisch-genetische Studien. I. Z. Neurol. 108:259, 1927. 32. Dawidenkow, S.: Uber dir neurotische Muskelatrophie Charcot-Marie: klinisch-genetische Studien. II. Z. Neurol. 108:344, 1927. 33. De Jonghe, P., Timmerman, V., Ceuterick, C., et al.: The Thr124Met mutation in the peripheral myelin protein zero (MPZ) gene is associated with a clinically distinct CharcotMarie-Tooth phenotype. Brain 122(Pt. 2):281, 1999. 34. de Waegh, S., and Brady, S. T.: Altered slow axonal transport and regeneration in a myelin-deficient mutant mouse: the trembler as an in vivo model for Schwann cell-axon interactions. J. Neurosci. 10:1855, 1990. 35. Déjérine, H., and Sottas, J.: Sur la nevritte interstitielle, hypertrophique et progressive de l’enfance. C. R. Soc. Biol. (Paris) 45:63, 1893. 36. Déjérine, J., and Thomas, A.: Sur la nevrite interstitielle hypertrophique et progressive de l’enfance (2 observations suivie d’autopsie). Nouv. Iconog. Salpetriere 19:477, 1906. 37. Donaghy, M., Sisodiya, S. M., Kennett, R., et al.: Steroid responsive polyneuropathy in a family with a novel myelin protein zero mutation. J. Neurol. Neurosurg. Psychiatry 69:799, 2000. 38. Dyck, P. J.: Histologic measurements and fine structure of biopsied sural nerve: normal, and in peroneal muscular atrophy, hypertrophic neuropathy, and congenital sensory neuropathy. Mayo Clin. Proc. 41:742, 1966. 39. Dyck, P. J.: Inherited neuronal degeneration and atrophy affecting peripheral motor, sensory and autonomic neurons. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 825, 1975. 40. Dyck, P. J., Beahrs, O. H., and Miller, R. H.: Peripheral nerves in hereditary neural atrophies: numbers and diameters of myelinated fibers. Paper presented at the Sixth International Congress of Electroencephalography and Clinical Neurophysiology, Vienna, Austria, September 5–10, 1965. 41. Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in Dejerine-Sottas disease. Mayo Clin. Proc. 45:286, 1970. 42. Dyck, P. J., and Gomez, M. R.: Segmental demyelination in Dejerine-Sottas disease: light, phase-contrast, and electron microscopic studies. Mayo Clin. Proc. 43:280, 1968. 43. Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:280, 1968.

Hereditary Motor and Sensory Neuropathies: Overview of Features 44. Dyck, P. J., Karnes, J. L., and Lambert, E. H.: Longitudinal study of neuropathic deficits and nerve conduction abnormalities in hereditary motor and sensory neuropathy type 1. Neurology 39:1302, 1989. 45. Dyck, P. J., and Lais, A. C.: Electron microscopy of teased nerve fibers: method permitting examination of repeating structures of same fiber. Brain Res. 23:418, 1970. 46. Dyck, P. J., Lais, A. C., and Offord, K. P.: The nature of myelinated nerve fiber degeneration in dominantly inherited hypertrophic neuropathy. Mayo Clin. Proc. 49:34, 1974. 47. Dyck, P. J., and Lambert, E. H.: Numbers and diameters of nerve fibers and compound action potential of sural nerve: control and hereditary neuromuscular disorders. Trans. Am. Neurol. Assoc. 91:214, 1966. 48. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. I. Neurologic, genetic, and electrophysiologic findings in hereditary polyneuropathies. Arch. Neurol. 18:603, 1968. 49. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. II. Neurologic, genetic, and electrophysiologic findings in various neuronal degenerations. Arch. Neurol. 18:619, 1968. 50. Dyck, P. J., Lambert, E. H., and Mulder, D. W.: Charcot Marie Tooth disease: nerve conduction and clinical studies of a large kinship. Neurology 13:1, 1963. 51. Dyck, P. J., Lambert, E. H., and O’Brien, P. C.: Pain in peripheral neuropathy related to rate and kind of fiber degeneration. Neurology 26:466, 1976. 52. Dyck, P. J., Lambert, E. H., Sanders, K., and O’Brien, P. C.: Severe hypomyelination and marked abnormality of conduction in Dejerine Sottas hypertrophic neuropathy: myelin thickness and compound action potential of sural nerve in vitro. Mayo Clin. Proc. 46:432, 1971. 53. Dyck, P. J., Low, P. A., Bartelson, J. D., et al.: Prednisoneresponsive hereditary motor and sensory neuropathy. Mayo Clin. Proc. 57:239, 1982. 54. Dyck, P. J., Ott, J., Moore, B., et al.: Linkage evidence for genetic heterogeneity among kinships with hereditary motor and sensory neuropathy, type I. Mayo Clin. Proc. 58:430, 1983. 55. Dyck, P. J., Oviatt, K. F., and Lambert, E. H.: Intensive evaluation of unclassified neuropathies yields improved diagnosis. Ann. Neurol. 10:222, 1981. 55a. Evgrafov, O. V., Mersiyanova, I., Irobi, J., et al.: Mutant small heat-shock protein 27 causes axonal Charcot-Marie-Tooth disease and distal hereditary motor neuropathy. Nat. Genet. 36:602, 2004. 56. Ferguson, F. R., and Critchley, M.: Leber’s optic atrophy and its relationship with the heredo-familial ataxias. J. Neurol. Psychopathol. 9:120, 1928. 57. Friedreich, N.: Uber progressive Muskelatrophie, uber wahre und falshe Muskelhypertrophie. Berlin, Hirschwald, 1873. 58. Gabreels-Festen, A. A., Joosten, E. M., Gabreels, F. J., et al.: Early morphological features in dominantly inherited demyelinating motor and sensory neuropathy (HMSN type I). J. Neurol. Sci. 107:145, 1992.

1655

59. Gombault, A.: Contribution a l’etude anatomique de la nevrite parenchymateuse subaigue et chronique nevrite segmentaire peri-axile. Arch. Neurol. 1:11, 1880–1881. 60. Gombault, A., and Mallet, R.: Un cas de tabes ayant debute dans l’enfance: autopsie. Arch. Med. Exp. Anat. Pathol. 1:385, 1889. 61. Griffiths, L. R., Zwi, M. B., McLeod, J. G., and Nicholson, G. A.: Chromosome I linkage studies in Charcot-MarieTooth neuropathy type I. Am. J. Hum. Genet. 42:756, 1988. 62. Gutmann, L., Fakadej, A., and Riggs, J. E.: Evolution of nerve conduction abnormalities in children with dominant hypertrophic neuropathy of the Charcot-Marie-Tooth type. Muscle Nerve 6:515, 1983. 63. Gutrecht, J. A., and Dyck, P. J.: Segmental demyelination in peroneal muscular atrophy: nerve fibers teased from sural nerve biopsy specimens. Mayo Clin. Proc. 41:775, 1966. 64. Guzzetta, F., Ferriere, G., and Lyon, G.: Congenital hypomyelination polyneuropathy: pathological findings compared with polyneuropathies starting later in life. Brain 105(Pt. 2):395, 1982. 65. Haites, N., Fairweather, N., Clark, C., et al.: Linkage in a family with X-linked Charcot-Marie-Tooth disease. Clin. Genet. 35:399, 1989. 66. Halliday, J., and Whiting, A. J.: The peroneal type of muscular atrophy: with an account of a family group of cases. Br. Med. J. 2:1114, 1909. 67. Harding, A. E.: From the syndrome of Charcot, Marie and Tooth to disorders of peripheral myelin proteins. Brain 118(Pt. 3):809, 1995. 68. Harding, A. E., and Thomas, P. K.: The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103:259, 1980. 69. Harding, A. E., and Thomas, P. K.: Genetic aspects of hereditary motor and sensory neuropathy (types I and II). J. Med. Genet. 17:329, 1980. 70. Hayasaka, K., Himoro, M., Sawaishi, Y., et al.: De novo mutation of the myelin P0 gene in Dejerine-Sottas disease (hereditary motor and sensory neuropathy type III). Nat. Genet. 5:266, 1993. 71. Hayasaka, K., Ohnishi, A., Takada, G., et al.: Mutation of the myelin P0 gene in Charcot-Marie-tooth neuropathy type 1. Biochem. Biophys. Res. Commun. 194:1317, 1993. 72. Head, H.: Peroneal atrophy. Proc. R. Soc. Med. 6:51, 1913. 73. Hoffman, J.: Ueber einen eigenartigen Symptomencomplex, eine combination von Angeborenem Schwachsinn mit progressive Muskelatrophie, als weiteren Beitrag zu den erblichen Nervenkrankherten. Dtsch. Z. Nervenheilkd. 6:150, 1895. 74. Hoyt, W. F.: Charcot Marie Tooth disease with primary optic atrophy: report of a case. Arch. Ophthalmol. 64:925, 1960. 75. Hsieh, S. T., Crawford, T. O., and Griffin, J. W.: Neurofilament distribution and organization in the myelinated axons of the peripheral nervous system. Brain Res. 642:316, 1994. 76. Hsieh, S. T., Kidd, G. J., Crawford, T. O., et al.: Regional modulation of neurofilament organization by myelination in normal axons. J. Neurosci. 14(11 Pt. 1):6392, 1994.

1656

Diseases of the Peripheral Nervous System

77. Ionasescu, V., Anderson, R., Burns, T. L., et al.: Evidence for linkage of Charcot-Marie-Tooth neuropathy (CMT1) to apolipoprotein A2 (Apo-A2). Am. J. Hum. Genet. 42:74, 1988. 78. Ionasescu, V. V., Ionasescu, R., and Searby, C.: Screening of dominantly inherited Charcot-Marie-Tooth neuropathies. Muscle Nerve 16:1232, 1993. 79. Ionasescu, V. V., Ionasescu, R., Searby, C., and Barker, D. F.: Charcot-Marie-Tooth neuropathy type 1A with both duplication and non-duplication. Hum. Mol. Genet. 2:405, 1993. 79a. Irobi, J., Van Impe, K., Seeman, P., et al.: Hot-spot residue in small heat-shock protein 22 causes distal motor neuropathy. Nat. Genet. 36:597, 2004. 80. Ismailov, S. M., Fedotov, V. P., Dadali, E. L., et al.: A new locus for autosomal dominant Charcot-Marie-Tooth disease type 2 (CMT2F) maps to chromosome 7q11-q21. Eur. J. Hum. Genet. 9:646, 2001. 81. Jani, A., Menichella, D., Jiang, H., et al.: Modulation of cell-mediated immunity prolongs adenovirus-mediated transgene expression in sciatic nerve. Hum. Gene Ther. 10:787, 1999. 81a. Jordanova, A., De Jonghe, P., Boerkoel, C. F., et al.: Mutations in the neurofilament light chain gene (NEFL) cause early onset severe Charcot-Marie-Tooth disease. Brain 126:590, 2003. 82. Kaku, D. A., Parry, G. J., Malamut, R., et al.: Nerve conduction studies in Charcot-Marie-Tooth polyneuropathy associated with a segmental duplication of chromosome 17. Neurology 43:1806, 1993. 83. Kalaydjieva, L., Gresham, D., Gooding, R., et al.: N-myc downstream-regulated gene 1 is mutated in hereditary motor and sensory neuropathy-Lom. Am. J. Hum. Genet. 67:47, 2000. 84. Kalaydjieva, L., Hallmayer, J., Chandler, D., et al.: Gene mapping in Gypsies identifies a novel demyelinating neuropathy on chromosome 8q24. Nat. Genet. 14:214, 1996. 85. Kamholz, J., Menichella, D., Jani, A., et al.: CharcotMarie-Tooth disease type 1: molecular pathogenesis to gene therapy. Brain 123(Pt. 2):222, 2000. 86. Karch, S. B., and Urich, H.: Infantile polyneuropathy with defective myelination: an autopsy study. Dev. Med. Child Neurol. 17:504, 1975. 87. Kennedy, W. R., Sung, J. H., and Berry, J. F.: A case of congenital hypomyelination neuropathy: clinical, morphological, and chemical studies. Arch. Neurol. 34:337, 1977. 88. Kennedy, W. R., Sung, J. H., Berry, J. F., and Mastri, A.: Hypertrophic neuropathy with primary failure of peripheral myelination. Trans. Am. Neurol. Assoc. 96:75, 1971. 89. Khoubesserian, P., Van Regemarter, N., Ohrn-Degueldre, O., et al.: Charcot-Marie-Tooth disease associated with retinal pigment dystrophy and protanopia: neurological, ophthalmological and genetic study of a family. J. Neurol. 222:1, 1979. 90. Killian, J. M., Tiwari, P. S., Jacobson, S., et al.: Longitudinal studies of the duplication form of CharcotMarie-Tooth polyneuropathy. Muscle Nerve 19:74, 1996. 91. Klein, C. J., Cunningham, J. M., Atkinson, E. J., et al.: The gene for HMSN2C maps to 12q23-24: a region of neuromuscular disorders. Neurology 60:1151, 2003.

92. Krajewski, K. M., Lewis, R. A., Fuerst, D. R., et al.: Neurological dysfunction and axonal degeneration in CharcotMarie-Tooth disease type 1A. Brain 123(Pt. 7):1516, 2000. 93. Lambert, E. H.: Electromyography and electric stimulation of peripheral nerves and muscle. In Clinical Examination in Neurology. Philadelphia, W. B. Saunders, p. 287, 1956. 94. Lambert, E. H.: Neurophysiological techniques useful in the study of neuromuscular disorders. In Adams, R. D., Eaton, L. M., and Shy, G. M. (eds.): Neuromuscular Disorders, Vol. 38: The Motor Unit and Its Disorders. Baltimore, Williams & Wilkins, p. 247, 1960. 95. Lambert, E. H., Bastron, J. A., and Mulder, D. W.: Conduction velocity of motor fibers of peripheral nerves in peroneal muscular atrophy (Charcot Marie Tooth disease). Paper presented at the annual meeting of the American Academy of Neurology, Philadelphia, April, 1958. 96. Leal, A., Morera, B., Del Valle, G., et al.: A second locus for an axonal form of autosomal recessive Charcot-Marie-Tooth disease maps to chromosome 19q13.3. Am. J. Hum. Genet. 68:269, 2001. 97. Lebo, R. V., Anderson, L., Lau, C., et al.: Linkage analysis of Charcot Marie Tooth syndrome. Muscle Nerve 9(Suppl.):235, 1986. 98. Lewis, R. A., and Shy, M. E.: Electrodiagnostic findings in CMTX: a disorder of the Schwann cell and peripheral nerve myelin. Ann. N. Y. Acad. Sci. 883:504, 1999. 99. Lewis, R. A., and Sumner, A. J.: The electrodiagnostic distinctions between chronic familial and acquired demyelinative neuropathies. Neurology 32:592, 1982. 100. Lewis, R. A., Sumner, A. J., and Shy, M. E.: Electrophysiological features of inherited demyelinating neuropathies: a reappraisal in the era of molecular diagnosis. Muscle Nerve 23:1472, 2000. 101. Liu, H., Nakagawa, T., Kanematsu, T., et al.: Isolation of 10 differentially expressed cDNAs in differentiated Neuro2a cells induced through controlled expression of the GD3 synthase gene. J. Neurochem. 72:1781, 1999. 102. Lupski, J. R., Montes de Oca-Luna, R., Slaugenhaupt, S., et al.: DNA duplication associated with Charcot-MarieTooth disease type 1A. Cell 66:219, 1991. 103. Lyon, G.: Ultrastructural study of a nerve biopsy from a case of early infantile chronic neuropathy. Acta Neuropathol. 13:131, 1969. 104. Magyar, J. P., Martini, R., Ruelicke, T., et al.: Impaired differentiation of Schwann cells in transgenic mice with increased PMP22 gene dosage. J. Neurosci. 16:5351, 1996. 105. Marie, M. P.: Forme speciale de nevrite interstitiells hypertrophique progressive de l’enfance. Rev. Neurol. (Paris) 14:557, 1906. 106. Marinesco, G.: De l’amyotrophie Charcot Marie. Arch. Med. Exp. Anat. Pathol. 6:921, 1894. 107. Massion-Verniory I., Dumont, E., and Potvin, A. M.: Retinite pigmentaire familiale complique d’une amyotrophie neurite. Rev. Neurol. (Paris) 78:561, 1946. 108. Mersiyanova, I. V., Perepelov, A. V., Polyakov, A. V., et al.: A new variant of Charcot-Marie-Tooth disease type 2 is probably the result of a mutation in the neurofilamentlight gene. Am. J. Hum. Genet. 67:37, 2000. 109. Middleton-Price, H. R., Harding, A. E., Berciano, J., et al.: Absence of linkage of hereditary motor and sensory

Hereditary Motor and Sensory Neuropathies: Overview of Features

110.

111.

112.

113.

114.

115. 115a.

116.

117.

118.

119.

120.

121.

122.

123.

neuropathy type I to chromosome 1 markers. Genomics 4:192, 1989. Millhorat, A. T.: Studies in diseases of muscles: XIV. Progressive muscular atrophy of peroneal type associated with atrophy of the optic nerves: report on a family. Arch. Neurol. 50:279, 1943. Myrianthopoulos, N. C., Lane, M. H., Silberberg, D. H., and Vincent, B. L.: Nerve conduction and other studies in families with Charcot Marie Tooth disease. Brain 87:589, 1964. Nelis, E., Timmerman, V. De Jonghe, P., and Van Broeckhoven, C.: Identification of a 5' splice site mutation in the PMP-22 gene in autosomal dominant Charcot-MarieTooth disease type 1. Hum. Mol. Genet. 3:515, 1994. Nelis, E., Van Vroeckhoven, C., De Jonghe, P., et al.: Estimation of the mutation frequencies in CharcotMarie-Tooth disease type 1 and hereditary neuropathy with liability to pressure palsies: a European collaborative study. Eur. J. Hum. Genet. 4:25, 1996. Nicholson, G., and Nash, J.: Intermediate nerve conduction velocities define X-linked Charcot-Marie-Tooth neuropathy families. Neurology 43:2558, 1993. Ott, J.: Analysis of Human Genetic Linkage. Baltimore, Johns Hopkins University Press, p. 185, 1985. Passage, E., Norreel, J. C., Noack-Fraissignes, P., et al.: Ascorbic acid treatment corrects the phenotype of a mouse model of Charcot-Marie-Tooth disease. Nat. Med. 10:396, 2004. Peiffer, J., Schlote, W., Bischoff, A., et al.: Generalized giant axonal neuropathy: a filament-forming disease of neuronal, endothelial, glial, and Schwann cells in a patient without kinky hair. Acta Neuropathol. (Berl.) 40:213, 1977. Pentao, L., Wise, C. A., Chinault, A. C., et al.: CharcotMarie-Tooth type 1A duplication appears to arise from recombination at repeat sequences flanking the 1.5 Mb monomer unit. Nat. Genet. 2:292, 1992. Perea, J., Robertson, A., Tolmachova, T., et al.: Induced myelination and demyelination in a conditional mouse model of Charcot-Marie-Tooth disease type 1A. Hum. Mol. Genet. 10:1007, 2001. Pinelli, P., Cosi, V., and Poloni, M.: EMG longitudinal studies in peroneal muscular atrophy: different findings in hypomyelination and neuronal degeneration. In Serratrice, G., and Roux, H. (eds.): Peroneal Atrophies and Related Disorders. New York, Masson Publishing, p. 153, 1979. Plante-Bordeneuve, V., Guiochon-Mantel, A., Lacroix, C., et al.: The Roussy-Levy family: from the original description to the gene. Ann. Neurol. 46:770, 1999 Popow, A.: Une famille atteinte d’une forme particulaire de maladie hereditaire (forme de Roussy-Lévy). Rev. Neurol. (Paris) 2:447, 1932 Raeymaekers, P., Timmerman, V., De Jonghe, P., et al.: Localization of the mutation in an extended family with Charcot-Marie-Tooth neuropathy (HMSN I). Am. J. Hum. Genet. 45:953, 1989. Raeymaekers, P., Timmerman, V., Nelis, E., et al.: Duplication in chromosome 17p11.2 in Charcot-MarieTooth neuropathy type 1a (CMT 1a). The HMSN Collaborative Research Group. Neuromuscul. Disord. 1:93, 1991.

1657

124. Roa, B. B., Garcia, C. A., Pentao, L., et al.: Evidence for a recessive PMP22 point mutation in Charcot-Marie-Tooth disease type 1A. Nat. Genet. 5:189, 1993. 125. Roa, B. B., Garcia, C. A., Suter, U., et al.: Charcot-MarieTooth disease type 1A: association with a spontaneous point mutation in the PMP22 gene. N. Engl. J. Med. 329:96, 1993. 126. Rogers, T., Chandler, D., Angelicheva, D., et al.: A novel locus for autosomal recessive peripheral neuropathy in the EGR2 region on 10q23. Am. J. Hum. Genet. 67:664, 2000. 127. Roussy, G. and Lévy, G.: A sept cas d’une maladie familiale particulaire. Rev. Neurol. (Paris) 33:427, 1926. 128. Roussy, G. and Lévy, G.: La dystasie areflexique hereditaire. Presse Med. 2:1733, 1932. 129. Roussy, G. and Lévy, G.: A propos de la dystasie areflexique hereditaire. Rev. Neurol. (Paris) 2:763, 1934. 130. Sahenk, Z., Chen, L., and Mendell, J. R.: Effects of PMP22 duplication and deletions on the axonal cytoskeleton. Ann. Neurol. 45:16, 1999. 131. Sambuughin, N., Sivakumar, K., Selenge, B., et al.: Autosomal dominant distal spinal muscular atrophy type V (dSMA-V) and Charcot-Marie-Tooth disease type 2D (CMT2D) segregate within a single large kindred and map to a refined region on chromosome 7p15. J. Neurol. Sci. 161:23, 1998. 132. Sancho, S., Magyar, J. P., Aguzzi, A., and Suter, U.: Distal axonopathy in peripheral nerves of PMP22-mutant mice. Brain 122(Pt. 8):1563, 1999. 133. Senderek, J., Bergmann, C., Stendel, C., et al.: Mutations in a gene encoding a novel SH2/TPR domain protein cause autosomal recessive Charcot-Marie-Tooth type 4C neuropathy. Am. J. Hum. Genet. 73:1106, 2003. 134. Senderek, J., Bergmann, C., Weber, S., et al.: Mutation of the SBF2 gene, encoding a novel member of the myotubularin family, in Charcot-Marie-Tooth neuropathy type 4B2/11p15. Hum. Mol. Genet. 12:349, 2003. 135. Sereda, M., Griffiths, I., Puhlhofer, A., et al.: A transgenic rat model of Charcot-Marie-Tooth disease. Neuron 16:1049, 1996. 136. Sereda, M. W., Meyer zu Horst, G., Suter, U., et al.: Therapeutic administration of progesterone antagonist in a model of Charcot-Marie-Tooth disease (CMT-1A). Nat. Med. 9:1533, 2003. 137. Shy, M., Garbern, J., and Kamholz, J.: Hereditary motor and sensory neuropathies: a biological perspective. Lancet Neurol. 1:110, 2002. 138. Shy, M. E., Jáni, A., Krajewski, K., et al.: Phenotypic clustering in MPZ mutations. Brain 127:371, 2004. 139. Skre, H.: Genetic and clinical aspects of Charcot-MarieTooth’s disease. Clin. Genet. 6:98, 1974. 140. Street, V. A., Bennett, C. L., Goldy, J. D., et al.: Mutation of a putative protein degradation gene LITAF/SIMPLE in Charcot-Marie-Tooth disease 1C. Neurology 60:22, 2003. 140a. Sun, X., Fontaine, J. M., Rest, J. S., et al.: Interaction of human HSP22 (HSPB8) with other small heat shock proteins. J. Biol. Chem. 279:2394, 2004. 141. Suter, U., Moskow, J. J., Welcher, A. A., et al.: A leucineto-proline mutation in the putative first transmembrane domain of the 22-kDa peripheral myelin protein in the trembler-J mouse. Proc. Natl. Acad. Sci. U. S. A. 89:4382, 1992.

1658

Diseases of the Peripheral Nervous System

142. Suter, U., Welcher, A. A., Ozcelik, T., et al.: Trembler mouse carries a point mutation in a myelin gene. Nature 356:241, 1992. 143. Symonds, C. P., and Shaw, M. E.: Familial claw-foot with absent tendon-jerks: a forme fruste of the Charcot Marie Tooth disease. Brain 49:387, 1926. 144. Tachi, N., Ishikawa, Y., Tsuzuki, A., et al.: Hereditary motor and sensory neuropathy type I in early childhood: clinical features and the findings of nerve pathology [in Japanese]. No To Hattatsu 17:318, 1985. 145. Takashima, H., Boerkoel, D. F., De Jonghe, P., et al.: Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51:709, 2002. 146. Tandan, R., Little, B. W., Emery, E. S., et al.: Childhood giant axonal neuropathy: case report and review of the literature. J. Neurol. Sci. 82:205, 1987. 147. Taylor, J.: Peroneal atrophy. Proc. R. Soc. Med. 6:50, 1913. 148. Thevenard, A., and Berdet, H.: Remarques sur l’hypertrophie des nerfs peripheriques: interet de l’exploration clinique de plexus cervical superficiel en particulier de sa branche auriculaire et de son examen histologique apres biopsie. Presse Med. 66:529, 1958. 149. Thomas, P. K., and Calne, D. B.: Motor nerve conduction velocity in peroneal muscular atrophy: evidence for genetic heterogeneity. J. Neurol. Neurosurg. Psychiatry 37:68, 1974. 150. Thomas, P. K., Marques, W. Jr., Davis, M. B., et al.: The phenotypic manifestations of chromosome 17p11.2 duplication. Brain 120(Pt. 3):465, 1997. 151. Timmerman, V., Raeymaekers, P., De Jonghe, P., et al.: Assignment of the Charcot-Marie-Tooth neuropathy type 1 (CMT 1a) gene to 17p11.2-p12. Am. J. Hum. Genet. 47:680, 1990. 152. Tooth, H.: The Peroneal Type of Progressive Muscular Atrophy. London, Lewis, 1886. 153. Tsubaki, T., and Ishida, H.: A pedigree of Roussy-Levy syndrome. Brain Nerve (Tokyo) 15:1142, 1963. 154. Valentijn, L. J., Baas, F., Wolterman, R. A., et al.: Identical point mutations of PMP-22 in Trembler-J mouse and Charcot-Marie-Tooth disease type 1A. Nat. Genet. 2:288, 1992. 155. Vance, J. M., Barker, D., Yamaoka, L. H., et al.: Localization of Charcot-Marie-Tooth disease type 1a (CMT1A) to chromosome 17p11.2. Genomics 9:623, 1991.

156. Vance, J. M., Nicholson, G. A., Yamaoka, L. H., et al.: Linkage of Charcot-Marie-Tooth neuropathy type 1a to chromosome 17. Exp. Neurol. 104:186, 1989. 157. Verhoeven, K., De Jonghe, P., Coen, K., et al.: Mutations in the small GTP-ase late endosomal protein RAB7 cause Charcot-Marie-Tooth type 2B neuropathy. Am. J. Hum. Genet. 72:722, 2003. 158. Virchow, R.: Ein Fall von progressive Muskelatrophie. Arch. Pathol. Anat. 8:537, 1855. 159. Vital, A., Vital, C., Lagueny, A., et al.: Inflammatory demyelination in a patient with CMT1A. Muscle Nerve 28:373, 2003. 160. Vizioli: Dell’atrofia progr. nervosa. Bolli. R. Acad. MedicoChir. Napoli, 1889. 161. Warner, L. E., Mancias, P., Butler, I. J., et al.: Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet. 18:382, 1998. 162. Watanabe, M., Yamamoto, N., Ohkoshi, N., et al.: Corticosteroid-responsive asymmetric neuropathy with a myelin protein zero gene mutation. Neurology 59:767, 2002. 163. Watson, D. F., Nachtman, F. N., Kunel, R. W., and Griffin, J. W.: Altered neurofilament phosphorylation and beta tubulin isotypes in Charcot-Marie-Tooth disease type 1. Neurology 44:2383, 1994. 164. Wise, C. A., Garcia, C. A., Davis, S. N., et al.: Molecular analyses of unrelated Charcot-Marie-Tooth (CMT) disease patients suggest a high frequency of the CMTIA duplication. Am. J. Hum. Genet. 53:853, 1993. 165. Yokomori, K.: Ueber neuritis interstitialis hypertrophica et progressiva (Dejerine und Sottas) mit einem Sektionbefund. Mitt. Med. Fakult. Univ. Tokyo 15:1, 1915. 166. Yudell, A., Dyck, P. J., and Lambert, E. H.: A kinship with the Roussy-Levy syndrome: a clinical and electrophysiologic study. Arch Neurol. 13:432, 1965. 167. Zhao, C., Takita, J., Tanaka, Y., et al.: Charcot-Marie-Tooth disease type 2A caused by mutation in a microtubule motor KIF1Bbeta. Cell 105:587, 2001. 167a. Zuchner, S., Mersiyanova, I. V., Muglia, M., et al.: Mutations in the mitochondrial GTPase mitofusin 2 cause Charcot-Marie-Tooth neuropathy type 2A. Nat. Genet. 36:449, 2004.

70 Hereditary Motor and Sensory Neuropathies Involving Altered Dosage or Mutation of PMP22: The CMT1A Duplication and HNPP Deletion JAMES R. LUPSKI AND PHILLIP F. CHANCE

The CMT1A Duplication Genetic Basis of CMT1A Clinical Findings in Patients with CMT1A Duplication Pathologic Features Electrophysiologic Features CMT1A Duplication Case Studies Management Increased PMP22 Gene Dosage as the Cause of CMT1A

PMP22 Point Mutations as the Cause of Demyelinating Neuropathies The HNPP Deletion Genetic Basis of HNPP Clinical Findings in Patients with HNPP Deletion Pathologic Features Electrophysiologic Features HNPP Deletion Case Studies Management

Charcot-Marie-Tooth disease type 1 (CMT1), or hereditary motor and sensory neuropathy type I (HMSN I), is a genetically heterogeneous group of chronic, slowly progressive demyelinating peripheral neuropathies affecting both motor and sensory nerves.18,100,162,167,182 CMT type 1A (CMT1A) is the most common form of CMT1. Hereditary neuropathy with liability to pressure palsies (HNPP; also called tomaculous neuropathy) is another autosomal dominant disorder that leads to recurrent, focal demyelinating neuropathies. For the majority of CMT1 and HNPP patients, the disease does not result from mutations of a specific gene. Instead, these neuropathies are most frequently associated with a DNA rearrangement on chromosome 17p12 leading to duplication (in the case of CMT1A) or reciprocal deletion (in the case of HNPP) of a 1.4-megabase (Mb) segment containing a dosagesensitive myelin gene, the peripheral myelin protein 22 (PMP22) gene. In rare instances, point mutations in

Decreased PMP22 Gene Dosage as the Cause of HNPP Animal Models for HNPP Inherited Brachial Plexus Neuropathies Structure of the CMT1A/HNPP Genomic Region Generation of the CMT1A Duplication/HNPP Deletion by Unequal Crossing Over Molecular Diagnostics for Detection of CMT1A Duplication/HNPP Deletion

PMP22 may account for CMT1A or HNPP. Taken together, the CMT1A duplication and HNPP deletion are responsible for a significant fraction of inherited, demyelinating neuropathies. An extensive clinical, electrophysiologic, and neuropathologic description of hereditary motor and sensory neuropathies is given in Chapter 69. This chapter focuses on the clinical manifestations, molecular mechanisms, and diagnostic methods for detection of the CMT1A duplication and HNPP deletion. The roles of PMP22 dosage alteration, as well as PMP22 point mutations, in both inherited and sporadic demyelinating neuropathy are described. Each of these molecular lesions is responsible for a primarily demyelinating process. However, it must be emphasized that the demyelination is not the direct cause of muscle atrophy and weakness accompanying progression of the disease, but instead appears to act by initiating axonal loss166 (see Chapter 69). 1659

1660

Diseases of the Peripheral Nervous System

THE CMT1A DUPLICATION Genetic Basis of CMT1A The genetic locus for CMT1A was mapped to chromosome 17203 and is most often associated with a 1.4-Mb duplication in band 17p12.99,144 The CMT1A duplication99,144 (Fig. 70–1) is responsible for approximately 70% of autosomal dominant CMT1124,210 and 76% to 90% of sporadic CMT1.62,124 The CMT1A duplication was initially identified in Americans of French-Acadian and Ashkenazi Jewish decent99 as well as Belgian and Dutch families,144 and subsequently in all other world populations investigated.1,11,23,54,59,61,62,115,124,125,129 The CMT1A duplication was identified in 52% of 153 patients with inherited neuropathy, consisting of CMT or related disease, who enrolled in a study prior to the availability of clinical molecular testing.19 This fraction rises to 70%, as has been previously reported,124,210 if only patients with electrodiagnostic features consistent with demyelinating neuropathy are considered.19 The CMT1A duplication is stable during both meiosis and mitosis.102 The presence of the CMT1A duplication correlates with the diagnostic criteria for CMT1, including reduced or slowed motor and sensory nerve conduction velocities (NCV).73 The duplication is an excellent biologic marker for CMT1A because it is the causative genetic

Normal 17p

Dup 17p

Normal 17p

13.3

13.2

13.2

13.1

13.1

1.4 Mb dup

Clinical Findings in Patients with CMT1A Duplication The majority of individuals with the CMT1A duplication present with a classic HMSN I or CMT1 phenotype (see Chapter 69), including a childhood or adolescent onset with slowly progressive, symmetrical demyelinating polyneuropathy, leading to variable degrees of distal muscle weakness, atrophy, and sensory disturbances (Table 70–1). However, wide variability in clinical expression has also been observed in patients with the CMT1A duplication,

Del 17p

13.3

12

event. The initial methods that were utilized to detect the CMT1A duplication included (1) dosage difference of restriction fragment length polymorphism (RFLP) alleles in heterozygous individuals99,144; (2) the presence of three alleles at a highly polymorphic locus detected by polymerase chain reaction (PCR) or Southern blot analysis99,144; (3) rearrangement-specific junction fragments detected by Southern blot analysis of pulsed-field gel electrophoresis (PFGE)–separated genomic DNA99; and (4) three fluorescing signals, a doublet from the duplicated chromosome and a unique signal from the normal chromosome 17 homologue, visualized in interphase nuclei of peripheral blood leukocytes by fluorescent in situ hybridization (FISH).99

12

11.2

11.2

11.2

11.2

CMT1A duplication

HNPP deletion

1.4 Mb del

FIGURE 70–1 Neuropathy causing rearrangements in the short arm of chromosome 17 (17p). An ideogram of the short arm of chromosome 17 homologues is shown with the light and dark bands numbered. Indentations refer to the centromere. On the left of each pair is shown the normal 17p copy while on the right of each pair are the rearranged chromosomes. The segment duplicated in the CMT1A duplication and deleted in the HNPP deletion is 1.4 Mb. This is submicroscopic in size and cannot be resolved by conventional clinical cytogenetics, but rather requires molecular technologies to be revealed. Note that the CMT1A duplication is a tandem or direct duplication (vertical arrows) that results in three copies of the PMP22 gene, one copy on the normal chromosome and two copies on the duplicated chromosome. The HNPP deletion results in the absence of one PMP22 copy and, thus, haploinsufficiency of this locus.

HMSNs Involving Altered Dosage or Mutation of PMP22

Table 70–1. Clinical Findings in a Series of Patients with the CMT1A Duplication Feature Total no. of patients Age at examination: mean (range) Onset in first decade Presenting signs Abnormal gait Motor delays Foot deformity Other Clinical findings Atrophy/weakness Upper and lower Lower only None detected Sensory loss Apparent None detected Tendon reflexes Absent Depressed Normal Plantar responses Flexor Extensor No response Pes cavus Scoliosis Hand deformity MNCV profile (m/s, range) Median nerve Peroneal/tibial nerve Onion bulbs (N ⫽ 10)

61 (32M/29F) 37 yr (3–73 yr) 46 (75%) 21 (34%) 10 (16%) 20 (32%) 10 (16%)

45 (75%) 8 (13%) 7 (12%) 43 (70%) 18 (30%) 46 (75%) 13 (21%) 2 (4%) 55 (91%) 3 (5%) 2 (4%) 44 (72%) 8 (13%) 4 (6%) 19.9 (5–34) 17.0 (10–22) 10.0 (100%)

F ⫽ female; M ⫽ male; MNCV ⫽ motor nerve conduction velocity. From Thomas, P. K., Marques, W. Jr., Davis, M. B., et al.: The phenotypic manifestations of chromosome 17p11.2 duplication. Brain 120(Pt. 3):465, 1997, with permission.

including that reported in monozygotic twins in which one co-twin was severely affected while the other twin was clinically asymptomatic.46 In addition to CMT1, six other variant clinical phenotypes have been associated with the CMT1A duplication. Déjérine-Sottas neuropathy (DSN) (Online Inheritance in Man [OMIM] database #145900; see www.ncbi.nih.gov/htbin-post/Omim/getmim) has been associated with heterozygous duplication173 as well as homozygous duplication.73,78,91,99,135,180 The CMT1A duplication has also been reported in patients with diagnoses of Roussy-Lévy syndrome7,190 (OMIM #180800), calf hypertrophy,86,196 and scapuloperoneal atrophy or Davidenkow syndrome (OMIM #181400).58,158 Pressure palsies have also been reported as the initial presentation

1661

in a case of late-onset CMT1A associated with the CMT1A duplication.2 In one family, the presentation consisted of several episodes of acute paralysis and a sensory-dominant polyneuropathy.117 Patients with the CMT1A duplication, like other patients with inherited neuropathy, are susceptible to severe acute polyneuropathy when given vincristine.50 Thus vincristine should be avoided, and it seems prudent to avoid other neurotoxic agents. Recently, coexistent inflammatory neuropathy has been described in eight atypical CMT patients who developed acute or subacute deterioration, seven of whom had the CMT1A duplication.48 Each patient had an acute or subacute deterioration following a long stable or asymptomatic period; seven had neuropathic pain or prominent positive sensory symptoms. Importantly, three of four patients with CMT1A duplication and coexistent inflammatory neuropathy responded to immunomodulatory therapy.48 Several large clinical studies (⬎50 patients each) collectively suggested general trends for patients with the CMT1A duplication14,44,73,190,210 (see Table 70–1). The onset of the first functional manifestations was in the first decade in 50% of cases and before the age of 20 years in 70% of cases.14

Pathologic Features Sural nerve biopsy101 (Fig. 70–2) reveals a paucity of myelinated fibers and increased endoneurial connective tissue, but the epineurium and perineurium are normal and appear not to participate in the pathologic process. There is also a consistent increase in the number of Schwann cell nuclei and a decrease in the density of large myelinated fibers. Onion bulbs, formed by circumferentially interdigitated and flattened Schwann cell processes that are separated by bundles of collagen fibers, are easily seen, especially in severe cases. Ultrastructural studies show onion bulbs surrounding myelinated internodes, demyelinated internodes, or former sites of demyelination. Teased fibers reveal widespread segmental demyelination with frequent paranodal loss of myelin. In addition, there can be hypertrophic changes of atypical appearances on nerve biopsy in CMT1A duplication patients who have superimposed diabetes mellitus.189

Electrophysiologic Features NCV studies in CMT1A are consistent with an inherited demyelinating polyneuropathy. They demonstrate marked, uniform slowing of conduction velocity in all nerves, typically less than 70% of the lower limit of normal. The findings in CMT1A are essentially indistinguishable from those usually seen in CMT type 1B and CMT type 1C.13 In most patients with CMT1A, median and ulnar conduction velocities range

1662

Diseases of the Peripheral Nervous System

A

B

C

FIGURE 70–2 CMT1A neuropathology. A, Plastic section of a sural nerve biopsy from a patient with CMT1A showing Schwann cell proliferation around a thinly myelinated fiber resulting in onion bulb formation. B, Paraffin section of a sural nerve biopsy stain with Bodiantrichrome showing a uniform loss of myelinated fibers and marked proliferation of Schwann cells around myelinated and unmyelinated fibers, resulting in onion bulb formation. See Color Plate C, Electron micrograph of a sural nerve biopsy from a patient with CMT1A showing a thinly myelinated fiber surrounded by concentric Schwann cell processes, resulting in onion bulb formation. (Courtesy of Dr. Carlos A. Garcia, Departments of Neurology and Pathology, Tulane University, New Orleans, LA.)

between 15 and 30 m/s, with occasional patients reaching a NCV of 40 m/s.73,74 Maximal slowing evolves during the first 3 to 5 years of life with apparently little change thereafter.44,79 Slowing of conduction velocities has been found in patients as young as age 6 months, indicating that nerve conduction studies may be helpful in early diagnosis. Hence, in a patient several months of age who is suspected of having CMT1A but who has normal conduction velocities, the diagnosis is essentially excluded. Distal latency may increase during the first 10 years. Compound muscle action potential (CMAP) amplitudes are typically absent or reduced in the lower extremities and borderline or low in the upper extremities. Conduction block is not a feature of CMT1 in general, or CMT1A in particular.73 Inability to obtain supramaximal stimulation at proximal sites, as a result of high thresholds, may account for some of the reductions in proximal amplitudes reported. Temporal dispersion (⬎20% increase in duration) with proximal stimulation is also extremely rare in CMT1. Sensory studies are often absent because of the more

profound effect of phase cancellation on sensory nerve action potentials than CMAPs. When recorded, they are often low in amplitude with slowing in latencies, similar to the motor latencies.85 There is typically little correlation between the degree of NCV slowing and degree of clinical disability. However, CMAP amplitude reduction is more closely related to disability because it is a marker of axonal loss.85 Also, there is no apparent relationship between conduction slowing and sex, age, severity of disease, and length of symptomatic disease.85 Needle electromyography in CMT1A typically shows little active denervation, with long, large motor unit action potentials consistent with distal reinnervation. In 50 of 61 adult CMT1A duplication patients, the presentation was a typical CMT phenotype, whereas 8 were classified as having Roussy-Lévy syndrome and 3 had associated pyramidal signs, including 1 patient with a complicated neuropathy with signs of cerebellar and bulbar

HMSNs Involving Altered Dosage or Mutation of PMP22

involvement.190 The average median and ulnar motor NCVs were 20.8 ⫾ 6 m/s and 19.9 ⫾ 7 m/s, respectively, and these were not significantly different from those observed for CMT1 patients without the CMT1A duplication.210 Median conduction velocities can vary widely, ranging from 10 to 42 m/s, and can even differ by as much as 27 m/s in CMT1A duplication patients within the same family.73,210 One longitudinal study spanning over 20 years in a family with CMT1A duplication revealed that motor conduction velocities and clinical motor exam did not change significantly during 22 years.79 Clinical and electrophysiologic evaluations of CMT1A in infancy and early childhood showed that all affected children had abnormal NCVs from 2 years of age, and most had symptoms that appeared in early childhood. However, the florid clinical picture did not usually occur until the second decade of life.44

CMT1A Duplication Case Studies Patient 1 A 17-year-old male presented with difficulty in walking that had become progressively worse during the preceding 2 to 3 years. He was noted to have had high arches at age 2 years, and at age 4 years was noted to stumble frequently and walk “pigeon-toed.” He was unable to participate in active sports after beginning school at age 6 years. Regarding family history, his mother noted that “weakness of the ankles” and “high arches” were present in a number of members on her side of the family. On exam, the patient had no abnormalities of the cranial nerves. There was no hearing loss, nystagmus, or pupillary abnormalities. Weakness and atrophy were pronounced in the tibialis anterior and peroneal muscles bilaterally, and there was some wasting of the intrinsic hand muscles and proximal forearms. Deep tendon reflexes were completely absent, and sensory testing revealed a significant loss of light touch and pain sensation in all extremities. The vibratory threshold was moderately increased at the ankles. Additionally, position sense was compromised at the ankle. Gait was characterized by bilateral footdrop with a tendency to invert the feet at the ankles. No scoliosis was present. The ulnar nerves were palpably enlarged at the elbows. Nerve conduction testing revealed that velocities of the ulnar and median nerves were 9 and 10 m/s, respectively. Molecular testing revealed the presence of the chromosome 17p duplication, consistent with CMT1A in the patient and in his mother. Patient 2 A 25-year-old female presented with a 2- to 3-year history of increasing fatigue during sports, cramping of the calf muscles, and progressive bilateral hammer toe formation. She had a history of normal developmental milestones, including no difficulties with any athletic activities during

1663

childhood and adolescence. Of interest is that at age 18, during college, she was a volunteer participant in a series of neurophysiologic experiments, and was found to have significantly reduced motor and sensory NCVs, the significance of which was unexplained. In her early 20s she began having muscle cramps, particularly in the lower legs following strenuous athletic activities. Over the past 2 years, she had noted increased foot deformity, but denied any specific gait abnormality. Her family history was only remarkable for “high-arched feet” in her father and in an older sister. On physical examination, the patient had no cranial nerve abnormalities. Gait was essentially normal; however, the patient could not heel walk. There was slight atrophy of the dorsal interosseous muscles and, additionally, slight atrophy of the thenar eminence bilaterally. In general, there was slight distal muscle weakness. Sensory testing only disclosed a mild elevation of the vibratory threshold at the ankles, and all deep tendon reflexes were diminished. There was pes cavus with hammer toe formation. Motor NCV values ranged from 25 to 32 m/s in ulnar, medial, or peroneal nerves. Molecular testing disclosed the presence of the chromosome 17p duplication consistent with CMT1A.

Management Therapy for CMT1A is the same as for other forms of CMT.45,100,162 Treatment is directed toward maintaining function of joints and muscles. Because CMT is a slowly progressive neurodegenerative disease, periodic evaluation and assessment are required. Because of the tremendous variability of clinical expression, management must be individualized.

Increased PMP22 Gene Dosage as the Cause of CMT1A The gene dosage model proposes that the presence of three copies of a “CMT gene” instead of the usual two is sufficient to cause the disease.99 Three copies result because of the CMT1A duplication: two copies on the duplicated chromosome and one on the normal homologue (see Fig. 70–1). The finding of a more severe neuropathy phenotype in patients homozygous for the duplication73,78,99 suggested that the copy number, or gene dosage, may be associated with the clinical manifestation of CMT1. Support for a gene dosage model was provided by the evaluation of patients with visible chromosomal duplications as well as by genetically engineered animals. To investigate further a potential model for gene dosage in the CMT1A phenotype, a patient with a cytogenetically visible duplication of the short arm of chromosome 17 [dup(17)p11.2p12], as opposed to the submicroscopic and only molecularly detectable CMT1A duplication, was examined for signs of CMT1 with electrodiagnostic

1664

Diseases of the Peripheral Nervous System

techniques. In such individuals harboring a chromosomal abnormality, the clinical picture may typically include developmental delay, mental retardation, and a variety of somatic abnormalities. Additionally in this individual, the motor NCVs of all nerves tested were reduced to values between 17 and 20 m/s, consistent with demyelination as typically seen in CMT1.103 Similarly, a patient with trisomy for the entire short arm of chromosome 17, resulting from an unbalanced translocation inherited from her phenotypically normal balanced carrier father, also had demyelinating neuropathy or CMT1 as shown by electrophysiologic studies and a neurologic examination.27 Subsequent studies on other patients with cytogenetically visible duplication of proximal 17p, who manifested slowed motor NCVs as part of their complex clinical phenotype, enabled mapping of the genomic region containing the dosage-sensitive “CMT1A gene” to 17p12.80,100,138,155,197 Parallel studies in mice revealed that the Trembler (Tr) and TremblerJ (Tr J), both considered to be animal models for CMT1 and/or the more clinically severe demyelinating DSN, result from point mutations in the Pmp22 gene.181,184 Tr and Tr J both display the electrophysiologic features of symmetrically slowed motor NCV and the pathologic findings of demyelinating neuropathy, including onion bulbs; each maps to a region of mouse chromosome 11 that maintains conserved synteny with human 17p. Subsequently, the human PMP22 gene was mapped simultaneously by several groups to within the CMT1A duplication region.108,137,191,200 At that juncture, an apparent conundrum existed because the mouse phenotype was associated with a point mutation of Pmp22, whereas the human disease appeared to be due to increased dosage of a region in 17p12 that contained a presumably wild-type copy of PMP22. However, approximately 30% of CMT1 patients did not have the CMT1A duplication.210 Analysis of the PMP22 coding sequences in a nonduplication family linked to 17p12 markers,198 and in

a sporadic CMT1 family,154 identified a co-segregating PMP22 alteration and a de novo PMP22 mutation, respectively. The inherited PMP22 mutation (L16P) was identical to that found in the Tr J mouse.198 PMP22 point mutation was also identified in patients with DSN152; one DSN patient had a missense amino acid substitution (G150D) identical to that observed in the Tr mouse.69 Thus the clinical phenotype of CMT1 in patients with the CMT1A duplication appears to result from having three copies, or increased dosage, of the PMP22 gene. Alternatively, heterozygous gain-of-function PMP22 mutations can cause a CMT1 or DSN phenotype.152,198 The observation that PMP22 mutants have a prolonged association with the endoplasmic reticulum chaperone protein calnexin, possibly depleting the pool of calnexin and causing neuropathy, provides a potential mechanism for the gain of function of such mutations.36 In contrast, deletion of one copy of PMP22 as a result of the HNPP deletion, or heterozygous frameshift or nonsense loss-of-function mutations, result in a HNPP phenotype (see below). Substantive data support the notion that PMP22 is the dosage-sensitive gene responsible for CMT1A and HNPP (Table 70–2). Each of these findings is anticipated by the PMP22 gene dosage hypothesis: 1. In extremely rare neuropathy cases with smaller duplications68,131,199 or deletion,30 the PMP22 gene was encompassed within the alternative-sized rearrangement. 2. PMP22 messenger RNA (mRNA) was elevated in CMT1A duplication patients as determined by reverse transcriptase–PCR of sural nerve biopsies in comparison to patients with nonduplication neuropathies and also with another control myelin gene.214 3. PMP22 mRNA was shown to be decreased in patients with the HNPP deletion,165 and the PMP22 mRNA

Table 70–2. Evidence Supporting the Proposal That PMP22 Gene Dosage Causes Neuropathy Features Examined

CMT1A

HNPP

Neuropathy severity PMP22 maps within common rearrangement PMP22 within smaller or larger rearrangements PMP22 copy number (normal ⫽ 2) PMP22 mRNA PMP22 protein PMP22 point mutation Rodent models

HMZ ⬎ HTZ duplication Duplication, yes

NA Deletion, yes

Yes

Yes

3 Increased Increased Absent in duplication Pmp22 overexpression recapitulates CMT1A (HMZ ⬎ HTZ) Neuropathy included with overexpression; abrogated with return to normal expression

1 Decreased Decreased NA Pmp22 knockout recapitulates HNPP NA

Mouse model with regulated Pmp22 overexpression

HMZ ⫽ homozygous; HTZ ⫽ heterozygous; NA ⫽ not applicable.

HMSNs Involving Altered Dosage or Mutation of PMP22

4.

5.

6.

7.

8.

expression level correlated with the neuropathy disability score.164 Quantitative immunohistochemical43 and ultrastructural immunogold labeling201 showed increased PMP22 protein in CMT1A duplication patients and decreased PMP22 protein in HNPP deletion patients. DNA sequence analysis of the PMP22 gene in the youngest patient harboring the CMT1A duplication from each of seven different families, and in two sporadic CMT1A duplication patients, revealed an absence of any point mutations.205 Animal models, including two different transgenic mice and a rat model, that overexpress Pmp22 recapitulate the phenotypic features of CMT1,64,105,169 and the levels of Pmp22 expression correlate with the degree of demyelination and reduction in motor NCV.65,156,157 Consistent with PMP22 haploinsufficiency causing HNPP, nondeletion HNPP patients have nonsense and frameshift PMP22 mutations that are presumably null alleles.126,215 Heterozygous Pmp22⫹/⫺ (knockout) mice recapitulate the human HNPP phenotype with a progressive demyelinating tomaculous neuropathy that is more severe in the homozygous state,3,4 and a transgenic antisense mouse model also displays the clinical, electrophysiologic, and pathologic hallmarks of HNPP.109 A transgenic mouse model in which mouse Pmp22 overexpression can be regulated reveals that upregulation is followed by active demyelination that can be largely corrected when the Pmp22 gene expression is returned to normal levels.141

Given that gene dosage is the mechanism responsible for disease in patients with the CMT1A duplication, some of the clinical variability may be related to allelic variations that have been shown to occur in normal human gene expression.212 Other potential causes for clinical variability include modifier genes, particularly those important to transcriptional regulation of PMP22 or processing and degradation of PMP22 protein. Finally, environment or other stochastic factors may influence clinical expression, as evidenced by discordant phenotypes in identical twins with the CMT1A duplication.46 Each of these pathways provides a potential opportunity for intervention and therapy.

PMP22 POINT MUTATIONS AS THE CAUSE OF DEMYELINATING NEUROPATHIES In general, PMP22 point mutations cause more severe neuropathy than CMT1A duplication /HNPP deletion. The genetic mechanism by which PMP22 causes demyelinating neuropathy in CMT1A duplication and HNPP deletion patients is now firmly established as a gene dosage effect.

1665

Less well understood are (1) the exact function(s) of PMP22, and (2) the pathophysiologic mechanism by which either altered PMP22 dosage or point mutation results in demyelination. Ultrastructural studies that revealed uncompacted myelin lamellae in nerves obtained from patients with HNPP suggest a potential structural role for PMP22 in myelin compaction.213 An alternative proposed function is in the regulation of Schwann cell growth and differentiation.56,216 A dual role as a structural protein important to myelin compaction and as a regulatory protein important in myelin development and maintenance has also been proposed.120 It is clear that PMP22 mutations can have dramatic effects on myelin maintenance and/or development. The PMP22 gene contains five exons, including four coding exons; spans about 35 kb; and is regulated by two alternatively used promoters.21,183 It is expressed predominately in the peripheral nerve,21,137 but also in other non-neural tissues, during development and in the adult.9 Regulatory elements for the myelin-specific promoter52 and that which directs temporal and spatial expression in development and regeneration of peripheral nerves have been described.106 The promoter and /or these other regulatory elements, as well as nerve-specific mRNA transcripts, may be important as targets to modulate gene expression through either triplex formation,51 antisense, ribozyme, or small inhibitory RNA (siRNA) molecules. PMP22 expression can be stimulated with progesterone in cultured Schwann cells.35 In wild-type rats, exogenous administration of progesterone results in the upregulation of Pmp22 mRNA in sciatic nerves.170 A transgenic rat model has been constructed for the most common form of CMT, which results from the CMT1A duplication.169 This model overexpresses Pmp22 and recapitulates the clinical, electrophysiologic, and neuropathologic features of CMT1A. Recent exciting experiments show that administration of a selective progesterone receptor antagonist reduced overexpression of Pmp22 and improved the CMT phenotype.170 Importantly, there were no obvious side effects in either the transgenic rats or wild-type control animals. RNA interference (RNAi) is an evolutionarily conserved mechanism that directs the degradation of mRNAs homologous to short double-stranded RNA or siRNA molecules.41,110 RNAi has been widely used as a research tool to specifically ablate expression of a given gene and thus annotate gene function; it holds promise as a gene-specific therapeutic modality. CMT1A is an ideal candidate for siRNA therapy because (1) the phenotype results from altered gene dosage caused by overexpression of the PMP22 gene; (2) there is an mRNA transcript, the one that utilizes the PMP22 exon 1A Schwann cell–specific promoter, that is tissue and pathology specific; and (3) the haploinsufficiency phenotype is less severe than that which results from gene overexpression.

1666

Diseases of the Peripheral Nervous System

PMP22 protein comprises 2% to 5% of peripheral nervous system myelin132 and is found uniformly in compacted, but not in noncompacted, regions of myelin.57,132 The PMP22 protein is an integral membrane glycoprotein with four putative transmembrane domains. It has a mass of approximately 22 kDa (an 18-kDa core polypeptide and 4 kDa of carbohydrates linked to an asparagine residue). The glycosolated portion contains an L2/HNK-1 epitope.175,176 This epitope is implicated in intercellular recognition, interactions, and adhesion processes. PMP22 point mutations have been associated with CMT1, 154,198 DSN,152 HNPP, 126,215 and congenital hypomyelinating neuropathy (CHN)42,174 (Fig. 70–3). The majority of point mutations are associated with more severe disease (i.e., DSN or CHN). Mutations include mostly missense amino acid substitutions, but deletion of

a single amino acid, splice site mutations, and both frameshift and nonsense mutations have been reported (Fig. 70–3). As anticipated, frameshift and nonsense mutations, which usually represent null alleles and result in haploinsufficiency, cause a HNPP phenotype (Fig. 70–3). However, even identical mutations (e.g., Gly94fs) have been reported with a different phenotype, perhaps reflecting overlaps in clinical diagnosis and /or the fact that ultimate clinical outcome is dependent on more than just genetically determined mutation. Most missense mutations occur in amino acids comprising the putative transmembrane domains of this tetraspan protein, and greater than one quarter of the individual mutations reported are de novo and associated with sporadic disease. The overwhelming majority of mutations cause disease in the heterozygous state and thus are dominant alleles.

FIGURE 70–3 PMP22 neuropathy–associated mutations. A tetraspan model for the PMP22 protein is shown with the four transmembrane domains spanning the cell membrane. Individual amino acids are represented by oblong circles. The disease key above depicts the symbol associated with each specific clinical phenotype. Below are shown the specific mutations associated with the given phenotypes (see molgen-www.uia.ac.be/CMTMutations/DataSource/Mutations.cfm).

1667

HMSNs Involving Altered Dosage or Mutation of PMP22

However, three apparently recessive mutations, Thr118Met,121,123,153 Arg157Trp,136 and Arg157Gly,128 have been described. The exact mechanism by which PMP22 point mutations cause disease is an area of active research. Nevertheless, several studies suggest that impaired intracellular trafficking is a common disease mechanism for PMP22 point mutations in peripheral neuropathies.39,119,121 Furthermore, protein aggregates70,127,160,193 may be important in the neuropathology, and the ubiquitin/proteasome degradation pathway may be at least partially responsible for the regulation of PMP22 protein levels.12,72 Protein aggregates could also potentially be involved in the neuropathology observed secondary to gene dosage and PMP22 overexpression in patients with the CMT1A duplication. Thus aggresomes may be involved in the pathogenesis of CMT and related neuropathies in a manner similar to that observed in other neurodegenerative disorders.24,187 These neurodegenerative disorders illustrate the cumulative effects of subtle mutations, dosage, and/or improper protein degradation acting over prolonged time periods in cells that do not regenerate.

THE HNPP DELETION Genetic Basis of HNPP The genetic locus for HNPP maps to chromosome 17p12, where it is most commonly associated with a large (1.4-Mb) DNA deletion.26 As outlined above, a duplication of exactly this same 1.4-Mb region had been shown to be associated with CMT1A.99,144 In a study of 156 patients with HNPP, 84% were found to have the associated DNA deletion.124 Rarely, HNPP deletions30 and CMT1A duplications68,131,199 of a different size have been observed. As mentioned earlier, the 1.4-Mb region, as well as the deletions and duplications of “aberrant” size, contain the PMP22 gene. Furthermore, nine PMP22 sequence mutations that include two 2-bp deletions, one single base pair deletion, a single base pair insertion, three point mutations, and two splice site mutations have been reported in nondeletion HNPP patients.15,20,53,93,113,126,134,161,186,215

Clinical Findings in Patients with HNPP Deletion HNPP (OMIM #162500) is an autosomal dominant disorder that most often leads to episodic, painless, recurrent, focal motor and sensory peripheral neuropathies. HNPP is an entrapment or compressive neuropathy first described by De Jong in 1947.34 Episodes may be preceded by minor compression or trauma of the affected peripheral nerve. Vulnerable sites of nerve compression are the median

nerve at the wrist, the ulnar nerve at the elbow, and the peroneal nerve at the fibular head. The brachial plexus may also be affected in some cases. A history of limb trauma or prolonged positioning of the limb may be obtained in some cases. The onset of HNPP is usually in childhood or adolescence, with a high degree of penetrance. However, clinically asymptomatic obligate gene carriers are sometimes noted. Nerve palsies may be debilitating in that they may last for days to weeks and may require installation of a lower limb brace or ankle-foot orthosis in the cases of prolonged peroneal palsies. Hypoactive deep tendon reflexes and mild pes cavus are commonly observed in clinically asymptomatic patients. The spectrum of clinical presentation in HNPP is broad and may range from clinically asymptomatic, or subclinical, to more typical recurrent palsies and, in some advanced cases, progressive residual deficits mimicking indolent forms of CMT189,112,208 (Table 70–3). Specific guidelines for the diagnosis of HNPP have been reported previously.38 In an analysis of 39 HNPP patients from 16 unrelated pedigrees, two thirds of patients had the typical presentation of acute mononeuropathy and the remaining subjects had features consistent with a more long-standing polyneuropathy. Furthermore, it was noted that over 40% of affected persons were unaware of their illness and 25% of patients were essentially symptom free at the time of observation.133 More recently, the spectrum of clinical and neurophysiologic findings in 99 patients with HNPP was documented.116 The majority of patients in this survey (70%) presented with a typical history of a single, focal episode of neuropathy; however, there were patients with short-term and chronic sensory syndromes, as well as asymptomatic gene carriers. Rarely, patients with HNPP Table 70–3. Clinical Findings in 70 Patients with Typical HNPP Feature Total no. of patients Age at examination: mean (range) Age at onset: mean (range) Acute nerve palsies Positional sensory symptoms Painless symptoms History of compression Pes cavus Generalized areflexia Absent ankle jerks Neurologic sequelae

70 (40M/30F) 38 (14–73) 23 (7–73) 70 (100%) 8 (11%) 70 (100%) 46 (65%) 13 (18%) 9 (12%) 26 (37%) 9 (15%)

F ⫽ female; M ⫽ male. From Mouton, P., Tardieu, S., Gouider, R., et al.: Spectrum of clinical and electrophysiologic features in HNPP patients with the 17p11.2 deletion. Neurology 52:1440, 1999, with permission.

1668

Diseases of the Peripheral Nervous System

may present with a more fulminant course in which palsies affecting more than one limb are seen.31 Other rare associations with HNPP include a report of central nervous system demyelination.6 Very few studies have addressed the epidemiology of HNPP; one such study from western Finland reported a prevalence of 16 per 100,000.111 In 51 patients presenting with multifocal neuropathies, the HNPP deletion was detected in 24, including 19 patients from 14 of 15 families in whom the diagnosis of HNPP had been considered likely on clinical, electrophysiologic, and pathologic grounds.195 Seven of the 19 (37%) index patients with the HNPP deletion had no affected relatives, and fewer than half had evidence of a generalized neuropathy on examination.195 One patient had unusual clinical features presenting with a progressive scapuloperoneal syndrome. In one rare family, the HNPP deletion co-segregated for three generations with dominant carpal tunnel syndrome.143 However, HNPP is not a major cause of idiopathic carpal tunnel syndrome. The HNPP deletion was not identified in any of 50 unrelated patients with carpal tunnel syndrome presenting for surgery, 40% of whom had carpal tunnel syndrome diagnosed in a first-degree relative.177

Pathologic Features In 1972, Behse et al. documented the pathologic features of HNPP.10 Histologic assessment of sural nerve biopsies reveals segmental de- and remyelination. The presence of tomacula or “sausage”-shaped structures is the pathologic signature of HNPP. Tomacula consist of massive redundancy or overfolding of variable-thickness layers in the myelin sheath.104 It should be noted that tomacula are likely not a pathognomonic feature of HNPP. They may also be seen in CMT neuropathy type 4B (with myelin outfolding), immunoglobulin M paraproteinemic neuropathy, chronic inflammatory demyelinating polyneuropathy, and DSN.163 Tomacula have been observed also in cases of various CMT1 types, CMT1 associated with extrapyramidal signs, neurogenic scapuloperoneal syndrome, and multiple sclerosis.5,37 While tomacula are generally considered an important diagnostic feature of HNPP (Fig. 70–4), rare patients showing axonal regeneration and lacking tomacula have been observed.171

Electrophysiologic Features The neurophysiologic features of HNPP were described by Earl et al. and are consistent with demyelination, showing mildly prolonged motor and sensory nerve conduction velocities in a symmetrical, generalized pattern.40 Conduction blocks are characteristic of affected segments in symptomatic nerves, especially over entrapment sites. Mildly reduced NCVs and prolonged distal motor latencies are not restricted to those nerves affected by a palsy, but are found in a generalized pattern, even in asymptomatic gene

FIGURE 70–4 Cross section of a sural nerve biopsy from a patient with HNPP showing a tomaculum. Thinly sliced myelinated axons are also seen. One-micron section prepared from plastic embedded specimen, toluidine blue staining. (Original magnification, ⫻400.) (Courtesy of Dr. Zarife Sahenk, Department of Neurology, Ohio State University, Columbus, OH.) See Color Plate

carriers. While slowing of motor and sensory NCVs is commonly observed in hereditary demyelinating neuropathies, the NCVs of patients with HNPP are typically normal or only slightly prolonged. In HNPP the parameters most commonly affected include (1) increased distal motor latencies in the median nerves, (2) slowed sensory NCVs in median nerves at the wrists, and (3) increased distal motor latencies and mild slowing of NCVs in the peroneal nerves.116

HNPP Deletion Case Studies Patient 1 A 9-year-old female presented with an acute onset of left footdrop. Five years previously she had experienced right footdrop. Neurologic examination revealed weakness of dorsiflexion of the left foot and decreased motor NCV in the left peroneal nerve. Her neurologic examination was otherwise normal. She had made minimal recovery from her peroneal palsy when seen again 6 weeks later. Neither parent had signs of weakness, atrophy, or sensory disturbances, nor did they have a history of pressure palsies. A maternal grandfather had experienced an episode of bilateral footdrop in his 20s and presently carries a diagnosis of CMT neuropathy. DNA analysis on this child, her mother, and her maternal grandfather detected a deletion on chromosome 17p12 associated with HNPP. Patient 2 During weight training, a previously healthy 19-year-old male experienced a “crunch” and a brief sharp pain in his left shoulder. Over a 2-week period he developed weakness

HMSNs Involving Altered Dosage or Mutation of PMP22

and numbness of the fourth and fifth digits and ulnar aspect of the left hand. Neurologic exam confirmed sensory loss in the left hand and diffuse weakness in the left upper arm, consistent with a brachial plexopathy. Passive movement of the left arm produced pain in the shoulder. Nerve conduction studies disclosed prolonged distal latencies in the affected arm. There was no prior history of brachial plexopathy or other peripheral neuropathy in the patient or his family. The patient had partial recovery from his brachial plexopathy over a 3-month period, and underwent a reevaluation 1 year later because of persistent numbness of the left fifth digit. Neurologic examination was normal with the exception of diminished sensation in the ventral aspect of the left fifth finger. Repeated electrophysiologic studies found prolonged distal latencies of most sensory action potentials with normal amplitudes and conduction velocities. HNPP was suspected. DNA analysis on the patient and his father detected a deletion on chromosome 17p12, confirming the clinical diagnosis of HNPP. Patient 3 A 20-year-old male presented with a 2-year history of transient weakness of the left hand and acute onset of bilateral footdrop. Diffuse nerve conduction abnormalities were obtained and a diagnosis of isolated bilateral peroneal palsies was made. Subsequent clinical evaluation revealed normal strength, hypoactive deep tendon reflexes, pes cavus, enlarged ulnar nerves, and mild ichthyosis. Motor NCVs were mildly prolonged in multiple nerves. Both parents were normal with the exception of pes cavus in the patient’s mother. A diagnosis of X-linked CMT neuropathy with ichthyosis was suspected. DNA analysis on the patient and his parents detected a de novo deletion on chromosome 17p12, associated with the spontaneous occurrence of HNPP in this family.

Management There is no specific treatment for HNPP, and current therapy consists of conservative management and symptomeasing measures. The first, and perhaps most important, element of the treatment is early detection and diagnosis of the disease. The knowledge that HNPP is often triggered by compression or trauma of the peripheral nerves gives the patient the possibility to avoid those movements or joint positions that most often evoke an HNPP episode. Several genetic tools have been designed for a reliable diagnosis.92,145,172 These methods (see below) have been applied also to the prenatal diagnosis of CMT1A76,122; however, prenatal diagnosis is not applicable to HNPP given the mild phenotype. The second element of the treatment is conservative management to avoid evoking HNPP episodes.94 Overall, two basic rules apply: (1) excessive force or repetitive movements should be reduced to a minimum; and (2) extreme,

1669

awkward, or static joint positions should be avoided. Because the wrist, elbow, knee, and shoulder are the most vulnerable sites, special management rules can be considered. At the wrist (carpal tunnel), episodes are mostly caused by forceful gripping, repetitive movements, or extreme wrist bending. Tools and gloves can be used to improve grip, and wrist splints can prevent extreme wrist bend during the night. At the elbow, episodes are mostly triggered by repetitive or sustained bending and by habitual leaning on the elbows. Management consists of elbow pads to reduce pressure, a headset for use during long telephone conversations, and eventually an elbow splint during the night. The most frequent trigger at the knee is habitual crossing of the legs. A lower leg brace can stabilize ankles in case this causes recurrent stretch injury at the knee. The complexity of the brachial plexus means that many factors can evoke an HNPP episode at the shoulder. Avoiding overhead work or sleeping with the arms overhead, and maintaining a good posture, can reduce the risk. A single case report mentions surgery to reduce pressure on the ulnar nerve; however, a follow-up study is not yet available.185 Although various surgical treatments of carpal tunnel syndrome exist,47 no reports have been published on such surgery in HNPP patients. Given the vulnerability of the peripheral nerves in HNPP patients, surgery is generally considered unfavorable. Because HNPP is mainly caused by the dosage effect of PMP22 resulting from the 1.4-Mb deletion on chromosome 17p12, one could reason that increasing the gene dosage might rescue the defect. The reverse would be true for CMT1A patients, who carry a duplication increasing the PMP22 expression. However, several hurdles have to be overcome. First of all, increasing PMP22 expression requires either stimulation of the endogenous PMP22 or a gene transfer introducing another copy of the PMP22 gene. Further study of the PMP22 promoters and the recent discovery of a positive regulatory element of PMP22 could provide a future avenue for therapeutic research.52 A potential additional therapeutic approach is to increase PMP22 expression of the remaining nondeleted allele using a selective progesterone agonist.170

Decreased PMP22 Gene Dosage as the Cause of HNPP Because the majority of the HNPP patients carry the same 1.4-Mb deletion on chromosome 17p12, they are haploinsufficient for all expressed genes within this region, including PMP22. They are therefore predicted to have only 50% of the normal expression of the PMP22 gene in peripheral nerves, whereas CMT1A patients (discussed above), carrying the 1.4-Mb duplication, have a 50% increase in PMP22 expression. Various sural nerve biopsy studies showed that, in comparison to unaffected persons, PMP22 mRNA levels are increased /decreased in CMT1A/HNPP patients,

1670

Diseases of the Peripheral Nervous System

respectively.165,214 PMP22 appears to be the only gene within this 1.4-Mb segment a dosage effect of which is contributing to the HNPP/CMT1A phenotype, as evidenced by the identification of rare deletions and duplications of alternate size that did still include PMP22.30,68,199 The identification of different PMP22 mutations in patients without the HNPP deletion confirms the critical role of PMP22 in the HNPP phenotype. These include a 1-bp insertion, as well as a 1-bp and two different 2-bp deletions in PMP22 that cause frameshift mutations,15,93,126,186,215 two point mutations that create an early stop codon,53,186 and two splice mutations that are predicted to cause exon skipping, or aberrant splicing.20,113 Each of these mutations likely represents loss-of-function alleles either at the mRNA or protein levels. The pathogenic mechanism of these PMP22 mutations is that they likely mimic the dosage effect of the HNPP deletion by creation of a null allele. Thus far, only one HNPP-associated missense mutation has been reported. This Val30Met mutation is predicted to occur at the border of the first transmembrane domain and the extracellular loop.161 Patient nerve xenografts onto mice sciatic nerve showed a marked delay in the onset of myelination with an impairment of regenerative capacity and an increased neurofilament density, in comparison with nerve xenografts from unaffected individuals.161 These observations demonstrated the effect of the HNPP point mutation on the ability of Schwann cells to myelinate axons and pointed to perturbations in the axonal cytoskeleton resulting from a hypothesized interaction between the mutant Schwann cells and their axons. The function of PMP22 will need to be clarified in order to reconcile the various mutational mechanisms involving this myelin component in HNPP and other associated peripheral neuropathies.

Animal Models for HNPP According to the hypothesis that the pathologic mechanism of HNPP is a dosage effect, it may be speculated that total

absence of PMP22 protein would also cause demyelinating neuropathy, likely more severe than HNPP. Such patients have not been described, but a murine model with this genotype has been reported.3,4 Homozygous deletion (Pmp22⫺/⫺) knockout mice develop a severe demyelinating neuropathy with very slow conduction velocities (7 m/s) and striking demyelination on pathologic examination, including the evolution of demyelinated axons as tomaculous myelin sheaths are replaced by thin to absent ones. Heterozygous deletion (Pmp22⫹/⫺) mice are less affected, with minimal slowing of conduction velocities and little evidence of demyelination in biopsies, although numerous tomacula eventually develop in affected nerve. These data confirm that the loss of the Pmp22 gene product (i.e., a null allele), and not another gene in the 1.4-Mb deletion, is detrimental for HNPP. The precise mechanism of demyelination in Pmp22⫹/⫺ mice is unknown. It can be speculated that these mice synthesize only half as much Pmp22 protein as their wild-type counterparts, thereby altering the stoichiometry of the proteins in compact myelin, which in turn leads to demyelination. A naturally occurring bovine illness appears to resemble HNPP. This bovine neuropathy shows tomaculous neuropathy with the clinical features of weak shuffling gait, dysphagia, and chronic ruminal bloat.60

INHERITED BRACHIAL PLEXUS NEUROPATHIES HNPP may be clinically confused with hereditary neuralgic amyotrophy (HNA; also called familial brachial plexus neuropathy) (Table 70–4). Approximately 10% of HNPP patients have had involvement of the brachial plexus, and in some patients a brachial plexopathy is the initial or only expression of HNPP.130 Furthermore, both disorders were earlier proposed to represent the same condition, or allelic variants of the same locus,22,104,107 justifying the

Table 70–4. Differential Criteria for HNPP and HNA Neuropathy Localization Phenotype Preceding events Electrophysiology

Histopathology Genetics

HNPP

HNA

Generalized with focal episodes Mainly entrapment sites Painless Compression/trauma, infections Slowed nerve conduction velocities; conduction block over entrapment sites also measurable in unaffected nerves Tomacula De- and re-myelination 1.4-Mb deletion at 17p12, or deletion of aberrant size, or null mutations in PMP22

Focal episodes, not generalized Brachial plexus Painful Immunizations, puerperium, stress, exercise Normal to slightly reduced nerve conduction velocities in affected nerves, normal elsewhere Minor, focal axonal degeneration distal to affected brachial plexus, normal elsewhere Linked to 17q25 Genetic heterogeneity

HMSNs Involving Altered Dosage or Mutation of PMP22

need for differential diagnostic tools to distinguish between HNPP and HNA. HNA is an autosomal dominant disorder that causes episodes of paralysis and muscle weakness initiated by severe pain.208 Episodes are often triggered by infections, immunizations, the puerperium, and stress.114 Electrophysiologic studies in HNA show normal or mildly prolonged motor NCVs distal to the brachial plexus. A generalized neuropathy is not present in HNA. Pathologic studies have found axonal degeneration in nerves examined distal to the plexus abnormality.194 In some HNA pedigrees there are mild dysmorphic features of hypotelorism and short stature.71 The prognosis for recovery of normal function of affected limbs in HNA is good, although recurrent episodes can cause residual deficits. HNA is genetically linked to chromosome 17q25,139,178,207 although genetic heterogeneity has been reported.87,202,206 The underlying pathogenesis of HNA is unknown. Three studies have shown that HNPP and HNA are clinically, electrophysiologically, and genetically distinct diseases.28,49,209 Diagnostic guidelines for both HNPP and HNA have been published.38,88 The main differential criteria are summarized in Table 70–4.

STRUCTURE OF THE CMT1A/HNPP GENOMIC REGION The CMT1A duplication was originally estimated to represent a large DNA rearrangement based on the physical distance between genetic markers revealing RFLP dosage differences in CMT1 patients.63,99,145 Physical mapping using PFGE to separate large-molecular-weight DNA fragments demonstrated that the duplication is a tandem repeat of an approximately 1.4-Mb monomer unit resulting in an approximately 3.0-Mb rearrangement.140 Physical analysis also revealed proximal 17p region–specific, low-copy repeats that were labeled CMT1A-REPs.140 The proximal and distal CMT1A-REP repeats flank the 1.4-Mb region and are 24,011 bp in length and approximately 99% identical in sequence.82,150 Studies in nonhuman primates determined that the origin of the CMT1A-REP occurred during speciation between chimpanzee and gorilla, and that the distal copy of CMT1A-REP is the ancestral copy.17,81,150 Subsequently it was found that each CMT1AREP copy contains the same coding exon, exon VI of the human COX10 gene. COX10 encodes the protein hemeA:farnesyltransferase.150 However, the normal or wild-type COX10 gene traverses only the distal CMT1AREP, and the last coding exon of COX10 is located within the genomic region that is duplicated in patients with the CMT1A duplication.118,150 Analyses of proximal CMT1AREP and surrounding DNA sequences show that it likely arose by duplication of the distal CMT1A-REP and that the COX10 exon within it is a “pseudo-exon.”150

1671

Remarkably, the proximal CMT1A-REP contains the 3⬘ end of a gene termed HREP that is transcribed from the opposite strand of the open reading frame in the COX10 pseudo-exon VI,77 and the proximal CMT1A-REP insertion also disrupted an ancient gene, as can be inferred from mouse EST databases.66 Thus the CMT1A-REP genomic duplication that occurred during primate speciation has resulted in the evolution of two novel genes with distinct patterns of transcription.67 Complete nucleotide sequencing of the genomic segment duplicated in the CMT1A duplication and deleted in the HNPP deletion revealed that the segment is 1,421,129 bp in length, or approximately 1.4 Mb.66 Gene annotation identified 21 genes, including PMP22. Interestingly, a short tandem repeat (STR) consisting of 12 copies of the CAG trinucleotide is present in the 20-kb intron located between exons III and IV of PMP22. Whether expansion of this trinucleotide repeat could be involved in modifying disease through altering gene expression and subsequent PMP22 protein levels remains unanswered. However, in a rare family with CMT and deafness84 accompanied by anticipation,83 no expansion of this triplet repeat was identified in the more severely affected individuals from the most recent generation.66

GENERATION OF THE CMT1A DUPLICATION/HNPP DELETION BY UNEQUAL CROSSING OVER In the first molecular description of a sporadic CMT1 patient with a de novo CMT1A duplication, studies of marker genotypes revealed that the duplication arose by meiotic unequal crossing over.144 This unequal crossing over is mediated by the CMT1A-REP low-copy repeats.140 During meiosis, if the proximal copy of CMT1A-REP pairs with the distal copy from the other chromosome homologue, recombination between nonallelic CMT1A-REP copies results in duplication (Fig. 70–5). This unequal crossing-over model predicts a tandem CMT1A duplication structure with the presence of three copies of CMT1A-REP, including a recombinant CMT1A-REP copy in which the crossover occurred. The presence of three copies of CMT1A-REP in the CMT1A duplication was shown by Southern blot analysis,140 and the copies were visualized directly by FISH on stretched chromosome fibers.146 The tandem (or direct) nature of the CMT1A duplication was shown by both PFGE140 and FISH.200 Furthermore, the unequal crossing-over model predicted a reciprocal deletion of the same genomic region that is duplicated in CMT1A140 (Fig. 70–5). The HNPP deletion was proposed to represent the product of the reciprocal recombination that led to the CMT1A duplication.26 Identification of both the predicted junction fragment in HNPP patients by PFGE and a single copy of the

1672

Diseases of the Peripheral Nervous System

Proximal CMT1A-REP A

Distal CMT1A-REP B

C

D

PMP22 A'

A

B

C

B'

PMP22

B'

C'

C'

D'

D'

A'

D

PMP22 JCT

JCT

CMT1A duplication

HNPP deletion

FIGURE 70–5 Reciprocal recombination model for generating the CMT1A duplication and HNPP deletion. Unequal crossing over results in CMT1A duplication and HNPP deletion. The closed rectangles with different shades represent homologous CMT1AREP sequences with the proximal and distal copies as depicted. Unique sequences that flank the CMT1A-REP low-copy repeat are designated by letters (i.e., A, B, C, D) and those on the chromosome homologue as a prime (i.e., A⬘, B⬘, C⬘, D⬘). The filled rectangle located between the CMT1A-REP low-copy repeats represents the PMP22 gene. The X marks the crossover and the arrows depict the products of recombination, which include the CMT1A duplication and the HNPP deletion. Note that the CMT1A duplication has three copies of the CMT1A-REP, including a recombinant copy (dual shaded rectangles) at the recombination junction (JCT), whereas the HNPP deletion retains only one recombinant CMT1A-REP copy. Because the COX10 gene spans distal CMT1A-REP, the HNPP deletion results in disruption of one copy of COX10.148 Thus the potential exists, albeit likely rare, for a myopathy secondary to cytochrome oxidase deficiency if such a HNPP deletion–bearing individual carries a heterozygous point mutation in the nondeleted allele.

CMT1A-REP on the HNPP deletion chromosome supported this hypothesis.25,140,153 Observations regarding the dosage of CMT1AREP–specific restriction fragments suggested that the CMT1A duplication and HNPP deletion arose from recombination events within a limited region of the CMT1A-REP.25 Thus, despite approximately 99% sequence identity, strand exchanges associated with the unequal crossing-over event resulting in recombinant CMT1AREP, and leading to the CMT1A duplication and reciprocal product HNPP deletion, have a positional preference for the recombination event.25,29,55,82,90,95,149,192,204,211 Direct observation of human recombination products was obtained by DNA sequence analysis of the recombinant CMT1A-REP in multiple patients.96,148 The positional preference for unequal crossover within CMT1A-REP in both HNPP deletion patients148 and CMT1A duplication patients96 overlapped and occurred in a segment of approximately 500 bp; the recombination products observed were consistent with a double-strand break repair model for homologous recombination.96,148 Analyses of the crossovers also potentially delineate minimal efficient processing segments, of approximately 300 bp of identity, for strand exchanges in homologous recombination during meiosis in humans.148

Interestingly, the unequal nonsister chromatid exchange occurs preferentially (approximately 10:1) during spermatogenesis.131 Several hypotheses have been proposed to explain the preponderance of paternally derived de novo duplication. One recent hypothesis proposes that regional reduced meiotic recombination may lead to an increased propensity to generate unequal reciprocal recombination products.66,67

MOLECULAR DIAGNOSTICS FOR DETECTION OF CMT1A DUPLICATION/HNPP DELETION Presently, in the United States both PFGE and FISH are utilized in clinical diagnostics98,147,151,172 (Fig. 70–6). In other parts of the world, the molecular diagnosis is usually obtained by PCR detection of three alleles using multiple polymorphic STRs (microsatellites),8,16,32,99,168,179 or quantitative PCR to examine for dosage97,142,159,188 (Fig. 70–6). In each of these methods, one screens not only for the CMT1A duplication, but also for the HNPP deletion. Thus this molecular diagnostic test can be extremely useful. Molecular testing establishes a specific and secure diagnosis;

1673

HMSNs Involving Altered Dosage or Mutation of PMP22

Kb

600 550

Kb 820 770

HNPP del

CMT1A dup

Q-PCR

NML

NML

NML

HNPP del

Microsatellite

CMT1A dup

FISH

HNPP del

CMT1A dup

NML

PFGE

CMT1A DUP

500 HNPP DEL

FIGURE 70–6 Molecular methods to detect CMT1A duplication and HNPP deletion. Pulsed-field gel electrophoresis enables the separation of high-molecular-weight (30 kb to several megabases) genomic DNA. Normal individuals (NML) have 600-kb and 550-kb cross-hybridizing fragments by genomic Southern blotting analysis (left lane 1). CMT1A duplication patients have an additional 500-kb fragment that represents a junction fragment associated with the recombinant CMT1A-REP. HNPP deletion patients have a different junction fragment that appears as a split signal of 770 and 820 kb. Fluorescence in situ hybridization reveals an extra PMP22 signal (filled circle) in patients with the CMT1A duplication, whereas in HNPP one PMP22 copy is missing compared to the control probe (open circle). Note that normal individuals (NML) have two copies of both the control probe and PMP22, one from each chromosome 17 homologue. The microsatellite analysis reveals two alleles at this heterozygous locus, one from each chromosome 17 copy. CMT1A duplication patients are triallelic (i.e., have three copies), whereas HNPP deletion patients are monoallelic or hemizygous for polymorphic short tandem repeats (STR) or microsatellite markers within the rearranged genomic segment. Quantitative polymerase chain reaction (Q-PCR) enables quantification of PMP22 copy number differences in patients with the CMT1A duplication (three copies) versus those with the HNPP deletion (one copy) or normal individuals with two copies (one from each chromosome 17 homologue).

enables genetic counseling, including recurrence risk estimates; and provides for the possibilities of both prenatal75,76,122 and preimplantation33 diagnosis. A review of testing results from one commercial laboratory in the United States (Athena Diagnostics, Worcester, MA) that utilizes PFGE for CMT1A duplication/HNPP deletion testing (kindly provided by Drs. Uma Ananth and Aidan Hennigan) revealed that, during the last 6 years, approximately 5000 tests/year have been performed (N ⫽ 32,000). A positive result was obtained in about 22% of samples; the CMT1A duplication was detected almost four times as often as the HNPP deletion (3.8:1). Given the molecular mechanism of reciprocal recombination resulting in the CMT1A duplication or the HNPP deletion, one might expect the frequency of de novo duplication and deletion events to be equivalent. The observation of an apparent approximately 4:1 preference for duplication may reflect ascertainment bias given that HNPP is usually a milder phenotype than CMT1A. PFGE is a technically challenging method and, thus, approximately 7.7% of tests remained inconclusive and required resampling. In a small academic laboratory that utilizes FISH to detect the CMT1A duplication, during a comparable time span approximately 200 tests/year were performed. An abnormal or positive test result was obtained in about 28%

of samples. The CMT1A duplication was detected eight times as often as the HNPP deletion (data kindly provided by Drs. Sau W. Cheung and Pawel Stankiewicz of the Kleberg Cytogenetics Laboratory, Baylor College of Medicine, Houston, TX).

REFERENCES 1. Aarskog, N. K., Aadland, S., Gjerde, I. O., and Vedeler, C. A.: Molecular genetic analysis of Charcot-Marie-Tooth 1A duplication in Norwegian patients by quantitative photostimulated luminescence imaging. J. Neurol. Sci. 188:21, 2001. 2. Abe, Y., Ikegami, T., Hayasaka, K., et al.: Pressure palsy as the initial presentation in a case of late-onset CharcotMarie-Tooth disease type 1A. Intern. Med. 36:501, 1997. 3. Adlkofer, K., Frei, R., Neuberg, D. H., et al.: Heterozygous peripheral myelin protein 22-deficient mice are affected by a progressive demyelinating tomaculous neuropathy. J. Neurosci. 17:4662, 1997. 4. Adlkofer, K., Martini, R., Aguzzi, A., et al.: Hypermyelination and demyelinating peripheral neuropathy in Pmp22-deficient mice. Nat. Genet. 11:274, 1995. 5. Alexianu, M., Macovei, M., Alexianu, M. E., et al.: Tomaculous neuropathy with unusual clinical aspects. Rom. J. Neurol. Psychiatry 33:229, 1995.

1674

Diseases of the Peripheral Nervous System

6. Amato, A. A., and Barohn, R. J.: Hereditary neuropathy with liability to pressure palsies: association with central nervous system demyelination. Muscle Nerve 19:770, 1996. 7. Auer-Grumbach, M., Strasser-Fuchs, S., Wagner, K., et al.: Roussy-Levy syndrome is a phenotypic variant of CharcotMarie-Tooth syndrome IA associated with a duplication on chromosome 17p11.2. J. Neurol. Sci. 154:72, 1998. 8. Badano, J. L., Inoue, K., Katsanis, N., and Lupski, J. R.: New polymorphic short tandem repeats for PCR-based Charcot-Marie-Tooth disease type 1A duplication diagnosis. Clin. Chem. 47:838, 2001. 9. Baechner, D., Liehr, T., Hameister, H., et al.: Widespread expression of the peripheral myelin protein-22 gene (PMP22) in neural and non-neural tissues during murine development. J. Neurosci. Res. 42:733, 1995. 10. Behse, F., Buchthal, F., Carlsen, F., and Knappeis, G. G.: Hereditary neuropathy with liability to pressure palsies: electrophysiological and histopathological aspects. Brain 95:777, 1972. 11. Bellone, E., Mandich, P., Mancardi, G. L., et al.: CharcotMarie-Tooth (CMT) 1a duplication at 17p11.2 in Italian families. J. Med. Genet. 29:492, 1992. 12. Bence, N. F., Sampat, R. M., and Kopito, R. R.: Impairment of the ubiquitin-proteasome system by protein aggregation. Science 292:1552, 2001. 13. Bird, T. D., Kraft, G. H., Lipe, H. P., et al.: Clinical and pathological phenotype of the original family with CharcotMarie-Tooth type 1B: a 20-year study. Ann. Neurol. 41:463, 1997. 14. Birouk, N., Gouider, R., Le Guern, E., et al.: CharcotMarie-Tooth disease type 1A with 17p11.2 duplication: clinical and electrophysiological phenotype study and factors influencing disease severity in 119 cases. Brain 120(Pt. 5):813, 1997. 15. Bissar-Tadmouri, N., Parman, Y., Boutrand, L., et al.: Mutational analysis and genotype/phenotype correlation in Turkish Charcot-Marie-Tooth type 1 and HNPP patients. Clin. Genet. 58:396, 2000. 16. Blair, I. P., Kennerson, M. L., and Nicholson, G. A.: Detection of Charcot-Marie-Tooth type 1A duplication by the polymerase chain reaction. Clin. Chem. 41:1105, 1995. 17. Boerkoel, C. F., Inoue, K., Reiter, L. T., et al.: Molecular mechanisms for CMT1A duplication and HNPP deletion. Ann. N. Y. Acad. Sci. 883:22, 1999. 18. Boerkoel, C. F., and Lupski, J. R.: Hereditary motor and sensory neuropathies. In Rimoin, D. L., Connor, J. M., Pyeritz, R. E., and Korf, B. R. (eds.): Principles and Practices of Medical Genetics. London, Harcourt, p. 3303, 2002. 19. Boerkoel, C. F., Takashima, H., Garcia, C. A., et al.: Charcot-Marie-Tooth disease and related neuropathies: mutation distribution and genotype-phenotype correlation. Ann. Neurol. 51:190, 2002. 20. Bort, S., Nelis, E., Timmerman, V., et al.: Mutational analysis of the MPZ, PMP22 and Cx32 genes in patients of Spanish ancestry with Charcot-Marie-Tooth disease and hereditary neuropathy with liability to pressure palsies. Hum. Genet. 99:746, 1997. 21. Bosse, F., Zoidl, G., Wilms, S., et al.: Differential expression of two mRNA species indicates a dual function of peripheral

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

myelin protein PMP22 in cell growth and myelination. J. Neurosci. Res. 37:529, 1994. Bradley, W. G., Madrid, R., Thrush, D. C., and Campbell, M. J.: Recurrent brachial plexus neuropathy. Brain 98:381, 1975. Brice, A., Ravise, N., Stevanin, G., et al.: Duplication within chromosome 17p11.2 in 12 families of French ancestry with Charcot-Marie-Tooth disease type 1a. The French CMT Research Group. J. Med. Genet. 29:807, 1992. Bucciantini, M., Giannoni, E., Chiti, F., et al.: Inherent toxicity of aggregates implies a common mechanism for protein misfolding diseases. Nature 416:507, 2002. Chance, P. F., Abbas, N., Lensch, M. W., et al.: Two autosomal dominant neuropathies result from reciprocal DNA duplication/deletion of a region on chromosome 17. Hum. Mol. Genet. 3:223, 1994. Chance, P. F., Alderson, M. K., Leppig, K. A., et al.: DNA deletion associated with hereditary neuropathy with liability to pressure palsies. Cell 72:143, 1993. Chance, P. F., Bird, T. D., Matsunami, N., et al.: Trisomy 17p associated with Charcot-Marie-Tooth neuropathy type 1A phenotype: evidence for gene dosage as a mechanism in CMT1A. Neurology 42:2295, 1992. Chance, P. F., Lensch, M. W., Lipe, H., et al.: Hereditary neuralgic amyotrophy and hereditary neuropathy with liability to pressure palsies: two distinct genetic disorders. Neurology 44:2253, 1994. Chang, J. G., Jong, Y. J., Wang, W. P., et al.: Rapid detection of a recombinant hotspot associated with Charcot-MarieTooth disease type IA duplication by a PCR-based DNA test. Clin. Chem. 44:270, 1998. Chapon, F., Diraison, P., Lechevalier, B., et al.: Hereditary neuropathy with liability to pressure palsies with a partial deletion of the region often duplicated in Charcot-MarieTooth disease, type 1A. J. Neurol. Neurosurg. Psychiatry 61:535, 1996. Crum, B. A., Sorenson, E. J., Abad, G. A., and Dyck, P. J.: Fulminant case of hereditary neuropathy with liability to pressure palsy. Muscle Nerve 23:979, 2000. Cudrey, C., Chevillard, C., Le Paslier, D., et al.: Assignment of microsatellite sequences to the region duplicated in CMT1A (17p12): a useful tool for diagnosis. J. Med. Genet. 32:231, 1995. De Vos, A., Sermon, K., De Rijcke, M., et al.: Preimplantation genetic diagnosis for Charcot-Marie-Tooth disease type 1A. Mol. Hum. Reprod. 9:429, 2003. De Jong, J.: Over families met hereditaire dispositie tot het optreden van neuritiden gecorreleered met migraine. Psychiatr. Neurol. B (Amst.) 50:60, 1947. Desarnaud, F., Thi, A., Brown, A., et al.: Progesterone stimulates the activity of the promoters of peripheral myelin protein-22 and protein zero genes in Schwann cells. J. Neurochem. 71:1765, 1998. Dickson, K. M., Bergeron, J. J., Shames, I., et al.: Association of calnexin with mutant peripheral myelin protein-22 ex vivo: a basis for “gain-of-function” ER diseases. Proc. Natl. Acad. Sci. U. S. A. 99:9852, 2002. Drulovic, J., Dozic, S., Levic, Z., et al.: Unusual association of multiple sclerosis and tomaculous neuropathy. J. Neurol. Sci. 157:217, 1998.

HMSNs Involving Altered Dosage or Mutation of PMP22 38. Dubourg, O., Mouton, P., Brice, A., et al.: Guidelines for diagnosis of hereditary neuropathy with liability to pressure palsies. Neuromuscul. Disord. 10:206, 2000. 39. D’Urso, D., Prior, R., Greiner-Petter, R., et al.: Overloaded endoplasmic reticulum-Golgi compartments, a possible pathomechanism of peripheral neuropathies caused by mutations of the peripheral myelin protein PMP22. J. Neurosci. 18:731, 1998. 40. Earl, C. J., Fullerton, P. M., Wakefield, G. S., and Schutta, H. S.: Hereditary neuropathy, with liability to pressure palsies: a clinical and electrophysiological study of four families. Q. J. Med. 33:481, 1964. 41. Elbashir, S. M., Harborth, J., Lendeckel, W., et al.: Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411:494, 2001. 42. Fabrizi, G. M., Simonati, A., Taioli, F., et al.: PMP22 related congenital hypomyelination neuropathy. J. Neurol. Neurosurg. Psychiatry 70:123, 2001. 43. Gabriel, J. M., Erne, B., Pareyson, D., et al.: Gene dosage effects in hereditary peripheral neuropathy: expression of peripheral myelin protein 22 in Charcot-Marie-Tooth disease type 1A and hereditary neuropathy with liability to pressure palsies nerve biopsies. Neurology 49:1635, 1997. 44. Garcia, A., Combarros, O., Calleja, J., and Berciano, J.: Charcot-Marie-Tooth disease type 1A with 17p duplication in infancy and early childhood: a longitudinal clinical and electrophysiologic study. Neurology 50:1061, 1998. 45. Garcia, C. A.: Familial neuropathies. In Johnson, R. T., and Griffin, J. W. (eds.): Current Therapy in Neurologic Disease. St. Louis, Mosby–Year Book, p. 386, 1997. 46. Garcia, C. A., Malamut, R. E., England, J. D., et al.: Clinical variability in two pairs of identical twins with the CharcotMarie-Tooth disease type 1A duplication. Neurology 45:2090, 1995. 47. Gerritsen, A. A., Uitdehaag, B. M., van Geldere, D., et al.: Systematic review of randomized clinical trials of surgical treatment for carpal tunnel syndrome. Br. J. Surg. 88:1285, 2001. 48. Ginsberg, L., Malik, O., Kenton, A. R., et al.: Co-existent hereditary and inflammatory neuropathy. Brain 127:193, 2003. 49. Gouider, R., LeGuern, E., Emile, J., et al.: Hereditary neuralgic amyotrophy and hereditary neuropathy with liability to pressure palsies: two distinct clinical, electrophysiologic, and genetic entities. Neurology 44:2250, 1994. 50. Graf, W. D., Chance, P. F., Lensch, M. W., et al.: Severe vincristine neuropathy in Charcot-Marie-Tooth disease type 1A. Cancer 77:1356, 1996. 51. Hai, M., Bidichandani, S. I., Hogan, M. E., and Patel, P. I.: Competitive binding of triplex-forming oligonucleotides in the two alternate promoters of the PMP22 gene. Antisense Nucleic Acid Drug Dev. 11:233, 2001. 52. Hai, M., Bidichandani, S. I., and Patel, P. I.: Identification of a positive regulatory element in the myelin-specific promoter of the PMP22 gene. J. Neurosci. Res. 65:508, 2001. 53. Haites, N. E., Nelis, E., and Van Broeckhoven, C.: 3rd Workshop of the European CMT Consortium: 54th ENMC International Workshop on Genotype/Phenotype Correlations in Charcot-Marie-Tooth Type 1 and Hereditary Neuropathy with Liability to Pressure Palsies, 28–30 November 1997,

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

1675

Naarden, the Netherlands. Neuromuscul. Disord. 8:591, 1998. Hallam, P. J., Harding, A. E., Berciano, J., et al.: Duplication of part of chromosome 17 is commonly associated with hereditary motor and sensory neuropathy type I (Charcot-Marie-Tooth disease type 1). Ann. Neurol. 31:570, 1992. Han, L. L., Keller, M. P., Navidi, W., et al.: Unequal exchange at the Charcot-Marie-Tooth disease type 1A recombination hot-spot is not elevated above the genome average rate. Hum. Mol. Genet. 9:1881, 2000. Hanemann, C. O., and Muller, H. W.: Pathogenesis of Charcot-Marie-Tooth 1A (CMT1A) neuropathy. Trends Neurosci. 21:282, 1998. Haney, C., Snipes, G. J., Shooter, E. M., et al.: Ultrastructural distribution of PMP22 in Charcot-MarieTooth disease type 1A. J. Neuropathol. Exp. Neurol. 55:290, 1996. Harding, A. E., and Thomas, P. K.: Distal and scapuloperoneal distributions of muscle involvement occurring within a family with type I hereditary motor and sensory neuropathy. J. Neurol. 224:17, 1980. Hertz, J. M., Borglum, A. D., Brandt, C. A., et al.: CharcotMarie-Tooth disease type 1A: the parental origin of a de novo 17p11.2-p12 duplication. Clin. Genet. 46:291, 1994. Hill, B. D., Prior, H., Blakemore, W. F., and Black, P. F.: A study of pathology of a bovine primary peripheral myelinopathy with features of tomaculous neuropathy. Acta Neuropathol. (Berl.) 91:545, 1996. Holmberg, B. H., Holmgren, G., Nelis, E., et al.: CharcotMarie-Tooth disease in northern Sweden: pedigree analysis and the presence of the duplication in chromosome 17p11.2. J. Med. Genet. 31:435, 1994. Hoogendijk, J. E., Hensels, G. W., Gabreels-Festen, A. A., et al.: De-novo mutation in hereditary motor and sensory neuropathy type I. Lancet 339:1081, 1992. Hoogendijk, J. E., Hensels, G. W., Zorn, I., et al.: The duplication in Charcot-Marie-Tooth disease type 1a spans at least 1100 kb on chromosome 17p11.2. Hum. Genet. 88:215, 1991. Huxley, C., Passage, E., Manson, A., et al.: Construction of a mouse model of Charcot-Marie-Tooth disease type 1A by pronuclear injection of human YAC DNA. Hum. Mol. Genet. 5:563, 1996. Huxley, C., Passage, E., Robertson, A. M., et al.: Correlation between varying levels of PMP22 expression and the degree of demyelination and reduction in nerve conduction velocity in transgenic mice. Hum. Mol. Genet. 7:449, 1998. Inoue, K., Dewar, K., Katsanis, N., et al.: The 1.4-Mb CMT1A duplication/HNPP deletion genomic region reveals unique genome architectural features and provides insights into the recent evolution of new genes. Genome Res. 11:1018, 2001. Inoue, K., and Lupski, J. R.: Molecular mechanisms for genomic disorders. Annu. Rev. Genomics Hum. Genet. 3:199, 2002. Ionasescu, V. V., Ionasescu, R., Searby, C., and Barker, D. F.: Charcot-Marie-Tooth neuropathy type 1A with both duplication and non-duplication. Hum. Mol. Genet. 2:405, 1993.

1676

Diseases of the Peripheral Nervous System

69. Ionasescu, V. V., Searby, C. C., Ionasescu, R., et al.: Dejerine-Sottas neuropathy in mother and son with same point mutation of PMP22 gene. Muscle Nerve 20:97, 1997. 70. Isaacs, A. M., Jeans, A., Oliver, P. L., et al.: Identification of a new Pmp22 mouse mutant and trafficking analysis of a Pmp22 allelic series suggesting that protein aggregates may be protective in Pmp22-associated peripheral neuropathy. Mol. Cell. Neurosci. 21:114, 2002. 71. Jacob, J. C., Andermann, F., and Robb, J. P.: Heredofamilial neuritis with brachial predilection. Neurology 11:1025, 1961. 72. Johnston, J. A., Ward, C. L., and Kopito, R. R.: Aggresomes: a cellular response to misfolded proteins. J. Cell. Biol. 143:1883, 1998. 73. Kaku, D. A., Parry, G. J., Malamut, R., et al.: Nerve conduction studies in Charcot-Marie-Tooth polyneuropathy associated with a segmental duplication of chromosome 17. Neurology 43:1806, 1993. 74. Kaku, D. A., Parry, G. J., Malamut, R., et al.: Uniform slowing of conduction velocities in Charcot-Marie-Tooth polyneuropathy type 1. Neurology 43:2664, 1993. 75. Kashork, C. D., Chen, K. S., Lupski, J. R., and Shaffer, L. G.: Prenatal diagnosis of Charcot-Marie-Tooth disease type 1A. Ann. N. Y. Acad. Sci. 883:457, 1999. 76. Kashork, C. D., Lupski, J. R., and Shaffer, L. G.: Prenatal diagnosis of Charcot-Marie-Tooth disease type 1A by interphase fluorescence in situ hybridization. Prenat. Diagn. 19:446, 1999. 77. Kennerson, M. L., Nassif, N. T., Dawkins, J. L., et al.: The Charcot-Marie-Tooth binary repeat contains a gene transcribed from the opposite strand of a partially duplicated region of the COX10 gene. Genomics 46:61, 1997. 78. Killian, J. M., and Kloepfer, H. W.: Homozygous expression of a dominant gene for Charcot-Marie-Tooth neuropathy. Ann. Neurol. 5:515, 1979. 79. Killian, J. M., Tiwari, P. S., Jacobson, S., et al.: Longitudinal studies of the duplication form of Charcot-Marie-Tooth polyneuropathy. Muscle Nerve 19:74, 1996. 80. King, P. H., Waldrop, R., Lupski, J. R., and Shaffer, L. G.: Charcot-Marie-Tooth phenotype produced by a duplicated PMP22 gene as part of a 17p trisomy-translocation to the X chromosome. Clin. Genet. 54:413, 1998. 81. Kiyosawa, H., and Chance, P. F.: Primate origin of the CMT1A-REP repeat and analysis of a putative transposonassociated recombinational hotspot. Hum. Mol. Genet. 5:745, 1996. 82. Kiyosawa, H., Lensch, M. W., and Chance, P. F.: Analysis of the CMT1A-REP repeat: mapping crossover breakpoints in CMT1A and HNPP. Hum. Mol. Genet. 4:2327, 1995. 83. Kovach, M. J., Campbell, K. C., Herman, K., et al.: Anticipation in a unique family with Charcot-Marie-Tooth syndrome and deafness: delineation of the clinical features and review of the literature. Am. J. Med. Genet. 108:295, 2002. 84. Kovach, M. J., Lin, J. P., Boyadjiev, S., et al.: A unique point mutation in the PMP22 gene is associated with CharcotMarie-Tooth disease and deafness. Am. J. Hum. Genet. 64:1580, 1999. 85. Krajewski, K. M., Lewis, R. A., Fuerst, D. R., et al.: Neurological dysfunction and axonal degeneration in

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

Charcot-Marie-Tooth disease type 1A. Brain 123(Pt. 7):1516, 2000. Krampitz, D. E., Wolfe, G. I., Fleckenstein, J. L., and Barohn, R. J.: Charcot-Marie-Tooth disease type 1A presenting as calf hypertrophy and muscle cramps. Neurology 51:1508, 1998. Kuhlenbaumer, G., Meuleman, J., De Jonghe, P., et al.: Hereditary neuralgic amyotrophy (HNA) is genetically heterogeneous. J. Neurol. 248:861, 2001. Kuhlenbaumer, G., Stogbauer, F., Timmerman, V., and De Jonghe, P., on behalf of the European CMT Consortium: Diagnostic guidelines for hereditary neuralgic amyotrophy or heredofamilial neuritis with brachial plexus predilection. Neuromuscul. Disord. 10:515, 2000. Kumar, N., Muley, S., and Pakim, A. S.: Phenotypic variability in hereditary neuropathy with liability to pressure palsy (HNPP). Neurology 50:A73, 1998. LeGuern, E., Gouider, R., Lopes, J., et al.: Constant rearrangement of the CMT1A-REP sequences in HNPP patients with a deletion in chromosome 17p11.2: a study of 30 unrelated cases. The French CMT Collaborative Research Group. Hum. Mol. Genet. 4:1673, 1995. LeGuern, E., Gouider, R., Mabin, D., et al.: Patients homozygous for the 17p11.2 duplication in Charcot-MarieTooth type 1A disease. Ann. Neurol. 41:104, 1997. LeGuern, E., Ravise, N., Gouider, R., et al.: Microsatellite mapping of the deletion in patients with hereditary neuropathy with liability to pressure palsies (HNPP): new molecular tools for the study of the region 17p12 : p11 and for diagnosis. Cytogenet. Cell Genet. 72:20, 1996. Lenssen, P. P., Gabreels-Festen, A. A., Valentijn, L. J., et al.: Hereditary neuropathy with liability to pressure palsies: phenotypic differences between patients with the common deletion and a PMP22 frame shift mutation. Brain 121(Pt. 8):1451, 1998. Liebelt, J., and Parry, G.: Conservative management of hereditary neuropathy with liability to pressure palsies. Available at www.hnpp.org/index.htm Lopes, J., LeGuern, E., Gouider, R., et al.: Recombination hot spot in a 3.2-kb region of the Charcot-Marie-Tooth type 1A repeat sequences: new tools for molecular diagnosis of hereditary neuropathy with liability to pressure palsies and of Charcot-Marie-Tooth type 1A. French CMT Collaborative Research Group. Am. J. Hum. Genet. 58:1223, 1996. Lopes, J., Tardieu, S., Silander, K., et al.: Homologous DNA exchanges in humans can be explained by the yeast double-strand break repair model: a study of 17p11.2 rearrangements associated with CMT1A and HNPP. Hum. Mol. Genet. 8:2285, 1999. Lorentzos, P., Kaiser, T., Kennerson, M. L., and Nicholson, G. A.: A rapid and definitive test for Charcot-Marie-Tooth 1A and hereditary neuropathy with liability to pressure palsies using multiplexed real-time PCR. Genet. Test. 7:135, 2003. Lupski, J. R.: DNA diagnostics for Charcot-Marie-Tooth disease and related inherited neuropathies. Clin. Chem. 42:995, 1996. Lupski, J. R., de Oca-Luna, R. M., Slaugenhaupt, S., et al.: DNA duplication associated with Charcot-Marie-Tooth disease type 1A. Cell 66:219, 1991.

HMSNs Involving Altered Dosage or Mutation of PMP22 100. Lupski, J. R., and Garcia, A.: Charcot-Marie-Tooth peripheral neuropathies and related disorders. In Scriver, C. R., Beaudet, A. L., Sly, W. S., et al. (eds.): The Metabolic and Molecular Bases of Inherited Diseases. New York, McGraw-Hill, p. 5759, 2001. 101. Lupski, J. R., and Garcia, C. A.: Molecular genetics and neuropathology of Charcot-Marie-Tooth disease type 1A. Brain Pathol. 2:337, 1992. 102. Lupski, J. R., Pentao, L., Williams, L. L., and Patel, P. I.: Stable inheritance of the CMT1A DNA duplication in two patients with CMT1 and NF1. Am. J. Med. Genet. 45:92, 1993. 103. Lupski, J. R., Wise, C. A., Kuwano, A., et al.: Gene dosage is a mechanism for Charcot-Marie-Tooth disease type 1A. Nat. Genet. 1:29, 1992. 104. Madrid, R., and Bradley, W. G.: The pathology of neuropathies with focal thickening of the myelin sheath (tomaculous neuropathy): studies on the formation of the abnormal myelin sheath. J. Neurol. Sci. 25:415, 1975. 105. Magyar, J. P., Martini, R., Ruelicke, T., et al.: Impaired differentiation of Schwann cells in transgenic mice with increased PMP22 gene dosage. J. Neurosci. 16:5351, 1996. 106. Maier, M., Berger, P., Nave, K. A., and Suter, U.: Identification of the regulatory region of the peripheral myelin protein 22 (PMP22) gene that directs temporal and spatial expression in development and regeneration of peripheral nerves. Mol. Cell. Neurosci. 20:93, 2002. 107. Martinelli, P., Fabbri, R., Moretto, G., et al.: Recurrent familial brachial plexus palsies as the only clinical expression of ‘tomaculous’ neuropathy. Eur. Neurol. 29:61, 1989. 108. Matsunami, N., Smith, B., Ballard, L., et al.: Peripheral myelin protein-22 gene maps in the duplication in chromosome 17p11.2 associated with Charcot-Marie-Tooth 1A. Nat. Genet. 1:176, 1992. 109. Maycox, P. R., Ortuno, D., Burrola, P., et al.: A transgenic mouse model for human hereditary neuropathy with liability to pressure palsies. Mol. Cell. Neurosci. 8:405, 1997. 110. McManus, M. T., and Sharp, P. A.: Gene silencing in mammals by small interfering RNAs. Nat. Rev. Genet. 3:737, 2002. 111. Meretoja, P., Silander, K., Kalimo, H., et al.: Epidemiology of hereditary neuropathy with liability to pressure palsies (HNPP) in south western Finland. Neuromuscul. Disord. 7:529, 1997. 112. Meuleman, J., and Chance, P. F.: Hereditary pressure palsy. In Noseworthy, J. (ed.): Neurological Therapeutics: Principals and Practice. London, Martin Dunitz, p. 2198, 2003. 113. Meuleman, J., Pou-Serradell, A., Lofgren, A., et al.: A novel 3⬘-splice site mutation in peripheral myelin protein 22 causing hereditary neuropathy with liability to pressure palsies. Neuromuscul. Disord. 11:400, 2001. 114. Meuleman, J., Timmerman, V., Van Broeckhoven, C., and De Jonghe, P.: Hereditary neuralgic amyotrophy. Neurogenetics 3:115, 2001. 115. Morocutti, C., Colazza, G. B., Soldati, G., et al.: CharcotMarie-Tooth disease in Molise, a central-southern region of Italy: an epidemiological study. Neuroepidemiology 21:241, 2002.

1677

116. Mouton, P., Tardieu, S., Gouider, R., et al.: Spectrum of clinical and electrophysiologic features in HNPP patients with the 17p11.2 deletion. Neurology 52:1440, 1999. 117. Murakami, T., Oomori, H., Hara, A., et al.: Atypical phenotype of Charcot-Marie-Tooth disease type 1A. Muscle Nerve 22:1593, 1999. 118. Murakami, T., Reiter, L. T., and Lupski, J. R.: Genomic structure and expression of the human heme A:farnesyltransferase (COX10) gene. Genomics 42:161, 1997. 119. Naef, R., Adlkofer, K., Lescher, B., and Suter, U.: Aberrant protein trafficking in Trembler suggests a disease mechanism for hereditary human peripheral neuropathies. Mol. Cell. Neurosci. 9:13, 1997. 120. Naef, R., and Suter, U.: Many facets of the peripheral myelin protein PMP22 in myelination and disease. Microsc. Res. Tech. 41:359, 1998. 121. Naef, R., and Suter, U.: Impaired intracellular trafficking is a common disease mechanism of PMP22 point mutations in peripheral neuropathies. Neurobiol. Dis. 6:1, 1999. 122. Navon, R., Timmerman, V., Lofgren, A., et al.: Prenatal diagnosis of Charcot-Marie-Tooth disease type 1A (CMT1A) using molecular genetic techniques. Prenat. Diagn. 15:633, 1995. 123. Nelis, E., Holmberg, B., Adolfsson, R., et al.: PMP22 Thr(118)Met: recessive CMT1 mutation or polymorphism? Nat. Genet. 15:13, 1997. 124. Nelis, E., Van Broeckhoven, C., De Jonghe, P., et al.: Estimation of the mutation frequencies in Charcot-MarieTooth disease type 1 and hereditary neuropathy with liability to pressure palsies: a European collaborative study. Eur. J. Hum. Genet. 4:25, 1996. 125. Nicholson, G. A., Kennerson, M. L., Keats, B. J., et al.: Charcot-Marie-Tooth neuropathy type 1A mutation: apparent crossovers with D17S122 are due to a duplication. Am. J. Med. Genet. 44:455, 1992. 126. Nicholson, G. A., Valentijn, L. J., Cherryson, A. K., et al.: A frame shift mutation in the PMP22 gene in hereditary neuropathy with liability to pressure palsies. Nat. Genet. 6:263, 1994. 127. Notterpek, L., Ryan, M. C., Tobler, A. R., and Shooter, E. M.: PMP22 accumulation in aggresomes: implications for CMT1A pathology. Neurobiol. Dis. 6:450, 1999. 128. Numakura, C., Lin, C., Oka, N., et al.: Hemizygous mutation of the peripheral myelin protein 22 gene associated with Charcot-Marie-Tooth disease type 1. Ann. Neurol. 47:101, 2000. 129. Ohnishi, A., Li, L. Y., Fukushima, Y., et al.: Asian hereditary neuropathy patients with peripheral myelin protein-22 gene aneuploidy. Am. J. Med. Genet. 59:51, 1995. 130. Orstavik, K., Skard Heier, M., Young, P., and Stogbauer, F.: Brachial plexus involvement as the only expression of hereditary neuropathy with liability to pressure palsies. Muscle Nerve 24:1093, 2001. 131. Palau, F., Lofgren, A., De Jonghe, P., et al.: Origin of the de novo duplication in Charcot-Marie-Tooth disease type 1A: unequal nonsister chromatid exchange during spermatogenesis. Hum. Mol. Genet. 2:2031, 1993. 132. Pareek, S., Suter, U., Snipes, G. J., et al.: Detection and processing of peripheral myelin protein PMP22 in cultured Schwann cells. J. Biol. Chem. 268:10372, 1993.

1678

Diseases of the Peripheral Nervous System

133. Pareyson, D., Scaioli, V., Taroni, F., et al.: Phenotypic heterogeneity in hereditary neuropathy with liability to pressure palsies associated with chromosome 17p11.2–12 deletion. Neurology 46:1133, 1996. 134. Pareyson, D., and Taroni, F.: Deletion of the PMP22 gene and hereditary neuropathy with liability to pressure palsies. Curr. Opin. Neurol. 9:348, 1996. 135. Pareyson, D., Testa, D., Morbin, M., et al.: Does CMT1A homozygosity cause more severe disease with root hypertrophy and higher CSF proteins? Neurology 60:1721, 2003. 136. Parman, Y., Plante-Bordeneuve, V., Guiochon-Mantel, A., et al.: Recessive inheritance of a new point mutation of the PMP22 gene in Dejerine-Sottas disease. Ann. Neurol. 45:518, 1999. 137. Patel, P. I., Roa, B. B., Welcher, A. A., et al.: The gene for the peripheral myelin protein PMP-22 is a candidate for Charcot-Marie-Tooth disease type 1A. Nat. Genet. 1:159, 1992. 138. Pellegrino, J. E., Pellegrino, L., Spinner, N. B., et al.: Developmental profile in a patient with monosomy 10q and dup(17p) associated with a peripheral neuropathy. Am. J. Med. Genet. 61:377, 1996. 139. Pellegrino, J. E., Rebbeck, T. R., Brown, M. J., et al.: Mapping of hereditary neuralgic amyotrophy (familial brachial plexus neuropathy) to distal chromosome 17q. Neurology 46:1128, 1996. 140. Pentao, L., Wise, C. A., Chinault, A. C., et al.: CharcotMarie-Tooth type 1A duplication appears to arise from recombination at repeat sequences flanking the 1.5 Mb monomer unit. Nat. Genet. 2:292, 1992. 141. Perea, J., Robertson, A., Tolmachova, T., et al.: Induced myelination and demyelination in a conditional mouse model of Charcot-Marie-Tooth disease type 1A. Hum. Mol. Genet. 10:1007, 2001. 142. Poropat, R. A., and Nicholson, G. A.: Determination of gene dosage at the PMP22 and androgen receptor loci by quantitative PCR. Clin. Chem. 44:724, 1998. 143. Potocki, L., Chen, K. S., Koeuth, T., et al.: DNA rearrangements on both homologues of chromosome 17 in a mildly delayed individual with a family history of autosomal dominant carpal tunnel syndrome. Am. J. Hum. Genet. 64:471, 1999. 144. Raeymaekers, P., Timmerman, V., Nelis, E., et al.: Duplication in chromosome 17p11.2 in Charcot-Marie-Tooth neuropathy type 1a (CMT 1a). The HMSN Collaborative Research Group. Neuromuscul. Disord. 1:93, 1991. 145. Raeymaekers, P., Timmerman, V., Nelis, E., et al.: Estimation of the size of the chromosome 17p11.2 duplication in Charcot-Marie-Tooth neuropathy type 1a (CMT1a). HMSN Collaborative Research Group. J. Med. Genet. 29:5, 1992. 146. Rautenstrauss, B., Fuchs, C., Liehr, T., et al.: Visualization of the CMT1A duplication and HNPP deletion by FISH on stretched chromosome fibers. J. Peripher. Nerv. Syst. 2:319, 1997. 147. Ravise, N., Dubourg, O., Tardieu, S., et al.: Rapid detection of 17p11.2 rearrangements by FISH without cell culture (direct FISH, DFISH): a prospective study of 130 patients with inherited peripheral neuropathies. Am. J. Med. Genet. 118A:43, 2003.

148. Reiter, L. T., Hastings, P. J., Nelis, E., et al.: Human meiotic recombination products revealed by sequencing a hotspot for homologous strand exchange in multiple HNPP deletion patients. Am. J. Hum. Genet. 62:1023, 1998. 149. Reiter, L. T., Murakami, T., Koeuth, T., et al.: A recombination hotspot responsible for two inherited peripheral neuropathies is located near a mariner transposon-like element. Nat. Genet. 12:288, 1996. 150. Reiter, L. T., Murakami, T., Koeuth, T., et al.: The human COX10 gene is disrupted during homologous recombination between the 24 kb proximal and distal CMT1A-REPs. Hum. Mol. Genet. 6:1595, 1997. 151. Roa, B. B., Ananth, U., Garcia, C. A., and Lupski, J. R.: Molecular diagnosis of Charcot-Marie-Tooth disease type 1A and hereditary neuropathy with liability to pressure palsies. LabMedica Int. 12:22, 1995. 152. Roa, B. B., Dyck, P. J., Marks, H. G., et al.: Dejerine-Sottas syndrome associated with point mutation in the peripheral myelin protein 22 (PMP22) gene. Nat. Genet. 5:269, 1993. 153. Roa, B. B., Garcia, C. A., Pentao, L., et al.: Evidence for a recessive PMP22 point mutation in Charcot-Marie-Tooth disease type 1A. Nat. Genet. 5:189, 1993. 154. Roa, B. B., Garcia, C. A., Suter, U., et al.: Charcot-MarieTooth disease type 1A: association with a spontaneous point mutation in the PMP22 gene. N. Engl. J. Med. 329:96, 1993. 155. Roa, B. B., Greenberg, F., Gunaratne, P., et al.: Duplication of the PMP22 gene in 17p partial trisomy patients with Charcot-Marie-Tooth type-1 neuropathy. Hum. Genet. 97:642, 1996. 156. Robaglia-Schlupp, A., Pizant, J., Norreel, J. C., et al.: PMP22 overexpression causes dysmyelination in mice. Brain 125:2213, 2002. 157. Robertson, A. M., Perea, J., McGuigan, A., et al.: Comparison of a new Pmp22 transgenic mouse line with other mouse models and human patients with CMT1A. J. Anat. 200:377, 2002. 158. Ronen, G. M., Lowry, N., Wedge, J. H., et al.: Hereditary motor sensory neuropathy type I presenting as scapuloperoneal atrophy (Davidenkow syndrome): electrophysiological and pathological studies. Can. J. Neurol. Sci. 13:264, 1986. 159. Ruiz-Ponte, C., Loidi, L., Vega, A., et al.: Rapid real-time fluorescent PCR gene dosage test for the diagnosis of DNA duplications and deletions. Clin. Chem. 46:1574, 2000. 160. Ryan, M. C., Shooter, E. M., and Notterpek, L.: Aggresome formation in neuropathy models based on peripheral myelin protein 22 mutations. Neurobiol. Dis. 10:109, 2002. 161. Sahenk, Z., Chen, L., and Freimer, M.: A novel PMP22 point mutation causing HNPP phenotype: studies on nerve xenografts. Neurology 51:702, 1998. 162. Saifi, G. M., Szigeti, K., Snipes, G. J., et al.: Molecular diagnosis, mechanisms, and rational approaches to management and therapy for Charcot-Marie-Tooth and related neuropathies. J. Invest. Med. 51:261, 2003. 163. Sander, S., Ouvrier, R. A., McLeod, J. G., et al.: Clinical syndromes associated with tomacula or myelin swellings in sural nerve biopsies. J. Neurol. Neurosurg. Psychiatry 68:483, 2000. 164. Schenone, A., Nobbio, L., Caponnetto, C., et al.: Correlation between PMP-22 messenger RNA expression

HMSNs Involving Altered Dosage or Mutation of PMP22

165.

166. 167.

168.

169.

170.

171.

172.

173.

174.

175.

176.

177.

178.

179.

180.

and phenotype in hereditary neuropathy with liability to pressure palsies. Ann. Neurol. 42:866, 1997. Schenone, A., Nobbio, L., Mandich, P., et al.: Underexpression of messenger RNA for peripheral myelin protein 22 in hereditary neuropathy with liability to pressure palsies. Neurology 48:445, 1997. Scherer, S.: Axonal pathology in demyelinating diseases. Ann. Neurol. 45:6, 1999. Scherer, S. S., Kleopa, K. A., Chance, P. F., and Benson, M. D.: Peripheral neuropathies. In Rosenberg, R. N., Prusiner, S. B., Di Mauro, S., and Barchi, R. L. (eds.): The Molecular and Genetic Basis of Neurologic and Psychiatric Disease. Philadelphia, Butterworth Heinemann, p. 435, 2003. Seeman, P., Mazanec, R., Zidar, J., et al.: Charcot-MarieTooth disease type 1A (CMT1A) and hereditary neuropathy with liability to pressure palsies (HNPP): reliable detection of the CMT1A duplication and HNPP deletion using 8 microsatellite markers in 2 muliplex PCRs. Int. J. Mol. Med. 6:421, 2000. Sereda, M., Griffiths, I., Puhlhofer, A., et al.: A transgenic rat model of Charcot-Marie-Tooth disease. Neuron 16:1049, 1996. Sereda, M., Meyer Zu Horste, G., Suter, U., et al.: Therapeutic administration of progesterone antagonist in a model of Charcot-Marie-Tooth disease (CMT-1A). Nat. Med. 9:1533, 2003. Sessa, M., Nemni, R., Quattrini, A., et al.: Atypical hereditary neuropathy with liability to pressure palsies (HNPP): the value of direct DNA diagnosis. J. Med. Genet. 34:889, 1997. Shaffer, L. G., Kennedy, G. M., Spikes, A. S., and Lupski, J. R.: Diagnosis of CMT1A duplications and HNPP deletions by interphase FISH: implications for testing in the cytogenetics laboratory. Am. J. Med. Genet. 69:325, 1997. Silander, K., Meretoja, P., Nelis, E., et al.: A de novo duplication in 17p11.2 and a novel mutation in the P0 gene in two Dejerine-Sottas syndrome patients. Hum. Mutat. 8:304, 1996. Simonati, A., Fabrizi, G. M., Pasquinelli, A., et al.: Congenital hypomyelination neuropathy with Ser72Leu substitution in PMP22. Neuromuscul. Disord. 9:257, 1999. Snipes, G. J., Suter, U., and Shooter, E. M.: Human peripheral myelin protein-22 carries the L2/HNK-1 carbohydrate adhesion epitope. J. Neurochem. 61:1961, 1993. Snipes, G. J., Suter, U., Welcher, A. A., and Shooter, E. M.: Characterization of a novel peripheral nervous system myelin protein (PMP-22/SR13). J. Cell Biol. 117:225, 1992. Stockton, D. W., Meade, R. A., Netscher, D. T., et al.: Hereditary neuropathy with liability to pressure palsies is not a major cause of idiopathic carpal tunnel syndrome. Arch. Neurol. 58:1635, 2001. Stogbauer, F., Young, P., Timmerman, V., et al.: Refinement of the hereditary neuralgic amyotrophy (HNA) locus to chromosome 17q24-q25. Hum. Genet. 99:685, 1997. Stronach, E. A., Clark, C., Bell, C., et al.: Novel PCR-based diagnostic tools for Charcot-Marie-Tooth type 1A and hereditary neuropathy with liability to pressure palsies. J. Peripher. Nerv. Syst. 4:117, 1999. Sturtz, F. G., Latour, P., Mocquard, Y., et al.: Clinical and electrophysiological phenotype of a homozygously duplicated

181.

182. 183.

184.

185.

186.

187. 188.

189.

190.

191.

192.

193.

194.

195.

196.

197.

1679

Charcot-Marie-Tooth (type 1A) disease. Eur. Neurol. 38:26, 1997. Suter, U., Moskow, J. J., Welcher, A. A., et al.: A leucineto-proline mutation in the putative first transmembrane domain of the 22-kDa peripheral myelin protein in the trembler-J mouse. Proc. Natl. Acad. Sci. U. S. A. 89:4382, 1992. Suter, U., and Scherer, S. S.: Disease mechanisms in inherited neuropathies. Nat. Rev. Neurosci. 4:714, 2003. Suter, U., Snipes, G. J., Schoener-Scott, R., et al.: Regulation of tissue-specific expression of alternative peripheral myelin protein-22 (PMP22) gene transcripts by two promoters. J. Biol. Chem. 269:25795, 1994. Suter, U., Welcher, A. A., Ozcelik, T., et al.: Trembler mouse carries a point mutation in a myelin gene. Nature 356:241, 1992. Taggart, T. F., and Allen, T. R.: Surgical treatment of a tomaculous neuropathy. J. R. Coll. Surg. Edinb. 46:240, 2001. Taroni, F., Bottis, S., and Sghirlanzoni, A.: A nonsense mutation in the PMP22 gene in hereditary neuropathy with liability to pressure palsies (HNPP) not associated with the 17p11.2 deletion. Am. J. Hum. Genet. 57:A229, 1995. Taylor, J. P., Hardy, J., and Fischbeck, K. H.: Toxic proteins in neurodegenerative disease. Science 296:1991, 2002. Thiel, C. T., Kraus, C., Rauch, A., et al.: A new quantitative PCR multiplex assay for rapid analysis of chromosome 17p11.2–12 duplications and deletions leading to HMSN/HNPP. Eur. J. Hum. Genet. 11:170, 2003. Thomas, P. K., King, R. H., and Bradley, J. L.: Hypertrophic neuropathy: atypical appearances resulting from the combination of type I hereditary motor and sensory neuropathy and diabetes mellitus. Neuropathol. Appl. Neurobiol. 23:348, 1997. Thomas, P. K., Marques, W. Jr., Davis, M. B., et al.: The phenotypic manifestations of chromosome 17p11.2 duplication. Brain 120(Pt. 3):465, 1997. Timmerman, V., Nelis, E., Van Hul, W., et al.: The peripheral myelin protein gene PMP-22 is contained within the Charcot-Marie-Tooth disease type 1A duplication. Nat. Genet. 1:171, 1992. Timmerman, V., Rautenstrauss, B., Reiter, L. T., et al.: Detection of the CMT1A/HNPP recombination hotspot in unrelated patients of European descent. J. Med. Genet. 34:43, 1997. Tobler, A. R., Liu, N., Mueller, L., and Shooter, E. M.: Differential aggregation of the Trembler and Trembler J mutants of peripheral myelin protein 22. Proc. Natl. Acad. Sci. U. S. A. 99:483, 2002. Tsairis, P., Dyck, P. J., and Mulder, D. W.: Natural history of brachial plexus neuropathy: report on 99 patients. Arch. Neurol. 27:109, 1972. Tyson, J., Malcolm, S., Thomas, P. K., and Harding, A. E.: Deletions of chromosome 17p11.2 in multifocal neuropathies. Ann. Neurol. 39:180, 1996. Uncini, A., Di Muzio, A., Chiavaroli, F., et al.: Hereditary motor and sensory neuropathy with calf hypertrophy is associated with 17p11.2 duplication. Ann. Neurol. 35:552, 1994. Upadhyaya, M., Roberts, S. H., Farnham, J., et al.: CharcotMarie-tooth disease 1A (CMT1A) associated with a maternal

1680

198.

199.

200.

201.

202.

203.

204.

205.

206.

Diseases of the Peripheral Nervous System

duplication of chromosome 17p11.2 : 12. Hum. Genet. 91:392, 1993. Valentijn, L. J., Baas, F., Wolterman, R. A., et al.: Identical point mutations of PMP-22 in Trembler-J mouse and Charcot-Marie-Tooth disease type 1A. Nat. Genet. 2:288, 1992. Valentijn, L. J., Baas, F., Zorn, I., et al.: Alternatively sized duplication in Charcot-Marie-Tooth disease type 1A. Hum. Mol. Genet. 2:2143, 1993. Valentijn, L. J., Bolhuis, P. A., Zorn, I., et al.: The peripheral myelin gene PMP-22/GAS-3 is duplicated in CharcotMarie-Tooth disease type 1A. Nat. Genet. 1:166, 1992. Vallat, J. M., Sindou, P., Preux, P. M., et al.: Ultrastructural PMP22 expression in inherited demyelinating neuropathies. Ann. Neurol. 39:813, 1996. van Alfen, N., van Engelen, B. G., Reinders, J. W., et al.: The natural history of hereditary neuralgic amyotrophy in the Dutch population: two distinct types? Brain 123(Pt. 4):718, 2000. Vance, J. M., Nicholson, G. A., Yamaoka, L. H., et al.: Linkage of Charcot-Marie-Tooth neuropathy type 1a to chromosome 17. Exp. Neurol. 104:186, 1989. Vandenberghe, A., Latour, P., Chauplannaz, G., et al.: Molecular diagnosis of Charcot-Marie-Tooth 1A disease and hereditary neuropathy with liability to pressure palsies by quantifying CMT1A-REP sequences: consequences of recombinations at variant sites on chromosomes 17p11.2–12. Clin. Chem. 42:1021, 1996. Warner, L. E., Roa, B. B., and Lupski, J. R.: Absence of PMP22 coding region mutations in CMT1A duplication patients: further evidence supporting gene dosage as a mechanism for Charcot-Marie-Tooth disease type 1A. Hum. Mutat. 8:362, 1996. Watts, G. D., O’Briant, K. C., Borreson, T. E., et al.: Evidence for genetic heterogeneity in hereditary neuralgic amyotrophy. Neurology 56:675, 2001.

207. Wehnert, M., Timmerman, V., Spoelders, P., et al.: Further evidence supporting linkage of hereditary neuralgic amyotrophy to chromosome 17q. Neurology 48:1719, 1997. 208. Windebank, A. J.: Inherited recurrent focal neuropathies. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1137, 1993. 209. Windebank, A. J., Schenone, A., and Dewald, G. W.: Hereditary neuropathy with liability to pressure palsies and inherited brachial plexus neuropathy—two genetically distinct disorders. Mayo Clin. Proc. 70:743, 1995. 210. Wise, C. A., Garcia, C. A., Davis, S. N., et al.: Molecular analyses of unrelated Charcot-Marie-Tooth (CMT) disease patients suggest a high frequency of the CMTIA duplication. Am. J. Hum. Genet. 53:853, 1993. 211. Yamamoto, M., Yasuda, T., Hayasaka, K., et al.: Locations of crossover breakpoints within the CMT1A-REP repeat in Japanese patients with CMT1A and HNPP. Hum. Genet. 99:151, 1997. 212. Yan, H., Yuan, W., Velculescu, V. E., et al.: Allelic variation in human gene expression. Science 297:1143, 2002. 213. Yoshikawa, H., and Dyck, P. J.: Uncompacted inner myelin lamellae in inherited tendency to pressure palsy. J. Neuropathol. Exp. Neurol. 50:649, 1991. 214. Yoshikawa, H., Nishimura, T., Nakatsuji, Y., et al.: Elevated expression of messenger RNA for peripheral myelin protein 22 in biopsied peripheral nerves of patients with CharcotMarie-Tooth disease type 1A. Ann. Neurol. 35:445, 1994. 215. Young, P., Wiebusch, H., Stogbauer, F., et al.: A novel frameshift mutation in PMP22 accounts for hereditary neuropathy with liability to pressure palsies. Neurology 48:450, 1997. 216. Zoidl, G., Blass-Kampmann, S., D’Urso, D., et al.: Retroviral-mediated gene transfer of the peripheral myelin protein PMP22 in Schwann cells: modulation of cell growth. EMBO J. 14:1122, 1995.

71 Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations MICHAEL E. SHY

Historical Background The MPZ Molecule MPZ in Myelination MPZ as an Adhesion Molecule MPZ in the Developing Myelin Sheath Patients with MPZ Mutations Epidemiology and Genetics Early-Onset Neuropathy Late-Onset Neuropathy Classic Phenotype Phenotypic Variability

Pathology Early-Onset Disease Late-Onset Neuropathy Pathogenesis Role of Intracellular Trafficking Tertiary Structure of the Extracellular Domain Truncation of the Cytoplasmic Domain Associated with Early-Onset Neuropathy

HISTORICAL BACKGROUND In the late 19th century, Charcot and Marie23 in France, Tooth143 in England, and Friedreich42 in Germany described the clinical and pathologic features of patients with hereditary motor and sensory neuropathies (HMSNs). A clinically more severe neuropathy with onset in early childhood was subsequently described by two students of Charcot, Dejerine and Sottas,28 while a motor and sensory neuropathy associated with action tremor was later identified by Roussy and Lévy.114 Dawidenkow made the first systematic attempt to classify the inherited neuropathies in the late 1920s.24,25 However, it was not until 40 years later, following the development of electrophysiology, that the current system of classification was first developed. In 1968, Dyck and Lambert separated patients into hypertrophic and neuronal groups using mode of inheritance, class of neurons (axons) affected, pathologic alterations, and electrophysiologic features.31,32 Harding and Thomas,54,55 Dyck and Lambert,31,32 and others17,19,20 measured nerve conduction velocities (NCVs) in a group of patients with HMSN, and found that they could be divided into at least two groups, regardless of the genetic pattern of inheritance. One group had slow NCVs (⬍38 m/s),

Schwann Cell–Axon Interactions and Axonal Degeneration Axonal Degeneration in Late-Onset Neuropathy Differential Diagnosis Management Decisions about Genetic Testing Physiatry Pulmonary Management Pain Management Future Directions

suggesting a primary demyelinating process as the cause of their neuropathy (HMSN type I), while a second group had NCVs greater than 38 m/s, suggesting a primary axonal process as the cause of their neuropathy (HMSN type II). Bradley and colleagues argued for a third group with intermediate slowing of NCVs.17 Subsequent studies have shown that the neuropathy in most patients with HMSN I is caused by a duplication of a 1.4-Mb sequence on chromosome 17p11.2, in a region that includes the gene encoding the 22-kDa peripheral myelin protein (PMP22).87,141 However, HMSN I is a genetically heterogeneous disorder, with multiple causes, as was shown in early mapping studies that identified HMSN I loci on both chromosomes 1 and 17. Those cases mapped to chromosome 1 constitute the subject matter for this chapter. In 1962, Bird and co-workers evaluated a 21/2-year-old girl, and her two brothers, who presented with footdrop, distal leg weakness, and a loss of deep tendon reflexes. Motor NCVs were 5 to 15 m/s. The original handwritten pedigree, leading to one of the first linkage studies of neurologic disease, is shown in Figure 71–1. By 1982, these studies had revealed linkage to the Duffy (Fy) blood group gene localized on chromosome 1.10 In 1993, Hayasaka and colleagues localized the gene encoding the major 1681

1682

Diseases of the Peripheral Nervous System

THE MPZ MOLECULE

FIGURE 71–1 Original handwritten pedigree in 1962 from the medical record of the original HMSN IB family showing the index cases and the prior diagnosis of possible polio.7 (Courtesy of Tom Bird.)

peripheral nervous system (PNS) myelin protein P0 (MPZ) to a region on chromosome 1 near the Duffy locus, making MPZ a candidate gene for HMSN IB.57 Subsequent research identified point mutations in MPZ in families with linkage to Duffy, and MPZ mutations were confirmed to be the cause of the inherited demyelinating neuropathies collectively known as known as HMSN IB, or Charcot-Marie-Tooth disease type 1B (CMT1B).59 Many of the early reported cases with MPZ mutations had NCVs between 10 and 15 m/s, similar to those reported in the initial HMSN IB family. Thus it was hypothesized that HMSN IB patients could be distinguished from patients with HMSN IA by nerve conduction testing, because the more common HMSN IA patients typically have faster NCVs, in the range of 20 to 25 m/s.11,78,140 However, this hypothesis has not proved to be correct. Presently, more than 95 distinct MPZ mutations have been shown to cause neuropathy (see http://molgen-www.uia.ac.be/CMTMutations/), and many of the affected patients have widely disparate nerve conductions, as well as widely different clinical phenotypes. Some patients have even had normal or near-normal NCVs with reduced motor and sensory evoked amplitudes, leading to a clinical diagnosis of HMSN type II (or CMT2) despite the fact that their neuropathy is caused by a mutation in a myelin-specific gene.26 At present, a fundamental question is why some mutations cause severe neuropathies and other mutations cause milder or later onset symptoms. Prior to investigating these genotype-phenotype correlations, it will be helpful to review the MPZ molecule, which is responsible for the neuropathies.

MPZ is the major myelin protein expressed by Schwann cells, comprising approximately 50% of all PNS myelin proteins,34,49 and is necessary for both normal myelin function and structure.45 MPZ is a primordial protein in that central nervous system (CNS) myelin of bony fishes (Actinopterygyn) and cartilaginous fishes (Chondrichthyes) contains glycosylated components that are recognized by antibodies to bovine MPZ.41,149,150 MPZ is a component of PNS myelin in the human, bovine, pig, rabbit, guinea pig, rat, mouse, and chicken, demonstrating its role in both the mammalian and avian PNS (reviewed by Uyemura et al.148). The gene encoding MPZ in rats, mice,84 and humans57 is divided into six exons distributed over 7 kb of DNA,84 is located on 1q22-q23 of chromosome 1 in humans,57 and is expressed exclusively by myelinating Schwann cells. Thus MPZ is not found in other tissues, including CNS myelin.18,76 The six exons encode a protein of 248 amino acids; the 29–amino acid signal peptide encoded by exon 1 is cleaved prior to the insertion of the final 219–amino acid protein into the myelin sheath.84 Because the cleaved amino acids are not present in the myelin sheath, they are not included in the subsequent numbering of amino acids in this chapter. The mature protein has an extracellular domain (amino acids 1 through 124), a single transmembrane domain (amino acids 125 through 150), and a cytoplasmic domain (amino acids 151 through 219).84,147,148 The amino acid sequence of MPZ is more than 95% identical in human, rat, and cow.58,148 Newly translated MPZ is processed through the rough endoplasmic reticulum (ER) and Golgi apparatus prior to being transported to the myelin sheath. In the ER and Golgi apparatus it undergoes extensive posttranslational modifications, including glycosylation, sulfation, and acylation.30,33 The glycosylation consists of the addition of a single oligosaccharide, N-linked to Asn93 in the extracellular domain (Fig. 71–2). This oligosaccharide is complex, is sulfated,142 and contains the HNK-1 moiety, also found in PMP-22 and myelin-associated glycoprotein (MAG).111 This oligosaccharide is necessary for MPZ to function as an adhesion molecule38 (see below). Palmitate-rich fatty acids are presumably incorporated onto Cys153 in the Golgi apparatus. Fatty acids are incorporated into rat Mpz, in vivo, from radioactive palmitic acid.2 The relative molar ratio of palmitate, stearate, and oleate linked to Mpz is 8:3:2.116 The localization of fatty acids near the hydrophobic transmembrane domain suggests that there is a possible interaction between fatty acids and the adjacent membrane lipids.147 However, the exact reason for the acylation of MPZ remains unknown. The cytoplasmic domain of MPZ is also phosphorylated at several residues, including Ser197, Ser199, Ser204, and Ser21461 and Tyr191.67,68 Some of the phosphorylation presumably occurs prior to the incorporation of MPZ into the myelin sheath,155 but some also appears to occur in preexisting

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

1683

FIGURE 71–2 Mutations in the open reading frame of MPZ associated with inherited neuropathies. Adhesive interface, fourfold interface, and head-to-head interface refer to amino acid residues deemed essential for cis and trans adhesion between adjacent myelin wraps. The numbering system for MPZ mutations does not include the 29–amino acid leader peptide, which is cleaved before insertion in the myelin sheath. Clinical phenotypes are those utilized by the authors reporting the cases. (From Shy, M., Garbern, J., and Kamholz, J.: Hereditary motor and sensory neuropathies: a biological perspective. Lancet Neurol. 1:110, 2002, with permission.) See Color Plate

myelin.142 MPZ phosphorylation is thought to involve both protein kinase C (PKC) and calmodulin-dependant kinase.134 Specifically, Ser181, Ser204, and Ser214 are phosphorylated by PKC, and Ser181 and Ser204, but not Ser214, are phosphorylated by calmodulin-dependent protein kinase

in in vitro studies.134 The characterization of the role of PKC in mediating MPZ phosphorylation is based on sciatic nerve preparations incubated with phorbol esters, which increased incorporation of 32P—an increase that was blocked by PKC inhibitors.115 Neither incubation with forskolin nor

1684

Diseases of the Peripheral Nervous System

incubation with 8-bromo–cyclic AMP stimulated phosphorylation, making it unlikely that protein kinase A was contributing to the phosphate incorporation.115 Prolonged exposure of nerve segments to phorbol esters abolished the increased phosphorylation of MPZ, and immunoblots of nerve proteins revealed a decrease in the content of the PKC.115 Since the ␣ isoform is the most prevalent form of PKC in PNS myelin, which also contains ␤I, ␤II, ␦, and ⑀ isoforms,14,115 these data suggest that it is PKC␣ that is playing an important role in regulating the phosphorylation of MPZ. The significance of this regulation is discussed later in the chapter, in the section on MPZ-mediated adhesion.

MPZ in Myelination MPZ is expressed in myelinating Schwann cells as part of the coordinated program of myelin gene expression. As with other myelin-specific genes, MPZ expression is induced by as yet unidentified axonal signals.83 When axons have established a one-to-one relationship with developing Schwann cells, the Schwann cells increase messenger RNA (mRNA) expression of a series of genes, including MPZ, PMP22, myelin basic protein (MBP), and MAG, as well as the mRNAs encoding the cholesterol biosynthetic enzymes, hydroxymethylglutaryl–coenzyme A reductase and oleoyl–coenzyme A synthase, so that sufficient amounts of both myelin structural proteins and membrane constituents are available during myelin synthesis.44,82 Once myelination has been completed, its maintenance also depends on continued Schwann cell–axon interactions. If a peripheral nerve is cut, severing the axon and its Schwann cells below the cut from the neuronal cell body, axons degenerate and demyelination occurs, initiating the process of Wallerian degeneration. During Wallerian degeneration, myelinating Schwann cells change their pattern of gene expression, turning off expression of MPZ and the other myelin-specific genes, and turning on a series of genes, such as the low-affinity nerve growth factor receptor (NGFR), previously expressed prior to myelination in immature Schwann cells. If the nerve is crushed, however, allowing regeneration to occur after Wallerian degeneration, Schwann cell differentiation and myelination are restored as the axons regenerate through the crushed segment, recontacting denervated Schwann cells (reviewed by Kamholz et al.70). Northern blots from lesioned rat sciatic nerve illustrating this point are shown in Figure 71–3. The coordinate program of myelination also involves proper localization of molecules in the developing myelin sheath. As shown in the cartoon of a myelinated axon and its node of Ranvier in Figure 71–4, the mature myelin sheath has two regions, compact and noncompact, each of which contains a unique, nonoverlapping set of protein constituents. MPZ is normally targeted to the compact region along with PMP22 and MBP, which participate in forming the highly organized myelin sheath and in

electrically insulating axons. The noncompact region, which does not contain MPZ, is composed of two subdomains, the paranode and the juxtaparanode.

MPZ as an Adhesion Molecule MPZ is a member of the immunoglobulin supergene family, with a region of the extracellular domain similar in structure to a single immunoglobulin variable region domain.82 In vitro studies have shown that MPZ can act as a homophilic adhesion molecule,39 as is the case with other members of the immunoglobulin supergene family. Supporting this hypothesis, crystallographic analysis of the MPZ extracellular domain demonstrates that it interacts in cis to form homotetramers, which then interact in trans with tetramers of MPZ extracellular domains from the opposing myelin wrap to form the intraperiod line of compact myelin122 (Fig. 71–5). Cis interactions occur by the formation of a doughnut structure by four MPZ extracellular domains around a large central hole. The “fourfold interface” where these domains meet requires direct interaction between nine “critical” amino acids (see Figs. 71–2 and 71–5). Trans interactions between opposing MPZ molecules occur at two additional interfaces of the extracellular domain, termed putative adhesive interfaces (AIs) and head-to-head interfaces (HHIs). AIs and HHIs require specific interactions between 10 and 7 critical amino acids, respectively (Figs. 71–2 and 71–5).122 Taken together, these data thus suggest that MPZ plays an essential adhesive role in myelination by holding together adjacent wraps of myelin membrane through homotypic interactions of the extracellular domain. The highly basic intracellular domain of P0 is also necessary for mediating adhesion by the extracellular domain of the molecule. Mutations in this region have been shown to interfere with homotypic adhesion in vitro,40 and can cause particularly severe forms of demyelinating peripheral neuropathy in patients (see below).89,151 Moreover, coexpression of both wild-type and cytoplasmically truncated MPZ causes a loss of in vitro adhesion, presumably as a result of a dominant-negative effect of the truncated protein.145 Deletion of a 14–amino acid sequence in the cytoplasmic domain of MPZ also abolished adhesion in in vitro assays.159 Point mutations in this sequence that alter a PKC␣ substrate motif (RSTK, between amino acids 198 and 201) and mutation of an adjacent serine residue (204) also abolish adhesion in this system. Furthermore, PKC␣ and the activated C kinase receptor 1 are coimmunoprecipitated from cells expressing wild-type MPZ but not MPZ bearing a deletion of the 14 amino acids between residues 192 and 206. Taken together, these results indicate that PKC-mediated phosphorylation is an important component of the regulation of MPZ-mediated adhesion. A patient with a mutation of the initial arginine of the PKC substrate domain (R198S) has also been identified,

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

1685

(A)

FIGURE 71–3 Northern blot analysis of rat sciatic nerve mRNA. A, RNA isolated distal to a permanent transection. B, RNA isolated distal to a crush injury. The number of days after transection (T) or crush (C) is indicated; the zero time point is from unlesioned nerves. In crushed nerves, the distal nerve stumps were divided into proximal (P) and distal (D) segments of equal lengths. The blots were successfully hybridized with radiolabeled complementary DNA probes encoding P0 (MPZ), MBP, PMP22, and MAG, expressed by myelinating Schwann cells, and the low-affinity nerve growth factor receptor (NGFR), expressed by denervated Schwann cells. Twelve days after crush injury, regenerating axons have contacted denervated Schwann cells and reinduced myelin gene expression. By 24 days after crush, regenerating axons have reinduced myelin gene expression in Schwann cells in the distal portion of the distal nerve stump. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA levels were used as a loading control. (From Kamholz, J., Menichella, D., Jani, A., et al.: Charcot-Marie-Tooth disease type 1: molecular pathogenesis to gene therapy. Brain 123[Pt. 2]:222, 2000, with permission.)

(B)

suggesting that disruption of adhesion is associated with demyelination.159 The cytoplasmic domain of MPZ also plays an important role in mediating compaction of the Schwann cell cytoplasm, and the formation of the major dense line in PNS myelin. This role appears to require an interaction between MPZ and MBP, because mice doubly deficient in the genes for Mpz and Mbp lack major dense lines,92 while some Schwann cells form major dense lines in Mpz knockout mice.45 Moreover, the PNS is normal in shiverer mutant mice that cannot express Mbp.73 How Mpz effects the compaction of the cytoplasmic domain is not known. Early speculation hypothesized that basic residues in the cytoplasmic domain of Mpz might interact with phospholipids of adjacent cytoplasmic aspects of Schwann cell membranes.30,73

MPZ in the Developing Myelin Sheath In the developing myelin sheath, spiral wrapping of the inner mesaxon forms two to six wraps of loosely compacted myelin around the axon in which there is a 12- to 14-nm gap between the extracellular leaflets and the Schwann cell cytoplasm separating the cytoplasmic leaflets of the wraps.145 Most of these wraps then compact so that the gap between extracellular leaflets of the myelin wraps is reduced to 2.5 nm and the cytoplasmic leaflets appear fused.108 MPZ is very specifically targeted to the compact region of myelin, where it is likely to participate in the process of compaction. Alternatively, MAG, a large (100,000-kDa) transmembrane glycoprotein with five immunoglobulin-like domains in its extracellular domain,

1686

Diseases of the Peripheral Nervous System

FIGURE 71–4 Schematic view of myelinated axon and myelinating Schwann cell in which the localization of proteins associated with some of the causes of inherited neuropathies are illustrated at the site of their localization in the cell. The region enclosed in the rectangle on the bottom panel is shown in detail on the top. MPZ is localized to compact myelin, whereas connexin 32 (Cx32) is localized to the paranodal loops, incisures, and inner mesaxon composed of noncompact myelin. (Modified with permission from a figure of Steven Scherer.) See Color Plate

appears specifically targeted to regions of the developing myelin that have not yet been, or will not be, compacted.60,145 Consistent with this separate targeting of the two molecules, during myelin assembly, MPZ and MAG are sorted into specific carrier vesicles as they leave the trans-Golgi network in a microtubule-dependent manner, and then inserted into the appropriate membrane domain of the developing myelin sheath.144 If MPZ is targeted

incorrectly to noncompact myelin, myelination is drastically disrupted. For example, both spiral wrapping of axons by Schwann cells and radial sorting of Schwann cells by axons are disrupted in transgenic mice overexpressing Mpz, probably because the inner mesaxon of myelinating Schwann cells is inappropriately expressing Mpz.161 While it is clear that MPZ has an adhesive function, it may also play a regulatory role in myelination. In the

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

1687

FIGURE 71–5 Molecular associations observed in crystals of the MPZ extracellular domain. A, The position of the fourfold axis is illustrated by a square with each of the MPZ extracellular domains interacting in cis given a different color. B, A perpendicular view of the fourfold interface. The C-termini are indicated pointing down out of the doughnut, while the N-termini are found midway through it. The side chains of Trp28 and Trp78 are shown in yellow and are labeled for the green protomer; the tryptophan side chain that seems to come between these two is from the blue protomer. C, A twofold side-by-side interaction that is hypothesized to correspond to the adhesive interface (AI) between extracellular domains of opposing membranes. The orientation is shown such that the Prp28 side chains might be positioned to intercalate into the opposing membranes. The C-termini shown do not reach the level of these membranes, but there are five additional C0-terminal residues that are disordered in the structure. D, The head-to-head interface (HHI), also thought to be involved in trans interactions between opposing MPZ extracellular domains. Side chains are shown for Trp28. (From Shapiro, L., Doyle, J. P., Hensley, P., et al.: Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17:435, 1996, with permission.) See Color Plate

homozygous Mpz knockout mouse (Mpz⫺/Mpz⫺), in which the protein is not expressed, the program of myelin-specific gene expression in Schwann cells is disrupted. Mag and proteolipid protein 1 (Plp1) mRNA and protein levels are increased approximately sevenfold, Pmp22 and quaking 6

and 7 (Qki-6 and Qki-7) expression is absent, and Mbp levels are unchanged. In addition to changes in gene expression, the intracellular localization of Mag and E-cadherin are also abnormal in the loosely compacted nerves of Mpz knockout mice. Mag is distributed homogeneously

1688

Diseases of the Peripheral Nervous System

throughout the Schwann cell cytoplasm rather than becoming localized to nodes and paranodes. Similarly, neither E-cadherin nor its binding partner, ␤-catenin, are localized at their normal sites at paranodes, but are instead found in small puncta throughout the Schwann cell.97 Thus the absence of MPZ, but not other myelin proteins, leads to complex changes in the pattern of myelin-specific gene expression and myelin morphogenesis, suggesting that MPZ is playing an important role, either directly or indirectly, in the regulation of myelination. Whether the regulatory functions of MPZ are associated with its adhesive properties, however, remains unknown.

PATIENTS WITH MPZ MUTATIONS Epidemiology and Genetics MPZ mutations account for approximately 5% of HMSN type I cases.103 There is no ethnic or sex predisposition to develop HMSN IB. At present, all mutations that change or delete amino acids have caused neuropathy with the exception of Arg215Leu, which has been reported to cause a benign polymorphism102 (see http://molgen-www.uia.ac.be/ CMTMutations/). While MPZ mediated neuropathies are transmitted in an autosomal dominant pattern, a review of the literature reveals that 42 of 83 cases presented as sporadic disorders. Thus the lack of a family history does not exclude the diagnosis of HMSN IB. Patients with MPZ mutations have been reported with clinically severe, moderate, or mild neuropathies, depending at least in part on the particular mutation (see Fig. 71–2). However, it has been difficult to determine which mutations cause the different phenotypes for several reasons. One reason is phenotypic variability, in which the same mutation may cause different phenotypes. This important issue is discussed below after a review of the various phenotypes associated with particular mutations. However, three other issues complicate genotype-phenotype correlations in patients with MPZ mutations. First, the diagnosis HMSN IB has been interpreted differently by different authors: some have used the term historically to identify all patients with MPZ mutations7; others have used HMSN IB to identify patients with MPZ mutations and slow NCVs26; and still others have used the term to identify patients with MPZ mutations who presented with “typical” HMSN I phenotypes,151 as described by Harding and Thomas.54 Second, there has been confusion about the use of the diagnoses congenital hypomyelination (CH) and DéjérineSottas disease (DSD). The term congenital hypomyelination, which is pathologically based, originally hypothesized a developmental failure of PNS myelination.71,72,88 Patients were noted to be hypotonic within the first year of life, to have developmental delays in walking, and, in some cases, to

have swallowing or respiratory difficulties.50 Alternatively, the term Déjérine-Sottas disease, which is genetically based, was originally used to identify an autosomal recessive disorder presenting as a severe demyelinating neuropathy in infancy, often with delayed motor milestones such as walking. With the advent of genetic testing for mutations, it became apparent that patients diagnosed with both CH89,137,151 and DSD56,113,151 can have de novo MPZ mutations. Moreover, these neuropathies were often subsequently transmitted in an autosomal dominant fashion.81 Because these were therefore not recessively inherited disorders, it has become difficult to determine which diagnosis to give certain patients. DSD patients have very slow NCVs (⬍10 m/s),4 as do CH patients,49 and patients classified as having either CH or DSD50,105,106 have severe pathologic changes on sural nerve biopsies, as discussed later in this chapter. Moreover, the same patient has been classified as having CH135 and DSD.56 Third, there have not been clear limits as to what constitutes a HMSN II–like phenotype. Some patients with this diagnosis have had normal NCV, others have had some nerves with mildly slowed NCV, and some patients have had some nerves with slow NCV and others with normal NCV. In this chapter, patients are initially grouped according to the age of onset of their symptoms. Thus patients with the onset of neuropathy in infancy are grouped together without an attempt to distinguish between CH, DSD, and HMSN IB, and patients with late-onset disease are grouped independently of their NCV. Patients who develop symptoms similar to “classic” patients with HMSN I will be grouped together. In the classic HMSN I phenotype, as described by Harding and Thomas,54 patients reach early milestones such as walking at the normal time but develop distal weakness, decreased tendon reflexes, and foot deformities within the first two decades of life, typically around the onset of adolescence. Sensory loss may or may not be present.130 Classic phenotype patients frequently require the use of ankle-foot orthoses (AFOs) to ambulate as they get older, but only rarely require the use of a wheelchair or walker.11,78,140

Early-Onset Neuropathy Thirty-seven mutations that cause an early onset of signs and symptoms have been resported and are illustrated in red in Figure 71–6, and a summary of their clinical traits is provided in Table 71–1. Long-term follow-up was not provided for most patients, making a predictive natural history difficult to provide. Patients with the early onset of symptoms usually have severe neuropathies with very slow NCVs. Clinically, all but 7 of 37 reported cases were unable to walk prior to 18 months of age, and in the remaining 7 cases details of walking were not provided, although the clinical diagnosis was made prior to the age of 4 years. Five of the children were described as floppy or hypotonic prior to 9 months of age. All patients, with the exception of two children who died prior to 2 years of age, were eventually

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

1689

FIGURE 71–6 Mutations in the open reading frame of MPZ that are associated with early-onset (red) or late-onset (blue) symptoms of neurology. Early-onset cases are defined as those in which there was a delayed onset of walking or other early milestones. Late-onset cases are defined as those in which the first symptoms of neuropathy occurred after the age of 21. (Modified from Shy, M. E., Jani, A., Krajewski, K, et al.: Phenotypic clustering in MPZ mutations. Brain 127:371, 2004.) See Color Plate

able to ambulate. Typically, affected children were able to walk by the age of 5 years, but rarely walked normally, and in many cases could never run. Two patients, with different mutations, did not walk until 10 and 12 years of age. At least six patients ultimately required the use of a wheelchair; in four this was prior to the age of 25. However wheelchair use was not invariable, and several patients

were able to ambulate into their ninth decade with the aid of a cane.9 Despite the severity of the neuropathy, life expectancy did not appear to be reduced in most patients with early-onset disease. The two patients who died in childhood did so because of respiratory compromise.43,65,96 The most likely outcome appeared to be a normal lifespan requiring significant assistance in ambulation.

DSD

CMT1B CMT1B CMT1B CMT1B CMT1B/DSD

Gln56_Ile60del/ insHisLeuPhe Asp61Glu Gly64Glu Arg69Ser Arg69Cys Arg69Cys

DSD

DSD DSD

DSD

DSD/CH

Ile85Thr||

Asn87His|| Asp89Tyr90del/ insPheTyr Thr95Phe96del

Cys98Tyr

Gly74Glu

DSD

CMT1B DSD

CMT1B

Abnl gait at age 7 yr Toe walking Delayed Hypotonia

CMT1B/DSS EO-CMT1 Severe CMT1B CMT1B/DSD DSD

Phe35del§ Tyr39Cys Ser49Leu His52Arg Tyr53Cys

Arg69Cys Arg69His Try72Cys Val73fs§

Infantile

DSD/CH

Ser34Cys

Delayed at 7 mo, dysphagia

Floppy infant, sat at 2 yr

Severe at 10 mo Delayed Age of walking Delayed sitting prior to 1 yr Delayed sitting prior to 1 yr Hypotonia

Age of walking Prior to age 1 yr Delayed Delayed at 9 mo

Age of walking

Prior to age 3 yr

DSD CMT1B DSD CMT1B

Ile1Met Thr5Ile Ile33Phe Ser34Phe

First Limitations

Author Diagnosis*

Amino Acid Change

10 yr

Never

⬎2 yr

4–5 yr

3.5 yr

Delayed

Never

12 mo but falls frequently Delayed in 6 pts 18 mo 3 yr 2 yr 2 yr

20 mo 20 mo

18 mo

3 yr

3 yr 3 yr 18–24 mo

Age Walked

Table 71–1. Features of Early-Onset Neuropathy (Literature Review)

Proximal-distal weakness at 13 yr

Died at 10 mo

WCB at 10 yr

Severe, WCB

Severe, WCB

Died at 22 mo

Abnl gait Abnl gait Abnl gait

Moderate

Never ran Severe at 25 yr

Worst possible clinical score Severe Severe

Severe at ⬎26 yr

AFO at 5 yr

Clinical Course†

U 3.8, P NR, S NR

P NR, M “delayed” M 6, T 6

Slowed

NR

M 10 NR

5–15 M 6, U 5 U 13, M 11 M7 P 7.5, S 5.1, M 8, T 4.9 U 8.5

Delayed NR U 4.7, M 6.2, T 3.3 M 10

M 13

M8

NCV‡ (m/s)

Sporadic

Mother (germline mosaicism)

Sister

Daughter similar Brother, parents

4 generations Dominant

Sporadic

Father

Siblings Mother (didn’t walk ’til 12)

Family History

35

117

152 65

152

36

151

43 96 81 107

9 63 151 151 77

128

43 131 128 123

106, 56

56 43 101 12

Reference

RLS

CMT1B/DSD DSD

CMT1B DSD

EO-CMT1 DSD

DSD DSD

CH

DSD

Asn102Lys

Ile106Leu Ile106Thr

Val117Phe Leu137fs222X

Gly138Ala Gly138Arg

Leu145fs Ala159fs206X

Glu186X

Ala192fs

Delayed at 6 mo Floppy infant at 6 mo, delayed Delayed at age 1 yr, hypotonia Boy at age 4 yr

Unsteady gait Delayed motor milestones Early onset Delayed motor development

Delayed development

Sat at 18 mo

3.5 yr

2–3 yr 4 yr

3 yr

30 mo

18 mo

2.5 yr 2 yr

5 yr

⬎20 mo

2.5 yr

Severe WCB at 18 yr, resp support WCB at 20 yr, resp distress Head control at 1 yr Severe at 8 yr (neonatal hypoxia)

Age 6 yr walks short distance Severe at 15 yr WCB at 25 yr

Severe at later age

WCB at 10 yr WCB at 7 yr Severe

15

M 3.6, P 4.2

Abnormal NR

U 8, P NR M5 U6

M7

⬍16

M 11 M 10 ⬍10

Sporadic

Sporadic

Sporadic Germline mosaicism Sister of above

Sporadic

Sporadic

Sister, mother, grandmother 5 of 7 family members

*Ch ⫽ congenital hypomyelination; CMT ⫽ Charcot-Marie-Tooth disease; DSD ⫽ Déjérine-Sottas disease; EO-CMT ⫽ early-onset CMT; RLS ⫽ Roussy-Lévy syndrome. † Abnl ⫽ abnormal; AFO ⫽ ankle-foot orthosis; resp ⫽ respiratory; WCB ⫽ wheelchair bound. ‡ Nerve conduction velocity: M ⫽ median; NR ⫽ nonrecordable; P ⫽ peroneal; S ⫽ sural; T ⫽ tibial; U ⫽ ulnar. § Homozygous mutation. || All three mutations in the same patient.

DSD DSD DSD CMT1

Asp99Asn|| Lys101Arg Lys101Arg Lys101Arg

112

89

151 137

138

131 138

146 104 146

43

109

152 43 43 136

1692

Diseases of the Peripheral Nervous System

While most early-onset patients had severe neuropathies, there has been variability in disability. For example, one patient who died, who had an Arg69Cys mutation, developed weakness at 6 months of age and a head lag at 14 months, could not cough, and cried without sound. She lost the ability to roll over, developed a flaccid quadriparesis, and died at the age of 22 months.43,96 More typical is the original MPZ family, in which atrophy of legs and intrinsic hand muscles developed prior to adolescence in most patients. Several affected members of this family developed proximal as well as distal leg weakness, and also developed large and small fiber sensory loss prior to adolescence.9 Only one patient developed what ultimately was characterized as a mild neuropathy.90 Virtually all the early-onset cases were areflexic in childhood. Only occasional patients have been reported with palpably enlarged nerves.104 Cerebrospinal fluid (CSF) protein levels have been reported to be as high as 167 mg/dL in a patient with a Leu172fs mutation137 and 169 mg/dL in a patient with an Arg69Cys mutation,151 though another patient with an Ar69Cys mutation had a CSF protein of only 50 mg/dL.77 Nerve Conduction Velocities Motor nerve conduction velocities (MNCVs) in earlyonset cases were below 15 m/s in the upper extremities for all but two patients, and in most patients were less than 10 m/s (see Table 71–1). Compound muscle action potentials (CMAPs) and sensory nerve action potentials (SNAPs) were reduced in all patients. Despite severe disability in the lower limbs, several patients had obtainable peroneal, tibial, or sural conduction velocities, all of which were similar to NCVs in their upper extremities. These data, taken together, demonstrate a severe, uniform slowing of NCV in nerves of early-onset patients with secondary axonal degeneration. Temporal dispersion and/or conduction block were not identified in any patient.

Late-Onset Neuropathy Twenty-one mutations causing a late onset of signs and symptoms are illustrated in blue in Figure 71–6, and a summary of their clinical traits is provided in Table 71–2. Most patients with late-onset disease developed symptoms around or after their fourth decade of life. Only occasional late-onset patients developed symptoms before age 30, and in many of these cases the patients were outliers with respect to others affected with the same mutation. For example, two sisters in a three-generation French family with a Thr95Met mutation developed symptoms of neuropathy at ages 18 and 20. However, the other five affected family members all developed symptoms after 30,21 and in 13 additional families with the same mutation, no patient developed symptoms until at least the age of 30.26,99 The presence of an adult-onset neuropathy has not necessarily implied that the disease will be mild. Six patients

with Thr95Met mutations became wheelchair bound in one series.26 In a second series with the same mutation, many patients developed difficulties walking by 50 years of age, and one patient required the use of a wheelchair by age 49. This patient subsequently died at 52 because of an aspiration pneumonia related to swallowing dysfunction. Some late-onset patients were asymptomatic at the time of diagnosis. Ikegami et al.64 reported a child with a homozygous Ser35del, in which both parents were unaware that they had neuropathy. The parents were first cousins and were completely asymptomatic. On neurologic examination, the father could not walk on his heels and had absent reflexes and moderate pes cavus. However, he had no weakness or muscle wasting. The parents of a patient with a Gly74fs were also asymptomatic.107,151 At age 34, the mother was noted to have normal muscle bulk and strength but had bilateral pes cavus with varus deformity. The father had mild atrophy of the gastrocnemius muscles and a distal loss of light touch, pain, and vibration sensation in his legs. The mother had absent reflexes in the lower extremities, and the father was areflexic. One patient was atypical for the late-onset group because she developed symptoms at a younger age, and had slower NCVs than the others. She also had a rapid increase in disability. This 26-year-old woman, with an Arg69His mutation, first developed symptoms of neuropathy at age 23, was noted to have peroneal and intrinsic hand atrophy at 26, and required foot surgery by 27. Her ulnar MNCV was 16.4 m/s. No further follow-up is provided, but her father was affected by a similar “severe” phenotype though the proband was classified as “moderate.”162 Nerve Conduction Velocities MNCV were greater than 20 m/s in the upper extremities for all but the one patient mentioned above; in many cases the velocities were above 30 m/s, and in some cases they were above 50 m/s (see Table 71–2). Some patients had NCVs in the normal or near-normal range, leading to their characterization as having HMSN II, or “CMT2,” despite the fact that MPZ is expressed by Schwann cells, not neurons.26,99 However, these patients usually had some slowing of NCV, although an occasional value was in the normal range. CMAPs and SNAPs were typically reduced, even in patients with normal or near-normal conduction velocities. Taken together, these results suggest that mutations causing late-onset neuropathy may disrupt interactions between Schwann cells and axons but only cause minimal or mild effects on normal saltatory conduction.

Classic Phenotype Occasional patients have been described with MPZ mutations and classic HMSN I phenotypes. These patients had normal early milestones prior to developing symptoms and signs of neuropathy during adolescence. Young et al.162 reported a Glu112stop mutation in 11- and 16-year-old

Normal

CMT1B CMT

CMT CMT2 CMT Normal

Age 27 Age 38 Age 45

Age 30

Age 40s

CMT2

Mild Mild Mild at 49

Mild to moderate

Severe Normal Mild Mild at 53

Mild Mild Severe at 72 Moderate Severe Severe

Mild Mild

Moderate to severe

Mild

Severe in 2 patients

Clinical Course

*CMT ⫽ Charcot-Marie-Tooth disease. † Nerve conduction velocity: M ⫽ median; P ⫽ peroneal; S ⫽ sural; T ⫽ tibial; U ⫽ ulnar. ‡ Homozygous mutation.

Gly134Arg Arg198Ser Lys207Del

Ser111Thr Glu112X

Age 44 Age 37–56

Normal Normal

Normal Normal Normal Normal

Age 18–50

CMT1B CMT1B CMT1B CMT1B CMT1B CMT1B CMT1B CMT2 CMT1B CMT2

Ser49Leu His52Tyr‡ Arg69His Arg69His Arg69His Ile70Thr Val73fs Tyr90Cys Asn93Ser Thr95Met

None None Age 32–36 Age 60–61

Age 45 Age 42 Age 50–59

First Limitation

Age 53 Age 33 Age 42 Age 35 Age 23 Age 47 Age 29 Age 58 Age 41

Normal

Normal Normal Normal Normal

Normal Normal Normal

Age Walked

CMT2

CMT1B

Ser49Leu

CMT2

Asp32Gly Ser34del Phe35del

Glu42stop Asp46Val

CMT1B CMT2 CMT1B

Asp6Tyr Ser15Phe Ser22Phe

CMT1B CMT1B CMT1B CMT2

Author Diagnosis*

Amino Acid Change

Table 71–2. Features of Late-Onset Neuropathy (Literature Review)

M 32, U 30 M 38, U 40, P 27, T 39

U 36–52 M 34–58 U 36–52 M 37, T 26 M 44.5 M 33, U 35, P 21

M 32 M 47–49 T 32–44 M 34–58

M 24 M 16, U 19, T 14 M 33, U 34 P 29, U 44, S 30, M 34

M 31, P 31 M 36, P 29 31–53, many M 44–52 T 42 M 28, P 23 M 29 P 21.6 M 35, U 37

U 26, P 31 U 16.4 M 36–43, T 27–31

M 24–40

NCV† (m/s)

Sporadic Dominant, similar

Dominant Dominant Dominant, similar phenotype

9 families

3 generations, 7 patients

4 families

9 patients Father Dominant

Dominant

Dominant Unclear

2 separate pedigrees

3 families

Father Mother

Dominant

10 patients 16 patients Siblings

Family History

Conduction block

Adie’s pupil Deaf

Argyll-Robinson pupil, deaf

Adie’s pupil

Steroid responsive Steroid responsive

Adie’s pupil

Deafness

Severe cramps

Other Features

159 132

132 163 132

26

22

162 6 162 80 153 29 107, 151 119 13 98

37

80 98

64

119

93 91 162

Reference

1694

Diseases of the Peripheral Nervous System

siblings with a neuropathy characterized as HMSN II because of ulnar MNCVs of 44 and 51 m/sec. Early milestones were described as normal in a 31-year-old patient with a Ser34del, but no clinical information is provided except for a median NCV of 12 m/s. Other reported cases may also have had classic phenotypes, although clinical information permitting this classification was, again, incomplete. For example, early milestones were normal in a patient with a Ser34del and a median MNCV of 12 m/s.43 A Tyr53Cys mutation caused deformed feet in a 13-year-old, with thenar wasting at 15 and weakness in the legs and feet at 16. However, no history about early milestones was provided.62 The propositus of a family with a Ser49Leu mutation developed symptoms around the time of walking, though she was not examined until age 32. Her daughter had developed difficulties walking and pes cavus by the age of 7.81 Finally, a 12-year-old boy with a Gly184fs had normal early milestones but developed weakness and sensory loss with a median MNCV of 21 m/s. His 40-year-old mother, with the same mutation, was asymptomatic, though she had pes cavus and conduction velocities between 21 and 26 m/s.120

Phenotypic Variability Phenotypic variability is always a concern in attempts to perform genotype-phenotype correlations, and it does occur in patients with MPZ mutations. Moreover, characterizing a particular phenotype can be difficult because it is often difficult to determine the precise age of onset of genetic disorders.8 The most obvious example of phenotypic variability in HMSN IB patients is the report of distinct phenotypes in two monozygotic twins with Asp99Asn mutations.90 The firstborn twin did not walk until 18 months, and walked poorly until 3 years of age. He was a slow runner and developed a Roussy-Lévy tremor at age 15. The second-born twin had normal early motor milestones and, while also poor at sports as a child, performed better than his brother. At the age of 15, he could run 4 miles per day while his brother could never run normally. Complicating the issue, however, was the fact that the mother had developed toxemia during the pregnancy, and had a threatened miscarriage, and the twins were delivered 6 weeks prematurely. It is difficult to determine how these events may have contributed to the early phenotypes of the boys. Interestingly, the median (16 m/s) and ulnar (16 m/s) MNCVs are identical for both twins and their peroneal MNCVs are similar (17 m/s and 12 m/s), though the more severely disabled twin had a slower amplitude (0.3 mV compared to 1.4 mV). Moreover, at age 34 both boys were reported to have similar mild phenotypes.90 A second example of phenotypic variability is the report cited above of a 12-year-old boy with a Gly177stop with a classic CMT phenotype (normal early milestones, mild disability in adolescence, and median NCV of 21 m/s) whose affected mother had pes cavus but was otherwise asymptomatic, with a median MNCV of 26 m/s at age 40.120

In spite of some phenotypic variability, most MPZ mutations appear to cause specific phenotypes. Several examples support this point. Several groups have identified patients with a Thr95Met mutation, all of whom have a late-onset neuropathy associated with pupillary abnormalities and hearing loss.22,26,99 No cases of Thr95Met mutations causing early-onset neuropathy or neuropathy with significant slowing of nerve conductions have been identified. Furthermore, Bird and colleagues evaluated 13 members of a single family with the same MPZ mutation and found that 10 of the 13 individuals could not keep up with their peers by age 6, most were clumsy runners during childhood, and at least 6 had a delayed onset of walking. NCVs were less than 15 m/s in all 9 patients who were evaluated, and only 1 of the 13 developed symptoms after the age of 21.9 All but one report120 of cytoplasmic mutations that truncate the RSTK domain at amino acid 198 are associated with severe early-onset disease. In addition, the four patients we have evaluated from three families with His10Pro mutations have all had late onset phenotypes.125 Finally, only two mutations (R69H and S49L) out of 58 cited in Tables 71–1 and 71–2 have patients presenting with both early-onset and late-onset disease. With respect to R69H, the available clinical data placing patients into either the early or late group are scant. The patient in the early-onset group did not walk until 22 months of age, but no additional information about the patient is provided.43 The patient with a later onset of neuropathy developed symptoms at age 23, but there is no information provided about early milestones.162 NCVs were similar in both of these patients (19 m/s43 and 16.4 m/s162) suggesting that the phenotypes may not, in fact, be dissimilar. With respect to the S49L mutation, one patient was diagnosed at 11/2 years of age for unclear reasons; specific data about early milestones are not provided.129 This patient had surgery at the age of 6, as did his father, who carried the same mutation. However, the proband in a separate family with the same mutation developed gait abnormalities only at age 53, with the affected daughter remaining asymptomatic at age 27. The proband of a third pedigree remained asymptomatic until she developed cramps and gait difficulties at age 33. Despite the late onset, her phenotype was classified as severe, with the ulnar MNCV unobtainable and a peroneal MNCV of 21 m/s with amplitudes less than 1 mV.162 Taken together these results suggest that genotypephenotype correlations can be made with MPZ mutations, though outliers for particular mutations do exist.

PATHOLOGY Early-Onset Disease Marked abnormalities of myelin are characteristic of nerve biopsies from early-onset patients. Electron microscope (EM) studies have identified two apparently divergent types of myelin abnormalities: one with areas of uncompacted

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

1695

FIGURE 71–7 A, Electron micrograph of myelinated fiber in a patient with a Thr5Ile mutation with uncompacted lamellae of the inner myelin layers. Bar: 500 nm. B, Electron micrograph of sural nerve illustrating two fibers with focal myelin folding (tomacula), as well as thinly myelinated fibers, several demyelinated fibers surrounded by basal lamina, and a few thin Schwann cell lamellae. Bar: 2 ␮m. (From Gabreels-Festen, A. A., Hoogendijk, J. E., Meijerink, P. H., et al.: Two divergent types of nerve pathology in patients with different P0 mutations in Charcot-Marie-Tooth disease. Neurology 47:761, 1996, with permission.)

myelin and a second with excessive areas of focally folded myelin (tomacula)43 (Fig. 71–7). The two abnormalities appear to be mutually exclusive.43 Regions of noncompact myelin often preferentially involve inner, adaxonal wraps,35,43,96 as has been reported with Gln186stop,89,151 Thr5Ile,43 Arg69Cys,43,77,96 Cys98Tyr,35 and Arg69His96 mutations. Alternatively, tomacula have predominated in patients with Lys101Arg,43 Ile106Leu,43 Asn102Lys,109 Asp32Asn,9 and Ile33Phe.101 Biopsies from virtually all patients have exhibited other typical features of demyelination and remyelination, including onion bulb formation,9,43,64,77,81,89,96,129,136,137,146,152,162 abnormally thinly myelinated axons,9,35,64,77,89,96,109,151 and, on occasion, amyelinated fibers.35,89,151 Teased fiber analysis has demonstrated demyelination and remyelination in all patients studied from the early-onset group.35,64,109,151 A loss of myelinated axons was frequent in sural nerve biopsies from early-onset cases.9,64,77,96,109,136,137,152 In some studies, the loss preferentially involved large-diameter myelinated axons.35,43

Late-Onset Neuropathy Sural nerve biopsies from late-onset cases have often demonstrated more features of axonal degeneration than demyelination, though features of both were typically present. Many cases demonstrated a marked loss of myelinated fibers of all calibers, with numerous clusters of regenerating axons. These cases included three independent reports of Thr95Met biopsies.22,26,119 In one report, 73% of the

myelinated axons had been lost.26 In another Thr95Met sural nerve biopsy, there was a selective loss of predominantly large-diameter fibers with extensive sprouting and atrophy of remaining fibers.99 Similar results were found in biopsies from patients with a double His52Tyr and Val84Phe mutation,6 and in patients with Asp32Gly and Tyr90Cys mutations.119 A biopsy from a patient with a Ser49Leu mutation demonstrated a decrease in the density of myelinated fibers, particularly the large-diameter fibers.37 An Ile70Thr mutation caused a loss of myelinated axons at both the light microscope and EM levels, with an increase in endoneurial connective tissue.22 A superficial peroneal biopsy from a patient with Glu42stop showed a marked loss of myelinated fibers and the presence of many lipid-laden macrophages.80 A sural nerve biopsy from an Asp6Tyr patient showed a nonselective dropout of myelinated fibers and evidence of axonal degeneration by light microscopy, EM, and teased fiber analysis.93 The sural biopsy from a patient with an Asp46Val mutation showed a moderate loss of myelinated fibers and prominent axonal sprouting, and teased fibers from the patient also showed axonal loss with little evidence of demyelination.99 Features of demyelination were present but less noticeable in many late-onset cases. For example, no onion bulbs were detected in four separate reports from patients with Thr95Met mutations,21,52,93,113 or in patients with Asp6Tyr,93 Asp32Gly,119 and Asp46Val99 mutations. Teased fiber studies demonstrated no myelin abnormalities in a patient with an Ile70Thr mutation and in another

1696

Diseases of the Peripheral Nervous System

with a Thr95Met mutation.22 In a teased fiber study from a patient with a Thr95Met mutation, 83% of myelinated fibers were normal, 10% showed axonal change, and 7% showed remyelination. EM studies revealed no segmental demyelination of internodes or paranodal retraction in patients with Asp32Gly, Tyr90Cys, and Thr95Met mutations.119 However, biopsies from patients with a Thr95Met mutation,26 a His52Tyr and Val84Phe double mutation,6 a Ser49Leu mutation,37 and an Asp6Tyr mutation93 had at least some thinly myelinated axons. In the Ser49Leu biopsy, many fibers had concentric or eccentric thickenings of myelin sheath; several fibers had onion bulbs, particularly those from the second pedigree; and virtually all internodes on teased fiber studies were abnormal. Evidence of abnormal compaction and tomacula formation was also found in late-onset cases, but these were less frequent than in early-onset cases. For example, areas of uncompaction were noted but not considered to be prominent features of sural nerve biopsies from a series of patients with Asp32Gly, Tyr90Cys, and Thr95Met mutations.119 Similarly, in a biopsy from a patient with a His52Tyr and Val84Phe double mutation, the authors detected regions of uncompacted myelin but calculated that the frequency of these regions was in the normal range.6 Several uncompacted fibers were detected in a superficial peroneal nerve biopsy from a patient with a Glu42stop mutation,80 and uncompacted fibers were mentioned briefly in a biopsy from a patient with an Ile70Thr mutation.29 No uncompacted regions were identified in other patients with Thr95Met,53 Ser49Leu,37 or Asp6Tyr93 mutations. Tomacula were described in the patient cited above with the Ser49Leu mutation37 but in no other of the late-onset patients. In the patient with the Asp6Tyr mutation, no abnormalities in compaction or tomacula were detected.93

PATHOGENESIS Role of Intracellular Trafficking Results from the literature suggest that early-onset mutations prevent cis and trans interactions between adjacent and opposing MPZ molecules. As a result, patients develop morphologic abnormalities similar to those described in the noncompacting neuropathies of homozygous Mpz⫺/ Mpz⫺ mice. How might this occur? One possible explanation is that certain mutations prevent MPZ from being incorporated into the myelin sheath because the mutation prevents the protein from being transported out of the ER or Golgi apparatus of the Schwann cell. For example, certain mutations in the PLP1 gene are thought to cause severe forms of Pelizaeus-Merzbacher disease by causing PLP1 and its alternatively spliced isoform, DM20, to be retained in the ER, where they cannot be properly folded.47,48 Ordinarily proteins unsuitable for transport out of the ER

are lysed in proteosomes as part of the ER-associated degradation pathway. However, in certain circumstances this system can be overwhelmed such that the oligodendrocyte cannot process all of the mutant PLP1 and DM20 in the ER, and the viability of the cell is threatened (reviewed by Gow46). Some PMP22 mutations have been proposed to cause peripheral neuropathy by a similar mechanism.100 Thus it is possible that selected MPZ mutations cause severe early-onset neuropathy by overwhelming the ability of the ER and Golgi apparatus in myelinating Schwann cells to fold and process the mutant MPZ. As an example, two independent studies have demonstrated that the MPZ Ser34del mutation, associated with early-onset disease, prevents the protein from being transported out of the ER.94,121 However, many other mutations causing both earlyonset (Ser34Cys, Ser34Phe, Ser49Leu, Lys101Arg, Gly138Arg, Glu186stop) and late-onset (Ser15Phe, Thr95Met) neuropathy are capable of being transported out of the ER to the plasma membrane,121 as we have also shown with Arg198Ser.159 Furthermore, Kirschner and colleagues have demonstrated by both x-ray diffraction74 and high-resolution EM75 that certain MPZ mutations disrupt both extracellular and cytoplasmic apposition of myelin wraps. Similar studies, using mutations not found in patients, have shown that Cys21Ala163,164 and Asp92Glu and Gly94Ala165 mutant MPZ is transported to the cell surface of Chinese hamster ovary cells. Studies with transgenic mice that express a myc tag on the Mpz carboxyl terminus develop a severe neuropathy in which the mutant protein is not retained in the ER. Immunohistochemical analysis identifies the mutant protein in the noncompacted myelin sheath.110 Finally, overexpressing Mpz causes congenital hypomyelination in mice by causing premature and excessive adhesion in the developing myelin sheath, not because the additional Mpz is retained in the ER.158 Taken together, all these results suggest that many MPZ mutations are in fact transported to the myelin sheath and that it is in this location that they disrupt myelin.

Tertiary Structure of the Extracellular Domain Based on the aforementioned in vitro studies and the known biology of MPZ, it seems reasonable to hypothesize that at least some mutations cause neuropathy by disrupting adhesion at the level of the myelin sheath. We first attempted to identify a relationship between phenotypes and adhesion by evaluating whether early- and late-onset mutations occur in amino acids thought to be critical for cis- or transinteracting sites for adhesion.122 We hypothesized that earlyonset cases were more likely to have abnormal adhesion because many of them had abnormalities in compaction on ultrastructural analysis, along with their delays in achieving early milestones. The results, however, shown in Tables 71–3 and 71–4, do not reveal an obvious relationship between

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

1697

Table 71–3. Features of Mutations Affecting Amino Acids Critical for Homotypic Interactions Interface

Early-Onset Neuropathy

Late-Onset Neuropathy

Fourfold Adhesive Head to Head Remaining Amino Acids

R69C, R69S, R69H, N87H I1M, C21del, Y39C, S49L, H52R C21del T5I, I32F, Del35F (homo), Y53C, Del56QPYI/InsHLF, D61E, G64E, W72C, V73fs (homo), G74E, I85T, Del89DY/InsFY, G94C, Del96FT, C98Y, K101R, I106T/L, N102K, V117F, G138R

S15F, R69H D46V, S49L, H52Y D32G D6Y, H10P, S22F, F35del, I70T, V73fs (hetero), Y90C, N93S, T95M, S111N, G134R, R198S, Del207K

these critical amino acids and patient phenotypes. In earlyonset cases, 10 mutations occur at sites deemed critical to form cis or trans homotypic interactions with other MPZ molecules. However, many other early-onset mutations do not occur at one of these critical sites, including Val3Phe, Thr5Ile, Ser25Cys, Thr53Cys, Gly64Glu, Lys67Glu, Trp72Cys, Gly74Glu, Ile83Thr, Ile85Thr, Cys98Tyr, Lys101Arg, Asn102Lys, and Val117Phe. Similarly, in lateonset cases, some mutations6 occur in amino acids critical for adhesion while others13 do not. We next investigated whether particular mutations that changed the charge or hydrophobicity of the amino acid had an effect on the phenotype. These results suggest that amino acid substitutions that introduce a negative (Gly64Glu, Gly74Glu, Asn102Lys) or positive (Asn87His) charge to the extracellular domain of MPZ cause early-onset disease, because no mutation introducing such a change caused lateonset disease. Removing a charged amino acid did not necessarily cause early-onset disease, because Asp6Tyr, His10Pro, Asp32Gly, Asp42Val, and His52Val mutations cause lateonset disease. There were no obvious factors based on whether an amino acid is neutral and polar or neutral and

hydrophobic that predicted whether a particular mutation would cause an early or late onset. Thus, in early-onset cases, three mutations change a neutral and hydrophobic amino acid to a neutral and polar amino acid (Trp72Cys, Ile85Thr, and Ile106Thr); three mutations change a neutral and polar amino acid to a neutral and hydrophobic amino acid (Thr5Ile, Ser34Phe, and Ser49Leu); one mutation change one neutral and hydrophobic amino acid for another (Val117Phe); three mutations change one neutral and polar amino acid for another (Tyr39Cys, Tyr53Cys, and Cys98Tyr); and two mutations change one basic amino acid for another (Lyr101Arg and His53Arg). Similarly, in late-onset cases, one mutation changes a neutral and hydrophobic amino acid to a neutral and polar amino acid (Ile70Thr); four mutations change a neutral and polar amino acid to a neutral and hydrophobic amino acid (Ser15Leu, Ser15Phe, Ser49Leu, and Thr95Met); two mutations change one neutral and polar amino acid for another (Tyr90Cys and Asn93Ser); and one mutation changes one basic amino acid for another (Arg69His). Cys-Cys disulfide bonds are postulated to stabilize the final conformation of many immunoglobulin superfamily adhesion molecules (reviewed by Zhang and Filbin163). We

Table 71–4. Features of Mutations Affecting Phenotypes Early Onset Extracellular

Transmembrane Cytoplasmic Late Onset Extracellular

Transmembrane Cytoplasmic

Introduction of new ⫹/⫺ charges Taking away ⫹/⫺ charges Deletion truncating the EC domain Disruption involving Cys Changes in polarity and hydrophobicity Frameshift that truncates TM and IC domain Frameshift that truncates IC domain

G64E, G74E, N87H, N102K, G138R R69S, R69C, D99N C21del, Q56del, D89del, F96del C21del, S34C, Y39C, Y52C, R69C, W72C, G94C, C98Y T5I, S34F, S49L, T85I, W72C, I106T G138fs, L145fs A159fs, A192fs, Q186stop

Taking away ⫹/⫺ charge Change at/near glycosylation site Introduction of Cys Changes in polarity and hydrophobicity

D6Y, H10P, D32G, D42V, H52Y N93S, Y90C, T95M Y90C D6Y, H10P, S15F, S22F, D32G, D46V, H52Y, I70T, T95M G134R R198S K207del

Change at RSTK site (PKC) Deletion

1698

Diseases of the Peripheral Nervous System

therefore speculated that mutations disrupting existing cysteines or introducing additional cysteines into the extracellular domains would be particularly likely to disrupt adhesion and cause early-onset neuropathy. This proved to be true in most cases. Cys98Tyr35 and Cys21–28del125 mutate the cysteine residues predicted to cause disulfide bonds in MPZ,82,112,163 and both of these mutations cause severe, early-onset disease in patients. Similarly, Ser34Cys, Val39Cys, Tyr53Cys, Arg69Cys, and Trp72Cys all cause early-onset disease. However, Tyr90Cys introduces a cysteine only eight amino acids removed from Cys98 but has caused a late-onset phenotype. Mutations that affect amino acids conserved throughout evolution (shark, chicken, bovine, rat, and human) also appear slightly more likely to cause early (17 of 23, or 74%) than late onset (10 of 17, or 59%) cases. Conserved amino acids causing early-onset disease include Ile1, Thr5, Ser49, His52, Tyr53, Gly64, Arg69, Trp72, Gly74, Ile85, Asn87, Gly94, Cys98, Asp99, Lys101, Asn102, and Val117; conserved amino acids causing late-onset disease include Ser15, Asp46, Ser49, Arg69, Val73, Tyr90, Asn93, Thr95, Ser111, and Gly134. Nonconserved amino acids causing early-onset disease are Ile33 (not conserved in cow and shark), Ser34 (not conserved in shark), Tyr39 (not conserved in chicken and shark), Asp61 (not conserved in shark), Ile106 (not conserved in shark), and Gly138 (not conserved in chicken). Nonconserved amino acids causing late-onset disease are Asp6 (not conserved in chicken and shark), His10 (not conserved in rat and chicken), Ser22 (not conserved in shark), Asp32 (not conserved in chicken), Phe35 (not conserved in chicken and shark),Glu68 (not conserved in shark), Ile70 (not conserved in chicken and shark), and Arg198 (not conserved in shark). Taken together, these results suggest that adding a charged amino acid, disrupting a cysteine, or changing an amino acid conserved throughout evolution permits one to predict whether a particular mutation in the extracellular domain would cause early- or late-onset disease. To address this issue more thoroughly will likely require detailed studies on how particular mutations affect adhesion and alter the proposed crystal structure of the extracellular domains of MPZ, in addition, of course, to determining whether the predicted mutant proteins are incorporated into the myelin sheath.

Truncation of the Cytoplasmic Domain Associated with Early-Onset Neuropathy As discussed previously, the cytoplasmic domain of MPZ is also necessary for MPZ-mediated adhesion156,157; therefore mutations that significantly truncate the cytoplasmic domain would be predicted to cause severe neuropathies by delaying or preventing adhesion during development. This prediction proves to be correct in most cases. Transmembrane domain mutations that, through frameshifts, eliminate the normal sequence of the MPZ cytoplasmic domain, cause severe,

early-onset disease. These mutations include Leu137fs,146 Val140fs,139 and Leu145fs.151 Similarly, mutations within the cytoplasmic domain that severely truncate the normal sequence, such as Ala159fs205stop,137 Gln186stop,89,151 and Ala191fs,112 also cause severe, early-onset forms of neuropathy. The one exception to this rule has been a 12 year-old boy with a mild neuropathy caused by a Gly177fs, which he inherited from his asymptomatic mother.120 While the mechanisms by which the cytoplasmic domain regulates adhesion remain to be determined, they are likely to involve PKC-mediated signaling at the RSTK site between amino acids 198 and 201, as well as Ser 204.15 Thus the Arg198Ser mutation is one of the few amino acid substitution mutations in the cytoplasmic domain known to cause neuropathy,159 whereas the Arg215Leu mutation, distal to this domain, is the sole amino acid–changing mutation of MPZ reported to be a benign polymorphism.15 Furthermore, supporting the putative role of serine 204, the Ser204fs mutation has been reported to cause a severe case of DSD.3

Schwann Cell–Axon Interactions and Axonal Degeneration Schwann cell–axon interactions are disrupted in virtually all demyelinating neuropathies, with resultant axonal degeneration (reviewed by Shy et al.124). Trembler mice, which have a dysmyelinating peripheral neuropathy caused by a point mutation in Pmp22,133 have significant changes in both axonal structure and function, including alterations in neurofilament phosphorylation, increased neurofilament density, and decreased axonal transport.27 Similar changes have been found in patients with HMSN I.154 Patients with HMSN type X, caused by mutations in the GJ␤1 gene encoding the gap junction protein connexin 32, also develop axonal loss, and like patients with the late-onset MPZ neuropathies, often have little evidence of demyelination pathologically or by nerve conduction studies.52 In all of these neuropathies signaling pathways between Schwann cells and axons are disrupted. While the details of signaling between Schwann cells and axons are outside the scope of this chapter, they are reviewed briefly because they are germane to the discussion. Signaling between Schwann cells and axons probably occurs, at least in part, between the nodal, paranodal, and juxtaparanodal regions of the Schwann cell and underlying axolemma. The paranodal region, the loops of Schwann cell membrane and interacting axonal membrane, contains the Schwann cell proteins MAG, connexin 32, and neurofascin 155 and the axolemmal proteins Caspr and contactin (see Fig. 71–4). These proteins participate in interactions between Schwann cells and axons, and they act to isolate the nodal region electrically. The juxtaparanodal region, the portion of Schwann cell and interacting axonal membrane adjacent to the paranode, contains potassium channels and Caspr 2, both expressed by axons. The results of these specializations

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

lead to the establishment of an electrically insulated node of Ranvier, capable of saltatory conduction. When these specializations are disrupted, saltatory conduction becomes disrupted as well. Thus, on the axolemmal side, when Caspr5 or contactin16 is deleted, the Shaker-type Kv1.1 and Kv1.2 potassium channels are displaced from their normal position in the juxtaparanodes, junctions between the axolemma and paranodal loops are disrupted, and saltatory conduction is disrupted. Similarly, mice whose Schwann cells cannot express sulfatide display similar abnormalities in their distribution of axolemmal potassium channels and disruption of axo-glial junctions, although myelin compaction is normal.66 Finally, occasional Schwann cells in Mpz⫺/Mpz ⫺ mice are no longer able to express Mag. In these internodes, Caspr is no longer localized normally in the axolemma.97 Taken together, these results suggest that disrupted signaling between Schwann cells and axons can occur without disrupting myelin compaction, and that these abnormalities can lead to changes in the axon even though the cause of the abnormalities is in the Schwann cell.

Axonal Degeneration in Late-Onset Neuropathy While mutations may cause the early onset of neuropathy by delaying the completion of myelination, it is not obvious why other mutations should cause late-onset symptoms. It is not simply a case of late-onset patients developing a milder neuropathy, because several patients with Thr95Met mutations required a wheelchair for ambulation within a decade of the onset of their symptoms.26 Rather, myelination appears to develop at least in a fairly normal fashion in these patients, but signaling between Schwann cells and axons becomes impaired, often dramatically. How might this occur? First, mutant MPZ might prevent the normal organization of the axolemma or Schwann cell paranodal loops. Over time this could lead to increased energy demands on the neuron and its ultimate degeneration. At present there are no detailed immunohistochemical studies on the paranodal regions of nerves from patients with MPZ mutations. Second, certain MPZ mutations might permit the formation of less stable tetramers in cis and trans, resulting in myelin wraps that are poorly anchored to each other. Thus myelination could proceed normally, but adjacent wraps would be more susceptible to shearing forces or mechanical stress. Over time this may lead to the disruption of the normal alignment between molecules in Schwann cells and the axolemma that are necessary for Schwann cell–axon interactions. Finally, it is possible that mutations might lead to unstable myelin that would then cause an inflammatory or immune process that would damage the nerve. Mpz heterozygous knockout mice (Mpz⫹/Mpz⫺) are clinically normal at a young age but subsequently develop severe, asymmetrical

1699

slowing of MNCVs, with temporal dispersion or conduction block. These findings are similar to those in patients with the presumably immune-mediated chronic inflammatory demyelinating polyradiculopathy (CIDP). Moreover, morphologic analysis of affected nerves from the mice reveals severe and selective demyelination of motor fibers, focal regions of demyelination, and inflammatory cells.126 Schmid and colleagues have rescued the neuropathy, at least in part, in Mpz⫹/Mpz⫺ mice by breeding the mice into null mutants for the recombinant activating gene 1 (Rag1), in which the mice cannot generate an immune response.118 These same investigators then reinduced the neuropathy by reconstituting the bone marrow of the Rag1-deficient mice with bone marrow from immunocompetent wild-type mice95 (see also Chapter 24). Clinical improvement in a late-onset patient treated with steroid therapy has also been reported.29 In this context, it has recently been demonstrated that sera from some patients with CIDP have antibodies to MPZ and that these antibodies caused slowing of NCVs and conduction block following injection into experimental animals.160

DIFFERENTIAL DIAGNOSIS The first differential concerns patients with demyelinating peripheral neuropathies who have a pedigree suggestive of an autosomal dominant inheritance pattern. The most frequent cause of neuropathy in this group is HMSN IA (discussed in Chapter 69), caused by a duplication on chromosome 17. Features suggesting HMSN IA include uniformly slowed NCVs,81 in the range of 20 to 25 m/s,11,78,140 although NCVs as high as 38 m/s have been reported for HMSN IA.86 In addition to HMSN IA, other dominantly inherited demyelinating neuropathies should be considered in the differential, including those caused by PMP22, the early growth response 2 (EGR2) gene, and neurofilament light gene point mutations (see http:/ /molgen-www.uia.ac.be/ CMTMutations/). X-linked HMSN cases can also resemble HMSN IB, particularly the late-onset cases, because many HMSN X patients have only mildly slowed NCVs.85 Many cases of HMSN IB are sporadic, without a clear family history. They must therefore be differentiated from recessively inherited neuropathies, including those caused by periaxin or ganglioside-induced differentiation–associated protein 1 mutations. Also, as alluded to previously, sporadic HMSN IB must sometimes be distinguished from CIDP, because some patients with HMSN IB have asymmetrically slowed NCVs85 and elevated CSF protein levels.

MANAGEMENT Effective treatment of patients with MPZ mutations requires a multidisciplinary approach involving genetic counseling,

1700

Diseases of the Peripheral Nervous System

physiatry, physical and occupational therapy, and pulmonary support. Genetic counseling helps patients and their families deal with the complex issues that arise when the diagnosis of a genetic disease is made. Nondirective genetic counseling can help patients not only understand who in their families is at risk for developing neuropathy but also make informed decisions that they feel are best for their lives. An informed neurologist, familiar with the natural history of the various forms of HMSN IB, can aid in this process by helping the patient understand what is known about the likely progression of his or her particular MPZ mutation. However, a challenge for many clinicians is that the approach to the patient with an inherited condition is often different from that to other patients. While many neurologists are trained to give advice regarding prognosis and therapy, a nondirective counseling approach is necessary for handling many of the complex issues that arise when the diagnosis of a inherited condition is made. Patients with inherited conditions often seek further information regarding various decisions, including those regarding family planning. All options available, including prenatal testing, nondisclosed prenatal testing, preimplantation genetic diagnosis, sperm and egg donation, adoption, having children without any testing, and having no children, must be explored. It is imperative that patients are aided in making decisions that best fit with their beliefs, values, culture, and lifestyle and are not influenced by the bias of the clinician. This approach can be difficult, especially when the clinician is faced with the inevitable question, “What would you do?” Although it may be very easy and tempting to answer this question with an opinion, it is important to keep in mind that the answer given may not be the best solution for a particular patient. Nondirective counseling is based on the belief that an individual is the only person who knows what decisions are best for his or her life and that that individual has the right to make autonomous decisions.

natural history studies of the various inherited neuropathies and accurate genotype-phenotype correlation between different HMSN I mutations are not yet available, and it is only by generating this information that accurately predicting a patient’s clinical course will be possible. Knowing the specific mutation in a family also allows for the risk to the patient’s children and other relatives to be determined and for the possibly of prenatal testing. Once one family member has been accurately genotyped, it is usually not necessary to routinely genotype other family members. Instead, they can be screened by clinical examination and nerve conduction studies. It is also useful to screen patients in whom no maleto-male transmission has been demonstrated, even if their NCVs are not clearly slow. Conduction velocities in patients with HMSN type X can resemble those of HMSN IB, and X-linked inheritance puts at risk a different population within the family than an autosomal dominant form of disease. In the case of testing for individuals at risk for MPZ mutations, it is worth emphasizing again the need for a multidisciplinary approach for the testing. Guidelines for predictive testing for conditions such as Huntington’s disease have been well established. At a minimum, it is recommended that a patient, along with a support person, be seen by a neurologist and genetic counselor and be counseled about the possible implications of testing regarding their relationships, insurance, and employability. Every effort to obtain documentation of the disease in a relative should be exhausted. The issues surrounding predictive testing are extremely complex and need to be very carefully addressed with the patient prior to testing. Testing is usually not recommended for asymptomatic children who are at risk for HMSN IB. If a child were clearly showing symptoms, then testing would be appropriate. Issues regarding family planning are also often of concern.

Physiatry Decisions about Genetic Testing The neurologist must also determine when to order genetic testing for MPZ mutations. It may seem self-evident to say that one needs to suspect a particular disease in order to test for it. However, the clinical presentations of many of the inherited neuropathies can be complex and may overlap, and it may sometimes be difficult to be certain of what test to order. Despite these complexities, it is essential that the clinician avoid the temptation to simply order an all-inclusive battery of available genetic tests for inherited neuropathies. These tests are expensive, and in many cases are not covered by insurance carriers. As a general rule, a reasonable approach is to order testing when the clinical phenotype, electrophysiology, or pathology suggests the disease to be tested, in this case HMSN IB. One approach is to order testing in patients whose evaluation suggests HMSN IB when other members of the family have not yet been tested. While some might feel this testing is unnecessary, careful

Physiatry and rehabilitation medicine are discussed in Chapter 118, but because they are essential for the management of HMSN IB, they will be briefly mentioned here. Most patients develop stability problems at their ankles as a consequence of distal weakness of foot muscles and a distal loss of position sense. As with many other forms of neuropathy, the weakness is usually more pronounced in muscles of dorsiflexion and foot eversion than in muscles of plantar flexion and foot inversion. Thus even mildly affected patients tend to roll their ankles laterally, sprain their ankles frequently, and be at risk for falling. Most patients require at least orthotics to stabilize their ankles, and often benefit from AFOs. Because disability may progress to the point that patients require further ambulation aids, including walkers or wheelchairs, a physiatrist can play an important role in assessing the impact of the impairment upon performance of functional skills, and developing a treatment plan that best aids the patient over time.

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations

1701

When hand intrinsic muscles develop symptoms, problems often arise with fine motor hand skills such as buttoning or zipping clothes, pulling on socks or tying shoes, holding a pen or pencil, or typing on a computer keyboard. The weak hand muscles will sometimes fatigue with activity, causing painful muscle cramps. An occupational therapist can help in such circumstances by providing adaptive equipment (e.g., a buttonholer or sock puller) and teaching techniques that minimize the stress on the intrinsic hand muscles or compensate for their lack of function. Occupational therapists can also help problem solve difficulties with performance of household chores by such methods as using a wheeled cart for carrying items around the house when hand function and balance have been affected. If driving skills become impaired by either lower or upper limb involvement, some occupational therapy departments offer driver training with adaptive equipment to allow continued safe operation of a motor vehicle.

where gene products can be actively taken up into peripheral nerves by retrograde transport.51 Trophic factors have been demonstrated to prevent motor neuron death and nerve degeneration. Genetically altered T cells have demonstrated an ability to introduce genes into peripheral nerve.79 Future prospects for the development of an effective and safe therapy for inherited neuropathies are thus promising.

Pulmonary Management

1. Acsadi, G., Massie, B., and Jani, A.: Adenovirus-mediated gene transfer into striated muscles. J. Mol. Med. 73:165, 1995. 2. Agrawal, H. C., Schmidt, R. E., and Agrawal, D.: In vivo incorporation of [3H]palmitic acid into P0 protein, the major intrinsic protein of rat sciatic nerve myelin. J. Biol. Chem. 258:6556, 1983. 3. Bellone, E., Mandich, P., James, R., et al.: Identification of a 4 bp deletion (1560del4) in P0 gene in a family with severe Charcot-Marie-Tooth disease. Hum. Mutat. 7:377, 1996. 4. Benstead, T. J., Kuntz, N. L., Miller, R. S., and Daube, J. R.: The electrophysiologic profile of Dejerine-Sottas disease (HMSN III). Muscle Nerve 13:586, 1990. 5. Bhat, M.A., Rios, J. C., Lu, Y., et al.: Axon-glia interactions and the domain organization of myelinated axons requires neurexin IV/Caspr/Paranodin. Neuron 30:369, 2001. 6. Bienfait, H. M., Baas, F., Gabreels-Festen, A. A., et al.: Two amino-acid substitutions in the myelin protein zero gene of a case of Charcot-Marie-Tooth disease associated with light-near dissociation. Neuromuscul. Disord. 12:281, 2002. 7. Bird, T. D.: Historical perspective of defining CharcotMarie-Tooth type 1B. Ann. N. Y. Acad. Sci. 883:6, 1999. 8. Bird, T. D., and Kraft, G. H.: Charcot-Marie-Tooth disease: data for genetic counseling relating age to risk. Clin. Genet. 14:43, 1978. 9. Bird, T. D., Kraft, G. H., Lipe, H. P., et al.: Clinical and pathological phenotype of the original family with Charcot-MarieTooth type 1B: a 20-year study. Ann. Neurol. 41:463, 1997. 10. Bird, T. D., Ott, J., and Giblett, E. R.: Evidence for linkage of Charcot-Marie-Tooth neuropathy to the Duffy locus on chromosome 1. Am. J. Hum. Genet. 34:388, 1982. 11. Birouk, N., Gouider, R., Le Guern, E., et al.: CharcotMarie-Tooth disease type 1A with 17p11.2 duplication: clinical and electrophysiological phenotype study and factors influencing disease severity in 119 cases. Brain 120(Pt. 5):813, 1997. 12. Blanquet-Grossard, F., Pham-Dihn, D., Dautigny, A., et al.: Charcot-Marie-Tooth type 1B neuropathy: third mutation of serine 63 codon in the major peripheral myelin glycoprotein P0 gene. Clin. Genet. 48:281, 1995.

While most patients do not develop overt respiratory symptoms, severely affected patients can develop breathing problems and one patient has died because of respiratory failure.91 Therefore, following pulmonary function tests is an important component of the management of severely affected patients. Selected patients may benefit from continuous positive airway pressure or biphasic positive airway pressure therapy.

Pain Management Neuropathic pain, such as burning or tingling of feet and hands, occurs in some patients with MPZ mutations, though these patients are the minority. Treatment for HMSN IB pain patients is similar to that for other patients with neuropathic pain and includes tricyclic antidepressant agents such as amitriptyline or nortriptyline, antiepileptic agents such as gabapentin, and topical analgesics. Particularly as they age, patients may develop aching arthritic pain in their feet, ankles, knees, hips, and/or back. These ailments often respond to nonsteroidal anti-inflammatory agents. As with other forms of inherited neuropathies, patients may develop symptoms in their hands typical of the carpal tunnel syndrome. These symptoms are usually treated with night splints or other conservative approaches.

Future Directions While at present there are no cures for most of the forms of inherited neuropathies, advances in understanding the genetics, molecular biology, and pathogenesis of HMSN (CMT) have made gene therapy a realistic possibility for the future. Adenoviral vectors have been shown to be able to introduce genes into Schwann cells69,127 or into muscles,1

ACKNOWLEDGMENTS The author would like to thank James Garbern for his help with the figures, Agnes Jani for her help with the tables, and John Kamholz, Karen Krajewski, Jack Lilien, and Janne Balsamo for their helpful suggestions.

REFERENCES

1702

Diseases of the Peripheral Nervous System

13. Blanquet-Grossard, F., Pham-Dihn, D., Dautigny, A., et al.: Charcot-Marie-Tooth type 1B neuropathy: a mutation at the single glycosylation site in the major peripheral myelin glycoprotein Po. Hum. Mutat. 8:185, 1996. 14. Borghini, I., Ania-Lahuerta, A., Regazzi, R., et al.: Alpha, beta I, beta II, delta, and epsilon protein kinase C isoforms and compound activity in the sciatic nerve of normal and diabetic rats. J. Neurochem. 62:686, 1994. 15. Bort, S., Nelis, E., Timmerman, V., et al.: Mutational analysis of the MPZ, PMP22 and Cx32 genes in patients of Spanish ancestry with Charcot-Marie-Tooth disease and hereditary neuropathy with liability to pressure palsies. Hum. Genet. 99:746, 997. 16. Boyle, M. E., Berglund, E. D., Murai, K. K., et al.: Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30:385, 2001. 17. Bradley, W. G., Madrid, R., and Davis, C. J.: The peroneal muscular atrophy syndrome: clinical genetic, electrophysiological and nerve biopsy studies. Part 3. Clinical, electrophysiological and pathological correlations. J. Neurol. Sci. 32:123, 1977. 18. Brockes, J. P., Raff, M. C., Nishiguchi, D. J., and Winter, J.: Studies on cultured rat Schwann cells. III. Assays for peripheral myelin proteins. J. Neurocytol. 9:67, 1980. 19. Brust, J. C., Lovelace, R. E., and Devi, S.: Clinical and electrodiagnostic features of Charcot-Marie-Tooth syndrome. Acta Neurol. Scand. Suppl. 68:1, 1978. 20. Buchthal, F., and Behse, F.: Peroneal muscular atrophy (PMA) and related disorders. I. Clinical manifestations as related to biopsy findings, nerve conduction and electromyography. Brain 100(Pt. 1):41, 1977. 21. Chapon, F., Diraison, P., Lechevalier, B., et al.: Hereditary neuropathy with liability to pressure palsies with a partial deletion of the region often duplicated in Charcot-MarieTooth disease, type 1A. J. Neurol. Neurosurg. Psychiatry 61:535, 1996. 22. Chapon, F., Latour, P., Diraison, P., et al.: Axonal phenotype of Charcot-Marie-Tooth disease associated with a mutation in the myelin protein zero gene. J. Neurol. Neurosurg. Psychiatry 66:779, 1999. 23. Charcot, J., and Marie, P.: Sue une forme particulaire d’atrophie musculaire progressive souvent familial debutant par les pieds et les jamber et atteingnant plus tard les mains. Rev. Med. 6:97, 1886. 24. Dawidenkow, S.: Uber die neurotische Muskelatrophie Charcot-Marie: Klinisch-genetische Studien. Z. Neurol. 108:259, 1927. 25. Dawidenkow, S.: Uber dir neurotische Muskelatrophie Charcot-Marie: Klinisch-genetische Studien. Z. Neurol. 108:344, 1927. 26. De Jonghe, P., Timmerman, V., Ceuterick, C., et al.: The Thr124Met mutation in the peripheral myelin protein zero (MPZ) gene is associated with a clinically distinct Charcot-Marie-Tooth phenotype. Brain 122(Pt. 2):281, 1999. 27. de Waegh, S., and Brady, S. T.: Altered slow axonal transport and regeneration in a myelin-deficient mutant mouse: the trembler as an in vivo model for Schwann cell-axon interactions. J. Neurosci. 10:1855, 1990.

28. Dejerine, H., and Sottas, J.: Sur la nevritte interstitielle, hypertrophique et progressive de l’enfance. C. R. Soc. Biol. Paris 45:63, 1893. 29. Donaghy, M., Sisodiya, S. M., Kennett, R., et al.: Steroid responsive polyneuropathy in a family with a novel myelin protein zero mutation. J. Neurol. Neurosurg. Psychiatry 69:799, 2000. 30. D’Urso, D., Brophy, P. J., Staugalitis, S. M., et al.: Protein zero of peripheral nerve myelin: biosynthesis, membrane insertion, and evidence for homotypic interaction. Neuron 4:449, 1990. 31. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. I. Neurologic, genetic, and electrophysiologic findings in hereditary polyneuropathies. Arch. Neurol. 18:603, 1968. 32. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. II. Neurologic, genetic, and electrophysiologic findings in various neuronal degenerations. Arch. Neurol. 18:619, 1968. 33. Eichberg, J., and Iyer, S.: Phosphorylation of myelin protein: recent advances. Neurochem. Res. 21:527, 1996. 34. Eylar, E. H., Uyemura, K., Brostoff, S. W., et al.: Proposed nomenclature for PNS myelin proteins. Neurochem. Res. 4:289, 1979. 35. Fabrizi, G. M., Cavallaro, T., Morbin, M., et al.: Novel mutation of the P0 extracellular domain causes a Dejerine-Sottas syndrome. J. Neurol. Neurosurg. Psychiatry 66:386, 1999. 36. Fabrizi, G. M., Ferrarini, M., Cavallaro, T., et al.: A somatic and germline mosaic mutation in MPZ/P(0) mimics recessive inheritance of CMT1B. Neurology 57:101, 2001. 37. Fabrizi, G. M., Taioli, F., Cavallaro, T., et al.: Focally folded myelin in Charcot-Marie-Tooth neuropathy type 1B with Ser49Leu in the myelin protein zero. Acta Neuropathol. (Berl.) 100:299, 2000. 38. Filbin, M. T., and Tennekoon, G. I.: The role of complex carbohydrates in adhesion of the myelin protein, P0. Neuron 7:845, 1991. 39. Filbin, M. T., Walsh, F. S., Trapp, B. D., et al.: Role of myelin P0 protein as a homophilic adhesion molecule. Nature 344:871, 1990. 40. Filbin, M. T., Zhang, K., Li, W., and Gao, Y.: Characterization of the effect on adhesion of different mutations in myelin P0 protein. Ann. N. Y. Acad. Sci. 883:160, 1999. 41. Franz, T., Waehneldt, T. V., Neuhoff, V., and Wachtler, K.: Central nervous system myelin proteins and glycoproteins in vertebrates: a phylogenetic study. Brain Res. 226:245, 1981. 42. Friedreich, N.: Uber progressive Muskelatrophie, uber wahre und falshe Muskelhypertrophie. Berlin, Hirschwald, 1873. 43. Gabreels-Festen, A. A., Hoogendijk, J. E., Meijerink, P. H., et al.: Two divergent types of nerve pathology in patients with different P0 mutations in Charcot-Marie-Tooth disease. Neurology 47:761, 1996. 44. Garbay, B., Boiron-Sargueil, F., Shy, M., et al.: Regulation of oleoyl-CoA synthesis in the peripheral nervous system: demonstration of a link with myelin synthesis. J. Neurochem. 71:1719, 1998. 45. Giese, K.P., Martini, R., Lemke, G., et al.: Mouse P0 gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 71:565, 1992.

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations 46. Gow, A.: Protein misfolding as a disease determinant. In Lazzarini, R. (ed.): Myelin—Biology and Disorders. Oxford, UK, Academic Press, 2003 (in press). 47. Gow, A., and Lazzarini, R.: A cellular mechanism governing the severity of Pelizaeus-Merzbacher disease. Nat. Genet. 13:422, 1996. 48. Gow, A., Southwood, C. M., and Lazzarini, R. A.: Disrupted proteolipid protein trafficking results in oligodendrocyte apoptosis in an animal model of Pelizaeus-Merzbacher disease. J. Cell Biol. 140:925, 1998. 49. Greenfield, S., Brostoff, S., Eylar, E. H., and Morell, P.: Protein composition of myelin of the peripheral nervous system. J. Neurochem. 20:1207, 1973. 50. Guzzetta, F., Ferriere, G., and Lyon, G.: Congenital hypomyelination polyneuropathy: pathological findings compared with polyneuropathies starting later in life. Brain 105(Pt. 2):395, 1982. 51. Haase, G., Kennel, P., Pettmann, B., et al.: Gene therapy of murine motor neuron disease using adenoviral vectors for neurotrophic factors. Nat. Med. 3:429, 1997. 52. Hahn, A. F., Ainsworth, P. J., Naus, C. C., et al.: Clinical and pathological observations in men lacking the gap junction protein connexin 32. Muscle Nerve 23(Suppl. 9):S39, 2000. 53. Hanemann, C. O., Gabreels-Festen, A. A., and DeJonghe, P.: Axon damage in CMT due to mutation in myelin protein P0. Neuromuscul. Disord. 11:753, 2001. 54. Harding, A. E., and Thomas, P. K.: The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103:259, 1980. 55. Harding, A. E., and Thomas, P. K.: Genetic aspects of hereditary motor and sensory neuropathy (types I and II). J. Med. Genet. 17:329, 1980. 56. Hayasaka, K., Himoro, M., Sawaishi, Y., et al.: De novo mutation of the myelin P0 gene in Dejerine-Sottas disease (hereditary motor and sensory neuropathy type III). Nat. Genet. 5:266, 1993. 57. Hayasaka, K., Himoro, M., Wang, Y., et al.: Structure and chromosomal localization of the gene encoding the human myelin protein zero (MPZ). Genomics 17:755, 1993. 58. Hayasaka, K., Nanao, K., Tahara, M., et al.: Isolation and sequence determination of cDNA encoding the major structural protein of human peripheral myelin. Biochem. Biophys. Res. Commun. 180:515, 1991. 59. Hayasaka, K., Ohnishi, A., Takada, G., et al.: Mutation of the myelin P0 gene in Charcot-Marie-tooth neuropathy type 1. Biochem. Biophys. Res. Commun. 194:1317, 1993. 60. Heath, J. W., Inuzuka, T., Quarles, R. H., and Trapp, B. D.: Distribution of P0 protein and the myelin-associated glycoprotein in peripheral nerves from Trembler mice. J. Neurocytol. 20:439, 1991. 61. Hilmi, S., Fournier, M., Valeins, H., et al.: Myelin P0 glycoprotein: identification of the site phosphorylated in vitro and in vivo by endogenous protein kinases. J. Neurochem. 64:902, 1995. 62. Himoro, M., Yoshikawa, H., Matsui, T., et al.: New mutation of the myelin P0 gene in a pedigree of Charcot-MarieTooth neuropathy 1. Biochem. Mol. Biol. Int. 31:169, 1993. 63. Ikegami, T., Ikeda, H., Mitsui, T., et al.: Novel mutation of the myelin P0 gene in a pedigree with Charcot-MarieTooth disease type 1B. Am. J. Med. Genet. 71:246, 1997.

1703

64. Ikegami, T., Nicholson, G., Ikeda, H., et al.: A novel homozygous mutation of the myelin P0 gene producing Dejerine-Sottas disease (hereditary motor and sensory neuropathy type III). Biochem. Biophys. Res. Commun. 222:107, 1996. 65. Ikegami, T., Nicholson, G., Ikeda, H., et al.: De novo mutation of the myelin Po gene in Dejerine-Sottas disease (hereditary motor and sensory neuropathy type III): two amino acid insertion after Asp 118. Hum. Mutat. Suppl. 1:S103, 1998. 66. Ishibashi, T., Dupree, J. L., Ikenaka, K., et al.: A myelin galactolipid, sulfatide, is essential for maintenance of ion channels on myelinated axon but not essential for initial cluster formation. J. Neurosci. 22:6507, 2002. 67. Iyer, S., Bianchi, R., and Eichberg, J.: Tyrosine phosphorylation of PNS myelin P(0) occurs in the cytoplasmic domain and is maximal during early development. J. Neurochem. 75:347, 2000. 68. Iyer, S., Rowe-Rendleman, C. L., Bianchi, R., and Eichberg, J.: Tyrosine phosphorylation of myelin protein P0. J Neurosci Res. 46:531, 1996. 69. Jani, A., Menichella, D., Jiang, H., et al.: Modulation of cellmediated immunity prolongs adenovirus-mediated transgene expression in sciatic nerve. Hum. Gene Ther. 10:787, 1999. 70. Kamholz, J., Menichella, D., Jani, A., et al.: Charcot-MarieTooth disease type 1: molecular pathogenesis to gene therapy. Brain 123(Pt. 2):222, 2000. 71. Karch, S. B., and Urich, H.: Infantile polyneuropathy with defective myelination: an autopsy study. Dev. Med. Child Neurol. 17:504, 1975. 72. Kennedy, W. R., Sung, J. H., and Berry, J. F.: A case of congenital hypomyelination neuropathy: clinical, morphological, and chemical studies. Arch. Neurol. 34:337, 1977. 73. Kirschner, D. A., and Ganser, A. L.: Compact myelin exists in the absence of basic protein in the shiverer mutant mouse. Nature 283:207, 1980. 74. Kirschner, D. A., and Saavedra, R. A.: Mutations in demyelinating peripheral neuropathies support molecular model of myelin P0-glycoprotein extracellular domain. J. Neurosci. Res. 39:63, 1994. 75. Kirschner, D. A., Szumowski, K., Gabreels-Festen, A. A., et al.: Inherited demyelinating peripheral neuropathies: relating myelin packing abnormalities to P0 molecular defects. J. Neurosci. Res. 46:502, 1996. 76. Kitamura, K., Suzuki, A., and Uyemura, K.: Purification and partial characterization of two glycoproteins in bovine peripheral nerve myelin membrane. Biochim. Biophys. Acta 455:806, 1976. 77. Komiyama, A., Ohnishi, A., Izawa, K., et al.: De novo mutation (Arg98:Cys) of the myelin P0 gene and uncompaction of the major dense line of the myelin sheath in a severe variant of Charcot-Marie-Tooth disease type 1B. J. Neurol. Sci. 149:103, 1997. 78. Krajewski, K. M., Lewis, R. A., Fuerst, D. R., et al.: Neurological dysfunction and axonal degeneration in CharcotMarie-Tooth disease type 1A. Brain 123(Pt. 7):1516, 2000. 79. Kramer, R., Zhang, Y., Gehrmann, J., et al.: Gene transfer through the blood-nerve barrier: NGF-engineered neuritogenic T lymphocytes attenuate experimental autoimmune neuritis. Nat. Med. 1:1162, 1995.

1704

Diseases of the Peripheral Nervous System

80. Lagueny, A., Latour, P., Vital, A., et al.: Mild recurrent neuropathy in CMT1B with a novel nonsense mutation in the extracellular domain of the MPZ gene. J. Neurol. Neurosurg. Psychiatry 70:232, 2001. 81. Latour, P., Blanquet, F., Nelis, E., et al.: Mutations in the myelin protein zero gene associated with Charcot-MarieTooth disease type 1B. Hum. Mutat. 6:50, 1995. 82. Lemke, G., and Axel, R.: Isolation and sequence of a cDNA encoding the major structural protein of peripheral myelin. Cell 40:501, 1985. 83. Lemke, G., and Chao, M.: Axons regulate Schwann cell expression of the major myelin and NGF receptor genes. Development 102:499, 1988. 84. Lemke, G., Lamar, E., and Patterson, J.: Isolation and analysis of the gene encoding peripheral myelin protein zero. Neuron 1:73, 1988. 85. Lewis, R. A., Sumner, A. J., and Shy, M. E.: Electrophysiological features of inherited demyelinating neuropathies: a reappraisal in the era of molecular diagnosis. Muscle Nerve 23:1472, 2000. 86. Lupski, J. R.: DNA diagnostics for Charcot-Marie-Tooth disease and related inherited neuropathies. Clin. Chem. 42:995, 1996. 87. Lupski, J. R., de Oca-Luna, R. M., Slaugenhaupt, S., et al.: DNA duplication associated with Charcot-Marie-Tooth disease type 1A. Cell 66:219, 1991. 88. Lyon, G.: Ultrastructural study of a nerve biopsy from a case of early infantile chronic neuropathy. Acta Neuropathol. 13:131, 1969. 89. Mandich, P., Mancardi, G. L., Varese, A., et al.: Congenital hypomyelination due to myelin protein zero Q215X mutation. Ann. Neurol. 45:676, 1999. 90. Marques, W. Jr., Hanna, M. G., Marques, S. R., et al.: Phenotypic variation of a new P0 mutation in genetically identical twins. J. Neurol. 246:596, 1999. 91. Marrosu, M. G., Vaccargiu, S., Marrosu, G., et al.: CharcotMarie-Tooth disease type 2 associated with mutation of the myelin protein zero gene. Neurology 50:1397, 1998. 92. Martini, R., Mohajeri, M. H., Kasper, S., et al.: Mice doubly deficient in the genes for P0 and myelin basic protein show that both proteins contribute to the formation of the major dense line in peripheral nerve myelin. J. Neurosci. 15:4488, 1995. 93. Mastaglia, F. L., Nowak, K. J., Stell, R., et al.: Novel mutation in the myelin protein zero gene in a family with intermediate hereditary motor and sensory neuropathy. J. Neurol. Neurosurg. Psychiatry. 67:174, 1999. 94. Matsuyama, W., Nakagawa, M., Takashima, H., and Osame, M.: Altered trafficking and adhesion function of MPZ mutations and phenotypes of Charcot-Marie-Tooth disease 1B. Acta Neuropathol. (Berl.) 103:501, 2002. 95. Maurer, M., Schmid, C. D., Bootz, F., et al.: Bone marrow transfer from wild-type mice reverts the beneficial effect of genetically mediated immune deficiency in myelin mutants. Mol. Cell. Neurosci. 17:1094, 2001. 96. Meijerink, P. H., Hoogendijk, J. E., Gabreels-Festen, A. A., et al.: Clinically distinct codon 69 mutations in major myelin protein zero in demyelinating neuropathies. Ann. Neurol. 40:672, 1996. 97. Menichella, D. M., Arroyo, E. J., Awaframani, R., et al.: Protein zero is necessary for E-cadherin-mediated adherens

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111. 112.

113.

114.

junction formation in Schwann cells. Mol. Cell. Neurosci. 18:606, 2001. Misu, K., Yoshihara, T., Shikama, Y., et al.: An axonal form of Charcot-Marie-Tooth disease showing distinctive features in association with mutations in the peripheral myelin protein zero gene (Thr124Met or Asp75Val). J. Neurol. Neurosurg. Psychiatry 69:806, 2000. Misu, K., Yoshihara, T., Yamamoto, M., et al.: Two families of Charcot-Marie-Tooth disease with Adie’s pupil, axonal neuropathy and the Thr124Met mutation in the peripheral myelin protein zero gene [in Japanese]. Rinsho Shinkeigaku 40:149, 2000. Naef, R., and Suter, U.: Impaired intracellular trafficking is a common disease mechanism of PMP22 point mutations in peripheral neuropathies. Neurobiol. Dis. 6:1, 1999. Nakagawa, M., Suehara, M., Saito, A., et al.: A novel MPZ gene mutation in dominantly inherited neuropathy with focally folded myelin sheaths. Neurology 52:1271, 1999. Nelis, E., Haites, N., and Van Broeckhoven, C.: Mutations in the peripheral myelin genes and associated genes in inherited peripheral neuropathies. Hum. Mutat. 13:11, 1999. Nelis, E., Van Broeckhoven, C., De Jonghe, P., et al.: Estimation of the mutation frequencies in Charcot-MarieTooth disease type 1 and hereditary neuropathy with liability to pressure palsies: a European collaborative study. Eur. J. Hum. Genet. 4:25, 1996. Ohnishi, A., Aoki, A., Yamamoto, T., and Tsuji, S.: A case of Charcot-Marie-Tooth disease 1 B with Val 146Phe mutation of myelin protein zero showing a severe clinical phenotype [in Japanese]. Rinsho Shinkeigaku 40:268, 2000. Ouvrier, R.: Correlation between the histopathologic, genotypic, and phenotypic features of hereditary peripheral neuropathies in childhood. J. Child Neurol. 11:133, 1996. Ouvrier, R. A., McLeod, J. G., and Conchin, T. E.: The hypertrophic forms of hereditary motor and sensory neuropathy: a study of hypertrophic Charcot-Marie-Tooth disease (HMSN type I) and Dejerine-Sottas disease (HMSN type III) in childhood. Brain 110(Pt. 1):121, 1987. Pareyson, D., Menichella, D., Botti, S., et al.: Heterozygous null mutation in the P0 gene associated with mild CharcotMarie-Tooth disease. Ann. N. Y. Acad. Sci. 883:477, 1999. Peters, A., Palay, S. L., and Webster, H. D.: The Fine Structure of the Nervous System: Neurons and Their Supporting Cells. New York, Oxford University Press, 1991. Plante-Bordeneuve, V., Guiochon-Mantel, A., Lacroix, C., et al.: The Roussy-Levy family: from the original description to the gene. Ann. Neurol. 46:770, 1999. Previtali, S. C., Quattrini, A., Fasolini, M., et al.: Epitopetagged P(0) glycoprotein causes Charcot-Marie-Tooth-like neuropathy in transgenic mice. J. Cell Biol. 151:1035, 2000. Quarles, R. H.: Glycoproteins of myelin sheaths. J. Mol. Neurosci. 8:1, 1997. Rautenstrauss, B., Nelis, E., Grehl, H., et al.: Identification of a de novo insertional mutation in P0 in a patient with a Dejerine-Sottas syndrome (DSS) phenotype. Hum. Mol. Genet. 3:1701, 1994. Roa, B. B., Dyck, P. J., Marks, H. G., et al.: Dejerine-Sottas syndrome associated with point mutation in the peripheral myelin protein 22 (PMP22) gene. Nat. Genet. 5:269, 1993. Roussy, G., and Lévy, G.: A sept cas d’une maladie familiale particulaire. Rev. Neurol. 33:427, 1926.

Hereditary Motor and Sensory Neuropathies Related to MPZ (P0) Mutations 115. Rowe-Rendleman, C. L., and Eichberg, J.: P0 phosphorylation in nerves from normal and diabetic rats: role of protein kinase C and turnover of phosphate groups. Neurochem. Res. 19:1023, 1994. 116. Sakamoto, Y. K., et al.: Fatty acid linked peptides from bovine P0 protein in peripheral nerve myelin. Biomed. Res. 7:261, 1986. 117. Schiavon, F., Rampazzo, A., Merlini, L., et al.: Mutations of the same sequence of the myelin P0 gene causing two different phenotypes. Hum. Mutat. Suppl. 1:S217, 1998. 118. Schmid, C. D., Stienekemeier, M., Oehen, S., et al.: Immune deficiency in mouse models for inherited peripheral neuropathies leads to improved myelin maintenance. J. Neurosci. 20:729, 2000. 119. Senderek, J., Hermanns, B., Lehmann, U., et al.: CharcotMarie-Tooth neuropathy type 2 and P0 point mutations: two novel amino acid substitutions (Asp61Gly; Tyr119Cys) and a possible “hotspot” on Thr124Met. Brain Pathol. 10:235, 2000. 120. Senderek, J., Ramaekers, V. T., Zerres, K., et al.: Phenotypic variation of a novel nonsense mutation in the P0 intracellular domain. J. Neurol. Sci. 192:49, 2001. 121. Shames, I., Fraser, A., Colby, J., et al.: Phenotypic differences between peripheral myelin protein-22 (PMP22) and myelin protein zero (P0) mutations associated with CharcotMarie-Tooth-related diseases. J. Neuropathol. Exp. Neurol. 62:751, 2003. 122. Shapiro, L., Doyle, J. P., Hensley, P., et al.: Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17:435, 1996. 123. Shizuka, M., Ikeda, Y., Watanabe, M., et al.: A novel mutation of the myelin P(0) gene segregating Charcot-MarieTooth disease type 1B manifesting as trigeminal nerve thickening. J. Neurol. Neurosurg. Psychiatry 67:250, 1999. 124. Shy, M., Garbern, J., and Kamholz, J.: Hereditary motor and sensory neuropathies: a biological perspective. Lancet Neurol. 1:110, 2002. 125. Shy, M. E., et al.: Early and late onset neuropathy in CMT1B. 2003 (submitted). 126. Shy, M. E., Arroyo, E., Sladky, J., et al.: Heterozygous P0 knockout mice develop a peripheral neuropathy that resembles chronic inflammatory demyelinating polyneuropathy (CIDP). J. Neuropathol. Exp. Neurol. 56:811, 1997. 127. Shy, M. E., Tani, M., Shi, Y. J., et al.: An adenoviral vector can transfer lacZ expression into Schwann cells in culture and in sciatic nerve. Ann. Neurol. 38:429, 1995. 128. Silander, K., Meretoja, P., Nelis, E., et al.: A de novo duplication in 17p11.2 and a novel mutation in the Po gene in two Dejerine-Sottas syndrome patients. Hum. Mutat. 8:304, 1996. 129. Silander, K., Meretoja, P., Juvonen, V., et al.: Spectrum of mutations in Finnish patients with Charcot-Marie-Tooth disease and related neuropathies. Hum. Mutat. 12:59, 1998. 130. Skre, H.: Genetic and clinical aspects of Charcot-MarieTooth’s disease. Clin. Genet. 6:98, 1974. 131. Sorour, E., and Upadhyaya, M.: Mutation analysis in Charcot-Marie-Tooth disease type 1 (CMT1). Hum. Mutat. Suppl. 1:S242, 1998. 132. Street, V. A., Meekins, G., Lipe, H. P., et al.: CharcotMarie-Tooth neuropathy: clinical phenotypes of four novel mutations in the MPZ and Cx 32 genes. Neuromuscul. Disord. 12:643, 2002.

1705

133. Suter, U., Welcher, A. A., Ozcelik, T., et al.: Trembler mouse carries a point mutation in a myelin gene. Nature 356:241, 1992. 134. Suzuki, M., Sakamoto, Y., Kitamura, K., et al.: Phosphorylation of P0 glycoprotein in peripheral nerve myelin. J. Neurochem. 55:1966, 1990. 135. Tachi, N., Ishikawa, Y., Tsuzuki, A., et al.: Hereditary motor and sensory neuropathy type I in early childhood: clinical features and the findings of nerve pathology [in Japanese]. No To Hattatsu 17:318, 1985. 136. Tachi, N., Kozuka, N., Ohya, K., et al.: A new mutation of the P0 gene in patients with Charcot-Marie-Tooth disease type 1B: screening of the P0 gene by heteroduplex analysis. Neurosci. Lett. 204:173, 1996. 137. Tachi, N., Kozuka, N., Ohya, K., et al.: A small direct tandem duplication of the myelin protein zero gene in a patient with Dejerine-Sottas disease phenotype. J. Neurol. Sci. 156:167, 1998. 138. Takashima, H., Nakagawa, M., Kanzaki, A., et al.: Germline mosaicism of MPZ gene in Dejerine-Sottas syndrome (HMSN III) associated with hereditary stomatocytosis. Neuromuscul. Disord. 9:232, 1999. 139. Taroni, F., et al.: PMP22 and MPZ mutations in Italian families with hereditary neuropathy with liability to pressure palsies (HNPP) and Dejerine-Sottas disease (DSD). Mediz. Genetik. 9:171, 1997. 140. Thomas, P. K., Marques, W. Jr., Davis, M. B., et al.: The phenotypic manifestations of chromosome 17p11.2 duplication. Brain 120(Pt. 3):465, 1997. 141. Timmerman, V., Nelis, E., Van Hul, W., et al.: The peripheral myelin protein gene PMP-22 is contained within the Charcot-Marie-Tooth disease type 1A duplication. Nat. Genet. 1:171, 1992. 142. Toews, A. D., Fischer, H. R., Goodrum, J. F., et al.: Metabolism of phosphate and sulfate groups modifying the P0 protein of peripheral nervous system myelin. J. Neurochem. 48:883, 1987. 143. Tooth, H.: The peroneal type of progressive muscular atrophy. London, Lewis, 1886. 144. Trapp, B. D., Kidd, G. J., Hauer, P., et al.: Polarization of myelinating Schwann cell surface membranes: role of microtubules and the trans-Golgi network. J. Neurosci. 15(3 Pt. 1):1797, 1995. 145. Trapp, B. D., and Quarles, R. H.: Presence of the myelinassociated glycoprotein correlates with alterations in the periodicity of peripheral myelin. J. Cell Biol. 92:877, 1982. 146. Tyson, J., Ellis, D., Fairbrother, U., et al.: Hereditary demyelinating neuropathy of infancy: a genetically complex syndrome. Brain 120(Pt. 1):47, 1997. 147. Uyemura, K., Asou, H., and Takeda, Y.: Structure and function of peripheral nerve myelin proteins. Prog. Brain Res. 105:311, 1995. 148. Uyemura, K., Kitamura, K., and Miura, M.: Structure and molecular biology of P0 protein. In Martenson, R. E. (ed.): Myelin: Biology and Chemistry. Boca Raton, FL, CRC Press, pp. 481, 1992. 149. Wachneldt, R. V., and Malotka, J.: Presence of proteolipid protein in coelacanth brain myelin demonstrates tetrapod affinities and questions a chondrichthyan association. J. Neurochem. 52:1941, 1989.

1706

Diseases of the Peripheral Nervous System

150. Wachneldt, T. V., Matthieu, J. M., and Jeserich, G.: Major central nervous system myelin glycoprotein of the African lungfish (Protopterus dolloi) cross reacts with myelin proteolipid antibodies, indicating a close phylogenetic relationship with amphibians. J. Neurochem. 46:1387, 1986. 151. Warner, L. E., Hilz, M. J., Appel, S. H., et al.: Clinical phenotypes of different MPZ (P0) mutations may include Charcot-Marie-Tooth type 1B, Dejerine-Sottas, and congenital hypomyelination. Neuron 17:451, 1996. 152. Warner, L. E., Shohat, M., Shorer, Z., and Lupski, J. R.: Multiple de novo MPZ (P0) point mutations in a sporadic Dejerine-Sottas case. Hum. Mutat. 10:21, 1997. 153. Watanabe, M., Yamamoto, N., Ohkoshi, N., et al.: Corticosteroid-responsive asymmetric neuropathy with a myelin protein zero gene mutation. Neurology 59:767, 2002. 154. Watson, D. F., Nachtman, F. N., Kunel, R. W., and Griffin, J. W.: Altered neurofilament phosphorylation and beta tubulin isotypes in Charcot-Marie-Tooth disease type 1. Neurology 44:2383, 1994. 155. Wiggins, R. C., and Morell, P.: Phosphorylation and fucosylation of myelin protein in vitro by sciatic nerve from developing rats. J. Neurochem. 34:627, 1980. 156. Wong, M. H., and Filbin, M. T.: The cytoplasmic domain of the myelin P0 protein influences the adhesive interactions of its extracellular domain. J. Cell Biol. 126:1089, 1994. 157. Wong, M. H., and Filbin, M. T.: Dominant-negative effect on adhesion by myelin P0 protein truncated in its cytoplasmic domain. J. Cell Biol. 134:1531, 1996.

158. Wrabetz, L., Feltri, M. L., Quattrini, A., et al.: P(0) glycoprotein overexpression causes congenital hypomyelination of peripheral nerves. J. Cell Biol. 148:1021, 2000. 159. Xu, W., Shy, M., Kamholz, J., et al.: Mutations in the cytoplasmic domain of P0 reveal a role for PKC- mediated phosphorylation in adhesion and myelination. J. Cell Biol. 155:439, 2001. 160. Yan, W. X., Archelos, J. J., Hartung, H. P., and Pollard, J. D.: P0 protein is a target antigen in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:286, 2001. 161. Yin, X., Kidd, G. J., Wrabetz, L., et al.: Schwann cell myelination requires timely and precise targeting of P(0) protein. J. Cell Biol. 148:1009, 2000. 162. Young, P., Grote, K., Kuhlenbaumer, G., et al.: Mutation analysis in Chariot-Marie Tooth disease type 1: point mutations in the MPZ gene and the GJB1 gene cause comparable phenotypic heterogeneity. J. Neurol. 248:410, 2001. 163. Zhang, K., and Filbin, M. T.: Formation of a disulfide bond in the immunoglobulin domain of the myelin P0 protein is essential for its adhesion. J. Neurochem. 63:367, 1994. 164. Zhang, K., and Filbin, M. T.: Myelin P0 protein mutated at Cys21 has a dominant-negative effect on adhesion of wild type P0. J. Neurosci. Res. 53:1, 1998. 165. Zhang, K., Merazga, Y., and Filbin, M. T.: Mapping the adhesive domains of the myelin P0 protein. J. Neurosci. Res. 45:525, 1996.

72 Hereditary Motor and Sensory Neuropathy Related to Early Growth Response 2 (EGR2) Gene LAURA E. WARNER AND JAMES R. LUPSKI

EGR2 Gene Identification and Protein Structure EGR2 Function EGR2 Mutations and Myelinopathies

Functional Consequences of EGR2 Mutations R1 Inhibitory Domain NAB Proteins

In 1998, the early growth response 2 (EGR2) gene was added to the list of genes associated with the myelinopathies, Charcot-Marie-Tooth disease type I (CMT1), DejerineSottas neuropathy (DSN), and congenital hypomyelinating neuropathy (CHN), through the identification of point mutations in patient samples.59 Unlike three other myelinopathy-associated genes (PMP22, MPZ, and Cx32) that are myelin membrane–bound proteins, EGR2 encodes a transcription factor that is required for myelin gene expression and is important for myelin development and maintenance. The association of different mutations within the same gene with varying clinical phenotypes is a common finding in this group of myelinopathies and supports the contention that these disorders represent a spectrum of related phenotypes caused by an underlying defect in myelination of the peripheral nervous system (PNS).

Zinc Finger Domain Conclusions

cells after serum stimulation.14 The human EGR2 gene maps to chromosome 10q21q22, contains two coding exons (Genbank AF139463),14,60 and encodes a protein with two identifiable domains, an R1 inhibitory domain (amino acids 243 through 279) that is important for regulation of EGR2 function and a zinc finger domain (amino acids 322 through 427) that confers DNA-binding specificity (Fig. 72–1). The mouse orthologue, Egr2 (also known as Krox20), maps to the syntenic region in the mouse (chromosome 10, band 5) and also contains two coding exons (Genbank M24376 and M24377).7 Human EGR2 and mouse Egr2/Krox20 share a nucleotide identity of 86% in the coding region and an amino acid identity of 89% for the entire sequence, which increases to 95% in the R1 inhibitory domain and to 100% in the zinc finger domain (see Fig. 72–1). This extensive similarity suggests that Egr2/Krox20 and EGR2 perform similar functions in mice and humans.

EGR2 GENE IDENTIFICATION AND PROTEIN STRUCTURE

EGR2 FUNCTION

EGR2 is a member of a family of early growth response (EGR) genes that includes EGR143 (also know as NGFI-A,24 Krox24,20 or Zif2689), EGR3,31 and EGR4 (NGFI-C).12 EGR2 was initially identified using low-stringency hybridization with a mouse Egr1 complementary DNA (cDNA) probe on a human cDNA library constructed from RNA extracted from

The EGR proteins contain a Cys2His2-type zinc finger domain homologous to the one identified in the Drosophila melanogaster Krüppel (Kr) gap gene.34,36 The similarity with Kr suggested that the EGR proteins were DNA-binding proteins involved in genetic control at the transcriptional level. Experimental evidence has shown 1707

1708

Diseases of the Peripheral Nervous System

FIGURE 72–1 Structural domains of EGR2 and position of disease-associated mutations. Shown are the two identified domains of EGR2 (R1 domain and zinc finger domain) and the percent homology between mouse and human of these domains. The checkered box indicates the R1 domain that binds the NAB corepressors; black boxes indicate the three tandem zinc fingers. Below the structural representation is shown the position of the EGR2 mutations in functional regions of the EGR2 protein, as indicated by the asterisk, along with the inheritance pattern and associated phenotype. dom ⫽ dominant; rec ⫽ recessive; Zn ⫽ zinc finger.

that the EGR proteins mediate transcription of genes in a variety of signaling pathways that are important for cellular growth and differentiation and responses to hormonal stimuli.19,41,46,53,54 Egr2/Krox20 is an immediate-early response gene encoding a protein that binds DNA in a sequencespecific manner and acts as a transcription factor.7,8,29,56 Stable expression of Egr2/Krox20 is specifically associated with the onset of myelination in the PNS.63 Several transcription factors have been implicated in the establishment of myelinating Schwann cells: PAX3, a paired domain–containing protein; SCIP (suppressed cAMPinducible POU), a POU-domain homeodomain protein; SOX10, a high-mobility group–containing protein; and EGR2, a Cys2His2 transcription factor.16,39 PAX3 is important for the differentiation of embryonic Schwann cells into myelinating or nonmyelinating forms.15 SCIP is transiently required in the promyelinating stage that immediately precedes the differentiation into either myelinating or nonmyelinating cells.25,26,40 SOX10 functions synergistically with SCIP and PAX3 and partially represses EGR2, suggesting that SOX10 acts as a transcriptional modulator of glial development.16 EGR2 expression is restricted to myelinating Schwann cells and is essential for the final differentiation of myelinating Schwann cells.62 Egr2⫺/⫺ (Krox20⫺/⫺) homozygous knockout mice display disrupted hindbrain segmentation and development41,46 and a block of Schwann cell differentiation at an early stage.53 Schwann cells ensheath individual axons, but fail to form the spiral of membrane that becomes the myelin sheath.53 In addition, there is reduced expression of some of the major components of compacted myelin, such as myelin protein zero (MPZ) and myelin basic protein (MBP).53 Based on these obser-

vations, Egr2/Krox20 is speculated to activate genes required for peripheral nerve myelination. Using microarray expression profiling, Nagarajan et al.28 have demonstrated in Schwann cells that Egr2/Krox20 expression alone is sufficient for the activation of a number of myelin proteins, such as MPZ, MBP, peripheral myelin protein 22 (PMP22), periaxin (PRX), connexin 32 (GJB1/Cx32), and myelin-associated glycoprotein (MAG). Surprisingly, Egr2/Krox20 also appears to regulate the expression of several enzymes responsible for the formation of critical lipid components of myelin, as well as growth factors and their receptors, which could provide trophic and mitogenic support to Schwann cells.28 Egr2/Krox20 expression resulted in a 3- to 59-fold induction of these genes, indicating that different promoters have varied Egr2/Krox20 responsiveness. Two EGR2 binding sites have been identified in the GJB1/Cx32 promoter.5,27 The sequence of the GJB1 EGR2 binding sites differ slightly from the consensus EGR2 binding site and therefore lowers the EGR2 binding affinity by twofold.27 Variations in EGR2 binding affinity to promoters may explain the observed differences in fold induction.28 In addition, Bondurand et al.5 have shown that EGR2 acts synergistically with SOX10 to activate GJB1 expression. Therefore, the regulation of EGR2-responsive genes appears to be a complex process that is affected by variations in the EGR2 binding site and interactions with other transcription cofactors and modulators.

EGR2 MUTATIONS AND MYELINOPATHIES To date, one recessive and six dominant mutations have been identified in EGR2 and are associated with a continuum of myelinopathies ranging from severe CHN to

1709

Hereditary Motor and Sensory Neuropathy Related to EGR 2 Gene

Table 72–1. EGR2 Mutations Gene Mutation

Amino Acid Substitution

Position

Phenotype

References Warner et al.59 Bellone et al.1 Boerkoel et al.3 Taroni et al.51 Timmerman et al.52 Timmerman et al.52 Takashima et al.50 Yoshihara et al.61 Latour et al.17 Pareyson et al.30

803T ⬎ A 1064A ⬎ T 1075C ⬎ T

I268N D355V R359W

R1 inhibitory domain 1st zinc finger 1st zinc finger

CHN CMT1—severe DSN—plus cranial nerve involvement

1086A ⬎ C

R362R

1st zinc finger

Polymorphism

1141C ⬎ T 1142G ⬎ A

R381C R381H

2nd zinc finger 2nd zinc finger

1146T ⬎ G; 1147G ⬎ T 1225C ⬎ T 1352G ⬎ T

S382R; D383Y R409W G451V

2nd zinc finger

CMT1—late onset CMT1—plus cranial nerve involvement and congenital hypotonia CHN

3rd zinc finger C-terminal region

CMT1 Polymorphism

Warner et al.59 Warner et al.59 Takashima et al.50

CHN ⫽ congenital hypomyelinating neuropathy; CMT1 ⫽ Charcot-Marie-Tooth disease, type 1; DSN ⫽ Dejerine-Sottas neuropathy.

typical CMT1 (Table 72–1; see Fig. 72–1). A subset of the clinical features for each mutation is listed in Table 72–2. The family with the recessive Ile268Asn (I268N) mutation had three affected siblings from a consanguineous marriage. All three affected children were diagnosed with CHN and presented with similar clinical histories.59 They were floppy at birth, had delayed motor milestones and now walk with the aid of crutches, have gait abnormalities because of bilateral footdrop, and have marked wasting of all muscle, particularly distally. Motor nerve conduction velocities (MNCVs), when recordable, were 3 m/s, and sural nerve biopsies showed the absense of myelin in virtually all fibers. The six dominant mutations are Asp355Val (D355V), Arg359Trp (R359W), Arg381Cys (R381C), Arg381His (R381H), Arg409Trp (R409W), and the complex allele Ser382Arg ⫹ Asp383Tyr (S382R ⫹ D383Y) (see Table 72–1). The de novo heterozygous D355V mutation is associated with a severe CMT1 phenotype. The patient developed symptoms late in the first decade of life. These symptoms included diffuse muscle weakness, progressive distal muscle wasting, pes cavus, absent deep tendon reflexes, MNCVs less than 19 m/s, and mild loss of myelinated fibers and onion bulb formation (OBF).1 Several patients have been identified with the R359W mutation.3,51,52 All the R359W mutations are sporadic and result from the same 1075C ⬎ T point mutation (see Table 72–1). The phenotype observed with the R359W mutation is most consistent with DSN, but can range from typical DSN to a more rapidly progressive neuropathy that can cause death by 6 years of age.3 Moreover, in contrast to patients with typical DSN, patients with the R359W mutation have more respiratory compromise and cranial nerve

involvement. Similarly, the R381H mutation is also associated with cranial nerve involvement. The father and daughter with the R381H mutation had a history of gait abnormalities from infancy, MNCVs of 16 to 28 m/s, progressive wasting and weakness of the distal limb muscles, severe loss of myelinated fibers, OBF, and cranial nerve involvement.30 An unrelated 9-year-old girl with the R381H mutation also showed cranial nerve involvement. She presented with a severe peripheral neuropathy, congenital hypotonia, and Duane syndrome, a congenital eyemovement disorder characterized by the failure of cranial nerve VI to develop normally.17 The clinical involvement of multiple cranial nerves is not normally associated with CMT1 or DSN and may indicate a role for EGR2 in brainstem development, similar to that attributed to Egr2/Krox20.41,46 However, magnetic resonance imaging studies of several patients have failed to show any overt anatomic brain abnormalities.30,59 Nevertheless, these findings set a precedent that patients with EGR2 mutations should be tested for cranial nerve involvement and patients with cranial nerve abnormalities along with a peripheral neuropathy should be tested for EGR2 mutations. A second mutation at the amino acid residue R381 (R381C) has been identified in a late-onset CMT1 patient. The patient presented at age 59 with numbness in the distal extremities, distal limb muscle weakness, pes cavus, MNCVs of 17.5 to 27.3 m/s, severe loss of myelinated fibers, and OBF. No cranial nerve involvement was present. The patients in the family segregating the R409W mutation exhibited typical CMT1, with disease onset in the second decade of life and MNCVs ranging from 26 to 42 m/s.59 The patient heterozygous for the complex

Sporadic

DSN

DSN

DSN

Late-onset CMT1

CMT1 ⫹ congenital hypotonia CMT1

1091/R359W Boerkoel et al.3

R359W Taroni et al.51

R359W Timmerman et al.52

R381C Yoshihara et al.61

R381H Latour et al.17

Sporadic

Dominant

CHN

CMT1

Dominant

Sporadic

Sporadic

Sporadic

Sporadic

15 yr

Infancy

Infancy

Infancy

59 yr

Birth

Infancy

Infancy

Birth

10 yr

Birth

Age at Onset

NA

NA

NA

NA

NA

RF (age 16)

NA

NA

RF (age 6)

NA

NA

Cause of Death†

Distal dominant, severe Distal dominant, severe NR

Distal dominant, severe Distal dominant, severe Distal dominant, severe Distal dominant, severe Distal dominant, severe Distal dominant, severe Distal dominant, mild NR

Motor Involvement

NR

NR

⬎2 yr NR

Yes

Yes

No

Yes

NR

Yes

Yes

NR

NR

CNI‡

NR

NR

Normal

3

NR

1.3 yr

Never

NR

Delayed

Age of Walking

*CHN ⫽ congenital hypomyelinating neuropathy; CMT1 ⫽ Charcot-Marie-Tooth disease, type 1; DSN ⫽ Dejerine-Sottas neuropathy. † NA ⫽ not applicable; RF ⫽ respiratory failure. ‡ CNI ⫽ cranial nerve involvement; NR ⫽ not reported. § MNCV ⫽ motor nerve conduction velocity. || MF ⫽ myelinated fibers; OBF ⫽ onion bulb formation.

1085/S382R, D383Y Warner et al.59 615/R409W Warner et al.59

R381H Pareyson et al.30

Sporadic

DSN

1083/R359W Boerkoel et al.3

Sporadic

Severe CMT1

D355V Bellone et al.1

Recessive

Inheritance

CHN

Diagnosis*

904-906/I268N Warner et al.59

Patient(s)/Mutation Reference

Table 72–2. Clinical Features of Patients with EGR2 Mutations

Absence of MF, OBF NR

ⱕ32

Severe loss of MF, OBF

ⱕ28 ⱕ8

NR

NR

Severe loss of MF, OBF

ⱕ27

Severe loss of MF, OBF

ⱕ8

Severe loss of MF, OBF

Severe loss of MF, OBF

ⱕ8

8

Severe loss of MF, OBF

Mild loss of MF, OBF

⬍19 Not detectable

Absence of MF

Peripheral Nerve Histopathology||

3

MNCV§ (m/s)

Hereditary Motor and Sensory Neuropathy Related to EGR 2 Gene

de novo mutation S382R ⫹ D383Y presented with infantile hypotonia, significant motor developmental delay, MNCVs of 8 m/s or less, and profound absence or loss of myelin in virtually all axons along with OBF.59 A benign polymorphism, R362R, was identified in the first zinc finger.50,52 A G451V variation in the extreme C-terminal region outside of the zinc finger domain was identified in two neuropathy patients.50 However, this mutation does not segregate with the disease phenotype in either family, and therefore is likely a benign polymorphism. Of the 11 unrelated patients reported with EGR2 mutations, eight are caused by transition mutations: four 1075C ⬎ T transitions causing R359W,3,51,52 two 1142G ⬎ A transitions causing R381H,17,30 one 1141C ⬎ T transition causing R381C,61 and one 1225C ⬎ T transition causing R409W.59 The other three mutations result from one 803T ⬎ A transversion (I268N),59 one 1064A ⬎ T transversion (D355V),1 and the 1146T ⬎ G and 1147G ⬎ T transversions (S382R ⫹ D383Y).59 All the transitions, as well as three of the five confirmed de novo mutations, occur at CpG dinucleotides. Methylation of mammalian DNA occurs on cytosines primarily in CpG dinucleotides. Spontaneous deamination of 5-methylcytosine to thymine at CpG dinucleotides is a common mutation mechanism,11 accounting for a third of the germline point mutations causing human disease10 and 73% of the EGR2 mutations. The increased frequency of mutations at 1075C could indicate that the 1075 cytosine is more frequently methylated or repaired less efficiently following spontaneous deamination.3

FUNCTIONAL CONSEQUENCES OF EGR2 MUTATIONS R1 Inhibitory Domain The differences in clinical severity and inheritance pattern seen with EGR2 mutations can conceivably be explained by the location of the mutations and their differing effects on protein structure and functions. The I268N mutation, which behaves as a recessive allele and causes CHN, occurs within the R1 domain that binds two transcriptional modulators, NAB138 and NAB2.44 The NAB proteins were initially identified as corepressors that inhibit the transcription activity of multiple members of the EGR gene family, including Egr1 and Egr2/Krox20.44,47 However, they appear to also act as coactivators in specific promoter contexts.42 A Phe substitution of the analogous Ile (Ile293) in Egr1 abolishes the inhibitory effect of NAB1 and NAB2 and increases the transcriptional activity of a reporter gene downstream of multiple EGR2 consensus sites by 15-fold.37 Similarly, the I268N mutation shows increased transcriptional activity of a luciferase reporter driven by multiple EGR2 consensus sites and abrogation of NAB repression.60 When the I268N

1711

mutant EGR2 protein was tested on the EGR2-responsive elements of the GJB1 promoter, no increase in transcriptional activity was observed.5 The difference in transcriptional activity from the different constructs could be caused by differing amounts of endogenous NAB proteins in the two cell lines (CV-1 or HeLa) used or lower binding affinity to the nonconsensus EGR2 binding sites of the GJB1 promoter. The loss of EGR2 transcriptional regulation suggested for the I268N mutation could result in increased/decreased expression of certain dosage-sensitive PNS gene(s), thereby affecting PNS myelination and causing the associated myelinopathy, CHN, in the same way that increased dosage of PMP22 expression by the CMT1A duplication results in CMT1.23

NAB Proteins The I268N mutation strongly indicates that the EGR2NAB interaction plays a critical role in regulating EGR2 target genes during the myelination process. Therefore, it was speculated that mutations within the NAB proteins might be associated with unexplained cases of human myelinopathies. The NAB proteins have two highly conserved domains, NAB conserved domain 1 (NCD1) and NAB conserved domain 2 (NCD2). The NCD1 domain mediates the interaction between the NAB proteins and the R1 domain of the EGR proteins,44 as well as the multimerization of NAB molecules.45 The NCD2 domain is critical for the repression function.47 The NAB repression mechanism is not one of interference in DNA binding by the EGR proteins. Instead, the NAB proteins have been shown to form a complex with EGR proteins that then binds to the binding sites.47 In addition, tethering experiments that bound NAB protein to a promoter in an EGRindependent manner showed active repression.47 Svaren et al.45 have identified dominant-negative mutations within the NCD1 region of NAB proteins that enhanced Egr1 activity by 20-fold. Theoretically, these NCD1 NAB mutations could have the same effect on transcriptional activity of EGR2 as mutations in the R1 domain of EGR2. Mutational analysis of the NCD1 region of both NAB1 and NAB2 in a number of patients with peripheral neuropathies revealed no mutations.55,57 Therefore, point mutations in the NCD1 region of the NAB proteins do not appear to play a role in this group of disorders.

Zinc Finger Domain All the reported dominant disease-associated mutations are found within the zinc finger domain and affect the DNA-binding properties of EGR2 (Fig. 72–2; see also Fig. 72–1). The EGR2 zinc finger domain is made up of three tandem Cys2His2 zinc fingers. Each Cys2His2 zinc finger consists of a conserved motif of 28 to 30 amino acids that includes an ␣-helix and H-C linker region

1712

Diseases of the Peripheral Nervous System

FIGURE 72–2 Placement of EGR2 mutations in the zinc finger domain. A, Circles represent individual amino acids in the zinc finger domain. C and H show the position of the cysteine and histidine residues, respectively, which bind the zinc ion (Zn⫹⫹). The black circles denote the location of conserved hydrophobic residues that are important for the structure of the zinc fingers. The boxed area in the first zinc finger shows the residues in each finger that form the ␣-helix and are important for DNA binding specificity. The amino acid substitutions caused by the identified mutations in EGR2 are shown. Standard one-letter abbreviations for amino acids are used. B, Shown are the amino acids within the zinc fingers that are important for DNA recognition within the major groove of DNA. Bold letters indicate the residues at positions 17 and 20 that determine type 1 and type 2 fingers. Pound signs (#) denote the residues that have been shown to directly bind DNA, and asterisks (*) indicate those that support DNA binding. The positions of the different myelinopathy-associated missense amino acid substitutions and their conveyed phenotypes are shown.

(Fig. 72–2A). The arginine (R) that immediately precedes the ␣-helix and some of the amino acids that make up the ␣-helix are important for DNA recognition within the major groove of DNA.2,6,13,18 Specific conserved amino acids adjacent to and within the ␣-helix of Egr2/Krox2029 and Egr1 (Zif268)32 have been shown by nuclear magnetic resonance modeling and crystallization to either hydrogen bond with the donor sites on guanine or support DNA

binding interactions (Fig. 72–2B). EGR2 is composed of two types of fingers arranged in the order type 1–type 2–type 1. Each finger type binds to a different 3-bp DNA sequence. The type 1 fingers bind to 5⬘-GCG-3⬘ and the type 2 finger binds to 5⬘-GGG-3⬘,29 thus making the EGR2 consensus binding site 5⬘-GCGGGGGCG-3⬘. However, some variability in the DNA binding sequence has been observed.48

Hereditary Motor and Sensory Neuropathy Related to EGR 2 Gene

The myelinopathy-associated mutations all occur in key zinc finger amino acids. The R359W, R381C, R381H, and R409W amino acid substitutions all affect arginine residues that have been shown to directly bind DNA (see Fig. 72–2). The R359W and R381H mutant proteins retain residual transcriptional activation on the GJB1 promoter and EGR2 consensus binding sites, suggesting they maintain some low-level DNA-binding activity.5,60 The R409W mutant protein is unable to bind the EGR2 consensus sequence or activate transcription, similar to a loss-offunction allele created from a nonsense mutation (R405X) that is missing the third zinc finger and C-terminal sequences.60 No experimental functional data are available for the R381C mutation; however, the milder phenotype and late onset associated with this mutation compared to the R381H mutation may imply that transcription of downstream genes by EGR2 may be less affected by the substitution to a cysteine at the R381 residue. Variations in clinical phenotype associated with different substitutions at the same amino acid position have also been observed in MPZ, another gene associated with this group of myelinopathies, and have been related to the differing effects each substitution has on protein function.58 The D355V and complex S382R ⫹ D383Y mutations occur at the aspartic acid residues that support DNA binding and as a result affect DNA-binding and transcriptional activity to a lesser extent. Musso et al.27 have shown that the D355V mutant protein shows a 3-fold reduction in DNA-binding activity on the EGR2 consensus site and a 10-fold reduction on the GJB1 promoter. The complex S382R ⫹ D383Y mutant protein shows reduced DNA binding affinity and a 50% reduction in transcriptional activity.5,60 As expected, the DNA-binding defect can be attributed solely to the mutation of the D383 amino acid residue that directly supports DNA binding.60 While the S382R ⫹ D383Y mutant protein can bind and activate transcription from the GJB1 promoter region or EGR2 consensus binding site, it appears unable to activate multiple myelin proteins (MAG, PRX, PMP22, and MPZ) from their chromosomal loci.28 This may result from the varied Egr2 responsiveness of the promoters, differences in mutant Egr2 binding affinity, and/or that the EGR2 consensus binding site and GJB1 promoter-reporter constructs used do not correctly indicate the effect the Egr2 mutant proteins have on the target promoters. The dominant nature of the EGR2 zinc finger domain mutations is interesting given that heterozygous Egr2/ Krox20⫺/⫺ knockout mice appear phenotypically normal.41,46 This suggests that, rather than acting as loss-of-function alleles, the zinc finger mutations may instead be acting as dominant-negative or gain-of-function alleles. Functional studies on the EGR2 zinc finger mutations using electrophoretic mobility shift assays reveal that disease severity correlates with the amount of residual DNA binding.60 Thus complete loss of binding generally is associated with CMT1,

1713

whereas the retention of residual binding causes a more severe myelinopathy (e.g., severe CMT1, DSN, and CHN). These observations lead to the hypothesis that mutations with residual DNA-binding activity result in a severe neuropathy via a dominant-negative mechanism.60 In support of a dominant-negative disease mechanism, dominant mutations of these same conserved amino acids in the zinc fingers of the WT1 gene have been associated with Denys-Drash syndrome (DDS)33 and are known to abolish the DNAbinding ability of the WT1 protein.21 The DDS mutations are proposed to act as dominant-negative alleles because patients with a deletion of the WT1 gene have a less severe phenotype22 and transcriptional activation by wild-type WT1 has been shown to be inhibited by coexpression of the DDS mutant WT1 protein.35 However, in vitro cotransfection of expression constructs of mutant and wild-type EGR2 in ratios up to 4:1, along with a reporter plasmid containing four EGR2 consensus binding sites upstream of the luciferase gene, in both CV-1 and rat Schwann cells failed to show any dominant-negative activity.28,60 One potential explanation for this result could be the inability of the promoter–luciferase reporter constructs to adequately recapitulate EGR2-mediated regulation of EGR2 target promoters from their native chromosomal loci. When the messenger RNA levels of Egr2/Krox20 target genes were evaluated after coexpression of wild-type Egr2/Krox20 and mutant Egr2/Krox20 (S382R ⫹ D383Y) in mouse Schwann cells, it was determined that the mutant Egr2/Krox20 protein was able to dramatically affect the ability of wild-type Egr2/Krox20 to regulate its target genes, even at a 1:1 ratio.28 The Egr2/Krox20 zinc finger mutant proteins show enhanced protein stability, which may explain how an equal gene dosage of mutant Egr2/Krox20 protein can effectively inhibit wild-type Egr2/Krox20 function in Schwann cells.28 The dominant-negative effect that mutant Egr2/Krox20 proteins have on wild-type Egr2/Krox20 activity is likely related to their ability to interact and sequester Egr2/Krox20 coactivators or corepressors, such as SOX10, NAB1, and NAB2, that are crucial for Egr2/Krox20 regulation of its various target genes.

CONCLUSIONS The myelinopathies associated with EGR2 mutations likely result from inappropriate expression of critical myelin proteins, such as PMP22, MPZ, GJB1, MBP, PRX, and MAG, which leads to defective myelination. The variations in clinical severity associated with these mutations may result from differences in the effect each mutation has on binding affinity or regulation of EGR2 target promoters. In general, mutations that retain DNA-binding activity act as dominant-negative alleles. The more severe DSN and CHN phenotypes conveyed by these dominant-negative mutations presumably reflect the ability of the mutant protein to

1714

Diseases of the Peripheral Nervous System

interfere with the normal wild-type copy. Complete loss-offunction alleles, as evidenced by absence of DNA binding, are associated with the less severe CMT1 phenotype. The observed haploinsufficiency effect of heterozygous loss-offunction mutations indicates a requirement for proper dosage of the EGR2 gene and suggests a role not only in myelin development, but also potentially in the maintenance of peripheral nerve function. The apparent absence of a neuropathy phenotype in heterozygous knockout mice could indicate either different transcription factor requirements for proper function in humans or that the Krox20 knockout animals may present with a subtle neuropathy at a later age. EGR2 mutations represent the first example of mutations of a transcription factor resulting in an inherited neuropathy and affirm the critical role of this protein in myelin development and maintenance. Careful quantification of mutant protein levels in patients with EGR2 mutations will reveal differences between the mutant alleles and may ultimately provide additional insight into the cause of the differing severity of the myelinopathies observed. Genes downstream of EGR2 (e.g., PRX4,49 and others that remain to be identified) are prime candidates for neuropathy “disease genes” in patients in whom a molecular genetic etiology for the disease process remains to be elucidated.

REFERENCES 1. Bellone, E., Di Maria, E., Soriani, S., et al.: A novel mutation (D355V) in the early growth response 2 gene is associated with severe Charcot-Marie-Tooth type 1 disease. Hum. Mutat. 14:353, 1999. 2. Berg, J. M.: Proposed structure for the zinc-binding domains from transcription factor IIIA and related proteins. Proc. Natl. Acad. Sci. U. S. A. 85:99, 1988. 3. Boerkoel, C. F., Takashima, H., Bacino, C. A., et al.: EGR2 mutation R359W causes a spectrum of Dejerine-Sottas neuropathy. Neurogenetics 3:153, 2001. 4. Boerkoel, C. F., Takashima, H., Stankiewicz, P., et al.: Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am. J. Hum. Genet. 68:325, 2001. 5. Bondurand, N., Girard, M., Pingault, V., et al.: Human connexin 32, a gap junction protein altered in the X-linked form of Charcot-Marie-Tooth disease, is directly regulated by the transcription factor SOX10. Hum. Mol. Genet. 10:2783, 2001. 6. Brown, R. S., and Argos, P.: Fingers and helices. Nature 324:215, 1986. 7. Chavrier, P., Janssen-Timmen, U., Mattéi, M.-G., et al.: Structure, chromosome location, and expression of the mouse zinc finger gene Krox-20: multiple gene products and coregulation with the proto-oncogene c-fos. Mol. Cell. Biol. 9:787, 1989. 8. Chavrier, P., Zerial, M., Lemaire, P., et al.: A gene encoding a protein with zinc fingers is activated during G0/G1 transition in cultured cells. EMBO J. 7:29, 1988.

9. Christy, B. A., Lau, L. F., and Nathans, D.: A gene activated in mouse 3T3 cells by serum growth factors encodes a protein with “zinc finger” sequences. Proc. Natl. Acad. Sci. U. S. A. 85:7857, 1988. 10. Cooper, D. N., and Youssoufian, H.: The CpG dinucleotide and human genetic disease. Hum. Genet. 78:151, 1988. 11. Coulondre, C., Miller, J. H., Farabaugh, P. J., and Gilbert, W.: Molecular basis of base substitution hotspots in Escherichia coli. Nature 274:775, 1978. 12. Crosby, S. D., Puetz, J. J., Simburger, K. S., et al.: The early response gene NGFI-C encodes a zinc finger transcriptional activator and is a member of the GCGGGGGCG (GSG) element-binding protein family. Mol. Cell. Biol., 11:3835, 1991. 13. Gibson, T. J., Postma, J. P. M., Brown, R. S., and Argos, P.: A model for the tertiary structure of the 28 residue DNA-binding motif (‘zinc finger’) common to many eukaryotic transcriptional regulatory proteins. Protein Eng. 2:209, 1988. 14. Joseph, L. J., Le Beau, M. M., Jamieson G. A. Jr., et al.: Molecular cloning, sequencing, and mapping of EGR2, a human early growth response gene encoding a protein with “zinc-binding finger” structure. Proc. Natl. Acad. Sci. U. S. A. 85:7164, 1988. 15. Kioussi, C., Gross, M. K., and Gruss, P.: Pax3: a paired domain gene as a regulator of PNS myelination. Neuron 15:553, 1995. 16. Kuhlbrodt, K., Herbarth, B., Sock, E., et al.: Sox10, a novel transcription modulator in glial cells. J. Neurosci. 18:237, 1998. 17. Latour, P., Gatignol, A., Boutrand, L., et al.: A R381H mutation in the EGR2 gene associated with a severe peripheral neuropathy with hypotonia. J. Periph. Nerv. Syst. 4:293, 1999. 18. Lee, M. S., Gippert, G. P., Soman, K. V., et al.: Threedimensional solution structure of a single zinc finger DNAbinding domain. Science 245:635, 1989. 19. Lee, S. L., Sadovsky, Y., Swirnoff, A. H., et al.: Luteinizing hormone deficiency and female infertility in mice lacking the transcription factor NGFI-A (Egr-1). Science 273:1219, 1996. 20. Lemaire, P., Revelant, O., Bravo, R., and Charnay, P.: Two mouse genes encoding potential transcription factors with identical DNA-binding domains are activated by growth factors in cultured cells. Proc. Natl. Acad. Sci. U. S. A. 85:4691, 1988. 21. Little, M., Holmes, G., Bickmore, W., et al.: DNA binding capacity of the WT1 protein is abolished by Denys-Drash syndrome WT1 point mutations. Hum. Mol. Genet. 4:351, 1995. 22. Little, M. H., Williamson, K. A., Mannens, M., et al.: Evidence that WT1 mutations in Denys-Drash syndrome patients may act in a dominant-negative fashion. Hum. Mol. Genet. 2:259, 1993. 23. Lupski, J. R.: Charcot-Marie-Tooth disease: a gene-dosage effect. Hosp. Pract. 32:83, 1997. 24. Milbrandt, J.: A nerve growth factor-induced gene encodes a possible transcriptional regulatory factor. Science 238:797, 1987. 25. Monuki, E. S., Kuhn, R., Weinmaster, G., et al.: Expression and activity of the POU transcription factor SCIP. Science 249:1300, 1990. 26. Monuki, E. S., Weinmaster, G., Kuhn, R., and Lemke, G.: SCIP: a glial cell POU domain gene regulated by cyclic AMP. Neuron 3:783, 1989. 27. Musso, M., Balestra, P., Bellone, E., et al.: The D355V mutation decreases EGR2 binding to an element within the Cx32 promoter. Neurobiol. Dis. 8:700, 2001.

Hereditary Motor and Sensory Neuropathy Related to EGR 2 Gene 28. Nagarajan, R., Svaren, J., Le, N., et al.: EGR2 mutations in inherited neuropathies dominant-negatively inhibit myelin gene expression. Neuron 30:355, 2001. 29. Nardelli, J., Gibson, T. J., Vesque, C., and Charnay, P.: Base sequence discrimination by zinc-finger DNA-binding domains. Nature 349:175, 1991. 30. Pareyson, D., Taroni, F., Botti, S., et al.: Cranial nerve involvement in CMT disease type 1 due to early growth response 2 gene mutation. Neurology 54:1696, 2000. 31. Patwardhan, S., Gashler, A., Siegel, M.G., et al.: EGR3, a novel member of the Egr family of genes encoding immediate-early transcription factors. Oncogene 6:917, 1991. 32. Pavletich, N. P., and Pabo, C. O.: Zinc finger-DNA recognition: crystal structure of a Zif268-DNA complex at 2.1Å. Science 252:809, 1991. 33. Pelletier, J., Bruening, W., Kashtan, C. E., et al.: Germline mutations in the Wilms’ tumor suppressor gene are associated with abnormal urogenital development in Denys-Drash syndrome. Cell 67:437, 1991. 34. Preiss, A., Rosenberg, U. B., Kienlin, A., et al.: Molecular genetics of Krüppel, a gene required for segmentation of the Drosophila embryo. Nature 313:27, 1985. 35. Reddy, J. C., Morris, J. C., Wang, J., et al.: WT1-mediated transcriptional activation is inhibited by dominant negative mutant proteins. J. Biol. Chem. 270:10878, 1995. 36. Rosenberg, U. B., Schröder, C., Preiss, A., et al.: Structural homology of the product of the Drosophila Krüppel gene with Xenopus transcription factor IIIA. Nature 319:336, 1986. 37. Russo, M. W., Matheny, C., and Milbrandt, J.: Transcriptional activity of the zinc finger protein NGFI-A is influenced by its interaction with a cellular factor. Mol. Cell. Biol. 13:6858, 1993. 38. Russo, M. W., Sevetson, B. R., and Milbrandt, J.: Identification of NAB1, a repressor of NGFI-A-and Krox20mediated transcription. Proc. Natl. Acad. Sci. U. S. A. 92:6873, 1995. 39. Scherer, S. S.: The biology and pathobiology of Schwann cells. Curr. Opin. Neurol. 10:386, 1997. 40. Scherer, S. S., Wang, D. Y., Kuhn, R., et al.: Axons regulate Schwann cell expression of the POU transcription factor SCIP. J. Neurosci. 14:1930, 1994. 41. Schneider-Maunoury, S., Topilko, P., Seitanidou, T., et al.: Disruption of Krox-20 results in alteration of rhombomeres 3 and 5 in the developing hindbrain. Cell 75:1199, 1993. 42. Sevetson, B. R., Svaren, J., and Milbrandt, J.: A novel activation function for NAB proteins in EGR-dependent transcription of the luteinizing hormone beta gene. J. Biol. Chem. 275:9749, 2000. 43. Sukhatme, V. P., Cao, X., Chang, L. C., et al.: A zinc fingerencoding gene coregulated with c-fos during growth and differentiation, and after cellular depolarization. Cell 53:37, 1988. 44. Svaren, J., Sevetson, B. R., Apel, E. D., et al.: NAB2, a corepressor of NGFI-A (Egr-1) and Krox20, is induced by proliferative and differentiative stimuli. Mol. Cell. Biol. 16:3545, 1996. 45. Svaren, J., Sevetson, B. R., Golda, T., et al.: Novel mutants of NAB corepressors enhance activation by Egr transactivators. EMBO J. 17:6010, 1998. 46. Swiatek, P. J., and Gridley, T.: Perinatal lethality and defects in hindbrain development in mice homozygous for a targeted mutation of the zinc finger gene Krox20. Genes Dev. 7:2071, 1993.

1715

47. Swirnoff, A. H., Apel, E. D., Svaren, J., et al.: Nab1, a corepressor of NGFI-A (Egr-1), contains an active transcriptional repression domain. Mol. Cell. Biol. 18:512, 1998. 48. Swirnoff, A. H., and Milbrandt, J.: DNA-binding specificity of NGFI-A and related zinc finger transcription factors. Mol. Cell. Biol. 15:2275, 1995. 49. Takashima, H., Boerkoel, C. F., De Jonghe, P., et al.: Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51:709, 2002. 50. Takashima, H., Boerkoel, C. F., and Lupski, J. R.: Screening for mutations in a genetically heterogeneous disorder: DHPLC versus DNA sequence for mutation detection in multiple genes causing Charcot-Marie-Tooth neuropathy. Genet. Med. 3:335, 2001. 51. Taroni, F., Pareyson, D., Botti, S., et al.: Mutations in the Schwann cell transcription factor EGR2/Krox-20 in patients with severe hereditary demyelinating neuropathies. Neurology 52(Suppl. 2):A258, 1999. 52. Timmerman, V., De Jonghe, P., Ceuterick, C., et al.: Novel missense mutation in the early growth response 2 gene associated with a Dejerine-Sottas syndrome phenotype. Neurology 52:1827, 1999. 53. Topilko, P., Schneider-Maunoury, S., Levi, G., et al.: Krox-20 controls myelination in the peripheral nervous system. Nature 371:796, 1994. 54. Tourtellotte, W. G., and Milbrandt, J.: Sensory ataxia and muscle spindle agenesis in mice lacking the transcription factor Egr3. Nat. Genet. 20:87, 1998. 55. Venken, K., Di Maria, E., Bellone, E., et al.: Search for mutations in the EGR2 corepressor proteins, NAB1 and NAB2, in human peripheral neuropathies. Neurogenetics 4:37, 2002. 56. Vesque, C., and Charnay, P.: Mapping functional regions of the segment-specific transcription factor Krox-20. Nucleic Acids Res. 20:2485, 1992. 57. Warner, L. E.: Mutational analysis and elucidation of the molecular mechanisms resulting in human myelinopathies. Ph.D. thesis, Baylor College of Medicine, Houston, TX, Chapter 8, p. 193, 1999. 58. Warner, L. E., Hilz, M. J., Appel, S. H., et al.: Clinical phenotypes of different MPZ (P0) mutations may include Charcot-Marie-Tooth type 1B, Dejerine-Sottas, and congenital hypomyelination. Neuron 17:451, 1996. 59. Warner, L. E., Mancias, P., Butler, I. J., et al.: Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet. 18:382, 1998. 60. Warner, L. E., Svaren, J., Milbrandt, J., and Lupski, J. R.: Functional consequences of mutations in the early growth response 2 (EGR2) gene correlate with severity of human myelinopathies. Hum. Mol. Genet. 8:1245, 1999. 61. Yoshihara, T., Kanda, F., Yamamoto, M., et al.: A novel missense mutation in the early growth response 2 gene associated with late-onset Charcot-Marie-Tooth disease type 1. J. Neurol. Sci. 184:149, 2001. 62. Zorick, T. S., and Lemke, G.: Schwann cell differentiation. Curr. Opin. Cell Biol. 8:870, 1996. 63. Zorick, T. S., Syroid, D. E., Arroyo, E., et al.: The transcription factors SCIP and Krox-20 mark distinct stages and cell fates in Schwann cell differentiation. Mol. Cell. Neurosci. 8:129, 1996.

73 HMSN II (CMT2) and Miscellaneous Inherited System Atrophies of Nerve Axon: Clinical–Molecular Genetic Correlates CHRISTOPHER J. KLEIN AND PETER J. DYCK

Definition and Overview Axonal Dysfunction and Pathologic Abnormality Caused by Neuronal (Axonal) or Schwann Cell Dysfunction Ascertainment of Axonally Affected Individuals Clinical and Molecular Mimickers of HMSN II Molecular Genetic Classification Dominant Inheritance: HMSN II Myelin Protein Mutations Leading to HMSN II

Neuronal (Axonal) Mutations Leading to HMSN II HMSN II Types with Chromosomal Localization without Gene Discovery Autosomal Recessive Inheritance: Axonal CMT HMSN IVA/CMT4A (GDAP1) Giant Axonal Neuropathy (GAN) Anderman’s Syndrome (SLC12A6) Friedreich’s Ataxia and Spinocerebellar Atrophy with Axonal Neuropathy (Frataxin [FRX] and TDP1)

DEFINITION AND OVERVIEW Hereditary motor and sensory neuropathy type II (HMSN II) is a group of autosomal dominant system atrophies of motor and sensory neurons (axons) caused by mutations in a number of genes. The group of disorders is also called Charcot-Marie-Tooth disease type 2 (CMT2), but that term is increasingly linked with molecular classifications and inheritance not strictly limited to autosomal dominant forms. Here we describe disorders designated as HMSN II and CMT2 and several miscellaneous inherited system atrophies of the nervous system with axonal neuropathy. The terms HMSN II and CMT2 are used interchangeably when appropriate. Common features of these system atrophies of nerve are as follows: (1) they are inherited as simple mendelian traits, with variability of clinical expression; (2) there is significant involvement of lower motor and primary sensory neurons with failure of development or slow atrophy of neurons

Autosomal Recessive Axonal Neuropathy (LMNA) Hereditary Axonal Neuropathies with Associated Conditions HMSN V (with Spastic Paraplegia) HMSN VI (with Optic Atrophy) HMSN VII (with Retinitis Pigmentosa) Infantile and Juvenile Neuroaxonal Dystrophy

(axons); (3) atrophy and degeneration of the nerve fibers typically begins distally and progresses from distal to proximal; and (4) a causative interstitial histopathologic process is not demonstrated. Based on clinical examination, patients with axonal disorders (HMSN II) are difficult to distinguish from those with demyelinating disorders (hereditary motor and sensory neuropathy type I [HMSN I]), the phenotypes being similar. Both have dorsal greater than plantar flexion weakness of the ankles with a variable degree of distal upper extremity weakness typically limited to the intrinsic hand muscles. High arches or hammer toes are common to both, as is limited sensory involvement. These patients typically do not have pain or positive sensory symptoms such as asleep-numbness or prickling. Although enlarged, palpable nerves may be seen in HMSN I, this is certainly not a reliable clinical finding. In type I, slow conduction velocity has proven to be the important clinically distinguishing test. In HMSN II the nerve conduction velocities are usually 1717

1718

Diseases of the Peripheral Nervous System

normal or barely abnormal in the nerves of the upper limbs (ⱖ38 m/s in ulnar motor forearm nerves) and only slightly slowed in the lower limbs, with reduced compound muscle action potential (CMAP) amplitudes. Among patients with HMSN I or other demyelinating forms, the conduction velocities of the upper extremity nerves are usually less than 38 m/s. On needle electromyography (EMG) examination, HMSN II patients have large, poorly recruited motor unit potentials in distal greater than proximal muscles with or without active denervation (e.g., fibrillation). The increased size (duration and amplitude) of motor units and the extent of fibrillation are generally more pronounced than in demyelinating forms, but this may depend on the age of the patient and the severity of the disease of the individual studied. In severe or late demyelinating disease, axonal involvement and similar needle EMG examination findings are seen, but typically nerve biopsies are not required to distinguish the disorders. A clinical diagnostic approach for the various forms of HMSN in

algorithm form is outlined in Figure 73–1 emphasizing the electrophysiology and the known genes.127 Significant advances have been made in identifying the chromosomal localization and characterizing the mutant genes in these disorders (Table 73–1). Combining this information with the increasing molecular understanding of nerve and other biologic systems has allowed for improved pathologic understanding. Despite the improved molecular understanding, the information has not yet allowed for better treatment, and the phenotypic characterization, including the pattern of inheritance, remains the mainstay for diagnosis and counseling. In general, disorders with dominant inheritance tend to have less multiple-system neurologic involvement beyond nerve. This is true among those with axonal or demyelinating phenotypes. The lifespan of patients with dominantly inherited axonal disorders is typically not reduced, and most of them can lead a near-normal life, in contrast to some dominantly inherited demyelinating

FIGURE 73–1 Summary of the various clinical and electrophysiologic correlates of different molecular genetic abnormalities of axonal forms of HMSN. (Modified from Kleopa, K. A., and Scherer, S. S.: Inherited neuropathies. Neurol. Clin. 20:679, 2002.)

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

1719

Table 73–1. HMSN II and CMT2 Inherited Neuropathies Disease

Locus

Gene

Function(s)

Autosomal Dominant (HMSN II and CMT2) A B C

1p35-36 3q13-22 12q24

KIF1B RAB7 —

Axonal transport Axonal transport —

D E & 1F

7p14 8p21

GARS NEFL

F I&J

7q11-21 1q22

— MPZ

tRNA synthetase Neurofilament organization — Myelin structural protein

P

3q13.1

—

—

Autosomal Recessive and X-Linked (CMT2) G&K H X1-linked

8q21.1 8q21.3 Xq13.1

GDAP1 — GJB1

Neuronal development — Gap junctions

X2-linked

Xp22.2

—

—

X3-linked

Xq26

—

—

Clinical

⫾ Foot ulcers ⫾ Diaphragm vocal paresis, several with infant onset Allelic to dSMA V ⫾ Hyperkeratosis

Allelic to HMSN IB J ⫽ pupillary involvement ⫾ deafness P ⫽ proximal involvement, elevated creatine kinase

⫾ Vocal cord involvement Pyramidal features ⫾ CNS, hearing loss, hand (thenar) atrophy Infantile onset, ⫾ mental retardation ⫾ Spasticity

CNS ⫽ central nervous system; dSMA V ⫽ distal spinal muscular atrophy type V; tRNA ⫽ transfer RNA.

forms such as Déjérine-Sottas or hypomyelinating syndromes, in which the phenotype may be severe. In HMSN II one large prospective study (43 patients from 27 families) performed over 5 years suggested no change in quality of life despite noted worsening of electrophysiologic and clinical examinations in some.210 Among the axonal recessively inherited forms, multiple neurologic system involvement, including involvement of brain and spinal cord, is common. These patients tend to have more disabilities with earlier, more severe involvement. Foot and leg orthotics may help preserve ambulation, especially when dorsiflexor weaknesses have become severe. Although sensory loss tends to be less problematic in these patients, good care of the feet is emphasized because the symptoms and signs may advance with increasing age. The complications of foot ulcers are preventable if patients emphasize (1) good footwear, wide and long enough, with adequate arch support to distribute body weight in those with high arches and avoidance of irregular shoe seams leading to callus and subsequent ulceration; (2) daily foot inspections; (3) absence of weight bearing on first sign of foot ulceration; and (4) avoidance of unnecessary cosmetic foot surgeries.

Axonal Dysfunction and Pathologic Abnormality Caused by Neuronal (Axonal) or Schwann Cell Dysfunction Early studies utilizing histopathologic and clinical correlates with nerve conduction studies suggested important, clinically relevant interactions between Schwann cell products (myelin) and axons.67,68 Patients with the demyelinating phenotypes HMSN I and Déjérine-Sottas disease were noted to have axonal atrophy. The regression lines from plotting axonal diameters (from area) relative to myelin thickness or numbers of myelin lamellae of sural nerves were found to be different from those of normal subjects. Diseased nerves had less steep regression lines and the x intercept was higher than in nerves of healthy subjects. We interpreted the findings as showing axonal atrophy of large-diameter internodes and the influence of remyelination in intermediate and small fibers.67 Additionally, in a longitudinal 15-year study of a group of 31 HMSN I patients, later identified to have peripheral myelin protein 22 (PMP22) duplication and myelin protein zero (MPZ) missense mutations, conduction velocity was found to follow an independent course (slowly improving during maturation, then plateauing and worsening). Both motor

1720

Diseases of the Peripheral Nervous System

CMAP amplitudes and conduction velocity reductions inversely correlated with degree of eventual neurologic disability. The reduction in motor nerve conduction velocities correlated with more rapidly progressive declines in the amplitudes of CMAPs and neurologic disabilities.68 The data suggested the pivotal role of myelin in axonal health, and the importance of maintenance of axonal integrity in disability. We recently plotted the axonal area against myelin thickness in healthy individuals and patients with specific MPZ and connexin 32 gene mutations having axonal electrophysiology (Fig. 73–2). The data supported the earlier suggestion that axonal atrophy precedes fiber degeneration

in Schwann cell disorders. Also, rodent transgenic models with PMP22 alteration suggest axonal degenerations are important in this Schwann cell disorder.181 Molecular biologic work is providing a rationale for these earlier clinical observations. Subspecialized myelin and axonal proteins important in axon-myelin interaction have been identified and localized to various regions of nerve: (1) adaxonal membranes (K⫹ channels, caspr); (2) paranodal myelin loops (neurofilament [NF]155, NF186); and (3) microvilli or perinodal astrocytes (NF155, NF186, Nr-CAM, tenascin-C, actin, ezrin, radixin, moesin, spectrin ␤IV⌺1, ankyrin, Na⫹ channels) (Fig. 73–3).186 Many

FIGURE 73–2 A scatter plot of the relationship of ln axonal area (␮m2 ⫹ 1) to ln myelin thickness (␮m ⫹ 1) of myelinated fibers of sural nerve from a healthy subject (first panel), from a patient with a myelin protein zero mutation (PZ) Thr124Met (second panel), and from a patient with a connexin 32 mutation (arginine to a stop codon at position 220) (third panel). The sural nerve epoxy sections from which the data for the second and third panels were derived are shown in Figures 73–6 and 73–7, respectively. The measurement of axonal area and myelin thickness was done on transverse semithin epoxy sections using the Imaging System for Nerve Morphometry (ISNM, developed by us). The data are reasonably distributed around linear regression lines. Fourth panel, Regression lines for this relationship of nerve from healthy subjects (n ⫽ 6) and from the nerves of patients with the two gene mutations. Axonal atrophy, especially in the nerve from the patient with the connexin mutation and in large fibers, is especially well seen. This information is strong evidence for axonal atrophy preceding fiber degeneration even in inherited demyelinating disorders as a result of genetic abnormalities of Schwann cells.

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

1721

FIGURE 73–3 Axonal protein associations at the node, paranode, and juxtaparanode sections with theorized protein cis and trans interactions. Established localizations within myelin are peripheral myelin protein 22, myelin protein zero, and connexin 32 proteins. The elaborate interactions suggest the important relationship between axonal and myelin components in the normal maintenance of healthy nerve. (Modified from Arroyo, E. J., and Scherer, S. S.: On the molecular architecture of myelinated fibers. Histochem. Cell Biol. 113:1, 2000.) See Color Plate

of these proteins are implicated in the structural integrity of nerve, but most of their developmental and dynamic regulatory functions remain unknown.10 Knockout mouse models (null mice) for the proteins caspr, cgt, and contactin, located at paranodal myelin, result in altered axonal microarchitecture and localization of paranodal and juxtaparanodal proteins.23,33,64,170 These alterations are functionally deleterious, resulting in altered nerve conduction.32,46 Molecular work using mouse models and the proteins known to cause human neuropathies most commonly, including connexin 32, MPZ, and PMP22, also suggest the importance of myelin-axon interactions in axonal health. How dynamic expression factors are altered between the axon and myelin with mutations of specific proteins is largely unknown. Axotomy experiments of myelinated nerves demonstrate Schwann cells to dedifferentiate, and multiple myelin proteins may undergo modified expression; these observations have been reviewed in detail elsewhere.187

The symbiosis of axonal and myelin components may make it difficult to clinically distinguish primary axonal from demyelinating pathogenesis. In general, however, the clinical features of the nerve conduction studies and sural nerve histopathology allow for localization of the primary pathogenesis. Gene defects altering Schwann cell products lead to demyelinating nerve conductions and pathology as outlined previously. Technical factors may limit the usefulness of interpreting nerve conductions. If the study is performed when the CMAP amplitudes have fallen below 0.5 mV, it is often difficult to determine whether the conduction velocity slowing is from a primary demyelinating or axonal process. Occasionally, when CMAP amplitudes are low, examining more proximal motor nerves (i.e., radial and musculocutaneous) may be helpful in the ascertainment of conduction velocity. The validity of nerve conductions is also improved when performed in labs emphasizing (1) proper stimulation and measuring techniques; (2) temperature regulation; and

1722

Diseases of the Peripheral Nervous System

(3) correction of the data using normative information, including age, height, and weight. From the nerve biopsy data in HMSN II, reported in the prior edition of this text, a number of observations suggest that a primary neuronal (axonal) process can be determined.66 Whether ultimately the primary pathogenesis lies within the perikaryon or the axon will depend on the specific mutated genes. The major observation in HMSN II, however, is the reduction of the large myelinated nerve fibers, especially of distal axons. This is in contrast to the features of de- and remyelination typically seen in HMSN I. There is relative preservation of normal internode lengths in contrast to what is found in demyelinating disorders, in which they tend to be quite variable in length and shortened. The frequency distribution and the diameters of myelinated fibers of sural and saphenous nerves are typically abnormal in that the height of the large fiber peak is smaller than in normal nerves at the proximal level (thigh and midcalf). At the ankle level a similar but more severe change is seen. Here the largest fibers are smaller in numbers than in control nerves. The most striking difference between HMSN I and HMSN II can often be found in teased fiber preparations, in which demyelinating and remyelinating features (C, D, and F fiber types) are common in the former. Infrequent scattered and small onion bulbs may be seen in HMSN II, but they are largely distinguishable from the Schwann cell changes seen in the hypertrophic varieties of HMSN I. In some chronic axonal conditions or those more acutely affected severe axonal processes, the loss of fibers and the associated axonal regeneration may make it difficult to distinguish them from the segmental remyelination related to a primary Schwannian process.

Ascertainment of Axonally Affected Individuals The difficulty in identifying mildly affected individuals in families with axonal neuropathies has hindered molecular understanding in this group of patients. Because of this, the molecular genetic classification of HMSN II is less well characterized than are those of demyelinating (HMSN I) varieties. In the demyelinating inherited varieties nerve conduction studies, showing diffuse and often uniform conduction velocity slowing,135 allow for identification of clinically affected individuals and distinguishing such abnormalities from acquired etiologies.69 In contrast, the components of nerve conduction used to measure axonal injury—CMAP, sensory nerve action potential (SNAP), and motor unit duration and size with needle EMG—are quite variable because age, height, sex, and body mass index may affect values.72 The poor ascertainment of axonally affected individuals in kindred studies has impaired progress in linkage analysis. More careful kindred studies are needed to identify affected persons in this group of disorders.

Attempts have been made to improve the ascertainment of asymptomatic but clinically affected individuals with inherited axonal neuropathies. By utilizing composite nerve conduction scores (e.g., summated five nerve conduction attributes expressed as normal deviates of percentiles [⌺ 5NC normal deviates]) and quantified clinical neurologic examinations such as the Neuropathy Impairment Score (NIS) (see Chapter 40), improved ascertainment of affected from unaffected individuals may be possible.71,74 The use of ⌺ 5NC normal deviates and the NIS was helpful in deciding who was, or was not, affected in a large kindred with HMSN IIC (see later), and resulted in determination of the gene locus.125,126

Clinical and Molecular Mimickers of HMSN II Distinguishing patients with HMSN II from those with inherited distal neuronopathies (also referred to as distal spinal amyotrophies and progressive distal muscular atrophies), those with distal myopathies, and those with varieties of hereditary sensory and autonomic neuropathies (HSANs) may be difficult. Equally challenging and likely a more frequent diagnostic problem is the distinction of acquired from inherited axonal disorders. Inherited as compared to acquired neuropathies have the following features: (1) more selective involvement of classes of nerve fibers; (2) a more chronic course; (3) less frequent occurrence of positive neuropathic sensory symptoms (P-NSSs) of asleep-numbness, prickling (paresthesias), or other P-NSSs of many kinds; (4) skeletal abnormalities (e.g., high arches, hammer toes, shallow hip sockets, arthropathy); and (5) a kinship history of similar disorder. Meeting with and examining “unaffected” family members is most useful in determining whether a patient’s disorder may be inherited.75 To recognize mimickers of HMSN II, we offer the following clinical patterns of abnormality, summarized in Figure 73–4. Lack of clinical sensory loss and no nerve conduction abnormality of sensory fibers may suggest myopathy or distal neuronopathy. Distal neuronopathies have large, poorly recruited motor unit potentials on needle EMG and compensated denervation allowing for relative preservation of the CMAP amplitudes. This is in contrast to forms of HMSN II, in which the CMAP amplitudes are reduced and typically the size of the motor unit potentials is less than in distal spinal muscular atrophies (SMAs). In myopathies, motor units are small with increased rate of recruitment. SNAPs allow for the objective identification of sensory involvement. Psychophysical testing (bedside sensory testing and quantitative sensation testing [QST]) are also important tools in the identification of sensory abnormalities. In general QST is superior for accurate identification of mild deficit or hyperalgesia. Lower extremity edema, prior foot trauma, and other technical issues can interfere with accurate assessment of sensory nerve function by all methods.

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

1723

FIGURE 73–4 An algorithm for the clinical differential diagnosis of HMSN II and mimickers of HMSN II. CMAP ⫽ compound muscle action potential; MUP ⫽ motor unit potentials; QST ⫽ quantitative sensation testing; SNAP ⫽ sensory nerve action potential.

Additionally, pain medications, calluses, and patient cooperation interfere with bedside and QST testing. QST provides for quantitative study of individual groups of fibers, including large A␣ (or ␤) and small sensory (A␦, C) fibers, components not studied in routine nerve conduction studies (see Chapter 42). In general, sensory abnormalities of HMSN II and other forms of HMSN have large A␣ (or ␤) fibers more affected than small myelinated (A␦) and unmyelinated (C) fibers. Primary involvement of A␦ and C fibers should raise a question as to the diagnosis of HMSN and suggests better classification as HSAN. HSAN is also likely the better classification for those patients in whom weakness is proportionately less than the extent of sensory loss. Autonomic

reflex testing may be helpful in diagnosing patients. Specifically, postganglionic sudomotor sweating abnormalities and cardiovagal autonomic involvement should not occur in pure motor neuronopathies (spinal amyotrophies) and myopathies, whereas in HSAN such involvement may be striking (see Chapter 78). In forms of HMSN these autonomic abnormalities are typically mild. Cutaneous sensory nerve biopsy, typically of the sural nerve, may provide for the most sensitive test in the identification of distal sensory nerve involvement. This sensitivity is realized when quantitative morphometric studies are performed on nerve fibers. Foremost, the nerve biopsy may allow for the identification of interstitial abnormalities (e.g., microvessel disease, amyloidosis, and others), and

1724

Diseases of the Peripheral Nervous System

such processes can be primary in the pathogenesis or additive of disability or impairments in inherited diseases. Nerve biopsy may be especially useful in individuals without established family history. The presence of significant mononuclear (lymphocyte) inflammatory infiltrates in endoneurial, epineurial, or perineurial tissues should bring skepticism to the diagnosis of HMSN II, and in such cases an empirical treatment trial may be helpful in understanding what component of impairment is acquired versus inherited. It is important to realize that patients with inherited neuropathies are not resistant to superimposed secondary illness. Rare individuals and their families have seen benefit from treatment when a primary inherited form was suspected.61,78 Despite the best attempts to classify the type of lower motor neuron involvement, it may not always be possible to be accurate. This in part relates to the fact that some molecular defects can produce muscle, nerve, or nerve cell body (neuronopathy) disease (see later). Complex genomic and external environmental factors influence the ultimate expression and type of nervous system tissue predominately affected, and such processes likely account for the observations in HMSN IID. In that disorder mutations of the gene for the enzyme glycyl–transfer RNA (tRNA) synthetase (GARS), a ubiquitously expressed gene, lead either to a motor-only neuronopathy (distal SMA type V [dSMA V]) or to HMSN IID with distinct sensory involvement.7 This nonstructural protein exerts its biologic affect in different neurologic tissues based on shared biologic demands. Additionally, mutations of the gene for RAB7, a small GTPase late endosomal protein with homology to Rasrelated GTPases, leads to phenotypes more consistent with HSAN type I (HSAN I) with mutilating ulcerations of the feet or phenotypes more characteristic of HMSN II.12,213 Similarly, sensory nerve involvement may occur in those designated myopathies or neuronopathies (e.g., myotubular myopathy184; spinal bulbar muscular atrophy [SBMA], also known as Kennedy’s disease8; amyotrophic lateral sclerosis77; and SMA91,161). The sensory features are typically mild and often best realized with electrophysiologic and histopathologic quantification. Nerve involvement may also occur within several congenital myopathies, as with mutations in the laminin family, namely Emery-Dreifuss muscular dystrophy, also known as emerin myopathy (lamin A/C),220 and merosin myopathy (lamin A2).200 The neuromyopathy phenotype may relate to metabolic or structural proteins shared between muscle and nerve. Additionally, one unique patient with congenital Duchenne’s muscular dystrophy (DMD) and dystrophin mutation has been reported with neuropathy.47 The mutation in that patient is thought to result in alteration of the proper expression of the nerve dystrophin isoform Dp116, which localizes in the dystroglycan complex of nerve also near the basal lamina. Although nerve conductions were reported as being normal within that patient, a nerve biopsy

revealed “severe reduction of myelinated fibers” and absence of nerve DP116 on Western blot compared to other DMD controls (see Fig. 73–3 for localization of DP116).

Molecular Genetic Classification The increasing identification of specific genetic defects in HMSN is leading to a greater emphasis on a molecular classification. As noted earlier, when the genetic defect is preferentially expressed in the Schwann cell myelin, the disorders are designated CMT (type 1) versus expression in the neuron-axon (CMT type 2), and this genetic classification has been used independent of forms of inheritance.143 Also noted in the new classification is the absence of the type 3 (Déjérine-Sottas or hypomyelinating) group, because it is not believed to be a separate genetic subtype, but rather a distinct phenotypic expression. Both the previous clinical classifications (HMSN II) and the new classification (CMT2) have their inherent limitations. The limitations of the genetic classification are being realized because many of the newly identified genes are equally expressed or produce pathologic insult at both the axon and myelin. It is likely that the classifications will continue to evolve. We hope that the molecular genetic classification will not overshadow the importance of careful phenotypic characterization of patients. The severity of the clinical manifestations may be quite different in persons with the same gene abnormality. Therefore, it should remain important to clinically evaluate patients to provide proper clinical counsel to them and their families and ultimately to understand the complex factors involved in the varied clinical expression. Determination of these factors may provide for interventional therapies currently not practical by gene therapy. The theoretical potential of such approaches is emphasized by consideration of mouse models in which mutations of the MPZ gene expressed in Schwann cells show retardation of neuropathy when placed in a T-cell–deficient background.189 Utilizing knowledge of the functions or proposed functions of some of the discovered proteins, a biochemical classification could also be used to characterize the various mutations leading to axonal neuropathy. For example, based on our current understanding (see Table 73–1; see also Table 73–2 later), one could consider axonal phenotypes secondary to defects in 1. Myelin structural proteins: myelin protein zero (gene: MPZ) 2. Myelin transport: connexin 32 (gene: gap junction B1 [GJB1]) 3. Axonal transport: neurofilament light (gene: NEFL); kinesin family, member 1B (gene: KIF1B); gigaxonin (gene: GAN) 4. Signaling proteins: ganglioside-induced differentiation– associated protein 1 (gene: GDAP1)

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

5. Axonal osmotic homeostasis: KCl cotransporter 3 (gene: solute carrier family 12, member 6 [SLC12A6]) 6. tRNA synthetase: glycyl-tRNA synthetase (gene: GARS) 7. DNA replication: tyrosyl-DNA phosphodiesterase 1 (gene: TDP1) 8. Basal lamin stability: lamin A/C (gene: LMNA) In this chapter we discuss the various gene abnormalities and the associated clinical phenotypes as well as consider other distinct axonal HMSN and CMT phenotypes with chromosomal localization without gene discovery. Because several other earlier HMSN forms were noted for axonal peroneal atrophy, we will also consider HMSN V (spastic paraparesis), HMSN VI (optic atrophy), HMSN VII (retinitis pigmentosa), and juvenile neuroaxonal dystrophy. The avalanche of new genetic information requires a constant review of the literature. Continuously updated information is available online through the compendium to Mendelian Inheritance in Man (www3.ncbi.nlm.nih.gov/Omim).143 Included are references linked to MEDLINE and search engines applicable to phenotype and gene localization. A multicollaborative effort between European and world clinician scientists has established a database of those mutations specific to the inherited neuropathies (molgen-www.uia.ac.be/CMTmutations/). Furthermore, at www.genetests.org/, search engines are available for specific gene tests, and details of the limitations and qualifications of the sites performing those tests. Finally, details of the human genome data sets and reported polymorphisms of each discovered gene can be located at genome.ucsc.edu/cgi-bin/hgGateway?org ⫽ human. Such tools are invaluable to the modern neurologist and neuromuscular specialist.

DOMINANT INHERITANCE: HMSN II Myelin Protein Mutations Leading to HMSN II Myelin Protein Zero (MPZ) Identification of the chromosome region where MPZ resides (1q22-23) provided the first localization of a demyelinating form.25 Subsequent work has noted that MPZ mutations may also be associated with axonal or intermediate forms when determined by nerve conduction studies. We limit our discussion to the axonal varieties; further discussion of MPZ is provided in Chapter 71. In the CMT classification these patients are referred to as CMT type 2I or J143; the J classification designates those with pupillary involvement. Consideration of MPZ sequencing should be made for patients thought to have axonal or intermediate phenotypes and/or afferent light and accommodation pupillary defects.

1725

MPZ, or protein zero (P0), is a glycoprotein made up of 248 amino acids and represents a major structural protein of compact myelin, accounting for over 50% of myelin by weight. It contains one extracellular domain with a disulfide bond, one transmembrane domain, and one intracellular domain. It functions as a true myelin adhesion compacting molecule via homophilic interaction and is expressed in Schwann cells. Opposed tetramers of P0 at the membrane surface hold together the intraperiod line, whereas intracellular carboxyl terminals hold by electrostatic interactions the major dense line.197 MPZ consists of six exons. The intron-exon boundaries are known, allowing for ease of sequence analysis and clinical testing.103,104 The extracellular domain has sequence homology with some immunoglobulins,132,134,148 and, interestingly, antibody directed against P0 has been observed in chronic acquired demyelinating polyneuropathy.222 Missense point mutations identified in the extracellular domain, including the immunoglobulin homolog, are typically responsible for the axonal and intermediate demyelinating forms, whereas the intracellular mutations have been more commonly associated with the severe demyelinating forms (i.e., Déjérine-Sottas disease). Of the greater than 90 micromutations of MPZ, approximately 7 have been linked to axonal disease (Ser44Phe,139 Asp61Gly,193 Asp75Val,148 Glu97Val,190 Tyr119Cys,193 Gln141stop,223 and Gly213Arg190) and 3 with intermediatedemyelinating or axonal forms (Asp35Tyr,141 Asp61Asn,19 and Thr124Met44,100,130,147,159,188,193) (Fig. 73–5). The mutation Ser44Phe is in the extracellular domain within the intraperiod line, and reported individuals have no significant conduction velocity slowing and prominent axonal loss.139 Among other specific micromutations within P0 with primary axonal or intermediate demyelinating forms, a number are also associated with varied cranial nerve or bulbar involvement. The mutations His81Tyr, Val113Phe, Thr124Met, Glu97Val, and Asp75Val have been associated with altered form, size, and function of the pupils, but not in all affected individuals.24,44,147,190,193 One patient was heterozygous for two mutations (His81Tyr, Val113Phe) and had pupillary light-near dissociation with axonalpredominant phenotype on nerve biopsy, whereas nerve conductions showed borderline demyelinating features (median and ulnar conduction velocities of 35 and 37 m/s, respectively); she had approximately a 10-year clinical history beginning in her mid-30s.24 We have followed one family with axonal electrophysiology and sural nerve pathology with the Thr124Met mutation and associated pupillary defects. Affected family members had accommodation and light afferent pupillary response consistent with Adie’s pupil, whereas others of the family, affected by neuropathy, had largely light afferent defects with preservation of accommodation. One had in addition asymmetrical hearing loss at young age and had undergone tympanic surgery for what had been labeled

1726

Diseases of the Peripheral Nervous System

FIGURE 73–5 Selected myelin protein zero (MPZ or P0) amino acid missense mutations and one introduced stop codon leading to axonal phenotypes (black), intermediate demyelinating phenotypes (gray), and those phenotypes in which both intermediate and axonal forms are described (gray boxes). The mutations leading to demyelination (HMSN IB) are not shown, but can be reviewed in Chapter 71. The Thr124Met missense change is also associated with cranial nerve deficits, namely afferent near and light pupil dissociation (Adie’s pupil), hearing loss, and pharyngeal weakness. The axonal phenotype occurring with this mutation is first described in this chapter, whereas Thr124Met, His81Tyr, Val12Phe, Flu97Val, and Asp75Val, previously described mutations with pupillary dysfunction, are typically demyelinating. One double mutation, His81Tyr and Val113Phe, has been associated with pupil defect and axonal phenotype. Axonal mutations to date have not occurred with disruption of the transmembrane domain (e.g., between codons 154 and 180). Sequence and conformational data were obtained from the National Center for Biotechnology Information (NCBI), available at www.ncbi.nlm.nih.gov (entry NP000521). The amino acid numbering includes the 29–amino acid leader peptide, which is cleaved prior to insertion of the protein into myelin, and represents the nomenclature of the original papers.

Ménière’s disease at age 20 years. Others had not had such difficulties, and the affected father needed hearing aids only later in life at age 70 years. Pharyngeal weakness with aspiration was also present in one most severely affected patient, requiring ankle-foot orthotic devices since her late 40s (now age 55 years). Her less severely affected sister, age 38 years, had undergone sural nerve biopsy, which showed reduced numbers of thinly myelinated fibers with increased regeneration (Fig. 73–6). As in our one family member, deafness and difficulties in swallowing have been described in other sporadic individuals with the Thr124Met mutation in addition to the afferent pupillary defects.55,147,193 Although exceptions are noted, most patients with the axonal form have noticed symptoms after age 30 years and may or may not be progressive. Why some MPZ mutations result in significant axonal involvement is of considerable interest but not understood.

The degree of preservation in nerve conduction velocity with reduced amplitudes does suggest the prominent involvement of the axonal neural tissue, but with superimposed myelin involvement suggested most strongly from the limited biopsies reported (including that in Fig. 73–6). MPZ knockout mice have severe demyelination and noted axonal degeneration.90 Transfected cells have been used to study the specific myelin adhesion properties of MPZ because frequently noncompact myelin is observed in patients and in mouse models of MPZ. By disrupting the disulfide bond in the extracellular space,224 truncating the cytoplasmic domain,221 or abolishing the N-linked glycosylation, the adhesion properties are noted as reduced.84 Heterozygous MPZ mice with demyelinating phenotype have reduction in degenerating and regenerating myelin fibers when invading cells of the immune system are reduced or eliminated.36,189 Nevertheless, the specific axonal features among some

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

1727

A

B

FIGURE 73–6 A, Transverse epoxy-embedded section stained with methylene blue of the sural nerve from a 38-year-old woman with afferent pupillary defects and axonal polyneuropathy (HMSN IIJ, or CMT2J) having myelin protein zero missense mutation Thr124Met, and whose axon-to-myelin ratio, plotted in Figure 73–2, suggests axonal atrophy. B, Frequency distribution of myelinated fibers shows a shift toward thinly myelinated fibers with increased regenerating fibers (arrows).

patients are not explained by this work, and the varied expression between individuals with similar mutations suggests the important role of factors external to the MPZ gene in determination of axonal or demyelinating phenotypes. Connexin 32 (GJB1) Like MPZ, connexin 32 is also expressed by myelinating Schwann cells, but differs by being a gap junction protein. The associated neuropathy, an X-linked dominant disorder, also has variable clinical expression. In the CMT classification, this neuropathy is referred to as CMT type X. We mention it among the forms of axonal HMSN II because it

often accounts for sporadic or dominant inheritance with reported axonal phenotypes (see Chapter 76). The limited size of families often precludes identification of X-linked inheritance, and therefore clinicians should consider it in the differential of presumably primary axonal disease. Using strict electrophysiologic criteria, presenting patients are more likely to be classified as having an axonal form with mild demyelinating features. Their electrophysiology is often described as “intermediate,” with nerve conduction velocities in the range of 30 to 38 m/s in ulnar motor responses. However, taken together with the electrophysiology, pathologic description, and animal models,178 connexin 32

1728

Diseases of the Peripheral Nervous System

neuropathy is a Schwann cell disorder that leads to axonal loss, with maldevelopment and loss of myelin. Mutation analysis of connexin 32 should be considered in those families without male-to-male transmission. This approach assisted in the diagnosis in one family we followed (Fig. 73–7). Nerve conduction studies and needle EMG suggested a primary axonal process, but when nerve biopsy was performed, a Schwann cell component was also observed (Fig. 73–7). In the proband (III-2) there were significant strength differences in the ankle dorsiflexor muscles without superimposed nerve entrapment at the knee or evidence of an L5 radiculopathy. She had developed left peroneal atrophy and weakness some 6 years before the right peroneal muscle was clinically affected. Her sister (III-1), also affected, had similar but less dramatic asymmetry, but of thenar musculature. A stop codon was identified in the connexin 32 gene at codon position 220 where arginine normally resides. The significant asymmetries in strength suggest the possibility for the phenomenon of lyonization. Lyonization occurs in females of lineages with X-linked disorders, whereby a disproportionate number of nonmutated X-chromosomes are methylated through the normal process of random X-inactivation. The clinical result is often patchy asymmetrical clinical involvement. This phenomenon has not been well documented in connexin 32–mutated female patients, but would be expected and clinically suggested by our patient; however, further molecular studies are required to provide certainty. In one man of this family, the significant median atrophy led to carpal tunnel release at an outside institution without clinical improvements. The predominance of thenar involvement has been noted previously, and may provide a clue to the clinical diagnosis.129,212 Why thenar involvement may be prominent is not understood. Also of clinical significance is the potential for central nervous system involvement, with white matter changes noted on magnetic resonance imaging (MRI).99,138,164,185,202 The transmembrane protein connexin 32 is a gap junction molecule encoded by the GJB1 gene and expressed at paranodal regions within the Schmidt-Lanterman incisures of peripheral nerve myelin, but also at oligodendrocytes and many other non-neuronal tissues.178 Alternate splicing within the one intron of this gene allows for tissue-specific expression in this two-exon gene. In nerve the protein has two extracellular loops, one intracellular loop, and two intracellular terminal domains. It appears to form reflexive gap junctions. These pathways probably provide diffusion pathways to transport ions, metabolites, and second messenger molecules through intracellular channels between axons and myelin.177 Genetic analysis for connexin 32 mutations has allowed for several important observations. This disorder is not rare and probably accounts for the second most common form of HMSN after HMSN type IA.115 Point mutations are by far the most common mutation, and over 150 different mutations have been identified in greater than

200 unrelated families.156 These data have provided for speculation about genotype-phenotype correlations. Nonsense mutations (frameshift mutations) tend to produce more severe phenotypes with earlier presentation compared to missense mutation.27,98,112 Exceptions are noted, however, including the most dramatic case in two male siblings with complete deletion of the connexin 32 coding sequence.154 These brothers were no more severely affected than others with typical missense mutation. Such observations suggest that factors external to the loci influence clinical expression and confuse whether connexin 32 mutations exert their affect by a primary loss of function or by a dominant-negative affect.1 Extensive ongoing work is attempting to answer such difficult questions.177 As with MPZ, the critical interaction between the axonal and Schwann cell elements likely explains why apparent axonal involvement is common. Histopathologic morphometric and ultrastructural analysis of nerve from 14 individuals with various connexin 32 mutations suggests that prominent paranodal myelin pathology exists, with widened nodes of Ranvier and axonal atrophy and with widening of Schmidt-Lanterman incisures.97

Neuronal (Axonal) Mutations Leading to HMSN II HMSN Type IIA (KIF1B) In contrast to MPZ and GJB1, the KIF1B gene expresses a protein in the axon. The gene consists of 47 exons and is around 200 kDa in size. KIF1B has binding domains to microtubules and ATP and functions in axonal vesicular transport. Mutations created in the gene of mice yield abnormal axonal transport.225 The first linkage reports of HMSN type IIA (HMSN IIA) came in 1993 from Othmane et al., who noted three of six families with axonal phenotype and linkage to 1p with autosomal dominant inheritance.162 Refinement of the chromosomal region was performed by Muglia et al., providing for a region at 1p35.152 Saito et al. also identified two Japanese families mapping to this region (1p35-36) and provided more detailed clinical information.180 Described were two families with mean ages of onset of 8.7 (range, 3 to 15) years in one and 9.7 (range, 7 to 10) years in the other. Nerves were not enlarged, and there was no significant conduction velocity slowing having axonal features on nerve conduction and needle examination. A total of 15 affected members were identified between the families. Difficulties in running and climbing stairs were the reported onset problems in both families. No weakness or atrophy was seen in upper extremity or girdle muscles of the lower extremity regardless of age, and only two members of one family had “mild sensory disturbance.” Specific comment as to the progressive nature of these disorders was not made. Additionally, one large Italian family mapped to the region, and the clinical picture was uniform and characterized by distal muscular weakness and atrophy

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

A

B

C

FIGURE 73–7 A, Pedigree of one family with an X-linked dominant form of Charcot-Marie-Tooth disease (CMTX1), and identified to have a causative mutation in connexin 32 (GJB1), changing an arginine to a stop codon at codon position 220. Note the absence of male-to-male transmission, a hallmark of X-linked disorders. Patient III-2 had noted asymmetries in peroneal and thenar weaknesses, without evidence for a superimposed peroneal neuropathy or L5 radiculopathy on EMG studies. The possibility of a lyonization was considered in this woman given the X-linked nature of this disorder. Significant thenar atrophy was present in several members of the family and in individuals III-1 and II-3. B, Transverse epoxy-embedded section stained with methylene blue of the sural nerve from patient III-2, age 41 years. Seen are thinly myelinated fibers with increased regenerating clusters (arrowhead), and infrequent rudimentary onion bulbs (arrows). C, Frequency distributions of myelinated fibers show the reduction in large myelinated fibers, and the axon-to-myelin thickness ratio, plotted in Figure 73–2, suggests axonal atrophy.

1729

1730

Diseases of the Peripheral Nervous System

in the lower limbs, reduced or absent tendon reflexes mainly in the lower limbs, and again only mild sensory impairment in the feet.152 Utilizing one of the families described by Saito et al.,180 Zhao et al.225 identified a heterozygous missense mutation in KIF1B transforming glutamine to leucine at amino acid position 98. This segment correlates with the predicted ATP binding site of the axonal transport protein. The amino acid alteration was found in four affected and not in eight unaffected members of the kinship. Additionally, 95 healthy Japanese controls did not carry the base change. The mutation was then created in mice, which developed neuropathy and demonstrated disruption of the protein’s known mechanochemical function in axonal transport. Specifically, there was reduction in the levels of cargo proteins synaptotagmin and SV2 in Kif1b⫹/⫺ mice. Although this mitochondrial kinesin superfamily motor protein isoform was previously known to be important in transporting synaptic vesicle precursors, the human data combined with the mouse work provided a specific example of the role of defects in this gene and its relevance to disease. The authors proposed that the distal-first theory of axonal neuropathies may occur because distal-longer axons are found with higher levels of motor molecules such as KIF1B.225 Additional identification of families with mutations in this gene will be important in the understanding of genotypephenotype correlation, and those previous families showing linkage to the area will be important to analyze for mutation within KIF1B. It is possible that an alternative responsible gene may be found in this region at 1p35. Within this superfamily of proteins, some 40 kinesin motor molecules are known, and it is likely that mutations in others will ultimately be found to cause axonal neuropathies. HMSN Type IIB (RAB7) The gene RAB7 produces a small GTPase late endosomal protein with homology to Ras-related GTPases, located at chromosome position 3q13-22. In humans it is expressed in multiple tissues, both neural and non-neural. Over 60 human RAB genes have been identified, with many different highly specialized functions predicted. RAB7’s function appears important in axonal vesicular transport. The protein localizes to endosomes, and by work on phagosomes seems critical in retrograde tubular extensions. Active RAB7 on phagosomal membranes associates with the effector protein RAB7interacting lysosomal protein (RILP), which in turn bridges phagosomes with dynein-dynactin, a microtubule-associated motor complex.35,122 Activation of RAB7 is important to allow recruitment of RILP, and consequent association of phagosomes with microtubule-associated motors.102,214 It is predicted that mutations of RAB7 ultimately affect axonal transport. Disruption of other specific microtubule motor complex proteins in nerve disease is known. For example, mutations of dynactin are believed to be causative of a motor neuron disease described in one family with associated vocal

cord paralysis and facial- and hand-predominant weakness.173 Patients are described without sensory symptoms, but detailed sensory testing is not reported. The described mutation is predicted to distort the folding of dynactin’s microtubule-binding domain and thereby interfere with retrograde axonal transport. The initial descriptions of the human work came from a large American family.131 Kwon et al. noted the difficulties in knowing who was affected or unaffected among one large kindred, and conventional nerve conduction studies and needle EMG examination were helpful in establishing affected status. The features were of a chronic axonal neuropathy. Patients’ symptoms began in early adulthood for most, and many had high-arched feet, hammer toes, absent or diminished ankle reflexes, poorly healing foot sores, and distal sensory impairment of the feet and legs. No specific comment was made as to weakness (rather, clumsiness was noted), but CMAPs as well as SNAPs were reported as reduced. Conduction velocities were normal or slightly reduced. Needle EMG examinations revealed denervation in distal muscles. Linkage analysis provided localization to chromosome position 3q13-22. Two other families, one of Scottish and the other of Austrian descent, were noted for similar chromosomal localization, and significant distal dorsiflexor weakness was noted at the ankles.13,56 Among these kindreds mild demyelinating features were seen typically in older individuals, but specific details as to the loss of CMAP amplitude were not discussed to know whether these changes were primary. Large and small fiber sensory modalities were equally affected. None of the families had shocklike pains as has been reported in the disorder characterized as HSAN I, with mutations of serine palmityl transferase long chain subunit 1 (SPTLC1) (see Chapter 78). Verhoeven et al. identified two causative mutations (Leu129Phe and Val162Met) in the RAB7 gene of the previously reported families.13,56,131 The amino acid changes occur at conserved amino acids across different species, and were not found in 200 control chromosomes. Haplotype analysis and race suggested that the three families were not related. Given the clinical similarities with HSAN I, the authors also sequenced the DNA of 24 unrelated HSAN patients for the two identified mutations. Three individuals within this group had the described mutations, two with Val162Met and one with Leu129Phe. It was not noted whether those individuals did or did not have lancinating pains, which were characteristically absent among the other families described. Genetic testing for RAB7 mutations is likely to become an important step in the evaluation of individuals with axonal neuropathies with or without foot injuries and or ulcerations. More molecular work will need to be done to understand how mutations in the gene RAB7 lead to neuropathy. Because mutations of this gene are associated with infectious ulcerations of the feet and neuropathy, it is an intriguing possibility that, in addition to disruption of axonal transport,

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

it involves a potential defect in the innate immune system through disruption of phagocytosis.102,214 Perhaps foot insensibility alone may not be adequate to explain acromutilations with infectious ulceration. HMSN Type IID and dSMA V (GARS) GARS is a 17-exon gene that localizes to 7p14. GARS is expressed ubiquitously in virtually all tissues, both neural and non-neural, in the human body. A member of the family of proteins known as aminoacyl-tRNA synthetases, it is responsible for assisting in the attachment of tRNAs with their appropriate glycyl amino acid. As a group these enzymes are essential in the fidelity of all protein formation carried out during translation. Implication of the potential role of tRNA transferases in human pathogenesis of neuromuscular disease comes from earlier work in myositis. Specifically, a number of aminoacyl-tRNA synthetases have been identified as autoantigens in inflammatory myositis.208 The antibodies directed against histidyl-tRNA synthetase are better known as “anti-Jo-1.” The GARS protein may be implicated in myositis as well. Previously, Ge et al. isolated the human GARS complimentary DNA by screening cell libraries against autoantibodies obtained from a patient with dermatomyositis.94 The work provided emphasis of the potential role of GARS as an autoantigen. The clinical accounts of the nerve disorders ultimately found to be caused by mutations of GARS can be separated into those with or without noted sensory involvement. Patients without sensory loss were characterized as having dSMA V, whereas those with sensory loss were labeled as having HMSN type IID (HMSN IID) or CMT disease type 2D (CMT2D). The first large kindred reported came from Ionasescu et al., and localization was made to chromosome 7P.113 Within this kindred affected individuals presented at ages 16 to 30 years, typically with onset in the hands, including atrophy. Most had high arches and hammer toes. Examination revealed severe weaknesses in the hands and moderate weaknesses of the lower extremities. Reflexes were absent in the upper extremities and reduced in the lower extremities. Touch, proprioception, and vibration were reduced, with a positive Romberg’s sign in the worst affected patients. Of 14 identified as affected, 12 were noted for slow progression. There were no bulbar, vocal cord, or diaphragmatic signs or symptoms. Nerve conduction velocities were normal, without abnormality in distal latencies. Large, poorly recruited motor unit potentials were noted with fibrillations. There were no hypertrophic nerves palpable, but no nerve histopathology was reported. Pericak-Vance et al. later described one family with a logarithm of odds (LOD) score greater than 3.0 to this 7p region as well as reporting on the genetic heterogeneity of HMSN II. From their data they reported that the IIA form was potentially more common than IID. They also observed no significant atrophy of the limbs in their IID kindred, and suggested variable expression in this form.167

1731

Prior to the work on HMSN IID, Christodoulou et al.45 described a large Bulgarian kindred that fit the classification proposed by Harding101 for dSMA. These patients, like the one family described by Ionasescu et al.113 with HMSN IID, also had prominent thenar and interosseous atrophy. When linkage analysis excluded localization of this disorder to chromosome 5 and SMA forms I, II, and III, the disorder was designated dSMA type V, and localization was discovered at the same region later found in HMSN IID. The 30 affected individuals of this five-generation family had no sensory symptoms and ambiguity to the sensory involvement limited to only 10%, with vibratory sensation reduction. The electrophysiology showed absent CMAP responses in the upper extremity and preserved SNAPs. It was not until a large Mongolian kindred was later published with haplotype data that it was suggested that dSMA V and HMSN IID disorders were in fact allelic to one another.181 Specifically, dSMA V was diagnosed in 11 members of the family and HMSN IID in 6 members of the same family. All 17 patients had significant thenar and interosseous atrophy with cold-induced muscle cramps, with onset from ages 12 to 36 years. Those labeled as having HMSN IID had stocking-glove sensory loss with reduced SNAPs. Utilizing the previously described four families and one additional family of Algerian Sephardic Jewish background, Antonellis et al.7 identified GARS missense mutations in all five families, with one conserved mutation between the families of Ionasescu et al.113 and Pericak-Vance et al.167 (G240R). The other mutations included L129P, E71G, and G526R missense mutations. Antonellis et al. were careful to exclude the possibility that these changes were simple polymorphisms by sequencing large appropriate race populations. Reviewing the specifics of each mutation, the authors proposed a loss or a decrease in catalytic activity as the result of each mutation. Several explanations were proposed for why a ubiquitously expressed protein defect should only affect nerve. They proposed that perhaps, in a typical cell, these defects may not give rise to a detectable phenotype, whereas in long axons there may be a reduction of proteins reaching distal axonal termini. Alternatively, they proposed that the axons affected in HMSN IID or dSMA may be more rich in glycine proteins. They also made the observation that lysyl-tRNA synthetase interacts with superoxide dismutase 1 gene, the gene implicated in one form of inherited amyotrophic lateral sclerosis. Although not discussed by Antonellis et al.,7 one additional mechanism for GARS pathogenesis might be autoimmunity produced by the introduced mutations. Consideration of this mechanism is given support by the earlier work in myositis showing antibody against GARS as well as other tRNA synthetases. Although unlikely to be the only cause of neural degeneration, consideration of autoimmunity is important as a potential mechanism amenable to current immunotherapies, because correction of the underlying genetic defect is currently not practical.

1732

Diseases of the Peripheral Nervous System

HMSN Type IIE and Type IF (NEFL) Neurofilament light (NF-L) protein lies within the cytoskeleton of myelinated axons and belongs to the group of intermediate neurofilaments. Among the neurofilament proteins are light (NF-L), mid-sized (NF-M), and heavy (NF-H) forms. Several mutations of the NF-H protein have been associated with amyotrophic lateral sclerosis.83,211 Additionally, antibodies to NF-L have been found in the cerebrospinal fluid of multiple sclerosis patients who are more severely affected.79 The NEFL gene localizes to chromosome position 8p21 and is encoded by a conserved five exons, with polymorphisms noted. The 543–amino acid NEFL protein consists of a head, rod, and tail domain. Knockout mice lacking NEFL demonstrate reduced axonal caliber and decreased regeneration of myelinated fibers after crush injury to peripheral nerves.227 Additionally, transgenic mice with a NEFL missense Leu394Pro mutation have massive degeneration of spinal motor neurons with abnormal neurofilament accumulation and severe neurogenic atrophy of skeletal muscles.123 The earliest human reports of human NEFL mutation emphasized the axonal phenotypes designated HMSN type IIE (HMSN IIE). In a large Russian family, Mersiyanova et al.145 described a Gln333Pro missense mutation. Affected family members had the typical HMSN II phenotype with dominant inheritance and nerve conduction velocities reported to be consistent with the previous criteria of 38 to 52 m/s, although other details of the conductions were not reported. The mutation was found in all affected individuals and not in 180 normal chromosomes. Several patients had hyperkeratosis, but the association with disease was not clear. Subsequently both demyelinating and axonal phenotypes have been described with mutations in the NEFL gene. The demyelinating phenotypes have been referred to as CMT disease type 1F or HMSN type IF (HMSN IF). Screening 323 unrelated idiopathic HMSN patients with both axonal and demyelinating phenotypes, Jordanova et al.121 reported six disease-causing missense mutations (Pro8Arg, Pro8Gln, Pro8Leu, Glu89Lys, Asn97Ser, and Glu528del). Mutations of the head region of the protein at codon 8 were most common, and four separate families had the Pro8Arg mutation. Nerve conduction studies were available among each of the mutations represented, but in only one patient among the most common Pro8Arg mutation. This patient’s symptoms were noticed at age 7 years, and the ulnar motor conduction velocity was 47 m/s with a CMAP amplitude of 0.9 mV (i.e., axonal type). This specific amino acid change had been reported previously by De Jonghe et al.54 in a multigeneration Belgian family, and variable conduction velocity slowing was noted. This family had acral hypoesthesia and severe below-ankle weakness in members more than 40 years old, and onset typically in the 20s. The conduction studies were performed on older and hence more severely affected individuals. SNAPs were typically absent. The ranges of motor

nerve conduction velocities were 25 to 39 m/s (n ⫽ 3) for the median and 30 to 42 m/s (n ⫽ 2) for the ulnar nerve. Amplitudes of the CMAP were reduced and ranged from 1.4 to 3.8 mV (normal, ⬎6 mV). Among Jordanova et al.’s121 remaining patients, all presented with symptoms before age 13 involving difficulties in ambulation. The nerve conduction velocities among them varied in the ulnar forearm from 19 to 33 m/s, with variable axonal loss. Also reported was the sural nerve biopsy of a patient with Glu89Lys mutation and demyelinating nerve conductions, with what was concluded to be axonal pathology, showing axonal degeneration and regeneration as well as small, infrequent onion bulbs. The authors concluded that NEFL gene testing should be considered for early-onset, often demyelinating conditions. Given these observations, mutations of the NEFL gene have been classified as either HMSN type IF (demyelinating) or HMSN IIE (axonal).

HMSN II Types with Chromosomal Localization without Gene Discovery HMSN Type IIC (12q23-24) HMSN type IIC (HMSN IIC), also called CMT disease type 2C, is a rare autosomal dominant axonal form of peroneal muscular atrophy with progressive muscle weakness and atrophy of limb, diaphragm, vocal cord, and intercostal muscles with variable degrees of acral sensory loss.73 Because of the sensory loss seen in patients, it has been classified as a form of HMSN. As with other forms of HMSN, sensory involvement may be mild and unassociated with positive neuropathic sensory symptoms (i.e., pain, asleepnumbness, and paresthesias). Among affected persons, the age of onset and clinical severity are extremely varied. Individuals presenting in childhood tend to have severe disease and a shortened lifespan from respiratory failure and complications. Initial evidence suggested that HMSN IIC was not genetically linked to HMSN IIA (1P36.2), IIB (3q13-22), or IID (7p14).153 A recent European kindred designated IIC has also not shown linkage to varieties of HMSN II or distal hereditary motor neuronopathy type VII (dHMN VII; 2q14).153,183 Because of the known variability in clinical severity of HMSN IIC (Fig. 73–8) several approaches, including composite scoring of nerve conduction abnormality, were used to identify individuals as affected or unaffected for linkage analysis. Composite scores of nerve conduction abnormalities include multiple attributes of nerve conduction (e.g., CMAP, distal latency, and conduction velocity), and these obtained values are then compared to values from normals of similar age and anthropomorphic features, including age, height, and weight.74 The data may be represented as a summated percentile of normal values.71 These data were combined with quantified clinical neurologic examination, scoring abnormalities of strength, sensation,

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

FIGURE 73–8 The degree of clinical severity was variable among persons in the studied kindred with HMSN IIC. Shown is the proband, IV-58, whose Neuropathy Impairment Score (NIS) was 164 points, whereas her mother’s (III-67) NIS was 50 points. One family member not shown had a NIS score of only 8, but the composite score of nerve conduction abnormalities was greater than the 97.5th percentile, and the conserved haplotype of affected individuals was also demonstrated in that person.125

and reflex (e.g., NIS76). Patients were designated as affected when composite nerve conductions were at or above the 97.5th percentile of normal values and NIS scores were greater than 2 points. Those designated as unaffected were greater than 23 years old, had NIS scores less than 2 points, and had composite nerve conductions less than the 50th percentile of normals. Utilizing this approach, localization of the gene to a 5-cM region at chromosome 12q23-24 was possible, whereas prior attempts with the same family utilizing traditional clinical examinations were unsuccessful in linkage analysis. All patients identified as affected had a conserved haplotype.

1733

Whereas the presence of diaphragmatic, vocal cord, and intercostal paralysis are distinguishing features of HMSN IIC, hoarseness and possibly diaphragm involvement may occur in other HMSN diseases, including some patients with HMSN type IVA (CMT disease type 4A [CMT4A]).50 Several autosomal dominant neuronopathies (motor neuron diseases) and distal myopathies may also have overlapping clinical features with HMSN IIC, including peroneal muscle atrophy and pharyngeal and/or vocal cord paralysis. The latter disorders are without sensory nerve involvement and include dHMN VII (locus at 2q14)142 and one distal myopathy (vocal cord and pharyngeal dysfunction with distal myopathy; locus at 5q31).81 Two additional forms of inherited motor neuron disease are worth mentioning because of the diaphragm involvement in one and the vocal cord paresis in the other. Specifically, SMA with respiratory distress type 1 is an autosomal recessive severe infantile form with diaphragmatic paresis leading to death. It is caused by various micromutations in the gene encoding immunoglobulin mu-binding protein 2 (IGHMBP2).95,96 How these mutations produce motor neuron degeneration is not known. Second, Puls et al. found in a North American family vocal cord paresis and mutation in the gene dynactin.173 As discussed earlier, these mutations likely interfere with axonal transport. These disorders with dynactin and IGHMBP2 mutations should not be classified as HMSN, because of the absence of reported sensory nerve involvement. An emerging literature also notes pharyngeal and vocal cord involvement in some spinocerebellar syndromes.9,16,117,166,199 Because spinocerebellar features are prominent in those disorders, they are unlikely to be mistaken for HMSN. Triplet repeat expansions in various genes account for the phenotype of spinocerebellar ataxia.51 The region linked to HMSN IIC is distinct from those identified in other HMSN syndromes34,136; however, several other lower motor–predominant neuromuscular disorders have been linked to this region on chromosome 12 (Fig. 73–9). The chromosomal regions for spinocerebellar ataxia type 2, scapuloperoneal SMA (SPSMA), and congenital distal SMA directly overlap the 5.0-cM region identified here for HMSN IIC. The latter two diseases are characterized as spinal amyotrophies. Clinically they are distinguished by several features, most notably the absence of sensory involvement. Patients with SPSMA also have laryngeal palsy, but with prominent scapular involvement.59 Congenital distal SMA patients are without vocal cord involvement and have congenital nonprogressive deficit.89 Whether the genes responsible for these spinal amyotrophies are allelic to HMSN IIC remains unknown. Separation of the disorder in our kindred from the disorders of reported families with progressive spinal amyotrophies is based largely on our finding of

1734

Diseases of the Peripheral Nervous System

HMSN IIF (7q11-21) A six-generation family from the Voronezh province of Russia was studied that had the typical nerve conductions and phenotype of axonal HMSN with autosomal dominant inheritance. Ages of clinical onset ranged from 15 to 25 years, and 14 individuals were diagnosed with the disorder; they had progressive problems with increasing severity of steppage gait, and subsequent clawing hand deformities years later. Linkage analysis identified a region at 7q11-q21 as responsible, with haplotype analysis narrowing the region to approximately a 14-cM span.116

FIGURE 73–9 Neuromuscular diseases localizing to chromosome region 12q21-24 (from linkage data), including HMSN IIC. Shown are the corresponding polymorphic markers, and genetic position from pter (cM). Diseases localizing to this region include scapuloperoneal muscular dystrophy (SPMD), scapuloperoneal spinal muscular atrophy (SPSMA), distal hereditary motor neuropathy (HMN II), congenital distal spinal muscular atrophy, and spinocerebellar ataxia (SCA2). Precise chromosomal localization of D12S101 and D12S1366 is not available. Ataxin-2, the gene responsible for SCA2, localizes within the 5.0-cM region between D12S105 and D12S1330, the region implicated in HMSN IIC. The allelic nature of these disorders will be best determined once the specific gene abnormality is identified.125

sensory involvement. Therefore, we remain open to the possibility that a common genetic abnormality explains these varied phenotypes, as has been realized with those patients with dSMA V and CMT2D, namely, mutations of GARS. Further studies are needed to demonstrate the specific gene alteration and its relationship with the other neuromuscular genes in this region.

HMSN with Proximal Involvement (3q13.1) HMSN with proximal involvement is a rare autosomal dominant phenotype reported in Okinawa, Japan. Patients have a phenotype very similar to that of SBMA. Takashima et al.206 reported on the clinical, pathologic, and genetic findings of 23 patients in eight families with this form and designated it a form of HSMN. Patients had proximal dominant neurogenic atrophy, distal sensory involvement, painful muscle cramps, fasciculations, and areflexia. Also noted were elevated creatine kinase, hyperlipidemia, and diabetes among some. The electrophysiologic and pathologic studies revealed an axonal neuropathy with involvement of the anterior horn cells and dorsal ganglion cells characteristic of a neuronopathy.206 Linkage analysis confirmed a distinct genetic disorder independent of the androgen receptor gene on the X-chromosome that is known to be responsible for SBMA. Instead, chromosome region 3q13.1 was identified as linked to these families, with a multipoint LOD score of 4.93.206,207 X-Linked CMT Disease Type 2 (Xp22.2) One large family with CMT disease was noted by Ionasescu et al.114,115 for X-linked recessive inheritance, infantile onset, atrophy with weakness of the lower leg muscles, areflexia, pes cavus, and mental retardation in two of five children.114,115 The nerve conduction details were not reported but were noted as both axonal and demyelinating. Multipoint linkage analysis showed a maximum LOD score of 3.48 at Xp22.2. The region is in close approximation to the locus identified by Fischbeck et al.86,87 in the family originally described by Cowchock et al.48 with infantile onset, deafness, and mental retardation. Deafness was not present in the families of Ionasescu et al., but a specific genetic cause has not been ascertained to better assess whether these two disorders are in fact allelic. X-Linked CMT Disease Type 3 (Xq26) Ionasescu et al.114,115 also described two families with a HMSN phenotype and X-linked recessive inheritance with chromosomal linkage to the region Xq26 with combined 2-point LOD scores of 2.32 and a multipoint LOD score of 1.81. The disease onset ranged from 10 to 14 years with distal atrophy and weakness with normal intelligence; no ocular

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

features were noted. Two boys of one family had spasticity, without such feature in the other family. Long-chain fatty acids were reported as normal. Female carriers had normal nerve conductions, and affected boys were reported with both axonal (denervation atrophy) and demyelinating (slowed conduction velocities) features, but unfortunately the details of the CMAP and motor conduction velocities were not reported. Pes cavus was absent in the family with spasticity but present in the other family. The data provided evidence of a unique chromosomal localization, different from that of X-linked HMSN type 1 (connexin 32 locus), and despite the spasticity seen in the one family, the gene for Pelizaeus-Merzbacher disease is predicted to be greater than 50 cM away at Xq21.

AUTOSOMAL RECESSIVE INHERITANCE: AXONAL CMT Because family histories are often inadequate in the ascertainment of autosomal dominant inheritance, it is important to have understanding of the additional phenotypes, loci, and genes involved in axonal-predominant forms with varied inheritance, including autosomal recessive disorders. The factors making knowledge of the pattern of inheritance difficult are (1) de novo mutations, (2) small family size, (3) social discord within families precluding contacting relatives, (4) misdiagnosis among additional family members, and (5) varied clinical expression. Several generalities can be made among the autosomal recessive forms: (1) they are often not identified as inherited and are rare; (2) they often occur in families with consanguineous marriage; (3) they often are clinically severe, presenting in childhood or infancy; and (4) other nervous system tissues are affected (e.g., brain and spinal cord), and occasionally other tissues as well. Because many of these disorders occur as syndromic

1735

illnesses, it is often easy to distinguish them from classic forms of HMSN, but because of variable expression, the features of spinal cord, brain, or other tissue involvement may be mild. Table 73–2 presents a summary of the various forms. We review briefly the various clinical forms; more detailed molecular descriptions can be found in Chapter 75 for some.

HMSN IVA/CMT4A (GDAP1) The GDAP1 gene encodes a protein with a glutathione-Stransferase domain, and therefore may be involved in the cells’ ability to handle toxins and/or participate in signal transduction pathways important in neuronal development based on sequence analogy to other signal transduction proteins. The gene is expressed in multiple tissues and appears more abundant in central than peripheral nerves, despite the absence of currently known central nervous system clinical involvement.50 Mutations in GDAP1 have been associated with both axonal and demyelinating phenotypes.155,195 Nerve biopsies among two families were thought to be more consistent with a predominantly axonal process,195 whereas biopsies from those with severe conduction velocity slowing showed onion bulbs and other more typical features of demyelination.155 Vocal cord and diaphragmatic paralysis may occur with this disorder. A number of different populations have been found to carry mutations in this gene believed to be responsible for disease: Tunisian17 and Spanish50 kindreds with hoarseness, and European and North African families with14 or without5,26,155,192 hoarseness. The disease appears to be one of childhood onset with progression and frequent wheelchair dependency by the 20s. In addition to vocal cord involvement, an axonal process of phrenic nerve and diaphragm may be seen.14,50,195 More patients and their families will need to be described before the genotype-phenotype correlations can predict the vocal cord and diaphragmatic

Table 73–2. Miscellaneous Conditions with Axonal Nerve Involvement and Gene Discovery Not Classified HMSN Disease Autosomal Recessive Giant axonal neuropathy Anderman’s disease Friedreich’s ataxia Spinocerebellar atrophy with axonal neuropathy Lamin A/C (emerin) neuropathy Spinal muscular atrophy (types 0, I, II, III) SMARD1 Autosomal Dominant Dynactin neuronopathy

Locus

Gene

Function(s)

Clinical

16q24.1 15q13-15 9q13, p23-11 14q31-32

GAN SLC12A6 FRX TDP1

Neurofilament organization Neuronal ion homeostasis Apoptosis, oxidative stress DNA damage repair enzyme

CNS and hair involvement Agenesis of corpus callosum Ataxia, dysarthria, cardiomyopathy Rapid progression

1q21

LMNA

Nuclear envelope stability

Allelic to Emery-Dreifuss myopathy

5q

SMN

Motor neuron apoptosis

11q13-q21

IGHMBP2

Immunoglobulin binding

Varied severity and distal involvement Infantile diaphragmatic weakness

2p13

dynactin

Axonal transport

Vocal cord paresis

CNS ⫽ central nervous system; SMARD1 ⫽ spinal muscular atrophy with respiratory distress type 1.

1736

Diseases of the Peripheral Nervous System

involvement seen. Factors other than the specific gene mutations may result in this particular phenotype, which is generally more severe. No carriers are reported to be affected even mildly later in life.

Giant Axonal Neuropathy (GAN) Mutations in the gigaxonin gene (GAN) lead to an autosomal recessive axonal neuropathy termed giant axonal neuropathy. The term originates from the observation that nerve biopsies from affected individuals have diffuse giant axons with accumulations of neurofilaments, with mitochondria and microtubules often pushed to the side.11 The gene was localized to 16q24.1 after a genomewide search of three unrelated Tunisian families20 and by studying five consanguineous families.88 Although the exact function is unknown, it is a probable cytoskeletal component that directly or indirectly plays an important role in neurofilament architecture. In the disease state there is a generalized disorganization and aggregation of cytoskeletal intermediate filaments in multiple tissue types: (1) neurofilaments (peripheral nerve); (2) vimentin (endothelial cells, Schwann cells); (3) glial

fibrillary acidic protein (astrocytes); and (4) keratin (hair). Neurofilaments are predominately affected. Patients are sometimes characterized as having a form of HMSN with mental retardation and curly kinky hair. The pattern of inheritance is autosomal recessive, and frameshift, nonsense, and missense mutations are linked to childhood onset.30,128 Berg and co-workers22 and Asbury11 described a 6-yearold girl with kinky hair who, since the age of 3 years, had progressive clumsiness first of the lower and then of the upper limbs, muscle weakness and atrophy, areflexia, and sensory loss. Axonal spheroids (Figs. 73–10 and 73–11) densely packed with neurofilaments were found in myelinated and unmyelinated fibers of sural nerve. These neurofilaments stream together in separate channels from microtubules and other organelles. These and later authors have commented on the similarity of the pathologic abnormality to those of various intoxications (i.e., acrylamide, ␤,␤⬘-iminodipropionitrile, methyl-n-butyl ketone, and n-hexane). Onset generally is reported as occurring in the third year of life or later but was progressive thereafter. Patients apparently did not experience discomfort but were clumsy of gait.

A

B

C

D

E

F

G

FIGURE 73–10 Consecutive lengths (A to G) of a single teased fiber from a typical child with giant axonal neuropathy, showing axonal enlargement to form a spheroid (well developed in E and less developed in C). Spheroids without myelin are probably present in D. Despite the prominent axonal involvement, segmental demyelination is encountered in D and remyelinated internodes in B, C, E, and F.

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

1737

FIGURE 73–11 Electron micrograph of transverse sections of myelinated fibers from the sural nerve of a patient with giant axonal neuropathy. In A and C the tightly packed neurofilaments are segregated from the microtubules (without neurofilaments), which are surrounded by the densely packed neurofilaments. The segregation of neurofilaments from microtubules is a striking feature of this disorder.

The patient of Ouvrier and co-workers163 and Boltschauser and associates29 had a syndrome identified as a spinocerebellar degeneration. The additional involvement of the central nervous system was suggested by the presence of electroencephalographic abnormalities and of an upgoing

toe on plantar stimulation in another case.110 By pathologic criteria, spheroids have been described in the central nervous system. In a 25-year-old man, Peiffer and co-workers165 described swollen axons with neurofilament accumulation in brain, cortex, and spinal cord. Others149,172 reported

1738

Diseases of the Peripheral Nervous System

cytoplasmic microfilaments in endoneurial fibroblasts, endothelial cells, Schwann cells, perineurial cells, and cultured skin fibroblasts, suggesting that this is a generalized disorder of cytoplasmic microfilament formation. This was confirmed in another study,93 and by Jedrzejowska and Drac,119 who reported a patient without kinky hair. The disorder is considered to be separate from neuroaxonal dystrophy because, in the latter disorder, onset usually occurs before the age of 1 year, is unassociated with kinky hair, leads quickly to a bedridden state and dementia, and shows differences in histology. Carpenter and associates, studying hair from a typical patient, reported a decrease in disulfides and an increase in thiol groups.38 Two families with giant axons have now been reported without the systemic signs of brain or hair abnormalities, but with axonal swellings and neurofilament accumulations on cutaneous sural nerve biopsy, and were considered as having unique forms of HMSN II.137,216 The families were not exposed to toxic agents such as industrial chemicals known to produce axonal swellings. Several patients of one kindred were also noted to have cardiomyopathy, but the feature was not present in all and not believed to be directly linked.216 Clinically patients also differed from those with “giant axonal neuropathy” associated with GAN mutation by having (1) autosomal dominant inheritance, (2) fewer axonal swellings, and (3) adult onset. In the one family genetic association with known HMSN II genes—that is, GJB1, MPZ, NEFL, and GAN, neurofilament mid-sized gene (NEFM), and PMP22 duplication or deletion—was excluded.137 It is predicted in both families that the eventually identified genes will be important in axonal integrity, including possibly axonal transport.

Anderman’s Syndrome (SLC12A6) HMSN associated with agenesis of the corpus callosum, also known as Anderman’s syndrome or agenesis of the corpus callosum with peripheral neuropathy, is a not uncommon disorder in French Canadian populations of northeastern Quebec,6,53 the overall incidence being 1 in 2117 live births in this region. Nevertheless, patients have been seen in multiple other populations, including populations in Italy, Austria, Tanzania, and Turkey. Children fail to meet developmental milestones and have progressive sensorimotor neuropathy with areflexia. The degree of agenesis of the corpus callosum is variable. Associated features include hypotonia, amyotrophy of extremities, cognitive impairment, and atypical psychosis. Also variably present are high-arched palates and syndactyly. The nerve pathology is characterized by prominent axonal swellings, with early edematous changes of the endoneurium noted, and some early onion bulbs seen as well.37,63 A predominance of large fiber involvement was seen in several sural nerve biopsies.63 The electrophysiologic features are reported as variable, with sensory action potentials of upper and lower

extremities being absent in most, and motor conductions varied in velocity from axonal to demyelinating ranges, but this study made no comment on CMAP amplitudes.63 Combining the pathology with the characteristics of nerve conductions, a primary axonal process may be favored over a primary demyelinating process, but this is less clear than desired.63 The SLC12A6 gene encodes the KCl cotransporter 3 (KCC3) protein, which is an ion cotransporter. The limited mutations to date have led to truncation of the protein.108 The abnormalities are predicted to lead to not only a progressive neurodegenerative condition (neuropathy, atypical psychosis), but also defective developmental embryogenesis (agenesis of corpus callosum). Earlier work suggested that KCC3 may be involved in ion homeostasis and specifically chloride equilibrium, with a role in cell proliferation via ion-sensitive kinases.198 Disruption of this homeostasis may lead to the types of axonal swellings seen. Further work will be needed to better understand this gene’s function in pathogenesis.

Friedreich’s Ataxia and Spinocerebellar Atrophy with Axonal Neuropathy (Frataxin [FRX] and TDP1) In general, system atrophies affecting the spinocerebellar tracks are inherited as both autosomal dominant and recessive disorders. Currently there are more than 18 dominant syndromes described and fewer known recessive forms.143 Of a limited few in which peripheral nerve studies have been described, a primary axonal pathology is thought to be predominant.65 Patients have progressive central ataxia with spinocerebellar degeneration and varied degrees of cognitive, ocular, autonomic, and hyperkinetic dysfunction. The peripheral nervous system is also often affected in these disorders. The central nervous system is typically the major site of clinical pathology, and for this reason less is known about peripheral nerve involvement than desired. Among the known autosomal recessive forms, the peripheral nerve involvement has more commonly been recognized and studied compared to dominant forms. Included among the autosomal recessive spinocerebellar ataxias are Friedreich’s ataxia (the most common spinocerebellar syndrome), spinocerebellar atrophy with axonal neuropathy (SCAN1), vitamin E deficiency (resulting from ␣-tocopherol transfer protein deficiency or abetalipoproteinemia), ataxia-telangiectasia, infantile-onset spinocerebellar ataxia, Marinesco-Sjögren syndrome, spastic ataxia of Charlevoix-Saguenay, Refsum’s disease, carbohydrate-deficiency ataxia, and Cayman Island ataxia, among others.143 We limit our comments to two autosomal recessive forms of spinocerebellar ataxia, Friedreich’s ataxia and SCAN1. Of the forms of spinocerebellar ataxia, Friedreich’s ataxia is arguably the most described, whereas SCAN1 has only recently been identified, and is rare but has a unique genetic

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

cause. An extensive review of the clinical and pathologic work in Friedreich’s ataxia is available in the third edition of this book65 and elsewhere.4 Clinically, patients with Friedreich’s ataxia typically have ataxia, dysarthria, areflexia, Babinski’s sign, and proprioception deficits, with the neuropathy affecting large myelinated axons predominantly. Other manifestations may include cardiomyopathy, scoliosis, and chorea and characteristic high arches and hammer toes. The peripheral nerve pathology in Friedreich’s ataxia shows an axonal atrophy leading to mild demyelination and remyelination, with large fiber sensory fibers most significantly affected (Figs. 73–12 and 73–13).70 Friedreich’s ataxia is the most common hereditary ataxia and is caused by expansion of a GAA triplet-repeat located within the first intron of the frataxin (FRX) gene,4,143 located

1739

at 9q13. There is correlation between size of the expanded repeat and severity of the phenotype. Friedreich’s ataxia is considered a nuclear-encoded mitochondrial disease. FRX is thought to be a nuclear-origin mitochondrial protein that plays a role in iron homeostasis. With deficiency of the gene product frataxin, there is an accumulation of iron in the mitochondria, poor mitochondrial enzyme function, enhanced sensitivity to oxidative stress, and eventually free radical–mediated cell death. Mutation of the gene that encodes the topoisomerase I–dependent DNA damage repair enzyme tyrosyl-DNA phosphodiesterase 1 (TDP1) causes a spinocerebellar syndrome with axonal neuropathy (SCAN). Only one family to date has been reported with a defect in this gene. The family, a large Saudi Arabian kindred, had nine affected

FIGURE 73–12 Numbers and sizes of myelinated and unmyelinated fiber of sural nerve from a 10-year-old boy without disease of nerves (A) and an 11-year-old boy with Friedreich’s ataxia (B). There is a pronounced decrease in number of myelinated fibers, selectively of large fibers. Large fibers are atrophied and therefore fall in smaller size categories as compared to healthy nerves, although their total number is normal. (From Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potentials in health, Friedreich’s ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb, W. A. [ed.]: Handbook of Electroencephalography and Clinical Neurophysiology, Vol. 9. Amsterdam, Elsevier, p. 83, 1971, with permission.)

1740

Diseases of the Peripheral Nervous System

FIGURE 73–13 Fourteen consecutive portions along the length of a teased myelinated fiber (from top to bottom) from the sural nerve of an 11-year-old patient with Friedreich’s ataxia, showing multiple remyelinated internodes (in strips 2 and 3, in strip 5, in strip 7, in strip 12, and in strip 14). The internodes are longer and have heavy myelin that is irregular, and therefore we believe a primary axonal process is responsible for the atrophic process seen. (From Dyck, P. J.: Neuronal atrophy and degeneration affecting primary sensory and autonomic neurons. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. [eds.]: Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1065, 1993, with permission.)

individuals, autosomal recessive inheritance inferred, steppage gate and mild dorsiflexor weakness, pes cavus, and central ataxia with dysarthria.205 Three individuals were examined thoroughly. One had history of seizures and all three had hypercholesterolemia and borderline hypoalbu-

minemia, with cholelithiasis in two. All were reported as having normal intelligence. Sagittal MRI examinations of the head showed moderate cerebellar atrophy, and sural nerve biopsy in one showed loss of large myelinated axonal fibers, but morphometric analysis was not reported. Motor

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

nerve conduction velocities in all three were normal in the upper extremities, and were reduced (i.e., 26.3 m/s vs. ⬎39 m/s reported in normals) in one individual’s lower extremity, but no CMAP amplitudes in either upper or lower extremities or needle EMG examination results were reported. The disorder was designated as an autosomal spinocerebellar condition with axonal neuropathy. A homozygous missense base pair change (A1478G) was found changing histidine to arginine at amino acid position 493. This mutation was not found in 340 control chromosomes, including 160 from unaffected Saudi Arabian families, and was predicted to be 15,000 to 3000 times less active than the wild-type protein. The disorder was considered to be rare when neither 93 individuals of mixed ethnicity with CMT disease nor 187 Japanese individuals with sporadic or autosomal recessive forms of hereditary ataxia were found to have TDP1 mutation. The function of TDP1 is not entirely understood, but much is known and speculated with regard to its importance in DNA repair and RNA maintenance. Of note is the observation that mutations in other DNA repair genes have also been found to be causative of hereditary ataxias, namely ataxia-telangiectasia, xeroderma pigmentosum, and Cockayne’s syndrome.60 TDP1 is implicated in the repair of single- and double-strand breaks of DNA in part created by topoisomerase I.111,171 Because the functions of topoisomerase I are also involved in healthy transcription, defects in TDP1 were suggested205 as a potential mechanism of pathogenesis in postmitotic Purkinje cells, dentate nuclei, anterior horn cells, and dorsal root ganglia, where such repair mechanisms are important in cell survival. The tolerance of these cells to such toxic events as triggers to apoptosis may be less than that of other tissues, hence the absence of significant other disease apart from possibly mild abnormalities (i.e., hypercholesterolemia). To date this family has not been reported to have greater incidence of primary cancers, as occurs in ataxia-telangiectasia and xeroderma pigmentosum.

Autosomal Recessive Axonal Neuropathy (LMNA) Lamin A/C is a nuclear envelope protein localizing to chromosome position 1q21. As a group the lamins are a large constituent of the nuclear lamina within the nuclear membrane. Their function is thought to be related directly to the proper handling of DNA chromatin, with functions of replicational organization in addition to stabilizing the nucleus and its envelope protein.203 There are four different lamin isoforms (A, A⌬, C, and C2), and the R298C substitution is predicted to affect all four known isoforms.58 Mutations in the LMNA gene lead to a variety of clinical phenotypes, frequently with neuromuscular involvement and with different tissues affected: (1) skeletal muscle (Emery-Dreifuss muscular dystrophy,31 dominant or recessive174 and autosomal

1741

dominant limb girdle muscular dystrophy150 [LGMD1B]); (2) cardiac conduction defects and cardiomyopathy (EmeryDreifuss muscular dystrophy,31 LGMD1B,150 and dilated cardiomyopathy type 1A [CMD1A] 80; (3) bone (mandibuloacral dysplasia158); (4) fat (partial lipodystrophy196); and (5) nerve.43,58 To date the mutations of LMNA responsible for each of these distinct phenotypes have occurred at different sites, suggesting separate functional domains in this nuclear envelope protein. Only one apparent founder mutation in Algerian patients has been associated with the axonal neuropathy: missense mutation 892C-to-T, whereby arginine is changed to cystidine at amino acid position 298 within a rod domain. This mutation is predicted to destabilize the nuclear membrane structure and function.58 As with other autosomal recessive disorders, the individuals affected with this form of axonal neuropathy are in part the result of consanguineous marriages. The four families identified with the 892C-to-T mutation are all of Algerian descent. Within the region of Algiers, phenotypes of HMSN have been determined to be inherited as autosomal recessive forms in 60% of individuals among three regional medical centers.43 This likely relates to the high rates of consanguineous marriages (i.e., 23%) in the general population. The phenotypes of the families reported with the LMNA mutations are noteworthy and distinct. The age of presentation of the individuals described ranged from 12 to 24 years. The authors had a mean follow up of 12.1 years. Patients had the more characteristic features of HMSN, with high arches and hammer toes in addition to distal peroneal weaknesses, but in addition prominent proximal weakness with atrophy including the pelvic girdle was noted in five of eight patients examined. The extent of sensory involvement could be correlated with the duration of disease and was typically of a stocking-glove distribution. Scoliosis was common and occurred within 4 years of diagnosis among 50% of patients. The rate of progression is unique among forms of HMSN by being rapid. Specifically, six of eight patients had severe disabilities within 5 years of first noticeable symptoms. Seven of eight patients had involvement of the hands with disability and amyotrophy several months to 2 years after onset. Five patients also had difficulties in standing in less than 4 years after onset of diagnosis. Of note, however, were three patients with only distal involvement at 4 years in two and 7 years in one.43,58 The nerve conductions of these eight patients showed normal conduction velocity of the median motor nerve with relative preservation in the peroneal motor response when one considers the drop in CMAP amplitude. The sural SNAPs were absent in six of eight patients. No needle EMG examinations were reported, nor were muscle biopsies. Nerve biopsy was available on one patient and showed reduction in large myelinated fibers with axonal degeneration, without Schwann cell proliferation (onion bulbs). No cardiac conduction abnormalities or dilated cardiomyopathy was seen, as does occur in Emery-Dreifuss muscular dystrophy.

1742

Diseases of the Peripheral Nervous System

How LMNA mutations lead to such a range of phenotypes is an intriguing question without current answer. The noted myopathies (Emery-Dreifuss muscular dystrophy, LGMD1B, CMD1A) caused by lamin mutations might suggest the potential for the identified neuropathy and muscle atrophy to also be associated with myopathy. To date needle EMG and muscle pathology of patients reported with the R298C mutation have yet to be reported. However, the findings on nerve biopsy, the sensory loss, and the absence of sural SNAPs all suggest the clear involvement of nerve. Also, previous work suggested this protein’s importance in neural development.168 Additionally, homozygous knockout mice have a comparable phenotype to the neuropathy patients described previously, with axonal atrophy, kyphoscoliosis, and locomotor difficulties.204 One might imagine a scenario like that in the myotubularin protein multigene family, whereby both myopathy and neuropathy can occur. Specifically, mutations of MTM1 cause myotubular myopathy,133 whereas mutations in its homolog MTMR2 cause a demyelinating form of CMT designated 4B.28

HEREDITARY AXONAL NEUROPATHIES WITH ASSOCIATED CONDITIONS HMSN V (with Spastic Paraplegia) The HMSN V classification has largely been overshadowed by work in the hereditary spastic paraplegias, specifically those designated “complicated” spastic paraplegia by having both spasticity and other features, including possibly neuropathy. Nevertheless, the HMSN V classification may be appropriate in individuals in whom the peroneal atrophy and lower motor neuron involvement is predominant over spasticity. The details of the historical accounts of this disor-

der(s) can be found in the third edition of this book.66 There is known clinical and genetic heterogeneity among these patients, and many of the originally described kindreds have benefited from the molecular work on hereditary spastic paraplegia.85 Specifically, the families and work described by Strumpell, Silver, and Troyer, and discussed in the previous edition of this book, are now designated SPG3A, SPG17, and SPQ20, respectively, with either chromosomal localization and or gene discovery. For a list of those with axonal neuropathy and gene discovery, see Table 73–3. We note that autosomal dominant hereditary spastic paraplegias are more commonly designated “uncomplicated,” by not having neuropathy or syndromic multiple neurologic system or systemic illness compared to their autosomal recessive and X-linked counterparts, in which these features are more common. It is likely that the degree of lower motor neuron (peripheral neuropathy) compared to upper motor neuron (spasticity) involvement in each of these disorders relates more to factors of varied expression than necessarily distinct gene mutations. Within the same family, HMSN V individuals may have prominent upper or lower motor neuron involvement. Because prior pathologic descriptions and electrophysiologic studies of nerve suggested an axonal peripheral nerve process69 among some, we have included them in this section. Spasticity with neuropathy may, however, also occur in demyelinating spastic paraplegias, including arylsulfatase deficiency (metachromatic leukodystrophy) and with mutation of proteolipid protein, an intrinsic myelin protein. Interestingly, both these disorders have been noted for axonal features as well as peripheral nerve demyelination. Unfortunately, the extent of examination of the peripheral nervous system in many forms of spastic paraplegia is not as complete as desired, likely owing to the predominance of upper motor neuron signs and symptoms. The original descriptions of HMSN V noted seconddecade onset, with autosomal dominant inheritance.66 These

Table 73–3. Complicated* Hereditary Spastic Paraplegias with Axonal Neuropathy and Gene Discovery Disease

Locus

Gene

Function(s)

Clinical

Autosomal Dominant SPG10 12q13

KIF5A

Axonal transport

⫾ Distal atrophy

Autosomal Recessive SPG7

16q

paraplegin

Mitochondrial function

SPG20

13q

spartin

Microtubule formation

⫾ Neuropathy, dysarthria, dysphagia, optic disc pallor Distal wasting: Troyer’s syndrome

X-Linked SPG2

Xq21

PLP

Intrinsic myelin structure ⫾ Varied CNS white matter disease

CNS ⫽ central nervous system. *Complicated refers to those with more than spasticity, and when neuropathy is prominent, they may also be referred to as HMSN type V.

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

patients do not have spinocerebellar or olivopontocerebellar atrophy, and therefore the spasticity and lower motor neuron involvement do not fit into those classes of disease (e.g., spinocerebellar ataxia and olivopontocerebellar atrophy). The course is typically insidiously progressive over many years. Although walking may become difficult owing to spasticity, many persons are never confined to a wheelchair, with those who are only becoming confined in the fourth decade or later. Some families may have limited upper motor neuron features, as in those described by Frith et al.,92 in whom autosomal dominant axonal neuropathy was associated with plantar extensor responses without spasticity. We have not seen shortened life expectancy in this group. Subjective sensory symptoms are typically not reported and sensory examinations are typically relatively preserved, although sensory nerve conductions and morphometric analysis of nerve are often abnormal.66,69 The conduction velocities of nerves are typically normal in the upper extremities and normal or low-normal in the lower extremities, with reduction in CMAP. On needle EMG examination, fibrillations and fasciculations are often seen distally with increased size of motor unit potentials. Among the current 10 autosomal dominant genetically characterized spastic paraplegias, several are noted for variable extremity weakness and atrophy (SPG9-10q23.3-q24.2, SPG17-11q12-q14, and SPG10-12q13). Detailed descriptions of the electrophysiology and nerve biopsy reports are lacking in these and other forms, including SPG3A, in which the specific genetic cause is known. Mutations of the gene designated SPG3A or atlastin151,226 have been found responsible among several families without peroneal atrophy noted. The protein has functional GTPase regions and structural homology similar to the dynamins. The dynamins have been implicated as important in (1) vesicular transport,39,120,157,217 (2) mitochondrial distribution,169 and (3) maintenance of the axonal cytoskeleton.160 It is understandable why mutations of this protein could lead to types of axonal degenerations centrally with resultant or simultaneous neuropathy. SPG17, also known as Silver syndrome,201 was described in the third edition of this text, characterized by spastic paraplegia with prominent upper extremity hand atrophy (e.g., “complicated”). No gene mutation has been identified in this syndrome. SPG9 is associated with cataracts, gastroesophageal reflux, and a motor neuronopathy also without gene identification. SPG10 has the closest similarity to the originally described patients with HMSN V, by having prominent distal atrophies including the peroneal muscles. Mutations of the kinesin heavy chain gene (KIF5A) have been found causative in this form. The mutation occurs at an invariant (between species) asparagine residue that, when mutated in orthologous kinesin heavy chain motor proteins, prevents stimulation of the motor ATPase by microtubule binding.176 Mutations are predicted to disrupt both axonal retrograde and anterograde transport, with resultant distal nerve degeneration in association with the spinal cord pathogenesis.

1743

There are no fewer than seven autosomal recessive forms of spastic paraplegia with identified chromosomal locations, and four are noted with either distal amyotrophy or neuropathy. Specific genes have been identified in SPG7-16q (paraplegin) and SPG20-13q (spartin). Both these forms appear to have variable peripheral nervous system involvement. Patients with paraplegin mutations (SPG7) have either complicated or uncomplicated phenotypes. Of those with the complicated form, presentations are variable, with skeletal muscle mitochondrial abnormalities (ragged red fibers and cytochrome oxidase–negative fibers)40 seen with dysarthria, dysphagia, optic disc pallor, and axonal neuropathy. Also seen on MRI imaging are global or cerebellar atrophy and vascular abnormalities.40,57 Paraplegin has homology to mitochondrial proteins, and localizes to the mitochondria.57 Less is known about spartin, the gene believed to be responsible for Troyer’s syndrome (SPG20), but sequence homology is present to the gene spastin, found to be responsible for SPG4, the most common cause of autosomal dominant spastic paraplegia. When mutated, both genes may disrupt microtubule formation and hence axonal cytoskeletal architecture.175 Troyer’s syndrome patients have prominent upper extremity atrophy.

HMSN VI (with Optic Atrophy) The molecular advances in this disorder since the prior edition of this text have been sparse, and in part this likely relates to the rarity of individuals reported. Many of the earlier reports predated electrophysiologic studies, and therefore the primary injury, either axonal or demyelinating, is not described. Reviewing many of the earlier cases, autosomal dominant disorder might be predicted, but among others X-linked or autosomal recessive forms are possible, and at least the latter has more recently been reported with axonal electrophysiology.42 In 1889, Vizioli215 described a kinship in which a father and two sons were found to have optic atrophy in association with peroneal muscular atrophy.52,109 Peripheral sensory neurons were probably also affected because of the lancinating pains that characteristically affected some patients with involvement of these neurons. Ages at onset were 59 years for the father, 26 years for one son, and 6 years for the other son. Loss of visual acuity, progressing to blindness, occurred in the father and in the 26-year-old son. Muscle weakness and atrophy began in leg muscles and later affected the hand muscles. Tendon reflexes were normal or decreased. Two of the patients had pes cavus. Dawidenkow recorded seven cases, including those of Vizioli, by 1927.52 Included were Taylor’s209 and Head’s105 cases of siblings who were affected by atrophy of the legs, optic atrophy, and loss of hearing. A further report of this association was given by Ferguson and Critchley.82

1744

Diseases of the Peripheral Nervous System

In the report by Milhorat, two brothers with peroneal muscular atrophy and hypertrophic neuropathy had optic atrophy.146 A sister with multiple sclerosis also had optic atrophy. Their grandparents were first cousins. Other reports of ocular manifestations and HMSN have been noted.15,124,144 Substantial advances have been made in our understanding of the most common inherited isolated optic atrophies, but without specific note of a more generalized neuropathy. Specifically, the gene OPA1, which codes for a dynamin-related GTPase, has been found to be responsible for some families with autosomal dominant optic atrophy. This gene is thought to exert its function in mitochondrial biogenesis and stabilization of mitochondrial membrane integrity.3 Abnormalities of mitochondria are therefore likely to result in peripheral neuropathy as well, but examples do not currently exist in which both have been recorded. Consideration of testing asymptomatic patients with optic atrophy is warranted given the observation of Chalmers and colleagues41 that, at least in one family with optic atrophy, asymptomatic neuropathy was documented with hyporeflexia, reduced sensation, and abnormal EMG examination with nerve conductions.

HMSN VII (with Retinitis Pigmentosa) The original reports cited in the prior edition of this book proposed the classification of a motor-predominant peroneal atrophy with retinitis pigmentosa.140 The designation of HMSN VII was chosen to represent this group. If one considers this specific form, there have been no significant molecular advances. However, advances in a similar and possibly allelic group are worth mentioning. Rare patients with retinitis pigmentosa and apparent sensory ganglionopathy with posterior column increased signal are noted and designated as having posterior column ataxia with retinitis pigmentosa.106,107 The disorder has been reported in several families with similar genetic localization and autosomal recessive inheritance. Patients may also be affected by scoliosis, achalasia, and gastrointestinal dysmotility. There is neurophysiologic and histiologic evidence of a sensory ganglionopathy. All sensory potentials were absent on nerve conductions, whereas motor conductions were without slowing or evidence for axonal loss.107 Chromosomal localization is at 1q32-q31, and the region designated AXPC1. The specific gene abnormality has not been identified.

Infantile and Juvenile Neuroaxonal Dystrophy Seitelberger described infants who, after a normal birth and early development, regressed in their psychomotor development, died in the first decade of life, and were found at autopsy to have axonal swellings in gray and white matter of the brain.191 Although the main clinical features

are related to chronic brain deterioration, peripheral nerves may be histologically affected and therefore the disorder is briefly described here. Typically the disorder begins late in the first or second year of life with a regression of mental and motor development. Recently, Seven et al. suggested that facial dysmorphisms may exist before onset of the clinical or neurologic signs.194 The infant or child stops speaking and walking and regresses in development. Usually the child is described as floppy; muscles may become atrophic and tendon reflexes disappear. Other children have been spastic. The mental dysfunction often overshadows the peripheral nerve involvement. The syndrome may therefore resemble metachromatic leukodystrophy or Krabbe’s disease. In the brain the axonal swellings (spheroids) are periodic acid–Schiff positive and argyrophilic. Cowen and Olmstead49 reported spheroids widely distributed in gray and white matter of brain of two young affected siblings. Biopsy of sural nerve was shown to aid in the diagnosis.62 Stacks of membranes, tubular and vesicular profiles, and abnormal mitochondria are typical of the nerve lesions.21 Skin nerves are also reported to be involved.218,219 The pathologic abnormality has been reported in an 8-day-old infant.118 The reported cases now number slightly more than 100.2,143 A relationship between infantile neuroaxonal dystrophy and giant axonal neuropathy has been suggested,18 but differences in natural history and pathologic abnormality make that possibility unlikely. No specific treatment is known, and the specific genetic cause remains unknown.

REFERENCES 1. Abrams, C. K., Oh, S., Ri, Y., and Bargiello, T. A.: Mutations in connexin 32: the molecular and biophysical bases for the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Brain Res. Rev. 32:203, 2000. 2. Aicardi, J., and Castelein, P.: Infantile neuroaxonal dystrophy. Brain 102:727, 1979. 3. Alexander, C., Votruba, M., Pesch, U. E., et al.: OPA1, encoding a dynamin-related GTPase, is mutated in autosomal dominant optic atrophy linked to chromosome 3q28. Nat. Genet. 26:211, 2000. 4. Alper, G., and Narayanan, V.: Friedreich’s ataxia. Pediatr. Neurol. 28:335, 2003. 5. Ammar, N., Nelis, E., and Merlini, G.: Identification of novel GDAP1 mutations causing autosomal recessive CharcotMarie-Tooth disease. Neuromuscul. Disord. 13:720, 2003. 6. Andermann, E., Andermann, F., Bergeron, D., et al.: Familial agenesis of the corpus callosum with sensorimotor neuronopathy. J. Neurol. Sci. 6:400, 1979. 7. Antonellis, A., Ellsworth, R. E., Sambuughin, N., et al.: Glycyl tRNA synthetase mutations in Charcot-Marie-Tooth disease type 2D and distal spinal muscular atrophy type V. Am. J. Hum. Genet. 72:1293, 2003.

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon 8. Antonini, G., Gragnani, F., Romaniello, A., et al.: Sensory involvement in spinal-bulbar muscular atrophy (Kennedy’s disease). Muscle Nerve 23:252, 2000. 9. Aragane, N., Katoh, O., Yamada, H., et al.: Respiratory failure due to vocal cord dyskinesia in olivo-ponto-cerebellar atrophy. Chest 96:1212, 1989. 10. Arroyo, E. J., and Scherer, S. S.: On the molecular architecture of myelinated fibers. Histochem. Cell Biol. 113:1, 2000. 11. Asbury, A. K.: Giant axonal neuropathy—a unique case with segmental neurofilamentous masses. Acta Neuropathol. (Berl.) 20:237, 1972. 12. Auer-Grumbach, M., De Jonghe, P., Verhoeven, K., et al.: Autosomal dominant inherited neuropathies with prominent sensory loss and mutilations: a review. Arch. Neurol. 60:329, 2003. 13. Auer-Grumbach, M., De Jonghe, P., Wagner, K., et al.: Phenotype-genotype correlations in a CMT2B family with refined 3q13-q22 locus. Neurology 55:1552, 2000. 14. Azzedine, H., Ruberg, M., Ente, D., et al.: Variability of disease progression in a family with autosomal recessive CMT associated with a S194X and new R310Q mutation in the GDAP1 gene. Neuromuscul. Disord. 13:341, 2003. 15. Baker, R. S., and Upton, A. R. M.: Variations on a phenotype in Charcot-Marie-Tooth disease. Neuropaediatrie 10:290, 1979. 16. Barbieri, F., Pellecchia, M. T., Esposito, E., et al.: Adult-onset familial laryngeal abductor paralysis, cerebellar ataxia, and pure motor neuropathy. Neurology 56:1412, 2001. 17. Baxter, R. V., Ben Othmane, K., Rochelle, J. M., et al.: Ganglioside-induced differentiation-associated protein-1 is mutant in Charcot-Marie-Tooth disease type 4A/8q21. Nat. Genet. 30:21, 2002. 18. Begeer, J. H., Houthoff, H. J., van Weerden, T. W., et al.: Infantile neuroaxonal dystrophy and giant axonal neuropathy: are they related? Ann. Neurol. 6:540, 1979. 19. Bellone, E., Cassandrini, D., Di Mari, E., et al.: Novel MPZ mutation in a sporadic CMT patient. J. Peripher. Nerv. Syst. 6:41, 2001. 20. Ben Hamida, C., Cavalier, L., Belal, S., et al.: Homozygosity mapping of giant axonal neuropathy gene to chromosome 16q24.1. Neurogenetics 1:129, 1997. 21. Berard-Badier, M., Gambarelli, D., Pinsard, N., et al.: Infantile neuroaxonal dystrophy and giant axonal dystrophy of Seitelberger’s disease: II. Peripheral nerve involvement: electron microscopic study in one case. Acta Neuropathol. (Berl.) 5(Suppl. 5):30, 1971. 22. Berg, B. O., Rosenberg, S. H., and Asbury, A. K.: Giant axonal neuropathy. Pediatrics 49:894, 1972. 23. Bhat, M. A., Rios, J. C., Lu, Y., et al.: Axon-glia interactions and the domain organization of myelinated axons require neurexin IV/caspr/paranodin. Neuron 30:369, 2001. 24. Bienfait, H. M., Baas, F., Gabreels-Festen, A. A., et al.: Two amino-acid substitutions in the myelin protein zero gene of a case of Charcot-Marie-Tooth disease associated with light-near dissociation. Neuromuscul. Disord. 12:281, 2002. 25. Bird, T. D., Ott, J., and Giblett, E. R.: Evidence for linkage of Charcot-Marie-Tooth neuropathy to the Duffy locus on chromosome 1. Am. J. Hum. Genet. 34:388, 1982.

1745

26. Birouk, N., Azzedine, H., Dubourg, O., et al.: Phenotypical features of a Moroccan family with autosomal recessive Charcot-Marie-Tooth disease associated with the S194X mutation in the GDAP1 gene. Arch. Neurol. 60:598, 2003. 27. Birouk, N., LeGuern, E., Maisonobe, T., et al.: X-linked Charcot-Marie-Tooth disease with connexin 32 mutations: clinical and electrophysiologic study. Neurology 50:1074, 1998. 28. Bolino, A., Muglia, M., Conforti, F. L., et al.: Charcot-MarieTooth type 4B is caused by mutations in the gene encoding myotubularin-related protein-2. Nat. Genet. 25:17, 2000. 29. Boltshauser, E., Bischoff, A., and Isler, W.: Giant axonal neuropathy: report of a case with normal hair. J. Neurol. Sci. 31:269, 1977. 30. Bomont, P., Cavalier, L., Blondeau, F., et al.: The gene encoding gigaxonin, a new member of the cytoskeletal BTB/kelch repeat family, is mutated in giant axonal neuropathy. Nat. Genet. 26:370–374, 2000. 31. Bonne, G., Di Barletta, M. R., Varnous, S., et al.: Mutations in the gene encoding lamin A/C cause autosomal dominant Emery-Dreifuss muscular dystrophy. Nat. Genet. 21:285–288, 1999. 32. Bosio, A., Binczek, E., and Stoffel, W.: Functional breakdown of the lipid bilayer of the myelin membrane in central and peripheral nervous system by disrupted galactocerebroside synthesis. Proc. Natl. Acad. Sci. U. S. A. 93:13280–13285, 1996. 33. Boyle, M. E., Berglund, E. O., Murai, K. K., et al.: Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30:385, 2001. 34. Brown, R. A. A.: Genetic and molecular basis for neuromuscular diseases. In Brown, W. F., Bolton, C. F., and Aminoff, M. J. (eds.): Neuromuscular Function and Disease. Philadelphia, W.B. Saunders, p. 619, 2002. 35. Cantalupo, G., Alifano, P., Roberti, V., et al.: Rab-interacting lysosomal protein (RILP): the Rab7 effector required for transport to lysosomes. EMBO J. 20:683, 2001. 36. Carenini, S., Maurer, M., Werner, A., et al.: The role of macrophages in demyelinating peripheral nervous system of mice heterozygously deficient in p0. J. Cell Biol. 152:301, 2001. 37. Carpenter, S.: The pathology of Andermann syndrome. In Lassonde, M., and Jeeves, M. (eds.): Callosal Agenesis: A Natural Split Brain? New York, Plenum, p. 27, 1994. 38. Carpenter, S., Karpati, G., Andermann, F., and Gold, R.: Giant axonal neuropathy: a clinically and morphologically distinct neurological disease. Arch. Neurol. 31:312, 1974. 39. Carroll, R. C., Beattie, E. C., Xia, H., et al.: Dynamindependent endocytosis of inotropic glutamate receptors. Proc. Natl. Acad. Sci. U. S. A. 96:14112, 1999. 40. Casari, G., De Fusco, M., Ciarmatori, S., et al.: Spastic paraplegia and OXPHOS impairment caused by mutations in paraplegin, a nuclear-encoded mitochondrial metalloprotease. Cell 93:973, 1998. 41. Chalmers, R. M., Bird, A. C., and Harding, A. E.: Autosomal dominant optic atrophy with asymptomatic peripheral neuropathy. J. Neurol. Neurosurg. Psychiatry 60:195, 1996. 42. Chalmers, R. M., Riordan-Eva, P., and Wood, N. W.: Autosomal recessive inheritance of hereditary motor and

1746

43.

44.

45.

46.

47.

48.

49. 50.

51.

52.

53.

54.

55.

56.

57.

Diseases of the Peripheral Nervous System

sensory neuropathy with optic atrophy. J. Neurol. Neurosurg. Psychiatry 62:385, 1997. Chaouch, M., Allal, Y., De Sandre-Giovannoli, A., et al.: The phenotypic manifestations of autosomal recessive axonal Charcot-Marie-Tooth due to a mutation in lamin A/C gene. Neuromuscul. Disord. 13:60, 2003. Chapon, F., Latour, P., Diraison, P., et al.: Axonal phenotype of Charcot-Marie-Tooth disease associated with a mutation in the myelin protein zero gene. J. Neurol. Neurosurg. Psychiatry 66:779, 1999. Christodoulou, K., Kyriakides, T., Hristova, A. H., et al.: Mapping of a distal form of spinal muscular atrophy with upper limb predominance to chromosome 7p. Hum. Mol. Genet. 4:1629, 1995. Coetzee, T., Fujita, N., Dupree, J., et al.: Myelination in the absence of galactocerebroside and sulfatide: normal structure with abnormal function and regional instability. Cell 86:209, 1996. Comi, G. P., Ciafaloni, E., de Silva, H. A., et al.: A G ⫹ 1 : A transversion at the 5⬘ splice site of intron 69 of the dystrophin gene causing the absence of peripheral nerve Dp116 and severe clinical involvement in a DMD patient. Hum. Mol. Genet. 4:2171, 1995. [Erratum appears in Hum. Mol. Genet. 5:562, 1996.] Cowchock, F. S., Duckett, S. W., Streletz, L. J., et al.: X-linked motor-sensory neuropathy type-II with deafness and mental retardation: a new disorder. Am. J. Med. Genet. 20:307, 1985. Cowen, D., and Olmstead, E. V.: Infantile neuroaxonal dystrophy. J. Neuropathol. Exp. Neurol. 22:175, 1963. Cuesta, A., Pedrola, L., Sevilla, T., et al.: The gene encoding ganglioside-induced differentiation-associate protein 1 is mutated in axonal Charcot-Marie-Tooth type 4A disease. Nat. Genet. 30:22, 2002. Cummings, C. J., and Zoghbi, H. Y.: Trinucleotide repeats: mechanisms and pathophysiology. Annu. Rev. Genomics Hum. Genet. 1:281, 2000. Dawidenkow, A.: Uber die neurotische Muskelatrophie Charcot-Marie: Klinish-genetische Studien. Z. Neurol. 107:259, 1927. De Braekeleer, M., Dallaire, A., and Mathieu, J.: Genetic epidemiology of sensorimotor polyneuropathy with or without agenesis of the corpus callosum in northeastern Quebec. Hum. Genet. 91:223, 1993. de Jonghe, P., Mersivanova, I., Nelis, E., et al.: Further evidence that neurofilament light chain gene mutations can cause Charcot-Marie-Tooth disease type 2E. Ann. Neurol. 49:245, 2001. de Jonghe, P., Timmerman, V., Ceuterick, C., et al.: The Thr124Met mutation in the peripheral myelin protein zero (MPZ) gene is associated with a clinically distinct CharcotMarie-Tooth phenotype. Brain 122:281, 1999. de Jonghe, P., Timmerman, V., FitzPatrick, D., et al.: Mutilating neuropathic ulcerations in a chromosome 3q13-q22 linked Charcot-Marie-Tooth disease type 2B family. J. Neurol. Neurosurg. Psychiatry 62:570, 1997. De Michele, G., De Fusco, M., Cavalcanti, F., et al.: A new locus for autosomal recessive hereditary spastic paraplegia maps to chromosome 16q24.3. Am. J. Hum. Genet. 63:135, 1998.

58. De Sandre-Giovannoli, A., Chaouch, M., Kozlov, S., et al.: Homozygous defects in LMNA, encoding lamin A/C nuclear-envelope proteins, cause autosomal recessive axonal neuropathy in human (Charcot-Marie-Tooth disorder type 2) and mouse. Am. J. Hum. Genet. 70:726, 2002. [Erratum appears in Am. J. Hum. Genet. 70:1075, 2002.] 59. DeLong, R., and Siddique, T.: A large New England kindred with autosomal dominant neurogenic scapuloperoneal amyotrophy with unique features. Arch. Neurol. 49:905, 1992. 60. Di Donato, S., Gellera, C., and Mariotti, C.: The complex clinical and genetic classification of inherited ataxias. II. Autosomal recessive ataxias. Neurol. Sci. 22:219, 2001. 61. Donaghy, M., Sisodiya, S. M., Kennett, R., et al.: Steroid responsive polyneuropathy in a family with a novel myelin protein zero mutation. J. Neurol. Neurosurg. Psychiatry 69:799, 2000. 62. Duncan, C.: Peripheral nerve biopsy as an aid to diagnosis in infantile neuroaxonal dystrophy. Neurology 20:1024, 1970. 63. Dupre, N., Howard, H. C., Mathieu, J., et al.: Hereditary motor and sensory neuropathy with agenesis of the corpus callosum. Ann. Neurol. 54:9, 2003. 64. Dupree, J. L., Girault, J. A., and Popko, B.: Axo-glial interactions regulate the localization of axonal paranodal proteins. J. Cell Biol. 147:1145, 1999. 65. Dyck, P.: Neuronal atrophy and degeneration affecting primary sensory and autonomic neurons. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1065, 1993. 66. Dyck, P., Chance, P. F., Lebo, R., and Carney, A.: Hereditary motor and sensory neuropathies. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1094, 1993. 67. Dyck, P. J., Ellefson, R. D., Lais, A. C., et al.: Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy: preliminary report of a lipid abnormality in nerve and liver in Dejerine-Sottas disease. Mayo Clin. Proc. 45:286, 1970. 68. Dyck, P. J., Karnes, J. L., and Lambert, E. H.: Longitudinal study of neuropathic deficits and nerve conduction abnormalities in hereditary motor and sensory neuropathy type 1. Neurology 39:1302, 1989. 69. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy: II. Neurologic, genetic, and electrophysiologic findings in various neuronal degenerations. Arch. Neurol. 18:619, 1968. 70. Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potentials in health, Friedreich’s ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb, W. A. (ed.): Handbook of Electroencephalography and Clinical Neurophysiology, Vol. 9. Amsterdam, Elsevier, p. 83, 1971. 71. Dyck, P. J., Litchy, W. J., Daube, J. R., et al.: Individual attributes versus composite scores of nerve conduction abnormality: sensitivity, reproducibility, and concordance with impairment. Muscle Nerve 27:202, 2003. 72. Dyck, P. J., Litchy, W. J., Lehman, K. A., et al.: Variables influencing neuropathic endpoints: the Rochester Diabetic

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85. 86.

87.

88.

89.

Neuropathy Study of Healthy Subjects [see comment]. Neurology 45:1115, 1995. Dyck, P. J., Litchy, W. J., Minnerath, S., et al.: Hereditary motor and sensory neuropathy with diaphragm and vocal cord paresis. Ann. Neurol. 35:608, 1994. Dyck, P. J., O’Brien, P. C., Litchy, W. J., et al.: Use of percentiles and normal deviates to express nerve conduction and other test abnormalities. Muscle Nerve 24:307, 2001. Dyck, P. J., Oviatt, K. F., and Lambert, E. H.: Intensive evaluation of referred unclassified neuropathies yields improved diagnosis. Ann. Neurol. 10:222, 1981. Dyck, P. J., Sherman, W. R., Hallcher, L. M., et al.: Human diabetic endoneurial sorbitol, fructose, and myo-inositol related to sural nerve morphometry. Ann. Neurol. 8:590, 1980. Dyck, P. J., Stevens, J. C., Mulder, D. W., and Espinosa, R. E.: Frequency of nerve fiber degeneration of peripheral motor and sensory neurons in amyotrophic lateral sclerosis: morphometry of deep and superficial peroneal nerves. Neurology 25:781, 1975. Dyck, P. J., Swanson, C. J., Low, P. A., et al.: Prednisoneresponsive hereditary motor and sensory neuropathy. Mayo Clin. Proc. 57:239, 1982. Eikelenboom, M. J., Petzold, A., Lazeron, R. H., et al.: Multiple sclerosis: neurofilament light chain antibodies are correlated to cerebral atrophy. Neurology 60:219, 2003. Fatkin, D., Dodds, G. A., Kisslo, J., et al.: Missense mutations in the rod domain of the lamin A/C gene as causes of dilated cardiomyopathy and conduction-system disease [see comment]. Am. J. Cardiol. 83:13H, 1999. Feit, H., Silbergleit, A., Schneider, L. B., et al.: Vocal cord and pharyngeal weakness with autosomal dominant distal myopathy: clinical description and gene localization to 5q31. Am. J. Hum. Genet. 63:1732, 1998. Ferguson, F. R., and Critchley, M.: Leber’s optic atrophy and its relationship with the heredofamilial ataxias. J. Neurol. Psychopathol. 9:120, 1928. Figlewicz, D. A., Krizus, A., Martinoli, M.G., et al.: Variants of the heavy neurofilament subunit are associated with the development of amyotrophic lateral sclerosis. Hum. Mol. Genet. 3:1757, 1994. Filbin, M. T., and Tennekoon, G. I.: Homophilic adhesion of the myelin P0 protein requires glycosylation of both molecules in the homophilic pair. J. Cell Biol. 122:451, 1993. Fink, J. K.: The hereditary spastic paraplegias: nine genes and counting. Arch. Neurol. 60:1045, 2003. Fischbeck, K. H., ar-Rushdi, N., Pericak-Vance, M., et al.: X-linked neuropathy: gene localization with DNA probes. Ann. Neurol. 20:527, 1986. Fischbeck, K. H., Ritter, A., Shi, Y., et al.: Linkage studies of X-linked neuropathy and spinal muscular atrophy: human gene mapping 9. Cytogenet. Cell Genet. 46:614, 1987. Flanigan, K. M., Crawford, T. O., Griffin, J. W., et al.: Localization of the giant axonal neuropathy gene to chromosome 16q24. Ann. Neurol. 43:143, 1998. Fleury, P., and Hageman, G.: A dominantly inherited lower motor neuron disorder presenting at birth with associated arthrogryposis. J. Neurol. Neurosurg. Psychiatry 48:1037, 1985.

1747

90. Frei, R., Motzing, S., Kinkelin, I., et al.: Loss of distal axons and sensory Merkel cells and features indicative of muscle denervation in hindlimbs of P0-deficient mice. J. Neurosci. 19:6058, 1999. 91. Frequin, S. T., Gabreels, F. J., Gabreels-Festen, A. A., and Joosten, E. M.: Sensory axonopathy in hereditary distal spinal muscular atrophy. Clin. Neurol. Neurosurg. 93:323, 1991. 92. Frith, J. A., McLeod, J. G., Nicholson, G. A., and Yang, F.: Peroneal muscular atrophy with pyramidal tract features (hereditary motor and sensory neuropathy type V): a clinical, neurophysiological, and pathological study of a large kindred. J. Neurol. Neurosurg. Psychiatry 57:1343, 1994. 93. Gambarelli, D., Hassoun, J., Pellissier, J. F., et al.: Giant axonal neuropathy: involvement of peripheral nerve, myenteric plexus and extra-neuronal area. Acta Neuropathol. (Berl.) 39:261, 1977. 94. Ge, Q., Trieu, E. P., and Targoff, I. N.: Primary structure and functional expression of human glycyl-tRNA synthetase, an autoantigen in myositis. J. Biol. Chem. 269:28790, 1994. 95. Grohmann, K., Schuelke, M., Diers, A., et al.: Mutations in the gene encoding immunoglobulin mu-binding protein 2 cause spinal muscular atrophy with respiratory distress type 1. Nat. Genet. 29:75, 2001. 96. Grohmann, K., Wienker, T. F., Saar, K., et al.: Diaphragmatic spinal muscular atrophy with respiratory distress is heterogeneous, and one form is linked to chromosome 11q13-q21. Am. J. Hum. Genet. 65:1459, 1999. 97. Hahn, A. F., Ainsworth, P. J., Bolton, C. F., et al.: Pathological findings in the X-linked form of Charcot-Marie-Tooth disease: a morphometric and ultrastructural analysis. Acta Neuropathol. (Berl.) 101:129, 2001. 98. Hahn, A. F., Bolton, C. F., White, C. M., et al.: Genotype/ phenotype correlations in X-linked dominant Charcot-MarieTooth disease. Ann. N. Y. Acad. Sci. 883:366, 1999. 99. Hanemann, C. O., Bergmann, C., Senderek, J., et al.: Transient, recurrent, white matter lesions in X-linked Charcot-Marie-Tooth disease with novel connexin 32 mutation. Arch. Neurol. 60:605, 2003. 100. Hanemann, C. O., Gabreels-Festen, A. A., and De Jonghe, P.: Axon damage in CMT due to mutation in myelin protein P0. Neuromuscul. Disord. 11:753, 2001. [Erratum appears in Neuromuscul. Disord. 12:432, 2002.] 101. Harding, A. E.: Inherited neuronal atrophy and degeneration predominantly of the lower motor neurons. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al.: (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1051, 1993. 102. Harrison, R. E., Bucci, C., Vieira, O. V., et al.: Phagosomes fuse with late endosomes and/or lysosomes by extension of membrane protrusions along microtubules: role of Rab7 and RILP. Mol. Cell. Biol. 23:6494, 2003. 103. Hayasaka, K., Nanao, K., Tahara, M., et al.: Isolation and sequence determination of cDNA encoding the major structural protein of human peripheral myelin. Biochem. Biophys. Res. Commun. 180:515, 1991. 104. Hayasaka, K., Ohnishi, A., Takada, G., et al.: Mutation of the myelin P0 gene in Charcot-Marie-Tooth neuropathy type 1. Biochem. Biophys. Res. Commun. 194:1317, 1993. 105. Head, H.: Peroneal atrophy. Proc. R. Soc. Med. 6:51, 1913.

1748

Diseases of the Peripheral Nervous System

106. Higgins, J. J., Kluetzman, K., Berciano, J., et al.: Posterior column ataxia and retinitis pigmentosa: a distinct clinical and genetic disorder. Mov. Disord. 15:575, 2000. 107. Higgins, J. J., Morton, D. H., Patronas, N., and Nee, L. E.: An autosomal recessive disorder with posterior column ataxia and retinitis pigmentosa [see comment]. Neurology 49:1717, 1997. 108. Howard, H. C., Mount, D. B., Rochefort, D., et al.: The K-Cl cotransporter KCC3 is mutant in a severe peripheral neuropathy associated with agenesis of the corpus callosum. Nat. Genet. 32:384, 2002. 109. Hoyt, W. F.: Charcot-Marie-Tooth disease with primary optic atrophy. Arch. Ophthalmol. 64:935, 1960. 110. Igisu, H., Ohta, M., Tabira, T., et al.: Giant axonal neuropathy: a clinical entity affecting the central as well as the peripheral nervous system. Neurology 25:717, 1975. 111. Inamdar, K. V., Pouliot, J. J., Zhou, T., et al.: Conversion of phosphoglycolate to phosphate termini on 3⬘ overhangs of DNA double strand breaks by the human tyrosyl-DNA phosphodiesterase hTdp1. J. Biol. Chem. 277:27162, 2002. 112. Ionasescu, V., Ionasescu, R., and Searby, C.: Correlation between connexin 32 gene mutations and clinical phenotype in X-linked dominant Charcot-Marie-Tooth neuropathy. Am. J. Med. Genet. 63:486, 1996. 113. Ionasescu, V., Searby, C., Sheffield, V. C., et al.: Autosomal dominant Charcot-Marie-Tooth axonal neuropathy mapped on chromosome 7p (CMT2D). Hum. Mol. Genet. 5:1373, 1996. 114. Ionasescu, V. V., Trofatter, J., Haines, J. L., et al.: Heterogeneity in X-linked recessive Charcot-Marie-Tooth neuropathy. Am. J. Hum. Genet. 48:1075, 1991. 115. Ionasescu, V. V., Trofatter, J., Haines, J. L., et al.: X-linked recessive Charcot-Marie-Tooth neuropathy: clinical and genetic study. Muscle Nerve 15:368, 1992. 116. Ismailov, S. M., Fedotov, V. P., Dadali, E. L., et al.: A new locus for autosomal dominant Charcot-Marie-Tooth disease type 2 (CMT2F) maps to chromosome 7q11-q21. Eur. J. Hum. Genet. 9:646, 2001. 117. Isozaki, E., Naito, R., Kanda, T., et al.: Different mechanism of vocal cord paralysis between spinocerebellar ataxia (SCA 1 and SCA 3) and multiple system atrophy. J. Neurol. Sci. 197:37, 2002. 118. Janota, I.: Neuroaxonal dystrophy in the neonate: a case report. Acta Neuropathol. (Berl.) 46:151, 1979. 119. Jedrzejowska, H., and Drac, H.: Infantile chronic peripheral neuropathy with giant axons. Acta Neuropathol. (Berl.) 37:213, 1977. 120. Jones, S. M., Howell, K. E., Henley, J. R., et al.: Role of dynamin in the formation of transport vesicles from the trans-Golgi network. Science 279:573, 1998. 121. Jordanova, A., de Jong, P., Boerkoel, C. F., et al.: Mutations in the neurofilament light chain gene (NEFL) cause early onset severe Charcot-Marie-Tooth disease. Brain 126:590, 2003. 122. Jordens, I., Fernandez-Borja, M., Marsman, M., et al.: The Rab7 effector protein RILP controls lysosomal transport by inducing the recruitment of dynein-dynactin motors. Curr. Biol. 11:1680, 2001. 123. Julien, J. P., Grosveld, F., Yazdanbaksh, K., et al.: The structure of a human neurofilament gene (NF-L): a unique

124.

125.

126.

127. 128.

129.

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

exon-intron organization in the intermediate filament gene family. Biochim. Biophys. Acta 909:10, 1987. Khoubesserian, P., Van Regemarter, N., Ohrn-Degueldre, O., et al.: Charcot-Marie-Tooth disease with retinal pigment dystrophy and protanopia. J. Neurol. 221:1, 1979. Klein, C. J., Cunningham, J. M., Atkinson, E. J., et al.: The gene for hereditary motor and sensory neuropathy type 2C maps to 12q24. Ann. Neurol. 52:S47, 2002. Klein, C. J., Cunningham, J. M., Atkinson, E. J., et al.: The gene for HMSN2C maps to 12q23-24: a region of neuromuscular disorders. Neurology 60:1151, 2003. Kleopa, K. A., and Scherer, S. S.: Inherited neuropathies. Neurol. Clin. 20:679, 2002. Kuhlenbaumer, G., Young, P., Oberwittler, C., et al.: Giant axonal neuropathy (GAN): case report and two novel mutations in the gigaxonin gene [see comment]. Neurology 58:1273, 2002. [erratum appears in Neurology 58:1444, 2002.] Kuntzer, T., Dunand, M., Schorderet, D. F., et al.: Phenotypic expression of a Pro 87 to Leu mutation in the connexin 32 gene in a large Swiss family with Charcot-Marie-Tooth neuropathy. J. Neurol. Sci. 207:77, 2003. Kurihara, S., Adachi, Y., Wada, K., et al.: An epidemiological genetic study of Charcot-Marie-Tooth disease in Western Japan. Neuroepidemiology 21:246, 2002. Kwon, J. M., Elliott, J. L., Yee, W. C., et al.: Assignment of a second Charcot-Marie-Tooth type II locus to chromosome 3q. Am. J. Hum. Genet. 57:853, 1995. Lai, C., Brow, M. A., Nave, K. A., et al.: Two forms of 1B236/myelin-associated glycoprotein, a cell adhesion molecule for postnatal neural development, are produced by alternative splicing. Proc. Natl. Acad. Sci. U. S. A. 84:4337, 1987. Laporte, J., Hu, L. J., Kretz, C., et al.: A gene mutated in X-linked myotubular myopathy defines a new putative tyrosine phosphatase family conserved in yeast. Nat. Genet. 13:175, 1996. Lemke, G., and Axel, R.: Isolation and sequence of a cDNA encoding the major structural protein of peripheral myelin. Cell 40:501, 1985. Lewis, R. A., and Sumner, A. J.: Electrophysiologic features of inherited demyelinating neuropathies: a reappraisal. Ann. N. Y. Acad. Sci. 883:321, 1999. Lewis, R. S., Sumner, A. J., and Shy, M. E.: Inherited neuropathies. In Brown, W. F., Bolton, C. F., and Aminoff, M. J. (eds.): Neuromuscular Function and Disease. Philadelphia, W. B. Saunders, p. 1017, 2002. Lus, G., Nelis, E., Jordanova, A., et al.: Charcot-Marie-Tooth disease with giant axons: a clinicopathological and genetic entity. Neurology 61:988, 2003. Marques, W. Jr., Sweeney, J. G., Wood, N. W., et al.: Central nervous system involvement in a novel connexin 32 mutation affecting identical twins. J. Neurol. Neurosurg. Psychiatry 66:803, 1999. Marrosu, M. G., Vaccargiu, S., Marrosu, G., et al.: CharcotMarie-Tooth disease type 2 associated with mutation of the myelin protein zero gene. Neurology 50:1397, 1998. Massion-Verniory, M. P., Dumont, E., and Potvin, A. M.: Retinite pigmentaire familiale complique d’une amyotrophie neurale. Rev. Neurol. (Paris) 78:561, 1946.

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon 141. Mastaglia, F. L., Nowak, K. J., Stell, R., et al.: Novel mutation in the myelin protein zero gene in a family with intermediate hereditary motor and sensory neuropathy. J. Neurol. Neurosurg. Psychiatry 67:174, 1999. 142. McEntagart, M., Norton, N., Williams, H., et al.: Localization of the gene for distal hereditary motor neuronopathy VII (dHMN-VII) to chromosome 2q14. Am. J. Hum. Genet. 68:1270, 2001. 143. McKusick, V.: Mendelian Inheritance in Man. Baltimore, Johns Hopkins University Press, 1998. Available at: www3.ncbi.nlm.nih.gov/Omim. 144. McLeod, J. G., Low, P. A., and Morgan, J. A.: A family with Charcot-Marie-Tooth disease and Leber’s optic atrophy. Proc. Aust. Assoc. Neurol. 12:23, 1975. 145. Mersiyanova, I. V., Perepelov, A. V., Polyakov, A. V., et al.: A new variant of Charcot-Marie-Tooth disease type 2 is probably the result of a mutation in the neurofilament-light gene. Am. J. Hum. Genet. 67:37, 2000. 146. Milhorat, A. T.: Studies in diseases of muscles: XIV. Progressive muscular atrophy of the optic nerves: report on a family. Arch. Neurol. 50:279, 1943. 147. Misu, K., Yoshihara, T., Shikama, Y., et al.: An axonal form of Charcot-Marie-Tooth disease showing distinctive features in association with mutations in the peripheral myelin protein zero gene (Thr124Met or Asp75Val). J. Neurol. Neurosurg. Psychiatry 69:806, 2000. 148. Misu, K., Yoshihara, T., Yamamoto, M., et al.: Two families of Charcot-Marie-Tooth disease with Adie’s pupil, axonal neuropathy and the Thr124Met mutation in the peripheral myelin protein zero gene. Rinsho Shinkeigaku 40:149, 2000. 149. Mizuno, Y., Otsuka, S., Takano, Y., et al.: Giant axonal neuropathy: combined central and peripheral nervous system disease. Arch. Neurol. 36:107, 1979. 150. Muchir, A., Bonne, G., van der Kooi, A. J., et al.: Identification of mutations in the gene encoding lamins A/C in autosomal dominant limb girdle muscular dystrophy with atrioventricular conduction disturbances (LGMD1B). Hum. Mol. Genet. 9:1453, 2000. 151. Muglia, M., Magariello, A., Nicoletti, G., et al.: A large family with pure autosomal dominant hereditary spastic paraplegia from southern Italy mapping to chromosome 14q11.2-q24.3. J. Neurol. 249:1413, 2002. 152. Muglia, M., Zappia, M., Timmerman, V., et al.: Clinical and genetic study of a large Charcot-Marie-Tooth type 2A family from southern Italy. Neurology 56:100, 2001. 153. Nagamatsu, M., Jenkins, R. B., Schaid, D. J., et al.: Hereditary motor and sensory neuropathy type 2C is genetically distinct from types 2B and 2D. Arch. Neurol. 57:669, 2000. 154. Nakagawa, M., Takashima, H., Umehara, F., et al.: Clinical phenotype in X-linked Charcot-Marie-Tooth disease with an entire deletion of the connexin 32 coding sequence. J. Neurol. Sci. 185:31, 2001. 155. Nelis, E., Erdem, S., Van Den Bergh, P. Y., et al.: Mutations in GDAP1: autosomal recessive CMT with demyelination and axonopathy [see comment]. Neurology 59:1865, 2002. 156. Nelis, E., Haites, N., and Van Broeckhoven, C.: Mutations in the peripheral myelin genes and associated genes in inherited peripheral neuropathies. Hum. Mutat. 13:11, 1999.

1749

157. Nicoziani, P., Vilhardt, F., Llorente, A., et al.: Role for dynamin in late endosome dynamics and trafficking of the cation-independent mannose 6-phosphate receptor. Mol. Biol. Cell 11:481, 2000. 158. Novelli, G., Muchir, A., Sangiuolo, F., et al.: Mandibuloacral dysplasia is caused by a mutation in LMNA-encoding lamin A/C. Am. J. Hum. Genet. 71:426, 2002. 159. Numakura, C., Lin, C., Ikegami, T., et al.: Molecular analysis in Japanese patients with Charcot-Marie-Tooth disease: DGGE analysis for PMP22, MPZ, and Cx32/GJB1 mutations. Hum. Mutat. 20:392, 2002. 160. Ochoa, G. C., Slepnev, V. I., Neff, L., et al.: A functional link between dynamin and the actin cytoskeleton at podosomes. J. Cell Biol. 150:377, 2000. 161. Omran, H., Ketelsen, U. P., Heinen, F., et al.: Axonal neuropathy and predominance of type II myofibers in infantile spinal muscular atrophy. J. Child Neurol. 13:327, 1998. 162. Othmane, K. B., Middleton, L. T., Loprest, L. J., et al.: Localization of a gene (CMT 2A) for autosomal dominant Charcot-Marie-Tooth disease, type 2 to chromosome 1p and evidence of genetic heterogeneity. Genomics 17:370, 1993. 163. Ouvrier, R. A., Prineas, J., Walsh, J. C., et al.: Giant axonal neuropathy—a third case. Proc. Aust. Assoc. Neurol. 11:137, 1974. 164. Panas, M.: Central nervous system involvement in four patients with Charcot-Marie-Tooth disease with connexin 32 extracellular mutations [see comment]. J. Neurol. Neurosurg. Psychiatry 65:947, 1998. 165. Peiffer, J., Schlote, W., Bischoff, A., et al.: Generalized giant axonal neuropathy: a filament-forming disease of neuronal, endothelial, glial, and Schwann cells in a patient without kinky hair. Acta Neuropathol. (Berl.) 40:213, 1977. 166. Perez-Garrigues, H., Burguera, J. A., and Lopez Arlandis Gea, A.: Laryngeal stridor caused by Gerhardt’s syndrome in olivopontocerebellar atrophy and a probable striatonigral degeneration [in Spanish]. An. Otorrinolaringol. Ibero Am. 17:5, 1990. 167. Pericak-Vance, M. A., Speer, M. C., Lennon, F., et al.: Confirmation of a second locus for CMT2 and evidence for additional genetic heterogeneity. Neurogenetics 1:89, 1997. 168. Pierce, T., Worman, H. J., and Holy, J.: Neuronal differentiation of NT2/D1 teratocarcinoma cells is accompanied by a loss of lamin A/C expression and an increase in lamin B1 expression. Exp. Neurol. 157:241, 1999. 169. Pitts, K. R., Yoon, Y., Krueger, E. W., and McNiven, M. A.: The dynamin-like protein DLP1 is essential for normal distribution and morphology of the endoplasmic reticulum and mitochondria in mammalian cells. Mol. Biol. Cell 10:4403, 1999. 170. Poliak, S., Gollan, L., Salomon, D., et al.: Localization of Caspr2 in myelinated nerves depends on axon-glia interactions and the generation of barriers along the axon. J. Neurosci. 21:7568, 2001. 171. Pouliot, J. J., Yao, K. C., Robertson, C. A., and Nash, H. A.: Yeast gene for a Tyr-DNA phosphodiesterase that repairs topoisomerase I complexes. Science 286:552, 1999. 172. Prineas, J. W., Ouvrier, R. A., Wright, R. G., et al.: Giant axonal neuropathy—a generalized disorder of cytoplasmic

1750

173.

174.

175.

176.

177.

178. 179.

180.

181.

182.

183.

184.

185.

186.

187.

188.

189.

Diseases of the Peripheral Nervous System

microfilament formation. J. Neuropathol. Exp. Neurol. 35:458, 1976. Puls, I., Jonnakuty, C., LaMonte, B. H., et al.: Mutant dynactin in motor neuron disease. Nat. Genet. 33:455, 2003. Raffaeledi Barletta, M., Ricci, E., Galluzzi, G., et al.: Different mutations in the LMNA gene cause autosomal dominant and autosomal recessive Emery-Dreifuss muscular dystrophy. Am. J. Hum. Genet. 66:1407, 2000. Rainier, S., Jones, S. M., Esposito, C., et al.: Analysis of microtubule-associated protein 1a gene in hereditary spastic paraplegia. Neurology 51:1509, 1998. Reid, E., Kloss, M., Ashley-Koch, A., et al.: A kinesin heavy dhan (KIF5A) mutation in hereditary spastic paraplegia (SPG10). Am. J. Hum. Genet. 71:1189, 2002. Ressot, C., and Bruzzone, R.: Connexin channels in Schwann cells and the development of the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Brain Res. Rev. 32:192, 2000. Rozental, R., Giaume, C., and Spray, D. C.: Gap junctions in the nervous system. Brain Res. Brain Res. Rev. 32:11, 2000. Sahenk, Z., and Chen, L.: Abnormalities in the axonal cytoskeleton induced by a connexin32 mutation in nerve xenografts. J. Neurosci. Res. 51:174, 1998. Saito, M., Hayashi, Y., Suzuki, T., et al.: Linkage mapping of the gene for Charcot-Marie-Tooth disease type 2 to chromosome 1p (CMT2A) and the clinical features of CMT2A. Neurology 49:1630, 1997. Sambuughin, N., Sivakumar, K., Selenge, B., et al.: Autosomal dominant distal spinal muscular atrophy type V (dSMA-V) and Charcot-Marie Tooth disease type 2D (CMT2D) segregate within a single large kindred and map to a refined region on chromosome 7P15. J. Neurol. Sci. 161:23, 1998. Sancho, S., Magyar, J. P., Aguzzi, A., and Suter, U.: Distal axonopathy in peripheral nerves of PMP22-mutant mice. Brain 122:1563, 1999. Santoro, L., Manganelli, F., Di Maio, L., et al.: CMT2C: A distinct genetic entity. Description of the first European family. Neuromuscul. Disord. 12:399, 2002. Sarnat, H. B., Roth, S. I., and Jimenez, J. F.: Neonatal myotubular myopathy: neuropathy and failure of postnatal maturation of fetal muscle. Can. J. Neurol. Sci. 8:313, 1981. Schelhaas, H. J., Van Engelen, B. G., Gabreels-Festen, A. A., et al.: Transient cerebral white matter lesions in a patient with connexin 32 missense mutation. Neurology 59:2007, 2002. Scherer, S. S., and Arroyo, E. J.: Recent progress on the molecular organization of myelinated axons. J. Peripher. Nerv. Sys. 7:1, 2002. Scherer, S. S., and Salaer, J. L.: Axon-Schwann cell interactions in peripheral nerve regeneration. In Jessen, K. R., and Richardson, W. D. (eds.): Glial Cell Development. Oxford, UK, Oxford University Press, p. 299, 1996. Schiavon, F., Rampazzo, A., Merlini, L., et al.: Mutations of the same sequence of the myelin P0 gene causing two different phenotypes. Hum. Mutat. Suppl 1:S217, 1998. Schmid, C. D., Stienekemeier, M., Oehen, S., et al.: Immune deficiency in mouse models for inherited peripheral neuropathies leads to improved myelin maintenance. J. Neurosci. 20:729, 2000.

190. Seemann, P., Huehne, K., Rautenstrauss, B., et al.: An axonal form of Charcot-Marie-Tooth disease beginning with deafness and abnormal pupillar reaction: new mutation in the myelin protein zero gene (MPZ) in a Czech family. Nervenheilkunde 20:S28, 2001. 191. Seitelberger, F.: Eine unbekannte Form von invantiler lipoidspeicher Krankheit des Gehirns. In Proceedings of the First International Congress of Neuropathology, Vol. 3. Turin, Rosenberg and Sellier, p. 323, 1952. 192. Senderek, J., Bergmann, C., Ramaekers, V. T., et al.: Mutations in the ganglioside-induced differentiationassociated protein-1 (GDAP1) gene in intermediate type autosomal recessive Charcot-Marie-Tooth neuropathy. Brain 126:642, 2003. 193. Senderek, J., Hermanns, B., Lehmann, U., et al.: CharcotMarie-Tooth neuropathy type 2 and P0 point mutations: two novel amino acid substitutions (Asp61Gly; Tyr119Cys) and a possible “hotspot” on Thr124Met. Brain Pathol. 10:235, 2000. 194. Seven, M., Ozkilic, A., and Yuksel, A.: Dysmorphic face in two siblings with infantile neuroaxonal dystrophy. Genet. Couns. 13:465, 2002. 195. Sevilla, T., Cuesta, A., Chumillas, M. J., et al.: Clinical, electrophysiological and morphological findings of Charcot-Marie-Tooth neuropathy with vocal cord palsy and mutations in the GDAP1 gene. Brain 126:2023, 2003. 196. Shackleton, S.: LMNA, encoding lamin A/C, is mutated in partial lipodystrophy [see comment]. Nat. Genet. 24:153, 2000. 197. Shapiro, L., Doyle, J. P., Hensley, P., et al.: Crystal structure of the extracellular domain from P0, the major structural protein of peripheral nerve myelin. Neuron 17:435, 1996. 198. Shen, M. R., Chou, C. Y., Hsu, K. F., et al.: The KCl cotransporter isoform KCC3 can play an important role in cell growth regulation. Proc. Natl. Acad. Sci. U. S. A. 98:14714, 2001. 199. Shiojiri, T., Tsunemi, T., Matsunaga, T., et al.: Vocal cord abductor paralysis in spinocerebellar ataxia type 1. J. Neurol. Neurosurg. Psychiatry 67:695, 1999. 200. Shorer, Z., Philpot, J., Muntoni, F., et al.: Demyelinating peripheral neuropathy in merosin-deficient congenital muscular dystrophy. J. Child Neurol. 10:472, 1995. 201. Silver, J. R.: Spastic paraplegia with amyotrophy of the hands. J. Neurol. Neurosurg. Psychiatry 29:135, 1966. 202. Stojkovic, T., Latour, P., Vandenberghe, A., et al.: Sensorineural deafness in X-linked Charcot-Marie-Tooth disease with connexin 32 mutation (R142Q). Neurology 52:1010, 1999. [Erratum appears in Neurology 52:1952, 1999.] 203. Stuurman, N., Heins, S., and Aebi, U.: Nuclear lamins: their structure, assembly, and interactions. J. Struct. Biol. 122:42, 1998. 204. Sullivan, T., Escanlant-Alcade, D., Bhatt, H., et al.: Loss of A-type lamin expression compromises nuclear envelope integrity leading to muscular dystrophy. J. Cell Biol. 147:913, 1999. 205. Takashima, H., Boerkoel, C. F., John, J., et al.: Mutation of TDP1, encoding a topoisomerase I-dependent DNA damage repair enzyme, in spinocerebellar ataxia with axonal neuropathy. Nat. Genet. 32:267, 2002.

HMSN II (CMT2) and Inherited System Atrophies of Nerve Axon 206. Takashima, H., Nakagawa, M., Nakahara, K., et al.: A new type of hereditary motor and sensory neuropathy linked to chromosome 3. Ann. Neurol. 41:771, 1997. 207. Takashima, H., Suehara, M., Saito, M., et al.: Gene for hereditary motor and sensory neuropathy (proximal dominant form) mapped to 3q13.1. Ann. N. Y. Acad. Sci. 883:449, 1999. 208. Targoff, I. N.: Update on myositis-specific and myositisassociated autoantibodies. Curr. Opin. Rheumatol. 12:475, 2000. 209. Taylor, J.: Peroneal atrophy. Proc R. Soc. Med. 6:50, 1913. 210. Teunissen, L. L., Notermans, N. C., Franssen, H., et al.: Disease course of Charcot-Marie-Tooth disease type 2: a 5-year follow-up study. Arch. Neurol. 60:823, 2003. 211. Tomkins, J., Usher, P., Slade, P., et al.: Novel insertion in the KSP region of the neurofilament heavy gene in amyotrophic lateral sclerosis (ALS). Neuroreport 9:3967, 1998. 212. Verhelst, H. E., Lofgren, A., and Van Coster, R. N.: A family with X-linked dominant Charcot-Marie-Tooth caused by a connexin32 mutation. Eur. J. Paediatr. Neurol. 4:235, 2000. 213. Verhoeven, K., De Jonghe, P., Coen, K., et al.: Mutations in the small GTP-ase late endosomal protein RAB7 cause Charcot-Marie-Tooth type 2B neuropathy. Am. J. Hum. Genet. 72:722, 2003. 214. Vieira, O. V., Bucci, C., Harrison, R. E., et al.: Modulation of Rab5 and Rab7 recruitment to phagosomes by phosphatidylinositol 3-kinase. Mol. Cell. Biol. 23:2501, 2003. 215. Vizioli: Dell’atrofia progr. nervosa. Boll. R. Acad. Medico-chir Napoli, 1889. 216. Vogel, P., Gabriel, M., Goebel, H. H., and Dyck, P. J.: Hereditary motor sensory neuropathy type II with neurofilament accumulation: new finding or new disorder? Ann. Neurol. 17:455–461, 1985. 217. Vogler, O., Bogatkewitsch, G. S., Wriske, C., et al.: Receptor subtype-specific regulation of muscarinic acetylcholine

218. 219.

220.

221.

222.

223.

224.

225.

226.

227.

1751

receptor sequestration by dynamin: distinct sequestration of m2 receptors. J. Biol. Chem. 273:12155–12160, 1998. Wilson, S. A. K.: Neurology, Vol. 1. (Bruce, A. N., ed.). London, Edward Arnold, p. 219, 1940. Wisniewski, K., and Wisniewski, H. M.: Diagnosis of infantile neuroaxonal dystrophy by skin biopsy. Ann. Neurol. 7:377, 1980. Witt, T. N., Garner, C. G., Pongratz, D., and Baur, X.: Autosomal dominant Emery-Dreifuss syndrome: evidence of a neurogenic variant of the disease. Eur. Arch. Psychiatry Neurol. Sci. 237:230, 1988. Wong, M. H., and Filbin, M. T.: Dominant-negative effect on adhesion by myelin P0 protein truncated in its cytoplasmic domain. J. Cell Biol. 134:1531, 1996. Yan, W. X., Archelos, J. J., Hartung, H. P., and Pollard, J. D.: P0 protein is a target antigen in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:286, 2001. Young, P., Grote, K., Kuhlenbaumer, G., et al.: Mutation analysis in Chariot-Marie Tooth disease type 1: point mutations in the MPZ gene and the GJB1 gene cause comparable phenotypic heterogeneity. J. Neurol. 248:410, 2001. Zhang, K., and Filbin, M. T.: Myelin P0 protein mutated at Cys21 has a dominant-negative effect on adhesion of wild type P0. J. Neurosci. Res. 53:1, 1998. Zhao, C., Takita, J., Tanaka, Y., et al.: Charcot-Marie-Tooth disease type 2A caused by mutation in a microtubule motor KIF1Bbeta. Cell 105:587, 2001. Zhao, X., Alvarado, D., Rainier, S., et al.: Mutations in a newly identified GTPase gene cause autosomal dominant hereditary spastic paraplegia. Nat. Genet. 29:326, 2001. Zhu, Q., Couillard-Despres, S., and Julien, J. P.: Delayed maturation of regenerating myelinated axons in mice lacking neurofilaments. Exp. Neurol. 148:299, 1997.

74 Hereditary Brachial Plexus Neuropathy CHRISTOPHER J. KLEIN AND ANTHONY J. WINDEBANK

Overview Historical Overview Clinical Features Natural History

Precipitating or Associated Factors Differential Diagnosis Laboratory Tests Pathologic Alterations of Biopsied Nerves

OVERVIEW Inherited neuropathies typically present with progressive, symmetrical sensory loss or weakness. They are underdiagnosed for complex reasons: (1) impairments may develop insidiously over years with only mild neurologic handicaps and therefore are not recognized as abnormal, (2) pain may be absent, or (3) simultaneously occurring diseases (e.g., diabetes mellitus, alcoholism, spine disease) may obscure the major inherited cause. Hereditary brachial plexus neuropathy (HBPN), by contrast, presents with attacks of multifocal nerve injury affecting the upper limbs. Attacks are acute or subacute in onset and often are associated with severe pain, significant morbidity, and disability. Typically attacks subside with recovery of impairments. Frequently, the inherited nature of HBPN is also overlooked and the attacks are mislabeled as radiculopathies, cervical rib, tumor, and musculoskeletal shoulder disorders. The pathophysiologic localization is unilateral or asymmetrical in the brachial plexus or nerves arising from the plexus or roots, without a generalized neuropathy, making it different from other inherited neuropathies with potential for focal injury (e.g., hereditary neuropathy with tendency to pressure palsies [HNPP]). HNPP may also have multifocal involvement of nerves but is associated with a generalized peripheral neuropathy and is genetically distinct from HBPN in having deletion mutations and micromutations in the gene encoding peripheral myelin protein 22 (PMP22), localized to 17q11.2. In some families, the locus for HBPN is at chromosome position 17q25. However, other loci are predicted to be associated with the disorder. Original descriptions of

Pathogenesis Pattern of Inheritance Molecular Genetics Treatment

HBPN postulated a potential inflammatory component. Recent study of nerve biopsies provides evidence for an inflammatory component in both acquired and inherited forms of brachial plexus neuropathy (BPN). Inflammatory infiltrates in nerve appear to explain the fiber degeneration of some attacks and provide a rationale for controlled trials of immune-modulating therapy.

HISTORICAL OVERVIEW In 1886, Dreschfeld described a 43-year-old woman who had suffered three discrete episodes of a painful upper limb weakness over a 3-year period.8 He used the term rheumatic peripheral neuritis and mentioned that the patient’s sister had similar attacks. He reported, “The sister of the patient has been under the care of Dr. Ross and has had similar attacks, seven in number; her case was looked upon by Dr. Ross as one of syringomyelia, and I have Dr. Ross’s authority that he now believes that his case, like mine, was really a case of peripheral neuritis.” Taylor, in 1960, studied a family in which five generations were affected with single or recurrent attacks in multiple individual nerves with predilection for the proximal brachial region.47 In 1961, Jacob et al. described in seven patients in two unrelated families severe attacks involving upper extremity nerve(s).22 The patterns of mendelian inheritance and the degree of penetrance and expressivity have been noted. Comments of genetic susceptibility to “hyperergic reaction” were made, and specific precipitating triggers were noted. Additional families have been described.9,15,19,21,37,42,49,51,54 From these cases and others, 1753

1754

Diseases of the Peripheral Nervous System

along with genetic and pathologic descriptions, a reasonable understanding of this disorder has emerged. Included in the features of HBPN are (1) autosomal dominant inheritance with varied penetrance and expressivity; (2) recurrent attacks of painful upper extremity weakness and atrophy with positive neuropathic sensory symptoms and sensory loss with varied severity, extent, and resolution; (3) association of attacks with constitutional stressors, including pregnancy and parturition, mild trauma, infection, and immunization; and (4) multifocal or patchy axonal degeneration of upper limb(s) nerves, plexus, and possibly roots with inflammation in areas of degenerating fibers.

CLINICAL FEATURES Children in the first decade of life may be affected by HBPN, but the more usual onset is in the second or third decade. The precipitating factors are pregnancy and parturition, mild trauma, immunization, and concomitant infection. Recurrence of acute episodes over a period of years with intervening recovery is most typical, but a single attack is also consistent with the diagnosis, especially when other family members are clearly affected. In almost all cases, the onset of weakness is preceded by prominent pain in the involved limb. The pain is often extreme. The quality of the pain is “deep” and “aching,” often located in the shoulder, scapula, or arm. It may radiate into the neck or forearm, and occasionally to the hand. The pain is commonly severe and has qualities described as “sharp,” “stabbing,” “throbbing,” or “burning.” Features of the pain appear sympathetically mediated, including allodynia (perception of pain from a stimulus that is normally not noxious) and hyperpathia (exaggerated subjective response to painful stimuli). Autonomic features may be more common in those patients with prolonged attacks. Pain may be worse at night. Swelling of the extremity and trophic skin changes such as sweating and vasomotor changes have been noted.50 As with the sporadic form (ParsonageTurner syndrome),14,33,48 we have frequently found the pain to be acute in onset. Many patients are awoken from sleep, and the course may be of such a rapidity as to suggest cervical disc extrusion or rotator cuff tears. The absence of worsening with the Valsalva maneuver may be helpful in distinguishing this pain from the pain of extruded disc(s). The neuropathic features of skin allodynia, paresthesias, and sensory alterations can help in exclusion of rotator cuff or other musculoskeletal causes of pain that mimic this condition. The pain is typically made worse by limb movement, particularly at the shoulder, and patients may voluntarily hold the arm immobile against the trunk with elbow flexion and shoulder adduction, a characteristic posture of rotator cuff tears. Patients and their physicians tend to gravitate toward splinting to help man-

age pain, and the ensuing weakness and atrophy may be mistakenly attributed to disuse. The duration of severe pain in the sporadic form is extremely varied, ranging from hours to months to years. The length of the severest pain often correlates with the ensuing degree of muscle weakness. Those rare individuals with painful attacks lasting only hours (so-called abortive attacks) may not develop weakness.19 Weakness and atrophy are prominent features of this disorder and may worsen as the pain abates. The areas of atrophy and occasional fasciculation often localize to sites of prior or ongoing pain; this feature is more common in the shoulder and arm (Fig. 74–1). Weakness may plateau several days to weeks after onset of the pain and usually does not worsen after 1 month. Exceptions occur. “Smoldering” cases with pain and weakness not resolving before the next attack have been termed chronic or chronic undulating1,19,50,51 and are thus distinguished from the classic or relapsing-remitting forms. Limited data initially suggested that the chronic undulating form may have unique pathogenesis because one family did not link to chromosome 17q25.50 However, these forms may not necessarily be genetically distinct. In our practice, we have noted individuals with both forms in the same family. Additionally, one group noted families with the classic form that did not have linkage to 17q25.52 Constitutional and environmental differences beyond the involved gene or genes and their abnormality(ies) are likely responsible for the varied course and presentations of muscle weakness and other features. The pattern of weakness is often useful in identifying the multifocal nature characteristic of this disorder. The muscles innervated by C5 and C6 segmental levels are most commonly affected, which may be helpful to distinguish this disorder from vertebral and disc diseases that tend to affect lower cervical levels (C7-C8). Similarly, the prominent involvement of shoulder musculature may help to distinguish this from other multifocal vasculitic processes in which distal hand and forearm musculature may be more commonly affected. Multifocality exists within nerves, plexus, and likely roots. For example, a single finger flexor (a C7 median-innervated muscle) may be weak, along with an isolated weakness at the infraspinatus (a C5 suprascapular-innervated muscle). This provides evidence that the lesions are separated by space occurring either at the cervical roots, the brachial plexus, or possibly branches of the median and suprascapular nerves because other muscles innervated by the same nerves are spared. Although termed brachial plexus neuropathy, the pattern of deficits often provides evidence of a process independent of the plexus proper. Included are examples of proximal involvement at isolated nerves: the long thoracic (C5-C7), suprascapular (C5-C6) (see Fig. 74–1), and phrenic (C3-C5). Individual distal nerves arising from the plexus (i.e., the anterior interosseous) may also be

Hereditary Brachial Plexus Neuropathy

FIGURE 74–1 Two patients having suffered an attack of HBPN and the subsequent atrophy, which is common. A, A patient with isolated painful weakness of his right infraspinatus, which came on 3 days after knee surgery and in the setting of a mild Staphylococcus aureus infection. Magnetic resonance imaging showed increased T2-weighted signal in not only the infraspinatus but also the supraspinatus and deltoid muscles, suggestive of the multifocal process characteristic of this disorder. Electromyography and clinical examination could not find neuropathic injury outside of the suprascapular nerve to the infraspinatus. No structural process was identified. At the time this photo was taken, he had resolution of his pain but continued to have improving weakness in the infraspinatus 1 year later. B, The mother of the patient in A had a more typical attack with multiple muscles affected and residual prominent deltoid atrophy. One sister, not shown, had attacks postoperatively regardless of form of anesthesia (i.e., epidural, general).

affected, sometimes in isolation. Some have noted cranial nerve palsies in the families they have studied. Most commonly affected are the tenth nerve, producing hoarseness and difficulty in swallowing; the seventh nerve, producing a typical Bell’s palsy; and the eighth nerve, producing a unilateral sensorineural hearing loss of subacute onset.

1755

The cranial nerve palsies more commonly are painless than are attacks of limb weakness, although the anterior interosseous attacks may also be painless. Presentations of individual nerve involvement are likely no less common in the inherited form, but the literature provides more examples in sporadic cases, including examples affecting cranial,38 phrenic,3,26,34,39 and forearm and shoulder14,35,43 nerves. Of additional note is the occasional involvement of the lower limb lumbosacral plexus. In some series this occurred in up to 33% of attacks.50 The lower extremity involvement can occur separately from or simultaneously with an attack of the upper limbs; isolated lumbosacral attacks may be more common in the chronic undulating form.50 The lower extremity attacks are similar to those described in lumbosacral plexus neuropathy.12,16,41 Loss of sensation in affected areas is less frequently reported by patients than the dramatic positive symptoms of pain, asleep-numbness (pins and needles), and prickling. Severe pain, including the characteristic allodynia, may obscure identification of specific sensory loss by patients and their physicians. In one large Dutch series following 101 attacks of HBPN, 58% of patients had “vital hypesthesia” and 33% “gnostic hypesthesia.” Vital hypesthesia refers to decreased pain and/or temperature sensation, whereas gnostic hypesthesia describes vibration sense deficiency. The data were obtained near the time of attacks, and long-term followup was not available.50 Sensory loss occurs in up to 66% of well-studied patients with the sporadic form.35,48 The sensory abnormalities, whether positive or negative, are often patchy and do not always correlate with the predicted nerve territory based on the motor examination. Such features provide further evidence to the clues of the underlying multifocal pathogenesis. In the sporadic form, the lateral shoulder is most commonly affected by loss of feeling, consistent with the prominent shoulder motor involvement.48 We have similar clinical experience with the loss of feeling in the hereditary condition; in addition, the loss of feeling in the shoulder may involve the lateral aspect of the forearm, which corresponds to the sensory branch (lateral antebrachial) originating from the upper trunk (C5-C6). Jacob and co-workers first remarked on the minor dysmorphic features associated with this condition.22 High arches and hammer toes or other features of inherited polyneuropathy are not found. Among the reported dysmorphic features, the most commonly found are hypotelorism (close-set eyes) (Fig. 74–2),9,15,18,19,21,22,42 epicanthic folds, short stature,22 cleft plate,15,22 and minor degrees of syndactly.18,19 Many of these early accounts fail to provide details of their determination of “hypotelorism” and “short stature.” More recently, studies of the intercanthal distance and/or interpupillary distance found that 67%50 to 76%23 of affected members of families had eye distances only 1 standard deviation

1756

Diseases of the Peripheral Nervous System

FIGURE 74–2 Left, The hypotelorism and epicanthal folds in some patients led at least one author to compare the dysmorphisms to the facial portraits of Italian painter and sculptor Amedeo Modigliani.9 We are impressed by how many patients without attacks but of the same family also have these features, suggesting the “dysmorphisms” may in fact relate to a simple familial tendency. Right, This patient with attacks had intercanthal, outer-canthal, and interpupillary distances within the 2nd to 25th percentile of normal, as did her sisters, who had not suffered from attacks. Mildly prominent epicanthal folds are also suggested. The family was not able to distinguish facial features of affected and unaffected members.

below the mean. An association with deep skinfolds and creases in the neck, limbs, and scalp has been reported in some families.23 Unfortunately, normative data for facial features and limb skinfolds are difficult to find and are in part racially determined, limiting the value of such observations. The ancestral origins of affected patients are increasingly emphasized given recent work suggesting that, in most U.S. pedigrees, a single mutation may have begun in one family lineage generations before geographic separation of later generations of the family (e.g., a founder effect).53 This may explain how different kindreds could have similar facial appearance and would suggest that these features are unlikely to relate to the primary pathogenesis (i.e., a connective tissue abnormality). This is supported by our observation that unaffected family members often have the same facial appearances of their affected family members. Rigorous attempts to address this specific question are lacking. Certainly, these “dysmorphisms” are not present universally. Ungley’s photographs of these patients do not appear to have the characteristic facial features.49 This absence of such features may imply different genetic cause. One such family is described without dysmorphism, and linkage excluded the 17q25 locus.52

NATURAL HISTORY The natural history of attacks is markedly varied, as emphasized by those patients labeled “chronic undulating” and, on the opposite extreme, “abortive,” as dis-

cussed earlier. Such variability may be seen among individuals both within and between families, and even the same individual may have markedly different severity between attacks. Most attacks, however, begin acutely or subacutely and are without development of new deficits by 2 to 4 weeks. As a rule, patients regain total function of the affected limb(s) within 1 year to 18 months, and many before 1 year, although reflex and clinical examination abnormalities may still be present after described “recovery.” Such examination deficits are more detectable in patients with recurrent or more severe events. The extent of improvement obviously depends on the severity of attack and the specific nerves affected. Although different nerves may be similarly affected, the specific localization provides for different morbidity and, therefore, perceived length of attacks by patients. As an example, the rare patient with diaphragmatic phrenic nerve involvement may have increased morbidity, although the actual injury to the nerve is no greater than in less critical nerves. Similarly, a laborer may describe a prolonged attack that is no different in duration from a “brief and mild” attack described by an office employee. Women who have postpartum attacks are most distressed and describe “horrific attacks” in which deficits prevent them from caring for their loved newborn child. In general, patients should be encouraged to understand that the disorder tends toward improvement, and it is rare that attacks lead to the types of neural destruction that prevent partial or total recovery. Attacks tend to become less frequent with increasing age, and it is unusual for patients in the sixth, seventh, and eighth decades to undergo attacks.47,50

Hereditary Brachial Plexus Neuropathy

PRECIPITATING OR ASSOCIATED FACTORS Understanding the range of precipitating factors implicated in attacks is important not only in the diagnosis of patients but likely in the understanding of the underlying pathogenesis. Both the sporadic or Parsonage-Turner variety and HBPN can have similar precipitants, although in both, clear precipitants may not be found. Taylor noted that, in the inherited group, infection often superseded attacks.47 Other authors have subsequently reported similar associations.9,19,22 Most common, however, is the association of attacks without any clear precipitant, which occurred in 73% of 101 attacks in one series.50 Even though the number of reported familial cases is increasing, it is not possible to make definitive epidemiologic statements. In the largest previously reported kinship,47 a total of 24 of 119 family members over five generations were affected by the disease. Many suffered multiple episodes, but in only six cases was it thought that infection preceded attacks. Although there are obvious problems with such

1757

comparison, 40 sporadic patients of 98 had an infection or immunization prior to their brachial plexus attack.48 These data may erroneously suggest that postinfectious attacks are less common in the inherited form. Pregnancy and parturition may be a more likely precipitant in the familial form compared to sporadic cases, but this is uncertain. We recently described a four-generation family with 17 affected members among whom 10 of 12 women with attacks had postpartum events (Fig. 74–3).24 This compares to 0 women of 29 who had sporadic attacks in the study of Tsairis et al.48 Other investigators note the occurrence of postpartum attacks in patients without family history and suggest it is no more common in the familial form.27 Unfortunately, much of the cited literature presented no discussion of family history, either positive or negative. One woman in our described kindred with postpartum events had her first attack after extended labor, including arm straining and prolonged Valsalva maneuvers. With a subsequent pregnancy, she underwent elective cesarean section under epidural anesthesia at 37 weeks of pregnancy to reduce the theoretical mechanical

FIGURE 74–3 This family affected by HBPN had prominent postpartum attacks occurring in 10 of the 12 affected women shown. The route of delivery or the specific conditions surrounding pregnancy or delivery did not seem to influence the development of attacks. One patient (III-8) underwent elective cesarean delivery under epidural anesthesia in an attempt to prevent the Valsalva maneuvers and arm strain characteristic of her prior delivery and subsequent attack. Regardless, within 3 hours of delivery she developed her typical attack. (Modified from Klein, C. J., Dyck, P. J. B., Friedenberg, S. M., et al.: Inflammation and neuropathic attacks in hereditary brachial plexus neuropathy. J. Neurol. Neurosurg. Psychiatry 73:45, 2002.)

1758

Diseases of the Peripheral Nervous System

forces of labor on the brachial plexus. Throughout delivery she had no neuromuscular symptoms; however, 3 hours after delivery she awoke with severe left shoulder pain and dysesthesia and paresthesias of the thumb and forearm (features characteristic of her previous attack). This and other observations suggest that postpartum attacks do not correlate with the specific events of labor, including type of delivery, degree of straining, anesthesia, and other factors.9,18,21,22,24,47 Similar observations have been noted in the sporadic form.27 In the familial form there appears to be a predominance of reported attacks in the first decade of life compared to the sporadic form.9,15,18,19,21,22,47,54 Several cases are recorded in which children born with brachial plexus lesions are not distinguished from the children with traumatic Erb’s or Klumpke’s paralysis. These children recover from their “birth injuries” but later show typical manifestations of the familial disease. The mothers of these children did not have peripartum or postpartum attack at the time of their affected child’s birth. We are unaware of any case of passive transfer of an attack to an infant among the many postpartum or peripartum attacks reported in both familial and sporadic cases. If an immunogenic mechanism is responsible, this would suggest that the primary abnormality more likely resides in the nerves or supporting tissues and not in an abnormal humoral response, but such inferences are less strong than one would hope. Many patients with episodes occurring in the first decade of life later are discovered to have a family history of attacks or themselves have subsequent attacks. Similarly, those with recurrent attacks should be suspected of having the familial form. Recurrence appears to be more common in the hereditary form and is likely explained by a genetic susceptibility to possible environmental triggers. There may be a slight predominance of this condition in men. If true, it may relate to occupations and lifestyles of men, which may predispose them to certain exposures (e.g., in war). Mild injuries to the affected limb have been thought to trigger familial,47 sporadic,48 and epidemic2 forms of the disease. From review of the prior literature, male preponderance is 1.27:1.56 However, if we include the kindreds described within this chapter, that ratio drops to 1.1:1. In the idiopathic form of acute BPN, Magee and DeJong30 found a male-to-female ratio of 10.5:1 and Tsairis and colleagues48 a ratio of 2.4:1. If the 122 patients from the two series are combined, the overall ratio is 2.9:1. If there is a predominance of HBPN in men given the known autosomal dominant inheritance, environmental or hormonal influences may be implicated in expression, but major difficulties in ascertainment limit the strength of such suggestions. Finally, emotional stress has been implicated in attacks. Individuals have had attacks in the setting of serious life emotional stressors. For example, in the previous edition of this text, Windebank described a woman who had developed an attack in the setting of her husband having been taken

as an innocent hostage in a standoff between police and a group of illicit drug smugglers.56 Whether such an event is purely coincidental or is causative is unclear.

DIFFERENTIAL DIAGNOSIS In HBPN the symptoms, distribution of neurologic findings, and course of attacks are not distinguishable from the sporadic form, known as “immune brachial plexus neuropathy” (Parsonage-Turner syndrome).30,35,48 Many of the earliest clinical reports of the sporadic form failed to mention either the presence or absence of a family history. Some probably were inherited, given the number with recurrent attacks. Tendency to recurrence, early age of onset, and postpartum and posttraumatic occurrence may be more common in the inherited form(s), but these features are probably not specific. Therefore, HBPN must remain in the differential of all patients diagnosed with the Parsonage-Turner syndrome. The most important distinguishing feature between HBPN and the sporadic form is family history, which is often hard to obtain given the difficulties in ascertainment of family history and likely de novo mutations. Identification of attacks and consideration of family history are the first steps in making the diagnosis. The variability in the signs and symptoms at presentation and their similarity to other neurologic and non-neurologic disorders may lead patients to present to internists, orthopedic surgeons, otolaryngologists, or pulmonologists before neurologists. Many patients are given incorrect diagnoses and the syndrome is missed. Once an attack is identified, casual discussion of family history is often inadequate; in our experience, less than half of affected families are aware of their own familial occurrence, even when other family members are clearly affected (Fig. 74–4). Many factors may contribute to obscuring genetic susceptibility in addition to de novo mutations: (1) small numbers or young ages of affected individuals within the family; (2) social discord, which prevents contacting relatives; and (3) misdiagnosed attacks in other family members. The last possibility is the most common. Therefore, meeting with and examining potentially affected family members is useful. Understanding the symptoms, signs, temporal course, and associated or precipitating events as described earlier can help to limit the differential diagnosis. Because some alternative diagnoses are correctable or have serious implications, it is important to consider and test for these possibilities. This is true even in individuals with prior attacks and known family history. During the early period of an initial attack, orthopedic consultation may be the first referral; this occurred in 64% of one group of patients with the sporadic form.17 This is especially true when pain is the initial major feature. Included among the orthopedic possibilities are plexus trauma or traction, rotator cuff injury, acute calcified

Hereditary Brachial Plexus Neuropathy

1759

FIGURE 74–4 Pedigrees of two families affected by HBPN but initially unaware of their familial tendency. Male-to-male transmission characteristic of autosomal dominant inheritance and incomplete penetrance is also illustrated. Some individuals of pedigree 2 clearly carried the gene (III-8, IV-2, and IV-6), but they were unaware of attacks despite aggressive questioning and clinical testing. In both families, varied diagnoses (rotator cuff injury, tendonitis, and bursitis) led to the families’ inability to recognize their familial tendency. Individual II-2 in pedigree 1 developed a paralytic shoulder after minor trauma in a sand-lot football game during grade school. He had been diagnosed with rotator cuff strain but recalled distinct paresthesias and severe weakness such that he could only lift a can of soup with his arm and shoulder, with the symptoms resolving over 1 year. Decades later he developed an anterior interosseus syndrome, which was labeled a complex nerve entrapment at the ligament of Struthers, but did not improve with surgical intervention. When his son (III-1) presented with recurrent attacks of arm paralysis, we performed electromyography on the father and found, in both the father and son, bilateral chronic diffuse denervation of multiple upper extremity muscles characteristic of brachial neuritis. Pedigree 2 illustrates that, even when multiple individuals are affected, the proband may not be aware. After formal interviews and clinical exams, 10 definitely affected individuals were identified. (Modified from Klein, C. J., Dyck, P. J. B., Friedenberg, S. M., et al.: Inflammation and neuropathic attacks in hereditary brachial plexus neuropathy. J. Neurol. Neurosurg. Psychiatry 73:45, 2002.)

tendonitis, adhesive capsulitis, and bursitis. As discussed earlier, families with this disease may have involvement outside the brachial plexus, mimicking idiopathic Horner’s syndrome, recurrent laryngeal nerve palsy, other cranial nerve palsies, or phrenic nerve paralysis. The latter are often referred to internists and pulmonologists. Bilateral diaphragmatic paralysis may be initially mistaken for acute asthma, pulmonary embolus, and congestive heart failure or various forms of diaphragmatic paralysis. As in the sporadic form, we have seen individuals with anterior interosseous syndrome or predominant suprascapular neuropathy (see Figs. 74–1 and 74–4). Of special note is the occasional occurrence of lumbosacral plexopathies that could be mistaken

for diabetic or nondiabetic forms of amyotrophy or lumbosacral radiculoplexus neuropathies.12 On rare occasions, patients with HNPP may develop a brachial plexopathy because of direct pressure (e.g., from a heavy backpack or sleeping on a hard surface). The HNPP condition, however, is clinically and genetically quite distinct from HBPN. The distinctions have been clarified by the discovery that deletions and rare point mutations in PMP22 result in HNPP that localizes to 17p11.2.57 Patients with HNPP develop compression neuropathies at common sites of entrapment and have a mild length-dependent sensory neuropathy. High arches and hammer toes are often also present in HNPP, a feature not seen in HBPN. Nerves

1760

Diseases of the Peripheral Nervous System

of patients with HNPP are frequently enlarged, although this is not strictly a hypertrophic neuropathy. Pain and atrophy of muscles is uncommon. Other diagnostic possibilities with episodic flares and lower extremity involvement would include relapsing inflammatory demyelinating neuropathy associated with inherited neuropathy.11 This usually presents with a more insidious onset and with prominent lower extremity involvement. As with sporadic cases, HBPN must be distinguished from compression or invasion of roots or plexus by a malignant process. Conversely, cervical radiculopathy caused by a herniated intervertebral disc should not be overlooked in the patient with a family history of BPN. Finally, it is worth mentioning the point made by Taylor that, even in families with a clear history of this disease, peripartum conversion disorder may be the correct diagnosis (R. A. Taylor, personal communication, 1984). Of note, however, is that several of our patients with postpartum pain had been diagnosed as suffering from functional or psychiatric disorders. Several were not aware of their family history. Each suffered from transient attacks lasting hours to days, and some had no demonstrable motor or sensory deficit. One woman had profoundly painful attacks without signs both following delivery and after three separate surgeries. Only after she developed a full-blown attack resulting in a paralytic arm did she become aware of her mother’s postpartum and brother’s postsurgical attacks (see Fig. 74–1). Sporadic BPN may occur in clusters. A regional environmental exposure or a predisposition of kindreds within the community could explain this phenomenon. Multiple family members affected at the same time should raise suspicion. The clustered forms seem more common in times of war, with widespread vaccination, and in the setting of unsanitary conditions. Occasionally such clustered cases occur in members of the same family. This observation may be in keeping with increased susceptibility but more likely may represent a heightened exposure to a specific trigger. Wyburn-Mason at the National Hospital, London, first suggested that BPN might occur in epidemic form.58 Between 1924 and 1940, a total of 13 cases of brachial neuritis were identified, whereas he personally saw 42 cases between December 1940 and August 1941, with a peak incidence in June and July 1941. At the same time, he noted an increased incidence of unexplained lateral popliteal nerve palsies. An occasional association between a febrile illness and the onset of symptoms was noted, and he postulated that the neuropathy was the result of a viral illness or an allergic response to such an infection. All of these cases occurred in London during the height of the bombing of that city during World War II. Bardos and Somodska, who recorded data on 265 patients with BPN, noted such a relationship to poor hygiene.2 The cases occurred in an industrial district of northeast Czechoslovakia between 1949 and 1953 and were concentrated among certain types of factory workers, with the morbidity rate being as high as 109 cases

per 1000 workers in some groups during the 5-year period. An etiologic agent was not identified, but common factors included crowded working conditions, living in dormitories under poor hygienic conditions, and being employed in a job involving sustained vigorous use of the upper limbs. Within these epidemics, a particular familial incidence was not noted. Other authors record the simultaneous occurrence of BPN in several members of a single family.7,31 In these cases there is no antecedent family history, and all the cases occur at the same time, suggesting that family members have been exposed to a common environmental agent, possibly a virus, and all have responded in the same way because of their genetic similarity. One report of cases occurring simultaneously in identical twins has been made.7 Recent patients with flavivirus (West Nile) infection have presented with segmental motor syndromes in the setting of flulike illnesses. The extent of involvement of the lower limbs has prevented confusion with brachial neuritis syndromes, but the absence of central nervous system symptoms and signs in some may suggest the potential for confusion with this disorder if an arm were selectively affected.20,28

LABORATORY TESTS No characteristic hematologic or biochemical changes are seen in association with HBPN; in particular, the white blood cell count and erythrocyte sedimentation rate are typically not elevated. Magnetic resonance imaging (MRI) should exclude structural process in roots, plexus, or nerves. Occasionally, increased MRI T2-weighted signal in various muscles can indicate neurogenic changes beyond the area of perceived involvement. Such imaging can be helpful in suggesting the multifocal involvement characteristic of this disorder, which sometimes is not seen on electrophysiologic testing. An example is the patient shown in Figure 74–1. His clinical examination and the results of electromyography (EMG) were consistent with an isolated suprascapular neuropathy. MRI of the shoulder, however, suggested a more diffuse process including the contralateral limb, where no symptoms or signs existed. The imaging abnormalities consisted of increased T2-weighted signal in the brachial plexus and suprascapular nerve along with various muscles (deltoid, biceps, and teres minor), characteristic of a multifocal denervation. The cerebrospinal fluid (CSF) is typically normal, but occasionally patients can have mild protein elevation, suggesting that the pathologic process may extend proximally to nerve roots. Such observations are more established in the sporadic form, in which CSF protein was elevated in 4 of 35 patients analyzed, and each was thought to have cervical root involvement, because CSF was collected at the time of a myelographic examination.48 An elevated CSF white blood cell count should raise questions about the diagnosis of HBPN.

Hereditary Brachial Plexus Neuropathy

Electrodiagnostic findings are the same as those in the sporadic form of the disease. The findings may conveniently be considered in two different categories: those related to the affected limb and those related to clinically unaffected nerves. Many authors report that the distal nerve conduction velocities in the weak limb are normal or slightly reduced.4,9,15,19,22,40 The amplitudes of the compound muscle action potential and of the sensory nerve action potential are often reduced in the distal segments of clinically affected nerves. In a few cases, distal motor latency from stimulation at Erb’s point has been used as a measure of conduction time through the plexus. In one patient with a typical history of HBPN, distal latency from Erb’s point to the mildly weak biceps muscle was normal.4 In a more detailed study focusing on the electrodiagnostic features of the condition, Dunn and co-workers reported on the findings in three children with the typical clinical syndrome.9 In one case, proximal stimulation during the convalescent phase of the illness showed normal conduction through the plexus. In another series, the lateral antebrachial cutaneous nerve of the forearm was the most commonly abnormal nerve (32% of cases examined) of the nerves tested with nerve conduction tests (NCTs).6 This finding would be consistent with the upper trunk predisposition of both the inherited and sporadic conditions. Needle electromyographic examination has been reported by various authors and has been performed in all of our patients. The findings are consistent with axonal interruption at the level of the brachial plexus or multiple individual nerves, more commonly affecting muscles innervated by the upper trunk. Most commonly there is a lack of electromyographic changes in the cervical paraspinal muscles. Some patients also had cervical paraspinal abnormalities as well as elevated CSF proteins, suggestive of root involvement. It is common to find subclinical electromyographic changes in the contralateral unaffected limb. This overlaps with the observation that a number of these patients have clinical symptoms or signs suggesting simultaneous bilateral plexus involvement. In a number of reports, family members are recorded as having symptoms suggestive of lower limb involvement. However, these patients have not been examined, and there are no reports of the clinical or electrophysiologic findings related to the lumbosacral plexus. A major question concerns the unaffected lower limbs of patients with HBPN. Do they show any signs of a generalized neuropathy? From the clinical viewpoint they do not, even when studied with computer-assisted sensory examination.55 This is true in these individuals both during and after attacks, provided they had no symptoms of the lower extremity. Several authors report normal motor nerve conduction velocities of the lower limbs.15,19,47 Dunn and co-workers found normal motor and sensory nerve conduction velocities in the legs and scattered abnormalities of motor unit potentials.9 They suggested that these lower limb findings

1761

are in keeping with a more broadly distributed mononeuritis multiplex in this condition. Patients do not develop subclinical electrophysiologic changes before attacks. We followed one patient, with a history of postpartum attacks, prospectively during pregnancy and postdelivery, when she developed a typical attack. Serial NCTs, EMG, and quantitative sensation tests (QSTs) were performed in the upper and lower extremities before her attack.24 NCT and QST results of the upper and lower limbs were normal throughout pregnancy at 16, 26, and 35 weeks’ gestation, just 2 weeks before her delivery. Initial needle EMG (16 weeks’ gestation) showed evidence of her old left brachial plexopathy (diffuse large motor unit potentials and no fibrillation potentials). At 2 weeks postpartum and after she had suffered a typical attack beginning 3 hours after delivery, her QST became abnormal, with elevation of vibratory thresholds (hyposensitivity) of the hands. NCTs and EMG were unchanged from the prepartum studies. Twenty weeks after the beginning of the attack, bilateral median and ulnar summated motor and sensory compound muscle action potentials were reduced 50% compared to earlier values. Cooling thresholds were raised (hyposensitivity) and heat-pain thresholds were decreased (hyperalgesia). At 40 weeks, her median and ulnar motor and sensory responses improved and tended toward normalcy, and needle EMG demonstrated multifocal upper extremity fibrillation potentials with neurogenic motor unit potentials. On QST she remained hyposensitive to vibration, and she no longer was hyperalgesic by heat-pain testing. In summary, the electrophysiologic evidence supports the concept of acute scattered axonal damage within the brachial plexus and isolated multiple nerves. EMG may provide evidence of such a process when clinical examination may only suggest involvement of an individual nerve or root. There is no evidence to suggest a generalized diffuse neuropathic process.

PATHOLOGIC ALTERATIONS OF BIOPSIED NERVES The pathologic alterations of the affected nerves during an attack were essentially unknown until recent studies.24 Previously the sural nerve biopsy findings were reported in a patient who may have been suffering from this condition.4,29 This patient developed pain, numbness, and weakness of the left arm with clinical and electrophysiologic features suggestive of a BPN 5 weeks after an influenzalike illness. An atypical finding was apparently asymptomatic weakness of the legs. In this patient, sural nerve biopsy showed focal and sausage-shaped thickenings of the myelin sheath. This patient was reported before genetic testing was available, and it is possible, perhaps likely, that he was suffering from HNPP.

1762

Diseases of the Peripheral Nervous System

The lack of extensive pathologic description is understandable. This is a rare, nonfatal disease, and affected individuals usually recover from episodes. Such episodes are rare after the age of 50 years and thus affected patients are not likely to die of unrelated diseases and come fortuitously to autopsy. Since, in most cases, there is no clinical evidence of generalized neuropathy, it has generally not been thought worthwhile to study this predominantly upper limb disease by sural nerve biopsy. However, the findings of Bradley and co-workers4,29 would bring this assumption into question because their sural nerve biopsy showed tomacula. We recently reported on upper limb nerve biopsies from patients of four separate kindreds during attacks.24 Fascicular radial sensory nerves were obtained in three and a fascicular median motor nerve in another. (The last patient was reported in an earlier edition of this text.56) The biopsies were obtained in order to justify continued treatment with steroids. In situ hybridization for common

viruses in nerve tissue (cytomegalovirus, herpes simplex, Epstein-Barr, varicella-zoster) and genetic testing for HNPP (tomaculous neuropathy) were performed and were negative. Upper extremity nerve biopsies in two patients showed prominent perivascular inflammatory infiltrates with vessel wall disruption (Fig. 74–5). The neuropathic abnormalities were characterized by prominent axonal degeneration (Fig. 74–6). Two patients showed no interstitial or neuropathic injury. Clearly, the fascicles chosen in these patients were not involved by the disease at that level. The negative findings yield the important information that there is no morphologic evidence of a generalized abnormality of nerve fibers or Schwann cells expressed at this level. It also suggests that the pathologic process is extremely multifocal. Pathologic study of biopsied nerve from our four patients also allows several other conclusions: (1) the inflammatory infiltrates were usually associated with epineurial microvessels, (2) axonal degeneration appears

FIGURE 74–5 A, Transverse paraffin section of superficial radial nerve (SRN) stained with hematoxylin and eosin (H&E) from patient III-8 shown in Figure 74–4 (at the time of a HBPN attack). Seen are prominent epineurial mononuclear cell infiltrates surrounding several small vessels adjacent to a nerve fascicle. B, Transverse section of a SRN from patient V-12 (see Fig. 74–4, pedigree 2) also shows epineurial mononuclear cell infiltrates, here histochemically reacted to CD45, a lymphocyte marker. Some of the inflammatory cells appear to infiltrate vessel walls. The degree of inflammation shown was also seen in the endoneurium and subperineurium, indicating that inflammation, probably immune, is playing a pathogenic role in the fiber degeneration. C–E, Serial paraffin sections of nerve microvasculature from patient III-8. C, Blood vessel involvement and tunica media disruption (H&E stain). D, Middle lower paraffin section reacted for CD45 (for leukocytes). E, Paraffin section reacted for smooth muscle actin; note the disruption of smooth muscle. (From Klein, C. J., Dyck, P. J. B., Friedenberg, S. M., et al.: Inflammation and neuropathic attacks in hereditary brachial plexus neuropathy. J. Neurol. Neurosurg. Psychiatry 73:45, 2002, with permission.)

1763

Hereditary Brachial Plexus Neuropathy

FIGURE 74–6 A, Semithin transverse epoxy section of superficial radial nerve (SRN) (obtained at the time of a HBPN attack) stained with methylene blue from patient V-12 (see Fig. 74–4) showing severe axonal degeneration, especially in large fibers, and empty nerve tubes with degenerating myelin. B, Closely applied teased fiber preparations of SRN from patient V-12 showing severe axonal degeneration at various stages. (From Klein, C. J., Dyck, P. J. B., Friedenberg, S. M., et al.: Inflammation and neuropathic attacks in hereditary brachial plexus neuropathy. J. Neurol. Neurosurg. Psychiatry 73:45, 2002, with permission.)

to be typical of the fiber alterations, and (3) mononuclear cell infiltrates occurred in distal nerves, thus providing evidence of multifocal lesions extending to peripheral nerves beyond the brachial plexus (a BPN). The pathologic findings are consistent with the clinical response. Two of our patients appeared to respond to intravenous (IV) methylprednisolone, at least as judged by improvement of symptoms (Neuropathy Symptoms and Change [NSC] score) and stabilization of signs (Neuropathy Impairment Score).24 Pain relief appeared to be the greatest benefit of treatment based on NSC evaluations. High-dose IV methylprednisolone (1 g/dose) appeared to be more beneficial than low-dose oral prednisone. We believe these biopsies provide evidence that inflammation, possibly immune mediated, appears to be pathogenic for some if not all attacks of HBPN. We cannot exclude the possibility that these pathologic findings are a reactive process, but the degree of inflammation, tendency to vessel wall disruption, and multifocal nature combined with at least a stabilization of signs and a dramatic reduction in pain with administration of high-dose steroids all

lead to the conclusion that inflammation is at least in part pathogenic for attacks. The microvessel wall disruption is characteristic of a small-vessel vasculitis. Inflammation has also been reported in the sporadic form.5,13,46

PATHOGENESIS More than inheritance appears to be involved in development of attacks of BPN. Other constitutional and environmental factors may also be involved. HBPN may be analogous to acute intermittent porphyria, in which the clinical expression of the neuropathic disorder requires an external stress on a defective enzyme system. In the familial predisposition to BPN, ingested toxins or drugs do not appear to play a significant role. Specifically, infection, vaccination, and other factors, as stated earlier, appear to be a precondition for some attacks of HBPN. Because antecedent strenuous use of an affected limb may predispose to attacks, a mechanical effect on nerves may be hypothesized. Alternatively, intense limb use and

1764

Diseases of the Peripheral Nervous System

BPN attacks may co-occur simply by chance. Because attacks often follow parturition, mechanical compression and distortion of the brachial plexus during vaginal delivery was considered. Specifically, one feature common to all vaginal deliveries is the performance of intense Valsalva maneuvers by the mother. This pressure to expel the baby is generated in the thorax, and it is possible that the plexus is in some way distorted or compressed as it passes over the apex of the lung. This mechanical effect might, in susceptible individuals, lead to the development of a plexopathy. With a putative role of the mechanical factors of parturition in mind, we scheduled an elective cesarean section for one of our patients, who subsequently developed an attack despite not having gone into labor, evidence against the hypothesis for mechanical stress on the brachial plexus. Because emotional stress has been linked with rare attacks, the influence of humoral changes in the hypothalamic-pituitary-adrenocortical axis should be considered. Whether these changes explain the association between parturition and the development of the neuropathy is unclear. Thus the stress associated with delivery may induce large changes in the circulating concentrations of endogenous corticosteroids, which in turn alter the sensitivity or response of the immune system. Several other mechanisms may be involved to explain the association with parturition. These would include either a direct or an indirect effect of the profound endocrine changes that occur at this time and an effect of the various drugs used at the time of delivery. Neither spinal anesthesia nor the use of nitrous oxide gas has been linked to the development of the neuropathy. Immunologic mechanisms are implicated not only by the association with immunization and infection but on the basis of several other lines of evidence. First, the multifocal involvement of nerves is typical of immunologic injury, including varieties of rheumatologic, diabetic, and other forms of vasculitis. Second, some attacks are associated with the unusual immunologic events that occur during pregnancy and parturition, at the time of attacks. Specifically, the mother’s immune system tolerates the presence of foreign tissue (the fetus) without attempting rejection. This tolerance ends abruptly at parturition. Another factor is that, toward the end of pregnancy and especially at the time of delivery, circulating fetal protein and cells enter the maternal circulation. These might act as triggers to the plexus-directed immune response. Finally, the pathologic alterations identified with attacks are associated with prominent microvessel wall inflammation, suggestive of other inflammatory processes of nerve tissue that are believed to be pathogenic. If immunogenic or inflammatory mechanisms are responsible, a primary question is whether the defect lies within the affected nerves or is the result of aberrant circulating humoral protein(s). The preferential involvement of the upper extremity nerves might support a tissue-

specific abnormality of nerve, but this is only conjecture. However, at this point passive transfer of an attack to an unborn child by any woman with a peripartum attack has not been reported, arguing against a humoral mechanism(s). So perhaps a loss of tolerance to an aberrant antigen in the upper extremity nerve tissue might be predicted during times of immune alteration or constitutional stressors. Other alternative mechanisms would include abnormality in diverse regulators of nerve apoptosis and metabolic processes of nerve maintenance, but no specific data exist to suggest these mechanisms. Identification of the gene or genes responsible and their abnormality(ies) will provide clarity to this issue.

PATTERN OF INHERITANCE The HPBN trait is inherited in an autosomal dominant fashion with varied expression and questionable penetrance (see Fig. 74–4). In one thoroughly evaluated kinship studied by Taylor47 and by one of us (A. J. W.), many of the individuals at risk in this family were seen and examined. In the kinship, 30% of individuals at risk of inheriting the disease were affected. This would compare with a theoretical value of 50% for pure autosomal dominance with 100% penetrance and 100% ascertainment. Thus the observed figure translates to a penetrance of 60%. Despite aggressive attempts at ascertainment, certain inherent pitfalls exist. The question of the degree of penetrance is difficult to address because of the problem of ascertainment of kinship history. Thus Dyck and co-workers have shown that, for inherited neuropathies in general, the apparent penetrance depends on how vigorously the affected kin are sought.10 The apparent degree of penetrance increases as one successively uses the ascertainment techniques of hearsay family evidence, clinical examination of kin, special sensory and physiologic evaluation of kin, and finally nerve biopsy of kin. In our experience, EMG and NCTs greatly increase the sensitivity in diagnosis of this condition, and upper extremity nerve biopsy is likely of lower yield in this patchy resolving condition. Accurate ascertainment is critical in linkage analysis.

MOLECULAR GENETICS Pellegrino et al., in 1996, established linkage of HBPN to the distal part of chromosome 17q in two pedigrees.36 The work was later confirmed in a large family, and flanking markers provided localization to a locus of 16 cM at 17q24q25.45 Homogeneity of this causal locus was originally suggested when four additional families showed linkage to the same region. Recently genetic heterogeneity was supported when two families did not show linkage to 17q24-25.52 The region at 17q has further been narrowed

1765

Hereditary Brachial Plexus Neuropathy

to approximately 4.0 cM.45 Most recently, Watts et al. have narrowed the region to a 1-Mb region at 17q25.53 A founder effect was suggested by genotyping 12 markers from seven separate U.S. pedigrees (Fig. 74–7). Specifically, haplotypes were similar among six of the seven pedigrees, and a minimally shared region suggested a further narrowing to a region of approximately 500 kb between markers 72GT1 and D17S939.53 Candidate gene approaches and cloning will be important to identify the responsible gene. Currently the use of haplotype analysis may allow for the genotyping of families that are too small for linkage analysis and may provide for a possible diagnostic tool prior to the cloning of the gene. Candidate gene approaches have greatest success when (1) the implicated chromosomal region is small, (2) the genes within the area are well characterized, and (3) pathogenesis is understood. In HBPN, the presumed localization provides for a relatively small region (500 kb to 1 Mb). However, a number of untranslated regions reside within this region, and other open reading frames without known

function make deductive candidate approaches more difficult. Based on the prominent perivascular epineurial inflammation seen on nerve biopsies, perhaps the genes implicated in disease are somehow important in regulation of the immune system at nerve tissue. Genes important in apoptosis, transcription regulators, binding proteins, and others must be considered as possibilities as well. Mild dysmorphic features, including epicanthic folds, hypotelorism, redundant skinfolds, and possibly an increased incidence of cleft palate, imply a possible connective tissue defect. Stogbauer et al.44 suggested, prior to the current localization at 17q25, that the proline 4-hydroxylase ␤-polypeptide (P4HB) gene, which is involved in connective tissue formation, may be responsible. This gene lies outside the current 500-kb localization, but given the plasticity of the known genetic maps, this region cannot be excluded. Meuleman and colleagues examined four candidate genes between markers at 17q25. The genes encoded (1) MLL, a septin-like fusion protein; (2) thymidine kinase 1, soluble; (3) N-acetylgalactosamine␣2,6-sialyltransferase and (4) SEC14-like protein.32 Several intronic and exonic polymorphisms were detected, but no disease-causing mutations were found. Only MLL, the septin-like fusion protein, localizes within the currently identified 500-kb region. Also previously examined were the general splicing factor gene SFRS2 and the c-myb ET locus, and both were without specific mutation.25 Other genes proposed as candidates also lie outside the 500-kb region. All are telomeric, including those for (1) galactosidebinding lectin (LGALS3BP); (2) interleukin enhancer binding factor 1 (ILF1); and (3) tissue inhibitor of metalloproteinase 2 (TIMP2).44 Provided the current mapping data are correct, these candidates seem unlikely. Better characterization of the known genes and gene fragments that lie within the putative causative site at 17q25 is needed. Such work will be essential in the identification of the causative gene of HBPN, and other families not localizing to this region will need genome-wide linkage analysis to identify new responsible chromosomal regions important in the eventual discovery of alternative causative genes. We observed, using the polymorphic markers described by Watts et al.,53 that 20% of well-studied HBPN kindreds do not carry the conserved haplotype at 17q25. These data provide emphasis to the identification of additional responsible loci.

FIGURE 74–7 Ideogram showing the polymorphic markers and the narrowed candidate region believed responsible for the most common genetic form of HBPN at 17q25. Both the narrowed region based on meiotic recombinants (A) and the founder effect (B) are shown. (Modified from Watts, G. D., O’Briant, K. C., and Chance, P. F.: Evidence of a founder effect and refinement of the hereditary neuralgic amyotrophy [HNA] locus on 17q25 in American families. Hum. Genet. 110:166, 2002.)

TREATMENT At this time the first line of management involves identification and education of individuals at risk. The risk of the child of an affected individual inheriting the trait is 50%, and the disease is expressed in approximately 60% of those who inherit the trait. Because penetrance is less than 100%, it is theoretically possible for clinically nonaffected parents to

1766

Diseases of the Peripheral Nervous System

carry the gene and bear affected children (see Fig. 74–4). Pregnancy and delivery are preventable risks for attacks, but this option has serious social implications. Counsel about the association of trauma, strain, possible emotional stress, and infections with attacks is reasonable, but prevention of events is difficult. When counseling families, it must be remembered that the condition is compatible with a likely normal lifespan, and individuals usually recover well from episodes. Important modes of treatment are the symptomatic management of pain by the use of adequate analgesia. Physical therapy is of importance because almost all patients recover functional use of the limb, but the period of profound weakness may be extended for months, during which joint and soft tissue mobility must be maintained. Clinical trials with IV methylprednisolone and possibly IV immunoglobulins need to be performed. Although patients treated with high-dose IV methylprednisolone appeared to have response, pain was the major improvement.24 As with all anecdotal reports, improvements must be interpreted cautiously. This is especially true in disorders such as HBPN, in which individual attacks have markedly variable courses. The morbidity of this disorder, the pathologic observations, and the results of open trials with IV methylprednisolone emphasize the need for clinical trials. It will be necessary to carefully distinguish inherited from cryptogenic variants to ensure that a potentially nonresponsive group does not interfere with recognition of a responding group. We believe this condition is more common than appreciated, and therefore clinical trials are feasible. Knowing if such therapies are beneficial is emphasized, because correction of the underlying gene(s) defect(s) may not be possible.

REFERENCES 1. Arts, W. F., Busch, H. F., Van den Brand, H. J., et al.: Hereditary neuralgic amyotrophy: clinical, genetic, electrophysiological and histopathological studies. J. Neurol. Sci. 62:261, 1983. 2. Bardos, V., and Somodska, V.: Epidemiologic study of a brachial plexus neuritis outbreak in northeast Czechoslovakia. World Neurol. 2:973, 1961. 3. Beydoun, S. R., and Rodriguez, R.: Neuralgic amyotrophy misdiagnosed as diaphragmatic rupture. Muscle Nerve 19:1181, 1996. 4. Bradley, W. G., Madrid, R., Thrush, D. C., and Campbell, M. J.: Recurrent brachial plexus neuropathy. Brain 98:381, 1975. 5. Cusimano, M. D., Bilbao, J. M., and Cohen, S. M.: Hypertrophic brachial plexus neuritis: a pathological study of two cases. Ann. Neurol. 24:615, 1988. 6. Cwik, V. A., Wilbourn, A. J., and Rorick, M.: Acute brachial neuropathy: detailed EMG findings in a large series. Muscle Nerve 13:859, 1990. 7. de Graaf, A. S.: Neuralgic amyotrophy in identical twins with Ehlers-Danlos syndrome. Electroencephalogr. Clin. Neurophysiol. 33:436, 1972.

8. Dreschfeld, J.: On some of the rarer forms of muscular atrophies. Brain 9:187, 1886. 9. Dunn, H. G., Daube, J. R., and Gomez, M. R.: Heredofamilial branchial plexus neuropathy (hereditary neuralgic amyotrophy with branchial predilection) in childhood. Dev. Med. Child Neurol. 20:28, 1978. 10. Dyck, P. J., Oviatt, K. F., and Lambert, E. H.: Intensive evaluation of referred unclassified neuropathies yields improved diagnosis. Ann. Neurol. 10:222, 1981. 11. Dyck, P. J., Swanson, C. J., Low, P. A., et al.: Prednisoneresponsive hereditary motor and sensory neuropathy. Mayo Clin. Proc. 57:239, 1982. 12. Dyck, P. J., and Windebank, A .J.: Diabetic and nondiabetic lumbosacral radiculoplexus neuropathies: new insights into pathophysiology and treatment. Muscle Nerve 25:477, 2002. 13. Dyck, P. J. B., Engelstad, J., and Suarez, G. A.: Biopsied upper limb nerves provide information about the distribution and mechanism in immune brachial plexus neuropathy [abstract]. Neurology 56:A395, 2001. 14. England, J. D., and Sumner, A. J.: Neuralgic amyotrophy: an increasingly diverse entity. Muscle Nerve 10:60, 1987. 15. Erikson, A.: Hereditary syndrome consisting in recurrent attacks resembling brachial plexus neuritis, special facial features, and cleft palate. Acta Paediatr. Scand. 63:885, 1974. 16. Evans, B. A., Stevens, J. C., and Dyck, P. J.: Lumbosacral plexus neuropathy. Neurology 31:1327, 1981. 17. Favero, K. J., Hawkins, R. H., and Jones, M. W.: Neuralgic amyotrophy. J. Bone Joint Surg. Br. 69:195, 1987. 18. Gardner, J. H., and Maloney, W.: Hereditary brachial and cranial neurities, special facial features, and cleft palate. Neurology 18:278, 1968. 19. Geiger, L. R., Mancall, E. L., Penn, A. S., and Tucker, S. H.: Familial neuralgic amyotrophy: report of three families with review of the literature. Brain 97:87, 1974. 20. Glass, J. D., Samuels, O., and Rich, M. M.: Poliomyelitis due to West Nile virus. N. Engl. J. Med. 347:1280, 2002. 21. Guillozet, N., and Mercer, R. D.: Hereditary recurrent brachial neuropathy. Am. J. Dis. Child. 125:884, 1973. 22. Jacob, J. C., Andermann, F., and Robb, J. P.: Heredofamilial neurities with brachial predilection. Neurology 11:1025, 1961. 23. Jeannet, P. Y., Watts, G. D., Bird, T. D., and Chance, P. F.: Craniofacial and cutaneous findings expand the phenotype of hereditary neuralgic amyotrophy. Neurology 57:1963, 2001. 24. Klein, C. J., Dyck, P. J. B., Friedenberg, S. M., et al.: Inflammation and neuropathic attacks in hereditary brachial plexus neuropathy. J. Neurol. Neurosurg. Psychiatry 73:45, 2002. 25. Kuhlenbaumer, G., Meuleman, J., and Schirmacher, A.: Mutation analysis of a putative sialyltransferase gene, the SFRS2 splicing factor gene and c-myb ET-locus in two families with hereditary neuralgic amyotrophy (HNA). Ann. Hum. Genet. 62:397, 1998. 26. Lahrmann, H., Grisold, W., Authier, F. J., and Zifko, U. A.: Neuralgic amyotrophy with phrenic nerve involvement. Muscle Nerve 22:437, 1999. 27. Lederman, R. J., and Wilbourn, A. J.: Postpartum neuralgic amyotrophy. Neurology 47:1213, 1996. 28. Leis, A. A., Stokic, D. S., Polk, J. L., et al.: A poliomyelitislike syndrome from West Nile virus infection. N. Engl. J. Med. 347:1279, 2002.

Hereditary Brachial Plexus Neuropathy 29. Madrid, R., and Bradley, W. G.: The pathology of neuropathies with focal thickening of the myelin sheath (tomaculous neuropathy). J. Neurol. Sci. 25:45, 1975. 30. Magee, K. R., and DeJong, R. N.: Paralytic brachial neuritis. JAMA 174:112, 1960. 31. Martinelli, P., Pazzaglia, P., Marchiori, L., and Lugaresi, E.: Simultaneous occurrence of neuralgic amyotrophy in three members of one family. Eur. Neurol. 19:316, 1980. 32. Meuleman, J., Kuhlenbaumer, G., Audenaert, D., et al.: Mutation analysis of 4 candidate genes for hereditary neuralgic amyotrophy (HNA). Hum. Genet. 108:390, 2001. 33. Misamore, G. W., and Lehman, D. E.: Parsonage-Turner syndrome (acute brachial neuritis). J. Bone Joint Surg. Am. 78:1405, 1996. 34. Mulvey, D. A., Aquilina, R. J., Elliott, M. W., et al.: Diaphragmatic dysfunction in neuralgic amyotrophy: an electrophysiologic evaluation of 16 patients presenting with dyspnea. Am. Rev. Respir. Dis. 147:66, 1993. 35. Parsonage, M. J., and Turner, J. W.: Neuralgic amyotrophy: the shoulder girdle syndrome. Lancet 1:973, 1948. 36. Pellegrino, J. E., Rebbeck, T. R., Brown, M. J., et al.: Mapping of hereditary neuralgic amyotrophy (familial brachial plexus neuropathy) to distal chromosome 17q. Neurology 46:1128, 1996. 37. Pellicot, R. J., and Chabert, A.: Forme familiale et recidivante de la paralysie amyotrophique de la ceinture scapulaire. Rev. Neurol. (Paris) 112:557, 1965. 38. Pierre, P. A., Laterre, C. E., and Van den Bergh, P. Y.: Neuralgic amyotrophy with involvement of cranial nerves IX, X, XI and XII. Muscle Nerve 13:704, 1990. 39. Pieters, T., Lambert, M., Huaux, J. P., and Nagant de Deuxchaisnes, C.: Hemidiaphragmatic paralysis, an unusual presentation of Parsonage-Turner syndrome. Clin. Rheumatol. 7:402, 1988. 40. Roos, D., and Thygesen, P.: Familial recurrent polyneuropathy: a family and a survey. Brain 95:235, 1972. 41. Sander, J. E., and Sharp, F. R.: Lumbosacral plexus neuritis. Neurology 31:470, 1981. 42. Smith, B. H., Ramakrishna, T., and Schlagenhauff, R. E.: Familial brachial neuropathy: two case reports with discussion. Neurology 21:941, 1971. 43. Spillane, J.: Localised neuritis of the shoulder-girdle. Lancet 2:532, 1941. 44. Stogbauer, F., Young, P., Kuhlenbaumer, G., et al.: Hereditary recurrent focal neuropathies: clinical and molecular features. Neurology 54:546, 2000.

1767

45. Stogbauer, F., Young, P., Timmerman, V., et al.: Refinement of the hereditary neuralgic amyotrophy (HNA) locus to chromosome 17q24-q25. Hum. Genet. 99:685, 1997. 46. Suarez, G. A., Giannini, C., Bosch, E. P., et al.: Immune brachial plexus neuropathy: suggestive evidence for an inflammatory-immune pathogenesis. Neurology 46:559, 1996. 47. Taylor, R. A.: Heredofamilial mononeuritis multiplex with brachial predilection. Brain 83:113, 1960. 48. Tsairis, P., Dyck, P. J., and Mulder, D. W.: Natural history of brachial plexus neuropathy: report on 99 patients. Arch. Neurol. 27:109, 1972. 49. Ungley, C. C.: Recurrent polyneuritis in pregnancy and the puerperium affecting three members of a family. J. Neurol. Psychopathol. 14:15, 1933. 50. van Alfen, N., van Engelen, B. G., Reinders, J. W., et al.: The natural history of hereditary neuralgic amyotrophy in the Dutch population: two distinct types? Brain 123:718, 2000. 51. Warot, P., Petit, H., Nuyts, J. P., and Meignie, S.: Familial brachial amyotrophic neuritis: study of 2 families [in French]. Rev. Neurol. (Paris) 128:281, 1973. 52. Watts, G. D., O’Briant, K. C., Borreson, T. E., et al.: Evidence for genetic heterogeneity in hereditary neuralgic amyotrophy. Neurology 56:675, 2001. 53. Watts, G. D., O’Briant, K. C., and Chance, P. F.: Evidence of a founder effect and refinement of the hereditary neuralgic amyotrophy (HNA) locus on 17q25 in American families. Hum. Genet. 110:166, 2002. 54. Weiderholt, W.: Familial brachial neuropathy. Excerpta Med. Int. Cong. Ser. 296:255, 1973. 55. Windebank, A. J.: Inherited recurrent focal neuropathies. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. P. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 1656, 1984. 56. Windebank, A. J.: Inherited recurrent focal neuropathies. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1137, 1993. 57. Windebank, A. J., Schenone, A., and Dewald, G. W.: Hereditary neuropathy with liability to pressure palsies and inherited brachial plexus neuropathy—two genetically distinct disorders. Mayo Clin. Proc. 70:743, 1995. 58. Wyburn-Mason, R.: Brachial neuritis occurring in epidemic form. Lancet 2:662, 1941.

75 Autosomal Recessive Hereditary Motor and Sensory Neuropathies ANNEKE GABREËLS-FESTEN AND P. K. THOMAS

Classification of Hereditary Motor and Sensory Neuropathies Historical Overview Recent Developments Autosomal Recessive HMSN Autosomal Recessive HMSN Type I Chromosome 5q32/KIAA1985 Chromosome 8q13-q21.1/GDAP1 Chromosome 8q24/NDRG1 Chromosome 10q21-q22/EGR2

Chromosome 11p15/SBF2 (MTMR13) Chromosome 11q23/MTMR2 Chromosome 17p11.2/PMP22 Chromosome 18q23-qter/CTDP1 (Autosomal Recessive Neuropathy) Chromosome 19q13/PRX Autosomal Recessive HMSN I with Unknown Gene Loci Autosomal Recessive HMSN Type II Chromosome 1q21.2/LMNA

CLASSIFICATION OF HEREDITARY MOTOR AND SENSORY NEUROPATHIES Historical Overview Dyck classified the hereditary motor and sensory neuropathies (HMSNs) on clinical, electrophysiologic, and morphologic criteria into demyelinating Charcot-MarieTooth (CMT) disease (HMSN I/CMT1), neuroaxonal Charcot-Marie-Tooth disease (HMSN II/CMT2), and a more severe, early-onset demyelinating Dejerine-Sottas disease (HMSN III).42 HMSN type I and type II are usually autosomal dominantly (AD) inherited, but some families show X-linked inheritance.127 The occurrence of autosomal recessive (AR) types I and II has been well documented by Harding and Thomas.61 Dejerine-Sottas disease, or HMSN type III, was considered to be of AR inheritance.42 Typical pathologic features in some demyelinating HMSN cases led to the designation of distinct subtypes. Congenital hypomyelination (amyelination) neuropathy (CHN) refers to a severe congenital or early infantile form of Dejerine-Sottas neuropathy (DSN) with no or extremely thin myelin sheaths and onion bulbs mainly composed of basal membranes in peripheral nerve.76,90 The abundant

Chromosome 5q /SMN1 (Autosomal Recessive Axonal Neuropathy) Chromosome 8q21.1/GDAP1 Chromosome 10q23.2 Chromosome 11q13-q21/IGHMBP2 Chromosome 15q13-q15/SLC12A6 Chromosome 16q24/Gigaxonin Chromosome 19q13.3 Autosomal Recessive HMSN II with Unknown Gene Loci

occurrence of focally folded myelin has been described in AD as well as AR forms of demyelinating HMSN, and clinical severity ranged from the milder CMT1 phenotype to the more severe DSN form.35,69,114 The AR variant of this subtype with myelin folding was designated CMT4B,14 although some patients from other families may show a DSN phenotype.50

Recent Developments The rapid progress in molecular genetic investigation of the HMSN disorders during the last decade of the 20th century has brought down many of the traditional views and concepts. It is now evident that defects in a myelin gene may result in CMT1, DSN, CHN, or the variant with focal myelin folding, and even CMT2, which constitute all phenotypes within the spectrum of HMSN. Each phenotype may be caused by defects in different genes, and each phenotype may result from AD- as well as AR-inherited mutations. Following the elucidation of the genetic defects, the traditional classification system of HMSN is being supplemented and partially replaced by a genetic classification. However, for the practicing clinician this has resulted in 1769

1770

Diseases of the Peripheral Nervous System

a classification system that is becoming more and more complex, confusing, unwieldy, and inadequate. A complicating factor is that different authors use different denominations for the same disorder, and one denomination may even be used for different genetic entities. Therefore, a classification system based on genetic defects or gene loci with reference to the range of phenotypic expressions seems to be more appropriate. In Table 75–1, the AR subtypes of HMSN are classified by the established genetic defects or gene loci, supplemented by the common denominations and the reported range of phenotypic presentations. The distinction between HMSN type I and type II depends on whether the genetic defect is preferentially expressed in a Schwann cell /myelin gene (type I) or in a neuroaxonal gene (type II). This roughly correlates with, but is not identical to, the former division into demyelinating and neuroaxonal disorders, which division was predominantly based on nerve conduction velocities and nerve pathology. An “axonal neuropathy” may occur with unequivocal myelin gene defects,144 and nerve conduction velocities below 38 m/s or even below 35 m/s have

been reported in primary axonal neuropathies. 51 However, it must be noted that nerve conduction velocity values are unreliable when evoked compound muscle action potentials (CMAPs) are very small (⬍0.5 mV).45 Moreover, it may be difficult in chronic neuropathies to distinguish on morphologic criteria whether the process is primarily demyelinating or axonal in nature. Numerous thinly myelinated axons in the presence of significant fiber loss suggest axonal regeneration rather than (segmental) remyelination. Also, it may be difficult to distinguish between a true onion bulb, resulting from chronic demyelination and remyelination, and a regenerative cluster with many Schwann cell sprouts containing unmyelinated axons. Finally, it is possible that true “intermediate” forms exist with genetic defects that are about evenly expressed in both Schwann cells and axons.77,132,152 Type III is no longer included in the classification because it represents not a genetic subtype, but rather a distinct phenotypic expression. It has to be noted that the present classification is not a static one, but is subject to future new developments and insights.

Table 75–1. Autosomal Recessive HMSN Gene Locus

Common Name

Autosomal Recessive HMSN Type I 5q32 CMT4C 8q24 HMSN-Lom 10q21-q22 — 11p15 11q23 17p11.2

CMT4B2 CMT4B1 CMT1A

18qter 19q13 Unknown Unknown

CCFDN CMT4F — CMT4B3

Autosomal Recessive HMSN Type II 1q21.2-q21.3 AR CMT2A 5q SMA type 0 8q21.1 CMT4A 8q21.1 CMT2K 8q21.1 CMT2H 10q23 HMSN-Russe 11q13-q21 SMARD1/ SIANRF 15q13-q15 ACCPN 16q24 GAN 19q13.3 — Unknown — Unknown — Unknown —

Genetic Defect

Phenotype

KIAA1985 NDRG1 EGR2 inhibitory domain SBF2/MTMR13 MTMR2 PMP22 cytoplasmic domain CTDP1 PRX — —

CMT1 and early severe scoliosis CMT1 and deafness DSN (CHN)

LMNA SMN1 GDAP1 GDAP1 GDAP1? — IGHMBP2

CMT2 Severe congenital spinal muscular atrophy CMT2 (CMT1) CMT2 and hoarseness CMT2 and pyramidal features CMT2 Severe infantile axonal neuropathy and diaphragmatic weakness CMT2 and agenesis of corpus callosum CMT2 and cerebellar, pyramidal features and kinky hair CMT2 Early childhood axonal neuropathy CMT2 and mental retardation, deafness CMT2 and acrodystrophy

SLC12A6 Gigaxonin — — — —

CMT1 with myelin folding ⫾ early-onset glaucoma CMT1 with myelin folding CMT1 CMT1 and cataract, facial dysmorphism, small stature CMT1–DSN and paresthesia, pain DSN (CHN) CMT1 with myelin folding

Autosomal Recessive Hereditary Motor and Sensory Neuropathies

AUTOSOMAL RECESSIVE HMSN HMSN with AR inheritance is less common than AD and X-linked forms, but it is encountered more frequently in communities where consanguineous marriages are customary. Although, in general, AD neuropathies result more frequently from mutations in structural proteins, AR forms are more likely to reveal defects in proteins involved in cell growth, differentiation, and signaling.100 The fast-expanding numbers of newly discovered genetic entities in AR HMSN are contributing significantly to the understanding of the molecular basis of neuropathies and to an increasing insight into the biology and pathophysiology of peripheral nerve. Axonal and demyelinating motor and sensory neuropathies are frequent manifestations of a spectrum of neurodegenerative diseases of the nervous system, multisystem disorders, or metabolic diseases. Involvement of peripheral nerves in these conditions is often subclinical and overshadowed by symptoms of the central nervous system or other features. A number of these complex forms of HMSN are referred to as HMSN type V (with spastic paraplegia), type VI (with optic atrophy), and type VII (with retinitis pigmentosa). A detailed description of these complex forms of HMSN is outside the scope of this chapter. Only those forms are discussed in which peripheral neuropathy is a main feature and the disorder is clearly defined by clinical or morphologic features or an established gene locus. Each form is classified by its gene locus and genetic defect, followed by a listing of the various common names for the disorder. Reference is made to the listing given in the Online Mendelian Inheritance in Man (OMIM) electronic database (see www.ncbi.nlm.nih.gov/Omim/) when available.

AUTOSOMAL RECESSIVE HMSN TYPE I A growing number of AR-inherited forms of demyelinating HMSN are being identified. It has to be noted that in this chapter the forms presenting with a DSN are also included under the heading of HMSN type I. Not included are those mutations in the peripheral myelin protein 22 (PMP22) and protein zero (P0) genes, which cause a mild phenotype in the heterozygous state but may result in a more severe phenotype in the homozygous form (see Chapters 70 and 71).

Chromosome 5q32/KIAA1985 Synonyms: CMT1 AR C,125 CMT4A2,63 CMT4C [OMIM 601596]. In 1996, LeGuern and co-workers identified a gene locus for an AR form of CMT1 to chromosome 5q23-q33 in two Algerian families.87 Subsequently, linkage to chromosome 5q32 was detected in AR HMSN type I kindred from different geographic areas, including five Dutch families, one

1771

French family, and three other families from Mediterranean countries, indicating that mutations of the responsible gene exist in the gene pool of other areas.47,56 A few years later, the responsible gene was identified: KIAA1985, encoding for a uncharacterized transcript. Protein-truncating mutations or missense mutations were found in this novel gene either in the homozygous or in the compound heterozygous state.133 Onset of this chronic progressive sensorimotor neuropathy is usually in early childhood, often interfering with early motor development, but in a minority of cases onset is in the second decade. Foot deformities, usually pes cavus, are present in all patients. Nearly all patients develop a severe and disabling scoliosis or kyphoscoliosis in the first decade, often as early as 3 years of age. The severity and progression of the spine deformities require corrective surgery in about two third of the patients, but this is not necessarily associated with a severe disability. In fact, several patients had consulted an orthopedic surgeon for operative correction of a severe scoliosis before the relatively mild neuropathy was diagnosed. Cranial nerve involvement (hypoacusis, facial weakness) is reported in some patients, and a few patients show a nystagmus. Usually, no palpably enlarged nerves are noted. Cerebrospinal fluid (CSF) protein is normal. In most patients neuropathic signs remain stationary or are only slightly progressive over many years. Some patients become wheelchair dependent or show respiratory insufficiency after a longer course of the disease.47,49,78,87,133 Motor nerve conduction slowing displays a wide range from 4 to 37 m/s in the median nerve (mean, 23.4 m/s). A remarkable difference in conduction slowing between nerves of the same limb may exist. Sensory action potentials are diminished in amplitude or absent.47,49,56,78,133 Unique pathologic changes in combination with an early and severe scoliosis were recognized as early as 1992 by Gabreëls-Festen and colleagues in families of Dutch and Turkish ancestry with an AR form of HMSN type I.49 Later, the same pathology was demonstrated in one of the Algerian patients.78 Subsequently, localization of the genetic defect to chromosome 5q32 was confirmed in Dutch and Turkish families.47 Nerve pathology is characterized by a reduction in the myelinated fiber population and evidence of demyelination /remyelination (Fig. 75–1A). In some biopsies there is a conspicuous variation in myelinated fiber density between fascicles of the same nerve. “Classic” onion bulbs are few, but multiple single or duplicated basal membranes surround many myelinated, demyelinated, or unmyelinated fibers (Fig. 75–1B). Schwann cells of both unmyelinated and myelinated fibers show multiple cytoplasmic protrusions. Occasionally, a broad ring of Schwann cell processes surrounds a myelinated or nonmyelinated fiber. Few focal myelin thickenings are present. In biopsies of some older patients, peculiar large denervated bands are found (Fig. 75–1C).47,49 A few earlier reports described probably identical cases (case 1,139 “intermediate HMSN,”101 and some sporadic Swedish cases58,111).

A

B

C

FIGURE 75–1 A, Sural nerve pathology of an 8-year-old boy with 5q32 AR HMSN I. Several nonmyelinated and small myelinated fibers are surrounded by multiple basal membranes (arrows). Bar: 5 ␮m. B, 5q32 AR HMSN I with single and duplicated basal membranes around nonmyelinated and myelinated fiber, and peculiar cytoplasmic protrusions of Schwann cell. Bar: 1 ␮m. C, 5q32 AR HMSN I with a large band of denervated Schwann cells and Schwann cells of unmyelinated fibers. Bar: 2 ␮m.

Autosomal Recessive Hereditary Motor and Sensory Neuropathies

KIAA1985 is strongly expressed in human brain, spinal cord, and peripheral nerve. The translated protein defines a new protein family of unknown function with putative orthologues in vertebrates. Members of this protein family contain multiple Src homology 3 (SH3) and tetratricopeptide repeat (TPR) domains likely involved in the formation of protein complexes.133

Chromosome 8q13-q21.1/GDAP1 Synonyms: CMT4A,13 CMT1 AR A,125 CMT4A/8q13,149 CMT4A163 [OMIM 214400]. The first AR form of HMSN was mapped on chromosome 8q13-q21.1 in four Tunisian families with a supposedly demyelinating neuropathy, although in a second paper the authors also considered the possibility of a primary axonal disorder.13,15 In 2002, the gene of the gangliosideinduced differentiation-associated protein-1 (GDAP1) was identified as the causative gene.11 This gene is also mutated in an axonal form of AR HMSN, under which heading both forms are discussed (see below).

Chromosome 8q24/NDRG1 Synonyms: HMSN-Lom,71 CMT1 AR D,125 CMT4C/ NDRG1,149 CMT4A3,63 CMT4D [OMIM 601455]. HMSN-Lom (HMSNL) was initially identified in Bulgarian gypsies living in the city Lom in Bulgaria. To date, the disease is known to occur in many patients from gypsy families across Europe. The disorder has been mapped to chromosome 8q24 and has subsequently been demonstrated to be due to a homozygous mutation in the N-myc downstream-regulated gene-1 (NDRG1).31,70,71 Nearly all patients show the same nonsense mutation (Arg148stop) and share conserved haplotypes in the HMSNL region, suggesting a single founder mutation.9,28,33,99 One patient with an identical phenotype showed a homozygous splice-site mutation of NDRG1.67 Onset is in the first decade of life with progressive lower limb weakness soon followed by upper limb weakness and severe distal atrophy. Distal sensory loss affecting all modalities is present, most evident in lower limbs. Tendon reflexes are absent, and in younger patients depressed reflexes may be elicited in upper limbs. Foot deformities are common, and some patients show a mild scoliosis. Hearing loss resulting from a neuropathy of the auditory nerve is noticed in the second or third decade.28,73 A few patients have dysphonia and some have mild involvement of facial muscles. One patient showed cerebellar atrophy on magnetic resonance imaging (MRI).72 Motor nerve conduction velocity (MNCV) is severely reduced in the upper limbs (8 to 16 m/s), with increased distal motor latency in young patients, indicating the demyelinating nature of the neuropathy. In the lower

1773

limbs and in older patients, motor responses are usually unobtainable, attesting to the rapid course of an axonal involvement. Sensory nerve action potentials are absent. Electromyography shows signs of severe denervation. Abnormal brainstem auditory evoked potential recordings, suggesting retrocochlear involvement of the central auditory pathways, are present in all patients before the onset of the subjective hearing loss.28,73 Nerve biopsy demonstrates a demyelinating neuropathy with an early and severe depletion of myelinated axons. The surviving fibers are of small size with thin myelin sheaths. Poorly developed onion bulbs are observed in early cases, which later regress in more advanced cases, probably secondary to axonal loss (Fig. 75–2). Ultrastructural features occasionally include curvilinear axonal inclusions resembling the distal axonopathy of experimental vitamin E deficiency, or pleomorphic inclusions in the adaxonal Schwann cell cytoplasm, and uncompaction of the innermost myelin lamellae. Endoneurial collagen is markedly increased.9,33,73,79,99 Preliminary studies localize NDRG-1 in the Schwann cell cytoplasm without evidence of axonal expression. NDRG-1 is possibly involved in growth arrest and cell differentiation during development and in the maintenance

FIGURE 75–2 Sural nerve pathology of a 13-year-old boy with AR HMSN-Lom reveals severe loss of myelinated fibers and a few poorly developed onion bulbs. Bar: 20 ␮m. See Color Plate

1774

Diseases of the Peripheral Nervous System

of the differentiated state in the adult, possibly as a Schwann cell–signaling protein necessary for axonal survival.70

Chromosome 10q21-q22/EGR2 Synonyms: CHN C,125 CMT4A /D.1,149 CMT4E [OMIM 129010]. Mutations in the gene of the transcription factor early growth response element 2 (EGR2), located on chromosome 10q21-q22, cause disorders of myelination. EGR2 is the human orthologue of the mouse Krox20 gene, both belonging to a family of genes encoding DNA-binding proteins that play a role in genetic control at the transcriptional level (see Chapter 72). Warner and colleagues demonstrated a homozygous mutation in the inhibitory domain of the EGR2 gene in three brothers with CHN. Their nonsymptomatic, consanguineous parents were heterozygous carriers of the mutation.154 Clinical, electrophysiologic, and pathologic features of this form of AR HMSN I are discussed in Chapter 72.

Chromosome 11p15/SBF2 (MTMR13) Synonyms: CMT1 AR B2,125 CMT4B/11p15,149 CMT4B263 [OMIM 604563]. A locus to chromosome 11p15 was mapped in a large Tunisian pedigree with focal myelin folding on nerve biopsy.14 Later the causative gene was identified in a Turkish family. This novel gene, named SET binding factor 2 (SBF2) or myotubularin-related protein-13 gene (MTMR13), is a member of the pseudophosphatase branch of myotubularins.134 At the same time, this gene was shown to be responsible for AR HMSN I with focal myelin folding, associated with early-onset glaucoma in two families from Tunisia and Morocco.7 This association was described before in a Brazilian and a Japanese family.5,80 Both phenotypes share a similar demyelinating neuropathy. Patients show normal early motor milestones. In childhood they start to complain of increasing difficulty in walking. All patients show a steppage gait as a result of severe weakness of distal leg muscles. Some patients become wheelchair bound early in the second decade. Weakness and atrophy of small hand muscles are found in the majority. Distal sensory loss is noticed in lower limbs. Most patients show total tendon areflexia; in a minority, upper limb reflexes are still elicitable. The CSF protein is moderately increased. In some families the neuropathy is associated with a congenital or early juvenile glaucoma.5,7,80 MNCV in the upper limbs is markedly reduced (median MNCV, 22.8 ⫾ 7.3 m/s). Sensory conduction is delayed,

with decreased action potentials, in the upper limbs; sensory responses are usually not obtainable in the lower limbs. Nerve biopsies reveal a severe decrease of myelinated fiber density. Many myelinated fibers show abnormal myelin thickenings and outfoldings, associated with numerous onion bulb formations.5,63,80,134 The function of myotubularin-related proteins in peripheral nerve is still largely unknown. The identification of a single gene involved in a dysfunction of both Schwann cells and cells of the trabecular meshwork of the eye is interesting because both types of cells are derived from the neural crest. Given the early onset of the symptoms, one could hypothesize that MTMR-13/ SBF-2 is involved both in the differentiation of Schwann cells during myelination and in the development of the trabeculum.7

Chromosome 11q23/MTMR2 Synonyms: CMT1 AR B1,125 CMT4B/11q23,149 CMT4B1,63 CMT4B [OMIM 601382]. In 1996, Bolino and co-workers localized a gene on chromosome 11q23 in a large Italian inbred family with an AR demyelinating neuropathy and focally folded myelin in nerve biopsy.21 The gene encoding myotubularin-related protein-2 (MTMR-2), a dual-specificity phosphatase, was identified as the causative gene in two Italian and two Saudi Arabian families.22,82 Next, mutations of this gene were also found in English, Indian, and Turkish families.65,108 In all patients the first symptoms appear during the second or third year of life. Patients show a markedly progressive neuropathy with prominent weakness beginning in the lower limbs, spreading proximally and later affecting the upper limbs. They show total areflexia and usually distal sensory impairment. Some patients developed mild scoliosis. In adulthood all patients are severely handicapped and wheelchair dependent. Additional features in some families are weakness of jaw, facial, and bulbar muscles; deafness; and diaphragmatic weakness. In the large Italian family, several patients had died by the end of the third or fourth decade, apparently as a result of respiratory failure.65,108,123,129 Median MNCVs in the reported families ranged from 9 to 20 m/s; in elderly patients motor responses were undetectable. Sensory nerve potentials were usually absent; a low-amplitude median nerve sensory action potential was recorded in a young patient. Widespread denervation is demonstrated in most patients. Auditory evoked potentials show abnormally increased I-to-III interpeak latencies.65,123,129 Nerve pathology displays a severe loss of myelinated fibers from an early age on. The remaining fibers are

Autosomal Recessive Hereditary Motor and Sensory Neuropathies

1775

Chromosome 18q23-qter/CTDP1 (Autosomal Recessive Neuropathy)

FIGURE 75–3 Sural nerve biopsy from a 4-year-old boy with chromosome 11q23/MTMR2 AR HMSN I. All the myelinated fibers have abnormally thin myelin, and many fibers have irregular myelin thickenings and abnormal profiles. Bar: 15 ␮m. See Color Plate

generally hypomyelinated but show focal regions with extensive foldings of the myelin sheath. Onion bulbs are small and occur infrequently (Fig. 75–3).65,108,123 The precise functions and substrate of MTMR-2 still need to be determined. Expression analysis of Mtmr2 in mice suggests particularly high levels in neurons. MTMR2 mutations lead to dramatically reduced phosphatase activity.17 How a lack of MTMR-2 results in demyelination and focal myelin folding is as yet uncertain.

Chromosome 17p11.2/PMP22 Most PMP22 mutations are inherited as an AD trait. At least one mutation is AR inherited. Parman and colleagues described a PMP22 mutation in the intracellular C-terminal domain of the protein that was silent in the heterozygous state of the parents but caused a DSN phenotype in the homozygous children.120 AR inheritance of a few other PMP22 mutations is disputed.110,113 Clinical, electrophysiologic, and pathologic features of this form of AR HMSN type I are discussed in Chapter 70.

Synonyms: CCFDN,3 CMT1 AR E125 [OMIM 604168]. A second form of AR demyelinating neuropathy among gypsy families from the northern part of Bulgaria was identified as part of a complex developmental disorder known as congenital cataracts, facial dysmorphism, and neuropathy (CCFDN). Therefore, it does not belong to HMSN in strict sense. The disease locus was assigned to the telomeric region of chromosome 18.3 The disease-causing mutation is a single nucleotide substitution in intron 6 of CTDP1, encoding the protein phosphatase FCP-1, an essential component of the eukaryotic transcription process. The mutation results in aberrant splicing and affects polymerase II–mediated transcription.150 The condition is recognized in infancy by the presence of congenital cataracts and microcorneas. The features of a predominantly motor neuropathy appear in early childhood, starting between 2 and 3 years in the lower limbs, later affecting the upper limbs, and leading to severe disability by the third decade of life. All sensory modalities are affected. Tendon reflexes are lost or depressed. Foot deformities, claw hands, and spinal deformities are common. Respiratory vital capacity is reduced in many patients secondary to kyphoscoliosis. Associated neurologic features include a mild, nonprogressive intellectual retardation in most affected individuals, strabismus, bilateral mild pyramidal signs, and in some mild chorea or ataxia. A few patients show bilateral sensorineural deafness. Non-neurologic features contribute to the characteristic phenotype: short stature, typical facial dysmorphism (prominent midface, thickening of perioral tissues, forwardly directed anterior dentition, hypognathism) (Fig. 75–4), and hypogonadotrophic hypogonadism. CSF protein is slightly elevated. Brain and spinal cord MRI scans demonstrate cerebral atrophy with enlargement of the lateral ventricles in nearly all patients and spinal cord atrophy in many patients.145 MNCVs are reduced into the demyelinating range and distal motor latencies are prolonged. MNCV slowing declines with age. Mean sensory action potentials are of normal amplitude, but conduction velocity is also markedly reduced.145 Nerve biopsies show normal myelinated fiber densities, but fibers are in general hypomyelinated. Occasionally, a demyelinated axon is seen (Fig. 75–5). In the older patients, infrequent small onion bulbs and few regenerative clusters are present. The unmyelinated fiber population appears to be normal. The morphologic changes in the peripheral nerve suggest a developmental defect of myelination followed by a superimposed degenerative process.146

1776

Diseases of the Peripheral Nervous System

A

B FIGURE 75–4 A, Typical facial appearance of CCFDN patient. See Color Plate B, Lateral skull radiograph of CCFDN patient.

Marinesco-Sjögren Syndrome with Myoglobinuria and CCFDN Syndrome Three unrelated gypsy families from Germany and Italy presented with classic symptoms of Marinesco-Sjögren syndrome (MSS; congenital cataract, cerebellar ataxia, developmental delay, mental retardation), and a demyelinating neuropathy associated with dramatic, often recurrent episodes of acute parainfectious muscle weakness with myoglobinuria. Linkage analysis of markers in the 18qter region demonstrated co-localization of the gene for this type of MSS with the CCFDN gene. Moreover, the patients with MSS shared the conserved marker haplotypes found in CCFDN chromosomes. This suggests that CCFDN and MSS with peripheral neuropathy and myoglobinuria are genetically identical and are caused by a single founder mutation.98,106

Chromosome 19q13/PRX

FIGURE 75–5 Sural nerve appearances in a 4-year-old boy with CCFDN. One fiber is demyelinated and several have rather thin myelin. Bar: 20 ␮m. See Color Plate

Synonyms: CMT4F,37 CMT1 AR F125 [OMIM 145900]. A disease locus on chromosome 19q13.1-q13.3 was reported in a large consanguineous family from Lebanon with an AR form of demyelinating HMSN.37 Shortly thereafter, nonsense and frameshift mutations were demonstrated in this family and in patients from several unrelated families of different ethnicity in the periaxin gene (PRX), a gene encoding two proteins, L- and S-periaxin.20,57 Patients with mutations in the PRX gene present with a delay in motor development.20,57,140 Areflexia and pes

Autosomal Recessive Hereditary Motor and Sensory Neuropathies

cavus are early signs; claw hand deformities and mild kyphoscoliosis occur in the second decade. Sensory involvement is more severe than is usually seen in CMT1 or DSN and includes deficits of position, vibration, pain, and temperature sensation. Sensory ataxia is common, and some patients complain of dysesthesia in both the upper and lower limbs. Patients show very low MNCVs (2 to 3 m/s) or a total absence of any sensory and motor evoked response. Sural nerve biopsies show a severe loss of myelinated fibers with many onion bulbs and fibers with hypermyelinated outfoldings, associated with axonal compression (Fig. 75–6). L-periaxin was undetectable by immunohistochemistry in the affected nerve using an antibody against L-periaxin (anti–C-terminal antibody).57,140 L- and S-periaxin are required for the maintenance of peripheral nerve myelin. As a cytoskeleton-associated protein, L-periaxin may facilitate integration of extracellular signaling through the cytoskeleton, essential for changes in Schwann cell shape and regulation of gene expression during axonal ensheathment. Prx-deficient mice show initially normal myelination of peripheral axons, but subsequently develop a severe demyelinating neuropathy with excess folded myelin in peripheral nerves associated with allodynia and hyperalgesia, supporting the view that the mouse mutant is a good model for the human disease.53

FIGURE 75–6 Sural nerve pathology of patient with periaxin gene mutation showing onion bulb of very thin Schwann cell processes around myelinated fiber. Many Schwann cells are denervated. Bar: 3 ␮m. (Courtesy of Dr. A. Mégarbané, Lebanon.)

1777

Autosomal Recessive HMSN I with Unknown Gene Loci Several AR forms of demyelinating HMSN have been described in which the genetic defect is still unknown. Some families show distinct clinical or morphologic features, whereas in others the phenotype is nonspecific.61 These subtypes await identification of the genetic defects that, in all probability, will lie ahead with the ongoing elucidation of the human genome. Autosomal Recessive Congenital Hypomyelination Neuropathy A number of authors have reported CHN in families with probably AR inheritance.48,68,75,76,89,119,131,147,153 The phenotypic features in these patients are largely the same as in cases of CHN caused by AD mutations of the PMP22 and P0 genes or the homozygous mutation in the EGR2 gene (see Chapters 70, 71, and 72). Patients are hypotonic in the neonatal period or they show a marked delay in early motor development. Unsupported walking is not achieved or is markedly delayed, and many patients become wheelchair dependent. CSF protein is usually increased. Conduction velocities are unrecordable or extremely slowed. Nerve biopsies show extremely thin or no myelin sheaths, and onion bulbs, mainly composed of paired basal laminae, surround nearly all fibers (Fig. 75–7). The final clinical outcome may vary considerably among patients, even within the same sibship. Gabreëls-Festen described a brother and sister with equal MNCV slowing (4 to 5 m/s) and similar nerve pathology, both consistent with CHN. However, the girl became wheelchair dependent at the age of 12 years, whereas the brother was still walking quite well at the age of 40 years.46 Some authors consider CHN a nonprogressive disorder solely on the basis of a defect in myelin formation.60,107,121 However, others have observed myelin breakdown products in CHN nerves or could demonstrate clear signs of demyelination and remyelination in teased fibers.46,68,89,90,103,148,153 Investigation of the DNA of three patients with AR congenital hypomyelination has excluded mutations in the PMP22, P0, and EGR2 genes.46 This implies that other as yet unknown gene(s) may be responsible for AR CHN. Autosomal Recessive HMSN I with Myelin Folding (CMT4B3) Ben Othmane and co-workers excluded the loci for CMT4B1 and CMT4B2 in two Tunisian families with demyelinating HMSN and focally folded myelin sheaths in nerve biopsy, suggesting the presence of at least a third locus for this morphologic phenotype.14

1778

Diseases of the Peripheral Nervous System

A

B FIGURE 75–7 A, Sural nerve pathology of a 3-year-old patient with AR CHN of unknown genetic defect. Most fibers with diameters in the myelinated range are without myelin or with a very thin myelin sheath. Multiple single and duplicated basal membranes surround these fibers. There is a marked increase of endoneurial collagen. Bar: 5 ␮m. B, Onion bulb comprising single and double basal membranes in AR CHN. Bar: 2 ␮m.

AUTOSOMAL RECESSIVE HMSN TYPE II AR HMSN cases in which the underlying pathology is an axonopathy or neuronopathy have been categorized as HMSN type II or CMT2. Harding and Thomas were the first to document AR inheritance in two families with HMSN type II. Age of onset was in the first decade and clinical involvement was more severe, with a greater severity of weakness, ataxia, tendon areflexia, and scoliosis than in AD type II cases.61 Later studies described series of AR severe axonal neuropathy with early childhood onset51,117 and cases of AR congenital axonal motor and sensory neuropathy with respiratory distress.102,151 Recently, additional families with AR axonal HMSN have become known, especially from communities with a high consanguinity rate, and the first genetic loci and genes have been identified.10,26,86,126,135

Chromosome 1q21.2/LMNA Synonyms: AR-CMT2A,26 CMT2 AR A,125 CMT3A / 1q21.3,149 CMT2B136 [OMIM 605588]. Bouhouche and co-workers described a large consanguineous Moroccan family with nine individuals affected by an axonal type of CMT disease. Markers on

chromosome 1q defined a region of homozygosity to an interval of 1.7 cM on chromosome 1q21.2-q21.3.26 In two Algerian families with the same phenotype and an identical disease locus, a unique homozygous mutation in LMNA, encoding lamin A /C, a component of the nuclear envelope, was identified in all affected members and in a third, unrelated family. 36 The mutation, Arg298Cys, was localized in the lamin A /C rod domain, thus affecting all four known isoforms of LMNA. Sitespecific amino acid substitutions of LMNA cause AD limb girdle muscular dystrophy with atrioventricular conduction disturbances (LGMD1B),104 AD or AR Emery-Dreifuss muscular dystrophy,25,124 AD dilated cardiomyopathy with atrioventricular block (CMD1A),43 AD partial lipodystrophy,137 and finally also AR CMT2. This suggests the existence of distinct functional domains in lamin A /C. The Moroccan and Algerian families show comparable clinical features. Age of onset is in the second decade (10 to 18 years). The first manifestations are weakness and wasting of distal limb muscles and tendon areflexia predominantly of lower limbs. Proximal limb muscles and limb girdle muscles become involved in many patients. Sensory examination is not mentioned. Pes cavus is a common feature; some patients show a severe kyphoscoliosis.

Autosomal Recessive Hereditary Motor and Sensory Neuropathies

Progression of the neuropathy is severe; many individuals lose the ability to walk early and become wheelchair dependent in the second decade.26,32 De Sandre-Giovannoli and colleagues mentioned explicitly that no cardiac conduction disturbances or dilatation was observed in the Algerian patients examined between the ages of 12 and 32 years.36 MNCV is normal or slightly reduced in all patients. Sensory nerve action potentials are decreased or absent. Nerve biopsy reveals a marked loss of large myelinated fibers and few26 or no regenerative clusters.32

Chromosome 5q/SMN1 (Autosomal Recessive Axonal Neuropathy) Synonyms: SMA type 0,40 severe congenital SMA.156 Recently, several authors described children with a severe congenital sensorimotor, axonal neuropathy who showed homozygous deletions/conversions of the SMN1 gene, localized on chromosome 5q. These children present in the neonatal period with asphyxia and generalized weakness, sometimes associated with arthrogryposis or facial weakness and external ophthalmoplegia. Motor and sensory nerves are inexcitable on electrophysiologic testing. The children usually die after withdrawal of ventilatory support.64,81,84,91,115 Sural nerve biopsies and postmortem examination show loss of myelinated fibers and evidence of axonal degeneration in sensory and mixed nerves. Many spinal motor neurons are chromatolytic, although their numbers are normal.81 These findings provide evidence for an unusually severe type of spinal muscular atrophy (SMA) associated with a congenital motor and sensory axonal neuropathy, probably related to an extended deletion in the region of the SMN1 gene.81 Dubowitz proposed designating this form of SMA type 0.40

Chromosome 8q21.1/GDAP1 Synonyms: CMT4A,13 CMT1 AR A,125 CMT4A/8q13,149 CMT4A163 [OMIM 214400]. The first locus in an AR form of HMSN was identified in 1993 by Ben Othmane and co-workers to chromosome 8q13-q21.1 in four Tunisian consanguineous families with a severe, supposedly demyelinating neuropathy.13 Later, two other AR HMSN variants were assigned to this disease locus, namely AR axonal CMT with hoarseness and vocal cord paresis (CMT4K [MIM 607831]) in three Spanish families,135 and AR axonal CMT with mild pyramidal involvement (CMT4H) in a Tunisian family.10 Mutations in the GDAP1 gene were found simultaneously in the Tunisian families (CMT4A)11 and in the Spanish families with hoarseness (CMT4K).34 Subsequently, additional GDAP1 mutations were reported in several European and North African families without 1,19,109,132 or with hoarseness.8

1779

Clinical presentation of the neuropathy is similar in the two variants. Onset of the disorder is by the end of the first year of life, often interfering with early motor development. Weakness and wasting of the feet is soon followed by involvement of the hands, leading to marked disability toward the end of the first decade. Most patients become wheelchair dependent in the second or third decade of life. Sensory involvement is mild, affecting all modalities of sensation. Tendon reflexes are usually absent in lower limbs and absent or diminished in upper limbs. Pes cavus, claw hands, and spinal deformities are common. CSF protein level is normal. Median MNCV is unobtainable in most adult patients. In several patients, mainly of younger age, upper limb MNCVs can still be registered, although with very small CMAP amplitudes. In some patients MNCVs are even reduced to values in the demyelinating range (⬍38 m/s), but in those cases CMAPs are extremely reduced (ⱕ0.5 mV). Sensory conduction velocity is likewise reduced, with a less severe decrease in action potential amplitudes.1,12,19,63,109,132 In the early stages of the disease (⬍3 years of age), nerve biopsies show a reduction in the large-diameter fiber population, and some fibers with thin myelin sheaths. In later stages large fiber loss has increased, and thinly remyelinated fibers, numerous clusters, and onion bulb–like structures are observed.12,19,109,132 Ben Othmane and co-workers13 considered the condition to be a demyelinating disorder based on nerve conduction slowing and pathology, although, according to the authors of a second paper, “an axonal defect with secondary demyelination cannot be ruled out.”15 Indeed, the minimal motor amplitudes (⬍0.5 mV) point to a severe loss of large-diameter fibers and make nerve conduction values unreliable.45 Other authors interpreted the neuropathy as axonal19 or “intermediate”132 in nature. Nerve pathology records report onion bulbs or onion bulb–like structures (pseudo–onion bulbs), likely depending on whether the pathogenetic process is considered primarily demyelinating or axonal in nature. This indistinctness might explain why patients with the same homozygous mutation (Ser194stop) are sometimes classified as having a demyelinating neuropathy11 and sometimes as having an axonal neuropathy.19,109 Based on the observation in a 3-year-old patient showing normal or mildly reduced MNCVs but severe reduction of CMAPs, Nelis and colleagues proposed that a severe loss of axons is the initial pathologic process, followed by or concomitant with demyelination.109 However, it cannot be excluded that specific GDAP1 mutations lead to a predominantly demyelinating neuropathy. Cuesta and co-workers demonstrated that GDAP1 transcripts are ubiquitously expressed in several human and mouse tissues with an apparently predominant expression in nervous tissue. GDAP1 expression also occurs in Schwann cells, but expression is higher in central nervous

1780

Diseases of the Peripheral Nervous System

tissue than in peripheral nerve.34 Little is known about the function of GDAP-1. The protein may be involved in a signal transduction pathway during neuronal development. In addition, the amino acid sequence of GDAP-1 shows strong similarity to those of glutathione S-transferases, enzymes that have a role in the detoxification of cells.34 Chromosome 8q21.1/GDAP1, with Hoarseness Synonym: CMT2K [OMIM 607831]. In some families with GDAP1 mutations, early-onset neuropathy is associated with hoarseness. In the second or third decade, patients complain of voice hoarsening. Laryngoscopic examination in two patients confirmed vocal cord paralysis, and in one patient chest radiography showed paralysis of one hemidiaphragm, resulting in restrictive respiratory insufficiency. Other cranial nerves were not involved. CMAPs from the diaphragm on stimulation of the phrenic nerves were reduced, but distal latencies in phrenic and axillary nerves were in the normal range.8,34,136 Sural nerve biopsy of two patients with GDAP1 mutations and hoarseness showed a pronounced loss of myelinated fibers, regenerative clusters, and signs of axonal degeneration. No demyelinated fibers and few concentric Schwann cell structures were seen.136 Electrophysiologic and morphologic features are in line with an axonal neuropathy. Further reports of families with GDAP1 mutations will be needed to determine if the hoarse voice results from particular mutations. Chromosome 8q21.1, with Pyramidal Involvement A third phenotype linked to chromosome 8q21 was described in a large Tunisian family with an AR form of axonal HMSN and mild pyramidal tract involvement, classified as CMT4C210 or CMT2H. Symptoms usually start between the ages of 4 and 8 years with difficulty in walking. Muscle weakness and wasting first appear in distal lower limbs, followed by involvement of distal muscles in the upper limbs. In patients over 40 years of age, severe distal atrophy is observed. Some patients become wheelchair dependent. Distal sensory loss is present in lower limbs. Mild pyramidal features are present from the early disease stages and remain rather stationary during the course of the disease. Pyramidal involvement is manifested by exaggeration of knee reflexes and of tendon reflexes in upper limbs. Ankle jerks are absent. There are no bulbar or pseudobulbar signs. Median MNCV is mildly reduced (40 m/s), and sensory conduction is comparably reduced with decreased action potential amplitudes. Nerve biopsy of two adult patients showed a marked loss of largediameter fibers and few signs of axonal regeneration.10 The disease locus in this family overlaps the GDAP1 locus.10 Whether this disorder is also a phenotypic expression of mutations in the GDAP1 gene awaits further investigation.

Chromosome 10q23.2 Synonyms: HMSN-Russe, HMSNR,126 CMT3B/10q23.2149 [OMIM 605285]. A novel form of AR HMSN was identified in large kindred of Bulgarian gypsies, and named HMSN-Russe after the city in northern Bulgaria where most of the affected individuals resided. Subsequently, similar features were identified in Romanian and Spanish gypsy families. The disorder was mapped to chromosome 10q23.2, and EGR2 has been excluded as the responsible gene.126,143 Distal lower limb weakness is noticed between the ages of 8 and 16 years, followed by upper limb involvement between the ages of 10 and 43 years. The disorder progresses steadily with age, leading to severe disability and total paralysis of lower limb muscles and also of small hand muscles and forearm extensor and flexor muscles by the fourth or fifth decade of life. Sensory loss affecting all modalities is a prominent feature, in two cases resulting in neuropathic joint degeneration. Tendon reflexes are absent. Foot and hand deformities are common, but spinal deformity is not observed. Minor cranial nerve involvement (ptosis, facial weakness, dysphonia) occurs in a few patients.126,143 MNCVs are unobtainable in the legs. In the upper limbs, motor nerve conduction is moderately reduced (32 m/s ⫾ 7 m/s) in combination with a markedly increased threshold for electrical nerve stimulation. Sensory action potentials are absent.143 Sural nerve biopsy shows a reduction of large-diameter fibers without active fiber degeneration. Numerous clusters of thinly myelinated regenerating fibers are present (Fig. 75–8). The morphology of unmyelinated fibers is normal.143

Chromosome 11q13-q21/IGHMBP2 Synonyms: SIANRF,155 distal hereditary motor neuronopathy type VI (dHMN VI),18 SMARD1 (diaphragmatic spinal muscular atrophy),55 severe infantile neuropathy with diaphragmatic weakness122 [OMIM 604320]. Several reports have described neonatal respiratory complications in cases defined as either diaphragmatic SMA, also known as spinal muscular atrophy with respiratory distress type 1 (SMARD1)18,95,96,97,112,138,156 or severe infantile axonal neuropathy with respiratory failure (SIANRF).4,52,151,155 A case reported by Hahn and colleagues was erroneously labeled as CHN, but electrophysiologic and nerve pathologic changes are clearly consistent with a sensorimotor neuropathy of axonal type.59 Cases occur sporadically or inheritance is consistent with an AR trait. SMARD1 shows a noncongenital onset of respiratory distress.18 Grohmann and colleagues demonstrated that SMARD1 is linked to chromosome 11q13-q21 and is caused by mutations in the gene encoding immunoglobulin ␮–binding protein-2 (IGHMBP2).54,55 One of the

Autosomal Recessive Hereditary Motor and Sensory Neuropathies

FIGURE 75–8 Nerve pathology of a 31-year-old man with HMSN-Russe. All the myelin sheaths are inappropriately thin for axon diameter, and many are in regenerative clusters. Bar: 20 ␮m. See Color Plate

patients of Wilmshurst and co-workers (case 3) with SIANRF155 was included in this publication.54 Pitt and colleagues described additional cases of severe infantile neuropathy showing IGHMBP2 mutations and presented a set of diagnostic criteria.122 The twin patients described by Mohan and colleagues,102 who showed marked abnormalities of autonomic function, were later found to have an IGHMBP2 mutation (R. H. M. King, personal communication). The IGHMBP-2 protein probably shares common functions with the survival motor neuron protein-1, important for motor neuron maintenance. Pregnancy, fetal movements, and birth are usually normal, but children often have a low birth weight. Soon after birth (1 to 12 weeks), the infants show respiratory failure resulting from diaphragmatic paralysis either unilaterally or bilaterally. They become ventilator dependent within less than 1 month of onset, with an inability to wean. Subsequently, paralysis and atrophy in distal limb musculature is observed, which spreads to involve proximal muscles. Death usually occurs before the age of 1 year, although the twins described by Mohan and co-workers were still alive on a respirator at this time.54,102,122 Neurophysiologic examination reveals markedly reduced MNCVs particularly in the legs, no or very small evoked

1781

CMAPs, and similar but less severe abnormalities of the sensory fibers, supporting a severe axonal motor and sensory neuropathy. Electromyographic examination often shows active denervation of the diaphragm and distal muscles. In the report of Grohmann and colleagues,54 the state of the peripheral sensory system was not mentioned. However, involvement of motor and sensory nerves in cases of IGHMBP2 mutations was clearly demonstrated by Pitt and colleagues122 and was previously described in case 3 of Wilmshurst and colleagues.155 Sural nerve biopsy shows only small myelinated fibers with, in general, no signs of regeneration or demyelination, and sparse evidence of active degeneration (Fig. 75–9). Electron microscopy may show recent axonal degeneration and bands of Büngner, indicating previous degeneration of myelinated axons. Unmyelinated fibers are apparently not involved. Postmortem examination of the spinal cord in two patients showed no conspicuous loss of or changes in the anterior horn cells or dorsal root ganglia.122 Pathologic features suggest a progressive loss of motor and sensory fibers, probably superimposed on faulty development. Identical histopathologic features of the sural nerve were found in some other children with early respiratory distress, who had tested negative for mutations in the IGHMBP2 gene. However, in these children the diaphragm was not involved and none showed the severe slowing of motor nerve conduction.122 Further studies are needed to further define the phenotype of IGHMBP2 mutations and to see whether this phenotype is genetically heterogeneous.

Chromosome 15q13-q15/SLC12A6 Synonyms: ACCPN, Andermann syndrome [OMIM 218000]. Agenesis of the corpus callosum with peripheral neuropathy (ACCPN), or Andermann syndrome, is an AR disorder characterized by agenesis of the corpus callosum, progressive peripheral sensorimotor neuronopathy, mental retardation, and dysmorphic features.2,85 It occurs preferentially in the French Canadian population of the Charlevoix-Saguenay region in the province of Quebec. Only a few cases outside French Canada have been reported.38,62,88 Genealogic studies suggest that the disorder in French Canada likely results from a single founder mutation. By genomic mapping, the ACCPN gene was localized within a 5-cM interval on chromosome 15q13-q15.30 AR spastic paraplegia type 11 maps to the same chromosomal region.93 In families with ACCPN, Howard and colleagues found mutations in the solute carrier family 12, member 6 (SLC12A6) gene, which encodes the K⫹-Cl⫺ cotransporter KCC3. Nearly all French Canadian patients were homozygous for the same frameshift mutation. In a Turkish and an Italian family, two distinct homozygous nonsense mutations in SLC12A6 were identified. The described mutations are predicted to result in nonfunctional proteins.66

1782

% fibers

Diseases of the Peripheral Nervous System

fiber diameter (im) B

A

FIGURE 75–9 A, Sural nerve pathology of a 15-month-old patient with severe infantile axonal neuropathy and respiratory failure. There is a lack of large-diameter fibers and few fibers are undergoing axonal degeneration. Bar: 4 ␮m. B, Myelinated fiber size histogram of sural nerve of the same patient (white area) in comparison with age-matched control (dotted area).

At birth, patients may present with minor dysmorphic features (elongated facies, hypertelorism, high-arched palate, hypoplastic maxilla, large angle of the mandible). Patients are hypotonic in infancy and show a delay in motor and mental development. Unsupported walking is markedly delayed or not acquired, followed by a loss of motor functions, starting in the legs with distal atrophy and pes cavus, and claw hands. Tendon reflexes are absent. Slight bilateral ptosis, convergent strabismus, mild facial diplegia, and a progressive scoliosis are often noted. Patients become wheelchair dependent in the second decade. Most patients show a mild to moderate mental retardation. Atypical psychosis is observed in many patients after 15 years of age. Death usually follows in the third or fourth decade from complications stemming from respiratory infections. Total or partial agenesis of the corpus callosum is an inconstant feature; in about one third of the patients the corpus callosum is present. Other developmental or degenerative midline anomalies occur in a high frequency.41,94 Neurophysiologic examination shows a reduction of motor and sensory nerve conduction velocities with markedly reduced amplitudes and absence of sensory action potentials. Somatosensory and visual evoked potentials are abnormal. Nerve biopsy reveals an almost total loss of large-diameter fibers, usually without clear signs of regeneration, and a sporadic wallerian degeneration.41 Slight onion bulb formation in some patients is interpreted as secondary demyelination and remyelination.85

It is not known how loss of function of the K⫹-Cl⫺ cotransporter KCC3 may lead to the complex phenotype of ACCPN, but it is evident that KCC3 is involved in the development and maintenance of both the peripheral and central nervous systems. Mice with a targeted deletion of Slc12a6 show a peripheral neuropathy, similar to the human disease, but a normal corpus callosum. Along with the variable presence of the corpus callosum in ACCPN patients, this suggests that a defect in SLC12A6 alone is not sufficient to produce agenesis of the corpus callosum, and that additional epigenetic factors are involved.66

Chromosome 16q24/Gigaxonin Synonym: Giant axonal neuropathy [OMIM 256850]. Giant axonal neuropathy (GAN) is a chronic progressive systemic disorder mainly affecting the peripheral and central nervous systems. In 1998, the gene responsible for GAN was localized to chromosome 16q24 in five consanguineous families,44 and 2 years later the gene was cloned.23 Subsequent reports confirmed that GAN is caused by mutations in the gigaxonin gene.24,83 The protein product of the gene, gigaxonin, is a novel protein of the cytoskeleton that directly or indirectly may have an important role in neurofilament architecture.23 Possibly, GAN is genetically heterogeneous. An Algerian family with three affected sibs showed absence of linkage to chromosome 16q24.1.141 The first descriptions of this AR disorder date from the early 1970s.6,16,29,118 Onset is usually in infancy with a

Autosomal Recessive Hereditary Motor and Sensory Neuropathies

delay in motor development, or the first symptoms appear just as the child has learned to walk independently. Although sensory ataxia may be a serious complaint in some patients with HMSN, in GAN it becomes the most important clinical feature. Gait becomes broad based, clumsy, and unsteady. Progressive distal weakness and wasting develop, starting in the lower limbs and subsequently spreading to the upper limbs. Slurred speech, nystagmus, intention tremor, dysmetria, and truncal ataxia gradually appear. Distal sensory functions are impaired. Facial weakness and involvement of other cranial nerves may occur. Tendon reflexes are diminished or absent. Often the plantar responses are extensor, indicating pyramidal tract involvement. The children are mildly mentally retarded. A striking abnormality, although not seen in all patients, is the presence of tightly curled hair. CSF protein is normal. Initially a polyneuropathy is the predominant clinical feature; after a few years symptoms of a spinocerebellar degeneration become dominant. In some cases epileptic seizures occur. Many patients show severe scoliosis with paraplegia, finally becoming wheelchair dependent. Death usually occurs in adolescence or in the third decade.116 MNCVs are normal or mildly decreased and sensory action potentials are usually absent, suggestive of an axonal neuropathy. Electroencephalography, brainstem evoked responses, somatosensory evoked responses, and visual evoked responses are often abnormal, indicating central nervous system involvement. Computerized tomography scanning or MRI may reveal widespread white matter changes.116 Nerve biopsies show giant axonal swellings measuring up to 50 ␮m, with densely packed neurofilaments, often with few or no neurotubules (Fig. 75–10). Giant axons show no or very thin myelin sheaths, and onion bulbs are present around some fibers.118 GAN is a generalized disorder, and accumulations of intermediate filaments are found also in other cell populations. Microscopic examination of the hair is normal, but scanning electron microscopy may show typical longitudinal grooves.

Chromosome 19q13.3 Synonyms: CMT3A /19q13.3,149 AR CMT2B [OMIM 605589]. In a large consanguineous Costa Rican family with 18 individuals affected by an AR form of HMSN type II, linkage to chromosome 19q13.3 was found.86 Age at onset of the disorder is 28 to 42 years. All patients present with symmetrical weakness of distal lower limbs muscles. Most patients show slight to severe involvement of small hand muscles. All patients have a symmetrical stocking-glove pattern of sensory loss. Ankle areflexia is common; some patients show total areflexia. MNCVs in the upper limbs are normal or slightly reduced; in most

1783

FIGURE 75–10 Sural nerve biopsy from a 7-year-old boy with giant axonal neuropathy. Many axons have unusually dense cytoplasm and some have undergone secondary demyelination. Onion bulbs have formed around two of these. Occasional arrays of concentrically arranged Schwann cell processes with no central fiber suggest axonal loss. Bar: 20 ␮m. See Color Plate

patients sensory conduction and MNCVs in the lower limbs cannot be recorded. Sural nerve biopsy has not been performed.86

Autosomal Recessive HMSN II with Unknown Gene Loci Several conditions with AR axonal motor and sensory neuropathy have been characterized for which the genetic defect or locus is still unknown. It is conceivable that these AR HMSN II forms are genetically heterogeneous. Some earlier published cases appear to be related to recently described genetic defects. Autosomal Recessive Early Childhood HMSN II Two authors have described a group of patients with an earlyonset severe axonal motor and sensory neuropathy.51,117 AR inheritance is suggested in six families by the presence of affected siblings born to healthy parents. A few earlier case reports have described similar patients.27,39,74

1784

Diseases of the Peripheral Nervous System

ation disturbance of peripheral motor and sensory neurons. In later stages a process of chronic axonal degeneration follows, concomitant with progressive clinical deterioration.

Onset of clinical symptoms is in the first years of life. A few patients already have limp feet at birth. Most children show a delay in motor development or have a clumsy gait with frequent falling from the beginning. Weakness and atrophy of distal lower limbs muscles are early features. Distal upper limbs become involved by the first decade with claw hand deformity, followed by weakness of proximal lower limb muscles at the beginning of the second decade, when most patients are no longer ambulatory. Many patients show mild sensory dysfunction in the lower limbs and slight sensory ataxia. Ankle jerks are usually absent, but total areflexia is uncommon. Several patients show pes cavus, and some develop scoliosis. CSF protein, whenever examined, is normal.51,117 Electrophysiologic investigations demonstrate mild to moderate slowing of motor nerve conduction, sometimes even below 30 m/s. Sensory action potentials are absent or reduced in amplitude. Electromyography shows spontaneous fibrillation and positive spikes in many patients, suggestive of denervation.51,117 Sural nerve biopsies show an extensive loss of largediameter fibers (Fig. 75–11A) and a shift to smaller diameters of the myelinated fiber size histogram (Fig. 75–11B). Regenerative clusters are almost absent, in contrast to most AD forms of HMSN type II. Insidious axonal degeneration and bands of Büngner occur in later stages of the disease.51 The lack of large-diameter fibers without overt signs of degeneration and regeneration, seen in biopsies of very young patients, suggests a primary matur-

% fibers

Autosomal Recessive HMSN II with Deafness, Mental Retardation, and Absence of Large Myelinated Fibers A distinct clinicopathologic form of HMSN type II with a congenital or early infantile onset has been described in three families from Southern Italy.92,128,130 The children, all boys, were born to healthy parents. The parents in one family were consanguineous, suggesting AR inheritance, although X-linked inheritance could not be ruled out. However, the recent recognition of a girl with the same phenotype favors AR inheritance.105 Patients show a slowly progressive motor and sensory neuropathy with an early infantile onset, congenital sensorineural deafness, and mental retardation (IQ 50 to 60). Electrophysiologic studies demonstrate a marked reduction of motor and sensory conduction velocities and absence of sensory action potentials. Nerve biopsies display normal fiber density but total absence of large myelinated fibers and signs of axonal atrophy, possibly resulting from a developmental disturbance of large motor and sensory neurons. A girl with the same phenotype showed at autopsy complete absence of large myelinated fibers in the peripheral nervous system, corresponding to a lack of large neurons in dorsal root ganglia and anterior horns of the spinal

fiber diameter (im) B

A

FIGURE 75–11 A, Sural nerve pathology of a 4-year-old patient with early-onset AR HMSN II. There is a lack of large-diameter fibers without signs of axonal degeneration. Bar: 5 ␮m. B, Myelinated fiber size histogram of sural nerve of the same patient (white area) in comparison with age-matched control (dotted area).

Autosomal Recessive Hereditary Motor and Sensory Neuropathies

cord.105 This observation is in favor of the presumed developmental disturbance of large motor and sensory neurons. Autosomal Recessive HMSN II with Acrodystrophy Thomas and colleagues described a consanguineous Turkish family with three affected siblings and healthy parents.142 Early development was normal, but during childhood the patients developed a progressive neuropathy with distal upper and lower limb weakness and wasting associated with severe distal sensory loss in both upper and lower limbs, leading to prominent mutilating changes. All sensory modalities were affected, but there were no signs of autonomic dysfunction. The mutilating acropathy led to limb amputation in all three patients. Two patients died from renal amyloidosis, secondary to chronic osteomyelitis. Electromyography showed complete denervation in distal upper and lower limbs and no motor and sensory response was obtained, except in one subject who showed normal median MNCV. Sural nerve biopsy showed no myelinated fibers or a grossly reduced density and a marked decrease in unmyelinated fiber density. Groups of Schwann cells appeared as either bands of Büngner or denervated stacks of unmyelinated Schwann cells. The severity of the motor involvement suggests a separate form of HMSN rather than hereditary sensory neuropathy.142

REFERENCES 1. Ammar, N., Nelis, E., Merlini, L., et al.: Identification of novel GDAP1 mutations causing autosomal recessive Charcot-Marie-Tooth disease. Neuromuscul. Disord. 13:720, 2003. 2. Andermann, E., Andermann, F., and Melançon, D.: Three familial midline malformation syndromes of the central nervous system. In de Grouchy, J. (ed.): Proceedings of the Fourth International Congress of Human Genetics (Excerpta Medica International Congress Series No.233). Amsterdam, Excerpta Medica, p. 15, 1971. 3. Angelicheva, D., Turnev, I., Dye, D., et al.: Congenital cataracts facial dysmorphism neuropathy (CCFDN) syndrome: a novel developmental disorder in Gypsies maps to 18qter. Eur. J. Hum. Genet. 7:560, 1999. 4. Appleton, R., Riordan, A., Tedman, B., et al.: Congenital peripheral neuropathy presenting as apnoea and respiratory insufficiency. Dev. Med. Child Neurol. 36:547, 1994. 5. Arruda, W. O., Comerlato, E. A., Scola, R. H., et al.: Hereditary motor and sensory neuropathy with congenital glaucoma: report on a family. Arq. Neuropsiquiatr. 57:190, 1999. 6. Asbury, A. K., Gale, M. K., Cox, S. C., et al.: Giant axonal neuropathy—a unique case with segmental neurofilamentous masses. Acta Neuropathol. (Berl.) 20:237, 1972. 7. Azzedine, H., Bolino, A., Taïeb, T., et al.: Mutations in MTMR13, a new pseudophosphatase homologue of

8.

9.

10.

11.

12.

13.

14.

15.

16. 17.

18.

19.

20.

21.

22.

1785

MTMR2 and Sbf1, in two families with an autosomal recessive demyelinating form of Charcot-Marie-Tooth disease associated with early-onset glaucoma. Am. J. Hum. Genet. 72:1141, 2003. Azzedine, H., Ruberg, M., Ente, D., et al.: Variability in disease progression in a family with autosomal recessive CMT associated with a S194X and new R310Q mutation in the GDAP1 gene. Neuromuscul. Disord. 13:341, 2003. Baethmann, M., Göhlich-Ratmann G., Schröder, J. M., et al.: HMSNL in a 13-year-old Bulgarian girl. Neuromuscul. Disord. 8:90, 1998. Barhoumi, C., Amouri, R., Ben Hamida, C., et al.: Linkage of a new locus for autosomal recessive axonal form of Charcot-Marie-Tooth disease to chromosome 8q21.3. Neuromuscul. Disord. 11:27, 2001. Baxter, R. V., Ben Othmane, K., Rochelle, J. M., et al.: Ganglioside-induced differentiation-associated protein-1 is mutant in Charcot-Marie-Tooth disease type 4A /8q21. Nat. Genet. 30:21, 2002. Ben Othmane, K.: Formes pures de la maladie de CharcotMarie-Tooth (HMSN) en Tunesie: étude clinique, généalogique, électrophysiologique et histologique de 82 familles. M.D. Thesis, Faculté de Médicine de Tunis, Tunisia, 1990. Ben Othmane, K., Hentati, F., Lennon, F., et al.: Linkage of a locus (CMT4A) for autosomal recessive Charcot-MarieTooth disease to chromosome 8q. Hum. Mol. Genet. 2:1625, 1993. Ben Othmane, K., Johnson, E., Menold, M., et al.: Identification of a new locus for autosomal recessive Charcot-Marie-Tooth disease with focally folded myelin on chromosome 11p15. Genomics 62:344, 1999. Ben Othmane, K., Loeb, D., Hayworth-Hodgte, R., et al.: Physical and genetic mapping of the CMT4A locus and exclusion of PMP-2 as the defect in CMT4A. Genomics 28:286, 1995. Berg, B. O., Rosenburg, S. H., and Asbury, A. K.: Giant axonal neuropathy. Pediatrics 49:894, 1972. Berger, P., Bonneick, S., Willi, S., et al.: Loss of phosphatase activity in myotubularin-related protein 2 is associated with Charcot-Marie-Tooth disease type 4B1. Hum. Mol. Genet. 11:1569, 2002. Bertini, E., Gadisseux, J. L., Palmieri, G., et al.: Distal infantile spinal muscular atrophy associated with paralysis of the diaphragm: a variant of infantile spinal muscular atrophy. Am. J. Med. Genet. 33:328, 1989. Birouk, N., Azzedine, H., Dubourg, O., et al.: Phenotypical features of a Moroccan family with autosomal recessive Charcot-Marie-Tooth disease associated with the S194X mutation in the GDAP1 gene. Arch. Neurol. 60:598, 2003. Boerkoel, C. F., Takashima, H., Stankiewicz, P., et al.: Periaxin mutations cause recessive Dejerine-Sottas neuropathy. Am. J. Hum. Genet. 68:325, 2001. Bolino, A., Brancolini, V., Bono, F., et al.: Localization of a gene responsible for autosomal recessive demyelinating neuropathy with focally folded myelin sheaths to chromosome 11q23 by homozygosity mapping and haplotype sharing. Hum. Mol. Genet. 5:1051, 1996. Bolino, A., Muglia, M., Conforti, F. L., et al.: CharcotMarie-Tooth type 4B is caused by mutations in the gene

1786

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

Diseases of the Peripheral Nervous System

encoding myotubularin-related protein-2. Nat. Genet. 25:17, 2000. Bomont, P., Cavalier, L., Blondeau, F., et al.: The gene encoding gigaxonin, a new member of the cytoskeletal BTB/kelch repeat family, is mutated in giant axonal neuropathy. Nat. Genet. 26:370, 2000. Bomont, P., Ioos, C., Yalcinkaya, C., et al.: Identification of seven novel mutations in the GAN gene. Hum. Mutat. 21:446, 2003. Bonne, G., Raffaele di Barletta, M., Varnous, S., et al.: Mutations in the gene encoding lamin A/C cause autosomal dominant Emery-Dreifuss muscular dystrophy. Nat. Genet. 21:285, 1999. Bouhouche, A., Benomar, A., Birouk, N., et al.: A locus for an axonal form of autosomal recessive Charcot-MarieTooth disease maps to chromosome 1q21.2-q21.3. Am. J. Hum. Genet. 65:722, 1999. Brust, J. C., Lovelace, R. E., and Devi, S.: Clinical and electrodiagnostic features of Charcot-Marie-Tooth syndrome. Acta Neurol. Scand. Suppl. 68:1, 1978. Butinar, D., Zidar, J., Leonardis, L., et al.: Hereditary auditory, vestibular, motor, and sensory neuropathy in a Slovenian Roma (Gypsy) kindred. Ann. Neurol. 46:36, 1999. Carpenter, S., Karpati, G., Andermann, F., et al.: Giant axonal neuropathy: a clinically and morphologically distinct neurological disease. Arch. Neurol. 31:312, 1974. Casaubon, L. K., Melanson, M., Lopes-Cendes, I., et al.: The gene responsible for a severe form of peripheral neuropathy and agenesis of the corpus callosum maps to chromosome 15q. Am. J. Hum. Genet. 58:28, 1996. Chandler, D., Angelicheva, D., Heather, L., et al.: Hereditary motor and sensory neuropathy-Lom (HMSNL): refined genetic mapping in Romani (Gypsy) families from several European countries. Neuromuscul. Disord. 10:584, 2000. Chaouch, M., Allal, Y., De Sandre-Giovannoli, A., et al.: The phenotypic manifestations of autosomal recessive axonal Charcot-Marie-Tooth due to a mutation in Lamin A/C gene. Neuromuscul. Disord. 13:60, 2003. Colomer, J., Iturriaga, C., Kalaydjieva, L., et al.: Hereditary motor and sensory neuropathy-Lom (HMSNL) in a Spanish family: clinical, electrophysiological, pathological and genetic studies. Neuromuscul. Disord. 10:578, 2000. Cuesta, A., Pedrola, L., Sevilla, T., et al.: The gene encoding ganglioside-induced differentiation-associated protein 1 is mutated in axonal Charcot-Marie-Tooth type 4A disease. Nat. Genet. 30:22, 2002. Dayan, A. D., Graveson, G. S., Robinson, P. K., et al.: Globular neuropathy : a disorder of axons and Schwann cells. J. Neurol. Neurosurg. Psychiatry 31:552, 1968. De Sandre-Giovannoli, A., Chaouch, M., Kozlov, S., et al.: Homozygous defects in LMNA, encoding lamin A/C nuclear-envelope proteins, cause autosomal recessive axonal neuropathy in human (Charcot-Marie-Tooth disorder type 2) and mouse. Am. J. Hum. Genet. 70:726, 2002. Delague, V., Bareil, C., Tuffery, S., et al.: Mapping of a new locus for autosomal recessive demyelinating CharcotMarie-Tooth disease to 19q13.1-13.3 in a large consanguineous Lebanese family: exclusion of MAG as a candidate gene. Am. J. Hum. Genet. 67:236, 2000.

38. Deleu, D., Bamanikar, S. A., Muirhead, D., et al.: Familial progressive sensorimotor neuropathy with agenesis of the corpus callosum (Andermann syndrome): a clinical, neuroradiological and histopathological study. Eur. Neurol. 37:104, 1997. 39. Dubowitz, V.: Muscle disorders in childhood. In Major Problems in Clinical Pediatrics, Vol. 16. London, W. B. Saunders, p.183, 1978. 40. Dubowitz, V.: Very severe spinal muscular atrophy (SMA type 0): an expanding clinical phenotype. Eur. J. Paediatr. Neurol. 3:49, 1999. 41. Dupré, N., Howard, H. C., Mathieu, J., et al.: Hereditary motor and sensory neuropathy with agenesis of the corpus callosum. Ann. Neurol. 54:9, 2003. 42. Dyck, P. J.: Inherited neuronal degeneration and atrophy affecting peripheral motor, sensory, and autonomic neurons. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 825, 1975. 43. Fatkin, D., MacRae, C., Sasaki, T., et al.: Missense mutations in the rod domain of the lamin A/C gene as causes of dilated cardiomyopathy and conduction-system disease. N. Engl. J. Med. 341:1715, 1999. 44. Flanigan, K. M., Crawford, T. O., Griffin, J. W., et al.: Localization of the giant axonal neuropathy gene to chromosome 16q24. Ann. Neurol. 43:143, 1998. 45. Franssen, H., Vermeulen, M., and Jennekens, F. G. I.: Chronic inflammatory neuropathies (Addendum to the electrophysiological criteria of the chronic inflammatory demyelinating neuropathies). In Emery, A. E. H. (ed.): Diagnostic Criteria for Neuromuscular Disorders, 2nd ed. London, Royal Society of Medicine Press, p. 57, 1997. 46. Gabreëls-Festen, A.: Dejerine-Sottas syndrome grown to maturity: overview of genetic and morphological heterogeneity and follow-up of 25 patients. J. Anat. 200:341, 2002. 47. Gabreëls-Festen, A., van Beersum, S., Eshuis, L., et al.: Study on the gene and phenotypic characterisation of autosomal recessive demyelinating motor and sensory neuropathy (Charcot-Marie-Tooth disease) with a gene locus on chromosome 5q23-q33. J. Neurol. Neurosurg. Psychiatry 66:569, 1999. 48. Gabreëls-Festen, A. A. W. M., Gabreëls, F. J. M., Jennekens, F. G. I., et al.: The status of HMSN type III. Neuromuscul. Disord. 4:63, 1994. 49. Gabreëls-Festen, A. A. W. M., Gabreëls, F. J. M., Jennekens, F. G. I., et al.: Autosomal recessive form of hereditary motor and sensory neuropathy type I. Neurology 42:1755, 1992. 50. Gabreëls-Festen, A. A. W. M., Joosten, E. M. G., Gabreëls, F. J. M., et al.: Congenital demyelinating motor and sensory neuropathy with focally folded myelin sheaths. Brain 113:1629, 1990. 51. Gabreëls-Festen, A. A. W. M., Joosten, E. M. G., Gabreëls, F. J. M., et al.: Hereditary motor and sensory neuropathy of neuronal type with onset in early childhood. Brain 114:1855, 1991. 52. Geller, T. J., Connolly, A. M., Kotagel, S., et al.: Progressive infantile axonal neuropathy. J. Child Neurol. 15:513, 2000. 53. Gillespie, C. S., Sherman, D. L., Fleetwood-Walker, S. M., et al.: Peripheral demyelination and neuropathic pain behaviour in periaxin-deficient mice. Neuron 26:523, 2000.

Autosomal Recessive Hereditary Motor and Sensory Neuropathies 54. Grohmann, K., Schuelke, M., Diers, A., et al.: Mutations in the gene encoding immunoglobulin ␮-binding protein 2 cause spinal muscular atrophy with respiratory distress type 1. Nat. Genet. 29:75, 2001. 55. Grohmann, K., Wienker, T. F., Saar, K., et al.: Diaphragmatic spinal muscular atrophy with respiratory distress is heterogeneous, and one form is linked to chromosome 11q13-q21. Am. J. Hum. Genet. 65:1459, 1999. 56. Guilbot, A., Ravisé, N., Bouhouche, A., et al.: Genetic, cytogenetic and physical refinement of the autosomal recessive CMT linked to 5q31-q33: exclusion of candidate genes including EGR1. Eur. J. Hum. Genet. 7:849, 1999. 57. Guilbot, A., Williams, A., Ravisé, N., et al.: A mutation in periaxin is responsible for CMT4F, an autosomal recessive form of Charcot-Marie-Tooth disease. Hum. Mol. Genet. 10:415, 2001. 58. Hagberg, B., Westerberg, B., Hagne, I., et al.: Hereditary motor and sensory neuropathies in Swedish children. III. De- and remyelinating type in 10 sporadic cases. Acta Paediatr. Scand. 72:537, 1983. 59. Hahn, J. S., Henry, M., Hudgins, L., et al.: Congenital hypomyelination neuropathy in a newborn infant: unusual cause of diaphragmatic and vocal cord paralyses. Pediatrics 108:E95, 2002. 60. Harati, Y., and Butler, I. J.: Congenital hypomyelinating neuropathy. J. Neurol. Neurosurg. Psychiatry 48:1269, 1985. 61. Harding, A. E., and Thomas, P. K.: Autosomal recessive forms of hereditary motor and sensory neuropathy. J. Neurol. Neurosurg. Psychiatry 43:669, 1980. 62. Hauser, E., Bittner, R., Liegl, C., et al.: Occurrence of Andermann syndrome out of French Canada—agenesis of the corpus callosum with neuropathy. Neuropediatrics 24:107, 1993. 63. Hentati, A., Ben Othmane, K., Barhoumi, C., et al.: Genotype phenotype relation analysis in autosomal recessive Charcot-Marie-Tooth disease in Tunisia. Acta Myol. 20:25, 2001. 64. Hergersberg, M., Glatzel, M., Capone, A., et al.: Deletions in the spinal muscular atrophy gene region in a newborn with neuropathy and extreme generalized muscular weakness. Eur. J. Paediatr. Neurol. 4:35, 2000. 65. Houlden, H., King, R. H. M., Wood, N. W., et al.: Mutations in the 5⬘ region of the myotubularin-related protein 2 (MTMR2) gene in autosomal recessive hereditary neuropathy with focally folded myelin. Brain 124:907, 2001. 66. Howard, H. C., Mount, D. B., Rochefort, D., et al.: The K-Cl cotransporter KCC3 is mutant in a severe peripheral neuropathy associated with agenesis of the corpus callosum. Nat. Genet. 32:384, 2002. 67. Hunter, M., Bernard, R., Freitas, E., et al.: Mutation screening of the N-myc downstream-regulated gene 1 (NDRG1) in patients with Charcot-Marie-Tooth disease. Hum. Mutat. 22:129, 2003. 68. Joosten, E., Gabreëls, F., Gabreëls-Festen A., et al.: Electron-microscopic heterogeneity of onion-bulb neuropathies of the Dejerine-Sottas type: two patients in one family with the variant described by Lyon (1969). Acta Neuropathol. (Berl.) 27:105, 1974. 69. Joosten, E., Gabreëls-Festen, A., Gabreëls, F., et al.: Idiopathic infantile chronic demyelinating neuropathy:

70.

71.

72. 73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

1787

evidence for a third type, morphologically distinct from the classical and Lyon type. In Abstracts of the 11th World Congress of Neurology (Excerpta Medica International Congress Series No. 427). Amsterdam, Excerpta Medica, p. 296, 1977. Kalaydjieva, L., Gresham, D., Gooding, R., et al.: N-myc downstream-regulated gene 1 is mutated in hereditary motor and sensory neuropathy-Lom. Am. J. Hum. Genet. 67:47, 2000. Kalaydjieva, L., Hallmayer, J., Chandler, D., et al.: Gene mapping in Gypsies identifies a novel demyelinating neuropathy on chromosome 8q24. Nat. Genet. 14:214, 1996. Kalaydjieva, L., King, R., Gresham, D., et al.: Hereditary motor and sensory neuropathy Lom. Acta Myol. 20:192, 2001. Kalaydjieva, L., Nikolova, A., Turnev, I., et al.: Hereditary motor and sensory neuropathy–Lom, a novel demyelinating neuropathy associated with deafness in gypsies: clinical, electrophysiological and nerve biopsy findings. Brain 121:399, 1998. Kalyanaraman, K., Cancilla, P. A., Munsat, T., et al.: Hereditary hypertrophic neuropathy: report of two cases of an autosomal recessive variant. Bull. Los Angeles Neurol. Soc. 35:58, 1970. Kasman, M., Bernstein, L., and Schulman, S.: Chronic polyradiculoneuropathy of infancy: a report of three cases with familial incidence. Neurology 26:565, 1976. Kennedy, W. R., Sung, J. H., and Berry, J. F.: A case of congenital hypomyelination neuropathy: clinical, morphological, and chemical studies. Arch. Neurol. 34:337, 1977. Kennerson, M. L., Zhu, D., Gardner, R. J. M., et al.: Dominant intermediate Charcot-Marie-Tooth neuropathy maps to chromosome 19p12-p13.2. Am. J. Hum. Genet. 69:883, 2001. Kessali, M., Zemmouri, R., Guilbot, A., et al.: A clinical, electrophysiologic, neuropathologic, and genetic study of two large Algerian families with an autosomal recessive demyelinating form of Charcot-Marie-Tooth disease. Neurology 48:867, 1997. King, R. H. M., Tournev, I., Colomer, J., et al.: Ultrastructural changes in peripheral nerve in hereditary motor and sensory neuropathy-Lom. Neuropathol. Appl. Neurobiol. 25:306, 1999. Kiwaki, T., Umehara, F., Takashima, H., et al.: Hereditary motor and sensory neuropathy with myelin folding and juvenile onset glaucoma. Neurology 55:392, 2000. Korinthenberg, R., Sauer, M., Ketelsen, U. P., et al.: Congenital axonal neuropathy caused by deletions in the spinal muscular atrophy region. Ann. Neurol. 42:364, 1997. Krajewski, K. M., Lewis, R. A., Fuerst, D. R., et al.: Neurological dysfunction and axonal degeneration in Charcot-Marie-Tooth disease type 1A. Brain 123:1516, 2000. Kuhlenbäumer, G., Young, P., Oberwittler, C., et al.: Giant axonal neuropathy (GAN): case report and two novel mutations in the gigaxonin gene. Neurology 58:1273, 2002. Kuo, A. A., Pulst, S. M., Eliashiv, D. S., et al.: Electrical inexcitability of nerves and muscles in severe infantile spinal muscular atrophy. J. Neurol. Neurosurg. Psychiatry 67:122, 1999. Larbrisseau, A., Vanasse, M., Brochu, P., et al.: The Andermann syndrome: agenesis of the corpus callosum

1788

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

Diseases of the Peripheral Nervous System

associated with mental retardation and progressive sensorimotor neuronopathy. Can. J. Neurol. Sci. 11:257, 1984. Leal, A., Morera, B., Del Valle, G., et al.: A second locus for an axonal form of autosomal recessive Charcot-Marie-Tooth disease maps to chromosome 19q13.3. Am. J. Hum. Genet. 68:269, 2001. LeGuern, E., Guilbot, A., Kessali, M., et al.: Homozygosity mapping of an autosomal recessive form of demyelinating Charcot-Marie-Tooth disease to chromosome 5q23-q33. Hum. Mol. Genet. 5:1685, 1996. Lesca, G., Cournu-Rebeix, I., Azoulay-Cayla, A., et al.: Le syndrome d’Andermann dans une famille algerienne: suggestion d’une homogénéite phénotypique et génétique. Rev. Neurol. 157:1279, 2001. Lütschg, J., Vassella, F., Boltshauser, E., et al.: Heterogeneity of congenital motor and sensory neuropathies. Neuropediatrics 16:33, 1985. Lyon, G.: Ultrastructural study of a nerve biopsy from a case of early infantile chronic neuropathy. Acta Neuropathol. (Berl.) 13:131, 1969. MacLeod, M. J., Taylor, J. E., Lunt, P. W., et al.: Prenatal onset spinal muscular atrophy. Eur. J. Paediatr. Neurol. 3:65, 1999. Mancardi, G. L., Di Rocco, M., Schenone, A., et al.: Hereditary motor and sensory neuropathy with deafness, mental retardation and absence of large myelinated fibers. J. Neurol. Sci. 110:121, 1992. Martinez Murillo, M., Kobayashi, H., Pegoraro, E., et al.: Genetic localization of a new locus for recessive familial spastic paraparesis to 15q13-15. Neurology 53:50, 1999. Mathieu, J., Bédard, F., Prévost, C., et al.: Neuropathie sensitivo-motrice héréditaire avec ou sans agénésie du corps calleux: étude radiologique et clinique de 64 cas. Can. J. Neurol. Sci. 17:103, 1990. McWilliam, R. C., Gardner-Medwin, D., Doyle, D., et al.: Diaphragmatic paralysis due to spinal muscular atrophy: an unrecognised cause of respiratory failure in infancy? Arch. Dis. Child. 60:145, 1985. Mellins, R. B., Hays, A. P., Gold, A. P., et al.: Respiratory distress as the initial manifestation of Werdnig-Hoffmann disease. Pediatrics 53:33, 1974. Mercuri, E., Goodwin, F., Sewry, C., et al.: Diaphragmatic spinal muscular atrophy with bulbar weakness. J. Paediatr. Neurol. 4:69, 2000. Merlini, L., Gooding, R., Lochmüller, H., et al.: Genetic identity of Marinesco-Sjögren /myoglobinuria and CCFDN syndromes. Neurology 58:231, 2002. Merlini, L., Villanova, M., Sabatelli, P., et al.: Hereditary motor and sensory neuropathy Lom type in an Italian Gypsy family. Neuromuscul. Disord. 8:182, 1998. Meuleman, J., Timmerman, V., Nelis, E., et al.: Molecular genetics of inherited peripheral neuropathies: who are the actors? Acta Neurol. Belg. 100:171, 2000. Midroni, G., and Bilbao, J. M.: Genetically determined neuropathies. In Midroni, G., and Bilbao, J. M. (eds.): Biopsy Diagnosis of Peripheral Neuropathies. Boston, Butterworth-Heinemann, p. 366, 1995. Mohan, U., Misra, V. P., Britto, J., et al.: Inherited early onset severe axonal polyneuropathy with respiratory failure and autonomic involvement. Neuromuscul. Disord. 11:395, 2001.

103. Moss, R. B., Sriram, S., Kelts, K. A., et al.: Chronic neuropathy presenting as a floppy infant with respiratory distress. Pediatrics 64:459, 1979. 104. Muchir, A., Bonne, M., van der Kooi, A., et al.: Identification of mutations in the gene encoding lamins A /C in autosomal dominant limb girdle muscular dystrophy with atrioventricular conduction disturbances (LGMD1B). Hum. Mol. Genet. 9:1453, 2000. 105. Müller, H., Mugler, M., Ramaekers, V., et al.: Hereditary motor and sensory neuropathy with absence of large myelinated fibers due to absence of large neurons in dorsal root ganglia and anterior horns, clinically associated with deafness, mental retardation, and epilepsy (HMSN-ADM). J. Peripher. Nerv. Syst. 5:147, 2000. 106. Müller-Felber, W., Zafiriou, D., Scheck, R., et al.: Marinesco-Sjögren syndrome with rhabdomyolysis: a new subtype of the disease. Neuropediatrics 29:97, 1998. 107. Nara, T., Akashi, M., Nonaka, I., et al.: Muscle and intramuscular nerve pathology in congenital hypomyelination neuropathy. J. Neurol. Sci. 129:170, 1995. 108. Nelis, E., Erdem, S., Tan, E., et al.: A novel homozygous missense mutation in the myotubularin-related protein 2 gene associated with recessive Charcot-Marie-Tooth disease with irregularly folded myelin sheaths. Neuromuscul. Disord. 12:869, 2002. 109. Nelis, E., Erdem, S., Van den Bergh, P. Y. K., et al.: Mutations in GDAP1: autosomal recessive CMT with demyelination and axonopathy. Neurology 59:1865, 2002. 110. Nelis, E., Holmberg, B., Adolfsson, R., et al.: PMP22 Thr(118)Met: recessive CMT1 mutation or polymorphism? Nat. Genet. 15:13, 1997. 111. Nordborg, C., Conradi, N., Sourander, P., et al.: Hereditary motor and sensory neuropathy of demyelinating and remyelinating type in children: ultrastructural and morphometric studies on sural nerve biopsy specimens from ten sporadic cases. Acta Neuropathol. (Berl.) 65:1, 1984. 112. Novelli, G., Capon, F., Tamisari, L., et al.: Neonatal spinal muscular atrophy with diaphragmatic paralysis is unlinked to 5q11.2-q13. J. Med. Genet. 32:216, 1995. 113. Numakura, C., Lin, C., Oka, N., et al.: Hemizygous mutation of the peripheral myelin protein 22 gene associated with Charcot-Marie-Tooth disease type 1. Ann. Neurol. 47:101, 2000. 114. Ohnishi, A., Murai, Y., Ikeda, M., et al.: Autosomal recessive motor and sensory neuropathy with excessive myelin outfolding. Muscle Nerve 12:568, 1989. 115. Omran, H., Ketelsen, U. P., Heinen, F., et al.: Axonal neuropathy and predominance of type II myofibers in infantile spinal muscular atrophy. J. Child Neurol. 13:327, 1998. 116. Ouvrier, R. A.: Giant axonal neuropathy: a review. Brain Dev. 11:207, 1989. 117. Ouvrier, R. A., McLeod, J. G., Morgan, G. J., et al.: Hereditary motor and sensory neuropathy of neuronal type with onset in early childhood. J. Neurol. Sci. 51:181, 1981. 118. Ouvrier, R. A., Prineas, J., Walsh, J. C., et al.: Giant axonal neuropathy—a third case. Proc. Aust. Assoc. Neurol. 11:137, 1974. 119. Pagès, M., Voisin, M., Echenne, B., et al.: Neuropathie congénitale par hypomyélinisation. Arch. Fr. Pediatr. 40:557, 1983.

Autosomal Recessive Hereditary Motor and Sensory Neuropathies 120. Parman, Y., Plante-Bordeneuve, V., Guiochon-Mantel, A., et al.: Recessive inheritance of a new point mutation of the PMP22 gene in Dejerine-Sottas disease. Ann. Neurol. 45:518, 1999. 121. Phillips, J. P., Warner, L. E., Lupski, J. R., et al.: Congenital hypomyelinating neuropathy: two patients with long-term follow-up. Pediatr. Neurol. 20:226, 1999. 122. Pitt, M., Houlden, H., Jacobs, J., et al.: Severe infantile neuropathy with diaphragmatic weakness and its relationship to SMARD1. Brain 126:2682, 2003. 123. Quattrone, A., Gambardella, A., Bono, F., et al.: Autosomal recessive hereditary motor and sensory neuropathy with focally folded myelin sheaths: clinical, electrophysiologic, and genetic aspects of a large family. Neurology 46:1318, 1996. 124. Raffaele di Barletta, M., Ricci, E., Galluzzi, G., et al.: Different mutations in the LMNA gene cause autosomal dominant and autosomal recessive Emery-Dreifuss muscular dystrophy. Am. J. Hum. Genet. 66:1407, 2000. 125. Reilly, M. M.: Classification of the hereditary motor and sensory neuropathies. Curr. Opin. Neurol. 13:561, 2000. 126. Rogers, T., Chandler, D., Angelicheva, D., et al.: A novel locus for autosomal recessive peripheral neuropathy in the EGR2 region on 10q23. Am. J. Hum. Genet. 67:664, 2000. 127. Rozear, M. P., Pericak-Vance, M. A., Fischbeck, K., et al.: Hereditary motor and sensory neuropathy, X-linked: a half century follow-up. Neurology 37:1460, 1987. 128. Sabatelli, M., Mignogna, T., Lippi, G., et al.: Hereditary motor and sensory neuropathy with deafness, mental retardation, and absence of sensory large myelinated fibers: confirmation of a new entity. Am. J. Med. Genet. 75:309, 1998. 129. Salih, M. A. M., Maisonobe, T., Kabiraj, M., et al.: Autosomal recessive hereditary neuropathy with focally folded myelin sheaths and linked to chromosome 11q23: a distinct and homogeneous entity. Neuromuscul. Disord. 10:10, 2000. 130. Santoro, L., Barbieri, F., Crisci, C., et al.: A second family with hereditary motor and sensory neuropathy with deafness, mental retardation and absence of large myelinated fibres, detected in the same geographic area as the first family. J. Neurol. 245:240, 1998. 131. Sawaishi, Y., Hayasaka, K., Goto, A., et al.: Congenital hypomyelination neuropathy: decreased expression of the P2 protein in peripheral nerve with normal DNA sequence of the coding region. J. Neurol. Sci. 134:150, 1995. 132. Senderek, J., Bergmann, C., Ramaekers, V. T., et al.: Mutations in the ganglioside-induced differentiation-associated protein-1 (GDAP1) gene in intermediate type autosomal recessive Charcot-Marie-Tooth neuropathy. Brain 126:642, 2003. 133. Senderek, J., Bergmann, C., Stendel, C., et al.: Mutations in a gene encoding a novel SH3 / TPR domain protein cause autosomal recessive Charcot-Marie-Tooth type 4C neuropathy. Am. J. Hum. Genet. 73:1106, 2003. 134. Senderek, J., Bergmann, C., Weber, S., et al.: Mutation of the SBF2 gene, encoding a novel member of the myotubularin family, in Charcot-Marie-Tooth neuropathy type 4B2/11p15. Hum. Mol. Genet. 12:349, 2003. 135. Sevilla, T., Cuesta, A., Chumillas, M. J., et al.: Clinical and genetic studies in three Spanish families with severe autosomal recessive Charcot-Marie-Tooth axonal neuropathy. Acta Myol. 20:49, 2001.

1789

136. Sevilla, T., Cuesta, A., Chumillas, M. J., et al.: Clinical, electrophysiological and morphological findings of CharcotMarie-Tooth neuropathy with vocal cord palsy and mutations in the GDAP1 gene. Brain 126:2023, 2003. 137. Shackleton, S., Lloyd, D. J., Jackson, S. N. J., et al.: LMNA, encoding lamin A/C, is mutated in partial lipodystrophy. Nat. Genet. 24:153, 2000. 138. Shapira, D., and Swash, M.: Neonatal spinal muscular atrophy presenting as respiratory distress: a clinical variant. Muscle Nerve 8:661, 1985. 139. Smith, T. W., Bhawan, J., Keller, R. B., et al.: CharcotMarie-Tooth disease associated with hypertrophic neuropathy: a neuropathologic study of two cases. J. Neuropathol. Exp. Neurol. 39:420, 1980. 140. Takashima, H., Boerkoel, C. F., De Jonghe, P., et al.: Periaxin mutations cause a broad spectrum of demyelinating neuropathies. Ann. Neurol. 51:709, 2002. 141. Tazir, M., Vallat, J. M., Bomont, P., et al.: Genetic heterogeneity in giant axonal neuropathy: an Algerian family not linked to chromosome 16q24.1. Neuromuscul. Disord. 12:849, 2002. 142. Thomas, P. K., Claus, D., and King, R. H. M.: Autosomal recessive type II hereditary motor and sensory neuropathy with acrodystrophy. J. Neurol. 246:107, 1999. 143. Thomas, P. K., Kalaydjieva, L., Youl, B., et al.: Hereditary motor and sensory neuropathy-Russe: new autosomal recessive neuropathy in Balkan Gypsies. Ann. Neurol. 50:452, 2001. 144. Timmerman, V., De Jonghe, P., Spoelders, P., et al.: Linkage and mutation analysis of Charcot-Marie-Tooth neuropathy type 2 families with chromosomes 1p35-p36 and Xq13. Neurology 46:1311, 1996. 145. Tournev, I., Kalaydjieva, L., Youl, B., et al.: Congenital cataracts facial dysmorphism neuropathy syndrome, a novel complex genetic disease in Balkan Gypsies: clinical and electrophysiological observations. Ann. Neurol. 45:742, 1999. 146. Tournev, I., King, R. H. M., Workman, J., et al.: Peripheral nerve abnormalities in the congenital cataracts facial dysmorphism neuropathy (CCFDN) syndrome. Acta Neuropathol. (Berl.) 98:165, 1999. 147. Tyson, J., Ellis, D., Fairbrother, U., et al.: Hereditary demyelinating neuropathy of infancy: a genetically complex syndrome. Brain 120:47, 1997. 148. Ulrich, J., Hirt, H.-R., Kleihues, P., et al.: Connatal polyneuropathy: a case with proliferated microfilaments in Schwann cells. Acta Neuropathol. (Berl.) 55:39, 1981. 149. Vance, J. M.: The many faces of Charcot-Marie-Tooth disease. Arch. Neurol. 57:638, 2000. 150. Varon, R., Gooding, R., Steglich, C., et al.: Partial deficiency of the C-terminal-domain phosphatase of RNA polymerase II is associated with congenital cataracts facial dysmorphism neuropathy syndrome. Nat. Genet. 35:185, 2003. 151. Vedanarayanan, V. V., Smith, S., Subramony, S. H., et al.: Lethal neonatal autosomal recessive axonal sensorimotor polyneuropathy. Muscle Nerve 21:1473, 1998. 152. Verhoeven, K., Villanova, M., Rossi, A., et al.: Localization of the gene for the intermediate form of Charcot-MarieTooth to chromosome 10q24.1-q25.1. Am. J. Hum. Genet. 69:889, 2001.

1790

Diseases of the Peripheral Nervous System

153. Vital, A., Vital, C., Riviere, J. P., et al.: Variability of morphological features in early infantile polyneuropathy with defective myelination. Acta Neuropathol. (Berl.) 73:295, 1987. 154. Warner, L. E., Mancias, P., Butler, I. J., et al.: Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet. 18:382, 1998.

155. Wilmshurst, J. M., Bye, A., Rittey, C., et al.: Severe infantile axonal neuropathy with respiratory dysfunction. Muscle Nerve 24:760, 2001. 156. Zerres, K., and Davies, K. E.: 59th ENMC International Workshop. Spinal Muscular Atrophies: Recent Progress and Revised Diagnostic Criteria, 17-19 April 1998, Soestduinen, The Netherlands. Neuromuscul. Disord. 9:272, 1999.

76 X-linked Charcot-Marie-Tooth Disease STEVEN S. SCHERER AND KLEOPAS A. KLEOPA

Discovery of X-linked Charcot-Marie-Tooth Disease Mutations That Cause X-linked Charcot-Marie-Tooth Disease Role of Connexins in Formation of Gap Junctions

Analysis of Cx32 Mutants Functional Gap Junctions in the Myelin Sheath Neuromuscular Manifestations Other Manifestations Clinical Electrophysiology

Pathologic Findings Genotype-Phenotype Correlations Medical Issues in CMTX Patients Animal Models of CMTX

DISCOVERY OF X-LINKED CHARCOT-MARIE-TOOTH DISEASE

MUTATIONS THAT CAUSE X-LINKED CHARCOT-MARIE-TOOTH DISEASE

Shortly after the first descriptions of autosomal dominant kindreds with inherited neuropathy by Charcot, Marie, and Tooth in 1886, Herringham46 described a family in which males were selectively affected. He noted the phenotypic similarity of the affected men to the individuals described by Charcot, Marie, and Tooth, and was struck by the finding that the women who appeared to pass the trait of their fathers to their sons were themselves unaffected:

After Herringham, X-linked kindreds with inherited neuropathy were only sporadically reported,2,19,25,28,32,45,102,110 and by the 1980s the existence of X-linked Charcot-MarieTooth disease (CMTX) was even questioned.44 Subsequent investigations confirmed X-linked inheritance in many kindreds.40,53,80,86 Early linkage studies excluded the distal short arm and the distal long arm of the X chromosome.25,53 Using restriction fragment length polymorphisms, many kindreds were linked to the proximal long arm of the X chromosome.7,30,33,34,41,50 Further recombination analyses in several large families refined the localization of CMTX to an approximately 1.5-Mb interval in Xq13.1,12,29,52,58,65 where three genes had been previously mapped, including GJB1. These candidate genes were screened by Northern blot analysis, reasoning that if CMTX was a demyelinating neuropathy, then the causal gene should be expressed by myelinating Schwann cells (peripheral nerve contains little neuronal messenger RNA [mRNA]). Of three candidate genes, the mRNA of GJB1 (also called connexin32 [Cx32]) was the only one detected in peripheral nerve. Sequencing GJB1/Cx32 in eight families led to the discovery of seven different mutations.11 Many groups have confirmed and extended this finding, and more than 240 different mutations affecting the open reading frame have been described (http://molgen-www.uia.ac.be/CMTMutations/ DataSource/MutByGene.cfms). As shown in Figure 76–1,

“This form makes one wonder what inheritance is. That the diseased tissues of a consumptive father should be so represented in his spermatozoon as to cause his child to fall into consumption is remarkable enough. But that the women of this family, themselves even uncommonly buxom and healthy, should be able to select their children, and transmit to the males alone tissues unlike their own, and endowed with a regular form of weakness which they do not themselves possess, is still more marvellous. It seems as if the daughter of a diseased father carried from the beginning of her life ova of two sexes, the female healthy, the male containing within it the representation of the father’s disease.”

Herringham’s analysis was profoundly prescient, because in 1889 Mendel’s discovery of autosomal inheritance was unheralded and Morgan’s demonstration of X-linked inheritance would not appear until 1910.

1791

1792

Diseases of the Peripheral Nervous System

FIGURE 76–1 GJB1 mutations associated with CMTX. Cx32 is an intrinsic membrane protein with four transmembrane domains, one intracellular and two extracellular loops, and an amino- and carboxy-terminal cytoplasmic tail. The positions of the amino acids affected by the known mutations are indicated. (From Kleopa, K. A., and Scherer, S. S.: Inherited neuropathies. Neurol. Clin. North Am. 20:679, 2002, with permission.)

X-linked Charcot-Marie-Tooth Disease

these mutations include missense (amino acid substitutions) and nonsense (premature stop codons) mutations, as well as deletions, insertions, and frameshifts, affecting all regions of the Cx32 protein. In three kindreds, GJB1 is deleted. Many of the mutations have been reported more that once; some of these probably represent founder effects, whereas others may represent mutational “hot spots” in GJB1. Beginning with the initial report,11 some CMTX kindreds were found not to have a mutation in the open reading frame.68 In these families, mutations might affect the GJB1 promoter, splice sites, or the untranslated portions of the mRNA. To understand how mutations in the noncoding regions of the GJB1 gene could cause CMTX, one needs to consider its structure (Fig. 76–2), which contains three alternative promoters; the second exon contains the entire open reading frame.69,70,99,100 Cx32 transcripts in peripheral nerve are mainly initiated at the promoter termed P2, the one nearest the second exon. Transcripts in the liver, embryonic stem cells, oocytes, and pancreas, in contrast, are initiated at the P1 and/or P3 promoters. Cx32 transcripts from the central nervous system (CNS) are initiated from both the P1 and P2 promoters. Thus transcripts initiated at the different promoters have divergent 5⬘ untranslated sequences, whereas the rest of the 5⬘ untranslated region, the coding region, and the 3⬘ untranslated sequence are identical. Ionasescu et al.51 reported noncoding region mutations in two CMTX kindreds that lacked mutations in the open reading frame. One mutation (⫺528T ⬎ G) was just proximal to the start site of transcription, and alters a putative binding site for SOX10,15 a transcription factor expressed in myelinating Schwann cells.57 In transient cotransfection assays, this mutation decreases the expression of the Cx32 gene.15 The other mutation (⫺458C ⬎ T) was in the 5⬘ untranslated region, and appears to abolish an internal ribosome entry site that is essential for the translation of Cx32 mRNA.47 The analysis of another 5⬘ untranslated region mutation indicates that the mutant mRNA is unstable, because Cx32 mRNA from sural nerve biopsies of two heterozygous carriers of this mutation revealed only wild-type Cx32 mRNA.31 A third putative promoter mutation (⫺713G ⬎ A)108 appears to be a polymorphism,10 and it does not decrease expression in transient cotransfection assays.15

Role of Connexins in Formation of Gap Junctions GJB1 encodes the gap junction protein Cx32. Gap junctions are intercellular channels, usually between adjacent cells, and are found in most tissues.17,109 Intercellular gap junctions have been postulated to be involved in a number of processes, including electrical conduction, metabolic cooperation, growth control, cellular differentiation, and

1793

pattern formation during development. Channels are composed of two apposed hemichannels (or connexons) that can form a contiguous pathway between the adjacent cells. Each connexon is composed of a hexamer of connexin molecules arranged around a central pore. All connexins are believed to have a similar structure, consisting of four transmembrane domains that link two extracellular loops and one intracellular loop as well as intracellular N- and C-termini (see Fig. 76–1). The transmembrane domains are ␣ helices; the third transmembrane domain probably forms the central pore, with polar residues lining the wall of the pore. The channel diameter is 1.2 nm, too small to allow transfer of proteins and nucleic acids but large enough to allow the diffusion of ions and other small molecules (⬍1000 Da). The N-terminal domain is involved in the insertion of the nascent polypeptide chain into the endoplasmic reticulum (ER), and, along with the first transmembrane domain, determines voltage gating. The extracellular loops regulate the connexon-connexon interactions, including heterotypic channel formation; each loop contains three cysteine residues (conserved among all connexins) that form essential intramolecular disulfide bonds. The intracellular loop and C-terminal domain regulate pH gating. In keeping with the diversity of tissues with gap junctions, more than 15 different connexins have been found in rodents and humans.17,109 Most connexins are expressed in more than one tissue, and most tissues express more than one connexin. All connexins are highly homologous, indicating that their structure and function were conserved as they evolved from a common ancestral gene. The possible interactions of different connexins are potentially complex, because connexons may be composed of a single connexin (homomeric connexons), or they may be heteromeric. Furthermore, homomeric connexons can couple with other homomeric connexons composed of the same connexin (homotypic junctions), with connexons composed of a different connexin (heterotypic junctions), or even with different combinations of heteromeric connexons. Not all combinations of hemichannels, however, can form heterotypic junctions. The possibility of interactions between different connexins may relate to the pathogenesis of CMTX. If myelinating Schwann cells express another connexin, then some Cx32 mutants could have dominantnegative effects on this connexin. Cx32 (or ␤1 connexin) was the first connexin to be cloned, and was named according to the predicted molecular mass of the protein, 32 kDa. It is highly conserved across mammalian species; the amino acid sequence of human Cx32 protein is 98% identical to those of the mouse and rat. Although Cx32 is most abundant in liver, it is expressed by many tissues, including kidney, intestine, lung, spleen, stomach, pancreas, uterus, testis, and brain (by oligodendrocytes and some kinds of neurons), and by myelinating Schwann cells.21,82,92,99 Despite this broad

1794

Diseases of the Peripheral Nervous System

FIGURE 76–2 The structure of the Cx32 gene and its promoter. There are three alternative promoters, P1-P3 (A), which give rise to three transcripts that differ only in part of their 5⬘ untranslated region (UTR); these are shown schematically in B. C, The nucleotide sequence of the promoter, exon 1B, and the most 5⬘ aspect of exon 2. The locations of SOX10 and EGR2 binding sites, a TATAA box (asterisk), a promoter mutation (at ⫺528), the start site of transcription (arrow), exon 1A (capitalized letters), a mutation in the 5⬘ UTR (at ⫺458), a 355-bp intron, and exon 2 (capitalized letters) are shown. Bases are named according to their position from the ATG initiation codon.

X-linked Charcot-Marie-Tooth Disease

1795

expression pattern, peripheral neuropathy is usually the sole clinical manifestation of GJB1 mutations. Why these other tissues are not affected is unclear. One reason may be the coexpression of another connexin(s), which could “protect” against the loss of Cx32. Myelinating Schwann cells in rodents express connexin 29 (Cx29),3,98 but this does not prevent the development of demyelinating neuropathy.4,93 Whether human myelinating Schwann cells express Cx29 remains to be determined.

Analysis of Cx32 Mutants Based largely on studies in oocytes, tentative structurefunction correlations for a number of the CMTX mutations have been proposed. For example, the third transmembrane domain is thought to form the gap junction pore,64 so that mutations in this domain would be expected to impair or block channel function. Substitution of a proline residue located at position 87 in the second transmembrane segment affects conformational changes associated with voltage gating.84 The N-terminal domain is involved in the insertion of the nascent polypeptide chain into the ER and, along with the first transmembrane domain, regulates voltage gating; many mutants affecting these residues have altered biophysical properties.75,83 The extracellular loops are thought to participate in interactions between apposed connexons, so that mutations in these regions might be expected to impair gap junction formation. Mutating any of the six cysteines (to serine) in the two extracellular loops leads to a loss of channel formation.23 Mutations of the intracellular loop and C-terminal domain may affect pH gating.20 The physiologic effects of different Cx32 mutants that cause CMTX have been studied in Xenopus oocytes. The R220stop mutation forms functional gap junctions; this activity is lost with further truncation of Cx32.81 The R142W, E186K, and 175 frameshift mutations do not form functional channels with wild-type Cx32 or connexin 26 (Cx26).18 Furthermore, the R142W, E186K, and 175 frameshift mutants have dominant-negative effects on wild-type Cx32 and Cx26, demonstrating the potential for such interactions in CMTX. Mutations in the carboxyl terminus, however, do not eliminate gap junction formation or function,20,81 although, compared with wild-type Cx32, some of these mutants are less able to form stable channels, and two (R238H and S281stop) have no known biophysical alterations.20 Another carboxy-terminal mutant that is associated with a severe phenotype has abnormal electrophysiologic characteristics suggesting a gain-offunction mutation.59 Expressing Cx32 mutants in mammalian cells leads to the concept that abnormal trafficking is often the main defect. As illustrated in Figure 76–3, four patterns of protein localization are found.26,55,61,76,106,111,112 No Cx32 is detected in cells transfected with the 175 frameshift muta-

FIGURE 76–3 Immunocytochemical localization of Cx32 mutants. These are images of HeLa cells that have been permanently transfected to express the indicated mutants. Y211stop is localized to the endoplasmic reticulum; R75P is localized to the Golgi apparatus; and R230L forms gap junction–like plaques (arrowheads) similar to wild-type (WT) Cx32.

tion, even though its mRNA is detected. Cells expressing R15Q, V35M, V63I, V139M, N205S, I213V, R215W, R219C/H, R220stop/G, R230C/L, R238H, L239I, C280G, and S281stop have a wild-type pattern, with gap junction–like plaques on the cell surface. Cells expressing G12S, M34K, A39V/P, T55I, A205I, E208K, and Y211stop have Cx32 immunoreactivity in the ER; cells expressing G12S, M34T, V38M, A40V, R75Q/P/W, M93V, R142W, R164Q/W, R183H/S/C, E186K, and C217stop mainly have Cx32 immunoreactivity in the Golgi apparatus. Taken together, these results indicate that mammalian cells may have stringent requirements for the normal trafficking of mutant proteins, because many mutants appear to be retained intracellularly. Of the mutants that reach the cell membrane, at least one (R220stop) forms functional gap junctions,76 as may others. The diversity of trafficking abnormalities has clinical implications. For example, how do the mutants that are able to traffic to the cell membrane (and even form functional gap junctions) cause disease? It seems possible that some of these mutants will not traffic properly in myelinating Schwann cells, but this remains to be directly demonstrated. Although some mutants may have dominant effects on wild-type Cx32, these are not physiologically relevant because normally only one GJB1 gene is expressed. Whether these, and/or other, mutants have dominant effects on Cx29, however, is a potentially important possibility. Finally, in one early study, there seemed to be a relationship between clinical severity and the trafficking of

1796

Diseases of the Peripheral Nervous System

Cx32 mutant protein, because mutant proteins that did not reach the cell surface seemed to be associated with a more severe phenotype.26

Functional Gap Junctions in the Myelin Sheath The localization of Cx32 to the incisures and paranodes, where gap junction–like structures were observed by freezefracture electron microscopy,88,105 suggested that Cx32 forms gap junctions between the layers of the Schwann cell myelin sheath.11 A radial pathway formed by gap junctions at these locations would be up to 1000-fold shorter than the circumferential pathway within the Schwann cell cytoplasm.92 To determine whether there is a direct radial pathway, living myelinating Schwann cells were injected with 5,6-carboxyfluorescein,6 a fluorescent dye of low molecular mass (376 Da) that passes through gap junctions. By injecting the perinuclear region and monitoring the injections by fluorescence microscopy, the pathway of dye diffusion across the incisures of the myelin sheath was documented. In contrast, injected large-molecular-mass fluorescent dyes, such as rhodamineconjugated dextran (10 kDa), did not reach the inner collar of cytoplasm. Furthermore, preincubating the fibers in 18-␣glycerrhetinic acid, a pharmacologic blocker of gap junctions, prevented 5,6-carboxyfluorescein from reaching the inner collar of cytoplasm. These results demonstrate that there are functional gap junctions within incisures that mediate the diffusion of small molecules across the myelin sheath. If Cx32 mediates diffusion across this radial pathway, then the loss of this radial pathway may damage myelinating Schwann cells and their axons, thereby causing neuropathy. The finding that 5,6-carboxyfluorescein diffuses across the myelin sheath in Gjb1/Cx32-null mice indicated that another connexin is present in the Schwann cell myelin sheath; this may be Cx29, which is localized in incisures.3

NEUROMUSCULAR MANIFESTATIONS In males, symptoms typically begin in childhood or adolescence. The initial symptoms include difficulty running and frequently sprained ankles; footdrop and sensory loss in the legs develop later. Depending on the tempo of the disease, the distal weakness may progress to involve the gastrocnemius and soleus muscles (Fig. 76–4), even to the point where assistive devices are required for ambulation. Weakness, atrophy, and sensory loss also develop in the hands, particularly in thenar muscles (Fig. 76–4). These clinical manifestations are the result of a chronic, lengthdependent neuropathy, and are nearly indistinguishable from those seen in patients with Charcot-Marie-Tooth disease types 1A or 1B (CMT1A or CMT1B), although atrophy (particularly of intrinsic hand muscles), positive

FIGURE 76–4 Length-dependent weakness and atrophy in a 61-year-old patient with the Y211stop mutation (patient IV.1 of Hahn et al.40). Note the loss of bulk in the muscles below the knees and in the hands. (Courtesy of Dr. Angelika Hahn; from Hahn, A. F., Brown, W. F., Koopman, W. J., and Feasby, T. E.: X-linked dominant hereditary motor and sensory neuropathy. Brain 113:1511, 1990, with permission of Oxford University Press.)

sensory phenomena, and sensory loss may be more prominent in CMTX patients.38,40,71 Neurologic examinations reveal weakness and atrophy, diminished to absent reflexes, and sensory impairment, all of which are length dependent and worsen insidiously over time but to varying degrees in different patients. Pes cavus, varus deformities, and “hammer toes” are frequently present. Symptoms in women, by comparison, are typically noted toward the end of the second decade, and at every age women are more mildly affected, even asymptomatic. The reason that female carriers are less affected is probably

X-linked Charcot-Marie-Tooth Disease

X-inactivation; only a fraction of their myelinating Schwann cells express the mutant GJB1 allele.93 A few kindreds have been reported to have “recessive” CMTX, but even in these kindreds, at least some obligate carriers have electrophysiologic evidence of peripheral neuropathy.39,73 These clinical manifestations, including their agerelated progression and greater manifestation in males, are elegantly described in a clinical vignette taken from Hahn et al.,40 who described four affected individuals, two men and two women.* This family had a Tyr211stop mutation. “IV.1 . . . was examined at age 61 yrs. The onset of symptoms dated back to early childhood when he noted awkwardness and difficulty in running. He always had high arched feet. In his teens, the muscles in his lower legs became thin and he frequently twisted his ankles. His penmanship was always poor and his fingers and hands were clumsy, particularly in cold weather. Over the years, his symptoms progressed quite insidiously and he developed pain and sensations of pins and needles in his feet and intermittently in his fingers. The hand muscles became progressively wasted, and his fingers clawed, so that he had very poor use of his hands and was now restricted to signing his name. A triple arthrodesis, performed in both feet at age 38, improved his walking. Yet in the last 10 yrs, walking had become more difficult and he has had to use two sticks. On examination there was marked wasting and weakness of the hand muscles, in particular the thenar muscles, and of all muscles below the knee. There was pes cavus and varus deformity. Tendon reflexes were depressed in the arms and at the knees and ankle jerks were absent. There was a graded sensory loss to just above the wrist and to the knee. His gait was unsteady with bilateral foot drop. VI.3 was examined at age 14 yrs. He had difficulty running even when in kindergarten, when his feet were already high arched. At age 8–10 his gait became increasingly awkward. He tripped easily and tended to walk on the lateral edge of the foot. He had difficulty using his hands in cold weather and needed help with shoelaces and buttons. On examination, there was mild wasting and weakness of the hand muscles, in particular the thenar muscles. His feet were high arched with marked hammer toe formation and extensive callosities on the lateral foot borders. Peroneal and anterior tibial muscles, and to a lesser extent the gastrocnemius muscles, were wasted and he was unable to dorsiflex his ankle. His tendon reflexes were depressed and the ankle jerks were absent. All modalities of sensation were reduced to wrist and midcalf levels. Cutaneous nerves were not enlarged. V.1, the 37-yr-old mother of VI.3, was clumsy when running as a child and was never able to ice skate. In her late teens her fingers felt awkward in cold weather and she had *Reprinted from Hahn, A. F., Brown, W. F., Koopman, W. J., and Feasby, T. E.: X-linked dominant hereditary motor and sensory neuropathy. Brain 113:1513–1515, 1990, with permission of Oxford University Press.

1797

trouble doing up buttons. Her feet were always high arched but recently she developed increasing hammer toes and she had a tendency to sprain her ankles. Her feet became numb. The examination showed moderate wasting of the intrinsic hand muscles and the thenar eminences and moderate wasting and weakness of the peroneal muscles and small foot muscles with pes cavus and hammer toe formation. Ankle jerks were absent and sensation was reduced to the wrists and ankles. She walked with a mild steppage gait. V.4 was examined at age 29 yrs. She had high arched feet since childhood, was unable to ice skate and had slight difficulty walking on uneven ground with a tendency to trip. The examination showed slight wasting and weakness of intrinsic hand muscles, highly arched feet and hammer toe formation, but only a little weakness in ankle dorsiflexion and eversion. Ankle jerks were absent and there was a mild impairment of sensation in her toes and soles. Her gait was normal.”

OTHER MANIFESTATIONS Some GJB1 mutations appear to be associated with clinical and magnetic resonance imaging (MRI) findings of CNS involvement. Subclinical involvement of the CNS is common, because the latencies of brainstem auditory evoked responses (BAERs) are delayed in a high proportion of CMTX patients, even in the absence of clinical involvement.72,95 These findings may not be limited to BAERs, because electrophysiologic testing of the visual and motor pathways also revealed abnormal responses in a family with the Asn205Ser mutation, in which affected members did not have clinical or MRI evidence of a CNS abnormality.5 Other mutations have been associated with clinical, electrophysiologic, and/or MRI abnormalities of CNS involvement, including A39V,60 T55I,77,78 R75W,104 M93V,8 E102del,42 R142W,79 R164Q,78 C168Y,89 and R183H.16 How commonly the CNS is affected in CMTX, and what subset of mutations is associated with CNS involvement, however, are unclear. Our initial expression studies of several of these Cx32 mutants associated with clinical CNS involvement showed that they have diverse trafficking abnormalities, similar to other mutations without CNS involvement.55 It is possible that these Cx32 mutants have trans-dominant effects on Cx47, which is also expressed by oligodendrocytes. Mice that lack either Cx32 or Cx47 alone have minimal CNS findings, whereas mice that lack both Cx32 and Cx47 develop devastating CNS demyelination.63,74 Whether partial dominant effects cause the CNS manifestations remains to be determined. One Cx32 mutant—R142Q—appears to causes a unique and unexpected phenotype, hearing loss plus neuropathy.101 Males and females can be affected even in

1798

Diseases of the Peripheral Nervous System

childhood. Why two other families with the same mutation are not reported to have hearing loss is unexplained. Perhaps expression of modifying genes in the cochlea, including several other connexin types (Cx26, Cx30, Cx31, and Cx43), may determine whether a given Cx32 mutation, perhaps in combination with a defect in another gene, will cause clinical hearing loss or only subclinical abnormalities.

CLINICAL ELECTROPHYSIOLOGY “Intermediate” nerve conduction velocities are characteristic of CMTX (Fig. 76–5).27 The forearm motor (median or ulnar) conduction velocities are typically 30 to 40 m/s in affected males and 30 to 50 m/s in affected females,14,39,71,85,95 which is faster than in most CMT1 patients and slower than in most patients with CharcotMarie-Tooth disease type 2 (CMT2). This intermediate slowing is so characteristic of CMTX that it should raise the consideration of CMTX as the diagnosis in an appropriate clinical setting.72 In addition to slowing, the distal motor latencies and F-wave latencies are mildly prolonged. In comparison with CMT1A, conduction slowing in CMTX is less uniform and dispersion is more pronounced.35,103 The intermediate degree of slowing is one of the principal reasons that CMTX was often considered an axonal neuropathy rather than a demyelinating neuropathy.37 This

misconception arose because CMTX was underrecognized during the period when the idea that 38-m/s forearm motor conduction velocities separated CMT1 from CMT was being developed. For example, Harding and Thomas43 did not identify any X-linked kindreds; Davis et al.24 found four kinships with intermediate conduction velocities in which “males tended to be more severely affected than females,” but did not consider them to have CMTX. When the 38-m/s criterion was initially promulgated, it had heuristic value because it facilitated the separation of CMT1 and CMT2. The distribution appeared bimodal, with one peak around 20 m/s (corresponding to CMT1A), and another between 40 and 60 m/s (corresponding to CMT2). Although the two groups overlapped, they appeared to intersect at 38 m/s. This imperfect distinction between CMT1 and CMT2 has been useful in the clinical setting, but an uncritical adherence to this criterion has contributed to the misconception that some inherited demyelinating neuropathies are “axonal,” exemplified by CMT1X. In keeping with the clinical finding of severe distal weakness and atrophy, electrophysiologic testing demonstrates that affected patients have distally accentuated axonal loss.14,39,40,71,85,86,95 The peroneal and tibial motor responses are frequently absent, the median and ulnar motor responses are reduced, and electromyography confirms the loss of motor units. These abnormalities become more common with age and may contribute to the slowing of the motor conduction velocity.

FIGURE 76–5 Intermediate slowing of motor nerve conduction velocities in male and female CMTX patients. (From Dubourg, O., Tardieu, S., Birouk, N., et al.: Clinical, electrophysiological and molecular genetic characteristics of 93 patients with X-linked Charcot-Marie-Tooth disease. Brain 124:1958, 2001; with permission of Oxford University Press.)

X-linked Charcot-Marie-Tooth Disease

PATHOLOGIC FINDINGS Nerve biopsies from CMTX patients have been reported,14,35,39,40,71,86,87,94,95,103,107 including those patients lacking the Cx32 gene.38,66 Biopsies appear similar; the most salient finding is age-related loss of myelinated fibers, paralleled by increasing numbers of clusters of regenerated axons (Fig. 76–6). Some workers have emphasized onion bulb–like structures,35,66,86,103 but these are seldom well developed. Many myelin sheaths are inappropriately thin for axonal diameter, suggesting chronic segmental demyelination and remyelination, although Vital et al.107 suggested that hypomyelination is the primary defect. This issue could be resolved by studying teased fibers from patients of different ages, because findings in Gjb1/Cx32-null mice strongly support the former interpretation. Some workers have noted enlargement of the adaxonal Schwann cell cytoplasm36,95 and increased packing density of neurofilaments,36 both features also seen in Gjb1/Cx32-null mice.4,93 In Gjb1/Cx32-null mice, macrophages may mediate some of the pathologic changes.56

GENOTYPE-PHENOTYPE CORRELATIONS Given the large number and variety of GJB1 mutations, the question arises as to whether different mutations cause different degrees of clinical involvement. This is clearly the case for mutations in the genes encoding other myelin proteins, such as proteolipid protein (PLP), peripheral myelin protein 22 (PMP22), and myelin protein P0 (P0),54,67 but whether different GJB1 mutations

FIGURE 76–6 Pathologic findings in CMTX. Electron micrograph of a biopsied peroneal sensory nerve from a 61-year-old man with the Y211stop mutation40; the tissue was kindly supplied by Dr. Angelika Hahn.

1799

cause different degrees of neuropathy is still unresolved. Some mutations appear to cause exceptionally severe phenotypes, typically with early onset—R22stop,13,48 the double mutation R22Q and V63I,96 147 frameshift,62 C201R,97 F235C,59 and the deletion nt 265-273 deletion/ frameshift.48 Moreover, some of these mutations— R22stop, 147 frameshift, P172L, and deletion nt 265-273 deletion/frameshift—appear to affect females at much younger ages than typical GJB1 mutations. How these mutations cause a more severe phenotype is unknown, but the mutant proteins probably have deleterious effects on myelinating Schwann cells that transcend a simple loss of function. Whether the unusually pronounced phenotype in these severely affected females is the result of skewed X-inactivation is also unknown. The main point for clinicians to consider is that GJB1 mutations can cause severe neuropathy in young children of either sex. Ionasescu and colleagues48,49 suggested the following clinical classification for CMTX kindreds: 1. Mild phenotype: “relatively good muscle strength, compatible with normal gait and ability to climb stairs . . . foot deformities and areflexia” 2. Moderate phenotype: “weakness of the tibialis anterior and peroneal muscles requiring ankle foot orthosis (braces), positive Romberg sign, distal hypesthesia, weakness of palmar and dorsal interossei, and scoliosis” 3. Severe phenotype: “proximal limb weakness in addition to distal weakness . . . need canes or wheelchairs . . . sensory and balance impairments . . . phrenic nerve involvement is sometimes present” They examined patients from 29 families with 24 different GJB1 mutations (17 missense mutations, 5 nonsense mutations, and 2 noncoding region mutations); 2 families showed linkage to the GJB1 locus, but no mutations have been identified. They reported a relationship between the type of mutation and the severity of the clinical phenotype: Deletions, frameshifts, and premature truncations appeared to be more deleterious than missense mutations. Hahn and colleagues39 also looked at this issue in 116 individuals from 13 families; each family had a different GJB1 mutation, including one family with a deletion of the Cx32 gene. They classified patients as follows: 0 ⫽ “none” 1 ⫽ “mild changes on examination only, no significant functional impairment” 2 ⫽ “obvious changes on exam, mild functional impairment, capable of gainful employment” 3 ⫽ “obvious changes on exam, moderate functional impairment, disabled for work, capable of self-care or activities of daily living” 4 ⫽ “severe changes on exam, severe functional impairment, requiring help with or incapable of self-care”

1800

Diseases of the Peripheral Nervous System

They found that all mutations produced a similar phenotype, which varied among males within the same kinship, suggesting that epigenetic factors modify the severity of disease. By the fourth to sixth decade, males were moderately (class 3) or severely (class 4) affected. Most mutations resulted in a severe phenotype, including deletion of the entire Cx32 gene. The three mutations that resulted in a moderate phenotype were all missense mutations in the second transmembrane domain and/or cytoplasmic loop. Hahn et al.39 suggested that mutant proteins encoded by these mutations may only have a partial loss of function. Dr. Michael Shy ([email protected]) and Dr. Angelika Hahn ([email protected]) are studying CMTX patients to determine the possible relationship between the genotype and phenotype. Based on initial observations, phenotype-genotype correlations are difficult to make, even within the same families. For example, three males of about the same age with the Q80R mutation had significantly different neuropathy severity scores (M. Shy and A. Hahn, personal communication, 2003). Study of a larger number of patients with different mutations may provide more conclusive phenotype-genotype correlations.

MEDICAL ISSUES IN CMTX PATIENTS There are no known treatments for CMTX. It is recommended that patients with inherited neuropathies avoid vincristine, which can cause acute worsening in CMT1 and CMT2 patients, but at least one CMTX patient was treated with cisplatin without noticeable clinical worsening.22 Finally, there is the worrisome possibility that individuals with CMTX are predisposed to develop cerebral edema at high altitudes, which has apparently happened to two males with different mutations (R142W and C168Y).79

ANIMAL MODELS OF CMTX All the inherited demyelinating neuropathies identified to date are caused by mutations in genes expressed by myelinating Schwann cells. Similarly, in the CNS, mutations in genes expressed by oligodendrocytes, such as PLP, cause inherited demyelinating diseases in humans and other animals.67 There is emerging evidence, largely based on the analysis of “knockout” mice, that the phenotypes of most naturally occurring mutations in PMP22/Pmp22, MPZ/Mpz, and PLP/Plp result from toxic effects and not a simple loss of function.9,90 Loss-of-function mutations of Gjb1/Cx32 have been created by the targeted deletion of the Cx32 gene in mice. These mice develop a progressive, demyelinating peripheral neuropathy beginning at about 3 months of age.4,93 For unknown reasons, motor fibers are much more affected than sensory fibers, a feature not noted in CMTX

patients. Like other genes of the X chromosome, Gjb1/Cx32 appears to be randomly inactivated, because in heterozygous females only some myelinating Schwann cells express Cx32.93 Heterozygous females have fewer demyelinated and remyelinated axons than age-matched Gjb1/Cx32-null females or males,93 which may explain why women who are obligate carriers of CMTX are often clinically unaffected. To determine whether some Cx32 mutants have more than a simple loss of function, transgenic mice expressing the 175 frameshift1 or the R142W mutation91 were generated. No Cx32 protein could be detected and no peripheral neuropathy was noted in 26 lines of mice expressing the 175 transgene, even though transgene mRNA was highly expressed in some lines. Mice expressing the R142W mutation, in contrast, expressed mutant Cx32 and developed a demyelinating peripheral neuropathy.91 The mutant protein remained in the perinuclear region and did not reach incisures or paranodes, the normal sites of Cx32 immunoreactivity in the myelin sheath.21,92 Furthermore, the expression of the mutant Cx32 reduced the level of the endogenous/wild-type Cx32, indicating that this mutant may have dominant-negative interactions with endogenous Cx32. Thus the expression and trafficking of these two Cx32 mutants is remarkably similar to that observed in transfected cells,26 indicating that altered trafficking of mutant connexins may be one of the fundamental perturbations caused by Cx32 mutations.

ACKNOWLEDGMENTS We thank our collaborators, especially Drs. Annette Abel, Rita Balice-Gordon, Linda Bone, Roberto Bruzzone, Kurt Fischbeck, Angelika Hahn, Albee Messing, David Paul, Michael Shy, and Klaus Willecke. S. S. S. was supported by grants from the National Institutes of Health, and K. A. K. was supported by the National Multiple Sclerosis Society.

REFERENCES 1. Abel, A., Bone, L. J., Messing, A., et al.: Studies in transgenic mice indicate a loss of connexin32 function in X-linked Charcot-Marie-Tooth disease. J. Neuropathol. Exp. Neurol. 58:702, 1999. 2. Allan, W.: Relation of hereditary pattern to clinical severity as illustrated by peroneal atrophy. Arch. Intern. Med. 63:1123, 1939. 3. Altevogt, B. M., Kleopa, K. A., Postma, F. R., et al.: Distinct expression of Cx29 by myelinating glial cells of the central and peripheral nervous system. J. Neurosci. 22:6458, 2002. 4. Anzini, P., Neuberg, D. H.-H., Schachner, M., et al.: Structural abnormalities and deficient maintenance of peripheral nerve myelin in mice lacking the gap junction protein connexin32. J. Neurosci. 17:4545, 1997.

X-linked Charcot-Marie-Tooth Disease 5. Bähr, M., Andres, F., Timmerman, V., et al.: Central visual, acoustic, and motor pathway involvement in a CharcotMarie-Tooth family with an Asn205Ser mutation in the connexin32 gene. J. Neurol. Neurosurg. Psychiatry 66:202, 1999. 6. Balice-Gordon, R. J., Bone, L. J., and Scherer, S. S.: Functional gap junctions in the Schwann cell myelin sheath. J. Cell Biol. 142:1095, 1998. 7. Beckett, J., Holden, J. J. A., Simpson, N. E., et al.: Localization of X-linked dominant Charcot-Marie-Tooth disease (CMT2) to Xq13. J. Neurogenet. 3:225, 1986. 8. Bell, C., Willison, H., Clark, C., and Haites, N.: CNS abnormalities in a family with a connexin32 mutation and peripheral neuropathy. Eur. J. Hum. Genet. 4:S136, 1996. 9. Berger, P., Young, P., and Suter, U.: Molecular cell biology of Charcot-Marie-Tooth disease. Neurogenetics 4:1, 2002. 10. Bergmann, C., Schröder, J. M., Rudnik-Schneborn, S., et al.: A point mutation in the human connexin32 promoter P2 does not correlate with X-linked dominant Charcot-Marie-Tooth neuropathy in Germany. Mol. Brain Res. 88:183, 2001. 11. Bergoffen, J., Scherer, S. S., Wang, S., et al.: Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262:2039, 1993. 12. Bergoffen, J., Trofatter, J., Pericak-Vance, M. A., et al.: Linkage localization of X-linked Charcot-Marie-Tooth disease. Am. J. Hum. Genet. 52:312, 1993. 13. Birouk, N., Gouider, R., Le Guern, E., et al.: CharcotMarie-Tooth disease type 1A with 17p11.2 duplication— clinical and electrophysiological phenotype study and factors influencing disease severity in 119 cases. Brain 120:813, 1997. 14. Birouk, N., Le Guern, E., Maisonobe, T., et al.: X-linked Charcot-Marie-Tooth disease with connexin32 mutations— clinical and electrophysiological study. Neurology 50:1074, 1998. 15. Bondurand, N., Girard, M., Pingault, V., et al.: Human connexin 32, a gap junction protein altered in the X-linked form of Charcot-Marie-Tooth disease, is directly regulated by the transcription factor SOX10. Hum. Mol. Genet. 10:2783, 2001. 16. Bort, S., Nelis, E., Timmerman, V., et al.: Mutational analysis of the MPZ, PMP22 and Cx32 genes in patients of Spanish ancestry with Charcot-Marie-Tooth disease and hereditary neuropathy with liability to pressure palsies. Hum. Genet. 99:746, 1997. 17. Bruzzone, R., White, T. W., and Paul, D. L.: Connections with connexins: the molecular basis of direct intercellular signaling. Eur. J. Biochem. 238:1, 1996. 18. Bruzzone, R., White T. W., Scherer, S. S., et al.: Null mutations of connexin32 in patients with X-linked CharcotMarie-Tooth disease. Neuron 13:1253, 1994. 19. Campeanu, E., and Morariu, M.: Les relations entre genotype et phenotype dans la maladie de Charcot-Marie-Tooth. Rev. Roum. Neurol. 7:47, 1970. 20. Castro, C., Gomez-Hernandez, J. M., Silander, K., and Barrio, L. C.: Altered formation of hemichannels and gap junction channels caused by C-terminal connexin-32 mutations. J. Neurosci. 19:3752, 1999. 21. Chandross, K. J., Kessler, J. A., Cohen, R. I., et al.: Altered connexin expression after peripheral nerve injury. Mol. Cell. Neurosci. 7:501, 1996.

1801

22. Cowie, F., and Barrett, A.: Uneventful administration of cisplatin to a man with X-linked Charcot-Marie-Tooth disease (CMT). Ann. Oncol. 12:422, 2001. 23. Dahl, G., Werner, R., Levine, E., and Rabadan-Diehl, C.: Mutational analysis of gap junction formation. Biophys. J. 62:172, 1992. 24. Davis, C. J., Bradley, W. G., and Madrid, R.: The peroneal muscular atrophy syndrome—clinical, genetic, electrophysiological and nerve biopsy studies. I. Clinical, genetic and electrophysiological findings and classification. J. Genet. Hum. 26:311, 1978. 25. de Weerdt, C. J.: Charcot-Marie-Tooth disease with sexlinked inheritance, linkage studies and abnormal serum alkaline phosphatase levels. Eur. Neurol. 17:336, 1978. 26. Deschênes, S. M., Walcott, J. L., Wexler, T. L., et al.: Altered trafficking of mutant connexin32. J. Neurosci. 17:9077, 1997. 27. Dubourg, O., Tardieu, S., Birouk, N., et al.: Clinical, electrophysiological and molecular genetic characteristics of 93 patients with X-linked Charcot-Marie-Tooth disease. Brain 124:1958, 2001. 28. Erwin, W. G.: A pedigree of sex-linked recessive peroneal atrophy. J. Hered. 35:24, 1944. 29. Fain, P. R., Barker, D. F., and Chance, P. F.: Refined genetic mapping of X-linked Charcot-Marie-Tooth neuropathy. Am. J. Hum. Genet. 54:229, 1994. 30. Fischbeck, K.H., ar-Rushdi, N., Pericak-Vance, M., et al.: X-linked neuropathy: gene localization with DNA probes. Ann. Neurol. 20:527, 1986. 31. Flagiello, L., Cirigliano, V., Strazzullo, M., et al.: Mutation in the nerve-specific 5⬘ non-coding region of Cx32 gene and absence of specific mRNA in a CMTX1 Italian family. Hum. Mutat. 12:361, 1998. 32. Fryns, J. P., and Van den Berghe, H.: Sex-linked recessive inheritance in Charcot-Marie-Tooth disease with partial clinical manifestation in female carriers. Hum. Genet. 55:413, 1980. 33. Gal, A., Mucke, J., Theile, H., et al.: X-linked dominant Charcot-Marie-Tooth disease: suggestion of linkage with a cloned DNA sequence from the proximal Xq. Hum. Genet. 70:38, 1985. 34. Goonewardena, P., Welinhinda, J., Anvret, M., et al.: A linkage study of the locus for X-linked Charcot-Marie-Tooth disease. Clin. Genet. 33:435, 1988. 35. Gutierrez, A., England, J. D., Sumner, A. J., et al.: Unusual electrophysiological findings in X-linked dominant CharcotMarie-Tooth disease. Muscle Nerve 23:182, 2000. 36. Hahn, A., Ainsworth, P. J., Bolton, C. F., et al.: Pathological findings in the X-linked form of Charcot-Marie-Tooth disease: a morphometric and ultrastructural analysis. Acta Neuropathol. 101:129, 2001. 37. Hahn, A. F.: Hereditary motor and sensory neuropathy: HMSN type II (neuronal type) and X-linked HMSN. Brain Pathol. 3:147, 1993. 38. Hahn, A. F., Ainsworth, P. J., Naus, C. C. G., et al.: Clinical and pathological observations in men lacking the gap junction protein connexin 32. Muscle Nerve 9:S39, 2000. 39. Hahn, A. F., Bolton, C. F., White, C. M., et al.: Genotype/phenotype correlations in X-linked CharcotMarie-Tooth disease. Ann. N. Y. Acad. Sci. 883:366, 1999.

1802

Diseases of the Peripheral Nervous System

40. Hahn, A. F., Brown, W. F., Koopman, W. J., and Feasby, T. E.: X-linked dominant hereditary motor and sensory neuropathy. Brain 113:1511, 1990. 41. Haites, N., Fairweather, N., Clark, C., et al.: Linkage in a family with X-linked Charcot-Marie-Tooth disease. Clin. Genet. 35:399, 1989. 42. Hanemann, C. O., Bergmann, C., Senderek, J., et al.: Transient, recurrent, white matter lesions in X-linked Charcot-Marie-Tooth disease with novel connexin 32 mutation. Arch. Neurol. 60:605, 2003. 43. Harding, A. E., and Thomas, P. K.: The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103:259, 1980. 44. Harding, A. E., and Thomas, P. K.: Genetic aspects of hereditary motor and sensory neuropathy (types I and II). J. Med. Genet. 17:329, 1980. 45. Heimler, A., Friedman, E., and Rosenthal, A. D.: Naevoid basal cell carcinoma syndrome and Charcot-Marie-Tooth disease: two dominant disorders segregating in a family. J. Med. Genet. 15:288, 1978. 46. Herringham, W. P.: Muscular atrophy of the peroneal type affecting many members of a family. Brain 11:230, 1889. 47. Hudder, A., and Werner, R.: Analysis of a Charcot-MarieTooth disease mutation reveals an essential internal ribosome entry site element in the connexin-32 gene. J. Biol. Chem. 275:34586, 2000. 48. Ionasescu, V., Ionasescu, R., and Searby, C.: Correlation between connexin 32 gene mutations and clinical phenotype in X-linked dominant Charcot-Marie-Tooth neuropathy. Am. J. Med. Genet. 63:486, 1996. 49. Ionasescu, V. V.: X-linked Charcot-Marie-Tooth disease and connexin32. Cell Biol. Int. 22:807, 1998. 50. Ionasescu, V. V., Burns, T. L., Searby, C., and Ionasescu, R.: X-linked dominant Charcot-Marie-Tooth neuropathy with 15 cases in a family: genetic linkage study. Muscle Nerve 11:1154, 1988. 51. Ionasescu, V. V., Searby, C., Ionasescu, R., et al.: Mutations of noncoding region of the connexin32 gene in X-linked dominant Charcot-Marie-Tooth neuropathy. Neurology 47:541, 1996. 52. Ionasescu, V. V., Trofatter, J., Haines, J. L., et al.: X-linked recessive Charcot-Marie-Tooth neuropathy—clinical and genetic study. Muscle Nerve 15:368, 1992. 53. Iselius, L., and Grimby, L.: A family with Charcot-MarieTooth disease showing a probable X-linked incompletely dominant inheritance. Heriditas 97:157, 1982. 54. Kleopa, K. A., and Scherer, S. S.: Inherited neuropathies. Neurol. Clin. North Am. 20:679, 2002. 55. Kleopa, K. A., Yum, S. W., and Scherer, S. S.: Cellular mechanisms of connexin32 mutations associated with CNS manifestations. J. Neurosci. Res. 68:522, 2002. 56. Kobsar, I., Maurer, M., Ott, T., and Martini, R.: Macrophage-related demyelination in peripheral nerves of mice deficient in the gap junction protein connexin 32. Neurosci. Lett. 320:17, 2002. 57. Kuhlbrodt, K., Herbarth, B., Sock, E., et al.: Sox10, a novel transcriptional modulator in glial cells. J. Neurosci. 18:237, 1998. 58. Le Guern, E., Ravise, N., Gugenheim, M., et al.: Linkage analyses between dominant X-linked Charcot-Marie-Tooth

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

disease, and 15 Xq11-Xq21 microsatellites in a new large family: three new markers are closely linked to the gene. Neuromuscul. Disord. 4:463, 1994. Lin, G. S., Glass, J. D., Shumas, S., et al.: A unique mutation in connexin32 associated with severe, early onset CMTX in a heterozygous female. Ann. N. Y. Acad. Sci. 883:481, 1999. Marques, W. Jr., Sweeney, M. G., Wood, N. W., et al.: Central nervous system involvement in a novel connexin 32 mutation affecting identical twins. J. Neurol. Neurosurg. Psychiatry 66:803, 1999. Martin, P. E. M., Mambetisaeva, E. T., Archer, D. A., et al.: Analysis of gap junction assembly using mutated connexins detected in Charcot-Marie-Tooth X-linked disease. J. Neurochem. 74:711, 2000. Meggouh, F., Benomar, A., Rouger, H., et al.: The first de novo mutation of the connexin 32 gene associated with X linked Charcot-Marie-Tooth disease. J. Med. Genet. 35:251, 1998. Menichella, D. M., Goodenough, D. A., Sirkowski, E., et al.: Connexins are critical for normal myelination in the central nervous system. J. Neurosci. 23:5963, 2003. Milks, L. C., Kumar, N. M., Houghten, R., et al.: Topology of the 32-kd liver gap junction protein determined by sitedirected antibody localizations. EMBO J. 7:2967, 1988. Mostacciuolo, M. L., Muller, E., Fardin, P., et al.: X-linked Charcot-Marie-Tooth disease: a linkage study in a large family by using 12 probes for the pericentromeric region. Hum. Genet. 87:23, 1991. Nakagawa, M., Takashima, H., Umehara, F., et al.: Clinical phenotype in X-linked Charcot-Marie-Tooth disease with an entire deletion of the connexin 32 coding sequence. J. Neurol. Sci. 185:31, 2001. Nave, K.-A., and Boespflug-Tanguy, O.: Developmental defects of myelin formation: from X-linked mutations to human dysmyelinating diseases. Neuroscientist 2:33, 1996. Nelis, E., Haites, N., and Van Broeckhoven, C.: Mutations in the peripheral myelin genes and associated genes in inherited peripheral neuropathies. Hum. Mutat. 13:11, 1999. Neuhaus, I. M., Bone, L., Wang, S., et al.: The human connexin32 gene is transcribed from two tissue-specific promoters. Biosci. Rep. 16:239, 1996. Neuhaus, I. M., Dahl, G., and Werner, R.: Use of alternative promoters for tissue-specific expression of the gene coding for connexin32. Gene 158:257, 1995. Nicholson, G. A., and Nash, J.: Intermediate nerve conduction velocities define X-linked Charcot-Marie-Tooth neuropathy families. Neurology 43:2558, 1993. Nicholson, G. A., Yeung, L., and Corbett, A.: Efficient neurophysiologic selection of X-linked Charcot-MarieTooth families. Neurology 51:1412, 1998. Niewiadomski, L. A., and Kelly, T. E.: X-linked CharcotMarie-Tooth disease: molecular analysis of interfamilial variability. Am. J. Med. Genet. 66:175, 1996. Odermatt, B., Wellershaus, K., Wallraff, A., et al.: Connexin 47 (Cx47)-deficient mice with enhanced green fluorescent protein reporter gene reveal predominant oligodendrocytic expression of Cx47 and display vacuolized myelin in the CNS. J. Neurosci. 23:4549, 2003.

X-linked Charcot-Marie-Tooth Disease 75. Oh, S., Ri, Y., Bennett, M. V. L., et al.: Changes in permeability caused by connexin 32 mutations underlie X-linked Charcot-Marie-Tooth disease. Neuron 19:927, 1997. 76. Omori, Y., Mesnil, M., and Yamasaki, H.: Connexin 32 mutations from X-linked Charcot-Marie-Tooth disease patients: functional defects and dominant negative effects. Mol. Biol. Cell 7:907, 1996. 77. Panas, M., Kalfakis, N., Karadimas, C., and Vassilopoulos, D.: Episodes of generalized weakness in two sibs with the C164T mutation of the connexin 32 gene. Neurology 57:1906, 2001. 78. Panas, M., Karadimas, C., Avramopoulos, D., and Vassilopoulos, D.: Central nervous system involvement in four patients with Charcot-Marie-Tooth disease with connexin 32 extracellular mutations. J. Neurol. Neurosurg. Psychiatry 65:947, 1998. 79. Paulson, H., Garbern, J. Y., Hoban, T. F., et al.: Transient CNS white matter abnormality in X-linked Charcot-MarieTooth disease. Ann. Neurol. 52:429, 2002. 80. Phillips, L. H., Kelly, T. E., Schnatterly, P., and Parker, D.: Hereditary motor-sensory neuropathy (HMSN): possible X-linked dominant inheritance. Neurology 35:498, 1985. 81. Rabadan-Diehl, C., Dahl, G., and Werner, R.: A connexin-32 mutation associated with Charcot-Marie-Tooth disease does not affect channel formation in oocytes. FEBS Lett. 351:90, 1994. 82. Ressot, C., and Bruzzone, R.: Connexin channels in Schwann cells and the development of the X-linked form of Charcot-Marie-Tooth disease. Brain Res. Rev. 32:192, 2000. 83. Ressot, C., Gomes, D., Dautigny, A., et al.: Connexin32 mutations associated with X-linked Charcot-Marie-Tooth disease show two distinct behaviors: loss of function and altered gating properties. J. Neurosci. 18:4063, 1998. 84. Ri, Y., Ballesteros, J. A., Abrams, C. K., et al.: The role of a conserved proline residue in mediating conformational changes associated with voltage gating of Cx32 gap junctions. Biophys. J. 76:2887, 1999. 85. Rouger, H., LeGuern, E., Birouk, N., et al.: Charcot-MarieTooth disease with intermediate motor nerve conduction velocities: characterization of 14 Cx32 mutations in 35 families. Hum. Mutat. 10:443, 1997. 86. Rozear, M. P., Pericak-Vance, M. A., Fischbeck, K., et al.: Hereditary motor and sensory neuropathy, X-linked: a half century follow-up. Neurology 37:1460, 1987. 87. Sander, S., Nicholson, G. A., Ouvrier, R. A., et al.: CharcotMarie-Tooth disease: histopathological features of the peripheral myelin protein (PMP22) duplication (CMT1A) and connexin32 mutations (CMTX1). Muscle Nerve 21:217, 1998. 88. Sandri, C., Van Buren, J. M., and Akert, K.: Membrane morphology of the vertebrate nervous system: a study with freeze-etch technique. Prog. Brain Res. 46:1,1977. 89. Schelhaas, H. J., Van Engelen, B. G., Gabreels-Festen, A. A., et al.: Transient cerebral white matter lesions in a patient with connexin 32 missense mutation. Neurology 59:2007, 2002. 90. Scherer, S. S.: Molecular genetics of demyelination: new wrinkles on an old membrane. Neuron 18:13, 1997. 91. Scherer, S. S., Bone, L. J., Deschênes, S. M., et al.: The role of the gap junction protein connexin32 in the pathogenesis of X-linked Charcot-Marie-Tooth disease. In Cardew, G. (ed.): Gap Junction-Mediated Intercellular Signalling in

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

1803

Health and Disease (Novartis Foundation Symposium 219). New York, John Wiley & Sons, p. 175, 1999. Scherer, S. S., Deschênes, S. M., Xu, Y.-T., et al.: Connexin32 is a myelin-related protein in the PNS and CNS. J. Neurosci. 15:8281, 1995. Scherer, S. S., Xu, Y.-T., Nelles, E., et al.: Connexin32-null mice develop a demyelinating peripheral neuropathy. Glia 24:8, 1998. Senderek, J., Bergmann, C., Quasthoff, S., et al.: X-linked dominant Charcot-Marie-Tooth disease: nerve biopsies allow morphological evaluation and detection of connexin32 mutations (Arg15Trp, Arg22Gln). Acta Neuropathol. 95:443, 1998. Senderek, J., Hermanns, B., Bergmann, C., et al.: X-linked dominant Charcot-Marie-Tooth neuropathy: clinical, electrophysiological, and morphological phenotype in four families with different connexin32 mutations. J. Neurol. Sci. 167:90, 1999. Silander, K., Meretoja, P., Juvonen, V., et al.: Spectrum of mutations in Finnish patients with Charcot-Marie-Tooth disease and related neuropathies. Hum. Mutat. 12:59, 1998. Sillen, A., Anneren, G., and Dahl, N.: A novel mutation (C201R) in the transmembrane domain of connexin32 in severe X-linked Charcot-Marie-Tooth disease. Hum. Mutat. 12(Suppl. 1):S8, 1998. Söhl, G., Eiberger, J., Jung, Y. T., et al.: The mouse gap junction gene connexin29 is highly expressed in sciatic nerve and regulated during brain development. Biol. Chem. 382:973, 2001. Söhl, G., Gillen, C., Bosse, F., et al.: A second alternative transcript of the gap junction gene connexin32 is expressed in murine Schwann cells and modulated in injured sciatic nerve. Eur. J. Cell Biol. 69:267, 1996. Söhl, G., Theis, M., Hallas, G., et al.: A new alternatively spliced transcript of the mouse connexin32 gene is expressed in embryonic stem cells, oocytes, and liver. Exp. Cell Res. 266:177, 2001. Stojkovic, T., Latour, P., Vandenberghe, A., et al.: Sensorineural deafness in X-linked Charcot-Marie-Tooth disease with connexin 32 mutations (R142Q). Neurology 52:1010, 1999. Swift, M. R., and Horowitz, S. L.: Familial jaw cysts in Charcot-Marie-Tooth disease. J. Med. Genet. 6:193, 1969. Tabaraud, F., Lagrange, E., Sindou, P., et al.: Demyelinating X-linked Charcot-Marie-Tooth disease: unusual electrophysiological findings. Muscle Nerve 22:1442, 1999. Taylor, R. A., Simon, E. M., Marks, H. G., et al.: The CNS phenotype of X-linked Charcot-Marie-Tooth disease: more than a peripheral problem. Neurology 61:1475, 2003. Tetzlaff, W.: Tight junction contact events and temporary gap junctions in the sciatic nerve fibres of the chicken during Wallerian degeneration and subsequent regeneration. J. Neurocytol. 11:839, 1982. VanSlyke, J. K., Deschênes, S. M., and Musil, L. S.: Intracellular transport, assembly, and degradation of wildtype and disease-linked mutant gap junction proteins. Mol. Biol. Cell. 11:1933, 2000.

1804

Diseases of the Peripheral Nervous System

107. Vital, A., Ferrer, X., Lagueny, A., et al.: Histopathological features of X-linked Charcot-Marie-Tooth disease in 8 patients from 6 families with different connexin32 mutations. J. Peripher. Nerv. Syst. 6:79, 2001. 108. Wang, H. L., Wu, T., Chang, W. T., et al.: Point mutation associated with X-linked dominant Charcot-Marie-Tooth disease impairs the P2 promoter activity of human connexin-32 gene. Mol. Brain Res. 78:146, 2000. 109. White, T. W., and Paul, D. L.: Genetic diseases and gene knockouts reveal diverse connexin functions. Annu. Rev. Physiol. 61:283, 1999.

110. Woratz, G.: Neurale Muskelatrophie mit Dominantem X-Chromosomalem Erbgang. Berlin, Akademie-Verlag, 1964. 111. Yoshimura, T., Satake, M., Ohnishi, A., et al.: Mutations of connexin32 in Charcot-Marie-Tooth disease type X interfere with cell-to-cell communication but not cell proliferation and myelin-specific gene expression. J. Neurosci. Res. 51:154, 1998. 112. Yum, S. W., Kleopa, K. A., Shumas, S., and Scherer, S. S.: Diverse trafficking abnormalities for connexin32 mutants causing CMTX. Neurobiol. Dis. 11:43, 2002.

77 Multiple Endocrine Neoplasia, Type 2B J. AIDAN CARNEY

Physical Appearance Face Skeleton

Alimentary Tract Ganglioneuromatosis Associated Neoplasms Medullary Thyroid Carcinoma

Multiple endocrine neoplasia type 2 (MEN 2) is an inherited multitumoral syndrome having three subtypes: MEN 2A (medullary thyroid carcinoma [MTC], pheochromocytoma, and parathyroid adenoma or hyperplasia); MEN 2B (MTC, pheochromocytoma, ganglioneuromatosis, and a variety of skeletal and connective tissue abnormalities); and familial MTC. The condition MEN 2B is included in this textbook because affected patients may have symptoms and signs of autonomic nerve dysfunction and because the phenotype has similarities to the syndrome of peroneal muscular atrophy (Charcot-Marie-Tooth disease). Patients with MEN 2B have a striking facies because of labial enlargement, apparent eyelid eversion, and a marfanoid habitus.4 The unusual appearance (Fig. 77–1) was described by Wagenmann14 more than 70 years ago and later meticulously detailed by Bazex and Dupré.3 However, it was not until 1966 that the significance of the physiognomy—namely, its association with two potentially lethal neoplasms (MTC and pheochromocytoma)—was recognized by Williams and Pollock.16 The designation “MEN 2B” arose as follows. By 1968, it was apparent that there were two distinct familial multiple endocrine neoplasia syndromes. The first, described by Wermer,15 was characterized by pituitary, parathyroid, and pancreatic islet tumors; the second, reported by Sipple,10 included MTC, pheochromocytoma, and parathyroid hyperplasia. To emphasize that these associations were genetically distinct (there had been some speculation that they might be related because of the occurrence of parathyroid disease in both), Steiner and co-workers12 sug-

Pheochromocytoma Neurologic Manifestations

gested that the Wermer and Sipple tumor combinations be designated “multiple endocrine neoplasia type 1 (MEN 1)” and “multiple endocrine neoplasia, type 2 (MEN 2),” respectively. Most patients with MEN 2 had a normal appearance, but a minority (those discussed in this chapter) had the unusual facial and body appearance mentioned earlier. To emphasize that the syndromes were genetically different, although they shared the same major endocrine tumors (MTC and pheochromocytoma), Sizemore and colleagues suggested that the disorders be designated “MEN 2A and 2B,” with the latter referring to the patients with the unusual appearance.5 MEN 2B has a worldwide distribution. About equal numbers of the cases are sporadic and familial. The disorder is recognizable at a young age because of its striking phenotype. It is due to germline mutation in the ret proto-oncogene, which is located in the pericentromeric region of chromosome 10, band q11.2.8 This gene encodes a member of the tyrosine kinase family of transmembrane receptors. The mutation involves substitution of a threonine for the methionine residue in exon 16 (codon 918). MEN 2B is transmitted as an autosomal dominant trait with parent-of-origin effects.2 Study of apparent de novo cases has shown that the new mutations are always of male parental origin.2 The sex ratio in both patients with de novo MEN 2B and the offspring of males transmitting MEN 2B is distorted from the expected. Also, there appears to be a different susceptibility of the ret gene to mutation in paternally and maternally derived DNA and a possible role for imprinting of the gene during development. 1805

1806

Diseases of the Peripheral Nervous System

FIGURE 77–2 Pedal abnormality in multiple endocrine neoplasia, type 2B. Pes cavus is common in the disorder and is sometimes accompanied by claw toes.

FIGURE 77–1 Facies of multiple endocrine neoplasia, type 2B. Features are elongation of the face, the appearance of eyelid eversion as a result of thickening of the tarsal plates, a broad nasal root, and nodular (upper) and diffuse (lower) thickening of the lips. (From Robertson, D. M., Sizemore, G. W., and Gordon, H.: Thickened corneal nerves as a manifestation of multiple endocrine neoplasia. Trans. Am. Acad. Ophthalmol. Otolaryngol. 79:772, 1975, with permission.)

PHYSICAL APPEARANCE Face Affected patients have a striking appearance produced by diffuse irregular thickening of their lips (bumpy lips), apparent eversion of the eyelids (creating a staring expression), dolichocephaly, and a broad nasal root (see Fig. 77–1). Disfiguring deformity of the lips and eyelids may require correction by plastic surgery.

Skeleton The syndrome includes a series of skeletal and connective tissue abnormalities that are very similar to those found in Marfan syndrome: excessive limb length, loosejointedness, scoliosis, and pectus excavatum.4 However, patients with MEN 2B do not exhibit the ectopia lentis and defective ascending aortic tunica media characteristic of Marfan syndrome. A high-arched palate and prog-

nathism are common. Redundancy and laxity of joint capsules, ligaments, tendons, and fasciae probably are responsible for hyperextensibility of joints, kyphoscoliosis, and hernias. Other common skeletal abnormalities are pes cavus (Fig. 77–2), talipes equinovarus, and genu valgum. Slipped femoral capital epiphysis requires surgical treatment.

ALIMENTARY TRACT GANGLIONEUROMATOSIS The fatal consequences of MEN 2B are usually due to metastatic MTC, less commonly to pheochromocytoma. Unfortunately, diagnosis of these neoplasms is usually not made in de novo cases until both have become well established. The thyroid neoplasm does not ordinarily cause symptoms or become palpable until it is already metastatic, and correct interpretation of the paroxysmal symptoms produced by the pheochromocytoma tends to be delayed. Therefore, it is very important to appreciate that patients with the syndrome have serious, chronic, and diverse alimentary tract manifestations, particularly severe constipation and megacolon, that antedate the symptoms of the endocrine neoplasms.3 These manifestations are usually the presenting complaints of affected patients. The cause of the gastrointestinal disorders is unknown; alimentary tract ganglioneuromatosis has been shown to be associated with poor contractility of the colon, and association between MTC and diarrhea is well known. Megacolon and diverticular disease of the colon may require surgical treatment.

1807

Multiple Endocrine Neoplasia, Type 2B

ASSOCIATED NEOPLASMS Medullary Thyroid Carcinoma Unilateral or bilateral thyroid masses located in the upper portion of the lateral lobes are present in more than half the patients. The tumor cells secrete calcitonin, and plasma levels of the substance are elevated. The neoplasm spreads by the lymphatic and vascular routes. Cervical nodal involvement is common at primary surgery. Blood-borne metastasis occurs to liver, lungs, and bone. The carcinoma is usually but not invariably aggressive11,13: approximately 20% of patients succumb to it. Therefore, it is important to recognize the abnormal phenotype to ensure prompt investigation for and treatment of MTC. Testing for the ret proto-oncogene provides an exquisitely sensitive molecular genetic diagnostic test for the tumor.8 Beginning at birth, patients thought to have MEN 2B should have testing for the RET mutation.9 Treatment of MTC is total thyroidectomy with central lymph node dissection; the parathyroid glands should be preserved.7 Locally recurrent tumor is usually treated surgically. There is no consensus on management of patients who have increased plasma levels of calcitonin after primary surgery but no clinically detectable disease. Only limited data are available on the effectiveness of chemotherapy for progressive metastasis. The patient’s primary relatives should be screened for the RET mutation and managed accordingly.

Pheochromocytoma In at least 50% of patients with MEN 2B, pheochromocytoma is manifested between the ages of 15 and 40 years.4 One half have the typical paroxysmal symptoms (headaches, sweating, palpitations, dizziness, nervousness, and tremor) and hypertension associated with catecholamine excess. The remainder are asymptomatic and normotensive, and the tumor or tumors are detected by measurement of urinary catecholamine content during routine screening in suspected instances. About 10% of the patients succumb to the neoplasm (as a result of hypertensive or hypotensive crisis, or cardiac arrhythmia). Patients with the syndrome should have measurement of 24-hour urinary excretion of total and fractionated catecholamines (epinephrine and norepinephrine) and their metabolites (vanillylmandelic acid and metanephrines). Tumor localization is accomplished by computed tomography, magnetic resonance imaging, and [131I]metaiodobenzylguanidine scan. In patients with symptoms, characteristically both adrenal glands are replaced by large multinodular pheochromocytomas. In asymptomatic patients, the abnormality may range from diffuse expansion of the medulla with preservation

of the shape of the gland (medullary hyperplasia) to small (⬍1 cm) medullary nodules with and without diffuse enlargement of the medulla. Treatment of the pheochromocytoma is bilateral total adrenalectomy and excision of any extra-adrenal paraganglioma. Bilateral total adrenalectomy is recommended because of the frequency of bilateral adrenal involvement, the possibility of recurrent pheochromocytoma after subtotal adrenalectomy, and the occasional occurrence of malignant disease.6

NEUROLOGIC MANIFESTATIONS In addition to the involvement of the autonomic nerve system (ganglioneuromatosis) already mentioned, somatic motor and sensory neurons may be affected in the syndrome.4 Four Mayo Clinic patients with the syndrome had examination of their nervous system. The results were varied: one patient was poorly muscled but had a normal neurologic examination; a second had diffuse weakness of all skeletal muscles and hypoactive deep tendon reflexes; a third had poorly developed musculature and distal muscle weakness; and the fourth patient had a sensory neuropathy affecting the lower limbs. Similarly inconsistent neurologic findings have been described in the literature, ranging from normal examination, through the presence of mild lower limb sensory abnormality, to a decrease in or absence of deep tendon reflexes. The nerves in the cubital fossa and neck were enlarged in two patients. Sural nerve biopsy in one case showed normal myelinated fibers and degeneration and regeneration of unmyelinated nerves, a process that was thought to be responsible for the increase in size of the nerve. Findings at postmortem examination indicate that certain symptoms can be attributed to a characteristic plaque of tissue composed of hyperplastic, interlacing bands of Schwann cells and myelinated fibers overlying the posterior columns of the spinal cord.6 There is no treatment for neural abnormality.

REFERENCES 1. Bazex, A., and Dupré, A.: Neuromes myéliniques muqueux a localisation centro-faciale et laryngée: neuromes des lèvres, de la langue, des paupières, des narines, du larynx (Entité nouvelle?). Ann. Dermatol. 85:613, 1958. 2. Carlson, K. M., Bracamontes, J., Jackson, C. E., et al.: Parent-of-origin effects in multiple endocrine neoplasia, type 2B. Am. J. Hum. Genet. 55:1076, 1994. 3. Carney, J. A., Go, V. L. W., Sizemore G. W, and Hayles, A. B.: Alimentary tract ganglioneuromatosis: a major component of the syndrome of multiple endocrine neoplasia, type 2b. N. Engl. J. Med. 295:1287, 1976. 4. Carney, J. A., Sizemore, G. W., and Hayles, A. B.: Multiple endocrine neoplasia, type 2b. Pathobiol. Ann. 8:105, 1978.

1808

Diseases of the Peripheral Nervous System

5. Chong, G. C., Beahrs, O. H., Sizemore, G. W., et al.: Medullary carcinoma of the thyroid gland. Cancer 35:695, 1975. 6. Dyck, P. J., Carney, J. A., Sizemore, G. W., et al.: Multiple endocrine neoplasia, type 2b: phenotype recognition, neurological features and their pathological basis. Ann. Neurol. 6: 302, 1979. 7. Giuffrida, D., and Gharib, H.: Current diagnosis and management of medullary thyroid carcinoma. Ann. Oncol. 9:695, 1998. 8. Heshmati, H. M., Gharib, H., Khosla, S., et al.: Genetic testing in medullary thyroid carcinoma syndromes: mutation types and clinical significance. Mayo Clin. Proc. 72:430, 1997. 9. Heshmati, H. M., Gharib, H., van Heerden, J. A., and Sizemore, G. W.: Advances and controversies in the diagnosis and management of medullary thyroid carcinoma. Am. J. Med. 103:60, 1997. 10. Sipple, J. H. The association of pheochromocytoma with carcinoma of the thyroid gland. Am. J. Med. 31:163, 1962.

11. Sizemore, G. W., Carney, J. A., Gharib, H., and Capen, C. C.: Multiple endocrine neoplasia type 2B: eighteen-year follow-up of a family. Henry Ford Hosp. Med. J. 40:236, 1992. 12. Steiner, A. L., Goodman, A. D., and Powers, S. R.: Study of a kindred with pheochromocytoma, medullary thyroid carcinoma, hyperparathyroidism and Cushing’s disease: multiple endocrine neoplasia, type 2. Medicine (Baltimore) 47:371, 1968. 13. Vasen, H. F. A., van der Feltz, M., Raue, F., et al.: The natural course of multiple endocrine neoplasia type 11b: a study of 18 cases. Arch. Intern. Med. 152:1250, 1992. 14. Wagenmann, A.: Multiple neurome des auges und der zunge. Ber. Dtsch. Opthalmol. Ges. 43:282, 1922. 15. Wermer, P.: Genetic aspects of adenomatosis of endocrine glands. Am. J. Med. 16:363, 1954. 16. Williams, E. D., and Pollock, D. J.: Multiple mucosal neuromata with endocrine tumors: a syndrome allied to von Recklinghausen’s disease. Pathol. Bacteriol. 91:71, 1966.

I. Hereditary Sensory and Autonomic Neuropathies

78 HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics CHRISTOPHER J. KLEIN AND PETER J. DYCK

Overview Indifference to Pain Historical Overview of HSAN Classification Autosomal Dominant HSAN (Type I) Differential Diagnosis Inheritance Distal Lower Limb Sensory and Autonomic Neuropathy ⫹ Acral Mutilation (“Typical”)

HSAN I with Acral Mutilation and Peroneal Weakness ⫾ Lancinating Pain ⫾ Hearing Loss ⫾ Hyperhidrosis (SPTLC1) HSAN I with Acral Mutilation and Peroneal Weakness (RAB7) Autosomal Dominant Forms of HSAN without Gene Discovery Autosomal Recessive HSAN Forms HSAN Type II: “HSN2” Gene and Theoretical Protein

OVERVIEW Hereditary sensory and autonomic neuropathies (HSANs) are inherited disorders of peripheral sensory and autonomic neurons (axons), which need to be distinguished from inherited neuropathies of large primary afferent neurons (spinocerebellar degeneration) on the one hand and from inherited disorders with prominent motor involvement (hereditary motor and sensory neurons) on the other hand. They are associated with a range of clinical presentations, pathologic alterations, electrophysiologic abnormalities, and increasingly specific biochemical or molecular genetic abnormalities. A length-dependent process affecting the lower limbs especially is typical of the dominantly inherited varieties, but a more generalized involvement occurs especially in recessively inherited disorders with congenital-onset

HSAN Type III (Familial Dysautonomia): IKBKAP HSAN Type IV: TRKA HSAN Type V: Possibly TRKA and NGFB Other Manifestations of Congenital HSAN Mechanisms and Treatment of Acral Mutilation and Arthropathy

varieties. Sensory loss, especially of pain and thermal sensation, is frequently associated with ulceration of the feet and hands with mutilation or amputations, whereas others may have prominent pain without marked sensory loss. Variable expression, especially in terms of severity, is typical. Among the most severely affected patients, typically those with HSAN of congenital onset, central nervous system involvement may be severe and may include mental retardation with premature death. Individuals with HSAN typically have greater morbidity and mortality than their hereditary motor and sensory neuropathy (HMSN) counterparts, described in Part H of this book. Although clinically heterogeneous, five combined clinical features set HSAN apart from other neuropathies and system atrophies. The characteristic features are (1) mendelian inheritance, (2) selective or predominant peripheral primary 1809

1810

Diseases of the Peripheral Nervous System

sensory ⫾ autonomic neuron (axon) involvement, and (3) failure to thrive or atrophy of affected neurons (axons). Not included with HSAN are genetic disorders with selective involvement of predominately large-diameter afferent neurons (axons) ⫾ spinocerebellar and cerebellar neurons, and genetic disorders affecting multiple non-neurologic tissues and small sensory fibers (e.g., ␣-galactosidase [Fabry’s disease] ␣-lipoproteinemia [Tangier disease], intermittent porphyria, amyloidosis, and many other inborn errors of metabolism). In individual cases, it is sometimes difficult to differentiate acquired chronic sensorimotor neuropathies from HSAN. Also, HSAN patients may be susceptible to developing neuropathic impairments from toxic, inflammatory, metabolic, and other acquired causes (i.e., diabetes, alcohol, or nutritional disorders); it is not uncommon for the primary neuropathic insult to wrongly be attributed to these other conditions. Family history is emphasized in arriving at the correct diagnosis. Figure 78–1 shows an algorithm of the steps that may be helpful in differential diagnosis. Increasingly, molecular biology is providing for improved understanding of the HSAN disorders, and currently defects in at least six genes are known to be causative of clinical varieties of HSAN: Defect in SPTLC1, the gene for serine palmitoyltransferase long chain subunit 1, and RAB7, the gene for a late endosomal GTPase protein, which cause HSAN type I (HSAN I) Defect in a gene tentatively named “HSN2”, which causes HSAN type II (HSAN II) Defect in IKBKAP, the gene for inhibitor of kappa light polypeptide gene enhancer in B cells, kinase complex–associated protein, which causes HSAN type III (HSAN III), usually called Riley-Day syndrome or familial dysautonomia Defect in TRKA, the gene for tyrosine receptor kinase A, which causes HSAN type IV (HSAN IV) and possibly HSAN type V (HSAN V) Defect in NGFB, the gene for nerve growth factor (NGF)-␤, which causes HSAN V Other distinct clinical forms of HSAN are predicted to have separate genetic causes, and as with HMSN, similar phenotypes may have different genetic causes (e.g., defects in SPTLC1 and RAB7 both cause HSAN I). The current molecular data support the diversity of histopathologic features seen and appear to support the earlier pathologic observations that the fundamental pathogenic abnormality must lie proximally in or near the soma of spinal ganglion neurons despite prominent distal phenotypic involvement (i.e., a “dying-back” phenomenon).35,79 The genetic data have provided strong evidence that these disorders are distinct from varieties of HMSN, although rarely patients with HMSN mutations such as PMP22 duplications (HMSN type IA) can have prominent small fiber sensory involvement.6,172

FIGURE 78–1 A diagnostic algorithm that might be helpful in differentiating the disorders of patients with predominant involvement of peripheral sensory neurons. (Modified from Dyck, P. J., and Klein, C. J.: Hereditary sensory and autonomic neuropathies. In Noseworthy, J. D. [ed.]: Neurological Therapeutics: Principles and Practice. London, Martin Dunitz, p. 2188, 2003.)

Despite the advances in molecular characterization, it remains important to carefully classify patients phenotypically. Currently, no gene therapies are available for varieties of HSAN, and selecting the correct genetic tests relies on the clinical context, including the particular sensory fibers affected, electrophysiologic testing, age of onset, associated clinical features, and the inheritance pattern. Although cutaneous sensory sural nerve biopsy is not required for all cases, it may be important for autosomal recessive varieties without clear family history, so as to recognize absence of fiber classes and to distinguish “indifference to pain” from insensitivity to pain (the disorders discussed here). Also, the nerve biopsy may allow for identification of interstitial abnormalities (e.g., microvessel disease, including vasculitis, amyloidosis, and others); such process can be primary in the pathogenesis or additive to disability or impairments in inherited diseases. The presence of significant mononuclear (lymphocyte) inflammatory infiltrates in endoneurial, epineurial, or perineurial tissues is not typical of HSAN and

1811

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

may justify an empirical treatment trial with immunosuppression. However, it should be appreciated that inherited neuropathies may also have an immune component. It is also likely that nerve tissue obtained will improve molecular understanding of the complex factors important in the ultimate disease expression in HSAN. Specifically, expression studies of diseased nerves will be important in understanding these disorders. For example, the varied severity seen both within and among different kindreds with the same or similar mutations emphasizes the importance of such unidentified molecular influences. Identifying those factors will be important in clinical treatment given the inherent difficulties in gene therapy. To date, the identified genetic defects have been associated with characteristic modes of inheritance, ages of presentation (congenital or adult onset), and patterns of involvement, including the specific sensory fiber types predominately affected (i.e., A␣, A␦, drC, sC) (see Table 78–1). Therefore, careful bedside and laboratory sensory testing is needed in characterizing the nature and distribution of sensory involvement. Autonomic testing with pulse response variability to deep breathing and to the Valsalva maneuver, tests for postural hypotension, quantitative sudomotor autonomic reflex screening, and thermoregulatory sweat testing are also helpful in diagnosing patients. The testing allows for the identification of postganglionic sudomotor sweating abnormalities and cardiovagal autonomic involvement, which typically are not present in forms of HMSN. The main emphasis in treatment of patients with HSAN is teaching parents and affected children about the possible bad outcomes that may result because of lack of protective

sensation and sudomotor defect. Knowing this, limb injury and acral mutilation can be minimized. Avoidance of injury, use of visual clues when sensory clues are not available, and special care of feet and footwear may prevent acral mutilation.

INDIFFERENCE TO PAIN There are rare persons who develop painless injuries despite having normal electrophysiologic tests (nerve conduction tests, autonomic tests) and cutaneous sensory nerve histopathology. The original characterization of these individuals was that of an “indifference to pain.” The concept of indifference to pain probably was based on Schilder and Stengel’s notion of an asymbolia for pain.148 Unfortunately, the distinction between an indifference to pain and a sensory neuropathy, as elucidated by Ogden and co-workers130 and by Winkelmann and associates,190 has not always been made. Particularly Ogden and co-workers,130 and others before them, made the point that patients with indifference to pain perceived stimuli correctly and at normal thresholds but failed to react in the usual defensive manner by withdrawing the offended part and failed to show the usual autonomic alterations such as pulse, respiration, and blood pressure acceleration. In an extensively studied case of this disorder, first reported by McMurray,119 Feindel found no abnormality of cutaneous nerves,57 and Baxter and Olszewski11 found no abnormality of cutaneous nerves, spinal cord, or thalamus. Unfortunately, with the techniques they employed, an abnormality of peripheral

Table 78–1. Hereditary Sensory and Autonomic Neuropathy (HSAN) Neurons (Axons) Disease

Onset

Inheritance

A␣

A␦

C

Sudomotor

Type I*

2nd ⫹ decade

AD

⫹

⫹⫹

⫹⫹

LS⫹

Type II Type III† Type IV

C C C

AR AR AR

⫹⫹ ⫹⫹ N

⫹⫹ ⫹⫹ ⫾

⫹ ⫹⫹ ⫹⫹

G G G

Type V

C

AR

N

⫹⫹

⫾

Loci

Gene

Function

9q22

SPTLC1

Sphingolipid biosynthesis

3q13 12p13 9q31 1q21

RAB7 “HSN2” IKBKAP TRKA

1q21‡

TRKA‡

Axonal transport Unknown Unknown Neurotrophin receptor Neurotrophin receptor

1p13

NGFB

N Nerve growth factor

*Many autosomal dominant forms without genetic cause (see text). † Ashkenazi Jews. ‡ A single patient reported. ⫹ ⫽ affected; ⫹⫹ ⫽ severely affected; ⫾ ⫽ may be affected; AD ⫽ autosomal dominant; AR ⫽ autosomal recessive; C ⫽ congenital; G ⫽ generalized; IKBKAP ⫽ inhibitor of kappa light polypeptide gene enhancer in B cells, kinase complex–associated protein; LS⫹ ⫽ lumbosacral plus; N ⫽ normal; NGFB ⫽ nerve growth factor-␤ gene (prominent C-fiber loss); SPTLC1 ⫽ serine palmitoyltransferase long chain subunit 1; TRKA ⫽ tyrosine receptor kinase A gene.

1812

Diseases of the Peripheral Nervous System

unmyelinated fibers could readily have gone undetected. There appear to be patients who have selective decrease or absence of the pain response yet appear to have pain reception and normal numbers of small myelinated (A␦) and unmyelinated (drC) fibers. Landrieu and co-workers,110 utilizing modern techniques, described a mother and a 5-year-old daughter (suggesting autosomal dominant inheritance) who had indifference to pain and no abnormality of peripheral nerve function. Number and size distribution of myelinated and unmyelinated fibers of the superficial peroneal nerve were normal. These findings led them to conclude that “true indifference to pain does exist and is not a variety of HSAN.” Within the literature some cases initially labeled as “indifference to pain” with dysautonomia were later identified to have cutaneous sensory histopathologic abnormalities.24,142 These patients were able to discriminate sharp from dull stimuli but did not experience pain and are now better classified as having forms of HSAN (see below). Because conventional cutaneous surface nerve conduction studies do not provide information about small myelinated (A␦) and unmyelinated (drC and sC) fibers of cutaneous or mixed nerves, cutaneous nerve excision and study has been previously employed to assess nerve electrophysiology in vitro (Fig. 78–2). This in vitro approach identifies abnormality of other classes of fibers than is possible using in vivo nerve conduction studies. The in vitro studies have not been reported in cases of

indifference to pain and would be helpful in assessing a histopathologic bias of abnormality. The causes of indifference to pain as defined by normal cutaneous nerve morphometrics and electrophysiology, including sensory testing, are likely broad: (1) central nervous system dysfunction, including an asymbolia for pain; (2) incomplete or nonstructural peripheral nerve lesion (i.e., neurochemical dysfunction without the characteristic features of fiber loss or degeneration); and (3) a proximal sensory abnormality, for example, dysfunction of secondorder pain fibers or associative central nervous system fibers. That a proximal lesion can cause a loss of sensation without showing abnormalities of nerve conduction or of peripheral nerve morphology is illustrated by previously reported cases with severe sensory loss resulting from abnormality of the posterior roots.155 These patients had chronic sensory loss with severe ataxia typically sparing small sensory fibers, with normal sensory nerve conduction studies and sural nerve biopsy. The localization to sensory roots was determined by root biopsy, lumbosacral magnetic resonance imaging, and somatosensory evoked potentials. This condition is likely a form of chronic inflammatory immune polyradiculopathy because patients have inflammation on root biopsies, albuminocytologic dissociation in spinal fluid, and frequently dramatic response to treatment. It is unlikely that such patients would be confused with those with indifference to pain given the prominence of large fiber sensory involvement, but prominent proximal

FIGURE 78–2 Action potentials of human sural nerves in vitro in HSAN II. From top to bottom, the records are from the father (case 5, unaffected), case 2, case 3, case 4, and case 1. The nerve from the father has well-developed A␣ (␤), A␦, and C fiber potentials. In the other nerves, no action potentials were elicited from myelinated fibers. Only the shock artifact is evident in the records. The C fiber potentials in these four nerves were decreased in amplitude and their latency was prolonged (From Ohta, M., Ellefson, R. D., Lambert, E. H., and Dyck, P. J.: Hereditary sensory neuropathy, type II: clinical, electrophysiologic, histologic, and biochemical studies of a Quebec kinship. Arch. Neurol. 29:23, 1973, with permission. Copyright 1973, The American Medical Association.)

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

small fiber involvement could also occur and be missed by distal histopathologic and electrophysiologic studies. These patients might best be characterized as having an “insensitivity to pain” that is difficult to diagnose. The potential of central nervous system dysfunction or emotional neglect of pain has been reviewed in the earlier edition of this text.43 Here we make several additions. The label “indifference to pain” implies cerebral inattention or other cognitive or emotional dysfunction. Some patients may have altered pain perception as a result of stagnant structural, neoplastic, or ischemic disease, preventing proper integration within limbic or cortical projections. Such defects may also occur globally, especially in some cases of mental retardation.43 Within the psychiatric literature are examples of schizophrenics and their first-degree relatives being afflicted by extreme tolerance of painful noxious stimulus and painless injury. The data have raised the possibility of indifference to pain as a marker of potential psychosis among affected families. Cited examples include individuals with perforation of bowel,146 broken bones, dental abscess, and acral burns having not perceived such events as painful.82 One of the authors (C.J.K.) cared for a young man with schizophrenia who, in a psychotic state, jumped from a four-story building and denied pain from his multiple injuries. His jump was witnessed, but he was not apprehended; he then climbed to the top of the same building and jumped again. It was only after his second jump, at which time he broke both legs, that he was apprehended. The sympathetic features of pain (i.e., tachycardia and hypertension) were absent, and he denied pain at the scene or in the postoperative setting with internal fixation of his multiple fractures. He did not utilize the intravenous patient-controlled analgesia provided him postoperatively. Both during and with resolution of his psychosis on neuroleptics, he could discriminate the different sensory modalities but had no response to pain as such. His family commented that in childhood he had a “high tolerance to pain.” His remaining neurologic examination was normal. The case sufficiently illustrates the problem of pain insensitivity among schizophrenics that, as noted in the surgical literature, concern regarding misdiagnosis in schizophrenic patients given the absence of pain has altered practice.15 Known abnormalities of central dopaminergic neurons potentially alter central processing of pain, but the mechanisms are not understood. These patients should be easily distinguished from those with HSAN because the losses of pain sensitivity in schizophrenics tend to be generalized and alteration of affect and emotional tone are usually obvious, although the fact that first-degree relatives with no history of psychosis can have heightened pain tolerance would bring this into question.82 Many of the earliest cases of indifference to pain simply were inadequately studied to state whether they had a neuropathologic basis. Such cases can now be studied using quantitative sensory examination (quantitative sen-

1813

sation testing; see Chapter 42), quantitative autonomic evaluation, and electrophysiologic and neuropathologic approaches.

HISTORICAL OVERVIEW OF HSAN CLASSIFICATION Dyck, Ohta, and co-workers41,132 suggested that there were at least four varieties of hereditary sensory neuropathies (HSNs): a dominantly inherited variety (HSN I); a congenital sensory neuropathy (HSN II); familial dysautonomia, a congenital sensory and autonomic neuropathy (HSN III); and a congenital insensitivity to pain, also a congenital sensory and autonomic neuropathy (HSN IV). In 1984 they suggested that, since autonomic involvement was present in HSN, the group should be designated HSAN. A fifth HSAN was later added to the classification (HSAN V).42 Reviewing their experience with 13 patients with congenital sensory and autonomic neuropathy, Axelrod and co-workers10 suggested the following varieties: congenital sensory neuropathy (HSN II), familial dysautonomia (HSN III), congenital sensory neuropathy with anhidrosis (HSN IV), sensory neuropathy with skeletal dysplasia, progressive paneuropathy with hypotonia, and congenital autonomic dysfunction with universal “insensitivity to pain.” Reporting on a Kashmiri family with three members affected by a congenital sensory and autonomic neuropathy and corneal opacification, Donaghy and co-workers38 proposed the following classification: I. Autosomal dominant inheritance 1. Dominantly inherited sensory neuropathy35 (HSAN I*) 2. Dominantly inherited sensory neuropathy with spastic paraplegia26 II. Autosomal recessive inheritance 1. Recessively inherited sensory neuropathy 132 (HSAN II) 2. Hereditary anhidrotic sensory neuropathy 142 (HSAN IV) 3. Familial dysautonomia141 (HSAN III) 4. Hereditary sensory neuropathy with neurotrophic keratitis38 5. Recessively inherited sensory neuropathy with spastic paraplegia26 III. X-linked recessive inheritance 1. X-linked recessive sensory neuropathy95 IV. Uncertain status 1. Hereditary sensory neuropathy with dysautonomia128 2. Congenital sensory neuropathy with selective loss of small myelinated fibers49,114 (HSAN V) *Parenthetical notations indicate the classification of Dyck et al.49

1814

Diseases of the Peripheral Nervous System

Classification of inherited neuropathies with predominant involvement (failure to develop or degeneration) of sensory neurons (axons) tends to divide them into three patterns of neurologic involvement: (1) varieties with predominant involvement of lower limbs, usually with onset in the second to fourth decades and slow progression (HSAN I); (2) more generalized involvement and congenital onset without evident progression (HSAN II through V); and (3) spinocerebellar degenerations. The current classification that we believe is most accurate combines the various clinical features of natural history, mode of inheritance, electrophysiologic characteristics, and the neuropathologic alterations, including histopathologic descriptions, with the genetic abnormalities (see Table 78–1). This approach seems most accurate because molecular genetic testing without clinical correlation is not as useful as desired. The details of the clinical, historical descriptions and molecular biologic discoveries are provided below for each class of HSAN, and it is likely that with time the classifications will continue to evolve as more is learned.

AUTOSOMAL DOMINANT HSAN (TYPE I) Although prospective epidemiologic studies are not available, the dominant HSAN forms are clearly the most common. We assume that HSAN I will be found to be due to multiple molecular genetic abnormalities (i.e., different disorders). Despite significant clinical variability, the conditions share the following characteristics: 1. Onset of neuropathic symptoms and deficits is typically in the second or later decades, with a slow progressive course thereafter. 2. Sensory symptoms and deficits or associated tissue complications (Fig. 78–3)23,66,79,141,177 overshadow autonomic and motor manifestations, but both are common. 3. There is symmetry (less than 25% variability between sides). 4. Feet (and legs) are generally much more affected than hands (and arms), with little or no involvement of proximal limbs, trunk, and head and neck. 5. The clinical neurophysiologic and neuropathologic features are those of chronic neuronal (axonal) atrophy, myelin remodeling, and axonal degeneration. 6. The central nervous system is usually not affected. 7. Only neural tissues are affected. The varied presentations include the following symptoms or signs: peroneal muscle atrophy or weakness,7,8,34,35,46,55,135 hearing loss,43,72,79 dementia,43 lancinating pain,35,43,79 burning feet,43 gastroesophageal reflux,104 neurogenic arthropathy,43 restless legs,164 and likely others. The names of the syndromes include these clinical features and the genetic abnormality when available.

Historically, many names have been given to this group of disorders: “mal perforant du pied,”69 “singular affection of os pied,”125 “familial symmetric gangrene with arthropathy,”23 “familial neural trophic disturbances of lower extremity,”65 “hereditary perforating ulcers,”150 “lumbosacral syringomyelia,”177 “ulcerative and mutilating acropathy,”169 “hereditary sensory radicular neuropathy,”35 “familial and sporadic neurogenic acro-osteolysis,”59 “acrodystrophic neuropathy,”162 and HSN I50 or HSAN I.42 For a detailed description of the historical accounts of these disorders, see earlier edition of this text.43 Unfortunately, for many of these earliest patients described, no tissue is available for molecular investigation.

Differential Diagnosis HSAN I must be separated from other varieties of HSAN. Three points help to distinguish HSAN I from HSAN II through V and other congenital HSANs. First, HSAN I usually begins in the second or third or even later decades, whereas other varieties probably begin before birth (congenital). Second, HSAN I is a slowly progressive clinical disorder, whereas in the other HSANs such progression is not demonstrated. Third, HSAN I affects lower limbs preferentially, whereas in the others lower and upper limbs and the trunk are variably affected. Spinocerebellar degenerations differ from HSAN I by three major features: (1) predominant involvement of kinesthetic and mechanoreceptor sensory loss, (2) cerebellar ataxia, and (3) no or only trivial involvement of small fiber sensory and autonomic dysfunction (except late in the course of the disorder). In inherited amyloidosis, sexual and sphincter dysfunctions often develop—not a characteristic feature of HSAN I. In intermittent porphyric neuropathy, the onset is much more acute and associated with delirium or other mental symptoms. The separation of HSAN I from HMSN type II (HMSN II) may be difficult. Fundamentally, sensory symptoms and deficits in HSAN I overshadow motor and autonomic ones, whereas in HMSN II, motor symptoms and deficits overshadow sensory ones. In some kindreds, however, the sensory and motor symptoms and deficits are approximately equivalent. These observations raise the issue of whether the separation of HSAN I and HMSN II is real. The identification of separate genetic defects for most cases supports the idea that HSAN I and HMSN II are separate genetic and clinical syndromes.

Inheritance The original and subsequent reports of inheritance showed a clear autosomal dominant pattern in most kindreds.13,33,38,46,56,67,79,92,98,140,143,177 Other kindreds had

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

1815

FIGURE 78–3 A, Slight degree of pes cavus and hammer toes in a 37-year-old woman with HSAN I. Although she had sensory loss similar in kind to that of her father and brothers, she did not have mutilating acropathy, perhaps mainly because she has not sustained injury of the feet and has been careful to treat injury. B, Marked degree of mutilation of the feet of a brother of the patient in A. Injury of the feet, neglect of injury, and inadequate care of the feet led to plantar ulcers and osteomyelitis, and osteolysis resulted. C, Roentgenograms of the feet of the patient in B.

possible autosomal recessive23,125,151 and possibly X-linked99 inheritance, but to our knowledge DNA analysis of the currently known genes (SPTLC1 and RAB7; see below) for these kindreds has not been reported to confirm the genetic cause and hence pattern of inheritance.

Regardless of whether different modes of inheritance exist, HSAN I is genetically heterogeneous. The quite different clinical presentations and the work noted below support the concept of genetic heterogeneity, as previously proposed.43

1816

Diseases of the Peripheral Nervous System

Distal Lower Limb Sensory and Autonomic Neuropathy ⫹ Acral Mutilation (“Typical”) Although many of the original HSAN I kindreds described do not have a reported genetic defect, the combined reports parallel the clinical and electrophysiologic findings seen in those with a genetic cause identified (see below). Schultze150 described a brother and two sisters who developed perforating ulcers of the feet and loss of sensation of the feet. Spontaneous pain and muscle weakness were not reported. Knee and ankle reflexes were absent. The parents were not evaluated, but a history of perforating ulcer was not reported. Göbell and Runge65 reported nine affected persons in two generations who had perforating plantar ulcers of the feet that discharged spicules of bone. Lancinating pains and deafness were not reported. In a family studied by Bruns,23 four of five siblings had plantar ulcers. Van Epps and Kerr177 reported a kindred with lumbosacral syringomyelia but who probably had HSAN I. In the “J” family, inheritance was from a grandmother to a son and to children. Affected persons had pes cavus, plantar ulcers, sensory loss, and shortening of feet, presumably as a result of bone loss through foot ulcers from osteomyelitis. Mild symmetrical leg weakness was reported. Spontaneous pain was not a feature—pain was reported only when the ulcer was enlarging. In a large kindred, described first by Tocantins and Reimann,174 again by Reimann and colleagues,140 and later by Whitaker and co-workers,188 affected members had plantar ulcers of feet and legs and sensory loss. Because direct transmission from generation to generation and from male to male was documented, autosomal dominant inheritance was reported. Whitaker and co-workers188 reported that typically the disorder began after childhood and adolescence. In young adults, hypalgesia and thermohypesthesia developed insidiously, followed by painless burns or injury. At later times hands might become affected. Spontaneous pain and hearing loss were not reported. Paresthesia was not a characteristic feature of this kinship’s disorder. Osseous necrosis was attributed to loss of bone spicules through ulcers. Clinical Features Typically the disorder is manifested in one of four ways: (1) foot complications, (2) spontaneous pain, (3) sensory and autonomic neuropathy, or (4) kinship evaluation. Foot complications,61,69,169,171 such as plantar ulcer, recurring paronychia of toes, stress fractures of foot bones,99 recurring cellulitis, osteomyelitis, and resorption of the foot bones,23,65 are a prominent feature in some kindreds (see Fig. 78–3). Pes cavus, when present, appears to be a risk factor for plantar ulcers. Typically plantar ulcers develop over metatarsal heads in patients with high arches, but they may develop at other pressure

points, such as the convexity of curled toes. The ulcer often develops at the site of a previous callus. Neglected ulcers do not heal and may become associated with recurring cellulitis, lymphangitis, and osteomyelitis.46 With chronic osteomyelitis, foot bones may gradually shorten and eventually disappear,23,46,65 by extrusion of sequestered bone from the ulcer. When this happens, the foot progressively shortens and disappears, leaving remnants of toes in a badly deformed stump. In one kindred we studied, the feet were so severely affected in some members that patients crawled on their knees. The smell of their sores was offensive, a situation that had been tolerated for years. Pain may be of two types: spontaneous or related to local events. Spontaneous pain is of three types: burning, aching, or lancinating. Burning or aching pain typically occurs in the feet, is worse with excessive use or weight bearing, and tends to be less at night. Sometimes it is worse with heat. Lancinating pain, to a severe degree, is a feature in only a small number of kindreds with HSAN I, but in these it may be disabling. It may occur in feet and legs, in thighs, and in shoulders. Individual jabs radiate for varying distances anatomically, are usually in deep tissues, and persist for 2 to 8 seconds. Usually lancinating pains have directionality— into or above a limb but usually only for short distances. Lancinating pains tend to occur in clusters and often recur in the same anatomic location. Local discomfort of the foot at pressure points, such as the region of the callus or ulcer over a prominent metatarsal head, may be the only discomfort of the disorder in affected persons. In other patients, such discomfort occurs only with local infection. Persistent paresthesia, or the sensation of being “asleep” (like a hand gone “asleep” from lying on it), is not a characteristic feature of HSAN I; in fact, its presence suggests another or a superimposed cause (usually acquired) of the neuropathy. The neurologic findings consist of changes in sensation, reflexes, autonomic function, and muscle strength. All modalities of sensation may be affected, but the most characteristic ones are loss of pain and thermal sensation. Joint position sense may be affected in some kindreds. The decreased sensation is over the feet and distal legs and in some cases over the fingers and hands. A selective loss of pain and temperature sensation with preserved tactile sensation over distal lower limbs was noted by some early observers.23,35,65,79,92,107,177 Because only simple hand-held instruments were used, mild tactile loss, given that it was present, might not have been detected. Thévenard reported three patterns of sensory loss: (1) all modalities of sensation affected (19 cases), (2) pain and temperature sensation loss with preservation of touch-pressure sensation, and (3) pansensory loss in acral parts and a dissociated loss in proximal limbs.170 Dissociated sensory loss was reported to be a stage in development of global loss.170

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

O’Brien and co-workers described a patient with severe impairment of nociception in feet and hands but with preserved thermal sensation.129 In a typical case of this disorder, we found that touchpressure threshold, thermal discrimination, and nociception all were elevated (abnormal) in the foot. Likewise, in the compound action potential of the sural nerve in vitro, the A␣, A␦, and C potentials were greatly decreased in amplitude (Fig. 78–4). As compared to control nerves, the C potential was perhaps most affected, followed by the A␦ potential and then by the A␣ potential. Counting the different size classes of fibers showed that the greatest loss was of unmyelinated fibers, then of small myelinated fibers, and then of large myelinated fibers (Fig. 78–5). In other patients with HSAN I, we have made similar observations—all fiber classes were affected, but small fibers appeared to be especially affected.132 Sensory loss is more affected in the feet and legs23,65,79 than in the arms or hands. In reevaluating one patient of the kinship described by Hicks,79 Denny-Brown found loss of pain sensation in the hands.35 No persons with the features of HSAN type I in the reports just mentioned had ulceration of the fingers. Such ulceration was not included in subsequent reports,34,61,92,134,176,183 in kinship “R” of Spillane and Wells,162 or in the HSAN type I kinship of Ohta and co-workers.132 Considering the severity of the fiber loss, it is perhaps surprising that these patients do not

FIGURE 78–4 Compound action potentials of sural nerve from a healthy person and from patients with Friedreich’s ataxia, HSAN I, and tabes dorsalis. Note the great decrease in amplitude of A␣ potential in Friedreich’s ataxia, the decrease in amplitude of all three peaks in HSAN I (greatest in C potential, less in A␦, and least in A␣), and the lack of abnormality in tabes dorsalis. (From Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerve in health, Friedreich ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb, W. A. [ed.]: Handbook of Electroencephalography and Clinical Neurophysiology, Vol. 9. Amsterdam, Elsevier, p. 83, 1971, with permission.)

1817

have ataxia. This may suggest that fibers from muscle and tendon receptors tend to be spared in this disorder, but the question has not been critically studied morphologically. A decrease in the Achilles tendon reflex is frequent in this disorder, probably followed next in frequency by a decrease in the quadriceps reflex.79,177 Denny-Brown also reported loss of reflexes in the upper limbs.35 An occasional author has reported single cases with spasticity, a mild increase in some tendon reflexes, and extensor toe signs.79 Clinically apparent nerve enlargement has not been noted, although Pallis and Schneeweiss reported a greatly enlarged sciatic nerve with interstitial fluid edema, many demyelinated fibers, and loss of myelin and of axis cylinders.134 Loss of sweating in the distal lower limbs is typical. As in other dominantly inherited neuromuscular disorders, the severity may be extremely variable within a kinship and between kinships. The neuropathic disorder is insidious in onset and very slowly progressive. The complications, by contrast, may begin suddenly and be overwhelming. Pathologic Alterations Jughenn and co-workers were the first to show that HSAN I was not due to syringomyelia.98 They found unequivocal abnormalities of peripheral nerves: foot nerves were almost without myelinated fibers and had increased numbers of Schwann cell nuclei; other skin nerves also had reduced numbers of fibers, thickened perineuria, and discontinuity

1818

Diseases of the Peripheral Nervous System

FIGURE 78–5 Histograms of myelinated and unmyelinated fibers of sural nerve of a 33-year-old man without nerve disease (A) and a 32-year-old man with HSAN I (B). Note markedly decreased numbers of unmyelinated, small myelinated, and large myelinated fibers, in order of decreasing severity. (Modified from Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerve in health, Friedreich ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb, W. A. [ed.]: Handbook of Electroencephalography and Clinical Neurophysiology, Vol. 9. Amsterdam, Elsevier, p. 83, 1971.)

of myelin. Muscles revealed changes typical of nerve degeneration. They stated that these pathologic findings placed this condition with the heredodegenerative disorders of peripheral neurons. Although Denny-Brown is usually credited with the observation that HSAN is caused by spinal ganglion neuron degeneration, Jughenn and associates’ clear description and conclusion clearly preceded Denny-Brown by several years. Denny-Brown, studying the postmortem examination tissue of an affected person from the family studied by Hicks,79 found marked alterations in posterior columns, posterior roots, spinal ganglia, and peripheral nerves.35 He considered the degeneration of the spinal ganglion neurons to be primary, hence representing a hereditary sensory radicular neuropathy. In the posterior column, there was loss of fibers; the Flechsig posterior root entry zone was especially affected, and there was loss of fibers in the dorsal roots and in peripheral nerves. There was a decrease in number of fibers in sections stained for myelin and axis cylinders. The spinal ganglion changes consisted of degenerative pathologic alterations

and of a decrease in number of perikarya. Degenerated cells were replaced by residual nodules of Nageotte. An additional feature, not subsequently confirmed, was the deposition of amyloid in the ganglia. More recently studies have focused principally on the class of fibers affected and pathologic alterations of sural nerve. As shown in Figure 78–4, the compound action potentials were decreased in amplitude but the C potential was most decreased, followed by the A␦ potential and last by the A␣ potential. Assessing the numbers and sizes of myelinated fibers and unmyelinated fibers of this nerve showed that all fiber classes were decreased in number (see Fig. 78–5). Fiber number was more decreased at the ankle (myelinated, 950/mm2, vs. unmyelinated, 8306/mm2) than at the midcalf level (myelinated, 2304/mm2, vs. unmyelinated, 11,368/mm2) (Figs. 78–6 and 78–7). The decrease in density of myelinated fibers between midcalf and ankle levels is greater than in normal nerves. Unmyelinated fibers, however, were not markedly decreased at ankle level. Teased fibers from midcalf and ankle levels (Fig. 78–8; see also

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

1819

FIGURE 78–6 Numbers and sizes of myelinated (left) and unmyelinated (right) fibers per square millimeter of sural nerve at the ankle level in health (top), at the midcalf level in a case of HSAN I (middle), and at the ankle level in the same case of HSAN I (bottom). Data in top left are based on seven nerves from persons 20 to 46 years old; data in top right are based on four nerves from persons 26 to 46 years old. Note in particular the markedly decreased numbers of myelinated fibers at ankle level as compared to midcalf level—evidence of the “dying-back” phenomenon.

Fig. 78–7) from a typical case are instructive about the nature of fiber degeneration. For both levels, approximately one quarter of the fibers were undergoing axonal degeneration. Approximately one fifth of the myelinated fibers showed myelin irregularity (condition B), and approximately one sixth showed demyelination and remyelination. How are these changes of the nerves to be interpreted? First, all size classes of fibers are affected but small may be more affected than large ones. Second, fibers appear to be undergoing atrophy, myelin wrinkling, demyelination and remyelination, and axonal degeneration—changes found in axonal atrophy and degeneration. Third, the changes may be more severe at the ankle than at the midcalf level. Fourth, axonal spheroids were not encountered, although proximal nerve levels were not studied. When results of the studies of Jughenn and associates and Denny-Brown and our studies are considered together, one assumes that lumbosacral spinal ganglion neurons

(nerves) are preferentially (but not exclusively) affected in this disorder. Individual neurons undergo atrophy of their axons that is sufficiently slow that myelin wrinkling and remodeling occur, but eventually axons degenerate into linear rows of myelin ovoids and balls. It is likely that the process is more severe distally, but detailed morphometric and teased-fiber studies of proximal to distal levels of nerve at death are needed. Whether axonal spheroids occur in proximal nerves remains unsettled.

HSAN I with Acral Mutilation and Peroneal Weakness ⫾ Lancinating Pain ⫾ Hearing Loss ⫾ Hyperhidrosis (SPTLC1) The clinical and electrophysiologic features of this group with mutation of SPTLC1 mimic the reports of much earlier “typical” kindreds described above. Missense mutations

1820

Diseases of the Peripheral Nervous System

FIGURE 78–7 Fourteen consecutive portions (from top to bottom) of a teased myelinated fiber from a patient with HSAN I. This fiber shows typical features of axonal atrophy: old internodes have marked irregularities and wrinkling of the myelin, and new internodes (as seen in strips 6 and 7 and strips 9 through 12) are of a lesser diameter, have thin myelin, and have fewer myelin irregularities. We assume that such a fiber is undergoing axonal atrophy and degeneration.

in the gene SPTLC1 have been reported in some kindreds.14,33,103,180 The gene resides at chromosome position 9q22.1-22.3 and consists of 15 exons and encodes a 473–amino acid protein that is a long chain base 1 (LCB1) subunit of the serine palmitoyltransferase (SPT) enzyme. In eukaryotes the SPT gene is a heterodimer, whereby LCB1 and LCB2 subunits are expressed by SPTLC1 and SPTLC2, respectively. Both are predicted to be critical to the enzyme’s function, but in individuals with the HSAN I phenotype, mutations have not been found in subunit 2 of SPTLC.32 The expressed enzyme is crucial in sphingolipid biosynthesis. It catalyzes the pyridoxal-5⬘-phosphate–dependent condensation of L-serine and palmitoyl-coenzyme A to generate 3-ketodihydrosphingosine. This is the first step in sphingolipid biosynthesis of sphingomyelin. In mammals, the enzyme is expressed in multiple tissues, both neurologic and non-neurologic, and its activity is primarily associated with the endoplasmic reticulum.80,117 In catabolism sphin-

gomyelin produces ceramide, which has apparent function in programmed cell death in multiple tissue types,105,138,154 including differentiating neuronal cells during neural tube closure.78 Abnormal neuronal apoptosis was suggested as causative in HSAN I when patients with SPTLC1 mutations were noted by Dawkins and associates33 to have increased de novo glucosylceramide synthesis in lymphoblast cell lines. Glucosylceramide synthesis in five HSAN I patients was 175% of the levels in eight controls. Two of the three identified HSAN mutations (C133Y and C133W) have been created in Chinese hamster ovary cell lines13 and a yeast system60 to confer dominant negative affect of the SPT enzyme. Further work will be needed before the exact pathogenic role of this enzyme defect is known. The fact that patients are apparently normal or typically subclinically affected in childhood or youth and develop accumulating impairments with time provides support to the idea of abnormal neuronal apoptosis.

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

1821

FIGURE 78–8 Teased fibers from a patient with HSAN I, showing advanced stages of axonal atrophy and degeneration. A, Seven consecutive portions (from top to bottom) of a teased fiber showing long regions without a myelin sheath and markedly abnormal portions of remaining internodes of myelin. The observation of repeated segmental demyelination in one teased fiber is in keeping with the idea of segmental demyelination secondary to axonal atrophy. Electron micrographs of single fibers lead us to believe that the axis cylinder is intact, although atrophic, at this level. B, Four consecutive portions along the length of a teased fiber with linear rows of myelin ovoids. The single fibers in this condition seen with the electron microscope have not shown a preserved axis cylinder.

The variability of clinical features of those HSAN I kindreds linked to chromosome 9q22 and eventually identified with SPTLC1 mutations is of note. Weakness in the acral legs more than the hands is a common feature, but the extent of acral mutilations is quite varied among individuals and different kindreds. In 1996, Nicholson and colleagues126 reported on one large and three smaller kindreds with HSAN I, providing mapping of the gene to an 8-cM region at 9q22. Linkage analysis was further refined to a region of approximately 25 kb by one large American kindred of German origin with 17 (and recently more) identified affected members spanning over five generations.12,14 Subsequently, with the identification of the gene, 13 kindreds, including one of our own, have been noted to carry one of several SPTLC1 mutations (C133Y, C133W, V144D, G387A). From these specific families a relative understanding of the genotype-phenotype correlation is possible. Wallace186 initially reported on the phenotype of a large Australian kindred. He referred to the condition as a hereditary sensory radiculopathy, with a description similar to that of the families described by Hicks79 and Dyck and colleagues.46 Patients had mutilating injuries to the feet with variable shooting painful discomfort with a progressive nature, and weakness of the ankles was common. A large Austrian family with the

C133Y missense mutation had six affected individuals and most had weakness, with the youngest presenting patient having this as the presenting feature initially being diagnosed as having HMSN II.6 Among the Austrian kindred: (1) five had sensory loss with injuries in young adulthood as the presenting symptom; (2) four had recurrent foot ulcers all of which went on to osteonecrosis, osteolysis, and digit or leg amputations; (3) one had pes cavus; (4) two had hearing loss later in life; and (5) three had hyperhidrosis. Our family, with C133W mutation (Fig. 78–9), was afflicted with shooting pains and features similar to those described above, barring absence of noted hearing loss and hyperhidrosis. One obligate female member in her 50s (III-1) had no clinical symptoms, suggesting variable penetrance or extremes of expression. Although the details of many kindreds’ phenotypes are not described, in general the other families are similarly affected with variable severity. The countries of origin among the described SPTLC1 mutated families include (to date) Australia,33 Austria,33 Belgium,180 Newfoundland,103 Canada,14 and an American kindred of German origin.14 It is of note, however, that among three English and three Australian families a common founder affect has been established for the C133W mutation.127 The commonality of the exon 5 and 6 SPTLC1 mutations among many suggests either other founder affects,

1822

Diseases of the Peripheral Nervous System

FIGURE 78–9 Left, The pedigree of one Newfoundland family with HSAN I whom we followed and identified to have missense mutation within SPTLC1.103 II-4 and II-5 had both undergone foot amputations, and the former had those amputations after having burned her feet against an oven without recognizing that she had burned herself. Members of the family had prominent peroneal atrophy and weakness. One member, an obligate carrier, had no symptoms (III-1). Right, The forward complementary missense mutation 399T : G [C133W] within exon 5 of SPTLC1. See Color Plate

or alternatively that the region confers a significant enzymatic functional domain, the latter possibility suggested by Bejaoui and co-workers.14 The description of additional families of different populations will be important in answering this question. In the original 24 HSAN I kindreds who underwent SPTLC1 sequencing, only 11 were identified to have mutations suggesting considerable genetic heterogeneity.14,33 Those chosen to be screened had significant peroneal weakness among most along with their predominant sensory loss, so a selection bias may have occurred, including those known to previously link to 9q.

HSAN I with Acral Mutilation and Peroneal Weakness (RAB7) Missense mutations in the gene for RAB7, a small GTPase late endosomal protein with homology to Ras-related GTPases, have been found to be causative of phenotypes characteristic of both the HMSN type designated IID (see Chapter 73) and HSAN I. Clinically patients have phenotypes similar to that of those with SPTLC1 mutation but without lancinating pains, hearing loss, or hyperhidrosis reported to date. The gene RAB7 in humans is expressed in multiple tissues, both neural and non-neural. RAB7’s function appears important in axonal vesicular transport. The protein localizes to endosomes and, as noted in work on phagosomes, seems critical in retrograde tubular extensions.25,97 The syndrome associated with mutations in RAB7 was originally classified as a form of HMSN (i.e., HMSN type IIB) because of prominent motor features; however, this classification has been correctly questioned given the

extent of motor weakness seen in other HSAN kindreds, including those with SPTLC1 mutations.178 More importantly, patients with RAB7 mutations commonly had ulceromutilating injuries of their feet.7,34,108,181 Genetic testing for RAB7 mutations is likely to become an important step in the evaluation of individuals with axonal neuropathies with or without foot injuries and/or ulcerations. To date two missense mutations have been reported among European families, V162M and L129P.181 Among the reported cases with RAB7 mutations, lancinating extremity pains have not been noted, which may be helpful in the clinical distinction from SPTLC1-mutated patients.6 We have screened 26 American HSAN I kindreds for SPTLC1 and RAB7 mutations and found only 1 to carry the SPTLC1 (C133W) mutation (see Fig. 78–9); none had mutation of RAB7.103 The kindred identified with SPTLC1 mutation and 10 others had peroneal weakness, with prominent foot injuries or Charcot joints, recurrent ulcers, or mutilating injuries leading to amputations, similar to the phenotypes described with SPTLC1 and RAB7 mutations. Our family with identified SPTLC1 mutation is from Newfoundland, Canada, but is clinically and genetically different from the Newfoundland family described by Ogryzlo,131 which has recently been identified as having a genetic defect of the putative gene called “HSN2”. Of our remaining kindreds studied, 12 had mutilating acropathy without peroneal atrophy or weakness; 1 had restless legs and burning feet; 1 had dementia, hearing loss, and peroneal atrophy (Fig. 78–10); 1 had demyelinating nerve conductions (i.e., ulnar motor nerve conductions in the 20-m/s range) with peroneal weakness and worse amputations associated with diabetes (see

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

Fig. 78–10); and 1 had worse presentations associated with positive antinuclear antibody and extractable nuclear antigen testing and diagnosis of Sjögren’s syndrome (see Fig. 78–10). Lancinating pains of the feet were common but not present in all patients. We concluded that HSAN I is clinically and genetically heterogenous, that SPTLC1 and RAB7 mutations appear to be an uncommon cause of HSAN I among these American kindreds, and that diabetes and Sjögren’s syndrome may influence clinical severity in HSAN I. We also examined 92 individuals labeled as having cryptogenic sensory neuropathy and found none of the known mutations in SPTLC1 and RAB7, suggesting that mutations in these genes are unlikely to cause sporadic idiopathic sensory neuropathy, but were reserved in our comment because the entire open reading frame has yet to be examined in those patients.

Autosomal Dominant Forms of HSAN without Gene Discovery HSAN I ⫾ Gastroesophageal Reflux ⫾ Cough (3p22-p24) Further evidence of genetic heterogeneity in HSAN I can be seen in those patients described by Kok, Spring, and colleagues.104,163 Although gastroesophageal reflux in the general population is common,124 Spring and co-workers163 reported a large Australian kinship with autosomal dominant neuropathy tracking with symptoms and testing consistent with gastroesophageal reflux–induced cough. Linkage studies were later performed in this family and one other to localize the abnormality to a region of 3.42 cM at 3p22-p24 with haplotype analysis.104 The nerve conductions and nerve biopsy suggested axonal sensory pathology, and all but one member of eight labeled as affected in one family had the constellation of reflux, cough, and neuropathy. Both myelinated and unmyelinated sensory fibers were affected among some biopsied. The symptoms of neuropathy typically occurred one to two decades after the upper gastrointestinal symptoms and included sensory loss, painless injury, and lancinating pains. Neuropathy symptoms began in the third and fifth decades. When performed, monitoring of the esophagus revealed a pH less than 4, with fluctuation at time of cough. Weakness was not specifically mentioned. Candidate gene analyses are underway. Neuropathic Arthropathy with Subclinical HSAN I That neuropathic arthropathy (Charcot joints) may be associated with loss of pain sensation has been known for a long time. This complication is encountered in tabes dorsalis, diabetic neuropathy, syringomyelia, and various sensory neuropathies. The affected joint may demonstrate fractures and bone fragmentation, subluxation, dislocation, effusion, and bone resorption. We have chosen to use the name “neuropathic arthropathy” or “neurogenic arthro-

1823

pathy” rather than “neurotropic arthropathy” because we assume that the joint dissolution is due, in some nerves, to insensitivity (a neuropathy) and use and abuse, rather than to lack of a tropic influence. Neurogenic arthropathy may occur in expressed HSAN I. The complication may be the major manifestation when neuropathy is not well expressed. We have described affected persons from several kinships51 in which the neuropathic arthropathy was the presenting symptom of a subclinical inherited sensory and autonomic neuropathy without conventional signs of peripheral neuropathy (muscle weakness, areflexia and sensory loss, and autonomic dysfunction). These patients developed Charcot joints at the ankle, knee, or shoulder. Evaluation of these patients with conventional nerve conduction and electromyographic examination revealed either no abnormality or only a mild neuropathy. Quantitatively evaluated detection thresholds of cutaneous sensation were also either normal or only slightly abnormal. The detection threshold of nociception of the periosteum of the lower one third of the tibia was sometimes unequivocally abnormal. Morphometric and quantitative teased-fiber evaluation of cutaneous nerves of these patients also provided evidence of a neuropathic abnormality. Histologic study of the tibia from a leg amputated because of acral mutilation, and strength of the bone under tension, were without abnormality. The disorder seemed to be dominantly inherited in some kinships, recessively inherited in others, and possibly sporadic in still others. It was our opinion that an abnormality of bone nociception played a role in this complication, but use and abuse, especially related to overweight and neglect of injury, probably were important risk factors. From these observations we draw the following conclusions. First, a history of degenerative arthropathy may indicate the presence of a subclinical neuropathy that may not be detected by the conventional clinical and neurologic neurophysiologic examination. Second, this subclinical neuropathy could be shown in several instances to be inherited. Third, the onset and pattern of inheritance suggested that there were different underlying genetic disorders. Fourth, although patients reported pain, the amount was less than anticipated from the amount of joint destruction observed. Fifth, observation for other evidence of a mild neuropathic abnormality and the elevated periosteal nociception threshold in some patients led us to believe that raised bone nociception threshold was an important factor in the development of the arthropathy. Sixth, overweight, often of serious proportion and from a young age, was frequently encountered in this group of patients, emphasizing the additional roles of use and abuse in the development of these neuropathic abnormalities. Finally, because of differences of natural history and patterns of inheritance, several types of HSAN may be involved. The factors that may be involved in the development of neuropathic arthropathy are shown in Figure 78–11.

1824

Diseases of the Peripheral Nervous System

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

1825

FIGURE 78–10 Examples of HSAN I kindreds with negative mutation testing for SPTLC1 and RAB7. Top, One large HSAN I kindred having dementia and hearing loss of variable severity, but tracking with peroneal atrophy and axonal sensory-predominant neuropathy with mutilating foot injuries; they were also negative for MPZ mutation. Middle, Kindred with demyelinating nerve conductions (ulnar motor forearm in the 20-m/s range) and dominant small fiber (A␦ and C) sensory loss leading to amputation; genetic testing was also negative for PMP22, GJB1, MPZ, NEFL, EGR2, and LITAF. Shown is the foot ulcer in one chronically affected family member (IV-10) with recent-onset diabetes, who with that diagnosis subsequently underwent foot amputations. Bottom, This kindred had variable severity of sensory neuropathy, with worst affected members having mutilating injuries of the feet. Shown is the most severely affected individual’s (II-2) healing fifth metatarsal fracture, which presented with “foot swelling” without pain. II-2 also had markedly elevated antinuclear antibody (ANA) and extractable nuclear antigen (ENA), with a clinical diagnosis of Sjögren’s syndrome, having dry eyes and mouth with inflammation on minor salivary gland lip biopsy. His identical twin brother (II-3) also had sensory-predominant neuropathy, but with normal ANA and no mutilation injuries of the feet. The combined observations of these kindreds and others suggest that HSAN I varieties are genetically and clinically heterogeneous, and a metabolic or immune inflammatory process may influence severity.

Burning Feet and Subclinical HSAN I Burning feet is a common, poorly understood symptom complex that has various causes (e.g., small fiber sensorimotor polyneuropathy, plantar fasciitis, and various arthritic conditions). Among the causes of hyperalgesic sensorimotor polyneuropathy are diabetes mellitus, old age deficiencies, toxicities of various kinds, and immune conditions. In a kindred on which we reported,44 burning feet were related to a dominantly inherited, usually subclinical sensory neuro-

FIGURE 78–11 Postulated factors and sequence of events leading to neurogenic arthropathy as described in the text. (From Dyck P. J., Mellinger, J. F., Reagan, T. J., et al.: Not “indifference to pain” but varieties of hereditary sensory and dysautonomic neuropathy. Brain 106:373, 1983, with permission.)

pathy. Typical symptoms were “burning” of the soles of the feet coming on in the second decade of life. The symptom was worsened with use of the feet and with excessive warmth and was relieved by cooling. At the onset of the syndrome, the discomfort in some affected patients was almost incapacitating, yet the conventional neurologic examination was normal. Nerve conduction studies of sensory nerve action potentials (SNAPs) also were normal. ComputerAssisted Sensory Examination (CASE) of touch-pressure,

1826

Diseases of the Peripheral Nervous System

vibratory sensation, and thermal cooling provided detection threshold values of the toe and foot within the range of normal values for site, age, and sex. Only mild neuropathologic abnormalities were found in sural nerves. Older persons from the same kindred, who had had similar burning of the feet when they were young, had mild abnormalities of sensory detection threshold of the foot by CASE, some slowing of conduction velocity of motor fibers, and some decrease in amplitude and conduction velocity of the evoked potentials of sensory fibers of limb nerves. Morphometric and quantitative assessment of teased fibers provided unequivocal evidence of sensory neuropathy. From these observations several conclusions were drawn: (1) the history of burning feet may indicate the presence of a subclinical neuropathy, which may not be detected by conventional neurologic and clinical neurophysiologic examination, (2) the neuropathy may be inherited (in our kinship by an autosomal dominant gene), and (3) cases with burning feet may represent a different type of expression of HSAN type I or possibly a unique sensory neuropathy. Linkage analysis in at least one large German family with burning feet and autosomal dominant inheritance failed to show association with HSAN I on chromosome 9q22 or Charcot-Marie-Tooth disease type 2B at 3q13-q22.164 Lancinating Pain and Restless Legs with Subclinical HSAN We have evaluated persons with lancinating pains of the legs and feet, coming on in one kinship near middle life, who during the first years of their disorder were without abnormality of the neurologic examination or of nerve conduction and electromyographic examinations. Biopsy of sural nerve revealed mild abnormalities detected with morphometric and quantitative teased-fiber evaluation. Later these patients had developed mild neuropathic and neurophysiologic abnormalities. Other members of the kinship also developed lancinating pains and were demonstrated to have mild neuropathic involvement. In another kinship, several members had pain characteristic of restless legs, which was associated with subclinical neuropathy. The inherited nature of restless legs had been recognized prior to adequate electrophysiologic and histopathologic testing.87 One five-generation family was described in which the condition began in adolescence and was relieved by cold soaks. Other families were subsequently described.18,54 Variable expressivity has been noted among some members with apparent autosomal dominant inheritance, including myoclonic-like movements.17 Among 12 affected monozygotic twins pairs studied, 10 were both affected, suggesting high penetrance.133 Among a large French-Canadian affected family, linkage analysis localized a potential causative gene to 12q12-q13,36 near a region of multiple neuromuscular disorders.102 Among those families in which genetic work

has been undertaken, neuropathic insult is lacking by electrophysiologic, histopathologic, and quantitative sensation testing.36

AUTOSOMAL RECESSIVE HSAN FORMS HSAN Type II: “HSN2” Gene and Theoretical Protein HSAN II is an autosomal recessive disorder with onset in infancy and childhood with often severe acral mutilating injures (Fig. 78–12). Among a number of Canadian kindreds a putatively responsible gene has been identified and designated “HSN2”.109 The extent to which other HSAN II kindreds of autosomal recessive inheritance typically of congenital or childhood onset will have defects in this gene is still undetermined. We think it unlikely that a single gene abnormality will explain all the kinship disorders following this pattern, but at this point this seems to be the case among the Canadian kinships examined to date.109 Genetic heterogeneity is suggested by the diversity of the original nomenclature, which has included the tissue complications or the presumed underlying mechanism. Names such as “Morvan’s disease,” “painless whitlows,”75 “syringomyelia of infancy,”19,28 “familial neuropathy of unknown etiology resembling Morvan’s disease,”131 “hereditary sensory neuropathy,”79,86,150 “congenital sensory neuropathy,”140,185 “progressive sensory neuropathy,”98 “syndrome of the neural crest,”20 and “acrodystrophic neuropathy”162 have been used. The descriptions provided allow for differences of the phenotypes, suggesting possible genetic heterogeneity. Recognizing that it was one of several varieties of hereditary sensory neuropathy (now HSAN), we referred to it as HSN II (now HSAN II). The clinical hallmarks of HSAN II are different from those of other recessive forms and include 1. Sporadic or sibship occurrence, suggesting autosomal recessive inheritance 2. Onset of symptoms in infancy or early childhood, suggesting congenital onset 3. Occurrence of paronychia; whitlows; ulcers of fingers (see Fig. 78–12); plantar ulcers; unrecognized fractures of the foot, hand, and, less commonly, limb bones; and Charcot joints 4. Sensory loss that affects all modalities of sensation and affects lower and upper limbs and perhaps the trunk as well, unlike the pattern found in HSAN I 5. Absence or diminution of tendon reflexes, usually in all limbs 6. Sweating loss, especially over acral parts, but without postural hypotension, sphincter disturbance, or impotence in the male 7. Absence of SNAPs 8. A virtual absence of myelinated fibers with decreased numbers of unmyelinated fibers in sural nerves

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

1827

FIGURE 78–12 Hands and feet of two teenage boys showing the typical mutilating acropathy of HSAN II.

Clinical Features Head’s case, a young child with healthy parents, developed painless whitlows and “knocked herself about” without discomfort.23,75 Mental development was normal. Because she exhibited Romberg’s sign, posterior column spinal cord involvement was assumed. In the original descriptions of Ogryzlo’s131 Newfoundland kinships, 4 of 12 children, with unaffected parents, had ulceration of the tips of the fingers and the soles of the feet and repeated infections, with the formation of paronychia and whitlows that produced loss of nails and digits. This same family was included in the chromosomal localization of the likely genetic abnormality at 12p13.33 (gene “HSN2”; see below).109 In our family132 the children had to empty their pockets to know what was in them—an indication of severe tactile sensory loss. Stress fractures, muscle atrophy, and acral mutilation with amputation were common. Muscle weakness was not found. Tendon reflexes were abolished. Abnormalities were noted in light touch, pain sensation, and temperature sensation (in order of decreasing severity). The sensory loss was greatest in distal limbs. Position sense was absent. Ogryzlo reasoned that this was a peripheral neuropathy selectively affecting the afferent peripheral nerve fibers. Some of these features were very similar to those of the subsequent kinships described by Murray123 and Davar and associates,31 but it is unsettled whether these are examples of HSAN II.31 It will be important to examine other kindreds for the “HSN2” mutations. In the kinship reported by Heller and

Robb, the children had difficulty with buttoning, developed frequent whitlows, and developed osteomyelitis that led to loss of the phalanges of fingers and toes.76 Tactile sensation was most severely affected, followed by pain sensation and temperature sensation. Repeated study40,84,132 of a Quebec kinship with HSAN II showed that parents were unaffected but four of six children were affected probably from infancy. It is likely that the gene “HSAN 2” will ultimately be found causative in this kinship given their likely French Canadian origin and the similarity in phenotype to the others described. Affected children could not tie their shoes, button their clothes, pick up small objects, or identify objects in their pockets. On attempting to remove objects from their pockets, their fingers would inadvertently include the lining of the pocket with the object to be extracted. Each of the children had sustained significant injuries to the hands or feet on multiple occasions without having recognized it—for example, one girl had walked on a doubled-over toe for several hours without realizing it. Foot swelling, unassociated with a roentgenographic abnormality, was found on subsequent roentgenograms to be due to a stress fracture. All the affected children, with the exception of one girl, had deformed, shortened, and broadened distal phalanges of fingers and toes (see Fig. 78–12). Chronic festering sores of various sizes were present around the nails and at the tips of the phalanges of hands. Ulcers were found on the plantar surface of the feet and especially on the toes and over the heads of the first and fifth metatarsal bones.

1828

Diseases of the Peripheral Nervous System

The neurologic examination was without abnormality except for cutaneous sensory and sudomotor loss. Mental, cranial nerve, cerebellar, and motor functions were normal. Specifically, patients did not exhibit weakness or muscle atrophy. Tendon reflexes were decreased or absent. The sensory loss affected more of the body than is characteristic of HSAN I. Both upper and lower limbs and perhaps the trunk, to a lesser degree, were affected. The distal limbs were most affected. All modalities of sensation (touch-pressure, vibration, cold, warm, and nociception) were affected. Vibratory, touch-pressure, and cold detection thresholds were studied in the great toe (foot), index finger (hand), or face in representative cases and compared to those in age-matched controls. Thresholds were abnormal for all modalities in the hands and feet, but not in the face. Ataxia in walking or in use of the limbs was not present. Among kinships variously described as having congenital sensory neuropathy or recessively inherited sensory neuropathy are ones that probably have HSAN II,19,75,184 as described below. Adams and co-workers described a 38-year-old Ashkenazi Jewish man with a severe pansensory familial neuropathy, possibly HSAN II, to our knowledge without IKBKAP mutation reported, and phenotypically different from familial dysautonomia (i.e., Riley-Day syndrome).1 Two of our patients had severe forms of congenital sensory neuropathy.51 One patient, in addition to a congenital diffuse pansensory neuropathy, experienced difficulty in learning to eat and swallow, had severe constipation, and had other dysautonomic features. The second patient had the additional features of retinitis pigmentosa. It is unclear whether these patients had HSAN II or a phenotypically similar but genetically different disorder. Miller and co-workers120 described two brothers with the features of HSAN II with tonic pupils. As in other patients with HSAN II, the onset may have been congenital, the disorder may have been recessively inherited, worsening was not observed, and the sural nerve had few myelinated fibers but many unmyelinated fibers (as in our cases of HSAN II). The pupils reacted only slightly to light but promptly and with magnitude to accommodation. Nordborg and associates128 described sporadic cases with many of the features of HSAN II. Assuming that these individuals had HSAN II, early development was delayed and they were floppy. Whether they had a severe form of HSAN II or another disorder is uncertain (see the later discussion of other congenital sensory and autonomic neuropathies). The case of Janzer and colleagues,93 a sporadic case with subtotal analgesia and other sensation loss in arms and legs, had tonic pupils. Course In reported cases the onset of the disorder is assumed to be in utero (congenital) or in infancy.21 There is no unequivocal

evidence of clinical worsening. As discussed later, the only evidence for worsening comes from the electromyographic and pathologic examination of sural nerves. Nerve Conduction and Electromyography Winkelmann and co-workers were unable to evoke an electrical response by stimulation of the sural nerve before biopsy.190 Absence of sensory potential from the lateral popliteal nerve was reported by Johnson and Spalding.96 In affected persons the major abnormality of the electromyographic examination is the absence of SNAPs of the ulnar, median, and sural nerves (invariably present in normal children of this age). In our Quebec kinship these observations were extended.132 SNAPs could not be recorded from ulnar, median, or sural nerves. Some minor involvement of motor fibers was also found. The conduction velocity of motor fibers of ulnar, median, and peroneal nerves was at or just below the lower limit of the normal range. The amplitude of the muscle action potential over the extensor digitorum brevis was slightly low in two persons. In the electromyograms of typical patients at ages 15 and 16 years, minimal fibrillation was found in the extensor digitorum brevis muscles with abnormal motor unit potentials (MUPs). MUPs were decreased in number, and an increased proportion were polyphasic. Other muscles examined in the lower limbs had essentially normal MUPs. We studied the compound action potential of the sural nerves of affected children in vitro. A␣ and A␦ potentials were absent (see Fig. 78–2). This corresponded to the virtually complete absence of large and small myelinated fibers (Fig. 78–13). Although low in amplitude, the C potential was present in sural nerves, but it was not detected in the superficial peroneal nerve. Genetics and Mechanisms Among the previously described HSAN II kindreds, along with several others from Newfoundland109,131 and Nova Scotia31,123 and other French Canadian kindreds,109,145 one tentatively causative gene—“HSN2”, residing at chromosome position 12p13.33—has been identified.109 The predicted gene is unique by residing within another gene’s (PRKWNK) intron. Interestingly, mutations of PRKWNK cause autosomal dominant pseudohypoaldosteronism type II (PHA II), characterized by severe hypertension, hyperkalemia, and sensitivity to thiazide diuretics, which may result from a chloride shunt in the renal distal nephron.147,189 The HSAN II and PHA II phenotypes are discordant, as are the mutations of the respective genes. These five Canadian kindreds have been identified to have mutations in the “HSN2” gene.109 All examined affected individuals from the Newfoundland families carried a homozygous mutation, 594delA, causing a frameshift and

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

1829

FIGURE 78–13 Representative transverse sections of sural nerve under phase contrast (A and B) and electron (C) microscopes. A, Healthy nerve, showing large and small myelinated fibers. B, Nerve from patient with HSAN II showing absence of myelinated fibers. C, Nerve from patient with HSAN II showing presence, albeit in decreased numbers, of unmyelinated fibers. (Modified from Ohta, M., Ellefson, R. D., Lambert, E. H., and Dyck, P. J.: Hereditary sensory neuropathy, type II: clinical, electrophysiologic, histologic, and biochemical studies of a Quebec kinship. Arch. Neurol. 29:23, 1973.)

truncation of the protein from 434 amino acids to 206 amino acids. A homozygous insertional mutation (918-919insA) was found in the Nova Scotia family leading to a truncation of the protein to 318 amino acids. The French Canadian kindreds had this conserved mutation, but in the heterozygote state, and had an additional homozygous 943C-to-T nonsense mutation. Such mutations were not found in a

large cohort of appropriate normal controls. Despite these very convincing studies, Northern blot assays looking for the predicted “HSN2” transcript in multiple human tissues and more sensitive reverse transcriptase assays utilizing polymerase chain reaction in various tissues, including dorsal root ganglion, have failed to conclusively identify the predicted product. Further work is needed before the exact

1830

Diseases of the Peripheral Nervous System

nature and cause-effect relationship of these base changes can be accepted as definitively causative. It is likely, however, that the difficulties in isolation relate to the selective expression in nerve and low levels. A signaling peptide sequence is predicted from the known sequence of “HSN2”, and the authors have speculated about the “HSN2” protein product potentially being important in nerve growth.109 Identification of the “HSN2” gene abnormality provides the first step in understanding one likely mechanism of pathogenesis. Other information can be gleaned from the prior histopathologic and family studies. Previously, we40,50,132 and others106 have suggested that congenital sensory neuropathy might be recessively inherited and a separate disorder(s) from dominantly inherited sensory neuropathy. Kondo and Horikawa divided previously reported sibships of HSAN into those with recorded abnormality in successive generations and those without successive abnormality.106 They evaluated 24 nonsuccessive sibships. Parental consanguinity was encountered in 83.3% of these, whereas in no instance was there involvement in successive kinships. For nonsuccessive kinships the segregation ratio probability (P) was 0.26 (near the expected value for an autosomal recessive gene). The value of P for successive kinships was 0.40, which suggests the operation of a dominant gene with incomplete penetrance, inadequate evaluation, or both. The disorder HSAN II therefore appears to be expressed by autosomal recessive genes. Because the disorder is present from birth and selectively affects myelinated more than unmyelinated neurons (axons), we have postulated that the selectively affected neurons (axons) do not develop or, alternatively, develop and degenerate prematurely. Pathology Ogryzlo found marked demyelination—fewer myelin profiles in transverse sections—in nerves from an amputated leg of a child with this condition.131 Loss of myelinated fibers in cutaneous nerves also was reported by others.122,184,185 Winkelmann and co-workers noted that only unmyelinated fibers remained in the sural nerve of their case.190 With electron microscopy, Dyck noted total absence of myelinated fibers and the presence of many unmyelinated fibers in the sural nerve of an affected child from the Quebec kinship.40 Abnormally vacuolated fibroblasts were described by Schoene and associates in sural nerves of patients,149 some of whom probably had this type of hereditary sensory and autonomic neuropathy. In our evaluation of sural nerve biopsy specimens from the Quebec kinship, we found a decrease in transverse fascicular area of sural nerve with an almost complete absence of myelinated fibers at the ankle level and with more myelinated fibers at the midcalf level (suggesting a distal preponderance of low fiber density).132 Unmyelinated fibers were less plentiful at the ankle than at the midcalf level. The peak of the frequency distribution of diameters was displaced to the left of that seen in healthy sural

nerves. The electron microscopic features were interpreted as showing loss of unmyelinated fibers, but the evidence was not conclusive. An important finding was an increased incidence of degenerating fibers in biopsy tissue from the lateral fascicle of the deep peroneal nerve. Approximately 7% of myelinated fibers showed segmental demyelination and remyelination. If trauma or compression of the nerve or of more proximal nerve can be excluded, this might favor an active, ongoing pathologic process. The following four lines of evidence may favor a degenerative process, but one beginning in utero or in infancy: (1) lesser numbers of fibers at the ankle than at the midcalf level; (2) occurrence of denervation (fibrillation) in the extensor digitorum brevis muscle; (3) segmental demyelination and remyelination, which may be secondary to axonal atrophy and degeneration; and (4) electron microscopic abnormalities within axis cylinders. The evidence, however, is not conclusive. Xenografts of sural nerve fascicles from patients with this disorder as compared with those from controls showed no difference in the degree of myelination. Therefore, we reasoned that the abnormality of myelination was not primarily due to a Schwann cell abnormality.48 Biochemistry Open liver biopsies were performed on two affected persons from the Quebec kinship, but no abnormality of lipids was found. Treatment Development of stress fractures, acral mutilation, and infection is difficult to prevent. Not only the feet but also the hands and other parts are insensitive. Furthermore, the onset is in infancy or young childhood, before the age at which full cooperation by the patient can be expected. By the age at which such cooperation is possible, stress fractures, mutilation, and infection have already developed. In addition to the care of the feet described for HSAN type I, special care of the hands is also needed. Patients must be instructed to use their eyes so that they do not bruise their fingers when they grasp objects. Patients must learn to judge when excessively hot objects might damage tissue. The patient him- or herself must be taught the nature and severity of the problem so that he or she may prevent injury. As in all patients with neuromuscular deficits of long duration and beginning at a young age, great emphasis on specialized training and on scholastic achievement is needed so that the person may have some options in obtaining a living. Genetic counseling regarding the inheritance and the natural history should be undertaken, because this will likely make it easier, in those with the “HSN2” mutations, to provide carrier and prenatal counseling.

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

HSAN Type III (Familial Dysautonomia): IKBKAP Familial dysautonomia is a recessively inherited disorder seen predominantly in Jewish infants and children. It is likely the most common form of autosomal recessive HSAN; however, with identification of the causative defective gene (IKBKAP)156 and ease of genetic mutation identification, combined with use of intrauterine and carrier assessment along with egg selection techniques, the incidence will likely decrease. The disease affects peripheral autonomic, sensory, and motor neurons, and likely other classes of central nervous system neurons. Other characteristic features of the disorder are typical onset in infancy, premature death (often in infancy or childhood), history of poor feeding, repeated episodes of vomiting and pulmonary infections, autonomic disturbances (defective lacrimation, defective temperature control, skin blotching, excessive perspiration, hypertension, and postural hypotension), insensitivity to pain, areflexia, corneal insensitivity, and absence of fungiform papillae of the tongue. The initial account was published by Riley and co-workers in 1949.141 Good reviews of the condition are available.9,24,30,63,115,121,142,160 Clinical Features Although there is an overlap of symptoms and signs in this disorder with those of other varieties of congenital sensory and autonomic neuropathy, this variety can be distinguished by the predominance of autonomic symptoms, absence of fungiform papillae, and occurrence among Ashkenazi Jews. Characteristically, the disorder is present at birth. Symptoms are poor sucking, failure to thrive, unexplained fever, and repeated episodes of pneumonia. The difficulty in feeding is often reported by the mothers as a difficulty in swallowing. The characteristic alacrima (absence of overflow tears), blotching of the skin, absence of fungiform papillae of tongue, and biochemical results may confirm the diagnosis even in infants. Presenting symptoms at an older age may include delay in development, stunted growth, corneal abrasions, kyphoscoliosis, clumsiness, excessive sweating and blotching of the skin with emotion, and decreased pain sensation. The abnormality of sucking, feeding, or swallowing is usually confined to infancy or early childhood. In one series of 16 dysautonomic patients, almost half had kyphosis, scoliosis, or both.3 Profound hypotension and prolonged respiratory depression have been reported after anesthesia in single cases.168 A poor gag reflex has been found. Esophageal motility studies have been reported as showing evidence of megaesophagus and pylorospasm. With cineradiography and manometric techniques, a delay in the coordinated relaxation of the upper esophageal sphincter (cricopharyngeus) and abnormal motility of the body of the esophagus have

1831

been implicated.24 Other reported gastrointestinal disturbances are gastric ulcer, jejunal distention, and megacolon. Vomiting is a common feature in the first years of life. Mahloudji and co-workers suggested that, because the disorder is associated with high fever and with disturbed consciousness and sleep, it may be of central nervous system origin.116 Although mental retardation has been mentioned, especially in early reports,141 later reports such as that of Brunt and McKusick24 have deemphasized its importance. Ocular abnormalities have been reviewed by Fogelson and co-workers.59 A variable degree of decreased visual acuity, pallor of optic discs, and abnormality of the reversed visual evoked potentials are evidence of optic nerve impairment.37,144 Two constant features are alacrima and corneal insensitivity. Parenteral administration of methacholine to these patients produces tears, an indication that the defect is neurogenic rather than glandular. The signs of autonomic dysfunction are many and superficially may appear to be conflicting. The color alterations of the skin may be divided into a generalized erythema (from exercise, eating, or overheating) and erythematous macules 2 to 5 cm in diameter on the trunk and extremities (with emotional upset).58 Scratching the skin or injecting histamine intradermally fails to produce the erythematous flare of the triple response. This indicates an abnormality of a local axonal reflex. Variability of blood pressure has been reported by many authors. Hypertension in response to emotion has been documented,22,58 as has postural hypotension with change of position from supine to erect by Solitare and Cohen.161 Excessive sweating and erratic sweating have been noted.116 Some of these anomalous responses, such as hypertension, hypotension, and excessive or decreased sweating, may be based on a denervation supersensitivity of blood vessels and sweat glands. Urinary and rectal incontinence and impotence and the associated patulous anus and lack of anal reflex are not features of this disorder. Of particular importance is the evidence of involvement of peripheral neurons. The clinical signs of hyporeflexia and kyphoscoliosis suggest such involvement. An abnormality of conduction velocity20 and a marked decrease of unmyelinated fibers of cutaneous nerves2 have been reported, providing evidence of such abnormality. The markedly decreased numbers of unmyelinated fibers may explain the abnormality of pain and temperature sensations. Axelrod and co-workers assessed the abnormalities of cutaneous sensation using hand-held instruments in 75 patients with familial dysautonomia.10 Abnormalities of cutaneous temperature discrimination and nociception were found for the majority of patients. Lesser numbers had abnormalities of joint position and vibratory sensation. These authors concluded that the sensory deficit increased with age— evidence of a progressive disorder. Although mutilation of

1832

Diseases of the Peripheral Nervous System

acral parts is not a prominent feature of this disorder, the occurrence of Charcot joints owing to pain insensitivity has been described.68 In view of the absence of tendon reflexes, an abnormality of peripheral sensory neurons may contribute to or account for the ataxia that is seen. Nerve Conduction Studies Motor nerve conduction velocities in limb nerves were shown to be somewhat below control values.2,20 Brown and Johns showed that a second action potential occurred20; Aguayo and co-workers reported a large decrease in amplitude of the compound action potential over the ulnar nerve on stimulation of digital nerves.2 Pathology Prior to the report by Cohen and Solomon,27 there had been nine postmortem examinations of children with dysautonomia; only two had nervous system lesions, and these were not sufficient to explain the disorder in either case. In the 31/2-year-old patient examined by Zimmerman, as reported by Cohen and Solomon,27 the pons and medulla were smaller than expected and there were lesions in the brainstem and particularly in the medulla. In the reticular substance, nerve fibers showed extensive damage (swollen myelinated fibers, myelin fragments, and phagocytosed lipid material in macrophages). The neurons of cranial nerve nuclei X and XI were shrunken and possibly decreased in number. Agenesis or destruction of some Purkinje cells also was noted. In the postmortem examination of a 2-year-old patient, Brown and co-workers found similar alterations in the brainstem reticular formation and also abnormalities of neurons of the cerebral cortex (poorly stained and vacuolated neurons of the third, fifth, and sixth layers) and of long tracts of the spinal cord (evidence of degeneration of myelinated fibers of the ventral and lateral spinothalamic tracts, of the spinocerebellar tract, and of posterior columns and degeneration of myelinated fibers in posterior roots).22 Absence of abnormalities of brainstem reticular structures with focal demyelination of the fasciculus gracilis (sacral to the cervical cord) and focal demyelination of dorsal roots (lumbar and sacral levels) in an 11-year-old boy was reported by Fogelson and associates.59 Morphometric studies of autonomic and spinal ganglia have shown that their gross volume is markedly decreased, as is the number of neuron cell bodies. A marked decrease in the number of neuron somas in cervical and thoracic sympathetic ganglia (estimated to be decreased to 27% of normal) and in gasserian and spinal ganglia (estimated to be 50% of normal) was reported by Pearson and Pytel.136 Active neuronophagia and many residual nodules of Nageotte were seen. Decreased numbers of neuron cell bodies have been demonstrated for the sphenopalatine ganglia, superior cervical ganglia and sympathetic ganglia, and spinal ganglia.136 Our morphometric studies of lumbar

motor neurons in one case indicated a selective absence of gamma motor neurons.45 A sural nerve biopsy specimen from a 25-year-old patient2 was found to contain an approximately normal density of myelinated fibers (8810/mm2; normal, 8000 to 12,000/mm2) but a greatly decreased number of unmyelinated fibers (3617/mm2; normal, 30,000 to 50,000/mm2). No active fiber degeneration was seen. Schwann cell cytoplasmic clusters were decreased in number as compared to that in other varieties of neuropathy. The findings were interpreted as consistent with an arrest in embryogenesis of the neural crest. Nemaline myopathy was reported in one patient with familial dysautonomia.4 Genetics Several mutations in the IKBKAP gene, located at 9q31, are causative of HASN III.16,156 Mutation of intron 19 at the donor splice site sequence (IVS2⫹6T-C) leads to inefficient production of a full-length transcript and consequently truncated IKAP protein.156 This noncoding mutation (IVS2⫹6T-C) accounts for 99.5% of all HSAN III cases.156 Why such defect should lead to a dysautonomia and sensory neuropathy is unknown, but poor development and maintenance of neurons are implicated given the neuropathologic and clinical progression often seen. Molecular examination of HSAN III patients’ brains suggests that neuronal cells are less able to produce a percentage of normal wild-type product, compared to other tissues in which normal populations are seen.29,156 Two rare IKBKAP mutations in HSAN III patients are predicted to reduce phosphorylation of the protein.112,156 Among only four affected kindreds is there a heterozygous arginine-to-proline amino acid change in exon 19 along with the common donor splice mutation in the heterozygous state.156 Another family having one non-Jewish parent had a proline-to-leucine mutation in exon 26, also with a heterozygous common mutation.112 Genetic testing for IKBKAP mutations has allowed for ease of carrier and affected assessment in a population (Ashkenazi Jews) in which a strong founder effect accounts for the limited responsible mutation. Inheritance Although a few patients with a similar phenotype have not been Jewish, most are.88,113 The sexes are affected approximately equally. Analysis of family pedigrees indicated that the disorder is inherited as an autosomal recessive trait.24 The gene frequency for this disorder in Israel has been estimated at 0.91 per 100,000; the carrier frequency in the same population has been established at 18 per 1000.121 In North American Jews, Brunt and McKusick estimated a gene frequency of 0.01 per 1000 at most.24 In a more recent study the incidence of dysautonomia in Israel was estimated to be higher (27 per 100,000) than among

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

American Ashkenazi Jews.115 With the identification of specific genetic (IKBAKP) mutations and population evaluations, the carrier frequency of the most common mutation in the Ashkenazi Jewish population has been reported to be between 1 in 27 and 1 in 32.39,165 Biochemistry With the advent of identification of the genetic cause of HSAN III within the IKBKAP gene, the earlier biochemical observations provide insight into the effects of such defect, but precise understanding of the gene’s mechanism of injury remains elusive. Some studies suggested that IKAP, the protein encoded by IKBKAP, was important in transcriptional efficiency of full-length messenger RNA.74 Current evidence suggests its role in response to microenvironmental stress through the c-Jun N-terminal kinase signaling pathway.71,81 The latter postulate lends itself to a theory of inadequate development, poor survival, or differentiation of neuronal cells. Smith and co-workers first demonstrated an abnormality of catecholamine metabolism in dysautonomia.158,159 Patients excreted increased amounts of homovanillic acid, decreased amounts of vanillylmandelic acid and methoxyhydroxyphenylglycol, and normal amounts of total metanephrines and normetanephrines. Gitlow and co-workers reasoned that a simple anatomic decrease in number of adrenergic neurons would be compatible with a decrease in vanillylmandelic acid and methoxyhydroxyphenylglycol but would not explain normal vanillylmandelic or increased homovanillic acid levels.64 Weinshilboum and Axelrod have shown that serum dopamine ␤-hydroxylase, the enzyme that converts dopamine to norepinephrine, was significantly decreased in the serum of patients with dysautonomia.187 Immunohistochemical studies of tyrosine hydroxylase antigen in sympathetic neurons are reported to show increased amounts of this antigen in familial dysautonomia.137 It was suggested that this increase might be a compensatory increase in surviving neurons. NGF was measured in cultured human skin fibroblasts from controls and from patients with familial dysautonomia and dystonia musculorum deformans (DMD).152 The NGF from familial dysautonomia cells was only approximately 10% as active per nanogram of immunoreactive protein as that of control and DMD cells. These results suggested to the investigators that the abnormality in familial dysautonomia might lie in the processing of the precursor or in the structure of the biologically active ␤ subunit of NGF. Recently, mutations of NGFB have been found to cause HSAN V in at least one family53 (see below).

HSAN Type IV: TRKA Also known as congenital insensitivity to pain with anhidrosis and familial dysautonomia type II, the characteristic features of HSAN IV are (1) congenital or infantile onset, (2) sporadic

1833

or sibling occurrence—recessively inherited, (3) repeated high fevers that may cause death, (4) decreased pain sensation and absence of sweating, (5) mental retardation in some patients, (6) virtual absence of unmyelinated fibers in sural nerve, and (7) among all of those described to date, genetic abnormality of TRKA gene. The initial consideration of this gene at 1q21-22 as a candidate for HSAN IV came from a mouse mutant with a similar phenotype.157 Subsequent human work has identified multiple mutations, including deletion, frameshift, splice site, and missense mutations—all in the tryosine kinase domain of TrkA.91 Other mutations in separate families of different race are noted for causing disruption of the extracellular domain involved in NGF binding.118 To date mutations have been found in Israel, Kuwait, Japan, the Netherlands, and the United States.89,90 Shatzky studied consanguineous Israeli-Bedouin groups, in which the disorder is relatively common, and found that 9 of 10 presumably unrelated families linked to the TRKA gene; in the one family that failed to show linkage, genetic heterogeneity (more than one mutation within the TRKA gene) was proposed.153 The gene is thought to be important for inducing neurite outgrowth and promotion of embryonic sensory and sympathetic neurons. The TRKA gene contains 17 exons, with alternative splicing noted.70 Hempstead and colleagues77 demonstrated that binding of NGF requires coexpression of the tyrosine kinase proto-oncogene (trkA) with NGF receptor. NGF stimulates phosphorylation of the TrkA protein in neural cell lines and in embryonic dorsal root ganglia. The TRKA gene product is a NGF receptor, and participates in signal transduction of NGF.100 Not only does the TrkA protein seem important in promotion of embryonic sensory and sympathetic neurons, but it is also reported that TrkA is expressed in monocytes and may also be an immunoregulatory cytokine acting on monocytes.52 Rare mutations of TRKA have been associated with medullary thyroid carcinoma without neuropathy.62 In 1963, Swanson described two brothers with insensitivity to pain, anhidrosis, and mild mental retardation.166 They had some of the features of HSAN II and HSAN III but were different in other ways. During infancy these boys had repeated episodes of high fever that usually were not related to infections but were related to environmental temperature. The family had found that the high fever could be prevented by sponging the child with cool water on a hot day. During infancy and childhood, significant injury occurred without evidence of pain—for example, biting off the tip of the tongue, burning the hands when reaching into a hot oven, breaking an ankle and sustaining many joint injuries, and purposefully burning a fingertip with an automobile cigarette lighter. One of the boys had fainted on arising quickly from a lying position. On examination, mental retardation (IQ in the 70s) was found. Muscle strength and tendon reflexes were normal. Pinprick, burning heat, compression of testicles or tendons,

1834

Diseases of the Peripheral Nervous System

ischemia, dilatation of the esophagus with a balloon, immersion of a limb in cold water, and a pneumoencephalographic procedure were not experienced as painful. Although the sharp end of a pin was usually distinguished from the dull end, warm was not reliably distinguished from cold. Clearly, a sensory abnormality existed. Testing of corneal sensation resulted in some discomfort. No comment was made regarding a greater severity of the sensation loss in acral parts, as is found in HSAN I and II. Although sweat glands were present in excised skin, sweating did not occur from these. On exposure to heat (in a heat-controlled box), some intertriginous sweating could be demonstrated. The lack of sweating was thought to separate this disorder from dysautonomia, in which excessive sweating was reported. One of the boys died at age 12 years.167 Morphometric evaluation was not undertaken, but an absence of small neurons in spinal ganglia, fewer small fibers in dorsal roots, absence of Lissauer tracts, and a decrease in size of the spinal tract of the trigeminal nerve were observed. Kinships with similar features have been described.21,139 Pinsky and DiGeorge distinguished familial sensory neuropathy with anhidrosis (our HSAN IV) from familial dysautonomia (HSAN III), from congenital sensory neuropathy (HSAN II), and from hereditary sensory radicular neuropathy (HSAN I).139 They found that myelinated fibers were present in a sural nerve biopsy specimen and that a normal dermal network was seen. Brown and Podosin attributed this disorder to an abnormality of differentiation of the neural crest.21 Goebel and co-workers described two infants with a virtual absence of unmyelinated fibers in HSAN IV.67 No unmyelinated fibers were found adjacent to sweat glands.111

HSAN Type V: Possibly TRKA and NGFB The type V classification arose out of necessity for a 6-yearold patient described by Low and co-workers who had congenital selective loss of pain sensation affecting the extremities,114 normal muscle strength, preservation of tendon reflexes, and normal tactile and mechanoreceptor sensation. She did not have the other features of the previously described HSAN forms. This patient might have been said to have the features of indifference to pain were it not for a selective severe decrease of the small myelinated fibers of the sural nerve (Fig. 78–14). Since that original patient, several others have been reported with varied detail as to the loss of small myelinated, A␦ fibers.53,85,175 In one of our patients, a girl 2 years and 2 months old, a severe abnormality of nociception was recorded from infancy. Tactile stimuli appeared to be recognized normally.49 As in Low and co-workers’ patient, muscle strength, tendon reflexes, and evoked cutaneous nerve compound action potentials were normal. The pattern of sensory loss and of sudomotor dysfunction had a peripheral neuropathy distribution. As in Low and co-workers’ case, we demonstrated

a severe reduction from normal of small myelinated fibers, but also seen was a small reduction of unmyelinated fibers (Figs. 78–15 and 78–16). It was our view that many earlier cases of “indifference to pain” may have had type IV or type V HSAN. Landrieu and co-workers110 report a mother and a daughter who had congenital indifference to pain but did not have sensory loss (other than nociception), autonomic dysfunction, or abnormality of SNAPs. Number and size distribution of myelinated and unmyelinated fibers of superficial peroneal nerves were not different from those in controls. They argued, convincingly we think, that the nociceptive loss in these patients may not be a hereditary sensory neuropathy but a true indifference to pain. In view of the findings of Landrieu and colleagues, a reappraisal of the mechanism of nociceptive loss in such cases in needed. Even in the Low and co-workers case and in our two cases, one would not have expected the degree of nociceptive loss from the demonstrated decrease of A␦ and, in our second case, of C fibers. It is now thought that many unmyelinated afferent fibers are polymodal nociceptors. In the Low and co-workers case and our cases, a peripheral nerve abnormality has been demonstrated. One therefore should ask whether nociceptor neurotransmitters and central connectivity are normal or the central nervous system sensory function is abnormal. Because one of Landrieu and colleagues’ patients had experienced pain from tooth removal, a complete asymbolia for pain cannot be inferred. It is evident that further research is needed to resolve these issues. Other similar patients were reported by Donaghy.38 One Pakistani patient diagnosed as having HSAN V has been identified to have an exon 8 Tyr359Cys TRKA mutation, predicted to yield partial loss of gene function.85 This patient was believed to phenotypically have HSAN V instead of type IV, in which TRKA mutations are typically seen. The patient was reported to have severe loss of small myelinated fibers. This patient also had anhidrosis, not seen in the earlier patient of Low and co-workers.114 Whereas in HSAN IV, the disorder associated with TRKA mutations, there is usually a virtual absence of small unmyelinated fibers on electron microscopy, the small unmyelinated density was only slightly reduced in this Pakistani patient (i.e., 26,920/mm2 compared to normal control value of 35,700/mm2). Others with the HSAN V phenotype will need to be identified before a firm conclusion about the allelic nature of HSANs IV and V can be made. Toscano and associates reported the absence of TRKA gene mutations in a patient they thought had HSAN V.175 The described patient had congenital and generalized absence of pain perception without neurologic findings apart from anosmia. Included in the testing was normal autonomic testing and nerve biopsy. The diagnosis of HSAN V seemed incorrect to us101 because their patient did not have peripheral nerve abnormality (i.e., reduced numbers of small myelinated sensory fibers

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

FIGURE 78–14 Representative section from a nerve of a patient with HSAN V reported by Low and co-workers.114 Upper, Transverse section of sural nerve of a control subject (left) and the patient (right). Note the relative complete absence of small myelinated fibers. Middle, Myelinated fiber density of a control subject (left) and the patient (right). Lower, Distribution of unmyelinated fiber diameters of a control subject (left) and the patient (right). (Modified from Low, P. A., Burke, W. J., and McLeod, J. G.: Congenital sensory neuropathy with selective loss of small myelinated fibers. Ann. Neurol. 3:179, 1978.)

1835

1836

Diseases of the Peripheral Nervous System

FIGURE 78–15 Diameter histogram of myelinated fibers from sural nerve from our patient with HSAN V. Note the severe absence of small-diameter myelinated fibers as compared with three normal nerves. (From Dyck P. J., Mellinger, J. F., Reagan, T. J., et al.: Not “indifference to pain” but varieties of hereditary sensory and dysautonomic neuropathy. Brain 106:373, 1983, with permission.)

[A␦]). Also, unmyelinated fibers were stated to be “regularly detected,” although morphometry of these fibers was not provided. We think that some evidence of a peripheral neuropathy needs to be demonstrated; otherwise indifference to pain might be a more accurate diagnosis.

One large consanguineous kindred from northern Sweden has been reported to have the HSAN V phenotype associated with a missense mutation (Arg661Trp) of the NGF-␤ gene (NGFB).53 Variable severity was noted in the family, with loss of deep pain perception and temperature

FIGURE 78–16 Diameter histogram of unmyelinated fibers from the sural nerve of our patient with HSAN V. Note the lower peak of fibers in our patient. (From Dyck P. J., Mellinger, J. F., Reagan, T. J., et al.: Not “indifference to pain” but varieties of hereditary sensory and dysautonomic neuropathy. Brain 106:373, 1983, with permission.)

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

with normal mental abilities, and with other neurologic responses intact, including sweating. Nerve conductions were normal. The reported nerve biopsy would be atypical to the original case described by Low and co-workers114 by having “severe loss of unmyelinated” fibers and mild reduction of small myelinated fibers. The histopathologic features would seem more in keeping with HSAN III. Unfortunately no TRKA mutation screen was performed, but certainly the other features of HSAN III were absent. The fact that mutations of NGFB were found as the cause of a small fiber sensory neuropathy in one family without apparent central nervous system involvement suggested to the authors that NGFB may play an important selective developmental role in peripheral versus central nervous system development. The mutation identified is thought to likely restrict but not entirely abolish the function of NGF in its ability to activate its receptor, TrkA.53 In contrast, the reported mutations of TRKA likely abolish function of NGF.

Other Manifestations of Congenital HSAN A large number of previously reported sporadic sensory neuropathies that are probably inherited would benefit from genetic analysis with the currently identified genes in definition of the spectrum of clinical expression of various mutations. We include below a partial historical list, and cannot find confirmation of genetic testing in these patients when reviewing recent literature. HSAN and Neurotrophic Keratitis (Possibly HSAN V) Donaghy and co-workers38 reported three siblings with congenital HSAN and selective absence of small myelinated fibers in sural nerve who had consanguineous Kashmiri parents. Neurotrophic keratitis was present. If one assumes that the corneal changes are secondary to autonomic deficit, inflammation, and scarring, then these cases may represent HSAN V with keratitis as a complication. Acromutilating, Paralyzing Neuropathy with Corneal Ulceration in Navajo Children (Probably an HMSN Variant) Appenzeller and co-workers5 described four Navajo children with severe mutilating and paralyzing neuropathy present from early age. They had severe anesthesia, corneal ulceration, painless fractures, acral mutilation, and scaling. Sural nerves were almost without myelinated fibers and had decreased number of unmyelinated fibers. The spinal fluid protein was elevated. This group might be more appropriately placed with HMSN. Hereditary Sensory Neuropathy with Spastic Paraplegia The five patients with HSN of Cavanagh and associates26 also had spastic paraplegia. The first two patients were siblings

1837

who had the same mother and probably different fathers. In infancy they had stiffness of the legs with hyperreflexia (ankle reflexes absent) and spasticity, development was delayed, acral mutilation of the toes and fingers occurred, and severe sensory loss was observed in the feet and legs as well as in the hands and arms. Mild abnormality of motor nerve conduction was present in the first case. Sensory nerve conduction was not reported. There was profound loss of myelinated and possibly of unmyelinated fibers. Cases 4 and 5, siblings with consanguineous parents, developed spasticity early in life. Mutilating acropathy affected the feet more than the hands. Nerve was not studied. The parents of case 3 were not consanguineous and there were no similarly affected members of the family. Stiffness of gait was noted at 2 years, frequent painless injuries occurred, and mutilating acropathy developed. Beginning at 9 years this patient developed intermittent urinary retention. He died at 29 years. In addition to lower limb spasticity, he had distal acral sensory loss with reduced pain and temperature sensation. Congenital Nonprogressive Sensory Neuropathy with Tonic Pupils (Possibly HSAN II) The patient of Janzer and colleagues93 had subtotal analgesia, with diminished temperature, vibration, and proprioceptive sensation in the arms and legs. SNAPs were absent. The patient had tonic pupils. In the biopsy specimen of sural nerve, the fascicular area was decreased, myelinated fibers were decreased in number, and unmyelinated fibers were greatly reduced in number. Tonic pupils and similar phenotypes have been recognized with certain mutation of myelin protein zero (see Chapter 71). Congenital Insensitivity to Pain with Anhidrosis (Possibly HSAN II) Thrush described a sibship of six children of whom three boys and one girl had what he called insensitivity to pain.173 The parents were thought to be unaffected. These patients were reported to have sustained numerous painless injuries of the head and upper and lower limbs, which resulted in bone fractures and Charcot joints. Although the authors reported no sensory abnormality other than pain loss, this seems inconsistent with the reported loss of large myelinated fibers in sural nerve. Also, the published photomicrographs appear to show a severe abnormality of myelinated fibers and a lesser abnormality of unmyelinated fibers. Whether these patients have a mild HSAN II disorder or another disorder remains unanswered. Vardy and co-workers described two children, of unaffected parents who were first cousins, with a syndrome consisting of mutilating acropathy, Charcot joints, insensitivity to pain and temperature, inability to sweat, and psychomotor retardation.179

1838

Diseases of the Peripheral Nervous System

Congenital Sensory Neuropathy (Possibly HSAN II) This disorder190 occurs among siblings with unaffected parents who may be consanguineous (possibly recessively inherited). The disorder is recognized in infancy or in early childhood (possibly congenital). The sensory loss tends to be generalized over the body and affects all modalities or nociception predominantly. Because of insensitivity, use, and abuse, mutilation of acral parts and Charcot joints may develop. Myelinated fibers may be markedly decreased or absent from cutaneous nerve.83 Recessive Form of Hereditary Sensory Neuropathy Jedrzejowska and Milczarek described siblings with a syndrome similar to HSAN I but that differed in the following respects: (1) the parents were unaffected and (2) axonal degeneration into linear rows of myelin ovoids was prominent in biopsy specimens of sural nerves, whereas other myelinated fibers appeared to be undergoing axonal atrophy and secondary demyelination.94 Although this may represent a new type of HSAN, it is more likely to be recessively inherited HSAN I. Pansensory Familial Neuropathy (Possibly HSAN II) The term pansensory familial neuropathy was used by Adams and colleagues to describe a familial sensory disorder of a 38-year-old Ashkenazi Jewish man. At least since the age of 18 years, he had loss of appreciation of all sensory modalities.1 The sensory deficit was greatest in the limbs but was also severe over the trunk, neck, and face. Areflexia, anhidrosis, postural hypotension, and ataxic gait were other clinical features described. No A␣, A␦, or C potentials could be elicited from an excised sural nerve in vitro, but a few unmyelinated fibers were seen in electron micrographs. A sister with a similar disorder died of complications of osteomyelitis at 29 years of age.

MECHANISMS AND TREATMENT OF ACRAL MUTILATION AND ARTHROPATHY We have recently reviewed the treatment approaches for the HSAN forms in the therapeutics textbook of Noseworthy.47 For specific detailed comment on the multiple approaches within familial dysautonomia (i.e., HSAN III), the reader is referred to Axelrod.9 The foot complications of HSAN may be plantar ulcers, osteomyelitis, osteolysis, Charcot joint, lymphangitis, thrombophlebitis, cellulitis, and loss of the foot and leg. The development of foot ulcers is related to risk factors other than sensory loss. These include use and abuse, neglect of injury, overweight, indifferent personality, and degree of foot care given (see Fig. 78–8). Additionally, we suspect there may be

certain increased risks of infection related to the specific gene abnormality independent of foot care, and insensible losses. Although not proven, certain genetic abnormalities affecting peripheral nerves may theoretically lead to reduced innate immune defense. Specifically, mutations of the gene RAB7 (described above and in Chapter 71) could theoretically interfere with macrophage phagocytosis.73,182 In a Virginia kinship,46 foot mutilation occurred in some members of the kinship but not in others. This was also the case for one obligate older female carrier in a family with the C133W mutation of the SPTLC1 gene (see Fig. 78–9). In the Virginia kindred, women tended to be more careful and prevented and cared for foot injury more than did men. In part this difference may have been related to occupation, in that the men did heavy physical labor while the women did not. Prevention of foot ulcers or their prompt treatment had an important influence on the course of the disease. If foot ulcers were ignored and weight bearing was continued, a triad of events occurred that led to the loss of metatarsal bones. Episodic infection, which sometimes results in septicemia, also occurred. The patient and the doctor must be made to understand that foot ulcers and the more serious consequences are often a failure of preventive care. The patient must be taught that his or her foot may have pressure points, which must be visually inspected at intervals. Because the foot may not sweat, the skin must be hydrated and sealed with petrolatum lotion. The main goal is the prevention of foot ulcers. Shoes should be roomy, firm, and comfortable, and they should be carefully inspected before the feet are placed in them. A forgotten sock, foreign object, or protruding tack may cause sufficient damage to produce an ulcer. Care should be taken that the sock does not double up and cause a pressure sore. These patients should not be doing hard labor that entails the use of the feet or in which bruising of the feet is likely. The patient should be given instruction regarding care of the insensitive feet and must be convinced that good care of the feet can prevent mutilation. It is important to soak the feet daily in lukewarm water to keep the skin pliable and prevent fissures through which infection can enter. An application of petrolatum lotion helps seal in the moisture. Whenever plantar ulcers develop, weight bearing should be discontinued. Proper cleansing of the ulcer and débridement may be necessary. Amputation is a measure of last resort and probably represents a failure of proper care. On occasion, when there is an extremely high arch with pressure points or when the foot is deformed so that the weight is borne on the lateral edge of the foot, corrective foot surgery to prevent ulcers may be indicated. Some thought should, however, be given to the possible benefit of the surgery as weighed against the risk. Antibiotics should be used after culture of the ulcer or culture of the blood when cellulitis or septicemia develops.

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

At present gene therapy is not available for any of the inherited sensory neuropathies. Although theoretically plausible, there is no evidence that any of the growth factors are useful. Because genetic abnormalities of the TrkA receptor appear to be involved in some congenital HSAN varieties, and NGF and TrkA receptors are involved in development of sympathetic nerve fibers and dorsal root C fibers, trials of NGF at very early times (perhaps even in utero) might be considered as potential treatment. For the present, the emphasis must be on avoidance of injury, prevention of acral mutilation, and treatment of positive neuropathic sensory symptoms (P-NSS). Prevention of acral mutilation is difficult, especially among affected children. This may relate to not only the difficulty in teaching children to be careful but the severity of their problem. As soon as the problem of loss of sensation is recognized, special preventative steps should be taken so as to try to prevent injury. The temperature of the hot water heater for the home should be turned down (e.g., to 42° C) so scalding will not occur even when the hot water only is turned on. Stops may be affixed to drawers and doors so that they cannot be slammed shut on insensitive acral anatomic parts. Measures may need to be taken so that hands cannot be placed on the hot burners of stoves. It is much more difficult to prevent a child from hurting him- or herself by chewing on insensitive lips, tongue, or cheek; jumping from excessive heights; or damaging insensitive parts to spite a parent or guardian. Ideally, the supervising adult will be able to set physical limits without resorting to physical punishment. The P-NSS are also difficult to treat. Prickling, pain (lancinating, burning, freezing, deep-ache, allodynia [pain caused by mechanical stimuli that are insufficient to cause pain]), asleep-numbness (like that induced by a local anesthetic; tight, constricted), and related symptoms are all troublesome. These symptoms may be improved by good sleep, relief of anxiety or depression, and physical treatment (e.g., cold soaks and pharmacologic treatment). For many individuals these treatments are inadequate, and simply learning how to cope with the problems via occasional assistance from pain clinics can be helpful. Injections and nerve resections are typically best avoided for such problems.

REFERENCES 1. Adams, R. D., Shahani, B. T., and Young, R. R.: A severe pansensory familial neuropathy. Trans. Am. Neurol. Assoc. 98:67, 1973. 2. Aguayo, A. J., Nair, C. P. V., and Bray, G. M.: Peripheral nerve abnormalities in the Riley-Day syndrome: findings in a sural nerve biopsy. Arch. Neurol. 24:106, 1971. 3. Albanese, S. A., and Bobechko, W. P.: Spine deformity in familial dysautonomia (Riley-Day syndrome). J. Pediatr. Orthop. 7:179, 1987.

1839

4. Amir, N., Jehnke, J., and Shapiro, Y.: Familial dysautonomia manifesting as neonatal nemaline myopathy. Pediatr. Neurol. 1:249, 1985. 5. Appenzeller, O., Kornfeld, M., and Snyder, R.: Acromutilating, paralyzing neuropathy with corneal ulceration in Navajo children. Arch. Neurol. 33:733, 1976. 6. Auer-Grumbach, M., De Jonghe, P., Verhoeven, K., et al.: Autosomal dominant inherited neuropathies with prominent sensory loss and mutilations: a review. Arch. Neurol. 60:329, 2003. 7. Auer-Grumbach, M., De Jonghe, P., Wagner, K., et al.: Phenotype-genotype correlations in a CMT2B family with refined 3q13-q22 locus. Neurology 55:1552, 2000. 8. Auer-Grumbach, M., Wagner, K., Timmerman, V., et al.: Ulcero-mutilating neuropathy in an Austrian kinship without linkage to hereditary motor and sensory neuropathy IIB and hereditary sensory neuropathy I loci. Neurology 54:45, 2000. 9. Axelrod, F. B.: Familial dysautonomia. Muscle Nerve 29:352, 2004. 10. Axelrod, F. B., Iyer, K., Fish, I., et al.: Progressive sensory loss in familial dysautonomia. Pediatrics 67:517, 1981. 11. Baxter, D. W., and Olszewski, J.: Universal indifference to pain. Brain 83:381, 1960. 12. Bejaoui, K., McKenna-Yasek, K., Hosler, B. A., et al.: Confirmation of linkage of type I hereditary sensory neuropathy to human chromosome 9q22. Neurology 52:510, 1999. 13. Bejaoui, K., Uchida, Y., Yasuda, S., et al.: Hereditary sensory neuropathy type I mutations confer dominant negative effects on serine palmitoyltransferase, critical for sphingolipid synthesis. J. Clin. Invest. 110:1301, 2002. 14. Bejaoui, K., Wu, C., Scheffler, M. D., et al.: SPTLC1 is mutated in hereditary sensory neuropathy, type I. Nat. Genet. 27:261, 2001. 15. Bickerstaff, L. K., Harris, S. C., Leggett, R. S., and Cheah, K. C.: Pain insensitivity in schizophrenic patients: a surgical dilemma. Arch. Surg. 123:49, 1988. 16. Blumenfeld, A., Slaugenhaupt, S. A., Axelrod, F. B., et al.: Localization of the gene for familial dysautonomia on chromosome 9 and definition of DNA markers for genetic diagnosis. Nat. Genet. 4:160, 1993. 17. Boghen, D., and Peyronnard, J. M.: Myoclonus in familial restless legs. Arch. Neurol. 33:368, 1976. 18. Bornstein, B.: Restless legs. Psychiatr. Neurol. 141:165, 1961. 19. Bousquet, D.: Un cas de syringomyélie chez l’enfant. Ann. Med. Chir. Infant. 10:673, 1906. 20. Brown, J. C., and Johns, R. J.: Nerve conduction in familial dysautonomia: Riley-Day syndrome. JAMA 201:200, 1967. 21. Brown, J. W., and Podosin, R.: A syndrome of the neural crest. Arch. Neurol. 15:294, 1966. 22. Brown, W. J., Beauchemin, J. A., and Linde, L. M.: A neuropathological study of familial dysautonomia (Riley-Day syndrome) in siblings. J. Neurol. Neurosurg. Psychiatry 27:131, 1964. 23. Bruns, O.: Aus den Gesellschaften. Neurol. Centralbl. 22:598, 1903. 24. Brunt, P. W., and McKusick, V. A.: Familial dysautonomia: a report of genetic and clinical studies, with a review of the literature. Medicine (Baltimore) 49:343, 1970.

1840

Diseases of the Peripheral Nervous System

25. Cantalupo, G., Alifano, P., Roberti, V., et al.: Rab-interacting lysosomal protein (RILP): the Rab7 effector required for transport to lysosomes. EMBO J. 20:683, 2001. 26. Cavanagh, N. P. C., Eames, R. A., Galvin, R. J., et al.: Hereditary sensory neuropathy with spastic paraplegia. Brain 102:79, 1979. 27. Cohen, P., and Solomon, N. H.: Familial dysautonomia: case report with autopsy. J. Pediatr. 46:663, 1955. 28. Cruchet, R., Petges, G., and Joulia: Un cas de maladie de Morvan chez un enfant de sept ans. Bull. Mem. Soc. Med. Chir. Bordeaux, 282, 1920. 29. Cuajungco, M. P., Leyne, M., Mull, J., et al.: Tissue-specific reduction in splicing efficiency of IKBKAP due to the major mutation associated with familial dysautonomia. Am. J. Hum. Genet. 72:749, 2003. 30. Dancis, J., and Smith, A. A.: Familial dysautonomia. N. Engl. J. Med. 274:207, 1966. 31. Davar, G., Shalish, C., Blumenfeld, A., and Breakfield, X. O.: Exclusion of p75NGFR and other candidate genes in a family with hereditary sensory neuropathy II. Pain 67:135, 1996. 32. Dawkins, J. L., Brahmbhatt, S., Auer-Grumbach, M., et al.: Exclusion of serine palmitoyltransferase long chain base subunit 2 (SPTLC2) as a common cause for hereditary sensory neuropathy. Neuromuscul. Disord. 12:656, 2002. 33. Dawkins, J. L., Hulme, D. J., Brahmbhatt, S. B., et al.: Mutations in SPTLC1, encoding serine palmitoyltransferase, long chain base subunit-1, cause hereditary sensory neuropathy type I. Nat. Genet. 27:309, 2001. 34. De Jonghe, P., Timmerman, V., FitzPatrick, D., et al.: Mutilating neuropathic ulcerations in a chromosome 3q13-q22 linked Charcot-Marie-Tooth disease type 2B family. J. Neurol. Neurosurg. Psychiatry 62:570, 1997. 35. Denny-Brown, D.: Hereditary sensory radicular neuropathy. J. Neurol. Neurosurg. Psychiatry 14:237, 1951. 36. Desautels, A., Turecki, G., Montplaisir, J., et al.: Identification of a major susceptibility locus for restless legs syndrome on chromosome 12q. Am. J. Hum. Genet. 69:1266, 2001. 37. Diamond, G. A., D’Amico, R. A., and Axelrod, F. B.: Optic nerve dysfunction in familial dysautonomia. Am. J. Ophthalmol. 104:645, 1987. 38. Donaghy, M., Hakin, R. N., Bamford, J. M., et al.: Hereditary sensory neuropathy with neurotrophic keratitis. Brain 110:563, 1987. 39. Dong, J., Edelmann, L., Bajwa, A. M., et al.: Familial dysautonomia: detection of the IKBKAPIVS20⫹6T to C and R696P mutations and frequencies among Ashkenazi Jews. Am. J. Med. Genet. 110:253, 2002. 40. Dyck, P. J.: Sural nerve biopsy studies in inherited neuropathy. Mayo Clin. Proc. 41:742, 1966. 41. Dyck, P. J.: Inherited neuronal degeneration and atrophy affecting peripheral sensory, and autonomic neurons. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 825, 1975. 42. Dyck, P. J.: Neuronal atrophy and degeneration predominantly affecting peripheral sensory and autonomic neuropathy. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 1557, 1984.

43. Dyck, P. J.: Neuronal atrophy and degeneration predominantly affecting peripheral sensory and autonomic neurons. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1065, 1993. 44. Dyck, P. J., and Bastron, J.: “Burning feet” as the only manifestation of dominantly inherited sensory neuropathy. Mayo Clin. Proc. 58:426, 1983. 45. Dyck, P. J., Kawamura, Y., Low, P. A., and Shimono, M.: The number and sizes of reconstructed peripheral autonomic, sensory and motor neurons in a case of dysautonomia. J. Neuropathol. Exp. Neurol. 27:741, 1978. 46. Dyck, P. J., Kennel, A. J., Magal, I. V., and Kraybill, E. N.: A Virginia kinship with hereditary sensory neuropathy: peroneal muscular atrophy and pes cavus. Mayo Clin. Proc. 40:685, 1965. 47. Dyck, P. J., and Klein, C. J.: Hereditary sensory and autonomic neuropathies. In Noseworthy, J. H. (ed.): Neurological Therapeutics: Principles and Practice. London, Martin Dunitz, p. 2188, 2003. 48. Dyck, P. J., Lais, A. C., Sparks, M. F., et al.: Nerve xenografts to apportion the role of axon and Schwann cell in myelinated fiber absence in hereditary sensory neuropathy, type II. Neurology 29:1215, 1979. 49. Dyck, P. J., Mellinger, J. F., Reagan, T. J., et al.: Not “indifference to pain” but varieties of hereditary sensory and autonomic neuropathy. Brain 106:373, 1983. 50. Dyck, P. J., and Ohta, M.: Neuronal atrophy and degeneration predominantly affecting peripheral sensory neurons. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 791, 1975. 51. Dyck, P. J., Stevens, J. C., O’Brien, P. C., et al.: Neuropathic arthropathy and recurring fractures with subclinical inherited neuropathy. Neurology 33:202, 1983. 52. Ehrhard, P. B., Ganter, U., Bauer, J., and Otten, U.: Expression of functional trk protooncogene in human monocytes. Proc. Nat. Acad. Sci. U. S. A. 90:5423, 1993. 53. Einarsdottir, E., Carlsson, A., Minde, J., et al.: A mutation in the nerve growth factor beta gene (NGFB) causes loss of pain perception. Hum. Mol. Genet. 13:799, 2004. 54. Ekbom, K. A.: Restless leg syndrome. Neurology 10:868, 1960. 55. Elliott, J. L., Kwon, J. M., Goodfellow, P. J., and Yee, W. C.: Hereditary motor and sensory neuropathy IIB: clinical and electrodiagnostic characteristics. Neurology 48:23, 1997. 56. England, A. C., and Denny-Brown, D.: Severe sensory changes, and trophic disorder, in peroneal muscular atrophy (Charcot-Marie-Tooth type). Arch. Neurol. Psychiatry 67:1, 1952. 57. Feindel, W.: Note on the nerve endings in a subject with arthropathy and congenital absence of pain. J. Bone Joint Surg. Br. 35:402, 1953. 58. Fellner, M. J.: Manifestations of familial autonomic dysautonomia: report of a case, with an analysis of 125 cases in the literature. Arch. Dermatol. 89:190, 1964. 59. Fogelson, M. H., Rorke, L. B., and Kaye, R.: Spinal cord changes in familial dysautonomia. Arch. Neurol. 17:103, 1967. 60. Gable, K., Han, G., Monaghan, E., et al.: Mutations in the yeast LCB1 and LCB2 genes, including those corresponding

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics

61. 62.

63. 64.

65.

66.

67.

68.

69. 70. 71.

72.

73.

74.

75. 76. 77.

78.

79. 80.

to the hereditary sensory neuropathy type I mutations, dominantly inactivate serine palmitoyltransferase. J. Biol. Chem. 277:10194, 2002. Giaccai, L.: Familial and sporadic neurogenic acro-osteolysis. Acta Radiol. (Stockh.) 38:17, 1952. Gimm, O., Greco, A., Hoang-Vu, C., et al.: Mutation analysis reveals novel sequence variants in NTRK1 in sporadic human medullary thyroid carcinoma. J. Clin. Endocrinol. Metab. 84:2784, 1999. Ginsberg, S. P.: Familial dysautonomia: the Riley-Day syndrome. Clin. Pediatr. (Phila.) 50:308, 1966. Gitlow, S. E., Bertani, L. M., Wilk, E., et al.: Excretion of catecholamine metabolites by children with familial dysautonomia. Pediatrics 46:513, 1970. Göbell, R., and Runge, W.: Zwei Fälle von familiärer Trophoneurose der unteren Extremitäten. Münch. Med. Wochenschr. 61:102, 1914. Göbell, R., and Runge, W.: Eine familiäre Trophoneurose der unteren Extremitäten. Arch. Psychiatr. Nervenkr. 57:297, 1917. Goebel, H. W., Veit, S., and Dyck, P. J.: Confirmation of virtual unmyelinated fiber absence in hereditary sensory neuropathy type IV. J. Neuropathol. Exp. Neurol. 39:670, 1980. Goldberg, M. F., Payne, J. W., and Brunt, P. W.: Ophthalmologic studies of familial dysautonomia: the Riley-Day syndrome. Arch. Ophthalmol. 80:732, 1968. Granger, M. E.: Sensory neuropathy with ulcerative mutilating acropathy. Neurology (Minneap.) 10:725, 1960. Greco, A., Villa, R., and Pierotti, M. A.: Genomic organization of the human NTRK1 gene. Oncogene 13:2463, 1996. Gupta, S., Campbell, D., Derijard, B., and Davis, R. J.: Transcription factor ATF2 regulation by the JNK signal transduction pathway. Science 267:389, 1995. Hageman, G., Hilhorst, B. G., and Rozeboom, A. R.: Is there involvement of the central nervous system in hereditary sensory radicular neuropathy? Clin. Neurol. Neurosurg. 94:49, 1992. Harrison, R. E., Bucci, C., Vieira, O. V., et al.: Phagosomes fuse with late endosomes and/or lysosomes by extension of membrane protrusions along microtubules: role of Rab7 and RILP. Mol. Cell. Biol. 23:6494, 2003. Hawkes, N. A., Otero, G., Winkler, G. S., et al.: Purification and characterization of the human elongator complex. J. Biol. Chem. 277:3047, 2002. Head, H.: Morvan’s disease. London Hosp. Gaz. Clin. Suppl. 10:5, 1903. Heller, I. H., and Robb, P.: Hereditary sensory neuropathy. Neurology (Minneap.) 5:15, 1955. Hempstead, B. L., Martin-Zanca, D., Kaplan, D. R., et al.: High-affinity NGF binding requires coexpression of the trk proto-oncogene and the low-affinity NGF receptor. Nature 350:678, 1991. Herget, T., Esdar, C., Oehrlein, S. A., et al.: Production of ceramides causes apoptosis during early neural differentiation in vitro. J. Biol. Chem. 275:30344, 2000. Hicks, E. P.: Hereditary perforating ulcer of the foot. Lancet 1:319, 1922. Holleran, W. M., Williams, M. L., Gao, W. N., and Elias, P. M.: Serine-palmitoyl transferase activity in cultured human keratinocytes. J. Lipid Res. 31:1655, 1990.

1841

81. Holmberg, C., Katz, S., Lerdrup, M., et al.: A novel specific role for IkappaB kinase complex-associated protein in cytosolic stress signaling. J. Biol. Chem. 277:31918, 2002. 82. Hooley, J. M., and Delgado, M. L.: Pain insensitivity in the relatives of schizophrenia patients. Schizophr. Res. 47:265, 2001. 83. Horvat, D.: Kongenitale sensorische neuropathie. Helv. Paediatr. Acta 29:555, 1974. 84. Hould, F., and Verret, S.: Neuropathie radiculaire héréditaire avec pertes de sensibilité: étude d’une famille CanadienneFrancaise. Laval Med. 38:454, 1967. 85. Houlden, H., King, R. H., Hashemi-Nejad, A., et al.: A novel TRK A (NTRK1) mutation associated with hereditary sensory and autonomic neuropathy type V. Ann. Neurol. 49:521, 2001. 86. Huang, Y. S., Marcel, Y. L., Vezina, C., et al.: Lecithin: cholesterol acyltransferase activity and fatty acid composition of erythrocyte phospholipids in Friedreich’s ataxia. Can. J. Neurol. Sci. 7:429, 1980. 87. Huizinga, J.: Hereditary acromelalgia (or “restless legs”). Acta Genet. Statist. Med. 7:121, 1957. 88. Hutchison, J. H., and Hamilton, W.: Familial dysautonomia in two siblings. Lancet 1:1216, 1962. 89. Indo, Y.: Molecular basis of congenital insensitivity to pain with anhidrosis (CIPA): mutations and polymorphisms in TRKA (NTRK1) gene encoding the receptor tyrosine kinase for nerve growth factor. Hum. Mutat. 18:462, 2001. 90. Indo, Y., Mardy, S., Miura, Y., et al.: Congenital insensitivity to pain with anhidrosis (CIPA): novel mutations of the TRKA (NTRK1) gene, a putative uniparental disomy, and a linkage of the mutant TRKA and PKLR genes in a family with CIPA and pyruvate kinase deficiency. Hum. Mutat. 18:308, 2001. 91. Indo, Y., Tsuruta, M., Hayashida, Y., et al.: Mutations in the TRKA/NGF receptor gene in patients with congenital insensitivity to pain with anhidrosis. Nat. Genet. 13:485, 1996. 92. Jackson, M.: Familial lumbo-sacral syringomyelia and the significance of developmental errors of the spinal cord and column. Med. J. Aust. 1:433, 1949. 93. Janzer, R. C., Spycher, M. A., Boltshauser, E., and Cserhatt, M.: A case of congenital non-progressive sensory neuropathy with tonic pupils. Clin. Neuropathol. 5:209, 1986. 94. Jedrzejowska, H., and Milczarek, H.: Recessive hereditary sensory neuropathy. J. Neurol. Sci. 29:371, 1976. 95. Jestico, J. V., Urry, P. A., and Efphimou, J.: An hereditary sensory and autonomic neuropathy transmitted as an Xlinked recessive trait. J. Neurol. Neurosurg. Psychiatry 48:1259, 1985. 96. Johnson, R. H., and Spalding, J. M.: Progressive sensory neuropathy in children. J. Neurol. Neurosurg. Psychiatry 27:125, 1964. 97. Jordens, I., Fernandez-Borja, M., Marsman, M., et al.: The Rab7 effector protein RILP controls lysosomal transport by inducing the recruitment of dynein-dynactin motors. Curr. Biol. 11:1680, 2001. 98. Jughenn, H., Krücke, W., and Wadulla, H.: Zur Frage der familiären Syringomyelie: klinisch-anatomische Untersuchungen über familiäre neuro-vasculäre Dystrophie der Extremitäten. Arch. Psychiatr. Nervenkr. 182:153, 1949.

1842

Diseases of the Peripheral Nervous System

99. Justice, J.: Congenital familial sensory neuropathy with anhidrosis. Arch. Neurol. 24:12, 1985. 100. Kaplan, D. R., Hempstead, B. L., Martin-Zanca, D., et al.: The trk proto-oncogene product: a signal transducing receptor for nerve growth factor. Science 252:554, 1991. 101. Klein, C. J.: Indifference rather than insensitivity to pain [comment]. Ann. Neurol. 53:417, 2003; author reply 53:418, 2003. 102. Klein, C. J., Cunningham, J. M., Atkinson, E. J., et al.: The gene for HMSN2C maps to 12q23-24: a region of neuromuscular disorders. Neurology 60:1151, 2003. 103. Klein, C. J., Wu, Y., Klein, D. M., et al.: SPTLC1 and RAB7 mutation analysis in dominantly inherited and cryptogenic sensory neuropathies [abstract]. Presented at the annual meeting of the American Academy of Neurology, San Francisco, 2004. 104. Kok, C., Kennerson, M. L., Spring, P. J., et al.: A locus for hereditary sensory neuropathy with cough and gastroesophageal reflux on chromosome 3p22-p24. Am. J. Hum. Genet. 73:632, 2003. 105. Kolesnick, R. N., and Kronke, M.: Regulation of ceramide production and apoptosis. Annu. Rev. Physiol. 60:643, 1998. 106. Kondo, K., and Horikawa, Y.: Genetic heterogeneity of hereditary sensory neuropathy. Arch. Neurol. 30:336, 1974. 107. Kuroiwa, Y., and Murai, Y.: Hereditary sensory radicular neuropathy: with special reference to conduction velocity study. Neurology (Minneap.) 14:574, 1964. 108. Kwon, J. M., Elliott, J. L., Yee, W. C., et al.: Assignment of a second Charcot-Marie-Tooth type II locus to chromosome 3q. Am. J. Hum. Genet. 57:853, 1995. 109. Lafreniere, R.G., MacDonald, M. L., Dube, M. P., et al.: Identification of a novel gene (HSN2) causing hereditary sensory and autonomic neuropathy type II through the study of Canadian genetic isolates. Am. J. Hum. Genet. 74:1064, 2004. 110. Landrieu, P., Said, G., and Allaire, C.: Dominantly transmitted congenital indifference to pain. Ann. Neurol. 27:574, 1990. 111. Langer, J., Goebel, H. H., and Veit, S.: Eccrine sweat glands are not innervated in hereditary sensory neuropathy type IV. Acta Neuropathol. (Berl.) 54:199, 1981. 112. Leyne, M., Mull, J., Gill, S. P., et al.: Identification of the first non-Jewish mutation in familial dysautonomia. Am. J. Med. Genet. 118A:305, 2003. 113. Linde, L. M.: Dysautonomia: case report of a variant. J. Pediatr. 46:453, 1955. 114. Low, P. A., Burke, W. J., and McLeod, J. G.: Congenital sensory neuropathy with selective loss of small myelinated fibers. Ann. Neurol. 3:179, 1978. 115. Maayan, C., Kaplan, E., Schacher, S., et al.: The incidence of familial dysautonomia in Israel 1977–1981. J. Pediatr. 32:106, 1987. 116. Mahloudji, M., Brunt, P. W., and McKusick, V. A.: Clinical neurological aspects of familial dysautonomia. J. Neurol. Sci. 11:383, 1970. 117. Mandon, E. C., Ehses, I., Rother, J., et al.: Subcellular localization and membrane topology of serine palmitoyltransferase, 3-dehydrosphinganine reductase, and sphinganine N-acyltransferase in mouse liver. J. Biol. Chem. 267:11144, 1992.

118. Mardy, S., Miura, Y., Endo, F., et al.: Congenital insensitivity to pain with anhidrosis: novel mutations in the TRKA (NTRK1) gene encoding a high-affinity receptor for nerve growth factor. Am. J. Hum. Genet. 64:1570, 1999. 119. McMurray, G. A.: Experimental study of a case with insensitivity to pain. Arch. Neurol. Psychiatry 64:650, 1950. 120. Miller, R. G., Nielsen, S. L., and Sumner, A. J.: Hereditary sensory neuropathy and tonic pupils. Neurology 26:931, 1976. 121. Moses, S. W., Rotem, Y., Jagoda, N., et al.: A clinical, genetic and biochemical study of familial dysautonomia in Israel. Isr. J. Med. Sci. 3:358, 1967. 122. Munro, M.: Sensory radicular neuropathy in a deaf child. Br. Med. J. 1:541, 1956. 123. Murray, T. J.: Congenital sensory neuropathy. Brain 96:387, 1973. 124. Nebel, O. T., Fornes, M. F., and Castell, D. O.: Symptomatic gastroesophageal reflux: incidence and precipitating factors. Am. J. Dig. Dis. 21:953, 1976. 125. Nélaton, M.: Affection singulière des os du pied. Gaz. Hop. Civils Militaires 4:13, 1852. 126. Nicholson, G. A., Dawkins, J. L., Blair, I. P., et al.: The gene for hereditary sensory neuropathy type I (HSN-I) maps to chromosome 9q22.1-q22.3. Nat. Genet. 13:101, 1996. 127. Nicholson, G. A., Dawkins, J. L., Blair, I. P., et al.: Hereditary sensory neuropathy type I: haplotype analysis shows founders in southern England and Europe. Am. J. Hum. Genet. 69:655, 2001. 128. Nordborg, C., Conradi, N., Sourander, P., and Westerberg, B.: A new type of non-progressive sensory neuropathy in children with atypical dysautonomia. Acta Neuropathol. (Berl.) 55:135, 1981. 129. O’Brien, B., Jackson, R., Tang-Wai, R., et al.: Hereditary sensory neuropathy: a case with pain and temperature dissociation. Can. J. Neurol. Sci. 7:73, 1980. 130. Ogden, T. E., Robert, F., and Carmichael, E. A.: Some sensory syndromes in children: indifference to pain and sensory neuropathy. J. Neurol. Neurosurg. Psychiatry 22:267, 1959. 131. Ogryzlo, M. A.: A familial peripheral neuropathy of unknown etiology resembling Morvan’s disease. Can. Med. Assoc. J. 54:547, 1946. 132. Ohta, M., Ellefson, R. D., Lambert, E. H., and Dyck, P. J.: Hereditary sensory neuropathy, type II: clinical, electrophysiologic, histologic, and biochemical studies of a Quebec kinship. Arch. Neurol. 29:23, 1973. 133. Ondo, W. G., Vuong, K. D., and Wang, Q.: Restless legs syndrome in monozygotic twins: clinical correlates. Neurology 55:1404, 2000. 134. Pallis, C., and Schneeweiss, J.: Hereditary sensory radicular neuropathy. Am. J. Med. 32:110, 1962. 135. Partsch, H.: Hereditäre sensorische Neuropathie (DennyBrown): familiäre Akroosteolyse. Wien. Klin. Wochenschr. 82:129, 1970. 136. Pearson, J., and Pytel, B. A.: Quantitative studies of sympathetic ganglia and spinal cord intermediolateral gray columns in familial dysautonomia. J. Neurol. Sci. 39:47, 1978. 137. Pearson, J., Brandeis, L., and Goldstein, M.: Tyrosine hydroxylase immunoreactivity in familial dysautonomia. Science 206:71, 1979.

HSANs: Clinical Features, Pathologic Classification, and Molecular Genetics 138. Perry, D. K., Carton, J., Shah, A. K., et al.: Serine palmitoyltransferase regulates de novo ceramide generation during etoposide-induced apoptosis. J. Biol. Chem. 275:9078, 2000. 139. Pinsky, L., and DiGeorge, A. M.: Congenital familial sensory neuropathy with anhidrosis. J. Pediatr. 68:1, 1966. 140. Reimann, H. A., McKechnie, W. G., and Stanisavljevic, S.: Hereditary sensory radicular neuropathy and other defects in a large family. Am. J. Med. 25:573, 1958. 141. Riley, C. M., Day, R. L., Greeley, D. M., and Langford, W. S.: Central autonomic dysfunction with defective lacrimation. I. Report of five cases. Pediatrics 3:468, 1949. 142. Riley, C. M., and Moore, R. H.: Familial dysautonomia differentiated from related disorders: case report and discussions of current concepts. Pediatrics 37:435, 1966. 143. Riley, H. A.: Syringomyelia or myelodysplasia. J. Nerv. Ment. Dis. 72:1, 1930. 144. Rizzo, J. F., Lessell, S., and Liebman, S. D.: Optic atrophy in familial dysautonomia. Am. J. Ophthalmol. 102:463, 1986. 145. Roddier, K., Thoma, T., Marleau, G., et al.: Two distinct mutations are responsible for the higher prevalence of hereditary sensory and autonomic neuropathy type II (HSAN2) in French Canadians [abstract]. Poster presented at the 53rd annual meeting of the American Society of Human Genetics, Los Angeles, California, November 4–8, 2003. 146. Rosenthal, S. H., Porter, K. A., and Coffey, B.: Pain insensitivity in schizophrenia: case report and review of the literature. Gen. Hosp. Psychiatry 12:319, 1990. 147. Schambelan, M., Sebastian, A., and Rector, F. C. Jr.: Mineralocorticoid-resistant renal hyperkalemia without salt wasting (type II pseudohypoaldosteronism): role of increased renal chloride reabsorption. Kidney Int. 19:716, 1981. 148. Schilder, P., and Stengel, E.: Asymbolia for pain. Arch. Neurol. Psychiatry 25:598, 1931. 149. Schoene, W. C., Asbury, A. K., Lanström, K. E., and Masters, R.: Hereditary sensory neuropathy: a clinical and ultrastructural study. J. Neurol. Sci. 11:463, 1970. 150. Schultze, F.: Familiar auftrendes Malumperforans der Fusse (familiare lumbale syringomyelia?). Dtsch. Med. Wochenschr. 43:545, 1917. 151. Schut, J. W.: Hereditary ataxia: clinical study through six generations. Arch. Neurol. Psychiatry 63:535, 1950. 152. Schwartz, J. P., and Breakfield, X. O.: Altered nerve growth factor in fibroblasts from patients with familial dysautonomia. Proc. Natl. Acad. Sci. U. S. A. 77:1154, 1980. 153. Shatzky, S.: Congenital insensitivity to pain with anhidrosis (CIPA) in Israeli-Bedouins: genetic heterogeneity, novel mutations in the TRKA/NGF receptor gene, clinical findings, and results of nerve conduction studies [see comment]. Am. J. Med. Genet. 92:353, 2000. 154. Shimabukuro, M., Higa, M., Zhou, Y. T., et al.: Lipoapoptosis in beta-cells of obese prediabetic fa/fa rats: role of serine palmitoyltransferase overexpression. J. Biol. Chem. 273:32487, 1998. 155. Sinnreich, M., Klein, C. J., Daube, J. R., et al.: Chronic immune sensory polyradiculopathy, a potentially treatable sensory ataxia. Neurology (in press). 156. Slaugenhaupt, S. A., Blumenfeld, A., Gill, S. P., et al.: Tissuespecific expression of a splicing mutation in the IKBKAP gene causes familial dysautonomia. Am. J. Hum. Genet. 68:598, 2001.

1843

157. Smeyne, R. J., Klein, R., Schnapp, A., et al.: Severe sensory and sympathetic neuropathies in mice carrying a disrupted trk/NGF receptor gene. Nature 368:246, 1994. 158. Smith, A. A., and Dancis, J.: Physiologic studies in familial dysautonomia [abstract]. J. Pediatr. 63:838, 1963. 159. Smith, A. A., Taylor, T., and Wortis, S. B.: Abnormal catecholamine metabolism in familial dysautonomia. N. Engl. J. Med. 268:705, 1963. 160. Smith, R. S., Maddox, S. F., and Collins, B. E.: Congenital alacrima. Arch. Ophthalmol. 79:45, 1968. 161. Solitare, G. B., and Cohen, G. S.: Peripheral autonomic nervous system lesions in congenital or familial dysautonomia: Riley-Day syndrome. Neurology (Minneap.) 15:321, 1965. 162. Spillane, J. D., and Wells, C. E. C.: Acrodystrophic Neuropathy: A Critical Review of the Syndrome of Trophic Ulcers, Sensory Neuropathy, and Bony Erosion, Together with an Account of 16 Cases in South Wales. London, Oxford University Press, 1969. 163. Spring, P., Ing, A. J., Nicholson, G. A., et al.: Autosomal dominant hereditary sensory neuropathy with gastro-oesophageal reflux and cough: clinical features of a family. J. Neurol. Sci. 199(Suppl.):S64, 2002. 164. Stogbauer, F., Young, P., Kuhlenbaumer, G., et al.: Autosomal dominant burning feet syndrome. J. Neurol. Neurosurg. Psychiatry 67:78, 1999. 165. Sugarman, E., and Allitto, B. A.: Familial dysautonomia mutation frequency: clinical testing of greater than 2700 specimens confirms high frequency in Ashkenazi Jews. Am. J. Hum. Genet. 71(Suppl.):387, 2002. 166. Swanson, A. G.: Congenital insensitivity to pain with anhydrosis. Arch. Neurol. 8:299, 1963. 167. Swanson, A. G., Buchan, G. C., and Alvord, E. C. Jr.: Anatomic changes in congenital insensitivity to pain: absence of small primary sensory neurons in ganglia, roots, and Lissauer’s tract. Arch. Neurol. 12:12, 1965. 168. Sweeney, B. P., Jones, S., and Langford, R. M.: Anaesthesia in dysautonomia: further complications. Anaesthesia 40:783, 1985. 169. Thévenard, A.: L’acropathie ulcéro-mutilante familiale. Rev. Neurol. (Paris) 74:193, 1942. 170. Thévenard, A.: Travaux originaux: l’acropathie ulcéromutilante familiale. Acta Neurol. Belg. 53:1, 1953. 171. Thévenard, A., and Coste, M.: Syringomyélie lombosacrée familiale probable et spina-bifida occulta sacré. Rev. Neurol. (Paris) 63:195, 1935. 172. Thomas, P. K., Marques, W. J., Davis, M. B., et al.: The phenotypic manifestations of chromosome 17p11.2 duplication. Brain 120:465, 1977. 173. Thrush, D. C.: Congenital insensitivity to pain: a clinical, genetic and neurophysiological study of four children from the same family. Brain 96:369, 1973. 174. Tocantins, L. M., and Reimann, H. A.: Perforating ulcers of feet, with osseous atrophy in a family with other evidence of dysgenesis (hare lip, cleft palate): an instance of probable myelodysplasia. J. Am. Med. Assoc. 112:2251, 1939. 175. Toscano, E., Simonati, A., Indo, Y., et al.: No mutation in the TRKA (NTRK1) gene encoding a receptor tyrosine kinase for nerve growth factor in a patient with hereditary sensory and autonomic neuropathy type V [see comment]. Ann. Neurol. 52:224, 2002.

1844

Diseases of the Peripheral Nervous System

176. Turkington, R. W., and Stiefel, J. W.: Sensory radicular neuropathy. Arch. Neurol. 12:19, 1965. 177. Van Epps, C., and Kerr, H. D.: Familial lumbosacral syringomyelia. Radiology 35:160, 1940. 178. Vance, J. M., Speer, M. C., Stajich, J. M., et al.: Misclassification and linkage of hereditary sensory and autonomic neuropathy type I as Charcot-Marie-Tooth disease, type 2B [comment]. Am. J. Hum. Genet. 59:258, 1996. 179. Vardy, P. A., Greenberg, L. W., Kachel, C., and Falewski de Leon, G.: Congenital insensitivity to pain with anhydrosis. Am. J. Dis. Child. 133:1153, 1979. 180. Verhoeven, K., Coen, K., De Vriendt, E., et al.: SPTLC1 mutation in twin sisters with hereditary sensory neuropathy type I. Neurology 62:1001, 2004. 181. Verhoeven, K., De Jonghe, P., Coen, K., et al.: Mutations in the small GTP-ase late endosomal protein RAB7 cause Charcot-Marie-Tooth type 2B neuropathy. Am. J. Hum. Genet. 72:722, 2003. 182. Vieira, O. V., Bucci, C., Harrison, R. E., et al.: Modulation of Rab5 and Rab7 recruitment to phagosomes by phosphatidylinositol 3-kinase. Mol. Cell. Biol. 23:2501, 2003.

183. Vignon, G., Mégard, M., and Marin, A.: Une observation anatomo-clinique d’acropathie ulcéro-mutilante. Presse Med. 64:1954, 1956. 184. Wadia, N. H., and Dastur, D. K.: Congenital sensory neuropathy. World Neurol. 1:409, 1960. 185. Walker, C. H. M.: Sensory radicular neuropathy: report on two cases. Great Ormond Str. J. 10:72, 1955–1956. 186. Wallace, D. C.: Hereditary sensory radicular neuropathy: a family study. Arch-dall Medical Monograph No 8. Sydney, Australian Medical Publishing Company, p. 13, 1970. 187. Weinshilboum, R. M., and Axelrod, J.: Reduced plasma dopamine-␤-hydroxylase activity in familial dysautonomia. N. Engl. J. Med. 285:938, 1971. 188. Whitaker, J. N., Falchuck, Z. M., Engel, W. K., et al.: Hereditary sensory neuropathy: association with increased synthesis of immunoglobulin A. Arch. Neurol. 30:359, 1974. 189. Wilson, F. H., Disse-Nicodeme, S., Choate, K. A., et al.: Human hypertension caused by mutations in WNK kinases. Science 293:1107, 2001. 190. Winkelmann, R. K., Lambert, E. H., and Hayles, A. B.: Congenital absence of pain: report of a case and experimental studies. Arch. Dermatol. 85:325, 1962.

J. Inborn Errors of Metabolism with Peripheral Nerve Involvement

79 Lysosomal and Peroxisomal Disorders P. K. THOMAS AND HANS H. GOEBEL

Lysosomal Disorders Sphingolipidoses The Lysosomal Leukodystrophies Other Lysosomal Sphingolipidoses Miscellaneous Disorders The Neuronal Ceroid-Lipofuscinoses Chédiak-Higashi Syndrome Glycogenosis Type II Wolman’s Disease

Mucopolysaccharidoses and Oligosaccharidoses Mucolipidosis IV

Peroxisomal Disorders Adrenoleukodystrophy and Adrenomyeloneuropathy Pathology Pathogenesis

Lysosomal and peroxisomal—as well as mitochondrial— disorders are inherited single-system multiorgan diseases. Although essentially organelle disorders, the clinical and pathologic effect is widespread. A multitude of clinical symptoms are confirmed by ultrastructural and biochemical abnormalities in various organs, including the central nervous system (CNS) and the peripheral nervous system (PNS). In fact, it was the combination of electron microscopic and biochemical abnormalities that led to the diagnosis and classification of individual lysosomal and peroxisomal disorders. Because these diseases often lead to premature death, a precise diagnosis is desirable for reliable genetic counseling and eventual prenatal diagnosis, and potentially effective treatment with bone marrow transplantation170,171,238 or enzyme replacement therapy.46 Transgenic mouse models278 and spontaneous animal models have now been created and vigorously explored to develop further therapeutic strategies. Involvement in lysosomal and peroxisomal disorders of the PNS occurs at different levels, such as the molecular, biochemical, morphologic, electrophysiologic, and clinical levels.

Treatment Neonatal Adrenoleukodystrophy Refsum Disease Clinical Aspects Pathology Infantile Refsum Disease Peroxisomal 2-Methylacyl-CoA Racemase Deficiency

Lysosomal Disorders Lysosomal disorders are marked by hereditary defects in the lysosomal catabolism of intracellular physiologic constituents and compounds such as sphingolipids, complex carbohydrates (e.g., glycolipids, glycosaminoglycans, glycoproteins), lipids, glycogen, and proteins. The lysosomal defect may reside within the lysosomal compartment (e.g., being a hydrolyzing lysosomal enzyme); in the lysosomal membrane (e.g., lysosome-associated membrane protein 2, causing Danon disease, or deficiency of the transmembrane protein mucolipin in mucolipidosis IV25); or in the prelysosomal mannose 6-phosphate pathway (e.g., in I-cell disease). The functional and biochemical significance of the respective physiologic lysosome-related factor—a lysosomal enzyme, an intralysosomal protein, a transmembrane protein, or the like—determines the functional involvement of various organs in the respective individual lysosomal disorder. Morphologic manifestations may be more widespread than functional deficits, occasionally even ubiquitous 1845

1846

Diseases of the Peripheral Nervous System

in almost any cell type and tissue (e.g., in the neuronal ceroid-lipofuscinoses [NCLs]). Thus morphologic and clinical manifestations may not always be correlated. The widespread morphologic changes, in conjunction with biochemical and molecular investigations, lead to success in making the diagnosis using the least invasive tissue biopsy, performed on tissues outside of the CNS, or even of the PNS.95,96 The morphologic hallmark of a lysosomal disorder is the accumulation of intralysosomal material, often well-defined chemical compounds. Such accruing intralysosomal compounds, especially sphingolipids, may produce a characteristic, sometimes lysosomal disorder–specific ultrastructure, which had served to confirm the diagnosis of a lysosomal disease in the prebiochemical and premolecular era. Because both neurons and Schwann cells are components of the PNS, morphologic manifestations of a lysosomal disorder in either one or both of these cell types of the PNS may be encountered.94,102,107,307 Accumulation of lysosomal disorder–related compounds or hydrolyzing enzyme substrates within perikarya of anterior horn cells or within neuronal perikarya of sensory and autonomic ganglia may impair both axoplasmic transport and functions of the respective nerve cells, resulting in their degeneration and disappearance. A not infrequent, nonspecific ultrastructural feature in lysosomal disorders is mild expansion of unmyelinated terminal axons by accumulating mitochondria and dense bodies,70,307 and occasionally also by lysosomal disorder–specific inclusions (e.g., membranous cytoplasmic bodies in gangliosidoses50 and mucolipidosis IV101) (Figs. 79–1 and 79–2). Impaired function of the PNS and respective abnormal ancillary data may originate from such intra-axonal and intraperikaryonal accumulation of lysosomal residual bodies. Another pathologic process may

affect Schwann cells, resulting in accretion of disease-characteristic lysosomal residual bodies in Schwann cells and subsequent damage to myelin and demyelination (see The Lysosomal Leukodystrophies later). Lysosomal accretion may also occur in Schwann cells of unmyelinated axons and in satellite cells to nerve cells in peripheral ganglia. Another electron microscopic lysosomal manifestation may be noted in mural cells of endoneurial and epineurial vessels, foremost in Fabry’s disease but also in many other lysosomal diseases. Because macrophages carry the same lysosomal defect as any other cell of the PNS, they may show lysosomal storage, an additional feature in the nerves of such patients. Moreover, certain lysosomal diseases affect perineurial cells (e.g., mucolipidosis IV).101 The extensive use of skin and conjunctival biopsies containing numerous small nerve fascicles has revealed a multitude of ultrastructural details in lysosomal disorders.96 These diverse morphologic manifestations in the PNS produce two major sequelae, axonal degeneration and/or demyelination-remyelination. Secondary damage to PNS components may derive from ischemia owing to primary lysosomal storage in mural vascular cells (e.g., Fabry’s disease) or from compression (e.g., carpal tunnel syndrome in mucopolysaccharidoses [MPSs]). Lysosomal disorders, as catabolic processes, are still classified according to accumulating biochemical compounds caused by various defects within the lysosomal system of the cell. Apart from eponyms (e.g., Gaucher’s and Fabry’s diseases, etc.), they encompass sphingolipids (sphingolipidoses), mucopolysaccharides (MPSs), oligosaccharides (mannosidoses, etc.), lipopigments (NCLs), lipid (Wolman’s disease), and glycogen (type II glycogenosis). Another conceivable classification would be either according to the site of the defect within the entire lysosomal

FIGURE 79–1 Gangliosidosis. An unmyelinated axon is considerably enlarged by mitochondria and mitochondriaderived dense bodies as well as by lamellar bodies (arrows) in a dermal nerve fascicle.

Lysosomal and Peroxisomal Disorders

1847

within the CNS and PNS, here in neuron-covering cells (i.e., Schwann cells) or within the neuronoaxonal cellular compartments. The former are so-called lysosomal leukodystrophies; the latter rarely result in prominent PNS disorders, with the exception of the well-recognized spinal muscular atrophy from juvenile GM2 gangliosidosis.

The Lysosomal Leukodystrophies

FIGURE 79–2 Niemann-Pick disease type C. In a dermal nerve fascicle, two unmyelinated axons are expanded by numerous dense bodies, a nonspecific feature in lysosomal diseases.

pathway, from endoplasmic reticulum to tertiary lysosome, or the type of lysosomal defect (e.g., lysosomal enzyme protein, activator protein, cofactor, transmembrane protein, transfer or trafficking factors). Lysosomal disorders may encompass groups of conditions that result from the blockage of individual catabolic intralysosomal steps, often appearing in a sequential order when affecting sphingolipids,258 mucopolysaccharides,218 oligosaccharides, or sialic acid compounds. Thus such closely related sequential degradative steps may explain clinical and morphologic similarities among different lysosomal disorders of the same biochemical group (e.g., sphingolipids or mucopolysaccharides).

SPHINGOLIPIDOSES Sphingolipids are complexes of ceramide and varying carbohydrates. They form an interconnected system in which catabolic degradation occurs in an orderly fashion.258 Defects in these degradative pathways result in respective sphingolipidoses and their morphologic manifestation

Leukodystrophies are hereditary metabolic or structural disorders affecting the myelin sheaths of the CNS and/or the PNS. Because myelin is composed of different lipids and different proteins, with the former being present in the PNS and the CNS, though to different degrees, such myelin lipid-related leukodystrophies affect both the PNS and the CNS, whereas mutant myelin proteins, so far, only seem to clinically and morphologically affect either the PNS or the CNS (e.g., mutant proteolipid protein the central myelin and mutant peripheral myelin protein 22 in the peripheral myelin, resulting in Pelizaeus-Merzbacher disease and in autosomal dominant Charcot-Marie-Tooth disease type 1, respectively). Among the lysosomal disorders, therefore, the two lysosomal leukodystrophies affecting both the CNS and the PNS are disorders of myelin sphingolipids: sulfatides in metachromatic leukodystrophy (MLD), and galactocerebroside in globoid cell leukodystrophy (Krabbe’s disease). Metachromatic Leukodystrophy Historical Development. MLD comprises a group of genetically determined disorders that primarily affect the nervous system. They are characterized by myelin degeneration in both the CNS and the PNS associated with the accumulation of galactosylsulfatide and other sulfatide lipids. They have therefore been classified as sulfatide lipidoses and are related to a deficiency of arylsulfatase A (ASA; cerebroside sulfatase). Late infantile, juvenile, and adult forms are recognized. An additional rare disorder that has been termed MLD variant is associated with multiple sulfatase deficiencies, and another, the AB variant, is related to a deficiency of activator protein. The term metachromatic leukodystrophy was first employed by Einarson and Neel.75 It describes the morphologic and histochemical finding of extensive and diffuse demyelination of the cerebral and cerebellar white matter with reactive fibrillar gliosis in association with the accumulation of a granular lipid substance in the nervous system and certain viscera, which stain metachromatically with acid aniline dyes. Historically, the first case of diffuse sclerosis corresponding to MLD, with the accumulation in the CNS of “prelipoid substances,” was described by Alzheimer5 and Nissl221 in 1910. Witte,322 in 1921, observed lipid deposition in the kidneys.

1848

Diseases of the Peripheral Nervous System

Greenfield112 initially suggested that the peripheral nerves are involved on the basis of the clinical observation of depressed tendon reflexes. This was later confirmed by Jacobi,140 Norman,222 Bertrand and co-workers,37 and others, both from pathologic studies on postmortem and nerve biopsy material1,4,30,121,140,223,323 and from nerve conduction studies.87,323 The first demonstration of the disorder in life by nerve biopsy was made by Thieffry and Lyon284 in a 4-year-old child with a progressive cerebral disorder. Austin 16,17 had earlier demonstrated that metachromatic granules may be detected in urine sediment and had pointed out that increased sulfatide could be found in urine. Diagnosis is now best made by the assay of ASA in leukocytes and fibroblasts. Late Infantile Form. The late infantile form is the most common variety of MLD. The incidence in northern Sweden was found to be 1:40,000.117 The disease begins between the ages of 1 and 4 years, usually between 15 and 18 months, with normal antecedent development. The course of the disease was divided into four stages by Hagberg (Fig. 79–3).119 After a normal postpartum development and early infancy, the first symptoms appear at the age of 15 to 18 months. Among the first signs are walking difficulties, weak valgus feet, an unsteady gait, and finally the inability to stand without support. Neurologic examination reveals weakness of the feet and hypotonia of the legs, and sometimes also of the arms. At the same time, retardation of mental development is observed. In the second phase, ataxia and hypotonia are more pronounced and flaccid diplegia or tetraplegia arises. The patient is unable to sit without support. The common symptoms of neuralgic pain and muscle cramps in the extremities may necessitate administration of large doses of analgesics. Tendon reflexes are diminished or absent. Cerebral deficiencies increase and include mental deterioration, spasticity, and tonic seizures.

In the third stage, the symptoms of CNS deterioration predominate. Speech impairment, cerebellar disturbances, increasing apathy as a result of loss of mental function, and loss of vision occur frequently. The patient becomes unable to stand or sit. The neuralgic pain decreases. In the fourth and final stage, the patient shows indications of decerebrate rigidity and bulbar symptoms, blindness resulting from optic atrophy, deafness, and hypertonic seizures (Fig. 79–4). Although these stages define the usual progress of the disorder, occasional atypical presentations can produce diagnostic difficulty. The disorder may begin as a neuropathy without signs of cerebral involvement,66,323 including mental retardation. It may present as a hypotonic paraparesis or tetraparesis with resemblances to a myopathy or infantile spinal muscular atrophy,118 and an elevated cerebrospinal fluid protein content may raise the possibility of an inflammatory polyneuropathy. Juvenile and Adult Cases. Cases with onset between the ages of 3 and 21 years have been classified as juvenile MLD, but clinically these may at times be difficult to separate from the adult-onset form. It seems likely, however, that the three forms are distinct. The clinical features of the juvenile form are, in general, intermediate between the late infantile and adult-onset varieties.126 Motor abnormalities tend to be the dominant early feature in the late infantile form. The adult form, which has arbitrarily been defined as having an onset after the age of 21 years, is characterized by a progressive intellectual deterioration with behavioral changes, often of a psychotic nature.241 Some cases with onset in late adolescence have also been accepted as being of the adult type. Neurologic signs may be absent,216 but bilateral pyramidal signs, disturbances of micturition,128 extrapyramidal and cerebellar signs, and visual loss120 have been described. Decreased nerve conduction velocity has been documented, indicating peripheral nerve involvement, including a

FIGURE 79–3 Stages of the clinical development of infantile metachromatic leukodystrophy. Diagram of the clinical findings in the infantile form according to the stages defined by Hagberg. (Data from Hagberg, B.: Clinical symptoms, signs and tests in metachromatic leukodystrophy. In Folch Pi, J., and Bauer, H. J. [eds.]: Brain Lipids and Lipoproteins and the Leucodystrophies. Amsterdam, Elsevier, p. 134, 1963.)

Lysosomal and Peroxisomal Disorders

1849

diagnosis. Molecular studies are now another reliable diagnostic procedure. Reduced nerve conduction velocity (NCV) is present in most of the late infantile cases and was evident in five of the seven cases reported by Fullerton.87 Sensory action potentials may be reduced in amplitude or absent.290 The severity of the reduction in motor conduction velocity may be marked, sometime with values of 10 to 20 m/s.87,290 Slowing to a similar degree may also be demonstrable in the juvenile cases.288 A slight reduction in motor NCV and in the amplitude of sensory nerve action potentials is detectable in some adult cases (personal observations); NCV may even be normal in adult patients.55 Nerve biopsy constitutes a reliable procedure for the diagnostic confirmation of MLD but is now rarely necessary. The detailed changes are discussed later, but metachromatic material is probably demonstrable in all cases.61,121,191,312 Rectal biopsy has yielded positive results, but is less reliable.293

FIGURE 79–4 Patient with infantile metachromatic leukodystrophy, showing hypertonic posture. (Courtesy of Dr. F. Vasella.)

possible preclinical case.242 Rarely, juvenile-onset cases in which the manifestations are solely those of a chronic progressive demyelinating peripheral neuropathy have been encountered.57,60,81,287 LABORATORY INVESTIGATIONS. Intracellular deposits of metachromatic material may be found on microscopic examination of urinary sediment,16,17,150 and increased amounts of sulfatide are excreted.122 A reduced level of ASA activity is demonstrable in blood leukocytes, serum, and cultured skin fibroblasts. Leukocyte and serum assays are the most satisfactory method for establishing the

PATHOLOGY. The major site of pathology in MLD is in the white matter of the CNS. This discussion is confined to peripheral nerve involvement. A number of abdominal viscera are also affected, including the kidneys and gallbladder. The peripheral nerves in late infantile cases show a reduction in the myelinated fiber population and evidence of segmental demyelination and remyelination.65,73 The loss of fibers, as discussed later, is usually less obvious in the juvenile and adult cases. Computer-assisted morphometry of biopsied sural nerves in different clinical forms suggested complete remodeling of the nerve fiber population.28 Small onion bulbs (concentric Schwann cell proliferation) may be encountered. The characteristic feature of the peripheral nerve pathology in MLD is the accumulation of granules of 0.5 to 1 ␮m diameter in the perinuclear cytoplasm of Schwann cells (Fig. 79–5). These stain metachromatically (i.e., brownish) with acid cresyl violet or toluidine blue. Material that stains similarly is also seen in Remak cells, within macrophages, and in the vicinity of endoneurial capillaries (Fig. 79–6). No definite correlation can be observed between the occurrence of segmental demyelination and the presence of inclusions.65,312 Electron microscopy was first undertaken by Webster312 in 1962 on sural nerve biopsy specimens. This preceded ultrastructural observations on brain biopsy material. Subsequent ultrastructural observations on peripheral nerve have been undertaken by numerous authors and have included investigations in late infantile, juvenile, and adultonset cases.8,13,49,62–64,97,105,166,191,237,290,293 It has been demonstrated that a variety of similar inclusions occur in both the PNS and CNS and that metachromasia is probably related to membrane-bound lamellated inclusions containing material exhibiting a periodicity of 5.6 to 5.8 nm.113,114,255 These

1850

Diseases of the Peripheral Nervous System

FIGURE 79–5 Infantile metachromatic leukodystrophy. Longitudinal section of peripheral nerve stained with acid cresyl violet, showing perinuclear accumulation of brown metachromatically stained granules in Schwann cells. N ⫽ Schwann cell nuclei. (Courtesy of Dr. A. Bischoff.)

inclusions are associated with acid phosphatase activity and are therefore classifiable as lysosomes.254 In peripheral nerve, Thomas and collaborators290 described two types of inclusion material. Both are consistently membrane bound. The first was seen in Schwann cells associated with myelinated axons and consisted of myelin figures composed of concentric lamellar material with major dense lines with an 8-nm spacing and an intervening electron-lucent space bisected by a line of lesser electron density. This material has a spacing that is quite different from that of other inclusions and is possibly derived from myelin breakdown. Myelin figures had been noted by Webster312 and by Cravioto and co-workers.62 All the other inclusions contained material with periodicity of 5.8 nm that, as already stated, corresponds to the metachromatic inclusion material detected by light microscopy. The dense lines represent layers of polar lipid separated from each other by an aqueous phase. This material is encountered in inclusions of three types290: 1. Zebra bodies. These contain groups of lamellae with a 5.8-nm spacing and are observed in Schwann cells and macrophages (Fig. 79–7).

FIGURE 79–6 Infantile metachromatic leukodystrophy. Semithin transverse section of peripheral nerve stained with acid cresyl violet, showing reduced density of myelinated nerve fibers. The Schwann cells associated with a number of the myelinated fibers contain accumulations of granular or globular material, which consists of dark orthochromatic globules (CM) and metachromatically stained bodies (TS). These are also observed in relation to unmyelinated fibers (asterisks). (Courtesy of Dr. A. Bischoff.)

2. Tuffstone bodies. These electron-dense structures are most numerous in Schwann cells, associated both with myelinated and unmyelinated axons. They contain irregularly oriented material, again, with a 5.8-nm spacing, and their name indicates their resemblance to volcanic limestone (Figs. 79–8 and 79–9). 3. Prismatic inclusions. The lamellar inclusions correspond to those described in the CNS by Grégoire and associates114 and consist of stacked discs with a 5.8-nm separation. Such inclusions were described in peripheral nerve by Anzil and colleagues,8 Weller and CervósNavarro,315 and Thomas and co-workers.290 Martin and associates191 conducted a comparative study on 10 cases, 6 of which were of the late infantile type, 2 of juvenile onset, and 2 of adult onset. These researchers concluded that demyelination was most pronounced in the late infantile form and in distal nerve. The juvenile and adult-onset cases show fewer Schwann cell inclusions and fewer macrophages, in keeping with their slower tempo.

Lysosomal and Peroxisomal Disorders

1851

FIGURE 79–7 Electron micrograph of a zebra body within a disintegrating endoneurial macrophage from a patient with late infantile metachromatic leukodystrophy. Inset, Inclusion material at higher magnification (⫻119,600). The groups of lamellae have a periodicity of 5.8 nm. (From Thomas, P. K., King, R. H. M., Kocen, R. S., and Brett, E. M.: Comparative ultrastructural observations on peripheral nerve abnormalities in late infantile, juvenile and late onset forms of metachromatic leukodystrophy. Acta Neuropathol. [Berl.] 39:237, 1977, with permission.)

Myelinated fiber density was less markedly diminished and frequently was not significantly different from controls. Thomas and co-workers290 found that small onion bulbs and regenerative clusters were more numerous in the later onset cases, this again paralleling the chronicity of their clinical course. Percy and colleagues237 showed that the nerves in the adult-onset cases may ultimately reach a more chronically demyelinated and fibrotic state than do those from late infantile MLD. Both Martin and co-workers191 and Thomas and associates290 failed to detect any differences in the types of inclusions between the different forms, although Joosten and associates147 did not observe the stacked discs (prismatic inclusions) in a patient with adult-onset disease. They and Luitjen and co-workers184 considered that tuffstone bodies were more common in the late infantile cases and zebra bodies in the juvenile and adult-onset forms.

The presence of inclusions in actively myelinating nerve without evidence of demyelination has been described in fetal cases diagnosed prenatally. Fiber type disproportion may be an associated feature in skeletal muscle.167 Neuronal perikarya of peripheral ganglia may contain lamellar inclusions of great diversity and quite unlike classic inclusions,97 occasionally even mixed with lipopigments.98 Transgenic ASA-deficient mice store sulfatides in their Schwann cells but do not show peripheral demyelination.130 PATHOGENESIS. The accumulation of sulfatides in cerebral white matter in MLD was described independently by Austin,18 Jatzkewitz,143 and Hagberg and Svennerholm121 and was later shown for myelin.224,279 The concentrations of other myelin lipids, such as cholesterol and sphingomyelin, are reduced, probably as a result of myelin loss. The

1852

Diseases of the Peripheral Nervous System

FIGURE 79–8 Electron micrograph of longitudinal section through unmyelinated nerve fiber from a patient with late infantile metachromatic leukodystrophy. The Schwann cell cytoplasm contains multiple dense inclusion bodies (db). ax ⫽ axon; Sn ⫽ Schwann cell nucleus. Bar: 0.1 ␮m.

reduction in cerebroside is particularly great. The sulfatide that accumulates is normal in composition. It was established by Austin and co-workers20,22 that the basic defect is deficient ASA activity, so that the tissues are unable to degrade p-nitrocatechol sulfate. A little later, Mehl and Jatzkewitz201 found that cerebroside sulfatase closely resembles ASA and that the tissues in MLD fail to degrade galactosylsulfatide. It was then established that the degradation of both p-nitrocatechol sulfate and galactosylsulfatide is catalyzed by the same enzyme (ASA, or cerebroside sulfatase), which is lacking in MLD.144,145,274 The metachromatic granules that accumulate in the brain have been shown to consist of galactose sulfatide (cerebroside sulfate).279 The mechanism of demyelination in MLD is still uncertain. One possibility relates to the finding of O’Brien and Sampson that the fatty acid composition of cerebrosides and sphingomyelin is abnormal, with a reduction in the proportion of long-chain fatty acids225,226 and an abnormal cerebroside-sulfatide ratio, with an excess of the latter.224 The increase in sulfatide may lead to myelin damage by attracting cations. In vitro studies suggest that this may have deleterious effects on myelin integrity.177 A second possibility is that the accumulation of sulfatide interferes with the function of oligodendrocytes and Schwann cells. The sulfatide inclusions are not secondary to the demyelination, because they are present in fetal cases before demyelination is evident. They also occur in Schwann cells associated with unmyelinated axons. The third possibility is that the demyelination is related to the accumulation of a substance other than sulfatide. It is known that sulfogalactosylsphingosine sulfatase activity is deficient in MLD. This could lead to the intracellular

accumulation of sulfogalactosylsphingosine, which is known to be cytotoxic.77 A further possibility relates to the temperature dependence of lipid bilayer stability. Lipid extracts from normal tissue spontaneously assemble to form bilayers but only at a critical temperature. The critical temperature for normal human myelin lipids is 37° C. Critical state theory dictates that the membrane bilayer will spontaneously degenerate if the lipid composition is inappropriate to the cells’ ambient temperature. The critical temperature for the myelin lipids from patients with MLD was found to be 30° C,92 suggesting a mechanism for myelin instability. GENETIC ASPECTS. All three forms of MLD are inherited as autosomal recessive traits. Most information is available for the late infantile form. In this type, segregation ratios of 0.25 and 0.22 have been found,117,266 and the incidence of parental consanguinity is increased.32,117 Leukocyte ASA activity in obligate carriers is reduced by approximately 50%. This is also true for the juvenile and adult-onset forms.88,241 Each clinical form of MLD is caused by either mutations in the ASA gene or by deficiency of the activator protein saposin B. The ASA gene, located at 22q13, consists of 8 exons and encodes a 507–amino acid precursor protein located in the rough endoplasmic reticulum, from where it undergoes modifications during maturational passage to the lysosome. Well over 16 mutations in the ASA gene have been recorded,306 some of them having ethnic predominance. Common mutations do not seem to exist, but only ones with “high frequency,” and often private mutations have been noted. The missense mutation Thr286Pro has repeatedly been encountered in isolated MLD neuropathy.60,81 Null

Lysosomal and Peroxisomal Disorders

1853

FIGURE 79–9 A, Electron micrograph of an inclusion body within an unmyelinated fiber Schwann cell from a patient with late infantile metachromatic leukodystrophy. This contains lamellar material with a 5.8-nm periodicity, which is irregularly oriented within the central part of the inclusion body. At the periphery, the material is organized into groups of lamellae oriented perpendicular to the limiting membrane. Bar: 0.1 ␮m. B, Electron micrograph of an inclusion body with endoneurial macrophage from the same patient, which contains stacks of lamellar discs with a periodicity of 5.8 nm. These are seen at higher magnification in the inset. Bar: 0.5 ␮m. (From Thomas, P. K., King, R. H. M., Kocen, R. S., and Brett, E. M.: Comparative ultrastructural observations on peripheral nerve abnormalities in late infantile, juvenile and late onset forms of metachromatic leukodystrophy. Acta Neuropathol. [Berl.] 39:237, 1977, with permission.)

mutations are apparently responsible for the severe late infantile form. ASA pseudodeficiency is defined by very low ASA enzyme activity but without clinical symptoms or morphologic manifestation of MLD. It is derived from polymorphisms in the ASA gene and needs to be remembered when interpreting biochemical tests and in genetic counseling. Patients with true MLD, however, may also have the

ASA pseudodeficiency.248 A knockout mouse model showing a mild phenotype, no demyelination, and typical MLD lysosomal residual bodies has recently been created.130 Prenatal diagnosis by amniocentesis has been performed in the late infantile and juvenile forms by ASA assay on cultured amniotic fluid cells or in amniotic fluid.178,299 Because of the delay in the onset of the disease, prenatal diagnosis

1854

Diseases of the Peripheral Nervous System

is unlikely to be relevant for adult-onset cases. There is no overlap between homozygotes and normal persons for late infantile cases in ASA levels, but at times obligate carriers are found to have enzyme values within the normal range. In approximately 20% of all families, one or both parents have levels of enzyme activity similar to those of their affected children. In prenatal diagnosis, differentiation would therefore not be possible between the affected fetus and a heterozygote.288 Based on the mutational status of the index patient, prenatal molecular testing is now a reliable procedure. TREATMENT. In the absence of specific therapy, treatment is confined to supportive measures. Replacement therapy with exogenous enzyme has earlier been unsuccessful,14 but bone marrow transplantation has succeeded, if not performed too late in the course of MLD, in ameliorating symptoms, including nerve conduction,270 slowing progression, and prolonging life.170,171,238 Attempts to diminish the rate of sulfatide accumulation with a low-sulfur diet212 or a diet free of vitamin A,202 as well as immunomodulation treatment,220 have been made. Vitamin A acts as a cofactor in the synthesis of sulfatide. Multiple Sulfatase Deficiency. Austin, 20 in 1963, described two siblings with a disorder that differed from the classic form of MLD in its clinical features and biochemistry. He referred to this as “metachromatic leukodystrophy variant,” but because it was later shown to be associated with a deficiency of arylsulfatases A, B, and C, the term multiple sulfatase deficiency (MSD) or mucosulfatidosis has been adopted.15 Additional cases were reported by Rampini and co-workers.249 The patients of Mossakowski and colleagues215 and Thieffry and associates285 probably had the same disorder. In contrast to classic MLD, MSD is characterized by the additional storage of ganglioside in cortical neurons and an elevation of mucopolysaccharide and cholesterol sulfate levels in various tissues. For this reason, Rampini and co-workers249 suggested the term mucosulfatidosis for this disorder, which was also employed by Couchot and associates.59 CLINICAL FEATURES. The inheritance is autosomal recessive and the age at onset and duration are similar to the pattern seen in late infantile MLD, but with a slightly earlier onset. Retardation of motor and intellectual development is followed by a rapidly progressive deterioration to complete dementia, accompanied by blindness, deafness, and decerebrate rigidity. The tendon reflexes are diminished. Many patients display ichthyosis, with thickened, dry skin (Fig. 79–10), and the facial features may be slightly coarse. In most cases there is hepatic and splenic enlargement. Skeletal deformities may be evident, with reduced joint mobility, an abnormal flaring of the ribs, and a lumbar kyphosis.

FIGURE 79–10 Patient with multiple sulfatase deficiency showing diplegia and ichthyosis of the upper part of the body. (Courtesy of Dr. S. Rampini.)

LABORATORY INVESTIGATIONS. There is an increased excretion of sulfatides and mucopolysaccharides in the urine, with excess of dermatan sulfate and heparan sulfate. Alder-Reilly granules are present in the bone marrow and peripheral blood leukocytes. Deficient activity of arylsulfatases A, B, and C and of steroid sulfatase is demonstrable in the urine, in leukocytes, and in cultured skin fibroblasts.

Lysosomal and Peroxisomal Disorders

Skeletal changes somewhat similar to those encountered in the MPSs may be found radiologically, including broad phalanges and a J-shaped sella turcica.15 Motor NCV is reduced, and nerve biopsy specimens show demyelination and metachromatic inclusions. PATHOLOGY. The cerebral changes are characterized by extensive diffuse demyelination, together with the presence of metachromatic material similar to that encountered in classic MLD and ballooned neurons containing an orthochromatic but periodic acid–Schiff (PAS)-positive substance.215,217 Peripheral nerve pathology was studied by Bischoff and Ulrich40 (Fig. 79–11). The findings resembled those of the classic form of MLD. Ultrastructural studies by Bischoff on sural nerve biopsy specimens from the three cases whose clinical features were reported by Rampini and associates249 did not reveal any changes that were not present in MLD.39 PATHOGENESIS. An interesting feature of MSD is that the combined enzyme defects depend on a single gene defect.58,69 This suggested that the sulfatases involved may be structurally related.217,275 The gene product, the formyl glycin–generating enzyme (FGE),58,69 normally confers structural stability on these sulfatases by transforming

FIGURE 79–11 Sural nerve biopsy from a patient with multiple sulfatase deficiency. Some dark orthochromatic globules are present in the cytoplasm of Schwann cells (arrow). (Courtesy of Dr. A. Bischoff.)

1855

an active catalytic site, cysteine 12,133 common to known sulfatases. Hence, FGE mutations, nine of which have been described in a single paper,69 affect multiple sulfatases, being a posttranslational defect that causes MSD, whereas mutations in genes encoding individual lysosomal sulfatases produce different lysosomal diseases (i.e., MLD and several MPS).133 Sulfatide Activator Protein-B Deficiency. A further MLD variant (the AB form) related to a deficiency of an activator protein has been described.267,273 The sulfatide activator protein (SAP) is a nonenzymatic glycoprotein, assisting ASA to degrade sulfatides. It is also termed saposin B, an 80–amino acid protein, and is one of four homologue proteins derived from the 524–amino acid precursor protein prosaposin, the gene for which is located at 10q21-23. Other such homologues are SAP-A, -C, and -D, with SAP-D deficiency causing juvenile Gaucher’s disease. Several mutations in the single gene for prosaposin and sites coding for the individual forms of homologue SAPs have been described, and the ones mutationally affecting the SAP-B region result in mutant SAP-B sphingolipid activator protein, responsible for an MLD-like clinical and morphologic picture.258 The patient reported by Hahn and associates124,125 was a 21-year-old woman who had a history of delayed intellectual and motor performance in early childhood. Progressive deterioration occurred during adolescence, with behavior disorder and dysarthria leading to severe dementia, a supranuclear bulbar palsy, muscle hypertonus in the arms, and signs of a moderately severe peripheral neuropathy in the lower limbs with hypotonic weakness, tendon areflexia, and sensory disturbance. Motor NCV was severely reduced (10 m/s in the peroneal nerve). Sensory nerve action potentials were not recordable, and electromyography revealed evidence of denervation. The cerebrospinal fluid protein concentration was elevated. A sural nerve biopsy showed changes of a chronic demyelinating neuropathy and the full range of inclusions that typify the classic form of MLD. In vitro activities of arylsulfatases A and B and cerebroside sulfatase were normal. Skin fibroblasts, cultured in a medium containing labeled sulfatide, showed sulfatide accumulation. A deficiency of the activator protein described by Fischer and Jatzkewitz83 was confirmed by Zhang et al.325 Galactosylceramide Lipidosis (Globoid Cell Leukodystrophy; Krabbe’s Disease) Beneke, in 1908,35 and Krabbe, in 1916,165 were responsible for describing the pathologic features of galactosylceramide lipidosis, which consist of extensive demyelination of the cerebral white matter, severe astrocytic gliosis, and variable axonal degeneration. The characteristic feature is the presence of multinucleated (i.e., giant) globoid cells of mesenchymal (i.e., microglia and macrophages) origin,

1856

Diseases of the Peripheral Nervous System

10 to 40 ␮m in diameter, with 5 to 15 marginal nuclei. These cells frequently accumulate around small vessels. Histochemically, they contain PAS-positive material. The globoid cells contain galactosylceramide (galactoceramide) and have been experimentally induced by psychosine, a putatively toxic lysosphingolipid, owing to failure in cytokinesis rather than cell-to-cell fusion.164 A second type of cell with similar histochemical characteristics, but of epithelial shape and with a single nucleus, is also seen. Globoid cells engorged with PAS-positive material have occasionally been encountered in the lungs, spleen, and lymph nodes. Demyelination has been shown to occur in the peripheral nerves.19,72,174,271 Electron microscopy of brain biopsy material reveals the presence of deposits with a crystalline or tubular appearance within the globoid cells.7,324 Similar prismatic structures are found in peripheral nerve.41,186 The diagnosis may be established by galactosylceramide-␤-galactosidase assay41 of the enzyme in leukocytes, serum, or cultured fibroblasts. Clinical Features. After normal neonatal development, the first signs of the disorder usually appear at the age of 3 to 6 months in the form of retarded intellectual development accompanied by irritability and episodes of crying and hypersensitivity to touch, sound, and visual stimuli. Unexplained vomiting may occur, as may convulsive seizures. This is stage I as defined by Hagberg (Fig. 79–12).118 In stage II, rapidly advancing motor and intellectual regression becomes evident, with hypertonicity of the limbs and axial musculature and exaggerated tendon reflexes. Optic atrophy may develop, and seizure activity continues. Stage III is reached after a course sometimes as short as a few months. The child is decerebrate, blind, and deaf. Death usually occurs within 2 years of the onset. Peripheral nerve involvement is not as obtrusive in Krabbe’s leukodystrophy as in MLD, but Dunn and associates72 found that the tendon reflexes usually became

depressed about 6 months after onset, although they may remain hyperactive even in the late stages of the disease. Cranial nerve and spinal nerve root enhancement has been documented by magnetic resonance imaging.36,303 Motor NCV has been found to be reduced in about 50% of cases,132,186 and electrophysiologic evidence of peripheral nerve involvement may be detectable as early as 7 weeks postnatally.179 Occasionally, peripheral neuropathy may be the initial manifestation in the infantile form.163 Cases with later onset also occur, with symptoms beginning at between 2 and 6 years of age and disease duration being 1 to even 15 years10 or longer.142 Such cases are characterized by progressive intellectual deterioration, optic atrophy, and pyramidal signs.63 The cerebral pathology includes the presence of globoid cells, and a deficiency of galactosylceramide-␤-galactosidase has been demonstrated. Peripheral neuropathy is often not evident,63 but may occasionally be present as the first clinical sign.189 Progressively reduced motor NCV10 and demyelination of peripheral nerve at autopsy have been described.180 A sural nerve biopsy is stated to have been normal. Recently, adult-onset Krabbe’s disease with normal NCV but abnormal evoked potentials and CNS symptoms has been reported in two brothers.296 Affected adults in another family with late-onset Krabbe’s disease showed only clinical paraparesis without CNS white matter abnormalities; PNS data were not given.27 True adult onset, in a 23-year-old woman, has been recognized by biochemical and molecular studies.67 An unusual case was reported by Thomas and coworkers.289 The patient presented in the fourth decade of life with the features of a slowly progressive spinocerebellar degeneration with an accompanying demyelinating sensorimotor neuropathy. He later developed a hemiparesis, and a cranial computed tomography scan revealed periventricular hypodense lesions. Nerve biopsy showed typical features of Krabbe’s disease (Fig. 79–13). The diagnosis was confirmed by enzyme studies. The patient’s parents had enzyme levels in the heterozygote range, as did his sister.

FIGURE 79–12 Diagram of the clinical findings in Krabbe’s disease. (Data from Hagberg, B.: The clinical diagnosis of Krabbe’s infantile leukodystrophy. Acta Paediatr. Scand. 52:213, 1963.)

Lysosomal and Peroxisomal Disorders

FIGURE 79–13 Transverse section from biopsied sural nerve from a patient with protracted Krabbe’s leukodystrophy. Myelin sheath thickness is uniformly reduced. Minor hypertrophic changes are present (short arrows), and inclusion material is visible in some Schwann cells (longer arrow). Semithin section stained with thionin and acridine orange (⫻600).

Kolodny and co-workers162 have recently reported other cases of late-onset Krabbe’s disease in which the clinical picture included a sensorimotor demyelinating neuropathy suggesting a clinical continuum.236 Bone marrow transplantation may lead to clinical improvement.170,171 Pathology. Sural nerve biopsy can provide confirmation of the diagnosis of Krabbe’s leukodystrophy41,132,186 but is

FIGURE 79–14 Krabbe’s disease. Electron micrograph of transverse section from sural nerve showing inclusion material (stars) within Schwann cell cytoplasm. Lead citrate and uranyl acetate. Bar: 0.1 ␮m. (Courtesy of Dr. R. H. M. King.)

1857

now usually no longer necessary. The pathologic changes seen by light microscopy are not very obtrusive in most cases, but a reduced myelinated nerve fiber density may be apparent (see Fig. 79–13). Segmental demyelination is common.174 Inclusion material is sometimes visible in Schwann cells and macrophages in semithin sections, as is dense material in the neighborhood of endoneurial blood vessels (Fig. 79–13). On electron microscopy it is evident that the peripheral nerves are regularly involved. About half of the Schwann cells of myelinated and unmyelinated nerve fibers contain profiles that are composed of nonoriented, straight or curved, prismatic or tubular inclusions in a clear matrix (Figs. 79–14 and 79–15),41 similar to those observed in globoid cells in the brain. The Schwann cell endoplasmic reticulum tends to be dilated; the distended cavities contain an electron-lucent substance, interspersed with finely granular material.41 Typical globoid cells are not observed in peripheral nerve, but endoneurial macrophages containing prismatic inclusions are present (Fig. 79–16), especially in the vicinity of blood vessels. Such inclusion material may also be observed lying free in the endoneurium, presumably related to the disruption of macrophages (see Fig. 79–16). In the skin, a tissue less invasively biopsied than sural nerve, dermal nerve fascicles may be affected, and secretory cells of eccrine sweat glands as well.76,100 Deciduous teeth and their nerves may also be diagnostic.42 In the late-onset form inclusions may appear ultrastructurally somewhat different, 99 and hypomyelination rather than segmental demyelination

1858

Diseases of the Peripheral Nervous System

FIGURE 79–15 Electron micrograph of endoneurial macrophages (mp) from a case of Krabbe leukodystrophy containing inclusion material (im). This is also visible extracellularly in the lower right of the figure. Bar: 0.1 ␮m. (Courtesy of Dr. R. H. M. King.)

and remyelination may prevail.196 The morphologic pattern of fiber type disproportion may be encountered in skeletal muscle.188 Globoid cells of typical form have been found in the brainstem and spinal cord in cases diagnosed prenatally by enzyme assay on cultured amniotic cells.135,193 Specific inclusions have also been detected in fetal peripheral nerve.193 Pathogenesis. In the brain there are pronounced decreases in all myelin lipids,3,19,240 but especially glycolipids, consistent with myelin loss, and an elevation of the ratio of cerebrosides to sulfatides to a value almost double that found in normal brain.240 The cause of the disorder is deficient activity of the enzyme galactosylceramide-␤galactosidase, or galactocerebrosidase,24 which normally converts galactocerebroside to ceramide and galactose. It has been suggested that oligodendrocytes and Schwann cells are

damaged by the accumulation of galactocerebroside, or possibly psychosine, a galactosphingosine also degraded by galactocerebrosidase. Globoid cells, with typical inclusion material, have been produced by the intracerebral inoculation of bovine cerebrosides into rat brain.23,228 Globoid cell leukodystrophy with the same enzyme defect has also been observed in some strains of dogs as an autosomal recessive disorder.21,85,199 As in the human disease, the peripheral nerves are affected.173,324 The mutant twitcher mouse has a recessively inherited demyelinating disorder of the CNS and PNS with pathologic features similar to those of Krabbe’s leukodystrophy in humans.71 The enzyme defect is the same.137,160 Genetic Aspects. The inheritance is autosomal recessive. It affects both sexes equally.321 The prevalence has been reported to be as high as 1.9 ⫻ 10⫺5 in Sweden,120 but the disorder may be more common in Scandinavia than in other

Lysosomal and Peroxisomal Disorders

1859

FIGURE 79–16 Krabbe’s leukodystrophy. Electron micrograph of inclusion material (im) within an endoneurial macrophage seen both in cross section and en face. Bar: 0.1 ␮m.

areas of the world. The gene is located at 14q24.3-32.1 and consists of 17 exons or 56 kb. More than 60 mutations spread across the gene have been identified. The 502T/del mutation has been found among some 50% of European patients and some 75% of Swedish patients, while the D270N mutation seems to cause a late onset and mild phenotype.316,317 The disease can be diagnosed prenatally by enzyme assay in cultured amniotic cells.135 Assay of galactosylceramide␤-galactosidase is also helpful in the detection of heterozygotes, but although the average activity in carriers is 50% of normal, there is considerable overlap with the normal population.280

Other Lysosomal Sphingolipidoses Ceramide Deficiency (Farber’s Disease) Farber’s lipogranulomatosis is a rare autosomal recessive disorder with onset most often in infancy, but other clinical forms have been delineated.209 The manifestations include joint enlargement and contractures, subcutaneous and periarticular nodules, hoarseness of the voice, and pulmonary consolidation.208 Affected tissues are infiltrated with macrophages and foam cells and show an increased ceramide content. This is related to deficiency of an acid ceramidase that degrades ceramide to sphingosine and fatty acid276; the gene for this enzyme is situated at 8p21.3/22, with more than 10 mutations spread across the 14 exons.214 Mental retardation may occur, but otherwise neurologic features indicate peripheral nerve involvement.208 Diminution or loss of tendon reflexes occurs, and the

affected children may show hypotonia and amyotrophy in the limbs with electromyographic evidence of denervation. Cutaneous hyperesthesia also occurs. At autopsy, PAS-positive storage material, probably glycolipid, is present in neurons, including anterior horn and autonomic and dorsal root ganglion cells. This is enclosed in lysosomal vacuoles and on electron microscopy is seen to be pleomorphic, the most characteristic inclusion being a comma-shaped curved tubular structure.208,256 Axonal loss in the peripheral nerves has been described with vacuoles in Schwann cells and histiocytes.305 Needle-shaped lysosomes have also been seen in dermal Schwann cells.51 Gangliosidoses Gangliosidoses are lysosomal disorders marked by intralysosomal accumulation of gangliosides, a major component of sphingolipids. They are divided into two categories, GM2 gangliosidoses resulting from hexosaminidase deficiencies, which are largely neuronal and CNS disorders, and GM1 gangliosidosis, affecting both CNS and extracerebral tissues, both mesenchymal and epithelial ones, as a result of deficiency of lysosomal ␤-galactosidase. GM2 Gangliosidoses. This group of disorders derives from deficiencies of hexosaminidases A and B, hexosaminidase A being composed of ␣ and ␤ subunits, and hexosaminidase B only containing ␤ subunit. The gene for the ␣ subunit is located on chromosome 15, with now more than 90 mutations, and that for the ␤ subunit on chromosome 5, with more than 25 mutations.111 A separate gene locus with four mutations exists on 5q32-33 coding for a GM2 activator

1860

Diseases of the Peripheral Nervous System

protein.258 Classic infantile Tay-Sachs disease (type I or B variant GM2 gangliosidosis) is an autosomal recessive disorder in which progressive motor and mental retardation develops at 3 to 6 months of age, later followed by blindness, deafness, seizures, spasticity, and decerebrate rigidity. Cherry-red spots are usually evident at the maculae. Pathologically there is lipid accumulation in neurons of the cerebral cortex, autonomic ganglia, and rectal mucosa. The ballooned neurons contain cytoplasmic membranous bodies with concentric lamellae composed of GM2 ganglioside. This disorder is related to an ␣-locus deficiency disorder with a lack of hexosaminidase A. Nerve biopsy specimens show changes confined to unmyelinated axons, some of which are enlarged because of multilamellar structures with a periodicity of 6.6 nm.39 In the late stages, some myelin loss may be evident.168 A particular biochemical form is the B1 variant, common in Portugal, which may show normal hexosaminidase A activity with a regular substrate, but specific enzyme deficiency with the sulfated substrate. Involvement of the peripheral nervous system in this B1 variant has recently been demonstrated176 affecting Schwann cells106 and terminal axons.50,93 In recent years there have been a number of reports of hexosaminidase A deficiency, often resulting from “juvenile” mutations,111 giving rise to a variety of different phenotypes with a more protracted course and usually multisystem neurologic involvement, including progressive ataxia, spasticity, neurogenic amyotrophy (as noted by Hund et al.,134 whose patients were compound heterozygotes for “adult” mutations), dystonia, dementia, seizures, and psychiatric disorders. Variants that give rise to a spinal muscular atrophy are further considered in Chapter 68. Barnes and co-workers31 have reported a patient with ataxia, internuclear ophthalmoplegia, amyotrophy, and a sensory neuropathy. Sural nerve biopsy showed an axonopathy with loss of both myelinated and unmyelinated axons. No axonal or Schwann cell inclusions were seen. Involvement of the peripheral autonomic nervous system may occur in the chronic juvenile form.257 Type II (O variant) GM2 gangliosidosis, or Sandhoff’s disease, is related to ␤-locus deficiency. Both hexosaminidases A and B are absent. The disorder usually resembles type I GM2 gangliosidosis, but more protracted juvenile cases occur. Some of these include spinal muscular atrophy as a salient feature or part of the clinical syndrome (see Chapter 68). Thomas and colleagues292 described a case that bore clinical resemblances to X-linked bulbospinal neuronopathy. Although there was no sensory loss, sensory nerve action potentials were absent. Rectal biopsy revealed the presence of membranous cytoplasmic bodies in axons. A severe sensory loss resulting from peripheral-central dyingback axonopathy has been found in three adult siblings,261 and evidence of autonomic dysfunctions, such as orthostatic hypotension, loss of sweating, and diarrhea, presented in another patient.207

Lack of an activator protein causes the AB variant, similar in severity to classic infantile or B variant. In GM2 gangliosidosis, membranous cytoplasmic bodies occur not only in the perikarya of peripheral nerve cells but, occasionally, also in terminal axons.50 In the B variant lysosomal storage appears largely to be confined to the neuronal compartment, whereas in the O variant, or Sandhoff’s disease, Schwann cells and mural cells of endo- and epineurial vessels may additionally show accumulation by lysosomal lamellar inclusions.34 GM1 Gangliosidosis. This lysosomal storage disorder caused by deficiency of ␤-galactosidase is allelic to MPS IVB, or Morquio syndrome type B, and thus affects both neuroectodermal, epithelial, and mesenchymal cells in respective organs. Schwann cell inclusions have been described in infantile (type I) GM1 gangliosidosis.39 They are also present in histiocytes and endothelial cells and consist of membranebound structures with an electron-lucent content. Such clear vacuoles may be seen in epithelial cells in skin biopsy specimens. This condition is a rare autosomal recessive disorder in which abnormalities are present at birth or develop shortly after. Progression to unresponsive decerebrate rigidity occurs, with death usually taking place by the age of 2 years. Associated features include cherry-red spots at the maculae, hepatosplenomegaly, and multiple dysostoses. In type II, neuronal perikarya and an occasional axon may be replete with membranous cytoplasmic and other lamellar lysosomal residual bodies,93,233 while Schwann cells may contain vacuololamellar lysosomes.161 The gene for ␤-galactosidase is located at 3p21.33, giving rise to several common mutations that often prevail in certain ethnic groups: R482H in Italian patients with infantile GM1 gangliosidosis, R208C in American patients with infantile GM1 gangliosidosis, R201C in Japanese patients with juvenile GM1 gangliosidosis, I51T in Japanese patients with adult GM1 gangliosidosis, and W273L in white patients with Morquio syndrome type B.281 Glucosylceramide Lipidosis (Gaucher’s Disease) Gaucher’s disease is divisible into a chronic non-neuropathic adult variety (type 1) and infantile and juvenile neuropathic variants (types 2 and 3, respectively). All three forms are related to deficient activity of glucosylceramide-␤glucosidase, the gene for which is at 1q21, with 11 exons having undergone numerous mutations, among them many frequent ones.38 The enzyme converts glucocerebroside to ceramide and glucose, and the three forms are all characterized by the presence in involved tissues of enlarged lipidcontaining macrophages derived from monocytes (Gaucher cells). The stored material consists of glucosylceramide and glucosylsphingosine. Peripheral nerve involvement has been described largely in the infantile variety, rarely and

Lysosomal and Peroxisomal Disorders

perhaps coincidentally in type 1.198 Intellectual regression and spasticity usually become manifest at about 6 months of age and are associated with hepatosplenomegaly. Ultrastructural examination of peripheral nerve biopsy specimens has shown irregular amorphous inclusion bodies in histiocytes and Schwann cells.40 The 13- to 15-nm tubular inclusions found in Gaucher cells in other tissues were not observed. No axonal loss or demyelination was detected. Enzyme replacement therapy has recently been successful in the non-neuronopathic form.47 Niemann-Pick Disease Niemann-Pick disease consists of a group of four disorders. In types A and B, now also combined as group I, there is an abnormal accumulation of sphingomyelin in a variety of tissues related to a deficiency of sphingomyelinase.48 Both are of autosomal recessive inheritance. Type B does not affect

FIGURE 79–17 Type A Niemann-Pick disease. Electron micrograph of transverse section through sural nerve biopsy specimen. Schwann cells associated with both myelinated and unmyelinated fibers are loaded with osmiophilic inclusions (arrows). Uranyl acetate and lead citrate. Bar: 2 ␮m. (From Landrieu, P., and Said, G.: Peripheral neuropathy in type A Niemann-Pick disease. Acta Neuropathol. [Berl.] 63:66, 1984, with permission.)

1861

the nervous system. In type A (Crocker type), referred to as the acute or infantile form, hepatosplenomegaly develops during the first year of life and is associated with intellectual deterioration, retinal cherry-red spots, and sometimes seizures. Peripheral neuropathy has been described in this form, but only rarely, although the occurrence of hypotonic limb weakness in previous descriptions suggests that it may have been overlooked in the past. In the cases described by Gumbinas and co-workers116 and Landrieu and Said,175 these features were associated with tendon areflexia and markedly reduced motor NCV.116 Morphologic studies on peripheral nerve116,175 have shown segmental demyelination and remyelination; lipid inclusions in Schwann cells and axons (Fig. 79–17), sometimes showing a 6-nm periodicity; and endothelial foam cells, also seen in a knockout mouse.172 Cultured Schwann cells display markedly deficient acid sphingomyelinase

1862

Diseases of the Peripheral Nervous System

activity.175 Types A and B are allelic disorders caused by mutations in the sphingomyelinase gene on chromosome 11p15.1-4,264 with a frequent 3-bp deletion of exon 6 in Niemann-Pick type B patients allowing for higher residual sphingomyelinase activity than in type A patients, which, however, may be difficult to correctly determine.110 In the past types C and D Niemann-Pick disease have been included as sphingomyelin lipidoses, but this is no longer justified. Sphingomyelinase activity is normal or only mildly reduced, and the activator protein is normal.86 There is an abnormality of cholesterol compartmentation and esterification, and the storage material consists of cholesterol and neutral lipids.197 Clinically, both types C and D Niemann-Pick disease are similar, with a wide clinical spectrum in type C (NPC) from early childhood to adulthood marked by hepatosplenomegaly, followed by psychomotor deterioration, ataxia, and vertical gaze palsy. Death occurs in adolescence. Both are of autosomal recessive inheritance. Neuropathy has been reported in the type D (Nova Scotia) variant,39 which is now considered allelic to NPC, and in NPC.123 Recently, it has been shown that intra-axonal cholesterol transport is impaired in deficient NPC neurons, with reduced amounts of cholesterol in distal axons but storage in neuronal perikarya.153,154 Some 95% of NPC patients link to the NPC1 locus on 18q11, which encodes a membrane glycoprotein in late endosome.302 The number of known mutations in the NPC1 gene is well above 100 and rising rapidly.232 Most of the remaining NPC patients are now found to have mutations in the NPC2 gene, located on 14q24, encoding for a cholesterol-binding small lysosomal protein, but extensive molecular studies in a large number

of NPC patients also revealed a few not linking to the NPC1 or NPC2 loci.232

MISCELLANEOUS DISORDERS The Neuronal Ceroid-Lipofuscinoses The NCLs are a group of progressive neurodegenerative disorders of childhood (rarely adulthood) of almost exclusively an autosomal recessive mode of inheritance. They are marked by continuous loss of nerve cells, resulting in premature death of the patients, and the intracellular accumulation of autofluorescent lysosomal lipopigments in neural and non-neural cells involving clinically the brain, but morphologically also many extracerebral cell types and tissues, including the PNS.108,146,168 Lipopigments accrue in nerve cell bodies (Fig. 79–18), many of which may be lost by the time of death from peripheral ganglia,168 which has rendered rectal biopsy a historically crucial diagnostic procedure, and in Schwann cells (Fig. 79–19) of unmyelinated axons more than of myelinated axons and mural cells of nerve vessels. In juvenile NCL, lysosomal vacuoles may also form in lymphocytes and other cells (Fig. 79–20). Clinical symptoms such as muscle weakness at the motor site, sensory deficits, or autonomic failure do not feature prominently. Because demyelination is not a consequence of lipopigment formation in myelinating Schwann cells, electrophysiologic studies of the PNS may give normal results. Cell types other than neurons and Schwann cells in peripheral ganglia or nerve cell– containing visceral organs

FIGURE 79–18 Juvenile neuronal ceroid-lipofuscinosis. Membranebound lysosomal residual bodies, partly vacuolar, partly filled with a mixture of curvilinear and fingerprint profiles in a sympathetic ganglion.

1863

Lysosomal and Peroxisomal Disorders

FIGURE 79–19 Late-infantile neuronal ceroid-lipofuscinosis. A curvilinear body (arrow) present in the Schwann cell of a myelinated axon in a postmortem sural nerve.

such as the intestine may also show NCL-specific lipopigment formation. Ubiquitous appearance of NCL-specific lipopigments that provide a granular pattern in the clinical infantile or the genetic CLN1 form, curvilinear profiles in the clinical late-infantile or genetic CLN2 form, and fingerprint profiles in the clinical juvenile or genetic CLN3 form allows less invasive biopsies than peripheral nerve biopsies (e.g., of blood, lymphocytes, skin, or conjunctiva). Currently, the NCLs are divided into eight different genetic forms according to six different gene loci (Table 79–1), with two lysosomal proteases being deficient in CLN1 and CLN2 (i.e., PPT1 and TPP1, respectively), and four transmembrane structural proteins underlying mutation-related defects in CLN3, CLN5, CLN6, and CLN8. On the Neuronal Ceroid Lipofuscinoses (NCL) Mutations website (www.ucl.ac.uk/ncl) there is information on all known mutations. Clinical symptoms comprise progressive cognitive impairment, seizures (not infrequently of the myoclonic type), visual impairment resulting in blindness and motor dysfunction, and often ataxia, though of different onset and in different sequential order in separate clinical and genetic forms of NCL. Treatment still is symptomatic, coping with epilepsy or behavioral abnormalities.104

FIGURE 79–20 Juvenile neuronal ceroid-lipofuscinosis. Numerous vacuoles in the Schwann cell of a myelinated axon from a dermal nerve fascicle.

Chédiak-Higashi Syndrome Chédiak-Higashi syndrome is a rare autosomal recessive disorder characterized by partial oculocutaneous albinism with giant melanosomes. Giant peroxidase-positive lysosomes are present in leukocytes. The disorder begins in childhood with defective hair pigmentation, anemia, leukopenia, thrombocytopenia, and a liability to lymphoreticular malignancy.43 There may be associated neurologic features including mental retardation and a clinical syndrome resembling a spinocerebellar degeneration. Peripheral neuropathy has been described in a number of patients. Early studies169,204 documented degeneration of nerve fibers and extensive infiltration of peripheral nerves with histiocytes. The first detailed study was by Lockman and co-workers,181 who reported the presence of giant lysosomes in Schwann cells. These were also recorded by Blume and Wolff43 and Pezeshkpour and associates.239 Sung and co-workers,277 in an autopsy study, again found histiocytic infiltration of nerves. A recent nerve biopsy study by Misra and colleagues205 reported loss of myelinated fibers, particularly those of large size, and of unmyelinated axons. Inflammatory cell infiltration and giant lysosomes were not observed. It was concluded that the appearances were those of an axonopathy. Its mechanism is uncertain. The gene for the Chédiak-Higashi syndrome (CHS1) on chromosome

1864

Diseases of the Peripheral Nervous System

Table 79–1. Current Nosographic Status of Neuronal Ceroid-Lipofuscinoses Genetic types

Chromosome

Gene Product

Clinical Types

CLN1 CLN2 CLN3

1p32 11p15 16p12

INCL, LINCL, JNCL, ANCL LINCL, JNCL JNCL

CLN4 CLN5 CLN6

Not known 13q31-32 15q21-23

Lysosomal palmitoyl protein thioesterase I Lysosomal tripeptidyl peptidase I Lysosomal transmembrane CLN3 protein (battenin) Not known Transmembrane CLN5 protein Transmembrane CLN6 protein

CLN7 CLN8

Not known 8

Not known Transmembrane CLN8 protein

ANCL LINCL Finnish variant EJNCL, LINCL Indian/Czech Gypsy variant LINCL Turkish variant “Northern” epilepsy LINCL Turkish variant

ANCL ⫽ adult neuronal ceroid-lipofuscinosis; EJNCL ⫽ early-juvenile neuronal ceroid-lipofuscinosis; INCL ⫽ infantile neuronal ceroid-lipofuscinosis; JNCL ⫽ juvenile neuronal ceroid-lipofuscinosis; LINCL ⫽ late-infantile neuronal ceroid-lipofuscinosis.

1q42, syntenic to the lyst gene in the beige mouse model, encodes a cytosolic 429-kDa protein138 composed to act as a lysosomal trafficking regulator, several mutant forms of which are now known.156 Bone marrow transplantation appears to be a therapy successful in eliminating nonneurologic symptoms.311

Glycogenosis Type II The infantile form of glycogenosis type II, or Pompe’s disease, presents with hypotonic weakness in the limbs and cardiac enlargement. It is related to a deficiency of ␣-1,4-glucosidase (acid maltase), the gene for which is located at 19q25.2-25.3. More than 40 mutations have now been verified.131 The peripheral nerves have shown extensive accumulations of glycogen in axons and Schwann cells89,103,192 unaccompanied by axonal loss or demyelination. Glycogen also accumulates in anterior horn cells.269 The glycogen was within membrane-bound inclusions according to Gambetti and co-workers,89 but was not membrane bound in the case described by Araoz and associates.9 No reports of similar CNS or PNS involvement are known in the more benign childhood cases or in the adult variant that presents with proximal muscle weakness without cardiomyopathy, although ubiquitous storage of lysosomal glycogen occurs in the juvenile and adult forms, even in blood lymphocytes. Enzyme replacement therapy has now been started in the infantile form.

Wolman’s Disease Wolman’s disease is an autosomal recessive disorder that becomes manifest in early infancy and commonly is fatal within the first 12 months of life. Its manifestation includes hepatosplenomegaly, adrenal calcification, intestinal dilatation, and steatorrhea.2 It is related to a

deficiency of acid cholesteryl ester hydrolase (acid lipase).234 Cholesteryl esters and triglycerides accumulate in most tissues, including Schwann and perineurial cells. 52 The ganglion cells of both Meissner’s and Auerbach’s plexus contain sudanophilic granules,54,151 and sudanophilic lipids have also been described in sympathetic neurons.148 Cholesterol ester storage disease is a mild allelic variant,6,182,183 the gene being located at 10q23.2-23.3 for both allelic conditions and affected by more than 10 mutations.11

Mucopolysaccharidoses and Oligosaccharidoses The mucopolysaccharidoses, now numbering I through IX with several subtypes (but types V and VIII have been abandoned) and 10 genes identified at different chromosomal loci,218 are a group of inherited conditions in which glycosaminoglycans (mucopolysaccharides) are stored within lysosomes in various tissues. Neurologic involvement, in which mental retardation is the main feature, is prominent only in the Hurler, Hunter, and Sanfilippo syndromes (MPS IH, II, and III, respectively), together with skeletal and other changes. Electron microscopy of nerve biopsy tissue may reveal membrane-bound striated inclusions in Schwann cells of pleomorphic pattern but often of zebra body type.39 They are also observed in fibroblasts.152 Such zebra bodies are only infrequently seen in Sanfilippo’s syndrome.39 Schwann cells may also contain lysosomal vacuoles, but not as many as endoneurial fibroblasts.243 Some axonal degeneration may be present.152 In the oligosaccharidoses—comprising genetically different ␣- and ␤-mannosidoses, fucosidosis, sialidosis or mucolipidosis I, aspartylglucosaminuria, mucolipidosis II or I-cell disease, and mild allelic mucolipidosis III and galactosialidosis—there is excessive urinary excretion of oligosaccharides. Schwann cell inclusions are a feature in

1865

Lysosomal and Peroxisomal Disorders

I-cell disease (inclusion-cell disease, or mucolipidosis II). The inclusions are large vacuolar lysosomes that may also contain membranous whorls.194 Peripheral nerve involvement has been described both in the cherry-red spot–myoclonus syndrome (juvenile or adult normosomatic type I sialidosis) and in childhood dysmorphic or type II sialidosis (Spranger syndrome). Both are oligosaccharidoses and are autosomal recessive disorders related to a deficiency of glycoprotein sialidase (neuraminidase). The residual activity of the enzyme in cultured fibroblasts is greater in the former than in the latter.227 The cherry-red spot–myoclonus syndrome begins in the second or third decade of life and is dominated by action myoclonus. A nonpigmentary retinal degeneration develops with cherry-red spots at the maculae. Tingling or burning paresthesias may be experienced in the extremities, the tendon reflexes may be depressed, and motor and sensory nerve conduction study results may be abnormal. Nerve biopsy has shown demyelination and inclusion material in Schwann cells.272 In childhood dysmorphic sialidosis, muscle weakness, sensory deficits, depressed tendon reflexes, and reduced NCV have suggested peripheral nerve involvement. In galactosialidosis and other sialic acid storage diseases, lysosomal vacuoles occur in Schwann cells and perineurial cells (Fig. 79–21) and lamellar bodies in neuronal perikarya of peripheral ganglia.139,159,206 Entrapment neuropathies, particularly the carpal and cubital tunnel syndromes, may be a feature in Hurler’s syndrome, probably related to ligamentous thickening because of deposition of mucopolysaccharides and collagen.152 Carpal tunnel compression of the median nerve is also common in Scheie’s syndrome200 (MPS IS), characterized by skeletal changes and other manifestations but without neurologic involvement, and in I-cell disease.

FIGURE 79–21 Numerous vacuolar lysosomes in a Schwann cell from a dermal nerve fascicle in Salla disease, a lysosomal sialic acid storage disorder.

Mucolipidosis IV This lysosomal disorder may appear as a severe form commencing in the first year of life, marked by psychomotor retardation, corneal clouding, retinal degeneration, and seizures,25 and a milder protracted form29 marked by microcephaly, mental retardation, tetraspasticity, corneal clouding, and retinal degeneration, thus similar to the severe form but with longer duration of the disease. Some 80% of all affected patients are Ashkenazi Jewish, most of them comprising two mutations in the gene MCOLN1, located at 19p13.2-13.3, which codes for the transmembrane protein mucolipin 1, a member of a new group of proteins, the mucolipins, analogous to the calcium channels of the TRP/TRPL family.25 Although clinical symptoms of the PNS may not be prominent or even apparent, widespread formation of lysosomal residual bodies, not only in the CNS but also in PNS neuronal perikarya,84 axons,307 Schwann cells of myelinated and unmyelinated axons, and perineurial cells101 (Fig. 79–22), attests to the involvement of the PNS. The ultrastructural spectrum in mucolipidosis IV encompasses vacuolar lysosomal residual bodies101 and a multitude of lamellar inclusions84,101 as well as lipopigments,29,84 reflecting deficiency of the miscoded transmembrane protein mucolipin 1, apparently affecting components of lipid, carbohydrate, and protein metabolism.

Peroxisomal Disorders Peroxisomes, like lysosomes, are single membrane-bound small intracellular organelles in almost all cells of the body, but particularly numerous in hepatic and renal cells. In their matrix they harbor more than 50 enzymes that are involved in both catabolic and anabolic processes. While

1866

Diseases of the Peripheral Nervous System

FIGURE 79–22 Mucolipidosis IV. Lamellar lysosomal inclusions (arrows) in Schwann cells and a perineural cell of a dermal nerve fascicle.

synthesis of phospholipids and isoprenoids encompasses the anabolic or synthetic component, both ␣- and ␤oxidations of particularly long-chain fatty acids represent the catabolic component. ␤-Oxidation of short-chain fatty acids is incomplete within peroxisomes, these only being completely oxidized within mitochondria. Unlike lysosomal disorders, the ultrastructural abnormalities in peroxisomal disorders are not pathognomonic. The changes merely reflect the number and size involved rather than being content specific. However, needle-like deposits of cholesterol are encountered in certain peroxisomal disorders, mostly in Schwann cells, macrophages, and other cells in adrenoleukodystrophy (ALD). Case reports from each entity are low in number, and morphologic, including electron microscopic, findings from extracerebral tissue biopsies lack diagnostic features. In general, peroxisomal disorders fall into two categories, those with defects in biogenesis of peroxisomes and those with deficiency of single peroxisomal enzymes.309 The latter group resemble lysosomal disorders. Peroxisomal membrane protein deficiency is found in ALD/adrenomyeloneuropathy (AMN), similar to the deficiency of single lysosomal enzymes in most of the lysosomal disorders. This group consists of ALD/AMN, Refsum disease, and most recently 2-methylacyl-coenzyme A (CoA) racemase deficiency, acyl-CoA oxidase deficiency, D-bifunctional protein deficiency, and peroxisomal thiolase deficiency. Peripheral neuropathy predominates in the clinical and morphologic features, usually developing during childhood or early adulthood. The infantile forms (i.e., neonatal ALD and infantile Refsum disease) belong to the peroxisomal biogenesis disorders. While there is marked elevation of phytanic as well as several other fatty acids in infantile Refsum disease,308 only phytanic acid is highly

elevated in Refsum disease. Probably because of the rapid progression resulting in early death, peripheral nerves are less involved or not involved at all clinically, although involvement could be documented at autopsy in infantile Refsum disease.26,294 Characteristic intracellular bilamellar inclusions were also identified by electron microscopy outside the PNS.294 The first group of peroxisomal disorders, which is marked by impaired biogenesis, has overlapping genetic defects and is largely composed of Zellweger syndrome, neonatal ALD, and infantile Refsum disease, all of autosomal recessive nature. The biochemical and molecular backgrounds are very complex, involving multiple genes. Impaired biogenesis results in multiple dysfunction or absence of proteins of both enzymatic and membranous types. Thus what are called individual disorders, such as Zellweger syndrome, are of genetic heterogeneity. The early onset implies very early interference with functional and developmental processes, particularly of the brain, thereby combining developmental and degenerative features in these disorders. Although there are morphologic abnormalities in the PNS, clinical involvement may be difficult to demonstrate in severe forms with a rapid fatal course. Only milder forms, such as infantile Refsum disease, have been found to be associated with peripheral neuropathy as reported in three affected Amish siblings.26

ADRENOLEUKODYSTROPHY AND ADRENOMYELONEUROPATHY ALD is a progressive disorder with X-linked recessive inheritance that affects both the nervous system and the adrenal cortex.213 The syndrome that was originally identified was a

1867

Lysosomal and Peroxisomal Disorders

progressive cerebral degeneration in young males, often associated with cortical blindness, deafness, optic atrophy, spasticity, and seizures. This was sometimes accompanied by evidence of adrenal failure. Autopsy studies showed extensive CNS demyelination and the presence of lamellar inclusions in brain macrophages, Schwann cells, and the adrenal glands.246 Cases were subsequently identified of males with a spastic paraplegia developing in the third or fourth decade of life and accompanied by a mild peripheral neuropathy.115 These cases were referred to as adrenomyeloneuropathy or adrenoleukomyeloneuropathy.190,313 They occurred in the same families as for typical ALD and therefore did not seem to be genetically distinct. In the Netherlands, AMN was found to be the most frequent phenotype.300 Moser and co-workers213 analyzed the clinical features in 409 cases of X-linked ALD/AMN. In 40% the presentation was a childhood-onset ALD, in 13% adolescent or adultonset ALD, in 20% AMN, and in 7% adrenal insufficiency without neurologic involvement. A further rare presentation is as a spinocerebellar syndrome presenting in early adult life.190,283 About 4% of female carriers have neurologic involvement, usually a spastic paraplegia, sometimes with an associated neuropathy. The gene has been mapped to the end of the long arm of the X chromosome,44,213 actually Xq28. Elongated, electron-lucent lamellar inclusions are present in histiocytes in the brain in patients with cerebral involvement, and such inclusions are also seen in cells of the adrenal cortex.260,313 Neuropathy is not a feature of the typical ALD in young males, but inclusions may be detected in Schwann cells in the absence of evidence of demyelination or axonal degeneration.246 Clinically evident peripheral neuropathy is present in cases of AMN. Patients with AMN show signs of adrenal insufficiency that begin in childhood, which are followed by progressive spastic paraparesis with an onset in the second or third decade. Other clinical findings are hypogonadism, sometimes scanty hair or baldness, and an occasional autonomic neuropathy.265 The peripheral neuropathy component of the disease may lead to lower motor neuron weakness in the legs and relative depression, but not loss, of the ankle jerks. NCV may be reduced. Mixed axonal-demyelination patterns have been found in extensive electrophysiologic studies.56,301

Pathology Nerve biopsy specimens in two cases of AMN examined by Griffin and co-workers115 showed loss of both myelinated and unmyelinated axons, an increase in endoneurial collagen, and numerous small onion bulb formations. These findings were also noted by Martin and associates.195 There was a loss of large myelinated fibers but a preservation of fiber density because of regenerative

activity. Some hypertrophic onion bulb changes were observed, and teased-fiber studies revealed evidence of chronic demyelination-remyelination. Fiber loss in distal nerves may at times be severe.259 On electron microscopy, Schwann cells may contain cytoplasmic inclusions of lamellar material (Fig. 79–23).246,247 Each lamella consists of two parallel electron-dense 2- to 5-nm leaflets separated by a space varying from 2 to 7 nm in width. Rows of glycogen particles and a clear space often surround these inclusions. Mitochondria in AMN dorsal root ganglion neurons showed many lipid inclusions, suggesting metabolic involvement of mitochondria as having a possible role in an impaired axoplasmic transport and a resulting axonopathy as the pathogenetic principle in AMN.245

Pathogenesis The salient biochemical feature of ALD is the accumulation of very-long-chain fatty acids (VLCAs), with an increase in those with 20- and 22-carbon chain lengths. The accumulation especially involves hexacosanoate, a saturated 26-carbon unbranched fatty acid, and is demonstrable in the cholesterol esters and gangliosides of the cerebral cortex and the adrenal cortex.136,203 It is also demonstrable in plasma210 and cultured skin fibroblasts.211 The metabolic defect involves the ␤-oxidation system for the degradation of these VLCAs.213,310 Degradation is impaired.213,310 This takes place in peroxisomes, and ALD/AMN is now considered to be a peroxisomal disorder.310 There is also an overproduction of VLCAs.295 The relationship between the elevation of VLCAs, clinical symptomatology, and the underlying genetic defect is still incompletely known. While the gene is located on Xq28, consisting of 10 exons and comprising 26 kb, the encoded ADL protein, a transmembrane protein, is located in the peroxisomal membrane, consisting of 745 amino acids. More than 400 mutations155 have been identified (available on the X-linked Adrenoleukodystrophy Database website: www.x-ald.nl), among them 68.5% private mutations. The entire mutational spectrum consists of some 53% missense mutations, 24% frameshift mutations, 5% in-frame deletions or insertions, and 2.5% splicing defects.214 Mutations in the ALD gene, ABCD1, severely alter the half-transporter function of the ALD protein because this protein is a member of the ATP-binding cassette (ABC) transporter protein family90 and as an ABC half-transporter imports VLCAs into the peroxisome.155 While genotype-phenotype correlation has largely failed,155,214,282 truncation of the ALD protein by 65 initial amino acids has suggested a particular correlation with AMN.231 Several mouse models have recently been produced,68 one deficient in very-long-chain acyl-CoA synthetase, the peroxisomal enzyme known to show reduced activity in ALD.127

1868

Diseases of the Peripheral Nervous System

FIGURE 79–23 Electron micrograph of a transverse section through sural nerve from a patient with adrenoleukodystrophy. Inclusion material (im) is present in the cytoplasm of a Schwann cell. ax ⫽ axon; Sn ⫽ Schwann cell nucleus. Bar: 0.1 ␮m.

Diagnostic assay is possible for affected persons using cultured skin fibroblasts employing the C26:C22 fatty acid ratio, and it will identify a high proportion of heterozygotes.211 The different phenotypic varieties appear to have similar ratios,211,229,230 and the reason for these different clinical patterns is uncertain. Environmental factors may be relevant, because some of the accumulated VLCAs may be of dietary origin.157 Because diagnostic tests may not be fully reliable in obligate heterozygotes, mutational analysis is the only safe procedure.155

Treatment Recent observations have suggested that diets with a very low content of VLCAs supplemented with glycerol trioleate (C18:1) and glycerol erucate (C22:1) may prevent neurologic progression in patients with asymptomatic ALD33 and lower VLCA levels, but not reduce clinical progression.214 Apart from now established effective bone marrow transplantation, hematopoietic stem cell transplantation (HSCT) early in the course of the clinical disease may stop central demyelination, but may also show unpredictable outcome.127 Magnetic resonance spectroscopy identifying increase in N-acetyl aspartate in

cerebral white matter after HSCT, and thus suggesting axonal preservation, may aid posttransplantation prognostication.320

NEONATAL ADRENOLEUKODYSTROPHY This is a separate disorder of autosomal recessive inheritance.12 The clinical course may be rapidly progressive, with death in early infancy, or survival until over the age of 2 years is possible. Affected children have a dysmorphic facies, are hypotonic, and have seizures. Hepatomegaly, pigmentary retinopathy, cataract, optic atrophy, and deafness may develop, as may diffuse amyotrophy. The tendon reflexes range from brisk to absent. Electromyography may reveal denervation, and motor NCV is occasionally reduced. Adrenal insufficiency does not occur. Autopsy studies have shown diffuse white matter degeneration in the cerebral and cerebellar hemispheres. Trilamellar inclusions may be found in the liver. Sural nerve biopsy has shown a “decreased density of myelin sheaths” but no abnormal inclusions.12 There are abnormalities of multiple peroxisomal enzymes, resulting in increased plasma levels of VLCAs

1869

Lysosomal and Peroxisomal Disorders

and, less regularly, of phytanic acid and increased trihydroxycoprostanic acid in bile fluid.12

REFSUM DISEASE Clinical Aspects Refsum disease is the other major peroxisomal disorder affecting the peripheral nerves. It is a multiorgan condition based on a single peroxisomal enzyme deficiency. In an earlier edition of this book268 it was termed “phytanic acid storage disease.” The name appears misleading because phytanic acid accumulation, a major feature in Refsum disease, also occurs in other peroxisomal disorders. The latter, such as infantile Refsum disease, can be distinguished by other different chemical aberrations. The age of onset in Refsum disease varies from infancy129 to late adulthood. The disease is progressive if not treated early and continuously. Although not necessarily the first clinical sign, polyneuropathy is a consistent feature in fully developed disease and has been part of the diagnostic tetrad since its first description, entailing polyneuropathy, pigmentary retinopathy, ataxia, and elevated cerebrospinal fluid protein content, giving rise to the two original terms heredoataxia hemeralopia polyneuritiformis251 and heredopathia atactica polyneuritiformis.252 It had been included as type IV hereditary motor and sensory neuropathy. All PNS divisions—motor, sensory, and autonomic—are affected, with muscle weakness and atrophy, disturbances of vibration and position senses, and occasional gastrointestinal, cardiac, and pupillary dysfunctions. The neuropathy commences distally and symmetrically. Deep tendon reflexes are reduced and finally abolished. Owing to the continuous demyelination and remyelination of peripheral nerves, several nerves may be palpably enlarged, assigning Refsum disease to the clinically heterogeneous group of hypertrophic neuropathies. Loss of deep tendon reflexes appears to be the most sensitive finding indicative of PNS involvement.268 Electrophysiologically, reduced conduction velocities are common, although a rare axonal neuropathy has been reported.91 A pigmented retinopathy is considered to be obligatory in Refsum disease. Retinal degeneration starts in the periphery, proceeding toward the center. Abnormal pigmentation in the fundus is said to be fine and has the pepper-and-salt appearance rather than coarse, “bony spicule” type.268 An abnormal electroretinogram accompanies this retinal degeneration. Night blindness is usually the presenting symptom and appears to be a constant feature among the various symptoms in Refsum disease.268 Optic atrophy and cataract as well as pupillary abnormalities are additional features. Cerebellar signs (e.g., nystagmus, intention tremor, and ataxia) in patients with Refsum disease have been

controversial. Although they originally formed one of the cardinal manifestations, morphologic changes in the cerebellum are limited. Other organs affected are the heart, by cardiomyopathy or conduction defects, occasionally resulting in sudden death; the skin, with dryness or ichthyosis; and the skeleton, manifested by epiphyseal abnormalities. Biochemically, Refsum disease is marked by highly elevated phytanic acid in the plasma that is also stored in liver and fatty tissue. Deficiency of the peroxisomal enzyme phytanoyl-CoA hydroxylase (PhyH) impairs adequate ␣-oxidation309,318 of phytanic acid and thereby prevents entry of the degradation product into the pathway of ␤-oxidation. Because phytanic acid exclusively derives from food rather than from endogenous production, respective dietary regimens may positively influence the clinical symptoms by accelerating reduction in phytanic acid plasma levels. However, these levels may only slowly be lowered by dietary means, offset by release of phytanic acid from body stores. Thus plasmapheresis has successfully been added to the therapeutic regimen. If treated early, long-term polyneuropathy may actually remain subclinical for even more than three decades.187 Molecularly, the gene PHYH on chromosome 10p13 for the peroxisomal enzyme phytanoyl-CoA hydroxylase has been found to be mutated in Refsum disease by missense mutations and deletions.141 However, not all patients with “Refsum symptomatology” are linked to this gene locus, because several families could be excluded by linkage studies,319 strongly suggesting genetic heterogeneity of Refsum disease. Hence a second gene, PEX7, recently was identified on chromosome 6q22-24, the mutations which are responsible not only for a mild form of Refsum disease but also for a more severe type of peroxisomal disorder, rhizomelic chondrodysplasia punctata type 1.298

Pathology Historical Aspects The early history of the way in which the histopathology of Refsum disease unfolded has been documented by Fardeau.78 He described it as possessing “much of the suspense of a good detective novel!” The neuropathology of Refsum first patient with heredopathia atactica polyneuritiformis was described by Jansen in 1938, but only the CNS was examined. Foci of myelin destruction were observed in the brainstem, and fatty deposits were seen in the leptomeninges, ependyma, choroid plexus, and pallidum. Swelling of neurons was noted at all levels examined. Cammermeyer undertook an autopsy on the sister of this patient in 1941. Fatty deposits were seen in the same distribution as in the first case. In addition, there was atrophy of scattered fiber tracts and chromatolysis of anterior horn cells.

1870

Diseases of the Peripheral Nervous System

Because of difficulties in Norway related to the war, sections containing fragments of retroperitoneal nerves were misplaced and were only discovered later when material was being assembled for shipment to the United States. The nervous system from a further patient, unrelated to the first two, was obtained in 1947. The inferior olivary nuclei were found to be atrophic and the olivocerebellar tracts had degenerated. The peripheral nerves had still not been examined. In 1948, a third adult from Refsum original series died and, at autopsy, Cammermeyer noticed for the first time that both the brachial and lumbosacral plexus were thickened throughout their entire length. He again observed an accumulation of fat in the meninges and pallidum and also noted a partial fiber loss in the gracile fasciculi. More importantly, he found, in the peripheral nerves, “onion bulb” changes characteristic of “interstitial hypertrophic neuropathy.” This new observation was included in a report by Refsum and co-workers253 that was mainly concerned with the clinical presentation of four new childhood cases. The first non-Norwegian case was documented in 1950 by Reese and Bareta250 in a North American patient. The clinical features corresponded closely to those described by Refsum. At postmortem examination, macroscopic thickening of the peripheral somatic and autonomic nerves was observed. Histologic study of the peripheral nerves demonstrated fiber loss and again showed “onion bulb” changes. It was not then known whether the concentrically proliferated cells were Schwann cells or fibroblasts, but the changes of “interstitial hypertrophic neuropathy” were considered at that stage to be pathognomonic of a single entity, Déjérine-Sottas disease. For this reason, Reese and Bareta questioned the claim that Refsum disease was a separate disorder and concluded that their case was a variant form of Déjérine-Sottas disease. When the sections that contained fragments of retroperitoneal nerves from the second case of Refsum original series that were shipped from Norway were recovered, they were examined by Cammermeyer, together with specimens of peripheral nerve from the third Norwegian case and another studied in Sweden by Kjellson in 1953.158 Cammermeyer concluded in a report published in 1956 that the major neuropathologic change in Refsum disease was hypertrophic peripheral neuropathy.53 The presence of an onion bulb neuropathy was confirmed in several subsequent reports,74,80,109,219 although in others only nonspecific changes were documented,235,262,304 including a publication by van Bogaert and co-workers297 on a portion of nerve from the patient investigated biochemically by Kahlke and Richterich.149 With the introduction of electron microscopy to neuropathology, it was later established that hypertrophic neuropathy represents a concentric proliferation of Schwann cells that is a nonspecific consequence of repeated segmental demyelination and remyelination.291,314

The first definite electron microscopic study in Refsum disease was undertaken in 1969 by Fardeau and Engel.80 It confirmed the schwannian nature of the hypertrophic changes in peripheral nerve. Histologic Changes The changes in peripheral nerve in classic Refsum disease as seen by light microscopy were reviewed by Fardeau.78 Involvement of the peripheral nerves is constant, but its severity is variable. Nerve hypertrophy may be diffuse and clinically evident,250 but usually the peripheral nerves are only slightly engaged. The enlargement may only be detectable post mortem because the nerve hypertrophy is most evident in the lumbosacral and brachial plexus. The spinal roots are sometimes affected. In one report they were believed to be thinned but the posterior root ganglia were considered to be enlarged, together with the spinal nerves immediately distal to the ganglia.109 The nerve hypertrophy is often irregular and has been described as forming localized swellings of gelatinous or cystic appearance. Macroscopically, transverse sections throughout these swellings have revealed that the nerve fibers are embedded in a grayish mucoid matrix. Similar appearances have occasionally been reported in nerve biopsy specimens.79 Transverse sections of paraffin- or epoxy-embedded nerves have shown depletion of myelinated fibers of varying severity, with normal or abnormally thin myelin sheaths on surviving fibers. Segmental demyelination has been documented in teased-fiber preparations.78 The degree of hypertrophic change is variable— sometimes it is striking, with densely packed whorls (Fig. 79–24). At times the whorls are separated by large endoneurial spaces. Many unmyelinated axons are present within the onion bulbs. The expanded endoneurial spaces may stain metachromatically with cresyl violet and toluidine blue,53 or may show slight positivity to PAS and a strong reaction with Alcian blue.79 Occasional fibroblasts, macrophages, and mast cells are present in the endoneurium. Refractile droplets have been noted in Schwann cells, but substantial deposits of lipid have not been observed. The basal lamina around capillaries and arterioles has been noted to be markedly thickened, and the endoneurium collagenized, but the perineurium has usually been considered to be normal. Electron microscopy has shown typical concentric Schwann cell proliferation in some cases78–80 (Fig. 79–25). The center of the onion bulb is occupied by either myelinated or unmyelinated axons or may merely contain denervated Schwann cells. As observed on light microscopy, the Schwann cells or the whorls are frequently associated with unmyelinated axons. In other specimens, hypertrophic changes may not be prominent. Fardeau and Engel80 commented on two types of inclusion in Schwann cells, the first being rounded osmiophilic bodies that are

Lysosomal and Peroxisomal Disorders

1871

FIGURE 79–24 Transverse section of paraffin-embedded peripheral nerve from a case of Refsum disease showing Schwann cell proliferation with the formation of numerous “onion bulbs.” Masson trichrome (⫻520). (Courtesy of Professor J. Lapresle.)

probably lipid in nature. Sometimes they display a lamellar pattern with a periodicity of about 15 nm. These were usually found in the vicinity of the nucleus. The second type of inclusion consisted of large crystalline bodies with a regular three-dimensional pattern. They were believed to be composed of parallel filaments, 6 nm in diameter, forming a regular hexagonal lattice with a 10.5- to 12-nm center-to-center spacing in a plane perpendicular to their long axis (Fig. 79–26). The inclusions were bounded by a double membrane, and occasionally internal cristae-like structures were seen. An origin from mitochondria was therefore suggested, and this has been the conclusion of subsequent authors who have commented on these bodies. The inclusions are nonspecific185,263 and have been observed in amyloid and diabetic neuropathies.286 Fardeau and Engel80 believed that they were more frequent in Refsum disease. In one case they found abnormally large

FIGURE 79–25 Electron micrograph of a transverse section through a sural nerve biopsy from a patient with Refsum disease showing two small onion bulb lesions surrounding myelinated nerve fibers (center and upper left). A regenerative cluster is present in lower right. (⫻5300.) (Courtesy of Dr. R. H. M. King.)

mitochondria with few cristae and a granular matrix in some unmyelinated axons.80 Fardeau78 also commented on the presence of typical lipofuscin granules in the cytoplasm of Schwann cells. These are found in Schwann cells associated with both unmyelinated and myelinated (Fig. 79–27) axons. This is unusual because normally they are confined to the Schwann cells of Remak fibers. No abnormal filamentous or other material has been detected in the endoneurium. Fardeau78 commented that electron microscopy confirmed thickening and reduplication of the basal laminae around endoneurial capillaries. This is also observable in hereditary motor and sensory neuropathy type I,45 in which hypertrophic change that may be secondary to endothelial cell hyperplasia also occurs. No ultrastructural abnormalities of the capillary endothelial cells have been commented on in Refsum disease.

1872

Diseases of the Peripheral Nervous System

FIGURE 79–26 A, Electron micrograph of a large crystalline inclusion within a rounded Schwann cell encircling an unmyelinated axon (⫻32,000). B and C, Serial sections of the inclusion, showing the regular arrangement of filaments, the presence of dense amorphous zones, and a crista-like structure (arrow). (⫻65,000.) (From Fardeau, M., Abelanet, R. Laudat, P., et al.: Maladie de Refsum : étude histologique ultrastructurale et biochimique d’une biopsie de nerf périphérique. Rev. Neurol. [Paris] 122:185, 1970. Reprinted by permission of Masson & Cie.)

INFANTILE REFSUM DISEASE Little information is available so far as to the pathology in the peripheral nerves in infantile Refsum disease. Nerve biopsy has been reported to show demyelination and loss of myelinated nerve fibers without hypertrophic changes.244

PEROXISOMAL 2-METHYLACYL-CoA RACEMASE DEFICIENCY Racemase is an enzyme located in peroxisomes and mitochondria converting R-stereoisomers to S-stereoisomers of phytanic and pristanic acids and, therefore, associated with peroxisomal ␤-oxidation. The gene for this enzyme has

Lysosomal and Peroxisomal Disorders

1873

FIGURE 79–27 Electron micrograph showing lipofuscin in cytoplasm of Schwann cell of a myelinated nerve fiber in a sural nerve biopsy from a case of Refsum disease (⫻49,600). (Courtesy of Dr. R. H. M. King.)

recently been cloned and mutations have been identified that are responsible for an adult-onset motor sensory neuropathy.82 This new form of peroxisome-related neuropathy developed in two unrelated adults, a man 44 years old at the time of examination, and a woman 48 years old at the time of clinical onset. They clinically as well as electrophysiologically suffered from axonal neuropathy and demyelinating neuropathy, respectively.82

REFERENCES 1. Abraham, K., and Lampert, P.: Intraneuronal lipid deposits in metachromatic leukodystrophy. Neurology 13:686, 1963. 2. Abramov, A., Schorr, S., and Wolman, M.: Generalised xanthomatosis with calcified adrenals. A. M. A. J. Dis. Child. 91:282, 1956. 3. Allen, N., and De Veyra, E.: Microchemical and histochemical observations in a case of Krabbe’s leukodystrophy. J. Neuropathol. Exp. Neurol. 26:456, 1967. 4. Allen, R. J., McCusker, J. J., and Tourtelotte, W. W.: Metachromatic leucodystrophy: clinical, histochemical and cerebrospinal fluid abnormalities. Pediatrics 30:629, 1962. 5. Alzheimer, A.: Beiträge zur Kenntnis der pathologischen Neuroglia und ihre Beziehungen zu den Abbauvorgängen im Nervengewebe. In Alzheimer, A. and Nissl, F. (eds.): Histologische und histopathologische Arbeiten, Vol. III. Jena, Gustav Fischer, p. 401, 1910. 6. Anderson, R. A., Bryson, G. M., and Parks, J. S.: Lysosomal acid lipase mutations that determine phenotype in Wolman and cholesterol ester storage disease. Mol. Genet. Metab. 68:333, 1999.

7. Andrews, J. M., and Cancilla, P. A.: Cytoplasmic inclusions in human globoid cell leukodystrophy: Krabbe’s disease. Arch. Pathol. 89:53, 1970. 8. Anzil, A. P., Blinzinger, K., and Martinius, J.: Ultrastructure of storage materials in metachromatic leukodystrophy peripheral neuropathy and in rabbit tissues after sulfatide application with special reference to pleated lamellar systems. J. Microsc. 18:173, 1973. 9. Araoz, C., Sun, C. N., Shenefelt, R., and White, H. J.: Glycogenosis type II (Pompe’s disease): ultrastructure of peripheral nerves. Neurology 24:739, 1974. 10. Arvidson, J., Hagberg, B., Mansson, J.-E., and Svennerholm, L.: Late onset globoid cell leukodystrophy (Krabbe’s disease)—Swedish case with 15 years of followup. Acta Paediatr. 84:218, 1995. 11. Assmann, G., and Seedorf, U.: Acid lipase deficiency: Wolman disease and cholesteryl ester storage disease. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3551, 2001. 12. Aubourg, P., Scotto, J., Rocchiccioli, F., et al.: Neonatal adrenoleukodystrophy. J. Neurol. Neurosurg. Psychiatry 49:77, 1986. 13. Aurebeck, G., Osterberg, K., Blaw, M., et al.: Ultrastructural observations on metachromatic leucodystrophy. J. Neuropathol. Exp. Neurol. 23:184, 1964. 14. Austin, H. J.: Studies in metachromatic leukodystrophy: XI. Therapeutic considerations. In Bergsma, O., Desnick, R. J., Bernlohr, P. W., and Krivit, W. (eds.): Enzyme Therapy in Genetic Diseases. Baltimore, Williams & Wilkins, p. 125, 1973. 15. Austin, H. J.: Studies in metachromatic leukodystrophy: XII. Multiple sulfatase deficiency. Arch. Neurol. 28:258, 1973.

1874

Diseases of the Peripheral Nervous System

16. Austin, J. H.: Metachromatic form of diffuse cerebral sclerosis: 1. Diagnosis during life by urine sediment examination. Neurology 7:415, 1957. 17. Austin, J. H.: Metachromatic form of diffuse cerebral sclerosis: 2. Diagnosis during life by isolation of metachromatic lipids from urine. Neurology 7:716, 1957. 18. Austin, J. H.: Observations in metachromatic leucoencephalopathy. Trans. Am. Neurol. Assoc. 83:149, 1958. 19. Austin, J. H.: Recent studies in the metachromatic and globoid cell forms of diffuse sclerosis. In Korey, S. R., Pope, A., and Robins, E. (eds.): Ultrastructure and Metabolism of the Nervous System (Association for Research and Nervous and Mental Diseases). Baltimore, Williams & Wilkins, p. 189, 1962. 20. Austin, J. H.: Recent studies in the metachromatic and globoid body forms of diffuse sclerosis. In Folch-Pi, J., and Bauer, H. J. (eds.): Brain Lipids and Lipoproteins and the Leucodystrophies. Amsterdam, Elsevier, p. 120, 1963. 21. Austin, J. H., Armstrong, D., and Margolis, G.: Studies of globoid leukodystrophy in dogs. Neurology 18:300, 1968. 22. Austin, J. H., Armstrong, D., and Shearer, L.: Metachromatic form of diffuse cerebral sclerosis. 5. The nature and significance of low sulfatase activity: a controlled study of brain, liver, and kidney in 4 patients with metachromatic leucodystrophy. Arch. Neurol. 13:593, 1965. 23. Austin, J. H., Lehfeldt, D., and Maxwell, W.: Experimental “globoid bodies” in white matter and chemical analysis in Krabbe’s disease. J. Neuropathol. Exp. Neurol. 20:284, 1961. 24. Austin, J. H., Suzuki, K., Armstrong, D. A., et al.: Studies in globoid (Krabbe) leukodystrophy (GLD) V. Controlled enzymatic studies in ten human cases. Arch. Neurol. 23:502, 1970. 25. Bach, G.: Mucolipidosis type IV: minireview. Mol. Genet. Metab. 73:197, 2001. 26. Bader, P. I., Dougherty, S., Cangany, N., et al.: Infantile Refsum’s disease in four Amish sibs. Am. J. Med. Genet. 90:110, 2000. 27. Bajaj, N. P. S., Waldman, A., Orrell, R., et al.: Familial adult onset of Krabbe’s disease resembling hereditary spastic paraplegia with normal neuroimaging. J. Neurol. Neurosurg. Psychiatry 72:635, 2002. 28. Bardosi, A., Friede, R. L., Ropte, S., and Goebel, H. H.: A morphometric study on sural nerves in metachromatic leucodystrophy. Brain 110:683, 1987. 29. Bargal, R., Goebel, H. H., Latta, E., and Bach, G.: Mucolipidosis IV: Novel mutation and diverse ultrastructural spectrum in the skin. Neuropediatrics 33:199, 2002. 30. Bargeton, E.: The metachromatic form of leucodystrophy and its relationship to lipidoses and demyelination in other metabolic disease. In Folch-Pi, J., and Bauer, H. J. (eds.): Brain Lipids and Lipoproteins and the Leucodystrophies. Amsterdam, Elsevier, p. 90, 1963. 31. Barnes, D., Misra, P., Young, E. P., et al.: An adult-onset hexosaminidase A deficiency syndrome with sensory neuropathy and intranuclear ophthalmoplegia. J. Neurol. Neurosurg. Psychiatry 54:1112, 1991. 32. Bass, N. H., Witmer, E. J., and Dreifuss, F. E.: A pedigree study of metachromatic leukodystrophy: biochemical identification of the carrier state. Neurology 20:50, 1970.

33. Bebin, E. M., Gomez, M. R., Forbes, G. S., and Moser, H. W.: Cerebral lesions of adrenoleukodystrophy remain static and patients asymptomatic during dietary therapy. Ann. Neurol. 28:436, 1990. 34. Beck, M., Sieber, N., and Goebel, H. H.: Progressive cerebellar ataxia in juvenile GM2-gangliosidosis type Sandhoff. Eur. J. Pediatr. 157:866, 1998. 35. Beneke, E.: Ein Fall hochgradigster ausgedehnter Sklerose des Zentralnervensystems bei einem 1 3/4 -jährigen Knaben. Arch. Kinderheilkd. 42:420, 1908. 36. Bernal, O. G., and Lenn, N.: Multiple cranial nerve enhancement in early infantile Krabbe’s disease. Neurology 54:2348, 2000. 37. Bertrand, I., Thieffry, S., and Bargeton, E.: Leucodystrophie familiale et détermination splénohépatique caractérisant un trouble général du métabolisme. Rev. Neurol. (Paris) 91:161, 1954. 38. Beutler, E., and Grabowski, G.A.: Gaucher disease. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3635, 2001. 39. Bischoff, A.: The peripheral nerves. In Johannesson, J. V. (ed.): Electron Microscopy in Human Medicine. New York, McGraw-Hill, p. 137, 1979. 40. Bischoff, A., and Ulrich, J.: Amaurotische Idiotie in Verbindung mit metachromatischer Leukodystrophie: Übergangsform oder Kombination? Acta Neuropathol. (Berl.) 8:292, 1967. 41. Bischoff, A., and Ulrich, J.: Peripheral neuropathy in globoid cell leucodystrophy (Krabbe’s disease): ultrastructural and histochemical findings. Brain 92:861, 1969. 42. Bloch-Zupan, A., Lecolle, S., and Goldberg, M.: Galactosylceramide lipidosis (Krabbe’s disease) and deciduous dental tissues: a case report. J. Submicrosc. Cytol. Pathol. 26:425, 1994. 43. Blume, R. S., and Wolff, S. M.: The Chédiak-Higashi syndrome: studies in four patients and a review of the literature. Medicine (Baltimore) 51:247, 1972. 44. Boué, J., Oberle, I., and Heilig, R.: First-trimester prenatal diagnosis of adrenoleukodystrophy by determination of very long chain fatty acid levels and by linkage analysis to a DNA probe. Hum. Genet. 69:727, 1985. 45. Bradley, J., Thomas, P. K., King, R. H. M., et al.: Morphometry of endoneurial capillaries in diabetic sensory and autonomic neuropathy. Diabetologica 33:611, 1990. 46. Brady, R. O.: Therapy for the sphingolipidoses. Arch. Neurol. 55:1055, 1998. 47. Brady, R. O.: Enzyme replacement therapy: conception, chaos and culmination. Philos. Trans. R. Soc. London B. Biol. Sci. 29:915, 2003. 48. Brady, R. O., Kanfer, J. N., Mock, M., and Frederickson, D. S.: The metabolism of sphingomyelin II: evidence of an enzyme deficiency in Niemann-Pick disease. Proc. Natl. Acad. Sci. U. S. A. 55:366, 1966. 49. Brion, S., Mikol, J., and Graveleau, J.: Leucodystrophie métachromatique de l’adulte jeune: étude clinique, biologique et ultrastructurale. Rev. Neurol. (Paris) 122:161, 1970. 50. Burck, U., Harzer, K., Goebel, H. H., et al.: Ultrastructural pathology of skin biopsy and fibroblast enzyme studies in a

Lysosomal and Peroxisomal Disorders

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61. 62.

63. 64.

65.

66. 67.

case of GM-2 gangliosidosis with deficient hexosaminidase A and thermolabile hexosaminidase B. Neuropädiatrie 11:161, 1980. Burck, U., Moser, H. W., Goebel, H. H., et al.: A case of lipogranulomatosis Farber: some clinical and ultrastructural aspects. Eur. J. Pediatr. 143:203, 1985. Byrd, J. C., and Powers, J. M.: Wolman’s disease: ultrastructural evidence of lipid accumulation in central nervous system. Acta Neuropathol. (Berl.) 54:37, 1979. Cammermeyer, J.: Neuropathological changes in hereditary neuropathies: manifestation of the syndrome heredopathia atactica polyneuritiformis in the presence of interstitial hypertrophic polyneuropathy. J. Neuropathol. Exp. Neurol. 15:340, 1956. Castro, M., Rosati, P., Boldrini, R., et al.: Wolman’s disease diagnosed by intestinal biopsy. Ital. J. Gastroenterol. Hepatol. 31:610, 1999. Cengiz, N., Özbenli, T., Onar, M., et al.: Adult metachromatic leukodystrophy: three cases with normal nerve conduction velocities in a family. Acta Neurol. Scand. 105:454, 2002. Chaudhry, V., Moser, H. W., and Cornblath, D. R.: Nerve conduction studies in adrenomyeloneuropathy. J. Neurol. Neurosurg. Psychiatry 61:181, 1996. Comabella, M., Wayne, J. S., Raguer, N., et al.: Late-onset metachromatic leukodystrophy clinically presenting as isolated peripheral neuropathy: compound heterozygosity for the IVS2 ⫹ 1G:A mutation and a newly identified missense mutation (Thr408Ile) in a Spanish family. Ann. Neurol. 50:108, 2001. Cosma, M. P., Pepe, S., Annunziata, I., et al.: The multiple sulfatase deficiency gene encodes an essential and limiting factor for the activity of sulfatases. Cell 113:421, 2003. Couchot, J., Pluot, M., Schmauch, M.-A., et al.: La mucosulfatidose: étude de trois cas familiaux. Arch. Fr. Pediatr. 31:775, 1974. Coulter-Mackie, M. B., Applegarth, D. A., Toone, J. R., et al.: Isolated peripheral neuropathy in atypical metachromatic leukodystrophy: a recurrent mutation. Can. J. Neurol. Sci. 29:159, 2002. Cravioto, H.: In vivo and in vitro studies of metachromatic leukodystrophy. J. Neuropathol. Exp. Neurol. 26:157, 1967. Cravioto, H., O’Brien, J. S., Landing, B. H., and Finck, B.: Ultrastructure of peripheral nerve in metachromatic leucodystrophy. Acta Neuropathol. (Berl.) 7:111, 1966. Crome, L., Hanefeld, F., Patrick, D., and Wilson, J.: Late onset globoid cell leukodystrophy. Brain 96:841, 1973. Czmok, E., Bischoff, A., Harzer, K., and Benz, H. O.: Metachromatische Leukodystrophie: Klinik und intravitale Diagnostik einer familiären adulten Form der metachromatischen Leukodystrophie (MLD). J. Neurol. 207:189, 1974. Dayan, A. D.: Peripheral neuropathy of metachromatic leucodystrophy: observations on segmental demyelination and remyelination and the intracellular distribution of sulphatide. J. Neurol. Neurosurg. Psychiatry 30:311, 1967. De Silva, K. L., and Pearce, J.: Neuropathy of metachromatic leukodystrophy. J. Neurol. Neurosurg. Psychiatry 36:30, 1973. De Stefano, N., Dotti, M. T., Mortilla, M., et al.: Evidence of diffuse brain pathology and unspecific genetic characterization in a patient with an atypical form of adult-onset Krabbe disease. J. Neurol. 247:226, 2000.

1875

68. Depreter, M., Espeel, M. F., and Roels, F.: Human peroxisomal disorders. Microsc. Res. Tech. 61:203, 2003. 69. Dierks, T., Schmidt, B., Borissenko, L. V., et al.: Multiple sulfatase deficiency is caused by mutations in the gene encoding the human C(alpha)-formylglycine generating enzyme. Cell 113:435, 2003. 70. Dolman, C. L., MacLeod, P. M., and Chang, E.: Fine structure of cutaneous nerves in ganglioside storage disease. J. Neurol. Neurosurg. Psychiatry 40:588, 1977. 71. Duchen, L. W., Eicher, E. M., Jacobs, J. M., et al.: Hereditary leucodystrophy in the mouse: the new mutant Twitcher. Brain 103:695, 1980. 72. Dunn, H. G., Lake, B. D., Dolman, C. L., and Wilson, J.: The neuropathy of Krabbe’s infantile cerebral sclerosis (globoid cell leucodystrophy). Brain 92:329, 1969. 73. Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased-fibre measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968. 74. Edström, R., Grontoft, O., and Sandring, H.: Refsum disease—three siblings, one autopsy. Acta Psychiatr. Scand. 34:39, 1959. 75. Einarson, L., and Neel, A. V.: Beitrag zur Kenntnis sklerosierender Entmarkungsprozesse im Gehirn mit besonderer Berücksichtigung der diffusen Sklerose. Acta Jutland. 10:1, 1938. 76. Elleder, M.: Eccrine gland involvement in Krabbe’s disease. Virchows Arch. [A] 421:351, 1992. 77. Eto, Y., Wiesman, U., and Herschkowitz, N. N.: Sulfogalactosylsphingosine sulfatase: characteristics of the enzyme and its deficiency in metachromatic leukodystrophy in human cultured skin fibroblasts. J. Biol. Chem. 249:4955, 1974. 78. Fardeau, M.: Pathology of Refsum disease. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 1693, 1984. 79. Fardeau, M., Abelanet, R., Laudat, P., and Bonduelle, M.: Maladie de Refsum: étude histologique, ultrastructurale et biochimique d’une biopsie de nerf périphérique. Rev. Neurol. (Paris) 122:185, 1970. 80. Fardeau, M., and Engel, W. K.: Ultrastructural study of a peripheral nerve biopsy in Refsum disease. J. Neuropathol. Exp. Neurol. 28:278, 1969. 81. Felice, K. J., Gomez, L. M., Natowicz, M., et al.: Adultonset MLD: a gene mutation with isolated polyneuropathy. Neurology 55:1036, 2000. 82. Ferdinandusse, S., Denis, S., Clayton, P. T., et al.: Mutations in the gene encoding peroxisomal ␣-methylacylCoA racemase cause adult-onset sensory motor neuropathy. Nat. Genet. 24:188, 2000. 83. Fischer, G., and Jatzkewitz, H.: The activator of cerebroside sulphatase: purification from human liver and identification as a protein. Hoppe-Seyler Z. Physiol. Chem. 356:605, 1975. 84. Folkerth, R. D., Alroy, J., Lomakina, I., et al.: Mucolipidosis IV: morphology and histochemistry of an autopsy case. J. Neuropathol. Exp. Neurol. 54:154, 1995. 85. Fankhauser, R., Luginbühl, H., and Hartley, W. J.: Leukodystrophie vom Typus Krabbe beim Hund. Arch. Tierheilkd. 105:198, 1963.

1876

Diseases of the Peripheral Nervous System

86. Fujibayashi, S., and Wenger, D. A.: Biosynthesis of the sulfatide/GM1 activator protein (SAP-1) in control and mutant skin fibroblasts. Biochim. Biophys. Acta 875:544, 1986. 87. Fullerton, P. M.: Peripheral nerve conduction in metachromatic leukodystrophy (sulphatide lipidosis). J. Neurol. Neurosurg. Psychiatry 27:100, 1964. 88. Gabreëls, F. J. M., Lamers, K., Kok, K., et al.: The biochemical differentiation between heterozygote carriers of metachromatic leucodystrophy and normal persons. Neuropädiatrie 2:461, 1974. 89. Gambetti, P., DiMauro, S., and Baker, L.: Nervous system in Pompe’s disease. J. Neuropathol. Exp. Neurol. 30:412, 1971. 90. Gärtner, J., Dehmel, T., Klusmann, A., and Roerig, P.: Functional characterization of the adrenoleukodystrophy protein (ALDP) and disease pathogenesis. Endocr. Res. 28:741, 2002. 91. Gelot, A., Vallat, J. M., Tabaraud, F., and Rocchiccioli, F.: Axonal neuropathy and late detection of Refsum’s disease. Muscle Nerve 18:667, 1995. 92. Ginsberg, L., and Gershfeld, N.: Membrane bilayer instability and the pathogenesis of disorders of myelin. Neurosci. Lett. 130:133, 1991. 93. Goebel, H. H.: Morphology of the gangliosidoses. Neuropediatrics 15:97, 1984. 94. Goebel, H. H.: Hereditary metabolic neuropathies. Zentralbl. Allg. Pathol. Pathol. Anat. 136:503, 1990. 95. Goebel, H. H.: Neurodegenerative diseases: biopsy diagnosis in children. In Garcia, J. H. (ed.): Neuropathology: The Diagnostic Approach. London, Mosby, p. 581, 1997. 96. Goebel, H. H.: Extracerebral biopsies in neurodegenerative diseases of childhood. Brain Dev. 21:435, 1999. 97. Goebel, H. H., Argyrakis, A., Shimokawa, K., et al.: Adult metachromatic leukodystrophy. IV. Ultrastructural studies on the central and peripheral nervous system. Eur. Neurol. 19:294, 1980. 98. Goebel, H. H., and Busch, H. F.: Abnormal lipopigments and lysosomal residual bodies in metachromatic leukodystrophy. In Porta, E. A. (ed.): Lipofuscin and Ceroid Pigments. New York, Plenum Press, p. 299, 1990. 99. Goebel, H. H., Harzer, K., Ernst, J. P., et al.: Late-onset globoid cell leukodystrophy: unusual ultrastructural pathology and subtotal ␤-galactocerebrosidase deficiency. J. Child. Neurol. 5:299, 1990. 100. Goebel, H. H., Kimura, S., Harzer, K., and Klein, H.: Ultrastructural pathology of eccrine sweat gland epithelial cells in globoid cell leukodystrophy. J. Child. Neurol. 8:171, 1993. 101. Goebel, H. H., Lenard, H. G., and Kohlschütter, A.: Morphologic and chemical biopsy findings in mucolipidosis IV. Clin. Neuropathol. 1:73, 1982. 102. Goebel, H. H., Lenard, H. G., and Kohlschütter, A.: Ultramorphology of peripheral nerves in lysosomal diseases. In Canal, N., and Pozza, G. (eds.): Peripheral Neuropathies. Amsterdam, Elsevier/North-Holland Biomedical Press, p. 427, 1978. 103. Goebel, H. H., Lenard, H. G., Kohlschütter, A., and Pilz, H.: The ultrastructure of the sural nerve in Pompe’s disease. Ann. Neurol. 2:111, 1977.

104. Goebel, H. H., Mole, S. E., and Lake, B. D. (eds.): The Neuronal Ceroid Lipofuscinoses (Batten Disease): Biomedical and Health Research. Amsterdam, IOS Press, 1999. 105. Goebel, H. H., Pilz, H., and Argyrakis, A.: Adult metachromatic leukodystrophy. II. Ultrastructural findings in peripheral nerve and skeletal muscle. Eur. Neurol. 15:308, 1977. 106. Goebel, H. H., Stolte, G., Kustermann-Kuhn, B., and Harzer, K.: B1 variant of GM2 gangliosidosis in a 12-year-old patient. Pediatr. Res. 25:89, 1989. 107. Goebel, H. H., Zeman, W., Patel, V. K., et al.: On the ultrastructural diversity and essence of residual bodies in neuronal ceroid-lipofuscinosis. Mech. Ageing Dev. 10:53, 1979. 108. Goebel, H. H., Zeman, W., and Pilz, H.: Ultrastructural investigations of peripheral nerves in neuronal ceroidlipofuscinoses (NCL). J. Neurol. 213:295, 1976. 109. Gordon, N., and Hudson, R. E. B.: Refsum’s syndrome— heredopathia atactica polyneuritiformis: a report of three cases, including a study of the cardiac pathology. Brain 82:41, 1959. 110. Graber, D., Salvayre, R., and Levade, T.: Accurate differentiation of neuropathic and non-neuronopathic forms of Niemann-Pick disease by evaluation of the effective residual lysosomal sphingomyelinase activity in intact cells. J. Neurochem. 63:1060, 1994. 111. Gravel, R. A., Kaback, M. M., Proia, R. L., et al.: The GM2 gangliosidoses. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3827, 2001. 112. Greenfield, J. G.: A form of progressive cerebral sclerosis in infants associated with primary degeneration of the interfascicular glia. J. Neurol. Psychopathol. 13:289, 1933. 113. Grégoire, A.: Ultrastructure des inclusions métachromatiques dans un cas de leucodystrophie. J. Microsc. 3:343, 1964. 114. Grégoire, A., Perier, O., and Dustin, P. J.: Metachromatic leucodystrophy, an electron microscopic study. J. Neuropathol. Exp. Neurol. 25:617, 1966. 115. Griffin, J. W., Goren, E., Schaumburg, H. H., et al.: Adrenomyeloneuropathy: a probable variant of adrenoleukodystrophy: I. Clinical and endocrinologic aspects. Neurology 27:1107, 1977. 116. Gumbinas, M., Larson, M., and Liu, H. M.: Peripheral neuropathy in classical Niemann-Pick disease: ultrastructure of nerves and skeletal muscles. Neurology 25:107, 1975. 117. Gustavson, K. H., and Hagberg, B.: The incidence and genetics of metachromatic leukodystrophy in Northern Sweden. Acta Paediatr. Scand. 60:585, 1971. 118. Hagberg, B.: The clinical diagnosis of Krabbe’s infantile leukodystrophy. Acta Paediatr. Scand. 52:213, 1963. 119. Hagberg, B.: Clinical symptoms, signs and tests in metachromatic leukodystrophy. In Folch-Pi, J., and Bauer, H. J. (eds.): Brain Lipids and Lipoproteins and the Leucodystrophies. Amsterdam, Elsevier, p. 134, 1963. 120. Hagberg, B., Kollberg, H., Sourander, P., and Åkesson, H. O.: Infantile globoid cell leukodystrophy (Krabbe’s disease): a clinical and genetic study of 32 Swedish cases 1953–1967. Neuropädiatrie 1:74, 1970.

Lysosomal and Peroxisomal Disorders 121. Hagberg, B., and Svennerholm, L.: Laboratory diagnostic tests in metachromatic leukodystrophy. Acta Paediatr. 48:632, 1959. 122. Hagberg, B., Svennerholm, L., and Wranne, B.: The excretion of urinary sulfatides in health and neurological diseases. Acta Paediatr. Scand. 54:408, 1965. 123. Hahn, A. F., Gilbert, J. J., Kwarciak, C., et al.: Nerve biopsy findings in Niemann-Pick type II (NPC). Acta Neuropathol. (Berl.) 87:149, 1994. 124. Hahn, A. F., Gordon, B. A., Feleki, V., et al.: A variant form of metachromatic leukodystrophy without arylsulfatase deficiency. Ann. Neurol. 12:33, 1982. 125. Hahn, A. F., Gordon, B. A., Gilbert, J. J., and Hinton, G. G.: The AB-variant of metachromatic leukodystrophy (postulated activator protein deficiency): light and electron microscopic findings in a sural nerve biopsy. Acta Neuropathol. (Berl.) 55:218, 1981. 126. Haltia, T., Palo, J., Haltia, M., and Icén, A.: Juvenile metachromatic leukodystrophy: clinical, biochemical and neuropathologic studies in nine new cases. Arch. Neurol. 37:42, 1980. 127. Heinzer, A. K., Watkins, P. A., Lu, J. F., et al.: A very longchain acyl-CoA synthase deficient mouse and its relevance to X-linked adrenoleukodystrophy. Hum. Mol. Genet. 12:1145, 2003. 128. Helmstaedt, E. R.: Über eine Beobachtung von metachromatischer Leukodystrophie. Dtsch. Zeitschr. Nervenheilkd. 184:21, 1963. 129. Herbert, M. A., and Clayton, P. T.: Phytanic acid alphaoxidase deficiency (Refsum disease) presenting in infancy. J. Inher. Metab. Dis. 17:211, 1994. 130. Hess, B., Saftig, P., Hartmann, D., et al.: Phenotype of arylsulfatase A-deficient mice: relationship to human metachromatic leukodystrophy. Proc. Natl. Acad. Sci. U. S. A. 93:1, 1996. 131. Hirschhorn, R., and Reuser, A. J. J.: Glycogen storage disease type II: acid ␣-glucosidase (acid maltase) deficiency. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3389, 2001. 132. Hogan, G. R., Gutmann, L., and Chou, S. M.: The peripheral neuropathy of Krabbe’s globoid leukodystrophy. Neurology 19:1094, 1969. 133. Hopwood, J. H., and Ballabio, A.: Multiple sulfatase deficiency and the nature of the sulfatase family. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3725, 2001. 134. Hund, E., Grau, A., Fogel, W., et al.: Progressive cerebellar ataxia, proximal neurogenic weakness and ocular motor disturbances: hexosaminidase A deficiency with late clinical onset in four siblings. J. Neurol. Sci. 145:25, 1997. 135. Ida, H., Rennert, O. M., Watabe, K., et al.: Pathological and biochemical studies of fetal Krabbe disease. Brain Dev. 16:480, 1994. 136. Igarashi, M., Schaumburg, H. H., Powers, J. M., et al.: Fatty acid abnormality in adrenoleukodystrophy. J. Neurochem. 26:851, 1976. 137. Igesu, H., Shimomura, K., Kishimoto, Y., and Suzuki, K.: Lipids of developing brain of twitcher mouse: an authentic murine model of Krabbe disease. Brain 106:405, 1983.

1877

138. Introne, W., Boissy, E. E., and Gahl, W. A.: Clinical, molecular, and cell biological aspects of Chediak-Higashi syndrome. Mol. Genet. Metab. 68:283, 1999. 139. Ishibashi, A., Tsuboi, R., and Shinmei, M.: ␤-Galactosidase and neuraminidase deficiency associated with angiokeratoma corporis diffusum. Arch. Dermatol. 120:1344, 1984. 140. Jacobi, M.: Über Leukodystrophie und PelizaeusMerzbacher’sche Krankheit. Virchows Arch. [A] 314:460, 1947. 141. Jansen, G. A., Ofman, R., Ferdinandusse, S., et al.: Refsum disease is caused by mutations in the phytanoyl-CoA hydrolase gene. Nat. Genet. 17:190, 1997. 142. Jardim, L. B., Giugliani, R., Pires, R. F., et al.: Protracted course of Krabbe disease in an adult patient bearing a novel mutation. Arch. Neurol. 56:1014, 1999. 143. Jatzkewitz, H.: Zwei Typen von CerebrosidSchwefelsäureestern als sog. “Prälipoide” und Speichersubstanzen bei der Leukodystrophie, Typ Scholz (metachromatische Form der diffusen Sklerose). HoppeSeyler Z. Physiol. Chem. 311:279, 1958. 144. Jatzkewitz, H., and Mehl, E.: Cerebroside-sulphatase and arylsulphatase A deficiency in metachromatic leucodystrophy (ML). J. Neurochem. 16:19, 1969. 145. Jerfy, A., and Roy, A. B.: The sulphatase of ox liver: XVI. A comparison of the arylsulfatase and cerebroside sulphatase activities of sulphatase A. Biochim. Biophys. Acta 293:178, 1973. 146. Joosten, E., Gabreëls, A., Stadhouders, A., et al.: Involvement of sural nerve in neuronal ceroid-lipofuscinoses: report of two cases. Neuropädiatrie 4:98, 1973. 147. Joosten, E., Hoes, M., Gabreëls, A., et al.: Electron microscopic investigation of inclusion material in a case of adult metachromatic leukodystrophy: observations on kidney biopsy, peripheral nerve and cerebral white matter. Acta Neuropathol. (Berl.) 33:165, 1975. 148. Kahana, D., Berant, M., and Wolman, M.: Primary familial xanthomatosis with adrenal involvement (Wolman’s disease): report of a further case with nervous system involvement and pathogenetic considerations. Pediatrics 42:70, 1968. 149. Kahlke, W., and Richterich, R.: Refsum’s disease (heredopathia atactica polyneuritiformis): an inborn error of the lipid metabolism with storage of 3,7,11,15-tetramethyl hexadecanoic acid: II. Isolation and identification of the storage product. Am. J. Med. 39:237, 1965. 150. Kamoshita, S.: Studies on metachromatic leucodystrophy. 2. Urine sediment examination for the diagnosis in lifetime. Acta Paediatr. Jpn. 31:16, 1964. 151. Kamoshita, S., and Landing, B. H.: Distribution of lesions in myenteric plexus and gastrointestinal mucosa in lipidoses and other neurological disorders of children. Am. J. Clin. Pathol. 49:312, 1968. 152. Karpati, G., Carpenter, S., Eisen, A. A., et al.: Multiple peripheral nerve entrapments: an unusual phenotypic variant of the Hunter syndrome (mucopolysaccharidosis II) in a family. Arch. Neurol. 31:418, 1974. 153. Karten, B., Vance, D. E., Campenot, R. B., and Vance, J. E.: Cholesterol accumulates in cell bodies, but is decreased in distal axons, of Niemann-Pick C1-deficient neurons. J. Neurochem. 83:1154, 2002.

1878

Diseases of the Peripheral Nervous System

154. Karten, B., Vance, D. E., Campenot, R. B., and Vance, J. E.: Trafficking of cholesterol from cell bodies to distal axons in Niemann Pick C1-deficient neurons. J. Biol. Chem. 278:4168, 2003. 155. Kemp, S., Pujol, A., Waterham, H. R., et al.: ABCD1 mutations and the X-linked adrenoleukodystrophy mutation database: role in diagnosis and clinical correlations. Hum. Mutat. 18:499, 2001. 156. King, R. A., Hearing, V. J., Creel, D. J., and Oetting, W. S.: Albinism. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 4. New York, McGraw-Hill, p. 5587, 2001. 157. Kishimoto, Y., Moser, H. W., Kawamura, N., et al.: Adrenoleukodystrophy: evidence that abnormal very long chain fatty acids of brain cholesterol esters are of exogenous origin. Biochem. Biophys. Res. Commun. 96:69, 1980. 158. Kjellson, L.: Refsums sjukdom. Nord Med 9:460, 1953. 159. Kobayashi, T., Ohta, M., Goto, I., et al.: Adult type mucolipidosis with ␤-galactosidase and sialidase deficiency. J. Neurol. 221:137, 1979. 160. Kobayashi, T., Yamanaka, T., Jacobs, J. M., et al.: The twitcher mouse: an enzymatically authentic model of human globoid cell leukodystrophy (Krabbe disease). Brain Res. 202:479, 1980. 161. Kohlschütter, A., Sieg, K., Schulte, F. J., et al.: Infantile cardiomyopathy and neuromyopathy with ␤-galactosidase deficiency. Eur. J. Pediatr. 139:75, 1982. 162. Kolodny, E. H., Raghavan, S. S., and Krivit, W.: Late-onset Krabbe disease (globoid cell leukodystrophy): clinical and biochemical features of 15 cases. Dev. Neurosci. 13:232, 1991. 163. Korn-Lubetzki, I., Dor-Wollman, T., Soffer, D., et al.: Early peripheral nervous system manifestations of infantile Krabbe disease. Pediatr. Neurol. 28:115, 2003. 164. Kozutsumi, Y., Kanazawa, T., Sun, Y., et al.: Sphingolipids involved in the induction of multinuclear cell formation. Biochim. Biophys. Acta 1582:138, 2002. 165. Krabbe, K.: A new familial infantile form of diffuse brain sclerosis. Brain 39:74, 1916. 166. Kraus-Ruppert, R., and Sommer, H.: The late form of metachromatic leukodystrophy II: ultrastructural correlations with morphological and neurological findings. J. Neurol. Sci. 17:383, 1972. 167. Krendel, D. A., Shutter, L. A., and Holt, P. J.: Fiber type disproportion in metachromatic leukodystrophy. Muscle Nerve 17:1352, 1994. 168. Kristensson, K., Olsson, Y., and Sourander, P.: Peripheral nerve change in Tay-Sachs and Batten-Spielmeyer-Vogt disease. Acta Pathol. Microbiol. Scand. [A] 70:630, 1967. 169. Kritzler, R. A., Terner, J. Y., Lindenbaum, J., et al.: ChédiakHigashi syndrome: cytologic and serum lipid observations in a case and family. Am. J. Med. 36:583, 1964. 170. Krivit, W., Aubourg, P., Shapiro, E. G., and Peters, C.: Bone marrow transplantation for globoid cell leukodystrophy, adrenoleukodystrophy, metachromatic leukodystrophy, and Hurler syndrome. Curr. Opin. Hematol. 6:377, 1999. 171. Krivit, W., Peters, C., and Shapiro, E. G.: Bone marrow transplantation as effective treatment of central nervous system disease in globoid cell leukodystrophy, metachromatic

172.

173.

174. 175.

176.

177.

178.

179.

180.

181. 182.

183.

184.

185.

186.

187.

leukodystrophy, adrenoleukodystrophy, mannosidosis, fucosidosis, aspartylglucosaminuria, Hurler, MaroteauxLamy, and Sly syndromes, and Gaucher disease type III. Curr. Opin. Neurol. 12:167, 1999. Kuemmel, T. A., Schroeder, R., and Stoffel, W.: Light and electron microscopic analysis of the central and peripheral nervous systems of acid sphingomyelinase-deficient mice resulting from gene targeting. J. Neuropathol. Exp. Neurol. 56:171, 1997. Kurtz, H. J., and Fletcher, T. F.: The peripheral neuropathy of canine globoid-cell leukodystrophy (Krabbe type). Acta Neuropathol. (Berl.) 16:226, 1970. Lake, B. D.: Segmental demyelination of peripheral nerves in Krabbe’s disease. Nature 217:171, 1968. Landrieu, P., and Said, G.: Peripheral neuropathy in type A Niemann-Pick disease. Acta Neuropathol. (Berl.) 63:66, 1984. Le Coz, P., Assouline, E., Vanier, M. T., et al.: GM2gangliosidosis variant B1 disclosed during adolescence by an isolated multi-system involvement of the central and peripheral nervous systems [in French]. Rev. Neurol. (Paris) 150:61, 1994. Leitch, G. J., Horrocks, L. A., and Samorajski, T.: Effects of cations on isolated bovine optic nerve myelin. J. Neurochem. 16:1347, 1969. Leroy, J. G., Van Elsen, A., and Martin, J.-J.: Infantile metachromatic leukodystrophy: confirmation of a prenatal diagnosis. N. Engl. J. Med. 288:1365, 1973. Lieberman, J. S., Oshtory, M., Taylor, R. G., and Dreyfus, P. M.: Perinatal neuropathy as an early manifestation of Krabbe’s disease. Arch. Neurol. 37:446, 1980. Liu, H. M.: Ultrastructure of globoid leukodystrophy (Krabbe’s disease) with reference to the origin of globoid cells. J. Neuropathol. Exp. Neurol. 29:441, 1970. Lockman, L. A., Kennedy, W. R., and White, J. G.: The Chédiak-Higashi syndrome. J. Pediatr. 70:942, 1967. Lohse, P., Maas, S., Lohse, P., et al.: Molecular defects underlying Wolman disease appear to be more heterogeneous than those resulting in cholesteryl ester storage disease. J. Lipid Res. 40:221, 1999. Lohse, P., Maas, S., Lohse, P., et al.: Compound heterozygosity for a Wolman mutation is frequent among patients with cholesteryl ester storage disease. J. Lipid Res. 41:23, 2000. Luitjen, J. A. F. M., Straks, W., Blikkendaal-Liefinck, L. F., et al.: Metachromatic leukodystrophy—a comparative study of the ultrastructural findings in the peripheral nervous system in three cases, one of the late infantile, one of the juvenile and one of the adult form of the disease. Neuropädiatrie 9:338, 1978. Lyon, G., and Evrard, P.: Sur la présence d’inclusions cristallines dans les cellules de Schwann dans diverse neuropathies périphériques. C. R. Acad. Sci. Paris 271:1000, 1970. Lyon, G., Jardin, L., and Aicardi, J.: Etude au microscope électronique d’un nerf périphérique dans un cas de leucodystrophie de Krabbe. J. Neurol. Sci. 12:263, 1971. Marcaud, V., Defontaines, B., Jung, P., and Degos, C. F.: Refsum disease: evolution 35 years after diagnosis [in French]. Rev. Neurol. (Paris) 158:225, 2002.

Lysosomal and Peroxisomal Disorders 188. Marjanovic, B., Cvetkovic, D., Dozic, S., et al.: Association of Krabbe leukodystrophy and congenital fiber type disproportion. Pediatr. Neurol. 15:79, 1996. 189. Marks, H. G., Scavina, M. T., Kolodny, E. H., et al.: Krabbe’s disease presenting as a peripheral neuropathy. Muscle Nerve 20:1024, 1997. 190. Marsden, D. D., Obeso, J. S., and Lang, A. E.: Adrenoleukomyeloneuropathy presenting as spinocerebellar degeneration. Neurology 32:1031, 1982. 191. Martin, J.-J., Ceuterick, C., Mercelis, R., and Joris, C.: Pathology of peripheral nerves in metachromatic leucodystrophy: a comparative study of ten cases. J. Neurol. Sci. 53:95, 1982. 192. Martin, J.-J., De Barsy, T., van Hoof, F., and Palladini, G.: Pompe’s disease: an inborn lysosomal disorder with storage of glycogen. Acta Neuropathol. (Berl.) 23:229, 1973. 193. Martin, J.-J., Leroy, J. G., Ceuterick, C., et al.: Fetal Krabbe leukodystrophy: a morphologic study of two cases. Acta Neuropathol. (Berl.) 53:87, 1981. 194. Martin, J.-J., Leroy, J. G., Farriaux, J. R., et al.: I-cell disease (mucolipidosis II): a report on its pathology. Acta Neuropathol. (Berl.) 33:285, 1975. 195. Martin, J.-J., Lowenthal, A., Ceuterick, C., and Gacoms, H.: Adrenomyeloneuropathy: a report of two families. J. Inher. Metab. Dis. 10:339, 1982. 196. Matsumoto, R., Oka, N., Nagahama, Y., et al.: Peripheral neuropathy in late-onset Krabbe disease: histochemical and ultrastructural findings. Acta Neuropathol. (Berl.) 92:635, 1996. 197. Mazière, C., Mazière, J. C., Mora, L., et al.: Alterations in cholesterol metabolism in cultured fibroblasts from patients with Niemann-Pick disease type C. J. Inher. Metab. Dis. 10:339, 1987. 198. McAlarney, T., Pastores, G. M., Hays, A. P., and Latov, N.: Antisulfatide antibody and neuropathy in a patient with Gaucher’s disease. Neurology 45:1622, 1995. 199. McGrath, J., Schutta, H. S., Yaseen, A., and Steinberg, A.: A morphologic and biochemical study of canine globoid leukodystrophy. J. Neuropathol. Exp. Neurol. 28:171, 1969. 200. McKusick, V. A., Neufeld, E. F., and Kelly, T. E.: The mucopolysaccharide storage disease. In Stanbury, J. B., Wyngaarden, J. B., and Frederickson, D. S. (eds.): The Metabolic Basis of Inherited Disease, 4th ed. New York, McGraw-Hill, p. 1282, 1978. 201. Mehl, E., and Jatzkewitz, H.: Evidence for the genetic block in metachromatic leukodystrophy. Biochem. Biophys. Res. Commun. 19:407, 1965. 202. Melchior, J. C., and Clausen, J.: Metachromatic leucodystrophy in early childhood: treatment with a diet deficient in vitamin A. Acta Paediatr. Scand. 57:2, 1968. 203. Menkes, J. H., and Corbo, L. M.: Adrenoleukodystrophy: accumulation of cholesterol esters with very long chain fatty acids. Neurology 27:928, 1977. 204. Meyers, J. P., Sung, J. H., Cowen, D., and Wolf, A.: Pathological changes in the central and peripheral nervous system in Chédiak-Higashi disease. J. Neuropathol. Exp. Neurol. 22:357, 1963. 205. Misra, V. P., King, R. H. M., Harding, A. E., et al.: Peripheral neuropathy in Chédiak-Higashi syndrome. Acta Neuropathol. (Berl.) 81:354, 1991.

1879

206. Miyatake, T., Atsumi, T., Obayashi, T., et al.: Adult type neuronal storage disease with neuraminidase deficiency. Ann. Neurol. 6:232, 1979. 207. Modigliani, R., Lemann, M., Melancon, S. B., et al.: Diarrhea and autonomic dysfunction in a patient with hexosaminidase B deficiency (Sandhoff disease). Gastroenterology 106:775, 1994. 208. Moser, H. W.: Ceramidase deficiency: Farber’s lipogranulomatosis. In Stanbury, J. B., Wyngaarden, J. B., and Fredrickson, D. S. (eds.): The Metabolic Basis of Inherited Disease, 4th ed. New York, McGraw-Hill, p. 707, 1978. 209. Moser, H. W., Linke, T., Fensom, A. H., et al.: Acid ceramidase deficiency: Farber lipogranulomatosis. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3573, 2001. 210. Moser, H. W., Moser, A. B., Frayer, K. K., et al.: Adrenoleukodystrophy: increased plasma content of very long chain fatty acids. Neurology 31:1241, 1981. 211. Moser, H. W., Moser, A. B., Krawamura, M., et al.: Adrenoleukodystrophy: elevated C26 fatty acids in cultured skin fibroblasts. Ann. Neurol. 7:542, 1980. 212. Moser, H. W., Moser, A. B., and McKhann, G. M.: The dynamics of a lipidosis: turnover of sulfatide, steroid sulfate and polysaccharide sulfate in metachromatic leukodystrophy. Arch. Neurol. 17:494, 1967. 213. Moser, H. W., Naidu, S., Kumar, A. J., and Rosenblum, A. E.: The adrenoleukodystrophies. C. R. C. Crit. Rev. Neurobiol. 3:29, 1987. 214. Moser, H. W., Smith, K. D., Watkins, P. A., et al.: X-linked adrenoleukodystrophy. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 2. New York, McGraw-Hill, p. 3257, 2001. 215. Mossakowski, M., Mathieson, G., and Cumings, J. N.: On the relationship of metachromatic leukodystrophy and amaurotic idiocy. Brain 84:585, 1961. 216. Müller, D., Pilz, J., and ter Meulen, V.: Studies on adult metachromatic leukodystrophy: 1. Clinical, morphological and histochemical observations on two cases. J. Neurol. Sci. 9:567, 1969. 217. Murphy, J. V., Wolfe, H. J., Balazs, A. E., and Moser, H.: A patient with deficiency of arylsulfatases A, B, C and steroid sulfatase, associated with storage of sulfatide, cholesterol sulfate and glycosamino-glycans. In Bernsohn, J., and Grossman, H. J. (eds.): Lipid Storage Diseases. New York, Academic Press, p. 67, 1971. 218. Neufeld, E. F., and Muenzer, J.: The mucopolysaccharidoses. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Diseases, Vol. 3. New York, McGraw-Hill, p. 3421, 2001. 219. Nevin, N. C., Cumings, J. M., and McKeown, F.: Refsum’s syndrome: heredopathia atactica polyneuritiformis. Brain 90:419, 1967. 220. Nevo, Y., Pestronk, A., Lopate, G., and Carroll, S. L.: Neuropathy of metachromatic leukodystrophy: improvement with immunomodulation. Pediatr. Neurol. 15:237, 1996. 221. Nissl, F.: Enzyklopädie der mikroskopischen Technik. In Ehrlich, P., Krause, R., Mosse, M., et al. (eds.):

1880

222.

223.

224. 225. 226. 227.

228.

229.

230.

231.

232.

233.

234. 235.

236. 237.

238.

239.

240.

Diseases of the Peripheral Nervous System

Nervensystem, 2nd ed. Berlin, Urban and Schwarzenberg, p. 284, 1910. Norman, R. M.: Diffuse progressive metachromatic leucoencephalopathy: a form of Schilder’s disease related to the lipidoses. Brain 22:234, 1947. Norman, R. M., Urich, H., and Tingey, A. H.: Metachromatic leucoencephalopathy: a form of lipidosis. Brain 83:369, 1960. O’Brien, J. S.: A molecular defect of myelination. Biochem. Biophys. Res. Commun. 15:484, 1964. O’Brien, J. S.: Stability of the myelin membrane. Science 147:1099, 1965. O’Brien, J. S., and Sampson, E. L.: Myelin membrane: a molecular abnormality. Science 150:1613, 1965. O’Brien, J. S., and Warner, T. G.: Sialidosis: delineation of subtypes by neuraminidase assay. Clin. Genet. 17:35, 1980. Olsson, R., Sourander, P., and Svennerholm, L.: Experimental studies on the pathogenesis of leucodystrophies: effect of intracerebrally injected sphingolipids in rat brain. Acta Neuropathol. (Berl.) 6:153, 1966. O’Neill, B. P., Moser, H. W., and Marmion, L. C.: Adrenoleukodystrophy: elevated C26 fatty acid in cultured skin fibroblasts and correlation with disease expression in three generations of a kindred. Neurology 32:540, 1982. O’Neill, B. P., Moser, H. W., and Saxena, K. M.: Familial Xlinked Addison disease as an expression of adrenoleukodystrophy (ALD): elevated C26 fatty acid in cultured skin fibroblasts. Neurology 32:543, 1982. O’Neill, G. N., Aoki, M., and Brown, R. H. J.: ABCD1 translation-initiator mutation demonstrates genotypephenotype correlation for AMN. Neurology 57:1956, 2001. Park, W. D., O’Brien, J. F., Lundquist, P. A., et al.: Identification of 58 novel mutations in Niemann-Pick disease type C: correlation with biochemical phenotype and importance of PTC1-like domains in NPC1. Hum. Mutat. 22:313, 2003. Patel, V., Goebel, H. H., Watanabe, I., and Zeman, W.: Studies on GM1-gangliosidosis, type II. Acta Neuropathol. (Berl.) 30:155, 1974. Patrick, A., and Lake, B. D.: Deficiency of an acid lipase in Wolman’s disease. Nature 222:1067, 1969. Pecker, J., Feuvrier, Y. M., and le Huerou, Y.: Maladie de Refsum: diagnostic avec certaines polyradiculonévrites à réchutes. Rev. Neurol. (Paris) 109:39, 1963. Percy, K. A.: Krabbe continuum or clinical conundrum? Neurology 49:1203, 1997. Percy, K. A., Kaback, M. M., and Herndon, R. M.: Metachromatic leukodystrophy: comparison of early and late-onset forms. Neurology 27:933, 1977. Peters, C., and Steward, C. G., for the NMDP I. and the Working Party on Inborn Errors of the EBMT: Hematopoietic cell transplantation for inherited metabolic disease: an overview of outcomes and practice guidelines. Bone Marrow Transplant. 31:229, 2003. Pezeshkpour, G., Kurent, J. S., Krarup, C., et al.: Peripheral neuropathy in Chédiak-Higashi syndrome. J. Neuropathol. Exp. Neurol. 45:353, 1986. Pilz, H.: Die Sphingolipidveränderungen bei der Leukodystrophie Typ Krabbe im Vergleich zum akuten und

241. 242.

243.

244.

245.

246.

247.

248.

249.

250. 251.

252.

253. 254.

255.

256.

257.

chronischen sudanophilen Markzerfall. Acta Neuropathol. (Berl.) 4:16, 1964. Pilz, H.: Late adult metachromatic leukodystrophy. Arch. Neurol. 27:87, 1972. Pilz, H., and Hopf, H. C.: A preclinical case of late adult metachromatic leukodystrophy? J. Neurol. Neurosurg. Psychiatry 33:360, 1972. Pilz, H., von Figura, K., and Goebel, H. H.: Deficiency of arylsulfatase B in 2 brothers aged 40 and 38 years (Maroteaux-Lamy syndrome, type B). Ann. Neurol. 6:315, 1979. Poulos, A., Pollard, A., Mitchell, J. D., et al.: Patterns of Refsum’s disease: phytanic acid oxidase deficiency. Arch. Dis. Child. 59:222, 1984. Powers, J. M., DeCiero, D. P., Cox, C., et al.: The dorsal root ganglia in adrenomyeloneuropathy: neuron atrophy and abnormal mitochondria. J. Neuropathol. Exp. Neurol. 60:493, 2001. Powers, J. M., and Schaumburg, H. H.: Adrenoleukodystrophy: similar ultrastructural changes in adrenal cortical and Schwann cells. Arch. Neurol. 30:406, 1974. Probst, A., Ulrich, J., Heitz, P. U., and Herschkowitz, N. N.: Adrenomyeloneuropathy: a protracted pseudosystematic variant of adrenoleukodystrophy. Acta Neuropathol. (Berl.) 49:108, 1980. Rafi, M. A., Coppola, S., Liu, S. L., et al.: Disease-causing mutations in cis with the common arylsulfatase A pseudodeficiency allele compound—the difficulties in accurately identifying patients and carriers of metachromatic leukodystrophy. Mol. Genet. Metab. 79:83, 2003. Rampini, S., Isler, W., Baerlocher, K., et al.: Die Kombination von metachromatischer Leukodystrophie und Mukopolysaccharidose als selbständiges Krankheitsbild (Mukosulfatidose). Helv. Paediatr. Acta 25:436, 1970. Reese, H., and Bareta, J.: Heredopathia atactica polyneuritiformis. J. Neuropathol. Exp. Neurol. 9:385, 1950. Refsum, S.: Heredoataxia hemeralopica polyneuritiformis— et tidligere ikke beskrevet familiaert syndrom? Nord. Med. 28:2682, 1945. Refsum, S.: Heredopathia atactica polyneuritiformis: a familial syndrome not hitherto described. A contribution to the clinical study of the hereditary diseases of the nervous system. Acta Psychiatr. Scand. Suppl. 38:1, 1946. Refsum, S., Salomonsen, L., and Skatvedt, M.: Heredopathia atactica. J. Pediatr. 35:335, 1949. Résibois, A.: Electron-microscopic study of metachromatic leucodystrophy. III. Lysosomal nature of the inclusions. Acta Neuropathol. (Berl.) 13:149, 1969. Résibois-Grégoire, A.: Electron microscopic studies of metachromatic leucodystrophy: II. Compound nature of the inclusions. Acta Neuropathol. (Berl.) 9:244, 1967. Rivel, J., Vital, C., Battin, J., et al.: La lipogranulomatose disséminée de Farber: étude anatomo-clinique et ultrastructurale, de deux observations. Arch. Anat. Cytol. Pathol. 25:37, 1977. Salman, M. S., Clarke, J. T., Midroni, G., and Waxman, M. B.: Peripheral and autonomic nervous system involvement in chronic GM2 gangliosidosis. J. Inher. Metab. Dis. 24:65, 2001.

Lysosomal and Peroxisomal Disorders 258. Sandhoff, K., Kolter, T., and Harzer, K.: Sphingolipid activator proteins. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3371, 2001. 259. Schaumburg, H. H., Powers, J. M., Raine, C. S., et al.: Adrenomyeloneuropathy: a probable variant of adrenoleukodystrophy. II. General pathologic, neuropathologic and biochemical aspects. Neurology 27:1114, 1977. 260. Schaumburg, H. H., Richardson, E. P., Johnson, P. C., et al.: Schilder’s disease: sex-linked recessive transmission with specific adrenal changes. Arch. Neurol. 27:458, 1972. 261. Schnorf, H., Gitzelmann, R., Bosshard, N. U., et al.: Early and severe sensory loss in three adult siblings with hexosaminidase A and B deficiency (Sandhoff disease). J. Neurol. Neurosurg. Psychiatry 59:520, 1995. 262. Schott, B., Masson, R., Quincy, C., et al.: Étude clinique et biologique d’une famille de maladie de Refsum-Thiébaut. Rev. Neurol. (Paris) 118:230, 1968. 263. Schröder, J. M., and Sommer, C.: Mitochondrial abnormalities in human sural nerves: fine structural evaluation of cases with mitochondrial myopathy, hereditary and nonhereditary neuropathies, and a review of the literature. Acta Neuropathol. (Berl.) 82:471, 1991. 264. Schuchmann, E. H., and Desnick, R. J.: Niemann-Pick disease types A and B: acid sphingomyelinase deficiencies. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3589, 2001. 265. Schulte-Mattler, W. J., Lindner, A., and Zierz, S.: Autonomic neuropathy in a patient with adrenomyeloneuropathy. Eur. J. Med. Res. 1:559, 1996. 266. Schutta, H. S., Pratt, R. T. C., Metz, H., et al.: A familial study of the late infantile and juvenile forms of metachromatic leucodystrophy. J. Med. Genet. 3:86, 1966. 267. Shapiro, L. J., Aleck, K., Kaback, M. M., et al.: Metachromatic leukodystrophy without arylsulfatase A deficiency. Pediatr. Res. 13:1179, 1979. 268. Skjeldal, O. H., Stokke, O., and Refsum, S.: Phytanic acid storage disease. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed., Vol. 2. Philadelphia, W. B. Saunders, p. 1149, 1993. 269. Smith, J., Zellweger, H., and Afifi, A. K.: Muscular form of glycogenosis type II (Pompe). Neurology 17:537, 1967. 270. Solders, G., Celsing, G., Hagenfeldt, L., et al.: Improved peripheral nerve conduction, EEG and verbal IQ after bone marrow transplantation for adult metachromatic leukodystrophy. Bone Marrow Transplant. 22:1119, 1998. 271. Sourander, P., and Olsson, Y.: Peripheral neuropathy in globoid cell leucodystrophy (Morbus Krabbe). Acta Neuropathol. (Berl.) 11:69, 1968. 272. Steinman, L., Tharp, B. R., Dorfman, L. J., et al.: Peripheral neuropathy in the cherry-red spot myoclonus syndrome (sialidosis type I). Ann. Neurol. 7:450, 1980. 273. Stevens, L. R., Fluharty, A. L., Kihara, H., et al.: Metachromatic leukodystrophy: a variant with apparent cerebroside sulfatase activator deficiency. Clin. Res. 27:104A, 1979. 274. Stinshof, K., and Jatzkewitz, H.: Comparison of the cerebroside sulphatase and arylsulfatase activity of human

275.

276.

277.

278. 279.

280.

281.

282.

283.

284.

285.

286.

287.

288.

289. 290.

1881

sulphatase A in the absence of activators. Biochim. Biophys. Acta 377:126, 1975. Stumpf, A. D., and Austin, H. J.: Qualitative and quantitative differences in sulfatase A in different forms of classical metachromatic leukodystrophy. In Bernsohn, J., and Grossman, H. J. (eds.): Lipid Storage Disease. New York, Academic Press, p. 203, 1971. Sugita, M., Dulaney, J. T., and Moser, H. W.: Ceramidase deficiency in Farber’s disease (lipogranulomatosis). Science 78:1100, 1972. Sung, J. H., Meyers, J. P., Stadlan, E. M., et al.: Neuropathological changes in Chédiak-Higashi disease. J. Neuropathol. Exp. Neurol. 28:86, 1969. Suzuki, K., Proia, R. L., and Suzuki, K.: Mouse models of human lysosomal diseases. Brain Pathol. 8:195, 1998. Suzuki, K., Suzuki, K., and Chen, G. C.: Isolation and chemical characterization of metachromatic granules from a brain with metachromatic leukodystrophy. J. Neuropathol. Exp. Neurol. 26:537, 1967. Suzuki, K., and Suzuki, Y.: Galactosylceramide lipidosis: globoid cell leukodystrophy (Krabbe disease). In Stanbury, J. B., Wyngaarden, J. B., and Frederickson, D. S. (eds.): The Metabolic Basis of Inherited Disease, 4th ed. New York, McGraw-Hill, p. 747, 1978. Suzuki, Y., Oshima, A., and Nanba, E.: ␤-Galactosidase deficiency (␤-galactosidosis): GM1 gangliosidosis and Morquio B disease. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3775, 2001. Takano, H., Koike, R., Onodera, O., and Tsuji, S.: Mutational analysis of X-linked adrenoleukodystrophy gene. Cell Biochem. Biophys. 32:177, 2000. Tan, E. K., Lim, S. H., Chan, L. L., et al.: X-linked adrenoleukodystrophy: spinocerebellar variant. Clin. Neurol. Neurosurg. 101:137, 1999. Thieffry, S., and Lyon, G.: Diagnostic d’un cas de leucodystrophie métachromatique (type Scholz) par la biopsie d’un nerf périphérique. Rev. Neurol. (Paris) 100:452, 1959. Thieffry, S., Lyon, G., and Maroteaux, P.: Leucodystrophie métachromatique (sulfatidose) et mucopolysaccharidose associées chez un même malade. Rev. Neurol. (Paris) 114:193, 1966. Thomas, P. K.: The ultrastructural pathology of unmyelinated nerve fibres. In Desmedt, J. E. (ed.): New Developments in Electromyography and Clinical Neurophysiology, Vol. 2. Basel, S. Karger, p. 227, 1973. Thomas, P. K.: Central and peripheral demyelination. In Serratrice, G., Pellissier, J.-F., Desnuell, D., and Pouget, J. (eds.): Myélopathies, Neuropathies et Myopathies. Paris, Expansion Scientifique Française, p. 141, 1989. Thomas, P. K., and Harding, A. E.: Peripheral neuropathies. In Emery, A. E. H., and Rimoin, D. L. (eds.): The Principles and Practice of Medical Genetics, 2nd ed. Edinburgh, Churchill Livingstone, p. 413, 1990. Thomas, P. K., King, R. H. M., Chiang, T. R., et al.: Neurofibromatous neuropathy. Muscle Nerve 13:93, 1990. Thomas, P. K., King, R. H. M., Kocen, R. S., and Brett, E. M.: Comparative ultrastructural observations on peripheral nerve

1882

291. 292.

293.

294.

295.

296.

297.

298.

299.

300.

301.

302. 303. 304.

305.

306.

307.

Diseases of the Peripheral Nervous System

abnormalities in the late infantile, juvenile and late onset forms of metachromatic leukodystrophy. Acta Neuropathol. (Berl.) 39:237, 1977. Thomas, P. K., and Lascelles, R. G.: Hypertrophic neuropathy. Q. J. Med. 36:223, 1967. Thomas, P. K., Young, E. P., and King, R. H. M.: Sandhoff disease mimicking adult-onset bulbospinal neuropathy. J. Neurol. Neurosurg. Psychiatry 52:1103, 1989. Toga, M., Bernard-Badier, M., Pinsard, N., et al.: Étude clinique, histologique et ultrastructurale de quatre cas de leucodystrophie métachromatique infantile et juvénile. Acta Neuropathol. (Berl.) 21:23, 1972. Torvik, A., Torp, S., Kase, B. F., et al.: Infantile Refsum’s disease: a generalized peroxisomal disorder. Case report with postmortem examination. J. Neurol. Sci. 85:39, 1988. Tsuji, S., Ohno, T., Miyatake, T., et al.: Fatty acid elongation activity in fibroblasts from patients with adrenoleukodystrophy (ALD). J. Biochem. (Tokyo) 96:1241, 1984. Turazzini, M., Beltramello, A., Bassi, R., et al.: Adult onset Krabbe’s leukodystrophy: a report of 2 cases. Acta Neurol. Scand. 96:413, 1997. van Bogaert, L., van Mechelen, P., Martin, J.-J., and Guazzi, G. C.: Sur la neuropathologie de la maladie de Refsum-Thiébaut (protocole de l’observation de Richterich, Kahlke, van Mechelen et Ross, 1963). Rev. Neurol. (Paris) 116:611, 1990. van den Brink, D. M., Brites, P., Haasjes, J., et al.: Identification of PEX7 as the second gene involved in Refsum disease. Am. J. Med. Genet. 72:471, 2003. Van der Hagen, C. B., Børreson, A. L., Molne, K., et al.: Metachromatic leukodystrophy I: Prenatal detection of arylsulphatase A deficiency. Clin. Genet. 4:256, 1973. van Geel, B. M., Assies, J., Weverling, G. J., and Barth, P. G.: Predominance of the adrenomyeloneuropathy phenotype of X-linked adrenoleukodystrophy in The Netherlands: a survey of 30 kindreds. Neurology 44:2343, 1994. van Geel, B. M., Koelman, J. H., Barth, P. G., and Ongerboer de Visser, B. W.: Peripheral nerve abnormalities in adrenomyeloneuropathy: clinical and electrodiagnostic study. Neurology 46:112, 1996. Vanier, M. T., and Millat, G.: Niemann-Pick disease type C. Clin Genet 64:269, 2003. Vasconcellos, E., and Smith, M.: MRI nerve root enhancement in Krabbe disease. Pediatr. Neurol. 19:151, 1998. Veltema, A. N., and Verjaal, A.: Sur un cas d’ hérédopathie ataxique polynévritique: maladie de Refsum. Rev. Neurol. (Paris) 104:15, 1961. Vital, C., Battin, J., Rivel, J., and Heunstre, J. P.: Aspects ultrastructuraux des lésions du nerf périphérique dans un cas de maladie de Farber. Rev. Neurol. (Paris) 132:419, 1976. von Figura, K., Gieselmann, V., and Jaeken, J.: Metachromatic leukodystrophy. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases of Inherited Disease, Vol. 3. New York, McGraw-Hill, p. 3695, 2001. Walter, S., and Goebel, H. H.: Ultrastructural pathology of dermal axons and Schwann cells in lysosomal diseases. Acta Neuropathol. (Berl.) 76:489, 1988.

308. Wanders, R. J., Jakobs, C., and Skjedal, O. H.: Refsum disease. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Bases Inherited Disease, Vol. 2. New York, McGraw-Hill, p. 3303, 2001. 309. Wanders, R. J., Jansen, G. A., and Lloyd, M. D.: Phytanic acid alpha-oxidation: new insights into an old problem. A review. Biochim. Biophys. Acta 1631:119, 2003. 310. Wanders, R. J. A., Van Roermund, C. W. T., Van Wijland, M. J. A., et al.: Peroxisomal very long chain fatty acid beta oxidation in human skin fibroblasts: activity in Zellweger syndrome and other peroxisomal disorders. Clin. Chim. Acta 166:255, 1987. 311. Ward, D. M., Shiflett, S. L., and Kaplan, J.-C.: ChédiakHigashi syndrome: a clinical and molecular view of a rare lysosomal storage disorder. Curr. Mol. Med. 2:469, 2002. 312. Webster, H. de F.: Schwann cell alteration in metachromatic leucodystrophy: preliminary phase and electron microscopic observations. J. Neuropathol. Exp. Neurol. 21:534, 1962. 313. Weis, G. M., Nelson, R. L., O’Neill, B. P., et al.: Use of adrenal biopsy in diagnosing adrenoleukomyeloneuropathy. Arch. Neurol. 37:634, 1980. 314. Weller, R. O.: An electron microscope study of hypertrophic neuropathy of Dejerine and Sottas. J. Neurol. Neurosurg. Psychiatry 30:111, 1967. 315. Weller, R. O., and Cervós-Navarro, J.: Pathology of Peripheral Nerves. London, Butterworth, 1977. 316. Wenger, D. A., Rafi, M. A., and Luzi, P.: Molecular genetics of Krabbe disease (globoid cell leukodystrophy): diagnostic and clinical implications. Hum. Mutat. 10:268, 1997. 317. Wenger, D. A., Rafi, M. A., Luzi, P., et al.: Krabbe disease: genetic aspects and progress toward therapy. Mol. Genet. Metab. 70:1, 2000. 318. Wierzbicki, A. S., Lloyd, M. D., Schofield, C. J., et al.: Refsum’s disease: a peroxisomal disorder affecting phytanic acid alpha-oxidation. J. Neurochem. 80:727, 2002. 319. Wierzbicki, A. S., Mitchell, J., Lambert-Hammill, M., et al.: Identification of genetic heterogeneity in Refsum’s disease. Eur. J. Hum. Genet. 8:649, 2000. 320. Wilken, B., Dechent, P., Brockmann, K., et al.: Quantitative proton magnetic resonance spectroscopy of children with adrenoleukodystrophy before and after hematopoietic stem cell transplantation. Neuropediatrics 34:237, 2003. 321. Wilson, J., Lake, B. D., and Dunn, H. G.: Krabbe’s leukodystrophy: some clinical, genetic and pathogenetic considerations. J. Neurol. Sci. 10:541, 1970. 322. Witte, F.: Über pathologische Abbauvorgänge im Zentralnervensystem. Münch. Med. Wochenschr. 68:69, 1921. 323. Yudell, A., Gomez, M. R., Lambert, E. H., and Dockerty, M. B.: The neuropathy of sulfatide lipidosis (metachromatic leukodystrophy). Neurology 17:103, 1967. 324. Yunis, E. J., and Lee, R. E.: The ultrastructure of globoid (Krabbe) leukodystrophy. Lab. Invest. 21:415, 1969. 325. Zhang, X.-L., Rafi, M. A., DeGala, G., and Wenger, D. A.: Insertion of the mRNA of a metachromatic leukodystrophy patient with sphingolipid activator protein-I deficiency. Proc. Natl. Acad. Sci. U. S. A. 87:1426, 1990.

80 Porphyric Neuropathy ANTHONY J. WINDEBANK AND HERBERT L. BONKOVSKY

Biosynthesis of Heme Effects of Drugs on Porphyrin Metabolism Other Factors Altering Porphyrin Metabolism Mechanism of Neuronal Damage Pathologic Changes in Porphyria

Findings from a Mouse Model of AIP Clinical Features of Porphyria Risk Factors and Inheritance Clinical Features of the Neurologic Disease Clinical Differential Diagnosis

Disorders of porphyrin metabolism are a rare cause of peripheral neuropathy. Porphyrins are intermediary metabolites in the synthesis of heme. Although all cells synthesize heme, quantitatively the most important tissues are the liver and bone marrow. Except for rare instances of severe protoporphyria after liver transplantation, porphyric neuropathy is associated with the acute or inducible hepatic porphyrias: acute intermittent porphyria (AIP), hereditary coproporphyria (HCP), and variegate porphyria (VP). Similar biochemical and clinical features may also occur in lead poisoning and hereditary tyrosinemia, type II. The final diagnosis of porphyric neuropathy depends upon demonstration of the appropriate metabolic abnormality. It is therefore important to understand the heme synthetic pathway.

BIOSYNTHESIS OF HEME The tetrapyrrole ring structure of porphyrin is shown in Figure 80–1. The tetrapyrrole intermediates in heme biosynthesis have substitutions on their side chains (carbons 2, 3, 7, 8, 12, 13, 17, and 18). For example, the side chains of protoporphyrin are four methyl groups, two vinyl groups, and two propionate groups. Iron is inserted into the protoporphyrin ring to form heme, which is an essential prosthetic group for many enzymes (e.g., cytochromes) and oxygen-carrying proteins (e.g., hemoglobin). These compounds are essential for all oxidative

Laboratory Studies Treatment Therapy for Acute Attacks Prognosis

metabolism, and a complete interruption of heme synthesis would be lethal. Sublethal defects occur that lead to abnormal accumulations of porphyrins and porphyrin precursors. The accumulation of metabolites is presumed to underlie the clinical manifestations of the diseases, and measurement of the metabolites or deficient enzymes is essential for diagnosis. The major steps in heme synthesis are set out in Figure 80–2. The first step involves succinyl coenzyme A and glycine combining to form ␦-aminolevulinic acid (ALA) in the mitochondrion. This reaction is catalyzed by ALA synthase and is usually the rate-limiting step in hepatic heme synthesis. ALA passes into the cytosol, and two molecules of ALA are condensed to form the monopyrrole porphobilinogen (PBG).41 Four PBG molecules are further condensed by PBG deaminase (also known as hydroxymethylbilane synthase or uroporphyrinogen I synthase) and uroporphyrinogen III synthase to form uroporphyrinogen III. The final cytoplasmic step is a stepwise decarboxylation of the four acetate side chains of uroporphyrinogen III to form coproporphyrinogen III. The enzyme responsible for this is uroporphyrinogen decarboxylase. Coproporphyrinogen III reenters the intermembrane space of the mitochondria and is converted by coproporphyrinogen oxidase to protoporphyrinogen IX. The final oxidation step to protoporphyrin IX is catalyzed by mitochondrial protoporphyrinogen oxidase. Insertion of divalent iron to form heme is catalyzed by ferrochelatase. Thus, for all steps of the pathway from the third to the last, 1883

1884

Diseases of the Peripheral Nervous System

FIGURE 80–1 The basic tetrapyrrole ring structure for porphyrin molecules. The number and letter designations are those recommended by the International Union of Pure and Applied Chemistry.

the actual intermediates are porphyrinogens, not porphyrins (Fig. 80–2). The porphyrinogens differ from the porphyrins because the pyrrole rings (A, B, C, and D in Fig. 80–1) are linked by fully saturated methylene bridges (no double bonds between carbons 5 and 6, 10 and 11, 14 and 15, and 19 and 20). The porphyrinogens are colorless and nonfluorescing. They may be rapidly oxidized nonenzymatically by contact with atmospheric oxygen to the intensely colored, fluorescing corresponding porphyrins. This explains why the urine of a patient excreting large amounts of porphyrinogen may change color upon standing at room temperature. Hepatic heme synthesis is controlled by a negative feedback loop. There is a small pool of free intracellular heme that regulates the rate of expression of the enzyme ALA synthase. Several sites of action of heme on expression of ALA synthase have been identified in experimental systems, including decreased rate of messenger RNA (mRNA) synthesis, increased rate of mRNA breakdown, and interference with the normal transfer of ALA synthase into mitochondria,4 the site at which ALA synthase normally acts. Intracellular content and activity of ALA synthase can vary 100-fold over a period of minutes. Because this is normally the rate-limiting step, there is a consequent similar rapid change in heme synthesis.4 Understanding of this feedback control loop is critical to understanding the biochemical abnormalities underlying the clinical manifestations of the acute porphyrias. Heme synthesis in erythropoietic tissue is not controlled in this way. Erythropoietic tissue contains a different isoform of ALA synthase the synthesis of which is not controlled by heme.45

FIGURE 80–2 The metabolic pathway of hepatic heme synthesis. Deficiencies of the outlined enzymes are responsible for the clinical types of porphyria. PBG deaminase (uroporphyrinogen I synthase), coproporphyrinogen oxidase, and protoporphyrinogen oxidase are usually deficient on a genetic basis and are associated with acute attacks of neurologic disease. Uroporphyrinogen decarboxylase deficiency may exist on a genetic basis or occur secondarily to toxic (usually alcohol-induced) liver disease. Porphyria cutanea tarda is not associated primarily with neurologic disease. The enzymes marked with an asterisk may be routinely assayed in erythrocytes in reference laboratories.

EFFECTS OF DRUGS ON PORPHYRIN METABOLISM The abnormalities in regulation of heme synthesis are best understood for AIP, and this condition is therefore used to illustrate the neuropathic porphyrias in general. AIP derives its name from the acute severe crises with intervening periods of normality. The disease was virtually unknown until the advent of modern pharmaceuticals. Virtually all drugs that induce the hepatic microsomal cytochrome P-450 system may induce attacks of AIP. The inherited enzyme defect in AIP is decreased activity of PBG deaminase (uroporphyrinogen I synthase). In most patients who are heterozygous for this deficiency, under basal conditions there is sufficient residual activity of PBG deaminase for normal or nearly normal rates of heme

Porphyric Neuropathy

synthesis to occur. When the cytochrome P-450 system is activated by a drug, there is increased incorporation of heme into cytochrome P-450 and a decrease in the regulatory pool of free heme. This in turn induces synthesis of ALA synthase by the mechanism discussed above. In addition, such drugs may increase expression of ALA synthase by direct interaction with promoter or enhancer elements of the gene. The underactive PBG deaminase then becomes the rate-limiting step in heme synthesis, and there is a rapid accumulation of the preceding substrates ALA and PBG. Some of these leak out of the hepatocytes, are taken up into other tissues, and are excreted in the urine.

OTHER FACTORS ALTERING PORPHYRIN METABOLISM Apart from drugs, there are two other major modifiers of heme synthesis: hormonal influences and glucose. ALA synthase is induced by several hormones, including progesterone, estrogen, and testosterone. This may explain why attacks of AIP rarely occur before puberty in men or women and are most likely to occur in women during the luteal phase of the menstrual cycle.26 Starvation also induces ALA synthase. The mechanism of this is not known but is thought to be a direct consequence of low intracellular glucose levels. This is the rationale for administration of glucose in the treatment of AIP.

MECHANISM OF NEURONAL DAMAGE In terms of morbidity and mortality, the neurologic manifestations of AIP, HCP, and VP are the most serious. Broadly, there are two types of abnormality. The first involves physiologic or pharmacologic changes producing psychiatric disturbances, seizures, disorders of consciousness, and autonomic changes. These are generally rapidly reversed when biochemical homeostasis returns. The mechanism of these changes is unknown. It has been suggested that PBG mimics the action of serotonin and acts as a false neurotransmitter. Other effects include inhibition of transmitter release at neuromuscular junctions, and interaction of ALA with receptors for ␥-aminobutyric acid (GABA). ALA is closely related to GABA structurally and is a partial agonist of GABA, the major inhibitory neurotransmitter in the vertebrate central nervous system (CNS). Thus some CNS dysfunction in acute porphyria, particularly the development of delirium and seizures, may be explained on the basis of ALA blocking normal GABA neurotransmission.5,48 The second type of abnormality involves structural neuronal damage, particularly for cells of the peripheral nervous system. The mechanism of production of these

1885

changes is not fully understood but may occur at several different levels. Because of the enzyme defect, there may be a deficiency of production of heme-containing proteins for oxygen transport, electron transport, and the cytochrome P-450 system. This in turn may lead to alterations in energy-dependent cell metabolism, particularly perhaps axonal transport, leading to distal axonal degeneration or cell death. Widespread disturbance of neuronal functions would be expected because of the high degree of dependence of neurons on oxidative energy metabolism. However, it is worth noting that the pathologic abnormalities of porphyria do not resemble those of hypoxia or hypoglycemia, which might induce a similar biochemical derangement. This mechanism would also require that neuronal heme synthesis be controlled in the same way as hepatic heme synthesis and that drugs causing porphyric symptoms would have the same effect on neurons as they have on hepatic metabolism. This is not the case. In the CNS, neuronal cytochrome P-450 and ALA synthase are not induced by exogenous drugs as are the liver enzymes.35,37 It has also been demonstrated that intravenous heme, which is used in the treatment of porphyria, does not cross the blood-brain barrier13,46 and, therefore, would not be able to substitute for intraneuronal heme unless there was disruption of the blood-brain barrier during attacks of AIP. Studies of heme penetration into the peripheral nervous system are not available, but blood-nerve barrier function is likely to be qualitatively similar to that of the blood-brain barrier. Heme administration is most effective when used early in the attack. If blood-brain barrier function is disrupted, it probably occurs late in the attack and only after neuronal structural damage has occurred. A second putative mechanism may involve the failure of the hepatic cytochrome P-450 system. Inability to detoxify a drug or metabolite may in turn lead to axonal damage. In general, however, drugs inducing AIP are not in themselves directly neurotoxic, and many of them have effects opposite those seen in AIP. Failure to detoxify barbiturates or diazepam would produce sedation and would have an anticonvulsant effect rather than producing the agitation and seizures seen in AIP. The final mechanism would require neurotoxicity of metabolites known to accumulate in attacks of AIP. ALA and PBG are produced and excreted in excess in all types of porphyric attacks associated with AIP, HCP, and VP. ALA and PBG, produced in excess in the hepatocyte, do leak into the blood and cross the blood-brain barrier. There is, however, very little direct evidence to demonstrate that either of these metabolites is directly neurotoxic in the acute porphyrias. Levels of these metabolites do interfere with neuronal function in experimental systems when present in concentrations of 10⫺2 to 10⫺4 M.16,27,30 However, measured levels of ALA and PBG (10⫺5 to 10⫺7 M) are much lower in the cerebrospinal fluid (CSF) of patients during attacks of AIP.36 CSF levels may not reflect intracellular levels,

1886

Diseases of the Peripheral Nervous System

particularly if those cells are unable to efficiently metabolize ALA or PBG because of the inherited defect in PBG deaminase activity. A detailed review of porphyrin metabolism and its cellular effects has recently been published.5

PATHOLOGIC CHANGES IN PORPHYRIA Interpretation of autopsy findings in patients with porphyria is complicated by the fact that patients have usually experienced repeated episodes over many years before death and, therefore, a mixture of acute and chronic changes is seen. Acute changes caused by porphyria and those caused by the terminal illness and respiratory failure must also be separated. Structural changes in cells of the CNS are sparse, and the most dramatic changes are in the peripheral nerves and supporting cells. The first probable autopsy case of porphyria was reported as acute polyneuritis following Sulfonal administration.20 Wallerian degeneration in the peripheral nervous system was described. There has been a prolonged controversy about whether the axon or the Schwann cell is the primary target for damage in porphyria.40 The most detailed pathologic11 and electrophysiologic1 studies strongly support the notion that the neuron or axon is the major target. The distribution of fiber degeneration in Cavanagh and Mellick’s studies suggested that the process involved dying back.11 In dyingback neuropathies caused by toxins, the longest axons are affected first; however, in acute porphyria, proximal muscles and cranial nerves may be affected very early in the illness, implying pathology at the root or perikaryal level. This is supported by the electrophysiologic findings.1 It is likely that the segmental demyelination2,44 and slowed nerve conduction34 seen in some cases reflect demyelination secondary to axonal changes resulting from repeated or chronic insults. Autonomic disturbances are a prevalent feature of attacks of AIP,21 but pathologic studies of the autonomic nervous system have been very limited. Hierons noted chromatolysis in celiac ganglion neurons.24 Detailed studies of autonomic involvement in AIP have not been carried out using the new tools available for studying the autonomic nervous system in patients.

FINDINGS FROM A MOUSE MODEL OF AIP Mice with targeted disruption of PBG deaminase have been developed.29 In these compound heterozygotes, one allele has a loss-of-function disruption and the other a partial disruption, such that residual enzymatic activity is 30% of normal. Although basal urinary excretions of ALA and PGB were normal in these mice, the mice developed marked increases in excretion during administration of phenobarbital. The mutant mice moved abnormally slowly,

were ataxic, and had shortened strides, impaired motor coordination, flat body posture, and impaired proprioception and nociception. In mutant mice older than 6 months, the sciatic nerves showed severe loss of axons and muscles showed atrophic bundles interspersed with hypertrophic bundles, findings typical of changes resulting from denervation.29 Such changes occurred in the absence of inducing drug treatment, suggesting that marked overproduction of ALA and/or PBG may not be essential for neuromuscular toxicity.33

CLINICAL FEATURES OF PORPHYRIA The primary disturbance of porphyrin metabolism may be most prominent in liver tissue (hepatic porphyria) or bone marrow (erythropoietic porphyria). However, only the hepatic porphyrias are associated with neurologic disease. The erythropoietic porphyrias cause photosensitivity and are not considered further here. The enzyme defects producing hepatic porphyria are illustrated in Figure 80–2. AIP is associated with acute attacks of severe neurologic dysfunction induced by drugs or hormonal changes. Skin changes do not occur. The metabolic defect involves decreased PBG deaminase. Therefore, tetrapyrroles (see Fig. 80–1) do not accumulate in the tissues of these patients. It is the presence of these ringed structures that cause photosensitivity. All porphyrins absorb light in the wavelength range of 400 nm (the Soret band). This absorbed energy activates resonating electrons in the conjugated double-bond system of the tetrapyrrole ring. This energy may be transferred to oxygen, which binds to the ring and produces oxygen free radicals and peroxides, which are cytotoxic. Similar neurologic manifestations occur in HCP and VP. In both, the inherited enzyme defect occurs after the condensation of the tetrapyrrole porphyrin ring, and therefore porphyrins may accumulate in the skin and cause photosensitivity. Porphyria cutanea tarda or symptomatic porphyria is associated with decreased activity of uroporphyrinogen decarboxylase (either genetic or acquired) and associated with iron overload and liver disease often related to chronic hepatitis and/or to ethanol abuse. These patients do not have acute attacks of neuropathy resulting from their porphyria.

Risk Factors and Inheritance The genetic defect underlying AIP is inherited as an autosomal dominant trait.21,25,26,43,47 Estimates of gene prevalence vary, but in the white population it is of the order of 1 in 10,000 to 1 in 100,000. The true incidence is difficult to assess because the disease is only recognized in individuals who suffer acute attacks of neurologic disease. Family members of affected individuals may have low levels of

1887

Porphyric Neuropathy

PBG deaminase and not suffer from clinically overt disease. The full explanation for this latent form of the disease is unknown, although several factors are known to play a role. As discussed above, acute attacks may be precipitated by drug ingestion (Table 80–1), hormonal changes, and nutritional deprivation. For the hormonal reasons discussed above, attacks rarely occur before puberty and are more frequent in women, particularly in the luteal phase of the menstrual cycle. HC was first clearly recognized as a distinct genetic entity in 1955.3 Although it was first described in a boy whose parents were first cousins, it is now recognized as

having an autosomal dominant pattern of inheritance.12,22 The gene probably occurs less frequently than that for AIP, although, because the disease is milder and more likely to occur in latent form,4 the true frequency is difficult to assess. It is not known to show any racial preference. Patients with this disorder present with either the skin manifestations or the acute attacks of neurologic disease, which are similar to, although usually less severe than, those occurring with AIP. Attacks are more common in women and are usually drug induced.10 VP occurs most frequently in South Africans of Dutch ancestry. The incidence has been estimated at 1 in 3000 in

Table 80–1. Drugs and Chemicals in Acute Hepatic Porphyrias* Reported to Exacerbate Disease

Theoretically Risky

Believed to Be Safe

Barbiturates Bemegride Chloramphenicol Chlordiazepoxide Chlorpropamide Chloroquine Danazol Diazepam Ergot preparations Estrogens Ethanol excess Eucalyptol (in mouthwash) Glutethimide Griseofulvin Halothane Hydantoins Imipramine

Alcuronium Allyl-containing compounds Bupivacaine Clonazepam (large doses) Clonidine Etidocaine Hydralazine Lidocaine Mepivacaine Methychlothiazide Parglyine Phenoxybenzamine Prilocaine Pyrocaine Spironolactone

Aspirin Atropine Bromides Calcium salts Chlorpromazine Chloral hydrate Corticosteroids D-Tubocurarine Cyclopropane Dicumarol Droperidol Ether Fentanyl Gallamine Guanethidine Mefanamic acid Meperidine

Ketamine Meprobamate Methyldopa Methylprylon Methsuccimide Nikethamide Oral contraceptives Pentazocine Phensuccimide Phenylbutazone Progestagens Pyrazinamide Pyrazolone derivatives Sulfonamides Theophylline derivatives Tolbutamide Troxidone Valproate

All agents known to induce cytochrome P-450 or to increase hepatic heme turnover

Methadone Morphine Neostigmine Nitrous oxide Oxylate/sodium Pancuronium Paraldehyde Penicillin Pentamethonium Procaine Promazine Promethazine Propoxyphene Propranolol Reserpine Succinylcholine Tetracycline Vitamins A, B, C, D, and E

*Among agents reported to exacerbate disease, those in italic type are incriminated most often. Those listed as theoretically risky have not been incriminated clearly in humans, but experimental studies indicate they have potential for damage. Those believed to be safe have been used in humans without ill effects, are not theoretically risky, and have not been reported to exacerbate porphyria.

1888

Diseases of the Peripheral Nervous System

this group. Most cases in South Africa have had their inheritance traced to a single Boer couple in the late 17th century.14,17 Macalpine and Hunter have reviewed evidence suggesting that this form of porphyria was responsible for the mental illness of King George III of England.31 The name “variegate porphyria” was coined because patients may have acute neurologic attacks only, skin manifestations only, a combination of both, or neither. The acute attacks are precipitated by the same factors as those in AIP and HCP.

Clinical Features of the Neurologic Disease As discussed above, the acute neurologic disease is indistinguishable among the different forms of porphyria. A similar syndrome occurs in hereditary tyrosinemia (type II), in which the fundamental genetic defect in fumaryl acetohydrolase leads to accumulation of 4,6-dioxoheptanoic acid, a potent inhibitor of ALA dehydratase. These findings strongly suggest that, if the neurologic manifestations are due to metabolite accumulation, the metabolite is ALA rather than PBG, because the latter compound does not accumulate with inhibition or deficiency of ALA dehydratase. A second possibility is that clinical disease is either a direct or indirect consequence of heme deficiency. Attacks are less severe in patients with HCP and VP than in those with AIP. Skin changes only occur in the former two diseases. In all types, attacks may be precipitated by drugs, hormonal changes, starvation, or other severe stress. During acute attacks, excess tetrapyrroles are excreted in the urine. The porphyrinogens are excreted colorless but are oxidized readily to the intensely colored corresponding porphyrin, causing the urine to turn red or purple. This is most apparent if the urine is exposed to light and air. In AIP, large excesses of porphyrins are not excreted. However, when high concentrations of PBG are present in urine, nonenzymatic condensation may occur and the same urine color changes may ensue. In addition, PBG may polymerize to form a nonporphyrin polymer, porphobilin, which has a black color. The common clinical pattern of symptom progression involves acute abdominal pain, followed by psychiatric disturbances, and then acute neuropathy. Abdominal pain is acute in onset and often colicky. It most frequently mimics acute bowel perforation or renal colic. Hence, narcotic analgesics are often used or the patient may be anesthetized for laparotomy. In the unrecognized case, the drugs used for anesthesia may promote the progression of the attack. The patient usually has constipation and tachycardia, which further simulates an acute abdominal emergency. It is not known whether abdominal symptoms are due to a local effect of heme precursors or heme deficiency on muscle or nerves of the gastrointestinal tract or to acute autonomic neuropathy. The psychiatric symptoms that develop in this context may be misinterpreted as a reaction to the abdominal

pain. Early symptoms include restlessness, agitation, and nightmares. This may progress to frank psychosis with delirium, hallucinations, coma, and seizures. All of these events and progression to neuropathy may be accelerated if inappropriate sedating drugs are used. The neuropathy in a full attack usually develops within 2 to 3 days of the onset of abdominal and psychiatric symptoms. It has many features resembling acute inflammatory polyradiculoneuropathy (Guillain-Barré syndrome). The first symptom may be back or limb pain. Typically, motor symptoms are the earliest and most prominent clinical manifestations. The weakness may be asymmetrical. It may begin in the upper limbs or cranial nerves, and proximal muscles are as likely to be affected as distal ones. Sensory symptoms are less prominent but may involve patchy or migratory unpleasant paresthesias. Progression to maximum deficit usually occurs in a few days but may occur in a stepwise, progressive fashion for several weeks.39,40 Facial weakness and swallowing difficulty are common cranial nerve abnormalities. In severe cases motor functions of all cranial nerves may be involved, with tongue weakness and paralysis of extraocular muscles. Autonomic features, especially sympathetic hyperactivity, are prominent. These may include pupillary dilatation, tachycardia, systemic arterial hypertension, hesitancy of micturition, and constipation. Limb weakness may be severe and progress to truncal weakness and respiratory muscle paralysis. The distribution of weakness may be patchy, with prominent proximal weakness and distal sparing or weakness of selected muscle groups, particularly in the upper limbs.40 Reflexes are usually lost in proportion to the degree of muscle weakness, which distinguishes the entity from Guillain-Barré syndrome, in which the reflexes are usually lost early in the course of the disease. An ascending pattern of weakness, as seen typically in Guillain-Barré syndrome or arsenic poisoning, is unusual in porphyria.39 Atrophy is seen early in the clinical course and may be severe. It is probably due to the combination of axonal degeneration and the abdominal disturbance, which lead to poor nutrition and generalized weight loss. Sensory loss is variable and may be either a symmetrical glove-and-stocking loss or a patchy proximal loss involving predominantly the upper limbs and trunk.39 The prognosis for recovery from an individual attack depends upon the severity. It is discussed in more detail in a later section.

Clinical Differential Diagnosis Early diagnosis and prompt management are essential for aborting or ameliorating the acute attack. Clinical diagnosis is not difficult in the known porphyric. In the undiagnosed patient, any history of urine color change should raise the

1889

Porphyric Neuropathy

suspicion of porphyria. Similarly, the association of chronic photosensitivity should alert the physician to the possibility of HCP or VP. The clinical differential diagnosis of acute or subacute, predominantly motor neuropathy is short. Guillain-Barré syndrome may be suspected. The gastrointestinal disturbance of porphyria is generally more severe, and the psychiatric manifestations or seizures do not occur in Guillain-Barré syndrome. Arsenic or thallium poisoning should be suspected in all cases of acute neuropathy preceded by severe nausea, vomiting, and abdominal pain. The metals may also cause psychiatric changes and, occasionally, seizures. The only reliable way to distinguish between acute metal poisoning and porphyria is to make the appropriate measurements in blood or urine.

LABORATORY STUDIES In acute porphyria in relapse, routine liver chemistries and hematologic indices are usually normal. Electrolyte changes with low sodium levels, low serum osmolality, and high urine sodium and osmolality are probably due to inappropriate antidiuretic hormone secretion.23 It has also been suggested that intermediate metabolites that accumulate, especially ALA, are directly nephrotoxic and cause a salt-losing nephropathy.18,19 Spinal fluid examination either is normal or reveals mildly elevated protein. Heme precursors may be detected in CSF but are not routinely sought in the clinical laboratory. Heme precursor and heme synthesis enzyme activity measurements are the key diagnostic tools. In all acute porphyric attacks there is a marked increase in urinary excretion of ALA (Table 80–2). In AIP, the increase in PBG excretion (in milligrams per gram of creatinine)

exceeds that of ALA, whereas in HCP or VP, ALA excretion is more often increased to a greater extent than that of PBG. In symptomatic AIP, increases in urinary porphyrins are modest and fecal porphyrins are normal or slightly increased. In contrast, in symptomatic HCP there are marked increases in urinary and fecal coproporphyrin excretion, and in VP in fecal protoporphyrin excretion. Increases in urinary or fecal excretion of coproporphyrin are modest in VP. In addition to patients with acute attacks, two other categories of patients have been described: those with some overexcretion of heme precursors but no symptoms (latent disease); and those without either, in whom the only recognizable abnormality is in activity of a heme synthetic enzyme (carrier state). In those with latent disease, increases in ALA, PBG, or porphyrin excretions are higher during an attack than during asymptomatic periods. Detection of carriers has rested on demonstration of a decrease in activity of one of the enzymes of heme biosynthesis (see Fig. 80–2). In most cases this can be done by measuring enzyme activities in erythrocytes (the soluble enzymes ALA dehydratase and PBG deaminase) or in leukocytes (the mitochondrial enzymes coproporphyrinogen oxidase and protoporphyrinogen oxidase). However, several families with typical AIP and normal erythrocyte PBG deaminase have been described.4 In such patients, the defect is apparently limited to the liver or “housekeeping” isozyme of ALA synthase and spares the erythroid isozyme. Since the genes for both isozymes have now been cloned and the abnormalities in some kindreds identified at the level of altered nucleic acid base sequence, more definitive means of diagnosis by DNA sequencing is possible but not currently commercially available. In HCP, urine and fecal coproporphyrin levels are usually elevated in between attacks. This should be contrasted with

Table 80–2. Quantitative Porphyrin and Enzyme Levels in the Porphyrias Associated with Neuropathy Urine ALA Acute intermittent porphyria Latent Active Hereditary coproporphyria Latent Active Variegate porphyria Latent Active

PBG

Uro

Fecal Copro

Uro

Copro

Erythrocyte Proto

Porphobilinogen Deaminase*

Coproporphyrinogen Oxidase

N or q qq

N or q N or q N or q N N N qq q q N or q N or q N or q

p (or N) p (or N)

N N

N or q qq

N or q N or q qq qq q qq

N qq

N qq

N q

q qq

N q N or q qq

N or q q

N N

p (or N) p (or N)

N q N or q q

q qq

N N

N N

N ⫽ normal; q ⫽ mild or moderate increase; qq ⫽ marked increased; p ⫽ decreased activity. *Also known as uroporphyrinogen I synthase or hydroxymethylbilane synthase.

1890

Diseases of the Peripheral Nervous System

drug-induced elevation of coproporphyrin levels, in which only urine levels are elevated. In HCP, coproporphyrinogen oxidase activity in leukocytes is usually but not invariably reduced. Therefore, in order to make a reliable diagnosis, urine and fecal levels of porphyrin should be quantified as well as estimating coproporphyrinogen oxidase levels. Latent VP is usually accompanied by normal urine porphyrin levels or moderate elevation of coproporphyrin levels. It is usually suspected on the basis of the skin photosensitivity, especially in those of Afrikaner heritage. In the feces, copro- and protoporphyrin levels are usually elevated between attacks. Protoporphyrinogen oxidase levels are not routinely measured at the present time. A common diagnostic question occurs when a patient is found on routine screening to have elevated levels of coproporphyrins in the urine. This screening is carried out in many cases of undiagnosed neuropathy. Elevated urine coproporphyrin levels should lead to the suspicion of a drug effect or metal intoxication. The latter should be excluded by measurement of urine metal excretion (especially lead, arsenic, and thallium) and by careful review of the clinical features. Uroporphyrinogen I synthase and coproporphyrinogen oxidase activities should be normal in metal poisoning. Many drugs that induce the cytochrome P-450 system in liver microsomes may, even in normal subjects, also induce ALA synthase activity and produce a mild secondary increase of coproporphyrin excretion in the urine. In these cases, fecal excretion of copro- and protoporphyrin should be normal. Finally, if the patient is taken off all medications and alcohol for 2 weeks, urine coproporphyrin excretion should return to normal. This laboratory observation of elevated urine coproporphyrin is not directly associated with neurologic disease. In porphyria cutanea tarda, which is not associated with neuropathy, uroporphyrinogen decarboxylase levels are decreased in erythrocytes in the inherited but not in the sporadic, apparently acquired, form of the disease. Urinary excretion of ALA is normal or slightly increased; that of PBG is normal. In contrast, excretions of uro- and heptacarboxylporphyrins are markedly increased. In the feces, isocoproporphyrin excretion is increased, producing a distinct increase in the ratio of isocoproporphyrin to coproporphyrin in the feces. Electrophysiologic studies do not show diagnostically distinctive changes. As with any acute neuropathy or polyradiculopathy, nerve conduction studies (NCS) and needle electromyography (EMG) may be normal during the first few days of the illness. However, because the primary process appears to involve axonal degeneration (see above), compound muscle action potentials should fall in proportion to the degree of muscle weakness. Prominent slowing or dispersion should not occur unless almost all motor axons are degenerated. Slowing and dispersion are more apparent during recovery, when axons

are regenerating and may be thinly myelinated before they are fully regenerated.1 Because proximal and cranial innervated muscles are often more prominently involved, the usually measured distal NCS may be normal. Sensory conductions may be well preserved if patients do not have prominent sensory involvement. EMG becomes abnormal in 5 to 10 days. Typical denervation changes are observed. In the same way, during the recovery phase, changes typical of reinnervation will occur. Because of the possible patchy nature of the muscle involvement, it is important to examine clinically involved muscles.

TREATMENT The major thrust of treatment for all forms of neuropathic porphyria is avoidance of acute attacks. The patient with AIP is normal and requires no treatment between episodes. Patients with HCP and VP only require treatment for their skin conditions. They should be instructed in sun avoidance, the use of opaque clothing, and the liberal use of opaque sunscreen to light-exposed areas of skin. Avoidance of acute attacks is almost entirely dependent upon avoidance of porphyrogenic drugs (see Table 80–1). Because many new drugs have not been in use for sufficiently long periods to assess their porphyrogenic activity, it is safest to avoid all drugs in patients with acute porphyria. If drugs are necessary, only those that are known to be safe should be used. As discussed previously, any drug known to induce the cytochrome P-450 system should be avoided if at all possible. Since all of the neuropathic porphyrias are known to be dominantly inherited, all at-risk relatives should be screened for latent disease (see Table 80–2) and instructed in drug avoidance strategies if appropriate. Because a decreased supply of glucose to the liver may increase ALA synthase activity (see above), patients with known acute porphyria should be advised to avoid periods of prolonged starvation or excessive physical stress without carbohydrate loading.

Therapy for Acute Attacks The first measure is to avoid all drugs that might further exacerbate the disease. Chlorpromazine may be used to treat the agitation and anxiety but does have the potential for decreasing seizure threshold. Meperidine or morphine may be used for pain relief. Treatment of seizures is difficult because the commonly used parenteral anticonvulsants (barbiturates and hydantoins) are potent inducers of porphyria. Clonazepam has been used.32 Both clonazepam and valproic acid have been shown to be porphyrinogenic in experimental systems and may exacerbate AIP.6 Older drugs such as the bromides and paraldehyde may be used, although their absorption from intramuscular injection is

1891

Porphyric Neuropathy

very variable. Paraldehyde is no longer available for clinical use in the United States. Thus bromides may be the therapy of choice. In the patient with an occasional and nonrepetitive seizure, it is probably safest to not treat the seizures directly but to concentrate on management of the underlying porphyria. Acute infection and hyponatremia should be treated aggressively. The hyponatremia should be corrected gradually to minimize the risk of development of central pontine myelinolysis. Carbohydrate nutrition is very important for the reasons discussed above. Glucose is the safest form for administration, and an infusion aimed to provide 300 to 500 g/24 h should be used.4 Insulin should not be used routinely because it does not accelerate hepatic uptake of glucose and has been found in experimental systems to facilitate drug-mediated induction of ALA synthase. The most direct and effective therapy for the underlying defect is the intravenous administration of heme.8,28,46 This is given in the hydroxylated form as hematin (Panhematin; Abbott Laboratories, Abbott Park, IL). A dose of 2 to 5 mg/kg/day is given by intravenous infusion over 30 to 60 minutes. This dose should be repeated each day for 3 to 14 days. In an individual patient, the 24-hour cumulative dose should not exceed 6 mg/kg. Solutions of hematin are unstable, and the dissolved preparation should be given within 1 hour of reconstitution. Because the final solution is not transparent, it is recommended that filters be used in the intravenous line to avoid infusion of particulate material. Higher doses have been implicated in causing renal damage.15 Hematin is more stable when reconstituted with albumin in a 1:1 molar ratio (1 mg hematin:100 mg albumin). This forms methemalbumin, which is stable in solution for up to 24 hours and has been shown by Bonkovsky to inhibit production of ALA and PBG.4 Use of hematin infusion early in the acute attack appears to produce a rapid and reproducible remission of symptoms.38 Because heme is presumed to act by suppressing the induction of ALA synthase and because it is rapidly cleared by the liver, it is possible for the attack to recur rapidly after cessation of therapy. If therapy is discontinued during an acute attack, ALA and PBG levels in serum may rise rapidly and require further hematin infusion for suppression.7 Hematin is effective in the management of all three types of neuropathic porphyria.

PROGNOSIS The mortality from acute attacks of porphyria was about 30%21,39 before the advent of hematin therapy and modern intensive care techniques. It is now less than 10%.4,43 Recovery from the psychiatric and autonomic features is usually rapid. Recovery from the neuropathic damage depends upon the severity of axonal and neuronal loss during the acute

attack. Good functional recovery is the rule, although recovery may be prolonged and incomplete as would be expected in a process involving axonal degeneration.9,42

ACKNOWLEDGMENTS The helpful comments of Dr. Gregory D. Cascino and the expert secretarial assistance of Jane M. Meyer are most gratefully acknowledged.

REFERENCES 1. Albers, J. W., Robertson, W. C. Jr., and Daube, J. R.: Electrodiagnostic findings in acute porphyric neuropathy. Muscle Nerve 1:292, 1978. 2. Anzil, A. P., and Dozic, S.: Peripheral nerve changes in porphyric neuropathy: findings in a sural nerve biopsy. Acta Neuropathol. (Berl.) 42:121, 1978. 3. Berger, H., and Goldberg, A.: Hereditary coproporphyria. Br. Med. J. 2:85, 1955. 4. Bonkovsky, H. L.: Porphyrin and heme metabolism and the porphyrias. In Zakim, D., and Boyer, T. D. (eds.): Hepatology: A Textbook of Liver Disease. Philadelphia, W. B. Saunders, p. 378, 1990. 5. Bonkovsky, H. L., and Schady, W.: Neurologic manifestation of acute porphyria. Semin. Liver Dis. 2:108, 1982. 6. Bonkovsky, H. L., Sinclair, P. R., Emery, S., and Sinclair, J. F.: Seizure management in acute hepatic porphyria: risks of valproate and clonazepam. Neurology 30:588, 1980. 7. Bonkovsky, H. L., Sinclair, P. R., and Sinclair, J. F.: Hepatic heme metabolism and its control. Yale J. Biol. Med. 52:13, 1979. 8. Bonkovsky, H. L., Tschudy, D. P., Collins, A., et al.: Repression of the overproduction of porphyrin precursors in acute intermittent porphyria by intravenous infusions of hematin. Proc. Natl. Acad. Sci. U. S. A. 68:2725, 1971. 9. Bosch, E. P., Pierach, C. A., Bossenmaier, I., et al.: Effect of hematin in porphyric neuropathy. Neurology 27:1053, 1977. 10. Brodie, M. J., Thompson, G. G., Moore, M. R., et al.: Hereditary coproporphyria: demonstration of the abnormalities in haem biosynthesis in peripheral blood. Q. J. Med. 46:229, 1977. 11. Cavanagh, J. B., and Mellick, R. S.: On the nature of the peripheral nerve lesions associated with acute intermittent porphyria. J. Neurol. Neurosurg. Psychiatry 28:320, 1965. 12. Connon, J. J., and Turkington, V.: Hereditary coproporphyria. Lancet 2:263, 1968. 13. De Matteis, F., Zetterlund, P., and Wetterberg, L.: Brain ␦-aminolaevulinate synthase: developmental aspects and evidence for regulatory role. Biochem. J. 196:811, 1981. 14. Dean G.: The Porphyrias: A Story of Inheritance and Environment, 2nd ed. London, Pitman Medical, 1971. 15. Dhar, G. J., Bossenmaier, I., Petryka, Z. J., et al.: Effects of hematin in hepatic porphyria: further studies. Ann. Intern. Med. 83:20, 1975.

1892

Diseases of the Peripheral Nervous System

16. Dichter, H. N., Taddeini, L., Lin, S., and Ayala, G. F.: Delta amino levulinic acid: effect of a porphyrin precursor on an isolated neuronal preparation. Brain Res. 126:189, 1977. 17. Eales, L., Day, R. S., and Blekkenhorst, G. H.: The clinical and biochemical features of variegate porphyria: an analysis of 300 cases studied at Groote Schuur Hospital, Cape Town. Int. J. Biochem. 12:837, 1980. 18. Eales, L., and Dowdle, E. B.: Electrolyte abnormalities in porphyria. Lancet 1:51, 1969. 19. Eales, L., Dowdle, E. B., and Sweeney, G. D.: The acute porphyric attack. I. The electrolyte disorder of the acute porphyric attack and the possible role of delta-aminolaevulic acid. S. Afr. Med. J. 17:89, 1971. 20. Erbslöh, W.: Pathology and pathological anatomy of toxic polyneuritis after application of Sulfonal [in German]. Dtsch. Z. Nervenheilkd. 23:197, 1903. 21. Goldberg, A.: Acute intermittent porphyria. Q. J. Med. 28:183, 1959. 22. Goldberg, A., Rimington, C., and Lochhead, A. C.: Hereditary coproporphyria. Lancet 1:632, 1967. 23. Hellman, E. S., Tschudy, D. P., and Bartter, F. C.: Abnormal electrolyte and water metabolism in acute intermittent porphyria: the transient inappropriate secretion of antidiuretic hormone. Am. J. Med. 32:734, 1962. 24. Hierons, R.: Changes in the nervous system in acute porphyria. Brain. 80:176, 1957. 25. Hierons, R.: Acute intermittent porphyria. Postgrad. Med. J. 43:605, 1967. 26. Kappas, A., Sassa, S., and Anderson, K. E.: The porphyrias. In Stanbury, J. B., Wyngaarden, J. B., Fredrickson, D. S., et al.: The Metabolic Basis of Inherited Disease. New York, McGraw-Hill, p. 1301, 1983. 27. Kramer, S., Viljoen, D., Becker, D., and Metz, J.: A possible explanation for the neurological disturbances in the porphyrias. S. Afr. Med. J. 17:103, 1971. 28. Lamon, J. M., Frykholm, B. C., Hess, R. A., and Tschudy, D. P.: Hematin therapy for acute porphyria. Medicine 58:252, 1979. 29. Lindberg, R. L., Porcher, C., Grandchamp, B., et al.: Porphobilinogen deaminase deficiency in mice causes a neuropathy resembling that of human hepatic porphyria. Nat. Genet. 12:195, 1996. 30. Loots, J. M., Becker, D. M., Meyer, B. J., and Goldstuck, N. K. S.: The effect of porphyrin precursors on monosynaptic reflex activity in the isolated hemisected frog spinal cord. J. Neural Transm. 36:71, 1975. 31. Macalpine, I., and Hunter, R.: George III and the Mad Business. New York, Pantheon, 1969.

32. Magnussen, C. R., Doherty, J. M., Hess, R. A., and Tschudy, D. P.: Grand mal seizures and acute intermittent porphyria: the problem of differential diagnosis and treatment. Neurology 25:1121, 1975. 33. Meyer, U. A., Mace, M., Schuurmanns, M., and Lindberg, R. L. P.: Acute porphyrias: pathogenesis of neurological manifestations. Semin Liver Dis. 18:43, 1998. 34. Mustajoki, P., and Seppäläinen, A. M.: Neuropathy in latent hereditary hepatic porphyria. Br. Med. J. 2:310, 1975. 35. Paterniti, J. R. Jr., Simone, J. J., and Beattie, D. S.: Detection and regulation of delta-aminolevulinic acid synthetase activity in the rat brain. Arch. Biochem. Biophys. 189:86, 1978. 36. Percy, V. A., and Shanley, B. C.: Porphyrin precursors in blood, urine and cerebrospinal fluid in acute porphyria. S. Afr. Med. J. 52:219, 1977. 37. Percy, V. A., and Shanley, B. C.: Studies on haem biosynthesis in rat brain. J. Neurochem. 33:1267, 1979. 38. Pierach, C. A.: Hematin therapy for the porphyric attack. Semin. Liver Dis. 2:125, 1982. 39. Ridley, A.: The neuropathy of acute intermittent porphyria. Q. J. Med. 38:307, 1969. 40. Ridley, A.: Porphyric neuropathy. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. P. (eds.): Peripheral Neurology, 3rd ed. Philadelphia, W. B. Saunders, p. 1704, 1984. 41. Shemin, D.: d-Aminolevulinic acid dehydratase. In Boyer, P. D. (ed.): The Enzymes, Vol. 3, 3rd ed. New York, Academic Press, p. 323, 1972. 42. Sorenson, A. W. S., and With, T. K.: Persistent pareses after porphyric attacks. S. Afr. J. Lab. Clin. Med. 17:101, 1971. 43. Stein, J. A., and Tschudy, D. P.: Acute intermittent porphyria: a clinical and biochemical study of 46 patients. Medicine 49:1, 1970. 44. Thomas, P. K.: The morphological basis for alterations in nerve conduction in peripheral neuropathy. Proc. R. Soc. Med. 64:295, 1971. 45. Watanabe, N., Hayashi, N., and Kikuchi, G.: d-Aminolevulinate synthase isozymes in the liver and erythroid cells of chicken. Biochem. Biophys. Res. Commun. 113:377, 1983. 46. Watson, C. J.: Hematin and porphyria. N. Engl. J. Med. 293:605, 1975. 47. Wetterberg, L.: A Neuropsychiatric and Genetical Investigation of Acute Intermittent Porphyria. Stockholm, Svenska Bokforlaget, 1967. 48. Yeung, A. C., Moore, M. R., and Goldberg, A.: Pathogenesis of acute porphyria. Q. J. Med. 63:377, 1987.

81 Fabry’s Disease ROSCOE O. BRADY AND RAPHAEL SCHIFFMANN

Historical Background Clinical Manifestations Genetics Pathology and Organ System Involvement Skin Cardiovascular and Pulmonary Systems Kidney Central and Peripheral Nervous Systems Eye

Reticuloendothelial System Skeletal Muscle Liver and Gastrointestinal System Histochemical Studies and Ultrastructural Changes Nature of the Accumulating Lipid Enzymatic Defect Source of the Accumulating Gb3 Molecular Defects Diagnosis

Intense, frequently incapacitating burning pains in the hands, feet, and lower portions of the legs with onset often in childhood are hallmarks of Fabry’s disease. The metabolic abnormality in this disorder is insufficient or absent activity of the lysosomal enzyme ceramide trihexosidase, now generally referred to as -galactosidase A (-Gal A).10 The lack of -Gal A causes the deposition of harmful quantities of ceramide trihexoside (globotriaosylceramide [Gb 3 ]) in many organs and tissues in patients with Fabry’s disease. Procedures are available for diagnosis, detection of heterozygous carriers, and monitoring of pregnancies at risk for this disorder by measuring -Gal A activity in appropriate cells and by genotyping. Since the preceding edition of Peripheral Neuropathy was published, an especially important advance has been made concerning reduction of the neuropathic pain and improvement of the quality of life of patients with Fabry’s disease by enzyme replacement therapy (ERT).46

HISTORICAL BACKGROUND Original descriptions of patients by Anderson5 and Fabry22 are chronicled in detail in the preceding editions of Peripheral Neuropathy.

Hemizygous Affected Males Heterozygous Females Prenatal Diagnosis Treatment Enzyme Replacement Therapy Gene Therapy Substrate Depletion Therapy Conclusions

CLINICAL MANIFESTATIONS Fabry’s disease should be considered when boys or young men present with pains in their feet and intense burning sensations in the hands and lower legs. The pains are often so severe that they cannot walk even moderate distances or participate in sports. Young patients frequently have some mild behavioral aberrations, and they are often bothered by persistent apprehensiveness. Most male patients with Fabry’s disease have reddish purple maculopapular angiokeratomas that may be restricted to the umbilical, scrotal, inguinal, and gluteal regions. There are also dark red angiectases under the nail beds, and the oral mucosa and conjunctiva may be involved. These dermatologic manifestations vary considerably, and careful scrutiny may be required to discover them. The principal cause of disability and death in patients with Fabry’s disease is progressive renal impairment. The initial manifestations are albuminuria and inability to concentrate urine. Chronic renal insufficiency and end-stage kidney disease occur later. Hypertension, premature strokes, and early myocardial infarctions also occur in patients with this disorder. Other manifestations include anhidrosis, corneal opacification, cardiomegaly, and persistent peripheral edema that may antedate the hypertension or kidney damage. 1893

1894

Diseases of the Peripheral Nervous System

Fabry’s disease heterozygotes have a varying degree of the manifestations of the disorder. Not infrequently, the carriers have peripheral neuralgia with onset later in life than that in the afflicted males. Heterozygotes may also have corneal opacities and electrocardiographic alterations, and some have obvious signs of renal impairment. Only rarely, however, does one encounter the severe renal failure that is characteristic of affected males.

GENETICS Fabry’s disease is an X-linked recessive disorder.41 Only the mother need be a carrier to have an affected (hemizygous) male child. On a statistical basis, half her sons will have the disorder and half her daughters will be carriers; the others will not be involved. It must be kept in mind that the same odds for having an affected child occur with each pregnancy, and therefore the disease may occur sequentially. When a heterozygous carrier is identified through family history, enzymatic assay, or genotyping, prenatal monitoring of the pregnancies is available.12 More than 200 mutations have been identified in the gene for -Gal A in patients with Fabry’s disease.4,9 Many of the mutations result in large deletions or stop codons that prevent the formation of the enzyme. Those patients are cross-reacting immunologic material negative. This situation caused some concern when ERT trials were undertaken because of the possibility of sensitizing the recipients to exogenous -Gal A. However, detrimental immunologic difficulties were not encountered.46

PATHOLOGY AND ORGAN SYSTEM INVOLVEMENT The stored Gb3 is found mainly in vascular elements, particularly in the arterial system, with preference for the endothelial, perithelial, and smooth muscle cells. Neurons are involved to a variable extent. Sinusoidal histiocytes of the reticuloendothelial system also contain this material. Unmetabolized crystalline substance is usually found to be greater on ultrastructural evaluation than can be appreciated on light microscopic observation.

Skin Widely dispersed or confluent angiokeratomas are considered to be telangiectasias in the epidermis associated with a proliferation of keratin and epidermal cells. These lesions are generally found over the lower part of the

trunk, buttocks, perineum, scrotum, and upper thighs.27,60 Angiokeratomas have, however, also been found on the lip and oral mucosa.32,42 Frozen sections may show lipid in the vascular (particularly arterial) elements in the dermis.26 Lipid has been demonstrated in the basal cells of the epidermis and sweat glands. Vacuolization of the sweat glands in the usual hematoxylin and eosin–stained sections would suggest the diagnosis. In frozen sections, the lipid can be identified in these cells as well as in the vascular elements coursing through the dermis. The histologic diagnosis can be made by skin biopsy (which may not necessarily include the angiokeratoma), in which one may demonstrate the typical ultrastructural inclusions in endothelial cells.19,26,45

Cardiovascular and Pulmonary Systems Crystalline material has been found in the central portions of cardiac muscle fibers as well as in the smooth muscle of the larger vascular elements, principally the arteries. The vascular deposition is the underlying cause for most of the organ system failure in this disease. In frozen sections the material can be seen in the endothelial cells, perithelial cells, and smooth muscle fibers. Cardiomegaly is also found and may be the result of the hypertension that these patients develop. Electrocardiographic and vectorcardiographic changes have been reported.33,36,44 Lung changes are primarily vascular but also involve alveolar lining cells and may predispose to pulmonary malfunction.14,43

Kidney Renal involvement is characterized by foam cells in both the endothelial cells of the glomeruli and the tubular cells, particularly evident in distal tubules. Shedding of these elements into the tubular lumen accounts for the appearance of birefringent lipid material in the urine sediment. Mesangial widening is an early aspect of renal damage.13

Central and Peripheral Nervous Systems Neurons are only slightly or unevenly involved, as are ganglion cells in the central nervous system, the spinal nerves, and the sympathetic nervous system. Ganglion cells in the peripheral nervous system are variably involved, and occasionally perineurial cells contain storage material (Fig. 81–1). The involvement of the ganglions of the peripheral nervous system and sympathetic nervous system may be the cause of anhidrosis or hypohidrosis and possibly of the spasmodic pain.15 Fiber loss and degenerative changes in unmyelinated nerve fibers and small myelinated fibers and relative preservation of large myelinated fibers have been described (Table 81–1)

Fabry’s Disease

FIGURE 81–1 A, Typical glycolipid granules of Fabry’s disease in the perineurial cells (arrow) from a transverse section of sural nerve. B, An enlargement of the rectangular area shown in A. C, Glycolipid granules in endothelial and perithelial cells of a small endoneurial capillary from a patient with Fabry’s disease. (From Ohnishi, A., and Dyck, P. J.: Loss of small peripheral sensory neurons in Fabry’s disease. Arch. Neurol. 31:120, 1974, with permission. Copyright 1974, The American Medical Association.)

1895

1896

Diseases of the Peripheral Nervous System

Table 81–1. Comparison of Sural Nerve Quantitation with Thigh Innervation Density

Control Patient 1 3 6

Large Myelinated Fibers (fibers/mm2)

Small Myelinated Fibers (fibers/mm2)

Unmyelinated fibers (fibers/mm2)

Thigh innervation density (fibers/mm)

2848–4187

4099–6556

31,311–41,642

18.5  3.5

ND 4956 3951

ND 2623 3134

26,514 21,063 26,645

7.6  1.8 13.9  1.8 4.0  1.3

ND  not done. From Scott, L. J., Griffin, J. W., Luciano, C., et al.: Qualitative analysis of epidermal innervation in Fabry disease. Neurology 52:1249, 1999, with permission.

(Fig. 81–2).8,29,48 Patients with Fabry’s disease have a length-dependent small fiber neuropathy that is characterized by normal nerve conduction velocity in the vast majority of patients with normal renal function, as well as an increased threshold of cold and warm perception with reduced epidermal innervation density48,53 (Fig. 81–3). The vascular elements supplying these systems contain quite a bit of the storage material (see Fig. 81–1C). Vascular involvement of the central nervous system (sometimes associated with hypertension) is one of the major causes of death from cerebrovascular accidents among these patients. Recent work on the nature of vasculopathy in Fabry’s disease showed the presence of a prothrombotic state17 coupled with a vascular endothelium– dependent hyperreactivity3 and increased cerebral blood flow that is probably due to non–nitric oxide pathway dysregulation.38

Eye Birefringent lipids are found in ocular tissues, particularly the vacuolated smooth muscle fibers of the media of the arterial elements, especially in the choroid, retina and ciliary body, iris, and limbus. Ultrastructural studies have demonstrated material in the basal cell layers of the cornea, corneal epithelium, pigmented epithelium of the iris, and epithelium of the lens.25 Ganglion cells of the uvea and retina are not usually involved.24 Aneurysmal dilatations and retinal vessel tortuosity are also frequent in affected males.49

Reticuloendothelial System Although lymphadenopathy and massive splenomegaly are not common, the cords of Billroth of the spleen as well as the sinusoidal and endothelial cells of the lymph node tissue are involved. Scattered clumps of cells are found in the bone marrow and may or may not be associated with endothelial cells.

Skeletal Muscle The vascular involvement is the primary manifestation, although there appears to be accumulation of material between myofibrils or under the plasma membrane.58

Liver and Gastrointestinal System The Kupffer cells in the liver are involved to some extent. There is, however, very little evidence of hepatocyte involvement. Vascular elements are affected as previously described. The same vascular involvement is found in the gastrointestinal tract. In addition, rectal neuronal cells of Meissner’s plexus and lamina propria smooth muscle cells and Brunner’s glands may show some changes.7 This may be the basis of the gastrointestinal signs and symptoms that occur in many Fabry’s disease patients.

HISTOCHEMICAL STUDIES AND ULTRASTRUCTURAL CHANGES Birefringent lipid is demonstrated in frozen sections of fresh and formalin-fixed material. In paraffin sections, however, most of this lipid has been removed. In the cholesterol that is present, the Schultz reaction or the modified Liebermann-Burchard reaction is moderately positive, and it has been reported that perchloric acid–naphthoquinone staining is also positive.32 The material also shows a positive reaction to some silver stains, including Gomori’s methenamine silver. Oil red O and Sudan black B reactions are positive. Luxol fast blue gives variable results in frozen tissue. The stored material is intensely periodic acid–Schiff positive, particularly in thicker sections. It is affected by digestion with diastase.26 It also shows positive acid mucopolysaccharide reactions, but this is true of most histiocytes and occurs in other sphingolipid storage diseases. Acid phosphatase stains react positively but not to the extent seen in Gaucher’s cells.

Fabry’s Disease

FIGURE 81–2 Transverse sections from a sural nerve of a patient with Fabry’s disease. A, A clump of filaments within the axis cylinder of a myelinated fiber. B, High-power view of these filaments. C, Membrane-bound aggregates of glycogen granules and of filaments in the axoplasm of a medium-sized myelinated fiber. D, Aggregation of mitochondria and osmiophilic particles in axoplasm of an unmyelinated fiber. E, Stacks of Schwann cell processes, probably indicating denervated clusters of unmyelinated fibers, adjacent to more normal-looking unmyelinated fibers and a somewhat dilated unmyelinated fiber (lower left). (From Ohnishi, A., and Dyck, P. J.: Loss of small peripheral sensory neurons in Fabry’s disease. Arch. Neurol. 31:120, 1974, with permission. Copyright 1974, The American Medical Association.)

1897

1898

Diseases of the Peripheral Nervous System

FIGURE 81–3 Kaplan-Meier curves for patients and controls for the three sensory modalities in the hand and foot. JND  just noticeable difference. (From Luciano, C. A., Russell, J. W., Banerjee, T. K., et al.: Physiological characterization of neuropathy in Fabry’s disease. Muscle Nerve 26:622, 2002, with permission.)

The stored material appears to consist of concentric laminated myelin figure–like inclusions that are occasionally called zebra bodies. These have a thickness of 140 to 210 nm. The lamellae have a fine internal periodicity of approximately 5 nm.42,45 They are found in the cytoplasm, compressing the nuclei (Fig. 81–4). The material reacts with acid phosphatase stain adapted for electron microscopy, suggesting that it is present in engorged lysosomes.26 It is also found in the cytoplasm of endothelial cells, smooth muscle cells, and perithelial cells of blood vessels and variably in neuronal cells, ganglion cells, and perineurial cells.40,50

molecules of galactose are joined to the primary alcohol group on carbon 1 of the sphingosine moiety of ceramide (ceramide-glucose-galactose-galactose). The terminal galactose is linked to the internal molecule of galactose by an -glycosidic bond. The internal galactose is joined to glucose and the glucose to ceramide through -glycosidic bonds. Although this is the characteristic accumulating lipid in most organs and tissues of patients with Fabry’s disease, a disaccharide consisting of ceramide-galactosegalactose is found in abnormally high concentrations in the kidney, urinary sediment, and pancreas of patients with this disorder. The anomeric bond between the two molecules of galactose in this lipid is also believed to be in the  configuration.

Nature of the Accumulating Lipid The substance that accumulates in the kidney, liver, spleen, adventitia of blood vessels of the skin, nerves, and other organs of patients with Fabry’s disease is ceramide trihexoside, now generally referred to as globotriaosylceramide (Gb3).52 The lipid consists of the long-chain amino alcohol sphingosine, to which a long-chain fatty acid is linked via an amide bond to the nitrogen atom on carbon 2 of sphingosine. This fatty acyl–sphingosine moiety is called ceramide. One molecule of glucose and two

Enzymatic Defect The enzymatic defect in Fabry’s disease is a deficiency of the enzyme that catalyzes the hydrolytic cleavage of the terminal molecule of galactose from the accumulating Gb3.10 Patients with Fabry’s disease frequently have very little or no -Gal A activity in their tissues. A major exception to this condition occurs in patients with the so-called cardiac variant of Fabry’s disease, whose pathologic signs

Fabry’s Disease

FIGURE 81–4 A, Angiokeratoma corporis diffusum of Fabry’s disease on the skin. This section is not diagnostic. (Hematoxylin and eosin; 8.) (AFIP Neg. No. 73-1276-1.) B, Foam cells in renal glomerulus in a patient with Fabry’s disease. (Hematoxylin and eosin; 250.) (AFIP Neg. No. 73-1276-2.) C, Frozen section of kidney demonstrating birefringent material in renal tubules and wall of arteriole (300). (AFIP Neg. No. 73-1276-3.) D, Electron micrograph of lamellated inclusions (“zebra bodies”) (14,000). (AFIP Neg. No. 73-1276-9.) E, Electron micrograph of macrophage in spleen (5100). Note compression of nucleus with lysosomal inclusions. (AFIP Neg. No. 73-1276-8.) F, High-power electron micrograph demonstrating internal periodicity of lamellae (approx. 150,000). (AFIP Neg. No. 73-1276-10.)

1899

1900

Diseases of the Peripheral Nervous System

are mainly referable to cardiac hypertrophy and dysfunction. In these individuals, -Gal A activity is about 10% to 12% of normal, which is considerably higher than that in the usual patient with Fabry’s disease. -Gal A activity in heterozygotes ranges from low to almost normal. This wide variation presumably accounts for the variability in the clinical presentation of carriers.

Source of the Accumulating Gb3 The principal source of the accumulating glycolipid appears to be the major neutral glycolipid in the stroma of erythrocytes called globoside (ceramide-glucose-galactosegalactose–N-acetylgalactosamine) and to a lesser extent Gb3 itself. It is apparent from studies with -Gal A knockout mice that formation of Gb3 occurs in many tissues in situ.39

MOLECULAR DEFECTS The -Gal A gene is on the long arm of the X chromosome in the region between Xq21.33 and Xq22. It has been cloned and a large number of mutations have been described in Fabry’s disease patients. These alterations have resulted in (1) changes in the amino acid sequence that reduce the catalytic activity of the enzyme, (2) complete lack of production of the enzyme, (3) synthesis of an unstable protein and absence of mature enzyme, (4) abnormal processing of the enzyme to the mature form, and (5) abnormal synthesis of the immature precursor of the processed enzyme.

DIAGNOSIS Hemizygous Affected Males -Gal A activity is usually assayed with commercially available substrates such as p-nitrophenyl--D-galactopyranoside and the fluorogenic compound 4-methylumbelliferyl--Dgalactopyranoside. N-acetylgalactosamine is included in the incubation mixtures to suppress the activity of -gal B. Leukocytes and cultured skin fibroblasts are often used for these determinations.

Heterozygous Females Radiolabeled Gb3 and artificial substrates are useful for the identification of heterozygotes.10,12 Hair follicles are sometimes used as a tissue source when equivocal values are obtained with leukocytes and cultured skin fibroblasts.51 Quantitation of glycolipids in urine sediment has been used for the diagnosis of Fabry’s disease,16,59 and may be helpful in certain situations in which carrier status is uncertain.16 The identification of precise molecular alterations in the -Gal A gene (genotyping) aids in the

detection of heterozygous carriers. However, it should be remembered that a significant percentage of cases of Fabry’s disease have been attributed to spontaneous (private) mutations in the -Gal A gene.

Prenatal Diagnosis Cultured amniotic cells from normal fetuses have considerable -Gal A activity; consequently it is feasible to monitor pregnancies at risk for Fabry’s disease.12 Assays of -Gal A activity in chorionic villus biopsy specimens provide for accelerated prenatal diagnosis.28

TREATMENT A number of patients with Fabry’s disease have obtained some relief from their peripheral neuralgia with moderate quantities of phenytoin,34 but this response has not been universal. Carbamazepine has been somewhat more effective in this regard. However, caution must be exercised with this agent because it can cause exacerbation of autonomic dysfunction.23 Gastrointestinal complaints, including early satiety, postprandial bloating, episodic diarrhea, and abdominal pain, are common in Fabry’s disease patients. Treatment with metoclopramide or pancrelipase brings about symptomatic improvement in many of these individuals.6 In the past, many of the male patients with this disorder have required hemodialysis or kidney transplantation because of renal difficulties as they reach their 40s.57 Several of them had relief from the peripheral neuralgia and a partial restoration of sweating. The mechanism of these responses is far from clear. It appears that there is a reduction of Gb3 in the blood soon after operation, but it usually returns gradually to the preoperative level. Furthermore, the implanted kidney does not appear to secrete -Gal A into the circulation.

Enzyme Replacement Therapy Fabry’s disease patients appeared to be reasonable candidates for ERT. An early investigation was carried out with -Gal A isolated from human placental tissue.30 -Gal A was rapidly cleared from the blood and was taken up by the liver to a major extent.11 The level of Gb3 in the circulation fell rapidly from a threefold elevation to the normal range in the two recipients. However, within 48 hours, it had returned to the preinfusion value in both patients. Similar results were obtained later by another group of investigators using -Gal A isolated from human spleen and from human plasma.18 Further ERT trials in Fabry’s disease were delayed many years until biotechnological procedures became available for the production of sufficient quantities of -Gal A. The first of these was the production of -Gal A by transduced human skin fibroblasts. -Gal A obtained in this manner

1901

Fabry’s Disease

BPI score

was employed in a Phase I safety and dose escalation trial in patients with Fabry’s disease.47 The enzyme was found to be safe, and its potential was indicated by significant reductions of Gb3 in the liver and in urine sediment. The latter was particularly encouraging because it signaled the ability of ERT to reduce the accumulation of Gb3 in the kidney and the possibility of reversing renal pathology in patients with Fabry’s disease. These auspicious findings led to the design and implementation of a double-blind, placebo-controlled clinical efficacy trial of ERT in patients with Fabry’s disease using -Gal A produced in a continuous human cell line by activation of the endogenous -Gal A gene. One of the principal outcome measures was the determination of the effect of ERT on the painful acroparesthesias in patients whose pain medication was withheld for several days before completing the Brief Pain Inventory (BPI), a validated test instrument for assessing pain.31 There was a significant reduction of pain in the treated cohort compared with patients in the placebo group (Fig. 81–5). However, no increase in epidermal innervation density was apparent in the enzyme-treated patients. Quality of life of the recipients also improved, judging from their responses to the interference items on the BPI. There was a reduction in the number of glomeruli with mesangial widening in the treated group. There was also an improvement in creatinine clearance in the treated patients. Patients who received the enzyme gained weight; and there was a dramatic improvement of the gastrointestinal difficulties in this cohort compared with the placebo group. Other indications of the effectiveness of ERT were the reduction in cerebral hyperperfusion that is characteristic of Fabry’s disease and in nitrotyrosine staining of dermal blood vessels

10 9 8 7 6 5 4 3 2 1 0

(Fig. 81–6).37,38 With continued administration of -Gal A, there was a significant improvement in the threshold for cold sensation in the hands and for warm sensation in the feet. Another ERT trial was carried out independently using -Gal A produced in Chinese hamster ovary cells. These investigators reported clearing of Gb3 from kidney vascular endothelial cells.20,21 Together, the findings indicate the feasibility and potential of ERT for the treatment of Fabry’s disease.

Gene Therapy Much of the accumulating Gb3 in Fabry’s disease appears to arise from the catabolism of sphingolipids present in the stroma of senescent erythrocytes, and their biodegradation occurs primarily in reticuloendothelial cells. Because these cells arise from the bone marrow, replacement of the mutated gene with its functioning counterpart in bone marrow stem or progenitor cells may arrest, and possibly reverse, the accumulation of Gb3. We have carried out pivotal investigations of gene therapy in the mouse -Gal A knockout model of Fabry’s disease with highly encouraging results.35,54–56 Considerable thought is now being given to gene therapy trials in patients with Fabry’s disease. We have the impression that a continuous supply of -Gal A produced in the body of patients by gene therapy may be more beneficial than intermittent infusions of this enzyme.

Substrate Depletion Therapy Inhibition of the biosynthesis of sphingolipids is under consideration as a potential alternative or as an adjunct to ERT. Blocking the synthesis of Gb3 reduced the quantity of this lipid in immortalized lymphocytes derived from Fabry’s disease patients1 and reduced tissue stores in the -Gal A knockout mouse model of Fabry’s disease.2 It is anticipated that substrate depletion will eventually be evaluated in patients as a single agent or, perhaps more realistically, in conjunction with ERT.

CONCLUSIONS 0

5

10

15

20

25

Time (weeks)

FIGURE 81–5 Brief Pain Inventory “Pain at Its Worst” scores for patients with Fabry’s disease in clinical efficacy trial of enzyme replacement therapy. Filled circles  patient cohort who received -Gal A; open circles  placebo control group. P  0.02. (From Schiffmann, R., Kopp, J. B., Austin, H. A. 3rd, et al.: Enzyme replacement therapy in Fabry disease: a randomized controlled trial. JAMA 285:2743, 2001, with permission.)

Development of simple procedures for the detection of patients and carriers might theoretically provide for control of the incidence of this heritable disorder through genetic counseling. It should be remembered, however, that many cases apparently arise from new, spontaneous mutations in the -Gal A gene. Substantial progress has been made toward the development of effective ERT for patients with Fabry’s disease. For a number of reasons, we deduce that, in the future, gene therapy may be superior to ERT.

1902

Diseases of the Peripheral Nervous System

FIGURE 81–6 Dermal nitrotyrosine (A, C, and E) and laminin (B, D, and F) immunohistochemical localization in control skin and in skin from Fabry’s disease patients before and after enzyme replacement therapy (ERT). A and B, Control skin showing nitrotyrosine staining (A; rhodamine filter, 10) and laminin staining (B; fluorescent filter, 10) of dermal vessels (arrows) and stratum basalis (same section). C and D, Nitrotyrosine staining before ERT reveals significant perivascular staining (arrows). E and F, After ERT, dermal vascular staining is no longer apparent. (From Moore, D. F., Scott, L. T., Gladwin, M. T., et al.: Regional cerebral hyperperfusion and nitric oxide pathway dysregulation in Fabry disease: reversal by enzyme replacement therapy. Circulation 104:1506, 2001, with permission.) See Color Plate

Fabry’s Disease

REFERENCES 1. Abe, A., Arend, L. J., Lee, L., et al.: Glycosphingolipid depletion in Fabry disease lymphoblasts with potent inhibitors of glucosylceramide synthase. Kidney Int. 57:446, 2000. 2. Abe, A., Gregory, S., Lee, L., et al.: Reduction of globotriaosylceramide in Fabry disease mice by substrate deprivation. J. Clin. Invest. 105:1563, 2000. 3. Altarescu, G., Moore, D. F., Pursley, R., et al.: Enhanced endothelium-dependent vasodilation in Fabry disease. Stroke 32:1559, 2001. 4. Altarescu, G. M., Goldfarb, L. G., Park, K. Y., et al.: Identification of fifteen novel mutations and genotypephenotype relationship in Fabry disease. Clin. Genet. 60:46, 2001. 5. Anderson, W. A.: A case of angiokeratoma. Br. J. Dermatol. 18:113, 1898. 6. Argoff, C. E., Barton, N. W., Brady, R. O., and Ziessman, H. A.: Gastrointestinal symptoms and delayed gastric emptying in Fabry’s disease: response to metoclopramide. Nucl. Med. Commun. 19:887, 1998. 7. Bagdade, J. D., Parker, F., Ways, P. O., et al.: Fabry’s disease: a correlative clinical, morphologic, and biochemical study. Lab. Invest. 18:681, 1968. 8. Bischoff, A., Fierz, U., Regli, F., and Ulrich, J.: Peripheral neurological disorders in Fabry’s disease (angiokeratoma corporis diffusum universale): clinical and electron microscopic findings in a case. Klin. Wochenschr. 46:666, 1968. 9. Blaydon, D., Hill, J., and Winchester, B.: Fabry disease: 20 novel GLA mutations in 35 families. Hum. Mutat. 18:459, 2001. 10. Brady, R. O., Gal, A. E., Bradley, R. M., et al.: Enzymatic defect in Fabry’s disease: ceramidetrihexosidase deficiency. N. Engl. J. Med. 276:1163, 1967. 11. Brady, R. O., Tallman, J. F., Johnson, W. G., et al.: Replacement therapy for inherited enzyme deficiency: use of purified ceramidetrihexosidase in Fabry’s disease. N. Engl. J. Med. 289:9, 1973. 12. Brady, R. O., Uhlendorf, B. W., and Jacobson, C. B.: Fabry’s disease: antenatal detection. Science 172:174, 1971. 13. Branton, M. H., Schiffmann, R., Sabnis, S. G., et al.: Natural history of Fabry renal disease: influence of a-galactosidase A activity and genetic mutations on clinical course. Medicine (Baltimore) 81:122, 2002. 14. Brown, L. K., Miller, A., Bhuptani, A., et al.: Pulmonary involvement in Fabry disease. Am. J. Respir. Crit. Care Med. 155:1004, 1997. 15. Cable, W. J., Kolodny, E. H., and Adams, R. D.: Fabry disease: impaired autonomic function. Neurology 32:498, 1982. 16. Cable, W. J., McCluer, R. H., Kolodny, E. H., and Ullman, M. D.: Fabry disease: detection of heterozygotes by examination of glycolipids in urinary sediment. Neurology 32:1139, 1982. 17. DeGraba, T., Azhar, S., Dignat-George, F., et al.: Profile of endothelial and leukocyte activation in Fabry patients. Ann. Neurol. 47:229, 2000. 18. Desnick, R. J., Dean, K. J., Grabowski, G., et al.: Enzyme therapy in Fabry disease: differential in vivo plasma

19.

20.

21.

22. 23.

24.

25.

26.

27. 28.

29. 30.

31.

32. 33.

34.

35.

36.

1903 clearance and metabolic effectiveness of plasma and splenic alpha-galactosidase A isozymes. Proc. Natl. Acad. Sci. U. S. A. 76:5326, 1979. Dvorak, A. M., Cable, W. J., Osage, J. E., and Kolodny, E. H.: Diagnostic electron microscopy. II. Fabry’s disease: use of biopsies from uninvolved skin. Acute and chronic changes involving the microvasculature and small unmyelinated nerves. Pathol. Annu. 16:139, 1981. Eng, C. M., Banikazemi, M., Gordon, R. E., et al.: A Phase 1/2 clinical trial of enzyme replacement in Fabry disease: pharmacokinetic, substrate clearance, and safety studies. Am. J. Hum. Genet. 68:711, 2001. Eng, C. M., Guffon, N., Wilcox, W. R., et al.: Safety and efficacy of recombinant human alpha-galactosidase A–replacement therapy in Fabry’s disease. N. Engl. J. Med. 345:9, 2001. Fabry, J.: Ein Beitrag zur Kenntnis der Purpura haemorrhagica nodularis. Arch. Dermatol. Syphilol. 43:187, 1898. Filling-Katz, M. R., Merrick, H. F., Fink, J. K., et al.: Carbamazepine in Fabry’s disease: effective analgesia with dose-dependent exacerbation of autonomic dysfunction. Neurology 39:598, 1989. Font, R. L., and Fine, B. S.: Ocular pathology in Fabry’s disease: histochemical and electron microscopic observations. Am. J. Ophthalmol. 73:419, 1972. Francois, J., Hanssens, M., and Teuchy, H.: Corneal ultrastructural changes in Fabry’s disease. Ophthalmologica 176:313, 1978. Hashimoto K., Gross, B. G., and Lever, W. F.: Angiokeratoma corporis diffusum (Fabry): histochemical and electron microscopic studies of the skin. J. Invest. Dermatol. 44:119, 1965. Imperial, R., and Helwig, E. G.: Angiokeratoma, a clinicopathological study. Arch. Dermatol. 95:166, 1967. Kleijer, W. J., Hussaarts-Odijk, L. M., Sachs, E. S., et al.: Prenatal diagnosis of Fabry’s disease by direct analysis of chorionic villi. Prenat. Diagn. 7:283, 1987. Kocen, R. S., and Thomas, P. K.: Peripheral nerve involvement in Fabry’s disease. Arch. Neurol. 22:81, 1970. Kusiak, J. W., Quirk, J. M., and Brady, R. O.: Purification and properties of the two major isozymes of alpha-galactosidase from human placenta. J. Biol. Chem. 253:184, 1978. Larue, F., Colleau, S. M., Brasseur, L., and Cleeland, C. S.: Multicentre study of cancer pain and its treatment in France. BMJ 310:1034, 1995. Lehner, T., and Adams, C. W.: Lipid histochemistry of Fabry’s disease. J. Pathol. Bacteriol. 95:411, 1968. Linhart, A., Lubanda, J. C., Palecek, T., et al.: Cardiac manifestations in Fabry disease. J. Inherit. Metab. Dis. 24:75; discussion 65, 2001. Lockman, L. A., Hunninghake, D. B., Krivit, W., and Desnick, R. J.: Relief of pain of Fabry’s disease by diphenylhydantoin. Neurology 23:871, 1973. Medin, J. A., Tudor, M., Simovitch, R., et al.: Correction in trans for Fabry disease: expression, secretion and uptake of alpha-galactosidase A in patient-derived cells driven by a high-titer recombinant retroviral vector. Proc. Natl. Acad. Sci. U. S. A. 93:7917, 1996. Mehta, J., Tuna, N., Moller, J. H., and Desnick, R. J.: Electrocardiographic and vectorcardiographic observations in Fabry’s disease. Adv. Cardiol. 21:220, 1978.

1904

Diseases of the Peripheral Nervous System

37. Moore, D. F., Altarescu, G., Ling, G. S., et al.: Elevated cerebral blood flow velocities in Fabry disease with reversal after enzyme replacement. Stroke 33:525, 2002. 38. Moore, D. F., Scott, L. T., Gladwin, M. T., et al.: Regional cerebral hyperperfusion and nitric oxide pathway dysregulation in Fabry disease: reversal by enzyme replacement therapy. Circulation 104:1506, 2001. 39. Ohshima, T., Schiffmann, R., Murray, G. J., et al.: Aging accentuates and bone marrow transplantation ameliorates metabolic defects in Fabry disease mice. Proc. Natl. Acad. Sci. U. S. A. 96:6423, 1999. 40. Onishi, A., and Dyck, P. J.: Loss of small peripheral sensory neurons in Fabry disease: histologic and morphometric evaluation of cutaneous nerves, spinal ganglia, and posterior columns. Arch. Neurol. 31:120, 1974. 41. Opitz, J. M., Stiles, F. C., Wise, D., et al.: The genetics of angiokeratoma corporis diffusum (Fabry’s disease), and its linkage with Xg(a) locus. Am. J. Hum. Genet. 17:325, 1965. 42. Rae, A. I., Lee, J. C, and Hopper, J. Jr.: Clinical and electron microscopic studies of a case of glycolipid lipoidosis (Fabry’s disease). J. Clin. Pathol. 20:21, 1967. 43. Rosenberg, D. M., Ferrans, V. J., Fulmer, J. D., et al.: Chronic airflow obstruction in Fabry’s disease. Am. J. Med. 68:898, 1980. 44. Rowe, J. W., and Caralis, D. G.: Accelerated atrioventricular conduction in Fabry’s disease: a case report. Angiology 29:562, 1978. 45. Sagebiel, R. W., and Parker, F.: Cutaneous lesions of Fabry’s disease: glycolipid lipidosis; light and electron microscopic findings. J. Invest. Dermatol. 50:208, 1968. 46. Schiffmann, R., Kopp, J. B., Austin, H. A. 3rd, et al.: Enzyme replacement therapy in Fabry disease: a randomized controlled trial. JAMA 285:2743, 2001. 47. Schiffmann, R., Murray, G. J., Treco, D., et al.: Infusion of alpha-galactosidase A reduces tissue globotriaosylceramide storage in patients with Fabry disease. Proc. Natl. Acad. Sci. U. S. A. 97:365, 2000. 48. Scott, L. J., Griffin, J. W., Luciano, C., et al.: Quantitative analysis of epidermal innervation in Fabry disease. Neurology 52:1249, 1999.

49. Sher, N. A., Letson, R. D., and Desnick, R. J.: The ocular manifestations in Fabry’s disease. Arch. Ophthalmol. 97:671, 1979. 50. Sima, A. A., and Robertson, D. M.: Involvement of peripheral nerve and muscle in Fabry’s disease: histologic, ultrastructural, and morphometric studies. Arch. Neurol. 35:291, 1978. 51. Spence, M. W., Goldbloom, A. L., Burgess, J. K., et al.: Heterozygote detection in angiokeratoma corporis diffusum (Anderson-Fabry disease): studies on plasma, leucocytes, and hair follicles. J. Med. Genet. 14:91, 1977. 52. Sweeley, C. C., and Klionsky, B. L.: Fabry’s disease: classification as a sphingolipidosis and partial characterization of a novel glycolipid. J. Biol. Chem. 238:PC3148, 1963. 53. Syed, N. A., Sandbrink, F., Luciano, C. A., et al.: Cutaneous silent periods in patients with Fabry disease. Muscle Nerve 23:1179, 2000. 54. Takenaka, T., Hendrickson, C. S., Tworek, D. M., et al.: Enzymatic and functional correction along with long-term enzyme secretion from transduced bone marrow hematopoietic stem/progenitor and stromal cells derived from patients with Fabry disease. Exp. Hematol. 27:1149, 1999. 55. Takenaka, T., Murray, G. J., Qin, G., et al.: Long-term enzyme correction and lipid reduction in multiple organs of primary and secondary transplanted Fabry mice receiving transduced bone marrow cells. Proc. Natl. Acad. Sci. U. S. A. 97:7515, 2000. 56. Takenaka, T., Qin, G., Brady, R. O., and Medin, J. A.: Circulating alpha-galactosidase A derived from transduced bone marrow cells: relevance for corrective gene transfer for Fabry disease. Hum. Gene. Ther. 10:1931, 1999. 57. Thadhani, R., Wolf, M., West, M. L., et al.: Patients with Fabry disease on dialysis in the United States. Kidney Int. 61:249, 2002. 58. Tome, F. M., Fardeau, M., and Lenoir, G.: Ultrastructure of muscle and sensory nerve in Fabry’s disease. Acta Neuropathol. (Berl.). 38:187, 1977. 59. Ullman, M. D., and McCluer, R. H.: Quantitative analysis of plasma neutral glycosphingolipids by high performance liquid chromatography of their perbenzoyl derivatives. J. Lipid Res. 18:371, 1977. 60. Wise, D., Wallace, H. J., and Jellinek, W. F.: Angiokeratoma corporis diffusum: a clinical study of eight families. Q. J. Med. 31:177, 1962.

82 Tangier Disease and Neuropathy GILMORE N. O’NEILL AND MASON W. FREEMAN

TD Lipoprotein Metabolism HDL and Cholesterol Transport in Plasma Epidemiology Clinical Manifestations and Classification Neuropathy Extraneural TD Familial HDL Deficiency (FHA) Diagnosis

Differential Diagnosis Pathology Pathogenesis Alterations of Cholesterol Homeostasis in TD and Familial HDL Deficiency (FHA) ABCA1—Genetics and Cell Biology Genetics

Tangier disease (TD) is a rare, autosomal recessively inherited, multisystem disease, characterized neurologically by a primary axonal degeneration affecting unmyelinated and small myelinated fibers.26 TD was first described when unusual inclusions were seen in a tonsillectomy specimen taken from a boy living on Tangier Island (Virginia) in the Chesapeake Bay in 1960.31 These inclusions were recognized as sequestered cholesterol esters in macrophages. The tonsillar enlargement occasioned by the cholesterol deposition was quickly linked to an extremely low level of serum highdensity lipoprotein (HDL) cholesterol, and this dyad serves as the classic clinical marker of the disease.6 Almost 40 years after the original phenotypic description of the disorder, the molecular genetic defect was uncovered when several groups reported that TD patients harbored mutations in both alleles encoding a lipid transporter, now called ATP-binding cassette A1 (ABCA1).8,11,14,54,85 A related lipid disorder, familial HDL deficiency (also called familial hypoalphalipoproteinemia [FHA]) is seen in individuals who have one normal and one mutant allele encoding ABCA1.11,44,58

TD LIPOPROTEIN METABOLISM The hydrophobic plasma lipids are rendered soluble in the blood stream because of their association with proteins and more polar lipid moieties. The resulting spherical,

Expression Intracellular ABCA1 Trafficking ABCA1 Regulation Additional ABCA1 Functions Knockout Mouse Models Human Mutations Speculation on Neuropathophysiology Treatment

proteolipid macromolecular complexes, called plasma lipoproteins, are responsible for transporting lipids from their sites of synthesis to their sites of utilization via the aqueous environment of the blood. The major components of lipoproteins are cholesterol, phospholipid, triglyceride, and apolipoproteins. It is the proportions of these lipids, as well as the specific apoprotein components, that define each class of lipoprotein. These different lipoprotein species are in dynamic equilibrium, interacting with one another through the active, enzyme-catalyzed exchange of apoproteins and lipids. Because they are hydrophobic, esterified cholesterol and triglyceride reside in the core of the particles, whereas the detergent-like (amphipathic) unesterified cholesterol and phospholipid cover the surface, at an interface between the hydrophobic core and blood. Apolipoproteins also occupy the particle surfaces, acting both as a structural component of the lipoproteins and as a targeting sequence that directs the lipoprotein to the appropriate interaction with a cellular receptor or lipid enzyme. Apolipoproteins and unesterified cholesterol have significant water solubility and can be exchanged between lipoproteins and other lipid-laden surfaces. Triglyceride and nonpolar esterified cholesterol, however, can only move across such interfaces by the action of exchange proteins. Historically, lipoproteins have been subclassified according to the laboratory technique by which they were separated. Density gradient ultracentrifugation produced 1905

1906

Diseases of the Peripheral Nervous System

the names by which we now commonly refer to most of the lipoprotein classes (e.g., low-density lipoprotein [LDL] and HDL), but other laboratory methods have contributed to a sometimes confusing nomenclature, particularly in the older medical literature. Lipoprotein mobility, as determined by paper electrophoresis, led to the terms ␣-lipoprotein, ␤-lipoprotein, and pre-␤-lipoprotein, and these terms are still commonly associated with clinical lipid disorders. As molecular biologic techniques have uncovered the genetic causes of these disorders, the terminology linking the older and newer literature can often be obscure. Table 82–1 provides a summary of the density, electrophoretic mobility, lipid content, and apolipoprotein composition of the serum lipoproteins. The lipid core and surface protein composition influence the density of these particles (higher density particles have a greater percentage of protein and lower percentage of lipid). Density is inversely related to the size of the lipoprotein.

HDL AND CHOLESTEROL TRANSPORT IN PLASMA The genetic defect in TD typically affects all of the serum lipoprotein levels, but HDL is most profoundly influenced by the ABCA1 mutation. HDL is composed of protein (50%), cholesterol (20%), phospholipid (25%), and triglyceride (5%). Apolipoproteins A-I and A-II make up 90% of normal HDL protein content in a ratio of 3:1 (A-I:A-II). HDLs range in size from 70 to 100 Å and in density from 1.063 to 1.21 g/mL. They have been subclassified by analytic ultracentrifugation into HDL2 (density 1.063 to 1.12 g/mL) and HDL3 (density 1.12 to 1.21 g/mL), with the latter being considered lipid-poor HDL and the former lipid-rich HDL. Apolipoprotein A-I (apo A-I) and apo A-II are the main structural apoproteins of HDL and are abundant in plasma. They are single polypeptide chains synthesized predominantly in the liver. A lesser protein component of HDL, apo A-IV, is synthesized in the small intestine, where some apo A-I is also made. The apoprotein content of HDL can be heterogeneous, with some particles composed of four apo A-I molecules, while others can have two molecules each of apo A-I and apo A-II. A third category of HDL is enriched with apolipoprotein E and actually floats in the density range lighter that HDL2. HDL appears to start as a phospholipid bilayer with some free cholesterol and a coating of apo A-I. This may arise by cellular secretion from the liver or intestine and perhaps nonenteric cells as well. Once in the circulation, it can acquire other lipid and protein components from the lighter lipoproteins, especially very-low-density lipoproteins (VLDLs) and chylomicrons. It is the fusion of apo A-I–phospholipid complexes and cholesterol with other

phospholipid vesicles containing apo A-II and/or apo A-IV that results in the assembly of the various HDL particle species. This cholesterol-poor HDL3 plays the key role in initiating cholesterol efflux/transfer from peripheral tissues via the ABCA1 transporter. HDL2 is derived from HDL3 through the accumulation of cholesterol and cholesterol ester, via the action of lecithin:cholesterol acyltransferase (LCAT). When unesterified or free cholesterol partitions into a nascent HDL particle, LCAT can esterify this molecule, resulting in the generation of the significantly more hydrophobic cholesteryl ester. This effectively traps the cholesterol in the interior of the enlarging HDL particle, because the movement of a cholesteryl ester molecule back across the plasma interface to the originating cell membrane is energetically unfavorable. The ester, which is predominantly oleic acid, is derived from lecithin, a phospholipid. Apo A-I serves as a cofactor for LCAT activity. As cholesteryl ester accumulates and moves into the lipophilic core, the HDL3 particles grow larger and accommodate C apoproteins and, in a subgroup, apoprotein E. This enrichment produces HDL2 particles, which can subsequently transfer cholesteryl esters, through the action of cholesteryl ester transfer protein, to apo B–containing lipoproteins (e.g., LDL, VLDL). These lower density lipoproteins can then transport the cholesteryl ester into cells via binding to the cellular receptors that recognize them (i.e., the LDL receptor or LDL receptor–related protein). Alternatively, an HDL receptor called scavenger receptor-B1 (SR-B1) can selectively transfer the cholesterol out of the HDL particle with return of a de-lipidated HDL protein to the circulation.95 The pathway of cholesterol migration back from the peripheral cell to the liver, for secretion and potential excretion in the bile, is termed the reverse cholesterol transport pathway. In TD, the defect in the reverse transport pathway occurs at the step in which cholesterol is transferred to the lipidpoor HDL.69 Lipid-depleted HDLs appear to be rapidly catabolized and cleared by the kidney, perhaps accounting, at least in part, for the extremely low HDL levels seen in TD patients. The reason that TD patients usually also have very low LDL cholesterol levels as well as high serum triglyceride values may well be a secondary consequence of the near-absence of circulating HDL, but the molecular explanations for these apparently downstream effects remain unknown. The predilection for cholesterol accumulation in tissues harboring high levels of macrophages (e.g., tonsils, spleen, liver, bone marrow) is thought to be due to the macrophage expression of scavenger receptors, such as SR-A and CD36, which, unlike the LDL receptor, are not downregulated with intracellular cholesterol accumulation. Thus the macrophage appears to rely heavily on its ability to efflux cholesterol as a mechanism for avoiding intracellular cholesterol accumulation. Cells that do not take up large quantities of lipid through scavenger receptors appear to be

0.93 0.93–1.006 1.006–1.019 1.029–1.063 1.063–1.125 1.125–1.210 1.28

Density

Origin Pre-␤ Slow pre-␤ ␤ ␣ ␣ Pre-␤

Electrophoretic Mobility 2 7 9 8 5 4 5

Surface Cholesterol 7 18 19 22 33 25 5

Surface Phospholipid 2 8 19 22 40 55 90

Surface Apolipoprotein 86 55 23 6 5 3 0

Core Triglyceride

Lipid Content (% Dry Mass)

3 12 29 42 17 13 0

Core Cholesteryl Ester

*HDL ⫽ high-density lipoprotein; IDL ⫽ intermediate-density lipoprotein; LDL ⫽ low-density lipoprotein; VLDL ⫽ very-low-density lipoprotein.

Chylomicron VLDL IDL LDL HDL2 HDL3 Pre-␤1-HDL

Class*

Table 82–1. Lipoproteins

A-I, A-IV, B-48 B-100, C-I, C-II, C-III, E B-100, C-I, C-II, C-III, E B-100 A-I, A-II, C-I, C-II, C-III, E A-I, A-II, C-I, C-II, C-III, E A-I, A-II, C-I, C-II, C-III, D, E

Apolipoprotein Types

1908

Diseases of the Peripheral Nervous System

less dependent on cholesterol efflux as a control mechanism and consequently do not accumulate as much lipid in the absence of normal ABCA1 transport function.

EPIDEMIOLOGY TD is a rare disorder, with fewer than 100 cases described in the literature.68,74,96 Approximately 50% of TD homozygotes will develop some form of peripheral neuropathy (see below).5,94 Additionally, homozygotes experience a sixfold increase in risk for the development of cardiovascular disease in middle and old age.89,94

CLINICAL MANIFESTATIONS AND CLASSIFICATION Neuropathy Tangier neuropathy has been subclassified into three subtypes.32 The first is an asymmetrical, relapsing-remitting sensorimotor mononeuropathy, usually associated with normal nerve conduction velocities. Occasionally, distal latencies may be prolonged. These mononeuropathies may involve cranial nerves and nerves in the upper or lower extremities and manifest as sensory loss, paresthesias, compromised deep tendon reflexes (diminished or absent), and extraocular movement abnormalities.24,26,75 The second subtype is a symmetrical, slowly progressive polyneuropathy that affects predominantly the lower extremities. Normal nerve conduction studies reflect the predominant pathologic involvement of small nerve fibers.24,36,50 There has been one report of an acute-onset polyneuropathy.25 The third subtype is a slowly progressive neuropathy that predominantly affects the face and upper limbs with weakness (distal hand weakness and wasting, facial diplegia) and dissociative sensory loss. Pain and temperature loss occurs early. All sensory modalities are compromised late in the disease. Sensory loss may eventually involve the entire trunk and lower extremities. Tendon reflexes tend to be preserved. This “pseudosyrinx” phenotype is the rarest and most severe Tangier neuropathy.4,22,32,49,75 Magnetic resonance imaging of one pseudosyrinx TD patient has demonstrated cervical cord atrophy.73 Despite the predominance of small fiber involvement in the different TD neuropathies, orthostatic hypotension has not been reported, and the morphometry of intermediolateral column neuron cell bodies in one TD autopsy was normal.56 The course of the neuropathy may be benign or severely debilitating. Different clinical phenotypes have certain neuropathologic distinctions, with the demonstration of a demyelinating-remyelinating pathology associated with the mononeuritic type and

axonal degeneration with the syringomyelic variant.74 These differences may represent a genetic heterogeneity among the various subtypes.

Extraneural TD Cholesterol also accumulates in extraneural tissues, including tonsils, spleen, liver, lymph nodes, skin fibroblasts, macrophages, vascular smooth muscle, bone marrow, and cornea.26,60 About half of all TD patients develop premature cardio- and cerebrovascular disease, although the accelerated, very-early-onset atherosclerosis of familial hypercholesterolemia is not seen.89,94 Estimates of the increase in cardiovascular disease risk are imprecise because of the small number of individuals affected by the disorder, but a four- to sixfold elevation in risk has been suggested.94 Corneal clouding may be the result of increased free cholesterol and phospholipid deposition in the corneal stroma or exposure keratopathy arising from facial diplegia and eye closure weakness.76,107 Hemolytic anemia is observed in association with erythrocyte membrane abnormalities, explained by the reduced cholesterol:phosphatidylcholine ratio.78 Lipid accumulation in other organs does not appear to cause clinical problems, although both hepatomegaly and splenomegaly may be detected.

Familial HDL Deficiency (FHA) Whereas the clinical (neuropathy) phenotype of TD is inherited in an autosomal recessive mode, the reduction in serum lipid and apoprotein levels is transmitted in an autosomal codominant mode. This is manifested by a 50% reduction of HDL cholesterol, apo A-I, and apo A-II in individuals who are obligate heterozygotes for TD mutations (e.g., the parents of a TD patient).5,13,39 These individuals do not have any overt clinical or pathologic abnormalities.88 However, most FHA patients have demonstrable abnormalities of cholesterol efflux,13,23,59 and recent data suggest that some have an increased risk for cardiovascular disease.19,94,100

DIAGNOSIS The diagnostic hallmark of TD is absence or severe deficiency of HDL cholesterol (⬍5 mg/dL) and apo A-I (⬍5 mg/dL) in plasma. The deficiency is associated with low or normal cholesterol levels (usually ⬍150 mg/dL), reduced LDL levels (approximately 40% of normal), elevated triglyceride (⬎300 mg/dL), and increased triglyceride-rich VLDL and abnormal chylomicron remnants.6 Lipoprotein electrophoresis fails to detect any ␣-mobile lipoprotein, and immunoblotting localizes apo A-I in the pre-␤-HDL fraction.

1909

Tangier Disease and Neuropathy

Normal total serum cholesterol levels do not, however, exclude TD.75 The most robust diagnostic method to determine if a patient with an undiagnosed multifocal neuropathy or pseudosyringomyelia has TD is to measure the serum HDL cholesterol level. The absence of HDL in the serum is a consequence of the specific TD genetic defect, because markedly reduced HDL levels are not seen in other disorders of intracellular lipid metabolism such as Wolman’s disease, Niemann-Pick disease type C, or Gaucher’s disease. Because there is no pathognomonic blood chemistry abnormality in individuals affected with TD, TD heterozygotes (carriers) cannot easily be distinguished by routine blood tests from individuals in the general population who have reduced HDL cholesterol levels from other causes. Obligate heterozygotes are biochemically characterized by serum apo A-I, apo A-II, and HDL cholesterol levels at 50% of normal.5,88 A family pattern of low HDL cholesterol levels consistent with an autosomal codominant inheritance of this trait would strengthen the likelihood that a mutation in the TD gene is present in the family.

DIFFERENTIAL DIAGNOSIS Tangier neuropathy should be considered in a patient presenting with a relapsing-remitting mononeuropathy, a length-dependent small fiber sensory neuropathy, or pseudosyringobulbia with oculomotor palsies. An extremely low or absent HDL cholesterol level, in association with one of these neurologic presentations, makes the diagnosis of TD likely. It should be emphasized, however, that the evaluation of a HDL-deficient neuropathy should exclude potentially reversible etiologies, such as lymphoproliferative disorders. While other inherited and acquired dyslipidemias can produce severe reductions of HDL cholesterol and apo A-I (Table 82–2), they are not associated with similar neurologic findings. Familial LCAT deficiency (corneal opacities, anemia, and proteinuria) and fish eye disease (dense corneal opacities) have clinical phenotypes that readily distinguish them from TD. Individuals with genetic mutations resulting in little or no apo A-I production (apo A-I deficiency, apo A-I/C-III deficiency, apo A-I/C-III/A-IV deficiency) also typically have denser corneal opacities and lack the neuropathy and organ enlargements seen in TD patients. The major distinguishing laboratory characteristic in TD that aids in differentiating it from other genetically inherited low HDL disorders is the low serum total cholesterol level, typically in association with a moderate serum triglyceride elevation. Because low HDL levels can develop following acute or chronic infections, as well as other chronic inflammatory and hematologic disorders (dysglobulinemias, lymphoproliferative disorders, and lymphohistiocytosis),1,3,35,47,51,83 a patient with a neuropathy who also acquires a HDL-lowering

disorder can present a diagnostic challenge. Recovery from the inflammatory condition or measurements of HDL levels in family members would likely establish the lack of heritability of the patient’s HDL disorder. Heterozygotes for ABCA1 mutations also have low HDL cholesterol levels, but they do not develop neuropathy as a consequence of their heterozygous mutation. Other disorders of cholesterol storage associated with neurologic disease (Gaucher’s disease, Niemann-Pick disease type C, and Wolman’s disease [lysosomal acid lipase deficiency]) are distinguished from TD by much higher HDL cholesterol levels. More importantly, the clinical presentations of these disorders differ from that of TD.

PATHOLOGY Ultrastructural examinations of radial and sural nerve biopsies (Fig. 82–1) have shown electron-lucent Schwann cell lipid deposits, endoneurial fibrosis, and loss of small myelinated and unmyelinated fibers.48 Vasa nervorum demonstrate a nonspecific duplication of the basement membrane. Whereas Schwann cell lipid deposits are seen in a number of disorders, including Wolman’s disease, Tay-Sachs disease, Niemann-Pick disease, metachromatic leukodystrophy, Fabry’s disease, Refsum’s disease, and Krabbe’s disease, the deposits seen in TD predominate in Schwann cells rather than neurons and are extralysomal (Fig. 82–2). Neuronal and macrophage lipid sequestration has not been demonstrated in the central nervous system in two TD autopsies (Ferrans and Fredrickson26; unpublished observation), but abnormal lipid accumulation has been demonstrated in neuron cell bodies in other reports.74,90 Clinicopathologic correlation has been demonstrated. Demyelination-remyelination morphology has been associated with relapsing-remitting mononeuropathy, clustering of affected internodes along individual fibers, and limitation of lipid inclusions to the Remak cells. Axonal degeneration with a more generalized distribution of lipid droplets has been described in association with the symmetrical polyneuropathy subtype.74,75 Necropsy of a pseudosyrinx TD patient described motor neuron loss in the cervical spinal cord and facial nucleus.4 Neuropathologically, the disease is characterized by linear osmophilic droplets seen along nerve fibers in teased fiber preparations, with lipidladen vacuoles in Remak cells, Schwann cells, perineurial cells, and the endothelial cells of vasa nervorum. Cholesteryl ester deposits are seen in the macrophage components of tonsils and adenoidal tissues, spleen, lymph nodes, liver, bile ducts, intestinal mucosa, bone marrow, blood vessel smooth muscle, and adventitia. Ultrastructural examination of Tangier macrophages and fibroblasts has demonstrated morphologic abnormalities of the Golgi apparatus with hyperplasia and dilated cisterns.82

⫹ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹/⫺ ⫹/⫺ ⫹/⫺

Neuropathy ⫹ ⫹/⫺ ⫺ ⫺ ⫹ ⫹/⫺ ⫺ ⫺ ⫺ ⫺ ⫹/⫺ ⫹/⫺ ⫹/⫺ ⫹/⫺ ⫹/⫺ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹

⫹

⫺ ⫹

⫺ ⫺

⫹

Extraneural Tissue Involvement

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

⫺

⫹⫹⫹ ⫺

⫹⫹⫹ ⫹⫹⫹

⫹

Corneal Opacity

Normal Moderately decreased

Normal Normal

Decreased

Total Cholesterol

Apo A-I ⫽ apolipoprotein A-I; HDL ⫽ high-density lipoprotein; LCAT ⫽ lecithin:cholesterol acyltransferase.

Niemann-Pick disease type C Wolman’s disease Gaucher’s disease Liver outlet obstruction Hepatic disease Malnutrition/enteropathy Dysglobulinemia Severe inflammation Hodgkin’s disease Lymphohistiocytosis

Apo A-I deficiency Familial LCAT deficiency Fish eye disease Abetalipoproteinemia

Tangier disease

Disease

Tonsillar Abnormality

Table 82–2. Clinical Phenotypes

Moderately decreased Moderately decreased

Markedly decreased Moderately decreased

Markedly decreased

HDL Cholesterol

Moderately decreased Moderately decreased

Markedly decreased Moderately decreased

Markedly decreased

Apo A-I

Increased Markedly decreased

Moderately increased Increased Increased

Triglycerides

Tangier Disease and Neuropathy

1911

FIGURE 82–1 Electron micrograph of transverse section of sural nerve of a patient with Tangier disease. Note the presence of clear vacuoles in the Schwann cell cytoplasm, and a myelin void (lower left). There are multiple vacuoles in the Schwann cell (inset), which nevertheless retains two intact unmyelinated fibers. (From Dyck, P. J., Ellefson, R. D., Yao, J. K., et al.: Adult-onset Tangier disease. I. Morphometric and pathologic studies suggesting delayed degradation of neutral lipids after fiber degeneration. J. Neuropathol. Exp. Neurol. 37:119, 1978, with permission.)

PATHOGENESIS Cellular cholesterol homeostasis is maintained in most cells by balancing the efflux of sterol against its input from de novo cholesterol biosynthesis and the uptake of cholesterol-containing ligands. The latter process in humans is chiefly controlled by the expression of LDL receptors,15 although other lipoprotein receptors, such as SR-B1, can also contribute to cholesterol uptake. Endogenous cholesterol synthesis takes place in the mitochondria and cytosol via a number of intermediates derived from acetyl-coenzyme A (CoA). The rate-limiting step is catalyzed by ␤-hydroxy-␤-methylglutaryl-CoA reductase. Cellular cholesterol effluxes at a rate of 0.1% of total cell cholesterol per minute. Intracellular cholesterol exists in an equilibrium between free cholesterol and cholesterol ester. Cell membrane free cholesterol is predominantly represented in the inner leaflet of the plasmalemma.

In the current model for ABCA1-mediated cholesterol vesicular efflux, lipid-poor HDL or free apo A-I binds to cells expressing ABCA1.28,70,72,98,105 The binding is in close proximity to ABCA1, but it has not been established if the transporter is itself in direct contact with the apolipoprotein. Binding appears to activate phosphatidylinositol-specific phospholipase C, phosphatidylcholine-specific phospholipases C and D, and protein kinase C, which, acting as a second messenger,21,64,103 induces a Golgi-dependent process that generates cholesterol- and phospholipid-enriched vesicles that move toward and fuse with the inner leaflet of the cell membrane.62,80 Whereas cholesterol can move freely between the inner and outer cell membrane leaflets, phospholipids are actively “flipped,” generating irregular protrusions on the outer membrane that can then interact with the lipid-poor apo A-I. The “flippase” activity of ABCA1 is inferred based on its structural similarity to P-glycoprotein, a well-characterized ABC transporter that can translocate lipids from the inner leaflet to the outer leaflet. Translocated

1912

Diseases of the Peripheral Nervous System

receptor pathways that are not repressed by intracellular overaccumulation of cholesterol.46 If their efflux mechanism is impaired, as in TD, then these cells and the tissues that most abundantly harbor them are at risk of accumulating excess intracellular cholesterol. It is evident that this efflux defect caused by mutations in ABCA1 specifically influences the serum HDL levels, because such low levels of HDL are not seen in other disorders of intracellular lipid metabolism, such as Wolman’s disease, Niemann-Pick disease type C, or Gaucher’s disease.6 How the defect in cholesterol efflux results in neurologic injury is unknown, but some speculations are offered below.

Alterations of Cholesterol Homeostasis in TD and Familial HDL Deficiency (FHA)

FIGURE 82–2 Electron micrograph of an oblique section of a small nerve from the skin of a patient with Tangier disease. Note the endoneurial collagenization and vacuoles (lipid deposits) in Schwann cells. A few axons contain accumulations of glycogen granules; others have small lipid droplets. There is thickening of basement membranes of Schwann and perineurial cells. (⫻5000.) (From Ferrans, V. J., and Frederickson, D. S.: The pathology of Tangier disease: a light and electron microscopic study. Am. J. Pathol. 78:101, 1975, with permission.)

Macrophages and fibroblasts from TD and FHA patients demonstrate a significant defect in lipid-poor apo A-I–stimulated cholesterol and phospholipid efflux.13,30,84,102 As a result, cholesteryl esters accumulate in membrane free droplets. Whereas phospholipid efflux is abnormal, total cell phospholipid content is not significantly altered. However, the cellular cardiolipin and lysocardiolipin fractions are three- to fivefold enriched.29 There is some evidence that HDL is internalized after cell surface binding and is resecreted in a lipid-enriched state.91,99 It has been reported that, in cells from TD patients, cell binding of HDL is increased with a significant decrease in HDL resecretion.91 While other workers have failed to duplicate this observation, recent work on internalization of HDL by the scavenger receptor SR-B1 suggests that lipid exchange involving HDL might indeed be occurring inside intracellular vesicles.95 Abnormal hypercatabolism of mature apo A-I but not pro-apo A-I has been demonstrated in both TD and FHA patients,7,88 thus implying that the deficiency is secondary to the cellular defect in cholesterol efflux. Abnormal dilatation of the Golgi cisternae speaks to a disruption of intracellular lipid trafficking through vesicular budding, with abnormal ceramide and glycerophospholipid turnover.71,92

ABCA1—GENETICS AND CELL BIOLOGY phospholipid and cholesterol interact with the lipid-poor apo A-I (HDL3) to produce nascent HDL2 particles.106 The fate of this HDL is discussed above (see HDL and Cholesterol Transport in Blood). HDL serves to prevent cholesterol overload of macrophages (and perhaps fibroblasts and smooth muscle cells) by facilitating the efflux of cholesterol into the extracellular compartment. These cells are not fully protected by the cellular cholesterol homeostasis mechanisms operative in other tissues that rely on the inhibition of cholesterol biosynthesis and LDL receptor expression. Thus these cells internalize cholesterol-rich lipoproteins via scavenger

Genetics Using a graphic linkage exclusion strategy (homozygosity screening) in one large informative pedigree, combined with conventional linkage analysis in three smaller families, Rust et al. linked the TD defect to an 8-cM locus in 9q31.86 Using two additional families with a highly informative recombination, they were able to refine the locus to 1 cM, and subsequently localized ABCA1 to this candidate locus. The candidacy of this gene was reinforced by its expression in an expressed sequence tags database restricted to foam cell–specific messenger RNAs (mRNAs). Mutation analysis

1913

Tangier Disease and Neuropathy

of ABCA1 from affected families revealed mutations that co-segregated with the disease.8,11,14,54,85 ABCA1 mutations were also detected in heterozygote FHA pedigrees.11 Thus TD and FHA are associated with mutations in the same gene. In addition to the strong genetic evidence linking mutations in ABCA1 to TD, other evidence confirms the role of ABCA1 in cholesterol efflux and TD. This evidence includes the following findings: 1. ABCA1 mRNA expression levels are regulated by intracellular cholesterol content.52,54 2. Transfection studies with antisense primers that inhibit the production of ABCA1 result in decreased cholesterol efflux.11,54 3. Genes encoding wild-type but not mutant forms of ABCA1 confer enhanced cholesterol efflux on cells transfected with those genes.28 4. Targeted ablation of the murine ABCA1 gene orthologue leads to the accumulation of lipid-laden macrophages in the mice, as well as reduced serum cholesterol levels, complete lack of HDL cholesterol and apo A-I, and reduced LDL cholesterol.61 Naturally occurring human mutations (missense, nonsense, and deletions) are scattered throughout the ABCA1 gene (Table 82–3).2,8,11,14,18,34,41–45,54,58,67,79,85,96,104,108 Mutational hot spots lie in or close to the ABCs and the amino terminus of the protein (Fig. 82–3). Many laboratories are currently investigating possible associations of polymorphisms in ABCA1 with variations in plasma HDL levels and cardiovascular disease risk.12,18,19,96,104

Expression ABCA1 is a member of a large protein family described as the ATP-binding cassette transporter family.33,40 This family consists of membrane proteins that cleave ATP to transport substrates between cellular compartments. All ABC transporters share homology for a nucleotide-binding domain flanked by two highly conserved motifs called Walker A and B. Functional transporters appear to require two membrane-spanning domains and two nucleotide-binding domains. ABCA1 encodes these four domains in a single peptide chain having 12 membrane-spanning domains (see Fig. 82–3). In ABCA1 there is a large regulatory domain that lies between the two six-membrane-spanning halves of the protein. Human ABCA1 contains 50 exons that range in size from 249 to 33 bp and span 70 kb.53,79,87 As originally described, the ABCA1 complementary DNA sequence encoded a 2201–amino acid peptide having an intracellular amino terminus.52,57 More recently, however, an upstream, in-frame, alternate translation initiation codon has been described.77 This initiator adds 60 amino acids to the peptide, which have been demonstrated to be critical

to stable expression of the protein and to position the first large N-terminal loop in the extracellular space (see Fig. 82–3).27 ABCA1 is expressed at low levels in all tissues, including brain and lymph node. It is preferentially expressed in liver, adrenals, and uterus.10,52

Intracellular ABCA1 Trafficking Immunocytochemistry has localized ABCA1 to the plasmalemma. Expression studies have shown that an ABCA1GFP fusion protein localizes to the cell membrane surface and also to early and late endosomes and lysosomes.27,65 These data suggest that lysosomal degradation of ABCA1 may regulate its plasmalemmal expression. ABCA1 does not appear to be associated with cholesterol- and sphingolipidrich raft subdomains of the plasmalemma, suggesting that lipid efflux to apo A-I does not preferentially take place within these specialized microdomains.63

ABCA1 Regulation ABCA1 function is highly regulated, and gene expression is influenced, at the transcriptional level, by a number of molecules. These include cyclic AMP (cAMP), interferon-␥, peroxisome proliferator–activated receptor (PPAR) isoform agonists, sterols, and cis-retinoic acid. Quiescent cells increase ABCA1 expression commensurate with the observation that the cholesterol requirement for nonproliferating cells is low.54 cAMP upregulates ABCA1 mRNA expression, coincident with increased ABCA1 protein expression and apo A-I–mediated cholesterol efflux, in certain cell types.9,54,55,70,97 ABCA1 mRNA expression can be increased by cholesterol loading (through LDL or cholesterol incubation)52,54 and decreased by cholesterol depletion.54 Oxysterol agonists, including 22-(R)-hydroxycholesterol, upregulate ABCA1 expression through the liver X receptor (LXR) nuclear receptor.81,101 Although a PPAR-responsive element has not been found in the ABCA1 promoter, activation of PPAR␥ results in stimulation of LXR␣ which in turn can induce the transcription of ABCA1 mRNA.17 ABCA1 promoter elements include a direct repeat spaced by four nucleotides (DR4) element, SP1, and E-box protein-responsive elements. The promoter binds and responds to LXR␣ and retinoid X receptor nuclear receptors.20,77 A number of polymorphisms have been reported in the region.

Additional ABCA1 Functions In addition to modulating cellular efflux of cholesterol, ABCA1 mediates engulfment of damaged cells by macrophages,37 signals for apoptotic cell engulfment by “flipping” phosphatidylserine to the extracellular lamella, and facilitates secretion of interleukin-1␤.38 The second and third observations may contribute to understanding

1914

Diseases of the Peripheral Nervous System

Table 82–3. Naturally Occurring Mutations in Human ABCA1* Amino Acid Change† P85L R230C A255T 277x 282X D528X R587W W590L W590S Q597R L608X 627X 635X ⌬L693 R909X 913X T929I N935H N935S A937V A1046D M1091T

Reference 42

Hong et al. Wang et al.104 Nishida et al.67 Brousseau et al.13 Altilia et al.2 Rust et al.85 Lawn et al.54 Hong et al.43 Bodzioch et al.8 Brooks-Wilson et al.11; Lawn et al.54 Bodzioch et al.8; Rust et al.85 Singaraja et al.96 Singaraja et al.96 Brooks-Wilson et al.11 Marcil et al.58; Zuchner et al.108 Singaraja et al.96 Clee et al.18 Guo et al.34 Bodzioch et al.8 Bodzioch et al.8 Wang et al.104 Marcil et al.58

Amino Acid Change†

Reference

D1099Y S1115X 1145X 1284X D1289N C1477R S1506L N1611D S1544X R1680W N1800H 1834X R1851X 1887X ⌬1893-4 1919X P2009S R2081W R2144X P2150L 2203X 2145 2215X

Ho Hong et al.41 Remaley et al.79 Singaraja et al.96 Huang et al.44 Brousseau et al.13 Brooks-Wilson et al.11 Singaraja et al.96 Nishida et al.67 Wang et al.104 Ishii et al.45 Brousseau et al.13 Bodzioch et al.8 Nishida et al.67 Lawn et al.54 Marcil et al.58 Lawn et al.54 Ho Hong et al.41 Huang et al.44 Marcil et al.58 Clee et al.18 Clee et al.18 Clee et al.18 Brousseau et al.13

*Neuropathy-associated mutations are in bold type. † The number of the affected amino acid may differ from that used in the original report. Not all mutations have been functionally verified.

the apparent paradox arising from the observation that FHA patients have an increased risk of atheroma but no abnormal accumulation of lipid in extravascular macrophages. Extravascular macrophages accumulate most cholesterol by engulfing dying cells in an ABCA1mediated pathway, whereas arterial lipid plaque-based macrophages acquire cholesterol by scavenger receptor– mediated LDL internalization.68

Knockout Mouse Models Murine ABCA1 knockouts (Abc1-null mice) demonstrate a biochemical and pathologic phenotype similar to that of TD patients.61 A neurologic phenotype has yet to be reported or characterized, and unpublished data conflict.93

Human Mutations Work is in progress to determine if there are any genotype-to-phenotype correlations that explain the predilections for developing neuropathy or cardiovascular disease in TD (reviewed by Singaraja et al.96). Other genetic and environmental modifiers likely contribute to the clinical phenotype. The absence of a clinical phenotype in FHA subjects suggests that a threshold depression (dose effect)

of cholesterol efflux of HDL levels is required to cause disease. To date more than 45 Tangier-associated ABCA1 mutations have been described (see Table 82–3 and Fig. 82–3). These are distributed throughout the gene. Described mutations include open reading frame point mutations, deletions, and insertions; large deletions and insertions; and splice site mutations. Precise mutational hot spots have not yet been identified, although other ABC transporter–related diseases are associated with increased frequency of mutations in the ATP-binding cassettes. Several of the described patients are compound heterozygotes (have two different ABCA1 mutant alleles). Only a small number of mutations have been described in association with a neuropathy (see Table 82–3). A correlation between ABCA1 expression levels, cholesterol efflux, and neuropathy or other extraneural disorders has yet to be definitively determined.

SPECULATION ON NEUROPATHOPHYSIOLOGY TD appears primarily to affect tissues and cell types that import LDL-derived cholesterol through unregulated scavenger receptor pathways. These cells and tissues can

1915

Tangier Disease and Neuropathy

POINT MUTATIONS IN ABCA1 A255T

S1506L R587W

W590S

R230C

S

S C1477R

N1611D

RQ597

N1800H

44

821

1370

1852

26

843

1348

1873

Out

P85L

In N 1

Walker A D1289N

2261 C

R1680W

NBD-1 T929I N935S N935H A937V

Walker A Walker B A1046D

D1099Y

P2150L F2009S

M1091T

NBD-2

Walker B R2081W Walker A1 = 926–942 Walker B1 = 1056–1065

Walker A2 = 1939–1955 Walker B2 = 2068–2077

FIGURE 82–3 The topologic arrangement of the ABCA1 transporter. The salient features are 12 membrane-spanning domains, two prominent extracellular loops, and two nucleotide-binding domains (NBD) associated with their respective conserved sequences, called Walker motifs. The amino acids connoting the beginning (1) and end (2261) of the protein are numbered, as are the residues that approximately initiate several of the key locations in the protein. The topologic arrangement of the two large extracellular loops has been established by Fitzgerald and colleagues,27,28 and the disulfide bond linking the two halves is inferred from data on a related ABC transporter, ABCR.16 Many of the point mutations identified in Tangier and FHA patients are shown mapped onto this transporter topology, but not all have been verified to be functionally relevant. Mutations that truncate the reporter prematurely are listed in Table 82–3.

continue to accumulate cholesterol when cholesterol efflux is severely impaired and when endogenous cholesterol synthesis is shut off. Whereas Schwann cell accumulation of lipids is very prominent, particularly in the demyelinating relapsing-remitting phenotype, one must also consider that abnormalities of membrane vesicular transport may contribute to a primary neuronal degenerative process. It has been said that the unique absence of HDL in TD and the association of neuropathy (central and peripheral) in other disorders of cholesterol trafficking (e.g., Niemann Pick disease type C) points toward cholesterol as the main initiator of disease. The observation that abetalipoproteinemia can associate with ataxia, retinopathy, and spasticity and that apo A-I deficiency is associated with neuropathy, retinopathy, and ataxia might suggest otherwise (note that heterozygotes for apo A-I lack the retinopathy). How the above metabolic derangements impinge on the neuron is unknown. The selectivity of the pathology for certain tissues reflects the dependence of those tissues on

cholesterol efflux for intracellular cholesterol homeostasis. This is not likely to be due to apo A deficiency, because neuropathy is not encountered in selective apo A deficiencies. As an aside, it is intriguing that amyloid fibrils in familial amyloid polyneuropathy type III are derived from a mutant (Gly-25-Arg) form of apo A-I apolipoprotein.66 It must be emphasized, however, that there is no evidence for amyloid protein accumulation in TD.

TREATMENT Specific therapies are not available for Tangier neuropathy. Symptomatic therapies for pain may therefore be indicated. While transgenic and gene reconstitution experiments suggest that replacement of a functional ABCA1 protein in the liver of animals can partially redress the serum lipid abnormalities, it is unknown if such intervention would have any beneficial impact on a neuropathy. Specific gene

1916

Diseases of the Peripheral Nervous System

targeting of ABCA1 to neuronal tissues has not been reported. Recognition of TD as the cause of a neuropathy allows surveillance and early intervention for cardiovascular disease to be undertaken in a patient.

REFERENCES 1. Akerlund, B., Carlson, L. A., and Jarstrand, C.: Dyslipoproteinemia in patients with severe bacterial infections. Scand. J. Infect. Dis. 18:539, 1986. 2. Altilia, S., Pisciotta, L., Garuti, R., et al.: Abnormal splicing of ABCA1 pre-mRNA in Tangier disease due to a IVS2 ⫹5G ⬎C mutation in ABCA1 gene. J. Lipid Res. 44:254, 2003. 3. Alvarez, C., and Ramos, A.: Lipids, lipoproteins, and apoproteins in serum during infection. Clin. Chem. 32:142, 1986. 4. Antoine, J. C., Tommasi, M., Boucheron, S., et al.: Pathology of roots, spinal cord and brainstem in syringomyelia-like syndrome of Tangier disease. J. Neurol. Sci. 106:179, 1991. 5. Assmann, G., Simantke, O., Schaefer, H. E., and Smootz, E. Characterization of high density lipoproteins in patients heterozygous for Tangier disease. J. Clin. Invest. 60:1025, 1977. 6. Assman, G., von Eckardstein, A., and Brewer, H. B.: Familial high density lipoprotein deficiency: Tangier disease. In Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic Basis of Inherited Disease, 3rd ed. New York, McGraw-Hill, p. 2053, 1995. 7. Batal, R., Tremblay, M., Krimbou, L., et al.: Familial HDL deficiency characterized by hypercatabolism of mature apo A-I but not proapo A-I. Arterioscler. Thromb. Vasc. Biol. 18:655, 1998. 8. Bodzioch, M., Orso, E., Klucken, J., et al.: The gene encoding ATP-binding cassette transporter 1 is mutated in Tangier disease [see comments]. Nat. Genet. 22:347, 1999. 9. Bortnick, A. E., Rothblat, G. H., Stoudt, G., et al.: The correlation of ATP-binding cassette 1 mRNA levels with cholesterol efflux from various cell lines. J. Biol. Chem. 275:28634, 2000. 10. Broccardo, C., Luciani, M., and Chimini, G.: The ABCA subclass of mammalian transporters. Biochim. Biophys. Acta 1461:395, 1999. 11. Brooks-Wilson, A., Marcil, M., Clee, S. M., et al.: Mutations in ABC1 in Tangier disease and familial high-density lipoprotein deficiency [see comments]. Nat. Genet. 22:336, 1999. 12. Brousseau, M. E., Bodzioch, M., Schaefer, E. J., et al.: Common variants in the gene encoding ATP-binding cassette transporter 1 in men with low HDL cholesterol levels and coronary heart disease. Atherosclerosis 154:607, 2001. 13. Brousseau, M. E., Eberhart, G. P., Dupuis, J., et al.: Cellular cholesterol efflux in heterozygotes for Tangier disease is markedly reduced and correlates with high density lipoprotein cholesterol concentration and particle size. J. Lipid Res. 41:1125, 2000. 14. Brousseau, M. E., Schaefer, E. J., Dupuis, J., et al.: Novel mutations in the gene encoding ATP-binding cassette 1 in four Tangier disease kindreds. J. Lipid Res. 41:433, 2000.

15. Brown, M. S., and Goldstein, J. L.: A receptor-mediated pathway for cholesterol homeostasis. Science 232:34, 1986. 16. Bungert, S., Molday, L. L., and Molday, R. S.: Membrane topology of the ATP binding cassette transporter ABCR and its relationship to ABC1 and related ABCA transporters: identification of N-linked glycosylation sites. J. Biol. Chem. 276:23539, 2001. 17. Chawla, A., Boisvert, W. A., Lee, C. H., et al.: A PPAR gamma-LXR-ABCA1 pathway in macrophages is involved in cholesterol efflux and atherogenesis. Mol. Cell 7:161, 2001. 18. Clee, S. M., Kastelein, J. J., van Dam, M., et al.: Age and residual cholesterol efflux affect HDL cholesterol levels and coronary artery disease in ABCA1 heterozygotes. J. Clin. Invest. 106:1263, 2000. 19. Clee, S. M., Zwinderman, A. H., Engert, J. C., et al.: Common genetic variation in ABCA1 is associated with altered lipoprotein levels and a modified risk for coronary artery disease. Circulation 103:1198, 2001. 20. Costet, P., Luo, Y., Wang, N., and Tall, A. R.: Steroldependent transactivation of the ABC1 promoter by the liver X receptor/retinoid X receptor. J. Biol. Chem. 275:28240, 2000. 21. Drobnik, W., Mollers, C., Resink, T., and Schmitz, G.: Activation of phosphatidylinositol-specific phospholipase C in response to HDL3 and LDL is markedly reduced in cultured fibroblasts from Tangier patients. Arterioscler. Thromb. Vasc. Biol. 15:1369, 1995. 22. Dyck, P. J., Ellefson, R. D., Yao, J. K., and Herbert, P. N.: Adult-onset of Tangier disease: 1. Morphometric and pathologic studies suggesting delayed degradation of neutral lipids after fiber degeneration. J. Neuropathol. Exp. Neurol. 37:119, 1978. 23. Eberhart, G. P., Mendez, A. J., and Freeman, M. W.: Decreased cholesterol efflux from fibroblasts of a patient without Tangier disease, but with markedly reduced high density lipoprotein cholesterol levels. J. Clin. Endocrinol. Metab. 83:836, 1998. 24. Engel, W. K., Dorman, J. D., Levy, R. I., and Fredrickson, D. S.: Neuropathy in Tangier disease: alpha-lipoprotein deficiency manifesting as familial recurrent neuropathy and intestinal lipid storage. Arch. Neurol. 17:1, 1967. 25. Fazio, R., Nemni, R., Quattrini, A., et al.: Acute presentation of Tangier polyneuropathy: a clinical and morphological study [see comments]. Acta Neuropathol. 86:90, 1993. 26. Ferrans, V. J., and Fredrickson, D. S.: The pathology of Tangier disease: a light and electron microscopic study. Am. J. Pathol. 78:101, 1975. 27. Fitzgerald, M. L., Mendez, A. J., Moore, K. J., et al.: ATPbinding cassette transporter A1 contains an NH2-terminal signal anchor sequence that translocates the protein’s first hydrophilic domain to the exoplasmic space. J. Biol. Chem. 276:15137, 2001. 28. Fitzgerald, M. L., Morris, A. L., Rhee, J. S., et al.: Naturally occurring mutations in the largest extracellular loops of ABCA1 can disrupt its direct interaction with apolipoprotein A-I. J. Biol. Chem. 277:33178, 2002. 29. Fobker, M., Voss, R., Reinecke, H., et al.: Accumulation of cardiolipin and lysocardiolipin in fibroblasts from Tangier disease subjects. FEBS Lett. 500:157, 2001.

Tangier Disease and Neuropathy 30. Francis, G. A., Knopp, R. H., and Oram, J. F.: Defective removal of cellular cholesterol and phospholipids by apolipoprotein A-I in Tangier disease. J. Clin. Invest. 96:78, 1995. 31. Fredrickson, D. S., Altrochhi, P. S., Avioli, L. V., et al.: Tangier disease—combined clinical staff conference at the National Institutes of Health. Ann. Intern. Med. 55:1016, 1961. 32. Gibbels, E., Schaefer, H. E., Runne, U., et al.: Severe polyneuropathy in Tangier disease mimicking syringomyelia or leprosy: clinical, biochemical, electrophysiological, and morphological evaluation, including electron microscopy of nerve, muscle, and skin biopsies. J. Neurol. 232:283, 1985. 33. Gottesman, M. M., and Ambudkar, S. V.: Overview: ABC transporters and human disease. J. Bioenerg. Biomembr. 33:453, 2001. 34. Guo, Z., Inazu, A., Yu, W., et al.: Double deletions and missense mutations in the first nucleotide-binding fold of the ATP-binding cassette transporter A1 ( ABCA1) gene in Japanese patients with Tangier disease. J. Hum. Genet. 47:325, 2002. 35. Hachem, H., Favre, G., Ghalim, N., et al.: Quantitative abnormalities of lipoprotein particles in multiple myeloma. J. Clin. Chem. Clin. Biochem. 25:675, 1987. 36. Hager, H., and Zimmermann, P.: Peripheral nerve changes in Tangier disease: comparative light-, electron-microscopic and cytometric studies [author’s transl.]. Acta Neuropathol. (Berl.) 45:53, 1979. 37. Hamon, Y., Broccardo, C., Chambenoit, O., et al.: ABC1 promotes engulfment of apoptotic cells and transbilayer redistribution of phosphatidylserine. Nat. Cell Biol. 2:399, 2000. 38. Hamon, Y., Luciani, M. F., Becq, F., et al.: Interleukin-1beta secretion is impaired by inhibitors of the ATP binding cassette transporter, ABC1. Blood 90:2911, 1997. 39. Henderson, L. O., Herbert, P. N., Fredrickson, D. S., et al.: Abnormal concentration and anomalous distribution of apolipoprotein A-I in Tangier disease. Metabolism 27:165, 1978. 40. Higgins, C. F.: ABC transporters: from microorganisms to man. Annu. Rev. Cell Biol. 8:67, 1992. 41. Ho Hong, S., Rhyne, J., Zeller, K., and Miller, M.: Novel ABCA1 compound variant associated with HDL cholesterol deficiency. Biochim. Biophys. Acta 1587:60, 2002. 42. Hong, S., Rhyne, J., Zeller, K., and Miller, M.: ABCA1(Alabama): a novel variant associated with HDL deficiency and premature coronary artery disease. Atherosclerosis 164:245, 2002. 43. Hong, S. H., Riley, W., Rhyne, J., et al.: Lack of association between increased carotid intima-media thickening and decreased HDL-cholesterol in a family with a novel ABCA1 variant, G2265T. Clin. Chem. 48:2066, 2002. 44. Huang, W., Moriyama, K., Koga, T., et al.: Novel mutations in ABCA1 gene in Japanese patients with Tangier disease and familial high density lipoprotein deficiency with coronary heart disease. Biochim. Biophys. Acta 1537:71, 2001. 45. Ishii, J., Nagano, M., Kujiraoka, T., et al.: Clinical variant of Tangier disease in Japan: mutation of the ABCA1 gene in hypoalphalipoproteinemia with corneal lipidosis. J. Hum. Genet. 47:366, 2002.

1917

46. Jessup, W., and Kritharides, L.: Metabolism of oxidized LDL by macrophages. Curr. Opin. Lipidol. 11:473, 2000. 47. Khovidhunkit, W., Memon, R. A., Feingold, K. R., and Grunfeld, C.: Infection and inflammation-induced proatherogenic changes of lipoproteins. J. Infect. Dis. 181(Suppl. 3):S462, 2000. 48. Kocen, R. S., King, R. H., Thomas, P. K., and Haas, L. F.: Nerve biopsy findings in two cases of Tangier disease. Acta Neuropathol. (Berl.) 26:317, 1973. 49. Kocen, R. S., Lloyd, J. K., Lascelles, P. T., et al.: Familial alpha-lipoprotein deficiency (Tangier disease) with neurological abnormalities. Lancet 1:1341, 1967. 50. Kummer, H., Laissue, J., Spiess, H., et al.: Familial analphalipoproteinemia (Tangier disease) [in German]. Schweiz. Med. Wochenschr. 98:406, 1968. 51. Kumon, Y., Suehiro, T., Ikeda, Y., et al.: Influence of serum amyloid A protein on high-density lipoprotein in chronic inflammatory disease. Clin. Biochem. 26:505, 1993. 52. Langmann, T., Klucken, J., Reil, M., et al.: Molecular cloning of the human ATP-binding cassette transporter 1 (hABC1): evidence for sterol-dependent regulation in macrophages. Biochem. Biophys. Res. Commun. 257:29, 1999. 53. Lapicka-Bodzioch, K., Bodzioch, M., Krull, M., et al.: Homogeneous assay based on 52 primer sets to scan for mutations of the ABCA1 gene and its application in genetic analysis of a new patient with familial high-density lipoprotein deficiency syndrome. Biochim. Biophys. Acta 1537:42, 2001. 54. Lawn, R. M., Wade, D. P., Garvin, M. R., et al.: The Tangier disease gene product ABC1 controls the cellular apolipoprotein-mediated lipid removal pathway [see comments]. J. Clin. Invest. 104:R25, 1999. 55. Lin, G., and Bornfeldt, K. E.: Cyclic AMP-specific phosphodiesterase 4 inhibitors promote ABCA1 expression and cholesterol efflux. Biochem. Biophys. Res. Commun. 290:663, 2002. 56. Low, P. A., Dyck, P. J., Okazaki, H., et al.: The splanchnic autonomic outflow in amyloid neuropathy and Tangier disease. Neurology 31:461, 1981. 57. Luciani, M. F., Denizot, F., Savary, S., et al.: Cloning of two novel ABC transporters mapping on human chromosome 9. Genomics 21:150, 1994. 58. Marcil, M., Brooks-Wilson, A., Clee, S. M., et al.: Mutations in the ABC1 gene in familial HDL deficiency with defective cholesterol efflux [see comments]. Lancet 354:1341, 1999. 59. Marcil, M., Yu, L., Krimbou, L., et al.: Cellular cholesterol transport and efflux in fibroblasts are abnormal in subjects with familial HDL deficiency. Arterioscler. Thromb. Vasc. Biol. 19:159, 1999. 60. Mautner, S. L., Sanchez, J. A., Rader, D. J., et al.: The heart in Tangier disease: severe coronary atherosclerosis with near absence of high-density lipoprotein cholesterol. Am. J. Clin. Pathol. 98:191, 1992. 61. McNeish, J., Aiello, R. J., Guyot, D., et al.: High density lipoprotein deficiency and foam cell accumulation in mice with targeted disruption of ATP-binding cassette transporter-1. Proc. Natl. Acad. Sci. U. S. A. 97:4245, 2000. 62. Mendez, A. J.: Cholesterol efflux mediated by apolipoproteins is an active cellular process distinct from efflux mediated by passive diffusion. J. Lipid Res. 38:1807, 1997.

1918

Diseases of the Peripheral Nervous System

63. Mendez, A. J., Lin, G., Wade, D. P., et al.: Membrane lipid domains distinct from cholesterol/sphingomyelin-rich rafts are involved in the ABCA1-mediated lipid secretory pathway. J. Biol. Chem. 276:3158, 2001. 64. Mendez, A. J., Oram, J. F., and Bierman, E. L.: Protein kinase C as a mediator of high density lipoprotein receptordependent efflux of intracellular cholesterol. J. Biol. Chem. 266:10104, 1991. 65. Neufeld, E. B., Remaley, A. T., Demosky, S. J., et al.: Cellular localization and trafficking of the human ABCA1 transporter. J. Biol. Chem. 276:27584, 2001. 66. Nichols, W. C., Dwulet, F. E., Liepnieks, J., and Benson, M. D.: Variant apolipoprotein AI as a major constituent of a human hereditary amyloid. Biochem. Biophys. Res. Commun. 156:762, 1988. 67. Nishida, Y., Hirano, K., Tsukamoto, K., et al.: Expression and functional analyses of novel mutations of ATP-binding cassette transporter-1 in Japanese patients with high-density lipoprotein deficiency. Biochem. Biophys. Res. Commun. 290:713, 2002. 68. Oram, J. F.: Tangier disease and ABCA1. Biochim. Biophys. Acta 1529:321, 2000. 69. Oram, J. F.: ATP-binding cassette transporter A1 and cholesterol trafficking. Curr. Opin. Lipidol. 13:373, 2002. 70. Oram, J. F., Lawn, R. M., Garvin, M. R., and Wade, D. P.: ABCA1 is the cAMP-inducible apolipoprotein receptor that mediates cholesterol secretion from macrophages. J. Biol. Chem. 275:34508, 2000. 71. Orso, E., Broccardo, C., Kaminski, W. E., et al.: Transport of lipids from Golgi to plasma membrane is defective in Tangier disease patients and Abc1-deficient mice. Nat. Genet. 24:192, 2000. 72. Panagotopulos, S. E., Witting, S. R., Horace, E. M., et al.: The role of apolipoprotein A-I helix 10 in apolipoprotein-mediated cholesterol efflux via ATP-binding cassette transporter A1 (ABCA1). J. Biol. Chem. 277:39477, 2002. 73. Pietrini, V., Pinna, V., and Milone, F. F.: Tangier disease: central nervous system impairment in a case of syringomyelialike syndrome. J. Neurol. Sci. 98:245, 1990. 74. Pietrini, V., Rizzuto, N., Vergani, C., et al.: Neuropathy in Tangier disease: a clinicopathologic study and a review of the literature. Acta Neurol. Scand. 72:495, 1985. 75. Pollock, M., Nukada, H., Frith, R. W., et al.: Peripheral neuropathy in Tangier disease. Brain 106:911, 1983. 76. Pressly, T. A., Scott, W. J., Ide, C. H., et al.: Ocular complications of Tangier disease. Am. J. Med. 83:991, 1987. 77. Pullinger, C. R., Hakamata, H., Duchateau, P. N., et al.: Analysis of hABC1 gene 5⬘ end: additional peptide sequence, promoter region, and four polymorphisms. Biochem. Biophys. Res. Commun. 271:451, 2000. 78. Reinhart, W. H., Gossi, U., Butikofer, P., et al.: Haemolytic anaemia in an alpha-lipoproteinaemia (Tangier disease): morphological, biochemical, and biophysical properties of the red blood cell. Br. J. Haematol. 72:272, 1989. 79. Remaley, A. T., Rust, S., Rosier, M., et al.: Human ATP-binding cassette transporter 1 (ABC1): genomic organization and identification of the genetic defect in the original Tangier disease kindred. Proc. Natl. Acad. Sci. U. S. A. 96:12685, 1999.

80. Remaley, A. T., Schumacher, U. K., Stonik, J. A., et al.: Decreased reverse cholesterol transport from Tangier disease fibroblasts: acceptor specificity and effect of brefeldin on lipid efflux. Arterioscler. Thromb. Vasc. Biol. 17:1813, 1997. 81. Repa, J. J., Turley, S. D., Lobaccaro, J. A., et al.: Regulation of absorption and ABC1-mediated efflux of cholesterol by RXR heterodimers. Science 289:1524, 2000. 82. Robenek, H., and Schmitz, G.: Abnormal processing of Golgi elements and lysosomes in Tangier disease. Arterioscler. Thromb. 11:1007, 1991. 83. Rodriguez Reguero, J. J., Iglesias Cubero, G., Vazquez, M., et al.: Variation in plasma lipid and lipoprotein concentrations in community-acquired pneumonia: a six-month prospective study. Eur. J. Clin. Chem. Clin. Biochem. 34:245, 1996. 84. Rogler, G., Trumbach, B., Klima, B., et al.: HDL-mediated efflux of intracellular cholesterol is impaired in fibroblasts from Tangier disease patients. Arterioscler. Thromb. Vasc. Biol. 15:683, 1995. 85. Rust, S., Rosier, M., Funke, H., et al.: Tangier disease is caused by mutations in the gene encoding ATP-binding cassette transporter 1 [see comments]. Nat. Genet. 22:352, 1999. 86. Rust, S., Walter, M., Funke, H., et al.: Assignment of Tangier disease to chromosome 9q31 by a graphical linkage exclusion strategy. Nat. Genet. 20:96, 1998. [published erratum appears in Nat. Genet. 20:312, 1998.] 87. Santamarina-Fojo, S., Peterson, K., Knapper, C., et al.: Complete genomic sequence of the human ABCA1 gene: analysis of the human and mouse ATP-binding cassette A promoter. Proc. Natl. Acad. Sci. U. S. A. 97:7987, 2000. 88. Schaefer, E. J., Blum, C. B., Levy, R. I., et al.: Metabolism of high-density lipoprotein apolipoproteins in Tangier disease. N. Engl. J. Med. 299:905, 1978. 89. Schaefer, E. J., Zech, L. A., Schwartz, D. E., and Brewer, H. Jr.: Coronary heart disease prevalence and other clinical features in familial high-density lipoprotein deficiency (Tangier disease). Ann. Intern. Med. 93:261, 1980. 90. Schmalbruch, H., Stender, S., and Boysen, G.: Abnormalities in spinal neurons and dorsal root ganglion cells in Tangier disease presenting with a syringomyelia-like syndrome. J. Neuropathol. Exp. Neurol. 46:533, 1987. 91. Schmitz, G., Assmann, G., Robenek, H., and Brennhausen, B.: Tangier disease: a disorder of intracellular membrane traffic. Proc. Natl. Acad. Sci. U. S. A. 82:6305, 1985. 92. Schmitz, G., Fischer, H., Beuck, M., et al.: Dysregulation of lipid metabolism in Tangier monocyte-derived macrophages. Arteriosclerosis 10:1010, 1990. 93. Schmitz, G., and Orso, E.: Intracellular cholesterol and phospholipid trafficking: comparable mechanisms in macrophages and neuronal cells. Neurochem. Res. 26:1045, 2001. 94. Serfaty-Lacrosniere, C., Civeira, F., Lanzberg, A., et al.: Homozygous Tangier disease and cardiovascular disease. Atherosclerosis 107:85, 1994. 95. Silver, D. L., Wang, N., Xiao, X., and Tall, A. R.: High density lipoprotein (HDL) particle uptake mediated by scavenger receptor class B type 1 results in selective sorting of HDL cholesterol from protein and polarized cholesterol secretion. J. Biol. Chem. 276:25287, 2001.

Tangier Disease and Neuropathy 96. Singaraja, R. R., Brunham, L. R., Visscher, H., et al.: Efflux and atherosclerosis. Arterioscler. Thromb. Vasc. Biol. 23:1322, 2003. 97. Smith, J. D., Miyata, M., Ginsberg, M., et al.: Cyclic AMP induces apolipoprotein E binding activity and promotes cholesterol efflux from a macrophage cell line to apolipoprotein acceptors. J. Biol. Chem. 271:30647, 1996. 98. Smith, J. D., Waelde, C., Horwitz, A., and Zheng, P.: Evaluation of the role of phosphatidylserine translocase activity in ABCA1-mediated lipid efflux. J. Biol. Chem. 277:17797, 2002. 99. Takahashi, Y., and Smith, J. D.: Cholesterol efflux to apolipoprotein AI involves endocytosis and resecretion in a calcium-dependent pathway [see comments]. Proc. Natl. Acad. Sci. U. S. A. 96:11358, 1999. 100. van Dam, M. J., de Groot, E., Clee, S. M., et al.: Association between increased arterial-wall thickness and impairment in ABCA1-driven cholesterol efflux: an observational study. Lancet 359:37, 2002. 101. Venkateswaran, A., Laffitte, B. A., Joseph, S. B., et al.: Control of cellular cholesterol efflux by the nuclear oxysterol receptor LXRalpha. Proc. Natl. Acad. Sci. U. S. A. 97:12097, 2000. 102. von Eckardstein, A., Chirazi, A., Schuler-Luttmann, S., et al.: Plasma and fibroblasts of Tangier disease patients are

103.

104.

105.

106.

107.

108.

1919

disturbed in transferring phospholipids onto apolipoprotein A-I. J. Lipid Res. 39:987, 1998. Walter, M., Reinecke, H., Gerdes, U., et al.: Defective regulation of phosphatidylcholine-specific phospholipases C and D in a kindred with Tangier disease: evidence for the involvement of phosphatidylcholine breakdown in HDL-mediated cholesterol efflux mechanisms. J. Clin. Invest. 98:2315, 1996. Wang, J., Burnett, J. R., Near, S., et al.: Common and rare ABCA1 variants affecting plasma HDL cholesterol. Arterioscler. Thromb. Vasc. Biol. 20:1983, 2000. Wang, N., Silver, D. L., Costet, P., and Tall, A. R.: Specific binding of ApoA-I, enhanced cholesterol efflux, and altered plasma membrane morphology in cells expressing ABC1. J. Biol. Chem. 275:33053, 2000. Wang, N., Silver, D. L., Thiele, C., and Tall, A. R.: ATP-binding cassette transporter A1 (ABCA1) functions as a cholesterol efflux regulatory protein. J. Biol. Chem. 276:23742, 2001. Winder, A. F., Alexander, R., Garner, A., et al.: The pathology of cornea in Tangier disease (familial high density lipoprotein deficiency). J. Clin. Pathol. 49:407, 1996. Zuchner, S., Sperfeld, A. D., Senderek, J., et al.: A novel nonsense mutation in the ABC1 gene causes a severe syringomyelia-like phenotype of Tangier disease. Brain 126:920, 2003.

83 Hereditary Amyloid Neuropathy MARY M. REILLY

History Classification TTR Amyloidosis Clinical Features Pathology Genetics/Epidemiology

Pathogenesis Differential Diagnosis Diagnosis Treatment Prognosis

The hereditary amyloid neuropathies, more commonly known as the familial amyloid polyneuropathies (FAPs), are a heterogeneous group of autosomal dominant disorders originally described by Andrade in Portuguese patients in 1952.10 FAP is one of a group of conditions called the amyloidoses that are characterized by deposition of a fibrillar protein with abundant beta pleated structure in the extracellular space.13 The staining of amyloid with Congo red dye and the subsequent viewing of it microscopically under crossed polarized light gives a characteristic apple-green birefringence. Many different proteins have been described to form amyloid, and it is the individual constituent protein that defines the various types of amyloidosis.74 Amyloidosis is usually either systemic or localized. The most common systemic forms are primary light chain (AL) amyloidosis (constituent protein: kappa or lambda light chain), secondary or reactive amyloidosis (constituent protein: serum amyloid A), and FAP (constituent proteins: transthyretin [TTR], apolipoprotein A-I, and gelsolin). There are also many localized forms of amyloidosis, including some that affect the nervous system (e.g., Alzheimer’s disease). Amyloid neuropathy, classically seen in FAP, can also be a feature of AL amyloidosis and dialysis-related amyloidosis (constituent protein: ␤2-microglobulin).75 In AL amyloidosis, two types of neuropathy are described, a compression neuropathy such as carpal tunnel syndrome and a generalized neuropathy with autonomic involvement.54 In patients on long-term dialysis, carpal tunnel syndrome, caused by ␤2-microglobulin amyloid, may be found.

Apolipoprotein A-I Amyloidosis Gelsolin Amyloidosis

In FAP, the most common fibril protein deposited as amyloid is a variant form of TTR22 (formerly known as prealbumin), but FAP can also occur secondary to apolipoprotein A-I deposition62 and to gelsolin deposition.31 Although there are now at least 70 amyloidogenic point mutations and one trinucleotide deletion in the TTR gene, the first described mutation, methionine 30 (Met 30), originally described in Portuguese patients with TTR-related FAP,27 is still the most common worldwide. Clinically the neuropathy in TTR-related FAP starts as a small fiber neuropathy and progresses to a more generalized sensorimotor neuropathy commonly associated with autonomic dysfunction, a cardiomyopathy, and vitreous involvement. In the last two decades major progress has been made in the understanding of FAP, particularly TTR-related FAP, and more recently in the treatment of TTR-related FAP with liver transplantation.

HISTORY In 1842, Rokitansky described a specific condition giving rise to an enlarged liver and spleen associated with certain chronic diseases, later known as amyloidosis. The term amyloid was originally coined in 1853 by Virchow99 following his studies on isomers of starch in humans and his observation of the cellulose-like nature of the sago grains in the sago spleens described by Christensen. The earliest descriptions of amyloid clearly refer to secondary amyloidosis, that is, amyloid associated with chronic infections 1921

1922

Diseases of the Peripheral Nervous System

such as tuberculosis and osteomyelitis. The association of multiple myeloma and amyloid (AL) was noted in 1867, and in the first part of the 20th century the association of Bence Jones protein with amyloid formation and the finding of increased bone marrow cells in primary amyloidosis were recorded.20 The modern method of staining for amyloid with Congo red dye was introduced by Bennhold in 1922, and shortly afterward it was found that all amyloid had positive green birefringence after Congo red staining. By 1959, Cohen and Calkins had shown that amyloid could be clearly visualized, isolated, and characterized as a fibrous protein.21 The first description of amyloidosis clinically and pathologically affecting the peripheral nerves was published by Königstein in 1925.52 He noted that the clinical involvement of the peripheral nerves seemed to be much greater than the histologic involvement, a finding now well recognized in FAP. In 1938, de Navasquez and Treble described a case of polyneuropathy secondary to amyloidosis with associated autonomic features.26 Hereditary amyloid neuropathy was first described by Andrade in 1952.10 His original description of FAP was of “a peculiar form of peripheral neuropathy” known for generations among the local people as mal dos pesinhos (foot disease). He gave a detailed description of 74 patients with a disease that started insidiously in the second or third decade of life and was characterized by a “progressive lowering of the general state of health, gastrointestinal disturbances, premature impotence, and a syndrome of involvement of the peripheral neurone starting and predominating in the lower extremities.”10 He demonstrated that the disease was often familial and that the average duration of the illness was 7 to 10 years. The size of the problem in Portugal has gradually come to light, with over 500 kindreds now known to be affected. The suggestion that the fibril protein in FAP was immunologically related to TTR was made in 1978,22 and the first point mutation (Met 30) in the TTR gene associated with FAP and accounting for the original Portuguese cases was described in 1983.27 More that 70 amyloidogenic TTR point mutations subsequently have been described, but TTR Met 30 associated with FAP is still the most common and best characterized. In 1969, Van Allen et al. described an Iowan kindred with a typical clinical picture of FAP.98 In this kindred, the constituent amyloid protein has subsequently been shown to be a variant of apolipoprotein A-I with an arginine-forglycine substitution at position 26.63 The third form of FAP was originally described by Meretoja in a Finnish kindred in 1969.59 Clinically this is different from classic FAP and is characterized by corneal lattice dystrophy and a progressive cranial neuropathy. The fibril protein in this disease is an abnormal fragment of gelsolin, usually associated with a substitution of asparagine for aspartic acid at position 15.55

CLASSIFICATION In the past FAP has been classified into four different types based on clinical presentation, but this classification has now been replaced by a classification based on the nature of the constituent amyloidogenic protein. The four types of FAP in the original classification were type I (lower limb onset), originally and mainly described in Portuguese,10 Japanese,11 and Swedish7 kinships; type II (upper limb onset), originally described in the Indiana/Swiss56 and in the German/Maryland64 kinships; type III (lower limb neuropathy, nephropathy, and gastric ulcers), originally described in an Iowa kinship98; and type IV (cranial nerve involvement with corneal lattice dystrophy), originally described in a Finnish kindred.59 In the more modern classification, FAP is divided into three types, TTR-related FAP, apolipoprotein A-I–related FAP, and gelsolin-related FAP. TTR-related FAP is so named because TTR is the constituent amyloid protein. This encompasses types I and II of the original classification, both of which are secondary to TTR deposition. Apolipoprotein A-I–related FAP, secondary to apolipoprotein A-I deposition, encompasses type III of the original classification. Gelsolin-related FAP, secondary to gelsolin deposition, is the same as type IV in the original classification. In TTR-related and apolipoprotein A-I–related amyloidosis, a neuropathy is not always present, so in this chapter the three different forms of hereditary amyloidosis that can be associated with a neuropathy will be referred to as TTR amyloidosis, apolipoprotein A-I amyloidosis, and gelsolin amyloidosis. Of the over 70 pathogenic TTR point mutations now described, most are associated with hereditary amyloid neuropathy, and it is clear with the ever-increasing number of different mutations being described that there are overlapping clinical syndromes between specific mutations. Nevertheless, in an individual patient, the ethnic origin or in some cases the presenting clinical feature may alert the clinician to suspect a particular mutation.36

TTR AMYLOIDOSIS TTR amyloidosis is the most common cause of inherited amyloid neuropathy, and since the recognition that the fibril protein in this type of amyloidosis was immunologically related to TTR was made in 1978,22 TTR amyloidosis has been recognized throughout the world. The cardinal clinical features in TTR amyloidosis associated with a neuropathy are a sensorimotor peripheral neuropathy usually with autonomic involvement, a cardiomyopathy, and to a lesser extent vitreous and renal involvement.74 Many other systems can be involved, as described below. TTR amyloidosis Met 30 is a major cause of morbidity and mortality in

Hereditary Amyloid Neuropathy

areas where it is endemic, such as in the Oporto region of Portugal.

Clinical Features TTR Amyloidosis Met 30 TTR amyloidosis Met 30 is the best characterized of all the amyloidosis-associated TTR point mutations because it occurs most commonly. The detailed original description of this type of amyloidosis by Andrade in 74 patients10 fits well with the subsequent description of further cases. As well as describing a peripheral and autonomic neuropathy starting in the second or third decade of life, Andrade noted that the disease was often familial and that the average duration of the disease was 7 to 10 years, with patients usually succumbing to cachexia, intercurrent infection, or cardiovascular collapse. In his series, patients usually presented with sensory symptoms including hypo- or analgesia, painless injuries, burns, paresthesias, or pain in the lower limbs. Pain was often a significant feature, especially secondary to slight cutaneous stimuli. From these presentations, Andrade concluded that there was a pattern to the sensory loss, starting with the loss of thermal sensibility and progressively involving pain, touch, and deep sensibility. He described the motor system becoming involved later, with the lower limbs again being involved first. As the disease progressed the reflexes were gradually lost. Andrade noted anisocoria; gastrointestinal disturbances, including constipation, diarrhea, abdominal distention, and progressive weight loss; and genitourinary disturbances, especially impotence. Postmortem studies carried out in two of these cases found amyloid deposited in the peripheral nerves, kidneys, heart, pancreas, skin, lungs, and stomach. In the peripheral nerves, amyloid was deposited within the nerve bundles and in spinal roots and ganglia. Subsequent descriptions of this type of amyloidosis, particularly in other Portuguese cases, are very similar. The cardinal feature remains a length-dependent sensory and motor neuropathy that usually starts in the lower limbs with symptoms of small fiber involvement predominating initially. This gives rise to the classic examination findings of a dissociated sensory loss more pronounced for pain and temperature than for light touch.1 Eventually all sensory modalities are involved, which can lead to complications secondary to painless injury to the feet, including ulcers, cellulitis, osteomyelitis, and Charcot joints.74 Motor involvement typically occurs later in the course of the disease, initially with wasting and weakness in the lower limbs but gradually progressing to the upper limbs, accompanied by a progressive loss of tendon reflexes. The neuropathy can be very severe and disabling in the latter stages of the disease. Carpal tunnel syndrome can occur in TTR amyloidosis Met 30 but is not usually the presenting feature,14 unlike in other forms of TTR amyloidosis, as described below. Neurophysiologic studies confirm an axonal neuropathy. In the early stages the only abnormalities detected may be

1923

in small fiber function, such as abnormal thermal thresholds (see Chapter 39 for an example of this). Later the sensory action potentials become progressively reduced and eventually absent, initially in the lower limbs but eventually in the upper limbs. The sensory conduction velocities are usually normal or close to normal. The motor amplitudes are normal at first but become progressively reduced and eventually are often absent, especially in the lower limbs. Motor conduction velocities are usually normal or close to normal. As Andrade suggested in his original description,10 autonomic involvement can be early and can be severe. Symptomatically this usually manifests with orthostatic hypotension, alternating constipation and diarrhea, severe gastric distention and retention, and genitourinary problems, including sexual impotence and urinary disturbances.74 The gastrointestinal symptoms can be very troublesome and can include diarrhea triggered by meals, explosive diarrhea, and nocturnal diarrhea. Recurrent vomiting can more rarely be a prominent symptom.1 Examination findings include scalloped pupils, postural hypotension, an abnormal Valsalva response, and a fixed pulse rate. Pupillography and formal autonomic testing are usually abnormal, especially as the disease progresses, demonstrating both sympathetic and parasympathetic involvement. The autonomic neuropathy can be very debilitating, and in the terminal stages of the disease patients can be bedbound as a result of a combination of a severe peripheral and autonomic neuropathy. The other frequently involved system in TTR amyloidosis Met 30 is the heart. Although pathologic involvement of the heart was shown in Andrade’s original cases, subsequent studies have confirmed the importance of cardiac involvement in the overall clinical picture. One Portuguese study of 60 cases showed that all patients had cardiac involvement in advanced disease.30 Clinically the cardiac involvement usually presents with an arrhythmia, heart block, or heart failure. Electrocardiographic abnormalities include widespread T-wave and Q-wave repolarization changes and various conduction disturbances. Echocardiography is abnormal in cardiac amyloidosis, showing a restrictive cardiomyopathy with thickened interventricular septum and thickened ventricular walls with greatly refractile echoes (described as a “granular sparkling appearance”).42 Serum amyloid P (SAP) component scans using 123I radiolabeling are also used to detect cardiac amyloid. Vitreous amyloid deposits are well recognized in TTR amyloidosis Met 30 and are often the presenting feature in Swedish patients.84 Other systems, including the central nervous system (CNS), may be involved in TTR amyloidosis Met 30, but such involvement is less common (see Other System Involvement below). TTR amyloidosis Met 30 is most commonly seen in patients of Portuguese and Swedish origin,70 but is now

1924

Diseases of the Peripheral Nervous System

recognized worldwide.75 It had been postulated that all patients with TTR amyloidosis Met 30 were Portuguese in origin, with a common founder, but haplotype studies have clearly shown many founders for this mutation.43,72,106 There is marked clinical heterogeneity in cases of TTR amyloidosis Met 30, with some consistency in each major cluster of the disease, in relation to both age of onset and the nature of the initial presentation. Portuguese patients and patients of Portuguese descent tend to present in a classic manner with a lower limb sensory neuropathy in their 20s and 30s, whereas Swedish patients with the same mutation, and surprisingly the same haplotype, usually present in their late 50s with vitreous involvement.84 In Japan, the TTR amyloidosis Met 30 cases in the two main endemic areas of Arao city and Ogawa village present early, like the Portuguese cases, but Japanese patients presenting outside these areas often present later (46 to 80 years), with a slightly different phenotype including less autonomic involvement.44 Worldwide variations in both age of onset and clinical presentation are well documented with TTR amyloidosis Met 30. Other TTR Amyloidosis Point Mutations A clinical picture similar to that seen with TTR amyloidosis Met 30 is described with many other TTR point mutations. No specific clinical picture predicts a specific mutation, but a few mutations deserve further comment. TTR Tyr 77, the second most common TTR point mutation associated with TTR amyloidosis, has been shown to have multiple founders.72 This mutation was originally described in a German kindred from Illinois but has since been described in families from many countries, including France, the United Kingdom, the United States, and Spain.73,100 Disease onset is usually later than in TTR amyloidosis Met 30, often in the 50s. The clinical features are otherwise similar to those of TTR amyloidosis Met 30 except for a high incidence of carpal tunnel syndrome (which can be the presenting feature) and the absence of vitreous involvement. Other TTR amyloidosis mutations that classically present in the upper limbs and often with carpal tunnel syndrome are Ser 8428 and His 58.64 This type of TTR amyloidosis (FAP type II in the original classification) was originally described in two kindreds, an Indiana family of Swiss origin (Ser 84) and a Maryland family of German origin (His 58). The clinical presentation is of a generalized neuropathy starting in the upper limbs, often associated with autonomic failure and occasionally with cardiomyopathy. This type of upper limb presentation has now been described with many TTR mutations.75 TTR amyloidosis Ala 60 has been found in the United States, Ireland, the United Kingdom, and Australia, and to date all patients who have had haplotype studies done have been shown to have a common founder in northwest Ireland.76,89 TTR amyloidosis Ala 60 has also been described

in Japan, but no haplotype data are available.53 The TTR amyloidosis Ala 60 families therefore represent the second largest group of families with a common mutation presenting in different areas to be traced to a common founder, the largest being those of known Portuguese origin with TTR amyloidosis Met 30. TTR amyloidosis Ala 60 typically presents late in the sixth or seventh decade, and both large fiber sensory loss and motor involvement are more prominent than in TTR amyloidosis Met 30. Cardiomyopathy is often a major feature and is commonly the presenting feature. A more aggressive and rapid disease course is described in TTR amyloidosis with some TTR mutations, including Pro 55,45 Lys 54,96 and Ser 25.104 CNS Involvement Until recently CNS involvement in TTR amyloidosis was considered to be very rare, although postmortem studies had shown leptomeningeal amyloid deposition that usually was asymptomatic. Goren et al.34 first described the rare syndrome of oculoleptomeningeal amyloidosis (OLMA) in 1980, and subsequently this has been shown to be associated with TTR point mutations.17 The clinical features vary even within a kindred and include vitreous opacities, progressive dementia, stroke, subarachnoid hemorrhage, ataxia, hydrocephalus, seizures, spasticity, and episodes of fluctuating consciousness often with focal neurologic signs.17 Magnetic resonance imaging (MRI) may show meningeal enhancement in the brain and spinal cord, and the cerebrospinal fluid (CSF) protein can be elevated. The other typical TTR amyloidosis features, including peripheral neuropathy and cardiomyopathy, are usually mild, although one patient with a TTR Pro 12 mutation was described with a severe peripheral neuropathy and prominent central manifestations.17 OLMA has been described in association with four TTR point mutations (Gly 30, Met 30, Gly 18, and Pro 12), and because it has been described with the common Met 30 mutation, OLMA should be considered a rare part of the TTR amyloidosis syndrome. Other System Involvement In postmortem studies, amyloid is often found to be widely distributed in TTR amyloidosis, and yet many systems are only rarely symptomatic. Kidney involvement can occur but is often asymptomatic and may only be picked up when patients are being considered for liver transplantation. Gastrointestinal involvement can occur independently from the gastrointestinal symptoms secondary to autonomic involvement and includes a protein-losing enteropathy.74 Rarely, symptomatic involvement of muscle, bone, and lungs occurs. Course TTR amyloidosis is a slowly progressive illness in most patients except for the more aggressive forms described above. The course is best described for TTR amyloidosis

1925

Hereditary Amyloid Neuropathy

Met 30 and has been divided into three stages.24 Stage 1 is characterized by a mainly sensory neuropathy in the lower limbs when the patient is still independently ambulatory. In stage 2, occurring after a mean interval of 5.6 years, there is a more generalized sensorimotor neuropathy affecting gait (patients may need walking aids) and beginning to affect the upper limbs. In stage 3, which occurs after a mean of 10.4 years, the patient is either confined to a wheelchair or bedridden. Death usually occurs after an interval of about 10 years with TTR amyloidosis Met 30 and is usually due to infections, cachexia, or cardiac or autonomic complications.1 The course of the disease varies somewhat for other TTR mutations.

Pathology All types of amyloid found in the peripheral nervous system occur as extracellular, amorphous, eosinophilic deposits.74 The most widely used technique to demonstrate amyloid is a combination of alkaline Congo red (Fig. 83–1A) and polarizing filters to demonstrate the characteristic applegreen birefringence (Fig. 83–1B) (see Chapter 32). It is important to identify the constituent fibril protein in amyloid, particularly to distinguish between TTR amyloidosis and AL amyloidosis because they can present with a very similar clinical picture. Immunohistochemical studies using monoclonal antibodies for TTR (Fig. 83–1C) and kappa or lambda light chains are used for this purpose but are not always reliable.1 A negative study from immunostaining for TTR does not exclude TTR amyloidosis. TTR amyloidosis Met 30 is the best characterized pathologically because it is the most common type of TTR amyloidosis. Amyloid deposits can be found in any part of the peripheral nervous system, including the nerve trunks, plexuses, and sensory and autonomic ganglia. In the ganglia, amyloid often forms deposits around the small ganglion cells (see Chapter 31). In the peripheral nerves, amyloid deposits occur extracellularly in the epi-, peri, or endoneurium. The size and number of deposits vary, with both large and small deposits described. The deposits are often found around blood vessels, where they may form a symmetrical annulate cuff. Larger deposits can be seen, unrelated to blood vessels, with a globular form surrounding collagen fibrils (Fig. 83–1D). One study of both postmortem nerves and sural nerve biopsies from patients with varying stages of TTR amyloidosis has demonstrated that amyloid deposition in the endoneurium starts angiocentrically in the small vessel walls and surrounding tissue.12 The same study showed that, in more advanced cases, deposits were also seen in the subperineurial and/or epineurial regions. Electron microscopy demonstrates that amyloid is composed of fibrils with a characteristic ultrastructural appearance. The fibrils are straight, nonbranching tubules with a diameter of 8 nm (Fig. 83–1E) that in cross section have a very dense, hard-edged profile. A recent study of the

ultrastructure of amyloid fibrils in TTR amyloidosis Met 30 showed that they were composed of two protofilaments twisted at 180 degrees to the right and left alternatively with a periodicity of 125 to 135 nm.46 The fibrils may form disorganized mats or radiating arrays, and they are often closely related to the basal lamina, both of Schwann cells and of perineurial cells. In TTR amyloidosis, it is well recognized that the neuropathy is due to progressive axonal loss characterized by fibers undergoing wallerian degeneration. The axonal loss initially affects unmyelinated and small myelinated fibers and later affects the larger myelinated fibers.29,95 The loss of unmyelinated axons is indicated by the occurrence of flat sheets of Schwann cell processes. Although occasionally recent wallerian-type degeneration and macrophage infiltration are seen in affected nerves, it is more usual not to demonstrate active pathology. Teased fiber studies have shown that there may also be segmental demyelination and remyelination.29 This could result from demyelination secondary to Schwann cell dysfunction or to axonal atrophy. There is usually little evidence of axonal regeneration. As the disease progresses, the number of remaining nerve fibers declines with the gradual replacement of the endoneurial contents by amyloid.

Genetics/Epidemiology The original suggestion that the fibril protein in TTR amyloidosis was immunologically related to TTR was made by Costa et al.22 The three-dimensional structure of plasma TTR is known. TTR is a 55-kDa tetramer composed of four identical monomer subunits. Each monomer shows extensive beta pleated structure with eight beta pleated sheets arranged in two parallel plates. Two monomers combine to form a dimer and the two dimers bond noncovalently to form the tetramer.14 Over 90% of TTR is produced in the liver, with the remainder produced in the choroid plexus and retina. The mature protein has two known functions: it is responsible for about 20% of plasma thyroxine binding, and it also binds with retinol-binding proteins. TTR is the product of a single-copy gene on chromosome 18 (18q11.2-q12.1).101 This gene has four exons that code for a 127–amino-acid mature protein and an 18-residue signal peptide. The protein is mainly encoded by exons 2, 3, and 4, whereas exon 1 encodes a signal peptide and the first three amino acids of the mature protein. The more than 80 point mutations reported in TTR are all in exons 2, 3, and 4, and most of these are pathogenic, usually presenting with the clinical features of TTR amyloidosis neuropathy and more rarely with just a cardiomyopathy.70 A case of TTR amyloidosis has been reported secondary to a trinucleotide deletion in exon 4.97 Several nonpathogenic point mutations have been described, including Thr 109, associated with euthyroid hyperthyroxinemia,60 and neutral

1926

Diseases of the Peripheral Nervous System

FIGURE 83–1 A, Photomicrograph of a paraffin section of a nerve from a patient with familial amyloidosis stained with Congo red and hematoxylin. Large masses of acellular Congo red–positive amyloid (*) are present in the endoneurium. Bar: 1 ␮m. B, The same area viewed under polarized light. The Congo red–stained amyloid emits a characteristic apple-green birefringence (*), whereas that of collagen is white. Bar: 10 ␮m. C, An adjacent section to that in A and B immunostained for transthyretin, showing that the amyloid deposits are immunoreactive (*). Bar: 10 ␮m. D, Electron micrograph of the biopsy shown in A, B, and C. Deposits of amyloid fibrils (*) replace or displace endoneurial cells and collagen and appear to erode the basal laminae of adjacent Schwann cells (arrows), particularly nonmyelinating ones. Further away from the amyloid, nonmyelinating Schwann cells still contain axons (arrowheads). Bar: 1 nm. E, Amyloid is composed of linear nonbranching fibrils 7 to 10 nm in width, often randomly oriented (⫻20,000).

polymorphisms detected in healthy people, including Met 119 and Ser 6. The various point mutations in the TTR gene described in TTR amyloidosis usually exist in the heterozygous state and are inherited in an autosomal dominant manner. Haplotype

studies of TTR Met 30 have shown multiple founders for this mutation, which may be partially explained by the occurrence of the mutation in a CpG dinucleotide hot spot.106 Homozygosity for the TTR Met 30 mutation has now been described in at least 18 people and is not associated

1927

Hereditary Amyloid Neuropathy

with either earlier onset or more aggressive disease than in the heterozygotes,44 unlike most autosomal dominant conditions, in which homozygotes are more severely affected than heterozygotes. Patients homozygous for the TTR Met 30 mutation have been reported to have later onset disease and a higher incidence of vitreous deposits.79 Some TTR Met 30 homozygotes have been documented to be asymptomatic up to the age of 70 years. These observations could be due to other genetic factors. The description of a discordant phenotype of TTR amyloidosis Met 30 in monozygotic twins suggests that acquired factors are also likely to affect the phenotype in TTR amyloidosis.38 Compound heterozygosity has been described for nonpathogenic and pathogenic TTR point mutations usually occurring on different alleles.81 The Ser 6 polymorphism has been described in association with many pathogenic TTR point mutations, but it does not seem to influence the clinical course of the disease.5 Conversely, if either of two TTR polymorphisms, Met 119 and His 104, occurs with Met 30, this provides a protective effect as demonstrated by a more benign disease course.19,94 The mechanism behind this has been shown to be increased tetramer stability conferred by the nonpathogenic mutations, as discussed below.4 The penetrance for TTR amyloidosis varies even for patients with the same mutation. Epidemiologic studies in northern Portugal found a TTR amyloidosis prevalence rate of 1 in 1000 and a gene carrier frequency of 1 in 538. In this group the mean age of onset is 33, with almost complete penetrance.85 In Japan, the patients with TTR Met 30 from the two endemic foci who have early-onset disease are similar to the Portuguese patients and have a higher rate of penetrance than the patients from the nonendemic foci, who present later with a different disease course, as described above.44 The penetrance rates for other mutations vary, but reduced penetrance has been reported with TTR Ala 60.76 TTR amyloidosis patients have been described with unaffected parents, but nearly all TTR mutations seem to be inherited, with reduced penetrance accounting for the unaffected parents. There was probable evidence for a de novo mutation in a patient with the Arg 47 TTR mutation because neither parent carried the mutation.61 There is also a recent report of a family with TTR Ser 25 that originated in a paternal germline mosaicism.104 Other interesting epidemiologic observations in patients with TTR Met 30 have yet to be fully understood. Anticipation has been described in Portuguese, Swedish, and early-onset Japanese TTR amyloidosis Met 30 patients,70 but unlike other neurodegenerative diseases demonstrating anticipation, the anticipation in TTR Met 30, at least in the Portuguese families, has been shown not to be due to triplet repeat expansions.83 Portuguese and Swedish families with TTR Met 30 both show more affected men than women, with women having a slightly later age of onset.84 These families also show an earlier age of onset when the disease is inherited from an affected

mother. Similarly, in late-onset TTR amyloidosis Met 30 in French patients70 and also in late-onset TTR amyloidosis Met 30 in Japanese patients, there is an unusually high male-to-female ratio, with asymptomatic carriers, usually females, detected late in life.44 In the French patients, the TTR Met 119 polymorphism has been excluded as a genetic modifier. All of these observations suggest strongly that other genetic and environmental factors affecting the phenotype in TTR amyloidosis have yet to be identified.

Pathogenesis TTR amyloid is composed of protein fibrils and nonfibrillary constituents such as SAP and glycosaminoglycans, present in all types of amyloid. As stated above, these fibrils exist in a beta pleated structure with eight beta pleated sheets arranged in two parallel plates, a structure allowing possibly infinite self-aggregation.41 The amyloidogenic potential of TTR is presumed to be partially due to its extensive beta pleated structure. Further evidence for this is seen in senile systemic amyloidosis, a condition affecting approximately 20% of people over 80 years of age,68 in which normal TTR is deposited in the heart as amyloid.102 More recently, it has also been observed that the TTR deposited in the heart in TTR amyloidosis is composed of both mutated TTR and normal TTR.103 X-ray crystallography studies of TTR secondary to various mutations have been carried out to investigate whether conformational changes caused by the various mutations could account for the increased amyloidogenicity of mutant TTR. The results show that, generally, TTR amyloidosis point mutations do not alter the native TTR folded structure significantly, except for the Leu55Pro TTR mutation,81 which is associated with a particularly aggressive form of TTR amyloidosis. The solved structure of other mutant TTRs points toward a destabilization of the tetrameric structure of the protein, suggesting that these mutations increase the amyloidogenicity of TTR by reducing formation of the physiologic soluble tetramer and increasing the formation of the proamyloidogenic monomer.41 The importance of TTR tetramer stability is further demonstrated in studies of patients who are compound heterozygotes for the TTR Met 30 and TTR Met 119 mutations. As stated above, it has been noted that these compound heterozygotic patients have a more benign disease course than FAP patients heterozygous for the TTR Met 30 mutation. Studies conducted using semidenaturing isoelectric focusing have shown that TTR Met 30 has a higher tendency for dissociation of the tetramer into monomers than wild-type TTR. Conversely, TTR Met 119 showed higher resistance to dissociation into monomers than wild-type TTR. When TTRs from compound Met 30/Met 119 heterozygotes were studied, they showed the same resistance to dissociation as wild-type TTR.4 Thus the TTR Met 119 mutation

1928

Diseases of the Peripheral Nervous System

appears to have a protective effect by counteracting the weaker subunit interactions of the Met 30 tetramers.80 Many clinical and epidemiologic observations, including the variations in phenotypes and age of onset in patients with the same mutation, are difficult to explain by limiting pathogenetic studies to TTR amyloidogenesis. Other genetic factors, including perhaps polymorphisms of the nonfibrillary amyloid constituents, and environmental factors that might modify the phenotype have yet to be described. Despite the advances in the understanding of the pathogenesis of TTR amyloidosis, many questions remain, particularly regarding the cause of the neuropathy in TTR amyloidosis. The initial hypothesis was that the neuropathy was due to ischemia based on the finding of amyloid deposits in the walls of vasa nervorum. This is now thought to be unlikely because ischemia would be expected to cause the loss of large myelinated fibers first, unlike in amyloid neuropathy, where the small myelinated and unmyelinated fibers are affected first and are known to be less sensitive to ischemia. The second hypothesis suggests that localized compression of the Schwann cells in the endoneurium could lead to the neuropathy. It is already known that carpal tunnel syndrome, seen in many types of TTR amyloidosis, is probably due to a compression neuropathy secondary to amyloid deposition in the closed space of the carpal tunnel. Against the compression hypothesis for the neuropathy generally is the fact that the unmyelinated fibers, which are primarily affected in TTR amyloidosis, are more resistant to compression that myelinated fibers. The third and more likely hypothesis speculates as to whether amyloid deposits could have a toxic or metabolic effect on the endoneurium. A recent study has suggested that TTR fibrils may bind to the receptor for advanced glycation end products on specific cell targets and induce a stress response that may result in peripheral nerve dysfunction.87 TTR deposition has been shown to occur in peripheral nerve years before amyloid fibril formation.86 There is also evidence that the toxicity may be related to early stages of fibril formation, whereas the mature full-length fibrils may represent an inert end stage.6 Transgenic mouse studies have been used to try to understand the pathogenesis of the neuropathy in TTR amyloidosis further, but transgenic mice, whether expressing the human TTR Met 30 mutation or the human TTR Met 30 mutation and various regulatory regions, have not yet shown amyloid deposition in the peripheral nerves. The pattern of amyloid deposition in these mice is otherwise similar to that seen in autopsy cases of human TTR amyloidosis.92 Further research is obviously needed to elucidate the exact mechanisms underlying the neuropathy in TTR amyloidosis.

Differential Diagnosis The characteristic combination of a predominantly small fiber neuropathy with autonomic involvement and cardiac disease should always raise the suspicion of hereditary

amyloid neuropathy, especially if there is a family history. Diagnostic difficulties arise early in the disease when the patient may present with only a sensory neuropathy mainly involving the small fibers and no apparent family history. The main differentials for this type of neuropathy are diabetes mellitus, leprosy, human immunodeficiency virus infection, toxic neuropathy resulting from neurotoxic drugs, occasionally vasculitic neuropathy, and idiopathic small fiber neuropathy. Amyloid becomes much more likely when there is a significant autonomic component, but diabetes is still a possibility, as are hereditary sensory and autonomic neuropathy, Fabry’s disease, and Tangier disease if the history is long. Most of the above differentials are easily excluded, but differentiation between the various types of amyloid neuropathy, especially AL amyloidosis and TTR amyloidosis, can be difficult without immunohistochemical staining and molecular diagnosis (see below).

Diagnosis The evaluation of a patient with TTR amyloidosis is initially concerned with establishing the diagnosis and then with evaluating the extent and stage of the disease. The diagnosis of TTR amyloidosis requires the diagnosis of amyloidosis first and then the characterization of the nature of the constituent amyloid fibril protein. The diagnosis requires a high level of clinical suspicion. It should always be considered in patients with a neuropathy and a known family history of TTR amyloidosis and in patients with a known family history of TTR amyloidosis who do not have a neuropathy but who present with a cardiomyopathy, carpal tunnel syndrome, or vitreous deposits. The diagnosis should also be kept in mind in patients with an unexplained small fiber neuropathy, especially with autonomic involvement, or a cardiomyopathy regardless of family history. Once the clinical suspicion is present, the diagnosis of amyloidosis can be made by direct examination of biopsy material from the rectum, peripheral nerve, heart, subcutaneous fat, or other tissue. The actual tissue biopsied will depend on the clinical presentation, but in practice a rectal biopsy is often the initial choice if amyloid is suspected, followed if necessary by a peripheral nerve biopsy. Rectal biopsy is considered more sensitive than nerve biopsy presumably because the patchy nature of amyloid deposits may make the diagnosis more difficult in a limited nerve biopsy. The sensitivity of gastrointestinal biopsies is thought to be 85%.41 The most widely used technique for diagnosis is a combination of alkaline Congo red and polarizing filters to demonstrate the characteristic apple-green birefringence.75 The nature of the amyloid fibril protein is characterized by immunohistochemistry using the indirect immunoperoxidase method, in which monoclonal antibodies are used that are directed against ␭ and ␬ light chain–derived amyloid, amyloid A, and TTR.

1929

Hereditary Amyloid Neuropathy

FIGURE 83–2 Serial posterior whole-body 123I-labeled serum amyloid P (SAP) component scintigraphy in a patient with familial amyloidosis associated with variant transthyretin, showing regression of amyloid deposits in the spleen and kidneys following orthotopic liver transplantation (OLT). From left to right: Before OLT, 15 months post-OLT, and 36 months post-OLT. (Courtesy of Professor Philip Hawkins, National Amyloidosis Centre, Royal Free Hospital, London, U.K.)

The method is not completely reliable because there may not be sufficient amyloid present and the antibodies, especially the light chain antibodies, may lack specificity.1 If AL amyloid is suspected and the antibodies to TTR are negative, it may be worth pursuing immunofixation electrophoresis of serum and urine even if the light chain antibodies were also negative. If TTR amyloid is confirmed or if there is no definite diagnosis of AL amyloidosis, the next step is to search for variant TTR. In practice this usually means proceeding to direct sequencing of the TTR gene, but other methods can be used initially. In heterozygotic patients with TTR amyloidosis, both normal and variant TTR circulate in the blood. Variant TTR can be detected by a variety of means, including isoelectric focusing in urea gradients,3 immunoassay using specific monoclonal antibodies,33 and electrospray ionization mass spectrometry after immunoprecipitation with polyclonal TTR antibodies.9 Direct sequencing of the TTR gene is used to characterize new mutations, although other methods to screen for the presence of a TTR mutation may be used initially. These methods are continuously being updated and include single-strand conformation polymorphism analysis,66 denaturing gradient gel electrophoresis,23 and the chemical cleavage of mismatch method.23 TTR point mutations in a family with a known mutation can be detected using the polymerase chain reaction with subsequent digestion with an appropriate restriction enzyme. The above molecular genetic methods can be used not only for diagnostic testing but also for presymptomatic and prenatal testing in families with known mutations. Both require pre- and post-test genetic counseling. One of the difficulties in counseling for presymptomatic testing, especially in the rarer TTR mutations, is the lack of accurate information for penetrance. Once the diagnosis has been made, the extent and stage of the disease need to be accurately assessed, especially if liver transplantation (see Treatment below) is being considered.

The extent of systemic amyloid deposition can be assessed using scintigraphic studies with 123I-labeled SAP, because SAP is deposited in all types of amyloidosis38 (Fig. 83–2). SAP scanning also can occasionally be used diagnostically, but it is not widely available. The main limitation of this technique is that it does not detect amyloid deposition in the peripheral nerves or CNS. Peripheral nerve involvement can be assessed clinically with nerve conduction studies, electromyography, and occasionally nerve biopsy. Detailed autonomic testing is especially important in pre–liver transplant patients. CNS involvement, if present, can be studied by MRI with gadolinium and CSF analysis. An ophthalmologic assessment is necessary to look for vitreous deposits. Cardiac assessment is important and requires at the least an electrocardiogram (ECG), an echocardiogram, and sometimes a 24-hour ECG. Occasionally a cardiac biopsy is necessary. Renal involvement, although not usually symptomatic, needs to be assessed. The patient’s general health, mobility, and nutritional status are also important.

Treatment Until the 1990s, the treatment of TTR amyloidosis was limited to rehabilitative measures and symptomatic therapies. These measures remain important in the overall management of the patient. Neuropathic pain can be very troublesome, especially early on in the disease, and is treated with the usual agents that are used for this type of pain, including various anticonvulsants and antidepressants. The autonomic neuropathy often requires treatment, including pharmacologic treatment of the postural hypotension with drugs such as fludrocortisone and midodrine.1 The known cardiac complications may need addressing, including the insertion of a pacemaker for cardiac conduction disturbances. Other non-neurologic manifestations, including the gastrointestinal, vitreous, and renal disturbances, may need addressing.

1930

Diseases of the Peripheral Nervous System

Since the early 1990s, liver transplantation has been used in the treatment of TTR amyloidosis. Liver transplantation was first suggested as a potential treatment for TTR amyloidosis because approximately 90% of TTR is produced in the liver. The first liver transplants for TTR amyloidosis were performed in two Swedish patients in 1990.40 As predicted, the biochemical effect of liver transplantation is good as demonstrated by a dramatic reduction in variant TTR in plasma.39 A SAP scintigraphy study has also demonstrated a diminished amyloid load postoperatively (see Fig. 83–2), with 50% of TTR amyloidosis Met 30 patients who had a liver transplant showing a decrease in activity indicating diminished amyloid load.77 As of 2001, more than 490 liver transplants have been done worldwide for TTR amyloidosis according to the FAP World Transplant Register.37 This register is based in Sweden, and its purpose is to monitor the international results of liver transplantation for TTR amyloidosis. The figure of 490 is an underestimate because not all transplants are reported to the register. The transplanted liver is usually taken from a cadaver, but both living donor transplants and domino transplants have been used. More than 20 Japanese TTR amyloidosis patients have had living donor transplants because full cadaver liver transplantation is uncommon in Japan, where brain death has not been widely accepted as a definition of death.44 Domino liver transplantation arose because liver amyloid deposition in TTR amyloidosis is minimal.82 In this procedure, after a TTR amyloidosis patient has been transplanted with a new liver, the liver from the TTR amyloidosis patient is then transplanted into a nonamyloidotic patient who needs a new liver usually for palliative reasons, such as a liver neoplasm.41 A recent follow-up of an average of 24 months in seven such patients who received liver transplants from TTR amyloidosis patients showed no evidence of development of early amyloidosis.16 There have now been 82 domino liver transplants done with livers from TTR amyloidosis patients, reflecting the overall shortage of livers for transplantation.37 Although there is now a consensus of opinion that liver transplantation halts or slows the progression of TTR amyloidosis in most TTR amyloidosis Met 30 patients, it is clear that better data are needed to establish efficacy. Prospective controlled trials have not been done. Most of the information is available for TTR amyloidosis Met 30 patients because these account for 80% of patients transplanted in the worldwide register.36 There is some suggestion that the patients with mutations other than Met 30 do not do as well, especially from the cardiac point of view (see below), but because the numbers for patients transplanted with differing (non–Met 30) mutations are small, this suggestion has to be interpreted with caution. General well-being, gastrointestinal symptoms, and nutritional state are the most frequent indices reported to improve.19

If the transplant is done at a relatively early stage in the disease, when patients have only a moderate neuropathy, there is clinical and electrophysiologic stability in the neuropathy in 76% of transplanted patients according to one series.1 The same study showed a marked reduction in axonal loss by a histometric comparative study between seven transplanted TTR amyloidosis patients and four control nontransplanted patients. In other studies the response of the peripheral neuropathy to transplantation varied from lack of neuropathy progression to some improvement post-transplantation.90 The conclusion to date is that a moderate improvement in the peripheral neuropathy is seen in some patients, particularly if transplanted early in the disease course. Early reports suggested that the symptoms of autonomic neuropathy improved post-transplantation, although this has not been confirmed in patients with short- or longterm follow-up by noninvasive autonomic testing.2 Despite this, some studies do report an improvement in some of the autonomic symptoms.15 Most TTR amyloidosis Met 30 patients seem to have unchanged cardiac amyloid deposition post-transplantation, and indeed improvement has been noted in some patients. Many patients require pacemakers preoperatively. There have been a number of worrying reports of progression of cardiac disease post-transplantation in patients with other TTR point mutations if there was echocardiographic evidence of ventricular wall or valve thickening preoperatively.71,88 This is thought to be due to wild-type TTR being deposited as amyloid on existing amyloid deposits.103 It may be necessary to consider combined liver and heart transplants in these patients. One recent report has also described cardiac progression postoperatively in TTR amyloidosis Met 30 patients.65 The global mortality from liver transplantation is about 21%, with a 5-year survival of 78%.37 The surgery is usually technically straightforward, but perioperative circulatory instability secondary to autonomic dysfunction can be problematic. Arrhythmias can also occur perioperatively, and pacemakers are often fitted preoperatively. Death is most common in the first 6 months postoperatively, usually as a result of infections or cardiocirculatory problems. Mortality has been related to the length of the disease and nutritional status preoperatively. Although symptomatic renal disease is not common preoperatively, renal dysfunction can be a major postoperative problem, so detailed preoperative renal assessment is mandatory. Because of the high mortality and the shortage of liver donors, liver transplantation should probably be directed more at the young and rapidly worsening cases than at the old and slowly advancing cases. There is a concern that, because some TTR is produced in the choroid plexus, CNS TTR syndromes may be seen post-transplantation, but this has not been documented to date. On the contrary, de novo vitreous amyloid deposits

1931

Hereditary Amyloid Neuropathy

have been documented post-transplantation, presumably resulting from local TTR synthesis in the retina.8 Liver transplantation has emerged as a treatment that should be considered in TTR amyloidosis patients. The important prognostic factors are modified body mass index, disease duration, and perhaps the presence of a non–Met 30 TTR mutation. Survival after transplantation appears to be closely related to nutritional status pretransplantation, having a disease duration of less than 7 years, and whether autonomic involvement is present.37 There probably is not yet enough information about transplants in patients with non–Met 30 TTR mutations to say that having such a mutation is a definite bad prognostic factor. Older patients may also not do as well if they have other medical problems. The optimum time for liver transplantation has yet to be determined, but the evidence suggests that it should be considered early.90 This is particularly important in younger patients who are otherwise healthy. In older patients, the duration and severity of the disease, the coexistence of other medical disorders, and the mortality of the procedure need all to be considered before a transplant is recommended. Although liver transplantation is the first treatment for TTR amyloidosis and should be considered in affected patients, it is associated with significant mortality, and other safer and more effective treatments are being researched. Both plasmapheresis and immunoadsorption have been tried with the aim of reducing the high blood levels of variant TTR, but neither treatment has been shown to be useful in the long term. A gene therapy approach using ribozymes to degrade variant and native TTR messenger RNA and thus suppress TTR production has been successful in vitro, and this approach may be a potential treatment for TTR amyloidosis.93 Currently various chemotherapeutic approaches for the treatment of TTR amyloidosis are being researched and look promising. Several nonsteroidal anti-inflammatory drugs, including flufenamic acid, diclofenac, flurbiprofen, and resveratrol, and other structurally similar compounds have been shown to inhibit the formation of amyloid fibrils in vitro.51 This may lead to the development of similar drugs that are useful therapeutically. Recently the drug R-1-[6-[R-2-carboxypyrrolidin-1-yl]-6-oxo-hexanoyl]pyrrolidine-2-carboxylic acid has been developed that both inhibits SAP binding to amyloid fibrils and leads to the rapid clearance of SAP by the liver with redistribution of SAP from the tissues to the plasma.67 Because SAP binds to fibrils in all types of amyloidosis and is known to protect amyloid fibrils from degradation, this approach is potentially very useful. Preliminary pilot studies in systemic amyloidosis, including some patients with TTR amyloidosis, have confirmed the rapid clearance of SAP from the blood following an infusion of this drug. This treatment needs further evaluation, but the development of this drug and the other developments discussed above raise the possibility of alternative

treatments to liver transplantation for TTR amyloidosis in the future.

Prognosis The average life expectancy for patients with untreated TTR amyloidosis associated with neuropathy is about 10 years. There are ethnic differences in survival in patients with the same mutation, with Swedish TTR amyloidosis Met 30 patients having an average life expectancy of 14 years while that of French patients with TTR amyloidosis Met 30 is less than 10 years.69 Some less common TTR point mutations have a more aggressive course, as described above. The 5-year survival for patients with TTR amyloidosis who have had a liver transplant is 78%.37 This figure is likely to improve further as the optimum conditions and timing for liver transplantation are determined and as newer genetic and chemotherapeutic treatments are developed.

APOLIPOPROTEIN A-I AMYLOIDOSIS In 1969, Van Allen described a kindred in Iowa with familial amyloid polyneuropathy associated with a high incidence of gastric ulcers and renal failure.98 The age of onset of the Iowa type of amyloidosis is in the 30s, and the neuropathy is similar to that of TTR amyloidosis, beginning in the lower limbs with small fiber symptoms and progressing to the upper limbs, with motor involvement occurring later than sensory involvement. Autonomic neuropathy is not prominent. In the original family, two features, a high incidence of both gastric ulcers and renal failure, distinguished the disease from TTR amyloidosis, which is why this type of amyloidosis was classified separately as FAP type III. Postmortem studies confirm amyloid deposition throughout the peripheral nervous system, including nerves, roots, and dorsal root ganglia.98 Apolipoprotein A-I, the gene for which is on chromosome 11, was found to be the constituent fibril protein in 1988,62 and in 1990 an arginine-for-glycine substitution at position 26 of the protein was reported as the causative mutation of this autosomal dominant condition.63 Apolipoprotein A-I is the major protein constituent of highdensity lipoprotein, and the constituent amyloid protein in apolipoprotein A-I amyloidosis is the amino-terminal 83 residues of the mature 243–amino-acid-long apolipoprotein A-I with the Gly26Arg mutation.63 It is of interest that, although apolipoprotein A-I itself is not predicted to have extensive beta pleated structure, the first 55 residues of the 83–amino-terminal fragment referred to above are predicted to have mostly beta pleated structure. This type of amyloidosis has only been described in one kindred, but there are now eight different apolipoprotein A-I mutations associated with systemic non-neuropathic amyloidosis,32 usually with marked major organ involvement including

1932

Diseases of the Peripheral Nervous System

liver, renal, and cardiac. The Gly26Arg apolipoprotein A-I mutation has also been described in non-neuropathic systemic amyloidosis. The diagnosis is established by demonstrating amyloid on an appropriate tissue biopsy with subsequent immunohistochemical staining for apolipoprotein A-I. Sequencing of the apolipoprotein A-I gene reveals a guanine-to-cysteine mutation giving rise to the Gly26Arg substitution. Symptomatic treatment (e.g., painful neuropathy) as described for TTR amyloidosis is important, but most patients eventually develop renal failure requiring dialysis. Although apolipoprotein A-I is partially produced in the liver, liver transplantation to reduce the production of the abnormal precursor protein is not yet in use for this type of amyloid neuropathy. In apolipoprotein A-I non-neuropathic systemic amyloidosis, organ transplantation is recommended for severe organ (e.g., liver or kidney) failure. A recent report of hepatorenal transplantation in a patient with hereditary non-neuropathic systemic amyloidosis and the apolipoprotein A-I Gly26Arg mutation (with severe renal and hepatic involvement) is of interest because the plasma levels of variant apolipoprotein A-I decreased by 50% post-transplantation, and there was evidence of amyloid regression in the heart and spleen postoperatively as quantified by iodine-123 SAP scanning.32 This raises the possibility that liver transplantation may be a future treatment for apolipoprotein A-I amyloidosis, but the authors suggested that this is not a treatment that can be recommended yet except for liver dysfunction.

GELSOLIN AMYLOIDOSIS Gelsolin amyloidosis was first described in a Finnish kindred by Meretoja in 1969.59 Clinically this type of amyloidosis usually presents in the 30s with corneal lattice dystrophy, which may be asymptomatic. This is due to amyloid deposition in the corneal branches of the trigeminal nerve.58 This is followed by a progressive cranial neuropathy. The facial nerve is the cranial nerve most commonly involved, with the upper fibers, as manifested by forehead muscle weakness, initially being affected. Other cranial nerves may also be affected, including the vestibulocochlear, hypoglossal, and trigeminal, with varying clinical manifestations.75 Peripheral nerve involvement does occur but is usually a mild sensory neuropathy. There also may be mild autonomic involvement. Neurophysiologic tests show a higher incidence of carpal tunnel syndrome than is suspected clinically. The neuropathy is predominantly axonal, but slow nerve conduction velocities, prolonged distal motor latencies, and conduction block all point toward an additional demyelinating element.49 The facial skin is involved, at first appearing thickened but becoming lax with time. Minor neuropsychological, MRI, and evoked potential abnormalities have been described suggesting

mild CNS involvement.50 A recent postmortem study has shown widespread spinal, cerebral, and meningeal amyloid angiopathy associated with deposition of gelsolin amyloid.48 Cardiac involvement is rare clinically but has been shown in autopsy studies.47 The fibril protein in this type of amyloidosis is an abnormal fragment of the plasma protein gelsolin, a calciumbinding protein that fragments actin filaments.105 Gelsolin is mainly derived from muscle. The gene for gelsolin is located on chromosome 9 and codes for two separate proteins, one intracellular and one extracellular with an additional 25 amino acid residues, by starting at two different transcriptional sites. There have been two mutations described in the gelsolin gene associated with this type of amyloidosis. The first is a substitution of asparagine for aspartic acid at residue 187 (position 15 of the amyloid protein).55 This has been described in over 200 Finnish kindreds and also in patients of Dutch,25 Japanese,91 and U.S. origin.31,35 We have seen a patient with this mutation of Russian ancestry (personal observation). Two homozygotes for this mutation have been described, both more severely affected than heterozygotes. The second mutation involves the same codon, with a tyrosine-for–aspartic acid substitution at residue 187, and has been reported in a Danish family and a Czech family.25 The amyloid fibrils in this type of amyloidosis are composed of the internal degradation products of an abnormal fragment of gelsolin that begins at position 173 of the plasma protein. An abnormal 65-kDa gelsolin species has been shown to be the circulating precursor of tissue amyloid in this type of amyloidosis. Residue 187 is thought to be a critical site where a substitution of an amino acid with an uncharged (asparagine) or hydrophobic (tyrosine or valine) side chain creates a conformation that is highly amyloidogenic.57 Recently it has been discovered that the protease furin is responsible for gelsolin protein proteolysis-induced amyloidosis in familial amyloidosis, Finnish type.18 The two described mutations prevent Ca2⫹ from binding to domain 2 and allow gelsolin to misfold, allowing furin access to a site not normally exposed in the wild-type protein.78 The diagnosis is established by biopsy with amyloid being demonstrated as described above. The gene can be sequenced to look for causative mutations. There is no specific treatment other than symptomatic treatment. Liver transplantation is not an option because gelsolin is not produced in the liver. There is interest in developing furin inhibitors as a therapeutic strategy.78

ACKNOWLEDGMENTS Dr. Michael Groves is kindly acknowledged for help in producing Figure 83–1. Professor Philip Hawkins is kindly acknowledged for supplying Figure 83–2.

Hereditary Amyloid Neuropathy

REFERENCES 1. Adams, D.: Hereditary and acquired amyloid neuropathies. J. Neurol. 248:647, 2001. 2. Adams, D., Samuel, D., Goulon-Goeau, C., et al.: The course and prognostic factors of familial amyloid polyneuropathy after liver transplantation. Brain 123:1495, 2000. 3. Altland, K., Becker, P., and Banzhoff, A.: Paraffin oil protected high resolution hybrid isoelectric focusing for the demonstration of substitutions of neutral amino acids in denatured proteins: the case of four human transthyretin (prealbumin) variants associated with familial amyloidotic polyneuropathy. Electrophoresis 8:293, 1987. 4. Alves, I. L., Hays, M. T., and Saraiva, M. J. M.: Comparative stability and clearance of [Met30] transthyretin and [met119] transthyretin. Eur. J. Biochem. 249:662, 1997. 5. Alves, I. L., Jacobson, D. R., Torres, M. F., et al.: Transthyretin Ser 6 as a natural polymorphism in familial amyloidotic polyneuropathy. Amyloid 3:242, 1996. 6. Andersson, K., Olofsson, A., Nielsen, E. H., et al.: Only amyloidogenic intermediates of transthyretin induce apoptosis. Biochem. Biophys. Res. Commun. 294:309, 2002. 7. Andersson, R.: Hereditary amyloidosis with polyneuropathy. Acta Med. Scand. 188:85, 1970. 8. Ando, Y., Ando, E., Tanaka, Y., et al.: De novo amyloid synthesis in ocular tissue in familial amyloidotic polyneuropathy after liver transplantation. Transplantation 62:1037, 1996. 9. Ando, Y., Ohlsson, P.-I., Suhr, O., et al.: A new simple and rapid screening method for variant transthyretin (TTR) related amyloidosis. Biochem. Biophys. Res. Commun. 228:480, 1996. 10. Andrade, C.: A peculiar form of peripheral neuropathy: familial atypical generalized amyloidosis with special involvement of the peripheral nervous system. Brain 75:408, 1952. 11. Araki, S., Mawatari, S., Ohta, M., et al.: Polyneuritic amyloidosis in a Japanese family. Arch. Neurol. 18:593, 1968. 12. Araki, S., and Shigehira, Y.: Pathology of familial amyloidotic polyneuropathy with TTR Met 30 in Kumamoto, Japan. Neuropathology 20:S47, 2000. 13. Benson, M. D. : Familial amyloid polyneuropathy. Trends Neurosci. 12:88, 1989. 14. Benson, M. D.: Inherited amyloidosis. J. Med. Genet. 28:73, 1991. 15. Bergethon, P., Sabin, T., Dewis, D., et al.: Improvement in the polyneuropathy associated with familial amyloid polyneuropathy after liver transplantation. Neurology 47:944, 1996. 16. Bittencourt, P. L., Couto, C. A., Leitao, R. M., et al.: No evidence of de novo amyloidosis in recipients of domino liver transplantation: 12 to 40 months (mean 24 month) follow-up. Amyloid 9:194, 2002. 17. Brett, M., Persey, M. R., Reilly, M. M., et al.: Transthyretin Leu12Pro is associated with systemic, neuropathic and leptomeningeal amyloidosis. Brain 122:183, 1999. 18. Chen, C. D., Huff, M. E., Matteson, J., et al.: Furin initiates gelsolin familial amyloidosis in the Golgi through a defect in Ca2⫹ stabilization. EMBO J. 20:6277, 2001. 19. Coehlo, T.: Familial amyloid polyneuropathy: new developments in genetics and treatment. Curr. Opin. Neurol. 9:355, 1996.

1933

20. Cohen, A. S.: History of amyloidosis. J. Intern. Med. 232:509, 1992. 21. Cohen, A. S., and Calkins, E.: Electron microscopic observations on a fibrous component of amyloid of diverse origins. Nature 183:1202, 1959. 22. Costa, P., Figueira, A. S., and Bravo, R. R.: Amyloid fibril protein related to pre-albumin in familial amyloidotic polyneuropathy. Proc. Natl. Acad. Sci. U. S. A. 75:4499, 1978. 23. Cotton, R. G. H.: Detection of single base changes in nucleic acids. Biochem. J. 263:1, 1989. 24. Coutinho, P., Martins Da Silva, A., Lopes Lima, J., and Resende Barbosa, A.: Forty years of experience with type 1 amyloid neuropathy: review of 483 cases. In Glenner, G. G., Pinho, P., Costa, E., and Falcao De Freitas, A. (eds.): Amyloid and Amyloidosis. Amsterdam, Excerpta Medica, p. 88, 1980. 25. de le Chapelle, A., Tolvanan, R., Boysen, G., et al.: Gelsolin derived familial amyloidosis caused by asparagine or tyrosine substitution for aspartic acid at residue 187. Nature 2:157, 1992. 26. de Navasquez, S., and Treble, H. A.: A case of primary generalized amyloid disease with involvement of the nerves. Brain 61:116, 1938. 27. Dwulet, F. E., and Benson, M. D.: Polymorphism of human plasma thyroxine binding prealbumin. Biochem. Biophys. Res. Commun. 114:657, 1983. 28. Dwulet, F. E., and Benson, M. D.: Characterization of a transthyretin (pre-albumin) variant associated with familial amyloidotic polyneuropathy, type II (Indiana/Swiss). J. Clin. Invest. 78:880, 1986. 29. Dyck, P. J., and Lambert, E. H.: Dissociated sensation in amyloidosis. Arch. Neurol. 20:490, 1969. 30. Fonseca, C., Ceia, F., Nogueira, J. S., et al.: Myocardiopathy caused by Portuguese-type familial amyloidotic polyneuropathy: sequential morphological and functional study of 60 cases. Rev. Port. Cardiol. 10:909, 1991. 31. Ghiso, J., Haltia, M., Prelli, F., et al.: Gelsolin variant (ASN 187) in familial amyloidosis, Finnish type. Biochem. J. 272:827, 1990. 32. Gillmore, J. D., Stangou, A. J., Tennent, G. A., et al.: Clinical and biochemical outcome of hepatorenal transplantation for hereditary systemic amyloidosis associated with apolipoprotein A1 Gly26Arg. Transplantation 71:986, 2001. 33. Goldsteins, S. G., Persson, H., Andersson, K., et al.: Exposure of cryptic epitopes on transthyretin only in amyloid and in amyloidogenic mutants. Proc. Natl. Acad. Sci. U. S. A. 96:3108, 1999. 34. Goren, H., Steinberg, M. C., and Farboody, G. H.: Familial oculoleptomeningeal amyloidosis. Brain 103:473, 1980. 35. Gorevic, P. D., Munoz, P. C., Gorgone, G., et al.: Amyloidosis due to mutation of the gelsolin gene in an American family with lattice corneal dystrophy type II. N. Engl. J. Med. 325:1780, 1991. 36. Harding, A. E., and Reilly, M. M.: Molecular genetics of inherited neuropathies. In Asbury, A. K., and Thomas, P. K. (eds.): Peripheral Nerve Disorders 2. Oxford, ButterworthHeinemann, p. 118, 1995. 37. Herlinus, G., Larsson, M., and Erizon, B. G.: FAP World Transplant Register and Domino/Sequential Register

1934

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52. 53.

54. 55.

Diseases of the Peripheral Nervous System

Update: Yearly Report. Stockholm, FAP World Transplant Register, 2001. Holmgren, G., Ando, Y., and Wikström, L.: Discordant symptoms in monozygotic twins with familial amyloidotic polyneuropathy (FAP TTR met30) amyloid. Int. J. Exp. Clin. Invest. 4:178, 1997. Holmgren, G., Erizon, B.-G., Groth, C.-G., et al.: Clinical improvement and amyloid regression after liver transplantation in hereditary transthyretin amyloidosis. Lancet 341:1113, 1993. Holmgren, G., Steen, L., Ekstedt, J., et al.: Biochemical effect of liver transplantation in two Swedish patients with familial amyloidotic polyneuropathy (FAP-met30). Clin. Genet. 40:242, 1991. Hund, E., Linke, R. P., Willig, F., and Grau, A.: Transthyretin-associated neuropathic amyloidosis. Neurology 56:431, 2001. Hungo, M., and Ikeda, S.: Echocardiographic assessment of the evolution of amyloid heart disease: a study with familial amyloid polyneuropathy. Circulation 73:249, 1986. Ii, S., and Sommer, S. S.: The high frequency of TTR Met 30 in familial amyloidotic polyneuropathy is not due to a founder effect. Hum. Mol. Genet. 2:1303, 1993. Ikeda, S., Nakazato, M., Ando, Y., and Sobue, G.: Familial transthyretin-type amyloid polyneuropathy in Japan. Neurology 58:1001, 2002. Jacobson, D. R., McFarlin, D. E., Kane, I., and Buxbaum, J. N.: Transthyretin Pro55, a variant associated with early-onset, aggressive, diffuse amyloidosis with cardiac and neurologic involvement. Hum. Genet. 89:353, 1992. Katsuragi, S., Miyakawa, T., Ando, Y., and Terazaki, H.: High-resolution ultrastructure of amyloid fibrils in familial amyloid polyneuropathy. J. Electron Microsc. (Tokyo) 49:579, 2000. Kiuru, S., Matikainen, E., Kupari, M., et al.: Autonomic nervous system and cardiac involvement in familial amyloidosis, Finnish type (FAF). J. Neurol. Sci. 126:40, 1994. Kiuru, S., Salonen, O., and Haltia, M.: Gelsolin-related spinal and cerebral amyloid angiopathy. Ann. Neurol. 45:305, 1999. Kiuru, S., and Seppalainen, A. M.: Neuropathy in familial amyloidosis, Finnish type (FAF): electrophysiological studies. Muscle Nerve 17:299, 1994. Kiuru, S., Seppalainen, A.-M., Salonen, S., et al.: CNS abnormalities in patients with familial amyloidosis of the Finnish type (FAF). Amyloid Int. J. Exp. Clin. Invest. 2:22, 1995. Klabunde, T., Petrassi, H. P., Oza, V. B., et al.: Rational design of potent human transthyretin amyloid disease inhibitors. Nat. Struct. Biol. 7:312, 2000. Königstein, H.: Über Amyloidose der haut. Arch. Dermatol. Syphilis 148:330, 1925. Kotani, N., Hatori, T., Yamagata, S., et al.: Transthyretin Thr60Ala Appalachian-type mutation in a Japanese family with familial amyloidotic polyneuropathy. Amyloid 9:31, 2002. Kyle, R. A.: Amyloidosis. J. Intern. Med. 232:507, 1992. Levy, E., Haltia, M., Fernandez-Madrid, I., et al.: Mutation in gelsolin gene in Finnish hereditary amyloidosis. J. Exp. Med. 172:1865, 1990.

56. Mahloudji, M., Teasdall, R. D., Adamkiewics, J. J., et al.: The genetic amyloidoses with particular reference to hereditary neuropathic amyloidosis, type II (Indiana or Rukavina type). Medicine 48:1, 1969. 57. Maury, C. P., Nurmiaho-Lassila, E. L., and Rossi, H.: Amyloid fibril formation in gelsolin derived amyloidosis: definition of the amyloidogenic region and evidence of accelerated amyloid formation of mutant Asn-187 and Tyr187 gelsolin peptides. Lab. Invest. 70:558, 1994. 58. Mendell, J. R.: Familial amyloid polyneuropathies. In Mendell, J. R., Kissel, J. T., and Cornblath, D. R. (eds.): Diagnosis and Management of Peripheral Nerve Disorders. New York, Oxford University Press, p. 477, 2001. 59. Meretoja, J.: Familial systemic paramyloidosis with lattice dystrophy of the cornea, progressive cranial neuropathy, skin changes and various internal symptoms. Ann. Clin. Res. 1:314, 1969. 60. Moses, A. C., Rosen, H. N., Moller, D. E., et al.: A point mutation in transthyretin increases affinity for thyroxine and produces euthyroid hyperthyroxinemia. J. Clin. Invest. 86:2025, 1986. 61. Murakami, T., Maeda, S., Yi, S., et al.: A novel transthyretin mutation associated with familial amyloidotic polyneuropathy. Biochem. Biophys. Res. Commun. 182:520, 1992. 62. Nichols, W. C., Dwulet, F. E., Liepnieks, J., and Benson, M. D.: Variant apolipoprotein A-1 as a major constituent of a human hereditary amyloid. Biochem. Biophys. Res. Commun. 156:762, 1988. 63. Nichols, W. C., Gregg, R. E., Bryan Brewer, H., and Benson, M. D.: A mutation in apolipoprotein A-1 in the Iowa type of familial amyloidotic polyneuropathy. Genomics 8:318, 1990. 64. Nichols, W. C., Liepnieks, J. J., McKusick, V. A., and Benson, M. D.: Direct sequencing of the gene for Maryland /German familial amyloidotic polyneuropathy type II and genotyping by allele-specific enzymatic amplification. Genomics 5:535, 1989. 65. Olofsson, B.-E., Backman, C., Karp, K., and Suhr, O. B.: Progression of cardiomyopathy after liver transplantation in patients with familial amyloidotic polyneuropathy, Portuguese type 1. Transplantation 73:745, 2002. 66. Orita, M., Iwahana, H., and Sekiya, T.: Detection of polymorphisms of human DNA by gel electrophoresis as singlestrand conformation polymorphisms. Proc. Natl. Acad. Sci. U. S. A. 86:2766, 1989. 67. Pepys, M. B., Herbert, J., Hutchinson, W. L., et al.: Targeted pharmacological depletion of serum amyloid P component for treatment of human amyloidosis. Nature 417:254, 2002. 68. Pitkanen, P., Westermark, P., and Cornwell, G. G. III: Senile systemic amyloidosis. Am. J. Pathol. 117:391, 1984. 69. Planté-Bordeneuve, V., Lalu, T., Misrahi, M., et al.: Genotypic-phenotypic variations in a series of 65 patients with familial amyloid polyneuropathy. Neurology 51:708, 1998. 70. Planté-Bordeneuve, V., and Said, G.: Transthyretin related familial amyloidotic polyneuropathy. Curr. Opin. Neurol. 13:569, 2000. 71. Pomfret, E. A., Lewis, W. D., Jenkins, R. L., et al.: Effect of orthotopic liver transplantation on the progression of

Hereditary Amyloid Neuropathy

72.

73.

74. 75. 76.

77.

78. 79.

80. 81. 82.

83.

84.

85.

86.

87.

88.

89.

familial amyloidotic polyneuropathy. Transplantation 65:918, 1998. Reilly, M., Adams, D., Davis, M. B., et al.: Haplotype analysis of French, British and other European patients with familial amyloid polyneuropathy. J. Neurol. 242:664, 1995. Reilly, M. M., Adams, D., Booth, D. R., et al.: Transthyretin gene analysis in European patients with suspected familial amyloid polyneuropathy. Brain 118:849, 1995. Reilly, M. M., and King, R. H. M.: Familial amyloid polyneuropathy. Brain Pathol. 3:165, 1993. Reilly, M. M., and Staunton, H.: Peripheral nerve amyloidosis. Brain Pathol. 6:163, 1996. Reilly, M. M., Staunton, H., and Harding, A. E.: Familial amyloid polyneuropathy (TTR ala 60) in north west Ireland: a clinical, genetic and epidemiology study. J. Neurol. Neurosurg. Psychiatry 59:45, 1995. Rhyd, A., Suhr, O., Hietala, S.-O., et al.: Serum amyloid P scintigraphy in familial amyloid polyneuropathy: regression of visceral amyloid following liver transplantation. Eur. J. Nucl. Med. 25:709, 1998. Sacchettini, J. C., and Kelly, J. W.: Therapeutic strategies for human amyloid diseases. Nat. Rev. 1:267, 2002. Sandgren, O., Holmgren, G., and Lundgren, E.: Vitreous amyloidosis associated with homozygosity for the transthyretin methionine-30 gene. Arch. Ophthalmol. 108:1584, 1990. Saraiva, M. J. M.: Transthyretin amyloidosis: a tale of weak interactions. FEBS Lett. 498:201, 2001. Saraiva, M. J. M.: Transthyretin mutations in hyperthyroxinemia and amyloid diseases. Hum. Mutat. 17:493, 2001. Shaz, B. H., Gordon, F., Lewis, W. D., et al.: Orthotopic liver transplantation for familial amyloidotic polyneuropathy: a pathological study. Hum. Pathol. 31:40, 2000. Soares, M., Buxbaum, J., Coehlo, T., et al.: Genetic anticipation in Portuguese kindreds with familial amyloidotic polyneuropathy is unlikely to be caused by triplet repeat expansions. Hum. Genet. 104:480, 1999. Sousa, A., Andersson, R., Drugge, U., et al.: Familial amyloidotic polyneuropathy in Sweden: geographical distribution, age of onset, and prevalence. Hum. Hered. 43:288, 1993. Sousa, A., Coehlo, T., Barros, J., and Sequeiros, J.: Genetic epidemiology of familial amyloidotic polyneuropathy (FAP)-type 1 in Povoa do Varzim and Vila do Conde (north of Portugal). Am. J. Hum. Genet. 60:512, 1995. Sousa, M. M., Cardoso, I., Fernandes, R., et al.: Deposition of transthyretin in early stages of familial amyloidotic polyneuropathy: evidence for toxicity of nonfibrillar aggregates. Am. J. Pathol. 159:1993, 2001. Sousa, M. M., Yan, S. D., Fernandes, R., et al.: Familial amyloid polyneuropathy: receptor for advanced glycation end products-dependent triggering of neuronal inflammatory and apoptotic pathways. J. Neurosci. 21:7576, 2001. Stangou, A. J., Hawkins, P. N., Heaton, N. D., et al.: Progressive cardiac amyloidosis following liver transplantation for familial amyloid polyneuropathy. Transplantation 66:229, 1998. Staunton, H., Dervan, P., Linke, R. P., and Kelly, P.: Hereditary amyloid polyneuropathy in North West Ireland. Brain 110:1231, 1987.

1935

90. Suhr, O. B., Herlinius, G., Friman, S., and Ericzon, B.-G.: Liver transplantation for hereditary transthyretin amyloidosis. Liver Transpl. 6:263, 2000. 91. Sunada, Y., Shimizu, T., Nakase, H., et al.: Inherited amyloid polyneuropathy type IV (gelsolin variant) in a Japanese family. Ann. Neurol. 33:57, 1993. 92. Takaoka, Y., Tashiro, F., Yi, S., et al.: Comparison of amyloid deposition in two lines of transgenic mouse that model familial amyloidotic polyneuropathy, type 1. Transgenic Res. 6:261, 1997. 93. Tanaka, K., Yamada, T., Ohyagi, Y., et al.: Suppression of transthyretin expression by ribozymes: a possible therapy for familial amyloidotic polyneuropathy. J. Neurol. Sci. 183:79, 2001. 94. Terzaki, H., Ando, Y., Misumi, S., et al.: A novel compound heterozygote (FAP ATTR Arg104His/Val30Met) with high serum transthyretin (TTR) and retinal binding protein (RBP) levels. Biochem. Biophys. Res. Commun. 264:365, 1999. 95. Thomas, P. K., and King, R. H. M.: Peripheral nerve changes in amyloid neuropathy. Brain 97:395, 1974. 96. Togashi, S., Watanabe, H., Nagasaka, T., et al.: An aggressive familial amyloidotic polyneuropathy caused by a new variant transthyretin Lys 54. Neurology 53:637, 1999. 97. Uemichi, T., Liepnieks, J. J., and Benson, M. D.: A trinucleotide deletion in the transthyretin gene (delta V 122) in a kindred with familial amyloidotic polyneuropathy. Neurology 48:1667, 1997. 98. Van Allen, M. W., Frohlich, J. A., and Davis, J. R.: Inherited predisposition to generalized amyloidosis: clinical and pathological study of a family with neuropathy, nephropathy and peptic ulcer. Neurology 19:10, 1969. 99. Virchow, R.: Découverte d’une substance qui donne lieu au mêmes reactions chimiques que la cellulose vegetable, dans le corps human. C. R. Acad. Sci. (Paris) 37:492, 1853. 100. Wallace, M. R., Dwulet, F. E., Williams, E. C., et al.: Identification of a new hereditary amyloidosis prealbumin variant, Tyr 77, and detection of the gene by DNA analysis. J. Clin. Invest. 81:189, 1988. 101. Wallace, M. R., Naylor, S. L., Kluve-Beckerman, B., et al.: Localization of the pre-albumin gene to chromosome 18. Biochem. Biophys. Res. Commun. 129:753, 1985. 102. Westermark, P., Sletten, K., Johansson, B., and Cornwell, G. G. III: Fibril in senile systemic amyloidosis derived from normal transthyretin. Proc. Natl. Acad. Sci. U. S. A. 87:2843, 1990. 103. Yazaki, M., Tokuda, T., Nakamura, A., et al.: Cardiac amyloid in patients with familial amyloid polyneuropathy consists of abundant wild-type transthyretin. Biochem. Biophys. Res. Commun. 274:702, 2000. 104. Yazaki, M., Yamashita, T., Kincaid, J. C., et al.: Rapidly progressive amyloid polyneuropathy associated with a novel variant transthyretin serine 25. Muscle Nerve. 25:244, 2002. 105. Yin, H. L., Kwaitkowski, D. J., Mole, J. E., et al.: Structure and biosynthesis of cytoplasmic and secreted variants of gelsolin. J. Biol. Chem. 259:5271, 1984. 106. Yoshioka, K., Furuya, H., Saraiva, M. J. M., et al.: Haplotype analysis of familial amyloidotic polyneuropathy. Hum. Genet. 82:9, 1989.

84 Peripheral Nerve Diseases Associated with Mitochondrial Respiratory Chain Dysfunction MICHAEL G. HANNA AND PAULA CUDIA

Basic Features of the Mitochondrial Respiratory Chain The Mitochondrial Respiratory Chain Characteristics of Mitochondrial Genetics Making the Diagnosis of Mitochondrial Disease: Histopathology, Biochemistry, and Genetics

Common mtDNA Disease Phenotypes How Common Is Peripheral Nerve Dysfunction in Mitochondrial Respiratory Chain Disease? Neuropathology of Peripheral Nerve in Patients with Mitochondrial Disease

Human diseases caused by impaired respiratory chain function have been increasingly recognized in recent years.31,59 Although neurologic diseases are the most common consequence of such respiratory chain dysfunction, it is now apparent that virtually any tissue in the body can be affected. This is perhaps unsurprising when one considers the central role of the respiratory chain in normal cell and tissue function.31,59 The neurologic diseases associated with respiratory chain impairment are collectively known as the mitochondrial encephalomyopathies, a term that reflects the common involvement of both the central nervous system (CNS) and skeletal muscle in these patients. However, it is recognized that peripheral nerve dysfunction is not an infrequent accompaniment of neurologic respiratory chain disease. Peripheral nerve dysfunction consequent upon impaired respiratory chain activity may be a minor or subclinical part of a complex mitochondrial disease phenotype or, less commonly, may be a major or presenting clinical feature. One of the first clearly characterized cases that provided strong evidence that mitochondrial dysfunction could associate with peripheral nerve disease was reported by Peyronnard et al.37 in 1980. They

Mitochondrial Diseases with Neuropathy as a Major Clinical Feature Demyelinating Neuropathy Axonal Neuropathy Conclusion

described a 37-year-old man with progressive external ophthalmoplegia and ragged red muscle fibers (RRF) who also had a severe sensorimotor peripheral neuropathy. Histologic and morphometric analysis of the sural nerve revealed a marked loss of large myelinated fibers and also degeneration of axons. In contrast to the muscle biopsy findings, no paracrystalline inclusions were seen in the nerve mitochondria. However, the numbers of mitochondria per square micron were increased in Schwann cell cytoplasm. Subsequent studies have confirmed that axonal neuropathy can be caused by mitochondrial dysfunction; however, demyelinating neuropathy is also recognized. Over the years there has been controversy over how best to classify mitochondrial respiratory chain disease.2 Currently a combined clinical-genetic classification is favored. One of the problems encountered by any classification system is the marked clinical and genetic heterogeneity exhibited by these disorders. For example, it is recognized that a single clinical mitochondrial phenotype can be genetically heterogeneous. Conversely, a single mitochondrial DNA genotype may associate with totally different clinical phenotypes. 1937

1938

Diseases of the Peripheral Nervous System

Although most mitochondrial respiratory chain proteins are encoded in the nucleus, to date the majority of genetic defects causing mitochondrial disease are in mitochondrial DNA (mtDNA). Most of this chapter therefore focuses upon the relationship between mtDNA mutations and peripheral nerve disease. Recently, some nuclear mitochondrial diseases have been elucidated at the genetic level, and this group is likely to increase in the future. Relatively little is published about the relationship between nuclear genetic mitochondrial disorders and peripheral nerve disease, but we have also included mention of this. For reasons outlined, it is difficult to devise a neat classification of peripheral neuropathy in patients with mtDNA mitochondrial disease. However, we suggest the guidelines based on clinical and genetic considerations outlined in Tables 84–1 and 84–2. From a clinical viewpoint, patients fall broadly into two categories: those in whom peripheral nerve disease is a minor or subclinical feature and those in whom it is a major feature. From a mitochondrial genetic viewpoint, it is possible to consider peripheral nerve disease in the context of the three classes of mtDNA mutations outlined in Table 84–2. In this chapter we briefly review the basic features of the mitochondrial respiratory chain and its genetic control. We then describe the common mtDNA disease phenotypes. Finally, we consider the clinical, genetic, and pathologic features of the different types of peripheral nerve dysfunction reported in association with mitochondrial disease.

BASIC FEATURES OF THE MITOCHONDRIAL RESPIRATORY CHAIN Normal respiratory chain function is dependent on an elaborate interplay between the mitochondrial genome and the nuclear genome. It follows that human mitochondrial diseases are genetically heterogeneous. Early genetic mitochondrial research focused upon the smaller mitochondrial genome (mtDNA). This has proved to be a rich source of mutations that cause human mitochondrial disease. To date over 100 separate primary pathogenic mutations in mtDNA have been described in association with a bewildering array of clinical phenotypes.59 Much more recently, the first nuclear genes associated with mitochondrial disease have been reported.59

Table 84–1. Broad Categories of Mitochondrial Peripheral Nerve Dysfunction—a Clinical Classification Group 1: Peripheral nerve disease a minor feature of complex mitochondrial phenotype Group 2: Peripheral nerve disease a major/sole feature of mitochondrial disease

Table 84–2. Broad Categories of Mitochondrial Peripheral Nerve Dysfunction—a Genetic Classification A

B

C

mtDNA mutations in which peripheral neuropathy is usually present and is usually a major feature (e.g., multiple deletions in mtDNA, A8993T) mtDNA mutations in which peripheral neuropathy is often present, but is usually a minor or subclinical feature (e.g., A3243G, A8344G, single deletions of mtDNA) mtDNA mutations in which neuropathy is rarely present and, when present, is subclinical (e.g., A11778G)

The Mitochondrial Respiratory Chain The respiratory chain comprises four multi-polypeptide enzyme complexes (Fig. 84–1)—complex I (NADHubiquinone reductase); complex II (succinate-ubiquinone reductase); complex III (ubiquinol–cytochrome c reductase); and complex IV (cytochrome c oxidase)—and two mobile electron carriers: ubiquinone and cytochrome c. Together with ATP synthase (complex V), these complexes and electron carriers comprise the oxidative phosphorylation system. Electrons generated from reducing equivalents (NADH and FADH2) are passed between the complexes, resulting in energy release that allows proton pumping at complexes I, III, and IV. The generation of a proton gradient across the inner mitochondrial membrane is utilized by complex V to generate ATP from ADP. This process of oxidative phosphorylation results in regeneration of the universal energy currency, ATP, which is essential for most cellular functions. When one considers the central role of ATP in cell survival, it is perhaps unsurprising that virtually any organ system in the human body can be disrupted by mitochondrial respiratory chain disease.

Characteristics of Mitochondrial Genetics The mitochondrial respiratory chain comprises more than 80 separate peptides. Thirteen of these peptides are encoded in mtDNA and the remainder in nuclear DNA. The nuclearencoded proteins are synthesized on cytosolic ribosomes and imported into the mitochondria, where they co-assemble with mtDNA-encoded counterparts. Mitochondrial DNA is a circular, double-stranded molecule that consists of 16,569 base pairs (Fig. 84–2). It contains 37 genes, comprising 13 protein-encoding genes, 22 transfer RNA genes, and 2 ribosomal RNA genes. All 13 protein-encoding genes are components of the mitochondrial respiratory chain, which is located in the inner membrane of the mitochondria. There are important differences between mitochondrial genetics and nuclear genetics, as outlined below. Maternal Inheritance Human mitochondria and mtDNA are maternally inherited. This is probably because the midpiece of the sperm, which

1939

Diseases Associated with Mitochondrial Respiratory Chain Dysfunction

H+

Intermembrane space

FIGURE 84–1 The mitochondrial respiratory chain. 1 ⫽ complex I (NADH-ubiquinone reductase); 2 ⫽ complex II (succinateubiquinone reductase); 3 ⫽ complex III (ubiquinol–cytochrome c reductase); 4 ⫽ complex IV (cytochrome c oxidase); 5 ⫽ complex V (ATP synthase); Q ⫽ ubiquinone; C ⫽ cytochrome c; ETF ⫽ electron-transferring flavoprotein; TCA ⫽ tricarboxylic acid. See Color Plate

Inner membrane

H+

Q Q Q

1

2

Q Q Q Q ETF

C C C

3

H+

H+

4

5

NAD  O2 H2O

FAD FAD FADH2 FADH2 Matrix

NADH

ADP + Pi e–

H+

TCA cycle

Pyruvate

β-oxidation

ATP

Fatty acids

is the only part containing mitochondria, does not penetrate the ovum during fertilization. There is some evidence for a small paternal contribution, but to date this does not seem to have any role in disease pathogenesis. Because mtDNA is maternally inherited, mtDNA mutations have the potential to be inherited in a matrilineal fashion. In practice, most mtDNA point mutations are maternally inherited. In contrast, large-scale rearrangements of mtDNA, such as large deletions, are not usually maternally inherited. There

is also a small group of recently recognized mtDNA point mutations that are not maternally inherited. The precise explanation why certain mutations are inherited maternally and others are not is unclear. For maternally inherited mutations, a mother harboring the mtDNA mutation will transmit the mutant mtDNA to all her offspring. However, this does not imply that all offspring will develop disease. A number of factors seem to be important in determining the penetrance of a mtDNA mutation in a given individual.

LSP HSP1

OH

HSP2

16

Thr

S

D -l o o p

Cy tb

S

H-strand

Phe

Pro Glu

ND1

5 ND

Leu(UUR)

6 ND

RN A

A 12 RN

Ile Met

OL

C

O

L-strand

Ser(UCN) I

Asp

C O II

III CO 6 e ATPas

Lys ATPase8

ND 4

ND2

Ala Asn Cys Tyr

Trp

FIGURE 84–2 The human mitochondrial DNA. See Color Plate

Leu(CUN) Ser(AGY) His

Gln

ND4L Arg ND3 Gly

1940

Diseases of the Peripheral Nervous System

These include the amount of the mutant mtDNA and its tissue distribution, the mtDNA haplotype, the mtDNA copy number, and the nuclear background.18 Heteroplasmy Individual cells contain many mitochondria, and each mitochondrion contains 2 to 10 copies of mtDNA. Hence, a large number of mtDNA molecules exist in a healthy cell. The DNA sequence in all these copies of mtDNA molecules is identical. This state is known as homoplasmy. In contrast, when there is a variation of the mtDNA sequence between different mtDNA molecules within an individual cell (i.e., there are two populations of mtDNA), this is known as heteroplasmy. Most pathogenic mtDNA mutations in humans are heteroplasmic. Exceptions include the common mtDNA point mutations associated with Leber’s hereditary optic neuropathy (LHON) and nonsyndromic sensorineural deafness, which are almost invariably homoplasmic. It should be noted that the heteroplasmic state is not always pathogenic. For example, heteroplasmic polymorphisms are observed in the mtDNA control region (the D-loop). Threshold Effect There is a large body of data, mainly from cell culture studies, indicating that mtDNA mutations exhibit a threshold effect. That is, a certain proportion of mutant mtDNA is required before there is a reduction in respiratory chain activity. The exact threshold does seem to vary among different mtDNA mutations. For example, most studies using hybrid cell lines indicate that the threshold is greater than 90% mutant for mtRNA point mutations and greater than 60% for large-scale deletions of mtDNA. However, there is evidence that the precise threshold may be influenced by other factors. For example, the nuclear background of a cell may change the threshold, and differences have been observed when comparing transformed and untransformed cell lines. For some mutations, it seems that virtually 100% mutant mtDNA is necessary but not always sufficient to produce a disease phenotype. For example, the common mutations that associate with LHON are generally homoplasmic in all body tissues. However, affected patients generally only develop disease in the optic nerve, and there are a significant number of patients homoplasmic for a LHON mutation who never develop disease at all. Hence, while the proportion of mutant mtDNA is likely to be an important determinant of threshold, other factors must be involved. Importantly, the mechanisms determining threshold for expression are likely to be different for different mtDNA mutations. Segregation In addition to the proportion of mutant mtDNA, the tissue distribution is a determinant of penetrance, expression, and phenotype. Until recently, both mitochondrial

division and mtDNA replication were believed to be stochastic processes unrelated to the cell cycle or to the timing of nuclear replication. Therefore, at cell division, a heteroplasmic cell was considered to transmit different proportions of mutant mtDNA to daughter cells in a random fashion. However, there is recent evidence suggesting that this process may in fact not be entirely random. For example, analysis of different tissues of a mouse harboring a heteroplasmic polymorphism suggested a tissue-specific distribution. The factors that might control such a process are unknown.59

MAKING THE DIAGNOSIS OF MITOCHONDRIAL DISEASE: HISTOPATHOLOGY, BIOCHEMISTRY, AND GENETICS A common histochemical feature in the skeletal muscle biopsies taken from many patients with mitochondrial encephalomyopathies is the so-called ragged red fiber. This represents an abnormal proliferation of mitochondria at the level of the single muscle fiber. The molecular trigger that initiates such mitochondrial proliferation is unknown. This is best seen on transverse section using the modified Gomori trichrome stain, which has affinity for mitochondria (Fig. 84–3). Electron microscopy of muscle often reveals abnormal mitochondrial morphology, including swelling and alterations in cristae and paracrystalline inclusions within the mitochondrial matrix space (Fig. 84–3). The presence of significant numbers of RRF in a muscle biopsy is a strong indicator that there is dysfunction of the respiratory chain, which is likely to be the cause of the patient’s disease. Indeed, most clinicians practicing in this area of medicine would regard the presence of RRF as one of the “gold standard” diagnostic tests for a mitochondrial encephalomyopathy. Indeed, various authors recognized the association between the RRF, mitochondrial respiratory chain dysfunction, and human neurologic diseases long before any genetic defects were discovered.1,2 As long ago as 1962, Luft et al. described the first case of a mitochondrial myopathy in a patient with nonthyroidal hypermetabolism with mild weakness (Luft’s disease).26 The morphologic abnormalities were limited to skeletal muscle, with abnormal accumulations of mitochondria in subsarcolemmal and intermyofibrillar spaces, typical of RRF. Elegant biochemical studies in this patient revealed a disturbance of respiratory chain function. Although Luft’s disease was the first description of a mitochondrial myopathy, only two such cases have been reported to date.9 It has also become apparent that, while the presence of RRF is very good evidence that a patient has a disease caused by respiratory chain dysfunction, the converse is not true. Hence, the absence of RRF does not rule out a

1941

Diseases Associated with Mitochondrial Respiratory Chain Dysfunction

A

B

FIGURE 84–3 A, Skeletal muscle histochemistry of patient with mitochondrial myopathy shows an abundance of ragged red fibers. B, Electron microscopy reveals abnormal elongated mitochondria containing paracrystalline inclusion bodies. (Histochemistry and electron microscopic studies were performed by Prof. D. N. Landon.) See Color Plate

mitochondrial disease. In current specialist practice, the diagnosis of human mitochondrial disease relies on a range of modalities, including clinical assessment, muscle biopsy histochemical and biochemical analysis, and DNA analysis (mainly mtDNA at present). This approach is particularly important in analyzing the patient with isolated neuropathy as described.

COMMON mtDNA DISEASE PHENOTYPES Patients with impaired respiratory chain function may exhibit a wide range of neurologic symptoms and experience significant morbidity and mortality.3,4 There is currently no curative treatment. At present the majority of genetically defined mitochondrial encephalomyopathies are caused by mutations in mtDNA, although nuclear gene defects are now being identified. Common neurologic clinical features include ophthalmoplegia, retinitis pigmentosa, deafness, seizures, disorders of movement, ataxia, strokelike episodes, dementia, myopathy, and peripheral neuropathy.1,2,4,6,15,23,33,35,59 In addition, patients may have non-neurologic presentations, including diabetes and other endocrinopathies, cardiomyopathy, renal dysfunction, liver dysfunction, hematologic dysfunction, and skin disorders. A number of distinctive neurologic syndromes have been shown to commonly associate with particular mtDNA point mutations. For example, the A3243G mutation is commonly associated with mitochondrial encephalomyopathy, lactic acidosis, and strokelike episodes (MELAS).55 This is an often fatal progressive encephalomyopathy usually beginning before the age of 20 years, although later onset is recognized. Strokelike episodes are a particularly characteristic feature. A recent study investigated 32 patients harboring the A3243G

mutation for the presence of peripheral neuropathy using clinical examination and neurophysiology. Seven patients (22%) had clinical and or electrophysiologic evidence of peripheral neuropathy. A mixed axonal-sensorimotor neuropathy was present in six cases, but one exhibited a uniform demyelinating sensorimotor neuropathy.24 The A8344G mutation is associated with myoclonic epilepsy with RRF (MERRF). MERRF is one of the progressive myoclonic ataxias. Patients typically exhibit myoclonus, seizures, and progressive ataxia. Some patients exhibit multiple lipomata, and mild axonal peripheral neuropathy is not uncommon but rarely clinically significant.15,32,43 The T8993G mutation is associated with neurogenic weakness, ataxia, and retinitis pigmentosa (NARP) and with Leigh disease (LD), both of which are described in detail later. Single large-scale deletions commonly associate with chronic progressive external ophthalmoplegia and with the Kearns-Sayre syndrome (KSS). KSS is a serious disorder characterized by chronic progressive external ophthalmoplegia (CPEO) and retinopathy. There is often heart block, which is a common cause of premature death. The age at onset is before 20 years. There may be other associated features such as ataxia, short stature, deafness, and endocrinopathy. The frequency of peripheral neuropathy in patients with CPEO and KSS is not clearly defined, but in our experience it seems to be less common than in MELAS, MERRF, or NARP. One study found that 3 of 10 patients with CPEO had clinical and/or subclinical evidence of an axonal sensorimotor peripheral neuropathy.8,10 There are reports of axonal neuropathy in KSS.57 A mutation at position 11778 in a mitochondrial complex I subunit is associated with LHON, the most common cause of permanent blindness in otherwise healthy young males. Typically there is sequential visual loss, with an intereye delay of 2 to 3 months.18 Peripheral neuropathy is very unusual in patients with isolated LHON.

1942

Diseases of the Peripheral Nervous System

HOW COMMON IS PERIPHERAL NERVE DYSFUNCTION IN MITOCHONDRIAL RESPIRATORY CHAIN DISEASE? Although this seems a relatively simple question, the answer varies somewhat depending on the criteria upon which a diagnosis of mitochondrial disease is based. Before 1991, large series of patients were defined on the basis of clinical features, usually in conjunction with the presence of RRF on muscle histochemistry. After 1991, reviews incorporating genetic data appeared. Large clinical series incorporating clinical features in combination with muscle histopathology as the basis for the diagnosis of mitochondrial disease include those of Eymard et al.,12 Petty et al.,36 and Yiannikas et al.,56 but a number of smaller studies have also been reported.5,7,27,28,29,38,49,51 The combined clinical data in the three large series indicate that 28 of 114 patients (24.5%) had clinical signs of a peripheral neuropathy. In most cases this was a clinically mild mixed sensorimotor axonal neuropathy. In most cases it was not the predominant clinical feature. Eymard and colleagues12 studied 28 patients with mitochondrial disease diagnosed on the basis of clinical phenotype and muscle biopsy. Of these, 25 cases had predominant ophthalmoplegia (including four with KSS) and three had predominant CNS disease (one with LD, one with MELAS, and one with MERRF). Nine of the 28 cases (32%) had clinical signs of a sensory peripheral neuropathy. Peripheral electrophysiology showed changes consistent with a sensory or a mixed sensorimotor axonal neuropathy. Yiannikas et al.56 identified 20 patients based on clinical and muscle histopathology. Five (25%) cases had clinical evidence of a mild sensorimotor peripheral neuropathy confirmed to be axonal by electrophysiologic studies. However, a further five had subclinical but electrophysiologically axonal neuropathy. In a larger clinical/histopathologic series from Petty and colleagues, 14 of 66 patients (21%) had clinical signs of a mild sensory or sensorimotor neuropathy.36 Electrophysiology was undertaken in 51 cases, of which 12 had features of a mild mixed axonal sensorimotor neuropathy. One case had marked slowing of conduction velocity into the demyelinating range (24 m/s in median nerve). A further, more recent electrophysiologic study of 33 mitochondrial disease patients identified changes of a sensorimotor axonal neuropathy in 6 cases (18%).16 A recent comprehensive study by Bouillot et al. has reviewed 100 cases of peripheral neuropathy in mitochondrial disease defined on the basis of pathogenic mtDNA mutations described in the literature.3 They noted that the neuropathy was described in detail in only 43 of these cases. As with earlier reports based on clinical and muscle histopathologic criteria, the study of Bouillot and colleagues, based on the presence of an mtDNA

mutation, confirmed that the most common type of neuropathy in mitochondrial patients is a sensorimotor axonal neuropathy. Of the 43 cases reported by Bouillot et al., peripheral neuropathy was the presenting feature in 12 (30%).3 The majority of the 43 cases had a chronic axonal sensorimotor peripheral neuropathy. In only three cases in this series was the neuropathy subclinical; in the rest it represented a mild or occasionally severe clinical problem. In addition to a chronic sensorimotor peripheral neuropathy, other presentations identified included sensory ataxia neuropathy, dysarthria, and ophthalmoplegia (SANDO) associated with multiple deletions in mtDNA, a motor neuronopathy associated with a cyclooxygenase I (COX I) gene microdeletion, NARP associated with the T8993G mutation, painful neuropathy in association with the A3243G mutation, and acute sensorimotor neuropathy and sensory neuropathy in patients with mitochondrial neurogastrointestinal encephalopathy (MNGIE) associated with multiple mtDNA deletions.3 Each of these phenotypes is described in more detail below. It is interesting to consider how often mitochondrial disease as a whole might present with peripheral neuropathy, before developing additional clinical features. Although not totally reliable, it is possible to derive an estimate of this by combining the data from the clinical studies outlined above with those from Bouillot et al.’s genetically defined study. Combining these data would indicate that, of the 24.5% of mitochondrial patients with a peripheral neuropathy, 30% of these might be expected to present with symptoms from their neuropathy, which would suggest a figure of 7% overall. Clearly, such a derived calculation must be regarded with caution. However, clinical experience in our mitochondrial disease clinic would indicate that it is certainly unusual, but recognized, for a patient with a mitochondrial disease to present with neuropathy, and we would certainly estimate this to occur in less than 10% of cases. It should be noted that this is an estimate of the proportion of patients who have a peripheral neuropathy as a presenting feature of a mitochondrial disease that will subsequently manifest other clinical features. It remains unknown what proportion of isolated chronic axonal sensorimotor neuropathy is accounted for by mitochondrial disease. Our impression is that mitochondrial disease that remains confined to peripheral nerve without other clinical manifestations is rare. However, certain proof of this impression is confounded by the absence of an unequivocal neuropathologic signature of mitochondrial peripheral nerve disease. That is, unlike muscle histopathology, there seems to be no unequivocal histopathologic hallmark in nerve tissue that indicates mitochondrial disease as the cause. This issue is described below in more detail.

Diseases Associated with Mitochondrial Respiratory Chain Dysfunction

NEUROPATHOLOGY OF PERIPHERAL NERVE IN PATIENTS WITH MITOCHONDRIAL DISEASE Because mitochondria are present in significant numbers in the axon and neuronal cytoplasm, as well as in the Schwann cell cytoplasm of myelinated nerves, it is perhaps not unexpected that both axonal and demyelinating patterns of histopathology have been described.47 However, compared with skeletal muscle, relatively less is known about the histopathologic features of peripheral nerve damage caused by mitochondrial disease. Furthermore, unlike the characteristic and specific histopathologic changes seen in skeletal muscle, there is, as yet, no unequivocally specific neuropathologic change in nerve tissue that can be regarded as diagnostic of mitochondrial nerve disease. In reviewing the existing neuropathology literature in this area, it is again important to be clear about the diagnostic criteria used for mitochondrial disease. We have chosen to focus upon more recent neuropathologic studies in which all patients were defined by the presence of a pathogenic mtDNA mutation. We consider that a genetic diagnosis is most objective from the point of view of analyzing and comparing clinical or pathologic features in groups of patients. However, there are important neuropathologic studies published prior to the availability of genetic diagnosis, such as the study by Schroder et al.44 They reported the results of an electron microscopic study that included 15 cases of mitochondrial myopathy that were defined on clinical criteria. Altered mitochondria were seen in the peripheral nerves in 13 of the 15 cases. Most of the mitochondrial abnormalities were observed in Schwann cells, although in three cases with the MELAS phenotype, an increase in the numbers of mitochondria was seen in the endothelial cells of the epineurial arterioles. Indeed, a subsequent study by Schroder et al. confirmed an increase of the mean number and an enlargement of the cross-sectional area of mitochondria in endothelial and smooth muscle cells of the endoneurial and epineurial arterioles and pericytes of capillaries in the sural nerves of 20 cases of mitochondrial diseases.30 The abnormalities of the mitochondria observed in Schwann cells were varied but were considered to be distinct from the changes typically seen in the mitochondria in skeletal muscle biopsies of such patients. Mitochondrial abnormalities observed in nerve included enlargement with an amorphous matrix and distorted cristae, hexagonal paracrystalline inclusions, and prominent cristae containing oblique striations.44 However, it is notable that similar mitochondrial morphologic changes were also observed in three cases of hereditary motor and sensory neuropathy with optic atrophy (HMSN type VI) and in 25 of 280 unselected patients with peripheral

1943

neuropathy. The specificity of the changes observed was therefore considered to be uncertain. Bouillot and colleagues3 reported neuropathologic findings in eight patients with proven mitochondrial DNA mutations. Two patients harbored the A3243G mutation, two harbored the A8344G mutation, one harbored the A11778G mutation, one had a single large deletion, and two harbored multiple deletions. In none of these patients was the neuropathy the sole clinical feature. A severe sensory peripheral neuropathy was the presenting clinical feature in one of the cases with multiple deletions. All cases had other non-neuropathic clinical features commonly associated with these mutations. Furthermore, all except the A11778G case had light or electron microscopic features of mitochondrial disease on muscle biopsy. Superficial peroneal nerve biopsies were undertaken and showed some form of abnormality in all cases. Axonal loss was present in all cases. Myelinated fiber loss was observed in association with the A3243G, A8344G, single deletion, and multiple deletion cases. There was evidence of a primary demyelinating process, with thinly myelinated axons and onion bulb formation, in four of the cases (one case with A3243G, one case with A8344G, and two cases with multiple deletions). Abnormal mitochondria that were enlarged and had abnormal cristae were seen in the cytoplasm of Schwann cells in patients with A8344G and with multiple deletions. These mitochondria were swollen and had crystalline inclusions. Similar mitochondrial abnormalities were also seen in the cytoplasm of unmyelinated fibers in a case with the A8344G mutation. In addition to axonal loss and onion bulb changes, other perhaps more specific features have been reported, although they remain a little controversial. These include abnormal mitochondria in the Schwann cell cytoplasm with distorted cristae. Unmyelinated fibers have been suggested to show two different abnormalities: mitochondria with abnormal cristae and mitochondria with crystalline-like inclusions. It turns out that the crystalline-like inclusions may in fact not be entirely specific for mitochondrial disease, having been reported in chronic inflammatory demyelinating polyneuropathy, Refsum’s disease, diabetes, and sarcoid. It has recently been proposed that the crystalline-like bodies may be an end product of a chronic axonal process.44 In summary, when carefully defined on the basis of a specific mtDNA mutation, it would seem that the presence of abnormalities of mitochondrial cristae in axons or Schwann cell cytoplasm is indicative of a mitochondrial cause for a neuropathy. However, in a patient in whom a neuropathy is the main or sole clinical feature, it remains difficult to be certain that the cause is mitochondrial on nerve biopsy histopathology alone. Muscle biopsy, muscle biochemistry, and genetics remain important and often essential diagnostic aids.

1944

Diseases of the Peripheral Nervous System

MITOCHONDRIAL DISEASES WITH NEUROPATHY AS A MAJOR CLINICAL FEATURE Although peripheral neuropathy is a common accompaniment of mitochondrial disease in up to 25% of mitochondrial disease cases, in most of these it is a relatively minor part of a more complex neurologic phenotype. However, there are certain examples in which the neuropathy is a major and sometimes overwhelming clinical feature, and these are now considered in some detail.

Demyelinating Neuropathy Although axonal neuropathy is more common, there are well-characterized examples of mitochondrial patients exhibiting a significant, predominantly demyelinating neuropathy. Teener and Sladky reported four patients from three families with a demyelinating neuropathy as a prominent feature of their mitochondrial diseases.50 In two patients the clinical signs and symptoms, the electrophysiologic findings, and the nerve histopathology were remarkably similar to that seen in acquired demyelinating neuropathy. Indeed, the patients were initially treated with immunosuppression. In two patients the nerve conduction velocities were slowed to the point of meeting diagnostic criteria for demyelinating neuropathy. Nerve biopsy confirmed features of a demyelinating process with thinly myelinated axons and onion bulb formation. No genetic data were included. Demyelinating neuropathy has been described in association with defined mtDNA mutations. Rusanen et al. described a 38-year-old man with cognitive and hearing impairment and distal limb weakness but who had preserved tendon reflexes.40 He harbored the A3243G mutations and was from a family with this mutation. A wide variability in the clinical phenotype was observed in the family. The proportion of the A3243G mutation was highest in his brother, who had the full-blown MELAS phenotype. Nerve conduction studies showed uniform reduction of conduction velocity over the length of motor and sensory nerves, unlike acquired demyelinating neuropathy. There was no conduction block or temporal dispersion. It was concluded that the electrophysiology was similar to that seen in the hereditary demyelinating neuropathies. A nerve biopsy was not undertaken.40 Fang14 described a Chinese patient with features of a demyelinating neuropathy in association with the A3243G mutation. The patient had typical clinical features of the MELAS phenotype, including short stature, headache, episodic vomiting, mental retardation, hearing impairment, seizures, hemiparesis, lactic acidosis, and characteristic brain computed tomography findings. However, in addition he exhibited generalized wasting and weakness, areflexia,

and pes cavus. He also developed acute painful peripheral paresthesia. Electrophysiology was consistent with a mixed axonal and demyelinating sensorimotor neuropathy. In addition to marked slowing of motor conduction velocities, there was temporal dispersion resembling that seen in acquired demyelinating neuropathy. A nerve biopsy was not undertaken.15 King et al. described a 17-year-old boy with a rapidly progressive and fatal neuromyopathy without clinical or radiologic evidence of CNS involvement. The cerebrospinal fluid (CSF) protein was elevated, as was the serum lactate. A sural nerve biopsy revealed marked thinning of myelin sheaths with significant onion bulb formation indicative of a demyelinating process.25 Demyelinating Neuropathy in Leigh Disease LD is one of the most common mitochondrial encephalomyopathies in infancy.21 It is a devastating disorder associated with a range of neurologic signs including psychomotor retardation, optic atrophy, ophthalmoplegia, ptosis, nystagmus, ataxia, tremor, dystonia, pyramidal signs, and abnormal breathing. Definite diagnosis depends on either characteristic neuroimaging or the postmortem neuropathologic features of bilateral symmetrical spongiform lesions localized especially in the thalamus, basal ganglia, and brainstem. These findings are microscopically associated with cystic cavitation, demyelination, vascular proliferation, and gliosis.21 The onset of LD is often in the first year of life; however, cases of juvenile- or adult-onset LD are occasionally described.19 A number of reports have indicated that impaired peripheral nerve myelination is present in cases of LD. Seitz et al. reported a pathologic study of a girl who died at the age of 8 weeks with LD.45 Neuropathologic examination of the sciatic nerve revealed marked demyelination that was termed a primary demyelinating neuropathy. There was a reduction in the number of large myelinated fibers. There were many demyelinated axonal segments and phagocytic cells. Goebel et al. identified peripheral neuropathy in four unrelated children between the ages of 15 months and 9 years whose autopsies revealed changes of LD, confirming the clinical suspicion.17 Three of the cases had marked reduction in conduction velocities into the demyelinating range, and the fourth case had low-normal conduction velocity. Sural nerve biopsy was reported to show loss of myelinated fibers without evidence of preceding demyelination in one of the cases. In the other three cases there was evidence of demyelination and remyelination on sural nerve biopsies. Specifically, in these three cases there were large axons without myelin sheaths or with only thin myelin sheaths, and incipient onion bulb formation was present. In addition, in two of the cases there was also loss of unmyelinated axons.

Diseases Associated with Mitochondrial Respiratory Chain Dysfunction

Jacobs et al. described peripheral neuropathy in three children from two families with LD.21 Sural nerve biopsies were reported to show myelin fiber densities in the normal range with no evidence of active degeneration or regeneration. The myelin sheaths were noted to be thin. In teased fibers, there was no evidence of demyelination in two cases but some in one case. It was concluded that the findings suggested hypomyelination of peripheral nerves eventually leading to some demyelination. LD is now known to be a genetically heterogeneous mitochondrial respiratory chain disease.39,59 Although the first genetic defect identified was in the ATPase 6 gene of the mtDNA (the T8993G mutation), LD is more commonly inherited by an autosomal recessive trait than as a maternally inherited trait.39,41,59 Interestingly, autosomal recessive LS families frequently have either isolated complex IV (cytochrome c oxidase), or complex I (NADHubiquinone reductase) deficiency or, less commonly, complex II (succinate-ubiquinone reductase) deficiency. Recently, nuclear genes encoding important mitochondrial proteins have been identified in some families with autosomal LD.59 Nuclear genes that are important in the assembly of complex IV of the respiratory chain have been shown to associate particularly with LD in combination with peripheral neuropathy. Two examples of such genes that code for proteins important in the assembly of complex IV include SURF1 and SCO2.22,42 Mutations in the SURF1 gene account for 75% of LD cases with COX deficiency.59 In addition to the typical CNS features of LD, patients harboring SURF1 gene mutations may also have peripheral neuropathy characterized by impaired myelination. Santoro et al.42 described a 5-year-old boy with clinical and imaging features of LD and a peripheral neuropathy. Skeletal muscle biopsy confirmed a decrease in complex IV activity as the basis of the LD, and SURF1 gene analysis

1945

identified a novel homozygous nonsense mutation predicting a truncation of the SURF-1 protein. Sural nerve biopsy revealed reduced myelinated fibers and complete absence of fibers of large diameter. Myelin sheaths were noted to be thin in relation to diameter in all fibers examined. The authors did not observe wallerian degeneration, myelinophagic cells, regeneration clusters, onion bulb formation, or basal membrane replication. Ultrastructural examination revealed axons that were very thinly myelinated. Increased numbers of mitochondria were observed in Schwann cells. The mitochondria were often enlarged with rounded cristae. Intra-axonal mitochondria were considered to be normal.42 We recently studied a 30-year-old woman with isolated complex IV deficiency and a mixed sensorimotor peripheral neuropathy. Clinically the neuropathy was mild, but there was dramatic slowing of motor conduction velocity, as low as 12 m/s. Nerve biopsy showed loss of large myelinated fibers and many thinly myelinated fibers (Fig. 84–4). Observations from this LD case and those outlined above suggest that complex IV activity of the respiratory chain is particularly important in peripheral nerve myelination. Demyelinating Neuropathy in MNGIE MNGIE is a unique autosomal recessive multisystem disorder associated with multiple deletions or partial depletion of mtDNA. It is characterized by gastrointestinal dysmotility, cachexia, ptosis, progressive external ophthalmoplegia, polyneuropathy, and leukoencephalopathy on brain magnetic resonance imaging (MRI).34,46,48 Onset may vary from 5 months to 43 years of age, although the majority of patients have experienced symptoms before 20 years of age. Patients often die by the fourth decade. Almost half of the patients develop gastrointestinal symptoms as the initial manifestation. Ophthalmoplegia is also a common initial presentation. Gastrointestinal features are the most

FIGURE 84–4 Electron microscopy of sural nerve biopsy from a control and a patient with complex IV deficiency. The patient nerve biopsy shows loss of large fibers and thinly myelinated fibers.

1946

Diseases of the Peripheral Nervous System

prominent clinical features and include borborygmi, abdominal pain and cramps, diarrhea, early satiety, nausea/ vomiting, diverticulosis, pseudo-obstruction, gastroparesis, and dysphagia. Cachexia is another prominent symptom, but despite the apparent muscle atrophy, muscle strength is well maintained. Neurologic symptoms are often mild. Peripheral neuropathy affecting both motor and sensory nerves, leading to glove-stocking sensory loss and mild distal limb weakness, is frequent. Patients occasionally have debilitating weakness. Hearing loss is observed in almost half of the patients. Pigmentary retinopathy and mental retardation are rare. Nerve conduction studies are invariably consistent with demyelinating neuropathy, but sometimes also show additional features compatible with axonal neuropathy. Electromyography sometimes shows a myogenic pattern in addition to more common neuropathic features. Increased CSF protein is observed in most patients. Brain MRI generally reveals diffuse leukoencephalopathy. It is common that there may be no clinical manifestations attributable to the leukoencephalopathy, such as pyramidal signs or dementia. COX-deficient muscle fibers are invariably observed, and almost two thirds of patients have RRF on muscle biopsy. Nishino et al.34 described mutations in the thymidine phosphorylase (TP) gene on chromosome 22q13.32-qter in 12 MNGIE probands. TP activity in leukocytes in MNGIE patients was found to be less than 5% of normal controls, indicating that loss-of-function mutations of the TP gene cause the disease. MNGIE caused by mutations in the TP gene is a clinically homogeneous disease. All patients developed (1) gastrointestinal dysmotility, (2) cachexia, (3) ptosis and ophthalmoplegia, (4) peripheral neuropathy, (5) leukoencephalopathy on brain MRI, and (6) laboratory evidence of mitochondrial dysfunction, including COX-deficient muscle fibers and RRF or multiple deletions of mtDNA.34

Axonal Neuropathy Acute and Subacute Axonal Neuropathy with A3234G Mutation The A3243G mtDNA mutation most frequently causes the MELAS phenotype. A mild chronic axonal sensorimotor neuropathy may be a minor feature, as described above.24 However, occasionally the A3243G mutation is associated with more aggressive neuropathic disease. Hara et al.19 described a 33-year-old woman who developed an acute sensorimotor neuropathy with significant limb weakness and facial weakness in her second pregnancy. She was considered to have Guillain-Barré syndrome but went on to develop severe lactic acidosis and rhabdomyolysis. Genetic analysis showed that she harbored the mtDNA A3243G mutation. Electrophysiologic studies indicated a severe acute axonal neuropathy. Sural nerve biopsy showed a severe decrease in

myelinated fibers. Teased nerve preparations showed myelin ovoid formation but no demyelination. Ciafaloni et al.6 described a 21-year-old man who developed subacute distal limb weakness of moderate severity. Electrophysiologic studies showed an axonal polyneuropathy that was confirmed on sural nerve biopsy. A year after the presentation with neuropathy, he developed a strokelike episode and generalized seizures and a diagnosis of MELAS was made. Genetic analysis confirmed the A3243G mutation. Van Domburg et al.53 reported six adults from three separate families with an unusual heredoataxic syndrome. In all cases the early phase of the disease was characterized by a severe sensory neuropathy. Sural nerve biopsies showed severe loss of myelinated axons at an early stage in the disease. The patients subsequently develop CPEO and later develop myoclonic epilepsy with a high incidence of status epilepsy and sudden death. Muscle biopsy showed RRF and CSF lactate was elevated, confirming a mitochondrial disease, although the genetic basis was not determined. Sensory Ataxic Neuropathy, Dysarthria, and Ophthalmoparesis Four sporadic patients from North America developed a distinct clinical syndrome presenting with severe sensory ataxic neuropathy, dysarthria, and late-onset progressive external ophthalmoplegia.13 Fadic et al. proposed the acronym SANDO and identified multiple deletions in the mtDNA in all patients.13 The most prominent features are sensory ataxic neuropathy and ophthalmoplegia often developing after the age of 30 years. Migraine and depression were also common in these cases. Nerve conduction studies showed a predominantly sensory axonal neuropathy. Muscle biopsies generally revealed RRF with or without COX-deficient fibers.13 Recently a mutation in the POLG gene has been shown to be the cause of SANDO in a sporadic patient. The POLG gene encodes the DNA polymerase gamma enzyme, which is crucial for mtDNA replication. The patient described was a compound heterozygote harboring a previously described recessive mutation (A467T) in combination with a novel R627W change. Interestingly, the A467T mutation had previously been described in autosomal recessive CPEO, suggesting that this disorder and SANDO are allelic.13,54 Neuropathy, Ataxia, and Retinitis Pigmentosa The first case of peripheral neuropathy to be described in association with a pathogenic mtDNA mutation was in a family with the NARP phenotype. NARP was originally described by Holt et al.20 in a family with maternal inheritance of a variable combination of retinitis pigmentosa, ataxia, neurogenic muscle weakness, developmental delay, seizures, dementia, optic atrophy, and sensory neuropathy. Holt et al. identified a heteroplasmic T8993G mutation in

Diseases Associated with Mitochondrial Respiratory Chain Dysfunction

the mitochondrial ATP synthase subunit 6 (ATPase 6) gene in all studied family members. The more severe clinical phenotypes and the cases with earlier age at onset correlated with higher proportions of the mutant mtDNA in blood. Individuals with levels of mutant mtDNA less than 70% are usually asymptomatic or mildly affected. NARP often associates with proportions of mutant mtDNA in muscle or blood ranging between 70% and 90%. Patients harboring proportions of mutant mtDNA above 90% generally develop one of the most severe and frequently lethal phenotypes, LD. Gait and limb ataxia of the cerebellar type is prominent, and there is usually cerebellar atrophy on brain scan. Limb weakness is often mild and may be obscured by limb ataxia. Axonal sensory neuropathy may involve any sensory modality, and tendon reflexes are absent. Muscle biopsies do not show the typical histochemical evidence of mitochondrial myopathy, such as RRF or absent of COX activity. Instead, the muscle biopsy simply shows changes of denervation of muscle as a consequence of the neuropathy.42,52 The combination of the described clinical features in association with maternal transmission is therefore an important clue for the diagnosis, which can be confirmed by mtDNA analysis. Autonomic Neuropathy Autonomic neuropathy appears to be uncommon in mitochondrial disease, although some cases are reported. Edwards-Lee et al.11 described a 2-year-old boy with congenital insensitivity to pain with anhidrosis. Biochemical analysis determined that cytochrome c oxidase activity was reduced (35% of control values), suggesting a possible mitochondrial basis for the neuropathy, although a genetic cause was not determined. Sural nerve biopsy showed almost absent small unmyelinated axons. Abnormal axonal structure with decreased microtubules and some abnormally enlarged mitochondria were noted. Zelnik et al.58 reported three children presenting with gastrointestinal dysmotility, anemic episodes, cardiac arrhythmias, decreased lacrimation, supersensitivity to methacholine, altered sweating, postural hypotension, and diminished tendon reflexes. Biochemical analysis indicated a partial complex IV defect in two cases and a complex III defect in one case, suggesting a mitochondrial cause.

CONCLUSION Mitochondrial respiratory chain disease is clinically and genetically heterogeneous. Mutations in nuclear or mitochondrial genes may be the cause. Peripheral neuropathy is not uncommon in mitochondrial disease, and to date most cases are associated with mtDNA mutations. Although entry criteria vary, most series indicate a frequency of between 25% and 40%.

1947

Axonal neuropathy is the most common form, and this is often a minor part of a more complex disease phenotype. From a practical viewpoint, the discovery of a mild axonal peripheral neuropathy in a patient with a complex CNS disease may be a useful clue to the possible mitochondrial etiology of the condition, thereby prompting further investigation. Less commonly, demyelinating neuropathy has been described in mitochondrial disease. Electrophysiologic features similar to either inherited or acquired forms of demyelination have been described in different patients. Recently, the nuclear gene SURF1, which is important in the assembly of complex IV of the respiratory chain, has been shown to associate with peripheral nerve hypomyelination. There are certain mitochondrial phenotypes in which neuropathy may be the presenting or a major feature. These include the MNGIE, NARP, and SANDO phenotypes. It is unclear why the mtDNA mutations that cause these diseases should particularly affect peripheral nerve. Mitochondrial dysfunction seems to be a relatively uncommon cause of isolated chronic axonal peripheral neuropathy, although certainty on this point is confounded by the absence of an entirely reliable neuropathologic mitochondrial signature. Although there are certain pathologic features on nerve biopsy that will lead to a suspicion of mitochondrial disease, other investigative data will usually be required to be certain that mitochondrial respiratory chain dysfunction is the cause.

REFERENCES 1. Bastiansen, L. A. K., Joosten, E. M. G., De Rooij, J. A. M., et al.: Ophthalmoplegia-plus, a real nosological entity. Acta Neurol. Scand. 58:9, 1978. 2. Berenberg, R. A., Pellock, J. M., Dimauro, S., et al.: A. Lumping or splitting? Ophthalmoplegia plus or Kearns Sayre syndrome? Ann. Neurol. 1:37, 1977. 3. Bouillot, S., Martin-Negrier, A., Vital, A., et al.: Peripheral neuropathy associated with mitochondrial disorders: 8 cases and a review of the literature. J. Periph. Nerv. Syst. 7:213, 2002. 4. Chinnery, P. F., Johnson, M. A., Wardell, T. M., et al.: The epidemiology of pathogenic mitochondrial DNA mutations. Ann. Neurol. 48:188, 2000. 5. Chu, C. C., Huang, C. C., Fang, W., et al.: Peripheral neuropathy in mitochondrial encephalomyopathies. Eur. Neurol. 37:110, 1997. 6. Ciafaloni, E., Ricci, E., Servidei, S., et al.: Widespread tissue distribution of a tRNALeu (UUR) mutation in the mitochondrial DNA of a patient with MELAS syndrome. Neurology 41:1663, 1991. 7. Colomer, J., Iturriaga, C., Bestue, M., et al.: Aspects of neuropathy in mitochondrial diseases. Rev. Neurol. 30:1117, 2000. 8. Croft, P. B., Cutting, J. C., Jewesbury, E. C. O., et al.: Ocular myopathy (progressive external ophthalmoplegia) with neuropathic complications. Acta Neurol. Scand. 55:169, 1977.

1948

Diseases of the Peripheral Nervous System

9. DiMauro, S., Bonilla, E., Lee, C. P., et al.: Luft’s disease: further biochemical and ultrastructural studies of skeletal muscle in the second case. J. Neurol. Sci. 27:217, 1976. 10. Drachman, D. A.: Ophthalmoplegia plus: the neurodegenerative disorders associated with progressive external ophthalmoplegia. Arch. Neurol. 18:654, 1968. 11. Edwards Lee, T. A., Cornford, M. E., and Yu, K. T. T.: Congenital insensitivity to pain and anhidrosis with mitochondrial and axonal abnormalities. Pediatr. Neurol. 17:356, 1997. 12. Eymard, B., Penicaud, A., Leger, J. M., et al.: Etude clinique et electrophysiologique du nerf peripherique dans 28 cas du maladie mitochondriale. Rev. Neurol. (Paris) 147:508, 1991. 13. Fadic, R., Russel, J. A., Vedanarayanan, V. V., et al.: Sensory ataxic neuropathy as the presenting feature of a novel mitochondrial disease. Neurology 49:239, 1997. 14. Fang, W.: Polyneuropathy in the mtDNA base pair 3243 point mutation. Neurology 46:1494, 1996. 15. Fang, W., Huang, C. C., Chu, N. S., et al.: Myoclonic epilepsy with ragged red fibers (MERRF) syndrome: report of a Chinese family with mitochondrial DNA point mutation in tRNA lys gene. Muscle Nerve 17:52, 1994. 16. Girlanda, P., Toscano, A., Nicolosi, C., et al.: Electrophysiological study of neuromuscular system involvement in mitochondrial cytopathy. Clin. Neurophysiol. 110:1284, 1999. 17. Goebel, H. H., Bardosi, A., Friede, R. L., et al.: Sural nerve biopsy studies in Leigh’s subacute necrotizing encephalomyelopathy. Muscle Nerve 9:165, 1986. 18. Hanna, M. G., and Morgan-Hughes, J. A.: Mitochondrial encephalomyopathies: the enigma of genotype versus phenotype. Biochem. Biophys. Acta 1410:125, 1999. 19. Hara, H., Wakayama, Y., Kouno, Y., et al.: Acute peripheral neuropathy, rhabdomyolysis and severe lactic acidosis associated with 3243 A to G mitochondrial DNA mutation. J. Neurol. Neurosurg. Psychiatry 57:1545, 1994. 20. Holt, I. J., Harding, A. E., Petty, R. K., and Morgan-Hughes, J. A.: A new mitochondrial disease associated with mitochondrial DNA heteroplasmy. Am. J. Hum. Genet. 46:428, 1990. 21. Jacobs, J. M., Harding, B. N., Lake, B. D., et al.: Peripheral neuropathy in Leigh’s disease. Brain 113:447,1990. 22. Jaksch, M., Horvarth, R., Horn, N., et al.: Homozygosity (E 140 K) in SCO2 causes delayed infantile onset of cardiomyopathy and neuropathy. Neurology 57:1440, 2001. 23. Jinnai, K., Yamada, H., Kanda, F., et al.: A case of mitochondrial myopathy, encephalopathy and lactic acidosis due to cytochrome c oxidase deficiency with neurogenic muscular changes. Eur. Neurol. 30:56, 1990. 24. Karppa, M., Syrjala, P., Tolonen, U., and Majamaa, K.: Peripheral neuropathy in patients with the 3243A⬎G mutation in mitochondrial DNA. J. Neurol. 250:216, 2003. 25. King, M. D., O’Neill, G., Poulton, J., et al.: Polyneuropathy in the mtDNA base pair 3243 point mutation. Neurology 46:1494, 1996. 26. Luft, R., Ikkos, D., Palmieri, G., et al.: A case of severe hypermetabolism of nonthyroid origin with a defect in the maintenance of mitochondrial respiratory control—a correlated clinical, biochemical, and morphological study. J. Clin. Invest. 41:1776, 1962. 27. Majamaa, K., Tolonen, U., Rusanen, H., et al.: Polyneuropathy in the mtDNA base pair 3243 point mutation. Neurology 46:1494, 1996.

28. Mizusawa, H., Ohkoshi, N., Watanabe, M., and Kanazawa, I.: Peripheral neuropathy of mitochondrial myopathies. Rev. Neurol. (Paris) 147:501, 1991. 29. Mizusawa, H., Watanabe, M., Kanazawa, I., et al.: Familial mitochondrial myopathy associated with peripheral neuropathy: partial deficiencies of complex I and complex IV. J. Neurol. Sci. 86:171, 1988. 30. Molnar, M., Neudecker, S., and Schroder, J. M.: Increase of mitochondria in vasa nervorum of cases with mitochondrial myopathy, Kearns Sayre syndrome, progressive external ophthalmoplegia and MELAS. Neuropathol. Appl. Neurobiol. 21:432, 1995. 31. Nardin, R. A., and Johns, D. R.: Mitochondrial dysfunction and neuromuscular disease. Muscle Nerve 24:170, 2001. 32. Naumann, M., Keifer, R., Toyka, K. V., et al.: Mitochondrial dysfunction with myoclonus epilepsy and ragged red fibers point mutation in nerve, muscle and adipose tissue of a patient with multiple symmetric lipomatosis. Muscle Nerve 20:833, 1997. 33. Nelson, I., Hanna, M. G., Alsanjari, N., et al.: A new mitochondrial DNA mutation associated with progressive dementia and chorea: a clinical, pathological and molecular genetic study. Ann. Neurol. 37:400, 1995. 34. Nishino, I., Spinazzola, A., and Hirano, M.: MNGIE: from nuclear DNA to mitochondrial DNA. Neuromuscul. Disord. 11:7, 2001. 35. Okamoto, T., Mizuno, K., Iida, M., et al.: Ophthalmoplegia plus: its occurrence with periventricular diffuse low density on computed tomography scan. Arch. Neurol. 38:423, 1981. 36. Petty, R. K. H., Harding, A. E., and Morgan-Hughes, J. A.: The clinical features of mitochondrial myopathy. Brain 109:915, 1986. 37. Peyronnard, J. M., Charron, L., Bellavance, A., and Marchand, L.: Neuropathy and mitochondrial myopathy. Ann. Neurol. 7:262, 1980. 38. Pezeshkpour, G., Krarup, C., Di Mauro, S., et al.: Peripheral neuropathy in mitochondrial disease. J. Neurol. Sci. 77:285, 1987. 39. Pulkes, T., and Hanna, M. G.: Human mitochondrial disease. Adv. Drug Deliv. Rev. 49:27, 2001. 40. Rusanen, H., Majamaa, K., Tolonen, A., et al.: Demyelinating polyneuropathy in a patient with the tRNAleu(UUR) mutation at base pair 3243 of the mitochondrial DNA. Neurology 45:1188, 1995. 41. Santorelli, F. M., Kurenai, T., Shanske, S., and DiMauro, S.: Heterogeneous clinical presentation of the mtDNA NARP/T8993G mutation. Neurology 49:270, 1997. 42. Santoro, L., Carrozzo, R., Malandrini, A., et al.: A novel SURF1 mutation results in Leigh syndrome with peripheral neuropathy cause by cytochrome c oxidase deficiency. Neuromuscul. Disord. 10:450, 2000. 43. Sasaki, H., Kuzuhara, S., Kanazawa, I., et al.: Myoclonus, cerebellar disorder, neuropathy, mitochondrial myopathy and ACTH deficiency. Neurology 33:1288, 1983. 44. Schroder, J. M., and Sommer, C.: Mitochondrial abnormalities in human sural nerves: fine structural evaluation of cases with mitochondrial myopathy, hereditary and non-hereditary neuropathies and review of the literature. Acta Neuropathol. 82:471, 1991.

Diseases Associated with Mitochondrial Respiratory Chain Dysfunction 45. Seitz, R. J., Langes, K., Frenzel, H., et al.: Congenital Leigh’s disease: panencephalomyelopathy and peripheral neuropathy. Acta Neuropathol. (Berl.) 64:167, 1984. 46. Simon, L. T., Horoupian, D. S., Dorfman, L. J., et al.: Polyneuropathy, ophthalmoplegia, leukoencephalopathy and intestinal pseudo-obstruction: POLIP syndrome. Ann. Neurol. 28:349, 1990. 47. Sladky, J. T.: Histopathological features of peripheral nerve and muscle in mitochondrial disease. Semin. Neurol. 21:293, 2001. 48. Takeda, K., Kawai, M., Sakuta, M., and Takemura, T.: Paralytic ileus and atonic bladder in a case of mitochondrial encephalomyopathy—electrophysiological, chemical and pathological study with evidence of the peripheral nerve involvement. Rinsho Shinkeigaku 29:643, 1989. 49. Tanaka, J., Taniike, M., Inui, K., et al.: Peripheral neuropathy in two children with mitochondrial encephalomyopathy. Rinsho Shinkeigaku 31:772, 1991. 50. Teener, J. W., and Sladky, J. T.: Demyelinating neuropathy as a feature of mitochondrial encephalomyopathy. Ann. Neurol. 36:508, 1994. 51. Torbergsen, T., and Stalberg, E.: Nerve-muscle involvement in a large family with mitochondrial cytopathy: electrophysiological studies. Muscle Nerve 14:35, 1991. 52. Uziel, G., Moroni, E., Lamantea, E., et al.: Mitochondrial disease associated with the T8993G mutation of the mitochondrial ATPase 6 gene: a clinical, biochemical and molecular study in six families. J. Neurol. Neurosurg. Psychiatry 63:16, 1997.

1949

53. Van Domburg, P. H. M. F., Gabreels-Festen, A. A. W. M., Gabreels, F. J. M., et al.: Mitochondrial cytopathy presenting as hereditary sensory neuropathy with progressive external ophthalmoplegia, ataxia and fatal myoclonic epileptic status. Brain 19:997, 1996. 54. Van Goethem, G., Marin, J. J., Dermaut, B., et al.: Recessive POLG mutations presenting with sensory and ataxic neuropathy in compound heterozygote patients with progressive external ophthalmoplegia. Neuromuscul. Disord. 13:133, 2003. 55. Yamamoto, M., Sato, T., Anno, M., et al.: Mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes with recurrent abdominal symptoms and coenzyme Q10 administration. J. Neurol. Neurosurg. Psychiatry 50:1475, 1987. 56. Yiannikas, C., McLeod, J. C., Pollard, J. D., and Baverstock, J.: Peripheral neuropathy associated with mitochondrial myopathy. Ann. Neurol. 20:249, 1986. 57. Zanssen, S., Molnar, M., Buse, G., and Schroder, J. M.: Mitochondrial cytochrome b gene deletion in Kearns-Sayre syndrome associated with a subclinical type of peripheral neuropathy. Clin. Neuropathol. 17:291, 1998. 58. Zelnik, N., Axelrod, F. B., Leshinsky, E., et al.: Mitochondrial encephalomyopathies presenting with features of autonomic and visceral dysfunction. Pediatr. Neurol. 14:251, 1996. 59. Zeviani, M., and Klopstock, T.: Mitochondrial disease. Curr. Opin. Neurol. 14:553, 2001.

K. Neuropathy Associated with Systemic Disease

85 Diabetic Neuropathies J. G. LLEWELYN, D. R. TOMLINSON, AND P. K. THOMAS

Historical Aspects Epidemiology, Clinical Course, and Risk Factors Minimum Criteria for Diagnosis Natural History Risk Factors Clinical Features Neuropathy Related to Impaired Glucose Tolerance and Hyperglycemia Distal Symmetrical Sensorimotor Polyneuropathy Acute Painful Neuropathy Diabetic Motor Neuropathy Autonomic Neuropathy

Focal and Multifocal Neuropathies Hypoglycemic Neuropathy Pathologic Changes Histopathologic Changes Autonomic Neuropathy Relationship of Pathologic Changes to Clinical Syndromes Pathogenesis and Clinical Trials Hyperglycemia Polyol Pathway and Aldose Reductase Inhibitors Advanced Glycation End Products Essential Fatty Acids Neurotrophic Factors

HISTORICAL ASPECTS In 1864, Marchal de Calvi197 recognized that diabetes affected the nervous system, and from then, the salient clinical features of diabetic neuropathy were delineated. The loss of tendon reflexes in the legs was described by Bouchard29 and the occurrence of spontaneous pains and hyperesthesia was stressed by Pavy.238 Motor manifestations in the legs were documented in 1890 by Buzzard,45 Bruns,42 and Charcot,56 and cranial nerve involvement in diabetic patients had been noted earlier by Ogle.229 With recognition of the highly variable nature of diabetic neuropathy, the first classifications were suggested by Leyden179 and Pryce252

Nerve Ischemia/Hypoxia Oxidative Stress Treatment of Neuropathic Complications Treatment of Painful Diabetic Neuropathy The “At-Risk” Diabetic Foot Animal Models of Causation Glucose Toxicity: The First Steps of Causation Biochemical Consequences of Raised Intracellular Glucose Coupling of Disordered Biochemistry to Dysfunction Abnormalities of Cell Physiology Summary

in 1893. The frequency with which the autonomic nervous system is affected was recognized much later.265 The studies by Fagerberg,96 Mulder and co-workers,214 and Pirart249 showed that the incidence of neuropathy increased with the duration of diabetes, and that there was a link between neuropathy and other complications of diabetes, namely retinopathy and nephropathy. Based on clinical (symptomatic) assessment only, about 10% of patients with diabetes have neuropathy,221,233 but with a more detailed assessment protocol that includes quantitative sensory and autonomic testing and nerve conduction studies, the figure is much higher: just under half (47.6%) of all diabetic patients have a neuropathy.85 1951

1952

Diseases of the Peripheral Nervous System

EPIDEMIOLOGY, CLINICAL COURSE, AND RISK FACTORS Epidemiologic studies of diabetic neuropathy have yielded variable results because of differences in defining diabetic neuropathy, the tests used to assess the neuropathy, and the type of patient population studied.

lence of diabetic neuropathy is of interest, the critical information to obtain is that related to the severity of the neuropathy, because the complications of a bad neuropathy in terms of pain, foot ulcers, and neuropathic joints are the greatest concern to the patient and doctor. This information is lacking at present.

Natural History Minimum Criteria for Diagnosis The distal symmetrical sensory neuropathy associated with diabetes mellitus is a complex disorder in that there exists a spectrum of clinical abnormalities reflecting the variable involvement of small and large fibers. A single test is therefore not going to be a sufficiently sensitive tool to provide accurate information on prevalence. Is multiple testing the answer? Pirart, in his 25-year prospective study between 1947 and 1973,249 used absent ankle reflexes and impaired vibration sensation as the diagnostic criteria and found that the prevalence of neuropathy rose from 7.5% at the time of diagnosis of diabetes to nearly 50% after 25 years. We now know that these clinical signs will vary with age, and in individual cases over the age of 60 years, it may be difficult to ascertain whether any changes found are due to diabetes or to aging. Using a combination of clinical symptoms and signs, quantitative sensory testing, nerve conduction studies, and heart rate variability (HRV), Dyck and colleagues85 found the prevalence of neuropathy to be 54% in type 1 and 45% in type 2 diabetic subjects in a population-based study (Rochester Diabetic Neuropathy Study). The reality is that, to assess large numbers of diabetic patients, the screening test has to be simple, quick to administer, and inexpensive. That is why most of the epidemiologic studies to date have used a variable mixture of symptoms and signs or one test, such as vibration threshold testing. Using these simpler markers of neuropathy, it appears that, for type 2 diabetics, the prevalence of neuropathy is about 30% in hospital-based populations and slightly less than that (~20%) when the population is derived from general practices and the local community. For type 1 diabetics, the trend is the reverse, with a 13% to 23% rate in hospital cases and a slightly higher rate (13% to 34%) in the community. These figures appear to be similar in the Middle East (Egypt and Saudi Arabia), but Mauritius has the lowest recorded prevalence rate290 at 8.3% of the total diabetic population. There is no clear explanation for this low figure. The issue of standardization of diagnostic tests is important, and the problems of trying to extrapolate prevalence data from various studies using different tests for neuropathy was highlighted in the Diabetes Control and Complications Trial (DCCT),75 wherein the prevalence varied from 0.3% to 21.8% depending if neuropathy was assessed using symptoms and signs or the presence of nerve conduction abnormalities in a minimum of two nerves, respectively. Although there are valid reasons why having some idea of the preva-

It is known that, if one takes one or more markers of neuropathy, such as absent reflexes or nerve conduction studies, over a period of time the frequency of abnormalities will increase with the duration of diabetes. It has been pointed out by certain authors79 that the various surrogate markers of neuropathy produce variable results when estimating the incidence of neuropathy, and more importantly provide little useful information on the clinical severity of the neuropathy. These authors have recommended that a better understanding of the natural history of diabetic polyneuropathy could be obtained by using composite scores of impairment (e.g., neurologic assessment of the legs and the results from seven different tests) or utilizing a rating scale to stage the severity of the neuropathy (Table 85–1). In a population-based study in North America, the Neuropathy Impairment Score identified that 10% of neuropathies had worsened over 2 years, 81% were unchanged, and 9% had shown improvement.80 The revised form of Table 85–1 has been discussed in detail by others.79 The N2 grade can be subdivided into N2a and N2b, with the latter being more severe in having significant distal lower limb muscle weakness, but not bad enough to cause immobility (grade N3).

Risk Factors The most important risk factor for the development of diabetic polyneuropathy is hyperglycemia, and several studies have clearly shown that its tight control delays the progression of neuropathy in patients with type 1 diabetes (e.g., the DCCT study75), although this does not appear

Table 85–1. Staging the Severity of Diabetic Polyneuropathy • N0: diabetic patient; no symptoms or signs of neuropathy • N1: asymptomatic; signs of neuropathy (e.g., distal sensory loss ⫾ absent ankle jerks) • N2: symptomatic neuropathy (e.g., a variety of pain symptoms or numbness) N2a: less severe N2b: more severe; with weakness of dorsiflexion • N3: disabling polyneuropathy—patient cannot walk because of weakness or sensory ataxia, or cannot use hands because of numbness, or has disabling neuropathic pain; also included is the presence of symptoms of autonomic failure

1953

Diabetic Neuropathies

Table 85–2. Important Risk Factors in Diabetic Polyneuropathy

Table 85–3. Classification of the Diabetic Neuropathies

• • • • •

• Impaired glucose tolerance and hyperglycemic neuropathy • Generalized neuropathies Sensorimotor Acute painful (including treatment induced) Autonomic Acute motor • Focal and multifocal neuropathies Cranial Thoracolumbar Lumbosacral radiculoplexus (Bruns-Garland syndrome) Focal limb (entrapment and compression) • Superimposed chronic inflammatory demyelinating neuropathy • Hypoglycemic neuropathy

Hyperglycemia Duration of diabetes Age Hypertension Smoking

to be the case for those with type 2 diabetes who were followed for a period of 12 years in the U.K. Prospective Diabetes Study (UKPDS).338 Other important risk factors are listed in Table 85–2. The duration of diabetes and the age of the patient were found by Young and colleagues374 to be two additional risk factors, although it is difficult to assess the effect of age on diabetic neuropathy per se, because in normal individuals loss of ankle reflexes and some impairment of vibration sensation at the toes may be found. Hypertension and smoking are risk factors for neuropathy in type 1 diabetic patients, but again, probably not so in type 2 diabetics.108 Other factors that have been linked to neuropathy include the patient’s height,2,318 level of alcohol consumption, and high cholesterol and triglyceride levels. The search for candidate genes is ongoing. In a 7-year prospective study, Dyck and co-workers81 identified microvessel disease in the form of retinopathy, chronicity of exposure to elevated blood glucose, and having type 1 diabetes as the main factors linked to severity of diabetic polyneuropathy.

CLINICAL FEATURES Diabetes mellitus affects the peripheral nervous system in a complex variety of ways, and the differing clinical patterns produced can occur in combination or in isolation in individual patients. The classification system outlined in Table 85–3 is based on that originally proposed by Thomas in 1973.321

Neuropathy Related to Impaired Glucose Tolerance and Hyperglycemia Neuropathy can be the presenting feature of impaired glucose tolerance (IGT). Studies looking at the prevalence of IGT in patients with “idiopathic neuropathy” have found that 25% to 36% of such cases will show IGT diagnosed with a 2-hour oral glucose tolerance test using a 75-g glucose load.297,312 The fasting blood glucose and hemoglobin A1 (HbA1) levels are normal. The neuropathy is likely to be painful,225 and this predominant small fiber pattern is supported by abnormal epidermal innervation

seen on skin biopsy.298 It is not known whether the neuropathy associated with IGT is reversible. What seems to be clear is that an oral glucose tolerance test is an important investigation of patients with a painful “idiopathic” neuropathy. Uncomfortable sensory symptoms of tingling paresthesias, pain, or hyperesthesias in the feet were described in the early literature159 as being clinical features in newly diagnosed or poorly controlled diabetics, and the symptoms were noted to resolve with normalization of the blood glucose level. Pain can also occur transiently after diabetic ketoacidosis. This has been called hyperglycemic neuropathy,353 and its potential reversibility is supported by improvement in nerve conduction studies seen with better diabetic control.128,350 A further feature of hyperglycemic neuropathy is an abnormal resistance to the effect of ischemia on nerve function,304 perhaps related to the fact that in this situation the diabetic nerve can switch to anaerobic glycolysis.

Distal Symmetrical Sensorimotor Polyneuropathy In the presence of established diabetes mellitus, this should not be a difficult diagnosis to make. Because of the strong relationship of neuropathy with retinopathy and nephropathy, it has been suggested79 that diabetic symmetrical distal polyneuropathy (DSDP) should only be diagnosed when these other diabetic complications are present. This may certainly hold true in clinical studies of DSDP, but in general neurologic practice the clinical diagnosis of DSDP is going to be made with an appropriate history and examination findings. The need for neurophysiologic investigation in most cases is debatable (see later). Typical Symptoms Symptoms are not an essential requirement to make a diagnosis of DSDP. Indeed, in the Rochester populationbased study, symptomatic polyneuropathy (grade N2)

1954

Diseases of the Peripheral Nervous System

was found in only 13% of the diabetic patients,85 and an overall figure of about 20% was quoted by Harris and colleagues.136 When sensory symptoms are present, they are usually constant but can be variable or intermittent. The feet are predominantly affected, and the symptoms can progress proximally. Of 105 patients with painful neuropathy, the sites most often involved were the feet (96%), balls of the feet (69%), toes (67%), dorsum of the foot (54%), hands (39%), sole of the foot (37%), calves (37%), and heels (32%); patients noted that symptoms were at their worst when in bed at night or when tired or stressed.111 It is possible to try to infer from the symptoms the involvement of particular nerve fibers—for example, numbness, tingling, pins and needles, and tightness of the feet suggest that the thickly myelinated A␤ fibers are damaged, whereas burning or cold sensations and pains of variable types with autonomic symptoms suggest that damage has been done to thinly myelinated A␥ and unmyelinated C fibers (“small fiber neuropathy”). From the patients’ perspective, the main symptoms include the following: 1. Numbness or a feeling of walking on cotton wool or wearing a thick sock 2. Pains that can be dull, constant and boring in type, or more spontaneous sharp, shooting, or stabbing in nature; a sensation as if walking on pebbles 3. Tingling, pins and needles 4. Hot or cold sensations (e.g., “burning feet”; “like walking on hot sand”) 5. Allodynia (pain caused by an otherwise nonpainful stimulus, such as light touch or stroking); this can be very troublesome at night when the feet and legs rub against the bedclothes 6. Cramps in the calves and foot muscles Findings on Examination There may be wasting of the extensor digitorum brevis, but significant distal weakness in the lower limbs is uncommon, and if present, should be a trigger to look carefully for other causes of neuropathy (see later). The ankle reflexes are absent, and this should be the cornerstone of the clinical findings for diagnosing DSDP. The sensory loss is in a length-related distribution, with the toes and feet being most affected.270 Typically there is loss or impairment of all sensory modalities, with vibration sense often the first to go. It is important to check for deep pain sensation by applying pressure either to the nail of the big toe or to the Achilles tendon. Rarely, the pattern of involvement is confined to pain and temperature sensation—a small fiber neuropathy producing a “pseudosyringomyelic” clinical picture, frequently leading to the occurrence of foot ulcers and neuropathic joints.41,275 The question of whether selective involvement in DSDP can be confined to vibration, joint

position, and light touch (“large fiber” neuropathy) is debatable, because there is usually some evidence of pain and temperature involvement in such cases. As the sensory loss extends proximally from a “sock” to a “stocking” distribution, the fingertips become involved. It is uncommon for numbness to extend to involve the whole of the legs and arms, but when this does happen, the midline of the front of the chest and abdomen are affected with a “breastplate” pattern of sensory loss. This may extend laterally around the chest and abdomen, that is, in a length-dependent manner.273,275 The examination should also include an assessment of autonomic function (resting pulse, lying and standing blood pressure, and pupillary responses). In a patient with diabetes, if there are clinical autonomic abnormalities, a DSDP is invariably present, whereas Tackmann and colleagues315 noted that, in their patient group with DSDP, only just over 50% of cases had autonomic neuropathy.

Acute Painful Neuropathy An acute neuropathy is a relatively rare occurrence in the course of diabetes mellitus. The phenomenon of acute painful neuropathy with weight loss is described here as a separate clinical entity as first described by Archer and colleagues.12 The clinical sequence starts when the patient initially loses a lot of weight over a short period of time, followed by a severe and unremitting burning pain mainly in the feet. Hyperalgesia (discomfort on touching the skin) can be very troublesome. There is no muscle weakness and sensory loss may be minimal; ankle reflexes may be present or absent. Reactive depression is common, as is impotence. Maintaining good glycemic control is important, and this will eventually lead to an increase in body weight—the first sign of recovery. As the patient regains body weight the pains subside, but recovery can take up to 10 months, and even then it may be incomplete. This is probably the same cohort of patients described by Ellenberg as having diabetic neuropathic cachexia,92 and recurrence is unlikely. An acute painful diabetic neuropathy associated with weight loss has also been described in girls with anorexia.303 Other authors have reported an acute symmetrical painful neuropathy occurring a few weeks after the diagnosis of diabetes, but not associated with dramatic weight loss.345 Nerve biopsy showed some evidence of acute axonal degeneration, with a lymphocytic infiltrate identified in one case and endoneurial mast cells in the other. Further work is needed to look into the possibility of these acute painful neuropathies being immune mediated. Treatment-Induced Neuropathy In 1933, Caravati54 described a diabetic patient in whom insulin treatment triggered an acute painful neuropathy, a phenomenon he termed “insulin neuritis.” Although not common, there have been a number of reports of this

Diabetic Neuropathies

treatment-induced neuropathy to indicate that probably it is not just a chance occurrence. Detailed studies have not been performed and its pathogenesis remains unclear. An inflammatory response has not been demonstrated, and the trigger may not necessarily be insulin, because this neuropathy has been reported with oral hypoglycemic agents.91 An associated acute proliferative retinopathy or autonomic neuropathy has been reported,184,317 as has the development of acute bilateral cataracts377 in a young man with newly diagnosed type 1 diabetes who developed what appeared clinically to be an acute painful neuropathy with weight loss linked to tightening of glucose control with insulin. Treatment-induced neuropathy is even rarer in children.361 Controlling the pain can be difficult, often requiring opiates, but by maintaining good glucose control, the outcome is favorable, with resolution of pain over a period of up to 12 months. The neurologic signs in two cases reported by Said and colleagues remained unchanged after 2 years.271 Examination may be normal, as may nerve conduction studies. If there are abnormalities, they are mild, but nerve biopsy shows evidence of a more chronic axonal neuropathy with prominent regenerative activity.184 Epineurial arteriovenous shunting with proliferation of “new vessels” that leaked more fluorescein than normal (similar to new vessels of the retina) was shown on the surface of sural nerves in the patients reported by Tesfaye and co-workers.317 Although the latter observation emphasizes the possible contribution of vascular factors in the pathogenesis of this neuropathy, the overwhelming evidence points to a metabolic trigger. Or could this entity also possibly be immune mediated?

Diabetic Motor Neuropathy Diabetic neuropathy is primarily a distal sensory polyneuropathy. A number of situations exist in which muscle weakness is the major feature, and the importance of considering this entity is that chronic inflammatory demyelinating polyneuropathy (CIDP) may be more common in diabetic patients,289,308 and of course potentially reversible with immunotherapy. • Acute motor neuropathy—Guillain-Barré syndrome (GBS) can occur in diabetic patients and probably explains this type of neuropathy. Patients should be treated with intravenous immunoglobulin (IVIG) or plasma exchange when indicated. The association of GBS with diabetes is probably fortuitous, but no studies have looked at whether there might be an association between the two conditions. • Subacute or chronic distal motor neuropathy—In some diabetic patients with good glycemic control, a progressive distal motor neuropathy was reported by Timperley and co-workers324 to be caused by multiple small infarcts secondary to advanced microvascular disease. Nerve

1955

conduction studies showed an axonal process. However, the pattern of motor weakness was asymmetrical in many of the cases, and perhaps reminiscent of a radiculoplexus neuropathy (Bruns-Garland syndrome). Of more importance is identifying if a diabetic patient might have CIDP; in a patient with an established diabetic polyneuropathy, if there is a more rapid rate of progression than expected, one should consider the possibility that there might be a superimposed CIDP. In their study of 100 diabetics with symptomatic neuropathy referred to a specialist peripheral nerve center in Paris, Lozeron and colleagues found that 9 cases turned out to have CIDP.192 Once the possibility of CIDP has been considered, the diagnosis is confirmed by neurophysiology showing at least three electrical criteria of demyelination; however, because of overlapping results, it has been suggested that, even when there are only two electrical changes of demyelination, a trial with immunotherapy should be considered.339 There do not appear to be any clinically distinguishing features of CIDP in nondiabetic and diabetic patients.121,135 Investigations Nerve conduction tests may show more evidence of axonal loss in the diabetic ⫹ CIDP group, presumably reflecting the presence of an underlying diabetic axonal polyneuropathy.121 Cerebrospinal fluid (CSF) examination is unlikely to be helpful because the protein content can be elevated in diabetic polyneuropathy. The presence of oligoclonal bands may provide a clue of an ongoing inflammatory process but is not diagnostic. Nerve biopsy does not help with distinguishing between diabetic polyneuropathy with or without superimposed CIDP.135,308 Positive immunoreactivity for matrix metalloproteinase-9 in sural nerve biopsy may help to identify those cases with superimposed CIDP.152 Treatment If the clinical features are those of a progressive, predominantly motor neuropathy with convincing changes of demyelination on nerve conduction studies, then a diagnosis of CIDP can be considered, and the initial treatment to consider is IVIG (standard dose of 0.4 g/kg/day for 5 days). One study showed significant improvement in over 80% of patients.288 Although the time course to improvement is similar in diabetic and nondiabetic patients with CIDP, the level of functional recovery may be comparatively less in those with diabetes.121 In the longer term, treatment options depend on the response and duration of improvement after IVIG. If there was no recovery, plasma exchange might be considered (depending on the patient’s cardiovascular status). If there was a long period of good improvement but then the neuropathy deteriorated, a further treatment with IVIG would be sufficient. Conversely, if the response period was short, oral immunosuppression with azathioprine should be

1956

Diseases of the Peripheral Nervous System

considered. If a more rapid immunosuppression is needed, then oral steroids may have to be considered as a short-term measure—or possibly oral cyclophosphamide or mycophenolate, even though there are no trial data on their efficacy in idiopathic CIDP, let alone CIDP in a diabetic. Differential Diagnosis In the Rochester Diabetic Neuropathy Study, 10% of diabetic patients had an alternative cause for neuropathy other than diabetes.85 In a specialist center, however, 26% of diabetic patients with symptomatic neuropathy had a cause other than diabetes.192 In the last group, the patients were referred because there was something unusual or severe about the neuropathy. Although there is no unique clinical picture to the mainly sensory polyneuropathy, the diabetic patient with a neuropathy should not present a major diagnostic problem, and therefore a few simple investigations are sufficient (Table 85–4). Diabetic polyneuropathy comes on gradually and is generally slowly progressive. The more acute painful “small fiber” neuropathy with associated weight loss is one variant with a different time course; otherwise, a neuropathy that comes on acutely or subacutely has to be investigated in more detail mainly to exclude acute inflammatory demyelinating polyradiculoneuropathy, vasculitis, or CIDP (Table 85–5). Nerve conduction studies hold the key in showing changes of demyelination. If the CSF protein is greater than 1.5 to 2 g/L, this adds to the suspicion of an underlying inflammatory neuropathy because in diabetic polyneuropathy CSF protein values of over 1.2 g/L are rare. Diabetic polyneuropathy is mainly sensory in type and invariably symmetrical in its distribution. Therefore, if the pattern, either on history of development of the neuropathy or on examination, is asymmetrical or if weakness—with or without muscle wasting—is a prominent feature, alarm bells should ring that there might be an alternative or an additional cause for the neuropathy. Again, vasculitis and CIDP come to the top of the differential diagnostic list. The other factor to consider when assessing the likelihood of the polyneuropathy being due to diabetes is the

Table 85–4. Investigations Recommended in a Diabetic Presenting with a Mainly Sensory Distal Polyneuropathy • Urinalysis for protein/glucose/microscopy—for evidence of nephropathy • HbA1/glucose—to assess control of diabetes • Urea and electrolytes • Liver function tests, including ␥-glutamyltransferase • Thyroid function tests • Serum protein electrophoresis • Vitamin B12 level

Table 85–5. Differential Diagnosis of Diabetic Polyneuropathy and Further Investigations To Be Considered Clinical Pattern

Possibility

Investigation

• Acute onset

AIDP Porphyria Vasculitis CIDP CIDP, vasculitis

NCS ⫾ CSF Urine porphyrins

• Subacute • Asymmetrical • Motor ⬎ sensory • Large ⬎ small fiber involvement

• Selective small fiber involvement

CIDP Paraproteinemic drugs (cisplatin, etc.) B12 deficiency Ganglionopathy

Amyloidosis HSAN

NCS ⫾ CSF Autoimmune screen, NCS NCS IgM paraprotein B12 level CxR, CT chest/abdomen Tumor markers TTR gene mutation, nerve biopsy Genetics?

AIDP ⫽ acute inflammatory demyelinating polyneuropathy; CIDP ⫽ chronic inflammatory demyelinating polyneuropathy; CSF ⫽ cerebrospinal fluid; CT ⫽ computed tomography; CxR ⫽ chest radiograph; HSAN ⫽ hereditary sensory and autonomic neuropathy; IgM ⫽ immunoglobulin M; NCS ⫽ nerve conduction studies; TTR ⫽ transthyretin.

pattern of nerve fiber involvement. The full spectrum, from small fiber to large fiber damage, can be encountered in diabetic polyneuropathy, but most commonly there is an even loss of both fiber types. If examination reveals significant vibration and joint position sensation loss (perhaps with pseudoathetosis and sensory ataxia), with relative sparing of pain and temperature sensation (i.e., large fiber neuropathy), then one has to consider, for example, an immunoglobulin M paraproteinemic neuropathy, or perhaps a sensory ganglionopathy. When the clinical findings point to a selective small fiber neuropathy with or without autonomic neuropathy, one should think of alternative possibilities such as amyloidosis or hereditary sensory and autonomic neuropathy—both usually with a prominent autonomic neuropathy.

Autonomic Neuropathy Symptoms relating to autonomic involvement in diabetic patients have been recognized since the last century,45 but the frequency with which the autonomic system is affected was first stressed by Jordan159 in 1936 and then by Rundles265 in 1945. Over the last few years there have been considerable advances in knowledge, resulting in a better understanding of the natural history of this variety of diabetic neuropathy.

Diabetic Neuropathies

Prevalence Because of the variety of clinical tests employed and the nonspecific symptoms of autonomic dysfunction, interpretation of data on the prevalence and incidence of autonomic neuropathy from epidemiologic studies of diabetic neuropathy is difficult. The two population-based studies, one in Oxford219 and the other in Pittsburgh,198 recorded mainly cardiovascular autonomic function tests (HRV and blood pressure responses). In the Oxford Community Diabetes Study, 43 type 1 and 202 type 2 diabetics were tested from a total of 412 diabetic patients identified in the community population of 29,873. One or more cardiac autonomic tests were abnormal in 20.9% of type 1 and 15.8% of type 2 patients, but the lack of data from a control population means that one cannot extrapolate an accurate estimation of the prevalence of autonomic neuropathy. The Pittsburgh Epidemiology of Diabetes Complications Study found orthostatic hypotension (drop in blood pressure of ⬎ 30 mm Hg) in 3.4% of patients, and the most common symptom suggestive of autonomic dysfunction was unawareness of hypoglycemia (26%). Other symptoms were uncommon (0 to 8% of patients). Multicenter clinic-based studies such as the EURODIAB IDDM Complications Study 305 found orthostatic hypotension in 5.9%, and abnormal HRV in 19.3% of patients. Ziegler and colleagues380 (DiaCAN study) reported that, in an unselected cohort of type 1 diabetics, there were borderline cardiac autonomic abnormalities in 8.5% and definite abnormalities in 16.8% of cases, and the prevalence in type 2 patients was 12.1% and 22.1%, respectively. It is also recognized that autonomic function tests can be abnormal at the time of diagnosis of diabetes (Table 85–6).177,378 Diabetes is the most common organic cause of male erectile dysfunction, and data from the Massachusetts Male Aging Study show that up to 64% of men with treated diabetes have some degree of erectile problem.102 Control and duration of diabetes, the presence of

Table 85–6. Clinical Manifestations Associated with Diabetic Autonomic Neuropathy • Cardiovascular

• Gastrointesinal

• Urogenital

• Sudomotor function • Respiratory • Pupillary function

Increased heart rate Impaired cardiac function Painless myocardial infarction Orthostatic hypotension Abnormal esophageal motility Gastroparesis Diarrhea and constipation Erectile dysfunction Impaired testicular pain Retrograde ejaculation Anhidrosis Gustatory sweating Sleep apnea Small, unreactive pupil

1957

neuropathy, retinopathy, or nephropathy, and smoking were all identified as risk factors for impotence.101 The gastrointestinal tract does not have the range of noninvasive investigations available for the cardiovascular system to assess autonomic function. The largest clinic-based studies have therefore concentrated on clinical symptoms,94,173,283 but there is considerable variability in their findings, with only early satiety, fullness after meals, and nausea being significant symptoms in both type 1 and type 2 diabetic patients compared with controls. Surprisingly, diarrhea was not more frequently encountered in type 1 and 2 diabetics in two European studies94,283 but was more prevalent (35% vs. 5%) in type 2 diabetic patients in a study from Hong Kong.173 Natural History No population study has been done with type 1 diabetic patients. A 10-year study of patients with type 2 diabetes in Finland334 found that cardiac autonomic (parasympathetic) function assessed with an E/I ratio (heart rate on expiration/heart rate on inspiration) rose from 5% (controls 2%) at the beginning to 65% (controls 28%) at the end of the study. Changes in systolic blood pressure on standing (sympathetic function) increased from 7% to 24% (controls 6% to 9%) over 10 years. The development of cardiac autonomic neuropathy correlated with poor glycemic control. This was also noted by Ziegler and co-workers383 in a clinic-based study in Germany, who found that progression of abnormal cardiac function tests (HRV) in type 1 diabetic patients was most marked in those with poor glycemic control. Further evidence supporting a gradual progression of autonomic neuropathy with time comes from clinic-based studies showing a deterioration in autonomic function test scores in 57% of type 2 diabetics over 5 years,254 and a separate 4-year study of older type 2 diabetic patients showed that autonomic function test abnormalities increased from a prevalence of 41% at baseline to 64%.216 Similar progression has been noted in children with type 1 diabetes, in whom the prevalence of HRV abnormality increased from 27% to 56% after 10 years.301 Although the overwhelming evidence shows that the prevalence of cardiac autonomic neuropathy increases with the duration of diabetes, one recent study showed no change in autonomic function test results in adults with type 1 diabetes over a 9-year follow-up.178 Risk What is the significance to the patient and doctor when the presence of autonomic neuropathy is confirmed? • Part of the high frequency of cardiac mortality in patients with diabetic neuropathy may be due to abnormal cardiovascular innervation. The syndrome of “sudden death at night” has been postulated to be due to autonomic dysfunction predisposing the patients to arrhythmias when they become hypoglycemic.

1958

Diseases of the Peripheral Nervous System

• Ventricular tachycardia and ischemic and nonischemic sudden death have been associated with reduced heart rate variability.26 • Autonomic neuropathy may contribute to the development of nephropathy.302,313 • In type 2 diabetic patients, the presence of cardiac autonomic neuropathy is an independent risk factor for stroke after 10 years.334 • Vigilance is required during anesthesia in diabetic patients with autonomic neuropathy, given the documentation of increased anesthetic-induced cardiorespiratory arrests and perioperative cardiac instability.181 Cardiovascular Disturbances Heart Rate Abnormalities. Typically, in a diabetic patient with autonomic neuropathy the resting heart rate is between 90 and 100 beats/min, and does not alter with standing, breathing, or mild exertion. Impaired Cardiac Function. Although abnormal ventricular function on exercise is an indicator of diabetic cardiomyopathy, there is evidence that impaired left ventricular function can occur in those patients with autonomic neuropathy in the absence of coronary artery disease.386 It is not known whether the presence of autonomic neuropathy influences the prognosis of diabetic cardiomyopathy. Painless Myocardial Infarction. Painless myocardial infarction appears to occur more commonly in the diabetic population than in controls (10% to 20% vs. 1% to 4%),174,205 and from autopsy and clinical studies, this has been attributed to autonomic neuropathy,6,95,215 although Koistinen and colleagues in a later study could not confirm this finding. 175 Patients with diabetic neuropathy and ischemic heart disease should therefore be educated to recognize other symptoms of impending myocardial ischemia, such as shortness of breath, a feeling of exhaustion, and nausea. Orthostatic Hypotension. The occurrence of orthostatic or postural hypotension, defined as a fall of 30 mm Hg or more on standing from a lying position, has long been recognized as a manifestation of established diabetic autonomic neuropathy. Postural syncope occurred in 8 of the 125 cases described by Rundles in 1945,265 but symptomatic hypotension generally only occurs if the drop is greater than 50 mm Hg and the final standing blood pressure is less than 60 to 70 mm Hg. It is mainly due to splanchnic vasomotor denervation,190 with a small additional effect from reduced peripheral vasoconstriction and impaired heart rate and cardiac output response to standing.352 Orthostatic hypotension can be aggravated by insulin.251,373 In normal subjects, insulin is both a vasodilator (via nitric oxide) and vasoconstrictor (via the sympathetic system),

but because vasoconstriction is poor in diabetic autonomic neuropathy, insulin will tend to cause the blood pressure to drop. It may also be exacerbated by drugs such as diuretics, tricyclic antidepressants, phenothiazines, and antihypertensives. In a 10-year follow-up study of type 1 diabetic patients, Sampson and co-workers276 noted that orthostatic hypotension can fluctuate over the years. It can also vary during the day, being more severe in the morning and after eating. Treatment of orthostatic hypotension is outlined in Table 85–7. The essence of management is to begin with simple nonpharmacologic advice and only if this is unsuccessful to consider drug treatment (discussed subsequently). Gastrointestinal Symptoms Gastrointestinal symptoms in patients with diabetes are likely to be multifactorial. Disordered motility, autonomic neuropathy, visceral hypersensitivity, Helicobacter pylori infection, drugs, and psychological factors are some of the variables that have to be considered. Although the prevalence of gastrointestinal symptoms is high in the general population, there is evidence that they are more common in diabetic patients, particularly women46; however, there is little evidence to support the idea that autonomic neuropathy makes matters worse.59 Studies have concluded that it is hyperglycemia and not autonomic neuropathy that has the most important impact on gastrointestinal symptoms.158,283 It is not clear by what mechanism raised glucose levels affect gastrointestinal symptoms. Abnormal esophageal motility was found to be more frequent in diabetic patients with evidence of either a somatic or an autonomic neuropathy,267 but de Boer and colleagues have clearly demonstrated that normal

Table 85–7. Management of Orthostatic Hypotension • Avoid exacerbation: for example, prolonged standing, exercising after meals or in the morning, hot baths, excessive alcohol intake, and carbohydrate-rich meals. Take a detailed drug history (antihypertensives, diuretics, tricyclic antidepressants) • Minimize symptoms by squatting, bending over, raising one leg onto a chair, or crossing the legs when standing • Tilt head of bed up by 20 degrees • High-salt diet • Fludrocortisone: starting dose of 0.1 mg at night; maximum dose 0.4 mg • Other drugs include desmopressin (DDAVP) given at night and midodrine (␣1 agonist) • Octreotide, a somatostatin analogue (dose range 25 ␮g bid to 50 ␮g tid) administered subcutaneously helps to prevent postprandial and exercise-induced hypotension in addition to orthostatic hypotension

Diabetic Neuropathies

individuals made hyperglycemic show abnormal esophageal function.71 The mechanisms underlying gastric emptying are poorly understood, and although delayed gastric emptying occurs in 30% to 50% of patients with long-standing type 1 and type 2 diabetes (the term gastroparesis being used when this emptying delay is severe141,385), others have reported on abnormally rapid gastric emptying in type 2 diabetic patients.284 Neither is specifically linked to the presence of an autonomic neuropathy, and it now looks as though the motility abnormalities in the stomach in diabetic patients are caused by hyperglycemia. Likewise, small bowel motility is often abnormal in diabetic patients, and probably contributes to “diabetic diarrhea” either because the transit time for food is too quick and absorption is impaired, or the transit time is too slow, thereby promoting the overgrowth of bacteria. Diarrhea is a problem encountered in 20% patients with poorly controlled type 1 diabetes103; although it has been attributed to autonomic denervation,76 there have been no large clinical studies to confirm this association, and the problem is likely to be multifactorial, with hyperglycemia playing an important role. Poor blood glucose control is also now considered the most likely contributor to impaired gallbladder motility and constipation. Autonomic denervation may contribute to abnormal anal sphincter function, but so does hyperglycemia.268 Urogenital Dysfunction Neuropathy, arteriosclerosis, and changes in the corpus cavernosal tissue contribute the organic components to the problem of erectile dysfunction, which affects 30% to 50% of male diabetic patients. Anxiety and depression are also important factors.193 Population studies have shown that, as in nondiabetic men, the prevalence of erectile dysfunction increases with age, and that it is much more common in diabetic men: 1.1% at age 21 to 30 years and 47% in those over 47 years of age.172 Other risk factors for include, the presence of neuropathic ulcers or joints, poor glycemic control, nephropathy, retinopathy, hypertension, and duration of diabetes.101 Two other sexual problems include impaired pain sensation in the testicles, which has been considered a sign of autonomic neuropathy, and retrograde ejaculation,393,127 which is presumed to occur because of sympathetic denervation of the internal sphincter of the bladder. Although sexual dysfunction is very common, it is more often than not a subject that is avoided by both patient and doctor.132 This neglect is best addressed by always asking the patient about erectile function and libido. If a problem is identified, ensure that it is not aggravated by medication such as an antidepressant or antihypertensive, and check serum testosterone in addition to other routine blood tests. The patient may be considered for a trial with sildenafil (Viagra). Should this not be effective, with the patient’s approval, a referral should then be made to the urology

1959

service, so that a more detailed clinical and psychological assessment can be undertaken before proceeding to further investigations (intracavernous injection of vasoactive drugs, cavernous electromyography [EMG], and duplex ultrasound of the cavernous arterial flow). Loss of sensation of bladder filling, reduced urinary flow, incomplete bladder emptying, and a tendency to recurrent bladder infections may all be symptoms of the “diabetic bladder.” It is believed to be due to loss of myelinated afferent (bladder sensation) and unmyelinated efferent (to detrusor muscle) fibers. Sudomotor Dysfunction Using thermoregulatory sweat tests in diabetic patients with clinically evident neuropathy, Fealey and co-workers100 showed that 48 of 51 patients had abnormal results and that, although the pattern of sweat loss was distal and symmetrical in 65%, segmental and dermatomal patterns of sweat loss were also identified, and the majority (78%) had a combination of two or more patterns. The severity of autonomic neuropathy correlated with the degree of body anhidrosis— those with complete anhidrosis had the most profound autonomic disturbance. The use of thermoregulatory sweat tests and other methods of assessment of sudomotor function are outlined in detail elsewhere (see Fealey99). The interesting phenomenon of “gustatory sweating” is now thought unlikely to be due to aberrant parasympathetic reinnervation, as was initially proposed, because the problem can disappear following renal transplantation.291 The trigger zone appears to be the taste buds, because food or drink has to be in the mouth to trigger the facial sweating. The mechanism of the link with nephropathy is not known. Pupillary Dysfunction Although abnormal pupillary responses are common in diabetic patients (typically the pupil size becomes smaller with a reduced response to light),299 it is important to remember that other lesions, from the optic nerve to the optic tract, also result in pupillary abnormalities. The prevalence of pupillary autonomic dysfunction is related to the duration of diabetes and therefore variable, with a range of 23% to 59%.309 Retinal function will influence the pupil responses, so if there is a severe retinopathy causing blindness, then the pupils will be unreactive. Glycemic control is the other recognized influence on the rate of deterioration of pupillary reaction.243 There is no information on whether the presence of pupillary autonomic dysfunction carries any prognostic value. Disturbances of Respiratory Control Although a number of abnormalities of respiratory function have been documented in diabetic autonomic neuropathy, their importance in the clinical management of patients does not seem to be significant. Bronchial tone is reduced77 and there is impairment of bronchoconstriction.137,342

1960

Diseases of the Peripheral Nervous System

Possible damage to the cholinergic connections between the carotid bodies and the brainstem is thought to be responsible for a reduced ventilatory response to hypoxia,300 but no clear pattern emerged from looking at the ventilatory response to hypercapnia.300 Because sleep apnea is associated with an increased risk of cardiovascular mortality and sudden death, could this be one mechanism to account for increased sudden death in patients with diabetic autonomic neuropathy? Probably. Although Catterall and co-workers55 did not find any difference in the sleep-related breathing pattern of diabetic patients with and without autonomic neuropathy, sleep apnea has been found to be more frequent in more recent studies.28,220 Unawareness of Hypoglycemia Hypoglycemia of sufficient magnitude normally gives rise to an autonomic response with bradycardia and mild hypotension, followed by an epinephrine response resulting in cutaneous vasoconstriction, sweating, and a feeling of anxiety. Impairment of this response in diabetic patients has been attributed to autonomic neuropathy, but this is not the case in most patients, because numerous studies have convincingly demonstrated that it is a tendency to preceding hypoglycemia that leads to the syndrome of hypoglycemic unawareness in type 1 diabetics.66,340 Recurrent episodes of mild hypoglycemia, caused mainly by the fact that absorption of subcutaneously administered insulin is variable, cause increased delivery of glucose to the brain, and so when the next episode of hypoglycemia occurs, the brain, with its now relatively normal level of glucose, is not driven to generate the regulatory response and autonomic symptoms. Therefore the normal response to that subsequent episode of hypoglycemia is blunted. Increased cortisol secretion may also play a role in this impaired response.70 Improving blood glucose control and prevention of hypoglycemia corrects the unawareness of hypoglycemia. In those with long-term diabetes (more than 15 years’ duration), autonomic neuropathy does seem to have a role in hypoglycemic unawareness, in that patients in this category do not lose the feeling of unawareness despite avoiding hypoglycemia.97 Treatment of Autonomic Dysfunction Cardiac. The management of postural hypotension comprises giving advice on lifestyle issues and, when needed, drug treatment. The patient should be advised to lift the head of the bed at night (by 20% to 30%), increase daily salt intake, and eat small meals frequently. Crossing the legs while standing also helps. The patient should try to avoid aggravating factors such as high environmental temperatures (including hot baths), highcarbohydrate meals, alcohol, and drugs such as diuretics, antihypertensives, nitrates, and tricyclic antidepressants.

The drug of choice is fludrocortisone (50 to 100 ␮g at night). Midodrine is an alternative drug (2.5 mg three times daily, increasing up to 10 mg three times daily if required). Gastrointestinal. Metoclopramide and domperidone increase gastric motility and are helpful in gastroparesis, as are erythromycin and cisapride, but the most important change to be made is to tighten glucose control. Diabetic diarrhea is often caused by bacterial overgrowth, and a broadspectrum drug such as neomycin or tetracycline can be tried.

Focal and Multifocal Neuropathies This group of neuropathies should not be considered in isolation because they often coexist with the distal sensory polyneuropathies. Cranial Neuropathies Although population studies have not been carried out that prove an association between diabetes and cranial nerve palsies, it does seem highly likely that isolated neuropathies of the extraocular nerves and the facial nerve occur at an increased rate in diabetic patients. Of 4278 cases of oculomotor, abducens, and trochlear nerve palsies (incorporating previous studies of Rucker264 and Rush and Younge266), 103 (2.4%) were related to diabetes.261 An abducens (VIth cranial nerve) palsy was the most common, followed by oculomotor (IIIrd cranial nerve) and then trochlear (IVth cranial nerve) palsies. Older age groups are most commonly affected. In the series of 24 cases reported by Zorilla and Kozak,387 all except 3 patients were over 50 years old at the time of onset of symptoms. Very occasionally recurrent or multiple palsies may occur.263,387 In the VIth nerve palsy, the onset is abrupt and usually painless. Detailed investigations are rarely required because full recovery occurs in the vast majority over 3 to 5 months. In those with a IIIrd nerve palsy, many will have evidence of a distal polyneuropathy. The onset of symptoms is again abrupt, with half of cases reporting retro-orbital aching pain that may persist for several days. It is recognized that the pain may occasionally antecede the paralysis by a few days. Complete resolution usually occurs in 3 to 6 months. In addition to the ptosis and difficulty with upward, downward, and medial movement of the eye, the “key” feature is sparing of pupillary function— attributed to the fact that pupillomotor fibers are located on the outer layer of the IIIrd nerve and that ischemia tends to affect the center of the nerve. Sparing of the pupil occurs in most cases,119 but subtle abnormalities of pupil size have been reported in about one third of diabetic cases with a IIIrd nerve palsy.150 Given that, in general, it is believed that an acute ischemic insult to a peripheral nerve results in loss of axons, the report by Asbury and colleagues14 of a pathologic study of one

Diabetic Neuropathies

Truncal Radiculoneuropathy This phenomenon is encountered in both type 1 and type 2 diabetic patients, usually in middle or old age, and appears to be more common in men than women.57,307 The onset is abrupt, usually with pain over a focal area of the chest/abdomen. The burning pain can be in a variety of distributions, reflecting the involvement of nerve root or intercostal nerve trunk, and is characteristically worse at night. Contact of clothing or bedclothes against the skin may be unpleasant. Examination findings may not be dramatic, but often it is possible to map out a well-demarcated area of altered sensation—either numbness or hyperpathia. The distribution of this sensory change is variable (Fig. 85–1) and may be missed unless the diagnosis is considered. Other associated clinical features include focal anterior abdominal wall weakness (Fig. 85–2) presenting as an abdominal hernia,30,236,356 and occasionally profound weight loss—a feature also encountered in diabetic acute painful neuropathy and diabetic radiculoplexus neuropathy. There may be evidence of an accompanying distal polyneuropathy and autonomic neuropathy24 and lumbosacral radiculoplexus neuropathy.21 Spontaneous recovery occurs over several months. Truncal radiculoneuropathy can, uncommonly, recur.307

patient with an acute IIIrd nerve palsy, showing a focal lesion with demyelination between the point of entry of two nutrient arteries into the nerve, is therefore surprising. One explanation for this apparent anomaly is that reperfusion after experimental ischemic nerve injury in animals can lead to focal demyelination.226 It is clear that more needs to be learned about the pathogenesis of the diabetic IIIrd nerve palsy. Is it a peripheral nerve trunk ischemic injury, or could it be more central as suggested by Hopf and Gutmann,140 who reported the presence of ischemic areas in the mesencephalon on magnetic resonance imaging (MRI) of the brain in a proportion of diabetics presenting with acute IIIrd nerve palsies? The differential diagnosis is with a posterior communicating artery aneurysm, but with aneurysm the pupil is invariably enlarged. It is advisable that computed tomographic angiography or magnetic resonance angiography be carried out in all cases of IIIrd nerve palsy with pupil involvement. Diabetic patients over the age of 50 with a pupil-sparing IIIrd nerve palsy do not need imaging and can be managed conservatively. The prevalence of diabetes or IGT in cases of Bell’s palsy in the general population is variable. There is probably enough information to conclude that diabetes is a predisposing factor, and that it does not adversely affect recovery.241 The only clinical variation from nondiabetic patients is the apparent lower incidence of impaired taste sensation in Bell’s palsy encountered in those with diabetes.234,239

A

1961

Investigation. This is a clinical diagnosis, and investigations rarely contribute to clinical management. Electromyography may show evidence of paraspinal or abdominal wall denervation. The site of the pathology remains unknown. Lauria and colleagues176 found loss of intraepidermal nerve

B FIGURE 85–1 Line drawings indicating the distribution of sensory loss in two episodes of diabetic truncal neuropathy in the same patient. (Data from Steward, J. D.: Diabetic truncal neuropathy: topography of the sensory deficit. Ann. Neurol. 25:233, 1989.)

1962

Diseases of the Peripheral Nervous System

FIGURE 85–2 Left abdominal pseudohernia caused by a truncal neuropathy in a diabetic patient. Presenting symptoms were left lower abdominal pain followed 6 months later by the pseudohernia. Cutaneous sensation over the pseudohernia was reduced, with a surrounding area of hyperpathia. The pseudohernia resolved after 12 months. (From Weeks, R. A., Thomas, P. K., and Gale, A. N.: Abdominal pseudohernia caused by diabetic truncal radiculoneuropathy. J. Neurol. Neurosurg. Psychiatry 66:405, 1999, with permission.) See Color Plate

fibers in skin biopsies taken from symptomatic regions, with evidence of histologic recovery with improvement of symptoms. They postulated that the pathology was either distal to or within the dorsal root ganglion. Treatment. Management involves controlling the pain. Topical capsaicin can be useful if the affected area is small. Differential Diagnosis. Depending on the site of the pain, patients may present as possible myocardial ischemia, cholecystitis, or appendicitis, and these can be ruled out by careful examination. In this mainly elderly group, herpes zoster infection needs to be considered, but with zoster there is usually a rash that may persist for up to 3 weeks. If the history goes back more than 6 months without any improvement, then the patient should have an MRI of the thoracolumbar spine to exclude disc disease or an intramedullary spinal cord abnormality. Lumbosacral Radiculoplexus Neuropathy (Bruns-Garland Syndrome) This clinical syndrome appears to be a distinct entity, more common in men with type 2 diabetes and occurring in middle to late life (see Chapter 86). It may be the presenting feature of diabetes and may predate the onset of

diabetes. Although the evolution of symptoms can be quite variable, the onset is marked by severe aching pain felt proximally in one (or both) legs or in the lower back. The symptoms are worse at night. Over a period of time, extending from a few days to several weeks, lower limb muscle weakness becomes the main clinical feature. Early on this is often proximal and unilateral, but when bilateral it is usually asymmetrical. When assessed carefully, involvement of distal lower limb muscles can be found, and if this is bilateral, it will again usually be asymmetrical. One or both knee jerks is depressed or absent, and the ankle jerks may also be absent if the weakness involves distal muscles or if there is evidence of an additional DSDP. The weakness may slowly progress to generalized lower limb paresis (“diabetic paraplegia”). Sensory loss is not prominent, but in addition to some numbness over the anterior thigh, there is often evidence of a stocking sensory impairment. Features of autonomic dysfunction are also common.237 Said and colleagues272 reported a predominantly sensory clinical picture in 4 of 10 cases. Weight loss is a common problem and may be dramatic. Evidence has now accumulated that this syndrome is caused by ischemic nerve injury and that this is secondary to a microvasculitis rather than a diffuse microangiopathy.79,164,186,272 The fact that at best only 20% to 30% of nerve biopsies show changes of microvasculitis, and that other studies have not encountered these changes, is in part probably due to the fact that the patchy distribution of inflammation means it may be missed in a limited nerve biopsy, and also the fact that biopsies have only been done on sensory nerves (the intermediate cutaneous nerve of the thigh and the sural nerve) in what is a predominantly motor disorder. This points up the limited value of routinely performing a nerve biopsy as an investigative tool in these cases. In a detailed clinical and pathologic study of lumbosacral radiculoplexus neuropathy occurring in nondiabetic subjects, Dyck and colleagues82 could find no difference between these cases and those with associated diabetes. A similar syndrome affecting the upper limbs (cervical radiculoplexus neuropathy) is uncommon, but was noted by Katz and co-workers162 in 9 of 60 patients who presented with a lumbosacral radiculoplexus neuropathy. The upper limb weakness was unilateral or asymmetrical (mainly distal) and could precede, coincide with, or follow the onset of lower limb weakness. Recovery was spontaneous over a period up to 9 months. Neuromuscular respiratory weakness has also been reported in association with diabetic lumbosacral radiculoplexus neuropathy.36 Investigations. The clinical diagnosis is often straightforward, and the range of investigations can be kept to a minimum (Table 85–8). It is worth confirming that markers of systemic inflammation are normal. Because the majority of patients will be elderly, most will have some

Diabetic Neuropathies

1963

Table 85–8. Investigation of Diabetic Lumbosacral Radiculoplexus Neuropathy

Table 85–9. Management of Diabetic Lumbosacral Radiculoplexus Neuropathy

• MRI of lumbosacral spine/pelvis • EMG/NCS

• Pain control; treatment of secondary depression • Maintain good diabetic control and nutrition • Physiotherapy, occupational therapy, and orthotic assessments

• CSF examination

• Nerve biopsy

To exclude nerve root compression and pelvic infiltration. Denervation changes in paraspinal, proximal, and distal leg muscles. There is often an associated distal polyneuropathy. Only helps if clinical picture is unusual. Protein level often raised ⬎ 1 g/L. No excess of lymphocytes. Should not be done routinely because it rarely alters management. To be considered in atypical cases.

CSF ⫽ cerebrospinal fluid; EMG ⫽ electromyography; MRI ⫽ magnetic resonance imaging; NCS ⫽ nerve conduction study.

degenerative changes in the lumbosacral spine, and care has to be taken not to overinterpret these changes as being the cause of the neurologic problem, leading to unnecessary spinal surgery.186 Infiltration of the pelvis by a malignancy has to be considered, particularly when there is dramatic weight loss, and computed tomography of the pelvis is the investigation of choice. Nerve conduction studies often show a coexisting axonal neuropathy, and EMG examination reveals evidence of denervation that is not necessarily confined to proximal lower limb muscles but may also be found in distal leg muscles. There may also be evidence of denervation in the other leg even though weakness is unilateral. CSF examination is probably not routinely necessary—the CSF is acellular and the protein content can be elevated. Neither is there a role for nerve and muscle biopsy unless carried out as part of a research study. Investigations to exclude other possible causes when the clinical picture is atypical are recommended and discussed later (see Table 85–11). Treatment. If the patient is bed bound, deep vein thrombosis prophylaxis should be instituted. Although some diabetologists convert patients from oral hypoglycemic drugs to insulin treatment at the onset of this neuropathy, there is no published randomized study that has looked at whether this is actually advantageous. The diabetes is often mild, and it is known that Bruns-Garland syndrome can occur in type 2 diabetics already on insulin.61 Pain is a major problem early on, and may be difficult to control. Treatments should ideally be tried one after the other, but after starting with analgesics such as tramadol or codeine phosphate, often amitriptyline or gabapentin, for example, will need to be added to maintain adequate pain control. Use of transcutaneous electrical nerve stimulation can be tried. Physiotherapy has an important role in helping the patient regain mobility (Table 85–9).

Having established that lumbosacral radiculoplexus neuropathy in diabetic (and nondiabetic) patients is probably immune mediated, the debate has opened on the potential role and use of immune-modulating therapies—steroids, IVIG, and plasma exchange in particular. All have different potential benefits and problems in this group of elderly diabetic patients. Clinical trials are underway, but until results become available, immunomodulating treatments cannot be recommended. The decision to embark on potentially problematic immune therapies has to be weighed against the natural history of this condition, which appears to be one of spontaneous recovery and therefore carries a good prognosis. But is this true? In a retrospective review of patients seen in a diabetic clinic, Coppack and Watkins61 reported that recovery began after 3 months and was usually “complete” by 18 months, although residual neurologic signs (reduced thigh circumference and absent ankle jerks) were documented. None of their 27 patients had long-term disability problems, and they concluded that the prognosis was very good. This is in contrast to the findings of a 12-year retrospective study of 29 patients seen in a neurology unit,237 which noted that over one third of patients made no significant recovery and that in under two thirds there was slow recovery but this was incomplete. The natural history of nondiabetic lumbosacral radiculoplexus neuropathy reveals a poor outcome, with only 3 of 42 patients making a full recovery in the study by Dyck and colleagues,82 and given the similarities in clinical presentation and peripheral nerve pathology, there is no reason to predict any better prognosis for the diabetic patients with this neurologic problem. Is it possible that the diabetologist and the neurologist are seeing patients from either end of the clinical spectrum of diabetic lumbosacral radiculoplexus neuropathy? To be able to give the patient sensible advice on the potential benefits of immune-modulating therapy, a clearer picture of the natural history of this condition is required. Differential Diagnosis. When faced with a diabetic patient who is losing weight and has progressive weakness of one or more of the lower limbs, an infiltrating pelvic malignancy has to be considered, and in most cases some imaging is required of the lumbosacral spine and pelvis (Tables 85–10 and 85–11). Malignant infiltration of the lumbosacral plexus tends to run a longer course, and the symptom of pain is often not prominent. If the possibility of malignancy is being considered, three or more cytospin

1964

Diseases of the Peripheral Nervous System

Table 85–10. Diabetic Lumbosacral Radiculoplexus Neuropathy: Differential Diagnoses • • • •

Lumbosacral nerve root compression Malignant infiltration of the pelvis Chronic inflammatory demyelinating polyneuropathy Systemic necrotizing vasculitis

examinations of the CSF are essential. Nerve injury following radiotherapy can occur at a variable time after initial treatment (from 4 weeks to about 30 years), and the main clinical features are that the time course is slow, with little in the way of pain, and that often both legs are involved. A retroperitoneal bleed produces an acute syndrome. Other diagnostic possibilities include CIDP (but the presence of prominent pain would be unusual) and a systemic vasculopathy, which should be associated with elevated inflammatory markers (erythrocyte sedimentation rate and C-reactive protein). Compression Neuropathies Diabetes mellitus and compression neuropathies are common problems, and therefore the possible chance occurrence of these two conditions together has to be considered. There are no good population-based studies that have looked at this possible association. Risk factors identified for carpal tunnel syndrome include diabetes, along with rheumatoid arthritis, pregnancy, Colles’ fracture, and manual labor occupation.306 There is therefore a general assumption that compression of the median nerve in the carpal tunnel occurs more commonly in patients with diabetes. The hospital-based study of

Table 85–11. Investigations To Be Considered in the Differential Diagnosis of Diabetic Lumbosacral Radiculoplexus Neuropathy Clinical Feature

Possible Cause

Investigation

Prolonged course ⫾ minimal thigh pain Acute onset

Pelvic malignancy

MRI/CT of pelvis CSF CT of pelvis

No pain Pain on movement

Retroperitoneal bleed Necrotizing vasculitis CIDP Prolapsed lumbar disc Lumbar canal stenosis Diseases of the hip

478 subjects by Perkins and colleagues244 showed that carpal tunnel syndrome was found in 14% of diabetics without neuropathy and 30% of diabetics with neuropathy, compared with a prevalence of only 2% in control subjects. The symptoms are similar in diabetic and nondiabetic patients. There is little information on the incidence of ulnar nerve compression at the elbow and common peroneal neuropathy at the neck of the fibula in the diabetic population. Investigation. With an isolated median or ulnar neuropathy, no specific investigations are required unless the physician and patient agree that decompression is indicated, at which stage nerve conduction studies should be requested. If the clinical picture is different (see later), further investigations to be considered are outlined in Table 85–12. Treatment. Decisions on treatment should be the same as for patients without diabetes. Carpal tunnel decompression should be considered for troublesome sensory symptoms that are not helped by wearing wrist splints, or if they recur after steroid injection into the carpal tunnel or there is weakness of the abductor pollicis brevis. The operation can clearly help, but perhaps not unexpectedly, the overall outcome may not be as good as in nondiabetic patients.232 With subacute or acute ulnar and common peroneal nerve palsies, one can justifiably wait for 12 to 16 weeks to see if there is some spontaneous recovery before considering surgical decompression. Based on evidence extrapolated from carpal tunnel surgery, it would seem reasonable to forewarn the patient that the results of surgery for these other mononeuropathies may not be as successful as in nondiabetic patients.

Table 85–12. Investigations To Be Considered for Recurrent or Multiple Mononeuropathies in a Diabetic Patient Clinical Feature

Possible Diagnosis

Recurrent

HNPP

Blood, nerve biopsy Neurophysiology MRI of lower spine

Multiple

Vasculitis

X-ray/MRI of hip

ACE ⫽ angiotensin-converting enzyme; ANA ⫽ antinuclear antigen; ANCA ⫽ anticytoplasmic antigen; CRP ⫽ C-reactive protein; CXR ⫽ chest radiograph; ESR ⫽ erythrocyte sedimentation rate; HIV ⫽ human immunodeficiency virus; HNPP ⫽ hereditary liability to pressure palsies; NCS ⫽ nerve conduction studies; PMP ⫽ peripheral myelin protein; RhF ⫽ rhesus factor.

CIDP ⫽ chronic inflammatory demyelinating neuropathy; CSF ⫽ cerebrospinal fluid; CT ⫽ computed tomography; MRI ⫽ magnetic resonance imaging.

Investigation PMP22 deletion NCS ESR, CRP, eosinophil count, ANA, RhF, c-ANCA, cryoglobulins, ACE, Lyme serology, HIV, hepatitis B serology, CXR, NCS, nerve biopsy

1965

Diabetic Neuropathies

Differential Diagnosis. When the clinical pattern is different from that expected, further investigations are indicated (see Table 85-12). If there is a history of recurrent but reversible mononeuropathies, particularly involving the lower limbs, then hereditary liability to pressure palsies has to be considered. Involvement of two or more peripheral nerves (multiple mononeuropathy) is very uncommon in diabetic patients, and such a pattern raises the strong possibility of a systemic or tissue-specific vasculitis.

Hypoglycemic Neuropathy The literature on humans developing a neuropathy related to hypoglycemia is small and mainly related to the presence of an insulinoma,138,154,325 when there are other clinical features of a neuropsychiatric nature.209 Although the neuropathy is usually sensorimotor, the motor component predominates, and the weakness can be worse in the arms than in the legs; wasting of the small hand muscles can be characteristic. Removal of the insulinoma can lead to recovery of the neuropathy. The mechanism that leads from hypoglycemia to neuropathy is not clear. Effects on axonal transport and microvascular changes have been proposed from experimental studies. Whether hypoglycemia occurring during the treatment of diabetes can lead to a neuropathy is not clear.

PATHOLOGIC CHANGES It has to be stressed at the outset that, in clinical practice, peripheral nerve biopsy has no role to play in the diagnosis of diabetic polyneuropathy, and should only be considered in the uncommon situation in which there is suspicion of an alternative or additional factor responsible for the neuropathy (e.g., vasculitis). If the decision is made that a nerve biopsy is justified, the procedure and the processing of the biopsied tissue should be done in a center with expertise in this area. At the end of the 19th century, the earliest reports on the pathology of diabetic neuropathy drew attention to degenerative changes in the peripheral nerves,89,252 and this was reaffirmed over 30 years later following the work of Woltman and Wilder.364 Ever since, the pathologic changes seen and their interpretation have sparked controversy. This is in part because it had been, and possibly still is, hoped that pathologic changes in peripheral nerves would provide the answer to the causation of diabetic neuropathy, but this remains elusive. Thus the proponents of a “metabolic” etiology followed one pathway, and along another pathway went the followers of a “vascular” etiology. After more than 100 years, the site of the primary disturbance in diabetic neuropathy is still under discussion.

The comments that follow relate to changes that have been mainly documented on microscopic examination of the sural nerve in patients with symmetrical diabetic polyneuropathy, and focus on three categories: 1. Histopathologic changes documented in (a) myelinated and unmyelinated nerve fibers, (b) blood vessels, and (c) connective tissues 2. Pathologic changes in the autonomic nervous system 3. Relationship of pathology to clinical diabetic neuropathy syndromes

Histopathologic Changes Myelinated Nerve Fibers It is clearly established that loss of axons occurs in the peripheral nerve trunks and this affects both myelinated and unmyelinated axons. This loss is most marked distally in the lower limbs. There is a clear gradient of myelinated fiber loss that is greatest at distal regions of the peripheral nerve, but even at the nerve root level some degree of fiber loss can be detected.84 Using teased fiber preparations, a variety of abnormalities including axonal degeneration, axonal regeneration, and segmental demyelination and remyelination have been found in diabetic polyneuropathy. None is diagnostic. Nerve fiber degeneration and regeneration is the predominant pathology seen (Fig. 85–3), and this has been interpreted as indicating that degeneration of axons is the primary event in diabetic neuropathy.86 Are the changes of demyelination all secondary to axonal degeneration, or does diabetes/hyperglycemia have a separate effect on both the axon and the Schwann cell/myelin sheath? From their teased-fiber studies in diabetic “small fiber” neuropathy, Said and co-workers275 found evidence of secondary demyelination occurring just proximal to axonal degeneration. The same authors also commented on the finding of occasional fibers showing primary demyelination. In these examples the processes of axonal degeneration and demyelination appear to be linked, but the prospect that this is the case in other diabetics with polyneuropathy is unlikely because no such correlation was found in other studies.22,183,368 Remyelination occurring after segmental demyelination has been reported183,341 and is recognized by an axon having a thinner myelin sheath than normal. Repeated episodes of segmental demyelination and remyelination are probably responsible for the appearance of “onion bulbs,” which are frequently seen20 but are not specific for diabetic neuropathy, and are most florid in appearance in inherited neuropathies (CharcotMarie-Tooth [CMT] disease) but also encountered in other acquired neuropathies. Prominent hypertrophic change in a biopsy from a diabetic patient with a progressive neuropathy would suggest that there was a superimposed CIDP or an inherited neuropathy (CMT type 1). During axonal (wallerian) degeneration, the Schwann cell basal

1966

Diseases of the Peripheral Nervous System

FIGURE 85–3 Electron micrograph of transverse section through a sural nerve biopsy from a patient with an acute, painful diabetic neuropathy, showing a loss of myelinated nerve fibers evidenced by a reduced fiber density and the presence of Büngner bands (Bb). Mf ⫽ myelinated nerve fiber. There is also prominent regenerative activity indicated by multiple regenerative clusters (cl). Bar: 2 ␮m. (From Llewelyn, J. G., Thomas, P. K., Fonseca, V., et al.: Acute painful diabetic neuropathy precipitated by strict glycaemic control. Acta Neuropathol. [Berl.] 72:157, 1986 with permission.)

lamina that surrounds the myelinated fiber normally persists, but as myelin and axon debris is removed the basal lamina collapses. King and co-workers171 showed that in diabetic neuropathy the basal laminal tubes have a characteristic appearance in that they remain circular after fiber degeneration (Fig. 85–4A), and they are also evident when the nerve fiber regenerates (Fig. 85–4B). The increased durability and “rigidness” of the basal laminae in diabetes may be due to increased collagen cross-linking as a result of nonenzymatic glycosylation. In sural nerve biopsies, myelinated fiber loss is not uniform,84,185 although exactly what this finding means is disputed. From their extensive postmortem study showing that nerve fiber loss was patchy even at very proximal levels, this suggested to Dyck and colleagues84 that ischemia was the cause of the neuropathy. This was also the conclusion reached in another postmortem study157 of diabetic patients with neuropathy and age-matched controls. Looking only at the sural nerve level, myelinated fiber loss was found to be nonuniform in younger diabetics with neuropathy, but the same pattern was also found in inherited neuropathy (CMT disease), in which a vascular etiology is unlikely.185 Whether this implies that ischemia only becomes a significant causative factor in the more elderly patient with diabetic polyneuropathy is unresolved. At the electron microscope level, there is clear loss of unmyelinated axons22 but no unique features associated with diabetes. The issue of whether there is selective loss of unmyelinated fibers in specific subtypes of diabetic neuropathy (“small fiber type”) is discussed later.

Blood Vessels The blood vessels that enter the endoneurium and those that supply the outer epineurium and perineurium are all abnormal in diabetic patients. The majority of studies have focused on the ultrastructure of endoneurial vessels, and the consistent changes include thickening or reduplication of the basal lamina (Fig. 85–5) around endothelial cells and pericytes343,344 and endothelial and pericyte degeneration.115,359 Capillary closure has been reported83,359 but not confirmed in subsequent studies.33,194,359 The work of Giannini and Dyck115 has also shown that the endoneurial microangiopathy is evident before the neuropathy appears, and they and other authors195,370 have demonstrated that the severity of microvascular change correlates with the severity of the polyneuropathy. Bradley and associates33 confirmed the previous observations and also that basal lamina thickening is seen in diabetic neuropathy and in other chronic neuropathies.22 In diabetic neuropathy, compared with CMT disease, the reduplicated basal lamina is thicker. Can these varied changes found in diabetic neuropathy be linked together? Gianinni and Dyck116 referred to “microvascular transformation,” a term that describes the evolving pathology in endoneurial vessels. The process begins with basal laminal reduplication and pericyte degeneration. This leads to thickening of the vessel wall and also separation of endothelial cells, which in turn causes intercellular gaps to appear. The end result is an alteration in the blood-nerve barrier, whose function has already been noted to be abnormal in diabetic neuropathy.250

Diabetic Neuropathies

A

1967

B FIGURE 85–4 A, The circular Schwann cell basal laminal tube without cellular contents suggests an abnormally stiff basal lamina (BL). Bar: 1 ␮m. B, In this regenerating cluster, the original basal laminal tube is still intact and circular. Bar: 1 ␮m.

Connective Tissues There is increased endoneurial collagenization in diabetic and other neuropathies34 mainly involving types I and III collagen. Types IV, V, and VI collagen are increased around endoneurial microvessels, and types IV and V are increased in the perineurium of diabetic patients. Several abnormalities have been noted in the perineurium. The basal laminae of perineurial cells are thickened35,171 and the outer layers of the perineurium contain an excess of calcium (hydroxyapatite) deposits.160,170 The structural changes of Schwann cell basal laminae were discussed earlier. What these connective tissue changes imply in terms of the pathogenesis of diabetic polyneuropathy is not clear. Are they simple reflections of a diabetic state, or is it possible that, for example, they do have a negative effect on nerve function and slow down nerve regeneration in diabetic neuropathy?

Autonomic Neuropathy Observations on the pathologic changes in the autonomic nervous system have not been extensive, and previous studies were limited because of the lack of comparison with control tissue. Postmortem examination of paravertebral

sympathetic ganglia from patients with diabetes showed changes that were similar to those related to ageing.278 Neurofilamentous aggregates in swollen terminal axons next to neuronal cell bodies, dystrophic axons, and aggregates of lymphocytes in the ganglionic tissue and perivascular regions represent the main findings. Neuroaxonal dystrophy also develops in dorsal root ganglia related to diabetes and aging, and although changes are more prominent in diabetic patients, there is no light or electron microscopic appearance or immunohistochemical change that is specific for diabetes.278,279 What is also of interest is that sympathetic ganglia are not affected equally, and further studies are required to try to correlate the pathologic changes with details of the severity of clinical autonomic neuropathy. From the information available, diabetic autonomic neuropathy is likely to be multifactorial in causation.

RELATIONSHIP OF PATHOLOGIC CHANGES TO CLINICAL SYNDROMES The variant syndromes that come under the umbrella of diabetic polyneuropathy include large fiber type, small fiber type, autonomic, painful, and ataxic syndromes. These

1968

Diseases of the Peripheral Nervous System

A

B FIGURE 85–5 Electron micrograph of transverse section through an endoneurial capillary in a diabetic (A) and a control (B) patient. Note reduplicated basal laminae surrounding the endothelial cells. Bar: 5 ␮m.

probably are not completely separate clinicopathologic entities, but rather represent the extreme ends of a continuum. Earlier studies that looked at the morphologic correlates of pain in diabetic polyneuropathy compared changes in the peripheral nerve between diabetic and nondiabetic controls, but when comparison is made between painful and nonpainful diabetic neuropathy, neither the amount of axonal degeneration nor regeneration explains the occurrence of pain.39,40,183 Said and colleagues273 found a severe loss of myelinated and unmyelinated fiber in patients with acute sensory and autonomic neuropathy similar to that in acute-onset diabetic neuropathy with profound muscle weakness, wasting, and distal sensory loss but only minimal autonomic abnormalities. For asymmetrical or multifocal diabetic neuropathies, which tend to affect more elderly patients, the clinicopathologic correlation is clearer. Although there is loss of both myelinated and unmyelinated fibers, the pattern is more patchy, in keeping with the clinical and neurophysiologic findings. There is also evidence of precapillary vessel involvement with perivascular mononuclear cell infiltration (Table 85–13).274

PATHOGENESIS AND CLINICAL TRIALS The diabetic neuropathies are made up of a collection of diverse clinical syndromes that can, from a pathogenetic viewpoint, be divided into two groups—those neuropathies that are symmetrical and those that are asymmetrical. The most common is DSDP, which accounts for over half of all diabetic neuropathies, and this will also usually be present in patients who present with asymmetrical or multifocal diabetic neuropathy variants. There is probably no single pathogenetic mechanism, and multiple factors are likely to be interacting to provide the variable clinical picture that is

Table 85–13. Main Pathologic Features in Diabetic Polyneuropathy • • • • •

Loss of myelinated and unmyelinated fibers Myelinated nerve fiber loss may be diffuse or patchy Clusters of regenerating fibers may be abundant Thickening of endoneurial blood vessels Increased durability and rigidity of Schwann cell basal laminae

1969

Diabetic Neuropathies

produced. Evidence points more clearly to the asymmetrical diabetic neuropathies having a predominantly vascular/ immune etiology. A number of possible causative factors remain under scrutiny in relation to symmetrical polyneuropathy, and it seems inescapable that these potential pathogenetic mechanisms will be interlinked. It is no longer helpful to consider the pathogenesis of diabetic polyneuropathy in terms of a “metabolic hypothesis” versus a “vascular hypothesis.” It is to studies on experimental models of diabetic neuropathy that one has to turn to understand the evidence that supports the inclusion of these factors, and this is outlined in detail in the subsequent section on animal models. In this section the focus is on studies carried out in human diabetic neuropathy.

Hyperglycemia Hyperglycemia is probably the most important initiator of the process that leads to neuropathy. A number of shortterm studies have documented improvement in tests of peripheral nerve function (mainly nerve conduction studies) following restoration of normal blood glucose levels.128,316,350 Some of these improvements occur acutely and must be related to functional rather than structural nerve change. Even in intermediate-length studies such as that of Ziegler and co-workers,379 in which continuous subcutaneous insulin infusion treatment was used to “normalize” blood glucose levels in poorly controlled diabetic patients, improvements in nerve conduction velocities were related to the level of glycemic control (HbA1 level) in some nerves but not in others, and this again raises uncertainties as to whether neurophysiologic changes are of any relevance to structural neuropathy. There are no large randomized trials looking specifically at the effect of tight glucose control on the development or progression of diabetic polyneuropathy, but information has been collected from long-term trials in which certain aspects of neuropathy were included for analysis. For type 1 diabetes, the DCCT,75 the Stockholm Diabetes Intervention Study,256 and the Oslo Study9 all showed that tight glucose control does slow the deterioration in nerve conduction studies in those patients with established neuropathy, and it also delays the development of both somatic polyneuropathy and autonomic neuropathy, but it does not prevent the appearance of neuropathy. The same does not appear so clearly to be the case for the more common type 2 diabetic patient. The UKPDS338 showed that the effect of intensive diabetic treatment on neuropathy in those patients with type 2 diabetes was minimal. The Kumamoto Study230 (insulin-treated type 2 diabetes) did show that, in the upper limbs, the deterioration in nerve conduction velocity (NCV) seen with conventional treatment was significantly worse in comparison with patients receiving intensive treatment. Lower limb NCV was not recorded. A follow-up to this study showed that the onset

of clinical neuropathy was delayed by a mean of 2.2 years in those patients who had received intensive insulin treatment.349 Intensive insulin treatment for 2 years had no effect on neuropathy in the Veterans Administration Cooperative Study.15 Severe hypoglycemia is a significant complication of intensive diabetic treatment and may of course damage nerve fibers. Pancreatic transplantation is one way, albeit not commonly used, of achieving normoglycemia without risk of hypoglycemia, and studies have consistently shown that, if normoglycemia can be maintained for 2 to 3 years, nerve conduction study abnormalities show no further worsening and could thereafter show some improvement.5,165,217 The effects on clinical symptoms and signs, however, are more variable. Other factors—some metabolic, genetic, and vascular, for example—must also have a role in diabetic neuropathy, because only 50% of patients with diabetes eventually develop neuropathy.249 What is yet to be established is whether the metabolic abnormalities have a direct injurious effect on peripheral nerve axons or Schwann cells or whether these structures are damaged mainly by ischemia produced by microvessel dysfunction arising from metabolic abnormalities.

Polyol Pathway and Aldose Reductase Inhibitors The polyol pathway hypothesis links the fact that hyperglycemia induces increased flux through the polyol pathway—via the enzyme aldose reductase (AR)—resulting in accumulation of sorbitol in nerve, and this subsequently leads to a reduction in nerve myo-inositol content and Na⫹, K⫹-ATPase activity, leading to reduced nerve conduction velocity. This has largely been based on observations in animals and is discussed in more detail later. The evidence to date is that the pivotal reduction in myo-inositol has not been observed in peripheral nerve tissue from diabetic patients with polyneuropathy even though sorbitol content is increased.87,133 The accumulation of experimental evidence supporting a role for the AR (polyol) pathway in diabetic neuropathy and the beneficial effects of aldose reductase inhibitors (ARIs) has led to a number of clinical trials using a variety of ARIs that now span a period of 20 years, but overall the results have been disappointing. The only consistent end point has been NCV, and a recent Cochrane Review4 concluded that, although ARIs could diminish the reduction in motor nerve conduction velocity (MNCV) seen in diabetic neuropathy, no effect was seen in sensory nerves. Greene and colleagues123 showed that treatment with the ARI zenarestat for 12 months reversed the loss of myelinated nerve fibers in the sural nerve, possibly by stimulating regeneration, because the main increase seen was in the density of small myelinated fibers. It was also evident that greater than 80% AR inhibition was required to produce these changes, but toxic side

1970

Diseases of the Peripheral Nervous System

effects limited further studies. This has been a feature with other ARIs. In Japan, studies of the ARIs epalrestat142 and more recently fidarestat,143 administered over a maximum period of 12 months, showed a significant improvement in a range of subjective symptoms compared with placebo. The findings on clinical examination were not assessed. If the issues of getting adequate AR inhibition together with an acceptable safety profile can be tackled, and by setting clinically meaningful end points for longer term trials, an ARI may yet emerge to play a role in the treatment of diabetic neuropathy. Polymorphisms in the promoter region of the gene for AR have been associated with risk of developing complications of diabetes, including neuropathy,73 and identification of the susceptibility genes may greatly assist the design of further trials.

Advanced Glycation End Products One other consequence of hyperglycemia is the formation of advanced glycation end products (AGEs) that have been shown to accumulate in diabetic neuropathy.269 Whether this damages nerve cells or alters connective tissue leading to disruption of neuronal metabolism, axonal transport mechanisms, or possibly impairment of nerve regeneration is not known. AGE inhibitors have not been tested in any clinical trial of diabetic neuropathy.

Essential Fatty Acids Essential fatty acid synthesis is reduced in experimental diabetes, with a negative effect on endoneurial vasculature. Clinical trials of ␥-linolenic acid (as evening primrose oil)31,151,163 have shown modest improvements in NCV, but results of the largest study31 have appeared only in abstract form.

Neurotrophic Factors Neurotrophic factors have for some time been the focus of interest in diabetic neuropathy, not least because nerve growth factor (NGF) is required to maintain dorsal root ganglion cells and sympathetic neurons, and both are affected in diabetic neuropathy. Serum levels of NGF are reduced in diabetic patients with neuropathy, perhaps related to crossreacting insulin antibodies.98 The experimental data supporting a role for NGF in diabetic neuropathy generated Phase II clinical trials that seemed to show a positive effect on symptoms, but disappointingly, a larger scale Phase III study was unable to reproduce earlier benefits.11

Nerve Ischemia/Hypoxia Microangiopathy leading to hypoxia is considered to be an important factor in the pathogenesis of diabetic polyneuropathy. This is based, in human studies, on evidence from

pathologic studies of peripheral nerves showing multifocal lesions at autopsy and a multifocal pattern of nerve fiber loss even at the sural nerve level, suggesting ischemia. Ultrastructural changes in the vasa nervorum (outlined earlier) have added to the weight of evidence. Interpretation of these data must be cautious, because some observations have been made in older patients, and the pathologic changes in the pattern of nerve fiber loss and microvessel ultrastructure have also been observed in neuropathies that are not of an ischemic/hypoxic origin. Newrick and colleagues221 found the sural nerve of patients with diabetic neuropathy to be hypoxic, although comparison was made with a younger age normal control group and patients with other neuropathies were not included. Blood flow in sural nerve of patients with mild diabetic neuropathy is normal but is reduced in those with more severe polyneuropathy, and in diabetics with the BrunsGarland syndrome.320 Ibrahim and co-workers147 used a spectrophotometric technique to measure nerve oxygen saturation and the rate of nerve fluorescence after an intravenous injection of fluorescein as a marker of nerve blood flow, and found both to be reduced in patients with mild to moderate diabetic neuropathy compared with diabetic patients without neuropathy and normal control subjects. The angiotensin-converting enzyme (ACE) inhibitor trandolapril did improve neurophysiologic parameters after 12 months of treatment in type 1 diabetic patients with mild neuropathy and normal blood pressure. Vibration perception threshold, autonomic function, and neuropathy symptom and deficit scores did not change.196

Oxidative Stress An excess of free radicals is associated with oxidative stress and, in Germany, use of the antioxidant ␣-lipoic acid (thioctic acid) in clinical trials has consistently shown benefit on neuropathic symptoms and clinical signs. The safety profile appears to be good (see Ziegler and colleagues384 for review). A longer term North American/European study (SYDNEY trial) looking at total symptom score as the primary end point found that 600 mg/day of ␣-lipoic acid given intravenously for 5 days and repeated over a period of 14 weeks significantly improved neuropathic symptoms, including pain.8

TREATMENT OF NEUROPATHIC COMPLICATIONS Maintaining the best possible level of glucose control is the cornerstone of management from the time of diagnosis of diabetes. This has clearly been demonstrated in type 1 diabetes, with a 69% reduction in the risk of developing polyneuropathy after 5 years. The same has not yet been demonstrated in type 2 diabetes. Even obtaining the tightest

1971

Diabetic Neuropathies

level of glucose control by standard intensive insulin treatment regimens does not allow complete prevention of the development of neuropathy, nor does it stop the slow and irreversible progression of the neuropathy once it has developed. Other than for perhaps the acute painful neuropathy variant, there is no evidence that attaining near normoglycemia results in any clinically significant improvements that the patient will notice.

TREATMENT OF PAINFUL DIABETIC NEUROPATHY The treatment of pain associated with diabetic neuropathy poses one of the most difficult aspects of management because there is no single effective treatment. When pain is persistent, secondary depression may become a complicating factor, making the assessment of pain even more difficult. The list of preparations that can be tried is ever increasing, but often backed only by evidence from short-term clinical trials, and many drugs will not have a specific license to be used in painful diabetic neuropathy. For example, at this present time in the United Kingdom the only drug with a specific license for use in neuropathic pain is gabapentin. With such a variety of drugs available, it is important to have a list of preferred choices, so that both the physician and patient have a plan of action to follow. This plan is more often than not dictated by personal experience, and the goal is to try to improve the patient’s symptoms without causing side effects. The aim should be for monotherapy, but in reality this is often not attainable. Patients may have already tried simple analgesics (paracetamol or nonsteroidal anti-inflammatory drugs), and surprisingly there are few clinical trial data on dosage or efficacy of these first-line treatments for painful diabetic neuropathy. The options thereafter are outlined in Table 85–14. Anticonvulsants Versus Tricyclic Antidepressants Tricyclic antidepressants are still considered by many as first-line agents for neuropathic pain, although their role can be limited by troublesome side effects. The anticonvulsant gabapentin now probably shares pole position as the first option, although there are few long-term studies and no clear guidelines on the starting dosage, rate of increase, or optimal final dosage. A review of the results of randomized control trials suggested a starting dose of 300 mg/day, increasing over 3 days to 900 mg/day initially and then gradually up to 1800 mg/day in two divided doses, and again up to a total daily dose of 3600 mg if needed.16 What is clear is that the titration of the drug has to be individualized, and this may lead the clinician to decide in certain cases (e.g., an older patient) on a lower starting dose and smaller, more widely spaced increments. There is no difference between

Table 85–14. Treatment Options in Painful Diabetic Polyneuropathy Treatment Generalized pain (i) simple analgesics: paracetamol, NSAID (ii) Gabapentin Carbamazepine Amitriptyline Nortriptyline Imipramine Desipramine (iii) Mexiletine, SSRI (Citalopram, Paroxetine), other anticonvulsants (Lamotrigine, Topiramate, Oxcarbazepine) (iv) ␣-Lipoic acid

Localized pain Capsaicin ointment (0.075%)

Dosage

600–3600 mg/day 100–800 mg/day 10–150 mg/day 10–150 mg/day 10–150 mg/day 10–150 mg/day

600 mg IV/day for 5 days/week over 14 weeks Applied 4 times daily; maximum 8 weeks’ duration

Alternatives OpSite dressing Acupuncture Pain psychologist/counselor IV ⫽ intravenous; NSAID ⫽ nonsteroidal anti-inflammatory drug; SSRI ⫽ selective serotonin reuptake inhibitor.

gabapentin and the tricyclic antidepressant amitriptyline in terms of efficacy in providing pain relief.212,215 Amitriptyline is usually taken at night, and the starting dose can be as low as 10 mg and increased as required up to 75 to 100 mg daily. A review of published studies noted that the side effect profile of gabapentin is better than that of amitriptyline.60 Amitriptyline, imipramine, and desipramine are all better than placebo in the treatment of painful diabetic neuropathy, but there is no information on whether one preparation is better than the others. There are no placebo-controlled studies of nortriptyline. There is no information on the benefits of using gabapentin with a tricyclic antidepressant, and yet this is a combination that many patients end up being prescribed. Alternative Therapies Evidence is being accumulated on other antiepileptic drugs such as lamotrigine, oxcarbazepine, topiramate, and levetiracetam.17,235 The antiarrhythmic mexiletine is an alternative option in patients who have failed to respond to or are unable to tolerate other drugs.153 Venlafaxine, a serotonin and weak noradrenaline reuptake inhibitor, helps reduce pain and may be as effective as imipramine.295 Data from one short trial on the use of the selective serotonin reuptake inhibitor citalopram showed

1972

Diseases of the Peripheral Nervous System

improved control of pain.296 Increasing clinical trial data support the beneficial effects of the antioxidant ␣-lipoic acid (see earlier). Administered by intravenous infusion (600 mg/day for 5 days over 14 weeks), it improves pain.8 Shorter term studies of oral treatment (1800 mg/day) also suggest efficacy.381 For refractory cases, long-term opioid treatment may be the only option available. Capsaicin ointment (0.075%) depletes the skin of substance P and reduces pain,319 but should not be used for longer than 8 weeks because more prolonged treatment can lead to increased insensitivity of the skin. A doubleblind trial found topical capsaicin to be equally effective as oral amitriptyline, and it had the advantage of fewer side effects.25 Results of double-blind trials of topical capsaicin must be viewed with some caution because in about 50% of patients it will produce a transient area of discomfort at the site of application. Another topical treatment is the use of OpSite surgical dressing (“plastic skin”). When applied to the most painful areas of the feet and lower legs, it reduces pain when compared with a control group receiving no treatment.109

The “At-Risk” Diabetic Foot An important part of the assessment of any patient with neuropathy is inspection of the feet. This is particularly so for patients with diabetes because 15% will develop foot ulceration at some time,7 and one of the main risk factors for developing this complication is the presence of neuropathy (Table 85–15). Ischemia is the other main problem that places the diabetic foot “at risk.”351 In a diabetic patient with polyneuropathy, identification of the at-risk foot is important. A recent community-based study in the north of England looking at 9710 diabetic patients over a 2-year period showed the annual incidence of foot ulcers to be 2.2% and found the main risk factors for developing new foot ulcers to be abnormal neuropathy disability score, presence of an existing ulcer, history of previous ulcer, reduced foot pulses, and foot deformity.1 The presence of peripheral edema is another feature considered by some to be an important risk factor.88 There are no tests—only careful inspection and examination of the feet. Does the skin appear healthy or are there signs of thinning, and perhaps cracking of excessively dry skin? Table 85–15. Risk Factors for Foot Ulcers • Polyneuropathy (somatic ⫾ autonomic) • Presence of an existing ulcer or history of previous ulcer • Altered peripheral circulation—ischemia or abnormally high blood flow • Foot deformity (clawing of toes, wasting of small foot muscle, high arches or flat feet, Charcot joint) • Callus (in areas of increased pressure, leads to ulceration) • Peripheral edema

Discoloration may suggest ischemia, but distended veins and “warm” skin suggest increased blood flow as a result of autonomic dysfunction. There may be trophic changes evident. Any foot or toe deformity needs to be noted because shoes may become ill-fitting, causing skin abrasion. Checking between the toes for “hidden” ulcers must not be forgotten. Callus arises as a result of repeated friction in a pressure point area where there is loss of sensation and dry skin. Once identified, it has to be dealt with promptly with referral to the podiatry service. A warm and swollen foot is likely to indicate a Charcot joint. This complication can arise in up to 15% of diabetics with neuropathy292 and usually affects the tarsometatarsal region but can also occur in metatarsophalangeal joints and at the ankle. There is often a history of preceding injury to the foot, which can be minor, and this seems to start off the process of bone resorption. Although the feet are numb, patients often complain of an aching bone pain. Early diagnosis is essential and management requires specialist advice from the diabetic foot care team. Avoiding weight bearing on the affected foot is the mainstay of treatment, with bedrest, non–weight-bearing crutches, or total distal leg contact casting. Early on, plain radiographs can be normal, but an isotope bone scan will show areas of high blood flow, which leads to a phase of bone destruction lasting weeks or months. MRI of the foot should be done in addition to a gallium-labeled white blood cell scan, to pick up areas of injury and infection.

ANIMAL MODELS OF CAUSATION Many sources review the spectrum of animal models of diabetes mellitus, and the reader is referred elsewhere for comprehensive accounts.287 Studies on defects of nerve structure and function in diabetic animals have predominantly used rodent models, with the rat as the favorite species. Many studies use chemically induced diabetes, with the older agent alloxan supplanted in modern times by streptozotocin (STZ). Again, the reader will find extensive sources for information on these diabetogenic drugs.255,257 In the past decade the spontaneously diabetic BB rat has been used by some groups for studies on experimental neuropathy. It has been argued that these animals show ultrastructural pathology of the nodes of Ranvier.293 There has also been the sustained argument that diabetes of genetic origin ought to represent higher fidelity modeling than that induced by a chemical agent. However, no in vivo neurotoxic effects of diabetogenic doses of STZ have been reported. In contrast, the BB rat is known to develop a genetic thyrotoxicosis in addition to diabetes.347 Given that the rat is so far removed from humans in many respects, it seems facile to quibble over particular characteristics of different forms of insulin deficiency in the same rodent species.

1973

Diabetic Neuropathies

Diabetes can be induced in other species with either alloxan or STZ, and a few studies have used chemically diabetic rabbits and mice. With the latter species, the genetically diabetic strain C57 BL/Ks (db:db) has also been used. These animals are insulin resistant rather than insulin deficient, and arguably reflect features of non–insulin-dependent diabetes mellitus. No model is perfect. The cheapness and relative ease of use of rodents have guaranteed their overwhelming adoption for studies on nerve dysfunction and structure. The style of modeling and experimentation has evolved into testing of hypotheses focused on particular biochemical, physiologic, or morphologic, anomalies. The hypotheses surrounding myo-inositol, which are discussed in this chapter, provide an example of this. The natural progression for studies that concentrate on particular abnormalities in a diabetic model is then to examine its relevance by testing for the presence of the anomaly in human nerve. That done, it is possible to return to the model and explore the consequences of the validated defect. Alternatively, lack of validation may prompt examination of other models. This is discussed in detail elsewhere in relation to myo-inositol in the nerves of diabetic mice, rats, and humans.144,326 There are studies in the literature describing the induction and management of diabetes in the two preferred large-animal generic models—pigs or minipigs and monkeys.110,118,129,182,246,310,360 However, studies on neuropathies in these models highlight the problems rather than the opportunities and do not indicate improved fidelity with respect to humans,371 though there are exceptions to this assertion.27,62,63,204

Glucose Toxicity: The First Steps of Causation The studies of Jean Pirart gave the first definitive evidence that poor control of diabetes predisposed patients to neuropathy.247,248 Subsequently, the DCCT broadened this observation and gave more detail.74 This definitive association between poor glycemic control and raised incidence and severity of diabetic neuropathies has vindicated hypotheses based on glucose as a primary causative factor. This does not preclude involvement of other aspects of poorly controlled diabetes, and it is likely, for example, that hyperlipidemia, via vascular disease, is an accelerator of neuropathy in type 2 diabetes.338 However, the search for causation must follow the trail back to the real source of disorder, and interest centers on the next level of anomaly—those biochemical disorders that are the secondary consequence of hyperglycemia. Three of these have received attention: (1) exaggerated flux through the polyol pathway; (2) nonenzymatic glycation of macromolecules, especially proteins; and (3) increased oxidative stress. Each has the capacity to initiate functional and ultimately structural changes, but they are interlinked and can synergize.

The possible relationship between primary hyperglycemia, secondary biochemical events, disturbed cell biology, and the final stages of the disease is represented in Figure 85–6. Thus a systematic account should focus on these steps.

Biochemical Consequences of Raised Intracellular Glucose The Polyol Pathway This is a highly conserved pathway that uses glucose as a substrate and forms sorbitol, an electrically neutral osmolyte that does not cross cell membranes readily, enabling cells to compensate for interstitial fluid hypertonicity via accumulation of sorbitol. The pathway comprises conversion of glucose to sorbitol by AR with oxidation of NADPH. Sorbitol is then converted to fructose by sorbitol dehydrogenase with reduction of NAD. Exposure of cells to hypertonic extracellular fluid provokes an increase in expression of AR and increased production of sorbitol.19 Other mechanisms also participate, and these use a range of osmolytes including myo-inositol, taurine, and betaine,18 such that failure of one of these systems can be covered by compensatory upregulation of others. In nerve in diabetes, the pathway is driven primarily by raised substrate rather than osmotic stress. There is coincidence between insulin-independent glucose uptake and susceptibility to complications,78 and peripheral nerve is firmly in this category. Under these conditions, the cell is forced to produce sorbitol without extracellular hypertonic stress and the generation of such an intracellular osmolyte provokes an inappropriate osmotic response. This might be termed intracellular osmotic stress. That this occurs is demonstrated by the fact that tissues that accumulate sorbitol in diabetes reduce their content of myo-inositol, by reducing expression of its carrier protein.43,44 This must compensate for the intracellular stress imposed by the accumulated sorbitol, but it is not possible to judge the efficacy of this compensation. The use of inhibitors of the enzyme enables a pragmatic identification of its involvement in the signs associated with diabetic neuropathy. The clearest of these is the prevention of the short-term conduction velocity deficit in diabetic rats by treatment with an ARI; since the first demonstration,329 this has become a sine qua non for pharmacologic efficacy of an ARI (see Tomlinson327 for review). No other neurophysiologic end points have emerged for ARIs with a similar degree of uniformity among investigators. Clinically, findings have been disappointing, but the development of neuropathy-associated conduction disorders is very slow in humans, necessitating trials of a duration that are well outside the norms usually practiced by pharmaceutical companies. Furthermore, in humans, the more meticulous control of dosage has limited the degree of inhibition available with some compounds, so that genuine pharmacologic efficacy against the enzyme in situ is in

1974

Diseases of the Peripheral Nervous System

FIGURE 85–6 Postulated cause-to-effect hierarchy linking diabetes-derived biochemical disturbance to end-stage diabetic neuropathy. Hyperglycemia leads to raised intracellular glucose in cells (neurons, Schwann cells) that are capable of insulin-independent glucose uptake. This in turn leads to secondary biochemical disturbances (polyol pathway, nonenzymatic glycation of macromolecules, and oxidative stress), which affect cellular processes. Initially these changes are reversible—on initiation of good glycemic control, for example—but they progressively alter cellular phenotype, establishing irreversible changes. Critical to this is activation of mitogen-activated protein (MAP) kinases, which can alter neuronal gene expression and/or promote Schwann cell death. An alteration of cell-to-cell mediators also compromises cell physiology. Collectively, these changes lead to structural and functional breakdown.

doubt for some trials. All in all, although no ARI has been sufficiently impressive to gain a license and remain on the market in Europe or North America, these drugs are by no means a lost cause. We await the inhibitor with a more effective balance of pharmacologic efficacy, tissue penetration, and low toxicity. A detailed up-to-date appraisal of the properties and potential of ARIs in treatment of diabetic neuropathy can be found elsewhere.227 Nonenzymatic Glycation This refers specifically to the nonenzymatic reaction of glucose, ␣-oxoaldehydes, and other saccharide derivatives with proteins, nucleotides, and lipids. Two major types of reaction are illustrated in Figure 85–7. The first to be described was the Maillard reaction, an initially reversible formation of a Schiff base, which spontaneously undergoes the Amadori rearrangement. In the example shown, involving combination of glucose with a lysine group, the fructosamine so formed is fructose

lysine. These can then cyclize in various ways to form AGEs, such as carboxymethyl lysine, pyrroline, or pentosidine (formed by cross-linkage of Arg-Lys). Reactive ␣-oxoaldehydes formed from glucose (glyoxal, methylglyoxal, and 3-deoxyglucosone) are also important precursors of AGEs in vivo (see Fig. 85–7). Early glycation of peripheral and central nervous system proteins of rats with alloxan-induced diabetes was described 20 years ago.346 The presence of AGEs in peripheral nerve tissue (sural, peroneal, and saphenous nerve) of human diabetic subjects was described by Sugimoto and associates.311 Protein glycation can occur both intracellularly and extracellularly. Its importance in the development of diabetes complications is assumed rather than proven, on the basis of prevalence. This area of investigation has been disadvantaged by lack of specific inhibitors of the process (see later). Because the formation of AGEs is a relatively slow process, its impact on the functional effects of short-lived intracellular proteins may be limited, but we know relatively little

Diabetic Neuropathies

1975

FIGURE 85–7 Schematic showing interactions between intracellular and extracellular processes in the elaboration of diabetic neuropathies. Hyperglycemia causes secondary biochemical changes (polyol pathway, nonenzymatic glycation of macromolecules, and oxidative stress) and, either independently or via oxidative stress, compromises neurotrophic support. This interacts with the biochemical changes in neurons to degrade neuronal phenotype and, in Schwann cells, to cause death. Independently, vascular changes cause ischemic hypoxia, possibly via activation of protein kinase C, subtype ␤II. Compromise of the axonal endoskeleton, via both hyperphosphorylation of material in situ and impaired synthesis of cytoskeletal proteins, interferes with axonal transport and the capacity of the neuron to maintain its projections, resulting in distal axonopathy.

about this. In neurons, the proteins forming the axonal endoskeleton are long lived, and these are glycated in diabetes with consequent effects on assembly of the cytoskeleton.68,201 This process may participate in the shrinkage of the axonal cylinder with long duration of diabetes. Glycation of long-lived extracellular proteins is more easily studied. These and their effects may be divided into two generic types—the glycation of static long-lived proteins forming extracellular matrix and connective tissues and the glycation of circulating proteins. Glycation of matrix proteins alters their functions by changing their physicochemical properties and their interactions with other molecules. The clearest involvement of this process in diabetes complications is in diabetic nephropathy, wherein glycation of basal lamina alters filtration properties and stimulates cellular proliferation.130,139,245 There is

no direct demonstration of analogous events in diabetic neuropathy, but communication between the axonal endoskeleton and extracellular matrix is critical for neuronal plasticity and the maintenance of the axonal extremities. Understanding of these processes is limited, but their disturbance via glycation is a safe prediction and a useful potential research topic. The discovery that cellular receptors exist for the putative internalization and clearance of circulating AGEs opened up new possible mechanisms by which this process can impinge on the development of diabetes complications. The receptor for advanced glycosylation end products (RAGE) is a member of the immunoglobulin superfamily of proteins.218,277 Two forms of RAGE with masses of 35 and 45 kDa are detectable by immunoblotting in most tissues. RAGE is thought to exist in a complex

1976

Diseases of the Peripheral Nervous System

with lactoferrin-like protein. In the central nervous system, RAGE also bound to amphoterin, a DNA-binding protein that is strongly expressed in developing neurons and mediates neurite outgrowth. Binding of AGE-BSA to RAGE induces oxidative stress, as indicated by increased thiobarbituric acid–reactive substances and expression of hemoxygenase. This was linked to the Ras pathway leading to activation of nuclear factor-␬B.146,369 Thus understanding is evolving of mechanisms involving glycation, via RAGE activation and via shifting properties of matrix molecules, that may have fundamental effects on the capacity of neurons to maintain distant axonal processes. Research is hampered only by the absence of a specific inhibitor of the glycation process. Aminoguanidine was developed as a potential therapeutic agent to inhibit glycation in the development of diabetic complications. It intervenes in glycation reactions to scavenge ␣-oxoaldehydes, particularly methylglyoxal, glyoxal, 3-deoxyglucosone, and ␣-dicarbonyls in glycated proteins.322 However, it is also an inhibitor of inducible nitric oxide synthase206,323 and amine oxidase,375 and has antioxidant activity.117 These other effects may well contribute to the various reports in the literature of functional benefits from aminoguanidine treatment in diabetic rats. Not all reports describe benefits. Aminoguanidine did not affect the incidence or severity of neuroaxonal dystrophy, a feature of autonomic neuropathy, in STZ-diabetic rats.282 A study in STZ-diabetic baboons showed no evidence of improvement by aminoguanidine of declining NCV and autonomic dysfunction.27 The critical judgment of all this animal work must be that not a single report shows a correlation between functional improvement and a measured decrease in any parameter of glycation. Hence we have an agent that may not work and, even if it does, may not work by the method proposed by its proponents. Oxidative Stress The potential complexity of molecular oxidants as participants in the causes of diabetes complications was first mapped out by Halliwell in 1987.134 Since then there have been several hundred publications extending, clarifying, complicating, convoluting, and extrapolating these links with oxidative stress. In spite of this, we are not much closer to either defining the role of oxidative stress or using the link to develop useful therapeutics. There are several problems. First, markers of oxidative stress (see, e.g., Asayama and co-workers,13 Gutteridge and Halliwell,131 and Monboisse and colleagues213) tend to reflect molecular history rather than current events, and studies concentrating on plasma markers provide no useful information at all, because the source of oxidized molecules is unknown. Additionally, the presence of abnormal amounts of oxidative markers does not discriminate between increased production and impaired scavenging, and this difference is important when trying to understand causes and how to intervene. The other

approach, namely treatment with antioxidants to ameliorate dysfunction, gives only superficial understanding of the role of oxidative stress. It provides no information about the sources of oxidants and little or no understanding of their molecular targets. Thus useful information is limited. First, there is no doubt that diabetes is associated with increased production of oxidatively derived free radicals,145,210,362 and these have the capacity to damage intracellular molecules and contribute to the development of macrovascular disease. Indicators of diabetes-associated free radical injury are of two types: (1) accumulation of the lipid peroxidation products malondialdehyde191 and 4-hydroxyalkenals228 and conjugated dienes,168,187 and (2) disruption of antioxidant defense mechanisms. Many of these processes are directly linked to the polyol pathway372 and/or protein glycation,363 and oxidative stress further enhances the latter process.277 A clear example of such a link would be increased consumption of NADPH by aldose reductase (the first enzyme of the polyol pathway) compromises glutathione recycling and impairs clearance of hydrogen peroxide to water. This has the potential to increase production of the damaging hydroxyl radical via the Fenten reaction. It appears, therefore, that agents that reduce oxidative stress, interfere with the production of free radicals, scavenge those free radicals, or interfere with their capacity to produce tissue damage, have therapeutic potential. It is probable that ␣-lipoic acid (thioctic acid), a naturally occurring agent available by prescription in Germany, acts by reducing oxidative stress and its outcomes.314 In clinical trials, this agent has demonstrated as much efficacy as any other.382,384

Coupling of Disordered Biochemistry to Dysfunction Mitogen-Activated Protein Kinases Manifestations of diabetes-induced nerve dysfunction are initially reversible, but then develop into a more permanent state, associated with clear structural changes. This sort of transition is a feature of the other diabetes complications, and the shift from a metabolic to a structural basis is the critical zone between feasible and unfruitful drug targets. In terms of cell biology, this zone represents the processes by which the metabolic changes discussed earlier (and possibly other unidentified metabolic consequences of hyperglycemia) cause a critical change in phenotype in neurons, Schwann cells, and other elements of the peripheral nervous system. This requires links between the metabolic events downstream of hyperglycemia and gene transcription. The most likely transducers capable of converting extracellular changes to altered gene expression are the mitogen-activated protein (MAP) kinases. Three distinct groups of MAP kinases have been identified in mammalian cells: extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK), and p38. They play a central

1977

Diabetic Neuropathies

role in diverse cellular responses, including cell survival, growth, differentiation, maintenance of phenotype, and cell death.114 Historically ERK has been classified as responding to mitogens and growth factors, implicated in growth and survival and referred to as a MAP kinase. JNK and p38 have been described as stress-activated protein kinases because they respond predominantly to different stress stimuli (e.g., ultraviolet radiation, oxidants, cytokines).69,231 MAP kinases are components of a three-kinase regulatory cascade,113,122,335 which is coupled to cell membrane events via ligand-operated or biophysical signals involving autophosphorylation or GTPases. MAP kinase activation requires dual phosphorylation, and the activated kinase migrates to the nucleus, where it activates transcription factors. The potential involvement of MAP kinases in the development of diabetes complications, including neuropathies, is reviewed in detail elsewhere.253,328 In this context, it is adequate to note that they are activated in neurons by glucose and oxidative stress, in an additive manner, in vitro, and they are activated in peripheral nerves from diabetic rats and in sural nerve biopsies from diabetic patients.253 Finally, recent data have shown that inhibition of the MAP kinase p38 attenuates the conduction deficit in rats with diabetes of 3 months’ duration, but does so selectively for sensory neurons.3 Modern techniques using functional genomics will enable the identification of the genes that are activated or inhibited when p38 MAP kinase is phosphorylated in diabetes, and this will provide a distinct step forward in the identification of new drug targets for neuropathy. Possible candidates are considered later in the section on conduction deficits. In the Schwann cell, diabetes or raised glucose is associated with activation of the JNK subgroup of MAP kinases.58 Downstream this may lead to activation of pro-apoptotic signals and Schwann cell death.58,72 Protein Kinase C Protein kinase C (PKC) has long been the center of hypotheses seeking to explain the pathogenesis of diabetic neuropathy. As is indicated later, early explanations suggested that the result of polyol pathway activity would be reduced activation of PKC, but this was never proven. More recently, it has been suggested that, at least in the retina, diabetes might provoke and increase in the activation of one of the PKC isoforms, PKC␤II. This proposal arose from the effects of an agent, LY333531, whose major pharmacologic property is inhibition of this isoform.149 The critical components of this more recent hypothesis are as follows. A range of non-neural tissues from diabetic rats and derived cultured cell lines exposed to elevated glucose show an increase in diacylglycerol (DAG) levels65,148,365; DAG is the major activator of membrane-bound PKC. Radiolabeling studies suggested that the mechanism involves increased de novo synthesis of DAG. Concomitant with this increase, a rise in

PKC␤II isoform expression was observed in vascular tissues and cells.148 Enhanced PKC activity was associated with localization of an increased proportion of total PKC in the membrane versus the cytosol fraction and this altered enzyme distribution occurred as a result of hypergalactosemia as well as hyperglycemia.365 This suggests a link to exaggerated flux through the polyol pathway. The precise ramifications of these changes are not yet clear, though a review of work to date can be found elsewhere.355 In essence, the activation of PKC␤II is associated with several structural and physiologic changes that impair microvascular perfusion and promote angiogenesis. It is also suggested that elaboration of extracellular matrix is consequent on PKC␤II-induced release of transforming growth factor-␤. Clearly, all these changes are represented and functionally significant in diabetic retinopathy and nephropathy. Whether they form part of the pathogenesis of neuropathies depends on the extent of vascular involvement in causation. This is discussed in detail later and an argument is presented to suggest that inadequate perfusion of peripheral nerve is of secondary importance. Pragmatically, there are studies suggesting that the PKC␤II inhibitor LY333531 prevents the decrease in nerve trunk blood flow seen in diabetic rats,50,64 but no clinical data are in the public domain, and the predictive value of these rat data is unknown.

Abnormalities of Cell Physiology Not all aberrations of cell biology or cell physiology derive from changes in gene expression and altered phenotype. Perhaps the best distinction is to separate those events deriving from prenuclear signals, leading to altered gene expression, from events deriving from extranuclear signals. Endoskeleton, Axonal Transport, and Neurotrophins The clearest extranuclear anomalies in diabetic neuropathy are those affecting the network of endoskeletal and motile proteins in the nerve axon. There are aberrations of neurofilament phosphorylation in both experimental106 and human280,281 diabetic neuropathy, which are superimposed on impaired expression of genes coding for the neurofilament proteins.208 The glucose-derived signals that promote these aberrations are unknown, but activation of the MAP kinase JNK must be a candidate for neurofilament phosphorylation.107 Perturbations of the axonal endoskeleton are more widespread in both human and experimental diabetic neuropathy. Tubulin is glycated,68 and this may lead to increased immunoreactivity.67 Neuronal actin is also glycated,242 and these posttranslational modifications may lead to impaired assembly of axonal endoskeleton with consequent breakdown of axonal integrity.202 There is a correlation between loss of neurofilaments and shrinkage of the axonal cylinder,367 and glycation has been implicated in this process by the presence of glycated protein adducts and the

1978

Diseases of the Peripheral Nervous System

effects of different antiglycation agents on structure and function of peripheral nerves in diabetic rats.311,348,366 Given these disturbances to the microtubule and endoskeletal network, it is not surprising that investigators have examined the competence of axonal transport processes in experimental diabetes. The early work in this area has been reviewed elsewhere.330 These studies revealed deficits in the amount of material carried by fast and slow components both of anterograde axonal transport and of retrograde transport in diabetic rats. However, most of these deficits can be explained by reduced synthesis in the cell body, with deficits in the amount of material loaded onto the anterograde transport system. Where it has been studied, the competence of the transport system per se does not appear to be compromised in experimental diabetes. This is typified by studies on the synthesis and anterograde transport of substance P, for which deficient synthesis explains the reductions in accumulation of axonally transported peptide proximal to constriction in ligated sciatic nerve. The velocity and, hence, the likely competence of translocation is normal.48,262,333 A similar picture emerges with retrograde transport of neurotrophic factors. The clear deficit in retrograde transport of NGF in sciatic nerves of diabetic rats is secondary to reduced synthesis of NGF by innervated target cells in skin and muscle and consequent reduced capture of NGF by the distal extremities of sciatic sensory neurons.104 This deficit is, of course, related to the substance P deficit, because the gene coding for the substance P precursor protein, preprotachykinin A, is regulated by NGF.180 Thus reduced capture and retrograde transport of NGF in primary afferent neurons is responsible for the deficit in synthesis and anterograde transport of substance P and calcitonin gene–related peptide (CGRP).104 Administration of exogenous NGF compensates for this deficit and normalizes expression and transport of both substance P and CGRP.104 The functional result of this neurotrophic deficit is regression of the nociceptive and vasodilator phenotype of C fibers in diabetic rats.23,112 Similar deficits in NGF, substance P, and CGRP are seen in diabetic patients.10,336 Thus it is rational to suppose that suppression of nociceptive and related sensory phenotypes as a result of neurotrophic deficits may form a part of the pathogenesis of diabetic neuropathy. In spite of this, an extensive Phase III clinical trial of subcutaneous injections of human recombinant NGF in diabetic neuropathy failed to show efficacy. A detailed analysis of this trial and its outcomes has been published.11 Essentially, the problem was that side effects, especially where the protein was concentrated at the site of injection, limited the dose that could be used. Second, the study required identification of an end point that could be depended on to resolve within the time frame of the trial. It was also desirable to select end points that were minimally vulnerable to variation in patients’ perception—a hard task with sensory fibers. Neither of these criteria was adequately satisfied. There was

an additional problem, in that it was not possible to discriminate between an effect of NGF to protect “its” neurons from degenerative assault and the capacity of NGF to stimulate representation of the phenotype in neurons that remain unprotected. In other words, NGF treatment may give us a progressively diminishing population of C fibers, with a “hyped-up” phenotype. Clearly side effects may be inevitable with disseminated administration of neurotrophic factors, which physiologically are produced at discrete sites in function-limiting amounts. A possible solution to this would come with drugs that stimulate production of neurotrophins at their proper site of expression. Animal work with these is so far promising,258–260 but we really need to know more about the general capacity of neurotrophins to offer protection or rescue in diabetic neuropathy. This will come when we know more about the spectrum of cellular effects of the various neurotrophins. Neurotrophin-3 is presumably also involved in diabetesinduced deficits,105 and it has the capacity at pharmacologic dosage to give functional neuroprotection in diabetic rats,207 but we do not know its gene targets yet. Much more work is needed using functional genomics. Electrophysiologic Studies of Nerve Function: Conduction Velocity Many studies in experimental animals have demonstrated a reduction in peripheral NCV. For the most part these have followed the first demonstration of a conduction deficit in diabetic rats,90 by concentrating on MNCV. Most of these studies have used the rat, initially with alloxan-induced diabetes and later with STZ as the diabetogenic agent of choice. It is possible to distill this work into a few salient, unequivocal findings. In the STZ-diabetic rat, a slowing of MNCV develops about 2 weeks after the onset of diabetes124,156,200,329 and persists for several months.211,358 Diabetic rats show conduction velocities of about 80% of their nondiabetic counterparts (see earlier citations). The deficit is preventable in the short term (generally about 1 month) by maintenance of tight glycemic control with insulin.124,200 A similar deficit in MNCV is present in the BB rat when control of glycemia is allowed to lapse to a minimal life-preserving level.125,203 Improvement of the quality of glycemic control attenuates the MNCV deficit.126 More detailed examination of nerve conduction in different nerve trunks of STZ-diabetic rats revealed velocity deficits in all branches of the innervation of the lower limb, including some that carry predominantly sensory fibers.52 Just as there is no disagreement that short-term diabetes in rats is associated with a slowing of MNCV, its prevention by ARIs is a similarly ubiquitous finding. Thus treatment of STZ-diabetic rats from the onset of diabetes for short periods (around 1 month) with an ARI results in an MNCV that is not significantly lower than that seen in control rats and is significantly higher than that measured in untreated diabetic littermates. This has been demonstrated for ICI

Diabetic Neuropathies

105552,329 sorbinil,376 zenarestat,169 tolrestat,224 ponalrestat,294 and fidarestat.161 Furthermore, a group of rats left untreated over the first 3 weeks of their STZ-induced diabetes, and displaying a significant MNCV deficit at that time, showed a return of MNCV to normal levels over 3 weeks of treatment with sorbinil.331 Thus in the short term the deficit is reversible. With STZ-induced diabetes of longer duration, the beneficial effects of ARIs against MNCV deficits are less marked. In rats with STZ-induced diabetes of 6 months’ duration, daily administration of the ARI ponalrestat had little effect on the conduction deficit.358 However, this study also revealed that administration of background insulin (twiceweekly doses of the long-acting preparation, heat-treated ultralente) enabled the ARI to maintain a conduction velocity significantly greater than that seen in the rats given insulin alone. The velocity was significantly less than that of the nondiabetic rats,358 indicating that factors other than exaggerated flux through the polyol pathway contribute to MNCV deficits in rats with chronic STZ-induced diabetes. In mice with either genetic (C57 BL/Ks db:db) or STZinduced diabetes, an early (⬍1 month) MNCV deficit is not seen.357 However, with longer durations of diabetes a deficit does develop, and may develop more readily in older animals.120,223 In these animals, the absence of the early MNCV deficit may be attributed to a lack of accumulation of polyol pathway metabolites or to an absence of the related depletion of myo-inositol.47,49 Mouse nerve does contain a polyol-forming enzyme that is susceptible to ARIs, because administration of galactose to these animals causes a marked accumulation of galactitol (dulcitol), and this is prevented by ponalrestat.49 Furthermore, galactose feeding causes a marked MNCV deficit in mice; this also is prevented by AR inhibition with either ponalrestat or sorbinil.47 It is important to note that this MNCV deficit in galactose-fed mice is not accompanied by a depletion of nerve myo-inositol. The assertion that pathogenetic factors other than polyol pathway flux and metabolites contribute to the development of a shortfall in MNCV in chronic diabetes receives support from work with gangliosides. Treatment of diabetic mice with a mixture of bovine brain gangliosides prevented the MNCV deficit in animals with long-term diabetes.223 Taken together, it is possible to deduce the following conclusions concerning MNCV deficits in diabetic rodents. First, the presence of a reduced MNCV early in the syndrome requires an accumulation of polyol pathway metabolites and appears to be related to AR. Second, other factors contribute to reduced MNCV with longer durations of diabetes. Their contribution may be prevented by ganglioside treatment. The way in which the polyol pathway might act to cause the development of the conduction velocity deficit has received much attention in the past. The only cogent neurophysiologic argument to explain the deficit has come

1979

from the suggestion of increased refractoriness of nodes of Ranvier in diabetes leading to deferred activation and propagation.37,38,155 This change may derive from impairment of voltage-gated Na⫹ channels, but it has also been attributed to reduced driving gradient for Na⫹, and this has been attributed in turn to impaired activation of Na⫹, K⫹ATPase.126 The hypothesis then suggests that the depletion of myo-inositol, derived from exaggerated polyol pathway flux, limits the synthesis of inositol-containing phospholipids and that DAG, which is derived from these lipids, regulates activation of subunits of the membrane Na⫹, K⫹ATPase.125 This hypothesis still lacks substantive proof, and the role of impaired Na⫹, K⫹-ATPase activity in the development of conduction deficits remains contentious. Others have suggested that the NCV deficit derives from molecular dislocation of the myelin loops from the axolemma in the paranodal region.293 This would result in a loss of the electrical insulation between the Na⫹ channels in the nodal membrane and the insulated K⫹ channels of the juxtaparanode.354 The result of this would be inappropriate rectification of the nodal action potential and inevitable slowing of conduction. The molecular organization of the node, paranode, and juxtaparanode is highly complex and has been reviewed elsewhere.240 The integrity of the junction between the myelin loops and the axolemma depends on the expression and docking of both Schwann cell proteins (neurofascin, periaxin, and ezrin) and axonal proteins (contactin, caspr). Failure of expression of these—for example in gene-deleted mice—either has been demonstrated to or is likely to result in a drop in NCV (see, e.g., Boyle and associates32). Thus, although the proposal of Sima and coworkers293 that axoglial dysjunction occurs in diabetic nerve is based on chance observations of electron micrographs, it could be verified by molecular biology and represents a cogent explanation for the development of a conduction deficit based on phenotype breakdown in either Schwann cell or neuron. Resistance to Ischemic Conduction Failure This phenomenon has received little attention in animal models of diabetes, probably because limb occlusion is awkward and measurement of quantitative reductions in compound nerve action potential (CAP) amplitude in the tail is less simple than in the sciatic system. Nevertheless, it has long been known that peripheral nerves in diabetic rats exhibit a marked increase in the time taken for ischemia of the trunk to result in a significant (usually 50%) reduction in CAP amplitude.285,286 The dimensional change is similar to that seen in humans, and the increased resistance to ischemic conduction failure (RICF) may be modeled in rats with good fidelity. This makes its cause the more interesting. Low and associates have argued cogently that RICF is a hallmark of endoneurial hypoxia; it is known to develop in nondiabetic rats maintained in a hypoxic atmosphere, thereby creating central hypoxemia.188 Recently, Cameron

1980

Diseases of the Peripheral Nervous System

and Cotter, working in Low’s laboratory, found that administration of guanethidine to diabetic rats prevented the development of both endoneurial ischemia and RICF.51 This implicates the former as a cause of the latter. This study also revealed that guanethidine prevented the fall in MNCV in the same rats.51 Completing the circle, we have the observation that the centrally hypoxemic nondiabetic rats studied by Low and associates also showed a deficit in MNCV.188 These arguments implicating reduced nerve blood flow and ischemic hypoxia as a cause of the NCV deficits in diabetes require special consideration, as follows. The Conundrum of Nerve Blood Flow Observations from the Mayo Clinic in the 1980s stimulated the suggestion that reduced nerve trunk blood flow might bring a primary ischemic cause to diabetic neuropathy.84,337 This was followed by the demonstration of endoneurial hypoxia in exposed sural nerves of diabetic patients with neuropathy.222 There has since followed a debate as to whether this decrease in nerve blood flow, which is undoubtedly associated with neuropathy in humans, is a primary causation, an independent accelerator, or an unrelated parallel phenomenon. Much of the evidence for the primary causation argument has come from the timing of the two events—onset of ischemia and onset of deficient conduction velocity—in diabetic rats, plus the associated effects of vasoactive drugs (see Cameron and colleagues53 for a review proposing the hypothesis). The evidence both from humans and especially from laboratory rats leaves much to be desired, and the hypothesis of primary causation for deficient nerve blood flow is unacceptable. In the first place, the supposition that endoneurial hypoxia of nerve trunks can generate all the features of diabetic symmetrical polyneuropathy must be naïve. The clinical picture is one of predominantly sensory neuropathy in most patients, and it is not possible to argue for the sparing of motor fibers adjacent to hypoxia-damaged sensory axons in ischemic nerve trunks. Second, the coexistence of somatosensory small fiber defects involving axons in trunks and small fiber autonomic neuropathy involving axons in diffused plexuses in the innervated tissues does not support a single vascular causation. Third, a detailed study of relations between sural nerve blood flow and sensory dysfunction showed conclusively that there was no correlation between the two across patients nor was there a correlation in the progression of the change within patients with time.320 In short, sensory function worsens without a change in blood flow. The diabetic rat is a distraction in this context, because the profound nature of the diabetes causes severe acute changes in nerve blood flow derived from muscle wasting.332 Studies by Kiff and colleagues demonstrated a reduction of about 50% in blood flow to the hindquarters in conscious ambulatory diabetic rats, using acoustic Doppler probes.166,167 This is in accordance with the reduction in hind limb muscle mass, so it appears that muscle perfusion, per unit tissue

mass, remains constant. Because the sciatic nerve cannot autoregulate its endoneurial circulation,189 the gross hind limb flow must influence flow in the nerve. Treatment of diabetic rats with anabolic agents increases sciatic nerve blood flow in accord with muscle mass, and dietary restriction in nondiabetic rats causes reduced sciatic nerve blood flow, again with good correlation to hind limb muscle mass.332 Thus measurements of neurovascular function in diabetic rats are unrealistic with reference to the human condition, and this explains the modest effects in humans196 of agents such as ACE inhibitors, which are reported to be profoundly effective in diabetic rats.199 Thus the available evidence suggests that nerve ischemia is associated with nerve dysfunction in diabetic neuropathy,147,222 although it cannot be considered as a primary or the major cause and is best regarded as an independent accelerator. The potential exists for therapy in humans using ACE inhibitors,196 but the benefits will be modest and such drugs will certainly not give the degree of prevention or reversal that we need.

SUMMARY The interactions and hypotheses advanced to explain the pathogenesis of diabetic neuropathies are summarized in Figure 85–7. It is clear that the number of potential targets for therapy is increasing and, with the application of functional genomics and proteomics, will become more specific. Although nothing of a practical nature has evolved to assist management of diabetic neuropathy since the last edition of this book, we may be optimistic about outcomes before the next edition.

REFERENCES 1. Abbott, C. A., Carrington, A. L., Ashe, H., et al.: The NorthWest Diabetes Foot Care Study: incidence of, and risk factors for, new diabetic foot ulceration in a community-based patient cohort. Diabet. Med. 19:377, 2002. 2. Adler, A. I., Boyko, E. J., Ahroni, J. H., et al.: Risk factors for diabetic peripheral sensory neuropathy: results of the Seattle Prospective Diabetic Foot Study. Diabetes Care 20:1162, 1997. 3. Agthong, S., and Tomlinson, D. R.: Inhibition of p38 MAP kinase corrects biochemical and neurological deficits in experimental diabetic neuropathy. Ann. N. Y. Acad. Sci. 973:359, 2002. 4. Airey, M., Bennett, C., Nicolucci, A., and Williams, R.: Aldose reductase inhibitors for the prevention and treatment of diabetic peripheral neuropathy. Cochrane Database Syst. Rev. CD002182, 2000. 5. Allen, R. D., Al Harbi, I. S., Morris, J. G., et al.: Diabetic neuropathy after pancreas transplantation: determinants of recovery. Transplantation 63:830, 1997.

Diabetic Neuropathies 6. Ambepityia, G., Kopelman, P. G., Ingram, D., et al.: Exertional myocardial ischemia in diabetes: a quantitative analysis of anginal perceptual threshold and the influence of autonomic function. J. Am. Coll. Cardiol. 15:72, 1990. 7. American Diabetes Association: Consensus Development Conference on Diabetic Foot Wound Care: 7–8 April 1999, Boston, Massachusetts. Diabetes Care 22:1354, 1999. 8. Ametov, A. S., Barinov, A., Dyck, P. J., et al.: The sensory symptoms of diabetic polyneuropathy are improved with alpha-lipoic acid: the SYDNEY trial. Diabetes Care 26:770, 2003. 9. Amthor, K. F., Dahl-Jorgensen, K., Berg, T. J., et al.: The effect of 8 years of strict glycaemic control on peripheral nerve function in IDDM patients: the Oslo Study. Diabetologia 37:579, 1994. 10. Anand, P., Terenghi, G., Warner, G., et al.: The role of endogenous nerve growth factor in human diabetic neuropathy. Nat. Med. 2:703, 1996. 11. Apfel, S. C.: Nerve growth factor for the treatment of diabetic neuropathy: what went wrong, what went right, and what does the future hold? Int. Rev. Neurobiol. 50:393, 2002. 12. Archer, A. G., Watkins, P. J., Thomas, P. K., et al.: The natural history of acute painful neuropathy in diabetes mellitus. J. Neurol. Neurosurg. Psychiatry 46:491, 1983. 13. Asayama, K., Nakane, T., Uchida, N., et al.: Serum antioxidant status in streptozotocin-induced diabetic rat. Horm. Metab. Res. 26:313, 1994. 14. Asbury, A. K., Aldredge, H., Hershberg, R., and Fisher, C. M.: Oculomotor palsy in diabetes mellitus: a clinico-pathological study. Brain 93:555, 1970. 15. Azad, N., Emanuele, N. V., Abraira, C., et al.: The effects of intensive glycemic control on neuropathy in the VA Cooperative Study on Type II Diabetes Mellitus (VA CSDM). J. Diabetes Complications 13:307, 1999. 16. Backonja, M., and Glanzman, R. L.: Gabapentin dosing for neuropathic pain: evidence from randomized, placebocontrolled clinical trials. Clin. Ther. 25:81, 2003. 17. Backonja, M. M.: Use of anticonvulsants for treatment of neuropathic pain. Neurology 59:S14, 2002. 18. Bagnasco, S. M., Balaban, R., Fales, H. M., et al.: Predominant osmotically active solutes in rat and rabbit renal medullas. J. Biol. Chem. 261:5872, 1986. 19. Bagnasco, S. M., Uchida, S., Balaban, R. S., et al.: Induction of aldose reductase and sorbitol in renal inner medullary cells by elevated extracellular NaCl. Proc. Natl. Acad. Sci. U. S. A. 84:1718, 1987. 20. Ballin, R. H. M., and Thomas, P. K.: Hypertrophic changes in diabetic neuropathy. Acta Neuropathol. (Berl.) 11:93, 1968. 21. Bastron, J. A., and Thomas, J. E.: Diabetic polyradiculopathy: clinical and electromyographic findings in 105 patients. Mayo Clin. Proc. 56:725, 1981. 22. Behse, F., Buchthal, F., and Carlsen, F.: Nerve biopsy and conduction studies in diabetic neuropathy. J. Neurol. Neurosurg. Psychiatry 40:1072, 1977. 23. Bennett, G. S., Garrett, N. E., Diemel, L. T., et al.: Neurogenic cutaneous vasodilatation and plasma extravasation in diabetic rats: effect of insulin and nerve growth factor. Br. J. Pharmacol. 124:1573, 1998. 24. Bertherat, J., Netter, J. M., Timsit, J., et al.: Diabetic thoracoabdominal neuropathy: clinical and electrophysiological

25.

26.

27.

28.

29. 30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

1981

study with evaluation of the autonomic nervous system [in French]. Presse Med. 22:201, 1993. Biesbroeck, R., Bril, V., Hollander, P., et al.: A double-blind comparison of topical capsaicin and oral amitriptyline in painful diabetic neuropathy. Adv. Ther. 12:111, 1995. Bikkina, M., Alpert, M. A., Mukerji, R., et al.: Diminished short-term heart rate variability predicts inducible ventricular tachycardia. Chest 113:312, 1998. Birrell, A. M., Heffernan, S. J., Ansselin, A. D., et al.: Functional and structural abnormalities in the nerves of type I diabetic baboons: aminoguanidine treatment does not improve nerve function. Diabetologia 43:110, 2000. Bottini, P., Dottorini, M. L., Cristina, C. M., et al.: Sleepdisordered breathing in nonobese diabetic subjects with autonomic neuropathy. Eur. Respir. J. 22:654, 2003. Bouchard, M.: Sur la perte des réflexes tendineux dans le diabète sucre. Progres Med. 12:819, 1884. Boulton, A. J., Angus, E., Ayyar, D. R., and Weiss, D. R.: Diabetic thoracic polyradiculopathy presenting as abdominal swelling. Br. Med. J. (Clin. Res. Ed.) 289:798, 1984. Boulton, A. J. M., Ziegler, D., Scarpello, J., et al.: A multicentre double-blind trial of gamolenic acid in diabetic peripheral sensorimotor neuropathy. Diabetologia 41(Suppl 1):A32, 1997. Boyle, M. E. T., Berglund, E. O., Murai, K. K., et al.: Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 30:385, 2001. Bradley, J., Thomas, P. K., King, R. H. M., et al.: Morphometry of endoneurial capillaries in diabetic sensory and autonomic neuropathy. Diabetologia 33:611, 1990. Bradley, J. L., King, R. H. M., Muddle, J. R., and Thomas, P. K.: The extracellular matrix of peripheral nerve in diabetic polyneuropathy. Acta Neuropathol. (Berl.) 99:539, 2000. Bradley, J. L., Thomas, P. K., King, R. H. M., and Watkins, P. J.: A comparison of perineurial and vascular basal laminal changes in diabetic neuropathy. Acta Neuropathol. (Berl.) 88:426, 1994. Brannagan, T. H., Promisloff, R. A., McCluskey, L. F., and Mitz, K. A.: Proximal diabetic neuropathy presenting with respiratory weakness. J. Neurol. Neurosurg. Psychiatry 67:539, 1999. Brismar, T.: Potential clamp experiments on myelinated nerve fibres from alloxan diabetic rats. Acta Physiol. Scand. 105:384, 1979. Brismar, T., and Sima, A. A. F.: Changes in nodal function in nerve fibres of the spontaneously diabetic BB-Wistar rat: potential clamp analysis. Acta Physiol. Scand. 113:499, 1981. Britland, S. T., Young, R. J., Sharma, A. K., and Clarke, B. F.: Association of painful and painless diabetic polyneuropathy with different patterns of nerve fiber degeneration and regeneration. Diabetes 39:898, 1990. Britland, S. T., Young, R. J., Sharma, A. K., and Clarke, B. F.: Acute and remitting painful diabetic polyneuropathy: a comparison of peripheral nerve fibre pathology. Pain 48:361, 1992. Brown, M. J., Martin, J. R., and Asbury, A. K.: Painful diabetic neuropathy: a morphometric study. Arch. Neurol. 33:164, 1976.

1982

Diseases of the Peripheral Nervous System

42. Bruns, L: Uber neuritische Lähmungen beim Diabetes Mellitus. Berl Klin Wochenschr 27:509, 1890. 43. Burger-Kentischer, A., Müller, E., Neuhofer, W., et al.: Expression of Na⫹/Cl⫺/betaine and Na⫹/myo-inositol transporters, aldose reductase and sorbitol dehydrogenase in macula densa cells of the kidney. Pflügers Arch. 436:807, 1998. 44. Burger-Kentischer, A., Müller, E., Neuhofer, W., et al.: Expression of aldose reductase, sorbitol dehydrogenase and Na⫹/myo-inositol and Na⫹/Cl⫺/betaine transporter mRNAs in individual cells of the kidney during changes in the diuretic state. Pflügers Arch. 437:248, 1999. 45. Buzzard, F.: Illustrations of some less known forms of peripheral neuritis, especially alcoholic monoplegia and diabetic neuritis. BMJ 1:1419, 1890. 46. Bytzer, P., Talley, N. J., Leemon, M., et al: Prevalence of gastrointestinal symptoms associated with diabetes mellitus: a population-based survey of 15,000 adults. Arch. Intern. Med. 161:1989, 2001. 47. Calcutt, N. A., Tomlinson, D. R., and Biswas, S.: Coexistence of nerve conduction deficit with increased Na⫹-K⫹-ATPase activity in galactose-fed mice: implications for polyol pathway and diabetic neuropathy. Diabetes 39:663, 1990. 48. Calcutt, N. A., Tomlinson, D. R., Willars, G. B., and Keen, P.: Axonal transport of substance P-like immunoreactivity in ganglioside-treated diabetic rats. J. Neurol. Sci. 96:283, 1990. 49. Calcutt, N. A., Willars, G. B., and Tomlinson, D. R.: Statilsensitive polyol formation in nerve of galactose-fed mice. Metabolism 37:450, 1988. 50. Cameron, N. E., and Cotter, M. A.: Effects of protein kinase Cbeta inhibition on neurovascular dysfunction in diabetic rats: interaction with oxidative stress and essential fatty acid dysmetabolism. Diabetes Metab. Res. Rev. 18:315, 2002. 51. Cameron, N. E., Cotter, M. A., and Low, P. A.: Nerve blood flow in early experimental diabetes in rats: relation to conduction deficits. Am. J. Physiol. Endocrinol. Metab. 261:E1, 1991. 52. Cameron, N. E., Cotter, M. A., and Robertson, S.: The effect of aldose reductase inhibition on the pattern of nerve conduction deficits in diabetic rats. Q. J. Exp. Physiol. 74:917, 1989. 53. Cameron, N. E., Eaton, S. E. M., Cotter, M. A., and Tesfaye, S.: Vascular factors and metabolic interactions in the pathogenesis of diabetic neuropathy. Diabetologia 44:1973, 2001. 54. Caravati, C. M.: Insulin neuritis: a case report. Va. Med. Mon. 59:745, 1933. 55. Catterall, J. R., Calverley, P. M., Ewing, D. J., et al.: Breathing, sleep, and diabetic autonomic neuropathy. Diabetes 33:1025, 1984. 56. Charcot, M.: Sur un cas paraplégie diabétique. Arch Neurol. (Paris) 19:318, 1890. 57. Chaudhuri, K. R., Wren, D. R., Werring, D., and Watkins, P. J.: Unilateral abdominal muscle herniation with pain: a distinctive variant of diabetic radiculopathy. Diabet. Med. 14:803, 1997. 58. Cheng, H. L., Steinway, M. L., Xin, X., and Feldman, E. L.: Insulin-like growth factor-I and Bcl-X(L) inhibit c-jun N-terminal kinase activation and rescue Schwann cells from apoptosis. J. Neurochem. 76:935, 2001.

59. Clouse, R. E., and Lustman, P. J.: Gastrointestinal symptoms in diabetic patients: lack of association with neuropathy. Am. J. Gastroenterol. 84:868, 1989. 60. Collins, S. L., Moore, R. A., McQuay, H. J., and Wiffen, P.: Antidepressants and anticonvulsants for diabetic neuropathy and postherpetic neuralgia: a quantitative systematic review. J. Pain Symptom Manage. 20:449, 2000. 61. Coppack, S. W., and Watkins, P. J.: The natural history of diabetic femoral neuropathy. Q. J. Med. 79:307, 1991. 62. Cornblath, D. R., Dellon, A. L., and Mackinnon, S. E.: Spontaneous diabetes mellitus in a rhesus monkey: neurophysiological studies. Muscle Nerve 12:233, 1989. 63. Cornblath, D. R., Hillman, M. A., Striffler, J. S., et al.: Peripheral neuropathy in diabetic monkeys. Diabetes 38:1365, 1989. 64. Cotter, M. A., Jack, A. M., and Cameron, N. E.: Effects of the protein kinase Cbeta inhibitor LY333531 on neural and vascular function in rats with streptozotocin-induced diabetes. Clin. Sci. (Lond.) 103:311, 2002. 65. Craven, P. A., and De Rubertis, F. R.: Protein kinase C is activated in glomeruli from streptozotocin diabetic rats: possible mediation by glucose. J. Clin. Invest. 83:1667, 1989. 66. Cryer, P. E.: Hypoglycemia of obscure cause. Hosp. Pract. (Off. Ed.) 27:119, 1992. 67. Cullum, N. A., Coleman, J. W., Casson, I. F., and McLean, W. G.: Antibodies to tubulin and microtubule-associated proteins: a study in diabetes mellitus, systemic lupus erythematosus and rheumatoid arthritis. Mol. Chem. Neuropathol. 15:159, 1991. 68. Cullum, N. A., Mahon, J., Stringer, K., and McLean, W. G.: Glycation of rat sciatic nerve tubulin in experimental diabetes mellitus. Diabetologia 34:387, 1991. 69. Davis, R. J.: Signal transduction by the JNK group of MAP kinases. Cell 103:239, 2000. 70. Davis, S. N., Shavers, C., Costa, F., and Mosqueda-Garcia, R.: Role of cortisol in the pathogenesis of deficient counterregulation after antecedent hypoglycemia in normal humans. J. Clin. Invest. 98:680, 1996. 71. De Boer, S. Y., Masclee, A. A., Lam, W. F., and Lamers, C. B.: Effect of acute hyperglycemia on esophageal motility and lower esophageal sphincter pressure in humans. Gastroenterology 103:775, 1992. 72. Delaney, C. L., Russell, J. W., Cheng, H. L., and Feldman, E. L.: Insulin-like growth factor-I and over-expression of Bcl-xL prevent glucose-mediated apoptosis in Schwann cells. J. Neuropathol. Exp. Neurol. 60:147, 2001. 73. Demaine, A. G.: Polymorphisms of the aldose reductase gene and susceptibility to diabetic microvascular complications. Curr. Med. Chem. 10:1389, 2003. 74. Diabetes Control and Complications Trial Research Group: The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulindependent diabetes mellitus. N. Engl. J. Med. 329:977, 1993. 75. Diabetes Control and Complications Trial Research Group: The effect of intensive diabetes therapy on the development and progression of neuropathy. Ann. Intern. Med. 122:561, 1995. 76. Dooley, C. P., el Newihi, H. M., Zeidler, A., and Valenzuela, J. E.: Abnormalities of the migrating motor complex in

Diabetic Neuropathies

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90. 91. 92. 93. 94.

diabetics with autonomic neuropathy and diarrhea. Scand. J. Gastroenterol. 23:217, 1988. Douglas, N. J., Campbell, I. W., Ewing, D. J., et al.: Reduced airway vagal tone in diabetic patients with autonomic neuropathy. Clin. Sci. (Lond.) 61:581, 1981. Dvornik, D.: Aldose Reductase Inhibition—An Approach to the Prevention of Diabetic Complications. New York, Biomedical Information Corporation (McGraw-Hill), 1987. Dyck, J. B., and Dyck, P. J.: Diabetic polyneuropathy. In Dyck, P. J., and Thomas, P. K. (eds.): Diabetic Neuropathy. Philadelphia, W. B. Saunders, p. 255, 1999. Dyck, P. J., Davies, J. L., Litchy, W. J., and O’Brien, P. C.: Longitudinal assessment of diabetic polyneuropathy using a composite score in the Rochester Diabetic Neuropathy Study cohort. Neurology 49:229, 1997. Dyck, P. J., Davies, J. L., Wilson, D. M., et al.: Risk factors for severity of diabetic polyneuropathy: intensive longitudinal assessment of the Rochester Diabetic Neuropathy Study cohort. Diabetes Care 22:1479, 1999. Dyck, P. J., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy (LSRPN): similarity to the diabetic variety (DLSRPN). J. Neuropathol. Exp. Neurol. 59:525, 2000. Dyck, P. J., Hansen, S., Karnes, J., et al.: Capillary number and percentage closed in human diabetic sural nerve. Proc. Natl. Acad. Sci. U. S. A. 82:2513, 1985. Dyck, P. J., Karnes, J. L., O’Brien, P., et al.: The spatial distribution of fiber loss in diabetic polyneuropathy suggests ischemia. Ann. Neurol. 19:440, 1986. Dyck, P. J., Kratz, K. M., Karnes, J. L., et al.: The prevalence by staged severity of various types of diabetic neuropathy, retinopathy, and nephropathy in a population-based cohort: the Rochester Diabetic Neuropathy Study. Neurology 43:817, 1993. Dyck, P. J., Lais, A., Karnes, J. L., et al.: Fiber loss is primary and multifocal in sural nerves in diabetic polyneuropathy. Ann. Neurol. 19:425, 1986. Dyck, P. J., Zimmerman, B. R., Vilen, T. H., et al.: Nerve glucose, fructose, sorbitol, myo-inositol, and fiber degeneration and regeneration in diabetic neuropathy. N. Engl. J. Med. 319:542, 1988. Edmonds, M. E., and Watkins, P. J.: Plantar neuropathic ulcer and Charcot joints: risk factors, presentation and management. In Dyck, P. J., and Thomas, P. K. (eds.): Diabetic Neuropathy. Philadelphia, W. B. Saunders, p. 398, 1999. Eichorst, H.: Beiträge zur Pathologie der Nerven und Muskeln. 3. Neuritis diabetica und ihre Beziehungen zum fehlenden Patbbellarsehnenreflex. Virchows Arch. Pathol. Anat. Physiol. 127:1, 1892. Eliasson, S. G.: Nerve conduction changes in experimental diabetes. J. Clin. Invest. 43:2353, 1964. Ellenberg, M.: Diabetic neuropathy precipitating after institution of diabetic control. Am. J. Med. Sci. 236:466, 1958. Ellenberg, M.: Diabetic neuropathic cachexia. Diabetes 23:418, 1974. Ellenberg, M., and Weber, H.: Retrograde ejaculation in diabetic neuropathy. Ann. Intern. Med. 65:1237, 1966. Enck, P., Rathmann, W., Spiekermann, M., et al.: Prevalence of gastrointestinal symptoms in diabetic patients and nondiabetic subjects. Z. Gastroenterol. 32:637, 1994.

1983

95. Faerman, I., Faccio, E., Milei, J., et al.: Autonomic neuropathy and painless myocardial infarction in diabetic patients: histologic evidence of their relationship. Diabetes 26:1147, 1977. 96. Fagerberg, S. E.: Diabetic neuropathy: a clinical and histological study on the significance of vascular affections. Acta Med. Scand. 164:1, 1959. 97. Fanelli, C. G., Epifano, L., Rambotti, A. M., et al.: Meticulous prevention of hypoglycemia normalizes the glycemic thresholds and magnitude of most of neuroendocrine responses to, symptoms of, and cognitive function during hypoglycemia in intensively treated patients with short-term IDDM. Diabetes 42:1683, 1993. 98. Faradji, V., and Sotelo, J.: Low serum levels of nerve growth factor in diabetic neuropathy. Acta Neurol. Scand. 81:402, 1990. 99. Fealey, R. D.: Thermoregulatory sweat test. In Daube, J. R. (ed.): Clinical Neurophysiology. Philadelphia, F. A. Davis, p. 396, 1996. 100. Fealey, R. D., Low, P. A., and Thomas, J. E.: Thermoregulatory sweating abnormalities in diabetes mellitus. Mayo Clin. Proc. 64:617, 1989. 101. Fedele, D., Coscelli, C., Santeusanio, F., et al.: Erectile dysfunction in diabetic subjects in Italy. Gruppo Italiano Studio Deficit Erettile nei Diabetici. Diabetes Care 21:1973, 1998. 102. Feldman, H. A., Goldstein, I., Hatzichristou, D. G., et al.: Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J. Urol. 151:54, 1994. 103. Feldman, M., and Schiller, L. R.: Disorders of gastrointestinal motility associated with diabetes mellitus. Ann. Intern. Med. 98:378, 1983. 104. Fernyhough, P., Diemel, L. T., Hardy, J., et al.: Human recombinant nerve growth factor replaces deficient neurotrophic support in the diabetic rat. Eur. J. Neurosci. 7:1107, 1995. 105. Fernyhough, P., Diemel, L. T., and Tomlinson, D. R.: Target tissue production and axonal transport of neurotrophin-3 are reduced in streptozocin-diabetic rats. Diabetologia 41:300, 1998. 106. Fernyhough, P., Gallagher, A., Averill, S. A., et al.: Aberrant neurofilament phosphorylation in sensory neurons of rats with diabetic neuropathy. Diabetes 48:881, 1999. 107. Fernyhough, P., and Schmidt, R. E.: Neurofilaments in diabetic neuropathy. Int. Rev. Neurobiol. 50:115, 2002. 108. Forrest, K. Y., Maser, R. E., Pambianco, G., et al.: Hypertension as a risk factor for diabetic neuropathy: a prospective study. Diabetes 46:665, 1997. 109. Foster, A. V., Eaton, C., McConville, D. O., and Edmonds, M. E.: Application of OpSite film: a new and effective treatment of painful diabetic neuropathy. Diabet. Med. 11:768, 1994. 110. Gabel, H., Bitter-Suermann, H., Henriksson, C., et al.: Streptozotocin diabetes in juvenile pigs: evaluation of an experimental model. Horm. Metab. Res. 17:275, 1985. 111. Galer, B. S., Gianas, A., and Jensen, M. P.: Painful diabetic polyneuropathy: epidemiology, pain description, and quality of life. Diabetes Res. Clin. Pract. 47:123, 2000. 112. Garrett, N. E., Malcangio, M., Dewhurst, M., and Tomlinson, D. R.: ␣-Lipoic acid corrects neuropeptide deficits in diabetic

1984

113.

114.

115.

116.

117.

118.

119.

120.

121.

122. 123.

124.

125.

126.

127.

128.

129.

Diseases of the Peripheral Nervous System

rats via induction of trophic support. Neurosci. Lett. 222:191, 1997. Garrington, T. P., and Johnson, G. L.: Organization and regulation of mitogen-activated protein kinase signaling pathways. Curr. Opin. Cell Biol. 11:211, 1999. Geilen, C. C., Wieprecht, M., and Orfanos, C. E.: The mitogen-activated protein kinases system (MAP kinase cascade): its role in skin signal transduction. A review. J. Dermatol. Sci. 12:255, 1996. Giannini, C., and Dyck, P. J.: Basement membrane reduplication and pericyte degeneration precede development of diabetic polyneuropathy and are associated with its severity. Ann. Neurol. 37:498, 1995. Giannini, C., and Dyck, P. J.: Pathologic alterations in human diabetic polyneuropathy. In Dyck, P. J., and Thomas, P. K. (eds.): Diabetic Neuropathy. Philadelphia, W. B. Saunders, p. 279, 1999. Giardino, I., Fard, A. K., Hatchell, D. L., and Brownlee, M.: Aminoguanidine inhibits reactive oxygen species formation, lipid peroxidation, and oxidant-induced apoptosis. Diabetes 47:1114, 1998. Gibbs, G., Wilson, R., and Ho, C. K.: Experimental diabetes in the monkey. Proc. 2nd Int. Congr. Primates 3:169, 1969. Goldstein, J. E., and Cogan, D. G.: Diabetic ophthalmoplegia with special reference to the pupil. Arch. Ophthalmol. 64:592, 1960. Gorio, A., Aporti, F., Di Gregorio, F., et al.: Ganglioside treatment of genetic and alloxan-induced diabetic neuropathy. Adv. Exp. Med. Biol. 174:549, 1984. Gorson, K. C., Ropper, A. H., Adelman, L. S., and Weinberg, D. H.: Influence of diabetes mellitus on chronic inflammatory demyelinating polyneuropathy. Muscle Nerve 23:37, 2000. Graves, J. D., and Krebs, E. G.: Protein phosphorylation and signal transduction. Pharmacol. Ther. 82:111, 1999. Greene, D. A., Arezzo, J. C., and Brown, M. B.: Effect of aldose reductase inhibition on nerve conduction and morphometry in diabetic neuropathy. Zenarestat Study Group. Neurology 53:580, 1999. Greene, D. A., De Jesus, P. V. Jr., and Winegrad, A. I.: Effects of insulin and dietary myoinositol on impaired peripheral motor nerve conduction velocity in acute streptozotocin diabetes. J. Clin. Invest. 55:1326, 1975. Greene, D. A., Lattimer, S. A., and Sima, A. A. F.: Sorbitol, phosphoinositides, and sodium-potassium-ATPase in the pathogenesis of diabetic complications. N. Engl. J. Med. 316:599, 1987. Greene, D. A., Yagihashi, S., Lattimer, S. A., and Sima, A. A. F.: Nerve Na⫹-K⫹-ATPase, conduction, and myoinositol in the insulin-deficient BB rat. Am. J. Physiol. Endocrinol. Metab. 247:E534, 1984. Greene, L. F., Kelalis, P. P., and Weeks, R. E.: Retrograde ejaculation of semen due to diabetic neuropathy: report of 4 cases. Fertil. Steril. 14:617, 1963. Gregersen, G.: Variations in motor conduction velocity produced by acute changes of the metabolic state in diabetic patients. Diabetologia 4:273, 1968. Grüssner, R., Nakhleh, R., Grüssner, A., et al.: Streptozotocininduced diabetes mellitus in pigs. Horm. Metab. Res. 25:199, 1993.

130. Gugliucci, A., and Bendayan, M.: Reaction of advanced glycation endproducts with renal tissue from normal and streptozotocin-induced diabetic rats: an ultrastructural study using colloidal gold cytochemistry. J. Histochem. Cytochem. 43:591, 1995. 131. Gutteridge, J. M. C., and Halliwell, B.: The measurement and mechanism of lipid peroxidation in biological systems. Trends Biochem. Sci. 15:129, 1990. 132. Hackett, G.: Managing erectile dysfunction. Practitioner, 242:632, 1998. 133. Hale, P. J., Nattrass, M., Silverman, S. H., et al.: Peripheral nerve concentrations of glucose, fructose, sorbitol and myoinositol in diabetic and non-diabetic patients. Diabetologia 30:464, 1987. 134. Halliwell, B.: Oxidants and human disease: some new concepts. FASEB J. 1:358, 1987. 135. Haq, R. U., Pendlebury, W. W., Fries, T. J., and Tandan, R.: Chronic inflammatory demyelinating polyradiculoneuropathy in diabetic patients. Muscle Nerve 27:465, 2003. 136. Harris, M., Eastman, R., and Cowie, C.: Symptoms of sensory neuropathy in adults with NIDDM in the U.S. population. Diabetes Care 16:1446, 1993. 137. Heaton, R. W., Guy, R. J., Gray, B. J., et al.: Diminished bronchial reactivity to cold air in diabetic patients with autonomic neuropathy. Br. Med. J. (Clin. Res. Ed.) 289:149, 1984. 138. Heckmann, J. G., Dietrich, W., Hohenberger, W., et al.: Hypoglycemic sensorimotor polyneuropathy associated with insulinoma. Muscle Nerve 23:1891, 2000. 139. Hogan, M., Cerami, A., and Bucala, R.: Advanced glycosylation endproducts block the antiproliferative effect of nitric oxide: role in the vascular and renal complications of diabetes mellitus. J. Clin. Invest. 90:1110, 1992. 140. Hopf, H. C., and Gutmann, L.: Diabetic 3rd nerve palsy: evidence for a mesencephalic lesion. Neurology 40:1041, 1990. 141. Horowitz, M., Harding, P. E., Maddox, A. F., et al.: Gastric and oesophageal emptying in patients with type 2 (noninsulin-dependent) diabetes mellitus. Diabetologia 32:151, 1989. 142. Hotta, N., Sakamoto, N., Shigeta, Y., et al.: Clinical investigation of epalrestat, an aldose reductase inhibitor, on diabetic neuropathy in Japan: multicenter study. Diabetic Neuropathy Study Group in Japan. J. Diabetes Complications 10:168, 1996. 143. Hotta, N., Toyota, T., Matsuoka, K., et al.: Clinical efficacy of fidarestat, a novel aldose reductase inhibitor, for diabetic peripheral neuropathy: a 52-week multicenter placebocontrolled double-blind parallel group study. Diabetes Care 24:1776, 2001. 144. Hounsom, L., and Tomlinson, D. R.: Does neuropathy develop in animal models? Clin. Neurosci. 4:381, 1997. 145. Hunt, J. V., Dean, R. T., and Wolff, S. P.: Hydroxyl radical production and autoxidative glycosylation: glucose autoxidation as the cause of protein damage in the experimental glycation model of diabetes mellitus and ageing. Biochem. J. 256:205, 1988. 146. Huttunen, H. J., Fages, C., and Rauvala, H.: Receptor for advanced glycation end products (RAGE)-mediated neurite outgrowth and activation of NF-kappaB require the

Diabetic Neuropathies

147.

148.

149.

150.

151.

152.

153.

154.

155.

156.

157. 158.

159. 160.

161.

162.

163.

164.

cytoplasmic domain of the receptor but different downstream signaling pathways. J. Biol. Chem. 274:19919, 1999. Ibrahim, S., Harris, N. D., Radatz, M., et al.: A new minimally invasive technique to show nerve ischaemia in diabetic neuropathy. Diabetologia 42:737, 1999. Inoguchi, T., Battan, R., Handler, E., et al.: Preferential elevation of protein kinase C isoform ␤II and diacylglycerol levels in the aorta and heart of diabetic rats: differential reversibility to glycemic control by islet cell transplantation. Proc. Natl. Acad. Sci. U. S. A. 89:11059, 1992. Ishii, H., Jirousek, M. R., Koya, D., et al.: Amelioration of vascular dysfunctions in diabetic rats by an oral PKC beta inhibitor. Science 272:728, 1996. Jacobson, D. M.: Pupil involvement in patients with diabetesassociated oculomotor nerve palsy. Arch. Ophthalmol. 116:723, 1998. Jamal, G. A., and Carmichael, H.: The effect of gammalinolenic acid on human diabetic peripheral neuropathy: a double-blind placebo-controlled trial. Diabet. Med. 7:319, 1990. Jann, S., Bramerio, M. A., Beretta, S., et al.: Diagnostic value of sural nerve matrix metalloproteinase-9 in diabetic patients with CIDP. Neurology 61:1607, 2003. Jarvis, B., and Coukell, A. J.: Mexiletine: a review of its therapeutic use in painful diabetic neuropathy. Drugs 56:691, 1998. Jayasinghe, K. S., Nimalasuriya, A., and Dharmadasa, K.: A case of insulinoma with peripheral neuropathy. Postgrad. Med. J. 59:189, 1983. Jefferys, J. G. R., and Brismar, T.: Analysis of peripheral nerve function in streptozotocin diabetic rats. J. Neurol. Sci. 48:435, 1980. Jefferys, J. G. R., Palmano, K. P., Sharma, A. K., and Thomas, P. K.: Influence of dietary myoinositol on nerve conduction and inositol phospholipids in normal and diabetic rats. J. Neurol. Neurosurg. Psychiatry 41:333, 1978. Johnson, P. C., Doll, S. C., and Cromey, D. W.: Pathogenesis of diabetic neuropathy. Ann. Neurol. 19:450, 1986. Jones, K. L., Horowitz, M., Berry, M., et al.: Blood glucose concentration influences postprandial fullness in IDDM. Diabetes Care 20:1141, 1997. Jordan, W. R.: Neuritic manifestations in diabetes mellitus. Arch. Intern. Med. 57:307, 1936. Kalimo, H., Maki, J., Paetau, A., and Haltia, M.: Microanalysis of perineural calcification in diabetic nephropathy. Muscle Nerve 4:228, 1981. Kato, N., Mizuno, K., Makino, M., et al.: Effects of 15-month aldose reductase inhibition with fidarestat on the experimental diabetic neuropathy in rats. Diabetes Res. Clin. Pract. 50:77, 2000. Katz, J. S., Saperstein, D. S., Wolfe, G., et al.: Cervicobrachial involvement in diabetic radiculoplexopathy. Muscle Nerve 24:794, 2001. Keen, H., Payan, J., Allawi, J., et al.: Treatment of diabetic neuropathy with gamma-linolenic acid. The gamma-Linolenic Acid Multicenter Trial Group. Diabetes Care 16:8, 1993. Kelkar, P., Masood, M., and Parry, G. J.: Distinctive pathologic findings in proximal diabetic neuropathy (diabetic amyotrophy). Neurology 55:83, 2000.

1985

165. Kennedy, W. R., Navarro, X., Goetz, F. C., et al.: Effects of pancreatic transplantation on diabetic neuropathy. N. Engl. J. Med. 322:1031, 1990. 166. Kiff, R. J., Gardiner, S. M., Compton, A. M., and Bennett, T.: The effects of NG-nitro-L-arginine methyl ester on regional haemodynamics in conscious rats with streptozotocin-induced diabetes mellitus. Br. J. Pharmacol. 103:1321, 1991. 167. Kiff, R. J., Gardiner, S. M., Compton, A. M., and Bennett, T.: Selective impairment of hindquarters vasodilator responses to bradykinin in conscious Wistar rats with streptozotocininduced diabetes mellitus. Br. J. Pharmacol. 103:1357, 1991. 168. Kihara, M., Schmelzer, J. D., Poduslo, J. F., et al.: Aminoguanidine effects on nerve blood flow, vascular permeability, electrophysiology, and oxygen free radicals. Proc. Natl. Acad. Sci. U. S. A. 88:6107, 1991. 169. Kikkawa, R., Hatanaka, I., Yasuda, H., et al.: Effect of a new aldose reductase inhibitor, (E)-3-carboxymethyl-5-[(2E)methyl-3-phenylpropenylidene]rhodanine (ONO-2235) on peripheral nerve disorders in streptozotocin-diabetic rats. Diabetologia 24:290, 1983. 170. King, R. H. M., Llewelyn, J. G., Thomas, P. K., et al.: Perineurial calcification. Neuropathol. Appl. Neurobiol. 14:105, 1988. 171. King, R. H. M., Llewelyn, J. G., Thomas, P. K., et al.: Diabetic neuropathy: abnormalities of Schwann cell and perineurial basal laminae. Implications for diabetic vasculopathy. Neuropathol. Appl. Neurobiol. 15:339, 1989. 172. Klein, R., Klein, B. E., Lee, K. E., et al.: Prevalence of selfreported erectile dysfunction in people with long-term IDDM. Diabetes Care 19:135, 1996. 173. Ko, G. T., Chan, W. B., Chan, J. C., et al.: Gastrointestinal symptoms in Chinese patients with type 2 diabetes mellitus. Diabet. Med. 16:670, 1999. 174. Koistinen, M. J.: Prevalence of asymptomatic myocardial ischaemia in diabetic subjects. BMJ 301:92, 1990. 175. Koistinen, M. J., Airaksinen, K. E., Huikuri, H. V., et al.: No difference in cardiac innervation of diabetic patients with painful and asymptomatic coronary artery disease. Diabetes Care 19:231, 1996. 176. Lauria, G., McArthur, J. C., Hauer, P. E., et al.: Neuropathological alterations in diabetic truncal neuropathy: evaluation by skin biopsy. J. Neurol. Neurosurg. Psychiatry 65:762, 1998. 177. Lehtinen, J. M., Uusitupa, M., Siitonen, O., and Pyorala, K.: Prevalence of neuropathy in newly diagnosed NIDDM and nondiabetic control subjects. Diabetes 38:1307, 1989. 178. Levitt, N. S., Stansberry, K. B., Wynchank, S., and Vinik, A. I.: The natural progression of autonomic neuropathy and autonomic function tests in a cohort of people with IDDM. Diabetes Care 19:751, 1996. 179. Leyden, E.: Beiträg zur Klinik des Diabetes melliotus. Wien. Med. Wochenschr. 43:926, 1893. 180. Lindsay, R. M., and Harmar, A. J.: Nerve growth factor regulates expression of neuropeptide genes in adult sensory neurons. Nature 337:362, 1989. 181. Linstedt, U., Jaeger, H., and Petry, A.: [The neuropathy of the autonomic nervous system: an additional anesthetic risk in diabetes mellitus]. Anaesthesist 42:521, 1993. 182. Litwak, K. N., Cefalu, W. T., and Wagner, J. D.: Streptozotocin-induced diabetes mellitus in cynomolgus

1986

183.

184.

185.

186.

187.

188.

189.

190.

191.

192.

193.

194.

195.

196.

197. 198.

199.

Diseases of the Peripheral Nervous System

monkeys: changes in carbohydrate metabolism, skin glycation, and pancreatic islets. Lab. Anim. Sci. 48:172, 1998. Llewelyn, J. G., Gilbey, S. G., Thomas, P. K., et al.: Sural nerve morphometry in diabetic autonomic and painful sensory neuropathy: a clinicopathological study. Brain 114(Pt. 2):867, 1991. Llewelyn, J. G., Thomas, P. K., Fonseca, V., et al.: Acute painful diabetic neuropathy precipitated by strict glycaemic control. Acta Neuropathol. (Berl.) 72:157, 1986. Llewelyn, J. G., Thomas, P. K., Gilbey, S. G., et al.: Pattern of myelinated fibre loss in the sural nerve in neuropathy related to type 1 (insulin-dependent) diabetes. Diabetologia 31:162, 1988. Llewelyn, J. G., Thomas, P. K., and King, R. H.: Epineurial microvasculitis in proximal diabetic neuropathy. J. Neurol. 245:159, 1998. Low, P. A., and Nickander, K. K.: Oxygen free radical effects in sciatic nerve in experimental diabetes. Diabetes 40:873, 1991. Low, P. A., Schmelzer, J. D., Ward, K. K., and Yao, J. K.: Experimental chronic hypoxic neuropathy: relevance to diabetic neuropathy. Am. J. Physiol. Endocrinol. Metab. 250:E94, 1986. Low, P. A., and Tuck, R. R.: Effects of changes of blood pressure, respiratory acidosis and hypoxia on blood flow in the sciatic nerve of the rat. J. Physiol. (Lond.) 347:513, 1984. Low, P. A., Walsh, J. C., Huang, C. Y., and McLeod, J. G.: The autonomic nervous system in alcoholic and diabetic neuropathy. Proc. Aust. Assoc. Neurol. 12:145, 1975. Lowitt, S., Malone, J. I., Salem, A. F., et al.: Acetyl-L-carnitine corrects the altered peripheral nerve function of experimental diabetes. Metabolism 44:677, 1995. Lozeron, P., Nahum, L., Lacroix, C., et al.: Symptomatic diabetic and non-diabetic neuropathies in a series of 100 diabetic patients. J. Neurol. 249:569, 2002. Lustman, P. J., and Clouse, R. E.: Relationship of psychiatric illness to impotence in men with diabetes. Diabetes Care 13:893, 1990. Malik, R. A., Newrick, P. G., Sharma, A. K., et al.: Microangiopathy in human diabetic neuropathy: relationship between capillary abnormalities and the severity of neuropathy. Diabetologia 32:92, 1989. Malik, R. A., Veves, A., Masson, E. A., et al.: Endoneurial capillary abnormalities in mild human diabetic neuropathy. J. Neurol. Neurosurg. Psychiatry 55:557, 1992. Malik, R. A., Williamson, S., Abbott, C., et al.: Effect of angiotensin-converting-enzyme (ACE) inhibitor trandolapril on human diabetic neuropathy: randomised double-blind controlled trial. Lancet 352:1978, 1998. Marchal de Calvi, J.: Recherches sur les Accidents Diabétiques. 1864. Maser, R. E., Pfeifer, M. A., Dorman, J. S., et al.: Diabetic autonomic neuropathy and cardiovascular risk. Pittsburgh Epidemiology of Diabetes Complications Study III. Arch. Intern. Med. 150:1218, 1990. Maxfield, E. K., Cameron, N. E., Cotter, M. A., and Dines, K. C.: Angiotensin II receptor blockade improves nerve function, modulates nerve blood flow and stimulates endoneurial angiogenesis in streptozotocin-diabetic rats and nerve function. Diabetologia 36:1230, 1993.

200. Mayer, J. H., and Tomlinson, D. R.: Axonal transport of cholinergic transmitter enzymes in vagus and sciatic nerves of rats with acute experimental diabetes mellitus: correlation with motor nerve conduction velocity and effects of insulin. Neuroscience 9:951, 1983. 201. McLean, W. G.: The role of the axonal cytoskeleton in diabetic neuropathy. Neurochem. Res. 22:951, 1997. 202. McLean, W. G., Roberts, R. E., and Mullins, F. H.: Posttranslational modifications of microtubule- and growthassociated proteins in nerve regeneration and neuropathy. Biochem. Soc. Trans. 23:76, 1995. 203. Mendell, J. R., Sahenk, Z., Warmolts, J. R., et al.: The spontaneously diabetic BB Wistar rat: morphologic and physiologic studies of peripheral nerve. J. Neurol. Sci. 52:103, 1981. 204. Mésangeau, D., Laude, D., and Elghozi, J. L.: Early detection of cardiovascular autonomic neuropathy in diabetic pigs using blood pressure and heart rate variability. Cardiovasc. Res. 45:889, 2000. 205. MiSAD Group: Prevalence of unrecognized silent myocardial ischemia and its association with atherosclerotic risk factors in noninsulin-dependent diabetes mellitus. Milan Study on Atherosclerosis and Diabetes (MiSAD) Group. Am. J. Cardiol. 79:134, 1997. 206. Misko, T. P., Moore, W. M., Kasten, T. P., et al.: Selective inhibition of the inducible nitric oxide synthase by aminoguanidine. Eur. J. Pharmacol. 233:119, 1993. 207. Mizisin, A., Calcutt, N. A., Tomlinson, D. R., et al.: Neurotrophin-3 reverses nerve conduction velocity deficits in streptozotocin-diabetic rats. J. Peripher. Nerv. Syst. 4:211, 1999. 208. Mohiuddin, L., Fernyhough, P., and Tomlinson, D. R.: Reduced levels of mRNA encoding endoskeletal and growth-associated proteins in sensory ganglia in experimental diabetes mellitus. Diabetes 44:25, 1995. 209. Mohseni, S.: Hypoglycemic neuropathy. Acta Neuropathol. (Berl.) 102:413, 2001. 210. Monboisse, J. C., Gillery, P., Maquart, F. X., and Borel, J. P.: Production of superoxide anion by glycated proteins: involvement in complications of diabetes mellitus. Adv. Exp. Med. Biol. 264:551, 1990. 211. Moore, S. A., Peterson, R. G., Felten, D. L., and O’Connor, B. L.: A quantitative comparison of motor and sensory conduction velocities in short- and long-term streptozotocin- and alloxan-diabetic rats. J. Neurol. Sci. 48:133, 1980. 212. Morello, C. M., Leckband, S. G., Stoner, C. P., et al.: Randomized double-blind study comparing the efficacy of gabapentin with amitriptyline on diabetic peripheral neuropathy pain. Arch. Intern. Med. 159:1931, 1999. 213. Mukherjee, B., Mukherjee, J. R., and Chatterjee, M.: Lipid peroxidation, glutathione levels and changes in glutathionerelated enzyme activities in streptozotocin-induced diabetic rats. Immunol. Cell Biol. 72:109, 1994. 214. Mulder, D. W., Lambert, E. H., Bastron, J. A., and Sprague, R. G.: The neuropathies associated with diabetes mellitus: a clinical and electromyographic study of 103 unselected diabetic patients. Neurology 11(4 Pt. 1):275, 1961. 215. Murray, D. P., O’Brien, T., Mulrooney, R., and O’Sullivan, D. J.: Autonomic dysfunction and silent myocardial ischaemia on exercise testing in diabetes mellitus. Diabet. Med. 7:580, 1990.

Diabetic Neuropathies 216. Mustonen, J., Uusitupa, M., Mantysaari, M., et al.: Changes in autonomic nervous function during the 4-year follow-up in middle-aged diabetic and nondiabetic subjects initially free of coronary heart disease. J. Intern. Med. 241:227, 1997. 217. Navarro, X., Sutherland, D. E., and Kennedy, W. R.: Longterm effects of pancreatic transplantation on diabetic neuropathy. Ann. Neurol. 42:727, 1997. 218. Neeper, M., Schmidt, A. M., Brett, J., et al.: Cloning and expression of a cell surface receptor for advanced glycosylation end products of proteins. J. Biol. Chem. 267:14998, 1992. 219. Neil, H. A., Thompson, A. V., Thorogood, M., et al.: Diabetes in the elderly: the Oxford Community Diabetes Study. Diabet. Med. 6:608, 1989. 220. Neumann, C., Martinez, D., and Schmid, H.: Nocturnal oxygen desaturation in diabetic patients with severe autonomic neuropathy. Diabetes Res. Clin. Pract. 28:97, 1995. 221. Newrick, P. G., Boulton, A. J., and Ward, J. D.: The distribution of diabetic neuropathy in a British clinic population. Diabetes Res. Clin. Pract. 2:263, 1986. 222. Newrick, P. G., Wilson, A. J., Jakubowski, J. A., et al.: Sural nerve oxygen tension in diabetes. BMJ 293:1053, 1986. 223. Norido, F., Canella, R., Zanoni, R., and Gorio, A.: Development of diabetic neuropathy in the C57BL/Ks (db/db) mouse and its treatment with gangliosides. Exp. Neurol. 83:221, 1984. 224. Notvest, R. R., and Inserra, J. J.: Tolrestat, an aldose reductase inhibitor, prevents nerve dysfunction in conscious diabetic rats. Diabetes 36:500, 1987. 225. Novella, S. P., Inzucchi, S. E., and Goldstein, J. M.: The frequency of undiagnosed diabetes and impaired glucose tolerance in patients with idiopathic sensory neuropathy. Muscle Nerve 24:1229, 2001. 226. Nukada, H., and McMorran, P. D.: Perivascular demyelination and intramyelinic oedema in reperfusion nerve injury. J. Anat. 185(Pt. 2):259, 1994. 227. Oates, P. J.: Polyol pathway and diabetic peripheral neuropathy. Int. Rev. Neurobiol. 50:325, 2002. 228. Obrosova, I. G., Fathallah, L., and Greene, D. A.: Early changes in lipid peroxidation and antioxidative defense in diabetic rat retina: effect of DL-alpha-lipoic acid. Eur. J. Pharmacol. 398:139, 2000. 229. Ogle, J. W.: On disease of the brain as a result of diabetes mellitus. St George’s Hosp. Rep. 1:157, 1866. 230. Ohkubo, Y., Kishikawa, H., Araki, E., et al.: Intensive insulin therapy prevents the progression of diabetic microvascular complications in Japanese patients with non-insulin-dependent diabetes mellitus: a randomized prospective 6-year study. Diabetes Res. Clin. Pract. 28:103, 1995. 231. Ono, K., and Han, J. H.: The p38 signal transduction pathway: activation and function. Cell. Signal. 12:1, 2000. 232. Ozkul, Y., Sabuncu, T., Kocabey, Y., and Nazligul, Y.: Outcomes of carpal tunnel release in diabetic and nondiabetic patients. Acta Neurol. Scand. 106:168, 2002. 233. Palumbo, P. J., Elveback, L. R., Chu, C. P., et al.: Diabetes mellitus: incidence, prevalence, survivorship, and causes of death in Rochester, Minnesota, 1945–1970. Diabetes 25:566, 1976.

1987

234. Paolino, E., Granieri, E., Tola, M. R., et al.: Predisposing factors in Bell’s palsy: a case-control study. J. Neurol. 232:363, 1985. 235. Pappagallo, M.: Newer antiepileptic drugs: possible uses in the treatment of neuropathic pain and migraine. Clin. Ther. 25:2506, 2003. 236. Parry, G. J., and Floberg, J.: Diabetic truncal neuropathy presenting as abdominal hernia. Neurology 39:1488, 1989. 237. Pascoe, M. K., Low, P. A., Windebank, A. J., and Litchy, W. J.: Subacute diabetic proximal neuropathy. Mayo Clin. Proc. 72:1123, 1997. 238. Pavy, F. W.: Introductory address to the discussion of the clinical aspect of glycosuria. Lancet 2:1085, 1885. 239. Pecket, P., and Schattner, A.: Concurrent Bell’s palsy and diabetes mellitus: a diabetic mononeuropathy? J. Neurol. Neurosurg. Psychiatry 45:652, 1982. 240. Pedraza, L., Huang, J. K., and Colman, D. R.: Organizing principles of the axoglial apparatus. Neuron 30:335, 2001. 241. Peitersen, E.: Bell’s palsy: the spontaneous course of 2,500 peripheral facial nerve palsies of different etiologies. Acta Otolaryngol. Suppl. (Stockh.) 549:4, 2002. 242. Pekiner, C., Cullum, N. A., Hughes, J. N., et al.: Glycation of brain actin in experimental diabetes. J. Neurochem. 61:436, 1993. 243. Pena, M. M., Donaghue, K. C., Fung, A. T., et al.: The prospective assessment of autonomic nerve function by pupillometry in adolescents with type 1 diabetes mellitus. Diabet. Med. 12:868, 1995. 244. Perkins, B. A., Olaleye, D., and Bril, V.: Carpal tunnel syndrome in patients with diabetic polyneuropathy. Diabetes Care 25:565, 2002. 245. Pfeiffer, A., and Schatz, H.: Diabetic microvascular complications and growth factors. Exp. Clin. Endocrinol. 103:7, 1995. 246. Phillips, R. W., Panepinto, L. M., Spangler, R., and Westmoreland, N.: Yucatan miniature swine as a model for the study of human diabetes mellitus. Diabetes 31(1 Pt. 2):30, 1982. 247. Pirart, J.: Diabetes mellitus and its degenerative complications: a prospective study of 4,400 patients observed between 1947 and 1973. Part 1. Diabetes Care 1:168, 1978. 248. Pirart, J.: Diabetes mellitus and its degenerative complications: a prospective study of 4,400 patients observed between 1947 and 1973. Part 2. Diabetes Care 1:252, 1978. 249. Pirart, J., Lauvaux, J. P., and Rey, W.: Blood sugar and diabetic complications. N. Engl. J. Med. 298:1149, 1978. 250. Poduslo, J., Curran, G. L., and Dyck, P. J.: Increase in albumin, IgG and IgM blood nerve barrier indices in human diabetic neuropathy. Proc. Natl. Acad. Sci. U. S. A. 85:4879, 1988. 251. Porcellati, F., Fanelli, C., Bottini, P., et al.: Mechanisms of arterial hypotension after therapeutic dose of subcutaneous insulin in diabetic autonomic neuropathy. Diabetes 42:1055, 1993. 252. Pryce, T. D.: On diabetic neuritis, with a clinical and pathological description of three cases of diabetic pseudo-tabes. Brain 16:416, 1893. 253. Purves, T. D., Middlemas, A., Agthong, S., et al.: A role for mitogen-activated protein kinases in the aetiology of diabetic neuropathy. FASEB J. 15:2508, 2001.

1988

Diseases of the Peripheral Nervous System

254. Quadri, R., Ponzani, P., Zanone, M., et al.: Changes in autonomic nervous function over a 5-year period in non-insulin-dependent diabetic patients. Diabet. Med. 10:916, 1993. 255. Rakieten, N., Rakieten, M. L., and Moreshwar, V. N.: Studies on the diabetogenic action of streptozotocin (NSC-37917). Cancer Chemother. Rep. 29:91, 1963. 256. Reichard, P., Pihl, M., Rosenqvist, U., and Sule, J.: Complications in IDDM are caused by elevated blood glucose level: the Stockholm Diabetes Intervention Study (SDIS) at 10-year follow up. Diabetologia 39:1483, 1996. 257. Rerup, C. C.: Drugs producing diabetes through damage of the insulin secreting cells. Pharmacol. Rev. 22:485, 1970. 258. Riaz, S., Malcangio, M., Miller, M., and Tomlinson, D. R.: A vitamin D3 derivative (CB1093) induces nerve growth factor and prevents neurotrophic deficits in streptozotocindiabetic rats. Diabetologia 42:1308, 1999. 259. Riaz, S. S., and Tomlinson, D. R.: Neurotrophic factors in peripheral neuropathies: pharmacological strategies. Prog. Neurobiol. 49:125, 1996. 260. Riaz, S. S., and Tomlinson, D. R.: Clenbuterol stimulates neurotrophic support in streptozotocin-diabetic rats. Diabet. Obes. Metab. 1:43, 1999. 261. Richards, B. W., Jones, F. R., and Younge, B. R.: Causes and prognosis in 4,278 cases of paralysis of the oculomotor, trochlear, and abducens cranial nerves. Am. J. Ophthalmol. 113:489, 1992. 262. Robinson, J. P., Willars, G. B., Tomlinson, D. R., and Keen, P.: Axonal transport and tissue contents of substance P in rats with long-term streptozotocin-diabetes: effects of the aldose reductase inhibitor ‘Statil’. Brain Res. 426:339, 1987. 263. Ross, A. T.: Recurrent cranial nerve palsies in diabetes mellitus. Neurology 12:180, 1962. 264. Rucker, C. W.: The causes of paralysis of the third, fourth and sixth cranial nerves. Am. J. Ophthalmol. 61:1293, 1966. 265. Rundles, R. W.: Diabetic neuropathy: general review with report of 125 cases. Medicine (Baltimore) 24:111, 1945. 266. Rush, J. A., and Younge, B. R.: Paralysis of cranial nerves III, IV, and VI: cause and prognosis in 1,000 cases. Arch. Ophthalmol. 99:76, 1981. 267. Russell, C. O., Gannan, R., Coatsworth, J., et al.: Relationship among esophageal dysfunction, diabetic gastroenteropathy, and peripheral neuropathy. Dig. Dis. Sci. 28:289, 1983. 268. Russo, A., Sun, W. M., Sattawatthamrong, Y., et al.: Acute hyperglycaemia affects anorectal motor and sensory function in normal subjects. Gut 41:494, 1997. 269. Ryle, C., and Donaghy, M.: Non-enzymatic glycation of peripheral nerve proteins in human diabetics. J. Neurol. Sci. 129:62, 1995. 270. Said, G.: Diabetic neuropathy: an update. J. Neurol. 243:431, 1996. 271. Said, G., Bigo, A., Ameri, A., et al.: Uncommon early-onset neuropathy in diabetic patients. J. Neurol. 245:61, 1998. 272. Said, G., Goulon-Goeau, C., Lacroix, C., and Moulonguet, A.: Nerve biopsy findings in different patterns of proximal diabetic neuropathy. Ann. Neurol. 35:559, 1994. 273. Said, G., Goulon-Goeau, C., Slama, G., and Tchobroutsky, G.: Severe early-onset polyneuropathy in insulin-dependent diabetes mellitus: a clinical and pathological study. N. Engl. J. Med. 326:1257, 1992.

274. Said, G., Lacroix, C., Lozeron, P., et al.: Inflammatory vasculopathy in multifocal diabetic neuropathy. Brain 126:376, 2003. 275. Said, G., Slama, G., and Selva, J.: Progressive centripetal degeneration of axons in small fibre diabetic polyneuropathy. Brain 106(Pt. 4):791, 1983. 276. Sampson, M. J., Wilson, S., Karagiannis, P., et al.: Progression of diabetic autonomic neuropathy over a decade in insulindependent diabetics. Q. J. Med. 75:635, 1990. 277. Schmidt, A. M., Hori, O., Brett, J., et al.: Cellular receptors for advanced glycation end products: implications for induction of oxidant stress and cellular dysfunction in the pathogenesis of vascular lesions. Arterioscler. Thromb. 14:1521, 1994. 278. Schmidt, R. E.: Neuropathology of human sympathetic autonomic ganglia. Microsc. Res. Tech. 35:107, 1996. 279. Schmidt, R. E.: Age-related sympathetic ganglionic neuropathology: human pathology and animal models. Auton. Neurosci. 96:63, 2002. 280. Schmidt, R. E., Beaudet, L. N., Plurad, S. B., and Dorsey, D. A.: Axonal cytoskeletal pathology in aged and diabetic human sympathetic autonomic ganglia. Brain Res. 769:375, 1997. 281. Schmidt, R. E., Dorsey, D., Parvin, C. A., et al.: Dystrophic axonal swellings develop as a function of age and diabetes in human dorsal root ganglia. J. Neuropathol. Exp. Neurol. 56:1028, 1997. 282. Schmidt, R. E., Dorsey, D. A., Beaudet, L. N., et al.: Effect of aminoguanidine on the frequency of neuroaxonal dystrophy in the superior mesenteric sympathetic autonomic ganglia of rats with streptozocin-induced diabetes. Diabetes 45:284, 1996. 283. Schvarcz, E., Palmer, M., Ingberg, C. M., et al.: Increased prevalence of upper gastrointestinal symptoms in long-term type 1 diabetes mellitus. Diabet. Med. 13:478, 1996. 284. Schwartz, J. G., Green, G. M., Guan, D., et al.: Rapid gastric emptying of a solid pancake meal in type II diabetic patients. Diabetes Care 19:468, 1996. 285. Seneviratne, K. N., and Peiris, O. A.: The effect of ischaemia on the excitability of sensory nerves in diabetes mellitus. J. Neurol. Neurosurg. Psychiatry 31:348, 1968. 286. Seneviratne, K. N., and Peiris, O. A.: The effects of hypoxia on the excitability of the isolated peripheral nerves of alloxandiabetic rats. J. Neurol. Neurosurg. Psychiatry 32:462, 1969. 287. Shafrir, E., and Renold, A. E.: Frontiers in Diabetes Research: Lessons from Animal Diabetes 2. London, J. Libbey, 1988. 288. Sharma, K. R., Cross, J., Ayyar, D. R., et al.: Diabetic demyelinating polyneuropathy responsive to intravenous immunoglobulin therapy. Arch. Neurol. 59:751, 2002. 289. Sharma, K. R., Cross, J., Farronay, O., et al.: Demyelinating neuropathy in diabetes mellitus. Arch. Neurol. 59:758, 2002. 290. Shaw, J. E., Hodge, A. M., Dowse, G. K., et al.: Diabetic neuropathy in Mauritius: prevalence and risk factors. Diabetes Res. Clin. Pract. 42:131, 1998. 291. Shaw, J. E., Parker, R., Hollis, S., et al.: Gustatory sweating in diabetes mellitus. Diabet. Med. 13:1033, 1996. 292. Shaw, J. E., Vileikyte, L., Connor, H., and Boulton, A. J.: The diabetic foot 1994. Diabet. Med. 12:88, 1995. 293. Sima, A. A. F., Lattimer, S. A., Yagihashi, S., and Greene, D. A.: Axo-glial dysjunction: a novel structural lesion that accounts

Diabetic Neuropathies

294.

295.

296.

297.

298.

299.

300.

301.

302.

303.

304.

305.

306.

307. 308.

309.

for poorly reversible slowing of nerve conduction in the spontaneously diabetic Bio-Breeding rat. J. Clin. Invest. 77:474, 1986. Sima, A. A. F., Prashar, A., Zhang, W. X., et al.: Preventive effect of long-term aldose reductase inhibition (ponalrestat) on nerve conduction and sural nerve structure in the spontaneously diabetic Bio-Breeding rat. J. Clin. Invest. 85:1410, 1990. Sindrup, S. H., Bach, F. W., Madsen, C., et al.: Venlafaxine versus imipramine in painful polyneuropathy: a randomized, controlled trial. Neurology 60:1284, 2003. Sindrup, S. H., Bjerre, U., Dejgaard, A., et al.: The selective serotonin reuptake inhibitor citalopram relieves the symptoms of diabetic neuropathy. Clin. Pharmacol. Ther. 52:547, 1992. Singleton, J. R., Smith, A. G., and Bromberg, M. B.: Painful sensory polyneuropathy associated with impaired glucose tolerance. Muscle Nerve 24:1225, 2001. Smith, A. G., Ramachandran, P., Tripp, S., and Singleton, J. R.: Epidermal nerve innervation in impaired glucose tolerance and diabetes-associated neuropathy. Neurology 57:1701, 2001. Smith, S. A., and Smith, S. E.: Reduced pupillary light reflexes in diabetic autonomic neuropathy. Diabetologia 24:330, 1983. Sobotka, P. A., Liss, H. P., and Vinik, A. I.: Impaired hypoxic ventilatory drive in diabetic patients with autonomic neuropathy. J. Clin. Endocrinol. Metab. 62:658, 1986. Solders, G., Thalme, B., Aguirre-Aquino, M., et al.: Nerve conduction and autonomic nerve function in diabetic children: a 10-year follow-up study. Acta Paediatr. 86:361, 1997. Spallone, V., Gambardella, S., Maiello, M. R., et al.: Relationship between autonomic neuropathy, 24-h blood pressure profile, and nephropathy in normotensive IDDM patients. Diabetes Care 17:578, 1994. Steel, J. M., Young, R. J., Lloyd, G. G., and Clarke, B. F.: Clinically apparent eating disorders in young diabetic women: associations with painful neuropathy and other complications. Br. Med. J. (Clin. Res. Ed.) 294:859, 1987. Steiness, I. B.: Vibratory perception in diabetes during arrested blood flow to the limb. Acta Med. Scand. 163:195, 1959. Stephenson, J. M., and Fuller, J. H.: Microalbuminuria is not rare before 5 years of IDDM. EURODIAB IDDM Complications Study Group and the WHO Multinational Study of Vascular Disease in Diabetes Study Group. J. Diabetes Complications 8:166, 1994. Stevens, J. C., Beard, C. M., O’Fallon, W. M., and Kurland, L. T.: Conditions associated with carpal tunnel syndrome. Mayo Clin. Proc. 67:541, 1992. Stewart, J. D.: Diabetic truncal neuropathy: topography of the sensory deficit. Ann. Neurol. 25:233, 1989. Stewart, J. D., McKelvey, R., Durcan, L., et al.: Chronic inflammatory demyelinating polyneuropathy (CIDP) in diabetics. J. Neurol. Sci. 142:59, 1996. Straub, R. H., Zietz, B., Palitzsch, K. D., and Scholmerich, J.: Impact of disease duration on cardiovascular and pupillary autonomic nervous function in IDDM and NIDDM patients. Diabetes Care 19:960, 1996.

1989

310. Stump, K. C., Swindle, M. M., Saudek, C. D., and Strandberg, J. D.: Pancreatectomized swine as a model of diabetes mellitus. Lab. Anim. Sci. 38:439, 1988. 311. Sugimoto, K., Nishizawa, Y., Horiuchi, S., and Yagihashi, S.: Localization in human diabetic peripheral nerve of Necarboxymethyllysine-protein adducts, an advanced glycation endproduct. Diabetologia 40:1380, 1997. 312. Sumner, C. J., Sheth, S., Griffin, J. W., et al.: The spectrum of neuropathy in diabetes and impaired glucose tolerance. Neurology 60:108, 2003. 313. Sundkvist, G., and Lilja, B.: Autonomic neuropathy predicts deterioration in glomerular filtration rate in patients with IDDM. Diabetes Care 16:773, 1993. 314. Suzuki, Y. J., Tsuchiya, M., and Packer, L.: Thioctic acid and dihydrolipoic acid are novel antioxidants which interact with reactive oxygen species. Free Radic. Res. Commun. 15:255, 1991. 315. Tackmann, W., Kaeser, H. E., Berger, W., et al.: Autonomic disturbances in relation to sensorimotor peripheral neuropathy in diabetes mellitus. J. Neurol. 224:273, 1981. 316. Terkildsen, A. B., and Christensen, N. J.: Reversible nervous abnormalities in juvenile diabetics with recently diagnosed diabetes. Diabetologia 7:113, 1971. 317. Tesfaye, S., Malik, R., Harris, N., et al.: Arterio-venous shunting and proliferating new vessels in acute painful neuropathy of rapid glycaemic control (insulin neuritis). Diabetologia 39:329, 1996. 318. Tesfaye, S., Stevens, L. K., Stephenson, J. M., et al.: Prevalence of diabetic peripheral neuropathy and its relation to glycaemic control and potential risk factors: the EURODIAB IDDM Complications Study. Diabetologia 39:1377, 1996. 319. The Capsaicin Study Group: Treatment of painful diabetic neuropathy with topical capsaicin: a multicenter, doubleblind, vehicle-controlled study. Arch. Intern. Med. 151:2225, 1991. 320. Theriault, M., Dort, J., Sutherland, G., and Zochodne, D. W.: Local human sural nerve blood flow in diabetic and other polyneuropathies. Brain 120(Pt. 7):1131, 1997. 321. Thomas, P. K.: Metabolic neuropathy. J. R. Coll. Physicians Lond. 7:154, 1973. 322. Thornalley, P. J., Yurek-George, A., and Argirov, O. K.: Kinetics and mechanism of the reaction of aminoguanidine with the alpha-oxoaldehydes glyoxal, methylglyoxal, and 3-deoxyglucosone under physiological conditions. Biochem. Pharmacol. 60:55, 2000. 323. Tilton, R. G., Chang, K., Hasan, K. S., et al.: Prevention of diabetic vascular dysfunction by guanidines: inhibition of nitric oxide synthase versus advanced glycation end-product formation. Diabetes 42:221, 1993. 324. Timperley, W. R., Boulton, A. J., Davies-Jones, G. A., et al.: Small vessel disease in progressive diabetic neuropathy associated with good metabolic control. J. Clin. Pathol. 38:1030, 1985. 325. Tintore, M., Montalban, J., Cervera, C., et al.: Peripheral neuropathy in association with insulinoma: clinical features and neuropathology of a new case. J. Neurol. Neurosurg. Psychiatry 57:1009, 1994. 326. Tomlinson, D. R.: Polyols and myo-inositol in diabetic neuropathy: of mice and men. Mayo Clin. Proc. 64:1030, 1989.

1990

Diseases of the Peripheral Nervous System

327. Tomlinson, D. R.: Aldose reductase inhibitors and the complications of diabetes mellitus. Diabet. Med. 10:214, 1993. 328. Tomlinson, D. R.: Mitogen-activated protein kinases as glucose transducers for diabetic complications. Diabetologia 42:1271, 1999. 329. Tomlinson, D. R., Holmes, P. R., and Mayer, J. H.: Reversal, by treatment with an aldose reductase inhibitor, of impaired axonal transport and motor nerve conduction velocity in experimental diabetes mellitus. Neurosci. Lett. 31:189, 1982. 330. Tomlinson, D. R., and Mayer, J. H.: Defects of axonal transport in diabetes mellitus—a possible contribution to the aetiology of diabetic neuropathy. J. Auton. Pharmacol. 4:59, 1984. 331. Tomlinson, D. R., Moriarty, R. J., and Mayer, J. H.: Prevention and reversal of defective axonal transport and motor nerve conduction velocity in rats with experimental diabetes by treatment with the aldose reductase inhibitor sorbinil. Diabetes 33:470, 1984. 332. Tomlinson, D. R., Riaz, S., and Stevens, E. J.: Dependence of sciatic nerve blood flow on hind-limb muscle mass: an explanation for the deficits in nerve perfusion in experimental diabetes? Diabet. Nutr. Metab. 9:258, 1996. 333. Tomlinson, D. R., Robinson, J. P., Willars, G. B., and Keen, P.: Deficient axonal transport of substance P in streptozocininduced diabetic rats: effects of sorbinil and insulin. Diabetes 37:488, 1988. 334. Toyry, J. P., Niskanen, L. K., Mantysaari, M. J., et al.: Occurrence, predictors, and clinical significance of autonomic neuropathy in NIDDM: ten-year follow-up from the diagnosis. Diabetes 45:308, 1996. 335. Treisman, R.: Regulation of transcription by MAP kinase cascades. Curr. Opin. Cell Biol. 8:205, 1996. 336. Tsigos, C., Diemel, L. T., White, A., et al.: Cerebrospinal fluid levels of substance P and calcitonin gene-related peptide: correlation with sural nerve levels and neuropathic signs in sensory diabetic neuropathy. Clin. Sci. (Colch.) 84:305, 1993. 337. Tuck, R. R., Schmelzer, J. D., and Low, P. A.: Endoneurial blood flow and oxygen tension in the sciatic nerves of rats with experimental diabetic neuropathy. Brain 107:935, 1984. 338. UK Prospective Diabetes Study Group: Intensive blood glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type II diabetes (UK PDS 33). Lancet 352:837, 1998. 339. Uncini, A., De Angelis, M. V., Di Muzio, A., et al.: Chronic inflammatory demyelinating polyneuropathy in diabetics: motor conductions are important in the differential diagnosis with diabetic polyneuropathy. Clin. Neurophysiol. 110:705, 1999. 340. Veneman, T., Mitrakou, A., Mokan, M., et al.: Induction of hypoglycemia unawareness by asymptomatic nocturnal hypoglycemia. Diabetes 42:1233, 1993. 341. Veves, A., Malik, R. A., Lye, R. H., et al.: The relationship between sural nerve morphometric findings and measures of peripheral nerve function in mild diabetic neuropathy. Diabet. Med. 8:917, 1991. 342. Vianna, L. G., Gilbey, S. G., Barnes, N. C., et al.: Cough threshold to citric acid in diabetic patients with and without autonomic neuropathy. Thorax 43:569, 1988.

343. Vital, C., Le Blanc, M., Vallat, J. M., et al.: [Ultrastructural study of peripheral nerve in 16 diabetics without neuropathy: comparisons with 16 diabetic neuropathies and 16 nondiabetic neuropathies.] Acta Neuropathol. (Berl.) 30:63, 1974. 344. Vital, C., Vallat, J. M., Le Blanc, M., et al.: [Peripheral neuropathies caused by diabetes mellitus: ultrastructural study of 12 biopsied cases.] J. Neurol. Sci. 18:381, 1973. 345. Vital, C., Vital, A., Dupon, M., et al.: Acute painful diabetic neuropathy: two patients with recent insulin-dependent diabetes mellitus. J. Peripher. Nerv. Syst. 2:151, 1997. 346. Vlassara, H., Brownlee, M., and Cerami, A.: Nonenzymatic glycosylation of peripheral nerve protein in diabetes mellitus. Proc. Natl. Acad. Sci. U. S. A. 78:5190, 1981. 347. Voorby, H. A. M., Van der Gaag, R. D., Jeucken, P. H. M., et al.: The goitre of the BB/O rat: an animal-model for studying the role of immunoglobulins stimulating growth of thyroid cells. Clin. Exp. Immunol. 76:290, 1989. 348. Wada, R., Nishizawa, Y., Yagihashi, N., et al.: Effects of OPB-9195, anti-glycation agent, on experimental diabetic neuropathy. Eur. J. Clin. Invest. 31:513, 2001. 349. Wake, N., Hisashige, A., Katayama, T., et al.: Cost-effectiveness of intensive insulin therapy for type 2 diabetes: a 10-year follow-up of the Kumamoto study. Diabetes Res. Clin. Pract. 48:201, 2000. 350. Ward, J. D., Barnes, C. G., Fisher, D. J., et al.: Improvement in nerve conduction following treatment in newly diagnosed diabetics. Lancet 1:428, 1971. 351. Watkins, P. J.: The diabetic foot. BMJ 326:977, 2003. 352. Watkins, P. J., and Edmonds, M. E.: Sympathetic nerve failure in diabetes. Diabetologia 25:73, 1983. 353. Watkins, P. J., and Thomas, P. K.: Diabetes mellitus and the nervous system. J. Neurol. Neurosurg. Psychiatry, 65:620, 1998. 354. Waxman, S. G., and Ritchie, J. M.: Organization of ion channels in the myelinated nerve fiber. Science 228:1502, 1985. 355. Way, K. J., Katai, N., and King, G. L.: Protein kinase C and the development of diabetic vascular complications. Diabet. Med. 18:945, 2001. 356. Weeks, R. A., Thomas, P. K., and Gale, A. N.: Abdominal pseudohernia caused by diabetic truncal radiculoneuropathy. J. Neurol. Neurosurg. Psychiatry 66:405, 1999. 357. Whiteley, S. J., and Tomlinson, D. R.: Motor nerve conduction velocity and nerve polyols in mice with short-term genetic or streptozotocin-induced diabetes. Exp. Neurol. 89:314, 1985. 358. Willars, G. B., Townsend, J., Tomlinson, D. R., et al.: Studies on peripheral nerve and lens in long-term experimental diabetes: effects of the aldose reductase inhibitor statil. Metabolism 37:442, 1988. 359. Williams, E., Timperley, W. R., Ward, J. D., and Duckworth, T.: Electron microscopical studies of vessels in diabetic peripheral neuropathy. J. Clin. Pathol. 33:462, 1980. 360. Wilson, J. D., Dhall, D. P., Simeonovic, C. J., and Lafferty, K. J.: Induction and management of diabetes mellitus in the pig. Aust. J. Exp. Biol. Med. Sci. 64:489, 1986. 361. Wilson, J. L., Sokol, D. K., Smith, L. H., et al.: Acute painful neuropathy (insulin neuritis) in a boy following rapid glycemic

Diabetic Neuropathies

362.

363.

364.

365.

366.

367.

368.

369.

370.

371.

372.

373.

374.

375.

control for type 1 diabetes mellitus. J. Child Neurol. 18:365, 2003. Wohaieb, S. A., and Godin, D. V.: Alterations in free radical tissue-defense mechanisms in streptozocin-induced diabetes in rat: effects of insulin treatment. Diabetes 36:1014, 1987. Wolff, S. P., Jiang, Z. Y., and Hunt, J. V.: Protein glycation and oxidative stress in diabetes mellitus and ageing. Free Radic. Biol. Med. 10:339, 1991. Woltman, H. W., and Wilder, R. M.: Diabetes mellitus: pathological changes in the spinal cord and peripheral nerves. Arch. Intern. Med. 44:576, 1929. Xia, P., Inoguchi, T., Kern, T. S., et al.: Characterization of the mechanism for the chronic activation of diacylglycerolprotein kinase C pathway in diabetes and hypergalactosemia. Diabetes 43:1122, 1994. Yagihashi, S., Kamijo, M., Baba, M., et al.: Effect of aminoguanidine on functional and structural abnormalities in peripheral nerve of STZ-induced diabetic rats. Diabetes 41:47, 1992. Yagihashi, S., Kamijo, M., and Watanabe, K.: Reduced myelinated fiber size correlates with loss of axonal neurofilaments in peripheral nerve of chronically streptozotocin diabetic rats. Am. J. Pathol. 136:1365, 1990. Yagihashi, S., and Matsunaga, M.: Ultrastructural pathology of peripheral nerves in patients with diabetic neuropathy. Tohoku J. Exp. Med. 129:358, 1979. Yan, S. D., Schmidt, A. M., Anderson, G. M., et al.: Enhanced cellular oxidant stress by the interaction of advanced glycation end products with their receptors/binding proteins. J. Biol. Chem. 269:9889, 1994. Yasuda, H., and Dyck, P. J.: Abnormalities of endoneurial microvessels and sural nerve pathology in diabetic neuropathy. Neurology 37:20, 1987. Yasuda, H., Taniguchi, Y., Huitian, Z., et al.: Chronically streptozocin-diabetic monkey does not closely mimic human diabetic neuropathy. Exp. Neurol. 104:133, 1989. Yeh, L. A., and Ashton, M. A.: The increase in lipid peroxidation in diabetic rat lens can be reversed by oral sorbinil. Metabolism 39:619, 1990. Yki-Jarvinen, H., and Utriainen, T.: Insulin-induced vasodilatation: physiology or pharmacology? Diabetologia 41:369, 1998. Young, M. J., Breddy, J. L., Veves, A., and Boulton, A. J.: The prediction of diabetic neuropathic foot ulceration using vibration perception thresholds: a prospective study. Diabetes Care 17:557, 1994. Yu, P. H., and Zuo, D. M.: Aminoguanidine inhibits semicarbazide-sensitive amine oxidase activity: implications for advanced glycation and diabetic complications. Diabetologia 40:1243, 1997.

1991

376. Yue, D. K., Hanwell, M. A., Satchell, P. M., and Turtle, J. R.: The effect of aldose reductase inhibition on motor nerve conduction velocity in diabetic rats. Diabetes 31:789, 1982. 377. Yuen, K. C., Day, J. L., Flannagan, D. W., and Rayman, G.: Diabetic neuropathic cachexia and acute bilateral cataract formation following rapid glycaemic control in a newly diagnosed type 1 diabetic patient. Diabet. Med. 18:854, 2001. 378. Ziegler, D., Dannehl, K., Volksw, D., et al.: Prevalence of cardiovascular autonomic dysfunction assessed by spectral analysis and standard tests of heart-rate variation in newly diagnosed IDDM patients. Diabetes Care 15:908, 1992. 379. Ziegler, D., Dannehl, K., Wiefels, K., and Gries, F. A.: Differential effects of near-normoglycaemia for 4 years on somatic nerve dysfunction and heart rate variation in type 1 diabetic patients. Diabet. Med. 9:622, 1992. 380. Ziegler, D., Gries, F. A., Spuler, M., and Lessmann, F.: The epidemiology of diabetic neuropathy. DiaCAN Multicenter Study Group. Diabet. Med. 10(Suppl. 2):82S, 1993. 381. Ziegler, D., Hanefeld, M., Ruhnau, K. J., et al.: Treatment of symptomatic diabetic peripheral neuropathy with the antioxidant alpha-lipoic acid: a 3-week multicentre randomized controlled trial (ALADIN Study). Diabetologia 38:1425, 1995. 382. Ziegler, D., Hanefeld, M., Ruhnau, K. J., et al.: Treatment of symptomatic diabetic polyneuropathy with the antioxidant ␣-lipoic acid—a 7-month multicenter randomized controlled trial (ALADIN III Study). Diabetes Care 22:1296, 1999. 383. Ziegler, D., Mayer, P., Muhlen, H., and Gries, F. A.: The natural history of somatosensory and autonomic nerve dysfunction in relation to glycaemic control during the first 5 years after diagnosis of type 1 (insulin-dependent) diabetes mellitus. Diabetologia 34:822, 1991. 384. Ziegler, D., Reljanovic, M., Mehnert, H., and Gries, F. A.: Alpha-lipoic acid in the treatment of diabetic polyneuropathy in Germany: current evidence from clinical trials. Exp. Clin. Endocrinol. Diabetes 107:421, 1999. 385. Ziegler, D., Schadewaldt, P., Pour, M. A., et al.: [13C]octanoic acid breath test for non-invasive assessment of gastric emptying in diabetic patients: validation and relationship to gastric symptoms and cardiovascular autonomic function. Diabetologia 39:823, 1996. 386. Zola, B., Kahn, J. K., Juni, J. E., and Vinik, A. I.: Abnormal cardiac function in diabetic patients with autonomic neuropathy in the absence of ischemic heart disease. J. Clin. Endocrinol. Metab. 63:208, 1986. 387. Zorilla, E., and Kozak, G. P.: Ophthalmoplegia in diabetes mellitus. Ann. Intern. Med. 67:968, 1967.

86 Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties P. JAMES B. DYCK

Overview Historical Review Clinical Features of Radiculoplexus Neuropathies Diabetic LRPN Nondiabetic LRPN

Other Forms of Radiculoplexus Neuropathies Clinical Testing in DLRPN and LRPN Pathology of Radiculoplexus Neuropathies Diabetic LRPN

OVERVIEW Separate clinical and pathologic varieties of neuropathy are associated with diabetes mellitus. Distinguishing features among the different subtypes of diabetic neuropathies are the distribution and degree of symmetry of the disorders. Diabetic sensorimotor polyneuropathy is a distal symmetrical condition with an insidious onset with predominantly distal symptoms and findings. In contrast, the diabetic radiculoplexus neuropathies (DRPNs) (discussed in this chapter) present with sudden or subacute onset of pain and weakness in an asymmetrical pattern, and proximal segments are frequently involved. The same clinical phenotypes of radiculoplexus neuropathies also occur in nondiabetic patients, although they may be less frequent. DRPN can be divided into three subtypes: diabetic lumbosacral radiculoplexus neuropathy (DLRPN) (a lower limb syndrome), diabetic thoracic radiculoneuropathy (DTRN) (a truncal syndrome), and diabetic cervical radiculoplexus neuropathy (DCRPN) (an upper limb syndrome). Similar syndromes occur in nondiabetic patients and can be classified in the same way. DLRPN is probably the most common subtype and often will present alone or in combination with another of the subtypes of DRPN. Because it is the most common and disabling of the subtypes, it is reviewed here in the most detail. Over the years, there has been some debate as to the underlying etiology of the

Nondiabetic LRPN Other Forms of Radiculoplexus Neuropathy Immunotherapy Symptomatic Treatment

different types of DRPN, particularly of DLRPN. Consequently, this condition has been known by multiple different names, including Bruns-Garland syndrome,6,13 diabetic myelopathy,31,33 diabetic mononeuropathy multiplex,47,48 diabetic polyradiculopathy,7 femoral or femoralsciatic neuropathy of diabetics,11,54 diabetic motor or paralytic neuropathy,9,41,42 proximal diabetic neuropathy,3,64 diabetic lumbosacral plexopathy,8 and the name preferred by the author, DLRPN.24 The many different names of this disorder demonstrate the diverse opinion concerning anatomic location and underlying cause. It probably does make sense to group DLRPN, DTRN, and DCRPN together in the category of DRPN because they have a similar clinical course. However, it is not clear that cranial neuropathies should be included with DRPN; in my experience, patients who have one variety of radiculoplexus neuropathy often eventually develop other varieties, including cranial neuropathy (e.g., third nerve palsy). Radiculoplexus neuropathies occur in nondiabetic as well as diabetic patients. The form that is most recognized is the cervical radiculoplexus neuropathies (brachial plexopathies), which occur frequently in nondiabetic patients. This condition is also known as Parsonage-Turner syndrome and neuralgic amyotrophy. Because the nondiabetic brachial plexopathies are discussed elsewhere in the book, they are only mentioned briefly here. Another variety of radiculoplexus neuropathy that occurs in nondiabetic 1993

1994

Diseases of the Peripheral Nervous System

patients is nondiabetic lumbosacral radiculoplexus neuropathy (LRPN); it is a relatively recently recognized condition that has not been studied in great detail.30,52 However, the clinical features appear to be very similar to the diabetic variety.26 Thoracic radiculopathies also occur in nondiabetic patients. Much remains to be learned about the diabetic and the nondiabetic radiculoplexus neuropathy syndromes. It seems best to group these conditions together under the heading of radiculoplexus neuropathies because of the patchy involvement in these disorders of roots, plexuses, and peripheral nerves based on abnormalities observed in clinical, electromyographic (EMG), laboratory, and histopathologic study of nerves. In addition, they are best grouped together given the strong clinical similarities of the syndromes: they begin with pain followed by weakness, they are asymmetrical, they are usually monophasic illnesses, often the same patient has several anatomic sites involved, and they are associated with weight loss. This review outlines the clinical, laboratory, electrophysiologic, and pathologic features of the radiculoplexus neuropathies and discusses natural history and treatment. Because these syndromes appear to be caused by a microvasculitis and ischemic injury, immunotherapy may be the best treatment, but this has not yet been demonstrated.

HISTORICAL REVIEW Over the years, DLRPN has gotten more attention than the other varieties of DRPNs. Most authors attribute the original descriptions of DLRPN to Bruns in 1890.9 However, the condition was also recognized by Auche,4 Buzzard,10 and Leyden41,42 at about the same time. In 1887, Leyden reported three types of diabetic “neuritis”: (1) hyperesthetic or neuralgic, (2) paralytic (motor), and (3) ataxic.41 As examples of the paralytic subtype (probably DLRPN), Leyden described cases with intense lower limb pain associated with weakness that improved with time. The syndrome was largely forgotten except for occasional cases49 until Garland and Tavener described it again in 1953, first calling it “diabetic myelopathy”34 because some of the patients had extensor plantar responses and then “diabetic amyotrophy”31,33 because of the uncertainty of the pathoanatomic localization. Garland described the condition as one of pain and weakness involving proximal lower limb muscles with associated weight loss. He emphasized that it had a monophasic course and that it was a motor neuropathy (because the clinical sensory examination was normal). Although he emphasized proximal involvement, some of Garland’s patients also had distal findings.33 Since this time, there has been debate about characteristic clinical features and underlying disease mechanisms: metabolic derangement (hyperglycemia), ischemic injury,

or immune mechanisms. Some have argued that DLRPN is one distinct entity, whereas others have maintained that it is part of a continuum with rapid onset and asymmetrical involvement (from ischemia) being at one extreme and a slowly evolving symmetrical disorder (from metabolic derangement) at the other extreme.3,15,37 Raff and coworkers47,48 emphasized the rapid onset and asymmetrical course and referred to the syndrome as “diabetic mononeuritis multiplex.” They reported the only detailed study of a case at autopsy, and described nerve infarcts and occluded blood vessels and concluded that these findings were due to nerve ischemia. Although they observed inflammatory infiltrates in the nerve, they judged them to be reactive. By contrast, Chokroverty and colleagues12–14 emphasized the symmetrical, bilateral, and slowly progressive nature of the disorder. They (like Garland) wrote that it was predominantly a motor neuropathy. Based on their data, they concluded that metabolic factors (hyperglycemia) and not ischemic injury best explained the disease pathophysiology. Both groups emphasized the proximal lower limb involvement. Other groups, however, have disagreed that it is predominantly proximal and have stressed that the disease is multifocal and widespread and can involve distal regions.6,7,24,37 Bastron and Thomas were perhaps the first to associate the different anatomic localizations together into one syndrome and called the condition “diabetic polyradiculopathy”; they noted that frequently there were concomitant truncal neuropathies in addition to the lower limb syndrome.7 They also demonstrated the widespread nature of the denervation and that the paraspinal muscles frequently showed fibrillation potentials. Barohn and coworkers6 reported that the condition began unilaterally and locally but evolved into a multifocal and widespread disorder involving proximal and distal lower limb segments. Recently, Said and co-workers have also observed that different anatomic regions can be affected and have called the syndrome multifocal diabetic neuropathy.51 The controversy about the underlying mechanism of DLRPN has continued, and various authors have proposed that different subtypes are due to different mechanisms. Bradley and associates8 argued that the causes of lumbosacral plexopathies were diverse and that those with elevated sedimentation rates were probably due to immune mechanisms. They found in six patients (three with diabetes mellitus) who had painful lumbosacral plexopathy and elevated sedimentation rates that there was multifocal fiber loss and perivascular inflammatory cell cuffing on nerve pathology. Said and co-workers50 agreed with earlier investigators that there were two subtypes of DLRPN. In their series, ischemic injury and vasculitis predominated in severe cases, whereas metabolic factors and demyelination were more important in mild cases. Barohn and co-workers6 argued that the cause of DLRPN was probably damage from ischemia because they found multifocal fiber degeneration; they did not comment about

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties

inflammatory infiltrates, and they did not find evidence of different subtypes. Krendel and co-workers40 suggested that some cases may be due to inflammatory demyelination and may be similar to chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). More recent authors and we have suggested that at least some cases are due to immune causes, including vasculitis.22,24,39,43 A similar subacute painful neuropathy that occurs in diabetic patients has been described as diabetic neuropathic cachexia.28 Subsequently, Archer and co-workers wrote about acute painful neuropathy in diabetes mellitus.2 These conditions may be the same, and have some features in common with DLRPN in that they tend to occur in type 2 diabetic patients, occur subacutely early in the course of diabetes, are monophasic, and are associated with pain and weight loss. Archer and co-workers emphasized that the distinction from DLRPN is that their patients had sensory symptoms without weakness. In contrast, some patients described as having “diabetic neuropathic cachexia” had weakness. It seems likely that painful diabetic neuropathy with weight loss and diabetic cachexia are part of the spectrum of DLRPN, but because the underlying pathology of these disorders has not been well studied, they should be classified as separate disorders until the pathologic alterations of acute painful diabetic neuropathy are better understood. LRPN, a clinical syndrome very similar to DLRPN, occurs in nondiabetic patients.26 Much less has been written about it because it probably occurs much less commonly than the diabetic variety (although its frequency is now still unknown). It was first described in 1981 when two groups recognized a subacute, painful, paralytic lower limb neuropathy that involved the lumbosacral plexus.30,52 Evans and colleagues reported that the condition was monophasic but that the morbidity was prolonged because of pain and weakness.30 The clinical, laboratory, electrophysiologic, and pathologic features of LRPN are much like those of DLRPN.26 The truncal radiculopathies of diabetes (DTRN) have been noted for a long time. Auche4 in 1890 described 11 diabetic patients with symptomatic neuropathy, one of whom was a 35-year-old diabetic man who developed pain in the region of his kidneys that radiated around his body. In the past 30 years, there have been many small series describing patients with DTRN.29,53,58,61 Bastron and Thomas emphasized that abdominal pain and weakness occurred frequently in patients who had lower limb pain and weakness and consequently classified DLRPN and DTRN into one syndrome that they called diabetic polyradiculopathy.7 Truncal radiculopathies have also been noted in nondiabetic patients, and they also often occur in patients who have LRPN (the lower limb syndrome).26 Sudden onset of pain and weakness involving the upper limbs of a diabetic patient (DCRPN) reminiscent of Parsonage-Turner syndrome or immune brachial plexopathy

1995

is the least well recognized form of DRPN. However, cases of DCRPN have been described for over 100 years. In 1890, Auche4 described two cases of DCRPN. One was a 48-year-old man who awakened one night with pain in his shoulder radiating to the elbow. The pain disappeared in 3 weeks but the weakness persisted for a long time. About the same time, Althaus reported a 56-year-old diabetic man who was awakened by sudden burning in his shoulder, followed by an inability to raise his arm.1 Since this time, there has been controversy as to whether the association of diabetes mellitus and the upper limb neurologic syndrome is causally related.32,62 It has been noted that cases of DCRPN and DLRPN can occur in the same patient.63 In a prospectively identified series of 33 consecutively seen DLRPN patients, we noted that one third of patients (n ⫽ 11) had problematic symptoms and findings of upper limb.24 Of these, eight had mononeuropathy and three (or about 10%) had DCRPN. Katz and colleagues also reviewed their experience in patients with DLRPN and found that many had coexisting DCRPN.38 The evidence from these studies and the observation that often several forms of radiculoplexus neuropathies occur in the same patients makes a chance association between diabetes mellitus and the upper limb syndrome seem unlikely.

CLINICAL FEATURES OF RADICULOPLEXUS NEUROPATHIES Diabetic LRPN Typically DLRPN develops in patients with type 2 diabetes mellitus in middle or old age. There is often a large associated weight loss. In a series of prospectively studied DLRPN patients, we found that 28 of 33 patients lost more than 10 pounds.24 There are usually no precipitating factors identified as causing DLRPN. However, some patients seem to induce the condition by an overly aggressive program of weight loss, exercise, and blood sugar normalization. Others develop it in a postoperative period. Patients’ glycemic dysregulation tended to be mild, and they usually had not been diabetic for a long period of time. In a minority of cases, diabetes mellitus was diagnosed at the time of onset of the neurologic disease (in our series, in 7 of 33 patients). The clinical syndrome of DLRPN is easily distinguished from diabetic polyneuropathy. The symptoms of DLRPN usually begin acutely or subacutely. Patients often can recall the exact day the condition began. In contrast, diabetic polyneuropathy begins insidiously. Furthermore, DLRPN usually begins focally and unilaterally and involves proximal or distal segments, whereas diabetic polyneuropathy begins and remains symmetrical and predominantly involves distal segments. Patients with DLRPN usually have few long-term complications of

1996

Diseases of the Peripheral Nervous System

diabetes mellitus such as retinopathy or nephropathy; in contrast, these complications are common in diabetic polyneuropathy. We compared our cohort of DLRPN patients to a population-based cohort of diabetic patients in Rochester, Minnesota (those from the Rochester Diabetic Neuropathy Study [RDNS]), which is generally representative of community diabetic patients both with and without neuropathy. We found that DLRPN patients had better diabetic control and fewer diabetic complications. Specifically, they had better glycemic control, had a lower body mass index, used less insulin, were more likely to have type 2 diabetes mellitus, had been diabetic for a shorter period of time, and had less retinopathy and less cardiovascular disease than did community diabetic patients (Table 86–1).27 Based on this evidence, it seems unlikely that the pathophysiology of DLRPN is due to the effects of prolonged excess glucose exposure (as in the case of diabetic polyneuropathy). Most commonly, the first symptom of DLRPN is pain. It usually comes in several forms. In almost all cases, pain is described as hurting or deep aching. Other frequently described pains include sharp, lancinating, or electrical shock–like pain, burning or firelike pain, and contact allodynia. With allodynia, patients often do not like to wear clothes or to have bedsheets touch them because these stimuli cause excessive pain. The pain in this disorder may be severe. In our series, essentially all patients received analgesic pain medication and 25 of 33 required narcotics. Most commonly the pain begins asymmetrically (usually unilaterally) and focally, involving proximal segments (the hip and thigh) somewhat more commonly than distal segments (the foot and leg) (Table 86–2). However, the disorder quickly spreads to involve the initially unaffected segments and the contralateral side so that, by the time of evaluation, the disorder is much more widespread and symmetrical than when it first began. Occasionally, it can begin bilaterally and symmetrically. In our series, the syndrome began unilaterally in 29 of 33 patients but eventually became bilateral 32 of 33.24 The median time to bilateral disease was 3 months (see Table 86–1). Consequently, early in the disease course, one segment (back, buttock, hip, thigh, leg, or toe) is often primarily involved. These initially involved segments can be proximal or distal. For example, it is not uncommon to find isolated weak thigh muscles and absent knee reflexes in one patient and footdrop and no proximal findings in another. The process is multifocal, so a patient may have proximal involvement of one lower limb and distal involvement of the other limb. Most physicians recognize the proximal presentations of the disorder as DLRPN but not the distal presentations (probably because of confusion with diabetic polyneuropathy). However, distal presentations of this condition are common; 12 of the 33 patients in our cohort presented with distal symptoms. Because of this bias, diabetic patients who present with thigh and hip pain and

weakness usually are classified as having DLRPN, whereas those who present with leg symptoms (footdrop) often are categorized as having other diseases. This separation is undoubtedly arbitrary and not pathophysiologically based. Fundamentally, there are no major differences between diabetic patients who develop proximal symptoms and those who develop distal symptoms except for the anatomic level of the nerve segments involved. The observation that over time the lower limb segment initially unaffected usually becomes involved supports the view that the proximal and distal syndromes should be categorized together. Because the Mayo Clinic is a tertiary referral center where more severely affected patients are likely to be evaluated, we may see a disproportionately high number of DLRPN patients with bilateral disease and so may overestimate the frequency of bilateral disease. Nonetheless, the observation that the condition begins focally and unilaterally but spreads over time to become bilateral, involving both proximal and distal segments, is true (see Table 86–2). Although pain is the most problematic early symptom, weakness soon becomes the biggest problem (see Table 86–2). As with the pain, the weakness usually begins focally either proximally or distally but progresses to become multifocal and bilateral. Consequently, both proximal weakness (difficulty getting out of low chairs or going up and down stairs) and distal weakness (footdrop and difficulty walking on toes) are frequent complaints from DLRPN patients. The thigh muscles are often affected, causing the knee to buckle and the patient to fall. Consequently, bone fractures, especially of the hips, can occur in patients with DLRPN. The weakness is usually severe; half of the patients in our series (16 of 33) were wheelchair bound at the time of evaluation and almost all of them required some form of aid to ambulate (brace, cane, or walker) (see Table 86–2). When the disease remains unilateral, patients usually can continue ambulating, although they often will need help. However, when the disease becomes bilateral, with predominant involvement of thigh muscles, patients usually become wheelchair dependent. As noted above, Garland believed that DLRPN is usually solely a motor neuropathy.31,33 However, in most cases of DLRPN, autonomic and sensory systems are prominent. Essentially all patients with DLRPN suffer from pain, which implies involvement of sensory fibers. Furthermore, 31 of 33 patients from our series had paresthesias and most had distal or proximal sensory loss on examination. Quantitative sensation testing showed that there is unequivocal sensory abnormality at different anatomic sites (foot, leg, and thigh) to all sensory modalities (large and small fibers, resulting in abnormalities of vibration, cooling, and heat-pain).24 Similarly, autonomic symptoms are common in DLRPN; one half of our patients had new autonomic symptoms at the time of their illness that included orthostatic intolerance, change in sexual function, urinary dysfunction, diarrhea, constipation,

25

13 1 18 13 31 30 25

No

7.0–87.0 7.0–62.0

0.0–60.5

1.0–11.5

44.0–214.0

5.1–12.9 17.8–36.7 ⫺120.0–0.0 0.7–3.4 56.0–130.0

0.0–60.0 75.0–225.0

35.8–75.9 0.0–35.8 1.4–42.0

Range

18.6 14.4

14.6

2.1

35.3

2.0 4.9 32.6 0.5 19.4

10.6 44.3

10.4 7.9 8.9

SD

8

6

40 50

29

Yes

36.5 33.5

13.5

1.0

66.5

5.5 25.1 ⫺15.0 1.0 63.0

42

34

28

No

6.0–106.3 6.0–77.0

0.0–62.0

0.0–12.0

18.0–283.0

4.3–7.1 17.8–35.4 ⫺90.0–0.0 0.6–1.9 48.0–88.0

0.0–72.0 69.0–124.0

0.5–60.0

7.0 3.0 96.0

27.8–86.5

Range

69.4

Median

57 NA NA

57 57

56

49

50

34 50 57 53 49

45 57

57 NA 57

n

LRPN

21.2 17.0

14.4

2.2

56.9

0.7 4.6 19.6 0.2 9.3

11.6 11.5

12.4

12.3

SD

195 194 195 195 195 195

195 195

195 195 195 195

NA 195

195 195 NA

n

96 148 134 128 20 73

Yes

0.0 0.0

9.8 28.6 0.4 1.0

169.9

65.0 15.0

Median

99 46 61 67 175 122

No

0.0–18.0 0.0–14.0

5.5–16.5 17.8–50.8 ⫺8.7–10.1 0.5–2.5

94.2–389.0

19.0–88.0 5.9–74.8

Range

RDNS

2.5 2.3

2.0 5.7 2.6 0.2

41.9

15.1 8.5

SD

NS

NS

NS

NS NS

0.02

NS

0.009

0.0001 NS 0.002 NS 0.0001

NS 0.0001

NS

0.03

DLRPN vs. LRPN

P*

NS .005 .007 .001 NS .001

.0001 .0001

⬍.001 .002 ⬍.001 .023

.001

NS ⬍.001

DLRPN vs. RDNS

DLRPN ⫽ diabetic lumbosacral radiculoplexus neuropathy; LRPN ⫽ lumbosacral radiculoplexus neuropathy; RDNS ⫽ Rochester Diabetic Neuropathy Study; NA ⫽ not applicable; NIS ⫽ Neuropathy Impairment Score; NS ⫽ not significant (P ⬎ .05); SD ⫽ standard deviation. *Wilcoxon rank sum test for continuous data and Fisher’s exact test for dichotomous data. † For DLRPN and LRPN, weight change is from onset to evaluation; for RDNS, weight change is over 1 year. ‡ Nonproliferative retinopathy. Modified from Dyck, P. J. B., Norell, J., and Dyck, P. J.: Non-diabetic lumbosacral radiculoplexus neuropathy: natural history, outcome and comparison with the diabetic variety. Brain 124:1197, 2001.

7

43.0 37.0

33 33

32

6.0

31

20 32 13 4‡ 2 3 2

1.0

26

33 33 31 17 33 33 27

89.5

26

Sex, male Diabetes, type 2 Insulin use Retinopathy Nephropathy Cardiovascular disease Rheumatoid factor, reactive Antinuclear antibody, positive

7.5 25.7 ⫺30.0 0.9 85.0

30 29 33 30 26

Yes

3.0 144.5

32 30

Dichotomous

65.4 4.1 6.7

33 33 33

Age (yr) Duration of diabetes (yr) Duration of neuropathy at evaluation (mo) Onset to bilateral (mo) Fasting plasma glucose (mg/dL) Glycated hemoglobin (%) Body mass index (kg/m2) Weight change (lb)† Creatinine (mg/dL) Cerebrospinal fluid glucose (mg/dL) Cerebrospinal fluid protein (mg/dL) Cerebrospinal fluid cells (cells/␮L) Erythrocyte sedimentation rate (mm/hr) NIS Total NIS Lower limb

Median

n

Continuous

DLRPN

Table 86–1. Clinical and Laboratory Characteristics of DLRPN and LRPN Patients and a Population-Based Diabetic Cohort (RDNS)

12 18 3 0

27 0 6 0

NS

⬍.001

16 14 3

14 19 0 0

13 0 20 0

⬍.001

⬍.001

0.05

3 16 12

24 5 0 2

6 2 21 2

21 33 3 0

49 1 7 0

NS

⬍.0001

Onset (n ⫽ 57)

28 28 1

27 30 0 0

10 0 47 0

⬍.0001

.0001

.03

Mayo Evaluation (n ⫽ 57)

Telephone Follow-up (n ⫽ 31)*

Mayo Evaluation (n ⫽ 33)

5 21 16

32 7 0 3

12 1 26 3

Telephone Follow-up (n ⫽ 42)*

NS

NS

Onset

NS

NS

.03

Mayo Evaluation

NS

NS

NS

Telephone Follow-up

DLRPN vs. LRPN

DLRPN ⫽ diabetic lumbosacral radiculoplexus neuropathy; LRPN ⫽ lumbosacral radiculoplexus neuropathy; NS ⫽ not significant (P ⬎ .05). *Three LRPN and two DLRPN patients reported they were recovered. † Fisher’s exact test. Modified from Dyck, P. J. B., Norell, J., and Dyck, P. J.: Non-diabetic lumbosacral radiculoplexus neuropathy: natural history, outcome and comparison with the diabetic variety. Brain 124:1197, 2001.

Significance

Aids in Ambulation Wheelchair Walker, cane, or brace None

Significance

Most Involved Site Foot or leg Hip or thigh Buttock or back None

Significance†

Most Severe Symptom Pain Prickling Weakness None

Onset (n ⫽ 33)

LRPN

DLRPN

Table 86–2. Severity of Symptoms, Anatomic Location, and Use of Ambulating Aids in DLRPN and LRPN

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties

or change in sweating. Quantitative autonomic testing showed abnormalities in all patients tested, and the median Composite Autonomic Severity Score (CASS) showed moderate to severe autonomic dysfunction.24 Consequently, DLRPN is not just a motor neuropathy; sensory and autonomic systems also are involved. One reassuring fact about DLRPN is that it is a monophasic illness and improvement occurs in almost all patients. However, the monophasic course and the degree of improvement are frequently overemphasized to patients by their physicians. Patients are often told that all of their pain, weakness, sensory loss, and autonomic symptoms will disappear within several months of their evaluation, and they become frustrated when this does not occur. This is a severe neuropathy; the median value of the Neuropathy Impairment Score (NIS)21 (which is a global score of neurologic impairment taking into account weakness, reflex change, and sensory loss) was 43 points (see Table 86–1), indicative of moderate to severe impairment. The recovery is usually delayed and is often incomplete as nerves regrow slowly and often incompletely and incorrectly. It is often difficult to know in a specific patient whether there still is active disease or whether there is residual damage to nerves and partial regeneration. There usually is long-term morbidity in this disorder from pain, sensory loss, and weakness. Of 33 prospectively studied DLRPN patients, only 2 believed that they had completely recovered after a median time of 2 years.24 Most patients do have substantial improvement of both pain and weakness, and few require the long-term use of wheelchairs (in our series, only 3 of the 15 patients who initially needed wheelchairs still used them 2 years later). However, some degree of ongoing pain, sensory loss, and weakness are common. In general, younger patients do better than older ones and proximal injury usually recovers earlier and more effectively than does distal injury. Consequently, footdrop tends to be a common long-term problem for DLRPN patients. We believe that the explanation for this observation is that reinnervation occurs sooner and more completely in proximal nerve segments. Many patients also have some longterm pain and sensory loss, which is usually much less severe than it was early in the disease course. Patients with long-standing DLRPN often resemble patients with diabetic polyneuropathy. During their active phase, severely affected DLRPN patients have widespread disease with proximal and distal symptoms and signs. After several years, the proximal segments become reinnervated, leaving symmetrical distal deficits. Without knowledge of the clinical course, such patients could easily be mistaken as having a diabetic sensorimotor polyneuropathy.

Nondiabetic LRPN Nondiabetic LRPN is a clinical syndrome that has received much less attention than its diabetic counterpart. Based on

1999

large cohorts of LRPN (n ⫽ 57) and of DLRPN (n ⫽ 33) patients that were both evaluated at our institution, we believe that the two conditions are essentially the same except for the occurrence of diabetes mellitus in DLRPN.22,24,26,27 The clinical features are almost indistinguishable. As in DLRPN, LRPN begins focally with pain involving the back, buttock, thigh, or leg. It starts in an acute or subacute fashion, and patients frequently will remember the day their symptoms began. Over time, the symptoms will spread so that within months the disorder usually involves both proximal and distal segments and becomes bilateral (see Table 86–1). Early in the course pain is the predominant symptom, whereas later weakness is more important (see Table 86–2). As in DLRPN, the proximal segments in LRPN are initially involved somewhat more frequently than are the distal segments. However, with time, most patients develop proximal and distal symptoms (in our series, 52 of 57) and have bilateral involvement (51 of 57). The NIS is similar to that of patients with DLRPN (see Table 86–1). Weight loss is also common26; 42 of the 57 patients in our series had weight loss exceeding 10 pounds. In DLRPN, the weight loss is often attributed to poorly controlled diabetes mellitus, but this explanation is not applicable in LRPN because these patients are not diabetic. As is the case for DLRPN, LRPN is not solely a motor neuropathy; sensory and autonomic systems are unequivocally involved. Essentially, all patients with LRPN have pain (a sensory symptom) and there is involvement of all classes of sensory fibers (vibration, cooling, and heat-pain) at different anatomic sites (foot, leg, and thigh). About one half of patients have new autonomic symptoms that include orthostatic hypotension, urinary dysfunction, diarrhea, constipation, change in sexual function, or change in sweating.26 Long-term morbidity is severe in LRPN as it is in DLRPN. From our series, only 3 of 42 LRPN patients reported (on telephone follow-up interview) that they had complete recovery after a median follow-up time of 3 years. The others still had residual weakness, pain, or sensory loss (see Table 86–2). However, as in DLRPN, real improvement had occurred in most patients. At the time of initial evaluation, 25 LRPN patients used a wheelchair, whereas only 5 used a wheelchair at the time of telephone follow-up. The most problematic long-term impairments were confined to the distal segments (legs and feet) in 25 of the 39 nonrecovered patients. This often was footdrop. Both DLRPN and LRPN can be recurrent conditions. In our LRPN cohort, about 17% of patients (7 of 42) had recurrent episodes of pain and weakness of the lower extremities. We have questioned whether all patients with LRPNs have glucose impairment and if those with the nondiabetic form have simply escaped detection of their glucose dysregulation. We addressed the issue of glucose dysregulation in our

2000

Diseases of the Peripheral Nervous System

LRPN cohort by inquiring if, after long-term follow-up, any of them had developed diabetes mellitus. Only 2 of 4226 eventually became diabetic. Consequently, glucose impairment does not seem to be common, and if it is present in LRPN patients, it does not usually lead to frank diabetes mellitus.

Other Forms of Radiculoplexus Neuropathies As with the DRPNs, the nondiabetic radiculoplexus neuropathies can be divided into cervical (CRPN), thoracic (TRN), and lumbosacral (LRPN) subtypes. In the diabetic populations, DLRPN and DTRN are the most common, whereas in the nondiabetic populations, CRPN probably is the most common. In diabetic patients, the DCRPN and DTRN subtypes frequently occur concomitantly with DLRPN. Thus Bastron and Thomas7 noted that DLRPN and DTRN frequently occurred together in the same patients. Because they considered that there was no good reason except anatomic localization for separating these conditions, these authors grouped them together into the entity of “diabetic polyradiculopathy.” We also have noted that frequently the lower limb and truncal syndromes occur together in the same patients. In our DLRPN cohort, 4 of 33 patients had additional symptoms and findings consistent with thoracic radiculopathy (chest or abdominal wall pain, sensory loss, or weakness). DTRN often begins subacutely with pain. Patients frequently will complain of similar types of pain as those experienced in DLRPN—aching or hurting pain, burning pain, sharp-stabbing or lancinating pain, and contact allodynia. The pain often is constricting and wraps around them. It frequently is in a dermatomal distribution and may be unilateral or bilateral. Patients often will notice that they are becoming “fatter” because they will have protuberance of the belly secondary to weakness of the abdominal wall musculature. They frequently do not want to wear shirts or brassieres because clothing causes pain when touching affected skin (allodynia). Often, patients with DTRN have undergone a large abdominal work-up for their problems with negative results. Because the pain occurs in the setting of weight loss and autonomic symptoms, their physicians are often very concerned that they have amyloidosis or a malignancy. DTRN occurs most commonly (as does DLRPN) in middle-aged or older patients with mild type 2 diabetes mellitus associated with weight loss. It may affect people of either gender, although some series report mostly men.55 It usually is a monophasic illness like DLRPN that will last months with spontaneous improvement. Recovery from this condition is usually good. It can occur in isolation or in association with other forms of DRPN. As in DLRPN, DTRN can be recurrent.

It is also common to find patients with DLRPN who have upper limb symptoms and findings. In our series, one third (11 of 33) of patients had upper-limb involvement. Most of these were mononeuropathies (ulnar neuropathy at the elbow or median neuropathy at the wrist), but about 10% (3 of 33) had a more diffuse disorder consistent with DCRPN.24 These patients had pain followed by weakness not confined to one nerve or one nerve root distribution. Their upper limb syndromes began focally and unilaterally but frequently spread to become more diffuse. Consequently, the clinical course was very similar to what we observed in DLRPN. We have now observed multiple patients from our clinical practice whose problems begin focally in a lower limb (DLRPN) but then spread to adjacent lower limb segments, become bilateral, and eventually involve thoracic and cervical segments. In the end, all four limbs are involved, and these patients appear to have a relatively symmetrical polyradiculoneuropathy that may be confused with CIDP. However, these patients do not have CIDP and should not be labeled as such. The clinical course, the multifocal involvement of nerve segments, the predominant axonal findings on electrophysiologic testing, the characteristic pathologic findings on nerve biopsy, and the prominent pain are helpful in making the differentiation between DRPN and CIDP. Other authors have also observed that different anatomic segments can be affected in the same diabetic patient. Katz and co-authors found that, in their cohort of 60 patients with DLRPN, 9 had cervicobrachial involvement that was unilateral in 5 and bilateral in 4.38 Said and co-workers also noted similar findings of coexisting involvement of upper and lower limb and thoracic nerve segments and labeled it multifocal diabetic neuropathy.51 Consequently, finding truncal or upper-limb involvement does not exclude the diagnosis of DLRPN. The encompassing terms of “diabetic polyradiculopathy” or “multifocal diabetic neuropathy” have appeal, but we prefer DRPN because it more accurately reflects the site of pathologic involvement. Although frequently occurring together, these conditions even more frequently occur separately, and so it is best to designate them by level of involvement (DLRPN, DTRN, or DCRPN). Furthermore, DTRN and DCRPN probably have the same pathophysiologic basis (ischemic injury from microvasculitis) as DLRPN. The findings are similar in the nondiabetic cohort. Nine of the 57 LRPN patients in our series had symptoms and signs consistent with TRN. TRN can also occur in isolation in patients without a lower extremity syndrome (LRPN). Almost half of the LRPN patients (26 of 57) had some upper-limb involvement. As in the diabetic cohort, most of these were mononeuropathies, probably from compression (ulnar neuropathy at the elbow and median neuropathy at the wrist), but about 10% (6 of 57) had a more widespread disorder affecting multiple upper-limb nerves consistent with a CRPN.26

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties

CLINICAL TESTING IN DLRPN AND LRPN The fasting blood glucose levels of patients with DLRPN, DTRN, and DCRPN are elevated (which is expected because these patients have diabetes). This observation was confirmed by comparison of patients in our cohorts of DLRPN and LRPN (see Table 86–1). Overall, the metabolic control of the DLRPN patients was reasonably good. When the DLRPN cohort was compared with the RDNS cohort (a population-based, epidemiologic study of diabetic patients living in Rochester, Minnesota),18 their glycated hemoglobin was significantly lower, their body mass index was significantly lower, and more had type 2 diabetes mellitus (see Table 86–1). Previous authors have argued that markers of altered immunity, especially elevated erythrocyte sedimentation rates, distinguish patients with immune-mediated lumbosacral plexopathies from ones with different etiologies.8 This has not been our experience, and we believe that those with elevated erythrocyte sedimentation rates do not differ markedly from the rest of the radiculoplexus neuropathy patients.26 Some of the patients from our DLRPN and LRPN cohorts had elevated sedimentation rates, reactive rheumatoid factors, positive antinuclear antibodies, or other markers of altered immunity, but most did not. In both DLRPN and LRPN, the cerebrospinal fluid protein concentration is significantly elevated (see Table 86–1), evidence that the disease process extends to the nerve root level. The nerve conduction and needle EMG abnormalities in DLRPN have been extensively studied.7,57 Bastron and Thomas emphasized the multifocal and patchy involvement of lumbosacral and thoracic myotomes.7 Subramony and Wilbourn proposed that there were two distinct electrophysiologic subtypes of DLRPN: one group in which there are focal abnormalities confined to the affected lower limb(s) (multiple lumbosacral radiculopathies) and a second group in which there is evidence of focal abnormalities of the lower limb(s) (multiple lumbosacral radiculopathies) superimposed on a more diffuse peripheral neuropathy.57 Although we believe that the observation of these authors is correct, we do not think that it is helpful to divide patients into two groups. There are not strong reasons to postulate a separate peripheral neuropathy caused by different mechanism than that for the lumbosacral radiculopathies. It is likely that the subgroup reported to have a separate polyneuropathy are the more severely affected patients and have more diffuse electrophysiologic abnormalities. The pathophysiology of DLRPN is patchy but widespread, and patients frequently have concomitant thoracic and upper limb symptoms and signs (as noted above). Consequently, observing widespread electrophysiologic changes is expected. The more severe electrophysiologic abnormalities are seen in segments that are clinically

2001

the most involved (usually lumbosacral levels). The danger of reporting the EMG findings as multiple lumbosacral radiculopathies in addition to a separate diffuse, length-dependent peripheral neuropathy is that it implies two separate pathologic processes. Two separate, coexisting neurologic disorders are unlikely for most patients; rather, it is more likely that there is unequal pathologic involvement of different nerve segments producing a widespread but asymmetrical radiculoplexus neuropathy. Some patients may have a preexisting diabetic polyneuropathy, but most patients with DLRPN probably do not have diabetic polyneuropathy because they have only been diabetic for a short period of time and have reasonable glycemic control. The occurrence of diabetic polyneuropathy has been shown to be highly associated with diabetic retinopathy and nephropathy.18 Few of our DLRPN patients had retinopathy or nephropathy, and so few would be expected to have diabetes-induced polyneuropathy. The nerve conduction and needle EMG findings in DLRPN and LRPN are more suggestive of axonal degeneration than of segmental demyelination (Table 86–3). In our cohorts of DLRPN and LRPN, there are marked reductions in amplitudes of the compound muscle action potentials and of the sensory nerve action potentials with only mild slowing of nerve conduction velocities.24,26 Needle EMG shows fibrillation potentials, decreased recruitment, and long duration, high-amplitude motor unit potentials in muscles innervated by multiple nerve roots and different peripheral nerves. Upper lumbar and lower lumbar levels (see Table 86–3) both tend to be involved. Paraspinal muscles are also usually involved. Fibrillation potentials in lumbosacral paraspinal muscles were seen in 29 of 30 DLRPN patients and in 33 of 48 LRPN patients (who had paraspinal muscles examined). The electrophysiologic abnormalities tend to be more widespread than are clinical deficits. The findings are essentially the same in the two conditions, with the diabetic group having slightly more severe abnormalities (see Table 86–3).24,26 Although axonal degeneration is the main abnormality, an occasional patient will show slow conduction velocities somewhat suggestive of demyelination. In general, the electrophysiologic abnormalities are in keeping with a pathologic process that involved the roots, plexus, and peripheral nerves (a radiculoplexus neuropathy) that usually predominantly involves the lumbosacral segments but often also involves the thoracic and cervical segments. As discussed earlier, DLRPN and LRPN are not just motor neuropathies. There is strong evidence that autonomic fibers and sensory fibers are also affected. Most patients with these neuropathies have sensory (pain and numbness) and autonomic symptoms.24,26 There is objective evidence of dysfunction of these systems provided by abnormality of quantitative sensation and autonomic

2002

Diseases of the Peripheral Nervous System

Table 86–3. Nerve Conduction, Electromyographic, and Quantitative Autonomic Testing in DLRPN Compared with LRPN at Mayo Clinic Evaluation DLRPN Variable Nerve Conduction Studies† Sural SNAP (␮V) CV (m/s) DL (ms) Peroneal CMAP (mV) CV (m/s) DL (ms) Tibial CMAP (mV) CV (m/s) DL (ms) Needle Electromyography†,‡ L2-L4 muscles Fibrillation potentials Long MUPs L5-S1 muscles Fibrillation potentials Long MUPs LS paraspinal muscles Fibrillation potentials Long MUPs Autonomic Reflex Screen (CASS score) Sudomotor Index Adrenergic Index Cardiovascular Index Total CASS

LRPN

n

Median

Range

n

Median

Range

P*

37 4 6

0.0 39.5 4.4

0.0–7.0 35.0–41.0 4.0–4.8

65 22 43

2.0 42.0 4.0

0.0–18.0 36.0–54.0 3.3–5.1

.0001 NS NS

38 27 27

0.2 36.0 5.6

0.0–7.5 24.0–46.0 4.0–14.1

80 56 57

0.5 40.0 4.7

0.0–8.3 23.0–52.0 3.1–7.9

NS .0009 .0001

31 25 25

0.7 36.0 5.2

0.0–11.4 20.0–48.0 3.8–10.8

72 60 60

1.4 41.0 4.7

0.0–14.0 33.0–53.0 3.3–7.5

NS .0003 .003

44 39

1.0 1.0

0.0–3.0 0.0–3.0

88 87

1.0 1.0

0.0–3.0 0.0–3.0

NS NS

56 55

1.4 1.1

0.0–2.8 0.0–2.0

101 97

1.0 1.0

0.0–3.0 0.0–3.0

NS NS

35 18

1.5 1.0

0.0–3.0 0.0–2.0

64 28

0.7 0.0

0.0–3.0 0.0–3.0

.004 NS

14 14 13 14

2.5 2.5 2.0 7.0

1.0–3.0 0.0–4.0 0.0–3.0 2.0–10.0

12 12 12 12

2.0 2.0 1.0 5.0

0.0–3.0 0.0–4.0 0.0–2.0 0.0–8.0

NS NS .009 NS

CASS ⫽ composite autonomic severity score; CMAP ⫽ compound muscle action potential; CV ⫽ conduction velocity; DL ⫽ distal latency; DLRPN ⫽ diabetic lumbosacral radiculoplexus neuropathy; LRPN ⫽ lumbosacral radiculoplexus neuropathy; MUP ⫽ motor unit potential; NS ⫽ not significant (P ⬎ .05); SNAP ⫽ sensory nerve action potential. *Wilcoxon rank sum test. † Patients were counted twice if both sides were sampled. ‡ Values listed are median per level on one side, and are recorded as 0 ⫽ normal, 0.5 ⫽ ⫹/⫺, 1 ⫽ ⫹, 2 ⫽ ⫹⫹, 3 ⫽ ⫹⫹⫹, and 4 ⫽ ⫹⫹⫹⫹. Absent responses are not included. Modified from Dyck, P. J. B., Norell, J., and Dyck, P. J.: Non-diabetic lumbosacral radiculoplexus neuropathy: natural history, outcome and comparison with the diabetic variety. Brain 124:1197, 2001.

testing in the cohorts of DLRPN and LRPN patients.24,26 Results of quantitative sensation tests show that there is sensory abnormality to all classes of fibers in the foot, leg, and thigh and that there were not significant differences between the DLRPN and LRPN cohorts (Table 86–4). Similarly, CASS44 scores derived from quantitative autonomic testing show a moderate to severe autonomic dysfunction overall for both DLRPN and LRPN patients (see Table 86–3). The CASS scores for both groups were

not significantly different, and both groups suffered from generalized autonomic dysfunction.24,26 Consequently, the autonomic dysfunction seen in DLRPN is probably not due to the diabetes itself because the nondiabetic LRPN patients have a similar degree of autonomic impairment. The abnormalities in the autonomic and sensory systems have received less attention than the motor deficits, which are severe; nonetheless, they are present and widespread.

17 8 2 27

17 3 2 22

17 6 3 26

Vibration Foot Leg Thigh Total

Cooling Foot Leg Thigh Total

Heat-Pain 5 Foot Leg Thigh Total 2 1 0 3

0 0 0 0

0 0 0 0

Hyperesthetic (ⱕ5th)*

9 2 2 13

8 0 1 9

4 3 1 8

Normal (6th–94th)*

6 3 1 10

9 3 1 13

13 5 1 19

Hypoesthetic (ⱖ95th)*

19 11 7 37

24 9 7 40

24 3 6 33

Total Tests

2 2 1 5

0 1 0 1

0 0 0 0

Hyperesthetic (ⱕ5th)*

10 6 6 22

13 3 5 21

10 3 4 17

Normal (6th–94th)*

LRPN Tests (24 patients)

7 3 0 10

11 5 2 18

14 0 2 16

Hypoesthetic (ⱖ95th)*

NS

NS

NS

P†

CASE IV ⫽ Computer-Assisted Sensory Examination System IV; DLRPN ⫽ diabetic lumbosacral radiculoplexus neuropathy; LRPN ⫽ lumbosacral radiculoplexus neuropathy; NS ⫽ not significant (P ⬎ .05). *Percentiles. † Fisher’s exact test. Modified from Dyck, P. J. B., Norell, J., and Dyck, P. J.: Non-diabetic lumbosacral radiculoplexus neuropathy: natural history, outcome and comparison with the diabetic variety. Brain 124:1197, 2001.

Total Tests

Evaluation

DLRPN Tests (17 Patients)

Table 86–4. Number of DLRPN and LRPN Quantitative Sensation Tests Having Defined Levels of Abnormality Using CASE IV at the Time of Tertiary Mayo Clinic Evaluation

2004

Diseases of the Peripheral Nervous System

PATHOLOGY OF RADICULOPLEXUS NEUROPATHIES Diabetic LRPN The first information about the pathology of DLRPN was from the only autopsy case, published in 1968.47,48 These studies showed multiple areas suggesting nerve ischemia within the lumbosacral plexus and obturator nerve, and an occluded blood vessel was identified. Inflammatory mononuclear infiltrates were shown around epineurial blood vessels and were judged to be reactive. Unequivocal vasculitis was not identified. A subsequent study of 10 biopsied cases from distal nerve confirmed ischemic damage by demonstrating multifocal fiber loss6; no comment was made about inflammatory lesions. In patients with elevated sedimentation rates, six patients (three with and three without diabetes mellitus) were identified with painful lumbosacral plexopathy and whose biopsies showed perivascular inflammatory cell cuffing and multifocal fiber loss.8 Some researchers did not believe that ischemic injury was the main etiology for DLRPN and argued that metabolic factors and demyelination were more important in the underlying pathophysiology.12,14 Other authors suggested that some cases may be due to ischemic injury, whereas other cases may be due to metabolic derangement.3 The idea that there may be varied subtypes of DLRPN resulting from different causes has been stressed in some studies.50 Said and co-workers argued that some cases of DLRPN are due to ischemic injury from vasculitis, whereas other cases are due to metabolic injury from hyperglycemia. Among their 10 DLRPN patients who had cutaneous thigh nerves biopsied, they found that the 3 most severely affected cases had multifocal fiber loss (ischemic injury), 2 of which were diagnostic of vasculitis. The other seven were more mildly affected and had changes interpreted as showing metabolic derangement.50 Other studies have also shown evidence of altered immunity in DLRPN.39,40,43,46 One study found vasculitic changes in 3 of 14 biopsied cutaneous thigh nerves,43 whereas another found evidence of polymorphonuclear vasculitis with immune complex and complement deposition.39 More recently, recent bleeding has been described and the authors argued that this is evidence of vessel damage.51 We studied the pathologic findings of DLRPN by prospectively evaluating all static or worsening personally seen cases over a 3-year period.24 All patients underwent a distal cutaneous nerve biopsy (usually sural) from the more recently or severely involved leg. We found strong evidence implicating ischemic injury (see Table 86–3). Most nerves showed multifocal fiber loss (Fig. 86–1), perineurial degeneration or thickening (Fig. 86–2; see also Fig. 86–1), neovascularization (see Fig. 86–2), and abortive regeneration of nerve fibers forming microfasciculi (injury neuroma),

FIGURE 86–1 Transverse semithin epoxy sections of sural nerves from patients with DLRPN stained with para-phenylenediamine (top) and with methylene blue (bottom) showing multifocal fiber degeneration and loss (top panel, arrows), abortive repair with injury neuroma (microfasciculation) (bottom panel, arrows), and thickened perineurium (bottom panel, between arrowheads). Top, The three affected fascicles (arrows) are essentially devoid of myelinated fibers, whereas the adjacent fascicles are relatively unaffected. Bottom, Abortive regeneration is found both within the original perineurium (which is damaged and thickened) and beyond it. Note also that the rest of the nerve is severely damaged and essentially no myelinated fibers remain. All of these findings are typical of ischemic injury and repair and were commonly seen in both DLRPN and LRPN nerves. See Color Plate

both within damaged perineurium and outside of it (see Fig. 86–1). When compared to nerves of patients with diabetic polyneuropathy, the DLRPN nerves showed many more ischemic changes (Table 86–5). Some nerve fibers from damaged regions showed enlarged dark axons (from accumulation of organelles) that contained light cores (areas of intact axoplasm). At the regions of swelling and organelle accumulation, these axons are frequently devoid of myelin as they are more distally at sites of axonal attenuation. More

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties

FIGURE 86–2 Transverse paraffin sections of sural nerves showing typical changes seen in diabetic and nondiabetic lumbosacral radiculoplexus neuropathies as described in text. Top, Inflammation in the wall of an epineurial microvessel (right upper), probably fibrinoid degeneration of the perineurium (arrow), and a region of neovascularization (arrowhead). (Masson’s trichrome stain.) Middle, Several fascicles surrounded by normal-thickness perineurium (right middle, between two arrowheads) and one fascicle with extremely thick perineurium (left middle, between two arrowheads). We attribute the latter to scarring and repair after ischemic injury (note that all fascicles are devoid of myelinated fibers). (Luxol fast blue/periodic acid–Schiff stain.) Bottom, Accumulation of hemosiderin (bright blue iron stains, arrow) deposited along the inner aspects of the perineurium (Turnbull blue stain). All of these pathologic features are frequently seen together and are best explained by ischemic injury. (From Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus [LSRPN]: similarity to the diabetic variety. J. Neuropathol. Exp. Neurol. 59:525, 2000, with permission.) See Color Plate

2005

distally frequently they are remyelinated. Similar dark axons with light cores that show demyelination have been described in experimental models of ischemic nerve damage.45 Teased fiber abnormalities showed that axonal degeneration was the main fiber abnormality. There were increased rates of axonal degeneration, empty nerve strands, and segmental demyelination when compared to normal controls (Table 86–6). Frequently the segmental demyelination was clustered on individual nerve fibers, suggesting a demyelinating disorder secondary to axonal pathology.22 A potential explanation for the high rate of ischemic damage in DLRPN is a vasculopathy caused by the chronic hyperglycemic state of diabetes mellitus (metabolic damage). However, when we compared our DLRPN group with a population-based cohort of diabetic patients (the RDNS cohort), we found that the DLRPN patients had better glycemic control, used less insulin, were more likely to have type 2 diabetes mellitus, had been diabetic for a shorter period of time, and had less vascular disease (fewer cardiovascular events) than did the population-based diabetic RDNS cohort (see Table 86–1). Furthermore, nerve biopsies from patients with diabetic polyneuropathy (which is associated with prolonged chronic hyperglycemia) have significantly fewer ischemic changes than do nerves of DLRPN patients (see Table 86–5). Also, the endoneurial microvessel disease (degeneration of pericytes and reduplication of basement membranes) that is associated with chronic hyperglycemia35 is rarely found in DLRPN nerves but frequently is seen in diabetic polyneuropathy nerves. Consequently, it seems unlikely that a diabetes-induced metabolic vasculopathy is the cause of DLRPN. Our evidence suggests that the primary event in DLRPN is an immune attack and probably a microvasculitis of nerve. We found evidence of inflammation in association with the ischemic damage.24 All nerves of our DLRPN patients showed inflammatory infiltrates, usually surrounding epineurial microvessels. The degree of inflammation was much greater than in healthy control nerves or in nerves from patients with diabetic polyneuropathy (see Table 86–5). In half of biopsies, the inflammatory cells disrupted vessel walls and were suggestive of microvasculitis (Fig. 86–3). Large nerve blood vessels (large arterioles and small arteries) typically were not involved. Instead, the inflammatory vasculopathy involved small arterioles, venules, or capillaries (a microvasculitis). Although occasionally seen, fibrinoid degeneration of vessel walls was rare. Evidence of previous bleeding as seen on iron stains adjacent to damaged blood vessels and in the perineurium was common; hemosiderin-laden macrophages were seen in about half of the cases25 (see Fig. 86–2 and Table 86–5). Our findings suggest that DLRPN is a uniform disease caused by one mechanism—ischemic injury from

2006

Diseases of the Peripheral Nervous System

Table 86–5. Pathologic Results of Distal Cutaneous Nerve Biopsies from DLRPN Compared with LRPN, Diabetic Polyneuropathy (DPN), and Healthy Control Nerves DLRPN Variable Characteristics of Patients Biopsied Number Sex, male Median age (yr) Paraffin and Epoxy Sections Endoneurial and Perineurial Abnormality Fiber degeneration or loss Multifocal fiber degeneration or loss Focal perineurial degeneration Focal perineurial thickening Injury neuroma† Interstitial Abnormality Perivascular inflammation Individual cells (⬍10 cells) Small collections (11–50 cells) Moderate collections (51–100 cells) Large collections (⬎100 cells) Inflammation of vessel wall Diagnostic of microvasculitis Hemosiderin in macrophages Neovascularization

LRPN

DPN

Healthy Controls

n

n

P*

n

33 20 65.4

47 24 67.2

NS NS

21 15 57.0

NS .002

25 19 6 24 12

37 31 7 27 16

NS NS NS NS NS

15 2 0 2 0

NS ⬍.001 NS ⬍.001 .002

0 0 0 2 0

⬍.001 ⬍.001 NS ⬍.001 .009

33 0 21 7 5 15 2 19 21

47 5 29 6 7 24 7 25 21

NS

6 5 1 0 0 0 0 0 1

⬍.001

2 2 0 0 0 0 0 0 1

⬍.001

NS NS NS NS

P*

n

P*

14 9 61.5

NS NS

⬍.001 NS ⬍.001 ⬍.001

.002 NS ⬍.001 ⬍.001

DLRPN ⫽ diabetic lumbosacral radiculoplexus neuropathy; LRPN ⫽ lumbosacral radiculoplexus neuropathy; NS ⫽ not significant (P ⬎ .05). *Wilcoxon rank sum test for continuous data and Fisher’s exact test for dichotomous data. † Injury neuroma ⫽ abortive regenerative activity outside of the original perineurium. Data from Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy (LSRPN): similarity to the diabetic variety (DLSRPN). J. Neuropathol. Exp. Neurol. 59:525, 2000; Dyck, P. J. B., Norell, J., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999.

microvasculitis. We do not find evidence that some cases are due to ischemic injury, whereas other cases are due to metabolic injury, as has been postulated.3,50 We are not concerned that all cases did not show vasculitic changes. We do not expect to see vasculitis in all cases because

(1) vasculitis disorders are by nature patchy; (2) the disease process involves roots, plexus, and proximal and distal peripheral nerve multifocally; and (3) only a short segment of distal cutaneous nerve is examined in a biopsy. In fact, when we attempted to define further the longitudinal

Table 86–6. Teased Fiber Results from LRPN Compared with DLRPN and Healthy Control Nerves LRPN

†

Normal: A, B (%) Demyelination: C, D (%) Remyelination: F, G (%) Axonal: E, H (%) Classifiable (no.) Empty (no.)

DLRPN

Healthy Controls

Median

Range

Median

Range

P*

Median

Range

P*

60.7 2.6 14.6 17.2 83.0 25.5

0–88 0–13 0–38 0–100 9–145 1–89

45.7 2.6 11.9 30.9 63.0 42.5

0–92 0–25 0–33 0–100 7–127 5–93

.25 .82 .32 .17 .04 .007

85.7 0.0 13.7 0.0 99.5 6.0

72–95 0–2 5–28 0–3 86–118 0–19

⬍.001 ⬍.001 .67 ⬍.001 .001 ⬍.001

DLRPN ⫽ diabetic lumbosacral radiculoplexus neuropathy; LRPN ⫽ lumbosacral radiculoplexus neuropathy. *Wilcoxon rank sum test. † Teased fiber types: A ⫽ normal, B ⫽ myelin wrinkling, C ⫽ demyelination, D ⫽ demyelination and remyelination, E ⫽ axonal degeneration, F ⫽ remyelination, G ⫽ myelin reduplication, H ⫽ regeneration after axonal degeneration, empty ⫽ unclassifiable strands without myelinated fibers. Modified from Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy (LSRPN): similarity to the diabetic variety (DLSRPN). J. Neuropathol. Exp. Neurol. 59:525, 2000.

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties

2007

FIGURE 86–3 Serial skip paraffin sections of a microvessel at (upper row) and distal to (bottom row) a region of microvasculitis from the sural nerve of a patient with DLRPN. The sections in the left column are stained with hematoxylin and eosin, the sections in the middle column are reacted to anti–human smooth muscle actin, and the sections in the right column are reacted with leukocyte common antigen (CD45). The smooth muscle or the tunica media in the region of the microvasculitis (upper row) is separated by mononuclear cells, fragmented, and decreased in amount. The changes are those of a focal microvasculitis and are seen in both diabetic and nondiabetic patients. See Color Plate

extent of the microvasculitis lesions by doing serial skip sections in eight nerves (four with DLRPN and four with LRPN), we found that the microvasculitis was more common than we had originally appreciated on routine preparations. This was because the inflammatory lesions were usually localized only to short segments along the vessels. By using immunohistochemical staining on serial skip

sections, we found evidence that lymphocytes separated and destroyed the smooth muscle cells of the microvessel wall in a focal pattern over a short length. In some places along vessels, the smooth muscle was separated, fragmented, displaced outward, and diminished in amount by lymphocytic infiltration, whereas at other levels the vessels were relatively unaffected (Fig. 86–4; see also Fig. 86–3).

FIGURE 86–4 Paraffin sections of sural nerve from patients with nondiabetic lumbosacral radiculoplexus (LRPN) (upper three panels) and diabetic lumbosacral radiculoplexus neuropathy (DLRPN) (lower panel) reacted for anti–human smooth muscle actin (Dako) and counterstained with hematoxylin and eosin showing microvasculitis of four nerve vessels (right column) and an unaffected level of the same vessel proximal or distal to the lesion (left column). Compared with the relatively unaffected regions of the same microvessels (left), in the regions of microvasculitis (right) the smooth muscle is separated by mononuclear cells, fragmented, displaced outward, and decreased in amount. We interpret these changes to be typical of microvasculitis. They were encountered most commonly in the epineurium but were also found in the endoneurium and perineurium. Note the similarity of the microvasculitis in the LRPN nerves (top three) and the DLRPN nerve (bottom). Note also the adjacent unaffected vessels are often larger in size. (From Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus [LSRPN]: similarity to the diabetic variety. J. Neuropathol. Exp. Neurol. 59:525, 2000, with permission.) See Color Plate

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties

Finding abnormal degrees of inflammation in almost all nerves studied and inflammatory cells separating the muscle layers of vessels in half of the nerves in association with ischemic injury is compelling evidence of an immune/ vasculitis mechanism in DLRPN. What, then, is the explanation for the increased rates of segmental demyelination seen? Given the high degree of inflammation, an inflammatory demyelinating mechanism would seem an attractive hypothesis. However, we do not think this explanation is likely. The DLRPN nerves that showed increased frequency of segmental demyelination were significantly associated with having ischemic injury (multifocal fiber loss).22 Furthermore, the demyelinating changes tended to be clustered on individual axons, which is typical of demyelination secondary to axonal dystrophy.20,45 Demyelination secondary to axonal dystrophy has been demonstrated in uremic neuropathy, Friedreich’s ataxia, and animal models of ischemic injury.17,19,45 Consequently, both the segmental demyelination and the axonal degeneration observed are likely to be due to ischemic injury from microvasculitis.

Nondiabetic LRPN Little has been written about the underlying pathologic mechanism of LRPN.8 However, because of the similarity in the clinical, laboratory, and electrophysiologic features of DLRPN and LRPN, it seems likely that the two conditions are due to the same pathophysiologic mechanisms. Consequently, we did a retrospective study identifying LPRN patients who had nerve biopsies in order to determine whether the pathologic findings in distal cutaneous nerve were the same in the two conditions. The pathologic findings in LRPN are strikingly similar to those of DLRPN (see Table 86–5).22 There was multifocal fiber loss, regions of abortive regeneration within or beyond the original fascicle forming microfasciculi (injury neuroma), focal degeneration or scarring of the perineurium (see Fig. 86–2), and epineurial neovascularization (see Fig. 86–2). There were also ultrastructural changes observed in individual nerve fibers, such as enlarged dark axons with light cores. As in DLRPN, these dark axons with light cores frequently showed demyelinating changes both at the site of organelle accumulation as well as distally at sites of axonal attenuation (Fig. 86–5). All of these changes are attributable to ischemic damage. As in DLRPN, the ischemic changes seemed to be caused by altered immunity and microvasculitis. Epineurial perivascular inflammatory collections were seen in all nerves. There were features suggestive of microvasculitis with inflammatory cells separating microvessel wall elements in half of nerves. Serial skip sections demonstrated that the microvasculitis regions were focal, with some areas affected and adjacent areas unaffected (see Fig. 86–4). Fibrinoid degeneration (typical of large vessel vasculitic) was rarely seen.

2009

Teased fiber preparations from LRPN nerves confirm the observation that fiber loss and axonal degeneration are the main pathologic abnormalities of myelinated fibers because empty nerve strands and axonal degeneration are the most common teased nerve fiber abnormalities (see Table 86–6). There is also an increased rate of segmental demyelination as compared to normal.22 As in DLRPN, the LRPN nerves that had increased frequencies of segmental demyelination were significantly associated with nerves that showed ischemic damage (multifocal fiber loss). The segmental demyelination tended to be clustered as individual nerve strands, which implies demyelination secondary to axonal dystrophy. In some cases, proximal portions of nerve fibers showed multiple areas of segmental demyelination, whereas distal portions showed axonal degeneration (Fig. 86–6). Fibers such as these give evidence that the demyelination is due to an underlying axonal dystrophy. These findings taken together suggest that both the axonal degeneration and the segmental demyelination in LRPN are due to the common mechanism of ischemic injury. Overall, the pathologic findings of DLRPN and LRPN are very much alike. When we compared the findings from our two cohorts, there were few differences. There were significantly more empty nerve strands (end products after axonal degeneration) in DLRPN nerves than in LRPN nerves. Possible explanations for this observation are that the diabetic cohort was somewhat more severely affected than the nondiabetic cohort or that a small minority of the DLRPN patients had a chronic, preexisting diabetic polyneuropathy (although most probably do not, as discussed earlier). Our findings provide evidence that the primary event in both DLRPN and in LRPN is nerve ischemia caused by a microvasculitis. We believe these conditions are forms of nonsystemic vasculitis of nerve. A vasculitic process could explain the clinical pattern of sudden-onset, painful, asymmetrical neuropathy that these patients have. Weight loss is common in vasculitis disorders (as it is in DLRPN and LRPN). The small size of blood vessels involved may help explain the occurrence of a patchy, widespread asymmetrical process, as opposed to the discrete multiple mononeuropathies more typical of large-nerve vessel necrotizing vasculitis. It is unclear why the lumbosacral roots, plexus, and nerves are particularly susceptible in these conditions, but as noted above, it is common to find coexisting truncal radiculopathies (DTRN and TRN) and upper limb neuropathies (DCRPN and CRPN) in patients with DLRPN and LRPN; furthermore, EMG changes in these patients tend to be more diffuse than clinical involvement would suggest. One of the most compelling arguments that DLRPN is not due to metabolic derangement from hyperglycemia is the similarity between the clinical course and pathology of DLRPN and that of LRPN. Because LRPN patients are not diabetic, it is difficult to implicate hyperglycemia and

2010

Diseases of the Peripheral Nervous System

FIGURE 86–5 Serial skip electron micrographs along the length of one nerve fiber from a nondiabetic lumbosacral radiculoplexus neuropathy nerve that are pictured from proximal (upper left) to distal (lower right) levels and show injury that we attribute to ischemia. Top left and right, Two levels of a markedly enlarged axon, with accumulated organelles and light cores (relatively normal axoplasm, arrows). Notice that the axon has become demyelinated. Bottom left, The same axon (arrow) taken from a more distal level is much smaller in size (probably attenuated) and still is without myelin. Bottom right, At an even more distal level, the axon (arrow) is somewhat larger and now is myelinated (probably remyelinated because the myelin is thin). In this case, the axon has not been transected. We suggest that axonal enlargement (presumably from interruption resulting from interference with rapid axonal transport caused by ischemia) or attenuation may lead to demyelination. The fate of the axon distal to organelle accumulation may be attenuation and restoration or attenuation and degeneration. (From Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus [LSRPN]: similarity to the diabetic variety. J. Neuropathol. Exp. Neurol. 59:525, 2000, with permission.)

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties

2011

Other Forms of Radiculoplexus Neuropathy

FIGURE 86–6 Consecutive portions (A–L), teased from proximal to distal, along the length of a single teased fiber from a patient with nondiabetic lumbosacral radiculoplexus neuropathy showing multiple regions of proximal demyelination (between arrows in B, D, E, and G), a region of remyelination (the myelin thickness is 50% or less of the normal myelin thickness [G–I]), and distal axonal degeneration (I–L). Finding proximal demyelination and distal axonal degeneration is strong evidence that the processes of demyelination and axonal degeneration are linked and that the demyelination is probably secondary to an axonal dystrophy. (From Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus [LSRPN]: similarity to the diabetic variety. J. Neuropathol. Exp. Neurol. 59:525, 2000, with permission.)

metabolic derangement as the cause of their neuropathy. Consequently, chronic hyperglycemia also probably does not play the major role in the pathogenetics of DLRPN. One could ask whether patients with LRPN really have undetected mild diabetes mellitus. This seems unlikely because only 2 of 42 LRPN patients eventually developed diabetes mellitus.26 It is possible that they have mild glucose impairment, but this has not been demonstrated. In spite of these findings, we believe that hyperglycemia probably is a risk factor for developing DLRPN, if not the direct cause. The frequency of DLRPN is about 1% among diabetic patients,18 whereas the frequency of LRPN among the general population is not known. Nonetheless, most investigators believe that LRPNs occur more commonly among diabetic patients. Consequently, the diabetic state (hyperglycemia) probably plays a role in the disease process, even if it is not the direct cause of the condition.

Much less has been written about the pathology in other forms of radiculoplexus neuropathies than about the pathology of DLRPN and LRPN. Given the difficult anatomic location of the nerves in the truncal radiculopathies, few thoracic root biopsies have been performed. We have done a small number of thoracic root biopsies that have shown some degree of inflammatory infiltration. Similarly, little has been written about the pathologic findings in patients with the upper limb syndromes (DCRPN or CRPN). There are two studies describing six patients with brachial plexus neuropathies who had biopsies taken from the plexus, all of which showed inflammatory mononuclear cell infiltration.16,56 We studied biopsies taken from 20 patients with immune brachial plexus neuropathy.23 Eight were taken from the brachial plexus and 12 from forearm nerves. In the forearm biopsies, there was fiber loss in 11 of 12 nerves (multifocal in 6), perineurial thickening in 9 of 12 nerves, neovascularization in 5 of 12 nerves, and inflammatory mononuclear cell infiltrates in 8 of 12 nerves. In 5 of 12 nerves, the inflammatory cells infiltrated and separated vessel wall elements and were suggestive of microvasculitis. All eight of the brachial plexus biopsies showed inflammatory mononuclear cell infiltrates, often in large collections in the endoneurium but also involving microvessels. We interpreted these findings to be suggestive of an immune or microvasculitic pathogenesis with some evidence of ischemic injury. Moreover, the pathologic findings of CRPN had many features in common with those of DLRPN and LRPN. These results are not surprising, given the similar clinical pattern of the upper limb and lower limb syndromes.

IMMUNOTHERAPY There are no proven treatments for any of the radiculoplexus neuropathies. Because of the evidence that DLRPN, LRPN, and CRPN are probably immune-mediated neuropathies caused by microvasculitis of nerve, immunotherapy might be useful if employed early when the inflammatory phase of the illness is active and before too much axonal degeneration has occurred. To date, no controlled studies in DLRPN have been published, although small uncontrolled case series exist. Two DLRPN patients with vasculitis on nerve biopsy reportedly showed improvement with prednisone.50 Fifteen DLRPN patients received immune therapy; seven got a combination of intravenous immune globulin (IVIG) and prednisone, five got IVIG alone, two got prednisone and cyclophosphamide, and one got prednisone alone.40 All patients were reported to be improved and five markedly so, although the authors did not comment on how they differentiated the spontaneous improvement expected in a monophasic disorder from improvement resulting from immunotherapy. In

2012

Diseases of the Peripheral Nervous System

another study,46 12 DLRPN patients who received treatment (prednisone, IVIG, or plasma exchange) were compared to 29 who received no treatment. The authors concluded that, on average, the group that received immune-modulating treatment had greater impairment before treatment but improved to a greater extent and at a faster rate than the untreated cohort. The results of these studies support the idea that immune-modulating therapies might be helpful in DLRPN. However, because DLRPN is a monophasic illness and spontaneous improvement is usual, these reports of efficacy must be viewed as preliminary. Furthermore, in another study of DLRPN, three patients had progression of their deficits despite being treated with immune therapy, including IVIG.65 Consequently, we have begun a double-blind, placebocontrolled trial in DLRPN with intravenous methylprednisolone. Included cases must be static or worsening and be within 6 months of disease onset, so as to include only patients with active disease. The neuropathy can be unilateral, bilateral, or more widespread, but the predominant disease must be in the lower extremities. The symptoms and findings cannot be confined to one nerve or to one nerve root distribution. Structural causes are excluded with magnetic resonance imaging of the lumbosacral spine and lumbosacral plexus. The study is multicentered and is being conducted at centers in the United States and Canada. Two thirds of patients receive active drug and one third receive placebo. The methylprednisolone is initially given at a dose of 1.0 g three times weekly and then at lesser frequencies and lower doses over a 12-week period. Patients are examined and a neurologic symptom questionnaire (Neuropathy Symptoms and Change Score)36 is administered on eight separate occasions over a 2-year period to determine whether symptoms are worsening or improving. The final endpoint of the study is time to improvement of the NIS by 4 points. Secondary end points include a functional gait scale; isometric muscle strength testing; the Neuropathy Symptoms and Change Score; an analogue pain scale; quantitative sensation testing performed on the foot, leg, and thigh; quantitative autonomic testing; and nerve conduction studies (bilateral femoral, peroneal, and tibial motor; sural sensory). The study is now finished enrolling and treating patients. However, it is still ongoing because it is a 2-year study to judge long-term recovery of patients. We do not have any preliminary results from the study and anticipate its completion in late summer of 2005. There have been concerns about giving diabetic patients intravenous steroids. However, DLRPN patients usually have type 2 diabetes mellitus and their blood sugar levels mostly are under good control. Most of the patients in our study have not had much disruption in their glycemic control after infusion of the study drugs, although transient fluctuation of blood sugar levels does occur. No patients to date have had major complications related to steroids. One

patient had a rectal abscess and another died of pneumonia several months after completing her study treatments. We concluded that this death was probably unrelated to the study drug. Little has been published about immunotherapy in LRPN. Four of six patients in one study were reported to have improved with prednisone,8 whereas a single case was thought not to improve with prednisone in another study.5 Two other small series (two patients in one and five patients in the other) reported that LRPN patients improved with IVIG.59,60 We conducted an open trial of weekly intravenous infusion of methylprednisolone in 11 LRPN patients over 8 to 16 weeks.25 All patients had marked improvement of pain; for many, the pain resolved completely. All patients had improvement of weakness, and 9 of 11 judged this improvement as marked. The NIS was significantly improved after the treatment, in half by at least 50% (Fig. 86–7). Before treatment, half of patients were in wheelchairs and all but one needed aid in ambulating, whereas after treatment one patient was still in a wheelchair and six walked independently. These treatment data are preliminary and are not definitive; there were no control patients, and LRPN can improve spontaneously. Nonetheless, we believe the treatment probably was helpful because the improvement began with initiation of treatment in all cases and the degree of improvement was marked. Prospective, placebocontrolled, double-blind studies (as are being done in DLRPN) will be needed to determine whether immunemodulating therapies are effective in treating LRPN.

SYMPTOMATIC TREATMENT All of the radiculoplexus neuropathies (including DLRPN and LRPN) are severe illnesses with great morbidity. Because they are usually monophasic and improve with time, physicians may underappreciate the suffering and disability experienced. Patients frequently are so weak and have such severe pain that they are unable to work; they often require hospitalization to achieve adequate pain control. Most patients with a radiculoplexus neuropathy will need narcotic medication at some point in their illness as well as treatment with other chronic neuropathic pain medications (such as tricyclic antidepressants and anticonvulsants). Patients should be warned that it is often impossible to relieve the pain completely, and the goal is to make the pain tolerable. Because the radiculoplexus neuropathies begin abruptly and severely, these conditions are often misdiagnosed, and many patients undergo unnecessary operations for suspected disc disease. Patients often worry that the disease will be progressive and fatal, and because of the associated weight loss, they or their physicians believe they may have a cancer or amyloidosis. Very commonly, patients with radiculoplexus neuropathies

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties

2013

and disability insurance when needed. Patients need to keep the perspective that, although the illness may get worse in the short term, nearly all patients with radiculoplexus neuropathy (especially those with DLRPN and LRPN) will improve to some degree, if incompletely, over time.

REFERENCES

FIGURE 86–7 The Neuropathy Impairment Scores (NISs) of 11 patients with LRPN plotted against time after a course of intravenous methylprednisolone therapy. Each line represents a different patient; large dots represent patient evaluations; and solid lines represent the treatment period. Without exception, the NIS improved (scores decreased) after the treatment period, sometimes dramatically. This improvement in NIS probably reflects efficacy of therapy, but the study was an open trial, bias could have influenced results, and improvement may occur spontaneously. (From Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Methylprednisolone may improve lumbosacral radiculoplexus neuropathy. Can. J. Neurol. Sci. 28:224, 2001, with permission.)

become depressed as a result of protracted pain, weakness, sensory loss, and loss of sleep. Patients should be advised that, although the disease may be self-limited and monophasic, its course tends to be long and recovery is often incomplete. Axonal degeneration is the main pathologic process, and although nerves do regenerate, they do so at a very slow rate. Patients who are younger and have predominantly proximal disease have a better prognosis. Depression should be treated with appropriate antidepressant medication, and patients should be encouraged to stay involved with hobbies and other interests. Physical therapy should be used and homes should be made as safe as possible. Bracing, such as ankle-foot orthoses, should be fitted and used if needed. Physicians need to be aggressive in helping patients with pain control and in obtaining health

1. Althaus, J.: Neuritis of the circumflex nerve in diabetes. Lancet 1:455, 1890. 2. Archer, A. G., Watkins, P. J., Thomas, P. K., et al.: The natural history of acute painful neuropathy in diabetes mellitus. J. Neurol. Neurosurg. Psychiatry 46:491, 1983. 3. Asbury, A. K.: Proximal diabetic neuropathy. Ann. Neurol. 2:179, 1977. 4. Auche, M. B.: Des alterations des nerfs peripheriques. Arch. Med. Exp. Anat. Pathol. 2:635, 1890. 5. Awerbuch, G. I., Nigro, M. A., Sandyk, R., and Levin, J. R.: Relapsing lumbosacral plexus neuropathy. Eur. Neurol. 31:348, 1991. 6. Barohn, R. J., Sahenk, Z., Warmolts, J. R., and Mendell, J. R.: The Bruns-Garland syndrome (diabetic amyotrophy) revisited 100 years later. Arch. Neurol. 48:1130, 1991. 7. Bastron, J. A., and Thomas, J. E.: Diabetic polyradiculopathy: clinical and electromyographic findings in 105 patients. Mayo Clin. Proc. 56:725, 1981. 8. Bradley, W. G., Chad, D., and Verghese, J. P.: Painful lumbosacral plexopathy with elevated erythrocyte sedimentation rate: a treatable inflammatory syndrome. Ann. Neurol. 15:457, 1984. 9. Bruns, L.: Ueber neuritsche Lahmungen beim diabetes mellitus. Berlin Klin. Wochenschr. 27:509, 1890. 10. Buzzard, F.: Illustrations of some less known forms of peripheral neuritis, especially alcoholic monoplegia and diabetic neuritis. Br. Med. J. 1:1419, 1890. 11. Calverley, J. R., and Mulder, D. W.: Femoral neuropathy. Neurology 10:963, 1960. 12. Chokroverty, S.: Proximal nerve dysfunction in diabetic proximal amyotrophy. Arch. Neurol. 39:403, 1982. 13. Chokroverty, S., Reyes, M. G., and Rubino, F. A.: BrunsGarland syndrome of diabetic amyotrophy. Trans. Am. Neurol. Assoc. 102:1, 1977. 14. Chokroverty, S., Reyes, M. G., Rubino, F. A., and Tonaki, H.: The syndrome of diabetic amyotrophy. Ann. Neurol. 2:181, 1977. 15. Chokroverty, S., and Sander, H. W.: AAEM Case Report #13: diabetic amyotrophy. Muscle Nerve 19:939, 1996. 16. Cusimano, M. D., Bilbao, J. M., and Cohen, S. M.: Hypertrophic brachial plexus neuritis: a pathological study of two cases. Ann. Neurol. 24:615, 1988. 17. Dyck, P. J., Johnson, W. J., Lambert, E. H., and O’Brien, P. C.: Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin. Proc. 46:400, 1971. 18. Dyck, P. J., Kratz, K. M., Karnes, J. L., et al.: The prevalence by staged severity of various types of diabetic neuropathy, retinopathy, and nephropathy in a population-based cohort: the Rochester Diabetic Neuropathy Study. Neurology 43:817, 1993.

2014

Diseases of the Peripheral Nervous System

19. Dyck, P. J., and Lais, A. C.: Evidence for segmental demyelination secondary to axonal degeneration in Friedreich’s ataxia. In Kakulas, B. A. (ed.): Clinical Studies in Myology. Amsterdam, Excerpta Medica, p. 253, 1973. 20. Dyck, P. J., Lais, A., Karnes, J., et al.: Peripheral axotomy induces neurofilament decrease, atrophy, demyelination and degeneration of root and fasciculus gracilis fibers. Brain Res. 340:19, 1985. 21. Dyck, P. J., Sherman, W. R., Hallcher, L. M., et al.: Human diabetic endoneurial sorbitol, fructose, and myo-inositol related to sural nerve morphometry. Ann. Neurol. 8:590, 1980. 22. Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy (LSRPN): similarity to the diabetic variety (DLSRPN). J. Neuropathol. Exp. Neurol. 59:525, 2000. 23. Dyck, P. J. B., Engelstad, J., Suarez, G. A., and Dyck, P. J.: Biopsied upper limb nerves provide information about distribution and mechanism in immune brachial plexus neuropathy. Neurology 56:A395, 2001. 24. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999. 25. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Methylprednisolone may improve lumbosacral radiculoplexus neuropathy. Can. J. Neurol. Sci. 28:224, 2001. 26. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Non-diabetic lumbosacral radiculoplexus neuropathy: natural history, outcome and comparison with the diabetic variety. Brain 124:1197, 2001. 27. Dyck, P. J. B., and Windebank, A. J.: Diabetic and non-diabetic lumbosacral radiculoplexus neuropathies: new insights into pathophysiology and treatment. Muscle Nerve 25:477, 2002. 28. Ellenberg, M.: Diabetic neuropathic cachexia. Diabetes 23:418–423, 1974. 29. Ellenberg, M.: Diabetic truncal mononeuropathy: a new clinical syndrome. Diabetes Care 1:10, 1978. 30. Evans, B. A., Stevens, J. C., and Dyck, P. J.: Lumbosacral plexus neuropathy. Neurology 31:1327, 1981. 31. Garland, H.: Diabetic amyotrophy. Br. Med. J. 2:1287, 1955. 32. Garland, H.: Neurological complications of diabetes mellitus: clinical aspects. Proc. R. Soc. Med. 53:137, 1960. 33. Garland, H.: Diabetic amyotrophy. Br. J. Clin. Pract. 15:9, 1961. 34. Garland, H. T., and Taverner, D.: Diabetic myelopathy. Br. Med. J. 1:1405, 1953. 35. Giannini, C., and Dyck, P. J.: Pathologic alterations in human diabetic polyneuropathy. In Dyck, P. J., and Thomas, P. K. (eds.): Diabetic Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 279, 1999. 36. Grant, I. A., and Dyck, P. J.: Differential diagnosis of diabetic neuropathies. In Dyck, P. J. and Thomas, P. K. (eds.): Diabetic Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 415, 1999. 37. Gregersen, G.: Diabetic amyotrophy: a well-defined syndrome? Acta Med. Scand. 185:303, 1969. 38. Katz, J. S., Saperstein, D. S., Wolfe, G., et al.: Cervicobrachial involvement in diabetic radiculoplexopathy. Muscle Nerve 24:794, 2001.

39. Kelkar, P. M., Masood, M., and Parry, G. J.: Distinctive pathologic findings in proximal diabetic neuropathy (diabetic amyotrophy). Neurology 55:83, 2000. 40. Krendel, D. A., Costigan, D. A., and Hopkins, L. C.: Successful treatment of neuropathies in patients with diabetes mellitus. Arch. Neurol. 52:1053, 1995. 41. Leyden, E.: Die Entzundung der peripheren Nerven. Dtsch. Militar. Z. 17:49, 1887. 42. Leyden, E.: Beitrag zur Klinik des diabetes mellitus. Wien Med. Wochenschr. 43:926, 1893. 43. Llewelyn, J. G., Thomas, P. K., and King, R. H. M.: Epineurial microvasculitis in proximal diabetic neuropathy. J. Neurol. 245:159, 1998. 44. Low, P. A.: Composite autonomic scoring scale for laboratory quantification of generalized autonomic failure. Mayo Clin. Proc. 68:748, 1993. 45. Nukada, H., and Dyck, P. J.: Acute ischemia causes axonal stasis, swelling, attenuation, and secondary demyelination. Ann. Neurol. 22:311, 1987. 46. Pascoe, M. K., Low, P. A., Windebank, A. J., and Litchy, W. J.: Subacute Diabetic proximal neuropathy. Mayo Clin. Proc. 72:1123, 1997. 47. Raff, M. C., and Asbury, A. K.: Ischemic mononeuropathy and mononeuropathy multiplex in diabetes mellitus. N. Engl. J. Med. 279:17, 1968. 48. Raff, M. C., Sangalang, V., and Asbury, A. K.: Ischemic mononeuropathy multiplex associated with diabetes mellitus. Arch. Neurol. 18:487, 1968. 49. Root, H. F., and Rogers, M. H.: Diabetic neuritis with paralysis. N. Engl. J. Med. 202:1049, 1930. 50. Said, G., Goulen-Goeau, C., Lacroix, C., and Moulonguet, A.: Nerve biopsy findings in different patterns of proximal diabetic neuropathy. Ann. Neurol. 35:559, 1994. 51. Said, G., Lacroix, C., Lozeron, P., et al.: Inflammatory vasculopathy in multifocal diabetic neuropathy. Brain 126:376, 2003. 52. Sander, J. E., and Sharp, F. R.: Lumbosacral plexus neuritis. Neurology 31:470, 1981. 53. Schulz, A.: Brennende Schmerzen an Rumpf und Oberschenkel mit Bauchdeckenparesen. Dtsch. Med. Wochenschr. 97:1568, 1972. 54. Skanse, B., and Gydell, K.: A rare type of femoral-sciatic neuropathy in diabetes mellitus. Acta Med. Scand. 155:463, 1956. 55. Stewart, J. D.: Diabetic truncal neuropathy: topography of the sensory deficit. Ann. Neurol. 25:233, 1989. 56. Suarez, G. A., Giannini, C., Bosch, E. P., et al.: Immune brachial plexus neuropathy: suggestive evidence for an inflammatory immune pathogenesis. Neurology 46:559, 1996. 57. Subramony, S. H., and Wilbourn, A. J.: Diabetic proximal neuropathy: clinical and electromyographic studies. J. Neurol. Sci. 53:293, 1982. 58. Sun, S. F., and Streib, E. W.: Diabetic thoracoabdominal neuropathy: clinical and electrodiagnostic features. Ann. Neurol. 9:75, 1981. 59. Triggs, W. J., Young, M. S., Eskin, T., and Valenstein, E.: Treatment of idiopathic lumbosacral plexopathy with intravenous immunoglobulin. Muscle Nerve 20:244, 1997. 60. Verma, A., and Bradley, W. G.: High-dose intravenous immunoglobulin therapy in chronic progressive lumbosacral plexopathy. Neurology 44:248, 1994.

Radiculoplexus Neuropathies: Diabetic and Nondiabetic Varieties 61. Wessely, P., and Schnaberth, G.: Kasuistischer Beitrag zu einer seltenen Lokalisation diabetischer Neuropathie. Wien Klin. Wochenschr. 85:710, 1973. 62. Wilbourn, A. J.: Diabetic entrapment and compression neuropathies. In Dyck, P. J., and Thomas, P. K. (eds.): Diabetic Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 481, 1999.

2015

63. Williams, A. J.: Diabetic neuralgic amyotrophy. Postgrad. Med. J. 57:450–452, 1981. 64. Williams, I. R., and Mayer, R. F.: Subacute proximal diabetic neuropathy. Neurology 26:108, 1976. 65. Zochodne, D. W., Isaac, D., and Jones, C.: Failure of immunotherapy to prevent, arrest or reverse diabetic lumbosacral plexopathy. Acta Neurol. Scand. 107:299, 2003.

87 Neuropathies Associated with Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness DOUGLAS W. ZOCHODNE

Neuropathies and Hepatic Disorders Somatic and Autonomic Polyneuropathy Associated with Chronic Hepatic Disease Acute Inflammatory Demyelinating Polyradiculoneuropathy (Guillain-Barré Syndrome) Associated with Viral Hepatitis Neuropathies Associated with Hepatitis C Neuropathies Associated with Hepatitis B Sensory Neuropathy and Primary Biliary Cirrhosis

Cholestatic Hepatic Disease and Vitamin E Deficiency Other Neuropathies Associated with Hepatic Disease Neuropathies and Respiratory Disease Neuropathy and Chronic Obstructive Pulmonary Disease Neuropathies and Other Respiratory Disorders Mechanisms of Hypoxic Injury in Peripheral Nerve Disease Neuropathies and Renal Disease Polyneuropathy

This chapter reviews neuropathies associated with disorders of the kidney, liver, and lung as well as “critical illness,” the systemic inflammatory response syndrome (SIRS) associated with sepsis and multiple organ failure. How these disorders lead to neuropathy is often uncertain, and much of the information about them is published as case series. Polyneuropathy associated with renal failure may be diminishing in prevalence. Other neuropathies, such as those associated with hepatitis C, are garnering increased attention. As advances in the treatment of many acute and chronic medical disorders emerge, the residual disabilities posed by neuropathies have assumed more importance in patients’ lives.

NEUROPATHIES AND HEPATIC DISORDERS In the third edition of Peripheral Neuropathy, Asbury9 pointed out that separating neuropathies developing as a consequence of hepatic disease from those diseases that target liver and nerve together can be difficult. Examples not addressed in this section are the tandem disorders of

Etiology of Polyneuropathy of Renal Failure Mononeuropathies of Renal Failure Neuropathies and Critical Illness Definitions Critical Illness Polyneuropathy (Bolton’s Neuropathy) Possible Mechanisms of Critical Illness Polyneuropathy Differential Diagnosis of Weakness in Patients with Critical Illness Burn Neuropathy Hopkins Syndrome

liver and nerve that occur in patients suffering from alcohol abuse. Another example is porphyria, which is discussed in Chapter 80. This section highlights some specific and unique associations in which peripheral neuropathy appears to develop as a complication of a primary hepatic disorder. These include autonomic and somatic polyneuropathy of chronic liver disease, particularly advanced cirrhotic disease; acute inflammatory demyelinating polyradiculoneuropathy (Guillain-Barré syndrome) in association with viral hepatitis; polyneuropathy associated with hepatitis C and cryoglobulinemia; sensory neuropathy complicating primary biliary cirrhosis (PBC); and cholestatic liver disease associated with vitamin E deficiency.

Somatic and Autonomic Polyneuropathy Associated with Chronic Hepatic Disease Most series of neuropathy and hepatic cirrhosis patients include alcoholics (Table 87–1). The risk of polyneuropathy developing in nonalcoholic cirrhosis can therefore be difficult to extract. Overall, neuropathy probably occurs in the 2017

2018

Diseases of the Peripheral Nervous System

Table 87–1. Case Series of Patients with Chronic Hepatic Disease and Neuropathies

Disorder

Number (%) with Neuropathy Clinical Signs/ Electrophysiologic [Total Studied] Symptoms Abnormalities

Cirrhosis ⫾ alcohol 17 (89%) [19] Cirrhosis ⫾ alcohol 41 (71%) [58] Cirrhosis ⫾ alcohol [23] Cirrhosis ⫾ alcohol [33] Cirrhosis ⫾ alcohol 31 (91%) [34]

90% 22%/29%

68%

79% 43%

43%

Autonomic Involvement

Other

48% (DB, Abnormal QST Valsalva, tilt) (62%)—cooling or vibration 73.9% 67% VPT abnormal in 62% Abnormal ischemia VPT (55%)

References 155 46

197 57 88

DB ⫽ deep breathing; QST ⫽ quantitative sensation test; VPT ⫽ vibration perception threshold.

majority of patients with chronic liver disease. The ChildPugh classification156 is an index of worsening hepatic failure based on scored points for degree of encephalopathy, ascites score, bilirubin level, albumin level, and prothrombin time, resulting in a ranking of A, B, or C. A “C” ranking indicates a poor prognosis for recovery from surgical procedures. The severity of neuropathy appears to correlate with the ChildPugh severity of hepatic cirrhosis.46,57,154 Symptoms of neuropathy in patients with chronic hepatic disease include paresthesias, cramps, loss of sensation, and weakness of the limbs, especially the legs. Signs include weakness (particularly of the lower limbs), loss of deep tendon reflexes, and sensory loss.154 Electrophysiologic findings are mild slowing of motor and sensory conduction velocities and declines in the amplitudes of the compound muscle action potentials (CMAPs) or sensory nerve action potentials (SNAPs).88,154 Reductions in sural SNAP amplitudes, and then in peroneal CMAPs, are the most common abnormalities.46 In patients with chronic liver disease, Kardel and Nielsen identified abnormal vibration sensitivity on the toe and a slowing of the rise of vibration sensitivity threshold during ischemia.88 In one series, one third of patients had electrophysiologic evidence of superimposed carpal tunnel syndrome.46 Autonomic neuropathy may portend a more serious prognosis in patients with advanced hepatic disease.57,72,196 Vagal neuropathy (abnormal R-R intervals on deep breathing, with Valsalva maneuver, and with standing; abnormal blood pressure [BP] response to standing) was described71 in 45% of patients with chronic hepatic disease of varying cause (alcohol abuse, chronic active hepatitis, or PBC). A similar rate of autonomic neuropathy was reported by Thuluvath and Triger in 64 patients with liver disease, and its prevalence was independent of ethanol use.194 Kempler et al.90 examined autonomic testing (R-R interval changes with Valsalva maneuver, deep breathing, and standing;

BP changes in response to standing and handgrip) and corrected Q-T (QTc) interval prolongation in patients with liver disease with (n ⫽ 131) and without (n ⫽ 49) alcohol abuse. Abnormalities in autonomic testing (suggesting both sympathetic and parasympathetic dysfunction) correlated with QTc prolongation.

Acute Inflammatory Demyelinating Polyradiculoneuropathy (Guillain-Barré Syndrome) Associated with Viral Hepatitis Acute inflammatory demyelinating polyradiculoneuropathy (Guillain-Barré syndrome) is covered in depth in Chapter 99. This disorder may, however, occur following the onset of either hepatitis B or C.100

Neuropathies Associated with Hepatitis C Peripheral neuropathy may be a complication of hepatitis C viral infection in up to 8% of patients.163 While diffuse symmetrical sensory polyneuropathy is the most common pattern observed, mononeuritis multiplex occurs in a third or more of instances. There is considerable overlap in the pattern of presentation.32,124,191 Accumulated episodes of mononeuropathy may also eventually render a clinical picture of a severe symmetrical axonal polyneuropathy. There are sporadic reports of isolated brachial plexopathy in patients with hepatitis C.49 Cryoglobulinemia accompanies hepatitis C and neuropathy. Type I cryoglobulinemia results from a monoclonal expansion of B lymphocytes. Type II or mixed cryoglobulinemia arises from both a monoclonal (immunoglobulin M [IgM] ␬) and a polyclonal expansion of immunoglobulinproducing B lymphocytes. Type III cryoglobulinemia is associated with a polyclonal B lymphocyte population

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness

expansion only. Hepatitis C–related neuropathies usually accompany type II cryoglobulinemia, or less commonly type III cryoglobulinemia. While not present in all patients, cryoglobulinemia, both symptomatic and asymptomatic, has been associated with electrophysiologic evidence of neuropathy in 37% of patients.212 This suggests it is clearly a risk factor in hepatitis C infection for the development of peripheral nerve disease. Symptoms in patients with hepatitis C infection and neuropathy include paresthesias of the legs and feet, dysesthetic pain, cramps, difficulty walking, weakness, restless legs, sensory loss, exacerbation of symptoms in the cold, footdrop, fatigue, arthralgias, myalgias, Raynaud’s syndrome, dry eyes, and skin rash.41 Signs include muscle atrophy, weakness of one or both lower limbs, absence of one or both ankle reflexes, asymmetrical or symmetrical loss of sensation, sensory ataxia, purpura, digital ulcers, and livedo.39

2019

Electrophysiologic studies identify axonal damage in multiple nerve territories, with reduced or absent CMAPs and SNAPs. Loss of these potentials may be diffuse and most prominent in distal muscles, or multifocal. Similarly, denervation of muscles occurs, with fibrillation potentials and positive sharp waves recorded by electromyography (EMG). Muscles with partial denervation have reduced numbers of voluntary motor unit action potentials, or, somewhat later, enlarged remodeled motor units that develop from collateral sprouting of motor terminals. Mild slowing of conduction velocity occurs from the loss of large, more rapidly conducting myelinated fibers.41,66,186 Occasional patients have more prominent features of primary demyelination such as severe conduction slowing, temporal dispersion, or conduction block.191 Sural nerve biopsies in patients with hepatitis C–associated neuropathies (Fig. 87–1) show generalized large or small

FIGURE 87–1 Sural nerve biopsy from a patient with vasculitic mononeuritis multiplex, cryoglobulinemia, and hepatitis C, described by Khella et al.92 Left, Mononuclear infiltration of an epineurial arteriole. Right, Fiber loss and myelin debris (arrow). Scale ⫽ 10 ␮m. (From Khella, S. L., Frost, S., Hermann, G. A., et al.: Hepatitis C infection, cryoglobulinemia, and vasculitic neuropathy. Treatment with interferon alfa: case report and literature review. Neurology 45:407, 1995, with permission.)

2020

Diseases of the Peripheral Nervous System

fiber loss, epineurial perivascular inflammatory infiltrates, mononuclear cell infiltration of epineurial vessel walls, medium-size epineurial vessel wall necrosis, fibrinoid necrosis, enlarged capillaries, endothelial cell swelling, basal lamina reduplication, smooth muscle proliferation, obliterated epineurial vessels, axonal degeneration, and axonal regenerative clusters.32,66,163,191 Vascular wall necrosis may be absent.113 Immunohistochemical staining shows immunoglobulin G (IgG), IgM, and C3 in capillary walls186 and perivascular lymphocytes, including CD8-positive cells.66 In situ reverse transcriptase–polymerase chain reaction has detected inflammatory cells around epineurial vessels containing hepatitis C virus RNA sequences.32 Detection of viral sequences in whole-nerve biopsies as reported, however, may be complicated by contaminated viremic blood. Skin biopsies in patients with hepatitis C–associated polyneuropathy and purpura may also demonstrate vasculitis. Patients having hepatitis C, cryoglobulinemia, and neuropathy may be divided into a malignant category with mononeuritis multiplex resembling polyarteritis nodosa and a more benign category with symmetrical polyneuropathy.39 Patients with mononeuritis multiplex have severe multifocal and eventual diffuse weakness from infarction of nerve trunks supplying major muscle groups. These patients also have more frequent systemic disease, such as cerebral infarction, abdominal ischemia, or mesenteric infarction associated with abdominal arterial microaneurysms and renal infarction. The erythrocyte sedimentation rate (ESR) is elevated. On nerve biopsy, vasculitis involves mediumsized arteries (⬎70 ␮m in diameter) with vessel necrosis and perivascular neutrophils. Patients with hepatitis C and a benign symmetrical, largely sensory neuropathy may have skin purpura but a normal ESR and no evidence of systemic ischemia. Nerve biopsy shows perivascular infiltrates of inflammatory cells surrounding small arterioles, capillaries, and veins without vessel wall necrosis. In some patients, however, mild polyneuropathy may deteriorate into more serious mononeuritis multiplex. Nephrotic syndrome, membranoproliferative glomerulonephritis, rheumatoid factor, cryoglobulins, low C4 levels, and circulating immune complexes may accompany hepatitis C infection, cryoglobulinemia, and neuropathy.41,186 The ESR may be normal (see above),66 and antineutrophil cytoplasmic antibodies are usually absent. Additional central nervous system changes may include encephalopathy, optic neuropathy, and magnetic resonance imaging (MRI) abnormalities in the deep white matter, subcortical nuclei, and brainstem.191 To date, therapy of neuropathy associated with hepatitis C is based only on experience from case series. Standard approaches include prednisone, cytotoxic drugs such as cyclophosphamide, and plasma exchange.41 Chronic prednisone therapy has been associated with paradoxical worsening, rises in cryoglobulins, and increased serum viral RNA

in some patients.66,186,191 Interferon alfa may be more likely to help than plasma exchange39,66,92,113,191 and should be given to patients with severe polyarteritis-like mononeuritis multiplex if standard approaches fail.39,41 Patients require close follow-up while on interferon alfa, however, because of reports that the agent may worsen preexisting neuropathy or induce de novo polyneuropathy, cranial neuropathy, mononeuritis multiplex, or acute autonomic and sensory neuropathy.83,98,99,187 Serologic markers of chronic inflammation do not seem to predict how the neuropathy will fare. In some patients, long-term prognosis seems to most closely parallel that of renal involvement, if it is present.

Neuropathies Associated with Hepatitis B While there have been a number of reports of neuropathy, particularly mononeuritis multiplex (multiple mononeuropathy), associated with hepatitis B, earlier reports may not have excluded coincident hepatitis C infection.18 Polyarteritis nodosa in association with hepatitis B is discussed in Chapter 106. Two reports by Tsukada et al.197,198 address evidence that the hepatitis B virus may directly invade peripheral nerves. These reports described hepatitis B invasion in chronic relapsing polyneuropathy, radiculopathy, mononeuritis multiplex, or a steroid-responsive “Guillain-Barré syndrome” type of polyneuropathy, but clinical details were sketchy. Hepatitis B surface antigen was detected by immunofluorescent microscopy on the intima of vasa nervorum and within the endoneurium of sural nerves. Deposits of viral material were associated with IgG and C3. Sural nerve biopsies demonstrated large myelinated fiber loss with features of vasculitis, and electron-dense material, thought to represent immune complexes in the endoneurium and in endoneurial vessels. Liver histology identified changes of acute hepatitis, chronic active hepatitis, or chronic persistent hepatitis.

Sensory Neuropathy and Primary Biliary Cirrhosis PBC is an autoimmune inflammatory disorder of the intrahepatic biliary system. Features include hepatomegaly, jaundice, pruritus, hyperlipidemia, clubbing, cutaneous xanthomatosis, and antimitochondrial autoantibodies. PBC may be associated with a sensory polyneuropathy.193 Symptoms of neuropathy include painful paresthesias, burning or spontaneous pain, sensory loss, and loss of balance. Examination discloses loss of deep tendon reflexes, distal lower and upper limb sensory loss to all modalities, pseudoathetosis, ataxia, and rombergism.45,82 Electrophysiologic studies identify borderline low or absent SNAPs, normal motor conduction, and normal EMG.45,193 Cerebrospinal fluid (CSF) protein is normal or elevated. Nerve biopsies have distinctive abnormalities. Thomas and Walker193 described cells distended with cholesterol within and separating the

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness

layers of the perineurium, and in the adjacent endoneurium and epineurium. Distended lipid-laden cells were found in accumulations, or xanthomas, within fascicles and distorting them. Nerve xanthomas were similar to those infiltrating the skin. Despite these changes, the axon fiber density was preserved with only evidence of mild myelin abnormalities. Vasa nervorum were normal. Axonal degeneration and loss of myelinated and unmyelinated axons without xanthomas may also occur in patients with PBC and sensory neuropathy.45,82 One report described isolated xanthoma deposits in fascicles of unmyelinated fibers within the liver of a patient with cholestatic hepatitis.117 Finally, PBC may be associated with Sjögren’s syndrome and with vasculitic polyneuropathy.125 Autonomic function tests were abnormal in 77% of patients (deep breathing heart rate; standing 30:15 R-R interval and BP; Valsalva maneuver and handgrip BP) and with a prolonged QTc,89 an index associated with an increased risk of sudden death.62

Cholestatic Hepatic Disease and Vitamin E Deficiency In children with chronic cholestatic hepatic disease, a multisystem disorder involving the spinal cord, cerebellum, and peripheral nerves is associated with low serum levels of vitamin E.166,185 Fifty percent of children between ages 1 and 3 and all children beyond the age of 3 had neurologic abnormalities if they had cholestatic hepatic disease with vitamin E deficiency.177 These patients have visual loss, ophthalmoparesis, paresthesias, loss of deep tendon reflexes, ataxia, and abnormal position and vibration sensation.195 Pathologic studies identify degeneration of the cuneate and gracile nuclei, degeneration of spinocerebellar tracts, cerebellar atrophy, loss of sensory neurons in the dorsal root ganglia, loss of large peripheral nerve myelinated fibers, and pigmentary retinopathy.166,169 Specific changes in the gracile and cuneate nucleus of patients with congenital biliary atresia and cystic fibrosis are axonal spheroids, astrogliosis, and degeneration of the respective gracile and cuneate dorsal columns.185 The lesions in children have resembled those of rat models with vitamin E deficiency.153 Levels of tocopherol in peripheral nerve are reduced,195 and clinical improvement accompanies vitamin E supplementation.178 Similarly, patients who receive vitamin E are much less likely to demonstrate neuropathologic changes.

Other Neuropathies Associated with Hepatic Disease Navajo neuropathy is a progressive, probably autosomal recessive, severe sensory and motor polyneuropathy of unknown cause in Navajo children.5,173 Patients have progressive liver disease leading to cirrhosis and hepatic failure. The mean age of onset is 13 months.78 Infants

2021

develop weakness, muscle atrophy, motor delay, corneal anesthesia with ulceration, and loss of deep tendon reflexes. In older children, symptoms and signs of polyneuropathy develop later, after the age of 4, following the onset of hepatic disease. Signs include distal motor atrophy and weakness, loss of deep tendon reflexes, distal sensory loss, acral mutilation, and corneal scarring. Electrophysiologic studies show conduction velocity slowing. Nerve biopsies have generalized loss of myelinated fibers. Additional features are elevated CSF protein, susceptibility to infections, poor weight gain, short stature, sexual infantilism, and central nervous system white matter lesions on MRI. Autoimmune hepatitis as part of Sjögren’s syndrome and sensorimotor polyneuropathy has been described.80 There is a report linking liver hydatid cyst and axonal polyneuropathy.149

NEUROPATHIES AND RESPIRATORY DISEASE Isolated disease of the respiratory system is associated with neuropathy. A number of conditions, such as vasculitis or sarcoidosis, are examples of concurrent disease of lung and nerve. Similarly, in critical illness (discussed below), denervation of the respiratory muscles is associated with secondary respiratory failure and difficulty with weaning from the ventilator. These patients also have a high incidence of pulmonary sepsis and adult respiratory distress syndrome. In this section, however, we consider neuropathies that appear to develop secondary to a primary respiratory disorder, in most instances chronic obstructive pulmonary disease (COPD).

Neuropathy and Chronic Obstructive Pulmonary Disease Appenzeller et al.6 first described an association between COPD and polyneuropathy in eight patients. Most had lost weight before evaluation, raising a concern that they may have had a superimposed nutritional neuropathy, but none had malignancy. Since this initial report, several other case series have been described.55,85,128,140,146,150 Paramelle et al.150 noted a relationship between the duration of hypoxia and findings of polyneuropathy. Faden et al.55 described an association between the severity of neuropathy and cigarette smoking. Jarratt et al.85 described a mainly distal sensory, axonal neuropathy using electrophysiologic studies in 18% of 89 patients with COPD, with an additional 24% exhibiting isolated compression and traumatic mononeuropathies. Clinical findings of polyneuropathy, however, were present in only a third of those patients. There was a correlation between age and the presence of neuropathy. A multicenter study reported in two separate papers140,155 identified 8% of 151 COPD patients to have polyneuropathy, particularly in

2022

Diseases of the Peripheral Nervous System

the lower limbs. Peripheral neuropathy was more likely in older patients with more severe hypoxia. Ozge et al.146 identified evidence of polyneuropathy in 45% of 49 COPD patients, particularly in those with severe hypoxemia. Polyneuropathy was identified by an abnormal neurologic symptom score (34%), neurologic disability score (42%), or vibration perception threshold (94%). In the third edition of this text, Asbury discussed the development of polyneuropathy in patients treated with almitrine,9 a respiratory stimulant agent used for COPD. Distinguishing neuropathy secondary to almitrine from that directly attributable to respiratory disease was difficult. In COPD patients with neuropathy, there may be tingling, aches, pins and needles “numbness,” cramps, or no symptoms at all.140 Clinical signs of neuropathy are relatively modest and include mild weakness of intrinsic foot muscles, variable loss of deep tendon reflexes, and sensory loss in the feet to light touch, temperature, vibration, and position sensation.6,55 Electrophysiologic abnormalities are slowing of motor and sensory conduction velocity, decreased SNAP amplitudes, and occasional reductions of CMAP amplitudes, distal muscle fibrillation potentials, and abnormal voluntary motor unit potentials, indicating denervation with reinnervation.6,55,130,150,200 H-reflex latencies may be delayed.43 Nerve biopsies show axonal degeneration6 with loss of myelinated and unmyelinated fibers. There is only limited information describing demyelination. Stoebner et al.184 described abnormalities of microvessels in patients with COPD: narrowed lumina, thickened basement membranes, and pericyte debris. A subclinical autonomic neuropathy in patients with hypoxemic COPD may occur.183 Prolongation of the QTc interval and abnormal autonomic test results (heart rate changes in response to deep breathing, standing, and Valsalva maneuver; changes in BP responses to standing and sustained handgrip) correlated with the severity of hypoxemia and with mortality.

Neuropathies and Other Respiratory Disorders There is very little information about other types of respiratory disease and secondary polyneuropathy. Mayer et al.121 reported reduced mixed proximal median nerve compound action potentials and resistance to ischemic conduction failure in patients with obstructive sleep apnea and nocturnal hypoxemia.

Mechanisms of Hypoxic Injury in Peripheral Nerve Disease Hypoxia may account for peripheral nerve damage in respiratory failure. Several large surveys (described above) identified clear relationships between the presence of

polyneuropathy and the degree of hypoxemia. There are also parallels between hypoxia-related polyneuropathy and diabetic polyneuropathy wherein hypoxia from microangiopathy may cause fiber damage.50,131,199,219 Rats reared in a hypoxic environment have smaller myelinated fibers and an impairment in fiber myelination.17 Somewhat surprising is the relatively modest clinical neuropathy that has been encountered in severely hypoxemic COPD patients, perhaps suggesting that other factors in diabetes are required to generate axon damage. Hypoxia and diabetes are both associated with resistance to ischemic conduction failure.116 Other neuropathies, however, such as those associated with renal failure (see below) but without hypoxia, also have resistance.42 Hypoxia may contribute to axonal damage in patients with critical illness polyneuropathy (discussed below), in which tissue ischemia results from microcirculatory failure, or vasoplegia in septic shock. An index of tissue exposure to ischemia, however, did not correlate with the severity of critical illness polyneuropathy.217

NEUROPATHIES AND RENAL DISEASE An extensive literature on the relationship between renal disease and neuropathy (uremic neuropathy) exists. Most neurologists today report that they rarely observe this condition. More recent prospective studies of renal patients are not available. Dialysis is now initiated earlier in patients with kidney disease, and transplantation is more aggressively administered with better antirejection regimens.28 While hemodialysis itself has not clearly been shown to improve uremic neuropathy once it has developed, it is possible that prevention of neuropathy with more aggressive hemodialysis may have lessened its prevalence. Many patients with end-stage renal disease undergoing hemodialysis are also diabetic, but this discussion addresses neuropathy of nondiabetic renal failure. Original descriptions of uremic neuropathy were made by Marin and Tyler, who described a severe disease in two patients who had hereditary interstitial nephritis.119 These patients had subacute weakness; distal sensory loss to pinprick, cold, light touch, vibration, and position; and loss of deep tendon reflexes. There was distal-predominant neuropathy at autopsy, with loss of myelin sheaths out of proportion to the loss of silver-stained axons. There was no evidence of nerve inflammation or a microvascular abnormality. Hegstrom and colleagues observed uremic polyneuropathy during early dialysis usage in the 1960s,67,68 but initially were unsure whether the condition might be caused by dialysis. Against this possibility was the observation that polyneuropathy also developed in patients who had never been dialyzed. Asbury and colleagues subsequently coined the term uremic polyneuropathy in clinical and pathologic descriptions of patients with this disorder.10,11

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness

Polyneuropathy has been suggested as an index of the severity of renal failure and perhaps the adequacy of therapy for it.28 For example, Nielsen found a direct correlation between declines in motor conduction velocity and falls in creatinine clearance.135 Uremic polyneuropathy is generally associated with renal failure rendering a serum creatinine of 5 mg/dL or higher.15,28,159 The prevalence of clinical signs of polyneuropathy in patients undergoing hemodialysis has ranged between 10% and 83%, with larger series in the range of 50% to 60%.28 Using electrophysiologic studies as a more sensitive index, the prevalence of polyneuropathy has been somewhat higher, ranging from 57% to 100%. Children with renal failure have a similar prevalence of polyneuropathy.1,7 A series of 135 hemodialyzed patients from Italy15 identified clinical features of severe polyneuropathy in only 2.7% of those patients who had been receiving long-term dialysis (over 10 years) and in only 0.7% of all patients. Mild polyneuropathy was present in 57% of the hemodialysis patients, and 39% had no evidence of polyneuropathy. There was a higher incidence of polyneuropathy in males but no relationship between the severity of neuropathy and age, ethnic background, or type of renal disease.

2023

A

Polyneuropathy Symptoms of polyneuropathy associated with chronic renal failure are restless legs (51%), cramps (65%), weakness (12%), paresthesias (40%), dysesthesia (24%), pain (7%), and burning feet (8%).132 Other descriptors include painful tingling or electric feelings, unpleasant sensations induced by touching, bandlike constrictive feelings around the feet and ankles, and sensations of swelling or twisting feelings in the limbs.9 The symptom of burning feet described in early series of patients treated with hemodialysis may have been related to failure to supplement B vitamins in the dialysate.28 Signs of polyneuropathy in patients with renal failure include muscle atrophy (13%), weakness (14%), loss of deep tendon reflexes (28%), loss of vibration perception (38%), and hypesthesia (16%). Sensory loss develops to two-point discrimination, position sensation, and light touch, pain, and temperature sensation.20 Distal nerves (toes and feet) are involved first with loss of ankle reflexes and vibration perception. Some patients develop severe quadriplegia. The electrophysiologic features of uremic neuropathy include prolonged distal motor latencies, conduction velocity slowing, and declines in the amplitudes of CMAPs and SNAPs.28,136 While these abnormalities suggest axon loss, there is additional evidence of demyelination: slowing of conduction velocity that is greater than that expected with large axon loss alone, and proximal CMAP dispersion21,28,134,135 (Fig. 87–2). In some instances, motor conduction velocity may fall to 50% of the mean control value.20 While the relationship between the prevalence of nerve conduction abnormalities and clinical signs of polyneuropathy is relatively poor, the relationship between

B FIGURE 87–2 Examples of CMAPs recorded from thenar median-innervated muscles on stimulation of the median nerve at the wrist and elbow. Bars: 2 mV; 2 ms. A, Slowing of motor conduction velocity (longer latency, greater separation of the potentials) and dispersion of the potentials during progressive uremia after 1 month of hemodialysis (top) and 9 months later while the patient continued to be treated with hemodialysis (bottom). B, Median conduction improves after successful renal transplantation. Top, At 3 months prior to transplantation. Bottom, At 27 months after transplantation. The deterioration in A despite hemodialysis and improvement in B with transplantation illustrate changes resulting from segmental demyelination and remyelination, respectively. (From Bolton, C. F., and Young, G. B.: Neurological Complications of Renal Disease. Boston, Butterworths, 1990, with permission.)

peroneal motor nerve conduction velocity and creatinine clearance is more robust.137 Sural nerve conduction slowing may be the most sensitive electrophysiologic parameter of early polyneuropathy,2 and there may be slowing of conduction velocity before clinical symptoms and signs of

2024

Diseases of the Peripheral Nervous System

polyneuropathy appear. Bolton has suggested that peroneal motor conduction velocity is the most sensitive index in gauging the benefits of renal transplantation.20 In moderate or severe neuropathy, needle EMG identifies features of denervation, namely fibrillation potentials and positive sharp waves, accompanied by reduced numbers of voluntary motor unit potentials and enlarged polyphasic remodeled units indicating motor terminal sprouting. Denervated muscles in renal failure associated with diabetes may fail to exhibit fibrillation potentials and positive sharp waves despite otherwise severe denervation.26,27 This unusual finding can mislead clinicians in their evaluation of weakness in diabetic renal patients. It may be that the combination of renal disease and diabetes renders denervated single muscle fibers inexcitable and resistant to spontaneous firing of fibrillating muscles. The latencies of “long loop” F waves and H reflexes are prolonged in uremic neuropathy.2,147,148 Single-fiber EMG studies identify normal fiber density but increased jitter,95,192 suggesting early disease of distal motor terminals at neuromuscular junctions. Motor unit numbers are reduced, suggesting a failure of collateral muscle reinnervation.21 Somatosensory evoked potentials are delayed in peripheral segments.59,112,144,167,202 Excised sural nerve in vitro compound action potentials may have reduced A␣ myelinated potential or C unmyelinated potential amplitudes.51 Quantitative sensation testing vibration thresholds in patients with chronic renal failure are elevated.133,160,190 Cold perception may be impaired or instead may induce a paradoxical sensation of heat.211 Nerve biopsies in uremia identify distal-predominant axonal degeneration, loss of fibers (distal predominant), paranodal myelin retraction, segmental demyelination, segmental remyelination, and axonal atrophy.51,168 Loss of both myelinated and unmyelinated fibers has been reported. Electron microscopic studies may show stacks of empty Schwann cell processes without unmyelinated fibers, or collagen pockets.51 Segmental demyelination may be clustered on individual axons, rather than diffuse. This change may indicate that individual axons undergo atrophy in uremia, followed by myelin wrinkling, demyelination, and then axonal degeneration. Biopsies also show myelin sheaths with splitting and irregular contours. Axons may have abnormal mitochondria containing osmiophilic inclusions, clustering of neurofilaments, and lamellar and dense bodies within them. There are decreases in the ratio of the axon circumference to the number of myelin lamellae, indicating atrophy. Because axon–Schwann cell signaling is a complex “two-way” effort, clustered demyelination on individual axons may indicate that these axons fail to support their complement of Schwann cells. Autopsy studies by Asbury and colleagues10,11 have added further pathologic information. Axonal degeneration became more intense in

distal nerve segments, with relative sparing of nerve roots. There was also central chromatolysis of anterior horn cells and pallor of the spinal cord dorsal columns. Spinal fluid protein in uremic neuropathy is usually normal but occasionally can be elevated to 1 to 2 g/L.9 Although biochemical indices that reflect the severity of renal failure (e.g., serum creatinine level) correlate with those of neuropathy, no other laboratory abnormalities have been linked. There are some additional interesting features of polyneuropathy associated with renal failure. For example, deteriorating polyneuropathy has been viewed as an indication to increase the amount of dialysis, but there is little evidence to support this idea. Indeed, investigations have suggested the opposite, identifying continued deterioration in spite of ongoing dialysis, including specific regimens to eliminate “middle molecules” (see below). Dyck et al.52 compared neurologic disability, muscle strength, touch-pressure threshold, and electrophysiology (CMAPs, conduction velocities, distal motor latencies, SNAPs, and sensory latencies) in renal failure patients treated with either a shorter or longer hemodialysis regimen over 12 months. Despite more prominent symptoms of uremia that developed in patients treated with shorter hemodialysis programs, indices of polyneuropathy did not change. Dialysis improved conduction velocity in children7 but not in other studies.20,136 There were worsened clinical features but improved vibration perception in a study of patients undergoing hemodialysis compared to peritoneal dialysis.189 Rapidly progressive uremic polyneuropathies also occur. In the original report by Hegstrom et al.,67 a progressive ascending symmetrical neuropathy involving all extremities developed shortly after starting intensive hemodialysis, and it was associated with an elevated spinal fluid protein. Guillain-Barré syndrome was considered. Ropper165 described an accelerated neuropathy of renal failure in four patients undergoing chronic peritoneal dialysis. These patients had severe progression over days to weeks with loss of deep tendon reflexes and sensation. They had more prominent features of primary demyelination than that expected of uremia and had elevated CSF protein. One improved with more aggressive dialysis. Whether some such patients actually had coincident inflammatory demyelinating neuropathy in the setting of renal failure is uncertain. Along similar lines, Said et al.168 described three patients with electrophysiologic features of acute “axonal” polyneuropathy, but who both also had segmental demyelination and axonal degeneration or biopsy. Improvement appeared to be associated with hemodialysis. Renal transplantation is associated with improvement in uremic neuropathy.22,137 Initial improvement sometimes occurs rapidly in well-being, distal numbness and tingling, ability to perform fine finger tasks, and ability to walk. Early

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness

improvement is followed by more prolonged and gradual sustained improvement. There may also be an action tremor on recovery, resembling that of inflammatory demyelinating polyneuropathies.28 Clinical symptoms and electrophysiologic abnormalities worsen during rejection episodes, but may subsequently improve with repeat successful transplantation.20 Electrophysiologic studies show substantial improvements in distal latencies and conduction velocities after renal transplantation, but also improvement in CMAPs and SNAPs with lessened dispersion. Recovery also accompanies the disappearance of fibrillation potentials on needle EMG. While electrophysiologic studies have suggested that both remyelination and axonal regeneration occur after successful transplantation, improvement in demyelination is more prominent. For example, improvements in conduction velocity slowing and dispersion were greater than that of CMAP and SNAP amplitudes in a series of 17 transplanted patients20 (see Fig. 87–2). In addition, there can be irreversible axon loss and therefore only modest improvement in patients with severe polyneuropathy. Successful transplantation may improve deafness associated with renal failure.122,127,157 In children, transplantation improved ulnar and peroneal motor conduction velocities, but peroneal motor recovery occurred much more slowly, over a 3-year period.7 There has been less benefit, and instead continued progressive loss of motor fibers, despite renal transplantation in diabetic patients.181,201 Patients with both pancreatic and renal transplantation did not have further improvement compared to those with renal grafts alone.181 It may be that diabetic uremic patients gain more benefits by eliminating uremia than by correcting their diabetes. In contrast, autonomic function did not improve at all in diabetic patients over 2 years following combined transplantation.181 The importance of autonomic abnormalities in renal failure has been debated. Overt clinical abnormalities are uncommon.21 Solders et al.179 noted substantial loss of R-R interval variation at rest and in deep breathing in patients with progressive renal disease despite relatively modest somatic polyneuropathy. Abnormalities were not improved by dialysis. The same authors also reported that transplantation improved somatic neuropathy but not abnormalities of R-R intervals.180 Vita et al.204 reported that R-R intervals with rest, deep breathing, Valsalva maneuver, and tilt and BP changes with standing and sustained handgrip worsened in eight patients over 56 months and then, for unclear reasons, improved for 30 months after the dialysate was changed from acetate to bicarbonate. A variety of additional studies of autonomic function in renal patients have identified abnormalities in cardiovascular function with Valsalva maneuver, tilt-table testing, and sustained handgrip.13,40,54,64,91,94,108,131,138,164,182,210,215

2025

Etiology of Polyneuropathy of Renal Failure The mechanism of nerve dysfunction in renal failure is not established. There are reductions in nerve excitability, loss of nodal sodium permeability, and resistance to ischemic conduction failure.21,38,188,208 Experimental renal failure in rats is associated with an early (within 48 hours) decline in motor conduction velocity.188 Polyneuropathy from renal failure has been attributed to impaired clearance of “middle molecules” by dialysis membranes.12 These molecules, in the range of 0.5 to 500 kDa in size, are thought to directly influence axonal conduction and to be toxic to axons.8,28,34,35,97,118 Evidence for middle molecule neurotoxicity, however, has been conflicting. Hemodialysis techniques meant to improve middle molecule clearance have generated different results: improvement of neuropathy in some33 and no change in others.115,170 Evidence that the following factors may explain the development of uremic neuropathy has not been substantiated: urea, creatinine, parathyroid hormone, guanidine, phenols, and myo-inositol. Similarly, there is no firm evidence that ischemia contributes to the polyneuropathy of renal failure.

Mononeuropathies of Renal Failure The mononeuropathies, or focal neuropathies, associated with renal failure include carpal tunnel syndrome and ischemic forearm neuropathy associated with the placement of arteriovenous shunts or fistulas. Carpal tunnel syndrome in renal patients can be a source of chronic disability.44 Symptoms are upper limb pain, usually localized to the hand or radiating proximally; paresthesias and dysesthesias; sensory loss; and, in late disease, thenar muscle weakness and wasting. Symptoms of carpal tunnel syndrome and other mononeuropathies may sometimes worsen during hemodialysis. Carpal tunnel syndrome is probably not a consequence of vascular shunts placed in the same limb, because progressive damage in either limb can apparently continue despite carpal tunnel release. Such progression has been linked to deposition of beta2-microglobulin–associated amyloid both within nerve and in the flexor tendon ligament that roofs the carpal tunnel.44 Despite this problem, carpal tunnel surgery can nonetheless lessen shoulder pain and stiffness and is worth doing. As might be expected, the absence of amyloid in the flexor retinacula is associated with a better outcome.44 Plasma beta2-microglobulin may accumulate 40- to 60-fold in renal failure, and its clearance is particularly impaired in patients hemodialyzed with cuprophan membranes.60 In the author’s institution, cuprophan membranes are no longer in use. Accumulation of beta2-microglobulin, however, can also occur with other dialysis membranes, and with chronic ambulatory peritoneal dialysis.19,28,44,60,214

2026

Diseases of the Peripheral Nervous System

Ischemic forearm neuropathy can complicate upper limb vascular access surgery. It was originally described by Bolton and colleagues in two patients undergoing a bovine shunt inserted between the radial or brachial artery and the cephalic vein for dialysis access.23 In one patient with more severe involvement, the neuropathy involved ischemic axonal degeneration of median, ulnar, and radial nerve motor and sensory fibers distal to the elbow. Symptoms included hand weakness, numbness, pain (particularly during dialysis), and paresthesias. On examination, there were findings of paralysis of intrinsic hand and forearm muscles; sensory loss to all modalities except deep pain in distal median, ulnar, and radial territories; and loss of the brachioradialis reflex. Electrophysiologic studies disclosed absent median and ulnar CMAPs; absent median, ulnar, and radial SNAPs; and distal denervation on needle EMG. Blood flow to the thenar muscle was markedly reduced. Surgical closure of the arteriovenous conduit restored blood flow and improved the neuropathies, but only slowly and incompletely, as might be expected from infarction of multiple nerves. A second patient had paresthesias and numbness of the hand after a graft was placed connecting the brachial artery and cephalic vein. In this patient, SNAPs were diminished in amplitude but symptoms recovered spontaneously 6 weeks later without closure of the shunt. A similar neuropathy was described following a proximal brachial artery–antecubital vein fistula in diabetic patients.209 In patients with more distal upper limb fistulas connecting the radial artery and cephalic vein in the forearm (Brescia-Cimino fistula), there may be median and ulnar neuropathies but they are uncommon.28 Patients have pain, numbness, paresthesias, and weakness.93 That local arteriovenous shunting could lead to ischemic neuropathy was supported by experimental work. Most models of experimental ischemia involve acute interruption of the local nerve arterial supply, embolization of the nerve, or application of an epineurial vasoconstrictor, such as endothelin.96,142,144,152,171,220 Diabetic nerves are more susceptible to acute ischemia.141,217,218 These models demonstrate multifocal loss of fibers in the ischemic territory that likely develops within a 3-hour ischemic time frame.143,171,217 Sladky et al.175 created a model of more chronic experimental endoneurial ischemia in sciatic nerves of rat hind limbs by creating local arteriovenous fistulas. This lesion rendered a 50% to 75% reduction in endoneurial blood flow and a 25% to 30% fall in conduction velocity. Other mononeuropathies described in patients with renal failure include those related to compression of the ulnar nerve at the elbow or the peroneal nerve at the fibular head in bedridden patients. Compartment syndromes involving peripheral nerves may occur from hemorrhage (e.g., iliopsoas hematoma involving the lumbosacral plexus) associated with coagulopathy of renal disease.28

NEUROPATHIES AND CRITICAL ILLNESS Despite initial descriptions and investigations in 1984,24 the concept that a previously unrecognized polyneuropathy could develop in the setting of sepsis and multiple organ failure generated skepticism among neurologists. A number of reasons probably accounted for its slow acceptance. These included the lesser likelihood that neurologists would visit intensive care units, the difficulty in clinically examining invasively monitored unresponsive patients on ventilators, reluctance to move electrophysiologic testing equipment into the electrically “noisy” intensive care environment, and the lack of an obvious explanation for the disorder. While the description of critical illness polyneuropathy is now reasonably rooted in the literature, with reports from a variety of investigators,4,81,107,139,205,206 its explanation remains unclear. Indeed, the only feature routinely associated with its development has been that of the syndrome of sepsis and multiple organ failure, identified now as the “systemic inflammatory response syndrome.” In addition to discussing current information about this entity, the differential diagnosis of flaccid weakness of patients in intensive care units, including critical illness myopathy, also is addressed.

Definitions “Critical illness” is defined as a condition that occurs within the context of SIRS, defined by Bone and others16,30,31 to include two of the following findings: (1) temperature of greater than 38° C or less than 36° C; (2) heart rate greater than 90 beats/min; (3) respiratory rate of greater than 20 breaths/min or arterial partial pressure of carbon dioxide less than 32 mm Hg; and (4) white blood cell count greater than 12,000 cells/␮L or less than 4000 cells/␮L, or greater than 10% immature (band) forms. SIRS can develop following several types of insult, including infection, pancreatitis, ischemia, multiple trauma with tissue injury, hemorrhagic shock, immune-mediated organ injury, and exogenous administration of inflammatory mediators. SIRS associated with infection is termed sepsis. Some associated terms can cause confusion. For example, bacteremia refers to the presence of bacteria in circulating blood, but not all patients with sepsis have positive blood cultures (only 45% in one study30). The term septicemia has been discarded. Septic shock refers to sepsis with hypotension, and multiple organ dysfunction syndrome (MODS) is the presence of altered organ function in an acutely ill patient with inability to maintain homeostasis. Secondary MODS occurs within the context of SIRS. Patients meeting these criteria are very ill and have a mortality rate approaching 60% within an intensive care unit. In the past, units that had a less aggressive philosophical approach to managing this patient population for prolonged periods likely did not see many instances of critical illness polyneuropathy. Similarly, it may

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness

be that newer emerging approaches to managing septic shock prevent MODS and associated critical illness neuropathy. It also may be that some patients who would have developed critical illness polyneuropathy during severe irreversible MODS now avoid futile ongoing aggressive management through protocols that predict whether it will help them.

Critical Illness Polyneuropathy (Bolton’s Neuropathy) In 1984, Bolton and colleagues24 described five patients without known neuromuscular disease admitted to an intensive care unit who developed an acute and severe axonal polyneuropathy in the context of overwhelming sepsis and multiple organ failure. These patients were referred to the neurology service when there was difficulty weaning them from mechanical ventilators, and they were noted to have flaccid and areflexic limbs. Concerns were raised that the disease might be a variant of Guillain-Barré syndrome. In subsequent work, however,25 this disorder was distinguished from Guillain-Barré syndrome on the basis of its predisposing illness, clinical picture, electrophysiologic features, and CSF analysis (discussed below). Critical illness polyneuropathy develops in patients without known neuropathic disease, such as those admitted with multiple trauma or head injuries, those undergoing repair of a ruptured aortic aneurysm or bypass surgery, and those with myocardial infarction and other conditions. These are patients who meet the criteria for SIRS or sepsis as discussed above. All the patients described in original reports were admitted to intensive care units and ventilated, but it is likely that a version of critical illness polyneuropathy can occur in patients not admitted to intensive care units. It has also been described in children.172 After admission to the intensive care unit, patients have a high likelihood of developing nosocomial sepsis or MODS, and there is an overall mortality rate of 58%.216 Symptoms of neuropathy are usually not detected because patients are intubated and sedated and may have encephalopathy. In some instances, patients have reported painful paresthesias during their recovery, and many patients note muscle weakness during their rehabilitation. Cranial nerves have generally been spared on clinical (but not electromyographic) examination, an important clinical finding distinguishing this disorder from Guillain-Barré syndrome and myopathies. Limbs have varying degrees of flaccid weakness and atrophy. Distal atrophy may be obscured by limb edema, common in patients with critical illness. The majority of patients have loss of previously normal deep tendon reflexes. Disappearance of these reflexes is an important sign that neuropathy has developed. Unfortunately, this simple neurologic test is still routinely omitted when patients are admitted with other severe medical and surgical disorders. While loss of reflexes is commonly blamed on sedatives and analgesics, there is little evidence that doses of agents

2027

commonly employed in the intensive care unit (excepting those used for barbiturate coma) can eliminate them. Neuromuscular blocking agents and propofol can cloud the bedside neurologic examination by rendering flaccid areflexic paralysis. Sensory examination is often unreliable. Distal loss to all sensory modalities in a stocking-and-glove pattern may become evident only if the patient has a clear sensorium, or perhaps later, during recovery. Patients with critical illness polyneuropathy may be awake, yet fail to grimace or withdraw their limbs in response to nail-bed pressure. Electrophysiologic studies are the diagnostic standard for identifying this disorder.81,216 Nerve conduction studies identify declines of the amplitudes of the CMAPs, with relatively preserved motor conduction velocities, distal motor latencies, and F-wave latencies. Specific recordings of diaphragm CMAPs from phrenic nerve stimulation also identify declines in amplitude.213 Not observed are features of primary demyelination such as conduction block or temporal dispersion. There is variable loss of SNAP amplitudes with preserved sensory nerve conduction velocities. One problem complicating interpretation of high-gain surface electrode recordings of sensory potentials is edema. Edema may increase the distance between the nerve and the surface recording electrode, accounting for some loss of amplitude. Near-nerve needle recordings have been used to help with this problem. Needle EMG has identified fibrillation potentials and positive sharp waves, often intense, in distal muscles, proximal muscles, intercostal muscles, and the diaphragm. Hund et al.81 also reported denervation in facial muscles. Voluntary motor unit action potentials may not be recruited, may be small and polyphasic (nascent units, compatible with very early denervation and reinnervation), or may appear normal. CSF has generally been normal or has had only mild elevations in protein. In Guillain-Barré syndrome, protein elevations are considerably higher if the patient is tested at least a week after the onset of the condition. In a prospective series of 132 patients at the University of Western Ontario referred for electrophysiologic studies (mean interval of 40 days after admission), Zifko et al.214 diagnosed critical illness polyneuropathy in 62, or approximately half of the referred cohort. All diagnosed patients had reduced CMAPs and evidence of denervation by EMG. SNAP amplitudes were reduced in 71%. None had evidence of a neuromuscular transmission defect. There was no correlation between the severity of polyneuropathy and the use of steroids or nondepolarizing muscle blocking agents. Acute axonal degeneration and axon loss is observed in nerve biopsies and autopsy studies from patients with critical illness polyneuropathy.81,216 Axonal degeneration involves both motor and sensory fibers, and both proximal and distal nerve segments (e.g., peroneal nerve or cauda equina) (Fig. 87–3). The pattern of fiber loss and axon

2028

Diseases of the Peripheral Nervous System

FIGURE 87–3 Histologic changes in critical illness polyneuropathy. A, Transverse Epon section illustrating severe myelinated fiber loss and active axonal degeneration of the superficial peroneal nerve. B, Central chromatolysis of lumbar anterior horn cells indicating interruption of distal motor axons. C, Longitudinal Epon-embedded section of superficial peroneal nerve illustrating myelinated fiber loss and fibers degenerating into myelin ovoids from axonal degeneration. D, Longitudinal paraffin section of cauda equina showing myelin ovoid digestion chambers indicating axonal degeneration of proximal nerve. E, Transverse paraffin section of iliopsoas muscle illustrating scattered fiber atrophy from denervation. (A, B, C, and E from Zochodne, D. W., Bolton, C. F., Wells, G. A., et al.: Critical illness polyneuropathy, a complication of sepsis and multiple organ failure. Brain 110:819, 1987, with permission.)

breakdown has been diffuse rather than multifocal but is often very severe, involving large proportions of sampled nerve trunks. Similarly, teased fiber analysis and semithin sections of plastic-embedded nerves identify axonal degeneration, without evidence of segmental or paranodal demyelination. Teased fiber specimens graded according to Dyck’s classification (see Chapter 32) had 4.5% to 94.4% type E fibers (axonal degeneration). No inflammatory infiltrates have been identified. Muscle biopsies have shown features of acute, or sometimes (depending on the stage of the disease when sampled) chronic, neurogenic atrophy with scattered muscle fiber atrophy or grouped atrophy, respectively. These changes have been observed in a variety of proximal limb, distal limb, trunk, and respiratory muscles, including tibialis anterior, quadriceps, peroneus longus, psoas, deltoid, triceps, extensor hallucis longus, diaphragm, and intercostal muscles.216 In some pathologic studies,

features of both neurogenic atrophy and myopathy are observed, likely in patients having suffered a combination of critical illness neuropathy and primary myopathy (discussed below). In other series, features of primary myopathy do not accompany neurogenic atrophy.81,216 Occasional single necrotic muscle fibers may reflect acute denervation or myopathy. Chromatolysis of anterior horn cells and loss of dorsal root ganglion cells also occur (Fig. 87–3).

Possible Mechanisms of Critical Illness Polyneuropathy The mechanism of critical illness polyneuropathy is unknown. It may arise as an end-organ complication of sepsis and SIRS.207,216 Neuropathy develops rapidly, within 14 days of admission to the intensive care unit, and it is not clearly linked to specific drug use. No correlation between

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness

aminoglycoside usage or levels and the severity of critical illness polyneuropathy was identified, and not all patients were exposed to aminoglycosides prior to diagnosis.216 Aminoglycosides have been linked to a presynaptic neuromuscular transmission defect, implying that the agents may be toxic to distal motor terminals.123 While all patients with critical illness polyneuropathy had evidence of sepsis, specific pathogens were not linked to it. Unlike the myopathies discussed below, it did not develop in association with the use of nondepolarizing muscle blocking agents or corticosteroids. The electrophysiologic features and clinical context distinguish it from Guillain-Barré syndrome.25 Autopsy studies have not identified features of Wernicke’s disease to suggest thiamine deficiency, and most patients with critical illness polyneuropathy develop it despite ongoing parenteral vitamin B therapy. Indeed, in one series,216 all patients received intensive parenteral nutrition, including B vitamins, chromium, linolenic acid, and linoleic acid, deficiencies of which have been linked to neuropathy,77,86 and all had normal phosphate levels.3 An appropriate model of critical illness polyneuropathy has not emerged and may not be simple to generate. Such a model would require SIRS, and multiple organ dysfunction lasting over 2 weeks. Diffuse tissue ischemia and defects of oxygen extraction, as well as mitochondrial function, are well described in tissues subjected to sepsis.58,65,73 In two patients studied in the early recovery phase of critical illness polyneuropathy, severe depletions of muscle phosphocreatine were identified using nuclear magnetic resonance spectroscopy, reflecting loss of high-energy substrate. Levels rose as the patient improved.29 Such changes paralleled those measured in a rat model of acute sepsis.84 There are multiple molecular players in the development of sepsis, each of which may have a role in influencing the development of neuropathy. Among these are interleukins 1, 2, 4, 6, and 8 and tumor necrosis factor-␣, platelet activating factor, leukotrienes, thromboxane A2, prostaglandins E2 and I2, interferon-␥, granulocyte-macrophage colony-stimulating factor, endothelin-1, nitric oxide (NO), complement fragments C3a and C5a, adhesion molecules, vascular endothelial growth factor, bradykinin, thrombin, fibrin, plasminogen activator inhibitors, endorphins, heat shock proteins, transforming growth factor-␤1, and chemokines.31,56 Rises in NO synthesis associated with SIRS105 might contribute to axonal degeneration. NO is generated in tissues through the conversion of arginine to citrulline by one of three enzymes known as nitric oxide synthases (NOSs).36,37 Two isoforms of this enzyme are constitutive: neuronal and endothelial NOS. Neuronal NOS is elaborated by primary sensory, hippocampal, cerebellar, spinal cord, and other neurons. Its levels rise in sensory neurons after axotomy,203 and it is thought to play a role in brain infarction.53 Endothelial NOS (eNOS) elaborates NO that in turn diffuses to arteriolar and venular smooth muscle cells, increasing levels of cyclic GMP and in turn causing

2029

their hyperpolarization and relaxation. NO maintains vasodilator “tone” as an endothelium-derived relaxing factor, and mediates responses to vasodilator molecules such as acetylcholine. The most relevant NOS isoenzyme in sepsis and SIRS may be inflammatory/inducible NOS (iNOS). iNOS is generated by macrophages, smooth muscle, and other cells in response to lipopolysaccharide released from bacterial cell walls, and other factors. It is this form of NOS that likely contributes to “vasoplegia,” or paralysis of arteriolar vasoreactivity, in septic shock.151 iNOS is also activated by interleukin-1, a cytokine linked to muscle proteolysis in sepsis.14,47,58,61 Excessive amounts of NO released from iNOS may diffuse widely and combine with the superoxide anion to generate peroxynitrite (ONOO⫺). Peroxynitrite is a very potent oxidizing species that damages thiol-containing proteins (e.g., SNAP-25 and GAP43/B50) and DNA. Exactly how it might damage peripheral nerves, however, is uncertain. Both iNOS and eNOS are found in injured peripheral nerves.109,111,221 iNOS appears to be elaborated by Schwann cells and macrophages during wallerian-like degeneration. eNOS, in turn, appears to have a unique localization in early axonal end bulbs that develop after axons are transected. Both isoenzymes likely influence nerve blood flow after injury. NO may also have direct actions on axons. For example, application of NO donor molecules to myelinated axons in vitro elicited acute transient conduction block, but also axonal degeneration.176 NO may also inhibit regeneration of axons and may contribute to the development of neuropathic pain.74,110,161,176,222 Whether there is a specific role for NO in generating critical illness neuropathy, however, is unknown. Within intensive care units, NO is now added to ventilatory gas mixtures of patients in respiratory failure.

Differential Diagnosis of Weakness in Patients with Critical Illness There is an important differential diagnosis to be considered when examining patients with diffuse weakness in intensive care units. Inappropriate labeling of critical illness polyneuropathy may misdirect care and interventions. A similar note of caution and attention to the differential diagnosis should be considered when examining children with critical illness.172 Patients with diffuse flaccid limb weakness but intact bulbar function may have high cervical cord myelopathy. Myelopathy (e.g., C1-C2 dislocation resulting from rheumatoid arthritis) may initially present with reflex loss and absent upper motor neuron findings, whereas a sensory level may be obscured if the patient is sedated or encephalopathic. Exclusion of cervical myelopathy using MRI studies is recommended in quadriparetic patients. Similarly, patients with central pontine myelinolysis may present with diffuse severe areflexic quadriparesis. A history of ethanol abuse or alterations in serum sodium level is an important clue that this condition may be

2030

Diseases of the Peripheral Nervous System

present. The author has examined two patients who had combined critical illness polyneuropathy and central pontine myelinolysis. Motor neuron disease, myasthenia gravis, polymyositis, myotonic dystrophy, Lambert-Eaton myasthenic syndrome, and Guillain-Barré syndrome may all present with respiratory failure. Clinical suspicion, careful bedside examination (e.g., tongue and limb fasciculations in motor neuron disease, and ptosis and ophthalmoplegia in myasthenia gravis; these conditions often have some historical clues of their presence before respiratory failure), and electrophysiologic testing are important in identifying these conditions. Severe hypokalemia may present with acute ascending weakness and respiratory failure. Severe thiamine deficiency may be associated with a subacute ascending motor paralysis. An area of ongoing debate is how often an acquired primary myopathy may account for flaccid weakness of patients in intensive care units. Such myopathies may clinically and electrophysiologically resemble critical illness polyneuropathy.101,104,158,223 For this condition we have suggested the term critical illness myopathy, although several other terms have been used: floppy person syndrome, necrotizing myopathy of the intensive care unit, thick filament myopathy, steroid-induced tetraplegia, and acute quadriplegic myopathy.69,75,87,101–104,162,174 Five patients described by Gorson and Ropper as having “acute respiratory failure neuropathy” with flaccid paralysis following respiratory failure, steroid use, and neuromuscular junction blocking agents were likely instances of critical illness myopathy.63 It is also likely that other series contain patients who may in fact have a primary myopathy rather than critical illness polyneuropathy.205 The similarities to critical illness neuropathy are its development in patients admitted to intensive care units without preexisting neuromuscular disease, its presentation with diffuse flaccid limb weakness and areflexia, and some electrophysiologic features: loss of CMAPs, abnormal spontaneous activity (fibrillations and positive sharp waves) on needle EMG, and relative selectivity for motor involvement. Some authors have suggested that myopathy is now more frequently encountered than critical illness polyneuropathy. Both polyneuropathy and myopathy may contribute to diffuse weakness of intensive care unit patients not previously diagnosed with a neuromuscular condition.145 Patients with acute quadriplegic myopathy or critical illness myopathy have most often (but not always76) received either nondepolarizing muscle blocking agents, high doses of corticosteroids, or both. Patients with critical illness polyneuropathy have a more prominent association with episodes of SIRS and multiple organ failure. We have suggested that patients given vecuronium infusions represent an extreme variant of myopathy with muscle rhabdomyolysis (severe weakness, high creatine kinase levels, myoglobinuric renal failure, and severe muscle fiber necrosis on histologic examination). In other patients there may be less necrosis, and biopsy of muscle may identify loss of thick

(myosin) filaments, type II fiber atrophy, and regenerating muscle fibers. Patients given infusions of vecuronium or pancuronium, especially in renal failure, may accumulate active metabolites, resulting in blockade of their neuromuscular junctions lasting up to 7 days. Patients with myopathy in the intensive care unit have more cranial nerve involvement (ophthalmoplegia occasionally174), and spared sensation if the patient can be examined (e.g., grimacing in response to nail-bed pressure without withdrawal). Myopathy may occur sooner in patients admitted to intensive care units than polyneuropathy. Patients with myopathy have relatively preserved SNAPs and less intense abnormal spontaneous activity (fibrillations and positive sharp waves) on EMG. The author223 encountered a 34-year-old male patient who developed severe critical illness myopathy related to steroid and muscle blockade within the first few weeks of his hospitalization in the intensive care unit for multiple trauma. Serial studies later indicated the subsequent development of a severe axonal motor and sensory polyneuropathy following repeated episodes of sepsis and multiple organ failure. Loss of thick filaments is observed in non-necrotic muscle fibers of patients with critical illness myopathy. What generates thick filament loss, however (e.g., superimposition of steroids on a paralyzed muscle), and how truly specific it is are not clear.48 Myosin may be targeted for specific degradation through the ubiquitin pathway.126 Larsson et al.106 noted that muscles with myosin protein loss also had declines in their messenger RNA, suggesting that the disorder involves rapid degradation of myosin but also failure to replenish it. Two patients that they studied with neuropathy did not have myosin loss.

Burn Neuropathy Henderson et al.70 reported evidence of polyneuropathy among patients with burns. Of 249 patients with varying burn severity, 36 had signs and symptoms of neuropathy that included weakness, sensory loss, tingling, and loss of hearing. Forty-four patients had motor conduction velocity slowing. More detailed electrophysiologic and histologic details were not provided. Marquez et al.120 identified 19 patients with signs and symptoms of neuropathy among 800 patients admitted to a tertiary care burn center. Eleven of these patients had severe burns of greater than 20% of body surface area. Most (69%) had mononeuritis multiplex; only one had an axonal motor and sensory polyneuropathy. Among the mononeuropathies, the median nerve was involved proximal to the carpal tunnel (10 of 13 median neuropathies), the ulnar nerve almost always at the elbow level (9 of 10 ulnar neuropathies), the radial nerve in the forearm (8 of 10 radial neuropathies), and the peroneal nerve at the knee. Most (69%) mononeuropathies were in the area of the maximal thermal burn. The authors postulated that burns cause neuropathy by direct injury, through occlusion of vasa nervorum,

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness

or indirectly, perhaps through a disseminated neurotoxin. Generalized polyneuropathy not directly at burn sites may represent variations of critical illness polyneuropathy.

Hopkins Syndrome Hopkins syndrome is distinct from critical illness polyneuropathy or critical illness myopathy.79,114,172 This disorder has been described in children and is thought to arise from localized damage to anterior horn cells. It develops within 2 weeks (1 to 11 days) of an episode of status asthmaticus, often associated with sepsis with corticosteroid use. The disorder most often renders monoplegia, but diplegia and hemiparesis have also been described. Recovery is limited. There may be mild CSF protein elevation and CSF pleocytosis. Segmental denervation is identified by EMG and is thought to represent a localized anterior horn cell disorder. Viral etiologies have been considered.

ACKNOWLEDGMENTS Brenda Boake provided expert secretarial assistance. D. W. Z. is a Senior Medical Scholar of the Alberta Heritage Foundation for Medical Research. Dr. Chris White kindly read the paper critically.

REFERENCES 1. Ackil, A. A., Shahani, B. T., and Young, R. R.: Sural nerve conduction studies and late responses in children undergoing hemodialysis. Arch. Phys. Med. Rehabil. 62:487, 1981. 2. Ackil, A. A., Shahani, B. T., Young, R. R., and Rubin, N. E.: Late response and sural conduction studies: usefulness in patients with chronic renal failure. Arch. Neurol. 38:482, 1981. 3. Agusti, A. G., Torres, A., Estopa, R., and Agustividal, A.: Hypophosphatemia as a cause of failed weaning: the importance of metabolic factors. Crit. Care Med. 12:142, 1984. 4. Alonso Alonso, P., Monasterio Chicharro, F., Lopez-Messa, J. B., et al.: Critically ill polyneuropathy. Crit. Care Med. 15:1162, 1987. 5. Appenzeller, O., Kornfeld, M., and Snyder, R.: Acromutilating, paralyzing neuropathy with corneal ulceration in Navajo children. Arch. Neurol. 33:733, 1976. 6. Appenzeller, O., Parks, R. D., and MacGee, J.: Peripheral neuropathy in chronic disease of the respiratory tract. Am. J. Med. 44:873, 1968. 7. Arbus, G. S., Barnor, N. A., Hsu, A. C., et al.: Effect of chronic renal failure, dialysis and transplantation on motor nerve conduction velocity in children. Can. Med. Assoc. J. 113:517, 1975. 8. Asaba, H., Alvestrand, A., Furst, P., and Bergstrom, J.: Clinical implications of uremic middle molecules in regular hemodialysis patients. Clin. Nephrol. 19:179, 1983.

2031

9. Asbury, A. K.: Neuropathies with renal failure, hepatic disorders, chronic respiratory insufficiency and critical illness. In Dyck, P. J., Thomas, P. K., Griffin, J.W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1251, 1993. 10. Asbury, A. K., Victor, M., and Adams, R. D.: Uremic polyneuropathy. Trans. Am. Neurol. Assoc. 87:100, 1962. 11. Asbury, A. K., Victor, M., and Adams, R. D.: Uremic polyneuropathy. Arch. Neurol. 8:413, 1963. 12. Babb, A. L., Popovich, R. P., Christopher, T. G., and Scribner, B. H.: The genesis of the square meter-hour hypothesis. Trans. Am. Soc. Artif. Intern. Organs 17:81, 1971. 13. Bach, C., Iaina, A., and Eliahou, H. E.: Autonomic nervous system disturbance in patients on chronic hemodialysis. Isr. J. Med. Sci. 9:761, 1979. 14. Baracos, V., Rodemann, H. P., Dinarello, C. A., and Goldberg, A. L.: Stimulation of muscle protein degradation and prostaglandin E2 release by leukocytic pyrogen (interleukin-1): a mechanism for the increased degradation of muscle proteins during fever. N. Engl. J. Med. 308:553, 1983. 15. Bazzi, C., Pagani, C., Sorgato, G., et al.: Uremic polyneuropathy: a clinical and electrophysiological study in 135 shortand long-term hemodialyzed patients. Clin. Nephrol. 35:176, 1991. 16. Benjamin, E., Leibowitz, A. B., Oropello, J., and Iberti, T. J.: Systemic hypoxic and inflammatory syndrome: an alternative designation for “sepsis syndrome.” Crit. Care Med. 20:680, 1992. 17. Benstead, T. J., Dyck, P. J., and Low, P.: Chronic hypoxia induces selective maldevelopment of peripheral myelin in rat. J. Neuropathol. Exp. Neurol. 47:599, 1988. 18. Berkompas, R. J.: A case of peripheral neuropathy and hepatitis. J. Tenn. Med. Assoc. 83:191, 1990. 19. Blumberg, A., and Burgi, W.: Behavior of beta 2-microglobulin in patients with chronic renal failure undergoing hemodialysis, hemodiafiltration and continuous ambulatory peritoneal dialysis (CAPD). Clin. Nephrol. 27:245, 1987. 20. Bolton, C. F.: Electrophysiologic changes in uremic neuropathy after successful renal transplantation. Neurology 26:152, 1976. 21. Bolton, C. F.: Peripheral neuropathies associated with chronic renal failure. Can. J. Neurol. Sci. 7:89, 1980. 22. Bolton, C. F., Baltzan, M. A., and Baltzan, R. B.: Effects of renal transplantation on uremic neuropathy: a clinical and electrophysiologic study. N. Engl. J. Med. 284:1170, 1971. 23. Bolton, C. F., Driedger, A. A., and Lindsay, R. M.: Ischaemic neuropathy in uraemic patients caused by bovine arteriovenous shunt. J. Neurol. Neurosurg. Psychiatry 42:810, 1979. 24. Bolton, C. F., Gilbert, J. J., Hahn, A. F., and Sibbald, W. J.: Polyneuropathy in critically ill patients. J. Neurol. Neurosurg. Psychiatry 47:1223, 1984. 25. Bolton, C. F., Laverty, D. A., Brown, J. D., et al.: Critically ill polyneuropathy: electrophysiological studies and differentiation from Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 49:563, 1986. 26. Bolton, C. F., McKeown, M. J., Chen, R., et al.: Subacute uremic and diabetic polyneuropathy. Muscle Nerve 20:59, 1997.

2032

Diseases of the Peripheral Nervous System

27. Bolton, C. F., Remtulla, H., Toth, B., et al.: Distinctive electrophysiological features of denervated muscle in uremic patients. J. Clin. Neurophysiol. 14:539, 1997. 28. Bolton, C. F., and Young, G. B.: Neurological Complications of Renal Disease. Boston, Butterworths, 1990. 29. Bolton, C. F., Young, G. B., and Zochodne, D. W.: Neurological changes during severe sepsis. In Dobb, G., Bion, J., Bjardhi, H., and Billinger, R. P. (eds.): Current Topics in Intensive Care. London, W. B. Saunders, p. 180, 1994. 30. Bone, R. C.: Let’s agree on terminology: definitions of sepsis. Crit. Care Med. 19:973, 1991. 31. Bone, R. C.: The sepsis syndrome: definition and general approach to management. Clin. Chest Med. 17:175, 1996. 32. Bonetti, B., Scardoni, M., Monaco, S., et al.: Hepatitis C virus infection of peripheral nerves in type II cryoglobulinaemia. Virchows Arch. 434:533, 1999. 33. Botella, J., Gea, T., Sanz-Guajardo, D., et al.: Blood levels of middle molecules and their effects on motor nerve conduction velocity. Artif. Organs 4(Suppl.):151, 1981. 34. Boudet, J., Man, N. K., Cueille, G., et al.: Relationship between plasma concentration of middle molecular weight fraction b, motor nerve conduction velocity, and plasma creatinine in experimental chronic renal failure rats. Artif. Organs 4(Suppl.):115, 1981. 35. Braguer, D., Chauvet-Monges, A. M., Sari, J. C., et al.: Inhibition in vitro of the polymerization of tubulin by uremic middle molecules: corrective effect of isaxonine. Clin. Nephrol. 20:149, 1983. 36. Bredt, D. S.: Targeting nitric oxide to its targets. Proc. Soc. Exp. Biol. Med. 211:41, 1996. 37. Bredt, D. S., and Snyder, S. H.: Nitric oxide: a physiologic messenger molecule. Annu. Rev. Biochem. 63:175, 1994. 38. Brismar, T.: Changes in electrical threshold in human peripheral neuropathy. J. Neurol. Sci. 68:215, 1985. 39. Cacoub, P., Maisonobe, T., Thibault, V., et al.: Systemic vasculitis in patients with hepatitis C. J. Rheumatol. 28:109, 2001. 40. Campese, V. M., Romoff, M. S., Levitan, D., et al.: Mechanisms of autonomic nervous system dysfunction in uremia. Kidney Int. 20:246, 1981. 41. Caniatti, L. M., Tugnoli, V., Eleopra, R., et al.: Cryoglobulinemic neuropathy related to hepatitis C virus infection: clinical, laboratory and neurophysiological study. J. Peripher. Nerv. Syst. 1:131, 1996. 42. Castaigne, P., Cathala, H. P., Beaussart-Boulenge, L., and Petrover, M.: Effect of ischaemia on peripheral nerve function in patients with chronic renal failure undergoing dialysis treatment. J. Neurol. Neurosurg. Psychiatry 35:631, 1972. 43. Chandran, A. P., Singal, K. K., Mahajan, S. K., et al.: H-reflex latency and sensory conduction in the assessment of neuropathy in patients of chronic obstructive lung disease. Electromyogr. Clin. Neurophysiol. 31:99, 1991. 44. Charra, B., Calemard, E., Uzan, M., et al.: Carpal tunnel syndrome, shoulder pain and amyloid deposits in long-term haemodialysis patients. Proc. Eur. Dial. Transplant Assoc. Eur. Ren. Assoc. 21:291, 1985. 45. Charron, L., Peyronnard, J. M., and Marchand, L.: Sensory neuropathy associated with primary biliary cirrhosis: histologic and morphometric studies. Arch. Neurol. 37:84, 1980.

46. Chaudhry, V., Corse, A. M., O’Brian, R., et al.: Autonomic and peripheral (sensorimotor) neuropathy in chronic liver disease: a clinical and electrophysiologic study. Hepatology 29:1698, 1999. 47. Clowes, G. H. Jr., George, B. C., Villee, C. A. Jr., and Saravis, C. A.: Muscle proteolysis induced by a circulating peptide in patients with sepsis or trauma. N. Engl. J. Med. 308:545, 1983. 48. Danon, M. J., and Carpenter, S.: Myopathy with thick filament (myosin) loss following prolonged paralysis with vecuronium during steroid treatment. Muscle Nerve 14:1131, 1991. 49. Del Giudice, G., Galli, M., Chemotti, M., et al.: Brachial paresis complicating acute non-A, non-B hepatitis. Boll. Ist. Sieroter. Milan 63:499, 1984. 50. Dyck, P. J.: Hypoxic neuropathy: does hypoxia play a role in diabetic neuropathy? The 1988 Robert Wartenberg lecture. Neurology 39:111, 1989. 51. Dyck, P. J., Johnson, W. J., Lambert, E. H., and O’Brien, P. C.: Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin. Proc. 46:400, 1971. 52. Dyck, P. J., Johnson, W. J., Lambert, E. H., et al.: Comparison of symptoms, chemistry and nerve function to assess adequacy of hemodialysis. Neurology 29:1361, 1979. 53. Eliasson, M. J., Huang, Z., Ferrante, R. J., et al.: Neuronal nitric oxide synthase activation and peroxynitrite formation in ischemic stroke linked to neural damage. J. Neurosci. 19:5910, 1999. 54. Ewing, D. J., and Winney, R.: Autonomic function in patients with chronic renal failure on intermittent haemodialysis. Nephron 15:424, 1975. 55. Faden, A., Mendoza, E., and Flynn, F.: Subclinical neuropathy associated with chronic obstructive pulmonary disease. Arch. Neurol. 38:639, 1981. 56. Fink, M. P.: Prostaglandins and sepsis: still a fascinating topic despite almost 40 years of research. Am. J. Physiol. Lung Cell. Mol. Physiol. 281:L534, 2001. 57. Fleckenstein, J. F., Frank, S., and Thuluvath, P. J.: Presence of autonomic neuropathy is a poor prognostic indicator in patients with advanced liver disease. Hepatology 23:471, 1996. 58. Fry, D. E.: Multiple system organ failure. Surg. Clin. North. Am. 68:107, 1988. 59. Ganji, S., and Mahajan, S.: Changes in short-latency somatosensory evoked potentials during hemodialysis in chronic renal failure. Clin. Electroencephalogr. 14:202, 1983. 60. Gejyo, F., Homma, N., Suzuki, Y., and Arakawa, M.: Serum levels of beta 2-microglobulin as a new form of amyloid protein in patients undergoing long-term hemodialysis. N. Engl. J. Med. 314:585, 1986. 61. Goldberg, A. L., Baracos, V., Rodemann, P., et al.: Control of protein degradation in muscle by prostaglandins, Ca2⫹ and leukocytic pyrogen (interleukin 1). Fed. Proc. 43:1301, 1984. 62. Gonin, J. M., Kadrofske, M. M., Schmaltz, S., et al.: Corrected Q-T interval prolongation as diagnostic tool for assessment of cardiac autonomic neuropathy in diabetes mellitus. Diabetes Care 13:68, 1990.

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness 63. Gorson, K. C., and Ropper, A. H.: Acute respiratory failure neuropathy: a variant of critical illness polyneuropathy. Crit. Care Med. 21:267, 1993. 64. Gulati, P. D., Goel, M. K., Kalra, J. M., et al.: Autonomic neuropathy in chronic uraemics. J. Indian Med. Assoc. 74:45, 1980. 65. Haupt, M. T., Balaan, M. R., Barber, M. E., and Carlson, R.: Effects of serum of patients with advanced sepsis on mitochondrial metabolism. Am. Rev. Respir. Dis. 135:A83, 1987. 66. Heckmann, J. G., Kayser, C., Heuss, D., et al.: Neurological manifestations of chronic hepatitis C. J. Neurol. 246:486, 1999. 67. Hegstrom, R. M., Murray, J. S., Pendras, J. P., et al.: Hemodialysis in the treatment of chronic uremia. Am. Soc. Artif. Intern. Organs 7:136, 1961. 68. Hegstrom, R. M., Murray, J. S., Pendras, J. P., et al.: Two year’s experience with periodic hemodialysis in the treatment of chronic uremia. Am. Soc. Artif. Intern. Organs 8:266, 1962. 69. Helliwell, T. R., Coakley, J. H., Wagenmakers, J. M., et al.: Necrotizing myopathy in critically ill patients. J. Pathol. 164:307, 1991. 70. Henderson, B., Koepke, G. H., and Feller, I.: Peripheral polyneuropathy among patients with burns. Arch. Phys. Med. Rehabil. 52:149, 1971. 71. Hendrickse, M. T.: QT interval, autonomic neuropathy, and alcoholic liver disease. Lancet 342:61, 1993. 72. Hendrickse, M. T., Thuluvath, P. J., and Triger, D. R.: Natural history of autonomic neuropathy in chronic liver disease. Lancet 339:1462, 1992. 73. Hersch, M., Troster, M., Bersten, A., et al.: Histopathological evidence of tissue ischemia in a hyperdynamic and nonhypotensive septic animal model [abstract]. Crit. Care Med. 16:421, 1988. 74. Hess, D. T., Patterson, S. I., Smith, D. S., and Skene, J. H.: Neuronal growth cone collapse and inhibition of protein fatty acylation by nitric oxide. Nature 366:562, 1993. 75. Hirano, M., Ott, B. R., Raps, E. C., et al.: Acute quadriplegic myopathy: a complication of treatment with steroids, nondepolarizing blocking agents, or both. Neurology 42:2082, 1992. 76. Hoke, A., Rewcastle, N. B., and Zochodne, D. W.: Acute quadriplegic myopathy unrelated to steroids or paralyzing agents: quantitative EMG studies. Can. J. Neurol. Sci. 26:325, 1999. 77. Holman, R. T., Johnson, S. B., and Hatch, T. F.: A case of human linolenic acid deficiency involving neurological abnormalities. Am. J. Clin. Nutr. 35:617, 1982. 78. Holve, S., Hu, D., Shub, M., et al.: Liver disease in Navajo neuropathy. J. Pediatr. 135:482, 1999. 79. Hopkins, I. J.: A new syndrome: poliomyelitis-like illness associated with acute asthma in childhood. Aust. Paediatr. J. 10:273, 1974. 80. Hoshino, S., Yoshizawa, T., Hayashi, A., et al.: A combination of autoimmune hepatitis, sensory-dominant peripheral neuropathy, and primary Sjogren’s syndrome in the same patient: a rare association. J. Med. 30:83, 1999. 81. Hund, E. F., Fogel, W., Krieger, D., et al.: Critical illness polyneuropathy: clinical findings and outcomes of a frequent cause of neuromuscular weaning failure. Crit. Care Med. 24:1328, 1996.

2033

82. Illa, I., Graus, F., Ferrer, I., and Enriquez, J.: Sensory neuropathy as the initial manifestation of primary biliary cirrhosis. J. Neurol. Neurosurg. Psychiatry 52:1307, 1989. 83. Irioka, T., Yamada, M., Yamawaki, M., et al.: Acute autonomic and sensory neuropathy after interferon alpha-2b therapy for chronic hepatitis C. J. Neurol. Neurosurg. Psychiatry 70:408, 2001. 84. Jacobs, D. O., Maris, J., Fried, R., et al.: In vivo phosphorus 31 magnetic resonance spectroscopy of rat hind limb skeletal muscle during sepsis. Arch. Surg. 123:1425, 1988. 85. Jarratt, J. A., Morgan, C. N., Twomey, J. A., et al.: Neuropathy in chronic obstructive pulmonary disease: a multicentre electrophysiological and clinical study. Eur. Respir. J. 5:517, 1992. 86. Jeejeebhoy, K. N., Chu, R. C., Marliss, E. B., et al.: Chromium deficiency, glucose intolerance, and neuropathy reversed by chromium supplementation in a patient receiving long term total parenteral nutrition. Am. J. Clin. Nutr. 30:531, 1977. 87. Kaplan, P. W., Rocha, W., Sanders, D. B., et al.: Acute steroidinduced tetraplegia following status asthmaticus. Pediatrics 78:121, 1986. 88. Kardel, T., and Nielsen, V. K.: Hepatic neuropathy: a clinical and electrophysiological study. Acta Neurol. Scand. 50:513, 1974. 89. Kempler, P., Varadi, A., Kadar, E., and Szalay, F.: Autonomic and peripheral neuropathy in primary biliary cirrhosis: evidence of small sensory fibre damage and prolongation of the QT interval. J. Hepatol. 21:1150, 1994. 90. Kempler, P., Varadi, A., and Szalay, F.: Autonomic neuropathy and prolongation of QT-interval in liver disease. Lancet 340:318, 1992. 91. Kersh, E. S., Kronfield, S. J., Unger, A., et al.: Autonomic insufficiency in uremia as a cause of hemodialysis-induced hypotension. N. Engl. J. Med. 290:650, 1974. 92. Khella, S. L., Frost, S., Hermann, G. A., et al.: Hepatitis C infection, cryoglobulinemia, and vasculitic neuropathy: treatment with interferon alfa: case report and literature review. Neurology 45:407, 1995. 93. Knezevic, W., and Mastaglia, F. L.: Neuropathy associated with Brescia-Cimino arteriovenous fistulas. Arch. Neurol. 41:1184, 1984. 94. Koch, K. M., Baldamus, C. A., Ernst, W., et al.: Autonomic circulatory regulation in uremia [in German]. Klin. Wochenschr. 58:1037, 1980. 95. Konishi, T., Nishitani, H., and Motomura, S.: Single fiber electromyography in chronic renal failure. Muscle Nerve 5:458, 1982. 96. Korthals, J. K., and Wisniewski, H. M.: Peripheral nerve ischemia. Part 1. Experimental model. J. Neurol. Sci. 24:65, 1975. 97. Kumegawa, M., Hiramatsu, M., Yamada, T., and Yajima, T.: Effects of intermediate-sized molecular components in uremic sera on nerve tissues in vitro. Brain Res. 198:234, 1980. 98. La Civita, L., Pasero, G., and Ferri, C.: Drs. La Civita, Pasero, and Ferri reply. J. Rheumatol. 24:1656, 1997. 99. La Civita, L., Zignego, A. L., Lombardini, F., et al.: Exacerbation of peripheral neuropathy during alphainterferon therapy in a patient with mixed cryoglobulinemia and hepatitis B virus infection. J. Rheumatol. 23:1641, 1996.

2034

Diseases of the Peripheral Nervous System

100. Lacaille, F., Zylberberg, H., Hagege, H., et al.: Hepatitis C associated with Guillain-Barre syndrome. Liver 18:49, 1998. 101. Lacomis, D., Giuliani, M. J., Van Cott, A., and Kramer, D. J.: Acute myopathy of intensive care: clinical, electromyographic, and pathological aspects. Ann. Neurol. 40:645, 1996. 102. Lacomis, D., Petrella, T., and Giuliani, M. J.: Causes of neuromuscular weakness in the intensive care unit: a study of ninety-two patients. Muscle Nerve 21:610, 1998. 103. Lacomis, D., Smith, T. W., and Chad, D. A.: Acute myopathy and neuropathy in status asthmaticus: case report and literature review. Muscle Nerve 16:84, 1993. 104. Lacomis, D., Zochodne, D. W., and Bird, S. J.: Critical illness myopathy. Muscle Nerve 23:1785, 2000. 105. Landry, D. W., and Oliver, J. A.: The pathogenesis of vasodilatory shock. N. Engl. J. Med. 345:588, 2001. 106. Larsson, L., Li, X., Edstrom, L., et al.: Acute quadriplegia and loss of muscle myosin in patients treated with nondepolarizing neuromuscular blocking agents and corticosteroids: mechanisms at the cellular and molecular levels. Crit. Care Med. 28:34, 2000. 107. Leijten, F. S. S., and de Weerd, A.: Critical illness polyneuropathy, facts and controversies. J. Peripher. Nerv. Syst. 1:28, 1996. 108. Levitan, D., Massry, S. G., Romoff, M. S., and Campese, V. M.: Autonomic nervous system dysfunction in patients with acute renal failure. Am. J. Nephrol. 2:213, 1982. 109. Levy, D., Hoke, A., and Zochodne, D. W.: Local expression of inducible nitric oxide synthase in an animal model of neuropathic pain. Neurosci. Lett. 260:207, 1999. 110. Levy, D., Tal, M., Hoke, A., and Zochodne, D. W.: Transient action of the endothelial constitutive nitric oxide synthase (ecNOS) mediates the development of thermal hypersensitivity following peripheral nerve injury. Eur. J. Neurosci. 12:2323, 2000. 111. Levy, D., and Zochodne, D. W.: Local nitric oxide synthase activity in a model of neuropathic pain. Eur. J. Neurosci. 10:1846, 1998. 112. Lewis, E. G., Dustman, R. E., and Beck, E. C.: Visual and somatosensory evoked potentials characteristics of patients undergoing hemodialysis and kidney transplantation. Electroencephalogr. Clin. Neurophysiol. 44:223, 1978. 113. Lidove, O., Cacoub, P., Maisonobe, T., et al.: Hepatitis C virus infection with peripheral neuropathy is not always associated with cryoglobulinaemia. Ann. Rheum. Dis. 60:290, 2001. 114. Liedholm, L. J. A., Eeg-Olofsson, O., Ekenberg, B. E. K., et al.: Acute postasthmatic amyotrophy (Hopkins’ syndrome). Muscle Nerve 17:769, 1994. 115. Lindsay, R. M., Bolton, C. F., Clark, W. F., and Linton, A. L.: The effect of alterations of uremic retention products upon platelet and peripheral nerve function. Clin. Nephrol. 19:110, 1983. 116. Low, P. A., Lagerlund, T. D., and McManis, P. G.: Nerve blood flow and oxygen delivery in normal, diabetic, and ischemic neuropathy. Int. Rev. Neurobiol. 31:355, 1989. 117. Ludwig, J., Dyck, P. J., and LaRusso, N. F.: Xanthomatous neuropathy of liver. Hum. Pathol. 13:1049, 1982. 118. Man, N. K., Cueille, G., Zingraff, J., et al.: Uremic neurotoxin in the middle molecular weight range. Artif. Organs 4:116, 1980.

119. Marin, O. S. M., and Tyler, H. R.: Hereditary interstitial nephritis associated with polyneuropathy [abstract]. Neurology 11:999, 1961. 120. Marquez, S., Turley, J. J. E., and Peters, W. J.: Neuropathy in burn patients. Brain 116:471, 1993. 121. Mayer, P., Dematteis, M., Pepin, J. L., et al.: Peripheral neuropathy in sleep apnea: a tissue marker of the severity of nocturnal desaturation. Am. J. Respir. Crit. Care Med. 159:213, 1999. 122. McDonald, T. J., Zincke, H., Anderson, C. F., and Ott, N. T.: Reversal of deafness after renal transplantation in Alport’s syndrome. Laryngoscope 88:38, 1978. 123. McQuillen, M. P., Cantor, H. E., and O’Rourke, J. R.: Myasthenic syndrome associated with antibiotics. Arch. Neurol. 18:402, 1968. 124. Mendez, P., Saeian, K., Reddy, K. R., et al.: Hepatitis C, cryoglobulinemia, and cutaneous vasculitis associated with unusual and serious manifestations. Am. J. Gastroenterol. 96:2489, 2001. 125. Min, J. K., Park, K. S., Yu, W. J., et al.: Systemic mononuclear inflammatory vasculopathy associated with Sjogren’s syndrome in a patient with primary biliary cirrhosis. Korean J. Intern. Med. 15:89, 2000. 126. Minetti, C., Hirano, M., Morreale, G., et al.: Ubiquitin expression in acute steroid myopathy with loss of myosin thick filaments. Muscle Nerve 19:94, 1996. 127. Mitschke, H., Schmidt, P., Kopsa, H., and Zazgornik, J.: Reversible uremic deafness after successful renal transplantation. N. Engl. J. Med. 292:1062, 1975. 128. Moore, N., Lerebours, G., Senant, J., et al.: Peripheral neuropathy in chronic obstructive lung disease. Lancet 2:1311, 1985. 129. Naik, R. B., Mathias, C. J., Wilson, C. A., et al.: Cardiovascular and autonomic reflexes in haemodialysis patients. Clin. Sci. 60:165, 1981. 130. Narayan, M., and Ferranti, R.: Nerve conduction impairment in patients with respiratory insufficiency and severe chronic hypoxemia. Arch. Phys. Med. Rehabil. 59:188, 1978. 131. Newrick, P. G., Wilson, A. J., Jakubowski, J., et al.: Sural nerve oxygen tension in diabetes. Br. Med. J. (Clin. Res. Ed.) 293:1053, 1986. 132. Nielsen, V. K.: The peripheral nerve function in chronic renal failure. II. Intercorrelation of clinical symptoms and signs and clinical grading of neuropathy. Acta Med. Scand. 190:113, 1971. 133. Nielsen, V. K.: The peripheral nerve function in chronic renal failure. IV. An analysis of the vibratory perception threshold. Acta Med. Scand. 191:287, 1972. 134. Nielsen, V. K.: The peripheral nerve function in chronic renal failure. V. Sensory and motor conduction velocity. Acta Med. Scand. 194:445, 1973. 135. Nielsen, V. K.: The peripheral nerve function in chronic renal failure. VI. The relationship between sensory and motor nerve conduction and kidney function, azotemia, age, sex, and clinical neuropathy. Acta Med. Scand. 194:455, 1973. 136. Nielsen, V. K.: The peripheral nerve function in chronic renal failure. VII. Longitudinal course during terminal renal failure and regular hemodialysis. Acta Med. Scand. 195:155, 1974.

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness 137. Nielsen, V. K.: The peripheral nerve function in chronic renal failure. VIII. Recovery after renal transplantation: clinical aspects. Acta Med. Scand. 195:163, 1974. 138. Nies, A. S., Robertson, D., and Stone, W. J.: Hemodialysis hypotension is not the result of uremic peripheral autonomic neuropathy. J. Lab. Clin. Med. 94:395, 1979. 139. Nijeholt, J. L. A., and Troost, J.: Critical illness polyneuropathy. In Vinken, P. J., Bruyn, G. W., and Klawans, H. L. (eds.): Neuropathies (Handbook of Clinical Neurology, rev. ser. 7, Vol. 51). Amsterdam, Elsevier, p. 575, 1987. 140. Nowak, D., Bruch, M., Arnaud, F., et al.: Peripheral neuropathies in patients with chronic obstructive pulmonary disease: a multicenter prevalence study. Lung 168:43, 1990. 141. Nukada, H.: Increased susceptibility to ischemic damage in streptozocin diabetic nerve. Diabetes 35:1058, 1986. 142. Nukada, H., and Dyck, P. J.: Microsphere embolization of nerve capillaries and fiber degeneration. Am. J. Pathol. 115:275, 1984. 143. Nukada, H., Dyck, P. J., and Karnes, J. L.: Spatial distribution of capillaries in rat nerves: correlation to ischemic damage. Exp. Neurol. 87:369, 1985. 144. Obeso, J. A., Marti-Masso, J. F., Asin, J. L., et al.: Conduction velocity through the somesthetic pathway in chronic renal failure. J. Neurol. Sci. 43:439, 1979. 145. Op de Coul, A. A. W., Lambregts, P. C. L. A., Koeman, J., et al.: Neuromuscular complications in patients given Pavulon (pancuronium bromide) during artificial ventilation. Clin. Neurol. Neurosurg. 87:17, 1985. 146. Ozge, A., Atis, S., and Sevim, S.: Subclinical peripheral neuropathy associated with chronic obstructive pulmonary disease. Electromyogr. Clin. Neurophysiol. 41:185, 2001. 147. Panayiotopoulos, C. P., and Lagos, G.: Tibial nerve H-reflex and F-wave studies in patients with uremic neuropathy. Muscle Nerve 3:423, 1980. 148. Panayiotopoulos, C. P., and Scarpalezos, S.: F-wave studies on the deep peroneal nerve. Part 2.—1. Chronic renal failure. 2. Limb-girdle muscular dystrophy. J. Neurol. Sci. 31:331, 1977. 149. Paolino, E., Granieri, E., and Tugnoli, V.: Peripheral neuropathy as a manifestation of hydatid disease: a case report. Acta Neurol. Scand. 9:256, 1987. 150. Paramelle, B., Vila, A., Pollak, P., et al.: Incidence of polyneuropathies in chronic obstructive bronchopneumopathies. Presse Med. 15:563, 1986. 151. Parratt, J. R.: Nitric oxide in sepsis and endotoxaemia. J. Antimicrob. Chemother. 41(Suppl. A):31, 1998. 152. Parry, G. J., and Brown, M. J.: Arachidonate-induced experimental nerve infarction. J. Neurol. Sci. 50:123, 1981. 153. Pentschew, A., and Schwarz, K.: Systemic axonal dystrophy in vitamin E deficient adult rats: with implication in human neuropathology. Acta Neuropathol. 1:313, 1962. 154. Perretti, A., Gentile, S., Balbi, P., et al.: Peripheral neuropathy in liver cirrhosis: a clinical and electrophysiological study. Ital. J. Gastroenterol. 27:349, 1995. 155. Pfeiffer, G., Kunze, K., Bruch, M., et al.: Polyneuropathy associated with chronic hypoxaemia: prevalence in patients with chronic obstructive pulmonary disease. J. Neurol. 237:230, 1990. 156. Pugh, R. N., Murray-Lyon, I. M., Dawson, J. L., et al.: Transection of the oesophagus for bleeding oesophageal varices. Br. J. Surg. 60:646, 1973.

2035

157. Quick, C. A.: Hearing loss in patients with dialysis and renal transplants. Ann. Otol. Rhinol. Laryngol. 85:776, 1976. 158. Ramsay, D. A., Zochodne, D. W., Robertson, D. M., et al.: A syndrome of acute severe muscle necrosis in intensive care unit patients. J. Neuropathol. Exp. Neurol. 52:387, 1993. 159. Raskin, N. H., and Fishman, R. A.: Neurologic disorders in renal failure (second of two parts). N. Engl. J. Med. 294:204, 1976. 160. Read, D. J., Feest, T. G., and Holman, R. H.: Vibration sensory threshold: a guide to adequacy of dialysis? Proc. Eur. Dial. Transplant Assoc. 19:253, 1983. 161. Redford, E. J., Kapoor, R., and Smith, K. J.: Nitric oxide donors reversibly block axonal conduction: demyelinated axons are especially susceptible. Brain 120:2149, 1997. 162. Rich, M. M., Raps, E. C., and Bird, S. J.: Distinction between acute myopathy syndrome and critical illness polyneuropathy. Mayo Clin. Proc. 70:198, 1995. 163. Ripault, M. P., Borderie, C., Dumas, P., et al.: Peripheral neuropathies and chronic hepatitis C: a frequent association? [in French]. Gastroenterol. Clin. Biol. 22:891, 1998. 164. Rockel, A., Hennemann, H., Sternagel-Haase, A., and Heidland, A.: Uraemic sympathetic neuropathy after haemodialysis and transplantation. Eur. J. Clin. Invest. 9:23, 1979. 165. Ropper, A. H.: Accelerated neuropathy of renal failure. Arch. Neurol. 50:536, 1993. 166. Rosenblum, J. L., Keating, J. P., Prensky, A. L., and Nelson, J. S.: A progressive neurologic syndrome in children with chronic liver disease. N. Engl. J. Med. 304:503, 1981. 167. Rossini, P. M., Treviso, M., Di Stefano, E., and Di Paolo, B.: Nervous impulse propagation along peripheral and central fibres in patients with chronic renal failure. Electroencephalogr. Clin. Neurophysiol. 56:293, 1983. 168. Said, G., Boudier, L., Selva, J., et al.: Different patterns of uremic polyneuropathy: clinicopathologic study. Neurology 33:567, 1983. 169. Satya-Murti, S., Howard, L., Krohel, G., and Wolf, B.: The spectrum of neurologic disorder from vitamin E deficiency. Neurology 36:917, 1986. 170. Savazzi, G. M., Buzio, C., and Migone, L.: Lights and shadows on the pathogenesis of uremic polyneuropathy. Clin. Nephrol. 18:219, 1982. 171. Schmelzer, J. D., Zochodne, D. W., and Low, P. A.: Ischemic and reperfusion injury of rat peripheral nerve. Proc. Natl. Acad. Sci. U. S. A. 86:1639, 1989. 172. Sheth, R. D., and Bolton, C. F.: Neuromuscular complications of sepsis in children. J. Child Neurol. 10:346, 1995. 173. Singleton, R., Helgerson, S. D., Snyder, R. D., et al.: Neuropathy in Navajo children: clinical and epidemiologic features. Neurology 40:363, 1990. 174. Sitwell, L. D., Weinshenker, B. G., Monpetit, V., and Reid, D.: Complete ophthalmoplegia as a complication of acute corticosteroid and pancuronium associated myopathy. Neurology 41:921, 1991. 175. Sladky, J. T., Tschoepe, R. L., Greenberg, J. H., and Brown, M. J.: Peripheral neuropathy after chronic endoneurial ischemia. Ann. Neurol. 29:272, 1991. 176. Smith, K. J., Kapoor, R., Hall, S. M., and Davies, M.: Electrically active axons degenerate when exposed to nitric oxide. Ann. Neurol. 49:470, 2001.

2036

Diseases of the Peripheral Nervous System

177. Sokol, R. J., Guggenheim, M. A., Heubi, J. E., et al.: Frequency and clinical progression of the vitamin E deficiency neurologic disorder in children with prolonged neonatal cholestasis. Am. J. Dis. Child. 139:1211, 1985. 178. Sokol, R. J., Guggenheim, M. A., Iannaccone, S. T., et al.: Improved neurologic function after long-term correction of vitamin E deficiency in children with chronic cholestasis. N. Engl. J. Med. 313:1580, 1985. 179. Solders, G., Persson, A., and Gutierrez, A.: Autonomic dysfunction in non-diabetic terminal uraemia. Acta Neurol. Scand. 71:321, 1985. 180. Solders, G., Persson, A., and Wilczek, H.: Autonomic system dysfunction and polyneuropathy in nondiabetic uremia: a one-year follow-up study after renal transplantation. Transplantation 41:616, 1986. 181. Solders, G., Wilczek, H., Gunnarsson, R., et al.: Effects of combined pancreatic and renal transplantation on diabetic neuropathy: a two-year follow-up study. Lancet 2:1232, 1987. 182. Soriano, G., and Eisinger, R. P.: Abnormal response to the Valsalva maneuver in patients on chronic hemodialysis. Nephron 9:251, 1972. 183. Stewart, A. G., Waterhouse, J. C., and Howard, P.: The QTc interval, autonomic neuropathy and mortality in hypoxaemic COPD. Respir. Med. 89:79, 1995. 184. Stoebner, P., Mezin, P., Vila, A., et al.: Microangiopathy of endoneurial vessels in hypoxemic chronic obstructive pulmonary disease (COPD): a quantitative ultrastructural study. Acta Neuropathol. 78:388, 1989. 185. Sung, J. H., Park, S. H., Mastri, A. R., and Warwick, W. J.: Axonal dystrophy in the gracile nucleus in congenital biliary atresia and cystic fibrosis (mucoviscidosis): beneficial effect of vitamin E therapy. J. Neuropathol. Exp. Neurol. 39:584, 1980. 186. Suzuki, H., Takei, T., Tsuji, H., and Nishikawa, T.: Membranoproliferative glomerulonephritis and demyelinating neuropathy caused by type II mixed cryoglobulinemia associated with HCV infection. Intern. Med. 39:397, 2000. 187. Tambini, R., Quattrini, A., Fracassetti, O., and Nemni, R.: Axonal neuropathy in a patient receiving interferon-alpha therapy for chronic hepatitis C. J. Rheumatol. 24:1656, 1997. 188. Tegner, R., and Brismar, T.: Experimental uremic neuropathy. Part 1. Decreased nerve conduction velocity in rats. J. Neurol. Sci. 65:29, 1984. 189. Tegner, R., and Lindholm, B.: Uremic polyneuropathy: different effects of hemodialysis and continuous ambulatory peritoneal dialysis. Acta Med. Scand. 218:409, 1985. 190. Tegner, R., and Lindholm, B.: Vibratory perception threshold compared with nerve conduction velocity in the evaluation of uremic neuropathy. Acta Neurol. Scand. 71:284, 1985. 191. Tembl, J. I., Ferrer, J. M., Sevilla, M. T., et al.: Neurologic complications associated with hepatitis C virus infection. Neurology 53:861, 1999. 192. Thiele, B., and Stalberg, E.: Single fibre EMG findings in polyneuropathies of different aetiology. J. Neurol. Neurosurg. Psychiatry 38:881, 1975. 193. Thomas, P. K., and Walker, J. G.: Xanthomatous neuropathy in primary biliary cirrhosis. Brain 88:1079, 1965.

194. Thuluvath, P. J., and Triger, D. R.: Autonomic neuropathy and chronic liver disease. Q. J. Med. 72:737, 1989. 195. Traber, M. G., Sokol, R. J., Ringel, S. P., et al.: Lack of tocopherol in peripheral nerves of vitamin E-deficient patients with peripheral neuropathy. N. Engl. J. Med. 317:262, 1987. 196. Trevisani, F., Sica, G., and Bernardi, M.: Autonomic neuropathy in advanced liver disease. Hepatology 24:1549, 1996. 197. Tsukada, N., Koh, C. S., Inoue, A., and Yanagisawa, N.: Demyelinating neuropathy associated with hepatitis B virus infection: detection of immune complexes composed of hepatitis B virus surface antigen. J. Neurol. Sci. 77:203, 1987. 198. Tsukada, N., Koh, C. S., Owa, M., and Yanagisawa, N.: Chronic neuropathy associated with immune complexes of hepatitis B virus. J. Neurol. Sci. 61:193, 1983. 199. Tuck, R. R., Schmelzer, J. D., and Low, P. A.: Endoneurial blood flow and oxygen tension in the sciatic nerves of rats with experimental diabetic neuropathy. Brain 107:935, 1984. 200. Valli, G., Barbieri, S., Sergi, P., et al.: Evidence of motor neuron involvement in chronic respiratory sufficiency. J. Neurol. Neurosurg. Psychiatry 47:1117, 1984. 201. Van Der Vliet, J. A., Navarro, X., Kennedy, W. R., et al.: Long-term follow-up of polyneuropathy in diabetic kidney transplant recipients. Diabetes 37:1247, 1988. 202. Vaziri, D., Pratt, H., Saiki, J. K., and Starr, A.: Evaluation of somatosensory pathway by short latency evoked potentials in patients with end-stage renal disease maintained on hemodialysis. Int. J. Artif. Organs 4:17, 1981. 203. Verge, V. M. K., Xu, Z., Xu, X. J., et al.: Marked increase in nitric oxide synthase mRNA in rat dorsal root ganglia after peripheral axotomy: in situ hybridization and functional studies. Proc. Natl. Acad. Sci. U. S. A. 89:11617, 1992. 204. Vita, G., Savica, V., Milone, S., et al.: Uremic autonomic neuropathy: recovery following bicarbonate hemodialysis. Clin. Nephrol. 45:56, 1996. 205. Wijdicks, E. F. M., Litchy, W. J., Harrison, B. A., and Gracey, D. R.: The clinical spectrum of critical illness polyneuropathy. Mayo Clin. Proc. 69:955, 1994. 206. Williams, A. C., Sturman, S., Kelsey, S., et al.: The neuropathy of the critically ill. Br. Med. J. (Clin. Res. Ed.) 293:790, 1986. 207. Witt, N. J., Zochodne, D. W., Bolton, C. F., et al.: Peripheral nerve function in sepsis and multiple organ failure. Chest 99:176, 1991. 208. Wright, E. A., and McQuillen, M. P.: Hypoexcitability of ulnar nerve in patients with normal motor nerve conduction velocities. Neurology 23:78, 1973. 209. Wytrzes, L., Markley, H. G., Fisher, M., and Alfred, H. J.: Brachial neuropathy after brachial artery-antecubital vein shunts for chronic hemodialysis. Neurology 37:1398, 1987. 210. Yashpal, and Issar, U.: Autonomic failure of circulation in chronic uremia. Indian J. Med. Sci. 36:69, 1982. 211. Yosipovitch, G., Yarnitsky, D., Mermelstein, V., et al.: Paradoxical heat sensation in uremic polyneuropathy. Muscle Nerve 18:768, 1995.

Renal Failure, Hepatic Disorders, Chronic Respiratory Disease, and Critical Illness 212. Zaltron, S., Puoti, M., Liberini, P., et al.: High prevalence of peripheral neuropathy in hepatitis C infected patients with symptomatic and asymptomatic cryoglobulinemia. Ital. J. Gastroenterol. Hepatol. 30:391, 1998. 213. Zifko, U. A., Zipko, H. T., and Bolton, C. F.: Clinical and electrophysiological findings in critical illness polyneuropathy. J. Neurol. Sci. 159:186, 1998. 214. Zingraff, J., Beyne, P., Urena, P., et al.: Influence of haemodialysis membranes on beta 2-microglobulin kinetics: in vivo and in vitro studies. Nephrol. Dial. Transplant. 3:284, 1988. 215. Zoccali, C., Ciccarelli, M., and Maggiore, Q.: Defective reflex control of heart rate in dialysis patients: evidence for an afferent autonomic lesion. Clin. Sci. 63:285, 1982. 216. Zochodne, D. W., Bolton, C. F., Wells, G. A., et al.: Critical illness polyneuropathy, a complication of sepsis and multiple organ failure. Brain 110:819, 1987. 217. Zochodne, D. W., and Cheng, C.: Diabetic peripheral nerves are susceptible to multifocal ischemic damage from endothelin. Brain Res. 838:11, 1999.

2037

218. Zochodne, D. W., Cheng, C., and Sun, H.: Diabetes increases sciatic nerve susceptibility to endothelin-induced ischemia. Diabetes 45:627, 1996. 219. Zochodne, D. W., and Ho, L. T.: Normal blood flow but lower oxygen tension in diabetes of young rats: microenvironment and the influence of sympathectomy. Can. J. Physiol. Pharmacol. 70:651, 1992. 220. Zochodne, D. W., Ho, L. T., and Gross, P. M.: Acute endoneurial ischemia induced by epineurial endothelin in the rat sciatic nerve. Am. J. Physiol. 263:H1806, 1992. 221. Zochodne, D. W., Levy, D., Zwiers, H., et al.: Evidence for nitric oxide and nitric oxide synthase activity in proximal stumps of transected peripheral nerves. Neuroscience 91:1515, 1999. 222. Zochodne, D. W., Misra, M., Cheng, C., and Sun, H.: Inhibition of nitric oxide synthase enhances peripheral nerve regeneration in mice. Neurosci. Lett. 228:71, 1997. 223. Zochodne, D. W., Ramsay, D. A., Saly, V., et al.: Acute necrotizing myopathy of intensive care: electrophysiological studies. Muscle Nerve 17:285, 1994.

88 Neuropathy in Diseases of the Thyroid and Pituitary Glands JOHN D. POLLARD

Hypothyroid Neuropathy Mononeuropathy Polyneuropathy

Hyperthyroid Neuropathy Acromegalic Neuropathy

This chapter reviews the subjects of neuropathies relating to hypothyroidism, hyperthyroidism, and acromegaly.

HYPOTHYROID NEUROPATHY Neurologic complications of hypothyroidism are well recognized. Slowness of mental and locomotor activities was noted in the earliest description of the disease43; ataxia, epilepsy, headache, coma, cranial nerve dysfunction, and prolonged relaxation of tendon reflexes have also been well described.67,86 Although symptoms suggestive of peripheral neuropathy have been reported to occur commonly,22 adequately documented cases of a diffuse neuropathy with objective signs are relatively few.68 Treatment for hypothyroidism, of course, developed early, so that advanced cases of the disease are very infrequently seen, but even in areas of the world where myxedematous cretinism is endemic, peripheral neuropathy is uncommon.9 Involvement of single nerves (mononeuropathy), particularly the median nerve within the carpal tunnel, occurs commonly in hypothyroidism.65,80 In the reports of neuropathy in myxedema, it was believed that nerve damage resulted from deposition of a mucinous material around the nerves and their endings66,73 and from a high incidence of Renaut bodies.55 Although mucopolysaccharide protein complexes (mucoid) may play a role in carpal tunnel compression,65 it is more likely that, in cases of generalized neuropathy, dysfunction results from metabolic derangement associated with lack of thyroid hormone.

Mononeuropathy Polyneuropathy

Mononeuropathy Carpal Tunnel Compression There is a high incidence of carpal tunnel compression of the median nerve in myxedema. In their original description of a series of such cases, Murray and Simpson65 noted that 26 of 35 consecutive myxedematous patients complained of acroparesthesias. In 20 patients both hands were involved. In a recent prospective series of patients with newly diagnosed hypothyroidism, 29% had carpal tunnel syndrome based upon clinical and electrophysiologic studies.29 Murray and Simpson65 concluded on the basis of clinical examination and electrophysiologic studies in their patients that the symptoms resulted from compression of the median nerve within the carpal tunnel. These findings were confirmed by Purnell and co-workers80; in both series no correlation was noted between presence of neuropathy and severity of myxedema, but in the majority of cases there was a loss of symptoms with return of the euthyroid state. Clinical Features. Clinical symptoms and signs are similar to those arising from carpal tunnel compression caused by other disorders.37 Murray and Simpson65 noted in their cases numbness or pins and needles that tended to occur either during activities such as sewing, knitting, or washing, or at night, when the tingling would awaken the patient from sleep or prevent sleep. Symptoms were usually present on awakening in the morning but passed off soon after rising. Both hands were involved in the majority of patients. Approximately one third of their 2039

2040

Diseases of the Peripheral Nervous System

patients had objective sensory impairment in the median nerve distribution and one third showed atrophy of the abductor pollicis brevis and opponens pollicis. Diagnosis and Differential Diagnosis. Carpal tunnel syndrome may result from other causes of local compression, including rheumatoid arthritis, osteoarthritis, and acromegaly; from pregnancy; or from underlying neuropathy such as hereditary liability to pressure palsies32 or diabetes.37 If the patient is shown to be hypothyroid, disease of the pituitary must be excluded. Sensory symptoms in the fingers and arm pain may result from compression of C6 or C7 cervical nerve roots, for example, as a result of cervical spondylosis. Cervical nerve root lesions usually produce weakness of arm or forearm muscles and reflex changes. Wasting of thenar muscles may result from brachial plexus compression by a cervical rib or band or by an apical lung tumor (Pancoast’s syndrome).37 These lesions usually cause sensory loss in the C8-T1 dermatomes and some wasting in the hypothenar muscles. The diagnosis should be confirmed by nerve conduction studies and electromyography. Nerve conduction studies confirm slowing of median nerve conduction across the region of the flexor retinaculum. This may be demonstrated by prolongation of distal motor latency and/or prolongation of sensory latencies when stimulation is applied to the index finger or palm. Prolongation may be measured by comparison to control values or to values for ipsilateral ulnar nerve or radial nerve or contralateral unaffected median nerve. In addition, there may be reduction in amplitudes of median nerve compound muscle action potentials or sensory action potentials.14,37,39,51,59,97–99 Electromyography should exclude cervical nerve root or brachial plexus lesions. Sampling of other arm muscles may demonstrate chronic partial denervation in C6- or C7-supplied muscles in cases of cervical spondylosis, and in thoracic outlet lesions such changes would be expected in hypothenar muscles together with impaired sensory conduction within the ulnar nerve.38,40 Pathology and Treatment. It is likely that the carpal tunnel syndrome in hypothyroidism results from local physical factors—reduced space within the flexor retinaculum.65 Although there is considerable doubt concerning the deposition of acid mucopolysaccharide protein complexes in nerve in myxedema, there is evidence for their increased presence in tendons and related synovial sheaths41; joint effusion and synovial thickening, which clear with thyroid replacement, have been described in hypothyroid patients.7 Hence the original hypothesis of Murray and Simpson65 that the carpal tunnel syndrome in myxedema results from myxedematous tissue beneath the flexor retinaculum is probably correct. This view is reinforced by the finding

that surgery is only rarely necessary in these patients; the majority lose their symptoms within weeks to months of successful replacement therapy.65,80 Schwartz and Ingold91 studied nine patients with myxedema and carpal tunnel syndrome and found electrophysiologic evidence of tarsal tunnel compression in four, three of whom were mildly symptomatic. Cranial Nerves Deafness in Myxedema. The occurrence of deafness in hypothyroidism was noted as early as 1888, when a committee of the Clinical Society of London20 recorded impaired hearing in 32 of 60 patients. In subsequent reports the incidence has varied between 30% and 85%.4,6,27,67,101 Deafness results from a sensorineural deficit,27,45,78 but the precise cause is unknown.101 Some experimental studies have shown changes in the spiral ganglion,27 and others in the organ of Corti.52,84 In a study of 48 myxedematous patients, Van’t Hoff and Stuart101 found that improvement of hearing occurred in 73% of deaf ears following return to the euthyroid state and normal hearing occurred in 23%. Other Cranial Nerves. Although dysarthria and hoarseness occur in many patients, there is no strong evidence to incriminate involvement of the IXth, Xth, or XIIth nerve as has been suggested.42 The evidence rather suggests that the cause of these features is local myxedematous infiltration of the tongue and vocal cords.83,86

Polyneuropathy The existence of an intrinsic diffuse sensorimotor neuropathy secondary to hypothyroidism has been accepted only relatively recently.29,34,67 In hypothyroid patients distal paresthesias occur quite commonly. Crevasse and Logue22 found that 31 of 65 patients had severe burning or lancinating extremity pains and/or troublesome paresthesias that resolved with adequate therapy. In their series, however, objective signs were difficult to find. Nevertheless, more recently objective distal sensory disturbances have been reported to occur in 17% to 60% of large series of hypothyroid patients.23,29,67,102 In a prospective group of newly diagnosed patients studied by Duyff and colleagues, 42% had signs of a sensorimotor neuropathy.29 In a series of 520 cases of generalized peripheral neuropathy reported from Taiwan, 2% of cases were due to hypothyroidism.57 Clinical Features In the majority of patients, sensory symptoms dominate the clinical picture. In the series described by Nickel and Frame,67 all patients complained of paresthesias and objective sensory loss was found in 60%. Signs of neuropathy

Neuropathy in Diseases of the Thyroid and Pituitary Glands

2041

develop early in the course of the disease.29 Sensory complaints include painful dysesthesias in the hands and feet and lancinating pains sometimes suggesting nerve root compression.22,46,64,66,68,77 On examination there may be distal glove-and-stocking sensory loss, and gait ataxia may be seen.21,64,68,79 Although weakness is a common complaint,68 objective evidence of this is less commonly found. Proximal limb weakness may result from hypothyroid myopathy.3 However, distal weakness and wasting has been reported in several series,31,64,77 and some patients complain of muscle cramping.3,47,66,68 Reflexes may be diminished, and particularly the ankle jerk may show the delayed relaxation phase that is characteristic of hypothyroidism.16,53 This reflex change is believed to be the result of disturbances in energy transfer within muscle rather than slowing within neural pathways.44,53 Transient swelling percussion of the skin (mounding phenomenon) may be observed.54,68 The cerebrospinal fluid protein content may be normal but is usually increased, and the gamma globulin fraction may also be elevated.8,13 Nerve Conduction Studies Fincham and Cape,34 in a study of 16 patients, showed reduction in the amplitude of ulnar and median sensory action potentials but very little change in motor conduction velocity. In the cases studied by Dyck and Lambert,31 motor conduction velocities were slowed mildly in one case and moderately in another; in this latter case, considerable prolongation of distal motor latency was found with considerable dispersion of the compound muscle action potential (Fig. 88–1). It may be noted from Figure 88–1 that treatment with thyroid hormone was followed by marked improvement in the latency, dispersion, and amplitude of the compound muscle action potential. In studies of compound nerve action potentials from biopsy specimens of sural nerves in vitro (Fig. 88–2), Dyck and Lambert31 found that the potentials of myelinated fibers were reduced in size and conduction velocity, whereas the C-fiber potential was reduced in amplitude but not velocity. These findings are in keeping with underlying demyelination. In two other cases Pollard and co-workers77 found similar changes of prolonged distal motor latencies and slowed motor conduction velocities suggestive of demyelination. In other reports64,66,92 motor conduction velocity was only mildly slowed, consistent with the finding of axonal degeneration. Abbott and colleagues1 suggested that some of the slowing of nerve conduction velocity reported in hypothyroidism may result from decreased body temperature, because in several studies in which reduced nerve conduction velocity was found no reference was made to limb temperature.34,82,89 However, in the study of Dyck and Lambert,31 slowing was recorded even in biopsy specimens of sural nerve fascicles that were studied in vitro with temperature controlled at 37° C.

FIGURE 88–1 Action potentials (case 1) detected by belly and tendon electrodes over hypothenar muscles after maximal electric stimulus to ulnar nerve at elbow (upper pair) and at wrist (lower pair). Top record in each pair was obtained prior to treatment. Bottom record of each pair was obtained after 9 months of treatment with thyroid hormone. Note longer latency and greater dispersion of response prior to treatment and nearly normal response after treatment. (From Dyck, P. J., and Lambert, E. H.: Polyneuropathy associated with hypothyroidism. J. Neuropathol. Exp. Neurol. 29:631, 1970, with permission.)

Morphology Light Microscopy. Ord,73 in the earliest observation of myxedema affecting peripheral nerve, concluded that mucinous material enveloped nerves and their terminations. Nickel and associates68 observed a mucopolysaccharideprotein complex within the endoneurium and perineurium, which they suggested may interfere with nerve function. However, more recent studies using modern histologic techniques have not observed such deposits in nerve.31,64,77,92 All studies agree that a reduction in the number of myelinated fibers, particularly those of large diameter, occurs in hypothyroid neuropathy.31,64,66,77,92 Langhans55 found a high incidence of Renaut bodies, a finding also noted by other workers.63,66 In teased fiber preparations from their patients, Dyck and Lambert31 found evidence of demyelination and remyelination, whereas in another

2042

Diseases of the Peripheral Nervous System

FIGURE 88–2 Compound action potential of fascicles of sural nerve from a healthy 21-year-old man (A), and from a 55-year-old man (case 1) (B) and a 71-year-old woman (case 2) (C) with hypothyroidism. In B and C, myelinated fiber potentials, A␣ and B␦, are reduced in amplitude and conduction velocity; amplitude of C-fiber potential is also reduced but conduction velocity is unaffected. Action potentials were recorded 20 mm from stimulating cathode. (From Dyck, P. J., and Lambert, E. H.: Polyneuropathy associated with hypothyroidism. J. Neuropathol. Exp. Neurol. 29:631, 1970, with permission.)

study77 the predominant abnormality was active axonal degeneration, with linear arrays of myelin ovoids found in about 25% of fibers teased. Electron Microscopy. Evidence for both demyelination31,92 and axonal degeneration63,66,77 has been reported. In the cases described by Dyck and Lambert,31 occasional axons were found to be either demyelinated or in the process of undergoing segmental demyelination. In addition, very occasional axonal degeneration was seen. The distinctive feature of the nerves in their study was the presence of excessive glycogen, which occurred in aggregates or scattered diffusely, in Schwann cell cytoplasm, in capillaries, or in the cytoplasm of perineurial cells (Fig. 88–3). Diffusely occurring glycogen was also seen within axons of myelinated or unmyelinated fibers (Fig. 88–4). Within Schwann cell cytoplasm glycogen aggregates were often associated with mitochondrial aggregates, Schwann cell lamellar bodies, or lipid droplets (see Fig. 88–3). The findings of excessive glycogen, lamellar bodies, and abnormal mitochondria have been confirmed in our laboratory77 and in other studies.63,66 In addition, these studies found evidence of axonal degeneration and atrophy. Meier and Bischoff 64 noted that most remaining large-diameter myelinated fibers appeared to have shrunken axons and therefore inappropriately thick myelin. In other fibers, disintegration of neurotubules and neurofilaments was seen, with accumulations of dense bodies, vacuoles, and mitochondria.63,66,77 Cluster formation suggesting degeneration and regeneration has also been reported.63,77

Pathology and Treatment In hypothyroid polyneuropathy there is evidence suggesting dysfunction of both the Schwann cell (leading to demyelination and slowed nerve conduction) and axon or neuron (evidenced by axonal degeneration). It is not clear how the metabolic changes consequent upon lack of thyroid hormone produce these changes. Thyroid hormone is known to be critical for the normal development of the central nervous system.26 Biochemical data on nerve cell differentiation indicate that thyroid hormones, among other functions, regulate microtubule assembly.70 Whether the mature nerve cell is as dependent on thyroid hormone as developing cells is not known; however, there is evidence that slow axonal transport is reduced in the sciatic nerves of hypothyroid rats,93 a finding consistent with reduction of microtubule assembly. Most, if not all, of the effects of thyroid hormones result from the interaction of the hormones with specific nuclear receptors.72 Such receptors have been identified on neurons, astrocytes, and oligodendrocytes,87 but whether they are present on Schwann cells is unknown. Nevertheless, it is clear that, with replacement therapy in patients with hypothyroid neuropathy, considerable improvement may occur,31,64,77,92 so that some degree of plasticity is associated with whatever basic metabolic changes underlie hypothyroid neuropathy. In cases in which marked axonal degeneration has occurred, this improvement will be delayed and incomplete.64 Hence early diagnosis and treatment are essential to avoid longterm disability.

Neuropathy in Diseases of the Thyroid and Pituitary Glands

2043

FIGURE 88–3 A, Aggregate of glycogen granules, mitochondria, and cisternae in Schwann cell cytoplasm at node of Ranvier (⫻31,590). B, Aggregate of glycogen granules and lipid droplets in Schwann cell cytoplasm at Schmidt-Lanterman incisure (⫻39,480). C, Glycogen granules, mitochondria, lipid droplets, and cisternae at polar end of Schwann cell nucleus (⫻25,290). D, Glycogen granules, cisternae, and large lipid-like body in Schwann cell cytoplasm (⫻49,410). (From Dyck, P. J., and Lambert, E. H.: Polyneuropathy associated with hypothyroidism. J. Neuropathol. Exp. Neurol. 29:631, 1970, with permission.)

HYPERTHYROID NEUROPATHY Myopathy caused by thyrotoxicosis is well recognized,5,81 but there is some uncertainty regarding polyneuropathy as a complication. A paraplegia-like weakness occurring in hyperthyroidism was first described by Charcot17,18 and was mentioned by several early European authors.88 More

recently, Ludin and co-workers61 reported eight patients in whom electrophysiologic evidence of denervation was found and who were considered to have a subclinical neuropathy. Chollet and colleagues19 reported two cases of thyrotoxicosis and polyneuropathy in which slowed motor nerve conduction was found. Feibel and Campa33 reported a further case in whom a severe polyneuropathy affected

2044

Diseases of the Peripheral Nervous System

FIGURE 88–4 A, Many glycogen granules are seen in axis cylinder of myelinated fiber with well-preserved myelin. Smaller collections of glycogen granules are in cytoplasm. (⫻37,930.) B, Glycogen granules in axis cylinder of myelinated fiber, especially near mitochondria and about vesicles. Several mitochondria appear to have empty spaces. In center of axis cylinder is large structure with double outer membranes and multiple paired membranes inside. (⫻25,2000.) Inset, Enlargement of area enclosed in square (⫻40,320). (From Dyck, P. J., and Lambert, E. H.: Polyneuropathy associated with hypothyroidism. J. Neuropathol. Exp. Neurol. 29:631, 1970, with permission.)

mostly the legs and nerve conduction studies showed marked slowing of motor conduction velocity. In several of these cases, however, the neuropathy cannot be clearly distinguished from acute idiopathic polyneuritis.12,33 In a recent prospective study of patients with newly diagnosed

hyperthyroidism, 19% of patients had predominantly sensory signs of a sensorimotor axonal neuropathy early in the course of the disease. These signs resolved rapidly and completely following treatment.29 No cases have been reported with confirmatory nerve pathology.

Neuropathy in Diseases of the Thyroid and Pituitary Glands

ACROMEGALIC NEUROPATHY Mononeuropathy Compression of the median nerve within the carpal tunnel has frequently been reported in acromegaly.10,15,49,50,58,71,90,94,103 Of 100 acromegalic patients reviewed by O’Duffy and colleagues,71 35 had carpal tunnel syndrome. All but one of these 35 patients had active acromegaly, and in all cases the neuropathy was bilateral. The symptoms and signs are similar to those already described for patients with hypothyroidism, and the diagnosis should be confirmed by electrophysiologic studies. The cause of median nerve compression in acromegaly may involve several factors. Synovial edema may be an important cause; hyperplasia of ligaments and tendons and proliferation of cartilage and bone may also contribute.49,74,96 However, in a recent study using magnetic resonance imaging, nine patients with acromegaly were studied, of whom four had clinical symptoms of neuropathy. Symptomatic patients had increased nerve size and signal intensity compared to asymptomatic patients, but the two groups did not differ in volume of carpal tunnel content.48 Hence the predominant pathology of median nerve neuropathy in acromegaly appears to be increased edema or hypertrophic changes within, nerve rather than extrinsic compression from increased volume of carpal tunnel contents. Relief of symptoms follows correction of the endocrine disorder in most cases.71,90,94

Polyneuropathy A diffuse neuropathy in acromegaly was recognized by Marie and Marinesco in 189162 and later by Arnold.2 Few cases of polyneuropathy well documented by electrophysiologic and histologic studies have been published, although pain and paresthesias in the hands and feet have been noted more frequently.25 The clinical features in addition to the features of acromegaly are those of a sensorimotor neuropathy, and peripheral nerves may be palpable.56,60,95 Reflexes are diminished and weakness may be severe.95 In none of the 11 cases studied by Low and colleagues60 was marked slowing of motor conduction velocity found; rather, there was a mild reduction of motor conduction velocity in 8 of the 11 acromegalic subjects studied. There was also a significant reduction in the mean amplitude of sensory action potentials and of mixed nerve action potentials compared to a control population. Electrophysiologic evidence of carpal tunnel compression was found in 9 of the 11 patients. On histologic examination of the nerves, there was a reduction in the number of myelinated fibers. An increase in endoneurial tissue and in one nerve a pronounced increase in subperineurial tissue were observed. The fascicular area in nerves of patients with acromegaly was increased compared

2045

with control subjects. Hypertrophic changes were recorded in the original description of the neuropathy of acromegaly; Marie and Marinesco62 described perineurial hypertrophy and increased subperineurial loose connective tissue. In a remarkable case described by Stewart,95 a number of peripheral nerves were enlarged and easily palpable. There was an increase in endoneurial and perineurial connective tissue, which was considered to cause compression of the nerve fibers. In another case described by Dinn,28 there was moderate proliferation of Schwann cells with increased perineurial and endoneurial connective tissue. Similar changes were also recorded by Sandbank and colleagues.85 Diffuse segmental demyelination was found by Dinn28 in teased fiber studies, and this was the predominant change found by Low and co-workers.60 Because marked slowing of nerve conduction was not found, however, it is uncertain whether the demyelination was primary or secondary.30,63 Nevertheless, small onion bulb formations were found (Fig. 88–5). The cause of the neuropathy in acromegaly is unclear. It has been suggested that the high incidence of diabetes mellitus that occurs in acromegaly may be responsible28; however, among the cases reported by Low and co-workers,60 two patients with neuropathy had normal glucose tolerance and other patients had only mild degrees of glucose intolerance. It may be concluded that, although diabetes may be a contributing factor to neuropathy in some cases, neuropathy in acromegaly may occur independently. The hypertrophic changes in nerve that have been reported in acromegaly by several workers28,60,95 are of considerable interest and may contribute to the neuropathy. Experimental studies have shown that an extract of bovine pituitary gland (glial growth factor) is a potent mitogen for Schwann cells and fibroblasts.11 Mitogenic responses are well-recognized effects of somatomedins or insulin-like growth factors (IGFs).24,35,36,75,76 Stimulation of cell division is mediated by type I IGF receptors,100,104 and levels of IGF (somatomedin C) are elevated in acromegaly because of induction by growth hormone of IGF-I messenger RNA69 and IGF-I production by liver cells. Proliferation of cartilage, bone, and muscle cells is well known in acromegaly because of the clinical features of disease, but because many diverse cells possess IGF-I receptors, it is likely that the proliferation of Schwann cells and fibroblasts that has been documented in acromegaly is another manifestation of mitogenesis mediated by growth hormone and IGF-I receptors. Glial growth factors are almost certainly related peptides. Hence the cause of neuropathy in acromegalic patients may be multifactorial. Hypertrophic changes within the endoneurium and perineurium may render nerves more susceptible to pressure and trauma and may even produce axonal damage by these changes within the nerves28; cellular proliferation and water and electrolyte retention in other tissues may render pressure sites such as the carpal tunnel more constrictive. O’Duffy and co-workers71 noted

2046

Diseases of the Peripheral Nervous System

FIGURE 88–5 A small onion bulb formation from a patient with acromegaly. Bar: 1 ␮m. (From Low, P. A., McLeod, J. G., Turtle, J. R., et al.: Peripheral neuropathy in acromegaly. Brain 97:139, 1974, permission.)

rapid relief of acroparesthesias in acromegalic patients in whom pituitary overactivity was controlled, but there are insufficient data on the response to therapy of patients with generalized neuropathy.

REFERENCES 1. Abbott, R. J., O’Malley, B. P., Barnett, D. B., et al.: Central and peripheral nerve conduction in thyroid dysfunction: the influence of L-thyroxine therapy compared to warming upon conduction abnormalities of primary hypothyroidism. Clin. Sci. 64:617, 1983. 2. Arnold, J.: Weitere Beiträge zur Akromegaliefrage. Arch. Pathol. Anat. Physiol. Klin. Med. 135:1, 1894. 3. Åström, K.-E., Kugelberg, E., and Muller, R.: Hypothyroid myopathy. Arch. Neurol. 5:472, 1961. 4. Barlow, R. A.: The study of vestibular nerve function in myxedema. Am. J. Med. Sci. 164:401, 1922. 5. Bathurst, W. L.: A case of Graves’ disease associated with idiopathic muscular atrophy. Lancet 2:529, 1895.

6. Bhatia, P. L., Gupta, O. P., Agrawal, M. K., and Mishr, S. K.: Audiological and vestibular function tests in hypothyroidism. Laryngoscope 57:2082, 1977. 7. Bland, J. H., and Frymoyer, J. W.: Rheumatic syndromes of myxedema. N. Engl. J. Med. 282:1171, 1970. 8. Bloomer, H. A., Papadopoulos, N. M., and McLane, J. E.: Cerebrospinal fluid gamma globulin concentration in myxedema. J. Clin. Endocrinol. Metab. 20:869, 1960. 9. Boyages, S. C., Halpern, J.-P., Maberly, G. F., et al.: A comparative study of neurological and myxoedematous endemic cretinism in western China. J. Clin. Endocrinol. Metab. 67:1262, 1988. 10. Brain, W. R., Wright, A. D., and Wilkinson, M.: Spontaneous compression of both median nerves in the carpal tunnel: six cases treated surgically. Lancet 1:277, 1947. 11. Brockes, J. P., Fryxell, K. J., and Lemke, G. E.: Studies on cultured Schwann cells: the induction of myelin synthesis, and the control of their proliferation by a new growth factor. J. Exp. Biol. 95:215, 1981. 12. Bronsky, D., Kaganiec, G. I., and Waldstein, S. S.: An association between the Guillain-Barré syndrome and hyperthyroidism. Am. J. Med. Sci. 247:196, 1964.

Neuropathy in Diseases of the Thyroid and Pituitary Glands 13. Bronsky, D., Shrifter, H., de la Huerga, J., et al.: Cerebrospinal fluid proteins in myxedema, with special reference to electrophoretic partition. J. Clin. Endocrinol. Metab. 18:470, 1958. 14. Buchthal, F., Rosenfalck, A., and Trojaborg, W.: Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J. Neurol. Neurosurg. Psychiatry 37:340, 1974. 15. Canon, B. W., and Love, J. G.: Tardy median palsy; median neuritis; median thenar neuritis amenable to surgery. Surgery 20:210, 1946. 16. Chaney, W. C.: Tendon reflexes in myxedema: a valuable aid in diagnosis. J. Am. Med. Assoc. 82:2013, 1924. 17. Charcot, J.: Leçons du mardi. Polyclinique 239, 1888. 18. Charcot, J.: Nouveaux signes de la maladie de Basedow. Bull. Med. 3:147, 1889. 19. Chollet, P. H., Rigal, J.-P., and Pignide, L.: Une complication méconnue de l’hyperthyroidie: la neuropathie périphérique [letter to the editor]. Presse Med. 79:145, 1971. 20. Committee of the Clinical Society of London: Report on the subject of myxoedema. Trans. Clin. Soc. Lond. Suppl. 21:1, 1888. 21. Cremer, G. M., Goldstein, N. P., and Paris, J.: Myxedema and ataxia. Neurology (Minneap.) 19:37, 1969. 22. Crevasse, L. E., and Logue, R. B.: Peripheral neuropathy in myxedema. Ann. Intern. Med. 50:1433, 1959. 23. Currier, F. P., and Brink, J. R.: Multiple neuritis and hypothyroidism. Dis. Nerv. Syst. 9:144, 1948. 24. Daughaday, W. H.: A personal history of the origin of the somatomedin hypothesis and recent challenges to its validity. Perspect. Biol. Med. 32:194, 1989. 25. Davidoff, L. M.: Studies in acromegaly. III. The anamnesis and symptomatology in one hundred cases. Endocrinology 10:461,1926. 26. De Nayer, P., and Dozin, B.: Thyroid hormone receptors in the developing brain. In Delong, D. R., Robbins, J., and Condliffe, P. G. (eds.): Iodine and the Brain. New York, Plenum Press, p. 51, 1989. 27. Des Vos, J. A.: Deafness in hypothyroidism. J. Laryngol. Otol. 77:390, 1963. 28. Dinn, J. J.: Schwann cell dysfunction in acromegaly. J. Clin. Endocrinol. Metab. 31:140, 1970. 29. Duyff, R. F., Van den Bosch, J., Laman, D. M., et al.: Neuromuscular findings in thyroid dysfunction: a prospective clinical and electrodiagnostic study. J. Neurol. Neurosurg. Psychiatry 68:750, 2000. 30. Dyck, P. J., Johnson, W. J., Lambert, E. H., and O’Brien, P. C.: Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin. Proc. 46:400, 1971. 31. Dyck, P. J., and Lambert, E. H.: Polyneuropathy associated with hypothyroidism. J. Neuropathol. Exp. Neurol. 29:631, 1970. 32. Earl, C. J., Fullerton, P. M., Wakefield, G. S., and Schutta, H. S.: Hereditary neuropathy with liability to pressure palsies. Q. J. Med. 33:481, 1964. 33. Feibel, J. H., and Campa, J. F.: Thyrotoxic neuropathy (Basedow’s paraplegia). J. Neurol. Neurosurg. Psychiatry 39:491, 1976. 34. Fincham, R. W., and Cape, C. A.: Neuropathy in myxedema: a study of sensory nerve conduction in the upper extremities. Arch. Neurol. 19:464, 1968.

2047

35. Froesch, E. R., Burgi, H., Lamsier, E. G., et al.: Antibodysuppressible and non-suppressible insulin-like activities in human serum and their physiological significance: an insulin assay with adipose tissue of increased precision and specificity. J. Clin. Invest. 42:1816, 1963. 36. Froesch, E. R., Schmid C, Schwander J, Zapf J: Actions of insulin-like growth factors. Annu. Rev. Physiol. 47:443, 1985. 37. Gilliatt, R. W., and Harrison, M. J. G.: Nerve compression and entrapment. In Asbury, A. K., and Gilliatt, R. W. (eds.): Peripheral Nerve Disorders: A Practical Approach. London, Butterworths, p. 243, 1984. 38. Gilliatt, R. W., Le Quesne, P. M., Logue, V., and Sumner, A. J.: Wasting of the hand associated with a cervical rib or band. J. Neurol. Neurosurg. Psychiatry 33:615, 1970. 39. Gilliatt, R. W., and Sears, T. A.: Sensory nerve action potentials in patients with peripheral nerve lesions. J. Neurol. Neurosurg. Psychiatry 21:109, 1958. 40. Gilliatt, R. W., Willison, R. G., Dietz, V., and Williams, I. R.: Peripheral nerve conduction in patients with a cervical rib and band. Ann. Neurol. 4:124, 1978. 41. Golding, D. N.: Hypothyroidism presenting with musculoskeletal symptoms. Ann. Rheum. Dis 29:10, 1970. 42. Grotting, J. K., and Pemberton, J. de J.: Fixation of vocal cords in acromegaly. Arch. Otolaryngol. 52:608, 1950. 43. Gull, W. W.: On a cretinoid state supervening in adult life in women. Trans. Clin. Soc. Lond. 7:180, 1874. 44. Hofmann, W. W., and Denys, E. H.: Effects of thyroid hormone at the neuromuscular junction. Am. J. Physiol. 223:283, 1972. 45. Howarth, A. E., and Lloyd, H. E. D.: Perceptive deafness in hypothyroidism. Br. Med. J. 1:431, 1956. 46. Hun, H., and Prudden, T. M.: Myxedema: four cases with two autopsies. Am. J. Med. Sci. 96:1, 140, 1888. 47. Ikeda, J. K., and Marvin, S. L.: Peripheral neuropathy in hypothyroidism. Bull. Los Ang. Neurol. Soc. 25:106, 1960. 48. Jenkins, P. J., Sohaib, S. A., Akker, S., et al.: The pathology of median neuropathy in acromegaly. Ann. Intern. Med. 133:197, 2000. 49. Johnson, A. W.: Acroparaesthesiae and acromegaly. Br. Med. J. 1:1616, 1960. 50. Kellgren, J. H., Ball, J., and Tutton, G. K.: The articular and other limb changes in acromegaly: a clinical and pathological study of 25 cases. Q. J. Med. 21:1405, 1952. 51. Kimura, J.: The carpal tunnel syndrome. Brain 102:619, 1979. 52. Kohenen, A., Jauhiainen, T., Liewendahl, K., et al.: Deafness in experimental hypo- and hyperthyroidism. Laryngoscope 81:947, 1971. 53. Lambert, E. H., Underdahl, L. O., Beckett, S., and Mederos, L. O.: A study of the ankle jerk in myxedema. J. Clin. Endocrinol. Metab. 11:1186, 1951. 54. Lanari, A.: Los musculos estriadios en el’hipotiroidismo. Medicina (B. Aires) 7:197, 1947. 55. Langhans, T. L.: Über Veränderungen in den peripherischen Nerven bei Kuchexia thyreopriva des Menschen und Affen, sowie bei Kretinismus. Virchow’s Arch. A 128:369, 1892. 56. Lewis, P. D.: Neuromuscular involvement in pituitary gigantism. Br. Med. J. 2:499, 1972. 57. Lin, K. P., Kwan, S. Y., Chen, S. Y., et al.: Generalized neuropathy in Taiwan: an etiologic survey. Neuroepidemiology 12:257, 1993.

2048

Diseases of the Peripheral Nervous System

58. List, C. F.: Doppelseitige Medianusneuritis bei Akromegalie: Gleichzeitig eine Literaturübersicht der bei der Akromegalie beobachteten Veränderungen des peripheren und spinalen Nervensystems. Dtsch. Z. Nervenheilkd. 124:279, 1932. 59. Loong, S. C., and Seah, C. S.: Comparison of median and ulnar sensory nerve action potentials in the diagnosis of carpal tunnel syndrome. J. Neurol. Neurosurg. Psychiatry 32:750, 1971. 60. Low, P. A., McLeod, J. G., Turtle, J. R., et al.: Peripheral neuropathy in acromegaly. Brain 97:139, 1974. 61. Ludin, H. P., Spiess, H., and Koenig, M. P.: Neuropathie et hyperthyreose. Rev. Neurol. (Paris) 120:424, 1969. 62. Marie, P., and Marinesco, G.: Sur l’anatomie pathologique de l’acromegalie. Arch. Med. Exp. Anat. Pathol. (Paris) 3:5, 1891. 63. McLeod, J. G., Prineas, J. W., and Walsh, J. C. W.: The relationship of conduction velocity to pathology in peripheral nerves. In Desmedt, J. E. (ed.): New Developments in Electromyography and Clinical Neurophysiology, Vol. 2. Basel, Karger, p. 248, 1973. 64. Meier, C., and Bischoff, A.: Polyneuropathy in hypothyroidism: clinical and nerve biopsy study of four cases. J. Neurol. 215:10:1977. 65. Murray, I. P. C., and Simpson, J. A.: Acroparaesthesia in myxoedema: a clinical and electromyographic study. Lancet 1:1360, 1958. 66. Nemni, R., Bottacchi, E., Fazio, R., et al.: Polyneuropathy in hypothyroidism: clinical, electrophysiological and morphological findings in four cases. J. Neurol. Neurosurg. Psychiatry 50:1454, 1987. 67. Nickel, S. N., and Frame, B.: Nervous and muscular systems in myxedema. J. Chronic Dis. 14:570, 1961. 68. Nickel, S. N., Frame, B., Bebin, J., et al.: Myxedema neuropathy and myopathy: a clinical and pathologic study. Neurology (Minneap.) 11:125, 1961. 69. Norstedt, G., and Moller, C.: Growth hormone induction of insulin-like growth factor 1 messenger RNA in primary cultures of rat liver cells. J. Endocrinol. 115:135, 1987. 70. Nunez, J.: Microtubules and brain development: the effects of thyroid hormones. Neurochem. Int. 7:959, 1985. 71. O’Duffy, J. D., Randall, R. V., and MacCarty, C. S.: Median neuropathy (carpal-tunnel syndrome) in acromegaly: a sign of endocrine overactivity. Ann. Intern. Med. 78:379, 1973. 72. Oppenheimer, J. H., Schwarz, H. L., Mariash, C. N., et al.: Advances in our understanding of thyroid hormone action at the cellular level. Endocr. Rev. 8:288, 1987. 73. Ord, W. M.: On myxoedema, a term proposed to be applied to an essential condition in the “cretinoid” affliction occasionally observed in middle-aged women. Med. Chir. Trans. Lond. 61:57, 1878. 74. Phelan, G. S.: The carpal tunnel syndrome. J. Bone Joint Surg. Am. 48:211, 1966. 75. Phillips, L. S., and Vassilopoulou-Sellin, R.: Somatomedins (first of two parts). N. Engl. J. Med. 302:371, 1980. 76. Phillips, L. S., and Vassilopoulou-Sellin, R.: Somatomedins (second of two parts). N. Engl. J. Med. 302:438, 1980. 77. Pollard, J. D., McLeod, J. G., Honnibal, T. G. A., and Verheiden, M. A.: Hypothyroid polyneuropathy. J. Neurol. Sci. 53:461, 1982.

78. Post, J. T.: Hypothyroid deafness: a clinical study of sensorineural deafness associated with hypothyroidism. Laryngoscope 74:221, 1964. 79. Price, T. R., and Netsky, M. G.: Myxedema and ataxia: cerebellar alterations and “neural myxedema bodies.” Neurology [Minneap.] 16:957, 1966. 80. Purnell, D. C., Daly, D. D., and Lipscomb, P. R.: Carpaltunnel syndrome associated with myxedema. Arch. Intern. Med. 108:751, 1961. 81. Ramsay, I. D.: Electromyography in thyrotoxicosis. Q. J. Med. 34:255, 1965. 82. Rao, S. N., Katiyar, B. C., Nair, K. R. P., and Misra, S.: Neuromuscular status in hypothyroidism. Acta Neurol. Scand. 61:167, 1980. 83. Ritter, F. N.: The effects of hypothyroidism upon the ear, nose and throat: a clinical and experimental study. Laryngoscope 77:1427, 1967. 84. Ritter, F. N., and Lawrence, M.: Reversible hearing loss in human hypothyroidism and correlated changes in the chick inner ear. Laryngoscope 70:393, 1960. 85. Sandbank, U., Bornstein, B., and Najenson, T.: Acidophilicadenoma of the pituitary with polyneuropathy. J. Neurol. Neurosurg. Psychiatry 37:324, 1974. 86. Sanders, V.: Neurologic manifestations of myxedema (concluded). N. Engl. J. Med. 266:599, 1962. 87. Sarlieve, L. L., Besnard, F., Labourdette, G., et al.: Investigation of myelinogenesis in vitro: transient expression of 3,5,3⬘-triiodothyronine nuclear receptors in secondary cultures of pure rat oligodendrocytes. In De Long, G. R., Robbins, J., and Cordliffe, P. G. (eds.): Iodine and the Brain. New York, Plenum Press, p. 113, 1989. 88. Sattler, H: Basedow’s Disease. [English translation of 1908 edition by G. W. Marchland and J. F. Marchland.] New York, Grune & Stratton, 1952. 89. Scarpalezos, S., Lygidakis, C., Papageorgiou, C., et al.: Neural and muscular manifestations of hypothyroidism. Arch. Neurol. 29:140, 1973. 90. Schiller, F., and Kolb, F. O.: Carpal tunnel syndrome in acromegaly. Neurology (Minneap.) 4:271, 1954. 91. Schwartz, N. B., and Ingold, A. H.: Effect of thyroxin on stimulus frequency-tension output relationship in the neuromuscular system [abstract]. Fed. Proc. 15:166, 1956. 92. Shirabe, T., Tawara, S., Terao, A., and Araki, S.: Myxoedematous polyneuropathy: a light and electron microscopic study of the peripheral nerve and muscle. J. Neurol. Neurosurg. Psychiatry 38:241, 1975. 93. Sidenius, P., Laurberg, P., Nagel, P., et al.: Reduced axonal transport velocity of slow component a in sciatic nerves of hypothyroid rats. In Proceedings of the Peripheral Neuropathy Association of America, Hilton Head, NC, 1986, p. 46. 94. Skanse, B.: Carpal-tunnel syndrome in myxedema and acromegaly. Acta Chir. Scand. 121:476, 1961. 95. Stewart, B. J.: The hypertrophic neuropathy of acromegaly: a rare neuropathy associated with acromegaly. Arch. Neurol. 14:107, 1966. 96. Sullivan, C. R., Jones, D. R., Bahn, R. C., and Randall, R. V.: Clinics on Endocrine and Metabolic Diseases. 9. Biopsy of the costochondral junction in acromegaly. Mayo Clin. Proc. 38:81, 1963.

Neuropathy in Diseases of the Thyroid and Pituitary Glands 97. Thomas, J. E., and Lambert, E. H.: Das Karpaltunnelsyndrom. Muench. Med. Wochenschr. 104:123, 1962. 98. Thomas, J. E., Lambert, E. H., and Cseuz, K. A.: Electrodiagnostic aspects of the carpal tunnel syndrome. Arch. Neurol. 16:635, 1967. 99. Thomas, PK: Motor nerve conduction in the carpal tunnel syndrome. Neurology (Minneap) 10:1045, 1960. 100. Van Wyk, J. J.: The somatomedins: biological actions and physiological control mechanisms. Hormonal Proteins Peptides 12:81, 1984.

2049

101. Van’t Hoff, W., and Stuart, D. W.: Deafness in myxoedema. Q. J. Med. 48:361, 1979. 102. Watanakisnakorn, C., Hodges, R. E., and Evans, T. G.: Myxedema: a study of 400 cases. Arch. Intern. Med. 116:183,1965. 103. Woltman, H. W.: Neuritis associated with acromegaly. Arch. Neurol. Psychiatry 45:680, 1941. 104. Zapt, J., Schmid, C. H., and Froesch, E. R.: Biological and immunological properties of insulin-like growth factors (IGF) I and II. Clin. Endocrinol. Metab. 13:3, 1984.

L. Neuropathy Associated with Nutritional and Vitamin Deficiencies

89 Polyneuropathy Caused by Nutritional and Vitamin Deficiency DAVID S. SAPERSTEIN AND RICHARD J. BAROHN

Thiamine Deficiency Pathogenesis Clinical Features Diagnosis Management Vitamin B12 Deficiency Pathogenesis Clinical Features

Diagnosis Management Vitamin E Deficiency Pathogenesis Clinical Features Diagnosis Management Vitamin B6 Deficiency

Nutritional deficiency polyneuropathies are currently uncommon, especially in developed countries. Aside from historical interest, however, these disorders still represent a significant percentage of the relatively small number of neuropathies that are potentially treatable. Although nutritional neuropathy patients are still seen in modern clinical practice, deficiency is more likely to be secondary to factors other than inadequate intake of nutrients. Malnutrition can still be seen in chronic alcoholics, but chronic illness, food fads, and weight loss surgery are now more likely causes for deficiency neuropathies. Deficiencies of some vitamins, such as vitamin B12 and vitamin E, may occur because of impaired gastrointestinal absorption rather than lack of the nutrient in the diet. In other cases, as with vitamin B6, deficiency may occur secondary to the effects of medications (e.g., isoniazid). Although the clinical and laboratory features of most nutritional polyneuropathies are similar to those of cryptogenic and other common polyneuropathies, certain features—such as a sudden onset of symptoms, cranial nerve deficits, or involvement of other parts of the

Vitamin B6 Toxicity Pellagra (Niacin Deficiency) Strachan’s Syndrome Gastric Surgery Emerging Syndrome: Copper Deficiency

nervous system—can serve as important indicators of a possible vitamin deficiency. Timely diagnosis is important because, for most deficiency polyneuropathies, prognosis depends on how early replacement therapy is begun.

THIAMINE DEFICIENCY Beriberi was the first nutritional deficiency disorder to be identified. Although it was not until the early 20th century that thiamine deficiency was identified as the etiology, by the 17th century beriberi was known as a cause of numbness, pain, and weakness. Thiamine deficiency is currently an uncommon cause of peripheral neuropathy in developed countries. It is now most likely seen as a consequence of chronic alcohol abuse, recurrent vomiting, total parenteral nutrition, and weight reduction surgery. Thiamine deficiency polyneuropathy can occur in normal healthy young adults who do not abuse alcohol but who engage in inappropriately restrictive diets.95 2051

2052

Diseases of the Peripheral Nervous System

Beriberi means “I can’t, I can’t” in Singhalese, the language of natives of what was once part of the Dutch East Indies (now Sri Lanka).29 Dry beriberi refers to neuropathic symptoms. The term wet beriberi is used when cardiac manifestations predominate (in reference to edema). By the 19th century, it was known that both forms of beriberi represented the same disorder. Beriberi was relatively uncommon until the late 1800s, when it became widespread among people for whom rice was a dietary mainstay. This epidemic was due to a new technique of processing rice that removed the germ from the rice shaft, rendering the so-called polished rice deficient in thiamine and other essential nutrients.17 The nutritional origin of beriberi was first appreciated in the 19th century by two naval physicians, one from Holland and one from Japan.18,67 The incidence of beriberi could be decreased by providing sailors diets similar to those given to officers.18,67 Beriberi was initially incorrectly attributed to protein deficiency. In 1897, Christiaan Eijkman, a Dutch military physician stationed in the Dutch East Indies, serendipitously discovered a disease similar to beriberi in chickens.17,18,67 When a supply of brown rice did not arrive, Eijkman used polished white rice and the chickens became ill. The camp superintendent later decided the white rice was too expensive and, after the chickens were put back on a diet of brown rice, they recovered. Eijkman surveyed the colony’s prisons and discovered that the incidence of beriberi was high in prisons where prisoners were fed white rice and low in prisons where prisoners were fed brown rice.17,18 He initially thought beriberi was due to a toxin. In 1911, Casimir Funk, a Polish biochemist, isolated a watersoluble substance from rice polishings.37,38 Birds fed a diet deficient in this substance developed a neuropathy similar to beriberi.17,44 This substance, initially named “vitamine” to indicate it was a vital amine, was later synthesized and named thiamine.17,44,145

Pathogenesis Thiamine, also called vitamin B1, is a water-soluble vitamin. It is present in most animal and plant tissues, but the greatest sources are unrefined cereal grains, wheat germ, yeast, soybean flour, and pork.89 Thiamine is absorbed in the small intestine by both passive diffusion and active transport. In the jejunum, thiamine is converted to thiamine pyrophosphate (TPP), which is the main coenzyme form of thiamine.89 Compared to other micronutrients, body stores of thiamine are low. In addition, thiamine’s half-life is only 10 to 14 days. Therefore, a continuous source of dietary thiamine is necessary to prevent deficiency. The recommended daily allowance ranges from 1 to 1.5 mg/day.89 Thiamine in excess of that stored is excreted in the urine. There is no toxicity from overingestion of thiamine. Thiamine and TPP play important roles within mitochondria in the decarboxylation of ␣-keto acids to coenzyme

A (CoA) moieties, an important process in ATP synthesis.89 TPP is a coenzyme in reactions involved in myelin formation.22,23 Thiamine may also have an effect on nerve conduction and transmission by altering membrane sodium channel function.22,119

Clinical Features Symptoms usually develop gradually over weeks to months, but sometimes they may come on rapidly over a few days.141,145 Fatigue, irritability, and muscle cramps may appear within days to weeks of nutritional deficiency.7 Symptoms of neuropathy follow more prolonged deficiency. These begin with mild sensory loss and/or burning dysesthesias in the toes and feet. There is often associated aching and cramping in the lower legs. Pain may be the predominant symptom. With progression, patients develop features of a nonspecific generalized polyneuropathy, with distal sensory loss in the feet and hands.141,145 There may be mild distal weakness of the extensors of the ankles, fingers, and wrists. Reflexes are decreased distally. Classic descriptions of beriberi include involvement of the recurrent laryngeal nerve, producing hoarseness.141,145 Cranial nerve involvement manifesting as tongue and facial weakness may also occur.141,145 Oculomotor muscle weakness and nystagmus have been attributed to beriberi, but these manifestations are more likely due to coexistent Wernicke’s disease. Optic neuropathy has also been described in beriberi, but these cases may have represented Strachan’s syndrome (see later).26

Diagnosis Blood and urine assays for thiamine are not reliable for diagnosis of deficiency.89 Erythrocyte transketolase activity and the percentage increase in activity (in vitro) following the addition of TPP may be more accurate and reliable.10,12,61,62 This test, however, has not been sufficiently studied to establish its precise sensitivity and specificity. Testing must be performed before thiamine supplementation is given. Because thiamine is often administered quickly when there is a possibility of deficiency, this test is of limited help. Elevations of lactate and pyruvate are seen in thiamine deficiency, but these abnormalities are not specific enough for diagnostic purposes.135,144 Electrodiagnostic testing shows nonspecific findings of an axonal sensorimotor polyneuropathy.27,28,53,95 Nerve biopsies show axonal degeneration.95,135

Management When a diagnosis of thiamine deficiency is made or suspected, thiamine replacement should be provided until proper nutrition is restored. Thiamine is usually given

Polyneuropathy Caused by Nutritional and Vitamin Deficiency

intravenously or intramuscularly at a dose of 100 mg/day. Although cardiac manifestations show a striking response to thiamine replacement, neurologic improvement is usually more variable and less dramatic.63,141 Slow but clear improvement may be seen over 6 to 12 months.80 Motor deficits may respond better than sensory.72 Some improvement is expected in most patients, but this typically occurs slowly. In patients with severe neuropathy, there may be permanent deficits.53

VITAMIN B12 DEFICIENCY The most well-described neurologic consequence of vitamin B12, or cobalamin, deficiency is the syndrome of subacute combined degeneration.107,141 Pathology occurs in the posterior columns and lateral cortical spinal tracts, producing a clinical picture of vibratory and position sense deficits along with paraparesis and extensor plantar responses. Although there is pathologic and electrophysiologic evidence for peripheral nerve involvement in vitamin B12 deficiency,1,24,32,33,64,73,87,88,138 there is controversy over how often polyneuropathy occurs in the absence of myelopathy.19,97,141 Vitamin B12 deficiency polyneuropathy is the deficiency neuropathy most likely to be seen by clinicians in developed countries. Vitamin B12 deficiency polyneuropathy may represent up to 8% of all polyneuropathy cases seen at tertiary centers.110 In contrast to most of the other deficiency neuropathies, vitamin B12 deficiency is only rarely due to inadequate dietary intake. Pernicious anemia and other conditions of malabsorption are the most common causes.

Pathogenesis Vitamin B12 is present only in animal products. Strict vegetarians (vegans) need to ingest some form of vitamin B12 supplement. If a vegetarian eats eggs or dairy products, vitamin B12 needs will be met. Vitamin B12 is an essential cofactor for only two enzymatic reactions in humans. One reaction is the conversion of homocysteine (Hcy) to methionine. In addition, vitamin B12 is vital for converting methylmalonyl-CoA to succinyl-CoA.41 It is not known precisely how vitamin B12 deficiency produces polyneuropathy, but impaired formation of myelin appears to play a role. One hypothesis is that impaired methionine synthesis causes depletion of S-adenosylmethionine, which is needed for methylation of myelin phospholipids. Another possibility is that precursors of succinyl-CoA (methylmalonic acid [MMA] and propionic acid) build up and serve as abnormal substrates for fatty acid synthesis, leading to the production of abnormal myelin.41 In the stomach, vitamin B12 is liberated from food and binds to R proteins (transcobalamin I and III) in saliva and

2053

gastric juice. It is separated from the R proteins in the duodenum and then binds to intrinsic factor, a glycoprotein produced by gastric parietal cells. The vitamin B12–intrinsic factor complex is absorbed in the terminal ileum. The carrier for vitamin B12 within the blood stream is transcobalamin II. An acidic environment in the stomach is needed for vitamin B12 to be released from food protein. Therefore, achlorhydria will prevent vitamin B12 absorption. Pernicious anemia is the most common cause of vitamin B12 deficiency.49,137 This autoimmune disorder is more common in African Americans and those of Northern European background. Onset is most often after age 60, although earlier onset is often seen in African American and Hispanic women.16 Conditions other than pernicious anemia producing vitamin B12 deficiency include dietary avoidance (vegetarians), gastrectomy, gastric bypass surgery, ileal disease, pancreatic insufficiency, bacterial overgrowth, fish tapeworm, abuse of nitrous oxide, and, possibly, histamine2 blockers and proton pump inhibitors.15 An underappreciated cause of vitamin B12 deficiency is food-cobalamin malabsorption. This normally occurs in older individuals and is due to an impaired ability to absorb vitamin B12 that is bound in food protein.14 In a significant number of patients with vitamin B12 deficiency, no specific cause is identified.83,102 Genetic disorders are rare causes of vitamin B12 deficiency, and most present in childhood. Only R protein deficiency produces polyneuropathy.13,123

Clinical Features When features of both polyneuropathy and myelopathy are found in a patient, the need to look for vitamin B12 deficiency is usually obvious. A less obvious matter is determining which polyneuropathy patients might have vitamin B12 deficiency. Most series addressing the neurologic complications of vitamin B12 deficiency have not distinguished between myelopathy and neuropathy. Certain findings, such as Babinski’s signs and spinal sensory levels, clearly indicate myelopathy, and decreased or absent reflexes suggest polyneuropathy. However, differentiation on clinical grounds can be difficult. The neuropathy associated with vitamin B12 deficiency has been the subject of a few small series, but, until recently, the clinical presentation has not been well described.110 It is important to keep in mind that the prevalence of both vitamin B12 deficiency and cryptogenic polyneuropathy increases with age.4,82,102 Therefore, in a polyneuropathy patient with vitamin B12 deficiency, the deficiency might be a chance co-occurrence rather than the cause of the neuropathy. In most cases, the clinical features of vitamin B12 deficiency polyneuropathy are indistinguishable from those of a cryptogenic sensory or sensorimotor polyneuropathy.110,147 Distal numbness and paresthesias are present in most patients. Concomitant involvement of both the upper and lower limbs is more likely in vitamin B12 deficiency

2054

Diseases of the Peripheral Nervous System

polyneuropathy than in cryptogenic polyneuropathy patients.110 The neuropathy may be painful. Weakness sometimes occurs but is typically mild. In contrast to other polyneuropathies, the polyneuropathy of vitamin B12 deficiency may begin in the hands or hands and feet simultaneously.49,110,143 Some patients with vitamin B12 deficiency manifest only numbness in their hands.109,143 In the few cases that have been described in detail, myelopathy was the presumed etiology because nerve conduction studies were normal and somatosensory evoked potentials and/or magnetic resonance imaging (MRI) indicated lesions in the cervical spinal cord.56,109 Therefore, it is possible that the increased incidence of upper limb involvement observed in vitamin B12 deficiency polyneuropathy patients is the result of concomitant myelopathy. The sudden onset of myelopathy or myeloneuropathy following exposure to nitrous oxide or general anesthesia in patients with previously unappreciated vitamin B12 deficiency has been well described.36,45,52,70,79,86,109,117,122 Nitrous oxide, which transforms vitamin B12 into a biologically inactive form, is used to dilute inhalational general anesthesia in most procedures. Vitamin B12–deficient patients may also present with just polyneuropathy following surgery.110 In addition, a sudden onset of neuropathy symptoms may occur in vitamin B12–deficient patients in the absence of any exposure to anesthesia or nitrous oxide.110 Among patients who present with what appears to be a vitamin B12 deficiency polyneuropathy, coexisting myelopathy cannot necessarily be excluded. Based on clinical examination, these patients appear to have only polyneuropathy. Brisk reflexes or brisk knee reflexes coupled with absent ankle reflexes are seen in a small percentage of vitamin B12 deficiency polyneuropathy patients but also in cryptogenic polyneuropathy patients.110 Assessing small versus large sensory fiber function will not distinguish vitamin B12 deficiency polyneuropathy patients from cryptogenic polyneuropathy patients.110 Vitamin B12 deficiency may also produce optic neuropathy and cognitive disturbances.49 Most vitamin B12–deficient patients with neurologic manifestations will not have macrocytosis or anemia.49

Diagnosis Commonly used serum vitamin B12 assays are not sufficiently sensitive or specific. A significant proportion of vitamin B12–deficient patients may have serum vitamin B12 levels that are within the normal range.15,115 Measurement of the serum metabolites MMA and Hcy can significantly improve diagnostic sensitivity.83,115 Among persons with serum vitamin B12 levels within the normal range, MMA and Hcy testing will reveal vitamin B12 deficiency in 5% to 10% of those having a serum vitamin B12 level less than 300 pg/mL, and in 0.1% to 1% of those with a serum vitamin B12 level greater than 300 pg/mL.83 In the authors’ series, if

serum metabolites had not been tested, 12 polyneuropathy patients with vitamin B12 deficiency but normal serum vitamin B12 levels would have been diagnosed as having cryptogenic polyneuropathy.110 These 12 patients represented 44% of 27 vitamin B12–deficient polyneuropathy patients identified over a 2-year period. Prior to testing of MMA and Hcy, these patients had been classified as cryptogenic and represented 15% of cryptogenic polyneuropathy patients seen over the 2 years. There were no differences between the B12 deficiency polyneuropathy patients with low serum vitamin B12 levels and those having normal vitamin B12 levels but elevated serum metabolites. It is unlikely that the diagnosis of vitamin B12 deficiency in this group of patients was spurious because the incidence of pernicious anemia was more than 25 times that expected in an aged-matched population.110 It has been recommended that MMA and Hcy be measured in polyneuropathy patients having serum vitamin B12 levels less than 300 pg/mL.41 Some authors argue for metabolic testing in patients with vitamin B12 levels less than 350 pg/mL.82,102 Among polyneuropathy patients with serum vitamin B12 levels above 350 pg/mL, measurement of MMA and Hcy should be considered in patients with an increased likelihood of vitamin B12 deficiency: those with a sudden onset of symptoms, symptoms beginning in the hands, findings suggestive of myelopathy, or risk factors for vitamin B12 malabsorption. Although increased serum levels of MMA or Hcy suggest vitamin B12 deficiency, there are other causes that need to be considered. Hypovolemia, renal insufficiency, and certain genetic disorders can produce elevations of either MMA or Hcy.15 Hypothyroidism, increased age, and deficiency of folate or pyridoxine may elevate Hcy.15 Genetic causes of MMA elevations are very rare, and vitamin B12 is the only deficiency that raises MMA levels. If there is a question as to whether an elevated MMA or Hcy level truly indicates vitamin B12 deficiency, empirical vitamin B12 replacement therapy can be given and metabolite levels repeated after a few weeks. If an elevated MMA or Hcy level is secondary to vitamin B12 deficiency, it will return to normal after 1 to 2 weeks of replacement therapy.130 In the absence of symptomatic gastrointestinal disease, it probably is not necessary to seek a diagnosis of pernicious anemia in a patient with vitamin B12 deficiency because this information will not alter management.130 In the appropriate clinical setting, an elevated MMA is very specific for vitamin B12 deficiency. The Schilling test is often performed to diagnose pernicious anemia. However, because of concerns about accuracy, cost, and radiation exposure, a number of hospitals have stopped performing this test.134 Anti–intrinsic factor antibodies are highly specific for pernicious anemia but are found in only approximately 50% of patients.20 Anti–parietal cell antibodies can be detected in up to 90% of pernicious anemia patients, but may also been found in healthy older

Polyneuropathy Caused by Nutritional and Vitamin Deficiency

2055

adults.20 Elevated serum levels of gastrin are approximately 70% sensitive and specific for pernicious anemia.92 The combination of elevated gastrin and anti–parietal cell antibodies may be used to diagnose pernicious anemia.90 Early electrodiagnostic series reported demyelinating neuropathy, but these measured only conduction velocities.24,87 Most studies employing modern nerve conduction techniques report findings of axonal neuropathy.32,33,64,88,110 The authors have reported the largest series of vitamin B12 deficiency polyneuropathy patients studied with modern electrodiagnostic techniques.110,112 Vitamin B12 deficiency polyneuropathy patients are less likely to have nerve conduction study and electromyography (EMG) abnormalities compared with cryptogenic polyneuropathy patients, but, when abnormal, these studies indicate axonal polyneuropathy. For the most part, electrodiagnostic results in vitamin B12 deficiency polyneuropathy patients resemble those observed in cryptogenic polyneuropathy patients.112 There are no findings that, when detected in an individual polyneuropathy patient, might suggest vitamin B12 deficiency. Quantitative sensation testing will be abnormal in the majority of vitamin B12 deficiency polyneuropathy patients, but no pattern of abnormalities is unique to these patients.110,111 Somatosensory evoked potentials will indicate spinal lesions in a significant percentage of vitamin B12 deficiency patients with neurologic symptoms.32,33,64,75,138,149 Somatosensory evoked potentials were abnormal in 50% of the authors’ vitamin B12 deficiency polyneuropathy patients.110 Therefore, evoked potentials may indicate a subclinical myelopathy in patients with vitamin B12 deficiency having symptoms and findings suggestive only of neuropathy. Nerve biopsies show axonal, not demyelinating, features.1,73,88 The spinal cord pathology in vitamin B12 deficiency shows demyelination, and this probably contributes to the misperception that the neuropathy of vitamin B12 deficiency is demyelinating.

randomized trial that compared treatment with 2000 mg oral vitamin B12 per day to 1000 mg intramuscular vitamin B12 per month.78 Both groups showed similar improvements in hematologic indices, serum MMA and Hcy, and neurologic symptoms. However, only 8 of 33 subjects had neurologic symptoms, and the methods for assessing improvement are not described. A superior clinical outcome was not found for either group, but the patients treated with oral vitamin B12 had significantly higher serum vitamin B12 and lower MMA levels at the end of the study (4 months).78 In the absence of more convincing data regarding patients with neurologic deficits, the authors’ practice is to use intramuscular vitamin B12 replacement therapy. However, a reasonable compromise may be to switch to oral therapy after 1 or more months and periodically monitor MMA and/or Hcy levels. The neurologic deficits of vitamin B12 deficiency are considered to respond rapidly to parenteral replacement therapy.49 In approximately 2% of patients, there may be a “coasting” phenomenon wherein sensory symptoms initially worsen during the first month of treatment.49 The reason for this is not known. The response to treatment of vitamin B12 deficiency polyneuropathy, separate from other neurologic complications of vitamin B12, has not been well studied. Patients with vitamin B12 deficiency polyneuropathy probably do not show an immediate response to treatment and may not respond at all.88,110 This makes sense because the underlying pathology of the neuropathy is axonal degeneration. The duration of symptoms is an important determinant of treatment response.49,55,140

Management

Pathogenesis

The ideal approach to vitamin B12 replacement is not known. A common regimen is 1000 ␮g intramuscularly (IM) for 5 to 7 days, followed by monthly 500- to 1000-␮g intramuscular injections.41 An alternate approach is to begin with 1000 ␮g IM/wk for 1 month, followed by 1000 ␮g IM/mo thereafter. It may be possible to treat vitamin B12 deficiency with oral replacement. Even individuals with pernicious anemia are able to absorb small amounts of orally administered vitamin B12.20 The daily requirement for vitamin B12 is 1 to 2 ␮g, and approximately 1% of orally administered vitamin B12 can be absorbed by patients with pernicious anemia.20,78 Consequently, an oral vitamin B12 dose of 1000 to 2000 ␮g/day should be sufficient. Several series reported successful oral treatment of vitamin B12 deficiency (reviewed by Layzer78). There has been a

The term vitamin E is usually used for ␣-tocopherol, the most active of the four main types of vitamin E. It is a lipidsoluble vitamin that is absorbed in the small intestine after being acted on by bile acids. Vitamin E travels with chylomicrons to the liver, where it is incorporated into verylow-density lipoproteins. This step requires ␣-tocopherol transfer protein. Deficiency of this enzyme can produce isolated vitamin E deficiency (see later).96 Vitamin E appears to have important roles in scavenging free radicals and in maintaining cell membrane structure.136 Vitamin E deficiency usually occurs secondary to lipid malabsorption or in uncommon disorders of vitamin E absorption or transport. One hereditary disorder is abetalipoproteinemia, a rare autosomal dominant disorder characterized by steatorrhea, pigmentary retinopathy,

VITAMIN E DEFICIENCY Because vitamin E is present in animal fat, vegetable oils, and various grains, deficiency is usually due to factors other than insufficient intake.125

2056

Diseases of the Peripheral Nervous System

acanthocytosis, and progressive ataxia.69,94 Patients with cystic fibrosis may also have vitamin E deficiency secondary to steatorrhea. There are genetic forms of isolated vitamin E deficiency not associated with lipid malabsorption.46,127 Mutations in the ␣-tocopherol transfer protein gene on chromosome 8q13 affect vitamin E absorption by impairing its incorporation into very-low-density lipoproteins.40,96 Vitamin E deficiency may also occur as a consequence of various cholestatic and hepatobiliary disorders as well as short bowel syndromes resulting from the surgical treatment of intestinal disorders.47,54,114 The principal pathologic features of vitamin E deficiency include swelling and degeneration of axons in the posterior columns and spinocerebellar tracts.60,105,133 There are also neuronal loss and lipofuscin accumulation in the gracile and cuneate nuclei. The basal ganglia may be involved as well.

Clinical Features Vitamin E deficiency is associated with a number of neurologic manifestations. Clinical features may not appear until many years after the onset of deficiency. The onset of symptoms tends to be insidious, and progression is slow. The main clinical features are spinocerebellar ataxia and polyneuropathy, thus resembling Friedreich’s ataxia, other spinocerebellar ataxias, and adult-onset hexosaminidase A deficiency.125 Symptoms of intestinal malabsorption can provide a clue to vitamin E deficiency, although vitamin E deficiency can occur in the absence of gastrointestinal symptoms.59 Patients manifest progressive ataxia and signs of posterior column dysfunction, such as impaired joint position and vibratory sensation. Because of the polyneuropathy, there is hyporeflexia, but plantar responses may be extensor as a result of the spinal cord involvement. Other neurologic manifestations may include ophthalmoplegia, pigmented retinopathy, night blindness, dysarthria, pseudoathetosis, dystonia, and tremor. Vitamin E deficiency may present as an isolated polyneuropathy, but this is very rare.6,98 The yield of checking serum vitamin E levels in patients with isolated polyneuropathy is extremely low, and this test should not be part of routine practice.

Diagnosis Diagnosis is made by measuring ␣-tocopherol levels in the serum. In hyperlipidemic states the vitamin E level may be normal even when deficiency is present.126 Therefore, the ratio of total serum vitamin E to the total serum lipid concentration has been suggested as a superior assessment of vitamin E status.126 Electrodiagnostic studies show sensory neuropathy or neuronopathy. Sensory responses are reduced or absent; motor responses and EMG are usually normal.5,46,59,105 Electrodiagnostic abnormalities can normalize following vitamin E replacement therapy.71 Somatosensory evoked potentials will usually show a delay in central conduction

as a result of the posterior column dysfunction.11,46,114 Visual evoked potentials may be abnormal in patients with pigmented retinopathy, but electroretinography is more sensitive.68,74,146 Sural nerve biopsies show nonspecific changes. Loss of large myelinated fibers, evidence of axonal degeneration, and regenerating clusters are noted.139,148

Management The primary goal is to prevent progression, but neurologic recovery does occur. Replacement is with oral vitamin E, but high doses are needed. The specific dose varies with the cause of deficiency.125 For patients with isolated vitamin E deficiency, treatment consists of 1500 to 6000 IU/day in divided doses. In cases of chronic cholestasis, treatment begins with 50 IU/kg/day. The dose is increased in 50-IU/kg increments up to a dose of 200 IU/kg/day to achieve normalization of the serum tocopherol-to-lipid ratio. In patients with cystic fibrosis who are receiving oral pancreatic enzyme therapy, doses of 5 to 10 IU/kg/day are usually sufficient. Cystic fibrosis patients with cholestasis should be treated with the chronic cholestasis regimen. Doses of 300 to 5400 IU/day have been used in patients with short bowel syndrome. Patients with abetalipoproteinemia are usually treated with 150 to 300 IU/kg/day in divided doses. Treatment with 15,000 to 20,000 IU/day of vitamin A should also be given.

VITAMIN B6 DEFICIENCY Vitamin B6, or pyridoxine, can produce neuropathic manifestations from both deficiency and toxicity. Vitamin B6 deficiency is most commonly seen in patients treated with the antituberculosis drug isoniazid or the antihypertensive drug hydralazine (which is chemically related to isoniazid).39,84,121 Vitamin B6 deficiency may also result from the malnutrition resulting from chronic alcoholism and may occur in patients receiving chronic peritoneal dialysis.93 The polyneuropathy of vitamin B6 deficiency is nonspecific, manifesting as a generic axonal sensorimotor polyneuropathy.39,84 Vitamin B6 deficiency can be detected by direct assay. Vitamin B6 supplementation with 50 to 100 mg/day is suggested for patients being treated with isoniazid or hydralazine.85 This same dose is appropriate for replacement in cases of nutritional deficiency.

VITAMIN B6 TOXICITY In contrast to the nondescript polyneuropathy of vitamin B6 deficiency, excessive intake of vitamin B6 can produce a striking clinical syndrome of ataxic sensory neuropathy.3,25,99,116 Numbness, paresthesias, and impaired balance are present in most cases. If the upper limbs are involved, there can be

Polyneuropathy Caused by Nutritional and Vitamin Deficiency

pseudoathetosis. Additional symptoms in some patients are pain, Lhermitte’s sign and perioral numbness. Although position and vibration sensation are significantly affected, touch, pain, and temperature also are abnormal in most patients. There are absent ankle jerks or complete areflexia. Strength remains intact, and patients do not have upper motor neuron findings. The site of the lesion appears to be the dorsal root ganglion. The presence of Lhermitte’s sign in some patients implicates spinal cord involvement as well. Electrodiagnostic studies show abnormal or absent sensory responses with sparing of motor conduction. Nerve biopsies show loss of large and small myelinated fibers. Patients can recover after cessation of high-dose vitamin B6 intake, although symptoms and deficits may persist. This syndrome occurs with vitamin B6 doses as low as 200 mg/day. Standard multivitamin pills usually contain only 2 mg (the recommended daily allowance). Patients with polyneuropathy from any cause may be sensitive to toxicity from vitamin B6 and, in these patients, daily intake of more than 100 mg is probably best avoided.

PELLAGRA (NIACIN DEFICIENCY) Pellagra is produced by deficiency of niacin. Although pellagra may be seen in alcoholics, this disorder has essentially been eradicated in most Western countries by means of enriching bread with niacin. Nevertheless, pellagra continues to be a problem in a number of underdeveloped regions, particularly in Asia and Africa, where corn is the main source of carbohydrate. Pellagra can also be seen in the genetic condition Hartnup disease as well as in association with carcinoid syndrome. Pellagra is characterized by the “three Ds”: dermatitis, diarrhea, and dementia. The skin rash consists of scaly, reddish brown lesions distributed over the face, neck, and dorsal surface of the arms and legs.58 It may resemble a sunburn.30 Neurologic manifestations are variable because problems can develop in the brain and spinal cord as well as peripheral nerves. There is some controversy over how often neurologic symptoms are due to neuropathy as opposed to other nervous system lesions.141 When peripheral nerves are involved, the neuropathy is usually mild and resembles beriberi.81,141 Peripheral nerve pathology is axonal degeneration. Treatment is with niacin, 40 to 250 mg/day, but this may not necessarily produce resolution of neuropathy symptoms.81,106,120

STRACHAN’S SYNDROME In 1888, Strachan described a syndrome of polyneuropathy, amblyopia, and orogenital dermatitis in Jamaican sugarcane workers.131,132 The polyneuropathy is painful and primarily sensory. The vision loss is characterized by central or centrocecal scotomas. A similar syndrome was identified in

2057

Allied prisoners during World War II.26,34 Pathology findings include demyelination and axon loss in the posterior columns of the spinal cord as well as optic nerve fiber loss in the so-called papillomacular bundle.34 Peripheral nerves were not carefully studied. An epidemic of a syndrome resembling Strachan’s syndrome occurred in Cuba in 1992 and 1993.104 In addition to polyneuropathy and optic neuropathy, patients had sensorineural hearing loss and dorsolateral myelopathy.104 Sural nerve biopsies from these patients showed an axonal neuropathy primarily affecting large myelinated fibers.8 These patients did not appear to be malnourished. However, a deficiency of micronutrients, including thiamine, cobalamin, folate, and sulfur amino acids, was postulated as the cause.104 Most patients improved after treatment with B vitamins and folate.

GASTRIC SURGERY Polyneuropathy may occur following gastric surgery for ulcer, cancer, or weight reduction. This usually occurs in the context of rapid, significant weight loss and recurrent, protracted vomiting.21,31,48,100,101,113,129 The clinical picture is one of acute or subacute sensory loss and weakness. Weakness may be quite significant, leading to inability to walk. Some cases are associated with features of WernickeKorsakoff syndrome. Neuropathy following weight loss surgery usually occurs in the first several months after surgery.21 The latency between surgery and symptoms can be months to years in patients who have had total or partial gastrectomy for ulcer or cancer.51,72 In the cases of polyneuropathy after weight loss surgery, intractable vomiting was almost always described. Weight reduction surgical procedures include gastrojejunostomy, gastric stapling, vertical banded gastroplasty, and gastrectomy with Roux-en-Y anastomosis. The initial manifestations are usually numbness and paresthesias in the feet. Although thiamine deficiency seems to be a factor (given the frequent co-occurrence of the Wernicke-Korsakoff syndrome), pain is not a prominent feature. Thiamine deficiency is not well documented in the reported cases. In some cases, one or more vitamin deficiencies are identified.66 In many cases, no specific deficiency is identified. Peripheral nerve pathology shows nonspecific axonal neuropathy.2 Management consists of parenteral vitamin supplementation to include, especially, thiamine. Improvement has been observed following supplementation, parenteral nutritional support, and reversal of the surgical bypass.2,66 Patients developing vomiting after weight reduction surgery should be supported with total parenteral nutrition and vitamins. There can be good recovery if treatment is initiated quickly after symptom onset. However, a number of patients can have persistent sensory loss and weakness. The duration and severity of deficits before identification and treatment of neuropathy are important predictors of final outcome.

2058

Diseases of the Peripheral Nervous System

EMERGING SYNDROME: COPPER DEFICIENCY A syndrome that has only recently been described is myeloneuropathy or myelopathy secondary to copper deficiency. In 2001, Schleper and Stuerenburg118 reported a woman with myelopathy presumed secondary to copper deficiency resulting from prior gastric surgery for peptic ulcer disease. In the following 2 years, seven similar cases were reported.42,50,76,77,103 The clinical presentation has been similar, with most patients presenting with lower limb paresthesias, weakness, spasticity, and gait difficulties. Large fiber sensory function is impaired, reflexes are brisk, and plantar responses are extensor. In some cases light touch and pinprick sensation are affected, and nerve conduction studies indicate sensorimotor axonal polyneuropathy in addition to myelopathy.50,77 Greenberg and Briemberg42 described a patient with myelopathy and severe motor axonopathy. In some cases,103 brain involvement was emphasized based on demyelinating lesions on brain MRI, but the clinical syndrome was primarily myelopathy. One of these patients also had saccadic dysmetria.103 There is a case reported in the hematology literature of a patient with copper deficiency and a “severe progressive peripheral neuropathy” and optic neuropathy, but no specific details are provided.43 In most cases, patients also had microcytic anemia and neutropenia42,77,103,118; one patient had pancytopenia.50 Hematologic abnormalities are a known complication of copper deficiency.9 There was abnormal T2-weighted signal on cervical spine MRI in two patients.42,118 In other cases, brain MRI showed features suggestive of demyelination but spine imaging was normal.103 Cerebrospinal fluid was normal or showed mildly elevated protein or immunoglobulin synthesis rate.42,50,103 In three patients copper deficiency was likely secondary to prior gastric surgery.77,118 Copper is absorbed in the stomach and proximal jejunum.128 In each of the other cases, low serum copper was accompanied by high serum zinc. Excess zinc is an established cause of copper deficiency, and myelosuppression caused by high zinc and low copper (in the absence of neurologic complications) has been described (reviewed by Irving et al.57). Zinc upregulates enterocyte production of metallothionein, which results in decreased absorption of copper.35 One patient was taking high-dose zinc supplements,76 but in the other cases the cause for high serum zinc levels remains undetermined.42,50,103 It does not appear that elevated zinc levels, by themselves, are toxic.124 Other potential causes of copper deficiency include malnutrition, prematurity, total parenteral nutrition, and ingestion of copper chelating agents.9,128 Following oral or intravenous copper replacement, some patients showed neurologic improvement relatively quickly76,77,118; in other cases improvement was seen only after months or years50,103 or not at all.42 Improvement was

only partial. In contrast, every patient’s hematologic indices completely normalized in response to copper replacement therapy. Based on the available evidence, there does appear to be a neurologic syndrome of copper deficiency, consisting of myeloneuropathy or myelopathy, with most patients having a concomitant hematologic syndrome characterized by anemia and neutropenia or pancytopenia. There is an analogous disorder seen in ruminant animals called swayback or endozootic ataxia.65 This is characterized by anemia and myelopathy with axonal degeneration and demyelination.65 Cerebral neuronal degeneration occurs as well. Thus far, there has been no autopsy or peripheral nerve biopsy from human patients. The authors have seen several patients with unexplained myelopathy and bone marrow findings of myelodysplastic syndrome.91 Unfortunately, all these patient died before the recent reports of copper deficiency and myelopathy, and copper levels were not checked. However, one patient showed full hematologic improvement in response to corticosteroids, arguing against copper deficiency as the cause. In addition, a recent patient with unexplained myeloneuropathy and erythropoietin-dependent anemia had normal serum copper levels.108 Therefore, it appears that there may be additional syndromes of concomitant myeloneuropathy and myelosuppression unrelated to copper deficiency.

REFERENCES 1. Abarbanel, J. M., Frishers, S., and Osimani, A.: Vitamin B12 deficiency neuropathy: sural nerve biopsy study. Isr. J. Med. Sci. 22:909, 1986. 2. Akhtar, M., Collins, M. P., and Kissel, J. T.: Acute postgastric reduction surgery (APGARS) neuropathy: a polynutritional, multisystem disorder. Neurology 58(Suppl.):A68, 2002. 3. Albin, R. L., Albers, J. W., Greenberg, H. S., et al.: Acute sensory neuropathy-neuronopathy from pyridoxine overdose. Neurology 37:1729, 1987. 4. Beghi, E., and Monticelli, M. L., for the Italian General Practitioner Study Group.: Chronic symmetric symptomatic polyneuropathy in the elderly: a field screening investigation in two Italian regions. I. prevalence and general characteristics of the sample. Neurology 45:1832, 1995. 5. Bertoni, J. M., Abraham, F. A., Falls, H. F., and Itabashi, H. H.: Small bowel resection with vitamin E deficiency and progressive spinocerebellar syndrome. Neurology 34:1046, 1984. 6. Binder, H. J., Solitare, G. B., and Spiro, H. M.: Neuromuscular disease in patients with steatorrhoea. Gut 8:605, 1967. 7. Blass, J. P.: Vitamin and nutritional deficiencies. In Siegel, G. J., Agranoff, B. W., Albers, R. W., and Molinoff, P. B. (eds.): Basic Neurochemistry, 5th ed. New York, Raven Press, p. 749, 1994. 8. Borrajero, I. , Perez, J. L., Dominguez, C., et al.: Epidemic neuropathy in Cuba: morphological characterization of

Polyneuropathy Caused by Nutritional and Vitamin Deficiency

9.

10. 11.

12.

13. 14. 15. 16.

17. 18. 19. 20. 21.

22. 23. 24. 25.

26.

27.

28.

29.

30.

peripheral nerve lesions in sural nerve biopsies. J. Neurol. Sci. 127:68, 1994. Bottomley, S. S.: Sideroblastic anemias. In Lee, G. R., Foerster, J., Lukens, J., et al. (eds.): Wintrobe’s Clinical Hematology, 10th ed. Baltimore, Williams & Wilkins, p. 1022, 1999. Brin, M.: Erythrocyte transketolase in early thiamine deficiency. Ann. N. Y. Acad. Sci. 98:528, 1962. Brin, M., Pedley, T. A., Lovelace, R. E., et al.: Electrophysiologic features of abetalipoproteinemia: functional consequences of vitamin E deficiency. Neurology 36:669, 1986. Brin, M., Tai, M., Ostashever, A. S., and Kalinsky, H.: The effect of thiamine deficiency on the activity of erythrocyte hemolysate transketolase. J. Nutr. 712:273, 1960. Carmel, R.: Plasma R binder deficiency. N. Engl. J. Med. 318:1401, 1988. Carmel, R.: Malabsorption of food cobalamin. Baillieres Clin. Hematol. 8:639, 1995. Carmel, R.: Current concepts in cobalamin deficiency. Annu. Rev. Med. 51:357, 2000. Carmel, R., and Johnson, C. S.: Racial patterns of pernicious anemia: early age at onset and increased frequency of intrinsic-factor antibody in black women. N. Engl. J. Med. 298:647, 1978. Carpenter, K. J., and Sutherland, B.: Eijkman’s contribution to the discovery of vitamins. J. Nutr. 125:155, 1995. Carter, K. C.: The germ theory, beriberi and the deficiency theory of disease. Med. Hist. 21:119, 1977. Chalk, C. H.: Acquired peripheral neuropathy. Neurol. Clin. 15:501, 1997. Chanarin, I.: The Megaloblastic Anemias, 2nd ed. Oxford, UK, Blackwell Scientific, 1979. Cirignotta, F., Manconi, M., Mondini, S., et al.: WernickeKorsakoff encephalopathy and polyneuropathy after gastroplasty for morbid obesity. Arch. Neurol. 57:1356, 2000. Collins, R. C., and Lonergan, E. T.: Transketolase and myelin. N. Engl. J. Med. 285:751, 1971. Cooper, J. R., and Pincus, J. H.: The role of thiamin in nervous tissue. Neurochem. Res. 4:223, 1979. Cox-Klazinga, M., and Endtz, L. J.: Peripheral nerve involvement in pernicious anemia. J. Neurol. Sci. 45:367, 1980. Dalton, K., and Dalton, M. J.: Characteristics of pyridoxine overdose neuropathy syndrome. Acta Neurol. Scand. 76:8, 1987. Denny-Brown, D. E.: Neurological conditions resulting from prolonged and severe dietary restrictions. Medicine (Baltimore) 26:41, 1947. Djoenaidi, W., and Notermans, S. L.: Electrophysiological evaluation of beri-beri polyneuropathy. Electromyogr. Clin. Neurophysiol. 30:97, 1990. Djoenaidi, W., Notermans, S. L., Gabreels-Festen, A. A., et al.: Experimentally induced beriberi polyneuropathy in chickens. Electromyogr. Clin. Neurophysiol. 35:53, 1995. Ericson, K.: Beriberi. Abilene, TX, AccessMed Health Information Library, Hendrick Health System, 2003. Available at: www.ehendrick.org/healthy/000193.htm Etheridge, E.: Pellagra. In Kiple, K. F. (ed.): The Cambridge World History of Human Disease. Cambridge, UK, Cambridge University Press, p. 918, 1993.

2059

31. Feit, H., Glasberg, M., Ireton, C., et al.: Peripheral neuropathy and starvation after gastric partitioning for morbid obesity. Ann. Intern. Med. 96:453, 1982. 32. Fine, E. J., and Hallett, M.: Neurophysiological study of subacute combined degeneration. J. Neurol. Sci. 45:331, 1980. 33. Fine, E. J., Soria, E., Paroski, M. W., et al.: The neurophysiological profile of vitamin B12 deficiency. Muscle Nerve 13:158, 1990. 34. Fischer, M.: Residual neuropathological changes in Canadians held prisoner of war by the Japanese (Strachan’s disease). Can. Serv. Med. J. 11:157, 1955. 35. Fiske, D. N., McCoy, H. E. III, and Kitchens, C. S.: Zincinduced sideroblastic anemia: report of a case, review of the literature, and description of the hematologic syndrome. Am. J. Hematol. 46:147, 1994. 36. Flippo, T. S., and Holder, W. D. Jr.: Neurologic degeneration associated with nitrous oxide anesthesia in patients with vitamin B12 deficiency. Arch. Surg. 128:1391, 1993. 37. Funk, C.: On the chemical nature of the substance which cures polyneuritis in birds induced by a diet of polished rice. J. Physiol. (Lond.) 43:395, 1911. 38. Funk, C.: The etiology of the deficiency diseases: beriberi, polyneuritis in birds, epidemic dropsy, scurvy, experimental scurvy in animals, infantile scurvy, ship beriberi, pellagra. J. State Med. 20:341, 1912. 39. Gammon, GD, Burge, FW, and King, G.: Neural toxicity in tuberculous patients treated with isoniazid (isonicotinic acid hydrazide). Arch. Neurol. Psychiatry 70:64, 1953. 40. Gotoda, T., Arita, M., Arai, H., et al.: Adult-onset spinocerebellar dysfunction caused by a mutation in the gene for ␣-tocopherol transfer protein. N. Engl. J. Med. 333:1313, 1995. 41. Green, R., and Kinsella, L. J.: Current concepts in the diagnosis of cobalamin deficiency. Neurology 45:1435, 1995. 42. Greenberg, S. A., and Briemberg, H. R.: A neurological and hematological syndrome associated with zinc and excess and copper deficiency. J. Neurol. 251:111, 2004. 43. Gregg, X. T., Reddy, V., and Prchal, J. T.: Copper deficiency masquerading as myelodysplastic syndrome. Blood 100:1493, 2002. 44. Griminger, P.: Casimir Funk, a biographical sketch (1884–1967). J. Nutr. 102:1105, 1972. 45. Hadzic, A., Glab, K., Sanborn, K. V., and Thys, D. M.: Severe neurologic deficit after nitrous oxide anesthesia. Anesthesiology 83:863, 1995. 46. Harding, A. E., Matthews, S., Jones S., et al.: Spinocerebellar degeneration associated with a selective defect of vitamin E absorption. N. Engl. J. Med. 313:32, 1985. 47. Harding, A. E., Muller, D. P. R., Thomas, P. K., and Willison, H. J.: Spinocerebellar degeneration secondary to chronic intestinal malabsorption: a vitamin E deficiency syndrome. Ann. Neurol. 12:419, 1982. 48. Harwood, S. C., Chodoroff, G., and Ellenberg, M. R.: Gastric partitioning complicated by peripheral neuropathy with lumbosacral plexopathy. Arch. Phys. Med. Rehabil. 68:310, 1987. 49. Healton, E. B., Savage, D. G., Brust, J. C., et al.: Neurologic aspects of cobalamin deficiency. Medicine 70:229, 1991.

2060

Diseases of the Peripheral Nervous System

50. Hedera, P., Fink, J. K., Bockenstedt, P. L., and Brewer, G. J.: Myelopolyneuropathy and pancytopenia due to copper deficiency and high zinc levels of unknown origin: further support for existence of a new zinc overload syndrome. Arch. Neurol. 60:1303, 2003. 51. Hoffman, P. M., and Brody, J. A.: Neurological disorders in patients following surgery for peptic ulcer. Neurology 22:450, 1972. 52. Holloway, K. L., and Alberico, A. M.: Postoperative myeloneuropathy: a preventable complication in patients with B12 deficiency. J. Neurosurg. 72:732, 1990. 53. Hong, C. Z.: Electrodiagnostic findings of persisting polyneuropathies due to previous nutritional deficiency in former prisoners of war. Electromyogr. Clin. Neurophysiol. 26:351, 1986. 54. Howard, L., Ovensen, L., Satya-Murti, S., and Chu, R.: Reversible neurological symptoms caused by vitamin E deficiency in a patient with short bowel syndrome. Am. J. Clin. Nutr. 36:1243, 1982. 55. Hyland, H. H., and Farquharson, R. F.: Subacute combined degeneration of the spinal cord in pernicious anemia. Arch. Neurol. Psychiatry 36:1166, 1936. 56. Imaiso, Y., Taniwaki, T., Yamada, T., et al.: Myelopathy due to vitamin B12 deficiency presenting only sensory disturbances in upper extremities: a case report. Rinsho Shinkeigaku 37:135, 1997. 57. Irving, J. A., Mattman, A., Lockitch, G., et al.: Element of caution: a case of reversible cytopenias associated with excessive zinc supplementation. CMAJ 169:129, 2003. 58. Isaac, S.: The “gauntlet” of pellagra. Int. J. Dermatol. 37:599, 1998. 59. Jackson, C. E., Amato, A. A., and Barohn, R. J.: Isolated vitamin E deficiency. Muscle Nerve 19:1161, 1996. 60. Jeffrey, G. P., Muller, D. P. R., Burroughs, A. K., et al.: Vitamin E deficiency and its clinical significance in adults with primary biliary cirrhosis and other forms of chronic liver disease. J. Hepatol. 4:307, 1987. 61. Jeyasingham, M. D., Pratt, O. E., Burns, A., et al.: The activation of red blood cell transketolase in groups of patients especially at risk from thiamin deficiency. Psychol. Med. 17:311, 1987. 62. Jeyasingham, M. D., Pratt, O. E., Shaw, G. K., and Thomson, A. D.: Changes in the activation of red blood cell transketolase of alcoholic patients during treatment. Alcohol. 22:259, 1987. 63. Jolliffe, N.: The diagnosis, treatment and prevention of vitamin B1 deficiency. Bull. N. Y. Acad. Med. 15:469, 1939. 64. Jones, S. J., Yu, Y. L., Rudge, P., et al.: Central and peripheral SEP defects in neurologically symptomatic and asymptomatic subjects with low vitamin B12 levels. J. Neurol Sci. 82:55, 1987. 65. Jubb, K. V. F., and Huxtable, C. R.: The nervous system. In Jubb, K. V. F., Kennedy, P. C., and Palmer, N. (eds.): Pathology of Domestic Animals, 4th ed. San Diego, Academic Press, p. 267, 1993. 66. Juhasz-Poscine, D., Archer, R. L., Rudnicki, S. A., et al.: Retrospective review of neurology complications associated with surgical treatment for morbid obesity. Neurology 58A:A247, 2002.

67. Jukes, T. H.: The prevention of and conquest of scurvy, beri-beri, and pellagra. Prev. Med. 18:877, 1989. 68. Kaplan, P. W., Rawal, K., Erwin, C. W., et al.: Visual and somatosensory evoked potentials in vitamin E deficiency with cystic fibrosis. Electroencephalogr. Clin. Neurophysiol. 71:266, 1988. 69. Kayden, H. L., Silber, R., and Kossman, C. E.: The role of vitamin E deficiency in the abnormal autohemolysis of acanthocytosis. Trans. Assoc. Am. Physicians 78:334, 1965. 70. Kinsella, L. J., and Green, R.: “Anesthesia paresthetica”: nitrous oxide-induced cobalamin deficiency. Neurology 45:1608, 1995. 71. Ko, H. Y., and Park-Ko, I.: Electrophysiologic recovery after vitamin E-deficient neuropathy. Arch. Phys. Med. Rehabil. 80:964, 1999. 72. Koike, H., Misu, K., Hattori, N., et al.: Postgastrectomy polyneuropathy with thiamine deficiency. J. Neurol. Neurosurg. Psychiatry 71:357, 2001. 73. Kosik, K. S., Mullins, T. F., Bradley, W. G., et al.: Coma and axonal degeneration in vitamin B12 deficiency. Arch. Neurol. 37:590, 1980. 74. Krendel, D. A., Gilchrist, J. M., Johnson, A. O., and Bossen, E. H.: Isolated deficiency of vitamin E with progressive neurologic deterioration. Neurology 37:538, 1987. 75. Krumholz, A., Weiss, H. D., Goldstein, P. J., and Harris, K. C.: Evoked responses in vitamin B12 deficiency. Ann. Neurol. 9:407, 1981. 76. Kumar, N., Gross, J. B., and Ahlskog, J. E.: Myelopathy due to copper deficiency. Neurology 61:273, 2003. 77. Kumar, N., McEvoy, K. M., and Ahlskog, J. E.: Myelopathy due to copper deficiency following gastrointestinal surgery. Arch. Neurol. 60:1782, 2003. 78. Kuzminski, A. M., Del Giacco, E. J., Allen, R. H., et al.: Effective treatment of cobalamin deficiency with oral cobalamin. Blood 92:1191, 1998. 79. Layzer, R. B.: Myeloneuropathy after prolonged exposure to nitrous oxide. Lancet 2:1227, 1978. 80. Layzer, R. B.: Neuromuscular Manifestations of Systemic Disease. Philadelphia, FA Davis, p. 319, 1985. 81. Lewy, F. H., Spies, T. D., and Aring, C. D.: The incidence of neuropathy in pellagra: the effect of carboxylase upon its neurologic signs. Am. J. Med. Sci. 199:840, 1940. 82. Lindenbaum, J., Rosenberg, I. H., Wilson, P. W., et al.: Prevalence of cobalamin deficiency in the Framingham elderly population. Am. J. Clin. Nutr. 60:2, 1994. 83. Lindenbaum, J., Savage, D. G., Stabler, S. P., and Allen, R. H.: Diagnosis of cobalamin deficiency: II. Relative sensitivities of serum cobalamin, methylmalonic acid, and total homocysteine concentrations. Am. J. Hematol. 34:99, 1990. 84. Lubing, H. N.: Peripheral neuropathy in tuberculosis patients treated with isoniazid. Am. Rev. Tuberc. 68:458, 1953. 85. Marcus, R., and Coulston, A. N.: Water-soluble vitamins. In Gilman, A. G., Goodman, L. S., Rall, T. W., and Murad, F. (eds.): Goodman and Gilman’s The Pharmacological Basis of Therapeutics, 7th ed. New York, Macmillan, p. 1551, 1985. 86. Marie, R. M., Le Biez, E., Busson, P., et al.: Nitrous oxide anesthesia-associated myelopathy. Arch. Neurol. 58:134, 2001.

Polyneuropathy Caused by Nutritional and Vitamin Deficiency 87. Mayer, R. F.: Peripheral nerve function in vitamin B12 deficiency. Arch. Neurol. 13:355, 1965. 88. McCombe, P. A., and McLeod, J. G.: The peripheral neuropathy of vitamin B12 deficiency. J. Neurol. Sci. 66:117, 1984. 89. McCormick, D. B., and Greene, H. L.: Vitamins. In Burtis, C. A., and Ashwood, E. R. (eds.): Tierz Textbook of Clinical Chemistry. Philadelphia, W. B. Saunders, p. 999, 1999. 90. Metz, J., Bell, A. H., Flicker, L., et al.: The significance of subnormal serum vitamin B12 concentration in older people: a case control study. J. Am. Geriatr. Soc. 44:1355, 1996. 91. Milford, C., Saperstein, D. S., McVey, A. L., et al.: Myeloneuropathy in the setting of myelodysplasia. J. Child Neurol. 18:212, 2003. 92. Miller, A., Slingerland, D. W., Cardarelli, J., and Burrows, B. A.: Further studies on the use of serum gastrin levels in assessing the significance of low serum B12 levels. Am. J. Hematol. 31:194, 1989. 93. Moriwaki, K., Kanno, Y., Nakamoto, H., et al.: Vitamin B6 deficiency in elderly patients on chronic peritoneal dialysis. Adv. Perit. Dial. 16:308, 2000. 94. Muller, D. P. R., Harries, J. T., and Lloyd, J. K.: The relative importance of the factors involved in the absorption of vitamin E in children. Gut 15:966, 1974. 95. Ohnishi, A., Tsuji, S., Igisu, H., et al.: Beriberi neuropathy: morphometric study of sural nerve. J. Neurol. Sci. 45:177, 1980. 96. Ouahchi, K., Arita, M., Kayden, H., et al.: Ataxia with isolated vitamin E deficiency is caused by mutations in the a-tocopherol transfer protein. Nat. Genet. 9:141, 1995. 97. Pallis, C. A., and Lewis, P. D.: The Neurology of Gastrointestinal Disease. Philadelphia, W. B. Saunders, 1974. 98. Palmucci, L., Doriguzzi, C., Orsi, L., et al.: Neuropathy secondary to vitamin E deficiency in acquired intestinal malabsorption. Ital. J. Neurol. Sci. 9:599, 1988. 99. Parry, G. J., and Bredesen, D. E.: Sensory neuropathy with low dose pyridoxine. Neurology 35:1466, 1985. 100. Paulson, G. W., Martin, E. W., Mojzisik, C., and Carey, L. C.: Neurologic complications of gastric partitioning. Arch. Neurol. 42:675, 1985. 101. Peltier, G., Hermreck, A. S., Moffat, R. E., et al.: Complications following gastric bypass procedure for morbid obesity. Surgery 86:648, 1979. 102. Pennypacker, L. C., Allen, R. H., Kelly, J. P., et al.: High prevalence of cobalamin deficiency in elderly outpatients. J. Am. Geriatr. Soc. 40:1197, 1992. 103. Prodan, C. I., Holland, N. R., Wisdom, P. J., et al.: CNS demyelination associated with copper deficiency and hyperzincemia. Neurology 59:1453, 2002. 104. Roman, G. C.: An epidemic in Cuba of optic neuropathy, sensorineural deafness, peripheral sensory neuropathy and dorsolateral myeloneuropathy. J. Neurol. Sci. 127:11, 1994. 105. Rosenblum, J. L., Keating, J. P., Prensky, A. L., and Nelson, J. S.: A progressive neurologic syndrome in children with chronic liver disease. N. Engl. J. Med. 304:503, 1981. 106. Ruffin, J. M., and Smith, D. T.: Treatment of pellagra with special reference to the use of nicotinic acid. South. Med. J. 32:407, 1939.

2061

107. Russell, J. S. R., Batten, F. E., and Collier, J.: Subacute combined degeneration of the spinal cord. Brain 23:39, 1900. 108. Saperstein, D. S., Grogan, P. M., Milford, C., et al.: Myeloneuropathy in the setting of myelodysplastic syndrome. Neurology 62(Suppl. 5):A196, 2004. 109. Saperstein, D. S., Jackson, C. E., Nations, S. P., et al.: Uncommon neurologic manifestations of pernicious anemia with cobalamin deficiency. Neurology 56(Suppl. 3):A417, 2001. 110. Saperstein, D. S., Wolfe, G. I., Gronseth, G. S., et al.: Challenges in the identification of cobalamin-deficiency polyneuropathy. Arch. Neurol. 60:1296, 2003. 111. Saperstein, D. S., Wolfe, G. I., Longeria, O., et al.: Quantitative sensory testing in a large cohort of neuropathy patients. Muscle Nerve 24:1417, 2001. 112. Saperstein, D. S., Wolfe, G. I., Nations, S. P., et al.: Electrodiagnostic features of cobalamin deficiency polyneuropathy. Muscle Nerve 26:574, 2002. 113. Sassaris, M. , Meka, R., Miletello, G., et al.: Neuropsychiatric syndromes after gastric partition. Am. J. Gastroenterol. 78:321, 1983. 114. Satya-Murti, S., Howard, L., Krohel., G, and Wolf, B.: The spectrum of neurologic disorder from vitamin E deficiency. Neurology 36:917, 1986. 115. Savage, D. G., Lindenbaum, J., Stabler, S. P., and Allen, R. H.: Sensitivity of serum methylmalonic acid and total homocysteine determinations for diagnosing cobalamin and folate deficiencies. Am. J. Med. 96:239, 1994. 116. Schaumburg, H., Kaplan, J., Windsbank, A., et al.: Sensory neuropathy from pyridoxine abuse: a new megavitamin syndrome. N. Engl. J. Med. 309:445, 1983. 117. Schilling, R. F.: Is nitrous oxide a dangerous anesthetic for vitamin B12-deficient subjects? JAMA 255:1605, 1986. 118. Schleper, B., and Stuerenburg, H. J.: Copper deficiencyassociated myelopathy in a 46-year-old woman. J. Neurol. 248:705, 2001. 119. Schoffeniels, E.: Thiamine phosphorylated derivatives and bioelectrogenesis. Arch. Int. Physiol. Biochim. 91:233, 1983. 120. Sebrel, W. H., and Butler, R. E.: Riboflavin deficiency in man (ariboflavinosis). Public Health Rep. 54:2121, 1939. 121. Selikoff, I. J., Robitzek, E. H., and Ornstein, C. G.: Treatment of pulmonary tuberculosis with hydrazide derivatives of nicotinic acid. J. Am. Med. Assoc. 50:973, 1952. 122. Sesso, R. M., Iunes, Y., and Melo, A. C.: Myeloneuropathy following nitrous oxide anaesthesia in a patient with macrocytic anaemia. Neuroradiology 41:588, 1999. 123. Sigal, S. H., Hall, C. A., and Antel, J. P.: Plasma R binder deficiency and neurologic disease. N. Engl. J. Med. 317:1330, 1988. 124. Smith, J. C., Zeller, J. A., Brown, E. D., and Ong, S. C.: Elevated plasma zinc: a heritable anomaly. Science 191:496, 1976. 125. Sokol, R. J.: Vitamin E and neurologic deficits. Adv. Pediatr. 37:119, 1990. 126. Sokol, R. J., Heubi, J. E., Iannaccone, S. T., et al.: Vitamin E deficiency with normal serum vitamin E concentrations in children with chronic cholestasis. N. Engl. J. Med. 310:1209, 1984.

2062

Diseases of the Peripheral Nervous System

127. Sokol, R. J., Kayden, H. J., Bettis, D. B., et al.: Isolated vitamin E deficiency in the absence of fat malabsorption. Familial and sporadic cases: characterization and investigation of causes. J. Lab. Clin. Med. 111:548, 1988. 128. Solomons, N. W.: Biochemical, metabolic, and clinical role of copper in human nutrition. J. Am. Coll. Clin. Nutr. 4:83, 1985. 129. Somer, H., Bergstrom, L., Mustajoki, P., and Rovamo, L.: Morbid obesity, gastric application and a severe neurological deficit. Acta Med. Scand. 217:575, 1985. 130. Stabler, S. P.: Screening the older population for cobalamin (vitamin B12) deficiency. J. Am. Geriatr. Soc. 43:1290, 1995. 131. Strachan, H.: Malarial multiple peripheral neuritis. Ann. Universal Med. Sci. 1:139, 1888. 132. Strachan, H.: On a form of multiple neuritis prevalent in the West Indies. Practitioner 59:477, 1897. 133. Sung, J. H., and Stadlan, E. M.: Neuroaxonal dystrophy in congenital biliary atresia. J. Neuropathol. Exp. Neurol. 25:341, 1966. 134. Swain, R.: An update of vitamin B12 metabolism and deficiency states. J. Fam. Pract. 41:595, 1995. 135. Takahashi, K., and Nakamura, H.: Axonal degeneration in beriberi neuropathy. Arch. Neurol. 33:836, 1976. 136. Tappel, A. L.: Vitamin E as the biological lipid oxidant. Vitam. Horm. 20:493, 1962. 137. Toh, B. H., van Driel, I. R., and Gleeson, P. A.: Pernicious anemia. N. Engl. J. Med. 337:14418, 1997. 138. Tomoda, H., Shibasaki, H., Hirata, I., and Oda, K.: Central vs peripheral nerve conduction before and after treatment of subacute combined degeneration. Arch. Neurol. 45:526, 1988. 139. Traber, M. G., Sokol, R. J., Ringel, S. P., et al.: Lack of tocopherol in peripheral nerves of vitamin E-deficient patients with peripheral neuropathy. N. Engl. J. Med. 317:262, 1987.

140. Ungley, C. C.: Subacute combined degeneration of the cord: I. Response to liver extracts. II. Trials with vitamin B12. Brain 72:382, 1949. 141. Victor, M.: Polyneuropathy due to nutritional deficiency and alcoholism. In Dyck, P. J., Lambert, E. H., Thomas, P. K., and Bunge, R. (eds.): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 1899, 1983. 142. Victor, M., and Lear, A. A.: Subacute combined degeneration of spinal cord. Am. J. Med. 20:896, 1956. 143. Walton, J. N.: Brain’s Diseases of the Nervous System, 8th ed. New York, Oxford University Press, 1977. 144. Williams, R. D., Mason, H. L., Power, M. H., and Wilder, R. M.: Induced thiamine (vitamin B1) deficiency in man: relation of depletion of thiamine to development of biochemical defect of polyneuropathy. Arch. Intern. Med. 71:38, 1943. 145. Williams, R. R.: Structure of vitamin B1. Am. Chem. Soc. 58:1063, 1936. 146. Willison, H. J., Muller, D. P. R., Matthews, S., et al.: A study of the relationship between neurological function and serum vitamin E concentrations in patients with cystic fibrosis. J. Neurol. Neurosurg. Psychiatry 48:1097, 1985. 147. Wolfe, G. I., Baker, N. S., Amato, A. A., et al.: Cryptogenic sensory polyneuropathy: clinical and electrophysiologic characteristics. Arch. Neurol. 56:540, 1999. 148. Yokota, T., Wada, Y., Furukawa, T., et al.: Adult-onset spinocerebellar syndrome with idiopathic vitamin E deficiency. Ann. Neurol. 22:84, 1987. 149. Zegers, de Beyl, D., Delecluse, F., Verbanck, P., et al.: Somatosensory conduction in vitamin B12 deficiency. Electroencephalogr. Clin. Neurophysiol. 69:313, 1988.

90 Tropical Myeloneuropathies GUSTAVO C. ROMAN

Historical Aspects Epidemic Outbreaks of TMNs TMNs in Soldiers and POWs during World War II

Classification Tropical Ataxic Neuropathy Tropical Spastic Paraparesis Etiology Nutritional Deficiencies

Tropical myeloneuropathies (TMNs) are characterized by funicular lesions of the spinal cord accompanied by axonal, predominantly sensory, peripheral neuropathy that may be due to a number of possible etiologies. The high prevalence of TMNs in tropical and subtropical regions has been recognized for more than a century.87 However, reports from Africa, the Far East, and the West Indies remained unnoticed until World War II, when myeloneuropathies in prisoners of war (POWs) detained in tropical camps made neurologists aware of the relationship between tropical climate and the epidemic occurrence of TMNs. The occurrence in 1993–1994 of an epidemic of nutritional neuropathy in Cuba that affected more than 50,000 people, and constituted one of the worst neurologic epidemics of the last decade,81,84 indicates the pervasiveness and importance of these problems. TMNs constitute a public health problem in affected regions because of the number of disabled individuals and the absence of facilities for diagnosis, care, and rehabilitation.25 However, the true magnitude of the problem in underdeveloped countries has not been evaluated accurately because of deficiencies of statistical indicators and lack of appropriate neuroepidemiologic studies. These entities have been the subject of productive research, as demonstrated by the discovery of the role of human T-lymphotropic virus type 1 (HTLV-1), the first human retrovirus, in tropical spastic paraparesis (TSP)79; the postulated effect of glutamate-receptor excitatory amino acids in motor system disorders induced by Lathyrus sativus and Cycas circinalis102; and the growing problem of neurotoxic pesticides in tropical countries.24,25,56

Neurotoxins HTLV-1 Infection Summary

HISTORICAL ASPECTS Historical accounts of TMN in Jamaica78 offer a glimpse of the multiple potential causes of this problem in the tropics. Early in the history of Jamaica, epidemic outbreaks of strange palsies were described among settlers and slaves. In 1672, Ellwood31 wrote: “The diseases generally are here, at first coming, the flux, which turns to the bloody flux with violent fever, (. . .) and the gripes, which in a very short time takes away the use of the limbs.” According to Dunn,29 The most mysterious Caribbean malady of the seventeenth century was the dry bellyache, or dry gripes, perhaps induced by drinking rum processed in lead pipes. The victim suffered excruciating cramps in the pit of his stomach and bowels. He lost the use of his limbs temporarily or permanently and often died. This description of the “dry bellyache” of Jamaica is indeed suggestive of lead poisoning, abdominal colic being a prominent symptom of plumbism in adults, followed by motor neuropathy. Bruyn and Poser6 provided an erudite essay on the history of “burning feet” and other forms of peripheral neuropathy in the tropics. According to this source, J. Grierson,44 a British medical officer in the Indian Army, first described burning feet in 1826. Toward the turn of the 19th century, Strachan and Scott, British physicians stationed in Jamaica, as well as the Cuban physician Domingo Mádan, reported epidemic outbreaks of peripheral and optic neuropathy. A review of the best known epidemics of TMNs follows next. 2063

2064

Diseases of the Peripheral Nervous System

Epidemic Outbreaks of TMNs Strachan’s Multiple Neuritis of the West Indies (1888) In 1888 and 1897, Henry Strachan,108,109 Senior Medical Officer in Jamaica, reported 510 cases of “a form of multiple neuritis prevalent in the West Indies,” observed at the Kingston Public Hospital over a span of 5 years. The disease occurred almost exclusively among blacks, and it affected young adults of both sexes. Most had incomplete recovery, and deaths were rare. Clinical Features. Patients complained of numbness and cramps, but the most striking symptom was the severe burning pain in the palms of the hands and soles of the feet. “At night the patient will be awake for hours, rubbing his feet and legs . . . and moaning with pain.”108 All patients walked with a “typically ataxic” gait, and in some, “any attempts to move the upper extremities resulted in a peculiar aimless jerk.”108 The knee reflexes were absent in more than half (53%), decreased in 23%, and normal in the rest. Pain, touch, and temperature sensation, “although delayed, were completely absent only in the more severe cases.”108 Some had decreased sight and hearing. In chronic cases, muscle atrophy developed “until the ‘claw’ hand and foot were marked features.”108 Skin Pigmentation. The skin changes were quite noticeable: On examination . . . there is slight excoriation of the edges of the eyelids, margins of the lips, and around the margins of the nostrils; the palpebral conjunctiva may be hyperaemic, as will be the lips. The heat of the hands complained by the patient will be . . . appreciable to the touch, and due to a hyperaemic condition of the palms. Desquamation, which is fine and branny, involves the whole of the palm, which becomes more and more deeply pigmented as the disease progresses. . . . In the feet the thick epidermis of the soles of such patients as walk barefoot will come off in large flakes, and the soles also will become pigmented. The change is most striking to the observer in the regions which are normally fairer than the rest. The colour may vary from brown to intense black.108 Neuropathology. In a few fatal cases, postmortem examination showed pigmentation of the brain, spinal cord, and nerve trunks. However, more specific neuropathologic changes were not documented. Liver and spleen had evidence of chronic malaria.108 Etiology. Strachan considered that his syndrome was due to malaria and called it “malarial multiple peripheral neuritis.” He ruled out intoxication by cassava, alcohol, and lead. However, over the years, Strachan’s syndrome came

to be considered a nutritional deficiency neuropathy. Nevertheless, the clinical picture of burning feet, sensorimotor neuropathy, and skin pigmentation is typical of arsenic intoxication. ARSENIC INTOXICATION. The clinical features of arsenic poisoning were first described in a 1900 epidemic in England caused by beer contaminated with arsenic.50 The epidemic was masterfully described by Reynolds,75 who noticed . . . a considerable number of cases presenting with unusual skin lesions, mainly dermatosis and pigmentation, diagnosed in some cases as Addison’s disease. In June 1900, many cases of the typical erythromelalgia or painful red neuralgia of Weir Mitchel began to occur along with an increase in the number of cases of herpes zoster, which appeared to be in epidemic form . . . during this period I saw probably more cases than I had seen altogether in the past two years. At the beginning of August quite an extraordinary number of cases of so-called ‘alcoholic paralysis’ were admitted. There was at the beginning of November, no longer any doubt that a serious epidemic existed not only in Manchester and Salford but also in the neighbouring towns. The clinical picture developed in the following sequence: vomiting and diarrhea after the consumption of beer; “furred” or “silvery” tongue; running eyes and nose with puffy eyelids, bronchial irritation, and hoarseness; and numbness and tingling of hands and feet, followed shortly by the onset of extremely painful “burning feet” or acrodynia. Late findings were the distal muscle atrophies, keratosis with branny desquamation, and a severe hyperpigmentation resembling that of Addison’s disease, although the scars and buccal surfaces were not involved. Twenty years before Mees,57 the characteristic transverse lines of arsenic poisoning were described by Reynolds75: In many cases the nails are affected. After the patients have stopped taking the beer for some weeks the best appearances are seen, for then there is a transverse white ridge across the nail; proximal to this the nail is normal, but distal to it the nail is whiter, cracked, thin, and towards the tip almost papery and much flattened. Reynolds75 remarked on the unusual number of cases of herpes zoster (a finding noted as well by Strachan108,109) and, remembering this to be a common side effect of arsenical therapy, soon deduced that arsenic poisoning was the cause of the epidemic. The sugar used in the brewing of the beer in England was one of the sources of arsenic (it contained 0.5% by weight of arsenious acid). Román78 speculated that these two epidemics so closely related in time—one occurring in

Tropical Myeloneuropathies

1887–1897 in tropical Jamaica, where sugar was produced, and the other one in 1900 in England, where sugar from the West Indies was consumed—could be linked to a common source of arsenic to be found in the sugarcane plantations. Here, sodium arsenite, arsenic pentoxide, lead and calcium arsenate, and copper acetoarsenite (Paris green) were extensively used as weed killers and pesticides. Domingo Mádan’s Retrobulbar Optic Neuritis in Cuba (1898) From 1886 to 1898, during the Cuban insurrection against Spain, harsh measures were taken by the colonial rulers. To restrain the independence movement, General Valeriano Weyler ordered the forced internment of hundreds of thousands of Cubans in concentration camps. With the war of independence, economic collapse, widespread misery, and malnutrition prevailed everywhere. The explosion of the Maine on February 15, 1898, led to the U.S. intervention and on April 25, 1898, a U.S. naval blockade initiated the Cuban-Spanish-American War. During this time, the Cuban physician Domingo Mádan55 described numerous cases of retrobulbar optic neuropathy that he considered identical to tobacco-alcohol amblyopia, “although none of these patients drank.” López52 called this condition “blockade amblyopia” and considered that it was due to the worsening of the nutritional conditions of the patients. A century later, a similar epidemic would affect the island of Cuba during a U.S. embargo.67 Santiesteban-Freixas and colleagues93 have suggested the eponym Strachan-Mádan for the combination of peripheral and optic neuropathy occurring in the tropics. Scott’s Spanish Town Disease (1917) In 1917, Henry Harold Scott,94 a government bacteriologist in Jamaica, reported 21 patients with neuropathy observed during an outbreak in the neighborhood of Spanish Town (Santiago, or St. Iago de la Vega), the former capital of Jamaica. Patients were adult black laborers on a sugar plantation; males and females were equally affected. The epidemic started during the cutting of the cane crop, and ceased almost as soon as the crop was finished. . . . In practically every instance the first symptom complained of is a sensation of “itching in the eyes.” This comes on with comparative suddenness. In some cases both eyes are attacked at about the same time, in others one eye is affected at first. At this early stage the conjunctiva is congested and there is photophobia, but not of much intensity. Within the next three days or so the conjunctiva, both ocular and palpebral, is in a swollen, red, oedematous condition, the edges of the lids show abrasions, and small superficial ulcers form with discharge of pus.94 Clinical Features. Conjunctivitis and stomatitis cleared up rapidly with no other symptoms until 14 days later,

2065

when patients developed constipation, numbness, painless tingling and burning of the feet, dimness of vision, decreased hearing, and ataxic gait, “the legs being thrown about with wild, exaggerated movement.”94 Of 21 patients reported, 9 died, 8 had severe chronic disease, and the remainder recovered without neurologic deficit. No neurologic symptoms developed in several hundred patients affected by the epidemic conjunctivitis. Neuropathology. In two fatal cases, significant damage to the posterior columns was seen, from the cervical to the lumbosacral regions, with active demyelination and presence of fat-laden macrophages along with degeneration of posterior roots and dorsal root ganglia. The spinocerebellar tracts showed similar lesions, but the lateral corticospinal tracts were spared. Optic nerves were extensively affected. Minimal changes were present in the auditory, sciatic, median, and ulnar nerves. Etiology. Scott94 ruled out pellagra, because of the absence of dermatitis, diarrhea, mental changes, or the typical lesions of Betz cells. He also excluded beriberi because of the absence of edema, cardiac symptoms, and muscle tenderness and atrophy, as well as the minimal involvement of peripheral nerves at autopsy. He further discarded malnutrition, since the diet of most of his patients was adequate, consisting of yams (Dioscorea alata), breadfruit, cocoa, beans, cornmeal, and salted fish, “although during the crop most of them were living almost exclusively on cane.” He concluded that the most likely cause was toxic. Sugarcane is indeed cyanogenic,58 and arsenic also might be implicated because the neurologic picture in some aspects resembled that of arsenic poisoning, and dorsal column degeneration has been reported in arsenic neuropathy.32 Nevertheless, the Spanish Town epidemic appears to be clinically a postinfectious acute sensory neuronopathy following an epidemic outbreak of acute hemorrhagic conjunctivitis, a viral infection caused by Enterovirus 70, an enzootic picornavirus. Neurologic complications of acute hemorrhagic conjunctivitis have been reported,115 although acute sensory neuronopathy has not been described among them. Peraita’s Spanish Civil War Neuropathy (1939) The Spanish Civil War (1936–1939) produced widespread famine and saw the occurrence of neuropathies, myelopathies, and other nutritional neurologic disorders.39 In Madrid, Manuel Peraita (1908–1950) described hundreds of cases of retrobulbar neuropathy accompanied by agonizingly painful burning feet and hands.72 The feet, usually insensible and cold during the daytime, became burning hot and painful at night. The patients could not bear the slightest touch on the feet (anesthesia dolorosa). Associated symptoms included pellagrous psychoses, apathy, mental depression, insomnia, forgetfulness, and impaired vision and hearing.38,77

2066

Diseases of the Peripheral Nervous System

Cruickshank’s Jamaican (Tropical) Myeloneuropathy (1956) In 1956, Eric K. Cruickshank20 described in Jamaica 100 patients with “a neuropathic syndrome of uncertain origin . . . with features which do not conform to any recognized pattern.” Only a minority had ataxic neuropathy, sometimes with evidence of involvement of the optic and auditory nerves, and rarely a motor neuropathy with peroneal muscle atrophy. More than 80% presented with a predominantly spastic paraparesis with minimal sensory deficits (TSP). This report alerted the neurologic community to the existence of an endemic pyramidal syndrome in the tropics in the absence of overt nutritional deficits. It inaugurated the era of modern studies of TMNs that led, 30 years later, to the discovery of the retroviral myelopathy caused by HTLV-1.60,89

these cases were clearly different from the familial form of Leber’s optic neuropathy described in Cuba. A similar optic nerve and macular problem has been observed in young adults from Tanzania.74

Cuban Epidemic Neuropathy (1992–1993) The 1992–1993 epidemic neuropathy in Cuba was one of the most widespread in recent history. At least 50,862 patients on the island were diagnosed and treated during the epidemic.81,84

SENSORINEURAL DEAFNESS. During this epidemic in Cuba, patients also had high-pitched tinnitus and hearing loss. Pure tone audiometry demonstrated high-frequency (4 to 8 kHz) hearing loss, bilateral and usually symmetrical. There were no associated vestibular symptoms.

Epidemiology. The overall cumulative incidence rate was 462 per 100,000, with balanced rates for predominantly optic forms (242) and peripheral forms (219). There were few cases in children, adolescents, or the elderly. Most cases (87%) occurred in persons between the ages of 25 and 64 years. Optic forms predominated in men in the 45- to 64-year age group. Peripheral forms were more common in women ages 25 to 44 years. The geographic distribution showed a west-to-east pattern of decreasing incidence. Highest rates were found in the tobaccogrowing province of Pinar del Rio. Risk factors included cigar smoking, alcohol, irregular diet, weight loss and excessive sugar consumption. Clinical Features. Clinical manifestations included retrobulbar optic neuropathy, sensorineural deafness, predominantly sensory and autonomic neuropathy, and dorsolateral myelopathy. Less often, dysphonia, dysphagia, and spastic paraparesis were present. Mixed forms were frequent. Neurologic symptoms were preceded by weight loss, anorexia, chronic fatigue, lack of energy, irritability, sleep disturbances, and difficulties with concentration and memory. OPTIC NEUROPATHY. This was characterized by blurred vision, photophobia, central and cecocentral scotomata, loss of color vision for red and green, and loss of axons in the maculopapillary bundle, with temporal disc pallor in advanced cases. Approximately one third of the patients presented with concomitant skin and mucous membrane lesions and neuromyelopathic involvement, and 20% had hearing loss. Marked improvement of vision was obtained with parenteral B group vitamins and folic acid. Clinically,

DORSOLATERAL MYELOPATHY. Some patients presenting with similar sensory symptoms also had complaints of leg muscle weakness, difficulty in walking, increased urinary frequency, and impotence in males. These cases often had brisk knee reflexes with crossed adductor responses, contrasting with decreased ankle reflexes. Spasticity and Babinski’s signs were usually absent. Proximal motor weakness was present in one third of these cases. A significant decrease in position sense in the feet and a positive Romberg’s sign were often present. Severe cases had sensory ataxia. Vitamin B12 deficiency was probably responsible for most of these cases.

PERIPHERAL NEUROPATHY. Symptoms included painful dysesthesias of the soles and palms, mainly burning feet, numbness, cramps, and paresthesias. Nerves were sensitive to pressure. Motor involvement was minimal. Objective signs were often mild and included loss or decreased perception of vibration, light touch, and pinprick distally in the limbs in a glove-and-stocking pattern. Achilles tendon reflexes were decreased or absent. Motor nerve conduction velocities were normal and sensory nerve potential amplitudes were decreased only in severe cases. Some patients presented with a sensation of body heat, with excessive sweating, as well as coldness and hyperhidrosis of the hands and feet. Sural nerve biopsies in 34 patients with different forms of the disease showed an axonal neuropathy with predominant loss of myelinated large-caliber fibers and a less important loss of small-caliber fibers (Figs. 90–1 and 90–2). Etiology. A large number of toxic agents was investigated and ruled out during the outbreak. Deficit of B group vitamins, mainly B12 and thiamine, compounded by lack of essential sulfur-containing amino acids and carotenoids such as lycopene in the diet, appears to have been the cause of the outbreak. Cuba eliminated childhood malnutrition in the early 1970s, and nutritional supplementation programs for children, pregnant women, and the elderly were maintained despite enormous difficulties. This explains the absence of cases in these groups, so often affected by nutritional TMNs in other tropical areas. Optic Neuropathy in Tanzania In a 1997 study, Plant and colleagues74 called attention to an epidemic of subacute bilateral visual failure that has affected large numbers of teenage and young adult Africans

Tropical Myeloneuropathies

2067

FIGURE 90–1 Sural nerve biopsy from patient with Cuban epidemic neuropathy. Transverse section showing decrease of large- and small-diameter myelinated fibers. Note scarcity of largecaliber myelinated fibers. (Semithin section stained with p-phenylenediamine; ⫻200.) (From Borrajero, I., Pérez, J. L., Domínguez, C., et al.: Epidemic neuropathy in Cuba: morphological characterization of peripheral nerve lesions in sural nerve biopsies. J. Neurol. Sci. 127:68, 1994, with permission.)

in coastal Tanzania since 1988. Previous work had indicated that many patients had sensory symptoms, but the nature of the neurologic involvement was uncertain. The primary objective of this study was to characterize the accompanying neurologic disorder. Furthermore, the nature of the visual loss was uncertain from previous reports because both retinopathy and optic neuropathy had been suggested. The characteristic fundus showed symmetrical temporal optic atrophy and thinning of the cecocentral nerve fiber layer (maculopapillary bundle). Neurologic examination and nerve conduction measurements showed evidence of a predominantly large-fiber peripheral neuropathy in 47% of the patients; also, 42% developed hearing loss. Urinary thiocyanate levels were uniformly low. Serum was negative for antibodies to HTLV-1. DNA analysis was negative for

FIGURE 90–2 Morphometric analysis documenting fiber counts of 14% for largediameter myelinated fibers (from a normal of 40% for nerve fibers with a diameter ranging from 9 to 12 ␮m). There was also a reduction of small-caliber fibers (less than 6 ␮m) to 53% from a normal of about 60%. (From Borrajero, I., Perez, J. L., Dominguez, C., et al.: Epidemic neuropathy in Cuba: morphological characterization of peripheral nerve lesions in sural nerve biopsies. J. Neurol. Sci. 127:68, 1994, with permission.)

Leber’s hereditary optic atrophy. This entity occurs predominantly in the young, and does not correspond closely to other tropical neurologic syndromes previously described from East Africa; it is clinically very similar to Strachan’s syndrome and to the Cuban epidemic of optic and peripheral neuropathy. The etiology is unclear, but micronutrient deficiency is likely.

TMNs in Soldiers and POWs during World War II The problem of TMN and other neurologic syndromes observed in POWs in tropical and subtropical regions, mainly in the Far East, during World War II was extensively reviewed by Cruickshank,19 Spillane,104,105 and

2068

Diseases of the Peripheral Nervous System

Denny-Brown,26 among others. Untold thousands of people among refugees, civilian internees, and soldiers developed TMNs related to malnutrition. At a single camp in Singapore, Cruickshank noted at least 500 cases.19 When the incidence of painful feet was at its maximum, cases of frank spastic paraplegia developed, and 23% of those affected had brisk tendon reflexes, perhaps indicating early compromise of the pyramidal tract. In England, 36 years after imprisonment, more than 5% of 898 former POWs continued to suffer from peripheral neuropathy, optic atrophy, and deafness; 89 had developed neurologic symptoms many years after release and return to normal nutritional conditions, 35 had a spinal cord syndrome similar to TSP, and the prevalence of Parkinson’s disease was higher than expected.40,41 Neurologic Syndromes The cardinal symptoms were similar to those observed in TMN and consisted of peripheral edema and aching and stabbing pains in the feet (“burning feet,” “painful feet,” “sore feet,” “happy feet”). These were followed by tingling, pins-and-needles sensation, hypoesthesia or anesthesia to pinprick and light touch, and profound loss of position sense with sensory ataxia and positive Romberg’s sign. Tenderness of calf muscles was seen early, but footdrop and atrophy were rare. Reflexes progressed from hyperactivity to absence. Frequently there was visual loss, sometimes with pallor of the discs. In some camps, up to two thirds of the prisoners lost their vision.26,43,46,101 Other neurologic syndromes included deafness and vertigo (“camp dizziness”), acute ophthalmoplegic myasthenia, and spastic paraplegia. Neurologic manifestations were commonly accompanied by diarrhea, stomatitis, scrotal dermatitis, pellagroid erythema, glossitis, angular stomatitis, keratosis, tachycardia, and hypertension. Neuropathology In a neuropathology study of Canadian ex-POWs from Hong Kong, Fisher36 demonstrated symmetrical axonal loss and demyelination in posterior columns, mainly in the fasciculus gracilis. The primary lesion was considered to be in the dorsal root ganglia. Four of the 11 subjects showed lesions of the maculopapillary bundle with axonal loss from a lesion of bipolar ganglion cells at the macula. Etiology Although traditionally considered to be purely nutritional in origin, these syndromes, so prevalent among POWs in tropical and subtropical regions, were noted by Cruickshank21 to be rare in temperate zones, where nutritional deprivation was probably more severe. In addition, curiously, such syndromes were not observed in the severely undernourished Dutch population during the war famine of 1944–1945.

A role for a combination of tropical malabsorption and malnutrition (TM-M) in the production of these syndromes has been proposed by Román and co-workers,82,83,88 and is discussed next. In some instances, epidemic outbreaks of spastic paraparesis in prisoners in times of famine and war were clearly due to consumption of Lathyrus sativus. Tropical Malabsorption and Malnutrition Hundreds of soldiers were affected by tropical malabsorption (TM) during World War II. Elder30 described 400 cases of TM in British soldiers in Bengal, Assam, and Burma, usually beginning after a short stay in the tropics. The onset was sudden in 15% and more insidious in the rest. Of 400 soldiers, 393 developed painful glossitis, angular stomatitis, and cheilosis, and 22% had macrocytic anemia. Numbness and tingling of the feet were common. Walters118 reported 475 cases of TM in an Indian battalion of 815 men, predominantly vegetarians. Diarrhea, bacillary or amebic dysentery, or malaria precipitated neurologic symptoms. In the 42 worst cases, 90% had macrocytic anemia and 64% developed a peripheral neuropathy. The complications were less frequent in meat eaters than in vegans. Stefanini107 reported 1069 cases of TM in 12,500 Italian POWs in the Himalayan foothills in India. Deficiency signs, macrocytic anemia, glossitis, red tongue, aphthae, and hypochlorhydria were noted 2 to 4 weeks after the onset of diarrhea. Over 17% developed leg paresthesias, and 27% had muscle cramps. Therefore, it is likely that many of the neurologic complications in POWs in the Far East were due to TM-M. TM responds rapidly to tetracycline, and this probably explains why there were few cases in the Vietnam War, when early treatment of diarrhea with antibiotics was commonplace.51

CLASSIFICATION Dumas27,28 reviewed the problems in the classification and nosology of TMN. He emphasized that TMN is a homogeneous clinical syndrome that remains after exclusion of tropical causes of myelopathy and neuropathy, such as leprosy, bilharzial myelopathy, beriberi, and pellagra, as well as other causes of neuropathy and myelopathy that are not exclusively tropical in distribution, such as alcoholic neuropathy, diabetes, and other disorders of genetic, infectious, toxic, and inflammatory cause. Some of the possible etiologies of myelopathies and neuropathies in the tropics are summarized in Table 90–1. Román and colleagues88 suggested the clinical division of TMNs in two basic groups: tropical ataxic neuropathy (TAN), with prominent sensory ataxia caused by loss of proprioception; and TSP, a predominantly spastic paraplegia with minimal sensory deficit (Table 90–2). Intermediate forms with various degrees of ataxia and spasticity attributed to dorsolateral TMNs as well as forms with predominant

Tropical Myeloneuropathies

2069

Table 90–1. Possible Causes of Tropical Myeloneuropathies Type of Lesion Injury Fracture, cervical disc hemiation, spondylosis, hematomyelia, others Tumor Extra- or intradural; bony or epidural metastasis Malabsorption/Malnutrition Vascular Problems Thrombosis of anterior spinal artery, arteriovenous malformations, vasculitis-lupus Multiple Sclerosis

Infections

Bacterial Fungal Parasitic Neurotoxicity Lathyrism Cyanide Clioquinol (Entero-Vioform) Organophosphates and pesticides Other Problems Familial paraplegia, syringomyelia

Observations Usually acute onset. History of trauma. Radiology (⫹). Cervical spondylosis rare in tropics. Symptoms of hemisection or central spinal cord lesion with radiculopathy. There is myelopathy as a result of paraneoplastic effect of carcinoma. May produce TAN, TSP, optic and hearing defects. Probably caused neurologic syndromes of POWs in the Far East during World War II. Acute onset, transverse myelitis. Generally fluctuating course with relapses. Evidence of multiple neurologic lesions. The association of myelopathy and optic neuritis is called Devic’s disease. Motoneuronal lesion in diseases caused by polioviruses as well as by some echoviruses and arboviruses. Also in acute hemorrhagic conjunctivitis. Transverse myelitis from HSV-2 (genital); combined degeneration in patients with HIV infection, HTLV-1 antibodies in HAM/TSP, other chronic myelopathies and MS-like disease. Tuberculosis (Pott’s disease), epidural abscesses, arachnoiditis, syphilis, yaws, brucellosis, mycoplasma, leptospirosis, borrelia (Lyme disease). Rarely cause myelopathies. Schistosomiasis (S. haematobium, S. mansoni), filariasis (Dracunculus medinensis), eosinophilic meningitis (angiostrongyliasis), neurocysticercosis, hydatidosis. Common cause of spastic paraplegia in the Indian subcontinent. Excessive cassava consumption, common in Nigeria and Tanzania; may cause tropical ataxic neuropathy and paraparesis (Konzo). Caused subacute myelo-optico-neuropathy in Japan. Triorthocresylphosphate may cause tropical ataxic neuropathy.

HAM ⫽ HTLV-1–associated myelopathy; HIV ⫽ human immunodeficiency virus; HSV-2 ⫽ herpes simplex virus type 2; HTLV-1 ⫽ human T-lymphotropic virus type 1; MS ⫽ multiple sclerosis; POWs ⫽ prisoners of war; TAN ⫽ tropical ataxic neuropathy; TSP ⫽ tropical spastic paraparesis. Modified from Román, G. C.: Tropical spastic paraparesis and HTLV-1 myelitis. In McKendall, R. R. (ed.): Handbook of Clinical Neurology, Vol. 12. Viral Disease. Amsterdam, Elsevier, p. 525, 1989.

sensorimotor neuropathy are also observed. However, even in cases of predominant TAN, Hugon and associates47 have demonstrated spinal cord involvement of pyramidal tracts detected by central evoked motor potentials. Likewise, nerve biopsy in clinically pure forms of TAN revealed axonal nerve damage, indicating a continuum in the clinical spectrum. These TMNs are probably multifactorial in origin, but even in instances in which a single toxic agent is implicated—such as in cassava neurotoxicity—the clinical picture changes, resulting in TSP or TAN according to the age of the patient, nutritional status, amount of cassava consumed, and duration of exposure.

Tropical Ataxic Neuropathy TAN is an indolent and predominantly sensory neuropathy characterized by loss of proprioception and ataxia, usually

affecting large populations in the tropics. TAN was described originally among the Ijebu-speaking Yoruba in southwestern Nigeria.63,69,70 Clinical Features TAN is characterized by burning feet and severe loss of sensation of position, vibration, touch, and pressure, and, less frequently, pain and temperature. Sensory loss is more pronounced distally in the legs and is accompanied by loss of knee and ankle jerks, with preservation of strength. There may be lower motor neuron signs, with wasting, weakness, and atrophy of distal limb muscles, especially the peroneals. Vallat and co-workers114 studied 20 patients with TAN from Abidjan, Ivory Coast. Patients complained chiefly of burning feet. Strength was preserved, but pronounced distal sensory loss in the lower limbs, with absent knee and ankle reflexes, was present in all cases. Motor

2070

Diseases of the Peripheral Nervous System

Table 90–2. Clinical Features of Tropical Myeloneuropathies

Signs and Symptoms “Burning feet” and other paresthesias and dysesthesias involving legs Abnormal gait Leg weakness Loss of proprioception and vibration sense distally in legs Loss of pain and temperature sense in legs Muscle atrophy (peroneal) Loss of knee and ankle jerks Brisk knee and ankle jerks Babinski’s sign Sphincter disturbances Lumbar girdle pains Spasticity

Tropical Ataxic Neuropathy (TAN)

Tropical Spastic Paraparesis (TSP)

⫹⫹⫹⫹

⫹

⫹⫹⫹ ⫹ ⫹⫹⫹⫹

⫹⫹⫹⫹ ⫹⫹⫹⫹ ⫹

⫹

⫹

⫹ ⫹⫹⫹⫹ ⫺ ⫹ ⫹ ⫹ ⫹

⫹ ⫺ ⫹⫹⫹⫹ ⫹⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹⫹

Key: ⫹ Less than 25% of patients. ⫹⫹ Up to 50%. ⫹⫹⫹ Up to 75%. ⫹⫹⫹⫹ More than 75%. ⫺ Usually not present. From Román, G. C., Spencer, P. S., and Schoenberg. B. S.: Tropical myeloneuropathies: the hidden endemias. Neurology 35:1158, 1985, with permission.

nerve conduction velocities in the lower limbs usually showed severe slowing or conduction block. Neuropathology Sensory nerve biopsies with light and electron microscopy studies of nerve sections and teased fibers revealed severe axonal involvement affecting both unmyelinated and myelinated fibers, with no signs of regeneration.119 TAN is neurologically different from Strachan’s syndrome, epidemic neuropathy in Cuba, or konzo (see later). Epidemiology Oluwole and colleagues63,64,66 studied the prevalence of TAN in Ososa, a semiurban community in southwestern Nigeria. This was an area endemic for TAN in the original survey by Osuntokun in 1969. A two-stage door-to-door survey of all subjects older than 10 years (3428 subjects) was completed. The diagnostic criteria for TAN included any two or more of the following: (1) bilateral optic atrophy, (2) bilateral sensorineural deafness, (3) sensory gait ataxia, and (4) distal symmetrical sensory neuropathy. A total of 206 cases of TAN were found for an overall prevalence of 6.0% (males, 3.9%; females, 7.7%). The highest age- and sex-specific prevalence was 24% in women 60 to

69 years old. This interesting study confirmed the existence of TAN in Ososa, occurring with higher prevalences than those reported 30 years ago. To study the incidence of TAN, Oluwole and colleagues65 followed for 2 years a cohort of 3167 healthy subjects older than 10 years in Ososa, Nigeria. The incidence of ataxic neuropathy was 64 per 10,000 person-years (31 for males and 93 for females). Multivariate odds ratios were 0.78 (95% confidence interval [CI], 0.23 to 2.61) for intake of the most common cassava food, and 1.64 (0.56 to 5.09) for concentration of thiocyanate in plasma. The concentration of thiols was less than the reference limits in two controls, but in none of the cases. The latency of P100 was prolonged in 20 cases (69%) compared with 14 controls (42%) (P ⬍ .05). Pathogenesis TAN appears to result from a primary lesion of dorsal root ganglia (somatic sensory neuronopathy), with secondary degeneration of centrally directed axons in dorsal columns and of peripheral sensory axons. Etiology Most patients with TAN have clear evidence of malnutrition and skin and mucosal lesions suggestive of multivitamin deficits. In Africa, TAN is probably the most common form of TMN and usually results from nutritional deficits or unknown toxic injuries (Table 90–3). However, TAN has been traditionally associated with cyanide intoxication from cassava consumption. Recently, Oluwole and colleagues63,65 compared the prevalence of TAN in Ososa with that in Jobele, a community located outside the endemic area but previously shown to have high exposure to cyanide from cassava foods. Using identical methods, the prevalences of ataxic neuropathy were 490 per 10,000 in Ososa, and only

Table 90–3. Clinical Features of Tropical Ataxic Neuropathy in Nigeria and Senegal Signs and Symptoms “Burning feet” Visual deficit Deafness Leg weakness Mucocutaneous lesions Mental/psychosis Posterior column signs Sensory ataxia of gait Muscle atrophy Stocking-type sensory loss Total no. patients

Nigeria (%)69,70

Senegal (%)16

85 81 36 35 40 9 83 60 12 16 375

80 19 13 37 15 Rare 79 79 Rare Rare 75

From Román, G. C., Spencer. P. S. and Schoenberg. B. S.: Tropical myeloneuropathies: the hidden endemias. Neurology 35:1158, 1985, with permission.

2071

Tropical Myeloneuropathies

17 per 10,000 in Jobele, with an age-adjusted prevalence ratio of 4 per 10,000 (95% CI, 0 to 9). The mean intake of all cassava foods in Jobele was 7 meals/person/week (95% CI, 6 to 8), while the mean intake of all cassava foods in Ososa was slightly higher at 10 meals/person/week (95% CI, 9 to 11). The concentration of thiocyanate in the plasma was above the reference limit in 65% (95% CI, 57 to 73) in Jobele and 40% (95% CI, 27 to 52) in Ososa. The intake of protein was significantly lower in Ososa than in Jobele, but the concentrations of glutathione, cysteine, and gammaglutamylcysteine in the plasma were within the same range in Jobele and Ososa. The authors concluded that, although the exposure to cyanide was high in this community, the occurrence of ataxic neuropathy was low. These and previous findings do not suggest that cyanide is the cause of endemic ataxic neuropathy in Nigeria; its cause remains unknown.

Tropical Spastic Paraparesis Patients with TSP have primary involvement of the pyramidal tracts, initially and predominantly at the lumbar level, resulting in problems with gait and sphincter control. Epidemiology The frequency of pyramidal tract disorders varies in different regions: 21% of 75 cases of nutritional TMNs studied by Collomb and co-workers16 in Senegal; 26% of 61 similar patients reported by Giordano and associates42 in Ivory Coast; and 17% of 375 cases reported by Osuntokun in Nigeria.70 Etiology HTLV-1 is the cause of TSP. Cases have been described throughout the Caribbean, particularly in Jamaica and Martinique,23 as well as in Colombia, Brazil, the Seychelles islands, and some African countries. TSP appears to be identical to a chronic myelitis described in the southern islands of Japan (HTLV-1–associated myelopathy, or HAM).68,85 The World Health Organization has recommended the name HAM/TSP for this particular form of TSP. However, in most of Africa and South India the etiology of TSP appears to be different, and remains unknown. Clinical Features Román and Román86 have reviewed the clinical characteristics of TSP in different parts of the world. TSP appears to be a uniform syndrome worldwide, with minor regional variations.49 Onset of the disease is gradual in more than 60% of cases. In most patients, one leg is affected more than the other at onset. Progression usually is slow and indolent, sometimes reaching a plateau and remaining at that level for several years. Weakness of the legs, back pain, and leg numbness and stiffness as well as dysesthesias of the feet are experienced by almost all patients. Also common are bladder dysfunction manifested mainly by

increased urinary frequency and urgency, constipation, and penile impotence. Findings on neurologic examination of TSP patients are also quite consistent from place to place. The core of the syndrome is defined by the presence of spastic paraparesis or paraplegia; increased knee reflexes, crossed adductor responses, ankle clonus, and extensor plantar responses indicate the bilateral pyramidal tract lesion affecting the lower limbs. There are minimal signs of involvement of the posterior columns, but other sensory deficits, such as loss of pain sensation at the thoracic sensory level, are not commonly present. Spasticity is moderate, predominating on the thigh adductors and to a lesser degree on the thigh extensors and gastrocnemius muscles. Motor weakness is proximal and affects mainly the gluteus medius and iliopsoas muscles. Gait is typically slow and scissoring, with dragging and shuffling of the feet.80,85,87 The severe spasticity of lathyrism, with lurching gait on the ball of the foot and toes, is not seen in TSP. Most patients ambulate with minimal support, but one third are bedridden either as a result of rapid onset of the paraplegia or after long progression of the disease. Up to half the patients exhibit brisk reflexes in the upper limbs, and one sixth have snout and brisk mandibular reflexes. Higher mental functions are normal. Examination of the cranial nerves is usually normal, except in Jamaica, where retrobulbar neuropathy and deafness occurred in 15% and 7%, respectively, of TSP patients.76 Cerebellar signs, in particular intention tremor and dysmetria, have been observed in up to 20% of patients.87 In addition to paresthesias in the feet, mild decrease of vibratory perception distally in the feet is found, usually without loss of proprioception. Ankle reflexes are absent in about 25%, and stocking-type loss of sensation in a pattern suggestive of peripheral neuropathy has been described in a third of patients with TSP.86

ETIOLOGY Among the possible causes of TMN, the most common ones include nutritional deficiencies, usually resulting from TM-M, and neurotoxins, in particular cassava toxicity, lathyrism, and—with growing frequency—pesticide poisoning. Finally, as mentioned earlier, infection with HTLV-1, the first human retrovirus, has been demonstrated to be an important cause of TSP and of some cases of polymyositis, peripheral neuropathy, and motor neuron disease.

Nutritional Deficiencies Tropical Malabsorption-Malnutrition Nutritional deficiency has long been associated with a tropical neurologic syndrome characterized by orogenital dermatitis, painful sensory neuropathy, and amblyopia. Stannus,106 in particular, recognized this triad among undernourished populations in Africa. Skin lesions, such as

2072

Diseases of the Peripheral Nervous System

angular stomatitis, cheilosis, glossitis, geographic tongue, scrotal and vulvar dermatitis, eczema of the nostrils and palpebral fissures, and corneal neovascularization, are very often observed among malnourished populations in the tropics. They are the result of deficiency or malabsorption of riboflavin, combined in many cases with vitamin A deficiency, producing xerophthalmia and keratomalacia; niacin deficit, with frank pellagroid dermatosis of exposed parts of the body and a painful scarlet tongue; and megaloblastic or hypochromic anemia resulting from deficiencies of folate, vitamin B12, and iron. Among those suffering these severe nutritional deficiencies, the neurologic syndrome of TMN, including TAN, sensorimotor neuropathy, dorsolateral myelopathy, and TSP, often develops. Collomb and co-workers,16 in Dakar, Senegal, studied 47 patients with TAN and 28 with TSP. Visual deficits were found in 19% and auditory symptoms in 13% of these cases: 68% occurred in women over age 30 years, and in 33% pregnancy was a precipitating factor. Chronic diarrhea preceded the onset of symptoms in 29%. Gastric achlorhydria or hypochlorhydria was found in 84% of the patients; in 66% liver biopsy was mildly abnormal. With parenteral vitamin B complex treatment, 65% of the patients improved. The syndrome was attributed to a nutritional deficit resulting from a combination of poor dietary intake and malabsorption. Giordano and associates42 reported 61 cases of African TMNs studied in Abidjan, Ivory Coast. The most common forms were ataxic (59%), including 25 cases of ataxic polyneuropathy and 11 of pure spinal ataxia. Sixteen patients had TSP (associated in only one with HTLV-1); eight had sensorimotor polyneuropathies, and one had dorsolateral sclerosis. In 47% (25 of 53) amblyopia or optic atrophy was present, and 27% (11 of 40) had hypoacusis. Abnormal motor nerve conduction velocity of the lateral peroneal nerves was documented in 67 patients: 20 had marked slowing and 13 a complete conduction block. Most of these patients were relatively young (mean age, 35 years), from very low socioeconomic environments in sub-Saharan Africa; symptoms lasted at least 6 months in half of them and more than 1 year in a third. In 25% malnutrition was evident, with frequent skin and mucosal changes. Pregnancy was a precipitating factor in more than half the affected women, and gastrointestinal abnormalities—mainly diarrhea—were frequently found. Cerebrospinal fluid changes were not present in these patients, who were considered to be suffering from a nutritional deficit. As mentioned before, the recent epidemic of optic neuritis in Cuba was caused by nutritional deficiency. Tropical Malabsorption Most cases of TM result from chronic bacterial colonization of the small intestine with enterotoxin-producing coliform organisms, hence the name “postinfective tropical malab-

sorption,” coined by Cook.17 Changes of the type usually associated with malabsorption develop in the jejunal mucosa, most likely as a result of repeated viral, bacterial, and parasitic infections. This is the so-called tropical enteropathy commonly encountered in the poor. Most patients have few or no gastrointestinal complaints, and chronic diarrhea or steatorrhea is uncommon.17 However, tropical enteropathy does result in subclinical malabsorption with abnormal D-xylose tests, flat glucose tolerance curves, and vitamin B12 and folate malabsorption. Folate deficiency, resulting from poor diet and exacerbated by pregnancy, the puerperium, or systemic infections, helps perpetuate the intestinal lesion of TM.17,18 Drought, malaria, and bacterial diarrhea or other infections easily precipitate symptomatic TM. Overt cutaneous and systemic signs of protein-calorie malnutrition as well as deficiencies of minerals and vitamins A, C, D, and E and especially folic acid and B-group vitamins then appear. Neurologic signs occur relatively late, when the combination of several factors, including chronic TM, increased nutritional requirements (pregnancy, breast-feeding, infection), and dietary deficits, leads to a deficiency of essential nutrients severe enough to injure the nervous system. The most sensitive elements (dorsal root ganglia, bipolar retinal neurons) are the first to suffer the damage and manifest symptoms earliest. Vitamin Deficiencies Pallis and Lewis71 indicated the pointlessness of seeking to incriminate one particular factor as the cause of a myeloneuropathy in a condition characterized by widespread malabsorption, such as TM-M. Nevertheless, B-group vitamin deficiencies have long been thought to be the cause of TAN. Neurologic signs occur relatively late during malnutrition. Symptoms appear when a combination of factors finally leads to a deficiency of essential nutrients that is severe enough to injure the nervous system, or when protective nutrients—such as sulfur-containing amino acids and antioxidant carotenoids such as lycopene—become unavailable. The most sensitive elements (dorsal root ganglia, large myelinated distal axons, bipolar retinal neurons, cochlear neurons) are the first to suffer damage and manifest symptoms earliest. Beriberi. Beriberi has been considered a potential cause of TMNs because of the diet of polished rice among POWs, although thiamine (B1) produced no improvement of symptoms. Until early in the 20th century, beriberi was a major public health problem in the Orient, among populations depending on polished rice for their staple diet (China, Japan, Indonesia, the Philippines), but it also occurred in Africa and tropical South America. In the Japanese Navy, from 1878 until 1883, beriberi affected annually between 23% and 40% of the total force of about 5000 men, with an average mortality of 5%. Admiral

2073

Tropical Myeloneuropathies

Kanehiro Takaki practically eliminated beriberi in Japanese ships by lowering the carbohydrate content of the diet and increasing the amount of protein. Food supplementation with thiamine practically eliminated beriberi around the world. Most cases are currently observed in alcoholic patients, malnourished elderly subjects, and more rarely in pregnant women as a result of hyperemesis gravidarum. However, beriberi continues to occur in the tropics when the conditions of low thiamine intake, high carbohydrate intake, and high energy expenditure are met. In 1991 to 1993, the author studied an outbreak of epidemic neuropathy that occurred in a military garrison in Colombia. It was characterized clinically by edema of the feet, dysesthesias, frequent footdrop, and a high fatality rate. The diagnosis of wet beriberi caused by thiamine deficiency was confirmed by the presence of typical lesions in the heart in two fatal cases with myocardial edema, central necrosis of muscle fibers, and mitochondrial lesions. There was axonal neuropathy in sural nerve biopsies, low serum levels of thiamine, and deficiency of thiamine in the diet. The enormous energy expenditure of soldiers in training in the humidity of the tropic and the possible presence of fish thiaminase in the diet probably contributed to this outbreak. There is no effective body storage of thiamine, and dietary deficiency leads to symptoms in 1 or 2 months. Thiamine plays a central role in energy production, and thiamine requirements depend on the metabolic rate of the body. Beriberi responds rapidly to parenteral thiamine, but it carries a high mortality if the diagnosis is not suspected. Other Vitamin Deficiencies. Riboflavin deficiency was not a likely cause of TMNs among POWs in Cruickshank’s experience.21 He used nicotinic acid (niacin, vitamin B3) with good results in 68% of 500 cases,19 but pellagra or niacin deficiency alone is not a probable cause of this syndrome. Pantothenic acid was useful in treating some cases of burning feet,2 and experimental deficiency of pantothenic acid in pigs produces dorsal myelopathy and a sensory neuropathy.37,120 Cobalamin (vitamin B12) deficiency can also cause the clinical picture of TAN, and malabsorption of vitamin B12 and folic acid deficiency seems to be a common feature of TM. Folic acid deficiency per se has been associated with cases of myelopathy and sensory neuropathy in humans,3 and the close metabolic relationship between folate and cobalamin4,22 perhaps indicates a more important role of folic acid in TMNs. Vitamin E deficiency in chronic malabsorption causes a large fiber sensory neuropathy and spinocerebellar disorder with prominent loss of proprioception, ataxic gait, areflexia, weakness, ophthalmoplegia, and pigmentary retinal degeneration.5,45,61,100 Tranchant and colleagues113 have described vitamin E deficiency in eight patients with TMN. The role of vitamin deficiencies caused by TM-M in the causation of TAN (probably the largest group of

TMNs) remains largely unexplored despite recent advances in this field.

Neurotoxins Only the most common toxic causes of TMNs (i.e., cassava, lathyrism, and organophosphate insecticide poisoning) are discussed. A recent treatise on neurotoxicology103 addresses this topic in extenso. Cassava Toxicity (Cyanide Neurotoxicity) A number of staple foods in the tropics contain large amounts of cyanogenic glycosides. These plants include cassava (Manihot esculenta Crantz: yuca in Spanish, manioc in French), yam, sweet potato, corn, millet (Sorghum sp.), bamboo shoots, and beans such as Phaseolus vulgaris, particularly lima beans and the small black lima beans that grow wild in Puerto Rico and Central America. Tobacco smoke (Nicotiana tabacum) also contains considerable amounts of cyanide (150 to 300 ␮g per cigarette). Hydrolysis of plant glycosides releases cyanide as hydrocyanic acid. Intoxication occurs by rapid cyanide absorption through the gastrointestinal tract or the lungs. Detoxification is mainly to thiocyanate in a reaction mediated by a sulfur transferase (rhodanase), which converts thiosulfate into thiocyanate and sulfite. Thiocyanate is a goitrogenic agent that may be responsible for endemic cretinism in some tropical areas.33 The sulfurcontaining essential amino acids, cystine, cysteine, and methionine, provide the sulfur for these detoxification reactions. Also important is vitamin B12, with conversion of hydroxocobalamin to cyanocobalamin. In 1936, Clark10 first noted the high content of cyanide in bitter varieties of cassava and reported its toxic effects. Cassava (M. esculenta, M. utilissima) contains linamarin, a cyanogenic glycoside that, by the enzymatic action of linamarinase or on hydrolysis, forms cyanide. Extensive studies in Nigeria, Tanzania, and Mozambique have demonstrated the crucial role of chronic cyanide poisoning from cassava consumption in the production of TAN. The normal procedures for the preparation of cassava in different communities remove cyanide and reduce the toxicity. Cassava is resistant to drought; during famine the amount of cassava in the diet increases while insufficient time may be allowed for detoxification, leading to excessive intake of cyanide. Yams, beans, maize, sugarcane, and millet also contain cyanogenic alkaloids.58 Epidemiology. Cassava is consumed in large quantities throughout the tropics and provides a major source of calories for some 300 million people. In Africa, it constitutes the staple diet in Western Nigeria, Zaire, Tanzania, Uganda, and Mozambique. The geographic distribution of TAN in Africa tends to parallel the consumption of large amounts of cassava. TAN affects men and women equally, with peak incidence in the fifth and sixth decades. According to

2074

Diseases of the Peripheral Nervous System

Osuntokun,69,70 in some villages where the diet depends almost exclusively on cassava (“gari”), the prevalence of TAN is 18 to 26 per 1000. Cassava-related TSP also has been described in Mozambique.9,11–13 The prevalence ratio ranged from 29 to 34 per 1000. The disease predominated in children (65%) and in lactating women. Women and children ate raw and dried uncooked cassava almost exclusively, whereas men normally ate well-cooked cassava, which explains the lower disease incidence in men. The population was least affected in the coastal areas of this region, where fish is available. A similar epidemic was reported in a cassava-staple area of Zaire.121 Over 200 patients were reported during the dry season, most of them women and children. The clinical picture was characterized by spastic paraplegia, and there was minimal recovery with nutritional and vitamin therapy. Clinical Features. The clinical features of over 400 Nigerian patients with cassava-related TAN studied by Osuntokun69 included painful neuropathy, ataxia, and blurred vision. Common visual loss and bilateral optic atrophy (81%) as well as sensorineural hearing loss (36%) were present. Only 12% of the patients showed peroneal muscle atrophy, with decreased peroneal nerve motor conduction velocities and segmental demyelination in nerve biopsies.8 In Mozambique9,12 and Zaire,8,121 perhaps because of the high levels of cyanide intake, the clinical picture in most cases was characterized by involvement of the pyramidal tracts with severe TSP, usually of sudden onset or developing over a period of 1 week. Pain and paresthesias in the legs, headache, dizziness, and vomiting were early signs, in addition to fever. Signs of malnutrition, diminished visual acuity, and spastic dysarthria were common; hearing loss was rarely found. All patients had hyperreflexia and clonus in the legs, and in 50% brisk reflexes were also found in the arms. Spasticity and extensor plantar responses were present in all cases. Distal sensory loss in the legs occurred in 13%. Little improvement was seen at follow-up. Pathogenesis. The importance of chronic cyanide intoxication was reflected in high plasma concentrations of thiocyanate in Nigerians with TAN as compared with controls (11.4 versus 0.1 ␮mol/dL).70 Low levels were found for cysteine and other sulfur-containing amino acids involved in the detoxification pathways of cyanide from cassava, probably because of poor dietary intake and increased use of cassava. The exact mechanism of toxicity is, however, unknown. Although cobalamins are very important in cyanide detoxification, controlled trials of hydroxocobalamin and riboflavin therapy alone, or combined with cysteine, proved disappointing. In Mozambique, serum levels of thiocyanate were quite high (range, 29.8 to 33.6 ␮mol/dL, with normal controls in

nonsmokers ranging from 1.0 to 4.0 ␮mol/dL).9 During the drought, the estimated daily intake was 15 to 31.5 mg of hydrocyanic acid (HCN).9 Furthermore, children with the lowest levels of cysteine and other sulfur-containing amino acids developed paraplegia, probably as a result of impaired detoxification of HCN. Although the pathophysiology of cassava neurotoxicity is not fully understood, malnutrition, age of the patient, and dose of cyanide ingested are major factors in the production of the dominant element of the neurologic syndrome. In adults who ingest moderate amounts of cassava for prolonged periods, involvement of dorsal ganglia neurons is the most striking feature, resulting in the typical manifestations of TAN. In children as in pregnant and lactating women, and in association with almost exclusive dietary consumption of large amounts of bitter cassava during droughts, the brunt of the toxic effect is borne by the pyramidal tracts and is manifested clinically as TSP. Konzo. This is the traditional name given in Kwango, Zaire, to a form of epidemic spastic paraparesis described during times of drought and famine in cassava-staple areas. Epidemics have occurred in rural areas of Mozambique, Tanzania, Zaire, and the Central African Republic. The disease predominates in children (63%) and lactating women, with prevalences ranging from 29 to 34 per 1000. Women and children eat raw and sun-dried, uncooked bitter cassava, whereas men normally eat well-cooked cassava. Traditional methods of cassava preparation in Africa such as soaking, fermentation, and sun drying leave substantial amounts of cyanogenic glycosides in the cassava meal. Signs of acute cyanide intoxication are present, followed by a clinical picture characterized by sudden onset of nonprogressive spastic paraparesis, with flexion contractures of the hamstrings and Achilles tendons developing later. Scissoring gait and toe walking are frequently present, and patients require one or two sticks to walk. Increased tone in the lower limbs, hyperreflexia, ankle clonus, and Babinski’s signs are usually present. There is no sensory loss to light touch or pinprick. Impairment of rapid movements and hyperreflexia in the upper limbs are also present. Nystagmus and dysarthria may occur. There is minimal recovery with a nutritious diet and vitamin therapy. Despite major advances in the understanding of the natural history of konzo, the disease continues to occur in Africa. The most recent outbreak was reported by Bonmarin and colleagues,3 who investigated an outbreak of alleged poliomyelitis in Kahemba, a southwestern region of the Democratic Republic of Congo. The correct diagnosis was made and 237 cases were diagnosed mainly in children and women, especially after childbirth. Cassava processing remained unchanged and the authors postulated as cofactor a probable deficiency of micronutrients.

Tropical Myeloneuropathies

Cliff and co-workers12,13—who originally described one of the largest outbreaks to date in the province of Nampula, in northern Mozambique—recently revisited the region to survey the occurrence of subclinical upper motor neuron damage in schoolchildren and to test a new kit to measure the concentration of urinary thiocyanate. The mean total cyanogen concentrations in cassava flour varied with seasons and years of survey but were always high, ranging from 26 to 186 ppm. Highest levels were probably due to low rainfall in the 1997–1998 season. The new picrate kit for urinary thiocyanate worked well; mean concentrations in schoolchildren ranged from 225 to 384 ␮mol/L. The authors found 27 new cases of konzo and examined 397 schoolchildren for ankle clonus in five communities in the three districts affected by konzo. Urine analyses for urinary thiocyanate, linamarin, and inorganic sulfate were performed in a sample of 131 children. The proportion of children with clonus ranged from 4% to 22%. Geometric mean thiocyanate, linamarin, and inorganic sulfate concentrations were 163, 60, and 4.4 ␮mol/L, respectively. Children with ankle clonus had higher urinary thiocyanate concentrations. The authors recommend prevention to reduce cyanide exposure and further monitoring of cyanide exposure and neurologic damage in these communities. Konzo and subclinical upper motor neuron damage persist in poor rural communities in northern Mozambique, associated with high cyanogen concentrations in cassava flour and high urinary thiocyanate concentrations in schoolchildren.34 Konzo was associated with optic neuropathy, and a few patients had nystagmus.62 Although the etiopathogenesis of this optic neuropathy remains to be elucidated, the symmetry of the involvement suggests a toxic origin. Cyanide is believed to cause the neuro-ophthalmologic damage in konzo. However, the optic neuropathy in konzo patients does not resemble the features of the epidemic optic neuropathy in Tanzania, Cuba, or Nigeria or of Leber’s hereditary optic neuropathy, tobacco-alcohol amblyopia, or vitamin B deficiency. Lathyrism The classic cause of epidemic outbreaks of spastic paraparesis in the tropics is lathyrism caused by excessive consumption of peas of the Lathyrus family, especially L. sativus (chickling pea). Lathyrism is still endemic in regions of India, Bangladesh, and Ethiopia and continues to be a public health problem.53 Previous outbreaks were described in Europe, North Africa, and parts of Asia, mainly in times of war, famine, or drought.14,15,102 ␤-N-Oxalylamino-L-alanine, a neurotoxic amino acid, is considered the neurotoxin responsible for the spastic paraparesis of human lathyrism.102 Lathyrism begins after weeks or months of consumption of the peas and usually occurs in a setting of protein-calorie malnutrition.53 Onset is acute, subacute, or insidious. Cramps in the calf muscles

2075

are an early sign, and this is followed by increased muscle tone, spasticity, brisk reflexes, and Babinski’s sign. Sensory loss is typically absent, although by clinical neurophysiology, slight damage to the dorsal columns is also present. The gait is characterized by slight flexion of hips and knees, retraction of Achilles tendons, adductor crossing, and feet in plantar flexion; paraplegia in flexion is the eventual outcome. The most obvious neuropathologic lesion is fiber loss in the pyramidal tracts, mainly in the lumbar cord, along with pallor of the fasciculi gracilis and axonal swelling in Goll’s nuclei.110 Pesticides The widespread use of toxic pesticides in the developing world has become a major medical problem,24,112 resulting in a large number of casualties and increasing reports of neurotoxic side effects. Lack of appropriate precautions in handling these products, suicide attempts, and accidental exposure are common causes of morbidity and mortality. Not infrequently, however, pesticides banned in industrialized countries are distributed freely in the tropics. Largescale epidemic outbreaks of pesticide poisoning have resulted from contamination of grain or foodstuffs.48 For example, thousands of cases with neurologic manifestations were reported in Turkey from grain contamination with hexachlorobenzene.73 The Bhopal disaster in India7 vividly illustrates the dangers involved in the industrial production of these agents. Currently, organophosphate insecticide poisoning is one of the most frequent causes of neuropathy in the tropics. Organophosphate Insecticides. Triorthocresylphosphate (TOCP) has been known to cause neuromyelopathy since the 1930s, when an epidemic outbreak in the United States affected at least 50,000 people who drank adulterated Jamaican ginger extract (jake leg paralysis).59 Extensive outbreaks have also been reported in Bombay117 and Durban,111 usually as a result of contamination of cooking oil. The mechanism is often the use of discarded mineral oil drums for transport or storage of cooking oil. An outbreak was described in Sri Lanka by Senanayake95,96 in adolescent girls and in women soon after childbirth. In this case the source of toxicity was gingili oil contaminated with TOCP. Gingili oil was ingested as part of the ritual diet of menarche and the postpartum period among the Tamils. The cardinal features of the syndrome include pain and weakness of hands and feet, with bilateral footdrop, claw hands, and wristdrop. Marked distal axonopathy was documented by clinical neurophysiology studies. Pyramidal tract signs were also present in 40% of the patients. Sensory signs were minimal. In contrast with Jamaican jake leg paralysis, remarkable improvement was seen 3 years after the intoxication. Senanayake and Johnson97 have reported 10 cases of delayed polyneuropathy resulting from exposure to

2076

Diseases of the Peripheral Nervous System

methamidophos (Tamaron), another organophosphate insecticide. About 2 weeks after the acute cholinergic symptoms, these patients experienced delayed neuropathy with calf pains, pins-and-needles sensation in the feet, weakness of the feet and hands, difficulty walking, muscle weakness and atrophy, and brisk reflexes; sensory examination was normal. Findings were consistent with a predominantly motor neuropathy with partial denervation of affected muscles and relatively normal motor nerve conduction velocities. There was clinical evidence of involvement of the pyramidal tract. An “intermediate” syndrome of organophosphate insecticide poisoning has been described by Senanayake and Karalliedde.98,99 In contrast with the acute cholinergic effects of insecticide poisoning and the delayed neurotoxicity appearing 2 to 3 weeks after the poisoning, these patients developed, shortly after the acute phase, paralysis of proximal limb muscles, neck flexors, motor cranial nerves, and respiratory muscles, lasting up to 18 days. Electromyography showed a postsynaptic defect. Among the insecticides involved were fenthion, monocrotophos, dimethoate, and methamidophos. The intermediate syndrome is a recognized complication in weaning patients from respiratory support after organophosphate poisoning.91

HTLV-1 Infection Although most cases of HTLV-1 infection of the nervous system have been associated with TSP, there are welldocumented instances of peripheral nerve involvement producing an inflammatory neuropathy.92,116 There are also cases of polymyositis and anterior horn cell damage in a pattern resembling amyotrophic lateral sclerosis (pseudoALS forms).90 Neuropathy in HAM/TSP Despite the minimal sensory deficit demonstrated clinically, close to 70% of patients with TSP have abnormalities of the somatosensory evoked potentials, especially after stimulation of the posterior tibial nerves, in a pattern that suggests bilateral involvement of the fasciculus gracilis.1 Involvement of the peripheral nerves was documented in one third of 19 patients with TSP in the Seychelles.54 Abnormalities consisted of minimal increases in distal latencies with moderate slowing of proximal (F waves) and distal conduction velocities along the peroneal and sural nerves. These findings are suggestive of patchy demyelination and remyelination along the nerves, probably induced by a process such as vasculitis involving the vasa nervorum, with minimal axonal loss. Also, a few patients with TSP present with muscle atrophy and electromyographic changes indicative of anterior horn cell involvement.

Pseudo-ALS The disease begins slowly and evolves in a chronic pattern for a period of years. Initial symptoms are those of a spastic paraparesis with signs of an upper motor neuron lesion affecting the lower extremities, producing difficulty in walking, leg weakness, and back pain. Bladder problems, including increased urinary frequency and incontinence, and sensory symptoms are present in a third of the patients. Abnormal sensations include numbness and burning of the feet and a bandlike feeling of tightness about the waist. The foregoing symptoms are practically constant in HAM/TSP. Weight loss and wasting of the muscles were reported in all patients. At the time of diagnosis, most patients complained of weakness of the upper extremities. The latter is an uncommon complaint in patients with the classic form of HAM/TSP. Most patients exhibit overall brisk reflexes, usually with bilateral Babinski’s signs. Marked muscle wasting and fasciculations are constant, usually involving small hand muscles. Atrophy of scapular and shoulder muscles and of the quadriceps has been rarely reported. Prominent atrophy and fasciculations of the tongue and laryngeal paralysis with dysphonia have been described,90 but dysphagia and dysarthria are uncommon. Muscle biopsies have shown small angulated fibers and group atrophy, indicative of lower motor neuron disease. In a black patient from the United States,35 the biopsy showed a combination of denervated angular fibers with type I fiber grouping, consistent with chronic denervation. In addition, this patient showed a typical inflammatory myopathy. No autopsy reports of the pseudo-ALS form of HAM/TSP are available. However, earlier descriptions of neuropathologic changes in the Jamaican form of TSP indicated the presence of inflammation surrounding anterior horn cells, and neuronophagia as well as degeneration and loss of lower motor neurons. Neuropathologic studies in Japan of patients with HAM/TSP have shown central chromatolysis of anterior horn cells and mononuclear cell infiltration.

SUMMARY TMNs occur in geographic isolates in developing countries with high prevalence. The two main clinical forms are a sensory ataxia (TAN) and a predominantly spastic paraplegia (TSP). However, there is overlap in their clinical manifestations and in other clinical features, such as optic neuropathy, sensorineural hearing loss, and sensorimotor peripheral neuropathy. The multifactorial etiology of most cases of TMNs is obscure, but TM-M seems to be common. Vegetarian diets, famine, or increased nutritional requirements (such as pregnancy, lactation, and infection) not matched by appropriate diet may be contributing or

Tropical Myeloneuropathies

precipitating factors. TM-M was also a probable cause of the neurologic syndromes observed in Far East POWs in World War II. In addition, neurotoxic agents are important. Lathyrus and cyanogenic plants (mainly cassava and perhaps millet) may be responsible for outbreaks of TAN and TSP that affect large populations in tropical regions of Africa and India. Organophosphate pesticide poisoning is rapidly becoming a common form of neuropathy in the developing world. However, clusters of TSP in India, Africa, the Seychelles, Jamaica, and Colombia have been linked to HTLV-1. This infection has been shown to cause inflammatory neuropathy and pseudo-ALS lesions. As demonstrated by the large outbreak of optic and peripheral neuropathy in Cuba, TMNs are serious public health problems for the affected nations but also offer unique opportunities for international cooperation as well as for clinical and neuroepidemiologic investigations.

13.

14.

15.

16.

17. 18.

REFERENCES 19. 1. Arimura, K., Rosales, R., Osame, M., and Igata, A.: Clinical electrophysiologic studies of HTLV-l associated myelopathy. Arch. Neurol. 44:609, 1987. 2. Bibile, S. W., Lionel, D. W., Rohini, D., and Gnana, P.: Pantothenol and the burning feet syndrome. Br. J. Nutr. 11:434, 1957. 3. Bonmarin, I., Nunga, M., and Perea, W. A.: Konzo outbreak, in the south-west of the Democratic Republic of Congo, 1996. J. Trop. Pediatr. 48:234, 2002. 4. Botez, M. I., and Reynolds, E. H. (eds.): Folic Acid in Neurology, Psychiatry, and Internal Medicine. New York, Raven Press, 1979. 5. Brin, M. F., Fetell, M. R., and Green, P. H. A.: Blind loop syndrome, vitamin E malabsorption, and spinocerebellar degeneration. Neurology 35:338, 1985. 6. Bruyn, G. W., and Poser, C. M.: The History of Tropical Neurology: Nutritional Disorders. Canton, MA, Science History Publications, p. 144, 2003. 7. Bucher, J. R.: Methyl isocyanate: a review of health effects research since Bhopal. Fundam. Appl. Toxicol. 9:367, 1987. 8. Carton, H., Kayembe, K., Kabeya, O., et al.: Epidemic spastic paraparesis in Bandundu (Zaire). J. Neurol. Neurosurg. Psychiatry 49:620, 1986. 9. Casadei, E., Jansen, P., Rodrigues, A., et al.: Mantakassa: an epidemic of spastic paraparesis associated with chronic cyanide intoxication in a cassava staple area of Mozambique. 2. Nutritional factors and hydrocyanic acid content of cassava products. Bull. World Health Organ. 62:485, 1972. 10. Clark, A.: Report on effects of certain poisons in food plants of West Africa upon health of native races. J. Trop. Med. Hyg. 39:269, 1936. 11. Cliff, J., Lundqvist, P., Martensson, J., et al.: Association of high cyanide and low sulphur intake in cassava-induced spastic paraparesis. Lancet 2:1211, 1985. 12. Cliff, J., Martelli, A., Molin, A., and Rosling, H. (Ministry of Health, Mozambique): Mantakassa: an epidemic of spastic

20. 21.

22.

23.

24. 25.

26.

27. 28.

29.

30. 31.

2077

paraparesis associated with chronic cyanide intoxication in a cassava staple area of Mozambique. I. Epidemiology and clinical laboratory findings in patients. Bull. World Health Organ. 62:477, 1972. Cliff, J., Nicala, D., Saute, F., et al.: Ankle clonus and thiocyanate, linamarin, and inorganic sulphate excretion in school children in communities with Konzo, Mozambique. J. Trop. Pediatr. 45:139, 1999. Cohn, D. F., and Streifler, M.: Human neurolathyrism, a follow-up study. Part I: Clinical investigation. Arch. Suiss Neurol. Neurochir. Psychiatr. 128:151, 1981. Cohn, D. F., and Streifier, M.: Human neurolathyrism, a follow-up study. Part II: Special investigations. Arch. Suiss Neurol. Neurochir. Psychiatr. 128:157, 1981. Collomb, H., Quere, M. A., Cros, J., and Giordano, C.: Les neuropathies dites nutritionelles au Sénégal. J. Neurol. Sci. 5:159, 1967. Cook, G. C.: Tropical Gastroenterology. Oxford, UK, Oxford University Press, 1980. Cook, G. C., Morgan, J. O., and Hoffbrand, A. V.: Impairment of folate absorption by systemic bacterial infections. Lancet 2:1416, 1947. Cruickshank, E. K.: Painful feet in prisoners-of-war in the Far East: review of 500 cases. Lancet 2:369, 1946. Cruickshank, E. K.: A neuropathic syndrome of uncertain origin. West Indian Med. J. 5:147, 1956. Cruickshank, E. K.: Effects of malnutrition on the central nervous system and the nerves. In Vinken, P. J., and Bruyn, G. W. (eds.): Handbook of Clinical Neurology, Vol. 28. Amsterdam, Elsevier, 1976. Dastur, D. K., Santhadevi, N., and Quadros, E. V.: The B-vitamins in malnutrition with alcoholism: a model of intervitamin relationships. Br. J. Nutr. 36:143, 1976. d’Auriol, L., Vernant, J-C., Ouka, M., et al.: Diagnosis of HTLV-I infected seronegative neurological patients by polymerase chain reaction amplification in Martinique. Nouv. Rev. Fr. Hematol. 32:133, 1990. Davies, J. E.: Changing profile of pesticide poisoning. N. Engl. J. Med. 316:807, 1987. del Brutto, O., Carod, F. J., Dumas, M., et al.: Continuum: Tropical Neurology. St. Paul, MN, American Academy of Neurology, 2002. Denny-Brown, D.: Neurological conditions resulting from prolonged and severe dietary restriction (case reports in prisoners-of-war, and general review). Medicine (Baltimore) 26:41, 1947. Dumas, M.: Neuro-myélopathies tropicales. Riv. Neurobiol. 29:155, 1983. Dumas, M., Giordano, C., and Ndiaye, I. P.: African tropical myeloneuropathies. In Chopra, J. S., Jagannathan, K., and Sawhney, I. M. S. (eds.): Advances in Neurology. Amsterdam, Elsevier, p. 343, 1990. Dunn, R. S.: Sugar and Slaves: The Rise of the Planter Class in the English West Indies, 1624–1713. New York, Norton, p. 306, 1972. Elder, H. H. A.: Clinical features, diagnosis, and treatment of sprue. J. Trop. Med. Hyg. 2:1157, 1966. Ellwood, G.: Royal Society Manuscripts, Jamaica, 15 June 1672. Cited by Pawson, M., and Buisseret, D.: Port Royal, Jamaica. Oxford, UK, Clarendon Press, p. 101, 1975.

2078

Diseases of the Peripheral Nervous System

32. Erlicki, A., and Rybalkin, J.: Ueber Arseniklahmung. Arch. Psychiatr. Nervenkr. (Berl.) 23:861, 1892. 33. Ermans, A. M., Mbulamoko, N. M., Delange, F., and Ahluwalia, R.: Role of Cassava in the Etiology of Endemic Goitre and Cretinism. Ottawa, International Development of Research Centre, IDRC-136e, 1980. 34. Ernesto, M., Cardoso, A. P., Nicala, D., et al.: Persistent konzo and cyanogen toxicity from cassava in northern Mozambique. Acta Trop. 82:357, 2002. 35. Evans, B. K., Gore, I., Harrell, L. B., et al.: HTLV-I–associated myelopathy and polymyositis in a US native. Neurology 39:1572, 1989. 36. Fisher, C. M.: Residual neuropathological changes in Canadians held prisoners of war by the Japanese (Strachan’s disease). Can. Serv. Med. J. 11:157, 1955. 37. Follis, K. H. Jr., and Wintrobe, M. M.: A comparison of the effects of pyridoxine and pantothenic acid deficiencies on the nervous tissue of swine. J. Exp. Med. 81:539, 1945. 38. García Jiménez, F., and Grande Covián, F.: Sobre los transtornos carenciales observados en Madrid durante Ia guerra. Rev. Clin. Esp. 4:1, 1940. 39. García-Albea Ristol, E.: Deficiency neuropathies in Madrid during the Civil War period. Neurologia 14:122, 1999. 40. Gibberd, F. B., and Simmonds, J. P.: Neurological diseases in ex–Far-East prisoners-of-war. Lancet 2:135, 1980. 41. Gill, G. V., and Bell, D. R.: Persisting nutritional neuropathy amongst former war prisoners. J. Neurol. Neurosurg. Psychiatry 45:861, 1982. 42. Giordano, C., Dumas, M. Hugon, J., et al.: Neuromyélopathies tropicales Africaines: 61 cas étudiés en Côte d’Ivoire. Rev. Neurol. (Paris) 144:578, 1988. 43. Goldsmith, H.: Ocular signs in POWs from the Far East. Br. Med. J. 1:407, 1946. 44. Grierson, J.: On the burning in the feet of natives. Trans. Med. Phys. Soc. Calcutta 2:275, 1826. 45. Harding, A. E., Muller, D. P. R., Thomas, P. K., and Willison, H. J.: Spinocerebellar degeneration secondary to chronic intestinal malabsorption: a vitamin E deficiency syndrome. Ann. Neurol. 12:419, 1982. 46. Hobbs, H. E., and Forbes, F. A.: Visual defects in prisonersof-war from the Far East. Lancet 2:149, 1946. 47. Hugon, J., Giordano, C., Dumas, M., et al.: Evoked motor potentials in patients with tropical myeloneuropathies. Neurology 37(Suppl. 1):363, 1987. 48. Karalliedde, L., and Senanayake, N.: Acute organophosphorous insecticide poisoning in Sri Lanka. Forensic Sci. Int. 36:97, 1988. 49. Kazadi, K., Garin, B., Goussard, B., et al.: Non-spastic paraparesis associated with HTLV-1. Lancet 2:260, 1990. 50. Kelynack, T. N., and Kirby, W.: Arsenical poisoning from beer-drinking. Lancet 2:1600, 1900. 51. Klipstein, F. A.: Tropical sprue. In Sleisinger, M. H., and Fordtran, J. S. (eds.): Gastrointestinal Disease: Pathophysiology, Diagnosis, Management, 4th ed. Philadelphia, W. B. Saunders, p. 1281, 1989. 52. López, E.: Ambliopía por desnutrición o ambliopía del bloqueo. Arch. Policlín. 8:85, 1900.

53. Ludolph, A. C., Hugon, J., Dwivedi, M. P., et al.: Studies on the aetiology and pathogenesis of motor neuron diseases. 1. Lathyrism: clinical findings in established cases. Brain 110:149, 1987. 54. Ludolph, A. C., Hugon, J., Román, G. C., et al.: A clinical neurophysiologic study of tropical spastic paraparesis. Muscle Nerve 11:392, 1988. 55. Mádan, D.: Notas sobre una forma sensitiva de neuritis periférica: ambliopía por neuritis óptica retrobulbar. Crónica Med. Quirúrg. Habana 24:81, 1898. 56. Maghazaji, H. I.: Psychiatric aspects of methylmercury poisoning. J. Neurol. Neurosurg. Psychiatry 37:954, 1974. 57. Mees, R. A.: Een verschijnsel bij polyneuritis arsenicosa. Ned. Tijdschr. Geneeskd. 1:391, 1919. 58. Montgomery, R. D.: Cyanogens. In Liener, I. E. (ed.): Toxic Constituents of Human Foodstuffs. New York, Academic Press, p. 143, 1969. 59. Morgan, J. P., and Penovich, P.: Jamaica ginger paralysis. Arch. Neurol. 35:530, 1978. 60. Morgan, O. S., Montgomery, R. D., and Rodgers-Johnson, P.: The myeloneuropathies of Jamaica: an unfolding story. Q. J. Med. 67:273, 1988. 61. Muller, D. P. R., Lloyd, J. K., and Wolff, O. H.: Vitamin E and neurological function. Lancet 1:225, 1983. 62. Mwanza, J. C., Tshala-Katumbay, D., Kayembe, D. L., et al.: Neuro-ophthalmologic findings in konzo, an upper motor neuron disorder in Africa. Eur. J. Ophthalmol. 13:383, 2003. 63. Oluwole, O. S.: Endemic ataxic polyneuropathy in Nigeria. Stockholm, Karolinska Institutet, 2002. 64. Oluwole, O. S., Onabolu, A. O., Cotgreave, I. A., et al.: Low prevalence of ataxic polyneuropathy in a community with high exposure to cyanide from cassava foods. J. Neurol. 249:1034, 2002. 65. Oluwole, O. S., Onabolu, A. O., Cotgreave, I. A., et al.: Incidence of endemic ataxic polyneuropathy and its relation to exposure to cyanide in a Nigerian community. J. Neurol. Neurosurg. Psychiatry 74:1417, 2003. 66. Oluwole, O. S., Onabolu. A. O., Link, H., and Rosling, H.: Persistence of tropical ataxic neuropathy in a Nigerian community. J. Neurol. Neurosurg. Psychiatry 69:96, 2000. 67. Ordoñez-Garcia, P. O., Nieto, F. J., Espinosa-Brito, A. D., and Caballero, B.: Cuban epidemic neuropathy, 1991 to 1994: history repeats itself a century after the “amblyopia of the blockade.” Am. J. Public Health 86:738, 1996. 68. Osame, M., Usuku, K., Izumo, S., et al.: HTLV-I–associated myeloneuropathy, a new clinical entity. Lancet 1:1031, 1986. 69. Osuntokun, B. O.: An ataxic neuropathy in Nigeria: a clinical, biochemical and electrophysiological study. Brain 91:215, 1968. 70. Osuntokun, B. O.: Cassava, diet, chronic cyanide intoxication and neuropathy in Nigerian Africans. World Rev. Nutr. Diet. 36:141, 1981. 71. Pallis, C. A., and Lewis, P. D.: The Neurology of Gastrointestinal Disease. London, W. B. Saunders, 1974. 72. Peraita, M.: Deficiency neuropathies observed in the Civil War (1936–9). Br. Med. J. 2:784, 1946. 73. Peters, H. A., Goomen, A., Cripps, D. J., et al.: Epidemiology of hexachlorobenzene-induced porphyria in Turkey: clinical and laboratory follow-up after 25 years. Arch. Neurol. 39:744, 1982.

Tropical Myeloneuropathies 74. Plant, G. T., Mtanda, A. T., Arden, G. B., and Johnson, G. J.: An epidemic of optic neuropathy in Tanzania: characterization of the visual disorder and associated peripheral neuropathy. J. Neurol. Sci. 145:127, 1997. 75. Reynolds, B. S.: An account of the epidemic outbreak of arsenical poisoning occurring in beer-drinkers in the north of England and the midland counties in 1900. Lancet 1:166, 1901. 76. Rodgers, P. E. B.: The clinical features and aetiology of the neuropathic syndrome in Jamaica. West Indian Med. J. 14:36, 1965. 77. Román, G.: Deficiency neuropathies during the Civil War period. Neurologia 15:141, 2000. 78. Román, G. C.: Epidemic neuropathies of Jamaica. Trans. Stud. Coll. Physicians Phila. 7:261, 1985. 79. Román, G. C.: Retrovirus-associated myelopathies. Arch. Neurol. 44:659, 1987. 80. Román, G. C.: Tropical spastic paraparesis and HTLV-I myelitis. In McKendall, R. R. (ed.): Handbook of Clinical Neurology, Vol. 12. Viral Disease. Amsterdam, Elsevier, p. 525, 1989. 81. Román, G. C.: An epidemic in Cuba of optic neuropathy, sensorineural deafness, peripheral neuropathy and dorsolateral myelopathy. J. Neurol. Sci. 127:11, 1994. 82. Román, G. C.: Tropical neuropathies. Baillieres Clin. Neurol. 4:469, 1995. 83. Román, G. C.: Tropical neuropathies: an update. Curr. Opin. Neurol. 11:539, 1998. 84. Román, G. C.: Mission to Cuba: The True Story of the Epidemic That Changed Cuba’s Bearings. Barcelona, Prous Science, 2000. 85. Román, G. C., and Osame, M.: Identity of HTLV-1–associated tropical spastic paraparesis and HTLV-I–associated myelopathy. Lancet 1:651, 1988. 86. Román, G. C., and Román, L. N.: Tropical spastic paraparesis: a clinical study of 50 patients from Tumaco (Colombia) and review of the worldwide features of the syndrome. J. Neurol. Sci. 87:121, 1988. 87. Román, G. C., Román, L. N., and Osame, M.: Human T lymphotropic virus type 1 neurotropism. Prog. Med. Virol. 37:190, 1990. 88. Román, G. C., Spencer, P. S., and Schoenberg, B. S.: Tropical myeloneuropathies: the hidden endemias. Neurology 35:1158, 1985. 89. Román, G. C., Vernant, J.-C., and Osame, M. (eds.): HTLV-I and the Nervous System. New York, Alan R. Liss, 1989. 90. Román, G. C., Vernant, J.-C., and Osame, M.: HTLV-I– associated motor neuron disease. In de Jong, J. M. B. V. (ed.): Handbook of Clinical Neurology, Vol. 15. Diseases of the Motor System. Amsterdam, Elsevier, p. 447, 1991. 91. Routier, R. J., Lipman, J., and Brown, K.: Difficulty in weaning from respiratory support in a patient with the intermediate syndrome of organophosphate poisoning. Crit. Care Med. 17:1075, 1989. 92. Said, G., Goulon-Goeau, C., Lacroix, C., et al.: Inflammatory lesions of peripheral nerve in a patient with HTLV-I–associated myelopathy. Ann. Neurol. 24:275, 1988. 93. Santiesteban-Freixas, R., Pamias-Gonzalez, E., Luis, S., et al.: Epidemic neuropathy: proposal and arguments to rename

94. 95.

96.

97.

98.

99.

100.

101.

102.

103.

104. 105.

106. 107.

108. 109. 110.

111.

112. 113.

114.

2079

Strachan disease as Strachan-Mádan disease. Rev. Neurol. 25:1950, 1997. Scott, H. H.: An investigation into an acute outbreak of “central neuritis.” Ann. Trop. Med. Parasitol. 12:109, 1918. Senanayake, N.: Tri-cresyl-phosphate neuropathy in Sri Lanka: a clinical and neurophysiological study with a three year follow up. J. Neurol. Neurosurg. Psychiatry 44:775, 1981. Senanayake, N., and Jeyaratnam, J.: Toxic polyneuropathy due to gingili oil contaminated with tri-cresyl-phosphate affecting adolescent girls in Sri Lanka. Lancet 1:88, 1981. Senanayake, N., and Johnson, M. K.: Acute polyneuropathy after poisoning by a new organophosphate insecticide. N. Engl. J. Med. 306:155, 1982. Senanayake, N., and Karalliedde, L.: Neurotoxic effects of organophosphorous insecticides: an intermediate syndrome. N. Engl. J. Med. 316:761, 1987. Senanayake, N., and Karalliedde, L.: Pattern of acute poisoning in a medical unit in central Sri Lanka. Forensic Sci. Int. 36:101, 1988. Sitrin, M. D., Lieberman, F., Jensen, W. E., et al.: Vitamin E deficiency and neurologic disease in adults with cystic fibrosis. Ann. Intern. Med. 107:51, 1987. Smith, D. A.: Nutritional neuropathies in the civilian internment camp, Hong Kong, 1942–August, 1945. Brain 69:209, 1946. Spencer, P. S., Hugon, J., Ludolph, A., et al.: Discovery and partial characterization of primate motor-system toxins. Ciba Found. Symp. 126:221, 1987. Spencer, P. S., Schaumburg, H. H., and Ludolph, A. C.: Experimental and Clinical Neurotoxicology, 2nd ed. New York, Oxford University Press, 2000. Spillane, J. D.: Nutritional Disorders of the Nervous System. Edinburgh, E & S Livingstone, 1947. Spillane, J. D., and Scott, G. I.: Obscure neuropathy in the Middle East: report on 112 cases in prisoners-of-war. Lancet 2:261, 1945. Stannus, H. S.: Pellagra and pellagra-like conditions in warm climates. Trop. Dis. Bull. 33:729, 815, 885, 1936. Stefanini, M.: Clinical features and pathogenesis of tropical sprue: observations in series of cases among Italian prisoners of war in India. Medicine (Baltimore) 27:379, 1948. Strachan, H.: Malarial multiple peripheral neuritis. Sajou’s Ann. Univ. Med. Sci. 1:139, 1888. Strachan, H.: On a form of multiple neuritis prevalent in the West Indies. Practitioner 58:477, 1897. Striefler, M., Cohn, D. F., Hirano, A., and Schujman, E.: The central nervous system in a case of neurolathyrism. Neurology 27:1176, 1977. Susser, M., and Stein, Z.: An outbreak of tri-ortho-cresyl phosphate (TOCP) poisoning in Durban. Br. J. Ind. Med. 14:111, 1957. Toro, G., Román, G. C., and Román, L. N.: Tropical Neurology. Bogotá, Printer Colombiana, 1983. Tranchant, D., Darracq, R., and Ticolat, R.: Carence en vitamine E chez 8 malades atteints de neuromyélopathie nutritionnelle tropicale. Presse Med. 15:1729, 1986. Vallat, J. M., Dumas, M., Giordano, C., et al.: Histologic study of peripheral nerve biopsies from 20 patients with tropical ataxic neuropathy. Neurology 37(Suppl. 1):255, 1987.

2080

Diseases of the Peripheral Nervous System

115. Vejjajiva, A.: Acute hemorrhagic conjunctivitis with nervous system complication. In McKendall, R. R. (ed.): Handbook of Clinical Neurology, Vol. 12. Viral Disease. Amsterdam, Elsevier, p. 349, 1989. 116. Vernant, J.-C., Bellance, R., Buisson, G. G., et al.: Peripheral neuropathies and myositis associated to HTLV-I infection in Martinique. In Blattner, W. A. (ed.): Human Retrovirus: HTLV. New York, Raven Press, p. 225, 1990. 117. Vora, D. D., Dastur, D. K., and Braganca, B. M.: Toxic polyneuritis in Bombay due to ortho-cresyl-phosphate poisoning. J. Neurol. Neurosurg. Psychiatry 25:234, 1962.

118. Walters, J. H.: Dietetic deficiency syndromes in Indian soldiers. Lancet 1:861, 1947. 119. Williams, A. O., and Osuntokun, B. O.: Peripheral neuropathy in tropical (nutritional) ataxia in Nigeria: light and electron microscopic study. Arch. Neurol. 21:475, 1969. 120. Wintrobe, M. M., Miller, M. H., Follis, R. H., et al.: Sensory neuron degeneration in pigs. IV. Protection afforded by calcium pantothenate and pyridoxine. J. Nutr. 24:345, 1942. 121. World Health Organization: Surveillance of peripheral neuropathies. Wkly. Epidemiol. Rec. 28:213, 1982.

M. Neuropathy Associated with Infections

91 Leprosy THOMAS D. SABIN, THOMAS R. SWIFT, AND ROBERT R. JACOBSON

History and Stigma of Leprosy Epidemiology General Evaluation and Classification of Patients Tuberculoid Leprosy Lepromatous Leprosy Borderline Leprosy Complications Leprosy Reactions Amyloidosis Socioeconomic Factors Deformity

The Syndrome of Superficial Neuropathy Tuberculoid Leprosy Lepromatous Leprosy Borderline Leprosy Neuritis in Leprosy Reactions Nerve Compression Differential Diagnosis Distal Polyneuropathy Mononeuropathy and Mononeuropathy Multiplex Nerve Enlargement Absorption

Leprosy (Hansen’s disease) is chronic infectious disease caused by Mycobacterium leprae, a microorganism to which only a small portion of any given population is susceptible. Although M. leprae can be found nearly anywhere in the body outside the central nervous system, even in the more severe types of leprosy it produces significant damage only in superficial nerves, the skin, the anterior third of the eye, the upper respiratory tract, and the testes. The precise clinical and pathologic manifestations are largely determined by the natural resistance of the host to invasion by M. leprae. There has been preliminary work determining genetic susceptibility loci. The disease is rarely fatal, but serious disability may result from neuropathy and eye involvement. Leprosy is also complicated by social dislocation, immunologically induced reactional states, and secondary amyloidosis. The formerly widely practiced segregation of patients and the cloud of folklore surrounding leprosy have obscured

Electrophysiologic Testing Neuropathology Tuberculoid Leprosy Lepromatous Leprosy Borderline Leprosy Leprosy Reactions in Nerves Muscle Involvement Treatment Antibacterial Treatment Treatment of Leprosy Reactions Surgical Treatment Management of Insensitivity Social Rehabilitation

certain features of the disease of compelling interest for the neurologist: leprosy remains a common treatable neuropathy on a worldwide basis, bacillary invasion of peripheral nerve occurs in all cases, and the neurologic examination provides extremely reliable diagnostic information.

HISTORY AND STIGMA OF LEPROSY This ancient disease known as leprosy has a remarkable history that is inextricably bound to the strong societal attitudes that the leprosy patient has endured. A disease called kushtha in the Verdas manuscripts of India (circa 1300 BC) showed several of the hallmarks of leprosy.17 Ancient Chinese manuscripts from 60 BC contain similar observations that probably referred to leprosy, and Galen accurately described the disease that became known as 2081

2082

Diseases of the Peripheral Nervous System

Greek elephantiasis. The Hebrew word tsara’ath was rendered as “leprosy” in the translations of the Bible. Among Western writers there has been a tendency to overemphasize the biblical references as the source of the stigma. Tsara’ath referred to a variety of skin disorders and other conditions that demanded separation from the tribe. As Skinsnes has indicated, leprosy stigma arose in areas of the world where the Bible had no influence93; the source of the stigma is in the visible disfiguring nature of the lesions distorting the face, hands, and feet. Leprosy appeared in Europe at the fall of the Roman Empire and reached peak prevalence during the Crusades before beginning a mysterious decline in the 14th century.17 During the Crusades, the care of leprosy patients became a fashionable imitation of the life of Christ, and Lazarist houses proliferated across Europe and the British Isles. These homes provided largely domiciliary care, and diagnostic criteria for admission were not strict. The end of the Crusades brought the return of leprosy stigma; the brief era of “leprophilia” gave way to the return of social opprobrium that has persisted to the present but has been gradually weakened by modern views of leprosy that originated in the late 19th century. Abundant examples of the extent of stigmatization of leprosy over the centuries are found in the law, folklore, art, and literature of all continents. It was often considered a supernaturally decreed punishment for some moral wrong, especially egotism and venery. In literature the disease embodies hopeless physical and moral degradation.92 It was believed to be hereditary or contagious. The violent aversion to and fear of the afflicted resulted in their segregation in leprosariums. Patients were often forced to wear bells or identifying emblems. Not all the horror stories arise from ancient history. In 1937, 57 patients were gathered together on the pretext of an Easter breakfast and shot to death in South China.45 Anyone who undertakes the care of a leprosy patient today will soon realize that the factor of stigma is still present. Panic may follow admission of such a patient to a general hospital, and even for the person with an arrested case, employment is difficult to find. Patients with leprosy justifiably abhor the term leper, for they are painfully aware that it implies much more than having a mycobacterial infection; many of them have suffered more from the ravages of the stigma than from the disease itself.

EPIDEMIOLOGY When the effectiveness of multidrug therapy was established in the 1980s, international leprosy agencies set the goal of eliminating leprosy as a public health problem in the world by the year 2000. This was defined as reduction in the prevalence of the disease to one case per 10,000 population or below throughout the world and in all countries. It was hoped that at this level the disease could be easily managed

within the primary health care system. The effort was successful in terms of overall world statistics and in most endemic countries. The widespread effort also resulted in earlier case detection and a reduction in disabilities among new cases, and by assuring that all cases are rapidly and adequately treated, long-term disability has also been reduced. Since cases were removed from the registry upon completion of treatment, the number of registered cases had declined from 5,069,232 in 1986 to 700,000 by 2001. The effort to eliminate leprosy as a public health problem has yet to succeed in the two most endemic countries, India and Brazil, and in Myanmar, Madagascar, Nepal, and Mozambique. India now accounts for 64% of worldwide cases, and their intensive leprosy elimination campaigns have detected many more cases than anticipated from the start of the effort.52 The World Health Assembly has extended the elimination of leprosy as a public health problem effort to 2005, and India is now evaluating a vaccine that may help diminish the problem.26 While leprosy is not a large problem in the United States, immigration from Asia and South America should keep this diagnosis in mind for American neurologists. The incidence of new cases from the onceendemic areas around the Gulf Coast has been eliminated. The most likely portal of entry is considered to be the upper respiratory tract,105 but other possibilities such as inoculation via the skin exist.40 Untreated patients with multibacillary disease daily produce nasal discharge containing millions of microorganisms that can remain viable outside of the human host. Infected armadillos in the wild have also been proposed as an occasional source of cases.55 In the United States, study of the problem is hampered by our inability to culture M. leprae. Over 95% of a given population are probably not susceptible to the disease after routine exposure, and although children appear more susceptible than adults do, it may be more of a matter of initial exposure usually occurring during childhood. Marital partners are at lower risk than household children,49 and the disease is rare among visitors to endemic areas, including soldiers and missionaries. The decline in leprosy that occurred in Europe and more recently in Korea and China correlates better with the changes generated by a rising standard of living than with any specific measures directed toward eradication of the disease. Where extreme public health measures such as enforced segregation were instituted in the past, a false decrease in incidence occurred because patients concealed the diagnosis out of fear of internment.

GENERAL EVALUATION AND CLASSIFICATION OF PATIENTS The complaints of patients with leprosy include skin lesions, sensory loss, nasal stuffiness, epistaxis, decreased vision, loss of body hair, male sterility, and gynecomastia.

Leprosy

The complications of leprosy are discussed in a separate section of this chapter. Evaluation of patients includes physical examinations, skin biopsy, and skin scrapings. The latter are obtained by scraping the edge of a tiny skin slit and spreading the material obtained on a glass slide over an area about 5 mm in diameter. Areas commonly tested in this fashion include the earlobes, elbows, knees, and selected skin lesions. A biopsy section and the skin scrapings are stained for acid-fast bacilli. Depending on factors such as drug treatment and duration of illness, the bacilli will appear either intact, smooth in outline, and evenly stained (“solids”) or degenerate with a beaded or irregular appearance. Between 1% and 10% of all bacilli typically appear intact prior to therapy. This percentage is referred to as the morphologic index (MI),102 and it correlates with viability dropping to 0% after the start of treatment. Bacilli are quantified using the bacterial index (BI), and this varies between 0 (no bacilli seen in 100 oil immersion fields) and 6⫹ (over 1000, i.e., uncountable numbers of bacilli in each oil immersion field). This index may be useful in following the response of the patient to drug therapy, typically falling at a rate of 0.5 to 1 per year.71 However the MI is very difficult to determine accurately even by trained pathologists or technicians, and the World Health Organization (WHO) advises that it no longer be routinely used in patient evaluations.102 Today essentially all leprosy patients are classified as paucibacillary (PB) or multibacillary (MB). The WHO defines paucibacillary cases as those having five or fewer lesions and negative skin scrapings.103 Multibacillary patients have six or more lesions and/or positive skin scrapings at any site. A third category called “single-lesion

FIGURE 91–1 The clinicopathologic spectrum of leprosy.

2083 paucibacillary” is used in some areas for patients having only a single PB lesion. Although the WHO classification has the advantage of being relatively easy to use under field conditions, to fully understand the disease and its histopathology, manifestations, and complications, it is useful to discuss it in terms of the Ridley-Jopling103 classification. This includes three major forms and one minor form (Fig. 91–1). The three major forms are tuberculoid (TT), borderline (BB), and lepromatous (LL) leprosy. The minor form is called indeterminate (I) leprosy. The borderline type is often further divided into borderline-tuberculoid (BT), mid-borderline (BB), and borderline-lepromatous (BL). The WHO classification under PB would include I, TT, and BT of the Ridley-Jopling classification, while MB would include BB, BL, and LL. Single-lesion paucibacillary would include I and some TT. Mycobacterium leprae is an acid-fast rod 1 to 8 ␮m long and 0.2 to 0.5 ␮m in diameter.14 It is morphologically indistinguishable from M. tuberculosis. The high cytoplasmic encapsulated lipoidal content accounts for the acid-fast staining. The genome has been sequenced and there is massive gene decay (reductive evolution) when compared with its cousin M. tuberculosis.18 At an optimum temperature for growth of 27° to 30° C, the bacillus multiplies intracellularly with a division time of 12.5 days in the logarithmic phase of growth as measured in the mouse footpad.87 It is the only bacterium that consistently invades peripheral nerves. Koch’s postulates remain unfulfilled despite the fact that M. leprae was linked with human disease in 1874.27 This problem is largely related to the difficulties of cultivating the bacillus. Growth on artificial media has not yet been accomplished, but limited growth

2084

Diseases of the Peripheral Nervous System

in the mouse footpad has been achieved,86 and this approach is utilized for drug sensitivity testing.88 Generalized disease has been recorded in the irradiated thymectomized infant mouse.70 Disseminated disease with peripheral nerve invasion occurs in armadillos48 and mangabey and rhesus monkeys.4 Although only a relatively small percentage of diagnosed cases are indeterminate, all patients may go through an indeterminate phase during the evolution of their disease, as indicated in Figure 91–2A and Plate 91–2A. The form of leprosy that develops does not depend on the bacillus, which is the same for all forms, but on the response of the host to the presence of bacilli. Individuals who develop leprosy have a relatively specific defect in their cell-mediated immune (CMI) response to M. leprae.28,83,85 Normally, when M. leprae bacilli enter the body and come in contact with macrophages, they are activated through a series of steps involving, in particular, thymus-derived (T) lymphocytes and are thereby rendered capable of destroying and clearing of bacilli. This is the sequence of events in perhaps 95% or more of the population, and when exposed this group does not develop leprosy. The specific defect that prevents this normal chain of events from occurring so that leprosy develops is not fully understood but may be the result of excessive suppressor T-cell activity. The source of the defect is uncertain, but genetic and acquired factors may be involved. Further understanding of the problem may yield a successful and safe immunotherapeutic approach to the management of this disease in line with the limited successes already reported with such immunotherapeutic agents as transfer factor,31 interferon-␣,61 and vaccination with a mixture of heat-killed M. leprae and bacille Calmette-Guérin (BCG).19 Whatever the exact nature of the minimal defect, its severity determines the type of leprosy that will occur. Those with a minimal defect keep the disease localized, developing tuberculoid leprosy, and those with a severely defective CMI response will allow the infection to become generalized and develop lepromatous leprosy. Borderline disease occurs in those with defects between these two extremes. The lepromin skin test, which utilizes a suspension of killed M. leprae, is given like a tuberculin test and provides a simple measure of the CMI response. It is positive in the overwhelming majority of the general population, because they are not susceptible to leprosy, and in TT and BT cases, and is negative in BB, BL, and LL patients. In classifying the type of leprosy present in an individual, the clinical manifestations, histologic response to the presence of bacilli, and number of bacilli must be considered.

Tuberculoid Leprosy In tuberculoid leprosy, proliferation and dissemination of bacilli are limited. This results in a well-localized, asymmetrical disease process. There is a skin lesion that is

sharply demarcated, is often hypopigmented, and has a tendency to slow enlargement with central clearing. The border of the lesion may be elevated and erythematous. The lesion may be circular, ovoid, or serpiginous. Its central part usually has a dry, scaly appearance (see Fig. 91–2B and Plate 91–2B). Only one or a few such lesions occur. They commonly occur on the extensor surfaces of the limbs, the face, or the buttocks; they do not involve the perineum, axillae, or scalp. The prognosis is good, and many cases go on to spontaneous healing without treatment. Skin tests with lepromin are positive. Biopsy of the active border of such a lesion reveals an epithelioid granuloma that extends from subcutaneous tissues up through the dermis and in some instances breaches the basal layer of the epidermis (see Fig. 91–2C and Plate 91–2C). The cellular response is most vigorous about dermal appendages. Cutaneous and subcutaneous nerves may be invaded or completely destroyed. Many epithelioid cells are seen, as well as foreign body and Langerhans giant cells. Caseation may be present in the center of the granuloma. Lymphocytes are scattered about the periphery. Acid-fast stain occasionally reveals no organisms, although diligent searching of many sections under oil immersion usually yields a few. Spread of bacilli occurs by direct contiguity. Blood-borne spread has never been demonstrated in tuberculoid leprosy.

Lepromatous Leprosy Unchecked bacillary proliferation and hematogenous dissemination occur in lepromatous leprosy. The clinical manifestations are due to extensive infiltration of the cooler body tissues rather than to a violent tissue response to a few bacilli as is seen in tuberculoid leprosy. The possible role of low tissue temperature in leprosy was suggested by Binford, who attempted to grow M. leprae in laboratory animals.5 Brand commented on the distribution of pathologic changes in leprosy and noted that the most affected tissues, although diverse in structure and function, were all several degrees below core body temperature.9 Subsequent studies have revealed that bacillary density in lepromatous leprosy is increased in cooler body areas32 and that optimal proliferation of M. leprae in the mouse footpad occurs at about 30° C.87 The skin lesions may be macules, papules, nodules, or plaques and resemble those of a host of dermatologic disorders. Occasionally bullous or ulcerated lesions are seen. There may be a solitary skin lesion at the outset, but there is a strong tendency for the lesions to be numerous and symmetrically disposed. Some patients develop no visible skin lesions but do have diffuse infiltration of the skin. There may be only a subtle erythema or waxy sheen to the skin in the most affected areas, with a slight swelling in the face and distal extremities. The very gross nodular infiltration of the face, called “leonine facies,” is one of the most widely known

Leprosy

FIGURE 91–2 A, Indeterminate leprosy. A biopsy has been taken from the border of the hypopigmented lesion. (From Trautman, J. R., and Enna, C. D.: Leprosy. In Practice of Medicine, Vol. 3. New York, Harper & Row, 1970, with permission.) B, Tuberculoid leprosy. Note raised erythematous border and depressed hypopigmented center. C, Tuberculoid leprosy. Skin biopsy showing epithelioid granuloma extending to basal layer of epidermis. Note Langerhans giant cell. (Hematoxylin and eosin, ⫻450.) D, Lepromatous leprosy. Skin biopsy in which numerous acid-fast bacilli are present, many in the form of globi. Note lepra cells. (Wade-Fite stain, ⫻1000.) E, Borderline leprosy. Note symmetrically distributed skin lesions. F, The relatively normal-appearing segment of the median nerve was under the forearm flexor muscles before they were retracted. When the nerve emerges from under these muscles to superficial position, it becomes grossly erythematous and enlarged. (From Sabin, T. D.: Neurologic features of lepromatous leprosy. Am. Fam. Physician 4:84, 1971, with permission.) G, The right hand appears normal; the left hand is powerless because of median, ulnar, and radial nerve involvement at the level of the elbow. H, After 10 years, the right hand retains good motor power but has undergone absorption. The left hand could not be used, but digit length is unchanged. (From Sabin, T. D.: Lessons from leprosy. Am. J. Occup. Ther. 33:73, 1969, with permission.) I, Lepromatous leprosy. Small cutaneous nerve showing vacuoles containing acid-fast organisms, hypercellularity, and splitting apart of perineurium to give an onionskin appearance. (Wade-Fite stain, ⫻450.) See Color Plate

2085

2086

Diseases of the Peripheral Nervous System

signs of lepromatous leprosy. This represents a most advanced stage of the disease, which has not often been seen since the advent of effective chemotherapy. The infiltration that is seen in earlier cases is most evident over the nose and malar areas. The earlobes are frequently enlarged and pendular. Infiltration of the upper respiratory tract results in nasal stuffiness and sometimes bleeding. The cartilaginous nasal septum may become perforated, resulting in collapse of the nose. The involvement may extend to the larynx but never to the lower respiratory system. Hoarseness may be progressive, and although actual respiratory obstruction rarely occurs nowadays, tracheostomy was not infrequent prior to the sulfone era. The anterior third of the eye also becomes infiltrated and may develop several complications as a result of leprosy reactions or nerve damage.30 Beading of the corneal nerves may be seen with the slit lamp.64 Ocular pain, photophobia, and glaucoma secondary to involvement of the ciliary body may occur. Invasion of the testes is frequent. Orchitic pain, sterility, and gynecomastia may result as the disease advances.36 Palpable inguinal and epitrochlear lymphadenopathy is often present. Constitutional symptoms in uncomplicated lepromatous leprosy are minimal or absent. Lepromin skin tests are negative. Biopsy of skin reveals few changes in the epidermis. The rete is flattened. A clear space separates the epidermis from an infiltrate of histiocytes, which have foamy cytoplasm and are most numerous about blood vessels, nerves, and skin appendages. Occasionally many plasma cells are present. With acid-fast stain great numbers of bacilli are seen within the histiocytes (see Fig. 91–2D and Plate 91–2D). Such cells are known as Virchow cells, and the masses of bacilli contained within them are referred to as “globi.” Several such cells may seem to fuse, producing giant globi. Depending on factors such as drug treatment and duration of illness, the MI may be between 1% and 10% and the BI is typically 4 to 6⫹. These indices are useful in following the response of the patient to drug therapy. Unlike the epithelioid granuloma of tuberculoid leprosy, which reflects an active tissue response, the tissue response in the lepromatous form is quite passive, and destruction of affected organs and tissues does not occur early in the disease. The disease rarely shows evidence of spontaneous arrest and may recur after chemotherapy is discontinued even though the BI has fallen to 0.34

Borderline Leprosy The pattern of disease in these cases develops from variations in host tissue immunity to M. leprae that result in a clinical picture ranging from tuberculoid-like to near lepromatous. When they resemble tuberculoid leprosy the lesions are multiple, and with declining resistance a tendency toward symmetry becomes evident (see Fig. 91–2E and Plate 91–2E). A single borderline lesion may extend over a large percentage of the body surface. The margins of a borderline lesion tend to be less distinct than those of

a tuberculoid patch, and central clearing is uncommon except in BT lesions. The entire lesion may be slightly elevated and often has a reddish-blue hue. Skin biopsy reveals both lepromatous and tuberculoid features. There may be both foam cells containing bacilli and epithelioid cells surrounding cutaneous structures. Lymphocytes may be abundant. If the disease is near the tuberculoid end of the spectrum, epithelioid cells will predominate and few bacilli will be present; if it is near the lepromatous end, foam cells and bacilli will be found in great numbers. Clinically and pathologically borderline leprosy is unstable, tending to move toward lepromatous (downgrading) without treatment or tuberculoid (reversal reaction) with treatment.

COMPLICATIONS Leprosy Reactions Type 1 (Reversal) Reactions Type 1 reactions occur in borderline and tuberculoid leprosy. This is a sudden flaring of preexistent skin lesions, which become erythematous and edematous and occasionally ulcerate. New lesions and neuritis involving one or more peripheral nerves may occur. These reactions usually appear during the first few months following the initiation of effective chemotherapy and are thought to indicate an increase in the intensity of the CMI response to M. leprae, because as the reaction clears the lesions heal, and there is a marked reduction or eradication of bacilli in some instances. This rapid transition to a higher resistance form of leprosy is generally a favorable sign, but permanent neurologic deficits may be precipitated unless antireaction treatment is promptly initiated. Downgrading Reactions Although immunologic downgrading in borderline disease is common in untreated patients, this usually occurs slowly with no evidence of inflammation such as erythema or edema of the skin lesions or fever accompanying it. Rarely conditions such as malignant neoplasms and severe intercurrent infections, which are known to depress cell-mediated immunity, may result in a more rapid shift toward the lepromatous pole in patients with borderline leprosy accompanied by signs of reaction. During the course of such a reactive episode, a once positive lepromin skin test may become negative and bacilli increase in number, resulting in symmetrical infiltrative lesions. Institution of antimicrobial therapy will usually promptly arrest this process. Type 2 Reactions (Erythema Nodosum Leprosum) This form of reaction is associated with lepromatous leprosy, though BL patients may also develop it. It results

2087

Leprosy

in a serious systemic illness with fever, malaise, and prostration that may last several days or many months. Erythema nodosum leprosum (ENL) develops most often when a decrease in the BI occurs after several months to a year or more of therapy, but it may occur in untreated patients. It may also be precipitated by febrile illness or reaction to nonspecific stresses. Acute lesions occur in sites where the bacilli are most numerous. Crops of painful, red, indurated cutaneous and subcutaneous nodules develop. Acute painful peripheral nerve damage, iridocyclitis, orchitis, and arthritis may also occur. Hepatosplenomegaly and lymphadenopathy are sometimes detected. The pathologic substrate of this distressing syndrome is vasculitis that involves medium-sized and small arteries and arterioles. Fibrinoid degeneration of the media is accompanied by a cuff of inflammatory cells surrounding the vessel. In certain Mexican and Latin American patients with a diffuse infiltrative lepromatous leprosy, progression to actual thrombosis of deeper subcutaneous arteries may occur with necrosis of skin and subcutaneous fat. When skin ulceration is severe, the underlying tendons and muscles may be exposed. This rare form of reaction (Lucio’s phenomenon) may have a fatal outcome. Healing results in a thin, white “cigarette paper” scar that resembles Degos’ disease. Whereas the CMI response to M. leprae is defective in lepromatous leprosy, the humoral antibody system functions adequately and appears to be involved in the pathogenesis of ENL, which may represent a classic example of immune complex disease. Although the whole process now appears more complex than originally thought,20 a simplified overview indicates that ENL results when a very large amount of mycobacterial antigen is suddenly made available, as occurs after drug treatment has been given. Antigen-antibody complexes are deposited in areas of high antigen concentration, and complement fixation and cell lysis occur. This provokes the intense polymorphonuclear infiltration seen about such vessels. Active lysosomal digestion proceeds to microabscess formation, which evokes continued polymorphonuclear response. Mycobacterial antigen, antibody, and complement have been demonstrated in the lesions of ENL101 and are similar to what is seen in the Arthus reaction.

A serum protein, serum amyloid-A (SAA), is also found that, although it is not an immunoglobulin, is antigenically related to the protein deposited in secondary amyloidosis. SAA levels rise in lepromatous patients during ENL. Nonlepromatous patients have elevated levels of SAA only with active neurotrophic ulcers. These data make a strong link between ENL and secondary amyloidosis.

SOCIOECONOMIC FACTORS Rejection and isolation by community and even family members lead to social and economic dislocation. In desperation many patients have sought the relative acceptance and anonymity offered by leprosariums. Even when the disease is finally controlled, the young patient with little or no deformity is often unwilling to reenter society and may instead become institutionalized. Nowadays, outpatient treatment is the rule rather than the exception, and coincidentally the stigma is slowly diminishing. Leprosariums are being phased out everywhere.

DEFORMITY As a result of modern chemotherapy and improved leprosy control efforts, deformity is slowly becoming less common, and many patients today who have no significant deformity at the time of diagnosis will never develop it. Most deformity in leprosy is due to the neuropathy. This is not limited to the contractures resulting from paralysis but is more related to the extensive areas of insensitivity, which permit recurrent injuries to the hands, feet, and eyes. The eventual outcome of this process is progressive absorption in the extremities (described in the section on Differential Diagnosis below) and blindness. Rarely, direct invasion of the phalanges by leprosy bacilli results in pathologic fractures of the fingers. Facial infiltration and collapse of the nasal septum are other examples of deformity primarily related to direct effects of massive proliferation of M. leprae.

Amyloidosis

THE SYNDROME OF SUPERFICIAL NEUROPATHY

Although now rare because of early treatment, secondary amyloidosis may occur in cases of long-standing leprosy and is indistinguishable from that complicating other chronic inflammatory illnesses, such as rheumatoid arthritis.91 The actual incidence of this complication is unknown and varies with the type of leprosy, presence of adequate chemotherapy, and other factors, such as race. The organs primarily involved clinically are the kidney and the liver, with other tissues involved less often. In contrast to other forms of amyloidosis, the peripheral nerves are spared.

The cardinal symptom of leprosy is sensory loss. The loss is most frequently discovered by the occurrence of painless injury. When the sensory loss affects the hands or feet, the patient is made aware of it sooner than with sensory loss elsewhere. Many patients have sensory loss long before other evidence of the disease evolves, and some, often children, do not become alarmed by the loss of sensation and may even amuse acquaintances by demonstrating their “tolerance” of pain. Paresthesias do not always precede the sensory loss, but some patients do report

2088

Diseases of the Peripheral Nervous System

pruritus and a burning sensation in areas of commencing sensory loss. Pain in portions of involved nerves that are swollen and enlarged is apt to occur if the nerve is struck or stretched. The pain in such nerve segments often remains remarkably localized to the affected area. Less often, patients may complain of radiating pain, as is typically seen more often in some other neurologic diseases. Sensory abnormalities precede paralysis in all types of leprosy. Most sensory loss in leprosy is due to intracutaneous nerve damage. We find that pain and temperature sensation are most strikingly decreased in early cases. The patient may readily discern the sharp from the blunt end of a pin, yet the pin simply does not hurt. When the patient clearly understands that only the painful quality of the pinprick is to be evaluated by continuing comparison between pairs of stimuli, the sensory defect in leprosy can readily be mapped out. Position and vibration sensations are entirely normal except in the unusual or very advanced case in which multiple major nerves in an extremity are affected. The modalities are dissociated, but not so precisely as in syringomyelic syndromes, because of the variable affection of touch sensation in leprosy.

The evolution, pattern, and extent of the sensory loss and paralysis are determined by the type of leprosy. A carefully made sensory map is a useful tool for placing the case accurately along the spectrum from high- to lowresistance disease. Leprosy of all types is characterized by sensory loss that is primarily due to intracutaneous destruction and therefore not in the pattern of any peripheral nerve or nerve root distribution.

Tuberculoid Leprosy In tuberculoid leprosy there may be only one or at most part a few patches of intracutaneous sensory loss. The skin lesion and the sensory loss develop together, so the sensory loss is present by the time the patient is seen. Nerve damage thus appears to be an inherent part of the host response to the bacilli and is not related to massive proliferation of bacilli. The sensory loss is invariably accompanied by anhidrosis, and is sharply delimited and coterminous with the skin lesion (Fig. 91–3). A nerve trunk coursing beneath such a tuberculoid patch may also be affected, and the sensorimotor deficit of that nerve will then be superimposed.

FIGURE 91–3 Tuberculoid leprosy sensory map.

2089

Leprosy

The major mixed nerves that are most commonly affected include the ulnar, median, peroneal, and facial (Fig. 91–4). Subcutaneous sensory nerves, particularly those near the tuberculoid patch, are usually also palpably enlarged, sometimes forming skeins of firm nerves. The superficial cutaneous radial, digital, posterior auricular, and sural nerves are commonly affected sensory nerves. The intense response to bacilli within nerves in high-resistance leprosy may lead to necrosis with the formulation of “cold” nerve abscesses.82 These are most common in the ulnar and peroneal nerves. Intense pain may result, and surgical drainage of the abscess is advised. Calcification of nerves in long-standing cases of high-resistance leprosy may be sufficiently intense to appear on radiographs.53,84

FIGURE 91–4 The major mixed nerves most commonly affected in leprosy.

Lepromatous Leprosy The sensory loss in lepromatous leprosy is not confined to the area of the skin lesion in the precise manner noted in tuberculoid leprosy. The sensation in the lepromatous macule or nodule is the same as in the immediately adjacent normal-appearing skin. Skin lesions are, however, more numerous and prominent over the body regions where sensory loss is greatest. Skin lesions may have normal sensation in lepromatous disease. Palpably enlarged nerves may be functioning well. This is another feature of low-resistance leprosy that sharply contrasts with tuberculoid leprosy, wherein nerve dysfunction seems to be a prompt inherent feature of the host-parasite relationship.

2090

Diseases of the Peripheral Nervous System

The infiltrated nerves in low-resistance leprosy do eventually fail. The dysfunction appears first in loci where the bacilli have proliferated most actively. Because M. leprae reproduces most vigorously in the cooler tissues, a unique pattern of neurologic deficits emerges. Small differences in tissue temperature have effects on the evolution of the neuropathy.74 Hematogenous spread of bacilli in lowresistance leprosy ensures symmetry in the distribution of neural deficits. Sensory loss in lepromatous leprosy has incorrectly been described as showing a “stocking-glove” distribution. The relative coolness of the distal portions of the limbs probably led to this erroneous conclusion. At all stages the evolution of the temperature-linked sensory loss of lepromatous leprosy can be distinguished from the distal symmetrical sensory loss with fading proximal borders (stocking-glove pattern) that is the hallmark of numerous sensory polyneuropathies.75 Temperatures over the surface of the body vary among individuals and within an individual according to a wide variety of factors. Ambient temperature, humidity, clothing, fat distribution, drugs, vascular disease, and emotional state are but a few of the influences on surface temperature. Yet if one examines many normal subjects with thermographic

scanning, a reasonably constant overall pattern of surface temperature emerges. This is a constancy of temperature gradient between different body surfaces rather than a constancy of absolute temperatures. The evolution of neurologic deficits in lepromatous leprosy as described in the following paragraphs is clearly related to this usual distribution of surface temperature gradients.77 Numerous individual variations may be discovered, but the overall pattern is always evident enough for diagnostic purposes once the general principle is familiar to the examiner. Figures 91–5 through 91–8 demonstrate the patterns of sensory loss in order of severity. Cutaneous sensory loss appears first in the pinnae of the ears, the dorsal surfaces of the hands, the dorsomedial surfaces of the forearms, the dorsal surfaces of the feet, and the anterolateral aspects of the legs (Fig. 91–5). The lobes and helices of the ears are most affected. Progression results in more extensive loss on the dorsal forearms to the elbow regions and over the medial leg and anterior knee areas (see Fig. 91–6). Commencing sensory loss is now detectable over the nose, malar areas, breasts, central abdomen, and buttocks. At this stage, sparing of the palms and soles is often present even when there is

FIGURE 91–5 Lepromatous leprosy sensory map: initial sensory loss.

Leprosy

2091

FIGURE 91–6 Lepromatous leprosy sensory map: progression of sensory loss.

profound cutaneous sensory loss over the dorsa of the hands and feet. Although thermographic scanning often shows slightly greater warmth of the palms and soles than of the other dorsal surfaces, the actual temperature of the bacillary milieu is further increased by the insulating effect of the thickened cuticle of the palms and soles. A distinct change in sensation is often detectable at the cuticular border. This feature may be especially noticeable when sensory loss and lepromatous skin lesions of the medial arch terminate abruptly along the border of the cuticle. When sensory loss has progressed to this stage, paralysis appears. Paralysis reflects the involvement of the nerves containing motor fibers at locations where they lie closest to the surface of the body (see Fig. 91–5). The increased growth rate of bacilli at these sites is the initiating factor in nerve dysfunction. Once the nerves in these superficial areas become enlarged, chronic trauma and compression may play a secondary role in the pathogenesis of the paralyses. Trauma is not the primary cause of the paralysis, however, because several nerves (e.g., facial, auricular, supraorbital) affected in leprosy are not in positions that

are particularly vulnerable to repeated trauma. The deficits are more stereotyped in lepromatous leprosy than in high-resistance types of the disease. Paralysis in the ulnar distribution usually appears first. The ulnar nerve is most affected over a 10- to 15-cm segment proximal to the olecranon groove (see Fig. 91–7). The nerve is greatly thickened and erythematous in this segment but may have a grossly normal appearance elsewhere. A 2- to 4-cm segment of abnormality also may appear proximal to the transverse carpal ligament. Job and Desikan have shown that the greatest bacillary density occurs in these more superficial portions of the nerve.39 The ulnar nerve of normal subjects in this vulnerable arm site is 2.3° C cooler than in the mid-forearm, where it lies deep to muscle.78 If ulnar damage supervenes in a patient with sparing of sensation in the palm, a complex situation arises with regard to the sensory status of the hand: the medial palm is insensitive only because of the ulnar nerve lesion, the lateral dorsum is insensitive only because of intracutaneous nerve terminal damage, the medial dorsum is insensitive because of both types of involvement, and the lateral

2092

Diseases of the Peripheral Nervous System

FIGURE 91–7 Lepromatous leprosy sensory map: continued progression of sensory loss with onset of paralysis.

palm may be spared. This complex pattern explains part of the difficulty in giving a constant sequence of sensory loss by modalities in leprosy. The most common and early obvious paralysis in lepromatous leprosy affects the ulnarinnervated intrinsic hand muscles, leading to weakness, atrophy, and an eventual claw hand deformity. Affection of a small peroneal branch at the ankle results in a less obvious paralysis of the extensor digitorum brevis, which occurs at about the same stage as ulnar involvement. Extension of the sensory loss of “less cool” areas continues in the untreated case of lepromatous leprosy. Corneal sensation is reduced. The face and ears become insensitive except at the creases of the nasolabial folds, the posterior attachment of the pinnae, and the perimeters of the alae nasi. When insensitivity of the forehead develops, there is an abrupt return to normal sensation at the hairline (see Fig. 91–7). This highly diagnostic “hairline” sign is an elegant demonstration of the intracutaneous nature of the sensory loss and is the result of an increase of 1.5° to 3° C in temperature under the scalp hair. Palpable enlargement of the supraorbital branch of the trigeminal nerve occurs in advanced cases of lepromatous leprosy. Painful neuralgia may occur. Even in these circumstances some sparing of

sensation under the scalp hair usually can be demonstrated. The scalp falsely appears to be a cool area on the Thermoscan of a normal individual in Figure 91–9 only because the hair blocks the infrared emission. The cooler areas of the face on this Thermoscan correlate with the most intense facial sensory loss shown in Figures 91–5 through 91–8. When facial sensory loss is widespread, a unique paralysis in the distribution of the seventh cranial nerve appears. Monrad-Krohn masterfully described the patchy bilateral paralysis that is due to selective affection of multiple small branches of the facial nerve.59 The branches are involved where they assume a superficial course. The initial dysfunction is usually weakness of eye closure resulting from damage to the branches coursing over the zygoma and orbital rim. Involvement of small twigs to the orbicularis oris muscle results in localized outpouching of the lips at the corners of the mouth (Fig. 91–10A). The medial segments of the frontalis muscle are paralyzed before the lateral. When such a patient attempts eyebrow elevation, only the lateral aspects rise, resulting in a characteristic V configuration (Fig. 91–10B). With progression of the facial nerve lesions, complete lagophthalmos and ectropion of the lip occur. The superficial muscles forming the nasolabial

2093

Leprosy

FIGURE 91–8 Thermoscan of the face of a normal subject. Warm areas appear light; cool areas appear dark. (Courtesy of AGA Thermovision, Secaucus, NJ.)

fold may be affected, and the tone of the intact deeper buccinator causes a series of concentric creases extending out from the corners of the mouth (buccinator smile). The paralysis in the upper half of the face is symmetrical, whereas asymmetry is often encountered in the lower half of the face. Total facial paralysis is not present even in the most advanced cases of lepromatous leprosy. These unique defects in the Vth and VIIth nerves are the only common cranial nerve signs seen in lepromatous leprosy. Sensory loss progresses in the upper extremities with completion of insensitivity in the palms. The antecubital fossae and the axillae uniformly retain sensation. The median nerve then becomes involved in a 5- to 7-cm segment proximal to the carpal tunnel. The nerve assumes a superficial position as it emerges from under the forearm flexor muscles and extends to the area of the carpal tunnel.78 Figure 91–2F and Plate 91–2F illustrate the demarcation of erythema and gross enlargement of the nerve at this point of emergence. The sensory loss becomes more profound in the palm, and the median nerve–innervated intrinsic muscles of the hand show atrophy and paralysis. An established paralysis of ulnar nerve–innervated hand intrinsics is present in lepromatous leprosy before the median deficits appear. Radial nerve motor deficits are less common in lepromatous leprosy. The segment that emerges from under the triceps muscle 4 to 6 cm proximal to the elbow is occasionally affected so that the triceps, brachioradialis, and anconeus

muscles remain unaffected while the extensors of the wrists and fingers develop atrophy and weakness. The lower extremities in the untreated case show a similar progression of neurologic deficits. The intracutaneous sensory loss spreads over the soles, leg, and thigh. Sparing may be detected between the toes, at the popliteal fossa, and in the proximal half of the medial thigh. The superficial segment of the peroneal nerve coursing laterally around the fibular head is infiltrated and enlarged. Footdrop is the second most frequent of the obvious paralytic deformities in leprosy. There is also affection of the segment of the posterior tibial nerve in the distal third of the leg, which results in paralysis of the intrinsic musculature of the volar surface of the foot and the completion of insensitivity over the soles. With extension of cutaneous sensory loss over the torso, the temperature-linked pattern of cutaneous sensory loss is more readily identified by the islands of preserved sensation. The inguinal creases, the perineum, the axillae, the sternal area, and a stripe of variable width up the center of the back from the intergluteal fold to the neck are areas that show consistent sparing. These warm areas coincide with the regions that are spared of dermatologic lesions in lepromatous leprosy.32 No paralysis of trunk or limb girdle musculature occurs in low-resistance leprosy. The typical fully developed pattern of paralysis does not interrupt the arcs for the usually elicited stretch reflexes. This is a most helpful sign in differentiating leprosy from other polyneuropathies. Autonomic fibers are damaged along with the sensory and motor fibers. Loss of sweating occurs in insensitive areas. The extremities become cool and dusky. Because the autonomic fibers that lie more deeply within the body are not involved, the postural hypotension, nocturnal diarrhea, abdominal crises, bladder disturbances, and impotence that characterize some other neuropathies and radiculopathies do not occur in leprosy. Despite the lack of clinical signs of central autonomic dysfunction in leprosy, several studies of autonomic function suggest abnormalities in neural regulation of cardiac and genital systems in borderline-lepromatous and lepromatous leprosy.51,65,66

Borderline Leprosy The neurologic syndromes of borderline leprosy are based on the changing roles of tissue temperature and host resistance in the pathogenesis of nerve dysfunction at various points along the spectrum between the polar forms of the disease. Temperature of the tissues is operative only in determining the possibility of a high-resistance lesion occurring at a given locus within the body. When the temperature is “permissive,” local factors supervene in molding the characteristics of the lesion. This is reflected in the vigorous epithelioid cell response, the paucity of bacilli, and the sharply circumscribed nature of the skin lesions and neural deficits. Furthermore, caseous necrosis of nerves (nerve abscesses) is far more common in high-resistance

2094

Diseases of the Peripheral Nervous System

FIGURE 91–9 Lepromatous leprosy sensory map: extension of sensory loss to warmer areas of skin.

than in low-resistance leprosy. In low-resistance leprosy, nerves containing numerous bacilli may continue functioning; the tissue immune response is not an immediate cause of nerve dysfunction. These factors explain the occurrence of certain atypical neurologic deficits in borderline leprosy. In a patient with borderline leprosy, bacilli may spread to a nerve site where the temperature is not especially favorable for proliferation but does not totally inhibit reproduction. These few bacilli may evoke a tissue response that is violent enough to result in dysfunction of the nerve. The invasion of leprosy bacilli into the same site in a patient with lepromatous leprosy would fail to evoke this destructive tissue reaction, and if the temperature were warm enough there would be insufficient bacillary proliferation to result in dysfunction of the nerve at that site. Although the nerve damage of borderline leprosy is largely limited to the sites described for lepromatous leprosy, the potential for production of unusual forms of paralysis is greater.13 Occasionally, total facial paralysis, brachial plexus involvement, and median nerve damage at the level of the elbow are found, and borderline cases have the greatest potential for devastation of the peripheral nervous system. These cases are characterized by a degree of host resistance

sufficient to result in prompt nerve dysfunction but tissue response inadequate to preclude hematogenous dissemination of the disease, and such patients show widespread nerve damage even early in the course of the disease and are the cases often reported as pure neural leprosy.43 The borderline near-tuberculoid (BT) case has several skin lesions that are larger and less distinctly demarcated than pure tuberculoid lesions. The lesions are asymmetrical, and the larger ones may have several satellites surrounding them. Insensitivity is detectable in all the lesions and is still coterminous with the skin lesions. The superficial mixed motor and sensory nerves coursing near the intracutaneous lesions are apt to be involved. These are generally the same as the named nerves affected by the polar forms of the disease, but the potential for atypical sites is present. The resulting paralyses are usually asymmetrical. In the midrange (BB) cases of borderline leprosy, large skin lesions with more rapidly advancing borders are found. The lesions are insensitive, but the border of insensitivity may not be precisely coterminous with the border of the skin lesion. Large lesions may coalesce, and indeed the insensitivity may occur over a large percentage of the body surface. The advancing borders pay little heed to smaller

2095

Leprosy

FIGURE 91–10 A, Patient attempting to close his eyes. The patchy nature of facial paralysis in lepromatous leprosy is indicated by the bilateral lagophthalmos and protrusion of the lateral part of the upper lip. There is also loss of the lateral eyebrows. B, Same patient attempting to raise his eyebrows. Paralysis of the medial forehead elevators preceded paralysis of the lateral elevators. (From Sabin, T. D.: Neurologic features of lepromatous leprosy. Am. Fam. Physician 4:84, 1971, with permission.)

temperature gradients, but the warmest body surfaces are spared. The pattern of sensory loss and paralysis is clearly asymmetrical. The sensory map indicates an intracutaneous nerve dysfunction, because the borders do not conform to the distribution of the peripheral nerves, the roots, or the distal symmetrical loss of the polyneuropathies. In low-resistance borderline (BL) cases, there may be numerous active lesions with a definite tendency toward symmetry. The insensitivity may “spill over” the perimeters of the skin lesions. Temperature-linked patterns of sensory loss emerge, but the boundaries of insensitivity are more distinct and the overall pattern less perfectly symmetrical than in lepromatous leprosy (Fig. 91–11). As noted, in contrast to the tendency toward stability in the polar forms of leprosy, in borderline leprosy the patient’s position along the intermediate spectrum may alter with time. General debilitation, prolonged administration of steroids, and lymphoma or leukemia may result in downgrading of the borderline case toward the lepromatous pole. Conversely, patients who have type 1 reactions or who have received appropriate antimicrobial therapy may develop features of higher resistance leprosy. The patient with borderline leprosy may thus present with neurologic deficits characteristic of both high- and

low-resistance disease. The estimation of the current status of this patient must then be based on knowledge of the temporal evolution of the patient’s deficits. The neurologist cognizant of these few principles may anticipate considerable success in placing a particular case along the borderline spectrum. The details of the neurologic examination will allow a diagnosis based on the TT, TB, BB, BL, LL classification of Ridley and Jopling.72

NEURITIS IN LEPROSY REACTIONS The slow evolution of neurologic deficits in uncomplicated progressive leprosy contrasts with the acute or subacute affection of peripheral nerves in leprosy reactions. The same peripheral nerves involved in reactional neuritis are involved in progressive leprosy. In ENL the manifestations of reaction, including neural involvement, are most intense in areas of greatest bacillary density. A nerve segment may become intensely painful and tender, and total paralysis may occur within hours. Leprosy reactions with nerve involvement must be recognized and treated promptly, for these deficits are often entirely reversible.60 The recovery of nerves rendered functionless by progressive leprosy is less certain.

2096

Diseases of the Peripheral Nervous System

FIGURE 91–11 Borderline leprosy sensory map.

NERVE COMPRESSION

DIFFERENTIAL DIAGNOSIS

The nerve is damaged not only by bacillary proliferation and immune responses, including the leprosy reactions, but also by compression. The superficial nerves are most affected, and their increased size is an additional risk factor. The ulnar nerve may be so enlarged that it bulges out of the cubital groove and is repeatedly traumatized. Chronic compression and trauma often result in progressive median nerve damage at the carpal tunnel and peroneal nerve dysfunction at the tendinous arch of origin of the peroneus longus muscle. Nerve release surgery should be considered for the leprosy patient with a progressive mononeuropathy that is not explained by untreated leprosy or leprosy reaction.

The intracutaneous sensory loss, enlarged nerves, and preserved reflexes occurring in the appropriate epidemiologic setting usually make the diagnosis evident once leprosy is considered.

Distal Polyneuropathy In the distal polyneuropathies the clinical sensory and motor deficits evolve according to the length of the nerves. The longest fibers show dysfunction first, and if the disease progresses, deficits become ever more proximal. The toxic, nutritional, metabolic, or genetic factors that produce these disorders are not local. The clinical

2097

Leprosy

deficits are therefore symmetrical, are most extreme distally, and tend to shade off to normal across a zone of partial dysfunction. In a sensory polyneuropathy the defects appear in the feet and then extend to the knees before sensory loss appears in the hands. When sensory loss spreads to the proximal arms and legs, the fibers traversing the body wall become involved, and a band of decreased sensation appears on the anterior torso.76 When this band widens sufficiently, loss of scalp sensation in the most distal distribution of the trigeminal nerve can be detected. This pattern can be ascertained at the bedside by estimating the location of the appropriate dorsal root ganglion cells and tape measuring to the proximal margin of sensory loss. The measurement will be the same for all four extremities, the torso, and the scalp. Motor polyneuropathies typically evolve with a similar relationship to nerve fiber length. Length of nerve fibers is thus critical in the determination of the distribution of clinical deficits, whether the mechanism of nerve damage is based on increasing vulnerability of axons with increasing distance from their cell bodies or the random scattering of numerous microscopic lesions throughout the peripheral nervous system. Sensory polyneuropathies and low-resistance leprosy may both produce extensive symmetrical sensory loss. The temperature-linked pattern of leprosy must be distinguished from the distance-linked pattern of the sensory polyneuropathies. The greater affection of pinprick and temperature sensation with relative preservation of position and vibration sensation in leprosy serves to eliminate certain sensory polyneuropathies. There is no pure motor syndrome in leprosy. The pattern of paralysis in sensorimotor polyneuropathy can also be distinguished from that in leprosy. The distribution of weakness in low-resistance leprosy can always be localized to major named peripheral nerves and does not evolve on the basis of nerve length. Paralysis of ulnar nerve–innervated hand intrinsic muscles is obvious before median nerve–innervated intrinsics are affected and is usually present before foot dorsiflexors or intrinsics are weakened. The sparing of the stretch reflexes in low-resistance leprosy contrasts with the loss of reflexes seen in most polyneuropathies. A suitably stained biopsy specimen of skin will show abundant bacilli; the sural nerve biopsy will usually show the typical picture of lepromatous leprosy. The polyradiculopathies are even easier to exclude because the generalized loss of deep tendon reflexes, weakness of both proximal and distal musculature, and tendency for the sensory loss to be greatest for position and vibratory modalities are all features not seen in leprosy. The spinal fluid protein in leprosy is normal, although abnormalities of the colloidal gold curve have been reported.50

Mononeuropathy and Mononeuropathy Multiplex Leprosy has been described as a mononeuropathy multiplex, and if one attends only to the pattern of paralysis this terminology appears correct.73 The motor examination alone may be inadequate to differentiate leprosy from this category of disease. The unique “patchy” paralysis of the face as well as the stereotyped location and sequence of motor nerve lesions are the most helpful diagnostic features of paralysis in leprosy. The mononeuropathy multiplex of diabetes and collagen-vascular disease frequently affects cranial and somatic nerves, which are never damaged by leprosy. Several of the sites of chronic compression of nerves are roughly similar to those of leprosy lesions. When the traumatized nerve becomes enlarged, confusion with leprosy is possible. A careful search for a nearby area of intracutaneous sensory loss that does not fall into the distribution of the nerve in question is the key. Such an area of sensory loss is nearly always present in leprosy. Sensory perineuritis affects subcutaneous sensory nerves, giving rise to paresthesias, pain, and partial sensory loss.2 The onset is frequently acute and there is often eventual recovery. Biopsy demonstrates an inflammatory response and nerve destruction that is limited to certain fascicles. High-resistance leprosy must be considered in the differential diagnosis. In sensory perineuritis there is no skin lesion; the sensory loss is partial and within the distribution of named subcutaneous nerves, and tends to improve but recur at a distant site. An occasional hysterical patient will present with a patch of sensory loss. These patients are sorted out by their tendency to have sharply circumscribed total loss of all modalities, including vibration, in the affected area and intact autonomic function.

Nerve Enlargement In leprosy, any nerve close to the surface of the body may become enlarged, hardened, and tender. After treatment the nerves may gradually shrink, but they remain abnormally firm even when they become smaller than normal. If the examiner detects peripheral nerve enlargement, the diagnostic possibilities are helpfully limited. The fusiform neuromegaly of leprosy is readily distinguished from the nodular enlargement of nerves in von Recklinghausen’s neurofibromatosis. The enlargement of a chronically traumatized nerve is more circumscribed than that of one with leprosy neuritis. For example, palpable enlargement of the leprous ulnar nerve extends 10 to 12 cm proximal to the olecranon groove area. The generalized neuropathies with nerve enlargement, such as amyloidosis, the hypertrophic neuropathies, and Refsum’s disease, are typically associated

2098

Diseases of the Peripheral Nervous System

with widespread areflexia, distal symmetrical clinical deficits, and often very elevated spinal fluid protein levels.

Absorption Insensitive but actively used limbs are liable to undergo progressive loss of tissue. The pathogenesis of absorption is not clear in every detail, but the observations of Brand strongly implicate repeated acceptance of tissue-damaging forces by the insensitive limb with repeated bouts of painless infection as the major factors.10,11 Absorption of the hands and feet in leprosy has given rise to the erroneous notion that parts of the body may “drop off” in this disease. As insensitivity of the hand deepens, frequent examinations will reveal a constant array of small wounds and injuries. The most common wound of the hand is a fingertip hematoma. After numerous such injuries the pulp of the fingertip becomes fibrotic and spindle shaped. The fingernail slopes downward toward the fingertip (Fig. 91–12). Unless recurrent injuries are prevented, the fingers will now lose length and after many years become small nubbins on a “mitten” hand. Even in grossly shortened fingers, radiographs often reveal that all the bones are still present, although tiny and severely distorted. The soft tissues and bone have been actively remodeled in response to repeated stress. This shortening is associated with the occurrence of open wounds and bouts of osteomyelitis. In the foot, calluses develop over high-pressure areas, and a hematoma at the base of the callus is the harbinger of a “trophic” ulcer. These ulcers usually occur under the metatarsal heads. When the longest metatarsal head becomes reabsorbed, an indolent ulcer appears under the next longest metatarsal head.

This insidious process is dramatically accelerated by invasion of pyogenic organisms through open wounds or cracks in the anhidrotic, friable corium of flexor creases. These painless infections result in cellulitis, lymphangitis, tenosynovitis, and osteomyelitis that seriously endanger the function of the hand or foot. Absorption is not primarily a result of nerve damage per se, because if the hand is rendered useless by complete paralysis, absorption does not occur (see Fig. 91–2G and H and Plate 91–2G and H).75 Leprosy may create ideal conditions for absorption. The long flexors and extensors of the wrist and fingers as well as the plantar flexors of the foot remain powerful even when the hands and feet are insensitive. The clawing of the fingers results in a pattern of use of the hand that leads to gripping with the tips rather than the pulps of the fingers. This produces a steep increase in tissue pressure at the tips for a given force of contraction of the finger flexors. The process of absorption is not halted by bacteriologic “cure” but can be prevented by avoiding recurrent injury to the insensitive hands and feet.10,11 Any neurologic disease that results in loss of pain sensation in the hands and feet without full paralysis may result in absorption. Evidence of this process must not be taken as a diagnostic sign of any specific disease; it may be seen in such diverse disorders as congenital insensitivity to pain, hereditary sensory neuropathy, syringomyelia, tabes dorsalis, diabetic neuropathy, and amyloid neuropathy.

ELECTROPHYSIOLOGIC TESTING Motor and sensory nerve conduction velocity studies have been performed in lepromatous, borderline, and tuberculoid leprosy. These studies have been valuable in confirming segmental slowing in the nerve trunks and branches at

FIGURE 91–12 Deformed insensitive hand. Note hematoma on tip of index finger and migration of nail on fourth finger.

Leprosy

sites of enlargement and have demonstrated improvement with treatment.60,81 Brand operated on segments affected by acute neuritis, usually in the ulnar nerve at the elbow. He found that electrical stimulation of the nerve above the involved segment produced no muscle response. Upon incision of the sheath, however, the nerve immediately began to conduct impulses (P. W. Brand, personal communication, 1969). Sheshkin and co-workers found slowing in 3 of 13 ulnar nerves in leprosy worsening during ENL.89 Following treatment of reaction with thalidomide, motor conduction velocity returned to normal in 10 days to 2 weeks. These findings have been confirmed.90 Examining 103 patients with leprosy, Magora and co-workers found conduction slowing in 24.3% of ulnar, 13% of median, 6.5% of radial, and 15.8% of peroneal nerves.56 Over long-term follow-up, slowing occurred during reaction and could be controlled with thalidomide. Patients who did not have reaction had little change in conduction times. In the study of Sohi and associates, it was noted that nerves in reaction showed a uniform decrease in conduction velocity.94

NEUROPATHOLOGY Although leprosy organisms have occasionally been reported from various areas of the human central nervous system, including Purkinje cells, anterior horn cells, and the cells of the Clarke column, actual neuropathologic changes in leprosy are limited to the peripheral nervous system. Hansen elucidated the common cause of nerve damage in all forms of leprosy when he described masses of bacilli within peripheral nerve.27 The gross and microscopic appearance of such nerve lesions depends on the type of leprosy present. Nerve damage is found in all cases of leprosy. The target of nerve invasion by M. leprae is the Schwann cell. Peripheral nerves are otherwise highly resistant to bacterial invasion, so recent contributions revealing how leprosy invades nerves are of special interest. A glycolipid designated as phenolic glycolipid-1 (PGL-1) is unique to M. leprae and binds to Schwann cell extracellular basal membrane protein, laminin-2, and its associated ␣-dystroglycan membrane constituent.67 The ␣-dystroglycan forms part of a transmembrane array that binds to intracellular actin and provides a potential route of entry for the bacilli. In cell culture systems M. leprae does not invade Schwann cells that are actively forming myelin. PGL-1 attachment to Schwann cell basal membrane induces demyelination.12,100 This in turn initiates dedifferentiation into a form of Schwann cell vulnerable to M. leprae invasion. The nonmyelinating Schwann cells associated with small autonomic and pain-temperature perception are readily invaded, and this correlates with the extensive pattern of intracutaneous nerve destruction that causes the loss of

2099 autonomic function and pain and temperature perception. These earliest stages of nerve damage occur in M. leprae– infected immunodeficient RAG-1 mice, indicating that early Schwann cell injury is not dependent on immune mechanisms that play an important role in nerve destruction later in the disease.68 The precise manner in which intraneural M. leprae causes nerve damage is not well understood, but certain facts are beginning to emerge that challenge the long-held beliefs that in leprosy there is an “ascending neuritis” and that intraneural spread occurs through axons. Dehio proposed the concept that bacilli from an infected area of skin gained access to cutaneous nerve endings and from there traveled proximally through larger and larger nerves to reach the main trunk.21 On the basis of a meticulous dissection of an entire leprous ulnar nerve by Gerlach,24 Dehio suggested that, when the involvement progressed proximally, both motor and sensory fibers would be involved and paralysis with additional sensory loss would occur peripheral to the original site. This view was expanded by Khanolkar, who on the basis of light microscopic studies suggested that bacilli traveled intra-axonally “like fish migrating upstream.”46,47 Electron microscopy demonstrates that the overwhelming majority of endoneurial bacilli are within Schwann cell cytoplasm.42 Schwann cell involvement leads to demyelination of internodal segments, and segmental demyelination is probably responsible for the slowing of nerve conduction velocity that occurs in diseased segments.96 Remyelination of such diseased internodes results in the formation of thinly myelinated and short internodal segments adjacent to normal ones (Fig. 91–13). A process may be hypothesized in which bacilli are delivered to the endoneurium via the blood and are taken up by Schwann cells (Fig. 91–14), which eventually lose their ability to maintain the myelin sheath. Schwann cell disruption occurs, releasing into the endoneurium bacilli that can then be taken up by additional Schwann cells. This process may occur longitudinally in nerve fascicles and result in heavy involvement in certain fascicles with relative sparing of others. Demyelination and Schwann cell proliferation may lead to onion bulb formation.25,37 Overall, however, there is a loss of myelinated fibers per square millimeter of fascicular area, and larger fibers are predominantly affected. An interesting association exists between neurofibromatosis and leprosy. The cells making up the tumor in neurofibromas are believed to derive from Schwann cells. Two patients with both diseases have been reported. In these cases M. leprae has been present within tumor cells making up the neurofibromas and has appeared to be concentrated in these tumors.95 Ultrastructural studies of leprosy peripheral nerve reveal a few intra-axonal bacilli, but these constitute only a small percentage of all endoneurial bacilli.6,7,42,107 The role

2100

Diseases of the Peripheral Nervous System

FIGURE 91–13 Ten consecutive portions along length of one teased fiber of sural nerve in a 33-year-old male patient with lepromatous leprosy. A short and thinly myelinated internodal segment and a short segment of thin myelin that appeared fragmented are present. The presence of internodal segments of varying lengths and thickness on a single fiber is consistent with segmental demyelination and remyelination. (Osmium, ⫻445.)

of intra-axonal bacilli in the pathogenesis of leprosy is unclear. Intra-axonal bacilli have been demonstrated in M. leprae–infected armadillo nerves.54

Tuberculoid Leprosy Cutaneous nerves and nerve trunks may be enlarged in the proximity of a skin lesion, as originally noted by Virchow. The enlargement is quite firm and nodular and may affect only the segment close to the surface of the skin lesions, to which it may be adherent. Microscopically the greatest change occurs within cutaneous and subcutaneous nerves in the area of the skin lesion. Most often these structures are invaded and destroyed beyond recognition by an epithelioid and giant cell granuloma. Both Langerhans and foreign body giant cells are found, and lymphocytes are scattered in pockets throughout (see Fig. 91–2C and Plate 91–2C). A few nerves may still be seen, with thickening and infiltration of the perineurium and often extension of the granuloma into the endoneurium. Axon stains of these areas show few or no remaining fibers. Involvement of

large nerves and nerve trunks may occur by spread of the granuloma through the subcutaneous tissue or along large nerve branches, in response to invasion of bacilli into the nerve. The histologic appearance of the large nerves is similar to that of smaller nerves. The epineurium and perineurium are thickened and contain epithelioid cells, which may breach the inner layer and spread into the nerve. Intraneural collagen is greatly increased and may appear to wall off certain fascicles. Some of the fascicles are totally destroyed, whereas others appear to be almost normal. Caseation may occur and produce large abscesses within the nerve. With healing, the granuloma subsides. In late stages such nerves show fibrosis and hyalinization of the endoneurium, retaining only a few recognizable normal elements and thick perineural and epineurial sheaths. Bacilli may be difficult, or frequently impossible, to find throughout this entire process. Occasionally one must resort to excising a large block of tissue or even the entire skin lesion for serial sectioning before one or two bacilli can be found. When bacilli are found, they are almost always in nerve.

2101

Leprosy

FIGURE 91–14 Osmium-stained nerve sections in lepromatous and borderline leprosy. A, Longitudinal section of sural nerve in lepromatous leprosy. Vacuoles containing many degenerating bacilli in endoneurium. (⫻830.) B, Cross section of same nerve as in A (⫻216). C, Lepromatous leprosy. Bacilli in Schwann cell surrounding myelinated fiber. (⫻2080.) D, Borderline leprosy. Epithelioid granuloma extending into endoneurium of sural nerve. Note perineurium at top. (⫻216.) E, Dimorphous leprosy. A group of small myelinated regenerated fibers in sural nerve. Note circumferentially oriented Schwann cell nuclei. (⫻2080.)

Lepromatous Leprosy In lepromatous leprosy bacilli are spread hematogenously and appear in nearly every tissue and organ.63 Gross enlargement of certain nerve trunks occurs, usually as a

firm, fusiform enlargement of specific segments, frequently with grossly normal appearance proximal and distal to the diseased segment.38,39 The nerve in an involved segment, such as the ulnar nerve just proximal to the elbow or the median nerve proximal to the wrist, reveals

2102

Diseases of the Peripheral Nervous System

increased numbers and size of blood vessels on its epineurial surface, and it may appear brownish or erythematous (see Fig. 91–2F and Plate 91–2F). Microscopically, the nerve endings in skin as well as the nerve trunks usually show greater preservation of the overall architecture than in high-resistance leprosy. In contrast to the situation in tuberculoid leprosy, many nerve endings usually may still be identified in the skin biopsy from an involved site. The perineurium may be split into layers by sheets of foamy cells or clear areas representing edema. This leads to an “onionskin” appearance of such nerves (see Fig. 91–2I and Plate 91–2I). These nerve endings are surrounded by the lepromatous infiltrate consisting of foamy histiocytes filled with masses of bacilli. This infiltrate also involves the perineurium but usually does not breach this structure. There may be increased numbers of mast and plasma cells in the endoneurium. The endoneurial cells appear more vacuolated and increased numbers of nuclei are present, but the architecture is otherwise preserved. Acidfast stain reveals great numbers of organisms within endoneurium as well as perineurium (see Fig. 91–14A and B). Globi may be present in both structures. Masses of organisms are seen in a space between perineurium and endoneurium and may fully encircle the latter. When viewed in longitudinal sections, the bacilli appear parallel to the long axis of the nerve in Schwann cells (see Fig. 91–14C). Many can be seen in endothelium of endoneurial blood vessels and, occasionally, in cells within the lumen of such vessels. Abnormalities in endothelial cell structure occur, with resulting breaches in the blood-nerve barrier.7 Even in long-standing cases of lepromatous disease, the fascicular structure of nerve trunks remains discernible. The epineurium is now infiltrated by huge numbers of foamy histiocytes, especially around blood vessels, and bacilli may be found through all elements of the nerve trunk. With time there is replacement of normal endoneurial elements by collagen, and some contraction of the endoneurial area occurs. Light and electron microscopic quantitative studies of lepromatous nerves are rare. One sural nerve studied in this way had histograms compatible with a combination of wallerian degeneration and segmental demyelination but with evidence of regenerative activity of both myelinated and unmyelinated fibers.25 Bacilli may persist for long periods within nerve trunks, even when treatment has resulted in the eradication of bacilli from skin sites. In some patients whose disease was considered inactive, evidence of progression of nerve lesions has been found, and in such patients nerve biopsy may reveal the presence of bacilli. Even after 20 years of treatment in lepromatous leprosy in the era of dapsone monotherapy, a high antigen load (using an anti-BCG antibody technique) has been demonstrated on nerve biopsy material.3 An interesting feature of nerve trunk involvement in lepromatous leprosy is the uneven involvement of fascicles by bacilli.

Borderline Leprosy In a borderline leprosy case the exact admixture of neuropathologic features of the polar forms depends on the degree of resistance to the bacillus in the individual case. Patients with high resistance have predominantly tuberculoid features; that is, nerve trunk involvement is asymmetrical, early and severe destruction of neural elements by an epithelioid granuloma occurs in both skin lesions and nerve trunks, and few bacilli are present (see Fig. 91–14D). In contrast to lepromatous leprosy, early and severe axonal damage may occur, and evidence of regeneration of axons also may be seen (see Fig. 91–14E). Patients with little resistance to the bacillus have predominantly lepromatous features, that is, symmetrical involvement of nerve networks and nerve trunks in all cool areas; relatively good preservation of nerve architecture, which may show infiltration of foamy histiocytes; and great numbers of bacilli present. In some patients, both lepromatous and tuberculoid changes may be found in different areas of a single nerve or even in adjacent areas. Varieties of leprosy referred to as “polyneuritic” or “pure neural” usually are histologically borderline. This is an interesting form of leprosy, imperfectly understood, in which multiple nerve trunk involvement occurs in the absence of obvious skin lesions.33,99 Because of the enlargement of nerves as well as the giant amount of intraneural collagen present, many of these patients are mistakenly believed to have another form of hypertrophic neuropathy. The pathologic diagnosis in such cases depends on the demonstration of acid-fast organisms within the nerve. In contrast to polar tuberculoid leprosy, hematogenous dissemination probably occurs in borderline disease.

Leprosy Reactions in Nerves Reaction plays an important role in the evolution of nerve trunk damage in leprosy.44 Acute nerve trunk lesions are not seen in downgrading reactions. However, reversal reactions in borderline or tuberculoid cases may severely damage nerve trunks. In these forms of reaction, new lesions may appear and preexisting lesions show an acute inflammatory process and become intensely indurated and swollen. As the inflammatory process subsides, involved nerve trunks show replacement of endoneurial elements by epithelioid granuloma and fibrosis. In ENL arteritic lesions occur within cutaneous nerves and nerve trunks, and foci of edema and acutely inflammatory cells may be seen surrounding small arteries and arterioles showing fibrinoid degeneration in epineurium, perineurium, and endoneurium. They are particularly prone to occur where vessels pierce the inner layer of perineurium. The polymorphonuclear infiltration, acid pH, lysosomal activity, vascular engorgement, and edema in such lesions lead to severe destruction of neural elements.

Leprosy

This process occurs within the confines of a relatively inelastic perineurium or epineurium and may lead to increase in pressure and resulting ischemia. This ischemia may lead to a block of conduction through the involved segments. Foci of acutely inflammatory cells surrounding arteritic lesions may coalesce and produce microscopic or, very rarely, gross abscesses within the nerve.22

Muscle Involvement Involvement of muscle in leprosy is rare, but examples of leprous myositis are recorded in the literature.29,41,57,81 Mycobacterium leprae is present within skeletal muscle cells, and there is muscle fiber necrosis with an inflammatory cell reaction and evidence of regeneration. In almost every case coincident denervation atrophy of muscle fibers is present.

TREATMENT Specific antileprosy treatment prevents or arrests progressive nerve damage and promotes healing of other lesions. Leprosy is, however, a complex chronic disease, and optimum management demands a comprehensive program that may require the skills of internists, orthopedic surgeons, ophthalmologists, and public health officials as well as a rehabilitation team of therapists, social workers, and/or vocational counselors. Programs optimistically aimed at the eradication of leprosy that were based solely on case finding and the administration of sulfones have failed. Patients may not be mindful of improvement in the BI but will certainly note that social dislocation, plantar ulcers, or any other complications they may have or develop are not directly helped by their regimen. Indeed, the occurrence of disabling leprosy reactions following the initiation of therapy may cause them to conclude that the treatment is actually harmful if they have not had appropriate health education regarding the disease.

Antibacterial Treatment Chaulmoogra oil was the first agent widely used in the treatment of leprosy. Response was inconsistent, however, and when the clearly superior activity of the sulfones against M. leprae was demonstrated at Carville in 1941,23 its use was gradually abandoned. Sulfone monotherapy, usually utilizing the parent compound dapsone, became the treatment of choice and remained so until the 1970s. At that time the rising tide of secondary and primary dapsone resistance led to the realization that combination drug therapy was as important for leprosy as for tuberculosis if drug resistance was to be avoided. Today dapsone, clofazimine (B 663, Lamprene), and rifampin are the most widely used drugs, although

2103 ofloxacin (and some but not all other fluoroquinolones), minocycline, clarithromycin, streptomycin, and some thioamides (ethionamide and prothionamide) are also highly active against M. leprae. Streptomycin must be given by injection, and it and to a lesser extent the thioamides are generally more toxic than the others and are thus generally not utilized nowadays. Side effects are not a large problem with any of the other antibacterials. Hemolysis is common with dapsone, but the anemia, if any, is generally mild and seldom necessitates its discontinuation. Agranulocytosis and serious skin eruptions are fortunately rare, as is the occasionally reported peripheral neuropathy, which is usually associated with larger doses than those used to treat leprosy.79,80 Clofazimine pigments the involved skin, resulting in a red to black coloration, and may produce gastrointestinal discomfort. Dapsone and clofazimine are weakly bactericidal against M. leprae, but their combination is highly bactericidal.35 Rifampin has a very powerful bactericidal effect. Its major side effect is hepatotoxicity (uncommon) and it also accelerates the clearance of other drugs such as prednisone, dilantin, and oral contraceptives. Ofloxacin, minocycline, and clarithromycin are also highly bactericidal against M. leprae but to a lesser extent than rifampin. In the United States the treatment recommended for paucibacillary patients is dapsone 100 mg daily plus rifampin 600 mg daily for 1 year. Multibacillary patients are given the same combination for 2 years combined with 50 mg of clofazimine daily. In 1982 a WHO Study Group recommended a radical new approach to leprosy chemotherapy.104 The U.S. regimen outlined above is an approach developed within the last decade, and it is based on the WHO recommendations, which in fact have been somewhat modified in the last two decades. The WHO currently recommends that PB patients receive 100 mg of dapsone daily, unsupervised, with rifampin 600 mg once monthly, supervised, for 6 months. MB patients are given the same combination for 1 year combined with 300 mg of clofazimine once monthly supervised and 50 mg daily unsupervised.1,106 Long-term results indicate an overall relapse rate of around 1% for both PB and MB disease16 when the MB regimen is given for 2 years, but long-term results are not available as yet on those receiving the 1-year MB regimen. Currently a trial of 6 months’ therapy for MB patients is being developed.105 If relapse does occur with any of these regimens, the patients are retreated with the same regimen because resistance of M. leprae to any of the drugs utilized has not been demonstrated in patients who have relapsed. Singlelesion PB cases have been treated with a regimen consisting of single doses of 600 mg rifampin plus 400 mg ofloxacin plus 100 mg of minocycline (ROM regimen). The results thus far have been good. Treatment with any of the foregoing regimens renders the patient noninfectious within days. This, plus the fact

2104

Diseases of the Peripheral Nervous System

that the overwhelming majority of people are not susceptible to leprosy, means that isolation of patients is not required either initially or when they are hospitalized for other problems. Experimental approaches to treatment have included various forms of immunotherapy, such as interferon-␣26,61 and transfer factor31; some success has been reported, but results to date are not sufficiently promising to merit their widespread use. Vaccines to prevent the disease are also under study, but field trials to evaluate them will not be completed for several years.

Treatment of Leprosy Reactions Reactions constitute a grave threat to the patient and must be promptly and effectively treated. Corticosteroids, thalidomide, and clofazimine are the major antireaction drugs. Type 1 reactions accompanied by neuritis or the danger of skin ulcerations must always be treated with corticosteroids. Patients are typically given 60 to 80 mg of prednisone per day. This should rapidly control the reactive process, but if it does not, higher doses can be used and occasionally the drug may have to be given more than once daily. Although symptoms may rapidly clear in paucibacillary patients, if a neuritis is present, prolonged (several months’) therapy may be necessary to recover any sensory or motor loss that has occurred. Patients with BB or BL disease sometimes develop a chronic type 1 reaction requiring a year or more of therapy. When this occurs an attempt should be made to change the prednisone to an alternate-day schedule to minimize side effects, and periodically attempts are made to taper the patient off it entirely. When using corticosteroids for prolonged intervals, adjusting the dose may be easier if the rifampin is changed to 600 mg once monthly as in the WHO regimens to avoid the adverse effects of rifampin on corticosteroid metabolism. Also, if reactions occur after the patient has completed treatment and corticosteroids must be given for more than 4 months, it is recommended that 50 mg of clofazimine daily be given until the corticosteroids are discontinued.16 Type 2 reactions (ENL) may require only symptomatic measures if they are mild. The treatment of choice for more severe episodes is thalidomide. A dose of 300 to 400 mg/day usually clears the reaction within 36 to 48 hours. Once the reaction is controlled, the dose is tapered to an average maintenance level of about 100 mg daily over a period of about 2 weeks. Many patients require only a few weeks of therapy, but frequent recurrences necessitate prolonged treatment (months to years) for others. The mechanism of action of thalidomide is unknown, and aside from its known teratogenic effects, side effects are minimal. Long-term administration of thalidomide may result in a generalized neuropathy.15 Thalidomide is a Food and Drug Administration–approved drug for the treatment of

ENL in the United States, but because of its serious side effects, multiple precautionary measures are required before a prescription can be filled and no more than a 4 weeks’ supply may be given to a patient each time. Detailed information on the drug, the various restrictions, and prescribing information are given in the Physicians’ Desk Reference97 and may also be obtained from the National Hansen’s Disease Center in Baton Rouge, LA (1-800-642-2477). Corticosteroids will also rapidly control severe ENL and should be given if a neuritis accompanies it. Clofazimine, in addition to being antibacterial, is effective in slowly controlling ENL using a dose of 200 to 300 mg daily initially. The dose is tapered to 100 mg daily after the reaction is controlled. Chloroquine (250 mg twice daily) and stibophen (Fuadin) are occasionally used to control reactions of moderate severity. Neuritis may occur independent of other signs of reaction and is treated as under type 1 reactions with prednisone. Splinting the involved extremity may also occasionally be useful. Intraneural injection of corticosteroids with local anesthetics and hyaluronidase has been advocated for acutely swollen and tender segments of nerves, such as the ulnar nerve at the elbow.98 These measures have never been shown to have long-term benefit greater than that produced by systemically administered corticosteroids. The authors have observed almost complete permanent ulnar paralysis following such an injection. Iridocyclitis is treated with steroid eyedrops. Atropine and other mydriatics should also be instilled to prevent miosis, iris fixation, and secondary glaucoma. Surgical intervention may become necessary to relieve intractable glaucoma.

Surgical Treatment Surgical treatment of acute nerve abscess consists of carefully incising the nerve sheath and draining the abscess. Surgical neurolysis and even fascicular dissection have been advocated to relieve intraneural pressure, but longterm results of these procedures are unknown.62 Transposition of the ulnar nerve to the anterior aspect of the elbow and burial beneath muscle may prevent or retard further progression of damage to the segment by relieving the effects of trauma, and by increasing its temperature, multiplication of organisms may be inhibited. This procedure in some cases undoubtedly relieves nerve pain, but it has not been systematically assessed.84 Neurovascular island pedicle transfers may restore some degree of protective sensation to fingers anesthetic from leprosy.69 McLeod and co-workers have reported improvement in sensation in 17 of 23 nerve grafts inserted in leprosy patients who had localized nerve lesions.58 Six grafts resulted in no subjective or objective sensory return. No

2105

Leprosy

return of motor function was observed in any patients, but several had a return of sweating function. In only nine patients was sensory function sufficient to provide any degree of protection to previously insensitive areas. This study suggests that some localized nerve lesions in certain patients may respond to nerve grafting. Surgical intervention is frequently of value in leprosy for cosmesis and to overcome the effects of paralysis. Cosmetic operations include excision of redundant skin in earlobes and eyelids, excision of excessive breast tissue in patients with gynecomastia, and implantation of islands of scalp hair to replace lost eyebrows. Nasal reconstruction to correct the effects of nasal septum destruction may be the most valuable procedure to improve facial appearance. Commonly used surgical procedures to improve motor function in cases with paralysis consist of tenodesis to stabilize joints, arthrodesis to correct clawing, and tendon transfer to replace paralytic muscles with functioning ones, especially to restore dorsiflexion to the foot, abductionopponens action of the thumb, and extension of the proximal interphalangeal joint and flexion at the metacarpophalangeal joint.8,11 An ingenious method has been found to correct lagophthalmos. A slip of temporalis muscle attached to tendon is tunneled through the lids and attached to the inner canthus. Reeducation involves closing the jaws to effect eye closure.

soft by soaking hands and feet in water morning and night and applying an occlusive cream or ointment afterward to lock in moisture. Once wounds have occurred in insensitive hands and feet, treatment is often prolonged and difficult. Common problems include ulceration of the plantar surfaces and sides of the feet and toes and on the hands. Lacerations, burns, abrasions, and hematomas are quite common on the hands. Shortening of digits on feet and hands may occur through destructive osteomyelitis, and fragments of bone may be discharged through ulcerated areas. Amputations may occur traumatically or may be carried out surgically in an effort to remove hopelessly diseased tissue. Treatment of plantar ulceration may be particularly difficult. If active infection is present with fever, enlarged tender lymph nodes in the groin, and purulent drainage from the ulcer, the patient should be put to bed with the foot elevated and at rest. Antibiotics and mild warm soaking of the foot should be instituted. After the acute infection has subsided and swelling is reduced, the leg and foot should be placed in a total-contact plaster cast with a bowler’s iron to permit ambulation. Healing will occur over several weeks to several months in plaster, after which suitable footwear must be obtained. If proper footwear is not obtained and regularly used by the patient, ulceration will recur.

Management of Insensitivity Many of the deformities seen in patients with leprosy and thought to be characteristic or pathognomonic of the disease in fact are due to insensitivity and are common to many other diseases in which some degree of motor power is retained in insensitive limbs. Most of the damage to insensitive tissues is preventable, and insensitive patients must be made to understand this. Contrary to popular opinion, wounds in insensitive tissues may heal as promptly as those in normal tissues. The physician must provide some of the same protections that pain demands in normal individuals. Every wound must be splinted and protected from recurrent injury. Efforts must be made to ensure that the patient inspects eyes, hands, and feet daily for evidence of mild changes that, if left untreated or continually aggravated, might result in disability. Suitable, nondamaging occupations must be found for such patients, and modifications of tools and everyday items such as door keys must be made in order to remove sources of high pressure and shear. Gloves should be worn for potentially damaging routines. Footwear must have softness and total contact and must avoid shearing and direct forces. For insensitive and dry corneas, goggles should be worn to prevent both drying and introduction of foreign bodies; in many cases tarsorrhaphy is of benefit. When sympathetic denervation exists in hands and feet, the brittle keratin that results from anhidrosis may be kept

Social Rehabilitation Few diseases present the spectrum of chronic disabilities of leprosy. Many organs and tissues are involved, and deforming lesions of face, eyes, nose, ears, and extremities are often quite apparent even to the laity. With early treatment many potential deformities may be prevented, but even then public attitudes toward the leprosy patient change slowly. The physician responsible for primary care of such patients must not overlook this important aspect of management. Patients must be fully informed about the disease, its bacterial origin, the low order of communicability, and possibilities of successful treatment. Family, friends, and employers must be made aware of these facts if acceptance of the patient as an individual is to be accomplished. When insensitivity or crippling damage to eyes, hands, and feet is present, consultation with an expert in vocational evaluation and placement is essential. Many patients with leprosy are in the lower social and economic groups, and even before disease occurred, many have had only a marginal life adjustment. The advent of this potentially crippling and stigmatizing disease may present insurmountable problems for the patient to cope with alone. Social service is of immeasurable help to such patients in achieving sources of financial support, home and community health services, and adequate housing.

2106

Diseases of the Peripheral Nervous System

REFERENCES 1. Action Programme for the Elimination of Leprosy: A Guide to Eliminating Leprosy as a Public Health Problem, 2nd ed. Geneva, World Health Organization, 1997. 2. Asbury, A. K., Picard, E. H., and Baringer, J. R.: Sensory perineuritis. Arch. Neurol. 26:302, 1972. 3. Barros, V., Shelty, V. P., and Antia, N. H.: Demonstration of mycobacterial antigens in nerves of tuberculoid leprosy. Acta Neuropathol. (Berl.) 73:387, 1987. 4. Baskin, G. B., Gormus, B. J., Martin, N. L., et al.: Experimental leprosy in a rhesus monkey: necropsy findings. Int. J. Lepr. Other Mycobact. Dis. 55:109, 1987. 5. Binford, C. H.: Comprehensive programs for inoculation of human leprosy into laboratory animals. Public Health Rep. 71:995, 1956. 6. Boddingius, J.: The occurrence of Mycobacterium leprae within axons of peripheral nerves. Acta Neuropathol. (Berl.) 27:237, 1974. 7. Boddingius, J.: Ultrastructural and histophysiological studies on the blood-nerve barrier and perineurial barrier in leprosy neuropathy. Acta Neuropathol. (Berl.) 64:282, 1984. 8. Brand, P. W.: Treatment of leprosy: the role of surgery. N. Engl. J. Med. 254:64, 1956. 9. Brand, P. W.: Temperature variation in leprosy deformity. Int. J. Lepr. 27:1, 1959. 10. Brand, P. W.: Deformity in leprosy. In Cochrane, R. G., and Davey, T. F. (eds.): Leprosy in Theory and Practice, 2nd ed. Bristol, UK, John Wright & Sons/Baltimore, Williams & Wilkins, p. 447, 1964. 11. Brand, P. W., and Ebner, J. D.: Pressure sensitive devices for denervated hands and feet. J. Bone Joint Surg. Am. 51:109, 1969. 12. Brophy, P. J.: Subversion of Schwann cells and the leper’s bell. Science 296:862, 1998. 13. Browne, S. G.: Some less common neurological findings in leprosy. J. Neurol. Sci. 2:253, 1965. 14. Carpenter, C. M., and Miller, J. N.: The bacteriology of leprosy. In Cochrane, R. G., and Davey, T. F. (eds.): Leprosy in Theory and Practice, 2nd ed. Bristol, UK, John Wright & Sons/Baltimore, Williams & Wilkins, p. 13, 1964. 15. Chaudhry, V., Cornblath, D. R., Corse, A., et al.: Thalidomide-induced neuropathy. Neurology 59:1872, 2002. 16. Chemotherapy of leprosy: report of a WHO Study Group. World Health Organ Tech Rep Ser 847:1, 1994. 17. Cochrane, R. G.: The history of leprosy and its spread throughout the world. In Cochrane, R. G., and Davey, T. F. (eds.): Leprosy in Theory and Practice, 2nd ed. Bristol, UK, John Wright & Sons/Baltimore, Williams & Wilkins, p. 1, 1964. 18. Cole, S. T., Eiglmeier, K., Parkhill, J., et al.: Massive gene decay in the leprosy bacillus. L. Nature 409:1007, 2001. 19. Convit, J., Aranzazu, N., Ulrich, M. et al.: Immunotherapy with a mixture of Mycobacterium leprae and BCG in different forms of leprosy and in Mitsuda-negative contacts. Int. J. Lepr. Other Mycobact. Dis. 50:415, 1982. 20. Cooper, C. L., Mueller, C., Sinchaisri, T., et al.: Analysis of naturally occurring delayed-type hypersensitivity reactions in leprosy by in situ hybridization. J. Exp. Med. 169:1565, 1989.

21. Dehio, K.: On the “Lepra Anesthetica.” In Proceedings of the International Scientific Leprosy Conference, Berlin, October 1897. (Reprinted in part in Lepr. India 1–2[24]:78, 1952.) 22. Enna, C. D., and Brand, P. W.: Peripheral nerve abscess in leprosy. Lepr. Rev. 41:175, 1970. 23. Faget, G. H., Pogge, R. C., Johansen, F. A., et al.: The Promin treatment of leprosy, a progress report. Public Health Rep. 38:1729, 1943. 24. Gerlach, W.: Die Beziehungen Zwischen Hauttlecken und der Nervenkrakung bei der lepra anesthetica. Virchows Arch. 25:126, 1891. 25. Gibbels, E., Henke, U., Klingmuller, G., and Haupt, W. F.: Myelinated and unmyelinated fibers in sural nerve biopsy of a case with lepromatous leprosy—a quantitative approach. Int. J. Lepr. Other Mycobact. Dis. 55:333, 1987. 26. Gupte, M. D.: South India immunoprophylaxis trial against leprosy: relevance of findings in the context of leprosy trends. Int. J. Lepr. Other Mycobact. Dis. 69(Suppl.):S10, 2001. 27. Hansen, G. A.: Under sogeiser angaende spedalskhedens arasger tiedels ud forte sammem med forstander Hartwig. Norske Mag. Laegevidensk. 4(Suppl.):1, 1874. 28. Harboe, M.: The immunology of leprosy. In Hastings, R. C. (ed.): Leprosy. Edinburgh, Churchill Livingstone, p. 53, 1985. 29. Harmon, D. J.: Mycobacterium leprae in muscle. Lepr. Rev. 39:197, 1968. 30. Harrell, J. D.: Ocular leprosy in the Canal Zone. Int. J. Lepr. Other Mycobact. Dis. 45:56, 1977. 31. Hastings, R. C.: Transfer factor as a probe of the immune defect in lepromatous leprosy. Int. J. Lepr. Other Mycobact. Dis. 45:281, 1977. 32. Hastings, R. C., Brand P. W., Mansfield, R. E., and Ebner, J. D.: Bacterial density in the skin in lepromatous leprosy as related to temperature. Lepr. Rev. 39:71, 1968. 33. Jacob, M., and Mathai, R.: Diagnostic efficacy of cutaneous nerve biopsy in primary neuritic leprosy. Int. J. Lepr. Other Mycobact. Dis. 56:56, 1988. 34. Jacobson, R. R., and Trautman, J. R.: The treatment of leprosy with the sulfones. I. Faget’s original 22 patients: a thirty year follow-up on sulfone therapy for leprosy. Int. J. Lepr. Other Mycobact. Dis. 39:726, 1971. 35. Ji, B., Perani, E. G., Petinom, C., and Grosset, J. H.: Bactericidal activities of combinations of new drugs in nude mice. Antimicrob. Agents Chemother. 40:393, 1996. 36. Job, C. K.: Gynecomastia and leprous orchitis. Int. J. Lepr. 29:423, 1961. 37. Job, C. K.: Mycobacterium leprae in nerve lesions in lepromatous leprosy: an electron microscopic study. Arch. Pathol. 89:195, 1970. 38. Job, C. K.: Pathology of peripheral nerve lesions in lepromatous leprosy: a light and electron microscopic study. Int. J. Lepr. Other Mycobact. Dis. 39:251, 1973. 39. Job, C. K., and Desikan, K. V.: Pathologic changes and their distribution in peripheral nerves in lepromatous leprosy. Int. J. Lepr. Other Mycobact. Dis. 36:257, 1968. 40. Job, C. K., Jayakumar, J., and Ashhoff, M.: “Large numbers” of Mycobacterium leprae are discharged from the intact skin of lepromatous patients: a preliminary report. Int. J. Lepr. Other Mycobact. Dis. 67:164, l999.

Leprosy 41. Job, C. K., Karot, A. B., Varat, S., and Mathau, M.: Leprous myositis: a histopathological and electron microscope study. Lepr. Rev. 40:9, 1969. 42. Job, C. K., and Veghese, R.: Electron microscopic demonstration of Mycobacterium leprae in axons. Lepr. Rev. 45:235, 1974. 43. Jopling, W. H., and Morgan-Hughes, J. A.: Pure neural tuberculoid leprosy. Br. Med. J. 2:788, 1965. 44. Karat, A. B. A., Furness, M. A., Karat, S., and Rao, P. S. S.: Patterns of neurological involvement in relation to chronic and/or recurrent erythema nodosum leprosum. Lepr. Rev. 40:49, 1969. 45. Kellersberger, E. R.: The social stigma of leprosy. Ann. N. Y. Acad. Sci. 34:126, 1951. 46. Khanolkar, V. R.: Studies in the Histology of Early Lesions in Leprosy (Special Reports Series No. 19). New Delhi, Indian Council of Medical Research, 1951. 47. Khanolkar, V. R.: Pathology in leprosy. In Cochrane, R. G., and Davey, T. F. (eds.): Leprosy in Theory and Practice, 2nd ed. Bristol, UK, John Wright & Sons/Baltimore, Williams & Wilkins, p. 127, 1964. 48. Kirchheimer, W. F., and Storrs, E. E.: Attempts to establish the armadillo (Dasypus novemcinctus) as a model for the study of leprosy. L. Report of lepromatoid leprosy in an experimentally infected armadillo. Int. J. Lepr. Other Mycobact. Dis. 39:693, 1971. 49. Kluth, F. C.: Leprosy in Texas: a study of occurrence. Texas State J. Med. 51:199, 1955. 50. Koenigstein, R. P., and Chaug, S. T.: Colloidal gold reactions in the spinal fluid in leprosy. Int. J. Lepr. 19:323, 1951. 51. Kyriakidis, M. D., Noutsis, C. G., Robinson-Kyriakidis, C. A., et al.: Autonomic neuropathy in leprosy. Int. J. Lepr. Other Mycobact. Dis. 51:331, 1983. 52. Leprosy: global situation. Wkly Epidemiol Rec 77:1, 2002. 53. Lichtman, D. M., Swafford, A. W., and Kerr, D. M.: Calcified abscess in the ulnar nerve in a patient with leprosy. J. Bone Joint Surg. Am. 61:620, 1979. 54. Liu, T. C., Ji, Z. M., and Skinsnes, O.: Light and electronmicroscopic study of M. leprae infected armadillo nerves. Int. J. Lepr. Other Mycobact. Dis. 57:65, 1989. 55. Lumpkin, L. R. III, Cox, G. F., and Wolf, J. D. Jr.: Leprosy in five armadillo handlers. J. Am. Acad. Dermatol. 9:899, 1983. 56. Magora, A., Sheskin, J., Sagher, F., and Goven, B.: The conduction of the peripheral nerve in leprosy under various forms of treatment. Int. J. Lepr. 9:639, 1971. 57. Mansour, S., Mehansen, A., and El-Ariny, A. F.: Muscular changes in lepromatous leprosy. Trans. R. Soc. Trop. Med. Hyg. 64:918, 1970. 58. McLeod, J. G., Hargrave, J. C., Gye, R. S., et al.: Nerve grafting in leprosy. Brain 98:203, 1975. 59. Monrad-Krohn, G. H.: The Neurological Aspect of Leprosy. Oslo, Jacob Dybwad, 1923. 60. Naafs, B., Pearson, J. M. H., and Baar, A. J. M.: A follow-up of nerve lesions in leprosy during and after reaction using motor nerve conduction velocity. Int. J. Lepr. Other Mycobact. Dis. 44:188, 1976. 61. Nathan, C. F., Keplan, G., Levis, W. R., et al.: Local and systemic effects of intradermal recombinant interferon in patients with lepromatous leprosy. N. Engl. J. Med. 315:6, 1986.

2107 62. Pandaya, J. J., and Anita, N. H.: Elective surgical decompression of nerves in leprosy. Lepr. Rev. 49:53, 1978. 63. Powell, C. S., and Swan, L. L.: Leprosy pathologic changes in 50 consecutive necropsies. Am. J. Pathol. 31:1131, 1955. 64. Prendergast, J. J.: Ocular leprosy in the United States. Arch. Ophthalmol. 23:112, 1940. 65. Ramachandran, A., and Neelan, P. N.: Autonomic neuropathy in leprosy. Indian J. Lepr. 49:277, 1987. 66. Ramachandran, A., and Neelan, P. N.: Autonomic neuropathy in leprosy. Indian J. Lepr. 49:405, 1987. 67. Rambukkana, A., Yamada, H., Zanazzi, G., et al.: Role of ␣-dystroglycan as a Schwann cell receptor for Mycobacterium leprae. Science 282:2076, 1998. 68. Rambukkana A., Zanazzi, G., Tapinos, N., and Salzer, J. L.: Contact-dependent demyelination of Mycobacterium leprae in the absence of immune cells. Science 296:927; discussion 1475, 2002. 69. Ranney, D. A., and Lennox, W. M.: The protective value of a neurovascular island pedicle transfer in hands partially anesthetic due to ulnar denervation in leprosy. J. Bone Joint Surg. Am. 60:328, 1978. 70. Rees, R. J. W., Waters, M. F. R., Weddell, A. G. M., and Palmer, E.: Experimental lepromatous leprosy. Nature 215:599, 1967. 71. Ridley, D. S.: Appendix III: bacterial indices. In Cochrane, R. G., and Davey, T. F. (eds.): Leprosy in Theory and Practice, 2nd ed. Bristol, UK, John Wright & Sons/Baltimore, Williams & Wilkins, p. 620, 1964. 72. Ridley, D. S., and Jopling, W. H.: A classification of leprosy for research purposes. Lepr. Rev. 33:19, 1962. 73. Rosenberg, R. N., and Lovelace, R. E.: Mononeuritis multiplex in lepromatous leprosy. Arch. Neurol. 19:310, 1968. 74. Sabin, T. D.: Temperature-linked sensory loss: a unique pattern in leprosy. Arch. Neurol. 20:257, 1969. 75. Sabin, T. D.: Neurologic features of lepromatous leprosy. Am. Fam. Physician 4:84, 1971. 76. Sabin, T. D.: Peripheral neuropathy: the long and the short of it. Muscle Nerve 9:711, 1986. 77. Sabin, T. D., and Ebner, J. D.: Patterns of sensory loss in lepromatous leprosy. Int. J. Lepr. Other Mycobact. Dis. 37:239, 1969. 78. Sabin, T. D., Hackett, E. R., and Brand, P. W.: Temperatures along the course of certain nerves affected in leprosy. Int. J. Lepr. Other Mycobact. Dis. 42:33, 1974. 79. Saqueton, A. C., Lorinez, A. L., Vick, N. A., and Homer, R. D.: Dapsone and peripheral motor neuropathy. Arch. Dermatol. 100:214, 1969. 80. Sebille, A., Cordoliani, G., Raffalli, M. J., et al.: Dapsoneinduced neuropathy compounds Hansen’s disease nerve damage: an electrophysiological study in tuberculoid patients. Int. J. Lepr. Other Mycobact. Dis. 55:16, 1987. 81. Sebille, A., and Gray, F.: Electromyographic recording and muscle biopsy in lepromatous leprosy. J. Neurol. Sci. 40:3, 1979. 82. Sehgal, V. N.: Nerve abscesses in leprosy in northern India. Lepr. Rev. 37:109, 1966. 83. Sehgal, V. N., Joginder, and Sharma, V. K.: Immunology of leprosy: a comprehensive survey. Int. J. Dermatol. 28:574, 1989.

2108

Diseases of the Peripheral Nervous System

84. Selby, R. C.: Neurosurgical aspects of leprosy. Surg. Neurol. 2:165, 1974. 85. Sheagren, J. N., Block, J. B., Trautman, J. R., and Wolf, S. M.: Immunologic reactivity in patients with leprosy. Ann. Intern. Med. 70:295, 1969. 86. Shepard, C. C.: Multiplication of Mycobacterium leprae in the foot-pad of the mouse. Int. J. Lepr. 30:291, 1962. 87. Shepard, C. C.: Temperature optimum of Mycobacterium leprae in mice. J. Bacteriol. 90:1271, 1965. 88. Shepard, C. C.: Experimental chemotherapy in leprosy, then and now. Int. J. Lepr. Other Mycobact. Dis. 41:307, 1973. 89. Sheshkin, J., Magora, A., and Sagher, F.: Motor conduction velocity studies in patients with leprosy reaction treated with thalidomide and other drugs. Int. J. Lepr. Other Mycobact. Dis. 37:359, 1969. 90. Sheshkin, J., and Sagher, F.: Five years experience with thalidomide treatment of leprosy reaction. Int. J. Lepr. Other Mycobact. Dis. 39:585, 1971. 91. Shuttleworth, J. S., and Ross, H. Sr.: Secondary amyloidosis in leprosy. Ann. Intern. Med. 45:23, 1956. 92. Skinsnes, O. K.: Leprosy in society. III. The relationship of the social to the medical pathology of leprosy. Lepr. Rev. 34:175, 1964. 93. Skinsnes, O. K., and Elvore, R. M.: Leprosy in society. V. Leprosy in occidental literature. Int. J. Lepr. Other Mycobact. Dis. 38:294, 1970. 94. Sohi, A. S., Kandheri, K. C., and Singh, N.: Motor nerve conduction studies in leprosy. Int. J. Dermatol. 10:151, 1971. 95. Swift, T.: Neurofibromatosis and leprosy. J. Neurol. Neurosurg. Psychiatry 34:743, 1971.

96. Swift, T. R.: Peripheral nerve involvement in leprosy: quantitative histologic aspects. Acta Neuropathol. (Berl.) 29:1, 1974. 97. THALOMID (Thalidomide): prescribing information. In Physicians’ Desk Reference, 55th ed. Montvale, NJ, Medical Economics, p. 1153, 2003. 98. Tio, T. H.: Neural involvement in leprosy: treatment with intraneural injections of prednisolone. Lepr. Rev. 37:93, 1966. 99. Uplekar, M. W., and Anita, N. H.: Clinical and histopathological observations on pure neuritic leprosy. Indian J. Lepr. 48:513, 1986. 100. Weinstein, D. E., Freedman, V. H., and Kaplan, G.: Molecular mechanism of nerve infection in leprosy. Trends Microbiol. 7:185, 1999. 101. Wemambu, S. N. C., Turk, J. L., Waters, M. F. R., and Rees, R. J. W.: Erythema nodosum leprosum: a clinical manifestation of the Arthus phenomenon. Lancet 2:933, 1969. 102. WHO Expert Committee on Leprosy: sixth report. World Health Organ Tech Rep Ser 768:1, 1988. 103. WHO Expert Committee on Leprosy: seventh report. World Health Organ Tech Rep Ser 874:1, 1998. 104. WHO Study Group: Chemotherapy of Leprosy for Control Programmes. World Health Organ Tech Rep Ser 675:1, 1982. 105. WHO Technical Advisory Group on Elimination of Leprosy: Report on the Fourth Meeting of the WHO Technical Advisory Group on Elimination of Leprosy. Geneva, World Health Organization, p. 32, 2002. 106. World Health Organization: A Guide to Leprosy Control, 2nd ed. Geneva, World Health Organization, 1988. 107. Yoshizumi, M. O., and Asbury, A. K.: Intra-axonal bacilli in lepromatous leprosy. Acta Neuropathol. (Berl.) 27:1, 1974.

92 Lyme Disease GÉRARD SAID

History Epidemiology Geographic Incidence The Spirochetes The Vector: Ixodes Ticks Neurobiologic Basis of Infection

Clinical Manifestations Skin Lesions Neurologic Manifestations Other Manifestations Pathologic Aspects Experimental Neuroborreliosis

Lyme borreliosis is the most commonly reported tick-borne infection in Europe and North America. The disease is a multisystem disorder that can affect a complex range of tissues, including the skin, heart, nervous system, and to a lesser extent the eyes, kidneys, and liver. The illness is caused by a spirochete (spiral-shaped bacteria), which is transmitted during the blood feeding of ticks of the genus Ixodes.

HISTORY The term Lyme disease was first used following investigation into a geographic cluster of juvenile rheumatoid arthritis in the town of Old Lyme, Connecticut, in the mid-1970s. Subsequent studies led to the isolation from the deer tick, Ixodes scapularis (syn. dammini), of a gram-negative spirochete, which was named Borrelia burgdorferi. The disease has, however, been known in Europe under a variety of names (including erythema migrans, acrodermatitis chronica atrophicans, and Bannwarth’s syndrome) since the 1880s.1,3,16,27 In 1909, Afzelius had associated a red rash (erythema migrans) with the tick Ixodes ricinus.1 Spirochetes were then observed in erythema migrans biopsies. It was then suggested that erythema migrans with associated meningitis was probably the result of an infection by a tick- or other insect-borne bacterium. However, erythema migrans was considered a relatively harmless condition with no connection made between the lesion and subsequent symptoms caused by the same bacterium.

Diagnosis Diagnostic Criteria Differential Diagnosis Treatment Summary

The recognition of Lyme disease as a separate entity occurred in 1977 because of a geographic clustering in Old Lyme, Connecticut.43 Lyme disease, or Lyme borreliosis, is a common tick-borne disease in temperate regions of Europe, Asia, and North America. The peripheral nervous system is involved in a small proportion of patients infected. Its severity is extremely variable, and some cases of Lyme disease mimic several other diseases.

EPIDEMIOLOGY Geographic Incidence Following, the discovery of B. burgdorferi as the causative agent of Lyme borreliosis in 1982,5 the disease emerged as the most prevalent arthropod-borne infection in temperate climate zones around the world. Lyme affects both genders equally, predominantly before the age of 60 years. Children are often affected.7 Lyme disease has been reported in more than 50 countries worldwide. In Europe, thousands of new cases of Lyme borreliosis occur each summer, particularly in Germany, Austria, Switzerland, France, Sweden, and Denmark. In the United States, Lyme borreliosis is mainly reported from three areas: Massachussets to Maryland in the northeast, Wisconsin and Minnesota in the midwest, and California and Oregon in the West.46,47 Approximately 15,000 cases are reported each year in the United States. An estimated attack rate reaches 2% to 3% of the population in highly endemic 2109

2110

Diseases of the Peripheral Nervous System

areas, with an asymptomatic infection rate of 20%.8 Most cases occur during May through September, with 57% in June and July.

Adult females feed for 7 to 10 days from mid-October through November. Adult male ticks do not feed.2

NEUROBIOLOGIC BASIS OF INFECTION The Spirochetes Five genospecies of the Borrelia genus are known: B. burgdorferi sensu stricto, B. garinii, B. afzelii, B. japonica, and B. andersonii. The first three of these species are common in Europe. In the United States, B. burgdorferi predominates, and the other two strains are found in Japan. Certain differences have been noted between American and European isolates of B. burgdorferi in morphology, outer surface proteins, plasmids, and DNA homology, which may account for some differences in symptomatic manifestations of Lyme disease.10,47 The genome is quite small (approximately 1.5 megabases) and consists of a highly unusual linear chromosome of 950 kilobases as well as 9 linear and 12 circular plasmids.15 The sequences of outer surface protein C (OspC) vary considerably among strains, and only a few of the groups of sequences are associated with disseminated disease.4 Infection with the spirochetes induces a strong immune reaction, but immune evasion occurs through antigenic variation in the VlsE protein.30

At the time the tick hypostome penetrates the skin, substances from the tick saliva begin to enter the host, and an active biologic process takes place in the tick. In unfed I. scapularis ticks infected with B. burgdorferi, the spirochetes are generally restricted to the lumen of the gut. During feeding the population of spirochetes expands. During tick engorgement, spirochetes move from the gut to the salivary glands. Transmission of spirochetes from tick to host is uncommon when a tick has fed for less than 48 hours.11 During the time they feed, the ticks alternately suck blood and regurgitate saliva. While the tick feeds, the spirochetes inside the tick alter the expression of genes coding for surface proteins, so the population of spirochetes becomes more heterogeneous. Spirochetes in unfed ticks express primarily outer surface protein A (OspA); as the tick engorges, most spirochetes downregulate OspA and upregulate OspC, an antigen that continues to be expressed in the host during early infection.

CLINICAL MANIFESTATIONS The Vector: Ixodes Ticks Borreliae are transmitted by Ixodes ticks: in the United States by I. scapularis (in the East) and I. pacificus (in the West) and in Europe and Asia by I. ricinus and I. persulcatus. A tick may wander about on the skin of the host for several hours before attaching. The skin lesions, namely erythema migrans, may thus be found anywhere in the body, and not necessarily on areas that may have had contact with grass. Ticks feed once during the three stages of their usually 2-year life. Larval ticks take one blood meal in late summer, nymphs feed during the following spring and early summer, and adults during that autumn. Feeding causes direct damage to the skin of the host. Disease-causing organisms are expelled during feeding. Ticks tend to feed for 2 to 15 days. During feeding the population of spirochetes of the gut of the tick expands. Tick bites are painless. While they are taking a blood meal, the weight of ixodic ticks may increase more than 100fold.46 In the United States, the preferred host for both the larval and nympheal stages of I. dammini is the whitefooted mouse, while the white-tailed deer are the preferred host of adult I. dammini. The percentage of infected nymphal deer ticks with B. burgdorferi ranges from 15% to 85% in highly endemic areas. Marked seasonal peaks of feeding activity are characteristic: larvae of I. scapularis feed mainly in August. Larvae feed for 2 to 5 days until engorged. Nymphs reach peak activity from May to July.

The course of the disease follows three stages that roughly correspond to local infection (stage 1), dissemination (stage 2), and late manifestations (stage 3). Sequentially, the clinical manifestations are cutaneous, then neurologic, cardiac, and rheumatologic.

Skin Lesions The disease starts with local infection characterized clinically by erythema migrans, which is often accompanied by flulike symptoms. At this stage B. burgdorferi is recovered by culture in 50% of blood samples.48 The typical patient first has erythema migrans, sometimes followed several weeks or months later by meningitis or facial palsy, and often months later still by arthritis. Erythema migrans is the major and almost pathognomonic sign of Lyme disease (Fig. 92–1). It is the best clinical criterion of an early infection with B. burgdorferi, but erythema migrans is present in only 40% to 60% of patients with validated borreliosis. Localized erythema migrans results from local spreading of B. burgdorferi in the skin. It starts as a red macule or papule at the site of the bite at day 7 to 10 after tick bite and expands to form a large red ring with central clearing. Increase in size over several days of observation is a key feature of erythema migrans. It is accompanied by fever, minor constitutional symptoms, or regional lymphadenopathy. During dissemination, multifocal erythema

Lyme Disease

2111

FIGURE 92–1 Secondary erythema migrans, approximately 1 month after tick bite. See Color Plate

migrans lesions may occur. In Europe erythema migrans is often an indolent, localized infection, whereas in the United States this lesion is associated with more intense inflammation and signs that often suggest dissemination of the spirochete.44 Within days or weeks after inoculation, the spirochete may spread in the patient’s blood to many sites and can be recovered from blood and from many organs during this stage. Secondary annular lesions, which resemble the primary erythema migrans, occur in about half the patients in association with migratory musculoskeletal and joint pain. Widely disseminated symptoms seem to be more common in the United States than in Europe. Erythema migrans seems to spread faster and inflammation seems more intense in the United States than in Europe, with a possible hematogenous dissemination in the United States versus a contiguous or regional spreading in Europe.41 By this time, the host starts to develop a strong immune response to B. burgdorferi antigens that results in destruction of spirochetes by complement activation through immune complexes. Both polymorphonuclear leukocytes and monocytes readily phagocytose the spirochete, and histologically lymphocytic infiltration is observed in all affected tissues with plentiful plasma cells and perivascular infiltrates.

Neurologic Manifestations After several weeks or months, less than 3 months in most cases, 15% to 20% of patients develop neurologic manifestations. The first of these is usually radicular pain, often of the burning type, with or without associated weakness, with little or no clinical signs of meningitis. Meningitis is actually the most common neurologic abnormality in Lyme disease. It can be the first symptom but is preceded in most cases by erythema migrans, usually

beginning after the skin lesions resolve. Papilledema and increased cerebrospinal fluid (CSF) pressure occur. CSF examination reveals a lymphocytic pleocytosis, usually a few tens or hundreds of cells per milliliter, with mild elevation of protein with a high proportion of immunoglobulins, as indicated by oligoclonal bands. The presence of activated B cells is suggestive of the diagnosis in this setting. The CSF glucose is usually normal but can be low.34,36 The spirochete has been cultured from the CSF on several occasions, with reports of isolation of spirochetes from acellular CSF during seemingly localized infection.8 Common patterns include painful thoracoabdominal sensory radiculitis, associated with distal involvement. In some series a cranial meningoradiculitis is more common in children and young adults.20,21,23 A painful spinal meningoradiculitis occurs predominantly in elderly patients.32 Painful radiculitis develops over an average of 4 to 6 weeks after the tick bite. Pains occur first, worse at night. They are described as burning, boring, biting, or tearing in nature and only weakly respond to common analgesic drugs.23 Maximum pain levels are often reached within a few days. In most patients further neurologic irritation and deficit occur within the following weeks, with asymmetrical weakness in many.26 Multifocal spinal root or cranial nerve involvement often develops within a few days or weeks, with uni- or bilateral facial palsy and asymmetrical sensorimotor radiculoneuropathy. Cranial neuropathy is present in 50% of patients with neurologic abnormalities.8,17,34,36 Facial palsy is often bilateral. The cranial nerves can be affected, in association with meningitis. Cranial nerve palsies resolve within weeks or months, sometimes incompletely. Sphincter disturbances occur. In children, the optic nerve may also be affected because of inflammation or increased intracranial pressure, which may lead to blindness.41 Treated or untreated, both weakness and sensory loss improve within a few weeks, but

2112

Diseases of the Peripheral Nervous System

recovery may take up to several months and often remains incomplete because of the intensity and proximal localization of axonal lesions. Electrophysiologic testing of patients with peripheral neuropathy has shown evidence both of demyelination and of axonal degeneration, but usually axonal lesions predominate.13,45

Other Manifestations Cardiac involvement occurs in 4% to 8% of patients. Manifestations include fluctuating atrioventricular node block, mild left ventricular dysfunction, and, rarely, cardiomegaly or fatal pancarditis. The duration of cardiac abnormalities is usually brief, and they do not necessitate permanent insertion of a pacemaker.28,42 Arthritis occurs in transient episodes, a mean of 6 months after the onset of the disease. They affect about 60% of the patients in the United States and are characterized by asymmetrical oligoarticular arthritis especially of the knee; one or a few joints are affected. The spirochete has occasionally been cultured from joint fluid. Arthritis seems less common in Europe. Occasional cases of distal symmetrical polyneuropathy have been reported in association with acrodermatitis chronica atrophicans.22 In clinical and electrophysiologic study, 17 patients with the late manifestations of acrodermatitis chronica atrophicans were found to have a sensory polyneuropathy with marked abnormality of vibration threshold. Sural nerve biopsy performed in 3 of the 17 patients showed an axonal neuropathy without signs of inflammation.25 A variety of late syndromes affecting the central nervous system have been described, including spastic paraparesis, ataxia, relapsing multiple sclerosis–like illness, bladder dysfunction, cognitive impairment, dementia, and subacute encephalitis.19,33 Although these patients had serologic evidence of Lyme disease, they did not have intrathecal

synthesis of antibody to B. burgdorferi or were not tested for it. Some minor central nervous system involvement is likely to occur in some patients with marked meningitis. After appropriate treatment, a small percentage of patients with Lyme disease still complain of musculoskeletal pain, neurocognitive difficulties, or fatigue—symptoms that may last for years. This disabling syndrome, which is sometimes called “chronic Lyme disease” or “post–Lyme disease syndrome,” is similar to chronic fatigue syndrome or fibromyalgia,12 two syndromes that lack a substantial pathologic or biologic basis. This postinfectious syndrome seems to occur more frequently in patients whose symptoms are suggestive of early dissemination of the spirochete to the nervous system, particularly if treatment is delayed.24,39 However, in a large study, the frequency of symptoms of pain and fatigue was no greater in patients who had had Lyme disease than in age-matched subjects who had not had this infection.38 In addition, the post-Lyme manifestations do not respond to antibiotic therapy, suggesting that they are not linked to persisting infection.

PATHOLOGIC ASPECTS Histologically, the nerve lesions are associated with lymphoplasmocytic inflammatory infiltrates that predominate in the meninges and nerve roots, along with axonal lesions. The nerve trunks are variably affected. Severe inflammatory infiltrates can be found in nerve trunks (Fig. 92–2) in some patients.45 In others the inflammatory infiltrate is restricted to the meninges, cranial nerves, and spinal roots (Fig. 92–3). The presence of B. burgdorferi has not been convincingly documented in the nerves. One of our patients with Lyme meningoradiculitis died of pulmonary embolism in spite of preventive treatment with low-molecular-weight heparin. At postmortem examination the peripheral nerves were normal,

FIGURE 92–2 Superficial peroneal nerve biopsy in a patient who presented with subacute, painful, paralysis of the right foot. There was no awareness of tick bite or erythema migrans. Note the significant inflammatory infiltrate of the epineurium and perineurium and around endoneurial capillaries. (Hematoxylin and eosin staining.) See Color Plate

2113

Lyme Disease

FIGURE 92–3 Dorsal spinal root at postmortem examination showing heavy inflammatory infiltrate with mononuclear cells in the subarachnoid space (stars) and in the endoneurium of the dorsal root (arrows). The sciatic nerve was normal on postmortem examination. This patient, who was being treated for subacute Lyme meningoradiculitis, died of pulmonary embolism in spite of preventive treatment with low-molecularweight heparin. (Trichrome staining.) See Color Plate

while there was significant inflammatory infiltration of the meninges around spinal roots and cranial nerves, especially around the facial nerves (Fig. 92–4). The central nervous system was not affected, except for the inflammatory infiltrate related to meningitis in Virchow-Robin spaces.

EXPERIMENTAL NEUROBORRELIOSIS The histopathology of early and late aspects of neuroborreliosis were studied in rhesus monkeys infected with the JD1 strain of B. burgdorferi,37 a model previously studied electrophysiologically.14 Three months after infection of the monkeys, the authors found neuritis affecting multiple nerves. The pathologic findings were similar to those observed in humans, with massive mononuclear cell infiltrates around intraneural blood vessels. Macrophages and B lymphocytes were prominently represented. Axonal

FIGURE 92–4 Same patient as in Figure 92–3. High magnification of a longitudinal section of the facial nerve (stained with hematoxylin and eosin), which is massively infiltrated with lymphocytes and plasma cells. See Color Plate

degeneration and demyelination were noted near foci of inflammation. Borrelia burgdorferi antigens were detected immunohistochemically within macrophages. Inflammatory lesions were less marked at 6 months and virtually absent at 46 months postinfection. In one “chronic” animal, B. burgdorferi DNA was amplified from a sciatic nerve. This monkey model, which closely reproduces human infection, can be successfully used to test drugs or vaccinations.

DIAGNOSIS Diagnostic Criteria From a neurologic point of view, presence of a subacute meningoradiculoneuritis with facial palsy and signs and symptoms suggesting a focal or multifocal radiculoneuropathy are highly suggestive of Lyme borreliosis. These findings

2114

Diseases of the Peripheral Nervous System

are diagnostic if erythema migrans has been confirmed by a physician, as emphasized at the Second National Conference on Serologic Diagnosis of Lyme Disease.35 Guidelines have also been established by the American Academy of Neurology for the diagnosis of neurologic Lyme disease.18 Serology is the only practical laboratory aid in diagnosis, but serologic testing is not yet standardized and the results from different laboratories may vary. The physician must be aware of the potential for false-negative, and more commonly of false-positive, results. Titers should increase fourfold or more between the erythema migrans phase and subsequent neurologic involvement. Many patients have asymptomatic B. burgdorferi infection; in addition, falsepositive results, particularly with immunoglobulin M (IgM), may occur both in healthy subjects and in patients with a variety of other diseases. In the United States, approximately 20% to 30% of patients have positive responses, usually of the IgM isotype, during early stages of Lyme disease. Two to 4 weeks later, about 70% to 80% have seroreactivity, even after antibiotic treatment. After 1 month, the majority of patients with active infection have immunoglobulin G antibody responses. Examination of a single specimen in time does not discriminate between previous and ongoing infection. There are numerous reports of the successful use of polymerase chain reaction (PCR) for detecting B. burgdorferi in the CSF, but the sensibility and specificity of PCR depend on genetic stability of the target locus. Results of PCR must be interpreted with caution and performed in specific laboratories. It cannot be used routinely for diagnostic purposes.6 The most helpful specific test is intrathecal anti–B. burgdorferi antibody production. Paired serum and CSF samples must therefore be collected.

ampicillin, and ceftriaxone but only moderately sensitive to penicillin. For early Lyme disease (localized stage 1 or disseminated stage 2 infection), oral doxycycline (100 mg orally twice a day for 10 to 30 days), a long-acting tetracycline that achieves good tissue levels, is recommended. Although intravenous penicillin is generally considered effective in the treatment of neurologic disease, ceftriaxone is now commonly used.9 Ceftriaxone crosses the blood-brain barrier and requires only once-a-day administration of 2 g intravenously (IV) for 14 to 28 days. In children, ceftriaxone (75 to 100 mg/kg/day IV, once a day for 14 to 28 days) is recommended. Corticosteroids may be given concomitantly with antibiotic treatment in some cases. A recent study compared prophylaxis with single-dose doxycycline for the prevention of Lyme disease after a recognized I. scapularis tick bite.31 In patients who had received a single dose of 200 mg of doxycycline, erythema migrans occurred in 0.4% of the subjects, versus 3.2% in nontreated individuals. Others have found that prophylaxis was not indicated in this setting.40 Various environmental protective measures have been suggested, as well as vaccination for Lyme disease of 15- to 70-year-old persons who have frequent exposure to I. scapularis ticks.41 People who have had Lyme disease remain at risk of reinfection afterward.

SUMMARY Lyme disease is a common infectious disease that can induce subacute meningoradiculoneuritis in a small proportion of patients. The condition is benign and self-limited, but the damage to the peripheral nervous system can result in severe focal or multifocal sensorimotor deficits. In addition, Lyme disease can pose difficult diagnostic problems when the awareness of tick bite or erythema migrans is missing.

Differential Diagnosis In many cases the awareness of tick bite and erythema migrans are lacking, which makes the diagnosis difficult, resting only on biologic testing and exclusion of other causes. Subacute focal or multifocal neuropathy can lead to consideration of other causes of neuropathy, including vasculitis, sarcoidosis, and leprosy. Malignant lymphoma infiltrating the meninges and spinal roots can also be mistaken for Lyme disease.29 Greater confusion may result from misinterpretation of CSF pleocytosis with activated B cells.

TREATMENT Treatment with high doses of penicillin gives good results at stage 1, but the results are not as good in patients with stage 2 neurologic abnormalities, and in patients with arthritis. Borrelia burgdorferi seems highly sensitive to tetracycline,

ACKNOWLEDGMENT I am grateful to Doctor Catherine Lacroix for her valuable contribution.

REFERENCES 1. Afzelius, A.: Erythema chronicum migrans. Acta Derm. Venerol. 2:120, 1921. 2. Anderson, J. F.: The natural history of ticks. Med. Clin. North Am. 86:205, 2002. 3. Bannwarth, A.: Zur Klinik und Pathogenese der “chronischen lymphocytaren Meningitis.” Arch. Psychiatr. Nervenkr. 117:161, 1944. 4. Baranton, G., Seinost, G., Theodore, G., et al.: Distinct levels of genetic diversity of Borrelia burgdorferi are associated with different aspects of pathogenicity. Res. Microbiol. 152:149, 2001.

Lyme Disease 5. Budgrofer, W., Barbour, A. G., Hayes, S. F., et al.: Lyme disease—a tick borne spirochetosis? Science 216:1317, 1982. 6. Bunikis, J., and Barbour, A. G.: Mabpratpru testing for suspected Lyme disease. Med. Clin. North Am. 86:311, 2002. 7. Christen, H. J., Hanefeld, F., Eifferet, H., and Thomssen, R.: Epidemiology and clinical manifestations of Lyme borreliosis in childhood. Acta Paediatr. Suppl. 386:1, 1993. 8. Coyle, P. K., and Schutzer, S. E.: Neurologic aspects of Lyme disease. Med. Clin. North Am. 86:261, 2002. 9. Dattwyler, R. J., Halperin, J. J., Volkman, D. J., and Luft, B. J.: Treatment of late Lyme borreliosis—randomised comparison of ceftriaxone and penicillin. Lancet 1:1191, 1988. 10. de Silva, A. M., and Fikrig, E.: Arthropod- and host-specific gene expression by Borrelia burgdorferi. J. Clin. Invest. 99:377, 1997. 11. Des Vignes, F., Piesman, J., Herrerman, R., et al.: Effect of tick removal on transmission of Borrelia burgdorferi and Ehrlichia phagocytophilia by Ixodes scapularis nymphs. J. Infect. Dis. 183:773, 2001. 12. Dinerman, H., and Steere, A. C.: Lyme disease associated with fibromyalgia. Ann. Intern. Med. 117:281, 1992. 13. Duray, P. H., and Steere, A. C.: Clinical pathologic correlations of Lyme disease by stage. Ann. N. Y. Acad. Sci. 539:65, 1988. 14. England, J. D., Bohm, R. P., Roberts, E. D., and Philipp, M. T.: Mononeuropathy multiplex in rhesus monkeys with chronic Lyme disease. Ann. Neurol. 41:375, 1997. 15. Fraser, C. M., Casjens, S., Huang, W. M., et al.: Genomic sequence of a Lyme disease spirochete, Borrelia burgdorferi. Nature 390:580, 1997. 16. Garin, C., and Bujadoux, C.: Paralysie par les tiques. J. Med. Lyon 3:765, 1922. 17. Halperin, J., Luft, B. J., Volkman, D. J., and Dattwyler, R. J.: Lyme neuroborreliosis: peripheral nervous system manifestations. Brain 113:1207, 1990. 18. Halperin, J. J., Logigian, E. L., Finkel, M. F., and Pearl, R. A.: Practice parameters for the diagnosis of patients with nervous system Lyme borreliosis (Lyme disease). Quality Standards Subcommittee of the American Academy of Neurology. Neurology 46:619, 1996. 19. Halperin, J. J., Luft, B. J., Anand, A. K., et al.: Lyme neuroborreliosis: central nervous system manifestations. Neurology 39:753, 1989. 20. Hansen, K., and Lebech, A. M.: Lyme neuroborreliosis: a new sensitive diagnostic assay for intrathecal synthesis of Borrelia burgdorferi–specific immunoglobulin G, A, and M. Ann. Neurol. 30:197, 1991. 21. Hansen, K., and Lebech, A. M.: The clinical and epidemiological profile of Lyme neuroborreliosis in Denmark 1985–1990: a prospective study of 187 patients with Borrelia burgdorferi specific intrathecal antibody production. Brain 115:399, 1992. 22. Hopf, H. C.: Peripheral neuropathy in acrodermatitis chronica atrophicans (Herxheimer). J. Neurol. Neurosurg. Psychiatry 38:452, 1975. 23. Kaiser, R.: Neuroborreliosis. J. Neurol. 245:247, 1998. 24. Kalish, R. A., Kaplan, R. F., Taylor, E., et al.: Evaluation of study patients with Lyme disease, 10–20-year follow-up. J. Infect. Dis. 183:453, 2001.

2115

25. Kindstrand, E., Nilsson, B. Y., Hovmark, A., et al.: Polyneuropathy in late Lyme borreliosis—a clinical, neurophysiological and morphological description. Acta Neurol. Scand. 101:47, 2000. 26. Kruger, H., Reuss, K., Pulz, M., et al.: Meningoradiculitis and encephalomyelitis due to Borrelia burgdorferi. J. Neurol. 226:322, 1989. 27. Lipschütz, B.: Weiterer Beitrag zur Kenntis des “Erythema chronicum migrans.” Arch. Dermatol. Syph. 143:365, 1923. 28. Marcus, L. C., Steere, A. C., Duray, P. H., et al.: Fatal pancarditis in a patient with coexistent Lyme disease and babesiosis: demonstration of spirochetes in the myocardium. Ann. Intern. Med. 103:374, 1985. 29. Mayfrank, L., Koher, J., Arnolds, B., and Lycking, C. H.: Lymphomatous meningitis appearing as Garin-BujadouxBarnwarth meningopolyneuritis. Lancet 1:106, 1986. 30. McDowell, J. V., Sung, S. Y., Hu, L. T., and Marconi, R. T.: Evidence that the variable regions of the central domain of VlsE are antigenic during infection with Lyme disease spirochetes. Infect. Immun. 70:4196–4203, 2002. 31. Nadelman, R. B., Nowakowski, J., Fish, D., et al.: Prophylaxis with single-dose doxycycline for the prevention of Lyme disease after an Ixodes scapularis tick bite. N. Engl. J. Med. 345:79, 2001. 32. Oschmann, P., Dorndorf, W., Hornig, C., et al.: Stages and syndromes of neuroborreliosis. J. Neurol. 245:262, 1998. 33. Pachner, A. R., Duray, P., and Steere, A. C.: Central nervous system manifestations of Lyme disease. Arch. Neurol. 46:790, 1989. 34. Pachner, A. R., and Steere, A. C.: The triad of neurologic manifestations of Lyme disease: meningitis, cranial neuritis, and radiculoneuritis. Neurology 35:47, 1985. 35. Recommendations for test performance and interpretation from the Second National Conference on Serologic Diagnosis of Lyme Disease. MMWR Morb. Mortal. Wkly. Rep. 44:590, 1995. 36. Reik, L. Jr., Burgdorfer, W., and Donaldson, J. O.: Neurologic abnormalities in Lyme disease without erythema chronicum migrans. Am. J. Med. 81:73, 1986. 37. Roberts, E. D., Bohm, R. P., Lowrie, R. C., et al.: Pathogenesis of Lyme neuroborreliosis in the rhesus monkey: the early disseminated and chronic phases of disease in the peripheral nervous system. J. Infect. Dis. 178:722, 1998. 38. Seltzer, E. G., Gerber, M. A., Carter, M. L., et al.: Long-term outcomes of persons with Lyme disease. JAMA 283:609, 2000. 39. Shadick, N. A., Phillips, C. B., Sangha, O., et al.: Musculoskeletal and neurologic outcomes in patients with previously treated Lyme disease. Ann. Intern. Med. 131:919, 1999. 40. Shapiro, E. D., Gerber, M. A., Holabird, N. B., et al.: A controlled trial of antimicrobial prophylaxis for Lyme disease after deer-tick bites. N. Engl. J. Med. 237:1769, 1992. 41. Steere, A. C.: Lyme disease. N. Engl. J. Med. 345:115, 2001. 42. Steere, A. C., Batsford, W. P., Weinberg, M., et al.: Lyme carditis: cardiac abnormalities of Lyme disease. Ann. Intern. Med. 93:8, 1980. 43. Steere, A. C., Malawista, S. E., Snydman, D. R., et al.: Lyme arthritis: an epidemic of oligoarticular arthritis in children and adults in three Connecticut communities. Arthritis Rheum. 20:7, 1977.

2116

Diseases of the Peripheral Nervous System

44. Strle, F., Nadelman, R. B., Cimperman, J., et al.: Comparison of culture-confirmed erythema migrans caused by Borrelia burgdorferi sensu stricto in New York State and by Borrelia afzelii in Slovenia. Ann. Intern. Med. 130:32, 1999. 45. Vallat, J. M., Hugon, J., Lubeau, M., et al.: Tick-bite meningoradiculoneuritis: clinical, electrophysiologic, and histologic findings in 10 cases. Neurology 37:749, 1987.

46. Wilson, M. E.: Prevention of tick-borne diseases. Med. Clin. North Am. 86:219, 2002. 47. Wilson, M. L., and Spielman, A.: Seasonal activity of immature Ixodes dammini. J. Med. Entomol. 22:408, 1985. 48. Wormser, G. P., Bittker, S., Cooper, D., et al.: Comparison of yields of blood cultures using serum or plasma from patients with early Lyme disease. J. Clin. Microbiol. 38:1648, 2000.

93 Herpesvirus Infection and Peripheral Neuropathy DONALD H. GILDEN AND KENNETH L. TYLER

Herpes Simplex Virus Neurologic Complications of Infection Genomic Organization Latency Bell’s Palsy Radiculopathy Diagnosis Varicella Zoster Virus Neurologic Complications of Reactivation Genomic Organization

Latency Herpes Zoster Postherpetic Neuralgia Zoster Sine Herpete VZV Neuropathy without Rash Epstein-Barr Virus Neurologic Complications of Infection Virology and Latency Neuropathy Treatment

There are more than 80 known herpesviruses that infect humans, cattle, monkeys, fish, and mosquitoes. All herpesviruses are DNA viruses. Eight human herpesviruses have been identified: herpes simplex virus (HSV)-1, HSV-2, varicella zoster virus (VZV), Epstein-Barr virus (EBV), cytomegalovirus (CMV), and human herpesvirus (HHV)-6, HHV-7, and HHV-8. Of the eight viruses, seven have been associated with neurologic disease and five herpesviruses (HSV-1, HSV-2, VZV, CMV, and EBV) cause neuropathy. This chapter focuses on neuropathy produced by the five human herpesviruses summarized in Table 93–1. For each virus, a brief overview of virologic features is provided. Because neurologic disease produced by herpesviruses usually occurs after reactivation from latency, information on the site and physical state of viral nucleic acid and gene expression in latently infected tissue is given. To provide some perspective on the scope of neurologic disease produced by each herpesvirus, the non–peripheral nervous system complications are cited, followed by detailed clinical features of the neuropathy produced by each of the five herpesviruses.

Cytomegalovirus Neurologic Complications of Infection Virology Cranial Neuropathies Myeloradiculopathy Mononeuritis Multiplex CMV and the Guillain-Barré Syndrome Diagnosis Treatment

HERPES SIMPLEX VIRUS Primary HSV infection typically involves the oropharynx and genitals. Most infections are subclinical and followed by latency in ganglionic neurons (see Chapter 31). Periodic reactivation produces recurrent painful vesicular eruptions around the mouth (HSV-1) or genitalia (HSV-2). Neonatal HSV-2 infections occur when an infant becomes infected in utero, intrapartum, or in the immediate postnatal period. Intrapartum infection accounts for approximately 80% of congenital HSV infections, and occurs when the infant contacts infected maternal genital secretions as it passes through the birth canal.

Neurologic Complications of Infection Neurologic complications of HSV infection include encephalitis (usually with HSV-1) or aseptic meningitis and recurrent radiculopathy (usually with HSV-2). Rarely, HSV-2 also causes myelitis, especially in immunocompromised individuals. HSV-1 has also been linked with Bell’s palsy. 2117

2118

Diseases of the Peripheral Nervous System

Table 93–1. Neuropathies Associated with Human Herpesvirus Infection HSV-1

HSV-2

VZV

EBV

CMV

Bell’s palsy

Radiculopathy, primarily lumbosacral; often recurrent

Zoster (radicular pain with rash) Zoster sine herpete (radicular pain without rash) Postherpetic neuralgia (radicular pain that persists months after resolution of rash)

Cranial neuropathies Radiculopathy with or without myelitis and encephalitis Acute autonomic neuropathy

Polyradiculopathy retinopathy Guillain-Barré syndrome Mononeuritis multiplex

Genomic Organization HSV-1 is a double-stranded DNA virus that is 152 kb in size. The entire 152,260-bp genome has been sequenced (accession number X14112) and consists of two covalently linked components, a unique long (UL) and a unique short (US) segment, each bracketed by inverted repeat sequences. Thus there are four possible isomers of the viral DNA molecule. There are about 90 unique transcriptional units, at least 84 of which encode proteins. The genes have been grouped into three general kinetic classes: immediate-early, early, and late. An updated assignment of function to various HSV genes as well as the structure of herpes simplex virions is detailed in Roizman and Knipe.108

Latency HSV-1 latency appears to be restricted to cranial nerve ganglia, as manifested by spontaneous, often recurrent outbreaks of herpetic vesicles on the mouth, or by rescue of HSV-1 from explants of trigeminal,9,10 nodose, and vagal134 and ciliary16 ganglia of humans after death. HSV-1 sequences have also been found in human thoracic ganglia85 and brain,8,43 but virus has never been recovered by explantation or co-cultivation of these tissues with indicator cells. Most humans have latent HSV-1 in their trigeminal ganglia.85 HSV-2 is latent primarily in lumbosacral ganglia; reactivation produces genital herpes. Studies of HSV-1 latency have revealed a viral copy number that varies from less than 10 to greater than 1000 copies per cell.17,112 HSV-1 DNA is extrachromosomal,92 existing as a circular episome.38 Neurons are the exclusive site of HSV-1 latency; a single latency-associated transcript is present in the nucleus of neurons.121,123 Unlike other herpesviruses, there is no evidence that any HSV proteins are produced during latency.

Bell’s Palsy Peripheral facial nerve palsy is the most common cranial neuropathy; the incidence approaches 40 per 100,000 individuals. The most compelling evidence that HSV might

cause Bell’s palsy is based on the detection of HSV DNA in endoneurial fluid of 79% of 14 patients with idiopathic Bell’s palsy, but not in patients with facial palsy associated with VZV infection.99 This important study has led to treatment of Bell’s palsy with antiviral agents, particularly when the diagnosis is made within 1 week of the onset of disease. Nevertheless, controlled trials of acyclovir in Bell’s palsy have shown little or no effect compared to placebo or steroid treatment alone.62,116,122 Patients with Bell’s palsy frequently show enhancement of the geniculate ganglion and facial nerve on gadolinium-enhanced T1-weighted magnetic resonance imaging.127

Radiculopathy After reactivation from lumbosacral ganglia, HSV-2 causes genital herpes. Extension of virus into the meninges may produce aseptic meningitis, often recurrent. The same mechanism probably accounts for isolated cases of HSV radiculitis or radicular pain,118 manifesting as recurrent sciatica98 or trigeminal-distribution facial pain,80 as well as recurrent conjugal neuralgia associated with outbreaks of genital herpes caused by HSV-2.66 Zosteriform lesions are also caused by VZV. However, recurrent zoster (shingles) is rare except in severely immunocompromised individuals. In contrast, HSV often recurs, frequently involving the same dermatome and in immunocompetent individuals.

Diagnosis Amplification of viral DNA from cerebrospinal fluid (CSF) is the procedure of choice to corroborate the diagnosis of HSV infection of the nervous system and should be used in cases of radiculopathy of unknown cause, particularly if pain is recurrent, and even in the absence of rash.

VARICELLA ZOSTER VIRUS VZV is a neurotropic human herpesvirus that infects nearly all humans and causes chickenpox (varicella). After chickenpox, VZV becomes latent in cranial nerve, dorsal

2119

Herpesvirus Infection and Peripheral Neuropathy

root, and autonomic nervous system ganglia along the entire neuraxis. Virus reactivation produces shingles (zoster), characterized by pain and rash usually restricted to one to three dermatomes. The increased incidence of zoster in elderly and immunocompromised individuals appears to be due to a VZV-specific host immunodeficiency. Zoster is often complicated by postherpetic neuralgia (PHN), pain that persists for months to years after rash resolves. PHN may reflect a chronic VZV ganglionitis.50

trigeminal ganglia revealed a wide variation of 19 to 3145 copies of VZV DNA per latently infected neuron,27 which may reflect the uneven amount of circulating virus encountered during primary infection. Latent VZV DNA exists as a circular episome,20 similar to the structure of latent HSV-1 DNA.38,107 At least five VZV genes and two VZV-specific proteins are transcribed during latency.25,26,83

Herpes Zoster Neurologic Complications of Reactivation Reactivation produces herpes zoster. Virus may also spread to the central nervous system (CNS) to produce two alarming complications: small vessel vasculopathy primarily in immunocompromised individuals, and large vessel granulomatous arteritis in immunocompetent patients. These disorders are described in detail by Gilden et al.48,54 VZV vasculopathy is due to productive virus infection in cerebral arteries. Central and peripheral nervous system disease caused by VZV may occur at any level of the neuraxis and in the absence of rash. This includes cases of encephalomyeloradiculoneuropathy, mixed small and large vessel vasculopathy, and myelopathy.33,49

Genomic Organization The entire 125,884-bp VZV genome has been sequenced and has a high degree of homology to the prototype human alphaherpesvirus, HSV-1.30,89–91,102 The VZV genome is composed of UL (104,836 bp) and US (5232 bp) DNA segments bound by inverted repeats. Computer-assisted analysis of the VZV genome has identified 71 predicted open reading frames (ORFs) encoding proteins 8 to greater than 300 kDa in size.30 Transcripts mapping to most of the ORFs have been identified in VZV-infected cells in culture, although less than 20 VZV genes have been analyzed in detail.77,82,100,106 The VZV genome is compact: The 71 ORFs are separated by an average of 211 bp, suggesting that the promoters are closely associated with the genes they control. During virus replication, the US region and its repeat sequences invert around the UL region.29,65,78,124 The biologic significance of DNA isomerization for VZV, as well as for other human herpesviruses, is unknown.

Latency Unlike HSV-1, latency of which is mostly restricted to cranial nerve ganglia, VZV is latent in dorsal root84 and autonomic nervous system ganglia,52 as well as in cranial nerve ganglia.56,67 VZV DNA is latent in ganglia of more than 90% of normal adults.85 The virus is found predominantly if not exclusively in ganglionic neurons.55,67,73,74,81 Real-time polymerase chain reaction (PCR) analysis of

Zoster is characterized by pain and vesicular rash on an erythematous base in one to three dermatomes (localized zoster). Most herpes zoster cases reflect VZV reactivation from ganglia secondary to a natural decline in cell-mediated immunity to VZV with aging.14,94 In rare instances, cases of zoster are clustered, supporting the notion that zoster can be acquired by exogenous reinfection. More than one-half million Americans develop zoster per year. Fewer than 10% of cases of zoster recur. Although chickenpox occurs primarily in the spring, zoster occurs any time of year. There is no sex bias. Zoster is 8 to 10 times more common in persons older than age 60, and approximately 50% of persons 80 years of age and older have had herpes zoster once. A continuing increase in the number of elderly persons will result in greater herpes zoster–associated morbidity and mortality. Zoster and its attendant complications are also more common in immunocompromised individuals, especially in transplant recipients and cancer and acquired immunodeficiency syndrome (AIDS) patients. Long-term, low-dose steroid therapy may also predispose patients to the serious complications of VZV reactivation, disseminated zoster and zoster vasculopathy.51 Overall, zoster may be viewed in the context of a continuum in immunodeficient individuals, ranging from a natural decline in VZV-specific immunity with age, to more serious immune deficits seen in cancer patients and transplant recipients, and ultimately to patients with AIDS. Zoster also occasionally develops in patients stressed by surgery.53 Clinical Features Zoster pain is severe, deep, burning, and jabbing and usually lasts 4 to 6 weeks. Examination reveals hyperesthesia (allodynia) in a dermatomal distribution with mixed hypopigmentation and hyperpigmentation. Except for its longer duration, acute zoster pain does not differ from the pain of PHN. Zoster may involve any dermatome and cranial nerve except the first. The trunk, represented by 12 pairs of thoracic ganglia, is the most frequently involved site. The face, which is supplied by the trigeminal nerve, is involved more often than any single thoracic dermatome. Trigeminal-distribution zoster occurs most frequently in the ophthalmic branch. Cranial neuropathies often occur weeks or months after cutaneous signs, perhaps because of

2120

Diseases of the Peripheral Nervous System

microinfarction of cranial nerves from occlusion of the vasa vasorum. The blood supply of the cranial nerves is part of the carotid circulation; because trigeminal afferent nerves innervate the large extracranial and intracranial blood vessels, virus can also spread along ganglionic afferent fibers to small vessels supplying cranial nerves. A similar scenario has been proposed to explain the pathogenesis of VZV-induced granulomatous arteritis. Peripheral facial weakness, an abnormality of the cranial nerves, is often associated with vesicles in the ear (herpes zoster oticus); the combination of these conditions constitutes the Ramsay Hunt syndrome. Zoster in cervical or lumbosacral dermatomes may be associated with muscle weakness. The incidence of zoster paresis varies from as low as 0.5% to as high as 31%. Bladder and bowel dysfunction are often associated with a sacral distribution of herpes zoster. Treatment of Acute Herpes Zoster Valacyclovir (1000 mg three times daily for 7 days) and famciclovir (500 mg three times daily for 7 days) appear to be equally effective in clearing skin lesions and reducing pain. Both drugs are safe and well tolerated. Because it is easier for patients to take medication three times daily instead of five times daily, valacyclovir and famciclovir are preferred to acyclovir. In combination with antiviral therapy, steroids may reduce the use of analgesics as well as the time to uninterrupted sleep and to resumption of usual activities.135

Postherpetic Neuralgia PHN, defined as pain persisting longer than 4 to 6 weeks after the appearance of rash, is the most common complication of zoster. Age is the most important factor in predicting the development of PHN. The incidence of PHN is greater than 40% in patients 60 years of age and older (reviewed by Gilden47). Analgesics are used to reduce the pain; if possible, narcotics should be avoided. Gabapentin (Neurontin), beginning with 300 mg daily and gradually increasing up to a maximum of 3600 mg daily, provides some relief. Starting with 25 mg at night, amitriptyline can be gradually increased to 150 mg daily. Side effects include postural hypotension, confusion, and urinary retention. Carbamazepine may help, but doses exceeding 600 mg daily are often required and side effects of sleepiness, confusion, and unsteadiness, particularly in the elderly, may be intolerable. Carbamazepine should be started at doses of 100 mg daily and gradually increased by 100 mg every third day until pain decreases. Lidocaine skin patches are often helpful, as is topical application of ketamine105 if applied alternatively throughout the day and before sleep. Transcutaneous electrical nerve stimulation units, nerve blocks, and rhizotomy do not usually provide long-term relief. Studies linking PHN with VZV persistence50

provide a rationale to treat with high-dose acyclovir for more than 10 days to reduce the virus burden. Further studies are needed to determine whether antiviral agents, with or without steroids and started at the onset of herpes zoster, prevent PHN and whether short-term, high-dose antiviral agents effectively treat PHN.

Zoster Sine Herpete Prolonged radicular pain without rash may indicate zoster sine herpete. Until recently, support for the notion that zoster sine herpete was caused by viral infection was wanting. This nosologic entity was verified in two men, ages 62 and 66, who developed radicular pain that persisted for years. The pain did not differ from that of PHN. VZV DNA, but not HSV DNA, was found in the blood mononuclear cells of one man and in the CSF of both men 5 to 8 months after the onset of pain.57 Treatment with intravenous acyclovir, 10 to 15 mg/kg every 8 hours for 14 days, followed by oral acyclovir relieved pain in both patients.

VZV Neuropathy without Rash Of particular interest are three cases of acute, chronic, and recurrent neuropathy without rash attributed to VZV based on detection of VZV DNA and VZV antibody in CSF, along with reduced serum:CSF ratios of VZV immunoglobulin G (IgG) compared to total IgG or albumin, or both.42 Two of these patients responded well to treatment with intravenous acyclovir, underscoring the value of aggressive testing for VZV in unusual cases of neuropathy or CNS disease that might be caused by VZV.

EPSTEIN-BARR VIRUS EBV is the cause of infectious mononucleosis and is latent in most adult humans. EBV can also be oncogenic under unusual circumstances, as evidenced by its close association with Burkitt’s lymphoma and nasopharyngeal carcinoma, and by its ability to immortalize lymphocytes in vitro and to induce lymphomas in marmosets.

Neurologic Complications of Infection Of the human herpesviruses, EBV produces the most protean neurologic syndromes. EBV may cause aseptic meningitis, cerebellar ataxia, encephalitis (including acute demyelinating encephalopathy), myelitis, neuritis or overlapping myeloneuritis, and encephalomyelitis or encephalomyeloneuritis. Chronic active EBV infection has been associated with basal ganglionic calcification.

2121

Herpesvirus Infection and Peripheral Neuropathy

Virology and Latency The EBV genome is a linear, double-stranded DNA 184 kb in size. Like HSV and VZV, EBV exhibits gene rearrangements, including repeat sequences at both termini. EBV, the first herpesvirus to be completely sequenced, does not hybridize to HSV, VZV, or CMV DNA. Most adults are seropositive for EBV and intermittently shed virus in saliva. The virus is present in a latent state in approximately 1 in 105 to 106 peripheral blood B lymphocytes in adults. The in vitro host range for efficient EBV infection is restricted to human B lymphocytes. After infection in vitro, approximately 10% of cells become latently infected and proliferate as immortalized cells, and EBV DNA becomes episomal. During latency, multiple EBV genes are expressed. Latent virus in infected cells can be detected with antibodies to any of eight different proteins that are characteristically expressed in long-term lymphoblastoid cell lines; six of the proteins are nuclear proteins and two are integral membrane proteins. Current intense study of the interaction of these proteins with B-lymphocyte proteins will increase our understanding of the mechanisms of EBV-induced normal B-lymphocyte activation, proliferation, and survival.

Neuropathy The EBV neuropathies may present as acute chiasmal neuritis104 or acute ophthalmoplegia15,111; the latter may occur in concert with ataxia and areflexia, which constitutes Fisher syndrome.117 Based on serologic evidence of recent infection,114 EBV may produce a lumbosacral radiculoplexopathy; in that study, pain was a prominent feature in all cases and the prognosis was generally good, with recovery in weeks to months. Acute autonomic neuropathy (AAN) has also been linked to EBV infection by the development of heterophilic or EBV-specific antibodies in serum.41,44,137 More specific confirmation that EBV causes AAN was provided by the detection of both EBV DNA and antibody to EBV in CSF of a patient with AAN.13 The pathology of EBV-associated neuropathy is unknown because cases are not fatal. In a few instances of EBV-associated autonomic and sensory neuropathy, sural nerve biopsy revealed severe loss of myelinated and unmyelinated nerve fibers.44,126 In one case of the cohort of Suarez et al.,126 biopsy revealed epineural perivascular inflammation.

Treatment No antiviral agents have been proven to be effective in any conditions associated with EBV infection. In one case of AAN in which nerve biopsy revealed inflammation, the patient was treated with steroids and failed to improve, but did respond to intravenous IgG.13 Optimal treatment of acute EBV neuropathy awaits further studies.

CYTOMEGALOVIRUS CMV infection in immunocompetent individuals either is asymptomatic or results in a mononucleosis-like syndrome. After primary infection, CMV becomes latent. CMV is the most common opportunistic pathogen in immunocompromised individuals, in whom reactivation leads to infection in multiple organs, particularly the lung, liver, and nervous system. CMV infection in children born to non–CMVimmune mothers may produce hearing loss and mental retardation.

Neurologic Complications of Infection CMV is the most common fetal infection. In newborns with congenital CMV infection, 1% to 5% have severe mental retardation associated with microcephaly, CNS calcification, sensorineural hearing loss, microphthalmia, and visual impairment resulting from chorioretinitis and optic atrophy. In adults, CMV infects both the central and peripheral nervous systems,18,88 almost exclusively in immunocompromised individuals such as organ transplant recipients and patients with human immunodeficiency virus (HIV) infection. The most common presentation is subacute or chronic encephalitis, typically associated with evidence of disseminated disease in the eye, lung, or gastrointestinal tract.5,97,132 Before the advent of highly active antiretroviral therapy (HAART), CMV infection was found in up to 25% of HIV-infected patients at autopsy.18,130 CMV meningoencephalitis is rare in immunocompetent individuals.24,79,125

Virology CMV has been formally designated human herpesvirus 5 (HHV-5). The viral genome consists of a linear DNA molecule containing approximately 230,000 nucleotide base pairs, the largest genome of any human virus, and greater than 200 potential ORFs. CMV has a highly restricted host range, and in cell culture replicates preferentially in human and some primate fibroblasts and endothelial cells. Viral nucleocapsids account for the basophilic “owl’s eye” nuclear inclusions seen by light microscopy in CMVinfected cells. Viral genome replication utilizes a virally encoded DNA polymerase and involves circularization of the viral genome within the host-cell nucleus and a “rolling circle” mode of replication. The CMV-encoded DNA polymerase is the target of all currently available anti-CMV drugs, including ganciclovir, foscarnet, and cidofovir. Because CMV does not encode a thymidine kinase required to generate the antivirally active form of acyclovir, this drug and its oral analogues do not effectively inhibit CMV infection.

2122

Diseases of the Peripheral Nervous System

Cranial Neuropathies The most common ophthalmologic manifestation of CMV reactivation is a necrotic retinitis, seen almost exclusively in immunocompromised patients. CMV-associated optic neuritis develops in HIV-infected patients and in approximately 4% of patients with CMV retinitis.87,101 Most patients have severe visual impairment, with funduscopic changes including blurred disc margins and disc edema. Visual acuity may improve after ganciclovir treatment, although relapses are common if therapy is stopped or reduced. An immunocompetent woman with bilateral CMV-associated optic neuropathy (papillitis) improved after treatment with ganciclovir and steroids.6

Myeloradiculopathy CMV myeloradiculopathy usually occurs in HIV-infected patients,12,23,32,39,58,86,96 and rarely in immunocompetent individuals.7,71,93,131 The incidence of CMV polyradiculopathy has decreased since HAART therapy for HIV infection.18,45 Clinical features include radicular back and perianal pain, rapidly progressive ascending flaccid leg weakness, and both urinary retention and bowel dysfunction with decreased or absent reflexes.1 Proprioceptive sensory loss is unusual, although paresthesias and decreased pain and thermal sensation may occur. The CSF often contains hundreds of cells,1,95 although rarely does the cell count exceed 1000.32,46,70,128 The predominant cell in CSF is either mononuclear or polymorphonuclear.95,120 CMV infection must always be ruled out in an HIV-infected patient whose CSF shows a polymorphonuclear pleocytosis.60 The CSF protein is usually elevated, and severe hypoglycorrhachia may occur.32 Magnetic resonance imaging often reveals enhancing lesions in the conus, cauda equina, and lumbosacral nerve roots.11,128 Electrophysiologic studies show evidence of both axonal injury and demyelination with reduced amplitude of both motor and sensory action potentials.18,45 A decline in sensory and motor amplitudes may be a clue to diagnosis, because this is unusual in most nerve root diseases, and reflects the propensity of CMV to injure dorsal root ganglion cells. In the acute phase of illness, before wallerian degeneration of the nerve roots develops, electrodiagnostic studies may be normal. Diagnosis is confirmed by detection of CMV DNA in CSF by PCR, which is positive in greater than 90% of cases.58,95 CMV can also be cultured from CSF. Typical pathologic changes include a prominent polymorphonuclear and mononuclear cell infiltrate involving the sacral spinal cord, cauda equina, and lumbosacral nerve roots associated with both demyelination and axonal destruction. CMV antigen is present in Schwann cells of involved roots and endothelial cells, and acute Schwann cell death may occur.12,18,28,39,110 CMV antigen can also be detected in

the spinal cord, often in association with necrotizing changes. Some patients improve or stabilize after treatment with ganciclovir or foscarnet or a combination of the two.1,2,31,34,35,45,46,61,70,72,76,96,120 Ganciclovir-resistant strains of CMV may emerge.119,129 The prognosis is poor for patients with CMV polyradiculopathy, in whom survival is usually 5.4 ⫾ 1.8 weeks in patients not receiving ganciclovir compared to 14.6 ⫾ 9.4 weeks in those receiving the drug.1

Mononeuritis Multiplex CMV in AIDS patients can present as mononeuritis multiplex, with infarction of cranial and peripheral nerves.63,110 Most patients have an initial onset of multifocal pain or asymmetrical paresthesias that frequently resolve spontaneously. Sensory symptoms are followed by progressive weakness of the arms and legs. Electrophysiologic abnormalities include low motor and sensory amplitudes, consistent with an axonal process. The CSF protein may be elevated, and approximately 30% of patients have a CSF pleocytosis. PCR reveals CMV DNA in greater than 90% of patients,109 and CSF culture is positive for CMV in 15%. Nerve biopsies show polymorphonuclear cells associated with multifocal necrotic endoneurial lesions and frank necrotizing arteritis with fibrinoid necrosis.109 CMV antigen and owl’s eye inclusions may be seen in both infiltrating inflammatory and Schwann cells.109,110 Sometimes, CMV inclusions are found predominantly in endothelial cells of the vasa nervorum rather than in Schwann cells. Unlike those with CMV radiculomyelitis, approximately 35% of mononeuritis multiplex patients improve or recover completely after treatment.109

CMV and the Guillain-Barré Syndrome CMV infection may precede the Guillain-Barré syndrome in both immunocompetent and immunocompromised individuals.37,64,113,133 Several large serosurveys showed that 13% to 33% of these patients had evidence of recent CMV infection.36,37,69 The mechanism by which CMV causes Guillain-Barré syndrome may be molecular mimicry. For example, CMV-infected fibroblasts express GM2-like ganglioside epitopes, and antibody directed against GM2 could lead to the development of the syndrome.3 In fact, anti-GM2 antibodies have been detected in sera from approximately 40% of Guillain-Barré patients with recent CMV infection.68,75

Diagnosis PCR has become the diagnostic procedure of choice for CMV-associated neurologic disease4,18,19,21,59,136 and is often positive when cultures are negative.68,75 Quantitative

2123

Herpesvirus Infection and Peripheral Neuropathy

PCR studies of CMV DNA in CSF reveal high DNA levels, exceeding 10 million copies of CMV DNA /mL of CSF, in patients with polyradiculopathy and ventriculoencephalitis.4,18,115 Information about the sensitivity of CSF PCR in CMV-associated distal symmetrical neuropathy and mononeuritis multiplex is lacking. Therefore, diagnosis depends on demonstration of CMV antigens by immunocytochemistry in biopsy or autopsy specimens of affected nerves.

Treatment Controlled clinical trials of antiviral therapy in CMVassociated neurologic disease are lacking; most reports are from individual cases.5,18,22,40,103 The two first-line drugs are ganciclovir and foscarnet, with the former generally preferred in most settings. A typical induction regimen for ganciclovir involves 14 to 21 days of intravenous therapy with doses of 5 mg/kg every 12 hours. Intravenous foscarnet at a dose of 90 mg/kg every 12 hours provides an acceptable alternative. Patients who fail to respond to either foscarnet or ganciclovir alone may respond to therapy with both drugs.72 Ganciclovir or foscarnet therapy for 14 to 21 days is likely to be sufficient in immunocompetent individuals and in most organ transplant recipients, assuming an adequate clinical and virologic response has occurred. Maintenance therapy is usually not required for CMV CNS infections in organ transplant recipients, although maintained vigilance for recurrent disease is essential. In patients with HIV infection, it is critical that HAART be initiated or optimized concomitantly with antiCMV therapy. HIV patients invariably require maintenance therapy with either intravenous ganciclovir or foscarnet. Recently, a valine ester of ganciclovir (valganciclovir) has become available that can be taken orally. Induction and maintenance doses of both intravenous and oral ganciclovir and intravenous foscarnet require adjustment in patients with renal insufficiency. The major doselimiting side effect is neutropenia with ganciclovir and nephrotoxicity with foscarnet. Nephrotoxicity can also occur with ganciclovir but less commonly than with intravenous acyclovir or foscarnet. Neurotoxicity can occur with both drugs, and in the case of foscarnet is frequently related to electrolyte alterations (especially hypocalcemia). Patients whose HIV infection responds to HAART and who consistently (⬎6 months) demonstrate nondetectable HIV viral loads and CD4 cell counts greater than 100 cells/mm3 may be able to discontinue maintenance therapy.

ACKNOWLEDGMENTS This work was supported in part by Public Health Service grants NS 32623 and AG 06127 from the National Institutes of Health (D. H. G.); and VA Merit and Reap

Grants, and U.S. Army Grant DAMD17-98-1-8615 (K. L. T.). We thank Marina Hoffman for editorial review and Cathy Allen for preparing the manuscript.

REFERENCES 1. Anders, H. J., and Goebel, F. D.: Cytomegalovirus polyradiculopathy in patients with AIDS. Clin. Infect. Dis. 27:345, 1998. 2. Anders, H. J., Weiss, N., Bogner, J. R., and Goebel, F. D.: Ganciclovir and foscarnet efficacy in AIDS-related polyradiculopathy. J. Infect. 36:29, 1998. 3. Ang, C. W., Jacobs, B. C., Brandenburg, A. H., et al.: Crossreactive antibodies against GM2 and CMV-infected fibroblasts in Guillain-Barre syndrome. Neurology 54:1453, 2000. 4. Arribas, J., Clifford, D. B., Fichtenbaum, C. J., et al.: Level of cytomegalovirus (CMV) DNA in cerebrospinal fluid of subjects with AIDS and CMV infection of the central nervous system. J. Infect. Dis. 172:527, 1995. 5. Arribas, J., Storch, G. A., Clifford, D. B., and Tselis, A. C.: Cytomegalovirus encephalitis. Ann. Intern. Med. 125:577, 1996. 6. Baglivo, E., Leuenberger, P. M., and Krause, K. H.: Presumed bilateral cytomegalovirus-induced optic neuropathy in an immunocompetent person. J. Neuroophthalmol. 16:14, 1996. 7. Baig, S. M., and Khan, M. A.: Cytomegalovirus-associated transverse myelitis in a non-immunocompromised patient. J. Neurol. Sci. 134:210, 1995. 8. Baringer, J. R., and Pisani, P.: Herpes simplex virus genomes in human nervous system tissue analyzed by polymerase chain reaction. Ann. Neurol. 36:823, 1994. 9. Baringer, J. R., and Swoveland, P.: Recovery of herpessimplex virus from human trigeminal ganglions. N. Engl. J. Med. 288:648, 1973. 10. Bastian, F.O., Rabson, A. S., Yee, C. L., and Tralka, T. S.: Herpesvirus hominis: isolation from human trigeminal ganglion. Science 178:306, 1972. 11. Bazan, C., Jackson, C., Jinkins, J. R., and Barohn, R. J.: Gadolinium enhanced MRI in a case of cytomegalovirus polyradiculopathy. Neurology 41:1522, 1991. 12. Behar, R., Wiley, C., and McCutchan, J. A.: Cytomegalovirus polyradiculoneuropathy in acquired immune deficiency syndrome. Neurology 37:557, 1987. 13. Bennett, J. L., Mahalingam, R., Wellish, M. C., and Gilden, D. H.: Epstein-Barr virus-associated acute autonomic neuropathy. Ann. Neurol. 40:453, 1996. 14. Berger, R., Florent, G., and Just, M.: Decrease of the lymphoproliferative response to varicella-zoster virus antigen in the aged. Infect. Immun. 32:24, 1981. 15. Brey, R. L.: Ophthalmoplegia polyneuropathy: another case in association with Epstein-Barr virus. Ann. Neurol. 58:403, 1984. 16. Bustos, D. E., and Atherton, S. S.: Detection of herpes simplex virus type 1 in human ciliary ganglia. Invest. Ophthalmol. 43:2244, 2002. 17. Chen, X. P., Mata, M., Kelley, M., et al.: The relationship of herpes simplex virus latency associated transcript expression to genome copy number: a quantitative study using laser capture microdissection. J. Neurovirol. 8:204, 2002.

2124

Diseases of the Peripheral Nervous System

18. Cinque, P., Cleator, G. M., Weber, T., et al.: Diagnosis and clinical management of neurological disorders caused by cytomegalovirus in AIDS patients. European Union Concerted Action on Virus Meningitis and Encephalitis. J. Neurovirol. 4:120, 1998. 19. Cinque, P., Vago, L., Brytting, M., et al.: Cytomegalovirus infection of the central nervous system in patients with AIDS: diagnosis by DNA amplification from cerebrospinal fluid. J. Infect. Dis. 166:1408, 1992. 20. Clarke, P., Beer, T., Cohrs, R., and Gilden, D. H.: Configuration of latent varicella-zoster virus DNA. J. Virol. 69:8151, 1995. 21. Clifford, D. B., Buller, R. S., Mohammed, S., et al.: Use of polymerase chain reaction to demonstrate cytomegalovirus DNA in CSF of patients with human immunodeficiency virus infection. Neurology 43:75, 1993. 22. Cohen, B. A.: Prognosis and response to therapy of cytomegalovirus encephalitis and meningomyelitis in AIDS. Neurology 46:444, 1996. 23. Cohen, B. A., McArthur, J. C., Grohman, S., et al.: Neurologic prognosis of cytomegalovirus polyradiculomyelopathy in AIDS. Neurology 43:493, 1993. 24. Cohen, J. I., and Corey, G. R.: Cytomegalovirus infection in the normal host. Medicine 64:100, 1985. 25. Cohrs, R. J., Barbour, M., and Gilden, D. H.: Varicella-zoster virus (VZV) transcription during latency in human ganglia: detection of transcripts mapping to genes 21, 29, 62, and 63 in a cDNA library enriched for VZV RNA. J. Virol. 70:2789, 1996. 26. Cohrs, R. J., Gilden, D. H., Kinchington, P. R., et al.: Varicella-zoster virus gene 66 transcription and translation in latently infected human ganglia. J. Virol. 77:6660, 2003. 27. Cohrs, R. J., Randall, J., Smith, J., et al.: Analysis of individual human trigeminal ganglia for latent herpes simplex virus type 1 and varicella-zoster virus nucleic acids using real-time PCR. J. Virol. 74:11464, 2000. 28. Cornford, M. E., Ho, W. W., and Vinters, H. V.: Correlation of neuromuscular pathology in acquired immune deficiency syndrome patients with cytomegalovirus infection and zidovudine treatment. Acta Neuropathol. 84:516, 1992. 29. Davison, A. J.: Varicella-zoster virus. The Fourteenth Fleming Lecture. J. Gen. Virol. 72:475, 1991. 30. Davison, A. J., and Scott, J. E.: The complete DNA sequence of varicella-zoster virus. J. Gen. Virol. 67:1759, 1986. 31. de Gans, J., Portegies, P., Tiessens, G., et al.: Therapy for cytomegalovirus polyradiculomyelitis in patients with AIDS: treatment with ganciclovir. AIDS 4:421, 1990. 32. de Gans, J., Tiessens, G., Portegies, P., et al.: Predominance of polymorphonuclear leukocytes in cerebrospinal fluid of AIDS patients with cytomegalovirus polyradiculomyelitis. J. Acquir. Immune Defic. Syndr. 3:1155, 1990. 33. de Silva, S. M., Mark, A. S., Gilden, D. H., et al.: Zoster myelitis: improvement with antiviral therapy in two cases. Neurology 47:929, 1996. 34. Decker, C. F., Tarver, J. H., Murray, D. F., and Martin, G. J.: Prolonged concurrent use of ganciclovir and foscarnet in the treatment of polyradiculopathy due to cytomegalovirus in a patient with AIDS. Clin. Infect. Dis. 19:548, 1994. 35. Domingo, P., Puig, M., Iranzo, A., et al.: Polyradiculopathy due to cytomegalovirus infection: report of a case in which an AIDS patient responded to foscarnet. J. Infect. Dis. 18:1019, 1994.

36. Dowling, P., Menonna, J., and Cook, S.: Cytomegalovirus complement fixation antibody in Guillain-Barre syndrome. Neurology 27:1153, 1977. 37 Dowling, P. C., and Cook, S. D.: Role of infection in GuillainBarre syndrome: laboratory confirmation of herpesviruses in 41 cases. Ann. Neurol. 9(Suppl.):44, 1981. 38. Efstathiou, S., Minson, A. C., Field, H. J., et al.: Detection of herpes simplex virus-specific DNA sequences in latently infected mice and in humans. J. Virol. 57:446, 1986. 39. Eidelberg, D., Sotrel, A., Vogel, H., et al.: Progressive polyradiculopathy in acquired immunodeficiency syndrome. Neurology 36:912, 1986. 40. Enting, R., de Gans, J., Reiss, P., et al.: Ganciclovir/foscarnet for cytomegalovirus meningoencephalitis in AIDS. Lancet 340:559, 1992. 41. Estañol-Vidal, B., Perez-Ortega, R., Vargas-Lugo, B., et al.: Acute autonomic neuropathy. Arch. Invest. Med. (Mex.) 10:53, 1979. 42. Fox, R. J., Galetta, S. L., Mahalingam, R., et al.: Acute, chronic, and recurrent varicella zoster virus neuropathy without zoster rash. Neurology 57:351, 2001. 43. Fraser, N. W., Lawrence, W. C., Wroblewska, Z., et al.: Herpes simplex type 1 DNA in human brain tissue. Proc. Natl. Acad. Sci. U. S. A. 79:6461, 1981. 44. Fujii, N., Tabira, T., Shibasaki, H., et al.: Acute autonomic and sensory neuropathy associated with elevated EpsteinBarr virus antibody titre. J. Neurol. Neurosurg. Psychiatry 45:656, 1982. 45. Fuller, G. N.: Cytomegalovirus and the peripheral nervous system in AIDS. J. Acquir. Immune Defic. Syndr. 5:S33, 1992. 46. Fuller, G. N., Gill, S. K., Guiloff, R. K., et al.: Ganciclovir for lumbosacral polyradiculopathy in AIDS. Lancet 335:48, 1990. 47. Gilden, D. H.: Herpes zoster with postherpetic neuralgia— persisting pain and frustration. N. Engl. J. Med. 303:932, 1994. 48. Gilden, D. H.: Varicella zoster virus vasculopathy and disseminated encephalomyelitis. J. Neurol. Sci. 195:99, 2002. 49. Gilden, D. H., Beinlich, B. R., Rubinstein, E. M., et al.: Varicella zoster virus myelitis: an expanding spectrum. Neurology 44:1818, 1994. 50. Gilden, D. H., Cohrs, R. J., Hayward, A. R., et al.: Chronic varicella zoster virus ganglionitis—a possible cause of postherpetic neuralgia. J. Neurovirol. 9:404, 2003. 51. Gilden, D. H., Dueland, A. N., Cohrs, R., et al.: Preherpetic neuralgia. Neurology 41:1215, 1991. 52. Gilden, D. H., Gesser, R., Smith, J., et al.: Presence of VZV and HSV-1 DNA in human nodose and celiac ganglia. Virus Genes 23:145, 2001. 53. Gilden, D. H., and Katz, R. I.: Surgical induction of contralateral homologous dermatomal distribution zoster. Arch. Neurol. 60:616, 2003. 54. Gilden, D. H., Mahalingam, R., Cohrs, R. J., et al.: The protean manifestations of varicella zoster virus vasculopathy. J. Neurovirol. 8(Suppl. 2):75, 2002. 55. Gilden, D. H., Rozenman, Y., Murray, R., et al.: Detection of varicella-zoster virus nucleic acid in neurons of normal human thoracic ganglia. Ann. Neurol. 22:377, 1987. 56. Gilden, D. H., Vafai, A., Shtram, Y., et al.: Varicella-zoster virus DNA in human sensory ganglia. Nature 306:478, 1983. 57. Gilden, D. H., Wright, R. R., Schneck, S. A., et al.: Zoster sine herpete, a clinical variant. Ann. Neurol. 35:530, 1994.

Herpesvirus Infection and Peripheral Neuropathy 58. Glass, J. D., and Erozan, Y. S.: Rapid diagnosis of cytomegalovirus polyradiculitis in a patient with acquired immunodeficiency syndrome. Ann. Neurol. 34:239, 1993. 59. Gozlan, J., Salord, J.-M., Roullet, E., et al.: Rapid detection of CMV DNA in CSF of AIDS patients with neurologic disorders. J. Infect. Dis. 166:1416, 1992. 60. Granter, S. R., Doolittle, M. H., and Renshaw, A. A.: Predominance of neutrophils in the cerebrospinal fluid of AIDS patients with cytomegalovirus radiculopathy. Am. J. Clin. Pathol. 105:364, 1996. 61. Graveleau, P., Perol, R., and Chapman, A.: Regression of cauda equina syndrome in AIDS patients being treated with ganciclovir. Lancet 2:511, 1989. 62. Grogan, P. M., and Gonseth, G. S.: Practice parameter. Steroids, acyclovir, and surgery for Bell’s palsy (an evidence based review): report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 56:830, 2001. 63. Harada, H., Tamaoka, A., Yoshida, H., et al.: Horner’s syndrome associated with mononeuritis multiplex due to cytomegalovirus as the initial manifestation in a patient with AIDS. J. Neurol. Sci. 154:91, 1998. 64. Hart, I. K., and Kennedy, P. G. E.: Guillain-Barre syndrome associated with cytomegalovirus infection. Q. J. Med. 67:425, 1988. 65. Hayakawa, Y., and Hyman, R. W.: Isomerization of the UL region of varicella-zoster virus DNA. Virus Res. 8:25, 1987. 66. Hinthorn, D. R., Baker, L. H., Romig, D. A., and Liu, C.: Recurrent conjugal neuralgia caused by herpesvirus hominis type 2. JAMA 236:587, 1976. 67. Hyman, R. W., Ecker, J. R., and Tenser, R. B.: Varicellazoster virus RNA in human trigeminal ganglia. Lancet 2:814, 1983. 68. Irie, S., Saito, T., Nakamura, K., et al.: Association of antiGM2 antibodies in Guillain-Barre syndrome with acute cytomegalovirus infection. J. Neuroimmunol. 68:19, 1996. 69. Jacobs, B. C., Rothbart, P. H., van der Meche, F. G., et al.: The spectrum of antecedent infections in Guillain-Barre syndrome: a case control study. Neurology 51:1110, 1998. 70. Jacobson, M. A., Mills, J., Rush, J., et al.: Failure of antiviral therapy for acquired immunodeficiency syndrome-related cytomegalovirus myelitis. Arch. Neurol. 45:1090, 1988. 71. Kabins, S., Keller, R., Naraqi, S., and Peitchel, R.: Viral ascending radiculomyelitis with severe hypoglycorrhachia. Arch. Intern. Med. 136:933, 1976. 72. Karmochkine, M., Molina, J.-M., Scieux, C., et al.: Combined therapy with ganciclovir and foscarnet for cytomegalovirus polyradiculomyelitis in patients with AIDS. Am. J. Med. 97:196, 1994. 73. Kennedy, P. G. E., Grinfeld, E., and Gow, J. W.: Latent varicella-zoster virus is located predominantly in neurons in human trigeminal ganglia. Proc. Natl. Acad. Sci. U. S. A. 95:4658, 1998. 74. Kennedy, P. G. E., Grinfeld, E., and Gow, J. W.: Latent varicella-zoster virus in human dorsal root ganglia. Virology 258:451, 1999. 75. Khalili-Shirazi, A., Gregson, N., Gray, I., et al.: Antiganglioside antibodies in Guillain-Barre syndrome after recent cytomegalovirus infection. J. Neurol. Neurosurg. Psychiatry 66:376, 1999.

2125

76. Kim, Y. S., and Hollander, H.: Polyradiculopathy due to cytomegalovirus: report of two cases in which improvement occurred after prolonged therapy and review of the literature. Clin. Infect. Dis. 17:32, 1993. 77. Kinchington, P. R., and Cohen, J. I.: Viral proteins. In Arvin, A. M., and Gershon, A. A. (eds.): Varicella-Zoster Virus. Cambridge, UK, Cambridge University Press, p. 74, 2000. 78. Kinchington, P. R., Reinhold, W. C., Casey, T. A., et al.: Inversion and circularization of the varicella-zoster virus genome. J. Virol. 56:194, 1985. 79. Klemola, E., Kaariainen, L., von Essen, R., et al.: Further studies on cytomegalovirus mononucleosis in previously healthy individuals. Acta Med. Scand. 182:311, 1967. 80. Krohel, G. B., Richardson, J. R., and Farrell, D. F.: Herpes simplex neuropathy. Neurology 26:596, 1976. 81. LaGuardia, J. J., Cohrs, R. J., and Gilden, D. H.: Prevalence of varicella-zoster virus DNA in dissociated human trigeminal ganglion neurons and nonneuronal cells. J. Virol. 73:8571, 1999. 82. Maguire, H. F., and Hyman, R. W.: Polyadenylated, cytoplasmic transcripts of varicella-zoster virus. Intervirology 26:181, 1986. 83. Mahalingam, R., Wellish, M., Cohrs, R., et al.: Expression of protein encoded by varicella-zoster virus open reading frame 63 in latently infected human ganglionic neurons. Proc. Natl. Acad. Sci. U. S. A. 93:2122, 1996. 84. Mahalingam, R., Wellish, M., Wolf, W., et al.: Latent varicella zoster virus DNA in human trigeminal and thoracic ganglia. N. Engl. J. Med. 323:627, 1990. 85. Mahalingam, R., Wellish, M. C., Dueland, A. N., et al.: Localization of herpes simplex virus and varicella zoster virus DNA in human ganglia. Ann. Neurol. 31:444, 1992. 86. Mahieux, F., Gray, F., Fenelon, G., et al.: Acute myeloradiculitis due to cytomegalovirus infection of peripheral nerves. J. Neurol. Neurosurg. Psychiatry 52:270, 1989. 87. Mansor, A. M., and Li, H. K.: Cytomegalovirus optic neuritis: characteristics, therapy, and survival. Ophthalmologica 209:260, 1995. 88. McCutchan, J. A.: Cytomegalovirus infections of the nervous system in patients with AIDS. Clin. Infect. Dis. 20:747, 1995. 89. McGeoch, D. J., Dalrymple, M. A., Davison, A. J., et al.: The complete DNA sequence of the long unique region in the genome of herpes simplex virus type 1. J. Gen. Virol. 69:1531, 1988. 90. McGeoch, D. J., Dolan, A., Donald, S., and Brauer, D. H.: 1986. Complete DNA sequence of the short repeat region in the genome of herpes simplex virus type 1. Nucleic Acids Res. 14:1727, 1986. 91. McGeoch, D. J., Dolan, A., Donald, S., and Rixon, F. J.: Sequence determination and genetic content of the short unique region in the genome of herpes simplex virus type 1. J. Mol. Biol. 181:1, 1985. 92. Mellerick, D. M., and Fraser, N. W.: Physical state of the latent herpes simplex virus genome in a mouse model system: evidence suggesting an episomal state. Virology 158:265, 1987. 93. Miles, C., Hoffman, W., Lai, C.-W., and Freeman, J. W.: Cytomegalovirus-associated transverse myelitis. Neurology 43:2143, 1993.

2126

Diseases of the Peripheral Nervous System

94. Miller, A. E.: Selective decline in cellular immune response to varicella-zoster in the elderly. Neurology 30:582, 1980. 95. Miller, R. F., Fox, J. D., Thomas, P., et al.: Acute lumbosacral polyradiculopathy due to cytomegalovirus in advanced HIV disease: CSF findings in 17 patients. J. Neurol. Neurosurg. Psychiatry 61:456, 1996. 96. Miller, R. G., Storey, J. R., and Greco, C. M.: Ganciclovir in treatment of progressive AIDS-related polyradiculopathy. Neurology 40:569, 1990. 97. Morgello, S., Cho, E.-S., Nielsen, S., et al.: Cytomegalovirus encephalitis in patients with acquired immunodeficiency syndrome: an autopsy study of 30 cases and a review of the literature. Hum. Pathol. 18:289, 1987. 98. Morris, H. H. III, and Peters, B. H.: Recurrent sciatica associated with herpes simplex. J. Neurosurg. 41:97, 1974. 99. Murakami, S., Mizobuchi, M., Nakashiro, Y., et al.: Bell palsy and herpes simplex virus: identification of viral DNA in endoneurial fluid and muscle. Ann. Intern. Med. 124:27, 1996. 100. Ostrove, J. M., Reinhold, W., Fan, C. M., et al.: Transcription mapping of the varicella-zoster virus genome. J. Virol. 56:600, 1985. 101. Patel, S. S., Rutzen, A. R., Marx, J. L., et al.: Cytomegalovirus papillitis in patients with acquired immunodeficiency syndrome. Ophthalmology 103:1476, 1996. 102. Perry, L. J., and McGeoch, D. J.: The DNA sequences of the long repeat region and adjoining parts of the long unique region in the genome of herpes simplex virus type 1. J. Gen. Virol. 69:2831, 1988. 103. Price, T. A., Digioia, R. A., and Simon, G. L.: Ganciclovir treatment of cytomegalovirus ventriculitis in a patient infected with human immunodeficiency virus. Clin. Infect. Dis. 15:606, 1992. 104. Purvin, V., Herr, G. J., and De Myer, W.: Chiasmal neuritis as a complication of Epstein-Barr virus infection. Arch. Neurol. 45:458, 1988. 105. Quan, D., Wellish, M., and Gilden, D. H.: Topical ketamine treatment of postherpetic neuralgia. Neurology 60:1391, 2003. 106. Reinhold, W. C., Straus, S. E., and Ostrove, J. M.: Directionality and further mapping of varicella zoster virus transcripts. Virus Res. 9:249, 1988. 107. Rock, D. L., and Fraser, N. W.: Latent herpes simplex virus type 1 DNA contains two copies of the virion DNA joint region. J. Virol. 55:849, 1985. 108. Roizman, B., and Knipe, D. M.: Herpes simplex viruses and their replication. In Knipe, D. M., and Howley, P. M. (eds.): Fields Virology, 4th ed. Philadelphia, Lippincott Williams & Wilkins, p. 2399, 2001. 109. Roullet, E., Assuerus, V., Gozlan, J., et al.: Cytomegalovirus multifocal neuropathy in AIDS: analysis of 15 consecutive cases. Neurology 44:2174, 1994. 110. Said, G., Lacroix, C., Chemouilli, P., et al.: Cytomegalovirus neuropathy in acquired immunodeficiency syndrome: a clinical and pathological study. Ann. Neurol. 29:139, 1991. 111. Salazar, A., Martinez, H., and Sotelo, J.: Ophthalmoplegic polyneuropathy associated with infectious mononucleosis. Ann. Neurol. 13:219, 1983.

112. Sawtell, N. M.: Comprehensive quantification of herpes simplex virus latency at the single-cell level. J. Virol. 71:5423, 1997. 113. Schmitz, H., and Enders, G.: Cytomegalovirus as a frequent cause of Guillain-Barre syndrome. J. Med. Virol. 1:21–27, 1977. 114. Sharma, K. R., Sriram, S., Fries, T., et al.: Lumbosacral radiculoplexopathy as a manifestation of Epstein-Barr virus infection. Neurology 43:2550, 1993. 115. Shinkai, M., and Spector, S. A.: Quantitation of human cytomegalovirus (HCMV) DNA in cerebrospinal fluid by competitive PCR in AIDS patients with different HCMV central nervous system diseases. Scand. J. Infect. Dis. 27:559, 1995. 116. Sipe, J., and Dunn, L.: Acyclovir for Bell’s palsy (idiopathic facial paralysis). Cochrane Database Syst. Rev. 4:CD001869, 2001. 117. Slavick, H. E., and Shapiro, R. A.: Fisher’s syndrome associated with Epstein-Barr virus. Arch. Neurol. 38:134, 1981. 118. Slavin, H. B., and Ferguson, J. J. Jr.: Zoster-like eruptions caused by the virus of herpes simplex. Am. J. Med. 8:456, 1950. 119. Smith, I. L., Shinkay, M., Freeman, W. R., and Spector, S. A.: Polyradiculopathy associated with ganciclovir-resistant cytomegalovirus in an AIDS patient: phenotypic and genotypic characterization of sequential virus isolates. J. Infect. Dis. 173:148, 1996. 120. So, Y. T., and Olney, R. K.: Acute lumbosacral polyradiculopathy in acquired immunodeficiency syndrome: experience in 23 patients. Ann. Neurol. 35:53, 1994. 121. Spivack, J. G., and Fraser, N. W.: Detection of herpes simplex virus type 1 transcripts during latent infection in mice. J. Virol. 61:3841, 1987. 122. Steiner, I., and Mattan, Y.: Bell’s palsy and herpes viruses: to (acyclo)vir or not to (acyclo)vir? J. Neurol. Sci. 170:19, 1999. 123. Stevens, J. G., Wagner, E. K., Davi-Rao, G. B., et al.: RNA complementary to a herpesvirus alpha gene mRNA is prominent in latently infected neurons. Science 235:1056, 1987. 124. Straus, S. E., Owens, J., Ruyechan, W. T., et al.: Molecular cloning and physical mapping of varicella-zoster virus DNA. Proc. Natl. Acad. Sci. U. S. A. 79:993, 1982. 125. Studahl, M., Ricksten, A., Sandburg, T., and Burgstrom, T.: Cytomegalovirus encephalitis in four immunocompetent patients. Lancet 340:1045, 1992. 126. Suarez, G. A., Fealey, R. D., Camilleri, M., and Low, P. A.: Idiopathic-autonomic neuropathy: clinical, neurophysiologic, and follow-up studies on 27 patients. Neurology 44:1675, 1994. 127. Suzuki, F., Furuta, Y., Ohtani, F., et al.: Herpes virus reactivation and gadolinium-enhanced magnetic resonance imaging in patients with facial palsy. Otol. Neurotol. 22:549, 2001. 128. Talpos, D., Tien, R. D., and Hesselink, J. R.: Magnetic resonance imaging of AIDS-related polyradiculopathy. Neurology 41:1995, 1991. 129. Tokumoto, J. I., and Hollander, H.: Cytomegalovirus polyradiculopathy caused by a ganciclovir-resistant strain. Clin. Infect. Dis. 17:854, 1993.

Herpesvirus Infection and Peripheral Neuropathy 130. Tselis, A., and Lavi, E.: Cytomegalovirus infection of the adult nervous system. In Davis, L. E., and Kennedy, P. G. E. (eds.): Infectious Diseases of the Nervous System. Oxford, UK, Butterworth-Heinemann, p. 109, 2000. 131. Tyler, K. L., Gross, R. A., and Cascino, G. D.: Unusual viral causes of transverse myelitis: hepatitis A and cytomegalovirus. Neurology 36:855, 1986. 132. Vinters, H. V., Kwok, M. K., Ho, H. W., et al.: Cytomegalovirus in the nervous system of patients with the acquired immunodeficiency syndrome. Brain 112:245, 1989. 133. Visser, L. H., van der Meche, F. G., Meulstee, J., et al.: Cytomegalovirus infection and Guillain-Barre syndrome: the clinical, electrophysiological, and prognostic factors. Dutch Guillain-Barre Study Group. Neurology 47:668, 1999.

2127

134. Warren, K. G., Brown, S. M., Wroblewska, Z., et al.: Isolation of latent herpes simplex virus from the superior cervical and vagus ganglions of human beings. N. Engl. J. Med. 298:1068, 1978. 135. Whitley, R. J., Weiss, H., Gnann, J. W. Jr., et al.: Acyclovir with and without prednisone for the treatment of herpes zoster: a randomized, placebo-controlled trial. Ann. Intern. Med. 125:376, 1996. 136. Wolf, D. G., and Spector, S. A.: Detection of human cytomegalovirus central nervous system disease in AIDS patients by DNA amplification from cerebrospinal fluid. J. Infect. Dis. 166:1412, 1992. 137. Yahr, M. D., and Frontera, A. T.: Acute autonomic neuropathy: its occurrence in infectious mononucleosis. Arch. Neurol. 32:132, 1975.

94 Peripheral Neuropathies in Human Immunodeficiency Virus Infection AHMET HÖKE AND DAVID R. CORNBLATH

Epidemiology HIV-Associated Sensory Polyneuropathies Clinical Features Electrophysiology Pathology Pathogenesis of Distal Symmetric Polyneuropathy Pathogenesis of Antiretroviral Toxic Neuropathy Treatment of DSP and ATN

Neuropathies Associated with HIV-1 Seroconversion Inflammatory Demyelinating Polyneuropathy Clinical Features Pathology Therapy Etiology Multiple Mononeuropathy Clinical Features Pathology

Since the introduction of highly active antiretroviral therapy (HAART) in 1996,9,133 central nervous system (CNS) complications of human immunodeficiency virus (HIV) infection have declined dramatically. However, the incidence and prevalence of peripheral nervous system (PNS) complications of HIV infection remain high153 and, in fact, may be increasing. Many types of peripheral neuropathies are seen in HIV patients, but only a few appear to be HIV specific (i.e., not seen in other patient populations). Starting with early observations, it appeared that there was a link between the type of neuropathy and the stage of HIV infection (see below) and that certain non–HIV-specific neuropathies may be overrepresented in HIV-infected patients. Additionally, while the first PNS complications of HIV were described in 1985,106,108 the pathogenesis of most peripheral neuropathies in HIV-infected patients remains unknown. Recently, numerous reviews have been published on the HIV-associated neuropathies,8,33,94,110,134,159,182 but critical reviews of the literature are desperately needed and mechanistic studies are lacking. In this chapter, we critically review the peripheral neuropathies seen in HIVinfected individuals and propose new avenues of research to better understand pathogenesis and promote new ther-

Therapy Etiology Autonomic Neuropathy Neuropathy in Diffuse Infiltrative Lymphomatosis Syndrome Opportunistic Infections of the PNS Cytomegalovirus Herpes Zoster Mycobacterium avium-intracellulare Cryptococcus neoformans

apies. Table 94–1 outlines the major classes of peripheral neuropathies in HIV infection. One of the striking features of the neuropathies that occur in the HIV-infected population is the relative specificity of the individual neuropathic syndromes for stage of HIV infection.40,41,103,124 This disease-stage specificity is likely to reflect different pathogenetic mechanisms of the various syndromes. For example, the inflammatory demyelinating polyneuropathies (IDPs), both acute (Guillain-Barré syndrome [GBS]) and chronic (chronic inflammatory demyelinating polyneuropathy [CIDP]), occur mainly during the early phases of HIV-1 infection, often when individuals are otherwise asymptomatic. These disorders are likely to be immunopathogenic. During the early stages of infection, the immune system is relatively competent, but is stimulated and has altered responsiveness. This may provide the setting for development of autoimmune disorders, including the IDPs. A vasculitic syndrome occurs mainly during the early symptomatic phase of HIV-1 infection and has been hypothesized to be due to circulating immune complexes of HIV-1 antibody and antigen that are deposited in vessel walls.41 Distal symmetrical polyneuropathy (DSP) occurs mainly in the late 2129

2130

Diseases of the Peripheral Nervous System

Table 94–1. Peripheral Neuropathies Associated with HIV Infection HIV-associated sensory neuropathies Distal symmetrical polyneuropathy Antiretroviral toxic neuropathy Inflammatory polyneuropathies Acute inflammatory demyelinating polyneuropathy Chronic inflammatory demyelinating polyneuropathy Mononeuritis multiplex Autonomic neuropathy Neuropathy in diffuse infiltrative lymphomatosis syndrome Neuropathies caused by opportunistic infections

phase of HIV-1 infection (i.e., acquired immunodeficiency syndrome [AIDS]), when there is severe immunoincompetence. At this stage, the mechanisms that have been proposed include productive infection of neural tissue with HIV-181 or cytomegalovirus (CMV),74,125,181 or toxic or metabolic abnormalities resulting from the advanced systemic disease. The syndromes associated with CMV (i.e., polyradiculoneuropathy13,58 and multifocal demyelinating neuropathy67) also occur in the late phase of HIV-1 infection and relate to the loss of immune function, which allows productive infection of neural tissue by CMV. Lastly, there is the toxic neuropathy associated with antiretroviral medications, seen in those taking these medications.

EPIDEMIOLOGY Several studies have examined the incidence of neuropathies at the time of seroconversion or early in the course of the disease. In one of the prospective studies, Alliegro and co-workers found that only 19 (3.1%, or 5.5/1000 personyears) among 621 seroconverters followed for 5.7 years developed neuropathy. Of the 19 patients, 15 developed DSP, 2 vasculitic neuropathy, and 2 GBS.1 In another study,69 Fuller and co-workers followed 1500 HIV-infected individual for 15 months. Involvement of the PNS was seen in 54 patients (about 3%). All of the cases with PNS involvement were patients with advanced HIV infection. DSP was seen in 38 of the 54 patients (11.5% of AIDS cases). Thirteen patients had lumbosacral polyradiculopathy caused by CMV infection, and three had mononeuritis multiplex. No cases of acute inflammatory demyelinating neuropathy (AIDP) or CIDP were seen. In a cross-sectional study of 798 individuals identified through U.S. Air Force HIV screening between 1985 and 1989, only 12 had neuropathy.12 Most of these patients either were asymptomatic or had early stages of HIV infection. All of the cases of neuropathy were seen in HIV patients with advanced disease (i.e., CD4 count ⬍ 300/mL). In addition to these epidemiologic studies, which primarily used patient complaints and physical examination as

criteria for the diagnosis of neuropathy and looked at the incidence of neuropathy at earlier stages of the HIV infection, others have used electrophysiologic studies and examined the incidence of neuropathy in patients with more advanced disease. In the pre-antiretroviral era, Gastaut and co-workers studied the incidence of neuropathy in a full spectrum of HIV-infected patients using electrophysiology and nerve biopsy.70 Overall, the incidence of neuropathy was 89% among 56 patients ranging from asymptomatic carriers to patients with AIDS. In asymptomatic carriers, the incidence of neuropathy was 40% (2 of 5 patients), in contrast to 100% among AIDS patients (29 of 29). Neuropathy was more likely to be asymptomatic (58%) than symptomatic (42%) and usually involved solely sensory fibers. Twenty-eight patients with the most advanced disease underwent sural nerve biopsy, and all were abnormal. The major finding was a reduction in the number of large (⬎9 ␮m) myelinated fibers. In a series of studies, Chavanet and co-workers examined the incidence of neuropathy in asymptomatic patients using electrophysiology.29,30 Overall, the incidence of at least one abnormal electrophysiologic parameter was 35% among the 57 patients studied. The incidence of electrophysiologic abnormalities was higher among the patients with more advanced disease: 60% among AIDS patients versus 25% among asymptomatic carriers. Since the introduction of zidovudine (azidothymidine [AZT]) and zalcitabine (2⬘,3⬘-dideoxycytidine [ddC]), there have been other systematic studies of the incidence and prevalence of neuropathies in HIV patients.10,18,63,69,79,120,149 From these studies, however, it is unclear if the incidence of neuropathy increased with the introduction of nucleoside analogue reverse transcriptase inhibitors (NRTIs). Most prospective studies determined the incidence of neuropathy at 30% to 40%, with advanced disease as the main risk factor.10,18,63 Others found that the development of neuropathy did not correlate with low CD4 counts (⬍200/mL).183 In many of these studies, neuropathy was assessed by a combination of clinical symptoms and electrophysiology. The prevalence of neuropathy or PNS involvement in autopsy series has been much higher, near 100%.50,76 However, these studies suffer from the bias of examining patients with the most advanced disease. The prevalence of neuropathy across all HIV-infected individuals in the HAART era remains unknown. Prior to introduction of HAART, the incidence and prevalence of all neurologic complications of HIV were similar between developed and developing countries. However, since the introduction of HAART, the picture has changed dramatically in the developed countries (reviewed by Sacktor153). The incidence of HIV-associated dementia and opportunistic infections of the CNS has decreased by about 50%.24,54,128,129 Paralleling these changes, the incidence of opportunistic infections of the PNS has also declined. With this, predominantly sensory

Peripheral Neuropathies in Human Immunodeficiency Virus Infection

2131

FIGURE 94–1 Incidence (A) and prevalence (B) of HIV-associated neurologic complications in the Johns Hopkins University HIV clinical cohort. Since the introduction of HAART in 1997, the incidence of neurologic complications has decreased except for ATN. The prevalence of sensory neuropathy (DSP ⫹ ATN) increased substantially, in part as a result of increased use of NRTIs and in part as a result of prolonged life expectancy with the HAART regimen. Despite a decline in incidence, the prevalence of HIV-associated dementia has remained high, in part as a result of prolonged life expectancy with the HAART regimen. (The incidence and prevalence are expressed per 100 patient-years.) (J. McArthur, unpublished data, 2002.)

polyneuropathies became the predominant form of PNS involvement in HIV infection (Fig. 94–1). There are two main forms of predominantly sensory polyneuropathies: neuropathy associated with advanced HIV infection, most commonly called distal symmetrical polyneuropathy, and neuropathy caused by antiretroviral neurotoxicity, most commonly called antiretroviral toxic neuropathy (ATN). Discussed more fully below, these disorders, while clinically and physiologically similar, are different in etiopathogenesis. In many studies they are not differentiated, and instead are usually lumped together as “DSP,” which may lead to confusion in understanding these reports. When information is given on the specific disorder discussed, this is stated. Currently in the HAART era, the incidence of clinically significant DSP, which is mainly manifest as a predominantly sensory polyneuropathy, is as high as 36%, a figure similar to that of the pre-HAART era.22,40,94,130,153 In many studies the main risk factors for development of DSP are low CD4 count, high HIV viral load, and use of certain NRTIs.34 Thus it is difficult to determine what percentage is actually DSP and what percentage is ATN. There are many well-characterized cohorts of HIV patients in the United States, and further studies are needed to determine the prevalence of neuropathy across all patient groups. Schifitto and co-workers examined the development of predominantly sensory neuropathies among a cohort of 272 subjects in the pre-HAART era using clinical examination.157 All subjects were part of a selected cohort in which the entry criteria included cognitive dysfunction. At baseline, 55% of the subjects had evidence of neuropathy (20% were asymptomatic and 35% were symptomatic). The Kaplan-Meier estimate of

the 1-year incidence of symptomatic sensory neuropathy was 36%. The risk factors identified in the study were AIDS-defining illnesses and low CD4 counts. Surprisingly, the use of ddC was not associated with development of sensory neuropathies. This may have been due to the low numbers of subjects with recent or current ddC use (about 25%) and the assignment of subjects with remote use of ddC to the ddC group. In addition, small fiber neuropathy was not assessed with objective measures such as quantitative sensation testing or skin biopsies.

HIV-ASSOCIATED SENSORY POLYNEUROPATHIES Clinical Features The clinical presentation of the predominantly sensory polyneuropathies DSP and ATN is very similar.11,15,40,100, 103,104,124,140,164,186 In both DSP and ATN, the initial symptoms are distal dysesthesias, which start in the toes and slowly progress up the lower extremities. They often involve the ankles by the time the patients seek medical attention. These dysesthesias are often painful and have a burning quality. In a recent study, severity of pain was related to HIV viral load.160 Patients may develop allodynia, in some cases so severe that walking or even the weight of the sheets at night is unbearable. As the disease progresses, dysesthesias may reach up to the knees but rarely involve the hands. At later stages, patients complain of numbness. The distal loss of sensory function is corroborated on examination. Sensory thresholds are raised to all modalities, especially to those testing the small fiber functions. The ankle reflex is almost

2132

Diseases of the Peripheral Nervous System

universally lost. The deep tendon reflexes at the knees are often normal but may be brisk. In such instances the diagnosis of concomitant myelopathy should be considered. Although patients may complain of weakness, weakness is rarely found on examination and when present is rarely due to neuropathy. The only clinical feature that may help differentiate between DSP and ATN is a recent history of NRTI use. In those without symptoms at the start of NRTI use, ATN may start soon after initiation of NRTI use, often within weeks, but at times may be delayed by months. Even if the offending drug is discontinued, most patients experience worsening of their symptoms for a month or two as a result of a phenomenon known as “coasting.” Later, most patients will see an improvement in their symptoms, but often they never return to an asymptomatic state. This observation has raised the hypothesis that NRTI use “unmasks” DSP in susceptible patients, although a failure to return completely to normal is a feature of many toxic axonal polyneuropathies.156 Laboratory investigations in DSP and ATN patients are relatively unrevealing. General screening tests for other etiologies of neuropathy are often negative. Spinal fluid analysis may show slightly elevated protein in a small number of patients. The cell counts are usually normal, and, in fact, cerebrospinal fluid pleocytosis should raise the possibility of concurrent opportunistic infection.

Electrophysiology The electrophysiologic findings include mainly features of axonal loss,11,30,40,45,124 and document a more widespread sensorimotor polyneuropathy than detected clinically. Evidence of axonal loss is most severe in the legs, frequently with absent sural sensory responses. Although these conditions are called predominantly sensory polyneuropathies, there are motor abnormalities, as evidenced by the frequent motor conduction abnormalities. These include reduced distal evoked amplitudes and minimally reduced conduction velocities in the legs, and prolonged F-wave latencies in both arms and legs. Needle electromyography demonstrates denervation potentials and evidence of chronic partial denervation and reinnervation usually confined to leg and foot muscles.

Pathology The dominant feature in nerve biopsy and autopsy series has been the distal axonal degeneration.50,76,151,183 Distal axonal loss resulting from wallerian-like degeneration affects predominantly the unmyelinated fibers, but the myelinated axons are also involved. Demyelination and remyelination are seen but are not prominent features. In about a third of the patients, there is perivascular lymphocytic infiltration.50 The distal “dying-back axonopathy” nature of the lesions is supported by the findings of degeneration of the rostral

gracile tracts in the spinal cord of some cases.147 Recent use of skin biopsies allows repeated evaluations of the degree of pathologic severity in HIV neuropathy patients. The data from these studies confirm the “length-dependent” nature of the lesions and show that the loss of epidermal nerve fiber density is worse distally.145 In a recent study, immunoglobulin (Ig) deposition was found in autopsy specimens of sural nerve from patients with DSP.150 Contrary to most studies of DSP, the prominent pathologic process was reported as demyelination, seen in all 11 cases. IgA was the most frequent Ig seen, occurring in 7 cases. This work requires confirmation. Despite the prominent loss of axons distally, there is often minimal degeneration of neurons in the dorsal root ganglia (DRGs).76 There is prominent macrophage infiltration in DRGs, yet this macrophage infiltration is not associated with significant HIV replication. Immunostaining often fails to demonstrate any HIV viral proteins in DRGs (C. A. Pardo, personal communication, 2002). Furthermore, the HIV DNA or viral particles are rarely found in biopsied nerves or in autopsy specimens,151,183 and when found may have been in lymphocytes in blood vessels.

Pathogenesis of Distal Symmetrical Polyneuropathy The mechanisms of HIV-associated DSP are unknown. Over the years various hypotheses have been put forward, but recent data suggest that multiple mechanisms are likely to play a role in neuronal or axonal injury observed in DSP. Prior to antiretroviral therapy, patients with HIV infection often died of AIDS with multiple opportunistic infections. It was not clear if DSP was due to coinfection with other viruses or to the HIV itself. In fact, as a result of the high association of CMV coinfection in other organs, Fuller and co-workers postulated that CMV infection of the sensory ganglia was responsible for DSP.68 However, other studies did not find any increased incidence of CMV in the DRGs or the peripheral nerves.74 Since these early clinical observations, most of the experimental studies on neuropathogenesis of HIV have focused on the CNS complications. There have been very few experimental studies on the role of HIV in causing toxicity to sensory neurons. Clinical observations and autopsy findings suggest a length-dependent neuropathy with dying-back axonopathy features (reviewed by Pardo et al.139). Neuronal infection by HIV occurs rarely,187 if at all. Thus, in a search for other mechanisms, attention fell on the HIV-1 envelope glycoprotein gp120. It has been hypothesized that neurotoxicity from gp120 plays a role in the pathogenesis of HIV-associated DSP via ligation of chemokine receptors on glial cells and neurons. Two studies showed that gp120 could bind to sensory neurons7 and cause complement-mediated cytotoxicity.6 Injection of gp120 into rat sciatic nerve causes

Peripheral Neuropathies in Human Immunodeficiency Virus Infection

neuropathic pain and allodynia.80 However, the concentration of gp120 used in these studies was high and not likely to be present at those levels in the sensory ganglia. Studies since then failed to show significant levels of circulating gp120 in the DRGs.168 In a recent study, the complex interplay between the sensory neuron and the supporting glial cells within DRGs was studied.94 In this in vitro study, gp120-induced neurotoxicity did not occur directly on the sensory neuron but was mediated by the chemokine receptors on the Schwann cells. Binding of the gp120 to CXCR4 chemokine receptor on Schwann cells caused release of a CCR5 ligand, RANTES (regulated upon activation, normal T-cell expressed and secreted), which in turn bound to the CCR5 receptors on the sensory neurons and induced a tumor necrosis factor-␣ (TNF-␣)–mediated neuritic degeneration and apoptotic cell death (Fig. 94–2). Only the interaction between the sensory neuron and the surrounding Schwann cells was studied, omitting the role that the infiltrating macrophages play. Often there is an increase in the number of activated macrophages in the DRGs of patients with HIV, and these macrophages immunostain with markers of HIV such as p24.139 Because there is little evidence for direct infection of the sensory neuron or the Schwann cell by HIV, the infiltrating infected macrophages may serve as a reservoir of secreted viral proteins such as gp120. In addition to this indirect neurotoxicity, it has been shown that gp120 can directly bind to chemokine receptors on sensory neurons and induce intracellular calcium fluxes.137 The gp120-responsive subset of neurons also

FIGURE 94–2 Simplified hypothesis of the mechanism of gp120-induced DRG neurotoxicity. CXCR4 binding on Schwann cells by SDF-1␣ or gp120 results in the release of RANTES, which induces TNF-␣ production by DRG neurons, and subsequent TNF receptor 1–mediated neurotoxicity in an autocrine/paracrine fashion. (Adapted from Keswani, S. C., Polley, M., Pardo, C. A., et al.: Schwann cell chemokine receptors mediate HIV-1 gp120 toxicity to sensory neurons. Ann. Neurol. 54:287, 2003.)

2133

expressed the capsaicin receptor vanilloid receptor type 1. When the paws of rats were injected with gp120, they developed neuropathic pain behavior. Similarly, it has been shown that intrathecal injection of gp120 induces neuropathic pain.126,127 The authors concluded that binding of gp120 to glial cells induced the proinflammatory cytokines interleukin-1 (IL-1) and TNF-␣. Antagonists of both IL-1 and TNF-␣ were able to block the pain behavior. At this stage, it is difficult to determine what role direct versus indirect neurotoxicity plays in HIV-associated DSP. Further in vivo studies are needed to corroborate the in vitro studies. What is needed is a reliable animal model of HIV infection that also shows PNS neurotoxicity. In recent years, an accelerated version of simian immunodeficiency virus infection in macaques has been used successfully as a model of HIV encephalitis.113,188 The pathology of the PNS in these macaques is largely unknown. However, recent unpublished studies demonstrate large numbers of infiltrating activated macrophages in the DRGs (C. A. Pardo, personal communication, 2003). Skin biopsies in the footpads of these animals, however, failed to show a reduction in the intraepidermal nerve fiber density (J. C. McArthur, unpublished data, 2000). Therefore, it is unclear whether these animals will serve as a good model of HIV neuropathy. Another potentially useful animal model of HIV neuropathy is the development of neuropathy in newborn cats infected with the feline immunodeficiency virus (FIV).91 FIV-infected cats develop a lower density of intraepidermal nerve fibers and have lower number of myelinated fibers in their sural nerves compared to control littermates. Furthermore, there is evidence of macrophage infiltration in their peripheral nerves and DRGs.91 It is not clear whether this animal can be a model for adult-onset DSP, because these cats were infected with the FIV at a time when they were developing. This model may be more applicable to the HIV infection that occurs in children. Further studies are needed, especially with infection of adult cats, to determine if FIV-induced neuropathy can be used as a model for DSP. A promising new development in the field has been the creation of a transgenic rat that expresses human CD4 and CCR5.92 These animals can be infected with HIV and show evidence of immune deficiency. The nervous system pathology has not been examined in detail. However, if these animals develop pathology in the PNS, they will serve as a very useful small animal model to study HIV neuropathogenesis.

Pathogenesis of Antiretroviral Toxic Neuropathy Initial cases of ATN were reported in AIDS patients receiving ddC.15,56,120–122,144,186 As the newer NRTIs came to market, it was clear that some of them, namely ddC, didanosine (2⬘,3⬘-dideoxyinosine [ddI]), and stavudine

2134

Diseases of the Peripheral Nervous System

FIGURE 94–3 NRTI-induced neurotoxicity in primary DRG sensory neurons. NRTIs that cause neuropathy in HIV patients (ddC, ddI, and d4T), but not AZT, cause degeneration of axons in primary sensory cultures in a dose-dependent manner. A, Control. B, ddC at 1 ␮M. C, ddC at 10 ␮M. D, ddC at 100 ␮M. (Adapted from Keswani, S. C., Chander, B., Hasan, C., et al.: FK506 is neuroprotective in a model of antiretroviral toxic neuropathy. Ann. Neurol. 53:57, 2003.)

(didehydrodeoxythymidine [d4T]), also caused a predominantly sensory painful neuropathy.37,99,135,165 Moore and co-workers130 found the incidence of neuropathy increased when two or more neurotoxic agents were used together or when two neurotoxic agents were used with hydroxyurea. While the exact etiology of this neurotoxic neuropathy remains unclear, inhibition of mitochondrial DNA polymerase-␥ was thought to play an important role.31,90,170 Theoretically, inhibition of mitochondrial DNA polymerase-␥ could result in reduced numbers of mitochondria and subsequent mitochondrial failure, causing degeneration of the axons/neurons. This line of argument has been promoted over the years27,152,159 with little evi-

dence that inhibition of DNA polymerase-␥ is the cause of neurotoxicity. In fact, although the mitochondrial DNA content in the subcutaneous fat tissue correlates well with NRTI use, there is no correlation of this reduced mitochondrial content with incidence of neuropathy.32 Furthermore, although all of the NRTIs inhibit mitochondrial DNA polymerase-␥ in vitro, only some cause neuropathy. Of the six NRTIs in clinical use, ddC, ddI, and d4T have been associated with clinical neuropathy. There are no clinical reports of AZT, dideoxy-3⬘-thiacytidine (3TC), or abacavir causing neuropathy.19,65 Recent in vitro studies also argue against the inhibition of mitochondrial DNA polymerase-␥ as the main cause of neurotoxicity. In a study by Cui and co-workers, although ddI and ddC toxicity to PC-12 cells correlated with inhibition of mitochondrial DNA synthesis, no change in mitochondrial DNA content was observed with d4T at doses that caused toxic effects on neurites.46 Furthermore, AZT inhibited the mitochondrial DNA polymerase-␥ and caused a reduction in DNA content in PC-12 cells but did not cause degeneration of neurites. This suggests that NRTI mitochondrial toxicity may be mediated by mechanisms other than mitochondrial DNA depletion. In a recent study using primary DRG neurons, the mechanism of neurotoxicity of NRTIs was examined.93 In this study, neurotoxicity exhibited by ddC, ddI, and d4T correlated with their incidence of neuropathy in clinical experience (ddC ⬎ ddI ⬎ d4T), and AZT, which does not cause neuropathy, was not toxic to DRG neurons even at very high doses (Fig. 94–3). The neurotoxicity of these NRTIs was mediated by mitochondrial toxicity and energy failure from loss of the electrical potential differential (⌬-␺) that exists across the inner mitochondrial membrane (Fig. 94–4). The loss of ⌬-␺ was not due to inhibition of mitochondrial DNA polymerase-␥, because the effect was very rapid and there was no loss of ⌬-␺ with AZT, even at very high doses. Another study arguing against the hypothesis of inhibition of mitochondrial DNA polymerase-␥ is the toxicity of ddC on cardiac myocytes. Skuta and colleagues noted that ddC induced rapid cardiotoxicity in rats, and this was associated with decreased activity of respiratory complexes, but not with mitochondrial DNA depletion.166 Despite the studies that argue against a pivotal role for inhibition of mitochondrial DNA polymerase-␥ in mediating the NRTI neurotoxicity, recent clinicopathologic data do suggest that abnormal mitochondria are a feature of ATN. Dalakas and co-workers demonstrated structural abnormalities in mitochondria in axons and Schwann cells in the sural nerve biopsies of HIV neuropathy patients exposed to ddC as compared to HIV neuropathy patients not exposed to ddC or non–HIV-infected patients with neuropathy.49 While a larger study would be helpful in confirming this observation, detailed mechanistic studies are needed to assess the primary abnormality or abnormalities.

Peripheral Neuropathies in Human Immunodeficiency Virus Infection

2135

association between serum carnitine levels and incidence of neuropathy.161

Treatment of DSP and ATN

FIGURE 94–4 NRTI-induced neurotoxicity in primary DRG sensory neurons is mediated by mitochondrial toxicity. NRTIs that cause neuropathy in HIV patients (ddC, ddI, and d4T) cause rapid loss of the mitochondrial membrane potential differential (⌬-␺) that exists in healthy mitochondria, resulting in necrotic cell death caused by mitochondrial energy failure. (Adapted from Keswani, S. C., Chander, B., Hasan, C., et al.: FK506 is neuroprotective in a model of antiretroviral toxic neuropathy. Ann. Neurol. 53:57, 2003.)

The studies on the mechanism of NRTI-induced neurotoxicity have also suffered from lack of a reliable in vivo model. In animal studies, ddC was shown to cause a neuropathy in rabbits after 16 weeks of exposure.3,61 However, this neuropathy was characterized by prominent changes in Schwann cells in the proximal segments of the sciatic nerves and in the ventral roots, with demyelination and remyelination.4 Primary demyelination is not a characteristic pathologic feature of NRTI-induced neuropathy in patients. Therefore, it is unclear if the ddC-induced neuropathy in rabbits is a model of human ATN. Based on patient studies, two other hypotheses have been put forward to explain ATN. A recent prospective study compared the serum lactate levels in 20 patients with ATN caused by d4T, 10 patients with DSP, 20 patients on stavudine without neuropathy, and 23 patients who were not on stavudine and did not have neuropathy.23 Elevated serum lactate levels discriminated between stavudineinduced ATN and DSP. This study needs confirmation with a larger study and should include patients who develop ATN as a result of ddC or ddI use before elevated serum lactate levels can be considered as a unifying pathogenic marker for all cases of ATN. Another study raised the possibility of low carnitine levels in ATN patients.60 In this small study with 12 ATN patients on dideoxynucleoside analogues, serum acetylcarnitine levels were lower compared to those in HIV-infected patients on dideoxynucleosides without neuropathy. However, a subsequent study with a larger cohort failed to show any

Currently, treatment of DSP and ATN is similar to that of many other neuropathies with predominantly painful sensory involvement.62,83,119 There is no proven therapy directed to the underlying pathology. However, one prospective study reported an improvement in subjective quantitative sensation testing in HIV-infected patients responding to HAART.114 The patients who did not respond to HAART did not show any improvement in their quantitative sensation testing. This 8-month pilot study suffered from a small number of patients and from a failure to evaluate the patients’ symptoms and signs. It is possible that suppression of viral load will slow the progression of DSP. In fact, some studies found a correlation between incidence of sensory neuropathy and viral load,34 or severity of sensory neuropathy and viral load.160 However, others did not find any correlation between plasma viral loads and incidence of DSP or ATN.23 Further studies with carefully selected patients are needed to determine if suppression of viral load, especially early in the disease course, will reduce the incidence of DSP. The selection of most drugs used clinically for symptomatic control of neuropathic symptoms in HIV-associated DSP and ATN has been based on studies in other painful neuropathies, especially diabetic neuropathy. However, there are a few small studies conducted in HIV-infected patients. The medications studied include gabapentin,98,136 lamotrigine,162,163 amitriptyline and acupuncture,158 and topical capsaicin.138 In these studies, only gabapentin and lamotrigine provided benefits over placebo, yet in the most recent study lamotrigine was only marginally better than placebo and failed to show benefit in the primary outcome measure. Neither topical capsaicin, amitriptyline, nor standardized acupuncture was better than placebo in controlling pain in HIV-infected patients with painful sensory polyneuropathies. Thus most clinicians use gabapentin, amitriptyline, tramadol, or lamotrigine in the symptomatic treatment of painful neuropathic symptoms. One word of caution is that ritonavir-containing antiretroviral regimens may induce hepatic metabolism of lamotrigine and decrease its levels. Amitriptyline is also useful as an adjunct agent in helping with nocturnal, pain-induced insomnia. Caution is advised in patients on protease inhibitors because these agents inhibit hepatic metabolism of tricyclic antidepressants and augment toxicity. Narcotic analgesics are effective in controlling the chronic pain associated with a variety of painful neuropathic conditions, including DSP and ATN. However, because of a high incidence of side effects and abuse potential, they should be reserved for patients who do not respond to other

2136

Diseases of the Peripheral Nervous System

modalities. The principle is to use a long-acting narcotic with regular dosing and restrict short-acting analgesics to use for “breakthrough” pain. Recent reviews on the treatment of painful sensory neuropathy are available.83,141 Although clinical trials designed specifically for HIV neuropathy patients have been few to date, this is likely to change as the mechanisms underlying different types of HIV-associated neuropathies are understood. The only clinical trial aimed at regenerative potential in the peripheral nerves was with recombinant human nerve growth factor (rhNGF).115 This Phase II trial showed a benefit in self-reported pain measures as well as in pin sensitivity as measured by the examiners. However, there was no difference in the quantitative sensation testing or intraepidermal nerve fiber density in a subset of patients. The major drawback of this study was the relative short duration to test changes in morphologic outcome measures. Despite the apparent usefulness of rhNGF in relieving neuropathic pain in DSP and ATN, there are currently no plans to develop this agent for clinical use. Prosaptide is a saposin C analogue with analgesic and nerve regenerative properties in animal models of diabetic neuropathy.26 It entered a large clinical trial in HIV neuropathy patients in 2003. A recent study with an in vitro model of ATN identified immunophilin ligands as potential neuroprotective drugs.93 In this study, FK506, an immunophilin ligand with both immunosuppressive and neuroprotective/neuroregenerative potential, was used. Since the publication of this study, nonimmunosuppressive analogues have also been shown to be neuroprotective against NRTI-induced neurotoxicity (S. C. Keswani, J. Steiner, and A. Hoke, unpublished data, 2003). These types of neuroprotective strategies are very well suited for ATNs because one can even envision using these compounds proactively in HIV patients who need to be on high doses of NRTIs as part of their HAART regimen.

NEUROPATHIES ASSOCIATED WITH HIV-1 SEROCONVERSION A generalized systemic illness may accompany HIV-1 seroconversion.38,82 Additionally, a number of acute neurologic illnesses, including aseptic meningitis,81,82 encephalopathy,28 myelopathy,53 GBS,143,175 unilateral facial palsy,14,184 bilateral facial palsy,180 bibrachial palsy,25 and sensory neuropathy,52 may occur at the time of HIV-1 seroconversion. Many of these disorders are indistinguishable from those that occur in association with other viral diseases. Thus HIV-1 testing should be considered at the time of presentation in individuals with these acute neurologic illnesses and appropriate risk factors. If the serology is negative and suspicion is high, a test for viral RNA should be performed or a repeat serology test should be done 3 months later.

INFLAMMATORY DEMYELINATING POLYNEUROPATHY The IDPs have been divided historically into two types, acute (AIDP, subtype of GBS), and chronic (CIDP). AIDP and CIDP share many features. In both, the motor findings usually predominate; sensory symptoms and signs are usually minor in comparison. Electrophysiologic studies demonstrate features of demyelinating polyneuropathy with variable degrees of axonal loss. Pathologically, both disorders show similar features, primarily macrophagemediated demyelination and lymphocytic infiltrates. There is frequently superimposed axonal loss. In most instances, spinal fluid protein is elevated. Both diseases have been treated with a variety of methods aimed at immunosuppression or immunomodulation. For AIDP both plasmapheresis and intravenous immunoglobulin (IVIg)84,148 are proven therapies, whereas for CIDP plasmapheresis, IVIg, and prednisone117,118,174 are proven therapies. An extensive body of data supports the hypothesis that both disorders are immune mediated. Both GBS and CIDP have been associated with HIV-1 infection.21,41,105,124,171 Most but not all cases of IDP, either acute or chronic, occur in HIV-1–infected patients who are otherwise asymptomatic. Indeed, in certain patients GBS may be a manifestation of seroconversion78,143,175 or may be the first manifestation of illness that leads to the identification of concurrent HIV-1 infection.41 Prospective studies have shown that neither GBS nor CIDP is overrepresented in the HIV-1–infected population.

Clinical Features The clinical features of HIV-1–infected individuals with IDP are not different from those of patients without HIV-1 infection.16,20,36,41,102,103,105,124,142,143,146,171 Presentation may be either acute, with progression to maximum illness in less than 4 weeks, or chronic, with disease onset occurring over several months. In the majority of cases, the symptoms are primarily motor. The cranial nerves may be involved, and respiratory paralysis has been described in the acute cases. Several features distinguish HIV-1–seronegative from HIV-1–seropositive individuals with IDP. The first is the presence of HIV-1 infection. As mentioned, GBS may occur during seroconversion with HIV-1, so that serial testing is required in cases with a high index of suspicion. Negative enzyme-linked immunosorbent assay testing may be seen early, at a time when Western blots are positive. Second, in cases with HIV-1 infection, circulating CD4:CD8 T-cell ratios are inverted. While inverted ratios may be present in individuals with GBS, they are usually a minority.42 Third, in contrast to the usual absence of cellular response in spinal fluid in HIV-1–seronegative IDP, patients with HIV-1–seropositive IDP frequently have cerebrospinal fluid pleocytosis,41 a feature useful in suggesting the association

Peripheral Neuropathies in Human Immunodeficiency Virus Infection

of HIV-1 infection and IDP. Fourth, many patients have polyclonal elevations of serum immunoglobulins that can range up to several grams per deciliter. This may not be specific to IDP, but rather may be an indication of chronic antigenic stimulation and immunoglobulin production in HIV-1–infected individuals. Numerous reports have described the electrophysiologic features in HIV-1–seropositive individuals with all forms of IDP, including AIDP and CIDP. These do not differ from those in HIV-1–seronegative patients with IDP. In reported cases, the primary electrophysiologic finding is demyelination, indicating that these cases are the AIDP form of GBS. This is manifest as some combination of prolonged distal and F-wave latencies, reductions in conduction velocity and motor evoked amplitudes, abnormal temporal dispersion, and partial conduction block. In almost all cases studied serially, there is associated axonal degeneration.

Pathology In active disease in these ADIP cases, the ongoing demyelination is characterized by myelin stripping by macrophages. Teased fibers confirm the prominence of demyelination. The extent of axonal degeneration varies, as does the degree of lymphocytic infiltration. All of these findings are typical of AIDP in seronegative cases.

Therapy There are no clinical trials to guide definitive decisions about therapy in HIV-1–seropositive patients with IDP. However, because the clinical disease is similar if not identical to that in HIV-1–seronegative patients with IDP, studies in presumably HIV-1–seronegative populations probably apply to the HIV-1–seropositive patient.

Etiology Because of the similarities to HIV-1–seronegative IDPs, disorders presumed to be immune mediated, it is likely that IDP occurring in HIV-1–seropositive individuals occurs on an immunopathogenic basis. The differences described above may relate solely to the presence of HIV-1 infection. Viral infections are known to precede GBS, and an association has been described between GBS and herpesvirus infections, specifically CMV and Epstein-Barr virus.55 It is reasonable to postulate that the sequence of events that occurs in patients with virus infections resulting in GBS or CIDP also occurs in those with HIV. Both HIV-1–seropositive and HIV-1–seronegative patients with IDP respond to the same sorts of immunotherapies, suggesting that they share features of immunopathogenesis. The T cells present in sural nerve biopsies from patients with HIV-1–seronegative and –seropositive IDP have been

2137

typed.39 Quantitatively, similar numbers of total T cells are present in HIV-1–seropositive and HIV-1–seronegative patients with either GBS or CIDP. There is, however, a quantitative difference in the types of T cells present. CD4⫹ cells are more numerous than CD8⫹ cells in HIV1–seronegative IDP, and the reverse is true in HIV-1– seropositive IDP. Further similarities exist in the presence of antibodies in these patients. Antibodies to ganglioside or Forssman antigens occur in GBS patients, both seropositive and seronegative85 (A. Pestronk, unpublished observation, 1992). This suggests that both groups may make antibodies to similar neural antigens.

MULTIPLE MONONEUROPATHY Multiple mononeuropathy caused by vasculitis occurring in the setting of symptomatic HIV-1 infection has been reported.41,48,57,71,100,111,155 These cases have occurred primarily in individuals with what was previously termed AIDS-related complex (ARC). In this situation, there is some evidence for a direct relationship of HIV-1 infection to etiopathogenesis. Another syndrome of multiple mononeuropathies not caused by vasculitis has also been described. The initial report by Lipkin and co-workers108 described 12 patients with predominantly sensory symptoms. The neuropathy was multifocal in 10. There was evidence of CNS involvement in 6 of the 12. Like the patients with multiple mononeuropathies caused by vasculitis, most of these patients had ARC. The clinical features are less well delineated than for the other neuropathic syndromes. Most of these patients have a good prognosis, with symptoms resolving spontaneously. However, some of these patients evolved from this multifocal sensory neuropathy into typical IDP.140 Pathology has shown a combination of axon loss and demyelination associated with perivascular inflammation. Similar cases have been reported.103

Clinical Features In most of the reported cases of multiple mononeuropathy caused by vasculitis in HIV-1–infected patients, the clinical details have been only briefly mentioned because the emphasis of these reports has been either pathology or immunology. However, the available data suggest that the clinical presentations are similar to those of other series of patients with vasculitic neuropathy.57 The only apparent feature that separates these cases from other patients with vasculitic neuropathy is the HIV-1 infection. Electrophysiology reports are limited. In two cases from Johns Hopkins University, nerve conduction studies revealed a multifocal pattern of severe reductions in evoked sensory and motor action potential amplitudes in one case and evidence of nerve entrapments in the setting

2138

Diseases of the Peripheral Nervous System

of a generalized neuropathy in the other case. In the first case, these findings were highly suggestive of an angiopathic process, while in the second case the underlying basis of the neuropathy was not suspected until nerve biopsy was performed. This latter case highlights the value of nerve biopsy in cases of undiagnosed and unusual neuropathies.

Pathology Because most of the reports of this syndrome have concentrated on the pathologic features, an excellent description of the pathology is possible. Biopsies from these patients show all the features previously described in patients with vasculitic neuropathy.57,97,155 These include cellular infiltrates, perineurial infarcts, inter- and intrafascicular variability in the degree of nerve injury, and active axonal degeneration. The hallmark is necrotizing vasculitis. The death of vascular cells may be followed by the local proliferation of smooth muscle cells.

Therapy There is limited information on therapy in these cases. Therefore, firm conclusions about the role of therapy cannot be made. Intense immunotherapy may worsen a compromised immune system, leading to opportunistic infections. Thus this form of therapy should be carefully considered.

Etiology The etiology of the syndrome is unknown. By analogy to the multiple mononeuropathy syndrome that occurs in individuals with hepatitis B,47 the vasculitic multiple mononeuropathy in HIV-1–infected individuals may be related to circulating immune complexes of HIV-1 antibody and antigen that then lodge in blood vessels, setting up the sequence of events that lead to vasculitis. Said and colleagues reported HIV-1 p24 antigen in the blood vessel walls in their cases of vasculitic mononeuropathy.154 Gherardi and co-workers71 reported HIV-1 in perivascular mononuclear cells in two cases of vasculitic neuropathy. Additionally, circulating immune complexes composed of HIV-1 antigen-antibody are present in the sera of patients with HIV-1 infection.116 That these complexes may cause tissue damage has been inferred from studies of a subgroup of patients with HIV-associated idiopathic thrombocytopenic purpura (ITP).177 In this group, circulating, aggregated platelets have been found to be enmeshed in immune complexes, presumably HIV-1 related. If, instead of adhering nonspecifically to platelets, these circulating immune complexes became lodged in epi- or endoneurial blood vessels, they could begin the cascade of events leading to vasculitis. If the systemic circulation could then be

cleared of immune complexes, the process might be selflimited. This speculation awaits further confirmation.

AUTONOMIC NEUROPATHY As mentioned above, the somatic PNS disorders associated with HIV-1 infection generally occur at specific stages of HIV-1 infection. The brief reports summarized below concerning the autonomic system dysfunction occurring in HIV-1 infection also suggest a disease-stage specificity with dysautonomia occurring in the late stage of infection, AIDS. Craddock and co-workers43 initially reported that four of five patients with AIDS and pulmonary infection had syncopal reactions during fine-needle aspiration of the lung. They then studied one of these patients and four further patients with AIDS using a battery of standardized autonomic tests. Their results suggested that dysautonomia was frequent in patients with AIDS. Subsequently, a number of case reports appeared of individuals with autonomic dysfunction and AIDS.5,59,107,123,176–178 Cohen and Laudenslager35 studied 10 HIV-1–infected individuals, across the full spectrum of disease, with tests of both the sympathetic and parasympathetic nervous system. Four individuals had more than one abnormal test, and all four had AIDS. This larger study confirms the earlier reports that dysautonomia occurs in the late stage of HIV-1 infection. However, the findings in all these studies must be tempered by the brief report of Lohmoller and co-workers.109 They studied even larger groups of patients and included their own controls. All autonomic testing took into account both age and resting heart rate. With these corrections, they could find no increase in the number of abnormal tests in individuals with AIDS. They suggested that other nonspecific features such as resting heart rate, weight loss, hypovolemia, or inactivity may better explain the results of autonomic function testing in those with AIDS. They speculated that the test results indicative of dysautonomia may be nonspecific or false positive. Villa and colleagues found more abnormal cardiovascular autonomic tests in a group of HIV-seropositive subjects compared to an age-matched HIV-seronegative control group.178 While this finding is interesting, it is uncertain if subjects were symptomatic or if there were functional consequences of these abnormal test results. Longitudinal studies are needed.

NEUROPATHY IN DIFFUSE INFILTRATIVE LYMPHOMATOSIS SYNDROME Involvement of the PNS in HIV infection can also occur in the setting of a systemic illness called diffuse infiltrative lymphomatosis syndrome (DILS). DILS is characterized

Peripheral Neuropathies in Human Immunodeficiency Virus Infection

by CD8 hyperlymphomatosis with visceral organ infiltration, including peripheral nerves.64,77,86,87,89,132 In 1988, Gold and colleagues reported four cases of HIV-infected individuals presenting with peripheral neuropathy caused by diffuse lymphoma. In two individuals autopsy showed prominent infiltration of the peripheral nerves.73 In subsequent studies the infiltrating cells were found to be CD8 and did not necessarily represent systemic lymphoma.72 Moulignier and colleagues reported 12 cases of peripheral neuropathy in patients with DILS.132 All patients had a sicca syndrome and multivisceral involvement. The neuropathy was often symmetrical, of acute or subacute onset, and almost always painful. The electrophysiology suggested an axonal process in 10 patients, and biopsies revealed axonal neuropathy with diffuse CD8 cell infiltrates. The HIV genome was detectable in the nerve biopsies. Most patients responded to either antiretroviral or steroid therapy.

OPPORTUNISTIC INFECTIONS OF THE PNS Cytomegalovirus CMV is an opportunistic viral infection. In those without AIDS, the clinical manifestations include encephalitis, pneumonia, and multiorgan disseminated infection. In the AIDS era, a specific neurologic syndrome, cytomegalovirus polyradiculoneuropathy, has been clearly associated with CMV.13,17,58,66,75,88,112,125,131,173 In almost all reported cases, this syndrome occurs in those with AIDS. The clinical features are distinctive. Over a short time period, a neurologically well individual develops a cauda equina syndrome that is asymmetrical and predominantly motor. Low back pain with radiation into one leg may be the earliest symptom, followed by urinary incontinence. Shortly thereafter, asymmetrical leg weakness and saddle sensory disturbance develop. In the majority of reported cases, the syndrome rapidly advances to a flaccid paraplegia with bowel and bladder disturbance. In most cases, the disease remains at this stage for some period of time. Occasionally, the disease may ascend with arm weakness or cranial nerve involvement. The majority of affected individuals die. Spinal fluid analysis reveals a consistent picture of polymorphonuclear pleocytosis, hypoglycorrhachia, and elevated protein content. Frequently, spinal fluid viral cultures will reveal CMV, but this usually occurs weeks after the lumbar puncture. Evidence of CMV is usually found elsewhere, including in retina, blood, and urine, more promptly. Electrodiagnostic studies reveal primarily evidence of axon loss in lumbosacral roots with fibrillations and positive sharp waves in leg muscles. There is little or no evidence of demyelination.

2139

In AIDS patients, polyradiculopathy may have causes other than CMV.169 Lymphomatous meningitis or syphilis101 may also present with a syndrome similar to that described above. Spinal fluid studies, with special attention to syphilis serology, cytopathology, and cultures, are most helpful in distinguishing these syndromes. Autopsy studies in CMV polyradiculoneuropathy reveal a uniform picture of acute and chronic inflammation in anterior and posterior spinal roots and occasionally in the spinal cord itself. CMV has been described in neural tissue, based on both the presence of cytomegalic inclusions and immunocytochemical studies. Sural nerve biopsies in this syndrome have been relatively unrevealing, with only minimal degrees of inflammation. Recognition of this clinically distinct entity has led to therapeutic intervention in several reported cases. Ganciclovir, an antiviral agent effective against CMV, has been used,66,96,125 and treatment resulted in either stabilization or improvement. However, ganciclovir-resistant cases have been reported.44,167,172 In such cases, foscarnet can be an alternate agent.2,51 CMV has also been reported as causing a multifocal neuropathy, predominantly affecting the arms.67 In two cases, electrodiagnostic studies suggested that the primary pathologic process was demyelination. Sural nerve biopsies revealed inflammation, positive CMV immunostaining, cytomegalic cells, and prominent demyelination. CMV was cultured from other sites, and both patients were treated with ganciclovir, with stabilization of the neuropathy.

Herpes Zoster Herpes zoster infections of peripheral nerve and spinal cord are well known complications of HIV-1 infection. Herpes infections may involve any of the cranial or spinal peripheral nerves and in occasional severely immunocompromised patients may also involve the brain or spinal cord. Clinically, the disorders are indistinguishable from those occurring in HIV-seronegative individuals, with the possible exception that in individuals with AIDS the disease may be more aggressive. Herpes zoster infection is uncommon in young healthy individuals, and the presence of a herpes infection in an otherwise asymptomatic young individual with appropriate risk factors should suggest that testing for HIV-1 infection be performed.

Mycobacterium avium-intracellulare Mycobacterium avium-intracellulare (MAI), an atypical mycobacterium, was relatively uncommon until the AIDS epidemic. Subsequently, disseminated MAI infection was seen frequently in those dying with AIDS. Diagnosis can be made ante mortem by appropriate culture. There are reports of MAI at autopsy in both nerve and muscle,185 but the clinical consequence is unknown.

2140

Diseases of the Peripheral Nervous System

Cryptococcus neoformans Cryptococcus neoformans is a fungus that, like MAI, was an uncommon pathogen until the AIDS epidemic. Subsequently, cases of C. neoformans fungemia and meningitis have become more frequent. Infiltration of the spinal roots and DRGs can occur. The organisms distorted nerve fibers, but produced strikingly little destruction or inflammatory reaction. The clinical import of this finding is unknown.

ACKNOWLEDGMENT The work of A. H. was supported by grants from the National Institutes of Health (NS43991, NS46262, and AI56297).

REFERENCES 1. Alliegro, M. B., Petrucci, A., Arpino, C., et al.: Peripheral neuropathies among patients with HIV infection. J. Neurol. Neurosurg. Psychiatry 64:414, 1998. 2. Anders, H. J., Weiss, N., Bogner, J. R., et al.: Ganciclovir and foscarnet efficacy in AIDS-related CMV polyradiculopathy. J. Infect. 36:29, 1998. 3. Anderson, T. D., Davidovich, A., Arceo, R., et al.: Peripheral neuropathy induced by 2⬘,3⬘-dideoxycytidine: a rabbit model of 2⬘,3⬘-dideoxycytidine neurotoxicity. Lab. Invest. 66:63, 1992. 4. Anderson, T. D., Davidovich, A., Feldman, D., et al.: Mitochondrial schwannopathy and peripheral myelinopathy in a rabbit model of dideoxycytidine neurotoxicity. Lab. Invest. 70:724, 1994. 5. Anonymous: Testing for neurological involvement in HIV infection. Lancet 2:1531, 1987. 6. Apostolski, S., McAlarney, T., Hays, A. P., et al.: Complement dependent cytotoxicity of sensory ganglion neurons mediated by the gp120 glycoprotein of HIV-1. Immunol. Invest. 23:47, 1994. 7. Apostolski, S., McAlarney, T., Quattrini, A., et al.: The gp120 glycoprotein of human immunodeficiency virus type 1 binds to sensory ganglion neurons. Ann. Neurol. 34:855, 1993. 8. Authier, F. J., and Gherardi, R. K.: Peripheral neuropathies in HIV-infected patients in the era of HAART. Brain Pathol. 13:223, 2003. 9. Autran, B., Carcelain, G., Li, T. S., et al.: Positive effects of combined antiretroviral therapy on CD4⫹ T cell homeostasis and function in advanced HIV disease. Science 277:112, 1997. 10. Bacellar, H., Munoz, A., Miller, E. N., et al.: Temporal trends in the incidence of HIV-1-related neurologic diseases: Multicenter AIDS Cohort Study, 1985–1992. Neurology 44:1892, 1994. 11. Bailey, R. O., Baltch, A. L., Venkatesh, R., et al.: Sensory motor neuropathy associated with AIDS. Neurology 38:886, 1988.

12. Barohn, R. J., Gronseth, G. S., LeForce, B. R., et al.: Peripheral nervous system involvement in a large cohort of human immunodeficiency virus-infected individuals. Arch. Neurol. 50:167, 1993. 13. Behar, R., Wiley, C., and McCutchan, J. A.: Cytomegalovirus polyradiculoneuropathy in acquired immune deficiency syndrome. Neurology 37:557, 1987. 14. Belec, L., Gherardi, R., Georges, A. J., et al.: Peripheral facial paralysis and HIV infection: report of four African cases and review of the literature. J. Neurol. 236:411, 1989. 15. Berger, A. R., Arezzo, J. C., Schaumburg, H. H., et al.: 2⬘,3⬘-Dideoxycytidine (ddC) toxic neuropathy: a study of 52 patients. Neurology 43:358, 1993. 16. Berger, J. R., Difini, J. A., Swerdloff, M. A., et al.: HIV seropositivity in Guillain-Barre syndrome. Ann. Neurol. 22:393, 1987. 17. Bishopric, G., Bruner, J., and Butler, J.: Guillain-Barre syndrome with cytomegalovirus infection of peripheral nerves. Arch. Pathol. Lab. Med. 109:1106, 1985. 18. Blum, A. S., Dal Pan, G. J., Feinberg, J., et al.: Low-dose zalcitabine-related toxic neuropathy: frequency, natural history, and risk factors. Neurology 46:999, 1996. 19. Bozzette, S. A., and Richman, D. D.: Salvage therapy for zidovudine-intolerant HIV-infected patients with alternating and intermittent regimens of zidovudine and dideoxycytidine. Am. J. Med. 88:24S, 1990. 20. Brannagan, T. H. 3rd, Nuovo, G. J., Hays, A. P., et al.: Human immunodeficiency virus infection of dorsal root ganglion neurons detected by polymerase chain reaction in situ hybridization. Ann. Neurol. 42:368, 1997. 21. Brannagan, T. H. 3rd, and Zhou, Y.: HIV-associated Guillain-Barre syndrome. J. Neurol. Sci. 208:39, 2003. 22. Breen, R. A., Lipman, M. C., and Johnson, M. A.: Increased incidence of peripheral neuropathy with co-administration of stavudine and isoniazid in HIV-infected individuals. AIDS 14:615, 2000. 23. Brew, B. J., Tisch, S., and Law, M.: Lactate concentrations distinguish between nucleoside neuropathy and HIV neuropathy. AIDS 17:1094, 2003. 24. Brodt, H. R., Kamps, B. S., Gute, P., et al.: Changing incidence of AIDS-defining illnesses in the era of antiretroviral combination therapy. AIDS 11:1731, 1997. 25. Calabrese, L. H., Proffitt, M. R., Levin, K. H., et al.: Acute infection with the human immunodeficiency virus (HIV) associated with acute brachial neuritis and exanthematous rash. Ann. Intern. Med. 107:849, 1987. 26. Calcutt, N. A., Freshwater, J. D., and O’Brien, J. S.: Protection of sensory function and antihyperalgesic properties of a prosaposin-derived peptide in diabetic rats. Anesthesiology 93:1271, 2000. 27. Carey, P.: Peripheral neuropathy: zalcitabine reassessed. Int. J. STD AIDS 11:417, 2000. 28. Carne, C. A., Tedder, R. S., Smith, A., et al.: Acute encephalopathy coincident with seroconversion for antiHTLV-III. Lancet 2:1206, 1985. 29. Chavanet, P., Solary, E., Giroud, M., et al.: Infraclinical neuropathies related to immunodeficiency virus infection associated with higher T-helper cell count. J. Acquir. Immune Defic. Syndr. 2:564, 1989.

Peripheral Neuropathies in Human Immunodeficiency Virus Infection 30. Chavanet, P. Y., Giroud, M., Lancon, J. P., et al.: Altered peripheral nerve conduction in HIV-patients. Cancer Detect. Prev. 12:249, 1988. 31. Chen, C. H., Vazquez-Padua, M., and Cheng, Y. C.: Effect of anti-human immunodeficiency virus nucleoside analogs on mitochondrial DNA and its implication for delayed toxicity. Mol. Pharmacol. 39:625, 1991. 32. Cherry, C. L., Gahan, M. E., McArthur, J. C., et al.: Exposure to dideoxynucleosides is reflected in lowered mitochondrial DNA in subcutaneous fat. J. Acquir. Immune Defic. Syndr. 30:271, 2002. 33. Cherry, C. L., McArthur, J. C., Hoy, J. F., et al.: Nucleoside analogues and neuropathy in the era of HAART. J. Clin. Virol. 26:195, 2003. 34. Childs, E. A., Lyles, R. H., Selnes, O. A., et al.: Plasma viral load and CD4 lymphocytes predict HIV-associated dementia and sensory neuropathy. Neurology 52:607, 1999. 35. Cohen, J. A., and Laudenslager, M.: Autonomic nervous system involvement in patients with human immunodeficiency virus infection. Neurology 39:1111, 1989. 36. Conlon, C. P.: HIV infection presenting as Guillain-Barre syndrome in Lusaka, Zambia. Trans. R. Soc. Trop. Med. Hyg. 83:109, 1989. 37. Cooley, T. P., Kunches, L. M., Saunders, C. A., et al.: Once-daily administration of 2⬘,3⬘-dideoxyinosine (ddI) in patients with the acquired immunodeficiency syndrome or AIDS-related complex: results of a Phase I trial. N. Engl. J. Med. 322:1340, 1990. 38. Cooper, D. A., Gold, J., Maclean, P., et al.: Acute AIDS retrovirus infection: definition of a clinical illness associated with seroconversion. Lancet 1:537, 1985. 39. Cornblath, D. R., Griffin, D. E., Welch, D., et al.: Quantitative analysis of endoneurial T-cells in human sural nerve biopsies. J. Neuroimmunol. 26:113, 1990. 40. Cornblath, D. R., and McArthur, J. C.: Predominantly sensory neuropathy in patients with AIDS and AIDS-related complex. Neurology 38:794, 1988. 41. Cornblath, D. R., McArthur, J. C., Kennedy, P. G., et al.: Inflammatory demyelinating peripheral neuropathies associated with human T-cell lymphotropic virus type III infection. Ann. Neurol. 21:32, 1987. 42. Cornblath, D. R., Mellits, E. D., Griffin, J. W., et al.: Motor conduction studies in Guillain-Barre syndrome: description and prognostic value. Ann. Neurol. 23:354, 1988. 43. Craddock, C., Pasvol, G., Bull, R., et al.: Cardiorespiratory arrest and autonomic neuropathy in AIDS. Lancet 2:16, 1987. 44. Crane, L. R., and Ebright, J. R.: Cytomegalovirus polyradiculopathy caused by a ganciclovir-resistant strain. Clin. Infect. Dis. 19:365, 1994. 45. Cruz Martinez, A., and Villoslada, C.: Electrophysiologic study in peripheral neuropathy associated with HIV infection. Electromyogr. Clin. Neurophysiol. 31:407, 1991. 46. Cui, L., Locatelli, L., Xie, M. Y., et al.: Effect of nucleoside analogs on neurite regeneration and mitochondrial DNA synthesis in PC-12 cells. J. Pharmacol. Exp. Ther. 280:1228, 1997. 47. Cupps, T. R., and Fauci, A. S.: The Vasculitides. Philadelphia, W. B. Saunders, 1981. 48. Dalakas, M. C., and Pezeshkpour, G. H.: Neuromuscular diseases associated with human immunodeficiency virus infection. Ann. Neurol. 23(Suppl.):S38, 1988.

2141

49. Dalakas, M. C., Semino-Mora, C., and Leon-Monzon, M.: Mitochondrial alterations with mitochondrial DNA depletion in the nerves of AIDS patients with peripheral neuropathy induced by 2⬘,3⬘-dideoxycytidine (ddC). Lab. Invest. 81:1537, 2001. 50. de la Monte, S. M., Gabuzda, D. H., Ho, D. D., et al.: Peripheral neuropathy in the acquired immunodeficiency syndrome. Ann. Neurol. 23:485, 1988. 51. Decker, C. F., Tarver, J. H. 3rd, Murray, D. F., et al.: Prolonged concurrent use of ganciclovir and foscarnet in the treatment of polyradiculopathy due to cytomegalovirus in a patient with AIDS. Clin. Infect. Dis. 19:548, 1994. 52. Denning, D. W.: The neurological features of acute HIV infection. Biomed. Pharmacother. 42:11, 1988. 53. Denning, D. W., Anderson, J., Rudge, P., et al.: Acute myelopathy associated with primary infection with human immunodeficiency virus. Br. Med. J. (Clin. Res. Ed.) 294:143, 1987. 54. Dore, G. J., Correll, P. K., Li, Y., et al.: Changes to AIDS dementia complex in the era of highly active antiretroviral therapy. AIDS 13:1249, 1999. 55. Dowling, P. C., and Cook, S. D.: Role of infection in GuillainBarre syndrome: laboratory confirmation of herpesviruses in 41 cases. Ann. Neurol. 9(Suppl.):44, 1981. 56. Dubinsky, R. M., Yarchoan, R., Dalakas, M., et al.: Reversible axonal neuropathy from the treatment of AIDS and related disorders with 2⬘,3⬘-dideoxycytidine (ddC). Muscle Nerve 12:856, 1989. 57. Dyck, P. J., Benstead, T. J., Conn, D. L., et al.: Nonsystemic vasculitic neuropathy. Brain 110(Pt. 4):843, 1987. 58. Eidelberg, D., Sotrel, A., Vogel, H., et al.: Progressive polyradiculopathy in acquired immune deficiency syndrome. Neurology 36:912, 1986. 59. Evenhouse, M., Haas, E., Snell, E., et al.: Hypotension in infection with the human immunodeficiency virus. Ann. Intern. Med. 107:598, 1987. 60. Famularo, G., Moretti, S., Marcellini, S., et al.: Acetylcarnitine deficiency in AIDS patients with neurotoxicity on treatment with antiretroviral nucleoside analogues. AIDS 11:185, 1997. 61. Feldman, D., Brosnan, C., and Anderson, T. D.: Ultrastructure of peripheral neuropathy induced in rabbits by 2⬘,3⬘-dideoxycytidine. Lab. Invest. 66:75, 1992. 62. Fialho, D., and Cornblath, D. R.: Treatment for human immunodeficiency virus-related distal symmetric polyneuropathy (Protocol for a Cochrane Review). Cochrane Database Syst. Rev. (in press). 63. Fichtenbaum, C. J., Clifford, D. B., and Powderly, W. G.: Risk factors for dideoxynucleoside-induced toxic neuropathy in patients with the human immunodeficiency virus infection. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 10:169, 1995. 64. Franco-Paredes, C., Rebolledo, P., Folch, E., et al.: Diagnosis of diffuse CD8⫹ lymphocytosis syndrome in HIV-infected patients. AIDS Read. 12:408, 2002. 65. French, M., Amin, J., Roth, N., et al.: Randomized, openlabel, comparative trial to evaluate the efficacy and safety of three antiretroviral drug combinations including two nucleoside analogues and nevirapine for previously untreated HIV-1 infection: the OzCombo 2 study. HIV Clin. Trials 3:177, 2002.

2142

Diseases of the Peripheral Nervous System

66. Fuller, G. N., Gill, S. K., Guiloff, R. J., et al.: Ganciclovir for lumbosacral polyradiculopathy in AIDS. Lancet 335:48, 1990. 67. Fuller, G. N., Greco, C. M., and Miller, R. G.: Cytomegalovirus and mononeuropathy multiplex in AIDS. Neurology 40(Suppl.1):301, 1990. 68. Fuller, G. N., Jacobs, J. M., and Guiloff, R. J.: Association of painful peripheral neuropathy in AIDS with cytomegalovirus infection. Lancet 2:937, 1989. 69. Fuller, G. N., Jacobs, J. M., and Guiloff, R. J.: Nature and incidence of peripheral nerve syndromes in HIV infection. J. Neurol. Neurosurg. Psychiatry 56:372, 1993. 70. Gastaut, J. L., Gastaut, J. A., Pellissier, J. F., et al.: Peripheral neuropathies in human immunodeficiency virus infection: a prospective study of 56 patients [in French]. Rev. Neurol. (Paris) 145:451, 1989. 71. Gherardi, R., Lebargy, F., Gaulard, P., et al.: Necrotizing vasculitis and HIV replication in peripheral nerves. N. Engl. J. Med. 321:685, 1989. 72. Gherardi, R. K., Chretien, F., Delfau-Larue, M. H., et al.: Neuropathy in diffuse infiltrative lymphocytosis syndrome: an HIV neuropathy, not a lymphoma. Neurology 50:1041, 1998. 73. Gold, J. E., Jimenez, E., and Zalusky, R.: Human immunodeficiency virus-related lymphoreticular malignancies and peripheral neruologic disease: a report of four cases. Cancer 61:2318, 1988. 74. Grafe, M. R., and Wiley, C. A.: Spinal cord and peripheral nerve pathology in AIDS: the roles of cytomegalovirus and human immunodeficiency virus. Ann. Neurol. 25:561, 1989. 75. Graveleau, P., Perol, R., and Chapman, A.: Regression of cauda equina syndrome in AIDS patient being treated with ganciclovir. Lancet 2:511, 1989. 76. Griffin, D. W., and McArthur, J. C.: Peripheral neuropathies associated with HIV infection. In Gendelman, H. E., Epstein, L., and Swindels, S. (eds.): The Neurology of AIDS. New York, Chapman & Hall, p. 275, 1998. 77. Guillon, J. M., Fouret, P., Mayaud, C., et al.: Extensive T8-positive lymphocytic visceral infiltration in a homosexual man. Am. J. Med. 82(3 Spec. No.):655, 1987. 78. Hagberg, L., Malmvall, B. E., Svennerholm, L., et al.: Guillain-Barre syndrome as an early manifestation of HIV central nervous system infection. Scand. J. Infect. Dis. 18:591, 1986. 79. Hall, C. D., Snyder, C. R., Messenheimer, J. A., et al.: Peripheral neuropathy in a cohort of human immunodeficiency virus-infected patients: incidence and relationship to other nervous system dysfunction. Arch. Neurol. 48:1273, 1991. 80. Herzberg, U., and Sagen, J.: Peripheral nerve exposure to HIV viral envelope protein gp120 induces neuropathic pain and spinal gliosis. J. Neuroimmunol. 116:29, 2001. 81. Ho, D. D., Rota, T. R., Schooley, R. T., et al.: Isolation of HTLV-III from cerebrospinal fluid and neural tissues of patients with neurologic syndromes related to the acquired immunodeficiency syndrome. N. Engl. J. Med. 313:1493, 1985. 82. Ho, D. D., Sarngadharan, M. G., Resnick, L., et al.: Primary human T-lymphotropic virus type III infection. Ann. Intern. Med. 103(6 Pt. 1):880, 1985.

83. Hoke, A., Keswani, S. C., and McArthur, J.: Current theraphy for HIV sensory neuropathy. Curr. Treatment Options Infect. Dis. 5:467, 2003. 84. Hughes, R. A., Raphael, J. C., Swan, A. V., et al.: Intravenous immunoglobulin for Guillain-Barre syndrome. Cochrane Database Syst. Rev. 3:CD002063, 2001. 85. Ilyas, A. A., Willison, H. J., Quarles, R. H., et al.: Serum antibodies to gangliosides in Guillain-Barre syndrome. Ann. Neurol. 23:440, 1988. 86. Itescu, S., Brancato, L. J., and Winchester, R.: A sicca syndrome in HIV infection: association with HLA-DR5 and CD8 lymphocytosis. Lancet 2:466, 1989. 87. Itescu, S., and Winchester, R.: Diffuse infiltrative lymphocytosis syndrome: a disorder occurring in human immunodeficiency virus-1 infection that may present as a sicca syndrome. Rheum. Dis. Clin. North Am. 18:683, 1992. 88. Jacobson, M. A., Mills, J., Rush, J., et al.: Failure of antiviral therapy for acquired immunodeficiency syndromerelated cytomegalovirus myelitis. Arch. Neurol. 45:1090, 1988. 89. Kazi, S., Cohen, P. R., Williams, F., et al.: The diffuse infiltrative lymphocytosis syndrome: clinical and immunogenetic features in 35 patients. AIDS 10:385, 1996. 90. Keilbaugh, S. A., Prusoff, W. H., and Simpson, M. V.: The PC12 cell as a model for studies of the mechanism of induction of peripheral neuropathy by anti-HIV-1 dideoxynucleoside analogs. Biochem. Pharmacol. 42:R5, 1991. 91. Kennedy, J. M., Hoke, A., Johnston, J. B., et al.: Peripheral neuropathy in feline immunodeficiency virus infection: evidence of inflammation and axonal injury. AIDS 2004 (in press). 92. Keppler, O. T., Welte, F. J., Ngo, T. A., et al.: Progress toward a human CD4/CCR5 transgenic rat model for de novo infection by human immunodeficiency virus type 1. J. Exp. Med. 195:719, 2002. 93. Keswani, S. C., Chander, B., Hasan, C., et al.: FK506 is neuroprotective in a model of antiretroviral toxic neuropathy. Ann. Neurol. 53:57, 2003. 94. Keswani, S. C., Pardo, C. A., Cherry, C. L., et al.: HIV-associated sensory neuropathies. AIDS 16:2105, 2002. 95. Keswani, S. C., Polley, M., Pardo, C. A., et al.: Schwann cell chemokine receptors mediate HIV-1 gp120 toxicity to sensory neurons. Ann. Neurol. 54:287, 2003. 96. Kim, Y. S., and Hollander, H.: Polyradiculopathy due to cytomegalovirus: report of two cases in which improvement occurred after prolonged therapy and review of the literature. Clin. Infect. Dis. 17:32, 1993. 97. Kissel, J. T., Riethman, J. L., Omerza, J., et al.: Peripheral nerve vasculitis: immune characterization of the vascular lesions. Ann. Neurol. 25:291, 1989. 98. La Spina, I., Porazzi, D., Maggiolo, F., et al.: Gabapentin in painful HIV-related neuropathy: a report of 19 patients, preliminary observations. Eur. J. Neurol. 8:71, 2001. 99. Lambert, J. S., Seidlin, M., Reichman, R. C., et al.: 2⬘,3⬘Dideoxyinosine (ddI) in patients with the acquired immunodeficiency syndrome or AIDS-related complex: a Phase I trial. N. Engl. J. Med. 322:1333, 1990. 100. Lange, D. J., Britton, C. B., Younger, D. S., et al.: The neuromuscular manifestations of human immunodeficiency virus infections. Arch. Neurol. 45:1084, 1988.

Peripheral Neuropathies in Human Immunodeficiency Virus Infection 101. Lanska, M. J., Lanska, D. J., and Schmidley, J. W.: Syphilitic polyradiculopathy in an HIV-positive man. Neurology 38:1297, 1988. 102. Leger, J. M., Bolgert, F., Bouche, P., et al.: The peripheral nervous system and HIV infection: 13 cases [in French]. Rev. Neurol. (Paris) 144:789, 1988. 103. Leger, J. M., Bouche, P., Bolgert, F., et al.: The spectrum of polyneuropathies in patients infected with HIV. J. Neurol. Neurosurg. Psychiatry 52:1369, 1989. 104. LeLacheur, S. F., and Simon, G. L.: Exacerbation of dideoxycytidine-induced neuropathy with dideoxyinosine. J. Acquir. Immune Defic. Syndr. 4:538, 1991. 105. Leport, C., Chaunu, M. P., Sicre, J., et al.: Peripheral neuropathy in relation to LAV/HTLV III retrovirus infection. A clinical, anatomical and immunological study: 5 cases [in French]. Presse Med. 16:55, 1987. 106. Levy, R. M., Bredesen, D. E., and Rosenblum, M. L.: Neurological manifestations of the acquired immunodeficiency syndrome (AIDS): experience at UCSF and review of the literature. J. Neurosurg. 62:475, 1985. 107. Lin-Greenberg, A., and Taneja-Uppal, N.: Dysautonomia and infection with the human immunodeficiency virus. Ann. Intern. Med. 106:167, 1987. 108. Lipkin, W. I., Parry, G., Kiprov, D., et al.: Inflammatory neuropathy in homosexual men with lymphadenopathy. Neurology 35:1479, 1985. 109. Lohmoller, G., Matuschke, A., and Goebel, F. D.: Falsepositive test of autonomic neuropathy in HIV infection and AIDS? Case control study of heart rate variability in 62 HIV positive patients [in German]. Med. Klin. 84:242, 1989. 110. Luciano, C. A., Pardo, C. A., and McArthur, J. C.: Recent developments in the HIV neuropathies. Curr. Opin. Neurol. 16:403, 2003. 111. Mahadevan, A., Gayathri, N., Taly, A. B., et al.: Vasculitic neuropathy in HIV infection: a clinicopathological study. Neurol. India 49:277, 2001. 112. Mahieux, F., Gray, F., Fenelon, G., et al.: Acute myeloradiculitis due to cytomegalovirus as the initial manifestation of AIDS. J. Neurol. Neurosurg. Psychiatry 52:270, 1989. 113. Mankowski, J. L., Clements, J. E., and Zink, M. C.: Searching for clues: tracking the pathogenesis of human immunodeficiency virus central nervous system disease by use of an accelerated, consistent simian immunodeficiency virus macaque model. J. Infect. Dis. 186(Suppl. 2):S199, 2002. 114. Martin, C., Solders, G., Sonnerborg, A., et al.: Antiretroviral therapy may improve sensory function in HIV-infected patients: a pilot study. Neurology 54:2120, 2000. 115. McArthur, J. C., Yiannoutsos, C., Simpson, D. M., et al.: A Phase II trial of nerve growth factor for sensory neuropathy associated with HIV infection. AIDS Clinical Trials Group Team 291. Neurology 54:1080, 2000. 116. McDougal, J. S., Hubbard, M., Nicholson, J. K., et al.: Immune complexes in the acquired immunodeficiency syndrome (AIDS): relationship to disease manifestation, risk group, and immunologic defect. J. Clin. Immunol. 5:130, 1985. 117. Mehndiratta, M. M., and Hughes, R. A.: Corticosteroids for chronic inflammatory demyelinating polyradiculoneuropathy. Cochrane Database Syst. Rev. 3:CD002062, 2002.

2143

118. Mehndiratta, M. M., Hughes, R. A., and Agarwal, P.: Plasma exchange for chronic inflammatory demyelinating polyradiculoneuropathy. Cochrane Database Syst. Rev. (in press). 119. Mendell, J. R., and Sahenk, Z.: Clinical practice: painful sensory neuropathy. N. Engl. J. Med. 348:1243, 2003. 120. Meng, T. C., Fischl, M. A., and Richman, D. D.: AIDS Clinical Trials Group: Phase I/II study of combination 2⬘,3⬘-dideoxycytidine and zidovudine in patients with acquired immunodeficiency syndrome (AIDS) and advanced AIDS-related complex. Am. J. Med. 88:27S, 1990. 121. Merigan, T. C., and Skowron, G.: Safety and tolerance of dideoxycytidine as a single agent: results of early-phase studies in patients with acquired immunodeficiency syndrome (AIDS) or advanced AIDS-related complex. Study Group of the AIDS Clinical Trials Group of the National Institute of Allergy and Infectious Diseases. Am. J. Med. 88:11S, 1990. 122. Merigan, T. C., Skowron, G., Bozzette, S. A., et al.: Circulating p24 antigen levels and responses to dideoxycytidine in human immunodeficiency virus (HIV) infections: a Phase I and II study. Ann. Intern. Med. 110:189, 1989. 123. Miller, R. F., and Semple, S. J.: Autonomic neuropathy in AIDS. Lancet 2:343, 1987. 124. Miller, R. G., Parry, G. J., Pfaeffl, W., et al.: The spectrum of peripheral neuropathy associated with ARC and AIDS. Muscle Nerve 11:857, 1988. 125. Miller, R. G., Storey, J. R., and Greco, C. M.: Ganciclovir in the treatment of progressive AIDS-related polyradiculopathy. Neurology 40:569, 1990. 126. Milligan, E. D., Mehmert, K. K., Hinde, J. L., et al.: Thermal hyperalgesia and mechanical allodynia produced by intrathecal administration of the human immunodeficiency virus-1 (HIV-1) envelope glycoprotein, gp120. Brain Res. 861:105, 2000. 127. Milligan, E. D., O’Connor, K. A., Nguyen, K. T., et al.: Intrathecal HIV-1 envelope glycoprotein gp120 induces enhanced pain states mediated by spinal cord proinflammatory cytokines. J. Neurosci. 21:2808, 2001. 128. Mocroft, A., Katlama, C., Johnson, A. M., et al.: AIDS across Europe, 1994–98: the EuroSIDA study. Lancet 356:291, 2000. 129. Moore, R. D., and Chaisson, R. E.: Natural history of HIV infection in the era of combination antiretroviral therapy. AIDS 13:1933, 1999. 130. Moore, R. D., Wong, W. M., Keruly, J. C., et al.: Incidence of neuropathy in HIV-infected patients on monotherapy versus those on combination therapy with didanosine, stavudine and hydroxyurea. AIDS 14:273, 2000. 131. Moskowitz, L. B., Gregorios, J. B., Hensley, G. T., et al.: Cytomegalovirus: induced demyelination associated with acquired immune deficiency syndrome. Arch. Pathol. Lab. Med. 108:873, 1984. 132. Moulignier, A., Authier, F. J., Baudrimont, M., et al.: Peripheral neuropathy in human immunodeficiency virusinfected patients with the diffuse infiltrative lymphocytosis syndrome. Ann. Neurol. 41:438, 1997. 133. Mouton, Y., Alfandari, S., Valette, M., et al.: Impact of protease inhibitors on AIDS-defining events and hospitalizations in 10 French AIDS reference centres. Federation National des Centres de Lutte contre le SIDA. AIDS 11:F101, 1997.

2144

Diseases of the Peripheral Nervous System

134. Moyle, G.: Clinical manifestations and management of antiretroviral nucleoside analog-related mitochondrial toxicity. Clin. Ther. 22:911; discussion 898, 2000. 135. Murray, H. W., Squires, K. E., Weiss, W., et al.: Stavudine in patients with AIDS and AIDS-related complex. AIDS Clinical Trials Group 089. J. Infect. Dis. 171(Suppl. 2):S123, 1995. 136. Newshan, G.: HIV neuropathy treated with gabapentin. AIDS 12:219, 1998. 137. Oh, S. B., Tran, P. B., Gillard, S. E., et al.: Chemokines and glycoprotein120 produce pain hypersensitivity by directly exciting primary nociceptive neurons. J. Neurosci. 21:5027, 2001. 138. Paice, J. A., Ferrans, C. E., Lashley, F. R., et al.: Topical capsaicin in the management of HIV-associated peripheral neuropathy. J. Pain Symptom Manage. 19:45, 2000. 139. Pardo, C. A., McArthur, J. C., and Griffin, J. W.: HIV neuropathy: insights in the pathology of HIV peripheral nerve disease. J. Peripher. Nerv. Syst. 6:21, 2001. 140. Parry, G. J.: Peripheral neuropathies associated with human immunodeficiency virus infection. Ann. Neurol. 23(Suppl.):S49, 1988. 141. Periquet, M. I., Novak, V., Collins, M. P., et al.: Painful sensory neuropathy: prospective evaluation using skin biopsy. Neurology 53:1641, 1999. 142. Persuy, P., Arnott, G., Fortier, B., et al.: Guillain-Barre syndrome with favorable outcome in a case of recent human immunodeficiency virus infection [in French]. Rev. Neurol. (Paris) 144:32, 1988. 143. Piette, A. M., Tusseau, F., Vignon, D., et al.: Acute neuropathy coincident with seroconversion for anti-LAV/HTLV-III. Lancet 1:852, 1986. 144. Pizzo, P. A.: Treatment of human immunodeficiency virusinfected infants and young children with dideoxynucleosides. Am. J. Med. 88:16S, 1990. 145. Polydefkis, M., Yiannoutsos, C. T., Cohen, B. A., et al.: Reduced intraepidermal nerve fiber density in HIV-associated sensory neuropathy. Neurology 58:115, 2002. 146. Przedborski, S., Liesnard, C., Voordecker, P., et al.: Inflammatory demyelinating polyradiculoneuropathy associated with human immunodeficiency virus infection. J. Neurol. 235:359, 1988. 147. Rance, N. E., McArthur, J. C., Cornblath, D. R., et al.: Gracile tract degeneration in patients with sensory neuropathy and AIDS. Neurology 38:265, 1988. 148. Raphael, J. C., Chevret, S., Hughes, R. A., et al.: Plasma exchange for Guillain-Barre syndrome. Cochrane Database Syst. Rev. 3:CD001798, 2002. 149. Rathbun, R. C., and Martin, E. S. 3rd.: Didanosine therapy in patients intolerant of or failing zidovudine therapy. Ann. Pharmacother. 26:1347, 1992. 150. Rickert, C. H., Evers, S., Heese, C., et al.: Immunoglobulin deposition in sural nerves of AIDS patients with distalsymmetric HIV-associated polyneuropathy. Eur. J. Med. Res. 7:472, 2002. 151. Rizzuto, N., Cavallaro, T., Monaco, S., et al.: Role of HIV in the pathogenesis of distal symmetrical peripheral neuropathy. Acta Neuropathol. (Berl.) 90:244, 1995. 152. Rossero, R., Asmuth, D. M., Grady, J. J., et al.: Hydroxyurea in combination with didanosine and stavudine in antiretroviral-

153.

154.

155.

156.

157.

158.

159.

160.

161. 162.

163.

164.

165.

166.

167.

168.

169.

experienced HIV-infected subjects with a review of the literature. Int. J. STD AIDS 14:350, 2003. Sacktor, N.: The epidemiology of human immunodeficiency virus-associated neurological disease in the era of highly active antiretroviral therapy. J. Neurovirol. 8(Suppl. 2):115, 2002. Said, G., Lacroix, C., and Andrieu, J. N.: Necrotizing arteritis in patients with inflammatory neuropathy and human immunodeficiency virus (HIV-III) infection. Neurology 37(Suppl. 1):176, 1987. Said, G., Lacroix-Ciaudo, C., Fujimura, H., et al.: The peripheral neuropathy of necrotizing arteritis: a clinicopathological study. Ann. Neurol. 23:461, 1988. Schaumberg, H. H., Spencer, P. S., and Thomas, P. K.: Disorders of Peripheral Nerves. Philadelphia, F. A. Davis, 1983. Schifitto, G., McDermott, M. P., McArthur, J. C., et al.: Incidence of and risk factors for HIV-associated distal sensory polyneuropathy. Neurology 58:1764, 2002. Shlay, J. C., Chaloner, K., Max, M. B., et al.: Acupuncture and amitriptyline for pain due to HIV-related peripheral neuropathy: a randomized controlled trial. Terry Beirn Community Programs for Clinical Research on AIDS. JAMA 280:1590, 1998. Simpson, D. M.: Selected peripheral neuropathies associated with human immunodeficiency virus infection and antiretroviral therapy. J. Neurovirol. 8(Suppl. 2):33, 2002. Simpson, D. M., Haidich, A. B., Schifitto, G., et al.: Severity of HIV-associated neuropathy is associated with plasma HIV-1 RNA levels. AIDS 16:407, 2002. Simpson, D. M., Katzenstein, D., Haidich, B., et al.: Plasma carnitine in HIV-associated neuropathy. AIDS 15:2207, 2001. Simpson, D. M., McArthur, J. C., Olney, R., et al.: Lamotrigine for HIV-associated painful sensory neuropathies: a placebo-controlled trial. Neurology 60:1508, 2003. Simpson, D. M., Olney, R., McArthur, J. C., et al.: A placebocontrolled trial of lamotrigine for painful HIV-associated neuropathy. Neurology 54:2115, 2000. Simpson, D. M., and Tagliati, M.: Nucleoside analogueassociated peripheral neuropathy in human immunodeficiency virus infection. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 9:153, 1995. Skowron, G.: Biologic effects and safety of stavudine: overview of Phase I and II clinical trials. J. Infect. Dis. 171(Suppl. 2):S113, 1995. Skuta, G., Fischer, G. M., Janaky, T., et al.: Molecular mechanism of the short-term cardiotoxicity caused by 2⬘,3⬘-dideoxycytidine (ddC): modulation of reactive oxygen species levels and ADP-ribosylation reactions. Biochem. Pharmacol. 58:1915, 1999. Smith, I. L., Shinkai, M., Freeman, W. R., et al.: Polyradiculopathy associated with ganciclovir-resistant cytomegalovirus in an AIDS patient: phenotypic and genotypic characterization of sequential virus isolates. J. Infect. Dis. 173:1481, 1996. Smith, K. M., Crandall, K. A., Kneissl, M. L., et al.: PCR detection of host and HIV-1 sequences from archival brain tissue. J. Neurovirol. 6:164, 2000. So, Y. T., and Olney, R. K.: Acute lumbosacral polyradiculopathy in acquired immunodeficiency syndrome: experience in 23 patients. Ann. Neurol. 35:53, 1994.

Peripheral Neuropathies in Human Immunodeficiency Virus Infection 170. Starnes, M. C., and Cheng, Y. C.: Cellular metabolism of 2⬘,3⬘-dideoxycytidine, a compound active against human immunodeficiency virus in vitro. J. Biol. Chem. 262:988, 1987. 171. Thornton, C. A., Latif, A. S., and Emmanuel, J. C.: GuillainBarre syndrome associated with human immunodeficiency virus infection in Zimbabwe. Neurology 41:812, 1991. 172. Tokumoto, J. I., and Hollander, H.: Cytomegalovirus polyradiculopathy caused by a ganciclovir-resistant strain. Clin. Infect. Dis. 17:854, 1993. 173. Tucker, T., Dix, R. D., Katzen, C., et al.: Cytomegalovirus and herpes simplex virus ascending myelitis in a patient with acquired immune deficiency syndrome. Ann. Neurol. 18:74, 1985. 174. Van Schaik, I. N., Winer, J. B., de Haan, R., et al.: Intravenous immunoglobulin for chronic inflammatory demyelinating polyradiculoneuropathy. Cochrane Database Syst. Rev. 3:CD001797, 2002. 175. Vendrell, J., Heredia, C., Pujol, M., et al.: Guillain-Barre syndrome associated with seroconversion for anti-HTLV-III. Neurology 37:544, 1987. 176. Villa, A., Foresti, V., and Confalonieri, F.: Autonomic neuropathy and HIV infection. Lancet 2:915, 1987. 177. Villa, A., Foresti, V., and Confalonieri, F.: Autonomic nervous system dysfunction associated with HIV infection in intravenous heroin users. AIDS 6:85, 1992. 178. Villa, A., Foresti, V., and Confalonieri, F.: Autonomic neuropathy and prolongation of QT interval in human immunodeficiency virus infection. Clin. Auton. Res. 5:48, 1995.

2145

179. Walsh, C. M., Nardi, M. A., and Karpatkin, S.: On the mechanism of thrombocytopenic purpura in sexually active homosexual men. N. Engl. J. Med. 311:635, 1984. 180. Wechsler, A. F., and Ho, D. D.: Bilateral Bell’s palsy at the time of HIV seroconversion. Neurology 39:747, 1989. 181. Wiley, C. A.: Neuromuscular diseases of AIDS. FASEB J. 3:2503, 1989. 182. Williams, D., Geraci, A., and Simpson, D. D.: AIDS and AIDS-treatment neuropathies. Curr. Pain Headache Rep. 6:125, 2002. 183. Winer, J. B., Bang, B., Clarke, J. R., et al.: A study of neuropathy in HIV infection. Q. J. Med. 83:473, 1992. 184. Wiselka, M. J., Nicholson, K. G., Ward, S. C., et al.: Acute infection with human immunodeficiency virus associated with facial nerve palsy and neuralgia. J. Infect. 15:189, 1987. 185. Wrzolek, M. A., Rao, C., Kozlowski, P. B., et al.: Muscle and nerve involvement in AIDS patient with disseminated Mycobacterium avium intracellulare infection. Muscle Nerve 12:247, 1989. 186. Yarchoan, R., Perno, C. F., Thomas, R. V., et al.: Phase I studies of 2⬘,3⬘-dideoxycytidine in severe human immunodeficiency virus infection as a single agent and alternating with zidovudine (AZT). Lancet 1:76, 1988. 187. Yoshioka, M., Shapshak, P., Srivastava, A. K., et al.: Expression of HIV-1 and interleukin-6 in lumbosacral dorsal root ganglia of patients with AIDS. Neurology 44:1120, 1994. 188. Zink, M. C., and Clements, J. E.: A novel simian immunodeficiency virus model that provides insight into mechanisms of human immunodeficiency virus central nervous system disease. J. Neurovirol. 8(Suppl. 2):42, 2002.

95 Diphtheritic Polyneuropathy DAVID PLEASURE AND ALBEE MESSING

Changing Epidemiology of Diphtheria Clinical Features of Diphtheria and Diphtheritic Polyneuropathy Laboratory Features of Diphtheria and Diphtheritic Polyneuropathy

Treatment of Diphtheria and Diphtheritic Polyneuropathy Pathology of Diphtheritic Polyneuropathy

Diphtheria causes the prototypic demyelinative polyneuropathy.38 The pathogenesis and pathophysiology of this neuropathy are well established, and its occurrence is entirely preventable by a program of initial and booster immunization with diphtheria toxoid. Yet small epidemics of diphtheritic polyneuropathy continue to occur, and to cause substantial mortality and long-term disability. An additional reason for us to be interested in diphtheria is that cell-targeted diphtheria toxin is now employed clinically to ablate malignant cells10,27 and experimentally to analyze developmental processes.26,36

CHANGING EPIDEMIOLOGY OF DIPHTHERIA Prior to the advent of widespread diphtheria toxoid vaccination, diphtheria was a common, endemic childhood disorder. Infection was usually asymptomatic or mild, but dangerous complications such as laryngeal obstruction and myocarditis were frequent enough to make diphtheria the second leading cause of death in children under 15 in the United Kingdom prior to World War II.51 Today, most children in industrial and postindustrial countries receive diphtheria toxoid. Without subsequent booster immunizations, however, adults progressively lose protective immunity. In one large study, again from the United Kingdom, more than 50% of normal subjects between ages 50 and 59 had titers of serum diphtheria antitoxin below the protective range.29 In the United States, 61% of those 6 years of age or older had protective levels of diphtheria antitoxin, but

Molecular Pathogenesis and Pathophysiology of Diphtheritic Polyneuropathy

this had fallen to 30% of those 70 or older.33 Consequently, diphtheria in countries in which childhood vaccination is prevalent is predominantly a disease of adults.3,6,30,51

CLINICAL FEATURES OF DIPHTHERIA AND DIPHTHERITIC POLYNEUROPATHY Among the clinical features that raise the index of suspicion of diphtheria, in addition to knowledge of a diphtheria outbreak in the community, are a febrile illness with pronounced leukocytosis and an exudative, grayish brown pseudomembrane in the nares, soft palate, larynx, or pharynx, sometimes also involving the tracheobronchial tree.4,11 Cardiomyopathy is a common early complication of diphtheria, can cause arrhythmias, and may progress to heart failure and death.2,16,28 Corynebacterium diphtheriae can also infect cutaneous wounds, particularly in the tropics.44 In those patients with diphtheria who go on to develop polyneuropathy, neuropathic symptoms first appear 2 to 7 weeks after onset of the infection. Typical initial symptoms include numbness of the tongue and face and dysphonia.28,41 Impaired swallowing and respiratory function may necessitate tracheal intubation and artificial ventilation. The majority of patients at this stage in their illness also show abnormalities of oculomotor function (e.g., diplopia or abnormal pupillary motor responses).41 In contrast to Guillain-Barré syndrome, in which evolution of deficits is typically monophasic,43 progression in diphtheritic polyneuropathy is often biphasic.28,41 That is, 2147

2148

Diseases of the Peripheral Nervous System

motor and sensory deficits in the limbs may become prominent only as the cranial nerve deficits have begun to abate. Most patients have impaired position, vibration, pain, and temperature perception, and often demonstrate sensory ataxia. Sweating abnormalities and other evidence of autonomic dysfunction are common. Alterations in blood pressure control and cardiac rhythm may reflect both autonomic involvement and the concurrent myocarditis.19,47 Occasional patients demonstrate pale optic discs and delayed visual evoked responses, and one third show impaired bladder function.28,41 Recovery of strength and sensation in the limbs usually begins somewhat later than in Guillain-Barré syndrome, at 2 to 3 months after the first appearance of neuropathic symptoms, and the incidence of significant residual motor deficits a year after onset is higher in diphtheritic polyneuropathy than in Guillain-Barré syndrome.28,41

LABORATORY FEATURES OF DIPHTHERIA AND DIPHTHERITIC POLYNEUROPATHY Throat cultures provide an accurate means for diagnosis of diphtheria, but can yield false-negative results if processing of the specimen is delayed. Enzyme immunoassays for diphtheria toxin in throat cultures permit rapid, definitive diagnosis,13 as can polymerase chain reaction demonstration of the C. diphtheriae tox gene.12,24 Laboratory studies consistent with diphtheritic cardiomyopathy include an abnormal electrocardiogram, elevated serum creatine kinase, and echocardiographic evidence of ventricular dilatation. As in the Guillain-Barré syndrome, slowing is sometimes most prominent in distal nerve segments early in the course, and may intensify as the patient begins to show signs of clinical recovery.28,46 Also as in the GuillainBarré syndrome, cerebrospinal fluid from patients with diphtheritic polyneuropathy typically shows albuminocyto-

logic dissociation, though moderate pleocytosis is sometimes present.8 Cerebrospinal fluid pressure is elevated in a substantial proportion of patients.41

TREATMENT OF DIPHTHERIA AND DIPHTHERITIC POLYNEUROPATHY Treatment measures for symptomatic acute diphtheria include diphtheria antitoxin and penicillin.22 Antitoxin administration, by blocking cellular binding and uptake of diphtheria toxin, is most apt to have a favorable effect early in infection. It has not been established whether antitoxin administration a week or more after initial symptoms of diphtheria is effective in diminishing the incidence or severity of polyneuropathy.28,41 Glucocorticoid administration diminishes the frequency of airway obstruction in acute laryngeal diphtheria,15 and is likely to be useful in treating the serum sickness sometimes caused by diphtheria antitoxin administration,7 but has not been demonstrated to alter the incidence of cardiomyopathy or polyneuropathy.48 In experimental animals, glucocorticoid administration has been reported to improve muscle responses to stimulation of nerve segments demyelinated by diphtheria toxin,1 probably by diminishing conduction block, but not to alter the overall severity of diphtheritic neuropathy.34,35

PATHOLOGY OF DIPHTHERITIC POLYNEUROPATHY In 1881, Meyer38 provided a still classic description of the neuropathology of diphtheria that remains remarkably accurate even today.31 Meyer described withdrawal of myelin from the paranodes and segmental demyelination, with a patchy distribution but confined to the peripheral nervous system (Fig. 95–1). There was minimal inflammation and

FIGURE 95–1 Single teased nerve fibers from the phrenic nerve of a patient dying with diphtheritic polyneuritis, showing segmental (top) and paranodal (bottom) demyelination. Osmium stain. (From Meyer, P.: Anatomische untersuchungen über diphtheritische Lahmung. Virchows Arch. Pathol. Anat. Physiol. Klin. Med. 85:181, 1881.)

Diphtheritic Polyneuropathy

2149

FIGURE 95–2 Transverse sections of tail nerve fascicles in normal (left) and P0-DT transgenic (right) mice. The control nerve contains many myelinated fibers, whereas the transgenic nerve shows marked myelin deficiency, with large-diameter axons devoid of myelin and occasional thinly myelinated fibers. Toluidine blue stain; 1-␮m Epon sections. (From Messing, A., Behringer, R. R., Hammang, J. P., et al., P0 promoter directs expression of reporter and toxin genes to Schwann cells in transgenic mice. Neuron 8:507, 1992, with permission.)

only rarely axonal degeneration. Over 75 years of subsequent study have largely confirmed these initial observations.14 In some studies the pathology is most severe in and around sensory ganglia, with more distal regions of peripheral nerve only occasionally affected.14 Diphtheria toxin induces similar pathology in virtually all species on which it has been tested, with only minor variations in the distribution of lesions, though the dosage required varies by orders of magnitude among species.35 The central nervous system (CNS) appears largely unaffected,31 though there is abundant experimental evidence that the toxin can affect oligodendrocytes as well as Schwann cells32,55 and, as noted above, occasional patients do show evidence of optic nerve pathology. The relative lack of CNS pathology may reflect failure of the toxin to cross the blood-brain barrier.52 The time course of demyelination is consistent, and much of the data are derived from work on experimental animals such as guinea pigs, rabbits, rats, and chickens. The earliest changes visible at the light microscopic level are widening at the nodes of Ranvier, with retraction of paranodal loops of Schwann cell myelin and fragmentation of myelin. At the ultrastructural level there are discontinuities of various Schwann cell membranes, and sometimes curious evaginations of axonal cytoplasm into the adjacent Schwann cell.17 The lesions progress to formation of myelin ovoids and phagocytosis by both Schwann cells and macrophages, with loss of entire internodes. At the same time there is considerable prolif-

eration of Schwann cells and perhaps macrophages,14 a feature noted previously by Meyer in 1881.38 Eventually remyelination may occur; in the case of mildly affected areas, the Schwann cells simply extend back to the original configurations around the node, whereas more severely affected regions respond by integration of new Schwann cells with classic examples of short internodes. Although some studies showed effects on unmyelinated fibers, most of the pathology is confined to the myelinating Schwann cells,17 and in a transgenic model of diphtheria-induced hypomyelination (Fig. 95–2), the nonmyelinating Schwann cells represented the majority of proliferating cells.36

MOLECULAR PATHOGENESIS AND PATHOPHYSIOLOGY OF DIPHTHERITIC POLYNEUROPATHY Diphtheritic polyneuropathy and cardiomyopathy are caused by a toxic protein secreted by C. diphtheriae. The toxin is encoded by the tox bacteriophage gene18; only those C. diphtheriae strains harboring this bacteriophage cause polyneuropathy and cardiomyopathy. Diphtheria toxin is a 58-kDa protein that is composed of two chains, A (21 kDa) and B (37 kDa). The B chain contains two functional domains. The first, a ligand for

2150

Diseases of the Peripheral Nervous System

plasma membrane heparin-binding epidermal growth factor (HB-EGF) receptors,39 facilitates toxin endocytosis.25 The second, the T domain, penetrates the endosomal membrane and translocates the toxin into cytosol.45 In cytosol, the toxin A chain catalyzes transfer of ADP-ribose from NAD⫹ to polypeptide elongation factor-2 (EF-2), thus inactivating EF-2 and blocking protein synthesis.27,42 The toxin-poisoned cells undergo death by apoptosis.23 The toxic effect of intracellular diphtheria toxin A chains is very high; a single molecule is sufficient to cause death of many eukaryotic cell types. Why, then, are Schwann cells selectively targeted in patients with diphtheria, and cultured Schwann cells more susceptible than other cell types to diphtheria toxin?40 Schwann cells express abundant HB-EGF receptors, which transduce the trophic effects on Schwann cells of neuregulin,37 and thus are more capable of binding diphtheria toxin than most other cell types. The development of truncated, soluble HBEGF molecules that retain diphtheria toxin–binding capacity may ultimately prove useful in treatment of diphtheria by protecting Schwann cells from the toxin while avoiding the serum sickness sometimes elicited by diphtheria antitoxin.5 Although diphtheria is classically thought of as a pure demyelinating disease, with Schwann cells the primary target for diphtheria toxin and with sparing of axons, there is continuing debate over the degree to which axons suffer even when the initial insult is confined to the Schwann cell. Clearly Schwann cells do regulate various properties of axons. Mice with mutations in peripheral myelin protein 22 (Trembler) exhibit decreases in axoplasmic flow, with corresponding reductions in axon caliber,9,21 and early experimental studies in chickens also implicated changes in axoplasmic flow.20 A similar reduction in axon caliber is also present in the diphtheria transgenic model described above, in which there was also a loss of approximately 25% of sensory neurons and motor axons.36 Schwann cells also are a major determinant of how axons concentrate sodium channels at the node but keep potassium channels to the juxtaparanode or internode.50,54 Trembler mutant mice increase their expression of axonal K⫹ channels, with atypical localization.53 Similarly, Vabnick et al.49 found poor nodal clustering with diffuse axonal staining of sodium channels in sciatic nerves of P0-DT transgenic mice. Abnormal exposure, expression, or localization of such axonal ion channels resulting from loss of Schwann cell myelin is likely to be a central feature of the physiologic defects exhibited by these nerves.50

REFERENCES 1. Arnason, B. G. W., and Chelmicka-Szorc, E.: Peripheral nerve segmental demyelination induced by intraneural diphtheria toxin injection. I. Effect of hydrocortisone as

2.

3.

4. 5.

6.

7.

8. 9.

10. 11. 12.

13.

14. 15. 16.

17. 18.

19. 20.

21.

22.

measured by muscle twitch tension. Arch. Neurol. 30:157, 1974. Bethell, D. B., Nguyen, M. D., Ha, T. L., et al.: Prognostic value of electrocardiographic monitoring of patients with severe diphtheria. Clin. Infect. Dis. 20:1259, 1995. Bisgard, K. M., Hardy, I. R., Popovic, T., et al.: Respiratory diphtheria in the United States, 1980 through 1995. Am. J. Public Health 88:787, 1998. Bisno, A. L.: Acute pharyngitis. N. Engl. J. Med. 344:205, 2001. Cha, J. H., Brooke, J. S., Chang, M. Y., and Eidels, L: Receptor-based antidote for diphtheria. Infect. Immun. 70:2344, 2002. Christenson, B., Hellstrom, U., Sylvan, S. P., et al.: Impact of a vaccination campaign on adult immunity to diphtheria. Vaccine 19:1133, 2000. Clark, R. F., McKinney, P. E., Chase, P. B., and Walter, F. G.: Immediate and delayed allergic reactions to Crotalidae polyvalent immune Fab (ovine) antivenom. Ann. Emerg. Med. 39:671, 2002. Creange, A., Meyrignac, C., Roualdes, B., et al.: Diphtheritic polyneuropathy. Muscle Nerve 18:1460, 1995. De Waegh, S., and Brady, S. T.: Altered slow axonal transport and regeneration in a myelin-deficient mutant mouse: the Trembler as an in vivo model for Schwann cell-axon interactions. J. Neurosci. 10:1855, 1990. Debinski, W.: Local treatment of brain tumors with targeted chimera cytotoxic proteins. Cancer Invest. 20:801, 2002. Dobey, R. A., and Tobey, D. N.: Clinical features of diphtheria in the respiratory tract. JAMA 242:2197, 1979. Efstratiou, A., Engler, K. H., Mazurova, I. K., et al.: Current approaches to the laboratory diagnosis of diphtheria. J. Infect. Dis. 181(Suppl. 1):S138, 2000. Engler, K. H., Efstratiou, A., Norn, D., et al.: Immunochromatographic strip test for rapid detection of diphtheria toxin: description and multicenter evaluation in areas of low and high prevalence of diphtheria. J. Clin. Microbiol. 40:80, 2002. Fisher, C. M., and Adams, R. D.: Diphtheritic polyneuritis—a pathological study. J. Neuropathol. Exp. Neurol. 15:243, 1956. Havaldar, P. V.: Dexamethasone in laryngeal diphtheritic croup. Ann. Trop. Paediatr. 17:1, 1997. Havaldar, P. V., Sankpal, M. N., and Doddannavar, R. P.: Diphtheritic myocarditis: clinical and laboratory parameters of prognosis and fatal outcome. Ann. Trop. Paediatr. 20:209, 2000. Hildebrand, C.: Myelin sheath remodelling in remyelinated rat sciatic nerve. J. Neurocytol. 18:285, 1989. Holmes, R. K.: Biology and molecular epidemiology of diphtheria toxin and the tox gene. J. Infect. Dis. 181(Suppl. 1):156, 2000. Idiaquez, J.: Autonomic dysfunction in diphtheritic neuropathy. J. Neurol. Neurosurg. Psychiatry 55:159, 1992. Kidman, A. D., Dolan, L., and Sippe, H. J.: Blockade of fast axonal transport by diphtheritic demyelination in the chicken sciatic nerve. J. Neurochem. 30:57, 1978. Kirkpatrick, L. L., and Brady, S. T.: Modulation of the axonal microtubule cytoskeleton by myelinating Schwann cells. J. Neurosci. 14:7440, 1994. Kneen, R., Pham, N. G., Solomon, T., et al.: Penicillin vs. erythromycin in the treatment of diphtheria. Clin. Infect. Dis. 27:845, 1998.

Diphtheritic Polyneuropathy 23. Komatsu, N., Oda, T., and Muramatsu, T.: Involvement of both caspase-like proteases and serine proteases in apoptotic cell death induced by ricin, modeccin, diphtheria toxin, and Pseudomonas toxin. J. Biochem. (Tokyo) 124:1038, 1998. 24. Komiya, T., Shibata, N., Ito, M., et al.: Retrospective diagnosis of diphtheria by detection of the Corynebacterium diphtheriae tox gene in a formaldehyde-fixed throat swab using PCR and sequencing analysis. J. Clin. Microbiol. 38:2400, 2000. 25. Lanzrein, M., Garred, O., Olsnes, S., and Sandvig, K.: Diphtheria toxin endocytosis and membrane translocation are dependent on the intact membrane-anchored receptor (HB-EGF precursor): studies on the cell-associated receptor cleaved by a metalloprotease in phorbol-ester-treated cells. Biochem. J. 310:285, 1995. 26. Lee, K. J., Dietrich, P., and Jessell, T. M.: Genetic ablation reveals that the roof plate is essential for dorsal interneuron specification. Nature 403:734, 2000. 27. Li, Y., McCadden, J., Ferrer, F., et al.: Prostate-specific expression of the diphtheria toxin A chain (DT-A): studies of inducibility and specificity of expression of prostate-specific antigen promoter-driven DT-A adenoviral-mediated gene transfer. Cancer Res. 62:2576, 2002. 28. Logina, I., and Donaghy, M.: Diphtheritic polyneuropathy: a clinical study and comparison with Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 67:433, 1999. 29. Maple, P. A., Efstratious, A., George, R. C., et al.: Diphtheria immunity in UK blood donors. Lancet 345:963, 1995. 30. Markina, S. S., Maksimova, N. M., Vitek, C. R., et al.: Diphtheria in the Russian Federation in the 1990s. J. Infect. Dis. 181(Suppl. 1):S27, 2000. 31. McDonald, W. I., and Kocen, R.: Diphtheritic neuropathy. In Dyck, P. J., Thomas, P. K., Griffin, J.W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1412, 1993. 32. McDonald, W. I., and Sears, T. A.: Effect of demyelination on conduction in the central nervous system. Nature 221:182, 1969. 33. McQuillan, G. M., Kruszon-Moran, D., Deforest, A., et al.: Serologic immunity to diphtheria and tetanus in the United States. Ann. Intern. Med. 136(Suppl.):660, 2002. 34. Mellick, R.: The blood vessel permeability of peripheral nerve during demyelination and the effect of ACTH therapy on the permeability and on functional disability. Proc. Aust. Assoc. Neurol. 6:133, 1969. 35. Mellick, R., Dolan, L., and Kidman, A. D.: Experimental diphtheritic neuropathy in the mouse: a study in cellular resistance. Br. J. Exp. Pathol. 56:471, 1975. 36. Messing, A., Behringer, R. R., Hammang, J. P., et al.: P0 promoter directs expression of reporter and toxin genes to Schwann cells in transgenic mice. Neuron 8:507, 1992. 37. Meyer, D., and Birchmeier, C.: Multiple essential functions of neuregulin in development. Nature 378:386, 1995. 38. Meyer, P.: Anatomische untersuchungen über diphtheritische Lahmung. Virchows Arch. Pathol. Anat. Physiol. Klin. Med. 85:181, 1881. 39. Mitamura, T., Umata, T., Nakano, F., et al.: Structurefunction analysis of the diphtheria toxin receptor toxin

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

2151

binding site by site-directed mutagenesis. J. Biol. Chem. 272:27084, 1997. Pappenheimer, A. M. Jr., Harper, A. A., Moynihan, M., and Brockes, J. P.: Diphtheria toxin and related proteins: effect of route of injection on toxicity and the determination of cytotoxicity for various cultured cells. J. Infect. Dis. 145:94, 1992. Piradov, M. A., Pirogov, V. N., Popova, L. M., and Avdunina, I. A.: Diphtheritic polyneuropathy: clinical analysis of severe forms. Arch. Neurol. 58:1438, 2001. Pleasure, D. E., Feldmann, B., and Prockop, D. J.: Diphtheria toxin inhibits the synthesis of myelin proteolipid and basic proteins by peripheral nerve in vitro. J. Neurochem. 20:81, 1973. Pleasure, D. E., Lovelace, R. E., and Duvoisin, R. C.: The prognosis of acute polyradiculoneuritis. Neurology 18:1143, 1968. Rahman, K. M., Khan, H. M., and Haq, J. A.: Incidence of cutaneous diphtheria in Bangladesh. Bangladesh Med. Res. Counc. Bull. 9:49, 1983. Ren, J., Kachel, K., Kim, H., et al.: Interaction of diphtheria toxin T domain with molten globule-like proteins and its implications for translocation. Science 284:955, 1999. Roche, S., Stone, S., Allen, C. M., and Lange, L. S.: Diphtheritic polyneuritis in an elderly woman: clinical and neurophysiological follow-up. Br. J. Clin. Pract. 44:285, 1990. Solders, G., Nennesmo, I., and Persson, A.: Diphtheritic neuropathy, an analysis based on muscle and nerve biopsy and repeated neurophysiological and autonomic function tests. J. Neurol. Neurosurg. Psychiatry 52:876, 1989. Thisyakorn, U., Wongvanich, J., and Kumpeng, V.: Failure of corticosteroid therapy to prevent diphtheritic myocarditis or neuritis. Pediatr. Infect. Dis 3:126, 1984. Vabnick, I., Messing, A., Chiu, S. Y., et al.: Sodium channel distribution in axons of hypomyelinated and MAG null mutant mice. J. Neurosci. Res. 50:321, 1997. Vabnick, I., Trimmer, J. S., Schwarz, T. L., et al.: Dynamic potassium channel distributions during axonal regeneration prevent aberrant firing patterns. J. Neurosci. 19:747, 1999. Vitek, C. R., and Wharton, M.: Diphtheria in the former Soviet Union: reemergence of a pandemic disease. Emerg. Infect. Dis. 4:539, 1998. Waksman, B. H.: Experimental study of diphtheritic polyneuritis in the rabbit and guinea pig. III. The bloodnerve barrier in the rabbit. J. Neuropathol. Exp. Neurol. 19:35, 1961. Wang, H., Allen, M. L., Grigg, J. J., et al.: Hypomyelination alters K⫹ channel expression in mouse mutants shiverer and Trembler. Neuron 15:1337, 1995. Wang, H., Kunkel, D. D., Martin, T. M., et al.: Heteromultimeric K⫹ channels in terminal and juxtaparanodal regions of neurons. Nature 365:75, 1993. Wisniewski, H., and Raine, C. S.: An ultrastructural study of experimental demyelination and remyelination. V. Central and peripheral nervous system lesions caused by diphtheria toxin. Lab. Invest. 25:73, 1971.

96 Parasitic Infections of the Peripheral Nervous System SUDHA POTTUMARTHY AND THOMAS R. FRITSCHE

Parasites Affecting the Nerves Innervating the Visceral Organs American Trypanosomiasis (Trypanosoma cruzi) Schistosomiasis (Schistosoma spp.) Strongyloidiasis (Strongyloides stercoralis) Parasites Affecting the Cranial Nerves Free-Living Amebae (Acanthamoeba spp.) Angiostrongyliasis (Angiostrongylus cantonensis)

Trichinosis (Trichinella spiralis) Parasites Affecting the Peripheral Nerves Leishmaniasis (Leishmania spp.) Giardiasis (Giardia lamblia) Cysticercosis (Taenia solium) Coenuriasis (Taenia [Multiceps] spp.) Hydatidosis (Echinococcus granulosis) Dracunculiasis (Dracunculus medinensis) Onchocerciasis (Onchocerca volvulus) Parasites Associated with Guillain-Barré Syndrome

Parasitic infections cause significant morbidity and mortality in many regions of the world. Because most infections are not reported to public health officials, the actual incidence of parasitic infections in the United States and elsewhere is unknown. Many parasitic infections were once considered exotic tropical diseases. However, because of the popularity of travel and increases in migration, the world is now a smaller place and these infections may be encountered in clinical practice in any part of the world. Detailed travel history accompanied by a strong clinical index of suspicion often help make the accurate clinical diagnosis. Furthermore, with the increase in the number of immunocompromised hosts as a result of human immunodeficiency virus (HIV) infection, cancer, chemotherapy, and the like, there is a large population at increased risk of serious sequelae following exposure to parasitic infections as well as other infectious diseases. Acanthamoeba species, Naegleria fowleri, Toxoplasma gondii, Cysticercus (larval Taenia solium), Leishmania species, Entamoeba histolytica, Gnathostoma species, and Trypanosoma gambiense are only a few of the several parasites that infect the central nervous system.

Malaria (Plasmodium spp.) Toxoplasmosis (Toxoplasma gondii) Cyclosporiasis (Cyclospora cayetanensis) Tick Paralysis Mimicking Guillain-Barré Syndrome

Conversely, the peripheral nervous system is infected by only a few parasites (Table 96–1). Parasites either infect the peripheral nervous system by direct invasion, with concurrent tissue damage, or indirectly affect host tissues and organs through metabolic disturbances, pressure effects, aberrant migration, toxins, and immunologic mechanisms. Neuropathies that may have only a chance or spurious association with parasitism are also recognized. In addition, isolated case reports of parasites causing neuropathies emphasize the importance of chance encounters that might occur as a result of either aberrant migration or unusual localization of either larval or adult parasitic forms. Of all parasitic infections affecting the peripheral nervous system, Chagas’ disease remains the most prevalent and important. In this chapter a variety of protozoal and helminthic infections producing either primary or secondary peripheral neuropathic lesions are described. The chapter is organized into sections to emphasize the effects of parasitic infections on various components of the peripheral nervous system—nerves innervating the visceral organs, cranial nerves, and peripheral nerves—and finally those parasites associated with Guillain-Barré syndrome 2153

Neuritis

Recurrent lumbosacral and brachial plexopathy Footdrop

Peripheral Nerves Mild sensorimotor peripheral neuropathy Median nerve neuritis

Abducens nerve paralysis Facial nerve paralysis

Cranial Nerves Radial keratoneuritis (enlarged corneal nerves) Facial nerve paralysis (lower motor neuron type)

Megasyndromes Esophagus Colon Stomach (rare) Duodenum (rare) Atonic dilated ureter

Visceral Organs Cardiomyopathy

Peripheral neuropathic lesions

Post–kala-azar dermal leishmaniasis

Trichinosis

Schistosomiasis

American trypanosomiasis (Chagas’ disease) Schistosomiasis

Trichinosis

Trichinella spiralis Leishmania donovani complex

Schistosoma haematobium and S. mansoni Schistosoma japonicum

Trypanosoma cruzi

Angiostrongylus cantonensis Trichinella spiralis

Angiostrongylus cantonensis

Angiostrongyliasis

Angiostrongyliasis

Acanthamoeba spp.

Schistosoma haematobium

Trypanosoma cruzi

Trypanosoma cruzi

Parasite

Acanthamoeba keratitis

Schistosomiasis

American trypanosomiasis (Chagas’ disease) American trypanosomiasis (Chagas’ disease)

Disease

India, China, and East Africa

Southwestern United States to Argentina and Chile Between 36° north and 34° south latitude, mainly in Africa Between 36° north and 34° south latitude in China and Southeast Asia Worldwide

Worldwide

”

Africa, China, Southeast Asia, India, and Pacific Basin, including Hawaii

Cosmopolitan, worldwide

Between 36° north and 34° south latitude in Africa and Mideast

Southwestern United States to Argentina and Chile Southwestern United States to Argentina and Chile

Geographic distribution

Table 96–1. Summary of Peripheral Neuropathic Lesions Caused by Various Parasites

Infiltration of the nerves with lymphohistiocytic cells and occasionally parasites

Peripheral neuritis

Parainfectious disimmune phenomenon

Segmental and paranodal demyelination and axonal loss Autoimmune reaction caused by parasite toxins or antigens

Peripheral neuritis

”

Granulomatous inflammation, radial infiltrates along corneal nerves Possibly inflammation caused by migration of living larvae or dead and degenerating parasites in the neural tissue

Schistosomal perineuritis or nonspecific perineuritis or a combination of the two

Focal inflammation, repair, and fibrosis of the myocardium Denervation involving submucosal (Meissner’s) and myenteric (Auerbach’s) plexuses

Cause

16

43

48

53

25, 65, 66, 78

43

6

6

36, 40, 51

26

13, 37, 57

37, 57, 59

References

Giardiasis

Cysticercosis

Cysticercosis

Hydatidosis

Hydatidosis

Dracunculiasis

Onchocerciasis

Distal paresthesias

Median nerve compression

Ulnar nerve compression

Distal paresthesias and weakness

Footdrop

Carpel tunnel syndrome

Median nerve compression

Guillain-Barré syndrome (mimicking)

Tick paralysis

Guillain-Barré Syndrome Guillain-Barré Malaria syndrome Guillain-Barré Toxoplasmosis syndrome Guillain-Barré Cyclosporiasis syndrome

Cutaneous leishmaniasis

Neuritis (diminished sensation or hyperesthesia)

Ixodes holocycles, I. rubicundus, Dermacentor andersoni, D. variabilis, and others

Cyclospora cayetanensis

Toxoplasma gondii

Plasmodium spp.

Onchocerca volvulus

Dracunculus medinensis

Echinococcus granulosus

Echinococcus granulosus

Larval form of Taenia solium

Larval form of Taenia solium

Giardia lamblia

Leishmania spp.

Noted in increasing number of countries in the world Worldwide in occurrence

45° north to 40° south of the equator Worldwide

Central Africa, Central America, northern South America, and Arabian peninsula

Old World—Africa, Asia, Mediterranean, and Middle East New World—Mexico and Central America Cosmopolitan, worldwide Mexico, rest of Latin America, Europe, Africa, India, and Asia Mexico, rest of Latin America, Europe, Africa, India, and Asia Sheep- and cattle-raising areas of the world Sheep- and cattle-raising areas of the world Africa and Asia

Salivary neurotoxin of ticks impairs conduction in the motor nerves by interference with the voltage-gated sodium channels of the axons

Possibly immunologically mediated Possibly immunologically mediated Possibly immunologically mediated

Entrapment neuropathy caused by abscessed, calcified aberrant worm Mechanical effect caused by adult worm

Mechanical compression caused by intrapelvic cysts

Mechanical effect of cysticercus on the deep branch of the nerve Possibly immunoallergic

Mechanical and inflammatory effects of larva adherent to nerve

Possibly malabsorption

Infiltration of the perineural space and/or the nerves with lymphohistiocytic cells and occasionally parasites

19, 70

61

4, 5

33, 68

67

2, 49

28, 30

56

69

55, 76

3

39

2156

Diseases of the Peripheral Nervous System

(GBS). The division is somewhat arbitrary because some parasites may affect more than one component.*

PARASITES AFFECTING THE NERVES INNERVATING THE VISCERAL ORGANS American Trypanosomiasis (Trypanosoma cruzi) American trypanosomiasis (Chagas’ disease) is a zoonosis caused by the hemoflagellate Trypanosoma cruzi, occurring in animals and humans throughout the Americas. Transmission of the parasite occurs via triatomine insects of the family Reduviidae that live in close association with human reservoirs, including dogs, cats, armadillos, opossums, raccoons, and rodents. The most ubiquitous sylvatic reservoir host is the opossum, Didelphis virginiana. Human infections occur mainly in rural areas where poor sanitary, socioeconomic, and housing conditions allow for breeding of the reduviid vectors with easy access to humans for blood meals and thereby transmission of the infection.71 Chagas’ disease is found in all countries of the Americas, from the southwestern United States to Argentina and Chile. However, there are various strains of T. cruzi that differ in terms of epidemiology, pathogenicity, immunogenicity, biochemistry, and response to treatment.45,57 The strains are assigned to two major T. cruzi groups, I and II. Trypanosoma cruzi I is highly associated with opossums and is frequently seen in human infections north of the Amazon region. Trypanosoma cruzi I is only associated with cardiac lesions. Trypanosoma cruzi II is primarily associated with rodents, and is associated with both cardiac and digestive tract lesions in human infections.57 Transmission At the time the reduviid insect is taking a blood meal, infective metacyclic trypomastigotes are released in the feces, contaminating the skin surface. These infective trypomastigotes enter the body as a result of scratching or rubbing at the bite site or through intact mucosal surfaces of the mouth or conjunctiva. Other modes of transmission have also been described, but are less frequent. Transfusion-associated Chagas’ disease is a serious problem in endemic areas. The estimated risk of acquiring infection is 13% to 23% for each unit of contaminated blood transfused.45 Congenital transmission occurs in about 1% of deliveries of seropositive mothers. Personto-person transmission can also occur by breast-feeding

*Although Trypanosoma cruzi, Schistosoma species, and Trichinella spiralis affect more than one component of the peripheral nervous system, and appear in more than one section in Table 96–1, in the text they are addressed only in the first section discussing a component that they affect.

or by organ transplantation. Occasionally, transmission can occur either by ingestion of food contaminated with bug feces or by ingestion of the bugs themselves.45 In addition, laboratory-acquired infections have been documented as a result of accidental ingestion or inoculation of the parasites.31 Life Cycle The life cycle of T. cruzi consists of the epimastigote form, which replicates in the intestinal tract of the insects; infective metacyclic trypomastigotes found in the vector’s hindgut; the circulating blood-stage trypomastigote in mammals; and, finally, the replicating intracellular amastigote form (Fig. 96–1). After inoculation into the mammalian host, the infective T. cruzi enter local reticuloendothelial and connective tissue cells, differentiate into amastigotes, and begin replication. When the cell is filled, the amastigotes transform into trypomastigotes, lyse the cells, and infect adjacent cells and/or enter the circulation. These circulating trypomastigotes disseminate the infection by penetrating muscle cells (cardiac, smooth, and skeletal), neurons, lymph nodes, liver, and spleen (Fig. 96–2D). Clinical Disease Acute Phase. Severity of acute illness may vary from being asymptomatic to fatal. Severe illness is common in young children and occurs following an incubation period of 7 to 14 days, during which there is onset of fever, elastic nonpitting edema, hepatosplenomegaly, and peripheral lymphadenopathy. About one half of the patients have an inoculation granuloma, with an area of dusky, erythematous induration of the skin known as a chagoma and satellite lymphadenopathy at the site of entry of the organism. Unilateral and bipalpebral chronic edema associated with local lymphadenopathy (Romaña’s sign) may occur in cases of inoculation of the eye (see Fig. 96–2A). A small percentage of individuals (⬍10%) develop severe illness with meningoencephalitis and myocarditis. Acute illness of this nature resolves in 4 to 8 weeks and patients then enter the indeterminate phase. Indeterminate Phase. Patients in the indeterminate phase remain asymptomatic but latently infected. Ten to 30% of these patients progress and develop signs and symptoms of chronic Chagas’ disease over the next 2 to 3 decades. Chronic Phase. In this phase patients present with symptoms of cardiac lesions, megasyndrome disease of the gastrointestinal tract, and nervous system involvement. While the involvement of the central nervous system is well recognized in the acute phase and in cases of immunosuppression, the occurrence of central nervous system lesions in chronic Chagas’ disease is still debatable.8,57

Parasitic Infections of the Peripheral Nervous System

2157

FIGURE 96–1 Life cycle of Trypanosoma cruzi. (From Trypanosomiasis, American. Parasites and Health. In DPDx: Laboratory Identification of Parasites of Public Health Concern. Atlanta, Centers for Disease Control and Prevention, 2003. Available at www.dpd.cdc.gov/DPDx/HTML/TrypanosomiasisAmerican.htm) See Color Plate

Disease Reactivation. Recrudescence of symptomatic disease is seen in seropositive patients in the presence of immunosuppression, secondary to chemotherapy or chronic steroid therapy, or following organ transplantation.45,57 Increased parasite proliferation, with necrotic or tumorlike lesions in the brain (75%) presenting as meningoencephalitis (unifocal or multifocal), and increasing myocarditis (50%) with arrhythmias and heart failure are often found in the presence of HIV coinfection in conjunction with a low peripheral blood CD4 count and in organ transplantation.8,20,45,57 Pathogenesis The pathogenesis of end-organ damage in Chagas’ disease is incompletely understood. Postulated mechanisms include parasite antigen-specific inflammation, parasiteinduced myocardial cell necrosis and repair, microvascular spasm, and autoimmunity.17,35 It is not clear if T. cruzi amastigotes or parasite DNA is required for tissue destruc-

tion.45 While the finding of tissue-dwelling amastigotes in infected tissue is infrequent by histologic examination, T. cruzi DNA has been detected by polymerase chain reaction.35,45 Immunohistochemistry has also been used successfully to detect foci of parasite antigen in tissue lesions of patients with chagasic heart disease.35 The absence of parasites has led to the belief that infection with T. cruzi stimulates autoimmunity, which results in chronic pathologic changes. Autoimmune lymphocytes and autoantibodies targeting the endocardium, interstitium of striated muscle, gangliosides in the neural tissues, and blood vessels have been detected in chagasic patients.1,17,24,35,45 Pathophysiologic Characteristics The chronic cardiac form of Chagas’ disease is characterized by a focal inflammatory process leading to progressive destruction of cardiac fibers followed by reparative, reactive fibrosis affecting multiple areas of the myocardium.37,59 The conduction system, mainly the

B

A

D

C

E

FIGURE 96–2 A, Periocular edema known as Romaña’s sign. B, Apical aneurysm resulting from chronic Chagas’ disease. (A and B from Zaiman, H.: A Pictorial Presentation of Parasites. Available at parasitepics.biosci.uga.edu/) C, Chagasic megacolon. (From Zaiman, H.: A Pictorial Presentation of Parasites: Available at parasitepics.biosci.uga.edu/) D, Trypomastigote of Trypanosoma species in stained blood films; note the nucleus, kinetoplast, and undulating membrane (oil immersion). (From Parasite Image Library. In DPDx: Laboratory Identification of Parasites of Public Health Concern. Atlanta, Centers for Disease Control and Prevention, 2003. Available at www.dpd.cdc.gov/DPDx/HTML/Image_Library.htm) E, Small bowel biopsy demonstrating numerous eggs of Schistosoma japonicum within the lamina propria and submucosa, associated with a chronic inflammatory response (hematoxylin-eosin, ⫻400 magnification). See Color Plate

Parasitic Infections of the Peripheral Nervous System

right bundle and left anterior fascicle and parasympathetic cardiac nerves, is preferentially involved. The focal myocardial fibrosis predisposes to cardiac dilatation and cardiac failure, and results in the formation of left ventricular aneurysms.59 Gastrointestinal dysfunction in chronic disease occurs as a result of loss of neurons.50 Both sympathetic and parasympathetic systems are involved. This denervation involves the submucosal (Meissner’s) and myenteric (Auerbach’s) plexuses. The degree of denervation and segments of the bowel affected vary among patients.13,37,57 Presence of preganglionic lesions and a decrease in the dorsal cells of the motor nucleus of vagal innervation have also been noted.37,46 Histologic changes in the skeletal muscle and peripheral nerves occur in a small proportion of individuals with chronic Chagas’ disease.66 Muscle histology shows type I and type II grouping, atrophic angular fibers, and targetoid muscle fibers. Segmental and paranodal demyelination and axonal loss are seen in sural nerve biopsies.66 Cardiac Lesions. Cardiac involvement is the most common and serious manifestation of chronic Chagas’ disease. It is arrhythmogenic by nature and is the most common cause of cardiomyopathy in South and Central America. It is the leading cause of cardiovascular death in patients between the ages of 30 and 50 years.59 Manifestations typically include arrhythmias, congestive heart failure, and thromboembolic events.37,57,59 Sudden death is a typical feature of Chagas’ heart disease. Both bradyarrhythmias and tachyarrhythmias can occur in Chagas’ heart disease, and ventricular tachyarrhythmias and atrioventricular (AV) conduction abnormalities can coexist in a patient. Electrocardiographic features may include second- and third-degree AV blocks, intraventricular blocks, sinus bradycardia, and complex premature ventricular contractions, including runs of ventricular tachycardia.57,59 Sustained ventricular tachycardia may also occur. Congestive heart failure develops slowly, presenting 20 or more years following acute infection, and is often biventricular.59 Left ventricular aneurysm with or without thrombus formation occurs in 19% of chagasic patients in the endemic area; frequency is related to the severity of the cardiomyopathy.57 Aneurysms are preferentially localized at the apex of the left ventricle (see Fig. 96–2B) and in 95% of cases are associated with electrocardiographic abnormalities. Thromboembolic manifestations arise from mural thrombi in cardiac chambers, with emboli found mainly in the lungs, kidney, spleen, and brain. Sudden death occurs in about 38% of patients, mainly in males between the ages of 38 and 46 years.57 In a majority of cases it is due to ventricular fibrillation, but other causes include bradyarrhythmias, thromboembolic phenomenon, and rupture of an apical aneurysm.59

2159

Gastrointestinal Disease (Megasyndromes). Fifteen to 20% of chagasic patients in endemic areas develop alterations in motility, secretion, and absorption of the digestive tract.57 Changes in motility arise first with slowing transit, followed by an increase in caliber and increased difficulty in emptying.57 ESOPHAGUS. The most common presenting symptom of chagasic gastrointestinal disease is dysphagia. The extent of dysphagia correlates with the degree of megaesophagus.37 Odynophagia, regurgitation, eructations, hiccoughs, and cough also occur.37,55,57 Recurrent pulmonary infections resulting from aspiration and malnutrition often contribute to death of untreated patients.37 Manometry in patients with megaesophagus demonstrates a nonrelaxing or partially relaxing lower esophageal sphincter and aperistalsis upon deglutition.13 Barium esophagography reveals absence of mucosal irregularities and a variably dilated esophagus with a narrow, conical segment of 1 to 4 cm at its lower end demonstrating a nonrelaxing sphincter.13 C OLON . The colon is the next most frequently affected segment of the gastrointestinal tract after the esophagus (see Fig. 96–2C).13,37 The sigmoid colon is dilated in all cases and the rectum in 80%, while dilatation of the more proximal segments of the colon is rare.13 Megacolon often presents as a slow, progressive constipation. Complications include impaction, fecaloma, “overflow” diarrhea, pseudo-obstruction, volvulus of the redundant sigmoid, peritonitis, and ulcers.13,37,57 Nonectasic colonopathy is demonstrated by manometry and response to cholinergic substances (methacholine).57 Barium enema is the diagnostic method of choice, with demonstration of increased diameter and length of the colon and lack of haustral markings.13,37 OTHER LESIONS OF THE GASTROINTESTINAL TRACT: SALIVARY GLANDS, STOMACH, SMALL INTESTINE, AND BILIARY TRACT. Hypertrophy of the salivary glands, mainly the parotids, is found in approximately 25% of patients with chagasic megaesophagus.37 This often results in increased salivation and in an increased response to stimuli that cause salivation. While the exact mechanism is not understood, denervation is thought to play a role.37 Stomach involvement is less common and has generally been noted in patients with involvement of the esophagus and colon.13,37 The denervation as noted is said to result in hypotonia, hypoperistalsis, delayed emptying, and decreased acid production.13,37 Abnormal dilatations of the small intestine are less frequent than those of either the esophagus or colon, with megaduodenum being more frequent than megajejunum.13 Patients with megaduodenum may present with symptoms of nonspecific dyspepsia, nausea, and vomiting. Chronic intermittent intestinal pseudoobstruction and small intestinal bacterial overgrowth

2160

Diseases of the Peripheral Nervous System

syndrome have been noted in patients with dilatation of both duodenum and jejunum.37 While destruction of intrinsic neurons of the gallbladder was found in all patients seropositive for Chagas’ disease, dilatation of the gallbladder and biliary tree are uncommon manifestations of chronic Chagas’ disease.13 Peripheral Nervous System. Studies have shown that about 10% of patients with chronic or indeterminate Chagas’ disease develop a clinical mild sensorimotor peripheral neuropathy.25,65,66,78 The main clinical features include sensory impairment in the form of paresthesia and distal limb hypesthesia and diminished tendon reflexes that occur in isolation or combined with sensory impairment. Abnormalities are most frequently found in the lower limbs, followed by all four limbs, and seldom are found in upper limbs alone.66 Electrophysiologic evaluation may show one or more of the following features: diminished interference pattern, with most of the remaining motor unit potentials being polyphasic; reduced number of functional motor units in the thenar, hypothenar, soleus, and extensor digitorum brevis muscles; slow sensory and motor nerve conduction velocities; low sensory action potential amplitude; and impairment of neuromuscular transmission.25,66

Schistosomiasis (Schistosoma spp.) Schistosomiasis, also known as bilharziasis or snail fever, is an infection of immense public health importance caused by blood flukes of the genus Schistosoma. The most important species of these digenetic trematodes infecting humans are S. haematobium, S. mansoni, and S. japonicum. It is estimated that 500 to 700 million individuals of 77 countries in the world are exposed to this infection and about 200 million are infected.72 The geographic distribution of schistosomiasis is between 36 degrees north and 34 degrees south latitude and coincides with the average freshwater temperatures of 25° to 30° C. The endemic human schistosomiasis is ecologically dependent on the presence of the appropriate snail intermediate host and deposition of schistosome ova into its warm freshwater habitat. Schistosoma haematobium occurs in many parts of Africa, the Middle East, and Madagascar; S. mansoni occurs in Africa, especially in the Nile delta, southern Africa, and Madagascar, as well as Brazil, Venezuela, Surinam, and certain Caribbean islands, including Puerto Rico; and S. japonicum occurs in China, Southeast Asia, and the Philippines.21 Life Cycle and Transmission The life cycles of the human schistosomes are similar. The adult worms reside in the venules of the portal system of the bowel (S. mansoni and S. japonicum) (see Fig. 96–2E)

or the venules in the wall of the urinary bladder (S. haematobium). The females deposit the eggs in the smallest venule, where a strong inflammatory response results in the extrusion of the eggs into the lumen of the intestine or the bladder and subsequent passage into the environment with either the feces or the urine. These embryonated eggs hatch in the freshwater and the miracidia undergo further asexual development in specific freshwater snails. Forktailed larvae known as cercariae emerge from the snail host and penetrate the skin of the susceptible hosts, including humans. The larvae lose their tails to become schistosomules, which then enter the circulation and pass through the lungs before reaching the mesenteric-portal vessels. Clinical Disease and Pathology The pathogenesis and pathology of schistosomiasis is one of a foreign body reaction to different stages of the schistosome parasite. However, the granulomatous response to the numerous eggs being deposited in the different viscera is the main cause of the clinical and pathologic manifestations of schistosomiasis. The stages of schistosomiasis can be divided into the migratory phase, with cercarial dermatitis and occasionally pneumonitis following penetration and migration of cercariae; an acute systemic phase (known as Katayama fever) that presents as serum sickness–like syndrome and occurs 4 to 7 weeks after infection as a result of the antigenic challenge at the time of first egg release; and finally chronic schistosomiasis, which is primarily due to the host’s granulomatous response to the schistosome eggs deposited in the tissues. In the acute stage the granuloma surrounding each egg is large and diffuse and is mainly composed of eosinophils, neutrophils, and mononuclear cells. In the chronic phase, egg granulomas are well circumscribed with a predominance of macrophages, lymphocytes, fibroblasts, and multinucleated giant cells.72 Collagen deposition leading to fibrosis and scarring are the key features of this phase. The chronic phase is divided into two basic patterns of disease: gastrointestinal (S. mansoni and S. japonicum) and urinary (S. haematobium) schistosomiasis. Gastrointestinal schistosomiasis presents as intestinal schistosomiasis, with symptoms of focal granulomatous large bowel disease leading to malaise, mucohemorrhagic diarrhea, and cramping abdominal pains; or as hepatosplenic schistosomiasis, with symptoms of hepatomegaly, splenomegaly with associated hypersplenism, and portal hypertension. Urinary schistosomiasis initially presents as dysuria, frequency, and hematuria, with chronic infections progressing to obstructive uropathy, chronic bacteriuria, and possibly bladder calcification and carcinoma of the bladder.72 Location of egg granulomas in the nervous system can present as cerebral schistosomiasis, with focal and generalized seizures, or spinal schistosomiasis, which usually presents as transverse myelitis. These complications are

Parasitic Infections of the Peripheral Nervous System

often produced by S. japonicum infections. This species produces many more eggs than the other species, and more commonly moves to ectopic sites. Peripheral Neuropathy. Mostafa et al. described early and late lesions of schistosomal neuritis.53 In the early phase of both mansoni and haematobium schistosomiasis, the terminal ramifications of the median nerve in the forearm were affected. In the chronic group, muscle biopsy specimens revealed endoneural fibrosis and paucity of nerve fibers. They concluded that denervation corresponded to the degree of chronicity, denervation was of sufficient duration to allow reinnervation and regeneration of the affected motor units by sprouting healthy terminal fibers, and peripheral lesions of the intramuscular nerves were present. Because no worms or eggs were in the biopsy specimens, it was concluded that the damage to the nerves was the result of an autoimmune reaction to the parasite toxins or the antigens. In a study from Egypt in which schistosome-infected ureters were examined histopathologically, ureteral nerves were found to be affected in 62 of the 80 ureters.26 The nerves in the schistosomal ureters were found to be affected by pure perineuritis (44.5%), in which the schistosomal ova penetrated the nerve sheath and, along with the associated granulomatous reaction, strangled the nerve bundles and caused atrophy; nonspecific perineuritis caused by microbial infection (38.8%); and a mixture of schistosomal and nonspecific perineuritis in 16.7% of cases. The difference between the nerve involvement by schistosomiasis in the dilated and nondilated segments of the ureter was statistically significant. While the nerve involvement with fibrosis and the resulting muscle atony and atrophy resulted in dilatation, in contrast, the strictured zones showed greatest schistosomal involvement of the nerves with extensive fibrosis and total ureteral wall replacement by calcified ova. A report from Wisconsin described a 42-year-old man who presented with clinical and electrophysiologic evidence of recurrent neuralgic amyotrophy involving both brachial and lumbar plexuses.48 The illness occurred in the context of rectal biopsy–proven S. japonicum infection. Because of the absence of evidence for direct parasitic infection and normal myelographic findings, the neuropathy was thought to be the result of a parainfectious dysimmune phenomenon.

Strongyloidiasis (Strongyloides stercoralis) Strongyloidiasis is an infection produced by the “threadworm” parasite Strongyloides stercoralis. Adult worms reproduce in the bowel, releasing larvae that may be passed in feces that can reinfect the host directly. It occurs in many areas of the tropics and subtropics, but also occurs in temperate regions and is endemic in areas of the southeastern United States. Hyperinfection syndrome, which

2161

occurs in individuals with compromised immunity, results in widespread dissemination of the infection with involvement of various organs, including the brain and meninges. However, peripheral neuropathy caused by strongyloidiasis has not been reported. Association of paralytic ileus with hyperinfection strongyloidiasis is thought to be the result of larvae migrating in the bowel wall causing inflammation and fibrosis, to the point of causing hypomotility and eventual splinting.58 Another report described a 35-year-old woman who had been taking steroids for chronic renal disease, and died of renal failure and hyperinfection strongyloidiasis.10 Autopsy revealed filariform larvae of S. stercoralis concentrated in the vessel walls of the ileum, and lesser numbers aligned with the small nerves. The larvae caused no inflammation of the nerves, and the only neuropathic symptom was that of ileus. The authors postulated a tropism between the larvae and the nerves with direct mechanical or toxic effects.

PARASITES AFFECTING THE CRANIAL NERVES Free-Living Amebae (Acanthamoeba spp.) Pathogenic small, free-living amebae (FLA) of the genera Acanthamoeba, Naegleria, and Balamuthia are ubiquitous, with a worldwide distribution. They have been isolated from diverse environmental sources including soil and water (both natural and man-made sources), where they feed on other microscopic organisms, especially bacteria and yeasts. FLA primarily cause three distinct clinical syndromes in humans: primary amebic meningoencephalitis (PAM), granulomatous amebic encephalitis (GAE), and amebic (acanthamebic) keratitis.15,21 PAM is caused by the ameboflagellate Naegleria fowleri, and affects healthy children and young adults who have been swimming or diving in warm freshwater lakes or pools. The ameboflagellate enters the brain via the cribriform plate and olfactory bulbs to reach the frontal lobes, where it produces a rapid, fatal acute hemorrhagic meningoencephalitis. GAE, in contrast, is a subacute or chronic infection of immunosuppressed or debilitated individuals resulting in death in weeks to months following the onset of illness. It is primarily caused by several species of Acanthamoeba (A. castellani, A. culbertsoni, A. polyphaga, A. astronyxis, A. palestinensis, and A. healyi among others), but a recently described leptomyxid ameba, Balamuthia mandrillaris, has also been implicated.15,21 The infection is thought to spread hematogenously from primary foci in the skin, pharynx, or respiratory tract. Within lesions both trophozoites and cysts are observed amidst a chronic granulomatous inflammatory response with multinucleated giant cells. The brain involvement is patchy but mainly involves the thalamus, diencephalon, brainstem, and posterior fossa structures.

2162

Diseases of the Peripheral Nervous System

Amebic (Acanthamebic) Keratitis and Keratoneuritis The first case of Acanthamoeba keratitis was reported in 1972 when a Texas rancher splashed contaminated river water into his eye. Additional cases associated with minor ocular trauma were reported in the next decade. A marked increase in the number of cases was noted in the mid1980s, and an association with the use of contact lenses was soon established.15,40 Acanthamoeba keratitis is estimated to affect 1 in 250,000 individuals in the United States, and greater than 400 cases have been reported in the United Kingdom alone.40 It is thought to be an important infectious cause of enlarged, prominent corneal nerves.36 Transmission Acanthamoeba species have been isolated from numerous sources: freshwater, seawater, chlorinated water from swimming pools, dental treatment units, and contact lens cases and solutions.40 The cysts of Acanthamoeba are airborne in dust, and their presence in the atmosphere is thought to be an important factor in the prevalence of keratitis. The victims are primarily young, healthy individuals who use daily-wear disposable soft contact lenses. Lenses create microscopic corneal abrasions, thus facilitating the entry of Acanthamoeba. Soaking lenses in homemade saline solution and wearing lenses while swimming are thought to be important risk factors. Life Cycle and Morphology Acanthamoeba are small, filose FLA with a trophozoite and a cyst stage (Fig. 96–3A). The trophozoites are uninucleate and actively motile, with fine acanthopodia, and require a liquid or highly humid environment with plentiful bacteria for their subsistence. Under adverse conditions they enter into a cyst stage, and revert to trophozoites following the return of suitable conditions. Pathogenesis and Pathology Antecedent trauma is known to contribute to the onset and development of amebic keratitis. Presence of bacterial biofilms on contact lenses as a result of improper and incomplete decontamination is thought to increase risk of aquisition. Following the adhesion of the trophozoite to the corneal epithelium, the pathogenic strains are thought to produce a variety of proteases resulting in parasitemediated cytolysis of the cornea.40 Microscopic pathology shows evidence of granulomatous inflammatory reaction similar to GAE, with a mixed inflammatory infiltrate and presence of both trophozoites and cysts. Initially a dendriform keratitis appears that progresses to form a paracentral ring infiltrate of the corneal stroma, which can evolve to destroy the cornea completely.15,21 The corneal nerves are preferentially affected, demonstrating the neurotropic tendencies of the organisms.36,51

Clinical Disease Initial clinical features include a foreign body sensation with tearing, photophobia, and severe ocular pain. As the disease progresses, blurring of vision, blepharospasm, and conjunctival injection set in. The infection is often confused with bacterial, fungal, or herpetic keratitis. Distinguishing characteristics of Acanthamoeba keratitis are radial keratoneuritis, a ring infiltrate, pain disproportionate to clinical features, an elevated epithelial line, and relative lack of neovascularization in the setting of chronic anterior segment disease.36,51 The deep linear stromal infiltrates, which may be radial, are thought to be localized along the corneal nerves.51 This neurotropic tendency is thought to explain the extreme ocular pain as well as decreased corneal sensation often leading to a misdiagnosis of herpetic keratitis.

Angiostrongyliasis (Angiostrongylus cantonensis) Two species in the genus Angiostrongylus commonly cause infections in humans: A. cantonensis, a neurotropic parasite causing eosinophilic meningitis, and A. costaricensis, causing abdominal angiostrongyloidiasis. Angiostrongylus cantonensis, the rat lung worm, is a widespread zoonosis, being found in Africa, most of Southeast Asia, India, and the Pacific Basin, including Hawaii.29 This distribution coincided with the expansion of the territory of the main intermediate host, the giant African land snail Achatina fulica. Human infections may occur sporadically, or regularly in endemic areas and occasionally in epidemic outbreaks.38,64 Life Cycle and Transmission Adult worms of A. cantonensis live in the pulmonary arteries of rodents, where they lay eggs. The eggs hatch in the lungs and migrate up the trachea and down the alimentary tract to be excreted in the feces. These larvae undergo two further stages of development following ingestion by terrestrial or aquatic snails and slugs, which serve as intermediate hosts. After the ingestion of third-stage larvae by rodents, the nematodes migrate to the brain, where they mature into adults. They then reach the pulmonary arteries via the venous system. Humans are accidentally infected by the ingestion of improperly cooked intermediate hosts such as snails and slugs or by eating salad and other vegetables contaminated by snails and slugs or their slime, in which the larvae can be entrapped. In humans the parasites do not undergo full development and usually die in the brain, the spinal cord, or the eye chambers; few reach the lungs.6,64 Clinical Disease and Pathology The neuropathology of human angiostrongyliasis is characterized by congestion of the brain and inflammation of the

Parasitic Infections of the Peripheral Nervous System

A

B

D

C

E

F FIGURE 96–3 A, Double-walled cysts of Acanthamoeba species in the corneal stroma (hematoxylin-eosin; oil immersion). B, Trichinella spiralis: cross section of a larva in the gastrocnemius muscle (hematoxylin-eosin, ⫻100). C, and D, Leishmania species amastigotes in impression smear of a cutaneous lesion. E, Duodenal biopsy demonstrating numerous trophozoites of Giardia lamblia (hematoxylin-eosin; oil immersion). F, Cross section through a cysticercus of Taenia solium demonstrating the protoscolex and fluid-filled bladder. See Color Plate

2163

2164

Diseases of the Peripheral Nervous System

leptomeninges. The inflammatory reaction is provoked by the development and migration of the living parasites as well as by dead and degenerating ones in the neural tissue and subarachnoid space. Numerous tracks or cavities representing the migration of the parasites are often noted in the brain and the spinal cord.6,29 The main histologic findings are the marked cellular infiltration of the meninges with lymphocytes, plasma cells, macrophages, and eosinophils, and granulomas surrounding the dead worms. Clinical disease may be benign and self-limiting, or lead to a severe illness with neurologic sequelae and even death. The severity of the illness directly correlates with the infecting larval load. Angiostrongyliasis produces several clinical syndromes in humans: eosinophilic meningoencephalitis, eosinophilic myeloencephalitis, eosinophilic radiculomyeloencephalitis, and ocular angiostrongyliasis.6 Eosinophilic meningoencephalitis is characterized by headache, photophobia, nuchal rigidity, visual impairment, facial paresthesias, and paralysis. It is often accompanied by vertigo, loss of balance, and meningismus. Cerebrospinal fluid (CSF) and blood eosinophilia are often noted. Eosinophilic myeloencephalitis is often fatal, and symptomatology and eosinophilia are more pronounced in these cases compared to eosinophilic meningoencephalitis. Eosinophilic radiculomyeloencephalitis presents as exquisite pain and paresthesias of the lower extremities accompanied by sensory symptoms in the arms, trunk, and other locations. In an outbreak of angiostrongyliasis involving 16 Korean fisherman following consumption of raw African snails (Achatina fulica), radiculomyeloencephalitis with the invasion of brain, spinal cord, and spinal nerve roots was the main clinical syndrome.38 Their symptomatology related to motor and sensory abnormalities in the lower extremities with severe pains, weakness, absent reflexes, and bowel and bladder dysfunction indicative of spinal cord and nerve root involvement. Ocular angiostrongyliasis, found in up to 16% of cases, presents often as diplopia and visual field abnormalities. Fundus examination reveals either an abnormal fundus, a blurred disc, or papilledema. Optic atrophy may occur, and living parasites have been recovered from various eye chambers. Peripheral Neuropathy. In addition to the optic nerve, facial and abducens nerves are occasionally affected in angiostrongyliasis. Unilateral facial paralysis of the lower motor neuron type and lateral abducens paralysis occur in less than 5% of patients.6 Schmutzhard et al. reported a 40-year-old man who presented 2 days after consumption of Pila species snails with severe headache, nausea, mild neck stiffness, and diplopia.64 The most prominent feature was the palsy of both abducens nerves. He had both CSF and peripheral blood eosinophilia, and serology for Angiostrongylus antibodies was strongly positive. He improved on steroids, and follow-up

6 months later revealed only residual mild left abducens palsy.

Trichinosis (Trichinella spiralis) Transmission and Life Cycle Trichinosis is a disease caused by the parasites of the genus Trichinella, and most human infections are due to T. spiralis. The disease occurs sporadically worldwide, but its incidence in the United States is on a steady decline, with fewer than 100 cases reported each year. Humans acquire the infection by consumption of raw, rare, or incompletely cooked pork, pork products, or, less commonly, bear meat infected with cysts of Trichinella. After the larvae are released in the stomach, they mature in the tissues of the small intestine, and following fertilization the females discharge live larvae for several weeks. These larvae enter the circulation via the lymphatics and are capable of invading any tissue, but can only survive in the striated skeletal muscle tissue, where they undergo further development and encapsulation (see Fig. 96–3B). The cysts eventually calcify, leading to the death of the encysted larvae. Pathology and Clinical Disease The severity of infection depends mainly on the larval burden in the human muscle, with severe, life-threatening infections occurring in the presence of 1000 or more larvae per gram of muscle. Other factors that play a role in the disease outcome include number of ingested cysts and patient’s resistance, size, age, and general health. Symptoms are initially gastrointestinal as a result of the development of the worms in the small intestine, followed by a phase of allergic and inflammatory responses caused by muscle invasion. This is heralded by the occurrence of “trichinosis syndrome” with fever, myalgia, periorbital edema, and eosinophilia. The myositis progresses steadily, and difficulties in breathing, chewing, and swallowing may develop.54 Involvement of the myocardium may result in sudden death as a result of arrhythmias or acute congestive heart failure. Migration of the larvae through the central nervous system can cause intracerebral hemorrhage, meningitis, seizures, monoparesis, or psychotic disturbances.54 Peripheral Neuropathy. Macandrew and Davis described a case of a 39-year-old woman with trichinosis who presented with right footdrop and right facial palsy.43 She habitually ate lightly cooked pork, and had experienced symptoms of gastrointestinal disturbance 2 weeks before. Neurologic examination showed complete right footdrop, full sensory loss in the L5 and S1 distribution, diminished ankle and knee jerks, and flexor plantar reflexes accompanied by paresis of facial muscles below the eye on

Parasitic Infections of the Peripheral Nervous System

the right side. She had marked eosinophilia and subungual “splinter hemorrhages.” Her facial palsy recovered rapidly, and at a 9-week follow-up a mild residual right footdrop was present. Other cases of facial palsy caused by trichinosis have also been described.

PARASITES AFFECTING THE PERIPHERAL NERVES Leishmaniasis (Leishmania spp.) Leishmaniases are a diverse group of clinical syndromes caused by kinetoplastid protozoan parasites of the genus Leishmania. They are transmitted by the sandflies belonging to the genera Phlebotomus in the Old World and Lutzomyia in the New World, and species of Leishmania that infect humans usually have canine or rodent reservoirs. These species are difficult to distinguish morphologically and are usually classified based on their geographic origin, the clinical syndrome they produce, and ecologic characteristics.44 Life Cycle and Transmission Leishmania parasites have a relatively simple life cycle with amastigotes (aflagellar forms) in mammalian hosts and promastigotes (flagellar forms) in insect vectors. The amastigotes are obligatory intracellular parasites of phagocytic cells that multiply by binary fission (see Fig. 96–3C and D). Sandflies feeding on infected individuals ingest the parasitized cells during a blood meal. These amastigotes transform into promastigotes, which multiply in the gut and migrate to the proboscis of the insect vector. The promastigotes are introduced into the skin of a new vertebrate host during the next blood meal. They invade the reticuloendothelial system, transform into amastigotes, multiply within the phagolysosomes, and invade other reticuloendothelial cells, thus completing the cycle. Clinical Disease and Pathology The three best known clinical syndromes of leishmaniasis are visceral leishmaniasis, cutaneous leishmaniasis, and mucocutaneous leishmaniasis. The form and severity of the disease vary with the infecting species, immune status of the host, and previous exposure. Visceral Leishmaniasis (Kala-azar). Visceral leishmaniasis is most often caused by species of the L. donovani complex that occur in different parts of the world: L. donovani in the Indian subcontinent; L. infantum in the Mediterranean, Africa, Middle East, and China; and L. chagasi in South America. The worldwide incidence of visceral leishmaniasis has been estimated to be around 100,000 per year, with the Ganges river basin and southern

2165

Sudan being the two major endemic areas. 44 The epidemiology is characterized by the interplay between the sandflies, reservoir hosts, and susceptible humans. At the site of the bite of an infected sandfly, a granuloma develops that consists of histiocytes filled with amastigotes surrounded by epithelioid and giant cells. Parasites then spread to the local lymph nodes and are hematogenously carried to liver, spleen, and bone marrow, where they stimulate a granulomatous inflammation to result in either a subclinical infection or progression to the clinical syndrome of visceral leishmaniasis. Visceral leishmaniasis is a chronic disease characterized by fever, splenomegaly, hepatomegaly, lymphadenopathy, weight loss, pancytopenia, and hypergammaglobulinemia. Post–kala-azar dermal leishmaniasis refers to the skin lesions that sometimes appear after successful treatment of visceral leishmaniasis. It occurs in up to 20% of patients with visceral leishmaniasis in India and 2% in East Africa, and is rare in China. However, recently it has been reported in greater than 50% of Sudanese visceral leishmaniasis patients treated.44 In India it occurs 2 to 10 years following successful treatment of visceral leishmaniasis, while in East Africa it appears within a few months of treatment. Macules and papules occur around the mouth and later spread to the rest of the face, and may also occur on the arms, trunk, and legs. Pathologically the lesions can range from a granulomatous reaction with a few parasites to numerous parasites with rare inflammatory cells. Cutaneous Leishmaniasis. Cutaneous leishmaniasis is characterized by ulcerative skin lesions and occasionally nodular, psoriasiform, and verrucous lesions. Old world cutaneous leishmaniasis (oriental sore) is caused mainly by L. tropica, L. major, and L. aethiopica, and occasionally by L. donovani and L. infantum. It occurs in Iran, India, Afghanistan, the Middle East, south Russia, north and east Africa, and southern Europe. Cutaneous leishmaniasis of the New World, in contrast, is caused by the species complexes L. mexicana and L. braziliensis. It is mostly acquired in the forested areas of Mexico and Central America. Mucocutaneous Leishmaniasis (Espundia). Mucocutaneous leishmaniasis is primarily caused by L. braziliensis, which is present in Mexico and Central and South America. Typical cutaneous lesions are produced that are aggressive, last longer, and disseminate to mucous membranes of the nasal, oral, and pharyngeal areas. In these locations disfiguring lesions may occur as a result of erosion of the soft tissue and cartilage. Peripheral Neuropathy. A report from Sudan describes four patients who developed post–kala-azar dermal leishmaniasis and neuritis 1 to 6 months

B

D

C

E

F

G

H

Parasitic Infections of the Peripheral Nervous System

following successful treatment of kala-azar.16 Macules, papules, or nodules were present affecting the face, extremities, and trunk. Pathology showed evidence of neuritis affecting the cutaneous nerves in the lesion with lymphohistiocytic infiltration of the nerves. In all four patients leishmaniae were identified in the lesions, and in two patients they were also present in the nerves. Despite the pathologic evidence of neuritis, clinical examination showed intact sensation and motor power. In a review of 288 skin biopsy specimens of patients with cutaneous leishmaniasis caused by L. major, 14 specimens (5%) had evidence of nerve involvement.39 Ten of these showed perineural inflammatory cell infiltrates with either lymphocytes or a mixture of lymphocytes, plasma cells, and macrophages. Of the remaining four specimens, three had inflammatory cell invasion of the nerves and in the fourth severe granulomatous inflammation was noted along with evidence of nerve destruction. Amastigotes were seen within the nerves in two specimens. Clinical examination revealed diminished pain, touch, and temperature sensation in one patient and diminished pain sensation in another patient with intact touch and temperature sensations. In another report, a patient with cutaneous leishmaniasis was noted to have several tender hyperesthetic skin nodules.63 Pathologically, a lymphohistiocytic inflammatory infiltrate was noted around and within the cutaneous nerves along with the presence of amastigotes in the perineural space.

Giardiasis (Giardia lamblia) Giardiasis is an infection of the lumen of the small intestine with the cosmopolitan flagellate protozoan Giardia lamblia. It causes endemic and epidemic disease worldwide.22,79 It is most common in the tropics, where the prevalence rates range from 10% to 20%. In the United States it is problematic for travelers, campers, children in day-care centers, and homosexual men.77 Waterborne outbreaks can occur and have been reported in Colorado, in Rome, New York, and in numerous other locales.

2167

Transmission Infection occurs following ingestion of viable cysts of the parasite in food and water. Person-to-person spread can also occur between children. A variety of infected animal reservoirs, including beavers and dogs, may also be a source of infective cysts.79 Waterborne outbreaks occur as a result of defective water treatment plants with malfunctioning filters, because the protozoa are not killed by the usual concentrations of chlorine.21,79 Life Cycle Once ingested, the cysts undergo excystation to release two trophozoites. Trophozoites are most commonly located in the crypts within the duodenum, where they actively divide by longitudinal binary fission (see Fig. 96–3E). Infective cysts are formed as the trophozoites move down the intestine with the secretion of cyst wall, condensation of cytoplasm, and duplication of internal structures. Clinical Features Clinical presentation varies with age of the host, immunologic status, and previous exposure to the parasite. Infection may be asymptomatic, or it may cause disease ranging from mild diarrhea to a frank malabsorption syndrome. After an incubation period of 12 to 15 days, symptoms of acute infection occur with anorexia, nausea, and lassitude followed by a sudden onset of explosive, foul-smelling, watery diarrhea. Abdominal distention and flatulence are usual; weight loss occurs because of anorexia and diarrhea. The acute phase may be followed by a chronic phase characterized by continuous or recurrent, mild to moderate gastrointestinal symptoms with cramping, bloating, and bulky stools containing fat and undigested food.22,79 Weight loss in this group is often due to malabsorption. Musculoskeletal Manifestations. Reactive arthritis secondary to giardiasis was first described in 1977, and typically involves large joints of the lower limbs.41,42,75 The knee joint is most commonly affected, followed by the shoulder, elbows, and wrist. Small joints of the hands and feet are rarely affected. A total of 70 pediatric and 2 adult

FIGURE 96–4 A, Computed tomography scan showing a hydatid cyst in the liver. (Courtesy of Washington Winn, Jr., M.D., M.B.A.). B, Echinococcus granulosus protoscolices within a unilocular hydatid cyst adjacent to the germinal epithelium. C, Adult female worm of Dracunculus medinensis being extracted from a leg ulcer. (From Zaiman, H.: A Pictorial Presentation of Parasites. Available at parasitepics.biosci.uga.edu/) D, Onchocerca volvulus: cross section of adult worm in silk nodule (hematoxylin-eosin; ⫻100). E, Peripheral smear showing a trophozoite and “banana-shaped” gametocyte of Plasmodium falciparum. F, Peripheral smear showing a trophozoite and gametocyte of Plasmodium vivax. G, Pseudocyst of Toxoplasma gondii in brain tissue (hematoxylin-eosin; oil immersion). H, Toxoplasma gondii tachyzoites producing pneumonia replicating in the lung in a human immunodeficiency virus–positive individual. See Color Plate

2168

Diseases of the Peripheral Nervous System

cases have been reported to date. The arthritis typically is refractory to anti-inflammatory drugs, but responds to antimicrobial treatment of giardiasis. Peripheral Neuropathy. Bassett et al. reported two patients with intestinal giardiasis and peripheral neuropathy.2 The patients were 27- and 33-year-old men who presented with progressive numbness and paresthesias involving the feet (fingers were also involved in one). Nerve conduction studies were normal, but electromyography showed evidence of denervation of distal muscles. Denervation atrophy was present in the peroneus brevis. Evidence of steatorrhea was present in one patient. Histologic examination of the sural nerve revealed excess collagen fibers, consistent with chronic neuropathy. The fact that comprehensive investigation of the patients failed to identify an alternative cause of neuropathy, and that the symptoms resolved when giardiasis was treated, suggests a causal relationship between giardiasis and peripheral neuropathy. The authors stated that perhaps malabsorption, which is common in giardiasis, is a possible explanation.2

Cysticercosis (Taenia solium) Cysticercosis is the human infection with the larval stage of the pork tapeworm Taenia solium. It has worldwide distribution and is especially prevalent in Mexico, the rest of Latin America, Europe, Africa, India, and Asia. Prevalence varies depending on the type of animal husbandry, hygiene, and dietary practices. While meat inspection and sewage disposal prevent infection of pigs and propagation of the life cycle in industrialized countries, sporadic cases are still encountered as a result of immigration from, and travel to, endemic areas. In the United States the number of locally acquired cases has increased in recent years, and cysticercosis accounts for 2% of neurologic/neurosurgical admissions in Southern California.18,60 Life Cycle and Transmission The adult tapeworm, T. solium, infects the small bowel of humans and produces thousands of highly infectious eggs contained in gravid proglottids passed out in the stool. When these eggs are ingested by the usual porcine intermediate host, they hatch and release active onchospheres that penetrate the bowel wall and are carried by the blood stream to various tissues, where they develop into fluidfilled cysticerci (see Fig. 96–3F). When humans ingest the undercooked, cysticercotic (measly) pork, the liberated larvae attach to the walls of the jejunum and grow into adult tapeworms. Human cysticercosis is acquired via the fecaloral route, and when humans ingest tapeworm eggs, the eggs develop into cysticerci in a variety of tissues and organ locations. Humans serve as an accidental intermediate host and cannot transmit this stage of the parasite further.

Clinical Disease Cysticerci can infect any tissue, but clinically significant lesions are those occurring in the brain and spinal cord (neurocysticercosis), eye, muscles, and skin. Symptoms depend on the number, size, and location of the cysticerci as well as the inflammatory response around them. Neurocysticercosis may cause epilepsy, obstructive hydrocephalus, vasculitis, cerebral infarcts, or encephalitis. Severe posterior fossa involvement may lead to IVth nerve palsies, facial myokymia, upbeat nystagmus, periodic alternating nystagmus, and rhythmic oculopalatal myoclonus.34 Visual loss can occur as a result of retinitis, retinal detachment, and vasculitis damaging the optic nerve and optic chiasm or as a result of retrochiasmal lesions.7 Subcutaneous cysticerci are usually asymptomatic, but they may be associated with pain and tenderness as the cysticerci degenerate. Large numbers of cysticerci may infect the muscle, causing massive muscular pseudohypertrophy.18

Peripheral Neuropathy. Isolated cases of cysticerci causing peripheral neuropathy have been reported. A 26year-old Vietnamese woman presented with a history of a painful, enlarging nodule close to the bicipital groove in the arm associated with paresthesias along the median nerve distribution.55 Clinically the lesion was thought to be a schwannoma. Nerve stimulation demonstrated that the mass was in direct continuity with the median nerve. Surgery revealed a highly vascular mass that could not be dissected from the nerve and had to be removed piecemeal. Histology showed an abundance of inflammatory tissue surrounding a gelatinous material that was identified as the parasite Cysticercus cellulosae. In another case, a cysticercus was removed from the left median nerve of a woman from Holland.76 A male epileptic from Thailand had widespread subcutaneous nodules, and biopsy revealed cysticerci. He also complained of a lump on his right wrist, and as the lump enlarged, his little finger was pulled up.69 Examination revealed a nodule on the radial aspect of the pisiform bone, and the right little finger was “clasped.” Weakness and wasting of the muscles of the hypothenar eminence, the interossei, the two most medial lumbricals, and the adductor pollicis was noted. These changes were thought to be due to the compression of the deep branch of the ulnar nerve by cysticercus.

Coenuriasis (Taenia [Multiceps] spp.) Coenuriasis is a zoonotic disease caused by infection with the larval stage (coenurus) of Taenia (Multiceps) species. Adults inhabit the small intestine of canids, usually dogs. Ingestion of eggs by susceptible herbivore intermediate hosts while

Parasitic Infections of the Peripheral Nervous System

feeding results in the formation of coenuri in the subcutaneous muscles and central nervous system. Humans are accidental hosts. Each coenurus consists of a large (up to 10-cm) transparent sac containing numerous scolices that bud off from a germinal membrane and invaginate into the fluidfilled sac. Taxonomy of coenuri may reflect their anatomic location.14 Taenia (Multiceps) multiceps is widely distributed in temperate climates and circulates in a cycle between dogs and herbivorous animals such as sheep, goats, cattle, and horses. Coenuri infect the brain and spinal cord of the intermediate host, and human neurocoenuriasis has been reported from the United States, England, France, Africa, and Brazil. Coenuri of T. (Multiceps) serialis are also found in intermuscular connective tissue of rabbits and squirrels in temperate regions. It has been rarely reported from humans in Canada, the United States, France, and Africa. Taenia (Multiceps) brauni, found in eastern Africa, and T. (Multiceps) glomeratus, found in Nigeria, result in coenuri in the subcutaneous tissues of their intermediate hosts or accidentally in humans. Clinical Disease As in cysticercosis, coenuri may develop in any organ, producing similar disease. Symptoms of coenuriasis are due to a single cyst resulting in a space-occupying lesion. Neural coenuri have been reported from the cerebrum, ventricles, brain stem, and spinal cord and among the cranial nerves.14 Involvement of the CSF pathways results in arachnoiditis and ependymitis. Basilar leptomeningitis leading to arteritis and transient hemiparesis can occur. Subcutaneous coenuri are usually found on the trunk, in the intercostal region, or on the anterior abdominal wall but may occasionally be found on the neck, head, and limbs. Ocular coenuri can occur in the vitreous, anterior chamber, and conjunctiva. In a series of 14 cases of coenuriasis from Uganda, most occurred in young people and the majority of the cysts were solitary and unilocular.74 Ten cases presented as subcutaneous nodules and four involved the eye. The preoperative diagnosis was often lipoma, ganglion, or neurofibroma until the nodule was resected and the scolices within the cyst were visualized. In a recent review of the six reported cases of North American coenuriasis, three pediatric cases developed central nervous system disease and the three adults had soft tissue involvement.32 The disease prevalence was not localized to any one geographic area, and all cases had significant exposure to dogs.

Hydatidosis (Echinococcus granulosis) Hydatid disease is a zoonosis in humans and other mammals caused by the larval stage of the various tapeworms of the Echinococcus species, the adult stages of which are parasites in wild and domestic dogs and cats. Hydatidosis

2169

in humans may take one of three forms: unilocular hydatid disease caused by E. granulosus, multilocular or alveolar hydatid disease caused by E. multilocularis, and polycystic hydatid disease caused by E. vogeli and E. oligathrus.23 Echinococcus granulosus is the most important species producing human disease, and is common in many sheepand cattle-raising areas of the world, where dogs are the definitive host. Hydatid infection is widely distributed in South America and focally in North America, areas bordering the Mediterranean Sea, south and central Russia, central Asia, Australia, and Africa. In South America, it is a major public health concern in Uruguay, Argentina, Chile, southern Brazil, and Peru.52 Life Cycle and Transmission Hydatid disease is endemic in sheep-raising areas of the world, where dogs are used to care for the flocks. Adult E. granulosus are very small (a few millimeters in length) and inhabit the intestine of the definitive canine host (usually dogs). What they lack in size they make up for in numbers, with hundreds or thousands of worms producing large numbers of eggs. The intermediate hosts are usually farm animals (sheep, cattle, swine, or horses), which acquire the infection by ingestion of eggs in the pasture. Humans, especially children, are infected by accidental ingestion of eggs from the environment. The risk of human infection is highly influenced by the nature and extent of association with dogs used to herd sheep. The infectious embryonated eggs hatch in the intestine, and the onchospheres penetrate the intestinal wall and are transported by the blood stream to the final location, often the liver, where the unilocular hydatid cyst develops (Fig. 96–4A). The cysts are lined by germinal epithelium, develop slowly over a period of years, and contain brood capsules with protoscolices in the thousands, surrounded by clear fluid (see Fig. 96–4B). In endemic regions the prevalent practice of giving raw, infected viscera of slaughtered livestock to dogs maintains the zoonotic transmission of E. granulosus. Clinical Features and Pathology Signs and symptoms of the hydatid cyst are mainly mechanical, resulting from the pressure exerted by the growing cyst, and thus depend upon the organ involved and the size of the cyst. The onset of symptoms varies from months to several years, with infections of the central nervous system becoming apparent earlier than infections at other sites.21,52 Cysts are most commonly located in the liver (52% to 77%), followed by the lungs (9% to 44%).52 Other locations include the spleen, heart, brain, kidney, soft tissues, and bone. Hydatid cysts have, however, been reported in almost every organ. Complications of the cysts include infection with formation of abscesses, with subsequent rupture leading to anaphylactoid reaction and formation of a number of secondary cysts.

2170

Diseases of the Peripheral Nervous System

Peripheral Neuropathy. Rarely footdrop and peripheral neuropathy caused by hydatid cysts have been reported.28,30,56 A 23-year-old man presented with a 6-month history of persistent footdrop and progressive limping associated with radicular pain and numbness below the knee.30 Clinical findings included wasting of the muscles of the right leg, with weakness of the abductors and extensors of the hip and knee flexors, and paralysis of ankle dorsiflexors and plantar flexors. Ankle jerk was absent, and sensory loss was evident over the L4, L5, S1, and S2 dermatomes. Computed tomography (CT) scan of the pelvis showed a large globular cystic lesion on the right side of the pelvis with a well-defined calcified wall extending from the S1 neural foramen and causing erosion of the S2 foramen. On operative removal the diagnosis was established by histopathology. Postoperatively, full recovery of the proximal muscle weakness was noted within a year; however, the recovery was only partial in the distal muscles. Similarly, another case of footdrop associated with pelvic hydatid cyst was reported in a 24-year-old woman, but in this instance there was an associated palpable extrapelvic component resulting in a gluteal swelling.28 Surgery revealed a dumbbell-shaped retroperitoneal cystic mass occupying the hollow of the sacrum and extending through the greater sciatic notch to the gluteal region, causing stretching of the lower lumbar and sacral nerve roots adherent to the wall of the cyst. Similar to the first case, diagnosis was confirmed by histopathologic methods, and near-complete recovery was noted in the proximal muscles.28 In another report, the authors described a case of peripheral neuropathy in association with a partially ruptured hydatid cyst in the liver.56 A 57-year-old man presented with a febrile illness with recurrent exacerbations associated with weakness, numbness, tingling, and burning of hands and feet. Neurologic examination revealed severely reduced deep tendon reflexes in the legs and moderately reduced reflexes in the arms, as well as impaired vibration, position, and pain and temperature sensations distally. Electromyographic evaluation revealed severe neurogenic damage in the muscles of the upper and lower limbs. Attacks of fever were resistant to usual antibiotics and were associated with worsening of peripheral neuropathic deficits. Ultrasound and CT revealed a 6-cm calcific lesion in the left lobe of the liver, and a partially ruptured hydatid cyst was resected. Operative removal of the cyst was associated with relief from attacks of pyrexia, and clinical and electrophysiologic recovery of the neuropathy was noted. The authors thought that the association between the neuropathy and the hydatid cyst was more than incidental and the likely pathogenic mechanism was immunoallergic in nature.56

Dracunculiasis (Dracunculus medinensis) Dracunculiasis, or guinea worm disease, is a painful, incapacitating disease caused by the nematode parasite

Dracunculus medinensis. More than 120 million people in rural Africa and 20 million people in Asia are at risk of infection with this parasite.62 In Africa, guinea worm disease is endemic in 17 countries and does not occur south of the equator. In Asia, the disease is present in India and Pakistan and residual disease is also present in parts of the Middle East. Life Cycle and Transmission Human infection is acquired by drinking water containing the zooplankton copepod Cyclops species infected with third-stage larvae. The crustaceans are killed in the gastric juices, and the liberated larvae penetrate the intestinal wall and migrate into the abdominal and thoracic cavities, where they mature into adults. After mating, the female worm continues to mature and its uterus becomes distended with millions of rhabditiform larvae. The females then migrate to the site of emergence, usually at the feet or legs, where a blister forms and larvae are liberated by contracting the uterus on contact with water. These liberated larvae continue their maturation following ingestion by the copepod intermediate host. Clinical Disease Infected persons remain asymptomatic for 10 to 14 months until just before the worm emerges. Clinical manifestations include allergic reactions, blister formation, and abscesses caused by secondary bacterial infection. The initial sign is the appearance of a reddish papule, with a vesicular center and an indurated margin, that soon develops into a blister from which the worm emerges and discharges larvae into the water. Appearance of the blister is preceded by profound systemic symptoms of erythema, urticarial rash, intense pruritus, nausea, vomiting, and syncope that occur in 30% to 80% of individuals as a result of allergic reactions. While 95% of the lesions occur on the lower extremities (see Fig. 96–4C), they may occur anywhere on the body, including the hand, arm, trunk, buttocks, scrotum, knee joints, calf, thigh, shoulders, and even the angle of the jaw. A principal complication is secondary bacterial infection. Acute abscesses, cellulitis, arthritis, synovitis, epididymoorchitis, fibrous ankylosis of joints, and contractures of tendons can occur.62 Nonemergent or aberrant worms usually die and calcify but may also cause abscesses. Peripheral Neuropathy. Entrapment neuropathy of the median nerve caused by guinea worm infection resulting in carpel tunnel syndrome has been reported.2,47,49 A 55-yearold woman from Madras, India, presented with swelling in the right forearm just above the wrist for several months. This was associated with wasting of the thenar eminence and diminished sensation of the lateral half of the palm. Surgery revealed a partially abscessed, calcified guinea worm incorporated into and under the sheath of the median nerve. The worm had apparently entered the nerve

Parasitic Infections of the Peripheral Nervous System

sheath before degenerating and eventually caused thickening resulting in motor and sensory deficits.

Onchocerciasis (Onchocerca volvulus) Onchocerciasis is a leading cause of blindness in the world and is caused by an infection with the filarial nematode Onchocerca volvulus. Onchocerciasis is a major public health problem and is found in 37 Central African countries, northern parts of South America, Central America, and the Arabian Peninsula.12 It is estimated that 17.8 million people are infected, with large numbers suffering from visual impairment and blindness.12 It is also known as “river blindness” because the black fly vector (genus Simulium) only develops in the presence of abundant clean running water found in streams and rivers. Life Cycle and Transmission Adult female and male worms lie in fibrous tissue capsules in the dermis and subcutaneous tissues, forming subcutaneous nodules or onchocercomas. Gravid females produce microfilariae that invade the skin, subcutaneous tissues, and eyes (see Fig. 96–4D). The black fly vectors ingest the microfilariae during a blood meal; these develop into infective thirdstage larvae in 10 to 12 days. At subsequent blood meals infective larvae are transmitted to the definitive host. Clinical Disease and Pathology In endemic areas onchocerciasis can cause severe morbidity ranging from troublesome skin disease to visual impairment and blindness. Tissue pathology is primarily caused by severe inflammatory reactions resulting from the death and degeneration of migrating microfilariae, producing dermal, lymphatic, ocular, and systemic manifestations. Onchocercomas are innocuous and mainly cause discomfort as a result of mechanical effects. Acute onchodermatitis occurs in the form of macular, maculopapular, or urticarial rashes associated with pruritus. Chronic onchodermatitis may include hypertrophic, atrophic, hyperpigmented, and depigmented lesions. Lymph node changes occur in areas draining onchodermatitis.12 Ocular Changes. Microfilariae enter the eye through the bulbar conjunctiva, along the sheaths of the scleral vessels and nerves, or by embolization in the choroidal or ciliary capillaries. Ocular manifestations include punctate keratitis, conjunctivitis, limbitis, sclerosing keratitis, anterior uveitis, chorioretinal disease, glaucoma, and optic nerve disease. Onchocercal optic atrophy is either consecutive, postneuritic, or primary. Microfilariae may be found within the optic nerve and its sheaths, and their death may initiate an inflammatory response with atrophic changes.

2171

Peripheral Neuropathy. Onchocerciasis presenting as a “cold mass” in the flexor compartment of the forearm leading to symptoms of compression of the median nerve has been reported.67 A 36-year-old man presented with a 5week history of an enlarging, painless, semifluctuant mass on the volar aspect of his forearm with symptoms of ring finger tenosynovitis and median nerve compression. Surgery revealed a brown, rubbery, indurated mass densely adherent to tendon and muscle in the superficial and deep flexor compartments that extended from the musculotendinous junction to the flexor retinaculum distally, compressing the median nerve. Histopathology revealed an adult O. volvulus worm surrounded by a granulomatous reaction and eosinophilic infiltrate. Connor et al. described excessive mucin between the nerve and the nerve sheath in the biopsy specimens of 9 of 24 patients with severe onchocercal dermatitis.11 They also described the presence of onchocercal microfilariae throughout the node, capsule, and walls of vessels, within the scar tissue, and in the small nerves of a lymph node excised from an individual with advanced onchocercal lymphadenitis.27 The presence of microfilariae was thought to be attributable to chance migration rather than neurotropism, and reveals that the nerves are not protected from migrating microfilariae.

PARASITES ASSOCIATED WITH GUILLAIN-BARRÉ SYNDROME Malaria (Plasmodium spp.) Malaria is an acute and occasionally chronic blood stream infection caused by obligate intracellular protozoan parasites of the genus Plasmodium. While there are more than 100 species of plasmodia found in the blood of birds, reptiles, and mammals, only 4 species cause human malaria: P. falciparum, P. vivax, P. malariae, and P. ovale. Malaria is spread by female Anopheles mosquitoes and occurs mainly between 45 degrees north and 40 degrees south of the equator. Plasmodium falciparum infection occurs mainly in the tropical areas worldwide; P. vivax occurs in both tropical and temperate regions; P. malariae has worldwide distribution but occurs less frequently than either P. falciparum or P. vivax; and P. ovale is the least common, with cases being mainly acquired in western Africa, South America, and India. The estimated global incidence of malaria is 300 to 500 million cases per year.73 While malaria transmission has been interrupted in the United States, cases are imported, and on average about 1000 cases are reported to the U.S. Centers for Disease Control and Prevention annually.73 Life Cycle and Transmission Malaria is transmitted primarily by the bite of an infected female Anopheles mosquito. Rarely, malaria can be transmitted by direct inoculation of infected red cells, as in

2172

Diseases of the Peripheral Nervous System

transfusion-induced malaria, congenital malaria resulting from maternal-fetal transfusion, and malaria acquired by use of contaminated needles. The life cycle of the Plasmodium parasite involves a sexual reproductive cycle (sporogony) in the mosquito that results in production of sporozoites, and two cycles of asexual replication (schizogony) in the vertebrate host. Infectious sporozoites are injected into the vertebrate host when the mosquito takes a blood meal. These sporozoites enter hepatic parenchymal cells, where they multiply asexually to produce thousands of merozoites, a process called exoerythrocytic schizogony. These merozoites initiate the blood stream infection, invade red blood cells, and undergo a process of asexual multiplication (erythrocytic schizogony) to release numerous merozoites that invade other uninfected red cells. In cases of P. vivax and P. ovale, relapses may occur weeks to months after initial disease as a result of reactivation and development of latent hepatic sporozoites called hypnozoites. Some of the merozoites in erythrocytes differentiate into sexual forms, the gametocytes (macro- and microgametocyte) (see Fig. 96–4E and F). These gametocytes are ingested by the female Anopheles mosquito when feeding on an infected individual. The gametocytes mature, and fusion of the microgamete with the macrogamete results in the formation of a motile ookinete. This ookinete migrates through the stomach wall and forms an oocyst within which numerous sporozoites are formed that then migrate to the salivary glands to complete the life cycle. Clinical Disease and Pathology The clinical disease of malaria is a result of the effects of schizont rupture and destruction of parasitized and unparasitized red cells. It typically presents as febrile paroxysms, with abrupt onset of chills followed by fever ending in profuse sweating, all lasting for 9 to 10 hours. It may be accompanied by headache, nausea, vomiting, cough, dizziness, myalgia, fatigue, and gastrointestinal symptoms. Symptoms include fever, tachycardia, soft splenomegaly, and accompanying hepatomegaly. Plasmodium falciparum causes the most severe form of malaria, with a 25% fatality rate of nonimmune individuals within 2 weeks of the initial attack. It can result in parasitemia of up to 50% or more, which leads to severe hemolysis accompanied by hemoglobinuria and anemia. Erythrocytes infected with growing trophozoites and schizonts of P. falciparum become sequestered in the small vessels of the body, resulting in occlusion and leading to obstruction of blood flow and tissue anoxia. Multisystem organ involvement is common, and involvement of the brain results in cerebral malaria, in which the patient is initially disoriented but can rapidly progress to delirium, coma, and death. Peripheral Neuropathy. Twelve cases of GBS or demyelinating polyneuritis following malaria have been reported to date, nine following infection with

P. falciparum and three following P. vivax.33,68 Nine cases were reported from India and one each from Thailand, Portugal, and Sri Lanka; ages of the patients ranged from 15 to 48. The interval between the onset of fever and symptoms of GBS was between 5 and 21 days, with a mean of 15 days. Only one patient had cerebral malaria, and the remaining cases were uncomplicated. Five patients, all infected with P. falciparum, developed respiratory failure as a result of GBS. Three of these cases died and two survived, one treated with intravenous immunoglobulin and the other with plasma exchange. Patients infected with P. vivax had a milder course of illness with complete recovery. All patients had a predominant proximal weakness with features of demyelinating motor neuropathy on nerve conduction studies. While GBS could be the result of microvascular occlusion of parasitized red cells caused by P. falciparum, which resulted in demyelination, or the result of antimalarial drug–induced neuropathy, because of its occurrence in three patients infected with P. vivax and also its development in three patients (infected with P. falciparum) who had not received antimalarials prior to its onset, the authors believed that it was probably immunologically mediated.33 They postulated that GBS following malaria was due to an immune-mediated mechanism triggered by cytokines released by the asexual stage of the malarial parasite.68 Given the incomplete understanding of this process, the pathogenesis of GBS following malarial infection needs further investigation.

Toxoplasmosis (Toxoplasma gondii) Transmission and Life Cycle Toxoplasmosis, caused by the apicomplexan protozoan parasite Toxoplasma gondii, is a worldwide infection of humans and domestic and wild carnivorous animals.21 Infection is transmitted by ingestion of oocysts shed in the feces of the definitive feline hosts, by ingestion of tissue cysts in inadequately cooked meat (mainly pork or lamb), and by tachyzoites in the blood. Transplacental transmission occurs if the woman is infected during pregnancy, and transmission can also occur through organ transplantation. The life cycle includes a sexual stage occurring in the intestinal epithelium of cats and other felines, which are the definitive hosts and which acquire the infection by carnivorism. In this enteroepithelial cycle, asexual schizogony and gametogony occur, and the resulting immature oocysts are passed in the feces and mature in the environment. These mature oocysts containing four sporozoites are infectious to susceptible vertebrate hosts, in whom the actively growing tachyzoites will infect any nucleated cell. The proliferating tachyzoites result in cell death and tissue injury. Once immunity develops, tissue cysts form that contain thousands of slow-growing bradyzoites

Parasitic Infections of the Peripheral Nervous System

(see Fig. 96–4G). All stages of the life cycle occur in cats, but only tachyzoites and tissue cysts are found in humans and intermediate hosts. Clinical Disease Clinical disease of toxoplasmosis is usually asymptomatic or mild in immunocompetent individuals. It commonly presents as fever, cervical lymphadenopathy, headache, myalgias, and rash. Ocular toxoplasmosis presents as retinochoroiditis that may be associated with optic neuritis, iridocyclitis, and occasionally panuveitis. Infection in utero results in serious congenital infection with sequelae or may result in prematurity, stillbirth, or abortion. In immunosuppressed individuals toxoplasmosis results in central nervous system involvement with encephalitis, and occasionally pneumonia (see Fig. 96–4H), myocarditis, and retinochoroiditis. Peripheral Neuropathy. Two cases of GBS associated with toxoplasmosis in immunocompetent patients have been reported. In one report, a 26-year-old immunocompetent individual developed GBS during a disseminated infection with T. gondii.4 He became infected after eating undercooked warthog and doe, and initially developed symptoms of acute toxoplasmosis. His disease persisted with the development of diffuse lymphadenopathy and hepatosplenomegaly, and 5 weeks later he developed lower limb weakness and facial diplegia. Acute left retinochoroiditis was noted. CSF examination showed albuminocytologic dissociation. Acute Toxoplasma infection was confirmed serologically with the presence of positive immunoglobulin (Ig) G, IgM, IgA, and IgE antibodies and high titers of IgG. Intraperitoneal injection of blood into mice resulted in acute infection with death in 3 days and demonstration of the parasite in the ascites. The patient was treated with pyrimethamine and folinic acid and completely recovered after 10 months. In the second report, a 26-yearold man initially presented with symptoms of acute toxoplasmosis and a week later developed symmetrical distal paresthesias that progressed to ascending weakness, culminating in areflexic flaccid tetraplegia.5 He later developed diplopia, facial weakness, and difficulty in swallowing and required ventilation. Blood and CSF examination confirmed the diagnosis of toxoplasmosis, and he made a complete recovery in a month.

Cyclosporiasis (Cyclospora cayetanensis) Cyclosporiasis is a recently recognized intestinal infection caused by Cyclospora cayetanensis. It causes diarrheal disease in immunocompetent and immunocompromised individuals. The life cycle of this organism has yet to be fully investigated. Waterborne transmission occurs, and the organisms are resistant to chlorination. Food-borne transmission has also been documented with the finding of the organisms on lettuce; a multistate outbreak was

2173

linked to imported raspberries.9 Cyclospora cayetanensis typically produces symptoms of small bowel pathology with predominant nausea, anorexia, and eructation; malabsorption has been confirmed pathologically, with evidence of inflammation of the small bowel on biopsy. Clinical illness is characterized by an acute phase of fever, vomiting, and watery diarrhea followed by a phase of remissions and exacerbations of fatigue, anorexia, nausea, and diarrhea that lasts for 6 to 7 weeks. Weight loss is common. Peripheral Neuropathy. A report from Ohio described a 58-year-old man who initially developed fever and diarrhea and a week later developed weakness and paresthesias in his hands and feet that soon progressed to result in quadriparesis, areflexia, and respiratory paralysis.61 CSF analysis showed evidence of albuminocytologic dissociation. Nerve conduction studies were typical for GBS, with evidence of axonal loss with or without distal demyelination. He was treated with plasmapheresis and intravenous immunoglobulin and recovered with residual weakness of his hands bilaterally after 2 months. History revealed consumption of raspberries imported from Guatemala and followed the development of a similar diarrheal illness among 17 co-workers and family members. Cyclospora cayetanensis oocysts were recovered from the patient and eight other symptomatic individuals. The authors postulated that a Cyclospora-induced immune response resulted in severe GBS in the patient, and they advocated performing active searches for this agent, which may be missed by routine stool screening, in cases of GBS.

Tick Paralysis Mimicking Guillain-Barré Syndrome Ticks are obligate blood-sucking ectoparasites of the class Arachnida, which also includes mites, spiders, and scorpions. Ticks are important vectors of viral, bacterial, and parasitic pathogens to humans and domestic animals.70 Of the three families of ticks, two include species that transmit pathogens to humans: Ixodidae, or hard ticks, and Argasidae, or soft ticks. While they are notorious primarily for the disease-causing pathogens they transmit, they also cause illness by secreting noxious substances that cause paralysis, allergic reactions, or toxemia. More than 40 species of soft and hard ticks produce paralysis in humans and animals as a result of a salivary neurotoxin.70 Tick paralysis has been reported worldwide, and the majority of the cases are caused by Ixodes holocyclus in Australia, Ixodes rubicundus in South Africa, Dermacentor andersoni in the western United States, and Dermacentor variabilis in the eastern United States.19,70 Young females are typically affected, but among adults men are infected more

2174

Diseases of the Peripheral Nervous System

often than women. Cases occur often in spring and summer, when ticks are most active.

Clinical Disease and Pathology A neurotoxin produced by the tick’s salivary glands is responsible for the paralysis. The toxin is transmitted by an engorged, gravid female tick, and it impairs the conduction in the motor nerves by interference with the voltage-gated sodium channels of the axons.19 Symptoms begin 2 to 7 days after attachment of the tick, and the patient becomes restless and irritable, which may be accompanied by numbness and tingling of the extremities, lips, throat, and face. Symmetrical weakness of the lower extremities progresses rapidly to an ascending flaccid paralysis, which can reach the bulbar centers to cause dysphagia, slurred speech, and diplopia. Death may occur from respiratory failure or aspiration pneumonia, and mortality rates of 10% to 12% have been reported.70 Sensory function is usually spared and the sensorium is clear. Blood and CSF examinations are normal. Nerve conduction velocity and compound muscle action potentials are decreased.70 Provided the paralysis is not too far advanced, rapid and complete recovery follows the removal of the tick. Diagnosis depends on the clinical history and the recovery of the tick. It is advised that any patients presenting with symptoms suggestive of GBS and normal CSF protein levels should undergo a meticulous inspection for engorged ticks in occult locations such as the scalp, axilla, and perineum.19

REFERENCES 1. Avila, J. L., Rojas, M., and Avila, A.: Increase in asialoganglioside- and monosialoganglioside-reactive antibodies in chronic Chagas’ disease patients. Am. J. Trop. Med. Hyg. 58:338, 1998. 2. Balasubramanian, V., and Ramamurthi, B.: An unusual location of guinea worm infestation: report of a case. J. Neurosurg. 23:537, 1965. 3. Bassett, M. L., Danta, G., and Cook, T. A.: Giardiasis and peripheral neuropathy. Br. Med. J. 2:19, 1978. 4. Bossi, P., Caumes, E., Paris, L., et al.: Toxoplasma gondiiassociated Guillain-Barre syndrome in an immunocompetent patient. J. Clin. Microbiol. 36:3724, 1998. 5. Bouchez, B., Poirriez, J., Arnott, G. E., et al.: Acute polyradiculoneuritis during toxoplasmosis. J. Neurol. 231:347, 1985. 6. Bunnag, T.: Angiostrongyliasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 793, 2000. 7. Chang, G. Y., and Keane, J. R.: Visual loss in cysticercosis: analysis of 23 patients. Neurology 57:545, 2001. 8. Chimelli, L., and Scaravilli, F.: Trypanosomiasis. Brain Pathol. 7:599, 1997.

9. Connor, B. A., and Shlim, D. R.: Cyclosporiasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 600, 2000. 10. Connor, D. H., and Manz, H. J.: Parasitic infections of the peripheral nervous system. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1380, 1993. 11. Connor, D. H., Morrison, N. E., Kerdel-Vegas, F., et al.: Onchocerciasis: onchocercal dermatitis, lymphadenitis and elephantiasis in the Ubangi territory. Hum. Pathol. 1:553, 1970. 12. Cooper, P. J., and Nutman, T. B.: Onchocerciasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 756, 2000. 13. de Oliveira, R. B., Troncon, L. E., Dantas, R. O., and Menghelli, U. G.: Gastrointestinal manifestations of Chagas’ disease. Am. J. Gastroenterol. 93:884, 1998. 14. Diaz, J. F., and Gilman, R. H.: Coenuriasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 877, 2000. 15. Drabick, J. J.: Free-Living Amebic Infections. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 705, 2000. 16. Elhassan, A. M., Ali, M. S., Zijlstra, E., et al.: Post-kala-azar dermal leishmaniasis in the Sudan: peripheral neural involvement. Int. J. Dermatol. 31:400, 1992. 17. Engman, D. M., and Leon, J. S.: Pathogenesis of Chagas heart disease: role of autoimmunity. Acta Trop. 81:123, 2002. 18. Evans, C. A. W., Garcia, H. H., and Gilman, R. H.: Larval cestode infections: cysticercosis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 862, 2000. 19. Felz, M. W., Swift, T. R., and Hobbs, W.: Tick paralysis in the United States: a photographic review. Arch. Neurol. 57:1071, 2000. 20. Ferreira, M. S., Nishioka, S. A., Silvestre, M. T., et al.: Reactivation of Chagas’ disease in patients with AIDS: report of three new cases and review of literature. Clin. Infect. Dis. 25:1397, 1997. 21. Fritsche, T. R., and Smith, J. W.: Medical parasitology. In Henry, J. B. (ed.): Clinical Diagnosis and Management by Laboratory Methods. Philadelphia, W. B. Saunders, p. 1196, 2001. 22. Garcia, L. S.: Intestinal protozoa: flagellates and ciliates. In Garcia, L. S. (ed.): Diagnostic Medical Parasitology. Washington, DC, ASM Press, p. 36, 2001. 23. Garcia, L. S.: Tissue cestodes: larval forms. In Garcia, L. S. (ed.): Diagnostic Medical Parasitology. Washington, DC, ASM Press, p. 386, 2001. 24. Gea, S., Ordonez, P., Cerban, F., et al.: Chagas’ disease cardioneuropathy: association of anti-Trypanosoma cruzi and anti-sciatic nerve antibodies. Am. J. Trop. Med. Hyg. 49:581, 1993. 25. Genovese, O., Ballario, C., Storino, R., et al.: Clinical manifestations of peripheral nervous system involvement in human chronic Chagas disease. Arq. Neuropsiquiatr. 54:190, 1996.

Parasitic Infections of the Peripheral Nervous System 26. Ghoneim, M. A., Ashamallah, A., and El-Bolkainy, M. N.: Role of perineuritis in the atonic dilated bilharzial ureter. Int. Surg. 61:411,1976. 27. Gibson, D. W., and Connor, D. H.: Onchocercal lymphadenitis: clinicopathologic study of 34 patients. Trans. R. Soc. Trop. Med. Hyg. 72:137, 1978. 28. Gupta, A., Kakkar, A., Chadha, M., and Sathaye, C. B.: A primary intrapelvic hydatid cyst presenting with foot drop and a gluteal swelling. J. Bone Joint Surg. Br. 80:1037, 1998. 29. Gutierrez, Y.: Strongylida—hookworms, Oesophagostomum, Trichostrongylus, Angiostrongylus and others. In Gutierrez, Y. (ed.): Diagnostic Pathology of Parasitic Infections with Clinical Correlations. New York, Oxford University Press, p. 314, 2000. 30. Hassan, F. O., and Shannak, A.: Primary pelvic hydatid cyst: an unusual cause of sciatica and foot drop. Spine 26:230, 2001. 31. Herwaldt, B. L.: Laboratory-acquired parasite infections from accidental exposures. Clin. Microbiol. Rev. 14:659, 2001. 32. Ing, M. B., Schantz, P. M., and Turner, J. A.: Human coenurosis in North America: case reports and review. Clin. Infect. Dis. 27:519, 1998. 33. Kanjalkar, M., Karnad, D. R., Narayana, R. V., and Shah, P. U.: Guillain-Barre syndrome following malaria. J. Infect. 38:48, 1999. 34. Keane, J. R.: Cysticercosis: unusual neuro-ophthalmologic signs. J. Clin. Neuroophthalmol. 13:194, 1993. 35. Kierszenbaum, F.: Chagas’ disease and autoimmunity hypothesis. Clin. Microbiol. Rev. 12:210, 1999. 36. Kim, S. K., and Dohlman, C. H.: Causes of enlarged corneal nerves. Int. Ophthalmol. Clin. 41:13, 2001. 37. Kirchhoff, L. V.: American trypanosomiasis (Chagas’ disease). Gastroenterol. Clin. North Am. 25:517, 1996. 38. Kliks, M. M., Kroenke, K., and Hardman, J. M.: Eosinophilic radiculomyeloencephalitis: an angiostrongyliasis outbreak in American Samoa related to Achatina fulica snails. Am. J. Trop. Med. Hyg. 31:1114, 1982. 39. Kubba, R., el-Hassan, A. M., Al-Gindan, Y., et al.: Peripheral nerve involvement in cutaneous leishmaniasis (Old World). Int. J. Dermatol. 26:527, 1987. 40. Kumar, R., and Lloyd, D.: Recent advances in the treatment of Acanthamoeba keratitis. Clin. Infect. Dis. 35:434, 2002. 41. Layton, M. A., Dzeidzic, K., and Dawes, P. T.: Sacroiliitis in an HLA B27-negative patient following giardiasis. Br. J. Rheumatol. 37:581,1998. 42. Letts, M., Davidson, D., and Lalonde, F.: Synovitis secondary to giardiasis in children. Am. J. Orthop. 27:451,1998. 43. Macandrew, M., and Davis, E.: Trichiniasis presenting with foot-drop and facial palsy. Lancet 1:141, 1948. 44. Magill, A. J.: Leishmaniasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 665, 2000. 45. Magill, A. J., and Reed, S. G.: American trypanosomiasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 653, 2000. 46. Mahler-Araujo, M. B., and Chimelli, L.: Autonomic dysfunction in Chagas disease: lack of participation of the vagus nerve. Trans. R. Soc. Trop. Med. Hyg. 94:405, 2000. 47. Marin, G. A.: Calcified guinea worms simulating various orthopedic conditions. Int. Surg. 51:117, 1969.

2175

48. Marra, T. A.: Recurrent lumbosacral and brachial plexopathy associated with schistosomiasis. Arch. Neurol. 40:586,1983. 49. Mascola, J. R., and Rickman, L. S.: Infectious causes of carpal tunnel syndrome: case report and review. Rev. Infect. Dis. 13:911, 1991. 50. McLeod, J. G.: Autonomic dysfunction in peripheral nerve disease. J. Clin. Neurophysiol. 10:51, 1993. 51. Moore, M. B., McCulley, J. P., Kaufman, H. E., and Robin, J. B.: Radial keratoneuritis as a presenting sign in Acanthamoeba keratitis. Ophthalmology 93:1310, 1986. 52. Moro, P. L., Gonzalez, A. E., and Gilman, R. H.: Cystic hydatid disease. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 866, 2000. 53. Mostafa, M., Habib, M. A., Abdel-Moniem, S., et al.: Peripheral neuropathy in bilharziasis. J. Egypt. Med. Assoc. 55:44, 1972. 54. Murrell, K. D.: Trichinosis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 780, 2000. 55. Nosanchuk, J. S., Agostini, J. C., Georgi, M., and Posso, M.: Pork tapeworm of Cysticercus involving peripheral nerve. JAMA 244:2191, 1980. 56. Paolino, E., Granieri, E., and Tugnoli, V.: Peripheral neuropathy as a manifestation of hydatid disease: a case report. Acta Neurol. (Napoli) 9:256, 1987. 57. Prata, A.: Clinical and epidemiological aspects of Chagas’ disease. Lancet Infect. Dis. 1:92, 2001. 58. Purtilo, D. T., Meyers, W. M., and Connor, D. H.: Fatal strongyloidiasis in immunosuppressed patients. Am. J. Med. 56:488, 1974. 59. Rassi, A. Jr., Rassi, A., and Little, W. C.: Chagas’ heart disease. Clin. Cardiol. 23:883, 2000. 60. Richards, F. O. Jr., Schantz, P. M., Ruiz-Tiben, E., and Sorvillo, F. J.: Cysticercosis in Los Angeles County. JAMA 254:3444, 1985. 61. Richardson, R. F. Jr., Remler, B. F., Katirji, B., and Murad, H.: Guillain-Barre syndrome after Cyclospora infection. Muscle Nerve 21:669, 1998. 62. Sarwari, A. R.: Dracunculiasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 775, 2000. 63. Satti M. B., el-Hassan, A. M., al-Gindan, Y., et al.: Peripheral neural involvement in cutaneous leishmaniasis: a pathologic study of human and experimental animal lesions. Int. J. Dermatol. 28:243, 1989. 64. Schmutzhard, E., Boongird, P., and Vejjajiva, A.: Eosinophilic meningitis and radiculomyelitis in Thailand, caused by CNS invasion of Gnathostoma spinigerum and Angiostrongylus cantonensis. J. Neurol. Neurosurg. Psychiatry 51:80,1988. 65. Sica, R. E., Genovese, O. M., and Gargia Erro, M.: Peripheral motor nerve conduction studies in patients with chronic Chagas’ disease. Arq. Neuropsiquiatr. 49:405, 1991. 66. Sica, R. E., Gonzalez Cappa, S. M., Sanz, O. P., and Mirkin, G.: Peripheral nervous system involvement in human and experimental chronic American trypanosomiasis. Bull. Soc. Pathol. Exot. 88:156, 1995. 67. Simmons, E. H., Van Peteghem, K., and Trammell, T. R.: Onchocerciasis of the flexor compartment of the forearm: a case report. J. Hand Surg. [Am.] 5:502, 1980.

2176

Diseases of the Peripheral Nervous System

68. Sithinamsuwan, P., Sinsawaiwong, S., and Limapichart, K.: Guillain-Barre’s syndrome associated with Plasmodium falciparum malaria: role of plasma exchange. J. Med. Assoc. Thai. 84:1212, 2001. 69. Sorasuchart, A., Khunadorn, N., and Edmeads, J.: Parasitic diseases of the nervous system in Thailand. Can. Med. Assoc. J. 98:859, 1968. 70. Spach, D. H., Liles, W. C., Campbell, G. L., et al.: Tickborne diseases in the United States. N. Engl. J. Med. 329:936,1993. 71. Starr, M. D., Rojas, J. C., Zeledon, R., et al.: Chagas’ disease: risk factors for house infestation by Triatoma dimidiata, the major vector of Trypanosoma cruzi in Costa Rica. Am. J. Epidemiol. 133:740, 1991. 72. Strickland, G. T., and Ramirez, B. L.: Schistosomiasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 804, 2000.

73. Taylor, T. E., and Strickland, G. T.: Malaria. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 614, 2000. 74. Templeton, A. C.: Anatomical and geographical location of human coenurus infection. Trop. Geogr. Med. 23:105, 1971. 75. Tupchong, M., Simor, A., and Dewar, C.: Beaver fever—a rare cause of reactive arthritis. J. Rheumatol. 26:2701, 1999. 76. Vinken, P. J., and Bruyn, G. W. (eds.): Handbook of Clinical Neurology, Vol. 35: Infections of the Nervous System, Part III. Amsterdam, North–Holland, 1978. 77. Wolfe, M. S.: Giardiasis. Clin. Microbiol. Rev. 5:93, 1992. 78. Woodhouse, J. I.: The prevalence of clinical peripheral neuropathies in human chronic Chagas disease. J. R. Army Med. Corps. 139:54, 1993. 79. Wright, S. G.: Giardiasis. In Strickland, G. T. (ed.): Hunter’s Tropical Medicine and Emerging Infectious Diseases. Philadelphia, W. B. Saunders, p. 589, 2000.

N. Inflammatory and Immune Neuropathy

97 Polyneuropathies and Antibodies to Nerve Components ALAN PESTRONK AND GLENN LOPATE

Antibody Testing Methodology General Issues Antibodies with Binding to Gangliosides GM1 Ganglioside GQ1b Ganglioside GT1a Ganglioside GM2 and GalNAc-GD1a Gangliosides GM1b Ganglioside GD1a Ganglioside

GD1b Ganglioside (and Other Gangliosides with Internal Disialosyl Groups) Antibodies with Binding to Peripheral Nerve Glycoproteins and Glycolipids Myelin-Associated Glycoprotein Sulfated Glucuronyl Paragloboside Peripheral Myelin Protein 22 and Myelin Protein Zero Sulfatide

In clinical practice, immune-mediated polyneuropathies are an important focus of diagnostic evaluations. Immunomodulating treatments often lead to improvement in function and quality of life in patients with these disorders. Categorization of immune-mediated polyneuropathies involves defining clinical patterns of involvement and tissue targets of disease. Diagnostically useful features of clinical patterns include the involvement of different functional classes of cells or axons, the anatomic distribution of deficits, and the time course of the polyneuropathy.129,132,190 Targets of the pathologic processes in immune-mediated polyneuropathies include myelin, axons, cell bodies, blood vessels, perineurium, and subcellular constituents. 129,132,190 Electrodiagnostic and histopathologic testing may provide further help in defining tissue targets of disease. Delineation of specific polyneuropathies by these clinical and laboratory features can be useful because each syndrome often has a distinct spectrum of response to therapeutic agents.

GALOP Antigen (Sulfatide in a Lipid Membrane Environment) Galactocerebroside Other Serum Antibodies Trisulfated Heparin Disaccharide ␤-Tubulin Heparan Sulfate Neurofilaments

In this chapter we discuss one feature associated with immune neuropathy syndromes: serum autoantibodies with binding to specific neural antigens.77,189 These serum autoantibodies may have diagnostic utility, suggest specific avenues of therapy, and provide clues to the pathogenesis of neuropathy syndromes (Tables 97–1 through 97–3). The search for serum factors that are associated with immunemediated polyneuropathies includes testing both for monoclonal antibodies (M proteins) without identifiable antigenic targets,88 and for monoclonal and polyclonal autoantibodies that bind to specific neural components.

ANTIBODY TESTING Methodology Enzyme-linked immunosorbent assay (ELISA), Western blot, immunocytochemistry, thin-layer chromatography, and immunofixation electrophoresis are the methods 2177

2178

Diseases of the Peripheral Nervous System

Table 97–1. Acute and Subacute Neuropathies: Clinical Features and Associated Autoantibodies Neuropathy Syndrome Guillain-Barré syndrome (GBS) Acute motor axonal neuropathy (AMAN)

Miller Fisher syndrome (MFS)

Pharyngeal-cervicalbrachial variant Acute sensory neuropathy GBS with prodromal CMV infection Paraneoplastic sensory neuronopathy

Clinical Features

Electrophysiology

Serum Autoantibodies

Treatment

Motor ⬎ sensory Weakness: Proximal ⫹ Distal Symmetrical Motor only Weakness: Distal ⬎ proximal Asymmetrical Prodromal C. jejuni infection Ophthalmoplegia Ataxia Tendon reflexes: absent ⫾ Brainstem encephalitis Bulbar & facial palsy Neck weakness ⫾ Ophthalmoplegia Sensory loss: diffuse ⫾ Autonomic dysfunction Mild weakness Facial paresis Prodromal CMV infection Sensory loss: severe Proximal ⫹ Distal CNS: Ataxia Limbic encephalitis

Slow NCV Prolonged distal & F-wave latencies Conduction block Axonal loss

IgG or IgM vs. heparan sulfate IgG vs. PMP22 or MPZ IgG or IgM vs. GalC IgG or IgM vs. GM1, GalNAc-GD1a or GD1a ganglioside IgG vs. GM1b ganglioside

IVIG or PE

Absent SNAPs Mild evidence of demyelination

IgG vs. GQ1b ganglioside

IVIG or PE May be self-limited

Axonal loss Rare demyelination

IgG vs. GT1a ganglioside May cross-react with GQ1b IgG vs. GD1b (disialosyl) ganglioside

? IVIG

Absent SNAPs Occasional demyelination Demyelination

Absent or small SNAPs

IVIG

? IVIG

IgM vs. GM2

? IVIG

IgG vs. 35- to 40-kDa neuronal nuclear proteins (Hu; ANNA-1)

Chemotherapeutic treatment of cancer ? Immunomodulation

ANNA-1 ⫽ antineuronal nuclear antibody-1; CMV ⫽ cytomegalovirus; CNS ⫽ central nervous system; Ig ⫽ immunoglobulin; IVIG ⫽ intravenous immune globulin; MPZ ⫽ myelin protein p0; NCV ⫽ nerve conduction velocity; PE ⫽ plasma exchange; PMP22 ⫽ peripheral myelin protein 22; SNAPs ⫽ sensory nerve action potentials.

commonly used to detect antibodies associated with neuropathy syndromes. The specific technique used in measuring serum antibodies by each method can influence the sensitivity and specificity of testing.14,98,131,132,138 Enzyme-Linked Immunosorbent Assay ELISA methods determine the presence and amount of antibodies in a tissue fluid by testing for binding to antigens immobilized in a plastic well.12,13,26,169 ELISA involves three operations: (1) binding of an antibody to a target antigen that has been attached to a solid phase, (2) generation of a signal that can identify the presence of any bound antibody, and (3) detection of the signal. The first step in an ELISA is to attach a defined amount of a target antigen to small plastic wells. Methods of attachment include incubating an antigencontaining solution in wells for a period of time or evaporating the antigen from solution, leaving the dried antigen attached to the plastic well. Special methods of attaching antigens to the ELISA wells are necessary when aspects of the conformation or charge of an antigen are important, or if

only small amounts of a purified antigen are available.26 These special binding methods may include linking of the antigen to the well via covalent bonds or attaching monoclonal antibodies to the well before adding the antigen. After attaching the antigen to the well, a solution containing albumin or animal serum is used to block nonspecific binding. Generation of a possible signal begins by adding dilutions of patient sera, or other tissue fluids, to the wells for a period of hours to allow binding of any antibodies to the antigen in the well. After incubation, washing removes unbound serum antibodies and allows detection of remaining antibodies bound to the wells by a second antibody, which is specific for the species tested (e.g., anti-human) and has an attached color or fluorescent marker. A reading device that detects amounts of light measures any change in color or fluorescence. With appropriate dilutions, the amount of antibody in the serum correlates with levels of color that develop in the solution. Specific antibody binding is calculated by subtracting any serum antibody binding to parallel sets of control wells that typically contain either no antigen or, when more

2179

Polyneuropathies and Antibodies to Nerve Components

Table 97–2. Chronic Immune Demyelinating Neuropathies: Clinical Features and Associated Autoantibodies Neuropathy Syndrome Chronic immune demyelinating polyneuropathy (CIDP)

Multifocal CIDP

Multifocal motor neuropathy (MMN)

Anti-MAG neuropathy

Antisulfatide neuropathy

Serum Autoantibody

M Protein (Serum)* Treatment

Antibody targets: ␤-tubulin, heparan sulfate, MPZ, PMP22 Antibody class: IgM or IgG Frequency: rare Heparan sulfate: 20% ?

15%

T-cell suppression: Corticosteroids Cyclosporine A Methotrexate IVIG PE

?

T-cell suppression: Corticosteroids IVIG

Motor only: Conduction block Axonal loss EMG: no paraspinous denervation

Antibody target: GM1 ganglioside Antibody class: IgM Frequency: 30%–80%

20%

IVIG B-cell suppression: Cyclophosphamide Rituximab

Motor ⫹ sensory Prolonged distal latency Slow NCV No conduction block Axonal loss Motor ⫹ sensory Slow NCV Prolonged distal latency Axonal loss Motor ⫹ sensory Slow NCV Prolonged distal latency No conduction block Axonal loss Motor ⫹ sensory Slow NCV Prolonged distal latency Conduction block

Antibody target: MAG Antibody class: IgM Frequency: ~100%

85%

B-cell suppression: Cyclophosphamide Rituximab

Antibody target: sulfatide 90% Antibody class: IgM

IVIG B-cell suppression: Cyclophosphamide

Antibody target: sulfatide in lipid membrane Antibody class: IgM

80%

IVIG B-cell suppression: Cyclophosphamide

Antibody target: GM2 & GalNAc-GD1a Antibody class: IgM

Common

IVIG

Clinical Features

Electrophysiology

Motor ⬎ sensory Weakness: Proximal & distal Symmetrical Sensory loss: distal Onset: 1–80 yr Chronic or relapsing Motor ⬎ sensory Weakness: Distal ⬎ proximal Asymmetrical, arms ⬎ legs Sensory loss: multifocal Onset: 15–75 yr Motor only Weakness: Distal ⬎ proximal Asymmetrical, arms ⬎ legs Onset: 25–60 yr Slowly progressive Sensory ⬎ motor Distal, symmetrical Gait disorder & tremor Onset: ⬎50 yr Slowly progressive Sensory ⬎ motor Distal, symmetric Onset: ⬎45 yr

Motor ⫹ sensory Slow NCV Conduction block Prolonged distal latency Prolonged F waves

GALOP syndrome Gait disorder Sensory ⬎ motor Distal, symmetrical Tremor Onset: ⬎50 yr Anti-GM2 & Sensory ⬎ motor GalNAc-GD1a Distal, symmetrical neuropathy Limb & gait ataxia Onset: adult

Motor ⫹ sensory Slow NCV Conduction block Prolonged distal latency Prolonged F waves

*Frequency based on testing by immunofixation methodology. EMG ⫽ electromyography; Ig ⫽ immunoglobulin; IVIG ⫽ intravenous immune globulin; M protein ⫽ monoclonal protein; MAG ⫽ myelin-associated glycoprotein; MPZ ⫽ myelin protein p0; NCV ⫽ nerve conduction velocity; PE ⫽ plasma exchange; PMP22 ⫽ peripheral myelin protein 22.

specificity is desired for the assay, an antigen with some structural similarities to the test antigen. Optimally, an additional control antigen, such as histone H3, that may itself have no strong clinical correlations, will detect any confounding polyspecific antibodies that bind to a variety of antigens. ELISAs are especially useful for sensitive screening of large numbers of sera, when the target antigen is available in purified form, or if a degree of quantitation is desired. In the evaluation of polyneuropathies, it is common to test for serum antibodies to glycolipid and glycoprotein antigens by ELISA methods. Difficulties with the sensitivity

and specificity of ELISA assays12,13,26,169 arise when tertiary structural properties on the antigen are required for antibody binding. Impure antigens applied to ELISA wells also reduce specificity. In addition, problems occur when sera to be tested contain polyspecific antibodies that bind nonspecifically to a variety of substrates, sometimes even including the plastic wells of the ELISA plate. The sensitivity and specificity of ELISA assays can depend on seemingly minor methodologic details. In clinical practice, ELISA methods usually vary among clinical testing services. This requires that each laboratory provide

2180

Diseases of the Peripheral Nervous System

Table 97–3. Chronic Immune Axonal Neuropathies: Clinical Features and Associated Autoantibodies Neuropathy Syndrome Distal LMN syndrome Antisulfatide neuropathy

CANOMAD syndrome

Serum Autoantibody

M Protein (Serum)*

Axonal loss

IgM or IgG vs. GM1 or GalNAc-GD1a

Occasional

IVIG Rituximab

Small fiber sensory neuropathy Axonal loss

Antibody target: sulfatide Antibody class: IgM (polyclonal)

Occasional

IVIG

Axonal, sensory neuropathy Occasional demyelination

IgM vs. GD1b and other gangliosides with disialosyl residues

Common

IVIG Rituximab

Clinical Features

Electrophysiology

Motor Distal ⬎ proximal Asymmetrical Sensory ⬎ motor Painful Distal Symmetrical Sensory Ataxia Distal Ophthalmoplegia Bulbar palsy Cold agglutination

Treatment

*Frequency based on testing by immunofixation methodology. CANOMAD ⫽ chronic ataxic neuropathy with ophthalmoplegia, M protein, agglutination, and disialosyl antibodies; Ig ⫽ immunoglobulin; IVIG ⫽ intravenous immune globulin; LMN ⫽ lower motor neuron; M protein ⫽ monoclonal protein.

its own data regarding the clinical correlations, as well as the sensitivity, specificity, and reproducibility of their positive and negative test values. Western Blot Western blotting is typically used to evaluate the ability of antibodies to bind selectively to tissue proteins of known molecular weights.169 Western blot methodology involves (1) an initial separation of target tissue proteins on an electrophoretic gel according to their molecular weight, (2) transfer of the separated proteins to a strip of paper, and (3) exposure to diluted serum. If antibodies to one of the tissue proteins are present in the serum, they will bind to a band at the appropriate molecular weight region on the paper. A second antibody with an attached marker stains and allows visualization of the bound antibody from patient serum. The proteins on the paper that remain unstained provide a useful inherent control that confirms antibody specificity. Western blotting is useful for detection of antibodies to protein antigens, especially to confirm the specificity of antibody binding to protein antigens detected by other methods such as ELISA or immunohistochemistry. In the evaluation of polyneuropathies, Western blot methods are commonly used in confirming antibody binding to myelin-associated glycoprotein (MAG),138 tubulin,29 and the Hu family of antigens.32 Western blotting may be a less sensitive antibody testing method than ELISA. Further reduction in sensitivity is possible if there is variability in the amount of target antigen in tissue applied to the gel, or if the antigen on the paper deteriorates with storage. Another problem with Western blot assays is difficulty providing quantitative data. The specificity of Western blot results is usually high. Exceptions can occur when the presence of several protein antigens at similar molecular weight levels, or nonspecific antibody binding, causes false-positive test results.

Immunocytochemistry Immunocytochemistry detects patterns of serum antibody binding to specific cells, subcellular structures, or other tissue components.30,169,172 Methodology involves first choosing a tissue that contains the target antigen. The tissue is then fixed or frozen, cut, attached to slides, and then exposed to diluted serum or other tissue fluids. A second antibody system with an attached visible, or fluorescent, marker allows localization of any bound antibodies from patient serum. If antibodies to a tissue protein are present in the serum, they will bind to the tissue section in an anatomically appropriate pattern. Immunocytochemistry is an inexpensive and rapid procedure, making it especially useful as an initial qualitative screening test. Testing of nuclear antigens by immunocytochemistry is common.169 Laboratories often use immunocytochemistry, with modifications to improve specificity, to confirm the presence of paraneoplastic antibodies, including anti-Hu96 and Yo.159 Using serial serum dilutions can provide semiquantitative results. Immunocytochemistry is particularly useful when test antigens are difficult to purify in quantities needed for other testing methods. Testing by immunocytochemistry may identify antibody binding to conformation-specific epitopes not preserved with ELISA or Western blot methods. A major difficulty with immunocytochemistry can be poor specificity. Several different antigens can have similar distributions in tissue, causing antibodies to be mistaken for one another. Thus results determined using only this methodology may not be specific for a particular tissue antigen. Thin-Layer Chromatography Thin-layer chromatography67,71 examines antibody binding to lipid antigens separated from other lipids based on solubility. Lipid mixtures placed at the bottom of flat plates

Polyneuropathies and Antibodies to Nerve Components

coated with silica gel move up along with a solvent that carries the different lipids with it at varying rates to distinct distances from the origin. Lipids that are more soluble move further up the plate. If antibodies to one of the lipids are present in patient serum, they will bind to a band at the appropriate region on the plate. A second antibody with an attached marker stains the bound antibodies from patient serum. Other lipids on the plate that remain unstained provide an inherent control for nonspecific antibody binding. Thin-layer chromatography is useful to test for, or confirm, antibody binding to lipid antigens. It is most commonly used to test for antibody binding to lipids, such as sulfated glucuronyl paragloboside (SGPG),58 that are difficult to purify because they are present in very small amounts in tissues. In clinical practice, thin-layer chromatography is cumbersome to perform and can produce variable results. Immunofixation Electrophoresis Immunofixation electrophoresis172 is the most sensitive method for identifying and characterizing M proteins. The specimen to be tested can be serum, urine, or other body fluids. Immunofixation involves two steps. First, six aliquots of the test specimen are applied to an agarose gel. Electrophoresis separates the proteins in the specimen according to net charge. In the second fixation steps, five of the electrophoresis lanes are individually stained with antibodies to immunoglobulin (Ig) G, IgA, IgM, and ␬ and ␭ light chains, while the sixth lane has a protein fixative applied. The formation of antigen-antibody complexes in the first five lanes precipitates, and fixes, the proteins in the gel. A deproteination solution of NaCl washes the gel to remove nonprecipitated proteins. Finally, an amido black solution stains the antigen-antibody complexes and the fixed proteins in the sixth lane. The sensitivity of testing for M proteins is highly dependent on methodology. In neuropathies associated with monoclonal gammopathy of unknown significance (MGUS), the quantity of serum M proteins may be very small. Immunofixation, since it is the most sensitive method for detecting M proteins, is the preferred method in the laboratory evaluation of neuropathies. Other methods, such as serum protein electrophoresis, may not detect some M proteins.

General Issues Testing for specific autoantibodies in serum or cerebrospinal fluid can be useful in the identification, diagnosis, and treatment of immune-mediated neuropathies.124,147 Several general issues affect the utility of antibody testing in the evaluation of polyneuropathies. Cost-effectiveness Focused evaluation of relevant antibodies is most costeffective, since, in the clinical setting, serum antibodies are

2181

often associated with quite specific syndromes. With proper characterization of the clinical features of a neuropathy, the physician can usually choose specific antibodies to search for, and avoid large expensive preselected testing panels.191 Varying Prevalence The prevalence of antiganglioside antibodies may vary greatly among different geographic regions. This is especially common for antibodies associated with acute immune neuropathies. For example, IgG antiglycolipid antibodies are relatively common in patients with acute immune neuropathies in Japan, China, and other regions in the Third World, where they are associated with specific prodromal infections.53 Routine antibody testing in acute neuropathies may be more generally useful in those regions with high prevalence of a disorder. These acute neuropathy syndromes and related antibodies are rare in most parts of the United States. In such areas of low prevalence, antibody testing is less likely to yield positive results unless there is strong clinical or laboratory evidence of the presence of an antibody-associated syndrome. Clinical Validation of Methodology Although the general principles of antibody testing are straightforward, the assays are often difficult to carry out reproducibly in practice. The results of antibody testing differ among laboratories.14,98,131,132,138 This is manifest as prominent interlaboratory variation. The “gold standards” for antibody testing are sensitivity, specificity, and reproducibility. Sensitivity may be increased by efforts to present antigens in their native conformation.132 Specificity is increased by eliminating false-positive results that occur as a result of polyreactive antibodies180 that may be produced by immature B cells and do not correlate well with specific syndromes. Results should be correlated both with positive and negative control samples within the laboratory, and with clinical features of the specific patient syndromes associated with positive test results. Each clinical testing service performing antibody testing should provide data on clinical correlations for the results obtained in their laboratory. Laboratories that cannot supply such clinical correlations for their methodology should not perform antibody testing for clinical use.129 Pathogenic Significance While serum antibodies can be useful laboratory diagnostic markers for clinical syndromes, for many, their pathogenic significance remains to be determined. It has been especially difficult to show any pathologic effects of antibodies directed against intracellular antigens, such as Hu.145 There is good evidence that IgM anti-MAG antibodies may play a pathogenic role in the development of the associated neuropathy.147 Some antibodies may be the result of antigenic cross-reactivity with a prodromal infectious agent. The best

2182

Diseases of the Peripheral Nervous System

documented example is the IgG binding to GM1 ganglioside that occurs in patients with acute motor axonal neuropathies that are associated with prior Campylobacter jejuni infection (see below). However, the precise pathogenic role of these antibodies in producing the associated neuropathy syndromes remains to be firmly established.

ANTIBODIES WITH BINDING TO GANGLIOSIDES Gangliosides are glycosphingolipids that contain sialic acid and are present in large amounts in the nervous system, especially on myelin (see Fig. 97–1). Gangliosides may act as signaling molecules, receptors, or receptor modulators and to effect neuronal differentiation.47 They can also be antigenic targets in a variety of autoimmune neuropathies, as discussed in detail below and in review articles.77,189

GM1 Ganglioside Antibody binding to GM1 ganglioside is associated with both acute and chronic motor neuropathies.78,133 IgM Anti-GM1 Antibodies High titers of serum IgM binding to GM1 ganglioside occur in a majority of patients with multifocal motor neuropathy (MMN),133 a motor neuropathy syndrome characterized by asymmetrical, progressive weakness that most commonly begins in the arms.112,128,130 Men develop MMN more often than women, and onset age ranges from 20 to 70 years, with the mean in the fifth decade. Progression of weakness develops slowly over a period of years. Disability is common as a result of the loss of dexterity in the hands and gait impairment from footdrop. Cramps and fasciculations occur in over 50% of patients, but sensory signs or symptoms are rare. On electrophysiologic testing, motor conduction block— a partial and focal reduction in compound muscle action potential (CMAP) amplitude at locations other than sites of entrapment—is present in most patients with MMN. The degree of reduction in CMAP amplitude that is defined as significant varies among authors, but the greatest specificity is achieved when reductions of at least 50% are found. Consensus criteria for the diagnosis of conduction block have been developed.125 Some patients with a typical MMN phenotype and anti-GM1 antibodies have no conduction block. Explanations may include conduction block that occurs in untested nerve segments, or disease resulting from motor axon loss without conduction block.68,176 Antibody Testing. The sensitivity and specificity of ELISA testing for IgM anti-GM1 antibodies in MMN has

varied prominently among laboratories.14,98,131,133,138 Testing for IgM binding to GM1 ganglioside molecules covalently linked to ELISA wells (Co-GM1) has yielded the highest reported sensitivity in MMN.133 Presumably, this method presents GM1 ganglioside in a configuration more easily recognized by serum antibodies than the usual methods of linking GM1 to plastic wells. Using this methodology, high titers of anti–Co-GM1 antibodies (titers ⬎ 1800) are detected in 73% to 80% of patients with MMN and conduction block. With Co-GM1 methodology, the rate of false-positive test results (in sera from disease control patients without motor neuropathies) is less than 0.3%. In routine clinical practice, positive test results for IgM binding to Co-GM1 have approximately 97% specificity for MMN or other motor neuropathies. IgM anti-GM1 antibodies occur only rarely in amyotrophic lateral sclerosis (ALS) or chronic inflammatory demyelinating polyneuropathy (CIDP) (see Table 97–2). Serum antibody binding to other gangliosides, including GM1b and N-acetylgalactosaminyl (GalNAc)-GD1a, has also been detected.83,89 IgM anti-GM1 antibodies can occur in about 30% of patients with acute motor neuropathy syndromes,136 but an association between acute motor neuropathies and serum IgG anti-GM1 antibodies is more clearly established (see below). Small variations in ELISA methodology can greatly alter the sensitivity and specificity of anti-GM1 antibody testing.14,98,131,133,138 Clinical testing for IgM anti-GM1 antibodies by ELISA is indicated in slowly progressive motor syndromes. Several features increase the likelihood of positive anti-GM1 testing, including patient age between 20 and 70 years of age, asymmetrical weakness beginning in the upper extremities, no upper motor neuron signs, selective motor conduction block detected on electrophysiologic testing, and IgM M protein in the serum.132 Treatment of Motor Neuropathies Based on Antibody Testing. Testing for IgM anti-GM1 antibodies is appropriate in the evaluation of motor system disorders. The presence of IgM anti-GM1 antibodies in the serum suggests that a motor neuropathy is immune mediated and treatable, and aids in distinguishing the syndrome from CIDP and ALS.109,133 Treatment with rituximab, cyclophosphamide, or intravenous immune globulin (IVIG) may produce functional benefit.21,38,130,137 The presence of serum IgM binding to GM1 ganglioside increases the likelihood that a motor neuropathy syndrome will respond to IVIG treatment,176 but IVIG treatment does not generally produce a change in antiGM1 antibody titers.115 Initial response to IVIG is usually rapid and favorable, but long-term treatment with repeated infusions is often necessary. While strength may be improved, ongoing axon loss may not be prevented.175 The beneficial effects of IVIG treatment may decline over time.

2183

Polyneuropathies and Antibodies to Nerve Components

Structure β(1-3)

β(1-4)

Ganglioside β(1-4)

β(1-1)

CER

GM1

CER

GM1b

CER

GM2

CER

GalNAc-GD1a

CER

GD1a

CER

GD1b

CER

GT1a

CER

GQ1b

α(2-3)

β(1-3)

β(1-4)

β(1-4)

β(1-1)

β(1-4)

β(1-4)

β(1-1)

α(2-3)

α(2-3)

β(1-4)

β(1-3)

β(1-4)

α(2-3)

β(1-3)

β(1-4)

α(2-8)

β(1-4)

α(2-8)

Legend

β(1-4)

α(2-3)

β(1-3) α(2-3)

= galactose;

β(1-4)

β(1-4)

β(1-1)

α(2-3)

β(1-4)

β(1-1)

α(2-3)

β(1-4)

α(2-8)

β(1-4)

β(1-1)

α(2-3)

= N-acetylgalactosamine;

= N-acetylneuraminic acid; CER = ceramide FIGURE 97–1 Structures of selected gangliosides.

β(1-1)

α(2-3)

α(2-8)

β(1-3)

β(1-1)

α(2-3)

α(2-3)

β(1-3)

β(1-4)

= glucose;

2184

Diseases of the Peripheral Nervous System

Anti-GM1 antibody testing may also be useful to determine sufficient treatment of MMN using cyclophosphamide.140 Benefit is only likely with reduction in antibody titers of 60% or more, and is unlikely if antibody titers remain unchanged. After treatment is completed, increasing antibody titers suggest the possibility that a relapse may begin to occur during the next several months. Corticosteroids and T-cell immunosuppressive treatments are rarely associated with improvement. Prednisone has produced exacerbations in weakness.130 Rituximab, an anti–B-cell monoclonal antibody, appears to be of therapeutic benefit. Some patients show progressive improvement in strength, and reduction in serum antibody titers, over 2 years of treatment.90,137 Association with Disease Pathogenesis. There is evidence that anti-GM1 antibodies may have pathogenic effects in MMN. GM1 ganglioside is present on both motor and sensory nerves,117,160 but a higher concentration in motor nerves has been reported.118 Using a variety of techniques, GM1 or its terminal disaccharide Gal(␤1-3)GalNAc has been localized to nodes of Ranvier, compact or outer myelin, the postsynaptic neuromuscular junction, and the axolemma.112 In vitro experiments have shown that antiGM1 antibodies can alter sodium and potassium currents at the node of Ranvier52,167 and can inhibit distal conduction at motor nerve terminals.152 The antibodies may alter ion channel function at these sites. In vivo studies include immunization of rabbits with GM1 ganglioside to cause a subclinical neuropathy with conduction block, mild axonal degeneration, and IgM deposits at nodes of Ranvier.170 Intraneural injection of anti-GM1 antibody–containing serum has produced conduction block in some, but not all studies.127,173 IgG Anti-GM1 Antibodies High titers of serum IgG binding to GM1 ganglioside are most strongly associated with acute motor axonal neuropathy (AMAN).55,78,147 This acute motor neuropathy syndrome has several features that distinguish it from classic Guillain-Barré syndrome (GBS), including rapid onset with mean nadir at 6 days, weakness that is most prominent distally, a lack of sensory signs, and electrodiagnostic evidence of axonal loss with few, or no, features of demyelination (see Table 97–1). AMAN is commonly associated with preceding C. jejuni, and less commonly Haemophilus influenzae, infection.106 It occurs most frequently in China, Japan, and Third World countries and is uncommon in most areas of the United States. Antibody Testing. We recommend clinical testing for IgG anti-GM1 antibodies by ELISA in acute motor syndromes. Testing is more likely to be positive in areas where AMAN

is more prevalent. There is an increased likelihood of positive test results with specific clinical features, such as distal weakness, no sensory signs or symptoms, and prodromal diarrheal illness.132 An association with acute motor neuropathies has also been found with IgG binding to GalNAc-GD1a, GM1b, and GD1a gangliosides (see below).4,49,195 Treatment of Motor Neuropathies and Antibody Testing. In acutely developing neuropathies, the finding of serum IgG anti-GM1 antibodies suggests that the weakness is due to a motor neuropathy and not to “classic” motor-sensory GBS. Two patterns of recovery, one rapid and one more prolonged, have been described in patients with AMAN and anti-GM1 antibodies.5,74 Patients with IgG subclass 1 antibodies more often have preceding C. jejuni gastroenteritis and slow recovery. Patients with IgG subclass 3 antibodies often have preceding respiratory infection and a more rapid recovery.74 Treatment with IVIG may be more effective than plasma exchange (PE) in acute motor neuropathies with IgG anti-GM1 antibodies.86 In contrast, PE and IVIG are similarly effective in classic GBS.177 It may be difficult, in the setting of an acutely evolving neuropathy, to obtain results of antibody testing rapidly enough to include them in therapeutic decision making. Association with Disease Pathogenesis. IgG anti-GM1 antibodies may play a role in the pathogenesis of the AMAN. Rabbits immunized with bovine brain gangliosides or GM1 developed a monophasic illness with flaccid paralysis resulting from an axonal neuropathy, high serum anti-GM1 antibody titers, IgG deposited on the axonal membrane, and periaxonal macrophages, all characteristics of AMAN.163,198 IgG anti-GM1 antibodies from patients with acute neuropathy syndromes can induce leukocyte degranulation and macrophage phagocytosis of GM1coated beads, presumably through interaction with the IgG receptor.178 In some, but not all models, IgG antibodies to GM1 antibodies or C. jejuni lipopolysaccharides can cause acute conduction block.33,56,181 Serum with anti-GM1 antibodies applied to a preparation of endoneurial microvascular endothelial cells increased the permeability of the membrane, suggesting these antibodies may be important in the disruption of the blood-nerve barrier in AMAN.66 There is evidence that IgG anti-GM1 antibodies may link C. jejuni infection with the precipitation of autoimmune neuropathy.157 Rabbits immunized with specific strains of C. jejuni lipopolysaccharides produce antiganglioside antibodies that could be predicted based on the structure of the lipopolysaccharide.1,2 Studies of the lipopolysaccharide coat of C. jejuni have shown the presence of ganglioside-like structures, and specific serotypes (e.g., HS:19) are more commonly isolated from patients

Polyneuropathies and Antibodies to Nerve Components

with AMAN compared to other C. jejuni serotypes.110 Taken together, these studies suggest that molecular mimicry between the C. jejuni lipopolysaccharide coat and peripheral nerve gangliosides may be responsible for the generation of IgG that binds to GM1 ganglioside.

GQ1b Ganglioside IgG antibodies that bind selectively to GQ1b were initially found in Japanese patients with the Miller Fisher syndrome (MFS).24 Since then a spectrum of clinical disorders with common features of acute-onset ataxia or ophthalmoplegia have been associated with serum IgG binding to GQ1b ganglioside. Associated syndromes include MFS, Bickerstaff’s brainstem encephalitis (BBE), GBS-like syndromes with ataxia and ophthalmoplegia, and selective, acute-onset ophthalmoplegia or ataxia.17,80,116,185 These syndromes are most commonly described in Japan and are unusual in most of the United States. The prototypical MFS, in which IgG anti-GQ1b antibodies are found in up to 90% of patients, is characterized by ataxia, ophthalmoplegia, and areflexia. Cases have also been described with weakness and sensory loss suggesting overlap with GBS, and with drowsiness as is seen in BBE.116 The presence of serum IgG anti-GQ1b IgG antibodies may suggest that immunomodulating treatment of associated clinical syndromes could be beneficial.116 Binding of IgG anti-GQ1b antibodies is presumably targeted, at least in part, to the disialosyl group (Neu5Ac[␣2-8]Neu5Ac) attached to the external sugar, since most anti-GQ1b antibodies cross-react with GT1a ganglioside, which also contains an external disialosyl residue. Some serum IgG (50%) cross-reacts with other gangliosides with disialosyl residues, including GD3 and GD1b.184

2185

terminal may be additional pathophysiologic effects of anti-GQ1b antibodies.10,121 Serum IgG anti-GQ1b antibodies bind to peripheral nerve and brainstem structures.22,79 GQ1b ganglioside was found to be in high concentrations in cranial nerves III, IV, and VI.23 The presence of IgG anti-GQ1b antibodies may reflect prodromal infections with C. jejuni or H. influenzae. GQ1blike epitopes are present in C. jejuni isolated from patients with MFS and anti-GQ1b antibodies,1 and specific C. jejuni serotypes (O:2 and O:10) are more frequent in patients with MFS.197 Serum IgG from H. influenzae–seropositive patients with anti-GQ1b antibodies binds to the lipopolysaccharide fraction from H. influenzae, suggesting that the lipopolysaccharide has ganglioside-like moieties.75 As with C. jejuni infection and GM1 antibodies in AMAN, C. jejuni or H. influenzae infection in MFS may cause the production of antibodies against their lipopolysaccharide coat, with subsequent immune attack on structurally similar gangliosides in peripheral nerve.157

GT1a Ganglioside IgG binding to GT1a ganglioside is associated with a spectrum of acute neuropathies with bulbar dysfunction (pharyngeal-cervical-brachial variants).73,103,193 There is often cross-reactivity to GQ1b ganglioside. The difference between the two gangliosides is that GT1a has a monosialosyl residue attached to its internal galactose, whereas GQ1b has a disialosyl residue. In a recent large series of acute immune neuropathy patients from Japan, 10% had anti-GT1a antibodies.72 Many patients with binding to both GT1a and GQ1b had ophthalmoplegia, facial palsy, bulbar palsy, and neck weakness, findings similar to those in other patients with anti-GQ1b antibodies. IgG anti-GT1a binding alone may predict a subgroup of patients with acute immune neuropathies that less commonly have ophthalmoplegia. Anti-GT1a antibodies with no binding to GQ1b are associated with syndromes including bulbar palsy, neck weakness, normal sensation, and recent C. jejuni infection. Ophthalmoplegia is unusual (20%).

Antibody Testing and Treatment The authors have found that testing for serum IgG binding to GQ1b ganglioside is most commonly useful in the setting of a recent-onset brainstem or cranial nerve syndrome with clinical features that may suggest either intrinsic or extrinsic disease. The finding of IgG binding to GQ1b ganglioside suggests that a syndrome with brainstem signs is immune mediated.116 These syndromes may respond to treatment with PE or IVIG, but usually have a monophasic course with spontaneous recovery. Testing is most often positive early in the course of these syndromes. Serum antibodies may disappear within 2 to 4 weeks after onset.79

Antibody Testing Testing for serum IgG binding to GT1a ganglioside, like binding to GQ1b, is most commonly positive in the setting of a recent brainstem or cranial nerve syndrome, especially in patients with oropharyngeal palsy. The presence of IgG binding to GT1a may suggest the presence of an immunemediated syndrome that responds to IVIG.72

Association with Disease Pathogenesis Antibodies that bind selectively to GQ1b ganglioside can cause a complement-dependent increase in neurotransmitter quantal release and subsequent presynaptic blockade of motor nerve terminals.11,143,151 Postsynaptic blockade of acetylcholine receptors and destruction of the motor

Association with Disease Pathogenesis Few data are available on the pathogenicity of anti-GT1a antibodies. Serum IgG from a patient with anti-GT1a antibodies stained human oculomotor, glossopharyngeal, and vagal nerves, suggesting that this ganglioside is present in these structures.72

2186

Diseases of the Peripheral Nervous System

GM2 and GalNAc-GD1a Gangliosides Polyclonal antibody binding to GM2 and GalNAc-GD1a gangliosides is associated with two different GBS variants. The specific association depends partly on the class of antibody. IgG anti–GalNAc-GD1a antibodies with cross-reactivity to GM2 are associated with motor neuropathy syndromes.49,61,195 Patients often have an acute motor neuropathy, similar to AMAN, with distal-predominant weakness, no sensory signs or symptoms, sparing of the cranial nerves, and evidence of antecedent C. jejuni infection. Electrodiagnostic features of axonal dysfunction and early reversible conduction failure may occur.61,120 IgG anti–GalNAc-GD1a antibodies without cross-reactivity to GM2 have also been described, usually in patients with chronic axonal motor neuropathies.64 IgM anti–GalNAc-GD1a antibodies with cross-reactivity to GM2 have been described in a GBS variant syndrome with mild weakness, facial paresis, and a good prognosis.61,62 Prodromal cytomegalovirus infections are common in this group of patients.59,179 IgM anti–GalNAc-GD1a antibodies have rarely been associated with a pure motor variant of GBS, especially for antibodies with no cross-reactivity with GM2 ganglioside.62 Monoclonal IgM antibodies with binding to both GM2 and GalNAc-GD1a gangliosides are associated with a chronic progressive demyelinating neuropathy.91,122 The neuropathy can be sensory ⬎ motor, with panmodal distal sensory loss, minimal weakness, and an ataxic gait, although a syndrome of chronic demyelinating motor ⬎ sensory neuropathy has also been described.126 Antibody Testing and Treatment Clinical testing for IgG or IgM binding to GM2 and GalNAc-GD1a gangliosides may be useful in the evaluation of acute and chronic neuropathy syndromes. In acute neuropathy syndromes, the identification of IgG or IgM anti-GM2 or anti–GalNAc-GD1a antibodies may distinguish the disorder from classical motor-sensory GBS. In chronic motor neuropathy syndromes, the identification of IgG or IgM anti–GalNAc-GD1a antibodies may be the only evidence that the disorder is immune mediated and treatable.64 When a monoclonal protein is present in the serum, the finding of selective IgM binding to GM2 and GalNAc-GD1a gangliosides distinguishes a chronic demyelinating neuropathy syndrome from CIDP and MAG neuropathies. Treatment with IVIG in some patients results in clear functional improvement, while prednisone and immunosuppressive medications may not be beneficial.91 Association with Disease Pathogenesis IgM anti-GM2 antibodies have been found to induce complement-mediated lysis of neuroblastoma and motor neuron cell lines.19,122 Serum from a patient with IgM

monoclonal anti–GM2/GalNAc-GD1a antibodies inhibited neurotransmitter release from normal neuromuscular junctions, suggesting a possible role in the production of weakness.126 Anti–GalNAc-GD1a antibodies can bind to ventral roots and intramuscular nerves where GalNAcGD1a is localized.63 The epitope for antibody binding by GalNAc-GD1a and GM2 is likely the terminal GalNAc(␤1-4) Gal(␣2-3)NeuAc trisaccharide moiety present on both GM2 and GalNAc-GD1a gangliosides.63,91,126 This epitope is also present internally on other gangliosides that have been associated with neuropathy syndromes, such as GM1 and GD1a.

GM1b Ganglioside Serum polyclonal IgG binding to GM1b ganglioside occurs in some GBS patients with acute motor neuropathy.83,119,194,195 Preceding infection with C. jejuni is common. Weakness is distal and severe with rapid onset and slow recovery, usually with sparing of cranial nerves and sensation. Many sera also show binding to GM1 or GalNAc-GD1a, suggesting there may be cross-reactive antibodies to common structures in these three gangliosides.194,195 As has been reported in patients with acute neuropathies and anti-GM1 antibodies, clinical response to IVIG treatment may be better than to PE.194

GD1a Ganglioside Several reports suggest that serum antibody binding to GD1a ganglioside is associated with motor neuropathy syndromes.16,54,196 The clinical phenotype is similar to that of AMAN, with rapidly progressive distal weakness, frequent preceding C. jejuni infection, and electrophysiologic evidence of an axonal neuropathy. However, preparations of GD1a often contain GalNAc-GD1a as a contaminant. None of the studies of anti-GD1a antibodies tested the possibility that antibody binding was actually to GalNAc-GD1a. The clinical spectrum associated with selective antibody binding to GD1a ganglioside remains to be determined, although, as with IgG binding to GM1, Gal-NAcGD1a, and GM1b, there may be an association with the AMAN subtype of GBS.

GD1b Gangliosides (and Other Gangliosides with Internal Disialosyl Groups) Antibodies that bind selectively to gangliosides with disialosyl groups (Neu5Ac[␣2-8]Neu5Ac) attached to an internal oligosaccharide sugar, including GD1b, GD3, GQ1b, and GT1b, are associated with acute and chronic, predominantly sensory neuropathies.144,164,183,184 Monoclonal IgM binding to GD1b ganglioside occurs in a chronic neuropathy syndrome that typically includes distal sensory loss (especially large fiber modalities), sensory ataxia, bulbar weakness, ophthal-

Polyneuropathies and Antibodies to Nerve Components

moplegia, and normal limb strength,144,164,187 which has been given the acronym CANOMAD syndrome (chronic ataxic neuropathy with ophthalmoplegia, M protein, agglutination, and disialosyl antibodies). These IgM antibodies cross-react with Pr2 antigens on red blood cells and may cause cold agglutination. Electrodiagnostic studies may show evidence of demyelination or axonal degeneration. Polyclonal IgG binding to GD1b ganglioside, without binding to GM1 ganglioside, occurs in acutely evolving sensory neuropathies that may be axonal or demyelinating.102,158,183,184 Panmodal sensory loss evolves over a period of weeks in proximal and distal regions. Strength is preserved and tendon reflexes are lost. Spontaneous improvement or response to immunomodulating therapy occurs in some patients, although recovery can be incomplete. Antibody Testing and Treatment Clinical testing for antibody binding to GD1b ganglioside is used in the evaluation of sensory neuropathy syndromes. The finding of monoclonal serum IgM antibodies to GD1b, and to other carbohydrate antigens with internal disialosyl groups, distinguishes an immunemediated chronic sensory neuropathy syndrome from other chronic ataxic sensory neuronopathies. There are a few reports of improvement in this neuropathy syndrome after IVIG144 and rituximab161 therapy. The finding of serum IgG antibodies to GD1b in a patient with an acute sensory neuropathy helps to direct the evaluation and treatment by distinguishing the disorder from paraneoplastic (anti-Hu) neuronopathy, Sjögren’s syndrome, and toxic sensory neuropathies (caused by cis-platinum or pyridoxine). Association with Disease Pathogenesis Anti-disialosyl antibodies may be involved in disease pathogenesis. Antibodies against GD1b ganglioside bind to dorsal root ganglion neurons, paranodal myelin, myelinated axons, and motor nerve terminals.45,84,186 A monoclonal anti-GD1b antibody generated in mice lacking complex gangliosides bound preferentially to large and medium dorsal root ganglion neurons, supporting the concept that anti-GD1b antibodies may cause selective injury to the subpopulation of large sensory neurons that are damaged in patients with these antibodies.45 Antibodies against disialosyl gangliosides can also cause cell death of dorsal root ganglion cells in culture and cause degeneration of rabbit sensory neurons after passive transfer.81,123 An affinity-purified anti-disialosyl antibody impaired nerve excitability and reduced quantal release of neurotransmitter from a phrenic nerve–hemidiaphragm preparation.186 Finally, rabbits immunized with GD1b ganglioside develop a syndrome of ataxia and areflexia with structural damage to sensory neurons, similar to the human acute sensory neuropathy with IgG anti-GD1b antibodies.82

2187

ANTIBODIES WITH BINDING TO PERIPHERAL NERVE GLYCOPROTEINS AND GLYCOLIPIDS Several different myelin proteins and glycolipids, the best characterized being MAG, are important targets of the immune system in some peripheral neuropathies. MAG and myelin protein zero (MPZ) are both members of the immunoglobulin superfamily of cell adhesion molecules that serve as recognition molecules by binding to other molecules of MAG or MPZ or to other immunoglobulinlike structures. In addition, MAG and MPZ, as well as peripheral myelin protein 22 (PMP22) and some sulfated glycolipids, all contain the HNK-1 epitope, a glucuronic acid–containing carbohydrate that may confer binding specificity of these molecules.

Myelin-Associated Glycoprotein Serum IgM binding to MAG is found in 50% to 60% of patients who present with a polyneuropathy and a serum IgM monoclonal protein of the MGUS type.37,111,138,139,153 The clinical anti-MAG polyneuropathy is a predominantly sensory syndrome that is slowly progressive, symmetrical, and predominantly distal. Legs are involved more severely than arms. Sensory loss, involving all modalities, occurs earlier than weakness. Weakness, most severe in the distal legs, eventually develops in 80% of patients with antiMAG antibody–associated polyneuropathies. Tendon reflexes are usually absent at the ankles, and hypoactive or absent elsewhere. Gait disorders, with inability to walk in tandem on examination, are a major cause of disability in a majority of patients. Tremor, predominantly in the hands, may also become disabling, particularly later in the disease course. Electrophysiologic testing suggests a polyneuropathy with demyelinative features in over 90% of patients (especially those with anti-MAG titers ⬎ 5000), although most have evidence of axonal degeneration as well.35,65,138 The earliest and most consistent change, in up to 90% of patients, is prolonged distal motor latencies, usually measured as a prolonged terminal latency index.6,18,65,171 Slowing of motor conduction velocities occurs in about 50% of patients, and becomes more common with increasing disease duration, although progressive axonal degeneration may account for ongoing disability.144 Conduction block is uncommon in anti-MAG antibody–related demyelinating polyneuropathies. Prolonged terminal latency index and the absence of conduction block differentiate this demyelinating neuropathy electrophysiologically from CIDP6,18,65,171 (see Table 97–2). Nerve biopsies often show both demyelination and axonal loss, but no inflammation. A characteristic pathologic abnormality in anti-MAG polyneuropathies, widening

2188

Diseases of the Peripheral Nervous System

of myelin lamellae,35,99,166 is best demonstrated on ultrastructural examination of nerve. Another electron microscopic finding in sural nerve biopsies of anti-MAG patients is reduced neurofilament spacing,97 which is not found in biopsies of patients with CIDP or GBS. Despite these distinctive ultrastructural abnormalities, nerve biopsy is not generally utilized in the diagnosis of the anti-MAG neuropathy syndrome. Antibody Testing and Treatment Clinical testing for IgM anti-MAG antibodies may be useful in slowly progressive sensory or sensorimotor neuropathies. Several features increase the likelihood of a positive test, including patient age above 50 years, gait disorder, tremor, and a demyelinating neuropathy with prolonged distal latencies but no conduction block.132 IgM M proteins are detected by immunofixation in serum in 85% of patients. The finding of IgM anti-MAG antibodies, especially when confirmed by Western blot methodology, is strongly associated with the presence of a sensory ⬎ motor demyelinating neuropathy syndrome. The clinical features of IgM anti-MAG antibody– associated neuropathies that distinguish them from other immune demyelinating neuropathies, including CIDP, include the relative preservation of proximal strength and the electrophysiologic finding of prolonged distal latencies without conduction block. Long-term retrospective studies of patients with anti-MAG antibody–associated neuropathies showed moderate disability developing after several years.46 The finding of anti-MAG antibodies helps to identify this syndrome with poor sustained responsiveness to many immunomodulating therapies. There is little response to treatment with PE, prednisone, IVIG, chlorambucil, or other immunosuppressive agents.46,114 AntiMAG antibody testing may be useful after treatment of the associated neuropathy using cyclophosphamide.8 Most patients who improve after treatment have titers of antiMAG antibodies reduced by at least 60%. Improvement is rare without a significant reduction in anti-MAG titers. Rising titers of IgM anti-MAG antibodies precede by 3 to 6 months the reappearance of symptoms. Rituximab may produce persistent benefit for 1 to 3 years in patients with anti-MAG–associated neuropathy, especially in the gait disorder and weakness.137,149 Association with Disease Pathogenesis The antigenic binding site on MAG is thought to be an HNK-1–like epitope containing a sulfate 3-glucuronate moiety that is also found on MPZ and PMP22. The sulfated glycolipids, SGPG and sulfated glucuronyllactosaminyl paragloboside (see below), also contain this epitope.147 There is substantial evidence that IgM antiMAG antibodies can cause a demyelinating neuropathy. Anti-MAG antibodies typically bind to noncompact

myelin, but can also bind to compact peripheral nervous system myelin and associated Schwann cell cytoplasm.40,92,166 Deposits of IgM and complement can be found on myelin sheaths in biopsies from patients with anti-MAG neuropathy.51,104 Direct injection into cat sciatic nerve of serum from patients with anti-MAG antibodies has produced widening of myelin lamellae and demyelination in experimental animals.50,105,188 Passive transfer of human anti-MAG IgM into chicks produces peripheral nerve pathology that is similar to the human disorder.168 Clinically, the presence of anti-MAG antibodies strongly correlates with the associated neuropathy.20,113 Anti-MAG antibodies found in the serum of asymptomatic patients predict that a typical demyelinating neuropathy will develop during the next few years.100 Long-term reduction in anti-MAG antibody titers correlates with improvement in the associated neuropathy.8,69,137 While much of the above data suggest that MAG is the pathogenic target of anti-MAG antibodies, effects of the antibodies could also be due, at least in part, to binding to other molecules with HNK-1, or other sulfated carbohydrate epitopes.

Sulfated Glucuronyl Paragloboside IgM binding to SGPG25,58,76 is usually found in patients with anti-MAG antibodies. However, serum IgM binding to SGPG is not specific to the demyelinating neuropathies associated with anti-MAG antibodies. AntiSGPG antibodies, especially without associated binding to MAG, are not strongly associated with a specific syndrome. They occur not only in demyelinating neuropathies, but also in motor syndromes and axonal peripheral neuropathies.138,154,174,182 Generation of antibodies to SGPG in Lewis rats does not produce evidence of neuropathy.57 Anti-MAG antibodies that cross-react with SGPG more often bind to axons, while anti-MAG antibodies without SGPG cross-reactivity usually bind to areas of noncompact myelin.92 There is no clear clinical indication to test for anti-SGPG antibodies.

Peripheral Myelin Protein 22 and Myelin Protein Zero Mutations in the genes for PMP22 and MPZ, both components of compact myelin, cause hereditary demyelinating neuropathies, but may also play a role in autoimmune neuropathies. Both PMP22 and MPZ can induce experimental allergic neuritis.44,101 Anti-MAG antibodies may cross react with PMP22 and MPZ, likely by binding to an HNK-1–like epitope.182 A few reports suggest that antibodies to both PMP22 and MPZ may be common in GBS or CIDP.42,70,192 Anti-PMP22 antibodies may occur at a higher frequency in patients with a PMP22 duplication, especially those with stepwise disease progression.43

Polyneuropathies and Antibodies to Nerve Components

2189

However, anti-PMP22 antibodies can also occur in several other neuropathy syndromes and may have little clinical specificity.87,150 The significance and frequency of selective antibody binding to PMP22 and MPZ in different neuropathy syndromes remains to be determined.

antibody–related neuropathies are limited to cyclophosphamide for syndromes including demyelination and monoclonal IgM antibodies, or IVIG for patients with polyclonal antisulfatide antibodies and a painful polyneuropathy.34

Sulfatide

Association with Disease Pathogenesis Several lines of evidence suggest that antisulfatide antibodies could play a role in the pathogenesis of the associated neuropathy. Polyclonal IgM antisulfatide antibodies can bind to dorsal root ganglia148 or peripheral nerve axons.94 Monoclonal IgM antisulfatide antibodies often bind to peripheral nerve myelin.39,94,142 Nerve biopsies from patients with serum IgM binding to sulfatide and demyelinating neuropathies show deposition of IgM and components of complement on myelin.39 This pattern of antibody binding could be related to the myelin widespacing that can occur in this syndrome. An experimental demyelinating neuropathy developed in guinea pigs after immunization with sulfatide,146 and in rabbits after passive transfer of serum from a patient with an IgM monoclonal antisulfatide antibody.107

Polyneuropathies associated with serum IgM antisulfatide antibodies are usually predominantly sensory.15,36,93,139 A distal, predominantly small fiber neuropathy is associated with polyclonal IgM antisulfatide antibodies.15,31 Examination shows distal symmetrical sensory loss in the legs and arms, primarily to pin prick and temperature sensations, but in some cases to large fiber modalities as well. Distal pain and paresthesias are the most frequent presenting complaints. More proximal pain and paresthesias, a pattern not typical for neuropathies, can also be present. Mild distal foot weakness occurs rarely. Electrophysiologic studies may be normal or show axonal features with absent sural sensory nerve action potentials. Several different features occur more frequently in patients with monoclonal IgM antisulfatide antibodies.15,93 Symptoms of large fiber sensory loss are more prominent. Gait ataxia, similar to that seen in the antiMAG syndrome, may be present. Pain is not a common feature. Distal weakness is more common than with polyclonal antisulfatide antibodies. A comparison of patients with IgM M proteins and antisulfatide or antiMAG antibodies showed that both groups had a chronic, slowly progressive distal sensory, or sensorimotor, neuropathy with hyporeflexia, tremor, and gait ataxia.36 Electrophysiologic studies in most patients show demyelinating features.15,93 Occasional patients with antisulfatide antibodies and serum M proteins have predominantly axonal features.31,36 Antibody Testing Clinical testing for IgM antisulfatide antibodies may be useful in evaluation of slowly progressive sensory or sensorimotor neuropathies. Several features increase the likelihood of finding antisulfatide antibodies: patient age above 50 years, disability from neuropathic symptoms, gait disorder, tremor, demyelinating neuropathy on electrophysiologic testing, and IgM M protein in serum.132 The finding of high titers (above 5000) of IgM antisulfatide antibodies has 80% to 90% specificity for a sensory ⬎ motor neuropathy syndrome.60,93 Antisulfatide antibody titers are often higher in patients with M proteins than in those with polyclonal antibodies.15,93 Moderate or low titers of antisulfatide antibodies have been reported in patients with infections, including human immunodeficiency virus95,141 and trypanosomiasis.7 Anecdotal reports suggest that treatment options for antisulfatide

GALOP Antigen (Sulfatide in a Lipid Membrane Environment) Patients with the GALOP syndrome (gait disorder, autoantibody, late-age onset, and polyneuropathy) have IgM binding to sulfatide that occurs when it is in a lipid membrane (cholesterol, galactocerebroside [GalC]) environment. The GALOP syndrome combines gait ataxia and a sensorimotor polyneuropathy with late-age onset (mean ⫽ 70 years of age).134 Weakness, when present, is mild and predominantly distal. Sensory loss is distal, is mild to moderate in degree, and involves large and small fiber modalities. The gait is moderately wide based, associated with a positive Romberg phenomenon, and more severe than can be explained by the degree of motor or sensory loss on physical examination. The gait disorder often causes severe functional disability and frequent falling. Electrophysiologic studies are heterogeneous, but 80% of patients have features suggesting demyelination, similar to those found in patients with anti-MAG neuropathies. Clinical testing for IgM anti-GALOP antibodies may be useful in slowly progressive gait disorders with an accompanying sensory or sensorimotor neuropathy in patients over 50 years of age. The finding of antibodies to the GALOP antigen distinguishes this sensory neuropathy syndrome from other gait disorders in the elderly, and from other ataxic sensory neuronopathies that have more severe and widespread sensory loss and rarely improve after treatment. In the GALOP syndrome, immunomodulating treatment with IVIG or cyclophosphamide may result in functional improvement.

2190

Diseases of the Peripheral Nervous System

Galactocerebroside

Heparan Sulfate

Polyclonal IgG or IgM antibody binding to GalC occurs in some patients with classical motor-sensory demyelinating GBS (1% to 12%).3,48 There are no clear correlations between anti-GalC binding and clinical parameters. Patients have been reported to have milder symptoms, an increased number of respiratory tract infections, more sensory deficits and paresthesias, and demyelinating features on electrophysiologic studies. Many patients with anti-GalC reactivity (50% to 73%) have serologic evidence of previous infection with Mycoplasma pneumoniae.3 Preincubation of serum samples with a M. pneumoniae surface antigen can inhibit anti-GalC binding, suggesting a cross-reactivity between anti-GalC antibodies and a cell surface molecule on M. pneumoniae.3,85 Immunization with GalC can lead to a demyelinating neuropathy,155 and anti-GalC antibodies have a demyelinating effect in vitro.41 However, the role of anti-GalC antibodies in the development of acute immune neuropathies remains to be determined.

Polyclonal IgM and IgG antibodies to heparan sulfate occur in 35% of patients with motor-sensory, demyelinating GBS.136 Monoclonal IgM binding to heparan sulfate occurs in some sera from patients with anti-MAG neuropathies. The role of testing for antibodies to heparan sulfate in acute neuropathy syndromes is not clearly established. In patients with rapidly progressive weakness, polyclonal IgM or IgG anti–heparan sulfate antibodies may suggest an immune disorder.

OTHER SERUM ANTIBODIES A few reports have described other antibodies to peripheral nerve antigens.

Trisulfated Heparin Disaccharide Five patients with polyneuropathy were found to have serum IgM binding to the trisulfated heparin disaccharide (TS-HDS) IdoA2S-GlcNS-6S, the most abundant disaccharide component of heparin oligosaccharides.135 The patients all had painful, predominantly sensory polyneuropathies. Sensory loss was distal and panmodal. Electrophysiologic and pathologic studies were consistent with axonal loss, especially of unmyelinated axons. Immunohistochemistry showed IgM and ␬ light chains deposited around the rim of intermediate-sized veins in the perimysium and epineurium. Serum IgM binding to TS-HDS was selective, present in high titer (⬎12,000), and limited to ␬ light chains. The role of IgM binding to TS-HDS in the pathogenesis of the neuropathy remains to be determined.

␤-Tubulin IgM monoclonal antibodies that bind selectively to a specific epitope on ␤-tubulin, amino acids 301–314, are associated with CIDP.28,29 Weakness is often slowly progressive and may be asymmetrical. Response to prednisone treatment is poor. Although only described in a few patients,28,165 clinical testing for IgM binding to ␤-tubulin may be useful in demyelinating neuropathies, especially those syndromes resistant to therapy.

Neurofilaments Antibody binding to the high-molecular-weight (200kDa) neurofilament subunit was described in a patient with a chronic sensory axonal neuropathy and an IgM M protein.9 In a patient with a sensorimotor neuropathy and a serum IgG M protein, IgG binding was found to the low-molecular-weight (68-kDa) neurofilament subunit.108 Four patients with MGUS had polyclonal antineurofilament antibody binding, and intraneural injection of these patient sera caused conduction block.162 Low-titer serum IgM binding to neurofilaments has no specific clinical associations and has been described in many different disorders, including spongiform encephalopathies, systemic lupus erythematosus, rheumatoid arthritis and neuropathy, ALS, Parkinson’s disease, Alzheimer’s disease, and childhood opsoclonusmyoclonus, as well as normal controls.27,156

REFERENCES 1. Ang, C. W., Laman, J. D., Willison, H. J., et al.: Structure of Campylobacter jejuni lipopolysaccharides determines antiganglioside specificity and clinical features of GuillainBarré and Miller Fisher patients. Infect. Immun. 70:1202, 2002. 2. Ang, C. W., Noordzij, P. G., De Klerk, M. A., et al.: Ganglioside mimicry of Campylobacter jejuni lipopolysaccharides determines antiganglioside specificity in rabbits. Infect. Immun. 70:5081, 2002. 3. Ang, C. W., Tio-Gillen, A. P., Groen, J., et al.: Cross-reactive anti-galactocerebroside antibodies and Mycoplasma pneumoniae infections in Guillain-Barré syndrome. J. Neuroimmunol. 130:179, 2002. 4. Ang, C. W., Yuki, N., Jacobs, B. C., et al.: Rapidly progressive, predominantly motor Guillain-Barré syndrome with anti-GalNAc-GD1a antibodies. Neurology 53:2122, 1999. 5. Annunziata, P., Figura, N., Galli, R., et al.: Association of anti-GM1 antibodies but not of anti-cytomegalovirus, Campylobacter jejuni and Helicobacter pylorus IgG with a poor outcome in Guillain-Barré syndrome. J. Neurol. Sci. 213:55, 2003. 6. Attarian, S., Azulay, J. P., Boucraut, J., et al.: Terminal latency index and modified F ratio in distinction of chronic

Polyneuropathies and Antibodies to Nerve Components

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

demyelinating neuropathies. Clin. Neurophysiol. 112:457, 2001. Avila, J. L., Rojas, M., and Carrasco, H.: Elevated levels of antibodies against sulphatide are present in all chronic chagasic and non-chagasic dilatory cardiomyopathy sera. Clin. Exp. Immunol. 92:460, 1993. Blume, G., Pestronk, A., and Goodnough, L. T.: Anti-MAG antibody associated polyneuropathies: improvement following immunotherapy with monthly plasma exchange and intravenous cyclophosphamide. Neurology 45:1577, 1995. Brindel, I., Preudhomme, J. L., Diaz, J. J., et al.: A human monoclonal IgM lambda specific for an epitope shared by the 200 kDa neurofilament protein, histones and ribosomal proteins. J. Autoimmun. 8:915, 1995. Buchwald, B., Bufler, J., Carpo, M., et al.: Combined preand postsynaptic action of IgG antibodies in Miller Fisher syndrome. Neurology 56:67, 2001. Buchwald, B., Weishaupt, A., Toyka, K. V., and Dudel, J.: Pre- and postsynaptic blockade of neuromuscular transmission by Miller-Fisher syndrome IgG at mouse motor nerve terminals. Eur. J. Neurosci. 10:281, 1998. Butler, J. E.: Solid phases in immunoassay. In Diamandis, E. P., and Christopoulis, T. K. (eds.): Immunoassays. Boca Raton, CRC Press, p. 205, 1996. Carpenter, A. B.: Enzyme linked immunoassays. In Rose, N. R., de Macario, J. L., Fahey, J. L., et al. (eds.): Manual of Clinical Laboratory Immunology. Washington, DC, American Society of Microbiology, p. 2, 1992. Carpo, M., Allaria, S., Scarlato, G., and Nobile-Orazio, E.: Marginally improved detection of GM1 antibodies by Covalink ELISA in multifocal motor neuropathy. Neurology 53:2206, 1999. Carpo, M., Meucci, N., Allaria, S., et al.: Anti-sulfatide IgM antibodies in peripheral neuropathy. J. Neurol. Sci. 176:144, 2000. Carpo, M., Nobile-Orazio, E., Meucci, N., et al.: AntiGD1a ganglioside antibodies in peripheral motor syndromes. Ann. Neurol. 39:539, 1996. Carpo, M., Pedotti, R., Lolli, F., et al.: Clinical and fine specificity of anti-GQ1b antibodies in peripheral neuropathy. J. Neurol. Sci. 155:186, 1998. Caspasso, M., Torrieri, F., Di Muzio, A., et al.: Can electrophysiology differentiate polyneuropathy with antiMAG/SGPG antibodies from chronic inflammatory demyelinating polyneuropathy? Clin. Neurophysiol. 113:346, 2002. Cavanna, B., Jiang, H., Allaria, S., et al.: Anti-GM2 IgM antibody-induced complement-mediated cytotoxicity in patients with disimmune neuropathies. J. Neuroimmunol. 114:226, 2001. Chassande, B., Leger, J. M., Ben Younes-Chennoufi, A., et al.: Peripheral neuropathy associated with IgM monoclonal gammopathy: correlations between M-protein antibody activity and clinical/electrophysiological features in 40 cases. Muscle Nerve 21:55, 1998. Chaudhry, V., Corse, A. M., Cornblath, D. R., et al.: Multifocal motor neuropathy: response to human immune globulin. Ann. Neurol. 33:237, 1993. Chiba, A., Kusunoki, S., Obata, H., et al.: Serum anti-GQ1b IgG antibody is associated with ophthalmoplegia in Miller

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34. 35.

36.

37.

2191

Fisher syndrome and Guillain-Barré syndrome: clinical and immunohistochemical studies. Neurology 43:1911, 1993. Chiba, A., Kusunoki, S., Obata, H., et al.: Ganglioside composition of human cranial nerves, with special reference to pathophysiology of Miller Fisher syndrome. Brain Res. 745:32, 1997. Chiba, A., Kusunoki, S., Shimizu, T., and Kanazawa, I.: Serum IgG antibody to ganglioside GQ1b is a possible marker of Miller Fisher syndrome. Ann. Neurol. 31:677, 1992. Chou, D. K., Ilyas, A. A., Evans, J. E., et al.: Structure of sulfated glucuronyl glycolipids in the nervous system reacting with HNK-1 antibody and some IgM paraproteins in neuropathy. J. Biol. Chem. 261:11717, 1986. Christopoulis, T. K., and Diamandis, E. P.: Immunoassay configurations. In Diamandis, E. P., and Christopoulis, T. K. (eds.): Immunoassay. San Diego, Academic Press, p. 227, 1996. Connolly, A. M., Pestronk, A., Mehta, S., et al.: Serum autoantibodies in childhood opsoclonus-myoclonus syndrome: an analysis of antigenic targets in neural tissue. J. Pediatr. 130:878, 1997. Connolly, A. M., Pestronk, A., Mehta, S., et al.: Serum IgM monoclonal autoantibody binding to the 301 to 314 amino acid epitope of beta-tubulin: clinical association with slowly progressive demyelinating polyneuropathy. Neurology 48:243, 1997. Connolly, A. M., Pestronk, A., Trotter, J. L., et al.: High titer selective anti-␤-tubulin antibodies in chronic inflammatory demyelinating polyneuropathy. Neurology 43:557, 1993. Corbo, M., and Hayes, A. C.: Immunopathological studies in immune-mediated neuropathies. In Latov, N., Wokke, J. H. J., and Kelly, J. J. (eds.): Immunology and Infectious Diseases of the Peripheral Nerves. Cambridge, UK, Cambridge University Press, p. 81, 1998. Dabby, R., Weimer, L. H., Hays, A. P., et al.: Antisulfatide antibodies in neuropathy: clinical and electrophysiologic correlates. Neurology 54:1448, 2000. Dalmau, J., Graus, F., Rosenblum, M. K., and Posner, J. B.: Anti-Hu-associated paraneoplastic encephalomyelitis/sensory neuronopathy: a clinical study of 71 patients. Medicine (Baltimore) 71:59, 1992. Dilley, A., Gregson, N. A., Hadden, R. D., and Smith, K. J.: Effects on axonal conduction of anti-ganglioside sera and sera from patients with Guillain-Barré syndrome. J. Neuroimmunol. 139:133, 2003. Donofrio, P. D.: Immunotherapy of idiopathic inflammatory neuropathies. Muscle Nerve 28:273, 2003. Ellie, E., Vital, A., Steck, A. J., et al.: Neuropathy associated with ‘benign’ anti-myelin-associated glycoprotein IgM gammopathy: clinical, immunological, neurophysiological, and pathological findings and response to treatment in 33 cases. J. Neurol. 243:34, 1996. Erb, S., Ferracin, F., Fuhr, P., et al.: Polyneuropathy attributes: a comparison between patients with anti-MAG and anti-sulfatide antibodies. J. Neurol. 247:767, 2000. Eurelings, M., Moons, K. G. M., Notermans, N. C., et al.: Neuropathy and IgM M-proteins; prognostic value of antibodies to MAG, SGPG, and sulfatide. Neurology 56:228, 2001.

2192

Diseases of the Peripheral Nervous System

38. Feldman, E. L., Bromberg, M. B., Albers, J. W., and Pestronk, A.: Immunosuppressive treatment in multifocal motor neuropathy. Ann. Neurol. 30:397, 1991. 39. Ferrari, S., Morbin, M., Nobile-Orazio, E., et al.: Antisulfatide polyneuropathy: antibody-mediated complement attack on peripheral myelin. Acta Neuropathol. (Berl.) 96:569, 1998. 40. Fluri, F., Ferracin, F., Erne, B., and Steck, A. J.: Microheterogeneity of anti-myelin-associated glycoprotein antibodies. J. Neurol. Sci. 207:43, 2003. 41. Fry, J. M., Weissbarth, S., Lehrer, G. M., and Bornstein, M. B.: Cerebroside antibody inhibits sulfatide synthesis and myelination and demyelinates in cord tissue cultures. Science 183:540, 1974. 42. Gabriel, C. M., Gregson, N. A., and Hughes, R. A. C.: AntiPMP22 antibodies in patients with inflammatory neuropathy. J. Neuroimmunol. 104:139, 2000. 43. Gabriel, C. M., Gregson, N. A., Wood, N. W., and Hughes, R. A. C.: Immunological study of hereditary motor and sensory neuropathy type 1a (HMSN1a). J. Neurol. Neurosurg. Psychiatry 72:230, 2002. 44. Gabriel, C. M., Hughes, R. A. C., Moore, S. E., et al.: Induction of experimental autoimmune neuritis with peripheral myelin protein-22. Brain 121:1895, 1998. 45. Gong, Y., Tagawa, Y., Lunn, M. P. T., et al.: Localization of major gangliosides in the PNS: implications for immune neuropathies. Brain 125:2491, 2002. 46. Gorson, K. C., Ropper, A. H., Weinberg, D. H., and Weinstein, R.: Treatment experience in patients with antimyelin-associated glycoprotein neuropathy. Muscle Nerve 24:778, 2001. 47. Hakomori, S. I., Yamamura, S., and Handa, K.: Signal transduction through glyco(sphingo)lipids: introduction and recent studies on glyco(sphingo)lipid-enriched microdomains. Ann. N. Y. Acad. Sci. 845:1, 1998. 48. Hao, Q., Saida, T., Kuroki, S., et al.: Antibodies to gangliosides and galactocerebroside in patients with Guillain-Barré syndrome with preceding Campylobacter jejuni and other identified infections. J. Neuroimmunol. 81:116, 1998. 49. Hao, Q., Saida, T., Yoshino, H., et al.: Anti-GalNAc-GD1a antibody-associated Guillain-Barré syndrome with a predominantly distal weakness without cranial nerve impairment and sensory disturbance. Ann. Neurol. 45:758, 1999. 50. Hays, A. P., Latov, N., Takatsu, M., and Sherman, W. H.: Experimental demyelination of nerve induced by serum of patients with neuropathy and an anti-MAG IgM M-protein. Neurology 37:242, 1987. 51. Hays, A. P., Lee, S. S., and Latov, N.: Immune reactive C3d on the surface of myelin sheaths in neuropathy. J. Neuroimmunol. 23:394, 1988. 52. Hirota, N., Kaji, R., Bostock, H., et al.: The physiological effect of anti-GM1 antibodies on saltatory conduction and transmembrane currents in single motor axons. Brain 120:2159, 1997. 53. Ho, T. W., Mishu, B., Li, C. Y., et al.: Guillain-Barré syndrome in northern China: relationship to Campylobacter jejuni infection and anti-glycolipid antibodies. Brain 118:597, 1995. 54. Ho, T. W., Willison, H. J., Nachamkin, I., et al.: Anti-GD1a antibody is associated with axonal but not demyelinating

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67. 68.

69.

70.

71.

forms of Guillain-Barré syndrome. Ann. Neurol. 45:168, 1999. Hughes, R. A. C., Hadden, R. D., Gregson, N. A., and Smith, K. J.: Pathogenesis of Guillain-Barré syndrome. J. Neuroimmunol. 100:74, 1999. Ifergone, G., Bersudsky, M., Kachko, L., and Wirguin, I.: Induction of conduction block by Campylobacter jejuni lipopolysaccharides and focal neural insult. J. Neurol. Sci. 213:111, 2003. Ilyas, A. A., Chen, Z. W., and Prineas, J. W.: Generation and characterization of antibodies to sulfated glucuronyl glycolipids in Lewis rats. J. Neuroimmunol. 127:54, 2002. Ilyas, A. A., Dalakas, M. C., Brody, R. O., and Quarles, R. H.: Sulfated glucuronyl glycolipids reacting with anti-myelinassociated glycoprotein monoclonal antibodies including IgM paraproteins in neuropathy: species distribution and partial characterization of epitopes. Brain Res. 385:1, 1986. Irie, S., Saito, T., Nakamura, K., et al.: Association of antiGM2 antibodies in Guillain-Barré syndrome with acute cytomegalovirus infection. J. Neuroimmunol. 68:19, 1996. Isoardo, G., Ferrero, B., Barbero, P., et al.: Anti-GM1 and anti-sulfatide antibodies in polyneuropathies: threshold titers and accuracy. Acta Neurol. Scand. 103:180, 2001. Kaida, K., Kusunoki, S., Kamakura, K., et al.: GuillainBarré syndrome with antibody to ganglioside, N-acetylgalactosaminyl GD1a. Brain 123:116, 2000. Kaida, K., Kusunoki, S., Kamakura, K., et al.: GuillainBarré syndrome with IgM antibody to the ganglioside GalNAc-GD1a. J. Neuroimmunol. 113:260, 2001. Kaida, K., Kusunoki, S., Kamakura, K., et al.: GalNAcGD1a in human peripheral nerve: target sites of antiganglioside antibody. Neurology 61:465, 2003. Kaji, R., Kusunoki, S., Mizutani, K., et al.: Chronic motor axonal neuropathy associated with antibodies monospecific for N-acetylgalactosaminyl GD1a. Muscle Nerve 23:702, 2000. Kaku, D. A., England, J. D., and Sumner, A. J.: Distal accentuation of conduction slowing in polyneuropathy associated with antibodies to myelin-associated glycoprotein and sulphated glucuronyl paragloboside. Brain 117:941, 1994. Kanda, T., Yamawaki, M., and Mizusawa, H.: Sera from Guillain-Barré patients enhance leakage in blood-nerve barrier model. Neurology 60:301, 2003. Kates, M.: Thin-layer chromatography. In Kates, M. (ed.): Techniques of Lipidology. Amsterdam, Elsevier, p. 232, 1986. Katz, J. S., Barhon, R. J., Kojan, S., et al.: Axonal multifocal motor neuropathy without conduction block or other features of demyelination. Neurology 58:615, 2002. Kelly, J. J., Adelman, L. S., Berkman, E., and Bhan, I.: Polyneuropathies associated with IgM monoclonal gammopathies. Arch. Neurol. 45:1355, 1988. Khalili-Shirazi, A., Atkinson, P., Gregson, N. A., and Hughes, R. A. C.: Antibody response to P0 and P2 myelin proteins in Guillain-Barré syndrome and chronic inflammatory demyelinating polyradiculoneuropathy. J. Neuroimmunol. 46:245, 1993. Kirchner, J.: Techniques of thin-layer chromatography. In Perry, E. S. (ed.): Thin-Layer Chromatography. St. Louis, Sigma Chemical Company, p. 3, 1990.

Polyneuropathies and Antibodies to Nerve Components 72. Koga, M., Yoshino, H., Morimatsu, M., and Yuki, N.: AntiGT1a IgG in Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 72:767, 2002. 73. Koga, M., Yuki, N., and Hirata, K.: Antiganglioside antibody in patients with Guillain-Barré syndrome who show bulbar palsy as an initial symptom. J. Neurol. Neurosurg. Psychiatry 66:513, 1999. 74. Koga, M., Yuki, N., Hirata, K., et al.: Anti-GM1 antibody IgG subclass: a clinical recovery predictor in Guillain-Barré syndrome. Neurology 60:1514, 2003. 75. Koga, M., Yuki, N., Tai, T., and Hirata, K.: Miller Fisher syndrome and Haemophilus influenzae infection. Neurology 57:686, 2001. 76. Kohriyama, T., Kusunoki, S., Ariga, T., et al.: Subcellular localization of sulfated glucuronic acid-containing glycolipids reacting with anti-myelin-associated glycoprotein antibody. J. Neurochem. 48:1516, 1987. 77. Kornberg, A. J., and Pestronk, A.: Antibody-associated polyneuropathy syndromes: principles and treatment. Semin. Neurol. 23:181, 2003. 78. Kornberg, A. J., Pestronk, A., Bieser, K., et al.: The clinical correlates of high-titer IgG anti-GM1 antibodies. Ann. Neurol. 35:234, 1994. 79. Kornberg, A. J., Pestronk, A., Blume, G., et al.: Selective staining of the cerebellar molecular layer by serum IgG in MillerFisher and related syndromes. Neurology 47:1317, 1996. 80. Kusunoki, S., Chiba, A., and Kanazawa, I.: Anti-GQ1b antibody is associated with ataxia as well as ophthalmoplegia. Muscle Nerve 22:1071, 1999. 81. Kusunoki, S., Hitoshi, S., Kaida, K., et al: Degeneration of rabbit sensory neurons induced by passive transfer of antiGD1b antiserum. Neurosci. Lett. 273:33, 1999. 82. Kusunoki, S., Hitoshi, S., Kaida, K., et al: Monospecific anti-GD1b IgG is required to induce rabbit ataxic neuropathy. Ann. Neurol. 45:400, 1999. 83. Kusunoki, S., Iwamori, M., Chiba, A., et al.: GM1b is a new member of antigen for serum antibody in Guillain-Barré syndrome. Neurology 47:237, 1996. 84. Kusunoki, S., Mashiko, H., Mochizuki, N., et al.: Binding of antibodies against GM1 and GD1b in human peripheral nerve. Muscle Nerve 20:840, 1997. 85. Kusunoki, S., Shiina, M., and Kanazawa, I.: Anti-Gal-C antibodies in GBS subsequent to mycoplasma infection: evidence of molecular mimicry. Neurology 57:736, 2001. 86. Kuwabara, S., Mori, M., Ogawara, K., et al.: Intravenous immunoglobulin therapy for Guillain-Barré syndrome with IgG anti-GM1 antibody. Muscle Nerve 24:54, 2001. 87. Kwa, M. S., van Schaik, I. N., Brand, A., et al.: Investigation of serum response to PMP22, connexin 32 and P(0) in inflammatory neuropathies. J. Neuroimmunol. 116:220, 2001. 88. Kyle, R. A.: Monoclonal proteins in neuropathy. Neurol. Clin. 10:713, 1992. 89. Lefaucheur, J. P., Gregson, N. A., Gray, I., et al.: A variant of multifocal motor neuropathy with acute, generalised presentation and persistent conduction blocks. J. Neurol. Neurosurg. Psychiatry 74:1555, 2003. 90. Levine, T. D., and Pestronk, A.: IgM antibody-related polyneuropathies: treatment with B cell depletion chemotherapy using rituximab. Neurology 52:1701, 1999.

2193

91. Lopate, G., Choksi, R., and Pestronk, A.: Severe sensory ataxia and demyelinating polyneuropathy with IgM anti-GM2 and GalNAc-GD1a antibodies. Muscle Nerve 25:828, 2002. 92. Lopate, G., Kornberg, A. J., Yue, J., et al.: Anti-myelin associated glycoprotein antibodies: variability in patterns of IgM binding to peripheral nerve. J. Neurol. Sci. 188:67, 2001. 93. Lopate, G., Parks, B. J., Goldstein, J. M., et al.: Polyneuropathies associated with high titre antisulphatide antibodies: characteristics of patients with and without serum monoclonal proteins. J. Neurol. Neurosurg. Psychiatry 62:581, 1997. 94. Lopate, G., Pestronk, A., Kornberg, A. J., et al.: IgM antisulfatide autoantibodies: patterns of binding to cerebellum, dorsal root ganglion and peripheral nerve. J. Neurol. Sci. 151:189, 1997. 95. Lopate, G., Streif, B., Reese, R., et al.: Anti-sulfatide antibodies in HIV infected patients with neuropathy. J. Neurol. Sci. 199(Suppl. 1):S52, 2002. 96. Lucchinetti, C. F., Kimmel, D. W., and Lennon, V. A.: Paraneoplastic and oncologic profiles of patients seropositive for type 1 antineuronal nuclear antibodies. Neurology 50:652, 1998. 97. Lunn, M. P. T., Crawford, T. O., Hughes, R. A. C., et al.: Anti-myelin-associated glycoprotein antibodies alter neurofilament spacing. Brain 125:904, 2002. 98. Marcus, D. M., Latov, N., Hsi, B. P., and Gillard, B. K.: Measurement and significance of antibodies against GM1 ganglioside. J. Neuroimmunol. 25:255, 1989. 99. Mendell, J. R., Sahenk, Z., Whitaker, J. N., et al.: Polyneuropathy and IgM monoclonal gammopathy: studies of the pathogenic role of the anti-myelin-associated glycoprotein antibody. Ann. Neurol. 17:243, 1985. 100. Meucci, N., Baldini, L., Cappellari, A., et al.: Anti-myelinassociated glycoprotein antibodies predict the development of neuropathy in asymptomatic patients with IgM monoclonal gammopathy. Ann. Neurol. 46:119, 1999. 101. Milner, P., Lovelidge, C. A., Taylor, W. A., and Hughes, R. A. C.: P0 myelin produces experimental allergic neuritis in Lewis rats. J. Neurol. Sci. 79:275, 1987. 102. Miyazaki, T., Kusunoki, S., Kaida, K., et al.: Guillain-Barré syndrome associated with IgG monospecific to ganglioside GD1b. Neurology 56:1227, 2001. 103. Mizoguchi, K., Hase, A., Obi, T., et al.: Two species of antiganglioside antibodies in a patient with a pharyngealcervical-brachial variant of Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 57:1121, 1994. 104. Monaco, S., Bonetti, B., Ferrari, S., et al.: Complementmediated demyelination in patients with IgM monoclonal gammopathy and polyneuropathy. N. Engl. J. Med. 322:649, 1990. 105. Monaco, S., Ferrari, S., Bonetti, B., et al.: Experimental induction of myelin changes by anti-MAG antibodies and terminal complement complex. J. Neuropathol. Exp. Neurol. 54:96, 1995. 106. Mori, M., Kuwabara, S., Miyake, M., et al.: Haemophilus influenzae has a GM1 ganglioside-like structure and elicits Guillain-Barré syndrome. Neurology 52:1282, 1999. 107. Nardelli, E., Bassi, A., Mazzi, G., et al.: Systemic passive transfer studies using IgM monoclonal antibodies to sulfatide. J. Neuroimmunol. 63:29, 1995.

2194

Diseases of the Peripheral Nervous System

108. Nemni, R., Feltri, M. L., Fazio, R., et al.: Axonal neuropathy with monoclonal IgG kappa that binds to a neurofilament protein. Ann. Neurol. 28:361, 1990. 109. Niebroj-Dobosz, I., Janik, P., and Kwiecinski, H.: Serum IgM anti-GM1 ganglioside antibodies in lower motor neuron syndromes. Eur. J. Neurol. 11:13, 2004. 110. Nishimura, M., Nukina, M., Kuroki, S., et al.: Characterization of Campylobacter jejuni isolates from patients with Guillain-Barré syndrome. J. Neurol. Sci. 153:91, 1997. 111. Nobile-Orazio, E.: Neuropathies associated with anti-MAG antibodies and IgM monoclonal gammopathies. In Latov, N., Wokke, J. H. J., and Kelly, J. J. (eds.): Immunology and Infectious Diseases of the Peripheral Nerves. Cambridge, UK, Cambridge University Press, p. 168, 1988. 112. Nobile-Orazio, E.: Multifocal motor neuropathy. J. Neuroimmunol. 115:4, 2001. 113. Nobile-Orazio, E., Manfredini, E., Carpo, M., et al.: Frequency and clinical correlates of anti-neural IgM antibodies in neuropathy associated with IgM monoclonal gammopathy. Ann. Neurol. 36:416, 1994. 114. Nobile-Orazio, E., Meucci, N., Baldini, L., et al.: Longterm prognosis of neuropathy associated with anti-MAG IgM M-proteins and its relationship to immune therapies. Brain 123:710, 2000. 115. Nobile-Orazio, E., Meucci, N., Barbieri, S., et al.: High dose intravenous immunoglobulin therapy in multifocal motor neuropathy. Neurology 43:537, 1993. 116. Odaka, M., Yuki, N., and Hirata, K.: Anti-GQ1b IgG antibody syndrome: clinical and immunological range. J. Neurol. Neurosurg. Psychiatry 70:50, 2001. 117. Ogawa-Goto, K., and Abe, T.: Gangliosides and glycosphingolipids of peripheral nervous system myelin—a minireview. Neurochem. Res. 23:305, 1998. 118. Ogawa-Goto, K., Funamoto, N., Ohta, Y., et al.: Myelin gangliosides of human peripheral nervous system: an enrichment of GM1 in the motor nerve myelin isolated from cauda equina. J. Neurochem. 59:1844, 1992. 119. Ogawara, K., Kuwabara, S., Koga, M., et al.: Anti-GM1b IgG antibody is associated with acute motor axonal neuropathy and Campylobacter jejuni infection. J. Neurol. Sci. 210:41, 2003. 120. Ogawara, K., Kuwabara, S., Mori, M., et al.: Axonal Guillain-Barré syndrome: relation to anti-ganglioside antibodies and Campylobacter jejuni infection in Japan. Ann. Neurol. 48:624, 2000. 121. O’Hanlon, G. M., Plomp, J. J., Chakrabarti, M., et al.: AntiGQ1b antibodies mediate complement-dependent destruction of the motor nerve terminal. Brain 124:893, 2001. 122. O’Hanlon, G. M., Veitch, J., Gallardo, E., et al.: Peripheral neuropathy associated with anti-GM2 ganglioside antibodies: clinical and immunopathological studies. Autoimmunity 32:133, 2000. 123. Ohsawa, T., Miyatake, T., and Yuki, N.: Anti-B-series ganglioside-recognizing autoantibodies in an acute sensory neuropathy patient cause cell death of rat dorsal root ganglion neurons. Neurosci. Lett. 157:167, 1993. 124. O’Leary, C. P., and Willison, H. J.: The role of antiglycolipid antibodies in peripheral neuropathies. Curr. Opin. Neurol. 13:583, 2000.

125. Olney, R. K.: Consensus criteria for the diagnosis of partial conduction block. Muscle Nerve 22 (Suppl. 8):S225, 1999. 126. Ortiz, N., Rosa, R., Gallardo, E., et al.: IgM monoclonal antibody against terminal moiety of GM2, GalNAc-GD1a and GalNAc-GD1b from a pure motor chronic demyelinating polyneuropathy patient: effects of neurotransmitter release. J. Neuroimmunol. 119:114, 2001. 127. Paparounas, K., O’Hanlon, G. M., O’Leary, C. P., et al.: Anti-ganglioside antibodies can bind peripheral nerve nodes of Ranvier and activate the complement cascade without inducing acute conduction block in vitro. Brain 122:807, 1999. 128. Perry, G.: AAEM case report #30: multifocal motor neuropathy. Muscle Nerve 19:269, 1996. 129. Pestronk, A.: Chronic immune polyneuropathies and serum autoantibodies. In Rolak, L. A., and Yadollah, H. (eds.): Neuroimmunology for the Clinician. Boston, ButterworthHeinemann, p. 237, 1997. 130. Pestronk, A.: Multifocal motor neuropathy: diagnosis and treatment. Neurology 51:S22, 1998. 131. Pestronk, A.: Testing for serum IgM binding to GM1 ganglioside in clinical practice. Neurology 54:2353, 2000. 132. Pestronk, A.: Neuromuscular Disease Center, Washington University School of Medicine, St. Louis, MO, 2001–2003. Available at: www.neuro.wustl.edu/neuromuscular 133. Pestronk, A., and Choksi, R.: Multifocal motor neuropathy: serum IgM anti-GM1 ganglioside antibodies in most patients detected using covalent linkage of GM1 to ELISA plates. Neurology 49:1289, 1997. 134. Pestronk, A., Choksi, R., Bieser, K., et al.: Treatable gait disorder and polyneuropathy associated with high titer serum IgM binding to antigens that copurify with myelinassociated glycoprotein. Muscle Nerve 17:1293, 1994. 135. Pestronk, A., Choksi, R., Logigian, E., and Al-Lozi, M. T.: Sensory neuropathy with monoclonal IgM binding to a trisulfated heparin disaccharide. Muscle Nerve 27:188, 2003. 136. Pestronk, A., Choksi, R., Yee, W. C., et al.: Serum antibodies to heparan sulfate glycosaminoglycans in Guillain-Barré syndrome and other demyelinating polyneuropathies. J. Neuroimmunol. 91:204, 1998. 137. Pestronk, A., Florence, J., Miller, T., et al.: Treatment of IgM antibody associated polyneuropathies using rituximab. J. Neurol. Neurosurg. Psychiatry 74:485, 2003 138. Pestronk, A., Li, F., Bieser, K., et al.: Anti-MAG antibodies: major effects of antigen purity and antibody cross-reactivity on ELISA results and clinical correlation. Neurology 44:1131, 1994. 139. Pestronk, A., Li, F., Griffin, J. W., et al.: Polyneuropathy syndromes associated with serum antibodies to sulfatide and myelin-associated glycoprotein. Neurology 41:357, 1991. 140. Pestronk, A., Lopate, G., Kornberg, A. J., et al.: Distal lower motor neuron syndrome with high-titer serum IgM antiGM1 antibodies—improvement following immunotherapy with monthly plasma exchange and intravenous cyclophosphamide. Neurology 44:2027, 1994. 141. Petratos, S., Turnbull, V. J., Papadopoulos, R., et al.: High titre anti-sulphatide antibodies in HIV-infected individuals. Neuroreport 10:2557, 1999. 142. Petratos, S., Turnbull, V. J., Papadopoulos, R., et al.: Hightitre IgM anti-sulfatide antibodies in individuals with IgM

Polyneuropathies and Antibodies to Nerve Components

143.

144.

145.

146.

147. 148.

149.

150.

151.

152.

153.

154.

155.

156.

157.

158.

159.

paraproteinemia and associated peripheral neuropathy. Immunol. Cell Biol. 78:124, 2000. Plomp, J. J., Molenaar, P. C., O’Hanlon, G. M., et al.: Miller-Fisher anti-GQ1b antibodies: alpha-latrotoxin-like effects on motor end plates. Ann. Neurol. 45:189, 1999. Ponsford, S., Willison, H. J., Veitch, J., et al.: Long-term clinical and neurophysiological follow-up of patients with peripheral neuropathy associated with benign monoclonal gammopathy. Muscle Nerve 23:164, 2000. Posner, J. B.: Anti-Hu autoantibody associated sensory neuropathy/encephalomyelitis: a model of paraneoplastic syndrome. Perspect. Biol. Med. 38:167, 1995. Qin, Z., and Guan, Y.: Experimental polyneuropathy produced in guinea pigs immunized against sulfatide. Neuroreport 8:2867, 1997. Quarles, R. H., and Weiss, M. D.: Autoantibodies associated with peripheral neuropathy. Muscle Nerve 22:800, 1999. Quattrini, A., Corbo, M., Dhaliwal, S. K., et al: Antisulfatide antibodies in neurological disease: binding to rat dorsal root ganglia neurons. J. Neurol. Sci. 112:152, 1992. Renaud, S., Gregor, M., Fuhr, P., et al.: Rituximab in the treatment of polyneuropathy associated with anti-MAG antibodies. Muscle Nerve 27:611, 2003. Ritz, M. F., Lechner, S. J., Scott, R. J., et al.: Characterization of autoantibodies to peripheral myelin protein 22 in patients with hereditary and acquired neuropathies. J. Neuroimmunol. 104:155, 2000. Roberts, M., Willison, H. J., Vincent, A., and Newsom Davis, J.: Serum factor in Miller-Fisher variant of GuillainBarré syndrome and neurotransmitter release. Lancet 343:454, 1994. Roberts, M., Willison, H. J., Vincent, A., and Newsom Davis, J.: Multifocal motor neuropathy human sera block distal motor nerve conduction in mice. Ann. Neurol. 38:111, 1995. Ropper, A. H., and Gorson, K. C.: Neuropathies associated with paraproteinemia. N. Engl. J. Med. 338:1601, 1998. Rowland, L. P., Sherman, W. H., Hays, A. P., et al: Autopsyproven amyotrophic lateral sclerosis, Waldenström’s macroglobulinemia, and antibodies to sulfated glucuronic acid paragloboside. Neurology 45:827, 1995. Saida, T., Saida, K., Dorfman, S. H., et al.: Experimental allergic neuritis induced by sensitization with galactocerebroside. Science 204:1103, 1979. Salih, A. M., Nixon, N. B., Dawes, P. T., and Mattey, D. L.: Prevalence of antibodies to neurofilament polypeptides in patients with rheumatoid arthritis complicated by peripheral neuropathy. Clin. Exp. Rheumatol. 16:689, 1998. Schwerer, B.: Antibodies against gangliosides: a link between preceding infection and immunopathogenesis of Guillain-Barré syndrome. Microbes Infect. 4:373, 2002. Serrano-Munuera, C., Gallardo, E., Rojas, R., et al: Antiganglioside antibodies in acute self-limiting ataxic neuropathy: incidence and significance. J. Neuroimmunol. 120:78, 2001. Shams’ili, S., Grefkens, J., de Leeuw, B., et al.: Paraneoplastic cerebellar degeneration associated with antineuronal antibodies: analysis of 50 patients. Brain 126:1409, 2003.

2195

160. Sheikh, K. A., Deerinck, T. J., Ellisman, M. H., and Griffin, J. W.: The distribution of ganglioside-like moieties in peripheral nerves. Brain 122:449, 1999. 161. Siddiqui, K., Cahalane, E., Keogan, M., and Hardiman, O.: Chronic ataxic neuropathy with cold agglutinins: atypical phenotype and response to anti-CD20 antibodies. Neurology 61:1307, 2003. 162. Stubbs, E. B., Lawlor, M. W., Richards, M. P., et al.: Antineurofilament antibodies in neuropathy with monoclonal gammopathy of undetermined significance produce experimental motor nerve conduction block. Acta Neuropathol. (Berl.) 105:109, 2003. 163. Susuki, K., Nishimoto, Y., Yamada, M., et al.: Acute motor axonal neuropathy rabbit model: immune attack on nerve root axons. Ann. Neurol. 54:383, 2003. 164. Susuki, K., Yuki, N., and Hirata, K.: Features of sensory ataxic neuropathy associated with anti-GD1b IgM antibody. J. Neuroimmunol. 112:181, 2001. 165. Tagawa, Y., Yuki, N., and Hirata, K.: The 301 to 314 amino acid residue of beta-tubulin is not a target epitope for serum IgM antibodies in chronic inflammatory demyelinating polyneuropathy. J. Neurol. Sci. 163:44, 1999. 166. Takatsu, M., Hays, A. P., Latov, N., et al.: Immunofluorescence study of patients with neuropathy and IgM M proteins. Ann. Neurol. 18:173,1985. 167. Takigawa, T., Yasuda, H., Kikkawa, R., et al.: Antibodies against GM1 ganglioside affect K⫹ and Na⫹ currents in rat myelinated nerve fibers. Ann. Neurol. 37:436, 1995. 168. Tatum, A. H.: Experimental paraprotein neuropathy, demyelination by passive transfer of human IgM antimyelin-associated glycoprotein. Ann. Neurol. 33:502, 1993. 169. Teodorescu, M., and Froelich, C. J.: ELISA and other immunologic techniques in rheumatology. In Teodorescu, M., and Froelich, C. J. (eds.): Advanced Immunoassays in Rheumatology. Boca Raton, FL, CRC Press, p. 4, 1994. 170. Thomas, F. P., Trojaborg, W., Nagy, C., et al.: Experimental autoimmune neuropathy with anti-GM1 antibodies and immunoglobulin deposits at the nodes of Ranvier. Acta Neuropathol. (Berl.) 82:378, 1991. 171. Trojaborg, W., Hays, A. P., Van den Berg, L. H., et al.: Motor conduction parameters in neuropathies associated with anti-MAG antibodies and other types of demyelinating and axonal neuropathies. Muscle Nerve 18:730, 1995. 172. Turgeon, M. L.: Labeling techniques in immunoassays. In Turgeon, M. L. (ed.): Immunology & Serology in Laboratory Medicine. St. Louis, Mosby, p. 141, 1996. 173. Uncini, A., Santoro, M., Corbo, M., et al: Conduction abnormalities induced by sera of patients with multifocal motor neuropathy and anti-GM1 antibodies. Muscle Nerve 16:610, 1993. 174. Van den Berg, L. H., Hays, A. P., Nobile-Orazio, E., et al.: Anti-MAG and anti-SGPG antibodies in neuropathy. Muscle Nerve 19:637, 1996. 175. Van den Berg-Vos, R. M., Franssen, H., Wokke, J. H. J., and Van den Berg, L. H.: Multifocal motor neuropathy: longterm clinical and electrophysiological assessment of intravenous immunoglobulin maintenance treatment. Brain 125:1875, 2002. 176. Van den Berg-Vos, R. M., Franssen, H., Wokke, J. H. J., et al.: Multifocal motor neuropathy: diagnostic criteria that predict

2196

177.

178.

179.

180.

181.

182.

183.

184.

185.

186.

187.

Diseases of the Peripheral Nervous System

the response to immunoglobulin treatment. Ann. Neurol. 48:919, 2000. Van der Meche, F. G. A., Schmitz, P. I. M., and the Dutch Guillain-Barré Study Group: A randomized trial comparing intravenous immune globulin and plasma exchange in Guillain-Barré syndrome. N. Engl. J. Med. 326:1123, 1992. Van Sorge, N. M., Van den Berg, L. H., Geleijns, K., et al.: Anti-Gm1 IgG antibodies induce leukocyte effector functions via Fc␥ receptors. Ann. Neurol. 53:570, 2003. Visser, L. H., Van der Meche, F. G. A., Meulstee, J., et al.: Cytomegalovirus infection and Guillain-Barré syndrome: the clinical, electrophysiologic, and prognostic features. Neurology 47:668, 1996. Wardemann, H., Yurasov, S., Schaefer, A., et al.: Predominant autoantibody production by early human B cell precursors. Science 301:1374, 2003. Weber, F., Rudel, R., Aulkemeyer, P., and Brinkmeier, H.: Anti-GM1 antibodies can block neuronal voltage-gated sodium channels. Muscle Nerve 23:1414, 2000. Weiss, M. D., Dalakas, M. C., Lauter, C. J., et al.: Variability in the binding of anti-MAG and anti-SGPG antibodies to target antigens in demyelinating neuropathy and IgM paraproteinemia. J. Neuroimmunol. 95:174, 1999. Wicklein, E. M., Pfeiffer, G., Yuki, N., et al.: Prominent sensory ataxia in Guillain-Barré syndrome associated with IgG anti-GD1b antibody. J. Neurol. Sci. 151:227, 1997. Willison, H. J., Almemar, A., Veitch, J., and Thrush, D.: Acute ataxic neuropathy with cross-reactive antibodies to GD1b and GD3 gangliosides. Neurology 44:2395, 1994. Willison, H. J., and O’Hanlon, G. M.: The immunopathogenesis of Miller-Fisher syndrome. J. Neuroimmunol. 100:3, 1999. Willison, H. J., O’Hanlon, G. M., Paterson, G., et al.: A somatically mutated human antiganglioside IgM that induces experimental neuropathy in mice is encoded by the variable region heavy chain gene, V1-V8. J. Clin. Invest. 97:1155, 1996. Willison, H. J., O’Leary, C. P., Veitch, J., et al: The clinical and laboratory features of chronic ataxic neuropathy with anti-disialosyl IgM antibodies. Brain 124:1968, 2001.

188. Willison, H. J., Trapp, B. D., Bacher, J. D., et al.: Demyelination induced by intraneural injection of human antimyelin-associated glycoprotein antibodies. Muscle Nerve 11:1169, 1988. 189. Willison, H. J., and Yuki, N.: Peripheral neuropathies and anti-glycolipid antibodies. Brain 125:2591, 2002. 190. Wokke, H. J., and Notermans, N. C.: Clinical evaluation and differential diagnosis. In Latov, N., Wokke, J. H. J., and Kelly, J. J. (eds.): Immunology and Infectious Diseases of the Peripheral Nerves. Cambridge, UK, Cambridge University Press, p. 55, 1998. 191. Wolfe, G. I., El-Feky, W. H., Katz, J. S., et al.: Antibody panels in idiopathic polyneuropathy and motor neuron disease. Muscle Nerve 20:1275, 1997. 192. Yan, W. X., Archelos, J. J., Hartung, H. P., and Pollard, J. D.: P0 is a target antigen in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:286, 2001. 193. Yoshino, H., Harukawa, H., and Asano, A.: IgG antiganglioside antibodies in Guillain-Barré syndrome with bulbar palsy. J. Neuroimmunol. 105:195, 2000. 194. Yuki, N., Ang, C. W., Koga, M., et al.: Clinical features and response to treatment in Guillain-Barré syndrome associated with antibodies to GM1b ganglioside. Ann. Neurol. 47:314, 2000. 195. Yuki, N., Ho, T. W., Tagawa, Y., et al.: Autoantibodies to GM1b and GalNAc-GD1a: relationship to Campylobacter jejuni pinfection and acute motor axonal neuropathy in China. J. Neurol. Sci. 164:134, 1999. 196. Yuki, N., Kuwabara, S., Koga, M., and Hirata, K.: Acute motor axonal neuropathy and acute motor-sensory axonal neuropathy share a common immunological profile. J. Neurol. Sci. 168:121,1999. 197. Yuki, N., Takahashi, M., Tagawa, Y., et al.: Association of Campylobacter jejuni serotype with antiganglioside antibody in Guillain-Barré syndrome and Fisher’s syndrome. Ann. Neurol. 42:28, 1997. 198. Yuki, N., Yamada M., Koga, M., et al.: Animal model of axonal Guillain-Barré syndrome induced by sensitization with GM1 ganglioside. Ann. Neurol. 49:712, 2001.

98 The Guillain-Barré Syndromes JOHN W. GRIFFIN AND KAZIM SHEIKH

Overview History 1843–1916: Description of the Syndrome 1916–1969: Setting Boundaries 1969–Present: Mechanisms and Treatments Clinical Manifestations Acute Inflammatory Demyelinating Polyneuropathy

Axonal Forms Differential Diagnosis Pathology Acute Inflammatory Demyelinating Polyneuropathy AMAN and AMSAN Fisher Syndrome Pathogenesis Axonal Forms Fisher Syndrome

OVERVIEW The Guillain-Barré syndromes (GBSs), the most frequent cause of acute flaccid paralysis worldwide, are also among the most dramatic neurologic emergencies. They can reduce a vigorous athlete to paralysis within days or even hours. Yet they have the potential for substantial or complete recovery, and the time required for recovery can be shortened by immunotherapies of proved efficacy. Some cases follow infections with agents that express antigens similar to those on normal nerve fibers. For this reason molecular mimicry is an attractive pathogenetic mechanism. At present the goals of full understanding of the immunopathogenesis and of providing specific therapies remain only partially fulfilled, but the progress has stimulated new excitement in the field.

HISTORY The history of the GBSs divides into three periods. Between 1843 and 1916 the hypothesis that peripheral nerve disease could be a cause for acute flaccid paralysis was advanced and confirmed pathologically. The period from 1916 to 1969 was dominated by defining GBS and debating the boundaries of the diagnosis. The modern era began in 1969, with the identification of the key pathologic features, the consensus that the disorder is autoimmune in

Acute Inflammatory Demyelinating Polyneuropathy Management Monitoring and Education in Early Stages of the Disease Critical Care of the Guillain-Barré Patients Immunotherapy

nature, and the understanding that several syndromes fit within the designation GBS. This review adopts the plural term the Guillain-Barré syndromes when designating this group of heterogeneous but related syndromes.

1843–1916: Description of the Syndrome The recognition that peripheral nerve disease could lead to paralysis is attributable to a report by Robert Graves in 1843. Graves had visited Paris in 1828, during an epidemic outbreak of a neuropathy dominated by abrupt onset of pain and hyperalgesia in the extremities followed by generalized loss of sensibility and paralysis.50 The cause of this outbreak remains unknown; the prominence of pain and hyperalgesia may be more suggestive of a toxic neuropathy than of any form of GBS. In any event, the pathologists identified no abnormalities in “brain, cerebellum, or spinal marrow.” Graves concluded, Here is a paralysis affecting all parts of the extremities as completely as if it had its origin in the central parts of the nervous system; and can any one, with such palpable evidence before him, hesitate to believe that paralysis or even hemiplegia, without any lesion of the brain or spinal cord, might arise from disease commencing and originating in the nervous extremities alone?50 2197

2198

Diseases of the Peripheral Nervous System

The argument that a paralytic disease without central lesions might be due to peripheral nerve involvement was extended by Jean Baptiste Octave Landry de Thezillat, an extravagantly talented young clinician in Paris. Under the title “Note sur Paralysie Ascendante Aiguë,” Landry in 1859 described five personal and five literature cases of “ascending paralysis.”109 He recognized that affected patients could develop generalized paralysis, including respiratory failure, within a 15-day period, and noted that paralysis could appear “sometimes in only a few hours,” as translated here: [T]he main problem is usually a motor disorder characterized by a gradual diminution of muscular strength with flaccid limbs and without contractures, convulsions or reflex movements of any kind. In almost all cases micturition and defecation remain normal. One does not observe any symptoms referable to the central nervous system, spinal pain or tenderness, headache or delirium. The intellectual faculties are preserved until the end. The onset of the paralysis can be preceded by a general feeling of weakness, pins and needles and even slight cramps. Alternatively the illness may begin suddenly and end unexpectedly. In both cases the weakness spreads rapidly from the lower to the upper parts of the body with a universal tendency to become generalised. . . . The progression can be more or less rapid. It was eight days in one and fifteen days in another case which I believe can be classified as acute. More often it is scarcely two or three days and sometimes only a few hours.109 Landry documented an ascending pattern of involvement, and noted the preservation of bowel and bladder function and the absence of bedsores. Two of these 10 cases were fatal. No abnormalities were found in the central nervous system. From this he concluded that the disease must be a peripheral neuropathy. The pathologic recognition of nerve fiber degeneration was dependent on Waller’s recognition in 1852 of the histologic changes after nerve transection that bear his name.191 In 1864, Duménil in Rouen reported the first pathologic demonstration of nerve disease in a paralyzed patient.28 This 61-year-old patient developed prickling paresthesias in the toes, followed within 2 weeks by rapidly evolving weakness with some asymmetry. This spread to complete paralysis and death after 41/2 months. An autopsy of the brain and spinal cord was normal, but there was extensive loss of nerve fibers with features consistent with advanced wallerian degeneration, including lipid-laden cells and excessive connective tissue within the nerves. The first description of lymphocytic inflammation in peripheral nerve was made by Eichhorst in Berlin in 1877,30 but the description of the histology is most suggestive of a severe vasculitic neuropathy. A more likely first autopsy description of inflammation in GBS

was that by Leyden in Germany in 1880.112 A 31-year-old merchant developed paresthesias of the feet followed by rapidly evolving paralysis, so that the legs were nearly completely paralyzed and the hands were weak. There was loss of tendon reflexes and sensory loss. He died with evidence of extensive degeneration in the nerves, and with perivascular and subperineurial inflammatory infiltrates. In the early 1890s, Osler described a series of cases of paralysis, usually ascending, leading to respiratory involvement and muscle wasting, but with potential for recovery.145 He used the term acute febrile polyneuritis, but emphasized the clinical identity with Landry’s cases. In 1893, the status of peripheral neuritis was reviewed in a little-recognized work by James Ross and Judson Bury of Manchester.168 They summarized 94 cases of acute ascending paralysis, the majority of which were almost certainly GBS. The clinical stages were classified into “the stage of invasion,” dominated by paresthesias and minor sensory symptoms; the stage of “advancing paralysis”; and the convalescent phase of “regressing paralysis.” Notably missing from the clinical descriptions to this time was the recognition of antecedent infectious illnesses. In their review of 38 autopsied cases, 24 were said to show “no neurologic involvement,” and abnormalities in peripheral nerves were identified in only 6 cases. Ross and Bury do not mention either inflammation in nerve or demyelination, even though by that time several reports had described inflammatory infiltrates within the nerves, and demyelination had been described by Gombault in experimental lead neuropathy 20 years earlier.45

1916–1969: Setting Boundaries During World War I two reports ushered in the second era. In 1916 Georges Guillain, Jean-Alexandre Barré, and Andre Strohl described two paralyzed French soldiers.59 Their novel contribution was the demonstration of elevated spinal fluid protein without cells. In 1917 Gordon Holmes described 12 cases of rapidly progressive paralysis including bulbar involvement, absent tendon reflexes, and recovery.82 Many of these patients, like Osler’s, were febrile, and all were seen over a few-month period in London. The nature of these cases has remained controversial, but the clinical descriptions are consistent with GBS. This same period coincided with the description of the first large epidemic of poliomyelitis, and the similarities and differences in the disorders make a fascinating comparison. The polio outbreak in 1916 in New York City affected some 20,000 individuals, and initiated the phase of public concern about the summertime epidemics. Both disorders can follow enteric infection: polio is due to infection with the poliomyelitis virus, and GBS can follow infection with Campylobacter jejuni and other agents. Because polio is a direct viral infection of motor neurons, whereas GBS

2199

The Guillain-Barré Syndromes

is immune mediated, successful vaccination has nearly eradicated polio worldwide, whereas treatment for the immune abnormalities that are like GBS has to date only been partially effective. During the succeeding 50 years the precise definition and syndromic boundaries of the disorder and the clinical distinctions from diphtheritic neuropathy, polio, toxic neuropathies, and other causes of acute flaccid paralysis were debated. This period saw a number of attempts to define GBS clinically. Based only on clinical findings, no scheme can be satisfactory, because similar clinical manifestations can be produced by a wide variety of other disorders. The advent of electrodiagnosis and the ability to recognize demyelinating physiology in the 1960s represented a landmark in diagnosis of GBS. The importance of demyelination in GBS was identified definitively by Haymaker and Kernohan in their survey of nerve tissue from the Armed Forces Institute of Pathology.71

1969–Present: Mechanisms and Treatments The last 35 years have been dominated by detailed pathologic and physiologic characterization and by exploration of mechanisms, recognition of the heterogeneity of GBS, and the development of immunotherapies. This era began with a pivotal pathologic report of 19 autopsies by Asbury and colleagues7 that underscored the lymphocytic inflammation in GBS and pointed out the analogy between GBS and experimental allergic neuritis (EAN) in animals. At that time EAN had recently been discovered. It can be induced by immunization of appropriate experimental animals with the whole peripheral nerve myelin and complete Freund’s adjuvant. The pathology typically includes lymphocytic inflammation and demyelination, resembling the findings in acute inflammatory demyelinating polyneuropathy. Ultrastructural studies underlined the similarities in macrophage-mediated demyelination in human demyelinating neuropathies and in EAN.156 On these bases, GBS was widely presumed to represent a human counterpart to EAN. Corticosteroids used alone had proved to be without value in the treatment of GBS. In the mid-1980s, the then new technique of plasmapheresis provided the first effective immunotherapy.60 Subsequent trials of infusion of intravenous immune globulin produced identical results, and made available a second form of effective immunotherapy. An increasingly prominent theme has been the likelihood that molecular mimicry is a mechanism underlying at least some cases. Yuki and colleagues first demonstrated the similarities in antigens within infectious organisms cultured from GBS patients and normal constituents of nerve.210 The best established examples of such molecular mimicry are in post-Campylobacter GBS, in which ganglioside-like moieties in the Campylobacter lipo-oligosaccharide are

mimics of normal gangliosides of peripheral nerve.210 At the present time, although molecular mimicry is not formally proved in GBS, it represents the most likely mechanism of production of the disorder in postinfectious cases. Indeed, post-Campylobacter GBS is one of the best studied models in which postinfectious molecular mimicry is likely to operate. Another theme in the modern era of GBS research is the heterogeneity of the syndrome. The most prevalent form of GBS in the United States, Europe, and much of the rest of the world is AIDP. It may account for as much as 97% of GBS cases in England and North America,63,64 and much of the clinical and epidemiologic writing of the past 20 years applied specifically to this form. The Fisher syndrome (FS) is defined by internal and external ophthalmoplegia, gait ataxia, and areflexia. Some cases have these findings in pure form and never develop weakness.37 Finally, one of the lessons of recent years has been that some forms of the GBSs lack features of demyelination, and fit best with acute axonal injuries. These cases are termed axonal GBS. Table 98–1 presents a provisional classification of the GBSs. Undoubtedly other variants will be added, particularly as specific immune markers are recognized. There are some patients with elements of more than one of these GBS classes, and some transition from one to the other. For example, some patients begin with FS that then evolves into typical AIDP.

CLINICAL MANIFESTATIONS Acute Inflammatory Demyelinating Polyneuropathy Ross and Bury’s division into a “stage of invasion” characterized by progressing paralysis, a “plateau phase,” and a “stage of regression” with lessening paralysis provides a convenient format for discussion.168 The “Stage of Invasion” The initial manifestations are often sensory and may include prickling paresthesias or typical “dead asleep” feelings. These frequently begin in the feet, but occasionally start in the hands or on the face. About 50% of patients198 experience a dull aching in the lumbar region throughout the back or in the neck. In children, this may be associated with neck stiffness and resistance to bending the neck.68 Often the patients dismiss the initial symptoms as a consequence of unusual recent exercise. This initial phase before weakness develops may last from hours to more than 1 week. Even at an early stage the muscle stretch reflexes are usually lost or diminished. Loss of the stretch reflexes when strength is preserved presumably reflects afferent involvement. Loss of reflexes is a sufficiently regular feature that the

2200

Diseases of the Peripheral Nervous System

Table 98–1. The Guillain-Barré Syndromes Syndrome

Pathology

Antecedent illness

Putative Antigen(s)

Acute inflammatory demyelinating polyneuropathy (AIDP)

Demyelination with variable axonal degeneration and lymphocytic inflammation No data

Herpesviruses, Campylobacter jejuni, Mycoplasma pneumoniae

Unknown

Occasionally Campylobacter jejuni; most cases have unknown antecedent

GQ1b and related gangliosides

Campylobacter jejuni

GD1a, GM1, GalNAc, GD1a, GM1b

Campylobacter jejuni

Probably gangliosides

Fisher syndrome (FS)

Axonal forms Acute motor axonal neuropathy (AMAN) Acute motor-sensory axonal neuropathy

Motor axonal degeneration with little inflammation or demyelination Similar, with sensory involvement

diagnosis of GBS is suspect in patients with weakness but preserved brisk reflexes. Abnormalities of special sensation (vision, vestibular function, and hearing) are rare, although optic neuropathy can occur.74,131 Both sensory findings and weakness are usually symmetrical. When weakness begins, it can involve virtually any muscles. Landry recognized that some cases describe an “ascending paralysis,”109 but in many patients weakness from the onset involves proximal and even axial muscles. Respiratory insufficiency is an especially ominous manifestation of proximal muscle weakness. This proximal involvement presumably correlates with the involvement of the spinal nerve roots pathologically (see later). In a small proportion of patients, the facial or bulbar muscles are almost exclusively affected. Involvement of the extraocular muscles occurs in less than 10% of patients with otherwise typical AIDP. When such patients with virtually complete ophthalmoplegia are asked to attempt to close their eyes, a vigorous Bell’s phenomenon is often elicited. Although this has raised the possibility that there might be a central cause for the ophthalmoplegia, it is likely that this represents a change in eye position related to minor changes in intraorbital pressures, rather than function of oculomotor nerves.91 The single most helpful diagnostic approach is electrodiagnosis. Electrodiagnostic data exclude many alternative diagnoses (myopathy, myasthenia, botulism), identify the demyelination characteristic of AIDP, and inform prognostic projections.24 Several lists of diagnostic criteria have been published (see Cummins and Waxman25). They share emphasis on identifying elements of demyelination, such as prolonged terminal latencies, prolonged F-wave latencies, and reduced nerve conduction velocities. Pathologically proved AIDP with nerve terminal demyelination can produce complete loss of elicitable compound muscle action potential (CMAP) and sensory nerve action potential (SNAP) responses.119,162

Progression and Plateau Phase The rate of evolution varies, but in over 90% of patients the nadir is reached within 1 month.159 Progression for more than a month suggests alternative diagnoses, such as a subacute or chronic inflammatory demyelinating polyneuropathy.177,198 The severity of involvement varies widely. Occasional patients have only mild weakness (e.g., in wrist extension) that begins to improve within days, associated with modest reduction in the tendon reflexes and evidence of demyelination on electrodiagnostic studies. Such patients comprise the mildest forms of AIDP. Conversely, AIDP can be a life-threatening neurologic emergency with weakness evolving from initial symptoms to respirator dependency within hours. Although this sequence is the exception, it is the reason that hospitalization is almost always required at the time that the initial diagnosis is made. In rare cases the paralysis is so severe that all voluntary movements cease and patients can be mistaken for brain-dead until electroencephalographic studies are done.41,179,188 Respiratory insufficiency requiring ventilation develops in 25% to 43% of cases.111,159 Autonomic dysfunction is a frequently overlooked and important aspect of the disease.114 Approximately two thirds of individuals have autonomic dysfunction, which is frequently manifest as a sinus tachycardia at rest. Bradycardia can occur. There is loss of heart period variability in a high proportion of cases. Orthostatic hypotension is also frequent and occurs at least in part on a neurogenic basis, with loss of sympathetically mediated vascular resistance. Autonomic dysfunction may contribute to mortality in the disease.114 Inappropriate secretion of antidiuretic hormone is a well-recognized complication of GBS as well,22,80,81 and can lead to symptomatic hyponatremia. Regression Phase In most patients, the cessation of progression initiates a plateau phase that blends into a phase of regression of symptoms. In most patients, progression of all manifestations ceases at the same time, but occasionally patients can

2201

The Guillain-Barré Syndromes

begin to improve in some areas, such as strength in the legs, at a time when other functions are in decline. The rate of improvement is highly variable. In the most favorable settings patients can go from virtually complete paralysis and respirator dependency to walking unassisted in a week. Prospective studies have identified several prognostic predictors.24,111,122,123,159 The most important are age, rate of progression, severity of weakness at the nadir, and whether effective immunotherapy was utilized. The mortality of GBS used to be as high as 15% and as late as the 1980s, a mortality of 13% was described. With modern critical care, mortality has been reduced to 1.25%167 to 2.5%.60,123 Even with immunotherapy, however, recovery may be measured in months and be profoundly incomplete. Approximately 15% of GBS patients have functionally important residua,38 and many more have abnormalities detectable for years on examination.

Axonal Forms Axonal GBS is uncommon in the United States and Europe31,32 but more frequently recognized in China, Japan, Mexico, Korea, and India.62,78,120,121,157,215,216 In retrospect, the first report was probably published by Ramos-Alvarez and colleagues in 1969.157 These authors had been involved with the early oral (Sabin) polio vaccine program in Mexico City. In spite of what they believed was successful immunization, a number of children continued to die with acute flaccid paralysis. The pathology showed that these cases did not have polio. Some had pathology typical of AIDP, but in others there was no inflammation or obvious demyelination in the nerves. The investigators’ attention was directed to changes in the nerve cell bodies. In many cases the motor nerve cell bodies were enlarged, with a pale-staining cytoplasm and an eccentrically placed nucleus. They termed such cases cytoplasmic neuronopathy. As noted later, recent review of the pathology suggests that most had axonal GBS. In 1986, Feasby and colleagues described a patient in whom the electrophysiology and pathology were interpreted to suggest a primary axonal disorder, reflecting an immunemediated attack on the axon.34,35 This suggestion was initially controversial because a degree of axonal degeneration is universal in severe cases of GBS, and axon degeneration in EAN could be induced by immunization with high doses of myelin antigens.69 Such cases are now termed acute motorsensory axonal neuropathy (AMSAN). Studies in China of summertime epidemics of GBS confirmed that primary axonal forms could occur.120 Acute motor axonal neuropathy (AMAN) is more frequent than AMSAN. Retrospective review of the pathologic material from Mexico City reported by Ramos-Alvarez and colleagues157 in 1969 suggests that “cytoplasmic neuronopathy” has the pathology of AMAN, a syndrome described 25 years later. Restaining these cases brought out the degree of wallerian-like degeneration of ventral roots and motor nerve fibers.

The axonal forms of GBS cannot be differentiated from AIDP without electrophysiology or extensive pathologic data. AMAN has exclusively motor findings, with weakness typically beginning in the legs, but in some individuals affecting arms or cranial muscles initially.120 Although sensory findings are absent, some affected children have resistance to neck flexion.121 Reflecting the lack of afferent involvement, tendon reflexes are preserved until weakness is severe enough to preclude phasic muscle contraction. Autonomic function has not been studied. The frequency of respiratory insufficiency is not defined, but appears to be at least equal to that in AIDP. In AMAN the electrodiagnostic data are characterized by loss of CMAP amplitudes, preserved SNAP amplitudes, early evidence of denervation on electromyography, and no evidence of demyelination in the terminal latencies, F waves, or conduction velocities.120 The spinal fluid can have elevated protein without cells, as in AIDP. The prognosis varies widely, as would be predicted from the range of pathologic severities described later, but overall the prognosis is similar to that of AIDP.76 AMSAN is clinically and physiologically similar, but with detectable sensory involvement. Some cases have inexcitable nerves, and were pathologically proved to be axonal in nature, with no evidence of antecedent demyelination. However, electrically inexcitable nerves can occur in severe demyelinating neuropathies,119,162 so that inexcitability is not diagnostic of AMSAN. The prognosis in AMSAN was considered, on the basis of the first reported cases, to be unusually poor,34,35 but subsequent data have suggested that some cases have a more favorable outlook.76

DIFFERENTIAL DIAGNOSIS The differential diagnosis of acute flaccid paralysis varies in different parts of the world. Historically, poliomyelitis was the major cause of acute flaccid paralysis, and the significance of the recognition of elevated spinal fluid protein without cells by Guillain, Barré, and Strohl59 lay in the fact that spinal fluid pleocytosis was the rule in polio. Even in parts of the world where polio has been virtually eradicated, poliomyelitis syndromes associated with other viruses, including Echo 70 and Coxsackie strains, continue to occur. An important new cause of viral poliomyelitis is the West Nile virus. Some of the first reported cases of West Nile encephalitis in the United States were initially diagnosed as axonal GBS.1,92 The electrodiagnostic studies are similar to AMAN syndrome, so that West Nile poliomyelitis must be in the differential of AMAN in endemic areas. Even though West Nile virus often produces an encephalitis, the neuromuscular manifestations can dominate the clinical picture. A rare but dramatic mimic of GBS is tick paralysis. Children are more frequently affected than adults, so that

2202

Diseases of the Peripheral Nervous System

this is a major differential possibility in childhood GBS. Paralysis is produced by a toxin produced within the tick. The molecular identity and pathophysiologic affects of this toxin are not fully understood, but it is likely that the effects occur at the neuromuscular junction or along the preterminal axon. Typically in the United States, removal of the tick is associated with improvement in strength over a period of hours. However, progression can continue for a period after tick removal, especially in Australia, where it is likely that a different toxin is more frequently encountered.49 The tick is often embedded in the scalp, and careful inspection and palpation may be needed to identify a deeply embedded tick. Paralytic rabies can mimic GBS. It has long been known that some cases of rabies produce a picture of acute flaccid paralysis that may precede the more typical encephalitic manifestations. Sheikh examined the spinal cord and peripheral nerves of a patient who had been bitten by a rabid dog years earlier, and then later developed rapidly progressive paralysis and death (K. A. Sheikh, personal communication). The initial clinical and pathologic diagnosis was AMSAN, and this appeared to be supported by the absence of inflammation in the spinal cord or nerves, and motor neuron death. However, there was abundant virus within motor neurons as well as a population of sensory neurons to support the mechanism for viral latency over long periods in such cases. The explosive onset of paralysis is not understood, but it is possible that the immune system drives this process in part. Botulism can also produce a rapidly evolving paralysis. In this case, the pathophysiology is very well understood; the toxin prevents release of acetylcholine from motor nerve terminals as a result of defects in docking and vesicle release (the specific molecular targets differ among different species of botulinum toxin). The clinical manifestations typically include internal and external ophthalmoplegia, bulbar paralysis, dry mouth, constipation, and postural hypotension. Sensory symptoms are absent. The diagnosis can be strongly suspected from demonstration of an incremental response with rapid repetitive nerve stimulation. Although most adult cases result from ingestion of preformed organisms in contaminated canned foods and similar food-borne sources, adults and especially infants can develop botulism from ingestion of spores and formation of toxin from within the gut. In such patients, the organism Clostridium botulinum may be cultured from the stool. Finally, wound botulinum caused by contamination of a wound with this anaerobic organism can produce full-blown botulism or more localized forms of the disease. Myasthenia gravis is rarely a difficult differential, but acute spinal cord and brainstem disorders can produce confusing pictures. Transverse myelitis, spinal cord compression, and vascular transverse myelopathies can begin abruptly and initiate a period of spinal shock in which tendon reflexes are lost. Electrodiagnosis is helpful in such patients because at least some abnormalities are expected in most cases of GBS.

Infiltration of the meninges by carcinomas or lymphoma cells can produce rapidly evolving polyradiculopathy suggestive of GBS. Electrodiagnostic data will not show reduction in nerve conduction velocity or prolongation of distal latencies, but may have prolonged F-wave latencies. Because the number of cells recovered in the spinal fluid may be not be elevated, repeated high-volume lumbar punctures with cytology may be necessary to establish the diagnosis.

PATHOLOGY Research into the mechanistic aspects of pathogenesis has been furthered by the pathologic observations made in human materials obtained from GBS cases. The classic descriptions of Asbury and colleagues7 established the pathologic hallmarks of AIDP, and more recent studies35,54–56 have characterized the key pathologic features seen in axonal variants of GBS.

Acute Inflammatory Demyelinating Polyneuropathy Lymphocytic inflammation, demyelination, variable degrees of axonal loss, and endoneurial edema are the major elements of the pathology of AIDP. The areas of most severe involvement are usually the spinal roots, including the mixed spinal roots, and the nerve terminals. These findings correlate with the prolonged F waves and prolonged distal latencies that are often found in motor nerve conduction studies. This distribution may reflect the relative deficiency of the blood-tissue barriers in these regions. Inflammation and demyelination can occur anywhere along the peripheral nerves. This may be facilitated by breakdown of the bloodnerve barrier. GBS sera have the capacity to alter bloodnerve barrier integrity.93 Inflammation Lymphocytes appear early in the disease, and can persist long after the active disease appears to have ceased.7 Cuffs around epineurial vessels are nonspecific, but can be seen in AIDP. More helpful diagnostically are lymphocytes in the endoneurial space, either in perivenular cuffs or scattered throughout the endoneurium. The lymphocytes may be numerous and concentrated in areas of demyelination, but often they are relatively sparse. The ratio of CD4 to CD8 lymphocytes varies, and can be affected by the ratio in the blood.23 Demyelination The extent of demyelination is best assessed by teased nerve fiber studies. Demyelination can affect nearly all fibers in a region of nerve,119 producing large plaques (Fig. 98–1). More often there is scattered involvement of fibers. Along any single fiber only one or a few internodes may be demyelinated. In some cases the demyelination

FIGURE 98–1 Pathology of acute inflammatory demyelinating polyneuropathy (AIDP). A, Transverse section of a lumbar ventral root from a fatal case of AIDP. There is a large demyelinated plaque in the center of the rootlet, with admixed demyelinated and intact fibers scattered throughout the remainder of the rootlet. The inset shows three demyelinated axons, one (identified by arrow) surrounded by myelin debris. Bar: 50 ␮m. B, A longitudinal section shows a central plaquelike region of demyelinated fibers, with normally myelinated fibers above and below. Bar: 20 ␮m. C, Mononuclear inflammatory infiltrate is seen in this ventral rootlet. Bar: 50 ␮m. D, Around the vessel to the lower left there is endoneurial edema and scattered mononuclear cells. The boxed fiber has a myelinated internode next to a wholly demyelinated internode. Bar: 50 ␮m. E, The boxed fiber in D, shown at higher power. Bar: 20 ␮m. (From Griffin, J. W., and Sheikh, K. A.: The Guillain Barre syndrome. In Lazzarini, R., Griffin, J. W., Lassman, H., et al. [eds.]: Myelin Biology and Disorders. Amsterdam, Elsevier, p. 887, 2004, with permission.)

2204

Diseases of the Peripheral Nervous System

is coherent, so that every affected fiber appears to be at about the same stage of demyelination or repair (Fig. 98–2; see also Fig. 98–1). However, in many cases remyelination can be seen adjacent to fibers in early stages of demyelination. Several studies have examined the ultrastructural aspects of demyelination, and the sequence has been filled in by extrapolating with data from studies of EAN. The studies of GBS suggest at least two patterns of demyelination. In the first pattern macrophages penetrate the Schwann cell basal lamina of apparently normal nerve fibers and beginning to “strip” myelin lamellae.83,155 These macrophages may enter the nerve fiber near the node or at any point along the internode.155 Several macrophages may attack individual internodes in this fashion. The second pattern involves vesicular myelin breakdown, in which the myelin sheath develops the appearance of “soap bubbles”67 (Figs. 98–3 and 98–4). In such cases macrophages enter affected fibers to remove obviously damaged myelin. As noted later, this pattern can correlate with the presence of immunoglobulin and complement activation products on the outside of the Schwann cell (see Fig. 98–4). The Schwann cell responds to myelin injury by amputating and separating from its myelin sheath, entering the cell cycle and dividing, and leaving the basal lamina of the fiber to produce a circlet of daughter Schwann cells around the demyelinated fiber. As myelin is cleared, several daughters reenter the basal lamina of the fiber to ensheath short segments of the fiber and begin production of several short internodes within the previous internode. These short internodes distributed between original long internodes provide a pathologic hallmark of earlier demyelination with remyelination. Axonal Degeneration The severity of axonal degeneration in AIDP varies widely.7 The factors that determine the extent of fiber loss

are not understood, but may include the severity of inflammation and the degree of edema in sites that cannot expand, such as the junction of ventral and dorsal roots with the mixed spinal nerves, where the perineurium is formed.13

AMAN and AMSAN Both axonal forms of GBS differ from AIDP in the paucity of lymphocytic inflammation and of demyelination.35,56,120 Wallerian-like degeneration dominates.54,120 AMAN and AMSAN differ from each other in the extent of sensory fiber involvement. In AMAN the cutaneous sensory nerves are normal or show only minimal wallerian-like degeneration, whereas in AMSAN sensory fiber degeneration can be extensive.54,55,77,120,202 In other regards their pathology is similar. In the most severe cases of AMAN, wallerian-like degeneration involves the ventral root exit zone and the motor fibers throughout their course in the roots and nerves.54 The motor neurons and the intraparenchymal motor fibers survive, so that the pathology resembles an immunologic axotomy at the level of the ventral root exit zone. The motor neuron cell bodies undergo prominent chromatolytic changes. This change may be the finding that was called “cytoplasmic neuronopathy” by Ramos-Alvarez and colleagues in their fatal cases of childhood flaccid paralysis in Mexico,157 described earlier. In some fatal cases of AMAN and AMSAN the pathology is much more limited.54,76,77 This observation suggests that conduction failure may occur in the absence of extensive wallerian-like degeneration of axons. This conduction failure could be due to conduction block that results from binding of antibody at nodes and consequent nodal changes; from opening of the periaxonal space, reducing the capacitance produced by “sealed” myelin; or by wallerian degeneration restricted to the terminal axons, a mechanism that has been supported by pathology in one reported case.77

FIGURE 98–2 Electron micrographs of demyelinating and demyelinated nerve fibers in spinal roots in AIDP. Several axons (A) are demyelinated in both panels A and B, while one is in the process of demyelination in A. M ⫽ macrophages.

2205

The Guillain-Barré Syndromes

FIGURE 98–3 Electron micrographs of vesicular demyelination and macrophage-mediated myelin removal in AIDP. A, This fiber has multiple macrophages involved in myelin phagocytosis. B, At higher magnification the macrophages (M) are beneath the basal lamina of the Schwann cell. The myelin throughout this internode shows vesicular change. A ⫽ axon. The inset is a transverse section of a fiber undergoing vesicular demyelination.

Fisher Syndrome The pathology of FS is incompletely defined, because in pure form it is not fatal. Some cases that have gone from ophthalmoparesis and ataxia to paralysis have had the pathologic features of AIDP at autopsy, and such cases also have evidence of demyelination by electrodiagnostic testing. An intriguing disorder that is closely related to FS is Bickerstaff’s brainstem encephalitis. Typically these cases have ophthalmoparesis associated with gadolinium-

enhancing posterior fossa lesions on magnetic resonance imaging. They can follow C. jejuni or Mycoplasma infection, and have anti-GQ1b antibodies (see later).139,209

PATHOGENESIS Postinfectious molecular mimicry has emerged as the foremost putative mechanism underlying the GBSs.144,176,197,210 Postinfectious molecular mimicry in the peripheral nervous

2206

Diseases of the Peripheral Nervous System

system (PNS) implies shared antigenic determinants between the infectious agents and peripheral nerve fibers. Antigenic stimulation by the organism results in an immune response that produces immune-mediated damage to nerve fibers. The support for this concept derives largely from both clinical and experimental studies on the AMAN and FS variants of GBS following Campylobacter infection. During the last 15 years clinical and experimental work has made significant progress in characterizing and understanding the immune effector mechanisms implicated in nerve damage during the acute phase of the illness. In contrast, how and why an immune response that injures self is generated by inciting events or infections that usually produce a benign response is understood only at a rudimentary level. Why do only a small fraction of individuals with C. jejuni infection develop GBS? Why is tolerance to self broken in these individuals? Why does the immune attack usually stop spontaneously to produce the characteristic monophasic, self-limited disease? These issues are only partly understood, but it is likely that both host factors and organism factors are involved.

Axonal Forms The mechanisms underlying axonal forms of GBS and FS are better understood than those of AIDP. The pivotal features are an autoimmune attack on nerve gangliosides, initiated by an immune response to ganglioside-like antigens in infectious organisms such as C. jejuni.

FIGURE 98–4 Complement activation and early vesicular demyelination in AIDP. A, Immunocytochemical demonstration of a marker of complement activation, C3d, on the surface of two nerve fibers (arrows). B, At the electron microscopic level the fiber identified by the arrow in A has early vesicular demyelination. Arrows identify two amputated lamellae. The barrel arrow identifies the outside of the Schwann cell. Note the vesicles scattered in the Schwann cell cytoplasm. C, Electron micrograph of a fiber undergoing advanced vesicular demyelination. Note the two macrophages (M) within the Schwann cell basal lamina. One of the macrophages has extended processes beneath the basal lamina (arrows). A ⫽ axon.

Antecedent Infection with Campylobacter Campylobacter jejuni infection is the most frequently recognized antecedent of the axonal forms, and can precede FS and AIDP as well. Campylobacter jejuni, a microaerophilic, gram-negative, non–spore-forming, nonflagellate rod, is the most prevalent cause of bacterial gastroenteritis worldwide, especially in children. It is a common commensal in birds, particularly chickens. The most important route of Campylobacter infection in the United States and other industrialized nations is the consumption and handling of chicken.3 Infection with C. jejuni typically results in an acute, self-limited gastroenteritis characterized by watery or bloody diarrhea, sometimes accompanied by fever and abdominal cramps. Infection can be asymptomatic. Bacteremia is rare and most likely to occur in immunocompromised individuals and children.3 In the context of GBS, the diagnosis of antecedent Campylobacter infection is usually made by serologic studies, because by the time neurologic symptoms are present the organism has largely been eliminated from the stool culture.125,163 Organism-related factors that increase the risk of GBS have been examined by two approaches. The first approach has been characterization of the cross-reactivity with neural antigens of the lipopolysaccharide (LPS) or lipo-oligosaccharide (LOS) from the isolates. The second

The Guillain-Barré Syndromes

has been typing the GBS and enteritis isolates. Classification schemes for serotyping are based on antigens, either heat stable (Penner serotyping) or heat labile (Lior serotyping), or on DNA markers.101,127,175 Penner serotyping, the most frequently used scheme, is based on differences in the saccharide structure of the heat-stable bacterial antigens, alternatively termed “O” or heat-stable (HS),127 and here called HS/O. HS/O antigens consist of bacterial cell wall LPS/LOS and/or capsular polysaccharide (CPS). Several C. jejuni serotypes have been associated with GBS. Some GBS-associated strains are uncommon in patients with uncomplicated gastroenteritis. HS/O:19 is highly overrepresented in Japanese GBS patients (up to 81%) compared to diarrhea isolates (~2%).42,101,207 HS/O:19 is also a common isolate among northern Chinese patients with GBS.129 This serostrain has also been isolated from GBS patients in Ireland and the United States,125 but HS/O:19 may not be overrepresented in all populations.89,161 In South Africa C. jejuni serostrain HS/O:41, an uncommon cause of diarrhea, was isolated from GBS patients.110 The overrepresentation of serostrains HS/O:19 and HS/O:41 in patients with GBS supports the notion that organism-related properties are critical in determining whether or not GBS follows C. jejuni infection. GBS-associated isolates are not restricted to these two rare serostrains (reviewed by Prendergast and Moran153). The LPS has three components: lipid A, the hydrophobic region within the membrane; an oligosaccharide core consisting of 10 to 15 sugars with inner and outer components; and O chains, which are externally directed polysaccharides. LOS contains only lipid A and oligosaccharide cores without O chains. In certain strains of C. jejuni the outer portion of core oligosaccharide shares structural similarities with human gangliosides.11,128 It is now believed that LPS/LOS alone or CPS alone contributes to the serospecificity determinant of some strains of C. jejuni, whereas in others both LPS/LOS and CPS are responsible for the heat-stable determinants.127 Biochemical, mass-spectroscopic, and solid-phase immune assays with antiganglioside antibodies and bacterial toxins and lectins have characterized the core regions of LPS/LOS of GBS- and diarrhea-associated C. jejuni strains. Moran and co-workers described the presence of sialic acid in C. jejuni strains in 1991.128 Yuki and co-workers first described the presence of a GM1-like structure in C. jejuni LPS isolated from a patient with AMAN,208 and subsequent studies have demonstrated the presence of GM1-, GD1a-, GalNAc-GD1a-, GM1b-, GT1a-, GD2-, GD3-, and GM2-like structures in different GBS isolates.5,8–11,128,130,137,152,176 GQ1b–cross-reactive gangliosides such as GT1a and GD3 are present in C. jejuni isolates from FS patients,171 although structural studies have so far failed to show the presence of GQ1b-like moieties. Both GBS- and diarrhea-associated isolates carry ganglioside-like moieties, but GBS-related organisms are more likely to do so.129,130 Furthermore,

2207

different C. jejuni isolates within the same serotype can have different ganglioside-like moieties,130,176 and a single isolate can have multiple ganglioside-like moieties in its LPS.8–11,58 Immunopathology Immunopathologic studies on Chinese post-Campylobacter/ AMAN cases support the role of antibody-mediated injury.66 The immunopathology is dominated by binding of immunoglobulin G (IgG) and complement activation products to axons.66 In AMAN this is exclusively on motor fibers,54,120 and in mild AMSAN it is predominantly on motor fibers.55 Such deposits can be predominantly on nodes, but paranodal and internodal axolemma can be affected as well.66 This finding suggests that the periaxonal space has been opened to entry of endoneurial fluid, presumably by disruption of the myelin sheath attachment sites.56 Macrophages typically approach and surround the IgG-positive nodes.56,66 In some fibers they extend processes beneath the myelin sheath attachment sites and enter the periaxonal space54,56,66 (Fig. 98–5). In this site they typically do not disrupt the myelin, but there is swelling and disappearance of the abaxonal Schwann cell cytoplasm.56 The axon appears condensed, and is often completely surrounded by macrophage processes56 (Fig. 98–5D). Such a sequence suggests that the axon must be capable of surviving for a period with complement activation on the axolemma and surrounded by periaxonal macrophages, but at some point typical wallerian degeneration ensues. Serologic Studies and Identification of Target Antigens Four gangliosides—GM1, GD1a, and the minor gangliosides GalNAc-GD1a and GM1b—have been implicated as potential target antigens in AMAN. The antibodies directed against these ganglioside antigens in AMAN are predominantly of the IgG isotype and generally of the complement-fixing subtypes IgG1 and IgG3.79,143,206 These antiganglioside antibody responses in GBS are almost always polyclonal and can have a broad range of cross-reactivity with related gangliosides.97,143,206 Yuki and colleagues first described the presence of hightiter IgG anti-GM1 antibodies in two Japanese patients who developed AMAN after Campylobacter infection.215 A recent Japanese report examined the relationship between antiganglioside antibodies and the axonal form of GBS. Eighty-six GBS patients were classified as AIDP (36%), AMAN (38%), or inexcitable/unclassifiable (26%) on the basis of electrophysiologic studies.142 Antiganglioside antibodies were almost exclusively present in the axonal group, and more than 50% of patients with AMAN had anti-GM1 antibodies. The association between anti-GD1a antibodies and AMAN was found in two large serologic studies of Japanese GBS cases,142,182 although both studies show that anti-GM1 antibodies are more frequent in AMAN than

2208

Diseases of the Peripheral Nervous System

GD1a antibodies. IgG anti-GD1a antibodies in Japanese patients are associated with severe motor axonal injury and are related to the severity of clinical disease and axonal injury.213,216 Ho and co-workers, in a large prospective series of Chinese GBS patients, found that 60% of AMAN and 4% of AIDP patients had IgG anti-GD1a antibodies.79 IgG anti-GM1 antibodies were frequently detected in both variants.79 In this series serologic evidence of recent C. jejuni infection was present in 80% of AMAN and 50% of AIDP cases. A recent study classified patients with AMAN and IgG anti-GM1 antibodies into two subgroups, one with rapid recovery and the other with slower recovery compared to a group of patients with AIDP.107 Thus antibodies to GD1a and GM1 are found in a proportion of patients with the motor axonal variant, but the frequency of these antibodies varies in different patient populations.

FIGURE 98–5 Pathology of AMAN. In each of these fibers a macrophage (M) has dissected into the periaxonal space and lies adjacent to the axon (A). A, Note that the macrophage is beginning to extend processes around the axon. B, These small processes are seen at higher power. C, The processes (arrows) extend around about half of the axonal circumference. D, Note the condensed-appearing axoplasm. The arrow identifies a clathrin-coated pit in the macrophage.

Distribution of Target Antigens The distribution in the PNS of specific gangliosides has been postulated to underlie the different clinical patterns, but it is now clear that the presence or absence of gangliosides in, for example, motor compared to sensory nerves cannot explain the different variants of GBS. Biochemical studies indicate that there are no consistent differences in the ganglioside content of motor and sensory nerves117,181 except that the fatty acid chain length in the ceramide portion of gangliosides from sensory nerves has a larger proportion of longchain C20–26 fatty acids than motor nerves.141 Qualitative differences in motor and sensory GD1a mobility on thin-layer chromatography was observed, consistent with differences in fatty acid length in the ceramide portion of sensory and motor GD1a.46 The implication of this finding is that antiGD1a antibodies could have different affinities for sensory versus motor GD1a. Several recently developed IgG antiganglioside monoclonal antibodies117 have been used for immunolocalization studies. An anti-GD1a monoclonal antibody preferentially stained the motor nerve axons and only a subpopulation of small unmyelinated sensory fibers.117 GalNAc-GD1a may only be expressed by motor neurons and axons.203 These results are consistent with a recent report of preferential motor nerve staining by serum from an AMAN patient with IgG reactivity to GD1a and GalNAc-GD1a.26,116 In contrast, GM1 is localized to nodal and internodal axolemma as well as paranodal myelin in both sensory and motor nerve fibers.44,46,135,174 No preferential motor axon binding with anti-GM1 antibodies has been recognized.46 Limitations to the immunolocalization studies include the fact that fixed tissues may have altered distribution of gangliosides, compared to living nerves; that the ligands (monoclonal antibodies, toxins, lectins, etc.) used for immunostaining can show either too little or too much depending on their affinity and avidity for the antigen and may not reflect the binding characteristics of antiganglioside antibodies in patient sera; that antiganglioside antibodies might cross-react with glycoproteins in peripheral nerves; and that crypticity of gangliosides

The Guillain-Barré Syndromes

may differ among gangliosides and in different nerves and different subcellular localizations. With these reservations, the immunolocalization studies provide some examples of preferential staining of motor nerve fibers by GD1a-related antibodies. Experimental Evidence The hypothesis of molecular mimicry has been supported by the following two complementary observations. First, several studies have demonstrated that either human GBS sera or purified human antiganglioside antibodies bind to ganglioside-like moieties contained in the core oligosaccharide region of the LPS.134,144,152,154,201,205 Second, immunization with C. jejuni LPSs has been successfully used to induce antiganglioside responses.4,16,47,200 These immune responses in experimental animals have generally been low-affinity non–T cell-dependent antibodies of immunoglobulin M (IgM) and IgG3 type, despite the use of adjuvants to recruit T-cell help.47,200 This reflects a high level of tolerance to self gangliosides that restricts antibody responses to C. jejuni LPSs.16 This high level of tolerance can be overcome in transgenic animals lacking complex gangliosides.16,117 Immunization of such animals with gangliosides has resulted in successful recruitment of T-cell help and production of T-cell–dependent IgG antibodies. The density of ganglioside-like moieties in LPS used for immunization and manipulation of CD40 on B cells can also influence antiganglioside antibody responses.16 These observations support the notion that under some conditions the high level of tolerance to self gangliosides can be overcome after C. jejuni infection, and suggest that breakdown of tolerance to self gangliosides is critical in the pathogenesis of post-Campylobacter GBS. Human studies in which gangliosides were administered provide circumstantial evidence for the role of antiganglioside antibodies in the pathogenesis of axonal forms of GBS. In parts of Europe, injection of brain gangliosides was a frequent treatment for various neurologic disorders. Several cases of acute neuropathy following parenteral administration of gangliosides were reported in the early 1990s.36,48,172 Although the incidence was low,48 there is a positive correlation between ganglioside treatment and GBS.158 The GBS patients had IgG antiganglioside antibodies.85 Sera from six patients reacted with GM1 and that from one with GD1a and GT1b. No individuals who did not develop GBS after ganglioside administration developed IgG antiganglioside reactivity. The clinical and electrophysiologic features of these GBS cases were consistent with AMAN.85 The disease onset varied between 5 and 15 days after the initiation of ganglioside therapy, a short time course for a primary immune response for IgG antiganglioside antibody production. A Phase I trial of murine monoclonal anti-GD2 antibody in metastatic melanoma patients provides the closest approximation to a human passive transfer experiment

2209

with antiganglioside antibodies.170 In this trial a monoclonal anti-GD2 antibody was administered. All patients receiving more than 10 mg of antibody developed acute pelvic and abdominal pain. Five patients developed delayed neuropathic-type pain after the third and fourth infusion, and four of the five patients developed neurologic side effects, including two patients with significant but reversible neuropathy. In a subsequent study, Yuki and colleagues reported that this antibody binds to human peripheral nerve myelin.214 These studies support the hypothesis that antiganglioside antibodies can cause peripheral nerve injury. Immunization with gangliosides or C. jejuni and passive transfer with human GBS sera or purified antiganglioside antibodies have been used by a number of investigators in an attempt to induce animal models of GBS. Nagai and colleagues induced paralytic disease following immunization with GM1132; the pathology was not described in detail. Kusunoki and co-workers induced sensory neuronopathy following immunization with GD1b106 (see later). Yuki and colleagues showed that rabbits repeatedly immunized with mixed brain gangliosides or GM1 developed clinical weakness, high titers of IgG anti-GM1 antibodies, and pathology similar to human AMAN, with periaxonal macrophages and axonal degeneration confined to ventral roots.212 The induction of antiganglioside antibodies in various experimental animals is not always associated with clinical symptoms and or pathology, as suggested by several negative studies.52,115,169,200 The generation of an animal model of C. jejuni–induced experimental neuropathy is complicated because of the absence of reproducible animal models of Campylobacter-related enteritis. Li and co-workers used a C. jejuni strain isolated from a child with AMAN and induced diarrhea, paralysis, and pathologic changes in the nerves, similar to those seen in AMAN, in chickens fed with this organism.113 The LPS of the organism used to induce this model carries ganglioside-like moieties.176 Serologic studies were not done on these animals. Campylobacter jejuni LPS extracted from different GBS-related strains has been successfully used to induce antiganglioside antibodies but without clinical or pathologic disease.4,115,151,200 A recent study proposed that the differences in the relative affinity of experimentally induced antiganglioside antibodies may relate to the dissociation of serologic responses and experimental disease.115 This observation provides a potential explanation for the disparate results reported by different laboratories. In vitro models have studied the electrophysiologic and morphologic effects of antiganglioside antibodies associated with AMAN. Some experimental findings have found no effect of antiganglioside antibodies.70,75,151 Other in vitro electrophysiologic studies have reported anti-GM1 antibody–mediated conduction failure at the nodes of Ranvier or motor nerve terminals,6,164,183 and this conduction failure could be secondary to blockade of voltage-gated sodium channels.183,193 The studies supporting the blockade of

2210

Diseases of the Peripheral Nervous System

sodium channels by anti-GM1 antibodies have given rise to the controversial concept that conduction failure could be on an axonal basis. These authors propose that AMAN patients with IgG anti-GM1 antibodies and rapid recovery have a reversible axonal conduction failure,183,193 whereas those with poor recovery and anti-GM1 antibodies have axonal degeneration. Consistent with this concept, some patients with AMAN recover rapidly,76,107 and the pathologic changes in some cases with AMAN are very mild despite severe clinical weakness.56 Zhang and colleagues have recently studied the pathogenic effects of anti-GD1a antibodies in two in vitro models.217 A cell lysis assay indicated that these antibodies can lyse neuronal cell lines by targeting gangliosides, and neuronal injury in this model is complement dependent. Overall, these studies indicate that antiganglioside antibodies have pathophysiologic effects in in vitro assays. The exact mechanisms of antiganglioside antibody–mediated neural injury needs further elucidation.

Fisher Syndrome FS represents approximately 5% of GBS cases. A small proportion of cases follow C. jejuni infection. The advances made during the last 10 years now strongly support the contention that this neuropathy is produced by antiganglioside antibodies. More than 85% of cases of FS have anti-GQ1b antibodies.19,197,208 FS sera contain IgM, immunoglobulin A, and IgG anti-GQ1b reactivity. The IgG anti-GQ1b antibodies, the most robust and persistent,99,173 are of complement-fixing isotypes.151,196 The fine specificity of anti-GQ1b antibodies in FS has been correlated with specific clinical features. Antibodies to GQ1b in FS almost always cross-react with a closely related minor ganglioside, GT1a.20,86 Serum reactivity to gangliosides such as GD3, GD1b, and/or GT1b is seen in about half the patients with FS. These sera thus recognize a disialosyl moiety shared by these gangliosides and GQ 1b . 194 Oropharyngeal weakness appears to be preferentially associated with GT1a reactivity, whereas ophthalmoplegia correlates with GQ1b reactivity.94,98,126,137 The presence of ophthalmoplegia in patients with AIDP has also been correlated with the presence of anti-GQ1b reactivity.165,184 As noted, antiGQ1b antibodies have also been reported in patients with Bickerstaff’s brainstem encephalitis, in which the clinical manifestations can include ophthalmoplegia, coma, pyramidal tract dysfunction, and in some cases responsiveness to plasmapheresis.96,140,204 Many investigators now consider Bickerstaff’s encephalitis and FS to be related conditions.140,209 The basis of ataxia seen in FS remains controversial; both cerebellar100 and proprioceptive loss20,126,194,197,208 has been implicated. Areflexia in FS is likely due to afferent defects.138,194

Inciting Events and Immune Response Infections are the most common triggering event preceding FS. Both upper respiratory infections and C. jejuni enteritis have been documented. The post-Campylobacter cases provide strong support for the concept of molecular mimicry. Structural studies have demonstrated GD3- and GT1a-like oligosaccharides in LPSs of C. jejuni isolates from both GBS and FS patients.171,178 However, a GQ1b-like structure has not been established by this technique. Antibody binding assays using human or murine monoclonal antibodies have shown the presence of GT1a– and GQ1b–cross-reactive moieties in C. jejuni LPSs.88,89,133,211,214 Mice immunized with C. jejuni LPS containing GT1a-like moieties produce antibodies that recognize GQ1b and other cross-reactive disialylated gangliosides. Antiganglioside antibodies generated in this experimental paradigm were mostly of the IgM type, typical of T-cell–independent responses to carbohydrate or glycolipid antigens. In contrast, serologic studies in FS indicate class switching to IgG and subclass restriction to IgG1 and IgG3, both usually features of T-cell help and atypical of the human anticarbohydrate antibody responses. Distribution of Target Antigens Chiba and colleagues initially showed binding of anti-GQ1b antibodies to paranodal myelin and nodes of Ranvier.20 In a quantitative analysis of extracted gangliosides, the same authors showed that the extraocular cranial nerves contain about twice the amount of GQ1b, compared to other cranial and peripheral nerves.21 GQ1b is present in other peripheral nerves, as shown by binding of disialosyl antibodies to the nodes of Ranvier in somatic nerves.47,195 Anti-disialosyl antibodies stain dorsal root ganglion (DRG) neurons in several species, including humans,47,105,118,136 and a GD1b-specific monoclonal antibody preferentially binds to large DRG neurons.46 Muscle spindles and intrafusal muscle fibers are also labeled by antibodies to disialylated antibodies.195 These distribution studies support the notion that FS-specific antibodies can injure the extraocular nerves and peripheral proprioceptive sensory connections, and alter the muscle stretch reflexes. Experimental Studies Antibodies with FS reactivity bind to neuromuscular junctions (NMJs) in both somatic and extraocular muscles.47,151,195 Anti-GQ1b and related antibodies can cause pathophysiologic changes.17,18,147,151 Although the few autopsy studies in patients with FS have indicated that segmental demyelination is the pathologic characteristic of this variant of GBS, most experimental studies have focused on demonstrating antibody-mediated injury at the NMJ. In phrenic nerve hemidiaphragm preparations, antiGQ1b–positive FS IgG caused a temporary increase in spontaneous quantal acetylcholine release at the NMJ, with subsequent nerve terminal conduction failure. Following passive transfer of a human IgM anti-disialosyl

The Guillain-Barré Syndromes

antibody, ex vivo hemidiaphragm preparations showed conduction failure.195 In in vitro studies of these hemidiaphragm preparations, Willison and colleagues have demonstrated that FS sera, FS IgG fractions, and a human monoclonal anti-GQ1b IgM antibody bind to NMJs, cause massive quantal release of acetylcholine, and eventually blockade of neuromuscular transmission primarily through presynaptic mechanisms resembling the effects of the paralytic neurotoxin ␣-latrotoxin.151 These effects on NMJs required complement activation. These findings were further extended to the monoclonal anti-diasialosyl antibodies generated by immunization with C. jejuni LPS, thus supporting a connection between the infecting organism, antiganglioside antibody, and nerve injury.47 Light and electron microscopic pathology confirmed degeneration of motor nerve terminals.136 A parallel set of studies by Buchwald and colleagues17,18 using a perfused macro patch-clamp electrode technique on the phrenic nerve hemidiaphragm preparation, IgG fractions from both GQ1b-positive and -negative FS cases blocked evoked acetylcholine release and depressed the amplitude of postsynaptic potentials, implicating both a pre- and postsynaptic blocking effect. This effect was reversible by washing out the IgGs and occurred in a complement-independent manner. This difference between the findings of the two groups may indicate that more than one mechanism of NMJ blockade can be mediated by these antibodies. Kusunoki and colleagues have induced a sensory ataxic neuropathy model in rabbits by immunization with the disialosyl ganglioside GD1b.103,105,106 Repeated injection of the ganglioside and KLH mixed with Freund’s adjuvant leads to sensory ataxia in a significant proportion of animals after 4 or more weeks of immunization. All animals developed anti-GD1b antibodies, but neuropathy was restricted to animals with antibody reactivity restricted to GD1b and not cross-reactive with the Gal(␤1-3)GalNAc epitope.103 Animals with clinical disease demonstrated degeneration of DRG neurons with secondary nerve fiber degeneration in dorsal roots and dorsal columns. This model not only is instructive for the acute and chronic sensory ataxic neuropathies associated with anti-GD1b antibodies but also provides support for the concept that antibodies with reactivity to other disialosyl gangliosides could cause a similar pattern of injury and sensory ataxia and areflexia.

Acute Inflammatory Demyelinating Polyneuropathy AIDP was long considered a prototypic T-cell–mediated disease, analogous to EAN. At present it is likely that there is heterogeneity of immunopathogenesis with the AIDP group. Immunopathologically, the extent of T-cell infiltration varies widely. As noted, in many cases macrophages invade normal-appearing myelin, whereas other cases have

2211

vesicular demyelination (see Figs. 98–3 and 98–4). Some reports have suggested antibody or complement localized with myelin sheaths, and others have identified antibody and complement activation on the outer perimeter of the Schwann cell67(see Fig. 98–4), suggesting binding to the basal lamina or Schwann cell plasmalemma. No single antigen has been identified in AIDP cases. Antibodies against gangliosides,53,160 basal lamina components,149 and proteins95,180,199 are each identified in a proportion of cases, making their role in tissue injury uncertain. With regard to gangliosides, IgG anti-GM1 antibodies are found in 14% to 25% of AIDP cases in developed countries. The presence of these antibodies correlates with prominent motor involvement, and more severe disease with a greater proportion of cases with inexcitable nerves than seronegative AIDP.90,160,192 The presence of antiganglioside antibodies has been frequently associated with preceding C. jejuni infection, although not all studies confirm this association.2,33,78,190 The presence of anti-GM1 antibodies probably correlates with a slower rate of recovery.160 LM1 is a major PNS myelin ganglioside, and occasional AIDP patients have high-titer monospecific anti-LM1 antibodies,87,180 but this finding is exceptional. Kusunoki and colleagues initially reported the presence of antibodies against the minor gangliosides GM1b and GalNAc-GD1a in Japanese AIDP patients.102,104 Subsequent studies have confirmed this association in Dutch and Chinese populations.14,15,25 Antibodies against these minor gangliosides are present in 12% to 20% of patients with AIDP. Patients with these antibodies are more likely to have distal motor weakness, lack sensory and cranial nerve involvement, and more frequently have preceding C. jejuni infection.25,27,29 In a recent comparative study of Japanese and Dutch patients with AIDP, Ang and co-workers found that Dutch patients had a stronger IgM response against GM1b and GalNAc-GD1a, whereas Japanese patients were more likely to mount an IgG response against these antigens.12 Sensory-predominant patterns of AIDP have different antibody associations. For example, AIDP cases with serologic evidence of preceding cytomegalovirus (CMV) infection are more likely to have prominent sensory symptoms, and findings compared to those preceded by C. jejuni infection.185,189 Antibodies to ganglioside GM2 are present in about half of the patients with CMV-associated GBS.65,148 Such antibodies to ganglioside GM2 can also be present after uncomplicated CMV and Epstein-Barr virus infection,65 suggesting that antibodies to GM2 are not sufficient to produce neuropathy and that other host or organism factors may play a role. Among 445 GBS patients, a subgroup of 9 patients had sensory disturbance, no electrophysiologic evidence of axonal injury, and monospecific IgG GD1b antibodies. Eight of these nine patients had a preceding respiratory infection and one had gastroenteritis.57 These findings support the notion that

2212

Diseases of the Peripheral Nervous System

motor-dominant syndromes are more likely to be preceded by C. jejuni infection, whereas patients with prominent sensory features are more likely to have preceding respiratory infections or CMV. It is likely that protein antigens play a role in some cases of AIDP. AIDP following Semple rabies vaccine is associated with anti–myelin basic protein antibody and T-cell responses.72 Serology has suggested that the PNS myelin constituent peripheral myelin protein 22 might be involved in some cases,43 and rare patients have reactivity to myelin protein zero or connexin 32.108

MANAGEMENT In the past 15 years, a number of large-scale controlled trials of immunotherapy of AIDP have been completed and published.40,51,60,84,186,187 AIDP is now one of the relatively few neurologic disorders in which treatment choices can be solidly based on evidence. Immunotherapies shorten the time to recovery and appear to prevent progression to more severe stages of disease. These successes, however, should not distract from the facts that the quality of critical care remains the biggest factor in influencing mortality and patient perception of disease experience, that the outcome is unsatisfactory for a substantial number of patients, and that we do not know the effectiveness of different immunotherapies in the axonal forms of GBS. The management of GBS can be divided into preparation for progression in the early stages, management of the full-blown disease with intensive care and immunotherapy, and rehabilitation.

Monitoring and Education in Early Stages of the Disease Because of the dramatic rapidity with which deterioration can occur, except the mildest cases, the presumptive diagnosis of GBS mandates hospitalization for observation. During the early stages, while diagnostic tests are being obtained and respiration and cardiovascular function monitored, teaching the patient and family about GBS can be initiated. After recovery from severe GBS, patients often reflect that they were not educated about the disease and its short- and longterm prognoses. With progression patients can lose the means of communicating complex questions or needs. For these reasons, at the time of admission it is useful to discuss the potential course of GBS and even to visit the intensive care unit. Seeing a critical care unit when walking or upright in a wheelchair can make transfer into the unit when paralyzed less ominous. Meeting individuals who have recovered from GBS can be helpful, and many former patients are pleased to participate. Teaching the patient and the family a communication system before it is required provides confidence even if it is not needed. A Lucite “letter board,” consisting of

the alphabet pasted on a transparent board, allows the caregiver or family member to stand on the other side of the board and “read” the patient’s meaning by following the movement of the patient’s eyes from letter to letter. With practice, this provides an efficient communication system that will be made available to all GBS patients except those with ophthalmoplegia. Monitoring of vital capacity, pulse, and blood pressure and swallowing is critical during the phase of progression. Vital capacity is monitored as frequently as dictated by the degree of impairment and the rate of change. If the vital capacity is decreasing, this may be every 2 hours. The decision for transfer to the intensive care unit should precede the need for intubation and ventilatory support. A vital capacity of less than 15 mL/kg has been taken as a threshold for intubation,166 but predictors of intubation and ventilation include the rate of change, facial and bulbar weakness, and dysautonomia.111 Arguably, any patient who develops significant hypercarbia or hypoxemia should have been intubated earlier. The operation of combinations of these factors can be illustrated by a case history. A patient with cardiovascular autonomic lability, marginal pharyngeal function, and a modestly reduced vital capacity aspirated a small volume while drinking. This triggered coughing, with consequent fatigue and hypoventilation coupled with alterations in pulse and blood pressure, requiring emergent intubation and cardiovascular stabilization. Monitoring of dysphasia can be done by watching the patient swallow water and then asking the patient to count, listening for the “bubbling” speech that represents retained fluid in the valleculae. Similarly, monitoring resting pulse and blood pressure, as well as orthostatic changes, and the pulse and blood pressure while ambulating on the floor are useful ways of detecting early autonomic insufficiency. Volume depletion or overexpansion is a particular threat to patients with autonomic insufficiency. Even early in GBS pain can be an important and underrecognized problem. If simple analgesics are usually without value, opiates are required. If communication is difficult, a signaling system for patient-controlled analgesia is essential. Constipation is a major side effect of opiate analgesia, so that monitoring of bowel movements and early intervention is important.

Critical Care of Guillain-Barré Patients The mortality of the GBSs before the 1980s is incompletely documented, but may have been as high as 15% to 20% for severely affected patients. Data from the large-scale multicenter studies that began in the early 1980s shows that, with modern critical care, the mortality is in the range of 1.25% to 2.5%.40,51,60,84,187 Death in critical care units is usually not the consequence of ventilatory failure but rather of intercurrent infection, myocardial infarction, or pulmonary embolism.

2213

The Guillain-Barré Syndromes

Although critical care is life-saving, complications are frequent during prolonged intensive care unit stays. With admissions of over 3 weeks, pneumonia, bacteremia, or other serious infections developed in 60% of patients.73 Decisions regarding prophylactic and supportive care are influenced by the patient’s age and other medical risk factors. For almost all patients, use of subcutaneous heparin, 5000 units two times a day, is recommended.73 A footboard and passive manipulation of paretic limbs also helps reduce the risk of phlebitis. In patients with a facial weakness, exposure keratitis can be a devastating and permanent sequela. Use of methylcellulose eyedrops and passive closure of the lids for portions of the day can prevent this complication.

Immunotherapy Both infusion of human immune globulin and plasmapheresis have been shown to be beneficial in AIDP. The best documented effect of these agents is to shorten the time to recovery. These conclusions are based on several controlled multicenter studies. The first trial of plasmapheresis randomized 242 patients in North America, entered into the trial with the criterion that they were unable to walk without assistance. Plasmapheresis substantially shortened the time to resume walking. Multifactorial analysis demonstrated that this effect was present for patients of greater degrees of severity on admission and for individuals of all ages.122,123 Subsequent randomized controlled studies from Scandinavia146 and France39 confirmed these effects. The sole negative study of plasmapheresis is an early North American study.124 The reason for its different outcome is not certain, but the study was small and there was a trend toward more severe disease at baseline in the plasmapheresis group. The subsequent French study suggested that early plasmapheresis prevented progression to more severe disease.39 Randomized studies of intravenous human immune globulin have demonstrated benefit.186 Two subsequent studies have performed head-to-head comparisons of plasmapheresis and human immune globulin. Although some preference for plasmapheresis was suggested in a Dutch study,186 an international study organized in England found equal benefit to plasmapheresis and intravenous immune globulin.150 There are also data suggesting that both of these therapies also prevent progression to more severe grades of dysfunction. Corticosteroids alone have been shown to produce no benefit in AIDP.61,84 An important recent trial compared combined intravenous immune globulin and methylprednisolone, 500 mg daily for 5 days, to an immune globulin/intravenous placebo group in a double-blind, randomized study.187 There was no difference in the primary outcome, the proportion of patients improved at 4 weeks.187 A suggestion of slight benefit came from

some secondary measures, and there was little definite evidence of toxicity beyond transient elevations in blood glucose. Four patients in the control group died compared to six in the methylprednisolone group. Five of these deaths were many weeks after randomization. At present there is little rationale for using corticosteroids in GBS. These data leave open the need for improvements in treatment, with the goal of further reducing the time to recovery and reducing the proportion of individuals with late residua.

REFERENCES 1. Ahmed, S., Libman, R., Wesson, K., et al.: Guillain-Barre syndrome: an unusual presentation of West Nile virus infection. Neurology 55:144, 2000. 2. Ali, Z., Ringkamp, M., Hartke, T. V., et al.: Uninjured C-fiber nociceptors develop spontaneous activity and alphaadrenergic sensitivity following L6 spinal nerve ligation in the monkey. J. Neurophysiol. 81:455, 1999. 3. Allos, B. M.: Campylobacter jejuni infections: update on emerging issues and trends. Clin. Infect. Dis. 32:1201, 2001. 4. Ang, C. W., de Klerk, M. A., Endtz, H. P., et al.: GuillainBarre syndrome– and Miller Fisher syndrome–associated Campylobacter jejuni lipopolysaccharides induce anti-GM1 and anti-GQ1b antibodies in rabbits. Infect. Immun. 69:2462, 2001. 5. Ang, C. W., Laman, J. D., Willison, H. J., et al.: Structure of Campylobacter jejuni lipopolysaccharides determines antiganglioside specificity and clinical features of Guillain-Barré and Miller Fisher patients. Infect. Immun. 70:1202, 2002. 6. Arasaki, K., Kusunoki, S., Kudo, N., et al: The pattern of antiganglioside antibody reactivities producing myelinated nerve conduction block in vitro. J. Neurol. Sci. 161:163, 1998. 7. Asbury, A. K., Arnason, B. G., and Adams, R. D.: The inflammatory lesion in idiopathic polyneuritis. Medicine 48:173, 1969. 8. Aspinall, G. O., Fujimoto, S., McDonald, A. G., et al.: Lipopolysaccharides from Campylobacter jejuni associated with Guillain-Barre syndrome patients mimic human gangliosides in structure. Infect. Immun. 62:2122, 1994. 9. Aspinall, G. O., McDonald, A. G., and Pang, H.: Lipopolysaccharides of Campylobacter jejuni serotype O:19: structures of O antigen chains from the serostrain and two bacterial isolates from patients with the Guillain-Barré syndrome. Biochemistry 33:250, 1994. 10. Aspinall, G. O., McDonald, A. G., Pang, H., et al.: Lipopolysaccharides of Campylobacter jejuni serotype O:19: structures of core oligosaccharide regions from the serostrain and two bacterial isolates from patients with the Guillain-Barré syndrome. Biochemistry 33:241, 1994. 11. Aspinall, G. O., McDonald, A. G., Raju, T. S., et al.: Chemical structures of the core regions of Campylobacter jejuni serotypes O:1, O:4, O:23, and O:36 lipopolysaccharides. Eur. J. Biochem. 213:1017, 1993.

2214

Diseases of the Peripheral Nervous System

12. Bauman, M. B., DiMase, J. D., Oski, F., Senior, J. R.: Brown bowel and skeletal myopathy associated with vitamin E depletion in pancreatic insufficiency. Gastroenterology 54:93, 1968. 13. Berciano, J., and Garcia, A.: Nerve ischemia in GuillainBarre syndrome: an alternative mechanism for early conduction failure. Rev. Neurol. (Paris) 158:364, 2002. 14. Black, J. A., Dib-Hajj, S., McNabola, K., et al.: Spinal sensory neurons express multiple sodium channel a-subunit mRNAs. Mol. Brain Res. 43:117, 1996. 15. Black, J. A., Langworthy, K., Hinson, A. W., et al.: NGF has opposing effects on Na⫹ channel III and SNS gene expression in spinal sensory neurons. Neuroreport 8:2331, 1997. 16. Bowes, T., Wagner, E. R., Boffey, J., et al.: Tolerance to self gangliosides is the major factor restricting the antibody response to lipopolysaccharide core oligosaccharides in Campylobacter jejuni strains associated with Guillain-Barre syndrome. Infect. Immun. 70:5008, 2002. 17. Buchwald, B., Toyka, K. V., Zielasek, J., et al.: Neuromuscular blockade by IgG antibodies from patients with Guillain-Barre syndrome: a macro-patch-clamp study. Ann. Neurol. 44:913, 1998. 18. Buchwald, B., Weishaupt, A., Toyka, K. V., et al.: Pre and postsynaptic blockade of neuromuscular transmission by Miller Fisher syndrome IgG at mouse motor nerve terminals. Eur. J. Neurosci. 10:281, 1998. 19. Carpo, M., Pedotti, R., Lolli, F., et al.: Clinical correlate and fine specificity of anti-GQ1b antibodies in peripheral neuropathy. J. Neurol. Sci. 155:186, 1998. 20. Chiba, A., Kusunoki, S., Obata, H., et al.: Serum anti-GQ1b antibody is associated with ophthalmoplegia in Miller Fisher syndrome and Guillain-Barre syndrome: clinical and immunohistochemical studies. Neurology 43:1911, 1993. 21. Chiba, A., Kusunoki, S., Obata, H., et al.: Ganglioside composition of the human cranial nerves, with special reference to pathophysiology of Miller Fisher syndrome. Brain Res. 745:32, 1997. 22. Cooke, C. R., Latif, K. A., Huch, K. M., et al.: Inappropriate antidiuresis and hyponatremia with suppressible vasopressin in Guillain-Barre syndrome. Am. J. Nephrol. 18:71, 1998. 23. Cornblath, D. R., Griffin, D. E., Welch, D., et al.: Quantitative analysis of endoneurial T-cells in human sural nerve biopsies. J. Neuroimmunol. 26:113, 1990. 24. Cornblath, D. R., Mellits, E. D., Griffin, J. W., et al.: Motor conduction studies in the Guillain-Barre syndrome: description and prognostic value. Ann. Neurol. 23:354, 1988. 25. Cummins, T. R., and Waxman, S. G.: Downregulation of tetrodotoxin-resistant sodium currents and upregulation of a rapidly repriming tetrodotoxin-sensitive sodium current in small spinal sensory neurons after nerve injury. J. Neurosci. 17:3503, 1997. 26. De Angelis, M. V., Di Muzio, A., Lupo, S., et al.: Anti-GD1a antibodies from an acute motor axonal neuropathy patient selectively bind to motor nerve fiber nodes of Ranvier. J. Neuroimmunol. 121:79, 2001. 27. Dixon, W. J.: Efficient analysis of experimental observations. Annu. Rev. Pharmacol. Toxicol. 20:441, 1980. 28. Duménil, L.: Paralysie peripherique du mouvement et du sentiment, portant sur les quatre membres—atrophie des ramaux nerveux des parties paralysees. Gaz. Hebd. Med. Chir. 11:203, 1864.

29. Eames, R., and Gamble, H.: Schwann cell relationships in normal human cutaneous nerves. J. Anat. 106:417, 1970. 30. Eichhorst, H.: Neuritis acuta progressiva. Virchows Arch. 268, 1877. 31. Emilia-Romagna Study Group on Clinical and Epidemiological Problems in Neurology: Campylobacter jejuni infection and Guillain-Barré syndrome: a case-control study. Neuroepidemiology 17:296, 1998. 32. Emilia-Romagna Study Group on Clinical and Epidemiological Problems in Neurology: Guillain-Barré syndrome variants in Emilia-Romagna, Italy, 1992–3: incidence, clinical features, and prognosis. J. Neurol. Neurosurg. Psychiatry 65:218, 1998. 33. Enders, U., Karch, H., Toyka, K. V., et al.: The spectrum of immune responses to Campylobacter jejuni and glycoconjugates in Guillain-Barre syndrome and in other neuroimmunological disorders. Ann. Neurol. 34:136, 1993. 34. Feasby, T. E., Gilbert, J. J., Brown, W. F., et al: An acute axonal form of Guillain-Barre polyneuropathy. Brain 109:1115, 1986. 35. Feasby, T. E., Hahn, A. F., Brown, W. F., et al.: Severe axonal degeneration in acute Guillain-Barre syndrome: evidence of two different mechanisms? J. Neurol. Sci. 116:185, 1993. 36. Figueras, A., Morales-Olivas, F. J., Capella, D., et al.: Bovine gangliosides and acute motor polyneuropathy. BMJ 305:1330, 1992. 37. Fisher, M.: An unusual variant of acute idiopathic polyneuritis (syndrome of ophthalmoplegia ataxia and areflexia). N. Engl. J. Med. 255:57, 1956. 38. Fletcher, D. D., Lawn, N. D., Wolter, T. D., et al.: Long-term outcome in patients with Guillain-Barre syndrome requiring mechanical ventilation. Neurology 54:2311, 2000. 39. French Cooperative Group on Plasma Exchange in Guillain-Barre Syndrome: Efficacy of plasma exchange in Guillain-Barre syndrome: role of replacement fluids. Ann. Neurol. 22:753, 1987. 40. French Cooperative Group on Plasma Exchange in Guillain-Barre Syndrome: Plasma exchange in GuillainBarre syndrome: one-year follow-up. Ann. Neurol. 32:94, 1992. 41. Friedman, Y., Lee, L., Wherrett, J. R., et al.: Simulation of brain death from fulminant de-efferentation. Can. J. Neurol. Sci. 30:397, 2003. 42. Fujimoto, S., Yuki, N., Itoh, T., et al.: Specific serotype of Campylobacter jejuni associated with Guillain-Barre syndrome [letter]. J. Infect. Dis. 165:183, 1992. 43. Gabriel, C. M., Gregson, N. A., and Hughes, R. A.: AntiPMP22 antibodies in patients with inflammatory neuropathy. J. Neuroimmunol. 104:139, 2000. 44. Ganser, A. L., Kirschner, D. A., and Willinger, M.: Ganglioside localization on myelinated nerve fibres by cholera toxin binding. J. Neurocytol. 12:921, 1983. 45. Gombault: Contribution a l’histoire anatomique de l’atrophie musculaire saturnine. Arch. Physiol. V:592, 1873. 46. Gong, Y., Tagawa, Y., Lunn, M. P., et al.: Localization of major gangliosides in the PNS: implications for immune neuropathies. Brain 125:2491, 2002. 47. Goodyear, C. S., O’Hanlon, G. M., Plomp, J. J., et al.: Monoclonal antibodies raised against Guillain-Barre syndromeassociated Campylobacter jejuni lipopolysaccharides react

The Guillain-Barré Syndromes

48.

49.

50. 51.

52.

53.

54.

55.

56.

57. 58.

59.

60. 61.

62.

63.

64.

65.

with neuronal gangliosides and paralyze muscle-nerve preparations. J. Clin. Invest. 104:697, 1999. [published erratum appears in J. Clin. Invest. 104:1771, 1999.] Granieri, E., Casetta, I., Govoni, V., et al.: Ganglioside therapy and Guillain-Barre syndrome: a historical cohort study in Ferrara, Italy, fails to demonstrate an association. Neuroepidemiology 10:161, 1991. Grattan-Smith, P. J., Morris, J. G., Johnston, H. M., et al.: Clinical and neurophysiological features of tick paralysis. Brain 120:1975, 1997. Graves, R. J.: Clinical Lectures on the Practice of Medicine. New Sydenham Soc. 1:578, 1884. Greenwood, R. J., Newsom-Davis, J. M., Hughes, R. A. C., et al.: Controlled trial of plasma exchange in acute inflammatory polyradiculoneuropathy. Lancet 1:877, 1984. Gregson, N. A., and Hammer, C. T.: Some immunological properties of antisera raised against the trisialoganglioside GT1B. Mol. Immunol. 19:543, 1982. Gregson, N. A., Koblar, S., and Hughes, R. A. C.: Antibodies to gangliosides in Guillain-Barre syndrome: specificity and relationship to clinical features. Q. J. Med. 86:111, 1993. Griffin, J. W., Li, C. Y., Ho, T. W., et al.: Guillain-Barre syndrome in northern China: the spectrum of neuropathologic changes in clinically defined cases. Brain 118:577, 1995. Griffin, J. W., Li, C. Y., Ho, T. W., et al.: Pathology of the motor-sensory axonal Guillain-Barre syndrome. Ann. Neurol. 39:17, 1996. Griffin, J. W., Li, C. Y., Macko, C., et al.: Early nodal changes in the acute motor axonal neuropathy pattern of the Guillain-Barre syndrome. J. Neurocytol. 25:33, 1996. Groher, M. E.: Dysphagic patients with progressive neurologic disease. Semin. Neurol. 16:355, 1996. Guerry, P., Szymanski, C. M., Prendergast, M. M., et al.: Phase variation of Campylobacter jejuni 81–176 lipooligosaccharide affects ganglioside mimicry and invasiveness in vitro. Infect. Immun. 70:787, 2002. Guillain, G., Barré, J. A., and Strohl, A.: Sur un syndrome de radiculonebrite avec hyperalbuminose du liquide cephalorachidien sans reaction cellulaire: remarques sur les caracteres cliniques et graphiques des reflexes tendineux. Bull. Soc. Med. Hop. Paris 40:1462, 1916. Guillain-Barre Study Group: Plasmapheresis and acute Guillain-Barre syndrome. Neurology 35:1096, 1985. Guillain-Barre Syndrome Steroid Trial Group: Double-blind trial of intravenous methylprednisolone in Guillain-Barre syndrome. Lancet 341:586, 1993. Gupta, S. K., Taly, A. B., Suresh, T. G., et al.: Acute idiopathic axonal neuropathy (AIAN): a clinical and electrophysiological observation. Acta Neurol. Scand. 89:220, 1994. Hadden, R. D., Cornblath, D. R., Hughes, R. A., et al.: Electrophysiological classification of Guillain-Barre syndrome: clinical associations and outcome. Plasma Exchange/Sandoglobulin Guillain-Barre Syndrome Trial Group. Ann. Neurol. 44:780, 1998. Hadden, R. D., Karch, H., Hartung, H. P., et al.: Preceding infections, immune factors, and outcome in Guillain-Barre syndrome. Neurology 56:758, 2001. Hadjivassiliou, M., Gibson, A., Davies-Jones, G. A., et al.: Does cryptic gluten sensitivity play a part in neurological illness? [see comments]. Lancet 347:369, 1996.

2215

66. Hafer-Macko, C., Hsieh, S.-T., Li, C. Y., et al.: Acute motor axonal neuropathy: an antibody-mediated attack on axolemma. Ann. Neurol. 40:635, 1996. 67. Hafer-Macko, C., Sheikh, K. A., Li, C. Y., et al.: Immune attack on the Schwann cell surface in acute inflammatory demyelinating polyneuropathy. Ann. Neurol. 39:625, 1996. 68. Hagberg, B.: Polyneuropathies in paediatrics. Eur. J. Pediatr. 149:296, 1990. 69. Hahn, A. F., Feasby, T. E., Steele, A., et al.: Demyelination and axonal degeneration in Lewis rat experimental allergic neuritis depend on myelin dosage. Lab. Invest. 59:115, 1988. 70. Harvey, G. K., Toyka, K. V., Zielasek, J., et al.: Failure of anti-GM1 IgG or IgM to induce conduction block following intraneural transfer. Muscle Nerve 18:388, 1995. 71. Haymaker, W., and Kernohan, J. W.: The Landry-GuillainBarre syndrome: a clinicopathologic report of fifty fatal cases and a critique of the literature. Medicine 28:59, 1949. 72. Hemachudha, T., Phanuphak, P., Johnson, R. T., et al.: Neurologic complications of Semple-type rabies vaccine: clinical and immunologic studies. Neurology 37:550, 1987. 73. Henderson, R. D., Lawn, N. D., Fletcher, D. D., et al.: The morbidity of Guillain-Barre syndrome admitted to the intensive care unit. Neurology 60:17, 2003. 74. Henderson, R. D., Ohlrich, G. D., and Pender, M. P.: Guillain-Barre syndrome and optic neuritis after Mycoplasma pneumoniae infection. Aust. N. Z. J. Med. 28:481, 1998. 75. Hirota, N., Kaji, R., Bostock, H., et al.: The physiological effect of anti-GM1 antibodies on saltatory conduction and transmembrane currents in single motor axons. Brain 120:2159, 1997. 76. Ho, T. W., Li, C. Y., Cornblath, D. R., et al.: Patterns of recovery in the Guillain-Barré syndromes. Neurology 48:695, 1997. 77. Ho, T. W., Hsieh, S.-T., Nachamkin, I., et al.: Motor nerve terminal degeneration provides a potential mechanism for rapid recovery in acute motor axonal neuropathy after Campylobacter infection. Neurology 48:717, 1997. 78. Ho, T. W., Mishu, B., Li, C. Y., et al.: Guillain-Barre syndrome in northern China: relationship to Campylobacter jejuni infection and anti-glycolipid antibodies. Brain 118:597, 1995. 79. Ho, T. W., Willison, H. J., Nachamkin, I., et al.: AntiGD1a antibody is associated with axonal but not demyelinating forms of Guillain-Barré syndrome. Ann. Neurol. 45:168, 1999. 80. Hochman, M. S., Kobetz, S. A., and Handwerker, J. V.: Inappropriate secretion of antidiuretic hormone associated with Guillain-Barre syndrome. Ann. Neurol. 11:322, 1982. 81. Hoffmann, O., Reuter, U., Schielke, E., et al.: SIADH as the first symptom of Guillain-Barre syndrome. Neurology 53:1365, 1999. 82. Holmes, G.: Acute febrile polyneuritis. Br. Med. J. 2:37, 1917. 83. Honavar, M., Tharakan, J. K. J., Hughes, R. A. C., et al.: A clinicopathological study of the Guillain-Barre syndrome. Brain 114:1245, 1991. 84. Hughes, R. A. C., Newsom-Davis, J. M., Perkin, G. D., et al.: Controlled trial of prednisolone in acute polyneuropathy. Lancet 2:750, 1978.

2216

Diseases of the Peripheral Nervous System

85. Illa, I., Ortiz, N., Gallard, E., et al.: Acute axonal GuillainBarre syndrome with IgG antibodies against motor axons following parenteral gangliosides. Ann. Neurol. 38:218, 1995. 86. Ilyas, A. A., Cook, S. D., Mithen, F. A., et al.: Antibodies to GT1a ganglioside in patients with Guillain-Barre syndrome. J. Neuroimmunol. 82:160, 1998. 87. Ilyas, A. A., Willison, H. J., Quarles, R. H., et al.: Serum antibodies to gangliosides in Guillain-Barre syndrome. Ann. Neurol. 23:440, 1988. 88. Jacobs, B. C., Endtz, H., van der Meché, F. G. A., et al.: Serum anti-GQ1b IgG antibodies recognize surface epitopes on Campylobacter jejuni from patients with Miller Fisher syndrome. Ann. Neurol. 37:260, 1995. 89. Jacobs, B. C., Hazenberg, M. P., Van Doorn, P. A., et al.: Cross-reactive antibodies against gangliosides and Campylobacter jejuni lipopolysaccharides in patients with Guillain-Barré or Miller Fisher syndrome. J. Infect. Dis. 175:729, 1997. 90. Jacobs, B. C., Van Doorn, P. A., Schmitz, P. I., et al.: Campylobacter jejuni infections and anti-GM1 antibodies in Guillain-Barré syndrome. Ann. Neurol. 40:181, 1996. 91. Jampel, R. S., and Haidt, S. J.: Bell’s phenomenon and acute idiopathic polyneuritis. Am. J. Ophthalmol. 74:145, 1972. 92. Jeha, L. E., Sila, C. A., Lederman, R. J., et al.: West Nile virus infection: a new acute paralytic illness. Neurology 61:55, 2003. 93. Kanda, T., Yamawaki, M., and Mizusawa, H.: Sera from Guillain-Barre patients enhance leakage in blood-nerve barrier model. Neurology 60:301, 2003. 94. Kashihara, K., Shiro, Y., Koga, M., et al.: IgG anti-GT1a antibodies which do not cross react with GQ1b ganglioside in a pharyngeal-cervical-brachial variant of Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 65:799, 1998. 95. Khalili-Shirazi, A., Atkinson, P., Gregson, N., et al.: Antibody responses to P0 and P2 myelin proteins in Guillain-Barre syndrome and chronic idiopathic demyelinating polyradiculoneuropathy. J. Neuroimmunol. 46:245, 1993. 96. Kikuchi, M., Tagawa, Y., Iwamoto, H., et al.: Bickerstaff’s brainstem encephalitis associated with IgG anti-GQ1b antibody subsequent to Mycoplasma pneumoniae infection: favorable response to immunoadsorption therapy. J. Child Neurol. 12:403, 1997. 97. Koga, M., Tatsumoto, M., Yuki, N., et al.: Range of cross reactivity of anti-GM1 IgG antibody in Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 71:123, 2001. 98. Koga, M., Yuki, N., Ariga, T., et al.: Is IgG anti-GT1a antibody associated with pharyngeal-cervical-brachial weakness or oropharyngeal palsy in Guillain-Barré syndrome? J. Neuroimmunol. 86:74, 1998. 99. Koga, M., Yuki, N., Takahashi, M., et al.: Close association of IgA anti-ganglioside antibodies with antecedent Campylobacter jejuni infection in Guillain-Barre and Fisher’s syndromes. J. Neuroimmunol. 81:138, 1998. 100. Kornberg, A. J., Pestronk, A., Blume, G. M., et al.: Selective staining of the cerebellar molecular layer by serum IgG in Miller-Fisher and related syndromes. Neurology 47:1317, 1996.

101. Kuroki, S., Saida, T., Nukina, M., et al.: Campylobacter jejuni strains from patients with Guillain-Barre syndrome belong mostly to Penner serogroup 19 and contain betaN-acetylglucosamine residues. Ann. Neurol. 33:243, 1993. 102. Kusunoki, S., Chiba, A., Kon, K., et al.: N-acetylgalactosaminyl GD1a is a target molecule for serum antibody in GuillainBarre syndrome. Ann. Neurol. 35:570, 1994. 103. Kusunoki, S., Hitoshi, S., Kaida, K., et al.: Monospecific anti-GD1b IgG is required to induce rabbit ataxic neuropathy. Ann. Neurol 45:400, 1999. 104. Kusunoki, S., Iwamori, M., Chiba, A., et al.: GM1b is a new member of antigen for serum antibody in Guillain-Barré syndrome. Neurology 47:237, 1996. 105. Kusunoki, S., Mashiko, H., Mochizuki, N., et al.: Binding of antibodies against GM1 and GD1b in human peripheral nerve. Muscle Nerve 20:840, 1997. 106. Kusunoki, S., Shimizu, J., Chiba, A., et al.: Experimental sensory neuropathy induced by sensitization with ganglioside GD1b. Ann. Neurol. 39:424, 1996. 107. Kuwabara, S., Asahina, M., Koga, M., et al.: Two patterns of clinical recovery in Guillain-Barre syndrome with IgG anti-GM1 antibody. Neurology 51:1656, 1998. 108. Kwa, M. S., van Schaik, I. N., Brand, A., et al.: Investigation of serum response to PMP22, connexin 32 and P0 in inflammatory neuropathies. J. Neuroimmunol. 116:220, 2001. 109. Landry, O.: Note sur paralysie ascendante aiguë. Gaz. Hebd. Med. Chir. 7:472, 1859. 110. Lastovica, A. J., Goddard, E. A., and Argent, A. C.: Guillain-Barre syndrome in South Africa associated with Campylobacter jejuni O:41 strains. J. Infect. Dis. 176(Suppl. 2):S139, 1997. 111. Lawn, N. D., Fletcher, D. D., Henderson, R. D., et al.: Anticipating mechanical ventilation in Guillain-Barre syndrome. Arch. Neurol. 58:893, 2001. 112. Leyden, E.: Ueber einen Fall von multipler Neuritis. Charite Annal. 206, 1880. 113. Li, C. Y., Xue, P., Gao, C. Y., et al.: Experimental Campylobacter jejuni infection in the chicken: an animal model of axonal Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 61:279, 1996. 114. Lichtenfeld, P.: Autonomic dysfunction in the Guillain-Barre syndrome. Am. J. Med. 50:772, 1971. 115. Lopez, P. H., Villa, A. M., Sica, R. E., et al.: High affinity as a disease determinant factor in anti-GM1 antibodies: comparative characterization of experimentally induced vs. disease-associated antibodies. J. Neuroimmunol. 128:69, 2002. 116. Lugaresi, A., Uncini, A., Torrieri, F., et al.: Acute motor axonal neuropathy with high titer IgG and IgA anti-GD1a antibodies following Campylobacter enteritis. J. Neurol. Sci. 147:193, 1997. 117. Lunn, M. P., Johnson, L. A., Fromholt, S. E., et al.: Highaffinity anti-ganglioside IgG antibodies raised in complex ganglioside knockout mice: reexamination of GD1a immunolocalization. J. Neurochem. 75:404, 2000. 118. Maehara, T., Ono, K., Tsutsui, K., et al.: A monoclonal antibody that recognizes ganglioside GD1b in the rat central nervous system. Neurosci. Res. 29:9, 1997.

The Guillain-Barré Syndromes 119. Massaro, M. E., Rodriguez, E. C., Pociecha, J., et al.: Nerve biopsy in children with severe Guillain-Barre syndrome and inexcitable motor nerves. Neurology 51:394, 1998. 120. McKhann, G. M., Cornblath, D. R., Griffin, J. W., et al.: Acute motor axonal neuropathy: a frequent cause of acute flaccid paralysis in China. Ann. Neurol. 33:333, 1993. 121. McKhann, G. M., Cornblath, D. R., Ho, T. W., et al.: Clinical and electrophysiological aspects of acute paralytic disease of children and young adults in northern China. Lancet 338:593, 1991. 122. McKhann, G. M., Griffin, J. W., Cornblath, D. R., et al.: Plasmapheresis and Guillain-Barre syndrome. I. Factors influencing outcome. Neurology 37(Suppl.):252, 1987. 123. McKhann, G. M., Griffin, J. W., Cornblath, D. R., et al.: Plasmapheresis and Guillain-Barre syndrome: analysis of prognostic factors and the effect of plasmapheresis. Ann. Neurol. 23:347, 1988. 124. Mendell, J. R., Kissel, J. T., Kennedy, M. S., et al.: Plasma exchange and prednisone in Guillain-Barre syndrome: a controlled randomized trial. Neurology 35:1551, 1985. 125. Mishu, B., Ilyas, A. A., Koski, C. L., et al.: Serologic evidence of previous Campylobacter jejuni infection in patients with the Guillain-Barre syndrome. Ann. Intern. Med. 118:947, 1993. 126. Mizoguchi, K., Hase, A., Obi, T., et al.: Two species of antiganglioside antibodies in a patient with a pharyngealcervical-brachial variant of Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 57:1121, 1994. 127. Moran, A. P., and Penner, J. L.: Serotyping of Campylobacter jejuni based on heat-stable antigens: relevance, molecular basis and implications in pathogenesis. J. Appl. Microbiol. 86:361, 1999. 128. Moran, A. P., Prendergast, M. M., and Hogan, E. L.: Sialosylgalactose: a common denominator of Guillain-Barré and related disorders? J. Neurol. Sci. 196:1, 2002. 129. Nachamkin, I., Engberg, J., Gutacker, M., et al.: Molecular population genetic analysis of Campylobacter jejuni HS:19 associated with Guillain-Barre syndrome and gastroenteritis. J. Infect. Dis. 184:221, 2001. 130. Nachamkin, I., Ung, H., Moran, A. P., et al.: Ganglioside GM1 mimicry in Campylobacter strains from sporadic infections in the United States. J. Infect. Dis. 179:1183, 1999. [Published erratum appears in J. Infect. Dis. 179:1593, 1999.] 131. Nadkarni, N., and Lisak, R. P.: Guillain-Barre syndrome (GBS) with bilateral optic neuritis and central white matter disease. Neurology 43:842, 1993. 132. Nagai, Y., Momoi, T., Saito, M., et al.: Ganglioside syndrome, a new autoimmune neurologic disorder, experimentally induced with brain gangliosides. Neurosci. Lett. 2:107, 1976. 133. Neisser, A., Bernheimer, H., Berger, T., et al.: Serum antibodies against gangliosides and Campylobacter jejuni lipopolysaccharides in Miller Fisher syndrome. Infect. Immun. 65:4038, 1997. 134. Neisser, A., Schwerer, B., Bernheimer, H., et al.: Gangliosideinduced antiganglioside antibodies from a neuropathy patient cross-react with lipopolysaccharides of Campylobacter jejuni associated with Guillain-Barre syndrome. J. Neuroimmunol. 102:85, 2000. 135. O’Hanlon, G. M., Paterson, G. J., Veitch, J., et al.: Mapping immunoreactive epitopes in the human peripheral nervous

136.

137.

138.

139.

140.

141.

142.

143.

144.

145. 146.

147.

148.

149.

150.

151.

2217

system using human monoclonal anti-GM1 ganglioside antibodies. Acta Neuropathol. 95:605, 1998. O’Hanlon, G. M., Paterson, G. J., Wilson, G., et al.: Anti-GM1 ganglioside antibodies cloned from autoimmune neuropathy patients show diverse binding patterns in the rodent nervous system. J. Neuropathol. Exp. Neurol. 55:184, 1996. O’Leary, C. P., Veitch, J., Durward, W. F., et al.: Acute oropharyngeal palsy is associated with antibodies to GQ1b and GT1a gangliosides. J. Neurol. Neurosurg. Psychiatry 61:649, 1996. O’Leary, C. P., and Willison, H. J.: Autoimmune ataxic neuropathies (sensory ganglionopathies). Curr. Opin. Neurol. 10:366, 1997. Odaka, M., Yuki, N., and Hirata, K.: Anti-GQ1b IgG antibody syndrome: clinical and immunological range. J. Neurol. Neurosurg. Psychiatry 70:50, 2001. Odaka, M., Yuki, N., Yamada, M., et al.: Bickerstaff’s brainstem encephalitis: clinical features of 62 cases and a subgroup associated with Guillain-Barre syndrome. Brain 126:2279, 2003. Ogawa-Goto, K., Funamoto, N., Abe, T., et al.: Different ceramide compositions of gangliosides between human motor and sensory nerves. J. Neurochem. 55:1486, 1990. Ogawara, K., Kuwabara, S., Mori, M., et al.: Axonal GuillainBarre syndrome: relation to anti-ganglioside antibodies and Campylobacter jejuni infection in Japan. Ann. Neurol. 48:624, 2000. Ogino, M., Orazio, N., and Latov, N.: IgG anti-GM1 antibodies from patients with acute motor neuropathy are predominantly of the IgG1 and IgG3 subclasses. J. Neuroimmunol. 58:77, 1995. Oomes, P. G., Jacobs, B. C., Hazenberg, M. P. H., et al.: AntiGM1 IgG antibodies and Campylobacter jejuni bacteria in Guillain-Barre syndrome: evidence of molecular mimicry. Ann. Neurol. 38:170, 1995. Osler, W.: The Principles and Practice of Medicine. New York, Appleton, 1892. Osterman, P. O., Lundemo, G., Pirskanen, R., et al.: Beneficial effects of plasma exchange in acute inflammatory polyradiculoneuropathy. Lancet 2:1296, 1984. Paparounas, K., O’Hanlon, G. M., O’Leary, C. P., et al.: Anti-ganglioside antibodies can bind peripheral nerve nodes of Ranvier and activate the complement cascade without inducing acute conduction block in vitro [see comments]. Brain 122(Pt. 5):807, 1999. Perkin, G. D., and Murray-Lyon, I.: Neurology and the gastrointestinal system. J. Neurol. Neurosurg. Psychiatry 65:291, 1998. Pestronk, A., Choksi, R., Yee, W. C., et al.: Serum antibodies to heparan sulfate glycosaminoglycans in Guillain-Barré syndrome and other demyelinating polyneuropathies. J. Neuroimmunol. 91:204, 1998. Plasma Exchange/Sandoglobulin Guillain-Barré Syndrome Trial Group: Randomised trial of plasma exchange, intravenous immunoglobulin, and combined treatments in Guillain-Barré syndrome. Lancet 349:225, 1997. Plomp, J. J., Molenaar, P. C., O’Hanlon, G. M., et al.: Miller Fisher anti-GQ1b antibodies: a-latrotoxin-like effects on motor end plates. Ann. Neurol. 45:189, 1999. [Published erratum appears in Ann. Neurol. 45:823, 1999.]

2218

Diseases of the Peripheral Nervous System

152. Prendergast, M. M., Lastovica, A. J., and Moran, A. P.: Lipopolysaccharides from Campylobacter jejuni O:41 strains associated with Guillain-Barre syndrome exhibit mimicry of GM1 ganglioside. Infect. Immun. 66:3649, 1998. 153. Prendergast, M. M., and Moran, A. P.: Lipopolysaccharides in the development of the Guillain-Barre syndrome and Miller Fisher syndrome forms of acute inflammatory peripheral neuropathies. J. Endotoxin Res. 6:342, 2000. 154. Prendergast, M. M., Willison, H. J., and Moran, A. P.: Human monoclonal immunoglobulin M antibodies to ganglioside GM1 show diverse cross-reactivities with lipopolysaccharides of Campylobacter jejuni strains associated with Guillain-Barre syndrome. Infect. Immun. 67:3698, 1999. 155. Prineas, J. W.: Acute idiopathic polyneuritis: an electron microscope study. Lab. Invest. 26:133, 1972. 156. Raine, C. S., Wisniewski, H., and Prineas, J.: An ultrastructural study of experimental demyelination and remyelination. Lab. Invest. 21:316, 1969. 157. Ramos-Alvarez, M., Bessudo, L., and Sabin, A.: Paralytic syndromes associated with noninflammatory cytoplasmic or nuclear neuronopathy: acute paralytic disease in Mexican children, neuropathologically distinguishable from LandryGuillain-Barre syndrome. JAMA 207:1481, 1969. 158. Raschetti, R., Maggini, M., Popoli, P., et al.: Gangliosides and Guillain-Barre syndrome. J. Clin. Epidemiol. 48:1399, 1995. 159. Ravn, H.: The Landry-Guillain-Barre syndrome. Acta Neurol. Scand. 30:8, 1967. 160. Rees, J. H., Gregson, N. A., and Hughes, R. A. C.: Antiganglioside GM1 antibodies in Guillain-Barre syndrome and their relationship to Campylobacter jejuni infection. Ann. Neurol. 38:809, 1995. 161. Rees, J. H., Soudain, S. E., Gregson, N. A., et al.: Campylobacter jejuni infection and Guillain-Barré syndrome. N. Engl. J. Med. 333:1374, 1995. 162. Reisin, R. C., Cersosimo, R., Garcia Alvarez, M., et al.: Acute “axonal” Guillain-Barre syndrome in childhood. Muscle Nerve 16:1310, 1993. 163. Rhodes, K. M., and Tattersfield, A. E.: Guillain-Barre syndrome associated with Campylobacter infection. Br. Med. J.(Clin. Res. Ed.) 285:173, 1982. 164. Roberts, M., Willison, H. J., Vincent, A., et al.: Multifocal motor neuropathy human sera block distal motor nerve conduction in mice. Ann. Neurol. 38:111, 1995. 165. Ropper, A. H.: Miller Fisher syndrome and other acute variants of Guillain-Barre syndrome. Baillieres Clin. Neurol. 3:95, 1994. 166. Ropper, A. H., and Kehne, S. M.: Guillain-Barre syndrome: management of respiratory failure. Neurology 35:1662, 1985. 167. Ropper, A. H., and Shahani, B. T.: Diagnosis and management of acute areflexic paralysis with emphasis on GuillainBarre syndrome. In Asbury, A. K., and Gilliatt, R. W. (eds.): Peripheral Nerve Disorders: A Practical Approach. London, Butterworths, p. 21, 1984. 168. Ross, J., and Bury, J. S.: A Treatise on Peripheral Neuritis. London, Charles Griffin, 1893. 169. Saez-Torres, I., Diaz-Villoslada, P., Martinez-Caceres, E., et al.: Gangliosides do not elicit experimental autoimmune

170.

171.

172. 173.

174.

175.

176.

177.

178.

179.

180.

181.

182.

183.

184.

185.

encephalomyelitis in Lewis rats and SJL mice. J. Neuroimmunol. 84:24–29, 1998. Saleh, M. N., Khazaeli, M. B., Wheeler, R. H., et al.: Phase I trial of the murine monoclonal anti-GD2 antibody 14G2a in metastatic melanoma. Cancer Res. 52:4342–4347, 1992. Salloway, S., Mermel, L. A., Seamans, M., et al.: Miller-Fisher syndrome associated with Campylobacter jejuni bearing lipopolysaccharide molecules that mimic human ganglioside GD3. Infect. Immun. 64:2945, 1996. Schonhofer, P. S.: Guillain-Barre syndrome and parenteral gangliosides. Lancet 338:757, 1991. Schwerer, B., Neisser, A., and Bernheimer, H.: Distinct immunoglobulin class and immunoglobulin G subclass patterns against ganglioside GQ1b in Miller Fisher syndrome following different types of infection. Infect. Immun. 67:2414, 1999. Sheikh, K. A., Deerinck, T. J., Ellisman, M. H., et al.: The distribution of ganglioside-like moieties in peripheral nerves. Brain 122(Pt. 3):449, 1999. Sheikh, K. A., Ho, T. W., Nachamkin, I., et al.: Molecular mimicry in Guillain-Barré syndrome. Ann. N. Y. Acad. Sci. 845:307, 1998. Sheikh, K. A., Nachamkin, I., Ho, T. W., et al.: Campylobacter jejuni lipopolysaccharides in Guillain-Barre syndrome: molecular mimicry and host susceptibility. Neurology 51:371, 1998. Shen, K. F., and Crain, S. M.: Nerve growth factor rapidly prolongs the action potential of mature sensory ganglion neurons in culture and this effect requires activation of Gs coupled excitatory kappa opioid receptors on these cells. J. Neurosci. 14:5570, 1994. Shin, J. E., Ackloo, S., Mainkar, A. S., et al.: Lipooligosaccharides of Campylobacter jejuni serotype O:10: structures of core oligosaccharide regions from a bacterial isolate from a patient with the Miller-Fisher syndrome and from the serotype reference strain. Carbohydr. Res. 305:223, 1997. Stojkovic, T., Verdin, M., Hurtevent, J. F., et al.: GuillainBarre syndrome resembling brainstem death in a patient with brain injury. J. Neurol. 248:430, 2001. Susuki, K., Yuki, N., Hirata, K., et al.: Fine specificities of anti-LM1 IgG antibodies in Guillain-Barre syndrome. J. Neurol. Sci. 195:145, 2002. Svennerholm, L., Bostrom, K., Fredman, P., et al.: Gangliosides and allied glycosphingolipids in human peripheral nerve and spinal cord. Biochim. Biophys. Acta 1214:115, 1994. Tagawa, Y., Yuki, N., and Hirata, K.: High anti-GM1 and anti-GD1a IgG antibody titers are detected in GuillainBarre syndrome but not in chronic inflammatory demyelinating polyneuropathy. Eur. Neurol. 48:118, 2002. Takigawa, T., Yasuda, H., and Kikkawa, R.: Antibodies against GM1 ganglioside affect K⫹ and Na⫹ currents in isolated rat myelinated nerve fibers. Ann. Neurol. 37:436, 1995. ter Bruggen, J. P., van der Meche, F. G., de Jager, A. E., et al.: Ophthalmoplegic and lower cranial nerve variants merge into each other and into classical Guillain-Barré syndrome. Muscle Nerve 21:239, 1998. Trojaborg, W.: Acute and chronic neuropathies: new aspects of Guillain-Barré syndrome and chronic inflammatory

The Guillain-Barré Syndromes

186.

187.

188.

189.

190.

191.

192.

193.

194.

195.

196.

197.

198.

199.

200.

demyelinating polyneuropathy, an overview and an update. Electroencephalogr. Clin. Neurophysiol. 107:303, 1998. van der Meché, F. G. A., and Schmitz, P. I. M., for the Dutch Guillain-Barre Study Group: A randomized trial comparing intravenous immune globulin and plasma exchange in Guillain-Barre syndrome. N. Engl. J. Med. 326:1123, 1992. van Koningsveld, R., Schmitz, P. I. M., van der Meche, F. G. A., et al.: Effect of methylprednisolone when added to standard treatment with intravenous immunoglobulin for Guillain-Barre syndrome: randomised trial. Lancet 363:192, 2004. Vargas, F., Hilbert, G., Gruson, D., et al.: Fulminant GuillainBarre syndrome mimicking cerebral death: case report and literature review. Intensive Care Med. 26:623, 2000. Visser, L. H., van der Meché, F. G. A., Meulstee, J., et al.: Cytomegalovirus infection and Guillain-Barré syndrome: the clinical, electrophysiologic, and prognostic features. Neurology 47:668, 1996. Vriesendorp, F. J., Mishu, B., Blaser, M., et al.: Serum antibodies to GM1, peripheral nerve myelin, and Campylobacter jejuni in patients with Guillain-Barre syndrome and controls: correlation and prognosis. Ann. Neurol. 34:130, 1993. Waller, A.: Experiments on the section of glossopharyngeal and hypoglossal nerves of the frog and observations of the alternatives produced thereby in the structure of their primitive fibers. Philos. Trans. R. Soc. Lond. 140:423, 1850. Walsh F. S., Cronin, M., Koblar, S., et al.: Association between glycoconjugate antibodies and Campylobacter infection in patients with Guillain-Barre syndrome. J. Neuroimmunol. 34:43, 1991. Weber, F., Rudel, R., Aulkemeyer, P., et al.: Anti-GM1 antibodies can block neuronal voltage-gated sodium channels. Muscle Nerve 23:1414, 2000. Willison, H. J., Almemar, A., Veitch, J., et al.: Acute ataxic neuropathy with cross-reactive antibodies to GD1b and GD3 gangliosides. Neurology 44:2395, 1994. Willison, H. J., O’Hanlon, G. M., Paterson, G., et al.: A somatically mutated human antiganglioside IgM antibody that induces experimental neuropathy in mice is encoded by the variable region heavy chain gene, V1–18. J. Clin. Invest. 97:1155, 1996. Willison, H. J., and Veitch, J.: Immunoglobulin subclass distribution and binding characteristics of anti-GQ1b antibodies in Miller Fisher syndrome. J. Neuroimmunol. 50:159, 1994. Willison, H. J., Veitch, J., Paterson, G., Kennedy, P. G.: Miller Fisher syndrome is associated with serum antibodies to GQ1b ganglioside. J. Neurol. Neurosurg. Psychiatry 56:204, 1993. Winer, J. B., Hughes, R. A. C., and Osmond, C.: A prospective study of acute idiopathic neuropathy. 1. Clinical features and their prognostic value. J. Neurol. Neurosurg. Psychiatry 51:605, 1988. Winer, J. B., Hughes, R. A. C., and Osmond, C.: A prospective study of acute idiopathic neuropathy. 3. Immunological studies. J. Neurol. Neurosurg. Psychiatry 51:619, 1988. Wirguin, I., Briani, C., Suturkova-Milosevic, L., et al.: Induction of anti-GM1 ganglioside antibodies by Campylo-

201.

202.

203.

204.

205.

206.

207.

208.

209.

210.

211.

212.

213.

214.

215.

216.

217.

2219

bacter jejuni lipopolysaccharides. J. Neuroimmunol. 78:138, 1997. Wirguin, I., Suturkova-Milosevic, L., Della-Latta, P., et al.: Monoclonal IgM antibodies to GM1 and asialo-GM1 in chronic neuropathies cross-react with Campylobacter jejuni lipopolysaccharides. Ann. Neurol. 35:698, 1994. Wu, H.-S., Liu, T. C., Lu, Z. L., et al.: A prospective clinical and electrophysiologic survey of acute flaccid paralysis in Chinese children. Neurology 49:1723, 1997. Yoshino, H.: Distribution of gangliosides in the nervous tissues recognized by axonal form of Guillain-Barre syndrome [in Japanese]. Neuroimmunology 5:174, 1997. Yuki, N.: Successful plasmapheresis in Bickerstaff’s brain stem encephalitis associated with anti-GQ1b antibody. J. Neurol. Sci. 131:108, 1995. Yuki, N., Handa, S., Tai, T., et al.: Ganglioside-like epitopes of lipopolysaccharides from Campylobacter jejuni (PEN 19) in three isolates from patients with Guillain-Barre syndrome. J. Neurol. Sci. 130:112, 1995. Yuki, N., Ichihashi, Y., and Taki, T.: Subclass of IgG antibody to GM1 epitope-bearing lipopolysaccharide of Campylobacter jejuni in patients with Guillain-Barre syndrome. J. Neuroimmunol. 60:161, 1995. Yuki, N., Sato, S., Fujimoto, S., et al.: Serotype of Campylobacter jejuni, HLA, and the Guillain-Barre syndrome. Muscle Nerve 16:968, 1992. Yuki, N., Sato, S., Tsuji, S., et al.: Frequent presence of anti-GQ1b antibody in Fisher’s syndrome. Neurology 43:414, 1993. Yuki, N., Sato, S., Tsuji, S., et al.: An immunologic abnormality common to Bickerstaff’s brain stem encephalitis and Fisher’s syndrome. J. Neurol. Sci. 118:83, 1993. Yuki, N., Taki, T., Inagaki, F., et al.: A bacterium lipopolysaccharide that elicits Guillain-Barre syndrome has a GM1 ganglioside-like structure. J. Exp. Med. 178:1771, 1993. Yuki, N., Taki, T., Takahashi, M., et al.: Molecular mimicry between GQ1b ganglioside and lipopolysaccharides of Campylobacter jejuni isolated from patients with Fisher’s syndrome. Ann. Neurol. 36:791, 1994. Yuki, N., Yamada, M., Koga, M., et al.: Animal model of axonal Guillain-Barre syndrome induced by sensitization with GM1 ganglioside. Ann. Neurol. 49:712, 2001. Yuki, N., Yamada, M., Sato, S., et al.: Association of IgG anti-GD1a antibody with severe Guillain-Barre syndrome. Muscle Nerve 16:642, 1993. Yuki, N., Yamada, M., Tagawa, Y., et al.: Pathogenesis of the neurotoxicity caused by anti-GD2 antibody therapy. J. Neurol. Sci. 149:127, 1997. Yuki, N., Yoshino, H., Sato, S., et al.: Acute axonal polyneuropathy associated with anti-GM 1 antibodies following Campylobacter jejuni enteritis. Neurology 40:1900, 1990. Yuki, N., Yoshino, H., Sato, S., et al.: Severe acute axonal form of Guillain-Barre syndrome associated with IgG anti-GD1a antibodies. Muscle Nerve 15:899, 1992. Zhang, G., Lopez, P. H., Li, C. Y., et al.: Anti-ganglioside antibody-mediated neuronal cytotoxicity and its protection by intravenous immunoglobulin: implications for immune neuropathies. Brain 127(Pt. 5):1085, 2004.

99 Chronic Inflammatory Demyelinating Polyradiculoneuropathy ANGELIKA F. HAHN, HANS-PETER HARTUNG, AND PETER J. DYCK

Clinical and Diagnostic Features Overview History Epidemiology Genetics Clinical Manifestations Preceding Infections or Immunizations Age of Onset and Its Influence on Clinical Features Early Symptoms and Signs Neurologic Findings

Diagnostic Criteria Differential Diagnostic Considerations Disease Course and Prognosis CIDP in Childhood Clinical Spectrum Laboratory Investigation Treatment History of Immunotherapy and General Remarks Review of RCTs and of Observations of Long-Term Use of Treatments

CLINICAL AND DIAGNOSTIC FEATURES Overview Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is an acquired peripheral neuropathy of presumed autoimmune etiology, which presents either as a chronically progressive or relapsing-remitting disorder. Symptoms develop subacutely or insidiously over weeks to months. Although spontaneous remissions may occur, as a rule the disease progresses steadily, but it will respond to immune-modulating therapies. Because of its progressive course, CIDP has been set apart from the self-limited acute inflammatory demyelinating polyneuropathy (AIDP) or “Guillain-Barré syndrome” (GBS). CIDP should also be separated from various disease-associated chronic polyradiculoneuropathies (see Table 99–4 later), specifically those associated with monoclonal proteins of unknown significance (MGUS) or with lymphoproliferative disorders; those occurring with diabetes mellitus, uremia, or intermittent porphyria; and those occurring with infections caused by hepatitis or the human immunodeficiency virus (HIV type 1). Hereditary neuropathies such as Charcot-Marie-

General Approaches to Management Pathology History Pathologic Hallmarks Other Pathologic Causes of Chronic Polyradiculopathies Sural Nerve Biopsy Pathogenesis Cellular Immune Responses Humoral Immune Responses Axonal Loss

Tooth disease (CMT) may need to be excluded by genomic DNA testing. CIDP can present in a spectrum of clinical manifestations.

History Early clinical reports, reviewed in previous editions, date back nearly a century and draw attention to recurrent or progressive nonfamilial hypertrophic neuropathies observed in adult patients. Harris and Newcomb76 gave a detailed description of the underlying pathology in nerve roots and major nerve trunks. They pointed out that, in the grossly enlarged nerves, axis cylinders were usually preserved but were devoid of myelin and surrounded by proliferating Schwann cells, giving an “onion bulb” appearance. They interpreted these histopathologic changes accurately as signs of chronically recurring demyelination. In a review of recurrent polyradiculoneuropathy, Austin10 surveyed 32 cases (including 2 of his own) with symmetrical and predominantly motor polyneuropathy of fluctuating course. Over weeks to months, paralysis of variable severity had developed to reach a plateau from which gradual 2221

2222

Diseases of the Peripheral Nervous System

improvement could occur, either spontaneously or in response to corticosteroids. Over 5 years, up to 20 relapses could be studied in a single patient who had been treated with adrenocorticotropic hormone (ACTH) or prednisone. In fact, the patient appeared to have become “steroiddependent,” because he worsened each time the dose was reduced or the drug was discontinued. Response to steroids had been ascertained by use of placebo and by quantitative measurements of neurologic function obtained in blinded assessments. During relapses, deterioration of motor functions could be fairly rapid, resembling AIDP. Austin noted, however, that unlike the acute forms, cases of chronically recurrent polyneuropathy were much more rarely associated with antecedent infections. He speculated on similarities of recurrent neuropathies and experimental allergic neuritis (EAN), a disease produced in animals by sensitization with either nerve extracts or myelin.228,229 Moreover, in a remarkably astute observation, Austin reasoned that the finding of muscle weakness in the absence of muscle atrophy meant that the propagation of nerve impulses had to be impaired by “conduction block” (CB), caused by focal or segmental demyelination. Hypertrophic nerve changes, seen in one third of cases, appeared to reflect the chronicity of the disorder. Adams and colleagues4 first reported a variant form of recurrent focal or multifocal neuropathy with hypertrophic nerve changes and a similar steroid responsiveness. Most commonly, the disease affected the upper limbs and often began with painful paresthesias, followed later by weakness in the territories of single or multiple nerves. Despite limited clinical manifestations, such patients suffered from a generalized or systemic demyelinating neuropathy as shown in sural nerve biopsies. In the late 1960s and the 1970s, improved neurophysiologic techniques and meticulous study of nerve biopsies, including the analysis of teased nerve fibers and examination by electron microscopy, allowed for a much better understanding of nerve pathology. Hinman and Magee81 described four cases of an acquired predominantly motor polyneuropathy that resembled GBS but had a much more chronic and sometimes fluctuating or relapsing course. They concluded that such cases represented a chronic variant of the acute syndrome. Dyck and colleagues43 reported two cases of chronic relapsing sensorimotor neuropathy with elevated cerebrospinal fluid (CSF) protein. Originally these had been diagnosed as GBS. Thomas and co-workers200 added observations of four juvenile patients with “recurrent” or “relapsing” GBS. Their attacks had all been triggered by antecedent infections and had been improved with prednisone. In one older patient the disease had been chronically progressive over many months, with only minor spontaneous fluctuations. In all five patients CSF protein levels were substantially elevated, and nerve biopsies or autopsy materials showed extensive demyelination, even though there was only limited break-

down or loss of axons. The changes were accompanied by perivenular collections of lymphocytes and were most prominent in nerve roots. Thomas and co-workers considered these cases to be variants of acute Guillain-Barré polyneuritis. CIDP as a Separate Nosologic Entity From a review of 53 patients they had evaluated personally, Dyck and associates46 first coined the descriptive term chronic inflammatory polyradiculoneuropathy by delineating the natural history, clinical and electrodiagnostic characteristics, typical CSF findings, and pathologic features of this separate entity. Their review included cases with predominantly motor, predominantly sensory, or mixed sensorimotor presentations, the latter being the most common. The course could be recurrent or relapsing, stepwise progressive, or monophasic and steadily progressive, eventually leading to marked disability. Dyck and co-workers50 also demonstrated that cases with a chronic progressive course would improve with corticosteroids as did those with the relapsing and remitting form.162 This supported the concept that both disease presentations form a single entity. Electrophysiologic examinations were typical for generalized and nonuniform slowing of conduction velocities in both motor and sensory nerves, with associated partial CB caused by multifocal demyelination observed in spinal roots, proximal nerve trunks, major nerve plexuses, and peripheral nerves. Macrophage-induced, segmental demyelination was usually accompanied by nerve edema and infiltrates of mononuclear inflammatory cells.46,161,162,202 Hence, the term was later changed to chronic inflammatory demyelinating polyradiculoneuropathy. The two inflammatory demyelinating polyradiculoneuropathies, AIDP and CIDP, typically show a cytoalbuminologic dissociation in CSF. The mechanisms underlying AIDP and CIDP are not yet fully known. It is possible that both syndromes are variants of the same disorder, as their shared pathologic features suggested. However, failing a complete understanding, there are cogent reasons for differentiating the two conditions, even though this distinction ultimately may prove invalid. First, AIDP and CIDP differ in temporal evolution. In AIDP neurologic deficits develop rapidly over a few days or weeks (usually ⬍ 4 weeks), crest, and then subside spontaneously and slowly. Recurrences are rare. Between episodes the neurologic recovery is usually complete, including the normalization of CSF protein.69 By contrast, in CIDP the neurologic deficits develop slowly over weeks, months, or even years (arbitrarily defined minimum progression of 8 weeks). A more rapid evolution, quasi simulating AIDP, can occasionally be seen in children and young adults. For such cases the correct diagnosis will be made often only on follow-up, when relapses have occurred.134,143,187 Rare cases, appropriately named “subacute idiopathic demyelinating polyradiculoneuropathy” (SIDP), may show a progressive phase of 6 to 8 weeks and

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

thus fall between the conventionally defined timing of CIDP and AIDP.84,147 As a rule, evolution and persistence of neurologic impairment are clearly different in CIDP and AIDP. Second, the two syndromes differ in their relationship to infections. In nearly 80% of AIDP cases one can identify an antecedent viral or bacterial infection, whereas fewer than 30% of CIDP patients remember an infection that might have occurred up to 3 months prior to the onset of their symptoms.46 A third reason for distinguishing between the two syndromes, clearly demonstrated in systematic reviews, is the difference in their response to corticosteroids. Steroids provide clear-cut benefit for most patients with CIDP,124 whereas they were shown to be ineffective in AIDP.86 In the last two decades, several large CIDP patient series from many parts of the world—Australia, North America, Europe, and Japan—have been published.13,19,50,80,115,120,139,183,185 They have served to delineate the diagnostic criteria, clinical findings, and natural history, as well as electrodiagnostic and pathologic features of idiopathic CIDP. Yet they also demonstrated heterogeneity in the clinical presentation, thus providing strong arguments in favor of a broader conceptualization of CIDP.

Epidemiology Few systematic population-based studies have examined the prevalence of CIDP using strictly defined diagnostic criteria.104,113,123 Estimates have ranged from a crude prevalence of 0.81 per 100,000 population in a Japanese prefecture, and 1.24 per 100,000 population in a region of southeast England, to approximately 1.9 per 100,000 population in New South Wales, Australia. In part these variances probably reflect accuracy of case ascertainment. In surveying the fairly uniform and stable population of nearly 6 million people of New South Wales, the Australian study calculated a crude annual incidence of 0.15 per 100,000 population, based on a total of 87 new CIDP cases reported during the decade from 1986 to 1995. This sample included cases with “probable” and “possible” CIDP, as defined by the American Academy of Neurology (AAN) Ad Hoc Subcommittee research criteria for the diagnosis of CIDP.6 Although the disease was shown to affect all ages, CIDP was more commonly found in males and in older individuals (mean age at onset, 48 years). For the 70- to 79-year age group, the age-specific prevalence overall was as high as 6.7 per 100,000 population, and for male patients it even reached 9.5 per 100,000 population.123 Thus onset of CIDP is common in the fifth to seventh decade, and in this age group the disease course is more likely monophasic and progressive. Patients with relapsing-remitting CIDP, who made up between 40% and 60% of cases in several large patient series, tended to have an earlier onset and responded very well to therapy.80,120,162,187

2223

Genetics Several studies have examined a possible genetically determined susceptibility for CIDP, yet they included only a small number of patients and only those with recurrent or chronically relapsing disease. Results derived from the study of 16 Australian patients indicated a strong association with human leukocyte antigen (HLA)–linked genetic factors, namely HLA-AW30, -AW31, and -DW3.194 However, these findings were not confirmed in a contemporaneous study of 14 English patients, which suggested instead an association with the HLA-A1, -B8, -DRw3, and -Dw3 haplotypes.3 The possible association of CIDP with HLA class I and II antigens was reexamined in a much larger cohort of 71 CIDP patients from New South Wales, Australia.55 Once again the results differed from prior observations insofar as they identified an increased frequency of the HLA-A3, -B7, and -DR2 haplotypes and a decreased frequency of HLA-44. A follow-up study of 31 CIDP patients from the United Kingdom differed again by showing the strongest association with HLA-Cw7.220 Moreover, a subsequent large case study from Holland indicated that the frequencies of certain HLA antigens in CIDP were not significantly increased when the P values were corrected for the number of HLA allotypes tested.216 We have made similar observations in a large series of Canadian CIDP patients, who showed no strong HLA associations (unpublished results). In all these studies the number of patients was probably not large enough to prove or disprove the association of susceptibility for CIDP and genes of the HLA major histocompatibility complex (MHC).

CLINICAL MANIFESTATIONS In typical CIDP motor and sensory deficits develop insidiously over weeks to months (arbitrarily defined minimum of 8 weeks) or even years, eventually leading to significant disability.13,19,22,46,115,120,183,186 Occasionally, the presentation may be more acute or subacute and not unlike GBS, and subsequently the disease will assume a relapsing or progressive course.84,134,143 This is particularly true for children, in whom the onset of symptoms is often more precipitous.80,100,120,187

Preceding Infections or Immunizations Whereas AIDP can often be linked to a preceding viral or bacterial infection, this association is much less apparent in CIDP. However, in view of the insidious onset and evolution of CIDP and the common delay in making the diagnosis (on average 6 to 12 months from onset of symptoms), patients may simply no longer recall such prodromal events. For instance, Dyck and colleagues46 recorded a respiratory tract infection in 4 of 53 patients, which had occurred within 3 to

2224

Diseases of the Peripheral Nervous System

6 months before the onset of CIDP; a further 6 patients had received either vaccinations, desensitization injections, or blood transfusions (total 19%). Yet McCombe and associates120 found a history of infections or vaccinations within 6 weeks from onset of symptoms in 29 of their 92 CIDP patients (32%), the great majority being upper respiratory tract infections. A similar frequency of antecedent events (29%) was recorded by Simmons and colleagues185 in their series of 77 adult patients with idiopathic CIDP: upper respiratory tract infections (12 patients), gastroenteritis (2), other infections (4), vaccination (2), surgery (2), and intraarticular cortisone injection (1). An upper respiratory tract infection was also the most common precipitant of CIDP in children. In 30% to 55% of reported cases, upper respiratory tract infections or gastroenteritis preceded the onset of neurologic symptoms.79,100,140,170 Of 100 adult CIDP patients studied by Said and colleagues,19,173 a total of 16 patients reported an infection within 6 weeks before onset of neurologic symptoms; 9 of these had a flulike syndrome and in the remaining 7 CIDP was associated with viral hepatitis. A further three patients had undergone a surgical procedure shortly before the onset of the neuropathy, and in one woman CIDP became manifest during the postpartum period. McCombe and colleagues119 had drawn attention to the fact, that in women of childbearing age, onset or relapse of CIDP could occur more frequently during pregnancy or the postpartum period. Such increased risk of developing CIDP under this circumstance may be related to the immunologic adjustments during or following the period of gestational immunosuppression. The observation that onset or relapse of CIDP was linked to infections or immunization in 19% to 32% of reported cases suggests that the association is higher than expected by chance alone. However, none of the reported studies had examined in parallel the incidence of infections in control populations. Therefore, a direct or indirect relationship between CIDP and the preceding events remains to be established. Often the question arises whether the child with CIDP may safely receive routine immunizations. A recent study found the risks of relapse of CIDP to be low with booster injections, unless the disease had been originally triggered by a vaccination.85 In susceptible CIDP patients, booster tetanus injections have, on occasion, led to a relapse.156 CIDP also occurs in the context of HIV type 1 infections, where it has been observed most commonly at the time of seroconversion or in early stages of the infection, before acquired immunodeficiency syndrome or overt immune suppression.150,222,236 Presentation and course do not differ from cryptogenic CIDP, yet in the HIV-infected patient one typically finds in CSF a lymphocytic pleocytosis of 10 to 50 cells/mm3, as well as elevated CSF protein. The pathology of HIV-associated CIDP, as observed in sural nerve biopsies, is similar to that of regular CIDP cases.190

Age of Onset and Its Influence on Clinical Features CIDP can occur at all ages. The disease is relatively uncommon in childhood, yet rare cases with onset at birth or in early infancy have been described.189 Commonly, children present with a symmetrical and predominantly motor polyneuropathy that manifests clinically as gait disturbances.30 In this age group, onset and progression of CIDP is often subacute, and afterward the disease may assume a relapsing-remitting course. Response to therapy and overall prognosis are usually excellent, although longterm immune modulation may occasionally be required.80 Prevalence of CIDP increases with age; it is particularly high in the fifth to seventh decade. In this age group, CIDP presents commonly as a symmetrical sensory and motor polyneuropathy of insidious onset and chronic progression. A relapsing disease course is much less common, and in general the prognosis is less favorable.19,80

Early Symptoms and Signs At presentation, most patients report symptoms that reflect a combination of proximal and distal limb muscle weakness and sensory loss. Cranial nerves may be affected, in particular the oculomotor and facial nerves. Usually, motor dysfunctions predominate and result in gait disturbances with a tendency to trip or fall, as well as difficulty with stairs or with rising from a seated position. Consequently, patients may require ambulatory aids such as a cane or a walker, or they may even become wheelchair dependent or bedbound. Weakness and clumsiness of arms and hands may lead to impaired dexterity and to difficulties in writing or in carrying out the tasks of personal hygiene. Frequently, patients report numbness, an “asleep” feeling, or stiffness and “deadness” in the hands or feet that leads to difficulties in buttoning a shirt or blouse, to trouble identifying or picking up coins, and to gait imbalance. In childhood, disturbance of gait or loss of ambulation is a common presenting feature of CIDP.30,141 Paresthesias such as tingling or prickling are felt frequently, whereas painful symptoms such as aching, burning, searing, or jabbing pains are much less common. In 5% to 8% of cases sensory symptoms predominate or are the only manifestation of CIDP, while coarse tremors and accompanying limb and gait ataxia reflect the impaired proprioception.46,120,146,213 Difficulty with balance will be noted particularly when going down stairs, because the patient will have trouble in gauging the distance of the steps and will eagerly seek support from the railing. Imbalance is also very evident in the shower or on attempting to go to the bathroom at night, when visual cues are reduced. On the whole, the severity of symptoms varies and may range from mild kinesthetic dysfunction to total helplessness.146,206

2225

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

Table 99–1 gives a summary of the clinical profiles of CIDP from five large patient series and of the presenting symptoms at the time of referral.

Neurologic Findings At the time of referral, observed neurologic signs usually correspond to symptoms mentioned earlier. Often motor findings predominate, and one is likely to find a symmetrical but variable degree of weakness in muscles of the shoulder and hip girdle. This is frequently accompanied by weakness in forearms and hands and in distal legs and feet. In fact, distal limbs are often most severely affected and grip strength is markedly reduced. Early in the disease, there is usually little muscle atrophy, yet weakness may be profound. Deep tendon reflexes are commonly absent or are markedly attenuated. On sensory testing, vibration and position sense, modalities mediated by large and heavily myelinated fibers, are most severely affected and are attenuated in a distal-to-proximal gradient. The loss of proprioception may lead to involuntary, so-called pseudoathetoid movements in the outstretched fingers, or else a coarse action tremor. In addition, stance and gait may be severely ataxic, which will be even more pronounced when eliminating visual cues on closing the eyes. Other modalities of sensation (touch, pinprick, and temperature sense) are usually reduced to a lesser degree. Involvement of cranial nerves (ophthalmoparesis or facial and bulbar weakness) may be observed in approximately 15% of cases. In very chronic cases one may find papilledema, whereby protein in the CSF is markedly elevated, presumably interfering with CSF absorption. Respiratory muscles may also be weakened, yet tracheal

intubation and assisted ventilation are seldom required.196 Eventually the patient may become wheelchair bound, or else the neuropathy advances to a point of rendering the patient bedbound. Disturbances of micturition (difficulty with voiding or urinary urgency) are seen in approximately 25% of CIDP cases and are mainly due to disturbed bladder sensation and detrusor areflexia.174 Moreover, patients with longstanding CIDP can have symptoms typical of lumbar stenosis and cauda equina dysfunction (posture-dependent lower back pain, lumbar radicular pain, sphincter and sexual dysfunctions, and exercise-related transient worsening in sensory and motor deficits). In these cases, massive hypertrophy of nerve roots produces crowding and entrapment of the roots in the lumbar thecal sack and the lumbar spinal canal, including the neural foramina.59,63,179 Signs of spinal cord compression from hypertrophied intradural nerve roots in the cervical or thoracic region may occasionally result in extensor plantar responses.128 In this circumstance, surgical decompression and a dural plasty may be considered.102

Diagnostic Criteria 1. Idiopathic CIDP presents characteristically as symmetrical polyradiculoneuropathy, with weakness in proximal and distal limb muscles as a predominant feature. Sensation, especially proprioception, is often impaired, and patients may experience numbness and paresthesias. 2. Both motor and sensory fibers are affected by multifocal inflammatory demyelination, resulting in a generalized neuropathy. Cranial nerves may also be involved.

Table 99–1. Clinical Profiles and Neurologic Symptoms of CIDP at Time of Referral in Five Large Series of Patients

Age at onset Range (years) Male/female ratio Motor symptoms: Weakness UL ⫽ LL Weakness UL ⬎ LL Sensory symptoms: Numbness/paresthesias Spontaneous pain Motor and sensory symptoms Cranial nerve involvement Bulbar dysfunction CSF protein ⬎ 45 mg/dL Relapsing vs. progressive (%)

Dyck et al.46 53 pts

McCombe et al.120 92 pts

Barohn et al.13 60 pts

Said173 100 pts

Hattori et al.80 124 pts

2–70 2/1

2–72 1.6/1

10–77 1.5/1

10–82 2/1

2–90 2/1

85%

94%

100%

83% 10%

72% 28%

79% 17% 84% 10% 10% 90% 36/64

64% 20% 72% 16% 6% N/R 65/35

86%

89% 8% 72% 22% 2% 86% N/R

71% N/R 72% 7%

86% 16% 95% 47/53

CSF ⫽ cerebrospinal fluid; LL ⫽ lower limbs; N/R ⫽ not recorded; pts ⫽ patients; UL ⫽ upper limbs.

100% 15/85

2226

Diseases of the Peripheral Nervous System

3. The course of CIDP may be continuously or stepwise progressive, or relapsing. Disease progression for longer than 8 weeks, or a relapsing course, are mandatory for diagnosis. 4. The diagnosis of CIDP is supported by the following laboratory findings: a. The protein content in the CSF is usually elevated, and the lymphocyte count should be less than 10/mm3. b. Electrodiagnostic examinations show unequivocal evidence of demyelination (see Laboratory Investigation later). c. Characteristic pathologic findings of inflammatory demyelination are found in the peroneal or sural nerve biopsies; usually some axonal degeneration is also present. In many instances the underlying pathology can be inferred from the clinical and electrophysiologic examinations.

5. At times, a trial of therapy may assist the diagnosis of CIDP, if quantitative clinical assessments and follow-up electrodiagnostic studies show unequivocal improvement. Proposed Diagnostic Criteria of CIDP for Clinical Trials From their retrospective review of a large number of cases, Barohn and co-workers13 formulated the first set of mandatory criteria for the diagnosis of CIDP (Table 99–2). These included (1) presence of proximal and distal weakness, hyporeflexia, or areflexia, and (2) disease progression for longer than 2 months. Utilizing the previously defined laboratory parameters (i.e., elevated protein concentration in the CSF, and evidence of demyelination in the electrodiagnostic examinations and in the nerve biopsy), three levels of confidence for the diagnosis of CIDP were formulated:

Table 99–2. Comparison of Diagnostic Criteria for Chronic Inflammatory Demyelinating Polyneuropathy Barohn et al.13 Mandatory Clinical Features Pattern of clinical Symmetrical proximal ⫹ distal involvement weakness

Reflexes Time course

Areflexia or hyporeflexia At least 2 months

Laboratory Features Electrodiagnostic Motor conduction velocity studies reduced ⬍ 70% of the lower limit of normal Cerebrospinal fluid CSF protein ⬎ 45 mg/dL studies Nerve biopsy features

Predominant features of demyelination†

AAN Ad Hoc Subcommittee6

Saperstein et al.178

Motor and/or sensory dysfunction involving more than one limb Areflexia or hyporeflexia At least 2 months

Major: symmetrical proximal ⫹ distal weakness Minor: exclusively distal weakness or sensory loss Areflexia or hyporeflexia At least 2 months

See Table 99–3

See Table 99–3

Mandatory: cell count ⬍ 10/mm3*; negative VDRL Supportive: elevated protein Unequivocal evidence of demyelination and remyelination

Mandatory: protein ⬎ 45 mg/dL

Inflammation Requirements for the Diagnostic Categories Definite Clinical, electrodiagnostic, CSF and biopsy

Supportive: cell count ⬍ 10/mm3* Predominant features of demyelination† Inflammation (not required)

Clinical, electrodiagnostic, CSF‡ and biopsy

Probable

Clinical and 2 out of 3 laboratory features

Clinical, electrodiagnostic and CSF‡

Possible

Clinical and 1 out of 3 laboratory features

Clinical and electrodiagnostic

Clinical major, electrodiagnostic, and CSF (biopsy supportive, but not mandatory) Clinical major, electrodiagnostic, and biopsy; or clinical major, CSF and biopsy Clinical major and 1 out of 3 laboratory features; or clinical minor and 2 out of 3 laboratory features

AAN ⫽ American Academy of Neurology; CSF ⫽ cerebrospinal fluid; VDRL ⫽ Venereal Disease Research Laboratory. *A CSF cell count greater than 10/mm3 (or ⬎ 50/mm3 if human immunodeficiency virus [HIV] seropositive) should prompt evaluation for HIV, Lyme disease and lymphomatous or leukemic infiltration of nerve roots. † To include segmental demyelination, remyelination, and onion bulb formation. ‡ The AAN criteria require only a normal cell count and a negative VDRL (elevated protein is not a mandatory feature). From Saperstein, D. S., and Barohn, R. J.: Current concepts and controversy in chronic inflammatory demyelinating polyneuropathy. Curr. Neurol. Neurosci. Rep. 3:57, 2003, with permission.

2227

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

“definite” (all three laboratory tests abnormal), “probable” (two abnormal tests), and “possible” (one abnormal test). These criteria were devised to assist clinicians in making a timely diagnosis of CIDP. In 1991, an Ad Hoc Subcommittee of the AAN6 modified these criteria to facilitate the selection of representative and fairly uniform cases for inclusion into clinical trials. The requirement of symmetrical distribution of muscle weakness was eliminated, CSF changes were deemphasized, and the electrodiagnostic findings were given greater importance (Table 99–3). Again three levels of confidence for the diagnosis of CIDP were defined. The AAN diagnostic criteria were adopted for most subsequent clinical trials. Recently they have been criticized for being on the one hand very specific but on the other hand not sensitive enough and too restrictive.83,142,178 For instance, in several retrospective analyses it was shown that only two thirds of patients who had been diagnosed with CIDP by a neuromuscular specialist fulfilled the stipulated AAN electrodiagnostic criteria for demyelination.22,23,142,177 Saperstein and colleagues178 revised the clinical and electrophysiologic criteria for the diagnosis of CIDP to be used not only as a research tool, but also for regular clinical

decision making (see Tables 99–2 and 99–3). Categories of “definite,” “probable,” and “possible” CIDP were upheld, as the AAN research criteria stipulated. However, patients of all categories were considered to suffer from a potentially treatable demyelinating neuropathy. Treatment response and overall prognosis can of course be predicted with greater confidence with a diagnostic certainty of definite or probable CIDP. Nonetheless, a trial of therapy is warranted even with a diagnosis of possible CIDP, although the response is less certain. Diagnostic criteria proposed by Saperstein and colleagues distinguish between “major” clinical criteria (presence of symmetrical proximal and distal weakness, typical for idiopathic CIDP) and “minor” criteria (signs of exclusively distal weakness and/or sensory loss). A diagnosis of definite or probable CIDP can only be made if the major clinical criteria are met and are supported by electrodiagnostic and CSF and/or nerve biopsy findings. In cases in which only the minor criteria are fulfilled, support from two of the three laboratory tests is required and only a diagnosis of possible CIDP can be made. Electrodiagnostic criteria for the presence of demyelination as proposed by Saperstein and colleagues are detailed in Table 99–3. They differ from those stipulated in the AAN

Table 99–3. Comparison of Electrodiagnostic Criteria for CIDP AAN Ad Hoc Subcommittee6 Definition of conduction block (CB) (% amplitude drop between proximal and distal stimulation sites) Definition of temporal dispersion (TD) (% increase in duration after proximal stimulation) Criteria for demyelination (in motor nerves)

Saperstein et al.178

Nicolas et al.142

⬎20%

Variable, depending on nerve; ⬎ 50% in most cases

⬎30%

⬎15%

⬎30%

⬎15%

3 of 4 abnormal:

2 of 4 abnormal:

One of the following:

CB/TD in ⱖ 1 nerve

CB/TD in ⱖ 1 nerve

CV* in ⱖ 2 nerves

CV* in ⱖ 2 nerves

DL† in ⱖ 2 nerves

DL† in ⱖ 2 nerves

FW‡ in ⱖ 2 nerves

FW‡ in ⱖ 2 nerves

CB/TD in ⱖ 3 nerves with abnormal CV*, DL†, or FW‡ in ⱖ 1 nerve CB/TD in 2 nerves with abnormal CV*, DL†, or FW‡ in ⱖ 1 nerve CB/TD in 1 nerve with abnormal CV*, DL†, or FW‡ in ⱖ 2 nerves No CB/TD but abnormal CV*, DL†, or FW‡ in ⱖ 3 different nerves

AAN ⫽ American Academy of Neurology; CIDP ⫽ chronic inflammatory demyelinating polyneuropathy; CMAP ⫽ compound motor action potential; CV ⫽ conduction velocity; DL ⫽ distal latency; FW ⫽ F wave; LLN ⫽ lower limit of normal; ULN ⫽ upper limit of normal. *Abnormal if CV less than 80% of the LLN and CMAP amplitude greater than 80% of the LLN, or CV less than 70% of the LLN and CMAP amplitude greater than 80% of the LLN. † Abnormal if DL greater than 125% of the ULN and CMAP amplitude greater than 80% of the LLN, or DL greater than 150% of the ULN and CMAP amplitude less than 80% of the LLN. ‡ Abnormal if minimum F-wave latency greater than 125% of the ULN and CMAP amplitude greater than 80% of the LLN, or minimum F-wave latency greater than 150% and CMAP amplitude less than 80% of the LLN, or if F-wave is absent. From Saperstein, D. S., and Barohn, R. J.: Current concepts and controversy in chronic inflammatory demyelinating polyneuropathy. Curr. Neurol. Neurosci. Rep. 3:57, 2003, with permission.

2228

Diseases of the Peripheral Nervous System

criteria only in the definition of “partial conduction block,” for which the more rigorous American Association of Electrodiagnostic Medicine (AAEM) consensus criteria149 are employed. A nerve biopsy is no longer considered mandatory for the diagnosis of definite CIDP; it is reserved for atypical CIDP cases in which there are diagnostic uncertainties.17 In general, the revised criteria accommodate better the varied clinical manifestations of CIDP and will assist clinicians in making a timely diagnosis. Recently, a further set of diagnostic criteria for CIDP have been published by a group of European researchers—the Inflammatory Neuropathy Cause and Treatment (INCAT) group.83 These are on the whole less restrictive and differ from previous definitions primarily in the neurophysiologic criteria for CIDP. Stipulated electrodiagnostic criteria for the definition of a “demyelinating” neuropathy are upheld. However, strong emphasis is placed on the presence or absence of partial CB (defined as ⬎ 30% drop in amplitude between proximal and distal stimulations) and of temporal dispersion (TD) of compound muscle action potentials (CMAPs). Nicolas and associates142 coined the index “conduction block/temporal dispersion” (CB/TD), which is used to represent presence or absence of such changes. The proposed electrodiagnostic criteria for CIDP were thus based on the number of nerves with CB/TD (for details see Table 99–3). To test the sensitivity and specificity of these criteria in their cohort of 40 patients with definite CIDP, Nicolas and associates compared their findings in this cohort to those in a group of 35 patients with a chronic axonal motor and sensory neuropathy, and also to findings in patients with various other chronic demyelinating neuropathies.142 In applying the new set of criteria to this cohort of CIDP patients, they found a sensitivity of 90% and a specificity of 97%. When applying the more stringent criteria for CB, stipulated in the AAEM consensus criteria,149 the sensitivity was reduced to 83% while the specificity remained at 97%. In contrast, only 63% of this CIDP cohort fulfilled the electrodiagnostic criteria of CIDP stipulated by the AAN Ad Hoc Subcommittee.6 Therefore, the new INCAT diagnostic criteria appear to be more sensitive while being sufficiently discriminating, although their validity has yet to be confirmed independently.

Differential Diagnostic Considerations In clinical practice, CIDP needs to be separated with appropriate laboratory testing from a number of chronic polyradiculoneuropathies with similar clinical presentations (Table 99–4). CIDP is distinguished from AIDP mainly by the much longer and continued disease progression of greater than 8 weeks, and also by the more constant complaints and findings of altered sensation. The distinction between the two conditions is important because they differ not only in their course, but also in their response to treatments

Table 99–4. Concurrent Illness Variants of Chronic Inflammatory Demyelinating Polyneuropathy HIV infection Lymphoma Osteosclerotic myeloma, POEMS, Crow-Fukase syndrome, Castleman’s disease Monoclonal gammopathy of undetermined significance Chronic active hepatitis, hepatitis C Inflammatory bowel disease Connective tissue disease Bone marrow and organ transplants Central nervous system demyelination Nephrotic syndrome Diabetes mellitus Hereditary neuropathy Thyrotoxicosis HIV ⫽ human immunodeficiency virus; POEMS ⫽ polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes.

(in particular to corticosteroids) and in overall prognosis (see later in the text). Despite certain similarities in clinical presentation, CIDP should also be separated from the acquired demyelinating polyneuropathies, the so-called CIDP-MGUS, which are associated with a monoclonal gammopathy involving immunoglobulin A (IgA), immunoglobulin G (IgG), or immunoglobulin M (IgM) (in particular those with positive anti–myelin-associated glycoprotein [anti-MAG] activity). Patients afflicted with these conditions tend to be older and more likely male, presenting with predominant and slowly progressive sensory deficits and imbalance. Any motor deficits usually involve distal limb muscles. In general, CIDP-MGUS follows a more indolent course but responds less well to treatment.29,95,138,185,186 It is also important to distinguish CIDP from the clinically similar neuropathies that may be encountered with lymphoproliferative or plasma cell disorders, such as solitary osteosclerotic myeloma or multiple myeloma, lymphoma, Waldenström’s macroglobulinemia, Castleman’s disease, cryoglobulinemia, or acquired amyloidosis. Therefore, the initial work-up of a CIDP patient must include a thorough immunologic and hematologic evaluation with the quantitative assessment of immunoglobulins in serum, a protein electrophoresis (preferably an immunofixation electrophoresis) of serum and urine, a radiologic skeletal survey including long bones, and a bone marrow aspirate and biopsy. CIDP is rarely seen in association with melanoma, whereby the serum may contain polyclonal antibody against the monosialoganglioside GM2 that cross-reacts with the melanoma tissues, suggesting molecular mimicry.8,16,231 This raises interesting considerations in regard to the pathogenesis of CIDP that are discussed later. In principle, it is advisable to separate idiopathic CIDP from the polyneuropathy seen with diabetes mellitus, so

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

long as the understanding of the pathogenesis of either condition remains incomplete. However, a concurrent diagnosis of CIDP should be considered in a diabetic patient if he or she reports a recent and subacute development of weakness that is greater proximally than distally and is seen together with sensory loss and ataxia. In this case, the electrophysiologic studies will show a typical slowing of motor conduction velocities, partial CB, and dispersion of the CMAPs, all indicating a demyelinating neuropathy. Finding of unequivocal CB in one or more nerves would lend strong support for a diagnosis of concurrent CIDP.1,205 Such patients tend to respond favorably to the various immune-modulating treatments.67,75,101,184,195 Idiopathic CIDP should also be set apart from the demyelinating polyradiculoneuropathy that may occur in HIV-seropositive individuals during the early stages of the disease, usually around the time of seroconversion. Characteristically, such patients have a greater degree of lymphocytic pleocytosis in CSF.12,222,236 The incidence of HIV-associated CIDP is unknown. Moreover, we currently have little understanding of how HIV and other infectious agents, such as the hepatitis viruses, cytomegalovirus, and Epstein-Barr virus, may trigger a CIDP-like polyradiculoneuropathy. HIV-associated CIDP responds to the usual therapies. Atypical presentations of CIDP have been recognized wherein the neurologic deficits are focal or multifocal and are often asymmetrical.68,111,145,199,211,223 In such variant cases, an acquired multifocal motor and sensory demyelinating neuropathy is supported by the finding on electrophysiologic examination of generally slow motor and sensory conduction velocities and of persistent partial CB in single nerves, indicating focally accentuated demyelination. Such patients respond favorably to the usual immune-modulating therapies. However, one needs to keep in mind that a similar clinical presentation may be seen with ischemic mononeuropathies, caused by a systemic or nonsystemic necrotizing vasculitis, or by a leukoclastic vasculitis that is seen with chronic hepatitis C and cryoglobulinemia. The differential diagnosis should also include the granulomatous or infiltrative pathologies. Therefore, apart from detailed electrophysiologic examinations, the work-up should include the appropriate laboratory blood tests, magnetic resonance imaging (MRI) of proximal nerve segments, and often pathology studies of a nerve biopsy.41,90,105,130,210 Rare forms of CIDP with purely sensory manifestations have been designated as “sensory CIDP.”146 These may be difficult to diagnose because demyelination may occur mainly at the level of sensory roots. Thus conduction velocities in sensory nerves and amplitudes of sensory nerve action potentials (SNAPs) may be normal while somatosensory evoked potentials are delayed or absent. CSF analysis will show a normal cell count but typically elevated protein content. Here, the differential diagnosis will include paraneoplastic (anti-Hu antibody positive) or

2229

nonmalignant sensory neuronopathy. The latter may be seen with Sjögren’s syndrome or it may be idiopathic. In both forms the sensory deficits are often patchy and can involve the face, trunk, and limbs, and neuropathic pain is common. In general, signs and symptoms will progress more rapidly and SNAPs are reduced or absent. CIDP should be separated from the multifocal motor neuropathy with conduction block (MMN-CB).94 In the latter condition, motor fibers are selectively affected by focal demyelination and upper limbs are preferentially affected; the two conditions differ in their natural history and response to treatment, in particular to corticosteroids.

Disease Course and Prognosis As mentioned, the course of CIDP may be either relapsingremitting or steadily or stepwise progressive. Often the pattern of disease cannot be predicted at the time of diagnosis and will only become evident with follow-up. However, certain directives can be taken from published series. For instance, in their review of 124 CIDP patients, Hattori and colleagues80 observed the relapsing-remitting form to be much more common among juvenile patients (age 20 years and younger), who often presented initially with subacute progression of motor weakness. Relapses were less frequent in the adult patients and they were distinctly uncommon in the elderly. Moreover, older patients tended to present with a chronic, insidiously progressive motor and sensory neuropathy. Despite the variable clinical presentation, electrophysiologic and pathologic studies of all patients showed the typical signs of demyelination in keeping with CIDP. As the disease progressed, there were often also signs of axonal degeneration, which varied in severity. The review showed that, during periods of maximal impairment, younger patients were often much more severely affected. Generally, they responded more favorably to the immune-modulating therapies and their overall prognosis was significantly better. Similar observations have been made in other large series.19,120,183,186,187 The long-term prognosis for older patients was less good, although they responded clearly to the usual treatments.80 Bouchard and colleagues19 gained a similar impression from their review of 100 French CIDP patients. They found that in older patients the less favorable outcome was correlated with a loss of axons in nerve biopsies. The pathologic changes in the biopsies were principally those of inflammatory demyelination. Yet accompanying loss of axons was so common that in 47% of patients the density of myelinated fibers had become reduced to less than half of normal. The authors concluded from these findings that, in long-standing CIDP, the cumulative loss of axons determined a poorer prognosis. Nonetheless, with the newer therapies the outlook for CIDP has become much more favorable. For instance, in a recent review of 90 CIDP patients cared for at our

2230

Diseases of the Peripheral Nervous System

institution, 79 (88%) had returned to normal or had only minor, nondisabling symptoms remaining (modified Rankin score of 0 or 1), and 55 patients (64%) had discontinued all medications and were in full remission at a mean follow-up of 5.7 years.221 However, early on, many of these patients had required aggressive therapy, often with a combination of several treatment modalities. Irrespective of the first treatment we used (prednisone, plasma exchange [PE], or intravenous immune globulin [IVIg]), the final outcome did not differ significantly. In recent years, we have preferred IVIg as the first-line treatment because improvements are usually noted within days (approximately two thirds of CIDP patients respond to IVIg) and the drug is generally well tolerated. Many patients, particularly those with a relapsing disease course, will require regular IVIg pulse treatments every 3 to 4 weeks. McCombe and co-workers120 also observed an overall good outcome in a series of 92 Australian CIDP patients. Among these, two thirds were young (mean age 26.8 years) and had shown a relapsing course, which predicted a better prognosis. This was indeed the case. At 10-year follow-up, 73% of 81 patients either were normal or had only minor functional deficits, 17% were moderately affected, 3% were unable to walk, and 6% (5 patients) had died as a result of their disease. Barohn and colleagues13 reviewed a series of 60 patients from a single center in the United States. The majority had experienced neuropathic symptoms for less than 1 year before being correctly diagnosed with CIDP. Approximately half (28 patients) had shown a relapsing course, whereas the remaining 32 patients had a monophasic illness. During a 10-year follow-up period, patients with a relapsing course experienced on average 0.7 ⫾ 0.4 relapses per year and only 5 of the 28 patients had achieved a full remission, while the others had required repeated treatments with each relapse. Generally, the patients had responded well to immunosuppressive therapies; 59 of 60 CIDP patients (~95%) had shown a favorable response to prednisone. A consistent regimen for the prescription of corticosteroids had been used, starting with prednisone 100 mg daily for 2 to 4 weeks, followed by a change to prednisone 100 mg on alternate days and gradual tapering off over many months. Signs of improvement were seen, on average, with a delay of 1.9 ⫾ 3.6 months, and patients had reached a stable clinical plateau at a mean 6.6 ⫾ 5.4 months, having improved substantially. Of course, the prolonged prescription of highdose prednisone carried the risk of serious side effects. With an average follow-up of 34 ⫾ 32.3 months, 19 of 32 CIDP patients (60%) with a monophasic disease course had achieved remission and no longer required treatment. Yet, in total only 24 of 60 CIDP patients (40%) were in complete remission, with the majority of patients with relapsing disease requiring ongoing treatments. The introduction of the additional treatment options of plasmapheresis and high-dose IVIg has improved the overall

prognosis of CIDP. Van der Meche and van Doorn212 reviewed their experience with short-term and long-term use of IVIg in 53 CIDP patients. IVIg, given as single therapy (usual dose of 2 g/kg), produced remarkable functional improvements in 41 (77%) of 53 CIDP patients. While being followed for a median duration of 4.7 years (range 0.5 to 14 years), 30 of the 41 patients (73%) continued to do well but they required periodic IVIg treatments. No serious side effects were observed. In view of the high cost of IVIg and the inconvenience of repeat infusions, it was recommended to add other immunosuppressant therapies if IVIg pulse treatments were required for 6 months or longer. One of us (P.J.D.) followed 63 patients with CIDP for periods of time up to 18 years, treating them with PE, IVIg, or oral immunosuppressants based on changes in periodic neuropathic examinations. All were better at last than at first examination, but only a small percentage had essentially returned to normal (Dyck and co-workers, unpublished results). Almost all patients still needed continuous treatment. Thus the outlook for idiopathic CIDP is considerably improved with immune-modulating treatments. Improvement and stabilization can usually be achieved for 80% of CIDP patients when the treatments are introduced early in the disease, are given consistently, and cover the complete range of available therapies, including physical and occupational therapies.

CIDP in Childhood Typical idiopathic CIDP may be observed in children of any age, but it is much less common than in adults. From the available epidemiologic data, Connolly30 estimated the prevalence to be roughly 0.48 per 100,000 population. Given the relatively rare occurrence, most published case series pooled the experience of several centers. Even then, the numbers are small.79,100,140,170,187–189 Altogether, these reports define, in general terms, the characteristic features of childhood CIDP. Clearly, a preceding infection (e.g., an upper respiratory infection, gastroenteritis, or simply a nonspecific febrile illness or vaccination) was more readily identified in children (in up to 50% of cases) than in adult patients. This may be due to the fact that children tended to present earlier in the course because their initial disease progression was generally more precipitous. For instance, a third of cases had reached their nadir within maximally 4 weeks from onset of the neuropathy, so that the clinical picture mimicked AIDP and was often diagnosed as such. Only subsequent relapses led to an accurate classification. A relapsing-remitting course was observed in nearly 80% of cases. Some children have a slower progression to maximal weakness over the course of 3 months or longer and tend to have a monophasic course. In general, children respond very well to the usual immune-modulating treatments and often achieve long-term remission.

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

As mentioned, children present usually with a symmetrical and mainly motor polyneuropathy that affects the lower limbs predominantly and results in gait disturbance as the salient clinical manifestation. Not infrequently, the child with CIDP temporarily loses the ability to walk, and the infant may present with the picture of developmental delay. Less commonly, weakness is predominant in the arms, leading to tremors and loss of dexterity. A few children (15%) develop diplopia, facial weakness, or difficulty chewing or swallowing and may eventually require intubation and assisted ventilation.32,40 One third of children may exhibit sensory symptoms, either paresthesias or pain, at the time of presentation. Laboratory tests for the diagnosis of childhood CIDP are the same as for adults. Electrophysiologic studies should include the examination of at least four motor nerves, and the results must be compared to the well-established normal childhood values. With the GBS-like presentations, the early finding of markedly slow motor nerve conduction velocities would point to the probable diagnosis of CIDP. At this stage, findings of decreased or absent median nerve SNAPs, yet still normal recordings from sural nerves, are not unusual.141 The response to immune-modulating therapies and overall prognosis for children with CIDP is excellent. The collective data are summarized in the recent review by Connolly.30 Prednisone in a dosage of 1 to 1.5 mg/kg/day resulted in improvement in the majority of children (93%), but IVIg or therapeutic PE is also clearly effective (70% and 90%, respectively). However, for this age group, no controlled comparative studies are available because the number of treated cases is small. Nevo and colleagues140 found that children with a monophasic course who progressed to their nadir over 1 to 3 months were likely to recover completely when treated with prednisone. Hattori and associates79 made similar observations and suggested that corticosteroids would be the drug of first choice for such patients. Simmons and co-workers188 compared their experience with long-term follow-up in 12 children with CIDP (mean disease duration 7.1 years) to that in their 60 adult CIDP patients. Ten children had a relapsingremitting course with an average of 5.4 relapses each (range 2 to 15). Episodes of deterioration developed more rapidly in children than in adults, but when treated with either IVIg or prednisone, children improved more rapidly and completely and were left with very little functional deficit. IVIg was usually very effective, and improvement commonly occurred in less than 1 week. This view is shared by Ryan and colleagues,170 who had followed 16 children on average for 10 years. The outcome was favorable in all. When last seen, 14 children were asymptomatic and 2 had mild residual deficits. The very-long-term outlook is not known because no children have been followed into adulthood. However, it would seem that the majority remain in remission.

2231

Clinical Spectrum Purely Motor Presentation Donaghy and colleagues39 drew attention to a rare form of CIDP with exclusively motor signs and symptoms, caused by a multifocal demyelinating motor neuropathy. They reported on four patients, three of whom had presented with an asymmetrical pattern of weakness that had affected predominantly the arms or had commenced in the arms before spreading to predominate in the legs. None had sensory symptoms or abnormal electrophysiologic recordings from sensory nerves. Kaji and colleagues91 had described two additional cases in whom they had observed bulbar involvement and dysarthria with atrophy and fasciculation of the tongue, so that the clinical picture resembled a motor neuron disease. However, on careful electrophysiologic study they found evidence of multifocal CB in recordings from motor nerves in the limbs. Recently, Sabatelli and co-workers171 reported a further four patients with such a purely motor presentation. None had sensory symptoms or signs on clinical or electrophysiologic examination. All four patients were young and had shown a relapsing course. Although in their clinical presentation these patients eventually have much in common with CIDP patients, in the early phases of their disease they can resemble patients with MMN-CB. Therefore, the nosologic position of this generalized and multifocal demyelinating motor neuropathy is still in debate. Despite the unresolved issues, it is important to remember that such patients, although responding favorably to IVIg or PE therapy, may worsen acutely in their motor functions when treated with corticosteroids.24 Purely Sensory Manifestation In approximately 5% to 8% of cases CIDP may present as a purely or predominantly sensory neuropathy with a subacute or chronic progression.46,146,148,169,241 Oh and colleagues146 described the typical clinical features of this variant from 10 such cases, all male patients whom they had studied over a period of 18 years. All had a slowly progressive and monophasic course and had presented originally with numbness in their feet and hands. Thereafter, the numbness had ascended in a typical stocking-glove distribution and had occasionally been accompanied by neuropathic pain that necessitated the prescription of strong analgesics. The sensory ataxia had resulted in unsteadiness of gait, as well as in action tremors and impaired dexterity. Although all their patients had retained normal muscle strength, the electrophysiologic studies revealed signs of demyelination not only in sensory but also in motor nerves, indicating the presence of a generalized demyelinating peripheral neuropathy. The widespread nature of the demyelinating pathology was shown also in the nerve biopsies, which supported the designation of “sensory” CIDP. Indeed,

2232

Diseases of the Peripheral Nervous System

the majority of these patients had only sensory dysfunctions remaining. However, some who initially had presented with purely sensory manifestations were shown later to have developed motor weakness.15,213 In general, patients responded well to a variety of immune-modulating therapies, including corticosteroids, therapeutic PE, and IVIg infusions, which also suggests an immune etiology.15,146 “Sensory” CIDP is now recognized as a variant form of the disease. In making this diagnosis, one needs to keep in mind that a similar sensory ataxic presentation can be observed with the acquired demyelinating neuropathy seen with IgM ␬ or ␭ monoclonal gammopathy, with or without anti-MAG activity.95 Therefore, this possibility needs to be excluded. Multifocal Motor and Sensory Demyelinating Neuropathy (Lewis-Sumner Syndrome) Lewis and colleagues111 observed in five of their patients an unusual pattern of a chronic demyelinating sensorimotor polyneuropathy that affected predominantly the upper limbs in an asymmetrical distribution. In meticulous electrophysiologic examinations they found evidence of focal motor and sensory CB in individual nerves that was persistent for years and accounted for the regional motor and sensory deficits. Early on in their disease, these patients had experienced numbness and painful paresthesias in the territories of single nerves. Later, this was followed by the development of insidiously progressive muscle weakness and atrophy. In every case, the neurologic deficits were persistent and eventually other limbs became similarly affected. Other causes of mononeuritis multiplex, in particular the different forms of vasculitis or granulomatous infiltrations, were ruled out with laboratory blood investigations and pathology study of sural nerve biopsies. Instead, the findings supported the diagnosis of an acquired, multifocal and generalized demyelinating polyradiculoneuropathy that was akin to CIDP. The analogy is strengthened by the elevated CSF protein and the favorable response to corticosteroids. Since this early report, many similar observations have been described in the literature with different designations: multifocal motor and sensory demyelinating mononeuropathy (MSDMM),145 upper limb–predominant multifocal chronic inflammatory neuropathy,68 multifocal acquired demyelinating sensory and motor neuropathy (MADSAM),176 and multifocal inflammatory demyelinating neuropathy.211 The following clinical picture has emerged. As in CIDP, males were more commonly affected and symptoms began usually between ages 40 and 50 years. In the majority of cases the disease was slowly progressive over many years; less often it appeared as a relapsing-remitting disorder. Initially, sensory symptoms prevailed, with tingling, numbness, and not uncommonly neuropathic pain localized to single nerves (i.e., the median, radial, and ulnar nerves, and/or the sural nerves). Subsequent motor deficits devel-

oped predominantly in the upper limbs (78%), often in an asymmetrical distribution. For years, the clinical deficits could remain localized to the upper limbs, yet laboratory studies were usually indicative of a more widespread subclinical nerve involvement. Deep tendon reflexes were often abolished, and patients reported a tendency to muscle cramping. The disease began less commonly in the lower limbs (22%); cranial nerves—optic, oculomotor, trigeminal, or facial nerves—were rarely affected (18%). After many years, some patients developed overt signs of a more generalized neuropathy, although by careful clinical examination one could still discern its multifocal nature. For example, a painful Tinel’s sign could be elicited at the sites of persistent CB, underscoring the focal accentuation of nerve pathology.145,176,211,223 In the course of their illness, some patients had developed localized thickening of the nerves, which were easily palpable or even visible in the supraclavicular region.4,20,199,211 These striking tumor-like appearances could be defined by nuclear magnetic resonance imaging (NMRI) of the brachial plexus, where they appeared as focal increased signal intensity on the T2-weighted sequences. No enhancement was seen after the administration of gadolinium-DTPA.211 On pathologic examination the regional nerve thickening corresponded to foci of chronic hypertrophic inflammatory demyelination in nerve trunks of the brachial plexus.20,33,128,199,210 Biopsies of more distal sensory nerves (radial, ulnar, and sural) showed signs of chronic low-grade demyelination and axonal loss, in keeping with the presence of a more generalized neuropathy seen also on electrophysiologic examination.20,111,118,145,199 Nerve conduction abnormalities were characterized by the finding of multiple mononeuropathies, and evidence of partial motor and sometimes sensory CBs, that were often localized to the forearm and could persist for years.111,211 In a series of 16 patients, Oh and co-workers145 observed diminished motor and sensory conduction velocities in 81% of cases, with abnormal TD of motor and sensory action potentials in 40%, prolongation of distal latencies in 19%, and prolongation of F-wave latencies in 25% of cases. Needle electromyography of affected muscles showed evidence of both active and chronic denervation. Combined, these findings indicated the presence of a generalized chronic demyelinating neuropathy with focal accentuations, the latter accounting mainly for the asymmetrical clinical presentation. Saperstein and colleagues176 argued that the universal reduction of recorded SNAP amplitudes distinguished this variant from MMN-CB. In nearly 60% to 80% of patients, CSF protein was mildly raised and serum antibodies to the GM1 ganglioside were not found, which again was clearly different from MMN-CB.145,176,211 In this variant form of CIDP, corticosteroids were very effective. Approximately two thirds of patients treated with prednisone in monotherapy showed a dramatic improvement and stabilization of their condition. The others did not respond. Indeed, two patients deteriorated with this

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

therapy—a negative response as seen more commonly in MMN-CB. Nowadays, the treatment of first choice is usually IVIg, because it has been found to be beneficial for greater than 70% of patients. As a matter of fact, a certain number of patients will require long-term IVIg pulse therapy. PE has not been used much in this variant of CIDP. Limited information indicates that it is not particularly effective.145,176,211 Discussion about the nosologic position of multifocal motor and sensory neuropathy remains lively. In recent editorials, the authors of the original descriptions have presented arguments for and against a clear distinction of the two conditions.110,151 Until the causal mechanisms are understood more precisely, the significance of the overlapping clinical and pathologic features will remain uncertain. Therefore, for the time being, it would seem more appropriate to keep the two conditions separate. CIDP and Multifocal Central Nervous System Involvement A few reports have drawn attention to the presence of a chronic demyelinating peripheral neuropathy in association with otherwise typical multiple sclerosis (MS).107,127,166,167,181,201 Thomas and colleagues201 reviewed in detail the findings of six patients and discussed the possibility of shared autoimmune responses by making reference to experimental models of central nervous system (CNS) and peripheral nervous system (PNS) inflammatory demyelination—experimental autoimmune encephalomyelitis and EAN. In the same year, Mendell and co-workers127 had examined 16 selected CIDP patients with the then relatively new imaging tool, NMRI. In six patients they disclosed typical multifocal and periventricular white matter lesions of MS. In three patients these lesions, identified by radiology, were clinically silent, although all six patients had overt symptoms and signs of CIDP. In fact, the clinical picture in all patients was dominated by signs of peripheral nerve dysfunction. The study shed no light on the possibility of a common etiology. Moreover, the true concurrence of the two conditions could not be judged from either of these two reports because both studies were motivated by a biased selection of cases with CNS involvement. In an unselected series of 19 well-documented cases with long-standing CIDP, Feasby and colleagues54 found NMRI lesions typical of MS to be very uncommon. In seven patients, none with clinical signs of CNS involvement or other supporting evidence of MS, one or more subcortical MRI white matter lesions were seen but none was typical of MS. Six patients were older than age 55 years, and the MRI abnormalities were considered to be due to ischemia. In only one 38-year-old woman could the two small white matter lesions pass for early MS, yet she had no other supportive evidence for this diagnosis. The study concluded that concurrence of CIDP and MS was distinctly uncommon.

2233

CIDP and Demyelinating Hereditary Motor and Sensory Polyneuropathy Occasionally, patients afflicted with a demyelinating form of CMT disease may suffer an acute or subacute worsening in motor dysfunction. Such patients will experience new-onset weakness in proximal leg muscles, superimposed on the chronic weakness in peroneal muscles resulting from CMT.58 Electrophysiologic examination may show new evidence of partial CB in motor nerves, a finding usually not seen in uncomplicated CMT. Pathology examination of sural nerves may reveal extensive subperineurial edema in addition to perivenular and endoneurial mononuclear cell infiltrates, which is in keeping with a superimposed inflammatory reaction. Typically, such a patient will improve with corticosteroids, whereas this is not the case in ordinary CMT.52,58,157,230 The pathogenesis of this superimposed inflammatory demyelinating neuropathy is not known, yet the responsiveness to corticosteroids suggests a possible immune reaction. In view of the documented therapy responses, one needs to be aware of these potentially concurrent pathologies.

Laboratory Investigation Routine hematology and biochemistry testing usually gives normal values. In CIDP cases with either focal or asymmetrical signs, one needs to consider and exclude (where appropriate) possible infections (HIV, hepatitis C virus, Lyme disease), diabetes mellitus, and all forms of vasculitis and sarcoidosis. The work-up should include a quantitative assessment of serum IgG, IgA, and IgM, and screening for monoclonal gammopathies in serum and in urine with high-resolution agarose gel immunoelectrophoresis or with immunofixation. If ␬ or ␭ light chains are detected, follow-up with a hematology consultation and bone marrow aspirate and biopsy is necessary to rule out a lymphoproliferative disorder or malignant plasma cell dyscrasia. In such circumstances a radiologic skeletal survey is also warranted to search for either solitary or multiple osteosclerotic or osteolytic myeloma. Cases of documented IgM monoclonal gammopathy require testing for anti-MAG antibody activity. In certain cases the clinician may order genomic DNA testing to rule out the common hereditary demyelinating neuropathies. In rare cases the differential diagnosis should include genetically determined metabolic disorders such as metachromatic leukodystrophy or mitochondrial neurogastrointestinal encephalomyopathy14 or other mitochondrial disorders. The diagnosis of CIDP is further confirmed with analysis of CSF obtained by lumbar puncture. Cell count is usually less than 10 white blood cells per cubic millimeter; a higher cell count should raise concerns about HIV infection. CSF protein is usually elevated (95% of patients in the series of Barohn and colleagues12 had an elevated CSF protein

2234

Diseases of the Peripheral Nervous System

[mean ⫾ SD, 1.34 ⫾ 1.12 g/L]); occasionally, early on it will still be normal. Oligoclonal banding was documented in 65% of cases.38 Nerve biopsy is performed only in exceptional cases, such as in those patients with an asymmetrical and multifocal Lewis-Sumner syndrome–like presentation and neuropathic pain, in whom there is concern about other pathologies such as a vasculitis, perineuritis, or granulomatous pathology.17,132 The findings in nerve biopsies are discussed in detail later in the section on pathology. Electrophysiologic Examination Nerve conduction studies are performed in at least two motor and two sensory nerves in each of one upper and one lower limb (preferably all four limbs) and should include proximal nerve segments. Limb temperature should be monitored carefully. If necessary, the limbs are warmed to 36° C. Electrodiagnostic findings that suggest a demyelinating neuropathy include the following: slowed motor conduction velocities, prolonged distal motor latencies, delayed or absent F waves, and presence of CB and/or TD. In typical CIDP these changes will all be present, while needle electromyography shows little or no evidence of acute or chronic denervation. Often, SNAPs cannot be recorded from ulnar, median, and superficial peroneal nerves, whereas early on the sural nerve SNAP may still be relatively normal. Abnormalities in distal motor latency, motor conduction velocity, and F-wave latency will also be seen in a primarily axonal neuropathy. Therefore, an accurate interpretation requires a balanced analysis of the entire set of recordings. Several sets of electrodiagnostic parameters have been formulated to delineate the nerve conduction abnormalities that indicate a primarily demyelinating neuropathy.5,6,142 These have been incorporated in the proposed electrodiagnostic criteria for CIDP listed in Table 99–3. Discussions continue about their sensitivity and specificity for accurately detecting CIDP in a clinical setting.23,133,168,177,209 In CIDP motor conduction velocities are generally below 80% of low-normal values while the CMAP amplitudes remain at greater than 80% of the lower limit of normal. Motor conduction velocities may be less abnormal early in the disease course, or when the lesions are located mainly in spinal roots. In this circumstance it is necessary to test specifically for the proximal motor conductions either by the determination of F-wave response latencies, or by directly stimulating the nerve roots. The amplitudes of the CMAPs are reduced (1) because of CB and TD of the CMAP on account of multifocal demyelination along the length of the nerves, including the distal nerve segments; and (2) because of loss of motor axons. Proposed criteria for the definition of partial CB (% CMAP amplitude drop between proximal and distal stimulation) have varied between 20% and 50%.6,178 Recently, a consensus committee of the AAEM published

guidelines for identifying partial CB in various motor nerves. AAEM criteria define CB as greater than 50% reduction in CMAP peak-to-peak amplitude or negative peak area, with less than 15% change in duration.149 Determinations of TD rely on comparing the duration of the proximally derived CMAP with that of the distal CMAP. Measurement of the distal CMAP dispersion (an index of very distally occurring demyelination) has also shown promise as a sensitive and specific adjunctive electrodiagnostic criterion for CIDP.198 Electrophysiologic criteria for the diagnosis of definite and probable CIDP are listed in Table 99–3 and are relevant even in the setting of clinical practice. Occasionally, a clinical diagnosis of possible CIDP will be made when not all criteria are met. In such circumstances, reliable predictions of treatment responses and prognosis are not possible. Additional support for the diagnosis of CIDP may be derived from a trial of immune-modulating therapy.

TREATMENT Current treatments for CIDP assume an immune pathogenesis that, although not yet proven, is well supported by medical and experimental evidence (discussed later in this chapter). This concept had already led to trials of therapy with corticosteroids in the early 1950s. The first anecdotal observations of the use of corticosteroids in CIDP were reviewed in detail in previous editions of this text and are referred to only briefly here. Efficacy of several immunemodulating therapies in CIDP is now firmly established.

History of Immunotherapy and General Remarks Corticosteroids In 1958 Austin10 first documented in a patient with relapsing CIDP the marked and reproducible improvements in neurologic impairment with each and over many treatment courses of either ACTH or prednisone. He also pointed out that, in spite of repeated beneficial effects with steroids, a cure was not obtained. Subsequent anecdotal reports echoed this view, but noted also that not all CIDP patients responded equally well to the therapy.144,200 From their study of 25 patients with a “dysimmune” neuropathy, Dalakas and Engel37 concluded that oral prednisone was effective and the treatment of first choice for the majority of such patients. A long-term prescription of high-dose oral prednisone was required to achieve recovery and stabilization. First signs of improvement were usually seen by 4 weeks, but occasionally it took up to 5 months to note a response. They recommended high doses of prednisone (1 to 1.5 mg/kg/day) for 3 to 4 weeks, followed by a gradual dosage change to alternate-day therapy and a further gradual reduction of the steroid dose over many months. The authors observed that sudden or rapid dose reduction could

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

lead to a prompt relapse or else to an incomplete therapy response and fixed neurologic deficits. Therefore, they suggested that relatively high steroid doses be prescribed until remission is achieved and that low-dose prednisone be maintained for several years to prevent recurrences. In some patients they added a single daily dose of azathioprine, 3 mg/kg, to reduce the intolerable side effects of prednisone. Benefit from corticosteroids for both adults and children with CIDP was documented subsequently in retrospective review of several large patient series13,46,100,120,131,188,189 and in one randomized, controlled trial (RCT). These are discussed in greater detail later. Although the prescription of prednisone is convenient and relatively cheap,122 prolonged use may result in hazardous side effects. Potential complications range from an altered appearance, weight gain, insomnia, mood change, and aggravation of hypertension or of carbohydrate intolerance to psychosis, peptic ulcers, cataracts, osteoporosis with subsequent vertebral compression fractures, and avascular necrosis of the hip. Some of these side effects may be diminished by appropriate countermeasures, such as the prescription of antacids (histamine2 receptor antagonists), a sodium-restricted low-carbohydrate and high-protein diet, and osteoporosis prophylaxis with supplemental calcium and bisphosphonates. The potentially serious and lasting complications make steroid therapy less desirable. Dose and duration of prednisone therapy can now be reduced with the added prescription of other immune-modulating treatments. Plasma Exchange Therapy In the search for alternate treatments, in the late 1970s therapeutic PE was first used in CIDP. It showed substantial short-term benefit for select patients, particularly those with relapsing CIDP.71,109,182 At times, improvement was observed within only a few days from the beginning of a series of PEs, although it was often not sustained. When PE was stopped, neurologic function worsened, at times rapidly, but improved again when PE was resumed. Thus PE emerged as useful adjuvant therapy and was particularly efficacious early in the disease course, when demyelination was the predominant pathology.155 Sustained improvement was obtained only by adding prednisone or other chronic immunosuppressive therapies.73,152 Anecdotally, response to PE was tested in several small groups of patients. However, these observations, though carefully documented, were derived from nonblinded assessments and therefore subject to bias.51,71,155 Moreover, the patients were often treated simultaneously with immunosuppressive drugs, which undoubtedly could have contributed to the clinical response. A more rigorous approach was taken by Dyck and colleagues at the Mayo Clinic in a randomized, double-blind, and sham apheresis–controlled study of 29 long-term CIDP patients,42 and in a similar multicenter Canadian trial of 18 newly diagnosed, untreated CIDP

2235

patients.73 These critical evaluations, which documented unequivocal short-term efficacy of PE, are reviewed later. The unique effectiveness of PE underscored the pathophysiologic importance of humoral factors (autoantibodies and/or T cell–derived cytokines) for the pathogenesis of CIDP, which is discussed later in this chapter. Intravenous Immune Globulin More recently, high-dose IVIg was introduced as immunemodulatory therapy for a number of autoimmune disorders, including CIDP. Vermeulen and colleagues224 summarized the first observations of its use in CIDP. They had observed substantial benefit in 13 of 17 patients (70%) with open-label treatment of IVIg (2 g/kg given over 2 to 5 days). Usually neurologic function began to improve within 1 week following the infusions and continued for several weeks. However, the benefit was not sustained in the majority of responders, so that they worsened again after a variable length of time. They improved again and to the same degree when the treatments were resumed. Van Doorn and colleagues215 critically assessed and confirmed the response to IVIg in a placebo-controlled, double-blind crossover study in seven CIDP patients who had previously responded favorably. All seven patients improved significantly in neurologic impairment and function, as well as in electrophysiologic studies. These observations were replicated in a number of case series31,217 and later confirmed in three RCTs.47,73,126 The results of the controlled trials and the observations with long-term use of IVIg are described later. The precise mechanisms by which IVIg exerts immunemodulating effects in CIDP are presently not known. Several possible interactions with various steps in the amplification and the effector phases of the immune response are being considered and are discussed by Stangel and co-workers.192 For a detailed overview and discussion of the use of IVIg in the treatment of the autoimmune neuropathies, the reader is referred to an excellent current review by Dalakas.35 IVIg is generally well tolerated, yet in approximately 30% of treatment courses patients will experience mild self-limited reactions (headaches, myalgias, chills). Rarely (~3%) serious complications (severe allergic reactions or thrombotic events) prohibit further use. Significant changes in laboratory findings (white blood cells, erythrocytes, hemoglobin, transaminases) may be encountered. They usually resolve spontaneously and are not clinically significant.191 Other Immunosuppressive Medications Azathioprine. Azathioprine, a broad-spectrum immunosuppressive agent, has frequently been added for CIDP patients with an incomplete response to prednisone or as a steroid-sparing agent for those suffering intolerable adverse effects.13,37,119 Although the drug is generally considered to be effective and there are sound reasons for its use

2236

Diseases of the Peripheral Nervous System

in a chronic condition such as CIDP, only uncontrolled observations are presently available. In a small, open-label RCT performed by Dyck and colleagues at the Mayo Clinic,49 the effects of 9 months’ treatment with azathioprine 2 mg/kg plus prednisone (n ⫽ 14) were compared to those with prednisone alone (n ⫽ 13). No measurable differences were detected in any of 16 outcome variables and therefore azathioprine was considered not to confer added benefits. The negative outcome of this small study is not altogether surprising because the trial was underpowered and both the dose and length of azathioprine treatment were likely not adequate. Therefore, no firm conclusions about the value of azathioprine for CIDP can presently be reached.

therapies. CsA was prescribed with a starting dose of 3 to 7 mg/kg/day; the dosage was usually adjusted later to 2 to 3 mg/kg and was maintained for a total of 6 months. Patients were closely monitored for nephrotoxicity and hypertension. In two patients the drug had to be discontinued early because of these side effects, which are usually dose dependent and reversible. Fourteen CIDP patients improved significantly in scored neurologic disability or had a reduction in annual relapse rate. Although such anecdotal observations hold promise, the experience with CsA in CIDP is very limited. The drug is therefore reserved for the exceptional circumstance of therapy-resistant CIDP.

Cyclophosphamide. Only anecdotal observations are available for use of oral or intravenous cyclophosphamide in relapsing CIDP that could otherwise not be stabilized. Prineas and McLeod162 were able to induce a sustained remission in four of five patients with relapsing CIDP with the prescription of oral cyclophosphamide (50 to 150 mg/day). In an open study of 15 CIDP patients treated with monthly pulses of intravenous cyclophosphamide (1 g/m2) in combination with dexamethasone (20 mg) for 6 months, 12 patients showed a remarkable improvement of neurologic function and either partial or complete longterm remission.65 The treatment was well tolerated, and improvements developed slowly over an average of 8.5 months (range 1 to 21 months). Recently, Brannagan and colleagues21 reported efficacy and safe use of high-dose cyclophosphamide (200 mg/kg over 4 days) without stemcell rescue in four CIDP patients refractory to all other therapies. All four patients showed sustained improvement in measured muscle strength and in functional status, which was confirmed in electrophysiologic assessments. Despite these encouraging results, use of cyclophosphamide requires very careful consideration and should be reserved for CIDP patients who are refractory or no longer responding to other less toxic medications. Potentially serious side effects such as bone marrow toxicity, azoospermia, and the late development of tumors are known complications.

Mycophenolate. Benefit from a long-term prescription of mycophenylate mofetil (CellCept) has been observed in a few therapy-resistant CIDP patients, and the drug is relatively well tolerated.25,135 Undoubtedly, lack of response has been observed by others but it remained unreported.

Autologous Stem Cell Transplantation. This technique has been successfully carried out in a single patient with a 10-year history of therapy-resistant relapsing and disabling CIDP. Vermeulen and colleagues226 provide a very detailed description of all considerations and procedures leading up to this unprecedented approach. The therapy was tolerated but it carried a mortality risk of 5%. Cyclosporin A. Cyclosporin A (CsA) has been used with some promise in selected patients who had failed other treatments.11,114 Barnett and colleagues11 summarized their experience with CsA in a retrospective analysis of 19 CIDP patients (13 with chronic progressive and 6 with relapsing courses) who had become refractory to all other

Interferons. Interferon-␤ (IFN-␤) is a naturally occurring cytokine involved in downregulating the inflammatory response. IFN-␤1a is a recombinant protein that replicates the human IFN-␤. Because the interferons have been shown to benefit patients with MS, there was good reason to try out these compounds in those with CIDP. Several reports suggested an apparent beneficial effect in a few treatment-resistant CIDP patients.27,66,116,172,207 However, no statistically significant improvement was documented in four patients with moderately severe CIDP treated prospectively for up to 10 months.103 The response to IFN␤1a in a further 10 patients with treatment-resistant CIDP was critically assessed in a double-blind, placebo-controlled trial with a crossover design, each period being 12 weeks separated by a 1-week washout period.72 There was no significant improvement in any of the outcome measures with this short-term treatment. However, in five CIDP patients IFN-␤1a appeared to be more effective when used in a higher dose and for up to 29 months.164 IFN-␣ has also been used in open assessment to treat a small number of therapy-resistant CIDP patients.66,172 IFN-␣ did not worsen the disease and was seemingly effective in a subset of patients. Although there are theoretical reasons for testing such new therapies, their value is currently very uncertain. IFN-␤ is being tested in a large RCT.

Review of RCTs and of Observations with Long-Term Use of Treatments Corticosteroids Only one RCT of 40 CIDP patients critically assessed the effect of prednisone in short-term use of 12 weeks.50 In this open study, 28 CIDP patients completed the trial. Of these, 14 patients had been assigned randomly to prednisone (120 mg/day for 1 week, then reducing the dosage by

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

20 mg/wk down to 20 mg/day, and a subsequent slower taper to 0), and 14 patients to no therapy. The two groups were well matched. The prednisone group contained an equal number of patients with primary progressive and with relapsing CIDP. On analysis, prednisone therapy had resulted in a small but significant and clinically meaningful improvement when compared to no treatment. Patients with progressive and with relapsing CIDP responded equally well to the drug. Improvement with prednisone was shown again in a second open-label RCT of 27 CIDP patients in which prednisone (initial dose 120 mg/day, tapering over 9 months; n ⫽ 13) was compared to prednisone plus azathioprine (2 mg/kg/day; n ⫽ 14). No added benefit was seen with the prescription of azathioprine.49 In a recent double-blind, controlled, crossover trial, oral prednisolone (initial dose 60 mg/day, tapering to 10 mg/day over 6 weeks) was compared to IVIg (2.0 g/kg given over 1 to 2 days) in 32 CIDP patients.83 Twenty-four of 32 patients completed both treatment periods. Prednisolone 60 mg daily (P ⫽ .005) as well as IVIg 2 g/kg (P ⫽ .012) were shown to reduce significantly disability and impairment when assessed at 2 weeks and at 6 weeks. There was no significant difference in the effects of the two treatments. On analysis of only the first trial period, which included all 32 randomized patients, a significant therapy effect was again documented for both treatments, with a slight but not significantly greater improvement with IVIg. Efficacy of prednisone in the treatment of CIDP was also shown in retrospective analysis of two large patient series, which documented improvement in 65%119 and 95%13 of patients. To achieve this benefit, high doses of prednisone (1 to 1.5 mg/kg/day) had been prescribed for a minimum of 4 weeks. Thereafter, the drug was gradually tapered off over many months. First signs of clinical improvement were observed with a mean delay of 1.9 ⫾ 3.6 months, and maximal benefit was attained only after a mean 6.6 ⫾ 5.4 months.13 Fewer than 50% of CIDP patients reached maximal benefit by 6 months; approximately 95% of patients reached maximal benefit by 12 months, and 5% were refractory to steroids. On follow-up over 10 years, about 60% of patients with chronically progressive CIDP had achieved full remission, and only 20% of patients with relapsing CIDP were not receiving any medications. Good functional outcome at long-term follow-up (Rankin score 0 to 2) was seen also by Simmons and colleagues186 in 90% of CIDP patients who had been treated with prednisone. Similar benefit from high-dose prednisone (1.5 mg/kg/day) had been documented in several uncontrolled studies in children.100,170,188,189 All children improved initially with prednisone; those with a relapsing disease course required additional immune-modulating therapies. Thus oral prednisone or prednisolone, in the thinking of some investigators, remains a mainstay of treatment for CIDP. Others (e.g., P.J.D.) find that oral prednisone usually does not improve clinical deficit as rapidly or as markedly

2237

as does IVIg or PE and is often associated with major complications (see earlier editions of this textbook). Efficacy of corticosteroids in CIDP was confirmed in a recent Cochrane meta-analysis.124 In a pilot study, Molenaar and colleagues131 treated 10 CIDP patients with six cycles of intravenous dexamethasone, 40 mg/day, for 4 consecutive days every 28 days. Six patients went into sustained remission, one patient improved but relapsed, two patients worsened, and one patient did not complete the study because of side effects. Encouraged by these results, the investigators have embarked on a RCT, which is in progress. Plasma Exchange Therapy Dyck and colleagues42 evaluated the therapy effect of PE in CIDP in a prospective double-blind, sham-controlled trial of 29 patients with either static or worsening disease. Patients were randomly assigned to receive six PE (n ⫽ 15) or six sham-exchange (n ⫽ 14) treatments over the course of 3 weeks. The investigators observed significant improvements in scored neurologic impairment (P ⫽ .025) for the PE group as a whole. They documented substantial benefit with PE in approximately one third of patients with longstanding CIDP. Yet, at the same time, some patients were also receiving other immunosuppressant drugs. Therefore, Hahn and co-workers,73 using a very similar trial design in a crossover study, examined the effect of PE in 18 CIDP patients who had not been previously treated. They recorded large and highly significant improvements in all outcome measures, including scores of neurologic impairment (P ⬍ .001), grip strength and clinical functional grading. This was also corroborated by the objective electrophysiologic measurements, which demonstrated among other findings a significant increase in proximal M-wave amplitudes, thereby indicating the reversal of CB.9 Twelve of the 15 patients (80%) who completed the blinded trial experienced substantial benefits, which usually began within days of starting PE. However, the improvements often were not sustained; 8 of 12 patients (66%) relapsed within 7 to 14 days following their last PE. They could be stabilized with repeat PE treatments and a more gradual tapering of the exchanges plus the added prescription of either prednisone or cyclophosphamide. Thus PE was regarded as a very effective adjuvant therapy. PE is considered to be safe and is remarkably free of complications. In patients with poor venous access, one may have to insert a temporary subclavian vein catheter, which carries a small risk of infection. Although PE can be carried out on an outpatient basis, the cost is nonetheless high. Moreover, the required expertise is not available at all medical centers. High-Dose IVIg Van Doorn and colleagues217 assigned seven CIDP patients, who seemingly had responded previously with IVIg, to a double-blind, controlled crossover trial. Patients

2238

Diseases of the Peripheral Nervous System

were randomly assigned to either IVIg (0.4 g/kg for 5 days) or placebo and later crossed over to the alternate therapy. All seven patients improved with IVIg and none with placebo. The observation underscored the value of IVIg for the treatment of a selected group of CIDP patients. Use and benefit of IVIg in the treatment of CIDP were further examined critically in three prospective RCTs with a double-blind, placebo-controlled design.74,126,225 Surprisingly, the first two-armed trial of 28 CIDP patients, carried out in the Netherlands, did not show benefit from IVIg and thereby contrasted remarkably with the investigators’ prior experience with open-label use of IVIg.216 On review of the patient data, it appeared that three patients of the placebo group had improved spontaneously, and that the clinical characteristics previously defined as predictors of a favorable response to IVIg were present in only 6 of the 15 patients who were randomized to active treatment, whereas they were present in 10 of 13 patients in the placebo group. On balance it was thought that these factors might have skewed the results of the trial. Nonetheless, the observations demonstrated clearly that only a subgroup of CIDP patients receive benefit from IVIg treatments. A second controlled, double-blind crossover study tested IVIg prospectively in 30 CIDP patients who had not previously been treated with this drug; 25 patients completed both arms of the trial.74 Treatment with IVIg resulted in significant improvements (P ⬍ .002) of the scored neurologic impairments grip strength and clinical functional grading, whereas with placebo scores were unchanged or worse. An analysis of only the first trial arm, including all 30 patients (16 had been randomly assigned to IVIg and 14 to placebo treatments), revealed that the improvements in scored neurologic impairment were of the same magnitude as observed with PE. Overall, significant benefit was demonstrated in 19 patients (63%): in 9 of 16 patients (56%) with chronic progressive CIDP and in 10 of 14 patients (71%) with the relapsing form. Usually, responders began to improve within 1 week of starting IVIg, and continued to improve to reach a plateau. For some patients, primarily those with a relapsing disease course, benefit was temporary, lasting only a median 6 weeks (range 3 to 22 weeks). However, improvement was restored with repeat IVIg treatment, and such patients could be stabilized with pulse IVIg treatments given at regularly scheduled intervals as single-day infusions (IVIg 1 g/kg or less). Six of the 11 patients who did not respond to IVIg subsequently improved with PE and prednisone, which illustrates the importance of trying alternate therapies. Five patients (17%) in this trial appeared to be resistant to all treatment modalities. All had a predominantly sensory presentation. Mendell and colleagues126 confirmed the value of IVIg as first-line treatment for newly diagnosed and untreated CIDP in a double-blind, controlled multicenter trial of 53 patients. Of these, 30 patients received IVIg and

23 placebo (5% albumin). The two groups were matched. Significant improvement of scored muscle strength was seen with IVIg treatment (P ⫽ .006) compared to placebo; 76% of IVIg-treated patients improved in their scored motor strength to a degree that was of functional significance. Results of the study supported the use of IVIg as the initial therapy for CIDP. In observer-blinded assessments, Dyck et al.47 compared the therapy effects of IVIg to those of PE in a crossover study and found that the two treatments conferred comparable benefits in individual patients. They were therefore considered to be equivalent in their efficacy in CIDP. This conclusion was also reached from the retrospective analysis of a large case series.26 A recent Cochrane review underscored the efficacy of IVIg in CIDP.219 Recently van Doorn and colleagues214 reviewed a 20-year experience with long-term use of IVIg in the treatment of 95 CIDP patients who were followed for a median 4.0 years (range 2.5 months to 19.5 years). IVIg had been prescribed as first-line therapy in all, and 77 of 95 CIDP patients (81%) had improved at least one grade on the modified Rankin disability scale. Of these 77 responders, 66 (86%) required additional pulse IVIg treatments for a minimum of 2 months to prevent worsening in neurologic function. To reduce the requirement for IVIg, 15 patients (20%) were treated in addition with immunosuppressant drugs. In following their CIDP patients long term, the investigators tried repeatedly to discontinue the IVIg treatments. This was possible at a median time of 2.1 years (mean 3.5 years), when patients were in stable remission. Retrospective analysis of patient data identified two factors that predict the need for ongoing IVIg therapy for more than 2 years: presence of profound weakness at the beginning of therapy, and an incomplete recovery with residual dysfunction (Rankin score ⬎ 0 to 1) after 6 months of treatment. Guided by these indicators, one would consider adding immunosuppressant drugs at approximately 6 months’ follow-up. Patients who had shown sensory and motor disturbances of relatively short duration at the start of IVIg therapy had a good chance of achieving a full remission early on. Approximately 10% of patients had been maintained in good function and on regular IVIg infusions every 2 to 6 weeks for more than 8.7 years (maximum 19.5 years). Treatment was tolerated well, and patients were not eager to switch to alternate therapy. This attitude reflects the relatively low incidence of significant adverse effects with IVIg therapy, so that recipients were willing to put up with the inconvenience and time commitment of the infusions. The cost of IVIg at current prices is very high, in particular when compared to the equally effective corticosteroid therapy. However, side effects commonly observed with a lengthy prescription of high-dose prednisone are potentially hazardous and may lead to excess morbidity and health care cost, which should not be underestimated. Formal economic modeling and cost-effectiveness analyses and comparisons of

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

the various drugs used in CIDP will have to address not only the effects but also the side effects with long-term prescription. A cost-utility comparison of IVIg and prednisone in short term use in the recent European INCAT RCT is the first of its kind and may serve as a model.122

General Approaches to Management Prior to formulating a treatment plan, it is important to thoroughly inform the patient about the heterogeneity and chronic nature of the disorder, the variable response to therapy, and the possible fluctuations in symptoms, which will have to be met with adjustments or changes in therapy. Disease activity may persist for years. Therefore, treatments are aimed at downregulating the abnormal immune responses in order to suppress the ongoing disease activity in the long term. As outlined earlier, this can be achieved with a variety of treatment modalities, which may be prescribed individually or in combination. Response to treatment varies and cannot be clearly predicted. Therefore, in a given patient, single treatments must be tested one by one and in adequate dose, monitoring the effects with a quantitative assessment of strength in proximal and distal arm and leg muscles using the Medical Research Council scale. There are currently no clear guidelines regarding the choice of the initial therapy. The order of preference may vary from patient to patient and may be dictated, at times, by medical circumstances. Initially, the choice is among the proven therapies (e.g., corticosteroids, IVIg, or PE). With this approach the chances of improvement are anywhere between 70% and 90%. Alternate immunosuppressant drugs (cytotoxic drugs) are reserved for treatment-resistant forms of CIDP. Early and aggressive medical and physical therapies are advised to restore function in a timely fashion. In general the outlook is good, yet a risk of relapse remains. Moreover, with longstanding disease or with delay in treatment, patients can develop muscle atrophy and weakness, which are caused by axonal loss and are no longer reversible.

PATHOLOGY History Prior to our report naming the disorder chronic inflammatory polyneuropathy, and later chronic inflammatory demyelinating polyradiculoneuropathy, there had been necropsy and nerve biopsy studies of patients with idiopathic, hypertrophic, and other chronic neuropathies (reviewed by Dyck and colleagues46). Some of the patients in these reports undoubtedly had CIDP, based on their clinical course, pathologic findings, and response to treatment. Support for an immune basis for both AIDP (or GBS) and CIDP came from the development of an animal model (EAN).229 Nerve or P2 protein and Freund’s adjuvant

2239

induced not only AIDP but also some cases of CIDP. Although there is no evidence that P2 protein induces AIDP or CIDP, many of the features of EAN appear to be homologous to the human disease. From our review of 53 cases of chronic inflammatory polyneuropathy (later renamed CIDP), we inferred that the neuropathy was symmetrical, that proximal nerves (sometimes even CNS) were involved, and that characteristic pathologic features were paranodal and internodal segmental demyelination, axonal degeneration, edema, epineurial and endoneurial inflammation, onion bulb formation, and scarring (Figs. 99–1 through 99–3).46 Investigators at Albert Einstein Medical School235 emphasized that macrophages were implicated in demyelination (Fig. 99–4).161,162,235 P. James B. Dyck and associates53 have reported that a particular mixed pattern of onion bulb distribution is characteristic of CIDP (Fig. 99–5). Various antibodies have been reported to be characteristic of CIDP, but generally they have failed to be very useful in diagnosis because sensitivity and specificity are still not as high as desirable (see Chapter 97).

Pathologic Hallmarks The three-dimensional pathologic alterations of the PNS in CIDP have not been extensively studied. Generally, it is a disease of the PNS, but otherwise typical cases may have changes in the CNS.46 Anatomic pathologic involvement in CIDP typically affects motor more than sensory fibers, large fibers more than small ones, and proximal (roots, plexus, and proximal nerves) more than distal nerves (see Fig. 99–1). Because the pathologic involvement is distributed multifocally, the neurologic impairment summates distally. Myelinated fibers appear to be more affected than unmyelinated fibers, resulting in weakness and especially mechanosensitive sensory loss. The characteristic lesions are a combination of fiber alterations and interstitial alterations. The typical fiber changes are segmental demyelination and remyelination, axonal degeneration and fiber loss, and reactions to these events (i.e., onion-bulb formation, fiber loss, and axonal sprouting) (see Figs. 99–2 and 99–3). The most characteristic lesion is segmental demyelination. In experimental lead neuropathy of guinea pigs, Gombault showed that the characteristic pattern of segmental demyelination was degeneration of myelin into small osmophilic or nonosmophilic lipid droplets of paranodal and internodal myelin, leaving the axon bare.64 In electron microscopic studies, the bare axon was found to be surrounded by Schwann cell cytoplasm—except for the nodes of Ranvier, naked axons are essentially never encountered. In wallerian degeneration, by contrast, large segments of myelin (ovoids) form. After segmental demyelination, new internodes of short length (0.2 to 0.3 mm) and thin myelin are formed (see

2240

Diseases of the Peripheral Nervous System

FIGURE 99–1 Transverse sections of nerve from patients with CIDP, as described in the text. A, An endoneurial mononuclear infiltrate and onion bulbs are shown. B, A semithin section showing extensive onion bulb development and marked hypomyelination. (From Dyck, P. J., Lais, A. C., Ohta, M., et al.: Chronic inflammatory polyradiculoneuropathy. Mayo Clin. Proc. 50:621, 1975, with permission.)

Figs. 99–2 and 99–3). In addition to primary segmental demyelination in CIDP, secondary demyelination also occurs, initiated by axonal atrophy.44,45,48 As a first step, myelin becomes irregular and wrinkled, followed by paranodal demyelination and then by the formation of intercalated internodes. Characteristically, demyelination occurs at multiple segments along the length of individual teased fibers, with the adjacent two fibers not showing such demyelination (see Fig. 99–2). Griffin and Price found that segmental demyelination could also occur in regions in which the axon was enlarged as a result of spheroids.70 Prineas thought that demyelination was typically initiated by macrophages, which entered Schwann cell tubes and insinuated a pseudopod into compact myelin, unraveling the lamellae of myelin and degrading it (see Fig. 99–4).161,162 Although this undoubtedly is the case, there may be other mechanisms by which segmental demyelination occurs. The mechanism by which axonal degeneration occurs also is not settled. Some investigators have attributed axonal degener-

ation to a bystander effect in the region of inflammation where cytokines are released. It is known that the inflammation in segmental demyelination causes a variable degree of CB and dispersion and axonal degeneration. The interstitial alterations include edema (Fig. 99–6; see also Fig. 99–5), inflammatory cell infiltrates, and scarring (Figs. 99–7 and 99–8; see also Fig. 99–1). The lesions tend to be quite variable by type, distribution, and severity. They are multifocal, affecting proximal to distal levels of nerves to a variable degree. It needs to be emphasized that the lesions vary considerably along the length of nerve and among fields of a transverse section. One region of a transverse section of nerve may have a relatively normal appearance, another may show only edema separating endoneurial contents (see Fig. 99–5), still another may show edema and onion bulbs (see Fig. 99–6), and still others may show scattered endoneurial lymphocytes and macrophages, florid segmental demyelination and remyelination, and even axonal degeneration.

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

2241

Length of internodes (mm)

FIGURE 99–2 Teased fiber abnormalities from the sural nerves of a 51-year-old man with chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). A, A small group of myelinated fibers with segmental demyelination and remyelination. B, A demyelinated segment, which may already have a few lamellae of new myelin. C and D, A demyelinated segment. E–G, Remyelinated segments with adjacent clumps of degenerating myelin. (From Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased-fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968, with permission.)

0.6

0.4

0.2

2

4

6

8

10

Diameter of the thickest internode (␮m)

FIGURE 99–3 Plot of internodal length versus diameter of thickest internode of myelinated nerve fibers in a case of chronic sensorimotor neuropathy (nerve 73-67). (From Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased-fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968, with permission.)

Foci of perivascular lymphocytes are frequently seen in the epineurium (see Figs. 99–7 and 99–8). Because other variants of neuropathy can show proximal and distal symmetrical involvement, as well as segmental demyelination and remyelination affecting large fibers, the question arises: What are the diagnostic pathologic features of CIDP? Endoneurial lymphocytes and macrophages, either accentuated focally or widely distributed, are a strong indicator of an inflammatory demyelinating process. Macrophages alone are less indicative of this. They may simply be a reactive change to fiber degeneration. Segmental demyelination, remyelination, or axonal degeneration is also not specific for CIDP alone. Although protein-rich edema is most reliably determined by noting the separation of endoneurial elements, separation of the space between the endoneurial contents and the perineurium is not reliable evidence of edema. By itself this is not diagnostic of CIDP, because it is also seen in other inflammatory and inherited neuropathies. Macrophage-induced segmental demyelination, a useful feature of inflammatory demyelination, is infrequently observed, and extensive electron microscopic study may be needed. The association of endoneurial

2242

Diseases of the Peripheral Nervous System

FIGURE 99–5 Mixed pattern of onion bulbs in sural nerve from a patient with CIDP. Many large onion bulbs (two examples are in left upper corner) do not contain myelinated fibers. Others contain a small or intermediate-sized myelinated fiber. Many myelinated fibers are unassociated with onion bulbs. In sural nerves, this pattern is more suggestive of CIDP than of inherited neuropathy. (Courtesy of P. James B. Dyck; modified from Dyck, P. J., Dyck, P. J. B., Giannini, C., et al.: Peripheral nerves. In Graham, D. I., and Lantos, P. L. [eds.]: Greenfield’s Neuropathology, 7th ed., Vol. 2. London, Arnold Publishing, p. 551, 2002.) See Color Plate FIGURE 99–4 Experimental allergic neuritis. A cytoplasmic process of a macrophage has inserted itself into the cytoplasm of a Schwann cell. Macrophages make contact with Schwann cells and are thought to participate in active demyelination in experimental allergic neuritis. (From Wisniewski, H., Prineas, J., and Raine, C. S.: An ultrastructural study of experimental demyelination and remyelination I. Acute experimental allergic encephalomyelitis in the peripheral nervous system. Lab. Invest. 21:105, 1969, with permission.)

lymphocytic infiltrates, segmental demyelination (with or without axonal degeneration), and the presence of onion bulb formations is the most typical histologic pattern of CIDP, but it is infrequently seen in a distal nerve biopsy specimen. The more common pattern is epineurial or endoneurial polyclonal lymphocyte infiltrates and onion bulbs. Although such findings are supportive of the diagnosis of CIDP, they are not diagnostic. P. James B. Dyck and colleagues have studied the distribution of onion bulbs in transverse sections of nerves with a view to recognize patterns that might be indicative of the underlying cause. They reported the following patterns of onion bulbs: generalized, focal, and mixed. In the mixed pattern, which was thought to favor the diagnosis of CIDP, well-developed, often large fibers were intermixed with large onion bulbs with multiple Schwann cell cytoplasmic leaflets, which is characteristic of CIDP.53 An example is shown in Figure 99–6.

Other Pathologic Causes of Chronic Polyradiculopathies The clinical pattern of chronic symmetrical polyradiculoneuropathy, involving motor and sensory fibers, low nerve conductions, and raised CSF proteins, may occur in disorders other than CIDP. The following need to be considered: amyloidosis, nonsystemic necrotizing vasculitis, sarcoidosis, inherited varieties of neuropathy, paraneoplastic disorders, plexiform neurofibromatosis, lymphoma, perineurioma, and other rare disorders. To make the distinction, the course, the pathologic alterations, and the response to immune-modulating therapy may need to be considered. CB and TD probably favors the diagnosis of CIDP, but in itself are not diagnostic. Nerve biopsy may help to confirm the diagnosis of CIDP but, perhaps more importantly, it will serve to rule out the previously mentioned pathologic entities. Patients with nonsystemic vasculitis may have prominent positive neuropathic symptoms of prickling, lancinating, or other neuropathic pain, and also numbness. The disease process is often in the distribution of a single nerve or a plexus, or it may be more broadly distributed. Small epineurial, perineurial, or endoneurial microvessels may show inflammation of their walls, as well as separation and fragmentation of mural elements (this may be demonstrated in sections reacted for smooth muscle actin). These pathologic changes may be accompanied by ischemic changes, resulting in bleeding and ischemic injury (e.g., degeneration of perineurium and endoneurium),

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

2243

FIGURE 99–6 A transverse section of sural nerve of a patient with CIDP demonstrating marked edema and early onion bulb formation. (From Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased-fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968, with permission.)

and also by evidence of regenerative activity (thickened and scarred perineurium), microfasciculation outside of old perineurium, and neovascularization. Hemosiderin is typically found adjacent to injured microvessels. In Sjögren’s neuropathy, the nonsystemic vasculitis tends to involve spinal ganglia and cutaneous nerves and salivary glands. In necrotizing vasculitis of large and small arterioles, the pattern tends to be a multiple mononeuropathy, and the systemic features of polyarteritis nodosa, rheumatoid arthritis, Wegener’s granulomatosis, and Churg-Strauss syndrome. In about 5% of these latter cases, necrotizing vasculitis may present as a symmetrical disorder.

Sural Nerve Biopsy Because CIDP is due to a multifocal interstitial process that usually affects roots and proximal to distal levels of peripheral nerve, a sural nerve biopsy may, or may not, show the worst or typical lesions of CIDP. Many authors have demonstrated increased endoneurial lymphocytes in sural nerve; others have noted that perivascular inflammation may be more noticeable, and still others that no inflammation may be found in otherwise typical cases.46,56,109,121,152,200 Varying degrees of segmental demyelination and remyelination and axonal degeneration are typical, although axonal degeneration may predominate.121,144,154,227,242 Some patients have prominent involvement of cranial nerves and spinal nerve roots. If one assumes that the lower limbs are affected, sural nerve will show fiber degeneration but it may not show the more characteristic features of CIDP, such as endoneurial inflammation, edema, segmental demyelination and remyelination, macrophage demyelination, or onion bulbs. More often in our practice, therefore, we perform a nerve biopsy not so much to confirm the diagnosis of CIDP but rather to

rule out other diseases such as amyloidosis, necrotizing vasculitis, sarcoidosis, and the like.

PATHOGENESIS Similarities with GBS, immunologic studies in patients, the beneficial response of patients to immunosuppressant therapies, and the existence of animal models provide strong evidence for an immune-mediated pathogenesis of CIDP. Animal models with a relapsing-remitting course (EAN) have been successfully established in rodents.62 Morphologic alterations of peripheral nerves are similar to those seen in human CIDP2,78 and include multifocal cellular infiltration, edema, macrophage-mediated demyelination, and onion bulb formation. These features are best reproduced in chronic relapsing EAN elicited by immunization of rabbits with galactocerebroside.

Cellular Immune Responses Evidence for systemic T-cell activation is available, although the antigen specificity of the immune response remains largely unknown.34,77,208 Studies in sural nerve biopsies from patients reveal that activated T lymphocytes invade the peripheral nerve. The T-cell populations found in situ are heterogeneous. They belong to the CD4 or CD8 subgroups and carry the ␣␤ or the ␥␦ T-cell receptor.18,87,88,180,234 The first step in focusing a systemic autoreactive response is homing to and migration through the blood-nerve barrier. This complex process of homing, adhesion, and transmigration has been extensively studied in EAN.60,193 In line with this observation, in CIDP patients, elevated serum or CSF levels, or both, of soluble adhesion molecules,159,160

2244

Diseases of the Peripheral Nervous System

FIGURE 99–7 Endoneurial inflammation in the sural nerve of a patient with CIDP. Top, A hematoxylin and eosin preparation. Middle, A CD45 preparation. Bottom, A CD68 preparation. With prominent infiltration of mononuclear cells as shown in the top frame, it is necessary to determine that the cells are polyclonal and not monoclonal (lymphoma). See Color Plate

chemokines,93,99,112,158 and matrix metalloproteinases98,108 can be detected, indicative of active T-cell migration across the blood-nerve barrier. Downregulation of the tight junction protein claudin-5 and altered location of the 220-kDa tight junction phosphoprotein Z0-1 can be observed in nerves of CIDP patients and reflect structural derangement of the blood-nerve-barrier.92 After encountering an antigen they recognize within the context of appropriate MHC molecules and co-stimulatory signals, these T cells within the PNS divide and undergo clonal expansion.60 These proliferating myelin–directed T-cells generate and release differentiation and proliferation signals, such as the helper T cell (Th) type 1 cytokines tumor necrosis factor (TNF)-␣, interferon (IFN)-␥, and interleukin-2.112,117,129 Via release of TNF-␣

FIGURE 99–8 Top, Longitudinal section of nerve root showing a slight degree of mononuclear cell infiltration adjacent to nerve fibers from case 6, a patient who died of CIDP. Middle, Longitudinal section of sural nerve from a patient with CIDP showing a low but increased frequency of mononuclear cells in endoneurium. Bottom, Patchy mononuclear cell infiltrate adjacent to a small vessel in sural nerve of a patient with CIDP. Such lesions were not associated with necrosis of the media of small arteries, with intimal proliferation, with aneurysm formation, or with hemorrhage, as is seen in angiopathic neuropathy. (Longitudinal sections in paraffin, stained with hematoxylin and eosin.) (From Dyck, P. J., Lais, A. C., Ohta, M., et al.: Chronic inflammatory polyradiculoneuropathy. Mayo Clin. Proc. 50:621, 1975, with permission.)

and IFN-␥, T cells activate resident endoneurial or immigrated macrophages to discharge an array of injurious and immunopotentiating molecules or to engage in extensive phagocytotic and cytotoxic activity against myelin or Schwann cells. By releasing interleukin-4, T cells particularly of the AV19-AJ33 T-cell receptor phenotype (V␣7.2-J␣33 invariant T cells) may locally stimulate and maintain autoantibody production by B cells.87,88 Conversely, specialized subpopulations of T cells (regulatory/suppressor T cells) may terminate the acute immuno-inflammatory process by the secretion of downregulatory Th2/Th3 cytokines (e.g., transforming growth factor-␤) or other molecules. In the nerve lesions of CIDP, messenger RNA of these anti-inflammatory cytokines, including interleukin-6,112 is upregulated. This

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

also holds for neurotrophic factors, such as nerve growth factor, glial-derived neurotrophic factor, and leukemia inhibitory factor, and their cognate receptors that might be involved in nerve regeneration.238 It is noteworthy that the local immune environment in peripheral nerve apparently provides for conditions that facilitate apoptotic cell death of invading auto-aggressive T cells.61 Macrophages represent another pathogenetically relevant cell population in CIDP nerves. They are derived from circulating monocytes that immigrate into the PNS or represent tissue resident cells.136 Macrophages are capable of releasing toxic mediators (oxygen radicals, nitric oxide metabolites, proteases, complement) targeting the peripheral nervous tissue and actively phagocytose myelin.82,97 Moreover, they serve as antigen-presenting cells and are critically important in the amplification and effector phase of immune-mediated demyelination.60,97 Their role as antigen presenters in CIDP was underscored by the observation of heightened expression of MHC class II and the MHC class–like molecule CD1a on macrophages in nerve biopsies.218 Co-stimulatory molecules such as B7-1, B7-2, and BB1 are essential for effective antigen presentation and may determine the differentiation of T lymphocytes into a Th1 or Th2 phenotype, thus modulating the local immune response and the clinical course of the disease. B7 molecules are expressed by macrophages in sural nerve biopsies from patients with CIDP.96 Schwann cells may be facultative antigen-presenting cells in CIDP because they express BB1 while the counter-receptors CTLA4 and CD28 are displayed on neighboring lymphocytes.137 Recently, Salomon and co-workers reported that the autoimmune diabetes–prone nonobese diabetic (NOD) mouse strain, deficient in B7-2 costimulation, develops a spontaneous autoimmune polyneuropathy, exhibiting clinical, electrophysiologic, and morphologic similarities to human CIDP.175 These observations emphasize the role of endoneurial macrophages as local antigen-presenting cells in the pathogenesis of inflammatory demyelination of the PNS. The cellular immune response within the PNS is tightly regulated at the transcriptional level. One of its key regulators, the transcription factor nuclear factor-␬B (NF-␬B), has been shown to be upregulated in CIDP, in contrast to its controlling inhibitory molecule, I␬B. Macrophages appear to be the major cell type expressing NF-␬B in chronic inflammatory demyelination.7 Normally, a network of immune cells and soluble factors provides a well-balanced immunoactivity that meticulously regulates the immune system within the local tissue compartment and protects its integrity. Such a protection against immune responses to host/self antigens is a key requisite for the maintenance of self-tolerance. In autoimmune diseases, self-tolerance breaks down and autoreactive T and B cells, which are part of the normal immune repertoire, are unleashed, turn auto-aggressive, and engender organ-specific damage.163 How autoreactivity

2245

escapes the regulatory mechanisms in CIDP is a critical yet to date unresolved phenomenon. One theory of special relevance to breaking tolerance in autoimmune neuropathies is the concept of molecular mimicry. This term refers to a process in which the host generates an immune response to an inciting factor, most usually an infectious organism that shares epitopic determinants with the host’s affected tissue.28 Whereas particularly in the acute motor axonal neuropathy subtype of GBS, certain targets for such an aberrant cross-reactive immune response have been identified,232,233 in CIDP such a strict correlation has not been convincingly shown. Nonetheless, some studies, exemplified by the observed association of CIDP with melanoma, point to molecular mimicry as a potentially relevant mechanism in the pathogenesis of CIDP. Although CIDP is rarely associated with carcinomas, the connection with melanoma is of great interest because both melanoma and Schwann cells derive from neural crest tissues and share common antigens. Some patients with melanoma, after therapy by vaccination with melanoma lysates, develop a subacute demyelinating polyneuropathy.16,231 Several cases of CIDP have now been reported in association with melanoma, and several carbohydrate epitopes shared by both the myelin and the tumor have been implicated as target antigens.204,231 Detection of AV19-AJ33T cell receptor–bearing T cells88 and possibly also of ␥␦ T cells in sural nerve biopsies234 raises the possibility of an enteric microbial infection that could trigger an immune-mediated assault on peripheral nerve. This may be the operative mechanism in the few cases of CIDP that follow an infective illness. More generally, present evidence suggests CIDP to be an organ-specific, immune-mediated disorder emerging from a synergistic interaction of cell-mediated and humoral immune responses directed against incompletely characterized peripheral nerve antigens.

Humoral Immune Responses Humoral immune responses to nerve antigen have long been considered fundamental to the pathogenesis of CIDP. This notion was based on the relative paucity of inflammatory infiltrates (although early morphologic studies did not include immunocytochemistry), the in vitro effects of transferred patient sera, the demonstration of immunoglobulin and complement deposition on myelinated nerve fibers,36 the presence of oligoclonal IgG bands in the CSF,38 and the therapeutic efficacy of PE. More recently, passive transfer experiments have demonstrated that serum or purified IgG from CIDP patients induces CB and demyelination in rat nerves.240 In these experiments, the 28-kDa myelin protein zero was identified as one of the putative target antigens.239 Furthermore, there is increasing albeit controversial evidence that gangliosides and other glycolipids might also represent potential autoantigens.125,203 Gangliosides are

2246

Diseases of the Peripheral Nervous System

widely distributed within the nervous system. Some of them, such as GM1, are present at the nodes of Ranvier and may be pathophysiologically strategic targets for an autoimmune attack. In one study, 15% (6 of 40) of the CIDP patients had serum antibodies against GM1125 and in another, 30% of CIDP patients had antibodies to various acidic glycolipids.89 IgM or IgG antibodies against sialosylneolactotetraosylceramide (LM1), the predominant ganglioside in human peripheral nerve myelin, have also been observed.125,237 Few patients harbor antibodies to the acidic glycolipid sulfated glucuronyl paragloboside (SGPG).243 Other studies demonstrated the presence of autoantibodies to peripheral myelin protein 22 in patients with hereditary but also with chronic inflammatory neuropathies.57,165 The potential pathophysiologic relevance of autoantibodies to gangliosides was highlighted by the observation that anti-GM1 antiserum from a CIDP patient significantly suppressed Na⫹ currents in rat single myelinated nerve fibers.197 Serum reactivity against presumably nonmyelin antigens on Schwann cells has recently been reported in a quarter of patients studied.106 Peripheral nerve demyelination and CB may also be the result of serum constituents other than myelin-specific antibodies, such as cytokines, complement, or other inflammatory mediators (e.g., nitric oxide). The low frequency of specific antibodies observed in CIDP patients suggests that different antibodies and separate mechanisms are critically involved in different individual patients.153

Axonal Loss CIDP, although a demyelinating polyneuropathy, is associated with a concomitant axonal loss. This has been attributed to the primary demyelinating process with subsequently disturbed axon-glia function.19,34 Alternatively, or in addition, it could be the result of collateral damage in the course of a vigorous inflammatory response.

REFERENCES 1. Abu-Shakra, S. R., Cornblath, D. R., Avila, O. L., et al.: Conduction block in diabetic neuropathy. Muscle Nerve 14:858, 1991. 2. Adam, A. M., Atkinson, P. F., Hall, S. M., et al.: Chronic experimental allergic neuritis in Lewis rats. Neuropathol. Appl. Neurobiol. 15:249, 1989. 3. Adams, D., Festenstein, H., Gibson, J. D., et al.: HLA antigens in chronic relapsing idiopathic inflammatory polyneuropathy. J. Neurol. Neurosurg. Psychiatry 42:184, 1979. 4. Adams, R. D., Asbury, A. K., and Michelsen, J. J.: Multifocal pseudohypertrophic neuropathy. Trans. Am. Neurol. Assoc. 90:30, 1965. 5. Albers, J. W., and Kelly, J. J. Jr.: Acquired inflammatory demyelinating polyneuropathies: clinical and electrodiagnostic features. Muscle Nerve 12:435, 1989.

6. American Academy of Neurology: Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP): report from an Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force. Neurology 41:617, 1991. 7. Andorfer, B., Kieseier, B. C., Mathey, E., et al.: Expression and distribution of transcription factor NF-kappaB and inhibitor IkappaB in the inflamed peripheral nervous system. J. Neuroimmunol. 116:226, 2001. 8. Anthoney, D. A., Bone, I., and Evans, T. R.: Inflammatory demyelinating polyneuropathy: a complication of immunotherapy in malignant melanoma. Ann. Oncol. 11:1197, 2000. 9. Ashworth, N. L., Zochodne, D. W., Hahn, A. F., et al.: Impact of plasma exchange on indices of demyelination in chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 23:206, 2000. 10. Austin, J. H.: Recurrent polyneuropathies and their corticosteroid treatment; with five-year observations of a placebocontrolled case treated with corticotrophin, cortisone, and prednisone. Brain 81:157, 1958. 11. Barnett, M. H., Pollard, J. D., Davies, L., and McLeod, J. G.: Cyclosporin A in resistant chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 21:454, 1998. 12. Barohn, R. J., Gronseth, G. S., LeForce, B. R., et al.: Peripheral nervous system involvement in a large cohort of human immunodeficiency virus-infected individuals. Arch. Neurol. 50:167, 1993. 13. Barohn, R. J., Kissel, J. T., Warmolts, J. R., and Mendell, J. R.: Chronic inflammatory demyelinating polyradiculoneuropathy: clinical characteristics, course, and recommendations for diagnostic criteria. Arch. Neurol. 46:878, 1989. 14. Bedlack, R. S., Vu, T., Hammans, S., et al.: MNGIE neuropathy: five cases mimicking chronic inflammatory demyelinating polyneuropathy. Muscle Nerve 29:364, 2004. 15. Berger, A. R., Herskovitz, S., and Kaplan, J.: Late motor involvement in cases presenting as “chronic sensory demyelinating polyneuropathy.” Muscle Nerve 18:440, 1995. 16. Bird, S. J., Brown, M. J., Shy, M. E., and Scherer, S. S.: Chronic inflammatory demyelinating polyneuropathy associated with malignant melanoma. Neurology 46:822, 1996. 17. Bosboom, W. M., van den Berg, L. H., Franssen, H., et al.: Diagnostic value of sural nerve demyelination in chronic inflammatory demyelinating polyneuropathy. Brain 124:2427, 2001. 18. Bosboom, W. M., van den Berg, L. H., Mollee, I., et al.: Sural nerve T-cell receptor Vbeta gene utilization in chronic inflammatory demyelinating polyneuropathy and vasculitic neuropathy. Neurology 56:74, 2001. 19. Bouchard, C., Lacroix, C., Plante, V., et al.: Clinicopathologic findings and prognosis of chronic inflammatory demyelinating polyneuropathy. Neurology 52:498, 1999. 20. Bradley, W. G., Bennett, R. K., Good, P., and Little, B.: Proximal chronic inflammatory polyneuropathy with multifocal conduction block. Arch. Neurol. 45:451, 1988. 21. Brannagan, T. H. 3rd, Pradhan, A., Heiman-Patterson, T., et al.: High-dose cyclophosphamide without stem-cell rescue for refractory CIDP. Neurology 58:1856, 2002.

Chronic Inflammatory Demyelinating Polyradiculoneuropathy 22. Briani, C., Brannagan, T. H. 3rd, Trojaborg, W., and Latov, N.: Chronic inflammatory demyelinating polyneuropathy. Neuromuscul. Disord. 6:311, 1996. 23. Bromberg, M. B.: Comparison of electrodiagnostic criteria for primary demyelination in chronic polyneuropathy. Muscle Nerve 14:968, 1991. 24. Busby, M., and Donaghy, M.: Chronic dysimmune neuropathy: a subclassification based upon the clinical features of 102 patients. J. Neurol. 250:714, 2003. 25. Chaudhry, V., Cornblath, D. R., Griffin, J. W., et al.: Mycophenolate mofetil: a safe and promising immunosuppressant in neuromuscular diseases. Neurology 56:94, 2001. 26. Choudhary, P. P., and Hughes, R. A.: Long-term treatment of chronic inflammatory demyelinating polyradiculoneuropathy with plasma exchange or intravenous immunoglobulin. Q. J. Med. 88:493, 1995. 27. Choudhary, P. P., Thompson, N., and Hughes, R. A.: Improvement following interferon beta in chronic inflammatory demyelinating polyradiculoneuropathy. J. Neurol. 242:252, 1995. 28. Christen, U., and von Herrath, M. G.: Induction, acceleration or prevention of autoimmunity by molecular mimicry. Mol. Immunol. 40:1113, 2004. 29. Cocito, D., Durelli, L., and Isoardo, G.: Different clinical, electrophysiological and immunological features of CIDP associated with paraproteinaemia. Acta Neurol. Scand. 108:274, 2003. 30. Connolly, A. M.: Chronic inflammatory demyelinating polyneuropathy in childhood. Pediatr. Neurol. 24:177, 2001. 31. Cornblath, D. R., Chaudhry, V., and Griffin, J. W.: Treatment of chronic inflammatory demyelinating polyneuropathy with intravenous immunoglobulin. Ann. Neurol. 30:104, 1991. 32. Costello, F., Lee, A. G., Afifi, A. K., et al.: Childhood-onset chronic inflammatory demyelinating polyradiculoneuropathy with cranial nerve involvement. J. Child Neurol. 17:819, 2002. 33. Cusimano, M. D., Bilbao, J. M., and Cohen, S. M.: Hypertrophic brachial plexus neuritis: a pathological study of two cases. Ann. Neurol. 24:615, 1988. 34. Dalakas, M. C.: Advances in chronic inflammatory demyelinating polyneuropathy: disease variants and inflammatory response mediators and modifiers. Curr. Opin. Neurol. 12:403, 1999. 35. Dalakas, M. C.: Mechanisms of action of IVIg and therapeutic considerations in the treatment of acute and chronic demyelinating neuropathies. Neurology 59(12 Suppl. 6):S13, 2002. 36. Dalakas, M. C., and Engel, W. K.: Immunoglobulin and complement deposits in nerves of patients with chronic relapsing polyneuropathy. Arch. Neurol. 37:637, 1980. 37. Dalakas, M. C., and Engel, W. K.: Chronic relapsing (dysimmune) polyneuropathy: pathogenesis and treatment. Ann. Neurol. 9(Suppl.):134, 1981. 38. Dalakas, M. C., Houff, S. A., Engel, W. K., et al.: CSF “monoclonal” bands in chronic relapsing polyneuropathy. Neurology 30:864, 1980. 39. Donaghy, M., Mills, K. R., Boniface, S. J., et al.: Pure motor demyelinating neuropathy: deterioration after steroid treatment and improvement with intravenous immunoglobulin. J. Neurol. Neurosurg. Psychiatry 57:778, 1994.

2247

40. Duarte, J., Martinez, A. C., Rodriguez, F., et al.: Hypertrophy of multiple cranial nerves and spinal roots in chronic inflammatory demyelinating neuropathy. J. Neurol. Neurosurg. Psychiatry 67:685, 1999. 41. Duggins, A. J., McLeod, J. G., Pollard, J. D., et al.: Spinal root and plexus hypertrophy in chronic inflammatory demyelinating polyneuropathy. Brain 122:1383, 1999. 42. Dyck, P. J., Daube, J., O’Brien, P., et al.: Plasma exchange in chronic inflammatory demyelinating polyradiculoneuropathy. N. Engl. J. Med. 314:461, 1986. 43. Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased-fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968. 44. Dyck, P. J., Johnson, W. J., Lambert, E. H., and O’Brien, P. C.: Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin. Proc. 46:400, 1971. 45. Dyck, P. J., and Lais, A. C.: Evidence for segmental demyelination secondary to axonal degeneration in Friedreich’s ataxia. In Kakulas, B. A. (ed.): Basic Research in Myology. Amsterdam, Excerpta Medica, p. 253, 1973. 46. Dyck, P. J., Lais, A. C., Ohta, M., et al.: Chronic inflammatory polyradiculoneuropathy. Mayo Clin. Proc. 50:621, 1975. 47. Dyck, P. J., Litchy, W. J., Kratz, K. M., et al.: A plasma exchange versus immune globulin infusion trial in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 36:838, 1994. 48. Dyck, P. J., Nukada, H., Lais, A. C., and Karnes, J. L.: Permanent axotomy: a model of chronic neuronal degeneration preceded by axonal atrophy, myelin remodeling and degeneration. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds.): Peripheral Neuropathy, 4th ed. Philadelphia, W. B. Saunders, p. 666, 1984. 49. Dyck, P. J., O’Brien, P., Swanson, C., et al.: Combined azathioprine and prednisone in chronic inflammatorydemyelinating polyneuropathy. Neurology 35:1173, 1985. 50. Dyck, P. J., O’Brien, P. C., Oviatt, K. F., et al.: Prednisone improves chronic inflammatory demyelinating polyradiculoneuropathy more than no treatment. Ann. Neurol. 11:136, 1982. 51. Dyck, P. J., Pineda, A., Swanson, C., et al.: The Mayo Clinic experience with plasma exchange in chronic inflammatorydemyelinating polyneuropathy (CIDP). Prog. Clin. Biol. Res. 106:197, 1982. 52. Dyck, P. J., Swanson, C. J., Low, P. A., et al.: Prednisoneresponsive hereditary motor and sensory neuropathy. Mayo Clin. Proc. 57:239, 1982. 53. Dyck, P. J. B., Engelstad, J., and Dyck, P. J.: Pattern of onion bulb distribution predicts acquired versus inherited hypertrophic neuropathy. Ann. Neurol. 46:482, 1999. 54. Feasby, T. E., Hahn, A. F., Koopman, W. J., and Lee, D. H.: Central lesions in chronic inflammatory demyelinating polyneuropathy: an MRI study. Neurology 40:476, 1990. 55. Feeney, D. J., Pollard, J. D., McLeod, J. G., et al.: HLA antigens in chronic inflammatory demyelinating polyneuropathy. J. Neurol. Neurosurg. Psychiatry 53:170, 1990. 56. Fowler, H., Vulpe, M., Marks, G., et al.: Recovery from chronic progressive polyneuropathy after treatment with

2248

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

Diseases of the Peripheral Nervous System

plasma exchange and cyclophosphamide. Lancet 2:1193, 1979. Gabriel, C. M., Gregson, N. A., and Hughes, R. A.: AntiPMP22 antibodies in patients with inflammatory neuropathy. J. Neuroimmunol. 104:139, 2000. Ginsberg, L., Malik, O., and Kenton, A. R., et al.: Coexistent hereditary and inflammatory neuropathy. Brain 127:193, 2004. Ginsberg, L., Platts, A. D., and Thomas, P. K.: Chronic inflammatory demyelinating polyneuropathy mimicking a lumbar spinal stenosis syndrome. J. Neurol. Neurosurg. Psychiatry 59:189, 1995. Gold, R., Archelos, J. J., and Hartung, H. P.: Mechanisms of immune regulation in the peripheral nervous system. Brain Pathol. 9:343, 1999. Gold, R., Hartung, H. P., and Lassmann, H.: T-cell apoptosis in autoimmune diseases: termination of inflammation in the nervous system and other sites with specialized immunedefense mechanisms. Trends Neurosci. 20:399, 1997. Gold, R., Hartung, H. P., and Toyka, K. V.: Animal models for autoimmune demyelinating disorders of the nervous system. Mol. Med. Today 6:88, 2000. Goldstein, J. M., Parks, B. J., Mayer, P. L., et al.: Nerve root hypertrophy as the cause of lumbar stenosis in chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 19:892, 1996. Gombault, A.: Contribution a l’etude anatomique de la nevrite parenchamenteuse subague et chronique: nevrite sementaire peri-axile. Arch. Neurol. (Paris) 1:11, 1880. Good, J. L., Chehrenama, M., Mayer, R. F., and Koski, C. L.: Pulse cyclophosphamide therapy in chronic inflammatory demyelinating polyneuropathy. Neurology 51:1735, 1998. Gorson, K. C., Allam, G., Simovic, D., and Ropper, A. H.: Improvement following interferon-alpha 2A in chronic inflammatory demyelinating polyneuropathy. Neurology 48:777, 1997. Gorson, K. C., Ropper, A. H., Adelman, L. S., and Weinberg, D. H.: Influence of diabetes mellitus on chronic inflammatory demyelinating polyneuropathy. Muscle Nerve 23:37, 2000. Gorson, K. C., Ropper, A. H., and Weinberg, D. H.: Upper limb predominant, multifocal chronic inflammatory demyelinating polyneuropathy. Muscle Nerve 22:758, 1999. Grand’Maison, F., Feasby, T. E., Hahn, A. F., and Koopman, W. J.: Recurrent Guillain-Barre syndrome: clinical and laboratory features. Brain 115:1093, 1992. Griffin, J. W., and Price, D. L.: Proximal axonopathies induced by toxic chemicals. In Spencer, P. S., and Schaumberg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 161, 1980. Gross, M. L., and Thomas, P. K.: The treatment of chronic relapsing and chronic progressive idiopathic inflammatory polyneuropathy by plasma exchange. J. Neurol. Sci. 52:69, 1981. Hadden, R. D., Sharrack, B., Bensa, S., et al.: Randomized trial of interferon beta-1a in chronic inflammatory demyelinating polyradiculoneuropathy. Neurology 53:57, 1999. Hahn, A. F., Bolton, C. F., Pillay, N., et al.: Plasma-exchange therapy in chronic inflammatory demyelinating polyneu-

74.

75.

76. 77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

ropathy: a double-blind, sham-controlled, cross-over study. Brain 119(Pt. 4):1055, 1996. Hahn, A. F., Bolton, C. F., Zochodne, D., and Feasby, T. E.: Intravenous immunoglobulin treatment in chronic inflammatory demyelinating polyneuropathy: a double-blind, placebocontrolled, cross-over study. Brain 119:1067, 1996. Haq, R. U., Pendlebury, W. W., Fries, T. J., and Tandan, R.: Chronic inflammatory demyelinating polyradiculoneuropathy in diabetic patients. Muscle Nerve 27:465, 2003. Harris, W., and Newcomb, W. D.: A case of relapsing interstitial hypertrophic polyneuritis. Brain 52:108, 1929. Hartung, H. P., Reiners, K., Schmidt, B., et al.: Serum interleukin-2 concentrations in Guillain-Barre syndrome and chronic idiopathic demyelinating polyradiculoneuropathy: comparison with other neurological diseases of presumed immunopathogenesis. Ann. Neurol. 30:48, 1991. Harvey, G. K., Pollard, J. D., Schindhelm, K., and Antony, J.: Chronic experimental allergic neuritis: an electrophysiological and histological study in the rabbit. J. Neurol. Sci. 81:215, 1987. Hattori, N., Ichimura, M., Aoki, S., et al.: Clinicopathological features of chronic inflammatory demyelinating polyradiculoneuropathy in childhood. J. Neurol. Sci. 154:66, 1998. Hattori, N., Misu, K., Koike, H., et al.: Age of onset influences clinical features of chronic inflammatory demyelinating polyneuropathy. J. Neurol. Sci. 184:57, 2001. Hinman, R. C., and Magee, K. R.: Guillain-Barre syndrome with slow progressive onset and persistent elevation of spinal fluid protein. Ann. Intern. Med. 67:1007, 1967. Hu, W., Mathey, E., Hartung, H. P., and Kieseier, B. C.: Cyclo-oxygenases and prostaglandins in acute inflammatory demyelination of the peripheral nerve. Neurology 61:1774, 2003. Hughes, R., Bensa, S., Willison, H., et al.: Randomized controlled trial of intravenous immunoglobulin versus oral prednisolone in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:195, 2001. Hughes, R., Sanders, E., Hall, S., et al.: Subacute idiopathic demyelinating polyradiculoneuropathy. Arch. Neurol. 49:612, 1992. Hughes, R. A., Choudhary, P. P., Osborn, M., et al.: Immunization and risk of relapse of Guillain-Barre syndrome or chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 19:1230, 1996. Hughes, R. A., and van Der Meche, F. G.: Corticosteroids for treating Guillain-Barre syndrome. Cochrane Database Syst. Rev. 3:CD001446, 2000. Illes, Z., Kondo, T., Newcombe, J., et al.: Differential expression of NK T cell V alpha 24J alpha Q invariant TCR chain in the lesions of multiple sclerosis and chronic inflammatory demyelinating polyneuropathy. J. Immunol. 164:4375, 2000. Illes, Z., Shimamura, M., Newcombe, J., et al.: Accumulation of Valpha7.2-Jalpha33 invariant T cells in human autoimmune inflammatory lesions in the nervous system. Int. Immunol. 16:223, 2004. Ilyas, A. A., Mithen, F. A., Dalakas, M. C., et al.: Antibodies to acidic glycolipids in Guillain-Barre syndrome and chronic inflammatory demyelinating polyneuropathy. J. Neurol. Sci. 107:111, 1992.

Chronic Inflammatory Demyelinating Polyradiculoneuropathy 90. Jeanjean, A. P., Duprez, T., and van den Bergh, P. Y.: Massive peripheral nerve hypertrophy in a patient with multifocal upper limb demyelinating neuropathy (Lewis-Sumner syndrome). Acta Neurol. Belg. 101:234, 2001. 91. Kaji, R., Shibasaki, H., and Kimura, J.: Multifocal demyelinating motor neuropathy: cranial nerve involvement and immunoglobulin therapy. Neurology 42(3 Pt. 1):506, 1992. 92. Kanda, T., Numata, Y., and Mizusawa, H.: Chronic inflammatory demyelinating polyneuropathy: decreased claudin-5 and relocated ZO-1. J. Neurol. Neurosurg. Psychiatry 75:765, 2004. 93. Kastenbauer, S., Koedel, U., Wick, M., et al.: CSF and serum levels of soluble fractalkine (CX3CL1) in inflammatory diseases of the nervous system. J. Neuroimmunol. 137:210, 2003. 94. Katz, J. S., and Saperstein, D. S.: Asymmetric acquired demyelinating polyneuropathies: MMN and MADSAM. Curr. Treat. Options Neurol. 3:119, 2001. 95. Katz, J. S., Saperstein, D. S., Gronseth, G., et al.: Distal acquired demyelinating symmetric neuropathy. Neurology 54:615, 2000. 96. Kiefer, R., Dangond, F., Mueller, M., et al.: Enhanced B7 costimulatory molecule expression in inflammatory human sural nerve biopsies. J. Neurol. Neurosurg. Psychiatry 69:362, 2000. 97. Kiefer, R., Kieseier, B. C., Stoll, G., and Hartung, H. P.: The role of macrophages in immune-mediated damage to the peripheral nervous system. Prog. Neurobiol. 64:109, 2001. 98. Kieseier, B. C., Kiefer, R., Clements, J. M., et al.: Matrix metalloproteinase-9 and -7 are regulated in experimental autoimmune encephalomyelitis. Brain 121:159, 1998. 99. Kieseier, B. C., Tani, M., Mahad, D., et al.: Chemokines and chemokine receptors in inflammatory demyelinating neuropathies: a central role for IP-10. Brain 125(Pt. 4):823, 2002. 100. Korinthenberg, R.: Chronic inflammatory demyelinating polyradiculoneuropathy in children and their response to treatment. Neuropediatrics 30:190, 1999. 101. Krendel, D. A., Costigan, D. A., and Hopkins, L. C.: Successful treatment of neuropathies in patients with diabetes mellitus. Arch. Neurol. 52:1053, 1995. 102. Kretzer, R. M., Burger, P. C., and Tamargo, R. J.: Hypertrophic neuropathy of the cauda equina: case report. Neurosurgery 54:515, 2004; discussion 518. 103. Kuntzer, T., Radziwill, A. J., Lettry-Trouillat, R., et al.: Interferon-beta1a in chronic inflammatory demyelinating polyneuropathy. Neurology 53:1364, 1999. 104. Kusumi, M., Nakashima, K., Nakayama, H., and Takahashi, K.: Epidemiology of inflammatory neurological and inflammatory neuromuscular diseases in Tottori Prefecture, Japan. Psychiatry Clin. Neurosci. 49:169, 1995. 105. Kuwabara, S., Nakajima, M., Matsuda, S., and Hattori, T.: Magnetic resonance imaging at the demyelinative foci in chronic inflammatory demyelinating polyneuropathy. Neurology 48:874, 1997. 106. Kwa, M. S., van Schaik, I. N., de Jonge, R. R., et al.: Autoimmunoreactivity to Schwann cells in patients with inflammatory neuropathies. Brain 126:361, 2003. 107. Lassmann, H., Budka, H., and Schnaberth, G.: Inflammatory demyelinating polyradiculitis in a patient with multiple sclerosis. Arch. Neurol. 38:99, 1981.

2249

108. Leppert, D., Hughes, P., Huber, S., et al.: Matrix metalloproteinase upregulation in chronic inflammatory demyelinating polyneuropathy and nonsystemic vasculitic neuropathy. Neurology 53:62, 1999. 109. Levy, R. L., Newkirk, R., and Ochoa, J.: Treating chronic relapsing Guillain-Barre syndrome by plasma exchange. Lancet 2:259, 1979. 110. Lewis, R. A.: Multifocal motor neuropathy and Lewis Sumner syndrome: two distinct entities. Muscle Nerve 22:1738, 1999. 111. Lewis, R. A., Sumner, A. J., Brown, M. J., and Asbury, A. K.: Multifocal demyelinating neuropathy with persistent conduction block. Neurology 32:958, 1982. 112. Lindenlaub, T., and Sommer, C.: Cytokines in sural nerve biopsies from inflammatory and non-inflammatory neuropathies. Acta Neuropathol. (Berl.) 105:593, 2003. 113. Lunn, M. P., Manji, H., Choudhary, P. P., et al.: Chronic inflammatory demyelinating polyradiculoneuropathy: a prevalence study in south east England. J. Neurol. Neurosurg. Psychiatry 66:677, 1999. 114. Mahattanakul, W., Crawford, T. O., Griffin, J. W., et al.: Treatment of chronic inflammatory demyelinating polyneuropathy with cyclosporin-A. J. Neurol. Neurosurg. Psychiatry 60:185, 1996. 115. Maisonobe, T., Chassande, B., Verin, M., et al.: Chronic dysimmune demyelinating polyneuropathy: a clinical and electrophysiological study of 93 patients. J. Neurol. Neurosurg. Psychiatry 61:36, 1996. 116. Martina, I. S., van Doorn, P. A., Schmitz, P. I., et al.: Chronic motor neuropathies: response to interferon-beta1a after failure of conventional therapies. J. Neurol. Neurosurg. Psychiatry 66:197, 1999. 117. Mathey, E. K., Pollard, J. D., and Armati, P. J.: TNF alpha, IFN gamma and IL-2 mRNA expression in CIDP sural nerve biopsies. J. Neurol. Sci. 163:47, 1999. 118. Matsuda, M., Ikeda, S., Sakurai, S., et al.: Hypertrophic neuritis due to chronic inflammatory demyelinating polyradiculoneuropathy (CIDP): a postmortem pathological study. Muscle Nerve 19:163, 1996. 119. McCombe, P. A., McManis, P. G., Frith, J. A., et al.: Chronic inflammatory demyelinating polyradiculoneuropathy associated with pregnancy. Ann. Neurol. 21:102, 1987. 120. McCombe, P. A., Pollard, J. D., and McLeod, J. G.: Chronic inflammatory demyelinating polyradiculoneuropathy: a clinical and electrophysiological study of 92 cases. Brain 110:1617, 1987. 121. McCombe, P. A., van der Kreek, S. A., and Pender, M. P.: The effects of prophylactic cyclosporin A on experimental allergic neuritis (EAN) in the Lewis rat: induction of relapsing EAN using low dose cyclosporin A. J. Neuroimmunol. 28:131, 1990. 122. McCrone, P., Chisholm, D., Knapp, M., et al.: Cost-utility analysis of intravenous immunoglobulin and prednisolone for chronic inflammatory demyelinating polyradiculoneuropathy. Eur. J. Neurol. 10:687, 2003. 123. McLeod, J. G., Pollard, J. D., Macaskill, P., et al.: Prevalence of chronic inflammatory demyelinating polyneuropathy in New South Wales, Australia. Ann. Neurol. 46:910, 1999. 124. Mehndiratta, M. M., and Hughes, R. A.: Corticosteroids for chronic inflammatory demyelinating polyradiculoneuropathy. Cochrane Database Syst. Rev. 3:CD002062, 2001.

2250

Diseases of the Peripheral Nervous System

125. Melendez-Vasquez, C., Redford, J., Choudhary, P. P., et al.: Immunological investigation of chronic inflammatory demyelinating polyradiculoneuropathy. J. Neuroimmunol. 73:124, 1997. 126. Mendell, J. R., Barohn, R. J., Freimer, M. L., et al.: Randomized controlled trial of IVIg in untreated chronic inflammatory demyelinating polyradiculoneuropathy. Neurology 56:445, 2001. 127. Mendell, J. R., Kolkin, S., Kissel, J. T., et al.: Evidence for central nervous system demyelination in chronic inflammatory demyelinating polyradiculoneuropathy. Neurology 37:1291, 1987. 128. Midroni, G., and Dyck, P. J.: Chronic inflammatory demyelinating polyradiculoneuropathy: unusual clinical features and therapeutic responses. Neurology 46:1206, 1996. 129. Misawa, S., Kuwabara, S., Mori, M., et al.: Serum levels of tumor necrosis factor-alpha in chronic inflammatory demyelinating polyneuropathy. Neurology 56:666, 2001. 130. Mizuno, K., Nagamatsu, M., Hattori, N., et al.: Chronic inflammatory demyelinating polyradiculoneuropathy with diffuse and massive peripheral nerve hypertrophy: distinctive clinical and magnetic resonance imaging features. Muscle Nerve 21:805, 1998. 131. Molenaar, D. S., van Doorn, P. A., and Vermeulen, M.: Pulsed high dose dexamethasone treatment in chronic inflammatory demyelinating polyneuropathy: a pilot study. J. Neurol. Neurosurg. Psychiatry 62:388, 1997. 132. Molenaar, D. S., Vermeulen, M., and de Haan, R.: Diagnostic value of sural nerve biopsy in chronic inflammatory demyelinating polyneuropathy. J. Neurol. Neurosurg. Psychiatry 64:84, 1998. 133. Molenaar, D. S., Vermeulen, M., and de Haan, R. J.: Comparison of electrodiagnostic criteria for demyelination in patients with chronic inflammatory demyelinating polyneuropathy (CIDP). J. Neurol. 249:400, 2002. 134. Mori, K., Hattori, N., Sugiura, M., et al.: Chronic inflammatory demyelinating polyneuropathy presenting with features of GBS. Neurology 58:979, 2002. 135. Mowzoon, N., Sussman, A., and Bradley, W. G.: Mycophenolate (CellCept) treatment of myasthenia gravis, chronic inflammatory polyneuropathy and inclusion body myositis. J. Neurol. Sci. 185:119, 2001. 136. Mueller, M., Wacker, K., Ringelstein, E. B., et al.: Rapid response of identified resident endoneurial macrophages to nerve injury. Am. J. Pathol. 159:2187, 2001. 137. Murata, K., and Dalakas, M. C.: Expression of the co-stimulatory molecule BB-1, the ligands CTLA-4 and CD28 and their mRNAs in chronic inflammatory demyelinating polyneuropathy. Brain 123:1660, 2000. 138. Mygland, A., and Monstad, P.: Chronic acquired demyelinating symmetric polyneuropathy classified by pattern of weakness. Arch. Neurol. 60:260, 2003. 139. Nagamatsu, M., Terao, S., Misu, K., et al.: Axonal and perikaryal involvement in chronic inflammatory demyelinating polyneuropathy. J. Neurol. Neurosurg. Psychiatry 66:727, 1999. 140. Nevo, Y., Pestronk, A., Kornberg, A. J., et al.: Childhood chronic inflammatory demyelinating neuropathies: clinical course and long-term follow-up. Neurology 47:98, 1996.

141. Nevo, Y., and Topaloglu, H.: 88th ENMC International Workshop: childhood chronic inflammatory demyelinating polyneuropathy (including revised diagnostic criteria), Naarden, The Netherlands, December 8–10, 2000. Neuromuscul. Disord. 12:195, 2002. 142. Nicolas, G., Maisonobe, T., Le Forestier, N., et al.: Proposed revised electrophysiological criteria for chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 25:26, 2002. 143. Odaka, M., Yuki, N., and Hirata, K.: Patients with chronic inflammatory demyelinating polyneuropathy initially diagnosed as Guillain-Barre syndrome. J. Neurol. 250:913, 2003. 144. Oh, S. J.: Subacute demyelinating polyneuropathy responding to corticosteroid treatment. Arch. Neurol. 35:509, 1978. 145. Oh, S. J., Claussen, G. C., and Kim, D. S.: Motor and sensory demyelinating mononeuropathy multiplex (multifocal motor and sensory demyelinating neuropathy): a separate entity or a variant of chronic inflammatory demyelinating polyneuropathy? J. Peripher. Nerv. Syst. 2:362, 1997. 146. Oh, S. J., Joy, J. L., and Kuruoglu, R.: “Chronic sensory demyelinating neuropathy”: chronic inflammatory demyelinating polyneuropathy presenting as a pure sensory neuropathy. J. Neurol. Neurosurg. Psychiatry 55:677, 1992. 147. Oh, S. J., Kurokawa, K., de Almeida, D. F., et al.: Subacute inflammatory demyelinating polyneuropathy. Neurology 61:1507, 2003. 148. Ohkoshi, N., Harada, K., Nagata, H., et al.: Ataxic form of chronic inflammatory demyelinating polyradiculoneuropathy: clinical features and pathological study of the sural nerves. Eur. Neurol. 45:241, 2001. 149. Olney, R. K.: Consensus criteria for the diagnosis of partial conduction block. Muscle Nerve 22(Suppl. 8):S225, 1999. 150. Pardo, C. A., McArthur, J. C., and Griffin, J. W.: HIV neuropathy: insights in the pathology of HIV peripheral nerve disease. J. Peripher. Nerv. Syst. 6:21, 2001. 151. Parry, G. J.: Are multifocal motor neuropathy and LewisSumner syndrome distinct nosologic entities? Muscle Nerve 22:557, 1999. 152. Pollard, J. D.: A critical review of therapies in acute and chronic inflammatory demyelinating polyneuropathies. Muscle Nerve 10:214, 1987. 153. Pollard, J. D.: Chronic inflammatory demyelinating polyradiculoneuropathy. Curr. Opin. Neurol. 15:279, 2002. 154. Pollard, J. D., King, R. H., and Thomas, P. K.: Recurrent experimental allergic neuritis: an electron microscope study. J. Neurol. Sci. 24:365, 1975. 155. Pollard, J. D., McLeod, J. G., Gatenby, P., and Kronenberg, H.: Prediction of response to plasma exchange in chronic relapsing polyneuropathy: a clinico-pathological correlation. J. Neurol. Sci. 58:269, 1983. 156. Pollard, J. D., and Selby, G.: Relapsing neuropathy due to tetanus toxoid: report of a case. J. Neurol. Sci. 37:113, 1978. 157. Prensky, A. L., and Dodson, W. E.: The steroid treatment of hereditary motor and sensory neuropathy. Neuropediatrics 15:203, 1984. 158. Press, R., Pashenkov, M., Jin, J. P., and Link, H.: Aberrated levels of cerebrospinal fluid chemokines in Guillain-Barre syndrome and chronic inflammatory demyelinating polyradiculoneuropathy. J. Clin. Immunol. 23:259, 2003.

Chronic Inflammatory Demyelinating Polyradiculoneuropathy 159. Previtali, S. C., Archelos, J. J., and Hartung, H. P.: Expression of integrins in experimental autoimmune neuritis and Guillain-Barre syndrome. Ann. Neurol. 44:611, 1998. 160. Previtali, S. C., Feltri, M. L., Archelos, J. J., et al.: Role of integrins in the peripheral nervous system. Prog. Neurobiol. 64:35, 2001. 161. Prineas, J. W.: Demyelination and remyelination in recurrent idiopathic polyneuropathy: an electron microscope study. Acta Neuropathol. (Berl.) 18:34, 1971. 162. Prineas, J. W., and McLeod, J. G.: Chronic relapsing polyneuritis. J. Neurol. Sci. 27:427, 1976. 163. Quattrini, A., Previtali, S. C., Kieseier, B. C., et al.: Autoimmunity in the peripheral nervous system. Crit. Rev. Neurobiol. 15:1, 2003. 164. Radziwill, AJ KT, Fuhr, P., et al.: Long term follow up in chronic inflammatory demyelinating polyneuropathy (CIDP) under interferon beta-1a (IFN-B1a) and intravenous immunoglobulin (IVIg). J. Neurol. Neurosurg. Psychiatry 70:273, 2001. 165. Ritz, M. F., Lechner-Scott, J., Scott, R. J., et al.: Characterisation of autoantibodies to peripheral myelin protein 22 in patients with hereditary and acquired neuropathies. J. Neuroimmunol. 104:155, 2000. 166. Ro, Y. I., Alexander, C. B., and Oh, S. J.: Multiple sclerosis and hypertrophic demyelinating peripheral neuropathy. Muscle Nerve 6:312, 1983. 167. Rodriguez-Casero, M. V., Shield, L. K., Coleman, L. T., and Kornberg, A. J.: Childhood chronic inflammatory demyelinating polyneuropathy with central nervous system demyelination resembling multiple sclerosis. Neuromuscul. Disord. 13:158, 2003. 168. Rotta, F. T., Sussman, A. T., Bradley, W. G., et al.: The spectrum of chronic inflammatory demyelinating polyneuropathy. J. Neurol. Sci. 173:129, 2000. 169. Rubin, M., and Mackenzie, C. R.: Clinically and electrodiagnostically pure sensory demyelinating polyneuropathy. Electromyogr. Clin. Neurophysiol. 36:145, 1996. 170. Ryan, M. M., Grattan-Smith, P. J., Procopis, P. G., et al.: Childhood chronic inflammatory demyelinating polyneuropathy: clinical course and long-term outcome. Neuromuscul. Disord. 10:398, 2000. 171. Sabatelli, M., Madia, F., Mignogna, T., et al.: Pure motor chronic inflammatory demyelinating polyneuropathy. J. Neurol. 248:772, 2001. 172. Sabatelli, M., Mignogna, T., Lippi, G., et al.: Interferon-alpha may benefit steroid unresponsive chronic inflammatory demyelinating polyneuropathy. J. Neurol. Neurosurg. Psychiatry 58:638, 1995. 173. Said, G.: Chronic inflammatory demyelinative polyneuropathy. J. Neurol. 249:245, 2002. 174. Sakakibara, R., Hattori, T., Kuwabara, S., et al.: Micturitional disturbance in patients with chronic inflammatory demyelinating polyneuropathy. Neurology 50:1179, 1998. 175. Salomon, B., Rhee, L., Bour-Jordan, H., et al.: Development of spontaneous autoimmune peripheral polyneuropathy in B7–2-deficient NOD mice. J. Exp. Med. 194:677, 2001. 176. Saperstein, D. S., Amato, A. A., Wolfe, G. I., et al.: Multifocal acquired demyelinating sensory and motor neuropathy: the Lewis-Sumner syndrome. Muscle Nerve 22:560, 1999.

2251

177. Saperstein, D. S., and Barohn, R. J.: Current concepts and controversy in chronic inflammatory demyelinating polyneuropathy. Curr. Neurol. Neurosci. Rep. 3:57, 2003. 178. Saperstein, D. S., Katz, J. S., Amato, A. A., and Barohn, R. J.: Clinical spectrum of chronic acquired demyelinating polyneuropathies. Muscle Nerve 24:311, 2001. 179. Schady, W., Goulding, P. J., Lecky, B. R., et al.: Massive nerve root enlargement in chronic inflammatory demyelinating polyneuropathy. J. Neurol. Neurosurg. Psychiatry 61:636, 1996. 180. Schmidt, B., Toyka, K. V., Kiefer, R., et al.: Inflammatory infiltrates in sural nerve biopsies in Guillain-Barre syndrome and chronic inflammatory demyelinating neuropathy. Muscle Nerve 19:474, 1996. 181. Schoene, W. C., Carpenter, S., Behan, P. O., and Geschwind, N.: ‘Onion bulb’ formations in the central and peripheral nervous system in association with multiple sclerosis and hypertrophic polyneuropathy. Brain 100:755, 1977. 182. Server, A. C., Lefkowith, J., Braine, H., and McKhann, G. M.: Treatment of chronic relapsing inflammatory polyradiculoneuropathy by plasma exchange. Ann. Neurol. 6:258, 1979. 183. Sghirlanzoni, A., Solari, A., Ciano, C., et al.: Chronic inflammatory demyelinating polyradiculoneuropathy: long-term course and treatment of 60 patients. Neurol. Sci. 21:31, 2000. 184. Sharma, K. R., Cross, J., Farronay, O., et al.: Demyelinating neuropathy in diabetes mellitus. Arch. Neurol. 59:758, 2002. 185. Simmons, Z., Albers, J. W., Bromberg, M. B., and Feldman, E. L.: Presentation and initial clinical course in patients with chronic inflammatory demyelinating polyradiculoneuropathy: comparison of patients without and with monoclonal gammopathy. Neurology 43:2202, 1993. 186. Simmons, Z., Albers, J. W., Bromberg, M. B., and Feldman, E. L.: Long-term follow-up of patients with chronic inflammatory demyelinating polyradiculoneuropathy, without and with monoclonal gammopathy. Brain 118:359, 1995. 187. Simmons, Z., Wald, J. J., and Albers, J. W.: Chronic inflammatory demyelinating polyradiculoneuropathy in children: I. Presentation, electrodiagnostic studies, and initial clinical course, with comparison to adults. Muscle Nerve 20:1008, 1997. 188. Simmons, Z., Wald, J. J., and Albers, J. W.: Chronic inflammatory demyelinating polyradiculoneuropathy in children: II. Long-term follow-up, with comparison to adults. Muscle Nerve 20:1569, 1997. 189. Sladky, J. T., Brown, M. J., and Berman, P. H.: Chronic inflammatory demyelinating polyneuropathy of infancy: a corticosteroid-responsive disorder. Ann. Neurol. 20:76, 1986. 190. Snider, W. D., Simpson, D. M., Nielsen, S., et al.: Neurological complications of acquired immune deficiency syndrome: analysis of 50 patients. Ann. Neurol. 14:403, 1983. 191. Stangel, M., Kiefer, R., Pette, M., et al.: Side effects of intravenous immunoglobulins in neurological autoimmune disorders—a prospective study. J. Neurol. 250:818, 2003. 192. Stangel, M., Toyka, K. V., and Gold, R.: Mechanisms of high-dose intravenous immunoglobulins in demyelinating diseases. Arch. Neurol. 56:661, 1999.

2252

Diseases of the Peripheral Nervous System

193. Steck, A. J., Schaeren-Wiemers, N., and Hartung, H. P.: Demyelinating inflammatory neuropathies, including Guillain-Barre syndrome. Curr. Opin. Neurol. 11:311, 1998. 194. Stewart, G. J., Pollard, J. D., McLeod, J. G., and Wolnizer, C. M.: HLA antigens in the Landry-Guillain-Barre syndrome and chronic relapsing polyneuritis. Ann. Neurol. 4:285, 1978. 195. Stewart, J. D., McKelvey, R., Durcan, L., et al.: Chronic inflammatory demyelinating polyneuropathy (CIDP) in diabetics. J. Neurol. Sci. 142:59, 1996. 196. Stojkovic, T., De Seze, J., Hurtevent, J. F., et al.: Phrenic nerve palsy as a feature of chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 27:497, 2003. 197. Takigawa, T., Yasuda, H., Terada, M., et al.: The sera from GM1 ganglioside antibody positive patients with GuillainBarre syndrome or chronic inflammatory demyelinating polyneuropathy blocks Na⫹ currents in rat single myelinated nerve fibers. Intern. Med. 39:123, 2000. 198. Thaisetthawatkul, P., Logigian, E. L., and Herrmann, D. N.: Dispersion of the distal compound muscle action potential as a diagnostic criterion for chronic inflammatory demyelinating polyneuropathy. Neurology 59:1526, 2002. 199. Thomas, P. K., Claus, D., Jaspert, A., et al.: Focal upper limb demyelinating neuropathy. Brain 119:765, 1996. 200. Thomas, P. K., Lascelles, R. G., Hallpike, J. F., and Hewer, R. L.: Recurrent and chronic relapsing Guillain-Barre polyneuritis. Brain 92:589, 1969. 201. Thomas, P. K., Walker, R. W., Rudge, P., et al.: Chronic demyelinating peripheral neuropathy associated with multifocal central nervous system demyelination. Brain 110:53, 1987. 202. Torvik, A., and Lundar, T.: A case of chronic demyelinating polyneuropathy resembling the Guillain-Barre syndrome. J. Neurol. Sci. 32:45, 1977. 203. Toyka, K. V., and Hartung, H. P.: Chronic inflammatory polyneuritis and neuropathies. Curr. Opin. Neurol. 9:240, 1996. 204. Tsuchida, T., Saxton, R. E., Morton, D. L., and Irie, R. F.: Gangliosides of human melanoma. Cancer 63:1166, 1989. 205. Uncini, A., De Angelis, M. V., Di Muzio, A., et al.: Chronic inflammatory demyelinating polyneuropathy in diabetics: motor conductions are important in the differential diagnosis with diabetic polyneuropathy. Clin. Neurophysiol. 110:705, 1999. 206. Uncini, A., Di Muzio, A., De Angelis, M. V., et al.: Minimal and asymptomatic chronic inflammatory demyelinating polyneuropathy. Clin. Neurophysiol. 110:694, 1999. 207. Vallat, J. M., Hahn, A. F., Leger, J. M., et al.: Interferon beta-1a as an investigational treatment for CIDP. Neurology 60(8 Suppl. 3):S23, 2003. 208. Van den Berg, L. H., Mollee, I., Wokke, J. H., and Logtenberg, T.: Increased frequencies of HPRT mutant T lymphocytes in patients with Guillain-Barre syndrome and chronic inflammatory demyelinating polyneuropathy: further evidence for a role of T cells in the etiopathogenesis of peripheral demyelinating diseases. J. Neuroimmunol. 58:37, 1995. 209. Van den Bergh, P. Y., and Pieret, F.: Electrodiagnostic criteria for acute and chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 29:565, 2004.

210. Van den Bergh, P. Y., Thonnard, J. L., Duprez, T., and Laterre, E. C.: Chronic demyelinating hypertrophic brachial plexus neuropathy. Muscle Nerve 23:283, 2000. 211. Van den Berg-Vos, R. M., Van den Berg, L. H., Franssen, H., et al.: Multifocal inflammatory demyelinating neuropathy: a distinct clinical entity? Neurology 54:26, 2000. 212. van der Meche, F. G., and van Doorn, P. A.: The current place of high-dose immunoglobulins in the treatment of neuromuscular disorders. Muscle Nerve 20:136, 1997. 213. van Dijk, G. W., Notermans, N. C., Franssen, H., and Wokke, J. H.: Development of weakness in patients with chronic inflammatory demyelinating polyneuropathy and only sensory symptoms at presentation: a long-term follow-up study. J. Neurol. 246:1134, 1999. 214. van Doorn, P. A., Dippel, D. W. J., van Burken, M. M. G., et al.: Longterm IV immunoglobulin treatment in CIDP. J. Peripher. Nerv. Syst. 8(Suppl. 1):70, 2003. 215. van Doorn, P. A., Brand, A., Strengers, P. F., et al.: High-dose intravenous immunoglobulin treatment in chronic inflammatory demyelinating polyneuropathy: a double-blind, placebocontrolled, crossover study. Neurology 40:209, 1990. 216. van Doorn, P. A., Schreuder, G. M., Vermeulen, M., et al.: HLA antigens in patients with chronic inflammatory demyelinating polyneuropathy. J. Neuroimmunol. 32:133, 1991. 217. van Doorn, P. A., Vermeulen, M., Brand, A., et al.: Intravenous immunoglobulin treatment in patients with chronic inflammatory demyelinating polyneuropathy: clinical and laboratory characteristics associated with improvement. Arch. Neurol. 48:217, 1991. 218. Van Rhijn, I., van den Berg, L. H., Bosboom, W. M., et al.: Expression of accessory molecules for T-cell activation in peripheral nerve of patients with CIDP and vasculitic neuropathy. Brain 123:2020, 2000. 219. Van Schaik, I. N., Winer, J. B., De Haan, R., and Vermeulen, M.: Intravenous immunoglobulin for chronic inflammatory demyelinating polyradiculoneuropathy. Cochrane Database Syst. Rev. 2:CD001797, 2002. 220. Vaughan, R. W., Adam, A. M., Gray, I. A., et al.: Major histocompatibility complex class I and class II polymorphism in chronic idiopathic demyelinating polyradiculoneuropathy. J. Neuroimmunol. 27:149, 1990. 221. Venance, S. L., Koopman, W. J., and Hahn, A. F.: Treatment approaches and long-term outcome in a Canadian cohort (1972–2002) with chronic inflammatory demyelinating polyneuropathy (CIDP). J. Peripher. Nerv. Syst. 8(Suppl. 1):71, 2003. 222. Verma, A.: Epidemiology and clinical features of HIV-1 associated neuropathies. J. Peripher. Nerv. Syst. 6:8, 2001. 223. Verma, A., Tandan, R., Adesina, A. M., et al.: Focal neuropathy preceding chronic inflammatory demyelinating polyradiculoneuropathy by several years. Acta Neurol. Scand. 81:516, 1990. 224. Vermeulen, M., van der Meche, F. G., Speelman, J. D., et al.: Plasma and gamma-globulin infusion in chronic inflammatory polyneuropathy. J. Neurol. Sci. 70:317, 1985. 225. Vermeulen, M., van Doorn, P. A., Brand, A., et al.: Intravenous immunoglobulin treatment in patients with chronic inflammatory demyelinating polyneuropathy: a double blind, placebo controlled study. J. Neurol. Neurosurg. Psychiatry 56:36, 1993.

Chronic Inflammatory Demyelinating Polyradiculoneuropathy 226. Vermeulen, M., and van Oers, M. H.: Successful autologous stem cell transplantation in a patient with chronic inflammatory demyelinating polyneuropathy. J. Neurol. Neurosurg. Psychiatry 72:127, 2002. 227. Vital, A., Vital, C., Brechenmacher, C., et al.: Chronic inflammatory demyelinating polyneuropathy in childhood: ultrastructural features of peripheral nerve biopsies in four cases. Eur. J. Pediatr. 149:654, 1990. 228. Waksman, B. H., and Adams, R. D.: Allergic neuritis: an experimental disease of rabbits induced by the injection of peripheral nervous tissue and adjuvants. J. Exp. Med. 102:213, 1955. 229. Waksman, B. H., and Adams, R. D.: A comparative study of experimental allergic neuritis in the rabbit, guinea pig, and mouse. J. Neuropathol. Exp. Neurol. 15:293, 1956. 230. Watanabe, M., Yamamoto, N., Ohkoshi, N., et al.: Corticosteroid-responsive asymmetric neuropathy with a myelin protein zero gene mutation. Neurology 59:767, 2002. 231. Weiss, M. D., Luciano, C. A., Semino-Mora, C., et al.: Molecular mimicry in chronic inflammatory demyelinating polyneuropathy and melanoma. Neurology 51:1738, 1998. 232. Willison, H. J., and Yuki, N.: Peripheral neuropathies and anti-glycolipid antibodies. Brain 125(Pt. 12):2591, 2002. 233. Wim Ang, C., Jacobs, B. C., and Laman, J. D.: The GuillainBarre syndrome: a true case of molecular mimicry. Trends Immunol. 25:61, 2004. 234. Winer, J., Hughes, S., Cooper, J., et al.: Gamma delta T cells infiltrating sensory nerve biopsies from patients with inflammatory neuropathy. J. Neurol. 249:616, 2002.

2253

235. Wisniewski, H., Prineas, J., and Raine, C. S.: An ultrastructural study of experimental demyelination and remyelination. I. Acute experimental allergic encephalomyelitis in the peripheral nervous system. Lab. Invest. 21:105, 1969. 236. Wulff, E. A., Wang, A. K., and Simpson, D. M.: HIV-associated peripheral neuropathy: epidemiology, pathophysiology and treatment. Drugs 59:1251, 2000. 237. Yako, K., Kusunoki, S., and Kanazawa, I.: Serum antibody against a peripheral nerve myelin ganglioside, LM1, in Guillain-Barre syndrome. J. Neurol. Sci. 168:85, 1999. 238. Yamamoto, M., Ito, Y., Mitsuma, N., et al.: Parallel expression of neurotrophic factors and their receptors in chronic inflammatory demyelinating polyneuropathy. Muscle Nerve 25:601, 2002. 239. Yan, W. X., Archelos, J. J., Hartung, H. P., and Pollard, J. D.: P0 protein is a target antigen in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:286, 2001. 240. Yan, W. X., Taylor, J., Andrias-Kauba, S., and Pollard, J. D.: Passive transfer of demyelination by serum or IgG from chronic inflammatory demyelinating polyneuropathy patients. Ann. Neurol. 47:765, 2000. 241. Yato, M., Ohkoshi, N., Sato, A., et al.: Ataxic form of chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). Eur. J. Neurol. 7:227, 2000. 242. Yuill, G. M., Swinburn, W. R., and Liversedge, L. A.: Treatment of polyneuropathy with azathioprine. Lancet 2:854, 1970. 243. Yuki, N., Tagawa, Y., and Handa, S.: Autoantibodies to peripheral nerve glycosphingolipids SPG, SLPG, and SGPG in Guillain-Barre syndrome and chronic inflammatory demyelinating polyneuropathy. J. Neuroimmunol. 70:1, 1996.

100 Neuropathy Associated with the Monoclonal Gammopathies ROBERT A. KYLE AND PETER J. DYCK

Overview The Monoclonal Gammopathies Recognition of Monoclonal Proteins Analysis of Serum for Protein Analysis of Urine for Protein Monoclonal Gammopathy of Undetermined Significance Characterization of MGUS Association of MGUS with Sensorimotor Peripheral Neuropathy

Etiology Pathology Experimental Approaches IgM Versus IgG and IgA MGUS Neuropathy Clinical Features Differential Diagnosis Electromyographic Studies IgM Monoclonal Proteins and Peripheral Neuropathy

OVERVIEW The presence of a monoclonal protein (immunoglobulin G, A, or M) in the serum of a patient with peripheral neuropathy should raise the suspicion of systemic amyloidosis or POEMS syndrome (polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes) (osteosclerotic myeloma) as well as of multiple myeloma, macroglobulinemia, or lymphoma. If these are excluded by appropriate studies, the patient is classified as having a monoclonal gammopathy of undetermined significance (MGUS; “benign” monoclonal gammopathy) with an associated neuropathy. It must be emphasized that most patients with MGUS do not have clinical evidence of peripheral neuropathy and that the causal relationship of monoclonal protein and the associated neuropathy is unclear. The neuropathy associated with immunoglobulin M (IgM) MGUS is increasingly separated from the neuropathy associated with immunoglobulin G (IgG) or immunoglobulin A (IgA) monoclonal proteins; clinical, pathologic, and electrophysiologic features and response to treatment may distinguish these conditions. There is considerable interest in the role of anti–myelin-associated glycoprotein (MAG) antibodies in the serum and nerves of

IgG and IgA Monoclonal Proteins and Peripheral Neuropathy Treatment Monoclonal Gammopathy and Motor Neuron Diseases Multiple Myeloma Solitary Plasmacytoma Waldenström’s Macroglobulinemia

patients with IgM MGUS, but the role, if any, of these antibodies in the pathogenesis of the neuropathy needs further study. It seems prudent to separate polyneuropathies associated with MGUS from chronic inflammatory demyelinating polyradiculopathy (CIDP). IgG or IgA MGUS neuropathies possibly are the same as CIDP, but because monoclonal proteins are found in the former, separation is recommended. Most commonly, MGUS neuropathies present as a symmetrical sensorimotor polyradiculopathy or neuropathy with nerve conduction and electromyographic features suggesting both segmental demyelination and axonal degeneration. There is an increase in cerebrospinal fluid protein but no increase in cells. Generally, the neuropathy is slowly progressive and often associated with considerable paresthesias and pain, but it seldom leads to complete paralysis, respiratory failure, or death. Treatment for MGUS neuropathy is inadequate. Individual or small groups of patients have shown improvement with chemotherapy or the use of plasma exchange. In a sham-controlled double-blind study, plasma exchange was shown to be efficacious for IgG and IgA MGUS but not for IgM MGUS.32 2255

2256

Diseases of the Peripheral Nervous System

THE MONOCLONAL GAMMOPATHIES The monoclonal gammopathies are a group of disorders in which each monoclonal protein (M protein, myeloma protein, or paraprotein) consists of two heavy polypeptide chains of the same class and subclass and two light polypeptide chains of the same type. The different monoclonal proteins are designated by letters that correspond to the class of their heavy chains, which are designated by Greek letters: ␥ in IgG, ␣ in IgA, ␮ in IgM, ␦ in immunoglobulin D (IgD), and ␧ in immunoglobulin E (IgE). The four subclasses of IgG are designated IgG1, IgG2, IgG3, and IgG4. There are two subclasses of IgA: IgA1 and IgA2. No subclasses of IgM, IgD, or IgE have been recognized. The light-chain types are kappa (␬) and lambda (␭). Each monoclonal protein is produced by a clone of plasma cells in the bone marrow. The cytoplasm of the plasma cells stains with a single class of heavy chain and a single class of light chain. The proliferative process may be low grade, as in MGUS, or malignant, as in multiple myeloma.76

RECOGNITION OF MONOCLONAL PROTEINS Analysis of the serum and urine for monoclonal proteins requires a sensitive, rapid, dependable screening method to detect the presence of a monoclonal protein and a specific assay to identify it according to its heavy-chain class and light-chain type.77 Electrophoresis with agarose gel is recommended for screening.67,157 Immunofixation should be used to confirm the presence of a monoclonal protein and to distinguish the immunoglobulin type and its lightchain class.77

Albumin

γ

FIGURE 100–1 Top, Densitometer tracing of electrophoretic pattern of agarose gel showing a modest-sized spike of ␥ mobility. Bottom, Electrophoresis of serum on agarose gel. Note dense localized band in ␥ region. Anode (albumin) is on the left. See Color Plate

tion of plasma cells. Patients with a polyclonal increase in immunoglobulins have an inflammatory or reactive process.30 In more than 5% of sera, there is an additional monoclonal protein of different immunoglobulin class, and this is designated as biclonal (double) gammopathy.78 It

Analysis of Serum for Protein Serum protein electrophoresis should be done for any patient 40 years or older with a peripheral neuropathy of unknown cause. It should also be performed when multiple myeloma, macroglobulinemia, or amyloidosis is suspected. In the electrophoretic pattern, a monoclonal protein is usually seen as a narrow peak (like a church spire) in the ␥-, ␤-␥-, ␤-, or ␣2-globulin region of the densitometer tracing or as a dense, discrete band on the agarose gel (Fig. 100–1). In contrast, an excess of polyclonal immunoglobulins (having one or more heavy-chain types and both ␬ and ␭ light chains) produces a broadbased peak or broad band. It is usually limited to the ␥ region (Fig. 100–2). One must differentiate between a monoclonal and a polyclonal increase in immunoglobulins because patients with a monoclonal protein have either a malignant process or a potentially malignant prolifera-

FIGURE 100–2 Top, Densitometer tracing of electrophoretic pattern showing a broad-based peak (polyclonal) of ␥ mobility. Bottom, Agarose gel showing a broad band extending throughout the ␥ area. Anode (albumin) is on the left. (From Kyle, R. A., and Rajkumar, S. V.: Monoclonal gammopathies of undetermined significance: a review. Immunol. Rev. 194:112, 2003, with permission of Blackwell Publishing.) See Color Plate

Neuropathy Associated with the Monoclonal Gammopathies

must be emphasized that a patient can have a monoclonal protein when the total protein concentration, ␤- and ␥-globulin levels, and quantitative immunoglobulin values are all within normal limits. A small monoclonal protein may be concealed among the normal ␤ or ␥ components and be overlooked. A monoclonal light chain (Bence Jones proteinemia) is rarely seen in the agarose gel. In several cases of IgD myeloma, the monoclonal protein is very small or not evident at all. In the heavy-chain diseases, the monoclonal component is often not apparent. Thus a normal value for the components of the electrophoretic pattern or a normal-appearing or nonspecific pattern may still contain a monoclonal protein; immunofixation or immunoelectrophoresis is critical. Immunofixation Immunofixation usually is used to identify monoclonal proteins (Fig. 100–3). A monoclonal protein produces a sharp, well-defined band with a single heavy-chain class and light-chain type. A polyclonal increase in immunoglobulins is characterized by broad, diffuse, heavily stained bands with heavy-chain antisera and both ␬ and ␭ antisera. Immunofixation is more sensitive than immunoelectrophoresis when one is searching for a small monoclonal immunoglobulin or light chain in the serum.72 Immunoglobulins may be quantified by radial immunodiffusion, but the technique is tedious and subject to spurious abnormalities and so is not recommended. The rate nephelometer is the preferred method for quantifying immunoglobulins. The degree of turbidity produced by antigen-antibody interaction is measured by nephelometry

SPE

IgG

IgA

IgM

Kappa

Lambda

FIGURE 100–3 Immunofixation of serum with antisera to IgG, IgA, IgM, ␬, and ␭. Antisera showing a discrete, localized band with IgM and ␬ antisera indicating an IgM ␬ monoclonal protein. SPE ⫽ serum protein electrophoresis. Anode (albumin) is on the left. See Color Plate

2257

in the near-ultraviolet region. It is not affected by the molecular size of the antigen, and 7S IgM, polymers of IgA, and aggregates of IgG are measured accurately. However, the levels of IgM measured by nephelometry may be 1000 to 2000 mg/dL higher than those expected on the basis of the serum protein electrophoretic tracing. IgG and IgA values may also be spuriously elevated.124 Serum Viscometry Serum viscometry should be done when the IgM monoclonal protein level is more than 4 g/dL or when the IgA or IgG level is more than 5 g/dL. It should also be performed when a patient has oronasal bleeding, blurred vision, or funduscopy findings suggestive of a hyperviscosity syndrome. Cryoglobulinemia Cryoglobulins are proteins that precipitate when cooled and dissolve when heated. They may be classified as follows: type I (monoclonal—IgG, IgM, IgA, or, rarely, only monoclonal light chains); type II (mixed—two or more immunoglobulins, of which one is monoclonal); and type III (polyclonal—in which no monoclonal immunoglobulin is found). Type I (monoclonal) cryoglobulins are usually of the IgM or IgG classes, but rarely IgA and Bence Jones cryoglobulins have been recognized. Many patients with large amounts of cryoglobulin are completely asymptomatic, whereas others with modest monoclonal cryoglobulinemia (in the range of 1 to 2 g/dL) may have pain, purpura, Raynaud’s phenomenon, cyanosis, or even ulceration and sloughing of the skin and subcutaneous tissues with exposure to the cold. The temperature at which the cryoglobulin precipitates is more important than the amount of monoclonal protein.86 Type II (mixed) cryoglobulinemia most commonly consists of monoclonal IgM and polyclonal IgG, but monoclonal IgG or monoclonal IgA may be seen with polyclonal IgM. Usually the quantity of the cryoglobulin is less than 0.2 g/dL. It may not reach the maximal amount for 7 days at 4° C. The serum protein electrophoretic pattern rarely shows a localized band and usually appears to be normal or like that of a polyclonal hypergammaglobulinemia. Clinically, type II cryoglobulinemia is characterized by purpura and polyarthralgias. Vasculitis involving the skin is common. Raynaud’s phenomenon, necrosis of the skin, and neurologic involvement may occur.100 Renal involvement manifested as glomerulonephritis is often seen.20 Proteinuria with microscopic hematuria is frequent. Renal insufficiency is not common, and the development of end-stage renal failure is infrequent. Rarely, a nephrotic syndrome may be seen. Mixed cryoglobulinemia may be associated with lymphoproliferative disorders and liver dysfunction, and serologic evidence of previous infection with hepatitis B virus

2258

Diseases of the Peripheral Nervous System

may be found.45 Hepatitis C virus is frequently associated with type II cryoglobulinemia.21 Valli et al.148 found peripheral neuropathy in 6 of 23 patients with essential cryoglobulinemia. All but two had mixed cryoglobulinemia (monoclonal IgM ␬ with polyclonal IgG). Eight additional patients had some clinical and neurophysiologic signs of mild peripheral neuropathy. No sera had reactivity with MAG.

Analysis of Urine for Protein Analysis of urine is essential for patients with monoclonal gammopathies. Sulfosalicylic acid or Exton reagent is best for the detection of protein. Dipstick tests are used in many laboratories to screen for protein, but they often do not recognize Bence Jones protein. Screening tests for the detection of Bence Jones proteins (monoclonal light chains) have made use of the unique thermal properties of these proteins. However, both false-positive and falsenegative results occur, and the heat test is not recommended. The recognition of Bence Jones proteinuria depends on the demonstration of a monoclonal light chain by immunofixation of an adequately concentrated urine specimen. Electrophoresis and immunofixation should be performed on the urine of all patients with a serum monoclonal protein concentration greater than 1.5 g/dL. These tests should be done in all instances of multiple myeloma, macroglobulinemia of Waldenström, primary amyloidosis, large MGUS, or suspicion of these entities. Immunofixation

Kappa

Lambda

FIGURE 100–5 Immunofixation of concentrated urine sample with ␬ and ␭ antisera. Note the dense, localized band with ␬ antisera indicating the presence of a monoclonal ␬ light chain (Bence Jones protein). See Color Plate

of urine should also be done in all patients older than 40 years who have a nephrotic syndrome of unknown cause. These tests should also be done on the urine of all patients older than 40 years with a sensorimotor peripheral neuropathy of unknown cause. A urinary monoclonal protein (Bence Jones protein) is seen as a dense, localized band on agarose gel or a tall, narrow, homogeneous peak on the densitometer tracing (Fig. 100–4). Immunofixation establishes the presence or absence of light chains and determines whether they are monoclonal or polyclonal (Fig. 100–5). Immunofixation is most helpful when a monoclonal light chain occurs in the presence of a polyclonal increase in light chains. It is also useful in detecting monoclonal heavy-chain fragments in the urine (Fig. 100–6). The need to collect a 24-hour urine specimen cannot be overemphasized. This allows the amount of monoclonal light chain (Bence Jones protein) excreted in the urine to be measured. This is an excellent indication of the effect of chemotherapy or evidence of progression of the disease. The amount of Bence Jones protein directly reflects the size of the plasma cell burden.

IgA

Kappa Albumin

γ

FIGURE 100–4 Top, Densitometer tracing of electrophoretic pattern of a concentrated urine specimen showing a tall, narrow-based spike of ␥ mobility. Bottom, Agarose gel showing dense band in ␥ region. Anode (albumin) is on the left. (From Kyle, R. A., and Rajkumar, S. V.: Monoclonal gammopathies of undetermined significance: a review. Immunol. Rev. 194:112, 2003, with permission of Blackwell Publishing.) See Color Plate

Lambda

FIGURE 100–6 Immunofixation of a concentrated urine specimen with IgA, ␬, and ␭ antisera. This shows two discrete ␬ bands, one of which corresponds to an IgA band. The findings represent a monoclonal ␬ protein plus an IgA ␬ fragment. See Color Plate

Neuropathy Associated with the Monoclonal Gammopathies

MONOCLONAL GAMMOPATHY OF UNDETERMINED SIGNIFICANCE The term monoclonal gammopathy of undetermined significance denotes the presence of a monoclonal protein in persons without evidence of multiple myeloma, macroglobulinemia, amyloidosis, or other related diseases. The term benign monoclonal gammopathy is misleading because one does not know at the time of diagnosis whether a monoclonal protein will remain stable and benign or will develop into symptomatic multiple myeloma, macroglobulinemia, or amyloidosis. Since Waldenström’s introduction of the term essential hyperglobulinemia in 1952,154 many similar terms have been used, including idiopathic, asymptomatic, benign, and nonmyelomatous monoclonal gammopathy; dysimmunoglobulinemia; idiopathic paraproteinemia; and asymptomatic paraimmunoglobulinemia. Waldenström155 stressed the constancy of the size of the protein spike obtained by electrophoresis of serum, contrasting it with the increasing quantity of the monoclonal protein in myeloma.

Characterization of MGUS MGUS is characterized by a serum monoclonal protein concentration less than 3 g/dL, less than 10% plasma cells in the bone marrow, no or only a small amount of monoclonal protein in the urine, absence of lytic bone lesions, anemia, hypercalcemia, renal insufficiency, and, most important, stability of the monoclonal protein and failure of development of other abnormalities. MGUS accounts for almost two thirds of patients with a serum monoclonal protein. The incidence of MGUS is 3% of patients older than 70 years in Sweden,1,51 the United States,73 and France.127 The incidence of a monoclonal protein in patients 50 years or older in a small Minnesota community was 1.25%, whereas in Finistère, France, 1.7% of patients 50 years or older had a monoclonal protein. In another report, a monoclonal protein was detected in 1.2% of 73,630 patients hospitalized in the United States.152 For both the patient and the physician, it is important to determine whether the monoclonal protein will remain benign or the patient will develop symptomatic multiple myeloma, amyloidosis, macroglobulinemia, or other lymphoproliferative disease. At the Mayo Clinic, we have followed 241 patients who had a monoclonal gammopathy between 1956 and 1970 but no evidence of multiple myeloma, macroglobulinemia, amyloidosis, lymphoma, or related diseases at the time the monoclonal protein was detected. After 24 to 38 years of follow-up, the number of patients whose monoclonal protein had remained stable and who could be classified as having “benign” monoclonal gammopathy had decreased to 30

2259

(12%). The monoclonal protein concentration increased to more than 3 g/dL in 23 patients (10%), but they did not develop symptomatic multiple myeloma, macroglobulinemia, or other disease and have not been treated. During follow-up, 126 patients (52%) have died, but none had multiple myeloma or macroglobulinemia that required therapy. Multiple myeloma, macroglobulinemia, amyloidosis, or a malignant lymphoproliferative process developed in 62 patients (actuarial rate was 16% at 10 years and 33% at 20 years). Of these 62 patients, 42 (68%) had multiple myeloma. The interval from the time of recognition of the monoclonal gammopathy to the diagnosis of multiple myeloma ranged from 23 to 251 months (median, 115 months [9.6 years]). Macroglobulinemia of Waldenström developed in seven patients. The median interval from the recognition of the monoclonal protein to the diagnosis of macroglobulinemia was 8.5 years. Systemic amyloidosis (amyloid light chain [AL] type) was found in eight patients 6 to 16.5 years after the recognition of the monoclonal protein (median, 8 years). Two patients developed an aggressive malignant lymphoma, and another patient had chronic lymphocytic leukemia. To confirm the studies of the original Mayo Clinic series, long-term follow-up was obtained for 1384 patients with MGUS who resided in southeastern Minnesota from 1960 to 1994. They were followed for a total of 11,009 person-years, during which 115 patients (8%) developed multiple myeloma, primary amyloidosis, lymphoma with an IgM serum monoclonal protein, macroglobulinemia, plasmacytoma, or chronic lymphocytic leukemia (Table 100–1). The cumulative probability of progression to one of these disorders was 10% at 10 years, 21% at 20 years, and 26% at 25 years, with a risk of progression of about 1% per year (Fig. 100–7). The number of patients with progression to a serious plasma cell disorder (115 patients) was 7.3 times the number expected on the basis of incidence rates for those conditions in the general population. The relative risk for development of multiple myeloma was increased 25-fold, macroglobulinemia 46-fold, and primary amyloidosis 8.4-fold. The risk for development of lymphoma was only moderately increased at 2.4-fold, but this risk is underestimated because only lymphoma associated with an IgM monoclonal protein was included in the observed number, whereas the incidence rates for lymphomas associated with IgG, IgA, and IgM monoclonal proteins were used to calculate the expected number. The 75 patients in whom multiple myeloma developed accounted for 65% of the 115 patients who had progression to a plasma cell disorder. In one third of patients, multiple myeloma was diagnosed more than 10 years after the monoclonal protein was initially detected, and in five patients (7%) it was diagnosed after 20 years of follow-up.

2260

Diseases of the Peripheral Nervous System

Table 100–1. Risk of Progression among 1384 Patients with Monoclonal Gammopathy of Undetermined Significance* No. of Patients Type of Progression

Observed

Expected†

75 19‡ 10 7 3§ 1 115

3.0 7.8 1.2 0.2 3.5 0.1 15.8

Multiple myeloma Lymphoma Primary amyloidosis Macroglobulinemia Chronic lymphocytic leukemia Plasmacytoma Total

Relative Risk (95% CI) 25.0 (20–32) 2.4 (2–4) 8.4 (4–16) 46.0 (19–95) 0.9 (0.2–3) 8.5 (0.2–47) 7.3 (6–9)

CI ⫽ confidence interval. * Mayo Clinic patients who were residents of southeastern Minnesota, diagnosis from 1960 to 1994. † Expected numbers of cases were derived from the age- and sex-matched white population in Iowa from Surveillance, Epidemiology, and End Results (SEER) Program: Public-Use Data (1973–1998). Bethesda, Maryland, National Cancer Institute, Cancer Statistics Branch, April 2001, except for primary amyloidosis, for which data are from Kyle, R. A., Linos, A., Beard, C. M., et al.: Incidence and natural history of primary systemic amyloidosis in Olmsted County, Minnesota, 1950 through 1989. Blood 79:1817, 1992. ‡ All 19 patients had serum IgM monoclonal protein. If the 30 patients with IgM, IgA, or IgG monoclonal protein and lymphoma were included, the relative risk would be 3.9 (95% CI, 2.6 to 5.5). § All three patients had serum IgM monoclonal protein. If all six patients with IgM, IgA, or IgG monoclonal protein and chronic lymphocytic leukemia were included, the relative risk would be 1.7 (95% CI, 0.6 to 3.7). From Kyle, R. A., Therneau, T. M., Rajkumar, S. V., et al.: A long-term study of prognosis in monoclonal gammopathy of undetermined significance. N. Engl. J. Med. 346:564, 2002, with permission. Copyright © 2002 Massachusetts Medical Society. All rights reserved.

25%

Cumulative probability (%)

30

30% Progression or increase

26%

20 12% 10

21% Progression only

10% 0 0

5

10

15

20

25

56

17

Years of follow-up No. at 1,384 risk

867

423

177

FIGURE 100–7 Probability of progression among 1384 residents of southeastern Minnesota in whom monoclonal gammopathy of undetermined significance was diagnosed from 1960 through 1994. The top curve shows the probability of progression to a plasma cell cancer (115 patients) or an increase in the monoclonal protein concentration to more than 3 g/dL or the proportion of plasma cells in bone marrow to more than 10% (32 patients). The bottom curve shows only the probability of progression of monoclonal gammopathy of undetermined significance to multiple myeloma, IgM lymphoma, primary amyloidosis, macroglobulinemia, chronic lymphocytic leukemia, or plasmacytoma (115 patients). The bars indicate 95% confidence intervals. (From Kyle, R. A., Therneau, T. M., Rajkumar, S. V., et al.: A long-term study of prognosis in monoclonal gammopathy of undetermined significance. N. Engl. J. Med. 346:564, 2002, with permission. Copyright © 2002 Massachusetts Medical Society. All rights reserved.)

The clinical and laboratory characteristics of these 75 patients were comparable to those of the usual Mayo Clinic referral patient with multiple myeloma except that the southeastern Minnesota patients were older (76 compared with 66 years) and less likely to be male (46% vs. 60%).79 The baseline variables evaluated with respect to predicting progression of MGUS included age; sex; hepatosplenomegaly; levels of hemoglobin, serum creatinine, and serum albumin; size of the serum monoclonal protein; type of serum monoclonal protein (IgG, IgA, IgM); presence, type, and amount of monoclonal urinary light chain; number of bone marrow plasma cells; and decrease in uninvolved immunoglobulins. Only the concentration and type of monoclonal protein were independent predictors of progression. The presence of a monoclonal urine protein (␬ or ␭) and a decrease in one or more uninvolved immunoglobulins were not risk factors for progression. Patients with IgM or IgA monoclonal protein had an increased risk of progression compared with patients who had IgG monoclonal proteins (P ⫽ 0.001). However, the initial concentration of the serum monoclonal protein was the most important risk factor for progression to a plasma cell disorder. The relative risk of progression was related directly to the concentration of the monoclonal protein in the serum at the time of diagnosis of MGUS. The risk of progression to multiple myeloma or a related disorder 20 years after diagnosis of MGUS was 15% for an initial monoclonal protein level of 0.5 g/dL or less, 16% for 1.0 g/dL, 25% for 1.5 g/dL, 41% for 2.0 g/dL, 49% for 2.5 g/dL, and 64% for 3.0 g/dL (Fig. 100–8). The

2261

9 7

64 48.8

5

65 50

41.2 3 24.6

25

15.6 1

risk of progression with a monoclonal protein level of 1.5 g/dL was almost twofold more than the risk of progression with a level of 0.5 g/dL, whereas the risk of progression with a level of 2.5 g/dL was 4.6 times more than with 0.5 g/dL. The number of plasma cells in the bone marrow also may be helpful in predicting progression. Cesana et al.15 reported that more than 5% bone marrow plasma cells was an independent risk factor for progression. Baldini et al.4 reported that the malignant transformation rate at 4 to 6 years was 6.8% when the bone marrow plasma cell level was less than 10%, compared with 37% for the group that met the criteria of MGUS but had a bone marrow plasma cell value of 10% to 30%. Patients with an apparently benign monoclonal gammopathy must be followed indefinitely because serious disease may develop 20 or more years after the recognition of the monoclonal protein. Although monoclonal gammopathy frequently exists without other abnormalities, certain diseases are associated with it, as would be expected in an older population. The association of two diseases depends on the frequency with which each occurs independently. Furthermore, there may be an association because of differences in the referral practice or in other selected groups of patients. The investigator must use valid epidemiologic and statistical methods in evaluating these associations. The need for appropriate control populations cannot be overemphasized. Among the many associations that have been reported, the following are the best established: lymphoproliferative disorders (especially non-Hodgkin’s lymphoma); hematologic disorders (including myelodysplasia, chronic neutrophilic leukemia, acquired von Willebrand’s disease, and Castleman’s disease); neurologic disorders (discussed below); dermatologic diseases (lichen myxedematosus, pyoderma gangrenosum, subcorneal pustular dermatosis, necrobiotic xanthogranuloma, and cutaneous lymphomas); immunosuppression associated with kidney, liver, bone marrow, or heart transplantation or acquired immunodeficiency syndrome; connective tissue disorders such as

13.6

13.6

0.5

1.0

1.5

2.0

Probability of full progression at 20 years (%)

FIGURE 100–8 Monoclonal gammopathy of undetermined significance (MGUS) in residents of southeastern Minnesota (1960 to 1994). The relative risk of full progression was related directly to the concentration of serum monoclonal protein at the time of diagnosis of MGUS. Error bars indicate 95% confidence intervals. (Modified from Kyle, R. A., and Dispenzieri, A.: Monoclonal gammopathy of undetermined significance. Rev. Clin. Exp. Hematol. 6:225, 2002, with permission of Blackwell Publishing.)

Relative risk of full progression

Neuropathy Associated with the Monoclonal Gammopathies

2.5

Serum M spike value

rheumatoid arthritis; and various unrelated conditions such as acquired deficiency of C1 esterase inhibitor and capillary leak syndrome.

Association of MGUS with Sensorimotor Peripheral Neuropathy In 1973, Forssman et al.39 described a patient with an IgM benign monoclonal gammopathy and peripheral neuropathy. The nerve was infiltrated with IgM-producing lymphocytes as well as IgM. The patient improved with chlorambucil therapy. Using agarose gel electrophoresis, Kahn et al.62 detected 58 cases (0.4%) of monoclonal gammopathy without evidence of multiple myeloma or Waldenström’s macroglobulinemia in 14,000 serum samples obtained from patients admitted to a large neurologic referral hospital. Two patients had a solitary plasmacytoma. Sixteen of the remaining 56 patients had peripheral neuropathy. Nine of these 16 patients had an idiopathic peripheral neuropathy with slow nerve conduction velocities, and the other 7 patients had less uniform electrophysiologic findings and other conditions known to be associated with peripheral neuropathy. Seven of the nine patients had IgM, and all had ␬ light-chain type. In a series of 21 consecutive patients with benign monoclonal gammopathy, 11 had mild sensory or motor symptoms of peripheral neuropathy and 4 others had electromyographic or electroneurographic findings. The distribution of monoclonal proteins was IgG in 16 patients, IgM in 3, and IgA in 2.112 Johansen and Leegaard60 found that 4 of 132 consecutive patients with a monoclonal gammopathy had peripheral neuropathy. Three of the four patients had an IgM monoclonal protein, and all four were of the ␬ light-chain class. During a 1-year period, 692 patients with a clinically apparent neuropathy were identified in the Electromyography Laboratory at the Mayo Clinic. Of these, 358 patients with neuropathy associated with a systemic disease known to

2262

Diseases of the Peripheral Nervous System

cause neuropathy, such as diabetes mellitus, alcoholism, or connective tissue disease, were eliminated. In the remaining 334 patients, no associated systemic condition was apparent. Of these patients, 279 had serum protein electrophoresis and 28 (10%) had a monoclonal protein. Sixteen (5.7%) of these cases were due to MGUS, and primary amyloidosis and multiple myeloma accounted for seven and three cases, respectively. One patient had Waldenström’s macroglobulinemia, and the remaining patient had ␥ heavy-chain disease. When compared with normal populations in Minnesota and Sweden, this represented a significant increase in the frequency of neuropathy among patients with monoclonal protein. However, because referral bias to the Mayo Clinic and to electromyographic examination could be involved, we do not consider the conclusions to have been proven. In another series of 239 patients with peripheral neuropathy or myopathy, 5.4% had a monoclonal protein.58 A total of 192 patients were registered with a CIDP study in Vest-Agder, Norway. Of these 192 patients, 8 had a monoclonal gammopathy and peripheral neuropathy without evidence of other disorders. The polyneuropathy was mainly sensory in five patients and sensorimotor in the others.101 Among 21 patients with MGUS, Krol-v Straaten et al.68 found only 1 patient with clinically distinctive symptoms and signs of polyneuropathy of the lower extremities. In another report, 8 of 19 patients with MGUS had a clinical polyneuropathy.153 In an unselected series of patients with MGUS, neuropathy was noted in 2 of 34 patients with IgG (6%), 2 of 14 with IgA (15%), and 8 of 26 with IgM (31%); the neuropathy was subclinical in 6 patients.105 Baldini et al.5 reported that 14 of 31 patients with an IgM MGUS had a peripheral neuropathy. In a recent report of 104 patients with polyneuropathy and monoclonal gammopathy, 23 patients had a hematologic malignancy, 6 had stage I multiple myeloma, 2 had stage III myeloma, 10 had low-grade lymphoma, 3 had a plasmacytoma, 1 had Castleman’s disease, and 1 had POEMS syndrome.36 Therefore, it appears that there is an association between MGUS and peripheral neuropathy. The incidence is variable and depends on patient selection, the vigor with which the presence of a monoclonal protein is sought, and whether peripheral neuropathy is diagnosed on clinical or electrophysiologic grounds. A monograph on neurologic disorders associated with plasma cell dyscrasias has been published,65 and neuropathy and monoclonal gammopathy have been reviewed recently.106,150

Etiology The cause of peripheral neuropathy with monoclonal gammopathy is unknown. The role of monoclonal IgM, IgG, and IgA is discussed below. A genetic effect may play a role in some cases. Jønsson et al.61 reported a demyelinating peripheral neuropathy in

a brother and sister with an IgM monoclonal protein. A mother and son were reported with an IgM protein and peripheral neuropathy. The mother had an IgM value of 490 mg/dL and a discrete band in the slow ␥ area. The son had an IgM monoclonal protein in the serum, but five of his children had no abnormality on immunoelectrophoresis.13 Manschot et al.90 reported three families with neuropathy associated with a monoclonal protein. In one family, the daughter had MGUS and her mother had Waldenström’s macroglobulinemia. In a second family, two brothers had MGUS, and in the third family, one brother had MGUS and the other had chronic lymphocytic leukemia. All six patients had an IgM monoclonal protein.

Pathology The IgM monoclonal protein with anti-MAG activity associated with peripheral neuropathy reacts to or binds to the myelin sheaths.27,151 In one patient, deposits of both ␮ heavy chains and ␭ light chains were found in the myelin sheath.145 Harbs et al.52 reported that sera from patients with macroglobulinemia reacted with peripheral nerve structures, mainly at the border of the myelin sheath as well as between the myelin and axon. However, there was no difference between the sera of patients with and without peripheral neuropathy. Loss of fibers, wallerian-like degeneration, and segmental demyelination were found in the peroneal superficial nerve of a patient with an IgM ␬ monoclonal gammopathy and peripheral neuropathy. Immunoperoxidase studies revealed binding of the IgM ␬ protein on the myelin sheath.102 Direct electron microscopic immunochemistry studies with colloidal gold showed deposition of IgM within the myelin and extending throughout the compact myelin in both large and small myelinated fibrils. This finding also suggests that the deposition of IgM monoclonal protein has a role in the pathogenesis of neuropathy.92 However, it remains unproved whether IgM protein initiates the myelin damage or simply precipitates in an already damaged nerve. Single-fiber preparations showed myelin swelling of internodes in a patient with IgM monoclonal protein reacting with MAG. These findings are similar to those of hereditary tomaculous neuropathy. The co-localization of myelin thickening and widening of myelin lamellae, which are typical of IgM neuropathies, is suggestive of a link between the two abnormalities.123 IgM and F(ab⬘)2 react with the Pi granules in the Schwann cells of patients with monoclonal gammopathy and peripheral neuropathy.85 In five patients with an IgM monoclonal protein and peripheral neuropathy, the IgM was bound to MAG in three patients. Two others had IgM activity against the endoneurium.61 Endothelial proliferation producing thickening of the vessel walls was found in the vasa nervorum in nine patients with a monoclonal gammopathy and peripheral

Neuropathy Associated with the Monoclonal Gammopathies

neuropathy. These microvascular changes could conceivably produce ischemia and result in the loss of axons. It has been postulated that alteration of vascular permeability may admit some proteins into the endoneurium, where they may directly affect the myelin sheath and produce demyelination.119 Surprisingly, labeling of the white matter within the cerebral and cerebellar cortex has been reported in patients with chronic demyelinating neuropathy and an IgM monoclonal protein with MAG activity.94 Is the neuropathy in MGUS the result of axonal degeneration or a demyelinating process? Many investigators studying immunologic localization in myelin have assumed it to be a demyelinating disorder.104 Favoring this idea is the dispersion of the compound muscle action potential found in a percentage of cases with MGUS neuropathy and its significantly higher frequency in IgM MGUS than in the other two varieties.47 However, maximal conduction velocity is only 10% to 20% below normal ranges, and there is evidence of abundant axonal degeneration in teased fibers of sural nerve and fibrillation and altered motor unit potentials in electromyographic examination. Some of the demyelination in teased fibers may be secondary to axonal events. A final answer as to the primary feature of the disease process awaits further study.

Experimental Approaches Intraperitoneal injection of mineral oil or pristane produced monoclonal immunoglobulin-producing tumors in 32 BALB/c mice. Morphologic and histometric studies of the sciatic nerve showed that 11 mice developed neuropathy. There was no significant amount of immunoglobulin or light chain in the endoneurium.103 Nerve lesions consisting of local loss of myelin sheaths and in some instances complete nerve fiber degeneration of the wallerian type were seen in mice with MOPC 47B, MOPC 104E, MOPC 315, and APC-5 tumors.25 Mice injected daily with purified monoclonal IgG from patients with polyneuropathy associated with myeloma or benign monoclonal gammopathy developed a demyelinating polyneuropathy with slowed nerve conduction velocities. Injection of Fab fragments from monoclonal IgG produced a similar demyelination.6 However, Bosch et al.10 injected serum from three patients with IgG ␬, IgM ␬, and IgA ␭ monoclonal proteins and chronic sensorimotor polyneuropathy intraneurally into rat sciatic nerves. None of the animals developed weakness, slowing of in vitro conduction of the sciatic nerve, or significant evidence of demyelination by light or electron microscopy or studies of teased single fibers. Furthermore, the serum from two patients with amyotrophic lateral sclerosis and chronic spinal muscular dystrophy with a monoclonal IgG ␬ did not inhibit terminal sprouting from motor nerve terminals in mouse muscles paralyzed by botulinum toxin. The mono-

2263

clonal proteins did not recognize a growth factor postulated to be implicated in the pathogenesis of motor neuron disease.31 Conversely, Lawlor et al.84 reported that intraneural injection of serum from a patient with MGUS and sensorimotor peripheral neuropathy produced myelin and axonal degeneration. Kvarnström et al.71 established cell lines from the peripheral blood of patients with MGUS and peripheral neuropathy. Two clones produced anti-P0 antibodies of the IgM ␭ type. The IgM protein reacted with P0 and MAG. Five of 7 patients with an IgG monoclonal gammopathy and 1 of 18 with an IgM monoclonal gammopathy and peripheral neuropathy had increased T-cell densities compared with controls. Patients with increased sural nerve T cells had a more progressive course.37 An IgM monoclonal protein that reacted with MAG in three patients with peripheral neuropathy produced demyelination of feline sciatic nerve following intraneural injection. Demyelination was greater than that seen in six patients with an IgM monoclonal protein unreactive with MAG. The myelinolytic effect required active human complement. The effect was abolished by removal of 90% of the monoclonal protein. Immunofluorescence studies demonstrated deposition of the injected monoclonal protein and complement on the surface of myelin sheaths. This suggests that the monoclonal protein reacted with an epitope of myelin exposed to the extracellular space.53

IgM Versus IgG and IgA MGUS Neuropathy We have reviewed our experience with 65 patients with MGUS and a sensorimotor peripheral neuropathy evaluated at the Mayo Clinic from 1980 to 1986. Patients not adequately evaluated, those with focal neuropathy, and those with neuropathies of other causes were excluded. Thirty-one patients had IgM, 24 had IgG, and 10 had IgA monoclonal proteins. Four features distinguished IgM from IgG and IgA neuropathies: (1) sensory ataxia occurred significantly more frequently in IgM, (2) 10 attributes of nerve conduction were significantly worse and none were better, (3) dispersion of the compound muscle action potential occurred significantly more frequently, and (4) IgM was overrepresented in the neuropathy group as compared with the group who had MGUS without neuropathy. Read et al.122 described three patients and Bleasel et al.7 reported on five patients who had an IgG monoclonal protein and peripheral neuropathy. In a comparison of 19 patients with IgM monoclonal gammopathy and 15 with IgG monoclonal gammopathy, Simovic et al.140 reported prolongation of distal latencies of the median and ulnar motor nerves, greater slowing of peroneal nerve conduction velocity, and more severe demyelination in IgM MGUS. IgA monoclonal gammopathy has been reported in several patients with peripheral neuropathy.3,29,139

2264

Diseases of the Peripheral Nervous System

In a review of 40 patients with polyneuropathy associated with IgM gammopathy, all but 1 had symmetrical polyneuropathy. It was purely sensory in 17 and predominantly sensory in 13. Demyelination was noted in the electrophysiologic studies in 83% of patients. Two thirds of the patients had anti-MAG antibodies, and they were associated only with demyelinating neuropathy.16 In a series of 52 symptomatic patients with IgM monoclonal gammopathy, symptomatic neuropathy occurred in 3 of the 4 patients with a high anti-MAG titer and in only 3 of the 21 with a low anti-MAG titer.97 In a series of 65 patients with polyneuropathy and IgM monoclonal gammopathy, multivariate analysis demonstrated that only the initial findings on electrophysiologic studies were independent prognostic factors. The presence of anti-MAG or anti–sulfoglucuronyl paragloboside specificity did not contribute to predicting outcome.34 As described below, pathologic and immunologic differences also appear to separate IgM from IgG or IgA neuropathies. In addition, IgG and IgA neuropathies have been responsive to plasma exchange, whereas IgM neuropathies appear to be less responsive. In a report of 39 patients with MGUS, it was suggested that those with neuropathy associated with IgM were different and should be distinguished from those with IgG or IgA neuropathy.146

Clinical Features Typically, MGUS neuropathy has the following features: (1) it is associated with the occurrence of a monoclonal protein in the serum; (2) amyloidosis, osteosclerotic myeloma (POEMS syndrome), or related disorders are absent; (3) a symmetrical sensorimotor polyradiculopathy or neuropathy that begins insidiously usually is slowly progressive over months or years; (4) it occurs especially in the fifth to seventh decades of life; (5) it affects males more frequently than females; (6) paresthesias, ataxia, and pain may be prominent; and (7) cranial nerves are not affected. The association of a monoclonal protein without myeloma or amyloidosis is a critical feature. In our series of 65 patients, the average age at initial symptoms was 55 years, but it ranged from 23 to 76 years. Men (68%) were involved more frequently than women. This predilection for males is not readily explained by referral bias because the percentage of male patients at our institution is slightly lower (47%) than that of female patients. In 75% of our patients, decreased or altered feeling, paresthesia, unsteadiness of gait, or pain was present initially. Later (usually months or years), symptomatic muscle weakness occurred. Symptoms began in the toes, feet, or legs, and although they might begin in one foot, they usually became symmetrical. On examination 2.5 years after the first symptoms, the distribution was confined to the lower limbs in 74% of patients, to the upper limbs in 13%, and to both lower and

upper limbs in 13%. In most, sensory impairment and muscle weakness could be demonstrated. Characteristically, muscle stretch reflexes were diminished or absent. Babinski signs were not found. The course was one of chronic progression in 71% and a stepwise progression or recurring symptoms in the others. Death was not attributed to the neuropathy in our series. The brachial plexus was studied with magnetic resonance imaging (MRI) in 21 patients with polyneuropathy associated with IgM monoclonal gammopathy and in 9 with IgG gammopathy.35 Of the 24 patients with demyelinating polyneuropathy, 23 had abnormal MRI findings. The authors concluded that demyelinating polyneuropathy associated with monoclonal gammopathy and distal symptoms and signs is more generalized than presumed. The protein content of the cerebrospinal fluid was approximately 100 mg/dL (range, 16 to 580 mg/dL). In about 85% of patients, the protein value was elevated above the normal range.

Differential Diagnosis MGUS neuropathy must be distinguished from CIDP, amyloidosis, myeloma, and paraneoplastic neuropathies as well as metabolic and toxic neuropathies. CIDP may occur at any age from infancy through old age. Motor symptoms tend to predominate over sensory symptoms, and there is a greater tendency for the course to be relapsing. Monoclonal proteins are not found. IgG and IgA MGUS neuropathies are similar in most respects, and they may actually be the same disorder. IgM MGUS neuropathies differ in several respects, as listed above. In a comparison of 77 patients with CIDP and 26 with CIDP and MGUS, the latter group had a more indolent course, less severe weakness, and more frequent sensory loss. The outcome of the two groups was similar.137 In a comparison of 45 patients with CIDP and 15 with MGUSassociated neuropathy, Gorson et al.46 reported that the latter group had less severe weakness, more ataxia, vibration loss in the hands, and absent median and ulnar sensory potentials. In another report, the clinical course was progressive in most patients with MGUS, whereas those with CIDP were more likely to have a relapsing course. Impairment appeared to develop more slowly in the patients with MGUS, and they also had less severe functional impairment and a lesser degree of weakness and sensory changes.138 In contrast, Maisonobe et al.89 noted no significant clinical or electrophysiologic difference between patients with CIDP and those with CIDP and MGUS. They noted that patients with anti-MAG antibodies had more pronounced slowing of the peroneal motor nerve conduction velocity, a lower frequency of conduction block, and distal accentuation of conduction slowing. Another report determined that the latency index was lower in MAG-associated neuropathy than in CIDP.17 In

2265

Neuropathy Associated with the Monoclonal Gammopathies

summary, CIDP may occur at any age, motor symptoms seem to predominate over sensory symptoms, and there is a greater tendency for the course to be relapsing. Monoclonal proteins, of course, are not found. The characteristics of peripheral neuropathy associated with MGUS, POEMS syndrome, multiple myeloma, and AL amyloidosis are shown in Table 100–2. MGUS neuropathies are similar to those in amyloidosis but differ in the following respects: (1) in MGUS the lower limbs are preferentially affected, but the upper and lower limbs tend to be affected to a greater degree in amyloidosis; (2) the course in amyloidosis is always slowly progressive; (3) although amyloidosis also may present as a sensorimotor neuropathy, autonomic features (postural hypotension, sphincter dysfunction, and anhydrosis) and organ (heart or kidney) failure are often seen, but these do not occur in MGUS neuropathy; and (4) amyloid can be recognized in biopsy tissue. Paraneoplastic or myeloma neuropathies may have many features overlapping those of MGUS neuropathies. Detection of the monoclonal protein or tumor is needed to differentiate them. In hypothyroidism or in persons exposed to certain industrial or medicinal toxins, a sensorimotor neuropathy may develop with features similar to those in MGUS neuropathy. Although it is doubtful that any basic difference exists in the pathology, pathophysiology, and treatment of monoclonal gammopathy associated with IgM or IgG and IgA, we discuss them separately below because most authors

have reported their findings on the basis of the type of monoclonal protein.

Electromyographic Studies Nerve conduction is characteristically abnormal in motor and sensory fibers and in the upper and lower extremities. Conduction velocity of motor fibers is decreased below the normal range by approximately 20%.47 Smith et al.141 also emphasized the reduction of motor nerve conduction velocity in 12 patients with neuropathy and an IgM protein reacting with human peripheral myelin. The amplitude of the compound muscle action potential is severely decreased in the lower limb nerves. Frequently F-wave latencies are prolonged, but probably not disproportionately to the degree of nerve conduction slowing in these patients. Frequently responses from the sensory fibers of the upper and lower limb nerves cannot be elicited. Of 35 patients with MGUS-associated neuropathy, 24 had asymmetrical electrodiagnostic findings, in contrast to only 4 of 29 patients with diabetes-associated peripheral neuropathy.159 Needle electromyography demonstrates denervation (increased insertional activity with fibrillation) in 80% of patients. Both demyelination and denervation may be present.64 The sural nerve is involved more severely than the median nerve. The myelin sheath or axons may be involved in monoclonal gammopathy and peripheral neuropathy.44 Separation of myelin lamellae has been reported as a characteristic feature of MGUS-associated

Table 100–2. Clinical and Laboratory Features of Plasma Cell Dyscrasia Associated with Peripheral Neuropathy Characteristic

MGUS

POEMS

MM

AL Amyloidosis

Peripheral neuropathy Sensory vs. motor predominance Autonomic involvement Organomegaly Skin involvement Other symptoms

~5% Mixed sensorimotor Minimal • • Asymptomatic

1%–8% Mixed sensorimotor Minimal ⫹ ⫹ Bone pain, fatigue, infections

Monoclonal heavy chain Monoclonal light chain Serum M-protein spike (g/dL) BM plasma cells Skeletal lesions

IgM ⬎ IgG ⬎ IgA ␬ 67% of cases ⬍3 ⬍10% •

15%–20% Mixed sensorimotor ⫹ to ⫹⫹ ⫹⫹ ⫹ Fatigue, edema, cardiomyopathy, nephrosis IgG ⬎ IgA ⬎ IgM ␭ 66% of cases Usually ⬍ 2 Usually ⬍ 10% •

Thrombocytosis Anemia

• •

100% Motor and sensory ⫹ ⫹⫹ ⫹⫹ Edema, fatigue, endocrine abnormalities IgG ⬎ IgA ⬎ IgM ␭ ⬎95% of cases Usually ⬍ 2 Usually ⬍ 5% ⫹⫹⫹ (sclerotic, mixed sclerotic, lytic) ⫹⫹ ⫹

IgG ⬎ IgA ␬ 66% of cases Usually ⬎ 3 ⬎ 10% ⫹⫹⫹ (lytic, osteoporotic, or fracture) • ⫹⫹

⫹ to ⫹⫹ ⫹

BM ⫽ bone marrow; M ⫽ monoclonal; MGUS ⫽ monoclonal gammopathy of undetermined significance; MM ⫽ multiple myeloma; POEMS ⫽ polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes. •, absent; ⫹, rare; ⫹⫹, occurs frequently; ⫹⫹⫹, almost always present. From Kyle, R. A., and Dispenzieri, A.: Neurologic aspects of monoclonal gammopathy of undetermined significance, multiple myeloma, and related disorders. In Gahrton, G., Durie, B., and Samson, D. (eds.): Multiple Myeloma. London, Edward Arnold Publishers, (in press), with permission of Edward Arnold Publishers.

2266

Diseases of the Peripheral Nervous System

FIGURE 100–9 Myelinated fiber showing different stages of splitting of myelin lamellae. Normal compact myelin is present only in the inner half of the sheath. Note irregularities of outer and inner lips of enveloping Schwann cell (straight black arrows). Curved arrows point to honeycomb-like inclusions derived from myelin breakdown. Accumulation of proteinaceous material and membranous debris in periaxonal location (lower right corner) and between dissociated lamellae of myelin (white arrow) (⫻65,000). Inset: Localization of ␬ light chain at the periphery of myelinated fibers (approximately ⫻200). (From Lach, B., Rippstein, P., Atack, D., et al.: Immunoelectron microscopic localization of monoclonal IgM antibodies in gammopathy associated with peripheral demyelinating neuropathy. Acta Neuropathol. 85:298, 1993, with permission of Springer-Verlag.)

polyneuropathy. In our experience, it is not found in all cases and may be found in patients without MGUSassociated neuropathy. A striking example is shown in Figure 100–9.

IgM Monoclonal Proteins and Peripheral Neuropathy Myelin-Associated Glycoprotein In 1980, Latov et al.83 described a patient with peripheral neuropathy and an IgM monoclonal protein. The monoclonal protein was directed against peripheral nerve myelin, as shown by complement fixation and immunoabsorption. MAG is found in both control and peripheral nerve myelin. It has a molecular weight of approximately 100,000 and consists of 30% carbohydrate. MAG shares an antigenic determinant (HNK-1 or Leu-7) with natural killer cells. The monoclonal IgM protein binds to a carbohydrate moiety in MAG.108 Bollensen et al.8 reported that only sera with reactivity with MAG also reacted with the major glycoprotein P0 in human peripheral nerves. IgM ␬ deposition has been noted in areas of splitting of the

myelin lamellae and bound to the internalized myelin debris80 (Fig. 100–10). Approximately one half of patients with peripheral neuropathy and IgM monoclonal gammopathy have IgM antibodies that bind to MAG.82,88,113,129 Binding of the IgM protein to MAG is apparently specific because it can be completely inhibited by MAG isolated from human myelin.144 The anti-MAG IgM titers in 16 patients with peripheral neuropathy and an IgM monoclonal gammopathy reacting with MAG ranged from 1:12,800 to 1:100,000. Low antibody titers to MAG (1:400 or less) were found in 8 of 24 patients with an IgM monoclonal protein without neuropathy. Very low titers (1:200 or less) were found in 17 of 100 control patients without a monoclonal gammopathy. In this study, high titers of anti-MAG IgM antibodies were always associated with neuropathy.107 It has been demonstrated conclusively that the serum immunoglobulins that bind MAG share idiotypic determinants with the patient’s monoclonal protein. This is strong evidence that it is the monoclonal protein that reacts with MAG.126 Brouet et al.11 showed that 9 of 10 IgM

Neuropathy Associated with the Monoclonal Gammopathies

FIGURE 100–10 Immunoreaction for ␬ light chain on light myelin, some cytoplasmic vesicles (arrow), and multivesicular honeycomb-like inclusions (upper left corner) (⫻57,600). (From Lach, B., Rippstein, P., Atack, D., et al.: Immunoelectron microscopic localization of monoclonal IgM antibodies in gammopathy associated with peripheral demyelinating neuropathy. Acta Neuropathol. 85:298, 1993, with permission of Springer-Verlag.)

monoclonal proteins with anti-MAG activity expressed a public cross-reactive idiotype. This shared idiotype is very likely involved in (or is very close to) the combining site of IgM because anti-idiotypic serum inhibited the binding of IgM to MAG. The monoclonal proteins do not bind to deglycosylated MAG, which supports the assumption that the reactive determinants contain carbohydrate moieties.134 Frail et al.40 also demonstrated that deglycosylation of MAG abolished the recognition of MAG by the patient’s IgM monoclonal protein. These data support the fact that the peripheral myelin sheath is involved in the autoimmune response directed against MAG. Complement may play a role in demyelination in patients with IgM monoclonal gammopathy and peripheral neuropathy. Six patients with peripheral neuropathy and an IgM monoclonal protein reacting with MAG had complement components C1q, C3d, and C5 along the myelin

2267

sheaths of peripheral nerve fibers. The authors concluded that demyelination may be mediated by complement.99 In one report, four patients with an IgM monoclonal protein reactive with MAG had predominantly motor symptoms and demyelination in the nerve biopsy specimen, whereas four other patients with IgM monoclonal protein but no anti-MAG activity had mainly sensory symptoms and axonal neuropathy.142 Steck et al.143 reported that demyelination was greater with an IgM protein reacting to MAG than that induced by sera with a monoclonal protein unreactive to MAG. The sera of patients with osteosclerotic myeloma and IgG monoclonal gammopathy with peripheral neuropathy do not react with MAG.50 Enzyme-linked immunosorbent assay is a simple and convenient method for detecting anti-MAG IgM antibodies. This method correlates with the demonstration of anti-MAG IgM antibodies by the immunoblot technique.109 In summary, antibody activity in patients with monoclonal gammopathies associated with peripheral neuropathy is almost exclusively found in monoclonal proteins of the IgM class. The most prevalent antibody activity is directed against MAG, which is found in both peripheral and central myelin. The monoclonal IgM reacts with the carbohydrate moiety of MAG and not with its polypeptide backbone. These antigenic determinants are shared by MAG and other glycolipids and glycoproteins in peripheral nerve. The carbohydrate moiety of MAG is also the target for the HNK-1 monoclonal antibody, but there is good evidence that the human anti-MAG monoclonal IgM and HNK-1 do not recognize the epitope. A review of monoclonal gammopathy and neuropathy has been published.136 Reaction of Monoclonal IgM with Components Other Than MAG In addition to the MAG determinants, other antigens, mostly glycolipids, have been found.69 Freddo et al.42 reported that the monoclonal IgM from 16 patients with anti-MAG activity bound to two glycolipids in peripheral nerve. IgM monoclonal protein nonreactive with MAG has also reacted specifically with a lipid component from human peripheral nerve.2 In one report, the IgM monoclonal protein reacted with specific sulfated glycolipids of the peripheral nervous system but, in addition, recognized a glycolipid antigen in the central nervous system white matter.59 Gangliosides represent other target antigens.143 Ilyas et al.56 reported that a monoclonal IgM protein in three patients with peripheral neuropathy bound to the carbohydrate portion of MAG and also to a single ganglioside of sciatic nerve. In another patient with an IgM ␬ monoclonal protein and peripheral neuropathy, there was binding to gangliosides containing disialosyl groups.55 In two patients, the antigenic glycolipids recognized by the monoclonal IgM protein also reacted with specific gangliosides.54 In another case, the IgM monoclonal protein bound preferentially to ganglioside GM1.70 Patients with biclonal

2268

Diseases of the Peripheral Nervous System

gammopathy and peripheral neuropathy also have a reaction with various gangliosides.57 Carpo et al.14 reported high titers of anti-GD1, a ganglioside antibody, in 5% of patients with CIDP, 18% with multifocal motor neuropathy, 3.8% with lower motor neuron disease, and 1.8% with amyotrophic lateral sclerosis. Clinical improvement was associated with a decreased antibody titer. In another report, antiganglioside antibodies were found in 15% of 87 patients with polyneuropathy associated with a monoclonal gammopathy. Antiganglioside antibodies have been associated with demyelinating neuropathy and IgM monoclonal gammopathy.33 The IgM protein in patients with peripheral neuropathy may be reactive with chondroitin sulfate C and produce axonal rather than myelin sheath damage. In one patient, clusters of three or more thinly myelinated fibers consistent with regeneration after axonal degeneration were seen.104 Other investigators131,161 also reported binding of IgM protein to chondroitin sulfate C. The monoclonal protein appears to bind chondroitin sulfate C–containing proteoglycans in the peripheral nerve.43 Three of seven patients with an IgM demyelinating polyneuropathy had antisulfatide antibody activity. Sciatic nerve sections showed intense myelin staining with direct immunofluorescence.116 One patient with an IgM ␬ monoclonal protein had antibody-like activity against muscle ␣-actinin.98 In another patient with an IgM monoclonal protein, peripheral neuropathy, and autonomic dysfunction, the monoclonal protein reacted with smooth and striated muscle.48 The antibodies associated with peripheral neuropathy have been reviewed.120,160

IgG and IgA Monoclonal Proteins and Peripheral Neuropathy Patients with IgG monoclonal gammopathy and peripheral neuropathy have been described. The relationship of an IgG monoclonal protein to peripheral neuropathy has not been well documented.65 Read et al.122 described three patients with a benign IgG monoclonal gammopathy and peripheral neuropathy. No direct immunologic role of the monoclonal protein was detected. IgG ␭ was found in the sural nerve of another patient with a benign IgG ␭ monoclonal protein and peripheral neuropathy. Binding of the purified IgG ␭ protein and its F(ab⬘)2 fragment with the nerve and brain tissue was demonstrated.130 A patient with an IgA monoclonal gammopathy and peripheral neuropathy has been described with binding of the monoclonal protein to normal human endoneurium. There was no binding with IgA ␬, IgG, or IgM protein.29 An IgA ␬ monoclonal protein associated with peripheral neuropathy and autonomic neuropathy has been described. Demyelination and axonal loss were documented. Amyloid stains of the sural nerve were negative.3

Treatment Treatment of patients with peripheral neuropathy and monoclonal gammopathy has shown promise but is still disappointing. In a randomized trial, 39 patients with MGUS and peripheral neuropathy received either plasma exchange twice weekly or sham plasma exchange in a double-blind trial. Those who had sham plasma exchange subsequently had plasma exchange in an open trial. The average neuropathy disability score improved by 2 points for the sham exchange group and by 12 points for the plasma exchange group. The neuropathy disability score, weakness score, and some compound muscle action potentials improved more with plasma exchange than with sham exchange. In both the double-blind and the open trials, IgG or IgA gammopathy had a better response to plasma exchange than IgM gammopathy, which had minimal benefit.32 In several other studies, plasmapheresis has been reported to be beneficial. Frayne and Stark41 reported that two patients—one with an IgG ␬ monoclonal protein and the other with an IgA ␬ monoclonal protein—had a response to plasmapheresis. In 10 patients with polyneuropathy and monoclonal gammopathy, Sherman et al.133 reported that plasma exchange produced improvement in neuropathy in 6 and stabilization in 3 others. Neuropathy progressed with cessation of plasma exchange. In another patient with anti-MAG monoclonal protein, plasmapheresis was beneficial. The patient’s signs and symptoms correlated with the IgM level during the year of therapy.49 Mazzi et al.93 reported that 8 of 13 patients with peripheral neuropathy and monoclonal gammopathy obtained benefit with plasma exchange. Immunoadsorption with sheep antibodies to IgM was used to remove IgM from the blood of a patient with peripheral neuropathy and IgM monoclonal protein. The serum level of IgM was decreased to 140 mg/dL. The treatment produced clinical improvement in the neuropathy, allowing the patient to walk without assistance.147 The combination of plasmapheresis and chemotherapy has been reported to be helpful. A woman with peripheral neuropathy and an IgM ␬ monoclonal protein had improvement with 2-L plasma exchanges performed weekly in combination with prednisone and cyclophosphamide therapy.117 Of 13 patients with peripheral neuropathy and an IgM or IgG monoclonal protein, 9 had improvement with plasma exchange and chemotherapy.142 Treatment with chlorambucil and plasmapheresis in one patient produced clinical and electrophysiologic improvement as well as a decrease in IgM deposits in the sural nerve.94 In a prospective study of 44 patients with IgM monoclonal gammopathy and peripheral neuropathy, patients were randomly assigned to treatment with either chlorambucil (0.1 mg/kg daily) orally or chlorambucil plus 15 courses of plasma exchange during the first 4 months of therapy. There was no difference in the response of the two treatment groups, suggesting that plasma exchange provided no additional benefit to chlorambucil therapy.111

2269

Neuropathy Associated with the Monoclonal Gammopathies

Table 100–3. Treatment of Peripheral Neuropathy Effect of Therapy Treatment

MGUS

POEMS

MM

Amyloidosis

Plasmapheresis

IgG and IgA more responsive than IgM Occasional impr

Not effective

Unknown

Not effective

Not effective

Unknown

Not effective

Impr in 50%, but risk of alkylator exposure Impr in some patients Impr in some patients Occasional impr Unknown

Effective in ⬍ 50%

Not applicable

Effective in ⬎ 50%

Stabil and impr possible Not applicable Not applicable Unknown Stabil and impr possible Stabil and impr possible

Stabil and impr possible Not applicable Unknown Unknown Stabil and impr possible Not applicable

Intravenous gamma globulin Oral or intravenous alkylators Fludarabine Rituximab Interferon-␣ High-dose chemotherapy with PBSCT Radiation to dominant bone lesion

Unknown Unknown Unknown Helpful in most

Impr ⫽ improvement; MGUS ⫽ monoclonal gammopathy of undetermined significance; MM ⫽ multiple myeloma; PBSCT ⫽ peripheral blood stem cell transplantation; POEMS ⫽ polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes; Stabil ⫽ stabilization. Data from Kyle, R. A., and Dispenzieri, A.: Neurologic aspects of monoclonal gammopathy of undetermined significance, multiple myeloma, and related disorders. In Gahrton, G., Durie, B., and Samson, D. (eds.): Multiple Myeloma. London, Edward Arnold Publishers (in press), with permission of Edward Arnold Publishers.

Dalakas et al.18 reported that a patient with Waldenström’s macroglobulinemia and polyneuropathy was treated successfully with chlorambucil and prednisone. In another study, 9 of 10 patients with peripheral neuropathy and IgM monoclonal protein had a response to treatment with prednisone, cyclophosphamide, chlorambucil, azathioprine, or plasmapheresis. Five patients had monoclonal proteins that reacted with MAG, and all had a response to treatment.63 We have also given chlorambucil to patients with IgM monoclonal gammopathy and melphalan to those with IgG and IgA monoclonal gammopathies and peripheral neuropathy. Some of these patients have obtained benefit. Fludarabine, a purine analogue, produced benefit in 7 of 10 patients with IgM monoclonal gammopathy and peripheral neuropathy.132 In another report, three of four patients with IgM monoclonal gammopathy and peripheral neuropathy obtained clinical and neurophysiologic benefit from fludarabine.158 Rituximab (Rituxan) is a monoclonal antibody directed against CD20, which is frequently expressed on the surface membrane of lymphocytes of patients with lymphoma. In one study, rituximab produced benefit in all five patients with IgM monoclonal gammopathy and peripheral neuropathy.87 Previous treatment with plasma exchange and cyclophosphamide had been successful, but these five patients had experienced relapse. Responses have been achieved with rituximab, but data are not sufficient to draw definite conclusions about its efficacy. Notermans et al.110 reported an improved response in two of six patients with neuropathy and monoclonal gammopathy treated with high-dose dexamethasone. However, four patients had serious side effects. Intravenous gamma globulin is beneficial in some patients with CIDP and may be given to patients with

sensorimotor peripheral neuropathy and monoclonal gammopathy.38 In a randomized crossover study comparing intravenous gamma globulin with placebo, strength improved in two patients and sensory symptoms decreased in one other.19 Antibody titers to MAG or ganglioside did not change. Less than 20% of the treated patients had any benefit.19 In a prospective randomized trial, only 1 of 10 patients with IgM-associated monoclonal gammopathy had a response to intravenous immune globulin, whereas 8 of 10 had a response to treatment with interferon-␣.91 In a single case, gabapentin was helpful for a patient with severe tremor resulting from chronic demyelinating neuropathy associated with IgM monoclonal protein.128 The Medical Outcome Study 36-item short-form health status scale is useful for measuring quality of life of various patients with immune-mediated polyneuropathies.95 The treatment of MGUS, POEMS syndrome, multiple myeloma, and AL amyloidosis is summarized in Table 100–3.

MONOCLONAL GAMMOPATHY AND MOTOR NEURON DISEASES Although several instances of monoclonal gammopathy and motor neuron disease have been reported, there is little evidence that a causal effect exists. Rowland et al.125 described a 48-year-old man with an IgM ␬ monoclonal protein and amyotrophic lateral sclerosis. The patient had widespread fasciculations and fibrillations and no evidence of sensory neuropathy. Sural nerve biopsy findings were normal. The patient died 14 months later, and at autopsy the number of anterior horn nerve cells was reported to be normal. Many neurons in the lumbar area showed central chromatolysis.

2270

Diseases of the Peripheral Nervous System

Latov81 reviewed from the literature 19 cases of monoclonal gammopathy and motor neuron disease. Anterior horn cell disease was reported in 11 patients with a monoclonal gammopathy. Eight had multiple myeloma or a malignant lymphoproliferative process. Three had benign monoclonal gammopathy.12 Poloni et al.118 found four patients with monoclonal gammopathy among 1200 admitted to a neurologic clinic. Only one had amyotrophic lateral sclerosis, and another had limb girdle muscular dystrophy. Patten114 described three patients with amyotrophic lateral sclerosis and an IgG ␬ monoclonal gammopathy and a fourth with an IgG ␭ protein. Merlini et al.96 reported three patients with amyotrophic lateral sclerosis—two had an IgG monoclonal protein and another had an IgG ␬/IgA ␭ biclonal gammopathy. In another report, 11 of 120 patients (9%) with motor neuron disease had an associated monoclonal protein. Of these 11 patients, 10 had amyotrophic lateral sclerosis and 1 had progressive spinal muscle atrophy.162 IgM monoclonal gammopathy has been reported in a patient with primary lateral sclerosis.28

MULTIPLE MYELOMA Multiple myeloma is characterized by the proliferation of a single clone of plasma cells engaged in the production of a specific immunoglobulin and accounts for about 1% of all types of malignant disease. The incidence of multiple myeloma is approximately 4 per 100,000 per year. The onset is usually between the ages of 40 and 80 years, with a peak incidence in the seventh decade of life. Bone pain, weakness, and fatigue are common symptoms. Renal insufficiency, hypercalcemia, and hyperuricemia may be seen.75 Anemia eventually occurs in almost every patient. Serum protein electrophoresis reveals a discrete band in 80% of patients. A serum monoclonal protein is found with immunofixation in more than 90% of patients. Immunofixation shows a monoclonal protein in the urine of 78% of patients. Of patients with multiple myeloma, 97% have a monoclonal protein in the serum or urine. Radiographs show abnormalities characterized by lytic lesions, osteoporosis, or fractures in 80% of patients. The diagnosis is made by the demonstration of increased numbers of plasma cells in the bone marrow. Neurologic involvement is manifested most often by root pain. Compression of the spinal cord or cauda equina, usually caused by myeloma arising in the marrow cavity of a vertebra and extending to the extradural space, occurs in about 5% of patients. This compression produces back pain with radicular features, weakness or paralysis of the lower extremities, and bowel or bladder incontinence. Peripheral neuropathy may occur but is usually associated with amyloidosis. Peripheral neuropathy associated with multiple myeloma is rare. Silverstein and Doniger135 found peripheral neuropathy

in 10 (3.5%) of 277 hospitalized multiple myeloma patients. Three of the 10 patients also had amyloidosis, which was the most likely cause of the neuropathy. Davison and Balser24 described a 49-year-old woman with multiple myeloma and peripheral neuritis of the upper extremities. She had bilateral atrophy of the thenar and hypothenar eminences and interosseous muscles. At autopsy, there was partial demyelination of one of the posterior dorsal roots. There was some demyelination of single fibers within the nerves of both brachial plexuses. There was no evidence of myelomatous invasion of the peripheral nerves. Peripheral neuropathy may involve both motor and sensory modalities without direct invasion or compression of the nerve in patients with multiple myeloma.149 In one series of 23 patients with multiple myeloma, 3 had clinical neuropathy and 6 had electrophysiologic evidence of peripheral neuropathy. None of the five with sural nerve biopsy had amyloidosis.156 Because peripheral neuropathy may precede the diagnosis of myeloma and dominate the clinical pattern, one must exclude multiple or solitary myeloma in all cases of peripheral neuropathy without known cause.23 We identified 10 patients with typical multiple myeloma and peripheral neuropathy between 1965 and 1978.66 Six patients had myeloma and neuropathy without amyloidosis. Four of the six patients presented with peripheral neuropathy, and myeloma was discovered during evaluation of the neuropathy. The six patients had neurologic symptoms for a median of 27 months before the diagnosis of neuropathy. One patient had neuropathy develop coincidentally with the discovery of myeloma, and one patient had known multiple myeloma for 5 years before the onset of neuropathy. Four patients had a relatively mild sensorimotor neuropathy. At presentation, these patients had slow progression (8 to 39 months) of distal numbness with mild involvement of all modalities of sensation and mild symmetrical distal weakness affecting the legs more than the arms. Neither pain nor autonomic involvement was prominent, although pain resulting from bone or nerve root involvement was present in two of the four patients. Slowing of motor conduction velocity was mild, with slightly low-amplitude compound muscle action potentials. Sensory nerve compound action potentials were low to absent. Fibrillation potentials and motor unit potential changes were mild and restricted to the distal muscles of the legs and arms. During follow-up, the neuropathy remained stable or progressed slowly. There was no evidence of clinical improvement from chemotherapy for myeloma despite improvement of the malignancy in two patients. The fifth patient had a predominantly sensory neuropathy of 24 months’ duration affecting the lower limbs, but no autonomic features or muscle weakness was found. The severity of the neuropathy remained unchanged with chemotherapy. The sixth patient, a 56-year-old man with multiple myeloma for 6 months, presented with a 5-month history

Neuropathy Associated with the Monoclonal Gammopathies

of severe, symmetrical distal and proximal weakness with bilateral facial palsies, respiratory insufficiency, and fasciculations involving the limbs and tongue. Nerve conduction studies showed moderate slowing of motor conduction velocities with an absent ulnar sensory nerve compound action potential. Needle electromyographic examination demonstrated fibrillation potentials in distal and proximal muscles and motor unit potential changes in distal muscles. A biopsy specimen from a proximal muscle showed evidence of neurogenic atrophy. No amyloid was found. Four of the 10 patients with multiple myeloma and peripheral neuropathy had systemic amyloidosis. It is important that biopsy specimens of appropriate tissues be examined in patients with multiple myeloma and peripheral neuropathy because many will have unrecognized primary amyloidosis. Delauche et al.26 described three patients with multiple myeloma and progressive sensorimotor polyneuropathy. None had a high tumor mass or overt multiple myeloma. Electrophysiologic and pathologic changes were consistent with a demyelinating process. The three patients did not have a response to melphalan and prednisone therapy. Borges and Busis9 described a patient with multiple myeloma who had ␭ light chains within neurons. The possibility that ␭ light chains were transported intra-axonally was raised. Sensorimotor peripheral neuropathy was reported in a woman with multiple myeloma producing monoclonal ␬ light chains. No amyloid was found in the sural nerve or in a muscle specimen biopsy.115

2271

fatigue, and oronasal bleeding are common presenting symptoms. Blurred or impaired vision, dyspnea, neurologic symptoms, and congestive heart failure may occur. Pallor, hepatosplenomegaly, and lymphadenopathy are common physical findings. Retinal lesions include hemorrhages, exudates, and venous congestion with vascular segmentation (“sausage” formation). Sudden deafness, progressive spinal muscular atrophy, and multifocal leukoencephalopathy have been reported with macroglobulinemia. A sensorimotor peripheral neuropathy is common. The symptoms and findings do not differ from those of the neuropathy associated with monoclonal gammopathy of the IgM class (see above). Almost all patients with Waldenström’s macroglobulinemia have moderate to severe normocytic, normochromic anemia. Lymphocytosis is not uncommon. The serum protein electrophoretic pattern, characterized by a tall, narrow peak or dense band, is almost always of ␥ mobility. The pattern is indistinguishable from that of multiple myeloma. Seventy-five percent of the IgM proteins have a ␬ light chain. The IgG and IgA levels are frequently reduced. Low-molecular-weight IgM (7S) is present and may account for a large part of the elevated IgM. A monoclonal light chain is found in the urine in 80% of patients. A bone marrow aspirate is often hypocellular, but biopsy specimens are hypercellular and extensively infiltrated with lymphocytes and plasma cells. Lytic bone lesions occur in less than 5% of cases.

REFERENCES SOLITARY PLASMACYTOMA Solitary plasmacytoma may be associated with peripheral neuropathy. In one instance, the patient developed severe peripheral neuropathy and subsequent studies showed a sacral plasmacytoma displacing the rectum anteriorly. The lesion was removed surgically and the site was irradiated. The peripheral neuropathy began to improve 1 week postoperatively, and during the 6-year follow-up there was no recurrence of either plasmacytoma or peripheral neuropathy.22 Read and Warlow121 reviewed 13 previously reported cases of solitary myeloma and peripheral neuropathy and added 3 of their own. The sensorimotor neuropathy may be the presenting feature of the illness. It occurs mainly in males and may improve after treatment.

WALDENSTRÖM’S MACROGLOBULINEMIA Macroglobulinemia is the result of an uncontrolled proliferation of lymphocytes and plasma cells in which a large monoclonal IgM protein is produced. The median age of patients is 63 years, and about 60% are males.74 Weakness,

1. Axelsson, U., Bachmann, R., and Hallen, J.: Frequency of pathological proteins (M-components) in 6,995 sera from an adult population. Acta Med. Scand. 179:235, 1966. 2. Baba, H., Miyatani, N., Sato, S., et al.: Antibody to glycolipid in a patient with IgM paraproteinemia and polyradiculoneuropathy. Acta Neurol. Scand. 72:218, 1985. 3. Bailey, R. O., Ritaccio, A. L., Bishop, M. B., and Wu, A. Y.: Benign monoclonal IgA␬ gammopathy associated with polyneuropathy and dysautonomia. Acta Neurol. Scand. 73:574, 1986. 4. Baldini, L., Guffanti, A., Cesana, B. M., et al.: Role of different hematologic variables in defining the risk of malignant transformation in monoclonal gammopathy. Blood 87:912, 1996. 5. Baldini, L., Nobile-Orazio, E., Guffanti, A, et al.: Peripheral neuropathy in IgM monoclonal gammopathy and Waldenström’s macroglobulinemia: a frequent complication in elderly males with low MAG-reactive serum monoclonal component. Am. J. Hematol. 45:25, 1994. 6. Besinger, U. A., Toyka, K. V., Anzil, A. P., et al.: Myeloma neuropathy: passive transfer from man to mouse. Science 213:1027, 1981. 7. Bleasel, A. F., Hawke, S. H., Pollard, J. D., and McLeod, J. G.: IgG monoclonal paraproteinaemia and peripheral neuropathy. J. Neurol. Neurosurg. Psychiatry 56:52, 1993.

2272

Diseases of the Peripheral Nervous System

8. Bollensen, E., Steck, A. J., and Schachner, M.: Reactivity with the peripheral myelin glycoprotein P0 in serum from patients with monoclonal IgM gammopathy and polyneuropathy. Neurology 38:1266, 1988. 9. Borges, L. F., and Busis, N. A.: Intraneuronal accumulation of myeloma proteins. Arch. Neurol. 42:690, 1985. 10. Bosch, E. P., Ansbacher, L. E., Goeken, J. A., and Cancilla, P. A.: Peripheral neuropathy associated with monoclonal gammopathy: studies of intraneural injections of monoclonal immunoglobulin sera. J. Neuropathol. Exp. Neurol. 41:446, 1982. 11. Brouet, J. C., Dellagi, K., Gendron, M. C., et al. Expression of a public idiotype by human monoclonal IgM directed to myelin-associated glycoprotein and characterization of the variability subgroup of their heavy and light chains. J. Exp. Med. 170:1551, 1989. 12. Brudon, F., Bady, B., Chauplannaz, G., et al.: Clinical amyotrophic lateral sclerosis syndromes in gammopathies: clinical, electrophysiologic and histoimmunologic aspects [in French]. Rev. Electroencephalogr. Neurophysiol. Clin. 15:369, 1986. 13. Busis, N. A., Halperin, J. J., Stefansson, K., et al.: Peripheral neuropathy, high serum IgM, and paraproteinemia in mother and son. Neurology 35:679, 1985. 14. Carpo, M., Nobile-Orazio, E., Meucci, N., et al.: Anti-GD1a ganglioside antibodies in peripheral motor syndromes. Ann. Neurol. 39:539, 1996. 15. Cesana, C., Klersy, C., Barbarano, L., et al.: Prognostic factors for malignant transformation in monoclonal gammopathy of undetermined significance and smoldering multiple myeloma. J. Clin. Oncol. 20:1625, 2002. 16. Chassande, B., Leger, J. M., Younes-Chennoufi, A. B., et al.: Peripheral neuropathy associated with IgM monoclonal gammopathy: correlations between M-protein antibody activity and clinical/electrophysiological features in 40 cases. Muscle Nerve 21:55, 1998. 17. Cocito, D., Isoardo, G., Ciaramitaro, P., et al.: Terminal latency index in polyneuropathy with IgM paraproteinemia and anti-MAG antibody. Muscle Nerve 24:1278, 2001. 18. Dalakas, M. C., Flaum, M. A., Rick, M., et al.: Treatment of polyneuropathy in Waldenström’s macroglobulinemia: role of paraproteinemia and immunologic studies. Neurology 33:1406, 1983. 19. Dalakas, M. C., Quarles, R. H., Farrer, R. G., et al.: A controlled study of intravenous immunoglobulin in demyelinating neuropathy with IgM gammopathy. Ann. Neurol. 40:792, 1996. 20. D’Amico, G., Colasanti, G., Ferrario, F., et al.: Renal involvement in essential mixed cryoglobulinemia: a peculiar type of immune-mediated renal disease. Adv. Nephrol. Necker Hosp. 17:219, 1988. 21. Dammacco, F., and Sansonno, D.: Antibodies to hepatitis C virus in essential mixed cryoglobulinaemia. Clin. Exp. Immunol. 87:352, 1992. 22. Davidson, S.: Solitary myeloma with peripheral polyneuropathy—recovery after treatment. Calif. Med. 116:68, 1972. 23. Davis, L. E., and Drachman, D. B.: Myeloma neuropathy: successful treatment of two patients and review of cases. Arch. Neurol. 27:507, 1972.

24. Davison, C., and Balser, B. H.: Myeloma and its neural complications. Arch. Surg. 35:913, 1937. 25. Dayan, A. D., and Stokes, M. I.: Peripheral neuropathy and experimental myeloma in the mouse. Nat. New Biol. 236:117, 1972. 26. Delauche, M. C., Clauvel, J. P., and Seligmann, M.: Peripheral neuropathy and plasma cell neoplasias: a report of 10 cases. Br. J. Haematol. 48:383, 1981. 27. Dellagi, K., Dupouey, P., Brouet, J. C., et al.: Waldenström’s macroglobulinemia and peripheral neuropathy: a clinical and immunologic study of 25 patients. Blood 62:280, 1983. 28. Desai, J., and Swash, M.: IgM paraproteinemia in a patient with primary lateral sclerosis. Neuromuscul. Disord. 9:38, 1999. 29. Dhib-Jalbut, S., and Liwnicz, B. H.: Binding of serum IgA of multiple myeloma to normal peripheral nerve. Acta Neurol. Scand. 73:381, 1986. 30. Dispenzieri, A., Gertz, M. A., Therneau, T. M., and Kyle, R. A.: Retrospective cohort study of 148 patients with polyclonal gammopathy. Mayo Clin. Proc. 76:476, 2001. 31. Donaghy, M., and Duchen, L. W.: Sera from patients with motor neuron disease and associated paraproteinaemia fail to inhibit experimentally induced sprouting of motor nerve terminals. J. Neurol. Neurosurg. Psychiatry 49:817, 1986. 32. Dyck, P. J., Low, P. A., Windebank, A. J., et al.: Plasma exchange in polyneuropathy associated with monoclonal gammopathy of undetermined significance. N. Engl. J. Med. 325:1482, 1991. 33. Eurelings, M., Ang, C. W., Notermans, N. C., et al.: Antiganglioside antibodies in polyneuropathy associated with monoclonal gammopathy. Neurology 57:1909, 2001. 34. Eurelings, M., Moons, K. G., Notermans, N. C., et al.: Neuropathy and IgM M-proteins: prognostic value of antibodies to MAG, SGPG, and sulfatide. Neurology 56:228, 2001. 35. Eurelings, M., Notermans, N. C., Franssen, H., et al.: MRI of the brachial plexus in polyneuropathy associated with monoclonal gammopathy. Muscle Nerve 24:1312, 2001. 36. Eurelings, M., Notermans, N. C., Van de Donk, N. W., and Lokhorst, H. M.: Risk factors for hematological malignancy in polyneuropathy associated with monoclonal gammopathy. Muscle Nerve 24:1295, 2001. 37. Eurelings, M., Notermans, N. C., Wokke, J. H., et al.: Sural nerve T cells in demyelinating polyneuropathy associated with monoclonal gammopathy. Acta Neuropathol. (Berl.) 103:107, 2002. 38. Faed, J. M., Day, B., Pollock, M., et al.: High-dose intravenous human immunoglobulin in chronic inflammatory demyelinating polyneuropathy. Neurology 39:422, 1989. 39. Forssman, O., Bjorkman, G., Hollender, A., and Englund, N. E.: IgM-producing lymphocytes in peripheral nerve in a patient with benign monoclonal gammopathy. Scand. J. Haematol. 11:332, 1973. 40. Frail, D. E., Edwards, A. M., and Braun, P. E.: Molecular characteristics of the epitope in myelin-associated glycoprotein that is recognized by a monoclonal IgM in human neuropathy patients. Mol. Immunol. 21:721, 1984. 41. Frayne, J., and Stark, R. J.: Peripheral neuropathy with gammopathy responding to plasmapheresis. Clin. Exp. Neurol. 21:195, 1985.

Neuropathy Associated with the Monoclonal Gammopathies 42. Freddo, L., Ariga, T., Saito, M., et al.: The neuropathy of plasma cell dyscrasia: binding of IgM M-proteins to peripheral nerve glycolipids. Neurology 35:1420, 1985. 43. Freddo, L., Sherman, W. H., and Latov, N.: Glycosaminoglycan antigens in peripheral nerve: studies with antibodies from a patient with neuropathy and monoclonal gammopathy. J. Neuroimmunol. 12:57, 1986. 44. Galassi, G., and Nemni, R.: Sensory action potentials and biopsy of the sural nerve in the neuropathy of nonmalignant IgM␬ plasma cell dyscrasia. Eur. Neurol. 25:1, 1986. 45. Gorevic, P. D., Kassab, H. J., Levo, Y., et al.: Mixed cryoglobulinemia: clinical aspects and long-term follow-up of 40 patients. Am. J. Med. 69:287, 1980. 46. Gorson, K. C., Allam, G., and Ropper, A. H.: Chronic inflammatory demyelinating polyneuropathy: clinical features and response to treatment in 67 consecutive patients with and without a monoclonal gammopathy. Neurology 48:321, 1997. 47. Gosselin, S., Kyle, R. A., and Dyck, P. J.: Neuropathy associated with monoclonal gammopathies of undetermined significance. Ann. Neurol. 30:54, 1991. 48. Green, A. R., and Whittaker, J. A.: IgM paraprotein reactive with muscle antigens associated with autonomic and peripheral neuropathy. Br. J. Haematol. 72:108, 1989. 49. Haas, D. C., and Tatum, A. H.: Plasmapheresis alleviates neuropathy accompanying IgM anti-myelin-associated glycoprotein paraproteinemia. Ann. Neurol. 23:394, 1988. 50. Hafler, D. A., Johnson, D., Kelly, J. J., et al.: Monoclonal gammopathy and neuropathy: myelin-associated glycoprotein reactivity and clinical characteristics. Neurology 36:75, 1986. 51. Hällen, J.: Frequency of “abnormal” serum globulins (Mcomponents) in the aged. Acta Med. Scand. 173:737, 1963. 52. Harbs, H., Arfmann, M., Frick, E., et al.: Reactivity of sera and isolated monoclonal IgM from patients with Waldenström’s macroglobulinaemia with peripheral nerve myelin. J. Neurol. 232:43, 1985. 53. Hays, A. P., Latov, N., Takatsu, M., and Sherman, W. H.: Experimental demyelination of nerve induced by serum of patients with neuropathy and an anti-MAG IgM M-protein. Neurology 37:242, 1987. 54. Ilyas, A. A., Li, S. C., Chou, D. K., et al.: Gangliosides GM2, IV4GalNAcGM1b, and IV4GalNAcGC1a as antigens for monoclonal immunoglobulin M in neuropathy associated with gammopathy. J. Biol. Chem. 263:4369, 1988. 55. Ilyas, A. A., Quarles, R. H., Dalakas, M. C., et al.: Monoclonal IgM in a patient with paraproteinemic polyneuropathy binds to gangliosides containing disialosyl groups. Ann. Neurol. 18:655, 1985. 56. Ilyas, A. A., Quarles, R. H., MacIntosh, T. D., et al.: IgM in a human neuropathy related to paraproteinemia binds to a carbohydrate determinant in the myelin-associated glycoprotein and to a ganglioside. Proc. Natl. Acad. Sci. U. S. A. 81:1225, 1984. 57. Ilyas, A. A., Willison, H. J., Dalakas, M. C., et al.: Identification and characterization of gangliosides reacting with IgM paraproteins in three patients with neuropathy associated with biclonal gammopathy. J. Neurochem. 51:851, 1988.

2273

58. Isobe, T., and Osserman, E. F.: Pathologic conditions associated with plasma cell dyscrasias: a study of 806 cases. Ann. N. Y. Acad. Sci. 190:507, 1971. 59. Jauberteau, M. O., Younes-Chennoufi, B., Rigaud, M., and Baumann, N.: IgM gammopathy and polyneuropathy react with an antigenic glycolipid present in human central nervous system. Neurosci. Lett. 97:181, 1989. 60. Johansen, P., and Leegaard, O. F.: Peripheral neuropathy and paraproteinemia: an immunohistochemical and serologic study. Clin. Neuropathol. 4:99, 1985. 61. Jønsson, V., Schroder, H. D., Staehelin Jensen, T., et al.: Autoimmunity related to IgM monoclonal gammopathy of undetermined significance: peripheral neuropathy and connective tissue sensibilization caused by IgM M-proteins. Acta Med. Scand. 223:255, 1988. 62. Kahn, S. N., Riches, P. G., and Kohn, J.: Paraproteinaemia in neurological disease: incidence, associations, and classification of monoclonal immunoglobulins. J. Clin. Pathol. 33:617, 1980. 63. Kelly, J. J., Adelman, L. S., Berkman, E., and Bhan, I.: Polyneuropathies associated with IgM monoclonal gammopathies. Arch. Neurol. 45:1355, 1988. 64. Kelly, J. J. Jr.: The electrodiagnostic findings in peripheral neuropathy associated with monoclonal gammopathy. Muscle Nerve 6:504, 1983. 65. Kelly, J. J. Jr., Kyle, R. A., and Latov, N.: Polyneuropathies Associated with Plasma Cell Dyscrasias. Boston, Martinus Nijhoff, 1987. 66. Kelly, J. J. Jr., Kyle, R. A., O’Brien, P. C., and Dyck, P. J.: Prevalence of monoclonal protein in peripheral neuropathy. Neurology 31:1480, 1981. 67. Keren, D. F., Alexanian, R., Goeken, J. A., et al.: Guidelines for clinical and laboratory evaluation of patients with monoclonal gammopathies. Arch. Pathol. Lab. Med. 123:106, 1999. 68. Krol-v Straaten, M. J., Ackerstaff, R. G., and De Maat, C. E.: Peripheral polyneuropathy and monoclonal gammopathy of undetermined significance. J. Neurol. Neurosurg. Psychiatry 48:706, 1985. 69. Kusunoki, S., Kohriyama, T., Pachner, A. R., et al.: Neuropathy and IgM paraproteinemia: differential binding of IgM M-proteins to peripheral nerve glycolipids. Neurology 37:1795, 1987. 70. Kusunoki, S., Shimizu, T., Matsumura, K., et al.: Motor dominant neuropathy and IgM paraproteinemia: the IgM M-protein binds to specific gangliosides. J. Neuroimmunol. 21:177, 1989. 71. Kvarnström, M., Sidorova, E., Nilsson, J., et al.: Myelin protein P0-specific IgM producing monoclonal B cell lines were established from polyneuropathy patients with monoclonal gammopathy of undetermined significance (MGUS). Clin. Exp. Immunol. 127:255, 2002. 72. Kyle, R. A.: Sequence of testing for monoclonal gammopathies. Arch. Pathol. Lab. Med. 123:114, 1999. 73. Kyle, R. A., Finkelstein, S., Elveback, L. R., and Kurland, L. T.: Incidence of monoclonal proteins in a Minnesota community with a cluster of multiple myeloma. Blood 40:719, 1972. 74. Kyle, R. A., and Garton, J. P.: The spectrum of IgM monoclonal gammopathy in 430 cases. Mayo Clin. Proc. 62:719, 1987.

2274

Diseases of the Peripheral Nervous System

75. Kyle, R. A., Gertz, M. A., Witzig, T. E., et al.: Review of 1027 patients with newly diagnosed multiple myeloma. Mayo Clin. Proc. 78:21, 2003. 76. Kyle, R. A., and Greipp, P. A.: Plasma cell dyscrasias: current status. Crit. Rev. Oncol. Hematol. 8:93, 1988. 77. Kyle, R. A., Katzmann, J. A., Lust, J. A., and Dispenzieri, A.: Immunochemical characterization of immunoglobulins. In Rose, N. R., Hamilton, R. G., and Detrick, B. (eds): Manual of Clinical Laboratory Immunology. Washington, DC, ASM Press, p. 71, 2002. 78. Kyle, R. A., Robinson, R. A., and Katzmann, J. A.: The clinical aspects of biclonal gammopathies: review of 57 cases. Am. J. Med. 71:999, 1981. 79. Kyle, R. A., Therneau, T. M., Rajkumar, S. V., et al.: A longterm study of prognosis in monoclonal gammopathy of undetermined significance. N. Engl. J. Med. 346:564, 2002. 80. Lach, B., Rippstein, P., Atack, D., et al.: Immunoelectron microscopic localization of monoclonal IgM antibodies in gammopathy associated with peripheral demyelinating neuropathy. Acta Neuropathol. 85:298, 1993. 81. Latov, N.: Plasma cell dyscrasia and motor neuron disease. Adv. Neurol. 36:273, 1982. 82. Latov, N., Hays, A. P., and Sherman, W. H.: Peripheral neuropathy and anti-MAG antibodies. Crit. Rev. Neurobiol. 3:301, 1988. 83. Latov, N., Sherman, W. H., Nemni, R., et al.: Plasma-cell dyscrasia and peripheral neuropathy with a monoclonal antibody to peripheral-nerve myelin. N. Engl. J. Med. 303:618, 1980. 84. Lawlor, M. W., Richards, M. P., Fisher, M. A., and Stubbs, E. B. Jr.: Sensory nerve conduction deficit in experimental monoclonal gammopathy of undetermined significance (MGUS) neuropathy. Muscle Nerve 24:809, 2001. 85. Leibowitz, S., Gregson, N. A., Kennedy, M., and Kahn, S. N.: IgM paraproteins with immunological specificity for a Schwann cell component and peripheral nerve myelin in patients with polyneuropathy. J. Neurol. Sci. 59:153, 1983. 86. Letendre, L., and Kyle, R. A.: Monoclonal cryoglobulinemia with high thermal insolubility. Mayo Clin. Proc. 57:629, 1982. 87. Levine, T. D., and Pestronk, A.: IgM antibody-related polyneuropathies: B-cell depletion chemotherapy using rituximab. Neurology 52:1701, 1999. 88. Lieberman, F., Marton, L. S., and Stefansson, K.: Pattern of reactivity of IgM from the sera of eight patients with IgM monoclonal gammopathy and neuropathy with components of neural tissues: evidence for interaction with more than one epitope. Acta Neuropathol. 68:196, 1985. 89. Maisonobe, T., Chassande, B., Verin, M., et al.: Chronic dysimmune demyelinating polyneuropathy: a clinical and electrophysiological study of 93 patients. J. Neurol. Neurosurg. Psychiatry 61:36, 1996. 90. Manschot, S. M., Notermans, N. C., van den Berg, L. H., et al.: Three families with polyneuropathy associated with monoclonal gammopathy. Arch. Neurol. 57:740, 2000. 91. Mariette, X., Chastang, C., Clavelou, P., et al. for the IgMassociated Polyneuropathy Study Group: A randomised clinical trial comparing interferon-alpha and intravenous immunoglobulin in polyneuropathy associated with monoclonal IgM. J. Neurol. Neurosurg. Psychiatry 63:28, 1997.

92. Mata, M., Kahn, S. N., and Fink, D. J.: A direct electron microscopic immunocytochemical study of IgM paraproteinemic neuropathy. Arch. Neurol. 45:693, 1988. 93. Mazzi, G., Raineri, A., Zucco, M., et al.: Plasma-exchange in chronic peripheral neurological disorders. Int. J. Artif. Organs 22:40, 1999. 94. Meier, C., Vandevelde, M., Steck, A., and Zurbriggen, A.: Demyelinating polyneuropathy associated with monoclonal IgM-paraproteinaemia: histological, ultrastructural and immunocytochemical studies. J. Neurol. Sci. 63:353, 1984. 95. Merkies, I. S., Schmitz, P. I., van der Meche, F. G., et al.: Quality of life complements traditional outcome measures in immune-mediated polyneuropathies. Neurology 59:84, 2002. 96. Merlini, G., Rutigliano, L., Masnaghetti, S., et al.: Neuromuscular disorders associated with monoclonal immunoglobulins. In Abstracts of the International Conference on Multiple Myeloma: Biology, Pathophysiology, Prognosis, and Treatment, Bologna, Italy, June 19–22, p. 215, 1989. 97. Meucci, N., Baldini, L., Cappellari, A., et al.: Anti-myelinassociated glycoprotein antibodies predict the development of neuropathy in asymptomatic patients with IgM monoclonal gammopathy. Ann. Neurol. 46:119, 1999. 98. Micouin, C., Seigneurin, J. M., Renversez, J. C., et al.: Monoclonal IgM with anti-muscle tissue activity in a patient with Waldenström’s macroglobulinemia: secretion of identical immunoglobulin by the established lymphoid cell line. Clin. Exp. Immunol. 58:677, 1984. 99. Monaco, S., Bonetti, B., Ferrari, S., et al.: Complementmediated demyelination in patients with IgM monoclonal gammopathy and polyneuropathy. N. Engl. J. Med. 322:649, 1990. 100. Montagnino, G.: Reappraisal of the clinical expression of mixed cryoglobulinemia. Springer Semin. Immunopathol. 10:1, 1988. 101. Mygland, A., and Monstad, P.: Chronic polyneuropathies in Vest-Agder, Norway. Eur. J. Neurol. 8:157, 2001. 102. Nardelli, E., Pizzighella, S., Tridente, G., and Rizzuto, N.: Peripheral neuropathy associated with immunoglobulin disorders: an immunological and ultrastructural study. Acta Neuropathol. (Berl.) Suppl. VII:258, 1980. 103. Nemni, R., Fazio, R., Corbo, M., et al.: Peripheral neuropathy associated with experimental plasma cell neoplasm in the mouse. J. Neurol. Sci. 77:321, 1987. 104. Nemni, R., Galassi, G., Latov, N., et al.: Polyneuropathy in nonmalignant IgM plasma cell dyscrasia: a morphological study. Ann. Neurol. 14:43, 1983. 105. Nobile-Orazio, E., Barbieri, S., Baldini, L., et al.: Peripheral neuropathy in monoclonal gammopathy of undetermined significance: prevalence and immunopathogenetic studies. Acta Neurol. Scand. 85:383, 1992. 106. Nobile-Orazio, E., and Carpo, M.: Neuropathy and monoclonal gammopathy. Curr. Opin. Neurol. 14:615, 2001. 107. Nobile-Orazio, E., Francomano, E., Daverio, R., et al.: Anti-myelin-associated glycoprotein IgM antibody titers in neuropathy associated with macroglobulinemia. Ann. Neurol. 26:543, 1989. 108. Nobile-Orazio, E., Hays, A. P., Latov, N., et al.: Specificity of mouse and human monoclonal antibodies to myelinassociated glycoprotein. Neurology 34:1336, 1984.

Neuropathy Associated with the Monoclonal Gammopathies 109. Nobile-Orazio, E., Vietorisz, T., Messito, M. J., et al.: AntiMAG IgM antibodies in patients with neuropathy and IgM M proteins: detection by ELISA. Neurology 33:939, 1983. 110. Notermans, N. C., Vermeulen, M., Lokhorst, H. M., et al.: Pulsed high-dose dexamethasone treatment of polyneuropathy associated with monoclonal gammopathy. J. Neurol. 244:462, 1997. 111. Oksenhendler, E., Chevret, S., Leger, J. M., et al. for the IgM-associated Polyneuropathy Study Group: Plasma exchange and chlorambucil in polyneuropathy associated with monoclonal IgM gammopathy. J. Neurol. Neurosurg. Psychiatry 59:243, 1995. 112. Osby, E., Noring, L., Hast, R., et al.: Benign monoclonal gammopathy and peripheral neuropathy. Br. J. Haematol. 51:531, 1982. 113. O’Shannessy, D. J., Ilyas, A. A., Dalakas, M. C., et al.: Specificity of human IgM monoclonal antibodies from patients with peripheral neuropathy. J. Neuroimmunol. 11:131, 1986. 114. Patten, B. M.: Neuropathy and motor neuron syndromes associated with plasma cell disease. Acta Neurol. Scand. 70:47, 1984. 115. Pellegrini, G., Scarlato, G., Moggio, M., et al.: Sensorimotor polyneuropathy in light chain multiple myeloma. Acta Neuropathol. (Berl.) Suppl. VII:255, 1980. 116. Petratos, S., Turnbull, V. J., Papadopoulos, R., et al.: Hightitre IgM anti-sulfatide antibodies in individuals with IgM paraproteinaemia and associated peripheral neuropathy. Immunol. Cell Biol. 78:124, 2000. 117. Pollard, J. D., McLeod, J. G., and Feeney, D.: Peripheral neuropathy in IgM kappa paraproteinaemia: clinical and ultrastructural studies in two patients. Clin. Exp. Neurol. 21:41, 1985. 118. Poloni, M., Rocchelli, B., Pinelli, P., and Scelsi, R.: Neuromuscular diseases associated with benign monoclonal gammopathy: study of CSF and serum proteins by isoelectric focusing. Acta Neurol. Scand. 65:154, 1982. 119. Powell, H. C., Rodriguez, M., and Hughes, R. A.: Microangiopathy of vasa nervorum in dysglobulinemic neuropathy. Ann. Neurol. 15:386, 1984. 120. Quarles, R. H., and Weiss, M. D.: Autoantibodies associated with peripheral neuropathy. Muscle Nerve 22:800, 1999. 121. Read, D., and Warlow, C.: Peripheral neuropathy and solitary plasmacytoma. J. Neurol. Neurosurg. Psychiatry 41:177, 1978. 122. Read, D. J., Vanhegan, R. I., and Matthews, W. B.: Peripheral neuropathy and benign IgG paraproteinaemia. J. Neurol. Neurosurg. Psychiatry 41:215, 1978. 123. Rebai, T., Mhiri, C., Heine, P., et al.: Focal myelin thickenings in a peripheral neuropathy associated with IgM monoclonal gammopathy. Acta Neuropathol. 79:226, 1989. 124. Riches, P. G., Sheldon, J., Smith, A. M., and Hobbs, J. R.: Overestimation of monoclonal immunoglobulin by immunochemical methods. Ann. Clin. Biochem. 28:253, 1991. 125. Rowland, L. P., Defendini, R., Sherman, W., et al.: Macroglobulinemia with peripheral neuropathy simulating motor neuron disease. Ann. Neurol. 11:532, 1982. 126. Saito, T., Sherman, W. H., and Latov, N.: Specificity and idiotype of M-proteins that react with MAG in patients with neuropathy. J. Immunol. 130:2496, 1983.

2275

127. Saleun, J. P., Vicariot, M., Deroff, P., and Morin, J. F.: Monoclonal gammopathies in the adult population of Finistère, France. J. Clin. Pathol. 35:63, 1982. 128. Saverino, A., Solaro, C., Capello, E., et al.: Tremor associated with benign IgM paraproteinaemic neuropathy successfully treated with gabapentin. Mov. Disord. 16:967, 2001. 129. Seligmann, M., Brouet, J.-C., and Dellagi, K.: Antibody activities of human monoclonal gammopathies with special emphasis on monoclonal IgM in patients with peripheral neuropathy. In Proceedings of the EURAGE Symposium on Monoclonal Gammopathies: Clinical Significance and Basic Mechanisms, Brussels, September 19–20, p. 49, 1985. 130. Sewell, H. F., Matthews, J. B., Gooch, E., et al.: Autoantibody to nerve tissue in a patient with a peripheral neuropathy and an IgG paraprotein. J. Clin. Pathol. 34:1163, 1981. 131. Sherman, W. H., Latov, N., Hays, A. P., et al.: Monoclonal IgM kappa antibody precipitating with chondroitin sulfate C from patients with axonal polyneuropathy and epidermolysis. Neurology 33:192, 1983. 132. Sherman, W. H., Latov, N., Lange, D., et al.: Fludarabine for IgM antibody-mediated neuropathies [abstract]. Ann. Neurol. 36:326, 1994. 133. Sherman, W. H., Olarte, M. R., McKiernan, G., et al.: Plasma exchange treatment of peripheral neuropathy associated with plasma cell dyscrasia. J. Neurol. Neurosurg. Psychiatry 47:813, 1984. 134. Shy, M. E., Vietorisz, T., Nobile-Orazio, E., and Latov, N.: Specificity of human IgM M-proteins that bind to myelinassociated glycoprotein: peptide mapping, deglycosylation, and competitive binding studies. J. Immunol. 133:2509, 1984. 135. Silverstein, A., and Doniger, D. E.: Neurologic complications of myelomatosis. Arch. Neurol. 9:102, 1963. 136. Simmons, Z.: Paraproteinemia and neuropathy. Curr. Opin. Neurol. 12:589, 1999. 137. Simmons, Z., Albers, J. W., Bromberg, M. B., and Feldman, E. L.: Presentation and initial clinical course in patients with chronic inflammatory demyelinating polyradiculoneuropathy: comparison of patients without and with monoclonal gammopathy. Neurology 43:2202, 1993. 138. Simmons, Z., Albers, J. W., Bromberg, M. B., and Feldman, E. L.: Long-term follow-up of patients with chronic inflammatory demyelinating polyradiculoneuropathy, without and with monoclonal gammopathy. Brain 118:359, 1995. 139. Simmons, Z., Bromberg, M. B., Feldman, E. L., and Blaivas, M.: Polyneuropathy associated with IgA monoclonal gammopathy of undetermined significance. Muscle Nerve 16:77, 1993. 140. Simovic, D., Gorson, K. C., and Ropper, A. H.: Comparison of IgM-MGUS and IgG-MGUS polyneuropathy. Acta Neurol. Scand. 97:194, 1998. 141. Smith, I. S., Kahn, S. N., Lacey, B. W., et al.: Chronic demyelinating neuropathy associated with benign IgM paraproteinaemia. Brain 106:169, 1983. 142. Smith, T., Sherman, W., Olarte, M. R., and Lovelace, R. E.: Peripheral neuropathy associated with plasma cell dyscrasia: a clinical and electrophysiological follow-up study. Acta Neurol. Scand. 75:244, 1987.

2276

Diseases of the Peripheral Nervous System

143. Steck, A. J., Murray, N., Dellagi, K., et al.: Peripheral neuropathy associated with monoclonal IgM autoantibody. Ann. Neurol. 22:764, 1987. 144. Steck, A. J., Murray, N., Meier, C., et al.: Demyelinating neuropathy and monoclonal IgM antibody to myelin-associated glycoprotein. Neurology 33:19, 1983. 145. Stefansson, K., Marton, L., Antel, J. P., et al.: Neuropathy accompanying IgM lambda monoclonal gammopathy. Acta Neuropathol. 59:255, 1983. 146. Suarez, G.A., and Kelly, J. J. Jr.: Polyneuropathy associated with monoclonal gammopathy of undetermined significance: further evidence that IgM-MGUS neuropathies are different than IgG-MGUS. Neurology 43:1304, 1993. 147. Toepfer, M., Schroeder, M., Muller-Felber, W., et al.: Successful management of polyneuropathy associated with IgM gammopathy of undetermined significance with antibody-based immunoadsorption. Clin. Nephrol. 53:404, 2000. 148. Valli, G., De Vecchi, A., Gaddi, L., et al.: Peripheral nervous system involvement in essential cryoglobulinemia and nephropathy. Clin. Exp. Rheumatol. 7:479, 1989. 149. Victor, M., Banker, B. Q., and Adams, R. D.: The neuropathy of multiple myeloma. J. Neurol. Neurosurg. Psychiatry. 21:73, 1958. 150. Vital, A.: Paraproteinemic neuropathies. Brain Pathol. 11:399, 2001. 151. Vital, A., Vital, C., Julien, J., et al.: Polyneuropathy associated with IgM monoclonal gammopathy: immunological and pathological study in 31 patients. Acta Neuropathol. 79:160, 1989. 152. Vladutiu, A. O.: Prevalence of M-proteins in serum of hospitalized patients: physician’s response to finding M-proteins in serum protein electrophoresis. Ann. Clin. Lab. Sci. 17:157, 1987.

153. Vrethem, M., Cruz, M., Wen-Xin, H., et al.: Clinical, neurophysiological and immunological evidence of polyneuropathy in patients with monoclonal gammopathies. J. Neurol. Sci. 114:193, 1993. 154. Waldenström, J.: Abnormal proteins in myeloma. Adv. Intern. Med. 5:398, 1952. 155. Waldenström, J.: Studies on conditions associated with disturbed gamma globulin formation (gammopathies). Harvey Lect. 56:211, 1961. 156. Walsh, J. C.: The neuropathy of multiple myeloma: an electrophysiological and histological study. Arch. Neurol. 25:404, 1971. 157. Whicher, J. T., Wallage, M., and Fifield, R.: Use of immunoglobulin heavy- and light-chain measurements compared with existing techniques as a means of typing monoclonal immunoglobulins. Clin. Chem. 33:1771, 1987. 158. Wilson, H. C., Lunn, M. P., Schey, S., and Hughes, R. A.: Successful treatment of IgM paraproteinaemic neuropathy with fludarabine. J. Neurol. Neurosurg. Psychiatry 66:575, 1999. 159. Wilson, J. R., Stittsworth, J. D. Jr., and Fisher, M. A.: Electrodiagnostic patterns in MGUS neuropathy. Electromyogr. Clin. Neurophysiol. 41:409, 2001. 160. Wolfe, G. I., and Nations, S. P.: Guide to autoantibody testing in peripheral neuropathies. Neurologist 7:195, 2001. 161. Yee, W. C., Hahn, A. F., Hearn, S. A., and Rupar, A. R.: Neuropathy in IgM lambda paraproteinemia: immunoreactivity to neural proteins and chondroitin sulfate. Acta Neuropathol. 78:57, 1989. 162. Younger, D. S., Rowland, L. P., Latov, N., et al.: Motor neuron disease and amyotrophic lateral sclerosis: relation of high CSF protein content to paraproteinemia and clinical syndromes. Neurology 40:595, 1990.

101 Multifocal Motor Neuropathy and Conduction Block BRUCE V. TAYLOR AND HUGH J. WILLISON

Background Clinical Features Differential Diagnosis Natural History Electrophysiologic Characteristics Conduction Block NCSs Outside of Areas of Conduction Block

Electromyographic Findings Other Electrophysiologic Techniques Laboratory Findings Antibodies to Ganglioside Epitopes in MMN-CB Other Laboratory Markers Pathologic Changes

BACKGROUND Current clinical practice regards multifocal motor neuropathy (MMN) as an acquired, autoimmune, chronic asymmetrical motor neuropathy with focal motor conduction block on electrophysiologic testing. Because of this latter feature, it is also termed multifocal motor neuropathy with persistent motor conduction block (MMN-CB). The first description of the clinical entity now recognized as MMN-CB occurred in the mid-1980s when three authors described cases of pure motor neuropathy with persistent conduction block on motor but not sensory nerves across the same segment.19,87,101 In 1988, two groups presented seven patients with what is now referred to as MMN-CB. These seven patients were probably the first clearly defined cases of MMN-CB published.88,96 Cases clearly existed prior to this and may have been referred to under diagnoses such as monomelic amyotrophy (Sobue’s syndrome)107 or motor-predominant chronic inflammatory demyelinating polyneuropathy (CIDP) (see Differential Diagnosis below). The five cases described by Parry and Clarke88 had asymmetrical, upper limb–predominant, slowly progressive wasting and weakness, with normal sensory nerve conduction studies (NCSs) and a normal sensory examination. Electrophysiologically, they displayed conduction

Imaging Findings Etiology and Pathogenesis Role of Ganglioside Antibodies Role of Campylobacter jejuni/coli Infection Treatment Treatment Recommendations

block in mixed nerves involving only the motor component of the nerve, with normal sensory conduction through the blocked motor segment. The authors postulated that they were seeing an unusual form of CIDP in which the demyelination was not only focal but also confined to motor fibers. The patients had normal cerebrospinal fluid (CSF) protein levels, and there was no clear involvement of sensory fibers either proximally or distally. Pestronk and colleagues96 published findings in two patients with clinical and electrophysiologic features very similar to those presented by Parry and Clarke. In both cases, sural nerve biopsies were performed and were both normal. In one case the patient had a motor point biopsy, in which muscle is biopsied at the point of insertion of the motor nerve. Staining revealed demyelination of the surviving axons. These patients were interesting with respect to two other features. First, both patients failed to respond to treatment with high-dose corticosteroids and plasma exchange, the standard therapies for CIDP at this time. However, both patients subsequently responded to intravenous immune globulin (IVIG), followed by oral cyclophosphamide, and made good recoveries. Second, these patients were found to have serum immunoglobulin G (IgG) and immunoglobulin M (IgM) antibodies that reacted with ganglioside epitopes, particularly GM1. 2277

2278

Diseases of the Peripheral Nervous System

It should be noted that, in 1982, prior to these publications, Lewis, Sumner, and colleagues had described an interesting group of five patients with a predominantly motor demyelinating neuropathy.62 The Lewis et al. paper raised particular interest because of the unusual presence of persistent motor conduction block on NCS. Some opinion has viewed this publication as the first description of MMN-CB, but there are important distinctions. All five patients had some degree of sensory involvement, from mild to moderate, two had optic neuritis, and all had widespread and significant demyelination on NCS. Sural nerve biopsies carried out in three of these patients showed evidence of demyelination and remyelination—pathology suggestive of CIDP. These patients also responded to the use of corticosteroids. Many of these disease features are now viewed as inconsistent with the diagnosis of MMN-CB. Most authorities now view the entity described by Lewis et al. as representing a variant of CIDP, or as a separate entity eponymously referred to as the Lewis Sumner syndrome, also termed multifocal acquired demyelinating sensory and motor (MADSAM) neuropathy,61,73–75,105,106,127 although there is yet to be an absolute consensus.66,86 Conduction block had been a well-recognized feature of compressive neuropathies, such as carpal tunnel syndrome, and also a feature of the early stages of acute inflammatory demyelinating polyneuropathy, the demyelinating form of Guillain-Barré syndrome (GBS).14,34 Conduction block had also been induced experimentally and widely studied by the use of a tourniquet. The fact that Lewis et al.’s patients had persistent conduction block that remained fairly constant for several years was unusual because in most patients with GBS or after decompression of an entrapment neuropathy, conduction block will start to resolve within 2 weeks as remyelination takes place.14,34 Therefore, it was suggested that some unknown factor was maintaining the conduction block by preventing remyelination. Because of the inability to demonstrate pathology, there was an assumption that the conduction block was due to demyelination because demyelination was the underlying feature of the previously described conduction block neuropathies, both clinical and experimental. This landmark study by Lewis and colleagues ignited a large amount of interest in the subject of conduction block neuropathies outside the field of GBS and foresaw the publication of the original case reports of MMN-CB neuropathy, and the subsequent widespread awareness of conduction block as an important feature of chronic neuropathies. Since the reports by Parry and Clarke88 and Pestronk et al.96 in 1988, several hundred cases of MMN-CB have been published, and MMN-CB is now well recognized as a distinct clinical entity.

CLINICAL FEATURES MMN-CB is an uncommon neuropathy, although the precise incidence is unknown. It has been estimated that 10 cases of motor neuron disease (MND) are seen for every case of MMN-CB at major tertiary referral centers.21 Because of the generally slow progression of the condition, which may extend over 30 years, the prevalence within the community is much higher than the incidence, and has been estimated at between 1 and 2 cases per 100,000 population.70 Recently several intermediate-size series of patients have been published,5,12,50,57,116 along with one excellent and detailed review analyzing the entire literature.70 Consequently, a clear clinical and electrophysiologic syndrome can now be defined through analysis of multiple cases. Generally MMN-CB has its onset between the ages of 20 and 40, with only 20% of cases arising over the age of 60. The oldest and youngest ages of onset reported have been 15 years and 72 years, respectively. Men are affected more often than women, with a reported male-to-female ratio of 2.6:1. The majority of cases follow a slowly progressive course, although more rapid onset has been reported.116 Likewise, stepwise progression has been noted occasionally,72 as has spontaneous improvement.12,19 Death directly attributed to MMN-CB is uncommon, with only seven deaths reported, of which two were directly from neuropathy,12,19,63,89 one from complications of therapy,22 and four from concomitant disease.7,9,10,18 Whether any of the reported cases were misdiagnosed cases of MND is conjectural. There has been one description of MMN-CB complicated by fatal secondary amyloidosis.10 The clinical characteristics of MMN-CB are of a pure motor neuropathy that affects multiple peripheral nerves in the limbs (multiple motor mononeuropathies), plexus, and roots. There is a predilection for the upper limbs, particularly the nerves innervating the forearm and intrinsic hand muscles. Thus the principle clinical feature that characterizes all cases of MMN-CB is lower motor neuron weakness that is asymmetrical at onset in the vast majority of cases. It is often confined to the distribution of one peripheral nerve at the onset, often erroneously resulting in the diagnosis of peripheral nerve entrapment. Sites outside the limbs can be rarely affected. Three cases of involvement of the phrenic nerve have been reported.9,11,63 Similarly, involvement of the cranial nerves is uncommon and has been reported only twice.47,98 However, neither finding has been noted in any of the moderate-size series published.57,116 Autonomic involvement has not been reported. Sensory involvement is minimal, and any significant clinical or electrophysiologic sensory abnormalities should raise major diagnostic doubt. Similarly, upper motor neuron signs comprising brisk reflexes, extensor plantar responses, or spasticity are not

2279

Multifocal Motor Neuropathy and Conduction Block

seen. Reflex asymmetry is common, however, with depressed reflexes at affected sites and normal reflexes elsewhere.116 Wasting is commonly seen, particularly as the illness progresses. However, in the early stages muscle bulk may appear normal even in severely weakened muscles. In rare cases hypertrophy may occur,12,78 and this has been attributed to continuous muscle activity arising through axonal hyperexcitability at sites of demyelination.12,78 Fasciculations and cramps are often seen, particularly following exercise of the affected muscle, and postactivity weakness may be marked. The presence of these features often raises concern as regards the diagnosis of MND.33,88 Pestronk et al.93 reviewed a group of 74 patients with asymmetrical progressive lower motor neuron syndromes. They were accrued using clinical and electrophysiologic criteria and the antiganglioside antibody profile of patients referred for testing. The authors were able to define three subgroups of patients. Of their 74 patients, one group of 25 cases comprised patients with MMN-CB with high titers of IgM anti-GM1 ganglioside antibodies and persistent motor conduction block. This group of 25 patients all had evidence of selective motor conduction block in one or more nerves; 32% were female, 68% had their disease onset before the age of 45, and all had more weakness in distal muscles than in proximal muscles, with the hands predominantly involved in 21 of 25 cases. Reflex loss outside the area of clinical weakness occurred at some time during the disease course in 59% of the patients. The other 49 patients had evidence of axonal loss as their primary abnormality, with 7 patients also having some evidence of demyelination. Bouche et al.12 presented a detailed assessment of 24 patients with MMN-CB and again confirmed previous findings of a markedly motor-predominant multiple mononeuropathy, principally affecting the upper limbs, with a male predominance and onset usually before the age of 40 years. Taylor et al.116 published a consecutive series of 46 patients with MMN-CB and confirmed the typical clinical phenotype of slowly progressive upper limb–predominant motor multiple mononeuropathies. There have been numerous other reported small series of patients with MMN-CB, most of which were published with the intent of describing the response to treatment using IVIG. These reports have often included patients with significant evidence of demyelination, and in many cases moderate sensory involvement that raises the possibility that many of these patients actually had CIDP or MADSAM neuropathy. This underlies the importance of recognizing the boundaries of the syndrome, with some specialists enforcing very strict criteria and others allowing more latitude, particularly with respect to the extent of sensory features. The clinical features of MMN-CB are summarized in Table 101–1. The clinical criteria for diagnosis are shown in Table 101–2. The typical appearance of the arm of a patient with established MMN-CB is shown in Figure 101–1.

Table 101–1. Physical Signs and Clinical Symptoms Seen in MMN-CB Cases at Diagnosis (Taylor et al.) and at Time of Reporting (Nobile-Orazio) Taylor et al. (N ⫽ 46)

Nobile-Orazio* (N ⫽ 294)

Symptom Weakness Wasting Cramps Fasciculations Sensory symptoms Pain Autonomic symptoms Cranial nerve involvement

100% 44% 39% 24% 11% 0% 0% 0%

100% NR 55% 58% 20% NR NR NR

Site of Onset Asymmetrical Onset in 1 limb Onset in arms Onset in hands Onset in legs Distal Proximal Distal & proximal

78% 78% 80% 68% 17% 98% 2% 0%

94% NR 79% NR 9% 87% 5% 8%

100% 98% 15%†

100% 86% 20%

0% 43% 57%

8% 20% 72%

100% 0%

NR NR

Physical Signs Weakness Atrophy Sensory findings Reflexes‡ Brisk Normal Reduced or absent Plantar responses Flexor Extensor

NR ⫽ not reported. * The figures quoted by Nobile-Orazio include the figures from Taylor et al. † Minimal sensory findings only. ‡ At sites of weakness. Data from Taylor, B. V., Wright, R. A., Harper, C. M., and Dyck, P. J.: Natural history of 46 patients with multifocal motor neuropathy with conduction block. Muscle Nerve 23:900, 2000; and Nobile-Orazio, E.: Multifocal motor neuropathy. J. Neuroimmunol. 115:4, 2001.

DIFFERENTIAL DIAGNOSIS MMN-CB, when well established, is a unique neuropathic process with a characteristic phenotype and clear electrophysiologic markers. The principle differentials include MADSAM neuropathy (Lewis Sumner syndrome),105,106 motor-predominant CIDP,58 monomelic MND,107 progressive muscular atrophy,30 multiple compressive neuropathies,8 and hereditary neuropathy with liability to pressure palsies.119

2280

Diseases of the Peripheral Nervous System

Table 101–2. Clinical and Laboratory Criteria for the Diagnosis of MMN-CB 1. A markedly motor-predominant, generally asymmetrical peripheral neuropathy or multiple mononeuropathies without upper motor neuron signs. Minor sensory symptoms are permitted but overt sensorimotor neuropathies are excluded. The process principally affects the distal limbs, particularly the hands, and is frequently asymmetrical. 2. Other causes of motor-predominant neuropathy must be excluded on clinical, electrophysiologic, and laboratory criteria, particularly inherited motor neuropathies or neuronopathies, hereditary neuropathy with liability to pressure palsies, motor neuron disease, MADSAM neuropathy (Lewis Sumner syndrome), and chronic or acute inflammatory demyelinating polyneuropathies. 3. High-titer antibodies to ganglioside epitopes, particularly IgM anti-GM1, are supportive of the diagnosis of MMN-CB when present, but do not exclude the diagnosis when absent. 4. CSF protein levels are normal or only mildly elevated. Data from Taylor, B. V., Wright, R. A., Harper, C. M., and Dyck, P. J.: Natural history of 46 patients with multifocal motor neuropathy with conduction block. Muscle Nerve 23:900, 2000; and Nobile-Orazio, E.: Multifocal motor neuropathy. J. Neuroimmunol. 115:4, 2001.

Careful clinical examination specifically looking for sensory involvement or upper motor neuron signs can eliminate many of the differentials. However, significant problems can arise with monomelic amyotrophy presentations of MND.107 The presence of high-titer IgM anti-GM1 antibodies can help in this situation, as can an objective response to therapy with IVIG. Distinguishing between MADSAM neuropathy, motor-predominant CIDP, and

MMN-CB can be important because of the possibility that MMN-CB patients will worsen with corticosteroid therapy or plasma exchange. An elevated CSF protein and the absence of high-titer IgM anti-GM1 antibodies may help, although these findings may not completely separate the conditions. Careful electrophysiologic assessment is by far the most useful tool in establishing the diagnosis of MMNCB, and repeating studies if unclear can be of significant benefit. The finding of focal changes on magnetic resonance imaging (MRI) of peripheral forearm nerves also has potential to aid in the diagnosis.

NATURAL HISTORY The natural history of MMN-CB is of a slowly progressive motor neuropathic process. Few patients have died from the condition, and the outlook for a normal lifespan in the absence of phrenic nerve dysfunction is good. Clearly most patients have inexorable progression over many years, with treatment having some effect but not curing the disease.116 The outlook for hand function is not as good. The striking predilection for the upper limbs and in particular the hands results in neurologic disability in the majority of patients. However, because of the asymmetrical nature of the disorder and its slow onset, most patients are not incapacitated and retain their ability to function, albeit with some degree of disability. Progression to generalized immobility in our experience is very rare.116

ELECTROPHYSIOLOGIC CHARACTERISTICS Clinically MMN-CB has certain features that allow its recognition as a unique motor neuropathy; however, it is the electrophysiologic characteristics of MMN-CB that define the condition as a separate entity. In particular, focal conduction block of motor fibers of mixed nerves is found at locations that are not common sites of compression, and normal sensory conduction studies are obtained through the sites of motor block.

Conduction Block

FIGURE 101–1 Appearance of the left arm of a 42-year-old man with typical MMN-CB.

Conduction block is generally recognized as constituting the electrophysiologic hallmark of MMN-CB. Conduction block is defined as a significant reduction in the evoked compound muscle action potential (CMAP) amplitude, or the area between distal and proximal sites of stimulation along a focal nerve segment, in the absence of abnormal temporal dispersion (Table 101–3; Figs. 101–2 and 101–3). Conduction blocks are well recognized in human neuropathies and are commonly seen with focal nerve

Multifocal Motor Neuropathy and Conduction Block

2281

Table 101–3. Electrophysiologic Criteria and Definition of Degrees of Partial Motor Conduction Block (PMCB) in MMN-CB Electrophysiologic Criteria for the Diagnosis of MMN-CB 1. At least one, and preferably two, areas of persisting motor conduction block should be present not at common sites of compression. Any grade of PMCB (see below) is acceptable, and other features are typical. Conduction block should not be assessed in isolation where the distal CMAP amplitude is less than 1 mV. Repeated studies over time may be required to demonstrate significant conduction block. 2. Diffuse demyelination is absent outside areas of PMCB. 3. Needle examination reveals evidence of chronic or more active partial denervation and reinnervation in muscles supplied by nerves with PMCB. In acute or new-onset conduction block, reduced recruitment may be the only abnormality seen. 4. Sensory conduction in mixed nerves through areas of PMCB should be normal. Criteria for Conduction Block (CB)*,† Definite CB A greater than 50% reduction in negative peak area or amplitude between distal and proximal stimulation sites for all motor nerves except the tibial. A ⬎ 75% reduction in negative peak area or amplitude between ankle and knee sites is required. OR For any motor nerve, a reduction in the negative peak area or amplitude of 20% or greater over a 10-cm or less segment of nerve, using sequential short-segment stimulation (the “inching” technique). Occurs in the absence of significant temporal dispersion. Probable CB A 30% to 50% reduction in negative peak area amplitude between distal and proximal stimulation sites in all motor nerves except the tibial nerve. A 50% to 75% reduction in negative peak area and/or amplitude between ankle and knee sites in the tibial nerve. Occurs in the absence of significant temporal dispersion. Possible CB A 20% to 30% reduction in negative peak area amplitude between distal and proximal stimulation sites of all motor nerves except the tibial nerve, in which possible conduction block cannot be assessed. * Abnormal temporal dispersion is defined as a ⬎ 25% increase in the negative peak duration of the proximally evoked CMAP, as compared to that evoked with distal stimulation. Borderline temporal dispersion is defined as a 15% to 25% increase in the negative peak duration of the proximally evoked CMAP, as compared to that evoked with distal stimulation. Less than a 15% increase in the negative peak duration of the proximally evoked CMAP is regarded as within normal limits. † Criteria used to suggest demyelination in segments of nerve not affected by CB: CMAP dispersion without CB, conduction velocity less than 70% of lower limit of normal with normal CMAP amplitude, conduction velocity less than 50% of lower limit of normal with CMAP amplitude 50% to 99% of normal, distal latency greater than 130% of normal, F-wave prolongation in excess of F estimates, or R1 blink reflex latency greater than 15 ms. Modified from Taylor, B. V., Wright, R. A., Harper, C. M., and Dyck, P. J.: Natural history of 46 patients with multifocal motor neuropathy with conduction block. Muscle Nerve 23:900, 2000.

compression, such as the median nerve within the carpal tunnel, the ulnar nerve at the cubital tunnel, or the peroneal nerve at the fibular head. Conduction block may also be seen in GBS, CIDP, and hereditary neuropathy with liability to pressure palsies. The pathologic basis of conduction block in human compressive neuropathies has been studied,34 and in all cases in which conduction block could be demonstrated electrophysiologically, subsequent nerve biopsies revealed up to 50% demyelination and remyelination, whereas patients with no conduction block had less than 5% demyelination and remyelination. The criteria used in these studies for definite conduction block were a greater than 20% reduction in amplitude with less than 15% temporal dispersion of the distal CMAP. Consequently, conduction block has traditionally been thought to be due to focal demyelination. Conduction block can vary significantly within patients and between patients, and also over time. Conduction block is often difficult to find electrophysiologically, and its detection depends on the skill of the electromyographer

and also on patient cooperation. For example, it is well recognized that conduction block can be difficult to define in certain nerve segments such as the tibial nerve, where a drop in amplitude of 50% between knee and ankle stimulating sites is considered within normal limits.29 Finding sites of proximal conduction block is also difficult and may result in the erroneous demonstration of proximal conduction block. This may be a particular problem in the brachial plexus, where submaximal stimulation at Erb’s point is common and more proximal stimulation is painful and generally poorly tolerated by patients. Needle stimulating electrodes are advocated as a means of overcoming this problem; however, their use is often difficult and requires the application of collision techniques, which are not in standard use.29 Rhee et al.99 described a computer simulation of conduction block that emphasized the importance of temporal dispersion in the recording of conduction block. They found that interphase cancellation between motor fibers conducting at different velocities could result in a measurement of up to 50% conduction block where no

2282

Diseases of the Peripheral Nervous System

5 ms

A 1

100 mA 1 mV

A1

320V 5 mV

A 2

100 mA 1 mV

A 3

98.1 mA 1 mV

A2

320V 5 mV

A 4

97.3 mA 1 mV Temp: 31.6°C

FIGURE 101–2 Motor nerve conduction study of the right radial nerve in a 52year-old man with typical MMN-CB, demonstrating definite conduction block of the motor fibers of the radial nerve between the axillary and supraclavicular stimulation sites. Note the absence of temporal dispersion. All sensory studies were normal. The left radial motor nerve study was normal and of similar amplitude and configuration; in particular, there was no amplitude drop across the same segment. (Courtesy of Dr. C. M. Harper.)

A3

320V 5 mV

A4

320V 5 mV

conduction block existed. Therefore, Rhee et al. emphasized the importance of measuring not only amplitude and area of evoked CMAPs, but also their duration to ensure that abnormal temporal dispersion does not occur. They regarded temporal dispersion of greater than 15% as being unacceptable; thus a CMAP generated from proximal stimulation should not have a duration greater than 1.15 times the duration of the distally evoked CMAP. Significant temporal dispersion is a frequent finding in patients with demyelinating neuropathies such as CIDP or MADSAM neuropathy.49 Conduction Block in MMN-CB Although widely and justifiably recognized as the gold standard for the electrophysiologic diagnosis of MMN-CB, conduction block represents, along with the role of ganglioside antibodies, one of the most difficult and controversial aspects of MMN-CB. Conduction block in MMN-CB has several unique features: 1. Pure motor fiber selectivity 2. Abrupt onset over focal nerve segments 3. Predilection for the distal segments of mixed nerves, particularly the forearm segments of the median and ulnar nerves 4. Lack of association with nerve compression 5. Pathologically, lack of association with focal demyelination (see Pathologic Changes below)

FIGURE 101–3 Median nerve motor conduction study in a 45-year-old man with typical MMN-CB, demonstrating definite conduction block within the forearm segment without temporal dispersion. Sensory studies of the median nerve through the same segment were normal.

One of the principle troubles is the definition of conduction block in MMN-CB, particularly what degree of decrease in amplitude or area of the CMAP represents true or definite conduction block, and what degree of temporal dispersion is acceptable. Exact criteria, when applied systematically, allow for the definition of a unique neuropathic disorder; when lesser degrees of certainty are used, the specificity falls significantly. Various authors have suggested criteria based on various degrees of conduction block and/or area decrease.12,55,70,84,97,116 Many patients with conditions such as MND, chronic axonal neuropathies, and motor and sensory neuropathies can have significant reductions in CMAP amplitude and alterations in distal and proximal CMAPs secondary to conduction delays and motor unit remodeling.97 Therefore, with inadequate NCS technique, including lack of supramaximal stimulation or lack of awareness of interphase cancellation, conduction block can be misdiagnosed.

Multifocal Motor Neuropathy and Conduction Block

Similarly, it is also extremely difficult to make a diagnosis of conduction block when the CMAP amplitude is less than 1 mV.97 Of all the techniques employed, the “inching” technique is probably the most useful in determining accurately whether a conduction block is present.55 The inching technique involves sequential short-segment stimulation of nerves in which the stimulating electrode is moved 1 to 2 cm along the tract of the nerve of interest, looking for an abrupt change in the distally evoked motor potential signifying the site of conduction block. Using meticulous electrophysiologic techniques such as inching, conduction block in MMN-CB can be localized to nerve segments 3 to 10 cm long in some cases, with significant focal conduction slowing with conduction velocities down to 10 m/s across these short segments. In these studies and others, sensory nerve conduction through segments of conduction block was assessed and found to not be different from that of control patients.55,84 Nobile-Orazio69 emphasized that many patients may be misdiagnosed as having conduction block because of the problem of interphase cancellation, first outlined by Rhee et al.,99 and the problem associated with submaximal proximal stimulation. They again supported the use of the inching technique and the finding of an abrupt change in CMAP amplitude over a short nerve segment, contrasting this finding with the gradual changes seen with demyelinating neuropathies or chronic axonal loss. Pfeiffer et al.97 examined the sensitivity and specificity of various criteria for conduction block and came to the conclusion that the variability seen in conditions in which motor unit restructuring and conduction delay were encountered meant that stricter criteria are required for definite diagnosis of conduction block. Particularly, the variability seen when the distal response amplitude fell below 1 mV increased significantly. They suggested that strict criteria allowing for amplitude of the distal response and the use of pooled data from healthy control and amyotrophic lateral sclerosis nerves as controls could provide sufficient sensitivity and specificity for conduction block in MMN-CB. They also suggested that amplitude criteria are more sensitive than area or duration criteria and recommended the use of a 50% amplitude criterion when the distal amplitude was greater than 1 mV and a 74% criterion when the distal amplitude was less than 1 mV. These figures specifically applied to the median, ulnar, and peroneal nerves and were most sensitive and specific for the forearm segment of the ulnar nerve. It has been suggested that early detection of conduction block (that does not meet the published criteria) may be useful in MMN-CB.15 By observing patients over long periods of time, Cappellari et al. found that conduction block was a very dynamic entity and that patients frequently progressed from having minimal conduction block to having significant conduction block over time. They

2283

went on to suggest less restrictive criteria for the diagnosis of conduction block.15 Similarly, Pakiam and Parry82 reported five patients with otherwise typical multifocal motor neuropathy but without overt conduction block, and found that these patients were not different in any other respect from patients with conduction block. Katz et al.48 described nine patients with what they termed axonal multifocal motor neuropathy without conduction block. Nobile-Orazio et al.71 demonstrated in 23 patients with MMN that the degree of conduction block was not associated with any significant clinical or immunologic differences except for a minor decrease in the response rate to IVIG. These authors concluded that failure to fulfill strict criteria for conduction block in patients with otherwise clinically typical MMN should not preclude the diagnosis or a treatment trial. Because of the widely varying criteria for defining conduction block, ranging from no block to greater than 50% reduction in amplitude, the American Association of Electrodiagnostic Medicine has published consensus criteria for conduction block,79 which were revised in 2003.80 These criteria are relatively strict, requiring a greater than 50% to 60% reduction of proximal versus distal CMAP amplitude without temporal dispersion for definite conduction block and a similar reduction in CMAP amplitude with increased temporal dispersion, or a 10% lesser reduction in CMAP amplitude with no temporal dispersion, for probable conduction block. Although definitely excluding spurious conduction block, these criteria are perhaps overly strict, and their applicability to routine clinical practice is difficult. A graded classification of certainty of conduction block, as suggested by Taylor et al.116 may be more useful in clinical practice because it allows for inclusion of less significant areas of conduction block where temporal dispersion is not encountered (see Table 101–3). Generally sites of conduction block follow the areas of clinical involvement and are more common in the upper limbs, particularly the forearm segments of the median and ulnar nerves.57,116 This may reflect that these sites are the easiest to study; however, the clinical findings in a significant majority of patients favor distal involvement of the upper limb motor nerves. The reason for the striking distal upper limb predilection is not known, nor is the reason for the site of the conduction block within the forearm segments of the upper limb nerves. Whether the site of conduction block occurs at areas of weakness of the blood-nerve barrier or at watershed sites is not known. The minimum number of sites of definite conduction block required for the diagnosis of MMN-CB has also been debated. Some authors have required the presence of two sites for the diagnosis,116 whereas in clinical practice the presence of one symptomatic site of conduction block may be sufficient to make the diagnosis if other causes of focal nerve compression are excluded.37 It is clear that MMN-CB is a dynamic condition and that new areas of

2284

Diseases of the Peripheral Nervous System

conduction block occur even when patients are aggressively treated with immunosuppression.15,126 Consequently, the presence of definite conduction block, although considered the electrophysiologic gold standard for the diagnosis of MMN-CB, should not be considered mandatory for the diagnosis of the condition when other features point to the diagnosis and treatment rather than observation is warranted.

NCSs Outside of Areas of Conduction Block Observations and discussion in this area have caused considerable controversy. With the recognition that MADSAM neuropathy (Lewis Sumner syndrome) is almost certainly a separate entity, it now seems clear that, in MMN-CB, NCSs outside of areas of conduction block should be essentially normal. Some degree of minor abnormalities of motor conduction is not uncommon, particularly mildly prolonged distal motor latencies or mildly prolonged F-wave latencies in nerves with motor conduction block. This is not out of keeping with the significant axonal loss and subsequent motor unit remodeling seen with decreased CMAP amplitudes in long-standing MMNCB. Similar minor changes can be seen in patients with MND. All sensory NCSs should be normal in MMN-CB, or at most minimally abnormal.70,116 The presence of demyelination outside areas of conduction block in MMN-CB is controversial. Many patients described in case series of MMN-CB have had significant evidence of demyelination outside of areas of conduction block along with significant temporal dispersion, and may not have MMN-CB as defined by stricter electrophysiologic criteria.25,50 Significant abnormalities on NCSs, particularly of diffuse motor demyelination, should raise the possibility of CIDP or MADSAM neuropathy as the diagnosis. This variation in case definition may have influenced some of the treatment reports and ganglioside antibody studies. Clearly MADSAM neuropathy and CIDP respond more favorably to immunomodulation than does MMN-CB.49

Electromyographic Findings Needle examination of muscles supplied by nerves with persistent motor conduction block almost invariably display evidence of active acute and chronic partial denervation, with the presence of fibrillation potentials and positive sharp waves. Axonal irritability (fasciculations) is often seen, along with evidence for motor unit remodeling with large polyphasic units. In very early stages of acute conduction block, decreased recruitment or markedly reduced recruitment with only minimal evidence of denervation may be the only electromyographic signs.116 Continuous motor unit activity, comprising multiple fasciculation potentials at a frequency of up to 30/min, with

markedly decreased recruitment has been demonstrated in muscles supplied by nerves with significant conduction block. This continuous motor unit activity has resulted in hypertrophy of weak muscles.78 The finding of electromyographic evidence of denervation in muscles supplied by individual peripheral nerves or plexuses rather than myotomes or regions is a useful method of distinguishing MMN-CB from MND.49 Similarly, the paraspinal muscles are not generally involved in MMN-CB.116 When examining patients with suspected MMN-CB using needle electromyography, the finding of denervation in muscles supplied by nerves without overt conduction block on standard NCS should induce a careful search for conduction block in more inaccessible regions of the nerve, including the use of proximal stimulation techniques where possible.

Other Electrophysiologic Techniques The use of magnetic stimulation of spinal roots may be the only method of determining very proximal conduction blocks.17 However, this technique is difficult to perform and has not been sufficiently quantified to allow routine use. It is nevertheless better tolerated than needle stimulation of spinal roots and certainly merits further investigation. Transcranial magnetic stimulation (TMS) has also been suggested as being useful in MMN-CB68; however, the diagnostic uncertainty of the patients studied in this report would tend to mitigate against the use of TMS as a routine diagnostic tool at this stage. TMS is generally well tolerated and may be a useful tool in the future.

LABORATORY FINDINGS Antibodies to Ganglioside Epitopes in MMN-CB One of several major landmarks in understanding MMNCB was the discovery by Pestronk and colleagues96 of antiGM1 IgM antibodies in two cases. This finding provided evidence for an autoimmune basis for MMN-CB and opened a large area of research into the relationship between anti-GM1 antibodies and motor system diseases, and the role of antiganglioside antibodies in disease pathogenesis. The background to this field is complex and at times controversial, as described below. It is now recognized that anti-GM1 antibodies are present in roughly 50% of typical cases of MMN-CB, and the assay is widely used in clinical practice for screening suspected cases. The role of antibodies directed to ganglioside epitopes in the diagnosis and pathogenesis of MMN-CB and other immune-based neuropathies has been a source of controversy for more than a decade. The issue remains far from

Multifocal Motor Neuropathy and Conduction Block

clear, and more research is required to better understand the complex nature of the interactions between ganglioside antibodies and the peripheral nervous system. Gangliosides are complex glycosphingolipids that are highly enriched in both peripheral and central nervous tissue. They are discussed in detail in Chapter 25. Glycosphingolipids are composed of the long-chain aliphatic amine sphingosine (acylated ceramide) attached to one or more sugar moieties. The ceramide portion is inserted in the membrane lipid bilayer, with the carbohydrate structure exposed extracellularly. Complex glycosphingolipids by definition must contain at least one sialic acid residue. Sialic acid is a generic term for N-acylneuraminic acid, or NeuNAc. The sialic acids in gangliosides are usually attached to either the terminal or the internal galactose of an oligosaccharide core composed of one to five sugars. Traditionally, gangliosides are named according to the convention of Svennerholm,45,110 with the following formula: G refers to “ganglio” and M, D, T, or Q refer to the number of sialic acid residues (one [mono], two [di], three [tri], and four [quad], respectively). Arabic numerals and lower case letters refer to the sequence of migration as determined by thin-layer chromatography. Bulky gangliosides with a longer oligosaccharide core containing more sialic acid moieties migrate more slowly than smaller gangliosides. For example, the monosialoganglioside GM1, with a four-sugar oligosaccharide core, migrates more slowly than the monosialoganglioside GM2, with a three-sugar oligosaccharide core. In MMN-CB, most interest has centered on antibodies that react with epitopes on the carbohydrate region of the ganglioside GM1. Because these carbohydrate structures are often present on several different glycolipids, glycoproteins, and other carbohydrate-rich molecules such as bacterial lipopolysaccharides, there is considerable scope for shared reactivity. Of note, the terminal Gal(beta 1–3)GalNAc epitope that is found in the GM1 molecule is also present on asialo-GM1 and GD1b. The complete GM1 carbohydrate sequence is also present in the lipopolysaccharide of the gram-negative bacterium Campylobacter jejuni.140 Prior infection with Campylobacter species has been implicated in the development of GBS,140 as discussed in Chapter 28 and elsewhere. Antiglycolipid antibodies or antiganglioside antibodies have been implicated in a number of peripheral nervous system disorders, including GBS.138 Braun et al.13 and Steck et al.108 recognized the fact that myelin-associated glycoprotein was an antigen for monoclonal IgM in patients with demyelinating polyneuropathies. This crossreactivity produced a benign late-onset demyelinating neuropathy, often associated with tremor.139 Chronic large fiber sensory neuropathy with ataxia has been linked to monoclonal IgM antibodies to gangliosides containing the sequence NeuAc(alpha 2–8)NeuAc.134,138

2285

With respect to patients with motor syndromes, antiganglioside antibodies were first identified as having specificity for serum IgM monoclonal proteins seen in rare patients with lower motor neuron syndromes.40,56 The recognition by Pestronk et al.96 that patients with a welldefined “lower motor neuron syndrome” with conduction block also had high levels of antiganglioside antibodies raised much interest. Previously, Pestronk et al.92 had demonstrated antibodies against GM1 ganglioside in 30% of patients with MND. The antibodies found were particularly prominent in patients with lower motor neuron signs. They did not find significant levels of ganglioside antibodies in normal controls or in patients with motor and sensory neuropathies. Latov et al.56 and Freddo et al.40 demonstrated similar findings in patients with MND. However, the patients who were positive for anti-GM1 antibodies also tended to have the lower motor neuron forms of MND. Subsequent to Pestronk et al.’s report of MMN-CB associated with anti-GM1 ganglioside antibodies,96 the testing of serum samples using enzyme-linked immunosorbent assays (ELISAs) became standard practice in patients with lower motor neuron syndromes. However, as larger numbers of patients were tested for GM1 ganglioside antibodies, it became clear that the antibodies lacked specificity and were found at detectable titers in neurologically normal patients, patients with MND, and also patients with sensory and motor neuropathy. Nevertheless, patients with clinically and electrophysiologically definite MMN-CB have a clear excess of IgM anti-GM1 antibodies in around 50% of cases.2,3,53,54,67,91,93–95,102,115,131,132 Three main patterns of anti-GM1 antibody specificity have been identified in MMN-CB. First, antibodies may react with the terminal Gal(beta 1–3)GalNAc structure and will therefore cross-react with GD1b and asialo-GM1. Second, antibodies may react with GM1 and GM2 via shared internal sialylated sugar moieties.77 Third, some antibodies are monospecific to GM1.76 The fine specificities of various gangliosides are demonstrated on chromatography in Figure 101–4. Although important efforts have been made to correlate these variations in fine specificity with clinical subgroups, no unifying patterns emerged. The sensitivity of IgM anti-GM1 antibodies for MMN-CB varies in these patient groups from 20% to 80%.60,94,131 However, when high titers of IgM anti-GM1 antibodies are defined, the specificity for potentially treatable neuropathies can exceed 90%.16,115,131 Pestronk and Choski94 reported significantly higher sensitivity and specificity for covalently linked GM1 ELISAs using a highly specialized technique. However, this finding was not confirmed by other groups,16 and the technique has proved difficult to use routinely. The significant variability and difficulty in performing ganglioside assays has led to attempts to standardize the methodology and to make available standard positive

2286

Diseases of the Peripheral Nervous System

appropriate, considering that in general the specificity of a high positive titer of IgM anti-GM1 antibodies for a potentially treatable neuropathy can exceed 90%. However, when the result is borderline or weakly positive, clinical judgment becomes important, and consideration should be given to a trial of treatment with clear understanding by the patient and the treating physician of the potential benefits, objective measures of progression, and when to cease or continue treatment. In rare cases in our experience, serial measurements of antibodies have demonstrated a change from negative to positive titers over time in clinically definite MMN-CB patients who initially tested negative, and thus there may be occasional grounds for repeat measurement.

Other Laboratory Markers

FIGURE 101–4 Thin-layer chromatography overlay of a panel of isolated gangliosides (from the top downward: GA1, GM1, GD1a, GD1b, GT1b, and GQ1b, whose migration positions on the chromatogram are indicated by the horizontal lines) immunostained with a human anti-GM1 IgM monoclonal antibody cloned from peripheral blood lymphocytes of a patient with MMN-CB. This antibody binds equally to GM1, GA1, and GD1b by virtue of their shared terminal Gal(beta1–3)GalNAc epitope, but does not bind to the other gangliosides that lack this terminal disaccharide structure. This pattern of binding is typically seen in the sera of patients with MMN-CB.

control sera.137 However, controversy still remains as to the role of measurement of antiganglioside antibody levels in the clinical setting. It is clear that elevated levels of antiganglioside antibodies can occasionally be found in patients with MND and in healthy controls. Similarly, up to 50% of patients with clinically and electrophysiologically definite MMN-CB can be negative for all classes of ganglioside antibodies. Therefore, relying solely on antiganglioside antibodies to establish a diagnosis of MMN-CB in difficult cases is not indicated. Many clinicians will be swayed by a clearly positive antiGM1 ganglioside antibody result to attempt treatment of cases in which there is either clinical or electrophysiologic doubt about a diagnosis of MND. This would seem

MMN-CB patients typically have normal parameters on all standard blood tests, including markers of inflammation and vasculitis. The creatine kinase level may be mildly elevated in a proportion of cases consistent with mild to moderate denervation23; however, significant or persisting elevation of creatine kinase levels should raise significant diagnostic doubt. CSF examination should reveal acellular fluid and mild to moderate elevation of the CSF protein in up to one third of cases.70,116 Marked elevation of the CSF protein is more in keeping with a diagnosis of CIDP or MADSAM neuropathy.

PATHOLOGIC CHANGES The pathologic basis of MMN-CB is not well understood. The widely accepted dogma is that conduction block represents focal demyelination. However, this has never been convincingly displayed by any pathologic study. Kaji et al.46 reported the findings of a pectoral nerve biopsy taken at exploration for a large supraclavicular mass in a patient with conduction block neuropathy across the brachial plexus on that side. This showed evidence of sparse demyelination with onion bulb formation. However, in this patient, the actual pathology was located in the brachial plexus mass, and this was not reported. This case is strikingly similar to the report by Cusimano et al.28 of two patients described as having hypertrophic brachial plexus neuritis or hypertrophic brachial plexopathy. A motor point biopsy of one case96 demonstrated demyelination of motor axons as they left the intramuscular nerves, as well as axonal degeneration. A sural nerve biopsy in this patient also revealed minimal changes of degeneration. Corse et al.27 reported very minimal changes of demyelination on sural nerve biopsies in patients with MMN-CB; however, because the sural nerve is neither clinically or electrophysiologically involved in MMN-CB, the utility of this observation is unclear.

Multifocal Motor Neuropathy and Conduction Block

2287

Bouche et al.12 reported changes in sural nerve and superficial radial nerve biopsies in 12 patients, finding only minimal changes of axonal de- and regeneration with some cluster formation and occasional thinly myelinated axons. There was no convincing evidence of demyelination in this series of patients. A major deficiency of these studies is that they have not addressed the principal problem of conduction block of motor fibers of mixed nerves. Recent work by Taylor et al.117 on the pathology of typical patients with MMN-CB at the site of conduction block of mixed upper limb nerves has added significantly to the understanding of the pathologic changes of MMN-CB. Eight fascicular biopsies were taken from seven typical patients with MMN-CB using intraoperative NCS to localize the site of conduction block in mixed upper limb nerves and biopsy blocked fascicles. The principle findings of this study were 1. Active axonal de- and regeneration leading to multifocal fiber loss in more severe cases (Figs. 101–5 and 101–6) 2. No evidence of de- or remyelination or hypertrophic changes consistent with CIDP 3. Minimal evidence of inflammation in two cases Nerve morphometry performed on all nerves displayed a shift to the left in the fiber histograms, with loss of large myelinated fibers and increased cluster formation leading to increased numbers of small, thinly myelinated fibers. Degenerating axonal profiles were also increased. The principle concern of Taylor et al. was that no adequate controls exist for this type of pathologic study because motor nerves are rarely biopsied in health and mixed motor nerves have not been studied in humans. However, the authors believe that the changes seen were clearly abnormal based on their knowledge of other nerve pathology. These pathologic findings do significantly aid in the understanding of the pathogenesis of MMN-CB (see Etiology and Pathogenesis below).

FIGURE 101–5 Paraphenylenediamine-stained epoxy section of two fascicles of the left ulnar nerve of a 47-year-old man with long-standing typical MMN-CB. The fascicles were biopsied using intraoperative NCS to define blocked fascicles at the site of conduction block within the left forearm. The fascicles display striking multifocal fiber loss with prominent clusters of regenerating fibers and degenerating profiles. There are no hypertrophic changes or features of demyelination. These findings are consistent with prominent axonal de- and regeneration in a multifocal pattern.

in inaccessible nerve regions may significantly help in the diagnosis of conduction block neuropathy. Further work in this area is required.

IMAGING FINDINGS ETIOLOGY AND PATHOGENESIS MRI of the brachial plexus has been reported in several cases of proximal conduction block.85,129,130 In some but not all cases, high signal on T2-weighted images pre- and postcontrast was seen in the brachial plexus in a focal manner. Similar findings have been seen in CIDP, but not in lower MND.130 Kaji et al.46 described focal MRI changes of the median nerve in the forearm at the site of conduction block. These findings may have more general applicability because MRI is a noninvasive technique and therefore may aid in the diagnostic process. There is no currently widely accepted role for MRI in MMN-CB, although the detection of focal sites of nerve abnormality

The etiology and pathogenesis of MMN-CB have been addressed in a large number of studies published over the last 10 years but remain unresolved. One barrier to a comprehensive understanding of the syndrome is that the underlying pathologic abnormality has not been well characterized until recently. The neuropathy runs an indolent course, rarely justifying motor nerve biopsy, and pathologic material is therefore scarce. Despite this, it is widely believed that MMN-CB is an autoimmune disorder, principally based on the presence of high-titer anti-GM1 serum antibodies in a proportion of cases, and a response to

2288

Diseases of the Peripheral Nervous System

5000

5000

MF–

No/1 um diameter/mm2

No/1 um diameter/mm2

MF

0

0 0

4

8 12 Diameter (um)

16

20

0

4

8 12 Diameter (um)

16

20

FIGURE 101–6 Paraphenylenediamine-stained epoxy sections and nerve morphometry of two patients with typical MMN-CB. In both cases, fascicular biopsies were performed of blocked fascicles of forearm nerves at the site of conduction block using intraoperative NCS. Nerve morphometry was performed on the nerves using an interactive system for nerve morphometry. Left, Nerve biopsy from the left ulnar nerve of a 64-year-old woman with definite MMN-CB who responded moderately well to treatment and had a disease duration of 6 years. The biopsy is clearly abnormal, with increased axonal de- and regeneration and a shift to the left in nerve morphometry consistent with a loss of large myelinated fibers. Right, Nerve biopsy from the left ulnar nerve of a 29-year-old man who had a dramatic response to therapy. The nerve is essentially normal, with a normal appearance on morphometry. In both cases there is no evidence of deor remyelination.

immunomodulating therapy with IVIG and/or cyclophosphamide.

Role of Ganglioside Antibodies The analysis of anti-GM1 antibodies in the pathogenesis of MMN-CB has yielded much information, but their specific role remains unproven. When both spinal motor neurons and spinal cord are exposed to antiganglioside antibodies, they tend to bind to the motor neuron and also diffusely to the gray matter of the spinal cord, but not to dorsal root ganglia.26 This pattern of binding of GM1 antibodies selectively to motor neurons may explain the specificity of MMN-CB for motor neurons, as opposed to sensory neurons. However, defining the pathogenic role of the antiGM1 antibodies has been somewhat more difficult. Santoro et al.103 demonstrated IgM binding at the nodes of Ranvier in the sural nerve biopsy of a patient with a lower motor neuron syndrome associated with conduction

block and IgM anti-GM1 antibodies. Serum IgM from the patient was injected into rat sciatic nerve and also bound to the nodes of Ranvier. The binding activity was removed by preincubation with GM1, suggesting that the GM1 antibodies may cause motor dysfunction by binding at the nodes of Ranvier. Subsequently, Santoro et al.104 found that injection of serum from a patient with high titers of IgM anti-GM1 antibodies into the sciatic nerve of a rat, along with fresh human complement, produced conduction block with temporal dispersion and deposits of immunoglobulin at the nodes of Ranvier. Demyelination was noted in 6.5% of the fibers on electron microscopic analysis and, again, preabsorption with GM1 abolished the conduction block. Of note, the authors found that the conduction block induced in the rat sciatic nerve was greatest at 3 days, whereas the morphologic changes, particularly of moderate demyelination, were not present or maximal until day 7. They suggested that minor morphologic or structural

Multifocal Motor Neuropathy and Conduction Block

changes in nodal regions are sufficient to block saltatory conduction. Using an isolated rat sciatic nerve model, Arasaki et al.4 noted that acute conduction block could be induced within hours of application of high-titer rabbit or human antiGM1 antibody sera. The effect was only seen on myelinated fibers, with unmyelinated nerve conduction remaining normal. However, they reported no pathologic or immune binding data. Uncini et al.121 demonstrated, again using a rat tibial nerve injection model, that ganglioside antibodies from a patient with MMN-CB induced a decrease in amplitude of the CMAP and dispersion, whereas sera from patients with progressive muscle atrophy and high titers of anti-GM1 antibodies did not induce any changes. They suggested that the fine differences in anti-GM1 antibody affinity combined with an ability to fix complement might be important. Additionally, they recognized the fact that another circulating factor not present in the progressive muscular atrophy patients, but present in the MMN-CB patients, may be important. Takigawa et al.,111 using a petroleum jelly voltage clamp technique on isolated single myelinated rat nerve fibers and anti-GM1 antisera obtained from immunized rabbits, were able to increase both the rate of rise and the amplitude of the K⫹ current elicited by step depolarization, with little effect on the sodium current in the absence of complement. When complement was added, anti-GM1 antibodies decreased the sodium current and caused a progressive increase of nonspecific leakage current. Complement alone did not produce these changes. They concluded that anti-GM1 antibodies may be able to uncover potassium channels in the paranodal regions, while GM1 antibodies in the presence of complement may form antibody complexes that block sodium channels and disrupt the membrane at the nodes of Ranvier. Roberts et al.100 demonstrated, in an isolated mouse phrenic nerve–diaphragm preparation, that direct application of plasma or serum from patients with MMN-CB could induce complete block of threshold potentials. They also noted that passive transfer of serum into the peritoneal cavity of rats would produce a similar but less marked effect over approximately 5 days. This effect was not seen with all patients’ serum. They concluded that serum factors in MMN-CB could block conduction in motor nerves in the mouse. Of note, two of the patients whose sera induced block did not have detectable titers of anti-GM1 antibodies, which raises the possibility that other factors within the serum are important, or that the detection methods for such antibodies are imperfect. Additionally, the same authors have reported that cloned antiganglioside antibodies from two of their patients blocked distal motor nerve conduction on direct application using the same technique.136 Conversely to the previously discussed studies, Harvey et al.41 were unable to demonstrate abnormalities using a

2289

rat nerve injection model. Immunoglobulin fractions with high anti-GM1, IgG, and IgM titers were prepared from patients with GBS and also MMN-CB. These fractions were injected intraneurally into rat tibial nerve with fresh human complement. Neither the anti-GM1 from MMNCB patients nor the IgG from GBS patients induced significant focal conduction block or slowing, compared to a pooled fraction prepared from five normal individuals. By way of a control, when the authors used rabbit experimental allergic neuritis serum, significant conduction block was found. Pathologically, no abnormalities were found on transverse sections of injected nerves, and in particular there was no evidence of demyelination. Harvey et al. were, however, able to demonstrate by cryostat injections the presence of anti-GM1 antibodies bound to the nodes of Ranvier for up to 6 days following intraneural transfer. They concluded that they were unable to corroborate previous reports demonstrating positive effects using this model. Hirota et al.42 looked at the physiologic effect of antiGM1 antibodies on the saltatory conduction of transmembrane currents in single motor axons using a single isolated motor axon preparation. Utilizing 140 single nerve fibers from 22 rat ventral roots and external longitudinal current measurements, they observed saltatory conduction for 4 to 12 hours after the application of high-titer anti-GM1 serum from patients with GBS or MMN-CB. In addition, antiGM1 rabbit serum was also used. By way of positive controls, they used antigalactocerebroside (anti-GalC), which caused acute demyelination and conduction block, and tetrodotoxin (TTX), a sodium channel blocker. These two agents produced conduction block in 82% of the fibers treated with anti-GalC and in 100% of those treated with TTX, but the authors found conduction block in only 2% of fibers treated with patients’ sera and in 5% of fibers treated with rabbit anti-GM1 sera. They concluded that physiologic action of the antibody alone is insufficient to explain clinically observed conduction block in human diseases. Paparounas et al.83 described the use of isolated unsheathed mouse sciatic nerve preparations to assess the effect of various antiganglioside antibody preparations on nerve conduction. They were unable to demonstrate any evidence of conduction block over the 4- to 6-hour time window of the experiment but were able to show clear binding of IgM to the nodes of Ranvier, with evidence for complement fixation. They hypothesized that the short time window of the experiment, along with the absence of circulating factors, may have contributed to the lack of development of conduction block. In summary, the physiologic, immunologic, and pathologic effects of anti-GM1 antibodies in either in vivo or ex vivo models of nerve injury have yet to produce convincing and universally accepted evidence that conduction block can arise. Thus there is evidence that anti-GM1 antibodies may be pathogenic, but equally well-performed studies

2290

Diseases of the Peripheral Nervous System

have shown no pathogenic role for them. The published data would, however, tend to support a pathogenic role for ganglioside antibodies with preliminary but not universally accepted evidence for an effect on saltatory conduction mediated through an effect on the nodes of Ranvier, rather than a morphologically visible demyelinating effect. This theory is supported by the pathologic changes noted by Taylor et al.117 at the site of motor conduction block. The finding by Kiernan et al.52 of hyperpolarization of the axon distal to the site of conduction block in MMN patients is also supportive of a functional block. They found focal hyperpolarization similar to that seen in axons released from ischemia and suggested that the distal hyperpolarization was due to focal depolarization at the site of conduction block resulting from altered K⫹ conductance at the nodal region. Kiernan et al. also suggested that the membrane depolarization could lead to accumulation of Na⫹ within the axon, followed by increased intracellular Ca2⫹ and axonal degeneration. Most neuromuscular physicians tend to believe that GM1 antibodies probably play some role in MMN-CB, whether they represent an epiphenomenon secondary to damage or are directly immunopathogenic to nerves. Further work is required on the nature of the antigenic targets of GM1 antibodies in peripheral nerves and also whether GM1 antibodies are the only humoral factor important in the pathogenesis of conduction block.

Role of Campylobacter jejuni/coli Infection The gram-negative bacterium Campylobacter has been implicated in the pathogenesis of GBS, Miller Fisher syndrome, and acute motor axonal neuropathy as described in Chapter 25.43,44,135,138 The mechanisms by which acute gastrointestinal infection with Campylobacter species can lead to human peripheral nervous system disease are being investigated in detail. Molecular mimicry between ganglioside epitopes and the bacterial lipopolysaccharides is the favored putative mechanism.140 In consideration of the suggested role for ganglioside antibodies in the pathogenesis of MMN-CB, there has been some interest in whether there is an association between Campylobacter infection and MMN-CB. There have been three case reports of acute conduction block neuropathy following Campylobacter infection1,109,133; two of these cases followed a monophasic course with acute onset in all four limbs, while the third case seemed more akin to MMN-CB.1 Additionally, there has been one report of elevated anti-Campylobacter antibodies in patients with MMN-CB114 that were not correlated with antiganglioside antibody levels. Conversely, Terenghi et al.118 found elevated titers of antiCampylobacter antibodies in only 1 of 20 patients studied. These findings raise the interesting issue of whether MMN-CB may be a chronic postinfectious neuropathy. Considerably more work is required in this field.

TREATMENT One of the principle reasons for the significant research and clinical interest in MMN-CB has been the awareness that a significant proportion of patients will respond to the use of immunomodulating therapy. The initial reports in 1988 by Pestronk et al.96 of a response to cyclophosphamide in two patients generated significant interest. Since then, there have been a number of reports dealing with various immunomodulating therapies for treatment of MMN-CB. Most interest has centered on the use of high-dose IVIG or the use of cyclophosphamide, both intravenous and oral, either as a sole treatment or in combination with IVIG. Corticosteroids have been tried in a number of patients with little success, and in some cases actually worsened the condition.31,36,124 Plasma exchange has been demonstrated to be beneficial in a small proportion of cases, but there have been reports of patients worsening during treatment with plasma exchange.17,25,32 Attempts at treatment with interferon beta-1a have shown some improvement in long-standing cases in which other therapies have been unsuccessful,64 but in other reports deterioration in a significant proportion of patients has been noted.128 CSF filtration and CSF immunoabsorption have not produced significant responses.38,39 Charles et al.20 produced the first case report of the use of IVIG in the treatment of MMN-CB, reporting a patient who, after 5 days of administration of 0.2 g/kg daily of IVIG, noted a striking improvement in muscle strength that lasted 6 to 7 weeks. Shortly afterward, Kermode et al.51 produced a second positive report of IVIG in the treatment of MMN-CB, documenting the results of the initial treatment of five patients. Over the next several years, multiple small trials describing the beneficial effects of IVIG as treatment for MMN-CB appeared.12,23,72 All these reports were of short-term treatment with IVIG in high dosages. Four small, short, double-blind trials of IVIG therapy have been published; all demonstrated a clear initial benefit with IVIG therapy.5,35,59,123 Federico et al.35 randomized 16 previously untreated patients to either IVIG, 0.4 g/kg daily for 5 days, or placebo in a double-blind crossover study. Among the 16 treated patients, there was a significant improvement in 9, mild or moderate improvement in 2, and no change in 5, for an overall response rate of 69% (56% significant). In this study, an improvement in conduction block was noted with IVIG therapy. Leger et al.59 described the use of IVIG in 19 patients in a double-blind, crossover, placebo-controlled study. Ten patients had never received IVIG and 9 had previously responded to IVIG. Of the 10 IVIG-naïve patients, 9 completed the study, with only 4 responding to IVIG and 2 to placebo. Eight of 9 of the previous responders again responded to IVIG. There were no significant differences on Medical Research Council grading or electrophysiologic parameters between placebo and IVIG groups, but significant differences on self-evaluation scores.

Multifocal Motor Neuropathy and Conduction Block

There have been four reports of longer term treatment of patients with IVIG.5,12,122,126 Azulay et al.5 reported the treatment with IVIG of 18 patients, followed for 9 to 48 months. They noted a clinical improvement in 67% of patients. However, they also noted that the response was transitory. Van den Berg et al.122 reported the long-term effect of IVIG in six MMN-CB patients followed for 24 to 48 months. One of the clearest findings was that treatment with IVIG produced an initial good response that lasted for up to 1 year in one patient, but generally lasted less than 12 weeks in the other five patients. During follow-up, despite regular weekly infusions of immunoglobulin, there was clear evidence of disease progression with loss of muscle power, development of new sites of conduction block, and worsening of some existing sites of conduction block. Van den Berg et al.126 updated these initial six patients and described the long-term effect of IVIG in a cohort of 11 patients followed for 4 to 8 years. These patients essentially responded to treatment but never returned to normal. Overall, there was a gradual decline in all parameters measured, again with the development of new areas of conduction block. These authors came to the conclusion that IVIG maintenance therapy has a beneficial effect on muscle strength and upper limb disability but may not prevent a decrease in muscle strength. They also concluded that IVIG did not affect established axon loss. This suggests that IVIG probably does not specifically alter the underlying pathologic process but may slow or partially control it. Bouche et al.12 also demonstrated this finding in 19 patients treated with IVIG. One third responded well to the use of IVIG. However, the response was not sustained and there was clear evidence of disease progression, with the development of new sites of conduction block, despite continuing therapy with IVIG. Severity of disease does not appear to correlate with a response to IVIG therapy. Some patients who have been symptomatic for many years may still respond to therapy, although complete remission is not possible because of established atrophy from axonal loss. This may indicate a persisting long-term immune process.125 Of the other immunosuppressive agents used, the only modality to have any real success has been the chemotherapeutic alkylating agent cyclophosphamide. Individual case reports have reported stabilization with azathioprine.12 However, the earliest reports of successful treatment for MMN-CB were with intravenous cyclophosphamide.96 Feldman et al.36 reported the use of an intensive regimen of intravenous cyclophosphamide, 3 g/m2 given over 8 days, followed by the institution of oral cyclophosphamide, 2 mg/kg, 1 month later, with a significant improvement in the patients studied. Tan et al.113 reported the benefit of cyclophosphamide, using 3 g/m2 intravenously followed by oral cyclophosphamide, and noted that 5 of the 11 patients with conduction block responded. Oral cyclophosphamide by itself has not been shown to be effective.47

2291

Recently Meucci et al.65 have advocated the use of a combination of IVIG and oral cyclophosphamide that allows for a gradual reduction in the frequency of the IVIG infusions. However, the long-term effectiveness of this therapy in modifying the underlying disease process is unknown. The use of cyclophosphamide is highly controversial in this condition because many of the patients are young and have a very slowly progressive disorder that is not life threatening and does not necessarily lead to significant impairment. Cyclophosphamide is associated with the development of secondary tumors (particularly transitional cell cancer of the bladder and lymphoma), hemorrhagic cystitis, and liver dysfunction and does expose the patient to the development of neutropenia.6,81,90 There has been one report of the development of acute myelogenous leukemia following prolonged intravenous cyclophosphamide therapy for MMN-CB.22 The use of oral or intravenous cyclophosphamide should always be accompanied by adequate pre- and postdosage hydration to prevent the development of hemorrhagic cystitis. The use of the bladder protective agent mesna is not advocated for the dosages of cyclophosphamide used to treat MMN-CB, provided there is no preexisting bladder condition and appropriate hydration is undertaken. A record of the total cumulative dosage of cyclophosphamide should be kept, because there is evidence that the risk of development of secondary tumors is greater once the total administered dosage reaches 85 g.112 Monitoring patients for the presence of microscopic nonglomerular hematuria may alert the clinician to those patients more likely to develop transitional cell cancers of the bladder. The Cochrane Collaboration has recently reviewed the use of immunosuppressive therapy in MMN-CB and came to the conclusion that there were no randomized controlled trials to indicate whether immunosuppressive agents are beneficial in MMN-CB.120

Treatment Recommendations All patients should receive an initial course of therapy with IVIG except in the unusual situation of a patient with established severe stable atrophy with no evidence of involvement of nerves elsewhere, in whom observation can be employed. The most efficacious dosage regimen for IVIG has not been established. Generally two regimens are used for initiating therapy that are probably of similar efficacy. The first involves the use of a loading dose of IVIG at 0.4 g/kg daily for 5 days, and the second the use of IVIG at 0.4 g/kg once weekly for 6 weeks. Generally the maximum clinical response is seen 3 weeks following administration in the first regimen and at the end of the 6-week cycle in the second. Typical response patterns as measured by myometry are shown in Figure 101–7. At this point most neuromuscular physicians will reassess the patient and make a decision as to whether therapy has been successful. Occasionally patients are seen in whom the initial loading dosage may produce a

2292

Diseases of the Peripheral Nervous System

FIGURE 101–7 The response to treatment with intravenous immune globulin (IVIG) in three typical patients with MMN-CB.

response lasting longer than 1 year.122 Maintenance therapy is almost invariably required, and the most appropriate maintenance regimen is controversial. Some authors advocate repeating the loading dosage of IVIG for 3 to 5 days

monthly or at other intervals depending on the patient’s response. Others prefer a decreasing regimen of weekly infusions beginning with 0.2 g/kg once weekly for 6 seconds and, if this maintains the patient, then continued weekly

Multifocal Motor Neuropathy and Conduction Block

administration with gradual reductions in dosage and lengthening of the interval until the required minimum maintenance dosage of IVIG is found. Because IVIG is in scarce supply in many countries, the second regimen is probably more useful because generally less IVIG is used. Most patients will progress despite appropriate therapy, with the development of new sites of conduction block and

increasing weakness and muscle atrophy. However, there is good evidence to suggest that the rate of decline is slowed and that some cases never return to a pretreatment nadir even after years of therapy.126 Intensifying the regimen of IVIG infusions does not appear to be of benefit. Oral cyclophosphamide can be used in combination with IVIG infusions at a dosage of 2 mg/kg lean body weight

Newly diagnosed MMN-CB See diagnostic criteria

Established non-progressive atrophy

Yes

Observe and consider therapy if new areas involved

No

Functionally impaired

No

Consider observation

Yes

Initiate therapy

Progression

Initiate therapy with intravenous immunoglobulin (IVIG) 0.4g/kg once weekly for 6 weeks and reassess within 1 week of end of therapy course. Or 0.4g/kg IV for 5 days reassess three weeks post infusion.

No improvement or progression

Objective improvement

Start maintenance IVIG regimen. Either reducing scale of weekly then 2nd weekly IVIG or regular monthly IVIG infusions depending on patient response. The minimum required dosage and time interval between infusions should be determined. Therapy may be required for long periods.

No

Progression

Yes

2293

Reassess diagnosis. If still consistent with MMN-CB and significant progressive illness consider use of other IVIG regime or addition or oral or IV cyclophosphamide.*

Consider adding cyclophosphamide to IVIG regimen. Either 2 mg/kg orally* daily or 10-15 mg/kg IV* monthly. Careful monitoring for response and side effects is essential. *Unproven potentially hazardous therapy

FIGURE 101–8 Treatment algorithm for MMN-CB. (Modified from Taylor, B. V.: The treatment of multifocal motor neuropathy with conduction block [MMN-CB]. In Noseworthy J. H. [ed.]: Neurological Therapeutics Principles and Practice. London, Martin Dunitz, p. 2111, 2003.)

2294

Diseases of the Peripheral Nervous System

orally daily, with oral hydration prior to the dosage. Treatment should be limited, if possible, to 1 year of therapy or a cumulative dose of less than 85 g of cyclophosphamide. The combination of intravenous cyclophosphamide given at monthly intervals at a dosage of 1 g/m2 of body surface area and IVIG has not been studied formally. However, it may have an advantage over regular oral cyclophosphamide because a lower overall dosage of cyclophosphamide is required. Patients treated with intravenous cyclophosphamide must be well hydrated (e.g., 2 to 3 L given 24 hours before and 48 hours after each infusion) to reduce the risk of hemorrhagic cystitis but do not require continued oral hydration between doses. In summary, all current data on treatment of MMN-CB are observational, and only four small double-blind trials exist. Clearly more information from large controlled, double-blind, multicenter trials is required. However, this may prove difficult because of the rareness of the condition. It is also clear that different and novel approaches to therapy are required because the current treatment options are not curative, and clear disease progression occurs despite seemingly adequate therapy. Additionally, cyclophosphamide has significant side effects, and there is no clear risk-benefit analysis for treatment with this medication. Consequently, treatment decisions need to be made on an individual basis. Careful monitoring of objective as well as subjective responses to therapy is highly desirable, particularly when the use of cyclophosphamide is considered. Fully informed consent is also required, particularly focusing on the risk of secondary tumors and potential mutagenesis of cyclophosphamide. As noted by the Cochrane Collaboration,120 there is currently no evidence that immunosuppressives of any type alter the course of MMN-CB. A flow chart for treatment of MMNCB is presented in Figure 101–8.

REFERENCES 1. Abbruzzese, M., Reni, L., Schenone, A., et al.: Multifocal motor neuropathy with conduction block after Campylobacter jejuni enteritis. Neurology 48:544, 1997. 2. Adams, D., Kuntzer, T., Burger, D., et al.: Predictive value of anti-GM1 ganglioside antibodies in neuromuscular diseases: a study of 180 sera. J. Neuroimmunol. 32:223, 1991. 3. Adams, D., Kuntzer, T., Steck, A. J., et al.: Motor conduction block and high titres of anti-GM1 ganglioside antibodies: pathological evidence of a motor neuropathy in a patient with lower motor neuron syndrome. J. Neurol. Neurosurg. Psychiatry 56:982, 1993. 4. Arasaki, K., Kusunoki, S., Kudo, N., and Kanazawa, I.: Acute conduction block in vitro following exposure to antiganglioside sera. Muscle Nerve 16:587, 1993. 5. Azulay, J. P., Rihet, P., Pouget, J., et al.: Long term follow up of multifocal motor neuropathy with conduction block under treatment. J. Neurol. Neurosurg. Psychiatry 62:391, 1997.

6. Baker, G. L., Kahl, L. E., Zee, B. C., et al.: Malignancy following treatment of rheumatoid arthritis with cyclophosphamide. Am. J. Med. 83:1, 1987. 7. Bentes, C., de Carvalho, M., Evangelista, T., and Sales-Luis, M. L.: Multifocal motor neuropathy mimicking motor neuron disease: nine cases. J. Neurol. Sci. 169:76, 1999. 8. Beydoun, S. R.: Multifocal motor neuropathy with conduction block misdiagnosed as multiple entrapment neuropathies. Muscle Nerve 21:813, 1998. 9. Beydoun, S. R., and Copeland, D.: Bilateral phrenic neuropathy as a presenting feature of multifocal motor neuropathy with conduction block. Muscle Nerve 23:556, 2000. 10. Beydoun, S. R., Rison, R. A., and Commins, D.: Secondary amyloidosis as a life-ending event in multifocal motor neuropathy. Muscle Nerve 24:1396, 2001. 11. Boonyapisit, K., and Katirji, B.: Multifocal motor neuropathy presenting with respiratory failure. Muscle Nerve 23:1887, 2000. 12. Bouche, P., Moulonguet, A., Younes-Chennoufi, A. B., et al.: Multifocal motor neuropathy with conduction block: a study of 24 patients. J. Neurol. Neurosurg. Psychiatry 59:38, 1995. 13. Braun, P. E., Prayel, D. E., and Latov, N.: Myelin associated glycoprotein is the antigen for a monoclonal IgM in polyneuropathy. J. Neurol. Chem. 39:1261, 1982. 14. Brown, W. F., and Feasby, T. E.: Conduction block and denervation in Guillain Barre polyneuropathy. Brain 107:219, 1984. 15. Cappellari, A., Nobile-Orazio, E., Meucci, N., et al.: Criteria for early detection of conduction block in multifocal motor neuropathy (MMN): a study based on control populations and follow-up of MMN patients. J. Neurol. 244:625, 1997. 16. Carpo, M., Allaria, S., Scarlato, G., and Nobile-Orazio, E.: Marginally improved detection of GM1 antibodies by Covalink ELISA in multifocal motor neuropathy [see comments]. Neurology 53:2206, 1999. 17. Carpo, M., Cappellari, A., Mora, G., et al.: Deterioration of multifocal motor neuropathy after plasma exchange. Neurology 50:1480, 1998. 18. Cavaletti, G., Zincone, A., Marzorati, L., et al.: Rapidly progressive multifocal motor neuropathy with phrenic nerve paralysis: effect of nocturnal assisted ventilation [letter]. J. Neurol. 245:613, 1998. 19. Chad, D. A., Hammer, K., and Sargent, J.: Slow resolution of multifocal weakness and fasciculation: a reversible motor neuron syndrome. Neurology 36:1260, 1986. 20. Charles, N., Benoit, P., Vial, C., et al.: Intravenous immunoglobulin treatment in multifocal motor neuropathy. Lancet 340:182, 1992. 21. Chaudhry, V.: Multifocal motor neuropathy. Semin. Neurol. 18:73, 1998. 22. Chaudhry, V., Corse, A., Cornblath, D., et al.: Maintenance immune globulin therapy for multifocal motor neuropathy. Ann. Neurol. 40:513, 1996. 23. Chaudhry, V., Corse, A. M., Cornblath, D. R., et al.: Multifocal motor neuropathy: response to human immune globulin [see comments]. Ann. Neurol. 33:237, 1993. 24. Chaudhry, V., Corse, A. M., Cornblath, D. R., et al.: Multifocal motor neuropathy: electrodiagnostic features. Muscle Nerve 17:198, 1994.

Multifocal Motor Neuropathy and Conduction Block 25. Claus, D., Specht, S., and Zieschang, M.: Plasmapheresis in multifocal motor neuropathy: a case report. J. Neurol. Neurosurg. Psychiatry 68:533, 2000. 26. Corbo, M., Quattrini, A., Lugaresi, A., et al.: Patterns of reactivity of human anti-GM1 antibodies with spinal cord and motor neurons. Ann. Neurol. 32:487, 1992. 27. Corse, A. M., Chaudhry, V., Crawford, T. O., et al.: Sensory nerve pathology in multifocal motor neuropathy. Ann. Neurol. 39:319, 1996. 28. Cusimano, M. D., Bilbao, J. N., and Cullen, S. M.: Hypertrophic brachial plexus neuritis: a pathological report of 2 cases. Ann. Neurol. 24:615, 1988. 29. Daube, J.: Clinical Neurophysiology. Philadelphia, F. A. Davis, 1996. 30. Di Bella, P., Logullo, F., Dionisi, L., et al.: Chronic multifocal demyelinating neuropathy simulating motor neuron disease. Ital. J. Neurol. Sci. 12:113, 1991. 31. Donaghy, M., Mills, K. R., Boniface, S. J., et al.: Pure motor demyelinating neuropathy deterioration after steroid therapy and improvement with intravenous immunoglobulin. J. Neurol. Neurosurg. Psychiatry 67:15, 1994. 32. Elliott, J. L., and Pestronk, A.: Progression of multifocal motor neuropathy during apparently successful treatment with human immunoglobulin. Neurology 44:967, 1994. 33. Ellis, C. M., Leary, S., Payan, J., et al.: Use of human intravenous immunoglobulin in lower motor neuron syndromes. J. Neurol. Neurosurg. Psychiatry 67:15, 1999. 34. Feasby, T. E., Brown, W. F., Gilbert, J. J., and Hahn, A. F.: The pathological basis of conduction block in human neuropathies. J. Neurol. Neurosurg. Psychiatry 48:239, 1985. 35. Federico, P., Zochodne, D. W., Hahn, A. F., et al.: Multifocal motor neuropathy improved by IVIg: randomized, doubleblind, placebo-controlled study. Neurology 55:1256, 2000. 36. Feldman, E. L., Bromberg, M. B., Albers, J. W., and Pestronk, A.: Immunosuppressive treatment in multifocal motor neuropathy. Ann. Neurol. 30:397, 1991. 37. Felice, K. J., and Goldstein, J. M.: Monofocal motor neuropathy: improvement with intravenous immunoglobulin. Muscle Nerve 25:674, 2002. 38. Finsterer, J., and Derfler, K.: Immunoadsorption in multifocal motor neuropathy. J. Immunother. 22:441, 1999. 39. Finsterer, J., Schwerer, B., Bittner, R. E., and Mamoli, B.: Cerebrospinal fluid filtration and immunoglobulins in multifocal motor neuropathy. Clin. Neuropathol. 18:31, 1999. 40. Freddo, L., Uu, R. K., Latov, N., et al.: Gangliosides GM1 and GD1B are antigens for IgM M protein in patients with motor neuron disease. Neurology 36:454, 1986. 41. Harvey, G. K., Toyka, K. V., Zielasek, J., et al.: Failure of anti-GM1 IgG or IgM to induce conduction block following intraneural transfer. Muscle Nerve 18:388, 1995. 42. Hirota, N., Kaji, R., Bostock, H., et al.: The physiological effect of anti-GM1 antibodies on saltatory conduction and transmembrane currents in single motor axons. Brain 120:2159, 1997. 43. Ho, T. W., Mishu, B., Li, C. Y., et al.: Guillain-Barre syndrome in northern China: relationship to Campylobacter jejuni infection and anti-glycolipid antibodies. Brain 118:597, 1995.

2295

44. Hughes, R. A., Hadden, R. D., Gregson, N. A., and Smith, K. J.: Pathogenesis of Guillain-Barre syndrome. J. Neuroimmunol. 100:74, 1999. 45. IUPAC-IUB Commission on Biochemical Nomenclature (CBN): The nomenclature of lipids. Eur. J. Biochem. 79:11, 1977. 46. Kaji, R., Oka, N., Tsuji, T., et al.: Pathological findings at the site of conduction block in multifocal motor neuropathy. Ann. Neurol. 33:152, 1993. 47. Kaji, R., Shibasaki, H., and Kimura, J.: Multifocal demyelinating motor neuropathy: cranial nerve involvement and immunoglobulin therapy. Neurology 42:506, 1992. 48. Katz, J. S., Barohn, R. J., Kojan, S., et al.: Axonal multifocal motor neuropathy without conduction block or other features of demyelination. Neurology 58:615, 2002. 49. Katz, J. S., and Saperstein, D. S.: Asymmetric acquired demyelinating polyneuropathies: MMN and MADSAM. Curr. Treat. Opin. Neurol. 3:119, 2001. 50. Katz, J. S., Wolfe, G. I., Bryan, W. W., et al.: Electrophysiologic findings in multifocal motor neuropathy [see comments]. Neurology 48:700, 1997. 51. Kermode, A. G., Laing, B. A., Carroll, W. M., and Mastaglia, F. L.: Intravenous immunoglobulin for multifocal motor neuropathy. Lancet 340:920, 1992. 52. Kiernan, M. C., Guglielmi, J. M., Kaji, R., et al.: Evidence for axonal membrane hyperpolarization in multifocal motor neuropathy with conduction block. Brain 125:664, 2002. 53. Kinsella, L. J., Lange, D. J., Trojaburg, W., et al.: Clinical and electrophysiological correlates of elevated anti-GM1 antibody titres. Neurology 44:1278, 1994. 54. Kornberg, A. J., and Pestronk, A.: The clinical and diagnostic role of anti-GM1 antibody testing. Muscle Nerve 17:100, 1994. 55. Krarup, C., Stewart, J. D., Sumner, A. J., et al.: A syndrome of asymmetric limb weakness with motor conduction block. Neurology 40:118, 1990. 56. Latov, N., Hayes, A. P., Donofrio, P. D., et al.: Monoclonal IgM with unique specificity to ganglioside GM1 and GD1b and to lacto-N tetraose associated with human motor neuron disease. Neurology 38:763, 1988. 57. Le Forestier, N., Chassanade, B., Moulonguet, A., et al.: Neuropathies motrice multifocales avec blocs de conduction: 39 cas. Rev. Neurol. 153:579, 1997. 58. Leger, J. M.: Multifocal motor neuropathy and chronic inflammatory demyelinating polyradiculoneuropathy. Curr. Opin. Neurol. 8:359, 1995. 59. Leger, J. M., Chassande, B., Musset, L., et al.: Intravenous immunoglobulin therapy in multifocal motor neuropathy: a double-blind, placebo-controlled study. Brain 124(Pt. 1):145, 2001. 60. Leger, J. M., and Salachas, F.: Diagnosis of motor neuropathy. Eur. J. Neurol. 8:201, 1997. 61. Lewis, R. A.: Multifocal motor neuropathy and Lewis Sumner syndrome: two distinct entities [letter]. Muscle Nerve 22:1738, 1999. 62. Lewis, R. A., Sumner, A. J., Brown, M. J., and Asbury, A. K.: Multifocal demyelinating neuropathy with persistent conduction block. Neurology 32:958, 1982.

2296

Diseases of the Peripheral Nervous System

63. Magistris, M., and Roth, G.: Motor neuropathy with multifocal persistent conduction blocks. Muscle Nerve 15:1056, 1992. 64. Martina, I. S., van Doorn, P. A., Schmitz, P. I., et al.: Chronic motor neuropathies: response to interferon-beta1a after failure of conventional therapies. J. Neurol. Neurosurg. Psychiatry 66:197, 1999. 65. Meucci, N., Cappellari, A., Barbieri, S., et al.: Long term effect of intravenous immunoglobulins and oral cyclophosphamide in multifocal motor neuropathy. J. Neurol. Neurosurg. Psychiatry 63:765, 1997. 66. Mezaki, T., Kaji, R., and Kimura, J.: Multifocal motor neuropathy and Lewis Sumner syndrome: a clinical spectrum [letter]. Muscle Nerve 22:1739, 1999. 67. Mizutamari, R. K., Wegandt, H., and Nores, G. A.: Characterization of anti-ganglioside antibodies present in normal human plasma. J. Neuroimmunol. 50:215, 1994. 68. Molinuevo, J. L., Cruz-Martinez, A., Graus, F., et al.: Central motor conduction time in patients with multifocal motor conduction block. Muscle Nerve 22:926, 1999. 69. Nobile-Orazio, E.: Multifocal motor neuropathy [editorial]. J. Neurol. Neurosurg. Psychiatry 60:599, 1996. 70. Nobile-Orazio, E.: Multifocal motor neuropathy. J. Neuroimmunol. 115:4, 2001. 71. Nobile-Orazio, E., Cappellari, A., Meucci, N., et al.: Multifocal motor neuropathy: clinical and immunological features and response to IVIg in relation to the presence and degree of motor conduction block. J. Neurol. Neurosurg. Psychiatry 72:761, 2002. 72. Nobile-Orazio, E., Meucci, N., Barbieri, S., et al.: High-dose intravenous immunoglobulin therapy in multifocal motor neuropathy [see comments]. Neurology 43:537, 1993. 73. Oh, S. J.: Multifocal inflammatory demyelinating neuropathy: a distinct clinical entity? Neurology 55:740, 2000. 74. Oh, S. J., Claussen, G. C., and Kim, D. S.: Motor and sensory demyelinating mononeuropathy multiplex (multifocal motor and sensory demyelinating neuropathy): a separate entity or a variant of chronic inflammatory demyelinating polyneuropathy? J. Peripher. Nerv. Syst. 2:362, 1997. 75. Oh, S. J., Claussen, G. C., Odabasi, Z., and Palmer, C. P.: Multifocal demyelinating motor neuropathy: pathologic evidence of ‘inflammatory demyelinating polyradiculoneuropathy.’ Neurology 45:1828, 1995. 76. O’Hanlon, G. M., Paterson, G. J., Wilson, G., et al.: AntiGM1 ganglioside antibodies cloned from autoimmune neuropathy patients show diverse binding patterns in the rodent nervous system. J. Neuropathol. Exp. Neurol. 55:184, 1996. 77. O’Hanlon, G. M., Veitch, J., Gallardo, E., et al.: Peripheral neuropathy associated with anti-GM2 ganglioside antibodies: clinical and immunopathological studies. Autoimmunity 32:133, 2000. 78. O’Leary, C. P., Mann, A. C., Lough, J., and Willison, H. J.: Muscle hypertrophy in multifocal motor neuropathy is associated with continuous motor unit activity. Muscle Nerve 20:479, 1997. 79. Olney, R. K.: Consensus criteria for the diagnosis of partial motor conduction block. Muscle Nerve 22:S225, 1999. 80. Olney, R. K., Lewis, R. A., Putnam, T. D., and Campellone, J. V. Jr., for the American Association of Electrodiagnostic

81.

82.

83.

84. 85. 86.

87.

88.

89. 90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

Medicine: Consensus criteria for the diagnosis of multifocal motor neuropthy. Muscle Nerve 27:117, 2003. Ortman, R. A., and Klippel, J. H.: Update on cyclophosphamide for systemic lupus erythematosus. Rheum. Dis. Clin. North Am. 26:363, 2000. Pakiam, A. S., and Parry, G. J.: Multifocal motor neuropathy without overt conduction block. Muscle Nerve 21:243, 1998. Paparounas, K., O’Hanlon, G. M., O’Leary, C. P., et al.: Antiganglioside antibodies can bind peripheral nerve nodes of Ranvier and activate the complement cascade without inducing acute conduction block in vitro. Brain 122:807, 1999. Parry, G. J.: Motor neuropathy with multifocal conduction block. Semin. Neurol. 13:269, 1993. Parry, G. J.: AAEM case report #30: multifocal motor neuropathy. Muscle Nerve 19:269, 1996. Parry, G. J.: Are multifocal motor neuropathy and LewisSumner syndrome distinct nosologic entities? Muscle Nerve 22:557, 1999. Parry, G. J., and Clarke, S.: Pure motor neuropathy with multifocal conduction block masquerading as motor neuron disease [abstract]. Muscle Nerve 8:167, 1985. Parry, G. J., and Clarke, S.: Multifocal acquired demyelinating neuropathy masquerading as motor neuron disease. Muscle Nerve 11:103, 1988. Parry, G. J., and Sumner, A. J.: Multifocal motor neuropathy. Neurol. Clin. 10:671, 1992. Pederson-Bjergaard, J., Ersboll, J., Hansen, V., et al.: Carcinoma of the urinary bladder after treatment with cyclophosphamide for non-Hodgkin’s lymphoma. N. Engl. J. Med. 318:1028, 1988. Pestronk, A.: Invited review: motor neuropathies, motor neuron disorders, and antiglycolipid antibodies. Muscle Nerve 14:927, 1991. Pestronk, A., Adams, R. N., Clawson, L., et al.: Serum antibodies to GM1 ganglioside in amyotrophic lateral sclerosis. Neurology 38:1457, 1988. Pestronk, A., Chaudhry, V., Feldman, E. L., et al.: Lower motor neuron syndromes defined by patterns of weakness, nerve conduction abnormalities, and high titers of antiglycolipid antibodies. Ann. Neurol. 27:316, 1990. Pestronk, A., and Choksi, R.: Multifocal motor neuropathy: serum IgM anti-GM1 ganglioside antibodies in most patients detected using covalent linkage of GM1 to ELISA plates. Neurology 49:1289, 1997. Pestronk, A., Choksi, R., Blume, G., and Lopate, G.: Multifocal motor neuropathy: serum IgM binding to a GM1 ganglioside-containing lipid mixture but not to GM1 alone. Neurology 48:1104, 1997. Pestronk, A., Cornblath, D. R., Ilyas, A. A., et al.: A treatable multifocal motor neuropathy with antibodies to GM1 ganglioside. Ann. Neurol. 24:73, 1988. Pfeiffer, G., Wicklein, E. M., and Wittig, K.: Sensitivity and specificity of different conduction block criteria. Clin. Neurophysiol. 111:1388, 2000. Pringle, C. E., Belden, J., Veitch, J. E., and Brown, W. F.: Multifocal motor neuropathy presenting as ophthalmoplegia. Muscle Nerve 20:347, 1997. Rhee, R. K., England, J. D., and Sumner, A. J.: A computer simulation of conduction block: effects produced by actual

Multifocal Motor Neuropathy and Conduction Block

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110. 111.

112.

113.

114.

115.

116.

block versus interphase cancellation. Ann. Neurol. 28:146, 1990. Roberts, M., Willison, H. J., Vincent, A., and NewsomDavis, J.: Multifocal motor neuropathy human sera block distal motor nerve conduction in mice. Ann. Neurol. 38:111, 1995. Roth, G., Rohr, J., Magistris, M. R., and Ochsner, F.: Motor neuropathy with proximal multifocal persistent conduction block, fasciculations and myokymia: evolution to tetraplegia. Eur. Neurol. 25:416, 1986. Sadiq, S. A., Thomas, F. P., Kilidireas, K., et al.: The spectrum of neurologic disease associated with anti-GM1 antibodies. Neurology 40:1067, 1990. Santoro, M., Thomas, F. P., Fink, M. E., et al.: IgM deposits at nodes of Ranvier in a patient with amyotrophic sclerosis and GM1 antibodies and multifocal motor conduction block. Ann. Neurol. 28:373, 1990. Santoro, M., Uncini, A., Corbo, M., et al.: Experimental conduction block induced by serum from a patient with anti-GM1 antibodies. Ann. Neurol. 31:385, 1992. Saperstein, D. S., Amato, A. A., Wolfe, G. I., et al.: Multifocal acquired demyelinating sensory and motor neuropathy: the Lewis-Sumner syndrome [see comments]. Muscle Nerve 22:560, 1999. Saperstein, D. S., Katz, J. S., Amato, A. A., and Barohn, R. J.: Clinical spectrum of chronic acquired demyelinating polyneuropathies. Muscle Nerve 24:311, 2001. Sobue, I., Saito, N., Iida, M., and Ando, K.: Juvenile type of distal and segmental muscular atrophy of upper extremities. Ann. Neurol. 3:429, 1978. Steck, A. J., Murray, N., Meier, C., et al.: Demyelinating neuropathy and monoclonal IgM antibody to myelin associated glycoprotein. Neurology 33:19, 1983. Sugie, K., Murata, K., Ikoma, K., et al.: A case of acute multifocal motor neuropathy with conduction block after Campylobacter jejuni enteritis. Rinsho Shinkeigaku 38:142, 1998. Svennerholm, L.: Chromatographic separation of gangliosides. J. Neurol. Chem. 10:613, 1963. Takigawa, T., Yasada, H., Kikkawa, R., et al.: Antibodies against GM1 ganglioside affect K⫹ and Na⫹ currents in isolated rat myelinated nerve fibers. Ann. Neurol. 36:436, 1995. Talar-Williams, C., Hijazi, Y. M., and Walther, M. M.: Cyclophosphamide induced cystitis and bladder cancer in patients with Wegner’s granulomatosis. Ann. Intern. Med. 124:477, 1996. Tan, E., Lynn, D. J., Amato, A. A., et al.: Immunosuppressive treatment of motor neuron syndromes. Arch. Neurol. 51:194, 1994. Taylor, B., Speed, B., Kaldor, J., et al.: Serological evidence for prior infection with the bacterium Campylobacter jejuni/coli in multifocal motor neuropathy. J. Clin. Neurosci. 5:33, 1998. Taylor, B. V., Gross, L., and Windebank, A. J.: The sensitivity and specificity of anti-GM1 antibody testing. Neurology 47:951, 1996. Taylor, B. V., Wright, R. A., Harper, C. M., and Dyck, P. J.: Natural history of 46 patients with multifocal motor neuropathy with conduction block. Muscle Nerve 23:900, 2000.

2297

117. Taylor, B., Dyck, P. J. B., Engelstad, J., and Grant, I.: Multifocal motor neuropathy: pathologic alterations at the site of conduction block. J. Neuropathol. Exp. Neurol. 63:129, 2004. 118. Terenghi, F., Allaria, S., Scarlato, G., and Nobile-Orazio, E.: Multifocal motor neuropathy and Campylobacter jejuni reactivity. Neurology 59:282, 2002. 119. Tyson, J., Malcolm, S., Thomas, P. K., and Harding, A. E.: Deletions of chromosome 17p11.2 in multifocal neuropathies. Ann. Neurol. 39:180, 1996. 120. Umapathi, T., Hughes, R. A., Nobile-Orazio, E., and Leger, J. M.: Immunosuppressive treatment for multifocal motor neuropathy. Cochrane Database Syst. Rev. 2:CD003217, 2002. 121. Uncini, A., Santoro, M., Corbo, M., et al.: Conduction abnormalities induced by sera of patients with multifocal motor neuropathy and anti-GM1 antibodies. Muscle Nerve 16:610, 1993. 122. Van den Berg, L. H., Franssen, H., and Wokke, J. H.: The long-term effect of intravenous immunoglobulin treatment in multifocal motor neuropathy. Brain 121:421, 1998. 123. Van den Berg, L. H., Kerkhoff, H., Oey, P. L., et al.: Treatment of multifocal motor neuropathy with high dose intravenous immunoglobulins: a double blind, placebo controlled study. J. Neurol. Neurosurg. Psychiatry 59:248, 1995. 124. Van den Berg, L. H., Lokhorst, H., and Wokke, J. H.: Pulsed high-dose dexamethasone is not effective in patients with multifocal motor neuropathy [see comments]. Neurology 48:1135, 1997. 125. Van den Berg-Vos, R. M., Franssen, H., Visser, J., et al.: Disease severity in multifocal motor neuropathy and its association with the response to immunoglobulin treatment. J. Neurol. 249:330, 2002. 126. Van Den Berg-Vos, R. M., Franssen, H., Wokke, J. H., and Berg, L. H.: Multifocal motor neuropathy: long-term clinical and electrophysiological assessment of intravenous immunoglobulin maintenance treatment. Brain 125:1875, 2002. 127. Van den Berg-Vos, R. M., Van den Berg, L. H., Franssen, H., et al.: Multifocal inflammatory demyelinating neuropathy: a distinct clinical entity? Neurology 54:26, 2000. 128. Van den Berg-Vos, R. M., Van den Berg, L. H., Franssen, H., et al.: Treatment of multifocal motor neuropathy with interferon-beta1A. Neurology 54:1518, 2000. 129. Van Es, H. W.: MRI of the brachial plexus. Eur. Radiol. 11:325, 2001. 130. Van Es, H. W., Van den Berg, L. H., Franssen, H., et al.: Magnetic resonance imaging of the brachial plexus in patients with multifocal motor neuropathy. Neurology 48:1218, 1997. 131. van Schaik, I. N., Bossuyt, P. M., Brand, A., and Vermeulen, M.: Diagnostic value of GM1 antibodies in motor neuron disorders and neuropathies: a meta-analysis. Neurology 45:1570, 1995. 132. Weller, M., Stevens, A., Sommer, N., et al.: Ganglioside antibodies: a lack of diagnostic specificity and clinical utility. J. Neurol. 239:455, 1992. 133. White, J. R., Sachs, G. M., and Gilchrist, J. M.: Multifocal motor neuropathy with conduction block and Campylobacter jejuni [see comments]. Neurology 46:562, 1996.

2298

Diseases of the Peripheral Nervous System

134. Willison, H. J.: Antiglycolipid antibodies in peripheral neuropathy: fact or fiction? J. Neurol. Neurosurg. Psychiatry 57:1383, 1994. 135. Willison, H. J., and O’Hanlon, G. M.: The immunopathogenesis of Miller Fisher syndrome. J. Neuroimmunol. 100:3, 1999. 136. Willison, H. J., Paterson, G., Kennedy, P. G., and Veitch, J.: Cloning of human anti-GM1 antibodies from motor neuropathy patients. Ann. Neurol. 35:471, 1994. 137. Willison, H. J., Veitch, J., Swan, A. V., et al.: Interlaboratory validation of an ELISA for the determination of

serum anti-ganglioside antibodies. Eur. J. Neurol. 6:71, 1999. 138. Willison, H. J., and Yuki, N.: Peripheral neuropathy and anti-glycolipid antibodies. Brain 125:1, 2002. 139. Yeung, K. B., Thomas, P. K., King, R. M. H., et al.: The clinical spectrum of peripheral neuropathies associated with benign monoclonal IgM, IgG, and IgA paraproteins. J. Neurol. 238:383, 1991. 140. Yuki, N., Taki, T., Inagaki, F., et al.: A bacterium lipopolysaccharide that elicits Guillain-Barre syndrome has a GM1 ganglioside-like structure. J. Exp. Med. 178:1771, 1993.

102 Immune Brachial Plexus Neuropathy GUILLERMO A. SUAREZ

Historical Background Epidemiology Clinical Features Symptoms Neurologic Examination

Antecedent Illnesses and Associated Conditions Diagnosis and Evaluation Investigations Differential Diagnosis

Immune brachial plexus neuropathy (IBPN) is a distinctive clinical syndrome characterized by the acute onset of pain localized to the scapular region followed by weakness and muscle atrophy. There is evidence for an inflammatory immune pathogenesis in this syndrome. The disorder is also known as neuralgic amyotrophy after the description of Parsonage and Turner in 1948 emphasizing the pain and weakness of the syndrome.61 Multiple other synonyms have been used to describe the syndrome highlighting some aspects of the illness. Some of these include Parsonage-Turner syndrome, brachial plexus neuritis, shoulder girdle neuritis, shoulder girdle syndrome, cryptogenic brachial neuropathy, acute shoulder neuritis, cervical radiculoplexus neuropathy, and amyotrophic paralysis of the periscapular muscles and serratus magnus palsy.73 Immune brachial plexus neuropathy (IBPN) is the term used here. Brachial plexus neuropathy occurs in sporadic and familial or hereditary forms. This chapter focuses on the acquired or sporadic brachial plexus neuropathy, which is presumed to be inflammatory-immune mediated. The hereditary form, hereditary brachial plexus neuropathy (HBPN) with genetic localization in some families to chromosome 17q25, is described in Chapter 74. Other causes of brachial plexus injury (trauma, radiation, cancer) are described in Chapter 55. Other reviews on the subject have been published.75,100,101

Pathogenesis and Pathology Treatment Prognosis

HISTORICAL BACKGROUND The initial descriptions of IBPN appeared in the literature in the late 1800s. In 1896, Leszynsky reported a case of bilateral brachial plexus neuritis following “acute croupous pneumonia” and provided the earliest photographic evidence of the periscapular muscle weakness that occurs in this illness.46 Feinberg in 1897 described a case of unilateral neuritis of the brachial plexus associated with influenza.26 However, Murray and Wilbourn stated that it may have been Hippocrates, in “Endemics,” who first mentioned this condition in association with an epidemic of cough at Perinthus.58 Large series of patients were reported years later. In 1941, Wyburn-Mason reported 42 patients with brachial neuritis and suggested the possibility of an epidemic form of the disease occurring over an 8-month period in London, England. No common etiology was found.103 Spillane described 46 cases of brachial plexopathy in military personnel during World War II.83 It was the report of Parsonage and Turner in 1948 that provided a detailed description of the clinical syndrome in a large group of 136 patients seen mainly among military personnel serving in the United Kingdom and India during World War II.61 Since then many additional cases and series of patients have been published, expanding the original clinical description to include cases affecting not only the brachial plexus but 2299

2300

Diseases of the Peripheral Nervous System

also isolated peripheral nerves, including the phrenic and cranial nerves.7,13,22,23,44,55,57,67,68,71,79,84–86,90

EPIDEMIOLOGY Epidemiologic studies of IBPN have been limited. Beghi et al., in the only population-based study of brachial plexus neuropathy, reported an annual incidence of 1.64 per 100,000 population in Olmsted County, Minnesota, from 1970 to 1981.5 This condition affects individuals of any age, from children, including newborns, to the elderly. Earlier reports emphasizing a predilection for young and middle-aged adults may have been biased, reflecting the younger age of military recruits. There is a slight male predominance with a ratio ranging from 2:1 to 3:1. In the large series of cases by Tsairis et al., the male-to-female ratio was 2.4:1.86 Most of the cases are sporadic, although there have been occasional reports suggesting epidemic forms of the disease. Wyburn-Mason in 1941 reported 42 patients who develop the disease over an 8-month period, but no common cause was established.103 Bárdos and Somodská reported an outbreak of the disease in northeast Czechoslovakia from 1949 to 1953 that involved 265 patients. A coxsackievirus infection was implicated in this epidemic.4 Augé and Velázquez reported an epidemiologic cluster of eight Native American patients from the southwest of the United States who develop the typical clinical syndrome over a 3-month period following a nonspecific antecedent febrile illness.3

CLINICAL FEATURES The clinical presentation is distinctive, as reported by Parsonage and Turner61: The essential clinical picture is simple: without any constitutional disturbance, pain starts suddenly across the top of the shoulder blade and may radiate down the outer side of the upper arm or into the neck. This pain lasts from a few hours to a fourth night or more, and then a flaccid paralysis of some of the muscles of the shoulder girdle and often of the arm develops. In some cases there is a patch of numbness over the outer side of the upper arm. When the paralysis appears, the severe pain usually stops, but a dull ache may persist considerably longer. This clinical picture is subject to modifications. As many additional cases have subsequently been published, this distinctive clinical picture has remained accurate, but variations in the clinical presentation have also been the norm.23

Symptoms The initial cardinal symptom is pain, which begins abruptly, may awaken the patient at night, and it is usually

described as “sharp,” “aching,” “stabbing,” “excruciating,” or “throbbing” in quality. It is constant, maximal at onset, and generally localized to the shoulder, scapula, and intrascapular region. In some cases, it is not restricted only to the shoulder but also affects the arm, forearm, and hand. The pain is not associated with constitutional symptoms, and fever is absent. The patient tends to immobilize the arm94 and, in contrast to the compressive radicular pain of a cervical radiculopathy, Valsalva-type maneuvers such as coughing and sneezing do not exacerbate the pain. In the series of Tsairis et al., pain was the initial symptom in 95 of the 99 cases.86 Unilateral pain correlated with subsequent unilateral weakness in 88% of patients. In 25 of 34 patients (75%), bilateral onset of pain was followed by bilateral weakness. Muscle tenderness was also observed in the area of the pain and was reported by 15% of the patients in this large series. In two other large series, pain was present in all 40 consecutive patients and in 86% of 37 cases.13,55 However, there are also reports of patients with IBPN presenting with minimal pain or no pain at onset.79,95 Two painless cases were noted in the original Parsonage and Turner description and other cases have been reported in other series, but overall these cases are the minority.86 Pain appears to be less frequent in children, with approximately two thirds of patients under 16 years of age presenting with pain.89,97 The duration of the pain is variable, but in general pain improves after 2 to 3 weeks. Muscle weakness appears when the pain is improving, although in some patients weakness develops suddenly, accompanying the intense painful onset of the syndrome. In the majority of patients (70%), muscle weakness is present by the third week of the illness. The distribution of weakness varies considerably from one patient to another, with a predilection for proximal shoulder girdle and arm muscles. The muscles more frequently involved include the deltoid, supraspinatus, infraspinatus, serratus anterior, biceps, and triceps brachii, but many other combinations occur. This patchy, multifocal involvement suggest compromise of parts of the brachial plexus, including upper, middle, and lower trunks; single or multiple peripheral nerves; and occasionally C5/6 cervical nerve roots. Bilateral plexus involvement has been reported in up to 34% of cases.86 In large series of cases, weakness limited to muscles innervated by an isolated peripheral nerve ranged from 10% to 35%, with the long thoracic, suprascapular, and axillary nerves most commonly involved. The axillary and suprascapular nerves were the most frequent combination of nerves affected together, with 10 patients reported in the Cruz-Martinez et al. series.13 There have been reports of selective involvement of single peripheral nerves in the upper extremity affecting the radial nerve, musculocutaneous nerve, or dorsal interosseous and anterior interossei branches of the median nerve.29,39,71,78,90,93 This presentation has been referred as mimicking a “mononeuritis multiplex” pattern.22,23 In another large

Immune Brachial Plexus Neuropathy

series of cases, single peripheral nerve involvement was absent.55 This likely reflects the variable clinical presentation but also the differences in case ascertainment and the criteria used to define involvement (clinical vs. electrophysiologic assessment). There are also reports of involvement of the phrenic nerve causing unilateral or bilateral diaphragmatic paralysis.32,44,59,81 These patients present with sudden onset of dyspnea, which may have been preceded by shoulder and neck pain. Lahrmann et al. describe in detail four patients with involvement of the phrenic nerves as part of IBPN.44 All patients had severe pain affecting one or both shoulders followed by dyspnea. Two patients had no other muscle weakness, and the remaining two patients had phrenic neuropathies associated with weakness of muscles innervated by the upper trunk of the brachial plexus and also the long thoracic nerve. Other single case reports and small series of patients have described similar findings.7,57,62,68 Although rare, these cases widen the spectrum of IBPN and include dyspnea as part of the symptom complex. The clinical spectrum is expanded further by a case report of a 55-year-old patient who had simultaneous involvement of the brachial plexus and multiple cranial nerves (IX, X, XI, and XII) and laryngeal nerves with vocal cord paralysis.67 Other cases presenting with vocal cord paralysis have been described.36,85 Sensory symptoms are often overshadowed by pain. In large series, paresthesias were reported by 20% to 35% of patients. “Tingling” and “numbness” were noted during the first 2 to 3 weeks of the illness and followed a variable distribution.

Neurologic Examination The predominant feature on neurologic examination is muscle weakness accompanied by variable atrophy. As mentioned before, there is significant variability in regard to the distribution of muscle weakness depending on which part of the brachial plexus or individual peripheral nerves sustains the brunt of the presumed inflammatory injury. Bilateral involvement is usually asymmetrical and reported in up to 34% of patients. Winging of the scapula as a result of weakness and atrophy of the serratus anterior may be the only neurologic deficit. Sparse fasciculations may be seen in some patients. Sensory deficits are variable and reported in 15% to 66% of patients. Sensory loss predominantly involves cutaneous sensation and does not always parallel the motor deficits. Objective sensory loss tends to affect the lateral aspect of the arm (axillary nerve distribution) and the forearm (radial nerve distribution). Deep tendon reflexes are either normal or hypoactive depending on the severity and extent of the lesion at the time of the examination.

2301

ANTECEDENT ILLNESSES AND ASSOCIATED CONDITIONS In some cases of IBPN, the onset of symptoms is preceded by an illness along the lines of a flulike syndrome or upper respiratory tract infection. In the series of Tsairis et al., 25% of patients had an antecedent flulike syndrome.86 In other series, up to 55% of patients reported an antecedent febrile illness.13 The onset of symptoms of IBPN occurs after the acute phase of the antecedent illness, usually 2 to 3 weeks thereafter. This is similar to other acquired postinfectious neuropathies such as Guillain-Barré syndrome. Many infectious agents have been postulated to be the precipitating event of IBPN, including viral,10,40,63,69,77 bacterial,9,12,24,34,60,88 and parasitic infections51 (Table 102–1). Multiple other conditions have been reported in association with IBPN, such as recent immunization,72,96 interferon6 and interleukin-2 therapy,49 serum sickness,100 medications,33 botulinum toxin administration,31,76 connective tissue disorders,1,2,8,38,56 giant cell arteritis,16,54 Hodgkin’s lymphoma,43,66 paraproteinemia,52 pregnancy,70 postpartum status,11,18,45 surgery,27,50 heroin use,25,74,98 and vigorous use of the arm(s)4,67,86 (see Table 102–1). Antecedent vaccination was reported by 14% of patients in one series.86 Different vaccines and antitoxin injections (especially tetanus) were associated with the disease.41,72 In the immunized patients, the symptoms of IBPN developed 6 to 21 days after the injection and affected either the injected or the noninjected arm. In the postoperative cases of IBPN, the onset of symptoms has been noted to occur immediately after awakening from anesthesia and up to 8 days after the surgical procedure.50 Well-described cases have also been reported during pregnancy and the puerperium.18,45,70 In summary, the data suggest that multiple conditions, rather than being directly causal or specific, can trigger an inflammatory-immune attack on the brachial plexus producing the syndrome of IBPN.

DIAGNOSIS AND EVALUATION The diagnosis of IBPN can be established with high certainty when all of the elements are present, specifically the typical clinical history of pain followed by weakness. As in many other conditions, the certainty of the diagnosis is low earlier in the course of the illness, especially when pain is the only feature of the syndrome. This may prompt an initial orthopedic consultation for the evaluation of shoulder pain. Another diagnostically challenging presentation is when the patient presents with dyspnea without other typical features of IBPN. Bilateral involvement of phrenic nerves may lead to orthopnea mimicking cardiac disease, leading to a cardiovascular evaluation.

2302

Diseases of the Peripheral Nervous System

Table 102–1. Antecedent Illnesses and Associated Conditions Infectious Upper respiratory infection Influenza Coxsackievirus Parvovirus B19 Epstein-Barr virus Human immunodeficiency virus Salmonella Escherichia coli sepsis Ehrlichiosis malaria schistosomiasis Immunizations Tetanus toxoid Hepatitis B Swine flu Medications Interferon Interleukin-2 Lamotrigine Botulinum toxin Connective Tissue Disorders Systemic lupus erythematosus Rheumatoid arthritis Scleroderma Polyarteritis nodosa Giant cell arteritis Ehlers-Danlos syndrome Postoperative (surgery at a distant site) Pregnancy and Puerperium (inherited brachial plexus neuropathy) Other Intravenous drug (heroin) abuse Infective endocarditis Paraproteinemia Hodgkin’s disease Strenuous exercise Associated with diabetes mellitus Modified from Rubin, D. I.: Neuralgic amyotrophy: clinical features and diagnostic evaluation. The Neurologist 7:350, 2001.

Investigations In addition to the clinical history and examination, electrophysiologic studies are important to localize and further characterize the disease process. Nerve conduction studies are often abnormal when performed 3 weeks after the onset of symptoms. Sensory nerve action potentials may be reduced or absent, indicating a lesion at or distal to the dorsal root ganglion. This is helpful to separate lesion(s) involving the plexus from lesion(s) affecting nerve roots or

spinal cord (anterior horn cells). In an electromyography (EMG) study of 16 patients with IBPN, Flaggman and Kelly reported abnormalities of sensory nerve conduction studies in 33% of patients.28 The most frequent finding was low-amplitude sensory nerve action potentials. Cwik et al., in an EMG study of 56 patients with IBPN, reported that routine nerve conduction studies were abnormal in only 15% of cases. However, when additional less commonly performed nerve conduction studies were done, the frequency of abnormalities increased. Low-amplitude or absent lateral antebrachial cutaneous nerve sensory response was the most frequent abnormality in 32% of cases.15 Median and ulnar motor nerve conduction studies may demonstrate low compound muscle action potential amplitudes if the lower trunk of the plexus is affected. In the Flaggman and Kelly series,28 only 3 of 16 patients showed abnormalities in these nerves; however, more frequently abnormal responses are encountered if additional motor nerve conduction studies are performed, especially of the musculocutaneous and axillary nerves, reflecting involvement of the upper trunk of the plexus. Needle EMG abnormalities are present in most patients with IBPN, but the findings will vary depending on the timing of the study in relationship to the onset of clinical symptoms. Needle EMG typically demonstrates a pattern of active denervation with fibrillation potentials and motor unit potential abnormalities in affected muscles during the acute phase of the disorder. Needle EMG of the cervical paraspinal muscles is generally normal, which is helpful in localizing the pathologic process distally in the brachial plexus. In a subset of patients there are changes on needle EMG on the unaffected contralateral upper extremity, suggesting that in some patients with IBPN there is subclinical involvement of the other arm. Wilbourn proposed a correlation of EMG findings with the severity of the axonal loss in IBPN.101 In mild cases of IBPN, the electrophysiologic findings may be limited to fibrillation potentials on needle EMG without nerve conduction abnormalities. With moderate lesions there is the additional reduction of sensory nerve action potential amplitudes on nerve conduction studies, and with more severe lesions, additional findings may include reduction of the compound muscle action potential amplitudes and reduced recruitment of voluntary motor unit potentials. In general, these findings are typically seen, but the following caveats must be taken into consideration. First, in many instances the neuropathic process is multifocal and may affect different parts of the brachial plexus not easily accessible to testing by conventional nerve conduction studies. Second, the timing of the study relative to the onset of the patient’s symptoms and recovery may influence the presence of active denervation. Third, inferences about pathologic interstitial changes cannot be made with certainty based solely upon electrodiagnostic evaluation.

Immune Brachial Plexus Neuropathy

Watson and colleagues described two patients with IBPN and electrophysiologic evidence of partial conduction block across the lower trunk of the plexus during the early phase of the illness.92 Three similar cases of possible focal conduction block have also been described, but the reported findings in these cases are less compelling.48 It is possible that partial conduction block may be present across the plexus in early stages of IBPN. However, current electrophysiologic methods do not allow for the determination of focal or partial conduction block in this anatomic region with a high level of confidence, especially with upper trunk lesions, because of the uncertainty about the exact location of electrical stimulation, the effects of volume conduction in this area, and assurance that supramaximal stimulation has been achieved. In summary, most of the currently available evidence from electrophysiologic studies suggests multifocal or patchy abnormalities consistent with axonal damage (loss) within the brachial plexus and isolated peripheral nerves of the upper extremities. Radiologic investigations such as computerized tomography and magnetic resonance imaging (MRI) are helpful in excluding other conditions that may present with similar clinical features, such as a herniated cervical disc and other brachial plexopathies (neoplastic infiltration, nerve tumors, and postradiation plexopathy). In some cases, MRI of the brachial plexus may provide supportive evidence of IBPN, showing signal alterations consistent with skeletal muscle denervation, especially in the subacute phase of the illness.17 In selected cases of IBPN, MRI may demonstrate increased T2 signal intensity within the brachial plexus.80,102 Laboratory studies are generally negative, including inflammatory and immunologic markers such as sedimentation rate, antinuclear antibodies, rheumatoid factor, antibodies to extractable nuclear antigens, and antineutrophil cytoplasmic antibodies, unless the disorder is associated with a specific infection or connective tissue disease. Cerebrospinal fluid (CSF) examination may show mildly increased protein without pleocytosis but, in general, CSF examination is normal. In one series, CSF examination was performed in 35 patients with only 4 patients demonstrating a modest elevation in CSF protein.86

DIFFERENTIAL DIAGNOSIS Neurologic and non-neurologic conditions need to be considered in the differential diagnosis of IBPN (Table 102–2). Cervical radiculopathies, especially involving C5/6 roots, can be confused with upper trunk involvement of IBPN. The pain of cervical radiculopathy may improve with immobilization and may be aggravated by some maneuvers such as rotation of the neck, coughing, or sneezing, whereas the pain of IBPN is constant and usually

2303

Table 102–2. Differential Diagnosis of Immune Brachial Plexus Neuropathy Neurologic Spinal cord tumors Inflammatory myelitis Acute poliomyelitis West Nile virus infection Cervical radiculopathies Isolated or multiple mononeuropathies Other brachial plexopathies Hereditary brachial plexus neuropathy Trauma Neoplastic (primary and secondary) Postradiation plexopathy Non-neurologic Conditions Rotator cuff lesions Other musculoskeletal injuries of the shoulder Pulmonary and cardiovascular conditions (dyspnea)

not modified by change in posture or Valsalva-type maneuvers. Ultimately, MRI and electrophysiologic studies are necessary to confirm the diagnosis and exclude a cervical radiculopathy. Traumatic mononeuropathies may be distinguished from IBPN by the clinical history and temporal profile. Multiple mononeuropathies of the upper limb can be difficult to separate from typical IBPN. Some investigators argue that single or multiple mononeuropathies are commonly encountered in the clinical spectrum of IBPN and prefer the term neuralgic amyotrophy because this name encompasses all of the variations in the clinical presentations of IBPN.22,23 Brachial plexus involvement as part of a cervicobrachial radiculoplexopathy associated with diabetes mellitus may need to be differentiated from IBPN, especially when there are no other manifestations of diabetic neuropathy. Katz et al. reported 9 patients with cervicobrachial involvement in a group of 60 patients with diabetic lumbosacral radiculoplexopathy (DLSRP).37 In most cases (seven), the upper limb involvement occurred after the onset of DLSRP. In one case the upper limb deficits developed simultaneously with DLSRP, and in another case the arm symptoms preceded the onset of the lumbosacral syndrome by 2 months. In the majority of cases the motor deficits involved the intrinsic hand muscles and flexors and extensors of the wrist; proximal arm involvement was rare. In the series of Dyck et al., 10% of patients (3 of 33) with DLSRP had cervicobrachial involvement.20 Brachial plexus neuropathy (cervical radiculoplexus neuropathy) associated with diabetes mellitus is also described in Chapter 86. In the early stages of the illness, spinal cord tumors and inflammatory myelitis may be confused with IBPN, but as the disease evolves, the presence of upper motor neuron signs and sphincter dysfunction allows a clear distinction

2304

Diseases of the Peripheral Nervous System

between these conditions. Acute poliomyelitis and the focal forms of West Nile virus infection (WNV) can superficially mimic IBPN.47,83 Several features from the history and examination usually permit a definite separation. During an attack of poliomyelitis or WNV, the neurologic deficits develop concomitantly with systemic symptoms and signs of infection such as fever, malaise, headache, and meningismus. In addition, localized pain is not a prominent feature of acute poliomyelitis or WNV. In IBPN, pain usually develops in the convalescent phase of any preceding viral illness, and there are generally no systemic symptoms or signs of an infection when the patient develops actual neurologic deficits. The distribution of weakness is also helpful for the differential diagnosis. It is segmental in poliomyelitis and multifocal or patchy and peripheral in IBPN. Furthermore, the CSF examination is generally abnormal in poliomyelitis and WNV, whereas in IBPN is normal in most cases. A recent case of WNV presenting with diffuse unilateral arm weakness mimicking IBPN was observed by the author (unpublished observation, 2003). The clinical findings helpful for the differential diagnosis were the absence of pain or sensory symptoms and the diffuse (proximal and distal) pattern of muscle weakness in this patient. Laboratory studies confirmed the diagnosis of WNV. Brachial plexus neuropathy also presents in an inherited form, HBPN.42,65,99 This is an autosomal dominant disorder with chromosomal localization in some families mapped to 17q25.64 Patients develop recurrent attacks of arm pain followed by weakness, which are essentially indistinguishable from the sporadic or acquired form of brachial plexus neuropathy. Some features may separate HBPN from sporadic IBPN. In HBPN, pregnancy and childbirth may trigger the attacks, some patients have mild dysmorphic features (hypotelorism, short stature, and epicanthic folds), and there are other family members affected by the condition. In some patients with HBPN, superficial radial nerve biopsies have demonstrated perivascular inflammation with axonal degeneration.42 HBPN is discussed in detail in Chapter 74. The clinical history, physical examination, and electrodiagnostic and radiologic studies permit differentiation of IBPN from other causes of brachial plexopathies such as traumatic injury, neoplastic infiltration, and postradiation plexopathy. In cases of traumatic brachial plexopathies, the separation is straightforward based on the history of significant localized trauma or traction of the brachial plexus. For instance, it is common to sustain traction-type injuries of the plexus during snowmobile or motorcycle accidents. Postradiation plexopathy presents with progressive arm weakness in a patient who has previously received radiation treatments usually as part of the treatment of breast cancer and less frequently lung cancer or lymphoma. Myokymic discharges may be seen in the EMGs of these patients. In cases of neoplastic infiltration of the plexus,

the differentiation may be more difficult and a high index of suspicion is necessary. Secondary neoplasms are more frequent than primary, and most cases are metastatic involvement from lung or breast cancer, but other neoplasms may also invade the brachial plexus. These are described in more detail in Chapter 55. IBPN may need to be differentiated from shoulder injuries, including rotator cuff tears and other musculoskeletal abnormalities of the shoulder joint.

PATHOGENESIS AND PATHOLOGY The pathogenesis of IBPN is partially understood, but current evidence suggests an underlying inflammatoryimmune process. There are several lines of evidence that suggest an inflammatory-immune mechanism. First, the clinical presentation with an acute onset and monophasic course, antecedent presumed viral infection,13,53,86 and association with immunizations,41,86,96 serum sickness, and treatment with immunomodulating agents such as interferon and interleukin-26,49 point toward a potential immune mechanism. Second, both humoral and cellular immune mechanisms have been implicated in the disease. Complement-fixing antibodies to peripheral nerve myelin (anti-PNM) have been reported in patients with IBPN. One report described three patients with elevated serum anti-PNM and soluble terminal complement activation products in the acute phase of IBPN and compared the results with 25 normal controls.91 In the affected patients, the level of antibodies decreased several months later in association with clinical recovery. Another study reported that blood lymphocytes from patients with IBPN are sensitized to extracts of human brachial plexus nerves.82 Pathologic studies, although limited, have provided strong evidence of cellular autoimmune mechanisms in the pathogenesis of IBPN. Cusimano et al. described two patients with recurrent IBPN with unusual fusiform enlargements of trunks of the brachial plexus with focal chronic inflammatory infiltrates, edema, and onion bulb formation on pathologic evaluation.14 We reported the brachial plexus biopsy findings of four additional patients with IBPN.84 Mononuclear inflammatory infiltrates were noted surrounding epineurial and endoneurial vessels without findings of necrotizing vasculitis. The infiltrates were mainly T lymphocytes. In one case, the lymphocytic infiltrate contained a reactive germinal center with CD20positive B lymphocytes (Fig. 102–1). Electron micrographs in one patient revealed demyelinated axons and onion bulb formation as noted in the previous report. In a study of nerve biopsy specimens from 16 additional cases of IBPN involving the brachial plexus (4) and more distal peripheral nerves, including the superficial radial (8), median (2), dorsal ulnar cutaneous (1), and lateral antebrachial cutaneous (1) nerves, we found the following spectrum of

Immune Brachial Plexus Neuropathy

2305

FIGURE 102–1 A, Longitudinal paraffin section, stained with hematoxylin-eosin, of a brachial plexus fascicle showing a large inflammatory cell collection surrounding an endoneurial small vessel. (Magnification ⫻ 4 before 32% reduction.) B, At higher magnification (⫻20 before 32% reduction), there is a reactive germinal center with small cleaved and noncleaved lymphocytes. (From Suarez, G. A., Giannini, C., Bosch, E. P., et al.: Immune brachial plexus neuropathy: suggestive evidence for an inflammatory-immune pathogenesis. Neurology 46:559, 1996, with permission.)

abnormalities.19 In the biopsies of individual nerves, fiber loss was present in most nerves (11 of 12 cases) that was focal in half of these patients. Perineurial thickening was noted in 9 of 12 nerves and perivascular inflammatory infiltrates in 5 cases, with features suggestive of microvasculitis in 3 (Fig. 102–2). In all of the brachial plexus biopsies there were perivascular epineurial or endoneurial inflammatory infiltrates. There appear to be

FIGURE 102–2 Paraffin section, stained with Luxol fast blue/periodic acid–Schiff, of a superficial radial nerve biopsy of a patient with IBPN. There is a mononuclear cell inflammatory infiltrate surrounding epineurial microvessels with probable involvement of the vessel wall (⫻50 magnification). (From Dyck, P. J. B., Englestad, J., Suarez, G., and Dyck, P. J.: Biopsied upper limb nerves provide information about distribution and mechanisms in immune brachial plexus neuropathy [abstract]. Neurology 56:A395, 2001, with permission.) See Color Plate

some pathologic similarities to well-described lower extremity plexopathies. Inflammatory infiltrates with microvasculitis have been observed in nondiabetic lumbosacral radiculoplexopathies as well as in DLSRP.21 It remains unknown why certain upper limb nerves and the brachial plexus are selectively affected in this disorder. It has been speculated that this selective anatomic vulnerability might be related to a permeable nerve-blood barrier

2306

Diseases of the Peripheral Nervous System

at specific branching points within the plexus or selective antigenic determinants in these anatomic areas. In summary, these observations support the hypothesis of an inflammatory-immune pathogenesis in IBPN.

TREATMENT Corticosteroids and immunomodulation with intravenous immunoglobulin or plasmapheresis have been used in patients with IBPN with some mixed results. Corticosteroids, especially intravenous methylprednisolone, may help to relieve the pain and stabilize the course of the illness, but definitive evidence of efficacy is lacking. In some of our patients, corticosteroids and immunomodulation appeared to be beneficial but treatment was uncontrolled, precluding definite conclusions.84 Other investigators have noted similar observations. Thus an adequately controlled clinical trial will be necessary to determine if immunotherapy is efficacious in IBPN.

PROGNOSIS In general, IBPN has a good prognosis for recovery. In the series of Tsairis et al., the rates of recovery, regardless of the severity or type of plexus lesion, were 36% within 1 year, 75% by the second year, and 89% at 3 years.86 Other large series have reported similar rates of recovery.13 Patients usually experience resolution of the pain several weeks after the onset of symptoms. The recovery of muscle weakness and atrophy depends on the extent and severity of the initial deficit and the distance between the site of the nerve lesion and its target muscle. Proximal muscles are reinnervated first, and this may explain why patients with upper trunk lesions may recover sooner than patients with other types of lesions. Superficial sensation typically returns concomitantly with motor recovery. Geertzen and co-workers reported a different outcome in a group of 16 patients with IBPN followed for an average of 8 years. Nine patients continued to complain of persistent pain and muscle weakness. Fifty percent of the patients had to change their occupation, and some continued to have significant impairment in activities of daily living.30 Recovery from diaphragmatic involvement may be slower and is more likely to be incomplete. In a study of 14 patients with diaphragmatic paralysis reassessed between 2 and 11 years after the onset, 50% of patients had significant recovery and only 2 remained impaired by dyspnea. However, the recovery rate was variable, with some evidence of improvement at 2 years and the earliest substantive recovery at 3 years for some patients. Elderly patients had less favorable recovery.35 In most cases IBPN is a monophasic illness with a good prognosis for full recovery. However, recurrences may occur and have been reported in 1% to 5% of cases.86,87 If this occurs in young patients or with repeated frequency, HBPN should be considered as a possible explanation.

REFERENCES 1. Allan, S. G., Towla, H. M., Smith, C. C., et al.: Painful brachial plexopathy: an unusual presentation of polyarteritis nodosa. Postgrad. Med. J. 58:311, 1982. 2. Allanore, Y., Zuber, M., and Kahan, A.: Brachial plexopathy associated with systemic sclerosis. Clin. Rheumatol. 21:401, 2002. 3. Augé, W. K. 2nd., and Velázquez, P. A.: Parsonage-Turner syndrome in the Native American Indian. J. Shoulder Elbow Surg. 9:99, 2000. 4. Bárdos, V., and Somodská, V.: Epidemiologic study of a brachial plexus neuritis outbreak in northeast Czechoslovakia. World Neurol. 2:973, 1961. 5. Beghi, E., Kurland, L. T., Mulder, D. W., and Nicolosi, A.: Brachial plexus neuropathy in the population of Rochester, Minnesota, 1970–1981. Ann. Neurol. 18:320, 1985. 6. Bernsen, P. L., Wong Chung, R. E., Vingerhoets, H. M., and Janssen, J. T.: Bilateral neuralgic amyotrophy induced by interferon treatment. Arch. Neurol. 45:449, 1988. 7. Beydoun, S. R., and Rodriguez, R.: Neuralgic amyotrophy misdiagnosed as diaphragmatic rupture. Muscle Nerve 19:1181, 1996. 8. Bloch, S. L., Jarrett, M. P., Swerdlow, M., and Grayzel, A. I.: Brachial plexus neuropathy as the initial presentation of systemic lupus erythematosus. Neurology 29:1633, 1979. 9. Bulgen, D. Y., Hazleman, B. L., and Warren, R. E.: Arthritis and neuralgic amyotrophy due to Yersinia enterocolitica. Br. Med. J. 1:1250, 1979. 10. Calabrese, L. H., Proffitt, M. R., Levin, K. H., et al.: Acute infection with the human immunodeficiency virus (HIV) associated with acute brachial neuritis and exanthematous rash. Ann. Intern. Med. 107:849, 1987. 11. Carmant, L., and Veilleux, M.: Anterior interosseous neuropathy in the postpartum period. Can. J. Neurol. Sci. 20:56, 1993. 12. Clay, S. A.: Osteomyelitis as a cause of brachial plexus neuropathy. Am. J. Dis. Child. 136:1054, 1982. 13. Cruz-Martinez, A., Barrio, M., and Arpa, J.: Neuralgic amyotrophy: variable expression in 40 patients. J. Peripher. Nerv. Syst. 7:198, 2002. 14. Cusimano, M. D., Bilbao, J. M., and Cohen, S. M.: Hypertrophic brachial plexus neuritis: a pathological study of two cases. Ann. Neurol. 24:615, 1988. 15. Cwik, V., Wilbourn, A. J., and Rorick, M.: Acute brachial neuropathy: detailed EMG findings in a large series. Muscle Nerve 13:859, 1990. 16. Dananchet, Y., and Besson, G.: Parsonage-Turner syndrome and giant-cell arteritis [in French]. Rev. Neurol. (Paris). 158:354, 2002. 17. Dill-Macky, M. J., Song, S., and Silbert, P. L.: Magnetic resonance imaging features of subacute idiopathic brachial neuritis. Australas. Radiol. 44:98, 2000. 18. Dumitru, D., and Liles, R. A.: Postpartum idiopathic brachial neuritis. Obstet. Gynecol. 73:473, 1989. 19. Dyck, P. J. B., Englestad, J., Suarez, G., and Dyck, P. J.: Biopsied upper limb nerves provide information about distribution and mechanisms in immune brachial plexus neuropathy [abstract]. Neurology 56:A395, 2001.

Immune Brachial Plexus Neuropathy 20. Dyck, P. J. B., Norell, J., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999. 21. Dyck, P. J. B., and Windebank, A. J.: Diabetic and nondiabetic lumbosacral radiculoplexus neuropathies: new insights into pathophysiology and treatment. Muscle Nerve 25:477, 2002. 22. England, J. D.: The variations of neuralgic amyotrophy. Muscle Nerve 22:435, 1999. 23. England, J. D., and Sumner, A. J.: Neuralgic amyotrophy: an increasingly diverse entity. Muscle Nerve 10:60, 1987. 24. English, P., and Maciver, D.: Neuralgic amyotrophy as a presenting feature of infective endocarditis. Postgrad. Med. J. 76:710, 2000. 25. Evans, P. A., and Millington, H. T.: Atraumatic brachial plexopathy following intravenous heroin use. Arch. Emerg. Med. 10:209, 1993. 26. Feinberg, J.: von Erg-Klumpke sher. Lahmung nach influenza. Centralblat 16:588, 1897. 27. Fibuch, E. E., Mertz, J., and Geller, B.: Postoperative onset of idiopathic brachial neuritis. Anesthesiology 84:455, 1996. 28. Flaggman, P. D., and Kelly, J. J. Jr.: Brachial plexus neuropathy: an electrophysiologic evaluation. Arch. Neurol. 37:160, 1980. 29. Gaitzsch, G., and Chamay, A.: Paralytic brachial neuritis or Parsonage-Turner syndrome anterior interosseous nerve involvement: report of three cases. Ann. Chir. Main 5:288, 1986. 30. Geertzen, J. H., Groothoff, J. W., Nicolai, J. P., and Rietman, J. S.: Brachial plexus neuropathy: a long-term outcome study. J. Hand Surg. Br. 25:461, 2000. 31. Glanzman, R. L., Gelb, D. J., Drury, I., et al.: Brachial plexopathy after botulinum toxin injections. Neurology 40:1143, 1990. 32. Gregory, R. P., Loh, L., and Newsom-Davis, J.: Recurrent isolated alternating phrenic nerve palsies: a variant of brachial neuritis? Thorax 45:420, 1990. 33. Hennessy, M. J., Koutroumanidis, M., and Elwes, R. D.: Neuralgic amyotrophy associated with hypersensitivity to lamotrigine. Neurology 51:1224, 1998. 34. Horowitz, H. W., Marks, S. J., Weintraub, M., and Dumler, J. S.: Brachial plexopathy associated with human granulocytic ehrlichiosis. Neurology 46:1026-1029, 1996. 35. Hughes, P. D., Polkey, M. I., Moxham, J., and Green, M.: Long-term recovery of diaphragm strength in neuralgic amyotrophy. Eur. Respir. J. 13:379-384, 1999. 36. Hyde, G. P., Postma, G. N., and Caress, J. B.: Laryngeal paresis as a presenting feature of idiopathic brachial plexopathy. Otolaryngol. Head Neck Surg. 124:575-576, 2001. 37. Katz, J. S., Saperstein, D. S., Wolfe, G., et al.: Cervicobrachial involvement in diabetic radiculoplexopathy. Muscle Nerve 24:794-798, 2001. 38. Kayed, K., and Kass, B.: Acute multiple brachial neuropathy and Ehlers-Danlos syndrome. Neurology 29:1620, 1979. 39. Kiloh, L., and Nevin, S.: Isolated neuritis of the anterior interosseus nerve. Br. Med. J. 1:850, 1952. 40. Kirchhoff-Moradpour, A., Huzly, D., Korinthenberg, R., and Berner, R.: Neuralgic amyotrophy associated with parvovirus B19 infection in a child. Eur. J. Pediatr. 160:200, 2001. 41. Kiwit, J. C.: Neuralgic amyotrophy after administration of tetanus toxoid. J. Neurol. Neurosurg. Psychiatry 47:320, 1984.

2307

42. Klein, C. J., Dyck, P. J., Friedenberg, S. M., et al.: Inflammation and neuropathic attacks in hereditary brachial plexus neuropathy. J. Neurol. Neurosurg. Psychiatry 73:45, 2002. 43. Lachance, D. H., O’Neill, B. P., Harper, C. M. Jr., et al.: Paraneoplastic brachial plexopathy in a patient with Hodgkin’s disease. Mayo Clin. Proc. 66:97, 1991. 44. Lahrmann, H., Grisold, W., Authier, F. J., and Zifko, U. A.: Neuralgic amyotrophy with phrenic nerve involvement. Muscle Nerve 22:437, 1999. 45. Lederman, R. J., and Wilbourn, A. J.: Postpartum neuralgic amyotrophy. Neurology 47:1213, 1996. 46. Leszynsky, W.: Bilateral neuritis of the brachial plexus following acute croupous pneumonia. N. Y. Med. J. 63:469, 1896. 47. Li, J., Loeb, J. A., Shy, M. E., et al.: Asymmetric flaccid paralysis: a neuromuscular presentation of West Nile virus infection. Ann. Neurol. 53:703, 2003. 48. Lo, Y. L., and Mills, K. R.: Motor root conduction in neuralgic amyotrophy: evidence of proximal conduction block. J. Neurol. Neurosurg. Psychiatry 66:586, 1999. 49. Loh, F. L., Herskovitz, S., Berger, A. R., and Swerdlow, M. L.: Brachial plexopathy associated with interleukin-2 therapy. Neurology 42:462, 1992. 50. Malamut, R. I., Marques, W., England, J. D., and Sumner, A. J.: Postsurgical idiopathic brachial neuritis. Muscle Nerve 17:320, 1994. 51. Marra, T. A.: Recurrent lumbosacral and brachial plexopathy associated with schistosomiasis. Arch. Neurol. 40:586, 1983. 52. Martinelli, P., Macri, S., Scaglione, C., et al.: Acute brachial plexus neuropathy as a presenting sign of peripheral nervous system involvement in paraproteinaemia. Acta Neurol. Scand. 95:319, 1997. 53. Martinelli, P., Pazzaglia, P., Marchiori, L., and Lugaresi, E.: Simultaneous occurrence of neuralgic amyotrophy in three members of one family. Eur. Neurol. 19:316, 1980. 54. Martinez Garcia, F. A., Martinez Vidal, S., Tortosa Conesa, D., et al.: Brachial plexitis as a manifestation of giant cell arteritis [in Spanish]. Rev. Neurol. 36:97, 2003. 55. Martinez-Salio, A., Porta-Etessam, J., Berbel, A., et al.: Amyotrophic neuralgia: review of 37 cases [in Spanish]. Rev. Neurol. 27:823, 1998. 56. Mouthon, L., Halimi, C., Dussaule, J. C., et al.: Brachial plexopathy associated with diffuse edematous scleroderma. Ann. Med. Interne (Paris). 151:303, 2000. 57. Mulvey, D. A., Aquilina, R. J., Elliott, M. W., et al.: Diaphragmatic dysfunction in neuralgic amyotrophy: an electrophysiologic evaluation of 16 patients presenting with dyspnea. Am. Rev. Respir. Dis. 147:66, 1993. 58. Murray, B., and Wilbourn, A. J.: Brachial plexus. Arch. Neurol. 59:1186, 2002. 59. Nardone, R., Bernhart, H., Pozzera, A., et al.: Respiratory weakness in neuralgic amyotrophy: report of two cases with phrenic nerve involvement. Neurol. Sci. 21:177, 2000. 60. Neundorfer, B., Claus, D., and Waller, D.: Brachial neuritis in salmonellosis. J. Neurol. 231:198, 1984. 61. Parsonage, M., and Turner, J.: Neuralgic amyotrophy: the shoulder-girdle syndrome. Lancet 1:973, 1948. 62. Patterson, D. L., DeRemee, R. A., and Hunt, L. W.: Severe asthma complicated by bilateral diaphragmatic paralysis attributed to Parsonage-Turner syndrome. Mayo Clin. Proc. 69:774, 1994.

2308

Diseases of the Peripheral Nervous System

63. Pellas, F., Olivares, J. P., Zandotti, C., and Delarque, A.: Neuralgic amyotrophy after parvovirus B19 infection. Lancet 342:503, 1993. 64. Pellegrino, J. E., George, R. A., Biegel, J., et al.: Hereditary neuralgic amyotrophy: evidence for genetic homogeneity and mapping to chromosome 17q25. Hum. Genet. 101:277, 1997. 65. Pellegrino, J. E., Rebbeck, T. R., Brown, M. J., et al.: Mapping of hereditary neuralgic amyotrophy (familial brachial plexus neuropathy) to distal chromosome 17q. Neurology 46:1128, 1996. 66. Pezzimenti, J. F., Bruckner, H. W., and DeConti, R. C.: Paralytic brachial neuritis in Hodgkin’s disease. Cancer 31:626, 1973. 67. Pierre, P. A., Laterre, C. E., and Van den Bergh, P. Y.: Neuralgic amyotrophy with involvement of cranial nerves IX, X, XI and XII. Muscle Nerve 13:704, 1990. 68. Pieters, T., Lambert, M., Huaux, J. P., and Nagant de Deuxchaisnes, C.: Hemidiaphragmatic paralysis, an unusual presentation of Parsonage-Turner syndrome. Clin. Rheumatol. 7:402, 1988. 69. Puechal, X., Hilliquin, P., Kahan, A., and Menkes, C. J.: Neuralgic amyotrophy and polyarthritis caused by parvovirus B19 infection. Ann. Rheum. Dis. 57:262, 1998. 70. Redmond, J. M., Cros, D., Martin, J. B., and Shahani, B. T.: Relapsing bilateral brachial plexopathy during pregnancy: report of a case. Arch. Neurol. 46:462, 1989. 71. Rennels, G. D., and Ochoa, J.: Neuralgic amyotrophy manifesting as anterior interosseous nerve palsy. Muscle Nerve 3:160, 1980. 72. Reutens, D. C., Dunne, J. W., and Leather, H.: Neuralgic amyotrophy following recombinant DNA hepatitis B vaccination. Muscle Nerve 13:461, 1990. 73. Richardson, J.: Serratus magnus palsy. Lancet 1:618, 1942. 74. Riggs, J. E., Schochet, S. S. Jr., and Hogg, J. P.: Focal rhabdomyolysis and brachial plexopathy: an association with heroin and chronic ethanol use. Mil. Med. 164:228, 1999. 75. Rubin, D. I.: Neuralgic amyotrophy: clinical features and diagnostic evaluation. The Neurologist 7:350, 2001. 76. Sampaio, C., Castro-Caldas, A., Sales-Luis, M. I., et al.: Brachial plexopathy after botulinum toxin administration for cervical dystonia. J. Neurol. Neurosurg. Psychiatry 56:220, 1993. 77. Scardigli, K., and Biller, J.: Acute inflammatory polyneuropathy and brachial plexopathy. IMJ Ill. Med. J. 168:36, 1985. 78. Schady, W., and Meara, R. J.: Brachial plexus neuropathy. Muscle Nerve 12:156, 1989. 79. Schott, G. D.: A chronic and painless form of idiopathic brachial plexus neuropathy. J. Neurol. Neurosurg. Psychiatry 46:555, 1983. 80. Sherrier, R. H., and Sostman, H. D.: Magnetic resonance imaging of the brachial plexus. J. Thorac. Imaging 8:27, 1993. 81. Shinder, N., Polson, A., Pringle, E., and O’Donnell, D. E.: Neuralgic amyotrophy: a rare cause of bilateral diaphragmatic paralysis. Can. Respir. J. 5:139, 1998. 82. Sierra, A., Prat, J., Bas, J., et al.: Blood lymphocytes are sensitized to branchial plexus nerves in patients with neuralgic amyotrophy. Acta Neurol. Scand. 83:183, 1991. 83. Spillane, J.: Localised neuritis of the shoulder girdle. Lancet 2:532, 1943.

84. Suarez, G. A., Giannini, C., Bosch, E. P., et al.: Immune brachial plexus neuropathy: suggestive evidence for an inflammatory-immune pathogenesis. Neurology 46:559, 1996. 85. To, W. C., and Traquina, D. N.: Neuralgic amyotrophy presenting with bilateral vocal cord paralysis in a child: a case report. Int. J. Pediatr. Otorhinolaryngol. 48:251, 1999. 86. Tsairis, P., Dyck, P. J., and Mulder, D. W.: Natural history of brachial plexus neuropathy: report on 99 patients. Arch. Neurol. 27:109, 1972. 87. Turner, J., and Parsonage, M.: Neuralgic amyotrophy (paralytic brachial neuritis) with special reference to prognosis. Lancet 2:209, 1957. 88. Tzur, A., and Shahin, R.: Bilateral brachial plexopathy after E. coli sepsis. Isr. J. Med. Sci. 33:687, 1997. 89. van Alfen, N., Schuuring, J., van Engelen, B. G., et al.: Idiopathic neuralgic amyotrophy in children: a distinct phenotype compared to the adult form. Neuropediatrics 31:328, 2000. 90. Vanneste, J. A., Bronner, I. M., Laman, D. M., and van Duijn, H.: Distal neuralgic amyotrophy. J. Neurol. 246:399, 1999. 91. Vriesendorp, F. J., Dmytrenko, G. S., Dietrich, T., and Koski, C. L.: Anti-peripheral nerve myelin antibodies and terminal activation products of complement in serum of patients with acute brachial plexus neuropathy. Arch. Neurol. 50:1301, 1993. 92. Watson, B. V., Nicolle, M. W., and Brown, J. D.: Conduction block in neuralgic amyotrophy. Muscle Nerve 24:559, 2001. 93. Watson, B. V., Rose-Innes, A., Engstrom, J. W., and Brown, J. D.: Isolated brachialis wasting: an unusual presentation of neuralgic amyotrophy. Muscle Nerve 24:1699, 2001. 94. Waxman, S. G.: The flexion-adduction sign in neuralgic amyotrophy. Neurology 29:1301, 1979. 95. Weikers, N. J., and Mattson, R. H.: Acute paralytic brachial neuritis: a clinical and electrodiagnostic study. Neurology 19:1153, 1969. 96. Weintraub, M. I., and Chia, D. T.: Paralytic brachial neuritis after swine flu vaccination. Arch. Neurol. 34:518, 1977. 97. Weller, S., Gartner, J., and Lenard, H. G.: Characteristic clinical features of idiopathic neuralgic amyotrophy in childhood. Neuropediatrics 32:110, 2001. 98. Wemeau, J., Montagne, B., and Hazzan Decarpentry, C.: Parsonage-Turner amyotrophic neuralgia in 2 heroin addicts [in French]. Presse Med. 26:165, 1997. 99. Wiederholt, W. C.: Hereditary brachial neuropathy: report of two families. Arch Neurol. 30:252, 1974. 100. Wilbourn, A. J.: Brachial plexus disorders. In Dyck P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 911, 1993. 101. Wilbourn, A. J.: Brachial plexopathies. In Brown, W. F., Bolton, C. F., and Aminoff, M. J. (eds.): Neuromuscular Function and Disease: Basic, Clinical, and Electrodiagnostic Aspects. Philadelphia, W. B. Saunders, p. 831, 2002. 102. Wittenberg, K. H., and Adkins, M. C.: MR imaging of nontraumatic brachial plexopathies: frequency and spectrum of findings. Radiographics 20:1023, 2000. 103. Wyburn-Mason, R.: Brachial neuritis occurring in epidemic form. Lancet 2:662, 1941.

103 Nonmalignant Inflammatory Sensory Polyganglionopathy BENN E. SMITH, ANTHONY J. WINDEBANK, AND PETER J. DYCK

Overview Historical Review Clinical Features Patterns

Symptoms and Signs Course Laboratory Features Radiologic Findings

OVERVIEW The dorsal root ganglion cell is one of the most remarkable neurons in the peripheral nervous system. Compared with the range of dimensions of other nerve cells, this primary sensory neuron often spans great distances. Afferent cells innervating pacinian corpuscles in the hallux stretch from the great toe to cell bodies in the proximity of the L5 intervertebral foramen, and from there, without synapse, enter the spinal cord through the posterior root and pass on into the gracile fasciculus, ascending to their first synapse in the nucleus gracilis at the craniocervical junction. In tall individuals this single cell may therefore span 2 m or more. While Santiago Ramon y Cajal labeled the dorsal root ganglion cell by its anatomic segments, namely the peripheral sensory axon, ganglion cell body, posterior root, and central sensory axon,74 different categories of disease appear to affect each dorsal root ganglion cell segment in isolation and occasionally in combination: 1. Peripheral sensory axon : sensory polyneuropathy 2. Ganglion cell body : sensory polyganglionopathy (sensory neuronopathy) 3. Posterior root : sensory polyradiculopathy 4. Central sensory axon : central sensory axonopathy (posterior column disease) 5. Peripheral axon and posterior root ( ganglion cell body) : sensory polyradiculoneuropathy

Electrophysiologic Findings Pathologic Findings Differential Diagnosis Treatment

Beginning distally, the main categories of dorsal root ganglion cell disease are sensory polyneuropathy, sensory polyganglionopathy, sensory polyradiculopathy, and central sensory axonopathy (also called posterior column disease). Multiple segments of the primary sensory neuron may be involved, leading to sensory polyradiculoneuropathy. The frequency of each of these five categories of disease, and perhaps our ability to recognize and distinguish them from one another, appear to lessen in a distal-to-proximal gradient. At the ends of the anatomic spectrum, distal sensory polyneuropathies are exceedingly common while central sensory axonopathies are considered to be quite rare. Disorders that rarely may primarily affect posterior column fibers are beyond the scope of this chapter, but include Sjögren syndrome, vitamin B12 deficiency, posterior column compressive myelopathy (associated with C3,4 spondylosis or thoracic meningioma), and multiple sclerosis.80 Many polyneuropathies affect sensory, motor, and sometimes autonomic fibers when the neuropathy has reached a moderate or advanced degree of clinical deficit. Several neuropathies that demonstrate this pattern present as a pure or predominantly sensory neuropathy at the beginning of their course, when neuropathic deficits are mild. Such a predominantly sensory polyneuropathy occurs in certain infections (e.g., Hansen’s disease [see Chapter 91], human immunodeficiency virus [HIV] infection [see Chapter 94], Lyme disease [see Chapter 92], and diphtheria [see Chapter 95]); metabolic derangements (e.g., diabetes mellitus [see Chapter 85], amyloidosis [see Chapter 108], 2309

2310

Diseases of the Peripheral Nervous System

uremia [see Chapter 87], hepatic failure [see Chapter 87], and hypothyroidism [see Chapter 88]); intoxications (e.g., arsenic, ethanol, cisplatin, and other toxins [see Chapters 112 through 114]); altered immunity (see Chapters 98 and 99); dysproteinemias (see Chapter 100); paraneoplastic syndromes (see Chapter 110); and inherited neuropathies (see Chapters 69 through 84). Sensory polyneuropathy can be further divided by the class of neurons (axons) affected (large, small, or both). Large-diameter fiber sensory neuropathy is characterized by kinesthetic sensory loss and ataxia. Small-diameter fiber sensory neuropathy may be painful and associated with pain and temperature sensory loss. In pure sensory polyradiculopathy, multiple posterior spinal roots are affected. This may occur in some cases of inflammatory demyelination and sometimes in lympho- or plasma-proliferative malignancies. In most cases with sensory polyradiculopathy, one is uncertain whether spinal ganglia, peripheral sensory nerves, or central sensory axons are not also affected. It is also difficult to make such distinctions using routine clinical and electrophysiologic tests alone. Pathologic studies that could resolve the issue can seldom be done. Sensory polyradiculopathy (with or without involvement of spinal ganglia and peripheral nerves) associated with motor polyradiculopathy occurs in acute (Guillain-Barré syndrome) and chronic (CIDP) inflammatory demyelinating polyradiculoneuropathies, diabetes mellitus, infections such as Lyme disease and HIV infection, lymphomatous and leukemic infiltrations, and meningeal carcinomatosis. In sensory polyganglionopathy, the major site of pathologic involvement is the spinal ganglion. Spinal ganglia may be preferentially affected in hereditary sensory and autonomic neuropathies (HSANs) and spinocerebellar degenerations (see Chapter 78), cancers (neoplastic infiltration and as a remote effect [see Chapters 110 and 111]), intoxications (see Chapters 112 through 114), and nonmalignant immune disorders of spinal ganglia. The subject of this chapter is the nonmalignant inflammatory sensory polyganglionopathies (NISP). We use the term nonmalignant inflammatory sensory polyganglionopathy because it is simple and parallels the terms used for roots (polyradiculoneuropathy) and nerves (polyneuropathy).

HISTORICAL REVIEW Patients with selective severe and widespread loss of sensation have been reported for more than a century. In 1858, Duchenne described patients with progressive incoordination of movement and preservation of muscular force as having l’ataxie locomotrice (locomotor ataxia).25 Hippocrates used the term tabes dorsalis (“wasting of the back”) to describe symptoms attributed to venereal excess.1 Stanley, in 1840, noted that incoordination was associated with posterior column disease.87 In 1847, Todd distinguished cases of incoordination without weakness from those with paraplegia and incoordination, inferring that posterior column

disease would explain such a dissociation, and verifying his assertion with two autopsy cases.92 Romberg described tabes dorsalis in 1851 but did not exclude weakness from the clinical findings.75 Gowers wrote in 1888 that syphilitic tabes dorsalis was a common chronic disease associated with degeneration in the posterior columns of the spinal cord, sensory roots, or both.35 He noted that other conditions sometimes presented with a similar profound loss of “afferent muscle-nerve” function, including diphtheria, arsenical poisoning, and alcoholic pseudotabes. As syphilis and its neurologic complications decreased in prevalence, diseases primarily affecting sensory peripheral nerve structures became less common. Denny-Brown described two cases with rapidly progressive severe sensory loss and ataxia without weakness, associated with myositis and bronchogenic carcinoma.22 Many similar cases have subsequently been reported.5,37,42 Wartenberg’s review of sensory neuritis discussed cases from the French literature with a sensory form of the ascending paralysis described by Guillain and Barré and Strohl. He used the terms ascending posterior polyradiculitis and sensory neuronitis to describe patients with acute or subacute ascending paresthesia.94 Denny-Brown later described “hereditary sensory radicular neuropathy,” characterized by chronic progressive symmetrical sensory loss and ulcers of the feet.23 He distinguished this condition from two similar disorders: paraneoplastic pure sensory neuropathy and pure sensory neuropathy, possibly inflammatory, related to Guillain-Barré syndrome. Various pharmacologic and industrial toxins are known to cause a sensory neuropathy. These include arsenic,56 doxorubicin,14 ethanol,93 metronidazole,7 misonidazole,95 cisplatin,55 pyridoxine,78 N-3-pyridylmethyl-N-p-nitrophenyl urea (PNU),72 paclitaxel,71 thalidomide,31 and tainted L-tryptophan.81 In 1968, Dyck et al.28 described the clinical and pathologic features of two cases of acute pure sensory neuropathy not associated with malignancy, similar to those in the entity discussed by Wartenberg,94 and concluded that multifocal radicular sensory loss with selective loss of large-diameter sural nerve fibers suggested dorsal root ganglia as the locus of disease. Sterman et al. subsequently reported three patients in whom no cancer was identified, who presented with subacute widespread paresthesia and dysesthesia, ataxia, severe loss of sensation, and poor recovery. They postulated lesions confined to dorsal root and gasserian ganglia, noting the similarity to the pattern of experimental pyridoxine and doxorubicin intoxications.88 Dalakas provided a longitudinal study of 15 patients with slowly progressive distal paresthesias and sensory ataxia with profound loss of largediameter fiber sensation, areflexia, and normal strength in whom neither malignancy nor antineuronal antibodies were found.19 Four of these patients had a monoclonal gammopathy. Dalakas suggested that immune mechanisms might have played a role. In 1990, Windebank et al. presented the clinical, laboratory, electrophysiologic, and pathologic findings in

2311

Nonmalignant Inflammatory Sensory Polyganglionopathy

42 patients with acute or subacute sensory neuropathy, discussing the natural history and response to treatment in individual cases.96 They concluded that the acute or subacute, often focal, onset of symptoms in the absence of inflammation in distal peripheral nerve suggests an immune-mediated or vascular process at the level of the posterior root or dorsal root ganglion. In another series published the same year, Griffin et al. described 13 patients with acute, subacute, or chronic ataxic sensory neuropathy, usually with autonomic features, associated with Sjögren syndrome.38 They reported clinical, laboratory, and electrodiagnostic data as well as sural nerve, posterior root, and dorsal root ganglion pathology, emphasizing the differences between their patients and those with either idiopathic or malignant sensory polyganglionopathy, and advocated that Sjögren syndrome be considered in patients with severe sensory and autonomic neuropathy with ataxia and proprioceptive sensory loss. Sobue and colleagues presented 11 patients with chronic progressive sensory ataxic neuropathy, 3 associated with Sjögren syndrome and 8 idiopathic. The authors delineated the symptoms, signs, and laboratory abnormalities in these subjects, including nerve biopsy results in eight and autopsy findings in one.85 Lauria and co-workers contributed 19 patients with NISP, focusing on the magnetic resonance imaging (MRI) findings in their cohort (see below).54 Additional series and case reports add to the literature total of more than 150 patients with detailed descriptions of acute, subacute, or chronic sensory neuropathy without cancer.4,10,13,16,32,41,44,47,48,50,52,53,59,62,63,66,70,84,97 Although associated medical conditions have not been identified in many of these patients, paraproteinemia,19,54,96 Sjögren syndrome,10,13,19,38,44,52,53,59,62,63,65,66,69,70,85,96 and celiac disease16 have been reported in some. This chapter discusses disorders of the dorsal root and spinal ganglia that are thought to be immune mediated, excluding cases associated with malignancy or toxins affecting primary sensory neurons. With these exclusions, the remaining conditions affecting proximal peripheral divisions of the dorsal root ganglion cell appear to be inflammatory. In the few cases in which tissue has been obtained, inflammation was prominent, implying immune mechanisms. Whether these disorders affect root, ganglion cell, segmental nerve, or combinations of these levels awaits further detailed study of patients with these disorders. Four lines of evidence may be useful in determining the localization of the disease process, which presumably is inflammation: (1) clinical, (2) electrophysiologic, (3) radiologic, and (4) pathologic.

which a patient falls into one or another type appears to depend on the population of neurons (axons) affected, the rapidity of involvement, and perhaps other factors. In the ataxic syndrome, predominantly large-diameter neurons (axons) are affected, giving rise to proprioceptive deficits and sensory ataxia. In the hyperalgesic syndrome, involvement of small-diameter neurons (axons) mediating pain and temperature sensation is assumed. In the latter syndrome, there may also be a variable degree of autonomic involvement. The majority of NISP patients, however, have combinations of ataxia and pain or dysesthesia. NISP patients presenting with ataxia were described by Sterman et al.,88 Dalakas,19 Windebank et al.,96 Griffin et al.,38 Kaufman et al.,47 Sobue et al.,85 Lauria et al.,54 and others.41,44,48,50,59,84 NISP patients with prominent hyperalgesia were discussed by Dyck et al.,28 Mitsumoto et al.,62 Windebank et al.,96 and Laloux et al.53

Symptoms and Signs In the ataxic type of NISP, presenting symptoms are numbness and tingling, unsteadiness of gait, diminished fine motor control, and occasionally pain, although this latter symptom is rarely prominent. In patients with the hyperalgesic type, shooting limb pains, burning numbness, and painful paresthesia sometimes are among the presenting symptoms. Patients with ataxic and hyperalgesic NISP develop difficulty walking, poor coordination, and loss of manual dexterity. In hyperalgesic NISP, pain, burning sensations, lancinating pains, and dysesthesia sometimes predominate. Neurologic examination typically shows normal strength, varying degrees of both distal and proximal diminished sensation (particularly vibration and joint position sense), and regional or generalized hypo- or areflexia. In severely affected individuals, the deafferentation is often so profound that limbs drift off into space without the patient knowing. In certain cases, nocturnal eye protection is needed to prevent inadvertent ocular injury from a wandering finger. This phenomenon is known as pseudoathetosis, and may initially be evidenced by slow, irregular movements of the outstretched fingers when the subject’s eyes are closed and the upper limbs extended. Adie pupils, loss of sinus arrhythmia, or abnormal Valsalva ratio values are sometimes apparent.38 Orthostatic hypotension, more commonly associated with malignant inflammatory sensory polyganglionopathy, can also occur in NISP.85

Course CLINICAL FEATURES Patterns Although there is considerable clinical heterogeneity, three major presentations occur—an ataxic syndrome, a hyperalgesic syndrome, and a mixture of the two. The extent to

Considering two of the largest series of cases, the onset and tempo of worsening are quite variable. Some patients present with a subacute or chronic progressive course, whereas others have an explosive onset over hours. The course is most often chronic and progressive, as reported in 20 of 38 patients by Windebank et al.96 and in 8 of 13 patients by

2312

Diseases of the Peripheral Nervous System

Griffin et al.38 In others the disease remains unchanged (10 of 38 patients96 and 4 of 8 patients38). Still others had a chronic or progressive monophasic course (8 of 38 patients96). One of Griffin et al.’s patients followed a chronic relapsing course. Mean follow-up periods for these series were 98 months (range 12 to 324 months)96 and 41 months (range 5 to 190 months).38

LABORATORY FEATURES Most NISP patients have normal to slightly elevated cerebrospinal fluid (CSF) protein values. Of the 15 patients reported by Dalakas, the mean CSF protein value was 42 mg/dL (range 19 to 128).19 In Windebank’s series, the mean value in 30 patients was 48.5 mg/dL (range 17 to 146), although 2 patients who presented acutely were found to have readings of 300 and 290 mg/dL.96 On repeat CSF analysis later in the course of disease, both of these individuals had normal CSF protein determinations. Of Sobue’s 10 patients, the mean CSF protein value was 39 mg/dL (range 10 to 67), with 3 patients in the abnormal category.85 Pleocytosis is not considered a typical feature of inflammatory demyelination, and when present suggests infiltrative or infectious conditions (e.g., HIV infection).17 One of Windebank et al.’s patients who presented acutely, however, initially had 115 lymphocytes/mm3 in the CSF, with a subsequent lumbar puncture showing a normal cell count.96 Elevated CSF protein in the absence of pleocytosis, as seen in the Guillain-Barré syndrome and CIDP, is thought to be associated with spinal or meningeal inflammation.2,29 In NISP, the CSF protein may be normal or slightly elevated and is most often without pleocytosis. In subacute monophasic cases, abnormal findings were restricted to the initial phase. Serologic markers (antinuclear antibodies, extractable nuclear antigens, rheumatoid factor, and antineutrophil cytoplasmic antibodies) may suggest specific connective tissue disorders. Many patients with NISP and Sjögren syndrome tested positive for extractable nuclear antigens such as ro (SS-A) and la (SS-B).38 Dalakas found four individuals with ataxic sensory findings suggesting NISP and paraproteinemia, three with immunoglobulin (Ig) M  and one with IgA  monoclonal proteins.19 Windebank et al. found one case each of IgM  and IgG  monoclonal paraproteinemia in their NISP cohort.96 The vast majority of patients with sensory polyganglionopathy and antineuronal antibodies either had previously known malignancy or subsequently were found to have cancer.9,18,49 Antineuronal antibodies have rarely been reported in NISP associated with Sjögren syndrome,65 although crossreactivity between antineuronal nuclear antibodies and the Ro-52K antigen, the latter a major antigen recognized by sera of some individuals with Sjögren syndrome, may account for at least some of these cases.60,73 Sound clinical

practice dictates that any patient with antineuronal nuclear antibodies must be considered to have undetected malignancy.64

RADIOLOGIC FEATURES Bang and colleagues reported a case of acute sensory ganglionopathy with diffuse MRI gadolinium enhancement of spinal roots, but with no appreciable intramedullary signal changes.4 Lauria et al. detected increased MRI T2-weighted signal changes in the posterior columns of 15 patients with chronic sensory ganglionopathy, reflecting involvement of the central sensory axons.54 Mori et al. demonstrated similar changes in the posterior columns of 12 of 14 patients with NISP associated with Sjögren syndrome.66 These findings are similar in appearance to those reported by Smith and Bosch in patients with isolated disease of the dorsal column pathways.80

ELECTROPHYSIOLOGIC FINDINGS The most common abnormality of nerve conduction in NISP is absent or low-amplitude sensory nerve action potentials (SNAPs) with normal or relatively preserved motor (compound muscle action potential) amplitudes. Needle electromyography (EMG) is usually normal, although occasionally it demonstrates mild abnormalities, indicating minimal involvement of motor neurons/axons. Of 41 patients in Windebank et al.’s series, all but 1 showed a decrease in the amplitude of SNAPs with only mild slowing of conduction velocity.96 No SNAP could be elicited from the majority of nerves tested. Only eight of these patients had sensory conduction velocities below the range of normal, five from the distal median nerve. Motor nerve conduction studies were abnormal much less often than sensory studies. Ten of the 41 patients had abnormal compound muscle action potential amplitudes or conduction velocities, most commonly in the median and fibular (peroneal) nerves. EMG was usually normal. Eleven of the patients had fibrillation potentials in intrinsic foot muscles, while an occasional patient demonstrated fibrillations in a leg or intrinsic hand muscle. Motor unit potential changes were more prominent, with abnormalities in 18 patients. In Griffin et al.’s series, 13 patients showed absent SNAPs in most of the clinically involved limbs; 7 of the patients had absent SNAPs in all nerves tested.38 Motor conduction abnormalities were limited to mild alterations in compound muscle action potential amplitudes and motor conduction velocities. EMG showed only minimal evidence of denervation in isolated muscles. Knazan et al. described a 19-year-old man who, after a “flu-like” illness, developed abrupt onset of severe sensory

Nonmalignant Inflammatory Sensory Polyganglionopathy

ataxia with marked disturbances of proprioception and vibration sense, minimal involvement of cutaneous sensation, areflexia, and normal strength.50 Extensive electrophysiologic testing revealed normal motor conductions in the median, ulnar, and posterior tibial nerves; absent SNAPs on stimulating the median, ulnar, or sural nerves; and normal EMG of upper and lower limb muscles. An H reflex was not elicited. Median and tibial somatosensory evoked potentials could be elicited, although the median N20 scalp response was prolonged and the tibial P40 waveform was either absent or markedly prolonged. Auger and colleagues reviewed the electrodiagnostic records of 60 patients with sensory polyganglionopathy using trigeminal “blink reflex” testing; unexpectedly, they found normal results in all 17 individuals with malignant inflammatory sensory polyganglionopathy and abnormal studies in 20 of 43 patients with NISP. The authors concluded that an abnormal “blink reflex” study favors a nonmalignant etiology for sensory polyganglionopathy.3 The absence of SNAPs argues for involvement at or below the level of the dorsal root ganglion. In isolated disease of spinal roots or more proximal sensory pathways, such as in traumatic root avulsion and disorders affecting central sensory pathways, SNAPs remain normal.15,80

PATHOLOGIC FINDINGS Morphologic study of the primary sensory neuron has focused on the peripheral (sural) nerve, with only a few reports of spinal ganglia or dorsal root alterations. Cellular infiltrates have been demonstrated in biopsy and autopsy material from a number of patients with clinical features of NISP. In the sural nerves of two patients, Dyck et al. found decreased myelinated fiber (MF) density (mean 4247/mm2, range 2286 to 6209) primarily affecting large MFs, a low rate of axonal degeneration, and no interstitial or endoneurial inflammation.28 These findings are consistent with but do not directly demonstrate disease of the spinal ganglia. Windebank et al. reported a pathologic study of 22 sural nerves.96 Although the mean MF density was significantly decreased at 4327 MFs/mm2 (range 52 to 9968) compared with the mean value from 58 controls (8451 MFs/mm2; range 5101 to 12995), no difference was found in the median MF diameter (5.32 m in patients vs. 5.20 m in controls). When the area of axons and myelin was assessed separately, however, striking differences were found. The median axon area was 4.24 m2 (range 2.19 to 7.71) compared with 7.09 m2 (range 3.15 to 12.63) in controls. The median myelin area was 17.79 m2 (range 6.81 to 24.22) compared with 14.20 m2 (range 7.78 to 29.55) in controls. Teased fiber analysis showed that the average percentage of normal fibers (condition A) was lower (58.6%, range 0 to 92%, vs. 89.4%, range 62% to 100%), the percent of fibers with de- and remyelination (conditions C, D,

2313

F, or G) was higher (11.1%, range 0 to 54%, vs. 1.6%, range 0 to 11%), and the percentage of teased fibers undergoing axonal degeneration (conditions E or H) was higher (29.2%, range 0 to 98%, vs. 8.5%, range 0 to 27%) than in normal controls. A small mononuclear cell perivascular infiltrate was found in 1 of 22 sural biopsy specimens. When taken with clinical and electrophysiologic findings, Windebank et al.’s studies showing axonal atrophy suggest a more proximal site of disease (e.g., spinal ganglia). Twelve of Griffin et al.’s 13 patients with Sjögren syndrome underwent sural nerve biopsy.38 Five had reduced or severely reduced MF density. The mean MF density of the other six was 3263 MFs/mm2 (range 688 to 7598). They found inflammatory cells around epineurial blood vessels in six sural nerve specimens. Griffin et al. reported the thoracic dorsal root and spinal ganglion biopsy findings in three patients.38 All three demonstrated loss of fibers in the dorsal roots. Inflammation was present in all three cases, with infiltrates around blood vessels in two cases and scattered in small numbers throughout the endoneurium. In the three dorsal root ganglia specimens, prominent mononuclear inflammatory cell infiltrates were noted around individual sensory ganglion cells. In one patient, immunohistochemical staining showed most of the inflammatory cells to be T lymphocytes, with occasional macrophages. Neuronal loss varied from case to case and appeared to relate to the presence of T cells and macrophages. We observed similar findings in lumbar dorsal root ganglion biopsy in another case of NISP associated with Sjögren syndrome. The posterior root appeared to be relatively unaffected but the sural nerve showed a markedly reduced density of myelinated fibers (Fig. 103–1). Sobue et al. reported detailed quantitative sural nerve biopsy findings in eight patients of their series on chronic NISP, two with Sjögren syndrome and six with no identifiable cause.85 The mean MF density of these eight biopsies was 2789 MFs/mm2 (range 694 to 5234). There was, however, a striking difference between the density of large fibers, at 309 MFs/mm2 (range 0 to 843), and small fibers, at 2580 MFs/mm2 (range 694 to 4662), with normal means being 3068 (standard deviation [SD] 294) and 5122 (SD 438), respectively.85 In addition, Sobue et al. performed quantitative studies on unmyelinated fiber in four cases, estimating a mean density of 30,762 unmyelinated fibers/mm2 (range 21,150 to 40,110), a value not significantly different from normal controls, with a mean density of 29,913 (SD 3457). Osmicated teased fiber studies of at least 50 fibers from seven individuals showed 89.7% to be normal (range 75.8% to 96.5%), 4.3% to indicate axonal degeneration (range 0 to 16.7%), and 6.0% to suggest segmental demyelination (range 0 to 15.1%). Paraffin sections of the sural nerve did not demonstrate inflammatory infiltrates or changes indicating vasculitis.85 In the same publication (as well as a previous one), these workers also

2314

Diseases of the Peripheral Nervous System

FIGURE 103–1 Top, Lumbar dorsal root ganglion section in a patient with NISP and Sjögren syndrome showing a mononuclear cell infiltrate in the interstitium and around ganglion cells. (Hematoxylin and eosin–stained paraffin section of aldehyde-fixed tissue; 25). Bottom, Transverse section of the sural nerve in the same patient with NISP demonstrating decreased density of myelinated fibers with a shift toward smaller diameter categories. (Paraphenylenediamine-stained semithin [0.75-m] transverse epoxy section of aldehyde- and osmium tetroxide–fixed sural nerve; 25).

described autopsy findings in a single case of NISP.84,85 In addition to “moderate to severe cellular infiltrates” in liver, stomach, lung, kidney, and thyroid, similar inflammatory changes were detected in proximal nerve trunks and dorsal root ganglia.84,85 Because of the risks involved and for technical reasons, spinal ganglion biopsy should be restricted to only a very select group of patients with severe disease in whom malignant involvement has to be ruled out.76 It should be done only in highly specialized medical institutions.

DIFFERENTIAL DIAGNOSIS The differential diagnosis involves separating NISP from other sensory neuropathies (see below) and from other sensory ganglioradiculoneuropathies (Table 103–1). A variety of sensory polyneuropathies must first be considered. They include (1) metabolic, inflammatory-demyelinating, toxic, deficiency, and other syndromes beginning with sensory manifestations but affecting other classes of neurons (axons) (e.g., motor and autonomic) at more advanced

Nonmalignant Inflammatory Sensory Polyganglionopathy

2315

Table 103–1. Differential Diagnosis of Sensory Polyradiculopathy/Polyganglionopathy/Polyganglioradiculoneuropathy Sensory Polyradiculopathy/ Polyganglionopathy/ Polyganglioradiculoneuropathy

History

Physical Findings

Laboratory Abnormalities

Malignant inflammatory sensory polyganglionopathy (MISP)

Smoking history, weight loss, may have autonomic/CNS symptoms

Antineuronal nuclear antibodies, abnormal chest imaging; lowamplitude sensory nerve action potentials; autonomic testing abnormalities

Nonmalignant inflammatory sensory polyganglionopathy (NISP), idiopathic

Hyperacute, acute, subacute, or chronic presentations

Nonmalignant inflammatory sensory polyganglionopathy (NISP), Sjögren associated

Xerophthalmia, xerostomia

Toxic sensory polyganglionopathy Pyridoxine Cisplatin Paclitaxel Semisynthetic penicillins (?)

Relevant exposure

Acute inflammatory demyelinating polyradiculoneuropathy (sensory variant)

Acute or subacute onset;  antecedent febrile illness

Distal and proximal upper and lower limb symmetrical or asymmetrical hypesthesia, loss of deep tendon reflexes, normal strength, sensory ataxia, pseudoathetosis Distal and proximal upper and lower limb symmetrical or asymmetrical hypesthesia, loss of deep tendon reflexes, normal strength, sensory ataxia, pseudoathetosis Distal and proximal upper and lower limb symmetrical or asymmetrical hypesthesia, loss of deep tendon reflexes, normal strength, sensory ataxia, pseudoathetosis Distal and proximal upper and lower limb symmetrical or asymmetrical hypesthesia, loss of deep tendon reflexes, normal strength, sensory ataxia, pseudoathetosis Distal  proximal ascending findings, typically symmetrical

Chronic inflammatory demyelinating polyradiculoneuropathy (sensory variant)

Chronic progressive course

Distal  proximal deficits, typically symmetrical

IgM paraproteinemic neuropathy

Chronic progressive course

Distal  proximal deficits, typically symmetrical, sensory ataxia

No distinctive features; low-amplitude sensory nerve action potentials (must aggressively exclude serologic and imaging evidence of occult malignancy) Elevated extractable nuclear antigen antibodies; positive Schirmer test; minor salivary gland inflammation; lowamplitude sensory nerve action potentials Normal CSF protein; lowamplitude sensory nerve action potentials

CSF cytoalbuminologic dissociation; nerve conductions suggesting demyelination; low-amplitude sensory nerve action potentials CSF cytoalbuminologic dissociation; nerve conductions suggesting demyelination; low-amplitude sensory nerve action potentials IgM monoclonal paraproteinemia; low-amplitude sensory nerve action potentials

CNS  central nervous system; CSF  cerebrospinal fluid; IgM  immunoglobulin M.

stages; (2) inherited neuropathies; (3) paraneoplastic sensory neuropathies; (4) toxic syndromes; (5) deficiency states; and (6) infections. As indicated above, many polyneuropathies begin with sensory symptoms and initial absence of motor or autonomic findings on physical examination. Later, motor and/or autonomic symptoms and signs may develop. This is the case for a number of polyneuropathies. The predilection for distal sensory involvement occurs in polyneuropathies associated with various metabolic conditions, including diabetes mellitus,8 hypothyroidism,33 primary biliary cirrhosis,91 and uremia90; in infectious polyneuropathies such as those reported with diphtheria,36 HIV,17 leprosy,20 and Lyme disease40; in inflammatory-dysimmune diseases, including dysproteinemias,82 nonsystemic vasculitis,27 and Sjögren syndrome61; in

inherited neuropathies such as Fabry disease,51 HSAN types I,43 II,68 IV,89 and V,58 inherited amyloidosis,83 and spinocerebellar degenerations30; and in toxic polyneuropathies such as those encountered with exposure to arsenic,56 dichlorophenoxyacetic acid,34 dimethylaminopropionitrile,79 ethylene oxide,39 gold,46 hexacarbons,86 isoniazid,6 misonidazole,95 phenytoin,57 platinum,55 polychlorinated biphenyls,67 PNU,72 thallium,12 tainted L-tryptophan,81 and vincristine.11 The predilection for early distal sensory involvement is usually explained by length-dependent involvement of peripheral nerve axons and the greater length of sensory versus motor axons. In evaluating a patient with symmetrical sensory symptoms and findings, the first decision to be made is whether the process is inherited or not. If inherited neuropathy can

2316

Diseases of the Peripheral Nervous System

be excluded, the list of possibilities is greatly reduced in length. The features that suggest that a sensory neuropathy may be inherited might be summarized as age at onset, course, nature of symptoms, population of neurons (axons) affected, type of pathologic alteration, kinship information, and associated laboratory findings. Inherited sensory neuropathy may be present at birth (HSAN types II through V [see Chapter 78]) or develop during the second or later decades. Perhaps more typical than age of onset is the slowly insidious development over many years that characterizes HSAN type I, spinocerebellar degeneration, Fabry disease, and perhaps Tangier disease. Generally patients with inherited sensory neuropathies do not experience paresthesia—the presence of prominent paresthesia therefore favors an acquired sensory neuropathy. Sometimes in inherited neuropathy, one class of neurons (axons) is especially vulnerable to being affected. Acquired sensory neuropathies tend to involve most classes of neurons. The nature of the differences in the pathologic alterations in inherited and acquired neuropathies is outlined in Chapter 31. A history of a similar disorder among kin or observing bony abnormalities or other physical findings (e.g., yelloworange discoloration of the pharyngeal tonsils in Tangier disease or angiokeratosis of the buttock, scrotum, and perineum in Fabry disease) or laboratory abnormalities (e.g., hypocholesterolemia and abnormal serum lipoprotein electrophoresis in abeta- and hypobetalipoproteinemia or low leukocyte -galactosidase in Fabry disease) suggesting a specific variety of inherited neuropathy obviously helps in identifying inherited neuropathy. Having established that the patient’s sensory symptoms are not likely to be inherited, the next step is to decide whether the sensory symptoms represent the onset of a more generalized polyneuropathy. General medical information about a patient provides special clues. Sensory symptoms associated with diabetes mellitus, uremia, hypothyroidism, acromegaly, hepatic failure, or severe hypoglycemic attacks are usually related to the systemic disease. The course and pattern of the symptomatology may suggest heavy metal intoxications, vitamin B12 deficiency, or a variety of other disorders. Clinical signs might also prove helpful, including microvascular or proliferative retinopathy (diabetes mellitus), macroglossia (amyloidosis, hypothyroidism), or sluggish deep tendon reflexes (hypothyroidism). Useful laboratory investigations include elevated fasting blood glucose and glycosylated hemoglobin (diabetes mellitus), increased thyroid-stimulating hormone (hypothyroidism), elevated serum creatinine (uremia), and serum or urine monoclonal gammopathy (paraproteinemia, amyloidosis). The paraneoplastic sensory neuronopathies associated with small cell bronchogenic carcinoma are not readily distinguished from nonmalignant inflammatory sensory polyganglionopathy. A history of smoking, constitutional symptoms (e.g., weight loss, fever, lymphadenopathy), or

findings of malignancy suggest the paraneoplastic disorders. Particularly in persons 50 years and older, one may wish to obtain a chest radiograph and an x-ray metastatic bone survey, assess for monoclonal proteins in plasma and urine, and, if appropriate, obtain bone marrow, nerve, or other tissue to make a diagnosis of myeloma (multiple or osteosclerotic), amyloidosis (familial or acquired), or monoclonal gammopathy of undetermined significance. Trigeminal “blink reflex” studies are often normal in paraneoplastic/malignant inflammatory sensory polyganglionopathy, but may be abnormal in patients with the nonmalignant variety.3 Any individual found to have a circulating antineuronal nuclear antibody (anti-Hu) must be presumed to harbor occult malignancy. The diagnosis of sensory ganglionopathy and discovery of antibodies may precede the diagnosis of cancer by months or years.9,49 Although there are few cases in the literature that describe antineuronal antibodies in patients with Sjögren syndrome,65 it should be noted that there is sometimes cross-reactivity between these antibodies and extractable nuclear antigen antibodies commonly associated with Sjögren syndrome.60,73 More importantly, a diligent search for neoplasm, including high-resolution computed tomography scanning or MRI of the chest, is indicated in such individuals. If a tumor is not found, an argument can be made for repeating the chest imaging every 3 to 6 months, or sooner should other symptoms intervene, for the purpose of discovering and treating the highly probable occult malignancy.73 Toxic syndromes may be suggested by their occurrence in clusters, exposure to putative toxins, distinctive physical findings (e.g., myokymia in gold intoxication, brawny induration of the limbs in eosinophilia-myalgia syndrome, palmoplantar cutaneous exfoliation in arsenic poisoning, or striate leukonychia (Mees lines) of the nails in metal intoxication), and suggestive laboratory abnormalities (e.g., peripheral eosinophilia in eosinophilia-myalgia syndrome associated with tainted L-tryptophan, peripheral basophilic stippling of erythrocytes in arsenical intoxication, or abnormal urine heavy metals in arsenic and thallium poisoning). Deficiency syndromes such as vitamin B12 deficiency can also present with combinations of characteristic symptoms (unsteady gait, especially in the dark), physical findings (limb and trunk ataxia, pseudoathetosis, and extensor plantar responses), and laboratory abnormalities (decreased cyanocobalamin levels with megaloblastic anemia). Infections such as tabes dorsalis, leprosy, and HIV infection might also be distinguished on the basis of symptoms and signs (e.g., tabetic “flapping” gait and Argyll Robertson pupils; leonine facies with hypesthetic ears in Hansen disease; or anorexia, lymphadenopathy, and cutaneous findings in acquired immunodeficiency syndrome) and laboratory testing (positive syphilis serology, demonstration of mycobacteria on tissue biopsy, or positive HIV serology, respectively). Other useful considerations include the size category of neurons (axons) affected and the clinical pattern of

Nonmalignant Inflammatory Sensory Polyganglionopathy

neurologic deficits. The specific sensory modalities involved may also prove helpful in establishing the diagnosis. Sensory neuropathies with prominent dysfunction or loss of large-diameter fiber modalities and with findings such as kinesthesia and ataxia include hypothyroid polyneuropathy, uremic neuropathy, tabes dorsalis, vitamin B12 deficiency, IgM paraproteinemic neuropathy, spinocerebellar degenerations such as Friedreich ataxia, toxic exposure to platinum or nitrous oxide, paraneoplastic sensory polyganglionopathies,26 toxic polyganglionopathies such as those associated with pyridoxine78 and doxorubicin,14 infiltrative disease such as hematogenous metastasis to dorsal root ganglia sparing roots and nerves,45 sicca48 and Sjögren syndrome,38 and idiopathic nonmalignant inflammatory sensory polyganglionopathies.19,96 Conditions with sensory neuropathies in which small-diameter fiber modalities may be prominently affected include amyloidosis, diabetes mellitus, Fabry disease, HSANs, immune dysautonomias, leprosy, Tangier disease, and toxin exposures including gold, misonidazole, and tainted L-tryptophan. Other sensory neuropathies may involve both large- and small-diameter fiber modalities. Even with detailed investigation, it may be difficult to arrive at a definite diagnosis. The differential diagnosis of sensory polyganglioradiculoneuropathy rests on the exclusion of an underlying malignancy. There is a female preponderance in NISP, while malignant sensory polyganglionopathies affect both sexes similarly. In NISP, clinical manifestations of Sjögren syndrome (e.g., dry eyes and mouth) may be present, while signs of severe systemic disease such as cachexia and weight loss (sometimes associated with cancers) are unusual. MRI can help to exclude structural intraspinal diseases as well as to demonstrate signal changes in the ganglia, roots, and posterior columns (see above). CSF for malignant cells can be useful. Although serologic testing such as for antinuclear and extractable nuclear antigens may be positive in NISP, antineuronal nuclear antibodies have been described only rarely in patients diagnosed with these conditions,65 although it can be difficult to exclude occult malignancy even with careful postmortem studies. In paraneoplastic polyganglioradiculoneuropathies, extractable nuclear antigen serologies are normal, and antineuronal nuclear9,49 or other autoantibodies are sometimes, though not invariably, demonstrated. As in the polyneuropathies, disorders that most commonly might be associated with both sensory and motor dysfunction on occasion may primarily affect the posterior roots, resulting in pure or relatively pure sensory polyradiculopathy. These include a pure sensory form of the Guillain-Barré syndrome,94 pure sensory CIDPs,29 lympho- or plasma-proliferative disease,21 meningeal lymphomatosis,24 and meningeal carcinomatosis.77 The pattern of deficits on neurologic examination may help distinguish polyganglioradiculoneuropathies from sensory polyneuropathies, but does not assist in dissecting

2317

out the specific causes of proximal sensory disease. Patients with sensory polyneuropathies have distal symmetrical “stocking-glove distribution” hypesthesia, absence or reduction of distal deep tendon reflexes, and normal muscle strength. Those with sensory polyradiculopathies and polyganglionopathies can demonstrate distal and proximal, symmetrical or asymmetrical sensory loss; widespread deep tendon hypo- to areflexia; and normal strength. Central sensory axonopathy (or central sensory syndrome) is important to distinguish from sensory polyganglionopathy of all causes, even though in some instances the disease processes resulting in each are the same. A diagnosis of central sensory axonopathy rests on the triad of (1) clinical symptoms and signs of sensory loss, (2) one or more normal sensory nerve action potentials in the territory of clinical sensory loss, and (3) objective (psychophysiologic) electrophysiologic evidence of sensory abnormality.80 The latter might include central abnormalities on somatosensory evoked potential testing (e.g., central interpeak latency prolongation) and abnormally elevated quantitative sensory detection thresholds by accepted methods.80 Conditions that have been associated with central sensory syndrome include compressive posterior column myelopathy (from C3,4 spondylosis or thoracic meningioma), vascular malformations in the spinal cord, multiple sclerosis, Sjögren syndrome, HIV infection, vitamin B12 deficiency, nitrous oxide intoxication, and paraneoplastic autoimmunity. In some cases no definite cause or association is found.80

TREATMENT Although immunotherapy has been given, trials of treatment have not been sufficiently rigorous to establish whether any are effective. Many patients have received corticosteroids, intravenous immunoglobulin, 63,70 or other immunomodulatory therapy,10 and some have undergone plasma exchange.13,19,38,96 To date no convincing data demonstrate that any treatment is more efficacious than no treatment, although in individual anecdotal cases stabilization in clinical course has been associated with corticosteroid treatment, plasma exchange, or intravenous immunoglobulin. Some patients appear to improve spontaneously.

REFERENCES 1. Adams, F.: The Genuine Works of Hippocrates. London, Sydenham Society, p. 76, 1849. 2. Arnason, B. G. W.: Acute inflammatory demyelinating polyradiculoneuropathies. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 2070, 1984.

2318

Diseases of the Peripheral Nervous System

3. Auger, R., Windebank, A. J., Lucchinetti, C. F., and Chalk, C. H.: Role of the blink reflex in the evaluation of sensory neuronopathy. Neurology 53:407, 1999. 4. Bang, M. S., Han, T. R., and Lim, J. Y.: Acute sensory neuronopathy identified with electrodiagnosis and magnetic resonance imaging. Am. J. Phys. Med. Rehabil. 77:494, 1998. 5. Bickerstaff, E. R.: Report of clinico-pathological conference held at the Midland Centre for Neurosurgery on 15th June, 1957 on the occasion of the visit to the centre of Dr. K.J. Zulch (Cologne). Birmingham Med. Rev. 20:358, 1958. 6. Biehl, J. P., and Nimitz, H. J.: Studies on the use of a high dose of isoniazid. Am. Rev. Tuberc. Pulm. Dis. 70:430, 1954. 7. Bradley, W. G., Karlsson, I. J., and Rassol, C. G.: Metronidazole neuropathy. Br. Med. J. 2:610, 1977. 8. Brown, M. J., Martin, J. R., and Asbury, A. K.: Painful diabetic neuropathy: a morphometric study. Arch. Neurol. 33:164, 1976. 9. Camdessanche, J.-P., Antoine, J.-C., Honnorat, J., et al.: Paraneoplastic peripheral neuropathy associated with anti-Hu antibodies: a clinical and electrophysiological study of 20 patients. Brain 125:166, 2002. 10. Caroyer, J.-M., Manto, M. U., and Steinfeld, S. D.: Severe sensory neuronopathy responsive to infliximab in primary Sjögren’s syndrome. Neurology 59:1113, 2002. 11. Casey, E. B., Jelliffe, A. M., Le Quesne, P. M., and Millett, Y. L.: Vincristine neuropathy: clinical and electrophysiological observations. Brain 96:69, 1973. 12. Cavanagh, J. B., Fuller, N. H., Johnson, H. R. M., and Rudge, P.: The effects of thallium salts, with particular reference to the nervous system changes. Q. J. Med. 43:293, 1974. 13. Chen, W.-H., Yeh, J.-H., and Chiu, H.-C.: Plasmapheresis in the treatment of ataxic sensory neuropathy associated with Sjögren’s syndrome. Eur. Neurol. 45:270, 2001. 14. Cho, E. S., Spencer, P. S., Jorther, B. S., and Schaumburg, H. H.: A single intravenous injection of doxorubicin (Adriamycin) induces sensory neuronopathy in rats. Neurotoxicology 1:583, 1980. 15. Conrad, B., and Benecke, R.: How valid is the distal sensory nerve action potential for differentiating between radicular and non-radicular nerve lesions? [abstract] Acta Neurol. Scand. 60(Suppl. 73):122, 1979. 16. Cooke, W. T., and Smith, W. T.: Neurological disorders associated with adult coeliac disease. Brain 88:683, 1966. 17. Cornblath, D. R., McArthur, J. C., Kennedy, P. G. E., et al.: Inflammatory demyelinating peripheral neuropathies associated with human T-cell lymphotropic virus type III infection. Ann. Neurol. 21:32, 1987. 18. Croft, P. B., Henson, R. A., Urich, H., and Wilkinson, P. C.: Sensory neuropathy with bronchial carcinoma: a study of four cases showing serological abnormalities. Brain 88:501, 1965. 19. Dalakas, M. C.: Chronic idiopathic ataxic neuropathy. Ann. Neurol. 19:545, 1986. 20. Dastur, D. K.: Cutaneous nerve in leprosy: the relationship between histopathology and cutaneous sensibility. Brain 78:615, 1955. 21. Davison, C., and Balser, B. H.: Myeloma and its neural complications. Arch. Surg. 35:913, 1937. 22. Denny-Brown, D.: Primary sensory neuropathy with muscular changes associated with carcinoma. J. Neurol. Neurosurg. Psychiatry 11:73, 1948.

23. Denny-Brown, D.: Hereditary sensory radicular neuropathy. J. Neurol. Neurosurg. Psychiatry 14:237, 1951. 24. Dickerman, R. C., and Chason, J. L.: Alterations in the dorsal root ganglia and adjacent nerves in the leukemias, the lymphomas, and multiple myeloma. Am. J. Pathol. 34:349, 1958. 25. Duchenne, G. B. A.: De l’ataxie locomotrice progressive. Arch. Gen. Med. (Paris) 1:641, 1858. 26. Dyck, P. J., Bailey, A. A., and Olszewski, J.: Carcinomatous neuromyopathy. Can. Med. Assoc. J. 79:913, 1958. 27. Dyck, P. J., Benstead, T. J., Conn, D. L., et al.: Nonsystemic vasculitic neuropathy. Brain 110:843, 1987. 28. Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968. 29. Dyck, P. J., Lais, A. C., Ohta, M., et al.: Chronic inflammatory polyradiculoneuropathy. Mayo Clin. Proc. 50:621, 1975. 30. Dyck, P. J., Lambert, E. H., and Nichols, P. C.: Quantitative measurement of sensation related to compound action potential and number and size of myelinated and unmyelinated fibers in health, Friedreich’s ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Cobb, W. A. (ed.): Handbook of Electroencephalography and Clinical Neurophysiology, Vol. 9. Amsterdam, Elsevier, p. 83, 1972. 31. Fullerton, P. M., and Kremer, M.: Neuropathy after intake of thalidomide (Distaval). Br. Med. J. 2:855, 1961. 32. Gherardi, R. K., Farcet, J.-P., Creange, A., et al.: Dominant T-cell clones of unknown significance in patients with idiopathic sensory neuropathies. Neurology 51:384, 1998. 33. Goldblatt, S.: Pallhypesthesia: depression of the appreciation of vibration in trauma and in disease; a preliminary report. Arch. Neurol. Psychiatry 59:292, 1948. 34. Goldstein, N. P., Jones, P. H., and Brown, J. R.: Peripheral neuropathy after exposure to an ester of dichlorophenoxyacetic acid. J. Am. Med. Assoc. 171:1306, 1959. 35. Gowers, W. R.: Locomotor ataxy. In A Manual of Diseases of the Nervous System. Philadelphia, P. Blakiston, p. 311, 1888. 36. Gowers, W. R.: Multiple neuritis. In A Manual of Diseases of the Nervous System. Philadelphia, P. Blakiston, p. 109, 1888. 37. Gray, K. W., Woolf, A. L., and Wright, E. A.: Two cases of primary sensory neuropathy associated with carcinoma. Guys Hosp. Rep. 104:157, 1955. 38. Griffin, J. W., Cornblath, D. R., Alexander, E., et al.: Ataxic sensory neuropathy and dorsal root ganglionitis associated with Sjögren’s syndrome. Ann. Neurol. 27:304, 1990. 39. Gross, J. A., Haas, M. L., and Swift, T. R.: Ethylene oxide neurotoxicity: report of 4 cases and review of the literature. Neurology (Minneap.) 29:978, 1979. 40. Halperin, J., Luft, B. J., Volkman, D. J., and Dattwyler, R. J.: Lyme neuroborreliosis: peripheral nervous system manifestations. Brain 113:1207, 1990. 41. Hankey, G. J., and Gubbay, S. S.: Peripheral neuropathy associated with sicca syndrome [letter]. J. Neurol. Neurosurg. Psychiatry 50:1085, 1987. 42. Henson, R. A., Russell, D. S., and Wilkinson, M.: Carcinomatous neuropathy and myopathy: a clinical and pathological study. Brain 77:82, 1954. 43. Hicks, E. P.: Hereditary perforating ulcer of the foot. Lancet 1:319, 1922.

Nonmalignant Inflammatory Sensory Polyganglionopathy 44. Hull, R. G., Morgan, S. H., Harding, A. E., and Hughes, G. R. V.: Sjögren’s syndrome presenting as a severe sensory neuropathy including involvement of the trigeminal nerve. Br. J. Rheumatol. 23:301, 1984. 45. Johnson, P. C.: Hematogenous metastases of carcinoma to dorsal root ganglia. Acta Neuropathol. (Berl.) 38:171, 1977. 46. Katrak, S. M., Pollock, M., O’Brien, C. P., et al.: Clinical and morphological features of gold neuropathy. Brain 103:671, 1980. 47. Kaufman, M. D., Hopkins, L. C., and Hurwitz, B. J.: Progressive sensory neuropathy in patients without carcinoma: a disorder with distinctive clinical and electrophysiological findings. Ann. Neurol. 9:237, 1981. 48. Kennett, R. P., and Harding, A. E.: Peripheral neuropathy associated with the sicca syndrome. J. Neurol. Neurosurg. Psychiatry 49:90, 1986. 49. Kimmel, D. W., O’Neill, B. P., and Lennon, V. A.: Subacute sensory neuronopathy associated with small cell lung carcinoma: diagnosis aided by autoimmune serology. Mayo Clin. Proc. 63:29, 1988. 50. Knazan, M., Bohlega, S., Berry, K., and Eisen, A.: Acute sensory neuronopathy with preserved SEPs and long-latency reflexes. Muscle Nerve 13:381, 1990. 51. Kocen, R. S., and Thomas, P. K.: Peripheral nerve involvement in Fabry’s disease. Arch. Neurol. 22:81, 1970. 52. Kumazawa, K., Sobue, G., Yamamoto, K., et al.: Segmental anhidrosis in the spinal dermatomes in Sjögren’s syndromeassociated neuropathy. Neurology 43:1820, 1993. 53. Laloux, P., Brucher, J. M., Guerit, J. M., et al.: Subacute sensory neuronopathy associated with Sjögren’s sicca syndrome. J. Neurol. 235:352, 1988. 54. Lauria, G., Pareyson, D., Grisoli, M., et al.: Clinical and magnetic resonance imaging findings in chronic sensory ganglionopathies. Ann. Neurol. 47:104, 2000. 55. LeQuesne, P. M.: Neuropathy due to drugs. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 2171, 1984. 56. LeQuesne, P. M., and McLeod, J. G.: Peripheral neuropathy following a single exposure to arsenic. J. Neurol. Sci. 32:437, 1977. 57. Lovelace, R. E., and Horwitz, S. J.: Peripheral neuropathy in long-term diphenylhydantoin therapy. Arch. Neurol. 18:69, 1968. 58. Low, P. A., Burke, W. J., and McLeod, J. G.: Congenital sensory neuropathy with selective loss of small myelinated fibers. Ann. Neurol. 3:179, 1978. 59. Malinow, K., Yannakakis, G. D., Glusman, S. M., et al.: Subacute sensory neuronopathy secondary to dorsal root ganglionitis in primary Sjögren’s syndrome. Ann. Neurol. 20:535, 1986. 60. Manley, G., Wong, E., Dalmau, J., et al.: Sera from some patients with antibody-associated paraneoplastic encephalomyelitis/sensory neuronopathy recognize the Ro-52k antigen. J. Neurooncol. 19:105, 1994. 61. Mellgren, S. I., Conn, D. L., Stevens, J. C., and Dyck, P. J.: Peripheral neuropathy in primary Sjögren’s syndrome. Neurology 39:390, 1989. 62. Mitsumoto, H., Wilbourn, A. J., and Massarweh, W.: Acquired “pure” sensory polyneuropathy (APSP): unique clinical features in 30 patients [abstract]. Neurology 35(Suppl. 1):295, 1985.

2319

63. Molina, J. A., Benito-Leon, J., Bermejo, F., et al.: Intravenous immunoglobulin therapy in sensory neuropathy associated with Sjögren’s syndrome [letter]. J. Neurol. Neurosurg. Psychiatry 60:699, 1996. 64. Molinuevo, J. L., Graus, F., Serrano, C., et al.: Utility of anti-Hu antibodies in the diagnosis of paraneoplastic sensory neuropathy. Ann. Neurol. 44:976, 1998. 65. Moll, J. W., Markusse, H. M., Pijnenburg, J. J., et al.: Antineuronal antibodies in patients with neurologic complications of primary Sjögren’s syndrome. Neurology 43:2574, 1993. 66. Mori, K., Koike, H., Misu, K., et al.: Spinal cord magnetic resonance imaging demonstrates sensory neuronal involvement and clinical severity in neuronopathy associated with Sjögren’s syndrome. J. Neurol. Neurosurg. Psychiatry 71:488, 2001. 67. Murai, Y., and Kuriowa, Y.: Peripheral neuropathy in chlorobiphenyl poisoning. Neurology (Minneap.) 21:1173, 1971. 68. Ogryzlo, M. A.: A familial peripheral neuropathy of unknown etiology resembling Morvan’s disease. Can. Med. Assoc. J. 54:547, 1946. 69. O’Leary, C. P., and Willison, H. J.: Autoimmune ataxic neuropathies (sensory ganglionopathies). Curr. Opin. Neurol. 10:366, 1997. 70. Pascual, J., Cid, C., and Berciano, J.: High dose i.v. immunoglobulin for peripheral neuropathy associated with Sjögren’s syndrome. Neurology 51:650, 1998. 71. Postma, T. J., Hoekman, K., van Riel, J. M., et al.: Peripheral neuropathy due to biweekly paclitaxel, epirubicin and cisplatin in patients with advanced ovarian cancer. J. Neurooncol. 45:241, 1999. 72. Prosser, P. R., and Karam, J. H.: Diabetes mellitus following rodenticide ingestion in man. JAMA 239:1148, 1978. 73. Quarles, R. H., and Weiss, M. D.: Autoantibodies associated with peripheral neuropathy. Muscle Nerve 22:800, 1999. 74. Ramon y Cajal, S.: In Kopsch, F. R.: Rauber-Kopsch: Lehrbuch und Atlas der Anatomie des Menschen, Band III: Nervensystem-Sinnesorgane. Leipzig, Georg Thieme-Verlag, p. 41, 1943. 75. Romberg, M. H.: A Manual of the Nervous Diseases of Man, Vol. II. London, Sydenham Society, p. 395, 1853. 76. Ross, A. T., and Tinker, J. H.: Anesthesia risk. In Miller, R. D. (ed.): Anesthesia. New York, Churchill Livingstone, p. 715, 1990. 77. Russell, D. S., and Rubinstein, L. J.: Pathology of Tumours of the Nervous System. Edinburgh, Edward Arnold, p.356, 1977. 78. Schaumberg, H., Kaplan, J., Windebank, A., et al.: Sensory neuropathy from pyridoxine abuse. N. Engl. J. Med. 309:445, 1983. 79. Schaumburg, H. H., and Spencer, P. S.: Human toxic neuropathy due to industrial agents. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 2102, 1984. 80. Smith, B. E., and Bosch, E. P.: Neurophysiological findings in patients with spinal sensory loss. Ann. Neurol. 50:S23, 2001. 81. Smith, B. E., and Dyck, P. J.: Peripheral neuropathy in the eosinophilia-myalgia syndrome associated with L-tryptophan. Neurology 40:1035, 1990. 82. Smith, I. S., Kahn, S. N., Lacey, B. W., et al.: Chronic demyelinating neuropathy associated with benign IgM paraproteinemia. Brain 106:169, 1983.

2320

Diseases of the Peripheral Nervous System

83. Sobue, G., Nakao, N., Murakami, K., et al.: Type I familial amyloid polyneuropathy. Brain 113:903, 1990. 84. Sobue, G., Yanagi, T., and Hashizume, Y.: Chronic progressive sensory ataxic neuropathy with polyclonal gammopathy and disseminated focal perivascular cellular collections. Neurology 38:463, 1988. 85. Sobue, G., Yasuda, T., Kachi, T., et al.: Chronic progressive sensory ataxic neuropathy: clinicopathological features of idiopathic and Sjögren’s associated cases. J. Neurol. 240:1, 1993. 86. Spencer, P. S., Couri, D., and Schaumburg, H. H.: n-Hexane and methyl n-butyl ketone. In Spencer, P. S., and Schaumberg, H. H. (eds): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 456, 1980. 87. Stanley, E.: A case of disease in the posterior columns of the spinal cord. Med. Chir. Trans. 5:80, 1840. 88. Sterman, A. B., Schaumburg, H., and Asbury, A. K.: The acute sensory neuronopathy syndrome: a distinct clinical entity. Ann. Neurol. 7:354, 1980. 89. Swanson, A. G.: Congential insensitivity to pain with anhidrosis. Arch. Neurol. 8:299, 1963. 90. Thomas, P. K., Hollinrake, K., Lascelles, R. G., et al.: The polyneuropathy of chronic renal failure. Brain 94:761, 1971.

91. Thomas, P. K., and Walker, J. G.: Xanthomatous neuropathy in primary biliary cirrhosis. Brain 88:1079, 1965. 92. Todd, R. B.: Cyclopaedia of Anatomy and Physiology, Vol. III. London, Longman, Brown, Green, Longmans, and Roberts, p. 721r, 1839–1847. 93. Victor, M.: Polyneuropathy due to nutritional deficiency and alcoholism. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. (eds): Peripheral Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 1913, 1984. 94. Wartenberg, R.: Sensory polyneuritis. In Neuritis, Sensory Neuritis, Neuralgia. New York, Oxford University Press, p. 160, 1958. 95. Wasserman, T. W., Phillips, T. L., VanRaalte, G., et al.: The neurotoxicity of misonidazole: potential modifying role of phenytoin sodium and dexamethasone. Br. J. Radiol. 53:172, 1980. 96. Windebank, A. J., Blexrud, M. D., Dyck, P. J., et al.: The syndrome of acute sensory neuropathy: clinical features and electrophysiologic and pathologic changes. Neurology 40:584, 1990. 97. Yasuda, T., Sobue, G., Hirose, Y., et al.: MR of acute autonomic and sensory neuropathy. AJNR Am. J. Neuroradiol. 15:114, 1994.

104 Cryptogenic Sensory Polyneuropathy IAN A. GRANT

Demographics Symptoms Clinical Findings Clinical Course Utility of Serial Evaluation

Diagnosis Electrodiagnostic Tests Quantitative Sensation Testing Autonomic Testing Sural Nerve Biopsy Skin Biopsy Serology

Despite intensive investigation, the cause of peripheral neuropathy cannot be determined in all patients. Among patients with polyneuropathy referred to specialized neuromuscular clinics for clinical assessment or nerve biopsy, 24% to 70% remain undiagnosed.9,33,38,48,51 The percentage of idiopathic cases is lower in more recent series, probably because of improved understanding of the causes of neuropathy and diagnostic advances such as genetic tests for some types of inherited neuropathy. However, idiopathic neuropathy remains a common clinical problem. Most patients with idiopathic polyneuropathy have a predominantly sensory disease. Many neurologists’ practices include a number of such individuals. An illustrative “typical” patient might be a woman presenting in her early 60s with long-standing pain of an aching or burning nature in both feet. Other distal sensory symptoms, such as numbness, are noted but are less bothersome. Physical examination reveals fairly mild, symmetrical distal sensory deficits in the lower limbs, without muscle weakness. Mild nerve conduction abnormalities consistent with an axonal, predominantly sensory polyneuropathy are found, although age-related issues of interpretation arise. Extensive laboratory tests show no evidence of diabetes, malignancy, or systemic disease. Over many years of observation, the patient’s pain persists, incompletely eased by a variety of agents, including tricyclic antidepressants. However, despite slight progression in clinical and electrodiagnostic findings, no significant neurologic disability develops. Laboratory testing remains unrevealing, and the lingering uncertainty about the cause of the problem is frustrating for both patient and physician.

Treatment Differential Diagnosis Diabetes Sjögren’s Syndrome Inherited Neuropathy Orthopedic Disease

This syndrome has been described under a variety of names, including burning feet syndrome,8,58,59 distal small fiber neuropathy (SFN),13,14,21 painful sensory neuropathy,45 sensory-predominant painful idiopathic neuropathy,25 idiopathic axonal polyneuropathy,19,40,42 and cryptogenic sensory polyneuropathy (CSP).31,68 The latter term is used in this chapter. Differences in terminology partly reflect differences in patient populations, such as the degree to which small and large sensory fibers are affected and whether motor involvement is present. To consider CSP as a distinctive disease is clearly artificial. The term encompasses a heterogeneous group of patients with respect to both the characteristics and causes of their neuropathies. Nevertheless, a number of published reports summarizing the features of this syndrome show that one can make certain generalizations that are clinically and prognostically useful.

DEMOGRAPHICS Idiopathic polyneuropathy is a disease of late middle age; median age of onset of symptoms is in the sixth decade in most series.14,21,25,36,40,45,58,68 This is generally true of both SFN and mixed sensory and sensorimotor types. However, symptoms may begin as early as the second decade or as late as the ninth. Symptoms have usually been present for several years before a diagnosis is made. In some cases, patients present with very longstanding symptoms going back two decades or more. 2321

2322

Diseases of the Peripheral Nervous System

Occasionally, the course is subacute over a period of months.21,58 There is disagreement about gender differences, perhaps because of other differences in patient populations in published series. In general, idiopathic peripheral neuropathy is more common in men. However, patients presenting with pain or other features of SFN are more likely to be female.14,21,25,45,58

SYMPTOMS A variety of sensory symptoms may occur, the specific symptoms depending on the affected population(s) of fibers. Sensory symptoms tend to occur early even in patients with sensorimotor neuropathies. Numbness and tingling paresthesias are particularly common, either alone or, more often, accompanied by pain. The patients with CSP described by Wolfe and colleagues most often complained of a combination of pain and other symptoms; only 10% complained of pain alone.68 Gait instability and other symptoms of severe large fiber dysfunction are uncommon. Among patients with purely sensory deficits and electrodiagnostic changes, a minority describes weakness or cramps, but these are seldom severe. Subjective weakness does not necessarily indicate disease of motor axons, and some patients will use the term “weakness” to describe numbness, incoordination, or other phenomena that actually arise from sensory dysfunction. In keeping with a length-dependent process, symptoms usually begin in the feet. Involvement of the hands is less common and does not usually occur until lower limb symptoms have spread beyond the feet. Upper limb onset is most unusual and should raise suspicion of a ganglionopathy, multiple mononeuropathies, perineuritis, or superimposed nerve entrapment. Symmetry is the rule, but mild side-to-side differences often occur, especially when symptoms first begin. Pain is the hallmark of SFN, and it is a required diagnostic criterion in some series. The quality of pain varies, and many patients experience more than one type. Common descriptions include burning, cold, aching, squeezing, sharp, or shooting sensations. Burning feet may be the only complaint. Pain is often worse with weight bearing or at night and may interfere with sleep. Allodynia is common, with non-noxious stimuli such as contact from bedsheets or socks perceived as painful. Coexisting nonpainful sensory symptoms may be present. Negative phenomena tend to be less prominent, but the patient may note a loss of ability to feel the temperature of bathwater on the leg. Unrecognized foot trauma is unusual. Autonomic symptoms are generally mild, even in patients with SFN.14,21,43,58 Some patients experience dryness of the eyes or mouth, increased or decreased sweating, or impo-

tence. Postural hypotension is rarely significant; Novak and colleagues obtained a history of orthostatic lightheadedness in 25% of patients with painful neuropathy, none of whom needed treatment.43

CLINICAL FINDINGS Examination reveals distal, symmetrical sensory deficits in a stocking distribution. Specific sensory modalities are affected to a variable degree depending on relative involvement of large myelinated, small myelinated, and unmyelinated populations of sensory axons. Vibration sense is reduced in most patients.36,40,68 Abnormal sensitivity to light touch and pinprick is also common, but joint position sense is affected in only a minority of patients, usually to a mild degree. Ankle reflexes are absent in most patients, but loss of upper limb reflexes is unusual. Selective small fiber neuropathy is not uncommon. Most patients with SFN show loss of pinprick and/or temperature sensation on the feet. Hyperalgesia to pinprick is occasionally the only abnormality found. Altered temperature sensation may be evident as an inability to appreciate a cold tuning fork. Mild loss of vibration sensitivity is consistently found in about half of cases.14,21,58 Tendon reflexes are usually preserved. The examination can be completely normal; in such cases, confirmation of neuropathy will require quantitative sensation testing (QST) or laboratory evidence of small fiber dysfunction. Skin ulceration and other destructive changes are rare. Assessment of motor findings is problematic. A rigid definition of sensory neuropathy would include only patients with exclusively sensory symptoms, signs, and electrodiagnostic abnormalities. Such pure sensory neuropathies are unusual; they often result from primary disease of sensory cell bodies in dorsal root ganglia and are properly termed neuronopathies or ganglionopathies, as discussed in Chapter 103. A looser but perhaps more practical definition would accept the presence of minor motor signs in patients whose findings are mainly sensory, or the appearance of mild weakness only after many years of sensory symptoms (technically implying sensorimotor neuropathy). In a group of patients with sensory neuropathy defined by symptoms, Wolfe and colleagues68 found distal weakness in 41%, mainly in small foot muscles; hand weakness was found in 18% and was almost always accompanied by lower limb weakness. Distal muscle wasting was noted in 15%.

CLINICAL COURSE CSP usually follows a benign course. Symptoms and neurologic deficits remain stable over a number of years after an initial period of progression,16,36,68 or continue

2323

Cryptogenic Sensory Polyneuropathy

to progress very slowly.68 A small number of patients experience spontaneous improvement.36 Even in chronic progressive cases, serious impairment (such as inability to walk unassisted) seldom results. In a 5-year follow-up study of patients with a sensory form of chronic idiopathic axonal neuropathy, Notermans and colleagues42 found no change in a quantitative grading score of sensory loss, minimal change in disability (as measured by the modified Rankin scale), and the need for a cane to walk in only 1 of 21 patients. The same investigators have also studied whether the duration of neuropathy relates to severity of deficits and disability, finding no difference in signs between those with symptoms of less than 10 years’ duration and those with more long-standing symptoms.67 Age of onset was more relevant; disability was greater in the early-onset group (before age 65), perhaps as a result of greater impact on occupation. However, no patients became severely disabled. Similar outcomes are seen in patients with SFN.14 However, chronic pain in these patients is often resistant to treatment, and quality of life may be poor. This issue has received little study, but it is likely that some patients’ quality of life is adversely affected, as is the case with other kinds of painful neuropathy such as that caused by diabetes.3,11,60

UTILITY OF SERIAL EVALUATION As with any idiopathic disorder, one might assume that long-term follow-up of patients with CSP with periodic reassessment and repetition of some or all screening tests might eventually establish the cause. The yield of this approach is relevant given the time and expense involved in reassessment. Grahmann et al.16 and McLeod et al.36 identified a cause in 29% and 36% of cases during mean follow-up periods of 2 and 3 years, respectively. In both series, reassessment permitted identification of cases of alcohol abuse, monoclonal gammopathy, malignancy and vitamin B12 deficiency. Both groups of investigators concluded that reassessment is a worthwhile endeavor. In contrast, Notermans and colleagues found a cause in only 4 of 75 patients followed for 5 years (Charcot-MarieTooth disease in 2, chronic inflammatory demyelinating polyneuropathy in 1, and alcohol abuse in 1).42 Based on this low yield, Notermans and colleagues concluded that routine reevaluation is not profitable and that follow-up testing should be done selectively, based on clinical developments. Perhaps the more important issue is how often a treatable cause of neuropathy is found; these limited data suggest that this occurs infrequently.

DIAGNOSIS Electrodiagnostic Tests The electromyography (EMG) findings in CSP vary with age and clinical subtype, but in general reflect an axonal length-dependent polyneuropathy affecting sensory fibers more than motor fibers.67,68 Evidence of demyelination is rare.36,40 Small or absent sensory nerve action potentials (SNAPs) are the most common finding, particularly the sural response. Mildly prolonged distal sensory latencies or reduced sensory conduction velocities may also be present. Motor nerve conduction abnormalities are less common, but not greatly so; Wolfe and co-workers68 found sensory and motor conduction abnormalities in 77% and 60% of patients, respectively. Motor conduction changes are often found in patients with purely sensory symptoms and signs. Subclinical involvement of motor axons is also evident in the fairly common finding of fibrillations or chronic motor unit abnormalities in distal leg muscles, particularly in the foot. Upper limb abnormalities are less common. Difficulties may arise in distinguishing nerve conduction findings caused by disease from those caused by aging. Most laboratories do not consider an absent sural SNAP to be definite evidence of neuropathy in patients greater than 60 years of age; the findings in elderly patients may therefore be indeterminate when only lower limb sensory parameters are affected. Fortunately, the majority of elderly patients with cryptogenic peripheral neuropathy and abnormal sural conduction also have other electrodiagnostic evidence of neuropathy.67,68 When viewed collectively, there are reasonably clear electrophysiologic differences between patients with idiopathic polyneuropathy and controls, even in older age groups. Comparing patients and controls older than 65 years of age, Vrancken and colleagues67 found significant differences for all motor and sensory parameters assessed in the median, tibial, and sural nerves, and for the presence of spontaneous activity on needle examination of the tibialis anterior muscle. Of these changes, reduced amplitudes and needle examination abnormalities are most likely to be apparent; the magnitude of conduction slowing is fairly small, and conduction velocities often remain within the normal range. Nerve conduction studies are often completely normal in patients with SFN. Indeed, some would consider normal nerve conduction studies to be a necessary criterion for this diagnosis.21,58 When a broader definition based mainly on clinical features such as pain is used, one will encounter a subset of patients with mild sensory greater than motor conduction abnormalities similar to those described above.14 Again, comparison of groups of SFN patients to controls reveals differences that imply that mild large fiber dysfunction is often present.23 Mild distal motor conduction changes are sometimes also present.14 Needle

2324

Diseases of the Peripheral Nervous System

EMG seldom shows abnormalities when nerve conduction studies are normal.23

Quantitative Sensation Testing QST has not received widespread attention in the evaluation of CSP. Even today, QST is not available at all neuromuscular centers, and the literature on idiopathic peripheral neuropathy mainly consists of older studies that predate the widespread use of QST. QST has mainly been used in the assessment of patients with SFN. Jamal and colleagues reported QST findings in 25 patients with SFN, including 10 idiopathic cases.23 Nerve conduction parameters were normal in all patients. Nonpainful heat and cold thresholds were measured on the ankle and wrist. The overall sensitivity of thermal testing was 100%, with abnormalities more often found in the lower limb than the upper limb. Heat threshold was abnormal more often than cold threshold (88% vs. 72%). Vibration thresholds were normal. In contrast, other investigators have demonstrated vibration threshold abnormalities in patients whose clinical findings suggest purely SFN,22,43 further evidence that the clinical features often underestimate the spectrum of affected fiber populations. When abnormal, thermal thresholds in patients with SFN are usually increased, that is, the patient is less sensitive than normal (hypalgesia or hypesthesia). However, the opposite can also occur; in patients with thermal hyperalgesia, heat-pain thresholds are low and the slope of the stimulus-response curve is increased. This finding is useful in identifying SFN and is the only abnormal test in some patients. Abnormal cooling thresholds have been compared to other laboratory tests of small fiber disease. The severity of thermal hypesthesia does not correlate well with sweating abnormalities or intraepidermal nerve fiber density,22,43 and QST appears to be a less sensitive detector of small fiber dysfunction than these other techniques. Comparing thermal thresholds to clinical findings and autonomic tests (quantitative sudomotor axon reflex testing [QSART] and cardiovascular reflexes), Tobin and colleagues62 found that the sensitivity of thermal threshold was 67%, lower than that of both autonomic tests and the clinical assessment. Cooling thresholds are less sensitive than intraepidermal nerve fiber density,45 but the difference is probably small. Sensitivity is not the only factor to consider; because there are clearly some patients with abnormalities of one or more tests of small fiber function but not others, these tests are probably complimentary.

Autonomic Testing Autonomic dysfunction has been studied mainly in patients with pain or other features of SFN. Prominent autonomic involvement occurs in neuropathy caused by

amyloidosis and diabetes, but cryptogenic neuropathy has received less study. Given the difficulties in testing small myelinated and unmyelinated somatic fibers, there is appeal in using autonomic testing because it examines similar-sized populations that are often affected by the same diseases that damage somatic small fibers. Dedicated autonomic labs remain few, and comprehensive autonomic assessment is not widely available. Sympathetic sudomotor function can be assessed with QSART or the thermoregulatory sweat test. Both are sensitive and provide information about the anatomic distribution of sweating abnormalities, but require specialized equipment. Assessment of cardiac vagal function is usually done by measuring changes in heart rate (R-R interval) and blood pressure during deep breathing or the Valsalva maneuver. Simple heart rate testing can be performed using a standard EMG machine.57 Adrenergic function is examined by monitoring changes in blood pressure during postural change using a tilt table. These and other testing methods have been reviewed.32,49 Autonomic changes in CSP are common, with abnormal test results in at least 90% of patients. Abnormalities are predominantly cholinergic. Common symptoms include dryness of the eyes, mouth, or skin; color changes in the skin; and impotence.43 Altered sweating is the most common laboratory finding, present in more than three quarters of patients, mainly in a distal pattern resembling that of the somatic neuropathy. Cardiac vagal dysfunction is less common, occurring in 28% to 64% of patients, and is often mild.43,58 Orthostatic hypotension and syncope do not occur, in contrast to other autonomic neuropathies. The reason for this is unknown. McLeod suggested that damage to the relatively short sympathetic fibers innervating the splanchnic vasculature is an important factor in the development of orthostatic hypotension35; because the pattern of sweating abnormalities in CSP suggests a length-dependent process, it may be that splanchnic nerves are spared as a result of their length. Somewhat surprisingly, hypertension was more common than expected in one series. 43 Dysautonomia is not limited to patients with selective SFN; Novak and colleagues found that autonomic dysfunction was more severe among patients with abnormal nerve conduction studies (implying more extensive neuropathy affecting large-diameter fibers) than among those with normal conduction.43 However, it is also clear that autonomic tests are abnormal in some patients with no other objective evidence of neuropathy and are therefore of clinical utility. Autonomic testing may be more sensitive than QST in detecting small fiber disease. QSART is abnormal more often than thermal threshold testing in patients with SFN.43,62

2325

Cryptogenic Sensory Polyneuropathy

Sural Nerve Biopsy Pathologic confirmation of sensory neuropathy is possible through sural nerve biopsy or through skin biopsy with measurement of epidermal nerve fiber density. Sural nerve biopsy is the older and more widely available method. Nerve biopsy is not necessary for diagnosis of sensory neuropathy in most cases. Its role is mainly to exclude specific underlying causes such as amyloidosis and inflammation.29 Nerve pathology in CSP has been described by several groups. The main finding is usually a loss of myelinated fibers, sometimes accompanied by active axonal degeneration.14,25,36,38,40,69 Clusters of small, thinly myelinated axons indicative of regeneration are sometimes present.36 Segmental demyelination is distinctly uncommon, but a few cases have been reported.69 In patients with SFN, nerve biopsy is often normal.14,21,58 In some cases there is selective loss of small myelinated and/or unmyelinated fibers.14,21 Detection of small fiber abnormalities requires examination of semithin plastic sections, either semiquantitatively or with morphometry using computerized image analysis software. Morphometric data are useful in demonstrating changes in the size distribution of myelinated fibers, but this technique has only occasionally been applied to CSP.21,38 Electron microscopy is necessary for assessment of unmyelinated fibers. Inflammation is absent in the vast majority of cases. However, Kelkar and colleagues25 found epineurial perivas-

FIGURE 104–1 Sural nerve pathology in a patient with cryptogenic sensory polyneuropathy. A, A small epineurial blood vessel with a perivascular collection of mononuclear inflammatory cells without evidence of fibrinoid necrosis (hematoxylin and eosin stain). B and C, Cross sections from the same nerve showing variability in the density of myelinated fibers between two different fascicles indicating inhomogeneous myelinated fiber loss, which is characteristic of a microangiopathic process (methylene blue, azure II, and basic fuchsin stain). Calibration bar: 50 ␮m. (From Kelkar, P., McDermott, W. R., and Parry, G. J.: Sensory-predominant, painful, idiopathic neuropathy: inflammatory changes in sural nerves. Muscle Nerve 26:413, 2002, with permission.)

cular inflammatory infiltrates in 8 of 11 patients with painful sensory neuropathy (Fig. 104–1). These patients had severe pain that was refractory to the usual drug treatments for neuropathic pain, such as tricyclic antidepressants and gabapentin. Multifocal fiber loss was found in three cases. No fibrinoid necrosis or transmural inflammation was found and thus necrotizing vasculitis was not apparent, but the findings did suggest an immunologically mediated process. Lacomis and co-workers30 reported two patients with SFN and evidence of vasculitis on nerve biopsy. Neither patient had a documented connective tissue disease, but one had an increased sedimentation rate and a raised titer of antinuclear antibodies. Nerve biopsy in both patients showed transmural lymphocytic inflammatory infiltrates without fibrinoid necrosis. Perhaps surprisingly given the putative vasculitic basis for their symptoms, axon loss was not apparent. One patient improved slightly with aggressive immunosuppression. Thus, although inflammation is rare in SFN, its presence in occasional cases with severe pain has obvious implications for treatment (see below), and biopsy should be considered in such cases.

Skin Biopsy Over the last decade there has been limited but increasing interest in skin biopsy as a means of quantifying cutaneous sensory nerves. Using a small skin sample obtained by punch biopsy, unmyelinated intraepidermal

2326

Diseases of the Peripheral Nervous System

nerve fibers (IENFs) can be labeled with an antibody against the pan-axonal marker protein gene product (PGP) 9.5 and their density measured under confocal microscopy. Normative data have been obtained for subjects across a wide range of ages and at multiple sites on the limbs.34 IENF density is well correlated with the density of small myelinated fibers in the sural nerve; correlation with unmyelinated fiber density is less clear.20 This technique has several advantages over full nerve biopsy: it is minimally invasive, IENF density is relatively independent of age and height, serial studies can be performed in the same patient, and more than one site can be sampled. It may be more sensitive than morphometric analysis of sural nerve biopsy in detecting fiber loss.27 IENF density is reduced in some patients with peripheral neuropathies associated with diabetes, human immunodeficiency virus (HIV) infection, and exposure to toxins.20–22,27,45 Abnormalities have also been described in patients with cryptogenic SFN.21,45 Mirroring their symptoms, these patients usually have a length-dependent pattern of reduction of IENF density with the greatest abnormality on the distal leg and milder changes more proximally.21 Holland and co-workers also described an uncommon pattern of fiber loss at both proximal and distal sites in patients with acute generalized burning pain, perhaps reflecting a polyganglionopathy.21 The sensitivity of this method is somewhat unclear. Absolute IENF density was reduced on the calf in 81% of patients with idiopathic SFN in one series.21 It is the only abnormal test in some patients with symptoms of SFN. Limited data suggest that IENF density is more sensitive than either QST or QSART in patients with painful neuropathy.21,45 Kennedy and colleagues have described an alternative method whereby IENFs are assessed in skin blisters.26 The blister is created with a suction capsule, and the roof is excised and fixed in whole mount. As in the punch biopsy technique, IENFs are visualized by immunostaining with an antibody against PGP 9.5. However, the two methods differ in that the blister roof is viewed from above rather than in cross section and confocal microscopy is not required. The blister technique has many of the same strengths as the punch biopsy technique, such as the ability to sample multiple sites. It has potential advantages over punch biopsy: it samples a larger area of tissue and is faster, less labor intensive, and slightly less invasive. Weaknesses include the potential for distortion of the tissue in creating the blister, and the limited normative data available. This method has been used to show changes in the number and distribution of cutaneous nerves in a small number of patients with diabetic and HIV-associated neuropathy, but its use has not yet been described in CSP.

Serology The presence of serologic abnormalities suggests that CSP may have an autoimmune cause in some patients. Early series reported a high prevalence of serum antibodies against nerve antigens, particularly sulfated glycoconjugates (sulfatides), in patients with axonal sensory polyneuropathies.39,46 Subsequent studies have not confirmed this.40,42,45,68 Wolfe and colleagues detected antisulfatide antibodies in none of their 41 patients with idiopathic sensory polyneuropathy.68 In a prospective study, serology for antibodies against sulfatide, myelin-associated glycoprotein, and GM1 ganglioside was negative in all 70 patients with idiopathic polyneuropathy, including those with pure sensory neuropathy.40 Testing for the paraneoplastic antibody anti-Hu is also almost invariably negative.14,45 Monoclonal gammopathy is occasionally found in patients with CSP.45,68 The prevalence of peripheral neuropathy is clearly increased in patients with monoclonal gammopathy, as has been extensively reviewed.50 The neuropathy is typically sensorimotor. Sensory deficits reflect mainly large fiber dysfunction, with ataxia and loss of joint position and vibration sense. Paraproteinemic neuropathy often has demyelinating features, particularly when the M protein is IgM, but axonal neuropathies also occur and may be predominantly sensory.15,41 Antinuclear antibodies or other non–organ-specific serologic abnormalities are sometimes present,14,45 but when in low titers and in the absence of clinical evidence of connective tissue disease (as is usually the case), their significance is uncertain.

TREATMENT Treatment of CSP mainly consists of drug therapy for neuropathic pain. As with other painful neuropathies, options include tricyclic antidepressants, gabapentin, carbamazepine, mexiletine, nonsteroidal anti-inflammatory drugs, opioids, and capsaicin. The rationale for the use of these drugs is based mainly on clinical experience and on inference from studies of other painful conditions, particularly diabetic neuropathy. Treatment response is often unsatisfactory. The principles of treatment of neuropathic pain are discussed in detail in Chapter 118. Immune-modulating therapies have been tried based on the possibility that an autoimmune or inflammatory process may be present. As previously noted, sural nerve biopsy in some patients with painful SFN has shown lymphocytic inflammatory infiltrates surrounding small epineurial arterioles.25,30 Nonsystemic vasculitic neuropathy presents with distal symmetrical symptoms and signs in some patients.5 Autoimmunity is also central to Sjögren’s syndrome, in which inflammation may be present in peripheral nerve or dorsal root ganglia.

2327

Cryptogenic Sensory Polyneuropathy

Treatment of CSP with immune-modulating drugs has not been adequately studied. No clinical trials have been performed, and currently it is not possible to draw firm conclusions about the efficacy of any drug. Empirical treatment has shown varied results. Gorson and Ropper14 found prednisone ineffective in patients with SFN, but four patients responded to intravenous immunoglobulin, with “near complete resolution” of burning pain in three. Improvement or stabilization has been reported in a small number of patients receiving prednisone or immunosuppressive agents for idiopathic30 and Sjögren-associated12,37 sensory neuropathy.

DIFFERENTIAL DIAGNOSIS Because a distal, predominantly sensory pattern is the most common kind of polyneuropathy encountered in neurologic clinics, the differential diagnosis is extensive (Table 104–1). It should be kept in mind that what remains as “cryptogenic” depends on the intensity of the evaluation. A complete discussion of the differential diagnosis of polyneuropathy, including the use of screening tests and indications for more specialized tests, is presented in Chapter 46. It is assumed that such an approach will have been followed and common disorders will have been excluded before labeling a patient with the diagnosis of CSP. This section covers several entities of particular importance because of their frequency and/or potential

Table 104–1. Differential Diagnosis of Chronic Sensory Polyneuropathies Alcohol abuse Connective tissue disease Diabetes Drugs Cis-platinum Isoniazid Metronidazole Nucleoside analogues (ddI, ddC) Paclitaxel Phenytoin* Pyridoxine Hypothyroidism Infections Human immunodeficiency virus Lyme disease Inherited disorders (see Table 104–3) Paraneoplastic syndromes Sarcoidosis Uremia Vasculitis (including nonsystemic vasculitic neuropathy) Vitamin deficiency *Rarely symptomatic.

difficulty in detection. Each of the disorders listed in Table 104–1 is discussed in detail elsewhere in this text.

Diabetes Fully developed diabetic polyneuropathy (DPN) clearly affects both sensory and motor fibers, but early DPN usually causes only sensory symptoms, including pain.61 Electrodiagnostic testing often reveals asymptomatic motor nerve conduction abnormalities in early DPN, but in some patients motor involvement will not yet be detectable. Selective SFN may occur,52 and small fiber dysfunction is an early feature of DPN generally.7 A distinctive, purely sensory form of acute, painful DPN occurs mainly in men in association with prominent weight loss.1 It is generally accepted that DPN occurs only after a period of hyperglycemia of sufficient duration and severity. Thus, type 1 diabetes mellitus (DM) (in which disease onset is clinically apparent) is almost always diagnosed years before the appearance of neuropathy. In contrast, the onset of type 2 DM is more difficult to determine, and patients may be hyperglycemic for years prior to diagnosis. Occasionally, neuropathy will develop during this period of unrecognized hyperglycemia. Measurement of fasting blood glucose (FBG) is a standard screening test for DM and should be done in all adults with newly discovered polyneuropathy. If DM is detected, the cause of neuropathy has probably been found, although a chance association could be present. A causal relationship is more likely if other end-organ complications are present, namely retinopathy or nephropathy.6,10 An elevated level of glycated hemoglobin A1c (HbA1c) also favors a diabetic cause; data from the Diabetic Control and Complications Trial show that the relative risk of DPN is proportional to HbA1c concentration.55 The dogma regarding the need for long-standing hyperglycemia as a prerequisite of DPN has recently been challenged. Two studies showed a higher than expected frequency of impaired glucose metabolism (identified mainly via an oral glucose tolerance test [OGTT]) in patients with idiopathic sensory neuropathy.44,54 The neuropathy in these patients was typically painful, with small fiber features. In contrast, glucose intolerance was no more common in patients with sensorimotor neuropathy than in an age-matched control population.54 These studies suggest that a low-level abnormality of glucose metabolism not meeting criteria for DM may still be capable of causing sensory neuropathy in some people. If so, an important clinical implication would be that measurement of FBG is an inadequate screening test in patients with CSP, and that an OGTT is needed to fully exclude abnormal glucose metabolism. This area requires further study.

2328

Diseases of the Peripheral Nervous System

Sjögren’s Syndrome Sjögren’s syndrome is a connective tissue disease characterized by the sicca complex of dry eyes (keratoconjunctivitis sicca) and dry mouth. The disease affects women more than men in a ratio of 9:1. In some patients, arthritis or other systemic symptoms also occur. An autoimmune cause is assumed, based on the presence of lymphocytic inflammatory infiltrates in lacrimal and salivary glands, serologic abnormalities, and an association with other immune-mediated disorders such as rheumatoid arthritis. The diagnosis cannot be made on the basis of any single symptom or test because none is sufficiently sensitive or specific. Diagnostic criteria consider symptoms of oral and ocular dryness, ophthalmologic tests of tear production (the Schirmer test) or keratoconjunctivitis (corneal or conjunctival staining with rose bengal), serologic abnormalities such as antibodies to SS-A (Ro) and SS-B (La), and the presence of lymphocytic inflammatory infiltrates in minor salivary glands biopsied from labial mucosa. Of these, lip biopsy has the highest sensitivity and specificity, but its modestly invasive nature precludes its use as a routine screening test. Several sets of diagnostic criteria have been published; the European classification criteria65,66 are validated and widely used and are listed in Table 104–2.

Sicca symptoms should be interpreted cautiously because dry eyes and mouth can result from other causes, including autonomic neuropathy and drugs. Autonomic dysfunction in idiopathic painful neuropathy is mainly cholinergic, and although the pattern tends to be distal, parasympathetic neuropathy can cause dryness of the eyes and mouth. Patients with painful neuropathy are often treated with amitriptyline or other drugs with anticholinergic side effects. Sarcoidosis can also cause sicca symptoms. Objective evidence is therefore needed to diagnose Sjögren’s syndrome, as described below. Sjögren’s syndrome is associated with several types of peripheral neuropathy. Sensory polyganglionopathy is a distinctive pattern resulting from lymphocytic infiltration of dorsal root ganglia and loss of sensory neurons.18 Acute or chronic in onset, ganglionopathy can cause profound asymmetrical large fiber sensory deficits, ataxia, and loss of reflexes with little or no weakness. In other patients, loss of small neurons results in a “hyperalgesic” form of ganglionopathy with spontaneous pain and loss of pain and temperature sensation.56 Vasculitic multiple mononeuropathies,47 trigeminal neuropathy,17,24 and polyradiculoneuropathy17 occur infrequently. Although frank autonomic failure is unusual in Sjögren’s syndrome,

Table 104–2. European Classification Criteria for Sjögren’s Syndrome I—Ocular symptoms: a positive response to at least one of the following three questions: 1. Have you had daily, persistent dry eyes for more than 3 months? 2. Do you have a recurrent sensation of sand or gravel in the eyes? 3. Do you use tear substitutes more than 3 times a day? II—Oral symptoms: a positive response to at least one of the following three questions: 1. Have you had a daily feeling of dry mouth for more than 3 months? 2. Have you had recurrently or persistently swollen salivary glands as an adult? 3. Do you frequently drink liquids to aid in swallowing dry food? III—Ocular signs: objective evidence of ocular involvement defined as a positive result in at least one of the following two tests: 1. Schirmer I test (ⱕ5 mm in 5 minutes) 2. Rose Bengal score (ⱖ4 according to Van Bijsterveld’s scoring system) IV—Histopathology Focus score ⱖ1 in a minor salivary gland biopsy (a focus is defined as an agglomerate of at least 50 mononuclear cells; the focus score is defined by the number of foci in 4 mm2 of glandular tissue) V—Salivary gland involvement: objective evidence of salivary gland involvement defined as a positive result in at least one of the following three tests: 1. Salivary scintigraphy 2. Parotid sialography 3. Unstimulated salivary flow (ⱕ1.5 mL in 15 minutes) VI—Autoantibodies: presence in the serum of the following autoantibodies: Antibodies to Ro (SS-A), La (SS-B), or both Rules for classification: In patients without any potentially associated disease, the presence of any four of the six items is indicative of primary Sjögren’s syndrome. In patients with a potentially associated disease (such as another connective tissue disease), item I or item II plus any two of items III–V is indicative of secondary Sjögren’s syndrome. Exclusion Criteria Preexisting lymphoma, acquired immunodeficiency syndrome (AIDS), sarcoidosis, graft-vs.-host disease, sialadenosis. Use of antidepressant and antihypertensive drugs, neuroleptics, parasympatholytic drugs. From Vitali, C., Bombardieri, S., Moutsopoulos, H. M., et al.: Preliminary criteria for the classification of Sjogren’s syndrome: results of a prospective concerted action supported by the European community. Arthritis Rheum. 36:340, 1993, with permission.

2329

Cryptogenic Sensory Polyneuropathy

Wright and colleagues found autonomic dysfunction in 90% of patients; in most cases this was mild or asymptomatic.70 Sensory or sensorimotor polyneuropathies are the most common type of Sjögren’s syndrome–associated neuropathy.12,17,37 A distal small fiber pattern sometimes occurs.53 The mechanism in these patients is not always clear, but an inflammatory process is likely given the frequent presence of epineurial perivascular mononuclear cells in the sural nerve,17,37 and the occasional presence of frank necrotizing vasculitis.17,37,47 Some patients with distal symmetrical neuropathies may actually have unrecognized ganglionopathies, since the distribution of symptoms depends on the subset of ganglia that are damaged. Sjögren’s syndrome presents particular problems with respect to the evaluation of CSP. Of greatest difficulty is the fact that sicca complex may be asymptomatic. In patients with Sjögren’s syndrome and chronic polyneuropathy, it is usually the neuropathic symptoms that prompt the patient to seek medical attention; sicca symptoms may be elicited only with specific questioning, if at all. Few such patients will have been previously diagnosed with Sjögren’s syndrome, and they have few symptoms of extraglandular disease or serologic abnormalities. Because of this contrast with classic descriptions of Sjögren’s syndrome, some authors prefer the less committal term sicca syndrome for such patients.17,28 The prevalence of neuropathy in Sjögren’s syndrome is unknown; estimates of 10% to 15% are probably low. Recognition of this association has increased over the last two decades, and such cases were not reported in older studies of causes of polyneuropathy. Undoubtedly some

patients with idiopathic peripheral neuropathies have unrecognized Sjögren’s syndrome. In the only prospective study to address this, van Dijk and colleagues systematically investigated patients with chronic idiopathic axonal polyneuropathy for Sjögren’s syndrome, 49 of whom submitted to a complete work-up, including minor salivary gland biopsy.63 Three of the 49 met criteria for Sjögren’s syndrome. All were females between 69 and 79 years of age, and the clinical features of their neuropathies were not distinctive. Notably, none complained spontaneously of sicca symptoms and two had no specific serologic abnormalities. The best method for detecting Sjögren’s syndrome in patients presenting with neuropathy remains uncertain, and procedures vary widely among clinics. These cases will not be found without a high index of suspicion. Given the very unequal sex ratio in Sjögren’s syndrome, it seems reasonable to focus investigations on women. Van Dijk and colleagues63 proposed that all women with idiopathic polyneuropathy be evaluated for Sjögren’s syndrome with a questionnaire, serology, and ocular tests, and also undergo minor salivary gland biopsy if any of these are positive. These authors suggest limiting tests in men to those with sicca symptoms (Fig. 104–2).

Inherited Neuropathy Inherited disorders account for some cases of initially undiagnosed sensory neuropathies. In Dyck et al.’s series, an inherited disorder was thought to be present in 42% of patients with previously undiagnosed neuropathy

CIAP

FIGURE 104–2 Diagnostic approach with respect to Sjögren’s syndrome in patients with chronic idiopathic axonal polyneuropathy (CIAP). RBS ⫽ rose bengal stain; SSG ⫽ sublabial salivary gland. (From van Dijk, G. W., Notermans, N. C., Kater, L., et al.: Sjögren’s syndrome in patients with chronic idiopathic axonal polyneuropathy. J. Neurol. Neurosurg. Psychiatry 63:376, 1997, with permission.)

Men

Women

Ocular or oral symptoms

Ocular or oral symptoms or abnormal Schirmer or RBS or Anti-Ro/SSA or anti-La/SSB

Yes

No

Yes

No

Abnormal Schirmer or RBS or Anti-Ro/SSA or anti-La/SSB

Stop

SSG biopsy

Stop

Yes

No

SSG biopsy

Stop

2330

Diseases of the Peripheral Nervous System

in whom a diagnosis could be made.9 In over half, review of the family history obtained from the patient was inconclusive, and the diagnosis was based on direct assessment of kin or on distinctive clinical, pathologic, or biochemical features in the index patient. Most inherited cases were classified as hereditary motor and sensory neuropathy (HMSN; i.e., Charcot-Marie-Tooth disease), a sensorimotor neuropathy. Because this study predated DNA testing, the proportion of inherited cases might have been underestimated. Follow-up studies of idiopathic polyneuropathy include a small number of cases eventually determined to be inherited, also usually HMSN. 36,42 Idiopathic sensory polyneuropathy has received less attention, but here too familial disorders are probably common; among patients with painful feet, Novak and colleagues found that 26% had a history of similar symptoms in a first-degree relative.43 Hereditary sensory and autonomic neuropathy type I (HSAN I) is the most common inherited sensory neuropathy. Onset is commonly in the second or third decades but can be later. Classically a rather severe disorder, it is characterized by distal sensory loss preferentially involving the lower limbs. All sensory modalities can be affected, but pain and temperature are impaired more often than proprioception. Lancinating or burning pain, pes cavus, and deafness are variable features. Distal weakness and muscle atrophy are mild if present.

Some patients have plantar ulcers or neurogenic arthropathy. Inheritance is usually autosomal dominant; several families have been found to have missense mutations in the SPTLC1 gene on chromosome 9q22, encoding serine palmitoyltransferase long chain base subunit 1.2,4 HSAN I encompasses a heterogeneous clinical spectrum, and diagnostic difficulty is most likely in individuals with limited phenotypic expression. An autosomal dominant syndrome of burning feet has been described in several families. Dyck and co-workers8 described a family with severe foot pain unaccompanied by clinical or quantitative sensory abnormalities. Based on the inheritance pattern and subtle pathologic and electrodiagnostic findings, the authors concluded that the disorder was an inherited sensory neuropathy, perhaps a mild form of HSAN I. Stogbauer and colleagues59 reported another kindred with burning feet and preferential loss of unmyelinated fibers on sural nerve biopsy, in which linkage analysis excluded HSAN I and Charcot-Marie-Tooth disease type 2B. Other inherited sensory neuropathies are much less common. Causes include familial amyloidosis, Fabry’s disease, Tangier disease, the spinocerebellar ataxias, and the leukodystrophies, each of which is discussed in detail elsewhere in this text. The distinguishing features of each are summarized in Table 104–3.

Table 104–3. Differential Diagnosis of Inherited, Predominantly Sensory Polyneuropathies Inheritance Pattern

Neuropathic Features

Other Features

Familial amyloid polyneuropathy (FAP)

Autosomal dominant

Small fiber Autonomic Carpal tunnel syndrome* Cranial, esp. CN 7†

Fabry’s disease

X-linked recessive

Small fiber Neuropathic pain

Tangier disease

Autosomal dominant

Spinocerebellar ataxia

Autosomal dominant Autosomal recessive

Variable: Distal symmetrical Relapsing asymmetrical Upper limb predominant Large fiber

Depend on subtype: Myopathy Cardiomyopathy Nephrotic syndrome Corneal lattice dystrophy† Cutaneous angiokeratomas Chronic renal failure Stroke, coronary artery disease Orange tonsils Hepatosplenomegaly Coronary artery disease

Leukodystrophies

Autosomal recessive X-linked recessive‡

*FAP type II (transthyretin mutations, usually Ser77Tyr). † FAP type IV (gelsolin mutations). ‡ Adrenoleukodystrophy/adrenomyeloneuropathy.

Large fiber

Depend on subtype: Dysarthria Eye movement abnormalities Retinopathy Extrapyramidal signs Upper motor neuron signs Cognitive impairment Depend on subtype: Upper motor neuron signs Cognitive impairment Extrapyramidal signs Adrenal insufficiency‡

Cryptogenic Sensory Polyneuropathy

Recognition of inherited neuropathy depends primarily on a detailed family history, but an affected individual may truly have no affected kin as a result of chance or a new mutation. In other cases, affected kin go unrecognized because they (or their physicians) incorrectly attribute their symptoms to other causes. Incorrect paternity may also confound the family history. Diagnosis rests on (1) considering an inherited cause in all cases of cryptogenic sensory neuropathy; (2) obtaining a detailed family history, including review of medical records if necessary; (3) maintaining a low threshold for assessment of family members, and (4) appropriate use of biochemical and (increasingly) genetic tests. Even in the absence of a positive family history, an inherited neuropathy should be considered in all individuals with unexplained sensory neuropathy, particularly when accompanied by pes cavus, plantar ulcers or other bony or soft tissue foot complications, hearing loss, or onset in early adulthood.

12.

Orthopedic Disease

13.

When signs of neuropathy are absent or unrecognized, neuropathic foot pain is sometimes incorrectly attributed to a musculoskeletal problem such as plantar fasciitis, flat feet, or bunions. Affected patients often see a variety of specialists, including orthopedic surgeons and podiatrists, and may be treated with orthotics, steroid injections, or surgery. Among 28 patients with chronic progressive SFN, Holland and colleagues noted that 7 had been misdiagnosed with orthopedic disorders and that 4 had undergone surgery.21 Referral to a neurologist is typically delayed until other symptoms such as numbness develop or neuropathy is finally considered after a process of elimination.

REFERENCES 1. Archer, A. G., Watkins, P. J., Thomas, P. K., et al.: The natural history of acute painful neuropathy in diabetes mellitus. J. Neurol. Neurosurg. Psychiatry 46:491, 1983. 2. Bejaoui, K., Wu, C., Scheffler, M. D., et al.: SPTLC1 is mutated in hereditary sensory neuropathy, type 1. Nat. Genet. 27:261, 2001. 3. Benbow, S. J., Wallymahmed, M. E., and MacFarlane, I. A.: Diabetic peripheral neuropathy and quality of life. Q. J. Med. 91:733, 1998. 4. Dawkins, J. L., Hulme, D. J., Brahmbhatt, S. B., et al.: Mutations in SPTLC1, encoding serine palmitoyltransferase, long chain base subunit-1, cause hereditary sensory neuropathy type I. Nat. Genet. 27:309, 2001. 5. Dyck, P. J., Benstead, T. J., Conn, D. L., et al.: Nonsystemic vasculitic neuropathy. Brain 110(Pt. 4):843, 1987. 6. Dyck, P. J., Kratz, K. M., Karnes, J. L., et al.: The prevalence by staged severity of various types of diabetic neuropathy, retinopathy, and nephropathy in a population-based cohort:

7.

8.

9.

10.

11.

14. 15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

2331

the Rochester Diabetic Neuropathy Study. Neurology 43:817, 1993. Dyck, P. J., Larson, T. S., O’Brien, P. C., and Velosa, J. A.: Patterns of quantitative sensation testing of hypoesthesia and hyperalgesia are predictive of diabetic polyneuropathy: a study of three cohorts. Nerve Growth Factor Study Group. Diabetes Care 23:510, 2000. Dyck, P. J., Low, P. A., and Stevens, J. C.: “Burning feet” as the only manifestation of dominantly inherited sensory neuropathy. Mayo Clin. Proc. 58:426, 1983. Dyck, P. J., Oviatt, K. F., and Lambert, E. H.: Intensive evaluation of referred unclassified neuropathies yields improved diagnosis. Ann. Neurol. 10:222, 1981. Fagerberg, S.-E.: Diabetic neuropathy: a clinical and histological study on the significance of vascular affections. Acta Med. Scand. Suppl. 345:1, 1959. Galer, B. S., Gianas, A., and Jensen, M. P.: Painful diabetic polyneuropathy: epidemiology, pain description, and quality of life. Diabetes Res. Clin. Pract. 47:123, 2000. Gemignani, F., Marbini, A., Pavesi, G., et al.: Peripheral neuropathy associated with primary Sjogren’s syndrome. J. Neurol. Neurosurg. Psychiatry 57:983, 1994. Giuliani, M. J., Stewart, J. D., and Low, P. A.: Distal small-fibre neuropathy. In Low, P. A. (ed.): Clinical Autonomic Disorders, 2nd ed. Philadelphia, Lippincott–Raven, p. 699, 1997. Gorson, K. C., and Ropper, A. H.: Idiopathic distal small fiber neuropathy. Acta Neurol. Scand. 92:376, 1995. Gorson, K. C., and Ropper, A. H.: Axonal neuropathy associated with monoclonal gammopathy of undetermined significance. J. Neurol. Neurosurg. Psychiatry 63:163, 1997. Grahmann, F., Winterholler, M., and Neundorfer, B.: Cryptogenetic polyneuropathies: an out-patient follow-up study. Acta Neurol. Scand. 84:221, 1991. Grant, I. A., Hunder, G. G., Homburger, H. A., and Dyck, P. J.: Peripheral neuropathy associated with sicca complex. Neurology 48:855, 1997. Griffin, J. W., Cornblath, D. R., Alexander, E., et al.: Ataxic sensory neuropathy and dorsal root ganglionitis associated with Sjogren’s syndrome. Ann. Neurol. 27:304, 1990. Gupta, S. K., Taly, A. B., Suresh, T. G., et al.: Acute idiopathic axonal neuropathy (AIAN): a clinical and electrophysiological observation. Acta Neurol. Scand. 89:220, 1994. Herrmann, D. N., Griffin, J. W., Hauer, P., et al.: Epidermal nerve fiber density and sural nerve morphometry in peripheral neuropathies. Neurology 53:1634, 1999. Holland, N. R., Crawford, T. O., Hauer, P., et al.: Small-fiber sensory neuropathies: clinical course and neuropathology of idiopathic cases. Ann. Neurol. 44:47, 1998. Holland, N. R., Stocks, A., Hauer, P., et al.: Intraepidermal nerve fiber density in patients with painful sensory neuropathy. Neurology 48:708, 1997. Jamal, G. A., Hansen, S., Weir, A. I., and Ballantyne, J. P.: The neurophysiologic investigation of small fiber neuropathies. Muscle Nerve 10:537, 1987. Kaltreider, H. B., and Talal, N.: The neuropathy of Sjogren’s syndrome: trigeminal nerve involvement. Ann. Intern. Med. 70:751, 1969. Kelkar, P., McDermott, W. R., and Parry, G. J.: Sensorypredominant, painful, idiopathic neuropathy: inflammatory changes in sural nerves. Muscle Nerve 26:413, 2002.

2332

Diseases of the Peripheral Nervous System

26. Kennedy, W. R., Nolano, M., Wendelschafer-Crabb, G., et al.: A skin blister method to study epidermal nerves in peripheral nerve disease. Muscle Nerve 22:360, 1999. 27. Kennedy, W. R., Wendelschafer-Crabb, G., and Johnson, T.: Quantitation of epidermal nerves in diabetic neuropathy. Neurology 47:1042, 1996. 28. Kennett, R. P., and Harding, A. E.: Peripheral neuropathy associated with the sicca syndrome. J. Neurol. Neurosurg. Psychiatry 49:90, 1986. 29. Lacomis, D.: Small-fiber neuropathy. Muscle Nerve 26:173, 2002. 30. Lacomis, D., Giuliani, M. J., Steen, V., and Powell, H. C.: Small fiber neuropathy and vasculitis. Arthritis Rheum. 40:1173, 1997. 31. Lauria, G., Pareyson, D., and Sghirlanzoni, A.: Chronic cryptogenic sensory polyneuropathy. Arch. Neurol. 57:759, 2000. 32. Low, P. A.: Autonomic nervous system function. J. Clin. Neurophysiol. 10:14, 1993. 33. Matthews, W. B.: Cryptogenic polyneuritis. Proc. R. Soc. Med. 45:667, 1952. 34. McArthur, J. C., Stocks, E. A., Hauer, P., et al.: Epidermal nerve fiber density: normative reference range and diagnostic efficiency. Arch. Neurol. 55:1513, 1998. 35. McLeod, J. G.: Autonomic dysfunction in peripheral nerve disorders. Curr. Opin. Neurol. Neurosurg. 5:476, 1992. 36. McLeod, J. G., Tuck, R. R., Pollard, J. D., et al.: Chronic polyneuropathy of undetermined cause. J. Neurol. Neurosurg. Psychiatry 47:530, 1984. 37. Mellgren, S. I., Conn, D. L., Stevens, J. C., and Dyck, P. J.: Peripheral neuropathy in primary Sjogren’s syndrome. Neurology 39:390, 1989. 38. Mitsumoto, H., and Wilbourn, A. J.: Causes and diagnosis of sensory neuropathies: a review. J. Clin. Neurophysiol. 11:553, 1994. 39. Nemni, R., Fazio, R., Quattrini, A., et al.: Antibodies to sulfatide and to chondroitin sulfate C in patients with chronic sensory neuropathy. J. Neuroimmunol. 43:79, 1993. 40. Notermans, N. C., Wokke, J. H., Franssen, H., et al.: Chronic idiopathic polyneuropathy presenting in middle or old age: a clinical and electrophysiological study of 75 patients. J. Neurol. Neurosurg. Psychiatry 56:1066, 1993. 41. Notermans, N. C., Wokke, J. H., van den Berg, L. H., et al.: Chronic idiopathic axonal polyneuropathy: comparison of patients with and without monoclonal gammopathy. Brain 119(Pt. 2):421, 1996. 42. Notermans, N. C., Wokke, J. H., van der Graaf, Y., et al.: Chronic idiopathic axonal polyneuropathy: a five year follow up. J. Neurol. Neurosurg. Psychiatry 57:1525, 1994. 43. Novak, V., Freimer, M. L., Kissel, J. T., et al.: Autonomic impairment in painful neuropathy. Neurology 56:861, 2001. 44. Novella, S. P., Inzucchi, S. E., and Goldstein, J. M.: The frequency of undiagnosed diabetes and impaired glucose tolerance in patients with idiopathic sensory neuropathy. Muscle Nerve 24:1229, 2001. 45. Periquet, M. I., Novak, V., Collins, M. P., et al.: Painful sensory neuropathy: prospective evaluation using skin biopsy. Neurology 53:1641, 1999.

46. Pestronk, A., Li, F., Griffin, J., et al.: Polyneuropathy syndromes associated with serum antibodies to sulfatide and myelin-associated glycoprotein. Neurology 41:357, 1991. 47. Peyronnard, J. M., Charron, L., Beaudet, F., and Couture, F.: Vasculitic neuropathy in rheumatoid disease and Sjogren syndrome. Neurology 32:839, 1982. 48. Prineas, J.: Polyneuropathies of undetermined cause. Acta Neurol Scand Suppl. 44:3, 1970. 49. Ravits, J. M.: AAEM minimonograph #48: autonomic nervous system testing. Muscle Nerve 20:919, 1997. 50. Ropper, A. H., and Gorson, K. C.: Neuropathies associated with paraproteinemia. N. Engl. J. Med. 338:1601, 1998. 51. Rose, F. C.: Discussion on neuropathies in rheumatic diseases and steroid therapy. Proc. R. Soc. Med. 53:51, 1960. 52. Said, G., Slama, G., and Selva, J.: Progressive centripetal degeneration of axons in small fibre diabetic polyneuropathy. Brain 106(Pt. 4):791, 1983. 53. Sandroni, P., and Low, P. A.: Autonomic peripheral neuropathies: clinical presentation, diagnosis, and treatment. J. Clin. Neuromuscul. Dis. 2:147, 2001. 54. Singleton, J. R., Smith, A. G., and Bromberg, M. B.: Painful sensory polyneuropathy associated with impaired glucose tolerance. Muscle Nerve 24:1225, 2001. 55. Skyler, J. S.: Diabetic complications: the importance of glucose control. Endocrinol. Metab. Clin. North Am. 25:243, 1996. 56. Smith, B. E., Windebank, A. J., and Dyck, P. J. Non-malignant inflammatory sensory polyganglionopathy. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1525, 1993. 57. Stalberg, E. V., and Nogues, M. A.: Automatic analysis of heart rate variation: I. Method and reference values in healthy controls. Muscle Nerve 12:993, 1989. 58. Stewart, J. D., Low, P. A., and Fealey, R. D.: Distal small fiber neuropathy: results of tests of sweating and autonomic cardiovascular reflexes. Muscle Nerve 15:661, 1992. 59. Stogbauer, F., Young, P., Kuhlenbaumer, G., et al.: Autosomal dominant burning feet syndrome. J. Neurol. Neurosurg. Psychiatry 67:78, 1999. 60. Teunissen, L. L., Eurelings, M., Notermans, N. C., et al.: Quality of life in patients with axonal polyneuropathy. J. Neurol. 247:195, 2000. 61. Thomas, P. K.: Classification, differential diagnosis, and staging of diabetic peripheral neuropathy. Diabetes 46(Suppl. 2):S54, 1997. 62. Tobin, K., Giuliani, M. J., and Lacomis, D.: Comparison of different modalities for detection of small fiber neuropathy. Clin. Neurophysiol. 110:1909, 1999. 63. van Dijk, G. W., Notermans, N. C., Kater, L., et al.: Sjogren’s syndrome in patients with chronic idiopathic axonal polyneuropathy. J. Neurol. Neurosurg. Psychiatry 63:376, 1997. 64. van Dijk, G. W., Wokke, J. H., Notermans, N. C., et al.: Indications for an immune-mediated etiology of idiopathic sensory neuronopathy. J. Neuroimmunol. 74:165, 1997. 65. Vitali, C., Bombardieri, S., Moutsopoulos, H. M., et al.: Preliminary criteria for the classification of Sjogren’s syndrome: results of a prospective concerted action

Cryptogenic Sensory Polyneuropathy supported by the European Community. Arthritis Rheum. 36:340, 1993. 66. Vitali, C., Bombardieri, S., Moutsopoulos, H. M., et al.: Assessment of the European classification criteria for Sjogren’s syndrome in a series of clinically defined cases: results of a prospective multicentre study. The European Study Group on Diagnostic Criteria for Sjogren’s Syndrome. Ann. Rheum. Dis. 55:116, 1996. 67. Vrancken, A. F., Franssen, H., Wokke, J. H., et al.: Chronic idiopathic axonal polyneuropathy and successful aging of

2333

the peripheral nervous system in elderly people. Arch. Neurol. 59:533, 2002. 68. Wolfe, G. I., Baker, N. S., Amato, A. A., et al.: Chronic cryptogenic sensory polyneuropathy: clinical and laboratory characteristics. Arch. Neurol. 56:540, 1999. 69. Wolfe, G. I., and Barohn, R. J.: Cryptogenic sensory and sensorimotor polyneuropathies. Semin. Neurol. 18:105, 1998. 70. Wright, R. A., Grant, I. A., and Low, P. A.: Autonomic neuropathy associated with sicca complex. J. Auton. Nerv. Syst. 75:70, 1999.

105 Neuropathies with Systemic Vasculitis MICHAEL P. COLLINS AND JOHN T. KISSEL

Overview of the Vasculitides Introduction to Neuropathies with Systemic Vasculitis Classification Epidemiology Pathogenesis Adhesion Molecules Chemokines Leukocyte Emigration Adhesion Molecules in Vasculitis Cytokines in Vasculitis Humoral and Cellular Immunity in Vasculitis Etiology Infections

Drugs Malignancy Systemic Necrotizing Vasculitis Classic Polyarteritis Nodosa Microscopic Polyangiitis Churg-Strauss Syndrome Wegener’s Granulomatosis Vasculitis Associated with Connective Tissue Diseases Rheumatoid Arthritis Systemic Lupus Erythematosus Sjögren’s Syndrome Scleroderma Behçet’s Disease Hypersensitivity Vasculitis

OVERVIEW OF THE VASCULITIDES Vasculitides are systemic or localized conditions in which blood vessel walls are infiltrated and damaged by inflammatory cells, with secondary ischemic injury to the involved tissues. Many classification schemes have been proposed for these clinically, pathologically, and etiologically heterogeneous diseases, but all are provisional, given incomplete understanding of first causes and pathogenic mechanisms. Most vasculitides are idiopathic and presumably mediated by autoimmune processes. Of the idiopathic vasculitic disorders, many occur as multiorgan-involving, potentially life-threatening diseases, categorized according to clinicopathologic and laboratory factors (e.g., polyarteritis nodosa, microscopic polyangiitis, and Wegener’s granulomatosis). Others occur as part of a systemic connective tissue disease (e.g., rheumatoid arthritis). Still others are restricted to a single organ such as the peripheral nervous system (e.g., nonsystemic vasculitic neuropathy). All of the idiopathic vasculitides ensue from cellular and humoral immune responses to specific antigens that trigger a complex cascade of immune and inflammatory reactions, in which T and B

Henoch-Schönlein Purpura Temporal Arteritis Nonsystemic Vasculitic Neuropathy Genetics Clinical Features Natural History Differential Diagnosis Diagnosis Treatment Antigen Removal Immunosuppressive Therapy Limitation of Adverse Effects Outcome

lymphocytes, immune effector cells, cytokines, chemokines, growth and angiogenic factors, procoagulant tendencies, activated endothelial cells, cellular adhesion molecules, reactive oxygen species, and degradative enzymes all play key roles. Immunopathogenic mechanisms relevant to vasculitis include deposition of circulating immune complexes, delayed-type hypersensitivity, and formation of antineutrophil cytoplasmic autoantibodies, cytotoxic T lymphocytes, and (possibly) anti–endothelial cell autoantibodies. The study of genetic influences on the incidence and manifestations of vasculitis remains in its infancy. Uncommonly, vasculitis results from specific etiologic triggers. For example, almost any infectious organism can produce an immune-complex vasculitis. Agents having wellestablished associations with vasculitic neuropathy include hepatitis B virus, hepatitis C virus, human immunodeficiency virus, cytomegalovirus, and Borrelia burgdorferi. Many drugs and cancers have also been implicated. Cryoglobulinemia related to hepatitis C virus, other chronic infections, connective tissue diseases, and lymphoproliferative diseases is often complicated by vasculitis involving peripheral nerves. 2335

2336

Diseases of the Peripheral Nervous System

Systemic vasculitides affecting small arteries and arterioles commonly produce peripheral neuropathies. Most patients with vasculitic neuropathy have a stereotypic onset and course, which was formerly designated “mononeuritis multiplex.” Today, the more understandable term is multiple mononeuropathy (or multifocal neuropathy). Typically, symptoms of “asleep” numbness, tingling, contact hypersensitivity, pain, and weakness begin in the territory of a single peripheral nerve. Sequentially, within days to weeks, other nerves become affected. If, as is often the case, evolution of the neuropathy is gradual rather than abrupt, regions of single nerve involvement “overlap,” creating an asymmetrical, distally accentuated polyneuropathy. Infrequently vasculitic neuropathy presents as a symmetrical polyneuropathy, but in these cases close attention to history and examination almost always reveals asymmetrical, non–length-dependent features. Painless, proximally predominant, or purely motor neuropathies are unlikely to be vasculitic in origin. In patients with vasculitic neuropathy secondary to a systemic disease, multiorgan manifestations soon follow the neuropathy, especially skin, renal, rheumatic, pulmonary, or gastrointestinal involvement. Weight loss, fatigue, myalgias, and arthralgias occur in both systemic and nonsystemic vasculitic neuropathy. To diagnose a vasculitic neuropathy, laboratory studies, electrodiagnostic testing, and nerve or nerve/muscle biopsy are usually required. Vasculitic neuropathy must be distinguished from other asymmetrical or multifocal neuropathies, in particular multifocal demyelinating neuropathies (e.g., Lewis-Sumner syndrome), multiple entrapments, and neoplastic infiltration. Laboratory studies are designed to assess for non-neurologic organ involvement, markers of systemic inflammation or autoantibody production, and specific disorders associated with vasculitis. In nonsystemic vasculitic neuropathy, blood work is generally less abnormal than in systemic vasculitides, but a mildly to moderately elevated sedimentation rate occurs in 60% of patients. Electrodiagnostic studies are crucial for documenting asymmetrical features of the neuropathy, excluding a demyelinating process, and assisting with the determination of which nerve to biopsy. The classic electrodiagnostic profile is that of an active, asymmetrical or multifocal, distal-predominant, axonal, sensorimotor neuropathy. Partial motor conduction blocks resulting from ongoing axonal degeneration and not segmental demyelination occur uncommonly, but generally disappear on repeat testing. For a definite tissue diagnosis of vasculitic neuropathy, sural nerve or superficial peroneal nerve/peroneus brevis muscle are most commonly biopsied. Muscle biopsy increases the diagnostic yield, at least in patients with systemic disease. Diagnostic sensitivity of these biopsies for vasculitis is approximately 60%. Pathologic diagnosis requires inflammatory cell infiltration within the vessel wall and signs of vascular injury such as fibrinoid necrosis.

Many pathologic alterations suggestive but not diagnostic of vasculitis have been identified; they should be applied in cases lacking definite necrotizing changes. Vasculitis in peripheral nerves is predilected for epineurial vessels. Active lesions are characterized by perivascular collections of T cells and macrophages and vascular deposits of immunoglobulin and complement proteins. Affected nerve fascicles show axon loss and wallerian-like degeneration, typically in an asymmetrical or multifocal pattern. Untreated primary systemic necrotizing vasculitides are almost always fatal. The associated neuropathies generally progress but sometimes stabilize, regress, or relapse. There are no natural history data on untreated nonsystemic vasculitic neuropathy. Traditionally, therapy of the vasculitides has been grounded on the results of retrospective studies and case series. However, more rational treatment strategies have emerged in the last 10 years, thanks to widely accepted classification systems, improved knowledge of pathogenic mechanisms, identification of prognostic factors, formation of multicenter collaborative study groups, and an increasing number of randomized controlled trials. Treatment of vasculitis always begins with removal of triggering antigens. Any recently prescribed drug should be discontinued. Hepatitis B–associated polyarteritis nodosa is generally treated with a short course of prednisone, plasma exchanges, and antiviral agents (interferon- or lamivudine). Patients with hepatitis C–related cryoglobulinemic vasculitis can also be treated with plasma exchanges and antiviral drugs (pegylated interferon- and ribavirin), but immunosuppressive therapy is favored for severe cases because interferon- can sometimes worsen the vasculitis or produce a new neuropathy. Human immunodeficiency virus–associated vasculitis is often self-limited; uncontrolled studies suggest efficacy for prednisone, antiretroviral drugs, plasma exchange, and/or intravenous immunoglobulins. Vasculitis in Lyme disease and cytomegalovirus infection responds to antimicrobial treatment of these conditions. Paraneoplastic vasculitic neuropathy usually improves with cancer-directed modalities, but in patients failing antineoplastic measures, combination immunosuppressive therapy is preferred. In patients for whom no inciting antigen is identified, immunosuppressive therapy is necessary. For systemic necrotizing vasculitides, combination therapy with prednisone and cyclophosphamide is usually indicated. It yields lower mortality rates and better outcomes than prednisone alone. Corticosteroid monotherapy, in contrast, is generally sufficient for temporal arteritis and “good-prognosis” polyarteritis nodosa and Churg-Strauss syndrome (no renal, cardiac, gastrointestinal, or central nervous system involvement). For nonsystemic vasculitic neuropathy, no controlled treatment trials have been conducted. Some investigators advocate corticosteroid monotherapy; others favor combination therapy. Based on data from two retrospective cohort surveys of patients with this condition, combination therapy is more effective that corticosteroid

Neuropathies with Systemic Vasculitis

monotherapy in inducing remission and improving disability. Despite the established efficacy of combination therapy in controlling vasculitis, newer treatment strategies have emerged, designed to reduce the likelihood of drug-related complications. In one approach, continuous oral cyclophosphamide is replaced by periodic intravenous pulses. A second approach involves replacing oral cyclophosphamide with a less toxic, maintenance agent once remission has been induced by combination therapy. Azathioprine is the most widely used remissionmaintaining drug. More rapid prednisone-tapering regimens have also been proposed. The optimal regimen for each type of vasculitis is not yet established. Patients with refractory or relapsing disease have been successfully treated with intravenous immunoglobulins, plasma exchanges, cyclosporine, etoposide, antithymocyte globulin, rituximab, infliximab, etanercept, mycophenolate mofetil, and autologous stem cell transplantation. An important aspect of vasculitic neuropathy management concerns the need for a program to limit the occurrence of adverse effects associated with immunosuppressive drugs. Five-year survival rates for systemic vasculitides and nonsystemic vasculitic neuropathy range from 70% to 90% in modern series. Nonsystemic vasculitic neuropathy is less likely to be lethal than most systemic vasculitides. Relapse rates for nonsystemic vasculitic neuropathy range from 25% to 45% but may be reduced by a longer course of combination therapy. Final neurologic outcome in nonsystemic vasculitic neuropathy is surprisingly good (e.g., only 15% of patients with moderately severe or severe disability), but most survivors have some degree of chronic pain.

INTRODUCTION TO NEUROPATHIES WITH SYSTEMIC VASCULITIS The term vasculitis refers to a pathologic process in which blood vessel walls are inflamed and partially destroyed, secondarily damaging the surrounding tissues through ischemic and inflammatory mechanisms.116,182 Multiple combinations of organ systems can be involved, resulting in a wide variety of clinical manifestations. Specific clinical features depend on the location and size of the involved vessels, the underlying etiology and pathogenesis, the host’s repertoire of immune responses, and the presence of comorbid conditions. Systemic vasculitides involving small to medium-sized arteries commonly affect epineurial vessels in the vasa nervorum and, thereby, produce peripheral neuropathies. In contrast, vasculitides with primarily large vessel or microvascular involvement tend to spare epineurial vessels and, thus, the peripheral nervous system (PNS). In some individuals, vasculitis remains confined to the peripheral nerves and manifests as a chronic, progressive, occasionally relapsing neuropathy. In others, vasculitic neuropathy heralds a life-threatening systemic illness.

2337

The pathologic depiction of vasculitis was made possible by several histopathologic advances in the 19th century.130 The first complete description of a patient with systemic vasculitis is credited to Kussmaul and Maier in 1866.367 They reported a patient with fever, anemia, rapidly progressive neuropathy, necrotizing enteritis, and nephritis whose autopsy showed widespread, nodular, inflammatory lesions of small and medium-sized arteries throughout the body. This new disease was called periarteritis nodosa. The term polyarteritis nodosa (PAN) was introduced by Ferrari in 1903 to recognize the fact that inflammatory infiltrates could be found in all layers of the vessel wall and were not confined to the periarterial region.186 For the next five decades, almost all cases of vasculitis were reported as periarteritis nodosa or PAN, irrespective of differentiating features. Although not immediately recognized as such, forms of vasculitis clinicopathologically distinct from PAN had already begun to appear. A patient with temporal arteritis (TA) was first described by Hutchinson in 1890.317 A form of systemic vasculitis characterized by connective tissue necrosis, granulomatous inflammation, upper and lower respiratory tract involvement, and necrotizing glomerulonephritis was first reported by Klinger in 1931 and, more rigorously, by Wegener in 1936.361,655 This condition was established as a distinct clinicopathologic entity and designated Wegener’s granulomatosis (WG) by Godman and Churg in 1954.234 A “microscopic” variant of PAN, characterized by predominant small vessel involvement and rapidly progressive glomerulonephritis, was described by Wohlwill in 1923 and Davson et al. in 1948.143,662 This disorder is now known as microscopic polyangiitis (MPA). Another type of small vessel vasculitis, dependent on exposure to foreign antigens such as sulfonamides and primarily involving the skin, was differentiated from PAN by Zeek and colleagues in 1948, and labeled hypersensitivity vasculitis (HSV).676 In 1951, Churg and Strauss described a syndrome of eosinophilia, asthma, necrotizing vasculitis, and extravascular granulomas, designated allergic granulomatous angiitis and now referred to as the Churg-Strauss syndrome (CSS).103 Vasculitis of the PNS associated with rheumatoid arthritis (RA) was first reported by Bannatyne in 1898.38 Little attention was paid to peripheral nerve involvement in vasculitis until the studies of Kernohan and Woltman in 1938, Boyd in 1940, and Lovshin and Kernohan in 1948.69,348,394 These workers made the important observations that vasculitic neuropathy may herald the systemic disease, spontaneously regress, and present in a distal, symmetrical fashion. Kernohan and Woltman394 were the first to report a case of PAN restricted to the PNS, a concept that then lay dormant until the 1980s. A seminal histologic study of vasculitic neuropathy was published by Dyck and colleagues in 1972.166 This study correlated the three-dimensional morphology of nerve fiber degeneration with sites of vascular pathology, providing persuasive evidence that ischemia accounts for the fiber

2338

Diseases of the Peripheral Nervous System

damage. In the last two decades, several series of patients with vasculitic neuropathy have been reported, better defining the clinical and immunopathologic spectrum of this disorder.67,91,96,99,118,120,142,165,279,285,286,290,360,447,472,495 Many of the recent series have described patients affected by vasculitis restricted to the PNS.

CLASSIFICATION An ideal classification system for vasculitis would distinguish one type of vasculitis from another by etiology and pathogenesis, permit prognostication, and guide therapeutic decisions. Such a system is not yet at hand, but significant progress has been made in the last 50 years. Many classification systems have been proposed based on such criteria as etiology, size and type of involved vessels, histopathologic characteristics, and clinical attributes.397 In most schemes, primary vasculitides—those with no known etiology or relationship to another disorder—have been distinguished from secondary vasculitides—those with an accepted etiology or predisposing disease.

In 1990, the American College of Rheumatology (ACR) Subcommittee on Classification of Vasculitis published diagnostic criteria for seven forms of vasculitis: PAN, CSS, WG, HSV, Henoch-Schönlein purpura (HSP), giant cell (temporal) arteritis (GCA), and Takayasu’s arteritis.314 These criteria were based on an analysis of 678 cases from 48 centers. They were formulated by comparing findings in one form of vasculitis with those in the others and determining features that best distinguished each type. While useful for classifying patients involved in studies, these criteria are not appropriate for diagnosing new patients with suspected vasculitis because they do not distinguish vasculitic from nonvasculitic diseases. One study of 198 possible vasculitis patients found positive predictive values of only 17% to 29% for WG, PAN, GCA, and HSV using this scheme.515 In 1994, the Chapel Hill Consensus Conference (CHCC) on the Nomenclature of Systemic Vasculitis proposed a new nomenclature that has achieved widespread acceptance.330 A consensus of experts selected names and definitions for 10 well-recognized expressions of vasculitis based on the size and histopathology of the involved vessels (Table 105–1 and Fig. 105–1). This nomenclature

Table 105–1. Names and Definitions of Vasculitides Adopted by the Chapel Hill Consensus Conference on the Nomenclature of Systemic Vasculitis330 Large Vessel Vasculitis Giant cell (temporal) arteritis

Takayasu’s arteritis Medium-Sized Vessel Vasculitis Polyarteritis nodosa (classic polyarteritis nodosa) Kawasaki disease

Small Vessel Vasculitis Wegener’s granulomatosis*

Churg-Strauss syndrome*

Microscopic polyangiitis*

Henoch-Schönlein purpura

Essential cryoglobulinemic vasculitis

Cutaneous leukocytoclastic angiitis

Granulomatous arteritis of the aorta and its major branches, with a predilection for the extracranial branches of the carotid artery. Often involves the temporal artery. Usually occurs in patients older than 50 and often is associated with polymyalgia rheumatica. Granulomatous inflammation of the aorta and its major branches. Usually occurs in patients younger than 50. Necrotizing inflammation of medium-sized or small arteries without glomerulonephritis or vasculitis in arterioles, capillaries, or venules. Arteritis involving large, medium-sized, and small arteries, associated with mucocutaneous lymph node syndrome. Coronary arteries are often involved. Aorta and veins may be involved. Usually occurs in children. Granulomatous inflammation involving the respiratory tract and necrotizing vasculitis affecting small to medium-sized vessels (e.g., capillaries, venules, arterioles, and arteries). Necrotizing glomerulonephritis is common. Eosinophil-rich and granulomatous inflammation involving the respiratory tract and necrotizing vasculitis affecting small to medium-sized vessels, associated with asthma and eosinophilia. Necrotizing vasculitis, with few or no immune deposits, affecting small vessels (i.e., capillaries, venules, or arterioles). Necrotizing arteritis involving small and mediumsized arteries may be present. Necrotizing glomerulonephritis is very common. Pulmonary capillaritis often occurs. Vasculitis with immunoglobulin A–dominant immune deposits, affecting small vessels (i.e., capillaries, venules, or arterioles). Typically involves skin, gut, and glomeruli and is associated with arthralgias or arthritis. Vasculitis, with cryoglobulin immune deposits, affecting small vessels (i.e., capillaries, venules, or arterioles) and associated with cryoglobulins in serum. Skin and glomeruli are often involved. Isolated cutaneous leukocytoclastic angiitis without systemic vasculitis or glomerulonephritis.

*Strongly associated with antineutrophil cytoplasmic autoantibodies.

2339

Neuropathies with Systemic Vasculitis

those resulting from immunologic mechanisms—systemic necrotizing vasculitides, hypersensitivity vasculitides, giant cell arteritides, and localized vasculitides (Table 105–2). This scheme encompasses most of the traditional vasculitides outlined above, but excludes Kawasaki disease and Takayasu’s arteritis because they do not cause a vasculitic neuropathy. Unlike the ACR and CHCC systems, this classification includes nonsystemic vasculitic neuropathy (NSVN), in which vasculitis is clinically restricted to the PNS.165

EPIDEMIOLOGY

FIGURE 105–1 Schematic diagram demonstrating spectrum of predominantly involved vessels for each of the vasculitides defined by the Chapel Hill Consensus Conference on the Nomenclature of Systemic Vasculitis. (Modified from Jennette, J. C., Falk, A. J., Andrassy, K., et al.: Nomenclature of systemic vasculitides: proposal of an international consensus conference. Arthritis Rheum. 37:187, 1994.)

had the advantage of providing simple, unambiguous pathologic definitions and was the first to recognize MPA as a separate diagnostic entity—a small vessel, pauciimmune, nongranulomatous vasculitis associated with rapidly progressive glomerulonephritis. One controversial proposal of this conference was that classic PAN (c-PAN) be restricted to cases affecting small to medium-sized arteries alone. Typical cases of PAN revealing any pathologic involvement of smaller vessels (e.g., in skin or nerves) were reclassified as MPA. In consequence, c-PAN has become rare in recent epidemiologic studies.241,523,652 In addition, the conference excluded HSV as a diagnostic entity. Most patients who would have been diagnosed with HSV under the ACR scheme were reassigned to MPA or cutaneous leukocytoclastic angiitis (CLA), a leukocytoclastic vasculitis confined to the skin. As might be predicted, the CHCC definitions and ACR criteria, when applied to the same group of patients, have a low concordance rate.76 The main drawback of the CHCC definitions is their reliance on histologic features, which can limit their clinical utility when appropriate biopsy material is missing or inconclusive.76,397 The CHCC definitions were not designed around, and did not take into account, nerve pathology and so may not be the best criteria to use in neuropathy patients. Vasculitic disorders associated with neuropathy can be divided into five groups, based on the size of involved vessels, presumed immunopathogenesis, and management considerations: those resulting from direct infection and

In general, the vasculitides are uncommon diseases (Table 105–3).239,241,523,652 TA and the predominantly cutaneous vasculitides occur most frequently, followed by rheumatoid vasculitis (RV), WG, and MPA. The incidence of vasculitic neuropathy, however, is quite different because of the variable prevalence of neuropathy in these conditions and the existence of an isolated form of PNS vasculitis (NSVN). Peripheral nerve involvement is common in the systemic necrotizing vasculitides and rare in TA and the predominantly cutaneous vasculitides. The relative prevalence of the different types of vasculitic neuropathy, compiled from series dedicated to such patients, is shown in Table 105–4. The three most commonly encountered vasculitic neuropathies are NSVN, PAN (now MPA)–associated neuropathy, and rheumatoid vasculitic neuropathy. Vasculitic neuropathy can occur at any age, but most commonly develops in the sixth through eighth decades. The mean age of onset ranges from 51 to 67 years in various series.165,286 For NSVN, pooled data from six studies yields an overall female:male ratio of 4:3.120,142,165,459,495,544 For the systemic vasculitides, men are somewhat more frequently affected than women in c-PAN, MPA, RV, and HSP; women are preferentially affected in TA, Takayasu’s arteritis, systemic lupus erythematosus (SLE)–related vasculitis, Behçet’s disease, and cryoglobulinemic vasculitis; and men and women are afflicted with equal frequency in WG and CSS.182,239

PATHOGENESIS Except for cases caused by direct invasion of vascular structures by microorganisms or neoplastic cells, vasculitides are assumed to result from autoimmune reactions, although the specific immunologic events producing vascular destruction are incompletely understood.35,135,590 Whereas progress has been made in elucidating effector mechanisms of the vasculitic immune response, knowledge of the initial, antigen-triggered processes is still primitive. Moreover, there is only rudimentary knowledge of the

2340

Diseases of the Peripheral Nervous System

Table 105–2. Classification of Vasculitides Associated with Neuropathy A. Vasculitides Resulting from Direct Infection 1. Bacterial (e.g., group A -hemolytic streptococcus, infective endocarditis, Lyme disease) 2. Viral (e.g., cytomegalovirus, HIV, HTLV-1, varicella-zoster virus) B. Vasculitides Resulting from Immunologic Mechanisms 1. Systemic Necrotizing Vasculitis a. Classic polyarteritis nodosa b. Antineutrophil cytoplasmic autoantibody-associated (1) Microscopic polyangiitis (2) Churg-Strauss syndrome (3) Wegener’s granulomatosis c. Hepatitis B–associated polyarteritis nodosa d. Vasculitis with connective tissue disease (1) Rheumatoid arthritis (2) Sjögren’s syndrome (3) Systemic lupus erythematosus (4) Mixed connective tissue disease (5) Relapsing polychondritis (6) Behçet’s disease 2. Hypersensitivity Vasculitis a. Henoch-Schönlein purpura b. Drug-induced vasculitis (e.g., amphetamines, cocaine)* c. Vasculitis associated with infections* d. Cryoglobulinemic vasculitis (hepatitis C)* e. Vasculitis associated with malignancy* f. Diabetic and nondiabetic (idiopathic) lumbosacral radiculoplexus neuropathy (?) 3. Giant Cell (Temporal) Arteritis 4. Localized Vasculitis a. Nonsystemic vasculitic neuropathy HIV  human immunodeficiency virus; HTLV-1  human T-lymphotropic virus 1. *Each of these entities can also produce a systemic necrotizing vasculitis.

Table 105–3. Annual Incidence Rates of Systemic Vasculitides (cases/year/1 million inhabitants) Norwich Health Authority, UK239,652,653 (persons age  15 yr)

Lugo, Spain238,239 (persons age  20 yr)

Germany523 (all ages)

13* 15.4 12.5 10.6† 8.4‡ 3.6 3.4‡ 2.7† 1.2 0.8 0‡ 0 0

58 (20 years) 46 (all ages) 28.6 12.2 4.8† 11.6‡ 7.9 7.9‡ 1.1† 4.8 4.2 1.1‡ 3.2 0

11.6 8.7 NA 6.2‡ 2.4‡ NA 9.0‡ 0.6‡ NA NA 1.1‡ NA 0.6

Temporal arteritis Cutaneous leukocytoclastic angiitis Rheumatoid vasculitis Wegener’s granulomatosis Microscopic polyangiitis Systemic lupus erythematosus Henoch-Schönlein purpura Churg-Strauss syndrome Essential cryoglobulinemic vasculitis Malignancy-related vasculitis Classic polyarteritis nodosa Sjögren’s syndrome Takayasu’s arteritis NA  not analyzed. *Data from Jonasson et al.337 † Based on American College of Rheumatology Criteria. ‡ Based on Chapel Hill Consensus Conference Criteria.

2341

Neuropathies with Systemic Vasculitis

Table 105–4. Relative Prevalence of Vasculitic Neuropathy Types Series

c-PAN NSVN or MPA

RV

UCTD CSS SLE SjS WG Other

Kissel et al. (1985)360 Harati and Naikan (1986)279 Dyck et al. (1987)165 Said et al. (1988)544

7 21 20 25

3 2 32 24

0 2 2 26

5 4 0 6

0 0 4 6

1 0 1 1

0 0 3 2

0 0 1 0

Panegyres et al. (1990)495 Hawke et al. (1991)290; McComb et al. (1987)421 Midroni and Bilbao (1995)435 Nicolai et al. (1995)472 Singhal et al. (1995)583 Ohkoshi et al. (1996)486 Collins et al. (2000)118 Sanchez et al. (2001)550

7 2

0 11

5 7

0 8

3 3

0 3

1 1

1 1

0 Scl (1) CA (3) Scl (1), HIV (1) Sarcoid (1), Mastocytosis (1), TA (1), PM (1), HIV (3), Scl (1) MGUS (1) CLL/Cryo (1)

4 2 7 1 11 5

3 2 6 6 6 12

8 1 0 1 8 0

10 7 4 0 0 0

3 4 2 3 0 3

2 0 4 2 1 2

1 0 0 0 0 0

0 0 1 0 2 2

HSV (1) Cryo (3), HIV (1) 0 0 MGUS (3), HSV (2), CA (2), Zoster (1) Sarcoid (1), Waldenström’s (1)

112

107

60

40

31

17

8

8

CA: 8 (2%); MGUS: 5 (1%); HIV: 5 (1%); Cryo: 3 (0.7%); HSV: 3 (0.7%); Scl: 3 (0.7%); Sarcoid: 2 (0.5%); TA: 1 (0.2%); PM: 1 (0.2%); Zoster: 1 (0.2%)

(27%)

(26%)

Total (n  415)

(14%) (10%) (7%) (4%) (2%) (2%)

CA  cancer; CLL  chronic lymphocytic leukemia; c-PAN  classic polyarteritis nodosa; Cryo  cryoglobulinemia; CSS  Churg-Strauss syndrome; HIV  human immunodeficiency virus; HSV  hypersensitivity vasculitis; MGUS  monoclonal gammopathy of unknown significance; MPA  microscopic polyangiitis; NSVN  nonsystemic vasculitic neuropathy; PM  polymyositis; RV  rheumatoid vasculitis; Scl  scleroderma; SjS  Sjögren’s syndrome; SLE  systemic lupus erythematosus; TA  temporal arteritis; UCTD  undifferentiated connective tissue disease–related vasculitis; WG  Wegener’s granulomatosis.

chemical and anatomic factors determining which organs and vessels are involved. Multiple triggering events, antigens, immunologic responses, and perpetuating agents probably exist. On a fundamental level, all autoimmune vasculitides are assumed to commence with antigen recognition by the host, triggering a humoral or cellular immune response (or both), which primes circulating leukocytes, activates endothelial cells, and upregulates antigen-presenting cells in and around blood vessel walls. This results in the adherence of selected leukocytes to the endothelial surface, an interaction that depends on a complex array of cellular adhesion molecules, cytokines, chemotactic agents, and other inflammatory regulators, which mediate leukocyte and endothelial cell activation, recruitment of additional effector cells, and transendothelial migration of leukocytes.349,456,610 Effector cells include neutrophils, eosinophils, monocytes/macrophages, T lymphocytes, and natural killer (NK) cells. They damage vessel walls by degradative enzyme and cytotoxic cytokine release, respiratory burst-induced production of toxic oxygen free radicals, perforin- or complement membrane attack complex–mediated cytolysis, and enzyme- or Fas ligand– induced apoptosis.131,536,608 Involved cells continue to release cytokines and expose tissue factors that stimulate coagulation, myointimal cell proliferation, vessel wall

fibrosis, and angiogenesis. Anti-inflammatory cytokines and growth factors eventually downregulate the inflammatory mediators, and repair gains dominance over destruction.133,607 The final result is a thickened, scarred, neovascularized, and often thrombosed or occluded vessel.

Adhesion Molecules Recent research in the vasculitides has focused on the chemical processes involved in leukocyte trafficking.349,613 Knowledge of these mechanisms will become increasingly important for the treatment of vasculitic neuropathy as pharmacologic agents are developed that modify this intricate process. Central to the process of vascular inflammation is a dramatic upregulation in both endothelial cells and leukocytes of the surface expression of molecules that support intercellular adhesion. Before this can develop, however, a primary immunologic event must occur that generates signals to “prime” circulating leukocytes and stimulate the involved endothelial cells.122 There are three large families of cellular adhesion molecules: the selectins, the integrins, and the immunoglobulin superfamily.350,610 Selectins are transmembrane glycoproteins that recognize a sialylated, fucosylated, complex carbohydrate determinant on their ligands.384,423,499 L-selectin is constitutively expressed on nearly all leukocytes and functions primarily

2342

Diseases of the Peripheral Nervous System

as a homing molecule. P-selectin is stored preformed in the alpha granules of platelets and Weibel-Palade bodies of endothelial cells. When platelets and endothelial cells are activated (e.g., by thrombin or histamine), P-selectin is mobilized to the plasma membrane, where it binds to neutrophils and monocytes. Similar to P-selectin, E-selectin is not expressed in resting endothelial cells. It is upregulated within 4 to 6 hours after stimulation of endothelial cells by tumor necrosis factor (TNF)-, interleukin (IL)-1, or interferon (INF)- and then binds to ligands on most leukocytes.610 Once ligated, the endothelial cell produces IL-8, a leukocyte chemoattractant that inhibits further E-selectin interactions, enabling the bound cell to move onto the next stage of adhesion. The immunoglobulin superfamily contains a large group of molecules central to the immune response, including adhesion receptors, co-stimulatory molecules, and members directly involved in antigen recognition, such as the major histocompatibility complex (MHC) class I and II molecules, T- and B-cell antigen receptors, CD4, and CD8.350 Intercellular adhesion molecule (ICAM)–1 and ICAM-2 are products of homologous genes that interact with 2-integrins, especially lymphocyte function–associated antigen-1 (LFA-1) and Mac-1, adhesion molecules found on leukocytes (see below).311,423,610 ICAM-1 has a wide tissue distribution. Resting endothelial cells express low levels of ICAM-1, but expression is upregulated by proinflammatory cytokines (TNF-, IL-1, and IFN-). ICAM-2 is constitutively expressed on endothelial cells at a higher level than ICAM-1, although its expression is not modified by inflammatory mediators. Thus ICAM-2 is the main receptor for LFA-1 on resting endothelium, and ICAM-1 assumes this role in inflamed tissues. Vascular cell adhesion molecule-1 (VCAM-1) is another immunoglobulin superfamily member whose endothelial expression is upregulated by inflammatory cytokines (IL-1, TNF-, and IL-4). It binds to 41 integrin (very late antigen-4 [VLA-4]) and 47 integrin, thereby supporting endothelial adhesion of monocytes, lymphocytes, and eosinophils.87 Integrins are heterodimeric transmembrane proteins composed of noncovalently bound  and  subunits.50,318 Every cell expresses multiple integrins, which mediate high-strength cell-cell and cell-matrix binding. The 2 and 4 integrins are important in leukocyte–endothelial cell interactions.350,610 LFA-1 is a 2 integrin that is expressed on all leukocytes and ligates both ICAM-1 and ICAM-2.673 Mac-1 is found on the cell surfaces of granulocytes and monocytes, with affinity for ICAM-1 but not ICAM-2. The 4 integrins include 41 (VLA-4) and 47.532,673 They are found on the surfaces of all leukocytes except mature neutrophils. VLA-4 binds preferably to VCAM-1 and fibronectin; 47 binds to the same ligands and to MadCAM-1. Integrin adhesiveness is regulated by chemokines, other inflammatory mediators, and the integrin-ligand interaction itself.617 Inflammatory cytokines

and chemokines also facilitate leukocyte adhesion by upregulating integrin expression, mobilizing integrin from cellular stores, and clustering integrin on the membrane.304

Chemokines Among the stimuli that activate leukocyte integrins, chemoattractant molecules are most important. Classic leukocyte chemoattractants are N-formyl peptides derived from bacterial processing, platelet activating factor (PAF), leukotriene B4 (LTB4), and complement factor C5a.12 More recently, a superfamily of chemoattractant cytokines, termed chemokines, has been elucidated.32,64,393,489 Chemokines are small polypeptide molecules produced by leukocytes, endothelial cells, fibroblasts, and other cell types. Two major (CXC, CC) and two minor (C, CX3C) subfamilies have been distinguished. One subset of the CXC chemokines acts predominantly on neutrophils, subserving an “inflammatory” function. A second subgroup—known as the “immune chemokines”—is highly specific for lymphocytes and dendritic cells and functions in the genesis, maintenance, and operation of the immune system. CC chemokines act preferentially on monocytes. T cells are influenced by both CXC and CC chemokines. Chemokines function by activating specific G protein– coupled transmembrane receptors. Most receptors recognize more than one chemokine, and several chemokines bind to more than one receptor. Chemokine-receptor binding transduces a signal that directs migration, promotes integrin adhesiveness, and stimulates degranulation and respiratory burst in phagocytic cells.

Leukocyte Emigration Leukocyte emigration from the vasculature at sites of inflammation is regulated by at least four sequential steps (Fig. 105–2).350,423,456,610 Endothelial cells and leukocytes are first activated or primed by inflammatory stimuli such as infectious agents and inflammatory cytokines (e.g., TNF-, IL-1, INF-, and IL-8) to increase surface expression of adhesion molecules (e.g., E- and P-selectins, carbohydrate selectin ligands, ICAM-1, VCAM-1, LFA-1, and VLA-4). The initial step of leukocyte–endothelial cell adhesion involves leukocyte margination and “rolling” along the endothelial surface, a low-affinity interaction mediated by the selectins, VLA-4, and their ligands. The second step requires the tethered leukocyte to become activated, a process mediated by chemokines and other chemoattractants.336 These mediators increase integrin adhesiveness by augmented receptor affinity and mobilization of integrin from cellular stores. This ushers in the third stage, in which the rolling leukocyte arrests, spreads, and firmly adheres to the endothelium, mediated by interactions between integrins and immunoglobulin

2343

Neuropathies with Systemic Vasculitis

FIGURE 105–2 Leukocyte emigration.

superfamily molecules (e.g., VLA-4/VCAM-1, LFA-1/ ICAM-1, and LFA-1/ICAM-2).385 During transendothelial migration—the final step—the leukocyte squeezes in an ameboid fashion between tightly apposed endothelial cells. This interaction involves homotypic interactions between two molecules found both on leukocytes and endothelial intercellular junctions: platelet/endothelial cell adhesion molecule (PECAM)-1, a member of the immunoglobulin gene superfamily, and CD99.456

Adhesion Molecules in Vasculitis Soluble forms of adhesion molecules are produced by proteolytic cleavage in situ or by alternatively spliced transcripts missing the transmembrane region.106 Assays for these circulating molecules in patients with vasculitides (including RV, WG, PAN, MPA, TA, and Behçet’s disease) have revealed consistently elevated levels of soluble ICAM-1 (sICAM-1) and sVCAM-1 and variable elevations of sE-selectin.24,30,106,115,600 Concentrations of these adhesion molecules generally correlate with disease activity and extent, but levels cannot be used to guide therapeutic decisions, because circulating adhesion molecules also increase with concurrent diseases (e.g., infections).106,115 In the only series devoted to vasculitic neuropathy, there was

a significant elevation of sE-selectin in patients with systemic and nonsystemic vasculitis of the nerve compared to normal and neurologic disease controls.282 An in situ study of muscle and nerve biopsies in patients with c-PAN revealed a dynamic pattern of adhesion molecule expression that varied according to the histopathologic state of the vascular lesion. Expression of constitutive adhesion molecules (ICAM-1, ICAM-2, PECAM-1, and P-selectin) was preserved and similar to controls in early PAN lesions, but tended to disappear in advanced lesions. E-selectin was expressed in the endothelium of early lesions, but not in advanced lesions or control specimens, suggesting a role for this molecule only in the initial stages of the inflammatory process. VCAM-1 was upregulated in endothelial lesions of various ages. A high percentage of infiltrating leukocytes expressed LFA-1 and VLA-4 integrins, but not L-selectin.121

Cytokines in Vasculitis Cytokines are a diverse group of soluble polypeptide molecules produced by multiple cell types that function primarily as growth factors and hormones of the immune and hematopoietic systems.64,133,393 The term encompasses the interleukins, chemokines, interferons, and

2344

Diseases of the Peripheral Nervous System

colony-stimulating factors. Many cytokines regulate immune and inflammatory responses. Cytokines act on many different cell types and induce many different effects on the same target cell, effects that may change under the influence of other cytokines. One cytokine frequently influences the synthesis of another, producing cytokine cascades. Major proinflammatory cytokines are IL-1, IL-8, INF-, and TNF-. Primary anti-inflammatory cytokines are IL-4, IL-10, and IL-13. Subsets of CD4 T helper (Th) lymphocytes can be distinguished by their cytokine profiles.64,133 T cells of Th1 type preferentially secrete INF-, IL-2, and IL-12, while Th2-type T cells secrete IL-4, IL-5, IL-10, and IL-13; these cytokine phenotypes reflect cell function. For example, IL-4, IL-10, and IL-13 from Th2 cells promote humoral and allergic responses by stimulating B-cell proliferation and inhibiting cytokine production by monocytes and INF- production by Th1 cells. Cytokines released by Th1 cells, in contrast, activate macrophages and promote T-cell proliferation, mediating cellular immune responses. Given the complexity of these interrelationships, specific patterns of cytokine release and function pertinent to vasculitides have been difficult to analyze.133,607 Most investigations to date have focused on TA and WG. In TA, elevated serum levels of sIL-6 are a consistent finding and correlate with disease activity.292 Peripheral blood monocytes predominantly express IL-6 and IL-1.246 A series of elegant studies analyzing cytokine expression and messenger RNA (mRNA) production in temporal artery specimens from patients with TA have shown upregulation of cytokines with a Th1 bias, notably INF-, produced (along with IL-2) by clonally expanded adventitial T cells.246,658 Adventitial macrophages secrete transforming growth factor (TGF)-1, IL-1, and IL-6, which support antigen presentation, T-cell activation, and a strong systemic inflammatory response. INF-, in turn, activates macrophages that migrate to the media and occasionally form giant cells. Macrophages, giant cells, and smooth muscle cells in the media and intima produce metalloproteinases, vascular endothelial growth factor (VEGF), and platelet-derived growth factor (PDGF). PDGF promotes myointimal cell proliferation, secondary vaso-occlusive disease, and ischemic symptoms. VEGF is a potent angiogenic factor, producing neovascularization of the adventitia and inflamed regions of the artery. Neovessels are a mixed blessing, reducing ischemic complications but providing a nidus for leukocytes to invade the vessel wall and a source of additional cytokines, chemokines, and growth factors, perpetuating the inflammatory process.104 The situation for other vasculitides is less well characterized. In WG and MPA, upregulated cytokines reflect a Th1 response bias, similar to TA.134,395,455 In CSS, there appears to be a predominantly Th2 cytokine response, because circulating levels of IL-10 are disproportionately increased and T-cell lines derived from peripheral blood

cells produce large amounts of IL-4 and IL-13.354,568 Three small studies of cytokine production in sural nerve biopsies in patients with vasculitic neuropathy revealed upregulated nerve growth factor (NGF), glial cell line–derived neurotrophic factor, and IL-6 mRNA; downregulated ciliary neurotrophic factor mRNA (synthesized in Schwann cells); and increased expression of IL-1, TNF-, IL-4, NGF, IL-3, IL-6, and IL-10.157,388,666 Another study comparing peripheral blood T-cell expression of INF- to IL-4 in patients with “mononeuritis multiplex” from various causes (including vasculitis) demonstrated an increased proportion of IL-4/ INF- T cells, reflective of a Th2 shift.307 These data suggest that humoral and cellmediated immune responses are both active in patients with vasculitic neuropathy, consistent with neuropathologic findings showing (1) immune deposits in epineurial vessel walls and (2) infiltration of nerve by cytotoxic T cells and macrophages in many patients with vasculitic neuropathy (see below).

Humoral and Cellular Immunity in Vasculitis Within this basic framework of antigen-driven adaptive immunity progressing to an effector phase marked by vascular inflammation and destruction, five predominant immunopathogenic mechanisms have been proposed, three involving humoral or antibody-mediated processes and two involving T-cell–mediated responses. Circulating Immune Complex–Mediated Vasculitis Of the humoral schemes, passive deposition of circulating immune complexes was the first to undergo extensive study (Fig. 105–3).112 In this model, antibodies interact with circulating antigens to form complexes of varying size, which activate complement, become coated with C3b and C4b, and prime circulating leukocytes.35,355,402,590 Under normal circumstances, circulating immune complexes bind to erythrocytes, mediated by an interaction between C3b and erythrocyte complement receptor CR1, and are then transported to the liver for clearance. Pathogenic deposition in vessel walls is apt to occur if the host’s clearance mechanisms are flawed or overcome, immune complexes acquire certain physical properties (intermediate-sized and cationic), blood flow is turbulent, or vessel walls are abnormally permeable. In situ immune complex deposits activate complement, generating chemoattractants such as C5a (which recruit neutrophils) and other inflammatory mediators. Further complement activation leads to assembly of the C5b-9 membrane attack complex that mediates endothelial cell lysis. Deposited immune complexes also become opsonized with C3b fragments. Neutrophils bind to the complexes via (1) complement receptors such as CR1 interacting with C3b, or CR3 and

Neuropathies with Systemic Vasculitis

2345

FIGURE 105–3 Circulating immune complex– mediated vasculitis.

CR4 interacting with C3dg and iC3b; and (2) Fc receptors (FcR) interacting with Fc portions of the complexed immunoglobulin G (IgG). Activated neutrophils release proteolytic enzymes and reactive oxygen metabolites that damage the vessel wall. Chemokines (e.g., IL-8) and other chemoattractants recruit and activate additional neutrophils and monocytes to the nidus of inflammation. This mechanism is believed to subserve (1) most types of HSV related to infections, drugs, and malignancies; (2) hepatitis C–associated mixed cryoglobulinemic vasculitis; and (3) hepatitis B–associated PAN.154,190,266,410 In addition, many vasculitides related to underlying connective tissue diseases may be mediated by autoantibodies directed against endogenous antigens such as DNA, nucleosomes, and immunoglobulins. However, circulating immune complexes are not typically found in WG, CSS, non–hepatitis B–associated PAN, MPA, and TA.35,590

Anti–Endothelial Cell Antibody–Mediated Vasculitis A second possible mechanism of humorally mediated vasculitis involves anti–endothelial cell antibodies (AECAs), which interact with antigenic determinants on endothelial cell membranes to form in situ immune complexes that have the potential to produce damage via complementmediated or antibody-dependent cellular cytotoxicity (ADCC) and induction of endothelial cell apoptosis.43,92,342,504 ADCC is a process whereby a specific antibody binds to a target cell and engages a NK cell through its FcR, resulting in cell lysis. Target antigens could be constitutive components of the vessel wall or adherent “planted” antigens.43,92 AECAs are detected in 50% to 80% of patients with antineutrophil cytoplasmic autoantibody (ANCA)–associated vasculitides such as WG, MPA, and CSS (see next section), 30% to 50% of patients

2346

Diseases of the Peripheral Nervous System

with TA, 90% of Takayasu’s arteritis patients, and 65% to 75% of patients with RV, but there is no convincing evidence to support a causal relationship.43,342,504 They may be epiphenomena, representative of an immune response directed against endothelial antigens exposed by vascular damage from other means. Indeed, AECAs are not unique to vasculitis, occurring in many other nonvasculitic conditions (e.g., diabetes mellitus, ulcerative colitis, thrombotic thrombocytopenic purpura, multiple sclerosis, and various connective tissue diseases).43,504 In vitro studies analyzing the ability of AECAs to mediate complement-dependent or antibody-dependent endothelial cytotoxicity have produced contradictory results.72,153,556 There is more compelling evidence that AECAs induce endothelial cell activation, upregulating expression of adhesion molecules (E-selectin, ICAM-1, VCAM-1), stimulating secretion of cytokines and chemokines (IL-1, IL-6, IL-8, monocyte chemotactic protein-1 [MCP-1]), and promoting thrombosis.88,342,457 Therefore, AECAs may function in vivo to sustain immune-mediated vascular inflammation initiated by other causes. ANCA-Mediated Vasculitis ANCAs represent a third potential mechanism of humorally mediated vascular damage, relevant to three vasculitides: WG, MPA, and CSS.35,40,341 ANCAs are autoantibodies directed against cytoplasmic proteins in neutrophils and monocytes. They are classified as cytoplasmic (c-ANCA), perinuclear (p-ANCA), or atypical (a-ANCA) based on their staining pattern on alcohol-fixed neutrophils.558,660 In vasculitis, c-ANCAs are almost always targeted to proteinase 3 (PR3) and p-ANCAs to myeloperoxidase (MPO). ANCAs are found in many other, nonvasculitic conditions, but in these diseases, antigens other than PR3 or MPO are usually recognized. PR3 and MPO are constituents of azurophilic granules in neutrophils and lysosomes in monocytes. c-ANCAs directed against PR3 are strongly associated with active WG, but also occur in MPA and CSS (Table 105–5). p-ANCAs targeting MPO are primarily seen in MPA, CSS, and renal-limited vasculitis.35,342,660

Multiple pathogenic mechanisms for ANCAs in vasculitis have been proposed (Fig. 105–4).35,40,342,506,537 The first concerns interactions of ANCAs with primed neutrophils and monocytes. Priming refers to preactivation of leukocytes by proinflammatory cytokines such as TNF-, IL-1, and IL-8, whereby small amounts of granule and secretory vesicle contents are transferred to the cell surface, permitting expression of PR3 and MPO, which are then accessible for interactions with ANCAs.122,280 When primed neutrophils are incubated in the presence of ANCAs, activation occurs, dependent on an interaction between ANCAs and FcRs on neutrophils. Activated neutrophils increase expression of adhesion molecules (especially 2 integrins) and adhere to the endothelium, degranulate, undergo respiratory burst, and lyse endothelial cells through release of reactive oxygen species, nitric oxide, lysosomal proteolytic enzymes, and inflammatory mediators such as TNF-, IL-1, IL-8, and LTB4.342,,513,514,537 ANCAs exert similar effects on primed monocytes.342,537 ANCAs can also directly activate endothelial cells.457,537,611 One effect of this activation is to upregulate expression of adhesion molecules, including E-selectin and VCAM-1, thereby rendering the endothelial membrane more adhesive for inflammatory cells. In addition, ANCAs cause endothelial cells to synthesize and release such inflammatory mediators as IL-1, IL-6, prostaglandin I2, and PAF. Furthermore, ANCAs increase endothelial permeability and promote endothelial expression of tissue factor. PR3 and MPO molecules released by ANCA-activated neutrophils and monocytes may also bind to the endothelial surface and serve as planted antigens, forming in situ immune complexes with ANCAs that attract neutrophils (opsonization), activate complement, and damage endothelial cells by ADCC.37,75,506,611 However, the significance of this mechanism to human ANCAassociated vasculitides is disputed because vascular deposits of immunoglobulins and complement are scant in these conditions.75 Incubation of endothelial cells with PR3 also induces chemokine production (IL-8, MCP-1), increased expression of ICAM-1, tissue factor release, and endothelial cell apoptosis and lysis.506,611

Table 105–5. Antineutrophil Cytoplasmic Autoantibodies (ANCAs) Associated with Systemic Vasculitis Target antigen Sensitivity: Associated vasculitides and sensitivities

Specificity for vasculitis (disease controls)

c-ANCA

p-ANCA

Proteinase 3 (PR3)

Myeloperoxidase (MPO)

Wegener’s granulomatosis 75%–80% (active/generalized) 40%–50% (inactive/generalized) 40%–50% (limited) Microscopic polyangiitis: 25%–35% Churg-Strauss syndrome: 25%–30% ~99%

Microscopic polyangiitis: 50%–60% Churg-Strauss syndrome: 25%–30% Wegener’s granulomatosis: 10%–15%

~99%

Neuropathies with Systemic Vasculitis

2347

FIGURE 105–4 ANCA-mediated vasculitis.

ANCAs might also affect the enzymatic properties of their antigens. 1-Antitrypsin (1-AT) is the primary physiologic inhibitor of PR3.281 Binding of anti-PR3 ANCA to PR3 may inhibit the irreversible inhibition of PR3 by 1-AT, thereby increasing the persistence of complexed PR3 in the involved tissues.342 Subsequent dissociation of PR3 from PR3-ANCA immune complexes could contribute to prolonged tissue-damaging capability. In addition, mechanisms involving apoptotic neutrophils have been adduced. PR3 and MPO are translocated to the surface of apoptotic neutrophils, permitting interactions with their respective ANCAs.175,280 Although ANCAs do not activate apoptotic neutrophils, they serve as opsonins, enhancing the recognition and phagocytosis of these cells by scavenger macrophages via FcR interactions. Whereas phagocytosis of apoptotic cells

usually occurs in an immunosuppressive fashion, phagocytosis of ANCA-opsonized apoptotic neutrophils triggers production of proinflammatory molecules such as IL-1, IL-8, TNF-, and thromboxane B2, amplifying and perpetuating the inflammatory response.280,448 At the opposite end of the immune response cascade, defective clearance and increased exposure to apoptotic neutrophils in an inflammatory environment may be the initiating mechanism for ANCA production and the subsequent development of vasculitis.175,518 A recent study offered the first convincing in vivo evidence that anti-MPO ANCAs are sufficient to cause a systemic, pauci-immune, small vessel vasculitis in a murine model, with pathologic changes identical to the alveolar capillaritis and necrotizing arteritis observed in human ANCA-associated vasculitides.665

2348

Diseases of the Peripheral Nervous System

Cellular Immunity in Vasculitis Although antibody-mediated mechanisms appear important in many vasculitides, compelling data support a role for cellular immunity as well.110,245,252 First, the granulomas seen in WG, CSS, and TA are prima facie evidence that T-cell responses are involved in these disorders, because granuloma formation is T-cell dependent.588 Second, phenotypic analyses of pulmonary and renal cellular infiltrates in patients with WG and temporal artery infiltrates in TA show a predominance of T cells and macrophages.225,620,658 Third, cytokine repertoires of TA and WG (described above) favor a predominant cellular immune response. Fourth, immunologic characterization of ANCAs in patients with ANCA-associated vasculitides indicates that T cells are involved in these conditions.110,252 Fifth, analysis of T-cell receptor (TCR) variable gene usage in patients with WG or PAN reveals overrepresentation of a common sequence in the V8 gene, corresponding to V8 CD4 cell expansions, suggesting exposure to a common antigen that elicited a cellular immune response.254 Sixth, anti–T-cell therapy using cyclosporine or T-cell–depleting biologic agents has been successful in patients with refractory systemic vasculitis.310,391 Cytotoxic T Lymphocyte–Mediated Vasculitis In cellular cytotoxicity–mediated vasculitis, the effector cells are cytotoxic T lymphocytes (CTLs) directed against constitutive or induced endothelial antigens, exogenous antigens processed by endothelial cells, or planted antigens (Fig. 105–5).245,536 Activated (mainly CD8) CTLs destroy antigen-bearing cells by at least two methods. In the granule exocytosis pathway, perforin monomers, granzymes, cytotoxic cytokines, and T-cell–restricted intracellular antigen-1 (TIA-1) are secreted in the direction of the target cell.20,601,627 After insertion into target cell membranes, perforin polymerizes into pore-forming aggregates that permit calcium influx and promote osmotic lysis of the target cell. Perforin-containing pores also allow granzymes and TIA-1 to enter the cell. Granzymes are proteolytic enzymes and TIA-1 a nuclear RNA–binding protein. In concert, they induce DNA fragmentation and apoptotic cell death. The second, nonsecretory pathway requires an interaction between upregulated Fas ligand on CTLs and Fas receptor on the target cell, resulting in delivery of a separate apoptotic signal.648 This mechanism is considered important in vasculitic neuropathies because CD4 and CD8 T cells and macrophages are the predominant cell types in epineurial infiltrates in sural nerve biopsies from such patients, whereas B cells, NK cells, and polymorphonuclear leukocytes are rarely detected.352,359,554 CTLs are activated after MHC-restricted binding to antigen-presenting cells.536 Consistent with such a process, patients with vasculitic neuropathy have upregulated dendritic cells around involved epineurial vessels.352 In addition, many of the

epineurial T cells have upregulated expression of perforin, granzyme A, granzyme B, TIA-1, Fas, and Fas ligand compared to nonvasculitic control nerves.294,554 However, arguing against a cellular cytotoxicity mechanism, one investigation of TCR V gene utilization in sural nerve biopsies of patients with vasculitic neuropathy revealed no clonally expanded T cells, suggesting that T cells are nonspecifically recruited to the PNS and not antigenically selected,65 and another study showed that apoptosis was restricted to inflammatory mononuclear cells, implying that apoptosis fulfills a regulatory role rather than mediating vascular injury.294 Nevertheless, cellular cytotoxicity might still occur in vasculitic neuropathy, mediated by perforininduced cytolysis, macrophage-derived toxic oxygen free radicals, and cytotoxic cytokines such as TNF-.487 Delayed-Type Hypersensitivity in Vasculitis Delayed, cell-mediated hypersensitivity vasculitis is relevant to the pathogenesis of TA, in which vasculitic lesions contain adventitial CD4 T cells that have undergone clonal expansion in response to an unknown antigen.658 These activated T cells produce INF-, which, in turn, induces macrophage activation and differentiation, with multinucleated giant cell formation and the emergence of granulomas near the internal elastic lamina.246 Macrophages in the media and intima produce matrix metalloproteinases, cytokines, growth factors, and reactive oxygen species that damage the vessel wall and promote myointimal proliferation and angiogenesis. In other granulomatous vasculitides, such as WG and CSS, similar mechanisms may be operative.590

ETIOLOGY Most vasculitic neuropathies are idiopathic, with a presumed autoimmune pathogenesis, whether occurring in isolation, as part of a primary multiorgan-involving vasculitis, or secondary to a connective tissue disease. Uncommonly, specific triggers are identified (e.g., infections, drugs, or malignancies). The immunopathogenic mechanisms in these cases serve as models for the idiopathic vasculitides.

Infections Virtually any organism can act as an antigenic stimulus leading to the formation of circulating or in situ immune complexes, which then mediate vasculitis.439,589,595 Microorganisms can also produce vasculitis by direct vascular invasion (e.g., rickettsia and herpesviruses), release of toxins acting as superantigens (e.g., staphylococci and streptococci), generation of vascular toxins (e.g., Pseudomonas), induction of cell-mediated or antibody-mediated immune responses against vascular

Neuropathies with Systemic Vasculitis

2349

FIGURE 105–5 Cytotoxic T lymphocyte–mediated vasculitis.

structures (e.g., ascariasis), and aberrant immunoregulation (e.g., human immunodeficiency virus [HIV]). While a variety of infections produce central nervous system (CNS) vasculitis by direct vascular invasion, invasion of peripheral vasa nervorum is not well documented.589 Most bacterial septicemias with vasculitis yield cutaneous, not neuropathic, manifestations. The only microbe-related vasculitides to involve peripheral nerves with direct cytopathic infection of endothelial cells are those caused by cytomegalovirus (CMV) and herpes zoster.100,542 Viruses with persistent replication, such as hepatitis B virus

(HBV), hepatitis C virus (HCV), HIV, and human Tlymphotrophic virus-1 (HTLV-1), have the strongest association with vasculitis in general and vasculitic neuropathies in particular. HBV-Associated Vasculitis A well-characterized infection-related vasculitis is HBVassociated PAN, in which immune complex deposition appears to be the primary disease mechanism.258,266,425,578,637 Depending on the population, circulating hepatitis B surface antigen is found in 10% to 55% of patients with

2350

Diseases of the Peripheral Nervous System

c-PAN. The clinical manifestations in these patients are similar to those in idiopathic PAN. In the largest reported series, more than 80% of patients had neuropathy.258,266 Vasculitis usually develops in the early stages of the hepatitis B infection and follows a monophasic, selflimited (albeit sometimes fatal) course. The median delay between the acute viral exposure and PAN is 2 months. Relapses are rare. Chronic hepatitis C,86 HIV,194 and parvovirus B19191 infections are less common causes of a PAN-like illness. HIV-Associated Vasculitis Approximately 2% of patients with HIV infection develop vasculitis, often with predominant peripheral nerve and muscle involvement.70,71,98,194,205,226,232,399,637,672 This figure likely underestimates the true incidence, however, because HIV-positive patients with distal, painful, sensorypredominant polyneuropathies can have pathologically proven vasculitis, and this group is not routinely biopsied, preempting ascertainment of some cases.70,672 Vasculitic neuropathies in HIV-infected individuals commonly develop during the early symptomatic stage, but can occur in patients with CD4 counts as low as 14/mL.194 Various neuropathy profiles are associated, including mononeuropathy, multiple mononeuropathy, cranial neuropathy, distal symmetrical polyneuropathy, and asymmetrical polyneuropathy. Unlike systemic necrotizing vasculitides, most cases of HIV vasculitis are self-limited. Constitutional symptoms such as fever, weight loss, myalgias, and arthralgias occur in one third of patients, with less common cutaneous, renal, CNS, gastrointestinal (GI), and cardiac involvement.194 Although ANCAs are positive in 20% to 40% of HIV-infected patients, they are not associated with vasculitis in this illness.129,559 Nerve and muscle biopsies in patients with HIV-associated vasculitic neuropathy reveal necrotizing vasculitis or non-necrotizing “microvasculitis” (perivascular and intramural inflammatory cell infiltration of vessels less than 70 m in diameter without fibrinoid necrosis).70,71,226,232,399,672 Although epineurial vessels are commonly affected, there is a characteristic and sometimes more significant involvement of the endoneurial and endomysial microvasculature, a finding atypical for PNS vasculitis. Both mononuclear and polymorphonuclear infiltrates are seen in the endoneurium. Epineurial infiltrates are composed primarily of CD8 T cells. Perivascular iron deposits in the endoneurium and endomysium and deposits of complement and immunoglobulin in small vessel walls are commonly observed.226,229 The pathogenesis of HIV-related vasculitis is not well understood. Speculated mechanisms include direct infection of endothelial cells by HIV or other organisms, immune complex deposition or in situ formation, release of proinflammatory cytokines (e.g., TNF-), and HIVmediated immune dysregulation.70,71,98,226,637 Circulating

immune complexes are often detected in HIV-positive patients, and deposition of these complexes in endoneurial and epineurial vessels is the best accepted explanation for HIV vasculitis. Although there is no evidence that HIV directly propagates in the vasa nervorum, HIV has been cultured from peripheral nerve homogenates, HIV antigens and genomic DNA have been detected in nerve and muscle perivascular cells, and HIV replication has been observed in endoneurial and epineurial inflammatory cell infiltrates in patients with PNS vasculitis.226,227,298 Tissue macrophages are believed to constitute the local source of replicant HIV. Other causes of an asymmetrical or multifocal neuropathy in HIV-infected patients include CMV infection of the PNS,535,542 neurolymphomatosis (PNS infiltration by lymphoma cells),236 and diffuse infiltrative lymphocytosis syndrome.454 CMV infection should be suspected in patients with CD4 counts less than 50/mL, CMV involvement of other organs such as the retina, and an asymmetrical, sensorimotor polyneuropathy that has evolved over weeks to months. CMV directly infects endothelial cells of endoneurial capillaries. Nerve biopsies show multifocal, necrotic, vasculitic endoneurial lesions with neutrophilrich inflammatory infiltrates. CMV is often cultured from blood, but not cerebrospinal fluid (CSF). Detection of CMV DNA in CSF by polymerase chain reaction (PCR) is the most sensitive diagnostic technique.535 HCV-Related Cryoglobulinemic Vasculitis More commonly than a PAN-like illness, HCV infection produces a mixed cryoglobulinemic vasculitis.80,163,190,208,369 Cryoglobulins are circulating immunoglobulins that precipitate when cooled to less than 37° C. They are divided into three types. Type I are monoclonal immunoglobulins or light chains, occurring in lymphoproliferative diseases. Type II consist of a mixture of monoclonal immunoglobulin M (IgM) rheumatoid factors (RFs)—having anti-IgG activity—and polyclonal IgG. Type III are polyclonal IgM RFs and polyclonal IgG. Types II and III are termed mixed cryoglobulins, which can be essential—without known cause—or secondary to a wide variety of connective tissue diseases, chronic infections, and lymphoproliferative disorders. It is now known that HCV infection underlies 80% to 90% of essential mixed cryoglobulinemia (EMC). Mixed cryoglobulins occur in approximately 40% of HCV-infected patients, but only 2% to 3% develop a cryoglobulinemic vasculitis.190 Unlike HBV-associated PAN, HCV-related cryoglobulinemic vasculitis is a chronic disease, evolving many years after the initial HCV infection. HCV chronically infects mononuclear peripheral blood and bone marrow cells, especially B lymphocytes, producing a non-neoplastic, oligoclonal B-cell expansion, resulting in the appearance of autoantibodies, RFs, and cryoglobulins.151 In symptomatic patients, mixed cryoglobulins induce immune

2351

Neuropathies with Systemic Vasculitis

complex–mediated, complement-dependent inflammation of small blood vessels, yielding a necrotizing, leukocytoclastic, or lymphocytic small vessel vasculitis.28 For patients with neuropathy, other possible mechanisms include (1) intravascular deposition of cryoglobulins causing an occlusive, endoneurial microangiopathy 507; (2) IgM binding to myelin, producing demyelination624; and (3) T-cell–mediated vascular injury.60 HCV genomic RNA is occasionally detected in nerve and muscle biopsies of patients with HCV-associated neuropathy, although no in situ replication of HCV occurs, arguing against direct infection of the nerve.28,150 The clinical syndrome is characterized by palpable purpura in 90% to 100% of patients, skin ulcers, arthralgias, Raynaud’s phenomenon, glomerulonephritis, liver function abnormalities, sicca symptoms, and adenopathy.80,163,190,208,369 RFs are found in 80% of patients and hypocomplementemia in 50% to 60%. About 10% develop an indolent B-cell lymphoma. The incidence of polyneuropathy in EMC is approximately 40% by clinical criteria190,209,223,316,634,675 and approximately 75% by electrodiagnostic testing.93,107,188,634 Most neuropathies are distal, sensory predominant, and symmetrical or mildly asymmetrical, but multifocal neuropathies also occur. One provocative study found cryoglobulins in 13% of idiopathic neuropathy patients, suggesting that all patients with unexplained neuropathies should be tested for cryoglobulins.222 Bacteria-Related Vasculitis Bacterial infection–related vasculitis only rarely involves the PNS.589,595 Exceptional cases have been reported as parainfectious phenomena with group A -hemolytic streptococcal,628 Mycoplasma pneumoniae,353 Salmonella typhi,492 leptospiral,277 and Pseudomonas aeruginosa616 infections. Vasculitic neuropathies may also complicate infective endocarditis and chronic dental infections.82,494 Lyme disease—a multisystem, tick-borne disease caused by Borrelia burgdorferi—is more commonly (35% to 65% of patients) associated with a wide variety of PNS manifestations, including cranial neuropathies, mono- or polyradiculopathies, brachial and lumbosacral plexopathies, multifocal neuropathies, distal symmetrical polyneuropathies, carpal tunnel syndrome, and GuillainBarré syndrome (GBS).274,357,428,491 Focal and multifocal neuropathies affecting cranial, spinal, and peripheral nerves predominate during the stage of early disseminated disease. Peripheral nerve involvement in later stages is more commonly diffuse and symmetrical. Sural nerve biopsies reveal mononuclear, epineurial, perivascular inflammatory infiltrates, perineurial thickening and neovascularization, and deposition of complement membrane attack complex in vessel walls, but necrotizing vasculitis is uncommon.563 Borrelia burgdorferi DNA has been detected in sural nerve tissue by PCR.400

Drugs A small vessel vasculitis has been associated with most of the commonly used pharmaceutical agents at one time or another.154,432,621 As many as 25% to 45% of cases of HSV may be due to drugs.213,410 Sulfonamides were the first agents recognized to cause vasculitis, but many other drugs have now been implicated. The most commonly reported drugs are propylthiouracil, hydralazine, colonystimulating factors, allopurinol, cefaclor, minocycline, D-penicillamine, phenytoin, isotretinoin, and methotrexate (MTX).621 Newer classes of drugs occasionally associated with vasculitis include leukotriene inhibitors, interferons, TNF- inhibitors, and quinolone antibiotics. Immune complex deposition is the suspected mechanism in most cases, with the drug forming a hapten-carrier conjugate and then triggering an immune response.248 However, all types of hypersensitivity reactions can be produced. Some drugs precipitate an anti-MPO ANCA-associated vasculitis, with clinical manifestations similar to those in the idiopathic ANCA-associated vasculitides.432,621 Most drug-induced vasculitides spare the PNS, but multifocal neuropathies or pathologically proven vasculitic neuropathies have been reported with penicillin,569 cromolyn,533 thiouracil,137 allopurinol,47 amantadine,580 INF-,500 naproxen,564 and montelukast.453 Illegal drugs can also cause vasculitis, especially amphetamines, cocaine, and heroin.442 PNS involvement occurs with these agents but is less common than CNS vasculitis.598

Malignancy Hypersensitivity vasculitis in the setting of lymphoproliferative disorders is well established.197,664 The tumor most frequently associated with vasculitis is hairy cell leukemia.283 Most malignancy-related vasculitides are restricted to the skin and demonstrate leukocytoclastic pathology. However, vasculitis with hairy cell leukemia occasionally assumes systemic necrotizing features, similar to c-PAN, albeit without reported PNS involvement.283 Waldenström’s macroglobulinemia, non-Hodgkin’s lymphomas, chronic lymphocytic leukemia, angioimmunoblastic lymphadenopathy, and multiple myeloma can also produce a mixed cryoglobulinemic vasculitis, sometimes involving the PNS.163,208,369,507,664 Cancer-related peripheral nerve vasculitis has been reported more frequently (two thirds of cases) with solid malignancies than hematologic disorders.22,26,55,228,275,287,312,334,366,381,414,484,485,618,626,677 Most patients present in the sixth to eighth decade (range 30 to 81 years). Men are twice as likely to be affected as women. In 50% of patients, neuropathy symptoms precede the diagnosis of malignancy. Most patients (80%) develop an asymmetrical or multifocal sensorimotor neuropathy; 20% have a distal symmetrical polyneuropathy. By clinical or

2352

Diseases of the Peripheral Nervous System

postmortem examination, most paraneoplastic vasculitides involving the PNS are confined to nerves and muscles, distinguishing them from the primary systemic vasculitides. Pathologically, 50% of nerve and muscle biopsies exhibit classic necrotizing vasculitis of small to medium-sized epineurial vessels, but the other 50% show “microvasculitis” or leukocytoclastic vasculitis of the endoneurial and endomysial microvasculature, similar to HIV-associated vasculitis. Mechanisms by which cancers produce vasculitis are unknown, but many theories have been proposed.197,366,664 Tumor antigens may elicit humoral immune responses, inducing in situ or circulating immune complex formation. Clearance of immune complexes by the reticular activating system may be disturbed by the tumor. Molecular mimicry between antigens on neoplastic and endothelial cells may trigger cross-reacting immune responses. Malignant cells may directly invade or embolize to vessel walls. Products released by cancer cells may deposit in vessel walls, causing inflammatory reactions. The cancer and the vasculitis may share a common viral etiology.

SYSTEMIC NECROTIZING VASCULITIS The systemic necrotizing vasculitides are a diverse group of idiopathic, potentially life-threatening, inflammatory diseases characterized by simultaneous or sequential vasculitic involvement of multiple organ systems. This group includes c-PAN, MPA, CSS, WG, and vasculitides secondary to underlying connective tissue diseases. Each disorder has distinguishing anatomic, pathologic, laboratory, and clinical features, which are summarized in Table 105–6. Most investigators now subdivide the systemic vasculitides by their association or nonassociation with ANCAs. The ANCA-associated vasculitides are WG, MPA, and CSS.83 These disorders share a predilection for small vessel involvement (arterioles, capillaries, and venules), paucity of immune deposits in vascular lesions, and potential for rapidly progressive glomerulonephritis and pulmonary capillaritis. ANCAs are found in 90% of patients with active, generalized WG; this figure drops to 40% when the disease is inactive or limited in distribution (see Table 105–5).35,83,660 ANCAs are also present in 75% to 85% of patients with MPA and 50% to 60% of CSS patients. Anti-PR3 c-ANCAs are strongly associated with active WG. Anti-MPO p-ANCAs predominate in MPA, CSS, and renal-limited vasculitis. Although p-ANCAs and, to a lesser extent, c-ANCAs have been described in many other conditions, PR3 and MPO ANCAs are both specific for vasculitis. For example, a multicenter study of 358 patients with vasculitis and 184 disease controls showed that anti-PR3 ANCAs were 87% and anti-MPO ANCAs 91% specific for vasculitis.270 When the results of immunofluorescent assay

(c-ANCA or p-ANCA) were combined with enzyme-linked immunosorbent assay (PR3 ANCA), specificity was enhanced. Combined anti-PR3 and combined p-ANCA/anti-MPO were both 99% specific for vasculitis.

those of or MPO c-ANCA/ positivity

Classic Polyarteritis Nodosa PAN has traditionally represented the most common systemic necrotizing vasculitis and the most frequent cause of vasculitic neuropathy. However, as previously discussed, its incidence has been severely curtailed by the CHCC nomenclature, which mandates that any microvessel involvement exclude the diagnosis, making MPA the much more prevalent disease. Dissenters have argued that cPAN does occasionally affect small vessels in the skin, peripheral nerves, and other tissues.255,263,386 They have proposed an alternate definition, wherein c-PAN is a systemic vasculitis with predominant, not exclusive, involvement of small and medium-sized arteries, characterized by renal artery vasculitis, renal infarcts, and angiographically demonstrated visceral microaneurysms, without glomerulonephritis, pulmonary vasculitis, cryoglobulins, or ANCAs. Classic PAN occurs primarily between the ages of 40 and 70 years.114,203,255,379,387,480,602 Unlike CSS, there is no allergic diathesis. The onset is typically subacute, with the gradual emergence of constitutional symptoms over weeks to months. Fevers occur in approximately 70% and weight loss in two thirds of patients. Symptoms referable to specific organ involvement soon follow. The most commonly affected tissues are peripheral nerves, joints, kidneys, skin, muscles, GI tract, heart, and testes (see Table 105–6). Lungs and spleen are rarely involved. Although skin vasculitis is classically observed in 50% of PAN patients, manifestations such as palpable purpura secondary to microvascular venulitis are now attributed to MPA rather than c-PAN. There are no large post-CHCC series of c-PAN patients because of the rarity of the CHCC-defined disease. However, based on a review of over 1000 patients in 13 old series, neuropathies occur in 60% of patients. Cranial neuropathies are uncommon ( 5% of patients). PAN is now believed to be a self-limited disease that rarely recurs once remission is induced; in HBV-related PAN, the relapse rate is only 6%.266 Deaths are most commonly due to vasculitis-induced GI bleeding and bowel perforation. Laboratory studies reflect the presence of a systemic inflammatory response, including a highly elevated erythrocyte sedimentation rate (ESR) in more than 85% of patients. Most patients have leukocytosis, normocytic anemia, and thrombocytosis. Antinuclear antibodies (ANAs), RFs, and hypocomplementemia occur uncommonly. ANCAs are typically absent in both idiopathic and HBV-associated PAN. Abdominal angiography shows

2353

Neuropathies with Systemic Vasculitis

Table 105–6. Clinical and Pathologic Features of Vasculitides Associated with Neuropathy

Disorder Group 1 Classic polyarteritis nodosa

Histopathology

Vessels Affected

Other Clinically Involved Organs (%)

Necrotizing vasculitis; mixed infiltrate

Small & medium arteries

Joints: ~50 Kidney: 45–50 Skin: ~45 GI: ~30 Testes: 2–29 Heart: ~10 ENT: rare Lungs: rare

Microscopic polyangiitis

Necrotizing vasculitis; mixed infiltrate*

Arterioles, capillaries, venules, veins  small & medium arteries

Churg-Strauss syndrome

Necrotizing vasculitis; eosinophils; granulomas*

Small & medium arteries, arterioles, capillaries, venules, veins

Kidney: ~75 Lungs: 35–40 Skin: 50–60 GI: ~40 Joints: ~40 ENT: ~25 Heart: 15–20 Testes: rare Lungs: 100 (asthma), ~60 (infiltrates) ENT (sinusitis, rhinitis): ~60 Skin: ~60 Joints: 40–50 GI: 40–50 Heart: ~40 Kidney: ~30

Wegener’s granulomatosis

Necrotizing vasculitis; palisading granulomas; collagen necrosis*

Small & medium arteries, arterioles, capillaries, venules, veins

Rheumatoid vasculitis (~20% of RA patients)

Necrotizing vasculitis; small vessel intimal hyperplasia*

All sizes & types

ENT: 90–95 Lungs: 70–80 Kidney: ~75 Eyes: 50–60 Skin: ~40 Joints: 70–75 Heart: ~15 GI: ~10 Skin: ~70 Muscle: ~55 Heart: 35–40 Lung: ~25 Kidney: ~25 Serositis: 20–25 GI: ~10

Laboratory Studies (%) qESR: ~85 qWBC: ~70 pHgb: ~60 qPlatelets: ~60 RF: ~30 Hep B: 20–30 pComp: ~25 ANA: ~15 Hep C: 5–10 ANCA: 10 Angio : 70 qESR: 90 ANCA: 75–85 (MPOPR3) RF: 25–50 ANA: 20–30 pComp: rare Hep B/C () Angio: rare qEos: 100 qESR: ~85 qIgE: ~75 ANCA: 50–60 (MPOPR3) RF: ~40 ANA: ~10 pComp: rare Hep B/C () Angio : 30 qESR: ~85 RF: 50–60 ANA: ~25 ANCA: 40–90 (PR3MPO) pHgb: ~75 qPlatelets: ~55 qWBC: ~35 RF: 90–95 qESR: ~85 qWBC: ~60 ANA: ~50 pComp: ~45 ANCA: ~40 pHgb: ~25 qEos: ~20 Angio: infreq

PNS, CNS Changes (% of patients) PNS: ~60 CNS: ~20

PNS: ~60 CNS: ~15

PNS: ~70 CNS: ~10

PNS: 15–20 CNS: ~5 CrN: ~10

PNS: 45–50 CNS: 10–15

Table continued on following page

2354

Diseases of the Peripheral Nervous System

Table 105–6. Clinical and Pathologic Features of Vasculitides Associated with Neuropathy—Continued Vessels Affected

Other Clinically Involved Organs (%)

Laboratory Studies (%)

PNS, CNS Changes (%)

Vasculitis: Skin: 80–85 GI: 1–29 Retina: ~10 Other organs: rare

SLE: ANA: 90–100 dsDNA: 60–70 pComp: 60 Antibodies: Sm: 30–40; Ro: 25–40; La: 10–45 RF: ~30 ANCA: ~15 Angio: infreq

PNS: ~13 CNS: Unknown

Arterioles, capillaries, venules

Skin: 100 Kidney: 7–37 GI: 5–20 Heart: 0–25

PNS: rare CNS: rare

LCV in skin; IgA/vascular deposits

Arterioles, capillaries, venules

Skin: 100 Joints: 60–80 GI: 65–75 Kidney: 20–50 Testes: ~20 Lungs: rare

Necrotizing vasculitis; occlusive microangiopathy; LCV in skin

Small arteries, arterioles, capillaries, venules

Skin: ~90 Joints: 70–90 Kidney: ~30 Salivary glands (sicca): ~30 Raynaud’s: 20–30 GI (pain): 10–20

qESR: ~65 qIgA: ~25 pComp: ~15 qWBC: ~15 RF: ~10 ANA: ~10 Cryos: ~10 pHgb: 5–10 ANCA () qESR: ~65 qIgA: 50–60 qWBC: 50–60 Cryos: 20–25 pComp: 5–10 pHgb: 5–10 RF: ~5 ANA: ~5 ANCA () Hep C: ~90 pComp: 80–90 RF: ~70–80 qESR: ~70 pHgb: ~70 ANA: ~55 Hep B: ~5 ANCA: 5 Angio: infreq

Granulomatous vasculitis; giant cells: ~50%

Medium & large arteries

PMR: ~50 Liver: ~30 Subclavian, axillary, brachial, vertebral, aorta (aneurysm): 10–15 Kidney: ~10

Disorder

Histopathology

Lupus vasculitis (~40% of SLE patients)

Necrotizing vasculitis*

Small & medium arteries, arterioles, capillaries, venules

LCV in skin

Henoch-Schönlein purpura

Cryoglobulinemic vasculitis

Group 2 Hypersensitivity vasculitis

Group 3 Giant cell arteritis

qESR: 95 pHgb: ~60 qWBC: ~20 ANCA: rare RF () ANA () Angio: infreq

PNS: rare CNS: rare

PNS: ~40 CNS: rare

Otologic: 60–90 Optic nerve: 15–20 PNS: ~15 CNS: ~5

Table continued on opposite page

2355

Neuropathies with Systemic Vasculitis

Table 105–6. Clinical and Pathologic Features of Vasculitides Associated with Neuropathy—Continued

Disorder Group 4 Nonsystemic vasculitic neuropathy

Miscellaneous Behçet’s disease

Histopathology

Vessels Affected

Other Clinically Involved Organs (%)

Laboratory Studies (%)

PNS, CNS Changes (%)

Necrotizing vasculitis: epineurial  endoneurial

Small arteries, arterioles

Muscle: 15

qESR: ~60 Anemia: ~30 ANA: ~30 qWBC: ~20 RF: ~15 pComp: ~10

PNS: 100 CNS: 0

Necrotizing vasculitis

All types & sizes (venous accentuation)

Oral ulcers: ~95–100 Genital ulcers: 60–80 Skin: 70–90 Eye: 60–70 Joints: 40–60 Phlebitis: ~30 (deep ~5) GI: 5–15 Testes: 5 Lung: 5

qESR: ~60 qWBC: ~50 pHgb: ~15 RF () ANA () ANCA () pComp ()

PNS: rare ( 1%) CNS: 5–10 (Middle East, Japan); 20–30 (Europe, North America)

*May also be associated with leukocytoclastic vasculitis in skin. ANA  antinuclear antibodies; ANCA  antineutrophil cytoplasmic autoantibodies; Angio  abdominal angiographically demonstrated microaneurysms; CNS  central nervous system; Comp  circulating complement factors; CrN  cranial nerve involvement; Cryos  cryoglobulins; dsDNA  double-stranded DNA; ENT  upper respiratory tract involvement; Eos  eosinophils; ESR  erythrocyte sedimentation rate; GI  gastrointestinal; Hep B  hepatitis B surface antigenemia; Hep C  hepatitis C antibodies or RNA; Hgb  hemoglobin; IgA  immunoglobulin A; IgE  immunoglobulin E; LCV  leukocytoclastic vasculitis; MPO  myeloperoxidase; PMR  polymyalgia rheumatica; PNS  peripheral nervous system; PR3  proteinase 3; RA  rheumatoid arthritis; RF  rheumatoid factor; SLE  systemic lupus erythematosus; WBC  white blood cell count.

microaneurysms in renal, hepatic, and mesenteric arteries in two thirds of patients and nonspecific occlusive changes in 98%.599 Pathologically, PAN is a focal, segmental, necrotizing vasculitis of small and medium-sized arteries and, less commonly, arterioles; it does not involve large arteries, venules, or pulmonary vessels.61,435 Granulomas and eosinophilic infiltrates are rarely observed.

Microscopic Polyangiitis MPA is a “microscopic” form of c-PAN that affects arterioles, capillaries, and venules primarily, without granulomas and with minimal to no immune deposits.8,83,108,255,260,531,557,579 The presentation of MPA is similar to PAN, but there are distinguishing features, including a frequent association with rapidly progressive glomerulonephritis (~80% of cases) and pulmonary involvement in 35% to 40% of patients (20% with alveolar hemorrhage resulting from capillaritis). Pathologically, vasculitis-related glomerulonephritis is a focal, segmental, necrotizing process associated with crescent formation but no or few immune deposits.331 Palpable purpura is also more common in MPA than c-PAN, because the usual pathologic correlate is venular leukocytoclastic vasculitis. MPA is further contrasted from c-PAN by the rarity of

HBV surface antigenemia and visceral aneurysms.260 The incidence of increased ESR, anemia, leukocytosis, ANAs, and RFs is similar to that in c-PAN, but hypocomplementemia is rarely encountered. Unlike c-PAN, ANCAs are positive in up to 90% of MPA patients, most of them anti-MPO p-ANCAS. Peripheral neuropathy was identified in only 20% of MPA patients in the older nephrology literature.8,531,557,579 However, in more recent and larger series, neuropathy occurred in 60%, a prevalence equal to that of c-PAN.108,260 MPA tends to be a more chronic, relapsing disease than c-PAN, with relapse rates of 25% to 35%. Death ensues most commonly from renal failure or pulmonary hemorrhage.

Churg-Strauss Syndrome CSS is another ANCA-associated disorder, characterized by asthma, eosinophilia, and histologic evidence of necrotizing small vessel vasculitis affecting the systemic and pulmonary circulations.83,155,256,264,375,412,474,521,575,594 Churg and Strauss reported three primary histopathologic alterations: (1) eosinophilic tissue infiltration, (2) necrotizing vasculitis, and (3) extravascular granulomas.103 However, although over 90% of autopsies reveal vasculitis, only 40% show granulomas or eosinophilic infiltrates and

2356

Diseases of the Peripheral Nervous System

only 24% contain all three lesions.375 Moreover, just 10% to 20% of biopsies of any organ reveal granulomas.102,412 Thus clinical diagnostic criteria have been developed. The Lanham criteria require (1) asthma, (2) eosinophilia (1500/mm3), and (3) vasculitis involving at least two nonpulmonary sites.375 The mean age of onset of asthma in CSS has ranged from 30 to 50 years in various series. Symptoms corresponding to the vasculitis are typically delayed 5 to 10 years from the onset of the asthma. Thus CSS tends to develop at a younger age than c-PAN or MPA. There are three phases of the illness.375 The first is a prodromal period, sometimes lasting many years, consisting of atopic manifestations such as asthma and allergic rhinosinusitis. The second phase is marked by peripheral eosinophilia and eosinophilic tissue infiltration, leading to eosinophilic pneumonia and gastroenteritis in some patients. The third and final phase is defined by the emergence of systemic vasculitis. Clinical features distinguishing CSS from c-PAN are (1) asthma (100% of patients) and pulmonary infiltrates (~60%); (2) prodromal allergic rhinitis and polyposis (~60%); (3) greater prevalence of skin lesions (~60%); (4) focal, necrotizing glomerulonephritis (~30%); and (5) increased incidence of congestive heart failure (25%). Constitutional, joint, and GI symptoms occur with equal frequency in CSS and c-PAN. After lung involvement, neuropathy is the second most common manifestation of CSS, developing in 70% of patients. This is the highest incidence of neuropathy in all of the systemic necrotizing vasculitides. Relapses occur in approximately 30% of patients, similar to MPA but greater than c-PAN. Fifty percent of deaths are secondary to vasculitis-related congestive heart failure. Elevated ESR, leukocytosis, and anemia occur in 80% to 90% of patients; ANAs, cryoglobulins, and complement are usually normal or negative. Immunoglobulin E levels are elevated in approximately 75% of patients285 and RFs in approximately 40%.375,594 ANCAs occur in 50% to 60% of patients, usually anti-MPO p-ANCAs. There is no association with HBV or HCV infection. Visceral angiograms reveal microaneurysms in 30%.256 In sural nerve biopsies from patients with CSS-associated neuropathy, necrotizing vasculitis of epineurial arteries is observed in 55%, eosinophilic infiltrates in 15%, and granulomas in 25%; immune deposits are sparse.285

Wegener’s Granulomatosis WG, the third ANCA-associated vasculitis, is a rare but unique disorder characterized by granulomatous inflammation of the respiratory tracts and a variable, disseminated, pauci-immune, small vessel vasculitis with a predilection for lungs and kidneys.234 It is defined pathologically by the triad of granulomatous inflammation,

necrosis, and vasculitis, in which necrosis and granulomas can be either intra- or extravascular.234,407,669 The earliest pathologic change in lung and nasal biopsies is focal fibrinoid degeneration of collagen, with microabscess formation, which later evolves into a pathognomonic palisading granuloma.405,669 The primary process involves multifocal necrosis of collagen and other mesenchymal tissues, not vasculitis. Necrotizing glomerulonephritis is a frequent accompaniment, but may be absent in the “limited” form of the disease. The vasculitis itself affects the microvasculature and small to medium-sized arteries and veins. WG can present at any age, but rarely before adolescence; the mean age of onset is about 40 years.18,58,59,83,253,259,302,396,522 Most patients pass through an indolent initial phase, characterized by granulomatous inflammation in the upper and lower respiratory tracts, lasting up to 20 years.567 Symptoms during this phase usually relate to upper respiratory involvement, including chronic sinusitis, bloody or purulent nasal drainage, nasal ulceration, septal perforation, saddle nose deformity, and otitis media. Transition to the generalized, vasculitic phase is typically marked by the appearance of constitutional symptoms, such as fevers in 50% of patients and arthralgias in 75%. Once the vasculitis is established, primary involvement occurs in the upper respiratory tract (90% to 95% of patients), lungs (75%), glomeruli (75%), and eyes (50% to 60%). Multiple, bilateral, cavitary pulmonary nodules occur in 60% of patients. Renal disease may smolder or progress fulminantly to end-stage failure in days to weeks. Any eye compartment can be affected, with proptosis resulting from retro-orbital pseudotumor a useful finding. Other potentially involved organs include the skin, heart, GI tract, and trachea (subglottic stenosis in 15%). This disease has the highest relapse rate of all the systemic necrotizing vasculitides, ranging from 37% to 67% in large series (mean, ~50%).62,522 Vasculitis of the kidneys or lungs is the usual cause of death.302,522 Retrospective studies have documented a relatively low (~15%) prevalence of neuropathy in WG.119,473 One prospective series revealed a much higher (44%) prevalence, including 24% distal symmetrical polyneuropathies and 20% multifocal neuropathies.149 However, other causes for neuropathy were not excluded, and no neuropathologic data were presented. Neuropathies occurred more commonly in older patients and those with renal insufficiency, both independent risk factors for a nonvasculitic polyneuropathy. Multifocal neuropathies responded to immunosuppressive therapy, but symmetrical neuropathies did not. For these reasons, the 20% prevalence of multifocal neuropathy is probably a better approximation of the true prevalence of vasculitic neuropathy in WG. Cranial neuropathies, often multiple and most commonly involving nerves II, VI, or VII, occur in 10% of patients, but these are usually produced by contiguous

Neuropathies with Systemic Vasculitis

extension from nasal or paranasal granulomas rather than vasculitis.19,473 Laboratory abnormalities in WG include markedly elevated ESR in 80% to 85% of patients, anemia in approximately 75%, thrombocytosis in 55%, leukocytosis in 35%, and RFs in 50%. ANCAs directed against PR3 occur in 90% of patients with active disease. When measured serially, ANCA titers are predictive of clinical relapses.62,253,276, Diagnostic yield of head and neck biopsies is relatively low, with the complete triad occurring in 15% of biopsies, two components of the triad in 20% to 30%, and a single component in 25% to 50%.152,158 Transbronchial biopsies are rarely diagnostic.302 Information on nerve pathology is sparse. In one series of 54 cases, 7.5% of postmortem examinations revealed necrotizing arteriolitis in peripheral nerves, without granulomas.650 Nevertheless, there are rare reports of sural nerve biopsies showing granulomatous inflammation.192 Open lung biopsies are the gold standard, yielding vasculitis, necrosis, and granulomas in 90% of specimens.629 In many centers today, the diagnosis is established by (1) an appropriate clinical picture, (2) positive ANCAs, and (3) a suggestive or diagnostic head and neck biopsy.253

VASCULITIS ASSOCIATED WITH CONNECTIVE TISSUE DISEASES Secondary forms of systemic necrotizing vasculitis, occurring in the setting of a diagnosed connective tissue disease, are distinguished from the primary forms of vasculitis just described. Connective tissue disorder–related vasculitic neuropathies are most commonly encountered in patients with RA and SLE, but may also develop in Sjögren’s syndrome, scleroderma, and Behçet’s disease. The clinical and pathologic features of the neuropathy and multiorgan involvement in these disorders resemble those of c-PAN or MPA, but a wider range of blood vessels can be affected, from capillaries to large arteries.

Rheumatoid Arthritis RA is a chronic, inflammatory, multisystem disorder, characterized by symmetrical synovitis of the peripheral joints.390 It is the most common connective tissue disease, affecting about 1% of the population. Most patients develop symptoms between the ages of 35 and 50 years. The cause of RA is unknown.587 According to one hypothesis, the disease occurs in genetically susceptible individuals as an exaggerated immune response to some antigenic stimulus, possibly an infection.193 This event could precipitate a persistent infection of the joint space, immune response against some joint component antigenically altered by the infection, cross-reactive

2357

immune attack on the joint arising from molecular mimicry between the infectious agent and an antigen in the joint, or stimulation of T cells by release of a superantigen microbial product. Despite the theoretical appeal of this hypothesis, no infectious agent or other environmental trigger has been identified. RA usually presents insidiously with pain, stiffness, and swelling in multiple joints, combined with constitutional symptoms such as fatigue, anorexia, and weight loss.390 Extra-articular involvement occurs in most patients with long-standing disease. One such manifestation is systemic RV, which develops more commonly in patients with high RF levels, decreased C3 levels, joint erosions, subcutaneous nodules, prior treatment with disease-modifying antirheumatic drugs, and Felty’s syndrome (RA, splenomegaly, and neutropenia).85,644,645 In autopsy series, approximately 20% of patients have vasculitis.46,57,132,347 Skin involvement is the most frequent (70% of cases) and often the initial feature of RV, presenting as a dermal, small vessel, necrotizing venulitis (palpable purpura, maculopapular erythema elevatum diutinum, or hemorrhagic bullae) or a small to medium vessel arteritis (subcutaneous nodules, livedo reticularis, atrophie blanche, or ulcers).97 Both of these lesions can presage more severe visceral disease. Neuromuscular manifestations are next most common. Neuropathies occur in 45% of RV patients, indicating that 10% of all RA patients will develop a vasculitic neuropathy. 358 Neuropathic signs and symptoms are similar to those in other vasculitic neuropathies, except for a higher proportion (50%) of predominantly sensory neuropathies.508 Other tissues sometimes affected include heart, lungs, kidneys, and GI tract.219,572,643,651 Laboratory findings are similar to those in the primary systemic vasculitides, although complement is more commonly decreased in this disorder. p-ANCAs directed against lactoferrin are found in 40% of patients.128 Peripheral neuropathies, in general, are relatively common in RA, with a 15% prevalence in clinical studies and 25% in electrodiagnostically based series.358 Most patients have a slowly progressive, distal, symmetrical, sensory or sensorimotor polyneuropathy. Electrodiagnostic studies reveal axonal motor involvement, even when clinical findings are mainly sensory. Information on nerve pathology of symmetrical polyneuropathies is sparse, although some biopsies show necrotizing vasculitis or IgM deposits in the walls of morphologically normal vessels (supportive of vasculitis).124,508 More frequently, postmortem and biopsy studies show nonspecific axonal changes and old healed vasculopathic lesions such as thrombosed vessels, neovascularization, myointimal hyperplasia, and segmental destruction of arterial walls.41,493,656 Similar bland vascular lesions are common in rheumatoid patients with multifocal neuropathies, arguing for a common pathogenesis in both symmetrical

2358

Diseases of the Peripheral Nervous System

and asymmetrical cases.123 In one autopsy study, 26 of 31 patients had perivascular, mononuclear, and occasionally nodular inflammation in the perineurium and endoneurium of peripheral nerves, suggesting that chronic subclinical vascular inflammation may produce low-grade vascular damage and secondary PNS ischemia.451

Systemic Lupus Erythematosus SLE is a chronic, multisystem disorder characterized clinically by exacerbations and remissions and immunologically by production of multiple autoantibodies, activation of complement, immune complex generation, and altered cellular immunity.273,503 It is the third most common connective tissue disorder, with a prevalence of about 50 per 100,000. Women are preferentially affected (90% of cases), especially during childbearing years. Pathogenically, SLE results from the initiation and propagation of an abnormal immune response, probably triggered by selfantigen (nucleosome) accumulation related to defective clearance of apoptotic cells or an environmental stimulus in a genetically susceptible host.207,524,667 The abnormal immune response features polyclonal and antigen-specific hyperactivation of B and T cells, yielding numerous pathogenic antibodies that react with self-antigens.273,351 These antibodies may bind directly to antigens in target tissues, forming in situ immune complexes, and then activate complement, recruit effector cells, and stimulate other immune mechanisms, resulting in tissue damage. Alternatively, autoantibodies may form circulating immune complexes that deposit in tissues and elicit similar pathogenic mechanisms. An inflammatory vasculopathy could also result from a process modeled by the Shwartzman phenomenon, wherein episodic, uncontrolled activation of circulating complement proteins induces interactions between primed neutrophils and activated endothelial cells.45 All of these processes might be relevant to lupus vasculitis. The clinical spectrum and prognosis of SLE is diverse.273,503,640 Lupus is a multisystem disorder, but patients often have manifestations in only one or two organs at any one time. The most common initial areas of involvement are joints in 50% of patients and skin in 25%. Other frequently affected organs include kidneys, blood, CNS, lungs, and heart. The 1982 ACR criteria for SLE have greatly simplified the diagnosis of this condition (Table 105–7), with a sensitivity of 97% and specificity of 98%.273,615 A positive ANA is detected in over 95% of SLE patients but is not specific to SLE, because many other autoimmune diseases, drugs, infections, and inflammatory disorders can also produce a positive ANA. More specific to SLE are antibodies directed against double-stranded DNA or the Smith (Sm) antigen.273

Clinically defined neuropathies occur in approximately 10% of SLE patients, increasing to approximately 25% when electrodiagnostic criteria are employed.358 Most common is a sensorimotor polyneuropathy with no distinguishing features.335,421,490 Patients develop distal tingling, numbness, and dysesthesias that spread proximally over weeks to months. Uncommonly, severe distal weakness emerges.56,408 At onset, the neuropathy may be mildly asymmetrical, but with time most patients develop a symmetrical pattern. Few neuropathologic examinations have been reported in SLE patients with symmetrical polyneuropathies. Some (primarily in patients with significant motor involvement) revealed vasculitis or changes suspicious for vasculitis; others showed nonspecific axon loss and degeneration.56,335,408,421 The full spectrum of nerve pathology in this group of patients has not been adequately investigated. Up to one third of SLE patients with neuropathy have a discrete or overlapping multiple mononeuropathy, probably secondary to vasculitis affecting small to mediumsized vessels.144,185,291,313,409 Although no autopsy series has addressed the incidence of vasculitis in SLE, biopsysupported clinical studies suggest that approximately 40% of patients develop vasculitis, with predominantly cutaneous involvement (80% of patients).164,185,195,233,299,640 The PNS is the second most commonly affected organ (10% to 15%), suggesting that approximately 5% of all patients with SLE will develop a vasculitic neuropathy.144,164,238 Mesenteric, retinal, renal, coronary, and pulmonary vessels are less frequently involved. Of the reported SLE patients with vasculitic neuropathy, 70% had an asymmetrical or multifocal neuropathy, 10% a symmetrical polyneuropathy with major motor manifestations, 10% a lumbosacral polyradiculopathy, and 10% a GBS-like disorder.358 Lupus vasculitis is believed to be an immune complex–mediated process, but vascular deposits of immunoglobulin and complement are not usually observed in nerve biopsies from these patients.313,421

Sjögren’s Syndrome Sjögren’s syndrome is a chronic autoimmune disease characterized by lymphocytic infiltration of exocrine glands, especially the lacrimal and salivary glands, producing dry mouth (xerostomia) and dry eyes (keratoconjunctivitis sicca).338,404 It primarily affects women (90% of cases) and has a prevalence ranging from 0.5% to 3%.136,623,625 The term primary Sjögren’s syndrome is used when the disorder occurs in isolation. Secondary Sjögren’s syndrome refers to a similar syndrome in patients with another connective tissue disease, most commonly RA or SLE. The glandular infiltrate is composed primarily of activated Th cells with a memory phenotype and less frequent B cells and plasma cells. Epithelial cells in the affected glands

Neuropathies with Systemic Vasculitis

2359

Table 105–7. Criteria for the Classification of Systemic Lupus Erythematosus273,615 Criterion* 1. Malar rash 2. Discoid rash 3. Photosensitivity 4. Oral ulcers 5. Arthritis 6. Serositis

7. Renal disorder 8. Neurologic disorder

9. Hematologic disorder 10. Immunologic disorder†

11. Antinuclear antibody

Definition Fixed erythema, flat or raised, over the malar eminences, tending to spare the nasolabial folds Erythematous raised patches with adherent keratotic scaling and follicular plugging; atrophic scarring may occur in older lesions Skin rash as a result of unusual reaction to sunlight, by patient history or physician observation Oral or nasopharyngeal ulceration, usually painless, observed by physician Nonerosive arthritis involving two or more peripheral joints, characterized by tenderness, swelling, or effusion (a) Pleuritis — convincing history of pleuritic pain or rub heard by a physician or evidence of pleural effusion; or (b) Pericarditis — documented by electrocardiogram or rub or evidence of pericardial effusion (a) Persistent proteinuria greater than 0.5 g/day or greater than 3 if quantification not performed; or (b) Cellular casts — may be red cell, hemoglobin, granular, tubular, or mixed (a) Seizures — in the absence of offending drugs or known metabolic derangements (e.g., uremia, ketoacidosis, or electrolyte imbalance); or (b) Psychosis — in the absence of offending drugs or known metabolic derangements (e.g., uremia, ketoacidosis, or electrolyte imbalance) (a) Hemolytic anemia — with reticulocytosis; or (b) Leukopenia — 4000/mm3 total on two or more occasions; or (c) Lymphopenia — 1500/mm3 on two or more occasions; or (d) Thrombocytopenia — 100,000/mm3 in the absence of offending drugs (a) Anti-DNA — antibody to native DNA in abnormal titer; or (b) Anti-Sm — presence of antibody to Sm nuclear antigen; or (c) Positive finding of antiphospholipid antibodies based on (1) an abnormal serum level of immunoglobulin G or M anti-cardiolipin antibodies, (2) a positive test result for lupus anticoagulant using a standard method, or (3) a false-positive serologic test for syphilis known to be positive for at least 6 months and confirmed by Treponema pallidum immobilization or fluorescent treponemal antibody absorption test An abnormal titer of antinuclear antibody by immunofluorescence or an equivalent assay at any point in time and in the absence of drugs known to be associated with “drug-induced lupus” syndrome

*1982 criteria of the American College of Rheumatology; four or more criteria are mandatory for diagnosis. † Modified in 1997 (Hochberg, M. C.: Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 40:1725, 1997).

express MHC class II molecules and secrete proinflammatory cytokines, possibly acting as antigen-presenting cells. B-cell hyperactivity is a consistent finding, reflected in autoantibody production, hypergammaglobulinemia, and oligo- or monoclonal immunoglobulin formation.338,404 Among the many autoantibodies encountered in patients with primary Sjögren’s syndrome are RFs and antibodies directed against two ribonuclear proteins, Ro/SS-A and La/SS-B. Anti-Ro antibodies occur in 60% to 75% of patients, anti-La antibodies in 40% to 60%, and RFs in 60% to 80%.338 Although characteristic, these antibodies are neither specific to Sjögren’s syndrome nor required for diagnosis. Sjögren’s syndrome may result from an aberrant immune response to a virus latently infecting the epithelial cells of salivary and other exocrine glands, triggering the cellular infiltration of the gland and the B-cell activation.125,338 In support of this theory, HTLV-1, HCV, and HIV infections are all known to produce a Sjögren’s syndrome–like disorder.365,464,581 The disease generally evolves over years to decades, with dry eye symptoms (burning, itching, redness, and foreign body sensation) and xerostomia (difficulty swallowing or speaking, change in taste, frequent need for water,

and “cotton mouth”).210,338,404,585 Fifty percent of patients develop chronic or recurrent salivary gland enlargement. Other exocrine glands may also be affected, producing dryness of the skin and mucosal surfaces of the upper respiratory tract, GI tract, and vagina. At least 50% of patients develop a multisystem, extraglandular disorder heralded by constitutional symptoms and characterized by arthritis, lymphadenopathy, Raynaud’s phenomenon, interstitial lung disease, glomerulonephritis, cutaneous vasculitis, liver disease, polymyositis, and nervous system involvement. Several different sets of classification criteria have been proposed for primary Sjögren’s syndrome. In 1993, a multicenter European Study Group published diagnostic criteria that have gained widespread acceptance.642 They were recently revised by an AmericanEuropean Consensus Group (Table 105–8).641 The prevalence of neuropathy in Sjögren’s syndrome is approximately 25%.21,221,297,417,444,646 Sensory neuronopathy caused by dorsal root ganglionitis is a distinctive but uncommon Sjögren’s-related neuropathy, occurring in 3% to 4% of patients and accounting for 15% to 20% of all neuropathies seen in the disease.221,247,297 Patients present with acute, subacute, or chronic paresthesias and

2360

Diseases of the Peripheral Nervous System

Table 105–8. Revised International Classification Criteria for Sjögren’s Syndrome641 Criterion I. Ocular symptoms II. Oral symptoms III. Ocular signs IV. Histopathology V. Salivary gland involvement VI. Autoantibodies Primary Sjögren’s syndrome: Secondary Sjögren’s syndrome: Exclusion criteria:

Definition Dry eyes every day for more than 3 mo, recurrent sensation of sand or gravel in the eyes, or use of tear substitutes more than three times a day Daily feeling of dry mouth for more than 3 mo, recurrent or persistently swollen salivary glands as an adult, or use of liquids to aid in swallowing dry food Abnormal Schirmer’s I test (5 mm in 5 min) or rose bengal score ( 4 according to van Bijsterveld’s scoring system) Focus score 1 in a minor salivary gland biopsy Positive result in one of the following tests: salivary scintigraphy, parotid sialography, salivary flow (1.5 mL in 15 min) Antibodies to Ro (SS-A) or La (SS-B) antigens (a) Any 4 of 6 items present (including IV or VI); or (b) 3 of 4 objective items present (III, IV, V, VI) (a) Another well-defined connective tissue disease; and (b) Item I or II present plus 2 items from among III, IV, and V Past head or neck radiation treatment Hepatitis C infection Human immunodeficiency virus infection Pre-existing lymphoma Sarcoidosis Graft-versus-host disease Use of anticholinergic drugs

dysesthesias involving the limbs, trunk, and face, with unsteady gait, loss of kinesthetic sensation in the extremities, and autonomic dysfunction. Examination reveals prominent large fiber sensory loss, areflexia, and normal strength.247,250,401,593 Although findings are usually asymmetrical at onset, they become symmetrical as the disease generalizes. More common is a distal, symmetrical, sensory-predominant polyneuropathy, which accounts for 75% of Sjögren’s neuropathies.358 Patients develop slowly progressive numbness and paresthesias in their feet, sometimes accompanied by pain and mild distal weakness. There are usually no associated autonomic features. The third and least common pattern is a sensorimotor multiple mononeuropathy. Vasculitis represents the primary pathologic mechanism for most multifocal neuropathies and some symmetrical polyneuropathies in Sjögren’s syndrome. Because systemic vasculitis occurs in 25% of Sjögren’s syndrome patients, and one third of these have neuropathy, it can be inferred that 5% to 10% of Sjögren’s syndrome patients will develop vasculitic neuropathy.51,210,443,501,630 The pathogenesis of the ataxic sensory neuronopathy involves cytotoxic/ suppressor T-cell infiltration of trigeminal sensory and dorsal root ganglia.247,250,401 The triggers for this T-cell–mediated attack against sensory ganglion cells are unknown, as are the mechanisms underlying most symmetrical sensory or sensorimotor polyneuropathies. While some nerve biopsies in these patients reveal definite

or probable vasculitis,344,430 others show nonspecific axonal or mixed axonal/demyelinating nerve alterations, epineurial perivascular inflammation, and abnormal thickening of the endoneurial microvasculature.221,247,358 Some clinically diagnosed sensory polyneuropathies probably represent unrecognized dorsal root ganglionopathies.

Scleroderma Scleroderma is a chronic disorder characterized by alterations of the microvasculature, disturbances of the humoral and cellular immune system, and massive proliferation of collagen and extracellular matrix proteins, resulting in fibrosis of the skin and visceral organs.231,576,632 Its reported prevalence ranges from 30 to 120 per 1 million; women are three to four times more commonly affected than men.582 The earliest pathologic change, occurring prior to fibrosis, is damage to small arteries and microvessels in the skin and viscera.231,576 Because skin biopsies in early stages of the illness also show perivascular and interstitial inflammatory infiltration of CD4 T cells and macrophages, scleroderma may result from a primary immune microvasculopathy, producing chronic skin and visceral organ ischemia and proliferation of fibroblasts and mast cells.340 The presence of autoantibodies in many patients also supports the importance of immune mechanisms. The etiologic trigger for the vascular injury is

2361

Neuropathies with Systemic Vasculitis

unknown, but occupational exposures may be operative in some cases.126 Cytokines such as TGF- and connective tissue growth factor appear to be important downstream profibrotic mediators.156 Scleroderma is classified as localized if restricted to the skin and systemic if other organs are affected.418,576 Systemic scleroderma is, in turn, subdivided into limited cutaneous and diffuse cutaneous forms. Limited cutaneous scleroderma is defined by skin involvement that does not extend proximal to the knees or elbows. If both proximal and distal limbs or the trunk are affected, the disorder is termed diffuse cutaneous scleroderma. Many patients with limited cutaneous scleroderma fulfill criteria for CREST syndrome (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, and telangiectasia). Patients with scleroderma typically present with swelling and tightness in the fingers, often accompanied by Raynaud’s phenomenon (95% of patients).231,576,632 Characteristic skin thickening then ensues over months to years, producing “sausage”-like fingers. In diffuse cutaneous scleroderma, fibrotic skin changes in the proximal limbs and joint contractures develop. Esophageal dysfunction emerges in most patients, resulting in dysphagia and gastroesophageal reflux. Intestinal involvement produces hypomotility and malabsorption syndromes. Serious pulmonary, cardiac, and renal manifestations may also occur. Ten percent of patients develop systemic vasculitis.140 The diagnosis of scleroderma is based on typical clinical features supported by laboratory findings. ANAs are present in almost all patients, but two autoantibodies— anticentromere and antitopoisomerase-1 (anti-Scl-70)— are more specific for scleroderma. Antitopoisomerase-1 antibodies are found in approximately 40% of patients with diffuse cutaneous disease and anticentromere antibodies in 60% to 80% of patients with limited cutaneous scleroderma.231 Antinucleolar antibodies occur in 20% to 30% of patients, and some of these (e.g., anti-RNA polymerases I, II, and III) are relatively specific for scleroderma. Prospective studies employing electrodiagnostic techniques indicate that 20% to 25% of scleroderma patients have a neuropathy.29,296,562 Trigeminal sensory neuropathy is most characteristic, occurring in 3% of patients.181,358 Similar neuropathies are seen in Sjögren’s syndrome, SLE, and mixed connective tissue disease. Most sensorimotor polyneuropathies in scleroderma begin asymmetrically, but 60% evolve into a distal, symmetrical pattern.358 In contrast to the sensory-predominant polyneuropathies seen in the other connective tissue disorders, polyneuropathy in scleroderma typically has motor involvement (90%), usually distal and sometimes severe. Although the asymmetrical onset in many patients suggests vasculitis, there are few nerve biopsy–proven cases of vasculitis in scleroderma.169,293,437,446,481 In biopsysupported cases, vasculitic neuropathy is strongly

associated with the CREST subtype (80% of patients).358 Vasculitic neuropathy occurs in 1% of patients with CREST syndrome169 and 0.5% of all scleroderma patients.482 In nonvasculitic cases, nerve biopsies reveal myelinated nerve fiber loss and axonal atrophy, increased endoneurial and perineurial connective tissue, and an endoneurial microvasculopathy.127,159,527 The neuropathy, in these patients, may result from a chronic noninflammatory vasculopathy.

Behçet’s Disease Behçet’s disease is a chronic inflammatory disorder characterized by recurrent oral and genital ulcers, uveitis, and cutaneous, arthritic, and neurologic manifestations.426,545,668 A mixed neutrophilic and lymphocytic large and small vessel vasculitis is its major pathologic feature.363 The pathogenesis of this disease is obscure. Genetic investigations have revealed a tendency for familial aggregation and a strong association with human leukocyte antigen (HLA)-B51.441 Neutrophil hyperreactivity has been demonstrated in some studies, suggesting that neutrophils may exist in a primed, preactivated state.162 T cells are also activated in Behçet’s disease, with a predominant Th1 cytokine profile (INF-, IL-2, IL-8, and IL-12). External and internal antigen-driven immune responses occur in patients with this disease. Behçet’s disease may thus arise from a genetic defect in a transcription factor or inflammatory mediator that predisposes to more intense neutrophilic and T-cell responses to antigenic stimuli.162 Behçet’s disease is more common in Mediterranean countries, India, and Japan than in North America, where its prevalence is 0.12 to 0.33 per 100,000.341,545,668 It occurs mainly in young adults and is more common and severe in males. The disorder usually presents with painful oral aphthous ulcers, which persist for 1 to 2 weeks and then resolve. Genital ulcers occur in 70% to 80% of patients. Uveitis or retinal vasculitis is often present at onset, but may be delayed for years. Blindness ensues in 10% to 15% of patients. Potential systemic manifestations include cutaneous, joint, large vessel (venous thromboses and arterial aneurysms), GI tract, and CNS involvement. Peripheral neuropathy is rare, affecting 0.5% of patients.561 An axonal sensorimotor polyneuropathy, multiple mononeuropathy, lumbosacral polyradiculopathy, and motor-predominant polyradiculoneuropathy have all been described.34,179,465,482,612

HYPERSENSITIVITY VASCULITIS The term hypersensitivity vasculitis refers to a wide spectrum of diseases having predominant involvement of small vessels in the skin, characterized by leukocytoclastic

2362

Diseases of the Peripheral Nervous System

angiitis.237,549 Predominant cutaneous vasculitis has a similar connotation. Pathologically, acute lesions feature neutrophilic infiltration of small-caliber vessels— especially postcapillary venules—with associated fibrin deposition, hemorrhage, and leukocytoclasia (nuclear debris from fragmented neutrophils). In lesions older than 72 hours, the infiltrate becomes progressively lymphocytic. Unlike MPA, exposure to a precipitating antigen is implied or apparent (e.g., organism, drug, tumor antigen, or self-protein such as DNA or immunoglobulin) and immune complex deposition is the presumed pathogenesis. IgG, IgM, and complement proteins are usually detected in and around vessel walls by immunofluorescent staining. The initial clinical manifestation of HSV is palpable purpura, most often localized to the legs.214,237,410,434,549 Other skin lesions may occur, including vesicles, pustules, bullae, urticaria, subcutaneous nodules, ulcers, and livedo reticularis. In some patients, constitutional symptoms such as fever, malaise, arthralgias, myalgias, and weight loss develop. Noncutaneous organs are uncommonly involved. The course is usually self-limited. As with the systemic necrotizing vasculitides, both primary and secondary forms can be distinguished. In primary HSVs, the triggering antigen is poorly characterized or unknown. Examples include HSP and CLA, the latter defined by the absence of systemic manifestations. Secondary forms occur in reaction to specific drugs, infections, malignancies, connective tissue diseases, and mixed cryoglobulins (see Etiology).

Henoch-Schönlein Purpura HSP is a clinical syndrome distinguished by the association of skin, joint, and GI symptoms with a small vessel, leukocytoclastic vasculitis having immunoglobulin A (IgA) immune deposits in the involved tissues.211,212,240,434,530,555 It occurs predominantly in children between the ages of 2 and 10, but less frequently affects adults. There is an antecedent respiratory infection in 60% of cases, with a variety of implicated pathogens, most commonly group A -hemolytic streptococcus. Typical manifestations include fever, palpable purpura, oligoarticular arthritis in 60% to 80% of patients, and GI involvement (abdominal pain, vomiting, and bleeding) in 65% to 75%. Glomerulonephritis develops in 20% to 50% of patients, but rarely (~1%) progresses to end-stage renal disease. Some studies have found nephritis to be more frequent and severe in adults with HSP.211 Pathologically, IgA is deposited in the walls of small vessels and the glomeruli. The alternate pathway of complement is activated. Serum IgA levels are elevated in 50% of patients. While headaches occur in approximately 30% of patients, CNS and PNS manifestations are both rare. There are a few reports of

mononeuropathy and polyradiculoneuropathy and isolated cases of multifocal neuropathy and brachial plexopathy.44,77,84,242

TEMPORAL ARTERITIS TA (giant cell arteritis) is a granulomatous vasculitis involving large and medium-sized arteries, characterized by giant cell formation in the media of affected blood vessels.301,469,546 Arterioles, capillaries, and veins are largely spared. Because giant cells are demonstrated in less than 50% of temporal artery biopsy specimens, their presence is not mandatory for diagnosis. The arteritis has a predilection for branches of the aortic arch and the extracranial carotid arteries, but a systemic disorder can evolve that affects arteries throughout the body. TA occurs rarely before age 50. The ESR is usually markedly elevated, but can be normal in 5% of patients.548 Typical symptoms include headache, weight loss, fevers, and jaw claudication. Polymyalgia rheumatica (PMR), a synovitis that produces severe aching pains and stiffness in the neck and torso, shoulders and proximal arms, or hips and proximal thighs, occurs in 40% to 50% of patients.546 Almost 90% have vestibular dysfunction, and approximately 60% report hearing loss.17 There is a 15% to 20% risk of permanent visual loss from acute ischemic optic neuropathy.301 In about half of these patients, premonitory visual disturbances occur. Thoracoabdominal aortic aneurysms develop as a late complication in 10% to 15% of patients.178 The disease generally runs a self-limited course, but at least 50% of patients require treatment with corticosteroids (CS) for more than 2 years.33,301 In several series of TA patients (including one large population-based study), neuropathy was identified in approximately 15% of patients.49,90 Because TA is the vasculitis with the highest incidence in adults, it is difficult to reconcile a 15% prevalence of TA-related neuropathy with the fact that TA is nearly absent in series dedicated to vasculitic neuropathies (see Table 105–4). Most neuropathies in elderly patients with TA probably have alternate etiologies and bear no direct relationship to the vasculitis. There are, nevertheless, over 100 cases of TA-associated neuropathy in the literature, including diffuse polyneuropathy (41% of patients), focal or multifocal neuropathy most commonly involving the distal median nerve (40%), C5/6 radiculopathy (15%), motor neuron disorder (3%), and sensory neuronopathy (1%).49,73,79,90,235,431,468,470,528 The few reports of nerve pathology suggest that both nutrient and epineurial arteries can be affected by vasculitis.73,90,431 Postmortem examination of a TA patient with a focal motor neuron disorder of the arms revealed florid arteritis of the lower cervical radicular arteries.79

2363

Neuropathies with Systemic Vasculitis

NONSYSTEMIC VASCULITIC NEUROPATHY NSVN is the most commonly reported type of vasculitis affecting the PNS (see Table 105–4). Neurologic signs and symptoms and nerve pathology are similar to PAN/MPArelated neuropathy, but the clinical course is free of non-PNS involvement. Since the initial description of PAN restricted to peripheral nerves by Kernohan and Woltman in 1938,348 multiple reports of relatively small numbers of patients have appeared.2,120,142,165,360,459,495,544,583 In 1987, Dyck and colleagues described 20 such patients and coined the term nonsystemic vasculitic neuropathy.165 In all series, patients were excluded if they had a specific etiology for their neuropathy, a defined connective tissue disease, or diabetes. The diagnosis has generally depended on sural nerve biopsy evidence of definite or probable vasculitis,142,165 but some groups have also found muscle involvement in patients with otherwise typical NSVN.120,544 Diagnostic criteria for this entity have been proposed (Table 105–9).120 The proper nosologic status of NSVN is uncertain. Organ-specific vasculitis has also been described for the CNS,278 GI tract,78 skin,141 reproductive tract,199 heart,609 lungs,332 and kidneys.427 Thus NSVN may represent one of several localized vasculitic diseases mediated by immune responses directed against tissue-specific antigens, with a pathogenesis and prognosis fundamentally different than the systemic vasculitides.512 However, many features of NSVN overlap with the systemic necrotizing vasculitides. First, a substantial minority of NSVN patients develops weight loss and other constitutional symptoms. Second, most patients have an elevated ESR. Third, anemia, leukocytosis, thrombocytosis, and autoantibodies occur in 20% to 40% of patients. Fourth, muscle biopsies occasionally

Table 105–9. Diagnostic Criteria for Nonsystemic Vasculitic Neuropathy Inclusions • Clinical evidence of neuropathy by history and examination • Electrodiagnostic findings consistent with neuropathy • Nerve or nerve/muscle biopsy diagnostic of or suspicious for necrotizing vasculitis Exclusions • Clinical, laboratory, radiologic, or pathologic evidence of organ involvement outside peripheral nervous system (except muscle) • Identified etiologic agent (drug exposure or infection, especially hepatitis B, hepatitis C, human immunodeficiency virus, cytomegalovirus, or varicella-zoster virus) • Underlying systemic condition predisposing to vasculitis (connective tissue disease, malignancy, diabetes mellitus, mixed cryoglobulinemia)

reveal coexisting muscle vasculitis and often show ischemic changes or inflammation.118,120,544 Fifth, neuromuscular pathology is similar in specimens from nonsystemic and systemic vasculitic neuropathy patients, except for a greater incidence of muscle involvement in the latter.118,359,360 Sixth, vasculitic involvement belatedly extends to the skin and other organs in a few NSVN patients.120,539 Seventh, NSVN and systemic vasculitides are both better controlled with combination therapy than CS alone and have similar relapse rates.120 These observations support the premise that NSVN is actually a systemic process, but unique in its predilection for peripheral nerve involvement.

GENETICS Although genetic influences may be important in the pathogenesis of some vasculitides, no single gene sufficient for producing vasculitis has been identified.309,343 Familial clusters of vasculitis have been reported, most commonly for Takayasu’s arteritis, Behçet’s disease, and TA/PMR, but also for WG, PAN, HSP, and MPA.309,333,364,382,476,479,534 Most clusters have involved siblings or parent-child pairs, making shared environmental exposures difficult to exclude. Familial vasculitis is, nevertheless, an unusual occurrence. For example, of 173 patients with PAN or WG in one series, only 2 had affected family members.534 Familial clustering of NSVN has not been reported. Most T-cell–dependent autoimmune disorders exhibit positive or negative associations with HLA genes. HLA gene–encoded proteins function as antigen-presenting molecules and play an important role in the generation of T-cell–mediated immune responses.467 HLA associations have been investigated in almost all vasculitides except NSVN. There are strong positive associations between (1) TA/PMR and HLA-DRB1,659 (2) Behçet’s disease and HLA-B51,441 (3) Takayasu’s arteritis and HLA-B52 and -B39.2 (C1-2-A allele),356 and (4) RV and HLADRB1*401.502 These genes serve as disease susceptibility loci, conferring increased risk of disease occurrence in individuals harboring the appropriate genotypes. HLA associations in ANCA-associated vasculitides are weaker and more variable, suggesting that T-cell responses are heterogeneous in these disorders.251,343,460,476,496,678 Other genetic alterations in T cells might also influence the development of vasculitis. For example, constitutive depletion of CD8 cytotoxic/suppressor T cells might represent a genetic trait relevant to pathogenesis of TA/PMR.333 In addition, longer (AT)n microsatellite polymorphisms in the CTL-associated antigen-4 (CTLA-4) gene are strongly associated with WG.308,574,680 The CTLA-4 gene product suppresses antigen-driven T-cell responses by inhibiting the CD28 pathway. The association between these polymorphisms and WG implicates uncontrolled

2364

Diseases of the Peripheral Nervous System

antigen-specific T-cell activation in the pathogenesis of this disorder. An established risk factor for ANCA-associated vasculitis is deficiency of 1-AT, the main plasma inhibitor of PR3.343 1-AT is encoded by a polymorphic protease inhibitor (Pi) gene on chromosome 14 with over 75 alleles, which define nondeficient and variably deficient phenotypes.177,281 The PiZ allele, which is the most common variant responsible for reduced plasma concentrations of 1-AT, is overrepresented in patients with ANCAassociated vasculitis.174 Moreover, PiZ heterozygosity confers increased risk of multiorgan involvement and fatal outcome in PR3-ANCA–positive vasculitis.176,573 Heritable 1-AT deficiencies may promote ANCA formation as a result of increased PR3 exposure to immunocompetent cells and may enhance disease activity as a result of reduced restraint of PR3-mediated proteolysis.177,281 Another genetic factor in PR3 accessibility is the degree of constitutive PR3 expression on the surface of neutrophils. Patients with a large subset of neutrophils expressing PR3 are at increased risk of developing and relapsing from ANCA-associated vasculitis.517 Genetic variation might also modify inflammatory response pathways in ways that promote vasculitis. For example, various hereditary complement abnormalities are associated with vasculitis.27,438 In addition, polymorphisms in FcR genes critically affect interactions of these receptors with IgG antibodies and are heritable risk factors for autoimmune diseases.309,635 Three classes of FcR are distinguished; class I FcRs are expressed on endothelial cells and classes II and III on leukocytes. One study showed that the R131 form of FcRIIa and F158 form of FcRIIIa were risk factors for disease relapse in WG.161 Cytokine studies have shown no association between polymorphisms in TNF-, TNF-, IL-1, IL-2, IL-4, and IL-6 genes and ANCA-associated vasculitis.224,308,679,680 However, a polymorphism in the IL-10 gene conducive to increased autoantibody production occurs with increased frequency in patients with WG,679 TA is strongly associated with the TNF2 polymorphism of the TNF- gene,416 and the TNF-1031C allele confers increased susceptibility to Behçet’s disease.10 Concerning adhesion molecules, a polymorphism in the ICAM-1 gene is strongly associated with TA/PMR,547 and several polymorphisms in the 2 integrin gene are associated with MPO-ANCA vasculitis.429

CLINICAL FEATURES The clinical presentation of a patient with PNS vasculitis depends on the distribution of affected neural vessels, disease severity, and pattern of extraneurologic involvement. 6 7 , 6 9 , 9 1 , 9 6 , 9 9 , 1 1 8 , 1 2 0 , 1 4 2 , 1 6 5 , 2 8 5 , 2 8 6 , 2 9 0 , 3 6 0 , 3 9 4 , 4 2 1 , 4 4 5 , 447,459,472,495,508,544,583,654 Multiorgan involvement typifies

systemic vasculitis or vasculitis caused by an underlying connective tissue disease. In patients with vasculitic neuropathy secondary to a systemic illness, neuropathic symptoms are a presenting feature in 70% of cases69,91,96,285,290,394,447,508; in the other 30%, systemic symptoms precede the neuropathy, usually by several weeks to months.290 The most commonly affected organ in systemic vasculitis involving the PNS is the skin (30% of patients). Less frequently involved are the kidneys, joints, lungs, and GI tract, with even less common spread to the heart, liver, and eyes. Patients with NSVN, by definition, have no clinical evidence of extra-PNS involvement. Constitutional symptoms such as weight loss, fatigue, and arthralgias occur in 70% of patients with systemic necrotizing vasculitis and 35% to 40% of patients with NSVN. Fever is a more specific sign, developing in 65% of patients with systemic vasculitis and 10% to 15% with NSVN. The neuropathy associated with necrotizing arteritis has a characteristic clinical picture whether it occurs alone or as part of a systemic vasculitis. Although neuropathic symptoms usually evolve over weeks to months, the mean delay from symptom onset to diagnosis ranges from 5 to 11 months in various series.142,360 Some patients present with an onset so acute and severe as to mimic GBS,66 while others are symptomatic for up to 8 years prior to diagnosis.120 Seventy-five percent of patients experience at least one acute attack and 50% to 60% follow an acute relapsing course, but 25% manifest a chronic, slowly progressive course from onset.120,165 The classic presentation, produced by a succession of ischemic insults to individual nerves, is a multifocal neuropathy (multiple mononeuropathy) (Fig. 105–6). Patients first develop pain, weakness, and sensory loss in the distribution of a single peripheral or cranial nerve, followed by stepwise progression from one nerve territory to the next. In other patients, involvement is so extensive that the mononeuropathies “overlap,” creating an asymmetrical polyneuropathy that is usually distal predominant. In a subset of these patients, involvement is restricted to lower extremity nerves, producing an asymmetrical lumbosacral plexopathy. The distribution most difficult to recognize as vasculitic is a distal, symmetrical polyneuropathy. This pattern presumably results from extensive small vessel ischemic events at many levels of multiple nerve trunks. The relative prevalence of these patterns in the overall vasculitic neuropathy literature116 and in several series dedicated to NSVN142,165,459,495,544 is approximately 50% multifocal neuropathy, 25% asymmetrical polyneuropathy, and 25% distal, symmetrical polyneuropathy. In contrast, one large cohort of 48 patients with NSVN revealed asymmetrical findings in 98% of patients (77% asymmetrical polyneuropathy, 13% multifocal neuropathy, 8% asymmetrical lumbosacral plexopathy, and 2% distal, symmetrical polyneuropathy).120 The rarity of symmetrical presentations in vasculitic neuropathy is also emphasized

2365

Neuropathies with Systemic Vasculitis

FIGURE 105–6 Schematic depiction of three patterns of neuropathic involvement encountered in patients with vasculitic neuropathy. Of these three, the symmetrical pattern is rarest.

in other recent studies, in which less than 5% of patients had symmetrical deficits.99,286 Reported discrepancies between multifocal and asymmetrical neuropathy frequencies are probably more semantic than substantive. Clinically, the distinction between a “discrete” and “overlapping” multiple mononeuropathy is irrelevant. Certain nerves appear to have a propensity for vasculitic involvement, probably as a result of poor collateral vascular supply (Table 105–10). The common peroneal nerve is the most frequently affected nerve overall, and the ulnar nerve is most commonly affected in the arm. In one large series, most patients had extensive, overlapping involvement of multiple individual nerves, with legs more commonly affected than arms and distal nerves more

commonly affected than proximal nerves.120 Although most vasculitic neuropathies are distally accentuated, proximal nerves and muscles are often affected. In one NSVN cohort, 60% of patients had proximal lower limb and 40% proximal upper limb weakness.120 Ninety percent of patients present with combined motor and sensory deficits; in 10%, findings are purely or primarily sensory, especially in small vessel vasculitides (e.g., RV, Sjögren’s syndrome, EMC, and HIV infection).91,99,118,120,165,285,290,447 Pure motor presentations are vanishingly rare. The great majority of vasculitic neuropathies are painful.116,120 Summarizing, patients with (1) no asymmetries, (2) pure motor involvement, (3) no pain, or (4) entirely proximal findings are unlikely to have a vasculitic neuropathy.

Table 105–10. Most Commonly Involved Motor Nerves in Vasculitic Neuropathy Frequency of Involvement (%) Collins et al. (2003)120 Nerve Peroneal Tibial Ulnar Femoral Superior gluteal Median Radial Musculocutaneous Axillary

Per Patient

Per Nerve

Literature Review* (Per Patient)

90 81 65 63 52 48 44 40 40

82 73 58 59 49 45 42 39 38

91 42 41 7 — 29 18 — —

*Based on review of 485 cases in nine vasculitic neuropathy series. 96,118,149,165,256,260,266,285,539

2366

Diseases of the Peripheral Nervous System

NATURAL HISTORY Untreated systemic necrotizing vasculitides are nearly always fatal. For patients not receiving an immunosuppressive agent, mean survival in WG is 5 months; only 20% survive 1 year.650 For PAN-type vasculitides, old series reveal 5-year survival rates of 12% to 13%.203,379 In MPA complicated by renal vasculitis, mean survival is only 4.5 months.143,662 The prognosis of vasculitic neuropathy associated with systemic vasculitis is poor, consistent with the natural history of the underlying illness. In 1940, Boyd reviewed the world’s literature on polyneuropathy in PAN, finding 129 cases.69 He concluded that the behavior of the neuropathy was highly variable, such that neuromuscular symptoms “may vanish permanently, recur, or remain a sequela when all other evidence of the syndrome has disappeared.” The best natural history data on PANrelated neuropathy come from a series of 15 patients reported by Lovshin and Kernohan in 1948.394 In four patients, “clearcut” regression of signs and symptoms occurred. In the other 11, the neuropathy progressed or remained static until death from other causes. Survival after onset of symptoms ranged from 3 to 10 months. Deaths ensued from renal failure, cardiac disturbances, GI bleeding, bowel infarction, cerebral disorders, respiratory failure, or pneumonia. Because NSVN was first recognized as a distinct entity in the 1980s, well after the introduction of CS and cytotoxic therapy for systemic vasculitis, no natural history data exist on this condition.

DIFFERENTIAL DIAGNOSIS Vasculitic neuropathy must be distinguished from other asymmetrical or multifocal neuropathies (Table 105–11). With systemic vasculitis, consideration of the nonneurologic features and laboratory findings usually permits an accurate diagnosis. In patients with NSVN, diagnosis is more difficult. Among the more common disorders to be considered are multifocal, sensorimotor, demyelinating neuropathy (Lewis-Sumner syndrome), diabetic or nondiabetic lumbosacral radiculoplexus neuropathy, neuralgic amyotrophy, sarcoidosis, Lyme disease, hereditary neuropathy with liability to pressure palsies (HNPP), and neoplastic meningitis or neural infiltration. Laboratory studies (Table 105–12) and nerve biopsy are usually required for accurate diagnosis of these patients. Electrodiagnostic testing is also crucial in differentiating the demyelinating features of Lewis-Sumner syndrome and HNPP from the axonal pattern of a vasculitic neuropathy. Patients with diabetic lumbosacral radiculoplexus neuropathy (also known as diabetic amyotrophy or proximal diabetic neuropathy) present subacutely with progressive weakness and pain of the pelvifemoral muscles that is

often bilateral but asymmetrical (see Chapter 86).171 The distal lower limbs and, uncommonly, upper limbs may also be affected. A variant with predominant distal lower limb involvement has also been described.543 Some have proposed that this is an autoimmune small vessel vasculitis, supported by nerve biopsy findings of T-cell–predominant perivascular inflammation, complement deposition in vessel walls, multifocal nerve fiber loss, leukocytoclastic vasculitis, and other features indicative of vasculitis.170,346,541,671 However, necrotizing vasculitis is rarely observed170,541 and recovery without treatment is the rule.171,540 Therefore, if truly vasculitic, the process bears more resemblance to HSV than PAN. In patients with the distal variant, nerve biopsies more commonly reveal endoneurial and epineurial necrotizing vasculitis.541

DIAGNOSIS To diagnose vasculitic neuropathy, laboratory studies and detailed electrodiagnostic testing are required. Blood work is designed to screen for non-neurologic organ involvement and measure nonspecific markers of inflammation and autoantibody production that may signal the presence of a systemic vasculitis or connective tissue disease (see Table 105–12). More specific studies are employed selectively. In the authors’ experience, leukocytosis, elevated ESR, and positive RF are the best laboratory predictors of biopsy-confirmed necrotizing vasculitis. In NSVN, blood work is generally less abnormal than in systemic vasculitis, although a mildly elevated ESR (mean, 40 mm/h) occurs in 60% of patients.99,120,142,165,360,459,508,544,583 Other laboratory markers of inflammation or immune activation are abnormal in a minority of patients, including positive ANA in 30%, anemia in 30%, leukocytosis in 20%, positive RF in 15%, and decreased complement in 10%. For patients with neuropathies caused by systemic vasculitis, laboratory abnormalities are more common, with elevated ESR in 85%, leukocytosis in 70%, anemia in 50%, positive RF in 45%, positive ANA in 30%, and hypocomplementemia in 30%. None of these tests is disease specific, and correlation with the clinical profile is mandatory. The CSF is almost never inflammatory in any type of vasculitic neuropathy ( 5% of patients), but mild CSF protein elevation occurs in 30%.2,67,91,118,120,165,290,472,495 Electrodiagnostic studies are indispensable for documenting non–length-dependent features of the neuropathy, excluding a primary demyelinating process, and assisting with the selection of which nerve to biopsy.109,488,498,654 Typical electrodiagnostic findings reflect the underlying pathology of multifocal axonal damage to motor and sensory fibers. Nerve conduction studies reveal low-amplitude or unelicitable compound sensory nerve and muscle action potentials. Sensory responses are more

Neuropathies with Systemic Vasculitis

2367

Table 105–11. Differential Diagnosis of Multifocal, Asymmetrical Neuropathy Ischemic Neuropathies Peripheral nerve vasculitis Diabetes mellitus Diabetic lumbosacral radiculoplexus neuropathy* Mononeuropathy multiplex (cranial nerve, thoracic, limb)* Inflammatory/Immune-Mediated Neuropathies Sarcoidosis* Multifocal sensorimotor demyelinating neuropathy (Lewis-Sumner syndrome) Multifocal motor neuropathy Multifocal variants of Guillain-Barré syndrome Idiopathic brachial or lumbosacral plexopathy* Neuropathy with various eosinophilic syndromes* Neuropathy with gastrointestinal conditions (Crohn’s disease, ulcerative colitis, celiac sprue)* Graft-versus-host disease Hashimoto’s thyroiditis Infectious Neuropathies* Leprosy Lyme disease Viral (HIV, HTLV-1, varicella-zoster, cytomegalovirus) Diffuse infiltrative lymphocytosis syndrome Infective endocarditis Other: (group A -hemolytic streptococcus, Pseudomonas aeruginosa, Salmonella typhi, Mycoplasma pneumoniae, leptospirosis, hepatitis A, Plasmodium falciparum, ascaris, trichinosis) Drug-Induced Neuropathies* Antibiotics (penicillin, sulfonamides) Cromolyn Thiouracil

Allopurinol Amantadine Naproxen Interferon- Montelukast Drugs of abuse (amphetamines, cocaine, heroin) Genetic Neuropathies Hereditary neuropathy with liability to pressure palsies Hereditary neuralgic amyotrophy Porphyria Tangier disease Krabbe disease Mechanical Neuropathies Multiple peripheral nerve injuries Multifocal entrapments not related to a genetic disorder Burns Neuropathies Related to Malignancy Direct infiltration of nerves by tumor (neurolymphomatosis) Multifocal mass lesions with external compression effect (e.g., neurofibromatosis 2) Lymphomatoid granulomatosis Intravascular lymphoma Neoplastic meningitis Systemic amyloidosis Miscellaneous Conditions Sensory perineuritis Cholesterol emboli syndrome Idiopathic thrombocytic purpura Atrial myxoma Hemophilia

HIV  human immunodeficiency virus; HTLV-1  human T-lymphotropic virus 1. *Conditions occasionally associated with vasculitis.

pervasively affected than motor responses, and lower limb studies are more frequently abnormal than upper limb studies. H reflexes and F waves are low in amplitude or unevocable with normal latencies. Conduction velocities are normal or mildly reduced and distal latencies normal or mildly prolonged. “Pseudo” conduction block, resulting from acute wallerian-like degeneration (and not segmental demyelination) is occasionally identified if a patient is studied in the first week after an ischemic lesion has occurred, because the distal stump remains capable of conducting motor impulses for up to 6 days and sensory impulses for up to 10 days.74,420 These apparent conduction blocks are transient, disappearing as distal axons degenerate. True but transient partial motor conduction blocks are more rarely encountered,321 comporting with experimental evidence that conduction block can develop in ischemic peripheral nerves, possibly as a result of axonal swelling and attenuation with secondary segmental

demyelination, structural changes at the nodes of Ranvier, reperfusion-induced segmental demyelination, and inactivation of nodal voltage-gated sodium channels by regional anoxia.389,477,478,586 Nevertheless, a predominance of demyelinating features should suggest an alternate diagnosis, because biopsy-proven cases of vasculitic neuropathy meeting electrodiagnostic criteria for a demyelinating neuropathy are vanishingly rare. Needle electromyographic examination reveals fibrillation potentials in more than 80% of patients and decreased recruitment of motor unit potentials (MUPs) in clinically weak muscles.118 At presentation, most vasculitic neuropathies are too nascent for remodeling of motor unit morphology to have occurred, but long-duration MUPs are observed in a minority of patients (those with long-standing disease or a preexisting neuropathy). Short-duration, myopathic MUPs are found in proximal muscles in approximately 40% of patients with systemic disease.67,638,654

2368

Diseases of the Peripheral Nervous System

Table 105–12. Laboratory Studies in Suspected Vasculitic Neuropathy Study

Associated Conditions

Nonspecific Measures of Inflammation, Autoantibody Production, & Organ Involvement Complete blood count Anemia: systemic vasculitides, ~50%; NSVN, ~30% Leukocytosis: systemic vasculitides, ~70%; NSVN, ~20% Thrombocytosis: systemic vasculitides, ~60%; NSVN, ~20% Erythrocyte sedimentation rate Mild q in ~60% of isolated PNS vasculitis; marked q in ~85% of all systemic vasculitides Chemistry panel q BUN, creatinine: renal involvement q LFTs: liver involvement; hepatitis B/C q Glucose: diabetes mellitus Urinalysis Proteinuria, hematuria, RBC casts: renal involvement, connective tissue disease, malignancy Chest films Pulmonary involvement, sarcoidosis, malignancy Antinuclear antibody  in  90% of connective tissue diseases, 20%–30% of primary systemic vasculitides and NSVN Serum complement p in 50%–90% SLE, RV, EMC, Sjögren’s; ~25% classic PAN; ~10% NSVN Rheumatoid factor  in 60%–90% RV, EMC, Sjögren’s; 30%–50% WG, PAN, MPA, CSS; ~15% NSVN Cryoglobulins  in EMC/hepatitis C, plasma cell dyscrasias, lymphoproliferative disorders, connective tissue diseases, hepatitis B, other chronic infections Eosinophil count q in CSS (MPA, PAN, WG), eosinophilia-myalgia syndrome, eosinophilic fasciitis, hypereosinophilic syndrome, some infections and drug reactions More Specific Studies Frequently Useful Protein immunofixation/electrophoresis Anti-dsDNA Anti-Ro and -La Anti-Sm Anti-Scl 70 and -centromere Hepatitis B serologies Hepatitis C antibodies or RNA Glycated hemoglobin 2-Hour oral glucose tolerance test HIV antibodies HTLV-1 antibodies Lyme antibodies c-ANCA (anti–proteinase 3) p-ANCA (anti-myeloperoxidase) Angiotensin-converting enzyme DNA for HNPP CSF

Monoclonal protein: MGUS, lymphoma, leukemia, MPD, amyloidosis  in SLE  in SLE and Sjögren’s  in SLE  in scleroderma  Surface antigen in hepatitis B–associated PAN  Antibodies and/or circulating RNA in hepatitis C–associated EMC q in diabetes mellitus q in diabetes mellitus  in HIV infection, CMV neuropathy  in HTLV-1 infection  in Lyme disease  in WG  MPA, CSS  in MPA, CSS  WG q in sarcoidosis Deletion of 17p11.2 (gene for PMP22) in HNPP q Protein in ~30% vasculitic neuropathy;  PCR or Ab index: CMV, Lyme, HIV, HTLV-1;  cytology: neoplastic meningitis

Ab  antibody; BUN  blood urea nitrogen; c-ANCA  cytoplasmic antineutrophil cytoplasmic autoantibody; CMV  cytomegalovirus; CSF  cerebrospinal fluid; CSS  Churg-Strauss syndrome; dsDNA  double-stranded DNA; EMC  essential mixed cryoglobulinemia; HIV  human immunodeficiency virus; HTLV-1  human T-lymphotropic virus 1; HNPP  hereditary neuropathy with liability to pressure palsies; LFTs  liver function tests; MGUS  monoclonal gammopathy of undetermined significance; MPA  microscopic polyangiitis; MPD  myeloproliferative disorder; NSVN  nonsystemic vasculitic neuropathy; PAN  polyarteritis nodosa; p-ANCA  perinuclear antineutrophil cytoplasmic autoantibody; PNS  peripheral nervous system; PCR  polymerase chain reaction; PMP22  peripheral myelin protein 22; RBC  red blood cell; RV  rheumatoid vasculitis; SLE  systemic lupus erythematosus; WG  Wegener’s granulomatosis.

Neuropathies with Systemic Vasculitis

Electrodiagnostic findings that support the presence of a multifocal, axon-loss process include (1) a more than twofold difference in amplitude between right and left motor or sensory compound action potentials of homologous nerves; (2) a significant amplitude reduction in one but not other nerves in the same limb; (3) a low-amplitude response in an arm nerve while amplitudes are normal in the legs; (4) active partial denervation (fibrillation potentials and MUP dropout) in any muscle not matched by similar findings in its contralateral homologue; (5) active partial denervation in upper but not lower limb muscles; and (6) active partial denervation in any proximal muscle.488,498 A three- or four-limb study is mandatory when assessing any patient with an unexplained polyneuropathy for an underlying multifocal process. Considering the many nonvasculitic causes of neuropathy in patients with vasculitis or other connective tissue diseases (e.g., nerve entrapment, drug effect, organ failure, critical illness, inflammatory demyelination) and the broad differential diagnosis of asymmetrical neuropathy (see Table 105–11), most patients with a suspected vasculitic neuropathy should undergo a nerve biopsy. The sural nerve is most commonly biopsied, assuming clinical or electrophysiologic involvement. However, based on reports of positive muscle biopsies in 50% of patients with PAN, combined nerve and muscle biopsies have emerged as a diagnostic alternative.196 In one such technique, superficial peroneal nerve (SPN) and peroneus brevis muscle (PBM) specimens are obtained through a single incision in the lower third of the leg.118,544 Although nerve is more commonly diagnostic than muscle in cases of suspected vasculitic neuropathy (especially NSVN), muscle biopsy increases the yield in patients with systemic vasculitis.109,118,120 Other procedures for single-incision combined muscle/nerve biopsies have

2369

been described (e.g., sural nerve/soleus muscle), but none has yet been applied to a series of vasculitis patients.315 Some advocate combined sural nerve and anterior tibialis muscle biopsies.109 In selected cases, superficial radial or intermediate femoral cutaneous nerve biopsy may be preferred. In patients with clinically suspected vasculitic neuropathy, definite pathologic evidence of vasculitis is found in only 20% of sural nerve biopsies109,291,516 and 30% of SPN/PBM biopsies,118 leaving 70% to 80% of patients undiagnosed. For these patients who lack definite vasculitic changes on biopsy, diagnostic criteria for clinically probable vasculitic neuropathy have been proposed (Table 105–13).118 By applying these criteria, a diagnostic sensitivity of 60% for SPN/PBM biopsy was derived.118 Analogous sensitivities for sural nerve and SPN/PBM biopsies, extracted from studies in which similar pathologic definitions of vasculitis were employed, are catalogued in Table 105–14. The pooled sensitivity of SPN/PBM biopsy is 64% versus 56% for sural nerve biopsy. A randomized, prospective trial is necessary to determine which of these two procedures has the greater diagnostic yield in PNS vasculitis. The pathologic diagnosis of vasculitis requires an inflammatory infiltrate within the vessel wall and signs of vascular destruction such as fibrinoid necrosis, hemorrhage, or endothelial cell disruption (Fig. 105–7).317,435 Perivascular or transmural inflammation alone, without accompanying structural damage, is not diagnostic (Fig. 105–8D). Many workers augment pathologic sensitivity by defining “probable,” “suspicious,” or “healed” vasculitis in specimens lacking definite necrotizing changes. Pathologic alterations supportive of active or healed vasculitic neuropathy include wallerian-like degeneration, asymmetrical or multifocal nerve fiber loss, thickening of

Table 105–13. Diagnostic Criteria for Clinically Probable Vasculitic Neuropathy in Patients Lacking Biopsy-Proven Necrotizing Vasculitis118 1. Clinical presentation typical for a vasculitic neuropathy: • Asymmetrical or multifocal, painful, sensorimotor neuropathy • Acute/subacute relapsing, progressive, or relapsing progressive course 2. Elevated sedimentation rate or other laboratory evidence of a systemic inflammatory state 3. Electrodiagnostic evidence of an active, asymmetrical, axonal, sensorimotor neuropathy 4. Suggestive neuromuscular pathology • Vascular thickening, narrowing or obliteration of the vascular lumen, thrombosis, periadventitial capillary proliferation, hemosiderin deposits, asymmetrical nerve fiber loss, or wallerian-like degeneration 5. Clinical response to immunosuppressive therapy 6. Clinicopathologic evidence of a systemic/secondary etiology • Concurrent condition known or suspected to predispose to vasculitis (connective tissue diseases, certain infections, certain drugs malignancies/paraproteinemias, cryoglobulinemia) • Simultaneous multiorgan, non–peripheral nerve involvement • Biopsy-proven vasculitis in other tissues Diagnosis of nonsystemic vasculitic neuropathy requires at least three of the first five criteria; criterion 6 also mandatory for systemic vasculitis.

2370

Diseases of the Peripheral Nervous System

Table 105–14. Diagnostic Sensitivity of Nerve Biopsy for Vasculitic Neuropathy Biopsy Site Superficial peroneal nerve/peroneus brevis muscle

Sural nerve

Reference Chia et al. (1996)

99

Collins et al. (2000)118 Collins et al. (2003)120 Combined total Dyck et al. (1987)165 Claussen et al. (2000)109 Hattori et al. (2002)286 Hattori et al. (2002)286 Collins et al. (2003)120 Combined total

Vasculitis Type*

Sensitivity

Systemic & NSVN

23/33 (70%)

Systemic NSVN Systemic & NSVN Systemic & NSVN Systemic & NSVN Churg-Strauss syndrome Microscopic polyangiitis NSVN Systemic & NSVN

15/25 (60%) 11/19 (58%) 49/77 (64%) 31/65 (48%) 27/45 (60%) 16/30 (53%) 21/26 (81%) 14/30 (47%) 109/196 (56%)

*NSVN  nonsystemic vasculitic neuropathy; Systemic  various systemic vasculitides.

vessel walls with narrowing or obliteration of the lumen, thrombosis with or without recanalization, epineurial capillary proliferation, perivascular hemosiderin deposits, focal perineurial necrosis and thickening, and injury neuroma (see Fig. 105–8).6,109,118,142,165,170,204,435,505,570,654 In one cohort survey of patients undergoing SPN/PBM biopsy for suspected vasculitic neuropathy, “suspicious” biopsies were defined by perivascular inflammation and at least one supportive pathologic alteration.118 Suspicious or positive biopsies had a sensitivity of 86% for vasculitis and a specificity of 85%. In this cohort, pathologic findings associated with necrotizing vasculitis were wallerian-like degeneration, asymmetrical nerve fiber loss, myofiber necrosis and regeneration, and vascular immune deposits (C3, IgM).117,118 Perls’ ferrocyanide test for iron deposits may be another sensitive and specific marker of vasculitic neuropathy.6 Peripheral nerve vasculitis has a predilection for epineurial vessels with diameters of 75 to 250 m.165,290,435 NSVN tends to involve smaller vessels ( 100 m) within this range.167 In certain conditions, however, vasculitis of endoneurial microvessels ( 50 m) can be observed. Microvascular changes in these disorders are typical necrotizing vasculitis, leukocytoclastic vasculitis, or, most commonly, non-necrotizing, lymphocytic “microvasculitis.” The differential diagnosis for endoneurial small vessel vasculitis includes infectious processes (e.g., HIV, CMV), paraneoplastic vasculitis, connective tissue diseases (e.g., RA, SLE, Sjögren’s syndrome), HSV, cryoglobulinemic vasculitis, and diabetic and nondiabetic lumbosacral radiculoplexus neuropathy (see Chapter 86).167–170,223,226,228,435,486,535 Active epineurial lesions are characterized by T cells and macrophages surrounding and invading vessel walls, with associated necrosis (Fig. 105–9).352,359,435,554 Affected fascicles show reduced myelinated nerve fiber density, ischemia-induced wallerian-like degeneration, and minimal demyelination/remyelination.505 Axonal degeneration

and nerve fiber loss are usually asymmetrical between fascicles because of the multifocal nature of vasculitis (see Fig. 105–8).204 Sectorial nerve fiber loss within fascicles is also typical. Large myelinated fibers are most susceptible to ischemia, but small myelinated and unmyelinated fibers can also be affected.204 In fulminant vasculitis, Schwann cells, fibroblasts, and all other cellular elements are lost. Direct immunofluorescent staining shows deposits of IgM, IgG, C3, and C5b-9 in epineurial vessel walls in 80% of patients (see Fig. 105–9).117,142,290,359,495 Ninety percent of patients with vasculitis undergoing SPN/PBM biopsy have vascular immune deposits in the nerve or muscle specimen, a finding fairly specific for vasculitis.117,565

TREATMENT The vasculitides are uncommon disorders that present with diverse clinical manifestations of variable severity. This clinical variability, combined with similarly heterogeneous etiologies, poorly understood pathogenic mechanisms, and nonuniform classification systems, has made it difficult to conduct large, controlled treatment trials in vasculitis. Therefore, recommendations for therapy have traditionally been grounded on retrospective studies and case series employing nonstandardized treatment protocols and definitions of disease. However, in the last 10 years new therapeutic strategies have evolved, thanks to development of widely applied classification systems, improved knowledge of disease processes, identification of prognostic factors, and formation of multicenter, collaborative study groups such as the European Vasculitis Study Group (EUVAS) and International Network for the Study of the Systemic Vasculitides (INSSYS).300,324 An increasing number of randomized, controlled trials (RCTs) has appeared, providing guidance for rational treatment of several vasculitic disorders. 322,371,520 Unfortunately,

Neuropathies with Systemic Vasculitis

2371

FIGURE 105–7 A, Superficial peroneal nerve biopsy specimen showing classic necrotizing vasculitis of epineurial artery, characterized by transmural mononuclear inflammatory cell infiltration and circumferential fibrinoid necrosis (epoxy resin embedded). B, Peroneus brevis muscle biopsy revealing more chronic vasculitic lesion of perimysial vessel, with intimal and adventitial thickening, disorganization of vessel wall, obliteration of lumen, vascular inflammation, and focal fibrinoid necrosis (hematoxylin and eosin). See Color Plate

treatment of vasculitic neuropathy has not been approached with nearly the same investigatory rigor as the systemic vasculitides. There are sparse data relevant to the treatment of any type of vasculitic neuropathy. Recommendations for the management of vasculitic neuropathy, therefore, depend largely on “clinical experience” and extrapolation from systemic vasculitis trials, an approach with questionable validity because therapy proven effective for one type of vasculitis may not yield similar benefits in other vasculitides. However, a ret-

rospective analysis of treatment responses in a large cohort of NSVN patients has provided the first controlled data concerning therapy of this condition.120 When treating a patient with vasculitis, several basic questions need to be addressed: 1. Is it possible to eliminate inciting antigens? 2. Is immunosuppressive therapy indicated, and, if so, what is the best regimen to induce remission? 3. What is the optimal regimen to maintain remission?

2372

Diseases of the Peripheral Nervous System

FIGURE 105–8 A, Superficial peroneal nerve specimen showing fascicle completely devoid of myelinated nerve fibers, accompanied by asymmetrically thickened perineurium and many epineurial neovessels, suggestive of healing vasculitis (hematoxylin and eosin; paraffin embedded). B, Sural nerve biopsy in patient with suspected vasculitic neuropathy, revealing supportive but nondiagnostic evidence of vasculitis in the form of asymmetrical myelinated nerve fiber loss within and between four fascicles (epoxy resin embedded). See Color Plate Figure continued on opposite page

4. How long should treatment be continued? 5. How should refractory patients be managed? 6. How can adverse treatment-related events be minimized?

Suggested first- and second-line treatment options for the major vasculitides, derived from an evidence-based review of the literature, are developed in the discussion that follows and summarized in Table 105–15.

Neuropathies with Systemic Vasculitis

2373

FIGURE 105–8 Continued C, Small perimysial artery with pronounced, noninflammatory intimal proliferation and narrowing of lumen, indicative of remote vascular insult, in peroneus brevis muscle specimen from patient with biopsy-proven nonsystemic vasculitic neuropathy (modified Gomori trichrome). D, Sural nerve biopsy showing epineurial artery surrounded and partially infiltrated by inflammatory cells, without fibrinoid necrosis, endothelial disruption, evident vascular remodeling, hemorrhage, or thrombus—a nonspecific finding in isolation (hematoxylin and eosin; paraffin embedded). See Color Plate

Antigen Removal Treatment of vasculitis begins with consideration given to removal of antigens that might be triggering the immunologic response. However, such inciting antigens are infrequently identified. This approach is feasible

only in vasculitis arising from drug exposures, cancers, or infections. Identification of a drug-induced vasculitis mandates immediate discontinuation of the drug; complete resolution of vasculitic symptoms usually follows within 1 to 4 weeks.621 For patients with severe, life-threatening

2374

Diseases of the Peripheral Nervous System

A

B manifestations, hemodialysis and combination immunosuppressive therapy are often implemented. No controlled trials of therapy have been reported for cancer-related (paraneoplastic) vasculitic neuropathy. Treatment decisions are thus based on case reports alone. Published cases of biopsy-supported paraneoplastic vasculitic neuropathy for which treatment information is available are summarized in Table 105–16, along with five of the authors’ unpublished cases. These reports reveal a high response rate (~85%), irrespective

FIGURE 105–9 A, Superficial peroneal nerve section stained for CD2 antigen (T-cell marker) in patient with clinically isolated peripheral nerve vasculitis, demonstrating focally intense T-cell infiltrate surrounding and invading small epineurial vessel (avidin-biotin-peroxidase). B, Direct immunofluorescent stain of superficial peroneal nerve biopsy in patient with rheumatoid vasculitis–associated neuropathy, revealing IgM deposits in two small epineurial vessels, with a third similarly sized vessel spared (fluoresceinconjugated anti-human IgM). See Color Plate

of treatment modality. Unlike paraneoplastic sensory neuronopathy, cancer-associated vasculitic neuropathy usually improves with anticancer therapy alone (surgery, radiation, or chemotherapy), presumably on the basis of reduced tumor antigen load.55,228,639,670 In patients for whom cancer therapy is not indicated, or for patients failing antineoplastic measures, trials with prednisone alone or combined with cyclophosphamide (CYC) are indicated. Combination therapy may be preferable, because all reported paraneoplastic vasculitic neuropathy

2375

Neuropathies with Systemic Vasculitis

Table 105–15. Summary of Treatment Options in Major Vasculitides Vasculitis Type

First-Line Treatment

Second-Line Options for Refractory Patients

NSVN

Induction (Standard Therapy): IV MP 15 mg/kg (1.0 g)/day for 3–5 days PRD 1.0 mg/kg/day, switched to qod after 2–4 wk Oral CYC 2.0 mg/kg/day Maintenance (after remission): Taper PRD over 6–12 mo Continue CYC for 7–12 mo, or Switch CYC to AZA 1.0–2.0 mg/kg/day

18–24 mo

• Replace oral CYC with pulse IV CYC 0.5–1.0 g/m2 q 3–4 wk 12 mo • IVIg 0.5 g/kg/day 4 days, then 0.5 g/kg/day q 3–4 wk 6–12 mo • Replace oral CYC with MTX 15–25 mg oral or IV q wk for 12–24 mo

Wegener’s granulomatosis

Induction (Standard Therapy): IV MP, CYC, and PRD as for NSVN Maintenance (after remission): Taper PRD over 6–12 mo; and 1) Continue CYC 2 mg/kg/day 12 mo, then stop or taper by 25 mg q 2–3 mo; or 2) Switch CYC to AZA 1.0–2.0 mg/kg/day

18–24 mo; or 3) Switch CYC to MTX 15–25 mg PO or IV q wk 18–24 mo

• Replace oral CYC with pulse IV CYC 0.5–1.0 g/m2 q 3–4 wk 12–24 mo • Replace CYC with MTX 15–25 mg oral or IV q wk 24 mo • IVIg (see NSVN) • PE 6–12 times • TMP/SMX 160/800 mg bid for WG

Classic PAN, MPA, and CSS

1) Patients with 1 poor-prognostic factors (creatinine  1.58 mg/dL; proteinuria  1 g/day; CNS, GI, or cardiac involvement): treat as Wegener’s granulomatosis 2) Patients with no poor-prognostic factors: PRD alone as in standard therapy

• See Wegener’s granulomatosis • INF- in treatment-refractory CSS

HBV-associated PAN

PRD 1 mg/kg/day 2 wk; PE 1–3/wk 10 wk INF- 3 million U tiw or lamivudine 100 mg/day until HBeAb seroconversion or 9–12 mo

• Higher doses of INF- (as in hepatitis B): 10 million U tiw or 5 million U/day • Famciclovir 500 mg tid • Adefovir dipivoxil 10 mg/day

HCV-associated cryoglobulinemic vasculitis

1) Severe disease manifestations: Induction: Oral CYC 2 mg/kg/day PRD 1.0 mg/kg/day, changing to qod after 2–4 wk, then tapering PE 1–3/wk 9 wk Maintenance (after remission): PegINF-2b 1.5 g/kg SC/wk and ribavirin 800 mg/day, or PegINF-2a 180 g SC/wk and ribavirin 1000–1200 mg/day 48 wk 2) Mild disease manifestations: PegINF-2b 1.5 g/kg SC/wk and ribavirin 800 mg/day, or PegINF-2a 180 g SC/wk and ribavirin 1000–1200 mg/day 48 wk PE 1–3/wk 9 wk

• See Wegener’s granulomatosis • Rituximab 375 mg/m2 IV/wk 4 wk

Ganciclovir: Induction—5 mg/kg IV q 12 h 3 wk Maintenance—1000 mg PO tid Start or maximize antiretroviral therapy

• Foscarnet: Induction—90 mg/kg IV q 12 h 3 wk Maintenance—90 mg/kg/day IV • PRD 0.5–1.0 mg/kg/day, tapered over 6–12 mo • PE 1–3/wk for 10 wk • IVIg (see NSVN)

Treat malignancy

• Standard therapy with PRD and CYC (see Wegener’s granulomatosis) • IVIg (see NSVN)

HIV-associated vasculitic neuropathy CMV-positive vasculitis CMV-negative vasculitis Cancer-associated vasculitic neuropathy

Table continued on following page

2376

Diseases of the Peripheral Nervous System

Table 105–15. Summary of Treatment Options in Major Vasculitides—Continued Vasculitis Type

First-Line Treatment

Second-Line Options for Refractory Patients

Temporal arteritis

PNS, CNS, or visual symptoms: PRD 1 mg/kg/day (60–100 mg/day) with slow taper after 4 wk No serious manifestations: PRD 0.5–1.0 mg/kg/day (20–60 mg/day) with slow taper after 2–4 wk

• IV MP 500–2000 mg/day 2–5 days • MTX 10 mg q wk added to PRD

ANCA  antineutrophil cytoplasmic autoantibody; AZA  azathioprine; CMV  cytomegalovirus; CNS  central nervous system; CSS  Churg-Strauss syndrome; CYC  cyclophosphamide; GI  gastrointestinal; HbeAb  hepatitis B e antibody; HBV  hepatitis B virus; HCV  hepatitis C virus; HIV  human immunodeficiency virus; INF-  interferon-; IV  intravenous; IVIg  intravenous immunoglobulins; MP  methylprednisolone; MPA  microscopic polyangiitis; MTX  methotrexate; NSVN  nonsystemic vasculitic neuropathy; PAN  polyarteritis nodosa; PE  plasma exchange; PNS  peripheral nervous system; PRD  prednisone; tiw  three times per week; TMP/SMX  trimethoprim/sulfamethoxazole; WG  Wegener’s granulomatosis.

patients treated with alkylating agents improved, in contrast to only five of nine patients treated with prednisone alone. Of the infectious vasculitides, antigen removal is especially important for those related to chronic viral infections, because immunosuppressive treatment may worsen the underlying infection.637 Many studies have shown that CS therapy in chronic HBV infection activates viral replication, increases serum HBV-DNA levels, amplifies expression of HBV genome and gene products, and accelerates hepatic damage and mortality.184,368,376 Therefore, the cornerstone of treatment for these patients is antiviral therapy. Complete eradication of HBV in chronically infected patients is generally not possible, so effective therapy is gauged by its ability to produce normalization of transaminases, sustained suppression of HBV-DNA, and stable hepatitis B e antigen (HBeAg) seroconversion.647 Standard therapy consists of a 4- to 6-month course of INF-, which has a response rate (sustained HBeAg seroconversion) of 33%, compared to 10% to 20% in untreated controls.647,663 A less expensive, better tolerated alternative is lamivudine, a nucleoside analogue that profoundly suppresses HBV replication.529 Controlled studies of lamivudine demonstrate a delayed, accumulating seroconversion rate of 20% after 1 year of treatment, 40% after 3 years, and 50% after 5 years.380 However, long-term use of lamivudine leads to the emergence of drug-resistant strains, whose incidence increases from approximately 20% after 1 year of therapy to approximately 70% after 5 years. Disease is still suppressed in most patients with these strains, but flares of hepatitis can occur. Adefovir dipivoxil is a newer agent that is effective against both wild-type and mutant HBV.405 In protocols designed by the French Vasculitis Study Group, patients with HBV-related PAN have been treated with antiviral agents, coupled with plasma exchanges (PEs) to remove circulating immune complexes and short courses of CS to control severe disease manifestations.266 Vidarabine, the antiviral agent employed in the initial

protocols, was later replaced by INF-. Based on uncontrolled data from this group, vasculitis is controlled in 80% of patients with this regimen. The same investigators and others have more recently used lamivudine, short-term CS, and PE to induce remission in 90% of patients with HBV-associated PAN.258,269 Hence, lamivudine is an acceptable alternative to INF- in these patients. Antiviral therapy is continued for 9 to 12 months or until seroconversion occurs. Antiviral therapy with INF- has also been successfully applied to HCV-associated EMC.163,190,208,369,636 As is true for chronic hepatitis B, CS stimulate viral replication and increase viremia in patients with chronic hepatitis C.139,398 There are also reports of immunosuppressive therapy– induced exacerbations of hepatitis and hepatic failure in HCV-infected patients,25,206 but this is rare.138,636 Four RCTs have shown significantly higher response rates with INF- than with no treatment in patients with HCVassociated cryoglobulinemia.139,189,378,440 Pooling data from these and several uncontrolled trials of INF- in EMC yields an overall short-term response rate of 70%.7,89,113,419,436 However, as a result of recurrent viremia, 85% of responders relapse when INF- is stopped. This situation mirrors that for the treatment of chronic hepatitis C itself, wherein INF- for 6 months produces a shortterm response in 40% to 50% of patients, but only a 10% to 15% sustained response rate.377 Treatment for 12 to 24 months increases the long-term response rate to 20% to 25%. Combining INF- with ribavirin and treating for 48 weeks yields a still higher sustained response rate of 40% to 50%.377 Pegylated interferon- (PegINF-) is derived from native INF- by covalent attachment of a polyethylene glycol moiety.345 Two forms of Peg INF- have been developed—PegINF-2a and PegINF-2b—with different pharmacokinetic and pharmacodynamic properties. Both have a longer half-life than unmodified INF- and can be administered once per week. PegINF-2a or PegINF-2b plus ribavirin has become the treatment of choice for chronic HCV infection based on two large RCTs

2377

Neuropathies with Systemic Vasculitis

Table 105–16. Paraneoplastic Vasculitic Neuropathy: Therapeutic Responses Report 381

Levine et al. (1967) Torvik and Berntzen (1968)626 Johnson et al. (1979)334 Patient 3 Tassin et al. (1980)618 Vincent et al. (1986)639 Case 1 Case 2

Tumor Type

Treatment

Neuropathy Response

Non–small cell lung CA Renal cell CA

Plasma exchange, penicillamine Prednisone

None None

Poorly differentiated lymphocytic lymphoma Waldenström’s macroglobulinemia

Chemotherapy

None

Plasma exchange, chlorambucil

Improved

Chemotherapy & surgical extirpation Chemotherapy, radiation therapy Corticosteroids Surgical extirpation, radiation therapy Cyclophosphamide

Improved

Non–small cell lung CA Hodgkin’s disease

Case 7 Gherardi et al. (1989)228

Immunoblastic lymphadenopathy Skull plasmacytoma

Oh et al. (1991)485 Hughes et al. (1994)312 Patient 1 Patient 2 Matsumuro et al. (1994)414 Younger et al. (1994)670 Blumenthal et al. (1998)55 Hatzis et al. (1998)287

Endometrial CA

Antoine et al. (1999)22 Patient 12 Patient 13 Aslangul-Castier et al. (2000)26 Zehnder et al. (2000)677 Personal observations Patient 1

Low-grade non-Hodgkin’s lymphoma Hodgkin’s disease Common bile duct CA Small cell lung CA Small cell lung CA Cholangiocarcinoma

Resolved in 18 mo Transiently improved Improved Nearly resolved in 15 mo

Corticosteroids Corticosteroids Prednisolone, cyclophosphamide Chemotherapy Chemotherapy Methylprednisolone, cyclophosphamide

Resolved Improved Improved Much imporved Resolved in 7 mo

Colon adenocarcinoma Tongue epidermoid Chronic myelomonocytic leukemia Colon adenocarcinoma

Corticosteroids Surgical extirpation Prednisone, cyclophosphamide Immunosuppressives

Improved to near-normal Improved Improved Improved

Chronic lymphocytic leukemia

Prednisone Prednisone/cyclophosphamide IVIg & surgical extirpation Surgical extirpation Prednisone Prednisone Surgical extirpation IVIg Prednisone Prednisone/cyclophosphamide

None Resolved over 1 yr Improved Progressed Improved Progressed Died Progressed Progressed Stabilized/improved (relapsed with taper)

Patient 2 Patient 3

Papillary thyroid CA Papillary thyroid CA

Patient 4

Colon adenocarcinoma

Patient 5

Small cell lung CA

Improved

CA  carcinoma; IVIg  intravenous immunoglobulins.

(one employing PegINF-2a and the other PegINF-2b) that demonstrated a significantly improved sustained virologic response of approximately 55% with this combination compared to INF- plus ribavirin.202,403 Thus antiviral therapy for HCV-related cryoglobulinemic vasculitis should probably comprise Peg-INF-2a/2b plus ribavirin.81 Unfortunately, the neuropathy associated with mixed cryoglobulinemia appears more refractory to antiviral

therapy than other disease manifestations (only 30% to 40% of patients improve with INF-).7,80,89,189,230 To further complicate matters, INF- can exacerbate a preexisting vasculitic neuropathy63 or produce a new neuropathy, including a sensorimotor, multifocal neuropathy typical of vasculitis500; distal, axonal, sensorimotor polyneuropathy511; acute, diffuse, axonal motor neuropathy466; chronic inflammatory demyelinating neuropathy411; acute autonomic and sensory neuropathy319; trigeminal

2378

Diseases of the Peripheral Nervous System

sensory neuropathy519; and ischemic optic neuropathy.268 INF- can also exacerbate other manifestations of the vasculitis, probably as a result of its immunomodulatory or antiangiogenic properties.105 In view of the limited efficacy of INF- in treating cryoglobulinemic neuropathies and the potential for causing PNS complications, immunosuppressive therapy with CS, CYC, and PE should be considered for all severe neuropathies and for mild neuropathies not immediately stabilized by INF-. This regimen was standard for EMC prior to discovery of the HCV association163,220,369 and was often effective in relieving PNS manifestations.631 If immunosuppressive therapy is used, antiviral treatment should be started once remission has been induced, because INF- may suppress the viremia associated with CS.139 Rituximab, a humanized murine monoclonal antibody (mAb) directed against CD20, which selectively depletes B cells, is an alternative to standard immunosuppression in patients resistant to INF-.551,674 Few reliable data are available on treatment of HIVassociated vasculitis. Vasculitis during HIV infection may be due to the HIV organism itself, immune complex deposition, altered immune responses, or opportunistic agents such as CMV. For treatment of CMV, an initial 3-week course of ganciclovir (5 mg/kg IV every 12 hours) or foscarnet (90 mg/kg IV every 12 hours) is followed by lifelong suppressive therapy with lower doses. Almost all patients with CMV-related vasculitic neuropathy respond to this regimen, but more than half later relapse in the PNS or CNS during maintenance therapy.535,542 For non–CMV-related HIV-associated vasculitis, treatment patterned after that in HBV-associated PAN has been proposed, employing antiviral therapy combined with PE.232 The efficacy of highly active antiretroviral therapy has not yet been investigated. Corticosteroid and cytotoxic agents must be used cautiously in patients with HIV infection because of the risk for worsening immunosuppression. Nevertheless, prednisone is the most commonly reported modality in HIV vasculitis, with an 80% response rate.71,95,194,413,566,633 However, interpretation of this and other uncontrolled data is confounded by the tendency of many HIV vasculitic neuropathies to spontaneously remit.592

Immunosuppressive Therapy Despite the intuitive appeal of antigen reduction strategies, no inciting antigen can be identified in the great majority of vasculitides, necessitating the use of immunosuppressive therapy. Immunosuppressive agents are also commonly used in combination with antigen reduction strategies to control severe manifestations of vasculitides related to persistent antigens. Standard Combination Therapy The traditional immunosuppressive regimen used for most primary or secondary systemic necrotizing vasculitides

involves a combination of CS and CYC and is referred to as “standard therapy.” This protocol was originally developed for the treatment of WG,183 a disease for which CS alone have limited efficacy.306 Although “standard,” the regimen was never validated in a RCT. Treatment is initiated with prednisone 1 mg/kg/day and oral CYC 2 mg/kg/day. For fulminant disease, higher doses of CYC (4 to 5 mg/kg/day), prednisone (2 mg/kg/day), or intravenous (IV) methylprednisolone (15 mg/kg or 1 g/day) are employed for the first 3 to 5 days, although there is no controlled evidence to support this practice. Daily prednisone is continued for 2 to 4 weeks and then tapered to alternate-day dosing (1 mg/kg qod) over 1 to 2 months. In NSVN, conversion to alternate-day prednisone can usually be abruptly implemented, without risk of disease exacerbation. The alternate-day CS regimen is maintained until stability of the neuropathy and other disease manifestations is assured, a process typically requiring 2 to 3 months. Prednisone is then slowly tapered over 6 to 12 months at a rate of 5 mg every 2 to 3 weeks. Relapses that occur during tapering mandate repeat pulsing with IV methylprednisolone or oral prednisone. Cyclophosphamide is continued for 1 year after remission and then stopped or tapered in 25-mg decrements every 2 to 3 months. Standard therapy has established efficacy in WG by comparison with natural history controls.302,650 It was also shown to be superior to CS monotherapy in a cohort of patients with ANCA-associated vasculitides (primarily MPA).303,462 In this study, patients treated with CS and CYC had a significantly higher remission rate, lower relapse rate, and lower mortality rate than patients receiving CS alone. The role of cytotoxic therapy in PAN and CSS is less clear. Many patients with these conditions are effectively managed with CS monotherapy.264 Cohort surveys and RCTs have shown no overall difference in survival and relapse rates between PAN and CSS patients treated with CS alone or CS plus CYC.217,262 However, when stratified into poor- and good-prognosis groups according to a Five-Factor Score (FFS) (creatinine  1.58 mg/dL, proteinuria 1 g/day, GI involvement, CNS involvement, and cardiomyopathy), survival rate is significantly prolonged in poor-prognosis patients (FFS 1) treated with combined CYC and CS.217 Therefore, initial treatment of poor-prognosis PAN and CSS should include both CYC and CS. For good-prognosis patients (FFS  0), treatment with CS alone is usually sufficient. TA is another vasculitis for which CS alone remains the mainstay of therapy.301,372,469,546 Prednisone rapidly controls the most common symptoms of TA within 24 to 72 hours. Cohort surveys and studies with natural history controls have also shown that CS are effective in preventing blindness.11,53,538 Although no RCT has compared starting doses of prednisone in TA, there is emerging evidence from observational studies that doses less than

Neuropathies with Systemic Vasculitis

60 mg/day may be adequate in patients free of visual, CNS, or other major organ manifestations.42,461,471 Most experts now recommend an initial prednisone dose of 0.5 to 1.0 mg/kg/day in uncomplicated cases.301,372,469,546 Patients with visual loss, CNS symptoms, or other serious organ dysfunction are treated with higher prednisone doses (60 to 100 mg/day) or pulse IV methylprednisolone. Several RCTs of MTX as a CS-sparing agent in TA have produced mixed results.300,339,597 Thus MTX’s role in the treatment of TA remains unestablished. No controlled study has addressed treatment of TA-related neuropathies, but case reports and series suggest that more than 90% of these neuropathies respond to CS monotherapy.73,90,235,468,470,528 An important but largely overlooked question concerns the appropriate treatment of patients with vasculitic neuropathy. There are no controlled data on treatment of neuropathy associated with any type of systemic vasculitis, apart from one small observational study in CSS-associated neuropathy.285 Because multiple mononeuropathy is not an adverse prognostic factor for survival in PAN, CSS, and MPA,114,198,267 combination therapy might not be necessary for patients with these vasculitides who manifest a neuropathy without renal, CNS, cardiac, or GI involvement. Nevertheless, even in good-prognosis disease, CS alone are not always sufficient to halt progression of the neuropathy.305 Thus some investigators advocate combination therapy for all vasculitic neuropathies, except in mild, indolent cases.116 Many others favor CS monotherapy.165,249,295,539 However, these recommendations lack evidentiary support and are based on extrapolation from systemic vasculitis trials or clinical experience. For NSVN, controlled data from a retrospective cohort survey are now available to assist with management decisions.120 In this study, long-term treatment outcomes were analyzed for all NSVN patients diagnosed and treated at one institution over 20 years. The key finding was that combination therapy (generally CS and CYC) is more effective than CS monotherapy in inducing remission (95% vs. 61%) and improving disability in NSVN, with trends toward reduced relapse rate (29% vs. 59%) and chronic pain (44% vs. 71%). These results are supported by data extracted from another NSVN series in which disability scores improved in 91% of patients administered combination therapy and only 55% receiving CS alone.142 The increased efficacy of combination therapy in NSVN parallels the aforementioned observational studies in ANCA-associated vasculitides. In the cited NSVN study, 60% of patients achieved remission with CS alone. Unfortunately, no clinical or laboratory variable predicted CS responsiveness. In particular, mild neuropathies were no more likely to respond than severe ones, and the use of IV methylprednisolone did not influence outcome (consistent with a previous report on CSS-associated vasculitic neuropathy).285 Thus, if a decision is reached to

2379

treat NSVN with prednisone alone, close follow-up is mandatory, with CYC added for any signs of neuropathy progression. The NSVN study also demonstrated that longer (6 months) courses of CYC are more effective in preventing relapses than shorter courses, a finding concordant with standard therapy for WG, in which oral CYC is continued for at least 12 months after remission is achieved.120,302 However, long-term follow-up studies of standard therapy–treated WG and NSVN have revealed a high incidence of drug-related morbidity.120,259,302,522 For example, in one series of 158 WG patients, treatment-associated side effects included serious infections (46%); CS-induced cataracts (21%), fractures (11%), diabetes (8%), and avascular hip necrosis (3%); and CYC-induced ovarian failure (57%), hemorrhagic cystitis (43%), bladder cancer (3%), and myelodysplasia (2%).302 There was a 2.4-fold increased malignancy rate in the treated patients, with a 33-fold increase in bladder cancers and 11-fold increase in lymphomas. Other studies have confirmed an increased risk of bladder, skin, and hematologic malignancies in WG patients receiving combination therapy.362,657 These sobering findings have provided a strong impetus to reduce patient exposure to CYC and, thereby, limit the occurrence of CYC-related toxicity and infections. Modifications of the Standard Regimen The perceived need for a less toxic “standard” regimen has prompted many proposed modifications.180,322,371,520 One approach has been to replace continuous oral CYC with intermittent IV pulses (0.35 to 1.0 g/m2 body surface area or 15 mg/kg every 1 to 4 weeks) for up to 2 years.525 This strategy is designed to decrease the cumulative CYC dose and minimize the risk of bladder toxicity, myelosuppression, and cancer. Several RCTs have shown that pulse CYC is as effective as continuous oral CYC in inducing remission in primary systemic vasculitides, with similar mortality rates and decreased risk of infections and CYC-related side effects, but at the expense of a higher relapse rate.9,145,216,259,289 One way to circumvent the relapse problem may be to administer more sustained (1 year) pulses, a tactic under investigation. A large RCT of pulse versus continuous oral CYC in ANCA-associated vasculitis is ongoing through the EUVAS.324 Until the results of this study become available, pulse CYC should remain a second-line option for patients with vasculitic neuropathy. A second approach to avoiding CYC-related toxicity is to completely eliminate CYC from the induction regimen, replacing it with a safer agent such as MTX. Three uncontrolled trials of MTX combined with CS in non–lifethreatening WG showed that MTX (15 to 25 mg/week) and prednisone (20 to 60 mg/day) are effective in inducing remission in 75% to 80% of patients, but with a high

2380

Diseases of the Peripheral Nervous System

(58%) relapse rate.146,373,591,604 Preliminary results from a 12-month EUVAS RCT comparing MTX /CS versus CYC/CS for induction and maintenance therapy of patients with MPA and WG lacking renal involvement revealed similar remission rates in the two groups, but a significantly delayed time to remission and higher relapse rate (74%) in the MTX limb.147 Further study with a longer period of immunosuppression is required. A third avenue to limiting CYC-related toxicity involves switching from oral CYC to an alternate immunosuppressive agent once remission has been induced by standard therapy, typically in 3 to 6 months.322,371,520 Azathioprine (AZA) is the most widely used remissionmaintaining drug, at doses of 1 to 2 mg/kg/day. Many uncontrolled studies have suggested that AZA may be effective in maintaining remission in patients with WG, MPA, and PAN, with an overall relapse rate of approximately 40% (compared to ~50% for standard therapy–treated WG).9,243,328,657 Prolonging AZA maintenance from 1 to 2 years may reduce the relapse rate to approximately 10%.509 One 18-month RCT compared CYC to AZA for maintenance of remission in ANCA-associated vasculitis. The study showed no differences between the two groups in relapse rate (15%), disease activity, organ damage, and incidence of adverse effects.323 The investigators concluded that AZA was as effective as continued CYC in maintaining remission in ANCA-associated vasculitis. Although there is no information on AZA in vasculitic neuropathies, the implication for NSVN is that a switch from CYC to AZA might be considered after remission has been induced with standard therapy, especially for patients poorly tolerant of CYC. Investigation of MTX as a CYC-sparing drug for maintenance therapy of systemic vasculitis has been limited to WG. Two uncontrolled studies demonstrated that patients with WG can be safely and effectively maintained on weekly MTX (15 to 25 mg or 0.3 mg/kg for up to 2 years) after remission is induced with CS and CYC.148,374 However, the relapse rate in the larger study was 52%, auguring caution for use of MTX as a remissionmaintaining agent in vasculitic neuropathy. Data are sparse on the use of other immunosuppressive agents for maintenance therapy in systemic vasculitides, although cyclosporine,288 mycophenolate mofetil,475 and leflunomide433 have all shown promise in small studies of patients with ANCA-associated vasculitis. A EUVAS-sponsored, 48-month RCT comparing mycophenolate mofetil to AZA for remission therapy in ANCA-associated vasculitis is ongoing. Yet another strategy for minimizing drug toxicity is to more rapidly taper CS. Modified CS dosing regimens have been published, but none is the focus of ongoing RCTs.520 For example, in most of the EUVAS-sponsored protocols for ANCA-associated vasculitis, prednisolone is started at 1.0 mg/kg/day but then rapidly tapered to 0.75 mg/kg/day

after 1 week, 0.5 mg/kg/day after 2 weeks, 0.4 mg/kg/day after 4 weeks, 0.25 mg/kg/day after 10 weeks, and 10 mg/day after 5 months, with more slowly reduced doses continued to 18 months.323,324 Another protocol employs prednisone 1 mg/kg/day for the first month, the alternateday dose then being tapered by 10 mg/week until alternate-day dosing is achieved (1 mg/kg qod). Tapering then continues apace until prednisone is stopped by the end of the fourth or fifth month.180 The usefulness of these protocols in patients with vasculitic neuropathy remains to be determined, but they seem reasonable alternatives in patients intolerant of prednisone. Other Immunosuppressive Strategies The search for less toxic therapies in the vasculitides has also led to consideration of nonpharmacologic modalities such as PE and intravenous immunoglobulin (IVIg). These techniques hold special interest for vasculitides with presumed humoral immunopathogenic mechanisms. Plasma Exchange. Plasma exchange has been tried in many types of vasculitis since the 1970s, under the theory that the procedure removes pathogenic circulating immune complexes, antibodies (including ANCAs), inflammatory mediators, and soluble cellular adhesion molecules.322,622 It has been investigated as an initial remission-inducing intervention, both alone and in combination with other agents, and in patients with refractory disease. The literature is replete with reports on PE as an induction modality in patients with vasculitis, but almost all are uncontrolled case reports or small series. Only a few RCTs have been performed. Two studies of 140 patients with PAN, MPA, or CSS randomized to treatment with drug therapy alone (prednisone or prednisone and IV CYC) versus drug therapy plus PE showed no significant difference between the two groups in initial disease control, relapse rate, long-term recovery, and survival.261,265 The investigators concluded that PE was of no added benefit in the initial treatment of patients with these vasculitides. Nevertheless, PE is still used in severe cases of vasculitis, especially those with major renal or pulmonary involvement.180,322 One small RCT showed that dialysis-dependent patients with rapidly progressive glomerulonephritis secondary to ANCA-associated vasculitides were more likely to recover if treated with PE.510 A larger RCT of adjunctive PE versus IV pulse methylprednisolone in patients with severe glomerulonephritis secondary to WG or MPA showed a benefit in favor of PE for the end point of renal independence at 3 months.215 Many case reports and small series in the 1970s and 1980s described the use of PE in patients with mixed cryoglobulinemic vasculitis (to reduce the cryocrit). Pooled results from reports in which PE was used alone or combined with low doses of CS reveal an approximately

Neuropathies with Systemic Vasculitis

80% response rate,48,187,200,244,424,584 including the PNS manifestations.23,187,458 Mixed cryoglobulinemic patients have also been successfully treated with PE combined with high-dose CS, immunosuppressive agents, and antiviral drugs.163,200,220,369,631 Rheumatoid vasculitis was treated with PE in many uncontrolled trials in the 1980s, with 80% to 90% response rates, but almost always combined with prednisone and immunosuppressive drugs, confounding the results.571,661 There are rare reports of PE as initial monotherapy of CSS-associated multiple mononeuropathy,320 HIV-related vasculitic neuropathy,399,606 hepatitis B–associated PAN,94 PAN-induced visual loss,552 and severe HSP.284 Harder to interpret are uncontrolled trials of PE combined with other modalities for initial therapy of HBVassociated PAN and HIV-related vasculitis.232,266 Although never studied in a controlled fashion, PE is also sometimes used as a rescue agent in patients with various types of vasculitis refractory to combination therapy.54,201,313 In summary, PE is not recommended as initial monotherapy for any form of vasculitis, with or without neuropathy. However, it is a reasonable adjunctive intervention for patients with severe rapidly progressive glomerulonephritis, alveolar hemorrhage, viral-induced vasculitis, cryoglobulinemic vasculitis, severe RV, severe HSP, or any treatment-refractory systemic vasculitis. Vasculitic neuropathies associated with these conditions generally improve in concert with the non-neurologic manifestations. There are no data on the use of PE in NSVN. Intravenous Immunoglobulin. In vasculitis, IVIg has been studied almost exclusively as rescue therapy for patients with refractory or relapsing disease.14,191,218,329,383,526,566 Most reports have described single cases or small series of patients with a wide variety of vasculitides refractory to CS monotherapy or combined CS/CYC. In these uncontrolled studies, 80% of patients improved and 40% attained a complete remission. Scheduled follow-up doses were important in reducing the relapse rate. The only RCT of IVIg in vasculitis involved 34 patients with ANCA-associated vasculitis who failed therapy with CS and CYC or AZA.327 Patients receiving IVIg (0.4 mg/kg/day for 5 days) had a response rate of 82% compared to 35% for the placebo group, a significant difference. Neuropathies improved in five of seven IVIg-treated patients. However, the effect was often short lived, with 36% of responders relapsing after 3 months, indicating a need for repeat dosing. Based on these reports, IVIg can be considered a second-line option for patients with vasculitic neuropathy refractory to standard therapy. Other Agents for Patients Refractory to Conventional Therapy. Cyclosporine has been used successfully in patients with WG, CSS, EMC, and HSP who were unresponsive to or intolerant of conventional therapy, but

2381

the evidence is confined to a few case reports and one small series.13,36,310,422 Several reports suggest that etoposide (VP-16) may be another highly effective agent in patients with WG who have failed treatment with CS and CYC.452,497 Interferon- has been used successfully as a rescue agent for patients with CSS refractory to standard therapy.619 T-lymphocyte depletion using mAbs or antithymocyte globulin (ATG) is a newer therapeutic alternative that has produced long-term remissions in several patients with resistant disease.271,325,391 The mAb studies utilized a humanized anti-CD52 mAb (CAMPATH-1H), which lyses lymphocytes and produces a prolonged depletion of CD4 and CD8 T cells.325,391 Humanized anti-CD4 mAbs were used adjunctively in relapsing patients. Rabbit ATG transiently depletes circulating T cells and modulates T-cell function.271 The EUVAS is conducting an open, multicenter pilot study of ATG in patients with progressive or relapsing ANCA-associated vasculitis despite standard treatment.324 Many other investigational approaches have benefited small numbers of refractory vasculitis patients, including etanercept,605 infliximab (humanized mAb against TNF),39 mycophenolate mofetil,463 15-deoxyspergualin,52 stem cell transplantation,31 tryptophan immunoabsorption,173 humanized mAbs against CD18 (2 integrin),391 and rituximab.596 Whether any of these agents will prove useful in vasculitic neuropathy remains to be seen. Projected future therapeutic strategies for refractory vasculitis include targeting of additional pathogenic cytokines, adhesion molecules, and coactivation signal pathways.326,603

Limitation of Adverse Effects An important aspect of vasculitic neuropathy management concerns the need for a program to limit adverse events associated with immunosuppressive drugs. Although control of the vasculitis with aggressive immunosuppression is crucial to achieving a good outcome, it is no less important to minimize the confounding effects of treatment-related morbidity, such as medication toxicity, infection, or malignancy, which occur in up to 70% of systemic vasculitis259,302 and 65% of NSVN patients.120 Prednisone side effects that need to be addressed as soon as therapy is initiated include weight gain, hypertension, glucose intolerance, electrolyte imbalance, cataracts, glaucoma, mood alterations, insomnia, osteoporosis, myopathy, and avascular necrosis of the femoral head and other bones (Table 105–17).68,449 All patients should begin a strict low-calorie, reduced sodium, low–simple sugar diet, with tracking of weight and blood pressure at every visit. Serum electrolyte and glucose determinations are critical during the early phases of treatment, but should be continued periodically throughout the treatment course. Ocular examinations, to include intraocular pressure

2382

Diseases of the Peripheral Nervous System

checks, should be performed every 6 to 12 months, with therapy initiated for elevated intraocular pressures. Patients should undergo tuberculin skin testing before therapy is begun, with positive responders started on antituberculous therapy. The use of short-term sedativehypnotics and antidepressants may be necessary to control insomnia and mood swings, which can interfere with patient compliance. Osteoporosis is one of the most insidious and serious complications of CS treatment.4,15,111,172 Glucocorticoidinduced bone loss is greatest in the first 6 to 12 months of therapy, but continues at a rate two to three times normal under chronic treatment.4,111 The reduced mobility seen in many patients with vasculitic neuropathy also contributes to bone loss. Fifteen percent of all patients and 20% of postmenopausal women suffer a vertebral compression fracture within 1 year of starting CS therapy.3,649 Bone mineral density should be determined before therapy is

begun and at least yearly as long as therapy continues. If bone loss occurs at a rate greater than 3% per year, osteoporosis treatment will need to be augmented.4 Laboratory tests to exclude secondary causes of osteoporosis are recommended (see Table 105–17). Supplementation with calcium (1000 to 1500 mg/day) and vitamin D (800 to 1000 IU/day) or calcitriol (0.25 to 0.50 g/day) should be started immediately in all patients.16,450 Postmenopausal women should be considered for estrogen replacement therapy under the direction of their primary physician, gynecologist, or endocrinologist, provided there are no contraindications. Similarly, testosterone replacement therapy should be used in hypogonadal men, assuming there is no prostate cancer or other contraindication. Bisphosphonates such as etidronate (400 mg/day for 14 days, repeated every 3 months), alendronate (10 mg/day or 70 mg/week), or risedronate (5 mg/day) inhibit bone resorption and have been shown to both prevent and treat

Table 105–17. Corticosteroids: Adverse Effects and Management Side Effects Common (10%)

Uncommon ( 10%)

Osteoporosis/fractures Fluid retention/edema Hypertension Hyperglycemia Menstrual irregularities Growth suppression Increased appetite; weight gain Suppression of hypothalamic-pituitary-adrenal axis Hirsutism Skin atrophy; bruising Impaired wound healing Acne Cataracts Moon facies; truncal adiposity Susceptibility to infections Insomnia Psychiatric reactions Nausea

Avascular necrosis of hips and other joints Tendon rupture Epidural lipomatosis Myopathy Hypokalemia Glaucoma Opportunistic infections Bowel perforation Encephalopathy

Monitoring Baseline

Follow-Up

CBC, electrolytes, glucose, glycated hemoglobin Blood pressure Eye examination PPD, chest radiograph, hepatitis B, hepatitis C, HIV Osteoporosis risk assessment • Calcium, phosphorus, creatinine, alkaline phosphatase, 25-OH vitamin D, testosterone level, TSH, serum protein electrophoresis, 24-hour urinary calcium • Bone densitometry

CBC, electrolytes, glucose Blood pressure Eye examination (yearly) Bone densitometry (yearly)

Table continued on opposite page

2383

Neuropathies with Systemic Vasculitis

Table 105–17. Corticosteroids: Adverse Effects and Management—Continued Preventative Measures Diet Calorie restricted, low sodium, low simple sugar Adequate calcium Eliminate tobacco and alcohol Infections No live vaccines Administer influenza and pneumococcal vaccines Pneumocystis carinii prophylaxis • With concomitant immunosuppressives • Trimethoprim/sulfamethoxazole 160/800 mg tiw Osteoporosis Prophylaxis Regular weight-bearing, resistance exercises Calcium, 1000–1500 mg/day Vitamin D, 800–1000 IU/day, or calcitriol, 0.25–0.50 g/day Thiazide diuretics if urinary calcium  300 mg/day Consider hormone replacement in postmenopausal women Consider testosterone replacement in hypogonadal men Bisphosphonates • Cyclic etidronate, 400 mg/day 2 wk q 3 mo; or • Alendronate, 5–10 mg/day (or 35–70 mg/wk); or • Risedronate, 5 mg/day CBC  complete blood count; HIV  human immunodeficiency virus; PPD  purified protein derivative of tuberculin; tiw  three times per week; TSH  thyroid-stimulating hormone.

CS-induced osteoporosis.3–5,15,649 Bisphosphonates are the only osteoporosis medications demonstrated in a RCT to reduce CS-related vertebral fractures.5,649 They should be considered first-line therapy for prevention and treatment of CS-induced osteoporosis. Activity as tolerated by the patient’s overall status, and in particular resistance exercise producing skeletal loading, limits osteoporosis and reduces fracture risk.406 The principal side effects of chronic CYC exposure are bone marrow suppression, opportunistic infections, bladder toxicity, gonadal failure, alopecia, pulmonary toxicity, and increased risk of malignancy (Table 105–18).302,370,560 A complete blood count should be obtained every 2 weeks early in treatment and monthly as long as CYC is continued. CYC-induced leukopenia usually reaches its nadir in 7 to 14 days, with recovery occurring in 3 to 4 weeks.370 The drug should be withheld for any leukocyte count less than 3500/mm3 and reinstituted at a lower dose once the count climbs above 4000/mm3. Bladder toxicity is caused by acrolein, a CYC metabolite that accumulates in the bladder and irritates the mucosa.370,520 The risk of drug-induced cystitis—which usually presents as gross hematuria but can be asymptomatic—is reduced by administering oral CYC as a single dose in the morning and encouraging patients to drink at least 2 quarts of water per day to ensure frequent urination. For patients

treated with IV pulses of CYC, coadministration of mesna (a compound that binds and detoxifies acrolein) is recommended.370,522 These measures probably also reduce the risk of bladder cancer, which is predicted to occur in 16% of patients within 15 years of first exposure to continuous oral CYC.614 Following CYC treatment, periodic screening urinalyses and urine cytologies should be obtained at least semiannually, with cystoscopy performed for nonglomerular hematuria or abnormal cytology. In patients with a history of hemorrhagic cystitis, routine cystoscopy should also be performed every 1 to 2 years.370,614 Because of a significantly increased risk of Pneumocystis carinii pneumonia in patients on combination CS/CYC therapy, prophylaxis with trimethoprim-sulfamethoxazole (TMP/ SMX) is recommended.101,259,483 Some investigators recommend TMP/SMX for all patients receiving the equivalent of at least 20 mg of prednisone daily for more than 1 month, irrespective of additional immunosuppressive therapy.577

OUTCOME Although no direct comparative study has been performed, NSVN appears to have a more benign prognosis for mortality and morbidity than neuropathies associated

2384

Diseases of the Peripheral Nervous System

Table 105–18. Adverse Effects and Management: Cyclophosphamide Side Effects Infertility • Oligo/aspermia: 50%–90% • Amenorrhea: 30%–60% Bone marrow suppression: 30%–50% Susceptibility to infection: 30%–50% Hemorrhagic cystitis: 10%–50% GI distress (nausea, diarrhea): 20%–45%

Alopecia: 15%–35% Rash: 1%–5% Malignancy • Bladder: 16% at 15 yr • Myelodysplastic syndrome: 2%–8% • Lymphoid cancer: 1%–2% • Skin (rare) Monitoring

Baseline

Follow-up

CBC, liver functions, urinalysis Creatinine (reduced dosing in renal failure)

CBC Urinalysis (including q 6 mo after therapy) Cystoscopy q 1–2 yr (if history of hemorrhagic cystitis)

Chest radiograph, PPD, hepatitis B, hepatitis C, HIV Preventative Measures Hemorrhagic Cystitis

Infections

Oral dosing • Administer as single morning dose • Frequent voiding • Consume 2–3 L fluids per day Intravenous (IV) dosing • IV fluids: 1–1.5 L before and after infusion • IV mesna: 60%–100% of cyclophosphamide dose

No live vaccines Give influenza and pneumococcal vaccines Pneumocystis carinii prophylaxis (see Table 106–17) Neutropenia • Neutropenia precautions • Consider recombinant human granulocyte-colony stimulating factor

CBC  complete blood count; GI  gastrointestinal; HIV  human immunodeficiency virus; PPD  purified protein derivative of tuberculin.

with systemic necrotizing vasculitis. The estimated 5-year survival of 87% in one recent NSVN cohort was similar to that of a prior report in which 1 of 25 patients died over a median 2.8 years of follow-up.120,142 In a third large series, 3 of 20 patients died, but duration of postdiagnosis followup and survival analysis were not reported.165 In modern systemic vasculitis cohorts, 5-year survival rates range from 70% to 90%.1,217,415,522,657 Ten percent of NSVN patients die as a result of the vasculitis or its treatment,120 contrasting with 12% to 17% of patients with systemic vasculitis.1,217,302,522 Thus the available evidence suggests that NSVN is less likely to be lethal than the systemic vasculitides. In the largest NSVN cohort, only 6% of patients developed signs of vasculitic involvement in non-neuromuscular tissues with extended follow-up, and in these patients only skin was affected.120 In the only other study to address this issue, 37% of patients with a clinically isolated vasculitic neuropathy followed for a mean of 6 years developed systemic manifestations.539 However, this cohort was incompletely described and included patients with systemic disease at the outset

(22% were chronic carriers of hepatitis B surface antigen).544 The observation of clinically restricted peripheral nerve and cutaneous involvement complements reports of “cutaneous periarteritis nodosa” in the dermatologic literature, in which vasculitis is confined to skin and nerves.141,160 These and other reports272,553 suggest an association between some NSVN and localized necrotizing vasculitis of the skin. More common than systemic spread is the emergence of new neuropathy symptoms following a sustained treatment response. Relapses of neuropathy occur in 24% to 46% of responders,120,142,165 typically after therapy has been stopped or tapered to low-dose prednisone, an argument for longer courses of CYC or an alternate agent. The spectrum of final neurologic disability in long-term survivors of NSVN is surprisingly good for a disease known to produce severe axon loss: 14% with no symptoms, 67% with mild to moderate disability, 16% with moderately severe disability (assistance with walking), and 3% with severe disability (nonambulatory).120,142 Although most NSVN patients have relatively mild neurologic

Neuropathies with Systemic Vasculitis

deficits at the end of follow-up, chronic pain is a persistent concern in 60%.120 Outcomes in NSVN are somewhat better than those in systemic vasculitis–associated neuropathy: 9% asymptomatic, 56% with mild to moderate disability, 22% with moderately severe disability, and 14% nonambulant.118,285,290,447 The good functional recovery seen in most patients with vasculitic neuropathy suggests that the neurologic deficits are not exclusively the result of ischemia-induced axonal degeneration.120,142 Structural or functional conduction blocks likely also contribute, supported by the observation that many patients show improved motor function (especially in proximal muscle groups) during the first 1 to 2 months of follow-up, prior to the expected effects of axonal regeneration. The largest share of clinical recovery, however, is delayed for several months and continues for 1 to 2 years, consistent with axonal regeneration.

REFERENCES 1. Aasarod, K., Iversen, B. M., Hammerstrom, J., et al.: Wegener’s granulomatosis: clinical course in 108 patients with renal involvement. Nephrol. Dial. Transplant. 15:611, 2000. 2. Abgrall, S., Mouthon, L., Cohen, P., et al.: Localized neurological necrotizing vasculitides: three cases with isolated mononeuritis multiplex. J. Rheumatol. 28:631, 2001. 3. Adachi, J. D., Bensen, W.G., Brown, J., et al.: Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N. Engl. J. Med. 337:382, 1997. 4. Adachi, J. D., and Papaioannou, A.: Corticosteroid-induced osteoporosis: detection and management. Drug Saf. 24:607, 2001. 5. Adachi, J. D., Saag, K. G., Delmas, P. D., et al.: Two-year effects of alendronate on bone mineral density and vertebral fracture in patients receiving glucocorticoids: a randomized, double-blind, placebo-controlled extension trial. Arthritis Rheum. 44:202, 2001. 6. Adams, C. W., Buk, S. J., Hughes, R. A., et al.: Perls’ ferrocyanide test for iron in the diagnosis of vasculitic neuropathy. Neuropathol. Appl. Neurobiol. 15:433, 1989. 7. Adinolfi, L. E., Utili, R., Zampino, R., et al.: Effects of long-term course of alpha-interferon in patients with chronic hepatitis C associated to mixed cryoglobulinaemia. Eur. J. Gastroenterol. Hepatol. 9:1067, 1997. 8. Adu, D., Howie, A. J., Scott, D. G., et al.: Polyarteritis and the kidney. Q. J. Med. 62:221, 1987. 9. Adu, D., Pall, A., Luqmani, R. A., et al.: Controlled trial of pulse versus continuous prednisolone and cyclophosphamide in the treatment of vasculitis. Q. J. Med. 90:401, 1997. 10. Ahmad, T., Wallace, G. R., James, T., et al.: Mapping the HLA association in Behcet’s disease: a role for tumor necrosis factor polymorphisms? Arthritis Rheum. 48:807, 2003. 11. Aiello, P. D., Trautmann, J. C., McPhee, T. J., et al.: Visual prognosis in giant cell arteritis. Ophthalmology 100:550, 1993.

2385

12. Ali, H., Haribabu, B., Richardson, R. M., and Snyderman, R.: Mechanisms of inflammation and leukocyte activation. Med. Clin. North Am. 81:1, 1997. 13. Allen, N. B., Caldwell, D. S., Rice, J. R., and McCallum, R. M.: Cyclosporin A therapy for Wegener’s granulomatosis. Adv. Exp. Med. Biol. 336:473, 1993. 14. Altmeyer, P., Seifarth, D., and Bacharach-Muhles, M.: Hochdosierte intravenose Immunoglobulin (IVIG)-Therapie bei therapieresistenter ANCA-negativer, nekrotisierender Vaskulitis. Hautarzt 50:853, 1999. 15. American College of Rheumatology Ad Hoc Committee on Glucocorticoid-Induced Osteoporosis: Recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis: 2001 update. Arthritis Rheum. 44:1496, 2001. 16. Amin, S., LaValley, M. P., Simms, R. T., and Felson, D. T.: The role of vitamin D in corticosteroid-induced osteoporosis: a meta-analytic approach. Arthritis Rheum. 42:1740, 1999. 17. Amor-Dorado, J. C., Llorca, J., Garcia-Porrua, C., et al.: Audiovestibular manifestations in giant cell arteritis: a prospective study. Medicine 82:13, 2003. 18. Anderson, G., Coles, E. T., Crane, M., et al.: Wegener’s granuloma: a series of 265 British cases between 1975 and 1985. A report by a sub-committee of the British Thoracic Society Research Committee. Q. J. Med. 83:427, 1992. 19. Anderson, J. M., Jamieson, D. G., and Jefferson, J. M.: Non-healing granuloma and the nervous system. Q. J. Med. 44:309, 1975. 20. Anderson, P.: TIA-1: structural and functional studies on a new class of cytolytic effector molecule. Curr. Top. Microbiol. Immunol. 198:131, 1995. 21. Andonopoulos, A. P., Lagos, G., Drosos, A. A., and Moutsopoulos, H. M.: The spectrum of neurological involvement in Sjogren’s syndrome. Br. J. Rheumatol. 29:21, 1990. 22. Antoine, J.-C., Mosnier, J.-F., Absi, L., et al.: Carcinoma associated paraneoplastic peripheral neuropathies in patients with and without anti-onconeural antibodies. J. Neurol. Neurosurg. Psychiatry 67:7, 1999. 23. Apartis, E., Leger J.-M., Musset, L., et al.: Peripheral neuropathy associated with essential mixed cryoglobulinemia: a role for hepatitis C virus infection? J. Neurol. Neurosurg. Psychiatry 60:661, 1996. 24. Ara, J., Mirapeix, E., Arrizabalaga, P., et al.: Circulating soluble adhesion molecules in ANCA-associated vasculitis. Nephrol. Dial. Transplant. 16:276, 2001. 25. Arend, S. M., Hagen, E. C., Kroes, A. C., et al.: Activation of chronic hepatitis C virus infection by cyclophosphamide in a patient with cANCA-positive vasculitis. Nephrol. Dial. Transplant. 10:884, 1995. 26. Aslangul-Castier, E., Papo, T., Amoura, Z., et al.: Systemic vasculitis with bilateral perirenal haemorrhage in chronic myelomonocytic leukaemia. Ann. Rheum. Dis. 59:390, 2000. 27. Ault, B. H., Stapleton, F. B., Rivas, M. L., et al.: Association of Henoch-Schonlein purpura glomerulonephritis with C4B deficiency. J. Pediatr. 117:753, 1990. 28. Authier, F.-J., Bassez, G., Payan, C., et al.: Detection of genomic viral RNA in nerve and muscle of patients with HCV neuropathy. Neurology 60:808, 2003. 29. Averbuch-Heller, L., Steiner, I., and Abramsky, O.: Neurologic manifestations of progressive systemic sclerosis. Arch. Neurol. 49:1292, 1992.

2386

Diseases of the Peripheral Nervous System

30. Aydintug, A. O., Tokgoz, G., Ozoran, K., et al.: Elevated levels of soluble intercellular adhesion molecule-1 correlate with disease activity in Behcet’s disease. Rheumatol. Int. 15:75, 1995. 31. Bacon, P. A., and Carruthers, D.: New therapeutic aspects: haemopoietic stem cell transplantation. Best Pract. Res. Clin. Rheumatol. 15:299, 2001. 32. Baggiolini, M.: Chemokines in pathology and medicine. J. Intern. Med. 250:91, 2001. 33. Bahlas, S., Ramos-Remus, C., and Davis, P.: Clinical outcome of 149 patients with polymyalgia rheumatica and giant cell arteritis. J. Rheumatol. 25:99, 1998. 34. Bakouche, P., and Guillard, A.: Polyradiculonevrite au cours d’une poussee evolutive de maladie de Behcet. Rev. Neurol. (Paris) 140:520, 1984. 35. Ball, G. V., and Bridges, S. L. Jr.: Pathogenesis of vasculitis. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 34, 2002. 36. Ballare, M., Bobbio, F., Poggi, S., et al.: A pilot study on the effectiveness of cyclosporine in type II mixed cryoglobulinemia. Clin. Exp. Rheumatol. 13(Suppl.):S201, 1995. 37. Ballieux, B. E., Zondervan, K. T., Kievit, P., et al.: Binding of proteinase 3 and myeloperoxidase to endothelial cells: ANCA-mediated endothelial damage through ADCC? Clin. Exp. Immunol. 97:52, 1994. 38. Bannatyne, G. A.: Rheumatoid Arthritis: Its Pathology, Morbid Anatomy and Treatment, 2nd ed. Bristol, John Wright & Sons, p. 72, 1898. 39. Bartolucci, P., Ramanoelina, J., Cohen, P., et al.: Efficacy of the anti-TNF- antibody infliximab against refractory systemic vasculitides: an open pilot trial on 10 patients. Rheumatology 41:1126, 2002. 40. Bartunkova, J., Tesar, V., and Sediva, A.: Diagnostic and pathogenetic role of antineutrophil cytoplasmic autoantibodies. Clin. Immunol. 106:73, 2003. 41. Beckett, V. L., and Dinn, J. J.: Segmental demyelination in rheumatoid arthritis. Q. J. Med. 41:71, 1972. 42. Behn, A. R., Perera, T., and Myles, A. B.: Polymyalgia rheumatica and corticosteroids: how much for how long? Ann. Rheum. Dis. 42:374, 1983. 43. Belizna, C., and Cohen Tervaert, J. W.: Specificity, pathogenicity, and clinical value of antiendothelial cell antibodies. Semin. Arthritis Rheum. 27:98, 1997. 44. Belman, A. L., Leicher, C. R., Moshe, S. L., and Mezey, A. P.: Neurologic manifestations of Schoenlein-Henoch purpura: report of three cases and review of the literature. Pediatrics 75:687, 1985. 45. Belmont, H. M., Abramson, S. B., and Lie, J. T.: Pathology and pathogenesis of vascular injury in systemic lupus erythematosus: interactions of inflammatory cells and activated endothelium. Arthritis Rheum. 39:9, 1996. 46. Bely, M., and Apathy, A.: Vasculitis rheumatoid arthritisben. Orv. Hetil. 137:1571, 1996. 47. Berbegaal, J., Morera, J., Andrada, E., et al.: Sindrome de hipersensibilidad al allopurinol: aportacion de un nuevo caso y revision de la literatura espanola. Med. Clin. 102:178, 1994. 48. Berkman, E. M., and Orlin, J. B.: Use of plasmapheresis and partial plasma exchange in the management of patients with cryoglobulinemia. Transfusion 20:171, 1980.

49. Berlit, P.: Clinical and laboratory findings in giant cell arteritis. J. Neurol. Sci. 111:1, 1992. 50. Berman, A. E., and Kozlova, N. I.: Integrins: structure and function. Membr. Cell Biol. 13:207, 2000. 51. Binder, A., Snaith, M. L., and Isenberg, D.: Sjögren’s syndrome: a study of its neurological complications. Br. J. Rheumatol. 27:275, 1988. 52. Birck, R., Warnatz, K., Lorenz, H. M., et al.: 15Deoxyspergualin in patients with refractory ANCA-associated vasculitis: a six-month open-label trial to evaluate safety and efficacy. J. Am. Soc. Nephrol. 14:440, 2003. 53. Birkhead, N. C., Wagener, H. P., and Shick, R. M.: Treatment of temporal arteritis with adrenal corticosteroids: results in fifty-five cases in which lesion was proved at biopsy. J. Am. Med. Assoc. 163:821, 1957. 54. Bletry, O., Bussel, A., Badelon, I., et al.: Interet des echanges plasmatiques au cours des angeites necrosantes: 11 cas. Nouv. Presse Med. 11:2827, 1982. 55. Blumenthal, D., Schochet, S., Gutmann, L., et al.: Small-cell carcinoma of the lung presenting with paraneoplastic peripheral nerve microvasculitis and optic neuropathy. Muscle Nerve 21:1358, 1998. 56. Bodi, I., Varadi, P., Pokorny, G., et al.: Polyneuropathy with endoneurial immune complex deposition as the first manifestation of systemic lupus erythematosus. Acta Neuropathol. (Berl.) 96:297, 1998. 57. Boers, M., Croonen, A. M., Dijkmans, B. A., et al.: Renal findings in rheumatoid arthritis: clinical aspects of 132 necropsies. Ann. Rheum. Dis. 46:658, 1987. 58. Boki, K. A., Dafni, U., Karpouzas, G. A., et al.: Necrotizing vasculitis in Greece: clinical, immunological and immunogenetic aspects. A study of 66 patients. Br. J. Rheumatol. 36:1059, 1997. 59. Bolley, R., Mistry-Burchardi, N., and Samtleben, W.: Wegener Granulomatose und mikroskopische Polyangiitis: diagnostische und klinische Ergebnisse bei 54 Patienten im Langzeitverlauf. Dtsch. Med. Wochenschr. 125:1519, 2000. 60. Bonetti, B., Invernizzi, F., Rizzuto, N., et al.: T-cell-mediated epineurial vasculitis and humoral-mediated microangiopathy in cryoglobulinemic neuropathy. J. Neuroimmunol. 73:145, 1997. 61. Bonsib, S. M.: Polyarteritis nodosa. Semin. Diagn. Pathol. 18:14, 2001. 62. Boomsma, M. M., Stegeman, C. A., van der Leij, M. J., et al.: Prediction of relapses in Wegener’s granulomatosis by measurement of antineutrophil cytoplasmic antibody levels: a prospective study. Arthritis Rheum. 43:2025, 2000. 63. Boonyapisit, K., and Katirji, B.: Severe exacerbation of hepatitis C-associated vasculitic neuropathy following treatment with interferon alpha: a case report and literature review. Muscle Nerve 25:909, 2002. 64. Borish, L. C., and Steinke, J. W.: Cytokines and chemokines. J. Allergy Clin. Immunol. 111:S460, 2003. 65. Bosboom, W. M., Van den Berg, L. H., Mollee, I., et al.: Sural nerve T-cell receptor V gene utilization in chronic inflammatory demyelinating polyneuropathy and vasculitic neuropathy. Neurology 56:74, 2001. 66. Bosch, X., Navarro, M., Lopez-Soto, A., et al.: Primary polyarteritis nodosa presenting as acute symmetric quadriplegia. Clin. Exp. Rheumatol. 17:232, 1999.

Neuropathies with Systemic Vasculitis 67. Bouche, P., Leger, J. M., Travers, M. A. et al.: Peripheral neuropathy in systemic vasculitis: clinical and electrophysiologic study of 22 patients. Neurology 36:1598, 1986. 68. Boumpras, D. T., and Wilder, R. L.: Corticosteroids. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 827, 2001. 69. Boyd, L. J.: IV. Periarteritis nodosa: neuromyositic manifestations. N. Y. Med. Coll. Flow. Hosp. Bull. 3:272, 1940. 70. Bradley, W. G., and Verma, A.: Painful vasculitic neuropathy in HIV-1 infection: relief of pain with prednisone therapy. Neurology 47:1446, 1996. 71. Brannagan, T. H.: Retroviral-associated vasculitis of the nervous system. Neurol. Clin. 15:927, 1997. 72. Brasile, L., Kremer, J. M., Clarke, J. L., and Cerilli, J.: Identification of an autoantibody to vascular endothelial cell-specific antigens in patients with systemic vasculitis. Am. J. Med. 87:74, 1989. 73. Bridges, A. J., Porter, J., and England, D.: Lower extremity peripheral neuropathy and ischemic ulcers associated with giant cell arteritis. J. Rheumatol. 16:1366, 1989. 74. Briemberg, H. R., Levin, K., and Amato, A. A.: Multifocal conduction block in peripheral nerve vasculitis. J. Clin. Neuromuscul. Dis. 3:153, 2002. 75. Brons, R. H., Kallenberg, C. S., and Cohen Tervaert, J. W.: Are antineutrophil cytoplasmic antibody-associated vasculitides pauci-immune? Rheum. Dis. Clin. North Am. 27:833, 2001. 76. Bruce, I. N., and Bell, A. L.: A comparison of two nomenclature systems for primary systemic vasculitis. Br. J. Rheumatol. 453, 1997. 77. Bulun, A., Topaloglu, R., Duzova, A., et al.: Ataxia and peripheral neuropathy: rare manifestations in HenochSchonlein purpura. Pediatr. Nephrol. 16:1139, 2001. 78. Burke, A. P., Sobin, L. H., and Virmani, R.: Localized vasculitis of the gastrointestinal tract. Am. J. Surg. Pathol. 19:338, 1995. 79. Burton, E. A., Winer, J. B., and Barber, P. C.: Giant cell arteritis of the cervical radicular vessels presenting with diaphragmatic weakness. J. Neurol. Neurosurg. Psychiatry 67:223, 1999. 80. Cacoub, P., Costedoat-Chalumeau, N., Lidove, O., and Alric, L.: Cryoglobulinemia vasculitis. Curr. Opin. Rheumatol. 14:29, 2002. 81. Cacoub, P., Lidove, O., Maisonobe, T., et al.: Interferon- and ribavirin treatment in patients with hepatitis C virus-related systemic vasculitis. Arthritis Rheum. 46:3317, 2002. 82. Cacoub, P., Sbai, A., Gatel, A., et al.: Vascularites systemiques post-infectieuses: guerison sans corticotherapie. J. Mal. Vasc. 22:29, 1997. 83. Calabrese, L. H., and Duna, G.: Vasculitis associated with antineutrophil cytoplasmic antibody. In Ruddy, S., Harris, E. D., and Sledge, C. B. (eds.): Kelley’s Textbook of Rheumatology, 6th ed. Philadelphia, W. B. Saunders, p. 1165, 2001. 84. Campbell, S. B., Hawley, C. M., and Staples, C.: Mononeuritis multiplex complicating Henoch-Schonlein purpura [letter]. Aust. N. Z. J. Med. 24:580, 1994. 85. Campion, G., Maddison, P. J., Goulding, N., et al.: The Felty syndrome: a case-matched study of clinical manifestations and outcome, serologic features, and immunogenetic associations. Medicine 69:69, 1990.

2387

86. Carson, C. W., Conn, D. L., Czaja, A. J., et al.: Frequency and significance of antibodies to hepatitis C virus in polyarteritis nodosa. J. Rheumatol. 20:304, 1993. 87. Carter, R. A., and Wicks, I. P.: Vascular cell adhesion molecule 1 (CD106): a multifaceted regulator of inflammation. Arthritis Rheum. 44:985, 2001. 88. Carvalho, D., Savage, C., Isenberg, D., and Pearson, J.: IgG anti-endothelial cell autoantibodies from patients with systemic lupus erythematosus or systemic vasculitis stimulate the release of two endothelial cell-derived mediators, which enhance adhesion molecule expression and leukocyte adhesion in an autocrine manner. Arthritis Rheum. 42:631, 1999. 89. Casato, M., Lagana, B., Antonelli, G., et al.: Long-term results of therapy with interferon-alpha for type II essential mixed cryoglobulinemia. Blood 78:314, 1991. 90. Caselli, R. J., Daube, J. R., Hunder, G. G., and Whisnant, J. P.: Peripheral neuropathic syndromes in giant cell (temporal) arteritis. Neurology 38:685, 1988. 91. Castaigne, P., Brunet, P., Hauw, J. J., et al.: Systeme nerveux peripherique et panarterite noueuse: revue de 27 cas. Rev. Neurol. (Paris) 140:343, 1984. 92. Castillo, S., Revelen, R., Bordron, A., et al.: Anti-endothelial cell reactivity, the unresolved dilemma . . . Int. J. Immunopathol. Pharmacol. 14:109, 2001. 93. Cavaletti, G., Petruccioli, M. G., Crespi, V., et al.: A clinicopathological and follow up study of 10 cases of essential type II cryoglobulinemic neuropathy. J. Neurol. Neurosurg. Psychiatry 53:886, 1990. 94. Chalopin, J. M., Rifle, G., Turc, J. M., et al.: Immunological findings during successful treatment of HBsAg-associated polyarteritis nodosa by plasmapheresis alone. Br. Med. J. 280:368, 1980. 95. Chamouard, J. M., Smadja, D., Chaunu, M. P., and Bouche, P.: Neuropathie par vasculite necrosanta au cours de l’infection par le VIH1. Rev. Neurol. (Paris) 149:358, 1993. 96. Chang, R. W., Bell, C. C., and Hallett, M.: Clinical characteristics and prognosis of vasculitic mononeuropathy multiplex. Arch. Neurol. 41:618, 1984. 97. Chen, K. R., and Miyakawa, S.: Clinical and histopathological spectrum of cutaneous vasculitis in rheumatoid arthritis. Br. J. Dermatol. 147:905, 2002. 98. Chetty, R.: Vasculitides associated with HIV infection. J. Clin. Pathol. 54:275, 2001. 99. Chia, L., Fernandez, A., LaCroix, C., et al.: Contribution of nerve biopsy findings to the diagnosis of disabling neuropathy in the elderly: a retrospective review of 100 consecutive patients. Brain 119:1091, 1996. 100. Chretien, F., Gray, E., Lescs, M. C., et al.: Acute varicellazoster virus ventriculitis and meningo-myelo-radiculitis in acquired immunodeficiency syndrome. Acta Neuropathol. (Berl.) 86:659, 1993. 101. Chung, J. B., Armstrong, K., Schwartz, J. S., and Albert, D.: Cost-effectiveness of prophylaxis against Pneumocystis carinii pneumonia in patients with Wegener’s granulomatosis undergoing immunosuppressive therapy. Arthritis Rheum. 43:1841, 2000. 102. Churg, A.: Recent advances in the diagnosis of the ChurgStrauss syndrome. Mod. Pathol. 14:1284, 2001. 103. Churg, J., and Strauss, L.: Allergic granulomatosis, allergic angiitis, and periarteritis nodosa. Am. J. Pathol. 27:277, 1951.

2388

Diseases of the Peripheral Nervous System

104. Cid, M. C., Cebrian, M., Font, C., et al.: Cell adhesion molecules in the development of inflammatory infiltrates in giant cell arteritis. Arthritis Rheum. 43:184, 2000. 105. Cid, M. C., Hernandez-Rodriguez, J., Robert, J., et al.: Interferon- may exacerbate cryoglobulinemia-related ischemic complications: an adverse effect potentially related to its anti-angiogenic activity. Arthritis Rheum. 42:1051, 1999. 106. Cinta-Cid, M. C., Coll-Vincent, B., and Bielsa, I.: Endothelial adhesion molecules. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 13, 2002. 107. Ciompi, M. L., Marini, D., Siciliano, G., et al.: Cryoglobulinemic peripheral neuropathy: neurophysiologic evaluation in twenty-two patients. Biomed. Pharmacother. 50:329, 1996. 108. Cisternas, M., Wainstein, E., Soto, L., et al.: Clinical features of systemic vasculitis in Chile: a ten-year study [abstract]. Cleve. Clin. J. Med. 69(Suppl. 2):SII170, 2002. 109. Claussen, G. C., Thomas, T. D., Goyne, C., et al.: Diagnostic value of nerve and muscle biopsy in suspected vasculitis cases. J. Clin. Neuromuscul. Dis. 1:117, 2000. 110. Clayton, A. R., and Savage, C. O.: What you should know about PR3-ANCA: evidence for the role of T cells in the pathogenesis of systemic vasculitis. Arthritis Res. 2:260, 2000. 111. Clowes, J. A., Peel, N., and Eastell, R.: Glucocorticoid-induced osteoporosis. Curr. Opin. Rheumatol. 13:326, 2001. 112. Cochrane, C., and Koffler, D.: Immune complex disease in experimental animals and man. Adv. Immunol. 16:185, 1973. 113. Cohen, P., Nguyen, Q. T., Deny, P., et al.: Treatment of mixed cryoglobulinemia with recombinant interferon alpha and adjuvant therapies: a prospective study of 20 patients. Ann. Med. Interne (Paris) 147:81, 1996. 114. Cohen, R. D., Conn, D. L., and Ilstrup, D. M.: Clinical features, prognosis, and response to treatment in polyarteritis. Mayo Clin. Proc. 55:146, 1980. 115. Cohen Tervaert, J. W., and Kallenberg, C. G.: Cell adhesion molecules in vasculitis. Curr. Opin. Rheumatol. 9:16, 1997. 116. Collins, M. P., and Kissel, J. T.: Vasculitis of the peripheral nervous system. In Noseworthy, J. H. (ed.): Neurological Therapeutics: Principles and Practice. London, Martin Dunitz, p. 2078, 2003. 117. Collins, M. P., Mendell, J. R., Periquet, M. I., et al.: Immunostaining in vasculitic neuropathy: sensitivity and specificity of findings [abstract]. Ann. Neurol. 42:415, 1997. 118. Collins, M. P., Mendell, J. R., Periquet, M. I., et al.: Superficial peroneal nerve/peroneus brevis muscle biopsy in vasculitic neuropathy. Neurology 55:636, 2000. 119. Collins, M. P., and Periquet, M. I.: Prevalence of vasculitic neuropathy in Wegener granulomatosis. Arch. Neurol. 59:1333, 2002. 120. Collins, M. P., Periquet, M. I., Mendell, J. R., et al.: Nonsystemic vasculitic neuropathy: insights from a clinical cohort. Neurology 61:623, 2003. 121. Coll-Vinent, B., Cebrian, M., Cid, M. C., et al.: Dynamic patterns of endothelial cell adhesion molecule expression in muscle and perineural vessels from patients with classic polyarteritis nodosa. Arthritis Rheum. 41:435, 1998.

122. Condliffe, A. M., Kitchen, E., and Chilvers, E. R.: Neutrophil priming: pathophysiological consequences and underlying mechanisms. Clin. Sci. 94:461, 1998. 123. Conn, D. L., McDuffie, F. C., and Dyck, P. J.: Immunopathologic study of sural nerves in rheumatoid arthritis. Arthritis Rheum. 15:135, 1972. 124. Conn, D. L., Schroeter, A. L., and McDuffie, F. C.: Cutaneous vessel immune deposits in rheumatoid arthritis. Arthritis Rheum. 19:15, 1976. 125. Cooke, S. P., Rigby, S. P., Griffiths, D. J., and Venables, P. J.: Viral studies in rheumatic disease. Ann. Med. Interne (Paris) 149:30, 1998. 126. Cooper, G. S., Miller, F. W., and Germolec, D. R.: Occupational exposures and autoimmune diseases. Int. Immunopharmacol. 2:303, 2002. 127. Corbo, M., Nemni, R, Iannaccone, S., et al.: Peripheral neuropathy in scleroderma. Clin. Neuropathol. 12:63, 1993. 128. Coremans, I. E., Hagen, E. C., Daha, M. R., et al.: Antilactoferrin antibodies in patients with rheumatoid arthritis are associated with vasculitis. Arthritis Rheum. 35:1466, 1992. 129. Cornely, O. A., Hauschild, S., Weise, C., et al.: Seroprevalence and disease association of antineutrophil cytoplasmic autoantibodies and antigens in HIV infection. Infection 27:92, 1999. 130. Crissey, J. T., and Parish, L. C.: Vasculitis: the historical development of the concept. Clin. Dermatol. 17:493, 1999. 131. Crow, M. K.: Structure and function of monocytes/ macrophages and other antigen-presenting cells. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 317, 2001. 132. Cruickshank, B.: The arteritis of rheumatoid arthritis. Ann. Rheum. Dis. 13:136, 1954. 133. Csernok, E., and Gross, W. L.: Cytokines and vascular inflammation. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 97, 2002. 134. Csernok, E., Trabandt, A., Muller, A., et al.: Cytokine profiles in Wegener’s granulomatosis: predominance of type 1 (Th1) in the granulomatous inflammation. Arthritis Rheum. 42:742, 1999. 135. Cuchacovich, R.: Immunopathogenesis of vasculitis. Curr. Rheumatol. Rep. 4:9, 2002. 136. Dafni, U. G., Tzioufas, A. G., Staikos, P., et al.: Prevalence of Sjogren’s syndrome in a closed rural community. Ann. Rheum. Dis. 56:521, 1997. 137. Dalgleish, P. G.: Polyarteritis nodosa after thiouracil. Lancet 2:319, 1952. 138. D’Amico, G.: Renal involvement in hepatitis C infection: cryoglobulinemic glomerulonephritis. Kidney Int. 54:650, 1998. 139. Dammacco, F., Sansonno, D., Han, J. H., et al.: Natural interferon- versus its combination with 6-methylprednisolone in the therapy of type II mixed cryoglobulinemia: a long-term, randomized, controlled study. Blood 84:3336, 1994. 140. D’Angelo, W. A., Fries, J. F., Masi, A. T., and Shulman, L. E.: Pathologic observations in systemic sclerosis (scleroderma):

Neuropathies with Systemic Vasculitis

141.

142.

143. 144.

145.

146.

147.

148.

149.

150.

151.

152.

153.

154.

155.

156.

157.

a study of fifty-eight autopsy cases and fifty-eight matched controls. Am. J. Med. 46:428, 1969. Daoud, M. S., Hutton, K. P., and Gibson, L. E.: Cutaneous periarteritis nodosa: a clinicopathological study of 79 cases. Br. J. Dermatol. 136:706, 1997. Davies, L., Spies, J. M., Pollard, J. D., and McLeod, J. G.: Vasculitis confined to peripheral nerves. Brain 119:1441, 1996. Davson, J., Ball, J., and Platt, R.: The kidney in periarteritis nodosa. Q. J. Med. 17:175, 1948. D’Cruz, D., Cervera, R., Aydintug, A. O., et al.: Systemic lupus erythematosus evolving into systemic vasculitis: a report of five cases. Br. J. Rheumatol. 32:154, 1993. de Groot, K., Adu, D., and Savage, C. O.: The value of pulse cyclophosphamide in ANCA-associated vasculitis: meta-analysis and critical review. Nephrol. Dial. Transplant. 16:2018, 2001. de Groot, K., Muhler, M., Reinhold-Keller, E., et al.: Induction of remission in Wegener’s granulomatosis with low dose methotrexate. J. Rheumatol. 25:492, 1998. de Groot, K., Rasmussen, N., Cohen Tervaert, J. W., and Jayne, D. R.: Randomized trial of cyclophosphamide versus methotrexate for induction of remission in “non-renal” ANCA-associated vasculitis [abstract]. J. Am. Soc. Nephrol. 13:A2, 2002. de Groot, K., Reinhold-Keller, E., Tatsis, E., et al.: Therapy for the maintenance of remission in sixty-five patients with generalized Wegener’s granulomatosis: methotrexate versus trimethoprim/sulfamethoxazole. Arthritis Rheum. 39:2052, 1996. de Groot, K., Schmidt, D. K., Arlt, A. C., et al.: Standardized neurologic evaluations of 128 patients with Wegener granulomatosis. Arch. Neurol. 58:1215, 2001. De Martino, L., Sampaolo, S., Tucci, C., et al.: Viral RNA in nerve tissues of patients with hepatitis C infection and peripheral neuropathy. Muscle Nerve 27:102, 2003. De Vita, S., De Re, V., Gasparotto, D., et al.: Oligoclonal non-neoplastic B cell expansion is the key feature of type II mixed cryoglobulinemia. Arthritis Rheum. 43:94, 2000. Del Buono, E. A., and Flint, A.: Diagnostic usefulness of nasal biopsy in Wegener’s granulomatosis. Hum. Pathol. 22:107, 1991. Del Papa, N. L., Meroni, P. L., Barcellini, W., et al.: Antibodies to endothelial cells in primary vasculitides mediate in vitro endothelial cytotoxicity in the presence of normal peripheral blood mononuclear cells. Clin. Immunol. Immunopathol. 63:267, 1992. Del Rosso, A., Generini, S., Pignone, A., and MatucciCerinic, M.: Vasculitides secondary to systemic diseases. Clin. Dermatol. 17:533, 1999. Della Rossa, A., Baldini, C., Tavoni, A., et al.: ChurgStrauss syndrome: clinical and serological features of 19 patients from a single Italian centre. Rheumatology 41:1286, 2002. Denton, C. P., and Abraham, D. J.: Transforming growth factor-beta and connective tissue growth factor: key cytokines in scleroderma pathogenesis. Curr. Opin. Rheumatol. 13:505, 2001. Deprez, M., Lubke, U., Verlaet, M., et al.: Detection of cytokines in human sural nerve biopsies: an immunohis-

158.

159.

160. 161.

162.

163. 164. 165. 166.

167.

168.

169.

170.

171.

172.

173.

174.

175.

2389

tochemical study. Acta Neuropathol. (Berl.) 101:393, 2001. Devaney, K. O., Travis, W. D., Hoffman, G., et al.: Interpretation of head and neck biopsies in Wegener’s granulomatosis: a pathologic study of 126 biopsies in 70 patients. Am. J. Surg. Pathol. 14:555, 1990. Di Trapani, G., Tulli, A., LaCara, A., et al.: Peripheral neuropathy in course of progressive systemic sclerosis: light and ultrastructural study. Acta Neuropathol. (Berl.) 72:103, 1986. Diaz-Perez, J. L., and Winkelmann, R. K.: Cutaneous periarteritis nodosa. Arch. Dermatol. 110:407, 1974. Dijstelbloem, H. M., Scheepers, R. H., Oost, W. W., et al.: Fcgamma receptor polymorphisms in Wegener’s granulomatosis: risk factors for disease relapse. Arthritis Rheum. 42:1823, 1999. Direskeneli, H.: Behcet’s disease: infectious aetiology, new autoantigens, and HLA-B51. Ann. Rheum. Dis. 60:996, 2001. Dispenzieri, A., and Gorevic, P. D.: Cryoglobulinemia. Hematol. Oncol. Clin. North Am. 13:1315, 1999. Drenkard, C., Villa, A. R., Reyes, E., et al.: Vasculitis in systemic lupus erythematosus. Lupus 6:235, 1997. Dyck, P. J., Benstead, T. J., Conn, D. L., et al.: Nonsystemic vasculitic neuropathy. Brain 110:843, 1987. Dyck, P. J., Conn, D. L., and Ozaki, H.: Necrotizing angiopathic neuropathy: three-dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972. Dyck, P. J. B., Engelstad, J. K., and Dyck, P. J.: Necrotizing vasculitis of biopsied nerves: heterogeneity of pathologic alterations and clinical syndromes [abstract]. J. Peripher. Nerv. Syst. 6:138, 2001. Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy (LSRPN): similarity to the diabetic variety (DLSRPN). J. Neuropathol. Exp. Neurol. 59:525, 2000. Dyck, P. J. B., Hunder, G. G., and Dyck, P. J.: A case-control and nerve biopsy study of CREST multiple mononeuropathy. Neurology 49:1641, 1997. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999. Dyck, P. J. B., and Windebank, A. J.: Diabetic and nondiabetic lumbosacral radiculoplexus neuropathies: new insights into pathophysiology and treatment. Muscle Nerve 25:477, 2002. Eastell, R., Reid, D. M., Compston, J., et al.: A UK Consensus Group on management of glucocorticoid-induced osteoporosis: an update. J. Intern. Med. 244:271, 1998. Elliot, J. D., Lockwood, C. M., Hale, G., and Waldmann, H.: Semi-specific immuno-absorption and monoclonal antibody therapy in ANCA positive vasculitis: experience in four cases. Autoimmunity 28:163, 1998. Elzouki, A. N., Segelmark, M., Wieslander, J., and Eriksson, S.: Strong link between the alpha 1-antitrypsin PiZ allele and Wegener’s granulomatosis. J. Intern. Med. 236:543, 1994. Esnault, V. L.: Apoptosis: the central actor in the three hits that trigger anti-neutrophil cytoplasmic antibody-related

2390

176.

177.

178.

179. 180.

181. 182.

183.

184.

185.

186.

187.

188.

189.

190.

191.

192.

Diseases of the Peripheral Nervous System

systemic vasculitis. Nephrol. Dial. Transplant. 17:1725, 2002. Esnault, V. L., Audrain, M. A., and Sesboue, R.: Alpha1-antitrypsin phenotyping in ANCA-associated diseases: one of several arguments for protease/antiprotease imbalance in systemic vasculitis. Exp. Clin. Immunogenet. 14:206, 1997. Esnault, V. L., Testa, A., Audrain, M., et al.: Alpha 1-antitrypsin genetic polymorphism in ANCA-positive systemic vasculitis. Kidney Int. 43:1329, 1993. Evans, J. M., O’Fallon, W. M., and Hunder, G. G.: Increased incidence of aortic aneurysms and dissection in giant cell (temporal) arteritis: a population-based study. Ann. Intern. Med. 122:502, 1995. Fadli, M. E., and Youssef, M. M.: Neuro-Behcet’s syndrome in the United Arab Republic. Eur. Neurol. 9:76, 1973. Falk, R. J., Nachman, P. H., Hogan, S. L., and Jennette, J. C.: ANCA glomerulonephritis and vasculitis: a Chapel Hill perspective. Semin. Nephrol. 20:233, 2000. Farrell, D. A., and Medsger, T. A. Jr.: Trigeminal neuropathy in progressive systemic sclerosis. Am. J. Med. 73:57, 1982. Fauci, A. S.: The vasculitis syndromes. In Braunwald, E., Fauci, A. S., Kasper, D. L., et al. (eds.): Harrison’s Principles of Internal Medicine, 15th ed. New York, McGraw-Hill, p. 1956, 2001. Fauci, A. S., Haynes, B. F., Katz, P., et al.: Wegener’s granulomatosis: prospective and therapeutic experience with 85 patients for 21 years. Ann. Intern. Med. 98:76, 1983. Fei, G. Z., Sylvan, S. P., Yao, G. B., and Hellstrom, U. B.: Quantitative monitoring of serum hepatitis B virus DNA and blood lymphocyte subsets during combined prednisolone and interferon-alpha therapy in patients with chronic hepatitis B. J. Viral Hepat. 6:210, 1999. Feinglass, E. J., Arnett, F. C., Dorsch, C. A., et al.: Neuropsychiatric manifestations of lupus erythematosus: diagnosis, clinical spectrum, and relationship to other features of the disease. Medicine 55:323, 1976. Ferrari, E.: Ueber polyarteritis acuta nodosa (sogenannte Periarteritis nodosa) und ihre Beziehungen zur Polymyositis und Polyneuritis acuta. Beitr. Pathol. Anat. 34:350, 1903. Ferri, C., Gremignai, G., Bombardieri, S., et al.: Plasmaexchange in mixed cryoglobulinemia : effects on renal, liver and neurological involvement. La Ricerca Clin. Lab. 16:403, 1986. Ferri, C., LaCivita, L., Cirafisi, C., et al.: Peripheral neuropathy in mixed cryoglobulinemia: clinical and electrophysiologic investigations. J. Rheumatol. 19:889, 1992. Ferri, C., Marzo, E., Longombardo, G., et al.: Interferonalpha in mixed cryoglobulinemia patients: a randomized, crossover-controlled trial. Blood 81:1132, 1993. Ferri, C., Zignego, A. L., Giuggioli, D., et al.: HCV and cryoglobulinemic vasculitis. Cleve. Clin. J. Med. 69(Suppl. 2): SII20, 2002. Finkel, T. H., Torok, T. J., Ferguson, P. J., et al.: Chronic parvovirus B19 infection and systemic necrotizing vasculitis: opportunistic infection or aetiological agent? Lancet 343:1255, 1994. Finkelman, R., Munsat, T., Mandell, H., et al.: Neuromuscular manifestations of Wegener’s granulomatosis: a case report. Neurology 43:617, 1993.

193. Firestein, G. S.: Etiology and pathogenesis of rheumatoid arthritis. In Ruddy, S., Harris, E. D., and Sledge, C. B. (eds.): Kelley’s Textbook of Rheumatology, 6th ed. Philadelphia, W. B. Saunders, p. 921, 2001. 194. Font, C., Miro, O., Pedrol, E., et al.: Polyarteritis nodosa in human immunodeficiency virus infection: report of four cases and review of the literature. Br. J. Rheumatol. 35:796, 1996. 195. Font, J., Cervera, R., Navarro, M., et al.: Systemic lupus erythematosus in men: clinical and immunological characteristics. Ann. Rheum. Dis. 51:1050, 1992. 196. Fort, J. G., Griffin, R., Tahmoush, A., and Abruzzo, J. L.: Muscle involvement in polyarteritis nodosa: report of a patient presenting clinically as polymyositis and review of the literature. J. Rheumatol. 21:945, 1994. 197. Fortin, P. R.: Vasculitis and malignancy. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 707, 2002. 198. Fortin, P. R., Larson, M. G., Watters, K., et al.: Prognostic factors in systemic necrotizing vasculitis of the polyarteritis nodosa group—a review of 45 cases. J. Rheumatol. 22:78, 1995. 199. Francke, M. L., Mihaescu, A., and Chaubert, P.: Isolated necrotizing arteritis of the female reproductive tract: a clinicopathologic and immunohistochemical study of 11 cases. Int. J. Gynecol. Pathol. 17:193, 1998. 200. Frankel, A. H., Singer, D. R., Winearls, C. G., et al.: Type II essential mixed cryoglobulinemia: presentation, treatment and outcome in 13 patients. Q. J. Med. 82:101, 1992. 201. Fregoni, V., Perseghin, P., Epsis, R., et al.: Churg-Strauss syndrome with peripheral polyneuropathy refractory to steroidal and immunosuppressive therapy successfully treated with plasma exchange. Recent Prog. Med. 86:353, 1995. 202. Fried, M. W., Shiffman, M. L., Reddy, K. R., et al.: Peginterferon alfa-2a plus ribavirin for chronic hepatitis C virus infection. N. Engl. J. Med. 347:975, 2002. 203. Frohnert, P. P., and Sheps, S. G.: Long-term follow-up study of periarteritis nodosa. Am. J. Med. 43:8, 1967. 204. Fujimura, H., LaCroix, C., and Said, G.: Vulnerability of nerve fibers to ischaemia: a quantitative light and electron microscopic study. Brain 114:1929, 1991. 205. Fuller, G. N., Jacobs, J. M., and Guiloff, R. J.: Nature and incidence of peripheral nerve syndromes in HIV infection. J. Neurol. Neurosurg. Psychiatry 56:372, 1993. 206. Funaoka, M., Kato, K., Komatsu, M., et al.: Fulminant hepatitis caused by hepatitis C virus during treatment for multiple sclerosis. J. Gastroenterol. 31:119, 1996. 207. Gaffney, P. M., Moser, K. L., Graham, R. R., and Behrens, T. W.: Recent advances in the genetics of systemic lupus erythematosus. Rheum. Dis. Clin. North Am. 28:111, 2002. 208. Galli, M., Invernizzi, F., Monteverde, A., and Monti, G.: Cryoglobulinemic vasculitis. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 495, 2002. 209. Garcia-Bragado, F., Fernandez, J. M., Navarro, C., et al.: Peripheral neuropathy in essential mixed cryoglobulinemia. Arch. Neurol. 45:1210, 1988. 210. Garcia-Carrasco, M., Ramos-Casals, M., Rosas, J., et al.: Primary Sjogren’s syndrome: clinical and immunologic

Neuropathies with Systemic Vasculitis

211.

212.

213.

214.

215.

216.

217.

218.

219.

220. 221.

222.

223.

224.

225.

disease patterns in a cohort of 400 patients. Medicine 81:270, 2002. Garcia-Porrua, C., Calvino, M. C., Llorca, J., et al.: Henoch-Schonlein purpura in children and adults: clinical differences in a defined population. Semin. Arthritis Rheum. 32:149, 2002. Garcia-Porrua, C., and Gonzalez-Gay, M. A.: Comparative clinical and epidemiological study of hypersensitivity vasculitis versus Henoch-Schonlein purpura in adults. Semin. Arthritis Rheum. 28:404, 1999. Garcia-Porrua, C., Gonzalez-Gay, M. A., and LopezLazaro, L.: Drug associated cutaneous vasculitis in adults in northwestern Spain. J. Rheumatol. 26:1942, 1999. Garcia-Porrua, C., Llorca, J., Gonzalez-Louzao, C., and Gonzalez-Gay, M. A.: Hypersensitivity vasculitis in adults: a benign disease usually limited to the skin. Clin. Exp. Rheumatol. 19:85, 2001. Gaskin, G., Jayne, D. R., and the European Vasculitis Study Group: Adjunctive plasma exchange is superior to methylprednisolone in acute renal failure due to ANCA-associated glomerulonephritis [abstract]. J. Am. Soc. Nephrol. 13:2A, 2002. Gayraud, M., Guillevin, L., Cohen, P., et al.: Treatment of good-prognosis polyarteritis nodosa and Churg-Strauss syndrome: comparison of steroids and oral or pulse cyclophosphamide in 25 patients. French Cooperative Study Group for Vasculitides. Br. J. Rheumatol. 36:1290, 1997. Gayraud, M., Guillevin, L., le Toumelin, P., et al.: Longterm followup of polyarteritis nodosa, microscopic polyangiitis, and Churg-Strauss syndrome: analysis of four prospective trials including 278 patients. Arthritis Rheum. 44:666, 2001. Gedalia, A., Correa, H., Kaiser, M., and Sorensen, R.: Case report: steroid sparing effect of intravenous gamma globulin in a child with necrotizing vasculitis. Am. J. Med. Sci. 309:226, 1995. Geirsson, A. J., Sturfelt, G., and Truedsson, L.: Clinical and serological features of severe vasculitis in rheumatoid arthritis: prognostic implications. Ann. Rheum. Dis. 46:727, 1987. Geltner, D.: Therapeutic approaches to mixed cryoglobulinemia. Springer Semin. Immunopathol. 10:103, 1988. Gemignani, F., Marbini, A., Pavesi, G., et al.: Peripheral neuropathy associated with Sjögren’s syndrome. J. Neurol. Neurosurg. Psychiatry 57:983, 1994. Gemignani, F., Melli, G., Inglese, C., and Marbini, A.: Cryoglobulinemia is a frequent cause of peripheral neuropathy in undiagnosed referral patients. J. Peripher. Nerv. Syst. 7:59, 2002. Gemignani, F., Pavesi, G., Manganelli, P., et al.: Peripheral neuropathy in essential mixed cryoglobulinemia. J. Neurol. Neurosurg. Psychiatry 55:116, 1992. Gencik, M., Borgmann, S., Zahn, R., et al.: Immunogenetic risk factors for anti-neutrophil cytoplasmic antibody (ANCA)-associated systemic vasculitis. Clin. Exp. Immunol. 117:412, 1999. Gephardt, G. N., Ahmad, M., and Tubbs, R. R.: Pulmonary vasculitis (Wegener’s granulomatosis): immunohistochemical study of T and B cell markers. Am. J. Med. 74:700, 1983.

2391

226. Gherardi, R., Belec, L., Mhiri, C., et al.: The spectrum of vasculitis in human immunodeficiency virus-infected patients. Arthritis Rheum. 36:1164, 1993. 227. Gherardi, R., Lebargy, F., Gaulard, C., et al.: Necrotizing vasculitis and HIV replication in peripheral nerves. N. Engl. J. Med. 321:685, 1989. 228. Gherardi, R. K., Amiel, H., Martin-Mondiere, C., et al.: Solitary plasmacytoma of the skull revealed by a mononeuritis multiplex associated with immune complex mediated vasculitis. Arthritis Rheum. 32:1470, 1989. 229. Gherardi, R. K., Mhiri, C., Baudrimont, M., et al.: Iron pigment deposits, small vessel vasculitis, and erythrophagocytosis in the muscle of human immunodeficiency virusinfected patients. Hum. Pathol. 22:1187, 1991. 230. Ghini, M., Mascia, M. T., Gentilini, M., and Mussini, C.: Treatment of cryoglobulinemic neuropathy with a-interferon. Neurology 46:588, 1996. Letter. 231. Gilliland, B. C.: Systemic sclerosis (scleroderma). In Braunwald, E., Fauci, A. S., Kasper, D. L., et al. (eds.): Harrison’s Principles of Internal Medicine, 15th ed. New York, McGraw-Hill, p. 1937, 2001. 232. Gisselbrecht, M., Cohen, P., Lortholary, O., et al.: Human immunodeficiency virus-related vasculitis: clinical presentation of and therapeutic approach to eight cases. Ann. Med. Interne (Paris) 149:398, 1998. 233. Gladman, D. D., and Urowitz, M. B.: Systemic lupus erythematosus. B. Clinical and laboratory features. In Klippel, J. H. (ed.): Primer on the Rheumatic Diseases, 11th ed. Atlanta, Arthritis Foundation, p. 251, 1997. 234. Godman, G. C., and Churg, J.: Wegener’s granulomatosis: pathology and review of the literature. Arch. Pathol. 58:533, 1954. 235. Golbus, J., and McCune, W. J.: Giant cell arteritis and peripheral neuropathy: a report of 2 cases and review of the literature. J. Rheumatol. 14:129, 1987. 236. Gold, J. E., Jimenez, E., and Zalusky, R.: Human immunodeficiency virus-related lymphoreticular malignancies and peripheral neurologic disease: a report of four cases. Cancer 61:2318, 1988. 237. Gonzalez, E. B., and Conn, D. L.: Hypersensitivity vasculitis (small-vessel cutaneous vasculitis). In Ruddy, S., Harris, E. D., and Sledge, C. B. (eds.): Kelley’s Textbook of Rheumatology, 6th ed. Philadelphia, W. B. Saunders, p. 1197, 2001. 238. Gonzalez-Gay, M. A., and Garcia-Porrua, C.: Systemic vasculitis in adults in northwestern Spain, 1988–1997: clinical and epidemiological aspects. Medicine 78:292, 1999. 239. Gonzalez-Gay, M. A., and Garcia-Porrua, C.: Epidemiology of the vasculitides. Rheum. Dis. Clin. North Am. 27:729, 2001. 240. Gonzalez-Gay, M. A., and Garcia-Porrua, C.: HenochSchonlein purpura. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 476, 2002. 241. Gonzalez-Gay, M. A., Garcia-Porrua, C., Guerrero, J., et al.: The epidemiology of the primary systemic vasculitides in northwest Spain: implications of the Chapel Hill Consensus Conference definitions. Arthritis Rheum. 49:388, 2003. 242. Goraya, J. S., Jayashree, M., Ghosh, D., et al.: GuillainBarre syndrome in a child with Henoch-Schonlein purpura. Scand. J. Rheumatol. 27:310, 1998.

2392

Diseases of the Peripheral Nervous System

243. Gordon, M., Luqmani, R. A., Adu, D., et al.: Relapses in patients with a systemic vasculitis. Q. J. Med. 86:779, 1993. 244. Gorevic, P. D., Kassab, H. J., Levo, Y., et al.: Mixed cryoglobulinemia: clinical aspects and long-term follow-up of 40 patients. Am. J. Med. 69:287, 1980. 245. Goronzy, J. J.: T cells in vascular disease. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 57, 2002. 246. Goronzy, J. J., and Weyand, C. M.: Cytokines in giant-cell arteritis. Cleve. Clin. J. Med. 69(Suppl. 2):SII91, 2002. 247. Grant, I. A., Hunder, G. G., Homburger, H. A., and Dyck, P. J.: Peripheral neuropathy associated with sicca complex. Neurology 48:855, 1997. 248. Griem, P., Wulferink, M., Sachs, B., et al.: Allergic and autoimmune reactions to xenobiotics: how do they arise? Immunol. Today 19:133, 1998. 249. Griffin, J. W.: Vasculitic neuropathies. Rheum. Dis. Clin. North Am. 27:751, 2001. 250. Griffin, J. W., Cornblath, D. R., Alexander, E., et al.: Ataxic sensory neuropathy and dorsal root ganglionitis associated with Sjögren’s syndrome. Ann. Neurol. 27:304, 1990. 251. Griffith, M. E., and Pusey, C. D.: HLA genes in ANCAassociated vasculitides. Exp. Clin. Immunogenet. 14:196, 1997. 252. Griffith, M. E., and Pusey, C. D.: Cellular aspects of vasculitis—T cell-mediated aspects. Springer Semin. Immunopathol. 23:287, 2001. 253. Gross, W. L., Schnabel, A., and Reinhold-Keller, E.: Wegener’s granulomatosis: clinical aspects. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 358, 2002. 254. Grunewald, J., Halapi, E., Wahlstrom, J., et al.: T-cell expansions with conserved T-cell receptor beta chain motifs in the peripheral blood of HLA-DRB1*0401 positive patients with necrotizing vasculitis. Blood 92:3737, 1998. 255. Guillevin, L.: Polyarteritis nodosa and microscopic polyangiitis. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 300, 2002. 256. Guillevin, L., Cohen, P., Gayraud, M., et al.: Churg-Straus syndrome: clinical study and long-term follow-up of 96 patients. Medicine 78:26, 1999. 257. Guillevin, L., Cohen, P., Gayraud, M., et al.: Hepatitis B virus-related polyarteritis nodosa (HBV-PAN): follow-up of 104 patients [abstract]. Arthritis Rheum. 43(Suppl.):S369, 2001. 258. Guillevin, L., Cohen, P., Larroche, C., et al.: Treatment of HBV-related polyarteritis nodosa (PAN) with lamivudine and plasma exchanges: a prospective, multicenter, pilot trial in 10 patients [abstract]. Arthritis Rheum. 44(Suppl.):S271, 2001. 259. Guillevin, L., Cordier, J.-F., Lhote, F., et al.: A prospective, multicenter, randomized trial comparing steroids and pulse cyclophosphamide versus steroids and oral cyclophosphamide in the treatment of generalized Wegener’s granulomatosis. Arthritis Rheum. 40:2187, 1997. 260. Guillevin, L., Durand-Gasselin, B., Cevallos, R., et al.: Microscopic polyangiitis: clinical and laboratory findings in eighty-five patients. Arthritis Rheum. 42:421, 1999. 261. Guillevin, L., Fain, O., Lhote, F., et al.: Lack of superiority of steroids plus plasma exchange to steroids alone in the

262.

263.

264.

265.

266.

267.

268.

269.

270.

271.

272.

273.

274.

275.

treatment of polyarteritis nodosa and Churg-Strauss syndrome: a prospective, randomized trial in 78 patients. Arthritis Rheum. 35:208, 1992. Guillevin, L., Jarrousse, B., Lok, C., et al.: Long-term follow-up after treatment of polyarteritis nodosa and Churg-Strauss angiitis with comparison of steroids, plasma exchange and cyclophosphamide to steroids and plasma exchange: a prospective randomized trial of 71 patients. J. Rheumatol. 18:567, 1991. Guillevin, L., Lhote, F., Amouroux, F., et al.: Antineutrophil cytoplasmic antibodies, abnormal angiograms and pathological findings in polyarteritis nodosa and Churg-Strauss syndrome: indications for the classification of vasculitides of the polyarteritis nodosa group. Br. J. Rheumatol. 35:958, 1996. Guillevin, L., Lhote, F., and Cohen, P.: Churg-Strauss syndrome: clinical aspects. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 399, 2002. Guillevin, L., Lhote, F., Cohen, P., et al.: Corticosteroids plus pulse cyclophosphamide and plasma exchanges versus corticosteroids plus pulse cyclophosphamide alone in the treatment of polyarteritis nodosa and Churg-Strauss syndrome patients with factors predicting poor prognosis. Arthritis Rheum. 38:1638, 1995. Guillevin, L., Lhote, F., Cohen, P., et al.: Polyarteritis nodosa related to hepatitis B virus: a prospective study with long-term observation of 41 patients. Medicine 74:238, 1995. Guillevin, L., Lhote, F., Gayraud, M.., et al.: Prognostic factors in polyarteritis nodosa and Churg-Strauss syndrome: a prospective study in 342 patients. Medicine 75:17, 1996. Gupta, R., Singh, S., Tang, R., et al.: Anterior ischemic optic neuropathy caused by interferon alpha therapy. Am. J. Med. 112:683, 2002. Gupta, S., Piraka, C., and Jaffe, M.: Lamivudine in the treatment of polyarteritis nodosa associated with acute hepatitis B. N. Engl. J. Med. 344:1645, 2001. Hagen, E. C., Daha, M. R., Hermans, J., et al.: Diagnostic value of standardized assays for anti-neutrophil cytoplasmic antibodies in idiopathic systemic vasculitis. Kidney Int. 53:743, 1998. Hagen, E. C., de Keizer, R. J., Andrassy, K., et al.: Compassionate treatment of Wegener’s granulomatosis with rabbit anti-thymocyte globulin. Clin. Nephrol. 43:351, 1995. Hagiwara, N., Sekijima, Y., Hattori, T., and Ikeda, S.: Two patients with different types of vasculitic neuropathy—a comparison between cutaneous polyarteritis nodosa and nonsystemic vasculitic neuropathy [in Japanese]. Rinsho Shinkeigaku 43:102, 2003. Hahn, B. H.: Systemic lupus erythematosus. In Braunwald, E., Fauci, A. S., Kasper, D. L., et al. (eds.): Harrison’s Principles of Internal Medicine, 15th ed. New York, McGraw-Hill, p. 1922, 2001. Halperin, J., Luft, B. J., Volkman, D. J., and Dattwyler, R. J.: Lyme neuroborreliosis: peripheral nervous system manifestations. Brain 113:1207, 1990. Hamidou, M. A., El Koiri, D., Audrain, M., and Grolleau, J.-Y.: Systemic antineutrophil cytoplasmic antibody vasculitis

Neuropathies with Systemic Vasculitis

276.

277. 278. 279.

280.

281. 282.

283. 284.

285.

286.

287.

288.

289.

290.

291.

292.

293.

associated with lymphoid neoplasia. Ann. Rheum. Dis. 60:293, 2001. Han, W. K., Choi, H. K., Roth, R. M., et al.: Serial ANCA titers: useful tool for the prevention of relapses in ANCAassociated vasculitis. Kidney Int. 63:1079, 2003. Hancox, R. J., Karalus, N., and Singh, V.: Mononeuritis multiplex in leptospirosis. Scand. J. Infect. Dis. 23:395, 1991. Hankey, G. J.: Isolated angiitis/angiopathy of the central nervous system. Cerebrovasc. Dis. 1:2, 1991. Harati, Y., and Niakan, E.: The clinical spectrum of inflammatory-angiopathic neuropathy. J. Neurol. Neurosurg. Psychiatry 49:1313, 1986. Harper, L., Cockwell, P., Adu, D., and Savage, C. O.: Neutrophil priming and apoptosis in anti-neutrophil cytoplasmic autoantibody-associated vasculitis. Kidney Int. 59:1729, 2001. Harper, L., and Savage, C. O.: Pathogenesis of ANCAassociated systemic vasculitis. J. Pathol. 190:349, 2000. Hartung, H. P., Reiners, K., Michels, M., et al.: Serum levels of soluble E-selectin (ELAM-1) in immune-mediated neuropathies. Neurology 44:1153, 1994. Hasler, P., Kistler, H., and Gerber, H.: Vasculitides in hairy cell leukemia. Semin. Arthritis Rheum. 25:134, 1995. Hattori, M., Ito, K., Konomoto, T., et al.: Plasmapheresis as the sole therapy for rapidly progressive Henoch-Schonlein purpura nephritis in children. Am. J. Kidney Dis. 33:427, 1999. Hattori, N., Ichimura, M., Nagamatsu, M., et al.: Clinicopathological features of Churg-Strauss syndromeassociated neuropathy. Brain 122:427, 1999. Hattori, N., Mori, K., Misu, K., et al.: Mortality and morbidity in peripheral neuropathy associated with Churg-Strauss syndrome and microscopic polyangiitis. J. Rheumatol. 29:1408, 2002. Hatzis, G. S., Papachristodoulou, A., Delladetsima, I. K., and Moutsopoulos, H. M.: Polyarteritis nodosa associated with cholangiocarcinoma. Lupus 7:301, 1998. Haubitz, M., Koch, K. M., and Brunkhorst, R.: Cyclosporin for the prevention of disease reactivation in relapsing ANCA-associated vasculitis. Nephrol. Dial. Transplant. 13:2074, 1998. Haubitz, M., Schellong, S., Gobel, U., et al.: Intravenous pulse administration of cyclophosphamide versus daily oral treatment in patients with antineutrophil cytoplasmic antibody-associated vasculitis and renal involvement. Arthritis Rheum. 41:1835, 1998. Hawke, S. H., Davies, L., Pamphlett, R., et al.: Vasculitic neuropathy: a clinical and pathological study. Brain 114:2175, 1991. Hellman, D. B, Laing, T. J., Petri, M., et al.: Mononeuritis multiplex: the yield of evaluations for occult rheumatic diseases. Medicine 67:145, 1988. Hernandez-Rodriguez, J., Segarra, M., Vilardell, C., et al.: Elevated production of interleukin-6 is associated with a lower incidence of disease-related ischemic events in patients with giant-cell arteritis. Circulation 107:2428, 2003. Herrick, A. L., Oogarah, P., Brammah, T. B., et al.: Nervous system involvement in association with vasculitis and anticardiolipin antibodies in a patient with systemic sclerosis. Ann. Rheum. Dis. 53:349, 1994.

2393

294. Heuss, D., Probst-Cousin, S., Kayser, C., and Neundorfer, B.: Cell death in vasculitic neuropathy. Muscle Nerve 23:999, 2000. 295. Heuss, D., Schlotter-Weigel, B., Sommer, C., and the German Center for Neuromuscular Disease: Klinik, Diagnostik und Therapie der vaskulitischen Neuropathie: bundeseinheitliche Konsensuspapiere der Muskelzentren im Auftrag der Deutschen Gesellschaft fur Muskelkranke e. V. (DGM). Fortschr. Neurol. Psychiatr. 71:172, 2003. 296. Hietaharju, A., Jaaskelainen, S., Kalimo, H., and Hietarinta, M.: Peripheral neuromuscular manifestations in systemic sclerosis (scleroderma). Muscle Nerve 16:1204, 1993. 297. Hietaharju, A., Yli-Kerttula, U., Hakkinen, V., and Frey, H.: Nervous system manifestations in Sjogren’s syndrome. Acta Neurol. Scand. 81:144, 1990. 298. Ho, D. D., Rota, T. R., Schooley, R. T., et al.: Isolation of HTLV-III from cerebrospinal fluid and neural tissues of patients with neurologic syndromes related to the acquired immunodeficiency syndrome. N. Engl. J. Med. 313:1493, 1985. 299. Hochberg, M. C., Boyd, R. E., Ahearn, J. M., et al.: Systemic lupus erythematosus: a review of clinico-laboratory features and immunogenetic markers in 150 patients with emphasis on demographic subsets. Medicine 64:285, 1985. 300. Hoffman, G. C., Cid, M. C., Hellmann, D. B., et al.: A multicenter, randomized, double-blind, placebo-controlled trial of adjuvant methotrexate treatment for giant cell arteritis. Arthritis Rheum. 46:1309, 2002. 301. Hoffman, G. S.: Treatment of giant-cell arteritis: where we have been and why we must move on. Cleve. Clin. J. Med. 69(Suppl. 2):SII117, 2002. 302. Hoffman, G. S., Kerr, G. S., Leavitt, R. Y., et al.: Wegener granulomatosis: an analysis of 158 patients. Ann. Intern. Med. 116:488, 1992. 303. Hogan, S. L., Nachman, P. H., Wilkman, A. S., et al.: Prognostic markers in patients with antineutrophil cytoplasmic autoantibody-associated microscopic polyangiitis and glomerulonephritis. J. Am. Soc. Nephrol. 7:23, 1996. 304. Hogg, N., Henderson, R., Leitinger, B., et al.: Mechanisms contributing to the activity of integrins on leukocytes. Immunol. Rev. 186:164, 2002. 305. Hoi, A. Y., and Morand, E. F.: Churg-Strauss syndrome: the use of cyclophosphamide in mononeuritis. Rheumatology 42:390, 2003. 306. Hollander, D., and Manning, R. T.: The use of alkylating agents in the treatment of Wegener’s granulomatosis. Ann. Intern. Med. 67:393, 1967. 307. Horiuchi, I., Ochi, H., Mura, H., et al.: Th2 shift in mononeuritis multiplex and increase of Th2 cells in chronic inflammatory demyelinating polyneuropathy: an intracellular cytokine analysis. J. Neurol. Sci. 193:49, 2001. 308. Huang, D., Giscombe, R., Zhou, Y., and Lefvert, A. K.: Polymorphisms in CTLA-4 but not tumor necrosis factoralpha or interleukin 1beta genes are associated with Wegener’s granulomatosis. J. Rheumatol. 27:397, 2000. 309. Huang, D., Zhou, Y., and Hoffman, G. S.: Pathogenesis: immunogenetic factors. Best Pract. Res. Clin. Rheumatol. 15:239, 2001.

2394

Diseases of the Peripheral Nervous System

310. Huang, D. C., Yang, Y. H., Lin, Y. T., and Chiang, B. L.: Cyclosporin A therapy for steroid-dependent HenochSchonlein purpura. J. Microbiol. Immunol. Infect. 36:61, 2003. 311. Hubbard, A. K., and Rothlein, R.: Intercellular adhesion molecule-1 (ICAM-1) expression and cell signaling cascades. Free Radic. Biol. Med. 28:1379, 2000. 312. Hughes, R. A., Britton, T., and Richards, M.: Effects of lymphoma on the peripheral nervous system. J. R. Soc. Med. 87:526, 1994. 313. Hughes, R. A., Cameron, J. S., Hall, S. M., et al.: Multiple mononeuropathy as the initial presentation of systemic lupus erythematosus—nerve biopsy and response to plasma exchange. J. Neurol. 228:239, 1982. 314. Hunder, G. G., Arend, W. P., Bloch, D. A., et al.: The American College of Rheumatology 1990 criteria for the classification of vasculitis: introduction. Arthritis Rheum. 33:1101, 1990. 315. Hurley, M. E., Davids, J. R., and Mubarak, S. J.: Singleincision combination biopsy (muscle and nerve) in the diagnosis of neuromuscular disease in children. J. Pediatr. Orthop. 14:740, 1994. 316. Hussein, M., El Menyawi, M., Habashi, R., et al.: Hepatitis C virus related cryoglobulinemic vasculitis in Egypt [abstract]. Cleve. Clin. J. Med. 69(Suppl. 2):SII74, 2002. 317. Hutchinson, J.: Diseases of the arteries: on a peculiar form of thrombotic arteritis of the aged which is sometimes productive of gangrene. Arch. Surg. (Lond.) 1:323, 1890. 318. Hynes, R. O.: Integrins: bidirectional, allosteric signaling machines. Cell 110:673, 2002. 319. Irioka, T., Yamada, M., Yamawaki, M., et al.: Acute autonomic and sensory neuropathy after interferon -2b therapy for chronic hepatitis C. J. Neurol. Neurosurg. Psychiatry 70:408, 2001. 320. Iwamoto, Y., Okuda, B., Tachibana, H., and Sugita, M.: A case of allergic granulomatous angiitis with beneficial effects of plasma exchange [in Japanese]. Rinsho Shinkeigaku 37:115, 1997. 321. Jamieson, P. W., Guiliani, M. J., and Martinez, A. J.: Necrotizing angiopathy presenting with multifocal conduction blocks. Neurology 41:442, 1991. 322. Jayne, D.: Evidence-based treatment of vasculitis. Rheumatology 39:585, 2000. 323. Jayne, D., Rasmussen, N., Andrassy, K., et al.: A randomized trial of maintenance therapy for vasculitis associated with antineutrophil cytoplasmic autoantibodies. N. Engl. J. Med. 349:36, 2003. 324. Jayne, D. R.: Update on the European Vasculitis Study Group trials. Curr. Opin. Rheumatol. 13:48, 2001. 325. Jayne, D. R.: CAMPATH 1-H (anti-CD52) for refractory vasculitis: retrospective Cambridge experience 1989–1999 [abstract]. Cleve. Clin. J. Med. 69(Suppl. 2):SII129, 2002. 326. Jayne, D. R.: Considerations for novel therapies in the future. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 37, 2002. 327. Jayne, D. R., Chapel, H., Adu, D., et al.: Intravenous immunoglobulin for ANCA-associated systemic vasculitis with persistent disease activity. Q. J. Med. 93:433, 2000.

328. Jayne, D. R., Gaskin, G., Pusey, C. D., and Lockwood, C. M.: ANCA and predicting relapse in systemic vasculitis. Q. J. Med. 88:127, 1995. 329. Jayne, D. R., and Lockwood, C. M.: Pooled intravenous immunoglobulin in the management of systemic vasculitis. Adv. Exp. Med. Biol. 336:469, 1993. 330. Jennette, J. C., Falk, A. J., Andrassy, K., et al.: Nomenclature of systemic vasculitides: proposal of an international consensus conference. Arthritis Rheum. 37:187, 1994. 331. Jennette, J. C., Thomas, D. B., and Falk, R. J.: Microscopic polyangiitis (microscopic polyarteritis). Semin. Diagn. Pathol. 18:3, 2001. 332. Jennings, C. A., King, T. E., Tuder, R., et al.: Diffuse alveolar hemorrhage with underlying isolated, pauciimmune pulmonary capillaritis. Am. J. Respir. Crit. Care Med. 155:1101, 1997. 333. Johansen, M., Elling, P., Elling, H., and Olsson, A.: A genetic approach to the aetiology of giant cell arteritis: depletion of the CD8 T-lymphocyte subset in relatives of patients with polymyalgia rheumatica and arteritis temporalis. Clin. Exp. Rheumatol. 13:745, 1995. 334. Johnson, P. C., Rolak, L. A., Hamilton, R. H., and Laguna, J. F.: Paraneoplastic vasculitis of nerve: a remote effect of cancer. Ann. Neurol. 5:437, 1979. 335. Johnson, R. T., and Richardson, E. P.: The neurological manifestations of systemic lupus erythematosus: a clinicalpathological study of 24 cases and review of the literature. Medicine 47:337, 1968. 336. Johnston, B., and Butcher, E. C.: Chemokines in rapid leukocyte adhesion triggering and migration. Semin. Immunol. 14:83, 2002. 337. Jonasson, F., Cullen, J. F., and Elton, R. A.: Temporal arteritis: a 14-year epidemiological, clinical and prognostic study. Scott. Med. J. 24:111, 1979. 338. Jonsson, R., Haga, H.-J., and Gordon, T. P.: Sjogren’s syndrome. In Koopman W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 1737, 2001. 339. Jover, J. A., Hernandez-Garcia, C., Morado, I. C., et al.: Combined treatment of giant-cell arteritis with methotrexate and prednisone: a randomized, double-blind, placebocontrolled trial. Ann. Intern. Med. 134:106, 2001. 340. Kahaleh, M. B., and LeRoy, E. C.: Autoimmunity and vascular involvement in systemic sclerosis (SSc). Autoimmunity 31:195, 1999. 341. Kaklamani, V. G., Vaiopoulos, G., and Kaklamanis, P. G.: Behcet’s disease. Semin. Arthritis Rheum. 27:197, 1998. 342. Kallenberg, C. G., and Cohen Tervaert, J. W.: Autoantibodies in vasculitis. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 37, 2002. 343. Kallenberg, C. G., Rarok, A., and Stegeman, C. A.: Genetics of ANCA-associated vasculitides. Cleve. Clin. J. Med. 69(Suppl. 2):SII61, 2002. 344. Kaltreider, H. B., and Talal, N.: The neuropathy of Sjogren’s syndrome: trigeminal nerve involvement. Ann. Intern. Med. 70:751, 1969. 345. Karnam, U. S., and Reddy, K. R.: Pegylated interferons. Clin. Liver Dis. 7:139, 2003.

Neuropathies with Systemic Vasculitis 346. Kelkar, P., Masood, M., and Parry, G. J.: Distinctive pathologic findings in proximal diabetic neuropathy (diabetic amyotrophy). Neurology 55:83, 2000. 347. Kemper, J. W., Baggenstoss, A. H., and Slocumb, C. H.: The relationship of therapy with cortisone to the incidence of vascular lesions in rheumatoid arthritis. Ann. Intern. Med. 46:831, 1957. 348. Kernohan, J. W., and Woltman, H. W.: Periarteritis nodosa: a clinicopathological study with special reference to the nervous system. Arch. Neurol. Psychiatry 39:655, 1938. 349. Kevil, C. G., and Bullard, D. C.: Roles of leukocyte/ endothelial cell adhesion molecules in the pathogenesis of vasculitis. Am. J. Med. 106:677, 1999. 350. Kevil, C. G., and Bullard, D. C.: Cell adhesion molecules in the rheumatic diseases. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 478, 2001. 351. Kewalramani, R., and Singh, A. K.: Immunopathogenesis of lupus and lupus nephritis: recent insights. Curr. Opin. Nephrol. Hypertens. 11:273, 2002. 352. Khalili-Shirazi, A., Gregson, N. A., Londei, M., et al.: The distribution of CD1 molecules in inflammatory neuropathy. J. Neurol. Sci. 158:154, 1998. 353. Kidron, D., Barron, S. A., and Mazliah, J.: Mononeuritis multiplex with brachial plexus neuropathy coincident with Mycoplasma pneumoniae infection. Eur. Neurol. 29:90, 1989. 354. Kiene, M., Csernok, E., Muller, A., et al.: Elevated interleukin-4 and interleukin-13 production by T cell lines from patients with Churg-Strauss syndrome. Arthritis Rheum. 44:469, 2001. 355. Kimberly, R. P.: Characteristics of immune complexes and principles of immune complex disease. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 549, 2001. 356. Kimura, A., Ota, M., Katsuyama, Y., et al.: Mapping of the HLA-linked genes controlling the susceptibility to Takayasu’s arteritis. Int. J. Cardiol. 75(Suppl. 1):S105, 2000. 357. Kindstrand, E., Nilsson, B. Y., Hovmark, A., et al.: Polyneuropathy in late Lyme borreliosis—a clinical, neurophysiological and morphological description. Acta Neurol. Scand. 101:47, 2000. 358. Kissel, J. T., and Collins, M. P.: Vasculitic neuropathies and neuropathies of connective tissue disorders. In Katirji, B., Kaminski, H. J., Preston, D. C., et al. (eds.): Neuromuscular Disorders in Clinical Practice. Boston, Butterworth Heinemann, p. 669, 2002. 359. Kissel, J. T., Riethman, J. L., Omerza, J., et al.: Peripheral nerve vasculitis: immune characterization of the vascular lesions. Ann. Neurol. 25:291, 1989. 360. Kissel, J. T., Slivka, A. P., Warmolts, J. R., and Mendell, J. R.: The clinical spectrum of necrotizing angiopathy of the peripheral nervous system. Ann. Neurol. 18:251, 1985. 361. Klinger, H.: Grenzformen der Periarteritis nodosa. Frankfurt Z. Pathol. 42:455, 1931. 362. Knight, A., Askling, J., and Ekbom, A.: Cancer incidence in a population-based cohort of patients with Wegener’s granulomatosis. Int. J. Cancer 100:82, 2002.

2395

363. Kobayashi, M., Ito, M., Nakagawa, A., et al.: Neutrophil and endothelial cell activation in the vasa vasorum in vasculoBehcet disease. Histopathology 36:362, 2000. 364. Kone-Paut, I., Geisler, I., Wechsler, B., et al.: Familial aggregation in Behcet’s disease: high frequency in siblings and parents of pediatric probands. J. Pediatr. 35:89, 1999. 365. Kordossis, T., Paikos, S., Aroni, K., et al.: Prevalence of Sjogren’s-like syndrome in a cohort of HIV-positive patients: descriptive pathology and immunopathology. Br. J. Rheumatol. 37:691, 1998. 366. Kurzrock, R., Cohen, P. R., and Markowitz, A.: Clinical manifestations of vasculitis in patients with solid tumors: a case report and review of the literature. Arch. Intern. Med. 154:334, 1994. 367. Kussmaul, A., and Maier, R.: Ueber eine bisher nicht beschriebene eigenthumliche Arterien-erkrankung (Periarteritis nodosa), die mit Morbus Brightii und rapid fortschreitender allgemeiner Muskellahmung einhergeht. Dtsch. Arch. Klin. Med. 1:484, 1866. 368. Lam, K. C., Lai, C. L., Trepo, C., and Wu, P. C.: Deleterious effect of prednisolone in HBsAg-positive chronic active hepatitis. N. Engl. J. Med. 304:380, 1981. 369. Lamprecht, P., Gause, A., and Gross, W. L.: Cryoglobulinemic vasculitis. Arthritis Rheum. 42:2507, 1999. 370. Langford, C. A.: Complications of cyclophosphamide therapy. Eur. Arch. Otorhinolaryngol. 254:65, 1997. 371. Langford, C. A.: Management of systemic vasculitis. Best Pract. Res. Clin. Rheumatol. 15:281, 2001. 372. Langford, C. A.: Takayasu’s arteritis, giant cell arteritis, and polymyalgia rheumatica. In Weisman, M. H., Weinblatt, M. E., and Louie, J. S. (eds.): Treatment of the Rheumatic Diseases: Companion to Kelley’s Textbook of Rheumatology. Philadelphia, W. B. Saunders, p. 353, 2001. 373. Langford, C. A., Fauci, A. S., Talar-Williams, C., and Sneller, M. C.: Treatment of Wegener’s granulomatosis with methotrexate and glucocorticoids: update on rate of relapse [abstract]. Arthritis Rheum. 39(Suppl.):S211, 1996. 374. Langford, C. A., Talar-Williams, C., Barron, K. S., and Sneller, M. C.: Use of a cyclophosphamide-induction methotrexate-maintenance regimen for the treatment of Wegener’s granulomatosis: extended follow-up and rate of relapse. Am. J. Med. 114:463, 2003. 375. Lanham, J. G., Elkon, K. B., Pusey, C. D., and Hughes, G. R.: Systemic vasculitis with asthma and eosinophilia: a clinical approach to the Churg-Strauss syndrome. Medicine 63:65, 1984. 376. Lau, J. Y., Bird, G. L., Alexander, G. J., and Williams, R.: Effects of immunosuppressive therapy on hepatic expression of hepatitis B viral genome and gene products. Clin. Invest. Med. 16:226, 1993. 377. Lauer, G. M., and Walker, B. D.: Hepatitis C virus infection. N. Engl. J. Med. 345:41, 2001. 378. Lauta, V. M., and De Sangro, M. A.: Long-term results regarding the use of recombinant interferon alpha-2b in the treatment of II type mixed essential cryoglobulinemia. Med. Oncol. 12:223, 1995. 379. Leib, E. S., Restivo, C., and Paulus, H. E.: Immunosuppressive and corticosteroid therapy of polyarteritis nodosa. Am. J. Med. 67:941, 1979.

2396

Diseases of the Peripheral Nervous System

380. Leung, N.: Treatment of chronic hepatitis B: case selection and duration of therapy. J. Gastroenterol. 17:409, 2002. 381. Levine, G. B., Mills, P. E. Jr., and Epstein, W. V.: Circulating IgG globulin anti-IgG globulin complex in a patient with carcinoma of the lung and severe neuromyopathy. J. Lab. Clin. Med. 69:749, 1967. 382. Levy, M.: Familial cases of Berger’s disease and anaphylactoid purpura: more frequent than previously thought. Am. J. Med. 87:246, 1989. 383. Levy, Y., Sherer, Y., George, J., et al.: Serologic and clinical response to treatment of systemic vasculitis and associated autoimmune disease with intravenous immunoglobulin. Int. Arch. Allergy Immunol. 119:231, 1999. 384. Ley, K.: Functions of selectins. Results Probl. Cell Differ. 33:177, 2001. 385. Ley, K.: Integration of inflammatory signals by rolling neutrophils. Immunol. Rev. 186:8, 2002. 386. Lie, J. T.: Nomenclature and classification of vasculitis: plus ca change, plus c’est la meme chose. Arthritis Rheum. 37:181, 1994. 387. Lightfoot, R. W., Michel, B. A., Bloch, D. A., et al.: The American College of Rheumatology 1990 criteria for the classification of polyarteritis nodosa. Arthritis Rheum. 33:1088, 1990. 388. Lindenlaub, T., and Sommer, C.: Cytokines in sural nerve biopsies from inflammatory and non-inflammatory neuropathies. Acta Neuropathol. (Berl.) 105:593, 2003. 389. Lindstrom, P., and Brismar, T.: Mechanism of anoxic conduction block in mammalian nerve. Acta Physiol. Scand. 141:429, 1991. 390. Lipsky, P. E.: Rheumatoid arthritis. In Braunwald, E., Fauci, A. S., Kasper, D. L., et al. (eds.): Harrison’s Principles of Internal Medicine, 15th ed. New York, McGraw-Hill, p. 1928, 2001. 391. Lockwood, C. M.: Refractory Wegener’s granulomatosis: a model for shorter immunotherapy of autoimmune diseases. J. R. Coll. Physicians Lond. 32:473, 1998. 392. Lockwood, C. M., Elliot, J. D., Brettman, L., et al.: Antiadhesion molecule therapy as an interventional strategy for autoimmune inflammation. Clin. Immunol. 93:93, 1999. 393. Lotz, M. K.: Cytokines and their receptors. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 437, 2001. 394. Lovshin, L. L., and Kernohan, J. W.: Peripheral neuritis in periarteritis nodosa: a clinicopathologic study. Arch. Intern. Med. 82:321, 1948. 395. Ludviksson, B. R., Sneller, M. C., Chua, K. S., et al.: Active Wegener’s granulomatosis is associated with HLA-DR CD4 T cells exhibiting an unbalanced Th1-type T cell cytokine pattern: reversal with IL-10. J. Immunol. 160:3602, 1998. 396. Luqmani, R. A., Bacon, P. A., Beaman, M., et al.: Classical versus non-renal Wegener’s granulomatosis. Q. J. Med. 87:161, 1994. 397. Luqmani, R. A., and Robinson, H.: Introduction to, and classification of, the systemic vasculitides. Best Pract. Res. Clin. Rheumatol. 15:187, 2001.

398. Magrin, S., Craxi, A., Fabiano, C., et al.: Hepatitis C viremia in chronic liver disease: relationship to interferon-alpha or corticosteroid treatment. Hepatology 19:273, 1994. 399. Mahadevan, A., Gayathri, N., Taly, A. B., et al.: Vasculitic neuropathy in HIV infection: a clinicopathological study. Neurol. India 49:277, 2001. 400. Maimone, D., Villanova, M., Stanta, G., et al.: Detection of Borrelia burgdorferi DNA and complement membrane attack complex deposits in the sural nerve of a patient with chronic polyneuropathy and tertiary Lyme disease. Muscle Nerve 20:969, 1997. 401. Malinow, K., Yannakakis, G. D., Glusman, S. M., et al.: Subacute sensory neuronopathy secondary to dorsal root ganglionitis in primary Sjögren’s syndrome. Ann. Neurol. 20:535,1986. 402. Mannik, M.: Serum sickness and pathophysiology of immune complexes. In Rich, R. R., Fleisher, T. A., Schwartz, B. D., et al. (eds.): Clinical Immunology: Principles and Practice. St. Louis, Mosby, p. 217, 1996. 403. Manns, M. P., McHutchison, J. G., Gordon, S. C., et al.: Peginterferon alfa-2b plus ribavirin compared with interferon alpha-2b plus ribavirin for initial treatment of chronic hepatitis C: a randomized trial. Lancet 358:958, 2001. 404. Manoussakis, M. N., and Moutsopoulos, H. M.: Sjogren’s syndrome: current concepts. Adv. Intern. Med. 47:191, 2001. 405. Marcellin, P., Chang, T.-T., Lim, S. G., et al.: Adefovir dipivoxil for the treatment of hepatitis B e antigen-positive chronic hepatitis B. N. Engl. J. Med. 348:808, 2003. 406. Marcus, R.: Role of exercise in preventing and treating osteoporosis. Rheum. Dis. Clin. North Am. 27:131, 2001. 407. Mark, E. J., Matsubara, O., Tan-Liu, N. S., and Fienberg, R.: The pulmonary biopsy in the early diagnosis of Wegener’s (pathergic) granulomatosis: a study based on 35 open lung biopsies. Hum. Pathol. 19:1065, 1988. 408. Markusse, H. M., Vroom, T. M., Heurkens, A. H., and Vecht, C. J.: Polyneuropathy as initial manifestation of necrotizing vasculitis and gangrene in systemic lupus erythematosus. Neth. J. Med. 38:204, 1991. 409. Martinez-Taboada, V. M., Alonso, R. B., Armona, J., et al.: Mononeuritis multiplex in systemic lupus erythematosus: response to pulse intravenous cyclophosphamide. Lupus 5:74, 1996. 410. Martinez-Taboada, V. M., Blanco, R., Garcia-Fuentes, M., and Rodriguez-Valverde, V.: Clinical features and outcome of 95 patients with hypersensitivity vasculitis. Am. J. Med. 102:186, 1997. 411. Marzo, M. E., Tintore, M., Fabregues, O., et al.: Chronic inflammatory demyelinating polyneuropathy during treatment with interferon-alpha. J. Neurol. Neurosurg. Psychiatry 65:604, 1998. 412. Masi, A. T., Hunder, G. G., Lie, J. T., et al. : The American College of Rheumatology 1990 criteria for the classification of Churg-Strauss syndrome (allergic granulomatosis and angiitis). Arthritis Rheum. 33:1094, 1990. 413. Massari, M., Salvarani, C., Portioli, I., et al.: Polyarteritis nodosa and HIV infection: no evidence of a direct pathogenic role of HIV. Infection 24:159, 1996. 414. Matsumuro, K., Izumo, S., Umehara, F., et al.: Paraneoplastic vasculitic neuropathy: immunohistochemical studies

Neuropathies with Systemic Vasculitis

415.

416.

417.

418. 419.

420.

421.

422.

423.

424.

425.

426.

427.

428.

429.

430.

431.

on a biopsied nerve and post-mortem examination. J. Intern. Med. 236:225, 1994. Matteson, E. L., Gold, K. N., Bloch, D. A., and Hunder, G. G.: Long-term survival of patients with Wegener’s granulomatosis from the American College of Rheumatology Wegener’s granulomatosis classification criteria cohort. Am. J. Med. 101:129, 1996. Mattey, D. L., Hajeer, A. H., Dababneh, A., et al.: Association of giant cell arteritis and polymyalgia rheumatica with different tumor necrosis factor microsatellite polymorphisms. Arthritis Rheum. 43:1749, 2000. Mauch, E., Volk, C., Kratzsch, G., et al.: Neurological and neuropsychiatric dysfunction in primary Sjogren’s syndrome. Acta Neurol. Scand. 89:31, 1994. Mayes, M. D.: Classification and epidemiology of scleroderma. Semin. Cutan. Med. Surg. 17:22, 1998. Mazzaro, C., Laccin, T., Moretti, M., et al.: Effects of two different alpha-interferon regimens on clinical and virological findings in mixed cryoglobulinemia. Clin. Exp. Rheumatol. 13(Suppl.):S181, 1995. McCluskey, L., Feinberg, D., Cantor, C., and Bird, S.: “Pseudo-conduction block” in vasculitic neuropathy. Muscle Nerve 22:1361, 1999. McCombe, P. A., McLeod, J. G., Pollard, J. D., et al.: Peripheral sensorimotor and autonomic neuropathy associated with systemic lupus erythematosus: clinical, pathological and immunologic features. Brain 110:533, 1987. McDermott, E. M., and Powell, R. J.: Cyclosporin in the treatment of Churg-Strauss syndrome. Ann. Rheum. Dis. 57:258, 1998. McEver, R. P.: Adhesive interactions of leukocytes, platelets, and the vessel wall during hemostasis and inflammation. Thromb. Haemost. 86:746, 2001. McLeod, B. C., and Sassetti, R. J.: Plasmapheresis with return of cryoglobulin-depleted autologous plasma (cryoglobulinpheresis) in cryoglobulinemia. Blood 55:866, 1980. McMahon, B. J., Heyward, W. L., Templin, D. W., et al.: Hepatitis B-associated polyarteritis nodosa in Alaskan Eskimos: clinical and epidemiologic features and long-term follow-up. Hepatology 9:97, 1989. Meador, R., Ehrlich, G., and Von Feldt, J. M.: Behcet’s disease: immunopathologic and therapeutic aspects. Curr. Rheumatol. Rep. 4:47, 2002. Medvedev, G., Al-Shamari, A. E., Copland, M. A., and Magil, A. B.: Isolated renal giant cell arteritis. Am. J. Kidney Dis. 40:658, 2002. Meier, C., Grahmann, F., Engelhardt, A., and Dumas, M.: Peripheral nerve disorders in Lyme-borreliosis: nerve biopsy studies from eight cases. Acta Neuropathol. (Berl.) 79:271, 1989. Meller, S., Jagiello, P., Borgmann, S., et al.: Novel SNPs in the CD18 gene validate the association with MPO-ANCA vasculitis. Genes Immun. 2:269, 2001. Mellgren, S. I., Conn, D. L., Steven, J. C., and Dyck, P. J.: Peripheral neuropathy in primary Sjögren’s syndrome. Neurology 39:390, 1989. Merianos, P., Smyrnis, P., Tsomy, K., and Hager, J.: Giant cell arteritis of the median nerve simulating carpal tunnel syndrome. Hand 15:249, 1983.

2397

432. Merkel, P. A.: Drug-induced vasculitis. In Hoffman G. S., and Weyand C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 727, 2002. 433. Metzler, C., Low-Friedrich, I., Reinhold-Keller, E., et al.: Leflunomide, a new, promising agent in maintenance of remission in Wegener’s granulomatosis [abstract]. Clin. Exp. Immunol. 112(Suppl. 1):56, 1998. 434. Michel, B. E., Hunder, G. G., Bloch, D. A., and Calabrese, L. H.: Hypersensitivity vasculitis and Henoch-Schonlein purpura: a comparison between the 2 disorders. J. Rheumatol. 19:721, 1992. 435. Midroni, G., and Bilbao, J. M.: Biopsy Diagnosis of Peripheral Neuropathy. Boston, Butterworth-Heinemann, p. 241, 1995. 436. Migliaresi, S., and Tirri, G.: Interferon in the treatment of mixed cryoglobulinemia. Clin. Exp. Rheumatol. 13(Suppl.):S175, 1995. 437. Miller, A., Ryan, P. F., and Dowling, J. P.: Vasculitis and thrombotic thrombocytopenic purpura in a patient with limited scleroderma. J. Rheumatol. 24:598, 1997. 438. Minta, J. O., Winkler, C. J., Biggar, W. D., and Greenberg, M.: A selective and complete absence of C1q in a patient with vasculitis and nephritis. Clin. Immunol. Immunopathol. 22:225, 1982. 439. Misiani, R.: Virus-associated vasculitides: pathogenesis. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 553, 2002. 440. Misiani, R., Bellavita, P., Fenili, D., et al.: Interferon alfa-2a therapy in cryoglobulinemia associated with hepatitis C virus. N. Engl. J. Med. 330:751, 1994. 441. Mizuki, N., and Inoko, H.: Behcet’s syndrome: immunogenetics. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 433, 2002. 442. Mockel, M., Kampf, D., Lobeck, H., and Frei, U.: Severe panarteritis associated with drug abuse. Intensive Care Med. 25:113, 1999. 443. Molina, R., Provost, T. T., Arnett, F. C., et al.: Primary Sjogren’s syndrome in men: clinical, serologic, and immunogenetic features. Am. J. Med. 80:23, 1986. 444. Moll, J. W., Markusse, H. M., Pijnenburg, J. J., et al.: Antineuronal antibodies in patients with neurologic complications of primary Sjogren’s syndrome. Neurology 43:2574, 1993. 445. Molyneux, P. D., Barker, R., Thomas, P. K., et al.: Nonsystemic vasculitic neuropathy presenting with a painful polyradiculopathy: a case report. J. Neurol. 247:645, 2000. 446. Moore, M. E., Burke, J. M., Hartman, J. H., and Korenzwitt, E.: Lumbosacral plexopathy in a woman with CREST syndrome and vasculitis. Arthritis Rheum. 32:661, 1989. 447. Moore, P. M., and Fauci, A. S.: Neurologic manifestations of systemic vasculitis: a retrospective study of the clinicopathologic features and responses to therapy in 25 patients. Am. J. Med. 71: 517, 1981. 448. Moosig, F., Csernok, E., Kumanovics, G., and Gross, W. L.: Opsonization of apoptotic neutrophils by anti-neutrophil cytoplasmic antibodies (ANCA) leads to enhanced uptake by macrophages and increased release of tumour necrosis factoralpha (TNF-alpha). Clin. Exp. Immunol. 122:499, 2000.

2398

Diseases of the Peripheral Nervous System

449. Moreland, L. W., and O’Dell, J. R.: Glucocorticoids and rheumatoid arthritis: back to the future? Arthritis Rheum. 46:2553, 2002. 450. Morgan, S. L.: Calcium and vitamin D in osteoporosis. Rheum. Dis. Clin. North Am. 27:101, 2001. 451. Morrison, L. R., Short, C. L., Ludwig, A. O., and Schwab, R. S.: The neuromuscular system in rheumatoid arthritis: electromyographic and histologic observations. Am. J. Med. Sci. 214:33, 1947. 452. Morton, S. J., Lanyon, P. C., and Powell, R. J.: Etoposide in Wegener’s granulomatosis. Rheumatology 39:810, 2000. 453. Moudgil, S. S., Sripathi, N., and Newman, D.: Eosinophilic vasculitis in a patient on leukotriene inhibitors [abstract]. Muscle Nerve 26:573, 2002. 454. Moulignier, A., Authier, F.-J., Baudrimont, M., et al.: Peripheral neuropathy in human immunodeficiency virusinfected patients with the diffuse infiltrative lymphocytosis syndrome. Ann. Neurol. 41:438, 1997. 455. Muller, A., Trabandt, A., Gloeckner-Hofmann, K., et al.: Localized Wegener’s granulomatosis: predominance of CD26 and INF-gamma production. J. Pathol. 192:113, 2000. 456. Muller, W. A.: Leukocyte-endothelial cell interactions in the inflammatory response. Lab. Invest. 82:521, 2002. 457. Muller Kobold, A. C., van Wijk, R. T., Franssen, C. F., et al.: In vitro up-regulation of E-selectin and induction of interleukin-6 in endothelial cells by autoantibodies in Wegener’s granulomatosis and microscopic polyangiitis. Clin. Exp. Rheumatol. 17:433, 1999. 458. Murai, H., Inaba, S., Kira, J., et al.: Hepatitis C virus associated cryoglobulinemic neuropathy successfully treated with plasma exchange. Artif. Organs 19:334, 1995. 459. Murthy, J. M., Sundaram, C., and Meena, A. K.: Nonsystemic vasculitic neuropathy. J. Assoc. Physicians India 46:204, 1998. 460. Murty, G. E., Mains, B. T., Middleton, D., et al.: HLA antigen frequencies and Wegener’s granulomatosis. Clin. Otolaryngol. 16:448, 1991. 461. Myles, A. B., Perera, T., and Ridley, M. G.: Prevention of blindness in giant cell arteritis by corticosteroid treatment. Br. J. Rheumatol. 31:103, 1992. 462. Nachman, P. H., Hogan, S. L., Jennette, J. C., and Falk, R. J.: Treatment response and relapse in antineutrophil cytoplasmic autoantibody-associated microscopic polyangiitis and glomerulonephritis. J. Am. Soc. Nephrol. 7:33, 1996. 463. Nachman, P. H., Joy, M. S., Hogan, S. L., et al.: Mycophenolate mofetil (MMF): preliminary results of a feasibility trial in relapsing ANCA small vessel vasculitis [abstract]. Clin. Exp. Immunol. 120(Suppl. 1):72, 2000. 464. Nakamura, H., Kawakami, A., Tominaga, M., et al.: Relationship between Sjogren’s syndrome and human T-lymphotropic virus type I infection: follow-up study of 83 patients. J. Lab. Clin. Med. 135:139, 2000. 465. Namer, I. J., Karabudak, R., Zileh, T., et al.: Peripheral nervous system involvement in Behçet’s disease: case report and review of the literature. Eur. Neurol. 26:235, 1987. 466. Negoro, K., Fukusako, T., Morimatsu, M., and Liao, C.: Acute axonal polyneuropathy during interferon--2a therapy for chronic hepatitis C. Muscle Nerve 17:1351, 1994.

467. Nepom, B. S.: Structure, function, and genetics of the HLA complex in rheumatic disease. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 578, 2001. 468. Nesher, G.: Neurologic manifestations of giant cell arteritis. Clin. Exp. Rheumatol. 18(Suppl. 20):S24, 2000. 469. Nesher, G.: Giant cell arteritis and polymyalgia rheumatica. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 255, 2002. 470. Nesher, G., Rosenberg, P., Shorer, Z., et al.: Involvement of the peripheral nervous system in temporal arteritispolymyalgia rheumatica: report of 3 cases and review of the literature. J. Rheumatol. 14:358, 1987. 471. Nesher, G., Rubinow, A., and Sonnenblick, M.: Efficacy and adverse effects of different corticosteroid dose regimens in temporal arteritis: a retrospective study. Clin. Exp. Rheumatol. 15:303, 1997. 472. Nicolai, A., Bonetti, B., Lazzarino, L. G., et al.: Peripheral nerve vasculitis: a clinico-pathological study. Clin. Neuropathol. 14:137, 1995. 473. Nishino, H., Rubino, F. A., DeRemee, R. A., et al.: Neurological involvement in Wegener’s granulomatosis: an analysis of 324 consecutive patients at the Mayo Clinic. Ann. Neurol. 33:4, 1993. 474. Noth, I., Strek, M. E., and Leff, A. R.: Churg-Strauss syndrome. Lancet 361:587, 2003. 475. Nowack, R., Gobel, U., Klooker, P., et al.: Mycophenolate mofetil for maintenance therapy of Wegener’s granulomatosis and microscopic polyangiitis: a pilot study in 11 patients with renal involvement. J. Am. Soc. Nephrol. 10:1965, 1999. 476. Nowack, R., Lehmann, H., Flores-Suarez, L. F., et al.: Familial occurrence of systemic vasculitis and rapidly progressive glomerulonephritis. Am. J. Kidney Dis. 34:364, 1999. 477. Nukada, H., and Dyck, P. J.: Acute ischemia causes axonal stasis, swelling, attenuation, and secondary demyelination. Ann. Neurol. 22:311, 1987. 478. Nukada, H., McMorran, P. D., and Shimizu, J.: Acute inflammatory demyelination in reperfusion nerve injury. Ann. Neurol. 47:71, 2000. 479. Numano, F., Kobayash, Y., Maruyama, Y., et al.: Takayasu arteritis: clinical characteristics and the role of genetic factors in its pathogenesis. Vasc. Med. 1:227, 1996. 480. Nuzum, J. W.: Polyarteritis nodosa: statistical review of one hundred seventy-five cases from the literature and report of a “typical” case. Arch. Intern. Med. 94:942, 1954. 481. Oddis, C. V., Eisenbeis, C. H., Reidbord, H. E., et al.: Vasculitis in systemic sclerosis: association with Sjögren’s syndrome and the CREST syndrome variant. J. Rheumatol. 14:942, 1987. 482. O’Duffy, J. D., Carney, J. A., and Deodhar, S.: Behçet’s disease: report of 10 cases, 3 with new manifestations. Ann. Intern. Med. 75:561, 1971. 483. Ognibene, F. P., Shelhaamer, J. H., Hoffman, G. S., et al.: Pneumocystis carinii pneumonia: a major complication of immunosuppressive therapy in patients with Wegener’s granulomatosis. Am. J. Respir. Crit. Care Med. 151:795, 1995.

Neuropathies with Systemic Vasculitis 484. Oh, S. J.: Paraneoplastic vasculitis of the peripheral nervous system. Neurol. Clin. 15:849, 1997. 485. Oh, S. J., Slaughter, R., and Harrell, L.: Paraneoplastic vasculitic neuropathy: a treatable neuropathy. Muscle Nerve 14:152, 1991. 486. Ohkoshi, N., Mizusawa, H., Oguni, E., and Shoji, S.: Sural nerve biopsy in vasculitic neuropathies: morphometric analysis of the caliber of involved vessels. J. Med. 27:153, 1996. 487. Oka, N., Akiguchi, I., Kawasaki, T., et al.: Tumor necrosis factor- in peripheral nerve lesions. Acta Neuropathol. (Berl.) 95:57, 1998. 488. Olney, R. K.: AAEM minimonograph #38: neuropathies in connective tissue disease. Muscle Nerve 15:531, 1992. 489. Olson, T. S., and Ley, K.: Chemokines and chemokines receptors in leukocyte trafficking. Am. J. Physiol. Regulatory Integrative Comp. Physiol. 283:R7, 2002. 490. Omdal, R., Henriksen, O. A., Mellgren, S. I., and Husby, G.: Peripheral neuropathy in systemic lupus erythematosus. Neurology 41:808, 1991. 491. Oschmann, P., Dorndorf, W., Hornig, C., et al.: Stages and syndromes of neuroborreliosis. J. Neurol. 245:262, 1998. 492. Ozen, H., Cemeroglu, P., Ecevit, Z., et al.: Unusual neurologic complications of typhoid fever (aphasia, mononeuritis multiplex, and Guillain-Barre syndrome): a report of two cases. Turk. J. Pediatr. 35:141, 1993. 493. Pallis, C. A., and Scott, J. T.: Peripheral neuropathy in rheumatoid arthritis. Br. Med. J. 1:1141, 1965. 494. Pamphlett, R., and Walsh, J.: Infective endocarditis with inflammatory lesions in the peripheral nervous system. Acta Neuropathol. (Berl.) 78:101, 1989. 495. Panegyres, P. K., Blumbergs, P. C., Leong, A. S., and Bourne, A. J.: Vasculitis of peripheral nerve and skeletal muscle: clinicopathological correlation and immunopathogenic mechanisms. J. Neurol. Sci. 100:193, 1990. 496. Papiha, S. S., Murty, G. E., Ad’Hia, A., et al.: Association of Wegener’s granulomatosis with HLA antigens and other genetic markers. Ann. Rheum. Dis. 51:246, 1992. 497. Papo, T., Le Thi Huong, D., Wiederkehr, J. L., et al.: Etoposide in Wegener’s granulomatosis. Rheumatology 38:473, 1999. 498. Parry, G. J.: AAEE case report #11: mononeuropathy multiplex. Muscle Nerve 8:493, 1985. 499. Patel, K. D., Cuvelier, S. L., and Wiehler, S.: Selectins: critical mediators of leukocyte recruitment. Semin. Immunol. 14:73, 2002. 500. Pateron, D., Fain, O., Sehonnou, J., et al.: Severe necrotizing vasculitis in a patient with hepatitis C virus infection treated with interferon. Clin. Exp. Rheumatol. 14:79, 1996. 501. Pavlidis, N. A., Karsh, J., and Moutsopoulos, H. M.: The clinical picture of primary Sjogren’s syndrome: a retrospective study. J. Rheumatol. 9:685, 1982. 502. Perdriger, A., Chales, G., Semana, G., et al.: Role of HLA-DR-DR and DR-DQ associations in the expression of extraarticular manifestations and rheumatoid factor in rheumatoid arthritis. J. Rheumatol. 24:1272, 1997. 503. Petri, M. A.: Systemic lupus erythematosus: clinical aspects. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 1455, 2001.

2399

504. Praprotnik, S., Blank, M., Meroni, P. L., et al.: Classification of anti-endothelial cell antibodies into antibodies against microvascular and macrovascular endothelial cells: the pathogenic and diagnostic implications. Arthritis Rheum. 44:1484, 2001. 505. Prayson, R. A., and Sedlock, D. J.: Clinicopathologic study of 43 patients with sural nerve vasculitis. Hum. Pathol. 34:484, 2003. 506. Preston, G. A., Yang, J. J., Xiao, H., and Falk, R. J.: Understanding the pathogenesis of ANCA: where are we today? Cleve. Clin. J. Med. 69(Suppl. 2):II51, 2002. 507. Prior, R., Schober, R., Scharffetter, K., and Wechsler, W.: Occlusive microangiopathy by immunoglobulin (IgM-kappa) precipitation: pathogenetic relevance in paraneoplastic cryoglobulinemic neuropathy. Acta Neuropathol. (Berl.) 83:423, 1992. 508. Puechal, X., Said, G., Hilliquin, P., et al.: Peripheral neuropathy with necrotizing vasculitis in rheumatoid arthritis: a clinicopathologic and prognostic study of thirty-two patients. Arthritis Rheum. 38:1618, 1995. 509. Pusey, C. D.: Microscopic polyangiitis. Clin. Exp. Immunol. 120(Suppl. 1):32, 2000. 510. Pusey, C. D., Rees, A. J., Evans, D. J., et al.: Plasma exchange in focal necrotizing glomerulonephritis without anti-GBM antibodies. Kidney Int. 40:757, 1991. 511. Quattrini, A., Comi, G., Nemni, R., et al.: Axonal neuropathy associated with interferon- treatment for hepatitis C: HLA-DR immunoreactivity in Schwann cells. Acta Neuropathol. (Berl.) 94:504, 1997. 512. Quinet, R. J., Zakem, J. M., and McCain, M.: Localized versus systemic vasculitis: diagnosis and management. Curr. Rheumatol. Rep. 5:93, 2003. 513. Radford, D. J., Lord, J. M., and Savage, C. O.: The activation of the neutrophil respiratory burst by antineutrophil cytoplasm autoantibody (ANCA) from patients with systemic vasculitis requires tyrosine kinases and protein kinase C activation. Clin. Exp. Immunol. 118:171, 1999. 514. Radford, D. J., Luu, T., Hewins, P., et al.: Antineutrophil cytoplasmic antibodies stabilize adhesion and promote migration of flowing neutrophils on endothelial cells. Arthritis Rheum. 44:2851, 2001. 515. Rao, J. K., Allen, N. B., and Pincus, T.: Limitations of the 1990 American College of Rheumatology classification criteria for the diagnosis of vasculitis. Ann. Intern. Med. 129:345, 1998. 516. Rappaport, W. D., Valenti, J., Hunter, G. C., et al.: Clinical utilization and complications of sural nerve biopsy. Am. J. Surg. 166:252, 1993. 517. Rarok, A. A., Stegeman, C. A., Limburg, P. C., and Kallenberg, C. G.: Neutrophil membrane expression of proteinase 3 (PR3) is related to relapse in PR3-ANCAassociated vasculitis. J. Am. Soc. Nephrol. 13:2232, 2002. 518. Rauova, L., Gilburd, B., Zurgil, N., et al.: Induction of biologically active antineutrophil cytoplasmic antibodies by immunization with human apoptotic polymorphonuclear leukocytes. Clin. Immunol. 103:69, 2002. 519. Read, S. J., Crawford, D. H., and Pender, M. P.: Trigeminal sensory neuropathy induced by interferon-alpha therapy. Aust. N. Z. J. Med. 25:54, 1995.

2400

Diseases of the Peripheral Nervous System

520. Regan, M. J., Hellmann, D. B., and Stone, J. H.: Treatment of Wegener’s granulomatosis. Rheum. Dis. Clin. North Am. 27:863, 2001. 521. Reid, A. J., Harrison, B. D., Watts, R. A., et al.: ChurgStrauss syndrome in a district hospital. Q. J. Med. 91:219, 1998. 522. Reinhold-Keller, E., Beuge, N., Latza, U., et al.: An interdisciplinary approach to the care of patients with Wegener’s granulomatosis: long-term outcome in 155 patients. Arthritis Rheum. 43:1021, 2000. 523. Reinhold-Keller, E., Herlyn, K., Wagner-Bastmeyer, R., et al.: No difference in the incidences of vasculitides between north and south Germany: first results of the German Vasculitis Register. Rheumatology 41:540, 2002. 524. Rekvig, O. P., and Nossent, J. C.: Anti-double-stranded DNA antibodies, nucleosomes, and systemic lupus erythematosus: a time for new paradigms? Arthritis Rheum. 48:300, 2003. 525. Richmond, R., McMillan, T. W., and Luqmani, R. A.: Optimisation of cyclophosphamide therapy in systemic vasculitis. Clin. Pharmacokinet. 34:79, 1998. 526. Richter, C., Schnabel, A., Csernok, E., et al.: Treatment of anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis with high-dose intravenous immunoglobulin. Clin. Exp. Immunol. 101:2, 1995. 527. Richter, R. B.: Peripheral neuropathy and connective tissue disease. J. Neuropathol. Exp. Neurol. 13:168, 1954. 528. Rivest, D., Brunet, D., Desbiens, R., and Bouchard, J.-P.: C-5 radiculopathy as a manifestation of giant cell arteritis. Neurology 45:1222, 1995. 529. Rivkina, A., and Rybalov, S.: Chronic hepatitis B: current and future treatment options. Pharmacotherapy 22:721, 2002. 530. Robson, W. L., and Leung, A. K.: Henoch-Schonlein purpura. Adv. Pediatr. 41:163, 1994. 531. Rodgers, H., Guthrie, J. A., Brownjohn, A. M., and Turney, J. H.: Microscopic polyarteritis: clinical features and treatment. Postgrad. Med. J. 65:515, 1989. 532. Rose, D. M., Han, J., and Ginsberg, M. H.: Alpha4 integrins and the immune response. Immunol. Rev. 186:118, 2002. 533. Rosenberg, J. L., Edlow, D., and Sneider, R.: Liver disease and vasculitis in a patient taking cromolyn. Arch. Intern. Med. 138:989, 1978. 534. Rottem, M., Cotch, M. F., Fauci, A. S., and Hoffman, G. S.: Familial vasculitis: report of 2 families. J. Rheumatol. 21:561, 1994. 535. Roullet, E., Assuerus, V., Gozlan, J., et al.: Cytomegalovirus multifocal neuropathy in AIDS: analysis of 15 consecutive cases. Neurology 44:2174, 1994. 536. Russell, J. H., and Ley, T. J.: Lymphocyte-mediated cytotoxicity. Annu. Rev. Immunol. 20:323, 2002. 537. Russell, K. A., and Specks, U.: Are antineutrophil cytoplasmic antibodies pathogenic? Experimental approaches to understand the antineutrophil cytoplasmic antibody phenomenon. Rheum. Dis. Clin. North Am. 27:815, 2001. 538. Russell, R. W.: Giant-cell arteritis. A review of 35 cases. Q. J. Med. 28:471, 1959. 539. Said, G.: Vasculitic neuropathy. Baillieres Clin. Neurol. 4:489, 1995.

540. Said, G., Elgrably, F., LaCroix, C., et al.: Painful proximal diabetic neuropathy: inflammatory nerve lesions and spontaneous favorable outcome. Ann. Neurol. 41:762, 1997. 541. Said, G., Goulon-Goeau, C., Lacroix, C., and Moulonguet, A.: Nerve biopsy findings in different patterns of proximal diabetic neuropathy. Ann. Neurol. 35:559, 1994. 542. Said, G., Lacroix, C., Chemouilli, P., et al.: Cytomegalovirus neuropathy in acquired immunodeficiency syndrome: a clinical and pathological study. Ann. Neurol. 29:139, 1991. 543. Said, G., Lacroix, C., Lozeron, P., et al.: Inflammatory vasculopathy in multifocal diabetic neuropathy. Brain 126:376, 2003. 544. Said, G., LaCroix-Ciaudo, C., Fujimura, H., et al.: The peripheral neuropathy of necrotizing arteritis: a clinicopathological study. Ann. Neurol. 23:461, 1988. 545. Sakane, T., Takeno, M., Suzuki, N., and Inaba, G.: Behcet’s disease. N. Engl. J. Med. 341:1284, 1999. 546. Salvarani, C., Cantini, F., Boiardi, L., and Hunder, G. G.: Polymyalgia rheumatica and giant-cell arteritis. N. Engl. J. Med. 347:261, 2002. 547. Salvarani, C., Casali, B., Boiardi, L., et al.: Intercellular adhesion molecule 1 gene polymorphisms in polymyalgia rheumatica/giant cell arteritis: association with disease risk and severity. J. Rheumatol. 27:1215, 2000. 548. Salvarani, C., and Hunder, G. G.: Giant cell arteritis with low erythrocyte sedimentation rate: frequency of occurrence in a population-based study. Arthritis Rheum. 45:140, 2001. 549. Sams, H., and Sams, W. M. Jr.: Cutaneous leukocytoclastic vasculitis. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 467, 2002. 550. Sanchez, J., Coll Canti, J., Ariza, A., et al.: Neuropatia por vasculitis necrotizante: estudio anatomoclinico, neurofisiologico y evolucion terapeutica de 26 pacientes. Rev. Neurol. 33:1033, 2001. 551. Sansonno, D., De Re, V., Lauletta, G., et al.: Monoclonal antibody treatment of mixed cryoglobulinemia resistant to interferon  with an anti-CD20. Blood 101:3818, 2003. 552. Saraux, H., Le Hoang, P., and Laroche, L.: Neuropathie optique ischeemique aigue anterieure et posterieure au cours d’une peri-arterite noueuse: traitement par plasmapherese. J. Fr. Ophtalmol. 5:55, 1982 553. Satoh, J., Yamamoto, T., and Kuroda, Y.: Vasculitis confined to the peripheral nerve and skin: a variant of non-systemic vasculitic neuropathy. Eur. J. Neurol. 6:241, 1999. 554. Satoi, H., Oka, N., Kawasaki, T., et al.: Mechanisms of tissue injury in vasculitic neuropathies. Neurology 50:492, 1998. 555. Saulsbury, F. T.: Henoch-Schonlein purpura in children: report of 100 patients and review of the literature. Medicine 78:395, 1999. 556. Savage, C. O., Pottinger, B. E., Gaskin, G., et al.: Vascular damage in Wegener’s granulomatosis and microscopic polyarteritis: presence of anti-endothelial cell antibodies and their relation to anti-neutrophil cytoplasm antibodies. Clin. Exp. Immunol. 85:14, 1991. 557. Savage, C. O., Winearls, C. G., Evans, D. J., et al.: Microscopic polyarteritis: presentation, pathology, and prognosis. Q. J. Med. 56:467, 1985.

Neuropathies with Systemic Vasculitis 558. Savige, J., Gillis, D., Benson, E., et al.: International consensus statement on testing and reporting of antineutrophil cytoplasmic antibodies (ANCA). Am. J. Clin. Pathol. 111:507, 1999. 559. Savige, J. A., Chang, L., and Crowe, S. M.: Anti-neutrophil cytoplasm antibodies in HIV infection. Adv. Exp. Med. Biol. 336:349, 1993. 560. Sawitzke, A. D., and Cannon, G. W.: Cytotoxic drugs. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 794, 2001. 561. Sbai, A., Wechsler, B., Duhaut, D. L., et al.: Neurological complications (dural sinus thrombosis excluded) in Behcet’s disease: evaluation of 109 patients [abstract]. Arthritis Rheum. 46(Suppl.):S180, 2002. 562. Schady, W., Sheard, A., Hassell, A., et al.: Peripheral nerve dysfunction in scleroderma. Q. J. Med. 80:661, 1991. 563. Schafers, M., Neukirchen, S., Toyka, K. V., and Sommer, C.: Borrelia neuropathy: histologic and immunohistochemical characterization [abstract]. J. Peripher. Nerv. Syst. 6:175, 2001. 564. Schapira, D., Balbir-Gurman, A., and Nahir, A. M.: Naproxen-induced leukocytoclastic vasculitis. Clin. Rheumatol. 19:242, 2000. 565. Schenone, A., DeMartini, I., Tabaton, M., et al.: Direct immunofluorescence in sural nerve biopsies. Eur. Neurol. 28:262, 1988. 566. Schifitto, G., Barbano, R. L., Kieburtz, K. D., et al.: HIV related vasculitic mononeuropathy: a role for IVIg? J. Neurol. Neurosurg. Psychiatry 63:255, 1997. 567. Schleiffer, T., Burkhard, B., Klooker, P., and Brass, H.: Clinical course and symptomatic prediagnostic period of patients with Wegener’s granulomatosis and microscopic polyangiitis. Ren. Fail. 20:519, 1998. 568. Schonermarck, U., Csernok, E., Trabandt, A., et al.: Circulating cytokines and soluble CD23, CD26 and CD30 in ANCA-associated vasculitides. Clin. Exp. Rheumatol. 18:457, 2000. 569. Schrier, R. W., Bulger, R. J., and VanArsdel, P. P.: Nephropathy associated with penicillin and homologues. Ann. Intern. Med. 64:116, 1966. 570. Schutz, G., and Schroder, J. M.: Number and size of epineurial blood vessels in normal and diseased human nerves. Cell Tissue Res. 290:31, 1997. 571. Scott, D. G., Bacon, P. A., Bothamley, J. E., et al.: Plasma exchange in rheumatoid vasculitis. J. Rheumatol. 8:433, 1981. 572. Scott, D. G., Bacon, P. A., and Tribe, C. R.: Systemic rheumatoid vasculitis: a clinical and laboratory study of 50 cases. Medicine 60:288, 1981. 573. Segelmark, M., Elzouki, A. N., Wieslander, J., and Eriksson, S.: The PiZ gene of alpha 1-antitrypsin as a determinant of outcome in PR3-ANCA-positive vasculitis. Kidney Int. 48:844, 1995. 574. Segelmark, M., Persson, U., Westman, K. W., et al.: Polymorphisms of candidate genes in ANCA positive vasculitis [abstract]. Cleve. Clin. J. Med. 69(Suppl. 2):SII155, 2002. 575. Sehgal, M., Swanson, J. W., DeRemee, R. A., and Colby, T. V.: Neurologic manifestations of Churg-Strauss syndrome. Mayo Clin. Proc. 70:337, 1995.

2401

576. Seibold, J. R.: Scleroderma. In Koopman, W. J. (ed.): Arthritis and Allied Conditions: A Textbook of Rheumatology, 14th ed. Philadelphia, Lippincott Williams & Wilkins, p. 1211, 2001. 577. Sepkowitz, K. A.: Pneumocystis carinii pneumonia without acquired immunodeficiency syndrome: who should receive prophylaxis? Mayo Clin. Proc. 71:102, 1996. 578. Sergent, J. S., Lockshin, M. D., Christial, C. L., and Gocke, D. J.: Vasculitis with hepatitis B antigenemia: long-term observations in nine patients. Medicine 55:1, 1976. 579. Serra, A., Cameron, J. S., and Turner, D. R.: Vasculitis affecting the kidney: presentation, histopathology and longterm outcome. Q. J. Med. 53:181, 1984. 580. Shulman, L. M., Minagar, A., Sharma, K., and Weiner, W. J.: Amantadine-induced peripheral neuropathy. Neurology 53:1862, 1999. 581. Siagris, D., Pharmakakis, N., Christofidou, M., et al.: Keratoconjunctivitis sicca and chronic HCV infection. Infection 30:229, 2002. 582. Silman, A. J.: Scleroderma—demographics and survival. J. Rheumatol. 48(Suppl.):58, 1997. 583. Singhal, B. S., Khadilkar, S. V., Gursahani, R. D., and Surya, N.: Vasculitic neuropathy: profile of twenty patients. J. Assoc. Physicians India 43:459, 1995. 584. Sinico, R. A., Fornasieri, A., Fiorini, G., et al.: Plasma exchange in the treatment of essential mixed cryoglobulinemia nephropathy: long-term follow up. Int. J. Artif. Organs 8(Suppl. 2):15, 1984. 585. Skopouli, F. N., Dafni, U., Ioannidis, J. P., and Moutsopoulos, H. M.: Clinical evolution, and morbidity and mortality of primary Sjögren’s syndrome. Semin. Arthritis Rheum. 29:296, 2000. 586. Sladky, J. T., Tschoepe, R. L., Greenberg, J. H., and Brown, M. J.: Peripheral neuropathy after chronic endoneurial ischemia. Ann. Neurol. 29:272, 1991. 587. Smith, J. B., and Haynes, M. K.: Rheumatoid arthritis—a molecular understanding. Ann. Intern. Med. 136:908, 2002. 588. Sneller, M. C.: Granuloma formation: implications for the pathogenesis of vasculitis. Cleve. Clin. J. Med. 69(Suppl. 2): SII40, 2002. 589. Sneller, M. C.: Vasculitis secondary to bacterial, fungal, and parasitic infection. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 599, 2002. 590. Sneller, M. C., and Fauci, A. S.: Pathogenesis of vasculitic syndromes. Med. Clin. North Am. 81:221, 1997. 591. Sneller, M. C., Hoffman, G. S., Talar-Williams, C., et al.: An analysis of forty-two Wegener’s granulomatosis patients treated with methotrexate and prednisone. Arthritis Rheum. 38:608, 1995. 592. So, Y. T., and Olney, R. K.: The natural history of mononeuropathy multiplex and simplex in patients with HIV infection [abstract]. Neurology 41(Suppl. 1):375, 1991. 593. Sobue, G., Yasuda, T., Kachi, T., et al.: Chronic progressive sensory ataxic neuropathy: clinicopathological features of idiopathic and Sjogren’s syndrome-associated cases. J. Neurol. 240:1, 1993.

2402

Diseases of the Peripheral Nervous System

594. Solans, R., Bosch, J. A., Perez-Bocanegra, C., et al.: ChurgStrauss syndrome: outcome and long-term follow-up of 32 patients. Rheumatology 40:763, 2001. 595. Somer, T., and Finegold, S. M.: Vasculitides associated with infections, immunization, and antimicrobial drugs. Clin. Infect. Dis. 20:1010, 1995. 596. Specks, U., Fervenza, F. C., McDonald, T. J., and Hogan, M. C.: Response of Wegener’s granulomatosis to anti-CD20 chimeric monoclonal antibody therapy. Arthritis Rheum. 44:2836, 2001. 597. Spiera, R. F., Mitnick, H. J., Kupersmith, M., et al.: A prospective, double-blind, randomized, placebo controlled trial of methotrexate in the treatment of giant cell arteritis (GCA). Clin. Exp. Rheumatol. 19:495, 2001. 598. Stafford, C. R., Bogdanoff, B. M., Green, L., and Spector, H. B.: Mononeuropathy multiplex as a complication of amphetamine angiitis. Neurology 25:570, 1975. 599. Stanson, A., Friese, J., Johnson, C., et al.: Polyarteritis nodosa: spectrum of angiographic findings. Radiographics 21:151, 2001. 600. Stegeman, C. A., Cohen Tervaert, J. W., Huitema, M. G., et al.: Serum levels of soluble adhesion molecules intercellular adhesion molecule 1, vascular cell adhesion molecule 1, and E-selectin in patients with Wegener’s granulomatosis. Arthritis Rheum. 37:1228, 1994. 601. Stepp, S. E., Mathew, P. A., Bennett, M., et al.: Perforin: more than just an effector molecule. Immunol. Today 21:254, 2000. 602. Stone, J. H.: Polyarteritis nodosa. JAMA 288:1632, 2002. 603. Stone, J. H.: Targeted therapies in systemic vasculitis. Cleve. Clin. J. Med. 69(Suppl. 2):SII124, 2002. 604. Stone, J. H., Tun, W., and Hellman, D. B.: Treatment of non-life threatening Wegener’s granulomatosis with methotrexate and daily prednisone as the initial therapy of choice. J. Rheumatol. 26:1134, 1999. 605. Stone, J. H., Uhlfelder, M. I., Hellmann, D. B., et al.: Etanercept combined with conventional treatment in Wegener’s granulomatosis: a six-month open-label trial to evaluate safety. Arthritis Rheum. 44:1149, 2001. 606. Strickler, R. B., Sanders, K. A., Owen, W. F., et al.: Mononeuritis multiplex associated with cryoglobulinemia in HIV infection. Neurology 42:2103, 1992. 607. Sundy, J. S., and Haynes, B. F.: Cytokines and adhesion molecules in the pathogenesis of vasculitis. Curr. Rheumatol. Rep. 2:402, 2000. 608. Suzuki, K.: Neutrophil functions of patients with vasculitis related to myeloperoxidase-specific anti-neutrophil antibody. Int. J. Hematol. 74:134, 2001. 609. Swalwell, C. I., Reddy, S. K., and Rao, V. J.: Sudden death due to unsuspected coronary vasculitis. Am. J. Forensic Med. Pathol. 12:306, 1991. 610. Szekanecz, Z., and Koch, A. E.: Biology of endothelial cells. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 19, 2002. 611. Taekema-Roelvink, M. E., van Kooten, C., Verburgh, C. A., and Daha, M. R.: Role of proteinase 3 in activation of endothelium. Springer Semin. Immunopathol. 23:299, 2001. 612. Takeuchi, A., Kodama, M., Takatsu, M., et al.: Mononeuritis multiplex in incomplete Behcet’s disease: a case

613. 614.

615.

616.

617.

618.

619.

620.

621.

622.

623.

624.

625. 626.

627.

628.

629.

630.

report and review of the literature. Clin. Rheumatol. 8:375, 1989. Takeuchi, T., and Abe, T.: Role of adhesion molecules in vasculitis syndrome. Intern. Med. 41:41, 2002. Talar-Williams, C., Hijazi, Y. M., Walther, M. M., et al.: Cyclophosphamide-induced cystitis and bladder cancer in patients with Wegener granulomatosis. Ann. Intern. Med. 124:477, 1996. Tan, E. M., Cohen, A. S., Fries, J. F., et al.: The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 25:1271, 1982. Tanaka, E., Tada, K., Amitani, R., and Kuze, F.: Systemic hypersensitivity vasculitis associated with bronchiectasis. Chest 102:647, 1992. Tanaka, Y.: Integrin activation by chemokines: relevance to inflammatory adhesion cascade during T cell migration. Histol. Histopathol. 15:1169, 2000. Tassin, S., Ferriere, G., and Laterre, E. C.: Mononevrite multiple dans un cas de macroglobulinemia de Waldenstrom. Acta Neurol. Belg. 80:287, 1980. Tatsis, E., Schnabel, A., and Gross, W. L.: Interferon-alpha treatment of four patients with the Churg-Strauss syndrome. Ann. Intern. Med. 129:370, 1998. ten Berge, I., Wilmink, J., Meyer, C., et al.: Clinical and immunologic follow up of patients with severe renal disease in Wegener’s granulomatosis. Am. J. Nephrol. 5:21, 1985. ten Holder, S. M., Joy, M. S., and Falk, R. J.: Cutaneous and systemic manifestations of drug-induced vasculitis. Ann. Pharmacother. 36:130, 2002. Tesar, V., Jelinkova, E., Masek, E., et al.: Influence of plasma exchange on serum levels of cytokines and adhesion molecules in ANCA-positive renal vasculitis. Blood Purif. 16:72, 1998. Thomas, E., Hay, E. M., Hajeer, A., and Silman, A. J.: Sjogren’s syndrome: a community-based study of prevalence and impact. Br. J. Rheumatol. 37:1069, 1998. Thomas, F. P., Lovelace, R. E., Ding, X.-S., et al.: Vasculitic neuropathy in a patient with cryoglobulinemia and antiMAG IgM monoclonal gammopathy. Muscle Nerve 15:891, 1992. Tomsic, M., Logar, D., Grmek, M., et al.: Prevalence of Sjogren’s syndrome in Slovenia. Rheumatology 38:164, 1999. Torvik, A., and Berntzen, A. E.: Necrotizing vasculitis without visceral involvement: postmortem examination of three cases with affection of skeletal muscles and peripheral nerves. Acta Med. Scand. 184:69, 1968. Trapani, J. A., and Smyth, M. J.: Functional significance of the perforin/granzyme cell death pathway. Nat. Rev. Immunol. 10:735, 2002. Traverso, F., Martini, F., Banchi, L., et al.: Vasculitic neuropathy associated with beta-hemolytic streptococcal infection: a case report. Ital. J. Neurol. Sci. 18:105, 1997. Travis, W. D., Hoffman, G. S., Leavitt, R. Y., et al.: Surgical pathology of the lung in Wegener’s granulomatosis: review of 87 open lung biopsies from 67 patients. Am. J. Surg. Pathol. 15:315, 1991. Tsokos, M., Lazaron, S. A., and Moutsopoulos, H. M.: Vasculitis in Sjögren’s syndrome: histologic classification and clinical presentation. Am. J. Clin. Pathol. 88:26, 1987.

Neuropathies with Systemic Vasculitis 631. Valbonesi, M., Montani, F., Mosconi, L., et al.: Plasmapheresis and cytotoxic drugs for mixed cryoglobulinemia. Haematologia 17:341, 1984. 632. Valentini, G., and Black, C.: Systemic sclerosis. Best Pract. Clin. Rheumatol. 5:807, 2002. 633. Valeriano-Marcet, J., Ravichandran, L., and Kerr, L. D.: HIV associated systemic necrotizing vasculitis. J. Rheumatol. 17:1091, 1990. 634. Valli, G., de Vecchi, A., Gaddi, L., et al.: Peripheral nervous system involvement in essential cryoglobulinemia and neuropathy. Clin. Exp. Rheumatol. 7:479, 1989. 635. van der Pol, W., and van de Winkel, J. G.: IgG receptor polymorphisms: risk factors for disease. Immunogenetics 48:222, 1998. 636. Vassilopoulos, D., and Calabrese, L. H.: Hepatitis C virus infection and vasculitis: implications of antiviral and immunosuppressive therapies. Arthritis Rheum. 46:585, 2002. 637. Vassilopoulos, D., and Calabrese, L. H.: Virus-associated vasculitides: clinical aspects. In Hoffman, G. S., and Weyand, C. M. (eds.): Inflammatory Diseases of Blood Vessels. New York, Marcel Dekker, p. 565, 2002. 638. Vilppula, A.: Muscular disorders in some collagen diseases. Acta Med. Scand. Suppl. 540:1, 1972. 639. Vincent, D., Dubas, F., Hauw, J. J., et al.: Nerve and muscle microvasculitis in peripheral neuropathy: a remote effect of cancer? J. Neurol. Neurosurg. Psychiatry 49:1007, 1986. 640. Vitali, C., Bencivelli, W., Isenberg, D. A., et al.: Disease activity in systemic lupus erythematosus: report of the Consensus Study Group of the European Workshop for Rheumatology Research. I. A descriptive analysis of 704 European lupus patients. Clin. Exp. Rheum. 10:527, 1992. 641. Vitali, C., Bombardieri, S., Jonsson, R., et al.: Classification criteria for Sjogren’s syndrome: a revised version of the European criteria proposed by the American-European Consensus Group. Ann. Rheum. Dis. 61:554, 2002. 642. Vitali, C., Bombardieri, S., Moutsopoulos, H. M., et al.: Preliminary criteria for the classification of Sjögren’s syndrome. Arthritis Rheum. 36:340, 1993. 643. Vollertson, R. S., Conn, D. L., Ballard, D. J., et al.: Rheumatoid vasculitis: survival and associated risk factors. Medicine 65:365, 1986. 644. Voskuyl, A. E., Hazes, J. M., Zwinderman, A. H., et al.: Diagnostic strategy for the assessment of rheumatoid vasculitis. Ann. Rheum. Dis. 62:407, 2003. 645. Voskuyl, A. E., Zwinderman, A. H., Westedt, M. L., et al.: Factors associated with the development of vasculitis in rheumatoid arthritis: results of a case-control study. Ann. Rheum. Dis. 55:190, 1996. 646. Vrethem, M., Lindvall, B., Holmgren, H., et al.: Neuropathy and myopathy in primary Sjogren’s syndrome: neurophysiological, immunological and muscle biopsy results. Acta Neurol. Scand. 82:126, 1990. 647. Wai, C. T., and Lok, A. S.: Treatment of hepatitis B. J. Gastroenterol. 37:771, 2002. 648. Wajant, H.: The Fas signaling pathway: more than a paradigm. Science 296:1635, 2002. 649. Wallach, S., Cohen, S., Reid, D. M., et al.: Effects of risedronate treatment on bone density and vertebral frac-

650. 651.

652.

653.

654.

655.

656.

657.

658.

659.

660.

661.

662.

663.

664. 665.

666.

2403

ture in patients on corticosteroid therapy. Calcif. Tissue Int. 67:277, 2000. Walton, E. W.: Giant cell granuloma of the respiratory tract (Wegener’s granulomatosis). Br. Med. J. 2:265, 1958. Wattiaux, M. J., Kahn, M. F., Thevenet, J. P., et al.: L’atteinte vasculaire dans la polyarthrite rhumatoide: etude retrospective de 37 polyarthrites rhumatoides avec atteinte vasculaire et revue de la litterature. Ann. Med. Interne (Paris). 138:566, 1987. Watts, R. A., Lane, S. E., Bentham, G., and Scott, D. G.: Epidemiology of systemic vasculitis: a ten-year study in the United Kingdom. Arthritis Rheum. 43:414, 2000. Watts, R. A., and Scott, D. G.: Classification and epidemiology of the vasculitides. Baillieres Clin. Rheumatol. 11:191, 1997. Wees, S. J., Sunwoo, I. N., and Oh, S. J.: Sural nerve biopsy in systemic necrotizing vasculitis. Am. J. Med. 71:525, 1981. Wegener, F.: Uber generalisierte, septische Gefasserkrankungen. Verh. Dtsch. Pathol. Ges. 29:202, 1936. Weller, R. O., Bruckner, F. E., and Chamberlain, M. A.: Rheumatoid neuropathy: a histological and electrophysiological study. J. Neurol. Neurosurg. Psychiatry 33:592, 1970. Westman, K. W., Bygren, P. G., Olsson, H., et al.: Relapse rate, renal survival, and cancer morbidity in patients with Wegener’s granulomatosis or microscopic polyangiitis with renal involvement. J. Am. Soc. Nephrol. 9:842, 1998. Weyand, C. M., and Goronzy, J. J.: Mechanisms of disease: medium- and large-vessel vasculitis. N. Engl. J. Med. 349:160, 2003. Weyand, C. M., Hicok, K. C., Hunder, G. G., and Goronzy, J. J.: The HLA-DRB1 locus as a genetic component in giant cell arteritis: mapping of a disease-specific motif to the antigen binding site of the HLA-DR molecule. J. Clin. Invest. 90:2355, 1992. Wiik, A. S.: Clinical use of serological tests for antineutrophil cytoplasmic antibodies: what do the studies say? Rheum. Dis. Clin. North Am. 27:799, 2001. Winkelstein, A., Starz, T. W., and Agarwal, A.: Efficacy of combined therapy with plasmapheresis and immunosuppressants in rheumatoid vasculitis. J. Rheumatol. 11:162, 1984. Wohlwill, F.: Uber die nur Mikroskopisch erkenbarre Form der Perarteritis nodosa. Arch. Pathol. Anat. 246:377, 1923. Wong, D. K., Cheung, A. M., O’Rourke, K., et al.: Effect of alpha-interferon treatment in patients with hepatitis B e antigen-positive chronic hepatitis B: a meta-analysis. Ann. Intern. Med. 119:312, 1993. Wooten, M. D., and Jasin, H. E.: Vasculitis and lymphoproliferative diseases. Semin. Arthritis Rheum. 26:564, 1996. Xiao, H., Heeringa, P., Hu, P., et al.: Antineutrophil cytoplasmic autoantibodies specific for myeloperoxidase cause glomerulonephritis and vasculitis in mice. J. Clin. Invest. 110:955, 2002. Yamamoto, M., Ito, Y., Mitsuma, N., et al.: Pathologyrelated differential expression regulation of NGF, GDNF,

2404

667.

668.

669. 670.

671.

672.

673.

Diseases of the Peripheral Nervous System

CNTF, and IL-6 mRNAs in human vasculitic neuropathy. Muscle Nerve 24:830, 2001. Yasutomo, K.: Pathological lymphocyte activation by defective clearance of self-ligands in systemic lupus erythematosus. Rheumatology 42:214, 2003. Yazici, H., Fresco, I., Tuns, R., and Melikoglu, M.: Behçet’s syndrome: pathogenesis, clinical manifestations and treatment. In Ball, G. V., and Bridges, S. L. Jr. (eds.): Vasculitis. New York, Oxford University Press, p. 406, 2002. Yi, E. S., and Colby, T. V.: Wegener’s granulomatosis. Semin. Diagn. Pathol. 18:34, 2001. Younger, D. S., Dalmau, J., Inghirami, G., et al.: Anti-Huassociated peripheral nerve and muscle microvasculitis. Neurology 44:181, 1994. Younger, D. S., Rosoklija, G., Hays, A. P., et al.: Diabetic peripheral neuropathy: a clinicopathologic and immunohistochemical analysis of sural nerve biopsies. Muscle Nerve 19:722, 1996. Younger, D. S., Rosoklija, G., Neinstedt, L. J., et al.: HIV-1 associated sensory neuropathy: a patient with peripheral nerve vasculitis. Muscle Nerve 19:1364, 1996. Yusuf-Makagiansar, H., Anderson, M. E., Yakovleva, T. V., et al.: Inhibition of LFA-1/ICAM-1 and VLA-4/VCAM-1 as a therapeutic approach to inflammation and autoimmune diseases. Med. Res. Rev. 22:146, 2002.

674. Zaja, F., De Vita, S., Mazzaro, C., et al.: Efficacy and safety of rituximab in type II mixed cryoglobulinemia. Blood 101:3827, 2003. 675. Zaltron, S., Puoti, M., Liberini, P., et al.: High prevalence of peripheral neuropathy in hepatitis C virus infected patients with symptomatic and asymptomatic cryoglobulinemia. Ital. J. Gastroenterol. Hepatol. 30:391, 1998. 676. Zeek, P. M, Smith, C. C., and Weeter, J. C.: Studies on periarteritis nodosa. III. The differentiation between the vascular lesions of periarteritis nodosa and of hypersensitivity. Am. J. Pathol. 24:889, 1948. 677. Zehnder, P., Jenni, W., Brandner, S., et al.: Vaskulitis und Mononeuritis multiplex. Schweiz. Rundsch. Med. Prax. 89:776, 2000. 678. Zhang, L., Jayne, D. R., Zhao, M. H., et al.: Distribution of MHC class II alleles in primary systemic vasculitis. Kidney Int. 47:294, 1995. 679. Zhou, Y., Giscombe, R., Huang, D., and Lefvert, A. K.: Novel genetic association of Wegener’s granulomatosis with the interleukin 10 gene. J. Rheumatol. 29:317, 2002. 680. Zhou, Y., Huang, D., and Hoffman, G. S.: Genetic polymorphisms in TNF, IL-1, IL-6 and cytotoxic T lymphocyteassociated antigen 4 (CTLA-4) in Wegener’s granulomatosis (WG) [abstract]. Cleve. Clin. J. Med. 69(Suppl. 2):SII64, 2002.

106 Microvasculitis P. JAMES B. DYCK, JANEAN ENGELSTAD, AND PETER J. DYCK

Overview Disease Association

Pathologic Alterations of Microvessels

OVERVIEW Necrotizing vasculitides are classified by the kind and size of blood vessels involved, organ involvement, disease associations, and underlying mechanisms. The Chapel Hill consensus Conference on the Nomenclature of Systemic Vasculitis developed a classification that may be interpreted as placing most vasculitides of nerve into the microscopic polyarteritis category. They recognize two categories of microvasculitis: Henoch-Schönlein purpura and essential cryoglobinemic vasculitis, and cutaneous leukoclastic angiitis12,16 (mainly of the skin). Based mainly on our peripheral nerve biopsy experience, we favor a separation of necrotizing vasculitis of nerve into two different groups: nerve large arteriole vasculitis (Figs. 106–1 and 106–2) and nerve microvasculitis (Figs. 106–3 to 106–6). The separation is based on differences in the size and kind of vessels involved, disease associations, and, perhaps, course and outlook and treatment considerations. A series of scientific studies support the existence of at least two varieties of nerve necrotizing vasculitides5,8,13,17,20 (also see Chapter 32). The first group (nerve large arteriole vasculitis) has involvement of small arteries and large arterioles (and a varying degree of smaller vessels may also be involved) and is usually associated with rheumatoid arthritis, polyarteritis nodosa, Churg-Strauss syndrome, or Wegener’s granulomatosis (see Chapter 105). The second group, nerve microvasculitis (discussed here), is less well defined but involves a different spectrum of vessels—the smallest arterioles, microvessels, and venules. This nerve microvasculitis occurs in diabetic and nondiabetic radiculoplexus neuropathies, immune sensorimotor polyneuropathies sometimes

Pathologic Alteration of Nerve

associated with sicca, classical Sjögren’s syndrome, paraneoplastic neuropathies, nonsystemic vasculitic neuropathies, and virus-associated neuropathies (some cases of human immunodeficiency virus, cytomegalovirus, hepatitis A and C, and perhaps others). Whereas fibrinoid necrosis of the tunica media is often prominent and characteristic in nerve large arteriole vasculitis (see Figs. 106–1 and 106–2), obvious fibrinoid necrosis is usually not found in nerve microvasculitis, but the latter group (see Fig. 106–3) there is inflammation of the vessel wall with separation, fragmentation, and necrosis of the thin tunica media. In both groups of necrotizing vasculitis, evidence of ischemic injury or repair is found (Figs. 106–7 to 106–9). For both varieties of vasculitis there is evidence of ischemic injury of nerve, but for microvasculitis, closure of the lumen has not yet been adequately demonstrated. There is anecdotal information on a favorable response to immune-modulating therapy for both varieties of necrotizing vasculitis.9,10 In this chapter, we focus on showing the histologic differences between the two nerve vasculitides.

DISEASE ASSOCIATION Microvasculitis of nerve has been extensively studied in nondiabetic and diabetic lumbosacral radiculoplexus neuropathy, also called femoral neuropathy, proximal diabetic neuropathy, Bruns-Garland syndrome, and by other names7,8 (also see Chapter 86). Although it appears that this disorder is more prevalent in diabetes mellitus, glycemic exposure does not appear to be the direct metabolic cause. The frequently 2405

2406

Diseases of the Peripheral Nervous System

FIGURE 106–1 Serial transverse paraffin sections of sural nerve illustrating a large epineurial arteriole undergoing necrotizing vasculitis with fibrinoid degeneration. Upper left, The vessel structure is markedly altered with cellular, almost complete occlusion of the lumen; necrosis of the muscle cells of the tunica media; fibrinoid degeneration of the inner portions of the media; and necrosis and inflammation of the outer tunica media and perivascular inflammation. (Hematoxylin and eosin stain.) Upper right, The red fibrinoid degeneration is especially prominent (trichrome stain). Lower, Immunohistochemical reactions—the left is reacted to CD45 (lymphocytes), the middle to CD68 (macrophages), and the right to smooth muscle actin. These sections illustrate the inflammatory cell involvement in necrotizing vasculitis and the separation, fragmentation, and disappearance of muscle cells. See Color Plate

associated weight loss may perhaps provide an indication of systemic involvement. We have questioned whether circulating cytokines might be the possible explanation for the weight loss. The pathologic basis of both noninherited and inherited immune brachial plexus neuropathy (a cervical radiculoplexus neuropathy) has not been as extensively studied as lumbosacral radiculoplexus neuropathy, but nerve microvasculitis has been demonstrated in some cases.14,20 It is of considerable interest that endocrine and immune factors play a role in the mutant gene abnormality hereditary brachial plexus neuropathy. Within hours of parturition, when the immune system changes to a less tolerant state, the patient with the mutant gene may develop an acute brachial plexus neuropathy. Biopsy of a superficial radial nerve during an attack has shown changes of microvasculitis.14 Large doses of intravenous methylprednisolone appeared to abort some features of involvement. Nerve microvasculitis has also been described in nonsystemic vasculitic neuropathy,5,13

immune sensorimotor polyneuropathy with sicca, 17 paraprotein neuropathy,21 paraneoplastic neuropathy,2 and various viral infections, notably hepatitis C with cryoglobulinemia.3,4,11

PATHOLOGIC ALTERATIONS OF MICROVESSELS The vessels involved in microvasculitis are usually arterioles without an internal elastic lamina (i.e., ⬍30 to 40 ␮m), microvessels, and venules. Typically the involvement is of epineurial vessels, but endoneurial vessels may also be involved. The lower frequency of endoneurial small arteriole involvement may be explained by their lower frequency of occurrence in the endoneurium. Mononuclear cells separate the muscle leaflets of the vessel wall. The vascular and perivascular inflammatory

2407

Microvasculitis

infiltrates are both CD45 (common leukocyte antigen) and CD68 (macrophages) reactive. Inflammatory cells separate muscle layers, but with increased severity there is separation of the muscle leaflets, which become fragmented and separated from the microvessel. Obvious occlusion of vessels has not been encountered, but recent or previous bleeding (hemosiderin in macrophages, in the Turnbull blue stain) is typical. The hemosiderin is typically found adjacent to affected microvessels. Fibrinoid degeneration of the media (seen to advantage in trichrome stain in paraffin sections), seen in nerve large arteriole vasculitis, is rarely observed in microvasculitis. Presumably the small amount of the tunica media may be the reason for its not being an obvious feature. Typical of microvessel inflammation is angioneogenesis— closely spaced, thin-walled microvessels in regions of previously ischemic areas.

PATHOLOGIC ALTERATION OF NERVE

FIGURE 106–2 Nerves from patients with large nerve vessel necrotizing vasculitis. Top, A recently occluded arteriole (hematoxylin and eosin stain). The former site of the lumen is replaced by granulation tissue and early recanalization. The dot shows relatively intact tunica media, whereas the asterisk shows an area where the media has disappeared. This focal necrosis of media is characteristic of vessel injury in necrotizing vasculitis. Middle, An arteriole with an inflammatory mononuclear cell infiltrate and fibrinoid necrosis of a sector of the wall (red) (trichrome stain). The adjacent vessel wall is intact. Bottom, An arteriole that has been previously damaged and occluded, although the internal elastic lamina remains intact. Within the original lumen there are multiple small channels of recanalization as well as neovascularization surrounding the outside of the arteriole. (Top, modified from Dyck, P. J., Conn, D. L., and Okazaki, H.: Necrotizing angiopathic neuropathy: three-dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972.) See Color Plate

The pathologic alterations of nerve may be used to infer underlying mechanisms—different patterns of involvement occur with acute compression, ischemia, inflammatory demyelination, and various parenchymatous alterations (see Chapter 32). We interpret the changes associated with microvasculitis as typical of various stages of ischemic injury and repair.1,6,18,19 The hallmarks of ischemic injury were delineated from three-dimensional studies of limb nerves in patients with nerve large arteriole vasculitis associated with rheumatoid arthritis.6 Ligation of large vessels15 and microsphere embolization of nerve with 15-␮m microsphere emboli18 provided the temporal and three-dimensional anatomicopathologic correlations. These are described in the chapter on Pathology of the Peripheral Nerve (see Chapter 32). The most characteristic features of ischemic lesions are the association with vessel disease; the kind and pattern of fiber degeneration, loss, and regeneration; and injury extending to involve the perineurium. The typical changes of ischemic injury found in human nerve large arteriole vasculitis also have been demonstrated for microvasculitis. As shown in the accompanying figures, quite dramatic multifocal fiber degeneration, fiber loss, and regeneration typical of ischemic injury is associated with nerve microvasculitis. A typical feature of an ischemic lesion is that it is not confined to the endoneurial compartment but extends to involve the perineurium. Associated with microvasculitis, we have found all stages of perineurial injury—from acute fibrinoid degeneration to thickening and scarring and regrowth of microfasciculi through the perineurium into the epineurium (injury neuroma). Although segmental demyelination may be found in acute ischemic injury, it is usually at the borders of such injury and may relate to axonal atrophy (distal to sites of axonal stasis) or to sites of axonal enlargement.

2408

Diseases of the Peripheral Nervous System

FIGURE 106–3 Serial skip paraffin sections of a microvessel above (upper panels), at (middle panels), and below (lower panels) regions of microvasculitis in the sural nerve of a patient with lumbosacral radiculoplexus neuropathy. The sections in the left column are stained with hematoxylin and eosin, the sections in the middle column are reacted with anti–human smooth muscle actin, and the sections in the right column are reacted with CD45 (lymphocytes). The smooth muscle of the tunica media in the regions of the microvasculitis (middle panels) are separated by mononuclear cells, fragmented, and decreased in amount. The changes are those of a focal microvasculitis. These changes are seen in both diabetic and nondiabetic conditions. (From Dyck, P. J. B., and Windebank, A. J.: Diabetic and non-diabetic lumbosacral radiculoplexus neuropathies: new insights into pathophysiology and treatment. Muscle Nerve 25:477, 2002.) See Color Plate

Microvasculitis

FIGURE 106–4 Serial paraffin sections from a sural nerve of a patient with microvasculitis of nerve. Upper, Fibrinoid necrosis of the media (inner red segment of the vessel) surrounded by a mononuclear inflammatory cell infiltrate, which disrupts and destroys the vessel wall (left, hematoxylin and eosin stain; right, Masson trichrome stain). Lower, Immunohistochemical reactions to lymphocytes (left, reacted with CD45), macrophages (middle, reacted with CD68), and smooth muscle (right, reacted with smooth muscle actin), demonstrating the destruction of the vessel wall. This example is somewhat atypical of microvasculitis in that it contains a microvessel with fibrinoid necrosis of the vessel wall (see text). See Color Plate

2409

2410

Diseases of the Peripheral Nervous System

FIGURE 106–5 Serial paraffin sections from a sural nerve demonstrating a granulomatous microvasculitis of nerve. Upper, A microvessel with fibrinoid degeneration of the media (arrow) surrounded by an adventitial inflammatory response (left, hematoxylin and eosin stain; right, Masson trichrome stain). Multinucleated giant cells (arrowhead) are present. Lower, Immunohistochemical reactions to lymphocytes (left, reacted with CD45) and macrophages (right, reacted with CD68). Note the paucity of reactivity for lymphocytes as compared to the robust reactivity to macrophages. These findings are typical of a granulomatous vasculitis. See Color Plate

Microvasculitis

FIGURE 106–6 Serial paraffin sections from a sural nerve demonstrating a B-cell lymphoma causing a microvasculitis. All sections are from the same microvessel—the upper row is at one level and the lower row is at a slightly distal level. Upper left, Disruption of the vessel wall by mononuclear inflammatory cells (hematoxylin and eosin stain). Upper middle, The intense mononuclear cell infiltration (reacted with CD45, lymphocytes). Upper right, The marked disruption, destruction, and separation of the smooth muscle layers by the inflammatory cells (reacted with smooth muscle actin). Lower left, Almost all of the mononuclear cells infiltrating the vessel wall are B cells (reacted with CD20). Lower right, Few are T cells (reacted with CD3). In a typical immune microvasculitis, one would expect the T-cell response to be greater than that of B cells. See Color Plate

2411

2412

Diseases of the Peripheral Nervous System

FIGURE 106–7 Different stages of bleeding caused by vasculitis in nerves of patients with necrotizing vasculitis. Top panel, Early blood breakdown pigments (gold color) interspersed on the right-hand side with red blood cells in a paraffin section from a sural nerve (hematoxylin and eosin stain). Finding the pigment change guarantees that the bleeding is not simply related to the act of the biopsy. Middle and bottom panels, Serial longitudinal paraffin sections (middle, hematoxylin and eosin stain; bottom, Turnbull blue stain) of sural nerve illustrating previous bleeding into tissue with formation of hemosiderin-laden macrophages (brown on hematoxylin and eosin and blue on Turnbull blue) interspersed with mononuclear inflammatory cells. See Color Plate

FIGURE 106–8 Transverse paraffin sections from nerves of patients with vasculitis showing new vessel formation (neovascularization). Top, A fascicle (top center) and multiple newly formed small epineurial blood vessels beneath it (Luxol fast blue/periodic acid–Schiff stain). Bottom, Multiple newly formed small blood vessels interspersed with mononuclear inflammatory cells (reacted with smooth muscle actin). Neovascularization is a typical change found in nerves with ischemic injury in both large nerve vessel (arteriole) and small nerve vessel (microvessel) vasculitis. See Color Plate

Microvasculitis

2413

FIGURE 106–9 Transverse semithin epoxy sections stained with methylene blue taken from sural nerves of patients with vasculitis demonstrating changes typically seen in nerves having undergone chronic ischemic damage. The panels in the upper row show multifocal fiber loss. Upper left, The fascicle in the left middle shows relative preservation of myelinated fibers, whereas the fascicle in the lower right is devoid of myelinated fibers (except for a few degenerating profiles). The arrowhead demonstrates a region of microvasculitis. Upper right, Demonstrates both intrafascicular and interfascicular multifocal fiber loss. Note that many of the remaining myelinated fibers are undergoing active axonal degeneration. The lower row demonstrates abortive regeneration in nerve after ischemic injury. Lower left, A “parent fascicle” (center) that is devoid of large myelinated fibers but still contains some small myelinated fibers (probably regenerating fibers). This fascicle is surrounded by very thickened perineurium (asterisk) and by injury neuroma forming microfascicles (abortive regeneration, arrows). Lower right, Section from the same nerve showing the injury neuroma and microfasciculation (arrows) at a higher power. These ischemic changes are found frequently in nerves of patients with either large nerve vessel (arteriole) or small nerve vessel (microvessel) vasculitis. See Color Plate

REFERENCES 1. Benstead, T. J., Dyck, P. J., and Sangalang, V.: Inner perineurial cell vulnerability in ischemia. Brain Res. 489:177, 1989. 2. Blumenthal, D., Schochet, S. Jr., Gutmann, L., et al.: Smallcell carcinoma of the lung presenting with paraneoplastic peripheral nerve microvasculitis and optic neuropathy. Muscle Nerve 21:1358, 1998. 3. Cacoub, P., Maisonobe, T., Thibault, V., et al.: Systemic vasculitis in patients with hepatitis C. J. Rheumatol. 28:109, 2001. 4. Daoud, M. S., el-Azhary, R. A., Gibson, L. E., et al.: Chronic hepatitis C, cryoglobulinemia, and cutaneous necrotizing vasculitis: clinical, pathologic, and immunopathologic study of twelve patients. J. Am. Acad. Dermatol. 34:219, 1996.

5. Dyck, P. J., Benstead, T. J., Conn, D. L., et al.: Nonsystemic vasculitic neuropathy. Brain 110:843, 1987. 6. Dyck, P. J., Conn, D. L., and Okazaki, H.: Necrotizing angiopathic neuropathy: three-dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972. 7. Dyck, P. J. B., Engelstad, J., Norell, J., and Dyck, P. J.: Microvasculitis in non-diabetic lumbosacral radiculoplexus neuropathy (LSRPN): similarity to the diabetic variety (DLSRPN). J. Neuropathol. Exp. Neurol. 59:525, 2000. 8. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Microvasculitis and ischemia in diabetic lumbosacral radiculoplexus neuropathy. Neurology 53:2113, 1999. 9. Dyck, P. J. B., Norell, J. E., and Dyck, P. J.: Methylprednisolone may improve lumbosacral radiculoplexus neuropathy. Can. J. Neurol. Sci. 28:224, 2001.

2414

Diseases of the Peripheral Nervous System

10. Guillevin, L., Cohen, P., Gayraud, M., et al.: Churg-Strauss syndrome: clinical study and long-term follow up of 96 patients. Medicine (Baltimore) 78:26, 1999. 11. Heckmann, J. G., Kayser, C., Heuss, D., et al.: Neurological manifestations of chronic hepatitis C. J. Neurol. 246:486, 1999. 12. Jennette, C., Falk, R. J., Andrassy, K., et al.: Nomenclature of systemic vasculitides: proposal of an international consensus conference. Arthritis Rheum. 37:187, 1994. 13. Kissel, J. T., Slivka, A. P., Warmolts, J. R., and Mendell, J. R.: The clinical spectrum of necrotizing angiopathy of the peripheral nervous system. Ann. Neurol. 18:251, 1985. 14. Klein, C. J., Dyck, P. J., Friedenberg, S. M., et al.: Inflammation and neuropathic attacks in hereditary brachial plexus neuropathy. J. Neurol. Neurosurg. Psychiatry 73:45, 2002. 15. Korthals, J. K., Gieron, M. A., and Wisniewski, H. M.: Nerve regeneration patterns after acute ischemic injury. Neurology 39:932, 1989.

16. Kumar, V., Cotran, R. S., and Robbins, S. L.: Basic Pathology, 6th ed. Philadelphia, W. B. Saunders, p.775, 1997. 17. Mellgren, S. I., Conn, D. L., Stevens, J. C., and Dyck, P. J.: Peripheral neuropathy in primary Sjogren’s syndrome. Neurology 39:390, 1989. 18. Nukada, H., and Dyck, P. J.: Microsphere embolization of nerve capillaries and fiber degeneration. Am. J. Pathol. 115:275, 1984. 19. Nukada, H., Dyck, P. J., and Karnes, J. L.: Spatial distribution of capillaries in rat nerves: correlation of ischemic damage. Exp. Neurol. 87:369, 1985. 20. Suarez, G. A., Giannini, C., Bosch, E. P., et al.: Immune brachial plexus neuropathy: suggestive evidence for an inflammatory immune pathogenesis. Neurology 46:559, 1996. 21. Turner, M. R., Warren, J. D., Jacobs, J. M., et al.: Microvasculitic paraproteinaemic polyneuropathy and B-cell lymphoma. J. Peripher. Nerv. Syst. 8:100, 2003.

107 Sarcoid Neuropathy GÉRARD SAID AND CATHERINE LACROIX

History Epidemiology Sarcoidosis Neurosarcoidosis Genetics Neurobiologic Basis Granuloma Formation Angiitis and Sarcoidosis Consequences of Granuloma Formation in the Peripheral Nervous System

Symptoms and Signs of Sarcoid Neuropathy Cranial Nerve Involvement Peripheral Neuropathy Sarcoidosis: A Multisystem Disease Central Nervous System and Meningeal Involvement Other Organ System Involvement Course Electrophysiology

Sarcoidosis is an inflammatory multisystem disease of unknown origin. The lungs are by far the most frequently affected tissue, but other organs can be involved, including skin, eyes, liver, and lymph nodes. Neurologic manifestations of sarcoidosis are relatively rare but constitute a treatable cause of central and peripheral neurologic manifestations. Cranial nerve involvement, particularly facial palsies and hypothalamic and pituitary lesions, is common. These lesions tend to occur early and respond favorably to treatment. Space-occupying masses, peripheral neuropathy, and symptomatic muscular involvement occur later and portend a chronic course. Regarding the peripheral nervous system, cranial nerves are predominantly affected, and peripheral facial nerve palsy, often bilateral, is the most common neurologic manifestation of sarcoidosis. Multifocal peripheral neuropathy is a rare event in sarcoidosis. In some cases however, peripheral neuropathy is the presenting manifestation and the peripheral nerves are seemingly the only system affected. Definite diagnosis of sarcoidosis rests ideally on histologic demonstration of sarcoid granulomas in tissue biopsy specimens. In patients who present with sarcoid neuropathy, nerve biopsy and biopsy of other organs affected, especially intrathoracic lymph nodes, can be diagnostic.

Pathology Nerve Biopsy Findings Muscle Biopsy Findings Diagnosis Treatment Conclusion

HISTORY The identification of sarcoidosis starts with Hutchinson’s description in 1877 of a patient with cutaneous patches of the hands and feet that developed over 2 years. Hutchinson suggested that this phenomenon represented a form of skin disease that had escaped recognition.26 It then took several decades to complete the concept of sarcoidosis.24 The earliest pathologic description of sarcoidosis of the peripheral nerves was that of Winkler in 1905.57 In 1909, the Danish ophthalmologist Heerfordt described uveoparotid fever in three patients (characterized by a chronic febrile course, enlarged parotid glands, and uveitis; two had unilateral facial nerve palsy).22 At the time, the syndrome was ascribed to mumps. The first mention of peripheral neuropathy, without histologic confirmation, dates back to Feiling and Viner in 1922.17

EPIDEMIOLOGY Sarcoidosis Sarcoidosis occurs throughout the world, affecting both sexes and all races and ages. The prevalence of sarcoidosis ranges from less than 1 to 64 per 100,000 worldwide. 2415

2416

Diseases of the Peripheral Nervous System

The disease shows a consistent predilection for adults less than 40 years of age, peaking in those 20 to 29 years old.20 In Scandinavian countries and Japan, there is a second peak incidence in women more than 50 years of age.4 Most studies suggest a slightly higher disease rate for women. In the United States, sarcoidosis is much more prevalent in blacks than in whites,40 while in Europe the disease affects mostly whites and is 6 to 20 times more common in Sweden than in other European countries. An accurate estimate of the prevalence or incidence of sarcoidosis is probably not available. The reasons for this are lack of consistent disease criteria, variability of disease presentation, and lack of sensitive and specific diagnostic tests, resulting in underrecognition and misdiagnosis of the disease.23 Significant heterogeneity in disease presentation and severity occurs among different ethnic and racial groups. Several studies suggest that sarcoidosis in blacks is more severe, while whites are more likely to present with asymptomatic disease.25 The overall mortality from sarcoidosis is 1% to 5%, especially from pulmonary and neurologic manifestations. Intriguing spatial clusters of illness have suggested person-to-person transmission or shared exposure to an environmental agent, but so far neither animal experiments nor human studies have yet proven these hypotheses.

Neurosarcoidosis Both the central and the peripheral nervous system can be affected in perhaps 5% of patients with sarcoidosis.8,9,13,31,36,48,51 Stern et al.51 published a large retrospective study on the occurrence of neurologic manifestations in a cohort of 649 patients with sarcoidosis seen at the Johns Hopkins Hospital from 1975 to 1980. Neurologic problems that could be attributed to neurosarcoidosis occurred in 5.1% of the patients. The presenting manifestation of sarcoidosis was neurologic in half of them. The average age at diagnosis of neurosarcoidosis was 33 years; 64% of the patients were female and 85% were black. Twenty-four patients (73%) developed neurosarcoidosis within the first 2 years of their illness. In another retrospective epidemiologic study of 285 patients with a definite diagnosis of sarcoidosis, only one case of peripheral neuropathy was identified.9 The proportion of patients in whom the peripheral nervous system is affected, not necessarily in isolation, is high, representative figures being 12 of 1813 and 25 of 3315 cases. Isolated peripheral seventh cranial nerve palsy was observed in nearly half the patients in one study.5 The 11 patients with noncranial peripheral neuropathy in our series account for 0.3% of the patients who underwent a nerve biopsy for symptomatic neuropathy in our center during the same period of time.43

GENETICS There are numerous reports of familial clustering of sarcoidosis. In the United States, familial clusters occur more commonly among blacks (with a rate of at least 19% in affected black families) than whites (with a rate of 5%).21 Human leukocyte antigen analyses of affected families suggest that the mode of inheritance of risk for sarcoidosis is likely polygenic. It is assumed that genetically predisposed hosts are exposed to antigens that trigger an exaggerated cellular immune response leading to granuloma formation.

NEUROBIOLOGIC BASIS Granuloma Formation The histologic hallmark of sarcoidosis is the presence of multiple noncaseating granulomas composed of clustered macrophages and lymphocytes, which provides evidence that they result from an immunologic problem. Because of this, it is hypothesized that unknown antigens, either nonself or self, initiate sarcoidosis.29,54 The first manifestation is an accumulation of mononuclear inflammatory cells, mostly helper T lymphocytes (Th cells) and mononuclear phagocytes, in affected organs. Initially acute inflammation fails to degrade the antigens and then nonspecific chronic inflammation ensues. Lymphocytes and mononuclear phagocytes are attracted to the inflamed area; in addition, mononuclear phagocytes migrate from the blood to the tissue, where they mature into macrophages and phagocytize the antigens in an attempt to neutralize them. These cells are important in the formation of noncaseating granulomas. T-cell lymphocytes are markedly increased in areas of active granulomatous disease, and the ratio of Th cells to T-suppressor cells is markedly increased. This is the opposite of the circulating blood, where the number of T cells is reduced and the ratio of Th cells to T-suppressor cells is decreased. This exaggerated T-cell activity indicates an altered immune response. The etiology of the altered T-cell response is not known, but inherited or acquired alteration of immune response genes may be an important factor, or the antigen may selectively affect the T-cell arm of the immune response. Granulomas consist of a central collection of epithelioid cells, which are derived from macrophages. As macrophages mature to epithelioid cells, they gain secretory and bactericidal capabilities but lose some phagocytic capability. Morphologically, epithelioid cells are large and polygonal and have an elliptical nucleus that contains fine chromatin and one to two nucleoli. The cytoplasm can be light or dark. There are often giant cells in the central part of the granuloma. Giant cells are a conglomeration of epithelioid cells that share the same cytoplasm and have multiple

2417

Sarcoid Neuropathy

nuclei. The central portion of the granuloma also contains predominantly CD4⫹ lymphocytes, whereas CD8⫹ lymphocytes are present in the peripheral zone.39 Sarcoid granulomas may develop fibrotic changes that usually begin at the periphery and travel centrally, ending with complete fibrosis and/or hyalinization. Granulomas may occasionally exhibit focal coagulative necrosis.39,49 On electron microscopy, well-developed epithelioid cells show numerous cytoplasmic projections with frequent interdigitations. The morphology suggests a secretory function. At first most have randomly distributed nuclei (foreign body–type giant cell), but later the nuclei are arranged in the periphery (Langhans type giant cell). Several types of cytoplasmic inclusion bodies are common in sarcoidosis but are nonspecific. Chemoattractant cytokines attract CD4⫹ Th cells at sites of granuloma formation, where they proliferate following releasing of interleukin-2. Several studies indicate that interleukin-2 acts as a local growth factor for T lymphocytes infiltrating the lung parenchyma and other involved sarcoid tissues. The process of recruiting Th cells is principally triggered by antigen-presenting cells, but the question of autoreactive lymphocytes that would attract Th cells without the need of antigen-presenting cells has been raised.29 T cells isolated from patients with active sarcoidosis also show elevated messenger RNA and protein levels of interferon-␥ with proliferation of activated T cells and are involved in the differentiation of Th0 cells into Th1 cells.33 Thus a Th1-type T-cell response (secreting interleukin-2, interleukin-12, interferon-␥, and tumor necrosis factor-␣) is likely to favor the granulomatous response at sites of disease activity.

Angiitis and Sarcoidosis Association of sarcoidosis with angiitis has been recognized at autopsy in the central nervous system (CNS) and the meninges.3,32 In Meyer et al.’s study,32 as in the two autopsy cases of patients with CNS, meningeal, and ophthalmologic signs reported by Caplan and co-workers,7 the granulomas were in the perivascular spaces and adventitia throughout the subarachnoid space and CNS. The granulomas often extended into the media, and small arteries often showed acute necrosis of all layers and recent thrombosis. Vasculitis was also found in patients with sarcoid neuropathy,35,56 and in 6 of 11 patients of our series, nerve biopsy specimens showed epineurial and perineurial granulomatous angiitis.43 Considering the small size of the specimens biopsied, necrotizing vasculitis appears to play an important role in sarcoid neuropathy. Epineurial necrotizing vasculitis induces nerve ischemia and subsequent axonal degeneration that contribute to the onset of symptomatic neuropathy. The six patients with demonstrated necrotizing vasculitis had no specific clinical presentation,

but axonal lesions were more pronounced than in the others. The occurrence of necrotizing vasculitis in association with granuloma stresses the need to exclude other causes of granulomatous angiitis, especially Wegener’s granulomatosis. However, clinical features and biologic markers sharply differentiate the two diseases.14,55

Consequences of Granuloma Formation in the Peripheral Nervous System Nerve granulomas and multinucleated giant cells found in sarcoidosis are similar to those encountered in tuberculoid leprosy and in the reversal upgrade reaction during treatment of multibacillary leprosy, which are both characteristic of a cell-mediated immune response.41 Demonstration of granulomas in the muscle biopsy specimens is helpful to exclude tuberculoid leprosy in patients at risk for leprosy, because leprous granulomas do not affect muscles, except, occasionally, around intramuscular nerves. Granulomas express a cell-mediated immune response to Mycobacterium leprae antigens in leprosy, and to unknown antigens in sarcoidosis. In these settings, memory T cells recruit and activate macrophages, which become epithelioid cells and may fuse to multinucleated giant cells. In the meantime, they gain secretory capabilities2 and release noxious secretory products, including proteolytic enzymes, that can damage neighboring fiber.42 When the granulomatous infiltrate is located in the epineurium, there is little or no nerve fiber damage, except if granuloma-related necrotizing vasculitis induces nerve ischemia. When the inflammatory infiltrate enters the endoneurium with its surrounding blood vessels, following tissue septa, the infiltrate can seriously damage the nerve fibers. Necrotizing arteritis can in turn induce ischemic nerve lesions with axonal degeneration, which come in addition to those secondary to endoneurial inflammation.

SYMPTOMS AND SIGNS OF SARCOID NEUROPATHY Involvement of the peripheral nervous system markedly predominates in cranial nerves. However, any part of the peripheral nervous system can be involved, and patients may present with a variety of peripheral neuropathies.

Cranial Nerve Involvement Multiple fluctuating and remitting cranial nerve palsies, so-called cranial polyneuritis, are relatively characteristic of sarcoidosis.31 In the fully developed form, the cranial nerves are affected successively and apparently at random over a period of several weeks or months. Acute presentation of sarcoidosis with the combination of uveitis, parotid gland enlargement, facial palsy, and fever

2418

Diseases of the Peripheral Nervous System

is called uveoparotid fever or Heerfordt’s syndrome. This syndrome was first described in 1909 by the Danish ophthalmologist Heerfordt as uveoparotid fever in three patients with a chronic febrile course, enlarged parotid glands, and uveitis. Two had unilateral facial nerve palsy.22 At the time, the syndrome was ascribed to mumps. Facial palsies may follow parotitis by days or months, precede it,13 or occur independently. Facial palsy is by far the most common peripheral nervous system manifestation in sarcoidosis.31,51 It was observed in 50% of the patients by Colover.11 Heerfordt’s syndrome was associated with numbness and tingling in the hands and unsteadiness in a patient reported by Feiling and Veiner,17 which shows that overlap is possible between classic Heerfordt’s syndrome and the more diffuse meningoradicular inflammatory involvement in sarcoidosis. Although they do not belong to the peripheral nervous system, it is of interest to note that the optic nerve and the chiasma are frequently affected. Ocular involvement may occur in approximately one fourth of patients with definite sarcoidosis. The eye is often affected by anterior or posterior uveitis, optic nerve lesions, papilledema, vitreous opacities, perivenous sheathing, “candle wax dripping,” retinal and choroidal exudates, and conjunctival deposits.7 In one series of 281 patients with verified sarcoidosis, uveitis (present in nearly 10% of the patients) and periphlebitis (present in 45%) in the fundus and conjunctival nodules verified histologically were the most common ophthalmic findings.28 Ophthalmic findings were present in 28% of the patients, but neurologic abnormalities were present only in 4.3%.28 Hearing loss, vertigo, and tinnitus may occur in the absence of other neurologic symptoms. Colover noted deafness in 8 of 118 patients,11 and Delaney noted it in 3 of his 23 patients.13 In such patients, performance of magnetic resonance imaging with gadolinium enhancement can reveal conspicuous infiltrates in the symptomatic areas, but also silent lesions in other areas, especially in the meninges and the hypothalamic and hypophyseal regions.

Peripheral Neuropathy Besides cranial nerve involvement, especially the facial nerve, peripheral neuropathy is extremely uncommon in sarcoidosis. In spite of its rarity, a wide spectrum of peripheral nerve disorders can be observed in this setting, including focal and multifocal sensorimotor neuropathy, Guillain-Barré syndrome (GBS), multifocal subacute or chronic sensorimotor neuropathy, chronic symmetrical sensory polyneuropathy, and subacute multifocal sensorimotor neuropathy with conduction block. Large series of peripheral neuropathies with histologic confirmation are very rare in the literature. In our series of 11 patients with demonstration of granulomas in the nerve biopsy specimen,43 the neuropathy was focal motor and sensory in one

patient, multifocal motor and sensory in three patients, and multifocal sensory only in one. A distal symmetrical sensory polyneuropathy was present in three patients. One patient manifested a multifocal deficit of the lower limbs without actual sensory loss. In all patients the course of the neuropathy was subacute and progressive. None of the patients had CNS manifestations. The occurrence of sensory loss in large areas of bizarre outline on the trunk was noted by Heerfordt, who described anesthesia of one quadrant of the abdomen.22 Multiple Mononeuropathy The clinical picture of multiple mononeuropathy of the spinal nerves may be seen in isolation or may be associated with major involvement of the CNS. In our series, however, none of the patients manifested signs or symptoms of CNS involvement either at referral or during follow-up.43 Reported instances include serratus anterior palsy combined with a peroneal nerve lesion, bilateral median nerve palsies, radial nerve palsies, and phrenic nerve palsy.31 A progressive lesion of the cauda equina, shown to be caused by infiltration of nerve roots, has been described as the sole clinical feature of sarcoidosis.19 Multifocal Neuropathy with Conduction Block One of our patients manifested gradual weakness of right foot dorsiflexion followed, within a month, by numbness of the second and third fingers of the right hand and in the left median nerve territory. Numbness extended gradually over the other fingers and to both feet, up to midleg. A moderate motor deficit was noted in the affected territories in addition to glove-and-stocking hypoesthesia. Cerebrospinal fluid (CSF) protein was 0.68 g/L with two lymphocytes per milliliter of CSF. Electrophysiologic studies showed an 80% conduction block in the left median nerve with conduction velocity at 27 m/s. A 60% conduction block was noted in the left ulnar nerve in the forearm, with compound muscle action potential at 2.7 mV after stimulation at the wrist level versus 1.2 mV after stimulation of the ulnar nerve below the elbow. Also noted were a 40% conduction block on the median nerve at the forearm level, a 50% conduction block in the left peroneal nerve at the knee level, and decreased sensory action potential in the left sural nerve (1.1 ␮V; coefficient of variation: 30.9 m/s). Numerous perineurial sarcoid granulomas and conspicuous lesions of granulomatous angiitis were present on biopsy specimens. The neurologic deficit gradually improved when the patient was placed on corticosteroids. After 6 months, only axonal changes were found on electromyography. Purely Motor Neuropathy One patient of our series manifested a multifocal motor neuropathy that was mistaken for motor neuron disease until the results of muscle and nerve biopsies.43

2419

Sarcoid Neuropathy

Guillain-Barré–like Sarcoid Polyneuropathy Presentation as a Guillain-Barré–like syndrome is uncommon in sarcoidosis. Although multiple symmetrical neuropathies roughly mimicked a subacute inflammatory polyneuritis with a progressive onset extending over 3 months in the patient reported by Strickland and Moser,53 a rapid onset as in GBS is very uncommon. Of the two patients reported by Schott et al.,46 one manifested a subacute, painful motor deficit restricted to the lower limbs, with normal nerve conduction velocity and pleocytosis in the CSF, which fits well with the diagnosis of sarcoid neuropathy but not with GBS. The patient recovered with corticosteroids. The second patient had a typical, but likely coincidental, GBS from which he recovered spontaneously in a matter of days. One of our patients manifested a GBSlike neuropathy.43 One week after a flulike episode, this 64-year-old man complained of hand numbness and clumsiness followed by walking difficulties within 2 days. Five days later, he was admitted to an intensive care unit because of rapid respiratory failure, which required ventilatory assistance for more than 2 months. He had facial diplegia, diffuse asymmetrical weakness more marked in the lower limbs, and generalized areflexia and minor sensory changes in the feet. One week later, he manifested bradycardia and severe postural hypotension. CSF examination showed 36 lymphocytes/mL and 1.28 g/L protein. He partially improved after infusion of high doses of immune globulins. Four months later he was able to walk with aid, but still complained of paresthesias that predominated in the left leg. He was then referred because of slight worsening of his neurologic condition. Sarcoid granulomas were demonstrated in the nerve, muscle, and gastric mucosa of this patient.43 One patient of another series also manifested a GBS-like syndrome with a relapse of acute, self-limited neuropathy 3 years later.58

SARCOIDOSIS: A MULTISYSTEM DISEASE Sarcoidosis is a multisystem disease. Because neurologic manifestations can be the presenting and only evidence of the disease, it is necessary to look for involvement of other organs when sarcoidosis is suspected and after nerve or muscle biopsy findings of sarcoid lesions. Respiratory symptoms are most common and include cough, chest discomfort, and dyspnea. After the thorax, the lymph nodes, skin, and eyes are most often involved. Other organs produce signs and symptoms much less often despite the fact that there are granulomas on histologic examination in many organs in the majority of patients. Only two patients in our series were known to have sarcoidosis before the onset of the neuropathy.43 General manifestations including fever, fatigue, and weight loss were present in three patients. Clinical manifestations that could be attributed to sarcoidosis, including erythema nodosum, arthritis, enlarged periph-

eral lymph nodes, liver dysfunction, cough and dyspnea, purpura in the legs, and bilateral uveitis, were found retrospectively in eight patients. The erythrocyte sedimentation rate was normal or mildly elevated in most cases. The angiotensin-converting enzyme (ACE) levels were increased in only two patients.

Central Nervous System and Meningeal Involvement Cerebrospinal Fluid Abnormalities Elevated protein levels, pleocytosis, and increased spinal fluid pressure have been reported in about half of the patients with cranial nerve palsies and peripheral neuropathy. Symptomatic meningitis is seldom observed. Other CSF features of neurosarcoidosis include elevated ACE (about half of patients), lysozyme, and beta2macroglobulin levels and an increased CD4:CD8 ratio.30 CSF analysis is also important in excluding tuberculosis and fungal infections. Central Nervous System Involvement In our experience, symptomatic CNS involvement is seldom associated with sarcoid neuropathy. CNS manifestations of sarcoidosis include seizures, intracranial mass lesions, encephalopathy and vasculopathy, and hydrocephalus.36 CNS manifestations are often associated with a more severe and progressive course. Neuroimaging The neuroradiologic features of neurosarcoidosis variously include meningeal enhancement with contrast agents, involvement of the hypothalamic and hypophyseal regions, and high signal intensity in the white matter on T2-weighted magnetic resonance imaging. Dural thickening and/or dural mass, leptomeningeal disease, enhancing or nonenhancing brain lesions, cranial nerve lesions, and spinal cord or nerve root lesions are also noted.10 The correlation between radiologic findings and clinical manifestations is not always good. Many images are clinically silent. Sarcoid Myopathy A high percentage of patients at any stage of sarcoidosis have lesions in the muscle, which remain asymptomatic in many cases.44,52 It is our experience that virtually all patients with sarcoid neuropathy have simultaneous involvement of muscle on biopsy specimens. Symptomatic involvement of muscle takes several forms. Proximal muscles are usually affected more than distal, but the opposite also occurs. Weakness, pain, and tenderness of the involved muscles are encountered.50 Chronic disease may induce muscular wasting and weakness that may mimic motor neuron disease.12

2420

Diseases of the Peripheral Nervous System

Other Organ System Involvement Lymphadenopathy Peripheral lymphadenopathy is common. Many patients have palpable nontender lymph nodes. There is thoracic lymphadenopathy in the large majority of patients. These enlarged lymph nodes seldom cause symptoms, but rare cases of venous obstruction, including the superior vena cava syndrome and bronchial obstruction producing atelectasis, have been reported.25 Lung Disease About one third of patients have a dry cough and shortness of breath. These signs and symptoms usually clear completely. Bilateral hilar lymphadenopathy, usually symmetrical, is observed in roughly 85% of patients with sarcoidosis, and unilateral hilar lymphadenopathy in 5%. In the 20% of patients who are left with stage IV disease, there is irreversible parenchymal lung disease. The lung parenchyma is initially abnormal on chest radiographs in about one half of patients. The percentage is higher on computed tomography, which can detect disease when the radiograph is normal. If lung tissue is obtained, however, there is histologic disease in almost all patients, including those who have no visible lung abnormalities on imaging studies. This is why transbronchial biopsy has such a high diagnostic yield and is the favored means of obtaining tissue for diagnosis. Most parenchymal lung disease resolves. A minority of patients develop irreversible pulmonary fibrosis with disability ranging from minimal to death. This stage IV disease is often most severe in the upper lobes. Pulmonary fibrosis can lead to cor pulmonale and right heart failure. Bullae can be colonized by fungi, most often an Aspergillus species. Computed tomography can show disease when radiographs do not and can also better demonstrate lymphadenopathy and the characteristic features of sarcoidosis involving the lung parenchyma. This can be important when the diagnosis is uncertain, but none of the radiographic or computed tomographic findings that can be produced by sarcoidosis is pathognomonic. Skin Lesions Erythema nodosum occurs in the context of acute Löfgren’s syndrome, lupus pernio, and papules, nodules, and plaques.37 They must be sought and biopsied to confirm the diagnosis. Hepatomegaly and Splenomegaly These conditions occur in about one third of patients. Liver lesions have been found in about 5%, and spleen lesions in 10% to 15%, of patients with sarcoidosis undergoing

computed tomography examinations. The liver and spleen lesions most often appear as small, widespread foci of decreased attenuation best seen when intravenous contrast material is utilized. Osteoarticular Lesions Acute polyarthritis is a common presenting symptom of sarcoidosis, but these patients do not have radiologic abnormalities. Visible bone lesions are associated with chronic disease and are rare unless the patient has skin lesions.27 The hands and feet are sites of lesions much more often than any other bone. The distal and middle phalanges are most often involved. The lesions are often few and inconspicuous, but rarely chronic disease leads to more severe and even destructive lesions. Exocrine Gland Involvement Decreased lacrimal gland secretions can lead to dry eyes. When the parotid gland is involved, the mouth is dry also. Severe dryness of the mouth and eyes in sarcoidosis can be similar to that produced by the idiopathic form of Sjögren’s syndrome.

COURSE Improvement of peripheral neuropathy often occurs under treatment with corticosteroids. However, some patients never respond while others, especially patients with Heerfordt’s syndrome, improve spontaneously. Two main patterns of disease course can thus be observed. An acute, often self-limited, course with spontaneous recovery is observed in Heerfordt’s syndrome and in other cases with early facial palsy. Conversely, subacute progressive onset with a chronic course characterizes the other pattern of peripheral neuropathy.

ELECTROPHYSIOLOGY Electrophysiologic studies are usually consistent with a generalized axonal sensorimotor neuropathy.43,47,58 Occasional cases of GBS show mixed axonal and demyelinating features. One of our patients manifested a multifocal motor and sensory neuropathy with motor conduction blocks on several nerves along with decreased nerve conduction velocity. The patient gradually improved on corticosteroids, and only residual axonal signs were present a few months later. This suggests that multifocal demyelination can be an early stage of nerve pathology induced by sarcoid granulomas.

2421

Sarcoid Neuropathy

PATHOLOGY Nerve Biopsy Findings Sarcoid granulomas are seldom found in nerve biopsy specimens, and mainly isolated cases have been reported.18,34,35,56 In our series of 11 patients with sarcoid neuropathy and characteristic nerve lesions in the nerve biopsy specimen, inflammatory lesions included epineurial and occasional endoneurial granulomas and perineurial inflammatory infiltrates associated with variable, asymmetrical involvement of nerve fascicles and axon loss (Fig. 107–1). Granulomas predominated around nerve blood vessels and were associated with lymphocytic necrotizing vasculitis in six patients in our series.43 In the nerve specimens with mild endoneurial infiltrates, the infiltrates predominated in the subperineurial space and around endoneurial capillaries. Mononuclear cells also adhered to the endothelium of some capillaries and sometimes occluded the lumen. Fiber loss varied markedly among nerve fascicles and among patients. Large epineurial and occasionally endoneurial granulomas can coexist with near-normal nerve fiber density in all nerve fascicles and no abnormality on teased fiber preparations (Fig. 107–2). By contrast, another patient in our series had lost most nerve fibers in all but two nerve fascicles, with a high proportion of fibers at different stages of wallerian degeneration and numerous Büngner’s bands.43 On electron microscopic examination, the unmyelinated fibers were also affected and their density was decreased, roughly in parallel with that of myelinated fibers. Immunohistochemical labeling revealed that the infiltrates were predominantly made of macrophages and T cells. CD68⫹ macrophages constituted the vast majority of cells in

FIGURE 107–1 Superficial peroneal nerve biopsy (plasticembedded specimen) showing conspicuous inflammatory infiltrate in the epineurium and endoneurium. (Thionine blue staining, 1-␮ section.) See Color Plate

the granulomas. Multinucleated giant cells, which result from fusion of activated macrophages, were heavily decorated by monoclonal anti-CD68 antibody. The proportion of CD4⫹ T cells was higher than that of CD8⫹, in keeping with a delayed-type hypersensitivity reaction.

Muscle Biopsy Findings Besides the nonspecific neurogenic atrophy, characteristic noncaseating sarcoid granulomas were found in 9 of 10 patients who underwent a muscle biopsy, and necrotizing vasculitis was found in 2 patients (Figs. 107–3 and 107–4).43 Thus muscle biopsy is a very useful complement to nerve biopsy in patients with multifocal neuropathy of unknown origin.

DIAGNOSIS The diagnosis of sarcoid neuropathy is based on demonstration of multiorgan involvement with noncaseating granulomatous infiltrates. In some patients a characteristic neurologic syndrome is associated with several manifestations of sarcoidosis, which makes the diagnosis easy. Subclinical mediastinal lymph nodes or pulmonary involvement was found in eight patients in our series,43 a proportion in agreement with that found in a large series of patients in which 78% of those with neurosarcoidosis had abnormal findings on chest radiographs.48 When spinal roots are affected, chronic meningitis is usually found and should be differentiated from other causes of chronic meningitis, such as lymphoma, adenocarcinoma, Wegener’s granulomatosis, idiopathic hypertrophic

2422

Diseases of the Peripheral Nervous System

FIGURE 107–2 Superficial peroneal nerve biopsy specimen in a patient with multifocal neuropathy. Endoneurial sarcoid granuloma with preserved density of nerve fibers is shown in this fascicle. (Thionine blue staining, 1-␮ section.) See Color Plate

cranial pachymeningitis, granulomatous infection, and leukemia. The Mantoux tuberculin skin test is negative in 60% of patients with sarcoidosis. The tuberculin test was negative in all patients except in one in our series.43 The value of the Kveim test in doubtful case is problematic, and the test is not used in most centers. The serum level of ACE is elevated in roughly 50% of patients with sarcoidosis in some studies.8,48 It was elevated in only 2 of the 11 patients with histologically demonstrated sarcoid neuropathy in our series.43 In practice, the ACE level in the serum or CSF is of no diagnostic value, but can be used to monitor treatment in patients with histologically confirmed sarcoidosis and high serum ACE. Serum ACE levels are actually of limited practical value in sarcoido-

sis.25 Gallium scans to look for silent involvement of organs by sarcoidosis lack specificity, and their usefulness in this setting is not established.1,38,45 In patients presenting with peripheral neuropathy, the nerve biopsy should be combined with biopsy of the adjacent muscle, which is virtually always affected in patients with sarcoid neuropathy.43 Demonstration of granulomas in the muscle specimen is helpful to exclude tuberculoid leprosy in patients at risk for leprosy, because leprous granulomas do not affect muscles. Histologic confirmation of multisystem involvement can be provided by biopsying another affected organ. When the neuropathy is the only manifestation, bronchial biopsy, alveolar lavage, or biopsy of salivary glands can be useful to demonstrate multisystem involvement.

FIGURE 107–3 Peroneal brevis muscle biopsy showing noncaseating granuloma with multinucleated giant cell in same patient as in Figure 108–1. (Hematoxylin & eosin staining.) See Color Plate

Sarcoid Neuropathy

2423

FIGURE 107–4 Muscle biopsy specimen of a patient with multifocal neuropathy illustrating necrotizing vasculitis that was associated with sarcoid granulomas. (Hematoxylin & eosin staining.) See Color Plate

TREATMENT Corticosteroids are the cornerstone of therapy for neurosarcoidosis. When indicated, treatment should start with prednisone, 1 mg/day/kg body weight. The dose of prednisone may vary and the treatment may be required for a long time. Many patients on long-term treatment with corticosteroids develop side effects related to the high dose and longer duration of therapy. Immunosuppressive treatments, particularly methotrexate, cyclosporine, and azathioprine, have been used in resistant cases. Thalidomide is recommended by some for treatment of chronic cutaneous sarcoidosis at doses of 50 to 200 mg/day.6,16 Because of its neurotoxicity, thalidomide should be used with greater care in patients with peripheral neuropathy.

CONCLUSION Peripheral neuropathy is a rare, yet treatable manifestation of sarcoidosis, a multisystem disorder characterized by the presence of noncaseating granulomas that are seldom found in nerve biopsy specimens. The course of the neuropathy can be acute and spontaneously reversible in young patients with an early-onset cranial neuritis. Conversely, a subacute progression of multifocal neuropathy is usually followed by a more chronic course. The response to treatment with corticosteroids is often good, but the long-term prognosis remains uncertain.

REFERENCES 1. Abe, T., Tanioka, D., Izumiyama, H., et al.: Usefulness of gallium-67 scintigraphy in neurohypophyseal germinoma. Acta Neurochir. (Wien) 142:655, 2000.

2. Adams, D. O., and Hamilton, T. A.: Phagocytic cells: cytotoxic activity of macrophages. In Gallin, J. I., Goldstein, I. M., and Snyderman, R. (eds.): Inflammation: Basic Principles and Clinical Correlates. New York, Raven Press, p. 471, 1988. 3. Alajouanine, T., Bertrand, I., Degos R., et al.: Sarcoidose ganglionnaire, cutanée et oculaire avec atteinte secondaire diffuse, périphérique et centrale du système nerveux. Rev. Neurol. (Paris) 99:421, 1958. 4. Alsbirk, P. H.: Epidemiologic studies on sarcoidosis in Denmark based on a nationwide central register: a preliminary report. Acta Med. Scand. 176:106, 1964. 5. Babin, R. W., Liu, C., and Aschenbrener, C.: Histopathology of neurosensory deafness in sarcoidosis. Ann. Otol. Rhinol. Laryngol. 93:389, 1984. 6. Baughman, R. P., Judson, M. A., Teirstein, A. S., et al.: Thalidomide for chronic sarcoidosis. Chest 122:227, 2002. 7. Caplan, L., Corbett, J., Goodwin, J., et al.: Neuroophthalmologic signs in the angiitic form of neurosarcoidosis. Neurology 33:1130, 1983. 8. Chapelon, C., Ziza, J. M., Piette, J. C., et al.: Neurosarcoidosis: signs, course and treatment in 35 confirmed cases. Medicine 69:261, 1990. 9. Chen, R. C., and McLeod, J. G.: Neurological complications of sarcoidosis. Clin. Exp. Neurol. 26:99, 1989. 10. Christoforidis, G. A., Spickler, E. M., Recio, M. V., et al.: MR of CNS sarcoidosis: correlation of imaging features to clinical symptoms and response to treatment. AJNR Am. J. Neuroradiol. 20:655, 1999. 11. Colover, J.: Sarcoidosis with involvement of the nervous system. Brain 71:451, 1948. 12. Crompton, M. R., and McDermot, V.: Sarcoidosis associated with progressive muscular wasting and weakness. Brain 84:62, 1961. 13. Delaney, P.: Neurologic manifestations of sarcoidosis. Ann. Intern. Med. 87:336, 1977. 14. DeReeme, R. A.: Sarcoidosis and Wegener’s granulomatosis: a comparative analysis. Sarcoidosis 11:7, 1994.

2424

Diseases of the Peripheral Nervous System

15. Emery, J.-P., Lasserre, P.-P., and Dryll, A.: Les manifestations neurologiques périphériques de la sarcoidose. Sem. Hop. Paris 48:3039, 1972. 16. Estines, O., Revuz, J., Wolkenstein, P., et al.: Sarcoidosis: thalidomide treatment in ten patients. Ann. Dermatol. Venereol. 128:611, 2001. 17. Feiling, A., and Viner, G.: Iridocyclitis-parotitis-polyneuritis: a new clinical syndrome. J. Neurol. Neurosurg. Psychiatry 2:253, 1922. 18. Galassi, G., Gibertoni, M., Mancini, A., et al.: Sarcoidosis of the peripheral nerve: clinical, electrophysiological and histological study of two cases. Eur. Neurol. 23:459, 1984. 19. Garcin, R.: Les atteintes neurologiques et musculaires dans la maladie de Besnier-Boeck-Schaumann. Folia Psychiatr. Neurol. Neurochir. Neerl. 63:285, 1960. 20. Gordis, L.: Sarcoidosis. In Epidemiology of Chronic Lung Diseases in Children. Baltimore, Johns Hopkins University Press, p. 53, 1973. 21. Harrington, D. W., Major, M., Rybicki, B., et al.: Familial sarcoidosis: analysis of 91 families. Sarcoidosis 11:240, 1994. 22. Heerfordt, C.: Uber eine Febris uveo-parotidea subchronica. Von Graefe’s Arch. Opthalmol. 70:254, 1909. 23. Hennessy, T. W., Ballard, D. J., DeReeme, R. A., et al.: The influence of diagnostic access bias on the epidemiology of sarcoidosis: a population-based study in Rochester, Minnesota, 1935–1984. J. Clin. Epidemiol. 41:565, 1988. 24. Hosoda, Y., and Odaka, M.: History of sarcoidosis. Semin. Respir. Med. 13:359, 1992. 25. Hunninghake, G. W., Costabel, U., Ando, M., et al.: Statement on sarcoidosis (American Thoracic Society). Am. J. Respir. Crit. Care Med. 160:736, 1999. 26. Hutchinson, J.: Case of livid papillary psoriasis. In Illustrations of Clinical Surgery, Vol. 1. London, J & A Churchill, p. 42, 1877. 27. James, D. G., Neville, E., and Carstairs, L. S.: Bone and joint sarcoidosis. Semin. Arthritis Rheum. 1:53, 1976. 28. Karmi, A.: Ophthalmic changes in sarcoidosis. Acta Ophthalmol. Suppl. 141:1, 1979. 29. Kataria, Y. P., and Holter, J. F.: Immunology of sarcoidosis. Clin. Chest Med. 18:719, 1997. 30. Lynch, J. P. III, Sharma, O. P., and Baughman, R. P.: Extrapulmonary sarcoidosis. Semin. Respir. Infect. 13:229, 1998. 31. Matthews, W. B.: Sarcoid neuropathy. In Dyck, P. J., Thomas, P. K., Lambert, E. H., and Bunge, R. P. (eds.): Peripheral Neuropathy, 2nd ed., Vol. 2. Philadelphia, W. B. Saunders, p. 2018, 1984. 32. Meyer, J., Foley, J., and Campagna-Pinto, D.: Granulomatous angiitis of the meninges in sarcoidosis. Arch. Neurol. Psychiatry 69:587, 1953. 33. Moller, D. R., Forman, J. D., Liu, M. C., et al.: Enhanced expression of IL-12 associated with Th1 cytokine profiles in active pulmonary sarcoidosis. J. Immunol. 156:4952, 1996. 34. Nemni, R., Galassi, G., Cohen, M., et al.: Symmetric sarcoid polyneuropathy: analysis of a sural nerve biopsy. Neurology 31:1217, 1981. 35. Oh, S. J.: Sarcoid polyneuropathy: a histologically proved case. Ann. Neurol. 7:178, 1980.

36. Oksanen, V. E.: Neurosarcoidosis. In James D. G. (ed.): Sarcoidosis and Other Granulomatous Disorders (Lung Biology in Health and Disease, Vol. 73). New York, Marcel Dekker, p. 285, 1994. 37. Olive, K. E., and Kataria, Y. P.: Cutaneous manifestations of sarcoidosis: relationships to other organ system involvement, abnormal laboratory measurements, and disease course. Arch. Intern. Med. 145:1811, 1985. 38. Rehm, P. K.: Gallium-67 scintigraphy in the management of Hodgkin’s disease and non-Hodgkin’s lymphoma. Cancer Biother. Radiopharm. 14:251, 1999. 39. Rosen, Y.: Sarcoidosis. In Dail, D. H., and Hammer, S. P. (eds.): Pulmonary Pathology, 2nd ed. New York, SpringerVerlag, p. 13, 1994. 40. Rybicki, B. A., Major, M., Popovich, J. Jr., et al.: Racial differences in sarcoidosis incidence: a 5-year study in a health maintenance organization. Am. J. Epidemiol. 145:234, 1997. 41. Said G.: Leprous neuropathy. In Mendell, J. R., Kissel, J. T., and Cornblath, D. R. (eds.): Diagnosis and Management of Peripheral Nerve Disorders. New York, Oxford University Press, p. 551, 2001. 42. Said, G., and Hontebeyrie-Joskowicz, M.: Nerve lesions induced by macrophage activation. Res. Immunol. 143:589, 1992. 43. Said, G., Lacroix, C., Plante-Bordeneuve, V., et al.: Nerve granulomas and vasculitis in sarcoid peripheral neuropathy: a clinicopathological study of 11 patients. Brain 125:264, 2002. 44. Sartoris, D. J., Resnick, D., Resnick, C., et al.: Musculoskeletal manifestations of sarcoidosis. Semin. Roentgenol. 20:376, 1985. 45. Schattner, A., Cohen, A., Wolfson, L., et al.: Unexplained systemic symptoms and gallium-67-guided decisions. Am. J. Med. Sci. 321:198, 2001. 46. Schott, B., Michel, D., Lejeune, E., et al.: Polyradiculonévrites au cours de la maladie de Besnier-Boeck-Schaumann. J. Med. Lyon 49:431, 1968. 47. Scott, T., Brillman, J., and Gross, J.: Sarcoidosis of the peripheral nervous system. Neurol. Res. 15:389, 1993. 48. Sharma, O. P.: Neurosarcoidosis: a personal perspective based on the study of 37 patients. Chest 112:220, 1997. 49. Sheffield, E. A., and Williams, W. J.: Pathology. In James, D. G. (ed.): Sarcoidosis and Other Granulomatous Disorders (Lung Biology in Health and Disease, Vol. 73). New York, Marcel Dekker, p. 45, 1994. 50. Silverstein, A., and Siltzbach, L. E.: Muscle involvement in sarcoidosis: asymptomatic, myositis and myopathy. Arch. Neurol. 21:235, 1969. 51. Stern, B. J., Krumholtz, A., Johns, C. J., et al.: Sarcoidosis and its neurologic manifestations. Arch. Neurol. 42:909, 1985. 52. Stjernberg, N., Cajander, S., Truedsson, H., et al.: Muscle involvement in sarcoidosis. Acta Med. Scand. 209:213, 1981. 53. Strickland, G. T., and Moser, K. M.: Sarcoidosis with a Landry-Guillain-Barré syndrome and clinical response to corticosteroids. Am. J. Med. 43:131, 1967. 54. Tajima, Y., Shinpo, K., and Itoh, Y.: Infiltrating cell profiles of sarcoid lesions in the muscle and peripheral nerves: an immunohistological study. Intern. Med. 36:876, 1997.

Sarcoid Neuropathy 55. Urich, H.: Neurosarcoidosis or granulomatous angiitis: a problem of definition. Mt. Sinai J. Med. 44:718, 1977. 56. Vital, C., Aubertin, J., Ragnault, J., et al.: Sarcoidosis of the peripheral nerve: a histological and ultrastructural study of two cases. Acta Neuropathol. 58:111, 1982.

2425

57. Winkler, M.: Beitrag zur Frageder “sarkoide” (Boeck) Resp der subkutanenen nudularen Tuberkulide (Darier). Arch. Dermatol. Syphilol. (Wien) 77:3, 1905. 58. Zuniga, G., Ropper, A., and Frank, J.: Sarcoid peripheral neuropathy. Neurology 41:1558, 1991

O. Neuropathy Associated with Malignancy

108 Amyloidosis and Neuropathy ROBERT A. KYLE, JOHN J. KELLY, AND PETER J. DYCK

Microscopic Appearance and Ultrastructure Classification of Amyloidosis Primary Amyloidosis Frequency Clinical Findings

Laboratory Findings Diagnosis Prognosis Therapy Familial Amyloidosis Familial Amyloid Polyneuropathy

Amyloidosis was probably first described more than 300 years ago,124,185 when early in the 18th century Wainewright described amyloid involving the liver.47 The term amyloid was introduced by Matthias Schleiden, a German botanist, in 1838 to describe a normal amylaceous constituent of plants. “Waxy” was used interchangeably with “lardaceous” to describe the substance, which was recognized as albuminous rather than fatty by William Budd, an English physician, in the 1840s.20 In 1856, Wilks described a 51-year-old man with “lardaceous viscera” in whom there was no evidence of syphilis, osteomyelitis, other osseous disease, or tuberculosis.220 The first case of amyloidosis associated with multiple myeloma was reported by Weber in 1867.218 The use of metachromatic stains—methyl violet and crystal violet—was introduced in 1875.8 In 1928, Divry and Florkin46 described the green birefringence that occurred when tissue from the brains of patients with Alzheimer’s disease was stained with Congo red and viewed under polarized light. Cohen and Calkins31 recognized that all types of amyloid had a fibrillar structure.

Pathogenesis Diagnosis Pathology Therapy

MICROSCOPIC APPEARANCE AND ULTRASTRUCTURE Amyloid appears homogeneous and amorphous under the light microscope and stains pink with hematoxylin and eosin and metachromatically with methyl violet or crystal violet. When stained with Congo red, it produces an apple-green birefringence when viewed under polarized light. Methods for staining amyloid have been reviewed.53 Although applegreen birefringence is usually taken as the definitive histologic stain for amyloidosis, we would caution that not all that appears to produce apple-green birefringence in Congo red staining is amyloid. The modification of alkaline Congo red dye by Puchtler and Sweat165 is the recommended stain for identifying amyloid. In electron micrographs, amyloid is made up of a mat of linear, nonbranching, aggregated fibrils that are 7.5 to 10 nm wide and of indefinite length.84 Each amyloid fibril consists of two to five filaments and is arranged in an antiparallel or cross ␤-pleated sheet configuration. This is thought to be responsible for the staining and optical 2427

2428

Diseases of the Peripheral Nervous System

features of amyloid. The fibrils are insoluble, generally resist proteolytic digestion, and destroy normal tissue. The fibrils consist of the variable portions of monoclonal (␬ or ␭) immunoglobulin light chains in primary amyloidosis (AL) or, very rarely, heavy chains. In secondary amyloidosis (AA), the fibrils consist of protein A, a nonimmunoglobulin, and in familial amyloidosis (AF), they are composed of mutant transthyretin (prealbumin) or, occasionally, fibrinogen, lysozyme, or apolipoprotein. In senile systemic amyloidosis, the fibrils consist of normal transthyretin. The amyloid fibrils associated with long-term dialysis arthropathy consist of beta2-microglobulin. Serum amyloid P (SAP) component is a glycoprotein composed of 10 identical glycosylated polypeptide subunits, each with a molecular weight of 23,500 and arranged as two pentamers. SAP component is produced by the liver. It is present in healthy persons and shows 50% to 60% homology with C-reactive protein. SAP is bound to the amyloid fibrils and is not an integral part of the structure of the fibril. SAP component is found in all types of amyloid, including the vessel walls in patients with Alzheimer’s disease. The physiologic function of SAP and its pathologic role in amyloidosis are unknown. Glycosaminoglycans are present in amyloid deposits. Their role is also unknown. The catabolism of the fibrils is an important factor in pathogenesis; however, little is known about the process. The mechanism for the deposition of amyloid fibrils in AL is also unclear. Amyloid fibrils have been demonstrated within plasma cells153 and in macrophages associated with multiple myeloma.49 Excessive production of amyloid precursor proteins and degradation or slow removal of amyloid fibrils are important factors in pathogenesis, but more work needs to be done to elucidate the mechanisms of these processes.133

CLASSIFICATION OF AMYLOIDOSIS Metachromatic stains (methyl violet and crystal violet), Congo red, and electron microscopy do not distinguish among the various types of amyloidosis,33 nor does organ

distribution of amyloid involvement allow one to differentiate the types of systemic amyloidosis. Immunohistochemical testing is useful for identification of protein-specific types of amyloid. Antiserum to SAP component reacts with all types of amyloid and is useful for its identification.38,186 Antisera for AL (␬ and ␭), AA, and AF permit accurate identification of the type of amyloid fibril (Table 108–1).140 The classification in Table 108–2 is based on the identification of the protein thought to be the major constituent of the amyloid fibril. Several of these types of amyloidosis are not discussed in this chapter because neuropathy is not a usual feature: AA; localized amyloidosis; AF of nephrotic, cardiopathic, or cutaneous type; senile cardiac amyloidosis (AScl); isolated atrial amyloid; Alzheimer’s disease; hereditary cerebral hemorrhage with amyloidosis (Icelandic and Dutch types); amyloid associated with endocrine abnormalities; and amyloidosis associated with dialysis.

PRIMARY AMYLOIDOSIS The relationship of Bence Jones protein to amyloid was first postulated by Magnus-Levy more than 70 years ago.143 Osserman et al.158 proposed that Bence Jones proteins were involved directly in the production of amyloid and that these light chains had a greater affinity for certain tissues. A major advance occurred when Glenner et al.85 demonstrated that amyloid fibrils from a patient with AL were virtually identical to the variable portion of the patient’s monoclonal light chain (Bence Jones protein). Thus the variable portion of a monoclonal light chain or the complete light chain constitutes the typical fibril of AL. Lambda (␭) light chains are more frequent than kappa (␬) (2:1) in AL. This ratio is the reverse of that in multiple myeloma and other monoclonal gammopathies. The ␭VI subgroup accounts for more than half of amyloidosis cases consisting of ␭ light chains.190 Almost all recognized ␭VI proteins have been associated with amyloidosis. It has been suggested that patients with amyloidosis have aberrant de novo synthesis of light chains or abnormal proteolytic

Table 108–1. Systemic Amyloidosis Amyloid Stains Amyloid Type Primary (AL) Secondary (AA) FMF (familial Mediterranean fever) Associated with long-term hemodialysis Familial (AF) Senile systemic (AS)

Congo Red

␬ or ␭

Serum Amyloid A

Beta2-microglobulin

Transthyretin (Prealbumin)

⫹ ⫹ ⫹

⫹ ⫺ ⫺

⫺ ⫹ ⫹

⫺ ⫺ ⫺

⫺ ⫺ ⫺

⫹

⫺

⫺

⫹

⫺

⫹ ⫹

⫺ ⫺

⫺ ⫺

⫺ ⫺

⫹ ⫹

Modified from Kyle, R. A., and Gertz, M. A.: Primary systemic amyloidosis: clinical and laboratory features in 474 cases. Semin. Hematol. 32:45, 1995.

2429

Amyloidosis and Neuropathy

Table 108–2. Clinical Classification of Amyloidosis* Amyloid Type

Classification

Major Protein Component

AL

1. Primary: no evidence of preceding or coexisting disease except multiple myeloma 2. Secondary: coexistence with other conditions such as rheumatoid arthritis or chronic infection 3. Localized: involvement of a single organ without evidence of generalized involvement 4. Familial A. Neurologic (type I) Portuguese Japanese Swedish Indiana (type II) Illinois (German) B. Nephrotic Familial Mediterranean fever Muckle-Wells C. Cardiopathic Danish Appalachian 5. Senile Senile cardiac amyloid Isolated atrial amyloid 6. Endocrine Medullary carcinoma Islets of Langerhans 7. Dialysis arthropathy

␬ or ␭ light chains

AA AL ATTR

AScl AANF AEl AEp A␤2M

Protein A ␬ or ␭ light chains

TTR Met 30† TTR Met 30 TTR Met 30 TTR Ser 84 TTR Tyr 77 Protein A Protein A TTR Met 111 TTR Ala 60 TTR (prealbumin) Atrial natriuretic peptide (ANP) Calcitonin Islet amyloid polypeptide (IAPP) Beta2-microglobulin

*AA, localized amyloidosis, Alzheimer’s disease, and hereditary cerebral hemorrhage with amyloidosis are not included. † TTR ⫽ transthyretin. Modified from Kyle, R. A., and Greipp, P. R.: Amyloidosis (AL): clinical and laboratory features in 229 cases. Mayo Clin. Proc. 58:665, 1983.

processing, or both.24 This concept is supported by the demonstration of large amounts of free monoclonal light chains and light-chain fragments in biosynthesis experiments in primary amyloidosis.164 An animal model for AL has not been found.

Frequency Primary amyloidosis was found in 6 (0.2%) of 3414 autopsies conducted from 1937 to 1946 at the Royal Victoria Hospital in Belfast, Ireland, and in 43 (0.4%) of 11,586 autopsies performed at the same institution from 1961 to 1970.203 During the 12-year period from 1981 through 1992, 1315 patients with amyloidosis were evaluated at Mayo Clinic. Almost three fourths had the primary type (Table 108–3). During 2002, 121 new cases of AL were seen. These accounted for 12% of the patients at Mayo Clinic with a plasma cell proliferative process. The predominance of AL is due, at least in part, to the tendency for patients with a rare disease to come to a tertiary care center where there is an interest in the disease and prospective therapeutic studies are in progress. The type of amyloidosis depends on the geographic area and the type of practice. The incidence of

AL in Olmsted County, Minnesota, is 0.89 per 100,000 (8.9/million person-years).133 Applying this rate to the U.S. population, one would expect approximately 2500 new cases annually. The median age at diagnosis is 65 years, with only 1% younger than 40 years. Patients with AL can be divided into two groups—AL and AL with multiple myeloma—on the basis of the number and appearance of plasma cells in the bone marrow, Table 108–3. Distribution of 1315 Patients According to Type of Amyloidosis (Mayo Clinic, 1981–1992) Type Primary (AL) Secondary (AA) Localized Familial Senile Total

No. of Patients

%

918 45 245 55 52 1315

70 3 19 4 4 100

Modified from Kyle, R. A., and Gertz, M. A.: Amyloidosis. In Wiernik, P. H., Canellos, G. P., Kyle, R. A., and Schiffer, C. A. (eds.): Neoplastic Diseases of the Blood, 2nd ed. New York, Churchill Livingstone, p. 525, 1991.

2430

Diseases of the Peripheral Nervous System

the amount of monoclonal protein in the serum and urine, and the presence or absence of skeletal lesions. However, these findings often overlap in patients with and those without multiple myeloma. Because the amyloid fibrils consist of the variable portion of a monoclonal light chain in both groups and because the presence of multiple myeloma does not influence survival during the first year after the diagnosis of amyloidosis, both categories should be considered together as AL.

Clinical Findings Cutaneous nerve biopsy (usually sural or superficial peroneal) is the only certain way to recognize amyloidosis in a peripheral nerve. Rajani et al.169 reported that 1.2% of 1098 sural nerve biopsy specimens had amyloid deposits. Of the 13 patients affected, the neuropathy was mainly sensory in 6, motor in 2, and mixed in the other 5. Amyloid was identified in the endoneurium of 12 of the 13 biopsy specimens, in the epineurium in 9, and in the perineurium in 2. Axonal loss was moderate or severe in 12 specimens, and in 8 of 10 specimens axonal degeneration predominated over demyelination. In our experience, the diagnosis of polyneuropathy associated with amyloidosis should be suspected when (1) the course is progressive; (2) the pattern is that of symmetrical polyradiculoneuropathy—all classes of fibers (motor, sensory, and autonomic) are affected, and proximal and distal limbs are involved (distal findings predominate); (3) patients have positive neuropathic symptoms (prickling, asleepnumbness, and various pains); and (4) there are constitutional symptoms (e.g., weight loss) or organ involvement.51 In a few cases and early in the course of disease, selective sensory loss, especially of small sensory and autonomic fibers, produces a pattern akin to that seen in syringomyelia, but it is due to involvement of small myelinated sensory fibers (A␦ fibers, subserving cold sensation), unmyelinated fibers (many subserving nociception), and autonomic fibers (several classes).52 Quattrini et al.167 also emphasized the loss of small myelinated and unmyelinated fibers. Of 26 patients evaluated at the Mayo Clinic who had sural nerve biopsy–proven amyloid neuropathy and monoclonal protein in the serum or urine, 81% had paresthesias, 65% had muscle weakness, and 58% had numbness. Autonomic neuropathy was present in 17 of the patients at the time of diagnosis of amyloidosis. The median duration of symptoms before diagnosis was 29 months, and the median survival was 25 months.170 We have recognized three separate families who had two first-degree relatives with AL.65 Consequently, a genetic predisposition or an environmental factor may play some role in AL. Primary amyloidosis presents in one or more of the following ways: systemic or generalized symptoms, peripheral nerve symptoms, cardiovascular involvement, renal dis-

ease (nephrotic syndrome or renal insufficiency), and disorders of various other organs. Systemic Symptoms and Physical Findings Weakness or fatigue and weight loss are the most frequent symptoms and occur in more than half of patients. Weight loss may be striking, with a median loss of more than 20 lb.127 The major initial physical findings include ankle edema, hepatomegaly, purpura, and macroglossia. The liver is palpable in about one fifth of patients. Splenomegaly is found in less than 5% of patients, and when present it is only of modest size. Macroglossia may be impressive (Fig. 108–1), but it is found in only one tenth of patients. Purpura is common and usually involves the neck and face, particularly the upper eyelids. The skin may be fragile and easily traumatized. Specific organ involvement or syndromes are cardinal features of AL and are discussed individually below. The incidence of various syndromes is shown in Figure 108–2. Peripheral Neuropathy This is found at the time of diagnosis in approximately one sixth of our patients with AL. Symptoms of peripheral neuropathy or carpal tunnel syndrome are usually present for at least 1 year before diagnosis. In contrast, in patients with congestive heart failure, orthostatic hypotension, and nephrotic syndrome, the symptoms are present for approximately 3 months before diagnosis. The experience of neurologists with a large tertiary referral base will be quite different. Amyloidosis patients are often referred with an

FIGURE 108–1 Macroglossia and AL (primary systemic amyloidosis). (From Kyle, R. A., and Greipp, P. R.: Amyloidosis [AL]: clinical and laboratory features in 229 cases. Mayo Clin. Proc. 58:665, 1983, with permission of Mayo Foundation.)

Amyloidosis and Neuropathy

FIGURE 108–2 Syndromes present at diagnosis and during follow-up of patients with AL. Some patients had more than two syndromes at presentation. CHF ⫽ congestive heart failure; Ortho hypo ⫽ orthostatic hypotension. (From Kyle R. A., and Gertz, M. A.: Systemic amyloidosis. Crit. Rev. Oncol. Hematol. 10:49, 1990, with permission of Elsevier Science Ireland.)

undiagnosed neuropathy, and these patients may or may not have systemic symptoms or other organ involvement. The neuropathy of AL may be quite variable, but it frequently is sensorimotor and tends to affect lower limbs more than upper limbs and distal limbs more than proximal limbs. Small sensory and autonomic fibers are also affected. This type of neuropathy with constitutional symptoms of weight loss and cardiac and other organ involvement should lead physicians to suspect AL. Many patients with AL complain of painful numbness and varying types of limb discomfort, including stabbing, shocklike lancinating pains, and burning of extremities. Pain and temperature sensation are usually affected more than discriminative perception (vibration, position, and touch localization) and are in a distal, stocking-glove distribution. Occasionally, discriminative sensation is more affected. Symptoms of autonomic dysfunction such as orthostatic dizziness and fainting, diarrhea, bladder dysfunction, and impotence are common. Postural blood pressure drop with a fixed pulse is often present, suggesting neurogenic hypotension. In a study of the mechanism of postural hypotension, Low et al.142 found that patients with amyloidosis and postural hypotension had a decreased number of intermediolateral cell column neurons (about 60% of normal at the T7 level of the spinal cord). An occasional patient may present with only autonomic involvement and no sensorimotor symptoms. Muscle weakness is less common at presentation and distal in distribution. Reflexes are lost, with a distal predominance. Many of these patients appear systemically ill, and medical complications of amyloidosis overshadow the neuropathy. However, in one study,113 about 40% of patients presented mainly with neuropathy and had varying degrees of involvement of other organs. The discovery of involvement of the heart, liver,

2431

kidneys, or gastrointestinal tract may suggest the diagnosis of amyloidosis. The presence of carpal tunnel syndrome, which occurred in about 25% of patients in our study,113 may also suggest the diagnosis when associated with neuropathy or systemic symptoms. In all cases, the neuropathy is inexorably progressive and associated with organ dysfunction. Patients often require medications to control pain and autonomic symptoms, but death is typically due to organ failure and not neuropathy.113 Li et al.138 reviewed the clinical characteristics of immunohistochemically confirmed (in sural nerve) cases of AL. In 18 patients with AL, loss of touch sensation, paresthesias, and pain (aching, burning, or lancinating) initiated the syndrome in 14 patients. In the four other patients, the first symptom was weakness in two, weight loss in one, and postural hypotension in one. The sensory loss began in the feet slightly more frequently than in the hands, but it soon was present in both the feet and the hands. Paresthesias (prickling numbness) were the first sensory symptom in more than half the patients, and lancinating pains (usually in distal limbs) occurred in approximately one fourth. Other pains were “deep aching” and superficial “burning.” Some patients reported excessive coldness as an initial symptom. Typically, all developed sensory, motor, and autonomic deficits over time. Both upper and lower limb structures and proximal and distal segments became affected. In Li et al.’s study,138 there was considerable similarity in the development of neuropathy between the primary and familial types of systemic amyloidosis. The age at onset and the sex ratio were not significantly different between transthyretin-positive and AL cases. In both varieties, sensory symptoms (loss of touch sensation, paresthesia, and pain—burning and lancinating) were early and prominent. The lower limbs tended to be more affected than upper limbs, and the distal segments of limbs were affected more than the proximal segments. Although survival is thought to be better for transthyretin-positive patients than for those with AL, this could not be studied in our cohort. The number and size of amyloid deposits in nerve and the rate of axonal degeneration in the sural nerves were not significantly different between the primary and the familial types. It should, however, be appreciated that phenotypic variation might be revealed if disorders caused by specific gene mutations were studied. Electromyographic Findings. Motor nerve conduction velocities are usually just below normal values, with decreased or absent compound muscle action potential amplitudes.112,113 Distal motor latencies tend to be normal except in the median nerves, where the presence of carpal tunnel syndrome often lengthens the latency. The amplitude of sensory nerve action potentials is more frequently abnormal than that of compound muscle action potentials. In most cases, sensory nerve action potentials cannot be recorded. In most patients, needle electromyography

2432

Diseases of the Peripheral Nervous System

demonstrates fibrillation and neurogenic motor unit potentials. In the proper setting, the findings of an axonal peripheral neuropathy on electrodiagnostic testing, especially if associated with a superimposed carpal tunnel syndrome, should suggest the possibility of amyloid neuropathy.112 Pathogenesis. The pathogenesis of neuropathy in AL is unclear. The absence of the blood-nerve barrier in the dorsal root and sympathetic ganglia may allow access of amyloidogenic proteins and produce damage to the neural structures.210 Macrophages have been noted near the amyloid deposits in a patient with AL and peripheral neuropathy. It was postulated that the monoclonal light chain was absorbed and processed by the macrophage to produce amyloid fibrils.211 Biclonal gammopathy (immunoglobulin M [IgM] ␬ and immunoglobulin G [IgG] ␭) was recognized in a patient with sensorimotor peripheral neuropathy. The IgM ␬ protein was bound to the Schwann cells and may have produced demyelinating changes. The IgG ␭ was fixed to the amyloid deposits between nerve fibers.105 Li et al.138 have shown that the predominant pathologic alteration of myelinated fibers is axonal degeneration. This and the electromyographic data are strongly against amyloidosis being a demyelinating disorder. The authors found that amyloidosis typically forms a cuff around endoneurial microvessels. Sometimes, multiple deposits adjacent to nerve vessels can be shown. Although deposits of amyloid indent fibers and even cause demyelination and possibly axonal degeneration, it seems unlikely that this is the major mechanism by which fiber injury occurs. Deposition of amyloid adjacent to the soma of primary afferent neurons might in certain instances explain selectivity of fiber involvement.52 Kernohan and Woltman115 assumed that the fiber degeneration reflected ischemic damage. Our finding of prominent cuffs of amyloid around endoneurial microvessels at least raises the possibility of endoneurial hypoxia as a putative mechanism. Pathology. Axonal degeneration, sometimes with predominant involvement of the small myelinated and unmyelinated fibers, is seen. In the presence of peripheral neuropathy, amyloid deposits, usually in globular or diffuse form, are found infiltrating epineural and endoneurial connective tissue. In addition, epineurial and endoneurial blood vessel walls are frequently thickened by amyloid deposits. There is a gross loss of myelin fibers, which often show signs of active axonal degeneration. Electron microscopy often reveals a marked loss of unmyelinated fibers. Teased fiber studies show a predominance of axonal degeneration. In a study by Sommer and Schroder,191 in four of five cases of AL with polyneuropathy, the amyloid deposits reacted with antisera to ␬ or ␭ light chains. Electron microscopy revealed bundles of immunogold-labeled

amyloid fibrils in coated and uncoated single membranebound vesicles of endoneurial macrophages. Schwann cells did not contain intracellular amyloid, but their processes were entangled in amyloid fibrils and their basement membranes were sometimes fused with fibrillar masses. The authors concluded that the monoclonal light chains precipitated and formed amyloid fibrils in the presence of, and presumably with the assistance of, endoneurial cells. They postulated that inefficiency of phagocytosis was one of the major factors in developing amyloidosis.191 In summary, amyloid infiltrates the proximal dorsal root and sympathetic ganglia early in the course of the disease. Amyloid also infiltrates the peripheral nerves, where it deposits between nerve fibers and compresses and distorts the nerves. The vessel lumens are frequently compromised by amyloid deposition. The nerve fibers develop axonal degeneration and, to a lesser extent, segmental demyelination. Carpal Tunnel Syndrome. Carpal tunnel syndrome was an initial presenting syndrome in one fourth of our patients with AL. However, amyloidosis is infrequently found in the tenosynovium of patients presenting only with carpal tunnel syndrome. In a series of 1215 patients undergoing carpal tunnel decompression at Mayo Clinic from 1930 to 1960, systemic disease existed in 26%. Rheumatoid arthritis (7.7%) and myxedema (6.3%) were the most common causes. Of the patients with systemic disease, more than half (54%) had rheumatoid arthritis or myxedema. Amyloid deposits were found in 12 (7%) of 177 patients in this series who had histologic evaluation of the tissue.222 The symptoms of carpal tunnel syndrome consist of paresthesias in the distribution of the median nerve, which supplies the skin of the first three fingers and the lateral half of the fourth finger on the palmar aspect. The symptoms are frequently worse at night, when paresthesias or pain awakens the patient. The symptoms may also be aggravated by use of the hands, especially when driving, talking on the telephone, or holding up a book or paper to read. The median nerve is entrapped by swelling from synovitis or deposition of amyloid in the tenosynovium or the transverse carpal ligament (flexor retinaculum). Later, atrophy of the thenar muscles may occur. The clinical diagnosis is made on the basis of the history, hyperflexion of the wrist (Phalen’s maneuver), electrodiagnostic studies, and relief of symptoms by surgery.193 Electromyography is helpful for differentiating carpal tunnel syndrome from a cervical root lesion. Patients younger than 40 years are at low risk for carpal tunnel syndrome.110 Although the frequency of amyloid deposits in the tenosynovium or carpal ligament is low, one must be alert for the possibility of amyloid. Kyle et al.125 reported 152 patients with amyloid in the tenosynovium at the time of carpal tunnel release. Of these patients, 24 had AL, 3 had

Amyloidosis and Neuropathy

AA, and 1 had AF. The remaining 124 patients (82%) had carpal tunnel syndrome with local deposition of amyloid and no evidence of systemic disease. Only 2 of the 124 patients developed systemic amyloidosis—9 and 10 years after recognition of tenosynovial amyloid. Of particular interest were 12 patients who had a monoclonal protein in the serum or urine. None of these developed systemic amyloidosis or multiple myeloma during a median follow-up of 14 years. It is apparent that amyloid may be localized to the tenosynovium and that systemic amyloidosis rarely develops during long-term follow-up.125 The amyloid localized to the carpal ligament is composed of transthyretin (prealbumin) rather than immunoglobulin light chains.129 One must consider the possibility of AL in any patient with carpal tunnel symptoms despite the fact that amyloid localized to the tenosynovium is more common in older persons. AL must be excluded when the patient presents with carpal tunnel syndrome and has a monoclonal light chain in the serum or urine or when the patient has cardiac, renal, or gastrointestinal abnormalities. Other Neurologic Features The neurologic findings in AL may be atypical. Widespread fasciculation, atrophy, fasciculation of the tongue, dysphasia, hyperreflexia, and Hoffmann’s sign consistent with a clinical diagnosis of amyotrophic lateral sclerosis were described in one patient with AL and typical amyloid neuropathy.1 In another patient with AL and congestive heart failure, the initial presentation was facial diplegia.146 Multiple bilateral cranial nerve deficits involving the trigeminal, facial, abducens, and hypoglossal nerves have been seen.141 Traynor et al.206 recognized six patients with AL and cranial neuropathy. In three of the six patients, multiple cranial nerves were involved. The cranial nerves most commonly affected were the facial, trigeminal, and oculomotor nerves. Rare patients may present with a very asymmetrical neuropathy affecting one leg much more than the other, suggesting a disorder such as diabetic or vasculitic neuropathy. Cardiovascular Involvement Congestive heart failure is present in approximately 25% of patients at the time of diagnosis of AL. An additional 10% develop congestive heart failure during the course of the disease. Echocardiographic evidence of cardiac involvement without congestive heart failure is frequently obtained. The presence of a monoclonal protein in the serum or urine of any patient with congestive heart failure should raise the possibility of amyloidosis. Low voltage in the limb leads or changes consistent with anterior septal infarction are common electrocardiographic findings. The patients who have findings consistent with infarction do not have evidence of myocardial infarction at autopsy.189 Atrial fibrillation, atrial or junctional tachycardia, heart block, and ventricular premature complexes are common electrocardiographic features.172

2433

Echocardiography is a valuable technique for the demonstration of amyloid heart disease. There is a direct correlation between increasing thickness of the ventricular wall and septum and the incidence of congestive heart failure.40 Abnormalities of diastolic function in AL are frequently seen with comprehensive Doppler echocardiographic evaluation.121 Abnormal relaxation is seen initially, whereas restrictive hemodynamics occur with advanced involvement. The ejection fraction is usually normal until late in the disease. Occasionally, patients have endomyocardial biopsy–proven amyloid when the echocardiogram is not suggestive of amyloid infiltration.66 The severity of cardiac infiltration in AL may be unrelated to the whole-body amyloid load when measured by SAP tracer studies. The combination of radiolabeled SAP and echocardiography permits the identification of patients who have severe cardiac disease and a low total body load.30 Claudication of the jaw occurs in almost 10% of patients with AL. Claudication involving the legs or arms may also be present.73 Orthostatic hypotension is found in about one sixth of our patients at the time of diagnosis. Lightheadedness and syncope may be severe enough to prevent ambulation.132 Up to one third of patients with cardiac amyloid have syncope, which is often precipitated by physiologic stress. Stress-precipitated syncope is associated with a poor prognosis (median survival, 2 months) and is frequently a precursor of sudden cardiac death.28 Occasionally, symptomatic ischemic heart disease can occur from amyloidosis even when coronary angiograms are normal.3 Renal Disease Nephrotic syndrome is present in one third of patients with AL at the time of diagnosis. The degree of proteinuria in nephrotic syndrome does not correlate well with the extent of amyloid deposition in the kidneys.217 The degree of proteinuria correlates best with the extent of tufts and detachment of epithelial cells.108 Nephrotic syndrome may be associated with minimal deposits of amyloid, and a kidney biopsy specimen may be interpreted as minimalchange glomerulopathy. The specimen must be stained carefully for amyloid, particularly in older patients.103 Gross hematuria is rare in AL. Deposits of amyloid and free light chains may occur.25 Other Organ Involvement The respiratory tract is frequently involved microscopically, but symptoms rarely occur. Diffuse interstitial or reticular nodular amyloid involving the lungs has been reported. Patients frequently have dyspnea, and median survival is shortened.208 The prevalence of pulmonary amyloidosis may be higher among patients who have IgM monoclonal gammopathy or Waldenström’s macroglobulinemia.226 Histologic involvement of the gastrointestinal tract is common, but most patients are asymptomatic.

2434

Diseases of the Peripheral Nervous System

Computed tomographic scans in 23 patients with AL showed bowel wall thickening in 4, bowel wall dilatation without thickening in 4, mesenteric soft tissue infiltration in 2, and mesenteric lymphadenopathy in 1; a completely normal scan was common.7 Abnormalities in motility and reduction of lower esophageal sphincter pressure have been reported in systemic amyloidosis.174 Decreased motor activity of the small bowel is not uncommon. In some cases, pseudo-obstruction occurs.135 The clinical presentation and radiographic features resemble those of acute small bowel pseudo-obstruction.123 Autonomic dysfunction rather than deposition of amyloid in the mucosa is a much more frequent cause of severe diarrhea and gastrointestinal symptoms.10 In one series, 13 of 58 patients with AL had gastrointestinal bleeding.223 Massive intestinal hemorrhage requiring surgical resection occurred in two patients with AL. The site of the hemorrhage was at the border between the muscularis mucosa and the overlying mucosa. Decreased motility and increased rigidity of the musculature may result in shear forces that can lead to tearing of the vessels.106 Occasionally, gastric deposits of amyloid may produce nausea, vomiting, and weight loss. Severe diarrhea may be a manifestation of AL.212 This may occur from mucosal infiltration, but autonomic dysfunction of the gastrointestinal tract is a major element. Malabsorption characterized by diarrhea or steatorrhea may be a prominent feature of AL.93 Liver involvement by amyloidosis is common. The liver is palpable in approximately one fifth of patients at the time of diagnosis. In a series of 80 patients with antemortem biopsy-proved hepatic AL, liver function tests were not helpful in the differential diagnosis.67 The presence of liver amyloidosis is suggested by the presence of proteinuria, a monoclonal protein in the serum or urine,

Howell-Jolly bodies in a peripheral blood smear, and hepatomegaly with less-than-expected liver function abnormalities.72 Spontaneous rupture of the liver is rare and usually fatal.22 Petechiae, ecchymoses, papules, plaques, nodules, tumors, bullous lesions, alopecia, and dystrophy of the nails may be seen in systemic amyloidosis.18 Deposition of amyloid in the extraocular muscles may produce ophthalmoplegia.109 Amyloidosis may involve para-articular structures and be mistakenly diagnosed as seronegative rheumatoid arthritis.159 Prolongation of the thrombin time is the most common coagulation abnormality. Low levels of factor X may also be seen.61 Occasionally, patients present with muscle infiltrated by amyloid. Myopathy caused by ischemia from extensive deposits of amyloid in the walls of small vessels has been observed.19 Most patients have diffuse muscle weakness related to muscle atrophy (a consequence of denervation or, possibly, chronic vascular occlusion). In a series of 12 patients with amyloid deposits in muscles, only 3 had skeletal muscle pseudohypertrophy.71 Pseudohypertrophy of skeletal muscles may occur because of extensive deposition of amyloid (Fig. 108–3).173 Heterogeneous signal intensities seen on magnetic resonance imaging have been reported with amyloid myopathy. The signal is generally lower than that in normal muscle, with intermingled areas of hyperintensity on T2weighted images.35 Progressive muscular stiffness from deposits of amyloid in the skeletal muscle has been described.137 This patient had areas of muscle atrophy presumably from ischemia caused by deposition of amyloid in blood vessels and capillaries, but neurogenic atrophy cannot be excluded. Muscle atrophy and weakness may also be due to amyloid deposits covering muscle fibers.102 Amyloid may involve a wide variety of organs.126

FIGURE 108–3 Pseudohypertrophy of skeletal muscles from infiltration by amyloid. (From Kyle, R. A., and Greipp, P. R.: Amyloidosis [AL]: clinical and laboratory features in 229 cases. Mayo Clin. Proc. 58:665, 1983, with permission of Mayo Foundation.)

2435

Amyloidosis and Neuropathy

Laboratory Findings Hematologic and Chemistry Values Anemia is not a prominent feature of AL. When present, it is usually due to multiple myeloma, renal insufficiency, or gastrointestinal bleeding. Leukocyte and differential counts are usually normal. Thrombocytosis (more than 500,000 platelets/mm3) occurs in 9% of patients with AL.127 Functional hyposplenism from amyloid deposition may be an important factor in thrombocytosis. Thrombocytopenia is not a feature of AL. The erythrocyte sedimentation rate is usually normal or modestly elevated. Approximately 80% of patients have proteinuria. A serum creatinine level of more than 1.3 mg/dL is seen in almost half of patients at the time of diagnosis. Approximately one fourth of all patients have a serum creatinine level of 2 mg/dL or more at diagnosis. Hypoalbuminemia is common with nephrotic syndrome. The serum alkaline phosphatase level is increased in about one fourth of patients. Hyperbilirubinemia is an infrequent finding and, when present, is an ominous sign. The factor X level is decreased in 15% of patients but is rarely the cause of bleeding. The prothrombin time is increased in approximately 15% of patients, but the thrombin time is prolonged in 60%.62 The carotene and vitamin B12 levels are each reduced in approximately 5% of patients. Serum Protein Findings A localized band is found in the serum protein electrophoretic pattern in one half of patients. Even when present, the protein spike is of modest size. Hypogammaglobulinemia is seen in nearly one fourth of patients, and the rest have a normal-appearing electrophoretic pattern. Serum IgG levels are reduced in almost half of patients, whereas immunoglobulin (Ig)A and IgM values are often normal. Two thirds of patients with AL have a monoclonal protein on immunoelectrophoresis or immunofixation. Almost 20% have a free monoclonal light chain (Bence Jones proteinemia), and 45% have a monoclonal intact immunoglobulin in the serum (Table 108–4). Immunoelectrophoresis or immunofixation is necessary because the serum protein electrophoretic patterns of patients with Bence Jones proteinemia rarely have a discrete band or spike. Almost three fourths of patients with AL have ␭ light chains in their sera, whereas two thirds of patients with multiple myeloma or monoclonal gammopathy of undetermined significance (benign monoclonal gammopathy) have ␬ chains. Urine Protein Findings Electrophoresis frequently shows an albumin peak. A globulin band is seen in more than two thirds of patients, but it is usually small, and immunoelectrophoresis or immunofixation is necessary. This is particularly true in patients with nephrotic syndrome.

Table 108–4. Results of Serum Immunoelectrophoresis at Diagnosis of Primary Amyloidosis (AL) Result

% of Patients

IgG IgA IgM IgD ␬ ␭ Negative Total

31 11 4 1 6 13 34 100

From Kyle, R. A., and Gertz, M. A.: Amyloidosis. In Wiernik, P. H., Canellos, G. P., Kyle, R. A., and Schiffer, C. A. (eds.): Neoplastic Diseases of the Blood, 2nd ed. New York, Churchill Livingstone, p. 525, 1991, with permission.

Two thirds of patients with AL have a monoclonal light chain in the urine on immunoelectrophoresis or immunofixation. A monoclonal ␭ protein is more than twice as common as ␬. In our experience, a monoclonal protein is present in the serum or urine in approximately 90% of patients with AL. A monoclonal protein is most common in nephrotic syndrome and least common in patients with peripheral neuropathy. Bone Marrow Findings Only 15% of our patients have more than 20% plasma cells in the bone marrow. Most of these patients do not have symptomatic myeloma despite the significant plasmacytosis. Almost three fourths of our patients have 10% or fewer plasma cells in the bone marrow. The median percentage of plasma cells is 6%. Radiographs of AL patients are normal unless patients have accompanying multiple myeloma. Occasionally, amyloidomas (large deposits of amyloid) may produce lytic lesions.

Diagnosis The diagnosis of AL depends on the demonstration of amyloid deposits in tissue.197 The most commonly used stain is Congo red, which produces an apple-green birefringence with polarizing light. Metachromatic stains may also be useful, particularly methyl violet staining of paraffin sections of sural nerve tissue. It is a useful survey stain because the metachromasia stands out from the background. It should be appreciated that smeared mast cells may bear a resemblance to deposits of amyloid. Electron microscopy has detected diagnostic amyloid fibrils in several cases in which staining of tissue with Congo red or metachromatic stains was negative, but electron microscopy is tedious and expensive. The presence of amyloid may be demonstrated by antiserum to SAP component because it reacts with all types of

2436

Diseases of the Peripheral Nervous System

amyloid.186 123I-labeled purified human SAP component was found in amyloid deposits of AL, AA, and beta2microglobulin types. There was complete absence of uptake in control subjects. High-resolution scintigraphic images revealed minor deposits of amyloid, such as in the carpal tunnel.92 However, the ability of 123I-labeled SAP scans to recognize amyloid deposits in the heart, digestive tract, and kidneys is limited.87 Even though radiolabeled SAP scans can detect amyloid deposits,83 diagnosis requires the demonstration of amyloid deposits in biopsy tissue.101 This is potentially of help in following the course of amyloidosis in a patient with systemic disease. Utilization of antisera to AA, A␬, A␭, AF, and AScl correctly identified the type of amyloid in 98% of 114 patients with systemic amyloidosis.140 Treatment with formic acid enhances immunoreactivity of amyloid tissues to anti–prion protein, anti–␤-protein in Alzheimer’s disease, anti–amyloid A, and anti-prealbumin antisera.117 Immunohistochemistry is useful for diagnosis because of the increasing availability of satisfactory antisera. Generally, AA amyloid and beta2-microglobulin amyloid lose their affinity for Congo red and their polarization characteristics after exposure to potassium permanganate, whereas AL, AScl, AF, and nodular pulmonary amyloid are resistant to potassium permanganate.221 However, the affinity of AL amyloid for Congo red may be decreased after treatment with potassium permanganate, so this test is not reliable.188 The possibility of AL must be considered in every patient who has a monoclonal protein in the serum or urine and who also has sensorimotor peripheral neuropathy—particularly if autonomic involvement is present. The presence of carpal tunnel syndrome, nephrotic syn-

drome, congestive heart failure, hepatomegaly, or idiopathic malabsorption with an associated monoclonal protein also suggests the possibility of AL. Appropriate tissue must be obtained for the histologic identification of amyloid. The initial diagnostic procedure should be an abdominal fat aspirate, which is positive in more than 70% of patients.80 Overstaining and underdecolorization are potential problems with this technique. One must take care to avoid confusion with collagen tissue. A bone marrow aspirate and biopsy are positive in about half of patients (Fig. 108–4). The bone marrow also provides an estimate of the number of plasma cells. If the bone marrow and abdominal fat are negative, a rectal biopsy may be performed. Rectal biopsy is positive in 70% of cases, but it must include the submucosa because that tissue is involved more frequently than mucosa. If these sites are negative, tissue should be obtained from the suspected involved organ. Renal biopsy results in a high frequency of positive findings and can be done with minimal risk. Liver biopsy is frequently positive, and in our experience bleeding is uncommon. The sural nerve is an excellent source of tissue when the patient has severe peripheral neuropathy. A sural nerve biopsy should be performed only if the peripheral neuropathy is severe, because the procedure results in loss of local sensation and may be associated with protracted discomfort. Tissue obtained at carpal tunnel decompression should always be examined for amyloid. Endomyocardial biopsy is a useful procedure for patients with cardiac involvement. Biopsy specimens from the small intestine, skin, prostate, and gastric mucosa are all potential sources for diagnosis. A biopsy specimen of a salivary gland from the lip shows amyloid in the majority of patients.86 Skin biopsies often reveal amyloid deposits in the blood vessels of the subcutaneous

FIGURE 108–4 Results of bone marrow biopsy. (From Kyle, R. A., and Gertz, M. A.: Amyloidosis. In Wiernik, P. H., Canellos, G. P., Kyle, R. A., and Schiffer, C. A. [eds.]: Neoplastic Diseases of the Blood, 2nd ed. New York, Churchill Livingstone, p. 525, 1991, with permission.)

2437

Amyloidosis and Neuropathy

FIGURE 108–5 Results of biopsy from sites of involvement in patients with AL. (From Kyle, R. A.: Amyloidosis. In Hoffman, R., Benz, E. J. Jr., Shattil, S. J., et al. [eds.]: Hematology: Basic Principles and Practice. New York, Churchill Livingstone, p. 1038, 1991, with permission.)

tissue.97 Results from our experience with biopsy of various sites are shown in Figure 108–5. We have been disappointed with the results of technetium-99m pyrophosphate uptake by amyloid. We found that only 3 of 20 patients with biopsy-proved cardiac amyloid had uptake, but 15 of 20 control patients without heart disease also had cardiac uptake.63 We have also been disappointed with the specificity of increased uptake of gallium-67 citrate in patients with nephrotic syndrome resulting from amyloidosis. In our experience, uptake of gallium-67 citrate was found in many patients without AL.64 Examination of the urine for the presence of amyloid fibrils is misleading and is not recommended. The Congo red test (Paunz test) is no longer used because anaphylactic reactions may occur and it is insensitive.

Prognosis The median duration of survival of 474 patients with AL evaluated at Mayo Clinic within 1 month of diagnosis was 13.2 months. Only 7% survived for 5 or more years.127 Survival varied greatly among the various syndromes (Table 108–5 and Fig. 108–6). For example, the median duration of survival was 4 months for 80 patients presenting with congestive heart failure. The clinical groups

were arranged in a hierarchical fashion, and patients were placed only in the group in which the syndrome first occurred. Thus none of the 114 patients who presented with renal disease had either congestive heart failure or orthostatic hypotension, but some may have had peripheral neuropathy. The 40 patients with peripheral neuropathy had no evidence of congestive heart failure, orthostatic hypotension, or nephrotic syndrome at diagnosis and had a median survival of 26 months. The Doppler-derived index of diastolic myocardial function is important in assessing the prognosis of a patient with amyloidosis.202 The most important echocardiographic predictors of survival are the Doppler index of diastolic myocardial performance, ejection fraction, mitral deceleration time, and the ability of the ventricle to relax. We have found that the level of serum beta2-microglobulin is helpful in predicting survival, but this has not been confirmed by others. Death of patients presenting with only peripheral neuropathy or carpal tunnel syndrome (or both) is usually due to involvement of other organs, particularly the heart. Duston et al.50 noted a median survival of 35 months for patients with peripheral neuropathy, in contrast to 16 months for patients with AL without peripheral neuropathy.

Therapy Table 108–5. Survival in Amyloidosis (Hierarchical) Patients

Congestive heart failure Orthostatic hypotension Nephrotic Peripheral neuropathy Total group

No.

%

Years

Months

80 41 114 40 474

17 9 24 8

0.3 1.0 1.4 2.1 1.1

4 12 16 26 13

From Kyle, R. A., and Gertz, M. A.: Primary systemic amyloidosis: clinical and laboratory features in 474 cases. Semin. Hematol. 32:45, 1995, with permission.

Treatment for AL is unsatisfactory. The most important measurement of the efficacy of therapy is prolongation of survival. A prospective randomized trial is essential because earlier diagnosis and improvement in supportive measures make the use of historical controls unreliable. Because amyloid fibrils consist of the variable portion of monoclonal light chains, which are synthesized by plasma cells, it is reasonable to attempt therapy with alkylating agents known to have an effect on plasma cell proliferative processes such as multiple myeloma. In a placebocontrolled, double-blind study of 55 patients with AL, the patients receiving melphalan-prednisone therapy continued

2438

Diseases of the Peripheral Nervous System

FIGURE 108–6 Survival in AL, by syndrome, arranged hierarchically. (From Kyle R. A., and Gertz, M. A.: Primary systemic amyloidosis: clinical and laboratory features in 474 cases. Semin. Hematol. 32:45, 1995, with permission.)

with treatment longer and received larger doses than did patients in the placebo group before the code was broken because of progressive disease. Almost one half of patients with nephrotic syndrome showed objective improvement.130 The development of a myelodysplastic syndrome or acute nonlymphocytic leukemia associated with alkylating agent therapy has been recognized. We found that 10 of 153 patients with biopsy-proven AL who were treated with melphalan developed cytogenetic abnormalities consistent with alkylator-induced damage to hematopoietic cells. Acute nonlymphocytic leukemia was recognized in four, a dysmyelopoietic syndrome was noted in five, and one had a nondiagnostic bone marrow examination. Eight of the 10 patients died as a direct result of pancytopenia. The median survival from the onset of acute leukemia or dysmyelopoietic syndrome was 8.1 months. The actuarial risk in patients surviving 3.5 years was 21% despite the fact that only 6.5% of the total group had leukemia or dysmyelopoietic syndrome.68 Colchicine inhibits casein induction of amyloidosis in mice. In patients with familial Mediterranean fever, colchicine has relieved abdominal pain and prevented amyloidosis. The mechanism of action is unknown, but colchicine may block the formation of amyloid-enhancing factor.17 In one study, patients treated with colchicine had a median survival of 17 months, whereas a historical control group had a median survival of 6 months.32 In a prospective randomized study of melphalan (0.15 mg/kg daily for 7 days) plus prednisone (0.8 mg/kg daily for the same 7 days) every 6 weeks versus daily colchicine therapy, 101 patients were stratified according to their dominant clinical manifestation. The other regimen was added if the disease progressed. The median survival of those randomized to melphalan and prednisone was 25.2

months, and with colchicine it was 18 months, but the difference in survival was not statistically significant. When the survival of patients who received only one regimen was analyzed or when survival was analyzed from the time of entry into the study to the time of death or progression of disease, significant differences (P ⫽ .001 and P ⬍ .0001, respectively) favoring melphalan-prednisone therapy were evident. The study suggests that melphalan-prednisone therapy may be superior to colchicine therapy.131 In a prospective study, 220 patients with biopsy-proven AL were randomly assigned to treatment with (1) colchicine, melphalan, and prednisone; (2) melphalan and prednisone; or (3) colchicine alone. They were stratified according to their major clinical manifestations: renal disease (105 patients), cardiac involvement (46 patients), peripheral neuropathy (19 patients), or other (50 patients). The median duration of survival after randomization was 8.5 months for the colchicine group; 18 months for the melphalan and prednisone group; and 17 months for the melphalan, prednisone, and colchicine group (P ⬍ .001).128 In an effort to determine whether more intensive chemotherapy was beneficial, 101 AL patients were randomly assigned to receive a combination of agents (either vincristine, carmustine, melphalan, cyclophosphamide, and prednisone [VBMCP] or melphalan and prednisone). The combination of alkylating agents did not produce a higher response rate or longer survival than melphalan and prednisone.79 We have also used dimethyl sulfoxide, ␣-tocopherol (vitamin E),69 and interferon alfa-2,70 but none was beneficial. High-dose dexamethasone for untreated or previously treated patients had a response rate of approximately 15%.77,78 The combination of vincristine, doxorubicin, and dexamethasone has been reported to be of benefit in a small number of patients.216 An

Amyloidosis and Neuropathy

anthracycline analogue, 4⬘-iodo-4⬘-deoxydoxorubicin, which binds to amyloid fibrils, is reportedly beneficial for AL.81 However, we have been disappointed in its efficacy.75 In AL, the definition of a response can be either organ based or hematologic. An organ-based response is defined as a 50% reduction in urinary protein excretion in the absence of progressive renal insufficiency, a reduction in the size of the liver by at least 2 cm and a 50% decrease in serum alkaline phosphatase, a 2-mm decrease in the echocardiographic wall thickness, or an increase of 20% in the ejection fraction when it is initially less than 50%. A hematologic response is defined as a 50% decrease in serum and urine monoclonal protein. Autologous peripheral blood stem cell transplantation is available for patients with AL. Transplantation requires adequate renal, cardiac, and pulmonary function as well as a good performance status (0 or 1), good liver function, and adequate nutrition. However, most patients with AL present with marked cardiac, renal, pulmonary, or liver dysfunction. Comenzo et al.34 reported that, of 25 AL patients given an autologous stem cell transplant, 17 were alive at 24 months. Of these 17 patients, 11 had improved amyloidrelated organ involvement. Comenzo et al.34 also have reported neurologic improvement after peripheral stem cell transplantation in four of five patients. The median age of the patients was 48 years, and 90% had a performance status of 1 or 2. Eight patients had echocardiographic evidence of amyloid, but none had congestive heart failure. Two of three patients with neurologic involvement had resolution of peripheral and autonomic neuropathy. Three of the 25 patients died before the 100th posttransplantation day. Gertz et al.76 reported that 12 (60%) of 20 patients with AL who received an autologous stem cell transplant had a hematologic or organ response. Echocardiography demonstrated an interventricular septal thickness of 15 mm or more in six patients; five of the six died after transplantation. Thirteen patients were alive, with follow-up of 3 to 26 months. Severe gastrointestinal adverse effects occurred in five patients. The authors concluded that stem cell transplantation had a much higher morbidity and mortality than transplantation for multiple myeloma. The best results occurred in patients who had nephrotic syndrome as the only manifestation of the disease. Patients who receive transplants for AL are selected on the basis of age, performance status, number and type of organs involved, and the absence of major cardiac involvement. Dispenzieri et al.45 found that only 229 of the 1288 patients with symptomatic AL amyloidosis evaluated at Mayo Clinic from 1983 to 1997 fulfilled the eligibility criteria for an autologous stem cell transplant. The criteria included biopsy-proven amyloid, symptomatic AL, absence of a clinical diagnosis of multiple myeloma, age 70 years or younger, cardiac interventricular septal thickness

2439

of 15 mm or less, cardiac ejection fraction greater than 55%, serum creatinine level of 2 mg/dL or less, and a direct bilirubin concentration of 2 mg/dL or less. The median age of the 229 patients was 56 years. Of the patients, 100 had early cardiac involvement, 41 had liver involvement, 167 had renal disease, and 39 had peripheral neuropathy. The 229 patients were treated with chemotherapy. Median survival was 42 months, with 5- and 10-year survival rates of 36% and 15%, respectively. Thus selection of patients for stem cell transplantation is important. A randomized trial is needed to assess the benefit of stem cell transplantation.45 General Therapeutic Measures Treatment of peripheral neuropathy is ineffective. In our experience, the neurologic symptoms and findings continue to worsen. Dysesthesias tend to disappear as the sensory involvement worsens. Analgesics, sedatives, and membrane-stabilizing agents may be helpful, depending on the kind and severity of symptoms. Tricyclic antidepressants such as amitriptyline in doses of 75 to 100 mg at bedtime can be helpful for nocturnal pain interfering with sleep. However, they may worsen orthostatic symptoms and bowel and bladder function. Despite the chronic and “benign” nature of amyloid neuropathy, the dysesthesias may be sufficiently distressing to require narcotics for control of symptoms. Gabapentin and topiramate in high doses can lessen pain in these patients, but the dose must be titrated slowly. Low doses of narcotics, such as codeine, at bedtime can help patients sleep. The risk of habituation and addiction in these patients is extremely low. However, narcotics can make autonomic symptoms worse, so they must be used carefully. Decompression of the carpal ligament relieves the pain of carpal tunnel syndrome, but residual sensory loss is frequent. Pain caused by periarticular deposition of amyloid in the shoulders, producing the “shoulder-pad” syndrome, has usually been unresponsive to injections of corticosteroids. Treatment of orthostatic hypotension is unsatisfactory, but elastic support extending to the waist may be of help. Patients should be instructed to rise slowly and to sit on the edge of the bed for a few minutes before walking. Fludrocortisone is often associated with increased fluid retention, and this limits its effectiveness. Investigative agents such as midodrine or L-threo-3,4-dihydroxyphenylserine are sometimes helpful (see Chapter 44). Octreotide may help in the management of both orthostatic hypotension and autonomic-induced diarrhea. In patients with syncope, it is always important to consider a cardiac cause. Sudden death is not rare and presumably is due to ventricular arrhythmias or heart block. Ambulatory electrocardiographic monitoring is necessary for patients with syncope because placement of a pacemaker may be life saving. Frequent ventricular arrhythmias occur in patients with cardiac amyloidosis when

2440

Diseases of the Peripheral Nervous System

monitored by ambulatory echocardiography. Symptoms are frequently absent, but these patients should be managed with antiarrhythmic agents. Nephrotic syndrome is best managed with salt restriction and diuretic agents as needed. Chronic renal dialysis is necessary when symptomatic azotemia develops. Kidney transplantation has been performed in several instances, and the results are promising.160 Amyloid has been detected in the transplanted kidney.94 Congestive heart failure should be managed with salt restriction and judicious administration of diuretics such as furosemide. Digitalis must be used with care because many patients with AL are unusually sensitive to the drug, and arrhythmias and heart block are common.21

FAMILIAL AMYLOIDOSIS Familial or inherited amyloidosis has been reviewed by Buxbaum and Tagoe.26 Autosomal dominant inheritance is found in all forms except amyloid associated with familial Mediterranean fever (autosomal recessive). Familial amyloidosis can be classified clinically as neuropathic, nephropathic, cardiopathic, or miscellaneous. Hereditary cerebral hemorrhage with amyloidosis has been recognized in Iceland155 and the Netherlands16 but is not discussed further here because peripheral neuropathy is not a prominent feature. Fibrinogen A ␣-chain mutation,88,207 lysozyme mutation,82 or apolipoprotein II mutation13 produces familial renal amyloidoses, but peripheral neuropathy is not a feature. The discussion in this chapter is limited to the types of amyloidosis associated with neuropathy (Table 108–6).

Familial Amyloid Polyneuropathy In 1929, DeBruyn and Stern44 described a 52-year-old man who had pain and numbness of his extremities for 3 years. Two brothers and a sister had died of a similar

Table 108–6. Amyloid and Neuropathy Amyloid

Classification (Clinical)

AH ATTR A Apo Al * A Gel * A * A Fib * A Cyst * A␤

Amyloid, heavy chain Renal/neuropathic Finnish Kidney

Major Protein Component Ig heavy chain Transthyretin Apolipoprotein A-I Gelsolin Lysozyme Fibrinogen ␣ chain Cystatin C

Alzheimer

*Not discussed in this chapter because they are not associated with peripheral neuropathy.

illness. At autopsy, microscopic examination showed a nonnucleated, homogeneous substance in the peripheral nerves that stained with aniline blue-orange G. The findings are consistent with familial amyloidosis. In 1952, Andrade6 described members of a Portuguese family from Porto who lost the sensation of pain and temperature in the lower extremities and then experienced paresthesias and loss of touch and vibratory sensation. Muscular weakness developed later. Impotence and disturbances of bladder and gastrointestinal function manifested by incontinence and diarrhea were severe and indicated autonomic dysfunction. Vitreous opacities were noted, and ulceration of the feet was frequent. Clinical involvement of the heart and kidney was uncommon. More than 1000 patients have been reported from the region of Porto, Portugal. Dyck and Lambert52 explored the nature of the dissociated sensory loss in two Greek patients and one Yugoslavian-American patient with familial amyloid polyneuropathy (FAP). In one patient, painless injury of the foot initiated the search for the cause of the neuropathy. On quantitative sensation assessment, nociception and cold and warm sensation were severely abnormal in the feet and legs, whereas touch-pressure and vibration sensation were relatively spared (see Chapter 42). Recording the compound action potential of the sural nerve in vitro, we found that the latency of the A␣ potential (large-diameter myelinated fiber) was not prolonged but the amplitude of the potential was reduced. By comparison, the A␦ potential (small-diameter myelinated fiber) and the C potential (unmyelinated fibers) were absent (see Chapter 39). Diameter histograms of myelinated and unmyelinated fibers confirmed that most unmyelinated and small myelinated fibers had degenerated (see Chapter 39). The only remains of Remak fibers (cluster of unmyelinated fibers) were the cluster of Schwann cell cytoplasmic processes. Large myelinated fibers (A␣, ␤) were also decreased in number or were undergoing degeneration. In addition to demonstrating the dissociated sensory loss that may occur in FAP, the study was important in confirming that various-sized classes of sensory fibers convey different sensory modalities. At a later stage the sensory loss is more dissociated, and sensory, motor, and autonomic deficits are found in both lower and upper limbs as well as proximal and distal segments. Symptoms usually begin in the third decade of life, and death occurs within 10 years. Late onset may occur in some patients, with symptoms not apparent until the sixth or seventh decade. However, patients with late onset have the same biochemical abnormality of transthyretin (prealbumin) (methionine substituted for valine at position 30 [ATTR Met 30]). The factors responsible for the delayed onset are unknown.181,182 Cranial neuropathy may also be a feature. The cranial autonomic nerves are affected first and produce pupillary

Amyloidosis and Neuropathy

changes and a decrease in salivary secretion during the first 3 years of the disease. Fasciculations of the tongue occur, followed by facial fasciculation and dysphonia and dysphagia. Sensory changes consisting of absence of the corneal reflexes and facial sensory loss do not occur until after 8 to 9 years.179 Electrocardiographic abnormalities were found in 87% of a series of 327 patients with FAP. The most common findings were low-voltage and conduction disturbances. In contrast to AL, echocardiography revealed no increased thickness of the septum or posterior ventricular wall.42 Andersson4 reported FAP in northern Sweden that was clinically similar to the Portuguese type except that the average age at onset was 55 years and malabsorption and peptic ulcer were more frequent. The greatest concentration of involved families is in Skelleftea near Umea in northern Sweden. Approximately 250 cases have been recognized. Families have been traced to the 16th century through church records.205 Although renal insufficiency is rare, glomerular clearance and concentrating capacity are frequently decreased.195 Sinus node dysfunction, atrioventricular conduction abnormalities, and supraventricular and ventricular arrhythmias are common in Swedish patients.55 Severe bradycardia or heart block may occur during anesthesia.54 Increased thickness of the interventricular septum was noted in 20 of 22 patients (91%).9 However, myocardial function was only mildly impaired and congestive heart failure was rare.156 Seven generations of a Swedish family in America presented with peripheral neuropathy in the fourth and fifth decades of life.11 Abnormalities of the autonomic nervous system were common. This family differed from Andersson’s4 group in that an unusual infiltration of amyloid was found in the meninges and renal insufficiency was a common cause of death. In 1981, residents of Kumamoto and Nagano prefectures in Japan were reported with FAP.201 The clinical features are similar and the biochemical abnormality is identical to that in the Portuguese family members. The disease usually begins between 25 and 35 years of age, but late onset may also occur. Sensorimotor peripheral neuropathy developing in the legs is the major feature. Autonomic dysfunction is prominent. Nausea, vomiting, constipation, diarrhea, orthostatic hypotension, urinary and fecal incontinence, impotence, and trophic changes of the skin are the major symptoms. Involvement of the heart producing atrioventricular block and renal involvement are uncommon. However, electrocardiographic abnormalities were found in 93% of one group of 14 patients with AF. Clinically overt heart disease was noted in only three. Left ventricular relaxation was impaired in early diastole and was seen before the development of clinically identifiable heart disease.122 Vitreous opacities and scalloped pupillary abnormalities are rare.

2441

High concentrations of variant transthyretin (albumin [Met 30]) are found in the choroid plexus of the brain as well as in the liver.148 Other investigators have reported variant transthyretin in the cerebrospinal fluid. 149 Kinetics analysis of the variant transthyretin is the same as that of normal transthyretin. This suggests that factors other than the rate of transfer of the variant transthyretin from plasma to an extravascular compartment have a critical role in the deposition of amyloid in these patients.5 Unique features in addition to peripheral and autonomic neuropathy have been found in some kinships. Cerebellar ataxia and pyramidal tract signs were present in one family. However, the possibility of a combination of AF and an unrelated hereditary disorder of the central nervous system cannot be excluded.59 The same mutation (TTR Met 30) may be found in patients with cerebellar ataxia, pyramidal tract signs, and polyneuropathy without a family history as in those with typical polyneuropathy.90 In another family, members presented with amyloid heart disease. Prominent nephropathy and proteinuria were noted in another kinship, and in still another symptoms began in the sixth decade.98 The typical variant transthyretin has been found in patients with typical features of FAP but no family history of the disorder. An asymptomatic parent may also have the variant transthyretin. It is obvious that apparent sporadic cases of amyloid neuropathy may actually be unrecognized AF.151 In a study that used recombinant DNA techniques and Southern blot hybridization in 49 patients from Japan, 5 patients who had clinical features of neuropathy but no family history, 5 patients without symptoms but who were members of a polyneuropathy family, and 2 patients with spinocerebellar degeneration and spastic paraplegia all had the same TTR Met 30 mutation.89 Extensive endoneurial deposition of amyloid and loss of myelinated nerve fibers in the vagus nerve and celiac ganglion were reported in patients with FAP.99 Abnormalities of the lower esophageal sphincter and simultaneous contractions of decreased amplitude in the body of the esophagus have been noted in AF. 23 Examination of abdominal fat aspirate may be a helpful procedure in the diagnosis of FAP. In one study, all 14 patients had positive aspirates but only half had earlystage disease.145 Gastrointestinal involvement is common in FAP. In one series, 42 of 52 patients had gastrointestinal symptoms.196 Evidence of peripheral neuropathy usually preceded gastrointestinal symptoms. Steatorrhea was found in more than one half of patients. It appeared that disruption of the autonomic nervous system rather than deposition of amyloid in the intestinal wall was responsible for the gastrointestinal symptoms. Adrenal hypofunction has also been recognized.157

2442

Diseases of the Peripheral Nervous System

Pathogenesis The amyloid fibrils in FAP of the Portuguese, Japanese, and Swedish types all consist of transthyretin (prealbumin).39 Methionine is substituted for valine at position 30 (TTR Met 30) in this variant transthyretin.180 More than 80 transthyretin mutations have been reported.37 Transthyretin is a transport molecule for thyroxin and retinol. It is a tetramer consisting of four identical noncovalently associated subunits. Each subunit consists of 127 amino acid residues and has a molecular weight of 14,000. Conformational changes resulting from amino acid substitutions are thought to destabilize the structure, causing aggregation and formation of amyloid fibrils. The structural changes of transthyretin and the formation of amyloid have been reviewed by Kelly et al.114 The gene coding for transthyretin is on human chromosome 18q11.2 : 18q12.1.192 Restriction fragment length polymorphism analysis may be used for diagnosis. DNA is extracted from whole blood and digested with a restriction enzyme. Electrophoresis in agarose is performed and the gel is subjected to Southern transfer. Autoradiography is performed after hybridization with a DNA probe labeled with 32P-deoxycitidine triphosphate14; it permits recognition of asymptomatic cases and children of affected individuals. It is of great interest to find the same mutation (TTR Met 30) in Portugal, Sweden, and Japan. The gene may have been transmitted by Vikings from Sweden who visited Western Europe beginning in the eighth century. The greatest prevalence in Portugal is in the region of Povoa de Varzim and Vila do Conde on the coast of Portugal near Porto and is consistent with this possibility. More than 1000 patients with FAP have been recognized in Portugal. Alternatively, a mutation could have arisen in Portugal and then been carried to Sweden by Portuguese traders who brought salt essential for the preservation of fish and game. Portuguese traders may also have introduced the gene into Japan in the 16th century. Coincidentally, St. Francis Xavier preached in Kagoshima (Kyushu) in 1549.205,224 This story is part of folklore and is entertaining, but it is not supported by fact. Haplotype analysis strongly suggests that multiple independent mutations occurred at hot spots.171,225 TTR Met 30 in Other Countries Familial amyloid polyneuropathy has been reported from many parts of the world. Ten of 13 European families with amyloid polyneuropathy carried the TTR Met 30 mutation.96 The familial features are frequently not recognized. For example, index cases from only three of nine Cypriot and Greek families were aware of similarly affected relatives. Amyloid polyneuropathy has been reported in families from Greece,183 Italy (Sicily),57,184 Brazil,12 and Hong Kong.29 In addition, two unrelated German kinships and one sporadic case of transthyretin-associated AF polyneuropathy were reported.58 Another American family of

German ancestry,154 a family from Herakleion (Crete),134 and seven similar cases from seven different families in northwest Ireland194 have been described. Because the Irish patients all lived near the sea in the district of Donegal in the northwest area of Ireland, it is interesting to speculate on the role of survivors from the Spanish Armada in these cases. The TTR Met 30 mutation was found in an American family of English origin who presented with proximal arm weakness before the development of sensorimotor peripheral neuropathy. Both patients died of cardiac failure, and postmortem examination revealed extensive cardiac amyloid. Other atypical features included onset after age 65 years.116 TTR Tyr 77 A large kinship of German-English ancestry residing in Texas with an autosomal dominant FAP was reported.139 The family immigrated to the United States early in the 19th century and settled originally in Arkansas. Onset of symptoms occurred in the sixth or seventh decade. Lower and upper extremity sensorimotor peripheral neuropathy was the major feature. Hypertrophic cardiomyopathy and orthostatic hypotension were seen. None of the patients had vitreous opacities or scalloping of the pupils. Affected members had a variant TTR with a tyrosine-for-serine substitution at position 77. Remarkably, DNA from 9 of 10 children of the four affected siblings did not have the mutant TTR gene.187 A patient of German descent from Illinois presented with peripheral neuropathy of the lower limbs, chronic diarrhea, and periorbital purpura. Several other family members had similar symptoms but were not available for study. This patient had the same mutation (TTR Tyr 77) as in the large Texas kindred.215 TTR Ser 84 (Indiana, Rukavina) Rukavina et al.175,176 described a family of Swiss origin residing in Indiana who had AF (type II). The syndrome was characterized by the development of carpal tunnel syndrome, which began in the fifth decade of life. Sensorimotor peripheral neuropathy often involving the lower extremities developed later in the course of the disease. Autonomic and central nervous system involvement was not seen. Vitreous opacities and impotence may occur. Nephropathy is not a feature of the disease, and cardiac involvement is not prominent. Occasionally, inflammatory arthritis and joint erosions occur.56 Spinal claudication resulting from amyloid deposition in the ligamentum flavum has been reported.91 Many patients survive into the sixth and seventh decades. These patients have a variant of transthyretin in which there is a serine-for-isoleucine substitution at amino acid residue 84. This abnormality has been verified at the DNA level and permits a direct test for the abnormality.213 Six carriers of the variant gene (TTR Ser 84) had no amyloid in skin, rectal, or carpal tunnel

2443

Amyloidosis and Neuropathy

biopsy specimens. The authors suggested that accumulation of amyloid did not begin until a few years before symptoms appeared.91 TTR Ala 60 (Appalachian) An Ala-60 mutation of transthyretin has been identified in patients with familial amyloidosis from the Appalachian region of the United States.15,214 Although the major manifestation is cardiac, many of these patients have a sensorimotor peripheral neuropathy or carpal tunnel syndrome. Other TTR Substitutions Mahloudji et al.144 reported 11 families with a similar clinical picture residing in Maryland and traceable to a German couple. A mutation of transthyretin consisting of histidine for leucine at position 58 has been identified. Neuropathy of the lower extremities, reduction of vision by vitreous opacities, and impotence occurred in an Ashkenazi Jew and his son.60 Initially, amyloid fibrils were thought to consist of TTR with a glycine-for-threonine substitution at position 49.163 However, the abnormality is actually a substitution of isoleucine for phenylalanine at position 33.150 Peripheral and autonomic neuropathy, vitreous opacities, cardiomyopathy, and thyromegaly were reported in a kindred of western European origin living in West Virginia. The transthyretin contained a valine-for-leucine substitution at position 55. A large kindred in Iowa with neuropathy involving all four extremities and renal involvement was described.209 The neuropathy began in the fourth decade of life. The average survival was 12 years after diagnosis, with death often resulting from renal insufficiency. Vitreous amyloid deposits and cardiomyopathy occurred in some cases. Amyloid fibrils consist of apolipoprotein A-I in which there is an arginine-for-glycine substitution at position 26.152 Meretoja147 reported on familial systemic amyloidosis that was characterized by corneal lattice dystrophy, progressive cranial neuropathy, skin changes, and various systemic symptoms. Widespread spinal, cerebral, and meningeal amyloid angiopathy has been reported.119 Although the majority of patients have been reported from Finland, patients from Denmark, the Netherlands, Japan, and the United States have been recognized.118 A mutant gelsolin gene producing a protein with asparagine in place of aspartic acid at position 187 has been found. It is thought that symptoms are due to the accumulation of amyloid derived from plasma gelsolin in the tissues.107 The amyloid fibrils consist of gelsolin with a G654A or G654T gelsolin gene mutation.

Diagnosis A positive family history of amyloidosis is helpful in the diagnosis, but in our experience approximately one half of

patients with AF have a negative family history. Almost one third of the patients with AF evaluated at our institution initially had been thought to have AL and some had inappropriately been given chemotherapy.74 Currently, only 2% of cases of systemic amyloidosis seen at Mayo Clinic are familial. Diagnosis depends on the demonstration of amyloid in tissue. Sural nerve biopsy or subcutaneous fat aspirates are suitable for diagnosis. Immunohistochemical staining shows transthyretin in the majority of specimens, but some contain apolipoprotein A-I or gelsolin. Screening for a transthyretin variant using polyacrylamide gel electrophoresis, followed by isoelectric focusing, under partially denaturing conditions detected 74 of 77 variants (96%), with a specificity of 100%. No false-positive results were obtained.36 Restriction fragment length polymorphism analysis may also be used for diagnosis as noted above. In a study of 38 nerve biopsy specimens, Li et al.138 performed immunohistochemical staining for amyloid protein. In nine, the amyloid was not sufficiently large for the responsible protein component to be determined reliably. In 18, ␬ or ␭ light chains were identified, and in 11, TTR positivity was identified. The TTR-positive patients were light-chain negative. A family history of amyloidosis was found in 8 of 11 patients, and the remaining 3 had relatives who had died early or represented single sibships, so that inheritance could have been missed. Monoclonal proteins were not detected in any of the 11 patients. Among the 11 TTR-positive cases, initial symptoms related to sensory fiber dysfunction were the first evidence of involvement in 6. Symptoms included decreased or absent cutaneous sensation, paresthesias, lancinating pain, and cramping of the hands or feet. Muscle weakness was the initial symptom in two patients. In another patient, sensory symptoms and postural hypotension were the initial findings. In still another, sensory symptoms and diarrhea initiated the disorder. In a final patient, frequency of urination and symptoms related to renal failure were found initially. Later, sensory, motor, and autonomic dysfunction and organ involvement were typical. In an earlier section, we compared the severity of the neuropathy, the type and location of amyloid deposition, and the severity of fiber degeneration in primary amyloidosis and TTR-positive amyloidosis.

Pathology Dyck and Lambert52 showed relatively selective loss of unmyelinated and small myelinated fibers of the sural nerve. Teased fiber studies revealed axons degenerating into linear rows of myelin ovoids. The largest and most numerous deposits of amyloid were within fascicles. In some cases, nodules of amyloid indented myelinated fibers, producing large sacculations of the fiber on both sides of the compression. In another report of teased fiber

2444

Diseases of the Peripheral Nervous System

studies in AF, distortion of the myelin sheaths, segmental demyelination, and wallerian degeneration adjacent to amyloid deposits were seen.178

Therapy Liver transplantation was introduced as a treatment for AF with transthyretin mutation a decade ago.95 It is a reasonable intervention because transthyretin is synthesized in the liver. In contrast, liver transplantation is not justified for apolipoprotein A-I amyloid because this is synthesized in both the intestine and the liver. Also, liver transplantation has no role in the treatment of amyloidosis caused by gelsolin or cystatin C. Of the first 20 patients in Sweden who had liver transplantation for AF caused by transthyretin mutation, 70% were alive 10 to 52 months after surgery. Ambulation, nutritional status, and gastrointestinal and urinary symptoms were improved.200 Amyloid deposits in the kidneys, adrenal glands, and spleen were reduced in five of eight patients who were evaluated with 123I-labeled SAP after liver transplantation.177 In another report, 5 of 11 patients had increased left ventricular thickening after liver transplantation. Congestive heart failure occurred in one patient and another had sudden cardiac death.48 Amyloid deposition in the conduction system of the heart was responsible for the death of 3 of 13 patients who had liver transplantation for AF.162 Gastric retention and malabsorption were adverse prognostic features in a group of patients with AF who had a liver transplant. Survival was improved if the patient’s nutritional state was good at the time of the transplantation.198 Quality of life improved in 10 of 12 patients who had a liver transplant.104 Adams et al.2 reported a series of 45 patients with symptomatic AF (43 had TTR Met 30). Motor nerve involvement stabilized in 7 of 11 patients with mild sensory motor neuropathy and in 2 of 8 with severe sensory motor neuropathy. It was noted that patients with urinary incontinence had a poor outcome.2 In a review of 35 liver transplants for TTR Met 30, the mortality rate was 14% in the first 6 months. The neurophysiologic score progressed in the first year following transplantation but did not increase significantly thereafter.41 In a review of liver transplantation, Suhr et al.199 concluded that the procedure was of proven value only for patients with TTR Met 30. In an editorial, Lewis136 concluded that liver transplantation was effective therapy for familial transthyretin-associated amyloidosis. Left lobar grafts from family members were performed in six patients with AF and were successful.111 A domino split-liver transplant was performed from a living donor.100 Flufenamic acid inhibits the conformational changes of transthyretin associated with amyloid fibril formation and may be useful.161 Only a small subset of the inhibitors

saturate the TTR binding sites in plasma.166 Inhibitors of amyloid formation are promising avenues of research.120,168 An ion exchange resin made of porous beads of polyvinyl alcohol gel covalently bound with dimethylaminoethanol decreased the concentration of both normal and variant transthyretin to about half of the preabsorption levels.204 It remains to be determined whether this selective transthyretin absorption column will be beneficial. Orthostatic hypotension is a major problem in AF. In eight of nine patients, L-threo-3,4-dihydroxyphenylserine, a synthetic precursor of norepinephrine, eliminated symptoms of orthostatic hypotension. The side effects included tachycardia and nausea.219 Carvalho et al.27 reported improvement in all 20 patients with orthostatic hypotension who were given L-threo-3,4-dihydroxyphenylserine. Double-blind, randomized, placebo-controlled studies with a larger number of patients and longer follow-up are necessary. Two multicenter European trials are underway.43

ACKNOWLEDGMENT This work was supported in part by Research Grant CA-62242 from the National Cancer Institute.

REFERENCES 1. Abarbanel, J. M., Frisher, S., and Osimani, A.: Primary amyloidosis with peripheral neuropathy and signs of motor neuron disease. Neurology 36:1125, 1986. 2. Adams, D., Samuel, D., Goulon-Goeau, C., et al.: The course and prognostic factors of familial amyloid polyneuropathy after liver transplantation. Brain 123:1495, 2000. 3. Al Suwaidi, J., Velianou, J. L., Gertz, M. A., et al.: Systemic amyloidosis presenting with angina pectoris. Ann. Intern. Med. 131:838, 1999. 4. Andersson, R.: Familial amyloidosis with polyneuropathy: a clinical study based on patients living in northern Sweden. Acta Med. Scand. Suppl. 590:1, 1976. 5. Ando, Y., Ikegawa, S., Miyazaki, A., et al.: Role of variant prealbumin in the pathogenesis of familial amyloidotic polyneuropathy: fate of normal and variant prealbumin in the circulation. Arch. Biochem. Biophys. 274:87, 1989. 6. Andrade, C.: A peculiar form of peripheral neuropathy: familial atypical generalized amyloidosis with special involvement of the peripheral nerves. Brain 75:408, 1952. 7. Araoz, P. A., Batts, K. P., and MacCarty, R. L.: Amyloidosis of the alimentary canal: radiologic-pathologic correlation of CT findings. Abdom. Imaging 25:38, 2000. 8. Aterman, K.: A historical note on the iodine-sulphuric acid reaction of amyloid. Histochemistry 49:131, 1976. 9. Backman, C., and Olofsson, B. O.: Echocardiographic features in familial amyloidosis with polyneuropathy. Acta Med. Scand. 214:273, 1983.

Amyloidosis and Neuropathy 10. Battle, W. M., Rubin, M. R., Cohen, S., and Snape, W. J. Jr.: Gastrointestinal-motility dysfunction in amyloidosis. N. Engl. J. Med. 301:24, 1979. 11. Benson, M. D., and Cohen, A. S.: Generalized amyloid in a family of Swedish origin: a study of 426 family members in seven generations of a new kinship with neuropathy, nephropathy, and central nervous system involvement. Ann. Intern. Med. 86:419, 1977. 12. Benson, M. D., Dwulet, F. E., Scheinberg, M. A., and Greipp, P.: Chemical classification of hereditary amyloidosis in Brazilian families and identification of gene carriers. J. Rheumatol. 13:927, 1986. 13. Benson, M. D., Liepnieks, J. J., Yazaki, M., et al.: A new human hereditary amyloidosis: the result of a stop-codon mutation in the apolipoprotein AII gene. Genomics 72:272, 2001. 14. Benson, M. D., and Wallace, M. R.: Genetic amyloidosis: recent advances. Adv. Nephrol. Necker Hosp. 18:129, 1989. 15. Benson, M. D., Wallace, M. R., Tejada, E., et al.: Hereditary amyloidosis. Description of a new American kindred with late onset cardiomyopathy: Appalachian amyloid. Arthritis Rheum. 30:195, 1987. 16. Bornebroek, M., Haan, J., and Roos, R. A.: Hereditary cerebral hemorrhage with amyloidosis–Dutch type (HCHWA-D): a review of the variety in phenotypic expression. Amyloid 6:215, 1999. 17. Brandwein, S. R., Sipe, J. D., Skinner, M., and Cohen, A. S.: Effect of colchicine on experimental amyloidosis in two CBA/J mouse models: chronic inflammatory stimulation and administration of amyloid-enhancing factor during acute inflammation. Lab. Invest. 52:319, 1985. 18. Breathnach, S. M., and Black, M. M.: Systemic amyloidosis and the skin: a review with special emphasis on clinical features and therapy. Clin. Exp. Dermatol. 4:517, 1979. 19. Bruni, J., Bilbao, J. M., and Pritzker, P. H.: Myopathy associated with amyloid angiopathy. Can. J. Neurol. Sci. 4:77, 1977. 20. Budd, G.: On Diseases of the Liver, 2nd ed. Philadelphia, Blanchard and Lea, 1853. 21. Buja, L. M., Khoi, N. B., and Roberts, W. C.: Clinically significant cardiac amyloidosis: clinicopathologic findings in 15 patients. Am. J. Cardiol. 26:394, 1970. 22. Bujanda, L., Beguiristain, A., Alberdi, F., et al.: Spontaneous rupture of the liver in amyloidosis. Am. J. Gastroenterol. 92:1385, 1997. 23. Burakoff, R., Rubinow, A., and Cohen, A. S.: Esophageal manometry in familial amyloid polyneuropathy. Am. J. Med. 79:85, 1985. 24. Buxbaum, J.: Aberrant immunoglobulin synthesis in light chain amyloidosis: free light chain and light chain fragment production by human bone marrow cells in short-term tissue culture. J. Clin. Invest. 78:798, 1986. 25. Buxbaum, J. N., Chuba, J. V., Hellman, G. C., et al.: Monoclonal immunoglobulin deposition disease: light chain and light and heavy chain deposition diseases and their relation to light chain amyloidosis: clinical features, immunopathology, and molecular analysis. Ann. Intern. Med. 112:455, 1990. 26. Buxbaum, J. N., and Tagoe, C. E.: The genetics of the amyloidoses. Annu. Rev. Med. 51:543, 2000.

2445

27. Carvalho, M. J., van den Meiracker, A. H., Boomsma, F., et al.: Improved orthostatic tolerance in familial amyloidotic polyneuropathy with unnatural noradrenaline precursor L-threo-3,4-dihydroxyphenylserine. J. Auton. Nerv. Syst. 62:63, 1997. 28. Chamarthi, B., Dubrey, S. W., Cha, K., et al.: Features and prognosis of exertional syncope in light-chain associated AL cardiac amyloidosis. Am. J. Cardiol. 80:1242, 1997. 29. Chang, C. M., Yu, Y. L., Wong, M., et al.: Type I familial amyloid polyneuropathy in a Chinese family. Acta Neurol. Scand. 79:391, 1989. 30. Clesham, G. J., Vigushin, D. M., Hawkins, P. N., et al.: Echocardiographic assessment of cardiac involvement in systemic AL amyloidosis in relation to whole body amyloid load measured by serum amyloid P component (SAP) clearance. Am. J. Cardiol. 80:1104, 1997. 31. Cohen, A. S., and Calkins, E.: Electron microscopic observations on a fibrous component in amyloid of diverse origins. Nature 183:1202, 1959. 32. Cohen, A. S., Rubinow, A., Anderson, J. J., et al.: Survival of patients with primary (AL) amyloidosis: colchicine-treated cases from 1976 to 1983 compared with cases seen in previous years (1961 to 1973). Am. J. Med. 82:1182, 1987. 33. Cohen, A. S., and Shirahama, T.: Electron microscopic analysis of isolated amyloid fibrils from patients with primary, secondary and myeloma-associated disease: a study utilizing shadowing and negative staining techniques. Isr. J. Med. Sci. 9:849, 1973. 34. Comenzo, R. L., Vosburgh, E., Falk, R. H., et al.: Doseintensive melphalan with blood stem-cell support for the treatment of AL (amyloid light-chain) amyloidosis: survival and responses in 25 patients. Blood 91:3662, 1998. 35. Comesana, L., del Castillo, M., Martin, R., et al.: Musculoskeletal amyloid disease: MRI features. Ann. Radiol. (Paris) 38:150, 1995. 36. Connors, L. H., Ericsson, T., Skare, J., et al.: A simple screening test for variant transthyretins associated with familial transthyretin amyloidosis using isoelectric focusing. Biochim. Biophys. Acta 1407:185, 1998. 37. Connors, L. H., Richardson, A. M., Theberge, R., and Costello, C. E.: Tabulation of transthyretin (TTR) variants as of 1/1/2000. Amyloid 7:54, 2000. 38. Coria, F., Castano, E., Prelli, F., et al.: Isolation and characterization of amyloid P component from Alzheimer’s disease and other types of cerebral amyloidosis. Lab. Invest. 58:454, 1988. 39. Costa, P. P., Figueira, A. S., and Bravo, F. R.: Amyloid fibril protein related to prealbumin in familial amyloidotic polyneuropathy. Proc. Natl. Acad. Sci. U. S. A. 75:4499, 1978. 40. Cueto-Garcia, L., Reeder, G. S., Kyle, R. A., et al.: Echocardiographic findings in systemic amyloidosis: spectrum of cardiac involvement and relation to survival. J. Am. Coll. Cardiol. 6:737, 1985. 41. de Carvalho, M., Conceicao, I., Bentes, C., and Luis, M. L.: Long-term quantitative evaluation of liver transplantation in familial amyloid polyneuropathy (Portuguese V30M). Amyloid 9:126, 2002. 42. de Freitas, A. F.: The heart in Portuguese amyloidosis. Postgrad. Med. J. 62:601, 1986.

2446

Diseases of the Peripheral Nervous System

43. de Freitas, F.: Treatment of orthostatic hypotension with L-threo-DOPS in familial amyloidotic polyneuropathy [editorial]. Amyloid 3:212, 1996. 44. DeBruyn, R. S., and Stern, R. O.: Case of the progressive hypertrophic polyneuritis of Dejerine and Sottas, with pathological examination. Brain 52:84, 1929. 45. Dispenzieri, A., Lacy, M. Q., Kyle, R. A., et al.: Eligibility for hematopoietic stem-cell transplantation for primary systemic amyloidosis is a favorable prognostic factor for survival. J. Clin. Oncol. 19:3350, 2001. 46. Divry, P., and Florkin, M.: Sur les propriétés optiques de l’amyloïde. C. R. Soc. Biol. (Paris) 97:1808, 1928. 47. Doyle, L.: Lardaceous disease: some early reports by British authors (1722–1879). J. R. Soc. Med. 81:729,1988. 48. Dubrey, S. W., Davidoff, R., Skinner, M., et al.: Progression of ventricular wall thickening after liver transplantation for familial amyloidosis. Transplantation 64:74, 1997. 49. Durie, B. G., Persky, B., Soehnlen, B. J., et al.: Amyloid production in human myeloma stem-cell culture, with morphologic evidence of amyloid secretion by associated macrophages. N. Engl. J. Med. 307:1689, 1982. 50. Duston, M. A., Skinner, M., Anderson, J., and Cohen, A. S.: Peripheral neuropathy as an early marker of AL amyloidosis. Arch. Intern. Med. 149:358, 1989. 51. Dyck, P. J., Karnes, J., O’Brien, P. C., and Swanson, C. J.: Neuropathy symptom profile in health, motor neuron disease, diabetic neuropathy, and amyloidosis. Neurology 36:1300, 1986. 52. Dyck, P. J., and Lambert, E. H.: Dissociated sensation in amyloidosis: compound action potential, quantitative histologic and teased-fiber, and electron microscopic studies of sural nerve biopsies. Arch. Neurol. 20:490, 1969. 53. Elghetany, M. T., and Saleem, A.: Methods for staining amyloid in tissues: a review. Stain Technol. 63:201, 1988. 54. Eriksson, P., Boman, K., Jacobsson, B., and Olofsson, B. O.: Cardiac arrhythmias in familial amyloid polyneuropathy during anaesthesia. Acta Anaesthesiol. Scand. 30:317, 1986. 55. Eriksson, P., Karp, K., Bjerle, P., and Olofsson, B. O.: Disturbances of cardiac rhythm and conduction in familial amyloidosis with polyneuropathy. Br. Heart J. 51:658, 1984. 56. Eyanson, S., and Benson, M. D.: Erosive arthritis in hereditary amyloidosis. Arthritis Rheum. 26:1145, 1983. 57. Ferlini, A., Romeo, G., Tassinari, C. A., et al.: Discrimination of peripheral polyneuropathies caused by TTR variant or diabetes in the same pedigree through protein studies. Adv. Neurol. 48:201, 1988. 58. Feurle, G. E., Linke, R. P., Kuhn, E., and Wagner, A.: Clinical value of immunohistochemistry with AF-antibody in the diagnosis of familial amyloid neuropathy. J. Neurol. 231:237, 1984. 59. Furuya, H., Yoshioka, K., Sasaki, H., et al.: Molecular analysis of a variant type of familial amyloidotic polyneuropathy showing cerebellar ataxia and pyramidal tract signs. J. Clin. Invest. 80:1706, 1987. 60. Gafni, J., Fischel, B., Reif, R., et al.: Amyloidotic polyneuropathy in a Jewish family: evidence for the genetic heterogeneity of the lower limb familial amyloidotic neuropathies. Q. J. Med. 55:33, 1985.

61. Gamba, G., Montani, N., Anesi, E., et al.: Clotting alterations in primary systemic amyloidosis. Haematologica 85:289, 2000. 62. Gastineau, D. A., Gertz, M. A., Daniels, T. M., et al.: Inhibitor of the thrombin time in systemic amyloidosis: a common coagulation abnormality. Blood 77:2637, 1991. 63. Gertz, M. A., Brown, M. L., Hauser, M. F., and Kyle, R. A.: Utility of technetium Tc 99m pyrophosphate bone scanning in cardiac amyloidosis. Arch. Intern. Med. 147:1039, 1987. 64. Gertz, M. A., Brown, M. L., Hauser, M. F., and Kyle, R. A.: Utility of gallium imaging of the kidneys in diagnosing primary amyloid nephrotic syndrome. J. Nucl. Med. 31:292, 1990. 65. Gertz, M. A., Garton, J. P., and Kyle, R. A.: Primary amyloidosis (AL) in families. Am. J. Hematol. 22:193, 1986. 66. Gertz, M. A., Grogan, M., Kyle, R. A., and Tajik, A. J.: Endomyocardial biopsy-proven light chain amyloidosis (AL) without echocardiographic features of infiltrative cardiomyopathy. Am. J. Cardiol. 80:93, 1997. 67. Gertz, M. A., and Kyle, R. A.: Hepatic amyloidosis (primary [AL], immunoglobulin light chain): the natural history in 80 patients. Am. J. Med. 85:73, 1988. 68. Gertz, M. A., and Kyle, R. A.: Acute leukemia and cytogenetic abnormalities complicating melphalan treatment of primary systemic amyloidosis. Arch. Intern. Med. 150:629, 1990. 69. Gertz, M. A., and Kyle, R. A.: Phase II trial of alphatocopherol (vitamin E) in the treatment of primary systemic amyloidosis. Am. J. Hematol. 34:55, 1990. 70. Gertz, M. A., and Kyle, R. A.: Phase II trial of recombinant interferon alfa-2 in the treatment of primary systemic amyloidosis. Am. J. Hematol. 44:125, 1993. 71. Gertz, M. A., and Kyle, R. A.: Myopathy in primary systemic amyloidosis. J. Neurol. Neurosurg. Psychiatry 60:655, 1996. 72. Gertz, M. A., and Kyle, R. A.: Hepatic amyloidosis: clinical appraisal in 77 patients. Hepatology 25:118, 1997. 73. Gertz, M. A., Kyle, R. A., Griffing, W. L., and Hunder, G. G.: Jaw claudication in primary systemic amyloidosis. Medicine (Baltimore) 65:173, 1986. 74. Gertz, M. A., Kyle, R. A., and Thibodeau, S. N.: Familial amyloidosis: a study of 52 North American-born patients examined during a 30-year period. Mayo Clin. Proc. 67:428, 1992. 75. Gertz, M. A., Lacy, M. Q., Dispenzieri, A., et al.: A multicenter Phase II trial of 4⬘-iodo-4⬘-deoxydoxorubicin (IDOX) in primary amyloidosis (AL) [abstract]. Blood 96 (II Pt. 2):288b, 2000. 76. Gertz, M. A., Lacy, M. Q., Gastineau, D. A., et al.: Blood stem cell transplantation as therapy for primary systemic amyloidosis (AL). Bone Marrow Transplant. 26:963, 2000. 77. Gertz, M. A., Lacy, M. Q., Lust, J. A., et al.: Phase II trial of high-dose dexamethasone for previously treated immunoglobulin light-chain amyloidosis. Am. J. Hematol. 61:115, 1999. 78. Gertz, M. A., Lacy, M. Q., Lust, J. A., et al.: Phase II trial of high-dose dexamethasone for untreated patients with primary systemic amyloidosis. Med. Oncol. 16:104, 1999. 79. Gertz, M. A., Lacy, M. Q., Lust, J. A., et al.: Prospective randomized trial of melphalan and prednisone versus vin-

Amyloidosis and Neuropathy

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

cristine, carmustine, melphalan, cyclophosphamide, and prednisone in the treatment of primary systemic amyloidosis. J. Clin. Oncol. 17:262, 1999. Gertz, M. A., Li, C. Y., Shirahama, T., and Kyle, R. A.: Utility of subcutaneous fat aspiration for the diagnosis of systemic amyloidosis (immunoglobulin light chain). Arch. Intern. Med. 148:929, 1988. Gianni, L., Bellotti, V., Gianni, A. M., and Merlini, G.: New drug therapy of amyloidoses: resorption of AL-type deposits with 4⬘-iodo-4⬘-deoxydoxorubicin. Blood 86:855, 1995. Gillmore, J. D., Booth, D. R., Madhoo, S., et al.: Hereditary renal amyloidosis associated with variant lysozyme in a large English family. Nephrol. Dial. Transplant. 14:2639, 1999. Gillmore, J. D., Hawkins, P. N., and Pepys, M. B.: Amyloidosis: a review of recent diagnostic and therapeutic developments. Br. J. Haematol. 99:245, 1997. Glenner, G. G.: Amyloid deposits and amyloidosis: the betafibrilloses (second of two parts). N. Engl. J. Med. 302:1333, 1980. Glenner, G. G., Terry, W., Harada, M., et al.: Amyloid fibril proteins: proof of homology with immunoglobulin light chains by sequence analyses. Science 172:1150, 1971. Hachulla, E., Janin, A., Flipo, R. M., et al.: Labial salivary gland biopsy is a reliable test for the diagnosis of primary and secondary amyloidosis: a prospective clinical and immunohistologic study in 59 patients. Arthritis Rheum. 36:691, 1993. Hachulla, E., Maulin, L., Deveaux, M., et al.: Prospective and serial study of primary amyloidosis with serum amyloid P component scintigraphy: from diagnosis to prognosis. Am. J. Med. 101:77, 1996. Hamidi Asl, L., Fournier, V., Billerey, C., et al.: Fibrinogen Aa chain mutation (Arg554 Leu) associated with hereditary renal amyloidosis in a French family. Amyloid 5:279:1998. Harada, T., Kito, S., Ishizaki, F., et al.: Clinical features and diagnosis by recombinant DNA techniques of familial amyloid polyneuropathy in Japan. Res. Commun. Chem. Pathol. Pharmacol. 65:237, 1989. Harada, T., Kito, S., Shimoyama, M., et al.: Genetic and clinical studies of Japanese patients with familial amyloid polyneuropathy. Eur. Neurol. 29:48, 1989. Harats, N., Worth, R., and Benson, M. D.: Spinal claudication in systemic amyloidosis. J. Rheumatol. 16:1003, 1989. Hawkins, P. N., Myers, M. J., Lavender, J. P., and Pepys, M. B.: Diagnostic radionuclide imaging of amyloid: biological targeting by circulating human serum amyloid P component. Lancet 1:1413, 1988. Hayman, S. R., Lacy, M. Q., Kyle, R. A., and Gertz, M. A.: Primary systemic amyloidosis: a cause of malabsorption syndrome. Am. J. Med. 111:535, 2001. Helin, H., Pasternack, A., Falck, H., and Kuhlback, B.: Recurrence of renal amyloid and de novo membranous glomerulonephritis after transplantation. Transplantation 32:6, 1981. Holmgren G., Ericzon, B.-G., Groth, C.-G., et al.: Clinical improvement and amyloid regression after liver transplantation in hereditary transthyretin amyloidosis. Lancet 341:1113, 1993.

2447

96. Holt, I. J., Harding, A. E., Middleton, L., et al.: Molecular genetics of amyloid neuropathy in Europe. Lancet 1:524, 1989. 97. Huang, C. Y., Wang, W. J., and Wong, C. K.: Skin biopsy gives the potential benefit in the diagnosis of systemic amyloidosis associated with cardiovascular involvement. Arch. Dermatol. 134:643, 1998. 98. Ikeda, S., Hanyu, N., Hongo, M., et al.: Hereditary generalized amyloidosis with polyneuropathy: clinicopathological study of 65 Japanese patients. Brain 110:315, 1987. 99. Ikeda, S., Yanagisawa, N., Hongo, M., and Ito, N.: Vagus nerve and celiac ganglion lesions in generalized amyloidosis: a correlative study of familial amyloid polyneuropathy and AL-amyloidosis. J. Neurol. Sci. 79:129, 1987. 100. Inomata, Y., Nakamura, T., Uemoto, S., et al.: Domino splitliver transplantation from a living donor: case reports of in situ and ex situ splitting. Liver Transpl. 7:150, 2001. 101. Jager, P. L., Hazenberg, B. P., Franssen, E. J., et al.: Kinetic studies with iodine-123-labeled serum amyloid P component in patients with systemic AA and AL amyloidosis and assessment of clinical value. J. Nucl. Med. 39:699, 1998. 102. Jennekens, F. G., and Wokke, J. H.: Proximal weakness of the extremities as main feature of amyloid myopathy. J. Neurol. Neurosurg. Psychiatry 50:1353, 1987. 103. Jones, B. A., Shapiro, H. S., Rosenberg, B. F., and Bernstein, J.: Minimal renal amyloidosis with nephrotic syndrome. Arch. Pathol. Lab. Med. 110:889, 1986. 104. Jonsen, E., Suhr, O., Athlin, E., and Wikström, L.: Qualityof-life after liver-transplantation in patients with familial amyloidotic polyneuropathy. Amyloid 3:124, 1996. 105. Julien, J., Vital, C., Vallat, J. M., et al.: IgM demyelinative neuropathy with amyloidosis and biclonal gammopathy. Ann. Neurol. 15:395, 1984. 106. Kaiserling, E., and Krober, S.: Massive intestinal hemorrhage associated with intestinal amyloidosis: an investigation of underlying pathologic processes. Gen. Diagn. Pathol. 141:147, 1995. 107. Kangas, H., Ulmanen, I., Paunio, T., et al.: Functional consequences of amyloidosis mutation for gelsolin polypeptide—analysis of gelsolin-actin interaction and gelsolin processing in gelsolin knock-out fibroblasts. FEBS Lett. 454:233, 1999. 108. Katafuchi, R., Taguchi, T., Takebayashi, S., and Harada, T.: Proteinuria in amyloidosis correlates with epithelial detachment and distortion of amyloid fibrils. Clin. Nephrol. 22:1, 1984. 109. Katz, B., Leja, S., Melles, R. B., and Press, G. A.: Amyloid ophthalmoplegia: ophthalmoparesis secondary to primary systemic amyloidosis. J. Clin. Neuroophthalmol. 9:39, 1989. 110. Katz, J. N., Larson, M. G., Sabra, A., et al.: The carpal tunnel syndrome: diagnostic utility of the history and physical examination findings. Ann. Intern. Med. 112:321, 1990. 111. Kawasaki, S., Makuuchi, M., Matsunami, H., et al.: Living related liver transplantation in adults. Ann. Surg. 227:269, 1998. 112. Kelly, J. J. Jr.: The electrodiagnostic findings in peripheral neuropathy associated with monoclonal gammopathy. Muscle Nerve 6:504, 1983.

2448

Diseases of the Peripheral Nervous System

113. Kelly, J. J. Jr., Kyle, R. A., O’Brien, P. C., and Dyck, P. J.: The natural history of peripheral neuropathy in primary systemic amyloidosis. Ann. Neurol. 6:1, 1979. 114. Kelly, J. W., Colon, W., Lai, Z., et al.: Transthyretin quaternary and tertiary structural changes facilitate misassembly into amyloid. Adv. Protein Chem. 50:161,1997. 115. Kernohan, J. W., and Woltman, H. W.: Amyloid neuritis. Arch. Neurol. Psychiatry 47:132, 1942. 116. Kincaid, J. C., Wallace, M. R., and Benson, M. D.: Lateonset familial amyloid polyneuropathy in an American family of English origin. Neurology 39:861, 1989. 117. Kitamoto, T., Ogomori, K., Tateishi, J., and Prusiner, S. B.: Formic acid pretreatment enhances immunostaining of cerebral and systemic amyloids. Lab. Invest. 57:230, 1987. 118. Kiuru, S.: Gelsolin-related familial amyloidosis, Finnish type (FAF), and its variants found worldwide. Amyloid 5:55, 1998 119. Kiuru, S., Salonen, O., and Haltia, M.: Gelsolin-related spinal and cerebral amyloid angiopathy. Ann. Neurol. 45:305, 1999. 120. Klabunde, T., Petrassi, H. M., Oza, V. B., et al.: Rational design of potent human transthyretin amyloid disease inhibitors. Nat. Struct. Biol. 7:312, 2000. 121. Klein, A. L., Hatle, L. K., Burstow, D. J., et al.: Doppler characterization of left ventricular diastolic function in cardiac amyloidosis. J. Am. Coll. Cardiol. 13:1017, 1989. 122. Kono, J., Hongo, M., Okubo, S., et al.: Left ventricular systolic and diastolic time intervals in familial amyloid polyneuropathy. Jpn. Circ. J. 53:291, 1989. 123. Koppelman, R. N., Stollman, N. H., Baigorri, F., and Rogers, A. I.: Acute small bowel pseudo-obstruction due to AL amyloidosis: a case report and literature review. Am. J. Gastroenterol. 95:294, 2000. 124. Kyle, R. A.: Amyloidosis: a convoluted story. Br. J. Haematol. 114:529, 2001. 125. Kyle, R. A., Eilers, S. G., Linscheid, R. L., and Gaffey, T. A.: Amyloid localized to tenosynovium at carpal tunnel release: natural history of 124 cases. Am. J. Clin. Pathol. 91:393, 1989. 126. Kyle, R. A., and Gertz, M. A.: Systemic amyloidosis. Crit. Rev. Oncol. Hematol. 10:49, 1990. 127. Kyle, R. A., and Gertz, M. A.: Primary systemic amyloidosis: clinical and laboratory features in 474 cases. Semin. Hematol. 32:45, 1995. 128. Kyle, R. A., Gertz, M. A., Greipp, P. R., et al.: A trial of three regimens for primary amyloidosis: colchicine alone, melphalan and prednisone, and melphalan, prednisone, and colchicine. N. Engl. J. Med. 336:1202, 1997. 129. Kyle, R. A., Gertz, M. A., and Linke, R. P.: Amyloid localized to tenosynovium at carpal tunnel release: immunohistochemical identification of amyloid type. Am. J. Clin. Pathol. 97:250, 1992. 130. Kyle, R. A., and Greipp, P. R.: Primary systemic amyloidosis: comparison of melphalan and prednisone versus placebo. Blood 52:818, 1978. 131. Kyle, R. A., Greipp, P. R., Garton, J. P., and Gertz, M. A.: Primary systemic amyloidosis: comparison of melphalan/prednisone versus colchicine. Am. J. Med. 79:708,1985. 132. Kyle, R. A., Kottke, B. A., and Schirger, A.: Orthostatic hypotension as a clue to primary systemic amyloidosis. Circulation 34:883, 1966.

133. Kyle, R. A., Linos, A., Beard, C. M., et al.: Incidence and natural history of primary systemic amyloidosis in Olmsted County, Minnesota, 1950 through 1989. Blood 79:1817, 1992. 134. Laskaratos, J., and Fronimopoulos, J.: Familial amyloidosis of the Andrade disease type in a Greek family [in German]. Klin. Monatsbl. Augenheilkd. 190:439, 1987. 135. Legge, D. A., Carlson, H. C., and Wollaeger, E. E.: Roentgenologic appearance of systemic amyloidosis involving gastrointestinal tract. Am. J. Roentgenol. Radium Ther. Nucl. Med. 110:406, 1970. 136. Lewis, W. D.: Liver transplantation: an effective treatment for familial ATTR amyloidosis. Amyloid 9:201, 2002. 137. Li, K., Hizawa, K., Nunomura, S., and Morizumi, H.: Systemic amyloid myopathy: light-microscopic and finestructural study of the skeletal muscles with histochemical and immunohistochemical study of amyloid. Acta Neuropathol. (Berl.) 64:114, 1984. 138. Li, K., Kyle, R. A., and Dyck, P. J.: Immunohistochemical characterization of amyloid proteins in sural nerves and clinical associations in amyloid neuropathy. Am. J. Pathol. 141:217, 1992. 139. Libbey, C. A., Rubinow, A., Shirahama, T., et al.: Familial amyloid polyneuropathy: demonstration of prealbumin in a kinship of German/English ancestry with onset in the seventh decade. Am. J. Med. 76:18, 1984. 140. Linke, R. P., Nathrath, W. B. J., and Eulitz, M.: Classification of amyloid syndromes from tissue sections using antibodies against various amyloid fibril proteins: report of 142 cases. In Glenner, G. G., Osserman, E. F., Benditt, E. P., et al. (eds.): Amyloidosis. New York, Plenum Press, p. 599, 1986. 141. Little, K. H., Lee, E. L., and Frenkel, E. P.: Cranial nerve deficits due to amyloidosis associated with plasma cell dyscrasia. South. Med. J. 79:677, 1986. 142. Low, P. A., Dyck, P. J., Okazaki, H., et al.: The splanchnic autonomic outflow in amyloid neuropathy and Tangier disease. Neurology 31:461, 1981. 143. Magnus-Levy, A.: Bence-Jones-Eiweiss und amyloid. Z. Klin. Med. 116:510, 1931. 144. Mahloudji, M., Teasdall, R. D., Adamkiewicz, J. J., et al.: The genetic amyloidoses with particular reference to hereditary neuropathic amyloidosis, type II (Indiana or Rukavina type). Medicine (Baltimore) 48:1, 1969. 145. Maruyama, K., Ikeda, S., Yanagisawa, N., and Nakazato, M.: Diagnostic value of abdominal fat tissue aspirate in familial amyloid polyneuropathy. J. Neurol. Sci. 81:11, 1987. 146. Massey, E. W., and Massey, J. M.: Facial diplegia due to amyloidosis. South. Med. J. 79:1458, 1986. 147. Meretoja, J.: Familial systemic paramyloidosis with lattice dystrophy of the cornea, progressive cranial neuropathy, skin changes and various internal symptoms: a previously unrecognized heritable syndrome. Ann. Clin. Res. 1:314, 1969. 148. Mita, S., Maeda, S., Shimada, K., and Araki, S.: Analyses of prealbumin mRNAs in individuals with familial amyloidotic polyneuropathy. J. Biochem. (Tokyo) 100:1215, 1986. 149. Nakazato, M., Kangawa, K., Kurihara, T., et al.: Variant transthyretin in cerebrospinal fluid in familial amyloidotic polyneuropathy. J. Neurol. Sci. 79:111, 1987.

Amyloidosis and Neuropathy 150. Nakazato, M., Kangawa, K., Minamino, N., et al.: Revised analysis of amino acid replacement in a prealbumin variant (SKO-III) associated with familial amyloidotic polyneuropathy of Jewish origin. Biochem. Biophys. Res. Commun. 123:921, 1984. 151. Nakazato, M., Tanaka, M., Yamamura, Y., et al.: Abnormal transthyretin in asymptomatic relatives in familial amyloidotic polyneuropathy. Arch. Neurol. 44:1275, 1987. 152. Nichols, W. C., Dwulet, F. E., Liepnieks, J., and Benson, M. D.: Variant apolipoprotein AI as a major constituent of a human hereditary amyloid. Biochem. Biophys. Res. Commun. 156:762, 1988. 153. Nomura, S., Kanoh, T., and Uchino, H.: Intracellular formation of amyloid fibrils in myeloma: cytochemical, immunochemical and electron microscopic observations [in Japanese]. Nippon Ketsueki Gakkai Zasshi 45:615, 1982. 154. O’Connor, C. R., Rubinow, A., Brandwein, S., and Cohen, A. S.: Familial amyloid polyneuropathy: a new kinship of German ancestry. Neurology 34:1096, 1984. 155. Olafsson, I., and Grubb, A.: Hereditary cystatin C amyloid angiopathy. Amyloid 7:70, 2000. 156. Olofsson, B. O., Bjerle, P., and Osterman, G.: Hemodynamic and angiocardiographic observations in familial amyloidosis with polyneuropathy. Acta Med. Scand. 212:77, 1982. 157. Olofsson, B. O., Grankvist, K., Boman, K., et al.: Assessment of thyroid and adrenal function in patients with familial amyloidotic polyneuropathy. J. Intern. Med. 225:337, 1989. 158. Osserman, E. F., Takatsuki, K., and Talal, N.: Multiple myeloma I: the pathogenesis of “amyloidosis.” Semin. Hematol. 1:3, 1964. 159. Pascali, E., Pezzoli, A., Melato, M., and Antonutto, G.: Pseudotumoral (para-articular) amyloidosis in nonmyelomatous monoclonal gammopathy. Pathol. Res. Pract. 168:215, 1980. 160. Pasternack, A., Ahonen, J., and Kuhlback, B.: Renal transplantation in 45 patients with amyloidosis. Transplantation 42:598, 1986. 161. Peterson, S. A., Klabunde, T., Lashuel, H. A., et al.: Inhibiting transthyretin conformational changes that lead to amyloid fibril formation. Proc. Natl. Acad. Sci. U. S. A. 95:12956, 1998. 162. Pomfret, E. A., Lewis, W. D., Jenkins, R. L., et al.: Effect of orthotopic liver transplantation on the progression of familial amyloidotic polyneuropathy. Transplantation 65:918, 1998. 163. Pras, M., Prelli, F., Franklin, E. C., and Frangione, B.: Primary structure of an amyloid prealbumin variant in familial polyneuropathy of Jewish origin. Proc. Natl. Acad. Sci. U. S. A. 80:539, 1983. 164. Preud’homme, J. L., Ganeval, D., Grundfeld, J. P., et al.: Immunoglobulin synthesis in primary and myeloma amyloidosis. Clin. Exp. Immunol. 73:389, 1988. 165. Puchtler, H., and Sweat, F.: Congo red as a stain for fluorescence microscopy of amyloid. J. Histochem. Cytochem. 13:693, 1965. 166. Purkey, H. E., Dorrell, M. I., and Kelly, J. W.: Evaluating the binding selectivity of transthyretin amyloid fibril inhibitors in blood plasma. Proc. Natl. Acad. Sci. U. S. A. 98:5566, 2001.

2449

167. Quattrini, A., Nemni, R., Sferrazza, B., et al.: Amyloid neuropathy simulating lower motor neuron disease. Neurology 51:600, 1998. 168. Quintas, A., Vaz, D. C., Cardoso, I., et al.: Tetramer dissociation and monomer partial unfolding precedes protofibril formation in amyloidogenic transthyretin variants. J. Biol. Chem. 276:27207, 2001. 169. Rajani, B, Rajani, V., and Prayson, R. A.: Peripheral nerve amyloidosis in sural nerve biopsies: a clinicopathologic analysis of 13 cases. Arch. Pathol. Lab. Med. 124:114, 2000. 170. Rajkumar, S. V., Gertz, M. A., and Kyle, R. A.: Prognosis of patients with primary systemic amyloidosis who present with dominant neuropathy. Am. J. Med. 104:232, 1998. 171. Reilly, M. M., Adams, D., Davis, M. B., et al.: Haplotype analysis of French, British and other European patients with familial amyloid polyneuropathy (Met 30 and Tyr 77). J. Neurol. 242:664, 1995. 172. Roberts, W. C., and Waller, B. F.: Cardiac amyloidosis causing cardiac dysfunction: analysis of 54 necropsy patients. Am. J. Cardiol. 52:137, 1983. 173. Roke, M. E., Brown, W. F., Boughner, D., et al.: Myopathy in primary systemic amyloidosis. Can. J. Neurol. Sci. 15:314, 1988. 174. Rubinow, A., Burakoff, R., Cohen, A. S., and Harris, L. D.: Esophageal manometry in systemic amyloidosis: a study of 30 patients. Am. J. Med. 75:951, 1983. 175. Rukavina, J. G., Block, W. D., and Curtis, A. C.: Familial primary systemic amyloidosis: experimental, genetic and clinical study. J. Invest. Dermatol. 27:111, 1956. 176. Rukavina, J. G., Block, W. D., Jackson, C. E., et al.: Primary systemic amyloidosis: review and experimental, genetic, and clinical study of 29 cases with particular emphasis on the familial form. Medicine 35:239, 1956. 177. Rydh, A., Suhr, O., Hietala, S. O., et al.: Serum amyloid P component scintigraphy in familial amyloid polyneuropathy: regression of visceral amyloid following liver transplantation. Eur. J. Nucl. Med. 25:709, 1998. 178. Said, G., Ropert, A., and Faux, N.: Length-dependent degeneration of fibers in Portuguese amyloid polyneuropathy: a clinicopathologic study. Neurology 34:1025, 1984. 179. Santos, M., Castro, L., Rio, E., and Coutinho, P.: Cranial nerve involvement in familial amyloidotic polyneuropathy (Portuguese type). Paper presented at the First International Symposium on Familial Amyloidotic Polyneuropathy and Other Transthyretin Related Disorders, Granja, Portugal, September 25–27, 1989. 180. Saraiva, M. J., Costa, P. P., and Goodman, D. S.: Biochemical marker in familial amyloidotic polyneuropathy, Portuguese type: family studies on the transthyretin (prealbumin)-methionine-30 variant. J. Clin. Invest. 76:2171, 1985. 181. Saraiva, M. J., Costa, P. P., and Goodman, D. S.: Genetic expression of a transthyretin mutation in typical and lateonset Portuguese families with familial amyloidotic polyneuropathy. Neurology 36:1413, 1986. 182. Saraiva, M. J., Costa, P. P., and Goodman, D. S.: Transthyretin (prealbumin) in familial amyloidotic polyneuropathy: genetic and functional aspects. Adv. Neurol. 48:189, 1988.

2450

Diseases of the Peripheral Nervous System

183. Saraiva, M. J., Sherman, W., and Goodman, D. S.: Presence of a plasma transthyretin (prealbumin) variant in familial amyloidotic polyneuropathy in a kindred of Greek origin. J. Lab. Clin. Med. 108:17, 1986. 184. Saraiva, M. J. M., Costa, P. P., Almeida, M. R., et al.: Familial amyloidotic polyneuropathy: transthyretin (prealbumin) variants in kindreds of Italian origin. Hum. Genet. 80:341, 1988. 185. Schwartz, P.: Amyloidosis: Cause and Manifestation of Senile Deterioration. Springfield, IL, Charles C Thomas, 1970. 186. Shirahama, T., Skinner, M., and Cohen, A. S.: Immunocytochemical identification of amyloid in formalinfixed paraffin sections. Histochemistry 72:161, 1981. 187. Skinner, M., Libbey, C. A., Skare, J. C., et al.: Tyrosine for serine 77 is the variant transthyretin (TTR) in a United States family of German/English ancestry with familial amyloid polyneuropathy. Paper presented at the First International Symposium on Familial Amyloidotic Polyneuropathy and Other Transthyretin Related Disorders, Granja, Portugal, September 25–27, 1989. 188. Smith, R. R., Hutchins, G. M., Moore, G. W., and Humphrey, R. L.: Type and distribution of pulmonary parenchymal and vascular amyloid: correlation with cardiac amyloid. Am. J. Med. 66:96, 1979. 189. Smith, T. J., Kyle, R. A., and Lie, J. T.: Clinical significance of histopathologic patterns of cardiac amyloidosis. Mayo Clin. Proc. 59:547, 1984. 190. Solomon, A., Kyle, R. A., and Frangione, B.: Light chain variable region subgroups of monoclonal immunoglobulins in amyloidosis-AL. In Glenner, G. G., Osserman, E. F., Benditt, E. P., et al. (eds.): Amyloidosis. New York, Plenum Press, p. 449, 1986. 191. Sommer, C., and Schroder, J. M.: Amyloid neuropathy: immunocytochemical localization of intra- and extracellular immunoglobulin light chains. Acta Neuropathol. (Berl.) 79:190, 1989. 192. Sparkes, R. S., Sasaki, H., Mohandas, T., et al.: Assignment of the prealbumin (PALB) gene (familial amyloidotic polyneuropathy) to human chromosome region 18q11.2q12.1. Hum. Genet. 75:151, 1987. 193. Spinner, R. J., Bachman, J. W., and Amadio, P. C.: The many faces of carpal tunnel syndrome. Mayo Clin. Proc. 64:829, 1989. 194. Staunton, H., Dervan, P., Kale, R., et al.: Hereditary amyloid polyneuropathy in north west Ireland. Brain 110:1231, 1987. 195. Steen, L., Wahlin, A., Bjerle, P., and Holm, S.: Renal function in familial amyloidosis with polyneuropathy. Acta Med. Scand. 212:233, 1982. 196. Steen, L. E., and Ek, B. O.: Familial amyloidosis with polyneuropathy: aspects of the relationship between gastrointestinal symptoms, EMG findings, and malabsorption studies. Scand. J. Gastroenterol. 19:480, 1984. 197. Stone, M. J.: Amyloidosis: a final common pathway for protein deposition in tissues. Blood 75:531, 1990. 198. Suhr, O., Danielsson, A., Rydh, A., et al.: Impact of gastrointestinal dysfunction on survival after liver transplantation for familial amyloidotic polyneuropathy. Dig. Dis. Sci. 41:1909, 1996.

199. Suhr, O. B., Herlenius, G., Friman, S., and Ericzon, B. G.: Liver transplantation for hereditary transthyretin amyloidosis. Liver Transpl. 6:263, 2000. 200. Suhr, O. B., Holmgren, G., Steen, L., et al.: Liver transplantation in familial amyloidotic polyneuropathy: follow-up of the first 20 Swedish patients. Transplantation 60:933, 1995. 201. Tawara, S., Araki, S., Toshimori, K., et al.: Amyloid fibril protein in type I familial amyloidotic polyneuropathy in Japanese. J. Lab. Clin. Med. 98:811, 1981. 202. Tei, C., Dujardin, K. S., Hodge, D. O., et al.: Doppler index combining systolic and diastolic myocardial performance: clinical value in cardiac amyloidosis. J. Am. Coll. Cardiol. 28:658, 1996. 203. Thornton, C.: Amyloid disease: an autopsy review of the decades 1937–46 and 1961–70. Ulster Med. J. 52:31, 1983. 204. Tokuda, T., Kondo, T., Hanaoka, N., et al.: A selective transthyretin-adsorption column for the treatment of patients with familial amyloid polyneuropathy. Amyloid 5:111, 1998. 205. Travels of a Gene. Guide to the exhibition shown during the First International Symposium on Familial Amyloidotic Polyneuropathy and Other Transthyretin Related Disorders, Granja, Portugal, September 25–27, 1989. 206. Traynor, A. E., Gertz, M. A., and Kyle, R. A.: Cranial neuropathy associated with primary amyloidosis. Ann. Neurol. 29:451, 1991. 207. Uemichi, T., Liepnieks, J. J., Yamada, T., et al.: A frame shift mutation in the fibrinogen A alpha chain gene in a kindred with renal amyloidosis. Blood 87:4197, 1996. 208. Utz, J. P., Swensen, S. J., and Gertz, M. A.: Pulmonary amyloidosis: the Mayo Clinic experience from 1980 to 1993. Ann. Intern. Med. 124:407, 1996. 209. Van Allen, M. W., Frohlich, J. A., and Davis, J. R.: Inherited predisposition to generalized amyloidosis: clinical and pathological study of a family with neuropathy, nephropathy, and peptic ulcer. Neurology 19:10, 1969. 210. Verghese, J. P., Bradley, W. G., Nemni, R., and McAdam, K. P.: Amyloid neuropathy in multiple myeloma and other plasma cell dyscrasias: a hypothesis of the pathogenesis of amyloid neuropathies. J. Neurol. Sci. 59:237, 1983. 211. Vital, A., and Vital, C.: Amyloid neuropathy: relationship between amyloid fibrils and macrophages. Ultrastruct. Pathol. 7:21, 1984. 212. Voulgarelis, M., Goutas, N., and Skopouli, F. N.: A 47-yearold woman with persistent watery diarrhea, proteinuria, proximal weakness and monoclonal gammopathy. Clin. Exp. Rheumatol. 17:351, 1999. 213. Wallace, M. R., Conneally, P. M., and Benson, M. D.: A DNA test for Indiana/Swiss hereditary amyloidosis (FAP II). Am. J. Hum. Genet. 43:182, 1988. 214. Wallace, M. R., Dwulet, F. E., Conneally, P. M. and Benson, M. D.: Biochemical and molecular genetic characterization of a new variant prealbumin associated with hereditary amyloidosis. J. Clin. Invest. 78:6, 1986. 215. Wallace, M. R., Dwulet, F. E., Williams, E. C., et al.: Identification of a new hereditary amyloidosis prealbumin variant, Tyr-77, and detection of the gene by DNA analysis. J. Clin. Invest. 81:189, 1988.

Amyloidosis and Neuropathy 216. Wardley, A. M., Jayson, G. C., Goldsmith, D. J., et al.: The treatment of nephrotic syndrome caused by primary (light chain) amyloid with vincristine, doxorubicin and dexamethasone. Br. J. Cancer 78:774, 1998. 217. Watanabe, T., and Saniter, T.: Morphological and clinical features of renal amyloidosis. Virchows Arch. [A] 366:125, 1975. 218. Weber, H.: Mollities ossium, doubtful whether carcinomatous or syphilitic. Trans. Pathol. Soc. Lond. 18:206, 1867. 219. Wikström L., Bjerle, P., Boman, K., et al.: L-threo-DOPS treatment of orthostatic hypotension in Swedish patients with familial amyloidotic polyneuropathy (TTR-Met 30). Amyloid 3:162, 1996. 220. Wilks, S.: Cases of lardaceous disease and some allied affections: with remarks. Guys Hosp. Rep. 2 (Ser. 3):103, 1856. 221. Wright, J. R., Calkins, E., and Humphrey, R. L.: Potassium permanganate reaction in amyloidosis: a histologic method

222. 223.

224.

225.

226.

2451

to assist in differentiating forms of this disease. Lab. Invest. 36:274, 1977. Yamaguchi, D. M., Lipscomb, P. R., and Soule, E. H.: Carpal tunnel syndrome. Minn. Med. 48:22, 1965. Yood, R. A., Skinner, M., Rubinow, A., et al.: Bleeding manifestations in 100 patients with amyloidosis. JAMA 249:1322, 1983. Yoshioka, K., Furuya, H., Sasaki, H., et al.: Haplotype analysis of familial amyloidotic polyneuropathy: an evidence for multiple origins of the Val:Met mutation. In Isobe, T., Araki, S., Uchino, F., et al. (eds): Amyloid and Amyloidosis. New York, Plenum Press, p. 371, 1988. Yoshioka, K., Furuya, H., Sasaki, H., et al.: Haplotype analysis of familial amyloidotic polyneuropathy: evidence for multiple origins of the Val:Met mutation most common to the disease. Hum. Genet. 82:9, 1989. Zatloukal, P., Bezdicek, P., Schimonova, M., et al.: Waldenström’s macroglobulinemia with pulmonary amyloidosis. Respiration 65:414, 1998.

109 POEMS Syndrome (Osteosclerotic Myeloma) ANGELA DISPENZIERI, GUILLERMO A. SUAREZ, AND ROBERT A. KYLE

Overview History Pathogenesis Clinical Features Polyneuropathy Skin Changes

Endocrine Abnormalities Bone Marrow and Bone Involvement Association with Castleman’s Disease (Giant Lymph Node Hyperplasia, Angiofollicular Lymph Node Hyperplasia)

OVERVIEW The major clinical feature in POEMS syndrome is a chronic progressive polyneuropathy with a predominant motor disability associated with osteosclerotic bone lesions. The acronym POEMS (polyneuropathy, organomegaly, endocrinopathy, M protein, and skin changes) captures several dominant features of the syndrome. Wellrecognized associated features not included in the acronym include sclerotic bone lesions, Castleman’s disease, papilledema, peripheral edema, ascites, polycythemia, thrombocytosis, fatigue, and clubbing.36,104,150 Other names for the syndrome include osteosclerotic myeloma, Crow-Fukase syndrome, PEP (plasma cell dyscrasia, endocrinopathy, and polyneuropathy) syndrome, and Takatsuki syndrome.40,104,154 Not all features are required to make the diagnosis; at a minimum, however, a patient must have the peripheral neuropathy; osteosclerotic myeloma (i.e., a clonal plasma cell dyscrasia and at least one sclerotic bone lesion) or Castleman’s disease; and at least one of the other features36 (Table 109–1). Though the majority of patients have osteosclerotic myeloma, these same patients usually have only 5% bone marrow plasma cells or less, and rarely have hypercalcemia or renal insufficiency. These characteristics and the superior median survival differentiate POEMS syndrome from multiple

Features Likely Related to POEMS Laboratory Features Diagnosis Treatment Strategies Course and Prognosis (Survival)

myeloma. Most patients have a  light chain containing M protein. The cerebrospinal fluid (CSF) is acellular and its protein level is elevated. Electromyography reveals slow motor and sensory nerve conduction velocities. Though the pathophysiologic mechanism is not well understood, there is a correlation between treating the underlying plasma cell proliferative disorder (clone) and clinical improvement. This observation clearly links the plasma cell clone to the peripheral neuropathy and other clinical features, though the mechanism remains obscure. Radiation therapy produces substantial improvement of the neuropathy in more than half of the patients who have a single lesion or multiple lesions in a limited area. If there are widespread lesions, high-dose chemotherapy and peripheral blood support may be helpful.

HISTORY Scheinker’s autopsy case in 1938 was the first report of what we now call POEMS syndrome, Crow-Fukase syndrome, PEP syndrome, or Takatsuki syndrome.40,104,134 His patient was a 39-year-old man with a solitary plasmacytoma, sensorimotor polyneuropathy, and localized patches of thickened and deeply pigmented skin on the chest.68,134 Moreover, as many as 50% of patients with 2453

2454

Diseases of the Peripheral Nervous System

Table 109–1. Criteria for the Diagnosis of POEMS Syndrome* Major Criteria Polyneuropathy Monoclonal plasma cell proliferative disorder

Minor Criteria

Known Associations

Possible Associations

Sclerotic bone lesions Castleman’s disease†

Clubbing Weight loss

Pulmonary hypertension Restrictive lung disease

Organomegaly (splenomegaly, hepatomegaly, or lymphadenopathy) Edema (edema, pleural effusion, or ascites) Endocrinopathy (adrenal, thyroid,‡ pituitary, gonadal, parathyroid, pancreatic‡) Skin changes (hyperpigmentation, hypertrichosis, plethora, hemangiomata, white nails) Papilledema

Thrombocytosis Polycythemia Hyperhidrosis

Thrombotic diatheses Arthralgias Cardiomyopathy (systolic dysfunction) Fever

†

Low vitamin B12 values Diarrhea

POEMS indicates polyneuropathy, organomegaly, endocrinopathy, M protein, skin changes. *Two major criteria and at least 1 minor criterion required for diagnosis. † Osteosclerotic lesion or Castleman’s disease is almost always present. ‡ Because of the high prevalence of diabetes mellitus and thyroid abnormalities, this diagnosis alone is not sufficient to meet this minor criterion. From Dispenzieri, A., Kyle, R. A., Lacy, M. Q., et al.: POEMS syndrome: definitions and long-term outcome. Blood 101:2496, 2003, with permission.

osteosclerotic myeloma have peripheral neuropathy,39,68,85 in contrast to 1% to 8% of multiple myeloma patients.44,118 The complexity of the interaction of plasma cell dyscrasia and peripheral neuropathy became increasingly evident in 1956 with Crow’s description of two patients with osteosclerotic plasmacytomas with neuropathy, and other “striking features,” which included clubbing, skin pigmentation, dusky discoloration of skin, white fingernails, mild lymphadenopathy, and ankle edema.32 Other authors reported patients with osteosclerotic myeloma and peripheral neuropathy with organomegaly, skin changes, endocrinopathy, edema, hypertrichosis, gynecomastia, and ascites.39,40,49,53,66,68,85,88,97,98,141,166 A syndrome distinct from multiple myeloma–associated neuropathy came to be recognized. In Iwashita et al.’s 1977 review of the literature, the 30 patients with osteosclerotic myeloma and peripheral neuropathy, as compared to the 29 patients without peripheral neuropathy, more commonly had hyperpigmentation, edema, skin thickening, hepatomegaly, hypertrichosis, and clubbing.68 In 1980, Bardwick coined the acronym POEMS to represent polyneuropathy, organomegaly, endocrinopathy, M proteins, and skin changes.9 This definition omits several important features, including sclerotic bone lesions, Castleman’s disease, papilledema, pleural effusion, edema, ascites, and thrombocytosis.82,92,104,150 Not all features are required to make a diagnosis. The “M” in POEMS should represent monoclonal plasma proliferative disorder, rather than “M protein” as originally suggested.36 In 1983, Takatsuki and Sanada reported a series of 109 Japanese cases of plasma cell dyscrasia with polyneuropathy and endocrine abnormalities.154 They described the diverse clinical manifestations of the syndrome and further separated this condition from multiple myeloma and “benign monoclonal gammopathy.”

PATHOGENESIS The pathogenesis of this multisystem disease continues to be unclear. The  light chain has been implicated because virtually all patients with the disorder have a monoclonal  plasma cell proliferative disorder. In the Mayo Clinic series, no patients with POEMS had a monoclonal  plasma cell proliferative disorder,36 and less than 5% of reported POEMS cases have been monoclonal .1,40,72,93,104,113,128,144,154 Hyperestrogenemia150 and human herpesvirus-811,12,21 infection have been proposed as unifying mechanisms for the syndrome, but without substantiation. The most appealing mechanistic hypotheses involve cytokines and vascular endothelial growth factor (VEGF).42,47,55,57,60,105,110,123,129,147 Soubrier et al.147 have shown that POEMS patients have higher levels of interleukin-1 (IL-1), tumor necrosis factor- (TNF-), and interleukin-6 (IL-6) than do classic multiple myeloma patients, as well as lower serum levels of transforming growth factor-1. Levels of IL-1 receptor antagonist and serum TNF receptor are also increased in POEMS.47,147 Elevated serum levels of IL-6,47,60,123 TNF,47,123 and IL-1 have been reported by some, but not by others.42,60,105,110,123,129 Elevated levels of IL-6 have been found in the CSF,110 pericardium,139 and ascites105 of POEMS patients. Some data suggest that levels of IL-6 correlate with disease activity.60 A pathogenic role for these cytokines is credible based on their known physiologic effects. Chronically elevated TNF- levels have been associated with inflammatory demyelinating neuropathy, organomegaly affecting liver and spleen, endocrine dysfunction, skin changes such as hypertrichosis and finger clubbing, edema, inhibition of key adipogenic enzymes, hypertriglyceridemia, and diarrhea.55 Moreover, high levels of IL-1 can cause cachexia with anorexia, sympathetic

2455

POEMS Syndrome (Osteosclerotic Myeloma)

activation of brown adipose tissue, and skin pigmentation through the activation of the propiomelanocortin gene. Overproduction of IL-1 can account for glucose intolerance and accelerated atherosclerosis.55 High IL-6 levels are known to be associated with plasma cell proliferation, thrombocytosis, Castleman’s disease, hemangiomas, and microangiopathic glomerulopathy. A major limitation of the IL-1 and TNF- hypotheses, however, is that both cytokines are potent osteoclast activating factors, which would appear to be antithetical to the osteosclerotic changes observed in POEMS patients.55 Further investigations suggest that VEGF may be a driving factor in this disorder. VEGF targets the endothelial cell, and induces a rapid and reversible increase in vascular permeability. It is important in angiogenesis, and osteogenesis is strongly dependent on angiogenesis. VEGF is expressed by osteoblasts and in bone tissue, and it could be an important regulator of osteoblastic differentiation. Both IL-1 and IL-6 have been shown to stimulate VEGF production.147 Plasma and serum levels of VEGF are markedly elevated in patients with POEMS57,147,167 and correlate with the activity of the disease.147 It has been postulated that the VEGF is secreted from the plasma cells168 and platelets,57 promoting vascular permeability, angiogenesis, and monocyte/macrophage migration, and potentially resulting in arterial obliteration. Hashiguchi et al. have demonstrated VEGF release from aggregated platelets in patients with POEMS.57 VEGF could account for the organomegaly, edema, and skin lesions. The role it would play in the polyneuropathy is less clear.168 Moreover, the VEGF hypothesis is supported by clinical observations from the finding of skin hemangiomata and mesangioproliferative changes on renal biopsy in some affected patients. One woman with POEMS who died of refractory ascites was found on autopsy to have the serosal surfaces of the bowel and the peritoneum thickened with a membrane characterized by increased numbers of proliferating capillaries, fibroblasts, lymphocytes, and plasma cells.81 Elevated levels of matrix metalloproteinases and tissue inhibitor of metalloproteinases (TIMP) have been observed in patients with POEMS. Serum levels of VEGF and TIMP-1 were strongly correlated with each other. The significance of these findings is not yet fully understood, but they may lead to a better understanding of the pathogenesis of POEMS syndrome.91 Most authors have not found monoclonal immunoglobulin associated with the nerve specimens,15,104,162 with the exception of Adams and Said1 and Broussolle et al.25 Adams and Said found M protein on the myelin sheath, in Schwann cells, and in endoneurial spaces by electron microscopy using an immunogold method indirect immunolabeling technique in three of four POEMS patients studied.1 Broussolle et al. observed a significant amount of light chain deposit on the basal lamina of some of the perineurial and endoneurial vessels.25 Microinjection of serum from several

of our patients with osteosclerotic myeloma into rat sciatic nerves failed to produce abnormal slowing of motor conduction velocity compared with that in control rats (J. J. Kelly, Jr., P. A. Low, and J. D. Schmelzer, unpublished observation, 1983). Autoantibodies against peripheral nerve antigens (SGPG and SGLPG glycolipids; GM1, GD1a, GD1b, and GT1b gangliosides) have not been demonstrated.123 Though the cytokine and angiogenesis hypotheses have not been completely validated, they offer an interesting context from which to examine the clinical features of the syndrome.

CLINICAL FEATURES The peak incidence of the POEMS syndrome is in the fifth and sixth decades of life, unlike multiple myeloma,36,40,154 which has a peak incidence in the seventh decade. Takatsuki and Sanada154 reported an average age at onset of 47 years. In the Mayo Clinic series, the average age at onset was 52 years. There also appears to be a male predominance, with a 3:2 ratio.36 Peripheral neuropathy is the dominant clinical feature of this disorder. Non-neurologic features are common (Table 109–2). Organomegaly is a frequent associated finding on examination, with hepatomegaly present in 24% to 78% of patients in the largest reported series. Splenomegaly (22% to 56%) and lymphadenopathy (26% to 69%) are also common findings. Hyperpigmentation is frequently present, but may be subtle and easily overlooked (Fig. 109–1). Hypertrichosis, characterized by the development of prominent coarse, stiff hair on the extremities and torso, must also be sought after by the physician because it may be overlooked or believed to be inconsequential by the patient. These two features on examination, hyperpigmentation and hypertrichosis, are important diagnostic clues in a patient who presents with a cryptogenic polyneuropathy. Gynecomastia and testicular atrophy may also be present. Clubbing of the fingers and toes and white nails may be prominent features. It is unclear whether the clubbing observed in association with POEMS is a function of undiagnosed pulmonary hypertension or an independent feature. The frequency of edema of the lower extremities varies between 24% and 89% according to different series.36,104,150 Ascites and pleural effusion are occasionally present,36,104,150 and tend to portend a worse outcome.36 Between 11% and 30% of POEMS patients have documented Castleman’s disease.36,104,150 This is likely a conservative estimate because many patients do not undergo lymph node biopsies. In contrast to multiple myeloma, bone pain occurs in only 15% of cases and fractures are rare. The percentage of patients with different features is related to reporting bias, and many patients will gradually develop additional features over time. In fact, as many as 25% of patients who

2456

Diseases of the Peripheral Nervous System

Table 109–2. Comparison of Clinical Characteristics of Patients in the Present Series and Two Previous Series % of Patients with Given Characteristic* Characteristic Polyneuropathy Peripheral neuropathy Cerebrospinal fluid protein 50 mg/dL Organomegaly Hepatomegaly Splenomegaly Lymphadenopathy Castleman’s disease Endocrinopathy Gonadal axis abnormality Adrenal axis abnormality Increased prolactin value Gynecomastia or galactorrhea Diabetes mellitus Hypothyroidism Hyperparathyroidism Monoclonal plasma cell dyscrasia M component on serum protein electrophoresis Skin changes Hyperpigmentation Acrocyanosis and plethora Hemangioma/telangiectasia Hypertrichosis Thickening Papilledema Extravascular volume overload Peripheral edema Ascites Pleural effusion Bone lesions Osteosclerotic only‡ Mixed sclerotic and lytic‡ Lytic only‡ Solitary lesion‡ 2 or 3 lesions‡ 3 lesions‡ Other features Weight loss 10 lb Fatigue Thrombocytosis Polycythemia Clubbing

Dispenzieri et al.36 (N  99)

Soubrier et al.150 (N  25)

Nakanishi et al.104 (N  102)

100 100† 50 24 22 26 11 67 55 16 5 18 3 14 3 100 54 68 46 19 9 26 5 29 29 24 7 3 97 47 51 2 45 23 32

100 100† — 68 52 52 24 — — — — — 36 36 — 100 100 — 48 — 32 24 28 40 — 80 32 24 68 41 59 0 41 — —

100 97† — 78 35 61 19 — — — — — 25 — — 75 70 — 93 — — 74 61 55 — 89 52 35 54 56 31 13 45 — —

37 31 54 18 5

— — 88 12 32

— — — 19 49

—  not reported. *Percentages are based on the total number of patients in the series (with the exception of cerebrospinal fluid protein, which includes the actual number tested). † Only 28, 23, and 73 patients, respectively, were tested; percentage represents percent positive of those tested. ‡ Of patients with bone lesions. From Dispenzieri, A., Kyle, R. A., Lacy, M. Q., et al.: POEMS syndrome: definitions and long-term outcome. Blood 101:2496, 2003, with permission.

POEMS Syndrome (Osteosclerotic Myeloma)

2457

FIGURE 109–1 Skin changes observed in a patient with POEMS syndrome. Left, Hyperpigmentation in patient’s hands. Right, Photograph of sister’s hands for comparison. See Color Plate

have been appropriately treated can relapse with additional new features.26,31,36,118,122,130,143,144,151,166

Polyneuropathy Symptoms of peripheral neuropathy usually precede the diagnosis of osteosclerotic myeloma by 1 or 2 years. The peripheral neuropathy begins in the lower extremities, is symmetrical, and ascends proximally. Patients often describe sensory symptoms including numbness, tingling, pins and needles, and a “cold sensation” in the feet. Sensory symptoms are followed by symmetrical weakness that begins distally over the feet with gradual and progressive proximal involvement. In Nakanishi et al.’s series, edema preceded the neuropathy in 12% of patients and coincided in 14%,104 and in the Mayo Clinic series, five patients had a known antecedent plasma cell proliferative disorder before the onset of neuropathic features.36 Though most patients have sensory impairment at onset, this is overshadowed by progressive motor impairment for the majority. The course is slowly progressive, and patients have difficulty climbing stairs and gripping objects with the hands and cannot ambulate independently. Some patients may be wheelchair bound by the time the diagnosis is established. Unlike amyloid neuropathy, autonomic symptoms are rare.72 Impotence may occur and likely is related to the endocrine dysfunction. Headache and visual symptoms such as transient obscurations of vision, scotomata, enlarged blind spots, and progressive constriction of visual fields have been described in patients with papilledema associated with the syndrome.20,24 These visual symptoms may be exacerbated or precipitated by sudden changes in posture and by fluorescent light. Physical examination shows a symmetrical sensorimotor peripheral neuropathy with sensory and motor deficits involving the extremities, with a distal predilection. Motor deficits with muscle weakness dominate the clinical pic-

ture. Severe muscle weakness with areflexia is seen in approximately one half of patients. Sensory deficits are generally present, with greater involvement of proprioception, vibration, and touch-pressure than pain and temperature modalities. Cranial nerve involvement is rare with the exception of papilledema, which is present in about 30% to 55% of reported cases.36,104,150 Ophthalmologic examination may reveal optic disc edema (usually bilateral) and blind spots.24 Cerebrospinal Fluid Patients have increased CSF protein, often greater than 100 mg/dL, and a normal cell count.36,104,150 Plasma cells are not present in the CSF. Electromyography Motor and sensory nerve conduction studies reveal moderate slowing of conduction velocities with prolonged distal latencies and reduction of compound muscle action potential (CMAP) amplitudes, especially in lower extremity nerves. In general, slowing of motor nerve conduction velocities is proportionately greater than reduction in the amplitude of the CMAP. Conduction block and temporal dispersion are rare. On needle electromyographic examination, distal fibrillation potentials and enlarged, polyphasic voluntary motor unit action potentials with decreased recruitment are found.72 The findings suggest a polyneuropathy with prominent demyelination as well as features of axonal degeneration. Sung et al.153 compared the electrodiagnostic features of 8 POEMS patients with 42 patients with chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). The patients with POEMS syndrome showed slowing of conduction velocities in the intermediate nerve segments with rare conduction block, and attenuation of the CMAPs in the lower extremities. In contrast, patients with CIDP revealed multifocal involvement of nerve segments with slowing in

2458

Diseases of the Peripheral Nervous System

proximal and distal nerve segments and a higher frequency of conduction block. Our results of the analysis of electrodiagnostic features in 84 patients with POEMS syndrome are similar to those of Sung et al. and other studies, with disproportionate involvement of lower extremity nerves, greater slowing of motor nerve conduction velocities than reduction in the amplitude of the CMAP, and rare temporal dispersion and conduction block.152 Pathologic Features In general, nerve biopsies demonstrate a combination of axonal degeneration and demyelination.72,128,161 Ohi et al. reported on a study of sural nerve biopsies of five patients with osteosclerotic myeloma and peripheral neuropathy demonstrating loss of myelinated fibers with an increased frequency of axonal degeneration on teased fibers.108 Diameter histogram analysis of myelinated fibers revealed a shift to intermediate and small myelinated fibers, suggesting fiber atrophy or fiber regeneration. There was also clustered distribution of segmental demyelination and remyelination and axonal attenuation relative to the thickness of the myelin. All of these pathologic features suggest axonal atrophy and secondary demyelination as the neuropathologic alterations of the sural nerves in this neuropathy. Umehara et al., in an autopsy case of POEMS syndrome, reported segmental demyelination and demyelination of the spinal roots with marked reduction of myelinated fibers and axonal degeneration in the sural nerve.161 Similar changes with endoneurial edema were reported by Sobue et al.145 Vital and colleagues reported the neuropathologic features of 22 patients with the syndrome.164 Superficial peroneal nerve biopsies were obtained in all cases. The main observation was the presence of uncompacted myelin lamella in 19 cases without any immunoglobulin deposition. In an additional study of five cases, uncompacted myelin lamella was associated with a reduction in the number of intra-axonal neurofilaments.165 Cellular infiltration is minimal. Individual or small foci of mononuclear cells are occasionally encountered in a perivascular location in the epineurium. Amyloid deposits are not present.16,34

Skin Changes Numerous cutaneous pathologic findings have been described (see Table 109–2) including hyperpigmentation in 32% to 93% of patients,104,150 hemangiomata (capillary and glomeruloid) in up to 26%,28,73,82,122,154,160,172 apparent skin thickening or scleroderma-like changes in 28% to 61%,104,150,159 hypertrichosis in 24% to 64%,104,150 opacity of fingernails,24,32 and Terry nails.24 Hypertrichosis may be generalized or limited to body parts. The presence of acrocyanosis and plethora is best captured in the Mayo Clinic series,36 and is described in a number of other case reports.32,104,130,150 Isolated cases

of Sweet’s syndrome and leukocytoclastic vasculitis have also been described.116,136,137,156

Endocrine Abnormalities Endocrine abnormalities are defining features of the syndrome9,154 (see Table 109–2). These abnormalities may be evident at presentation or develop through the course of the disease. Care should be taken in assigning a relation between hypothyroidism or hyperglycemia and the syndrome, because these two conditions are quite prevalent in the general population. Hyperglycemia resulting from treatment with corticosteroids should not be considered an endocrinopathy related to the syndrome in this context. Gonadal axis abnormalities such as impotence, gynecomastia, lactorrhea, low testosterone levels, and prolactinemia are not uncommon. Chemical aspects of primary9,18,22,132,150 and secondary9,132,138,150 hypothyroidism,22,132,150,151 hypogonadism,18,22,138,151 adrenocortical insufficiency,9 and diabetes mellitus9 have all been described. Hyperhidrosis may also occur. Abnormalities of the calcium parathyroid axis may also be part of the syndrome.26,39,70,83 Because most patients do not have a thorough endocrine evaluation, the true incidence of endocrinopathy in patients with POEMS is unknown. The mechanism of the multiple endocrinopathies is not understood. Endocrine glands studied at necropsy have appeared architecturally normal and without defining characteristics.54,151 Speculation that the hypophysis is the primary target of the pathologic antibodies has not been substantiated.90,119

Bone Marrow and Bone Involvement The definition of POEMS requires the presence of a monoclonal plasma cell proliferative disorder (or Castleman’s disease). Though the “M” in Bardwick’s eponym POEMS stands for “M proteins,” it should more appropriately represent monoclonal plasma cell proliferative disorder. If one looks at the two largest English language series, 75% of Nakanishi et al.’s patients had an M component104 and 85% of the Mayo Clinic series’ patients had a detectable monoclonal protein in their serum by immunofixation36 (Fig. 109–2). In the latter series, of the 84 patients with a positive immunofixation of the blood, the serum protein electrophoresis patterns of 24 patients were normal, and 7 appeared polyclonal. Had immunofixation not been done, the monoclonal protein would have been missed in these 31 patients. The quantity of monoclonal protein is small both in serum and in urine, with a median serum M protein spike of 1.1 g/dL.36 In those patients who do not have a detectable monoclonal protein in their serum or urine, a clonal plasma cell proliferative disorder should be demonstrated by immunohistochemical stains of biopsy specimens of sclerotic bone lesions and/or of bone marrow. The

POEMS Syndrome (Osteosclerotic Myeloma)

2459

FIGURE 109–2 Serum protein electrophoresis (negative study) and immunofixation (positive study).

vast majority of patients’ plasma cell clones and monoclonal light chains are ; in our experience, none of the 99 patients had a clonal  process driving their POEMS syndrome. When present, the heavy chain will be immunoglobulin A (IgA), immunoglobulin G (IgG), and occasionally immunoglobulin M (IgM) in 44%, 40%, and 1%, respectively.36 The bone marrow findings are typically not striking. Bone marrow may appear normal, hypercellular, or “reactive appearing.” The plasma cell clone can easily be missed if immunohistochemical stains, plasma cell labeling index, or flow cytometry are not used to demonstrate a -restricted plasma cell proliferation. In one series, only 14% and 4% of patients with POEMS had more than 10% and 20% bone marrow plasma cells, respectively.36 Other studies have had similar findings.104,150 Sclerotic Bone Lesions A major defining feature of POEMS is osteosclerotic bone lesions. Not all patients with osteosclerotic myeloma have POEMS, but sclerotic bone lesions are recognized in 68% to 95% of patients with POEMS (Fig. 109–3; see Table 109–2). The metastatic bone survey may have solitary (~40%) or multiple (~45% to 66%) lesions, which may be sclerotic (37% to 56%), mixed sclerotic and lytic (31% to 52%), or rarely lytic only (0 to 13%).36,104,150 Nakanishi et al. emphasized that, other than a higher incidence of splenomegaly and impotence in patients without bone

lesions as compared to those with bone lesions, there is no significant difference between the clinical features of the two groups.104 In their series of 25 patients with POEMS, Soubrier et al. reported that prognosis was significantly better for individuals with solitary bone lesions than for those with other types of plasma cell dyscrasia and was worse in the absence of bone lesions.150 In our experience, survival is independent of the number of bone lesions.36 These lesions may be commonly misdiagnosed radiologically as benign because lesions are slow growing and are often small. Sclerotic lesions may be misdiagnosed as a nonossifying fibroma, fibrous dysplasia, or aneurysmal bone cyst.

Association with Castleman’s Disease (Giant Lymph Node Hyperplasia, Angiofollicular Lymph Node Hyperplasia) Angiofollicular lymph node hyperplasia is seen in 11% to 30% of patients with POEMS.36,104,150 This is likely a conservative estimate because many patients with lymphadenopathy do not undergo lymph node biopsies. The association between Castleman’s disease and POEMS syndrome is not fully understood, though their association is well recognized.2,11–13,17,18,22,24,27–30,35,41,42,48,54,56,63,70,74,75,78,84,100,101,104,105,112, 113,121,142,150,157,164,171 Several published cases of “interesting features” associated with Castleman’s disease are likely cases of POEMS syndrome.19,46,52,59 Whether POEMS with

2460

Diseases of the Peripheral Nervous System

FIGURE 109–3 Osteosclerotic bone lesion as seen on computed tomography scan and plain radiograph.

Castleman’s disease and that with osteosclerotic myeloma are the same entity has not yet been determined.

including digital clubbing, periostosis of tubular bones, pachydermia, hyperhidrosis, and acro-osteolysis.87

Features Likely Related to POEMS

Cardiovascular Involvement Eight cases of congestive heart failure and cardiomyopathy in association with POEMS have been described.36,65,67,118,133,140 This cardiomyopathy can respond to treatment of the underlying POEMS syndrome. Both arterial and venous thromboses have been described in the setting of POEMS. In the Mayo Clinic series, 18 patients suffered serious events such as stroke, myocardial infarction, and Budd-Chiari syndrome.36 Lesprit et al. observed 4 of 20 patients to have arterial occlusion.79 Affected vessels included iliac, celiac, carotid, subclavian, mesenteric, and femoral arteries. An additional 16 patients can be found in the literature with gangrene, ischemia, myocardial infarction, splenic infarcts, and strokes.4,5,23,48,62,64,65,86,127,135,146,151,158,163,173 Vasospastic angina has been reported,71,95 and serious thrombotic events, including pulmonary embolism and Budd-Chiari syndrome, have also been described.77,170 Whether the thrombotic tendency is due to patients’ paresis,89 to the use of corticosteroids or chemotherapy, and/or elevations of proinflammatory cytokines,79 or to VEGF148 is unknown. Further study is required to determine whether this is a true association.

It is likely that restrictive lung disease, pulmonary hypertension, cardiomyopathy, thrombotic tendency, and renal insufficiency are a part of the syndrome. Greater attention to and awareness of these entities and better understanding of the pathogenesis of POEMS should clarify the question in the future.36 Pulmonary Involvement The first description of pulmonary hypertension in a POEMS patient was by Mufti et al. in 1983.99 At least 19 cases of pulmonary hypertension in POEMS patients have been described in prior series.24,67,73,80,99,107,109,111,120,123,162 Moreover, restrictive lung disease was recognized in another 5% of our patients, and 20% of deaths were related to respiratory failure.36 In another series of 20 patients with POEMS, followed over a 10-year period, 25% manifested pulmonary hypertension.80 At autopsy, classic pulmonary hypertension, including eccentric intimal fibrosis, medial hypertrophy, and marked dilatation of arteries and arterioles, has been seen.80 Improvement of POEMS-associated pulmonary hypertension after therapy has been reported.67,80,107 The levels of TNF-, IL-1, IL-6, and VEGF observed in these patients appeared to correlate with disease activity.80,107 Whether the digital clubbing seen in POEMS is a reflection of underlying pulmonary hypertension and/or parenchymal disease is yet to be determined. Martinez-Lavin et al. have drawn parallels between hypertrophic osteoarthropathy and POEMS,

Renal Involvement Renal failure in patients with POEMS is an exceedingly rare presentation. In our experience it tends to be a preterminal event, often associated with a capillary leak–like syndrome.36 Renal failure in the context of POEMS appears to be more common in patients with coexisting

POEMS Syndrome (Osteosclerotic Myeloma)

Castleman’s disease.45,50,51,76,94,96,102,103,105,106,131,149,151,155,162 Light chain deposition is observed in neither the tubules nor the glomeruli, making the renal pathology distinct from that seen in other plasma cell dyscrasias. Instead, membranoproliferative features and evidence of endothelial injury are characteristic. On both light and electron microscopy, mesangial expansion, narrowing of capillary lumina, basement membrane thickening, subendothelial deposits, widening of the subendothelial space, swelling and vacuolization of endothelial cells, and mesangiolysis predominate.94,106,131,151,155,162 Standard immunofluorescence is negative,50,162 though one author described immunoglobulin deposition within the subendothelial space of glomerular capillaries by immunoelectron microscopic localization.94 Rarely infiltration by plasma cells nests or Castleman-like lymphoma can be seen.103 IL-6 staining in the mesangium (normal) and in other cells, including the capillaries and some of the tubular cells in the interstitium, has been demonstrated.50 On electron microscopy there was deposition of amorphous material and reticulation of the mesangial matrix.50 The dominant hypothesis is that the underlying pathology of these renal findings is chronic endothelial injury.51,106

LABORATORY FEATURES Routine laboratory abnormalities are not prominent. Anemia is not characteristic. Thrombocytosis is present in at least 50% of patients, and polycythemia in nearly a fifth.36 In contrast to multiple myeloma, hypercalcemia and renal insufficiency are rare in patients with osteosclerotic myeloma. Approximately 50% of patients will have an M protein on serum protein electrophoresis, but nearly 90% will have a detectable monoclonal protein if immunofixation of serum and the urine are performed. Typically, the size of the M protein is quite small; it is greater than 2 g/dL in fewer than 10% of patients. In our experience approximately 50% of patients have an IgA and another 50% have an IgG immunoglobulin heavy chain. IgM heavy chains occur rarely. None of our 99 patients had a monoclonal  light chain.36 Abnormal levels of sensitive thyroid-stimulating hormone (thyrotropin), thyroxine, corticotropin, cortisol, fasting glucose, estradiol, testosterone, gonadotropin-stimulating hormones, prolactin, and parathyroid hormone are not uncommon. Some patients, especially those with Castleman’s disease, may have elevated C-reactive protein levels.

DIAGNOSIS Patients with osteosclerotic myeloma generally present with a sensorimotor peripheral neuropathy. Screening for the syndrome with serum protein electrophoresis alone is

2461

inadequate—immunofixation of serum and urine and roentgenographic bone survey are essential. The skeletal lesions may be small, subtle, and easily confused with benign disorders such as fibrous dysplasia. In the appropriate context, open or needle-guided biopsy of suspicious bone lesions is critical to establish the diagnosis. If the index of suspicion is high enough, bone marrow aspirate and biopsy with immunostains may also be required. The predominance of monoclonal  cannot be overemphasized because there are only a handful of convincing cases of monoclonal  POEMS syndrome in the literature.1,40,72,93,104,113,128,144,154 Skin, viscera, lymph nodes, and optic fundi should be carefully examined. A thorough endocrine evaluation should be performed, including measurements of serum thyrotropin, thyroxine, corticotropin, cortisol (in response to Cortrosyn stimulation test), glucose, glycosylated hemoglobin, prolactin, luteinizing hormone, follicle-stimulating hormone, estradiol, testosterone, and parathyroid hormone if the diagnosis is strongly suspected. If the patient has significant respiratory symptoms, pulmonary function testing and echocardiography should be performed. Decisions about the necessity of computed tomography and lymph node biopsies should be made on an individual basis. Table 109–1 includes the features associated with POEMS and a model of minimal criteria to establish the diagnosis. Two major criteria and at least one minor criterion, one of which must be either osteosclerotic myeloma or Castleman’s disease, must be satisfied to make the diagnosis of POEMS. The number of features in a given patient does not predict survival (Fig. 109–4).36

FIGURE 109–4 Survival based on number of features at presentation. (From Dispenzieri, A., Kyle, R. A., Lacy, M. Q., et al.: POEMS syndrome: definitions and long-term outcome. Blood 101:2496, 2003, with permission.)

2462

Diseases of the Peripheral Nervous System

The differential diagnosis for POEMS includes CIDP, monoclonal gammopathy of undetermined significance (MGUS)–associated neuropathy, multiple myeloma– associated neuropathy, Waldenström’s macroglobulinemia neuropathy, primary systemic amyloidosis, and cryoglobulinemia. All of these conditions, with the relative exception of CIDP, may be associated with a monoclonal protein and neuropathy. The major differences of these disorders are outlined in Table 109–3. The MGUS-associated neuropathy is a diagnosis of exclusion. Overall, the “interesting features” (sclerotic bone lesions, organomegaly, papilledema, skin changes, endocrinopathies, clubbing, thrombocytosis, and erythrocytosis) in Tables 109–1 and 109–2 should not be seen in any of these other conditions with the following exceptions. Diabetes mellitus and thyroid disorders in the context of a MGUS and a neuropathy should not automatically be considered POEMS syndrome, given their prevalence in a normal population. Patients with Waldenström’s macroglobulinemia may have lymphadenopathy and organomegaly. Anemia, renal sufficiency, and lytic bone lesions, though very common in multiple myeloma, are quite unusual in POEMS. Purpura and renal insufficiency are common in cryoglobulinemia, but not in POEMS. If there is pinch purpura, autonomic insufficiency, elevations in the alkaline phosphatase level, macroglossia, heavy proteinuria, restrictive cardiomyopathy, or carpal tunnel syndrome, one should strongly consider primary systemic amyloidosis. CIDP is another condition with clinical and laboratory similarities that should be considered in the differential diagnosis of POEMS syndrome. The neuropathy of

POEMS, with symmetrical distal and proximal weakness, areflexia, and sensory loss, resembles that seen in CIDP. However, differences in the clinical presentation, associated clinical features, laboratory findings, electrodiagnostic features, course, and response to treatment are helpful to separate the two conditions. Patients with POEMS neuropathy develop distal sensory symptoms followed by muscle weakness often affecting the feet first, with symptoms and deficits ascending up the lower extremities. Patients with CIDP may have similar onset of symptoms, but in typical cases proximal segments of the arms and legs are usually affected simultaneously. Any other “interesting clinical features,” especially skin changes, gynecomastia, and organomegaly, should alert the physician to consider POEMS syndrome as a possible diagnosis. The CSF findings are similar in both conditions and not helpful to differentiate them. The electrodiagnostic findings reveal a mixed axonal and demyelinating polyneuropathy with certain features153 helpful to separate the two conditions, but these findings may also overlap. Because of these similarities, it is important to emphasize the need to evaluate the serum and urine for monoclonal gammopathy and exclude a bone lesion with a radiographic skeletal survey. The course of the neuropathy of POEMS tends to be slowly progressive, but rapid decline is also possible. In contrast, patients with CIDP may experience spontaneous improvement or worsening unrelated to therapy. An additional feature helpful in the differential diagnosis is the response to treatment. A therapeutic trial with plasma exchange (PE) or intravenous immune globulin (IVIG) is frequently used to secure the diagnosis of CIDP. Improvement is typical of

Table 109–3. Clinical and Laboratory Features of Plasma Cell Dyscrasia–Associated Peripheral Neuropathy Characteristic Peripheral neuropathy Sensory versus motor predominance Organomegaly Skin involvement Other symptoms

Monoclonal heavy chain Monoclonal light chain Serum M spike (g/dL) BM plasma cells (%) Skeletal lesions

Thrombocytosis Anemia

MGUS ~5% Sensory, ataxia (IgM) sensorimotor   Asymptomatic IgM  IgG  IgA  75% of cases

3

10   

POEMS Syndrome

Multiple Myeloma

100% Predominantly motor Edema, fatigue, endocrine abnormalities IgG  IgA  IgM   95% of cases Usually 2 Usually 5 (sclerotic, mixed sclerotic & lytic)

1%–8% Sensory Bone pain, fatigue, infections IgG  IgA  75% of cases Usually 3 10 (lytic, osteoporotic, or fracture) 

BM  bone marrow; MGUS  monoclonal gammopathy of undetermined significance. , Absent; , rare; , occurs frequently; , almost always present.

AL Amyloidosis

Cryoglobulinemia

15%–20% Sensory, sensorimotor Fatigue, edema, cardiomyopathy, nephrosis IgG  IgA  IgM  75% of cases Usually 2 Usually 10 

~25% Predominantly sensory Purpura, arthralgia, hepatitis, nephritis

to



IgM IgG  IgA  75% of cases Usually 2 Usually 5 

POEMS Syndrome (Osteosclerotic Myeloma)

CIDP. In our experience, the neuropathy of POEMS does not improve with PE or IVIG. One should have a high level of suspicion in cases of “unresponsive CIDP” and consider the neuropathy of POEMS syndrome as an alternative diagnosis.

2463

with potential morbidity and mortality.37,61,69,115,126,169 There have been isolated reports of the benefits of interferon-,30,128 tamoxifen,10,43 and trans-retinoic acid.8

COURSE AND PROGNOSIS (SURVIVAL) TREATMENT STRATEGIES Treatment for this disorder is not standardized. For an isolated plasmacytoma, radiation is the preferred treatment.25,33,68,97,117 Response of systemic symptoms and skin changes tends to precede that of the neuropathy, with the former beginning to respond within a month and the latter within 3 to 6 months. Continued response of neuropathy to radiation can be seen even 1 to 2 years after completion of therapy. One should postpone the use of other systemic therapies in those patients with a solitary osteosclerotic lesion who have been treated with radiation. Because slow responders are not uncommon, systemic therapies should be reserved for only those patients who demonstrate progression after radiation therapy. In addition to radiation, a multitude of strategies have been employed, including plasmapheresis, IVIG, interferon-a, corticosteroids, alkylators, azathioprine, autologous stem cell transplantation, tamoxifen, and trans-retinoic acid.3,6,7,14,22,25,26,29,30,31,33,38,58,61,63,68,70,75,77,82,92,97,104,110,114,117–119, 124–126,128,131,143,151

In the Mayo Clinic series of 99 patients, 73 patients had at least some response (improvement) to therapy.36 Fourteen were refractory to the treatments chosen. Radiation therapy was an effective treatment, with at least half of patients responding as long as they had a dominant plasmacytoma. Of the 13 patients who did not respond to radiation, 9 had received less than 4000 cGy. Responses to prednisone and melphalan/prednisone occurred in about 22% to 56% of patients, respectively. Responses to alkylators with or without corticosteroids has previously been reported to be effective in some70,75,118 but not all92 patients. In our series, only those patients receiving prednisone along with plasmapheresis responded, suggesting that plasmapheresis itself is ineffective. A critical review of the literature supports our observation.143 Those reports purporting efficacy of PE combine the treatment with corticosteroids and/or other immunomodulators, confounding interpretation.7,75 No response was observed with IVIG in the 9 patients treated at Mayo Clinic.36 Again the literature supports this observation; the few reports claiming effectiveness simultaneously used other immunosuppressants or treatments.14,58,63,124 The only responses to azathioprine and cyclosporine were in conjunction with prednisone.36 To date, within the literature, 15 patients have been treated with high-dose chemotherapy with peripheral blood stem cell support. Responses have occurred in the majority of patients, but the procedure can be associated

The prognosis of POEMS had been reported to be universally poor, with estimated median survivals of 12 to 33 months,39,40,104 but this has not been our experience.36 In the large Mayo Clinic series, the median survival was 13.8 years (Fig. 109–5). Individual reports of patients with the disease for more than 5 years are not unusual,26,118,122,130,143,166 and in one French study, at least 7 of 15 patients were alive for more than 5 years, with the longest survivor alive at 25 years.55 In addition, the number of POEMS features did not affect survival. Soubrier et al. found that prognosis was no different between patients with four features versus five features,150 and the Mayo Clinic data clearly demonstrate that survival is not dependent on increasing number of features.36 Only fingernail clubbing and extravascular volume overload (i.e., effusions, edema, and ascites) were significantly associated with a shorter overall survival. Other variables, such as number of POEMS features, age, alkylator use, number of bone lesions, endocrine involvement, weight loss, lymphadenopathy, Castleman’s disease, organomegaly, papilledema, skin, gender, serum M protein, urine M protein, thrombocytosis, and hemoglobin, had no predictive value for overall survival.36 Patients who received radiation (P .04) and/or had a good quality of response to treatment (P .0001) had superior survival. Using the Bardwick criteria (five features), median survival for those patients with two, three, four, and five features at presentation was greater than 174, 85, greater than 302, and 96 months, respectively (differences were not significant). Using the nine criteria listed in the first and second columns of Table 110–1, median survival for patients with three, four, five, six,

FIGURE 109–5 Overall survival. (From Dispenzieri, A., Kyle, R. A., Lacy, M. Q., et al.: POEMS syndrome: definitions and long-term outcome. Blood 101:2496, 2003, with permission.)

2464

Diseases of the Peripheral Nervous System

seven, eight, and nine features was greater than 174, 85, greater than 302, 49, greater than 197, 103, and 50 months, respectively (differences were not significant) (Fig. 109–5). In those patients with prolonged survival, additional features may arise over time.26,31,36,82,122,130,143,144,151,166 In the Mayo Clinic series, 17 of 99 patients acquired new features of POEMS during follow-up. Another eight developed symptoms and signs that were likely related to POEMS. New features can develop more than a decade after initial presentation.36,130 On multivariate analysis, only the presence of a detectable urinary monoclonal protein by immunofixation was predictive of developing future features (proportional hazards  4.82, confidence interval 1.71 to 13.8, P  .0029).36 The most common causes of death are cardiorespiratory failure, progressive inanition, infection, capillary leak–like syndrome, and renal failure.36,104 Even those patients with multiple bone lesions or those with more than 10% plasma cells do not progress to classic multiple myeloma. The renal failure observed is different from that of classic myeloma (i.e., cast nephropathy).45,50,51,76,94,96,102,105,106,131,151,155,162 Typically those patients who die with renal failure have coexistent ascites and a capillary leak–like syndrome. Patients rarely, if ever, develop bone fractures or bone pain. The neuropathy may be unrelenting and contribute to progressive inanition and eventual cardiorespiratory failure and pneumonia. Stroke and myocardial infarction, which may or may not be related to the POEMS syndrome, are also observed causes of death. Patients do not die of classic myeloma, that is, progressive bone marrow failure or hypercalcemia or pathologic fractures.

7.

8.

9.

10.

11.

12.

13.

14.

REFERENCES 15. 1. Adams, D., and Said, G.: Ultrastructural characterisation of the M protein in nerve biopsy of patients with POEMS syndrome. J. Neurol. Neurosurg. Psychiatry 64:809, 1998. 2. Adelman, H. M., Cacciatore, M. L., Pascual, J. F., et al.: Case report: Castleman disease in association with POEMS. Am. J. Med. Sci. 307:112, 1994. 3. Ako, S., Kaneko, Y., Higuchi, M., et al.: Crow-Fukase syndrome—immunoadsorption plasmapheresis effectively lowers elevated interleukin-6 concentration. Nephrol. Dial. Transplant. 14:419, 1999. 4. Amiel, J., Machover, D., and Droz, J.: Plasma cell dyscrasia with arteriopathy, polyneuropathy, and endocrine syndrome: a Japanese disease in an Italian patient [in French]. Ann. Med. Interne (Paris) 126:745, 1975. 5. Anonymous: Case Records of the Massachusetts General Hospital. Weekly Clinicopathological Exercises. Case 10–1987: a 59-year-old woman with progressive polyneuropathy and monoclonal gammopathy. N. Engl. J. Med. 316:606, 1987. 6. Antela Lopez, A., Requena Caballero, I., Masa Vazquez, L., et al.: Polyneuropathy associated with osteosclerotic

16.

17.

18.

19.

20.

21.

myeloma: excellent response to chemotherapy [in Spanish]. Neurologia 5:102, 1990. Atsumi, T., Kato, K., Kurosawa, S., et al.: A case of CrowFukase syndrome with elevated soluble interleukin-6 receptor in cerebrospinal fluid: response to double-filtration plasmapheresis and corticosteroids. Acta Haematol. 94:90, 1995. Authier, F. J., Belec, L., Levy, Y., et al.: All-trans-retinoic acid in POEMS syndrome: therapeutic effect associated with decreased circulating levels of proinflammatory cytokines. Arthritis Rheum. 39:1423, 1996. Bardwick, P. A., Zvaifler, N. J., Gill, G. N., et al.: Plasma cell dyscrasia with polyneuropathy, organomegaly, endocrinopathy, M protein, and skin changes: the POEMS syndrome. Report on two cases and a review of the literature. Medicine 59:311, 1980. Barrier, J. H., Le Noan, H., Mussini, J. M., and Brisseau, J. M.: Stabilisation of a severe case of P.O.E.M.S. syndrome after tamoxifen administration [letter]. J. Neurol. Neurosurg. Psychiatry 52:286, 1989. Belec, L., Authier, F. J., Mohamed, A. S., et al.: Antibodies to human herpesvirus 8 in POEMS (polyneuropathy, organomegaly, endocrinopathy, M protein, skin changes) syndrome with multicentric Castleman’s disease. Clin. Infect. Dis. 28:678, 1999. Belec, L., Mohamed, A. S., Authier, F. J., et al.: Human herpesvirus 8 infection in patients with POEMS syndromeassociated multicentric Castleman’s disease. Blood 93:3643, 1999. Belec, L., Salmon-Ceron, D., Blanche, P., et al.: POEMS syndrome (Polyneuropathy, Organomegaly, Endocrinopathy, Monoclonal gammopathy, Skin changes), multicentric Castleman disease, renal chromophobe carcinoma and herpes virus type 8 infection [letter; comment] [in French]. Ann. Pathol. 19:373, 1999. Benito-Leon, J., Lopez-Rios, F., Rodriguez-Martin, F. J., et al.: Rapidly deteriorating polyneuropathy associated with osteosclerotic myeloma responsive to intravenous immunoglobulin and radiotherapy. J. Neurol. Sci. 158:113, 1998. Bergouignan, F. X., Massonnat, R., Vital, C., et al.: Uncompacted lamellae in three patients with POEMS syndrome. Eur. Neurol. 27:173, 1987. Besinger, U. A., Toyka, K. V., Anzil, A. P., et al.: Myeloma neuropathy: passive transfer from man to mouse. Science 213:1027, 1981. Bhatia, M., and Maheshwari, M. C.: Angiofollicular lymphoid hyperplasia presenting as POEMS syndrome. J. Assoc. Physicians India 42:751, 1994. Bitter, M., Komaiko, W., and Franklin, W.: Giant lymph node hyperplasia with osteoblastic bone lesions and the POEMS (Takatsuki’s) syndrome. Cancer 56:188, 1985. Black, D. A., Forgacs, I., Davies, D. R., and Thompson, R. P.: Pseudotumour cerebri in a patient with Castleman’s disease. Postgrad. Med. J. 64:217, 1988. Bolling, J. P., and Brazis, P. W.: Optic disk swelling with peripheral neuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes (POEMS syndrome). Am. J. Ophthalmol. 109:503, 1990. Bosch, E. P., and Smith, B. E.: Peripheral neuropathies associated with monoclonal proteins. Med. Clin. North Am. 77:125, 1993.

POEMS Syndrome (Osteosclerotic Myeloma) 22. Bosco, J., and Pathmanathan, R.: POEMS syndrome, osteosclerotic myeloma and Castleman’s disease: a case report. Aust. N. Z. J. Med. 21:454, 1991. 23. Bova, G., Pasqui, A. L., Saletti, M., et al.: POEMS syndrome with vascular lesions: a role for interleukin-1beta and interleukin-6 increase—a case report. Angiology 49:937, 1998. 24. Brazis, P. W., Liesegang, T. J., Bolling, J. P., et al.: When do optic disc edema and peripheral neuropathy constitute poetry? Surv. Ophthalmol. 35:219, 1990. 25. Broussolle, E., Vighetto, A., Bancel, B., et al.: P.O.E.M.S. syndrome with complete recovery after treatment of a solitary plasmocytoma. Clin. Neurol. Neurosurg. 93:165, 1991. 26. Cabezas-Agricola, J. M., Lado-Abeal, J. J., Otero-Anton, E., et al.: Hypoparathyroidism in POEMS syndrome [letter]. Lancet 347:701, 1996. 27. Carcaterra, A., Santini, R., Sozzi, G., and Zuccoli, E.: CrowFukase syndrome (POEMS syndrome): the first Italian presentation of a case and review of the literature [in Italian]. G. Ital. Dermatol. Venereol. 125:97, 1990. 28. Chan, J. K., Fletcher, C. D., Hicklin, G. A., and Rosai, J.: Glomeruloid hemangioma: a distinctive cutaneous lesion of multicentric Castleman’s disease associated with POEMS syndrome. Am. J. Surg. Pathol. 14:1036, 1990. 29. Chang, Y. J., Huang, C. C., and Chu, C. C.: Intravenous immunoglobulin therapy in POEMS syndrome: a case report. Chung Hua Hsueh Tsa Chih 58:366, 1996. 30. Coto, V., Auletta, M., Oliviero, U., et al.: POEMS syndrome: an Italian case with diagnostic and therapeutic implications. Ann. Ital. Med. Int. 6:416, 1991. 31. Crisci, C., Barbieri, F., Parente, D., et al.: POEMS syndrome: follow-up study of a case. Clin. Neurol. Neurosurg. 94:65, 1992. 32. Crow, R. Peripheral neuritis in myelomatosis. Br. Med. J. 2:802, 1956. 33. Davis, L., and Drachman, D.: Myeloma neuropathy. Arch. Neurol. 27:507, 1972. 34. Dayan, A. D., and Stokes, M. I.: Peripheral neuropathy and experimental myeloma in the mouse. Nat. New Biol. 236:117, 1972. 35. Del Rio, R., Alsina, M., Monteagudo, J., et al.: POEMS syndrome and multiple angioproliferative lesions mimicking generalized histiocytomas. Acta Derm. Venereol. 74:388, 1994. 36. Dispenzieri, A., Kyle, R. A., Lacy, M. Q., et al.: POEMS syndrome: definitions and long-term outcome. Blood 101:2496, 2003. 37. Dispenzieri, A., Lacy, M. Q., Litzow, M. R., et al.: Peripheral blood stem cell transplant (PBSCT) in patients with POEMS syndrome is associated with significant morbidity and mortality. Blood 98:5437a, 2001. 38. Donofrio, P. D., Albers, J. W., Greenberg, H. S., and Mitchell, B. S.: Peripheral neuropathy in osteosclerotic myeloma: clinical and electrodiagnostic improvement with chemotherapy. Muscle Nerve 7:137, 1984. 39. Driedger, H., and Pruzanski, W.: Plasma cell neoplasia with osteosclerotic lesions: a study of five cases and a review of the literature. Arch. Intern. Med. 139:892, 1979. 40. Driedger, H., and Pruzanski, W.: Plasma cell neoplasia with peripheral polyneuropathy: a study of five cases and a review of the literature. Medicine 59:301, 1980.

2465

41. Dworak, O., Tschubel, K., Zhou, H., and Meybehm, M.: Angiofollicular lymphatic hyperplasia with plasmacytoma and polyneuropathy: a case report with immunohistochemical study [in German]. Klin. Wochenschr. 66:591, 1988. 42. Emile, C., Danon, F., Fermand, J. P., and Clauvel, J. P.: Castleman disease in POEMS syndrome with elevated interleukin-6 [letter; comment]. Cancer 71:874, 1993. 43. Enevoldson, T. P., and Harding, A. E.: Improvement in the POEMS syndrome after administration of tamoxifen [letter]. J. Neurol. Neurosurg. Psychiatry 55:71, 1992. 44. Evison, G., and Evans, K. T.: Sclerotic bone deposits in multiple myeloma [letter]. Br. J. Radiol. 56:145, 1983. 45. Fam, A. G., Rubenstein, J. D., and Cowan, D. H.: POEMS syndrome: study of a patient with proteinuria, microangiopathic glomerulopathy, and renal enlargement. Arthritis Rheum. 29:233, 1986. 46. Feigert, J. M., Sweet, D. L., Coleman, M., et al.: Multicentric angiofollicular lymph node hyperplasia with peripheral neuropathy, pseudotumor cerebri, IgA dysproteinemia, and thrombocytosis in women: a distinct syndrome [see comments]. Ann. Intern. Med. 113:362, 1990. 47. Feinberg, L., Temple, D., de Marchena, E., et al.: Soluble immune mediators in POEMS syndrome with pulmonary hypertension: case report and review of the literature. Crit. Rev. Oncog. 10:293, 1999. 48. Forster, A., and Muri, R.: Recurrent cerebrovascular insult—manifestation of POEMS syndrome? [in German]. Schweiz. Med. Wochenschr. 128:1059, 1998. 49. Fukase, M., Kakimatsu, T., Nishitani, H., et al.: Report of a case of solitary plasmacytoma in the abdomen presenting polyneuropathy and endocrinological disorders [abstract]. Clin. Neurol. (Tokyo) 9:657, 1969. 50. Fukatsu, A., Ito, Y., Yuzawa, Y., et al.: A case of POEMS syndrome showing elevated serum interleukin 6 and abnormal expression of interleukin 6 in the kidney. Nephron 62:47, 1992. 51. Fukatsu, A., Tamai, H., Nishikawa, K., et al.: The kidney disease of Crow-Fukase (POEMS) syndrome: a clinicopathological study of four cases. Clin. Nephrol. 36:76, 1991. 52. Gaba, A. R., Stein, R. S., Sweet, D. L., and Variakojis, D.: Multicentric giant lymph node hyperplasia. Am. J. Clin. Pathol. 69:86, 1978. 53. Getaz, P., Handler, L., Jacobs, P., and Tunley, I.: Osteosclerotic myeloma with peripheral neuropathy. S. Afr. Med. J. 48:1246, 1974. 54. Gherardi, R., Baudrimont, M., Kujas, M., et al.: Pathological findings in three non-Japanese patients with the POEMS syndrome. Virchows Arch. A Pathol. Anat. Histopathol. 413:357, 1988. 55. Gherardi, R. K., Belec, L., Soubrier, M., et al.: Overproduction of proinflammatory cytokines imbalanced by their antagonists in POEMS syndrome. Blood 87:1458, 1996. 56. Gherardi, R. K., Malapert, D., and Degos, J. D.: Castleman disease-POEMS syndrome overlap [letter; comment]. Ann. Intern. Med. 114:520, 1991. 57. Hashiguchi, T., Arimura, K., Matsumuro, K., et al.: Highly concentrated vascular endothelial growth factor in platelets in Crow-Fukase syndrome. Muscle Nerve 23:1051, 2000.

2466

Diseases of the Peripheral Nervous System

58. Henze, T., and Krieger, G.: Combined high-dose 7S-IgG and dexamethasone is effective in severe polyneuropathy of the POEMS syndrome [letter] [see comments]. J. Neurol. 242:482, 1995. 59. Hineman, V. L., Phyliky, R. L., and Banks, P. M.: Angiofollicular lymph node hyperplasia and peripheral neuropathy: association with monoclonal gammopathy. Mayo Clin. Proc. 57:379, 1982. 60. Hitoshi, S., Suzuki, K., and Sakuta, M.: Elevated serum interleukin-6 in POEMS syndrome reflects the activity of the disease. Intern. Med. 33:583, 1994. 61. Hogan, W. J., Lacy, M. Q., Wiseman, G. A., et al.: Successful treatment of POEMS syndrome with autologous hematopoietic progenitor cell transplantation. Bone Marrow Transplant. 28:305, 2001. 62. Hori, T., Tsuboi, Y., Okubo, R., et al.: Crow-Fukase syndrome associated with Castleman disease showing hypertrophic cranial pachymeningitis and bilateral internal carotid artery occlusion [in Japanese]. Rinsho Shinkeigaku 39:456, 1999. 63. Huang, C. C., and Chu, C. C.: Poor response to intravenous immunoglobulin therapy in patients with Castleman’s disease and the POEMS syndrome [letter; comment]. J. Neurol. 243:726, 1996. 64. Huth, M., Gordon, D., Verrier, E. D., and Otto, C. M.: Aortic valvular fibroma as a source of systemic emboli in POEMS syndrome. J. Am. Soc. Echocardiogr. 4:401, 1991. 65. Ichikawa, Y., Nakata, T., Ohhata, J., et al.: Crow-Fukase syndrome with ischemic cardiomyopathy. Intern. Med. 40:726, 2001. 66. Imawari, M., Akatsuka, N., Ishibashi, M., et al.: Syndrome of plasma cell dyscrasia, polyneuropathy and endocrine disturbances: report of a case. Ann. Intern. Med. 81:490, 1974. 67. Iwasaki, H., Ogawa, K., Yoshida, H., et al.: Crow-Fukase syndrome associated with pulmonary hypertension. Intern. Med. 32:556, 1993. 68. Iwashita, H., Ohnishi, A., Asada, M., et al.: Polyneuropathy, skin hyperpigmentation, edema, and hypertrichosis in localized osteosclerotic myeloma. Neurology 27:675, 1977. 69. Jaccard, A., Royer, B., Bordessoule, D., et al.: High-dose therapy and autologous blood stem cell transplantation in POEMS syndrome. Blood 99:3057, 2002. 70. Judge, M. R., McGibbon, D. H., and Thompson, R. P.: Angioendotheliomatosis associated with Castleman’s lymphoma and POEMS syndrome. Clin. Exp. Dermatol. 18:360, 1993. 71. Kato, T., Kaneko, E., Numano, F., et al.: Vasospastic angina in Crow-Fukase syndrome. Am. Heart J. 124:505, 1992. 72. Kelly, J. J. Jr., Kyle, R. A., Miles, J. M., and Dyck, P. J.: Osteosclerotic myeloma and peripheral neuropathy. Neurology 33:202, 1983. 73. Kishimoto, S., Takenaka, H., Shibagaki, R., et al.: Glomeruloid hemangioma in POEMS syndrome shows two different immunophenotypic endothelial cells. J. Cutan. Pathol. 27:87, 2000. 74. Kobayashi, H., Ii, K., Sano, T., et al.: Plasma-cell dyscrasia with polyneuropathy and endocrine disorders associated with dysfunction of salivary glands. Am. J. Surg. Pathol. 9:759, 1985. 75. Ku, A., Lachmann, E., Tunkel, R., and Nagler, W.: Severe polyneuropathy: initial manifestation of Castleman’s disease

76.

77.

78.

79.

80.

81.

82. 83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

associated with POEMS syndrome. Arch. Phys. Med. Rehabil. 76:692, 1995. Kubota, T., Yoshizawa, N., Takeuchi, A., and Niwa, H.: Microangiopathic hypothesis of POEMS tissue lesions [letter]. Clin. Nephrol. 38:115, 1992. Kuwabara, S., Hattori, T., Shimoe, Y., and Kamitsukasa, I.: Long term melphalan-prednisolone chemotherapy for POEMS syndrome. J. Neurol. Neurosurg. Psychiatry 63:385, 1997. Lapresle, J., Lacroix-Ciaudo, C., Reynes, M., and Madoule, P.: Crow-Fukase syndrome (POEMS syndrome) and osseous mastocytosis secondary to Castleman’s angiofollicular lymphoid hyperplasia [in French]. Rev. Neurol. (Paris) 142:731, 1986. Lesprit, P., Authier, F. J., Gherardi, R., et al.: Acute arterial obliteration: a new feature of the POEMS syndrome? Medicine 75:226, 1996. Lesprit, P., Godeau, B., Authier, F. J., et al.: Pulmonary hypertension in POEMS syndrome: a new feature mediated by cytokines. Am. J. Respir. Crit. Care Med. 157(3 Pt. 1):907, 1998. Loeb, J. M., Hauger, P. H., Carney, J. D., and Cooper, A. D.: Refractory ascites due to POEMS syndrome. Gastroenterology 96:247, 1989. Longo, G., Emilia, G., and Torelli, U.: Skin changes in POEMS syndrome. Haematologica 84:86, 1999. Ludescher, C., Grunewald, K., Fend, F., et al.: Osteosclerotic myeloma with polyneuropathy and hypocalcemia. Blut 58:207, 1989. Mandler, R. N., Kerrigan, D. P., Smart, J., et al.: Castleman’s disease in POEMS syndrome with elevated interleukin-6 [see comments]. Cancer 69:2697, 1992. Mangalik, A., and Veliath, A. J.: Osteosclerotic myeloma and peripheral neuropathy: a case report. Cancer 28:1040, 1971. Manning, W. J., Goldberger, A. L., Drews, R. E., et al.: POEMS syndrome with myocardial infarction: observations concerning pathogenesis and review of the literature. Semin. Arthritis Rheum. 22:151, 1992. Martinez-Lavin, M., Vargas, A. S., Cabre, J., et al.: Features of hypertrophic osteoarthropathy in patients with POEMS syndrome: a metaanalysis [letter]. J. Rheumatol. 24:2267, 1997. Mayo, C. M., Daniels, A., and Barron, K. D.: Polyneuropathy in the osteosclerotic form of multiple myeloma. Trans. Am. Neurol. Assoc. 93:240, 1968. McKinley, W., Jackson, A., Cardenas, D., and DeVivo, M.: Long-term medical complications after traumatic spinal cord injury: a regional model systems analysis. Arch. Phys. Med. Rehabil. 80:1402, 1999. Meier, C., Reulecke, M., Kesselring, J., et al.: Polyneuropathy, organomegaly, endocrinopathy and skin changes in a case of solitary myeloma [in German]. Schweiz. Med. Wochenschr. 116:1326, 1986. Michizono, K., Umehara, F., Hashiguchi, T., et al.: Circulating levels of MMP-1, -2, -3, -9, and TIMP-1 are increased in POEMS syndrome. Neurology 56:807, 2001. Milanov, I., and Georgiev, D.: Polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy and skin changes (POEMS) syndrome. Can. J. Neurol. Sci. 21:60, 1994.

POEMS Syndrome (Osteosclerotic Myeloma) 93. Miralles, G. D., O’Fallon, J. R., and Talley, N. J.: Plasmacell dyscrasia with polyneuropathy: the spectrum of POEMS syndrome. N. Engl. J. Med. 327:1919, 1992. 94. Mizuiri, S., Mitsuo, K., Sakai, K., et al.: Renal involvement in POEMS syndrome. Nephron 59:153, 1991. 95. Mochizuki, Y., Yoshihashi, H., Oishi, M., et al.: CrowFukase syndrome: a case associated with vasospastic angina. Acta Neurol. 16:170, 1994. 96. Modesto-Segonds, A., Rey, J. P., Orfila, C., et al.: Renal involvement in POEMS syndrome. Clin. Nephrol. 43:342, 1995. 97. Morley, J. B., and Schwieger, A. C.: The relation between chronic polyneuropathy and osteosclerotic myeloma. J. Neurol. Neurosurg. Psychiatry 30:432, 1967. 98. Moya-Mir, M. S., Martin-Martin, F., Barbadillo, R., et al.: Plasma cell dyscrasia with polyneuritis and dermatoendocrine alterations: report of a new case outside Japan. Postgrad. Med. J. 56:427, 1980. 99. Mufti, G. J., Hamblin, T. J., and Gordon, J.: Melphalaninduced pulmonary fibrosis in osteosclerotic myeloma [letter]. Acta Haematol. 69:140, 1983. 100. Munoz, G., Geijo, P., Moldenhauer, F., et al.: Plasmacellular Castleman’s disease and POEMS syndrome. Histopathology 17:172, 1990. 101. Myers, B. M., Miralles, G. D., Taylor, C. A., et al.: POEMS syndrome with idiopathic flushing mimicking carcinoid syndrome. Am. J. Med. 90:646, 1991. 102. Nakamoto, Y., Imai, H., Yasuda, T., et al.: Mesangiolytic glomerulonephritis associated with Takatsuki’s syndrome: an analysis of five biopsy specimens. Hum. Pathol. 20:243, 1989. 103. Nakamoto, Y., Imai, H., Yasuda, T., et al.: A spectrum of clinicopathological features of nephropathy associated with POEMS syndrome. Nephrol. Dial. Transplant. 14:2370, 1999. 104. Nakanishi, T., Sobue, I., Toyokura, Y., et al.: The CrowFukase syndrome: a study of 102 cases in Japan. Neurology 34:712, 1984. 105. Nakazawa, K., Itoh, N., Shigematsu, H., and Koh, C. S.: An autopsy case of Crow-Fukase (POEMS) syndrome with a high level of IL-6 in the ascites: special reference to glomerular lesions. Acta Pathol. Jpn. 42:651, 1992. 106. Navis, G. J., Dullaart, R. P., Vellenga, E., et al.: Renal disease in POEMS syndrome: report on a case and review of the literature. Nephrol. Dial. Transplant. 9:1477, 1994. 107. Niimi, H., Arimura, K., Jonosono, M., et al.: VEGF is causative for pulmonary hypertension in a patient with Crow- Fukase (POEMS) syndrome. Intern. Med. 39:1101, 2000. 108. Ohi, T., Kyle, R. A., and Dyck, P. J.: Axonal attenuation and secondary segmental demyelination in myeloma neuropathies. Ann. Neurol. 17:255, 1985. 109. Okura, H., Gohma, I., Hatta, K., and Imanaka, T.: Thiamine deficiency and pulmonary hypertension in Crow-Fukase syndrome. Intern. Med. 34:674, 1995. 110. Orefice, G., Morra, V. B., De Michele, G., et al.: POEMS syndrome: clinical, pathological and immunological study of a case. Neurol. Res. 16:477, 1994. 111. Paciocco, G., Bossone, E., Erba, H., and Rubenfire, M.: Reversible pulmonary hypertension in POEMS syn-

112.

113.

114.

115.

116. 117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127. 128.

2467

drome—another etiology of triggered pulmonary vasculopathy? Can. J. Cardiol. 16:1007, 2000. Papo, T., Soubrier, M., Marcelin, A. G., et al.: Human herpesvirus 8 infection, Castleman’s disease and POEMS syndrome [letter]. Br. J. Haematol. 104:932, 1999. Pareyson, D., Marazzi, R., Confalonieri, P., et al.: The POEMS syndrome: report of six cases. Ital. J. Neurol. Sci. 15:353, 1994. Parra, R., Fernandez, J. M., Garcia-Bragado, F., et al.: Successful treatment of peripheral neuropathy with chemotherapy in osteosclerotic myeloma. J. Neurol. 234:261, 1987. Peggs, K. S., Paneesha, S., Kottaridis, P. D., et al.: Peripheral blood stem cell transplantation for POEMS syndrome. Bone Marrow Transplant. 30:401, 2002. Perniciaro, C.: POEMS syndrome. Semin. Dermatol. 14:162, 1995. Philips, E. D., el-Mahdi, A. M., Humphrey, R. L., and Furlong, M. B. Jr.: The effect of the radiation treatment on the polyneuropathy of multiple myeloma. J. Can. Assoc. Radiol. 23:103, 1972. Reitan, J. B., Pape, E., Fossa, S. D., et al.: Osteosclerotic myeloma with polyneuropathy. Acta Med. Scand. 208:137, 1980. Reulecke, M., Dumas, M., and Meier, C.: Specific antibody activity against neuroendocrine tissue in a case of POEMS syndrome with IgG gammopathy. Neurology 38:614, 1988. Ribadeau-Dumas, S., Tillie-Leblond, I., Rose, C., et al.: Pulmonary hypertension associated with POEMS syndrome. Eur. Respir. J. 9:1760, 1996. Rolon, P. G., Audouin, J., Diebold, J., et al.: Multicentric angiofollicular lymph node hyperplasia associated with a solitary osteolytic costal IgG lambda myeloma: POEMS syndrome in a South American (Paraguayan) patient. Pathol. Res. Pract. 185:468, 1989. Rongioletti, F., Gambini, C., and Lerza, R.: Glomeruloid hemangioma: a cutaneous marker of POEMS syndrome. Am. J. Dermatopathol. 16:175, 1994. Rose, C., Zandecki, M., Copin, M. C., et al.: POEMS syndrome: report on six patients with unusual clinical signs, elevated levels of cytokines, macrophage involvement and chromosomal aberrations of bone marrow plasma cells. Leukemia 11:1318, 1997. Rotta, F. T., and Bradley, W. G.: Marked improvement of severe polyneuropathy associated with multifocal osteosclerotic myeloma following surgery, radiation, and chemotherapy. Muscle Nerve 20:1035, 1997. Rousseau, J. J., Franck, G., Grisar, T., et al.: Osteosclerotic myeloma with polyneuropathy and ectopic secretion of calcitonin. Eur. J. Cancer 14:133, 1978. Rovira, M., Carreras, E., Blade, J., et al.: Dramatic improvement of POEMS syndrome following autologous haematopoietic cell transplantation. Br. J. Haematol. 115:373, 2001. Rushton, D.: Peripheral sensorimotor neuropathy associated with a localized myeloma. Br. Med. J. 2:203, 1965. Saida, K., Kawakami, H., Ohta, M., and Iwamura, K.: Coagulation and vascular abnormalities in Crow-Fukase syndrome. Muscle Nerve 20:486, 1997.

2468

Diseases of the Peripheral Nervous System

129. Saida, K., Ohta, M., Kawakami, H., and Saida, T.: Cytokines and myelin antibodies in Crow-Fukase syndrome. Muscle Nerve 19:1620, 1996. 130. Sakemi, H., and Okada, H.: An autopsy case of CrowFukase syndrome which developed 18 years after the first manifestation of plasmacytoma. Intern. Med. 31:50, 1992. 131. Sano, M., Terasaki, T., Koyama, A., et al.: Glomerular lesions associated with the Crow-Fukase syndrome. Virchows Arch. A Pathol. Anat. Histopathol. 409:3, 1986. 132. Sasano, T., Sakurai, S. I., and Hara, Y.: Improvement in gonadotropin secretion after corticosteroid therapy in a case of POEMS syndrome. Endocr. J. 45:413, 1998. 133. Sato, H., Kawamata, T., Kanaya, N., and Namiki, A.: Anesthetic management of a patient with Crow-Fukase syndrome complicated with severe heart failure [in Japanese]. Masui 50:552, 2001. 134. Scheinker, L.: Myelom und Nervensystem: uber eine bisher nicht beschrieben mit eigentumlichen Hautveranderungen einhergehende Polyneuritis bei einem plamazellularen Myelom des Sternums. Dtsch. Z. Nervenheilkd. 147:247, 1938. 135. Semble, E. L., Challa, V. R., Holt, D. A., and Pisko, E. J.: Light and electron microscopic findings in POEMS, or Japanese multisystem syndrome. Arthritis Rheum. 29:286, 1986. 136. Sharabi, Y., Raanani, P., Shenkar, A., et al.: Plasma cell dyscrasia with polyneuropathy—POEMS syndrome presenting with vasculitic skin lesions and responding to combination chemotherapy. Leuk. Lymphoma 40:209, 2000. 137. Shibata, M., Yamada, T., Tanahashi, N., et al.: POEMS syndrome with necrotizing vasculitis: a novel feature of vascular abnormalities. Neurology 54:772, 2000. 138. Shichiri, M., Iwashina, M., Imai, T., et al.: Abnormal FSH hypersecretion as an endocrinological manifestation of POEMS syndrome. Endocr. J. 45:131, 1998. 139. Shikama, N., Isono, A., Otsuka, Y., et al.: A case of POEMS syndrome with high concentrations of interleukin-6 in pericardial fluid. J. Intern. Med. 250:170, 2001. 140. Shimizu, N., Goya, M., Akimoto, H., et al.: Cardiomyopathy in a case of Crow-Fukase syndrome. Jpn. Heart J. 38:877, 1997. 141. Shimpo, S., Nishitani, H., and Tsunermura, T.: Solitary plasmacytoma with polyneuritis and endocrine disturbance [in Japanese]. Nippon Rinsho 26:2444, 1968. 142. Shirabe, S., Kishikawa, M., Mine, M., et al.: Crow-Fukase syndrome associated with extramedullary plasmacytoma. Jpn. J. Med. 30:64, 1991. 143. Silberstein, L. E., Duggan, D., and Berkman, E. M.: Therapeutic trial of plasma exchange in osteosclerotic myeloma associated with the POEMS syndrome. J. Clin. Apheresis 2:253, 1985. 144. Simmons, Z., Wald, J., Albers, J. W., and Feldman, E. L.: The natural history of a “benign” rib lesion in a patient with a demyelinating polyneuropathy and an unusual variant of POEMS syndrome. Muscle Nerve 17:1055, 1994. 145. Sobue, G., Doyu, M., Watanabe, M., et al.: Extensive demyelinating changes in the peripheral nerves of CrowFukase syndrome: a pathological study of one autopsied case. Acta Neuropathol. 84:171, 1992.

146. Solomons, R., and Gibbs, D.: Plasma cell dyscrasia with polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy and skin changes: the POEMS syndrome. J. R. Soc. Med. 75:553, 1982. 147. Soubrier, M., Dubost, J. J., Serre, A. F., et al.: Growth factors in POEMS syndrome: evidence for a marked increase in circulating vascular endothelial growth factor. Arthritis Rheum. 40:786, 1997. 148. Soubrier, M., Guillon, R., Dubost, J. J., et al.: Arterial obliteration in POEMS syndrome: possible role of vascular endothelial growth factor. J. Rheumatol. 25:813, 1998. 149. Soubrier, M., Sauron, C., Souweine, B., et al.: Growth factors and proinflammatory cytokines in the renal involvement of POEMS syndrome. Am. J. Kidney Dis. 34:633, 1999. 150. Soubrier, M. J., Dubost J. J., and Sauvezie, B. J.: POEMS syndrome: a study of 25 cases and a review of the literature. French Study Group on POEMS Syndrome. Am. J. Med. 97:543, 1994. 151. Stewart, P. M., McIntyre, M. A., and Edwards, C. R.: The endocrinopathy of POEMS syndrome. Scott. Med. J. 34:520, 1989. 152. Suarez, G. A., Dispenzieri, A., Gertz, M. A., and Kyle, R. A.: The electrophysiologic findings of the peripheral neuropathy associated with POEMS. Clin. Neurophysiol. 113(Suppl. 1):S9, 2002. 153. Sung, J. Y., Kuwabara, S., Ogawara, K., et al.: Patterns of nerve conduction abnormalities in POEMS syndrome. Muscle Nerve 26:189, 2002. 154. Takatsuki, K., and Sanada, I.: Plasma cell dyscrasia with polyneuropathy and endocrine disorder: clinical and laboratory features of 109 reported cases. Jpn. J. Clin. Oncol. 13:543, 1983. 155. Takazoe, K., Shimada, T., Kawamura, T., et al.: Possible mechanism of progressive renal failure in Crow-Fukase syndrome [letter]. Clin. Nephrol. 47:66, 1997. 156. Tal, A., and Jones, R. K.: POEMS syndrome in association with Sweet’s syndrome—a new variant? J. Tenn. Med. Assoc. 86:145, 1993. 157. Thajeb, P., Chee, C. Y., Lo, S. F., and Lee, N.: The POEMS syndrome among Chinese: association with Castleman’s disease and some immunological abnormalities. Acta Neurol. Scand. 80:492, 1989. 158. Tobin, M., and Fitzgerald, M.: The Japanese plasma cell dyscrasia syndrome: case report and theory of pathogenesis. Postgrad. Med. J. 58:786, 1982. 159. Toussaint, P., Sibaud, V., Labbe, L., and Geniaux, M.: POEMS syndrome revealed by a scleroderma-like skin thickening [in French]. Ann. Dermatol. Venereol. 127:73, 2000. 160. Tsai, C. Y., Lai, C. H., Chan, H. L., and Kuo, T.: Glomeruloid hemangioma—a specific cutaneous marker of POEMS syndrome. Int. J. Dermatol. 40:403, 2001. 161. Umehara, F., Izumo, S., Zyounosono, M., and Osame, M.: An autopsied case of the Crow-Fukase syndrome: a neuropathological study with emphasis on spinal roots. Acta Neuropathol. 80:563, 1990. 162. Viard, J. P., Lesavre, P., Boitard, C., et al.: POEMS syndrome presenting as systemic sclerosis: clinical and pathologic study of a case with microangiopathic glomerular lesions. Am. J. Med. 84(3 Pt. 1):524, 1988.

POEMS Syndrome (Osteosclerotic Myeloma) 163. Virmani, R., Popovsky, M., and Roberts, W.: Thrombocytosis, coronary thrombosis and acute myocardial infarction. Am. J. Med. 67:498, 1979. 164. Vital, C., Gherardi, R., Vital, A., et al.: Uncompacted myelin lamellae in polyneuropathy, organomegaly, endocrinopathy, M-protein and skin changes syndrome: ultrastructural study of peripheral nerve biopsy from 22 patients. Acta Neuropathol. 87:302, 1994. 165. Vital, C., Vital, A., Ferrer, X., et al.: Crow-Fukase (POEMS) syndrome: a study of peripheral nerve biopsy in five new cases. J. Peripher. Nerv. Syst. 8:136, 2003. 166. Waldenstrom, J. G., Adner, A., Gydell, K., and Zettervall, O.: Osteosclerotic “plasmocytoma” with polyneuropathy, hypertrichosis and diabetes. Acta Med. Scand. 203:297, 1978. 167. Watanabe, O., Arimura, K., Kitajima, I., et al.: Greatly raised vascular endothelial growth factor (VEGF) in POEMS syndrome [letter]. Lancet 347:702, 1996. 168. Watanabe, O., Maruyama, I., Arimura, K., et al.: Overproduction of vascular endothelial growth factor/vascular

169.

170.

171.

172.

173.

2469

permeability factor is causative in Crow-Fukase (POEMS) syndrome. Muscle Nerve 21:1390, 1998. Wiesmann, A., Weissert, R., Kanz, L., and Einsele, H.: Long-term follow-up on a patient with incomplete POEMS syndrome undergoing high-dose therapy and autologous blood stem cell transplantation. Blood 100:2679, 2002. Wong, V. A., and Wade, N. K.: POEMS syndrome: an unusual cause of bilateral optic disk swelling. Am. J. Ophthalmol. 126:452, 1998. Yang, S. G., Cho, K. H., Bang, Y. J., and Kim, C. W.: A case of glomeruloid hemangioma associated with multicentric Castleman’s disease. Am. J. Dermatopathol. 20:266, 1998. Zea-Mendoza, A. C., Alonso-Ruiz, A., Garcia-Vadillo, A., et al.: POEMS syndrome with neuroarthropathy and nodular regenerative hyperplasia of the liver. Arthritis Rheum. 27:1053, 1984. Zenone, T., Bastion, Y., Salles, G., et al.: POEMS syndrome, arterial thrombosis and thrombocythaemia. J. Intern. Med. 240:107, 1996.

110 Paraneoplastic Neuropathy JUDITH M. SPIES AND JAMES G. MCLEOD

Historical Background Incidence Types of Neuropathy Subacute Sensory Neuronopathy Sensorimotor Neuropathy

Demyelinating Neuropathy Vasculitic Neuropathy Motor Neuropathy Small Fiber Sensory Neuropathy Autonomic neuropathy

Paraneoplastic neuropathies are a remote effect of cancer, not related to treatment or deficiency states. They may occur in isolation, but more commonly there is concomitant involvement of both the central and peripheral nervous systems. In the peripheral nervous system neuronal cell bodies (sensory, motor, or autonomic neurons), the peripheral nerves, neuromuscular junction, or muscle may be affected. Mild peripheral neuropathy is very common in the later stages of malignancy. Clinically significant acute or subacute paraneoplastic neuropathies that occur before or soon after the diagnosis of malignancy are far less common and are frequently part of complex syndromes involving both central and peripheral nervous systems. Although there is abundant evidence to indicate that the paraneoplastic neuropathies represent disordered autoimmunity, the exact pathogenetic mechanisms are yet to be elucidated.

HISTORICAL BACKGROUND Several authors reported an association between malignancy and peripheral neuropathy, independent of the presence of metastases or direct invasion, in the early 20th century. However, it was not until Denny-Brown’s description in 1948 of sensory neuropathy resulting from degeneration of dorsal root ganglion (DRG) cells in two patients with carcinoma of the lung that neuropathy as a remote or paraneoplastic effect of malignancy was clearly recognized.44 Subsequently a variety of disorders of nerve and muscle were recognized and grouped under the label “carcinomatous neuromyopathy.”21 A group of workers at the London Hospital extensively

Immunopathogenesis Detection of Underlying Malignancy Treatment

characterized the clinical features, prevalence, and prognosis of paraneoplastic neuropathies in the 1950s and 1960s,20,21,34–37,75,158 recognizing that sensory neuropathy and subacute cerebellar degeneration constituted well-defined clinicopathologic states but that there was “a loose linkage of various clinical syndromes in which lesions may occur at almost any level of the nervous system and also in muscle.”75 The variable distribution of pathologic lesions giving rise to various clinical syndromes led Croft and Wilkinson36 to comment that “as more examples of neuromyopathy are seen, it becomes apparent that any type of neuromyopathy may occur with any type of malignant disease,” although the frequent association of sensory neuronopathy and encephalomyelitis, particularly in patients with small cell lung cancer (SCLC), was established.74,76 Wilkinson first described nervous system–specific antibodies in the serum of patients with cancer-associated sensory neuropathy in 1964.159 In the mid-1980s investigators at Memorial Sloan-Kettering Cancer Center described antibodies in the serum and cerebrospinal fluid (CSF) of patients with SCLC and sensory neuropathy that were subsequently shown to recognize specific antigens expressed on neural cells and the underlying tumor.61,63 A number of such nervous system–specific autoantibodies have now been described in a variety of paraneoplastic neurologic diseases. These antibodies were initially detected by indirect immunofluorescence and described according to their staining pattern, but as antibody specificities and target antigens have been better characterized and recombinant target proteins have been developed, a nomenclature based on the antigenic target has developed (Table 110–1). Both nomenclature 2471

2472

Diseases of the Peripheral Nervous System

Table 110–1. Nervous System–Specific Antibodies Present in Paraneoplastic Neurologic Disease Neuropathy

Antibody

Associated Malignancies

Subacute sensory neuronopathy

anti-Hu/ANNA-1 (~80%)

Sensorimotor neuropathy

antiamphiphysin anti-CV2 anti-Hu/ANNA-1 anti-Yo/PCA-1

SCLC (in 70%–80%), occasionally with wide variety of other carcinomas, including breast, prostate, gastrointestinal tract, thymoma, other neuroendocrine SCLC, breast, others Breast, colon, prostate, thymoma, sarcoma, neuroendocrine, others As above, but predominance of SCLC less marked

Autonomic neuropathy

Motor neuronopathy Demyelinating neuropathy

anti-CV2 anti-Ri/ANNA-2 anti-Hu /ANNA-1 anti–ganglionic acetylcholine receptor (30%) anti-Hu/ANNA-1 anti-Yo/PCA-1 anti-Hu/ANNA-1 antiganglioside

Breast, ovarian carcinoma Breast, colon, prostate, thymoma, sarcoma, neuroendocrine, etc. Lung, breast, others SCLC Thymoma, SCLC, bladder, gastrointestinal tract

Wide variety of carcinomas (renal cell, lung, breast), thymoma Breast, ovarian, other Wide variety of carcinomas Malignant melanoma

systems remain in current use, so that the antibody originally described by Graus and colleagues is designated as either antineuronal nuclear antibody type 1 (ANNA-1) or anti-Hu antibody.2,42,63,92 The association between anti-Hu antibodies, subacute sensory neuronopathy/paraneoplastic encephalomyelitis (SSN/PEM), and SCLC, and that between anti–Purkinje cell antibodies (PCA-1 or anti-Yo) and paraneoplastic cerebellar degeneration and breast or gynecologic malignancies, are the best characterized. In the majority of patients the nervous system involvement in paraneoplastic disorders is multifocal. Although there are strong associations such as those mentioned above, it is becoming clearer that in many instances antibody specificity is not absolutely predictive of either neurologic illness or tumor type, and that there are true paraneoplastic neuropathies that are antibody negative.

INCIDENCE A wide variety of solid tumors as well as hematologic malignancies (see Chapters 109, 111, and 115) have been associated with paraneoplastic neuropathies. Peripheral neuropathy is most common in carcinoma of the lung, but occurs with a wide variety of other carcinomas, including breast, prostate, stomach, and colon, and in other solid tissue tumors such as melanoma. It is difficult to be certain of the true frequency of peripheral neuropathy in malignancy because it depends on the pathologic type and site of the tumor, the stage and duration of illness, the techniques employed for detection of the neuropathy, and the criteria for diagnosis. The reported frequency of clinical neuropathy in lung cancer ranges from 1.7% to 16%.21,34–36,49,142 If

electrophysiologic studies, quantitative sensation testing, or histopathology is used to detect neuropathy, the incidence is significantly higher. However, in most of these cases the neuropathy is mild, appears after the diagnosis of the malignancy, and is more common in patients with significant weight loss. The paraneoplastic neuropathies that occur early, often before diagnosis of the tumor, can be more debilitating than the malignancy and are much less common. Using electrophysiologic criteria, Moody106 found evidence of neuropathy in 14 of 82 patients (17%) with a variety of malignancies; two thirds of these were sensorimotor neuropathies. In a prospective study of 71 patients with SCLC, Hawley and co-workers73 found no increased incidence of peripheral neuropathy in the initial stages of the illness but evidence of a mild peripheral neuropathy in all patients who had lost 15% or more of their body weight. Graus and colleagues64 found clinical evidence of a mild peripheral neuropathy in 8 of 62 patients (13%) with lung cancer; with nerve conduction studies the incidence increased to 18%. The incidence of neuropathy was strongly correlated with degree of weight loss. Thirteen percent of patients with less than 15% weight loss, compared with 50% of those who had lost greater than 15% of their body weight, had evidence of neuropathy, and 45% of the patients with neuropathy had lost more than 15% of their body weight. If mild changes indicative of denervation on electromyography (EMG) or biopsy are considered, the incidence is higher again. Hildebrand and Coers77 reported electrophysiologic or histologic evidence of neuropathy in 19 of 46 patients (41%) with carcinomas of the lung, breast, gastrointestinal tract, ovary, and uterus. None of these patients reported sensory symptoms and only two had predominantly distal weakness on examination. The most common clinical finding was

2473

Paraneoplastic Neuropathy

generalized muscle wasting, which the authors related to the degree of weight loss. Histologic evidence of denervation was present in 63% of patients with weight loss and only 15% of those without. Similarly, Trojaborg and colleagues142 found EMG evidence of denervation in 17 of 47 patients (36%) with lung cancer and no clinical evidence of a neuromuscular disorder, and Lenman and co-workers91 found spontaneous activity consistent with mild denervation in 35% of patients. Performing quantitative sensation testing on 29 patients with a variety of malignancies and no other identifiable risk factors for neuropathy, Lipton and colleagues99 were able to demonstrate elevated thermal or vibration thresholds in 50% and 38%, respectively, leading the authors to suggest that small fiber neuropathy may be more commonly associated with cancer than had been previously recognized. Clinically significant acute or subacute neuropathy occurring before or soon after tumor diagnosis is less common. In a prospective study of 150 consecutive patients with SCLC, Elrington and colleagues49 found only one patient with SSN. In that series, 16% of patients had some sensory loss; 9% had reduced reflexes and 11% had postural hypotension, consistent with mild somatic or autonomic peripheral neuropathy, respectively. Van Oosterhout and colleagues144 assessed 203 consecutive patients with SCLC and found peripheral neuropathy with no identifiable cause in 5 (2.5%). A further five patients had well-defined paraneoplastic neurologic syndromes (two cases of Lambert-Eaton myasthenic syndrome, two of limbic encephalitis, and one of subacute cerebellar degeneration) associated with serum antineuronal antibodies. There is less information regarding the incidence of neuropathy in breast and gynecologic malignancies. Croft and Wilkinson35 found evidence of “neuromyopathy” in 4.4% of 250 patients with breast cancer, suggesting an incidence three to four times less than that in carcinoma of the lung. Cavaletti and colleagues30 found clinical evidence of neuropathy in none of 12 and 2 of 46 patients (4%) with stage I or III ovarian cancer prior to chemotherapy. Electrophysiologic evidence of neuropathy was present in 33% and 72%, respectively. The prevalence of malignancy associated neuropathy in patients with peripheral neuropathy is not accurately known. Windebank and colleagues160 followed 42 patients with acute or subacute clinically pure sensory neuropathy for 2 to 35 years and found malignancy in 2 patients who died of metastatic cancer 11 and 19 years after the onset of neuropathy, suggesting a coincidental association. Similarly, Notermans and co-workers110 found no underlying malignancy in their study of 75 patients with neuropathy of undetermined cause. Lin and colleagues98 reviewed 520 cases of neuropathy and found that 2.3% were associated with malignancy. Camerlingo and colleagues28 searched for occult malignancy in 51 patients with sensory neuropathy and no identifiable cause. Malignancy was diagnosed in 35%, 3 to 72 months after the diagnosis of neuropathy (mean 27.9 months). Patients

with monoclonal gammopathy of uncertain significance but no reactivity to myelin-associated glycoprotein were included in the idiopathic group, which may in part explain the higher frequency of associated malignancy in this study. Antoine and co-workers9 found 26 patients with carcinoma but no infiltration, drug toxicity, or cachexia among 422 patients with neuropathy (6%). Twenty-one (5%) of these patients had either antineuronal antibodies or a short interval between onset of neuropathy and diagnosis of malignancy, or a clinical course consistent with a paraneoplastic etiology. The remaining five patients had a long interval between onset of neuropathy and malignancy and a chronic progressive course suggesting a coincidental association. In this same study, 47% of patients with SSN, 10% with multiple mononeuropathy or chronic inflammatory demyelinating polyradiculoneuropathy (CIDP), 4.5% with axonal polyneuropathy, and only 1.7% of patients presenting with Guillain-Barré syndrome had an associated carcinoma. SCLC is the underlying malignancy in 50% to 75% of patients with paraneoplastic neuropathy; this association is particularly strong for SSN and least true for remitting-relapsing sensorimotor neuropathy and motor neuronopathy.21,36,37,41,100,130

TYPES OF NEUROPATHY Subacute Sensory Neuronopathy SSN is the best characterized of the paraneoplastic neuropathies. Denny-Brown’s original description was of sensory neuronopathy as the only remote manifestation of lung cancer in two patients,44 but it soon became evident that sensory neuronopathy often occurs as part of more widespread involvement of both peripheral and central nervous systems. In a clinicopathologic review of 69 patients with paraneoplastic neurologic disease, Henson and Urich recognized that sensory neuronopathy and encephalomyelitis frequently developed together, particularly in association with SCLC.76 With the detection of specific antineuronal antibodies in the serum of patients with paraneoplastic encephalitis or sensory neuronopathy,3,41,61,63 it was recognized that these were different manifestations of the same underlying disease process. These anti-Hu or ANNA-1 antibodies react with a family of protein antigens (HuD, HuC, HuR, and Hel-N1) of 35 to 40 kDa present in neurons and some tumors, especially SCLC. These proteins show significant identity with neural specific proteins with RNA binding domains in Drosophila, and it is postulated that they are involved in regulation of cell proliferation and regulate messenger RNAs encoding protein products involved in neuronal development and function.117 All antiHu sera react with HuD and HuC antigens, but only HuD is present in associated tumors, indicating this is the important

2474

Diseases of the Peripheral Nervous System

protein in triggering the immune response.102 In normal tissues expression of Hu antigens is highly restricted to central and peripheral neurons as well as chromaffin cells of the adrenal and ganglion cells of the bronchus.38 Immunoreactivity is predominantly in the nuclei, sparing the nucleoli and with weaker staining of the cytoplasm61 (Fig. 110–1). All SCLCs and approximately 50% of neuroblastomas express the Hu antigen, as do some Merkel cell tumors and sarcomas and occasional melanomas, non–small cell lung tumors, and other carcinomas.25,38,40,63,66 Although all SCLCs express the antigen, only a small percentage of patients develop anti-Hu antibodies.25,39 Although SSN/PEM is strongly associated with anti-Hu antibodies and SCLC, it occurs with a wide variety of malignancies and is associated with a number of autoantibodies (see below). SSN is the most common manifestation of antiHu–associated paraneoplastic syndromes, but in most cases there is evidence of more widespread nervous system pathology with coexistent limbic or brainstem encephalitis, autonomic failure, cerebellar degeneration, or other features. Clinical Features Sensory loss, paresthesia, and dysesthesia may occur alone, or more typically, in combination; pain is often a prominent feature.27,31,44 Proprioceptive loss is a constant feature, leading to sensory ataxia, and pseudoathetosis of the hands and

FIGURE 110–1 Anti-Hu antibody binding to trigeminal ganglion cells. The nuclei of neurons are strongly positive, with sparing of the nucleoli. (From Graus, F., Cordon-Cardo, C., and Posner, J. B.: Neuronal antinuclear antibody in sensory neuronopathy from lung cancer. Neurology 35:538, 1985, with permission.)

fingers is common in severe cases. Moderate to severe loss of pain sensation is seen in the majority.31 Typically, reflexes are globally absent. Strength is preserved unless there is more widespread neurologic involvement. The sensory disturbance usually starts distally but the onset is often asymmetrical. The earliest symptoms may be in proximal limbs, trunk, or face and may predominantly involve the arms.27,31,60,67 Loss of taste and hearing may be evident.67 Like other paraneoplastic syndromes, SSN usually presents in middle or older age, with a mean age of onset of 59.6 years (range 45 to 73 years) in one representative series.31 In several large series, women have been affected more commonly than men, although the most common underlying neoplasm is significantly more common in males.31,37,41,66,80 The course of the neuropathy is typically subacute, with onset over a few weeks to several months, although occasional patients have an indolent course with slow progression over many months to years or a very acute course with onset over a few days.27,31,41,60,130 A mild or indolent course may be more common in patients with SSN without coincidental central nervous system (CNS) involvement130 (see below). Arrest or remission of the neuropathy after a variable period of progression is characteristic, although this often occurs only after severe disability has developed.75,130 Dalmau and colleagues41 found that sensory symptoms were the presenting complaint in 59% of their 71 patients with anti-Hu–associated SSN/PEM, and SSN developed over the course of the illness in 74%. In 62% of patients sensory neuropathy remained the dominant clinical syndrome, and in one third of these was the only manifestation. In their series of 26 patients with paraneoplastic sensory neuronopathy, Chalk and colleagues reported evidence of more widespread neurologic illness in 58%.31 Several other studies have subsequently confirmed that, although sensory neuronopathy is the predominant clinical manifestation on presentation in approximately 50% of patients with anti-Hu paraneoplastic neurologic syndromes, one half to three quarters of these patients develop more widespread neurologic involvement, including encephalitis, autonomic neuropathy, cerebellar degeneration, and Lambert-Eaton myasthenic syndrome.27,41,66,71,100,130 Several different forms of neuropathy may be present in an individual patient, most commonly sensory neuronopathy and enteric neuropathy or sensory neuronopathy and sensorimotor neuropathy. The onset of neuropathy precedes diagnosis of cancer in the majority of cases (71% to 88%), with a median interval from onset of symptoms to tumor diagnosis of 3.5 to 6 months (range 1 to 47 months) in various series.31,37,41,66,130 The 3-month, 1-year, and 3-year survival rates after onset of neurologic symptoms have been reported as 90%, 56%, and 23%, respectively.130 Age above 60 years, neurologic disability score at diagnosis, number of areas of the nervous system affected, presence of an underlying malignancy, and treatment predict neurologic outcome and survival.66,130

Paraneoplastic Neuropathy

Underlying Malignancy Several large series31,41,66,100,130 have found that 85% of patients with sensory neuronopathy/encephalomyelitis and high-titer anti-Hu antibodies have a detectable underlying neoplasm, most commonly SCLC (in approximately 70% to 80%). However, a wide variety of cancers, including breast, prostate, ovarian, endometrial and undefined adenocarcinomas, neuroblastoma, and germ cell tumors, have been reported.12,41,66,100,104,130 In approximately 15% of patients with otherwise typical anti-Hu–associated SSN/PEM, no tumor is detected despite extensive investigation. Some studies have suggested that a higher proportion of cases without detectable cancer, particularly over a prolonged follow-up period, have symptoms limited to the peripheral nervous system.66,130 Characteristically the paraneoplastic syndrome develops while the tumor is still small, often making detection difficult. The extent of disease is limited to the chest in 76% to 96% of SCLC associated with SSN/PEM.41,100,130 Laboratory Investigations CSF protein is elevated in approximately 80% of cases, and a pleocytosis may be present.27,31,67 The latter is more common if there is coincident CNS involvement but can be seen with isolated peripheral neuropathy. Oligoclonal bands of immunoglobulin may be detected in the CSF. Erythrocyte sedimentation rate and other systemic markers are usually normal.31 Autoantibodies. The most helpful laboratory investigation is the detection of antibodies directed against neural antigens in serum and/or CSF. The most characteristic of these are antibodies directed against neuronal nuclear antigens, anti-Hu or ANNA-1 antibodies. The presence of these antibodies not only indicates the paraneoplastic nature of the neuropathy but also in significant degree predicts the nature of the underlying tumor. However, the absence of anti-Hu antibodies does not entirely rule out a paraneoplastic etiology of the neuropathy; a variety of other antibodies have subsequently been described in association with SSN/PEM and a significant percentage of cases may not have detectable autoantibodies. Molinuevo and colleagues104 detected anti-Hu antibodies in 40 of 49 patients (82%) with paraneoplastic sensory neuropathy and 1 of 25 patients (4%) with idiopathic sensory neuropathy, yielding a specificity of 96% and sensitivity of 82%. The anti-Hu–negative cases were less likely to have involvement of other areas of the nervous system, and had a lower frequency of SCLC (44% compared to 79%) compared to the anti-Hu–positive cases. Care is needed in the interpretation of anti-Hu or other autoantibody reactivity. On both immunohistochemistry and Western blotting, false positives may be detected when substrates other than purified human Purkinje cells or purified recombinant Hu protein are used. Anti-Hu reactivity has been reported in the sera of patients with

2475

Sjögren’s syndrome105; however, on more detailed testing, two of the three positive sera did not react with purified recombinant Hu protein, nor did the sera of 13 other patients with Sjögren’s syndrome.131 Anti–Purkinje cell antibodies, without anti-Yo immunoreactivity, have also been described in patients with SSN with no evidence of tumor on prolonged follow-up.108 Antibodies against the synaptic vesicle protein amphiphysin were first described in paraneoplastic stiff-man syndrome and breast cancer but have subsequently been demonstrated in a variety of paraneoplastic neurologic disorders, including SSN and SCLC or breast carcinoma with or without coincident anti-Hu antibodies.5,47,94,123 AntiCV2 (collapsin response mediator protein-5 [CRMP-5]) antibodies are associated with cerebellar degeneration, uveitis, and peripheral neuropathy and are seen with a wide variety of neoplasms, including SCLC, prostate and colon carcinoma, uterine sarcoma, thymoma, and undifferentiated neuroendocrine tumors.7,8,78,164 These antibodies recognize a 66-kDa protein belonging to the Unc-33–like phosphoprotein (Ulip)/CRMP protein family involved in axonal growth and guidance.26,58,154 In immunohistochemical studies they bind to sensory neurons and myelinated and unmyelinated peripheral nerve axons.7 Although the neuropathy is more commonly sensorimotor (see below) unless there is coincident anti-Hu antibody positivity,7 anti-CV2 seropositivity has been reported in cases of sensory neuropathy associated with SCLC, uterine sarcoma, and undifferentiated carcinoma in the absence of anti-Hu antibodies.78,164 Differential Diagnosis The main differential diagnoses of paraneoplastic sensory neuronopathy are sensory neuronopathy associated with Sjögren’s syndrome and idiopathic sensory neuronopathy. The presence of sicca symptoms or characteristic serum autoantibodies are the clue to the former, although neither may be evident at initial presentation. The incidence of the paraneoplastic and idiopathic forms is probably similar.31,160 The evolution is characteristically subacute in both. Rapid progression, the presence of increased CSF protein, prominent pain, coexistent autonomic or CNS involvement, and the presence of specific antineuronal antibodies should prompt a search for underlying malignancy. Electrophysiology Characteristically, there is a marked disparity between sensory and motor studies, with absent or markedly reduced sensory action potential amplitudes and relatively normal motor conduction.31,46,89 Compound muscle action potential amplitudes and motor conduction velocities can be mildly reduced, especially in severe cases. Needle EMG is characteristically normal in patients with isolated SSN, but mild denervation changes may be evident late in the disease or when there is more widespread neurologic illness.31,89,120

2476

Diseases of the Peripheral Nervous System

Donofrio and colleagues have reported the evolution of electrophysiologic abnormalities in a typical case.46 Studies done 1 month after the onset of symptoms showed no abnormality. The earliest abnormality was absent H reflexes at day 39, followed by absent or reduced sensory amplitudes at day 68 and absent sensory responses with mild slowing of motor conduction by 3 to 4 months. Needle EMG remained normal until the final study 31/2 months after onset of symptoms, when sparse denervation potentials were recorded. The mild motor abnormalities late in the course have been attributed to the development of a subclinical sensorimotor neuropathy or disuse effect. Motor nerves may be subclinically involved more often than was previously thought. Camdessanche and co-workers27 reviewed 20 patients with anti-Hu antibodies and peripheral neuropathy. Although 14 (70%) had a clinically pure sensory neuropathy, only 4 (29%) of these had an electrophysiologic pattern typical of sensory neuronopathy, with abnormal motor conduction studies in most patients indicating that pathology often extends beyond the DRGs and sensory nerves. Pathology The characteristic findings in the DRGs are inflammation and neuron somal loss with formation of residual nodules or nodules of Nageotte (clusters of supporting cells).44,75,80 There is often significant variation in the degree of neuronal loss at different spinal segmental levels81 (Fig. 110–2). There is secondary wallerian degeneration of sensory nerves, dorsal roots, and dorsal columns. Sural nerve biopsies show severe but nonspecific loss of myelinated fibers and axonal degeneration with occasional mononuclear cell cuffing.31,120 The majority of cells in the DRG infiltrates are T lymphocytes, predominantly CD3- and CD8-positive cells, with scattered CD68-positive macrophages.68,155 Perivascular infiltrates contain CD20-positive B cells, T cells, and monocytes/macrophages.83,155 Anti-Hu immunoglobulin (IgG) deposits can be detected in and around neurons in the DRGs,22,63,68 with limited correlation between the principal clinical symptoms, areas of major tissue injury, and quantitative anti-Hu IgG distribution.40 The content of anti-Hu IgG in the DRGs is greater than that of serum in the majority of cases studied.40 Treatment Treatment is generally disappointing. Antitumor therapy is more effective than immunodulatory therapy in stabilizing the disease (see below).

Sensorimotor Neuropathy Sensorimotor neuropathy represents approximately 25% of neuropathies associated with anti-Hu antibodies, although there is some variability between series depending on

whether clinical or electrophysiologic criteria are used to determine motor involvement, since subclinical mild motor electrophysiologic abnormalities are common in clinically pure sensory neuropathy.27,66,100,130 The onset is frequently asymmetrical, and monomelic presentation has been reported. Sensory or motor symptoms may predominate.27 Hu proteins are not expressed in peripheral nerve, raising the possibility that other nerve proteins are the targets of the paraneoplastic immune response. As mentioned above, peripheral neuropathy occurs in approximately 50% of patients with anti-CV2–associated paraneoplastic syndromes.7,78,164 Antoine and co-workers7 reported the clinical and electrophysiologic features of neuropathy in nine patients with CV2 antibodies. They found sensorimotor neuropathy in all six patients who had only anti-CV2 antibodies and in one of three patients with both anti-CV2 and anti-Hu antibodies. The remaining two patients with both antibodies had a pure sensory neuropathy. The onset of the sensorimotor neuropathy was typically subacute and symmetrical, but in one third of patients was asymmetrical. The disability was usually mild except in one patient with a severe motor-predominant neuropathy. In approximately half the patients positive for anti-CV2 antibodies, electrophysiology showed a pattern the authors call mixed axonal and demyelinating, with motor nerve conduction velocities slower than expected for the reduction in compound muscle action potential amplitude without reaching the demyelinating range. Demyelinating features were evident at biopsy or autopsy in two of four cases studied. In the patients with only anti-CV2 antibodies, sensory and motor cell bodies were unaffected in the two cases in which they were studied, whereas in the case with both anti-Hu and anti-CV2 antibodies both sensory ganglionitis typical of anti-Hu–associated SSN and lesions in peripheral nerves similar to those in the other CV2-positive patients were evident. Other series have suggested a higher frequency of sensory neuropathy, but in the largest of these there is no information on coincident antibodies and the neuropathies are not reported in any detail.78,164 As in the anti-Hu paraneoplastic syndromes, most patients with anti-CV2 antibodies had multifocal nervous system involvement. Neuropathy was the only feature at presentation in one third; however, in all but one other neurologic features such as cerebellar degeneration, uveitis, autonomic dysfunction, and Lambert-Eaton myasthenic syndrome developed.7 Subacute sensorimotor neuropathies, especially if associated with cerebellar dysfunction or uveitis, therefore warrant further investigation for anti-CV2 antibodies and an underlying neoplasm. Although type 2 antineuronal nuclear (ANNA-2 or anti-Ri) antibodies are typically associated with opsoclonusmyoclonus and cerebellar ataxia,101 sensorimotor neuropathy is also seen with these antibodies. Pittock and colleagues reported neuropathy in 7 of 34 patients with anti-Ri antibodies, in association with lung, breast, or other tumors.118

2477

Paraneoplastic Neuropathy

FIGURE 110–2 Sections of dorsal root ganglia (A–C) and posterior roots (D–F) at L3 (A, D), L4 (B, E), and S1 (C, F) from a patient with paraneoplastic sensory neuronopathy. There is loss of neurons and residual nodules of Nageotte (arrows) in the ganglia and fiber loss in the corresponding posterior roots. The severity of the changes varies between the spinal levels, but the extent of fiber loss in posterior roots corresponds to the degree of neuronal loss at each level. Bars: A–C, 50 ␮m; D–F, 20 ␮m. (From Ichimura, M., Yamamoto, M., Kobayashi, Y., et al.: Tissue distribution of pathological lesions and Hu antigen expression in paraneoplastic sensory neuronopathy. Acta Neuropathol. 95:641, 1998, with permission.)

PCA-1 (anti-Yo) antibodies are characteristically associated with breast and gynecologic neoplasms and paraneoplastic cerebellar degeneration,70 but there are reports of sensorimotor neuropathy as the predominant clinical syndrome in patients with breast cancer and anti-Yo antibodies.88,103 These antibodies do bind to DRG homogenates.108 Sensorimotor neuropathy has also been reported in association with antiamphiphysin antibodies, although a

SSN/PEM presentation is more typical and, in the sensorimotor case reported, autopsy showed coincident typical pathologic features of PEM/SSN.123

Demyelinating Neuropathy Although demyelinating neuropathies are more commonly associated with Hodgkin’s disease and other lymphoproliferative diseases (see Chapter 111), demyelinating neuropathy

2478

Diseases of the Peripheral Nervous System

does occur in association with carcinoma, and early series of sensorimotor neuropathy associated with carcinoma included cases with prominent evidence of demyelination associated with a wide variety of neoplasms.37,51,127 Croft and Wikinson described 13 cases of remitting and relapsing sensorimotor neuropathy in their series of patients with carcinomatous neuromyopathy.37 The two cases in which electrodiagnostic studies were reported demonstrated markedly reduced motor conduction velocities, consistent with a demyelinating neuropathy, and prominent demyelination was evident pathologically in several of these relapsing cases as well as in a number of patients in their acute and subacute sensorimotor neuropathy group.34 Both acute and subacute or chronic presentations have been reported. An acute Guillain-Barré–like presentation with ascending paralysis, respiratory and bulbar weakness, and dysautonomia has been reported with a variety of underlying neoplasms,34,72,161 including a manifestation of anti-Hu–associated paraneoplastic syndrome.27 Antoine and co-workers10 reported a concomitant carcinoma in 3 of 33 consecutive patients with CIDP. In all patients electrophysiologic studies showed decreased motor conduction velocity and either increased distal motor latency or conduction block, consistent with demyelination. In one patient, with cholangiocarcinoma, an immunoglobulin M kappa monoclonal protein was detected. Nerve biopsy was performed in two patients and showed perivascular mononuclear cell infiltrates, thin myelin sheaths, some demyelination on teased fiber preparation, varying degrees of axonal degeneration, and onion bulbs in one. In all three, the diagnosis of CIDP preceded recognition of the carcinoma by several months. One patient improved with intravenous immune globulin (IVIG) and one with corticosteroids before antitumor therapy, and the third improved after surgery to remove the tumor. Reports that carcinomas can express antigens shared by Schwann cells13,90 suggest one possible pathogenetic mechanism for these neuropathies. CIDP has also been reported in association with malignant melanoma.15,157 Because melanoma cells express a number of glycoconjugate antigens also present in peripheral nerve, it is postulated that this association is based on molecular mimicry.109,157 This hypothesis is supported by the occurrence of demyelinating neuropathy after immunotherapy with melanoma cell lysates or monoclonal antibody against melanoma-derived antigens.55,124

Vasculitic Neuropathy Occasionally the sensorimotor neuropathy associated with carcinoma is found to be due to vasculitis.17,84,111,152 Recognition of this rare form of paraneoplastic neuropathy is important, because it appears to respond to treatment more consistently than other forms. Torvik and Berntzen141 initially reported the association between vasculitis restricted to nerve and muscle and malignancy in a single patient found to have a small renal cell carcinoma at

autopsy. Johnson and colleagues84 subsequently reported two patients with clinically unrecognized SCLC who presented with mononeuritis multiplex and progressed rapidly to death within several months, and a third case with a 2-year history of asymmetrical polyneuropathy subsequently found to have Hodgkin’s disease. Nerve biopsy in all three cases showed a predominantly epineurial microvasculitis, leading the authors to propose that vasculitic neuropathy represented a new paraneoplastic neurologic syndrome. In both the carcinoma-associated cases there was pathologic evidence of associated limbic and/or brainstem encephalitis and sensory neuronopathy, but the vasculitic neuropathy was the dominant clinical feature. Clinical Features Only 25% of patients with paraneoplastic vasculitic neuropathy present with a multiple mononeuropathy pattern, approximately 30% with an asymmetrical polyneuropathy and 40% with a symmetrical polyneuropathy. Although usually presenting with an axonal sensorimotor neuropathy, histologic evidence of peripheral nerve vasculitis has been reported in patients with predominantly sensory neuropathy and anti-Hu antibodies or with a demyelinating picture, and there may be evidence of coincident vasculitic neuropathy in cases in which other clinical features, such as autonomic failure, encephalitis, or Lambert-Eaton myasthenic syndrome, predominate.4,48,84 Laboratory Investigations Cerebrospinal fluid protein is elevated in 90% of cases,84,111,152 usually without an increase in cells; however, a mononuclear pleocytosis was present in cases found to have encephalitis at autopsy.84 Systemic inflammatory markers such as the erythrocyte sedimentation rate or C-reactive protein are elevated in approximately 80% of patients.111,152 Electrophysiology reveals prominent reduction in amplitudes with only mild slowing of conduction and evidence of active denervation on electromyography.84,113 Conduction block may be present. Pathology The characteristic pathologic features are a marked fiber loss, active axonal degeneration, and a mononuclear inflammatory cell infiltrate surrounding and invading the walls of small vessels, usually predominantly the epineurial arterioles, and sparing arteries greater than 70 ␮m in diameter.84,152 Necrotizing vasculitis of larger vessels may also be seen.111 There may be significant fascicle-to-fascicle variation in the degree of fiber loss or axonal degeneration, and perineurial deposits of immunoglobulin may be evident. Underlying Malignancy Approximately one quarter of the reported cases have been associated with lymphoproliferative diseases and the remainder with a wide variety of carcinomas.111 The

2479

Paraneoplastic Neuropathy

symptoms of neuropathy were present 1 to 18 months before the diagnosis of malignancy or appeared 2 to 29 months after the diagnosis. The incidence of paraneoplastic vasculitic neuropathy is not known. Vincent and colleagues152 reported an associated malignancy in 7 of 50 cases (14%) of nerve and/or muscle microvasculitis, and subsequent series have found 1.4% of biopsy-proven vasculitic neuropathy111 and 5% of vasculitis125 to be paraneoplastic. Response to Treatment The response to treatment is somewhat variable in published reports; however, some cases do respond to immunotherapy. Antitumor therapy has led to an improvement in neuropathy in approximately 50% of reported cases.4,17,48,111 Various immunomodulatory therapies have been used. Cyclophosphamide, used alone or in combination with corticosteroids, led to improvement or stabilization in the three cases in which its use has been reported.111 Corticosteroids have not been effective in most cases.4,48,111

Motor Neuropathy Lower motor neuron involvement is reported in up to 20% to 25% of patients with anti-Hu–associated PEM/SSN,3,41,56 but motor neuron disease as the only or predominant paraneoplastic neurologic manifestation is much less common. In 1944, Weschler and colleagues156 reported a patient with motor neuron disease in association with pancreatic carcinoma. Brain and colleagues19 described 11 patients with motor neuron disease and cancer, most commonly carcinoma of the lung or breast, and in Croft and Wilkinson’s series37 motor neuron disease represented the second most frequent form of “carcinomatous neuromyopathy.” Shy and Silverstein128 reported an associated malignancy in 9 of 186 patients with a clinical picture compatible with motor neuron disease. Since that time there have been a number of cases reported with isolated or predominant lower motor neuron disease in association with lymphoproliferative disorders (see Chapter 111) and a variety of carcinomas, including small cell lung, ovarian, renal, and prostate carcinoma and thymoma.50,54,86,146 A single case of bilateral phrenic nerve palsy associated with renal cell carcinoma has also been reported.122 Anti-Hu and anti-Yo antibodies may be detected in the serum and CSF protein may be elevated, and there may be a variable pleocytosis.54,86,146 Although breast cancer is more characteristically associated with a pure upper motor neuron syndrome, a lower motor neuron syndrome has more recently been reported with serum antibodies that reacted with nodes of Ranvier and axon initial segments, but did not react with amphiphysin 2 on Western blot.52 There are numerous reports of improvement in the lower motor neuron syndrome after removal of the associated tumor.19,50,52,122

Pathologic studies show loss of anterior horn cells, variable inflammatory infiltrates in the spinal cord gray matter, and normal corticospinal tracts in the spinal cord and brainstem.146 Although the vast majority of cases of motor neuron disease are not malignancy associated, it has been suggested that further investigation may be warranted in patients with atypical presentations, that is, predominant lower motor neuron syndromes, persistent sensory symptoms, or other evidence of more widespread nervous system involvement, because treatment may benefit both the tumor and the neurologic disorder.163 A subacute motor axonal neuropathy, with serum antiGQ1b antibodies, has been reported in association with recurrent melanoma.87 The neuropathy progressed after immunization with tumor-derived antigen and improved after steroids and plasma exchange, suggesting shared immunopathogenetic mechanisms with the melanomaassociated demyelinating neuropathies described above.

Small Fiber Sensory Neuropathy Lipton and colleagues99 were able to demonstrate elevated thermal thresholds in 50% of patients with a variety of malignancies, compared to elevated vibration thresholds in 38%, leading them to suggest that small fiber neuropathy may be more commonly associated with cancer than had been previously recognized. In their series of patients with anti-Ri–associated paraneoplastic neurologic disease, Pittock and co-workers report one patient with an isolated small fiber neuropathy and non-SCLC.118 We have seen a single case presenting with painful dysesthesias, no clinical or electrophysiologic evidence of large fiber neuropathy, and morphologic changes in the epidermal nerve fibers consistent with a small fiber neuropathy who we initially thought had a postinfectious etiology134 but who developed limbic encephalitis 3 months later and was found to have a non-SCLC, suggesting that the neuropathy was paraneoplastic. The patient’s neuropathic symptoms resolved completely after a single course of IVIG.

Autonomic Neuropathy The association between chronic intestinal pseudoobstruction and underlying malignancy was first reported in 1948 by Ogilvie114; however, he believed that the “only probable” explanation was that the demonstrated subdiaphragmatic metastases had interrupted the sympathetic supply to the colon. Several authors subsequently described orthostatic hypotension, often with more widespread autonomic impairment, as a complication of bronchogenic and other cancers.82,129,140 When Lhermitte and colleagues96 demonstrated characteristic pathologic changes in both sympathetic ganglia and DRGs, in a patient with undifferentiated bronchogenic carcinoma, orthostatic hypotension, and peripheral neuropathy, and

2480

Diseases of the Peripheral Nervous System

Park and colleagues116 reported a patient with urinary retention, orthostatic hypotension, and gastrointestinal dysfunction that resolved after radiotherapy for an inoperable small cell carcinoma of the lung, the paraneoplastic nature of the autonomic failure was established. Clinical Features Autonomic dysfunction is a common finding in patients with anti-Hu–associated paraneoplastic syndromes. Symptoms or signs of significant autonomic impairment, most commonly gastrointestinal dysmotility, orthostatic hypotension, and bladder dysfunction, are present in 4% to 14% of patients at the time of diagnosis and develop in 14% to 30% over the course of the illness.27,41,66,67,71,100,130 Autonomic dysfunction is most frequently seen in association with neuropathy or neuronopathy,27,130 although it is also a common finding in paraneoplastic limbic encephalitis. Approximately two thirds of patients with autonomic impairment in the setting of these multifocal paraneoplastic syndromes have gastrointestinal dysmotility,27,130 characteristically presenting with persistent nausea and/or constipation and with radiologic evidence of pseudo-obstruction; one half to two thirds have significant orthostatic hypotension and one third have symptoms of bladder dysfunction. Autonomic neuropathy may occur as an isolated paraneoplastic syndrome or as the predominant manifestation. Although the usual presentation is subacute, acute autonomic neuropathy has also been described in patients with SCLC.32 Enteric neuropathy is well described with both SCLC and, less frequently, carcinoid.1,33,57,97,126,133 Most commonly this presents with subacute or chronic pseudo-obstruction characterized by constipation, nausea, vomiting, weight loss, abdominal pain, and distention. Symptoms may remain restricted to the gastrointestinal tract, but more widespread autonomic dysfunction, including orthostatic hypotension, pupillary abnormalities, bladder dysfunction, and altered sweating, and a sensory or sensorimotor neuropathy are present or develop in approximately 50% of patients.33,67,95,126,133 Gastrointestinal symptoms usually precede the tumor diagnosis, by 1 to 30 months in reported series, or begin within months after the diagnosis.33,95 Pathology The characteristic histologic findings are infiltration of the myenteric or myenteric and submucous plexuses with lymphocytes and plasma cells and a striking loss of enteric ganglion cells.1 The lymphocytic infiltration is T-cell predominant.95 The findings are similar to those seen in other areas of the nervous system in paraneoplastic disorders and are essentially identical to those seen in DRGs in sensory neuronopathy, which is often also evident in these patients. More recently a single case of SCLC-associated paraneoplastic gastrointestinal dysmotility has been described with relatively intact myenteric plexus but a sparse and disorganized interstitial cell of Cajal network.115

Autoantibodies Schuffler and colleagues, though unable to detect antibody reactive with myenteric plexus in a case of enteric neuropathy and widespread autonomic failure secondary to SCLC, hypothesized that the neuropathy was likely the result of shared antigenic determinants on tumor cells and neurons of the myenteric plexus.126 Lennon and co-workers95 subsequently demonstrated IgG antibodies reactive with enteric neurons, later recognized to represent anti-Hu immunoreactivity, in four of five patients with SCLC and paraneoplastic enteric neuropathy. No antibody was detected in the patient whose tumor had been successfully treated 2 years previously, in patients with SCLC without gastrointestinal dysmotility, or in patients with chronic idiopathic intestinal pseudo-obstruction.95 Since identical clinical and pathologic findings are seen in idiopathic myenteric ganglionitis,43 anti-Hu antibodies are a useful marker of paraneoplastic autonomic neuropathy.95,100 As in most other paraneoplastic neurologic syndromes, occasional cases have been reported with typical clinical and pathologic features and serum autoantibodies but no identifiable malignancy.23,60 More recently, antibodies more specific for autoimmune autonomic neuropathy have been described. Vernino and colleagues147 detected autoantibodies specific for the neuronal ganglionic acetylcholine receptor in 5 of 12 patients (42%) with subacute autonomic neuropathy. In three of these antibody-positive cases there was an underlying carcinoma. Ganglionic acetylcholine receptor antibodies are also present, usually at lower titer, in other thymoma- and SCLC-associated paraneoplastic neurologic disorders, including Lambert-Eaton myasthenic syndrome and Isaacs’ syndrome.147,148 Subsequent studies have demonstrated these antibodies in almost 30% of patients with paraneoplastic autonomic neuropathy and 50% with idiopathic autonomic neuropathy.149 Blocking, as opposed to binding, antibodies appear to be specific for autoimmune autonomic neuropathy, whether idiopathic or paraneoplastic. In this same study, anti-Hu antibodies were detected in 56% of the paraneoplastic group, predominantly in patients with SCLC. The ganglionic receptor antibody–positive patients had a variety of underlying cancers, including thymoma, SCLC, bladder, and rectal carcinoma. There are several lines of evidence to indicate that these antibodies are pathogenetic. High levels of antibodies are significantly correlated with more severe autonomic failure,149 and patients with paraneoplastic autonomic neuropathy who improved after tumor treatment showed a corresponding reduction in antibody level, whereas treated patients whose autonomic failure did not improve showed no change in titer. Ganglionic acetylcholine receptor protein has been demonstrated in SCLC lines derived from patients with paraneoplastic neurologic disorders,93 supporting molecular mimicry as a possible mechanism. Finally, rabbits immunized with neuronal nicotinic acetylcholine receptor

Paraneoplastic Neuropathy

develop profound gastrointestinal hypomotility, dilated pupils, and distended bladders, with the severity of the autonomic deficit paralleling antibody titer.93,150

IMMUNOPATHOGENESIS Although an autoimmune pathogenesis for these disorders is indicated by the presence of specific antineuronal antibodies in serum and CSF, lymphocytic pleocytosis and oligoclonal bands in CSF, and inflammatory infiltrates in affected areas of the nervous system, the exact pathogenetic mechanism(s) remain uncertain. The association between paraneoplastic sensory neuropathy and serum or CSF antibodies reactive against normal human brain was first reported in the mid-1960s by Wilkinson, who was able to demonstrate complement-fixing antibodies specific for DRGs and brain in five cases of malignancy-associated sensory neuropathy.159 He speculated that, “the tumor in a patient with sensory neuropathy may contain antigenic determinants not present in other tumors . . . shared by some constituent of the central nervous system. An immune reaction against such tumor determinants might then accidentally cause damage to the central nervous system.” Specific antineuronal antibodies that also react with the underlying tumor have subsequently been demonstrated in a variety of paraneoplastic neurologic disorders,8,24,53,61,63,70,78,101,123 and the concept of molecular mimicry, whereby expression of neuronal antigens by the tumor leads to a breakdown of immune tolerance and subsequent immune-mediated neuronal damage, remains the central theory of pathogenesis. Initial studies focused on the role of anti-Hu and other autoantibodies in paraneoplastic syndromes, and several lines of evidence suggested that these antibodies might be pathogenetic. Firstly, high antibody titers correlate strongly with clinical paraneoplastic syndromes, whereas low titers of anti-Hu antibodies are found in approximately 15% of patients with SCLC and no paraneoplastic neurologic disease.39,41 Also, intrathecal synthesis of anti-Hu antibody is positively correlated with the development of CNS disease; CSF antibody titers are higher than serum titers in most patients with encephalitis but in only a small percentage of patients with isolated sensory neuronopathy.56,69,145 Serum antibody titers are also higher in encephalitis than in isolated sensory neuronopathy, suggesting that DRG neurons represent an easier target than other areas of the CNS because of the partial absence of the blood-brain barrier.145 However, although occasional cases have been reported in which a fall in CSF antibody titer with treatment corresponded to an improvement in neurologic disease,162 most studies have not shown a correlation between antibody titer and disease course.16,85,112,143 A lack of consistent response to plasmapheresis also suggests that these are not entirely antibody-mediated diseases, although plasmapheresis

2481

generally only reduces serum, not CSF, antibody titers.59,69 More importantly, there is no direct evidence that anti-Hu antibodies are cytotoxic.79,151 Furthermore, attempts to develop animal models of anti-Hu–associated paraneoplastic syndromes have been unsuccessful. Both passive transfer and immunization experiments have failed to transfer clinical or pathologic disease in spite of the development of hightiter anti-Hu antibodies.45,132 The same is true for anti-Yo antibodies,65,138,139 but in the case of ganglionic acetylcholine receptor antibodies, immunization with the target antigen produces both high-titer antibodies and clinical disease that correlates with antibody titer.93,150 Subsequent studies have therefore investigated the role of cell-mediated immunity in these disorders. A large percentage of the T lymphocytes in the brains of patients with SSN/PEM recognize Hu antigen,136 indicating that the inflammatory infiltrates are directly involved in pathogenesis. Cytotoxic T cells have been demonstrated in close contact with sensory neurons in DRGs, attached to and indenting the surface of neurons, with or without IgG deposition, consistent with T-cell–mediated immune reaction.119,155 A limited T-cell receptor V␤ repertoire usage and clonal expansion of T-cell receptors in patients with SSN/PEM suggests that an antigen-driven cytotoxic T-lymphocyte response has a role in pathogenesis.153 Peripheral blood lymphocyte phenotype analysis in antiHu–positive patients with SSN/PEM and seronegative SCLC patients or healthy controls has shown increased antigenspecific proliferation of peripheral blood mononuclear cells with HuD stimulation and expansion of the memoryhelper subset in seropositive patients.14 Activated CD8 cells from a patient with anti-Hu antibodies have also been shown to lyse the patient’s own fibroblasts after injection of recombinant HuD protein into the fibroblasts and induction of human lymphocyte antigen class I molecules on their surface.137 These results indicate that HuD protein is an antigenic target for autoreactive T lymphocytes directly involved in cell-mediated injury to the nervous system. Hu antigens are expressed in all SCLCs, but only a small percentage of patients develop high-titer anti-Hu antibodies and paraneoplastic neurologic disease.25,39 Low-titer antibodies are evident in approximately 15% of SCLC patients without paraneoplastic disorders.39 The factors that determine whether patients mount an immune response against Hu antigens and how vigorous this response is remain unknown. Tumor expression of antigen is clearly necessary but not sufficient to cause a clinically overt immune response. It has been suggested that the immune response in patients with paraneoplastic syndromes may have some relevance to tumor progression. In the vast majority of patients with paraneoplastic neuropathies, the cancer is limited41,100 and in fact is often difficult to diagnose. There are multiple anecdotal reports of spontaneous remission of cancer and of prolonged survival in individual patients. Anti-Hu seropositivity has been reported to be a strong

2482

Diseases of the Peripheral Nervous System

predictor of response to chemotherapy and to be associated with longer survival in SCLC,62 and immunization with HuD DNA inhibits growth of an implanted neuroblastoma cell line in mice.29 Overall, there does appear to be improvement in the course of the tumor at the expense of often devastating neurologic disease.

DETECTION OF UNDERLYING MALIGNANCY Diagnosis of the underlying malignancy may be difficult because the tumor is often small and difficult to demonstrate with conventional imaging, such as computed tomography (CT) scanning. It has been estimated that, in 50% to 60% of cases with anti-Hu–associated paraneoplastic neurologic syndromes, the initial radiologic work-up is negative and in 5% to 15% the tumor is not detected despite repeated investigations.6 Early diagnosis of the tumor is important because it improves the prognosis for treatment of both the paraneoplastic syndrome and the malignancy. Positron emission tomography (PET) scanning enables detection of metabolically active tumor tissue with a resolution of 6 to 8 mm and therefore has superior sensitivity and specificity for diagnosing small tumors.121 Two published series have shown the usefulness of [18F] fluorodeoxyglucose (FDG) PET scanning in the diagnosis of tumors in patients with paraneoplastic neurologic disorders. Antoine and co-workers6 investigated 15 patients with paraneoplastic neurologic syndromes and anti-Hu antibodies. Initial investigations including chest radiography, bronchoscopy, and chest, abdominal, and pelvic CT led to a diagnosis of cancer in 12 of the 15 patients. All three of the remaining patients had abnormal mediastinal uptake of FDG, and in two the malignancy was confirmed on subsequent investigations. In the third patient repeat CT scanning after 18 months again failed to detect any mass that could be biopsied. In a larger series, Rees and colleagues121 reported positive PET scans in 16 of 43 patients (37%) with suspected paraneoplastic neurologic disease and negative conventional imaging, including CT. In patients with a suggestive clinical syndrome, especially patients over the age of 60 with risk factors for malignancy such as a significant smoking history, serial imaging studies are warranted, even in the absence of specific antineuronal antibodies.

SSN/PEM,41,71,100,130 although in the largest reported series 7 of 34 patients (21%) treated with immunotherapy alone showed improvement or stabilization compared with 5 of 43 untreated patients (12%). Treatment with IVIG has been reported to yield sustained improvement, especially when started within a few weeks of the onset of neurologic symptoms, or stabilization in small series of patients.16,143 Plasmapheresis was not effective in a single small retrospective series of seven patients69; however, there is a single case report of improvement in anti-Hu–associated autonomic neuropathy with relapse and subsequent repeated improvement after a further course of plasma exchange.18 Isolated cases responding to high-dose intravenous steroids alone or in combination with azathioprine and IVIG,112 and effective combined IVIG and cyclophosphamide therapy in another,107 have also been reported. A trial of pulse therapy with IVIG, high-dose intravenous steroids, and cyclophosphamide has been conducted in 17 patients with SSN/PEM or paraneoplastic cerebellar degeneration.85 None of the patients improved, but one third of those patients still ambulant at the commencement of therapy stabilized. The significance of stabilization in these uncontrolled studies is difficult to assess given the tendency of the neurologic deficit to plateau in untreated cases and the occurrence of occasional spontaneous remission.130 The results of antitumor therapy are slightly more promising. In their series of 71 cases, Dalmau and colleagues41 found no improvement in the neurologic disorder with treatment of the tumor by surgery, chemotherapy, or irradiation, although details of the response of the malignancy to treatment are not given. There are a number of isolated case reports of improvement after successful antitumor therapy,11,50,135 and several large series have demonstrated improvement or stabilization of the neurologic syndrome in 38% to 80% of patients receiving antitumor therapy (with or without immunotherapy), compared to none to 27% given no treatment and none to 21% receiving immunotherapy alone.31,66,71,130 The suggestion has been made that cases with isolated involvement of the peripheral nervous system may have better likelihood of response to therapy.85 The response to therapy may also be better in vasculitic and demyelinating neuropathies (see above), as might be expected since neuronal cell bodies are largely spared unless there is coincident subacute sensory SSN.

TREATMENT In general the response to treatment is disappointing, but antitumor therapy seems to be slightly more effective than immunomodulatory therapy in stabilizing the neurologic deficit.66,69,85,130 Immunotherapy with steroids, IVIG, plasma exchange, and cyclophosphamide alone or in combination have been largely ineffective in large series of patients with

REFERENCES 1. Ahmed, M. N., and Carpenter, S.: Autonomic neuropathy and carcinoma of the lung. Can. Med. Assoc. J. 113:410, 1975. 2. Altermatt, H. J., Rodriguez, M., Scheithauer, B. W., et al.: Paraneoplastic anti-Purkinje and type-I anti-neuronal nuclear antibodies bind selectively to central, peripheral and autonomic nervous system cells. Lab. Invest. 65:412, 1991.

Paraneoplastic Neuropathy 3. Anderson, N. E., Rosenblum, M. K., Graus, F., et al.: Autoantibodies in paraneoplastic syndromes associated with small-cell lung cancer. Neurology 38:1391, 1988. 4. Ansari, J., Nagabhushan, N., Syed, R., et al.: Small cell lung cancer associated with anti-Hu paraneoplastic sensory neuropathy and peripheral nerve microvasculitis: case report and literature review. Clin. Oncol. (R. Coll. Radiol.) 16:71, 2004. 5. Antoine, J. C., Absi, L., Honnorat, J., et al.: Antiamphiphysin antibodies are associated with various paraneoplastic neurological syndromes and tumors. Arch. Neurol. 56:172, 1999. 6. Antoine, J. C., Cinotti, L., Tilikete, C., et al.: [18F]Fluorodeoxyglucose positron emission tomography in the diagnosis of cancer in patients with paraneoplastic neurological syndrome and anti-Hu antibodies. Ann. Neurol. 48:105, 2000. 7. Antoine, J. C., Honnorat, J., Camdessanche, J. P., et al.: Paraneoplastic anti-CV2 antibodies react with peripheral nerve and are associated with a mixed axonal and demyelinating peripheral neuropathy. Ann. Neurol. 49:214, 2001. 8. Antoine, J. C., Honnorat, J., Vocanson, C., et al.: Posterior uveitis, paraneoplastic encephalomyelitis and auto-antibodies reacting with developmental protein of brain and retina. J. Neurol. Sci. 117:215, 1993. 9. Antoine, J. C., Mosnier, J. F., Absi, L., et al.: Carcinoma associated paraneoplastic peripheral neuropathies in patients with and without anti-onconeural antibodies. J. Neurol. Neurosurg. Psychiatry 67:7, 1999. 10. Antoine, J. C., Mosnier, J. F., Lapras, J., et al.: Chronic inflammatory demyelinating polyneuropathy associated with carcinoma. J. Neurol. Neurosurg. Psychiatry 60:188, 1996. 11. Batson, O. A., Fantle, D. M., and Stewart, J. A.: Paraneoplastic encephalomyelitis: dramatic response to chemotherapy alone. Cancer 69:1291, 1992. 12. Bechich, S., Graus, F., Arboix, A., et al.: Anti-Hu-associated paraneoplastic sensory neuropathy and breast cancer. J. Neurol. 247:552, 2000. 13. Bell, C. E. Jr., and Seetharam, S.: Identification of the Schwann cell as a peripheral nervous system cell possessing a differentiation antigen expressed by a human lung tumor. J. Immunol. 118:826, 1977. 14. Benyahia, B., Liblau, R., Merle-Beral, H., et al.: Cellmediated autoimmunity in paraneoplastic neurological syndromes with anti-Hu antibodies. Ann. Neurol. 45:162, 1999. 15. Bird, S. J., Brown, M. J., Shy, M. E., et al.: Chronic inflammatory demyelinating polyneuropathy associated with malignant melanoma. Neurology 46:822, 1996. 16. Blaes, F., Strittmatter, M., Merkelbach, S., et al.: Intravenous immunoglobulins in the therapy of paraneoplastic neurological disorders. J. Neurol. 246:299, 1999. 17. Blumenthal, D., Schochet, S. Jr., Gutmann, L., et al.: Smallcell carcinoma of the lung presenting with paraneoplastic peripheral nerve microvasculitis and optic neuropathy. Muscle Nerve 21:1358, 1998. 18. Bohnen, N. I., Cheshire, W. P., Lennon, V. A., and Van Den Berg, C. J.: Plasma exchange improves function in a patient with ANNA-1 seropositive paraneoplastic autonomic neuropathy [abstract]. Neurology 48:A131, 1997. 19. Brain, L., Croft, P. B., and Wilkinson, M.: Motor neurone disease as a manifestation of neoplasm (with a note on the

20. 21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32. 33.

34.

35.

36.

37.

2483

course of classical motor neurone disease). Brain 88:479, 1965. Brain, R.: The neurological complications of neoplasms. Lancet 1:179, 1963. Brain, R., and Henson, R. A.: Neurological syndromes associated with carcinoma: the carcinomatous neuromyopathies. Lancet 2:971, 1958. Brashear, H. R., Caccamo, D. V., Heck, A., et al.: Localization of antibody in the central nervous system of a patient with paraneoplastic encephalomyeloneuritis. Neurology 41:1583, 1991. Briellmann, R. S., Sturzenegger, M., Gerber, H. A., et al.: Autoantibody-associated sensory neuronopathy and intestinal pseudo-obstruction without detectable neoplasia. Eur. Neurol. 36:369, 1996. Budde-Steffen, C., Anderson, N. E., Rosenblum, M. K., et al.: An anti-neuronal autoantibody in paraneoplastic opsoclonus. Ann. Neurol. 23:528, 1988. Budde-Steffen, C., Anderson, N. E., Rosenblum, M. K., et al.: Expression of an antigen in small cell lung carcinoma lines detected by antibodies from patients with paraneoplastic dorsal root ganglionopathy. Cancer Res. 48:430, 1988. Byk, T., Dobransky, T., Cifuentes-Diaz, C., et al.: Identification and molecular characterization of Unc-33-like phosphoprotein (Ulip), a putative mammalian homolog of the axonal guidance-associated unc-33 gene product. J. Neurosci. 16:688, 1996. Camdessanche, J. P., Antoine, J. C., Honnorat, J., et al.: Paraneoplastic peripheral neuropathy associated with anti-Hu antibodies: a clinical and electrophysiological study of 20 patients. Brain 125:166, 2002. Camerlingo, M., Nemni, R., Ferraro, B., et al.: Malignancy and sensory neuropathy of unexplained cause: a prospective study of 51 patients. Arch. Neurol. 55:981, 1998. Carpentier, A. F., Rosenfeld, M. R., Delattre, J. Y. et al.: DNA vaccination with HuD inhibits growth of a neuroblastom in mice. Clin. Cancer Res. 4:2819, 1998. Cavaletti, G., Bogliun, G., Marzorati, L., et al.: The incidence and course of paraneoplastic neuropathy in women with epithelial ovarian cancer. J. Neurol. 238:371, 1991. Chalk, C. H., Windebank, A. J., Kimmel, D. W., et al.: The distinctive clinical features of paraneoplastic sensory neuronopathy. Can. J. Neurol. Sci. 19:346, 1992. Chiappa, K. H., and Young, R. H.: A case of paracarcinomatous pandysautonomia [abstract]. Neurology 23:423, 1973. Chinn, J. S., and Schuffler, M. D.: Paraneoplastic visceral neuropathy as a cause of severe gastrointestinal motor dysfunction. Gastroenterology 95:1279, 1988. Croft, P. B., Urich, H., and Wilkinson, M.: Peripheral neuropathy of sensorimotor type associated with malignant disease. Brain 90:31, 1967. Croft, P. B., and Wilkinson, M.: Carcinomatous neuromyopathy: its incidence in patients with carcinoma of the lung and carcinoma of the breast. Lancet 1:184, 1963. Croft, P. B., and Wilkinson, M.: The incidence of carcinomatous neuromyopathy in patients with various types of carcinoma. Brain 88:427, 1965. Croft, P. B., and Wilkinson, M.: The course and prognosis in some types of carcinomatous neuromyopathy. Brain 92:1, 1969.

2484

Diseases of the Peripheral Nervous System

38. Dalmau, J., Furneaux, H. M., Cordon-Cardo, C., et al.: The expression of the Hu (paraneoplastic encephalomyelitis/ sensory neuronopathy) antigen in human normal and tumor tissues. Am. J. Pathol. 141:881, 1992. 39. Dalmau, J., Furneaux, H. M., Gralla, R. J., et al.: Detection of the anti-Hu antibody in the serum of patients with small cell lung cancer—a quantitative Western blot analysis. Ann. Neurol. 27:544, 1990. 40. Dalmau, J., Furneaux, H. M., Rosenblum, M. K., et al.: Detection of the anti-Hu antibody in specific regions of the nervous system and tumor from patients with paraneoplastic encephalomyelitis/sensory neuronopathy. Neurology 41:1757, 1991. 41. Dalmau, J., Graus, F., Rosenblum, M. K., et al.: AntiHu–associated paraneoplastic encephalomyelitis/sensory neuronopathy: a clinical study of 71 patients. Medicine (Baltimore) 71:59, 1992. 42. Dalmau, J., and Posner, J. B.: Neurologic paraneoplastic antibodies (anti-Yo; anti-Hu; anti-Ri): the case for a nomenclature based on antibody and antigen specificity. Neurology 44:2241, 1994. 43. De Giorgio, R., Barbara, G., Stanghellini, V., et al.: Clinical and morphofunctional features of idiopathic myenteric ganglionitis underlying severe intestinal motor dysfunction: a study of three cases. Am. J. Gastroenterol. 97:2454, 2002. 44. Denny-Brown, D.: Primary sensory neuropathy with muscular changes associated with carcinoma. J. Neurol. Neurosurg. Psychiatry 11:73, 1948. 45. Dick, D. J., Harris, J. B., Falkous, G., et al.: Neuronal antinuclear antibody in paraneoplastic sensory neuronopathy. J. Neurol. Sci. 85:1, 1988. 46. Donofrio, P. D., Alessi, A. G., Albers, J. W., et al.: Electrodiagnostic evolution of carcinomatous sensory neuronopathy. Muscle Nerve 12:508, 1989. 47. Dropcho, E. J.: Antiamphiphysin antibodies with small-cell lung carcinoma and paraneoplastic encephalomyelitis. Ann. Neurol. 39:659, 1996. 48. Eggers, C., Hagel, C., and Pfeiffer, G.: Anti-Hu-associated paraneoplastic sensory neuropathy with peripheral nerve demyelination and microvasculitis. J. Neurol. Sci. 155:178, 1998. 49. Elrington, G. M., Murray, N. M., Spiro, S. G., et al.: Neurological paraneoplastic syndromes in patients with small cell lung cancer: a prospective survey of 150 patients. J. Neurol. Neurosurg. Psychiatry 54:764, 1991. 50. Evans, B. K., Fagan, C., Arnold, T., et al.: Paraneoplastic motor neuron disease and renal cell carcinoma: improvement after nephrectomy. Neurology 40:960, 1990. 51. Evans, C. C., and Kaufman, H. D.: Unusual presentation of seminoma of the testis. Br. J. Surg. 58:703, 1971. 52. Ferraci, F., Fassetta, G., Butler, M. H., et al.: A novel antineuronal antibody in a motor neuron syndrome associated with breast cancer. Neurology 53:852, 1999. 53. Folli, F., Solimena, M., Cofiell, R., et al.: Autoantibodies to a 128-kd synaptic protein in three women with the stiff-man syndrome and breast cancer. N. Engl. J. Med. 328:546, 1993. 54. Forsyth, P. A., Dalmau, J., Graus, F., et al.: Motor neuron syndromes in cancer patients. Ann. Neurol. 41:722, 1997.

55. Fuller, G. N., Spies, J. M., Pollard, J. D., et al.: Demyelinating neuropathies triggered by melanoma immunotherapy. Neurology 44:2404, 1994. 56. Furneaux, H. F., Reich, L., and Posner, J. B.: Autoantibody synthesis in the central nervous system of patients with paraneoplastic syndromes. Neurology 40:1085, 1990. 57. Gerl, A., Storck, M., Schalhorn, A., et al.: Paraneoplastic chronic intestinal pseudoobstruction as a rare complication of bronchial carcinoid. Gut 33:1000, 1992. 58. Goshima, Y., Nakamura, F., Strittmatter, P., et al.: Collapsininduced growth cone collapse mediated by an intracellular protein related to UNC-33. Nature 376:509, 1995. 59. Graus, F., Abos, J., Roquer, J., et al.: Effect of plasmapheresis on serum and CSF autoantibody levels in CNS paraneoplastic syndromes. Neurology 40:1621, 1990. 60. Graus, F., Bonaventura, I., Uchuya, M., et al.: Indolent antiHu-associated paraneoplastic sensory neuropathy. Neurology 44:2258, 1994. 61. Graus, F., Cordon-Cardo, C., and Posner, J. B.: Neuronal antinuclear antibody in sensory neuronopathy from lung cancer. Neurology 35:538, 1985. 62. Graus, F., Dalmau, J., Rene, R., et al.: Anti-Hu antibodies in patients with small-cell lung cancer: association with complete response to therapy and improved survival. J. Clin. Oncol. 15:2866, 1997. 63. Graus, F., Elkon, K. B., Cordon-Cardo, C., et al.: Sensory neuronopathy and small cell lung cancer: antineuronal antibody that also reacts with the tumor. Am. J. Med. 80:45, 1986. 64. Graus, F., Ferrer, I., and Lamarca, J.: Mixed carcinomatous neuropathy in patients with lung cancer and lymphoma. Acta Neurol. Scand. 68:40, 1983. 65. Graus, F., Illa, I., Agusti, M., et al.: Effect of intraventricular injection of an anti-Purkinje cell antibody (anti-Yo) in a guinea pig model. J. Neurol. Sci. 106:82, 1991. 66. Graus, F., Keime-Guibert, F., Rene, R., et al.: Anti-Huassociated paraneoplastic encephalomyelitis: analysis of 200 patients. Brain 124:1138, 2001. 67. Graus, F., and Rene, R.: Paraneoplastic neuropathies. Eur. Neurol. 33:279, 1993. 68. Graus, F., Ribalta, T., Campo, E., et al.: Immunohistochemical analysis of the immune reaction in the nervous system in paraneoplastic encephalomyelitis. Neurology 40:219, 1990. 69. Graus, F., Vega, F., Delattre, J. Y., et al.: Plasmapheresis and antineoplastic treatment in CNS paraneoplastic syndromes with antineuronal autoantibodies. Neurology 42:536, 1992. 70. Greenlee, J. E., and Brashear, H. R.: Antibodies to cerebellar Purkinje cells in patients with paraneoplastic cerebellar degeneration and ovarian carcinoma. Ann. Neurol. 14:609, 1983. 71. Gultekin, S. H., Rosenfeld, M. R., Voltz, R., et al.: Paraneoplastic limbic encephalitis: neurological symptoms, immunological findings and tumour association in 50 patients. Brain 123(Pt 7):1481, 2000. 72. Halls, J., Bredkjaer, C., and Friis, M. L.: Guillain-Barre syndrome: diagnostic criteria, epidemiology, clinical course and prognosis. Acta Neurol. Scand. 78:118, 1988. 73. Hawley, R. J., Cohen, M. H., Saini, N., et al.: The carcinomatous neuromyopathy of oat cell lung cancer. Ann. Neurol. 7:65, 1980.

Paraneoplastic Neuropathy 74. Henson, R. A., Hoffman, H. L., and Urich, H.: Encephalomyelitis with carcinoma. Brain 88:449, 1965. 75. Henson, R. A., Russell, D. S., and Wilkinson, M.: Carcinomatous neuropathy and myopathy: a clinical and pathological study. Brain 77:82, 1954. 76. Henson, R. A., and Urich, H.: Encephalomyelitis with cancer. In Henson, R. A. and Urich, H. (eds): Cancer and the Nervous System. Oxford, UK, Blackwell Scientific Publications, p. 314, 1982. 77. Hildebrand, J., and Coers, C.: The neuromuscular function in patients with malignant tumours: electromyographic and histological study. Brain 90:67, 1967. 78. Honnorat, J., Antoine, J. C., Derrington, E., et al.: Antibodies to a subpopulation of glial cells and a 66 kDa developmental protein in patients with paraneoplastic neurological syndromes. J. Neurol. Neurosurg. Psychiatry 61:270, 1996. 79. Hormigo, A., and Lieberman, F.: Nuclear localization of anti-Hu antibody is not associated with in vitro cytotoxicity. J. Neuroimmunol. 55:205, 1994. 80. Horwich, M. S., Cho, L., Porro, R. S., et al.: Subacute sensory neuropathy: a remote effect of carcinoma. Ann. Neurol. 2:7, 1977. 81. Ichimura, M., Yamamoto, M., Kobayashi, Y., et al.: Tissue distribution of pathological lesions and Hu antigen expression in paraneoplastic sensory neuronopathy. Acta Neuropathol. (Berl.), 95:641, 1998. 82. Ivy, H. K.: Renal sodium loss and bronchogenic carcinoma: associated autonomic neuropathy. Arch. Intern. Med. 108:47, 1961. 83. Jean, W. C., Dalmau, J., Ho, A., et al.: Analysis of the IgG subclass distribution and inflammatory infiltrates in patients with anti-Hu-associated paraneoplastic encephalomyelitis. Neurology 44:140, 1994. 84. Johnson, P. C., Rolak, L. A., Hamilton, R. H., et al.: Paraneoplastic vasculitis of nerve: a remote effect of cancer. Ann. Neurol. 5:437, 1979. 85. Keime-Guibert, F., Graus, F., Fleury, A., et al.: Treatment of paraneoplastic neurological syndromes with antineuronal antibodies (Anti-Hu, anti-Yo) with a combination of immunoglobulins, cyclophosphamide, and methylprednisolone. J. Neurol. Neurosurg. Psychiatry 68:479, 2000. 86. Khwaja, S., Sripathi, N., Ahmad, B. K., et al.: Paraneoplastic motor neuron disease with type 1 Purkinje cell antibodies. Muscle Nerve 21:943, 1998. 87. Kloos, L., Sillevis, S. P., Ang, C. W., et al.: Paraneoplastic ophthalmoplegia and subacute motor axonal neuropathy associated with anti-GQ1b antibodies in a patient with malignant melanoma. J. Neurol. Neurosurg. Psychiatry 74:507, 2003. 88. Konanc, D. A., O’Neill, B. P., Caselli, R. J., and Lennon, V. A.: Peripheral neuropathic presentation in 3 patients with seropositivity for type 1 anti-Purkinje cell antibodies (PCA-1) and ovarian or related Müllerian carcinoma [abstract]. Neurology 45(Suppl 4):A320, 1995. 89. Krarup, C., and Crone, C.: Neurophysiological studies in malignant disease with particular reference to involvement of peripheral nerves. J. Neurol. 249:651, 2002. 90. Kudo, M., and Nogoshi, T.: Immunoreactive myelin basic protein in tumour cells associated with carcinomatous neuropathy. Am. J. Clin. Pathol. 84:741, 1985.

2485

91. Lenman, J. A., Fleming, A. M., Robertson, M. A., et al.: Peripheral nerve function in patients with bronchial carcinoma: comparison with matched controls and effects of treatment. J. Neurol. Neurosurg. Psychiatry 44:54, 1981. 92. Lennon, V. A.: Paraneoplastic autoantibodies: the case for a descriptive generic nomenclature. Neurology 44:2236, 1994. 93. Lennon, V. A., Ermilov, L. G., Szurszewski, J. H., et al.: Immunization with neuronal nicotinic acetylcholine receptor induces neurological autoimmune disease. J. Clin. Invest. 111:907, 2003. 94. Lennon, V. A., Manley, H. A., Kim, K., and Parisi, J. E.: Amphiphysin antibodies: a paraneoplastic serological marker of breast and lung cancer related encephalomyeloradiculoneuritides but not classical stiff man syndrome [abstract]. Neurology 48(Suppl. 1):A434, 1997. 95. Lennon, V. A., Sas, D. F., Busk, M. F., et al.: Enteric neuronal autoantibodies in pseudoobstruction with small-cell lung carcinoma. Gastroenterology 100:137, 1991. 96. Lhermitte, F., Gautier, J. C., Escourolle, R., et al.: Orthostatic hypotension and peripheral neuropathy in the course of bronchial carcinoma. Presse Med. 78:257, 1970. 97. Lhermitte, F., Gray, F., Lyon-Caen, O., et al.: Paralysis of digestive tract with lesions of myenteric plexuses: a new paraneoplastic syndrome [author’s transl.]. Rev. Neurol. (Paris) 136:825, 1980. 98. Lin, K. P., Kwan, S. Y., Chen, S. Y., et al.: Generalized neuropathy in Taiwan: an etiologic survey. Neuroepidemiology 12:257, 1993. 99. Lipton, R. B., Galer, B. S., Dutcher, J. P., et al.: Large and small fibre type sensory dysfunction in patients with cancer. J. Neurol. Neurosurg. Psychiatry 54:706, 1991. 100. Lucchinetti, C. F., Kimmel, D. W., and Lennon, V. A.: Paraneoplastic and oncologic profiles of patients seropositive for type 1 antineuronal nuclear autoantibodies. Neurology 50:652, 1998. 101. Luque, F. A., Furneaux, H., Ferziger, R., et al.: Anti-Ri: an antibody associated with paraneoplastic opsoclonus and breast cancer. Ann. Neurol. 29:241, 1991. 102. Manley, G. T., Smitt, P. S., Dalmau, J., et al.: Hu antigens: reactivity with Hu antibodies, tumor expression, and major immunogenic sites. Ann. Neurol. 38:102, 1995. 103. McNamara B., Boniface, S. J., Ray, J., et al.: Paraneoplastic sensory neuropathy and Purkinje cell antibodies. Muscle Nerve 22:1466, 1999. 104. Molinuevo, J. L., Graus, F., Serrano, C., et al.: Utility of anti-Hu antibodies in the diagnosis of paraneoplastic sensory neuropathy. Ann. Neurol. 44:976, 1998. 105. Moll, J. W., Markusse, H. M., Henzen-Logmans, S. C., et al.: Anti-neuronal antibodies in Sjogren’s syndrome and in paraneoplastic neurological disease. Lancet 343:1299, 1994. 106. Moody, J. F.: Electrophysiological investigations into the neurological complications of carcinoma. Brain 88:1023, 1965. 107. Mowzoon, N., and Bradley, W. G.: Successful immunosuppressant therapy of severe progressive cerebellar degeneration and sensory neuropathy: a case report. J. Neurol. Sci. 178:63, 2000. 108. Nemni, R., Camerlingo, M., Fazio, R., et al.: Serum antibodies to Purkinje cells and dorsal root ganglia neurons in

2486

109.

110.

111. 112.

113.

114. 115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

Diseases of the Peripheral Nervous System

sensory neuronopathy without malignancy. Ann. Neurol. 34:848, 1993. Noronha, A. B., Harper, J. R., Ilyas, A. A., et al.: Myelinassociated glycoprotein shares an antigenic determinant with a glycoprotein of human melanoma cells. J. Neurochem. 47:1558, 1986. Notermans, N. C., Wokke, J. H., Franssen, H., et al.: Chronic idiopathic polyneuropathy presenting in middle or old age: a clinical and electrophysiological study of 75 patients. J. Neurol. Neurosurg. Psychiatry 56:1066, 1993. Oh, S. J.: Paraneoplastic vasculitis of the peripheral nervous system. Neurol. Clin. 15:849, 1997. Oh, S. J., Dropcho, E. J., and Claussen, G. C.: Anti-Huassociated paraneoplastic sensory neuropathy responding to early aggressive immunotherapy: report of two cases and review of literature. Muscle Nerve 20:1576, 1997. Oh, S. J., Slaughter, R., and Harrell, L.: Paraneoplastic vasculitic neuropathy: a treatable neuropathy. Muscle Nerve 14:152, 1991. Ogilvie, H.: Large-intestine colic due to sympathetic deprivation. Br. Med. J. 2:671, 1948. Pardi, D. S., Miller, S. M., Miller, D. L., et al.: Paraneoplastic dysmotility: loss of interstitial cells of Cajal. Am. J. Gastroenterol. 97:1828, 2002. Park, D. M., Johnson, R. H., Crean, G. P., et al.: Orthostatic hypotension in bronchial carcinoma. Br. Med. J. 3:510, 1972. Peng, S. S., Chen, C. Y., Xu, N., et al.: RNA stabilization by the AU-rich element binding protein, HuR, an ELAV protein. EMBO J. 17:3461, 1998. Pittock, S. J., Lucchinetti, C. F., and Lennon, V. A.: Antineuronal nuclear autoantibody type 2: paraneoplastic accompaniments. Ann. Neurol. 53:580, 2003. Plonquet, A., Gherardi, R. K., Creange, A., et al.: Oligoclonal T-cells in blood and target tissues of patients with anti-Hu syndrome. J. Neuroimmunol. 122:100, 2002. Pourmand, R., and Maybury, B. G.: AAEM Case Report #31: paraneoplastic sensory neuronopathy. Muscle Nerve 19:1517, 1996. Rees, J. H., Hain, S. F., Johnson, M. R., et al.: The role of [18F]fluoro-2-deoxyglucose-PET scanning in the diagnosis of paraneoplastic neurological disorders. Brain 124:2223, 2001. Rijnders, B., and Decramer, M.: Reversibility of paraneoplastic bilateral diaphragmatic paralysis after nephrectomy for renal cell carcinoma. Ann. Oncol. 11:221, 2000. Saiz, A., Dalmau, J., Butler, M. H., et al.: Anti-amphiphysin I antibodies in patients with paraneoplastic neurological disorders associated with small cell lung carcinoma. J. Neurol. Neurosurg. Psychiatry 66:214, 1999. Saleh, M. N., Khazaeli, M. B., Wheeler, R. H., et al.: Phase I trial of the murine monoclonal anti-GD2 antibody 14G2a in metastatic melanoma. Cancer Res. 52:4342, 1992. Sanchez-Guerrero, J., Gutierrez-Urena, S., Vidaller, A., et al.: Vasculitis as a paraneoplastic syndrome: report of 11 cases and review of the literature. J. Rheumatol. 17:1458, 1990. Schuffler, M. D., Baird, H. W., Fleming, C. R., et al.: Intestinal pseudo-obstruction as the presenting manifestation of small-cell carcinoma of the lung: a paraneoplastic

127.

128. 129.

130.

131.

132.

133.

134.

135.

136.

137.

138.

139.

140.

141.

neuropathy of the gastrointestinal tract. Ann. Intern. Med. 98:129, 1983. Scully, R. E., Mark, E. J., McNeely, W. F., et al.: Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 20-1997: a 74-year-old man with progressive cough, dyspnea, and pleural thickening. N. Engl. J. Med. 336:1895, 1997. Shy, G. M., and Silverstein, I.: A study of the effects upon the motor unit by remote malignancy. Brain 88:515, 1965. Siemsen, J. K., and Meister, L.: Bronchogenic carcinoma associated with severe orthostatic hypotension. Ann. Intern. Med. 58:669, 1963. Sillevis Smitt, P., Grefkens, J., de Leeuw, B., et al.: Survival and outcome in 73 anti-Hu positive patients with paraneoplastic encephalomyelitis/sensory neuronopathy. J. Neurol. 249:745, 2002. Sillevis-Smitt, P., Manley, G., Moll, J.W., et al.: Pitfalls in the diagnosis of autoantibodies associated with paraneoplastic neurologic disease. Neurology 46:1739, 1996. Sillevis Smitt, P. A., Manley, G. T., and Posner, J. B.: Immunization with the paraneoplastic encephalomyelitis antigen HuD does not cause neurologic disease in mice. Neurology 45:1873, 1995. Sodhi, N., Camilleri, M., Camoriano, J. K., et al.: Autonomic function and motility in intestinal pseudoobstruction caused by paraneoplastic syndrome. Dig. Dis. Sci. 34:1937, 1989. Spring, P. J., Krupka, E., and Spies, J. M.: Post-infectious sensory neuropathy responding to intravenous immunoglobulin [abstract]. J. Peripher. Nerv. Syst. 6:179, 2001. Suzuki, M., Kimura, H., Tachibana, I., et al.: Improvement of anti-Hu-associated paraneoplastic sensory neuropathy after chemoradiotherapy in a small cell lung cancer patient. Intern. Med. 40:1140, 2001. Szabo, A., Dalmau, J., Manley, G., et al.: HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and Sex-lethal. Cell 67:287, 1991. Tanaka, K., Tanaka, M., Inuzuka, T., et al.: Cytotoxic T lymphocyte-mediated cell death in paraneoplastic sensory neuronopathy with anti-Hu antibody. J. Neurol. Sci. 163:159, 1999. Tanaka, K., Tanaka, M., Onodera, O., et al.: Passive transfer and active immunization with the recombinant leucine-zipper (Yo) protein as an attempt to establish an animal model of paraneoplastic cerebellar degeneration. J. Neurol. Sci. 127:153, 1994. Tanaka, K., Tanaka, M., Onodera, O., et al.: Trial to establish an animal model of paraneoplastic cerebellar degeneration with anti-Yo antibody. 1. Mouse strains bearing different MHC molecules produce antibodies on immunization with recombinant Yo protein, but do not cause Purkinje cell loss. Clin. Neurol. Neurosurg. 97:95, 1995. Thomas, J. P., and Shields, R.: Associated autonomic dysfunction and carcinoma of the pancreas. Br. Med. J. 4:32, 1970. Torvik, A., and Berntzen, A. E.: Necrotizing vasculitis without visceral involvement: postmortem examination of three cases with affection of skeletal muscles and peripheral nerves. Acta Med. Scand. 184:69, 1968.

Paraneoplastic Neuropathy 142. Trojaborg, W., Frantzen, E., and Andersen, I.: Peripheral neuropathy and myopathy associated with carcinoma of the lung. Brain 92:71, 1969. 143. Uchuya, M., Graus, F., Vega, F., et al.: Intravenous immunoglobulin treatment in paraneoplastic neurological syndromes with antineuronal autoantibodies. J. Neurol. Neurosurg. Psychiatry 60:388, 1996. 144. van Oosterhout, A. G., van de Pol, M., ten Velde, G. P. and Twijnstra, A.: Neurologic disorders in 203 consecutive patients with small cell lung cancer: results of a longitudinal study. Cancer 77:1434, 1996. 145. Vega, F., Graus, F., Chen, Q. M., et al.: Intrathecal synthesis of the anti-Hu antibody in patients with paraneoplastic encephalomyelitis or sensory neuronopathy: clinicalimmunologic correlation. Neurology 44:2145, 1994. 146. Verma, A., Berger, J. R., Snodgrass, S., et al.: Motor neuron disease: a paraneoplastic process associated with anti-Hu antibody and small-cell lung carcinoma. Ann. Neurol. 40:112, 1996. 147. Vernino, S., Adamski, J., Kryzer, T. J., et al.: Neuronal nicotinic ACh receptor antibody in subacute autonomic neuropathy and cancer-related syndromes. Neurology 50:1806, 1998. 148. Vernino, S., and Lennon, V. A.: Neuronal ganglionic acetylcholine receptor autoimmunity. Ann. N. Y. Acad. Sci. 998:211, 2003. 149. Vernino, S., Low, P. A., Fealey, R. D., et al.: Autoantibodies to ganglionic acetylcholine receptors in autoimmune autonomic neuropathies. N. Engl. J. Med. 343:847, 2000. 150. Vernino, S., Low, P. A., and Lennon, V. A.: Experimental autoimmune autonomic neuropathy. J. Neurophysiol. 90:2053, 2003. 151. Verschuuren, J. J., Dalmau, J., Hoard, R., et al.: Paraneoplastic anti-Hu serum: studies on human tumor cell lines. J. Neuroimmunol. 79:202, 1997. 152. Vincent, D., Dubas, F., Hauw, J. J., et al.: Nerve and muscle microvasculitis in peripheral neuropathy: a remote

153.

154.

155.

156.

157.

158.

159. 160.

161.

162.

163. 164.

2487

effect of cancer? J. Neurol. Neurosurg. Psychiatry 49:1007, 1986. Voltz, R., Dalmau, J., Posner, J. B., et al.: T-cell receptor analysis in anti-Hu associated paraneoplastic encephalomyelitis. Neurology 51:1146, 1998. Wang, L. H. and Strittmatter, S. M.: A family of rat CRMP genes is differentially expressed in the nervous system. J. Neurosci. 16:6197, 1996. Wanschitz, J., Hainfellner, J. A., Kristoferitsch, W., et al.: Ganglionitis in paraneoplastic subacute sensory neuronopathy: a morphologic study. Neurology 49:1156, 1997. Wechsler, I. S., Sapirstein, M. R., and Stein, A.: Primary and symptomatic amyotrophic lateral sclerosis. Am. J. Med. Sci. 208:70, 1944. Weiss, M. D., Luciano, C. A., Semino-Mora, C., et al.: Molecular mimicry in chronic inflammatory demyelinating polyneuropathy and melanoma. Neurology 51:1738, 1998. Wilkinson, M., Urich, H., and Croft, P. B.: The remote effects of cancer on the nervous system. Proc. R. Soc. Med. 60:683, 1967. Wilkinson, P. C.: Serological findings in carcinomatous neuropathy. Lancet 41:1301, 1964. Windebank, A. J., Blexrud, M. D., Dyck, P. J., et al.: The syndrome of acute sensory neuropathy: clinical features and electrophysiologic and pathologic changes. Neurology 40:584, 1990. Wolmorans, M., Shakir, T., and Ridley, S. A.: An unusual presentation of paraneoplastic neuropathy. Anaesthesia 54:616, 1999. Yagi, Y., Inuzuka, T., Takada, K., et al.: Changes in anti-HuD antibody titers in the long-term course in paraneoplastic sensory neuropathy. Eur. Neurol. 37:122, 1997. Younger, D. S.: Motor neuron disease and malignancy. Muscle Nerve 23:658, 2000. Yu, Z., Kryzer, T. J., Griesmann, G. E., et al.: CRMP-5 neuronal autoantibody: marker of lung cancer and thymomarelated autoimmunity. Ann. Neurol. 49:146, 2001.

111 Peripheral Neuropathy Associated with Lymphoma, Leukemia, and Myeloproliferative Disorders E. PETER BOSCH, THOMAS M. HABERMANN, AND AYALEW TEFFERI

Lymphoma Chemotherapy-Induced Peripheral Neuropathy Lymphomatous Nerve Infiltration (Neurolymphomatosis) Intravascular Lymphomatosis Paraneoplastic Vasculitis Acute and Chronic Inflammatory Demyelinating Polyradiculoneuropathies

Paraproteinemia in Lymphoproliferative Disorders Motor Neuron Diseases Paraneoplastic Neuropathies Neuromyotonia Paraneoplastic Cerebellar Degeneration Chronic Lymphocytic Leukemia T-Cell and Natural Killer Cell Leukemia

LYMPHOMA Lymphoma is a term used to describe a heterogeneous group of malignant neoplasms derived from lymphoid cells. The predicted number of new cases for 2002 in the United States was 53,900. Between 86% and 90% of them will be non-Hodgkin lymphoma (NHL), with the rest being Hodgkin lymphoma (HL). The number of predicted deaths per year is approximately 24,000.118 The classification of lymphomas proposed by the World Health Organization is based on histologic, immunophenotypic, and genotypic criteria.53 NHLs are a diverse group of diseases that vary in their natural history, response to treatment, and survival rate. HL has the second lowest death rate in cancer. With modern chemotherapy, 75% of patients are cured. Neurologic complications result from (1) direct involvement of the leptomeninges; (2) direct involvement of the brain; (3) compression of the spinal cord or nerve roots by epidural masses; (4) bacterial, fungal, and viral infections; (5) complications of treatment (e.g., chemotherapy, radi-

Chronic Myeloid Leukemia Acute Leukemias Acute Lymphoblastic Leukemia Acute Myelogenous Leukemia Polycythemia Vera Erythromelalgia Peripheral Neuropathies Related to Bone Marrow Transplantation

ation, bone-marrow transplantation); and (6) remote effects.112,117 In a series of nearly 1000 patients with lymphoma, spinal cord or root compression was the most frequent neurologic disorder (5%), followed by varicella zoster virus infections (4%), and toxic neuropathies related to chemotherapy (4%).24 Peripheral neuropathy unrelated to neurotoxic drugs is uncommon in both HL and NHL. Its reported frequency depends on the methods used for detection. If clinical criteria alone are used, 0.5% to 8% of patients have signs of mild neuropathy.26,152 Ten percent to 35% of patients have minor nerve conduction abnormalities if electrophysiologic studies are performed.42,152 Table 111–1 lists the many different potential causes of peripheral nerve disorders in malignant lymphomas. This classification scheme provides a framework to aid the clinician in the formulation of a differential diagnosis. Varicella-zoster virus infection and treatment complications, including radiation and chemotherapy, may result in peripheral nerve disorders that are fully discussed in other chapters. 2489

2490

Diseases of the Peripheral Nervous System

Table 111–1. Peripheral Nervous System Involvement in Lymphoma Drug neurotoxicity Herpes zoster infection Radiation Root, plexus, or nerve compression Lymphomatous infiltration: neurolymphomatosis Intravascular lymphomatosis Vasculitis Acute and chronic demyelinating polyradiculoneuropathies Paraproteinemia Motor neuron diseases Paraneoplastic syndromes Sensory neuronopathy or polyganglionopathy Brachial plexopathy Neuromyotonia Cerebellar degeneration Idiopathic neuropathy

Table 111–2. Chemotherapeutic Agents Used in Lymphoproliferative Disorders with Peripheral Neurotoxicity Drug

Clinical Features

Vinca alkaloids: vincristine, vinblastine, vindesine, vinorelbine

Sensory ⬎ motor neuropathy, autonomic symptoms, cranial mononeuropathies Large fiber sensory neuropathy/neuronopathy, hearing loss Sensory neuropathy GBS-like syndrome Sensory neuropathy Paresthesia Sensory neuropathy, brachial plexus neuropathy Sensory neuropathy, neuronopathy

Cisplatin, carboplatin

Taxanes: paclitaxel, docetaxel High-dose cytarabine (Ara-C) Etoposide (VP-16) Procarbazine Interferon alfa Thalidomide GBS ⫽ Guillain-Barré syndrome.

Varicella-zoster reactivation affects up to 20% of patients with lymphoma at some time during the course of the disease.35 The increased risk of zoster in patients with lymphoma is associated with impaired T-cell immunity against the varicella-zoster virus. Cutaneous herpes zoster requires prompt therapy with an antiviral agent to prevent dissemination and serious neurologic complications. Radiation therapy may result in injury of the brachial or lumbosacral plexus that typically develops after a latent period of 6 months or longer. The clinical features consist of slowly progressive sensorimotor deficits in the distribution of the affected plexus. Pain is prominent in brachial but not lumbosacral plexopathies, in which lower motor neuron weakness predominates.18,140 Myokymia is frequently seen on needle electromyography (EMG) in radiation-induced nerve lesions.52 Pathologic changes are characterized by epineurial fibrosis, hyalinized blood vessels, axonal loss, and demyelination.

Chemotherapy-Induced Peripheral Neuropathy Peripheral neurotoxicity is a dose-limiting side effect of a number of drugs, including vinca alkaloids, platinum compounds, and taxanes (Table 111–2).157 Vincristine is the most frequent drug to cause peripheral neuropathy in patients with lymphoma. The combination chemotherapeutic protocols CHOP (cyclophosphamide, hydroxydaunomycin [doxorubicin], Oncovin [vincristine], and prednisone) and CVP (cyclophosphamide, vincristine, and prednisone), used for the treatment of diffuse large cell and low-grade NHL, contain vincristine. Vinca alkaloids cause length-dependent axonal injury because they interfere with microtubule assembly. Vincristine pro-

duces a peripheral neuropathy in a dose-dependent manner. It has become accepted practice that individual doses of vincristine should not exceed 2 mg. It is reported that patients with hereditary neuropathies are at increased risk of developing devastating neurotoxic effects from vincristine.40 Enhanced vincristine neurotoxicity has also been described in combination with cyclosporine,12 isoniazid, and hematopoietic colony-stimulating factors.153 Cisplatin is used in aggressive combination regimens for patients in relapse. Cisplatin causes a dose-dependent, predominantly sensory neuropathy at cumulative doses of 400 mg/m2 body surface area. Typical manifestations include paresthesias, Lhermitte’s sign, proprioceptive sensory loss, and sensory ataxia combined with hearing loss, tinnitus, and autonomic symptoms. Platinum compounds have been shown to induce apoptosis in dorsal root ganglion neurons in animal models in vivo and in vitro.157 Etoposide and teniposide are semisynthetic derivatives of podophyllotoxin. Both agents may produce a mild sensory neuropathy in high doses.59,103 Procarbazine may produce paresthesia and decreased tendon reflexes, but neuropathy is not a dose-limiting factor.154 Cytosine arabinoside has been associated with a severe, demyelinating motor neuropathy when used in high doses.107 Taxanes have been evaluated in patients with refractory lymphoma.39 Paclitaxel and docetaxel promote aggregation of microtubules, thereby disrupting axonal transport. A sensory neuropathy occurs at doses greater than 200 mg/m2 body surface area.98 Thalidomide is used in the management of graft-versushost disease and certain neoplasms because of its immunomodulatory properties and effect on antiangiogenesis. It produces a sensory neuropathy or ganglionopathy.

Lymphoma, Leukemia, and Myeloproliferative Disorders

Prospective neurophysiologic studies have shown the reduction of sensory nerve action potential amplitudes to be a useful parameter for detecting early nerve damage.94

Lymphomatous Nerve Infiltration (Neurolymphomatosis) Neoplastic lymphocytic infiltration of spinal roots, dorsal root ganglia, brachial or lumbosacral plexus, and peripheral nerves is an uncommon complication of B-cell and rarely T-cell lymphomas. Currie and Henson26 reported only one case in their review of 774 patients with lymphoreticular malignancies. It may have become more common because improved chemotherapeutic regimens produce longer survival of patients. In a retrospective autopsy study of malignant lymphoma, there was histologic evidence of peripheral nerve involvement in almost 40% of cases.61 Recent clinical reviews of the subject described 40 and 52 patients, respectively.29,102 The term neurolymphomatosis was first used in the French literature to describe a patient with lymphomatous infiltrations of peripheral nerves resembling Marek’s disease or neurolymphomatosis gallinarum.48 In this paralytic disease of fowl, a T-cell lymphoma induced by a herpesvirus leads to infiltration of peripheral nerves by lymphoid cells and active demyelination.57 Currently, the term is used in the context of peripheral neuropathies that are caused by malignant lymphomatous cell infiltration of nerves, plexus, or nerve roots seen by nerve biopsy or at autopsy.29 Neurolymphomatosis occurs in four clinical settings. First, lymphoma can spread to peripheral nerves in patients with widespread extranodal disease, including central nervous system (CNS) and leptomeningeal metastases.10,29,38,63,150 Second, lymphomatous infiltration of peripheral nerves may develop in patients with systemic lymphoma after chemotherapy or as the only manifestation of relapse in patients in hematologic remission.72,95,128,143,151 This situation may arise because the intact nerve-blood barrier limits the access of chemotherapeutic agents, thereby providing a sanctuary for sequestered malignant cells. Third, infiltrated peripheral nerves may be the predominant or sole site of lymphoma involvement, without any prior systemic disease.48,60,102,120,142,150 Some experts suggest that the term neurolymphomatosis should be reserved for these cases.149 Fourth, primary leptomeningeal lymphoma may invade nerve roots, producing progressive polyradiculoneuropathies or cauda equina lesions.65,78 Neurolymphomatosis has no typical clinical presentation; depending on the site of peripheral nervous system involvement, it can present as acute ascending symmetrical paralysis closely mimicking Guillain-Barré syndrome (GBS),58 progressive polyradiculoneuropathy,48,65,142,150 cauda equina syndrome with or without bowel or bladder incontinence, cranial nerve lesions,69 distal painful sensori-

2491

motor polyneuropathy,8,33,73,161 sensory ataxic neuropathy,45 multiple mononeuropathies,72,142,151 brachial or lumbosacral plexopathies,95,134 and finally an isolated mononeuropathy.115,120,143 In some patients lymphoma is confined for months to years to a single nerve, frequently the sciatic nerve, before systemic disease evolves.32,67,110,115,120 A median mononeuropathy was reported to be the primary site of tumor recurrence after systemic chemotherapy.143 High-resolution magnetic resonance imaging (MRI) combined with fat suppression T2-weighted imaging sequences is a helpful technique to uncover such isolated peripheral nerve tumors.95 Magnetic resonance (MR) images reveal fusiform enlargement of the involved nerve segment with hyperintense signal on T2-weighted images. In solitary nerve lesions, exploration and fascicular biopsy of the peripheral nerve tumor are necessary for diagnosis. In some of the older cases from the literature, in which immunohistochemistry of the nerve infiltrates was not carried out, it is difficult to differentiate lymphoid neoplastic cells from reactive inflammation. One remarkable patient had a chronic relapsing polyneuropathy evolving over 30 years. Autopsy demonstrated lymphoid cell infiltrations of peripheral nerves and leptomeninges.16 Two patients were reported in whom a steroid-responsive chronic demyelinating neuropathy preceded the diagnosis of NHL in one and chronic lymphatic leukemia in the other.133 Distal nerve infiltration by an immunoglobulin M (IgM) ␬ B-cell lymphoma has been associated with a chronic demyelinating neuropathy.60 These polyradiculoneuropathies, characterized by lymphomatous nerve infiltrates associated with macrophage-induced demyelination, resemble more closely Marek’s disease and suggest immune-mediated demyelination in association with occult lymphoma.150 Direct neurologic complications of T-cell lymphomas consisting of parenchymal brain and leptomeningeal metastases were seen in 8% of patients with peripheral T-cell lymphoma and 2% of patients with cutaneous T-cell lymphoma in a large retrospective series from Mayo Clinic.69 Half of the patients with leptomeningeal seeding presented with polyradiculopathies. Peripheral nerve infiltration by lymphoma causing a painful distal sensorimotor polyneuropathy or bilateral facial palsies has been described in patients with cutaneous T-cell lymphoma and adult T-cell leukemia8,33,146,161 (Fig. 111–1). Primary leptomeningeal lymphoma (e.g., neoplastic meningitis without evidence of parenchymal CNS or systemic disease) is often associated with infiltration of cranial nerves and spinal nerve roots, with a predilection for the cauda equina.65 In a series of nine such patients, seven presented with a flaccid paraplegia, sensory loss, and cranial nerve palsies.79 The clinical diagnosis of neurolymphomatosis requires a high index of suspicion. Several diagnostic studies may help in clarifying the nature of the neuropathy. Electrophysiologic studies show an axonal neuropathy or,

2492

Diseases of the Peripheral Nervous System

FIGURE 111–1 A, A 55-year-old woman presented with progressive bilateral facial weakness, temporal wasting, and labial edema. The facial appearance was reminiscent of gelsolin-related familial amyloidosis. B, On buccal biopsy, there is diffuse lymphocytic infiltration. Immunohistochemistry demonstrated a cutaneous T-cell lymphoma (CD8 positive). (From Van Gerpen, J. A., Dyck, P. J. B., and Dyck, P. J.: Cutaneous T-cell lymphoma imitating gelsolin amyloidosis. Neurology 60:861, 2003, with permission.) See Color Plate

in cases of paraspinal denervation, axonal radiculoneuropathy in most patients.29 Slowed motor nerve conduction studies and conduction block have been described in the exceptional cases fulfilling diagnostic criteria for chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) associated with lymphoma.16,60,139 Cerebrospinal fluid (CSF) results may be normal, but moderate lymphocytic pleocytosis and increased protein levels are found in

half of the reported cases.29 CSF cytology is positive for tumor cells in less than one third.102 MR neurography has become a valuable technique in identifying infiltrated nerve segments by demonstrating nerve enlargement and possible sites for biopsy in patients with isolated mononeuropathies, plexopathies, or cauda equina syndrome.95,120,134 The diagnosis rests on identifying tumor cells in a peripheral nerve. Lymphoma cells are found within the

Lymphoma, Leukemia, and Myeloproliferative Disorders

endoneurium, epineurium, and perineurium associated with varying degrees of axonal degeneration.29,142,150 Immunohistochemical analysis of the cellular infiltrates is necessary to identify these cells as malignant monoclonal cells by their cell surface phenotype.142 Lymphoma tissue from 22 patients examined with cell marker studies stained positive for B-cell markers in 14, T-cell markers in 8, and natural killer (NK) cell markers in 1.102 Negative sural nerve biopsy findings have been reported in cases of autopsy-proven neurolymphomatosis caused by focal or proximal involvement of nerves.144 The possibility of infection by oncogenic viruses linked to the development of lymphoma, including human immunodeficiency virus, human T-lymphotropic virus 1 (HTLV-1), human herpesvirus-8, and Epstein-Barr virus (EBV), should be investigated.36,73 Treatment and Outcome Extranodal disease is an adverse prognostic factor in NHL,127 but because of the limited number of cases, involvement of the peripheral nervous system has not been identified as being a worse prognostic factor than disease at other extranodal sites. Neurolymphomatosis is a potentially fatal complication of lymphoma because the bloodnerve barrier theoretically limits access of cytotoxic agents. Indeed, few patients have responded to combination chemotherapy.8,60,72 If leptomeningeal lymphoma is present, intrathecal therapy is indicated. Isolated peripheral nerve lymphomas behave aggressively, and require radiation therapy combined with chemotherapy.120

Intravascular Lymphomatosis Intravascular lymphoma, or angiotropic large cell lymphoma, is a rare, high-grade lymphoma characterized by proliferation of malignant lymphocytic cells within small vessels throughout the body. The disease was originally called malignant angioendotheliomatosis before immunohistochemical studies of the malignant intravascular cells demonstrated a monoclonal lymphoid origin of either B-cell or T-cell lineage.156 The reason for the intravascular localization is poorly understood but may be explained by the absence of lymphocyte homing receptors for endothelial membrane antigens.28 There is a predilection for widespread vascular occlusions affecting the brain, spinal cord, nerve roots, peripheral nerves, and skin.34 Two thirds of patients present with neurologic disease. Systemic symptoms such as fever or weight loss and skin lesions in the form of raised hyperpigmented nodules or plaques may accompany the neurologic manifestations. Four main neurologic manifestations of intravascular lymphomatosis include multifocal cerebrovascular events, subacute encephalopathy, myelopathy with or without cauda equina syndrome, and peripheral and cranial neuropathies.34 Isolated peripheral nerve complications are rare

2493

and include progressive cauda equina lesions,3,77,97 ascending polyradiculoneuropathies,84 multiple mononeuropathies,122 symmetrical polyneuropathy,104 and isolated mononeuropathy.148 Anemia, elevated erythrocyte sedimentation rate, increased serum lactate dehydrogenase, nonspecific CSF findings of increased protein level, and moderate lymphocytic pleocytosis are frequent laboratory abnormalities.11 Lymph nodes and bone marrow are spared until late in the disease. There are usually no circulating lymphoma cells in the blood. Patients tend to have multifocal cerebral infarcts with an affinity for the deep white matter and occasional meningeal enhancement on MRI. Intravascular lymphoma may be confirmed by biopsy of involved tissues such as skin, muscle, or peripheral nerve. Without histologic confirmation, the differentiation from other vasculitic disorders is difficult. Muscle biopsy has been reported to show diagnostic changes in 72% of a small series of 18 patients.97 On nerve biopsy specimens, epineurial and perineurial medium-sized vessels and endoneurial capillaries are filled with atypical lymphomatous cells, leading to nerve ischemia with multifocal nerve fiber loss and axonal degeneration.84 Untreated, the disease is rapidly fatal, with a reported median survival of 5 months. Survival of over 2 years has been reported after CHOP chemotherapy.28,84,97,104

Paraneoplastic Vasculitis Vasculitis is recognized to occur in association with malignancies as a remote effect of cancer and usually presents as cutaneous vasculitis in leukemia and lymphoma.123 Peripheral nerve microvasculitis, defined by perivascular and intramural inflammatory infiltrates of small epineurial arterioles, is a distinctly uncommon manifestation of HL and NHL.51,62,105,150 Vincent et al.147 reported seven patients with malignancies including HL and immunoblastic lymphadenopathy in an unselected series of 50 examples of microvasculitis affecting peripheral nerves. Other cancers involved have been small cell lung cancer associated with antineuronal nuclear antibodies or anti-Hu antibodies159; adenocarcinoma of the lung, prostate, and endometrium; and renal cell carcinoma.105 Patients present with either an asymmetrical, painful sensorimotor polyneuropathy or multiple mononeuropathy that precedes the discovery of the tumor in most. Perivascular infiltrates with focal mononuclear cell invasion of the walls of small epineurial arterioles and acute axonal degeneration are the typical changes found on nerve biopsy.147 The association of microvasculitis with anti-Hu antibodies supports the possible paraneoplastic autoimmune etiology of these rare neuropathies.159 In contrast to those with malignant sensory ganglionopathies, two thirds of patients with paraneoplastic vasculitis have responded to either specific chemotherapy

2494

Diseases of the Peripheral Nervous System

of the underlying malignancy or cyclophosphamide with or without corticosteroids.105

Acute and Chronic Inflammatory Demyelinating Polyradiculoneuropathies Acute neuropathies fulfilling the diagnostic criteria for acute inflammatory demyelinating polyradiculoneuropathy (AIDP), or GBS, have sporadically been linked to lymphoma, especially HL.87,116,142,150 This association is rare. In a series of nearly 1000 lymphoma patients specifically studied for the presence of peripheral nerve complications, GBS occurred in one patient with HL.24 There have been at least 10 case reports of GBS associated with HL. The association is attributed to impaired cell-mediated immunity, which is characteristic of HL.87 In none of the anecdotal case studies were sophisticated investigations performed to look for specific antecedent viral or bacterial infections, the presence of antiglycolipid antibodies, or abnormalities of cytokines to dissect the pathogenetic mechanism. CIDP has been reported in a few patients with NHL and HL.44,133,142 A progressive demyelinating neuropathy consistent with CIDP was the first manifestation of EBVrelated lymphoma in a patient without known immunodeficiency.68 This interesting association raises questions about potential autoimmune complications of EBV-related lymphoproliferative disease. Recent reviews list lymphoma as one of the concurrent disorders that may be seen in CIDP.125 At this time there is no information about possible molecular mimicry of common antigens shared by both lymphoma and Schwann cells. This is in contrast to the more frequently reported association of CIDP with melanoma, wherein melanoma cells share common glycolipid epitopes with peripheral nerve myelin.27 The pathogenesis of both AIDP and CIDP remains unclear, and a fortuitous association with lymphoma cannot be dismissed. Nevertheless, it is important to recognize AIDP or CIDP in the clinical setting of lymphoma because both are responsive to immunomodulatory therapies.

Paraproteinemia in Lymphoproliferative Disorders The peripheral neuropathies associated with monoclonal proteins (monoclonal gammopathy of undetermined significance [MGUS], multiple myeloma, macroglobulinemia, primary amyloidosis, cryoglobulinemias, POEMS syndrome [osteosclerotic myeloma], and Castleman’s disease) are discussed in other chapters of this text. Approximately 8% of patients with a monoclonal gammopathy have a low-grade lymphoplasmacytic lymphoma or chronic lymphocytic leukemia (CLL).75 In patients with IgM MGUS, the relative risk of developing lymphoma is

2.4-fold increased as compared with a control population in the Midwestern United States.76 Even primary CNS lymphoma has been described in association with IgM monoclonal gammopathy and neuropathy.96 Over half of IgM paraproteins react with myelin-associated glycoprotein and several other components of peripheral nerve, including gangliosides and complex glycosphingolipids. These paraproteins with anti-nerve antibodies have been implicated in chronic immune-mediated neuropathies.121 The neuropathy of lymphoproliferative disorders associated with monoclonal gammopathy overlaps in clinical, electrophysiologic, and pathologic characteristics with IgM MGUS–related neuropathies. Pathologic changes of the peripheral nerve are characterized by fiber loss and demyelination. Electron microscopy shows widening of the outer myelin lamellae.142 Preliminary open trials are evaluating the role of rituximab, a mouse-human chimeric anti-CD20 monoclonal antibody that eliminates mature and malignant B cells.85,132,162 Rituximab has been approved for the management of B-cell lymphomas.

Motor Neuron Diseases Lymphoproliferative disorders seem to be overrepresented in patients with motor neuron disease (MND) compared to its rare association with other malignancies. This association was initially restricted to a purely lower motor neuron syndrome named “subacute motor neuronopathy,” occurring as a remote effect of HL and NHL.126 These patients had progressive lower motor neuron weakness affecting the legs more than the arms, with minimal or no sensory symptoms. The neurologic deficits improved spontaneously after months or years in 7 of the 10 reported patients. Pathologic studies showed anterior horn cell degeneration and demyelination of anterior roots and posterior columns. The cause of this lower motor neuron syndrome remains unclear, although radiation therapy may have been a contributing factor. Since this original observation, more than 60 patients have been reported with both diseases. More than 80% of these patients also have upper motor neuron signs qualifying them for the diagnosis of amyotrophic lateral sclerosis (ALS).37,160 The lymphoproliferative disorders associated with MND include mainly HL and NHL, CLL, and few cases of macroglobulinemia and multiple myeloma. The hematologic disorder may precede the onset of neurologic symptoms by years or be detected at presentation of the neurologic disease. Motor nerve conduction studies show reduced compound muscle action potential amplitudes with preserved conduction velocities and normal sensory conduction velocities. Needle EMG shows chronic active denervation. One half of patients have a monoclonal gammopathy, and

Lymphoma, Leukemia, and Myeloproliferative Disorders

one third an increased CSF protein level and CSF oligoclonal bands. High titers of antiganglioside antibodies may be found. A bone marrow examination should be considered in patients presenting with atypical lower motor neuron syndromes with either elevated CSF protein levels or monoclonal gammopathy to detect occult lymphoproliferative diseases.88 The diagnosis of ALS implies a poor outcome even when the lymphoma is responsive to therapy. Only a few patients with pure lower motor neuron syndromes have improved following treatment of the concurrent lymphoproliferative disorder.37,101

Paraneoplastic Neuropathies Paraneoplastic neuropathy occurring as a remote effect of lymphoma unrelated to treatment, infiltration by malignant cells, or nutritional deficiency states is a rare complication. A prospective study of 62 patients with malignant lymphoma found clinical evidence of a mild neuropathy in 8% while 35% had nerve conduction abnormalities consistent with a distal axonal neuropathy.152 Another prospective survey described nerve conduction abnormalities in 10% of patients with lymphoma.42 A prospective study of 51 patients with sensory neuropathy of unknown cause discovered underlying NHL in 6% of patients during followup 72 months after the onset of neurologic symptoms.20 The etiology of these mild sensorimotor or predominantly sensory neuropathies remains undetermined.142 Subacute Sensory Neuronopathy or Polyganglionopathy Subacute sensory neuronopathy is a distinctive paraneoplastic syndrome with asymmetrical, progressive loss of proprioceptive and kinesthetic sensation and ataxia that antedates the diagnosis of small cell lung cancer in more than 80% of cases.43 Some patients develop encephalomyelitis with widespread evidence of CNS and autonomic involvement. Although small cell lung cancer is the culprit in the majority of patients, a small proportion have lymphoma.22,56,89 The type 1 antineuronal nuclear or anti-Hu antibody detected in serum and CSF is a specific marker for paraneoplastic encephalomyelitis/sensory neuronopathy and small cell lung cancer. These autoantibodies recognize a family of RNAbinding proteins expressed in nuclei of both neurons and cancer cells. Lymphoma should be considered in seronegative anti-Hu patients with sensory neuronopathy.93 The syndrome of subacute sensory neuronopathy is fully discussed in Chapter 111. Brachial Plexus Neuropathy An immune-mediated brachial plexus neuropathy developed in two patients with HL during chemotherapy and mantle radiation.78,135 A surgical exploration and fascicular biopsy of the brachial plexus yielded endoneurial T-cell

2495

infiltrates and associated demyelination.78 One of the two patients improved with corticosteroids.

Neuromyotonia Acquired neuromyotonia is characterized by spontaneous and continuous muscle fiber activity resulting from hyperexcitability of motor nerve terminals. Patients may experience muscle stiffness, myokymia, cramps, delayed relaxation, and increased sweating. An autoimmune etiology was suspected based on its association with other autoimmune disorders (e.g., myasthenia gravis, CIDP, penicillamine induction, chronic graft-versus-host disease) and a favorable response to immunosuppressive therapy.99 The detection of serum autoantibodies against voltage-gated potassium channels establishes this disorder as an antibody-mediated ion channel syndrome.54 Paraneoplastic neuromyotonia occurs in several types of neoplasms of the immune system, including thymoma, plasmacytoma, and HL. Two cases of neuromyotonia associated with HL have been reported. Neuromyotonia was the presenting symptom in one; the other, while in remission, developed a sensory neuropathy with paresthesia and neuromyotonia.21,81 The needle electromyogram is characterized by sporadic or recurrent runs of motor unit potentials firing at rates of 150 to 300 Hz.49 Acquired neuromyotonia responds to phenytoin or carbamazepine in some but not all patients. In nonresponders, plasma exchange and immunosuppression have been successful.

Paraneoplastic Cerebellar Degeneration The most characteristic paraneoplastic manifestation of HL is cerebellar degeneration related to anti–Purkinje cell antibodies named anti-Tr antibodies.41

CHRONIC LYMPHOCYTIC LEUKEMIA B-cell CLL is the most common type of leukemia in the United States, affecting mainly patients older than 50 years. The disease is not curable but is responsive to chemotherapy. The overall neurologic complication rate is low in CLL (11%), and most complications develop in advanced stages (e.g., Rai stages 3 and 4) of the disease. In a series of nearly 1000 patients with B-cell CLL, herpes zoster infection was the most frequent neurologic disorder, followed by opportunistic infections, treatment-related complications, and direct leukemic involvement of the brain, leptomeninges, and spinal roots.19 Peripheral nervous system complications are rare (⬍1%), and consist of leukemic nerve infiltrations and immune-mediated neuropathies.

2496

Diseases of the Peripheral Nervous System

Leukemic infiltrations of peripheral nerves and spinal roots with or without associated leptomeningeal involvement may result in a fulminant axonal polyradiculoneuropathy or a sensory ataxic neuropathy.46,50,139,141 Malignant monoclonal B cells are found in perineurial, endoneurial, and perivascular infiltrates of peripheral nerves, spinal roots, and dorsal root ganglia at autopsy or, in exceptional cases, in nerve biopsy specimens.139,145 Autoimmune manifestations such as Coombs-positive autoimmune hemolytic anemia and idiopathic thrombocytopenic purpura are found with increased frequency in CLL as a result of immunoregulatory abnormalities.64 Acute GBS and chronic and relapsing CIDP have been described in patients with CLL and hairy cell leukemia.25,113,133,136,150 In a study of inflammatory neuromuscular disorders associated with CLL, Créange and co-workers25 reported another patient with AIDP and two patients presenting with mononeuropathy multiplex caused by necrotizing vasculitis. The authors pointed out the difficulty in distinguishing inflammatory from neoplastic infiltrates by conventional histologic criteria. Polymerase chain reaction analysis of immunoglobulin heavy chain gene detected clonal B-cell rearrangement in all nerve specimens, in contrast to immunohistochemistry of the lymphoid infiltrates in these nerve biopsies. This finding suggests that tumor B cells may participate in immune-mediated inflammation in lymphoproliferative disorders. A monoclonal protein is detected in about 8% of patients with CLL. Accordingly, chronic demyelinating neuropathies associated with immunoglobulin G monoclonal protein, IgM antiganglioside, and anti–myelin-associated glycoprotein antibodies have been described.5,31,92 In practice, careful evaluation, including neurophysiologic studies, CSF cytologic examination, and nerve biopsy, is necessary to determine the pathogenetic mechanism of the neuropathy of patients with CLL. Immunocytochemical studies of the inflammatory infiltrates are necessary to allow differentiation of polyclonal inflammatory cells from monoclonal lymphoid cells.139

T-Cell and Natural Killer Cell Leukemia Non–B-cell lymphocytic leukemias/lymphomas come in various types, including T-cell prolymphocytic leukemia, T-cell large granular lymphocytic leukemia, chronic NK cell lymphocytosis, hepatosplenic ␥/␦ T-cell lymphoma, Sézary’s syndrome, and adult T-cell leukemia/lymphoma. However, peripheral neuropathy in patients with these diseases is rare. A few patients with NK lymphocytosis might present with a painful axonal sensorimotor neuropathy or mononeuropathy multiplex caused by prominent inflammatory nerve infiltrates.14,130 Leukemic nerve infiltration by CD56-positive NK cells was proved in one case.80

HTLV-1 has been isolated from lymphocytes of patients of adult T-cell leukemia. A painful distal axonal sensorimotor polyneuropathy caused by distal peripheral nerve infiltration with leukemic cells has been reported in patients with adult T-cell leukemia/lymphoma.73 Treatment of the leukemia may result in improvement of the neurologic symptoms. The neuropathy associated with NK lymphocytosis has responded to high-dose corticosteroids and pulsed intravenous cyclophosphamide in anecdotal cases.130

CHRONIC MYELOID LEUKEMIA Chronic myeloid leukemia (CML) is a clonal stem cell disorder characterized by increased myeloid cells in blood and bone marrow and progressive splenomegaly. Its specific cytogenetic abnormality, the Philadelphia chromosome, is known to represent a reciprocal translocation between the long arms of chromosomes 9 and 22, designated t(9;22). CML typically follows a chronic phase, followed by a short accelerated phase and terminal blast crisis. Peripheral neuropathy is a rare complication. In the chronic phase, AIDP129 and subacute demyelinating polyneuropathy associated with monoclonal gammopathy have been reported.70,90 The role of the leukemia in the pathogenesis of these demyelinating neuropathies is uncertain but may be related to defects in humoral and cell-mediated immunity. Patients have responded to therapeutic plasmapheresis or corticosteroids. Leukemic infiltration of distal nerves has been confirmed by nerve biopsy in a patient who presented with a sensory neuropathy.17

ACUTE LEUKEMIAS Neurologic complications from acute leukemias stem from hemorrhage into the brain or isolated nerve trunks; leukemic infiltration of the brain, leptomeninges, cranial nerves, spinal roots, and peripheral nerves; CNS infections; and chemotherapy-related neurotoxicity.13

Acute Lymphoblastic Leukemia Acute lymphoblastic leukemia (ALL) is an aggressive neoplasm that has been defined by the presence of more than 30% lymphoblasts in bone marrow.82 ALL occurs most frequently in children, and accounts for 20% of acute leukemias in adults. About 30% of patients develop clinical evidence of CNS involvement at some stage of their disease. Most commonly, single or multiple cranial nerve deficits or spinal root lesions occur, with asymmetrical motor and sensory deficits. The CSF

Lymphoma, Leukemia, and Myeloproliferative Disorders

contains leukemic cells. The pathology is one of leptomeningeal and spinal root infiltration by leukemic blast cells.111 Rarely, a polyradiculopathy with leptomeningeal leukemia may antedate the diagnosis of ALL by several months.9 The incidence of CNS and meningeal leukemia has been greatly reduced by the routine administration of prophylactic intrathecal therapy or CNS radiotherapy. In contrast, leukemic peripheral nerve infiltration is rare. Involvement of the inferior alveolar and mental nerves causing unilateral or bilateral numbness of the chin and lower lip can be the presenting symptom of ALL.55 An acute axonal polyneuropathy with facial diplegia mimicking GBS has been reported as the initial manifestation of a leukemic relapse.15 The CSF examination was repeatedly normal, but the peripheral nerve biopsy demonstrated leukemic cells in a subperineurial location. The peripheral nervous system with its blood-nerve barrier may have provided a pharmacologic sanctuary for leukemic cells in this unusual case.

Acute Myelogenous Leukemia Acute myelogenous leukemia (AML) is characterized by proliferation and accumulation of immature hematopoietic cells in bone marrow and peripheral blood. Myeloid leukemic blasts are classified by morphology, immunohistochemistry, and immunophenotyping. About 5% to 7% of patients have meningeal involvement at presentation. High blast counts and monocytic morphology, particularly the M4 subtype with abnormal eosinophils, predispose to CNS involvement.13 Symmetrical sensorimotor polyradiculoneuropathies caused by leukemic infiltration of peripheral nerves have been described in single case reports in acute monocytic, myelomonocytic, and megakaryocytic subtypes.72,100 Patients in hematologic remission may develop acute axonal polyneuropathy leading to quadriplegia and respiratory failure or multiple mononeuropathies caused by leukemic infiltrations of distal nerves as the initial presentation of leukemia in relapse.72,83 Granulocytic sarcoma, or chloroma, is a solid tumor formed by myeloid precursor cells associated with AML. Granulocytic sarcomas can develop in the epidural space, leading to spinal cord or cauda equina compression.124 A few cases of AIDP have been described in association with AML.109,119 Differentiation from the more common polyradiculoneuropathies caused by leukemic nerve infiltration is of utmost clinical importance.

POLYCYTHEMIA VERA Polycythemia vera (PV) is a chronic, clonal myeloproliferative disorder characterized by erythropoietin-independent erythrocytosis that is often accompanied by splenomegaly,

2497

thrombocytosis, and leukocytosis. The disease is associated with a progressive risk of transformation to either myelofibrosis with myeloid metaplasia or AML.138 Untreated patients have a high incidence of arterial and venous thrombotic events, and are at increased risk for myocardial infarction and ischemic strokes.47 Although positive sensory symptoms such as pruritus, paresthesias, and burning feet are common complaints, well-documented peripheral neuropathy is thought to be a rare complication of the disease.26 PV-associated pruritus, which is typically precipitated by contact with water, occurs in 50% of patients with PV, and may be an agonizing aspect of the disease.30 Its pathogenesis is unclear, although the beneficial response to selective serotonin reuptake inhibitors suggests a possible role for serotonin in PV-associated pruritus. Paroxetine and fluoxetine have been used effectively to alleviate pruritus.137 Among the cytoreductive agents, only interferon alfa has been reported to be effective in controlling pruritus. A mild, distal sensory neuropathy has been described in two patient series with PV.114,158 Paresthesias and burning feet have been reported by 27% to 46% of patients, whereas fewer had clinical signs of mild neuropathy (e.g., absent distal tendon reflexes and distal sensory loss). In a prospective study of 28 patients, 70% had nerve conduction abnormalities consisting of reduced sural nerve action potential amplitudes and mildly prolonged distal motor latencies.114 Sural nerve biopsy showed diffuse myelinated fiber loss and low-grade axonal degeneration consistent with a mild axonal process.158 The pathogenesis by which PV induces these alterations is unknown, although most authors postulate an ischemic mechanism resulting from increased blood viscosity.158 It is unknown whether phlebotomy or cytoreductive therapy has any beneficial effect on the mild sensory neuropathy in PV.

Erythromelalgia Secondary erythromelalgia occurs most commonly in association with myeloproliferative disorders with elevated platelet counts, including PV and essential thrombocythemia.2,91 Erythromelalgia is the term used to describe a condition characterized by intense burning pain of the feet and hands associated with erythema and increased skin temperature.66 Symptoms are aggravated by exposure to heat and relieved by cooling. Erythromelalgia may precede the diagnosis of a myeloproliferative disorder by several months to years.74 There are no reports of neurophysiologic studies in this condition. Skin biopsies have concentrated on microvascular abnormalities, describing arteriolar narrowing and vascular occlusions by platelet thrombi.145 Investigations

2498

Diseases of the Peripheral Nervous System

of epidermal nerve fiber density have not been published. A single dose of aspirin (500 mg) provides prompt relief for several days. Low-dose aspirin (50 mg) or cytoreductive therapy that is aimed to lower platelet counts to normal values prevents recurrences.145 Either hydroxyurea or anagrelide can be used to lower platelet counts effectively.138

PERIPHERAL NEUROPATHIES RELATED TO BONE MARROW TRANSPLANTATION Bone marrow transplantation or peripheral blood stemcell transplantation has become the standard of care in relapsed intermediate or high-grade NHL, recurrent HL, and a number of leukemias.7 Neurologic complications have been reported to occur in 50% to 70% of patients undergoing allogeneic bone marrow transplantation and in a smaller proportion of patients after autologous bone marrow or peripheral blood stem-cell transplantation.108,131 Peripheral neuropathy is an uncommon complication. In a prospective study of 115 patients with leukemia undergoing allogeneic bone marrow transplantation, 4% developed neuropathy in the first 3 months after transplantation.6 Yet there are multiple causes of potential neuromuscular complications developing in the posttransplant period: neurotoxicity of drugs used in conditioning regimens, critical illness neuropathy, steroid myopathy, and immune-mediated polyradiculoneuropathies associated with graft-versus-hostdisease106 (Table 111–3). Neurotoxicity of the conditioning regimens (e.g., etoposide, cisplatin, paclitaxel) and thalidomide treatment for graft-versus-host disease may

result in predominantly axonal sensory neuropathies.59 A severe motor neuropathy together with cerebellar toxicity has been attributed to pretreatment with high-dose cytosine arabinoside.107 In these patients peripheral nerve demyelination is documented based on either nerve conduction studies or nerve biopsy findings. Because of the increased risk of sepsis and multiorgan failure, critical illness neuropathy must be considered in the posttransplant period. Some patients require treatment with megadose corticosteroids for a variety of transplant-related complications, and thus are at risk of developing acute quadriplegic myopathy or steroid myopathy. AIDP and CIDP typically occur in the clinical setting of graft-versus-host disease. In graft-versus-host disease, immunologically competent donor T cells attack host tissues, mainly the skin, gastrointestinal system, and liver.7 Polymyositis, myasthenia gravis, neuromyotonia, and CIDP have been described in association with chronic graft-versus-host disease.1,4,86 The observed predominantly motor polyradiculoneuropathy meets clinical and laboratory criteria for a diagnosis of CIDP.23 Patients improve after immunomodulatory treatment consisting of either plasmapheresis or intravenous immune globulin, or immunosuppressant therapy. Experience is too limited to suggest specific regimens. AIDP or GBS has been reported in patients after both allogeneic and autologous bone marrow transplantation.155 The increased susceptibility to infections and defects in both cell-mediated and humoral immunity may increase the risk of developing AIDP in these patients. Intravenous immune globulin or plasmapheresis is considered effective treatment.

Table 111–3. Neuromuscular Complications Related to Bone Marrow Transplantation Etiology

NC Studies

Clinical Features

Treatment

Drug toxicity High-dose Ara-C Cisplatin Etoposide (VP16) Paclitaxel Thalidomide Critical illness neuropathy

Demyelinating Axonal Axonal Axonal Axonal Axonal

Supportive care Supportive care Supportive care Supportive care Supportive care Supportive care

IVIG Plasmapheresis IVIG Plasmapheresis Immunosuppression

Acute quadriplegic myopathy GBS

Normal

GBS-like Sensory Sensory Sensory Sensory Difficulty weaning from respirator Flaccid weakness

Demyelinating

Motor ⬎ sensory PRN

CIDP

Demyelinating

GVHD, motor ⬎ sensory PRN

Reduction of steroids

Ara-C ⫽ cytosine arabinoside; CIDP ⫽ chronic inflammatory polyradiculoneuropathy; GBS ⫽ Guillain-Barré syndrome; GVHD ⫽ graft-versus-host disease; IVIG ⫽ intravenous immune globulin; NC ⫽ nerve conduction; PRN ⫽ polyradiculoneuropathy.

Lymphoma, Leukemia, and Myeloproliferative Disorders

REFERENCES 1. Adams, C., August, C. S., Maguire, H., and Sladky, J. T.: Case reports: neuromuscular complications of bone marrow transplantation. Pediatr. Neurol. 12:58, 1995. 2. Alarcon-Segovia, D., Babb, R. R., Fairbairn, F. II, and Hagedorn, A. B.: Erythermalgia: a clue to the early diagnosis of myeloproliferative disorders. Arch. Intern. Med. 117:511, 1966. 3. Al-Shahi, R., Warlow, C. P., Jansen, G. H., et al.: A 59-yearold man with progressive spinal cord and peripheral nerve dysfunction culminating in encephalopathy: Edinburgh Advanced Clinical Neurology Course, 1999. J. Neurol. Neurosurg. Psychiatry 71:696, 2001. 4. Amato, A. A., Barohn, R. J., Sahenk, Z., et al.: Polyneuropathy complicating bone marrow and solid organ transplantation. Neurology 43:1513, 1993. 5. Andres, E., Vinzio, S., Maloisel, F., et al.: Autoimmune peripheral neuropathies with anti-MAG antibodies and hematological disorders: five cases. Ann. Med. Interne (Paris) 152:147, 2001. 6. Antonini, G., Ceschin, V., Morino, S., et al.: Early neurologic complications following allogeneic bone marrow transplant for leukemia. Neurology 50:1441, 1998. 7. Armitage, J. O.: Bone marrow transplantation. N. Engl. J. Med. 330:827, 1994. 8. Atiq, O. T., DeAngelis, L. M., Rosenblum, M., and Portlock, C. S.: Cutaneous T-cell lymphoma presenting with diffuse lymphomatous infiltration of the peripheral nerves: response to combination chemotherapy. Am. J. Clin. Oncol. 15:212, 1992. 9. Banerjee, P., MacCallum, P. K., Walker, R. W. H., and Lister, T. A.: Polyradiculopathy complicating acute lymphoblastic leukemia. Hematol. Oncol. 13:45, 1995. 10. Batchelor, T. T., and Louis, D. N.: Case records of the Massachusetts General Hospital. Case 39–1997: a 67-yearold woman with cauda equina syndrome. N. Engl. J. Med. 337:1829, 1997. 11. Beristain, X., and Azzarelli, B.: The neurological masquerade of intravascular lymphomatosis. Arch. Neurol. 59:439, 2002. 12. Bertrand, Y., Capdeville, R., Bulduck, N., and Phillipe, N.: Cyclosporin A used to reverse drug resistance increases peripheral neuropathy. Am. J. Hematol. 40:158, 1992. 13. Blumenthal, D. T., and Glenn, M. J.: Neurologic manifestations of hematologic disorders. Neurol. Clin. 20:265, 2002. 14. Bobker, D. H., and Deloughery, T. G.: Natural killer cell leukemia presenting with a peripheral neuropathy. Neurology 43:1853, 1993. 15. Boiron, J.-M., Ellie, E., Vital, A., et al.: Isolated peripheral nerve relapse masquerading as Guillain-Barré syndrome in a patient with acute lymphoblastic leukemia. Leuk. Lymphoma 10:489, 1993. 16. Borit, A., and Altrocchi, P. H.: Recurrent polyneuropathy and neurolymphomatosis. Arch. Neurol. 24:40, 1971. 17. Bouche, P., Raphael, M., Hauw, J. J., and Derouesne, C.: Neuropathie périphérique révélatrice d’une leucose myéloïde chronique. Nouv. Presse Med. 8:2201, 1979. 18. Bowen, J., Gregory, R., Squier, M. and Donaghy, M.: The post-irradiation lower motor neuron syndrome: neuronopathy or radiculopathy? Brain 119:1429, 1996.

2499

19. Bower, J. H., Hammack, J. E., McDonnell, S. K., and Tefferi, A.: The neurologic complications of B-cell chronic lymphocytic leukemia. Neurology 48:407, 1997. 20. Camerlingo, M., Nemni, R., Ferraro, B., et al.: Malignancy and sensory neuropathy of unexplained cause. Arch. Neurol. 55:981, 1998. 21. Caress, J. B., Abend, W. K., Preston, D. C., and Logigian, E. L.: A case of Hodgkin’s lymphoma producing neuromyotonia. Neurology 49:258, 1997. 22. Chalk, C. H., Windebank, A. J., Kimmel, D. W., and McManis, P. G.: The distinctive clinical features of paraneoplastic sensory neuronopathy. Can. J. Neurol. Sci. 19:346, 1992. 23. Cornblath, D. R., Asbury, A. K., Albers, J. W., et al.: Research criteria for diagnosis of chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). Neurology 41:617, 1991. 24. Correale, J., Monteverde, D. A., Bueri, J. A., and Reich, E. G.: Peripheral nervous system and spinal cord involvement in lymphoma. Acta Neurol. Scand. 83:45, 1991. 25. Créange, A., Théodorou, I., Sabourin, J.-C., et al.: Inflammatory neuromuscular disorders associated with chronic lymphoid leukemia: evidence for clonal B cells within muscle and nerve. J. Neurol. Sci. 137:35, 1996. 26. Currie, S., and Henson, R. A.: Neurological syndromes in the reticuloses. Brain 94:307, 1971. 27. Dalakas, M. C.: Advances in chronic inflammatory demyelinating polyneuropathy: disease variants and inflammatory response mediators and modifiers. Curr. Opin. Neurol. 12:403, 1999. 28. Demirer, T., Dail, D. H., and Aboulafia, D. M.: Four varied cases of intravascular lymphomatosis and a literature review. Cancer 73:1738, 1994. 29. Diaz-Arrastia, R., Younger, D. S., Hair, L., et al.: Neurolymphomatosis: a clinicopathologic syndrome re-emerges. Neurology 42:1136, 1992. 30. Diehn, F., and Tefferi, A.: Pruritus in polycythaemia vera: prevalence, laboratory correlates and management. Br. J. Haematol. 115:619, 2001. 31. Drake, W. M., Monson, J. P., Trainer, P. J., et al.: Acute polyneuropathy with chronic lymphocytic leukaemia and paraproteinaemia: response to chlorambucil and prednisolone. J. Neurol. Neurosurg. Psychiatry 64:564, 1998. 32. Eusebi, V., Bondi, A., Cancellieri, A., et al.: Primary malignant lymphoma of sciatic nerve: report of a case. Am. J. Surg. Pathol. 14:881, 1990. 33. Gherardi, R., Gaulard, P., Prost, C., et al.: T-cell lymphoma revealed by a peripheral neuropathy: a report of two cases with an immunohistologic study on lymph node and nerve biopsies. Cancer 58:2710, 1986. 34. Glass, J., Hochberg, F. H., and Miller, D. C.: Intravascular lymphomatosis: a systemic disease with neurologic manifestations. Cancer 71:3156, 1993. 35. Goffinet, D. R., Glatstein, E. J., and Merigan, T. C.: Herpes zoster-varicella infections and lymphoma. Ann. Intern. Med. 76:235, 1972. 36. Gold, J. E., Jimenez, E., and Zalusky, R.: Human immunodeficiency virus-related lymphoreticular malignancies and peripheral neurologic disease: a report of four cases. Cancer 61:2318, 1988.

2500

Diseases of the Peripheral Nervous System

37. Gordon, P. H., Rowland, L. P., Younger, D. S., et al.: Lymphoproliferative disorders and motor neuron disease: an update. Neurology 48:1671, 1997. 38. Gordon, P. H., and Younger, D. S.: Neurolymphomatosis. Neurology 46:1191, 1996. 39. Goss, P., Stewart, A. K., Couture F., et al.: Combined results of two Phase II studies of Taxol (paclitaxel) in patients with relapsed or refractory lymphoma. Leuk. Lymphoma 34:295, 1999. 40. Graf, W. D., Chance, P. F., Lensch, M. W., et al.: Severe vincristine neuropathy in Charcot-Marie-Tooth disease type 1 A. Cancer 77:1356, 1996. 41. Graus, F., Dalmau, J., Valldeoriola, F., et al.: Immunological characterization of a neuronal antibody (anti-Tr) associated with paraneoplastic cerebellar degeneration and Hodgkin’s disease. J. Neuroimmunol. 74:55, 1997. 42. Graus, F., Ferrer, I., and Lamarca, J.: Mixed carcinomatous neuropathy in patients with lung cancer and lymphoma. Acta Neurol. Scand. 68:40, 1983. 43. Graus, F., Keime-Guibert, F., Reñe, R., et al.: Anti-Huassociated paraneoplastic encephalomyelitis: analysis of 200 patients. Brain 124:1138, 2001. 44. Griggs, J. J., Commichau, C. S., Rapoport, A. P., and Griggs, R. C.: Chronic inflammatory demyelinating polyneuropathy in non-Hodgkin’s lymphoma. Am. J. Hematol. 54:332, 1997. 45. Grisold, W., Drlicek, M., and Jellinger, K.: Neurolymphomatosis. Neurology 43:1628, 1993. 46. Grisold, W., Jellinger, K., and Lutz, D.: Human neurolymphomatosis in a patient with chronic lymphatic leukemia. Clin. Neuropathol. 9:224, 1990. 47. Gruppo Italiano Studio Policitemia: Polycythemia vera: the natural history of 1213 patients followed for 20 years. Ann. Intern. Med. 123:656, 1995. 48. Guberman, A., Rosenbaum, H., Braciale, T., and Schlaepfer, W. W.: Human neurolymphomatosis. J. Neurol. Sci. 36:1, 1978. 49. Gutmann, L., Libell, D., and Gutmann, L.: When is myokymia neuromyotonia? Muscle Nerve 24:151, 2001. 50. Haberland, C., Cipriani, M., Kucuk, O., et al.: Fulminant leukemic polyradiculoneuropathy in a case of B-cell prolymphocytic leukemia. Cancer 60:1454, 1987. 51. Harati, Y., and Niakan, E.: The clinical spectrum of inflammatory-angiopathic neuropathy. J. Neurol. Neurosurg. Psychiatry 49:1313, 1986. 52. Harper, M. C., Thomas, J. E., Cascino, T. L., and Litchy, W. J.: Distinction between neoplastic and radiation-induced brachial plexopathy, with emphasis on the role of EMG. Neurology 39:502, 1989. 53. Harris, N. L., Jaffe, E. S., Diebold, J., et al.: World Health Organization classification of the hematopoietic and lymphoid tissues: report of the clinical advisory committee meeting—Airlie House, Virginia, November 1997. J. Clin. Oncol. 17:3835, 1999. 54. Hart, I. K., Water, C., Vincent, A., et al.: Autoantibodies detected to expressed K channels are implicated in neuromyotonia. Ann. Neurol. 41:238, 1997. 55. Hiraki, A., Nakamura, S., Abe, K., et al.: Numb chin syndrome as an initial symptom of acute lymphocytic leukemia. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 83:555, 1997.

56. Horwich, M. S., Cho, L., Porro, R. S., and Posner, J. B.: Subacute sensory neuropathy: a remote effect of carcinoma. Ann. Neurol. 2:7, 1977. 57. Hughes, R. A. C.: Animal models of Guillain-Barré syndrome. In Guillain-Barré Syndrome. Heidelberg, Springer Verlag, p. 249, 1990. 58. Hughes, R. A. C., Britton, T., and Richards, M.: Effects of lymphoma on the peripheral nervous system. J. R. Soc. Med. 87:526, 1994. 59. Imrie, K. R., Couture, F., Turner, C. C., et al.: Peripheral neuropathy following high-dose etoposide and autologous bone marrow transplantation. Bone Marrow Transplant. 13:77, 1994 60. Ince, P. G., Shaw, P. J., Fawcett, P. R. W., and Bates, D.: Demyelinating neuropathy due to primary IgM kappa B cell lymphoma of peripheral nerve. Neurology 37:1231, 1987. 61. Jellinger, K., and Radaszkiewicz, T.: Involvement of the central nervous system in malignant lymphomas. Virchows Arch. A 370:345, 1976. 62. Johnson, P. C., Rolak, L. A., Hamilton, R. H., and Laguna, J. F.: Paraneoplastic vasculitis of nerve: a remote effect of cancer. Ann. Neurol. 5:437, 1979. 63. Jones, H. R., and Edgar, M. A..: Case records of the Massachusetts General Hospital. Case 8–1995: a 75-yearold man with multiple cranial-nerve defects and weakness and pain in the extremities. N. Engl. J. Med. 332:730, 1995. 64. Jønsson, V., Svendsen, B., Vorstrup, S., et al.: Multiple autoimmune manifestations in monoclonal gammopathy of undetermined significance and chronic lymphocytic leukemia. Leukemia 10:327, 1996. 65. Julien, J., Vital, C., Rivel, J., et al.: Primary meningeal B lymphoma presenting as a subacute ascending polyradiculoneuropathy. J. Neurol. Neurosurg. Psychiatry 54:610, 1991. 66. Kalgaard, O. M., Seem, E., and Kvernebo, K.: Erythromelalgia: a clinical study of 87 cases. J. Intern. Med. 242:191, 1997. 67. Kanamori, M., Matsui, H., and Yudoh, K.: Solitary T-cell lymphoma of the sciatic nerve: case report. Neurosurgery 36:1203, 1995. 68. Kasamon, Y. L., Nguyen, T. N., Chan, J. A., and Nascimento, A. F.: EBV-associated lymphoma and chronic inflammatory demyelinating polyneuropathy in an adult without overt immunodeficiency. Am. J. Hematol. 69:289, 2002. 69. Kaufman, D. K., Habermann, T. M., Kurtin, P. J., and O’Neill, B. P.: Neurological complications of peripheral and cutaneous T-cell lymphomas. Ann. Neurol. 36:625, 1994. 70. Konstadoulakis, M. M., Papasavvas, P., Bitsaktis, A., et al.: Acute demyelinating peripheral neuropathy in a patient with double monoclonal gammopathy and chronic myelomonocytic leukemia. Muscle Nerve 16:431, 1993. 71. Krendel, D. A., Albright, R. E., and Graham, D. G.: Infiltrative polyneuropathy due to acute monoblastic leukemia in hematologic remission. Neurology 37:474, 1987. 72. Krendel, D. A., Stahl, R. L., and Chan, W. C.: Lymphomatous polyneuropathy: biopsy of clinically involved nerve and successful treatment. Arch. Neurol. 48:330, 1991. 73. Kuroda, Y., Nakata, H., Kakigi, R., et al.: Human neurolymphomatosis by adult T-cell leukemia. Neurology 39:144, 1989.

Lymphoma, Leukemia, and Myeloproliferative Disorders 74. Kurzrock, R., and Cohen, P. R.: Erythromelalgia and myeloproliferative disorders. Arch. Intern. Med. 149:105, 1989. 75. Kyle, R. A. Monoclonal proteins in neuropathy. Neurol. Clin. 10:713, 1992. 76. Kyle, R. A., Therneau, T. M., Rajkumar, S. V., et al.: A longterm study of prognosis in monoclonal gammopathy of undetermined significance. N. Engl. J. Med. 346:564, 2002. 77. Lachance, D. H., and Louis, D. N.: Case Records of the Massachusetts General Hospital. Weekly Clinicopathological Exercises. Case 31–1995: a 43-year-old man with multifocal neurologic problems and confusion. N. Engl. J. Med. 333:992, 1995. 78. Lachance, D. H., O’Neill, B. P., Harper, C. M., et al: Paraneoplastic brachial plexopathy in a patient with Hodgkin’s disease. Mayo Clin. Proc. 66:97, 1991. 79. Lachance, D. H., O’Neill, B. P., Macdonald, D. R., et al.: Primary leptomeningeal lymphoma: report of 9 cases, diagnosis with immunocytochemical analysis, and review of the literature. Neurology 41:95, 1991. 80. Lackowski, D., Koberda, J. L., DeLoughery, T. G., and So, Y.: Natural killer cell leukemia as a cause of peripheral neuropathy and meningitis: case report. Neurology 51:640, 1998. 81. Lahrmann, H., Albrecht, G., Drlicek, M., et al.: Acquired neuromyotonia and peripheral neuropathy in a patient with Hodgkin’s disease. Muscle Nerve 24:834, 2001. 82. Lai, R., Hirsch-Ginsberg, C. F., and Bueso-Ramos, C.: Advances in the treatment of adult acute lymphocytic leukemia: pathologic diagnosis of acute lymphocytic leukemia. Hematol. Oncol. Clin. North Am. 14:1209, 2000. 83. Lekos, A., Katirji, M. B., Cohen, M. L., et al.: Mononeuritis multiplex: a harbinger of acute leukemia in relapse. Arch. Neurol. 51:618, 1994. 84. Levin, K. H., and Lutz, G.: Angiotropic large-cell lymphoma with peripheral nerve and skeletal muscle involvement: early diagnosis and treatment. Neurology 47:1009, 1996. 85. Levine, T. D., and Pestronk, A.: IgM antibody-related polyneuropathies: B-cell depletion chemotherapy using rituximab. Neurology 52:1701, 1999. 86. Liguori, R., Vincent, A., Avoni, P., et al.: Acquired neuromyotonia after bone marrow transplantation. Neurology 54:1390, 2000. 87. Lisak, R. P., Mitchell, M., Zweiman, B., et al.: GuillainBarré syndrome and Hodgkin’s disease: three cases with immunological studies. Ann. Neurol. 1:72, 1977. 88. Louis, E. D., Hanley, A. E., Brannagan, T. H., et al.: Motor neuron disease, lymphoproliferative disease, and bone marrow biopsy. Muscle Nerve 19:1334, 1996. 89. Lucchinetti, C. F., Kimmel, D. W., and Lennon, V. A.: Paraneoplastic and oncologic profiles of patients seropositive for type 1 antineuronal nuclear autoantibodies. Neurology 50:652, 1998. 90. Maeda, T., Ashie, T., Kikuiri, K., et al.: Chronic myelomonocytic leukemia with polyneuropathy and IgA-paraprotein. Jpn. J. Med. 28:709, 1989. 91. Michiels, J. J., and ten Kate, F. J. W.: Erythromelalgia in thrombocythemia of various myeloproliferative disorders. Am. J. Hematol. 39:131, 1992.

2501

92. Mitsui, Y., Kusunoki, S., Hiruma, S., et al.: Sensorimotor polyneuropathy associated with chronic lymphocytic leukemia, IgM antiganglioside antibody and human T-cell leukemia virus I infection. Muscle Nerve 22:1461, 1999. 93. Molinuevo, J. L., Graus, F., Serrano, C., et al.: Utility of anti-Hu antibodies in the diagnosis of paraneoplastic sensory neuropathy. Ann. Neurol. 44:976, 1998. 94. Molloy, F. M., Floeter, M. K., Syed, N. A., et al.: Thalidomide neuropathy in patients treated for metastatic prostate cancer. Muscle Nerve 24:1050, 2001. 95. Moore, K., Blumenthal, D. T., Smith, A. G., and Ward, J. H.: Neurolymphomatosis of the lumbar plexus: high-resolution MR neurography findings. Neurology 57:740, 2001. 96. Naccache, L., Thibault, S., Hoang-Xuan, K., et al.: Cerebral lymphoma in a patient with immunoglobulin paraproteinemic polyneuropathy. Muscle Nerve 20:122,1997. 97. Nakahara, T., Saito, T., Muroi, A., et al.: Intravascular lymphomatosis presenting as an ascending cauda equina: conus medullaris syndrome: remission after biweekly CHOP therapy [short report]. J. Neurol. Neurosurg. Psychiatry 67:403, 1999. 98. New, P. Z., Jackson, C. E., Rinaldi, D., et al.: Peripheral neuropathy secondary to docetaxel (Taxotere). Neurology 46:108, 1996. 99. Newsom-Davis, J., and Mills, K. R.: Immunological associations of acquired neuromyotonia (Isaacs’ syndrome): report of five cases and literature review. Brain 116:453, 1993. 100. Nishi, Y., Yufu, Y., Shinomiya, S., et al.: Polyneuropathy in acute megacaryoblastic leukemia. Cancer 68:2033, 1991. 101. Nobile-Orazio, E., Carpo, M., and Meucci, N.: Are there immunologically treatable motor neuron diseases? Amyotroph. Lateral Scler. Other Motor Neuron Disord. 2(Suppl. 1):S23, 2001. 102. Odabasi, Z., Parrott, J. H., Reddy, V. V. B., and Oh, S. J.: Neurolymphomatosis associated with muscle and cerebral involvement caused by natural killer cell lymphoma: a case report and review of literature. J. Peripher. Nerv. Syst. 6:197, 2001. 103. O’Dwyer, P. J., Leyland-Jones, B., Alonso, M. T., et al.: Drug Therapy. Etoposide (VP-16–213): current status of an active anticancer drug. N. Engl. J. Med. 312:692, 1985. 104. Oei, M. E., Kraft, G. H., and Sarnat, H. B.: Intravascular lymphomatosis. Muscle Nerve 25:742, 2002. 105. Oh, S. J.: Paraneoplastic vasculitis of the peripheral nervous system. Neurol. Clin. 15:849, 1997. 106. Openshaw, H.: Peripheral neuropathy after bone marrow transplantation. Biol. Blood Marrow Transplant. 3:202, 1997. 107. Openshaw, H., Slatkin, N. E., Stein, A. S., et al.: Acute polyneuropathy after high-dose cytosine arabinoside in leukemia. Cancer 78:1899, 1996. 108. Patchell, R. A.: Neurological complications of organ transplantation. Ann. Neurol. 36:688, 1994. 109. Phanthumchinda, K., Intragumtornchai, T., and Kasantikul, V.: Guillain-Barré syndrome and optic neuropathy in acute leukemia. Neurology 38:1324, 1988. 110. Pillay, P. K., Hardy, R.W. Jr., Wilbourn, A. J., et al.: Solitary primary lymphoma of the sciatic nerve: case report. Neurosurgery 23:370, 1988.

2502

Diseases of the Peripheral Nervous System

111. Pochedly, C.: Neurologic manifestations in acute leukemia. III. Peripheral neuropathy and chloroma. N. Y. State J. Med. 75:878, 1975. 112. Pollard, J. D., and Young, G. A. R.: Neurology and the bone marrow. J. Neurol. Neurosurg. Psychiatry 63:706, 1997. 113. Powles, R. L., and Malpas, J. S.: Guillain-Barré syndrome associated with chronic lymphatic leukemia. Br. Med. J. 3:286, 1967. 114. Poza, J. J., Cobo, A. M., and Martí-Massó, J. F.: Peripheral neuropathy associated with polycythemia vera. Neurología 11:10, 1996. 115. Quiñones-Hinojosa, A., Friedlander R. M., Boyer, P. J., et al.: Solitary sciatic nerve lymphoma as an initial manifestation of diffuse neurolymphomatosis: case report and review of the literature. J. Neurosurg. 92:165, 2000. 116. Re, D., Schwenk, A., Hegener, P., et al.: Guillain-Barré syndrome in a patient with non-Hodgkin’s lymphoma. Ann. Oncol. 11:217, 2000. 117. Recht, L. D.: Neurological complications of lymphoma. In Wiley, R. G. (ed.): Neurological Complications of Cancer (Neurological Diseases and Therapy, Vol. 37). New York, Marcel Decker, p. 465, 1995. 118. Ries, L. A. G., Eisner, M. P., Kosary, C. L., et al. (eds.): SEER Cancer Statistics Review, 1975–2000. Bethesda, MD, National Cancer Institute, 2003. Available at: seer.cancer.gov/csr/1975_2000 119. Rodrigues, A., Monteiro, A., Viana, J., et al.: Acute nonlymphoblast leukaemia presenting as a Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 56:935, 1993. 120. Roncaroli, F., Poppi, M., Riccioni, L., and Frank, F.: Primary non-Hodgkin’s lymphoma of the sciatic nerve followed by localization in the central nervous system: case report and review of literature. Neurosurgery 40:618, 1997. 121. Ropper, A. H., and Gorson, K. C.: Current concepts: neuropathies associated with paraproteinemia. N. Engl. J. Med. 338:1601, 1998. 122. Roux, S., Grossin, M., De Bandt, M., et al.: Angiotropic large cell lymphoma with mononeuritis multiplex mimicking systemic vasculitis. J. Neurol. Neurosurg. Psychiatry 58:363, 1995. 123. Sánchez-Guerrero, J., Gutiérrez-Ureña, S., Vidaller, A., et al.: Vasculitis as a paraneoplastic syndrome: report of 11 cases and review of the literature. J. Rheumatol. 17:1458, 1990. 124. Sandhu, G. S., Ghufoor, K., Gonzalez-Garcia, J., and Elexpuru-Camiruaga, J. A.: Granulocytic sarcoma presenting as cauda equina syndrome. Clin. Neurol. Neurosurg. 100:205, 1998. 125. Saperstein, D. S., Katz, J. S., Amato, A. A., and Barohn, R. J.: Clinical spectrum of chronic acquired demyelinating polyneuropathies. Muscle Nerve 24:311, 2001. 126. Schold, S. C., Cho, E.-S., Somasundaram, M., and Posner, J. B.: Subacute motor neuronopathy: a remote effect of lymphoma. Ann. Neurol. 5:271, 1979. 127. Shipp, M. A.: Prognostic factors in aggressive nonHodgkin’s lymphoma: who has “high-risk” disease? Blood 83:1165, 1994. 128. Shoenfeld, Y., Aderka, D., Sandbank, U., et al.: Fatal peripheral neurolymphomatosis after remission of histiocytic lymphoma. Neurology 33:243, 1983.

129. Silber, E., Lane, R. J. M., and Hopkinson, N. S.: Demyelinating polyneuropathy in a patient with chronic myeloid leukemia. Muscle Nerve 21:974, 1998. 130. Smith, B.: Multifocal neuropathy associated with natural killer cell leukemia: response to therapy with intravenous methylprednisolone and cyclophosphamide. J. Neurol. Sci. 187:S379, 2001. 131. Snider, S., Bashir, R., and Bierman, P.: Neurologic complications after high-dose chemotherapy and autologous bone marrow transplantation for Hodgkin’s disease. Neurology 44:681, 1994. 132. Steck, A. J., Renaud, S., Fuhr, P., et al.: An evaluation of a monoclonal anti-CD20 antibody (Rituximab) in the treatment of the anti-MAG associated polyneuropathy. J. Peripher. Nerv. Syst. 6:179, 2001. 133. Sumi, S. M., Farrell, D. F., and Knauss, T. A.: Lymphoma and leukemia manifested by steroid-responsive polyneuropathy. Arch. Neurol. 40:577, 1983. 134. Swarnkar, A., Fukui, M. B., Fink, D. J., and Rao, G. R.: MR imaging of brachial plexopathy in neurolymphomatosis. AJR Am. J. Roentgenol. 169:1189, 1997. 135. Symonds, R. P., Hogg, R. B., and Bone, I.: Paraneoplastic neurological syndromes associated with lymphomas. Leuk. Lymphoma 15:487, 1994. 136. Tayal, S. C., Rowbotham, D. S., and Bansal, S. K.: GuillainBarré syndrome in a patient with hairy cell leukaemia. J. R. Soc. Med. 84:238, 1991. 137. Tefferi, A., and Fonseca, R.: Selective serotonin reuptake inhibitors are effective in the treatment of polycythemia vera-associated pruritus. Blood 99:2627, 2002. 138. Tefferi, A., Solberg, L. A., and Silverstein, M. N.: A clinical update in polycythemia vera and essential thrombocythemia. Am. J. Med. 109:141, 2000. 139. Thomas, F. P., Vallejos, U., Foitl, D. R., et al.: B cell small lymphocytic lymphoma and chronic lymphocytic leukemia with peripheral neuropathy: two cases with neuropathological findings and lymphocyte marker analysis. Acta Neuropathol. 80:198, 1990. 140. Thomas, J. E., Cascino, T. L., and Earle, J. D.: Differential diagnosis between radiation and tumor plexopathy of the pelvis. Neurology 35:1, 1985. 141. Ulrich, J., and Müller, P.: Neurologische Komplikationen der Leukämien. Schweiz. Med. Wochenschr. 98:580, 1968. 142. Vallat, J. M., De Mascarel, H. A., Bordessoule, D., et al.: Non-Hodgkin malignant lymphomas and peripheral neuropathies—13 cases. Brain 118:1233, 1995. 143. Van den Bent, M. J., de Bruin, H. G., Beun, G. D. M., and Vecht, C. J.: Neurolymphomatosis of the median nerve. Neurology 45:1403, 1995. 144. Van den Bent, M. J., de Bruin, H. G., Bos, G. M. J., et al.: Negative sural nerve biopsy in neurolymphomatosis. J. Neurol. 246:1159, 1999. 145. Van Genderen, P. J. J., and Michiels, J. J.: Erythromelalgia: a pathognomonic microvascular thrombotic complication in essential thrombocythemia and polycythemia vera. Semin. Thromb. Hemost. 23:357, 1997. 146. Van Gerpen, J. A., Dyck, P. J. P., and Dyck, P. J.: Cutaneous T-cell lymphoma imitating gelsolin amyloidosis. Neurology 60:861, 2003.

Lymphoma, Leukemia, and Myeloproliferative Disorders 147. Vincent, D., Dubas, F., Hauw, J. J., et al.: Nerve and muscle microvasculitis in peripheral neuropathy: a remote effect of cancer? J. Neurol. Neurosurg. Psychiatry 49:1007, 1986. 148. Vital, C., Heraud, A., Vital, A., et al.: Acute mononeuropathy with angiotropic lymphoma. Acta Neuropathol. 78:105, 1989. 149. Vital, C., and Vital, A.: T-cell leukemic lymphomatosis. Neurology 39:1272, 1989. 150. Vital, C., Vital, A., Julien, J., et al.: Peripheral neuropathies and lymphoma without monoclonal gammopathy: a new classification. J. Neurol. 237:177, 1990. 151. Walk, D., Handelsman, A., Beckmann E., et al.: Mononeuropathy multiplex due to infiltration of lymphoma in hematologic remission. Muscle Nerve 21:823, 1998. 152. Walsh, J. C.: Neuropathy associated with lymphoma. J. Neurol. Neurosurg. Psychiatry 34:42, 1971. 153. Weintraub, M., Adde, M. A., Venzon, D. J., et al.: Severe atypical neuropathy associated with administration of hematopoietic colony-stimulating factors and vincristine. J. Clin. Oncol. 14:935, 1996. 154. Weiss, H. D., Walker, M. D., and Wiernik, P. H.: Neurotoxicity of commonly used antineoplastic agents. N. Engl. J. Med. 291:75, 1974.

2503

155. Wen, P. Y., Alyea, E. P., Simon, D., et al.: Guillain-Barré syndrome following allogeneic bone marrow transplantation. Neurology 49:1711, 1997. 156. Wick, M. R., Mills, S. E., Scheithauer, B. W., et al.: Reassessment of malignant “angioendotheliomatosis”: evidence in favor of its reclassification as “intravascular lymphomatosis.” Am. J. Surg. Pathol. 10:112, 1986. 157. Windebank, A.: Chemotherapeutic neuropathy. Curr. Opin. Neurol. 12:565, 1999. 158. Yiannikas, C., McLeod, J. G., and Walsh, J. C.: Peripheral neuropathy associated with polycythemia vera. Neurology 33:139, 1983. 159. Younger, D. S., Dalmau, J., Inghirami, G., et al.: Anti-Huassociated peripheral nerve and muscle microvasculitis. Neurology 44:181, 1994. 160. Younger, D. S., Rowland, L. P., Latov, N., et al.: Lymphoma, motor neuron diseases, and amyotrophic lateral sclerosis. Ann. Neurol. 29:78, 1991. 161. Zuber, M., Gherardi, R., Imbert, M., et al.: Peripheral neuropathy with distal nerve infiltration revealing a diffuse pleomorphic malignant lymphoma. J. Neurol. 235:61, 1987. 162. Pestronk, A., Florence, J., Miller, T., et al.: Treatment of IgM antibody associated polyneuropathies using rituximab. J. Neurol. Neurosurg. Psychiatry 74:485, 2003.

P. Neuropathy Associated with Industrial Agents, Metals, and Drugs

112 Human Toxic Neuropathy Caused by Industrial Agents ALAN R. BERGER AND HERBERT H. SCHAUMBURG

Acrylamide Use and Sources of Exposure Toxicity and Metabolism Clinical Features Morphology Electrophysiology Pathogenesis Allyl Chloride (3-Chlorophene) Carbamates Use and Source of Exposure Toxicity and Metabolism Clinical Features Carbon Disulfide Use and Sources of Exposure

Toxicology and Metabolism Clinical Manifestations Morphology Electrophysiology Pathogenesis Ethylene Oxide Hexacarbons: n-Hexane, Methyl n-Butyl Ketone Use and Sources of Exposure Toxicology and Metabolism Clinical Features Morphology Electrophysiology Pathogenesis

Peripheral neuropathy is one of the most common reactions of the nervous system to toxic chemicals. Enormous numbers of chemicals are currently deployed in the workplace, and several have been identified or implicated as causes of peripheral neuropathy, usually of the distal axonopathy type. It is probable that the reported cases of industrial neuropathy are only a fraction of those that occur. It appears likely that many industrial compounds widely regarded as “safe” may have significant neurotoxic potential, and tragic outbreaks of human disease may occur. Paradoxically, once a substance is showed to be neurotoxic, proper industrial hygiene can ensure its safe handling, and the incidence of human illness may actually decline.

Methyl Bromide Organophosphates Use and Sources of Exposure Toxicology Clinical Features of Cholinergic Reactions Clinical Features of OPIDP Pathology Physiology Pathogenesis Polychlorinated Biphenyls Pyriminil (Vacor) Trichlorethylene and Dichloracetylene

The following criteria establish the relationship of a suspected toxin to peripheral neuropathy: a characteristic clinical picture, definite and dose-related exposure, and neuropathy reproducible in experimental animals. Only some substances covered in this chapter—hexacarbons, carbon disulfide, ethylene oxide (EO), organophosphates, allyl chloride, pyriminil, and acrylamide—fulfill these criteria. Clinical reports, often isolated, frequently form the sole basis for identification of many of the alleged chemically induced peripheral neuropathies. Only those agents consistently associated with human neuropathy are included in this chapter. Industrial neuropathy following exposure to arsenic and lead is discussed in Chapter 113. 2505

2506

Diseases of the Peripheral Nervous System

ACRYLAMIDE Use and Sources of Exposure Acrylamide (CH2 "CHCONH2), a vinyl monomer, is a man-made compound synthesized by the hydration of acrylonitrile with sulfuric acid monohydrate. Commercial polymerization of the neurotoxic monomer into nontoxic polymers has been a principal use of acrylamide. The polymers have been used as flocculents in the treatment of wastewaters, as a grouting agent in waterproofing, and in electrophoresis chromatography. Most instances of toxicity have involved workers handling the monomer in preparation for polymerization either in the factory or in grouting; skin contact appears to have been a greater industrial hazard than inhalation. Instances of acrylamide neurotoxicity are now rare because the monomer is no longer a commercial product in North America.

Toxicity and Metabolism The monomer is an odorless, white, crystalline solid, very soluble in water. It is readily absorbed by all tissues, including skin, and is distributed throughout body water within minutes following intravenous administration; it disappears rapidly from the serum, with minute amounts persisting in nervous tissue for 14 days. Less than 1% is bound to nervous system tissues. The monomer undergoes oxidation at the double bond, which is required for neurotoxicity.63 Detoxification is via epoxy hydratase to 2,3-dihydroxypropenamide.160 Elimination is rapid via glutathione conjugation, although significant amounts persist in rats for 14 days after injection.154 Attempts to monitor acrylamide exposure in workers by urine measurements have been unsuccessful. One report suggests that hemoglobin adducts may be an accurate measure of industrial exposure.164 The nervous system is the principal toxic target organ for acrylamide in humans. Repeated exposures are necessary to produce human peripheral neuropathy.52 There is little evidence of general human dysfunction from even large exposures to acrylamide, aside from fatigue and weight loss. Acrylamide in large doses is carcinogenic and genotoxic in rats and mice.

Clinical Features Extremely high exposures, in suicide attempts, are associated with generalized seizures; moderately high exposure (e.g., 40 ppm acrylamide daily for several months in well water) causes an encephalopathy (memory loss, hallucinations) and cerebellar signs followed by peripheral neuropathy.73 Chronic low-level exposure can cause distal symmetrical peripheral neuropathy. Initial complaints are numbness and increased sweating in the hands and feet and an

unsteady gait.45,48,66,117,153 Positive sensory symptoms, such as paresthesias, are rare. Difficulty in walking and with balance (noticed in climbing ladders, etc.) is disproportionate to weakness or sensory loss. Objective signs are present in all symptomatic individuals and probably reflect peripheral nerve and, possibly, cerebellar dysfunction. Repetitive exposure to toxic doses of acrylamide eventuates in weakness of foot dorsiflexion and mild weakness of hand intrinsic muscles accompanied by unusually widespread loss of tendon reflexes. Loss of vibration sense in the feet and hands is inevitable. Muscle tenderness occasionally occurs, but the condition is not painful. Clumsiness and occasional intention tremor may be present in the upper limbs, and a broad-based, swaying gait is an early finding. Autonomic dysfunction, prominent in some experimental studies, is not a feature of human acrylamide neuropathy. Neither postmortem studies in acrylamide neuropathy nor nerve biopsies during the active phase of the illness have been reported. Sural nerve biopsies from two recovering patients displayed diminished numbers of large-diameter fibers. The amplitudes of sensory nerve action potentials usually are markedly reduced in the upper and lower limbs; this has been suggested as a sensitive electrophysiologic test for the detection of early cases.89 Most individuals with acrylamide neuropathy display only slight reduction of motor and sensory nerve conduction velocity. There is no known treatment for acrylamide neuropathy. Removal from exposure in the early stages of neuropathy results in gradual and, eventually, complete symptomatic recovery. Careful examination may reveal persistent depression of vibration sense as the sole abnormality. In more severely affected individuals, improvement continues for several months, but frequently there is residual distal weakness, gait ataxia, and impaired vibration sense. Persistent sensory dysfunction may stem from degeneration of axons in dorsal columns, and ataxia from changes in cerebellar efferent and afferent fibers. There is considerable interest in detecting subclinical acrylamide neuropathy in exposed workers. Estimation of sensory nerve action potential amplitude appears to be a sensitive electrophysiologic measurement. Quantitative assessment of distal extremity vibration sense is also a useful measure, an idea supported by studies of vibration sense and of degeneration in pacinian corpuscles in acrylamidedosed monkeys.105,140 Routine diagnostic laboratory tests, including cerebrospinal fluid (CSF) examination, are usually unremarkable. The diagnosis of acrylamide neuropathy is not difficult in an individual with proven industrial or occupational exposure. The presence of gait ataxia, moist peeling hands, and peripheral neuropathy leaves little room for doubt. A detailed occupational history is the most important diagnostic procedure because acrylamide neuropathy almost never stems from environmental exposure.

Human Toxic Neuropathy Caused by Industrial Agents

Morphology Repeated exposures are necessary to produce acrylamide distal axonopathy. Experimental animal studies in several mammalian species have generally utilized either shortterm, high-level daily administration (25 to 90 mg/kg for up to 14 days) or prolonged, low-level dosing (0.5 to 10 mg/kg for 6 weeks to 3 years). Each of these two regimens results in distinct spatiotemporal patterns of dysfunction and fiber degeneration, especially in the central nervous system (CNS). Low Doses The initial ultrastructural change in rats, cats, and monkeys is an accumulation of 10-nm neurofilaments, most marked at the distal ends of peripheral nerves.132,140 Paranodal accumulations of neurofilaments may cause axonal swelling and retraction of myelin at the paranodes. In general, neurofilaments initially accumulate at multiple paranodal sites in the distal ends of long and large-diameter axons. Early changes in feline pacinian corpuscles and annulospiral endings correlate with the profound loss of vibration sense and depressed tendon reflexes in the human neuropathy.140,154 Unmyelinated fibers in somatic nerves and in sympathetic nerve trunks degenerate in advanced stages of acrylamide intoxication.115 There is evidence of axonal regeneration during low-level intoxication of all species. Widespread, tract-oriented distal axonal degeneration in the CNS accompanies the peripheral nerve changes.162 Preterminal axonal degeneration in the gracile nucleus is an early event and has been observed in one study that utilized unusually low-level dosing for an extreme interval.137 Other vulnerable areas are the rostral regions of the dorsal spinocerebellar tracts and caudal regions of the long descending motor pathways in the lumbar spinal cord. Rats and monkeys dosed for longer durations display axonal swellings in shorter tracts, such as the optic nerve terminals in the lateral geniculate body. High Doses Administration of 30 mg/kg daily causes chromatolytic change in dorsal root ganglion cells and degeneration of Purkinje cells within 1 week. Two weeks later, widespread swellings of peripheral motor and sensory nerve endings appear, as well as multifocal terminal swelling in the spinal cord, brainstem, and cerebellum. Several studies have described impaired axonal regeneration in rats dosed with high levels of acrylamide. Nerve ligation together with acrylamide dosing causes retrograde axonal degeneration from the point of ligation, even extending into proximal sites in the nerve that are usually spared in distal axonopathies.19 A single dose of 75 mg/kg causes swelling in dorsal and ventral roots, suggesting impaired transport of all slow-component proteins.50 Repeated doses of 30 mg/kg

2507

cause diffuse proximal axonal atrophy that progresses in a distal direction at a time when there is little distal terminal change; this suggests that repeated high doses cause an axotomy-like reaction in the nerve cell body. Treatment for 5 days with levels of 30 to 50 mg/kg causes subcellular reorganization in sensory ganglia, autonomic ganglia, and Purkinje cells prior to peripheral axonal degeneration. It is suggested that this type of cellular reorganization represents a direct toxic effect of high-level administration of acrylamide.20,157,158

Electrophysiology There have been many electrophysiologic studies of experimental acrylamide-intoxicated animals. Early reports showed reduced peripheral nerve conduction in rats, cats, and primates.46,159 Single sensory nerve fiber conduction studies demonstrated the earliest change to be a failure of response of muscle stretch receptors. Studies of the responses of primary and secondary endings of muscle spindles revealed that the earliest abnormalities are elevated threshold and diminished response of spindle endings. It is suggested that the attenuated dynamic response of primary muscle spindles may be the basis for depressed tendon reflexes in the human neuropathy. It also seems likely that disordered peripheral sensory mechanisms, acting together with dysfunction in the cerebellum, may contribute to the severe gait ataxia that characterizes chronic neurotoxicity in humans. The spinal cord component of the monosynaptic stretch reflex is also clearly abnormal. One report of serial conduction measurements of central and peripheral projections of dorsal root ganglion cells describes persistent central slowing even after recovery of clinical abnormalities. Electrophysiologic studies of the autonomic system have revealed no changes in unmyelinated fiber conduction but slowed conduction in myelinated sympathetic nerve fibers.129,130 One study demonstrated impairment of neuronal control of the mesenteric vascular bed in severely affected animals, and it was concluded that both preganglionic myelinated and postganglionic unmyelinated fibers were damaged.130 Several studies support the view that physiologic changes antedate nerve fiber degeneration and behavioral dysfunction. One such study of pacinian corpuscles in the mesentery of cats treated with acrylamide demonstrated loss of detectable generator potentials prior to appearance of terminal axon changes as monitored by electron microscopic examination of chemically fixed or freeze-fractured, physiologically tested corpuscles.152 A study of somatosensory evoked potentials from the lower extremities in intoxicated primates revealed changes in the short-latency (gracile) components at a time when peripheral conduction and behavior were normal and clinical signs of neuropathy were absent.137 No morphologic changes were detected in the

2508

Diseases of the Peripheral Nervous System

peripheral nervous system (PNS) or CNS of these monkeys. The potentials returned to normal within 2 months of cessation of intoxication. Some monkeys, dosed at levels of 1.0 and 2.0 mg/kg/day (below the reported no-effect level), displayed onset of dysfunction only after extreme delay (940 days). This finding suggests that the current permissible levels of human exposure to toxins of this type should be reassessed.

Pathogenesis The primary biochemical lesion responsible for the development of acrylamide neuropathy is unknown, as are its relationships to the disorders of axonal transport and metabolism described in the many experimental studies of the past two decades. Initial studies on the mechanism by which acrylamide produces its neurotoxicity focused on the role of energy metabolism. However, subsequent studies reporting the failure of acrylamide to inhibit adenosine triphosphate production in vitro or to impair brain mitochondrial metabolism in vitro do not support the energy hypothesis.106,148 Acrylamide axonopathy is associated with impairment of both fast and slow axonal transport systems.50,75,109,110,149 The impairment in slow transport observed following a single injection most likely underlies the pathogenesis of the formation of the accumulations of neurofilamentous axonal swellings. Changes in axonal transport following repeated exposures, in contrast, reproduce those following axotomy and are likely to arise secondary to axonal injury and/or a defect in delivery to the nerve cell body’s response to a lost trophic support. Impaired fast axonal transport is the earliest known physiologic alteration following administration of acrylamide. The accumulation of tubovesicular profiles most likely is the morphologic correlate of early impairment of fast transport.26 The reduction in both retrograde and anterograde fast transport occurs within hours following a single intraperitoneal injection of acrylamide. There is considerable speculation that acrylamide impairs the function of microtubule-based bidirectional axonal movement, and that the microtubule-based fast anterograde transport is the specific neuronal target underlying the development of acrylamide-induced axonal degeneration.52,62,163 It is unclear how the axon undergoes degeneration in acrylamide neuropathy. Electron microprobe x-ray studies suggest that ion regulation in distal paranodal axon regions is impaired by diminished axolemmal sodium/potassium ATPase activity.93–95 It is alleged that this decreased ATPase activity is the result of aberrant cell body processing or deficient axonal transport. Subsequent membrane depolarization and accumulation of axoplasmic sodium favors the reverse influx of calcium and eventual axonal degeneration. It is uncertain whether these changes in distal axon elemental composition are a primary cause of or

reflect nonspecific secondary/compensatory responses to injury.

ALLYL CHLORIDE (3-Chlorophene) Allyl chloride (CH2CHCH2Cl) is a colorless, chlorinated hydrocarbon that is liquid at room temperature, volatile, and flammable. It has an irritating, unpleasant odor. Allyl chloride is primarily used in the manufacture of epichlorohydrin and glycerin pesticides, epoxy resins, monomers of polyacrylonitrile, and allyl compounds such as sodium allyl sulfonate. The threshold limit value in the United States is 1 ppm. Allyl chloride may be absorbed via the oral or inhalation route. Neurotoxic effects of allyl chloride intoxication include a distal axonopathy. Several outbreaks have been reported in Chinese workers who developed distal lower limb numbness, paresthesias, and weakness.65 Sensory loss is in a stocking-glove distribution, with early loss of the Achilles tendon reflexes.64 Electrophysiology showed denervation in distal leg muscles and delayed motor latencies. Recovery is usual after removal from exposure. Experimental animal studies demonstrated multifocal accumulations of axonal neurofilaments, with motor greater than sensory fiber degeneration in distal segments of the peripheral nerves and dorsolateral spinal cord tracts. Diagnosis of allyl chloride neuropathy is suggested by historical evidence of toxic exposure along with clinical and electrophysiologic evidence of a distal sensorimotor neuropathy.

CARBAMATES Use and Source of Exposure The underlying structure of carbamates consists of either carbamic acid, the monoamide of carbon dioxide, or dithiocarbamic acid, the monoamide of carbon disulfide. Carbamates are highly unstable compounds that decompose rapidly. The salts and esters of carbamic and dithiocarbamic acids are widely used as insecticides, herbicides, and fungicides. The compound disulfiram is used in treating alcoholism. Many of these compounds are neurotoxic, with exposure through occupational, accidental, or suicidal intoxication. Most carbamate insecticides are N-monomethylcarbamate esters, which vary in their degree of lipid solubility, species specificity, and potency as an anticholinesterase. More recently developed carbamate insecticides are derivatives of aliphatic oximes and resemble aldehydes or ketones in structure. Carbamate insecticides are efficiently absorbed from the gastrointestinal tract. The degree to which carbamates are distributed to various organs depends on their lipophilicity. Although there is little

Human Toxic Neuropathy Caused by Industrial Agents

body storage of carbamates, there are several case reports of episodic or prolonged effects, raising the question of organ storage with intermittent release into the blood stream. Excretion of the biotransformed or degradation products is predominantly through the urine and feces.

Toxicity and Metabolism The predominant effect of carbamates is to inhibit acetylcholinesterase (AChE), resulting in the accumulation of the unmetabolized neurotransmitter, acetylcholine (ACh). The result is excessive ACh accumulation at nerve endings of parasympathetic and sympathetic ganglia, CNS ganglia (nicotinic actions), postganglionic parasympathetic nerve endings (muscarinic actions), and the neuromuscular junction of somatic motor nerves. Inhibition involves both true AChE and pseudocholinesterase. Clinical symptoms do not last beyond 3 to 6 hours because there is a rapid, spontaneous reactivation of the inhibited enzyme, with destruction of accumulated ACh.

Clinical Features Acute high-level exposure to carbamates produces immediate neurologic dysfunction. Occasionally, there are unusual neurologic signs of a delayed and/or prolonged nature.37 Early reports of carbamate intoxication with Aldicarb-induced poisoning were reported in the United States and Canada in the 1970s and 1980s.37,53 Subjects ingesting Aldicarb-contaminated fruit and vegetables exhibited rapid-onset bradycardia, hypotension, vomiting, diarrhea, lacrimation, salivation, and muscular twitchings. An incident in Jamaica was reported in which methomyl was substituted for common salt in unleavened bread. Three of the five exposed individuals died. Two subjects displayed the muscarinic and nicotinic signs of carbamate toxicity.37 Neurotoxic symptoms appear rapidly, and tend to dissipate within a short time as a result of reactivation of AChE. The widespread availability of carbaryl has resulted in numerous reports of intoxication.31 The severity of symptoms is dose related, with early-onset nausea, vomiting, lacrimation, blurred vision, salivation, and headache. High-dose poisoning may be fatal; lower dose intoxication begins with early-appearing muscarinic and nicotinic signs, reaching peak severity within 2 to 3 hours of exposure. There have been reports of carbamate neurotoxicity that was either delayed in onset or had prolonged effects following a severe single exposure to carbamate insecticides.37 Early symptoms were as described above. Approximately 3 to 4 weeks later limb weakness and paresthesias, diffuse areflexia, limb incoordination, gait ataxia, and generalized fatigue appeared. Unusual persistent neurologic signs included intense headache, poor anger control, mood

2509

swings, impaired recent memory, and syncopal episodes.37 Neurotoxicity caused by chronic carbamate exposure is rare. One such case, in which there was exposure over a 6- to 8-month period, displayed typical early symptoms of carbamate toxicity. Persistent neurologic symptoms, after exposure was terminated, included a peripheral neuropathy with distal limb sensory loss and mild muscle weakness, disrupted sleep patterns, cognitive changes, tinnitus, and headache.12 Carbamates rarely cause a severe, delayed peripheral neuropathy, similar to that produced by organophosphate esters. This may be because of their inability to age the inhibited neuropathy target esterase compound (see Organophosphates below). There are fewer reports of human neurotoxicity with carbamates used as fungicides; the most common group of neurotoxic carbamate fungicides is the dithiocarbamates. Most carbamate fungicides are not readily absorbed through the skin or gastrointestinal tract. The usual route of intoxication is pulmonary, from inhaled powder or dusts. The most common metabolite is carbon disulfide, most of which is unexhaled. A degradation product of carbon disulfide, carbonyl sulfide may also contribute to neurotoxicity. Carbamate fungicides are very irritating to mucous membranes and skin, causing contact dermatitis and allergic hypersensitivity in exposed workers. These chemicals also cause thyroid dysfunction via the formation of ethylene urea with the release of atomic sulfur, which binds to thyroid molecules, diminishes iodine uptake, and iodinates tyrosine.72 Goiters can occur with repeated exposures. Of the two well-documented cases of carbamate fungicide toxicity, one involved workers exposed to a commercial mixture of maneb and zineb who initially exhibited behavioral changes, dizziness, fatigue and weakness. A second exposure led to the rapid appearance of muscle weakness, limb ataxia, disorientation, dysarthria, depressed mental status, and seizures.74 The worker eventually completely recovered by the fourth day. A second report documented parkinsonian-like signs and symptoms in two agricultural workers chronically exposed to maneb over 4 to 5 years.42 Dysfunction included tremor, rigidity, gait difficulties, shuffling steps, postural tremor, bradykinesia, hyperreflexia, and emotional lability. A subsequent study of 50 workers exposed to maneb revealed similar, but less severe findings. It is suggested that the parkinsonian-like symptoms are the result of carbon disulfide, a common breakdown product of most dithiocarbamates.151 Atropine can counteract the muscarinic symptoms.37 Atropine may also have a beneficial central effect, from a direct action on the respiratory system. Care should be taken regarding the amount of atropine given so as to avoid atropine toxicity, which may occur if there is reactivation of inhibited AChE. Treatment with diazepam can diminish muscle fasciculations, relieve anxiety, and counteract some CNS symptoms not affected by atropine. Pralidoxime, an axime reactivator, should be avoided in

2510

Diseases of the Peripheral Nervous System

most cases of carbamate poisoning, particularly with carbanyl, in which the agent enhances toxicity. Intoxication with Aldicarb may, however, be treated with pralidoxime; in this setting, it may enhance the beneficial effects of atropine.37

CARBON DISULFIDE Use and Source of Exposure Carbon disulfide is most commonly employed in the production of cellophane films145 and viscose rayon fibers.167 Other sources of carbon disulfide are dithiocarbamates used in cancer chemotherapy, as well as those employed in agriculture. Carbon disulfide is a major metabolite of the drug disulfiram (Antabuse), which is used as a deterrent for alcohol abuse.82 Most human intoxications result from inhalation exposure. Carbon disulfide intoxication has been a persistent problem in Scandinavia, Japan, and southern Europe.30

Toxicology and Metabolism Carbon disulfide is a colorless liquid that rapidly evaporates at room temperature. It is readily absorbed through the lungs and gastrointestinal tract; it passes less well through skin. The high fat solubility of carbon disulfide results in rapid disappearance from blood and greater concentrations in brain and liver. Toxicity is attributed to its reactivity with amine, sulfhydryl, and hydroxyl groups in biologic systems.57 Reactive sulfur atoms form and suppress the activity of cytochrome P-450 enzymes. Di- and trithiocarbamates can chelate copper and can inactivate metalloenzymes such as dopamine ␤-hydroxylase, vital to norepinephrine production. These compounds are also metabolized to isothiocyanates that can covalently bind and cross-link with cytoskeletal proteins. This may account for the giant axonal swellings in experimental studies. Carbon disulfide can be transformed through reaction with amino groups and can also be converted to carbon dioxide in a two-stage desulfuration reaction.36 Human studies have shown that only 1% of absorbed carbon disulfide is eliminated in the urine; approximately 90% is subject to biotransformation. Urinary levels of 2-trithioazolidine-4-carboxylic acid are a sensitive measure of carbon disulfide exposure.108

when exposure is low (10 to 40 ppm). As the concentration increases (20 to 60 ppm), a progressive sensorimotor distal asymmetrical polyneuropathy emerges. Clinical findings initially include leg weakness, loss of ankle and patellar reflexes, and diminished pain, touch, and vibration sensation in the distal lower limbs. Upper limb dysfunction appears following continued exposure. In some patients, recovery may be slow and incomplete, possibly because of residual CNS axonal damage.30,143 This notion is supported by experimental animal studies that demonstrate tractoriented distal axonal degeneration in the spinal cord as well as changes in distal peripheral nerve fibers.143,161 CNS manifestations with prolonged exposure include headache, dizziness, depression, memory impairment, and impaired sexual arousal.40 Exposed persons occasionally display spasticity as well as extrapyramidal signs with tremor, bradykinesia, and cogwheel rigidity. The diagnosis of carbon disulfide neurotoxicity is suggested by a combination of both CNS and PNS dysfunction in a setting of exposure. Abnormalities on electroencephalographic and psychological testing are common. There is no characteristic clinical profile, and routine laboratory tests, including CSF examination, are usually normal.

Morphology Human postmortem studies have revealed small, scattered foci of necrosis in the brain. Unfortunately, there are no adequate histologic studies of postmortem peripheral nerve changes and no reports of nerve biopsies in carbon disulfide toxicity. Rats exposed to 750 ppm of carbon disulfide develop hind limb weakness and ataxia after 6 weeks.144 Dogs intoxicated with carbon disulfide eventually lose tendon reflexes and become spastic in the lower extremities.4 Investigations in rats and rabbits, utilizing contemporary histologic techniques, have clearly demonstrated that carbon disulfide produces primary axonal degeneration, characterized by giant fusiform axonal swellings containing accumulations of 10-nm neurofilaments.81,92,143 The spatiotemporal evolution of these changes in the distal segments of both long peripheral nerve axons and long CNS fiber tracts indicates that prolonged exposure to carbon disulfide causes a centralperipheral distal axonopathy. The distribution and nature of the axonal changes caused by carbon disulfide resemble those produced by the hexacarbons.55 Retinogeniculate degeneration may also occur.38

Clinical Manifestations Acute or subacute high-level exposures cause signs of CNS dysfunction: confusion, hallucinations, memory impairment, and emotional lability.54,91 Chronic, low-level exposure causes a combination of peripheral neuropathy and CNS abnormalities.1 The neuropathy may be asymptomatic and detected only by electrophysiologic testing

Electrophysiology Nerve conduction studies are useful in detecting subclinical neuropathy in workers exposed to low levels of carbon disulfide. Seppalainen and Tolonen documented slowed motor conduction in the legs of exposed subjects.145 Knave and associates found that age-weighted motor conduction

Human Toxic Neuropathy Caused by Industrial Agents

velocities were successively decreased among heavily exposed workers with 5 to 30 years of exposure.86 Mild slowing of digital sensory nerve conduction may be the only abnormality in the early stages of neuropathy, and fibrillation potentials are present in distal leg muscles.165 Nonspecific electroencephalographic abnormalities are allegedly more common in carbon disulfide–exposed workers than in the general population, as are abnormalities in standard psychological test batteries.143 Nerve conduction velocities in rats and rabbits exposed to carbon disulfide are slowed; motor conduction slowing may precede weakness, and fibrillation potentials can be detected in paretic muscles.144 Abnormalities of visual, brainstem, auditory, and somatosensory evoked potentials are detectible in intoxicated rats.165

Pathogenesis The pathogenesis of carbon disulfide neurotoxicity may reflect the reaction with amines to form dithiocarbamates. Dithiocarbamates chelate copper and zinc, decreasing the availability of these two cofactors and resulting in enzyme inhibition. Alternatively, vitamin B6 deficiency could result from dithiocarbamate formation on pyridoxamine.133 The similarity of the axonal changes of carbon disulfide and hexacarbon neuropathy suggests a common pathogenesis, specifically the ability of both to cross-link neurofilaments. Neurofilament cross-linking, with resultant neurofilament accumulations, may underlie the axonal swellings seen in carbon disulfide and hexacarbon neuropathy. In experimental animals, the initial focal axonal swellings develop just proximal to nodes of Ranvier. Myelin retraction from nodes of Ranvier is followed by segmental demyelination, and eventually distal axonal degeneration. The longest and largest axons appear the most vulnerable in both peripheral nerve and spinal cord. The basic biochemical mechanisms underlying the neurotoxicity of carbon disulfide remain unclear.9,32 There are three main hypotheses: (1) a chelating effect of carbon disulfide metabolites on metals essential for enzyme function, especially copper and zinc,29,141; (2) direct inhibition of enzymes such as glyceraldehyde-3-phosphate dehydrogenase, monoamine oxidase, lactic acid dehydrogenase,104 and dopamine dehydroxylase9; and (3) release of free radicals following cleavage of the carbon-sulfur bond.9

2511

for medical supplies and fumigant for furs and some foods. Experimental and epidemiologic studies suggest EO to be a potential carcinogen and mutagen.134,147 EO clearly causes a distal symmetrical polyneuropathy in humans and in experimental animals following chronic or intermittent peak exposure at levels in excess of 250 ppm.58,122,124 Prolonged extreme low-level exposure may also be hazardous to the nervous system; one study suggests that residual EO in dialysis tubing may contribute to peripheral neuropathy in patients on long-term hemodialysis.172 Another report implicates potential neurotoxicity of EO in the solvent vehicle for cyclosporine.173 It is also alleged that sterilizer workers may develop overt or subclinical, EO-induced CNS and PNS dysfunction. Humans with symptomatic polyneuropathy initially experience symptoms of distal extremity numbness and weakness, accompanied by evidence of diminished sensation in the feet and hands.60 Tendon reflexes are diminished throughout and ankle jerks are absent. Motor and sensory nerve conduction velocities are mildly diminished. Encephalopathic symptoms and cognitive impairment may accompany peripheral neuropathy. Sural nerve biopsies exhibit axonal degeneration. There are no reports of postmortem findings in humans with EO intoxication. Gradual recovery from peripheral neuropathy commences following withdrawal from exposure. One report of three individuals describes normalization of motor and sensory nerve conductions and improved neurologic function at a 4-year follow-up examination.58 Experimental animals chronically exposed to 250 ppm of EO develop widespread sensory nerve fiber degeneration in the distribution of a central-peripheral distal axonopathy,71,124 as do rats chronically dosed with 1500 ppm of propylene oxide, a chemically similar substance.123 Intact animal studies demonstrate that EO exposure causes impaired fast axoplasmic transport.118 An in vitro report describes degeneration in dorsal root ganglion cell neurite extensions following EO exposure.172 The pathogenesis and biochemical basis of EO neurotoxicity are unknown.

HEXACARBONS: n-Hexane, Methyl n-Butyl Ketone Use and Sources of Exposure

ETHYLENE OXIDE EO is a gas at room temperature; it is water soluble and widely distributed to all organs shortly following inhalation exposure. EO is a powerful alkylating agent and reacts with virtually all cellular components, including DNA, and with histidine in hemoglobin.122 Its principal industrial use is in the synthesis of ethylene glycol. EO is an effective sterilant

Hexacarbon neuropathy was first discovered in humans through industrial exposure and subsequently encountered in individuals practicing inhalant abuse. n-Hexane (CH3CH2CH2CH2CH2CH3) and methyl n-butyl ketone (CH3COCH2CH2CH3; MnBK) are considered together in this chapter because each is metabolized to 2,5-hexanedione (2,5-HD), a ␥-diketone that appears to be responsible for much of the neurotoxicity of these substances.

2512

Diseases of the Peripheral Nervous System

Hexacarbon toxicology and clinical neurology have been extensively reviewed.35 n-Hexane is still widely used as an inexpensive solvent and is a minor component of gasoline and its combustion products. Occupational neuropathy has usually been associated with industries utilizing n-hexane in poorly ventilated quarters (shoemaking, furniture finishing) with airborne levels in excess of 60 to 240 ppm. MnBK is now rarely used in its pure form, but it still may be present in some ketone mixtures and also occurs as a by-product of some wood pulping operations. Methyl ethyl ketone (MEK) sometimes is present in solvent mixtures containing either n-hexane or MnBK. MEK does not cause neuropathy139; it is able to speed the development of hexacarbon neuropathy in animals and potentiate n-hexane neurotoxicity in humans.5 Methyl isobutyl ketone is associated with memory loss but not with peripheral neuropathy.59

Toxicology and Metabolism n-Hexane and MnBK are both colorless liquids and almost insoluble in water. Their common neurotoxic metabolite, 2,5-HD, is highly water soluble. For this reason, 2,5-HD is widely used as the most convenient compound for experimental studies of hexacarbon neuropathy. MnBK is readily absorbed through the skin and respiratory tract. Cutaneous absorption of n-hexane may be a major route of entry in humans, but appropriate experimental studies have not been performed to confirm this. n-Hexane is less well absorbed via the respiratory tract than MnBK. The metabolism of n-hexane and MnBK has been determined in experimental animals and in humans. n-Hexane can undergo oxidation to form 2-hexanol, which may be metabolized either to 2,5-hexanediol or to MnBK, both of which may be converted to 5-hydroxy-2-hexanone and then to 2,5-HD. MnBK likewise may yield 2,5-HD in vivo. Study of the metabolism of MnBK and n-hexane has identified 2,5-HD as the most persistent neurotoxic metabolite of both substances.35

Clinical Features Both males and females are affected, and the age of onset ranges from adolescence to late middle age. Individuals in different countries have been exposed to a wide variety of solvent mixtures and, in many instances, the contents and methods of chemical analysis are poorly described. The quality of the documentation of neurologic, clinical, electrophysiologic, and laboratory data varies considerably, and long-term follow-up examinations are scarce. Meticulous clinical studies of glue sniffing neuropathy have been done both in Europe and in North America.5,87 The Ohio MnBK study is widely held as a paradigm of the correct evaluative approach to an outbreak of an industrial toxic neuropathy.3

The most common initial complaint, both in industrial cases and among glue sniffers, is an insidious onset of numbness of the toes and fingers. This type of distal sensory neuropathy is generally the only clinical feature in the least severe industrial cases. The pattern of sensory abnormality is characteristically symmetrical and involves only the hands and feet, rarely extending as high as the knees. Moderate loss of touch, pain, vibration, and thermal sensation is usually evident and may be accompanied by loss of the Achilles tendon reflexes; the other tendon reflexes are spared. In mild cases, there is preservation of position sense and no sensory ataxia, periosteal pain, cranial nerve abnormalities, or autonomic dysfunction. In more severe industrial cases, weakness and weight loss occur, occasionally accompanied by anorexia, abdominal pain, and cramps in the lower extremities. Reflex loss is usually less than that observed in other polyneuropathies and, even in the moderate to severe cases, may be confined to the Achilles tendon reflexes and finger jerks. Weakness most commonly involves the intrinsic muscles of the hands and long extensors or flexors of the digits. A common complaint in these individuals is difficulty with pinching movements, grasping objects, and stepping over curbs. Instances of pure motor neuropathy are unusual in industrial cases. Vibration and position sense is only mildly impaired, and pinprick and tactile sensory loss is usually confined to the hands and feet. As the neuropathy becomes more severe, weakness and atrophy dominate the clinical picture and extend to involve proximal limb muscles. Glue-sniffing patients may display a subacute, distal-toproximal progression of weakness early in the course of the disease.5 In a few glue sniffers, blurred vision has been a symptom, but objective evidence of visual loss has not been documented. Seizures, toxic delirium, cerebellar ataxia, tremor, or cholinergic symptoms are not described. No predisposing conditions exist for hexacarbon neurotoxicity, although one report describes a high incidence of polyneuropathy in older workers and slowed motor nerve conduction in “normal” individuals in a factory with documented cases of solvent neuropathy.13 This strengthens the notion that subclinical and readily reversible hexacarbon nerve damage may be an unrecognized industrial problem. Autonomic disturbances have been reported among the glue sniffers but not in the industrial cases. Prominent among these disturbances is hyperhidrosis of the hands and feet, occasionally followed by anhidrosis. Blue discoloration of the hands and feet, reduced extremity temperature, and Mees’ lines are sometimes present. Impotence occasionally occurs among glue sniffers with moderate or severe neuropathy, but its relationship to nervous system dysfunction is unestablished.5 Slow progression is the hallmark of industrial cases. In most instances, this reflects low-level, intermittent

Human Toxic Neuropathy Caused by Industrial Agents

exposure. In some glue sniffers, especially those with excessive abuse, a subacute course develops, leading in severe cases to quadriplegia within 2 months of the first symptoms. Acute inflammatory demyelinating polyradiculopathy has been a serious diagnostic consideration in some of these patients.5 A universal feature of hexacarbon neurotoxicity is the continuous progression of disability (“coasting”) after removal from exposure. Coasting usually lasts for 1 to 4 months. The degree of recovery in most cases correlates directly with the intensity of the neurologic deficit. Individuals with a mild or moderate sensorimotor neuropathy usually recover completely within 10 months of cessation of exposure. Severely affected patients with industrial exposure also improve; some retain mild to moderate residual neuropathy on follow-up examination as long as 3 years after exposure. Such individuals, on occasion, display hyperactive knee jerks. This reflex change may reflect the degeneration in the corticospinal tracts accompanying the peripheral axonal degeneration.139 Differential diagnosis of n-hexane neuropathy is based upon clinical signs that indicate distal axonopathy, an unusual degree of slowing of peripheral nerve conduction, and, most importantly, a history of solvent exposure. Without a clear exposure history, peripheral neuropathy from other metabolic or toxic causes can be difficult to exclude. Routine clinical laboratory tests, including CSF examination, usually yield normal results. Assays for determination of blood and urinary levels of 2,5-HD are now commercially available and can confirm current exposure.56 This test is of little use in persons with n-hexane neuropathy in whom exposure terminated months previously. A biologic exposure index for occupational n-hexane exposure has been determined based on urinary 2,5-HD levels.

Morphology Humans Nerve biopsies in mild cases may be normal,69 but in individuals with overt neuropathy, nerve fibers show focally swollen axons filled with 10-nm neurofilaments and displaying myelin retraction.3,87 Giant axonal changes also are reported in the fasciculus gracilis in postmortem material from a glue sniffer. These changes in the PNS and CNS are indistinguishable from those in experimental hexacarbon neuropathy (see later discussion). Experimental Animals The neurotoxic hexacarbons have produced centralperipheral distal axonopathy in several species of experimental animal. In the PNS of rats, changes first appear in the large axons of the tibial nerve branches to the calf muscles. Subsequently, distal nerve fiber changes develop

2513

in the plantar nerves.155 With time, more proximal portions of the sciatic nerve become affected. The cell bodies of sensory and motor neurons remain intact. This pattern of axonal pathology has been well documented in several studies of experimental hexacarbon neuropathy following prolonged (weeks to months) low-level intoxication. The fundamental alteration is a focal condensation of neurotubules, mitochondria, and smooth endoplasmic reticulum, coupled with a massive local increase in the number of 10-nm neurofilaments. These axoplasmic changes initially develop on the proximal sides of nodes of Ranvier in distal regions of vulnerable nerve fibers; subsequently the swellings are associated with displacement of paranodal myelin.154 The spatiotemporal sequence of hexacarbon distal axonopathy has been documented in in vitro living peripheral nerve fibers that are one component of structurally and functionally coupled cord-ganglia-muscle explants. These explants have been chronically exposed to each of the six interrelated hexacarbon metabolites.166 Prolonged in vitro exposure (months) to any of these agents yields a sequence of distal change in selected large-diameter motor nerve fibers; giant axonal swellings containing excessive 10-nm neurofilaments first appear at the extreme distal tip of the broad, unbranched portion of the axon, and on the proximal side of a slightly widened tri- or multipartite node of Ranvier, which marks the origin of the terminal intramuscular branches. Subsequently, giant axonal swellings appear both proximally and distally. Prominent axonal swellings sometimes affect entire internodes and often precipitate paranodal myelin retraction. The denuded axon recovers a more normal diameter and undergoes demyelination if the toxin is removed at this stage. Continued intoxication precipitates breakdown of a distal length of nerve fiber. Cytons of motor neurons innervating affected fibers cannot be monitored in the living culture, but equivalent changes of sensory axons are not accompanied by any specific changes in their corresponding cytons. The long ascending and descending tracts of the spinal cord are particularly vulnerable in hexacarbon neuropathy.155 In the rat, nerve fiber swellings first appear in the rostral part of the gracile tract and the caudal parts of ventral and ventrolateral tracts. With time, degenerative changes spread retrogradely along the affected fiber pathways. Giant axonal changes also occur in the anterior cerebellar vermis, lateral geniculate, and mammillary bodies, corresponding to distal breakdown of spinocerebellar, optic, and fornical tracts, respectively.138 There appears to be a species-specific selective vulnerability to degeneration in the large retinal ganglion cells of the rat; there are no similar findings in the primate visual system. One study claims that the primate optic tract is less vulnerable to 2,5-HD than to carbon disulfide or acrylamide.103

2514

Diseases of the Peripheral Nervous System

Electrophysiology Humans In general, the degree of electrophysiologic abnormality, both electromyographic and in nerve conduction, parallels the severity of the clinical illness. The electromyographic abnormalities usually are symmetrical and greater in distal than in proximal muscles. In patients with minimal involvement, minor denervation changes are often the only findings. Nerve conduction times and clinical examination are usually normal at this stage. In the Ohio MnBK epidemic, 37 cases were detected without clinical evidence of nervous system disease, but with the abnormal electrodiagnostic findings mentioned earlier; these cases are considered to represent subclinical MnBK neuropathy.3 The more severe cases generally display more obvious electromyographic evidence of fiber degeneration. With recovery or stabilization of the condition, the electromyographic changes disappear. Residual findings include reduction in motor unit numbers and an increase in the amplitude of motor unit action potentials. In cases with minimal involvement, motor and sensory nerve conduction is usually normal or in the low-normal range for velocity. As the clinical illness intensifies, progressive slowing of conduction ensues and, in the most extreme cases, distal peroneal nerve conduction cannot be elicited. It was noted in the Ohio outbreak that five cases displayed prolonged distal latencies in the presence of normal peroneal nerve conduction velocity.3 There may be slowing of motor nerve conduction in “normal workers” and a direct correlation between the duration of solvent exposure and the slowing of motor nerve conduction velocity. One study of glue sniffers described severe slowing of motor nerve conduction, seemingly out of proportion to the degree of weakness. This report, which also incorporates a detailed study of a nerve biopsy specimen, concluded that, although such profound slowing is usually associated with primarily demyelinating neuropathies, in n-hexane neuropathy it probably reflects the paranodal myelin changes caused by axonal swelling.87 There have been few reports on nerve conduction in individuals recovering from hexacarbon neuropathy. In the Ohio MnBK cases showing clinical recovery, motor nerve conduction velocity tended to return to normal.3 A comprehensive study that analyzed visual, somatosensory, and brainstem auditory evoked potentials describes abnormalities in all modalities.

physiologic consequences of terminal giant axonal changes found in the optic tract of animals with hexacarbon neuropathy.138 Several studies report early (preclinical) and progressive reduction in conduction velocities of sciatic and ulnar nerves in animals receiving neurotoxic hexacarbons.77,127 These changes are ascribed to the focal demyelination accompanying giant axonal swelling, although one report emphasized that nerve conduction changes precede demyelination.77

Pathogenesis The neurotoxic mechanism of action of n-hexane, and more specifically the ␥-diketones, has been extensively studied in the 30 years since the initial human outbreaks of peripheral neuropathy. A recent review stated that “although a number of hypotheses have been proposed and many important aspects of the mechanism have been clarified, the molecular determinants and pathological significance of axonal neurofilament accumulation and the ultimate mechanism of axonal degeneration remain unclear.”35 In general, most studies indicate that 2,5-HD exposure impairs axonal transport. It is widely held that the neurofilament itself is the target of ␥-diketones, and alterations in filament structure and metabolism are the ultimate causes of axonal dysfunction and degeneration. This notion is supported by experiments that detect neurofilament accumulations at sites along the axon following the direct application of 2,5-HD at those loci.128 It has been suggested that the initial step is formation of a 2,5-dimethylpyrrole adduct from covalent binding of 2,5-HD to amine groups in neurofilaments. This is the fundamental reaction in the “pyrrole hypothesis.”34,49,135,174 Once formed, pyrrole adducts can undergo additional oxidative reactions to yield secondary, cross-linked derivatives. It is suggested that the cross-linking of neurofilaments results in disruption of neurofilament metabolism and the accumulation of neurofilaments at the internodes. Other studies have suggested targets other than the neurofilament protein because there is some evidence that distal axonal degeneration can occur in the absence of neurofilament accumulation. Covalent binding of 2,5-HD to microtubule-associated protein occurs in vivo, and 2,5-HD alterations of elemental composition and water content may implicate a primary attack on axonal homeostasis.96

METHYL BROMIDE Experimental Animals Various physiologic studies have been performed in experimental animals, including visual evoked responses and peripheral electrodiagnostic studies. One study recorded visual evoked potentials from monkeys exposed to MnBK and described increased latencies after 4 months.22 The authors suggested that these findings represent the electro-

Methyl bromide (CH3Br) is a colorless, nonflammable gas that is odorless at low concentrations but has a chloroformlike odor at high concentrations. It is used as an insecticidal fumigant, refrigerant, and fire extinguisher; as a solvent for oil extraction from nuts, flowers, and seeds; and as an industrial methylating agent. Intoxication is usually through

2515

Human Toxic Neuropathy Caused by Industrial Agents

the lungs, but absorption is also possible via the gastrointestinal tract and skin. Methyl bromide is rapidly and almost completely metabolized in the liver. Maximum allowable inhalation is reported to be 5 ppm. Acute intoxication initially results in mucosal irritation followed by malaise and gastrointestinal distress. Within hours there ensue signs and symptoms of severe CNS dysfunction, including headache, dizziness, dysarthria, visual impairment, delirium or psychosis, seizures, and myoclonus. Recovery is common after low-dose exposure, but high-dose intoxication may cause coma with eventual death. Chronic high-level exposure may result in a syndrome characterized by dysfunction of pyramidal tracts, cerebellum, and peripheral nerves.2,70 Behavioral abnormalities may coexist. There are few well-detailed reports of methyl bromide peripheral neuropathy. Most describe distal, symmetrical sensorimotor neuropathy developing over 3 to 7 months of exposure.83 Neuropathy is heralded by acral paresthesias and pain, followed by distal leg weakness, hand clumsiness, and gait ataxia. Findings include a stocking distribution of pain and touch sensory loss, distal leg weakness, and tender calf muscles. The Achilles tendon reflexes are lost. Mild optic neuropathy may be present, with early loss of color vision.16,23 The overall clinical pattern most closely suggests a distal axonopathy, but there are few confirmatory pathologic specimens. The CSF is normal. In one case, electrophysiologic studies showed a distal, predominantly motor axonopathy. Sensory conduction studies may potentially be normal.16 A sural nerve biopsy showed loss of predominantly large myelinated fibers.16 Postmortem findings in a fatal case following high-dose acute exposure demonstrated neuronal loss in dorsal root ganglia and axonal degeneration in nerve roots and proximal nerve segments.156 Gradual improvement, with complete recovery in milder cases, occurs within a year after withdrawal from exposure. The mechanism of methyl bromide neurotoxicity remains uncertain. Toxicity appears to be due to methyl bromide itself, rather than its metabolite, methanol, or the bromide ion.2,114,119 The finding of segregated microtubules in the nodes of Ranvier in one study suggests that methyl bromide may impair axonal flow, resulting in a distal axonopathy. There may also be a potential pathogenetic role for the methylglutathione metabolites methanethiol and formaldehyde. This is suggested by two individuals with comparable acute methyl bromide exposure, one of whom, with normal erythrocyte glutathione transferase activity, developed severe neurotoxicity, while the other, who lacked detectable enzyme activity, displayed minimal symptoms.47

ORGANOPHOSPHATES Over 20,000 compounds are classified as organophosphates (OPs). Their physicochemical properties are varied and they exist in solid, liquid, and gaseous forms. Some organophos-

phorus esters cause a distal axonopathy characterized by widespread CNS and PNS degeneration (organophosphateinduced delayed peripheral neuropathy [OPIDP]). This axonal effect does not involve inhibition of AChE, has a delayed onset following a single exposure, and has been responsible for devastating epidemics of neurotoxic injury. Neuropathy has been most widely associated with the OP tri-o-cresyl phosphate (TOCP), although other OPs, including leptophos, mipafox, chlorphos, and trichlorfon, have also been reported to cause OPIDP. Diisopropylfluorophosphate (DFP), haloxon, and butafox may produce similar neurotoxic changes in experimental animals. Persistent health complaints in Gulf War veterans have raised the question of whether low-level, prolonged exposure to a variety of OPs could cause peripheral nerve damage.98 Distal axonopathy is thought to be a relatively uncommon consequence of OP exposure, especially when compared with the well-known AChE effects. Experimental studies, however, indicate that OP-induced neuropathy may be more common than formerly thought, and that low-level exposure to neurotoxic OPs may produce neuropathy if combined with other less neurotoxic OPs.97

Uses and Sources of Exposure Organophosphates are most commonly used as petroleum additives, insecticides, lubricants, antioxidants, flame retardants, and plastic modifiers. Intoxication may occur as a result of accidental pesticide exposure from agricultural spraying. Exposure may occur in those individuals mixing or applying the pesticide or through dermal exposure from those working in the fields shortly after spraying.107 Most OPs are quickly degraded in the environment. The majority of epidemic TOCP intoxications have resulted from inadvertent adulteration of food, drink, or cooking oil. Prominent outbreaks occurred in the United States from drinking contaminated Jamaica ginger extract (“jake leg paralysis”), and in Morocco from eating food cooked in contaminated oil.84,150 Pesticides containing OPs are also intentionally ingested in suicide attempts.

Toxicology Absorption of OPs may take place through the respiratory and gastrointestinal tracts, and the skin. The rate of absorption and excretion differs among the various OPs. Once absorbed, OPs irreversibly inhibit, through phosphorylation, true AChE, located in erythrocytes and nervous tissue, and plasma pseudocholinesterase. Only depression of erythrocyte AChE is specific for OP exposure. Reduced pseudocholinesterase level is a nonspecific finding and may be seen in many naturally occurring diseases and toxic conditions. Acetylcholinesterase inhibition causes excessive accumulation of ACh with resultant overstimulation of muscarinic and nicotinic receptors. There are acute (type I)

2516

Diseases of the Peripheral Nervous System

and intermediate (type II) syndromes (see below) that reflect which receptors are preferentially affected by the excessive ACh. Cholinergic symptoms, caused by muscarinic receptor overstimulation, may be apparent within hours of OP exposure and are always evident within 1 day. Different OPs produce cholinergic signs and symptoms of varying intensity. Mild cholinergic symptoms include gastrointestinal distress such as diarrhea and vomiting, combined with lacrimation, salivation, sweating, and muscle fasciculations. Severe cases show behavioral changes, widespread weakness, and respiratory depression. Atropine may reverse the aforementioned cholinergic symptoms resulting from excessive muscarinic receptor stimulation, but is ineffective in reversing muscle weakness caused by nicotinic receptor blockade. Atropine’s effect may be transient and cholinergic symptoms may reoccur, occasionally requiring respiratory assistance. Severe exposure may cause death.44 Exposed subjects may be prone to ventricular arrhythmias when atropine is given in the presence of respiratory insufficiency, and atropine should not be administered until adequate ventilation has been established.44 Pralidoxime may be of clinical benefit by accelerating the hydrolysis of inhibited AChE.

Clinical Features of Cholinergic Reactions The nature and severity of neurologic symptoms following OP exposure is partly dependent upon the type of OP, the degree of exposure, and the extent of absorption. A transient, clinically heterogeneous cholinergic response usually occurs following a single OP exposure. The acute type I syndrome169 includes cholinergic symptoms caused by excessive muscarinic receptor stimulation, which are evident shortly after exposure. Symptoms include gastrointestinal distress, miosis, lacrimation, salivation, diarrhea, and bradycardia. Weakness is not a component of the type I syndrome. The type II or intermediate syndrome, so named because of its temporal appearance between the early type I syndrome and the later OPIDP, occurs 12 to 96 hours after exposure and is due to excessive nicotinic receptor stimulation. Despite lesser clinical renown, the type II syndrome is thought to occur more frequently than the more heralded OPIDP. The type II syndrome may result from exposure to many different OPs, including dimethoate, fenthion, methamidophos, malathion, and fenitrothion, and includes fasciculations, limb and respiratory muscle weakness, tachycardia, and hypertension.33 Patients may be symptom free between 1 and 4 days following OP exposure before type II symptoms occur. Respiratory distress may be the initial symptom, followed by weakness of proximal limb and neck flexor muscles. Although distal extremity strength is usually preserved, cranial nerve function may be impaired, including extraocular muscles. Sensory function is normal, but dystonic limb posturing may be present. Respiratory and cardiac failure may occur

in severe cases. Occasionally CNS signs appear, including anxiety, confusion, blurred vision, impaired memory, tremor, convulsions, respiratory depression, and coma. Recovery usually begins about 5 to 15 days after exposure and proceeds in the reverse order to symptom appearance, with cranial nerve muscles recovering first and neck flexors last. The rate of recovery depends on the type of OP, the extent of exposure, and the manner of treatment. Occasional patients are slow to eliminate OPs, resulting in impaired AChE recovery and persistent clinical deficits for several weeks.51 Persistent behavioral changes may be evident after acute poisoning. Atropine will modify symptoms caused by excessive muscarinic receptor stimulation (type I), but not those resulting from nicotinic receptor overstimulation (type II). There is no correlation between the occurrence of the intermediate syndrome and the later appearance of OPIDP. An early, objective sign of OP intoxication is electrophysiologic abnormality of neuromuscular transmission, its nature correlated with exposure severity.168 Single motor stimuli, applied shortly after OP exposure, may induce spontaneous repetitive motor action potentials (SRMAPs) following the primary motor action potential.10 This electrophysiologic finding is a sensitive marker of OP exposure, but does not correlate with the severity of intoxication. The onset and progression of muscle weakness are associated with a decremental pattern with repetitive motor nerve stimulation. The maximum decrement of motor potential amplitude occurs by the second potential, unlike myasthenia gravis, in which the fourth potential is usually of lowest amplitude. The characteristics of repetitive stimulation abnormalities reflect the intensity of intoxication. Repetitive stimulation studies in the setting of mild weakness or improving strength are characterized by (1) a decrement in the amplitude of the second potential, evident only with rapid repetitive stimulation; (2) a progressive restoration of motor potential amplitude in the following motor action potentials (decrement-increment response); and (3) the presence of SRMAPs following the initial few motor potentials.10 In more severe cases with greater weakness, the decremental response is evident with slower stimulation rates and SRMAPs are not observed.

Clinical Features of OPIDP OPIDP is much less common than the more frequently occurring cholinergic symptoms after OP exposure, but results in considerable morbidity. The underlying pathology of OPIDP is central-peripheral axonal degeneration, with clinical symptoms appearing after a latent period of 7 to 21 days. Organophosphates capable of producing OPIDP almost invariably cause preceding cholinergic symptoms, although these may be subtle and unappreciated.98 Peripheral neuropathy is heralded by early muscle cramping and calf pain, along with tingling and burning

2517

Human Toxic Neuropathy Caused by Industrial Agents

sensations in the feet and occasionally the hands. Weakness is an early and invariable finding, and established cases may have predominant motor deficits with minimal sensory complaints. Distal muscles are the earliest and most severely affected, although proximal muscles may occasionally be involved in severe cases. The Achilles reflexes are depressed or absent, while more proximal reflexes may be depressed, or even increased if CNS dysfunction is present. Cranial nerve and autonomic function are preserved. Neuropathy progression is subacute, being fully expressed within a few days. Chronic progression is not seen, unlike in other immune-related, inflammatory, or metabolicrelated distal axonopathies. There is one report of an acute demyelinating motor neuropathy that developed after exposure to merphos.43 Whether this neuropathy was actually acute inflammatory demyelinating neuropathy or an unusual presentation of OP intoxication is unclear. Symptoms and signs of CNS dysfunction may be present in severe cases and represent damage to distal ends of motor and sensory tracts within the spinal cord. CNS dysfunction is usually inapparent early in the neuropathy as signs of peripheral nerve damage predominate. As time passes and peripheral nerves recover, signs of CNS dysfunction may emerge, including hyperreflexia, increased motor tone, and spastic gait. In its most severe expression, OPIDP manifests upper and lower motor neuron involvement. A common physical finding in a 30-year follow-up study of individuals with TOCP poisoning was the combination of spastic paraparesis and distal leg atrophy. Evidence of corticospinal tract dysfunction is a feature that distinguishes OP neurotoxicity from most other distal axonopathies and may be misleading; for example, in a recent outbreak of leptophos intoxication, several cases were erroneously diagnosed as encephalomyelitis or multiple sclerosis. The prognosis in mildly affected individuals is usually good, with most making nearly a complete recovery. Others with a more severe initial deficit are left with varying degrees of morbidity, which include sequelae of both PNS (atrophy, claw hands, footdrop) and CNS (spasticity, ataxia) damage. In many cases the ultimate prognosis depends more on the degree of CNS than of peripheral nerve dysfunction. There is no specific treatment once the acute cholinergic crisis has resolved.8 Methylprednisolone, given to cats following exposure to the OPDFP, prevented the development of OPIDP.76 Its applicability in humans is unproven. There has been recent concern regarding the occurrence of chronic neurologic and neurobehavioral dysfunction related to acute or prolonged, low-level OP exposure. Media attention was focused on Gulf War veterans who were exposed to a variety of OPs during their military duty. The syndrome, labeled chronic OP-induced neuropsychiatric disorder, has been linked to acute, low-level exposure (phenomenon 1), as well as long-term exposures

(phenomenon 2) and includes a variety of neurologic and behavioral symptoms.98 A critical analysis of the literature related to acute and chronic low-level OP exposure concludes that reports have inconsistent symptoms, poor or absent correlation with OP exposure, or significant methodologic deficiencies. To date, no consistent relationship has been found between low-level OP exposure and peripheral neuropathy in either humans or experimental animals.98 Routine clinical laboratory studies in OPIDP are usually normal.168 Depressed erythrocyte AChE levels suggest exposure to OPs, and early severe weakness appears to be associated with AChE levels less than 20% of normal. AChE levels provide little information on the likelihood of developing OPIDP. Because erythrocyte AChE levels regenerate at the rate of about 1%/day, patients with previous exposure may have normal levels by the time they come to medical attention. Plasma pseudocholinesterase levels have little diagnostic value. The CSF is usually acellular, and the protein levels are normal or only minimally elevated. The diagnosis of OP-related neuropathy is simple if there is a clear indication of ingestion at about 2 weeks before the illness, including the presence of cholinergic symptoms. Should such evidence be lacking, this condition becomes almost impossible to establish with certainty. In the authors’ experience, cases of OP neurotoxicity have been erroneously diagnosed as multiple sclerosis. The subacute onset of a myeloneuropathy syndrome in a patient with possible OP exposure should raise the question of OPIDP.

Pathology Humans Sural nerve biopsy specimens demonstrate that the longest and largest diameter axons are most vulnerable, with varying degrees of axonal degeneration and regeneration. Demyelination is not a pathologic feature of OPIDP.76 Postmortem studies from the 1930 Jamaica ginger extract TOCP outbreak revealed wallerian degeneration in peripheral nerves. Evidence of CNS dysfunction included distal degeneration in the gracile fasciculus and the corticospinal tract, indicative of central-peripheral distal axonopathy.7 Experimental Animals Experimental animal studies have conclusively demonstrated that TOCP induces a distal axonopathy in several species.11 The neurotoxic effects of OPs vary according to the species of animal exposed. For example, neuropathy is clinically apparent in chickens, cats, and certain primates, but not in the rat, rabbit, and guinea pig. Histopathologic studies have demonstrated distal axonal degeneration in chickens,17 cats,21,131 and primates68 exposed to OPs. Such variability may be more apparent than real, however, because histopathologic studies revealed mild evidence of

2518

Diseases of the Peripheral Nervous System

neuropathy in many instances in which clinical signs were lacking. Degeneration of the long ascending and descending fiber tracts within the spinal cord and the ends of the dorsal spinocerebellar tracts in the medulla oblongata and cerebellum are present in pathologic specimens from OP-exposed experimental animals.17 In the cat PNS, large-diameter, heavily myelinated fibers appear most vulnerable.18 Axonal degeneration initially involves focal but nonterminal areas of the axon and spreads in a somatofugal direction to involve the entire distal axon.11 Ultrastructural studies of the axonal alterations describe nonspecific changes; accumulations of vesicular and membranous structures appear in the axoplasm before actual axonal degeneration.131

Physiology Humans Electromyography in individuals with OPIDP usually shows evidence of acute denervation and chronic motor unit reinnervation. Neurogenic abnormalities tend to be limited to distal limb muscles in mild instances, but may be found in more proximal leg muscles in severe cases. Motor nerve conduction is either normal or only mildly slowed in the legs and normal in the arms. Sensory nerve action potentials are usually absent in the legs and of reduced amplitude in the arms. The earliest electrophysiologic changes involve diminution of sensory, rather than motor, potential amplitudes. Electrophysiologic studies appear to be a sensitive test for screening patients exposed to OPs.89 Experimental Animals Unilateral OP neuropathy has been produced in cats by injecting DFP into the ipsilateral femoral artery.102 There is a profound loss of indirectly evoked post-tetanic potentiation of indirect twitch tension in the soleus muscle. This suggests that the lesion is located in the distal axon and that OP esters have a local toxic effect on nerve fibers.

Pathogenesis The initial biochemical event in the development of OPIDP involves inhibition of neuropathy target esterase (NTE, formerly neurotoxic esterase) by phosphorylation.79 This enzyme is distinct from AChE. The function of NTE is unknown; it is present in brain, spinal cord, and peripheral nerve, as well as non-neural tissues such as spleen, muscle, and lymphocytes. Not all OPs that inhibit NTE are neurotoxic, because some OPs that are able to inhibit NTE do not produce OPIDP. However, all neurotoxic OPs inhibit NTE, a characteristic that is used to predict whether an OP is potentially neurotoxic.

Phosphorylation of NTE is a rapid event, probably occurring within hours of exposure. A sufficient amount of NTE must be inhibited for OPIDP to develop (e.g., in the hen, about 70% to 80% of brain NTE must be inhibited).78 The amount of inhibited human NTE that is required before OPIDP occurs is not yet known, but appears to be less than required in the hen.101 The degree of NTE inhibition varies according to the OP and the source of NTE. For example, in the hen, brain NTE is most susceptible to inhibition by di-n-butyl-2,2-dichlorovinyl phosphate, followed by spinal cord NTE, and then nerve NTE.113 Inhibition of peripheral nerve NTE is required to develop OPIDP, and no clinical deficits result if only brain NTE is inhibited. Inhibition of cord, but not nerve, NTE results in a spastic spinal syndrome, without peripheral neuropathy. The ability of neuropathic OPs to inhibit NTE from multiple sources underlies the rationale for assaying lymphocyte NTE as an industrial screening procedure to predict the likelihood for developing OPIDP in individuals exposed to neurotoxic OPs.99 Studies in animals have documented the correlation between depressed lymphocyte NTE and, on postmortem examination, reduced nervous tissue NTE after intoxication by the combination of 2,4-dichlorophenoxyacetic acid, mecoprop, and chlorpyrifos.125 Lymphocyte NTE assays are still not widely used in industry. Although inhibited NTE is required to develop OPIDP, its presence alone is insufficient for the development of neuropathy. The development of OPIDP requires, in most cases, “aging” of the inhibited NTE enzyme, a nonenzymatic reaction involving cleavage of a lateral side chain from the phosphorylated NTE enzyme, leaving a charged monosubstituted phosphorous acid residue at the active site.27,28,80,171 This critical step apparently occurs in the nerve fiber rather than the cell body. The amount of aged inhibited NTE cannot be measured but is assumed to be proportional to the amount of inhibited NTE. The exact mechanism by which aged inhibited NTE causes OPIDP is unclear. Animal studies have correlated the presence of aged inhibited NTE with impaired retrograde transport of radioactive125 I-labeled tetanus toxin within the sciatic nerve.113 Maximum reduction in axonal transport occurs 7 days after exposure, prior to morphologic evidence of axonal degeneration. Retrograde transport fails to show progressive deterioration once axonal degeneration occurs, suggesting that transport dysfunction does not result from the axonal degeneration itself.112 Anterograde slow and rapid transport systems remain intact. Animals with insufficient NTE inhibition do not develop abnormalities of retrograde transport or clinical evidence of OPIDP. There has been some evidence suggesting that aging of NTE is not invariably necessary to develop OPIDP. Some OPs that inhibit but do not age NTE have been shown to cause neuropathy when given at extremely high doses, with almost complete inhibition of NTE. It is paradoxical that the severity of neuropathy in these cases is mild, despite the great amount of

Human Toxic Neuropathy Caused by Industrial Agents

inhibited NTE. Severity of the OPIDP does not appear to correlate with the amount of aged NTE.100 The development of OPIDP is, in most instances, associated with those OPs that inhibit and age NTE. Failure to age is probably related to the stability of phosphoruscarbon bonds of some OPs. Pretreatment with an OP capable of inhibiting NTE by aging protects against the subsequent development of OPIDP with exposure to other neurotoxic OPs. Infusion of such a prophylactic OP into a single limb of an animal, before administration of a neurotoxic OP, protects that appendage from neuropathy, while other extremities develop OPIDP.15 Animals pretreated with a prophylactic OP fail to demonstrate abnormalities of retrograde transport when exposed to neurotoxic OPs, supporting a link between aged inhibited NTE, impaired retrograde transport, and subsequent axonal degeneration.113 Impairment of retrograde transport may account for the accumulation of large vesicular and membranous structures seen in the nerve axoplasm before actual axonal degeneration.131 Many OPs that appear protective when given prior to the administration of a neurotoxic OP can potentiate neuropathy (promotion) if given shortly after exposure to a neurotoxic OP. The phenomenon by which the same OP can either protect against or promote neuropathy is possible because each condition is due to effects on different targets. Protection appears to involve submaximal inhibition of NTE, while promotion requires interaction with an as yet unknown target, but probably not NTE.111

POLYCHLORINATED BIPHENYLS Polychlorinated biphenyls (PCBs) have been extensively used as plasticizers and in electrical insulation. There has been considerable environmental pollution of waterways and marine life with PCBs. It is alleged, but still unproven, that developmental neurotoxicity has appeared in such populations following low-level, long-term environmental exposure.85,142 Outbreaks of mildly symptomatic sensory neuropathy occurred in 1968 in Japan,116 in 1978 in Taiwan24,25 when cooking oil became contaminated with PCBs, and in Finland146 among workers exposed following an explosion of capacitors. Reports of the Taiwanese and Finnish episodes feature meticulous longitudinal peripheral neurophysiologic evaluations. Affected individuals in Taiwan and Japan were repeatedly exposed from PCB-contaminated cooking oil; they developed acne, brown-pigmented nails, and a discharge from the meibomian glands. In the Taiwanese cases, blood levels of PCBs and their derivatives, polychlorinated quaterphenyls and polychlorinated dibenzofurans, were elevated at the initial evaluation, 2 years following exposure; they declined moderately after 4 years. A few persons experienced distal numbness and displayed absent tendon

2519

reflexes and slightly diminished pain and tactile sensation in the feet. Sensory nerve conduction was mildly slowed at the 2-year follow-up, and improved to near-normal values at the 4-year assessment. The Finnish exposure, a monophasic event, was unaccompanied by skin changes and only vague complaints of odd “pin prick sensations, itching or odd temperature feelings in arms and legs.”146 There was significant slowing of sensory nerve conduction values when measured 2 months after exposure, but sural nerve action potential amplitudes were unaffected. The nerve conduction abnormalities returned to near-normal levels 6 months later. A report of sensorimotor neuropathy in three transformer maintenance workers is unconvincing because of inadequate neurophysiologic documentation.6 Rats intoxicated with tetrachlorobiphenyl develop slowed motor conduction.121 The pathogenesis of PCB neuropathy is unknown. Biphenyl itself has also been reported to induce manifestations of sensorimotor neuropathy (thought to be demyelinating in type) and psychological changes in individuals using this compound as a fungistatic agent to preserve citrus fruits.61

PYRIMINIL (VACOR) Pyriminil (Vacor), a structural analogue of nicotinamide, is most commonly used as a rodenticide. Ingestion of Vacor results in a rapidly progressive, severe, distal axonopathy with accompanying diabetes mellitus, autonomic dysfunction, and encephalopathy.90 Diabetes is due to necrosis of pancreatic beta cells. Within hours of high-dose suicidal ingestion, distal limb weakness, sensory loss with impaired postural reflexes, and cranial nerve dysfunction ensue. Accompanying autonomic disturbances include hypotension, hypothermia, gastric hypomotility, and urinary retention. Occasionally there may be signs of CNS dysfunction, including visual impairment, myoclonic jerks, nystagmus, cerebellar dysfunction, and encephalopathy.126 Early administration of nicotinamide and gastric lavage may be helpful. Gradual recovery from the neuropathy occurs, but many patients remain severely diabetic with residual autonomic impairment. Animals given high-dose Vacor demonstrate early degeneration of distal peripheral nerve axons, possibly related to impaired fast antegrade axonal transport in distal nerve segments.170 The impairment of fast axonal transport in somatic and autonomic nerves may account for the rapid appearance of neuropathy after Vacor ingestion. The underlying pathogenetic mechanism remains unclear. Because early administration of nicotinamide can prevent severe dysfunction, it is postulated that Vacor inhibits a NAD-dependent enzyme essential to energy metabolism.

2520

Diseases of the Peripheral Nervous System

TRICHLORETHYLENE AND DICHLORACETYLENE Trichlorethylene (TCE) is used in dry cleaning and rubber production and as a degreasing agent. It was formerly widely employed as a general anesthetic. It is unclear whether chemically pure TCE is neurotoxic. It is suggested that cranial neuropathy stems from a decomposition product, dichloracetylene (DCA), resulting from interaction of TCE with alkaline materials. Acute industrial exposure or anesthetic exposure to TCE/DCA has been associated with a unique neurotoxic syndrome: prompt dysfunction of the trigeminal and, to a lesser extent, the facial, oculomotor, and optic nerves.39 Peripheral limb nerves have never been definitively implicated. The effect of TCE/DCA on the trigeminal nerve is so predictable that, for a time, victims of tic douloureux were intentionally exposed to it. Cranial nerves are affected after a period of exposure to high concentrations. Trigeminal neuropathy usually includes loss of sensory modalities in the entire distribution of the nerve, accompanied by weakness of muscles of mastication. Recovery occurs over a period of months, but patches of malar hypalgesia and absent corneal reflexes may be detected 12 years later.41 One postmortem report described extensive axonal and myelin degeneration in the trigeminal nerve and its spinal and main sensory nuclei; lesser degrees of neuronal degeneration were present in facial, abducens, and vestibular nuclei and in the nucleus of the tractus solitarius.14 Although most instances of trigeminal neuropathy have occurred following exposure to a mixture of TCE and its degradation product, several case reports strongly suggest that either agent alone may cause cranial neuropathy. Two reports describe vertigo and delayed onset of cranial neuropathy following inhalation of pure TCE at room temperature in the absence of alkali.88,120 Exposure to DCA alone has also caused facial sensory loss.67,136

REFERENCES 1. Aaserud, O., Hommeren, O. J., Tvedt, B., et al.: Carbon disulfide exposure and neurotoxic sequelae among viscose rayon workers. Am. J. Ind. Med. 18:25, 1990. 2. Alexeef, G. V., and Kilgore, W. W.: Methyl bromide. Residue Rev. 88:101, 1983. 3. Allen, N., Mendell, J. R., Billmaier, D. J., et al.: Toxic polyneuropathy due to methyl n-butyl ketone: an industrial outbreak. Arch. Neurol. 32:209, 1975. 4. Alpers, B. J., and Lewey, F. H.: Changes in the nervous system following carbon disulfide poisoning in animals and man. Arch. Neurol. Psychiatry 44:725, 1940. 5. Altenkirch, H., Mager, J., Stoltenburg, G., and Helmbrecht, J.: Toxic polyneuropathies after sniffing a glue thinner. J. Neurol. 214:152, 1977.

6. Altenkirch, H., Stoltenburg, G., Haller, D., et al.: Clinical data on three cases of occupationally induced PCB-intoxication. Neurotoxicology 17:639, 1996. 7. Aring, C. D.: The systemic nervous affinity of triorthocresyl phosphate (Jamaica ginger palsy). Brain 65:34, 1942. 8. Baker, T., and Stanec, A.: Methylprednisone treatment of an organophosphorous-induced delay neuropathy. Toxicol. Appl. Pharmacol. 79:348, 1985. 9. Beauchamp, R. O., Bus, J. S., Popp, J. A., et al.: A critical review of the literature on carbon disulfide toxicity. CRC Crit. Rev. Toxicol. 11:169, 1983. 10. Besser, R., Gutmann, L., Dillman, U., et al.: End-plate dysfunction in acute organophosphate intoxication. Neurology 39:561, 1989. 11. Bouldin, T. W., and Cavanagh, J. B.: Organophosphorous neuropathy. 1. A teased-fiber study of the early stages of axonal degeneration. Am. J. Pathol. 94:253, 1979. 12. Branch, R. A., and Jacqz, E.: Subacute neurotoxicity following long-term exposure to carbaryl. Am. J. Med. 80:741, 1986. 13. Buiatti, E., Cecchini, S., Ronchi, O., et al.: Relationship between clinical and electromyographic findings and exposure to solvents, in shoe and leather workers. Br. J. Ind. Med. 35:168, 1978. 14. Buxton, P. H., and Hayward, M.: Polyneuritis cranialis associated with industrial trichlorethylene poisoning. J. Neurol. Neurosurg. Psychiatry 30:511, 1967. 15. Caroldi, S., Lotti, M., and Masutti, A.: Intra-arterial injection of DFP or PMSF (phenylmethanesulfonyfluoride) produces respectively unilateral neuropathy or protection. Vet. Hum. Toxicol. 24:218,1982. 16. Cavalleri, F., Galassi, G., Ferrari, S., et al.: Methyl bromide induced neuropathy: a clinical, neurophysiological, and morphological study. J. Neurol. Neurosurg. Psychiatry 58:383, 1995. 17. Cavanagh, J. B.: The toxic effect of tri-ortho-cresyl phosphate on the nervous system. J. Neurol. Neurosurg. Psychiatry 17:163, 1954. 18. Cavanagh, J. B.: Peripheral nerve changes in ortho-cresyl phosphate poisoning in the cat. J. Pathol. Bacteriol. 87:365, 1964. 19. Cavanagh, J. B.: The pathokinetics of acrylamide intoxication: a reappraisal of the problem. Neuropathol. Appl. Neurobiol. 8:315, 1982. 20. Cavanagh, J. B., and Gysbers, M. F.: Ultrastructural features of the Purkinje cell damage caused by acrylamide in the rat: a new phenomenon in cellular neuropathology. J. Neurocytol. 12:413, 1983. 21. Cavanagh, J. B., and Patangia, G. N.: Changes in the central nervous system in the cat as a result of tri-o-cresyl phosphate poisoning. Brain 88:165, 1965. 22. Chang, Y. C.: Neurotoxic effects of n-hexane on the human central nervous system: evoked potential abnormalities in n-hexane polyneuropathy. J. Neurol. Neurosurg. Psychiatry 50:269, 1987. 23. Chavez, C. T., Hepler, R. S., and Straatsma, B. R.: Methyl bromide optic neuropathy. Am. J. Ophthalmol. 99:715, 1985. 24. Chia, L., and Chu, F.: Neurological studies on polychlorinated biphenyl (PCB)–poisoned patients. Am. J. Ind. Med. 5:117, 1984.

Human Toxic Neuropathy Caused by Industrial Agents 25. Chia, L., and Chu, F.: A clinical and electrophysiological study of patients with polychlorinated biphenyl poisoning. J. Neurol. Neurosurg. Psychiatry 48:894, 1985. 26. Chretien, M., Patey, G., Souyri, F., and Droz, B.: Acrylamideinduced neuropathy and impairment of axonal transport of proteins. II. Abnormal accumulations of smooth endoplasmic reticulum as sites of focal retention of fast transported proteins: electron microscopic radioautographic study. Brain Res. 205:15, 1994. 27. Clothier, B., and Johnson, M. K.: Rapid aging of neurotoxic esterase after inhibition by diisopropyl phosphorofluoridate. Biochem. J. 177:549, 1979. 28. Clothier, B., and Johnson, M. K.: Reactivation and aging of neurotoxic esterase inhibited by a variety of organophosphorous esters. Biochem. J. 185:739, 1980. 29. Cohen, A. E., Scheel, L. D., Kopp, J. F., et al.: Biochemical mechanisms in chronic carbon disulfide poisoning. Am. Ind. Hyg. J. 20:303, 1959. 30. Corsi, G., Maestrelli, P., Picotti, G., et al.: Chronic peripheral neuropathy in workers with previous exposure to carbon disulfide. Br. J. Ind. Med. 40:209, 1983. 31. Cranmer, M. F.: Carbaryl: a toxicological review and risk analysis. Neurotoxicology 7:247, 1986. 32. Dalvi, R. R.: Mechanism of the neurotoxic and hepatotoxic effects of carbon disulfide. Drug Metab. Drug Interact. 6:275, 1988. 33. DeBleecker, J.: The intermediate syndrome in organophosphate poisoning: an overview of experimental and clinical observations. Clin. Toxicol. 33:683, 1995. 34. DeCaprio, A. P., Briggs, R. G., Jackowski, S. J., and Kim, J. C. S.: Comparative neurotoxicity and pyrrole-forming potential of 2,5-hexanedione and perdeuterio-2,5-hexanedione in the rat. Toxicol. Appl. Pharmacol. 92:75, 1988. 35. DeCaprio, A. P.: n-Hexane, metabolites, and derivatives. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. New York, Oxford University Press, p. 633, 2000. 36. DeMatteis, F., and Seawright, A. A.: Oxidative metabolism of carbon disulfide by the rate: effect of treatments which modify the liver toxicity of carbon disulfide. Chem. Biol. Interact. 7:375, 1973. 37. Ecobichon, D. J.: Carbamic acid ester insecticides. In Ecobichon, D. J., and Joy, R. M. (eds.): Pesticides and Neurologic Diseases, 2nd ed. Boca Raton, FL, CRC Press, p. 251, 1994. 38. Eskin, T. A., Merigan, W. H., and Wood, R. W.: Carbon disulfide effects on the visual system: II: Retinogeniculate degeneration. Invest. Ophthalmol. Vis. Sci. 29:519, 1988. 39. Feldman, R. G.: Trichlorethylene. In Vinken, P. J., and Bruyn, G. W. (eds.): Handbook of Clinical Neurology, Vol. 36: Intoxications of the Nervous System. Amsterdam, NorthHolland, p. 457, 1979. 40. Feldman, R. G.: Carbon disulfide. In Occupational and Environmental Toxicology. Philadelphia, Lippincott–Raven Publishers, p. 400, 1999. 41. Feldman, R. G., White, R. F., Currie, J. N., et al.: Longterm follow-up after single toxic exposure to trichloroethylene. Am. J. Ind. Med. 8:119, 1985. 42. Ferraz, H. B., Bertolucci, P. H. F., Pereira, J. S., et al.: Chronic exposure to the fungicide maneb may produce symptoms and

43. 44.

45. 46.

47.

48. 49.

50.

51.

52.

53.

54. 55.

56.

57.

58. 59.

60.

61.

2521

signs of CNS manganese intoxication. Neurology 38:550, 1988. Fisher, J. R.: Guillain-Barre syndrome following organophosphate poisoning. JAMA 238:1950, 1977. Freeman, G., and Epstein, M. A.: Therapeutic factors in survival after lethal cholinesterase inhibition by phosphorous insecticides. N. Engl. J. Med. 253:266, 1955. Fujita, A., Shibata, M., Kato, H., et al.: Clinical observations on acrylamide poisoning. Nippon Iji Shimpo 1869:27, 1960. Fullerton, P. M., and Barnes, J. M.: Peripheral neuropathy in rats produced by acrylamide. Br. J. Ind. Med. 23:210, 1966. Gamier, R., Rambourg-Schepens, M. O., Muller, A., and Hallier, E.: Glutathione transferase activity and formation of macromolecular adducts in two cases of acute methyl bromide poisoning. Occup. Environ. Med. 53:211, 1996. Garland, T. O., and Paterson, M. W. H.: Six cases of acrylamide poisoning. Br. Med. J. 4:134, 1967. Genter St. Clair, M. B., Amarnath, V., Moody, M. A., et al.: Pyrrole oxidation and protein cross-linking as necessary steps in the development of ␥-diketone neuropathy. Chem. Res. Toxicol. 1:179, 1988. Gold, B. G., Griffin, J. W., and Price, D. L.: Slow axonal transport in acrylamide neuropathy: different abnormalities produced by single-dose and continuous administration. J. Neurosci. 5:1755, 1985. Gold, B. G., and Halleck, M. M.: Axonal degeneration and axonal caliber alterations following combined ␤,␤⬘iminodipropionitrile (IDPN) and acrylamide administration. J. Neuropathol. Exp. Neurol. 48:653, 1989. Gold, B. G., and Schaumburg, H. H.: Acrylamide. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology, 2nd ed. New York, Oxford University Press, p. 124, 2000. Goldman, L. R., Smith, D. F., Neutra, R. R., et al.: Pesticide food poisoning from contaminated watermelons in California. Arch. Environ. Health 45:222, 1990. Gordy, S. T., and Trumper, M.: Carbon disulfide poisoning with a report of six cases. JAMA 110:1543, 1938. Gottfried, M. R., Graham, D. G., Morgan, M., et al.: The morphology of carbon disulfide neurotoxicity. Neurotoxicology 6:89, 1985. Governa, M., Calisti, R., Coppa, G., et al.: Urinary excretion of 2,5-hexanedione and peripheral polyneuropathies in workers exposed to hexane. J. Toxicol. Environ. Health 20:219, 1987. Graham, D. G., Amarnath, V., Valentine, W. M., et al.: Pathogenic studies of hexane and carbon disulfide neurotoxicity. Crit. Rev. Toxicol. 25:91, 1995. Greenberg, M. K., and Swift, T. R.: Ethylene oxide peripheral neuropathy: four year follow-up. Muscle Nerve 5:557, 1982. Grober, E., and Schaumburg, H. H.: High level occupational exposure to methyl isobutyl ketone causes permanent impairment in working memory. Neurology 54:1853, 2000. Gross, J. A., Haas, M. L., and Swift, T. R.: Ethylene oxide neuro-toxicity: report of four cases and review of the literature. Neurology (Minneap.) 29:978, 1979. Hakkinen, I., Siltanen, E., Hernberg, S. D., et al.: Diphenyl poisoning in fruit paper production: a new health hazard. Arch. Environ. Health 26:70, 1973.

2522

Diseases of the Peripheral Nervous System

62. Harris, C. H., Gulati, A. K., Freidman, M. A., and Sickles, D. W.: Toxic neurofilamentous axonopathies and fast axonal transport. V. Reduced bi-directional vesicle transport in cultured neurons by acrylamide and glycidamide. J. Toxicol. Environ. Health 42:343, 1994. 63. Hashimoto, K., Sakamoto, J., and Tanil, H.: Neurotoxicity of acrylamide and related compounds and their effects on male gonads in mice. Arch. Toxicol. 47:179, 1981. 64. He, F., Jacobs, J. M., and Scaravilli, F.: The pathology of allyl chloride neurotoxicity in mice. Acta Neuropathol. 55:125, 1981. 65. He, F., Lu, B., Zhang, S., et al.: Chronic allyl chloride poisoning: an epidemiological, clinical, toxicological and neuropathological study. G. Ital. Med. Lav. 7:5, 1985. 66. He, F., Zhang, S., Wang, H., et al.: Neurological and electroneuromyographic assessment of the adverse effects of acrylamide on occupationally exposed workers. Scand. J. Work Environ. Health 15:125, 1989. 67. Henschler, D., Broser, F., and Hopf, H. C.: “Polyneuritis cranialis” following poisoning with chlorinated acetylenes while handling vinylidene chloride copolymers. Arch. Toxicol. 26:62, 1970. 68. Hern, J. E. C. (as quoted by Hopkins, A. P.: Toxic neuropathy due to industrial agents. In Dyck, P. J., Thomas, P. K, and Lambert, E. H. [eds.]: Peripheral Neuropathy, 2nd ed., Vol. II. Philadelphia, W. B. Saunders, p. 1213, 1975.) 69. Herskowitz, A., Ishii, N., and Schaumburg, H. H.: n-Hexane neuropathy: a syndrome occurring as a result of industrial exposure. N. Engl. J. Med. 285:82, 1971. 70. Herzstem, J., and Cullen, M. R.: Methyl bromide intoxication in four field workers during removal of soil fumigation sheets. Am. J. Ind. Med. 17:21, 1990. 71. Hollingsworth, R. L., Rowe, U. K., and Oyene, F.: Toxicity of ethylene oxide determined on experimental animals. Arch. Ind. Health 13:217, 1956. 72. Hunter, A., and Neal, R. A.: Response of the hepatic mixed function oxidase enzyme system to thionosulfur-containing compounds. Pharmacologist 16:239, 1974. 73. Igisu, H., Goto, I., Kawamura, Y., et al.: Acrylamide encephalopathy due to well water pollution. J. Neurol. Neurosurg. Psychiatry 38:581, 1975. 74. Israeli, R., Sculsky, M., and Tiberin, P.: Acute intoxication due to exposure to maneb and zineb: a case with behavioral and central nervous system changes. Scand. J. Work Environ. Health 9:47, 1983. 75. Jakobsen, J., and Sidenius, P.: Early and dose-dependent decrease of retrograde axonal transport in acrylamideintoxicated rats. J. Neurochem. 40:447, 1983. 76. Jedrzejowsha, H., Rowinska-Marcinska, K., and Hoppe, B.: Neuropathy due to phytosol (Agritox): report of a case. Acta Neuropathol. (Berl.) 49:163, 1980. 77. Johnson, B. L., Anger, W. K., Setzer, J. V., et al.: Neurobehavioral effects of methyl n-butyl ketone and methyl n-amyl ketone in rats and monkeys: summary of NIOSH investigators. In Proceedings 1st Annual NIOSH Scientific Symposium: Toxicological and Carcinogenic Health Hazards in the Workplace. Chicago, Pathotox Publishers, 1978. 78. Johnson, M. K.: The target for initiation of delayed neurotoxicity by organophosphate ester: biochemical studies and

79.

80.

81.

82. 83.

84.

85.

86.

87. 88.

89.

90. 91.

92.

93.

94.

toxicological applications. Rev. Biochem. Toxicol. 4:141, 1982. Johnson, M. K.: Receptor or enzyme: the puzzle of NTE and organophosphate-induced delayed polyneuropathy. Trends Pharmacol. Sci. 8:174, 1987. Johnson, M. K.: Contemporary issues in toxicology. Organophosphates and delayed neuropathy—is NTE alive and well? Toxicol. Appl. Pharmacol. 102:385, 1990. Juntunen, J., Linnoila, I., and Haltia, M.: Histochemical and electron microscopic observations on the myoneural junctions of rats with carbon disulfide induced polyneuropathy. Scand. J. Work Environ. Health 3:36, 1977. Kane, F. J. Jr.: Carbon disulfide intoxication from overdose of disulfiram. Am. J. Psychiatry 127:690, 1970. Kantarjian, A. D., and Shaheen, A. S.: Methyl bromide poisoning with nervous system manifestations resembling polyneuropathy. Neurology 13:1054, 1963. Kid, J. G., and Langworthy, O. R.: Paralysis following the injection of Jamaica ginger extract adulterated with triorthocresyl phosphate. Johns Hopkins Med. J. 52:39, 1933. Kimbrough, R. O., Doemland, M. L., and Krouskzs, C. A.: Analysis of research studying the effects of polychlorinated biphenyls and related chemicals on neurobehavioral development in children. Vet. Hum. Toxicol. 43:330, 2001. Knave, B., Kolmodin-Hedman, B., Persson, H. E., and Goldberg, J. M.: Chronic exposure to carbon disulfide: effects on occupationally exposed workers with special reference to the nervous system. Work Environ. Health 11:49, 1974. Korobkin, R., Asbury, A. K., Sumner, A. J., and Nielsen, S. L.: Glue-sniffing neuropathy. Arch. Neurol. 32:158, 1975. Lawrence, W. H., and Partyka, E. K.: Chronic dysphagia and trigeminal anesthesia after trichloroethylene exposure. Ann. Intern. Med. 95:710, 1981. LeQuesne, P. M.: Neurophysiological investigation of subclinical and minimal toxic neuropathies. Muscle Nerve 1:392, 1978. LeWitt, P.: The neurotoxicity of the rat poison Vacor. N. Engl. J. Med. 302:73, 1980. Lieben, J., and Williams, R. A.: Five years of experience with CS2. In Xintras, C., Johnson, B. L., and deGroot, I. (eds.): Behavioral Toxicology: Early Detection of Occupational Hazards (HEW Pub. No. [NIOSH] 74–126). Washington, DC, U.S. Dept. of Health, Education, and Welfare, p. 50, 1974. Linnoila, I., Haltia, M., Seppalainen, A. M., and Palo, J.: Experimental carbon disulfide poisoning: morphological and neurophysiological studies. In Kornyey, S., Tarishka, S., and Gosztonyi, G. (eds.): Proceedings of the VII International Congress of Neuropathology. Amsterdam, Excerpta Medica Foundation, p. 383, 1975. LoPachin, R. M., Castiglia, C. M., and Saubermann, A. J.: Acrylamide disrupts elemental composition and water content of rat tibial nerve. I. Myelinated axons. Toxicol. Appl. Pharmacol. 115:21, 1992. LoPachin, R. M., Castiglia, C. M., and Saubermann, A. J.: Acrylamide disrupts elemental composition and water content of rat tibial nerve. II. Schwann cells and myelin. Toxicol. Appl. Pharmacol. 115:35, 1992.

Human Toxic Neuropathy Caused by Industrial Agents 95. LoPachin, R. M., and Lehning, E. J.: Acrylamide induced distal axon degeneration: a proposed mechanism of action. Neurotoxicology 15:247, 1994. 96. LoPachin, R. M., Lehning, E. J., Stack, E. C., et al.: 2,5Hexanedione alters elemental composition and water content of rat peripheral nerve myelinated axons. J. Neurochem. 63:2266, 1994. 97. Lotti, M.: Promotion of organophosphate induced delayed polyneuropathy by certain esterase inhibitors. Chem. Biol. Interact. 119–120:519, 1999. 98. Lotti, M.: Low-level exposure to organophosphorous esters and peripheral nerve function. Muscle Nerve 25:492, 2002. 99. Lotti, M., Becker, C. E., and Aminoff, M. J.: Organophosphate polyneuropathy: pathogenesis and prevention. Neurology 34:658, 1984. 100. Lotti, M., Moretto, A., Capodicase, E., et al.: Interactions between neuropathy target esterase and its inhibitors and the development of polyneuropathy. Toxicol. Appl. Pharmacol. 122:165, 1993. 101. Lotti, M., Moretto, A., Zoppellari, R., et al.: Inhibition of lymphocytic neuropathy target esterase predicts the development of organophosphate-induced delayed polyneuropathy. Arch. Toxicol. 59:176, 1986. 102. Lowndes, H. E., Baker, T., and Riker, W. F.: Motor nerve dysfunction in delayed DFP neuropathy. Wur. J. Pharmacol. 29:66, 1974. 103. Lynch, J. J., Merigan, W. H., and Eskin, T. A.: Subchronic dosing of macaques with 2,5-hexanedione causes longlasting motor dysfunction but reversible visual loss. Toxicol. Appl. Pharmacol. 98:166, 1989. 104. Manu, P., Lancranjan, I., and Zamfirescu, M.: Investigation of serum LDH and its isoenzymatic model in carbon disulfide polyneuritis. Int. Arch. Arbeitsmed. 26:45, 1970. 105. Maurissen, J. P., Weiss, B., and Cox, C.: Vibration sensitivity recovery after a second course of acrylamide intoxication. Fund. Appl. Toxicol. 15:93, 1990. 106. Medrano, C. J., and LoPachin, R. M.: Effects of acrylamide and 2,5-hexanedione on brain mitochondrial respiration. Neurotoxicology 10:249, 1989. 107. Metcalf, D. R., and Holmes, J. H. V. II: Toxicology and physiology: EEG, psychological and neurological alterations in humans with organophosphorous exposure. Ann. N. Y. Acad. Sci. 160:357, 1969. 108. Meuling, W. J., Bragt, P. D., and Braun, C. L.: Biological monitoring of carbon disulfide. Am. J. Ind. Med. 17:247, 1990. 109. Miller, M. S., Miller, M. J., Burks, T. F., and Sipes, I. G.: Altered retrograde axonal transport of nerve growth factor after single and repeated doses of acrylamide in the rat. Toxicol. Appl. Pharmacol. 69:96, 1983. 110. Miller, M. S., and Spencer, P. S.: Single doses of acrylamide reduce retrograde transport velocity. J. Neurochem. 43:1401, 1984. 111. Moretto, A., Bertolazzi, M., and Lotti, M.: The phosphorothioic acid-O-(2-chloro-2,3,3-trifluorocyclobutyl)-O-ethyl S-propyl ester promotes organophosphate polyneuropathy without inhibition of neuropathy target esterase. Toxicol. Appl. Pharmacol. 129:133, 1994. 112. Moretto, A., Lotti, M., Sabri, M. I., and Spencer, P. S.: Progressive deficit of retrograde axonal transport is associated

113.

114.

115

116. 117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

2523

with the pathogenesis of di-n-butyl dichlorvos axonopathy. J. Neurochem. 49:1515, 1987. Moretto, A., Lotti, M., and Spencer, P. S.: In vivo and in vitro regional differential sensitivity of neuropathy target esterase to di-n-butyl-2-2-dichlorovinyl phosphate. Arch. Toxicol. 63:469, 1989. Moses, H., and Klawans, H. L.: Bromide intoxication. In Vinken, P. J., and Bruyn, G. W. (eds.): The Handbook of Clinical Neurology. Elsevier, Amsterdam, p. 291, 1979. Munch, G., Lincoln, J., Maynard, K. I., et al.: Effects of acrylamide on cotransmission in perivascular sympathetic and sensory nerves. J. Auton. Nerv. Syst. 49:197, 1994. Murai, Y., and Kuroiwa, Y.: Peripheral neuropathy in chlorobiphenyl poisoning. Neurology 21:1173, 1971. Myers, J. E.: Acrylamide neuropathy in a South African factory: an epidemiologic investigation. Am. J. Ind. Med. 19:487, 1991. Nagata, H., Ohkoshi, N., Kanazawa, I., et al.: Rapid axonal transport velocity is reduced in experimental ethylene oxide neuropathy. Med. Chem. Neuropathol. 17:209, 1992. Nishimura, M., Urneda, M., Ishizu, S., and Sato, M.: Effect of methyl bromide on cultured mammalian cells. J. Toxicol. Sci. 5:321, 1980. Noseworthy, J. H., and Rice, G. P. A.: Trichloroethylene poisoning mimicking multiple sclerosis. Can. J. Neurol. Sci. 15:87, 1988. Ogawa, M.: Electrophysiological and histological studies of experimental chlorobiphenyl poisoning. Fukuoka Acts Med. 62:74, 1971. Ohnishi, A.: Ethylene oxide. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. New York, Oxford University Press, p. 563, 2000. Ohnishi, A., Inoue, N., Yamamoto, T., et al.: Ethylene oxide neuropathy in rats: exposure to 250 ppm. J. Neurol. Sci. 74:215, 1986. Ohnishi, A., Yamamoto, T., Murai, Y., et al.: Propylene oxide causes central-peripheral distal axonopathy in rats. Arch. Environ. Health 43:353, 1988. Osterloh, J., Lotti, M., and Pond, S. M.: Toxicologic studies in a fatal overdose of 2,4-D, MCPP and chlorpyrifos. J. Anal. Toxicol. 7:125, 1983. Osterman, J., Zmyslinski, R. W., Hopkins, C. B., et al.: Full recovery from severe orthostatic hypotension after Vacor rodenticide ingestion. Arch. Intern. Med. 141:1505, 1981. Perbellini, L., DeGrandis, D., Semenzato, F., et al.: An experimental study on the neurotoxicity of n-hexane metabolites: hexanol-1 and hexanol-2. Toxicol. Appl. Pharmacol. 46:421, 1978. Politis, M. I., Pellegrino, R. G., and Spencer, P. S.: Ultrastructural studies of the dying-back process. 5. Axonal neurofilaments accumulate at sites of 2,5-hexanedione application: evidence of nerve fibre dysfunction in experimental hexacarbon neuropathy. J. Neurocytol. 9:505, 1980. Post, E. J., and McLeod, J. G.: Acrylamide autonomic neuropathy in the cat. Part I. Neurophysiological and histological studies. J. Neurol. Sci. 33:353, 1977. Post, E. J., and McLeod, J. G.: Acrylamide autonomic neuropathy in the cat. Part II. Effects on mesenteric vascular control. J. Neurol. Sci. 33:375, 1977.

2524

Diseases of the Peripheral Nervous System

131. Prineas, J.: The pathogenesis of dying-back polyneuropathies. Part I. An ultrastructural study of experimental tri-orthocresyl phosphate intoxication in the cat. J. Neuropathol. Exp. Neurol. 28:571, 1969. 132. Prineas, J.: The pathogenesis of dying-back polyneuropathies. Part II. An ultrastructural study of experimental acrylamide intoxication in the cat. J. Neuropathol. Exp. Neurol. 28:598, 1969. 133. Rebert, C. S., and Becker, E.: Effects of inhaled carbon disulfide on sensory-evoked potentials of Long-Evans rats. Neurobehav. Toxicol. Teratol. 8:533, 1986. 134. Rice, J. M., and Wilbourn, J. D.: Tumors of the nervous system in carcinogenic hazard identification. Toxicol. Pathol. 28:202, 2000. 135. Rosenberg, C. K., Genter, M. B., Szakal-Quin, G., et al.: d,lVersus meso- 3,4-dimethyl-2,5-hexanedione: a morphometric study of the proximo-distal distribution of axonal swellings in the anterior root of the rat. Toxicol. Appl. Pharmacol. 87:363, 1986. 136. Saunders, R. A.: A new hazard in closed environmental atmospheres. Arch. Environ. Health 14:380, 1967. 137. Schaumburg, H. H., Arezzo, J. C., and Spencer, P. S.: Delayed onset of distal axonal neuropathy in primates after prolonged low-level administration of a neurotoxin. Ann. Neurol. 26:576, 1989. 138. Schaumburg, H. H., and Spencer, P. S.: Environmental hydrocarbons produce degeneration in cat hypothalamus and optic tract. Science 199:199, 1978. 139. Schaumburg, H. H., and Spencer, P. S.: Clinical and experimental studies of distal axonopathy—a frequent form of brain and nerve damage produced by environmental chemical hazards. Ann. N. Y. Acad. Sci. 329:14, 1979. 140. Schaumburg, H. H., Wisniewski, H. M., and Spencer, P. S.: Ultra-structural studies of the dying-back process. I. Peripheral nerve terminal and axon degeneration in systemic acrylamide intoxication. J. Neuropathol. Exp. Neurol. 33:260, 1974. 141. Scheel, L. D.: Biological changes involving metal ion shifts. Am. Ind. Hyg. Assoc. J. 26:585, 1965. 142. Schell, J. D., Budinsky, R. A., and Wernke, M. J.: PCB’s and neurodevelopmental effects in Michigan children: an evaluation of exposure and dose characterization. Reg. Toxicol. Pharmacol. 33:300, 2001. 143. Seppalainen, A. M., and Haltia, M.: Carbon disulfide. In Spencer, P. S. and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 356, 1980. 144. Seppalainen, A. M., and Linnoila, I.: Electrophysiological findings in rats with experimental carbon disulfide neuropathy. Neuropathol. Appl. Neurobiol. 2:209, 1976. 145. Seppalainen, A. M., and Tolonen, M.: Neurotoxicity of longterm exposure to carbon disulfide in the viscose rayon industry: a neurophysiologic study. Work Environ. Health 11:145, 1974. 146. Seppalainen, A. M., Vuojolahti, P., and Elo, O.: Reversible nerve lesions after accidental polychlorinated biphenyl exposure. Scand. J. Work Environ. Health 11:91, 1985. 147. Shore, R. E., Gardner, M. J., and Pannett, B.: Ethylene oxide: an assessment of the epidemiologic evidence on carcinogenicity. Br. J. Ind. Med. 50:971, 1993.

148. Sickles, D. W., Fowler, S. R., and Testino, A. R.: Effects of neuro-filamentous axonopathy-producing neurotoxicants on in-vitro production of ATP by brain mitochondria. Brain Res. 528:25, 1990. 149. Sidenius, P., and Jakobsen, J.: Anterograde axonal transport in rats during intoxication with acrylamide. J. Neurochem. 40:697, 1983. 150. Smith, H. V., and Spalding, J. M. K.: Outbreak of paralysis in Morocco due to orthocresyl phosphate poisoning. Lancet 2:1019, 1959. 151. Spencer, P. S., and Butterfield, P.: Environmental agents and Parkinson’s disease. In: Ellenberg, H. J., and Koller, R. (eds.): Etiology of Parkinson’s Disease. New York, Marcel Dekker, p. 319, 1995. 152. Spencer, P. S., Sabri, M. I., Schaumburg, H. H., and Moore, C. L.: Does a defect of energy metabolism in the nerve fiber underlie axonal degeneration in polyneuropathies? Ann. Neurol. 5:501, 1979. 153. Spencer, P. S., and Schaumburg, H. H.: A review of acrylamide neurotoxicity. Part I. Properties, uses and human exposure. Can. J. Neurol. Sci. 1:143, 1974. 154. Spencer, P. S., and Schaumburg, H. H.: Ultrastructural studies of the dying-back process. III. The evolution of experimental peripheral giant axonal degeneration. J. Neuropathol. Exp. Neurol. 36:276, 1977. 155. Spencer, P. S., and Schaumburg, H. H.: Ultrastructural studies of the dying-back process. IV. Differential vulnerability of PNS and CNS fibers in experimental central-peripheral distal axonopathies. J. Neuropathol. Exp. Neurol. 36:300, 1977. 156. Squier, M. V., Thompson, J., and Rajgopalan, B.: Case report: neuro-pathology of methyl bromide intoxication. Neuropathol. Appl. Neurobiol. 18:579, 1992. 157. Sterman, A. B.: Altered sensory ganglia in acrylamide neuropathy: quantitative evidence of neuronal reorganization. J. Neuropathol. Exp. Neurol. 42:166, 1983. 158. Sterman, A. B.: Acrylamide-induced remodeling of perikarya in rat superior cervical ganglia. Neuropathol. Appl. Neurobiol. 10:221, 1984. 159. Sumner, A. J., and Asbury, A. K.: Physiological studies of the dying-back phenomenon: muscle stretch afferents in acrylamide neuropathy. Brain 98:91, 1975. 160. Sumner, S. C. J., MacNeela, J. P., and Fennell, T. R.: Characterization and quantitation of urinary metabolites of (1,2,3-13C) acrylamide in rats and mice using 13C nuclear magnetic resonance spectroscopy. Chem. Res. Toxicol. 5:81, 1992. 161. Szendzikowski, S., Stetkiewicz, J., Wronska-Nofer, T., and Zdrajkowska, I.: Structural aspects of experimental carbon disulfide neuropathy. I. Development of the neurohistological changes in chronically intoxicated rats. Int. Arch. Arbeitsmed. 31:135, 1973. 162. Takahashi, A., Mizutani, M., Agr, B., et al.: Acrylamideinduced neurotoxicity in the central nervous system of Japanese quails. J. Neuropathol. Exp. Neurol. 53:276, 1994. 163. Tilson, H. A.: The neurotoxicity of acrylamide: an overview. Neurobehav. Toxicol. Teratol. 3:445, 1981. 164. Tornquist, H. L., Nordander, C., Rosen, I., et al.: Health effects of occupational exposure to acrylamide using hemoglobin adducts as biomarkers of internal dose. Scand. J. Work Environ. Health 27:219, 2001.

Human Toxic Neuropathy Caused by Industrial Agents 165. Vasilescu, C.: Sensory and motor conduction in chronic carbon disulfide poisoning. Eur. Neurol. 14:447, 1976. 166. Veronesi, B., Peterson, E. R., Bornstein, M. B., and Spencer, P. S.: Ultra-structural studies of the dying-back process. VI. Examination of nerve fibers undergoing giant axonal degeneration in organotypic culture. J. Neuropathol. Exp. Neurol. 42:153, 1983. 167. Vigliani, E. C.: Carbon disulfide poisoning in viscose rayon factories. Br. J. Ind. Med. 11:235, 1954. 168. Wadia, R. S., Chitra, S., Amin, R. B., et al.: Electrophysiologic studies in acute organophosphate poisoning. J. Neurol. Neurosurg. Psychiatry 50:1442, 1987. 169. Wadia, R. S., Sadagopan, C., Amin, R. B., and Sardesai, H. V.: Neurological manifestations of organophosphorous insecticide poisoning. J. Neurol. Neurosurg. Psychiatry 37:841, 1974.

2525

170. Watson, D. F., and Griffin, J. W.: Vacor neuropathy: ultrastructural and axonal transport studies. J. Neuropathol. Exp. Neurol. 46:96, 1987. 171. Williams, D. G.: Intramolecular group transfer is a characteristic of neurotoxic esterase and is independent of the tissue source of the enzyme. Biochem. J. 209:817, 1983. 172. Windebank, A. J., and Blexrud, M. D.: Residual ethylene oxide in hollow fiber hemodialysis units is neurotoxic in vitro. Ann. Neurol. 26:63, 1989. 173. Windebank, A. J., Blexrud, M. D., and de Groen, P. C.: Potential neurotoxicity of the solvent vehicle for cyclosporine. J. Pharmacol. Exp. Ther. 268:1051, 1994. 174. Zhu, M., Spink, D. C., Yan, B., et al.: Formation and structure of crosslinking and monomeric pyrrole autoxidation products in 2,5-hexanedione-treated amino acids, peptides, and protein. Chem. Res. Toxicol. 7:551, 1974.

113 Metal Neuropathy ANTHONY J. WINDEBANK

Lead Clinical Aspects Laboratory Studies Pathology Treatment Arsenic Clinical Aspects Laboratory Studies Pathology

Treatment Mercury Clinical Aspects Laboratory Studies Pathology Treatment Thallium Clinical Aspects Laboratory Studies

The inorganic and organic compounds of heavy metals produce peripheral nerve disease in humans. Exposure is often related to industrial or agricultural use, although mercury, thallium, arsenic, gold, and platinum have been administered as therapeutic agents. The most common causes of arsenic and thallium intoxication are now attempted suicide and homicide. Platinum (atomic number [AN] 78), gold (AN 79), mercury (AN 80), thallium (AN 81), and lead (AN 82) are classified as “heavy metals” because of their atomic weights (195 to 207.2). Arsenic compounds are toxic and usually considered with this group, although arsenic (AN 33; atomic weight 74.9) is not a “heavy” metal. In terms of clinical presentation and pathologic changes in the nervous system, there are significant differences among the various elements. However, the general rules of toxicology apply; the onset of disease can be temporally related to exposure, and removal from exposure leads to a cessation of progression or to remission. Heavy metals are deposited in a variety of organs, particularly bones. Slow release from this depot back into the circulation may delay recovery. The toxic effects of metals are not limited to the peripheral nervous system. Thus the neuropathy usually appears in the context of a systemic disease involving the gastrointestinal system, the hematopoietic system, and not infrequently the central nervous system. Because of the diversity of the effects of the metals, they cannot be considered as a single group. Rather, each element produces its own spectrum of disease, which may be recognized as a distinct clinical entity.

Tissue Distribution and Pathology Treatment Gold Platinum Clinical Aspects Laboratory Studies Pathology Treatment

LEAD The classic description of lead neuropathy, that of asymmetrical scattered weakness, wristdrop, and little sensory involvement, dates from the 19th century. However, the association of colic, pallor, and weakness as a result of lead poisoning has been recognized since the time of Hippocrates. In ancient times this reflected the use of metallic lead as a lining for wine vats or containers. The elemental metal is particularly suitable for such containers because of its malleability and relative inertness. Because of its low solubility, lead was used in water systems from Roman times until the middle of the 20th century.68

Clinical Aspects Sources of Lead As with many of the metals, lead is toxic to the nervous system in both organic and inorganic forms. The organic tetraalkyl derivatives are fat soluble and cross the blood-brain barrier with ease. Peripheral neuropathy has not been reported as a direct toxic effect of these organic compounds. Lead poisoning reached epidemic proportions during the industrial revolution in the 19th century as a result of the widespread use of lead in industry, its use as a paint pigment, and the increased mining of the metal to meet these demands without any concept of industrial safety precautions. Lead and silver mining produced significant lead-associated disease. Using hospital and medical records of the time, Richards estimated that 85,000 cases 2527

2528

Diseases of the Peripheral Nervous System

of lead poisoning occurred in the state of Utah between 1872 and 1904 as a direct result of silver mining.169 Virtually all of our clinical information and much of the pathologic data come from this period of time. With the realization by industries of the toxicity of lead and the increasing use of safety precautions to avoid contamination of the air with lead dust and fumes, the occurrence of overt lead neuropathy has become a rarity. Thus, in a review of cases at Mayo Clinic between 1949 and 1989, during which period about 3.5 million new patients were seen, there were fewer than 10 cases of lead neuropathy in adults. Although overt lead poisoning has decreased, this does not reflect a decrease in the use of lead. It has been estimated that the worldwide annual production of lead was about 80,000 tons during the height of the Roman Empire and about 1 million tons at the beginning of the 20th century, and is about 3 to 4 million tons at the present time.97,186 In 1980, it was estimated that 400,000 metric tons of lead were annually dispersed into the atmosphere, predominantly in the form of exhaust emissions from automobile engines powered by fuels containing lead “antiknock” compounds.186 Tetraalkyl lead additives produced a particulate postcombustion emission with lead in the oxide, chloride, and bromochloride forms. In the presence of air and sunlight, the halides are converted to oxides and carbonates.4,97,205 The inorganic lead salts are deposited in the dust in the vicinity of streets and highways. This has been reduced steadily in countries that have reduced or removed lead from gasoline. However, many countries continue to use lead additives in gasoline, and environmental pollution is still a major problem.2,135,166,176 Overall changes in the atmospheric content of lead have been determined by measuring the lead concentration at different levels of the polar ice cap. A 400-fold increase between prehistoric times and the present has been demonstrated. Studies on human remains buried before the widespread use of lead have shown an approximately 100-fold increase in the lead content of bone between 3000 BC and the present.56,79,187 Exposure to inorganic lead therefore occurs in all individuals. Because no physiologic function has been demonstrated, there are no a priori reasons for assuming that there is a safe lower limit to this exposure. However, overt neuropathy is rarely, if ever, seen in individuals with blood lead levels below 80 ␮g/dL.27 Such levels are not approached in the general population, even among individuals who are chronically exposed to high levels of automobile exhaust or gasoline fumes.87 Workers in a number of industrial processes utilizing lead may have values that approach or exceed this “safe” level6,27,184 (Table 113–1). The particular occupations in which excessive lead exposure may be expected are battery manufacture; smelting of leadcontaining ores; manufacture of pipe, sheet lead, solder, and lead alloys; use of industrial paints containing lead; and lead shot manufacture.14,37,43,54,65,227 There are also

Table 113–1. Commonly Accepted Values for Whole-Blood and Urine Lead Concentrations Whole Blood (␮g/dL) Normal* Industrial safe exposure Borderline Industrial harmful exposure

⬍40 ⬍70 60–100 ⬎80

Urine (␮g/24 h) ⬍100 ⬍300 200–400 ⬎300

*The term “normal” should be interpreted cautiously because a low threshold value for biologic safety has not been demonstrated conclusively.

many individual reports of unusual cases of lead poisoning resulting from the burning of batteries to provide heat80 and the use of lead pipes or lead-soldered radiators to distill illicit liquor.11,218 It has also been reported occasionally in association with lead shot embedded in the body98 and in people working on indoor firing ranges. In children with pica, the lead-containing paint flakes from old houses should be suspected as a source of exposure. Clinical Manifestations Lead neuropathy usually occurs in the context of a systemic illness, which may be acute, subacute, or chronic. The typical clinical triad is that of abdominal pain and constipation, anemia, and neuropathy. The abdominal colic is thought to be due to an effect of ingested lead on gastrointestinal mobility, although this has never been demonstrated directly. Anemia is probably always present in cases with clinically demonstrable neuropathy. Additional systemic features include nephropathy and gout. An unusual finding in chronic severe lead poisoning is a bluish discoloration of the gums just below the teeth. This is often associated with periodontal disease. Biopsy of the affected gum shows deposition of lead in the dermis.206 The general systemic effects of lead toxicity have been well reviewed by others.78 Lead poisoning in children usually presents with encephalopathy rather than neuropathy. In adults, there is a painless progressive motor neuropathy in which the upper limbs are affected more than the lower limbs with localized weakness that may be asymmetrical. In contrast to other metal neuropathies, there is a lack of sensory symptoms or signs, and tenderness of the nerves is not present. If sensory changes are present, they are overshadowed by the motor neuropathy. In typical cases of wristdrop from lead neuropathy, weakness commences with the proximal extensors of the middle and ring fingers, with later involvement in the index and little fingers, followed by weakness of extension of the thumb.226 With more involvement, wrist extensors become weak. The weakness and atrophy are often not confined to the distribution of the radial nerve and can include thenar muscles, especially

Metal Neuropathy

the abductor pollicis brevis, and interossei. In these latter cases, the weakness suggests involvement of the C8 and T1 nerve roots rather than the radial and median nerves themselves. Weakness of the upper limbs may involve limb muscles supplied by the C5 and C6 nerve roots; rarely, proximal muscles of the shoulder girdle are affected. Weakness in the lower limbs, usually footdrop, is found in children more than in adults.179,185 Tendon reflexes are decreased or absent. Occurrence of corticospinal tract signs would raise the possibility of a diffuse motor neuron disease.

Laboratory Studies In order to make a diagnosis of lead-induced neuropathy, a number of criteria should be fulfilled: 1. The patient should have a neuropathy demonstrated on a clinical, pathologic, or electrophysiologic basis. 2. There should be evidence of involvement of other organ systems. 3. There should be suitable exposure corresponding to progression of disease; removal from exposure should result in stabilization or remission. 4. There should be evidence of an increased tissue or body fluid content of lead. There is still considerable discussion concerning diagnostic electrophysiology in human lead neuropathy. One group of reports relates to nerve conduction velocity in lead-exposed workers. Most reports6,27,31,184 suggest that there are minor abnormalities of nerve conduction in workers with blood lead levels in the region of 50 to 140 ␮g/dL. These levels straddle the usually accepted upper limit of safety for industrial workers of 70 to 80 ␮g/dL. However, different authors employed different techniques, and therefore comparison of findings is difficult. Catton and colleagues found no difference in the maximal motor or sensory nerve conduction velocity when lead-exposed workers were compared with age-matched controls.31 However, they found a statistically significant decrease in the ratio of the amplitude of the compound muscle action potential (CMAP) of the extensor digitorum brevis muscle when stimulated at the knee compared with the ankle. This implied some minor degree of conduction block or dispersion without fiber loss between these points. The possibility that the different occupations of the subjects versus the controls might have resulted in exposure of the former group to unrelated low-grade traumatic peroneal palsy was not explored. These workers appeared to have had very significant lead exposure. Seven of 14 had lead levels greater than 120 ␮g/dL, seven had hemoglobin values less than 12 g/dL, and three had lead lines on the gums. The group studied by Seppäläinen and colleagues was composed of 26 storage battery factory workers with blood lead levels less than 70 ␮g/dL.183 When compared with the

2529

controls, there was a slight but significant slowing of maximal motor conduction velocity for the ulnar and median nerves and in the conduction velocity of the slower motor fibers. Although they were relatively reduced, the values were still within the normal range. Sensory conduction parameters were normal. Buchthal and Behse studied a group of 20 workers with maximal blood lead levels of 70 to 140 ␮g/dL.27 They also found minor abnormalities of nerve conduction, which they considered of doubtful clinical significance. Motor conduction velocity in the median nerve was reduced by a mean of 5.5 m/s relative to controls, which was highly significant in statistical terms, although both groups remained within the laboratory’s normal range. Abnormalities of sensory conduction to a similar degree were demonstrated. Electrophysiologic changes in lead-exposed asymptomatic populations have been repeatedly demonstrated.14,37,43,54,65,175,227 Most studies did not find any correlation between blood lead level and the degree of slowing of nerve conduction velocity. However, in serial measurements of workers treated with calcium disodium ethylenediaminetetraacetic acid (calcium disodium EDTA) chelation therapy, Araki and colleagues found a significant correlation between the fall in blood lead level and an increase in median nerve motor conduction velocity of 4.0 m/s.6 The group showing improvement was selected on the basis of their motor nerve conduction velocity (MNCV) response to EDTA, which was then correlated with change in blood lead level and compared with a group in which the MNCV did not respond. Initial blood lead levels did not exceed 85 ␮g/dL in any of their subjects. In spite of difference in techniques of these studies, the conclusion is inescapable that at or just above the upper limit of safe occupational exposure, as judged by blood lead levels, abnormalities of nerve conduction begin to appear. It is also noteworthy that, in all of the affected workers who did not have a clinical neuropathy, electromyographic evidence of denervation (fibrillation potentials or alterations in motor unit potential configuration) was absent. These findings contrast with the few cases in which electrophysiologic data in patients with overt lead neuropathy have been presented. The majority of these find electromyographic signs of denervation accompanied by decreased amplitude of motor responses and relatively little slowing of conduction velocity.19,27,49,93,154,175,185,189 Hematologic changes are the most sensitive indicators of lead exposure and intoxication. A diagnosis of clinically significant lead neuropathy in a patient without anemia should be made with great caution. The anemia of lead poisoning is usually microcytic and hypochromic secondary to decreased hemoglobin synthesis. Although a coincident iron deficiency state may exist, the red cell abnormalities may be seen with normal serum iron levels.125 Basophilic stippling of red cells is due to clusters of ribosomes retained within mature erythrocytes.105 Stippling may be more prominent

2530

Diseases of the Peripheral Nervous System

in red cell precursors in the bone marrow than in peripheral red cells.215 This phenomenon is not specific to lead and does not correlate well with levels of lead exposure.81 The anemia is produced, at least in part, by inhibition of heme synthesis, although there is experimental evidence that erythrocyte fragility is increased, with a consequent decrease in cell survival.127,177 Lead appears to have a direct effect on the synthesis of hemoglobin, principally by its effect on porphyrin metabolism.210 A sensitive indicator of lead absorption is inhibition of erythrocyte ␦-aminolevulinic acid dehydrase, which correlates with blood lead levels down to values considered “normal” for urban populations.95 The physiologic significance of this is not clear. The substrate of this enzyme, ␦-aminolevulinic acid (ALA), accumulates and starts to appear in the urine in increased quantities at blood lead levels of 40 to 50 ␮g/dL, which is below the level required to produce the subclinical electrophysiologic changes described above. Therefore, ALA should always be increased in urine if lead neuropathy is diagnosed. With further increases in lead level, the excretion

of coproporphyrin in urine and the accumulation of free erythrocyte protoporphyrin occur.210 Rubens and colleagues175 suggested that acute lead neuropathy with abdominal pain is due to secondary porphyria. Effects on renal function are usually seen in cases of chronic, moderately severe intoxication. Other indications of chronic exposure, such as lead lines in the bones (Fig. 113–1), are rarely seen in modern practice. The third part of the diagnosis rests upon demonstrating a history of appropriate exposure. As discussed previously, although the environment is widely contaminated, this does not produce overt neuropathy. The most common mode of exposure is industrial, or from ingestion of old paint by children. Other rare sources of excessive lead exposure have been mentioned. The final criterion for diagnosis is the demonstration of excessive tissue concentration of lead. The first requirement for this is accurate and reproducible methods for estimation. A major problem is that environmental contamination with lead is so great that specimen collection and handling before analysis may spuriously introduce more lead than the original

FIGURE 113–1 Lead lines in the bones. A and B, Radiographs from the upper and lower limbs, respectively, of a child age 2 years 5 months. They were taken at the time of presentation with an episode of acute encephalopathy that had been preceded for 3 months by the child’s eating flakes of lead paint. A marked increase in metaphyseal density is seen in the growing ends of the long bones. C, Radiograph taken 10 months later, after cessation of exposure. The previous metaphyseal line is now in the shaft of the bone and is less distinct.

Metal Neuropathy

sample contained. For clinical purposes, flameless atomic absorption spectrophotometry (AAS) combined with the collection and laboratory precautions described by Nixon et al.152 should provide reliable measurements. Estimates of low-level contamination (as opposed to pathologic concentrations) require higher resolution techniques, such as isotope dilution–mass spectrometric methods.58,186 The tissues most commonly analyzed are urine, whole blood, serum, bone, and teeth. Lead in bone and teeth has been estimated to have a biologic half-life of 20 years, whereas there is a rapidly exchanging pool in blood and soft tissues with a half-life of about 20 days. Experimental studies in humans and animals demonstrate that, in the normal state, about 98% of the lead is bound to erythrocytes. The rest is bound to plasma proteins. The proportion bound to erythrocytes becomes less as the total blood lead concentration increases, suggesting saturable binding to red cells.152 Serum lead has a half-life for excretion of 30 hours compared to 20 days for whole-blood lead.152 Lead excretion in the urine relates directly to rate of intake and to blood lead levels in the steady state. Thus blood and urine levels represent recent exposure to lead over a time of weeks to several months, whereas levels in bone and teeth relate to exposure over years and may be used to estimate the total “body burden” of lead.153 Commonly accepted “normal” values for urine and whole blood are shown in Table 113–1. Provocative tests have been used in patients in whom excessive lead exposure in the past is suspected but not demonstrable with blood and urine estimations. In these, EDTA is injected as a single intravenous dose (usually 1 g) and the urine collected for up to 4 days. Rieders considered an output of more than 1000 ␮g in the first 24 hours to be indicative of harmful exposure.170 In healthy volunteers not occupationally exposed to lead, less than 600 ␮g was excreted after such a dose of calcium EDTA.55 Presumably, similar results would be obtained using oral penicillamine, but this has not been studied systematically. Cerebrospinal fluid (CSF) either is normal or shows a modest increase in protein concentration without pleocytosis.

Pathology Lead may be absorbed from foodstuffs, from drinking water, and from the air. Individuals not exposed to abnormally large amounts of lead obtain most of their lead by way of food contamination. Those working in atmospheres with high automobile exhaust emissions have increased blood levels as a result of inhalation of lead-containing aerosols.87 About 10% of lead in the diet is absorbed from the gastrointestinal tract,114 with the remainder being excreted in the feces. The proportion absorbed depends on a variety of factors, including overall concentration of lead in the diet and the concentration of lead relative to other metals, such as iron and calcium. Absorption of inhaled lead is more

2531

efficient. It occurs directly from the pulmonary tree and by clearing and swallowing bronchial secretions. Absorbed lead is excreted predominantly by way of the kidney. The proportion of dietary intake excreted is variable, but since the whole-body content of lead increases throughout life, it is assumed that output is less than intake.78,210 The mechanism of lead absorption is not fully understood. The observed dietary interaction of lead and calcium suggests that they share a common transport mechanism.16 Lead enters all tissues, but most lead absorbed in the diet enters the skeleton and is deposited in the insoluble crystalline matrix. Whether significant mobilization of the lead occurs in disease states associated with osteoporosis is unknown. In humans it is not known how well blood lead levels reflect the lead content in individual target tissues. In animals it has been shown that lead has a particular affinity for the peripheral nervous system.225 At the level of the individual cell, lead has been demonstrated to have a number of different effects. In 1937, Blackman first described the appearance of spiculated intranuclear inclusion bodies as characteristic of lead-poisoned cells.21 These bodies have been identified in many different tissues, including the Schwann cell149 and the endoneurial capillary endothelial cell. The inclusions are separate from the chromatin217 and appear to contain lead.77,145 The significance of these bodies is unclear, but it has been suggested that they exert a protective effect by sequestering lead in an insoluble, and hence metabolically inactive, lead-protein complex.76 Experimental animal studies suggest that the mitochondria may be the critical site at which lead acts. In a cell culture system, Walton and Buckley found early mitochondrial changes.217 Exposure to lead concentrations of 10⫺7 M induced the accumulation of electron-dense particles in cytoplasmic vesicles and mitochondria. Microincineration suggested that these particles contained lead, and a mechanical interference with normal mitochondrial membrane conformational changes was postulated. Changes in mitochondrial structure have also been noted in renal tubular cells78 and in cultured muscle cells.89 This hypothesis is strengthened by the observation that the radioactive isotope 203Pb was rapidly concentrated in the renal mitochondria of the rat.15 At present, little is known about the subcellular effects of lead on the peripheral nervous system, and there is continued controversy over the pathologic changes produced by lead intoxication. There are two reasons for this. The first is that almost all of our gross and microscopic pathologic knowledge of the human disease was accumulated from autopsy cases around the turn of the century. Many of these patients were known to be suffering from other diseases, such as diabetes, syphilis, and alcoholism. Others may have had diseases that were not fully recognized at the time, such as acute inflammatory demyelinating polyradiculoneuropathy (Guillain-Barré syndrome), multifocal neuropathy with

2532

Diseases of the Peripheral Nervous System

conduction block, or amyotrophic lateral sclerosis (ALS). This last point is illustrated by Wilson’s patients, who had a disease course indistinguishable from that of ALS, but who also had clinical evidence of lead intoxication.220 Wilson speculated on the coincidence and the possible interrelationship of these two entities. Most pathology reports do not have such excellent clinical descriptions, which makes it impossible to reevaluate the cases. The second problem relates to the generalized conclusions drawn from animal experiments. There are clearly real species differences in the susceptibility to lead of different parts of the nervous system at different ages. In adult rats, chronic lead poisoning produces a peripheral neuropathy characterized by almost pure segmental demyelination.123,157,225 In suckling rats, an encephalomyelopathy is produced.162 In guinea pigs, there is a mixed picture of axonal degeneration and segmental demyelination,66 whereas in baboons no neuropathy could be demonstrated in spite of high blood lead levels for periods up to 1 year.99 In tissue culture, lead clearly interferes with myelination more extensively and at lower concentrations than produce axonal or neuronal changes.222 In rat models, lead accumulates in the endoneurial compartment of the peripheral nerve (Figs. 113–2 and 113–3)225 and appears to cause Schwann cell hypertrophy,144 and is preferentially bound to myelin (Fig. 113–4).224 For a systemic toxin with widespread cellular effects, it is not surprising that, even though all cells of the nervous system might be affected by lead to a greater or lesser extent, different clinical patterns result from the different

FIGURE 113–2 The accumulation of lead in the epi-, peri-, and endoneurium of rat sciatic nerves during the process of intoxication with 4% lead carbonate. Each point represents the mean ⫾ 1 SE of estimations on eight nerves. With the scale used, corresponding values for control animals would form a straight line parallel and very close to the x axis. (From Windebank, A. J., McCall, J. T., Hunder, H. G., and Dyck, P. J.: The endoneurial content of lead related to the onset and severity of segmental demyelination. J. Neuropathol. Exp. Neurol. 39:692, 1980, with permission.)

susceptibilities of populations of cells in different species and at different ages. In part this different susceptibility may depend on the relative integrity of the blood-brain barrier and blood-nerve barrier to lead. This has been used to explain why young animals and children tend to develop encephalopathy, whereas adults are more likely to suffer from a neuropathy. There may also be a dose dependence of effects. The only systematic modern pathologic study of low-level lead intoxication27 demonstrated a statistically significant increase in the incidence of paranodal demyelination and internodal remyelination. These changes are identical to the early change found in chronically intoxicated rats.225 However, the rats go on to develop an almost pure segmentally demyelinating neuropathy with no signs of axonal involvement,53,157,225 whereas all the available evidence suggests that, in the advanced neuropathy in humans, axonal degeneration is the primary feature. Extensive reviews of the early literature have been undertaken by others.74,102 It is not clear whether the axonal degeneration seen is due to a primary effect on the anterior horn cell or to an effect on the distal process of the motor neurons. There is still no adequate explanation or hypothesis for the preferential involvement of the motor neurons of the upper limb, particularly those of the seventh cervical segment, which produce the typical wristdrop. The possibility that the motor neuropathy is induced by secondary porphyria has been raised.175 The question as to whether low-level lead intoxication plays a role in the development of ALS is controversial. The coincidence of progressive motor neuron disease and clinical signs of lead intoxication was first remarked on by Aran, who described a form of progressive muscular atrophy that we now recognize as ALS.7 Three of Aran’s 11 patients had suffered excessive occupational exposure to inorganic lead salts, but the pyramidal signs distinguished the clinical picture in these patients from that seen in typical saturnine (lead-induced) palsy. Later Wilson described four cases of ALS of toxic origin.220 He stated, “We do not assert that a prima facie case has been made out for the toxic origin of amyotrophic lateral sclerosis, but we cannot avoid the conclusion that the facts suggest the probability of such an origin.” However, Wilson also distinguished this disorder from the typical lead-induced distal neuropathy associated with wristdrop because of the presence of pyramidal signs. With the high incidence of lead poisoning in the 19th and early 20th centuries, both Aran and Wilson hesitated to conclude that lead poisoning was a common etiologic factor in motor neuron disease rather than a coincidence. Over the years this coincidence has been remarked on by a number of authors.28,29,60,130 They usually described this association as chronic lead intoxication mimicking motor neuron disease, thus implying that, in general, the two are separate. It is only in the last 25 years that the relationship has been tested by quantitative estimation of the lead content of tissues in patients with ALS. The first

Metal Neuropathy

FIGURE 113–3 Endoneurial and connective tissue labeling of the peroneal nerve 72 hours after a single intravenous dose of 100 ␮Ci 210Pb. A–C, Normal animal. D–F, Animal after 70 days of lead intoxication. These micrographs demonstrate the striking difference in endoneurial labeling of the chronically intoxicated animal. Top, Low-power phase contrast micrographs. Middle, Sections from the same nerves viewed by darkfield microscopy. This highlights the tracks produced in the autoradiographic emulsion by individual disintegration of atoms of 210Pb. Bottom, The same sections viewed under high power. At this magnification, individual tracks may readily be counted. (From Windebank, A. J., and Dyck, P. J.: Kinetics of 210Pb entry into the endoneurium. Brain Res. 225:67, 1981, with permission.)

2533

2534

Diseases of the Peripheral Nervous System

FIGURE 113–4 Labeling of subcellular compartments the rat peroneal nerve 72 hours after a single intravenous injection of 100 ␮Ci of 210Pb. These ratios were derived by comparing predicted random distribution against the observed autoradiographic superimposition of silver grains over cellular structures. (From Windebank, A. J., and Dyck, P. J.: Localization of lead in rat peripheral nerve by electron microscopy. Ann. Neurol. 18:197, 1985, with permission.)

such attempt was made by Currier and Haerer, who studied a group of 37 patients with the clinical features of ALS.46 Twenty-four patients in this group were thought, by history, to have an excessive occupational exposure to lead. However, there was no abnormality in urine lead excretion and no clinical response to chelation therapy. Two of the patients came to autopsy, and lead estimations were carried out on spinal cord tissue. Neither the method of estimation nor the numerical values for the results were reported, but the findings for both patients were said to be normal. The authors concluded that, if there is a toxic effect of lead, it must be an early or permissive effect, leading to degeneration that progresses in the absence of the toxin. Subsequent trials of chelation therapy in patients with ALS and a history of lead exposure have shown no treatment effect.44 Campbell and associates studied 26 patients retrospectively and 74 prospectively.30 All had clinically definite motor neuron disease. Seventy-four age- and sex-matched control patients were also investigated. Using a standard questionnaire, these investigators determined that 11 patients versus only 4 controls had a significant history of exposure to lead. They used two measures in attempting to quantitate this exposure: the blood lead level and the lead content of bone from an iliac crest bone biopsy, the latter being used as a measure of long-term exposure to lead. No difference was found in the results from patients with motor neuron disease who had significant exposure to lead. A careful series of experiments was carried out by Conradi and colleagues.41,42 Using flameless AAS, they investigated the CSF and plasma levels of lead in patients

with ALS and in neurologic controls. They found a small but statistically significant increase in lead levels in both of these body fluids from the ALS patients. A later study of CSF in 16 ALS patients, 22 neurologic disease controls, and 24 controls found no difference in lead levels.109 Louwerse and colleagues investigated the relationship of lead exposure and ALS by chelating patients with ALS, patients with spinal muscular atrophy, and controls with dimercaptosuccinic acid (DMSA; 20 mg/kg). They found no difference in lead or mercury excretion between the groups.131 Several epidemiologic studies of ALS and meta-analyses have shown a small but significant increase in history of lead exposure.8,106,172,212 Other studies have failed to identify an association.83

Treatment As with all intoxications, identification of and removal from the source of exposure is the first and most important step. The benefit of attempting to increase excretion of lead by the use of chelating agents is controversial. Boyd and colleagues, reporting on the treatment of tetraethyl lead poisoning,23 mentioned that intravenous EDTA was more effective than oral penicillamine in promoting the urinary excretion of lead. Boulding and Baker treated two lead-poisoned patients, one of them with a neuropathy, with penicillamine and observed a satisfactory remission of symptoms as well as a marked increase in the urinary excretion of lead.22 Harris compared the use of EDTA intravenously and penicillamine orally and found that both agents produced a satisfactory increase in urinary lead excretion.91 He concluded that, although EDTA produced a slightly greater excretion of lead than penicillamine, the latter was preferred because of the oral administration and lesser toxicity. Ohlsson also compared oral penicillamine and intravenous EDTA and concluded that both agents produce about the same increase in the urinary excretion of lead155; however, he thought that EDTA had a greater mobilizing effect on lead stored in the body because of the increase in blood lead concentration that was found after each intravenous administration. Goldberg and colleagues treated lead-exposed adults with penicillamine orally and found a satisfactory increase in the urinary lead excretion, with concurrent decreases in urinary coproporphyrin and ALA levels.73 One patient developed proteinuria after 4 months of treatment with penicillamine, so these authors advised periodic testing for proteinuria in patients being so treated. They also advised that an individual course of treatment with penicillamine should not be longer than 4 weeks. Wyllie and colleagues used a provocative dose of penicillamine (900 mg) on a single day and noted an increase in urinary lead excretion and a decrease in total urinary porphyrin excretion in subjects with industrial exposure to lead but with no signs of lead intoxication.228

2535

Metal Neuropathy

Selander and associates found that the symptoms of lead poisoning disappeared within 1 week of the start of oral treatment with penicillamine.183 Despite return of blood lead concentrations and urinary ALA excretion to normal, there was a significant urinary excretion of lead with continued treatment with penicillamine. Selander also compared the effect of calcium disodium EDTA with that of penicillamine, both intravenously and orally.182 The increase in urinary excretion of lead was greatest with intravenous EDTA, followed in descending order by intravenous penicillamine, oral penicillamine, and oral EDTA. Selander concluded that oral penicillamine therapy should be satisfactory for most patients with lead poisoning. If intravenous therapy was needed in severe cases, he preferred penicillamine to EDTA because of the greater toxicity of the latter. Chisolm noted that, in acute lead intoxication in children, treatment with EDTA alone caused transient exacerbation of the illness38; he favored combined treatment with British antilewisite (BAL) and EDTA rather than with EDTA alone. He chose to use BAL instead of penicillamine because BAL was thought to remove lead more completely and more rapidly than penicillamine. Chisolm concluded, however, that oral treatment with penicillamine was satisfactory for treatment of chronic lead poisoning, that the blood lead concentration could be kept within normal limits, and the urinary output of ALA could also be maintained within normal limits with penicillamine. Freeman reported that, for the previous 18 months, all new cases of lead intoxication in Sydney, Australia, had been treated satisfactorily with long-term use of penicillamine.63 In contrast, St. George chose to use a combination of penicillamine, BAL, and EDTA in two men with lead poisoning and recommended continuation of this triple therapy until the urinary excretion of lead was less than 300 ␮g/24 h.197 Sachs and colleagues treated 1155 children who had blood levels of lead of more than 50 ␮g/dL.179 If the blood level of lead was higher than 70 ␮g/dL, they used intramuscular injection of EDTA, but if it was less they used penicillamine given orally. When treatment with EDTA decreased the blood level to 70 ␮g/dL or less, they changed the treatment to penicillamine. More recently, chelation therapy with oral DMSA has been advocated as safer and more effective for the treatment of lead poisoning.75 As can be seen from the diversity of these reports, there is a lack of agreement with regard to the optimal mode of treatment of lead intoxication. This is also reflected in review articles on treatment of lead poisoning.39,85,90

ARSENIC Arsenic has an atomic number of 33 and mass of 74.9. The therapeutic use of arsenic compounds was advocated for many diseases from the time of Hippocrates to the middle

of the present century. Over the years it was recommended as specific treatment of arthritis, asthma, chorea, diabetes, malaria, many nutritional disorders, skin diseases, stomach ulcers, and tuberculosis.192 The therapeutic use was expanded enormously following Ehrlich’s description in 1905 of the specific action of organic arsenicals against trypanosomes. By 1937, the properties of over 8000 arsenic compounds were described.188 Use in the treatment of syphilis resulted in nervous system complications of optic neuritis, encephalopathy, and peripheral neuropathy. Old “home remedies” such as Fowler’s solution (potassium arsenite) and Donovan’s solution (arsenic iodide) may occasionally be encountered. A number of traditional medicines, particularly those from the Indian subcontinent, may contain high levels of arsenic and lead pigments. There has been renewed interest in arsenic trioxide as a potential therapeutic agent for the treatment of myeloma or other malignancies.9,47,142 Occupational or industrial exposure to arsenic compounds may occur in copper and lead smelting process workers. The major commercial use in the past has been as a pesticide in the form of lead arsenate, calcium arsenate, and copper acetoarsenate (Paris green). Although agricultural use in developed countries has declined in recent years, some farmers still carry stocks of the chemicals dating back to their widespread use in the elimination of grasshoppers, which were particularly prevalent in the midwestern United States in the 1950s. Other uses of arsenic have included wood preservatives, glass manufacture, cattle and sheep dips, leather manufacture, paper dyes, and chemical warfare gases such as lewisite.211 Accidental contamination of foodstuffs and beverages has led to occasional epidemics of arsenic poisoning. One of the largest and most well documented was that occurring in Lancashire and Staffordshire, England, in 1900.115 In this case sulfuric acid was manufactured from pyrites containing arsenic impurities. This acid contained 1.4% arsenous acid and was used in the preparation of sugar, which in turn was used for brewing beer; the beer produced ultimately contained 15 ppm arsenic. It was this epidemic, in which 6000 persons became ill and 70 died, that first allowed the precise documentation of the clinical features of arsenic intoxication. With greater quality control and the replacement of arsenic compounds in agriculture and medicine by more specific agents, the incidence of accidental arsenic poisoning has declined. Most cases recognized recently stem from homicide or suicide attempts.74,124

Clinical Aspects As with other metal intoxications, the neuropathy associated with arsenic ingestion presents in the context of a systemic illness. The most typical history is that of an acute gastrointestinal illness, followed several days later by the onset of

2536

Diseases of the Peripheral Nervous System

burning, painful paresthesias in the hands and feet, with progressive distal muscle weakness. Recent case reports have suggested that it is the most commonly recognized heavy metal neuropathy.36,74,96,124 An important clinical point made by LeQuesne and McLeod is that the neuropathic symptoms appear between 10 days and 3 weeks and may progress for up to 5 weeks following a single dose of arsenic.124 The initial symptoms are usually gastrointestinal and consist of abdominal pain, nausea, vomiting, and diarrhea. These occur minutes to hours after ingestion. In severe poisoning, urine output may fall secondary to systemic vascular collapse. Central nervous system symptoms develop rapidly in the context of acute poisoning, particularly in the case of organic arsenicals. This may involve drowsiness, confusion, stupor, organic psychosis, and delirium. With large doses, signs may progress rapidly to death within 24 hours of ingestion. If the patient survives this phase, more chronic signs of intoxication appear. It is important to note that chronic signs may appear after a single exposure, suggesting significant tissue binding or body storage of the toxin. Skin changes are commonly seen. Hyperkeratosis and sloughing of the skin on the palms and soles may occur several weeks after ingestion and be followed by a more chronic state of redness and swelling of the hands and feet. The skin on the rest of the body may develop areas of increased pigmentation and depigmentation, which may form “raindrop” or “teardrop” shapes, particularly on the trunk. Nail changes were described independently by a number of different authors3,140,164 and bear the eponym of Mees’ lines. The pale transverse bands appear parallel to the lunula in all fingernails and toenails 4 to 6 weeks after ingestion.104 They grow out with the nail, and multiple bands may be correlated with repeated episodes of acute intoxication. Mees’ lines are not pathognomonic for arsenic. They are seen in thallium poisoning and a variety of other medical conditions.104 Typically, hematologic changes occur with chronic poisoning. These usually have the characteristics of an aplastic anemia with pancytopenia.122 Basophilic stippling may be seen. The neuropathy is fairly uniform in its clinical characteristics. It begins 5 to 10 days after ingestion of a single dose of arsenic124 and progresses for several weeks after ingestion. In chronic or repeated intoxication, progression will occur over a longer interval. The first symptoms are usually aching or burning pain and tingling or numbness beginning in the fingers and toes and spreading proximally over several days. In severe cases there may be sensory symptoms in the distribution of the trigeminal nerve. Weakness occurs first in distal limb muscles and spreads proximally. In mild cases, a distal glove-and-stocking sensory loss affecting all modalities may be the only finding. In severe cases, there may be severe weakness or paralysis of limb and trunk muscles requiring ventilation. Tendon

reflexes are reduced proportionate to the degree of sensory loss. Recovery occurs slowly over a matter of months to years. It is complete in milder cases, but permanent, residual motor and sensory signs may follow severe peripheral nervous system involvement.

Laboratory Studies The major step in the diagnosis of arsenic neuropathy is consideration of the possibility. Detection of arsenic in tissue is straightforward. Arsenic is not in widespread industrial use, and the problem of sample contamination is not as great as with lead. Similarly, the population as a whole is exposed to little arsenic, and therefore the question of low-level toxicity only arises in the small group of industrial workers working with arsenic. This rarely raises any clinical difficulty. Simple qualitative screening tests have been available for many years. These include the generation of arsine gas (Marsh’s test) or reaction with silver nitrate (Gutzeit method). The latter may be quantitated.211 In more recent years, estimation of microgram quantities has become routine using flameless AAS and mass spectrometry. Use of the appropriate tissue is important in determining arsenic exposure. The mean blood level in normal subjects is of the order of 0.004 ␮g/g.26 Clearance from the blood is rapid and occurs over hours, with 70% to 80% excreted in the urine and the remainder distributed throughout body tissues.133,139 Thus blood arsenic levels are not usually of diagnostic importance. Mealey and colleagues demonstrated that a single intravenous injection of radiolabeled arsenite is excreted in the urine in three different phases with half-lives of 2 hours, 8 hours, and 8 days.139 Thus urinary excretion may provide an index of continuing low-level arsenic exposure or provide a clue to a single large dose of arsenic in the past. Urine levels may be raised several months after such a dose. Administration of penicillamine may increase the tissue mobilization and hence the urinary excretion of arsenic. Urinary concentration in persons without exposure has been demonstrated as 0.01 to 0.03 ␮g/L.24 Pinto and colleagues found an average urinary arsenic concentration of 0.053 ␮g/L among nonexposed workers and 0.174 ␮g/L among exposed workers at the same factory.163 Diet is an important factor in determining the source of slightly raised urine arsenic values, because seafood contains about 10 times the concentration of arsenic as other foods. In practice, our laboratory has used a value of less than 25 ␮g/24 h as a normal level of arsenic excretion. Some laboratories have the ability to differentiate between inorganic “toxic” arsenic and the organic derivatives from seafood that are not toxic.151 An alternative approach in the case of urine levels between 25 and 100 ␮g/24 h (which may be seen with seafood) is to ask patients to abstain from any fish or seafood for 3 days and repeat the urine estimation.

Metal Neuropathy

Arsenic is bound to the keratin of growing skin, hair, and nails, and because of the slow turnover of these tissues, it may be detected months or years later. In a large population study, the median concentration of arsenic in the hair was 0.51 ␮g/g, with 80% of individuals having a level below 1 ␮g/g and none of 1250 individuals exceeding levels of 10 ␮g/g.128,190 Similar values probably apply to the level in nail parings. Measurement of increased arsenic levels in hair and nails may be diagnostically important. An occasional patient with neuropathy may be found to have a modest elevation of 24-hour urine arsenic excretion (25 to 200 ␮g/24 h). While this may be seen in an individual following seafood ingestion, it may also represent the “tail end” of urine excretion of a massive dose, reflecting a past acute intoxication. AAS measures elemental arsenic and does not distinguish between different oxidative states. The (⫹5) oxidation state (arsenate) is the predominant form in seafood, is relatively less toxic and does not bind to hair and nails (see below). In contrast, the (⫹3) oxidative state (arsenite) is more toxic and binds avidly to sulfhydryl groups in the keratin of hair and nails. A moderate CSF protein elevation (⬍100 mg/dL) without pleocytosis has been consistently reported in arsenic intoxication. Neurophysiologic findings are those of an axonal sensorimotor neuropathy. Nerve conduction studies show absent or reduced amplitude of sensory potentials and reduction in amplitude of CMAP with relative preservation of MNCV. Electromyographic examination typically shows increased insertion activity and fibrillation potentials. Motor unit potentials may be reduced in number in the acute phase. In the chronic phase, increasing numbers of polyphasic units with long duration are seen. A single report has demonstrated that, early in the course of arsenic neuropathy, electrophysiologic changes may suggest proximal demyelination.50 This may further mislead the clinician into suspecting a diagnosis of Guillain-Barré syndrome. Chronic changes are consistently compatible with axonal degeneration.

Pathology Biochemical Mechanism Arsenic is in Group Vb of the periodic table. Arsenates (⫹5 oxidation state) resemble phosphates but are generally less stable. Arsenic in this state is relatively less toxic. This is the form usually found in shellfish and, because it does not form bonds with thiol groups, it is not incorporated into hair or nails. In isolated cell and enzyme systems, arsenic is a potent uncoupler of oxidative phosphorylation. Arsenic also interferes with several steps in glycolysis. Arsenate (5⫹) substitutes for inorganic phosphate and forms arsenic analogues of high-energy phosphates that are unstable and break down, regenerating inorganic arsenate.101,211 Arsenite compounds (⫹3 oxidation state) react directly with the sulfhydryl groups of proteins. Arsenite (3⫹) may react with a single sulfhydryl group, producing an arsenic-thiol

2537

mercaptide, or with two sulfhydryls, forming a stable ring. Such ring formation is thought to be important in the arsenic-mediated inhibition of various enzyme systems, including pyruvate dehydrogenase. Arsenic induces DNA damage and oxidative stress that may also be involved in cellular toxicity.17,92,118 Cellular Pathology Axonal degeneration is the major feature of altered peripheral nerve morphology in human arsenic intoxication. There are no modern autopsy reports of changes in the nervous system related to arsenic poisoning. The best older report is that of Erlicki and Rybalkin, who described the changes in a patient who died from pneumonia during recovery from arsenic poisoning.57 The clinical features were typical of a moderately severe arsenic neuropathy. At autopsy in this case, pathologic changes were found in the spinal cord and peripheral nerves. Anterior horn cells were decreased in number both in particular groups of the ventral gray matter and diffusely. The remaining cells had lost angular contours, were rounded, and often lacked processes. Cytoplasm contained granular yellow or yellow-brown pigment. This may reflect the presence of lipofuscin, an abnormality occurring in spinal cords unaffected by specific diseases. Nuclei were ragged and sometimes pale. In the cervical and lumbar enlargements of the spinal cord, there was no marked abnormality except that posterior and anterior columns were markedly thinned and had decreased numbers of myelinated fibers. The radial and peroneal nerves were evaluated. The majority of myelinated fibers were abnormal, but the morphologic changes were not described in detail. In the authors’ opinion, some fibers had remained without degeneration, some were degenerating, and others were regenerating. There are a number of reports concerning biopsies in patients with arsenic neuropathy36,52,74,96,111,124,158 in which histologic, morphometric, electron microscopic, and teased fiber changes are described in the sural nerve or a branch of the superficial peroneal nerve. The changes reported have been consistent with axonal degeneration. Chhuttani and colleagues36 and Heyman and associates96 reported increased endoneurial cellularity and interstitial fibrous tissue with some increase in the thickness of the perineurium. All authors reported a decrease in the number of myelinated fibers, and Dyck and colleagues demonstrated that this occurs equally across the range of all fiber diameters.52 Several authors have suggested that, in the case of a single dose of arsenic, all fibers are at the same stage of degeneration into linear rows of myelin ovoids. However, LeQuesne and McLeod have shown in two cases, each with a documented single exposure, that fibers at all stages of axonal degeneration may be seen in the biopsy specimen.124 Segmental demyelination is either absent or present at a very low frequency.72 Occasional onion bulbs are seen, which traditionally have been thought to indicate a chronic

2538

Diseases of the Peripheral Nervous System

process involving repeated episodes of demyelination and remyelination. Ohta has pointed out that, at the electron microscope level, the core of these onion bulbs appears to be a degenerated axon and myelin debris, suggesting that they may represent a Schwann cell response to axonal degeneration.158 This study also mentioned that unmyelinated fibers showed little change. Quantitative studies of unmyelinated fibers have not been performed. In tissue culture, arsenic appears to specifically affect axonal structure rather than producing demyelination.222

disadvantage is that, because of low lipid solubility, they may not enter the nervous system. There are no reports that help to understand whether these newer agents are more effective in preventing neuropathy in patients with arsenic poisoning. Because the recovery phase from severe arsenic neuropathy may be very slow, appropriate physical therapy and rehabilitative measures are essential. The prognosis for recovery is related, in general, to the severity and duration of symptoms and to the success with which the source of exposure is removed.

Treatment As with all intoxications, the first step in treatment is identification of and removal from the source of exposure. However, in the case of arsenic there is ample evidence, as discussed earlier, that toxic symptoms may progress for weeks after a single episode of poisoning. It would therefore seem reasonable to try to reverse these effects even in the absence of continued exposure. Two lines of treatment might be predicted on a theoretical basis: interfering with the binding of arsenic to thiol groups or increasing the rate of removal of arsenic from target tissue and from the body as a whole. The first of these options formed the basis for the development of BAL. It was found that monothiols would reverse the action of arsenic on certain enzymes, but not on the critical cerebral pyruvate dehydrogenase system. To reverse this, a dithiol was required, which would form a more stable ring than the arsenate-lipoate ring.198,199 BAL also increases the rate of urinary excretion of arsenic.96 However, there are only two studies in which comparison was made between patients with arsenic neuropathy who were treated with BAL and those who received nonspecific treatments. In the first, the study was descriptive and retrospective and does not have the implications of a randomized, controlled study with quantifiable end points. However, the 22 patients who received BAL and the 13 who did not were thought to be comparable in terms of the duration and severity of the neuropathy and period of follow-up. BAL did not appear to benefit the treated group with regard to either rate or extent of recovery. In the retrospective report by Chhuttani and colleagues, a similar comparison was made between 20 treated and 20 untreated patients.36 Although some of the patients treated with BAL appeared to recover more rapidly, the groups were significantly different in terms of duration of neuropathy. From these two studies, it is clear that BAL does not have a dramatic effect on the prognosis for recovery from neuropathy. The drug, which must be given by intramuscular injection and has many side effects, is not indicated for the treatment of established arsenic neuropathy. More recently, the use of two other dithiols, meso-2,3-DMSA and 2,3-dimercaptopropanesulfonate sodium, has been advocated. Both are water soluble and have less toxicity than BAL. They may be administered orally or intravenously.86,88,146 A potential

MERCURY Mercury has an atomic number of 80 and atomic mass of 200.6. The effects of mercury on the nervous system, more than those of any other metal, depend upon the chemical state of the metal: Elemental mercury, inorganic mercury salts, short-chain alkyl mercury compounds, and complex organic mercurials all behave differently. The complex organic mercurials have been used as diuretics for many years, and they are still available and occasionally used in some countries. There are no clear-cut cases of peripheral nervous system disease attributed to their use. Mercuric chloride is an inorganic compound that is converted into methylmercury by microorganisms. This compound then enters the aquatic food chain, as in the Minimata Bay outbreak of mercury poisoning.121 Mercury is concentrated in the bodies of fish and shellfish, which served as a staple item in the diet of the local population. The observation that cats and seabirds also feeding on the fish developed a neurologic disease led to the positive identification of methylmercury as the toxin. A second major outbreak arose in an Iraqi population after grain treated with ethylmercury paratoluene sulfonanilide as a fungicide was used in making bread.10,108 The grain had been intended for sowing, and the fungicide-treated grain was identified by a dye. However, the dye was easily washed off and it was erroneously believed that washing also removed the fungicide; this was not the case. Separation of the effects of elemental mercury from those of mercuric salts is difficult because most cases arise from occupational exposure to both forms of mercury.20 It is probable that inorganic mercury, because of its greater water solubility, is best absorbed through the gut and produces prominent systemic as well as nervous system effects. Elemental mercury, because of its greater lipid solubility, may produce nervous system disease with almost no systemic disease.

Clinical Aspects Because the various forms of mercury may produce different clinical pictures, they are considered separately.

2539

Metal Neuropathy

Organic Mercurials The features discussed here refer almost exclusively to methylmercury intoxication. Ethylmercury probably produces similar effects, but it has not been widely studied. During chronic exposure to methylmercury, prominent central nervous system disease develops, with constriction of visual fields, ataxia, dysarthria, decreased hearing, tremor, and mental impairment. However, one of the earliest complaints in many patients is distal paresthesias,203 which progress proximally and may involve the tongue. This is probably related to selective toxic effects on the neurons of the dorsal root and trigeminal sensory ganglia. Inorganic Mercury Poisoning Mercury salts are used in a number of industrial processes. Historically their use in the production of felt hats was responsible for the popular expression “mad as a hatter.” The association of mental impairment and tremor was characteristic.150 The major route of absorption of mercury salts is by way of the gastrointestinal tract. Gastrointestinal symptoms and a nephrotic syndrome may be prominent systemic features, especially in acute poisoning. In 1981, Svare et al.201 published a brief report that described an increase in mercury vapor released into the breath by chewing. The authors raised the possibility that biotransformation might occur and that low levels of ingested or inhaled mercury might be converted into more toxic methylmercury. In spite of the lack of direct evidence for this biotransformation, the role of dental amalgam in the production of neurologic disease has repeatedly arisen. No published studies have ever been able to link fillings to neurologic disease, and no trial has ever demonstrated a benefit from removing fillings. A variety of practitioners have reported anecdotally that patients with a variety of disorders respond to filling removal. Cases have included peripheral neuropathy, motor neuron diseases, dementing disorders, and chronic fatigue syndrome. None of these reports has appeared in the cited medical literature. For these reasons, it must be concluded that, at present, there is no evidence to link amalgam fillings to specific neurologic disease. Elemental mercury is absorbed poorly from the gastrointestinal tract, so that the swallowing of mercury (e.g., from broken oral thermometers) generally is not harmful. Metallic mercury is, however, volatile at room temperature. A saturated atmosphere of mercury vapor contains 18 mg Hg/m3 at 24 °C. Friberg and Nordberg demonstrated that chronic exposure to levels above 0.1 mg Hg/m3 in the air are associated with the development of toxicity.64 This is of particular significance because small quantities of metallic mercury may remain hidden for long periods of time in poorly ventilated areas. We have observed occasional domestic exposures in which children played with mercury droplets from broken thermometers, and unusual exposures, such as mercury electrical junctions used in model train sets. Because of its greater lipid solubility, this toxicity is often manifested first by

central nervous system symptoms of fatigue associated with weight loss, anorexia, and disturbance of gastrointestinal function, which is sometimes referred to as micromercurialism. With more severe exposure, the syndrome of mercurial erethism develops. This is characterized by personality and behavioral change, insomnia, and hyperexcitability, often associated with a coarse facial, head, and limb tremor. A few cases of predominantly peripheral nervous system involvement associated with exposure to organic and elemental mercury have been recorded.174,202 A Guillain-Barré–like neuropathy has been described in children who had coincident high exposure to mercury vapor.208,223 All were children who were playing with liquid mercury. Two of the cases were siblings, suggesting that mercury exposure was the cause of the neuropathy. Each case developed rapidly progressive severe weakness with minimal sensory involvement. It is not clear whether this is an idiosyncratic immune-mediated response or a direct result of mercury toxicity to the peripheral nervous system. In all cases, removal from exposure was associated with recovery.

Laboratory Studies The diagnosis of mercury intoxication depends upon the demonstration of an appropriate neurologic syndrome in the context of exposure to mercury or mercury compounds. In the case of methylmercury intoxication, circumstantial evidence may be the only form available. This compound has such a high affinity for the nervous system that neurologic disease may be present in the absence of elevated urine mercury levels. In the Japanese outbreak, methylmercury was incriminated by demonstrating high levels in the sediment and fish of Minamata Bay. It was also observed that laboratory cats fed fish from this bay developed a neurologic disease identical to that occurring simultaneously in the human inhabitants and cats of the area. There was a dose-response relationship between the amount of fish eaten and the degree of neurologic disease.121 Inorganic or metallic mercury intoxication can usually be confirmed by estimation of the 24-hour urinary mercury excretion using flameless AAS. Only limited data are available, but it appears that mercury is excreted very slowly from the body. One estimate is that the biologic half-life for mercury excretion is on the order of 60 days.20 In our trace metal laboratory, the mean level for nonexposed individuals is less than 5 ␮g/24 h. Occupationally exposed individuals may excrete 25 to 30 ␮g/24 h; levels above 50 ␮g/24 h suggest toxicity related to mercury exposure.

Pathology Following the outbreak of methylmercury intoxication in Minamata Bay, there were extensive pathology reports. These primarily involved the central nervous system, in which prominent atrophy and degeneration of the cerebellar

2540

Diseases of the Peripheral Nervous System

and calcarine cortex were reported.203 In the peripheral nerves of humans, axonal degeneration in the sural nerve and lumbar dorsal roots has been demonstrated.204 In experimental animal studies, the major pathologic change appears to be axonal degeneration without segmental demyelination. This has been shown in the L6 dorsal root of the rat but not in the ventral root. The rate of axonal degeneration in the sural nerve is the same at mid-thigh and ankle levels, and therefore the dorsal root ganglion cell is thought to be the primary site of action in the peripheral nervous system.103,143,156 Somjen and colleagues have demonstrated in rats that methylmercury administration resulted in reduction of the compound action potential evoked in dorsal roots, whereas the ventral root action potential was virtually normal.193,194 This physiologic action was correlated with the finding that the major site of uptake of radioactive 203Hg methylmercury in the peripheral nervous system was the spinal dorsal root ganglia. Thus, in both humans and animals, it appears that the major pathologic effect of methylmercury is on the dorsal root ganglion cells. No similar data are available for inorganic or metallic mercury poisoning. Because of the different chemical properties and different clinical manifestations, it is likely that these latter compounds exert their action at a different site. At the cellular level, it is thought that mercury exerts its effects by combining with enzymatically or structurally important sulfhydryl groups. In tissue culture experiments with dorsal root ganglion neurons, mercury has been demonstrated to be directly toxic to the neuron at very low concentrations.222 More recently, microarray studies have suggested that oxidative stress may induce apoptosis in dorsal root ganglion neurons in vitro.219

the cholinergic sympathetic innervation of the sweat glands. During use for this indication, the depilatory effect of thallium was noted. Hair loss resulting from thallium intoxication is usually most marked in the head hair, with lesser effect on the slower growing pubic hair and eyebrows. Thallium salts became widely used in pediatric practice for the treatment of ringworm. The very high incidence of toxic side effects, including systemic effects, from topical application was pointed out in a review of the literature by Munch in 1934.147 He noted a 5% mortality rate and an extremely high morbidity rate in 692 cases following clinical use of thallium. Since that time these salts have disappeared from the pharmacopeia; however, they are still in use for rodent and insect pest control, and reports of accidental or homicidal poisoning occur regularly.13,33,82,117,120,148,161,178,180 A list of current or previously marketed insecticides and rodenticides containing thallium can be found in articles by Grossman82 and Bank.12 Thallium intoxication almost uniformly produces symptoms and signs related to involvement of the peripheral nervous system. As with other metals, these occur in the context of an acute or subacute systemic illness. Alopecia, the hallmark of thallium poisoning, occurs several weeks after exposure.178 It is likely, therefore, that many cases are overlooked because patients die before the loss of hair occurs. This was well illustrated by the cases reported by Cavanagh and colleagues, in whom three persons were poisoned by a single individual.33 The first two died before hair loss occurred and an antemortem clinical diagnosis was not made. In the third and milder case, the patient survived and developed typical alopecia 19 days after the onset of limb paresthesias. A similar cluster of cases was reported by Rusyniak and colleagues in 2002.178

Clinical Aspects Treatment The major part of treatment involves identifying the source of exposure and taking appropriate measures to control this. Chelating agents, such as BAL and DMSA, increase the urinary excretion of mercury, but it is not known whether this increases the mobilization from nervous system tissue. Adequate clinical evidence is not available to substantiate that chelation increases the rate of recovery from the neurologic effects of mercury intoxication. However, isolated case reports suggest that this might be the case.1,132,138,202,208

THALLIUM Thallium is in group IIIa of the periodic table; it has an atomic number of 81 and atomic weight of 204.37. Discovered in 1861, its inorganic salts were used initially to treat the night sweats of tuberculosis. This use may have had a sound basis, because one toxic effect is probably damage to

Because the industrial use of thallium compounds is very limited,112 virtually all clinical cases are due to acute accidental or homicidal intoxication. The symptoms and signs in all of the case reports are rather similar. Gastrointestinal symptoms develop within hours. Nausea, vomiting, epigastric pain, and diarrhea are the most frequent. With massive doses (greater than 2 g), this progresses rapidly to cardiovascular shock, coma, and death within 24 hours. With moderate doses (0.1 to 2 g), peripheral nervous system symptoms develop within 24 to 48 hours. They initially consist of limb and joint pain accompanied by distal paresthesias. Depending on dose, this may progress rapidly over days, with increasing distal-to-proximal limb sensory loss accompanied by distal limb weakness and, within several weeks, muscle atrophy. Relative preservation of proximal limb tendon reflexes may occur. In severe cases, the weakness and sensory loss may involve the cranial nerves and muscles of respiration. Ptosis is common. It appears that, as with other metals, the neuropathy may progress for several weeks after ingestion of a single dose of thallium.

2541

Metal Neuropathy

Central nervous system manifestations occur with increasing frequency as the quantity ingested increases. The most common is anxiety, followed by psychosis or behavioral changes, convulsions, and coma. Individual case reports have mentioned choreoathetosis, ataxia, tremor, and abnormal chewing movements.33,82 Optic neuropathy has been reported, particularly in children.167 Autonomic involvement was commented on by various authors,147 with tachycardia, hypertension, fever, and loss of sweating being the most common signs. Manifestations in other systems include nephropathy with proteinuria, anemia without basophilic stippling, and abnormal liver function. The hallmark of thallium intoxication is alopecia. It is the sign that usually leads to suspicion of thallium poisoning. The scalp hair is preferentially affected, but it is commonly noted that pubic hair and eyebrows, while not falling spontaneously, may be plucked easily and painlessly. A brownish pigment may be deposited in the roots of affected hair.147,178 Typical Mees’ lines, similar to those of arsenic intoxication, may be seen in the fingernails and toenails.161 A comparison of these clinical features with those of arsenic poisoning reveals many similarities, although the alopecia is a late but constant distinguishing feature. The degree of limb pain and preservation of proximal limb reflexes help to distinguish thallium poisoning clinically from acute inflammatory demyelinating polyradiculoneuropathy (Guillain-Barré syndrome). The association of acute abdominal pain and neuropathy with tachycardia, hypertension, and mental changes necessitates exclusion of porphyria by laboratory studies.

Laboratory Studies General laboratory studies may show anemia, azotemia, and abnormalities of liver function; in particular, there may be a raised serum glutamic-pyruvic transaminase concentration.33 Electrocardiographic findings of sinus tachycardia and T-wave abnormalities of the type seen in potassium depletion may be observed. CSF may show a modest elevation of the total protein content (rarely greater than 150 mg/dL) without pleocytosis. Clinical electrophysiologic findings have not been reported in detail. The expected findings would be those of a distal axonal neuropathy. Definite diagnosis depends upon the demonstration of thallium in body tissues or fluids. Methods of analysis have been reviewed.112 From the clinical point of view, estimation of thallium in the urine, usually by AAS, is the practical method. Qualitative colorimetric methods for screening urine are available. Forensic examination of all autopsy tissues and even cremation ash33 is possible. False-positive analytic results are rare because thallium concentrations are relatively simple to estimate, and this element is not a common contaminant in nature.

Tissue Distribution and Pathology Thallium salts are absorbed rapidly from the gastrointestinal and respiratory tracts. Formerly, thallium creams were applied topically as depilatory agents. Munch reviewed the cases of 59 patients, 50 of whom had evidence of “polyneuritis”; these patients had side effects from such topical application, suggesting strongly that skin absorption leads to significant systemic side effects. In modern cases, oral administration in a sweetened beverage is the usual route. Following absorption, maximum tissue concentrations are found in the kidney and brain. Excretion occurs predominantly in the urine. No data are available to document the half-life for excretion of a single dose in humans. In rats the half-life is 3 to 4 days,112 but it is of interest that, in the surviving patient of Cavanagh and colleagues, thallium was detected in the urine 7 weeks after administration of a single dose.33 At the cellular level, there is considerable evidence13 that thallium ions (Tl⫹) behave in a similar way to potassium ions. In physical chemistry terms, thallium and potassium salts may form mixed crystals, and the ionic radii of the two elements are similar. In many systems the two ions act interchangeably, particularly with respect to transport across membranes and in Na⫹,K⫹-dependent ATPase systems. In organotypic cell culture systems, swelling of axonal mitochondria is the first morphologic change associated with exposure to thallium ions (Fig. 113–5).195 However, as has been pointed out by Spencer and colleagues, although Tl⫹ has a strong affinity for the K⫹ site in the membranebound Na⫹,K⫹-ATPase system, it is not clear that binding produces inhibition.195 As with other metals, the biologic effect may be more significantly related to binding of sulfhydryl groups or oxidative stress. Mitochondrial swelling noted in this culture system was not accompanied by functional change in the neuron; however, it was suggested that the swelling might be the precursor of the apparently selective axonal damage that occurs in thallium intoxication.195,196,222 Schoental and Cavanagh speculated, because of pathologic similarities, that thallium may act by interfering with vitamin B (especially riboflavin) cofactors to enzyme systems within the axon.181 The most complete studies are the autopsy cases reported by Cavanagh and colleagues, which describe the nervous system changes in three patients.33,117 In a limited study of the nervous system in one patient, who died 8 days after the onset of neurologic symptoms, no significant changes were seen in the central nervous system or sciatic nerve. In a second patient, who died 19 days after the onset of neurologic symptoms, there were, qualitatively, few changes in the sciatic nerve but prominent changes of axonal degeneration in the sural nerve, vagus nerve, and each of several intramuscular nerves examined. This was seen in both longitudinal sections and teased fiber preparations. Ventral and dorsal spinal roots were normal, but dorsal root ganglia and dorsal

2542

Diseases of the Peripheral Nervous System

A

B

C FIGURE 113–5 A, Largely normal peripheral nervous system axonal mitochondrion 2 hours after introduction of thallium sulfate into the nutrient fluid of combination tissue culture. (⫻56,000.) B, Two axonal mitochondria from peripheral nerve fiber of a combination culture exposed to thallium sulfate for 24 hours. The large mitochondrion on the left displays many pleomorphic cristae and a granular matrix space. In the mitochondrion on the right, the matrix space is swollen and each crista (arrow) is shortened. (⫻34,000.) C, Peripheral nerve fiber from dorsal root of combination tissue culture exposed to thallium sulfate for 48 hours. The axon contains numerous vacuoles, each derived from a swollen mitochondrion (M). Smaller vacuoles are surrounded by both limiting mitochondrial membranes, while the larger vacuoles possess a single membrane. Note unaffected mitochondrion in an adjacent Schwann cell process (arrow). (⫻30,000.) (From Spencer, P. S., Peterson, E. R., Madrid, A. R., and Raine, C. S.: Effects of thallium salts on neuronal mitochondria in organotypic cord-ganglia-muscle combination cultures. J. Cell Biol. 58:79, 1973, with permission.)

columns showed postmortem changes. Little chromatolysis was seen in ventral gray matter of the spinal cord, but prominent chromatolytic changes were observed in the motor nuclei of cranial nerves III, VI, and VII. In the third patient, who died approximately 1 month after the onset of neurologic symptoms, extensive degenerative changes were seen

in all peripheral nerves examined and in the lumbar dorsal spinal roots. In the dorsal columns, the most prominent changes were in the fasciculus gracilis (Fig. 113–6). Chromatolytic changes were seen in dorsal root ganglia and in lumbosacral ventral horn cells, but no changes were observed in lumbar roots. The observations of distal sensory

2543

Metal Neuropathy

and motor changes with relative preservation of tendon reflexes, recovery of function in mild cases,178 more prominent changes in distal nerves, and the greater involvement of long tracts in the central nervous system suggest that this is a “dying-back” neuropathy. However, whether this is due to a direct effect on the terminal axon or to a primary effect on the cell body that is first manifested in the distal axon is not clear. Interference with perikaryal protein (or other macromolecular) synthesis or interference with synthesis of components required for anterograde axonal transport might occur in the cell body but first become manifest as structural changes in the distal axon.

Treatment Chelating agents (DMSA) have been used empirically in the management of thallium poisoning.178 Grossman reviewed the unconvincing benefits of BAL treatment82 and Kazantzis those relating to other chelating agents.112 Other types of treatment have been suggested that relate to the interaction of Tl⫹ and K⫹. Administration of K⫹ intravenously or orally increases the excretion rate of thallium; however, various authors have pointed out that this may transiently exacerbate the clinical situation.13,160 Presumably, this is due to mobilization of thallium previously deposited in tissue stores. Since thallium intoxication usually represents a single or small number of discrete exposures, and since thallium is excreted rapidly, there seems little theoretical or practical benefit to be gained from the use of potassium salts. The second type of treatment has been suggested for use in acute thallium poisoning soon after ingestion. Kamerbeek and colleagues used potassium ferric ferrocyanide II (Prussian blue) by mouth.107 This compound is not absorbed from the gut, but the K⫹ is replaced by Tl⫹, which is strongly bound and thus rendered unabsorbable. If this is to be effective, constipation must be prevented. It is not clear whether this treatment has any role after the dose of thallium has passed through the gastrointestinal tract.112 General circulatory supportive measures and good diuresis promote renal excretion of thallium without increasing its tissue availability from the serum. The prognosis for recovery depends upon the severity of the peripheral and central nervous system involvement.178

FIGURE 113–6 Transverse sections from spinal cord in a case of thallium poisoning. A, C2 level. Note the equal density of granules in gracile and cuneate tracts. (Marchi staining; ⫻11.) B, T1 level. Note that the gracile tract is more completely affected than the cuneate tract; few degenerating fibers enter the root zone. (Marchi staining; ⫻10.) C, L5 level. Note abundant degenerating fibers entering the root entry zone. (Marchi staining; ⫻12.) (From Kennedy, P., and Cavanagh, J. B.: Spinal changes in the neuropathy of thallium poisoning. J. Neurol. Sci. 29:295, 1976, with permission.)

GOLD Gold salts, usually sodium aurothiomalate, are used almost exclusively in the treatment of rheumatoid and psoriatic arthritis. These disorders may be associated with the development of peripheral neuropathy.51 For this reason, descriptions of gold neuropathy must be approached cautiously. Also, because the number of case reports is extremely small, the possibility that development of neuropathy is independent of the underlying disease or

2544

Diseases of the Peripheral Nervous System

therapy should be considered. In the most suggestive reports available, a neuropathy has developed coincidentally with gold therapy.59,68,110,119,126,216 The first symptoms, usually distal limb paresthesias, have appeared both after the first injection of gold and well into a course of prolonged therapy. In some cases, the neuropathy developed in the context of a systemic hypersensitivity reaction. In most cases the clinical course has been one of progressive symmetrical distal paresthesias followed by distal symmetrical limb muscle weakness and atrophy. Burning discomfort of the face has been described. The progression has usually been rapid, over days to weeks, and may lead to severe motor involvement with bilateral footdrop; in one case there was almost complete tetraparesis.70 Sensory findings involve distal loss of sensitivity to all modalities. The most suggestive feature that the neuropathy is due to gold is that cessation of therapy has led to regression of symptoms. Remission of neuropathy would be unusual in the type of angiopathic neuropathy usually seen with rheumatoid arthritis. The degree of recovery has depended upon the degree of severity of the maximum neurologic deficit. An interesting feature of several case reports is the association of clinical and electrophysiologic evidence of myokymia associated with gold therapy.110,141 Autonomic involvement has also been described.141 In cases in which it has been measured, the CSF protein level has been elevated at the height of the neuropathy and has fallen as the disease improved.110 One report describes an elevated CSF protein at 290 mg/dL with 45 lymphocytes/mm.126 Electrophysiologic findings have included a decreased amplitude of sensory nerve action potentials and CMAPs with slowing of conduction velocity. Increased irritability and spontaneous discharges have been seen on needle electromyography. Electrophysiologic evidence of myokymia has been described in muscles affected by clinical evidence of myokymia. Pathologic studies of the peripheral nervous system are limited but show evidence of both axonal degeneration and segmental demyelination. In the most detailed study of sural nerve biopsy material, there was an increase in segmental demyelination and remyelination in one case and of axonal degeneration in two cases.110 Similar changes occurred in hens treated with intraperitoneal injections of sodium aurothiomalate. The frequency of abnormalities in the animal study appeared to relate to the dose of gold administered. From the data available at present, it is not clear whether the effect of gold is the result of a direct toxic action on neurons, producing axonal degeneration and secondary segmental demyelination, or the effect is immunologically mediated. Many of the cases described fit the clinical pattern of an acute monophasic inflammatory demyelinating polyradiculoneuropathy. An immunologic process might be triggered by the use of gold in individuals with rheumatoid arthritis, or this might represent a completely independent process.

Treatment with BAL has been described,40,110 but in so few cases that it is not clear whether BAL increases the rate of spontaneous recovery. Certainly it would seem prudent to discontinue the use of gold therapy in the patient who develops a progressive neuropathy in the appropriate clinical context.

PLATINUM Over the years, metal compounds have occupied a prominent place in the pharmacopeia and have then been removed because of unacceptable side effects and the discovery of more effective agents. Thus mercury, arsenic, and thallium compounds, once in widespread use, are now regarded as therapeutic anachronisms. In the last 40 years, as gold salts found a role in the treatment of rheumatoid arthritis, platinum compounds have been introduced as effective chemotherapeutic agents.

Clinical Aspects Platinum compounds were introduced in the early 1970s as experimental agents following the observation that they inhibit bacterial cell division.173 They are used as therapy in the treatment of malignancies of the ovary, bladder, testes, and head, and neck. They may also increase survival in some cases of squamous cell carcinoma and osteosarcoma.25,165,214,221 The most common platinum compounds used for cancer chemotherapy are cisplatin and carboplatin. Both compounds bind to DNA, and it is postulated that this binding is responsible for the antimitotic activity.165 Reports of peripheral neuropathy appearing in association with the use of cisplatin have been made.18,25,45,48,84,113,134,159,165,168,171,207,213,214 All of the cases involve patients with advanced cancer, who have usually received several different chemotherapeutic agents. The underlying disease and many of these agents, especially the vinca alkaloids, are associated with the development of neuropathy. Discovering whether neuropathy is due to direct invasion or pressure on nerves, is a paraneoplastic effect, or is drug induced may be challenging. Special caution is required because the sensory neuropathy is similar to that described as a paraneoplastic process.35,100 Clinical histories of platinum neuropathy are uniform. Initial symptoms are paresthesias in the distal limbs progressing to sensory loss for all modalities in a glove-and-stocking distribution. Sensory loss and incoordination from proprioceptive loss dominate the clinical presentation. Symptoms usually appear after one to seven doses of intravenous cisplatin and progress over a number of weeks. Motor and autonomic involvement is not prominent. The most suggestive evidence for a causative relationship between cisplatin and neuropathy is that cessation of chemotherapy leads to

2545

Metal Neuropathy

stabilization or remission of symptoms and signs.84,113,168 In a number of reports, withdrawal of the drug did not result in improvement of the neuropathy,84,134 even with long-term follow-up.32 In addition, as with other metal-related neuropathies, symptoms and signs may progress for several weeks after the drug is discontinued. Carboplatin is an analogue of cisplatin now in routine use as a chemotherapeutic agent. Sensory neuropathy associated with treatment is similar to the neuropathy produced by cisplatin.136 Cold-induced paresthesias are a distinct early symptom of carboplatin neurotoxicity. Patients note that holding a cold object induces uncomfortable tingling in the hand and that swallowing cold liquids induces tingling and aching in the throat. The symptom is transient but may precede the typical diffuse sensory neuropathy.

Laboratory Studies Diagnostic studies have been limited, but in general, where electrophysiologic examinations are reported, there is absence of sensory nerve action potentials, normal MNCV and CMAPs, and normal needle electromyographic findings. These are similar to findings in paraneoplastic sensory neuropathy. Spinal fluid is normal or protein is mildly elevated.

Pathology Pathologic changes have been described in sural nerve biopsy.134,207,213 The descriptions suggest that axonal degeneration is the primary process. In animal models, cisplatin primarily induces changes in dorsal root ganglion neurons.34,61 Rodent and other animal models have been difficult to produce because animals often succumb to renal failure before developing significant structural change in peripheral nerve. Renal failure confounds changes seen in nerve that may be due to uremia. Cisplatin causes apoptosis of dorsal root ganglion neurons in tissue culture models. Neuronal death is associated with aberrant reentry into the first part (G1) of the cell cycle.71,137 This mechanism may be common in stress-induced and toxic neuronal death.129

Treatment Approximately 20% of patients cannot complete their full treatment course with cisplatin because of peripheral neurotoxicity. In model systems, a variety of agents prevent or diminish cisplatin-induced changes. These include nerve growth factor,5,62,71,94,137 neurotrophin-3,67 neurotrophin-4/5,229 and retinoic acid.209 Glutathione analogues have been used in human studies with equivocal benefit.115,190,199 The major treatment is still discontinuation of the drug.

ACKNOWLEDGMENT The expert secretarial assistance of Jane M. Meyer is most gratefully acknowledged.

REFERENCES 1. Aaseth, J., Jacobsen, D., Andersen, O., and Wickstrom, E.: Treatment of mercury and lead poisonings with dimercaptosuccinic acid and sodium dimercaptopropanesulfonate: a review. Analyst 120:853, 1995. 2. Albalak, R., Noonan, G., Buchanan, S., et al.: Blood lead levels and risk factors for lead poisoning among children in Jakarta, Indonesia. Sci. Total Environ. 301:75, 2003. 3. Aldrich, C. J.: Leuconychia striata arsenicalis transversus, with report of 3 cases. Am. J. Med. 127:702, 1904. 4. Anonymous: Report of Interdepartmental Working Group on Heavy Metals. London, Her Majesty’s Stationery Office, 1974. 5. Apfel, S. C., Arezzo, J. C., Lipson, L., and Kessler, J. A.: Nerve growth factor prevents experimental cisplatin neuropathy. Ann. Neurol. 31:76, 1992. 6. Araki, S., Honma, T., Yanagihara, S., and Ushio, K.: Recovery of slowed nerve conduction velocity in lead-exposed workers. Int. Arch. Occup. Environ. Health 46:151, 1980. 7. Aran, F.-A.: Recherches sur une maladie non encore décrite du système musculaire (atrophie musculaire progressive). Arch. Gen. Med. 24:172, 1850. 8. Armon, C., Kurland, L. T., Daube, J. R., and O’Brien, P. C.: Epidemiologic correlates of sporadic amyotrophic lateral sclerosis. Neurology 41:1077, 1991. 9. Bahlis, N. J., McCafferty-Grad, J., Jordan-McMurry, I., et al.: Feasibility and correlates of arsenic trioxide combined with ascorbic acid-mediated depletion of intracellular glutathione for the treatment of relapsed/refractory multiple myeloma. Clin. Cancer Res. 8:3658, 2002. 10. Bakir, F., Damluji, S. F., Amin-Zaki, L., et al.: Methylmercury poisoning in Iraq. Science 181:230, 1973. 11. Ball, G. V.: Two epidemics of gout. Bull. Hist. Med. 45:401, 1971. 12. Bank, W. J.: Thallium. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 570, 1980. 13. Bank, W. J., Pleasure, D. E., Suzuki, K., et al.: Thallium poisoning. Arch. Neurol. 26:456, 1972. 14. Barats, M. S., Gonick, H. C., Rothenberg, S., et al.: Severe lead-induced peripheral neuropathy in a dialysis patient. Am. J. Kidney Dis. 35:963, 2000. 15. Barltrop, D., Barrett, A. J., and Dingle, J. T.: Subcellular distribution of lead in the rat. J. Lab. Clin. Med. 77:705, 1971. 16. Barton, J. C., Conrad, M. E., Harrison, L., and Nuby, S.: Effects of calcium on the absorption and retention of lead. J. Lab. Clin. Med. 91:366, 1978. 17. Bau, D. T., Wang, T. S., Chung, C. H., et al.: Oxidative DNA adducts and DNA-protein cross-links are the major DNA lesions induced by arsenite. Environ. Health Perspect. 110:753, 2002.

2546

Diseases of the Peripheral Nervous System

18. Becher, R., Schütt, P., Osieka, R., and Schmidt, C. G.: Peripheral neuropathy and ophthalmologic toxicity after treatment with cis-dichlorodiaminoplatinum II. J. Cancer Res. Clin. Oncol. 96:219, 1980. 19. Behse, F., Pach, J., and Dorndorf, W.: Bleipolyneuropathie Klinische, elektrophysiologische und bioptische Befunde [Polyneuropathy due to lead poisoning: clinical, electrophysiological and biopsy findings]. Z. Neurol. 202:209, 1972. 20. Berlin, M.: Mercury. In Friberg, L., Nordberg, G., and Vouk, V. (eds.): Handbook on the Toxicology of Metals. Amsterdam, Elsevier, p. 503, 1979. 21. Blackman, S. S.: The lesions of lead encephalitis in children. Bull. Johns Hopkins Hosp. 61:1, 1937. 22. Boulding, J. E., and Baker, R. A.: The treatment of metal poisoning with penicillamine. Lancet 2:985, 1967. 23. Boyd, P. R., Walker, G., and Henderson, I. N.: The treatment of tetraethyl lead poisoning. Lancet 1:181, 1957. 24. Braman, R. S., and Foreback, C. C.: Methylated forms of arsenic in the environment. Science 182:1247, 1973. 25. Bruckner, H. W., Cohen, C. J., Wallach, R. C., et al.: Treatment of advanced ovarian cancer with cis-dichlorodiammineplatinum(II): poor-risk patients with intensive prior therapy. Cancer Treat. Rep. 62:555, 1978. 26. Brune, D., Samsahl, K., and Wester, P. O.: A comparison between the amounts of As, Au, Br, Cu, Fe, Mo, Se and Zn in normal and uraemic human whole blood by means of neutron activation analysis. Clin. Chim. Acta 13:285, 1966. 27. Buchthal, F., and Behse, F.: Electrophysiology and nerve biopsy in men exposed to lead. Br. J. Ind. Med. 36:135, 1979. 28. Campbell, A. M. G.: Calcium versenate in motor-neurone disease. Lancet 2:376, 1955. 29. Campbell, A. M. G., and Williams, E. R.: Chronic lead intoxication mimicking motor neurone disease. Br. Med. J. 4:582, 1968. 30. Campbell, A. M. G., Williams, E. R., and Barltrop, D.: Motor neurone disease and exposure to lead. J. Neurol. Neurosurg. Psychiatry 33:877, 1970. 31. Catton, M. J., Harrison, M. J. G., Fullerton, P. M., and Kazantzis, G.: Subclinical neuropathy in lead workers. Br. Med. J. 2:80, 1970. 32. Cavaletti, G., Bogliun, G., Marzorati, L., et al.: Longterm peripheral neurotoxicity of cisplatin in patients with successfully treated epithelial ovarian cancer. Anticancer Res. 14:1287, 1994. 33. Cavanagh, J. B., Fuller, N. H., Johnson, H. R. M., and Rudge, P.: The effects of thallium salts, with particular reference to the nervous system changes: a report of three cases. Q. J. Med. 43:293, 1974. 34. Cece, R., Petruccioli, M. G., Cavaletti, G., et al.: An ultrastructural study of neuronal changes in dorsal root ganglia (DRG) of rats after chronic cisplatin administrations. Histol. Histopathol. 10:837, 1995. 35. Chalk, C. H., Windebank, A. J., Kimmel, D. W., and McManis, P. G.: The distinctive clinical features of paraneoplastic sensory neuronopathy. Can. J. Neurol. Sci. 19:346, 1992. 36. Chhuttani, P. N., Chawla, L. S., and Sharma, T. D.: Arsenical neuropathy. Neurology 17:269, 1967.

37. Chia, S. E., Chia, H. P., Ong, C. N., and Jeyaratnam, J.: Cumulative blood lead levels and nerve conduction parameters. Occup. Med. 46:59, 1996. 38. Chisolm, J. J. Jr.: The use of chelating agents in the treatment of acute and chronic lead intoxication in childhood. J. Pediatr. 73:1, 1968. 39. Chisolm, J. J. Jr.: Treatment of acute lead intoxication—choice of chelating agents and supportive therapeutic measures. Clin. Toxicol. 3:527, 1970. 40. Cohen, A., Goldman, J., and Dubbs, A. W.: The treatment of acute gold and arsenic poisoning: use of BAL (2,3-dimercaptopropanol, British Anti-Lewisite). JAMA 133:749, 1947. 41. Conradi, S., Ronnevi, L.-O., Nise, G., and Vesterberg, O.: Abnormal distribution of lead in amyotrophic lateral sclerosis: reestimation of lead in the cerebrospinal fluid. J. Neurol. Sci. 48:413, 1980. 42. Conradi, S., Ronnevi, L.-O., and Vesterberg, O.: Increased plasma levels of lead in patients with amyotrophic lateral sclerosis compared with control subjects as determined by flameless atomic absorption spectrophotometry. J. Neurol. Neurosurg. Psychiatry 41:389, 1978. 43. Cordeiro, R., Lima Filho, E. C., and Salgado, P. E.: Neurological disorders in workers with low levels of lead in the blood. I: Peripheral neuropathy. Rev. Saude Publica 30:248, 1996. 44. Couratier, P., Bernet-Bernady, P., Truong, T., et al.: Lead intoxication and amyotrophic lateral sclerosis. Rev. Neurol. 154:345, 1998. 45. Cowan, J. D., Kies, M. S., Roth, J. L., and Joyce, R. P.: Nerve conduction studies in patients treated with cisdiamminedichloroplatinum(II): a preliminary report. Cancer Treat. Rep. 64:1119, 1980. 46. Currier, R. D., and Haerer, A. F.: Amyotrophic lateral sclerosis and metallic toxins. Arch. Environ. Health 17:712, 1968. 47. Dalton, W. S.: Targeting the mitochondria: an exciting new approach to myeloma therapy. Clin. Cancer Res. 8:3643, 2002. 48. DeKoning, P., Neijt, J. P., Jennekens, F. G. I., and Gispen, W. H.: Evaluation of cisdiamminedichloroplatinum (II) (cisplatin) neurotoxicity in rats. Toxicol. Appl. Pharmacol. 89:81, 1987. 49. Delwaide, P. J., and Chantraine, A.: Participation spinale dans l’intoxication saturnine: arguments electromyographiques. Electroencephalogr. Clin. Neurophysiol. Suppl. 25:643, 1965. 50. Donofrio, P. D., Wilbourn, A. J., Albers, J. W., et al.: Acute arsenic intoxication presenting as Guillain-Barré-like syndrome. Muscle Nerve 10:114, 1987. 51. Dyck, P. J., Conn, D. L., and Okazaki, H.: Necrotizing angiopathic neuropathy: three-dimensional morphology of fiber degeneration related to sites of occluded vessels. Mayo Clin. Proc. 47:461, 1972. 52. Dyck, P. J., Gutrecht, J. A., Bastron, J. A., et al.: Histologic and teased-fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clin. Proc. 43:81, 1968. 53. Dyck, P. J., O’Brien, P. C., and Ohnishi, A.: Lead neuropathy: 2. Random distribution of segmental demyelination among “old internodes” of myelinated fibers. J. Neuropathol. Exp. Neurol. 36:570, 1977.

Metal Neuropathy 54. Ehle, A. L.: Lead neuropathy and electrophysiological studies in low level lead exposure: a critical review. Neurotoxicology 7:203, 1986. 55. Emmerson, B. T.: Chronic lead nephropathy: the diagnostic use of calcium EDTA and the association with gout. Australas. Ann. Med. 12:310, 1963. 56. Ericson, J. E., Shirahata, H., and Patterson, C. C.: Skeletal concentrations of lead in ancient Peruvians. N. Engl. J. Med. 300:946, 1979. 57. Erlicki, A., and Rybalkin, J.: Ueber arseniklähmung. Arch. Psychiatr. Nervenkr. 23:861, 1892. 58. Everson, J., and Patterson, C. C.: “Ultra-clean” isotope dilution/mass spectrometric analyses for lead in human blood plasma indicate that most reported values are artificially high. Clin. Chem. 26:1603, 1980. 59. Fam, A. G., Gordon, D. A., Sarkozi, J., et al.: Neurologic complications associated with gold therapy for rheumatoid arthritis. J. Rheumatol. 11:700, 1984. 60. Felmus, M. T., Patten, B. M., and Swanke, L.: Antecedent events in amyotrophic lateral sclerosis. Neurology 26:167, 1976. 61. Fischer, S. J., McDonald, E. S., Gross, L., and Windebank, A. J.: Alterations in cell cycle regulation underlie cisplatin induced apoptosis of dorsal root ganglion neurons in vivo. Neurobiol. Dis. 8:1027, 2001. 62. Fischer, S. J., Podratz, J. L., and Windebank, A. J.: Nerve growth factor rescue of cisplatin neurotoxicity is mediated through the high affinity receptor: studies in PC12 cells and p75 null mouse dorsal root ganglia. Neurosci. Lett. 308:1, 2001. 63. Freeman, R.: Chelating agents in treatment of lead intoxication [letter]. J. Pediatr. 74:322, 1969. 64. Friberg, L., and Nordberg, G. F.: Inorganic mercury— relation between exposure and effects. In Friberg, L., and Vostal, J. (eds.): Mercury in the Environment. Cleveland, CRC Press, p. 113, 1972. 65. Fukaya, Y., Matsumoto, T., Fujiwara, N., and Tokudome, S.: Relationship between vibratory sense threshold and blood lead concentration in ceramic color workers and transfer printing manufacturers. Nippon Eiseigaku Zasshi 50:748, 1995. 66. Fullerton, P. M.: Chronic peripheral neuropathy produced by lead poisoning in guinea-pigs. J. Neuropathol. Exp. Neurol. 25:214, 1966. 67. Gao, W.-Q., Dybdal, N., Shinsky, N., et al.: Neurotrophin-3 reverses experimental cisplatin-induced peripheral sensory neuropathy. Ann. Neurol. 38:30, 1995. 68. Garrido, J. A., Torres-Pardo, M. V., Sola, D., et al.: Myokimia induced by gold salts. Neurologia 10:235, 1995. 69. Gilfillan, S. C.: Lead poisoning and the fall of Rome. J. Occup. Med. 7:53, 1965. 70. Gilg, G., and Vraa-Jensen, G.: Histological examination of the central nervous system after gold treatment. Acta Psychiatr. Neurol. Scand. 33:174, 1958. 71. Gill, J. S., and Windebank, A. J.: Cisplatin-induced apoptosis in rat dorsal root ganglion neurons is associated with attempted entry into the cell cycle. J. Clin. Invest. 101:2842, 1998. 72. Goebel, H. H., Schmidt, P. F., Bohl, J., et al.: Polyneuropathy due to acute arsenic intoxication: biopsy studies. J. Neuropathol. Exp. Neurol. 49:137, 1990.

2547

73. Goldberg, A., Smith, J. A., and Lochhead, A. C.: Treatment of lead-poisoning with oral penicillamine. Br. Med. J. 1:1270, 1963. 74. Goldstein, N. P., McCall, J. T., and Dyck, P. J.: Metal neuropathy. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy. Philadelphia, W. B. Saunders, p. 1227, 1975. 75. Gordon, J. N., Taylor, A., and Bennett, P. N.: Lead poisoning: case studies. Br. J. Clin. Pharmacol. 53:451, 2002. 76. Goyer, R. A.: Lead and the kidney. Curr. Top. Pathol. 55:147, 1971. 77. Goyer, R. A., May, P., Cates, M. M., and Krigman, M. R.: Lead and protein content of isolated intranuclear inclusion bodies from kidneys of lead-poisoned rats. Lab. Invest. 22:245, 1970. 78. Goyer, R. A., and Rhyne, B. C.: Pathological effects of lead. Int. Rev. Exp. Pathol. 12:1, 1973. 79. Grandjean, P., Nielsen, O. V., and Shapiro, I. M.: Lead retention in ancient Nubian and contemporary populations. J. Environ. Pathol. Toxicol. 2:781, 1979. 80. Greengard, J., Rowley, W., Elam, H., and Perlstein, M.: Lead encephalopathy in children: intravenous use of urea in its management. N. Engl. J. Med. 264:1027, 1961. 81. Griggs, R. C.: Lead poisoning: hematologic aspects. Progr. Hematol. 4:117, 1964. 82. Grossman, H.: Thallotoxicosis: report of a case and a review. Pediatrics 16:868, 1955. 83. Guidetti, D., Bondavalli, M., Sabadini, R., et al.: Epidemiological survey of amyotrophic lateral sclerosis in the Province of Reggio Emilia, Italy: influence of environmental exposure to lead. Neuroepidemiology 15:301, 1996. 84. Hadley, D., and Herr, H. W.: Peripheral neuropathy associated with cis-dichlorodiammineplatinum (II) treatment. Cancer 44:2026, 1979. 85. Haley, T. J.: Saturnism, pediatric and adult lead poisoning. Clin. Toxicol. 4:11, 1971. 86. Hall, A. H.: Chronic arsenic poisoning. Toxicol. Lett. 128:69, 2002. 87. Hammond, P. B.: Lead poisoning: an old problem with a new dimension. Essays Toxicol. 1:115, 1969. 88. Hantson, P., Haufroid, V., Buchet, J. P., and Mahieu, P.: Acute arsenic poisoning treated by intravenous dimercaptosuccinic acid (DMSA) and combined extrarenal epuration techniques. J. Toxicol. Clin. Toxicol. 41:1, 2003. 89. Harary, I., and Berliner, J.: Electron microscopic examination of lead treated L6 skeletal muscle line cells in culture. J. Environ. Pathol. Toxicol. 4:305, 1980. 90. Hardy, H. L., Chamberlin, R. I., Maloof, C. C., et al.: Lead as an environmental poison. Clin. Pharmacol. Ther. 12:982, 1971. 91. Harris, C. E. C.: A comparison of intravenous calcium disodium versenate and oral penicillamine in promoting elimination of lead. Can. Med. Assoc. J. 79:664, 1958. 92. Hartwig, A., and Schwerdtle, T.: Interactions by carcinogenic metal compounds with DNA repair processes: toxicological implications. Toxicol. Lett. 127:47, 2002. 93. Hausmanowa-Petrusewicz, I., Emeryk, B., Sobkawicz, H., et al.: Electromyographic records in lead poisoning. Pol. Tyg. Lek. 17:1405, 1962.

2548

Diseases of the Peripheral Nervous System

94. Hayakawa, K., Itoh, T., Niwa, H., et al.: NGF prevention of neurotoxicity induced by cisplatin, vincristine and Taxol depends on toxicity of each drug and NGF treatment schedule: in vitro study of adult rat sympathetic ganglion explants. Brain Res. 794:313, 1998. 95. Hernberg, S., and Nikkanen, J.: Enzyme inhibition by lead under normal urban conditions. Lancet 1:63, 1970. 96. Heyman, A., Pfeiffer, J. B. Jr., Willett, R. W., and Taylor, H. M.: Peripheral neuropathy caused by arsenical intoxication: a study of 41 cases with observations on the effects of BAL (2,3, dimercapto-propanol). N. Engl. J. Med. 254:402, 1956. 97. Hilburn, M. E.: Environmental lead in perspective. Chem. Soc. Rev. 8:63, 1979. 98. Hillman, F. E.: A rare case of chronic lead poisoning: polyneuropathy traced to lead shot in the appendix. Ind. Med. Surg. 36:488, 1967. 99. Hopkins, A.: Experimental lead poisoning in the baboon. Br. J. Ind. Med. 27:130, 1970. 100. Horwich, M. S., Cho, L., Porro, R. S., and Posner, J. B.: Subacute sensory neuropathy: a remote effect of carcinoma. Ann. Neurol. 2:7, 1977. 101. Hughes, M. F.: Arsenic toxicity and potential mechanisms of action. Toxicol. Lett. 133:1, 2002. 102. Hyslop, G. H., and Kraus, W. M.: The pathology of motor paralysis by lead. Arch. Neurol. Psychiatry 10:444, 1923. 103. Jacobs, J. M., Carmichael, N., and Cavanagh, J. B.: Ultrastructural changes in the distal root and trigeminal ganglion of rats poisoned with methyl mercury. Neuropathol. Appl. Neurobiol. 1:1, 1975. 104. Jenkins, R. B.: Reynolds, Aldrich or Mees? A consideration of transverse striate leukonychia. Am. J. Med. Sci. 246:707, 1963. 105. Jensen, W. N., Morena, G. D., and Bessis, M. C.: An electron microscopic description of basophilic stippling in red cells. Blood 25:933, 1965. 106. Kamel, F., Umbach, D. M., Munsat, T. L., et al.: Lead exposure and amyotrophic lateral sclerosis. Epidemiology 13:311, 2002. 107. Kamerbeek, H. H., Rauws, A. G., ten Ham, M., and van Heijst, A. N. P.: Dangerous redistribution of thallium by treatment with sodium diethyldithiocarbamate. Acta Med. Scand. 189:149, 1971. 108. Kantarjian, A. D.: A syndrome clinically resembling amyotrophic lateral sclerosis following chronic mercurialism. Neurology 11:639, 1961. 109. Kapaki, E., Segditsa, J., Zournas, C., et al.: Determination of cerebrospinal fluid and serum lead levels in patients with amyotrophic lateral sclerosis and other neurological diseases. Experientia 45:1108, 1989. 110. Katrak, S. M., Pollock, M., O’Brien, C. P., et al.: Clinical and morphological features of gold neuropathy. Brain 103:671, 1980. 111. Kawasaki, S., Yazawa, S., Ohnishi, A., and Ohi, T.: Chronic and predominantly sensory polyneuropathy in Toroku Valley where a mining company produced arsenic. Clin. Neurol. 42:504, 2002. 112. Kazantzis, G.: Thallium. In Friberg, L., Nordberg, G. F., and Vouk, V. B. (eds.): Handbook on the Toxicology of Metals. Amsterdam, Elsevier, p. 599, 1979.

113. Kedar, A., Cohen, M. E., and Freeman, A. I.: Peripheral neuropathy as a complication of cis-dichlorodiammineplatinum (II) treatment: a case report. Cancer Treat. Rep. 62:819, 1978. 114. Kehoe, R. A.: The metabolism of lead in man in health and disease. I. The normal metabolism of lead. R. Inst. Public Health Hyg. 24:81, 1961. 115. Kelynack, T. N., Kirby, W., and Delepine, S.: Arsenical poisoning from beer-drinking. Lancet 2:1600, 1900. 116. Kemp, G., Rose, P., Lurain, J., et al.: Amifostine pretreatment for protection against cyclophosphamide-induced and cisplatin-induced toxicities: results of a randomized control trial in patients with advanced ovarian cancer. J. Clin. Oncol. 14:2101, 1996. 117. Kennedy, P., and Cavanagh, J. B.: Spinal changes in the neuropathy of thallium poisoning: a case with neuropathological studies. J. Neurol. Sci. 29:295, 1976. 118. Kitchin, K. T., and Ahmad, S.: Oxidative stress as a possible mode of action for arsenic carcinogenesis. Toxicol. Lett. 137:3, 2003. 119. Koh, W. H., and Boey, M. L.: Polyneuropathy following intra-muscular sodium aurothiomalate for rheumatoid arthritis—a case report. Ann. Acad. Med. Singapore 21:821, 1992. 120. Kubis, N., Talamon, C., Smadja, D., and Said, G.: Peripheral neuropathy caused by thallium poisoning. Rev. Neurol. 153:599, 1997. 121. Kurland, L. T., Faro, S. N., and Siedler, H.: Minamata disease: the outbreak of a neurologic disorder in Minamata, Japan, and its relationship to the ingestion of seafood contaminated by mercuric compounds. World Neurol. 1:370, 1960. 122. Kyle, R. A., and Pease, G. L.: Hematologic aspects of arsenic intoxication. N. Engl. J. Med. 273:18, 1965. 123. Lampert, P. W., and Schochet, S. S. Jr.: Demyelination and remyelination in lead neuropathy: electron microscopic studies. J. Neuropathol. Exp. Neurol. 27:527, 1968. 124. Le Quesne, P. M., and McLeod, J. G.: Peripheral neuropathy following a single exposure to arsenic: clinical course in four patients with electrophysiological and histological studies. J. Neurol. Sci. 32:437, 1977. 125. Leikin, S., and Eng, G.: Erythrokinetic studies of the anemia of lead poisoning. Pediatrics 31:996, 1963. 126. Leiper, E. J. R.: A case of polyneuritis due to gold. Br. Med. J. 2:119, 1946. 127. Levander, O. A., Fisher, M., Morris, V. C., and Ferretti, R. J.: Morphology of erythrocytes from vitamin E-deficient leadpoisoned rats. J. Nutr. 107:1828, 1977. 128. Liebscher, K., and Smith, H.: Essential and nonessential trace elements: a method of determining whether an element is essential or nonessential in human tissue. Arch. Environ. Health. 17:881, 1968. 129. Liu, D. X., and Greene, L. A.: Neuronal apoptosis at the G1/S cell cycle checkpoint. Cell Tissue Res. 305:217, 2001. 130. Livesley, B., and Sissons, C. E.: Chronic lead intoxication mimicking motor neurone disease. Br. Med. J. 4:387, 1968. 131. Louwerse, E. S., Buchet, J. P., Van Dijk, M. A., et al.: Urinary excretion of lead and mercury after oral administration of meso-2,3-dimercaptosuccinic acid in patients with motor neurone disease. Int. Arch. Occup. Environ. Health 67:135, 1995.

Metal Neuropathy 132. Ly, B. T., Williams, S. R., and Clark, R. F.: Mercuric oxide poisoning treated with whole-bowel irrigation and chelation therapy. Ann. Emerg. Med. 39:312, 2002. 133. Mappes, R.: Experiments on excretion of arsenic in urine [author’s transl.]. Int. Arch. Occup. Environ. Health 40:267, 1977. 134. Marin, A. C., and Rierson, B.: Peripheral neuropathy secondary to cis-dichlorodiammino-platinum (II) (Platinol): treatment for advanced ovarian cancer. Ariz. Med. 36:898, 1979. 135. Mayan, O. N., Henriques, A. T., and Calheiros, J. M.: Childhood lead exposure in Oporto, Portugal. Int. J. Occup. Environ. Health 7:209, 2001. 136. Mayerhofer, K., Bodner-Adler, B., Bodner, K., et al.: Paclitaxel/carboplatin as first-line chemotherapy in advanced ovarian cancer: efficacy and adverse effects with special consideration of peripheral neurotoxicity. Anticancer Res. 20:4047, 2000. 137. McDonald, E. S., Windebank, A. J.: Cisplatin-induced apoptosis of DRG neurons involves Bax redistribution and cytochrome c release but not Fas receptor signaling. Neurobiol. Dis. 9:220, 2002. 138. McFee, R. B., and Caraccio, T. R.: Intravenous mercury injection and ingestion: clinical manifestations and management. J. Toxicol. Clin. Toxicol. 39:733, 1995. 139. Mealey, J. Jr., Brownell, G. L., and Sweet, W. H.: Radioarsenic in plasma, urine, normal tissues, and intracranial neoplasms: distribution and turnover after intravenous injection in man. AMA Arch. Neurol. Psychiatry 81:310, 1959. 140. Mees, R. A.: Een verschijnsel bij polyneuritis arsenicosa. Ned. Geneeskd. 1:391, 1919. 141. Meyer, M., Haecki, M., Ziegler, W., et al.: Autonomic dysfunction and myokymia in gold neuropathy. In Canal, N., and Pozza, G. (eds.): Peripheral Neuropathies: Proceedings of the International Symposium on Peripheral Neuropathy, Milan, 1978. Amsterdam, Elsevier/North-Holland Biomedical Press, p. 475, 1978. 142. Miller, W. J., Schipper, H. M., Lee, J. S., et al.: Mechanisms of action of arsenic trioxide. Cancer Res. 62:3893, 2002. 143. Miyakawa, T., Deshimaru, M., Sumiyoshi, S., et al.: Experimental organic mercury poisoning—pathological changes in peripheral nerves. Acta Neuropathol. (Berl.) 15:45, 1970. 144. Monton, F., and Coria, F.: Reversible Schwann cell hypertrophy in lead neuropathy. Neuropathol. Appl. Neurobiol. 17:231, 1991. 145. Moore, J. F., Goyer, R. A., and Wilson, M.: Lead-induced inclusion bodies: solubility, amino acid content, and relationship to residual acidic nuclear proteins. Lab. Invest. 29:488, 1973. 146. Muckter, H., Liebl, B., Reichl, F. X., et al.: Are we ready to replace dimercaprol (BAL) as an arsenic antidote? Hum. Exp. Toxicol. 16:460, 1997. 147. Munch, J. C.: Human thallotoxicosis. J. Am. Med. Assoc. 102:1929, 1934. 148. Munch, J. C., Ginsburg, H. M., and Nixon, C. E.: The 1932 thallotoxicosis outbreak in California. J. Am. Med. Assoc. 100:1315, 1933. 149. Myers, R. R., Powell, H. C., Shapiro, H. M., et al.: Changes in endoneurial fluid pressure, permeability, and peripheral

150.

151.

152.

153.

154. 155. 156.

157.

158. 159. 160.

161. 162.

163.

164.

165.

166.

2549 nerve ultrastructure in experimental lead neuropathy. Ann. Neurol. 8:392, 1980. Neal, P. A., and Jones, R. R.: Chronic mercurialism in the hatters’ fur-cutting industry. J. Am. Med. Assoc. 110:337, 1938. Nixon, D. E., and Moyer, T. P.: Arsenic analysis II: rapid separation and quantification of inorganic arsenic plus metabolites and arsenobetaine from urine. Clin. Chem. 38:2479, 1992. Nixon, D. E., Moyer, T. P., Windebank, A. J., et al.: Lack of correlation of low levels of whole blood and serum lead in humans: an experimental evaluation of this relationship in animals. Trace Substances Environ. Health XIX:248, 1986. Nordberg, G. F.: Effects and dose-response relationships of toxic metals: a report from an international meeting. Scand. J. Work Environ. Health 2:37, 1976. Oh, S. J.: Lead neuropathy: case report. Arch. Phys. Med. Rehabil. 56:312, 1975. Ohlsson, W. T. L.: Penicillamine as lead-chelating substance in man. Br. Med. J. 1:1454, 1962. Ohnishi, A., Murai, Y., Ikeda, M., and Kuroiwa, Y.: Experimental methylmercury intoxication: morphometric analysis of the nerve fiber of the sural nerve, lumbar spinal roots, and Goll’s tract. Shinkei Naika Neurol. Med. 9:564, 1978. Ohnishi, A., Schilling, K., Brimijoin, W. S., et al.: Lead neuropathy. 1) Morphometry, nerve conduction, and choline acetyltransferase transport: new finding of endoneurial edema associated with segmental demyelination. J. Neuropathol. Exp. Neurol. 36:499, 1977. Ohta, M.: Ultrastructure of sural nerve in a case of arsenical neuropathy. Acta Neuropathol. (Berl.) 16:233, 1970. Ongerboer de Visser, B. W., and Tiessens, G.: Polyneuropathy induced by cisplatin. Prog. Exp. Tumor Res. 29:190, 1985. Papp, J. P., Gay, P. C., Dodson, V. N., and Pollard, H. M.: Potassium chloride treatment in thallotoxicosis. Ann. Intern. Med. 71:119, 1969. Passarge, C., and Wieck, H. H.: Thallium-polyneuritis. Fortschr. Neurol. Psychiatr. 33:477, 1965. Pentschew, A., and Garro, F.: Lead encephalo-myelopathy of the suckling rat and its implications on the porphyrinopathic nervous diseases: with special reference to the permeability disorders of the nervous system’s capillaries. Acta Neuropathol. (Berl.) 6:266, 1966. Pinto, S. S., Varner, M. O., Nelson, K. W., et al.: Arsenic trioxide absorption and excretion in industry. J Occup Med. 18:677, 1976. Politis, M. J., Schaumburg, H. H., and Spencer, P. S.: Neurotoxicity of selected chemicals. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 613, 1980. Prestayko, A. W., D’Aoust, J. C., Issell, B. F., and Crooke, S. T.: Cisplatin (cis-diamminedichloroplatinum II). Cancer Treat. Rev. 6:17, 1979. Rahbar, M. H., White, F., Agboatwalla, M., et al.: Factors associated with elevated blood lead concentrations in children in Karachi, Pakistan. Bull. World Health Organ. 80:769, 2002.

2550

Diseases of the Peripheral Nervous System

167. Reed, D., Crawley, J., Faro, S. N., et al.: Thallotoxicosis: acute manifestations and sequelae. JAMA 183:96, 1963. 168. Reinstein, L., Ostrow, S. S., and Wiernik, P. H.: Peripheral neuropathy after cis-platinum (II) (DDP) therapy. Arch. Phys. Med. Rehabil. 61:280, 1980. 169. Richards, R. T.: Lead poisoning. Rocky Mt. Med. J. 64:59, 1967. 170. Rieders, F.: Current concepts in the therapy of lead poisoning. In Sever, M. J. (ed.): Metal Binding in Medicine. Philadelphia, J. B. Lippincott, p. 143, 1960. 171. Roelofs, R. I., Hrushesky, W., Rogin, J., and Rosenberg, L.: Peripheral sensory neuropathy and cisplatin chemotherapy. Neurology 34:934, 1984. 172. Roelofs-Iverson, R. A., Mulder, D. W., Elveback, L. R., et al.: ALS and heavy metals: a pilot case-control study. Neurology 34:393, 1984. 173. Rosenberg, B., Van Camp, L., and Krigas, T.: Inhibition of cell division in E. coli by electrolysis products from a platinum electrode. Nature 205:698, 1965. 174. Ross, A. T.: Mercuric polyneuropathy with albumino-cytologic dissociation and eosinophilia. JAMA 188:830, 1964. 175. Rubens, O., Logina, E., Kravale, I., et al.: Peripheral neuropathy in chronic occupational inorganic lead exposure: a clinical and electrophysiological study. J. Neurol. Neurosurg. Psychiatry 71:200, 2001. 176. Rubin, C. H., Esteban, E., Reissman, D. B., et al.: Lead poisoning among young children in Russia: concurrent evaluation of childhood lead exposure in Ekaterinburg, Krasnouralsk, and Volgograd. Environ. Health Perspect. 110:559, 2002. 177. Rubino, G. F.: The role of lead in porphyrin metabolism. Panminerva Med. 4:340, 1962. 178. Rusyniak, D. E., Furbee, R. B., and Kirk, M. A.: Thallium and arsenic poisoning in a small midwestern town. Ann. Emerg. Med. 39:307, 2002. 179. Sachs, H. K., Blanksma, L. A., Murray, E. F., and O’Connell, M. J.: Ambulatory treatment of lead poisoning: report of 1,155 cases. Pediatrics 46:389, 1970. 180. Schaumburg, H. H., and Berger, A.: Alopecia and sensory polyneuropathy from thallium in a Chinese herbal medication. JAMA 268:3430, 1992. 181. Schoental, R., and Cavanagh, J. B.: Mechanisms involved in the “dying-back” process—an hypothesis implicating coenzymes. Neuropathol. Appl. Neurobiol. 3:145, 1977. 182. Selander, S.: Treatment of lead poisoning: a comparison between the effects of sodium calcium edetate and penicillamine administered orally and intravenously. Br. J. Ind. Med. 24:272, 1967. 183. Selander, S., Cramer, K., and Hallberg, L.: Studies in lead poisoning. Oral therapy with penicillamine: relationship between lead in blood and other laboratory tests. Br. J. Ind. Med. 23:282, 1966. 184. Seppäläinen, A. M., Tola, S., Hernberg, S., and Kock, B.: Subclinical neuropathy at “safe” levels of lead exposure. Arch. Environ. Health 30:180, 1975. 185. Seto, D. S. Y., and Freeman, J. M.: Lead neuropathy in childhood. Am. J. Dis. Child. 107:337, 1964. 186. Settle, D. M., and Patterson, C. C.: Lead in albacore: guide to lead pollution in Americans. Science 207:1167, 1980.

187. Shapiro, I. M., Mitchell, G., Davidson, I., and Katz, S. H.: The lead content of teeth: evidence establishing new minimal levels of exposure in a living preindustrialized human population. Arch. Environ. Health 30:483, 1975. 188. Sidgwick N. V.: The Chemical Elements and Their Compounds 1. Oxford, UK, Clarendon Press, 1950. 189. Simpson, J. A., Seaton, D. A., and Adams, J. F.: Response to treatment with chelating agents of anaemia, chronic encephalopathy, and myelopathy due to lead poisoning. J. Neurol. Neurosurg. Psychiatry 27:536, 1964. 190. Smith, H.: The interpretation of the arsenic content of human hair. Forensic Sci. Soc. J. 4:192, 1964. 191. Smyth, J. F., Bowman, A., Perren, T., et al.: Glutathione reduces the toxicity and improves quality of life of women diagnosed with ovarian cancer treated with cisplatin: results of a double-blind, randomised trial. Ann. Oncol. 8:569, 1997. 192. Sollman, T.: A Manual of Pharmacology. Philadelphia, W. B. Saunders, 1957. 193. Somjen, G. G., Herman, S. P., and Klein, R.: Electrophysiology of methyl mercury poisoning. J. Pharmacol. Exp. Ther. 186:579, 1973. 194. Somjen, G. G., Herman, S. P., Klein, R., et al.: The uptake of methyl mercury (203Hg) in different tissues related to its neurotoxic effects. J. Pharmacol. Exp. Ther. 187:602, 1973. 195. Spencer, P. S., Peterson, E. R., Madrid, R., and Raine, C. S.: Effects of thallium salts on neuronal mitochondria in organotypic cord-ganglia-muscle combination cultures. J. Cell Biol. 58:79, 1973. 196. Spencer, P. S., and Schaumburg, H. H.: Central-peripheral distal axonopathy: the pathology of dying-back polyneuropathies. Prog. Neuropathol. 3:253, 1976. 197. St. George, I. M.: Two men with lead poisoning. N. Z. Med. J. 71:294, 1970. 198. Stocken, L. A., and Thompson, R. H. S.: British anti-lewisite: I. Arsenic derivatives of thiol proteins. Biochem. J. 40:529, 1946. 199. Stocken, L. A., and Thompson, R. H. S.: British anti-lewisite: II. Dithiol compounds as antidotes for arsenic. Biochem. J. 40:535, 1946. 200. Sumiyoshi, Y., Hashine, K., Kasahara, K., and Karashima, T.: Glutathione chemoprotection therapy against CDDP-induced neurotoxicity in patients with invasive bladder cancer [in Japanese]. Jpn. J. Cancer Chemother. 23:1506, 1996. 201. Svare, C. W., Peterson, L. C., Reinhardt, J. W., et al.: The effect of dental amalgams on mercury levels in expired air. J. Dent. Res. 60:1668, 1981. 202. Swaiman, K. F., and Flagler, D. G.: Penicillamine therapy of the Guillain-Barré syndrome caused by mercury poisoning. Neurology 21:456, 1971. 203. Takeuchi, T.: Neuropathology of Minamata disease in Kumamoto: especially at the chronic stage. In Roizin, L., Shiraki, H., and Grcevic, N. (eds.): Neurotoxicology. New York, Raven Press, p. 235, 1977. 204. Takeuchi, T., and Eto, K.: Pathology and pathogenesis of Minamata disease. In Tsubaki, T., and Irukayama, K. (eds.): Minamata Disease: Methylmercury Poisoning in Minamata and Nigata, Japan. New York, Elsevier, p. 103, 1977. 205. Ter Haar, G. L., and Bayard, M. A.: Composition of airborne lead particles. Nature 232:553, 1971.

Metal Neuropathy 206. Thomas, H. M.: A case of generalized lead paralysis, with a review of the cases of lead palsy seen in the hospital. Johns Hopkins Hosp. Bull. 15:209, 1904. 207. Thompson, S. W., Davis, L. E., Kornfeld, M., et al.: Cisplatin neuropathy: clinical, electrophysiologic, morphologic, and toxicologic studies. Cancer 54:1269, 1984. 208. Tominack, R., Weber, J., Blume, C., et al.: Elemental mercury as an attractive nuisance: multiple exposures from a pilfered school supply with severe consequences. Pediatr. Emerg. Care 18:97, 2002. 209. Tredici, G., Tredici, S., Fabbrica, D., et al.: Experimental cisplatin neuronopathy in rats and the effect of retinoic acid administration. J. Neurooncol. 36:31, 1998. 210. Tsuchiya, K.: Lead. In Friberg, L., Nordberg, G. F., and Vouk, V. B. (eds.): Handbook on the Toxicology of Metals. Amsterdam, Elsevier, p. 541, 1979. 211. Vallee, B. L., Ulmer, D. D., and Wacker, W. E. C.: Arsenic toxicology and biochemistry. AMA Arch. Ind. Health 21:56, 1960. 212. Vanacore, N., Corsi, L., Fabrizio E., et al.: Relationship between exposure to environmental toxins and motor neuron disease: a case report. Med. Lav. 86:522, 1995. 213. Von Hoff, D. D., Reichert, C. M., Cuneo, R., et al.: Demyelination of peripheral nerves associated with cisdiamminedichloroplatinum (II) (DDP) therapy. Proc. Annu. Meeting Am. Assoc. Cancer Res. 63:1527, 1979. 214. Von Hoff, D. D., Schilsky, R., Reichert, C. M., et al.: Toxic effects of cis-dichlorodiammineplatinum(II) in man. Cancer Treat. Rep. 63:1527, 1979. 215. Waldron, H. A.: The anaemia of lead poisoning: a review. Br. J. Ind. Med. 23:83, 1966. 216. Walsh, J. C.: Gold neuropathy. Neurology 20:455, 1970. 217. Walton, J., and Buckley, I. K.: The lead-poisoned cell: a fine structural study using cultured kidney cells. Exp. Mol. Pathol. 27:167, 1977. 218. Whitfield, C. L., Chien, L. T., and Whitehead, J. D.: Lead encephalopathy in adults. Am. J. Med. 52:289, 1972.

2551

219. Wilke, R. A., Kolbert, C. P., Rahimi, R. A., and Windebank, A. J.: Methylmercury induces apoptosis in cultured rat dorsal root ganglion neurons. Neurotoxicology 24:369, 2003. 220. Wilson, S. A. K.: The amyotrophy of chronic lead poisoning: amyotrophic lateral sclerosis of toxic origin. Rev. Neurol. Psychiatry 5:441, 1907. 221. Wiltshaw, E., and Kroner, T.: Phase II study of cis-dichlorodiammineplatinum(II) (NSC-119875) in advanced adenocarcinoma of the ovary. Cancer Treat. Rep. 60:55, 1976. 222. Windebank, A. J.: Specific inhibition of myelination by lead in vitro: comparison with arsenic, thallium, and mercury. Exp. Neurol. 94:203, 1986. 223. Windebank, A. J.: Metal neuropathy. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1549, 1993. 224. Windebank, A. J., and Dyck, P. J.: Localization of lead in rat peripheral nerve by electron microscopy. Ann. Neurol. 18:197, 1985. 225. Windebank, A. J., McCall, J. T., Hunder, G. G., and Dyck, P. J.: The endoneurial content of lead related to the onset and severity of segmental demyelination. J. Neuropathol. Exp. Neurol. 39:692, 1980. 226. Woltman, H. W.: Neuritis and other diseases of peripheral nerves. In Tice, F., and Sloan, L. H. (eds.): Practice of Medicine, Vol. 9. Hagerstown, MD, W. F. Price Co., p. 289, 1962. 227. Wu, P. B., Kingery, W. S., and Date, E. S.: An EMG case report of lead neuropathy 19 years after a shotgun injury. Muscle Nerve 18:326, 1995. 228. Wyllie, J., Petermann, H., and Petermann, E.: Effect of penicillamine in promoting lead excretion. Can. Med. Assoc. J. 88:1155, 1963. 229. Zheng, J. L., Stewart, R. R., and Gao, W. Q.: Neurotrophin-4/5 enhances survival of cultured spiral ganglion neurons and protects them from cisplatin neurotoxicity. J. Neurosci. 15:5079, 1995.

114 Neuropathy Caused by Drugs STEVEN HERSKOVITZ AND HERBERT H. SCHAUMBURG

Almitrine Clinical Features Electrophysiology Pathology Amiodarone Clinical Features Electrophysiology Pathology Experimental Studies Chloroquine Clinical Features Electrophysiology Pathology Experimental Studies Colchicine Clinical Features Electrophysiology Pathology Experimental Studies Dapsone Clinical Features Electrophysiology Pathology Experimental Studies Disulfiram Clinical Features Electrophysiology Pathology Experimental Studies Ethambutol Clinical Features Electrophysiology Experimental Studies Isoniazid Clinical Features Electrophysiology Pathology Experimental Studies

Metronidazole Clinical Features Electrophysiology Pathology Experimental Studies Misonidazole Clinical Features Electrophysiology Pathology Experimental Studies Nitrofurantoin Clinical Features Electrophysiology Pathology Experimental Studies Nitrous Oxide Clinical Features Electrophysiology Pathology Experimental Studies Nucleoside Analogues Clinical Features Electrophysiology Pathology Experimental Studies Perhexilene Clinical Features Electrophysiology Pathology Experimental Studies Phenytoin Clinical Features Electrophysiology Pathology Experimental Studies Podophyllin Clinical Features Electrophysiology

New pharmaceutical agents are constantly identified or implicated as causes of human peripheral neuropathies; most such drugs appear to produce distal axonopathy, usually after prolonged use. In many cases, drugs are estab-

Pathology Experimental Studies Pyridoxine Clinical Features Electrophysiology Pathology Experimental Studies Suramin Clinical Features Electrophysiology Pathology Experimental Studies Tacrolimus Clinical Features Electrophysiology Pathology Experimental Studies Taxanes Clinical Features Electrophysiology Pathology Experimental Studies Thalidomide Clinical Features Electrophysiology Pathology Experimental Studies Tryptophan Clinical Features Electrophysiology Pathology Experimental Studies Vinca Alkaloids Clinical Features Electrophysiology Pathology Experimental Studies

lished or assumed to produce neuropathy by impairing axonal transport or other cell functions. There is more individual variation in nervous system vulnerability to some classes of drugs (antibiotics, anticancer agents) than 2553

2554

Diseases of the Peripheral Nervous System

to most occupational or environmental toxins. Several factors underlie this enhanced vulnerability. For one, persons exposed to occupational toxins are generally in good health with normal alimentary function, whereas individuals with acute systemic illnesses or chronic conditions (diabetes, asymptomatic hereditary neuropathy) receiving drugs may display increased vulnerability.67 There are relatively few careful experimental studies of the neurotoxicity of pharmaceutical agents, and satisfactory animal models exist for only a handful. Clinical reports are often the sole basis for many of the alleged drug-induced peripheral neuropathies. Some instances doubtless reflect peripheral nervous system dysfunction secondary to other, coincident conditions. This chapter discusses only those pharmaceutical conditions consistently associated with neuropathy.

ALMITRINE Almitrine bimesylate is a peripheral chemoreceptor agonist used as a respiratory stimulant. It is not approved for use in the United States.

AMIODARONE Amiodarone is a diiodinated benzofuran derivative initially developed in the 1960s as an antianginal agent, and later as a cardiac antiarrhythmic agent in the management of supraventricular and ventricular arrhythmias. It is used for the treatment of refractory, life-threatening ventricular arrhythmias. It is a class III antiarrhythmic agent and shares the capacity to prolong the duration of action potentials and refractoriness in Purkinje fibers and ventricular muscle.301 Neurotoxic side effects consist principally of a sensorimotor polyneuropathy, tremor, and ataxia. Infrequently, myopathy, optic neuropathy, basal ganglia dysfunction, encephalopathy, and pseudotumor cerebri occur. The exact mechanism of amiodarone tissue toxicity is uncertain. Like other amphiphilic cationic drugs such as perhexilene maleate, amiodarone forms intralysosomal lipid complexes leading to the inclusions observed in multiple tissues. An immunologic mechanism may explain observations such as early-onset toxicity, possible steroid responsiveness in pulmonary toxicity, and various immune perturbations.381

Clinical Features Clinical Features Almitrine therapy is associated with a sensory distal axonopathy. Numbness and pain may appear in the distal lower extremities after 2 months of therapy following conventional doses of 150 mg/day, or 12 months at 75 mg/day. Pain is severe and disabling.33,34,71,136 The upper limbs are rarely affected. Examination discloses diminished perception of pinprick, touch, and vibration senses in a stocking distribution; proprioception is usually spared and weakness is not prominent. Recovery is very slow following drug withdrawal.

Electrophysiology The amplitudes of both sensory and motor action potentials are diminished, consistent with axonal loss. Conduction velocities are usually normal.71,136

Pathology Sural nerve biopsies show degeneration of large-diameter fibers with sparing of unmyelinated axons.71,136 The mechanism of almitrine neuropathy is unknown. Rats treated for prolonged periods with diflurobenzhydrylpiperadine, a detrianzyl metabolite of almitrine, reportedly develop a spastic or ataxic gait. Lamellated and crystalloid bodies are present in sensory neurons, satellite cells, Schwann cells, endothelial cells of thoracic and lumbar dorsal root ganglia (DRGs), and hind limb muscles and muscle spindles.406,407

The reported incidence of neurotoxicity is widely variable, ranging from 5% to 74% in large series.381 Peripheral neuropathy correlates poorly with daily dose, total dose, or treatment duration. Most reported cases are on moderate to large doses for months to years, but symptoms may occur on as little as 200 mg/day and with durations of 1 month.205,278 The clinical presentation of neuropathy from chronic amiodarone therapy is subacute to chronic, symmetrical sensorimotor polyneuropathy with distal predominance; some reports note proximal weakness and asymmetry, and occasionally the neuropathy is predominantly motor.1,5,65,120,143,212,235,246,278,288 Evolution may be rapid, mimicking Guillain-Barré syndrome. Glove-stocking loss of all sensory modalities, depressed reflexes, and an ataxic gait are seen. There is one report of autonomic neuropathy.231 The cerebrospinal fluid (CSF) shows normal or mildly elevated protein without pleocytosis.278 A case of acute suicidal self-administration of 8 g of oral amiodarone resulted in profuse sweating, prolonged Q-T interval, and sinus bradycardia, but no peripheral neuropathy during 3 months of follow-up.13 A few patients have developed a clinical myopathy, with or without neuropathy, with a proximal pattern of weakness and myalgia.49,78,120,246,288 Tremor, with or without neuropathy, is the most frequent neurotoxic side effect of amiodarone therapy, occurring in 43% and 39% of patients in two large series.65,273 A 6- to 10-Hz action tremor develops early in the course of therapy; it is bilateral and symmetrical, and clinically indistinguishable from essential tremor.65 Gait

2555

Neuropathy Caused by Drugs

ataxia is also common, occurring in 7% and 37% of cases in these same series. It is not always attributable to neuropathy, and may be associated with limb ataxia and other findings that suggest cerebellar dysfunction.1,65 Optic neuropathy may occur; it is characterized by insidious onset and slow progression of bilateral visual loss, usually associated with optic disc swelling.119,225,232,262 Additional suggested neurotoxicity, based on anecdotal case reports, includes parkinsonian features of rest tremor, bradykinesia, and rigidity.388 Other evidence of nervous system dysfunction includes myoclonus, hemiballism, and dyskinesia273,388; encephalopathy with electroencephalographic slowing273; and pseudotumor cerebri.32 All are reversible with discontinuation of the drug. Both mild and severe cases of neuropathy improve with lowering the dose or discontinuing the drug235; persistent disability may remain but is rare.5 Improvement occurs over weeks to months.

Electrophysiology Reports of nerve conduction studies and needle electromyography (EMG) vary, with some describing predominant axonopathy with reduced amplitudes and distal denervation, and others prominent conduction slowing suggesting primary demyelination.5,124,175,235,246,278 The demyelinating electrophysiologic features can be multifocal and mimic chronic inflammatory demyelinating polyneuropathy.143 Needle EMG may show myopathic features in instances of myopathy.49,78,120,246,288

Pathology Sural nerve biopsy specimens display either predominant axonal degeneration affecting fibers of all sizes,246 predominant or almost pure demyelination,175 or a mixture.288 Numerous lysosomal inclusions are present in Schwann cells, axoplasm, fibroblasts, capillary endothelial and perithelial cells, and perineural cells.1,175,212,246,278,288 In one reported case with peripheral neuropathy, concentrations of amiodarone and desethylamiodarone in a sural nerve biopsy were 80 times higher than in serum.124 Examination of a retrobulbar optic nerve enucleated for melanoma demonstrated intracytoplasmic lamellar inclusions in large axons.232 Muscle biopsies may reveal vacuolar changes, lipid inclusions, marked accumulation of amiodarone and desethylamiodarone, or denervation atrophy.49,78,120,246,288 One report described necrotizing myopathy.78

Experimental Studies Rats and mice fed large doses of amiodarone for up to 90 days develop dose-related weight loss, decreased motor activity, weakness, and tremor.83 Cytoplasmic lysosomal

lipid inclusions are widespread but excluded from regions with intact blood-brain and blood-nerve barriers. Inclusions appear in many cell types, including autonomic ganglia and DRGs, myenteric plexus, and centrally in the area postrema, choroid plexus, ocular tissues, and pituitary. Increased blood-nerve permeability by sciatic nerve crush results in short-lived appearance of inclusions but does not affect the rate of axonal regeneration. Rats treated with amiodarone intraperitoneally at 20 mg/kg/day for up to 6 weeks develop incoordination and increased pain thresholds, suggesting the appearance of peripheral neuropathy.295 Amiodarone injected endoneurially into rat tibial nerve at low concentrations (25 ␮g/mL) produces electrophysiologic evidence of demyelination with conduction block and pathologic evidence of segmental demyelination.318 Higher concentrations result in increasingly severe axonal degeneration; this implies that the different pathologic changes noted in human neuropathy may be related to variable blood-nerve barrier efficacy, leading to different drug concentrations in nerve. Chronic administration of amiodarone to mice also produces a myopathy with autophagic vacuolation and phospholipid inclusions; denervation induces necrosis of mainly type II fibers and a significant increase in amiodarone and desethylamiodarone concentrations in muscle.84 Cytoplasmic lipid inclusions, similar to those in humans, can be reproduced experimentally in many ocular cell types of rats following both local application and oral administration of amiodarone.31 Desethylamiodarone, a metabolite of amiodarone, can induce lipidosis in rat alveolar macrophages.297 Amiodarone inhibits lysosomal phospholipases in rat lung tissue.157 Studies in rat brain synaptosomes suggest that neurotoxicity might be mediated by effects on ATPase activity or calcium homeostasis.192,296 A study of grape seed proanthocyanidins indicated protection against amiodarone-induced pulmonary toxicity in mice.15

CHLOROQUINE Chloroquine is a 4-aminoquinoline derivative used as an antimalarial, and, at higher doses, as an anti-inflammatory agent in the treatment of immune-mediated connective tissue and dermatologic disorders. Hydroxychloroquine, a related compound, is also effective with similar though lesser toxicity. The most serious adverse effect of chloroquine is dose-related retinopathy by damage to the retinal pigment epithelium.248,326 Central nervous system (CNS) toxicity may include psychotic symptoms, encephalopathy, acute extrapyramidal syndrome with torticollis and dystonia in children, and ototoxicity. Peripheral nervous system toxicity includes myopathy, polyneuropathy, and a myasthenialike syndrome. Chloroquine is a cationic, amphiphilic drug and, like amiodarone and perhexilene, can cause lysosomal damage.

2556

Diseases of the Peripheral Nervous System

Clinical Features Myopathy and/or neuropathy tend to occur after long-term administration (months to years) of chloroquine at high doses, in the 250- to 750-mg/day range. Individuals develop painless, progressive, symmetrical proximal limb weakness.115,352,385,389 Occasionally there are paresthesias in the feet. Quadriceps muscles may be particularly weak. Deep tendon reflexes are depressed or absent. Sensory examination is normal or shows some distal vibratory loss or hypoesthesia. Some individuals develop respiratory failure or cardiomyopathy.115,334 The serum creatine kinase (CK) is normal or mildly elevated. One report describes an elevated CSF protein of 89 mg/dL, causing diagnostic confusion with chronic inflammatory demyelinating polyneuropathy.385 Patients typically improve substantially over months with drug withdrawal. Persons receiving chloroquine sometimes develop a myasthenia-like syndrome, including fatigable ptosis, extraocular muscle paresis, bulbar dysfunction, proximal weakness, and response to edrophonium.300,337 One case, in which symptoms appeared rapidly after introduction of chloroquine and resolved rapidly after drug withdrawal with no detectable acetylcholine receptor antibody, suggests a direct drug effect on neuromuscular transmission. Another case evolved and improved over weeks to months; this was associated with elevated acetylcholine receptor antibody titer that subsequently normalized, suggesting that chloroquine may provoke reversible immune system derangement, similar to penicillamine. In both cases, symptoms recurred with brief reintroduction of chloroquine.

Electrophysiology Results of nerve conduction studies range from normal to findings consistent with a sensorimotor axonopathy; prominent demyelinating features are not described.115,350,385 In the myasthenia-like cases, abnormalities are demonstrable either with decrement in proximal and facial muscles on repetitive stimulation, or by jitter on stimulated single-fiber EMG.300,337 Needle EMG findings reflect mostly the myopathy, with myopathic motor unit and recruitment abnormalities, usually in association with spontaneous activity including fibrillations, positive sharp waves, and complex repetitive or myotonic discharges.

and myeloid bodies in the presynaptic portion of the axons in neuromuscular junctions.337 Similar inclusions are seen with amiodarone and perhexilene. Overall, the morphologic studies suggest primary involvement of the Schwann cells.272,362 Muscle biopsies usually show a vacuolar myopathy with intracellular myeloid or curvilinear bodies,115,337 and occasionally neurogenic atrophy. Endomyocardial biopsies in patients with cardiomyopathy have similar findings. Reports of follow-up biopsies months after drug withdrawal describe resolution of vacuolar changes, though curvilinear bodies persist.

Experimental Studies Chloroquine administered to rats produces a vacuolar myopathy with inclusions resembling those in humans, as well as myeloid bodies in Schwann cells and within the axoplasm of myelinated intermuscular nerves, and in terminal nerve endings at the end-plate region.226 Chloroquine and hydroxychloroquine accumulate in lysosomes and cause a rise in intralysosomal pH. It is suggested that this inhibits acidic lysosomal hydrolases, with resultant accumulation of phospholipid, glycogen, and curvilinear bodies.115 Coadministration of the potent cysteine protease inhibitor EST and chloroquine prevents induction of chloroquine myopathy in the rat.353

COLCHICINE Colchicine is a tricyclic alkaloid extracted from the meadow saffron or autumn crocus plant (Colchicum autumnale); it has been employed for centuries as an anti-inflammatory agent in the treatment of gouty arthritis. In addition to acute treatment and prophylaxis of gout, it has found use in the treatment of familial Mediterranean fever, Behçet’s syndrome, and primary biliary cirrhosis, among many other conditions. Colchicine can cause combined myopathy and polyneuropathy in patients on long-term therapy with underlying renal insufficiency. The mechanism of the neuropathy is held to be defective axonal transport resulting from colchicine binding to tubulin, impairing microtubule assembly.276

Clinical Features Pathology Sural nerve biopsies show mild to severe loss of myelinated fibers, with segmental demyelination and remyelination.115,272,362 Cytoplasmic dense and lamellated/myeloid inclusions are present on electron microscopy in Schwann, perineurial, and endothelial cells, as are occasional curvilinear profiles. One report of a motor point biopsy described membranous bodies in the intramuscular nerve axoplasm

Myopathic features predominate.91,103,198,199,293,309,360,413 Patients present with occasionally subacute (over 1 to 2 weeks) but more commonly chronic (over months) progressive proximal weakness, distal areflexia, and mild distal sensory loss to pinprick and vibration. They may have prominent myalgia, and some have paresthesias. An occasional patient has percussion and grip myotonia.99 Typically, individuals are taking customary doses in the range of

2557

Neuropathy Caused by Drugs

0.6 mg once or twice daily, and have at least mild chronic renal insufficiency, which likely results in elevated colchicine blood levels. The serum CK level is invariably elevated, from mildly to as much as 44-fold, and tends to parallel the severity of disease. Sometimes the CK is elevated unaccompanied by symptoms. This clinical presentation is most frequently misdiagnosed as polymyositis, or uremic myopathy and neuropathy. In addition to some helpful electrophysiologic and pathologic features, the clinical course aids differentiation, because weakness and elevated CK resolve over about 4 weeks after drug withdrawal; mild neuropathic signs remain. One report described a multisystem syndrome resulting from repeated acute colchicine intoxications from a bizarre self-dosing schedule over 5 years. Neuropathic features were severe, with prominent distal upper more than lower extremity weakness, severe large fiber sensory loss, and areflexia.299

Electrophysiology Nerve conduction studies are consistent with a mild axonal sensorimotor polyneuropathy, with reduced sensory and motor amplitudes as predominant features.200 Needle EMG of proximal muscles shows myopathic features, with lowamplitude, short-duration, polyphasic motor units. Early recruitment and associated spontaneous activity, including fibrillations, positive sharp waves, and complex repetitive discharges, are present. Myotonic discharges may be prominent and can be a helpful differentiating feature.309 All these EMG abnormalities may dissipate, along with weakness, after drug withdrawal. Needle EMG of distal muscles may also show spontaneous activity, but with motor unit potentials that are large amplitude, long duration, and polyphasic, with neuropathic recruitment patterns, indicating chronic denervation. These changes usually persist.

longitudinal geometry; myotubes fragment and myoblasts are arrested at metaphase.140 The pathogenesis of the axonopathy is likely related to defective axonal transport as a result of impaired microtubule assembly.276 Myopathy is also suggested to result from disruption of a cytoskeletal microtubular network.198 In an attempt to develop a model of colchicine neuropathy, rats were injected intraperitoneally with 0.2 mg/kg for 5 days per week for 7 months; they demonstrated significant gait abnormalities at 9 months on a walking track analysis.59 Muscle biopsies were unremarkable, and morphometric analysis of sciatic nerve biopsies only showed a statistically significant difference for the mean axon:myelin ratio compared to controls.

DAPSONE Dapsone (4,4⬘-diaminodiphenylsulfone) is a sulfone derivative used as an antibacterial or anti-inflammatory agent in the treatment of leprosy, malaria, Pneumocystis carinii pneumonia in acquired immunodeficiency syndrome, Crohn’s disease associated with Mycobacterium paratuberculosis, dermatitis herpetiformis, and rheumatoid arthritis.149 It is metabolized in the liver by N-acetylation and N-hydroxylation. The most common adverse effect is dose-dependent hematologic toxicity, including methemoglobinemia, sulfhemoglobinemia, and hemolytic anemia, particularly in the presence of glucose-6-phosphate dehydrogenase deficiency. Uncommon for a toxic neuropathy, dapsone produces an axonal, primarily motor neuropathy. It is suggested that this reflects a “dying back” of motor axons.150 The mechanism and exact site of neurotoxicity, and the reason for the predilection for motor axons, are unestablished.

Clinical Features Pathology Sural nerve biopsies are nonspecific, with mild to marked fiber loss and degenerating and regenerating axons.91,198,299 Biopsies of proximal muscles demonstrate a distinctive vacuolar myopathy, with a profuse accumulation of lysosomes and autophagic vacuoles.198 There are perinuclear aggregates of densely packed filamentous material, similar to the perinuclear caps of intermediate filaments seen in other eukaryotic cells treated with antimicrotubule agents.169 Distal muscles may also show mild denervation atrophy.

Experimental Studies Colchicine induces a myopathy with similar pathologic features in animals, as does vincristine, another microtubule disaggregator associated with a similar neuropathy.39,335 In vitro, colchicine causes developing muscle fibers to lose their

Dapsone neuropathy is uncommon and not clearly dose dependent.150,193,345 It is rare at the conventional dose for leprosy of 100 mg/day, and most cases are taking regimens ranging from 200 to 600 mg daily, for periods of as little as 2 weeks to many years. One patient developed neuropathy after 16 years on 100 mg/day.382 Acute suicidal ingestion of a massive dose (6 g) produced footdrop within 5 days.261 Slow acetylators may be at risk for developing neurotoxicity. The typical clinical presentation is subacute to chronically progressive, symmetrical distal weakness and atrophy, with no fasciculations, no upper motor neuron signs, and few sensory symptoms or signs. Weakness may involve predominantly the distal legs, both the distal arms and legs, or, unusual for a toxic neuropathy, just the distal arms; in one reported case, mostly proximal muscles were weak. Some asymmetry may be present.345 Reflexes are usually normal, or ankle jerks may be absent. Mild numbness or paresthesias are reported in a few patients, and sensory loss may include

2558

Diseases of the Peripheral Nervous System

just vibration or all modalities.114,150,193,261,400 There is a single report of optic neuropathy.170 Neuropathy invariably improves or resolves within months to a year of discontinuing the drug or markedly reducing the dose. Initial improvement in some cases has been rapid, even occurring after 1 week.193

Electrophysiology Nerve conduction studies are usually characterized by lowamplitude or absent compound muscle action potentials in involved areas, with velocities that are normal or mildly reduced as a result of loss of fast axons, and normal sensory potentials.150,193 Needle EMG of involved muscles shows spontaneous activity and motor unit loss; with time, enlarged, reinnervated motor units appear. The findings are compatible with a predominantly motor axonopathy. Serial studies over 5 years following drug withdrawal show progressive enlargement of compound muscle action potentials. Only one report documented low-amplitude sural sensory potentials.261

Pathology Sural nerve biopsy performed in one patient 30 days after acute suicidal ingestion of 6 g of dapsone demonstrated axonal degeneration.261 In another case with associated prominent sensory symptoms and signs and reduced sensory amplitudes, sural nerve biopsy showed axonal degeneration; it is unclear if this patient with dermatitis herpetiformis had a sensorimotor neuropathy associated with celiac sprue.105 A biopsy of the motor nerve supplying the extensor digitorum brevis muscle was performed in an 18-year-old female with leprosy and bilateral foot weakness after 5 months of dapsone at 200 mg/day.345 Ultrastructural studies suggested primary degeneration of the myelinated axons. The superficial peroneal sensory nerve in the same patient was normal on light microscopy.

a 5- to 10-fold rise in the blood concentration of acetaldehyde, causing unpleasant, and occasionally severe or even fatal, features of the disulfiram-ethanol reaction.113,190 Independent of this reaction, disulfiram has both central and peripheral neurotoxicity, principally producing an axonal sensorimotor polyneuropathy. Metabolites of disulfiram include diethyldithiocarbamate and carbon disulfide, which is suggested as the neurotoxic agent.

Clinical Features The neuropathy is dose dependent and progressive with continued treatment, appearing at or above the usual maintenance doses of 250 to 500 mg daily.89,104,129,255,256,270,304,386 Symptoms usually begin within a few months, though they may appear as quickly as 10 days with high doses.378 Initial symptoms include acral numbness and paresthesias, sometimes pain, and unsteady gait, followed by distal weakness, with evolution to include the hands. There is involvement of all sensory modalities; diminished or absent tendon reflexes and foot dorsiflexion weakness are common. Severe cases may have tetraparesis and facial weakness.330 Slow improvement over months follows drug withdrawal, but may be incomplete depending upon severity of axon loss. With or without associated polyneuropathy, some patients develop encephalopathy, psychosis, parkinsonism, and dystonia after acute high doses or subacute treatment regimens, with associated basal ganglia lesions on neuroimaging.147,195,203

Electrophysiology Nerve conduction studies are characteristic of an axonal sensorimotor polyneuropathy, with reduced sensory and motor amplitudes and normal or slightly reduced conduction velocities. Needle EMG shows distal denervation.270,386 Asymptomatic chronic alcoholics taking 250 mg, but not 12 mg, of disulfiram daily may display progressive, subclinical electrophysiologic abnormalities.274

Experimental Studies There is no experimental animal model of dapsone neuropathy. Attempts to reproduce nerve toxicity in rats, mice, and guinea pigs have failed.150,180,393 Dapsone binds to bovine brain tubulin with inhibition of microtubule assembly, a possible mechanism of neurotoxicity.291

DISULFIRAM Disulfiram (tetraethylthiuram disulfide) is an inhibitor of the enzyme aldehyde dehydrogenase, and has been used since the late 1940s for alcohol aversion therapy in chronic alcoholism.131 Alcohol ingestion by someone who has taken disulfiram, even up to 2 weeks earlier, results in

Pathology In most cases, sural nerve biopsies feature primary axonal degeneration with loss of large and small myelinated fibers, and occasionally also unmyelinated fibers.8,23,36,266,330,386 One report described segmental demyelination, suggesting toxicity to Schwann cells as well.266 Swollen myelinated and unmyelinated axons are present with accumulations of neurofilaments.8,23,27,330

Experimental Studies There are no well-established animal models of disulfiram neuropathy; one report described axonal degeneration in rats.9 Because carbon disulfide is a metabolite of disulfiram

2559

Neuropathy Caused by Drugs

and produces a neuropathy with neurofilamentous axonal swellings, this is held as the mechanism of disulfiram toxicity. However, in the most recent study of disulfiram in rats, the drug failed to produce either the morphologic changes or the intermolecular protein cross-linking characteristic of carbon disulfide.367 The predominant changes consisted of vacuoles within the Schwann cell cytoplasm and segmental demyelination. Other studies, in vitro, suggest mitochondrial injury or inhibition of DNA topoisomerases as mechanisms of toxicity.16,402 The effect of disulfiram on rat brain biogenic amines—inhibition of dopamine ␤-hydroxylase—may explain its central effects.185

ETHAMBUTOL Ethambutol is a derivative of ethylenediamine, a chelator of zinc and other metal ions. It is used to treat tuberculosis in combination with other agents. The most serious side effect of ethambutol is retrobulbar optic neuropathy; peripheral neuropathy is rare.

Clinical Features Peripheral neuropathy is usually associated with prolonged dosage in excess of 20 mg/kg/day; it mostly appears in the elderly, in those with alcoholism or diabetes, and in subjects with renal impairment. Initial symptoms are numbness and tingling in the feet and may precede the onset of visual disturbance; weakness is rare.259,370 Some individuals report incoordination of fine movements of the fingers. Signs include diminished position and vibration senses in the toes, with less impairment of pinprick and touch senses; ankle jerks are diminished or absent. Muscle wasting and weakness are present in extreme cases. Recovery from neuropathy generally occurs following drug withdrawal; recovery from optic neuropathy is more variable, especially in advanced cases.

Electrophysiology There is little information concerning the electrophysiologic changes save for reduction in sensory nerve potentials and mild conduction slowing. There are no reports of nerve biopsies.

stores of zinc. Several in vitro studies suggest that ethambutol exerts its retinal toxicity through a glutamatedependent mechanism, and that this effect can be blocked by pretreatment with a caspase inhibitor.160,236,326

ISONIAZID Isoniazid (INH), the hydrazide of isonicotinic acid, has been used in the prophylaxis and treatment of tuberculosis since its fortuitous discovery in the early 1950s.229 Its propensity to cause an axonal sensorimotor polyneuropathy was soon recognized, particularly in susceptible individuals having the slow acetylator genotype, an autosomal recessive trait.25,26,173,179,404 The standard dose of INH is 3 to 5 g/kg/day. It is acetylated in the liver before being excreted in the urine. Slow acetylators maintain high blood levels of INH for longer periods. INH inhibits pyridoxal phosphokinase, preventing required phosphorylation of the vitamin B6 group (pyridoxine, pyridoxal, and pyridoxamine), and also chelates with pyridoxal phosphate. This pyridoxine-deficient state is presumably responsible for the toxicity; the neuropathy is preventable with 10 to 50 mg/day of pyridoxine.30,186,303,340,343 Larger doses of pyridoxine for prolonged periods may also cause sensory neuropathy.

Clinical Features The neuropathy is a dose-related, chronic effect, occurring in about 2% of individuals on standard doses and 17% of those on 6 mg/kg/day, and increasing in incidence with higher doses.229 Symptoms appear after about 6 months on low doses and within a few weeks on high doses. Initial symptoms are sensory, with numbness and tingling of the feet, followed by complaints of weakness or unsteady gait. Aching cutaneous pain in the calf muscles is common. There is distal loss of touch, pain, temperature, and vibration sensations, and distal reflexes are diminished. Distal weakness and atrophy can be later features in more severe cases. Recovery after drug withdrawal is rapid in mild, early cases, but can take months to years in more severe instances. Pyridoxine treatment does not affect the rate of recovery.30,134 Severe optic neuropathy can occasionally occur.35

Electrophysiology Experimental Studies Experimental studies in rats describe nonspecific axonal degeneration in the sciatic nerve and axonal swellings in the optic nerves.213,236 A primate study utilizing prolonged dosage produced nerve fiber degeneration in the optic and pyramidal tracts as well as chromatolysis in the anterior horn cells of the lumbar cord.327 Ethambutol is widely held to exert its toxic effects by binding and lowering the body

The electrophysiologic features of INH-associated neuropathy have not been adequately characterized.

Pathology Sural nerve biopsy and postmortem studies clearly establish INH neuropathy as a distal axonopathy. There is axonal degeneration of myelinated and unmyelinated

2560

Diseases of the Peripheral Nervous System

axons along with regeneration in sural nerve biopsies, and axonal degeneration in both peripheral nerve and dorsal columns along with muscle denervation atrophy at autopsy.30,267

Experimental Studies Neuropathy similar to that seen in humans is readily reproduced in the rat with either acute large or chronic small doses. There is dose-dependent distal sensory and motor axon degeneration, as well as fiber loss in the rostral gracile tracts.24,55,76,176 In one study, whorls, extracellular debris, macrophages, and reduced numbers of myelinated axons were the pathologic features seen following chronic INH administration to rats.328 Sequential morphometric studies of pathologic changes suggest that, prior to axonal degeneration, there are focal accumulations of axonal organelles, particularly smooth endoplasmic reticulum.268 Chronic INH intoxication in dogs and birds causes a vacuolar leukoencephalopathy.30 In a culture of mouse DRG neurons, hydrazine, a metabolite of INH, rather than INH itself, was the predominant toxic moiety.317

white matter, corpus callosum, and inferior basal ganglia and around the fourth ventricle.2,171 A predominantly sensory polyneuropathy may follow large, cumulative doses37,85,208,316,414; one patient had only intermittent-dose therapy of 2 g/day for 5 days every other month.98 A case with associated limb ataxia and generally absent sensory nerve action potentials suggests that this may represent a sensory neuronopathy.151 A maximum daily dose of 800 mg is well tolerated in long-term controlled trials of patients with Crohn’s disease.349 In another study, 1200 mg/day for 2 to 4 weeks resulted in subclinical but statistically significant nerve conduction changes.184 The prevalence of abnormal neurologic exams or nerve conduction studies was 85% in a study of long-term metronidazole in pediatric Crohn’s disease.102 Clinical features include distal paresthesias; pain; little or no weakness; glove-stocking sensory loss to touch, pinprick, and vibration; and occasionally absent ankle jerks. Symptoms and signs usually improve or resolve within weeks or months after drug discontinuation, but occasionally persist for years.37

Electrophysiology METRONIDAZOLE Metronidazole is a nitroimidazole derivative used as an antimicrobial in the treatment of protozoal and anaerobic bacterial infections, and in Crohn’s disease.363 It is effective in the treatment of trichomoniasis, amebiasis, giardiasis, and anaerobic bacterial infections, and as part of a regimen for Helicobacter pylori–associated peptic ulcer disease. It has also been used as a radiosensitizer. Neurotoxic adverse effects include encephalopathy, cerebellar dysfunction, and an axonal polyneuropathy.

Clinical Features The most common side effects are headache, nausea, xerostomia, a metallic taste, and occasionally vomiting, diarrhea, or abdominal discomfort. A disulfiram-like effect can follow alcohol ingestion. Single suicidal overdoses with up to 15 g result in nausea, vomiting, and ataxia; doses from 3.6 to 19.5 g are tolerated without significant morbidity.363 Neurotoxicity with long-term metronidazole is related to cumulative dose and duration of treatment. Most reported cases received weeks to months of therapy. Reversible CNS dysfunction may include lethargy, disorientation, confusion, dysarthria, and cerebellar ataxia.2,171,201 High cumulative doses cause seizures.130,152 One study of 37 patients receiving high-dose metronidazole as a radiosensitizer reported a 25% incidence of dizziness, tremors, ataxia, and confusion.374 Magnetic resonance imaging in two cases with encephalopathy and ataxia showed reversible T2-weighted hyperintensities in the cerebellar nuclei, supratentorial

Nerve conduction studies suggest sensory axonal degeneration, with low-amplitude or absent sensory potentials and minimal, if any, involvement of motor fibers. A small fiber profile may predominate occasionally, with normal nerve conduction studies and abnormal quantitative sensation testing (QST) or quantitative sudomotor axon reflex testing.414

Pathology Sural nerve biopsies from patients with sensory polyneuropathy, including teased fiber studies and electron microscopy, demonstrate primary axonal pathology with degeneration of both myelinated and unmyelinated fibers.40,151,316,358 Occasionally, mild loss of small myelinated axons predominates.380 There are no human CNS studies.

Experimental Studies Long-term, high-dose metronidazole in dogs produces Purkinje cell loss and seizures.321 In rats, very large doses result in behavioral changes and histologic lesions in the brainstem and cerebellum similar to those of Wernicke’s encephalopathy.380 Long-term studies with lower doses approximating those used in humans failed to show degenerative changes in the central or peripheral nervous system.40 14C-labeled metronidazole accumulates in the cerebellum and hippocampus of mice.280 The mechanism of neurotoxicity is unclear. Suggested mechanisms include inhibition of neuronal protein synthesis by binding to RNA,40 and thiamine antagonism.4

2561

Neuropathy Caused by Drugs

MISONIDAZOLE Misonidazole, a 2-nitroimidazole compound, is a hypoxic cell radiosensitizer used to sensitize radioresistant tumor cells. Neurotoxicity includes dose-limiting, predominantly sensory polyneuropathy and, at high doses, encephalopathy. A related compound, metronidazole (a 5-nitroimidazole compound), is also associated with sensory neuropathy.414

Clinical Features High doses of nitroimidazoles in humans can produce encephalopathy with lethargy, confusion, seizures, and, rarely, focal abnormalities.93,320 The incidence of peripheral neuropathy correlates with the total dose and fractionation schedule. It is frequent at a total dose of more than 18 g.373 Thirty-nine percent of patients receiving 11 g/m2 developed neuropathy.249,277 Clinical features are those of a predominantly sensory polyneuropathy affecting distal legs more than arms. Pain is common. There is distal loss of touch, pain, vibration, and position sense to varying degrees. Tendon reflexes are usually preserved.228,249,250 Mild distal weakness in the feet occurs in some cases.277 Improvement, sometimes incomplete, occurs over months after treatment is discontinued.228,249 It has been suggested that steroids or phenytoin might protect against the development of misonidazole neuropathy.250,277 Other studies indicate that phenytoin, despite causing increased misonidazole metabolism, is not protective.178 The metabolite desmethylmisonidazole also appears to be neurotoxic.319

Electrophysiology Nerve conduction studies are consistent with an axonal polyneuropathy; sural nerve amplitudes are prominently affected.228,249

Pathology Sural nerve biopsies show axonal degeneration of predominantly large fibers, with secondary demyelination.249

Experimental Studies Prolonged exposure to misonidazole in rats produces distal axonal degeneration, most severely in the sensory terminals of intrafusal fibers, and edema in DRGs.145 In the CNS, there are spongy changes progressing to necrosis and petechiae in the lateral and superior vestibular nuclei, superior olives, cochlear nuclei, and cerebellar roof nuclei. Changes in brainstem auditory evoked potentials correlate with the histopathologic findings.107 These changes show topographic and pathologic similarities to those seen in

thiamine-deficient rats. Somewhat analogous findings occur in short-term experiments in mice exposed to desmethylmisonidazole; tremors and seizures also may occur.63 Studies of nitroimidazole neurotoxicity in cultured neuronal cell lines suggest the mechanism may be disruption and degradation of the neurofilament lattice.352 Other postulated mechanisms include impaired membrane transport from reactions with sulfhydryl groups, and inhibition of protein synthesis.228 Although thiamine antagonism is suggested by the pathologic features, administration of thiamine does not prevent neuropathy in animals or humans.93,145

NITROFURANTOIN Nitrofurantoin is a water-soluble, synthetic hydantoin derivative used for prevention and treatment of urinary tract infections. Initial enthusiasm for nitrofurantoin treatment stemmed from the ready achievement of persistent high drug concentrations in the urine with only modest tissue and plasma levels. Serious pulmonary toxicity and axonal neuropathy now limit its use.

Clinical Features Sensorimotor neuropathy is a rare but clearly delineated feature of nitrofurantoin treatment.19,79,111,368,369 It is more common in the elderly, in females, and in individuals with renal dysfunction. However, renal dysfunction is not prerequisite. Neuropathy does not bear a strong relationship to dose or duration; it may appear between 7 and 336 days of treatment. Cumulative doses range from 1.4 to 67.0 g; in some individuals, neuropathy appears following cessation of therapy. Nitrofurantoin neuropathy is heralded by lower limb paresthesias followed by the rapid onset and progression of severe weakness and subsequent muscle wasting. The speed and progression of weakness are almost unique among the toxic neuropathies, and may suggest an erroneous diagnosis of Guillain-Barré syndrome. Some cases eventuate with footdrop, wristdrop, and clawed hands. Severity of weakness is not always dose related. Most subjects display a stockingglove sensory loss with especially profound large fiber dysfunction. Should medication be continued after commencement of weakness, neuropathy is incapacitating and poorly reversible. Recovery in such cases is poor; it is related to severity, not to dose.368

Electrophysiology Electrophysiologic changes clearly indicate an axonal neuropathy and are especially useful in differential diagnosis of cases resembling Guillain-Barré syndrome. Sensory action potentials are markedly reduced or absent. Motor conduction is mildly decreased, but compound action potentials

2562

Diseases of the Peripheral Nervous System

are unobtainable in severe cases. Electromyography shows evidence of widespread denervation in distal muscles with only mild changes in proximal muscles.111

Pathology Sural nerve biopsy in moderate cases discloses widespread acute axonal degeneration with no distinctive features; in two advanced instances there were no surviving nerve fibers.411 One autopsy study reported widespread axonal degeneration in peripheral nerves and spinal roots.217

Experimental Studies The mechanism underlying nitrofurantoin neuropathy is unclear. The mechanism most favored is inhibition of the synthesis of acetyl-coenzyme A in the citric acid cycle by furan derivatives. There are no convincing experimental animal studies that utilize contemporary histopathologic or neurophysiologic techniques in attempts to reproduce nitrofurantoin neuropathy. Existing studies in which extremely high doses of 20 to 100 mg/kg were administered to rodents claimed axonal changes after only 2 days of dosing.17,18

NITROUS OXIDE Nitrous oxide is an inorganic gas used as anesthetic in brief surgical or dental procedures and as a propellant in food (e.g., whipped cream) dispensers.314 Nitrous oxide inhalant abuse stems from its euphoriant effects; it is especially common in adolescents and among health care professionals.207,302 Repeated self-administration inactivates cobalamin and causes a myeloneuropathy identical to that of vitamin B12 deficiency. Nitrous oxide is widely available in small propellant cylinders and is not a controlled substance in North America.

Clinical Features Most reports describe single cases and strongly support a dose-effect relationship.148,206,207,302,313 Prolonged heavy exposure (inhalation of 100 to 200 propellant cartridges daily for 3 months) causes a mild myeloneuropathy syndrome that steadily worsens with continued abuse. Six months of this level of inhalation causes disabling sensory ataxia of the lower limbs. Less frequent exposure to lower levels has caused symptoms commencing only after 5 years. There are rare reports of individuals with subclinical B12 deficiency who develop myeloneuropathy following general anesthesia with nitrous oxide; symptoms usually commence within a month of the surgical procedure.234,242,324 It is unlikely that normal individuals are at risk for postsurgical myeloneuropathy, even following prolonged nitrous oxide anesthesia. Several studies claim that dentists with daily

exposure in an office setting can develop symptoms of myeloneuropathy. Although these reports suggest a doseresponse relationship, it is unclear if any subject had objective evidence of myelopathy or neuropathy. In addition, it is possible that some participants abused nitrous oxide.43,148 Symptoms usually appear following an episode of heavy abuse of nitrous oxide. The neurologic disorder resembles subacute combined degeneration of the spinal cord, although features of peripheral nerve disease are often more prominent. Initial symptoms include acral paresthesias, unsteady gait, and Lhermitte’s sign. Subsequently, hand clumsiness, leg weakness accentuated distally, impotence, bladder dysfunction, mood disturbance, and truncal numbness appear. Early signs often are only distal diminution of touch and vibratory perception and depressed lower limb tendon reflexes. Later, a combination of myelopathic and neuropathic signs is present, including Babinski’s signs, distal weakness, severe sensory gait ataxia, proprioceptive sensory loss, and hyperreflexia.207,302,313 Improvement follows withdrawal from exposure and can take from a few weeks to 6 months. Especially severe cases are left with residual proprioceptive loss and an unsteady gait. Serum vitamin B12 levels are frequently normal, as are Schilling tests, because most inhalant abusers do not experience dietary deficiency or malabsorption of B12. Presumably, levels of homocysteine and methylmalonic acid would reflect the effects of abnormal cobalamin, but clinical reports have not included determinations of these substances. Treatment with B12 and methionine is rational, but probably of little help if individuals continue the pattern of abuse because the inhaled nitrous oxide will continue to inactivate cobalamin stores.348

Electrophysiology Findings of nerve conduction studies and EMG are consistent with a predominantly distal, axonal sensorimotor neuropathy. Somatosensory evoked responses demonstrate prolonged or absent cortical potentials, more severe with tibial than median nerve stimulation. This may reflect myelopathic or lower extremity peripheral nerve dysfunction.164,223,379

Pathology There are no postmortem studies. Sural nerve biopsy in a disabled patient showed severe axonal depletion223; in a less advanced case, biopsy disclosed only axonal enlargement and nonspecific abnormalities within Schwann cell cytoplasm.313

Experimental Studies Nitrous oxide oxidizes the monovalent cobalt moiety of cobalamin (vitamin B12) to an inactive trivalent state.90,194 This causes a cobalamin-deficient state with reduction in

2563

Neuropathy Caused by Drugs

cobalamin-dependent enzyme activity (e.g., methionine synthetase) in humans and in experimental animals. Thus there are usually normal levels of serum B12 and no malabsorption as measured by the Schilling test. Experimental animal studies demonstrate species variability in vulnerability to nitrous oxide. Following inhalation exposure, myeloneuropathy appears in monkeys and fruit bats,238,332 but not in rats, although rodents display markedly diminished levels of methionine synthetase.106,194 Monkeys exposed to a continuous flow of 15% nitrous oxide for 10 weeks develop limb ataxia and progressive clumsiness of the extremities. Ultimately they are unable to stand. Histopathologic study convincingly demonstrates vacuolar myelopathy and axonal degeneration of the posterior and lateral columns of the spinal cord.238,332 These findings are virtually identical to those of subacute combined degeneration. The peripheral nerves of experimental animals have not been rigorously studied.

NUCLEOSIDE ANALOGUES Nucleoside analogue reverse transcriptase inhibitors (NRTIs) are an essential component of retroviral therapy. Three analogues—zalcitabine (dideoxycytidine; ddC), didanosine (dideoxyinosine; ddI), and stavudine (didehydrodeoxythymidine; d4T)—are used in treating human immunodeficiency virus (HIV) infection. Peripheral neuropathy is associated with each analogue; zalcitabine was the first NRTI used, and most studies are focused on its neurotoxic profile.21,100,122,251,399,409

Clinical Features Zalcitabine is associated with a painful sensory distal polyneuropathy that is sometimes dose limiting. Neuropathy is evident after about 2 months in all HIV-positive patients taking the highest dose (0.06 mg/kg every 4 hours for up to 12 weeks). The predominant presenting symptom is abrupt onset of burning or shooting pain in the feet; discomfort is extreme and disabling in some instances.21,251 Pain is soon followed by paresthesias and numbness; occasionally muscle cramps and weakness of foot dorsiflexion appear. Examination discloses distal symmetrical loss of temperature and light touch sensations accompanied by allodynia to touch. Vibration sense is relatively intact. Loss of the ankle jerk is usual, but other tendon reflexes are spared. The incidence of neuropathy is 33% with the current lower dose of 0.005 mg/kg. Significant recovery occurs in 75% of persons withdrawn from the higher dose level following a “coasting” period of intensification for months. Elevated risk at the lower dose occurs in persons with diabetes, in those with previous treatment with NRTIs, and in alcoholics. Differentiation between NRTI-induced neuropathy and

the painful neuropathy of HIV can be difficult; in general, the HIV neuropathies have a more gradual onset, and NRTI neuropathy improves following withdrawal. NRTIinduced neuropathy is rare in children with HIV infection.

Electrophysiology Nerve conduction studies in persons receiving the highest dose reveal absent or diminished sensory potentials in lower limb nerves; motor conduction and F waves are normal.

Pathology There are no nerve biopsy reports; one study demonstrated nerve fiber degeneration in skin punch biopsy specimens.239

Experimental Studies In vitro studies indicate that the mechanism of toxicity reflects mitochondrial dysfunction.72,73 It is suggested that NRTIs exert toxicity by directly inhibiting mitochondrial bioenergetic function.403 One study demonstrated that the immunophilin ligand FK-506 is neuroprotective in an in vitro model of zalcitabine-induced DRG cell neurotoxicity.188 There is no satisfactory experimental animal model of human NRTI neuropathy. Rabbits dosed for 18 weeks develop weakness and a severe demyelinating neuropathy. Nerve conduction is slowed, and histopathologic study discloses diffuse splitting of peripheral nerve myelin and intramyelinic edema with relative sparing of axons.6 A primate study showed electrophysiologic evidence of sensory nerve dysfunction unaccompanied by significant morphologic changes.7

PERHEXILENE Perhexilene maleate is an orally administered prophylactic agent introduced in the 1970s for the treatment of angina pectoris and variant (Prinzmetal’s) angina.224,289,398 More recently, a beneficial effect has been suggested in elderly patients with severe aortic stenosis.372 It is not approved for use in the United States. Its actions include coronary vasodilatation by a direct effect on vascular smooth muscle and blocking of exercise-induced tachycardia. The principal neurotoxic effect of perhexilene is a predominantly demyelinating peripheral neuropathy, uncommon in toxic neuropathies. There are additional reports of raised intracranial pressure with papilledema, and myopathy. The toxic mechanism is likely related to accumulation of lysosomal inclusions in Schwann cells and other tissues. Like amiodarone, perhexilene is a cationic amphiphilic drug that can penetrate lysosomes. It is postulated that such compounds form tight but not irreversible complexes

2564

Diseases of the Peripheral Nervous System

with polar lipids, leading to impairment of degradation and to intralysosomal accumulation.222 The inclusions resemble those found in patients with inherited storage diseases.

Clinical Features Neuropathy associated with perhexilene in humans usually develops gradually after treatment periods of several months to years at the commonly used dosages of 200 to 400 mg daily.125,216,219,315,391 The incidence of symptomatic neuropathy may be low,279,333 but subclinical electrophysiologic abnormalities are reported in up to two thirds of patients.333,339 Large interindividual variations in plasma half-lives, and thereby the risk of toxic drug accumulation, may be the result of genetically determined differences in rate of metabolism of the drug; poor debrisoquine hydroxylators (poor metabolizer phenotype) are at greater risk of developing neuropathy.82,338,344 The risk of toxicity can be reduced by maintaining plasma drug concentrations below 600 ng/mL.172 Symmetrical sensorimotor polyneuropathy begins with acral pain and paresthesias and evolves to both proximal and distal weakness. There may be facial diplegia, perioral numbness, or autonomic dysfunction,125,315 as well as hypoor areflexia and stocking hypoesthesia. After drug withdrawal, recovery evolves gradually over months, and may be incomplete in severe cases.125,216,219,315,391 The CSF protein is typically mildly to moderately elevated, up to several hundred milligrams per deciliter.216,219,315,351,391 Perhexilene may be associated with raised intracranial pressure and papilledema, with or without peripheral neuropathy.137,351 Improvement occurs within a few weeks of stopping the drug. A single patient has been described with acute proximal weakness, rash, and an electromyogram with myopathic motor units and recruitment pattern; the presence of myopathy is tenuous because the CK was normal and no muscle biopsy was performed.366 Resolution occurred within 2 months of drug discontinuation. Systemic effects of perhexilene include weight loss, nausea, dizziness, and hepatotoxicity. Keratopathy with cytoplasmic inclusions in the conjunctiva and corneal epithelium may occur.137

five patients reported segmental demyelination of 16% to 90% of fibers and wallerian degeneration of 3% to 20%.353 Axons show few structural changes, and there is no inflammatory infiltrate.117,315 Numerous polymorphic lysosomal inclusions are present mainly in Schwann cells; they also are prominent in endothelial cells, fibroblasts, perineurial cells, phagocytes, and pericytes, as well as in hepatocytes, skin, and muscle fibers.117,154,315 Biochemical analysis of biopsy samples in four patients suggested that the accumulated perhexilene-induced lipid consists mainly of gangliosides.281

Experimental Studies Perhexilene given orally at 80 mg/kg/day for 4 to 10 weeks to dark Agouti rats, who are poor debrisoquine hydroxylators, caused cytoplasmic lipid inclusions similar to those in humans.247 These were most prominent in sensory and autonomic ganglia, but were also prominent in Schwann cells; the inclusions were accompanied by primary demyelination in spinal roots and mixed demyelination/axonal degeneration in distal nerve segments. The numbers of inclusions were related to the duration of treatment and to the tissue and plasma concentrations of perhexilene. There was mild prolongation of motor latencies in sciatic nerve. Withdrawal of the drug 2 weeks before sacrificing resulted in a decrease in inclusion bodies, indicating at least some degree of reversibility. Sprague-Dawley rats, which are vigorous hydroxylators, showed only mild accumulations of inclusion bodies and only after prolonged duration of treatment. In another study of adult Swiss NMRI-strain mice fed perhexilene for 10 to 18 weeks at 100 to 200 mg/kg/day, inclusions were present in Schwann and perineurial cells of sciatic nerve as well as in muscle biopsies.117 In both studies there were no changes in weight, behavior, or nonquantitative measures of neurologic function. Perhexilene administration induces polymorphous inclusions in the neurons, satellites, Schwann cells, and fibroblasts of cultured mice spinal ganglia.154

PHENYTOIN Electrophysiology Patients exhibit mild to marked prolongation of distal motor latencies and slowing of motor and sensory conduction velocities to under 20 m/s, consistent with primary demyelination, along with associated axonal changes.216,219,315,391

Pathology Pathologic features of sural nerve biopsies, including teased fiber studies, are primarily segmental demyelination along with loss of large myelinated fibers. A study of

Phenytoin (diphenylhydantoin) has been used as an anticonvulsant since 1938, primarily in management of generalized tonic-clonic and partial seizures, and also in neuropathic pain syndromes.230 It affects sodium permeability by binding to and stabilizing inactivated sodium channels across neuronal membranes, slowing the rate of membrane recovery to its resting state. Potential adverse effects are many and include, uncommonly, a clinically significant acute reversible or chronic axonal sensorimotor polyneuropathy. The mechanism of phenytoin-induced neuropathy is unknown.

2565

Neuropathy Caused by Drugs

Clinical Features There is an acute effect of phenytoin on nerve conduction parameters; high serum levels are associated with reversible slowing of sensory conductions,341 and there are rare reports of acute, reversible, clinically significant distal sensory and motor dysfunction, even within hours of oral administration.245,412 Most reported series of chronic phenytoin therapy in epileptics suggest that it specifically, as a sole agent, causes a subclinical and occasionally symptomatic axonal sensorimotor polyneuropathy, with widely varying incidence.74,95,109,221,253,292,341,355,361 Some studies suggest correlations with duration of therapy greater than 10 years, total dosage, or chronically high serum levels (over 20 ␮g/mL). Cases also occur with documented therapeutic levels and much shorter periods of treatment.95 Subclinical cases have absent ankle, or both knee and ankle, reflexes along with mild distal sensory loss predominantly involving vibration sense, and no weakness. Symptomatic individuals experience the insidious development of distal paresthesias, unsteady gait, reduced or absent reflexes, sensory loss involving all modalities, and mild distal weakness. Gradual clinical recovery follows drug withdrawal.

Electrophysiology Nerve conduction studies in acute cases show reversible mild slowing of sensory conduction.341,412 Studies of asymptomatic epileptic adults and children on chronic phenytoin show mild but significant effects on sensory amplitudes, sensory conductions, or motor conductions, correlating with loss of reflexes and most consistent with axonal dysfunction.74,109,221,244,254,341 Despite no weakness, needle EMG may show distal active and chronic denervation.221 Symptomatic cases show characteristic features of a sensorimotor axonopathy.253,292 Electrophysiologic abnormalities may improve with drug withdrawal.

Pathology In the only well-documented case, a sural biopsy performed in a patient with symptomatic neuropathy after 30 years of phenytoin treatment with documented high blood levels displayed a loss of large myelinated fibers, axonal shrinkage, and secondary demyelination.292

Experimental Studies Acute high levels of phenytoin achieved in the rat through intraperitoneal injection resulted in conduction slowing and reduced motor amplitudes within hours.233 Guinea pigs with phenytoin levels above 50 ␮g/mL over 3 to 4 days demonstrated reversible slowing of motor conductions.214

PODOPHYLLIN Podophyllin resin is derived from the rhizomes of the mayapple plant (Podophyllum peltatum). The active component is the highly lipid-soluble compound podophyllotoxin. It is a topical agent for condylomata acuminata,182 and has been used as a purgative by Native Americans and in Chinese medicine for centuries as an herbal preparation (giujiu and Bajiaolian) for a variety of conditions. Podophyllotoxin has a colchicine-like effect, reversibly binding to microtubules and disrupting cellular mitosis and axonal transport, and is an inhibitor of DNA, RNA, and protein synthesis.60–62,395,397,408 It causes an axonal sensorimotor neuropathy. There are two semisynthetic podophyllotoxin derivatives, etoposide and teniposide, which have a broad spectrum of antitumor activity and some neurotoxic potential.

Clinical Features Toxicity can occur via topical, oral (accidental/suicidal), ingestion (Chinese herbal broth, herbal laxative tablets), intralesional, and intramuscular routes, and may be fatal.62,64,75,94,121,126,142,181,265,271,305 As little as 350 mg of ingested podophyllin is lethal.101 Podophyllin is among the more common adulterants in toxic Chinese herbal preparations.58 Toxicity with standard topical application to treat condylomata acuminata is rare. Topical application may be a problem when podophyllin is applied over a large area for a prolonged period, or to friable skin and bleeding mucosa. Systemic toxicity begins within one or more hours, with nausea, vomiting, abdominal pain, and severe diarrhea. There may be associated pancytopenia and hepatic and renal dysfunction. Reversible encephalopathy ensues in some, ranging from mild confusion to coma, occasionally with seizures, lasting days to weeks. The electroencephalogram features generalized slowing or diffuse high-amplitude arrhythmic delta activity.265 High-dose etoposide can cause acute encephalopathy in patients treated for recurrent or resistant glioma.210,365 Symptoms and signs of a sensorimotor polyneuropathy appear within hours to days, or occasionally after over 1 week, and may progress over weeks before slowly improving over weeks to months.62,64,75,94,121,126,142,265,271 Distal paresthesias are pronounced, with sensory loss involving all modalities in a stocking-glove pattern, especially large fiber functions, resulting in pseudoathetosis, sensory ataxia, and areflexia suggesting a neuronopathy. Though the neuropathy is predominantly sensory, some individuals have severe distal weakness. Autonomic disturbances are common and may include paralytic ileus, orthostatic hypotension, urinary retention, and sinus tachycardia. Cases with severe axon loss can have persistent deficits. A series of severe neuropathies associated with combination chemotherapy including vincristine and etoposide suggests additive effects.364 A single case report described a reversible necrotizing myopathy,

2566

Diseases of the Peripheral Nervous System

in addition to encephalopathy and neuropathy, after suicidal ingestion of podophyllin.269

Electrophysiology Nerve conduction studies are consistent with an axonal polyneuropathy affecting sensory more than motor fibers. Abnormalities may be restricted to low-amplitude or absent sensory potentials and H reflexes, a pattern that, along with severe proprioceptive loss, is supportive of involvement at the DRG level. A case with preserved sensory potentials but abnormal H reflex and somatosensory evoked potentials suggests dorsal root localization.142 With motor involvement, there may be distal denervation on needle EMG.

Pathology Sural nerve biopsies display axonal degeneration affecting predominantly large myelinated fibers.64,75,265,271

Experimental Studies Rats given sublethal injections of podophyllotoxin developed neuronal changes including edematous changes of anterior horn motor neurons, more serious changes in DRG neurons and axons with severe depletion of Nissl substance, and doserelated increase in thickness of axons diffusely in the CNS.60,61 Fatal parenteral doses resulted in flaccidity, slow labored respiration, somnolence, unconsciousness, and finally death within 12 to 15 hours.354 Quercetin, a flavonoid present in podophyllin resin, is also toxic at high doses. In cell cultures, podophyllotoxin causes a dose-response inhibition of protein, RNA, and DNA synthesis.408 It binds reversibly to microtubules, causing depolymerization of mitotic spindles and disruption of axonal cytoarchitecture.395,397

PYRIDOXINE Pyridoxine (vitamin B6) is a water-soluble vitamin contained in grains, cereals, meat, and milk. The recommended daily requirement of 2 mg is readily available in the North American diet. Pyridoxine supplements of 100 mg daily are routinely administered to individuals at risk for vitamin deficiency (alcoholics, food faddists), and to persons receiving hydralazine, oral contraceptives, or isoniazid. Most instances of pyridoxine neuropathy occur following self-administration of megavitamin doses for the premenstrual syndrome.322

Clinical Features Oral administration of 0.5 to 10 g/day causes sensory axonopathy. Individuals taking 500 mg daily are asymptomatic for several years, while those taking 7 to 10 g daily generally experience symptoms within months.322 An

unconvincing study claimed that 2 years of consumption of 117 mg daily is associated with sensory symptoms88; it has also been alleged that fetal phocomelia can follow maternal abuse of pyridoxine. Pyridoxine sensory neuropathy is remarkably stereotyped; all affected individuals develop distal symmetrical numbness of the feet accompanied by unsteady gait. Sensory loss and clumsiness of the hands follows, frequently heralded by Lhermitte’s sign. Neurologic examination reveals a stocking-glove distribution of sensory loss; large fiber modalities appear especially affected, but strength is preserved.230,322 A study of deliberate, controlled administration to normal volunteers has demonstrated that subtle elevations in acral vibratory sense thresholds precede onset of symptoms.22 Neurologic disability may worsen for a month (coasting) and then gradually improves following withdrawal; those persons examined after a prolonged follow-up period have made a satisfactory recovery.275,322 Acute sensory neuronopathy has been described in two individuals receiving intravenous doses of 132 and 183 g of pyridoxine for 3 days as treatment for mushroom poisoning.3 Nystagmus developed within 2 days of injection, followed by diffuse paresthesias and disabling appendicular ataxia. Within a week, tendon reflexes disappeared and diffuse sensory loss and autonomic dysfunction developed. Strength was only transiently diminished, and autonomic function eventually recovered. The patients were permanently disabled by severe sensory ataxia.

Electrophysiology Nerve conduction studies of distal nerves indicate profoundly diminished or absent sensory amplitudes and normal motor conduction and amplitudes.

Pathology Sural nerve biopsy reveals widespread, nonspecific axonal degeneration.

Experimental Studies Massive doses produce a sensory neuronopathy in rats and dogs. Rats dosed at 600 mg/kg/day and dogs at 1 g/kg/day developed an unsteady gait within a few days.196,197 Histopathologic examination demonstrates vacuolar changes in the axon hillocks of DRG cells with acute somal chromatolytic reaction. Many sensory axons of peripheral nerves appear to be “transected” near their junctions with the DRG cells.258 The dorsal columns of the spinal cord and the descending fibers of the trigeminal tract in the brainstem also display severe axonal destruction. Detailed analysis of the spatiotemporal pattern of degeneration in rats dosed with 1200 mg/kg/day showed that accumulations of mitochondria, vesicles, and multilaminar and dense bodies appear in the nodal and distal paranodal axons of myelinated

2567

Neuropathy Caused by Drugs

fibers in the L6 DRG on the second day; this preceded degeneration of both peripheral and sensory axons.405 It is suggested that the accumulation of vesicular structures reflects a blockade of fast axonal transport in the proximal axon and cell body of the primary sensory neuron that causes degeneration of its peripheral and central projections. Moderately high parenteral doses (200 mg/kg in the rat and guinea pig) caused an unsteady gait after 6 weeks, progressing to an inability to walk; recovery commenced after pyridoxine was discontinued.396,401 Examination of peripheral nerves and DRGs from these animals showed severe axonal degeneration in sensory nerves with only minor vacuolar changes in the ganglion cells. A study of early neuronal changes in dogs dosed with 200 mg/kg/day found accumulations of neurofilaments in proximal axons and microtubule-neurofilament dissociation in the cytons of DRGs. The authors suggested that this indicates increased neurofilament protein synthesis resulting in obstruction of axonal transport.258 A study of 300 mg/kg/day for 12 weeks in mice failed to cause neuropathy, suggesting variable intermurine susceptibility.401 Two experimental studies convincingly demonstrated neurotropic factor protection by neurotropin-3 (NT-3) in rats given large doses (800 mg/kg for 8 days) of pyridoxine. Both studies showed that NT-3 can attenuate the behavioral, electrophysiologic, and morphologic sequelae of pyridoxine intoxication. One study used daily injection of 2 to 20 mg/kg of NT-3.159 The other study used a nonreplicating recombinant genomic herpes simplex virus vector expressing NT-3 delivered to the DRGs by subcutaneous inoculation in the footpad.66

SURAMIN Suramin is a polysulfonated naphthylurea used for decades as an antiparasitic agent for treatment of African trypanosomiasis and onchocerciasis.45 Antineoplastic activity is demonstrated in a number of cancers, including hormonerefractory metastatic prostate cancer; relapsed nodular lymphoma; ovarian, adrenocortical, and renal carcinomas; and HIV-associated Kaposi’s sarcoma and lymphoma.110 Neurotoxicity is common, potentially severe, and dose limiting and hampers the drug’s clinical utility.

Clinical Features Suramin can cause a variety of systemic adverse reactions, more severe in debilitated individuals and in those receiving high-dose (antineoplastic) regimens. Nephrotoxicity with albuminuria is common; other reactions include headache, upper gastrointestinal distress, myelosuppression, adrenal insufficiency, and coagulopathy. CNS toxicity is not a significant problem because this molecule is a large polar anion that does not normally cross the blood-brain barrier.

Peripheral neurotoxicity is common and dose limiting; there is high risk with blood peak levels over 0.35 mg/mL (88% develop neuropathy). The therapeutic target level is 0.3 mg/mL,29,69,204 though neuropathy is also seen with plasma concentrations over 0.2 mg/mL.29 Two clinical patterns are seen: a mild, distal, axonal sensorimotor polyneuropathy and a severe, subacute demyelinating neuropathy.69,204,347 The axonopathy is characterized by distal, length-dependent paresthesias and weakness, with pinprick and vibratory loss, mild toe extensor weakness, and reduced or absent ankle reflexes. The demyelinating neuropathy, which occurs in about 10% to 15% of patients (and is associated with higher suramin levels), reaches its peak in 2 to 9 weeks and resembles Guillain-Barré syndrome. There is diffuse, mild to severe distal and proximal weakness; distal paresthesias; sensory loss; and areflexia. Bulbar dysfunction and respiratory failure may ensue. The CSF protein is elevated, without pleocytosis. Worsening may continue for several weeks after suramin is discontinued (coasting), perhaps related to the drug’s long half-life of 40 to 50 days. Recovery ensues over weeks to months after the drug is stopped, either spontaneously or with plasmapheresis, which is employed in many cases. Improvement may be incomplete in severe cases.

Electrophysiology The mild axonopathy is characterized by reduced sensory and motor amplitudes with relative preservation of latencies and conduction velocities.69 QST shows abnormalities of vibration and cooling thresholds. The subacute demyelinating neuropathy has the electrophysiologic characteristics of an acquired demyelinating polyneuropathy, including occasional conduction block.69,204,347

Pathology The few sural nerve biopsies reported in patients with the fulminant neuropathy show a variety of abnormalities, from only loss of myelinated fibers to occasional examples of segmental demyelination, lymphocytic infiltration, or increased macrophages.69,204 In a single reported advanced cancer patient with proximal weakness following several weeks of suramin, a muscle biopsy showed a clinically reversible mitochondrial myopathy, with subsarcolemmal collections of abnormal mitochondria and markedly decreased cytochrome c oxidase activity, suggesting a mechanism of toxicity supported by in vitro studies.290

Experimental Studies In a rat model, suramin induced a length-, dose-, and timedependent axonal sensorimotor polyneuropathy.307 Only one of the animals in this study showed changes compatible with prominent demyelination and remyelination, and no

2568

Diseases of the Peripheral Nervous System

inflammatory changes were found. In addition, there was accumulation of glycolipid lysosomal inclusions in DRG neurons. Suramin is associated with inhibition of lysosomal enzymes involved in sphingolipid and mucopolysaccharide degradation,80 and this is a postulated mechanism for its neurotoxicity, perhaps with neuronal ceramide accumulation leading to cell death.139 Suramin blocks receptors for a number of growth factors. In an in vitro model of rat DRG neurons, suramin competitively inhibited binding by nerve growth factor (NGF) to the NGF receptor, and neurite outgrowth inhibition could be reversed by high concentrations of NGF.308 It is unclear if effects on receptor function are important for neurotoxicity. In the same model, another study showed that suramin causes injury to both axons and Schwann cells that is not prevented by NGF and produces lamellar inclusion bodies composed primarily of GM1 ganglioside.138

TACROLIMUS Tacrolimus (FK-506) is a macrolide antibiotic extracted from the soil microorganism Streptomyces tsukubaensis.323 It is a potent immunosuppressant used to prevent rejection following solid organ transplantation. The immunosuppressant and neurotoxic effects are mediated by binding to a cytosolic binding protein (FKBP-12), which results in the inhibition of calcineurin activity, leading to a decrease in interleukin-2 and reduction in T-cell proliferation. Tacrolimus is also a member of a class of compounds, the neuroimmunophilin ligands, with both neuroprotective and neuroregenerative actions mediated by a different binding protein (FKBP-52).141 Central neurotoxic effects are common, including encephalopathy, seizures, akinetic mutism, tremors, and headaches; these respond to dose reduction or discontinuation.108,390 Tacrolimus is associated with the reversible posterior leukoencephalopathy syndrome.167 There are three reports of tacrolimus neuropathy, two of which suggest that it causes a dysimmune demyelinating neuropathy.14,44,141

Clinical Features In an ongoing, prospective study of neurologic complications in some 1000 adult transplant recipients, 3 have been identified with a severe, multifocal sensorimotor neuropathy resembling chronic inflammatory demyelinating polyneuropathy.394 Asymmetrical sensory and motor dysfunction appeared 2 to 10 weeks after first exposure to tacrolimus. Paresthesias or painful dysesthesias were initial symptoms; eventually asymmetrical weakness appeared. Later, there was more symmetrical, diffuse reflex and distal sensory loss, with sparing of cranial and respiratory function. The CSF was acellular, with elevated protein (89 to 131 mg/dL). One patient improved with plasmapheresis and the others with

immunoglobulin, though recovery was slow. A second report described a single patient with a similar demyelinating neuropathy beginning 2 months into treatment with tacrolimus and substantially improving within 5 weeks of drug withdrawal and substitution with cyclosporine.44 In another report, two patients with elevated plasma concentrations of tacrolimus developed encephalopathy along with severe flaccid quadriparesis, hyporeflexia, and preserved sensation 8 days after liver transplantation and the start of tacrolimus therapy; one also had bifacial weakness and ptosis.14 Weakness resolved very rapidly, within days of drug withdrawal, and in one case returned with reintroduction of the drug. The clinical and electrophysiologic (lowamplitude motor responses) features in these cases are interpreted as reflecting a motor axonal neuropathy as a direct toxic effect of the drug. However, the clinical course, features, and limited data do not allow a clear assessment of the pathophysiology, which may reflect demyelination at very proximal or distal nerve segments, or effects at the neuromuscular junction or axonal channels. Bilateral optic neuropathy has been reported in one patient.42

Electrophysiology Nerve conduction studies in one report showed a multifocal, predominantly demyelinating, sensorimotor polyneuropathy. Needle EMG displayed active denervation primarily in distal muscles.394 Similar features were present in another report.44

Pathology Sural nerve biopsy in one patient revealed severe demyelination and axonal loss, with no evidence of vasculitis or inflammation.394

Experimental Studies There are no animal studies that document neurotoxicity. Cyclosporin, whose mechanism of immunosuppression closely resembles that of tacrolimus, can induce a chronic relapsing form of experimental allergic neuritis in rats.240 Hypotheses concerning a dysimmune neuropathy with tacrolimus include alteration of T-cell subsets by suppression of apoptosis in thymic cells causing an increase in circulating autoimmune cells, and enhancement of or failure to inhibit T-cell activation in response to peripheral myelin antigen.394

TAXANES The taxanes, paclitaxel and docetaxel, are novel antineoplastic agents used in the treatment of a variety of solid tumors.11,305 They are diterpene alkaloids isolated from the yew tree, paclitaxel from the bark and needles of

2569

Neuropathy Caused by Drugs

Taxus brevifolia, and docetaxel semisynthetically from needles of T. baccata. The taxanes bind to tubulin, and unlike the vinca alkaloids, which induce microtubule disassembly, they promote polymerization of tubulin and inhibit the disassembly of microtubules. This results in stable but dysfunctional microtubules, disrupting cell functions including mitosis and intracellular transport. Neurotoxicity is considered to result from disruption of axonal transport. It is unclear if the primary site of neurotoxicity is the nerve cell body (neuronopathy) or the axon (distal axonopathy). Systemic toxicities include severe neutropenia, hypersensitivity reactions, cardiac arrhythmias, nausea, diarrhea, mucositis, and alopecia.305 Because granulocyte colonystimulating factor can abrogate neutropenia, neuropathy is the principal dose-limiting toxicity.

Clinical Features Taxane-induced neuropathy is dose dependent, occurring with higher cumulative dose and higher dose per cycle. The reported incidence varies widely, but some 55% to 100% of patients will develop neuropathy from paclitaxel at doses greater than 200 to 250 mg/m2.218,284 At doses below 200 mg/m2, and probably also between 200 and 250 mg/m2, neuropathy is frequent but usually not clinically significant or dose limiting. With docetaxel, mild neuropathy may appear over a wide dose range and as early as 50 mg/m2 cumulative dose, but can be severe above 600 mg/m2.10,165,264 Additional risk factors for neuropathy include prior or concomitant exposure to known neurotoxic agents such as cisplatin, and preexisting neuropathy from diabetes or alcoholism.51,70,174 Symptoms often begin within 1 to 3 days of a single large dose, progressing after each subsequent treatment.123,218 Worsening can progress for weeks to months after the end of treatment (coasting) before improvement.118,166,377 Initial symptoms are usually symmetrical numbness and tingling or burning dysesthesias, most often beginning in the feet but sometimes simultaneously in the hands and feet. There may be initial asymmetry, and perioral and tongue numbness. These features have suggested that the toxicity may be at the level of the ganglion cell body. Examination reveals a distal, symmetrical pattern of sensory loss involving all modalities (particularly vibration) depending on severity, early ankle reflex loss or areflexia, and ataxia in severe cases. Lhermitte’s sign may appear.166,376 Autonomic dysfunction with orthostatic hypotension is rare.177,218,325 Although sensory dysfunction predominates, distal and/or proximal weakness occurs, particularly with high doses.70,127,218,264 Transient myalgias are common. One patient has been reported with bilateral facial palsy following a single high dose on a background of a large cumulative dose and distal neuropathy.209 Optic neuropathy is rare.48

Mild symptoms usually improve within weeks to months with dose reduction or drug discontinuation. Areflexia can resolve completely, but more severe cases may have residual symptoms and deficits.183 Steroid co-medication does not reduce the development of docetaxel-induced neuropathy.287 Despite several clinical trials, there are no established neuroprotective agents in humans; a recent study suggests glutamine may be effective.375

Electrophysiology Most studies are consistent with a symmetrical, lengthdependent, axonal sensorimotor polyneuropathy involving mostly sensory nerves, beginning with the sural.70,264 Needle EMG in cases with weakness may show distal denervation. Studies in patients with proximal weakness suggest either myopathic or neurogenic features.70,127,264 CK levels are normal. QST shows increased vibratory threshold in almost all individuals, correlating with the cumulative dose of paclitaxel in one study,70 but not with docetaxel in another.165 QST is not always as sensitive as the clinical exam.123

Pathology Sural biopsies are few. Two cases with paclitaxel neuropathy showed severe nerve fiber loss, axonal atrophy, secondary demyelination, and no signs of axonal regeneration or microtubule aggregation.312,377 One sural and one superficial peroneal nerve biopsy with docetaxel showed a preferential loss of large myelinated fibers, in one case also with axonal atrophy and evidence of active axonal regeneration.118,264

Experimental Studies In cultured sensory neurons from chick embryos, paclitaxel caused microtubules to aggregate into bundles and inhibited neurite initiation and outgrowth.215 Local injection of paclitaxel subepineurially into rat sciatic nerve induced local accumulation of microtubules in axons and Schwann cells, with associated axonal degeneration and demyelination.306 Similar changes were noted following injection into rat facial nerve, as was impairment of retrograde axonal transport.263 An in vivo chamber system with rat sciatic nerve showed inhibition of fast axonal transport.260 Systemic administration of paclitaxel in the rat intraperitoneally or intravenously induced a neuropathy resembling that in humans.52,53 The pathologic changes occurred mainly in peripheral nerves but also in dorsal and ventral spinal rootlets and spinal dorsal column fibers, with sparing of DRGs and spinal cord motor neurons. There was prominent axonal microtubule accumulation. This study suggests that paclitaxel directly damages axons. Measured paclitaxel tissue concentration after intravenous administration in rats is highest in the DRG, lowest in the brain, and intermediate in sciatic nerve and spinal cord.54

2570

Diseases of the Peripheral Nervous System

Other studies in mice and mature rats show some similar dose- and schedule-dependent pathologic features without involvement of ventral roots, and predominantly large fiber sensory neuropathic features by electrophysiologic and behavioral tests.77,252 Still other animal models display predominantly small fiber dysfunction in mice,12 and peripheral neuropathy with abnormal behavioral tests for pain using mechanical and thermal stimuli in rats.282 Docetaxel also causes neuropathy in mice.28 Several animal studies suggest that different factors may protect against taxane-induced neuropathy, including NGF,12,155 insulin-like growth factor-I,81 glutamate,38 the adrenocorticotropic hormone-(4-9) analogue ORG 2766,153 leukemia inhibitory factor,189 and the neurotrophic factor prosaptide.47 The WldS mouse, a mutant strain with the unique phenotype of resistance to axonal degeneration, shows resistance to paclitaxel neuropathy by behavioral, physiologic, and pathologic measures.383,384

THALIDOMIDE Thalidomide (Thd) is a synthetic substance that was marketed as a sedative and hypnotic in 1957; it was withdrawn in 1961 following an epidemic of dysmyelia in children born to women who took the drug while pregnant. Thalidomide has recently been returned to practice because of its immunomodulating and angiogenic inhibitory effects. The U.S. Food and Drug Administration has approved it for use in management of erythema nodosum leprosum,86 and its immunomodulatory properties have encouraged therapeutic trials in a host of conditions, including discoid lupus erythematosus, recurrent aphthosis, erythema multiforme, graft-versus-host disease, rheumatoid arthritis, Behçet’s syndrome, and multiple myeloma.211,298 Sedation and sensory peripheral neuropathy are potential dose-limiting factors in Thd therapy. Deployment of periodic screening electrophysiologic examinations designed to detect the early onset of nerve dysfunction has minimized the risk of disabling neuropathy.68,135,257,285,294,310

Clinical Features Robust dose-effect and dose-duration relationships are not established for Thd neuropathy. However, recent carefully monitored studies suggest a relationship among cumulative dose, symptomatology, and electrophysiologic changes.68,257 Healthy individuals receiving 100 to 200 mg daily for sedation develop neuropathy after 2 to 9 months. A recent study of persons with metastatic prostate cancer receiving 200 mg daily described neuropathic symptoms in 6 of 8 persons after 6 months of treatment, and in all after 9 months.257 Initial symptoms include tingling paresthesias and numbness in the feet with prominent cramps in the legs. Paresthesias, often painful and distressing, spread up the

lower limbs to the knees and then involve the hands after 2 months. A few individuals have developed mild proximal lower limb weakness; most remain strong. Tendon reflexes are affected late; eventually patellar and ankle reflexes are lost. All sensory modalities are impaired in a stocking-glove distribution. Some patients also develop palmar erythema, brittle nails, and mild tremor. Recovery is rapid and satisfactory if the drug is stopped soon after the onset of symptoms, emphasizing the importance of close monitoring of patients receiving Thd. Earlier reports of poor recovery, with persistent painful paresthesias after 2 to 6 years, may reflect the continuation of treatment after the onset of neuropathy in these individuals.132,133

Electrophysiology Electrodiagnostic studies consistently display loss or diminution of sural nerve and other sensory action potentials, with only minor abnormalities of motor conduction.68,135,163,202,257 Clinical symptoms and decline in sensory nerve action potentials generally are concurrent. Two studies suggest that prolongation of F-wave minimum latencies in the lower limbs precedes abnormalities in sensory nerve action potentials.294,310

Pathology Nerve biopsies during intoxication display wallerian-like degeneration and, after recovery, fiber loss.68 A study of postmortem material disclosed loss of sensory nerve fibers, DRG cells, and dorsal column fibers in the spinal cord.191

Experimental Studies The pathogenesis of Thd neuropathy is unknown, and there is no satisfactory experimental animal model. Administration of high levels (200 to 1000 mg/kg daily) of Thd for periods up to 1 year in two canine studies failed to cause abnormalities in sensory nerve conduction or convincing histopathologic changes.92,116 One study in the rabbit has produced slowing of nerve conduction without histopathologic alterations; another study in the rabbit produced slight alterations in myelin sheath thickness.329,331

TRYPTOPHAN Consumption of the amino acid L-tryptophan (LT) was widespread in the United States in the 1980s; it was used for insomnia, depression, and premenstrual syndrome. In 1989, an epidemic of individuals with severe myalgia and eosinophilia with no identifiable cause was recognized, and the condition was called the eosinophilia-myalgia syndrome (EMS).20,56,162 An epidemiologic association with

2571

Neuropathy Caused by Drugs

LT consumption was quickly established, and LT was removed from the market, but by August 10, 1991, over 1500 cases of EMS were identified, including 36 deaths.356 The epidemic was traced to LT produced by a single Japanese manufacturer. It remains unclear, however, whether EMS was caused by two contaminants identified in LT batches—1,1⬘-(ethylidenebis)-tryptophan (EBT) and 3-(phenylamino)alanine (PAA)—or the disorder can be produced by reagent-grade LT regardless of origin. The pathophysiology is unestablished. Similarities between EMS and other autoimmune diseases such as scleroderma and eosinophilic fasciitis, some clinical features such as persistence or progression of EMS after LT withdrawal or the first appearance of the syndrome months after drug discontinuation, the occasional demyelinating neuropathy, and the demonstration of cellular immune response directed against a connective tissue component in EMS all suggest that an instigating agent may evoke a persistent immune-mediated response.112,128 EMS also has clinical and pathologic features in common with the toxic oil syndrome (TOS), which afflicted over 20,000 people and resulted in over 1200 deaths in Spain in 1981. This epidemic was the result of adulterated rapeseed oil, although the causal agent has not been established.283 A possible biochemical link between EMS and TOS was suggested by the observation that the LT contaminant PAA has chemical similarities to 3-(phenylamino)-1,2-propanediol, an aniline derivative implicated as a contaminant in TOS.237 EMS is a chronic, multisystem disorder characterized by a variety of features, including severe myalgia, fatigue, arthralgia, peripheral edema, cough or dyspnea, fever, various rashes, and sclerodermiform skin, along with striking eosinophilia. Neurologic involvement is common, is an important cause of morbidity and mortality, and may be the presenting or predominant feature, including peripheral neuropathy and myopathy, and neurocognitive symptoms in the chronic phase.

Clinical Features EMS has a subacute onset, evolving over weeks, months, or years in patients taking a wide range of LT doses (median dose 1.5 g/day) from a few weeks to many years.243 Neuropathy was reported in 27% of a large series, and was associated with the majority who died from EMS.356,357 Neuropathic features may take several forms, including a sensorimotor polyneuropathy with acral hypoesthesia and paresthesia and distal-predominant weakness, painful predominantly sensory polyneuropathy, mononeuropathy multiplex, and severe, debilitating cases with proximal greater than distal weakness or diffuse quadriparesis and respiratory insufficiency.46,96,128,158,336,346,371 The severe motor presentation in particular may be associated with either an axonal or a demyelinating electrophysiology. The CSF protein is not significantly elevated in the reported demyelinating cases.

Reflexes are often absent in severe cases. Concurrent or independent myopathic dysfunction may be present, with or without serum CK or aldolase elevation.311,359 Patients may stabilize or improve after LT withdrawal, but many continue to worsen. After 2 years, aside from neuropathy and cognitive changes, most other features of EMS appear to improve.162 While corticosteroids may decrease the eosinophil count, may help the general features in the acute phase of the syndrome, and occasionally are reported to be of benefit in patients with neuropathy, both steroids and the occasional reported use of plasmapheresis, immunoglobulin, or other immunosuppressants has not been uniformly helpful.128 Many survivors of EMS have significant residual disability, and rehabilitation has had limited success.97

Electrophysiology Most studies have been consistent with either an axonal sensorimotor polyneuropathy, an axonal mononeuropathy multiplex, or a demyelinating sensorimotor polyneuropathy including conduction block, temporal dispersion, prolonged F waves, and slowing of motor potentials with associated severe axonal involvement.46,96,128,158,336,346 In myopathic cases, needle EMG shows myopathic motor unit features, along with spontaneous activity.311,359 One study of chronic EMS reported tremor, myoclonus, and myokymia.187

Pathology Sural nerve biopsy features include axonal degeneration, epineurial and perivascular mononuclear inflammation occasionally with eosinophils, inflammatory vasculopathy with luminal narrowing and angiogenesis, and nerve fiber regeneration.346 One biopsy displayed patchy loss of myelin, Schwann cell hyperplasia, and endoneurial fibrosis.96 In two clearly demyelinating cases, perineuritis and extensive demyelination and remyelination were apparent.128 Muscle biopsies show inflammatory infiltrates with lymphocytes, histiocytes, plasma cells, and eosinophils in perimysial and epimysial connective tissue, with perifascicular muscle fiber atrophy and fibrosis.46,96,161,311,346,359 There may be neuropathic features of fiber type grouping.128 Inflammation is demonstrable in skin and fascia.

Experimental Studies Two studies in rats and mice produced inflammation surrounding intramuscular nerve fibers, similar to that seen in human EMS.87,387 Most studies found that only LT from the implicated batch, or the contaminant EBT, were able to induce a disorder similar to EMS, although one report claimed that reagent-grade, “pure” LT also caused similar changes.86,220,342,387

2572

Diseases of the Peripheral Nervous System

VINCA ALKALOIDS Vincristine (VCR), vinblastine, vindesine, and vinorelbine, synthetic derivatives of the periwinkle plant (Vinca rosea), are important chemotherapeutic agents with similar neurotoxic actions. VCR is more widely used than the others, and is usually administered parenterally in combination with other agents to treat lymphomas, leukemia, and solid tumors. Common regimens are CHOP (cyclophosphamide, hydroxydaunorubicin, Oncovin [VCR], and prednisone) and MOPP (mechlorethamine, Oncovin [VCR], procarbazine, and prednisone). VCR is usually given as a bolus once weekly. To reduce the incidence and severity of toxic neuropathy, a single dose should not exceed 2 mg regardless of body area.

Clinical Features VCR neuropathy occurs in a stereotyped manner, is manifest to some extent in most treated patients, and has clinical features suggesting an unusual distribution and evolution of axonopathy.41,50,146,241 Symptoms and signs of peripheral neuropathy commence within 2 months in almost all subjects receiving the standard regimen. Paresthesias herald the onset, develop in about 90% of patients, and frequently appear in the fingers before the feet. Reduced or absent ankle jerks are initial findings in virtually everyone receiving VCR. High-dose regimens are associated with earlier onset and more severe disability.241 Pinprick and touch sensory loss is generally more severe than position sense loss; the pattern of sensory dysfunction usually remains confined to the distal extremities, and rarely progresses to the high stocking-glove patterns associated with other toxic neuropathies. Weakness and autonomic dysfunction are prominent and disabling features of VCR neuropathy. Clumsiness of the hands and leg cramps following exercise are initial motor symptoms. Weakness is most severe in extensors of the fingers and wrists and in foot dorsiflexors; it is frequently incapacitating and may develop rapidly (7 to 10 days). Some patients are unable to walk. Signs of autonomic dysfunction, usually ileus, constipation, and urinary retention, often follow the onset of weakness. Cranial nerve palsies (extraocular, facial nerve, dysphagia) are rare. Hearing loss may appear suddenly and usually recovers gradually.227 Muscle pain in the jaw or legs frequently appears soon after an intravenous injection; it is generally transient. Recovery from peripheral neuropathy is usual except for extreme cases; weakness improves rapidly if the drug is stopped in an early stage of disability. Paresthesias usually abate within 4 weeks of ending therapy, but most patients experience residual mild distal sensory loss. Tendon reflexes eventually reappear, save for the ankle jerks. Following recovery from VCR neuropathy, therapy can

recommence at a lower dosage and continue for several months without recurrence of symptoms. Patients with systemic neoplasms appear no more susceptible to VCR neuropathy than those with primary CNS tumors. Attempts to ameliorate VCR neuropathy by pretreatment with neuronotropic protective techniques have failed. Individuals with hereditary neuropathies have increased susceptibility and are at risk for early development of severe neuropathy, as are patients receiving other chemotherapeutic agents associated with peripheral neuropathy.168 One report suggests that preexisting diabetic neuropathy does not increase vulnerability.50

Electrophysiology Sensory nerve conduction studies in digital nerves disclose reduction in amplitude commensurate with drug dosage and duration, and conduction velocities are only mildly slowed, findings consistent with an axonal neuropathy.50,146,241 Abnormalities in nerve conduction studies may precede onset of symptoms. Electromyographic study of distal muscles shows evidence of distal denervation with fibrillation and reduced numbers of motor units; these changes may precede detectable weakness.

Pathology Sural nerve biopsies in advanced cases disclose nonspecific axonal degeneration; alterations in cytoskeletal components similar to those in experimental studies are not described in human nerve biopsies.241 Neurofilament accumulations in CNS neurons develop in subjects receiving high-dose intravenous therapy or following accidental intrathecal administration.392

Experimental Studies In vitro studies demonstrate that vinca alkaloids are attracted to two high-affinity binding sites on brain tubulin and can prevent microtubule polymerization.286 At high concentrations, they also can react with preformed tubulin. In vitro studies also indicate that VCR can disrupt fast axoplasmic transport.57 In general, it has proven difficult to create an experimental animal model that corresponds to human VCR neuropathy.41 There are no experimental animal models of chronic systemic administration of VCR that demonstrate convincing histopathologic evidence of polyneuropathy in intact animals. One study has shown definite electrophysiologic evidence of axonal degeneration.41 Several investigators have used topical administration or have given VCR systemically to animals following sciatic nerve crush. These studies demonstrated impaired axoplasmic transport and diminished density and disorientation of microtubules.144,313 Studies of the animals with crushed sciatic

Neuropathy Caused by Drugs

nerves also showed impairment of axonal repair and regeneration. One report suggested that local injection of VCR into damaged facial nerves may be useful in preventing the appearance of synkinesis.410 Experimental studies also indicate some neuroprotective effect with administration of NGF and the WldS protein.156,384

REFERENCES 1. Abarbanel, J. M., Osiman, A., Frisher, S., and Herishanu, Y.: Peripheral neuropathy and cerebellar syndrome associated with amiodarone therapy. Isr. J. Med. Sci. 23:893, 1987. 2. Ahmed, A., Loes, D. J., and Bressler, E. L.: Reversible magnetic resonance imaging findings in metronidazole-induced encephalopathy. Neurology 45:588, 1995. 3. Albin, R. L., Greenberg, H. S., Townsend, J. B., et al.: Acute sensory neuropathy from pyridoxine overdose. Neurology 36:125, 1986. 4. Alston, T. A.: Neurotoxicity of metronidazole [letter]. Ann. Intern. Med. 103:161, 1985. 5. Anderson, N. E., Lynch, N. M., and O’Brien, K. P.: Disabling neurologic complications of amiodarone. Aust. N. Z. J. Med. 15:300, 1985. 6. Anderson, T. D., Davidovich, A., Arceo, R., et al.: Peripheral neuropathy induced by 2⬘,3⬘-dideoxycytidine neurotoxicity. Lab. Invest. 66:63, 1991. 7. Anderson, T. D., Davidovich, A., Arezzo, J., and Brosnan, C.: Comparative neurotoxicity of dideoxycytidine in monkeys and rabbits. Toxicol. Lett. 191(Suppl.):1, 1992. 8. Ansbacher, L. E., Bosch, E. P., and Cancilla, P. A.: Disulfiram neuropathy: a neurofilamentous distal axonopathy. Neurology 32:424, 1982. 9. Anzil, A. P., and Dozic, S.: Disulfiram neuropathy: an experimental study in the rat. J. Neuropathol. Exp. Neurol. 37:585, 1978. 10. Apfel, S. C.: Docetaxel neuropathy. Neurology 46:2, 1996. 11. Apfel, S. C.: Taxoids. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. New York, Oxford University Press, p. 1135, 2000. 12. Apfel, S. C., Lipton, R. B., Arezzo, J. C., and Kessler, J. A.: Nerve growth factor prevents toxic neuropathy in mice. Ann. Neurol. 29:87, 1991. 13. Aronson, J. K.: Positive inotropic drugs and drugs used in dysrhythmias. In Dukes, M. N. G. (ed.): Myeler’s Side Effects of Drugs, 12th ed. Amsterdam, Elsevier, p. 397, 1992. 14. Ayres, R. C. S., Dousset, B., Wixon, S., et al.: Peripheral neurotoxicity with tacrolimus. Lancet 343:862, 1994. 15. Bagchi, D., Ray, S. D., Patel, D., and Bagchi, M.: Protection against drug- and chemical-induced multiorgan toxicity by a novel IH636 grape seed proanthocyanidin extract. Drugs Exp. Clin. Res. 27:3, 2001. 16. Balakirev, M. Y., and Zimmer, G.: Mitochondrial injury by disulfiram: two different mechanisms of the mitochondrial permeability transition. Chem. Biol. Interact. 138:299, 2001. 17. Behar, A., Livni, N., and Soffer, D.: Fine structure of sciatic nerves in nitrofurantoin polyneuropathy in rats. Neurotoxicology 4:13, 1977.

2573

18. Behar, A., Rachmilewistz, F., Rahamimoff, R., et al.: Experimental nitrofurantoin polyneuropathy in rats. Arch. Neurol. 13:160, 1965. 19. Bell, D. M.: Neuropathy associated with furaltadone. JAMA 176:808, 1961. 20. Belongia, E. A., Hedberg, C. W., Gleich, G. J., et al.: An investigation of the cause of the eosinophilia-myalgia syndrome associated with tryptophan use. N. Engl. J. Med. 323:357, 1990. 21. Berger, A. B., Arezzo, J. C., Schaumburg, H. H., et al.: 2⬘,3⬘-Dideoxycytidine (ddC) toxic neuropathy: a study of 52 patients. Neurology 43:358, 1993. 22. Berger, A. B., Schaumburg, H. H., Schroeder, E., et al.: Dose response, coasting, and differential fiber vulnerability in human toxic neuropathy: a prospective study of pyridoxine neurotoxicity. Neurology 42:1367, 1992. 23. Bergouignan, F. X., Vital, C., Henry, P., and Eschapasse, P.: Disulfiram neuropathy. J. Neurol. 235:382, 1988. 24. Beuche, W., and Friede, R. L.: Remodeling of nerve structure in experimental isoniazid neuropathy in the rat. Brain 109:759, 1986. 25. Biehl, J. P., and Skavlem, J. H.: Toxicity of isoniazid. Am. Rev. Tuberc. 68:296, 1953. 26. Biehl, J. P., and Vilter, R. W.: Effects of isoniazid on pyridoxine metabolism. J. Am. Med. Assoc. 156:1549, 1954. 27. Bilbao, J. M., Briggs, S. J., and Gray, T. A.: Filamentous axonopathy in disulfiram neuropathy. Ultrastruct. Pathol. 7:295, 1984. 28. Bissery, M. C., Guenard, D., Gueritte-Voegelein, F., and Lavelle, F.: Experimental antitumor activity of taxotere (RP56976, NSC 628503), a Taxol analogue. Cancer Res. 15:4845, 1991. 29. Bitton, R. J., Figg, W. D., Venzon, D. J., et al.: Pharmacologic variables associated with the development of neurologic toxicity in patients treated with suramin. J. Clin. Oncol. 13:2223, 1995. 30. Balkemore, W. F.: Isoniazid. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. Baltimore, Williams & Wilkins, p. 476, 1980. 31. Bockhardt, H., Drenckhahn, D., and Lullmann-Rauch, R.: Amiodarone-induced lipidosis-like alterations in ocular tissues of rats. Albrecht Von Graefes Arch. Klin. Exp. Ophthalmol. 207:91, 1978. 32. Borruat, F. X., and Regli, R.: Pseudotumor cerebri as a complication of amiodarone therapy. Am. J. Ophthalmol. 116:776, 1993. 33. Bouche, P., Leger, J. M., Lacomblez, L., et al.: Peripheral neuropathy during treatment with almitrine bismesylate: report of 28 cases. Muscle Nerve 9(Suppl.):122, 1986. 34. Bouche, P., Leger, J. M., Lacomblez, L., et al: Peripheral neuropathies during treatment of almitrine: report of 46 cases. J. Neurol. 236:29, 1989. 35. Boulanouar, A., Abdallah, E., el Bakkali, M., et al.: Severe toxic optic neuropathies caused by isoniazid: apropos of 3 cases. J. Fr. Ophthalmol. 18:183, 1995. 36. Bouldin, T. W., Hall, C. D., and Krigman, M. R.: Pathology of disulfiram neuropathy. Neuropathol. Appl. Neurobiol. 6:155, 1980. 37. Boyce, E. G., Cookson, E. T., and Bond, W. S.: Persistent metronidazole-induced peripheral neuropathy. DICP 24:19, 1990.

2574

Diseases of the Peripheral Nervous System

38. Boyle, F. M., Wheeler, H. R., and Shenfield, G. M.: Amelioration of experimental cisplatin and paclitaxel neuropathy with glutamate. J. Neurooncol. 41:107, 1999. 39. Bradley, W. G.: The neuromyopathy of vincristine in the guinea pig. J. Neurol. Sci. 10:133, 1970. 40. Bradley, W. G., Karlsson, I. J., and Rassol, C. G.: Metronidazole neuropathy. Br. Med. J. 2:610, 1977. 41. Bradley, W. G., Lassman, L. P., Pearce, G. W., and Walton, J.: The neuromyopathy of vincristine in man: clinical, electrophysiological and pathological studies. J. Neurol. Sci. 10:107, 1970. 42. Brazis, P. W., Spivey, J. R., Bolling, J. P., and Steers, J. L.: A case of bilateral optic neuropathy in a patient on tacrolimus (FK506) therapy after liver transplantation. Am. J. Ophthalmol. 129:536, 2000. 43. Brodsky, J. B., Cohen, E. N., Brown, B. W., et al.: Exposure to nitrous oxide and neurologic disease among dental professionals. Anesth. Analg. 60:297, 1981. 44. Bronster, D. J., Yonover, P., Stein, J., et al.: Demyelinating sensorimotor polyneuropathy after administration of FK506. Transplantation 59:1066, 1995. 45. Burch, T., and Ahburn, L.: Experimental therapy of onchocerciasis with suramin and hetrazan: results of a three year study. Am. J. Trop. Med. Hyg. 31:617, 1951. 46. Burns, S. M., Lange, D. J., Jaffe, I., and Hays, A. P.: Axonal neuropathy in eosinophilia-myalgia syndrome. Muscle Nerve 17:293, 1994. 47. Campana, W. M., Eskeland, N., Calcutt, N. A., et al.: Prosaptide prevents paclitaxel neuroxicity. Neurotoxicology 19:237, 1998. 48. Capri, G., Munzone, E., Tarenzi, E., et al.: Optic nerve disturbances: a new form of paclitaxel neurotoxicity. J. Natl. Cancer Inst. 20:1099, 1994. 49. Carella, F., Riva, E., Morandi, L., et al.: Myopathy during amiodarone treatment: a case report. Ital. J. Neurol. Sci. 8:605, 1987. 50. Casey, E. B., Jellife, A. M., LeQuesne, P. M., and Millet, Y. L.: Vincristine neuropathy: clinical and electrophysiological observations. Brain 96:69, 1973. 51. Cavaletti, G., Bogliun, G., Marzorati, L., et al.: Peripheral neurotoxicity of Taxol in patients previously treated with cisplatin. Cancer 75:1141, 1995. 52. Cavaletti, G., Cavaletti, E., Montaguti, P., et al.: Effect on the peripheral nervous system of the short-term intravenous administration of paclitaxel in the rat. Neurotoxicology 18:137, 1997. 53. Cavaletti, G., Cavaletti, E., Oggioni, N., et al.: Distribution of paclitaxel within the nervous system of the rat after repeated intravenous administration. Neurotoxicology 21:389, 2000. 54. Cavaletti, G., Tredici, G., Braga, M., and Tazzari, S.: Experimental peripheral neuropathy induced in adult rats by repeated intraperitoneal administration of Taxol. Exp. Neurol. 133:64, 1995. 55. Cavanagh, J. B.: On the pattern of change in peripheral nerves produced by isoniazid intoxication in rats. J. Neurol. Neurosurg. Psychiatry 30:219, 1964. 56. Centers for Disease Control: Eosinophilia-myalgia syndrome: New Mexico. MMWR Morb. Mortal. Wkly. Rep. 38:765, 1989.

57. Chan, S. Y., Worth, R., and Ochs, S.: Block of axonplasmic transport in vitro by vinca alkaloids. J. Neurobiol. 11:251, 1980. 58. Chan, T. Y., and Critchley, J. A.: Usage and adverse effects of Chinese herbal medicines. Hum. Exp. Toxicol. 15:5, 1996. 59. Chang, E., Dellon, A. E., Dellon, E. S., and Hendrickson, M. E.: Developing a model of colchicine neuropathy. Microsurgery 22:46, 2002. 60. Chang, L. W., Yang, C. M., Chen, C. F., and Deng, J. F.: Experimental podophyllotoxin (bajiaolian) poisoning: I. Effects on the nervous system. Biomed. Environ. Sci. 5:283, 1992. 61. Chang, L. W., Yang, C. M., Chen, C. F., and Deng, J. F.: Experimental podophyllotoxin (bajiaolian) poisoning: effects on the liver, intestine, kidney, pancreas and testis. Biomed. Environ. Sci. 5:293, 1992. 62. Chang, M. H., Lin, K. P., Wu, Z. A., and Liao, K. K.: Acute ataxic sensory neuronopathy resulting from podophyllin intoxication. Muscle Nerve 15:513, 1992. 63. Chao, C. F., Subjeck, J. R., and Johnson, R. J.: Behavioral and ultrastructural studies of desmethylmisonidazoleinduced neurotoxicity in mice. Br. J. Radiol. 56:27, 1983. 64. Chapon, F., Dupuy, B., Gosset, S., et al.: Accidental poisoning with podophyllin: a case with study of peripheral nerve. Rev. Neurol. (Paris) 147:240, 1991. 65. Charness, M. E., Moradi, F., and Scheinman, M. M.: Frequent neurologic toxicity associated with amiodarone therapy. Neurology 34:669, 1984. 66. Chattopadhyay, M., Wolfe, D., Huang, S., et al.: In vivo gene therapy for pyridoxine-induced neuropathy by herpes simplex virus-mediated gene transfer of neurotropin-3. Ann. Neurol. 51:19, 2002. 67. Chaudhry, V., Chaudry, M., Crawford, T. M., et al.: Toxic neuropathy in patients with pre-existing neuropathy. Neurology 60:337, 2003. 68. Chaudhry, V., Cornblath, D., Corse, A., et al.: Thalidomideinduced neuropathy: clinical and electrophysiological features. Neurology 59:1872, 2002. 69. Chaudhry, V., Eisenberger, M. A., Sinibaldi, V. J., et al.: A prospective study of suramin-induced peripheral neuropathy. Brain 119:2039, 1996. 70. Chaudhry, V., Rowinsky, E. K., Sartorius, S. E., et al.: Peripheral neuropathy from Taxol and cisplatin combination chemotherapy: clinical and electro-physiological studies. Ann. Neurol. 35:204, 1994. 71. Chedru, F., Nodzenski, R., Dunand, J. F., et al.: Peripheral neuropathy during treatment with almitrine. Br. Med. J. (Clin Res. Ed.) 290:896, 1985. 72. Chen, C., Vazquez-Padua, M., and Cheng, Y.: Effect of antihuman immunodeficiency virus nucleoside analogs on mitochondrial DNA and its implication for delayed toxicity. Am. Soc. Pharmacol. Exp. Ther. 39:625, 1991. 73. Cherry, C. L., Gahan, M. E., McArthur, J. C., et al.: Exposure to dideoxynucleosides is reflected in lowered mitochondrial DNA in subcutaneous fat. J. AIDS 30:271, 2002. 74. Chokroverty, S., and Sayeed, Z. A.: Motor nerve conduction study in patients on diphenylhydantoin therapy. J. Neurol. Neurosurg. Psychiatry 38:1235, 1975.

Neuropathy Caused by Drugs 75. Chu, C. C., Huang, C. C., and Chu, N. S.: Sensory neuropathy due to Bajiaolian (podophyllotoxin) intoxication. Eur. Neurol. 44:121, 2000. 76. Chua, C. L., Ohnishi, A., Tateishi, J., and Kuroiwa, Y.: Morphometric evaluation of degenerative and regenerative changes in isoniazid-induced neuropathy. Acta Neuropathol. (Berl.) 60:183, 1983. 77. Cliffer, K. D., Siuciak, J. A., Carson, S. R., et al.: Physiological characterization of Taxol-induced large-fiber sensory neuropathy in the rat. Ann. Neurol. 43:46, 1998. 78. Clouston, P. D., and Donnely, P. E.: Acute necrotising myopathy associated with amiodarone therapy. Aust. N. Z. J. Med. 19:483, 1989. 79. Collings, H.: Polyneuropathy associated with nitrofurantoin therapy. Arch. Neurol. 3:656, 1960. 80. Constantopoulos, G., Rees, S., Cragg, B. G., et al.: Effect of suramin on the activities of degradative enzymes of sphingolipids in rats. Res. Commun. Chem. Pathol. Pharmacol. 32:87, 1981. 81. Contreras, P. C., Steffler, C., Gruner, J. A., et al.: Insulinlike growth factor-I (RHIGF-I) prevents the peripheral neuropathy induced by paclitaxel, cisplatin, and vincristine [abstract]. Ann. Neurol. 38:315, 1995. 82. Cooper, R. G., Evans, D. A., and Whibley, E. J.: Polymorphic hydroxylation of perhexilene maleate in man. J. Med. Genet. 21:27, 1984. 83. Costa-Jussa, F. R., Guevara, A., Brook, G. A., et al.: Changes in denervated skeletal muscle of amiodarone-fed mice. Muscle Nerve 11:627, 1988. 84. Costa-Jussa, F. R., and Jacobs, J. M.: The pathology of amiodarone neurotoxicity. I. Experimental studies with reference to changes in other tissues. Brain 108:735, 1985. 85. Coxon, A., and Pallis, C. A.: Metronidazole neuropathy. J. Neurol. Neurosurg. Psychiatry 39:403, 1976. 86. Crawford, C. L.: Use of thalidomide in leprosy. Toxicol. Rev. 13:177, 1994. 87. Crofford, L. J., Rader, J. I., Dalakas, M. C., et al.: L-tryptophan implicated in human eosinophilia-myalgia syndrome causes fasciitis and perimyositis in the Lewis rat. J. Clin. Invest. 86:1757, 1990. 88. Dalton, K., and Dalton, M. J. T.: Characteristics of pyridoxine overdose neuropathy syndrome. Acta Neurol. Scand. 76:8, 1987. 89. Dano, P., Tamman, D., Brosset, C., and Bregigeon, M.: Peripheral neuropathies caused by disulfiram. Rev. Neurol. (Paris) 152:294, 1996. 90. Deacon, R., Perry, J., Lumb, M., et al.: Selective inactivation of vitamin B12 in rats by nitrous oxide. Lancet 2:1023, 1978. 91. DeDeyn, P. P., Ceuterick, C., Saxena, V., et al.: Chronic colchicine-induced myopathy and neuropathy. Acta Neurol. Belg. 95:29, 1995. 92. Diezel, P. B.: Morphologische veranderungen bei experimenteller Thalidomid-Neuropathie und thalidomidebedingten: Extremitatenmissbildungen. Munchn. Med. Wochenschr. 105:2265, 1963. 93. Dische, S., Saunders, M. I., Lee, M. E., et al.: Clinical testing of the radio-sensitizer Ro-07–0582: experience with multiple doses. Br. J. Cancer 35:567, 1977.

2575

94. Dobb, G. J., and Edis, R. H.: Coma and neuropathy after ingestion of herbal laxative containing podophyllin. Med. J. Aust. 140:495, 1984. 95. Dobkin, B. H.: Reversible subacute peripheral neuropathy induced by phenytoin. Arch. Neurol. 34:189, 1977. 96. Donofrio, P. D., Stanton, C., Miller, V. S., et al.: Demyelinating polyneuropathy in eosinophilia-myalgia syndrome. Muscle Nerve 15:796, 1992. 97. Draznin, E., and Rosenberg, N. L.: Intensive rehabilitation approach to eosinophilia myalgia associated with severe polyneuropathy. Arch. Phys. Med. Rehabil. 74:774, 1993. 98. Dreger, L. M., Gleason, P. P., Chowdhry, T. K., and Gazzuolo, D. J.: Intermittent-dose metronidazole-induced peripheral neuropathy. Ann. Pharmacother. 32:267, 1998. 99. Duarte, J., Cabezas, C., Rodrigues, F., et al.: Colchicineinduced myopathy with myotonia. Muscle Nerve 21:550, 1998. 100. Dubinsky, R. M., Yarchoan, R., Dalakas, M., and Broder, S.: Reversible axonal neuropathy from the treatment of AIDS and related disorders with 2⬘,3⬘-dideoxycytidine (ddC). Muscle Nerve 12:856, 1989. 101. Dudley, W. H.: Fatal podophyllin poisoning. Med. Rec. 37:409, 1890. 102. Duffy, L. F., Daum, F., Fisher, S. E., et al.: Peripheral neuropathy in Crohn’s disease patients treated with metronidazole. Gastroenterology 88:681, 1985. 103. Dupont, P., Hunt, I., Goldberg, L., and Warrens, A.: Colchicine myoneuropathy in a renal transplant patient. Transpl. Int. 15:374, 2002. 104. Dupuy, O., Flocard, F., Vial, C., et al.: Disulfiram (Esperal) toxicity: apropos of 3 original cases. Rev. Med. Interne 16:67, 1995. 105. DuVivier, A., and Fowler, T.: Possible dapsone-induced peripheral neuropathy in dermatitis herpetiformis. Proc. R. Soc. Med. 67:439, 1974. 106. Dyck, P. J., Grina, L. A., Lambert, E. H., et al.: Nitrous oxide neurotoxicity studies in man and rat. Anesthesiology 53:205, 1980. 107. Edwards, M. S., Powers, S. K., Baringer, R. A., et al.: Evoked potentials in rats with misonidazole toxicity. I. Brain stem auditory evoked potentials. J. Neurooncol. 1:115, 1983. 108. Eidelman, B. H., Abu-Elmagd, K., Wilson, J., et al.: Neurologic complications of FK-506. Transplant. Proc. 23:3175, 1991. 109. Eisen, A. A., Woods, J. F., and Sherwin, A. L.: Peripheral nerve function in long-term therapy with diphenylhydantoin: a clinical and electrophysiologic correlation. Neurology 24:411, 1974. 110. Eisenberger, M. A., Reyno, L. M., Jodrell, D. I., et al.: Suramin, an active drug for prostate cancer: interim observations in a Phase 1 trial. J. Natl. Cancer Inst. 85:611, 1993. 111. Ellis, F. G.: Acute polyneuritis after nitrofurantoin therapy. Lancet 2:1136, 1962. 112. Emslie-Smith, A. M., Engel, A. G., Duffy, J., and Bowles, C. A.: Eosinophilia myalgia syndrome: 1. Immunocytochemical evidence for a T-cell mediated immune effector response. Ann. Neurol. 29:524, 1991. 113. Eneanya, D. I., Bianchine, J. R., Duran, D. O., et al.: The actions and metabolic fate of disulfiram. Annu. Rev. Pharmacol. Toxicol. 21:575, 1981.

2576

Diseases of the Peripheral Nervous System

114. Epstein, F. W., and Bohm, M.: Dapsone-induced peripheral neuropathy. Arch. Dermatol. 112:1761, 1976. 115. Estes, M. L., Ewing-Wilson, D., Chou, S. M., et al.: Chloroquine neuromyotoxicity. Am. J. Med. 82:447, 1987. 116. Evans, T. S., Brochman, E. J., Allan, D., et al.: Lack of peripheral neuropathy in beagle dogs after 53 weeks oral administration of thalidomide. Hum. Exp. Toxicol. 19:615, 2000. 117. Fardeau, M., Tome, F. M. S., and Simon, P.: Muscle and nerve changes induced by perhexilene maleate in man and mice. Muscle Nerve 2:24, 1979. 118. Fazio, R., Quattrini, A., Bolognesi, A., et al.: Docetaxel neuropathy: a distal axonopathy. Acta Neuropathol. 98:651, 1999. 119. Feiner, L. A., Younge, B. R., Kazmier, F. J., et al.: Optic neuropathy and amiodarone therapy. Mayo Clin. Proc. 62:702, 1987. 120. Fernando-Roth, R., Itabashi, H., Louie, J., et al.: Amiodarone toxicity: myopathy and neuropathy. Am. Heart J. 119:1223, 1990. 121. Filley, C. M., Graff-Radford, N. R., Lacy, J. R., et al.: Neurologic manifestations of podophyllin toxicity. Neurology 32:308, 1982. 122. Fischl, M. A., Olson, R. M., Follansbee, S. E., et al.: Zalcitabine compared with zidovudine in patients with advanced HIV-1 infection who received the previous zidovudine therapy. Ann. Intern. Med. 118:762, 1993. 123. Forsyth, P. A., Balmaceda, C., Peterson, K., et al.: Prospective study of paclitaxel-induced peripheral neuropathy with quantitative sensory testing. J. Neurooncol. 35:47, 1997. 124. Fraser, A. G., McQueen, I. N. F., Watt, A. H., and Stephens, M. R.: Peripheral neuropathy during long-term high-dose amiodarone therapy. J. Neurol. Neurosurg. Psychiatry 48:576, 1985. 125. Fraser, D. M., Campbell, I. W., and Miller, H. C.: Peripheral and autonomic neuropathy after treatment with perhexilene maleate. Br. Med. J. 2:675, 1977. 126. Freeman, M. C., Weimer, L. H., Arnaudo, E., and Brust, J. C. M.: Neuropathic and autonomic effects of intramuscular podophyllin. Ann. Neurol. 42:416, 1997. 127. Freilich, R. J., Balmaceda, C., Seidman, A. D., et al.: Motor neuropathy due to docetaxel and paclitaxel. Neurology 47:115, 1996. 128. Freimer, M. L., Glass, J. D., Chaudhry, V., et al.: Chronic demyelinating polyneuropathy associated with eosinophiliamyalgia syndrome. J. Neurol. Neurosurg. Psychiatry 55:352, 1992. 129. Frisoni, G. B., and DiMonda, V.: Disulfiram neuropathy: a review (1971–88) and report of a case. Alcohol Alcohol. 24:429, 1989. 130. Frytak, S., Moertel, C. G., Childs, D. S., and Albers, J. W.: Neurologic toxicity associated with high-dose metronidazole therapy. Ann. Intern. Med. 88:361, 1978. 131. Fuller, R. K., Branchey, L., Brightwell, D. R., et al.: Disulfiram treatment of alcoholism: a Veterans Administration Cooperative Study. JAMA 256:1449, 1986. 132. Fullerton, P. M., and Kremer, M.: Neuropathy after intake of thalidomide (Distaval). Br. Med. J. 2:855, 1961. 133. Fullerton, P. M., and O’Sullivan, D. J.: Thalidomide neuropathy: a clinical, electrophysiological, and histological

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

150.

151.

follow-up study. J. Neurol. Neurosurg. Psychiatry 31:543, 1968. Gammon, G. D., Burge, F. W., and King, G.: Neural toxicity in tuberculous patients treated with isoniazid (isonicotinic acid hydrazide). Arch. Neurol. Psychiatry 70:64, 1953. Gardner-Medwin, J. M., Smith, N. J., and Powell, R. J.: Clinical experience with thalidomide in the management of severe oral and genital ulceration in conditions such as Behcet’s disease: use of neurophysiological studies to detect thalidomide neuropathy. Ann. Rheum. Dis. 53:828, 1994. Gherardi, R., Benevenuti, C., Lejong, J. L., et al.: Peripheral neuropathy in patients treated with almitrine dimesylate. Lancet 1:1247, 1985. Gibson, J. M., Fielder, A. R., Garner, A., and Millac, P.: Severe ocular side effects of perhexilene maleate: case report. Br. J. Ophthalmol. 68:553, 1984. Gill, J. S., Hobday, K. L., and Windebank, A. J.: Mechanism of suramin toxicity in stable myelinating dorsal root ganglion cultures. Exp. Neurol. 133:113, 1995. Gill, J. S., and Windebank, A. J.: Suramin induced ceramide accumulation leads to apoptotic cell death in dorsal root ganglion neurons. Cell Death Differ. 5:876, 1998. Godman, G. C., and Murray, M. R.: Influence of colchicine on the form of skeletal muscle in tissue culture. Proc. Soc. Exp. Biol. 84:668, 1953. Gold, B. G.: Neuroimmunophilin ligands: evaluation of their therapeutic potential for the treatment of neurological disorders. Expert Opin. Invest. Drugs 9:2331, 2000. Gorin, F., Kindall, D., and Seyal, M.: Dorsal radiculopathy resulting from podophyllin toxicity. Neurology 39:607, 1989. Grant, A. C., and Krendel, D. A.: Multifocal demyelinating neuropathy due to amiodarone [abstract]. Ann. Neurol. 40:539, 1996. Green, L. S., Donoso, J. A., Heller-Bettinger, I. E., and Samson, F. E.: Axonal transport disturbances in vincristine-induced peripheral neuropathy. Ann. Neurol. 1:255, 1977. Griffin, J. W., Price, D. L., Keuthe, D. O., and Goldberg, A. M.: Neurotoxicity of misonidazole in rats. I. Neuropathol. Neurotoxicol. 1:299, 1979. Guiheneuc, P., Ginet, J., Groleau, J. Y., and Rojonan, J.: Early phase of vincristine neuropathy in man: electrophysiological evidence for a dying-back phenomenon, with transitory enhancement of spinal transmission of the monosynaptic reflex. J. Neurol. Sci. 45:355, 1980. Guillaume, S., and Joachim, M.: Optic neuropathy from disulfiram (Antabuse): an observation. Bull. Soc. Belge Ophthalmol. 268:161, 1998. Gutman, L., and Johnson, D.: Nitrous oxide-induced myeloneuropathy: report of cases. J. Am. Dent. Assoc. 103:239, 1981. Guttmann, L.: Dapsone. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. New York, Oxford University Press, p. 464, 2000. Guttmann, L., Martin, J. D., and Welton, W.: Dapson motor neuropathy—an axonal disease. Neurology 26:524, 1976. Hahn, A. F., and Feasby, T. E.: Metronidazole neuropathy— a sensory neuronopathy. Neurology 39(Suppl. 1):289, 1989.

Neuropathy Caused by Drugs 152. Halloran, T. J.: Convulsions associated with high cumulative doses of metronidazole. Drug Intell. Clin. Pharmacol. 16:409, 1982. 153. Hamers, F. P., Pette, C., Neijt, J. P., and Gispen, W. H.: The ACTH-(4-9) analog, ORG 2766, prevents Taxol-induced neuropathy in rats. Eur. J. Pharmacol. 233:177, 1993. 154. Hauw, J. J., Mussini, J. M., Boutry, J. M., et al.: Perhexilene maleate induced lipidosis in human peripheral nerve and tissue culture: ultrastructural and biochemical changes. Clin. Toxicol. 18:1405, 1981. 155. Hayakawa, K., Itoh, T., Niwa, H., et al.: Nerve growth factor prevention of aged-rat sympathetic neuron injury by cisplatin, vincristine, and Taxol—in vitro explant study. Neurosci. Lett. 274:103, 1999. 156. Hayakawa, K., Sobue, G., Itoh, T., and Mitsuma, T.: Nerve growth factor prevents neurotoxic effects of cisplatin, vincristine, and Taxol on rat sympathetic ganglion explants in vitro. Life Sci. 55:519, 1994. 157. Heath, M. F., Costa-Jussa, F. R., Jacobs, J. M., and Jacobson, W.: The induction of pulmonary phospholipidosis and the inhibition of lysosomal phospholipases by amiodarone. Br. J. Exp. Pathol. 66:391, 1985. 158. Heiman-Patterson, T. D., Bird, S. J., Parry, G. J., et al.: Peripheral neuropathy associated with eosinophilia-myalgia syndrome. Ann. Neurol. 28:522, 1990. 159. Helgren, M. E., Cliffer, K. D., Torrento, K., et al.: Neurotropin-3 administration attenuates deficits of pyridoxine-induced large-fiber sensory neuropathy. J. Neurosci. 17:322, 1997. 160. Heng, J. E., Vorwerk, C. K., Lessell, E., et al.: Ethambutol is toxic to retinal ganglion cells via an excitotoxic pathway. Invest. Ophthalomol. Vis. Sci., 40:190, 1999. 161. Herrick, M. K., Chang, Y., Horoupian, D. S., et al.: L-tryptophan and the eosinophilia-myalgia syndrome: pathologic findings in eight patients. Hum. Pathol. 22:12,1991. 162. Hertzman, P. A., Blevins, W. L., Mayer, J., et al.: Association of the eosinophilia-myalgia syndrome with the ingestion of tryptophan. N. Engl. J. Med. 322:869, 1990. 163. Hess, C. W., Hunziker, T., Kupfer, A., and Ludin, H. P.: Thalidomide-induced peripheral neuropathy: a prospective clinical, neurophysiological and pharmacogenetic evaluation. J. Neurol. 233:83, 1986. 164. Heyer, E. J., Simpson, D. M., Bodis-Wollner, I., and Diamond, S. P.: Nitrous oxide: clinical and electrophysiological investigation of neurological complications. Neurology 36:1618, 1986. 165. Hilkens, P. H. E., Verweij, J., Stoter, G., et al.: Peripheral neurotoxicity induced by docetaxel. Neurology 46:104, 1996. 166. Hilkens, P. H. E., Verweij, J., Vecht, C. J., et al.: Clinical characteristics of severe peripheral neuropathy induced by docetaxel (Taxotere). Ann. Oncol. 8:187, 1997. 167. Hinchey, J., Chaves, C., Appignani, B., et al.: A reversible posterior leukoencephalopathy syndrome. N. Engl. J. Med. 334:494, 1996. 168. Hogan-Dann, C., Fellmeth, W. G., McGuire, A. S., and Kiley, V. A.: Polyneuropathy following vincristine therapy in two patients with Charcot-Marie-Tooth syndrome. JAMA 252:2862, 1984. 169. Holtzer, H., Forry-Schaudies, S., Dlugosz, A., et al.: Interactions between IF’s, microtubules, and myofibrils in

170.

171.

172.

173.

174.

175.

176.

177.

178.

179. 180.

181.

182. 183.

184.

185.

186.

187.

2577

fibrogenic and myogenic cells. Ann. N. Y. Acad. Sci. 455:106, 1985. Homeida, M., Babikr, A., and Daneshmend, T. K.: Dapsoneinduced optic atrophy and motor neuropathy. Br. Med. J. 281:1180, 1980. Horlen, C. K., Seifert, C. F., and Malouf, C. S.: Toxic metronidazole-induced MRI changes. Ann. Pharmacother. 34:1273, 2000. Horowitz, J. D., Sia, S. T., MacDonald, P. S., et al.: Perhexilene maleate treatment for severe angina pectoris— correlations with pharmacokinetics. Int. J. Cardiol. 13:219, 1986. Hughes, H. B., Biehl, J. P., Jones, A. P., and Schmidt, L. H.: Metabolism of isoniazid in man as related to the occurrence of peripheral neuritis. Am. Rev. Tuberc. Pulm. Dis. 70:266, 1954. Iniguez, C., Larrode, P., Mayordomo, J. I., et al.: Reversible peripheral neuropathy induced by a single administration of high-dose paclitaxel. Neurology 51:868, 1998. Jacobs, J. M., and Costa-Jussa, F. R.: The pathology of amiodarone neurotoxicity. II. Peripheral neuropathy in man. Brain 108:753, 1985. Jacobs, J. M., Miller, R. H., and Cavanagh, J. B.: The distribution of degenerative changes in INH neuropathy: further evidence for focal axonal lesions. Acta. Neuropathol. (Berl.) 48:1, 1979. Jerian, S. M., Sarosy, G. A., Link, C. J., et al.: Incapacitating autonomic neuropathy precipitated by Taxol. Gynecol. Oncol. 51:277, 1993. Jones, D. H., Bleehen, N. M., Workman, P., and Smith, N. C.: The role of microsomal enzyme inducers in the reduction of misonidazole neurotoxicity. Br. J. Radiol. 56:865, 1983. Jones, W. A., and Jones, G. P.: Peripheral neuropathy due to isoniazid: report of two cases. Lancet 1:1073, 1953. Kamala, A. N., Shetty, V. P., and Antia, N. H.: Evaluation of the neurotoxicity of dapsone in the mouse model. Indian J. Lepr. 56:251, 1984. Kao, W. F., Hung, D. Z., Tsai, W. J., et al.: Podophyllotoxin intoxication: toxic effect of Bajiaolian in herbal therapeutics. Hum. Exp. Toxicol. 11:480, 1992. Kaplan, I. W.: Condyloma acuminata. New Orl. Med. Surg. J. 94:388, 1942. Kaplan, J. G., Einzig, A. I., and Schaumburg, H. H.: Taxol causes permanent large fiber peripheral nerve dysfunction: a lesson for preventative strategies. J. Neurooncol. 16:105, 1993. Kapoor, K., Chandra, M., Nag, D., et al.: Evaluation of metronidazole toxicity: a prospective study. Int. J. Clin. Pharmacol. Res. 19:83, 1999. Karamanakos, P. N., Pappas, P., Stephanou, P., and Marselos, M.: Differentiation of disulfiram effects on central catecholamines and hepatic ethanol metabolism. Pharmacol. Toxicol. 88:106, 2001. Kasik, J. E.: Tuberculosis and other mycobacterial diseases. In Rakel, R. E. (ed.): Conn’s Current Therapy. Philadelphia, W. B. Saunders, p. 217, 1994. Kaufman, L. D., Kaufman, M. A., and Krupp, L. B.: Movement disorders in the eosinophilia-myalgia syndrome: tremor, myoclonus, and myokymia. J. Rheumatol. 22:157, 1995.

2578

Diseases of the Peripheral Nervous System

188. Keswani, S. C., Chander, B., Hasan, C., et al.: FK506 is neuroprotective in a model of antiretroviral toxic neuropathy. Ann. Neurol. 53:57, 2003. 189. Kilpatrick, T. J., Phan, S., Reardon, K., et al.: Leukemia inhibitory factor abrogates paclitaxel-induced axonal atrophy in the Wistar rat. Brain Res. 911:163, 2001. 190. Kitson, T. M.: The disulfiram-ethanol reaction. J. Stud. Alcohol 38:96, 1977. 191. Klinghardt, G. W.: Ein beitrag der experimentellen neuropathology zur toxizitatsprufung neuer chemotherapeutica. Mitt. Max Planck Ges. 3:142, 1965. 192. Kodavanti, P. R. S., Pentlyala, S. N., Yallpragada, R., and Desaiah, D.: Amiodarone and desethylamiodarone increase intrasynaptosomal free calcium through receptor mediated channel. Naunyn Schmeidebergs Arch. Pharmacol. 345:213, 1992. 193. Koller, W. C., Gehlmann, L. K., Malkinson, F. D., and Davis, F. A.: Dapsone-induced peripheral neuropathy. Arch. Neurol. 36:644, 1977. 194. Kondo, H., Osborne, M. L., Kollhouse, J. F., et al.: Nitrous oxide has multiple deleterious effects on cobalamin metabolism and causes decreases in activity of both mammalian cobalamin dependent enzymes in rats. J. Clin. Invest. 67:1270, 1981. 194. Krauss, J. K., Mohadjer, M., Wakhloo, A. K., and Mundinger, F.: Dystonia and akinesis due to pallidoputaminal lesions after disulfiram intoxication. Mov. Disord. 6:166, 1991. 196. Krinke, G., Naylor, D. C., and Skorpil, V.: Pyridoxine megavitaminosis: an analysis of the early changes induced with massive doses of vitamin B6 in rat primary sensory neurons. J. Neuropathol. Exp. Neurol. 44:117, 1985. 197. Krinke, G., Schaumburg, H. H., Spencer, P. S., et al.: Pyridoxine mega-vitaminosis produces degeneration of peripheral sensory neurons (sensory neuronopathy) in the dog. Neurotoxicology 2:13, 1981. 198. Kuncl, R. W., Cornblath, D. R., Avila, O., and Duncan, G.: Electrodiagnosis of human colchicine myoneuropathy. Muscle Nerve 12:360, 1989. 199. Kuncl, R. W., and Duncan, G.: Chronic human colchicine myopathy and neuropathy. Arch. Neurol. 45:245, 1988. 200. Kuncl, R. W., Duncan, G., Watson, D., et al.: Colchicine myopathy and neuropathy. N. Engl. J. Med. 316:1562, 1987. 201. Kusumi, R. K., Plouffe, J. F., Wyatt, R. H., and Fass, R. J.: Central nervous system toxicity associated with metronidazole therapy. Ann. Intern. Med. 93:59, 1980. 202. Lagueny, A., Rommel, A., Vignolly, B., et al.: Thalidomide neuropathy: an electrophysiological study. Muscle Nerve 9:837, 1986. 203. Laplane, D., Attal, N., Sauron, B., et al.: Lesions of basal ganglia due to disulfiram neurotoxicity. J. Neurol. Neurosurg. Psychiatry 55:925, 1992. 204. LaRocca, R. V., Meer, J., Gilliatt, R. W., et al.: Suramininduced polyneuropathy. Neurology 40:954, 1990. 205. Latini, R., Tognoni, G., and Kates, R. E.: Clinical pharmacokinetics of amiodarone. Clin. Pharmacokinet. 9:136, 1984. 206. Layzer, R. B.: Myeloneuropathy after prolonged exposure to nitrous oxide. Lancet 2:1227, 1978.

207. Layzer, R. B., Fishman, B. A., and Schafer, J. A.: Neuropathy following abuse of nitrous oxide. Neurology 28:504, 1978. 208. Learned-Coughlin, S.: Peripheral neuropathy induced by metronidazole. Ann. Pharmacother. 28:536, 1994. 209. Lee, R. T., Oster, M. W., Balmaceda, C., et al.: Bilateral facial nerve palsy secondary to the administration of highdose paclitaxel. Ann. Oncol. 10:1245, 1999. 210. Leff, R. S., Thompson, J. M., Daly, M. B., et al.: Acute neurologic dysfunction after high-dose etoposide therapy for malignant glioma. Cancer 62:32, 1988. 211. Lehman, T. J., Striegel, K. H., and Onel, K. B.: Thalidomide therapy for recalcitrant systemic onset juvenile rheumatic arthritis. J. Pediatr. 140:125, 2002. 212. Lemaire, J. F., Autret, A., Biziere, K., et al.: Amiodarone neuropathy: further arguments for human drug-induced neurolipidosis. Eur. Neurol. 21:65, 1982. 213. Lessell, S.: Histopathology of experimental ethambutol intoxication. Invest. Ophthalmol. 15:765, 1976. 214. LeQuesne, P. M., Goldberg, V., and Vajda, F.: Acute conduction velocity changes in guinea pigs after administration of diphenylhydantoin. J. Neurol. Neurosurg. Psychiatry 39:995, 1976. 215. Letourneau, P. C., and Ressler, A. H.: Inhibition of neurite initiation and growth by Taxol. J. Cell Biol. 98:1355, 1984. 216. Lhermitte, F., Fardeau, M., Chedru, F., and Mallecourt, J.: Polyneuropathy after perhexilene maleate therapy. Br. Med. J. 1:1256, 1976. 217. Lhermitte, F., Fritel, D., Cambier, J., et al.: Polynevrites au cours traitements par la nitrofurantoine. Presse Med. 71:767, 1963. 218. Lipton, R. B., Apfel, S. C., Dutcher, J. P., et al.: Taxol produces a predominantly sensory neuropathy. Neurology 39:368, 1989. 219. Lorentz, I. T., and Shortall, M.: Perhexilene neuropathy: a report of two cases. Aust. N. Z. J. Med. 13:517, 1983. 220. Love, L. A., Rader, J. I., Crofford, L. J., et al.: Pathological and immunological effects of ingesting L-tryptophan and 1,1⬘-ethylidenebis (L-tryptophan) in Lewis rats. J. Clin. Invest. 91:804, 1993. 221. Lovelace, R. E., and Horwitz, S. J.: Peripheral neuropathy in long-term diphenylhydantoin therapy. Arch. Neurol. 18:69, 1968. 222. Lullmann, H., and Lullmann-Rauch, R.: Perhexilene induces generalized lipidosis in rats. Klin. Wochenschr. 56:309, 1978. 223. Lunsford, J. M., Wynn, M. H., and Kwan, W. H.: Nitrous oxide-induced myeloneuropathy. J. Foot Surg. 22:222, 1983. 224. Lyon, L. J., Nevins, M. A., Fisch, S., and Henry, S.: Perhexilene maleate in treatment of angina pectoris. Lancet 1:1272, 1971. 225. Macaluso, D. C., Shults, W. T., and Fraunfelder, F. T.: Features of amiodarone induced optic neuropathy. Am. J. Ophthalmol. 127:610, 1999. 226. MacDonald, R. D., and Engel, A. G.: Experimental chloroquine myopathy. J. Neuropathol. Exp. Neurol. 29:479, 1973. 227. Mahajan, S. L.: Acute acoustic nerve palsy associated with vincristine therapy. Cancer 42:2404, 1981. 228. Mamoli, B., Wessely, P., Kogelnik, H. D., et al.: Electroneurographic investigations of misonidazole polyneuropathy. Eur. Neurol. 18:405, 1979.

Neuropathy Caused by Drugs 229. Mandell, G. L., and Sande, M. A.: Antimicrobial agents: drugs used in the chemotherapy of tuberculosis and leprosy. In Gilman, A. G., Rall, T. W., Nies, A. S., and Taylor, P. (eds.): Goodman and Gilman’s The Pharmacological Basis of Therapeutics, 8th ed. New York, McGraw-Hill, p. 1146, 1993. 230. Mann, A., Sinnett, M. J., and Shinnar, S.: Phenytoin. In Spencer, P. S., Schaumburg, H. H., and Ludolph, A. C. (eds.): Experimental and Clinical Neurotoxicology, 2nd ed. New York, Oxford University Press, p. 988, 2000. 231. Manolis, A. S., Tordjman, T., Mack, K. D., and Estes, N. A. 3rd: Atypical pulmonary and neurologic complications of amiodarone in the same patient: report of a case and review of the literature. Arch. Intern. Med. 147:1805, 1987. 232. Mansour, A. M., Puklin, J. E., and O’Grady, R.: Optic nerve ultrastructure following amiodarone therapy. J. Clin. Neuroophthalmol. 8:231, 1988. 233. Marcus, D. J., Swift, T. R., and McDonald, T. F.: Acute effects of phenytoin on peripheral nerve function in the rat. Muscle Nerve 4:48, 1981. 234. Marie, R. M., LeBiez, E., Busson, P., et al.: Nitrous oxide anesthesia-associated myelopathy. Arch. Neurol. 57:380, 2000. 235. Martinez-Arizala, A., Sobol, S. M., McCarty, G. E., et al.: Amiodarone neuropathy. Neurology 33:643, 1983. 236. Matsuoka, Y., Mukoyama, M., and Sobue, I.: Histopathological study of experimental ethambutol neuropathy. Clin. Neurol. 12:453, 1972. 237. Mayeno, A. N., Belongia, E. A., Lin, F., et al.: 3-(Phenylamino)alanine, a novel aniline-derived amino acid associated with the eosinophilia-myalgia syndrome: a link to the toxic oil syndrome? Mayo Clin. Proc. 67:1134, 1992. 238. McCann, S. R., Weir, D. G., Dinn, J., et al.: Neurological damage induced by nitrous oxide in the monkey in the absence of any detectable haematological change. Br. J. Haematol. 43:496, 1979. 239. McCarthy, B. G., Hsieh, S. T., Stocks, A., et al.: Cutaneous innervation in sensory neuropathies: evaluation by skin biopsy. Neurology 45:1848, 1995. 240. McCombe, P. A., van der Krecke, S. A., and Pender, M. P.: The effects of prophylactic cyclosporine A on experimental allergic neuritis (EAN) in the Lewis rat: induction of relapsing EAN using low-dose cyclosporine A. J. Neuroimmunol. 28:131, 1990. 241. McLeod, J. G., and Penny, R.: Vincristine neuropathy: an electrophysiological and histological study. J. Neurol. Neurosurg. Psychiatry 32:297, 1969. 242. McMorrow, A. M., Adams, R. J., and Rubenstein, M. N.: Combined system disease after nitrous oxide anesthesia: a case report. Neurology 45:1224, 1995. 243. Medsger, T. A.: Tryptophan-induced eosinophilia-myalgia syndrome. N. Engl. J. Med. 322:926, 1990. 244. Mehrotra, T. N., Srivastava, S. S., Goel, V. K., et al.: A study of nerve conduction velocity in patients on diphenylhydantoin therapy. J. Assoc. Physicians India 40:619, 1992. 245. Meienberg, O., and Bajc, O.: Acute polyneuropathy caused by diphenylhydantoin intoxication. Dtsch. Med. Wochenschr. 100:1532, 1975. 246. Meier, C., Kauer, B., Mulle, R. U., and Ludin, H. P.: Neuromyopathy during chronic amiodarone treatment: a case report. J. Neurol. 220:231, 1979.

2579

247. Meier, C., Wahllaender, A., Hess, C. W., and Preisig, R.: Perhexilene-induced lipidosis in the dark Agouti (DA) rat: an animal model of genetically determined neurotoxicity. Brain 109:649, 1986. 248. Meier-Ruge, W. A., and Schaumburg, H. H.: Chloroquine. In Spencer, P. S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. New York, Oxford University Press, p. 375, 2000. 249. Melgaard, B., Hansen, H. S., Kamieniecka, Z., et al.: Misonidazole neuropathy: a clinical, electrophysiological, and histological study. Ann. Neurol. 12:10, 1982. 250. Melgaard, B., Kohler, O., Sand Hansen, H., et al.: Misonidazole neuropathy: a prospective study. J. Neurooncol. 6:227, 1988. 251. Merigan, T. C., and Skowron, G.: Safety and tolerance of dideoxycytidine as a single agent: results of early-phase studies in patients with acquired immunodeficiency syndrome (AIDS) or advanced AIDS-related complex. Am. J. Med. 88:11S, 1990. 252. Mimura, Y., Kato, H., Eguchi, K., and Ogawa, T.: Schedule dependency of paclitaxel-induced neuropathy in mice: a morphological study. Neurotoxicology 21:513, 2000. 253. Mitrabhakdi, E., Krauss, G. R., and Chaudhry, V.: Clinical and electro-physiological features of phenytoin-induced polyneuropathy. Muscle Nerve 18:1069, 1995. 254. Mochizuki, Y., Suyehiro, Y., Tanizawa, A., et al.: Peripheral neuropathy in children on long-term phenytoin therapy. Brain Dev. 3:375, 1981. 255. Moddel, G., Bilbao, J. M., Payne, D., and Ashby, P.: Disulfiram neuropathy. Arch. Neurol. 35:658, 1978. 256. Mokri, B., Ohnishi, A., and Dyck, P. J.: Disulfiram neuropathy. Neurology 31:730, 1981. 257. Molloy, F. M., Floeter, M. K., Syed, N. A., et al.: Thalidomide neuropathy in patients treated for metastatic prostate cancer. Muscle Nerve 24:1050, 2001. 258. Montpetit, V. J. A., Clapin, D. F., Tryphonas, L., and Dancea, S.: Alteration of neuronal cytoskeletal organization in dorsal root ganglia associated with pyridoxine neurotoxicity. Acta Neuropathol. 76:71, 1988. 259. Nair, V. S., LeBrun, M., and Kass, I.: Peripheral neuropathy associated with ethambutol. Chest 77:98, 1980. 260. Nakata, T., and Yorifuji, H.: Morphological evidence of the inhibitory effect of Taxol on the fast axonal transport. Neurosci. Res. 35:113, 1999. 261. Navarro, J. C., Rosales, R. L., Ordinario, A. T., et al.: Acute dapsone-induced peripheral neuropathy. Muscle Nerve 12:604, 1989. 262. Nazarian, S. M., and Jay, W. M.: Bilateral optic neuropathy associated with amiodarone therapy. J. Clin. Neuroopthalmol. 8:25, 1988. 263. Nennesmo, I., and Reinholt, F. P.: Effects of intraneural injection of Taxol on retrograde axonal transport and morphology of corresponding nerve cell bodies. Virchows Arch. B Cell. Pathol. Mol. Pathol. 55:241, 1988. 264. New, P. Z., Jackson, C. E., Rinaldi, D., et al.: Peripheral neuropathy secondary to docetaxel (Taxotere). Neurology 46:108, 1996. 265. Ng, T. H., Chan, Y. W., Yu, Y. L., et al.: Encephalopathy and neuropathy following ingestion of a Chinese herbal broth containing podophyllin. J. Neurol. Sci. 101:107, 1991.

2580

Diseases of the Peripheral Nervous System

266. Nukada, H., and Pollock, M.: Disulfiram neuropathy: a morphometric study of sural nerve. J. Neurol. Sci. 51:51, 1981. 267. Ochoa, J.: Isoniazid neuropathy in man: quantitative electron microscopic study. Brain 93:831, 1970. 268. Ohnishi, A., Chua, C. L., and Kuroiwa, Y.: Axonal degeneration distal to the site of accumulation of vesicular profiles in the myelinated fiber axon in experimental isoniazid neuropathy. Acta. Neuropathol. (Berl.) 67:195, 1985. 269. Olivera, A. S., Calia, L. C., Kiyomoto, B. H., et al.: Acute necrotizing myopathy and podophyllin toxicity: report of a fatal case. Arq. Neuropsiquiatr. 54:288, 1996. 270. Olney, R. K., and Miller, R. G.: Peripheral neuropathy associated with disulfiram administration. Muscle Nerve 3:172, 1980. 271. O’Mahoney, S., Keohane, C., Jacobs, J., et al.: Neuropathy due to podophyllin intoxication. J. Neurol. 237:110, 1990. 272. Pages, M., and Pages, A. M.: Peripheral nerve lesions in chloroquine-induced neuromyopathies. Ann. Pathol. 4:289, 1984. 273. Palakurthy, P. R., Iyer, V., and Meckler, R. J.: Unusual neurotoxicity associated with amiodarone therapy. Arch. Intern. Med. 147:881, 1987. 274. Palliyath, S. K., Schwartz, B. D., and Gant, L.: Peripheral nerve functions in chronic alcoholic patients on disulfiram: a six month follow up. J. Neurol. Neurosurg. Psychiatry 53:227, 1990. 275. Parry, G. J., and Bresden, D. E.: Neuropathy due to pyridoxine intoxication. Neurology 35:1466, 1985. 276. Paulson, J. C., and McClure, W. O.: Inhibition of axoplasmic transport by colchicine, podophyllotoxin, and vinblastine: an effect on microtubules. Ann. N. Y. Acad. Sci. 253:517, 1975. 277. Paulson, O. B., Melgaard, B., Hansen, H. S., et al.: Misonidazole neuropathy. Acta Neurol. Scand. Suppl. 100:133, 1984. 278. Pellissier, J. F., Pouget, J., Cros, D., et al.: Peripheral neuropathy induced by amiodarone chlorhydrate. J. Neurol. Sci. 63:251, 1984. 279. Pilcher, J., Chandrasekhar, K. P., and Russell Rees, J.: Long term assessment of perhexilene maleate in angina pectoris. Postgrad. Med. J. 49:115, 1973. 280. Placidi, G. F., Masouka, D., Alcaraz, A., et al.: Distribution and metabolism of 14C-metronidazole in mice. Arch. Int. Pharmacodyn. Ther. 188:168, 1970. 281. Pollet, S., Hauw, J. J., Escourolle, R., and Baumann, N.: Peripheral nerve lipid abnormalities in patients on perhexilene maleate. Lancet 1:1258, 1977. 282. Polomano, R. C., Mannes, A. J., Clark, U. S., and Bennett, G. J.: A painful peripheral neuropathy in the rat produced by the chemotherapeutic drug, paclitaxel. Pain 94:293, 2001. 283. Posada de la Paz, M., Philen, R. M., and Borda, A. I.: Toxic oil syndrome: the perspective after 20 years. Epidemiol. Rev. 23:231, 2001. 284. Postma, T. J., Vermoken, J. B., Liefting, A. J., et al.: Paclitaxel-induced neuropathy. Ann. Oncol. 6:489, 1995. 285. Powell, R. J., and Gardner-Medwin, J. M.: Guidelines for the clinical use and dispensing of thalidomide. Postgrad. Med. 70:901, 1994.

286. Prakash, V., and Timasheff, S. N.: Aging of tubulin at neutral pH: the destabilizing effect of vinca alkaloids. Arch. Biochem. Biophy. 295:137, 1992. 287. Pronk, L. C., Hilkins, P. H., van den Bent, M. J., et al.: Corticosteroid co-medication does not reduce the incidence and severity of neurotoxicity induced by docetaxel. Anticancer Drugs 9:759, 1998. 288. Pulipaka, U., Lacomis, D., and Omalu, B.: Amiodaroneinduced neuromyopathy: three cases and a review of the literature. J. Clin. Neuromuscul. Dis. 3:97, 2002. 289. Raabe, D. S. Jr.: Treatment of variant angina pectoris with perhexilene maleate. Chest 75:152, 1979. 290. Rago, R. P., Miles, J. M., Sufit, R. L., et al.: Suramininduced weakness from hypophosphatemia and mitochondrial myopathy: association of suramin with mitochondrial toxicity in humans. Cancer 73:1954, 1994. 291. Rajagopalan, R., and Gurnani, S.: Inhibition of tubulin assembly by anti-leprosy drug dapsone. Biochem. Biophys. Res. Commun. 116:128, 1983. 292. Ramirez, J. A., Mendell, J. R., Warmolts, J. R., and Griggs, R. C.: Phenytoin neuropathy: structural changes in the sural nerve. Ann. Neurol. 19:162, 1986. 293. Rana, S. S., Giuliani, M. J., Oddis, C. V., and Lacomis, D.: Acute onset of colchicine myoneuropathy in cardiac transplant recipients: case studies of three patients. Clin. Neurol. Neurosurg. 99:266, 1997. 294. Rao, D. G., Kane, N. K., and Oware, A.: Thalidomide neuropathy: role of F-wave monitoring. Muscle Nerve 23:1301, 2000. 295. Rao, K. S. P., Fernando, J. C., Ho, I. K., and Mehendale, H. M.: Neurotoxicity in rats following subchronic amiodarone treatment. Res. Commun. Chem. Pathol. Pharmacol. 52:217, 1986. 296. Rao, K. S. P., Rao, S. B., Camus, P. H., and Mehendale, H. M.: Effect of amiodarone on Na⫹,K⫹-ATPase and Mg2⫹-ATPase activities in rat brain synaptosomes. Cell Biochem. Funct. 4:143, 1986. 297. Reasor, M. J., Ogle, C. L., Walker, E. R., and Kacew, S.: Amiodarone-induced phospholipidosis in rat alveolar macrophages. Am. Rev. Respir. Dis. 137:510, 1988. 298. Richardson, P., Hideshima, T., and Anderson, K.: Thalidomide: emerging role in cancer medicine. Annu. Rev. Med. 53:629, 2002. 299. Riggs, J. E., Schochet, S. S., Gutmann, L., et al.: Chronic human colchicine neuropathy and myopathy. Arch. Neurol. 43:521, 1986. 300. Robberecht, W., Bednarik, J., Bourgeois, P., et al.: Myasthenic syndrome caused by direct effect of chloroquine on neuromuscular junction. Arch. Neurol. 46:464, 1989. 301. Roden, D. M.: Antiarrhythmic drugs. In Goodman, L. S., and Gilman, A. (eds.): The Pharmacologic Basis of Therapeutics, 9th ed. New York, Pergamon, p. 858, 1996. 302. Rosenberg, H., Orkin, F. K., and Springstead, J.: Abuse of nitrous oxide. Anesth. Analg. 58:104, 1979. 303. Ross, R. R.: Use of pyridoxine hydrochloride to prevent isoniazid toxicity. J. Am. Med. Assoc. 168:73,1958. 304. Rothrock, J. F., Johnson, P. C., Rothrock, S. M., and Merkley, R.: Fulminant polyneuritis after overdose of disulfiram and ethanol. Neurology 34:357, 1984.

Neuropathy Caused by Drugs 305. Rowinsky, E. K., and Donehower, R. C.: Paclitaxel (Taxol). N. Engl. J. Med. 332:1004, 1995. 306. Roytta, M., and Raine, C. S.: Taxol-induced neuropathy: chronic effects of local injection. J. Neurocytol. 15:483, 1986. 307. Russell, J. W., Gill, J. S., Sorenson, E. J., et al.: Suramininduced neuropathy in an animal model. J. Neurol. Sci. 192:71, 2001. 308. Russell, J. W., Windebank, A. J., and Podratz, J. L.: Role of nerve growth factor in suramin neurotoxicity studied in vitro. Ann. Neurol. 36:221, 1994. 309. Rutkove, S. B., DeGirolami, U., Preston, D. C., et al.: Myotonia in colchicine myoneuropathy. Muscle Nerve 19:870, 1996. 310. Sadoh, D. R., Hawk, J. L. M., and Panayiotopoulos, C. P.: Chronodispersion in patients on thalidomide. Clin. Neurophysiol. 110:735, 1999. 311. Sagman, D. L., and Melamed, J. C.: L-tryptophan-induced eosinophilia-myalgia syndrome and myopathy. Neurology 40:1629, 1990. 312. Sahenk, Z., Barohn, R., New, P., and Mendell, J. R.: Taxol neuropathy: electrodiagnostic and sural nerve biopsy findings. Arch. Neurol. 51:726, 1994. 313. Sahenk, Z., Brady, S. T., and Mendell, J. R.: Studies on the pathogenesis of vincristine-induced neuropathy. Muscle Nerve 10:80, 1987. 314. Sahenk, Z., Mendell, J. R., Couri, D., and Nachtman, J.: Polyneuropathy from inhalation of N2O cartridges through a whipped cream dispenser. Neurology 28:485, 1978. 315. Said, G.: Perhexilene neuropathy: a clinicopathological study. Ann. Neurol. 3:259, 1978. 316. Said, G., Goasguen, J., and Laverdant, C.: Neuropathy in long term treatment with metronidazole. Rev. Neurol. 134:515, 1978. 317. Sanfeliu, C., Wright, J. M., and Kim, S. U.: Neurotoxicity of isoniazid and its metabolites in cultures of mouse dorsal root ganglion neurons and hybrid neuronal cell line. Neurotoxicology 20:935, 1999. 318. Santoro, L., Barbieri, F., Nucciotti, R., et al.: Amiodaroneinduced experimental acute neuropathy in rats. Muscle Nerve 15:788, 1992. 319. Sasai, K., Shibamoto, Y., Takahashi, M., et al.: Pharmacokinetics of 2-nitroimidazole hypoxic cell radiosensitizers in rodent peripheral nerve tissue. Int. J. Radiat. Biol. 57:971, 1990. 320. Saunders, M. E., Dische, S., Anderson, P., and Flockhart, I. R.: The neurotoxicity of misonidazole and its relationship to dose, half-life and concentration in the serum. Br. J. Cancer 37(Suppl. III):268, 1978. 321. Scharer, K.: Selective alterations of Purkinje cells in the dog after oral administration of high doses of nitroimidazole derivatives [author’s transl.]. Verh. Dtsch. Ges. Pathol. 56:407, 1972. 322. Schaumburg, H. H., Kaplan, J., Windebank, A., et al.: Sensory neuropathy from pyridoxine abuse: a new megavitamin syndrome. N. Engl. J. Med. 309:445, 1983. 323. Schaumburg, H. H., Ludolph, A. C., and Gold, B. G.: Tacrolimus. In Spencer, S., and Schaumburg, H. H. (eds.): Experimental and Clinical Neurotoxicology. New York, Oxford University Press, p. 1131, 2000.

2581

324. Schilling, R. F.: Is nitrous oxide a dangerous anesthetic for vitamin B12-deficient subjects? JAMA 255:1605, 1986. 325. Schilling, T., Henrich, B., Kau, R., et al.: Paclitaxel administered over 3 hr. followed by cisplatin in patients with advanced head and neck squamous cell carcinoma: a clinical Phase I study. Oncology 54:89, 1997. 326. Schindler, K. S., Zurakowski, D., and Dryer, E. B.: Caspase inhibitors block zinc-chelator induced death or retinal ganglion cells. Neuroreport 11:2299, 2000. 327. Schmidt, I. G., and Schmidt, L. H.: Studies of the neurotoxicity of ethambutol and its racemate for the rhesus monkey. J. Neuropathol. Exp. Neurol. 25:40, 1966. 328. Schmued, L. C., Albertson, C. M., Andrews, A., et al.: Evaluation of brain and nerve pathology in rats chronically dosed with ddI or isoniazid. Neurotoxicol. Teratol. 18:555, 1996. 329. Schroder, J. M., and Matthiesen, T.: Experimental thalidomide neuropathy: the morphological correlate of reduced conduction velocity. Acta Neuropathol. 65:285, 1985. 330. Schutz, H. J., Busse, O., and Vuia, O.: Polyneuropathy caused by disulfiram poisoning. Arch. Psychiatr. Nervenkr. 233:1, 1983. 331. Schwab, B. W., Arezzo, J. C., Paladino, A. M., et al.: Rabbit sural nerve responses to chronic treatment with thalidomide and supidimide. Muscle Nerve 7:362, 1984. 332. Scott, J. M., Wilson, P., Dinn, J. J, and Weir, D. G.: Pathogenesis of subacute combined degeneration: a result of methyl group deficiency. Lancet 2:362, 1981. 333. Sebille, A.: Prevalence of latent perhexilene neuropathy. Br. Med. J. 2:1321, 1978. 334. Seguin, P., Camus, C., Leroy, J. P., et al.: Respiratory failure associated with hydroxychloroquine neuromyopathy. Eur. Neurol. 35:236, 1995. 335. Seiden, D.: Effects of colchicine on myofilament arrangement and the lysosomal system in skeletal muscle. Z. Zellforsch. 144:467, 1973. 336. Selwa, J. F., Feldman, E. L., and Blaivas, M.: Mononeuropathy multiplex in tryptophan-associated eosinophilia-myalgia syndrome. Neurology 40:1632, 1990. 337. Sghirlanzoni, A., Mantegazza, R., Mora, M., et al.: Chloroquine myopathy and myasthenia-like syndrome. Muscle Nerve 11:114, 1988. 338. Shah, R. R., Oates, N. S., Idle, J. R., et al: Impaired oxidation of debrisoquine in patients with perhexilene neuropathy. Br. Med. J. (Clin. Res. Ed.) 1:295, 1982. 339. Shah, R. R., Oates, N. S., Idle, J. R., et al.: Prediction of subclinical perhexilene neuropathy in a patient with inborn error of debrisoquine hydroxylation. Am. Heart J. 105:159, 1983. 340. Shields, R. W.: Peripheral neuropathies. In Rakel, R. E. (ed.): Conn’s Current Therapy. Philadelphia, W. B. Saunders, p. 915, 1996. 341. Shorvon, S. D., and Reynolds, E. H.: Anticonvulsant peripheral neuropathy: a clinical and electrophysiological study of patients on single drug treatment with phenytoin, carbamazepine or barbiturates. J. Neurol. Neurosurg. Psychiatry 45:620, 1982. 342. Silver, R. M., Sutherland, S. E., Carreira, P., and Heyes, M. P.: A murine model of the eosinophilia-myalgia syndrome induced by 1,1⬘-ethylidenebis (L-tryptophan). J. Clin. Invest. 93:1473, 1994.

2582

Diseases of the Peripheral Nervous System

343. Singh, N. T., and Pesantl, E. L.: Tuberculosis and other mycobacterial disease. In Rakel, R. E. (ed.): Conn’s Current Therapy. Philadelphia, W. B. Saunders, p. 226, 1996. 344. Singlas, E., Goujet, M. A., and Simon, P.: Pharmacokinetics of perhexilene maleate in anginal patients with and without peripheral neuropathy. Eur. J. Clin. Pharmacol. 14:195, 1978. 345. Sirsat, A. M., Lalitha, V. S., and Pandya, S. S.: Dapsone neuropathy—report of three cases and pathologic features of a motor nerve. Int. J. Lepr. Other Mycobact. Dis. 55:23, 1987. 346. Smith, B. E., and Dyck, P. J.: Peripheral neuropathy in the eosinophilia-myalgia syndrome associated with Ltryptophan ingestion. Neurology 40:1035, 1990. 347. Soliven, B., Dhand, U. K., Kobayashi, K., et al.: Evaluation of neuropathy in patients on suramin treatment. Muscle Nerve 20:83, 1997. 348. Stacy, C. B., DiRocco, A., and Gould, R. J.: Methionine in the treatment of nitrous-oxide induced neuropathy and myeloneuropathy. J. Neurol. 239:401, 1992. 349. Stahlberg, D., Barany, F., Elinarsson, K., et al.: Neurophysiologic studies of patients with Crohn’s disease on long-term treatment with metronidazole. Scand. J. Gastroenterol. 26:219, 1991. 350. Stein, M., Bell, M. J., and Ang, L. C.: Hydroxychloroquine neuromyotoxicity. J. Rheumatol. 27:2927, 2000. 351. Stephens, W. P., Eddy, J. D., Parson, I. M., and Singh, S. P.: Raised intracranial pressure due to perhexilene maleate. Br. Med. J. 1:21, 1978. 352. Stevenson, M. A., Calderwood, S. K., and Coleman, C. N.: Effects of nitro-imidazoles on neuronal cells in vitro. Int. J. Radiat. Oncol. Biol. Phys. 16:1225, 1989. 353. Sugita, H., Higuchi, I., Sano, M., and Ishiura, S.: Trial of a cysteine proteinase inhibitor, EST, in experimental chloroquine myopathy in rats. Muscle Nerve 10:516, 1987. 354. Sullivan, M., Follis, R. H., and Hilgartne, M.: Toxicology of podophyllin. Proc. Soc. Exp. Biol. Med. 77:269, 1951. 355. Swift, T. R., Gross, J. A., Ward, L. C., and Crout, B. O.: Peripheral neuropathy in epileptic patients. Neurology 31:826, 1981. 356. Swygert, L. A., Back, E. E., Auerbach, S. B., et al.: Eosinophilia-myalgia syndrome: mortality data from the U.S. National Surveillance System. J. Rheumatol. 20:1711, 1993. 357. Swygert, L. A., Maes, E. F., Sewell, L. E., et al.: Eosinophilia-myalgia syndrome: results of national surveillance. JAMA 264:1698, 1990. 358. Takeuchi, H., Yamada, A., Touge, T., et al.: Metronidazole neuropathy: a case report. Jpn. J. Psychiatry Neurol. 42:291, 1988. 359. Tanhehco, J. L., Weichers, D. O., Golbus, J., and Neeley, S. E.: Eosinophilia-myalgia syndrome: myopathic electrodiagnostic characteristics. Muscle Nerve 15:561, 1992. 360. Tapal, M. F.: Colchicine myopathy. Scand. J. Rheumatol. 25:105, 1996. 361. Taylor, J. W., Murphy, M. J., and Rivey, M. P.: Clinical and electrophysiologic evaluation of peripheral nerve function in chronic phenytoin therapy. Epilepsia 26:416, 1985. 362. Tegner, R., Tome, F. M., Godeau, P., et al.: Morphological study of peripheral nerve changes induced by chloroquine treatment. Acta Neuropathol. 75:253, 1988.

363. Tester-Dalderup, C. B. M.: Metronidazole and related compounds. In Dukes, M. N. G. (ed.): Myeler’s Side Effects of Drugs, 12th ed. Amsterdam, Elsevier, p. 705, 1992. 364. Thant, M., Hawley, R. J., Smith, M. T., et al.: Possible enhancement of vincristine neuropathy by VP-16. Cancer 49:859, 1982. 365. Tomczak, R. L., and Hake, D. H.: Near fatal systemic toxicity from local injection of podophyllin for pedal verrucae treatment. J. Foot Surg. 31:36, 1992. 366. Tomlinson, I. W., and Rosenthal, F. D.: Proximal myopathy after perhexilene maleate treatment. Br. Med. J. 1:1319, 1977. 367. Tonkin, E. G., Erve, J. C., and Valentine, W. M.: Disulfiram produces a non-carbon disulfide-dependent schwannopathy in the rat. J. Neuropathol. Exp. Neurol. 59:786, 2000. 368. Toole, J. F., Gergen, J. A., Hayes, D. M., et al.: Neural effects of nitrofurantoin. Arch. Neurol. 18:680, 1968. 369. Toole, J. F., and Parrish, M. L.: Nitrofurantoin polyneuropathy. Neurology 23:554, 1973. 370. Tugwell, P., and James, S. L.: Peripheral neuropathy with ethambutol. Postgrad. Med. 48:667, 1972. 371. Turi, G. K., Solitaire, G. B., James, N., and Dicker, R.: Eosinophilia-myalgia syndrome (L-tryptophan-associated neuromyopathy). Neurology 40:1793, 1990. 372. Unger, S. A., Robinson, M. A., and Horowitz, J. D.: Perhexilene improves symptomatic status in elderly patients with severe aortic stenosis. Aust. N. Z. J. Med. 27:24, 1997. 373. Urtasan, R. C., Chapman, J. D., Feldstein, M. L., et al.: Peripheral neuropathy related to misonidazole: incidence and pathology. Br. J. Cancer 37(Suppl. III):271, 1978. 374. Urtasan, R. C., Rabin, H. R., and Partington, J.: Human pharmacokinetics and toxicity of high-dose metronidazole administered orally and intravenously. Surgery 93:145, 1983. 375. Vahdat, L., Papadopoulos, K., Lange, D., et al.: Reduction of paclitaxel-induced peripheral neuropathy with glutamine. Clin. Cancer Res. 7:1192, 2001. 376. van den Bent, M. J., Hilkens, P. H., Sillevis Smitt, P. A., et al.: Lhermitte’s sign following chemotherapy with docetaxel. Neurology 50:563, 1998. 377. van den Bent, M. J., van Raaij-van den Aarssen, V. J., Verweij, J., et al.: Progression of paclitaxel-induced neuropathy following discontinuation of treatment. Muscle Nerve 20:750, 1997. 378. Van Rossum, J., Ross, R. A., and Bots, G. T.: Disulfiram polyneuropathy. Clin. Neurol. Neurosurg. 86:81, 1984. 379. Vishnubhakat, S. M., and Beresford, H. R.: Reversible myeloneuropathy of nitrous oxide abuse: serial electrophysiological studies. Muscle Nerve 14:22, 1991. 380. von Rogulja, P., Kovac, W., and Schmid, H.: Metronidazole encephalopathy in rats. Acta Neuropathol. (Berl.) 25:36, 1973. 381. Vrobel, T. R., Miller, P. E., Mostow, N. D., and Rakita, L.: A general overview of amiodarone toxicity: its prevention, detection, and management. Prog. Cardiovasc. Dis. 31:393, 1989. 382. Waldinger, T. P., Siegle, R. J., Weber, W., and Voorhees, J. J.: Dapsone-induced peripheral neuropathy: case report and review. Arch. Dermatol. 120:356, 1984. 383. Wang, M. S., Davis, A. A., Culver, D. G., and Glass, J. D.: WldS mice are resistant to paclitaxel (Taxol) neuropathy. Ann. Neurol. 52:442, 2002.

Neuropathy Caused by Drugs 384. Wang, M. S., Fang, G., Culver, D. G., et al.: The WldS protein protects against axonal degeneration: a model of gene therapy for peripheral neuropathy. Ann. Neurol. 50:773, 2001. 385. Wasay, M., Wolfe, G. E., Herrold, J. M., et al.: Chloroquine myopathy and neuropathy with elevated CSF protein. Neurology 51:1226, 1998. 386. Watson, C. P., Ashby, P., and Bilbao, J. M.: Disulfiram neuropathy. Can. Med. Assoc. J. 123:123, 1980. 387. Weller, A., Rosenberg, N., Schlitz, P., et al.: Tryptophan causes EMS-like skin changes and induction of ICAM and LFA in SJL/J mice. J. Immunol. 150:172A, 1993. 388. Werner, E. G., and Olanow, C. W.: Parkinsonism and amiodarone therapy. Neurology 25:630, 1989. 389. Whisnant, J. P., Espinosa, R. E., Kierland, R. R., and Lambert, E. H.: Chloroquine neuromyopathy. Proc. Mayo Clin. 38:501, 1963. 390. Wijdicks, E. F. M., Wiesner, R. H., Dahlke, L. J., and Krom, R. A. F.: FK-506-induced neurotoxicity in liver transplantation. Ann. Neurol. 35:498, 1994. 391. Wijesekera, J. C., Critchley, E. M. R., Fahim, Y., et al.: Peripheral neuropathy due to perhexilene maleate. J. Neurol. Sci. 46:303, 1980. 392. Williams, M. E., Walker, A. N., Bracikowski, J. P., et al.: Ascending myelo-encephalopathy due to intrathecal vincristine sulfate. Cancer 51:2041, 1983. 393. Williams, M. H., and Bradley, W. G.: An assessment of dapsone toxicity in the guinea pig. Br. J. Dermatol. 86:650, 1972. 394. Wilson, J. R., Conwit, R. A., Eidelman, B. H., et al.: Sensorimotor neuropathy resembling CIDP in patients receiving FK506. Muscle Nerve 17:528, 1994. 395. Wilson, L., Bamburg, J. R., Mizel, S. B., et al.: Interaction of drugs with microtubule proteins. Fed. Proc. 33:158, 1974. 396. Windebank, A. J., Low, P. A., and Blexrud, M. D.: Pyridoxine neuropathy in rats: degeneration of sensory axons. Neurology 35:1617, 1985. 397. Wisniewski, H., Shelanski, M. L., and Terry, R. D.: Effects of mitotic spindle inhibitors on neurotubules and neurofilaments in anterior horn cells. J. Cell Biol. 38:224, 1968. 398. Wright, G. J., Leeson, G. A., Zeiger, A. V., and Lang, J. F.: The absorption, excretion, and metabolism of perhexilene maleate by the human. Postgrad. Med. 49:8, 1973. 399. Wulff, E. A., Wang, A. K., and Simpson, D. M.: HIVassociated peripheral neuropathy: epidemiology, pathophysiology, and treatment. Drugs 59:1251, 2000. 400. Wyatt, E. H., and Stevens, J. C.: Dapsone induced peripheral neuropathy. Br. J. Dermatol. 86:521, 1972.

2583

401. Xu, Y., Sladky, J. T., and Brown, M. B.: Dose-dependent expression of neuropathy after experimental pyridoxine intoxication. Neurology 39:1077, 1989. 402. Yakisich, J. S., Siden, A., Eneroth, P., and Cruz, M.: Disulfiram is a potent in vitro inhibitor of DNA topoisomerases. Biochem. Biophys. Res. Commun. 289:586, 2001. 403. Yamaguchi, T., Katoh, I., and Kurata, S.: Azidothymidine causes functional and structural destruction of mitochondria, glutathione deficiency and HIV-promotor sensitization. Eur. J. Biochem. 269:2782, 2002. 404. Yamamoto, M., Sobue, G., Mukoyama, M., et al.: Demonstration of slow acetylator genotype of N-acetyltransferase in isoniazid neuropathy using an archival hematoxylin and eosin section of a sural nerve biopsy specimen. J. Neurol. Sci. 135:51, 1996. 405. Yamamoto, T.: Pathologic processes of lumbar primary sensory neurons produced by high doses of pyridoxine in rats— morphometric and electron microscopic studies. Sangyo Ika Daigaku Zasshi 13:109, 1991. 406. Yamanaka, Y., Sakamoto, E., Sakuma, Y., et al.: Lipidosis of the dorsal root ganglia in rats with almitrine metabolite. Arch. Toxicol. 69:391, 1995. 407. Yamanaka, Y., Shimada, T., Mochizuki, R., et al.: Neuronal and muscular inclusions in rats with hindlimb dysfunction after treating with difluorobenzyhydrylpiperadine. Toxicol. Pathol. 25:150, 1997. 408. Yang, C. M., Deng, J. F., Chen, C. F., and Chang, L. W.: Experimental podophyllotoxin (bajiaolian) poisoning: III. Biochemical bases for toxic effects. Biomed. Environ. Sci. 7:259, 1994. 409. Yarchoan, R., Perno, C. R., Thomas, R. V., et al.: Phase I studies of 2⬘,3⬘-dideoxycytidine in severe human immunodeficiency virus infection as a single agent and alternating with zidovudine (AZT). Lancet 1:76, 1988. 410. Yiahn, C. H., Paniello, R. C., and Gershon-Spector, J.: Inhibition of motor nerve regeneration in a rabbit facial nerve mode. Laryngoscope 111:786, 2001. 411. Yiannikas, C., Pollard, J. D., and McLeod, J. G.: Nitrofurantoin neuropathy. Aust. N. Z. J. Med. 11:400, 1981. 412. Yoshikawa, H., Abe, T., and Oda, Y.: Extremely acute phenytoin-induced peripheral neuropathy. Epilepsia 40:528, 1999. 413. Younger, D. S., Mayer, S. A., Weimer, L. H., et al.: Colchicine-induced myopathy and neuropathy. Neurology 41:943, 1991. 414. Zivkovic, S. A., Lacomis, D., and Giuliani, M. J.: Sensory neuropathy associated with metronidazole: report of four cases and review of the literature. J. Clin. Neuromuscul. Dis. 3:8, 2001.

Q. Peripheral Nerve Tumors

115 Tumors and Tumor-like Conditions of Peripheral Nerve CATERINA GIANNINI

Benign Tumors of Peripheral Nerve Schwannoma and Its Variants Neurofibroma Perineurioma Miscellaneous Benign Peripheral Nerve Tumors

Malignant Tumors of Peripheral Nerve Malignant Peripheral Nerve Sheath Tumor and Its Variants Other Malignant Peripheral Nerve Tumors Tumor-like Conditions Affecting Peripheral Nerve

The peripheral nervous system (PNS) tumor spectrum encompasses a wide variety of tumors, largely exhibiting differentiation along cell types intrinsic and specific to peripheral nerve. The most common primary tumors therefore show differentiation along neurogenic lines. Neoplastic cells resemble in their differentiation Schwann cells, the most numerous among endoneurial cells and/or perineurial cells—the highly specialized cells that, with their well-developed tight junctions, contribute significantly to maintenance of the blood-nerve barrier. Tumors primary in nerve may also derive from mesenchymal elements, present in all three nerve compartments (endo-, peri-, and epineurium), such as fibroblasts, endothelial cells, and pericytes, as well as adipose cells. Finally, although less frequently, the PNS may be involved by tumors originating at other sites and secondarily involving peripheral nerve by metastasis or continuous spread. The currently accepted classification of peripheral nerve tumors is summarized in Table 115–1. This classification follows the general guidelines of the World Health Organization (WHO),54 as well as the recently published monograph on peripheral nerve tumors in the third Armed

Traumatic Neuroma Palisaded Encapsulated Neuroma Lipofibromatous Hamartoma and Neuromuscular Choristoma

Forces Institute of Pathology (AFIP) series, Atlas of Tumor Pathology.90 More frequently that any other group of central nervous system tumors, peripheral nerve neoplasms occur in conjunction with familial tumor syndromes, in particular with the neurofibromatoses (NFs). Neurofibromatosis type 1 (von Recklinghausen’s disease, peripheral NF, NF1) and type 2 (central NF, NF2) are both associated with an increased frequency of peripheral nerve tumors, benign and malignant. A discussion of the genetic basis and clinical spectrum of these disorders appears elsewhere in the volume. In addition to tumors, the PNS may also be affected by a variety of reactive, inflammatory, hyperplastic, and malformative processes that clinically and by imaging may simulate a peripheral nerve neoplasm. The spectrum of these “tumor mimics,” also referred to as tumor-like conditions of peripheral nerve, is summarized in Table 115–2. In order to correctly interpret the variety of lesions described in this chapter, as well as distinguish neoplastic from non-neoplastic conditions, familiarity with normal PNS anatomy, as well as with the basic mechanisms of response to injury exhibited by its highly specialized 2585

2586

Diseases of the Peripheral Nervous System

Table 115–1. Peripheral Nerve Tumors Benign

Malignant

Schwannoma Conventional Cellular Plexiform Melanotic (psammomatous)

MPNST Epithelioid With divergent differentiation Melanotic (malignant melanotic schwannoma) Arising in schwannoma Arising in ganglioneuroma/ ganglioneuroblastoma Arising in paraganglioma Perineurial

Neurofibroma Diffuse Localized Plexiform Massive soft tissue Perineurioma Intraneural Soft tissue Nerve sheath myxoma Neurothekeoma Granular cell tumor Ganglioneuroma Others: paraganglioma, lipoma, hemangioma, “angiomatosis,” hemangioblastoma, etc.

PNET/neuroepithelioma Malignant granular cell tumor Lymphoma Secondary tumors Carcinoma Melanoma Sarcoma Lymphoma

MPNST ⫽ malignant peripheral nerve sheath tumor; PNET ⫽ primitive neuroectodermal tumor.

Table 115–2. Tumor-like Conditions Affecting Peripheral Nerve Reactive Lesions Traumatic neuroma Localized interdigital neuritis Pacinian neuroma Nerve cysts Endometriosis of sciatic nerve Heterotopic ossification of nerve Inflammatory and Infectious Lesions Inflammatory pseudotumor of nerve Sarcoidosis of nerve Leprous neuropathy Hyperplastic Lesions Palisaded encapsulated neuroma Mucosal neuroma, neuromatosis, and ganglioneuromatosis of MEN IIB Mucosal neuroma, ganglioneuroma/ganglioneuromatosis and neuronal intestinal dysplasia unassociated with MEN IIB Localized hypertrophic neuropathy Hamartoma and Choristoma Lipofibromatous hamartoma of nerve Neuromuscular choristoma MEN IIB ⫽ multiple endocrine neoplasia type IIB.

components is essential. The reader is directed to the appropriate chapters on anatomy, histology, and response to injury in the first volume of this text. Finally, the reader is reminded that, as for every diagnosis in surgical neuropathology, critical and attentive correlation of morphologic data with clinical, imaging, and surgical data is imperative to arrive at a correct diagnosis.

BENIGN TUMORS OF PERIPHERAL NERVE Schwannoma and Its Variants By definition, schwannomas are benign, usually encapsulated tumors composed of differentiated neoplastic Schwann cells. According to the WHO, schwannomas correspond histologically to grade I tumors (WHO grade I).114 The use of the old synonyms neurilemoma and neurinoma is discouraged. Schwannoma is the most common of peripheral nerve tumors, generally occurring in the head and neck region and in the extremities, more frequently along the flexor surfaces.90 Schwannoma also accounts for approximately 8% of intracranial tumors, largely involving the vestibular division of the eighth cranial nerve,14 and less than a third of primary spinal tumors.40 Most conventional schwannomas are solitary, sporadic tumors. Multiple schwannomas typically occur in patients with NF2. Recently, a genetically distinct disorder, schwannomatosis, has been described48,68 that is also associated with multiple, generally subcutaneous schwannomas, but without vestibular tumors or other manifestations of NF2. Mutations of the NF2 gene have also been found in a large proportion of sporadic schwannomas.49 This gene, which is located on chromosome 22q12, encodes for Merlin, a protein closely related to the protein 4.1 family of cytoskeleton-associated proteins.54,65,67 Rarely schwannomas have been associated with radiation.87,93,99 These are solitary lesions histologically indistinguishable from conventional schwannomas. Schwannoma can occur at any age, although its incidence appears to peak between the fourth and sixth decades. The tumor occurs with equal frequency in males and females, with intracranial lesions, which favor females with a ratio of 2:1, being the only exception. Schwannomas arising intracranially and intraspinally most frequently affect sensory nerves, such as the vestibular nerve(s)55 and dorsal spinal nerve roots. Infrequently, schwannomas can be entirely intraparenchymal14 or intramedullary,40 with rare reported cases having an intraventricular location.14,78,105 Schwannomas occurring intraspinally in the cervicothoracic area frequently involve the intervertebral foramen and show an intra- and an extradural component with a characteristic “dumbbell” configuration. At the level of the cauda, they most frequently

Tumors and Tumor-like Conditions of Peripheral Nerve

present as sausage-shaped, entirely intradural masses, located distal to the conus medullaris and displacing surrounding nerve roots. Schwannomas arising from nerve roots may be located in the paraspinal region, posterior mediastinum, and retroperitoneum. Peripheral schwannomas most frequently occur in the head and neck region and the flexor surfaces of the extremities. They are generally deep in location. Only in a minority of cases is the parent nerve identified. Rare schwannomas are “intraosseous” in location,25 with the mandible, wherein they arise at the level of the dental foramen,97 and the vertebral body79 being the most frequently affected sites. Very large schwannomas displacing and in part destroying the sacrum, sometimes referred to as giant sacral schwannomas because of their appearance, have been reported.90 Rare visceral schwannomas, most frequently affecting the gastrointestinal tract (particularly the stomach) have also been reported.90 On magnetic resonance imaging (MRI), schwannomas appear as well-circumscribed masses, solid or cystic. They frequently show heterogeneous enhancement, reflecting the presence of degenerative changes such as cystic degeneration. Macroscopic Features Schwannomas are typically globular or ovoid, well circumscribed, and smooth surfaced (Fig. 115–1). Peripheral tumors are usually surrounded by a well-formed capsule, often thick and fibrocollagenous. Their size ranges from a few centimeters up to 10 cm in greatest diameter. A parent nerve is identifiable in less than half of cases, most often when the tumor arises in a sizable nerve (Fig. 115–2). Infrequently, tumors are multinodular or plexiform, especially when located in skin or subcutaneous tissue. The cut surface typically varies in color from tan to bright yellow, depending on the degree of lipid accumulation. Cystic degeneration and hemorrhage are common, particularly in large tumors (see Fig. 115–1). Areas of necrosis are typically not present.

FIGURE 115–1 Macroscopic features of two examples of schwannoma on cut surface. A, Areas of cystic degeneration are prominent in this example of a large retroperitoneal schwannoma. B, Cystic changes and hemorrhage are present in this pelvic schwannoma.

2587

Microscopic Features Schwannomas consist entirely of neoplastic Schwann cells. The advanced schwannian differentiation is well recognizable at the light microscopic, immunohistochemical, and ultrastructural level. Histologically, neoplastic Schwann cells are arranged in two basic patterns termed Antoni A and B (Fig. 115–3). The Antoni A pattern is characterized by closely packed spindled cells with large, elongate nuclei and ample eosinophilic cytoplasm with no discernible cell borders. Nuclei tend to palisade, a helpful diagnostic future. Verocay bodies, resulting from prominent palisading of nuclei and cell processes, represent an extreme example of this phenomenon. Collagen is rather abundant. The Antoni B pattern consists of loosely textured areas with cells with multipolar processes arranged within a pale mucopolysaccharide-rich matrix. Lipidization and accumulation of foamy macrophages is frequent in Antoni B areas. Microcyst formation is commonly seen in these areas, as are thick-walled, hyalinized, and often ectatic vessels. Vascular thrombosis and perivascular hemosiderin deposition, resulting from repeated episodes of hemorrhage, are also common. The proportions of Antoni A and B patterns and the transition between them vary greatly from tumor to tumor, generating an ample morphologic spectrum. Schwannomas grow in a solid, noninfiltrative fashion and lie eccentric to their parent nerve. Nerve fibers therefore are not dispersed within the substance of the tumor, but, when identifiable, lie in bundles adjacent to or within its capsule. Nuclear degenerative atypia, with nuclear enlargement, marked pleomorphism, and hyperchromasia, is not uncommon in schwannomas (“ancient schwannomas”) and is of no prognostic significance. Cartilaginous, osseous, and even adipose metaplasia is occasionally seen in schwannomas. Pseudoepithelial change in the lining of degenerative cysts is a frequent finding, while the reported occurrence of “glandular” differentiation in cutaneous schwannomas probably represents entrapped cutaneous adnexa rather than divergent epithelial differentiation.90 Immunoreactivity for S100 protein is uniform and strong in schwannomas.50,51 Approximately 50% of schwannomas

2588

Diseases of the Peripheral Nervous System

FIGURE 115–2 Multiple (three) schwannomas arising along a single nerve root. The largest lesion, approximately 1 cm in diameter, has a nodular appearance eccentric to the parent nerve.

show immunoreactivity for glial fibrillary acidic protein.50,51 In addition, collagen IV and laminin stains highlight the presence of a pericellular basement membrane.70,90 Ultrastructurally, characteristic features of Schwann cells are the presence of cytoplasmic intermediate filaments, intertwining cytoplasmic processes occasionally surrounding collagen bundles (“mesaxon formation”), and pericellular basement membranes that are often reduplicated.27,90 Longspaced stromal collagen fibrils (130-nm periodicity), referred to as Luse bodies, are frequently seen. Cellular Schwannoma This variant of schwannoma is characterized histologically by high cellularity, an exclusive or predominant Antoni A pattern, and the absence of well-formed Verocay bodies. This tumor was originally described by Woodruff et al.111 Cellular schwannoma is a benign tumor and should not be confused with a well-differentiated malignant peripheral nerve sheath tumor (MPNST). In contrast to MPNST, treatment is strictly surgical and allows conservative surgery with sparing of the parent nerve. No adjuvant therapy is required. Cellular schwannomas occur at any age, with a

peak of incidence in the fourth decade. They are more frequent in females than males, with a ratio of approximately 2:1. The vast majority of the tumors are solitary, and there is no significant association with NF.49,90 Cellular schwannoma occurs more frequently in the paravertebral region of the mediastinum, retroperitoneum, and pelvis.15 Cranial nerves may also be affected, in particular the fifth and the eighth. Similarly to conventional schwannoma, cellular schwannoma is macroscopically well circumscribed and encapsulated. Its average size is 5 cm, but tumors up to 20 cm in diameter have been reported. On cut surface, cellular schwannoma is usually solid and more fleshy than conventional schwannoma. Areas of cystic degeneration and/or hemorrhage are infrequent. Microscopically, the high cellularity, resulting from the predominance of Antoni A tissue without Verocay body formation, may be an alarming feature when seen in a small biopsy coming from the center of the tumor (Fig. 115–4). On resection, however, the circumscribed, encapsulated nature of the tumor is apparent (Fig. 115–4). Lymphocytic infiltration just beneath its capsule and/or around vessels is quite characteristic. The cytologic features are those of schwannian cells with uniform, elongate nuclei as well as eosinophilic cytoplasm. Mitotic figures, as a rule, are readily identified. Mitotic indices of 1 to 4 per 10 high-power fields (HPF) are commonly seen, with exceptional examples showing 10 or more mitoses per 10 HPF. Cellular schwannomas show strong and diffuse immunoreactivity for S100 protein,15,17,29,106 a feature generally not seen in MPNST. Plexiform Schwannoma This variant of schwannoma is characterized by a plexiform or multinodular pattern of growth, presumably resulting from involvement of a nerve plexus. Plexiform schwannoma is generally superficial in location, affecting dermis and subcutaneous tissue. The most frequent location is in the extremities, followed by the head and neck region and trunk.90,115 Prognosis is excellent following local excision. Recurrence is infrequent. An association of plexiform schwannoma has been noted with NF2 but not NF1, and also in non-NF2 patients with multiple schwannomas (schwannomatosis).28,47 Melanotic Schwannoma Melanotic schwannoma is defined as a tumor composed of cells with ultrastructural and immunophenotypic features of Schwann cells, but in addition containing melanosomes and showing immunoreactivity with melanoma markers. Melanotic schwannoma is rare, its peak incidence being a decade earlier than conventional schwannomas. It occurs slightly more frequently in females, with a female-to-male ratio of 1.4:1. Melanin production is variable. In some cases, pigmentation is so great as to obscure cytologic details and to require hydrogen peroxide “bleaching.” Approximately 50% of melanotic schwannomas contain psammoma bodies

Tumors and Tumor-like Conditions of Peripheral Nerve

A

B

C

D

E

F

2589

FIGURE 115–3 Schwannoma: histologic features. Schwannomas are composed of neoplastic spindle cells with elongated nuclei. A, In Antoni A pattern, the cells may be haphazardly arranged. B and C, Their nuclei may align closely, generating an alternating pattern of palisading nuclei and eosinophilic cytoplasm. D, The transition from Antoni A areas (right) to loosely arranged, macrophage-rich Antoni B areas (left). E, Thick vessels with hyalinized walls are a characteristic and common finding. F, Cystic degeneration is not uncommon.

and are therefore called “psammomatous” melanotic schwannomas. Recognition of psammomatous melanotic schwannoma is important because approximately 50% of psammomatous tumors are associated with Carney’s complex. This is a dominantly inherited autosomal disorder characterized by spotty lentiginous pigmentation, often facial; myxomas of the heart, skin, or breast; and endocrine overactivity, including Cushing’s syndrome resulting from bilateral multinodular adrenal cortical hyperplasia and acromegaly resulting from pituitary adenoma.11–13 The vast majority of non-psammomatous tumors involve spinal nerves, while psammomatous tumors frequently affect autonomic nerves of the gastrointestinal tract and heart.

Cranial nerves may also be affected. Melanotic schwannomas are relatively demarcated but generally not encapsulated. Histologically, they are composed of variably pigmented spindle or epithelioid tumor cells (Fig. 115–5). A common feature is the presence of stromal macrophages, often more heavily pigmented than the tumor cells. Uniform immunoreactivity for S100 protein is the rule. The collagen IV and laminin staining pattern varies from dense and pericellular to lobular, surrounding clusters of neoplastic cells. At the ultrastructural level, neoplastic cells have features of well-differentiated Schwann cells, with prominent, continuous as well as reduplicated basement membranes. Melanosomes are present by definition.

2590

Diseases of the Peripheral Nervous System

A

B

C

D

FIGURE 115–4 Cellular schwannoma: gross and microscopic appearance. A, The well-circumscribed gross appearance of a cellular schwannoma. Focal cystic degeneration is present. B, Microscopically, in low power, a thick collagenous capsule is recognizable. The Antoni A pattern predominates (C), and mitotic figures are not uncommon (D).

A

B

C

D

FIGURE 115–5 Psammomatous melanotic schwannoma: histologic appearance. A and B, Cells with epithelioid morphology. Nuclei are large and nucleoli prominent. Also shown are a mitotic figure (A) and a small psammoma body (B). C, Heavily pigmented stromal macrophages are present. D, Collagen IV surrounds small clusters of neoplastic cells, producing a lobular pattern.

Tumors and Tumor-like Conditions of Peripheral Nerve

2591

Approximately 10% of melanotic schwannomas follow a malignant clinical course characterized by local recurrence and/or metastasis. Although it is difficult to predict clinical behavior based on histologic appearance, histologic features that have been associated with clinically malignant tumors include large vesicular nuclei, prominent macronucleoli, frequent mitoses, and ample necrosis.

Neurofibroma Neurofibroma is a benign peripheral nerve sheath tumor composed of a mixture of Schwann cells, perineurial cells, and fibroblasts, as well as cells with intermediate features. In contrast to schwannomas, neurofibromas grow in a permeative fashion and overrun myelinated and unmyelinated fibers of their parent nerve. According to the WHO, neurofibromas correspond histologically to grade I tumors (WHO grade I).114 Neurofibromas are common tumors and in the vast majority of cases are solitary, sporadic lesions. In a minority of cases (⬍10%), they occur in patients with NF1 as solitary or more often multiple lesions. They occur in patients of any age and show no sex predilection. Most frequently they present as superficial, dermal and/or subcutaneous nodules. Neurofibromas involving a major nerve trunk as either a circumscribed intraneural mass or a plexiform enlargement of multiple fascicles are uncommon. Based on their site of occurrence and pattern of growth, neurofibromas have been categorized into five distinct forms: localized and diffuse cutaneous neurofibromas, localized and plexiform intraneural neurofibromas, and a massive soft tissue variant. Localized cutaneous neurofibromas typically present as painless, freely movable, soft, nodular or polypoid masses. They are often 1 to 2 cm in greatest dimension. Diffuse cutaneous neurofibromas are characterized by diffuse permeation of dermis and subcutaneous tissue and often form large, ill-defined plaques. Their texture is similar to that of the localized form. It is not uncommon, especially in NF1, to find lesions with a mixed pattern of growth. Localized intraneural neurofibromas, such as those involving nerve trunks, present as fusiform enlargement of the nerve trunk. These lesions come to clinical attention as visible or palpable soft tissue masses when superficially located, or by causing tingling or pain in a nerve distribution when deeply located. Tumors associated with NF1 are often proximal and frequently involve spinal nerve roots, often at the cervical level.37,39,112 Intraneural neurofibromas are whitish and translucent in appearance (Fig. 115–6). Plexiform intraneural neurofibroma, which may affect small distal nerve branches as well as sizable nerves, has a characteristic appearance (Fig. 115–7). By definition, the neoplastic proliferation involves and spreads through multiple separate nerve fascicles. When the involved fascicles are nonbranching, it results in a ropy lesion in which twisted strands remain apposed side to side.90 When the

A

B FIGURE 115–6 Gross appearance of two examples of intraneural neurofibroma: the fusiform external appearance, with a smooth surface (A), and the cut surface, which appears whitish and glistening (B).

involved fascicles branch and/or bridge, as commonly seen in peripheral nerve, the tumor forms a complex tangle of intertwined fascicles whose appearance has been often compared to a “bag of worms.” Large plexiform neurofibromas may produce marked deformity of the involved regions. Visceral involvement occurs. The diagnosis of plexiform neurofibroma, when a major nerve is involved, is pathognomonic of NF1. Plexiform neurofibromas may undergo malignant transformation in approximately 5% of cases. Massive soft tissue neurofibromas, which commonly include a diffuse as well as a plexiform component, are also highly indicative of NF1. Localized gigantism or the formation of pendulous masses of neurofibromatous tissue, particularly around limb girdles, may result from the massive soft tissue enlargement. The term elephantiasis neuromatosa was used in the past to indicate these lesions. Histologically, neurofibroma is composed of an admixture of cells with schwannian and perineurial features as well as fibroblasts and cells with transitional features. Microscopically, the composition of the various forms of neurofibromas is similar (Fig. 115–8). Typical neurofibromas are generally hypocellular. The cells are widely

2592

Diseases of the Peripheral Nervous System

A

B

C FIGURE 115–7 Plexiform neurofibroma. Gross appearance of a large (A) and a small (C) example. Multiple nerve fascicles are involved by the process. They are distorted and deformed, forming complex multilobulated masses. Lesions such as the one shown in A, when palpated, produce a characteristic sensation described as a “bag of worms.” B, On cut sections, multiple separate juxtaposed fascicles involved by the neurofibromatous process are seen.

A

B

C

D

FIGURE 115–8 A–C, The histologic appearance of neurofibroma shows variable, but generally low cellularity. Cells have wavy elongated nuclei, which end in a tapered manner. The background is loosely arranged and myxoid (B). Bands of collagen of variable thickness are often present, especially near the center of the tumor (C). D, Different degrees of differentiation toward tactile organelles, such as this meissnerian differentiation, can be seen.

Tumors and Tumor-like Conditions of Peripheral Nerve

distributed in an abundant myxoid/mucoid background matrix. Cell nuclei are characteristically ovoid to spindle shaped, often curved, and tapered at their ends. They are smaller than the nuclei of schwannoma. Cytoplasmic processes are extremely thin and long and poorly discernible in hematoxylin and eosin (H&E) preparations. By immunohistochemistry, S100 protein–positive cells are present in all neurofibromas, but their number varies and is generally considerably lower than in schwannomas.42,90 Epithelial membrane antigen (EMA) stain is generally negative, presumably because of the extremely thin processes of perineurial-like cells. Reactivity is limited to residual perineurium at the edge of nerve fascicles. CD34 stain is generally positive. In the intraneural form of neurofibroma, preexisting, residual nerve fibers may be visible at the center of the fascicles, when they are cut perpendicularly or in favorable midlongitudinal sections. Neurofilament stain usually highlights these residual axons. Not infrequently neurofibromas may show “tactile differentiation” with presence of pseudomeissnerian corpuscles (see Fig. 115–8D) and/or whorls reminiscent of pacinian corpuscles. Schwannoma-like nodules can also be seen. The mixed population of cells composing neurofibroma is best seen at the ultrastructural level. Perineurial-like cells, so difficult to identify at the light microscopic level, are easily recognized. They have long, thin processes and frequent pinocytotic vesicles as well as short stretches of basement membranes. These cells, admixed with cells with schwannian and fibroblastic differentiation, are very characteristic of neurofibromas.27,90 The terms atypical neurofibroma and cellular neurofibroma have been used to identify neurofibromas that display either the presence of atypical, hyperchromatic, enlarged nuclei or diffuse hypercellularity.90,112 These features in neurofibroma should not be mistaken for malignancy. The presence of atypical cells, especially in the absence of mitotic activity, is considered the morphologic equivalent of “degenerative atypia” in schwannoma. According to Scheithauer et al.,90 the distinction of lowgrade MPNST from cellular neurofibroma is based on a triad of findings: (1) definite cell crowding; (2) nuclear enlargement, at least three times the size of ordinary neurofibroma nuclei; and (3) hyperchromasia. Low-level mitotic activity can be accepted in a cellular neurofibroma.

Perineurioma The term perineurioma is used to indicate benign tumors composed entirely of well-differentiated neoplastic perineurial cells. Two distinct varieties of perineurioma are recognized. One is an intraneural tumor with distinctive clinicopathologic features, also called intraneural perineurioma. The other is a soft tissue neoplasm known as soft tissue perineurioma.24,35,36,90 The recent cytogenetic

2593

demonstration of a clonal chromosome 22 abnormality in both lesions supports the view that both are part of a spectrum of perineurial neoplasia.24,35 Perineuriomas correspond histologically to grade I tumors (WHO grade I).88 Intraneural Perineurioma Intraneural perineurioma is a tumor characterized by intrafascicular proliferation of neoplastic perineurial cells. The pattern generated by the endoneurial growth is unique and produces a microscopic picture that, however superficially, resembles the morphologic picture of hypertrophic neuropathy of inherited and acquired types. Because of this resemblance, originally intraneural perineurioma was called localized hypertrophic neuropathy, a term that should be reserved for an extremely rare form of Schwann cell proliferation with “true” onion bulb formation, localized to a nerve segment.90 Intraneural perineurioma most frequently affects adolescent and young adults and shows no sex predilection. Among the nerves most commonly involved are the radial, tibial, and sciatic nerve. Cranial nerves are only rarely involved. Muscle weakness, slowly progressing over a period of months to years, is the typical presenting symptom. Sensory symptoms are unusual. On physical examination, muscle atrophy may be apparent. Electromyography commonly demonstrates chronic denervation. On imaging, both computed tomography and MRI, segmental fusiform enlargement of the affected nerve trunk is usually present. Macroscopically, the nerve shows a fusiform enlargement that can extend over several centimeters of its length. Once the epineurium is opened, the single nerve fascicles appear enlarged and pale. Microscopically, the lesion is characterized by marked proliferation of perineurial cells, which, in cross section, appear to surround single nerve fibers and even capillaries and form large, complex whorls (Fig. 115–9). Crossing from one whorl to the other is continuous. The term pseudo–onion bulb formation has been used to indicate this peculiar arrangement of the perineurial cells, which represents a grotesque imitation of the onion bulbs seen in hypertrophic neuropathy. In longitudinal sections, the endoneurium appears quite cellular, with parallel bundles of perineurial cells, a pattern more difficult to interpret and that may be misinterpreted as that of either a cellular neurofibroma or potentially even a low-grade MPNST. In long-standing lesions, both axons and myelin may be severely reduced in number. Occasional tumors may become remarkably sclerotic. We have recently observed a unique example in which crisscrossing between whorls was minimal and limited to very thin perineurial strands. This case was indistinguishable on H&E staining from hypertrophic neuropathy. The pattern of immunoreactivity of intraneural perineurioma is quite characteristic. The perineurial cells forming the leaflets of the whorls show immunoreactivity for EMA, while residual axons and their accompanying Schwann cells, highlighted respectively by

2594

Diseases of the Peripheral Nervous System

diagnosis. Nerves affected should be preserved even though they exhibit only partial function. Resection is curative, but should be avoided in order to maintain nerve function as long as possible.

A

B

C

D

FIGURE 115–9 Intraneural perineurioma: histologic appearance. A and B, The typical appearance of intraneural perineurioma in cross (A) and longitudinal (B) hematoxylin and eosin–stained sections. Large whorls centered around nerve fibers mimicking onion bulbs are present. They are in close contact with each other, and crisscrossing of processes from one whorl to the other is present. When cut longitudinally, however, the characteristic architectural appearance is lost and one may be induced, given the cellularity and the haphazard arrangement of the nuclei, to consider the possibility of a cellular neurofibroma or even a low-grade MPNST. C, The epithelial membrane antigen immunoreactivity in leaflets of pseudo–onion bulbs formed by the neoplastic cells. D, Neurofilament stain highlights the presence of axons at the center of the whorls.

neurofilament and S100 protein stains, remain at the center of the whorls. Ultrastructurally, the neoplastic cells closely resemble normal perineurium, even though their number and spatial distribution are strikingly abnormal. Residual nerve fibers are easily recognized. Intraneural perineurioma is a benign neoplasm that grows quite slowly and causes progressive loss of nerve function. Biopsy of a nerve fascicle alone is sufficient for

Soft Tissue Perineurioma Soft tissue perineurioma is a soft tissue tumor composed entirely of well-differentiated perineurial cells. This tumor was originally described under the name “storiform perineurial fibroma.” Its perineurial nature has been confirmed both by immunohistochemistry and electron microscopy (EM).62,82 The tumor most commonly presents as a soft tissue mass located in subcutaneous tissue, or less frequently in deep soft tissue of the extremities or trunk. Rare examples have been reported in the maxillary sinus and kidney, and even in intraventricular locations.35,36,103 Women are more frequently affected than men (4:1 ratio). The tumor occurs most frequently in middle age. Soft tissue perineuriomas present as round or ovoid, welldemarcated masses, but are not encapsulated. They are generally 1 to 5 cm in diameter, but examples up to 20 cm have been reported. They are not associated with an identifiable nerve trunk. Microscopically, soft tissue perineuriomas are composed of elongate tumor cells with extremely long and thin processes. Cells tend to align in bundles, forming interweaving fascicles or loose whorls with a vague storiform pattern. Cell nuclei are elongate with tapered ends. EMA immunoreactivity in these long, thin cytoplasmic processes may be quite difficult to demonstrate. Gross total removal is curative.

Miscellaneous Benign Peripheral Nerve Tumors Nerve Sheath Myxoma and Neurothekeoma Nerve sheath myxoma and neurothekeoma are two distinct benign multinodular cutaneous tumors that for a long time had been considered, respectively, a myxoid and a cellular variant of the same tumor.32,80 Based on ultrastructural and immunohistochemical studies, nerve sheath myxoma is a spindle cell neoplasm showing Schwann cell differentiation,3–5 and it could be considered a myxoid variant of schwannoma. Neurothekeoma does not stain for S100 protein, and by ultrastructure its cells closely resemble fibroblasts. Nerve sheath myxoma primarily occurs in adults, most frequently in the third to fifth decades. Sites most frequently affected are the skin of the hands, back, arms, head, and neck. Rarely it may involve oral mucosa. By contrast, neurothekeoma affects mainly children and young adults, and most frequently involves the face. Both tumors occur more frequently in women, with a female-to-male ratio of 2:1. They both present as single, firm, and well-demarcated, often multinodular lesions, ranging in size from 0.5 to 3 cm.

Tumors and Tumor-like Conditions of Peripheral Nerve

Nerves are not affected. Microscopically, nerve sheath myxoma is generally hypocellular and composed of spindle and/or stellate cells with long processes immersed in an abundant myxoid matrix. Neurothekeoma, in contrast, is moderately to markedly cellular, composed of both spindle and epithelioid cells with abundant cytoplasm. Multinucleated cells are frequent. Neurothekeoma cells cluster and may form whorls, also interspersed in abundant myxoid stroma. Mitoses are frequently seen, and the mitotic index may be brisk (up to 10 ⫻ 10 HPF), although rarely. A cellular variant of neurothekeoma with little or no myxoid stroma has also been described.9 Although sometimes difficult to distinguish on H&E sections, as mentioned above, nerve sheath myxoma and neurothekeoma can be easily distinguished based on immunohistochemical and EM features. Both tumors are benign and treated by complete surgical excision. Granular Cell Tumor Granular cell tumor is a unique tumor composed of sheets of cells with abundant eosinophilic, granular cytoplasm. The cytoplasm, which is periodic acid–Schiff positive, ultrastructurally contains a large numbers of secondary lysosomes. Neoplastic cells display a characteristic immunoreactivity for S100 protein and CD68.90 These features and its association with craniospinal nerves support the neurogenic nature of granular cell tumor. Synonyms for this tumor include the terms granular cell schwannoma and granular cell neurogenic tumor. Granular cell tumor occurs in patients of any age, although it is most frequent between the fourth and sixth decades. Women are more frequently affected than men. The tumor is also more common in blacks.98 Skin and subcutaneous tissue of the head and neck region are the most common sites.1,77 The single most frequent site of occurrence is the tongue, accounting for approximately one fourth of all granular cell tumors.58,72 The larynx, gastrointestinal tract, and breast are involved less frequently.90 Pseudoepitheliomatous hyperplasia of overlying squamous epithelium, at times so marked as to closely mimic squamous cell carcinoma, is a common finding and well-known diagnostic pitfall of granular cell tumor. Granular cell tumors generally present as solitary nodules, relatively circumscribed, rarely measuring more than 3 cm in greatest diameter. Up to 10% of tumors can be multiple. The vast majority of granular cell tumors are benign and cured by local excision. However, it is important to obtain negative surgical margins because, after incomplete resection, recurrence is seen in up to 50% of cases. Granular cell tumors show an aggressive clinical behavior, with local recurrence and metastases to regional lymph nodes and distant sites, in less than 10% of cases. The lung is the most common site of distant metastasis. Although a

2595

clinically malignant behavior cannot be predicted based solely on morphologic grounds, high cellularity, cellular pleomorphism, prominent nucleoli, readily identifiable mitoses, and foci of necrosis are generally present in malignant examples.102 Ganglioneuroma Ganglioneuroma is a benign neurogenic tumor composed of mature ganglion cells admixed with abundant nerve fibers, predominantly unmyelinated, and accompanying Schwann cells.90 The tumor occurs most frequently in children and young adults, with a distinct female predilection. The association with NF1 is well established but uncommon.54 Although a rare event, some ganglioneuromas have been reported to secrete hormones or neurotransmitters such as vasoactive intestinal polypeptide or catecholamines, causing endocrine or functional symptoms.90 They arise most often at sites where sympathetic ganglia are found, including the posterior mediastinum, paraspinous region, retroperitoneum, and adrenal medulla. They generally present as well-circumscribed, solid masses with a thin pseudocapsule. They rarely show marked calcifications. The ganglion cell component is mature and often associated with accompanying satellite cells, mimicking normal ganglion cells. Multinucleated ganglion cells are usually seen. Cells possess long axonal processes, often out of proportion to the number of ganglion cells present. Schwann cells are numerous and strictly associated with axons. Myelination, however, is generally limited to entrapped normal nerve fibers. The number and distribution of ganglion cells may be highly variable. Extensive tumor sampling is necessary to fully document the histologic features of the tumor as well as to exclude the presence of an immature, neuroblastic component as seen in immature ganglioneuroma and ganglioneuroblastoma. Differences in patient age and tumor distribution support the view that ganglioneuromas and neuroblastic tumors are generally separate, distinct entities. However, maturation of ganglioneuroblastoma to fully differentiated ganglioneuroma is known to occur either spontaneously or after adjuvant therapy, at either primary or metastatic sites. This suggests that at least some ganglioneuromas might represent fully matured and differentiated neuroblastic tumors. The occurrence of ganglioneuroma in association with pheochromocytoma, forming a “composite tumor” with two distinct components, has been reported in the adrenal medulla.75 A well-differentiated ganglion cell component may also occur in paragangliomas of the cauda equina and duodenum.89,91,95 A diffuse ganglioneuromatous proliferation, involving all bowel layers, may occur in the gastrointestinal tract in multiple endocrine neoplasia type IIB. Ganglioneuromas are benign tumors and cured by surgical resection. Only on very rare occasions may a MPNST arise in a ganglioneuroma.55,90

2596

Diseases of the Peripheral Nervous System

Miscellaneous Non-neurogenic Tumors A variety of benign non-neurogenic tumors may occasionally arise in peripheral nerve roots and trunks, including meningioma,16,39 paraganglioma,94,95,107 hemangioma,26,84,104 hemangioblastoma,34 and lipoma.85,100 Quite rarely, adrenal cortical adenomas have also been reported to occur in nerve roots, presumably arising in ectopic adrenal tissue.52,74 Amyloidoma of peripheral nerve is a rarity and, as with central nervous system amyloidomas, is not associated with systemic amyloidosis.31,59,76

MALIGNANT TUMORS OF PERIPHERAL NERVE Malignant Peripheral Nerve Sheath Tumor and Its Variants The WHO defines as a MPNST any malignant tumor arising from a peripheral nerve or showing nerve sheath differentiation.113 Mesenchymal tumors arising from the epineurium or the peripheral nerve vasculature, composed of fibroblasts, endothelial cells, and pericytes, are excluded. Although many MPNSTs arise in identifiable nerve trunks, this is not a rule. Approximately 50% of MPNSTs occur in patients with NF1. The vast majority of MPNSTs, even when poorly differentiated, show some features, either at the light or the ultrastructural level, of schwannian differentiation. However, MPNSTs showing evidence of differentiation along other neurogenic lines, such as perineurial, have been demonstrated.43 Because of this, the old term malignant schwannoma used as a synonym for MPNST appears today to be incorrect and should not be used. As an additional reason to discourage the use of this term, Woodruff et al.112 suggested that it may invite confusion with cellular schwannoma and promote the idea that a MPNST may actually arise in transition/progression from this or even conventional schwannomas. In a similar way, the terms neurogenic sarcoma and neurofibrosarcoma, historically attributed to the tumor because of the spindle cell morphology of the neoplastic cells and the collagen production but implying a mesenchymal origin/differentiation for a tumor that is truly neuroectodermal, should not be used. The term malignant peripheral nerve sheath tumor, encompassing the wide spectrum of malignant tumors of nerve sheath, is the recommended designation. According to the WHO, histologically the vast majority of MPNSTs correspond to grade III or IV lesions. MPNST is an uncommon tumor and accounts for approximately 5% of malignant soft tissue tumors.63 Although a significant proportion of MPNSTs arise de novo in subjects without a genetic predisposition, in approximately 50% to 75% of cases one may find evidence of a preexisting neurofibroma, often plexiform, and/or the

patient is affected by NF1.22,46,56,96 MPNST may only very rarely arise in transition from other neuroectodermal tumors, such as schwannoma,116 ganglioneuroma/ganglioneuroblastoma,83 or pheochromocytoma.71,73,86 The clinical and pathologic features of these rare tumors are often unique compared to conventional MPNST. MPNST occurs most often in adults, with a peak incidence between the third and sixth decades of life. Patients affected by NF1 who develop MPNST are generally approximately 10 years younger.22,46 Women are slightly more frequently affected than men. MPNSTs arising in a field of prior radiation account for approximately 10% of all tumors.20,30,56,117 MPNSTs generally involve medium- and large-size nerves, with the brachial and lumbosacral plexuses as well as the buttocks and thighs being the most commonly affected sites. The sciatic nerve is the single most commonly affected nerve, while cranial or visceral nerves are only rarely involved. MPNST generally presents as a firm, spherical or fusiform, pseudoencapsulated mass. Frequently an identifiable nerve trunk is present (Figs. 115–10 and 115–11). Most

FIGURE 115–10 Malignant peripheral nerve sheath tumor: gross appearance. The tumor appears as a fusiform enlargement of the sciatic nerve. On cut surface, it shows a fleshy appearance with focal areas of hemorrhage.

Tumors and Tumor-like Conditions of Peripheral Nerve

2597

FIGURE 115–11 Malignant peripheral nerve sheath tumor: gross appearance. A, This tumor has the appearance of a large globular mass along the sciatic nerve. Patchy areas of necrosis are present. B, This tumor shows extensive confluent necrosis.

MPNSTs are larger than 5 cm. On cut surface, typically they have a cream or gray color with frequent areas of necrosis and/or hemorrhage (see Fig. 115–10). Histologically, the majority of MPNSTs are cellular tumors composed of tightly packed hyperchromatic spindle cells (Fig. 115–12). Nuclei are large, usually greater than three times the size of those of neurofibroma, with tapered ends. Commonly, cells are disposed in fascicles with interweaving, sinuous or herringbone arrangements, mimicking fibrosarcoma. Diffuse, patternless growth with alternating low- and high-cellularity areas, concentrated perivascular cellularity, and even a hemangiopericytoma-like appearance of the vasculature can be seen. Although the majority of MPNSTs show brisk mitotic activity and necrosis and are histologically high grade, low-grade tumors occur (Fig. 115–13). S100 protein immunostaining is generally positive in MPNST, but the number of positive cells is quite variable and often limited to scattered tumor cells.90 Leu-7 and/or myelin basic protein, markers that have been reported to be positive in MPNST, are generally not used as routine diagnostic markers. Immunostaining with p53 protein is often found in MPNST and has been associated with an unfavorable prognosis.38,53 Ultrastructurally, MPNSTs, especially of high grade, show little or no differentiation, limiting the diagnostic utility of EM in obviously high-grade tumors. EM may be useful in lower grade tumors, which often show more

A

obvious differentiation. Schwannian differentiation is manifested by the presence of cytoplasmic intermediate filaments, cell processes, rudimentary junctions, and often discontinuous basement membrane.41 In a minority of cases, tumors show distinct fibroblastic or perineurial features.43 The recurrence rate of MPNST following surgical excision is high, with rates varying from 40% for tumors located in the extremities to 68% for those located paraspinally.22,54 The rate of metastasis is also high (⬎60%) and is not affected by primary tumor location. The most common sites of metastasis include lung, bone, pleura, soft tissue, liver, and brain. Prognosis is poor, with overall 5- and 10-year survival rates varying from 34% to 52% and from 23% to 34%, respectively.54 Epithelioid MPNST Epithelioid MPNST is the term used to describe tumors that have a markedly epithelial-like cytology. The cells are polygonal, with abundant cytoplasm and large nuclei with prominent nucleoli. These tumors represent less than 5% of all MPNSTs. They occur at any age, show no sex predilection, and are not associated with NF1.7,18,60,64 Typically the tumors are located in the limbs. Tumors located superficially generally are smaller and have a better prognosis than deep-seated tumors. Deep-seated tumors also appear to undergo distant

B

FIGURE 115–12 Malignant peripheral nerve sheath tumor: histologic appearance. Spindle nuclei, approximately three times the size of those seen in neurofibroma, are disposed in fascicles with an interwoven or herringbone arrangement mimicking fibrosarcoma. Mitotic figures are frequent.

2598

A

Diseases of the Peripheral Nervous System

B

metastasis more frequently than superficial lesions.60 As are conventional MPNSTs, they are associated with an identifiable nerve trunk in approximately 50% of cases. Epithelioid MPNSTs show definite evidence of schwannian differentiation immunohistochemically and ultrastructurally. S100 protein staining is strongly and diffusely positive in the majority of epithelioid MPNSTs, while melanoma-associated markers such as HMB-45 are consistently negative. Cytokeratin and/or EMA positivity have been occasionally reported. Ultrastructurally, despite the lack of processes, cells show schwannian features, especially often a wellformed basal lamina.2,18,64,99 Frequently areas with a spindle cell component as seen in conventional MPNST are present

A

B

FIGURE 115–13 Malignant peripheral nerve sheath tumor of low grade arising in a neurofibroma. The characteristic appearance of a classic neurofibroma is observed in the upper portion of A. The cellularity and degree of atypia observed in the lower portion of A and in B, together with the presence of mitotic activity, are features of malignant transformation.

in epithelioid MPNST. Epithelioid MPNST appears to have a better prognosis than conventional MPNST.60 MPNST with Divergent Differentiation Divergent differentiation along mesenchymal and/or epithelial lines can be seen in approximately 15% of MPNSTs.21 Features of heterologous mesenchymal differentiation include obviously malignant components such as rhabdomyosarcoma, chondrosarcoma, osteosarcoma, and less commonly angiosarcoma. The term malignant triton tumor refers to tumor with myoid differentiation and represents the most common example of divergent differentiation (Fig. 115–14).108 The term

FIGURE 115–14 Malignant peripheral nerve sheath tumor with rhabdomyoblastic differentiation (malignant triton tumor). In this case, rhabdomyoblasts with abundant eosinophilic cytoplasm (B) are arranged in nests intermixed with the classic spindle cell MPNST areas.

Tumors and Tumor-like Conditions of Peripheral Nerve

A

2599

B

FIGURE 115–15 Malignant peripheral nerve sheath tumor with divergent (glandular) differentiation. A, Glandular differentiation in MPNST is generally represented by the presence of well-differentiated glandular structures with mucinous differentiation. B, Goblet cells are mucicarmine positive (lower left). Cytokeratin and polyclonal carcinoembryonic antigen are positive in glandular structures (upper left and lower right). Neuroendocrine differentiation with the presence of chromogranin-positive basal cells is frequent (upper right).

glandular MPNST is used to indicate tumors containing epithelial elements. These generally have a benign histologic appearance and include mucin-producing intestinaltype glands, squamous nests, and neuroendocrine cells (Fig. 115–15).110 MPNSTs may simultaneously show “multidirectional” differentiation with epithelial and mesenchymal elements, generally represented by glandular and rhabdomyosarcomatous components. In comparison to conventional MPNSTs, MPNSTs with divergent differentiation are more frequently associated with NF1, the frequency varying from approximately 60% in malignant triton tumors to 75% in glandular MPNSTs.90 The prognosis of these tumors is poorer than that of tumors lacking divergent differentiation. Survival rates are, respectively, 12% and 21% at 5 years. MPNST in Transformation from Tumors Other Than Neurofibroma MPNSTs may rarely arise in transformation from schwannoma. Transformation of ganglioneuroma/ganglioneuroblastoma or pheochromocytoma into MPNST is

exceedingly rare. There are well-documented examples of each of these reported in the literature.71,73,83,86,116 Fewer than 10 genuine and verifiable cases of MPNST arising in transformation from schwannoma have been reported.90 All cases described were in adults, with an age range from 31 to 75 years. No association was noted with NF1. A history of a preceding mass, present often for many years, was common. All tumors showed morphologic evidence of an underlying classic schwannoma, with Antoni A and Antoni B areas, adjacent to malignant areas. In the majority of the cases, the MPNST component was purely epithelioid by morphology. In two cases, the malignant component resembled a poorly differentiated small round cell “primitive neuroectodermal tumor.” We are aware of two additional cases of MPNST, one showing rhabdomyosarcomatous differentiation57 and one showing chondro- and osteosarcomatous differentiation (personal observation). The outcome of the reported cases was very poor, with locally aggressive behavior as well as distant metastases in 50% of cases leading to patient demise.

2600

Diseases of the Peripheral Nervous System

A

B

C

D

FIGURE 115–16 Metastatic squamous cell carcinoma involving trigeminal nerve and ganglion.

Other Malignant Peripheral Nerve Tumors Primitive neuroectodermal tumor arising in peripheral nerve, including the so-called peripheral neuroepithelioma, is very rare in comparison to examples occurring in soft tissue.90 A variety of nerves have been involved, including sciatic, median, ulnar, radial, intercostal, and lateral popliteal nerves as well as sacral roots. Malignant granular cell tumor has been mentioned above. A variety of malignant tumors may secondarily involve peripheral nerve trunks either by direct extension from lesions arising in nearby tissues or by metastasis. Carcinoma, sarcoma, melanoma, and leukemia/lymphoma may all involve peripheral nerve, with the pattern of involvement depending largely on tumor type.90 Carcinoma and melanoma typically spread along the perineurium and the endoneurium, occasionally for long distances. Metastatic involvement of ganglia is also a well-known occurrence. In the head and neck region, squamous cell carcinoma and adenoid cystic carcinoma can spread proximally following cranial nerve branches, especially the fifth nerve, and gain access to the central nervous system (Fig. 115–16).8 Cranial neuropathies resulting from such spreading may be difficult to diagnose. Distant metastasis can also manifest with focal cranial nerve involvement, as in the “numb chin” in breast cancer.45 In sarcomas, nerve involvement is generally limited to the epineurium, although we have rarely observed spreading along a nerve branch, as in a recurrent angiosarcoma of the mandible. Lymphoma/leukemia typically

involves the subperineurial space, endoneurial septa, and endoneurium.

TUMOR-LIKE CONDITIONS AFFECTING PERIPHERAL NERVE A variety of non-neoplastic lesions may involve peripheral nerves. Table 115–2 provides a summary of them as they were presented in the recently published third series AFIP fascicle Tumors of the Peripheral Nervous System.90 A brief discussion of the most frequently encountered lesions is provided here, and the reader is referred to the AFIP fascicle for an in-depth discussion.

Traumatic Neuroma Traumatic neuroma is a reactive, mass-like lesion that develops at sites of partial or complete nerve injury/transection. Traumatic neuroma develops as a consequence of the attempt of injured nerve fibers to regenerate and reconnect to the distal stump and ultimately to their target. The greater the loss of nerve continuity and damage to the nerve sheaths, the more disorderly is the regeneration, resulting at times in an entangled skein of microfascicles, also called an amputation neuroma.23,66 Following nerve transection, nerve fibers of the distal stump undergo wallerian degeneration, resulting in complete degeneration

Tumors and Tumor-like Conditions of Peripheral Nerve

and removal of axons and myelin debris.33 Axons in the proximal nerve stump undergo sprouting and start elongating in an attempt to regenerate and ultimately reconnect with their original and/or a functionally meaningful target. Axonal sprouting takes place at the level of the nodes of Ranvier immediately proximal to the site of transection, and each fiber may give rise to as many as 25 sprouts. Nerve fibers successful in making functional connections with the original and/or a new target mature and remain functional. The others undergo progressive atrophy and ultimately may disappear. As discussed in detail in Chapter 58, a variety of factors influence regeneration. The type of injury is critically important and strongly influences the probability that a regenerating axon might enter a Büngner’s band and be guided to the original or to a meaningful target, guaranteeing successful regeneration. Minimal mismatch of nerve fibers and Büngner’s bands occurs in experimental crush injury, which largely results in a transection limited to axons.101 When the continuity of the nerve is partially or completely lost, with disruption of the epineurium, perineurium, and vascular structures, bundles of regenerating nerve fibers, accompanied by proliferation of Schwann and perineurial cells, outgrow in surrounding soft tissue in a haphazard manner. The result is the traumatic neuroma, a tangle of disordered microfascicles containing a variable number of fibers and surrounded by perineurium.

2601

Traumatic neuromas generally develop following sporadic and at times insignificant injuries. They may develop in the postsurgical setting at either somatic or visceral sites of injury. The neuroma often presents as a firm, tender, and frequently painful nodule. To the surgeon it appears as a localized globular expansion at the end or on the side of an injured nerve stump (Fig. 115–17A). Histologically, intertwined microfascicles of different size, containing disorderly regenerating axons, are present in a variably collagenized background (Fig. 115–17B to D). Both fiber size and degree of myelination are markedly decreased compared to the parent nerve. Hemosiderin deposits resulting from previous hemorrhage, granulation tissue deposition, and fibrosis may be present. Localized Interdigital Neuritis Localized interdigital neuritis (also known as Morton’s neuroma and plantar neuroma) is related to traumatic neuroma and results from chronic/repeated nerve injury.61,81,92 The typical site of occurrence is the second or third interdigital space. The lesion affects primarily adult females and manifests clinically with paroxysmal episodes of pain. Although an obvious enlargement of the plantar digital nerve is often present clinically, changes seen histologically are primarily degenerative rather than proliferative. All the nerve compartments are affected, with presence of marked nerve fiber loss and fibrosis.

A

B

C

D

FIGURE 115–17 Traumatic neuroma: gross (A) and microscopic (B–D) appearance. A, A bulbous expansion of the proximal nerve stump is present. B, Numerous haphazardly arranged microfascicles replace the normal structure of the nerve fascicle. On S100 protein (C) and neurofilament (D) immunostaining, a large number of nerve fibers are present, representing an aborted attempt at nerve regeneration.

2602

Diseases of the Peripheral Nervous System

Palisaded Encapsulated Neuroma Palisaded encapsulated neuroma is a lesion that occurs primarily on the face.6,19 It presents as a small, nodular, circumscribed cutaneous lesion. Histologically it is composed of axons and Schwann cells, surrounded by a delicate perineurium. It is also thought to be related to traumatic neuroma.

Lipofibromatous Hamartoma and Neuromuscular Choristoma Lipofibromatous hamartoma is a rare lesion characterized by extensive overgrowth of primarily adipose and to a variable extent fibrous tissue in the nerve epineurium. This proliferation causes progressive marked enlargement of the nerve trunk and may be associated with slow and progressive loss of function (Fig. 115–18). The single most frequently involved nerve is the median nerve in the hand. Clinically, marked enlargement of the involved fingers is evident, resulting in apparent macrodactyly. The term true macrodactyly is reserved to cases associated with enlargement of underlying bone structures.100 Peculiar to lipofibromatous hamartoma are the endoneurial changes often seen in associated nerve fascicles, with prominent perineurial proliferation mimicking changes seen in intraneural perineurioma

FIGURE 115–18 Lipofibromatous hamartoma involving median nerve: gross appearance. The nerve epineurium is largely replaced by exuberant adipose tissue, conferring on the nerve a distorted, ropy, fatty appearance.

(Fig. 115–19). These changes, most likely reactive to this slowly progressive epineurial lesion, are often segmental along the course of nerve fascicles (personal observation). Neuromuscular choristoma is characterized by the disordered proliferation of mature skeletal muscle fibers in intimate association with peripheral nerve fibers.69,90

A

B

C

D

FIGURE 115–19 Lipofibromatous hamartoma: histologic appearance of nerve fascicles. A, Nerve fascicles are spread apart by exuberant adipose and fibrovascular tissue. Changes seen inside the nerve fascicles often mimic those seen in intraneural perineurioma. Pseudo–onion bulbs forming whorl-like structures around axons are perineurial in nature (B–D). These changes, which often appear to be focal and segmental in nerve fascicles (personal observation), are most likely reactive in nature and may result from chronic injury to the perineurium.

Tumors and Tumor-like Conditions of Peripheral Nerve

REFERENCES 1. Alessi, D. M., and Zimmerman, M. C.: Granular cell tumors of the head and neck. Laryngoscope 98(8 Pt. 1):810, 1988. 2. Alvira, M. M., Mandybur, T. K., and Menefee, M. G.: Light microscopic and ultrastructural observations of a metastasizing malignant epithelioid schwannoma. Cancer 38:1977, 1976. 3. Angervall, L., Kindblom, L. G., and Haglid, K.: Dermal nerve sheath myxoma: a light and electron microscopic, histochemical and immunohistochemical study. Cancer 53:1752, 1984. 4. Argenyi, Z. B., Kutzner, H., and Seaba, M. M.: Ultrastructural spectrum of cutaneous nerve sheath myxoma/cellular neurothekeoma. J. Cutan. Pathol. 22:137, 1995. 5. Argenyi, Z. B., LeBoit, P. E., Santa Cruz, D., et al.: Nerve sheath myxoma (neurothekeoma) of the skin: light microscopic and immunohistochemical reappraisal of the cellular variant. J. Cutan. Pathol. 20:294, 1993. 6. Argenyi, Z. B., Santa Cruz, D., and Bromley, C.: Comparative light-microscopic and immunohistochemical study of traumatic and palisaded encapsulated neuromas of the skin. Am. J. Dermatopathol. 14:504, 1992. 7. Axiotis, C. A., Merino, M. J., Tsokos, M., et al.: Epithelioid malignant peripheral nerve sheath tumor with squamous differentiation: a light microscopic, ultrastructural, and immunohistochemical study. Surg. Pathol. 3:301, 1990. 8. Ballantine, A. J., McCarten, A. B., and Ibanez, M. L.: The extension of cancer of the head and neck through peripheral nerves. Am. J. Surg. 106:651, 1963. 9. Barnhill, R. L., and Mihm, M. C. J.: Cellular neurothekeoma: a distinctive variant of neurothekeoma mimicking nevomelanocytic tumors. Am. J. Surg. Pathol. 14:113, 1990. 10. Bigner, S. H., Burger, P. C., Wong, A. J., et al.: Gene amplification in malignant human gliomas: clinical and histopathological aspects. J. Neuropathol. Exp. Neurol. 50:2347, 1988. 11. Carney, J. A.: Psammomatous melanotic schwannoma: a distinctive, heritable tumor with special associations, including cardiac myxoma and the Cushing syndrome. Am. J. Surg. Pathol. 14:206, 1990. 12. Carney, J. A., Gordon, H., Carpenter, P. C., et al.: The complex of myxomas, spotty pigmentation, and endocrine overactivity. Medicine (Baltimore) 64:270, 1985. 13. Carney, J. A., Hruska, L. S., Beauchamp, G. D., et al.: Dominant inheritance of the complex of myxomas, spotty pigmentation, and endocrine overactivity. Mayo Clin. Proc. 61:165, 1986. 14. Casadei, G. P., Komori, T., Scheithauer, B. W., et al.: Intracranial parenchymal schwannoma: a clinicopathological and neuroimaging study of nine cases. J. Neurosurg. 79:217, 1993. 15. Casadei, G. P., Scheithauer, B. W., Hirose, T., et al.: Cellular schwannoma: a clinicopathologic, DNA flow cytometric, and proliferation marker study of 70 patients. Cancer 75:1109, 1995. 16. Coons, S. W., and Johnson, P. C.: Brachial plexus meningioma: report of a case with immunohistochemical and ultrastructural examination. Acta Neuropathol. (Berl.) 77:445, 1989.

2603

17. Deruaz, J. P., Janzer, R. C., and Costa, J.: Cellular schwannoma of the intracranial and intraspinal compartment: morphological and immunological characteristics compared with classical benign schwannomas. J. Neuropathol. Exp. Neurol. 52:114, 1993. 18. DiCarlo, E. F., Woodruff, J. M., Bansal, M., et al.: The purely epithelioid malignant peripheral nerve sheath tumor. Am. J. Surg. Pathol. 10:478, 1986. 19. Dover, J. S., From, L., and Lewis, A.: Palisaded encapsulated neuromas: a clinicopathologic study. Arch. Dermatol. 125:386, 1989. 20. Ducatman, B. S., and Scheithauer, B. W.: Postirradiation neurofibrosarcoma. Cancer 51:1028, 1983. 21. Ducatman, B. S., and Scheithauer, B. W.: Malignant peripheral nerve sheath tumors with divergent differentiation. Cancer 54:1049, 1984. 22. Ducatman, B. S., Scheithauer, B. W., Piepgras, D. G., et al.: Malignant peripheral nerve sheath tumors: a clinicopathologic study of 120 cases. Cancer 57:2006, 1986. 23. Dyck, P. J., Giannini, C., and Lais, A.: Pathologic Alteration of Nerves, 3rd ed. Philadelphia, W. B. Saunders, 1993. 24. Emory, T. S., Scheithauer, B. W., Hirose, T., et al.: Intraneural perineurioma: a clonal neoplasm associated with abnormalities of chromosome 22. Am. J. Clin. Pathol. 103:696, 1995. 25. Emory, T. S., and Unni, K. K.: Intraosseous neurilemmoma: a clinicopathologic study of twenty-six cases. Am. J. Clin. Pathol. 100:328, 1993. 26. Ergin, M. T., Druckmiller, W. H., and Cohen, P.: Intrinsic hemangiomas of the peripheral nerves: report of a case and review of the literature. Conn. Med. 62:209, 1998. 27. Erlandson, R. A.: Diagnostic Transmission Electron Microscopy of Tumors. New York, Raven Press, 1994. 28. Fletcher, C. D., and Davies, S. E.: Benign plexiform (multinodular) schwannoma: a rare tumour unassociated with neurofibromatosis. Histopathology 10:971, 1986. 29. Fletcher, C. D., Davies, S. E., and McKee, P. H.: Cellular schwannoma: a distinct pseudosarcomatous entity. Histopathology 11:21, 1987. 30. Foley, K. M., Woodruff, J. M., Ellis, F. T., et al.: Radiationinduced malignant and atypical peripheral nerve sheath tumors. Ann. Neurol. 7:311, 1980. 31. Gabet, J. Y., Durand, D. V., Bady, B., et al.: Pseudo-tumor amyloide du neerf sciatique. Rev. Neurol. 145:872, 1989. 32. Gallager, R. L., and Helwig, E. B.: Neurothekeoma—a benign cutaneous tumor of neural origin. Am. J. Clin. Pathol. 74:759, 1980. 33. Giannini, C., and Dyck, P. J.: The fate of Schwann cell basement membranes in permanently transected nerves. J. Neuropathol. Exp. Neurol. 49:550, 1990. 34. Giannini, C., Scheithauer, B. W., Hellbusch, L. C., et al.: Peripheral nerve hemangioblastoma. Mod. Pathol. 11:999, 1998. 35. Giannini, C., Scheithauer, B. W., Jenkins, R. B., et al.: Softtissue perineurioma: evidence for an abnormality of chromosome 22, criteria for diagnosis, and review of the literature. Am. J. Surg. Pathol 21:164, 1997. 36. Giannini, C., Scheithauer, B. W., Steinberg, J., et al.: Intraventricular perineurioma: case report. Neurosurgery 43:1478; discussion 1481, 1998.

2604

Diseases of the Peripheral Nervous System

37. Halliday, A. L., Sobel, R. A., and Martuza, R. L.: Benign spinal nerve sheath tumors: their occurrence sporadically and in neurofibromatosis types 1 and 2. J. Neurosurg. 74:248, 1991. 38. Halling, K. C., Scheithauer, B. W., Halling, A. C., et al.: p53 expression in neurofibroma and malignant peripheral nerve sheath tumor: an immunohistochemical study of sporadic and NF1-associated tumors. Am. J. Clin. Pathol. 106:282, 1996. 39. Harkin, J. C., and Reed, R. J.: Tumors of the Peripheral Nervous System. Washington, DC, Armed Forces Institute of Pathology, 1969. 40. Herregodts, P., Vloeberghs, M., Schmedding, E., et al.: Solitary dorsal intramedullary schwannoma: case report. J. Neurosurg. 74:816, 1991. 41. Hirose, T., Hasegawa, T., Kudo, E., et al.: Malignant peripheral nerve sheath tumors: an immunohistochemical study in relation to ultrastructural features. Hum. Pathol. 23:865, 1992. 42. Hirose, T., Sano, T., and Hizawa, K.: Ultrastructural localization of S-100 protein in neurofibroma. Acta Neuropathol. (Berl.) 69:103, 1986. 43. Hirose, T., Scheithauer, B. W., and Sano, T.: Malignant perineurioma: a study of eight cases [abstract]. Mod. Pathol. 10, 1997. 44. Hodson, J. J.: An intra-osseous tumor combination of biological importance—invasion of a melanotic schwannoma by an adamantinoma. J. Pathol. Bacteriol. 82:257, 1961. 45. Horton, J., Means, E. D., Cunningham, T. J., et al.: The numb chin in breast cancer. J. Neurol. Neurosurg. Psychiatry 36:211, 1973. 46. Hruban, R. H., Shiu, M. H., Senie, R. T., et al.: Malignant peripheral nerve sheath tumors of the buttock and lower extremity: a study of 43 cases. Cancer 66:1253, 1990. 47. Iwashita, T., and Enjoji, M.: Plexiform neurilemmoma: a clinicopathological and immunohistochemical analysis of 23 tumours from 20 patients. Virchows Arch. A Pathol. Anat. Histol. 411:305, 1987. 48. Izumi, A. K., Rosato, F. E., and Wood, M. G.: Von Recklinghausen’s disease associated with multiple neurilemomas. Arch. Dermatol. 104:172, 1971. 49. Jacoby, L. B., MacCollin, M., Louis, D. N., et al.: Exon scanning for mutation of the NF2 gene in schwannomas. Hum. Mol. Genet. 3:413, 1994. 50. Johnson, M. D., Glick, A. D., and Davis, B. W.: Immunohistochemical evaluation of Leu-7, myelin basicprotein, S100-protein, glial-fibrillary acidic-protein, and LN3 immunoreactivity in nerve sheath tumors and sarcomas. Arch. Pathol. Lab. Med. 112:155, 1988. 51. Kawahara, E., Oda, Y., Ooi, A., et al.: Expression of glial fibrillary acidic protein (GFAP) in peripheral nerve sheath tumors: a comparative study of immunoreactivity of GFAP, vimentin, S-100 protein, and neurofilament in 38 schwannomas and 18 neurofibromas. Am. J. Surg. Pathol. 12:115, 1988. 52. Kepes, J. J., O’Boynick, P., Jones, S., et al.: Adrenal cortical adenoma in the spinal canal of an 8-year-old girl. Am. J. Surg. Pathol. 14:481, 1990. 53. Kindblom, L. G., Ahlden, M., Meis-Kindblom, J. M., et al.: Immunohistochemical and molecular analysis of p53, MDM2, proliferating cell nuclear antigen and Ki67 in

54.

55.

56.

57.

58.

59.

60.

61.

62.

63. 64.

65.

66. 67.

68.

69.

70.

71.

benign and malignant peripheral nerve sheath tumours. Virchows Arch. 427:19, 1995. Kleihues, P., and Cavanee, W. K.: Pathology and Genetics: Tumors of the Nervous System. Lyon, International Agency for Research on Cancer, 2000. Kodet, R., Kasthuri, N., Marsden, H. B., et al.: Gangliorhabdomyosarcoma: a histopathological and immunohistochemical study of three cases. Histopathology 10:181, 1986. Kourea, H. P., Bilsky, M. H., Leung, D. H., et al.: Subdiaphragmatic and intrathoracic paraspinal malignant peripheral nerve sheath tumors: a clinicopathologic study of 25 patients and 26 tumors. Cancer 82:2191, 1998. Kurtkaya-Yapicier, O., Scheithauer, B. W., Woodruff, J. M., et al.: Schwannoma with rhabdomyoblastic differentiation: a unique variant of malignant triton tumor. Am. J. Surg. Pathol. 27:848, 2003. Lack, E. E., Worsham, G. F., Callihan, M. D., et al.: Granular cell tumor: a clinicopathologic study of 110 patients. J. Surg. Oncol. 13:301, 1980. Laeng, R. H., Altermatt, H. J., Scheithauer, B. W., et al.: Amyloidomas of the nervous system: a monoclonal B-cell disorder with monotypic amyloid light chain lambda amyloid production. Cancer 82:362, 1998. Laskin, W. B., Weiss, S. W., and Bratthauer, G. L.: Epithelioid variant of malignant peripheral nerve sheath tumor (malignant epithelioid schwannoma). Am. J. Surg. Pathol. 15:1136, 1991. Lassmann, G., Lassmann, H., and Stockinger, L.: Morton’s metatarsalgia: light and electron microscopic observations and their relation to entrapment neuropathies. Virchows Arch. A Pathol. Anat. Histol. 370:307, 1976. Lazarus, S. S., and Trombetta, L. D.: Ultrastructural identification of a benign perineurial cell tumor. Cancer 41:1823, 1978. Lewis, J. J., and Brennan, M. F.: Soft tissue sarcomas. Curr. Probl. Surg. 33:817, 1996. Lodding, P., Kindblom, L. G., and Angervall, L.: Epithelioid malignant schwannoma: a study of 14 cases. Virchows Arch. A Pathol. Anat. Histopathol. 409:433, 1986. Louis, D. N., Ramesh, V., and Gusella, J. F.: Neuropathology and molecular genetics of neurofibromatosis 2 and related tumors. Brain Pathol 5:163, 1995. Lundborg, G.: Nerve regeneration and repair: a review. Acta Orthop. Scand. 58:145, 1987. MacCollin, M., Ramesh, V., Jacoby, L. B., et al.: Mutational analysis of patients with neurofibromatosis 2. Am. J. Hum. Genet. 55:314, 1994. MacCollin, M., Woodfin, W., Kronn, D., et al.: Schwannomatosis: a clinical and pathologic study. Neurology 46:1072, 1996. Maher, C. O., Spinner, R. J., Giannini, C., et al.: Neuromuscular choristoma of the sciatic nerve: case report. J. Neurosurg. 96:1123, 2002. McComb, R. D., and Bigner, D. D.: Immunolocalization of laminin in neoplasms of the central and peripheral nervous systems. J. Neuropathol. Exp. Neurol. 44:242, 1985. Miettinen, M., and Saari, A.: Pheochromocytoma combined with malignant schwannoma: unusual neoplasm of the adrenal medulla. Ultrastruct. Pathol. 12:513, 1988.

Tumors and Tumor-like Conditions of Peripheral Nerve 72. Miller, A. S., Leifer, C., Chen, S. Y., et al.: Oral granular-cell tumors: report of twenty-five cases with electron microscopy. Oral Surg. Oral Med. Oral Pathol. 44:227, 1977. 73. Min, K. W., Clemens, A., Bell, J., et al.: Malignant peripheral nerve sheath tumor and pheochromocytoma: a composite tumor of the adrenal. Arch. Pathol. Lab. Med. 112:266, 1988. 74. Mitchell, A., Scheithauer, B. W., Sasano, H., et al.: Symptomatic intradural adrenal adenoma of the spinal nerve root: report of two cases. Neurosurgery 32:658; discussion 661, 1993. 75. Moore, P. J., and Biggs, P. J.: Compound adrenal medullary tumor. South. Med. J. 88:475, 1995. 76. O’Brien, T. J., McKelvie, P. A., and Vrodos, N.: Bilateral trigeminal amyloidoma: an unusual case of trigeminal neuropathy with a review of the literature. Case report. J. Neurosurg. 81:780, 1994. 77. Peterson, L. J.: Granular-cell tumor: review of the literature and report of a case. Oral Surg. Oral Med. Oral Pathol. 37:728, 1974. 78. Pimentel, J., Tavora, L., Cristina, M. L., et al.: Intraventricular schwannoma. Childs Nerv. Syst. 4:373, 1988. 79. Polkey, C. E.: Intraosseous neurilemmoma of the cervical spine causing paraparesis and treated by resection and grafting. J. Neurol. Neurosurg. Psychiatry 38:776, 1975. 80. Pulitzer, D. R., and Reed, R. J.: Nerve-sheath myxoma (perineurial myxoma). Am. J. Dermatopathol. 7:409, 1985. 81. Reed, R. J., and Bliss, B. O.: Morton’s neuroma: regressive and productive intermetatarsal elastofibrositis. Arch. Pathol. 95:123, 1973. 82. Reed, R. J., and Harkin, J. C.: Tumors of the Peripheral Nervous System (Atlas of Tumor Pathology, 2nd Ser., Fasc. 3, Suppl.). Washington, DC, Armed Forces Institute of Pathology, 1983. 83. Ricci, A. J., Parham, D. M., Woodruff, J. M., et al.: Malignant peripheral nerve sheath tumors arising from ganglioneuromas. Am. J. Surg. Pathol. 8:19, 1984. 84. Roncaroli, F., Scheithauer, B. W., and Krauss, W. E.: Hemangioma of spinal nerve root. J. Neurosurg. 91(2 Suppl.):175, 1999. 85. Rusko, R. A., and Larsen, R. D.: Intraneural lipoma of the median nerve—case report and literature review. J. Hand Surg. Am. 6:388, 1981. 86. Sakaguchi, N., Sano, K., Ito, M., et al.: A case of von Recklinghausen’s disease with bilateral pheochromocytoma–malignant peripheral nerve sheath tumors of the adrenal and gastrointestinal autonomic nerve tumors. Am. J. Surg. Pathol. 20:889, 1996. 87. Salvati, M., Ciappetta, P., Raco, A., et al.: Radiationinduced schwannomas of the neuraxis: report of three cases. Tumori 78:143, 1992. 88. Scheithauer, B. W., Giannini, C., and Woodruff, J. M.: Perineurioma. In Kleihues, P., and Cavanee, W. K. (eds.): Pathology and Genetics of Tumors of the Nervous System. Lyon, IARC Press, p. 167, 2000. 89. Scheithauer, B. W., Nora, F. E., LeChago, J., et al.: Duodenal gangliocytic paraganglioma: clinicopathologic and immunocytochemical study of 11 cases. Am. J. Clin. Pathol. 86:559, 1986.

2605

90. Scheithauer, B. W., Woodruff, J. M., and Erlandson, R. A.: Tumors of the Peripheral Nervous System (Atlas of Tumor Pathology, 3rd Ser., Fasc. 24). Washington, DC, Armed Forces Institute of Pathology, 1999. 91. Schmitt, H. P., Wurster, K., Bauer, M., et al.: Mixed chemodectoma-ganglioneuroma of the conus medullaris region. Acta Neuropathol. (Berl.) 57:275, 1982. 92. Scotti, T. M.: The lesion of Morton’s metatarsalgia (Morton’s toe). Arch. Pathol. 63:91, 1957. 93. Shore-Freedman, E., Abrahams, C., Recant, W., et al.: Neurilemomas and salivary gland tumors of the head and neck following childhood irradiation. Cancer 51:2159, 1983. 94. Singh, R. V., Yeh, J. S., and Broome, J. C.: Paraganglioma of the cauda equina: a case report and review of the literature. Clin. Neurol. Neurosurg. 95:109, 1993. 95. Sonneland, P. R., Scheithauer, B. W., LeChago, J., et al.: Paraganglioma of the cauda equina region: clinicopathologic study of 31 cases with special reference to immunocytology and ultrastructure. Cancer 58:1720, 1986. 96. Sordillo, P. P., Helson, L., Hajdu, S. I., et al.: Malignant schwannoma—clinical characteristics, survival, and response to therapy. Cancer 47:2503, 1981. 97. Spjut, H. J., Dorfman, H. D., Fechner, R. E., et al.: Tumors of Bone and Cartilage (Atlas of Tumor Pathology, 2nd Ser., Fasc. 5). Washington, DC, Armed Forces Institute of Pathology, 1971. 98. Strong, E. W., McDivitt, R. W., and Brasfield, R. D.: Granular cell myoblastoma. Cancer 25:415, 1970. 99. Taxy, J. B., and Battifora, H.: Epithelioid schwannoma: a diagnosis by electron microscopy. Ultrastruct. Pathol. 2:19, 1981. 100. Terzis, J. K., Daniel, R. K., Williams, H. B., et al.: Benign fatty tumors of the peripheral nerves. Ann. Plast. Surg. 1:193, 1978. 101. Thomas, P. K., and Haftek, J.: Electron-microscope observation on the effects of localized crush injuries on the connective tissues of peripheral nerve. J. Anat. 103:233, 1968. 102. Usui, M., Ishii, S., Yamawaki, S., et al.: Malignant granular cell tumor of the radial nerve: an autopsy observation with electron microscopic and tissue culture studies. Cancer 39:1547, 1977. 103. Val-Bernal, J. F., Hernando, M., Garijo, M. F., et al.: Renal perineurioma in childhood. Gen. Diagn. Pathol. 143:75, 1997. 104. Vigna, P. A., Kusior, M. F., Collins, M. B., et al.: Peripheral nerve hemangioma: potential for clinical aggressiveness. Arch. Pathol. Lab. Med. 118:1038, 1994. 105. Weiner, H. L., Zagzag, D., Babu, R., et al.: Schwannoma of the fourth ventricle presenting as a hemifacial spasm: a report of two cases. J. Neurooncol. 15:37, 1993. 106. White, W., Shiu, M. H., Rosenblum, M. K., et al.: Cellular schwannoma: a clinicopathologic study of 57 patients and 58 tumors. Cancer 66:1266, 1990. 107. Wolansky, L. J., Stewart, V. A., Pramanik, B. K., et al.: Giant paraganglioma of the cauda equina in adolescence: magnetic resonance imaging demonstration. J. Neuroimaging 6:54, 1996. 108. Wong, W. W., Hirose, T., Scheithauer, B. W., et al.: Malignant peripheral nerve sheath tumor: analysis of treatment outcome. Int. J. Radiat. Oncol. Biol. Phys. 42:351, 1998.

2606

Diseases of the Peripheral Nervous System

109. Woodruff, J. M., Chernik, N. L., Smith, M. C., et al.: Peripheral nerve tumors with rhabdomyosarcomatous differentiation (malignant “Triton” tumors). Cancer 32:426, 1973. 110. Woodruff, J. M., and Christensen, W. N.: Glandular peripheral nerve sheath tumors. Cancer 72:3618, 1993. 111. Woodruff, J. M., Godwin, T. A., Erlandson, R. A., et al.: Cellular schwannoma: a variety of schwannoma sometimes mistaken for a malignant tumor. Am. J. Surg. Pathol. 5:733, 1981. 112. Woodruff, J. M., Horten, B. C., and Erlandson, R. A.: Pathology of Peripheral Nerves and Paragangliomas. New York, John Wiley, 1983. 113. Woodruff, J. M., Kourea, H. P., Louis, D. N., et al.: Malignant peripheral nerve sheath tumour (MPNST). In Kleihues, P., and Cavanee, W. K. (eds.): Pathology and

114.

115.

116.

117.

Genetics of Tumors of the Nervous System. Lyon, IARC Press, p. 172, 2000. Woodruff, J. M., Kourea, H. P., Louis, D. N., et al.: Schwannoma. In Kleihues, P., and Cavanee, W. K. (eds.): Pathology and Genetics of Tumours of the Nervous System. Lyon, IARC Press, p. 164, 2000. Woodruff, J. M., Marshall, M. L., Godwin, T. A., et al.: Plexiform (multinodular) schwannoma: a tumor simulating the plexiform neurofibroma. Am. J. Surg. Pathol. 7:691, 1983. Woodruff, J. M., Selig, A. M., Crowley, K., et al.: Schwannoma (neurilemoma) with malignant transformation: a rare, distinctive peripheral nerve tumor. Am. J. Surg. Pathol. 18:882, 1994. Yakulis, R., Manack, L., and Murphy, A. I. J.: Postradiation malignant triton tumor: a case report and review of the literature. Arch. Pathol. Lab. Med. 120:541, 1996.

R. Management

116 Management of Patients with Acute Neuromuscular Disease in the Intensive Care Unit EELCO F. M. WIJDICKS

ICU Admission Criteria General Aspects of ICU Care Skin, Eye, and Mouth Care Deep Venous Thrombosis Prophylaxis Gastrointestinal and Urinary Tract Complications Nutrition and Hydration Considerations

Management of Acute Dysautonomia Management of Respiratory Failure Principles of Respiratory Physiology Effects of Acute Neuromuscular Disease on Respiration Monitoring Pulmonary Function

Seriously ill patients with acute neuromuscular disease are treated in critical care facilities. These patients have a better chance of survival if they are treated in intensive care units (ICUs) where respiratory failure and life-threatening dysautonomia can be managed. The development of intensive care and mechanical ventilation is linked historically to treatment of patients with acute neuromuscular disease. The poliomyelitis epidemics spurred development of positive pressure ventilators. In the United States tank respirators were used, but in Europe shortages of materials led to use of other methods.26,113 Ibsen51 has been credited with the invention of the first manual ventilator, which used an oxygen source, a humidifier, soda lime to absorb carbon dioxide, and an endotracheal tube, and medical students were paid to “bag” patients who needed mechanical ventilation.58 Mortality resulting from respiratory failure was reduced from near 90% to 37%. Contemporary ICUs admit patients with neuromuscular disease caused by Guillain-Barré syndrome (GBS), amyotrophic lateral sclerosis (ALS), toxic neuropathies,

Intubation and Mechanical Ventilation Tracheostomy Weaning from Ventilation Management of Systemic Complications Withdrawal of Respiratory Support and Palliation Conclusion

and disorders of neuromuscular function. Triage of certain patients with a neuromuscular disorder to an ICU is needed for two reasons. First, acute neuromuscular disease of any cause may involve the patient’s respiratory system and increase the patient’s propensity toward respiratory failure.43,45–47,60 Appropriate monitoring can be achieved only in ICUs. Second, polyneuropathy may be a manifestation of a systemic illness (e.g., vasculitis), and a different medical manifestation (e.g., intrinsic pulmonary disease) may prompt the physician to transfer the patient to the ICU. For some patients, admission to the ICU is due to respiratory failure of unknown cause. These patients are admitted with acute hypercarbia and hypoxemia, they are intubated, and a neurologic consultation is requested to assess a possible underlying neurologic disorder that had been underdiagnosed. A large proportion of these patients eventually are diagnosed as having ALS. The care of many patients with neuromuscular disease is entrusted to medical intensivists. More recently, with the development of neurologic ICUs, patients with acute 2607

2608

Diseases of the Peripheral Nervous System

neuromuscular disease are admitted to these specialized units. Patients with acute neuromuscular disease have unique needs. Most of their care is focused on management of respiratory failure resulting from diaphragmatic weakness and management of pulmonary morbidity associated with mechanical ventilation.42,104 Our recent review of patients with severe GBS admitted to the ICU emphasized the high prevalence of pulmonary morbidity and systemic infection in one fifth of the patients, and the prevalence was more frequent with increasing duration of ICU stay.42 This chapter addresses the core of management of patients with GBS, ALS, and other acute neuromuscular disease in the ICU, principally decisions regarding intubation, mechanical ventilation, tracheostomy, and weaning, as well as some everyday issues such as nutrition, bladder and bowel care, and use of prophylactic medication. Some aspects of acute dysautonomia are explained, but the management of patients with autonomic failure is discussed further in Chapter 122.

ICU ADMISSION CRITERIA The decision to transfer a patient to an ICU spans the entire spectrum between ICU admission of patients with obvious clinical signs that suggest critical illness and situations in which concern regarding progression of disease guides ICU admission. Table 116–1 outlines criteria to apply to ICU admission of a patient with any type of acute neuromuscular disease. These criteria, a combination of clinical judgment and laboratory tests, are not absolute, and physician discretion should be exercised. Care of patients admitted to the neurology service with polyneuropathy and in rapidly worsening condition is enhanced by monitoring in the ICU, even for a short time (e.g., overnight observation). Rapid development of inability to support one’s own weight, involvement of upper extremities, and compromised bulbar function are legitimate reasons for ICU admission.40,60 Obviously a clearer and more compelling criterion is the presence of abnormal pulmonary function. In children, rapid onset of symptoms (within 8 days) and cranial nerve dysfunction are strong indicators of further worsening.88 However, data are limited on acute rapid deterioration of the condition of patients with acute neuromuscular disease. In our review of 114 patients with severe GBS admitted to the ICU, a large proportion had rapid decompensation (within 1 to 2 days).60 Urgent intubation was necessary in 16% of them. Some patients required intubation in the middle of the night and early morning hours. Whether reduction in centrally mediated respiratory drive during sleep or compromised respiratory mechanics in a supine position play a role is not clear. Early intervention may reduce pulmonary morbidity.77

Table 116–1. Admission Criteria to Intensive Care Unit Pulmonary infiltrates Tachypnea, hypoxemia or increased alveolar-arterial oxygen tension gradient Oropharyngeal dysfunction VC ⬍ 20 mL/kg PI max ⬍ ⫺30 cm H2O PE max ⬍ 40 cm H2O ⬎30% reduction in serial pulmonary function measurements Acute dysautonomia (tachyarrhythmia, blood pressure instability) PE max ⫽ maximum expiratory pressure; PI max ⫽ maximum inspiratory pressure; VC ⫽ vital capacity.

Dysautonomia in patients with GBS may cause cardiac arrest. In a review of fatal cases of GBS, we found no such incidences in the early progression of the disease and no documented cases of acute dysautonomia or cardiac arrhythmia that predated resuscitation.61 More commonly, the time of death was months into the illness and often associated with ventilator-associated pneumonia or acute respiratory distress syndrome. Data concerning other acute neuromuscular disorders are not available. These observations suggest that neurologists should promptly refer patients with an acute neuromuscular disorder in unstable condition to an ICU, but further research is needed to refine the admission criteria. Factors that are difficult to measure but may be associated with adverse outcomes are staffing levels of nurses (weekends vs. weekdays), sophistication of care, and volume of patients.6 We found that delayed transfer of critically ill, ventilated patients with GBS to our institution affected clinical outcomes.31 With administration of intravenous immune globulin, transfer to specialized centers may be deferred.

GENERAL ASPECTS OF ICU CARE Patients with severe acute neuromuscular disease require specialized care. Aggressive support of vital functions is the mainstay, but management in the ICU may relate also to complications of the disease and treatment. Patients with neuromuscular disorders are predisposed to have considerable morbidity. Prolonged mechanical ventilation may increase the risk of ventilator-associated pneumonia in these patients. Plasma exchange used as treatment may lead to hypotension or cardiac arrhythmias, and catheters to access veins may be misplaced. Most patients with neuromuscular disease who are admitted to the ICU have rapidly progressing GBS.6,42,49,77,80,96 In fact, much of the currently available morbidity data are

Management of Patients in the Intensive Care Unit

Table 116–2. Causes of Morbidity in 114 Patients with Guillain-Barré Syndrome Admitted to an Intensive Care Unit Patients Morbidity Respiratory Tracheobronchitis Pneumonia Pneumonia with effusion, cavitation, or abscess Pneumothorax Tracheostomy related Infection Bacteremia Sepsis related to line placement

No.

%

33 18 9

29 16 8

6 2

5 2

21 3

18 3

6 5 4 4 3 2 2 7

5 4 4 4 3 2 2 6

75

66

28 5

25 4

Miscellaneous Major Morbidity Gastrointestinal tract hemorrhage Deep venous thrombosis Multisystem morbidity Detection of underlying malignancy Pulmonary embolus Hepatitis Pseudomembranous colitis Other* Minor Morbidity Positive urinary cultures of a pathogenic organism Hyponatremia (⬍130 mmol/L) Hyperglycemia requiring treatment

*Other causes of morbidity included stroke, perforated ulcer, membranous nephropathy, ventricular tachycardia on intubation with succinylcholine chloride, heparin-induced thrombocytopenic purpura, and severe anoxic encephalopathy. From Henderson, R. D., Lawn, N. D., Fletcher, D. D., and Wijdicks, E. F.: The morbidity of Guillain-Barré syndrome admitted to the intensive care unit. Neurology 60:17, 2003, with permission of the American Academy of Neurology.

from a series on morbidity in 114 patients with GBS (Table 116–2).42 Major morbidity occurred in approximately one half of the patients and was more common if the ICU stay was prolonged because the patient’s condition failed to improve. Pneumonia and tracheobronchitis often were associated with prolonged mechanical ventilation. Complications caused by specific treatment (e.g., plasma exchange) or catheter-related complications (e.g., subclavian catheter) were uncommon.

2609

special bed (air fluidization) may be required, but modern ICU beds have a support surface sufficient to retard formation of decubitus ulcers. Frequent repositioning and daily skin inspection also prevent decubitus ulcer formation.3,53 Nonblanchable discoloration may develop in quadriplegic patients, and subsequent partial-thickness skin loss may involve the epidermis, particularly when systemic infection occurs or nutrition is insufficient.99 Use of artificial tears is necessary when bilateral facial palsy is present. Mouth care involves normal tooth brushing and daily inspection for the development of oral candidiasis.

Deep Venous Thrombosis Prophylaxis Immobilization is a risk factor for development of deep venous thrombosis (DVT).8,64 Patients with leg weakness from acute neuromuscular disease are at risk. DVT typically develops in a paralyzed leg, but any degree of weakness predisposes to clot formation because of reduced calf muscle function. The true incidence is not known because many patients with DVT are asymptomatic, and in them DVT can be documented only noninvasively by using duplex ultrasonography.23,32 Extension to proximal veins increases the risk of pulmonary emboli; a risk of 10% to 20% exists for isolated calf vein thrombus.54,74 Prophylaxis against thrombosis in patients with acute neuromuscular disease has not been studied systematically, nor has prevention in children, who have a low incidence of DVT. The most effective methods of prophylaxis include intermittent use of compression devices, alone or in combination with aspirin or a low dose of heparin. Studies of patients with orthopedic, general surgical, or spinal cord injuries suggest subcutaneous administration of heparin (5000 U twice a day) to prevent DVT. However, these recommendations cannot be extrapolated to the population of patients with acute neuromuscular disease.20,21,100,106 Low-molecular-weight heparin is twice as effective as unfractionated heparin, but its cost is substantial.106,107 Subcutaneous heparin (5000 U twice daily) is an acceptable method of prevention. However, patients with a much higher incidence of DVT, particularly those with prior venous insufficiency or varicose veins, probably should be treated with subcutaneous enoxaparin (40 mg once daily).92 Fever and calf tenderness are often the sole presenting signs of DVT. Swelling occurs later.

Skin, Eye, and Mouth Care

Gastrointestinal and Urinary Tract Complications

To prevent decubitus keratitis and Candida infections in the oral cavity, general care involves skin, eye, and mouth care, all important to provide optimal comfort to patients who are bedridden and quadriplegic. For some patients, a

Prophylaxis for stress gastric ulcers should include use of a proton pump inhibitor (e.g., omeprazole, 40 mg daily per nasogastric tube); the cost is reasonable and it provides the best protection against gastrointestinal (GI) tract bleeding.

2610

Diseases of the Peripheral Nervous System

The use of ranitidine is not recommended in elderly patients with renal failure because of its tendency to induce an acute confusional state. GI tract prophylaxis probably should be considered for high risk patients on mechanical ventilation, patients with underlying coagulopathy, and patients taking corticosteroids. Adynamic ileus is more likely to occur in severely affected patients with acute neuromuscular disease.16 Abdominal pain and nausea are uncommon, although they may indicate food-borne botulism, particularly if nausea precedes muscle weakness and bulbar signs are prominent. Constipation occurs frequently in bedridden patients. In GBS patients, approximately one half of patients may develop adynamic ileus in the acute phase, often in conjunction with other features of dysautonomic disease. Long-term immobilization, use of opioids for pain control, or immobility in critical illness that reduces mesenteric blood flow and produces imbalance of the sympathetic and parasympathetic regulation may cause adynamic ileus.18 It may also be caused by an increasing dose of narcotics, use of tricyclic antidepressants, and phenothiazine use. Commonly, pain, distention, and tympany are obvious symptoms, but colonic distention may occur before these symptoms appear. Recognition of adynamic ileus hinges further on routine daily measurement of abdominal girth and abdominal auscultation (paucity of sounds or quiet), particularly in patients with pain from abdominal distention. Plain abdominal radiography may document dilated colonic loops and fluid and air levels. Management of adynamic ileus may include suspension of enteral feeding, replacement of fluids and electrolytes (particularly potassium), nasogastric suctioning, and use of promotility agents. In particular, the use of erythromycin or neostigmine has been effective in the treatment of adynamic ileus, although these agents have not been tested in patients with acute neuromuscular disease. In patients with GBS, dysautonomia may be a contraindication to use these drugs with a tendency to cause cardiac arrhythmias. Maintaining bladder function in patients with acute neuromuscular disease in the ICU is complicated. Most patients are catheterized to avoid bladder distention and to maintain good hygiene. Management of bladder dysfunction includes maintenance of a sterile, closed urinary drainage system, integrity of the seals to obtain urinary samples, and irrigation of the bladder. In any febrile illness, urine cultures and aggressive management of infection are necessary. Bladder areflexia is common in patients who become quadriplegic, and catheterization is advised at an early stage.91

Nutrition and Hydration Considerations Fluid intake in adult patients who are in ICUs should always include a daily 2-L intravenous infusion of isotonic saline, but fluid administration should be adjusted

when enteric feeding is started. In any patient with mechanical ventilation and a fever, a higher fluid intake is necessary. In all patients, the insensible loss of fluid is underappreciated. Hypovolemia may cause hypotension that should not be attributed to dysautonomia. Patients on mechanical ventilation may lose liters of fluid despite adequate humidification, and fluid loss resulting from fever may amount to 1 L/day. Blood products and plasma exchange may shift intravascular volume. Typically, these shifts are anticipated and corrected, but inadequate or hypo-oncotic volume replacement may cause hypovolemia and even hypotension. The use of 5% albumin with continuous flow separations and matched input and output sufficiently circumvents this problem. The fluid status is best assessed by urinary output, which should be more than 0.5 mL/kg/h. Generally, fluid intake should approximate 30 mL/kg/day, and a positive fluid balance of at least 1 L/day should be maintained. Regular measurement of body weight could guide fluid intake. Laboratory measures such as hematocrit, serum sodium, and osmolality are also important to judge the overall fluid status. Early attention to nutritional status is imperative, and if the patient’s GI tract is functional, it is the best means of supplying energy. Patients with acute neuromuscular disease rarely display a severe sustained catabolic response but still may become underfed rapidly. In patients with diminished swallowing mechanisms, a nasogastric tube is placed and daily residuals are monitored. Full-strength enteral nutrition is used, and a stool softener (bisacodyl, 2 tablets a day) is helpful. Before a nasogastric tube is placed, a plain abdominal radiograph should be obtained to exclude the possibility of a paralytic ileus. The total nutritional needs of a patient with an acute neuromuscular disorder in the ICU do not necessarily include a stress factor, which typically is the basal energy expenditure in calories plus 20%. The Harris-Benedict formula is based on kilograms of weight (W), centimeters of height (H), and years of age (A). For men, the formula is 66.5 ⫹ 13.8 W ⫹ 5.0 H ⫺ 6.8 A; for women, the formula is 655 ⫹ 9.6 W ⫹ 1.8 H ⫺ 4.7 A. Enteral feeding includes continuous infusion with use of a volumetric pump. Various compositions of commercial formulas are available to meet the caloric needs of individual patients. Gastrostomy is indicated in patients who need prolonged ventilator support, such as patients with end-stage ALS or poorly recovering GBS. The decision to place a percutaneous endoscopic gastrostomy (PEG) is arbitrary, and it probably should not be inserted within the first month of ICU stay. It certainly should be considered if permanent ventilator support is anticipated and bulbar dysfunction does not improve. PEG placement requires antibiotic prophylaxis. Typically, a single 1-g dose of cefazolin is used before a gastroenterologist places the PEG.

Management of Patients in the Intensive Care Unit

Complications of placement are low, but pneumoperitoneum may occur in up to 40% of the patients. Recent studies have specifically addressed the placement of a PEG in ALS patients; use of noninvasive positive pressure ventilation facilitated placement and reduced compromise of patient breathing.9,37,73

MANAGEMENT OF ACUTE DYSAUTONOMIA In clinical practice, acute dysautonomia is encountered most often in patients with severe forms of GBS.28,29,67,82,111,116 Other acute neuromuscular disorders associated with ileus and blood pressure fluctuations are acute intermittent porphyria, botulism, thallium and lead poisoning, and chemotherapy-induced toxic neuropathy. The most common manifestations are blood pressure fluctuation, arrhythmia, bronchial smooth muscle dysfunction, and disordered thermoregulation. The pathophysiology of each of these conditions is discussed elsewhere in this text. Prediction of cardiac arrhythmias is difficult.94,103 Flachenecker et al.30 suggested that bilateral pressure on eyeballs caused vagal overreactivity and thus induced bradycardia. However, the predictive value of vagal overreactivity for cardiac arrest or the need for a pacemaker remains unresolved, and inducing bradycardia is dangerous without direct availability of intravenous atropine. A 24-hour heart rate power spectrum did detect bradyarrhythmias in patients with GBS and may become useful,28,29 but it has not been used widely in ICUs. Temporary transvenous pacemaker placement should be considered when vagal spells are frequent and associated with hypotension or syncope. Sinus tachyarrhythmias may respond well to treatment with ␤-blockers such as esmolol.17 Most challenging are uncontrolled blood pressure fluctuations. They are caused by afferent baroreflex abnormalities and may be exacerbated by poor central venous filling pressures or changes in systemic vascular resistance. A SwanGanz catheter should be placed to assess intravascular volume and subsequently to optimize it. Marked hypotension not responding to placing the patient in Trendelenburg’s position should be treated with phenylephrine (10 to 30 mcg/min) combined with dopamine (1 to 5 mcg/kg/min) to reduce reflex bradycardia. Hypertensive surges may require urgent treatment with nicardipine hydrochloride (2.5 to 5 mg/hr infusion) or fenoldopam (0.03 to 0.1 ␮g/kg/min) with titration to effect. We also found IV monfine helpful (a potent vasodilator both arterial and venous). One should be familiar with the hypersensitivity to antihypertensive and pressor drugs, possibly attributable to dysfunction of the baroreflex. Marked blood pressure excursions are difficult to treat, and blunting of the extremes may be the best option until symptoms disappear within a few days.

2611

MANAGEMENT OF RESPIRATORY FAILURE Respiratory failure is the most common reason for admission to the ICU.34 It is useful to review some of the essentials of respiratory muscle physiology to understand the clinical manifestations of respiratory failure.

Principles of Respiratory Physiology The respiratory pump is determined by the diaphragm and intercostal muscles.24 The diaphragm is the most important muscle of inspiration; it uses the abdomen as a fulcrum and it can expand the rib cage. The fibers of the diaphragm insert anteriorly to the posterior portion of the xiphoid process and posteriorly to the first three lumbovertebral bodies. During inspiration, the diaphragmatic dome descends and displaces the abdominal contents caudad and outward. Thus, during inspiration, both the rib cage and abdominal contents move outward. The intercostal muscles extend obliquely caudad and backward from the rib above to the rib below and have an additional inspiratory function. The scalene muscles that originate from the transverse process at the lower five cervical vertebrae and insert on the upper surface of the first rib and the second rib also expand the rib cage but are active only when other respiratory muscles fail. Expiration occurs as a result of passive recoil, although the abdominal muscles can be recruited if necessary.68,69 Abdominal contraction during expiration not only assists in the diaphragmatic function but also enhances coughing to clear secretions from the bronchial tracts. Neuromuscular respiratory failure is accompanied by a poor coughing mechanism, and improvement of respiratory function can often be measured by improvement of the abdominal strength. This is also reflected in measurement of pulmonary function, in which the maximum expiratory pressure (PE max) seems to have a much better predictive value than the maximum inspiratory pressure (PI max).38

Effects of Acute Neuromuscular Disease on Respiration The majority of patients with acute neuromuscular disease, because of their youth, have normal airways and pulmonary parenchyma. Pulmonary dysfunction is purely a consequence of failure of the respiratory mechanics. Vital capacity declines progressively when no strength is mounted.19 However, in long-term disorders of neuromuscular dysfunction, particularly ALS, in which motor neurons in the anterior horns of the spinal cord might be lost, respiratory dysfunction may involve spasticity of the rib cage musculature and prevent outward movement of the chest wall.15 Cough efficacy is diminished because of decreased expiratory muscle strength.84,85 The anatomy

2612

Diseases of the Peripheral Nervous System

FIGURE 116–2 Clinical and laboratory tests to assess neuromuscular respiratory function. PE max ⫽ maximum expiratory pressure; PI max ⫽ maximum inspiratory pressure; VC ⫽ vital capacity.

FIGURE 116–1 Pulmonary muscle pump. Inserts show positions of diaphragm: normal (A); normal inspiration (B); and paradoxical inspiration (C). Arrows indicate direction of diaphragm, chest, and abdominal movement. (From Wijdicks, E. F. M.: Clinical Practice of Critical Care Neurology, 2nd ed. New York, Oxford University Press, 2003, with permission of Mayo Foundation.)

and changes in motion in acute neuromuscular failure are depicted in Figure 116–1. Clinically, neuromuscular respiratory failure is evident when the patient attempts to increase the number of respirations to compensate for the inability to increase tidal volume.55,86,89,90,98 Tachycardia and sweating have been interpreted as an adrenergic response to the increased work of breathing. Contraction of the sternocleidomastoid and scalene muscles can be palpated and is synchronous with any breathing effort. The patient complains of inability to breathe and becomes restless.108 Symptoms and signs are more common when the patient is supine, because the action of the diaphragm on the rib cage is reduced and the inspiratory rib cage muscles are less active. Breathing movements in the supine posture are largely abdominal.12 It is advisable to assess respiratory excursions in a patient both supine and upright; the results of pulmonary function tests may differ in each posture.105

Monitoring Pulmonary Function Devices are available to monitor pulmonary function, both its mechanics and gas exchange measurements. Some of

these devices can be used while the patient is on mechanical ventilation. The most commonly used device to measure acute neuromuscular failure is a bedside spirometer. The most important spirometer data are vital capacity, PI max, and PE max (Fig. 116–2). Vital capacity is the volume of gas measured from a slow, complete forced expiration after maximum inspiration. It is usually recorded in liters or milliliters per kilogram. Vital capacity has been used to determine diaphragmatic function.83 However, it can be reduced by additional airway and pulmonary problems, such as mucus plugging and bronchial narrowing. In fact, a forced vital capacity decrease is a common feature of restrictive pulmonary disease. PI max is recorded when a patient inspires forcefully against an occluded airway. Typically, PI max is measured at the end of maximum expiration and, therefore, has a negative value as a result of the inspiratory effort in the presence of an occluded airway. It is measured as centimeters of water or millimeters of mercury. PE max is the maximum pressure generated by the patient during a forceful expiratory effort from an occluded airway. Technique is important in spirometry, and results are critically dependent on the effort of the patient. A scuba device is used to provide an adequate seal. The patient is connected to this apparatus, a nose clip is placed, and the airway is occluded by blocking a port in the valve or by closing a shutter. The manometer records ⫺10 to ⫺200 cm H2O. The instructions to the patient are important. Patients who have generalized weakness from myasthenia gravis or GBS usually need coaching to achieve adequate measurements. The second effort is often the patient’s best effort, and multiple efforts yield worsening values. Many patients have a tendency to produce a Valsalva maneuver in which the required pressure is not generated and values are falsely low. Lack

2613

Management of Patients in the Intensive Care Unit

of understanding by the patient of how to perform this test remains a major problem in obtaining accurate spirometric values. Normal adults can generate at least ⫺60 cm H2O of PI max, and inspiratory pressures are decreased in patients who have weakness of the mechanical function of the rib cage. However, PI max can also be decreased in patients with hyperinflation disorders such as emphysema. This condition in advanced stages flattens the diaphragm because of gas trapped in the lungs. PE max is largely a function of the abdominal and accessory muscles of respiration and some elastic recoil of the lungs. Normal adults can easily generate pressures higher than 200 cm H2O. The analysis of gas exchange can be obtained by using a pulse oximeter. In principle, the pulse oximeter measures the amount of oxygen bound to hemoglobin. The concentration of light beaming through a hemoglobin sample is a logarithmic function of the amount of light absorbed by the sample (normal, ⬎90% saturation). The oxyhemoglobin and deoxyhemoglobin can be differentiated by shining, respectively, 660- and 940-nm wavelengths of light through the index finger. The limitations are motion artifacts, intravascular dyes, low perfusion states, skin pigmentation, and nail polish. Methemoglobinemia is not measured. (A patient with methemoglobinemia can be markedly cyanotic with no signs of respiratory distress, and the pulse oximeter may indicate oxygen saturation of 85%. Methemoglobinemia is seen occasionally in the ICU after use of lidocaine, particularly in spray form.)

Intubation and Mechanical Ventilation The criteria for endotracheal intubation remain arbitrary, but the indications should be obvious in any patient with dyspnea, tachycardia, and sweating on the forehead. The time course of declining respiratory function is unpredictable, but the results of pulmonary function tests can be used. Fatigue of respiratory muscles is seen most commonly in patients with rapidly progressive acute neuromuscular disease, and asynchronous movement of the chest and abdomen and use of accessory muscles are major hallmarks in clinical examination. It is far better to perform endotracheal intubation electively rather than to wait for respiratory function to deteriorate further.110 Patients with oropharyngeal dysfunction have difficulty clearing secretions, which, in addition, may compromise gas exchange and increase the risk of aspiration. Patients with pulmonary infiltrates or severe atelectasis may require intubation and mechanical ventilation earlier. Aggressive physical therapy does not necessarily prevent the need for mechanical ventilation, although physical therapy is usually performed routinely. Physical therapy may include percussion; coughing exercises; use of mucolytic agents, bronchodilators, or nebulizers; and incentive spirometry.2

Table 116–3. Laboratory Criteria for Mechanical Ventilation Criterion

Value Necessitating Mechanical Ventilation

Decreased Oxygenation PaO2* P(A–a)O2

⬍70 mm Hg ⬎70 mm Hg

Decreased Ventilation pH PaCO2 VD /VT

⬍7.25 ⬎55 mm Hg ⬎0.6

Decreased Mechanics VC PI max PE max Respiratory frequency

⬍20 mL/kg ⬍⫺30 cm H2O ⬍40 cm H2O ⬎35/min

PaCO2 ⫽ partial pressure of carbon dioxide; PaO2 ⫽ partial pressure of oxygen; P(A–a)O2 ⫽ alveolar-arterial difference in partial pressure of oxygen; PE max ⫽ maximum expiratory pressure; PI max ⫽ maximum inspiratory pressure; VC ⫽ vital capacity; VD ⫽ dead space volume; VT ⫽ tidal volume. *With oxygen prongs on face mask providing 2 to 4 L of oxygen.

The basic laboratory criteria for mechanical ventilation are shown in Table 116–3. Often a combination of these factors is present. The need for mechanical ventilation in neurologic patients is different from that in patients in other medical ICUs. In many neurologic patients, the lungs were normal before the neurologic event occurred, although the incidence of aspiration of gastric acid contents is considerably higher. Patients with long-term neuromuscular disease have reset respiratory center thresholds and can maintain adequate respiration. However, oxygen delivery may cause marked hypercarbia and carbon dioxide narcosis, even with oxygen flow of 2 L/min.35 The ideal modes of mechanical ventilation in patients with acute neuromuscular failure have not been established, but in many cases, intermittent and mandatory ventilation and pressure support are required102 (Table 116–4). Controlled mechanical ventilation is used rarely in patients with acute neuromuscular failure, only when extreme pulmonary morbidity arises. The development of diffuse pneumonitis may change mechanical ventilation into pressure-controlled ventilation. In this setting, the tidal volume is set, as well as the frequency and timing of each breath. In the assist-controlled ventilation mode, a preset number of breaths per minute is delivered to the patient. When no patient effort exists, this mode is synchronized to avoid breath stacking. In fact, triggered breaths delay the next machine-determined breath. Pressure support is present only if the patient makes a respiratory effort, and then a set pressure is provided.

2614

Diseases of the Peripheral Nervous System

Table 116–4. Modes of Mechanical Ventilation Mode

Advantages

AC

Reduces work of breathing Patient increases mechanical breaths Hemodynamic stability Reduced intrathoracic pressure Patient comfortable Reduced sensation of being on ventilator

SIMV

PS

CMV

Resting pulmonary muscles

Major Concerns Common hypocapnia

Patient-ventilator dyssynchrony Usually combined with SIMV, may be insufficient when used alone Sedation and blockade of neuromuscular function required

AC ⫽ assist controlled; CMV ⫽ controlled mechanical ventilation; PS ⫽ pressure support; SIMV ⫽ synchronized intermittent mandatory ventilation.

and hypercarbia, or in patients who cannot tolerate the tightly strapped device. The ventilators provide positive airway pressure and expiratory positive airway pressure. The patient needs to trigger the ventilator to obtain the respiratory-assisted breath. Use of noninvasive ventilation in the ICU typically is a trial that is followed by endotracheal intubation if it fails. A recent study of patients with acute respiratory failure in motor neuron disease found that, when patients received noninvasive ventilation acutely, they rarely became independent of the device.13 However, several patients were able to receive intermittent positive pressure ventilation through a tracheostomy. Few, however, were able to breathe successfully through a face mask. Noninvasive ventilation in patients with more long-standing symptoms seemed to be successful up to 15 months in one study,36 and quality of life appeared to be influenced positively.66

Tracheostomy Continuous positive airway pressure (CPAP) is an ideal weaning mode, because it unloads the inspiratory muscles and reduces inspiratory work. Volume ventilation is the preferred mode for adult ventilation, with tidal volumes ranging from 100 to 500 mL. The usual recommended frequency ranges from 8 to 10 breaths/min. Typically, positive end-expiratory pressure or CPAP ranges from 0 to 30 cm H2O with flow rates between 10 and 180 L/min. Usually flow waveforms are constant, with descending waveform tracings. The available modes of ventilation (see Table 116–4) include assistcontrolled ventilation, synchronized intermittent mandatory ventilation (SIMV), CPAP, and pressure support. Tidal volume of 12 to 15 mL/kg maintains alveolar pressure at less than 35 cm H2O, respiration at 10 to 20 breaths/min, and good alveolar carbon dioxide tension. Fractionated inspired oxygen of 1.0 is good initially, but this value must be titrated downward to reduce oxygen toxicity. High inspiratory flow is preferred, and positive end-expiratory pressure is added (typically between 0 and 5 cm H2O). Noninvasive ventilation has become a potential alternative to invasive mechanical ventilation27,44 and is particularly useful in patients with longer term neuromuscular failure and a good option for patients with ALS.1 Noninvasive ventilation includes an assist device for use by patients with ALS,76 but it has not been used systematically in patients with acute neuromuscular disorders.5 It may not be useful in patients with GBS because of the rapid and profound decline in their condition. Noninvasive ventilation is not a substitute for endotracheal intubation and mechanical ventilation, and the choice of which form of ventilation to use must be made carefully. Noninvasive ventilation is not used in patients with excessive respiratory secretions, in patients who have had severe hypoxemia

A tracheostomy is necessary when prolonged mechanical ventilation or maintenance of the upper airway is anticipated. A tracheostomy is often contemplated in patients with anticipated long-term mechanical ventilation for neuromuscular disease and may be permanent in patients with progressive neuromuscular disease such as ALS. Generally, a tracheostomy should be postponed up to 3 weeks after intubation. We have been able to predict the duration of mechanical ventilation and tailor the use of tracheostomy in patients with GBS. In a study of 37 patients with GBS who received mechanical ventilation, calculating a pulmonary function ratio (defined as the sum of vital capacity, PI max, and PE max, divided at day 12 by the pulmonary function score at the day of endotracheal intubation) can predict the need for tracheostomy. With a pulmonary function ratio of less than 1, it is highly unlikely that patients can be weaned from mechanical ventilation within 3 weeks. The sensitivity of a pulmonary function ratio of less than 1 to predict duration of ventilation as more than 3 weeks is 70%, and the specificity or positive predictive value is 100%. If the pulmonary function ratio is more than 1, tracheostomy should be deferred an additional week to allow the patient an attempt to wean of the ventilator.62 The benefits of tracheostomy are notable. It allows better airway safety, and patients can suction their own tracheostomies. Furthermore, patients are able to eat, which eliminates the need for gastric tube feeding, and tracheostomy may facilitate weaning by reducing the volume of dead space (inability of certain lung regions to participate, thus causing wasted ventilation). In the event that mechanical ventilation is needed again, reintubation is not necessary.

Management of Patients in the Intensive Care Unit

Current endotracheal tubes are well tolerated, and placement for 3 weeks does not increase the likelihood of tracheal stenosis.33,41 The risks of tracheostomy are substantial, however, with bleeding at the site, accidental extubation and failure to find an airway for reintubation. In many patients, scarring is noticeable when a traditional surgical tracheostomy is performed.109 In patients with a disorder such as GBS that usually has a good outcome, a tracheostomy can cause permanent disfigurement. Percutaneous dilatation on tracheostomy rather than traditional surgical tracheostomy may be preferable in patients who need short-term endotracheal intubation and mechanical ventilation.71,72,95 The procedure involves a small skin incision, insertion of the cannula to the trachea, and insertion of dilators of gradually increasing size until a tracheostomy tube of appropriate size can be accommodated. Percutaneous tracheostomy may reduce the risk of accidental intubation because the tube fits more snugly around the stoma, and the smaller skin incision produces a better outcome.

Weaning from Ventilation Protocols to wean patients from mechanical ventilation include intermittent T-piece trials, SIMV, pressure support, CPAP, or a combination of these. Noninvasive positive pressure ventilation has become a preferred mode of weaning in patients with neuromuscular failure. However, weaning protocols have problems. No criteria exist that specify when a patient is ready to try breathing spontaneously, nor are there standards to judge a trial successful. Moreover, the criteria for extubation are usually poorly defined and biased by physician preferences. Many studies, however, suggest that multiple daily trials of T-piece breathing or pressure support might be superior to gradually reducing the frequency of SIMV. In patients with neurologic disease who are receiving mechanical ventilation, weaning criteria have not been defined, and prolonged mechanical ventilation may be attributable to clinician procrastination rather than inability of the patient to breathe without the ventilator. In patients with neuromuscular disease, measurement of pulmonary function is often discontinued inappropriately. Judicious use of pulmonary function testing in patients with GBS may indicate readiness for extubation after 1 to 2 weeks of mechanical ventilation, and tracheostomy may not be necessary. Mechanical ventilation is temporary in patients with GBS, but it is necessary in many other acute neuromuscular disorders such as ALS. The prognosis for GBS patients on mechanical ventilation is poor, with a mortality of approximately 20%.31 Outcome is largely determined by the duration of time on the ventilator; ventilation for longer than 4 months is invariably associated with a poor outcome, and all patients have a severe permanent disabil-

2615

ity. Time on the ventilator markedly increases pulmonary morbidity; thus aggressive management may reduce these complications. In summary, weaning should be guided by overall clinical improvement and improvement in pulmonary function. Pulmonary function should continue to be tested while the patient is intubated. Weaning protocols are not available, but generally CPAP is initiated.

MANAGEMENT OF SYSTEMIC COMPLICATIONS Little is known about the morbidity of patients admitted to ICUs.60,61,78 Recently, a survey of patients with GBS indicated that pulmonary complications are quite prevalent when mechanical ventilation is extended beyond 3 weeks.31 Nosocomial ventilator-associated pneumonia may evolve to a severe pulmonary dysfunction such as acute respiratory distress syndrome, which is difficult to manage and has high mortality. This type of lung injury is driven by inflammatory mediators; free radicals lead to often-irreversible tissue damage, and lung compliance is dramatically reduced.4,7,114 An interesting research question is whether mechanical ventilation can induce acute respiratory distress syndrome. Early increased dead space fraction resulting in shunting and hypoxemia is a major predictor of mortality.79 Hypoxemia in itself may promote the inflammatory responses that lead to the downward spiral of a patient’s condition, despite maximum support.97 Nosocomial bronchopneumonia is difficult to diagnose for several reasons. First, the sensitivity and specificity of infiltrates seen on chest radiography are poor. Second, microbiologic diagnosis often reveals multiple types of infection, and third, clinical examination findings may be inconclusive. Criteria for the diagnosis of nosocomial ventilator-associated pneumonia have been proposed.87 They include presence of fever (⬎39 °C), increased blood leukocyte count (⬎11 ⫻ 109/L), presence of purulent tracheal secretions, poor oxygenation despite maximal delivery, localized infiltrates on chest radiography, progression of findings, and presence of pathogenic bacteria in sufficient quantity.87 The risk factors are shown in Table 116–5. Pseudomonas aeruginosa and Staphylococcus aureus are typically implicated, and antibiotic therapy should reflect appropriate coverage for these pathogens. A third-generation cephalosporin (e.g., ceftazidime, 1 g every 8 hours intravenously) or vancomycin (2 g every 12 hours intravenously) should be considered initially. Progression to acute respiratory distress syndrome is a serious occurrence and complicated to manage. Treatment includes prone positioning to improve oxygenation, tidal volume reduction with increasing positive end-expiratory pressure, and fluid restriction. None of these measures has been tested

2616

Diseases of the Peripheral Nervous System

Table 116–5. Common Systemic Complications in an Intensive Care Unit Disorder

Risk Factor

Nosocomial ventilatorassociated pneumonia

Age ⬎ 70 yr Mechanical ventilation Use of histamine blockers Prior lung disease Reintubation Paraplegia Prior varices Absent prophylaxis for DVT Indwelling catheter Female sex Length of stay Anticoagulation Prior alcohol abuse, prior NSAID or aspirin use Mechanical ventilation Enteral nutrition Narcotic agents Antibiotic agents Antacid use Fluid overload Diuretic administration Prolonged insufficient nutrition Vomiting, loss via GI tract Critical illness Fat-deficient diet Citrate from plasma exchange

Pulmonary embolus

Nosocomial urinary tract infection GI tract hemorrhage

Diarrhea

Hyponatremia Hypokalemia Hypocalcemia

DVT ⫽ deep venous thrombosis; GI ⫽ gastrointestinal; NSAID ⫽ nonsteroidal anti-inflammatory drug.

systematically in randomized controlled trials, partly because of the many interacting variables.70 Another common complication is urinary tract infection. It occurs more commonly in women and is related to time in the ICU and duration of catheterization. Bacteriuria occurs in virtually all patients when a catheter is present for 2 to 3 weeks. Infection with Enterococcus species, Candida albicans, and Escherichia coli is frequent, but antibiotic resistance is low.59 Candiduria should be considered if patients have been treated with immunomodulating agents such as corticosteroids or cyclosporine. Aminoglycosides are the preferred drugs (e.g., gentamicin, 2 to 3 mg/kg every 8 hours). Prolonged immobilization with quadriplegia or any state that causes a patient to be bedridden increases venous clotting. Prolonged mechanical ventilation often confines patients to bed, although walking while ventilated may be permitted in some patients and circumstances. A recent study found DVT in almost one of four patients despite prophylactic measures (unfractionated heparin) when they were immobilized because of mechanical ventilation and

received neuromuscular blocking agents.50 Extrapolation of these data to a population of patients with neuromuscular disease is not possible. Moreover, pulmonary emboli are uncommon in patients with GBS. In our study, we found an incidence of 3%.42 Pulmonary emboli may underlie sudden onset of dyspnea and marginal oxygenation. Low-grade fever, tachypnea, and tachycardia are helpful clinical cues, but pulmonary emboli remain difficult to predict clinically. Even when clinical suspicion and probability are high, 30% of patients have normal pulmonary angiograms. Spiral computed tomography (CT) may be more cost-effective and, with a 1-year follow-up, had a 98% negative predictive value (in one study, three deaths occurred within 6 months of the CT study).101 Spiral CT may replace pulmonary angiography after normal findings on screening with ventilation-perfusion scan and ultrasonography of leg veins.81,93 Treatment in most patients involves correction of hypoxemia (with face mask at 40% oxygen), pain control with codeine, and immediate anticoagulation with heparin, aiming at an activated partial thromboplastin time twice the baseline value. DVT can be managed with high-pressure stockings, leg elevation, and rest. Anticoagulation with warfarin for 6 months should follow. Gastrointestinal tract complications are common.22,52,57,65 ICU-associated GI tract hemorrhage, mostly associated with the “stress” of mechanical ventilation, is evident by either coffee-grounds aspirate or dark, tarry stools. Prior alcoholic gastritis and use of aspirin and nonsteroidal antiinflammatory drugs predispose patients to GI tract hemorrhage in the ICU. Diarrhea is common in patients receiving enteral nutrition, antibiotics (cephalosporins and clindamycin), and antacids and may occur as fecal impaction resolves after use of narcotic agents. Stool analysis is advisable, but a change in the dilution of the enteral solution and infusion rate may suffice. Paralytic ileus may occur in any immobilized patient and is usually mild. Perforation of the distended colon is life threatening but uncommon. Excruciating pain and profound abdominal distention are leading indicators. Every patient in the ICU develops electrolyte abnormalities.25,48,115 Body stores of electrolytes are easily depleted because of insufficient supplementation or infection. Iatrogenic fluid loading remains the most common cause of hyponatremia and may be related to specific therapy such as plasma exchange or be caused by any critical illness that disturbs water homeostasis.39,75,112 We also noted the occurrence of pseudohyponatremia because of the large load of protein with intravenous immune globulin infusion.22,63 Hypocalcemia may be caused by citrate use with plasma exchange and can be managed with prophylactic calcium administration. Hypomagnesemia may be an epiphenomenon of any critical illness.48 Common complications and risk factors are summarized in Table 116–5.

Management of Patients in the Intensive Care Unit

WITHDRAWAL OF RESPIRATORY SUPPORT AND PALLIATION Discontinuation of mechanical ventilation as a palliative effort in a patient with ALS is a complex issue.11 In many countries, specific legal considerations are involved. It requires careful thought and a supportive environment, including nursing staff and clergy when appropriate.14,56 Morphine, propofol, or midazolam is administered to support discontinuation of the ventilator before the patient lapses into a hypercarbic coma. In a recent study that suggested the use of midazolam before ventilator withdrawal, a bolus dose of 2 to 4 mg of midazolam was followed by 5 to 10 mg of morphine before withdrawal.10 The protocol recommends continuous morphine infusion and 50% of the bolus dose of morphine each hour if distress occurs during weaning, with increased infusion as needed.

CONCLUSION A range of practical problems occur during ICU treatment of patients with acute neuromuscular disorders. Certain disorders, particularly GBS, may resolve completely, and failure to appreciate their difference from other medical illnesses may lead to overaggressive care (e.g., tracheostomy placement). Many care issues pertain to patient immobilization, the consequences of mechanical ventilation, and treatment of intercurrent infections. The major concern in patients with acute neuromuscular disorders is the management of respiratory failure. Clinical recognition of its time course, need for tracheostomy, and methods of mechanical ventilation require expert opinion. Dysautonomia may become challenging and life threatening and is more common than previously known. Future research should focus on timing of intubation, time on the ventilator, prophylactic use of antibiotics in patients at high risk of ventilator-associated pneumonia, and pulmonary rehabilitation.

REFERENCES 1. Abou-Shala, N., and Meduri, U.: Noninvasive mechanical ventilation in patients with acute respiratory failure. Crit. Care Med. 24:705, 1996. 2. Aldrich, T. K., Karpel, J. P., Uhrlass, R. M., et al.: Weaning from mechanical ventilation: adjunctive use of inspiratory muscle resistive training. Crit. Care Med. 17:143, 1989. 3. Allman, R. M., Laprade, C. A., Noel, L. B., et al.: Pressure sores among hospitalized patients. Ann. Intern. Med. 105:337, 1986. 4. Atabai, K., and Matthay, M. A.: The pulmonary physician in critical care. 5. Acute lung injury and the acute respiratory distress syndrome: definitions and epidemiology. Thorax 57:452, 2002.

2617

5. Bach, J. R., and Alba, A. S.: Management of chronic alveolar hypoventilation by nasal ventilation. Chest 97:52, 1990. 6. Bell, C. M., and Redelmeier, D. A.: Mortality among patients admitted to hospitals on weekends as compared with weekdays. N. Engl. J. Med. 345:663, 2001. 7. Bernard, G. R., Artigas, A., Brigham, K. L., et al.: The American-European Consensus Conference on ARDS: definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am. J. Respir. Crit. Care Med. 149:818, 1994. 8. Bick, R. L., Jakway, J., and Baker, W. F. Jr.: Deep vein thrombosis: prevalence of etiologic factors and results of management in 100 consecutive patients. Semin. Thromb. Hemost. 18:267, 1992. 9. Boitano, L. J., Jordan, T., and Benditt, J. O.: Noninvasive ventilation allows gastrostomy tube placement in patients with advanced ALS. Neurology 56:413, 2001. 10. Borasio, G. D., and Voltz, R.: Discontinuation of mechanical ventilation in patients with amyotrophic lateral sclerosis. J. Neurol. 245:717, 1998. 11. Borasio, G. D., Voltz, R., and Miller, R. G.: Palliative care in amyotrophic lateral sclerosis. Neurol. Clin. 19:829, 2001. 12. Borel, C. O., and Guy, J.: Ventilatory management in critical neurologic illness. Neurol. Clin. 13:627, 1995. 13. Bradley, M. D., Orrell, R. W., Clarke, J., et al.: Outcome of ventilatory support for acute respiratory failure in motor neurone disease. J. Neurol. Neurosurg. Psychiatry 72:752, 2002. 14. Brody, H., Campbell, M. L., Faber-Langendoen, K., and Ogle, K. S.: Withdrawing intensive life-sustaining treatment: recommendations for compassionate clinical management. N. Engl. J. Med. 336:652, 1997. 15. Brooks, B. R.: Natural history of ALS: symptoms, strength, pulmonary function, and disability. Neurology 47:S71, 1996. 16. Burns, T. M., Lawn, N. D., Low, P. A., et al.: Adynamic ileus in severe Guillain-Barré syndrome. Muscle Nerve 24:963, 2001. 17. Calleja, M. A.: Autonomic dysfunction and Guillain-Barré syndrome: the use of esmolol in its management. Anaesthesia 45:736, 1990. 18. Camilleri, M.: Autonomic regulation of gastrointestinal motility. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management, 2nd ed. Philadelphia, Lippincott–Raven, p. 135, 1997. 19. Chevrolet, J. C., and Deleamont, P.: Repeated vital capacity measurements as predictive parameters for mechanical ventilation need and weaning success in the Guillain-Barré syndrome. Am. Rev. Respir. Dis. 144:814, 1991. 20. Clagett, G. P., and Reisch, J. S.: Prevention of venous thromboembolism in general surgical patients: results of meta-analysis. Ann. Surg. 208:227, 1988. 21. Collins, R., Scrimgeour, A., Yusuf, S., and Peto, R.: Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin: overview of results of randomized trials in general, orthopedic, and urologic surgery. N. Engl. J. Med. 318:1162, 1988. 22. Cook, D. J., Griffith, L. E., Walter, S. D., et al.: The attributable mortality and length of intensive care unit stay of clinically important gastrointestinal bleeding in critically ill patients. Crit. Care 5:368, 2001.

2618

Diseases of the Peripheral Nervous System

23. Cornuz, J., Pearson, S. D., and Polak, J. F.: Deep venous thrombosis: complete lower extremity venous US evaluation in patients without known risk factors: outcome study. Radiology 211:637, 1999. 24. Derenne, J. P., Macklem, P. T., and Roussos, C.: The respiratory muscles: mechanics, control, and pathophysiology. Part III. Am. Rev. Respir. Dis. 118:581, 1978. 25. DeVita, M. V., Gardenswartz, M. H., Konecky, A., and Zabetakis, P. M.: Incidence and etiology of hyponatremia in an intensive care unit. Clin. Nephrol. 34:163, 1990. 26. Drinker, P. A., and McKhann, C. F. III: The iron lung: first practical means of respiratory support. JAMA 255:1476, 1986. 27. Elliott, M. W., Steven, M. H., Phillips, G. D., and Branthwaite, M. A.: Non-invasive mechanical ventilation for acute respiratory failure. BMJ 300:358, 1990. 28. Flachenecker, P., Hartung, H. P., and Reiners, K.: Power spectrum analysis of heart rate variability in GuillainBarré syndrome: a longitudinal study. Brain 120:1885, 1997. 29. Flachenecker, P., Lem, K., Mullges, W., and Reiners, K.: Detection of serious bradyarrhythmias in Guillain-Barré syndrome: sensitivity and specificity of the 24-hour heart rate power spectrum. Clin. Auton. Res. 10:185, 2000. 30. Flachenecker, P., Mullges, W., Wermuth, P., et al.: Eyeball pressure testing in the evaluation of serious bradyarrhythmias in Guillain-Barré syndrome. Neurology 47:102, 1996. 31. Fletcher, D. D., Lawn, N. D., Wolter, T. D., and Wijdicks, E. F.: Long-term outcome in patients with Guillain-Barré syndrome requiring mechanical ventilation. Neurology 54:2311, 2000. 32. Forbes, K., and Stevenson, A. J.: The use of power Doppler ultrasound in the diagnosis of isolated deep venous thrombosis of the calf. Clin. Radiol. 53:752, 1998. 33. Freezer, N. J., and Robertson, C. F.: Tracheostomy in children with Guillain-Barré syndrome. Crit. Care Med. 18:1236, 1990. 34. Fulgham, J. R., and Wijdicks, E. F.: Guillain-Barré syndrome. Crit. Care Clin. 13:1, 1997. 35. Gay, P. C., and Edmonds, L. C.: Severe hypercapnia after low-flow oxygen therapy in patients with neuromuscular disease and diaphragmatic dysfunction. Mayo Clin. Proc. 70:327, 1995. 36. Gelinas, D. F., O’Connor, P., and Miller, R. G.: Quality of life for ventilator-dependent ALS patients and their caregivers. J. Neurol. Sci. 160(Suppl. 1):S134, 1998. 37. Gregory, S., Siderowf, A., Golaszewski, A. L., and McCluskey, L.: Gastrostomy insertion in ALS patients with low vital capacity: respiratory support and survival. Neurology 58:485, 2002. 38. Griggs, R. C., Donohoe, K. M., Utell, M. J., et al.: Evaluation of pulmonary function in neuromuscular disease. Arch. Neurol. 38:9, 1981. 39. Guglielminotti, J., Pernet, P., Maury, E., et al.: Osmolar gap hyponatremia in critically ill patients: evidence for the sick cell syndrome? Crit. Care Med. 30:1051, 2002. 40. Hahn, A. F.: The challenge of respiratory dysfunction in Guillain-Barré syndrome. Arch. Neurol. 58:871, 2001. 41. Heffner, J. E.: Timing tracheotomy: calendar watching or individualization of care? Chest 114:361, 1998.

42. Henderson, R. D., Lawn, N. D., Fletcher, D. D., and Wijdicks, E. F.: The morbidity of Guillain-Barré syndrome admitted to the intensive care unit. Neurology 60:17, 2003. 43. Hewer, R. L., Hilton, P. J., Smith, A. C., and Spalding, J. M.: Acute polyneuritis requiring artificial respiration. Q. J. Med. 37:479, 1968. 44. Hillberg, R. E., and Johnson, D. C.: Noninvasive ventilation. N. Engl. J. Med. 337:1746, 1997. 45. Hughes, R. A., Wijdicks, E. F., Barohn, R., et al.: Practice parameter: Immunotherapy for Guillain-Barre syndrome: Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 61:736, 2003. 46. Hughes, R. A., Wijdicks, E. F., Barohn, R., et al.: Supportive therapy in Guillain-Barre syndrome. Arch. Neurol. 2004 (in press). 47. Hughes, R. A., and Bihari, D.: Acute neuromuscular respiratory paralysis. J. Neurol. Neurosurg. Psychiatry 56:334, 1993. 48. Huijgen, H. J., Soesan, M., Sanders, R., et al.: Magnesium levels in critically ill patients: what should we measure? Am. J. Clin. Pathol. 114:688, 2000. 49. Hund, E. F., Borel, C. O., Cornblath, D. R., et al.: Intensive management and treatment of severe Guillain-Barré syndrome. Crit. Care Med. 21:433, 1993. 50. Ibrahim, E. H., Iregui, M., Prentice, D., et al.: Deep vein thrombosis during prolonged mechanical ventilation despite prophylaxis. Crit. Care Med. 30:771, 2002. 51. Ibsen, B.: From anaesthesia to anaesthesiology: personal experiences in Copenhagen during the past 25 years. Acta Anaesthesiol. Scand. Suppl. 61:1, 1975. 52. Inayet, N., Amoateng-Adjepong, Y., Upadya, A., and Manthous, C. A.: Risks for developing critical illness with GI hemorrhage. Chest 118:473, 2000. 53. Inman, K. J., Dymock, K., Fysh, N., et al.: Pressure ulcer prevention: a randomized controlled trial of 2 risk-directed strategies for patients surface assignment. Adv. Wound Care 12:72, 1999. 54. Kazmers, A., Groehn, H., and Meeker, C.: Acute calf vein thrombosis: outcomes and implications. Am. Surg. 65:1124, 1999. 55. Kelly, B. J., and Luce, J. M.: The diagnosis and management of neuromuscular diseases causing respiratory failure. Chest 99:1485, 1991. 56. Kirchhoff, K. T.: Promoting a peaceful death in the ICU. Crit. Care Nurs. Clin. North Am. 14:201, 2002. 57. Kupfer, Y., Cappell, M. S., and Tessler, S.: Acute gastrointestinal bleeding in the intensive care unit: the intensivist’s perspective. Gastroenterol. Clin. North Am. 29:275, 2000. 58. Lassen, H. C. A.: A preliminary report on the 1952 epidemic of poliomyelitis in Copenhagen with special reference to the treatment of acute respiratory insufficiency. Lancet 1:37, 1953. 59. Laupland, K. B., Zygun, D. A., Davies, H. D., et al.: Incidence and risk factors for acquiring nosocomial urinary tract infection in the critically ill. J. Crit. Care 17:50, 2002. 60. Lawn, N. D., Fletcher, D. D., Henderson, R. D., et al.: Anticipating mechanical ventilation in Guillain-Barré syndrome. Arch. Neurol. 58:893, 2001. 61. Lawn, N. D., and Wijdicks, E. F.: Fatal Guillain-Barré syndrome. Neurology 52:635, 1999.

Management of Patients in the Intensive Care Unit 62. Lawn, N. D., and Wijdicks, E. F.: Tracheostomy in GuillainBarré syndrome. Muscle Nerve 22:1058, 1999. 63. Lawn, N. D., Wijdicks, E. F., and Burritt, M. F.: Intravenous immune globulin and pseudohyponatremia [letter]. N. Engl. J. Med. 339:632, 1998. 64. Legere, B. M., Dweik, R. A., and Arroliga, A. C.: Venous thromboembolism in the intensive care unit. Clin. Chest Med. 20:367, 1999. 65. Liolios, A., Oropello, J. M., and Benjamin, E.: Gastrointestinal complications in the intensive care unit. Clin. Chest Med. 20:329, 1999. 66. Lyall, R. A., Donaldson, N., Polkey, M. I., et al.: Respiratory muscle strength and ventilatory failure in amyotrophic lateral sclerosis. Brain 124:2000, 2001. 67. Lyu, R. K., Tang, L. M., Hsu, W. C., et al.: A longitudinal cardiovascular autonomic function study in mild GuillainBarré syndrome. Eur. Neurol. 47:79, 2002. 68. Macklem, P. T.: Respiratory muscle dysfunction. Hosp. Pract. (Off. Ed.) 21:83, 1986. 69. Macklem, P. T.: The mechanics of breathing. Am. J. Respir. Crit. Care Med. 157:S88, 1998. 70. Macnaughton, P. D., and Evans, T. W.: Management of adult respiratory distress syndrome. Lancet 339:469, 1992. 71. Malthaner, R. A., Telang, H., Miller, J. D., et al.: Percutaneous tracheostomy: is it really better? Chest 114:1771, 1998. 72. Marsh, H. M., Gillespie, D. J., and Baumgartner, A. E.: Timing of tracheostomy in the critically ill patient. Chest 96:190, 1989. 73. Mazzini, L., Corra, T., Zaccala, M., et al.: Percutaneous endoscopic gastrostomy and enteral nutrition in amyotrophic lateral sclerosis. J. Neurol. 242:695, 1995. 74. Meissner, M. H., Caps, M. T., Bergelin, R. O., et al.: Early outcome after isolated calf vein thrombosis. J. Vasc. Surg. 26:749, 1997. 75. Milionis, H. J., Liamis, G. L., and Elisaf, M. S.: The hyponatremic patient: a systematic approach to laboratory diagnosis. CMAJ 166:1056, 2002. 76. Miller, R. G., Rosenberg, J. A., Gelinas, D. F., et al.: Practice parameter: the care of the patient with amyotrophic lateral sclerosis (an evidence-based review). Report of the Quality Standards Subcommittee of the American Academy of Neurology: ALS Practice Parameters Task Force. Neurology 52:1311, 1999. 77. Newton-John, H.: Prevention of pulmonary complications in severe Guillain-Barré syndrome by early assisted ventilation. Med. J. Aust. 142:444, 1985. 78. Ng, K. K., Howard, R. S., Fish, D. R., et al.: Management and outcome of severe Guillain-Barré syndrome. Q. J. Med. 88:243, 1995. 79. Nuckton, T. J., Alonso, J. A., Kallet, R. H., et al.: Pulmonary dead-space fraction as a risk factor for death in the acute respiratory distress syndrome. N. Engl. J. Med. 346:1281, 2002. 80. Parry, G. J.: Guillain-Barré Syndrome. New York, Thieme Medical Publishers, 1993. 81. Paterson, D. I., and Schwartzman, K.: Strategies incorporating spiral CT for the diagnosis of acute pulmonary embolism: a cost-effectiveness analysis. Chest 119:1791, 2001. 82. Pfeiffer, G., Schiller, B., Kruse, J., and Netzer, J.: Indicators of dysautonomia in severe Guillain-Barré syndrome. J. Neurol. 246:1015, 1999.

2619

83. Polkey, M. I., Green, M., and Moxham, J.: Measurement of respiratory muscle strength. Thorax 50:1131, 1995. 84. Polkey, M. I., Lyall, R. A., Green, M., et al.: Expiratory muscle function in amyotrophic lateral sclerosis. Am. J. Respir. Crit. Care Med. 158:734, 1998. 85. Polkey, M. I., Lyall, R. A., Moxham, J., and Leigh, P. N.: Respiratory aspects of neurological disease. J. Neurol. Neurosurg. Psychiatry 66:5, 1999. 86. Pontoppidan, H., Geffin, B., and Lowenstein, E.: Acute respiratory failure in the adult. 2. N. Engl. J. Med. 287:743, 1972. 87. Pugin, J.: Clinical signs and scores for the diagnosis of ventilator-associated pneumonia. Minerva Anestesiol. 68:261, 2002. 88. Rantala, H., Uhari, M., Cherry, J. D., and Shields, W. D.: Risk factors of respiratory failure in children with GuillainBarré syndrome. Pediatr. Neurol. 13:289, 1995. 89. Ropper, A. H., and Kehne, S. M.: Guillain-Barré syndrome: management of respiratory failure. Neurology 35:1662, 1985. 90. Ropper, A. H., Wijdicks, E. F. M., and Truax, B. T.: GuillainBarré Syndrome. Philadelphia, F. A. Davis, 1991. 91. Sakakibara, R., Hattori, T., Kuwabara, S., et al.: Micturitional disturbance in patients with Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 63:649, 1997. 92. Samama, M. M., Cohen, A. T., Darmon, J. Y., et al.: A comparison of enoxaparin with placebo for the prevention of venous thromboembolism in acutely ill medical patients. Prophylaxis in Medical Patients with Enoxaparin Study Group. N. Engl. J. Med. 341:793, 1999. 93. Schoepf, U. J., Holzknecht, N., Helmberger, T. K., et al.: Subsegmental pulmonary emboli: improved detection with thin-collimation multi-detector row spiral CT. Radiology 222:483, 2002. 94. Smith, S. A., and Smith, S. E.: Heart rate variations in the Guillain-Barré syndrome [letter]. Br. Med. J. 281:1009, 1980. 95. Suh, R. H., Margulies, D. R., Hopp, M. L., et al.: Percutaneous dilatational tracheostomy: still a surgical procedure. Am. Surg. 65:982, 1999. 96. Sunderrajan, E. V., and Davenport, J.: The Guillain-Barré syndrome: pulmonary-neurologic correlations. Medicine (Baltimore) 64:333, 1985. 97. Tamura, D. Y., Moore, E. E., Partrick, D. A., et al.: Acute hypoxemia in humans enhances the neutrophil inflammatory response. Shock 17:269, 2002. 98. Teitelbaum, J. S., and Borel, C. O.: Respiratory dysfunction in Guillain-Barré syndrome. Clin. Chest Med. 15:705, 1994. 99. Theaker, C., Mannan, M., Ives, N., and Soni, N.: Risk factors for pressure sores in the critically ill. Anaesthesia 55:221, 2000. 100. Thromboembolic Risk Factors (THRIFT) Consensus Group: Risk of and prophylaxis for venous thromboembolism in hospital patients. BMJ 305:567, 1992. 101. Tillie-Leblond, I., Mastora, I., Radenne, F., et al.: Risk of pulmonary embolism after a negative spiral CT angiogram in patients with pulmonary disease: 1-year clinical follow-up study. Radiology 223:461, 2002. 102. Tobin, M. J.: Advances in mechanical ventilation. N. Engl. J. Med. 344:1986, 2001. 103. Tuck, R. R., and McLeod, J. G.: Autonomic dysfunction in Guillain-Barré syndrome. J. Neurol. Neurosurg. Psychiatry 44:983, 1981.

2620

Diseases of the Peripheral Nervous System

104. Tzeng, A. C., and Bach, J. R.: Prevention of pulmonary morbidity for patients with neuromuscular disease. Chest 118:1390, 2000. 105. Varrato, J., Siderowf, A., Damiano, P., et al.: Postural change of forced vital capacity predicts some respiratory symptoms in ALS. Neurology 57:357, 2001. 106. Wade, W. E., and Chisholm, M. A.: Venous thrombosis after acute spinal cord injury: cost analysis of prophylaxis guidelines. Am. J. Phys. Med. Rehabil. 79:504, 2000. 107. Weingarden, S. I.: Deep venous thrombosis in spinal cord injury: overview of the problem. Chest 102:636S, 1992. 108. Wijdicks, E. F.: Short of breath, short of air, short of mechanics. Pract. Neurol. 2:208, 2002. 109. Wijdicks, E. F., Lawn, N. D., and Fletcher, D. D.: Tracheostomy scars in Guillain-Barré syndrome: a reason for concern? [letter]. J. Neurol. 248:527, 2001. 110. Wijdicks, E. F. M., Henderson, R. D., and McClelland, R. I.: Emergency intubation for respiratory failure for GuillainBarré syndrome. Arch. Neurol. 60:947, 2003.

111. Winer, J. B., Hughes, R. A., and Osmond, C.: A prospective study of acute idiopathic neuropathy, I: clinical features and their prognostic value. J. Neurol. Neurosurg. Psychiatry 51:605, 1988. 112. Wong, L. L., and Verbalis, J. G.: Systemic diseases associated with disorders of water homeostasis. Endocrinol. Metab. Clin. North Am. 31:121, 2002. 113. Young, J. D., and Sykes, M. K.: Assisted ventilation, I: artificial ventilation: history, equipment and techniques. Thorax 45:753, 1990. 114. Zhang, H., Slutsky, A. S., and Vincent, J. L.: Oxygen free radicals in ARDS, septic shock and organ dysfunction. Intensive Care Med. 26:474, 2000. 115. Zivin, J. R., Gooley, T., Zager, R. A., and Ryan, M. J.: Hypocalcemia: a pervasive metabolic abnormality in the critically ill. Am. J. Kidney Dis. 37:689, 2001. 116. Zochodne, D. W.: Autonomic involvement in Guillain-Barré syndrome: a review. Muscle Nerve 17:1145, 1994.

117 Rehabilitation Management of Neuropathies MARY L. DOMBOVY

General Overview and Principles of Rehabilitation Rehabilitation Interventions and Techniques Therapeutic Exercise Orthoses Electrical Stimulation Adaptive Equipment, Assistive Devices, Wheelchairs, and Environmental Modifications

Specific Complications and Their Rehabilitation Management Weakness Sensory Loss Autonomic Dysfunction and Pain Neurogenic Bladder and Bowel Bowel Incontinence Sexual Dysfunction The Consequences of Immobility and Deconditioning

GENERAL OVERVIEW AND PRINCIPLES OF REHABILITATION Neuropathy as a category includes a broad spectrum of disorders with multiple causes. The pathologic involvement can range from a single nerve or root to multiple nerves and deficits from weakness, sensory loss, pain, and autonomic dysfunction in varying combinations. These deficits combine to lead to functional loss, risk of injury, and disability. The goal of rehabilitation is to restore function and to reduce the impact that the remaining impairments have on a person’s activities of daily living (ADLs). Individualized rehabilitation programs include modification of risk factors, lifestyle changes, and education on health awareness and fitness, as well as an approach that combines restorative and adaptive techniques. Restorative approaches attempt to influence the course of recovery and regain lost function. This might include therapeutic exercise, the application of electrical stimulation, or first having the patient complete a motion with gravity eliminated or with assistance from the therapist. Adaptive techniques include learning a new way of performing a task, the use of a brace to support weak muscle or provide proper positioning, and special equipment to enable the person to perform a task despite persistent deficits. The application of modalities such as transcutaneous electrical nerve stimulation (TENS) or ultrasound in combination

Specific Disorders Generalized Sensorimotor Peripheral Neuropathy Acute and Subacute Polyradiculoneuropathy Brachial Plexus Injuries and Disease Postpolio Syndrome Motor Neuron Disease Radiculopathy

with medication, stretching, and range-of-motion (ROM) exercises can be useful in relieving pain and increasing function. In complicated or complex illness involving multiple extremities and/or multiple organ systems, or an illness that creates disability in the performance of basic ADLs, a coordinated team approach is essential to achieving optimal outcomes. Physical therapists generally address mobility skills and gross motor function (strength, balance, coordination, posture, and body mechanics). They may use various modalities (e.g., heat, ultrasound, electrical stimulation) to alleviate pain and spasm and to promote muscle contraction. Occupational therapists focus on improving ADL skills (e.g., eating, dressing, bathing, toileting), as well as improving fine motor function and community living skills. Speech therapists may be needed when there is bulbar involvement to address swallowing and speech production and assess the need for augmentative communication devices. Case managers/social workers help the patient and family access and coordinate services and resources. Other rehabilitation specialists and vocational specialists or education professionals are added to the team as needed. The role that the physician plays may vary depending on the needs of the patient and the background and training of the physician. Physicians trained in the classic “medical model” tend to focus on diagnosis and treatment of symptoms such as sensory loss and weakness. However, 2621

2622

Diseases of the Peripheral Nervous System

the patients are more concerned about their ability to dress, walk, drive, and participate in their usual role within the family and in society. Because neuropathy is often a chronic or progressive process, this can create a “disconnect” between the physician and the patient. Physicians are encouraged to inquire about the patient’s and family’s concerns and goals, as well as to communicate with other members of the rehabilitation team regarding the patient’s progress. Chronic diseases are best managed with a consistent, continuing, and coordinated approach that focuses on achievement of realistic behaviors and outcomes that are important to patient and family.

REHABILITATION INTERVENTIONS AND TECHNIQUES Therapeutic Exercise Exercise is one of the most frequently prescribed modalities in rehabilitation. Understanding the physiology of exercise and its effects on physical performance is important to designing and implementing an appropriate program. Just as one would not write a prescription for a medication without giving the dose and frequency of administration, one should not tell a patient to “exercise” without delineating the type, intensity, duration, and frequency. Although in many situations the details of the exercise program are left to the physiatrist or physical therapist, a basic understanding of exercise physiology will help the clinician evaluate the program’s effectiveness in improving the patient’s functional status. A detailed review is beyond the scope of this chapter, and the reader is referred to Frontera25 for a concise review, and to McArdle et al.47 for more detail. In general, exercise in rehabilitation is aimed at improving one or more of the following parameters: (1) strength, (2) endurance, (3) coordination and balance, (4) speed and quickness, and (5) flexibility and range of motion. To be effective, strength training programs must be progressive in nature.15,16 That is, the intensity (training load) must be increased as strength increases to promote continued gain. In addition, the exercise must be integral to or approximate the tasks the person wishes to be able to perform.14,39 Because eccentric (lengthening) muscle contractions are as common in everyday tasks and mobility as concentric (shortening) muscle contractions, their inclusion is important to optimize the effects of training.36 Intermediate programs with medium resistance to allow for the completion of 15 to 30 repetitions will train simultaneously for strength and endurance. Endurance training must activate large muscle groups for relatively long periods of time (15 to 60 minutes) from three to five times per week. Although the recommended intensity for a training effect in the general population is from 40% to 80% of the maximum heart rate, in patient populations the intensity may initially be

lower. Depending on the situation, walking, aquatic exercise, a stationary bicycle, a treadmill, a rowing machine, and an arm ergometer may all have their applications. Physiologic adaptations are both central (cardiovascular) and peripheral (muscle oxidative capacity), and to some extent specific to the activity.42 Coordination, balance, and the acquisition of complex motor skills depend on repetitive practice of the correct pattern of movements.43 Speed and quickness are developed by exercises that require rapid changes in movement direction in response to an external stimulus. Flexibility is the ability to move a joint throughout a full ROM, and is important in the prevention of injuries and in the performance of most daily activities.37,66 Flexibility is frequently impaired in neuromuscular disease, and must be improved along with strength, endurance, and coordination to improve function. Various stretching regimens, passive and active ROM, and proprioceptive neuromuscular facilitation are among the approaches used to maintain and improve flexibility.63 Neuromuscular lesions create weakness and immobility that limit the stretching normally encountered by the muscle and connective tissue. Contracture ensues, leading to altered biomechanics and increased effort required for movement, further limiting activity. When inactivity limits muscle contraction to less than 20% of maximal tension, disuse atrophy develops, further exacerbating the weakness. In any neuromuscular disease or injury, preventative programs should be initiated in the acute phase to maintain ROM and prevent further weakness when possible. Preventative measures include proper positioning, splinting, ROM exercises, strengthening and mobility exercises for non- or minimally involved muscles, and frequent inspection and protection of sensory-impaired areas. Research has demonstrated the beneficial effects of exercise in neuromuscular disease, including low-intensity strength training and aerobic training.20,34 However, one must be careful, because overwork damage may occur if the exercise is too near the muscle’s maximal attainable tension. In this situation, overwork may lead to further muscle weakness in both muscles with “normal” physiology and muscles impaired by neuromuscular disease.33 Weakened muscle is more susceptible to overwork damage,18,33 leading to a narrow therapeutic window that requires close monitoring. When neuromuscular disease is severe, basic ADLs may overtax the muscles, and activities should be limited until fair to good muscle power is achieved without complaints of fatigue that last into the next day.34 Resistance training should not be attempted until the muscle has at least antigravity strength through the full ROM. In peripheral nerve lesions and injuries, there is no evidence to suggest that therapeutic exercise expedites the processes of reinnervation or remyelination.34 Exercise maintains the body, muscles, and joints in an optimal state and enhances the patient’s functional recovery by

2623

Rehabilitation Management of Neuropathies

maximizing the strength and endurance of the remaining functional motor units and muscle fibers.

Orthoses Orthotic devices have a number of applications in the management of neuropathies. Orthotic devices are generally classified by the joints they cross. For example, a WHO is a wrist-hand orthosis, a KAFO is a knee-ankle-foot orthosis, and a TLSO is a thoracic-lumbar-sacral orthosis. Many splints also bear the name of the person who designed them; they may also have names that say something about their function (e.g., a WHO tenodesis splint) or names that reflect their appearance. Orthotic devices generally serve one of three purposes: 1. Protection or immobilization: These orthoses are generally used for stabilization of bony fragments; after tendon or ligament injury; after surgical repair of tendons, joints, and the like; and in situations in which stabilization is required for joints supplied by significantly weakened muscles to allow for performance of activities (e.g., an AFO in footdrop to allow for safe ambulation). 2. Correction: Both dynamic and static splints can apply forces to joints that over time prevent and even correct deformities and subluxations. 3. Functional assistance: Orthoses, through biomechanical principles, can assist function by compensating for weak muscles or deformities. Examples of functional orthoses include the deltoid aid, which provides assistance for proximal arm weakness, and the balanced forearm orthosis, which allows trace movements at the shoulder to control arm movements. Many AFOs fall into this category because they can be set to provide varying degrees of dorsiflexion assist, plantar flexion stop, and mediolateral control, all aimed at facilitating an optimal gait pattern for the patient. Health care professionals involved in the prescription, fabrication, and application of orthoses need a thorough understanding of biomechanics, anatomy, and the response of tissues to healing. They also need creativity and technical skills to allow fabrication of devices that meet specific indications, and win patient acceptance. Often this requires a collaborative effort between physician, orthotist, and either the occupational or physical therapist. Orthoses are frequently used in nerve injuries and peripheral neuropathies to provide both support and functional assistance. For example, in radial nerve injuries distal to the spiral groove of the humerus, the most common presenting problem is wristdrop and finger drop, and an orthosis can be designed that will hold the wrist in a functional position of 30 degrees’ dorsiflexion and provide a spring assist for finger extension. Severe peripheral neuropathies or a peroneal nerve injury can result in footdrop that can be assisted by an AFO. An AFO can also be

designed that can compensate for quadriceps and ankle weakness by placing the foot in slight plantar flexion. This places the ground contact forces behind the knee, facilitating knee extension. It is important to note that the design of an orthosis used for someone with a peripheral nerve injury and low tone will be very different than an orthosis designed to compensate for the deforming forces of spasticity. Acceptance by the patient is crucial, because an orthotic device that is not worn is useless. The orthosis must be in line with the patient’s goals, and he or she must be able to appreciate how it will benefit function. Involvement of the patient in the design and appearance, assessment of skin tolerance and adjustment for comfort, and, when possible, an acceptable schedule for wearing the splint will all facilitate acceptance. Orthotic prescriptions should include the diagnosis or problem to be addressed and the function or motion to be achieved. The reader is referred to McKee and Morgan,48 Patel et al.,55 and Hennessey and Johnson32 for further detail.

Electrical Stimulation Although early in medical history the application of a variety of electricity-generating devices was advocated for use in the treatment of several different medical conditions, its association with questionable practices limited general acceptance by the medical field. Both research and clinical experience have helped to identify the benefits of electrical stimulation and facilitated its acceptance and use by many specialties, including rehabilitation, neurology, urology, gynecology, orthopedics, pain management, and dermatology. The use of functional electrical stimulation (FES) dates back to the early 1960s, when Liberson et al.45 applied electrical stimulation to the peroneal nerve of hemiplegic patients to produce dorsiflexion and reports were published of the use of FES to facilitate standing in patients with spinal cord injury.41 Melzack and Wall49 described the gate theory of pain modulation in 1965. This theory has provided the basis for the development of TENS as well as for the use of implanted dorsal column and conus medullaris root stimulators for neuropathic pain and neurogenic bladder control. Neuromuscular Electrical Stimulation for Muscle Strengthening Although electrical stimulation has been shown to be effective in increasing the force generated by a muscle,9 there is no evidence that it is any more effective in improving muscle strength beyond what can be achieved through voluntary isokinetic or isometric exercise programs carried out at maximal intensity.12 The same is true regarding endurance training.67 However, because neuromuscular electrical stimulation (NMES) preferentially activates type II fibers, which are important to maximal force generation,

2624

Diseases of the Peripheral Nervous System

NMES leads to greater overall increase in strength of the muscle at submaximal training intensity.13 In addition, NMES was able to produce changes in muscle strength without the accompanying increase in heart rate and blood pressure caused by isometric strength training.10 Thus NMES strengthening may offer some advantages over voluntary contraction training programs in those patients with significant cardiovascular disease or other conditions that limit voluntary contraction. NMES is useful in preventing disuse atrophy associated with immobilization, such as might occur following knee ligament surgery or fractures. In addition to improving strength, NMES has been shown to improve performance when applied to the quadriceps muscle in activities such as sprinting and vertical jump.71 Thus NMES may allow a quicker return to function following injury, which may be beneficial to those whose activities and/or work require a rapid return to peak physical performance. The benefits of the use of NMES in denervated muscle are controversial. Its ability to maintain muscle bulk and enzymatic function is questionable.7 Stimulation parameters that have been shown to be the most effective in maintaining muscle bulk in denervated muscle in experimental animals are short duration and high voltage or long duration and low voltage, both of which are uncomfortable and often impractical.34 There is no evidence that NMES speeds the process of reinnervation,65 with some studies suggesting that stimulation actually inhibits terminal sprouting and reinnervation.19 However, animal studies suggest that NMES may be beneficial in improving functional outcome by preventing atrophy in situations in which reinnervation is expected to occur over a relatively long period of time, such as following surgical nerve repair.70 Functional Electrical Stimulation Electrical stimulation of the pelvic floor muscles is a useful adjunct to the conservative management of stress/urge urinary incontinence, particularly in those who cannot voluntarily stop urine flow and have poor contraction of the pelvic floor muscles.53 Various applications of electrical stimulation have been tried in patients with spinal cord and peripheral nerve damage with varying degrees of success in restoring continence and varying acceptance by patients. Further research is underway in this area.53 FES has many successful applications in spinal cord injury, including phrenic nerve pacing in ventilatory insufficiency (high cervical injury); electroejaculation, which has led to successful pregnancy; and enabling exercise to maintain cardiovascular fitness. Emerging applications in spinal cord injury include producing functional gait as well as functional use of the upper extremity in quadriplegic patients. FES has also been used in hemiplegia following stroke to improve gait and upper extremity function. Unfortunately, these approaches require an intact peripheral nerve, which does not exist in peripheral neuropathies.

Transcutaneous Electrical Nerve Stimulation TENS is beneficial in the treatment of many types of pain syndromes, but patients who respond best are those with pain of neurogenic and/or musculoskeletal origin and a relatively minor psychological component.23 There is no current research to support the use of one type of TENS (various strengths/durations, burst, modulated, etc.) over another, and patients should be encouraged to adjust the impulse as well as to try various lead placements to obtain optimal relief. In most situations, physical therapists work with patients using TENS. Adverse reactions to TENS are infrequent, the most common being skin hypersensitivity. TENS is contraindicated in patients with demand pacemakers and in those who are pregnant.

Adaptive Equipment, Assistive Devices, Wheelchairs, and Environmental Modifications Adaptive equipment, assistive devices, wheelchairs, and environmental modifications all may play a major role in enhancing a patient’s functional abilities. A basic understanding of the range of devices available is essential to appropriate management of patients with impairing neuropathy. Ambulation Aids Indications for ambulatory aids include weakness of the trunk or lower extremities, poor balance, and decreased ability to bear weight as a result of disease or injury (e.g., an infection or ulcer in a patient with diabetic neuropathy). Assistive devices make walking safer by providing a wider base of support, by redistributing weight to the upper extremities, and by allowing for larger shifts in the center of gravity before balance is lost. Unfortunately, physicians frequently write prescriptions for assistive devices without a prior physical therapy evaluation. In this scenario, fully two thirds of the devices are either inappropriate or misused by the patient. When a properly trained physical therapist is involved, the device is appropriate and correctly used 85% of the time. Selection of a particular device is determined by the patient’s strength, stability, coordination, cardiovascular capacity, and cognitive status. Maximum stability and support are provided by a walker, followed by axillary crutches, forearm crutches, two canes, a quad cane, and a single-tipped cane. Wheelchairs A wheelchair may be needed in cases of polyradiculoneuropathy, motor neuron disease, or postpolio syndrome, and in other severe neuropathies in which the patient has limited cardiovascular capacity such that he or she cannot meet the additional energy cost of ambulating with a cane and a brace. There are over 10,000 models of wheelchairs available. The physician prescribing wheelchairs should be

2625

Rehabilitation Management of Neuropathies

familiar with standard models and options, and should recognize when evaluation by a specialist for a custom chair is needed. For patients who use wheelchairs, the seating system should maximize independence and minimize pressure over vulnerable areas while providing support for proper posture. Patients who use their wheelchairs for prolonged periods of time cannot get by with the standard sling seat and back because the hammock effect of the sling seat results in poor posture and unequal distribution of pressure, which can lead to scoliosis, kyphosis, back pain, and pressure sores. At a minimum, some type of solid seat with a cushion is needed. For anyone who spends more than an hour per day in a wheelchair, regular monitoring of areas of pressure on the skin is a requirement. Many wheelchairs are made in components, so that an appropriately sized frame is chosen initially and the component parts are ordered to meet the specific needs of the patient. Adaptive Equipment and Environmental Modifications Many forms of adaptive equipment are available to facilitate self-care in patients with neurologic disabilities. The most frequently used equipment is briefly reviewed here. Several types of adaptive eating utensils are available, including built-up handles, rocker knives for one-handed use, and utensils with cuff attachments (used when grip is absent). Removable grips and cuffs are also available for a variety of other things, including pens, toothbrushes, and the like. Long-handled reachers and shoehorns, buttonhooks, and special holders for cards and books are also frequently used. Clothing can be adapted with Velcro closures. Bath and shower benches, grab bars, and commodes are commercially available. Creative therapists are able to fashion custom devices to meet the individual needs of their patients. A wide range of adaptive equipment is also available for sports and leisure activities. Simple home modifications include the installation of raised toilet seats, grab bars, handheld shower heads, lever door handles, and stair rails, as well as relocation of commonly used items and rearrangement of furniture. More complex modifications include installing accessible doorways (side-to-side clearance of 30 inches), enlarging a bathroom to allow wheelchair access, and installing an entrance ramp. Sophisticated environmental control systems can be designed to allow those with quadriparesis to function independently much of the day. Various computer interfaces, telephone equipment, and other desk/workplace modifications can allow the significantly impaired individual to return to productive work. The need for and selection of home and workplace modifications are best assessed by a physical and an occupational therapist who understand the patient’s limitations, have knowledge of the vast array of available devices, and can also visit the home and work settings.

SPECIFIC COMPLICATIONS AND THEIR REHABILITATION MANAGEMENT Muscle weakness, sensory loss/abnormal sensation, and autonomic dysfunction of varying degrees and combinations are the deficits found in peripheral neuropathies.

Weakness Muscle weakness in neuropathies leads to inability to perform daily activities, muscle shortening, and contractures. The relative inactivity that follows further compounds the weakness and resultant functional loss. Uninvolved muscles and involved muscles with better than antigravity strength may benefit from closely monitored strengthening programs (see Therapeutic Exercise above). Where practical, general conditioning should also be maintained. ROM and musclestretching exercises carried out once or twice daily can help prevent contractures. Depending on the degree of weakness, these exercises can be passive, active-assisted, or active. Severely weakened or paralyzed muscles may require splinting and proper rest positioning to maintain the joint in a mobile and functional position until recovery occurs. During recovery, or when functional recovery is incomplete, functional orthoses may allow the patient to resume activities.

Sensory Loss Decreased ability to perceive pain and touch compromise the patient’s ability to protect the involved extremity from injury. Injury may lead to difficult-to-heal skin ulcerations and infections. In generalized peripheral neuropathy, the feet are most commonly affected, but the same problems can also occur in the upper extremity affected by brachial plexus injury. When deep pain sensation is lost and kinesthetic and joint perception sense are impaired, neuroarthropathic joints may develop from repetitive trauma. Patients with loss of kinesthetic and joint position sense, for example, cannot judge accurately the placement of their foot and the position of the foot and ankle, not only creating an ataxic gait but increasing the risk of ankle sprain that is repeated over time, leading to permanent joint injury. (The assessment of sensation loss is also described in Chapters 40 and 42.) The basics of treatment for sensory loss begin with education of the patient and a preventative program consisting of a careful daily skin examination and gentle cleaning and soaking techniques, followed by the application of a lubricant to prevent skin fissuring. Removing pressure from red areas and use of appropriate custom shoes or shoe inserts also play a role. Special care should be taken in those wearing orthoses, with several inspections per day as well as periods of time when the orthosis is removed.

2626

Diseases of the Peripheral Nervous System

Autonomic Dysfunction and Pain Autonomic dysfunction can be seen in neuropathies affecting small fibers, diabetic neuropathy, and polyradiculoneuropathy (Guillain-Barré syndrome [GBS]), and is postulated to play a role in neurogenic pain and in regional pain syndromes (e.g., reflex sympathetic dystrophy). Symptoms include decreased sweating, orthostatic hypotension, gastroparesis, constipation, vomiting, diarrhea, impotence, and a flaccid urinary bladder (the last two are covered later in this chapter). Cardiac arrhythmias are commonly seen in GBS. Neurogenic and regional pain syndromes can include dysesthesias; hyperesthesia; atrophic skin, hair, and nail changes; edema; and skin temperature and color changes. (The assessment of autonomic dysfunction is also discussed in Chapter 44.) Patients with orthostatic hypotension can be advised to arise more slowly and may also benefit from compression stockings. If these simple measures do not work, a mineralocorticoid may be prescribed, but patients then run the risk of supine hypertension. Gastric motility disturbances may be improved with diet and a proper bowel regimen and/or the addition of motility-enhancing drugs. Patients in the initial stages of acute GBS require cardiac monitoring. Hyperesthesia may be lessened with desensitization techniques (e.g., rubbing the affected area with a washcloth), or the wearing of a compression garment or glove over the affected area. Narcotics are often minimally effective in controlling neurogenic pain. Tricyclic and other antidepressants, carbamazepine, gabapentin, baclofen (Lioresal), phenytoin, and mexiletine are other choices. Drug selection is often a trial-and-error process. Care should be taken to give a selected medication an adequate trial at a therapeutic dose before concluding that a particular drug is ineffective. Because neurogenic pain is difficult to treat, these patients often end up on multiple drugs, and the physician needs to be aware of potential drug interactions.69 One should not forget that the simple conservative measures mentioned above, the use of TENS and topical capsaicin, and the benefits of a daily exercise program may lead to a decrease in the amount of needed medication. (The management of pain is discussed in Chapter 118.)

Neurogenic Bladder and Bowel Bladder and bowel dysfunction occur in a variety of neurologic diseases, resulting in incontinence, urinary retention, and infections (see also Chapters 13 through 15, 119, and 121). Serious complications include upper urinary tract infections, renal stones, renal damage, and skin breakdown. Incontinence and the need to learn techniques such as selfcatheterization may result in embarrassment and loss of self-esteem, and are often thought of as burdensome. It should be emphasized that bladder and bowel incontinence

are major factors in determining whether or not a patient can remain in a home setting. Bladder Incontinence Anatomically, the bladder empties into the urethra via two sets of sphincters: the internal sphincter, which is under involuntary control, mediated by the sympathetic nervous system, and the external sphincter. The external sphincter is composed of two distinct components: the pelvic floor musculature, which surrounds the urethra as a sling, and the intrinsic urethral striated muscle, which surrounds the distal urethra. The pelvic floor muscles receive somatic innervation, while the urethral external sphincter receives somatic, sympathetic, and parasympathetic innervation. At rest, the detrusor is usually relaxed and the sphincters tonically contracted to ensure continence. During micturition, the detrusor contracts and the sphincters relax in a coordinated fashion, thus preventing the development of high intravesical pressures and allowing for complete emptying (with normal residuals of ⬍25 mL). Detrusor contraction is mediated by the parasympathetic nervous system, originating from the intermediolateral cell column in cord levels S2-S4. Both motor and sensory components course through the pelvic nerves. Parasympathetic afferents are responsible for the sensation of bladder filling. Sympathetic neurons travel to the internal sphincter via the hypogastric nerve and arise from cord levels T11-L2. Sympathetic fibers also terminate on receptors in the detrusor, bladder neck, and urethra and on parasympathetic ganglia, providing reciprocal innervation and control. Activation of the sympathetic nervous system results in relaxation of the detrusor (␤-adrenergic transmission), contraction of the bladder neck and internal sphincter (␣-adrenergic), and inhibition of parasympathetic outflow. Somatic efferents to the external sphincter arise in cord levels S2-S4 and travel via the pudendal nerve. During micturition, the parasympathetics are activated and the sympathetic and somatic systems inhibited, resulting in detrusor muscle contraction and sphincter relaxation. These peripheral pathways are coordinated by descending pathways originating in the pontine micturition center. The pontine micturition center receives input from cortical and subcortical structures, thus allowing for voluntary control of bladder function. Diseases of and injury to the conus medullaris, cauda equina, and peripheral nerves lead to an areflexic, insensate bladder and incompetent sphincter. Examples include trauma, disc disease, spinal stenosis, and neuropathies affecting the autonomic nervous system (diabetes being the most common). Pelvic floor innervation is frequently affected in conus lesions, which can lead to incontinence, especially in women. In cauda equina lesions, the nonmyelinated pelvic nerve roots are usually damaged, leaving the pelvic floor innervation more intact than detrusor

Rehabilitation Management of Neuropathies

innervation. Autonomic nervous system involvement, most commonly seen in diabetes, most often first affects the detrusor afferents, resulting in decreased bladder sensation. Detrusor efferents are involved later. Decreased sensation contributes to detrusor overstretching, further compromising contractility. Overflow incontinence is the usual problem resulting from lower motor neuron lesions. Urodynamic studies are strongly recommended to accurately delineate the function of the detrusor, sphincters, and pelvic floor prior to embarking on a treatment program. The usual initial treatment is intermittent catheterization, and this may end up being the continuing treatment of choice. Credé’s maneuvers and straining may be helpful voiding techniques, but women often develop severe stress incontinence. These women and others with an incompetent sphincter may be candidates for a pelvic fascial sling or an artificial sphincter. Occasionally, men can benefit from ␣-adrenergic agents to enhance internal sphincter function. Cholinergic drugs such as bethanecol may be useful in inducing bladder contractions in those patients with partial lesions, as indicated by the presence of low-amplitude contractions on urodynamic studies. Electrical stimulation of the dorsal columns and sacral roots has yielded mixed results.

Bowel Incontinence Neurogenic bowel dysfunction results from autonomic and somatic denervation, resulting fecal incontinence, constipation, and difficulty with evacuation. Neurogenic bowel dysfunction is often hidden from public view, but is a major disability for many individuals with stroke, spinal cord injury, diabetes, and neuropathy. When sensation and mobility are limited, the person’s ability to anticipate bowel movements and participate in a bowel care program may be further limited. Bowel management should be pursed as a part of rehabilitation. The primary objectives of a bowel care program are (1) to prevent unplanned bowel movements; (2) to promote regular and predictable evacuations; and (3) to prevent complications such as infection, impaction, and perforation. The colon, the terminal segment of the intestine, is designed for fecal collection, formation, storage, and defecation. The colon and anorectal mechanisms receive parasympathetic, sympathetic, and somatic innervation and contain the intrinsic enteric nervous system between the muscular layers and under the mucosa. Neurogenic bowel results from involvement of any of these; however, in most cases the intrinsic enteric nervous system remains intact, except in diabetes and Hirschsprung’s disease. The intact enteric nervous system often serves as the basis for bowel training programs, because it can continue to modulate and integrate bowel function even in the absence of autonomic and somatic nervous system input. The colon is a reservoir for food waste until the time is convenient for

2627

elimination, which occurs when the colonic pressure exceeds the anal sphincter mechanism. The colon also reabsorbs water and provides an environment that facilitates bacterial breakdown of food waste products. The intrinsic enteric nervous system is the key to proper functioning of the gastrointestinal (GI) tract. It is divided into two layers, the submucosal (Meissner’s) plexus, and the intramuscular myenteric (Auerbach’s) plexus. The segment-to-segment function of the GI tract is largely regulated by the enteric nervous system, with modulation by the parasympathetic and sympathetic nervous systems. Sympathetic stimulation tends to promote storage by enhancing tone and inhibiting colonic contractions, although little clinical deficit results from bilateral sympathectomy.17 Parasympathetic activity enhances colonic motility, and loss of its input often results in constipation, impaction, and pseudo-obstruction.17 The rectum is usually empty until just prior to defecation. Continence is maintained by the anal sphincter mechanism, which consists of the internal anal sphincter, the external anal sphincter, and the puborectalis muscle.46 Anal pressure can be voluntarily increased by contracting the external sphincter and the puborectalis muscle. The external sphincter is innervated by S2-S4 nerve roots via the pudendal nerve, and the puborectalis muscle is directly innervated by branches of the S1-S4 nerve roots. Normal defecation begins with reflexes triggered by rectosigmoid distention. A rectorectal reflex occurs, resulting in relaxation of the bowel segment distal to the bolus, to facilitate propulsion further down the tract. The internal sphincter also reflexively relaxes. One then voluntarily contracts the levator to relax the external sphincter and puborectalis muscles, and defecation occurs. The external sphincter generally tenses in response to small rectal distentions via a spinal reflex centered in the conus medullaris, but will relax in response to greater distention. The gastrocolic reflex refers to the increased colonic motility that occurs 30 to 60 minutes following a meal. It is often used to promote regular bowel evacuation after spinal cord injury. Polyneuropathy, conus medullaris or cauda equina lesions, pelvic surgery, vaginal delivery, or even chronic straining during defecation can impair the somatic innervation of the anal sphincter mechanism. These conditions can also impair parasympathetic and sympathetic innervation. If an isolated pudendal insult has occurred, colonic transit is normal and fecal incontinence is the result. If parasympathetic supply is also lost, colonic sluggishness will occur, and difficulty with evacuation and possibly obstruction may occur. The result is often liquid stool leaking around a solid impaction, flowing uncontrollably out of the incompetent sphincter. Fecal incontinence is also likely to occur during increases in intra-abdominal pressure caused by normal daily activities. Patients with lower motor neuron lesions have decreased anal tone and a decreased or absent anocutaneous reflex on examination.

2628

Diseases of the Peripheral Nervous System

Maintaining an acceptable degree of continence is often difficult in peripheral nerve lesions that affect the bowel, owing to the usual incompetence of the sphincter. Use of fiber to add bulk and create a medium formed stool, combined with the use of a suppository that makes use of the gastrocolic reflex, is often the initial approach. Stool softeners may be avoided because they create a more pasty consistency that creates more risk of intermittent soiling. At times, digital rectal stimulation or the addition of a weekly or twice-weekly enema is needed. Careful review and alteration of diet may also help. Fewer long-term complications will ensue if rectal overdistention, rectal trauma, and overuse of oral stimulant medications are avoided. Diabetic patients may also be afflicted with colonic bacterial overgrowth, which can lead to periodic explosive diarrhea. Antibiotic treatment or prophylactic low-dose antibiotics may help with this problem. Electrical stimulation and artificial anal sphincters remain investigational and to date have not produced clinically reliable results.46 Satisfactory bowel care regimens can be achieved in 70% to 80% of patients with lower motor neuron lesions.68 In cases in which bowel care time is excessive, or there have been frequent impactions or frequent soiling, a colostomy may be a reasonable choice of last resort.

Sexual Dysfunction Because issues of medical care, prognosis, and recovery usually dominate the initial phases of a disease or injury, sexual dysfunction often becomes a major issue in the rehabilitation phase. Peripheral neuropathy can affect sexual function directly or contribute to problems via its effects on bowel and urinary function. (Chapter 14 provides additional information on the autonomic systems to the urinary bladder and sexual organs.) In men, the parasympathetic erectile and afferent fibers travel through the pelvic nerves that originate in the S2-S4 segments of the spinal cord. The sympathetic motor supply involved in psychogenic erections originates in the T10-L2 level and travels directly via the nerve roots. Somatic input controlling ejaculation travels in the pudendal nerve (S2-S4). In patients with spinal cord injury above T10, reflex and spontaneous erections are common once the period of spinal shock has passed. However, they are often not adequate or significantly long lasting to allow intercourse. In response to a vibrator stimulus, reflex ejaculation may occur beginning 6 months or more after cord injury. Successful pregnancies have resulted from electroejaculation techniques in spinal cord injury. Lesions involving the sacral cord, cauda equina, and peripheral nerves usually result in impaired erection as a result of damage to sensory or motor fibers, or both. In women, the parasympathetic pelvic nerves (S2-S4) control tumescence of the clitoris and vaginal secretions.

Sympathetic and splanchnic nerves (T10-L2) innervate the smooth muscle of the fallopian tubes and uterus. Contraction of the pelvic floor is controlled by somatic pudendal nerves emanating from S2-S4. Loss of the ability to achieve orgasm and impaired vaginal lubrication occur in both spinal cord injury and peripheral nerve damage. Intercourse-induced urinary tract infections may become a problem. In men with peripheral nerve damage, fertility is permanently reduced when compared to normal. Women may experience a transient cessation of menses and ovulation in the face of spinal cord injury or other serious systemic illness (e.g., GBS), but fertility generally returns to normal. As an example of peripheral neuropathy, consider diabetes, the most common cause of neuropathy in this country. Within 5 years of diagnosis of diabetes mellitus, one third to one half of men will experience some sexual dysfunction.38 This is usually lack of ability to attain or sustain an erection secondary to a combination of neurologic and vascular factors. Sexual difficulties are often further complicated by the resulting psychological distress. Oral sildenafil is both effective and well tolerated by men with diabetes-related erectile dysfunction.59 Other choices include the direct instillation of vasoactive drugs into the penis and the implantation of a penile prosthesis. Consultation with a urologist is strongly suggested. Women with diabetes experience impaired sensation/lack of orgasm, decreased vaginal lubrication, and an increased risk of urinary tract and vaginal infections. Voiding immediately prior to intercourse and use of a vaginal lubricant may help with some of these symptoms. A thorough review of the patient’s medications is also warranted because many commonly prescribed drugs, such as certain antidepressants and antihypertensives, may impair desire, sexual function, or both. The importance of addressing sexuality and sexual function in neurologically impaired patients in a sensitive manner, and the need to provide accurate, factual information, cannot be overstressed. Counseling should be provided to address psychological and behavioral issues, and to teach different techniques and approaches that are helpful in dealing with physical impairments affecting sexual performance. It is essential that all concerned recognize the importance of sexuality in restoring an acceptable quality of life.

The Consequences of Immobility and Deconditioning Immobility and deconditioning and their consequences pose major problems for the rehabilitation specialist. Combined with improper positioning, immobility can lead to pressure neuropathies, pressure sores, and loss of strength and ROM. Contractures may begin to form within as little

2629

Rehabilitation Management of Neuropathies

as 3 days of immobility. Edema hastens contracture formation, and spasticity and unequal weakness complicate its management. Hip flexion, knee flexion, and plantar flexion contractures are fostered by the common placement of a relatively immobile patient in bed on his or her back with the head of the bed elevated and a break in the bed at the knees. Frequent position changes, appropriate splinting, and ROM exercises will all serve to reduce the incidence and severity of contractures. In at-risk joints, ROM should be measured weekly. Immobility also results in reduction of muscle strength by as much as 20% per week,53 as well as loss of mineral from bone with resultant osteoporosis and greater susceptibility to fractures. In patients with severe GBS and quadriparesis, calcium loss from bone can be sufficiently severe as to produce symptomatic hypercalcemia and hypercalciuria. Other common complications of immobility include orthostatic hypotension (which may be further compromised by autonomic neuropathy), cardiovascular deconditioning, and thromboembolic phenomena. In patients who are not ambulating a minimum of 50 feet, one should begin deep venous thrombosis (DVT) prophylaxis with subcutaneous heparin or fractionated heparin unless there are strong contraindications. In those cases, alternating compression boots are the next best choice. Measurement of leg circumference and ultrasound studies when the circumference increases by greater than 1 cm are reasonable monitoring techniques in highrisk patients. Patients with overt DVT should be treated with full-dose intravenous heparin, followed by 3 months of therapeutic coumadin (6 months in cases of pulmonary embolus). The development of a decubitus ulcer will occur over bony prominences when the pressure exceeds the skin tolerance. The occiput, sacrum, coccyx, greater trochanter, heel, lateral malleolus, and foot are common locations for pressure sores. Friction plays a significant role, as does decreased sensation and decreased skin tolerance, as well as maceration and infection. Skin tolerance and skin healing are reduced in the presence of a neuropathy (further complicated by vascular compromise in diabetics and those with peripheral vascular disease).29,54 Skin breakdown also occurs in the presence of splints and casts. Frequent inspection of the skin in high-risk areas, frequent turning of the immobile patient combined with the use of a pressure-relief mattress, appropriate padding of splints and chairs, and appropriate footwear are all helpful. Cleanliness and the use of barrier creams such as zinc oxide are beneficial in the setting of incontinence. Optimal nutrition is also important to wound healing. Decubiti and chronic wounds add considerably to morbidity, mortality, and the cost of medical care. The reader is referred to Salcido et al.64 for a detailed review of the prevention and management of chronic wounds and pressure ulcers.

SPECIFIC DISORDERS Generalized Sensorimotor Peripheral Neuropathy The most common form of peripheral neuropathy is a diffuse sensorimotor polyneuropathy. Common causes include diabetes mellitus, alcoholism, systemic diseases such as collagen-vascular disease, vasculitides, and malignancy, as well as metabolic phenomena, genetic predisposition, and the toxic effects of drugs and chemicals. Diabetes deserves special mention, because it is the most common cause of peripheral neuropathy and because it is also associated with several other neuropathies, including autonomic neuropathy, multiple mononeuropathy, thoracolumbosacral polyradiculopathy, and acute cranial neuropathies. Diffuse polyneuropathy also predisposes the patient to pressure neuropathies, most commonly of the median, ulnar, and peroneal nerves. Close attention to the function-impairing effects of the patient’s comorbidities is crucial to effective rehabilitation management. An interdisciplinary team approach will maximize access to resources and facilitate return to maximal function. The rehabilitative program should be individualized to address specific symptoms and functional problems that will occur over the course of the patient’s life. Sensory loss, pain, and weakness, with the addition of autonomic dysfunction, are the common symptoms of generalized sensorimotor neuropathy, and their management is covered earlier in this chapter. Symptoms usually begin in the feet and lower extremities, and by the time sensory symptoms reach the knees, the hands may be affected. A few additional comments are made here. Therapeutic success is most likely to occur with an educated patient and family. For example, a study of recurrent neuropathic ulcers revealed that nearly half were solely attributable to failure to comply with proposed treatment.31 In spite of extensive foot ulceration, patients may ignore a non–weight-bearing recommendation! For anyone with a distal neuropathy and sensory loss, principles of skin inspection and care should be reviewed, the household hot water should be turned down below 120 °F, and barefoot walking is not allowed. Footwear for an insensate foot should generally attempt to accommodate rather than correct deformities. This includes an adequate toe box, extra depth, removable insoles, a rocker-bottom sole to disperse foot contact forces, and construction material that allows heat dissipation. When orthoses are needed to stabilize joints or compensate for weakness, they should be designed with as long a lever arm as possible to minimize the forces exerted at the ends of the orthosis, and to distribute pressure at contact points over as large an area as possible. Patellar tendon–bearing orthoses are designed to transfer floor reaction forces to the tibial plateau, and are used during ambulation after the acute phase of a neuropathic

2630

Diseases of the Peripheral Nervous System

arthropathy. They should be worn until radiographs demonstrate resolution of osteopenia and reconstruction of normaldensity bone.50 When paralysis of the interossei and lumbricals of the hands occurs, a bar across the dorsum of the proximal phalanges facilitates extension of the interphalangeal joints of the fingers via the long extensors. A universal hand cuff consists of a flat palmar pocket into which common utensils can be inserted. A C-bar, thumb post, or metacarpophalangeal stop may be added. Opponens orthoses assist with positioning of the thumb in opposition to the other fingers. Therapeutic heat and cold must be used with caution in patients with sensory loss, and in practice are rarely used in treatment of peripheral neuropathy. As noted earlier in this chapter, a combined approach of TENS, desensitization techniques, titration of medications, and a rational exercise program is often helpful in decreasing neuropathic pain. One should never lead the patient to believe that all pain will be eliminated, although, as the neuropathy progresses, the neurogenic pain may lessen or abate.73 Exercise is helpful in patients with neuropathy to counter the negative effects of disuse and deconditioning.27 Any exercise program should be closely monitored for signs of overwork weakness and cardiovascular compromise. For general conditioning, nonimpact or low-impact programs will help prevent foot injury. Foot inspections should always be performed after exercise sessions. As a general rule, fatigue or weakness that persists into the next day implies that the exercise has been too aggressive and the patient should take a day off and then resume the exercise at a reduced intensity. Use of canes, crutches, and walkers provides stability and transfers weight to the upper extremities. Consequences of this weight transfer may include injury to the shoulder girdle, wrist, and elbow; aggravation of degenerative arthritis; and the development of median and ulnar neuropathy of the wrist.

sensory involvement). The etiology is unknown but thought to be an autoimmune process triggered by infections, immunizations, and other events.3 In about 20% of cases a chronic, relapsing form occurs. Patients with acute GBS should be hospitalized and monitored closely for signs of impending respiratory dysfunction and cardiac arrhythmias. Patients who are immobile are at risk for all of the complications associated with immobility and bedrest (see above), and appropriate preventative measures should be taken (ROM exercise, proper positioning, splinting, DVT prophylaxis, pressurerelief mattress and frequent turning, and meticulous pulmonary toilet). Plasmapheresis has been shown to reduce the overall duration of ventilation time and hasten recovery and is the treatment of choice for those with ambulationimpairing weakness,30 unless contraindicated by recent myocardial infarction, active angina, or sepsis. The rehabilitation program is advanced as the patient improves. Orthostatic hypotension may be problematic and is treated via gradual reaccommodation to the upright posture (e.g., use of a tilt table), pressure stockings, and an abdominal binder. Mineralocorticoids may be needed in severe cases. GBS patients are particularly susceptible to overwork weakness,5 and any progressive strengthening program requires careful monitoring. Programs should be nonfatiguing until the muscles reach antigravity strength. Orthoses and assistive devices are useful in returning patients to independent ambulation and to independence in ADLs. Such devices often require modification as the patient improves. Neurogenic pain may benefit from use of TENS and antidepressant/anticonvulsant medications. Fully 80% to 90% of patients with acute GBS achieve complete recovery.61 Needle electromyography is the best prognostic indicator,58 and those with signs of denervation, indicating axonal degeneration rather than just demyelination, will recover more slowly, and often incompletely.

Brachial Plexus Injuries and Disease Acute and Subacute Polyradiculoneuropathy Acute inflammatory demyelinating polyradiculoneuropathy (GBS) usually initially affects the lower extremities, beginning with paresthesias in the toes and feet followed by ascending weakness.21 It can, however, begin with proximal weakness, and the Miller-Fisher variant comprises ataxia, areflexia, and ophthalmoplegia.24 The weakness typically progresses over days to weeks, and in some cases involves the respiratory muscles and produces a flaccid quadriplegia. Autonomic dysfunction is frequent and includes cardiac arrhythmias, orthostatic hypotension, gastroparesis, urinary retention, and altered sweating. Pain in the back and extremities is common acutely, and neurogenic extremity pain frequently accompanies the reinnervation phase. Sensation to touch, vibration, and joint position may be impaired to varying degrees (large fiber

Brachial plexus injuries occur in motor vehicle accidents, sports, and other activities that result in either distraction of the head and neck from the shoulder or falling on the arm with the shoulder in an abducted position. Such injuries are often accompanied by other life-threatening trauma and masked by the symptoms of brain or spinal cord injury, and may not be apparent until the patient is in rehabilitation and can cooperate with a neurologic examination. This underscores the importance of reexamination by the rehabilitation physician, with careful attention to deficits that do not fit the usual patterns seen in brain and spinal cord injury (e.g., hyporeflexia in an arm; upper extremity monoparesis; greater weakness at the elbow or in upper extremity flexors in brain injury; sensory loss in the arm). Brachial plexus stretch injury can also occur when an arm that is hemiplegic from a stroke or brain injury is stretched during an improper

Rehabilitation Management of Neuropathies

transfer or other activities (weakness of the shoulder girdle predisposes to this injury). Brachial plexus injury can also occur at birth or as a result of primary or metastatic tumors, and brachial plexitis can occur idiopathically after an infection or vaccination. Few professionals see a sufficient number of patients with brachial plexus injuries to attain a high level of expertise in treatment, and this results in many patients receiving less than the standard of care. Evaluation and treatment of patients with brachial plexus lesions requires challenging detective work, meticulous attention to detail, and consultation with many other specialists to develop an integrated team approach to treatment. Traumatic Brachial Plexus Injury A plexus injury can occur at any point from the beginning of the spinal roots at the cord level to distally at the point where the peripheral nerves arise from the cords of the plexus. Plexus injuries are classified as supraclavicular or infraclavicular. The supraclavicular portion of the brachial plexus includes the plexus elements from the intradural spinal roots distally to the divisions. The infraclavicular portion of the plexus includes the cords and peripheral branches. Supraclavicular injuries usually result from landing on the head and shoulder, causing distraction of the head and neck and stretching or tearing of the plexus. If avulsion fractures of the cervical transverse processes or a first rib fracture are seen on radiographs, by inference the corresponding nerve roots have been avulsed from the spinal cord. Other signs suggesting supraclavicular injury are evidence of injury to nerves arising from the supraclavicular portion (long thoracic and dorsal scapular), Horner’s syndrome, and swelling in the supraclavicular fossa. Injuries to the supraclavicular portion comprise about 75% of brachial plexus injuries, and in most cases severe damage occurs, commonly with root avulsions and a poor prognosis for recovery.44 The upper plexus (C5-C7) is involved in 20% to 25% of injuries, while isolated lower plexus injury (C8-T1) represents only 3%. Middle plexus (C7) involvement is associated with either upper or lower plexus injury. Involvement of all plexus elements is most common. Infraclavicular injuries, accounting for 25% of all plexus injuries, usually involve the posterior cord and/or the nerves that arise from the posterior cord. The mechanism of injury often involves landing on an outstretched arm. Associated injuries include shoulder dislocation and humeral fracture. Infraclavicular injuries are usually milder owing to the restricted movement of the humerus and the fact that the traction occurs further from the point of anatomic anchorage. Isolated injury to the axillary nerve is frequently seen because it is at the point closest to its anchorage. Neurapraxia, axonotmesis, and neurotmesis can all be seen with plexus injuries. Neurapraxic injuries carry the

2631

best prognosis, with complete recovery usual within a few weeks. In axonotmesis, the nerve sheath is intact but axons are disrupted, and recovery is incomplete. Proximal muscles and sensation are more likely to recover than distal muscles and sensation. With neurotmesis, spontaneous recovery is not likely, and surgical intervention may be undertaken. Meticulous history and examination along with electrodiagnostic studies and magnetic resonance imaging (MRI) or computed tomography will delineate the location and extent of injury. Presence of the sensory nerve action potential in the face of absent sensation is indicative of root avulsion. Knowledge of muscle innervation and a thorough muscle survey will help place the lesion at the root, trunk, or division level. In the first few weeks and months following injury, the rehabilitation goals are prevention of further injury and maintenance of function. The humeral head should be supported to prevent stretch of the rotator cuff, and the arm elevated when possible to prevent edema. ROM exercises are essential. Electrical stimulation of muscles that show weak volitional activity may be used in addition to active assisted exercise. In situations in which recovery is expected to be incomplete, teaching of one-handed techniques should begin. Brachial plexus surgery should also be considered early in the course of treatment. The timing of surgery depends on the severity of injury and expected recovery, with procedures undertaken earlier if neurotmesis is confirmed or strongly suspected. When brachial plexus repair is performed more than 6 months postinjury, the chances of optimal recovery are reduced.2 Pain is a significant problem following brachial plexus injuries, occurring in 90% of patients and persisting at 3 years in one third.72 Treatment with anticonvulsants/ antidepressants or TENS is often quite helpful. In cases with a sympathetic component, stellate blocks or Bier blocks may provide some benefit. Epidural stimulation and dorsal root entry zone procedures have been successful in some, but convincing research is lacking. Orthotics play a role in ongoing management of plexus injuries. If distal hand function is preserved and shoulder movement and elbow flexion are impaired, an orthosis is designed to stabilize the shoulder and provide for elbow flexion via a control cable. When hand function is impaired, a wrist-driven flexor tenodesis splint can achieve reasonable pinch, and if C7 function is present, gross grasp and release can be achieved. In addition to nerve repair and grafting procedures, several orthopedic procedures are used to improve function.60 These include stabilization procedures and tendon transfers. It is important to continue aggressive physical and occupational therapy following both neurosurgical and orthopedic procedures. Some patients with a flail, insensate, and/or painful arm have elected amputation. If the entire arm is insensate, a

2632

Diseases of the Peripheral Nervous System

transhumeral amputation and stabilization of the glenohumeral joint, followed by application of a functional prosthesis, can achieve an acceptable result. Amputation is often not successful in eliminating pain, especially in the setting of root avulsion. In injuries involving partial axonotmesis, recovery may take place over 18 months or more, and deferral of orthopedic reconstructive procedures should be delayed until recovery has plateaued. Obstetric Brachial Plexus Palsy Brachial plexus palsy at birth occurs in 1 to 2 babies per 1000 live births,28 and most have significant recovery on their own. The most common level of injury is C5-C6 (Erb’s palsy), with some infants showing involvement of C7. Involvement of the lower plexus (C8-T1; Klumpke’s palsy) is most commonly seen as a residual of initial total plexus involvement.11 MRI is often employed to assist in diagnosis because the examination of newborns is often limited. Electrodiagnostic testing is carried out 1 to 2 weeks after birth to assist in localization and to establish a prognosis. Occupational and physical therapy are begun in the first week of life, the main goal being education of parents and those caring for the infant in techniques of proper positioning and careful ROM exercises. Orthoses are used to prevent, correct, or minimize the occurrence of contractures. As strength returns, the uninvolved arm is periodically immobilized to promote use of the involved arm. When the arm is significantly involved, the infant may not recognize the extremity as a useful body part. At times, even when recovery occurs, the infant will not use the extremity. Encouraging the infant to look at and play with the arm, and forced use as described above, may help. Persistent pain is generally not a common sequela of birth brachial plexus injuries. If motor recovery does not improve over the first 3 months, surgical repair should be considered.51 Tendon transfers and other orthopedic procedures are considered when neurologic recovery is incomplete and function remains impaired. Idiopathic Brachial Plexitis Idiopathic brachial plexitis may occur following immunization or viral or other illness. The proximal portion of the plexus is most commonly affected, and about 20% of cases are bilateral. Symptoms usually begin with neck pain radiating into the shoulder, followed by the development of weakness and paresthesias over several hours to a few days. Most individuals experience near complete recovery, which may occur over 18 months. Because periscapular muscles are frequently impaired, a figure-of-8 harness may provide some support and prevent overstretching of the muscles. Depending on the location and severity of the weakness, as well as the prognosis for recovery, the same rehabilitation interventions described for traumatic injuries apply. A short course of steroids or other anti-inflammatory medication

may help with the acute pain. About 20% of cases will have a recurrence later in life.

Postpolio Syndrome Although acute poliomyelitis has been virtually eliminated in the United States, there are several hundred thousand polio survivors,26 many of whom are experiencing increasing or new symptoms as they age. Common symptoms include pain, increased weakness, fatigue, and respiratory and swallowing dysfunction.1 Pain is often related to overuse syndromes in compensation for muscular weakness, including degenerative arthritis, tendonitis, myofascial pain, and progressive scoliosis. Various theories have been proposed for the increased weakness, including age-related loss of motor neurons, premature aging of normal or partially damaged motor neurons, chronic poliovirus infection, and enhanced susceptibility of motor neurons to other insults. Overuse weakness may contribute in some cases.4 It is interesting to note that most patients do not appreciate muscle weakness until approximately 50% of muscle strength is lost, and many patients experience subclinical involvement that is more widespread than they report. Fatigue is difficult to objectively assess and is likely attributed to a combination of factors, including cardiovascular deconditioning and peripheral neuromuscular and/or emotional factors. The above emerging symptoms combine to produce new dysfunction in the performance of daily activities such as walking, dressing, and performing household chores or work activities. Pain and ADL or mobility dysfunction are the usual reasons postpolio patients seek medical attention. Because postpolio patients present with complex issues in combinations that are unique to each patient, it is imperative to carefully identify the patient’s impairments, primary functional concerns, and emotional status. The ultimate goal of treatment is to enhance the individual’s function and independence as much as possible. As noted above, pain in postpolio patients is often multifactorial. Many do not wear orthoses that provide optimal support for ambulation, which leads to chronic joint pain and damage.57 Most improve with the fitting of a properly designed orthosis. Shoulder pain may result from overuse or improper use of assistive devices, or improper transfer techniques. In patients with unequal trunk muscle weakness, scoliosis is a common problem that is often worsened by a poorly designed wheelchair seating system that needs modification. Many postpolio patients continue in the performance of activities until they have persistent pain or fatigue, and will benefit from the use of assistive devices or frequent rest breaks. Although anti-inflammatory medications may help with symptoms, identification and modification of the underlying problem is key to optimal management. Carefully monitored muscle strengthening and cardiovascular fitness exercises have been shown to improve function in polio survivors.22,40

Rehabilitation Management of Neuropathies

Patients with respiratory complaints should be referred for pulmonary function studies, or sleep studies when indicated. Likewise, complaints of swallowing difficulties should be thoroughly evaluated by videofluoroscopy. Appropriate referral for treatment (e.g., nighttime bilevel positive airway pressure for sleep apnea; consultation with a speech pathologist to address compensatory techniques in dysphagia) is essential for proper management. Because most postpolio patients are in the 50- to 75year age range, it is important to identify the effect of comorbidities and advancing age on the patient’s function. Comprehensive assessment and coordinated treatment of interacting causative factors will yield optimal results. This discussion represents merely a brief summary of issues encountered in postpolio patients; the reader is referred to an excellent comprehensive review by Agre and Rodriguez1 for further information.

Motor Neuron Disease Once the diagnosis of motor neuron disease or amyotrophic lateral sclerosis (ALS) is made, it is essential to educate and support the patient in a consistent manner, to treat new medical problems as they appear, and to utilize a variety of supportive interventions to enhance the quality of life. Management of ALS may be considered in four domains: educational, rehabilitational, psychosocial, and pharmacologic. An informed patient and family are vital. Diagnostic statements regarding prognosis and rate of deterioration should be avoided; the patient should be informed that the disease is progressive, but the rate of progression is unpredictable. The patient should be encouraged to do things sooner rather than later, while strength is at its best. The patient and family should be told that therapeutic advances will be communicated to them, and they are encouraged to call when they read of breakthroughs on the Internet or in other sources. Patients should be encouraged to participate in research trials. Continuing employment for as long as feasible may result in better life satisfaction. Inactivity in patients with ALS results in further functional deterioration, and exercises to optimize strength and conditioning should be begun early in the disease. A few sessions with a physical therapist will help the patient avoid overwork injury and weakness. When spasticity is severe, isotonic exercises combined with stretching and ROM are recommended. Isometric and resistive exercises may produce cramps and fatigue. As weakness progresses, concentration on stretching and ROM becomes important in preventing contractures related to unequal muscle weakness and immobility. Prevention of contractures is important because, even late in the disease, contractures will prevent adequate hygiene, dressing, and seating and can result in significant pain. AFOs help to maintain a safe

2633

gait for those with footdrop, and canes or walkers may be useful gait aids. Dysarthria and dysphagia occur frequently in ALS. Alternative means of communication often become necessary, and a wide range of options exist. A modified barium swallow will assist in dysphagia management. Thickened liquids, double swallow, avoidance of dry crumbly foods, and other compensatory techniques may minimize the risk of aspiration. Alternative feeding techniques (e.g., a feeding tube) should be addressed early in the course of the disease. Weakness of the respiratory muscles ultimately occurs in all patients with ALS, leading to hypercapnic ventilatory failure. The early identification of respiratory failure may be difficult. Increased respiratory rate and use of accessory muscles may appear, but not necessarily early, and arterial blood gases show little change in early respiratory failure: CO2 retention occurs only when respiratory strength is less than 30% of normal. Clinical signs of CO2 retention often begin with morning headache. Patients usually do not complain of breathlessness until capacity falls to 40% to 50% of predicted. The most sensitive measures of respiratory strength are the forced inspiratory and expiratory pressures (PI and PE). PI less than ⫺20 cm H2O associated with hypercapnia, PE less than 40 cm H2O with ineffective cough and difficulty raising secretions, and a vital capacity of less than 20% of predicted all signal impending respiratory failure. Serial testing is recommended to identify respiratory failure in advance of the need for ventilatory assistance. Issues regarding the use of ventilatory assistance should be addressed early in the disease as well as repeatedly during the course of the disease. Ventilatory assistance should be considered when pulmonary function testing indicates impending respiratory failure, when the patient is dyspneic at rest, or when dyspnea impairs sleep. Initially, nighttime ventilation with a mask may be effective; however, patients with bulbar dysfunction and an inability to clear secretions are at risk for aspiration when noninvasive ventilation is used, and a tracheostomy may be considered, depending on the wishes of the patient. Pneumonia often acutely precipitates the need for respiratory support, so prior discussions with patient and family are critical. Long-term survival is possible with mechanical ventilation. Eventual quadriplegia results, and careful attention to skin care, positioning, hygiene, and ROM become paramount in preserving function and preventing infections and pain. For those who chose to attempt long-term survival with assisted ventilation, a myriad of electronic devices, powered wheelchairs, and environmental controls are available to allow continuation of a productive life. Views on what constitutes quality of life, and the circumstances under which an individual would choose to continue on in the face of severe impairments, are quite varied and often intertwined with religious and other beliefs. The

2634

Diseases of the Peripheral Nervous System

physician must be careful not to impose his or her belief system on the patient and family, and be supportive of whatever path they choose.

Radiculopathy The etiology, presentation, and diagnosis of radiculopathy are covered elsewhere in this text. The therapeutic approach to radiculopathy includes relative rest, nonsteroidal antiinflammatory drugs, epidural steroid injections, physical therapy, and surgery. Activity limitations generally involve avoidance of strenuous activities, as well as activities that cause increased pain. Early consultation with a physical therapist is beneficial in ensuring that the patient is aware of proper posture and body mechanics, and knows how to employ modalities such as heat, ultrasound, massage, and stretching that may help to alleviate associated myofascial pain. Traction is useful in alleviating pain associated with cervical radiculopathy, but does not influence the rate of recovery.35 Traction is not accepted as a treatment modality in lumbar radiculopathy because of the amount of force needed to overcome the weight and resistance of the lower half of the body. TENS is also helpful. As the radicular symptoms improve, strengthening may begin and activity is usually increased as tolerated. Between 80% and 90% of patients respond to conservative management.8,56 Epidural steroids are useful in both cervical and lumbar radiculopathies, with resultant improvement in both pain and the neurologic deficit, especially in acute patients.6 Referral for surgery is indicated when pain is persistent, or the patient has overt or progressive neurologic deficits. Coexisting spinal stenosis is a major adverse outcome predictor for both conservative and surgical therapy.62 Numerous outcome studies exist on the effects of various treatment approaches in both cervical and lumbar radiculopathy, but the results are difficult to interpret because of variations in the study subjects and the contributing effects of psychological and behavioral motivation factors, as well as superimposed pain of musculoskeletal origin. Many patients seen in office practice have been out of work for many months, are depressed, and have the behavioral components of a chronic pain syndrome, making treatment less successful.

REFERENCES 1. Agre, J. C., and Rodriguez, A. A.: Poliomyelitis and postpolio syndrome. Arch. Phys. Med. Rehabil. 87:1591, 2000. 2. Aminoff, M. J., Olney, R. K., Parry, G. J., et al.: Relative utility of different electrophysiologic techniques in the evaluation of brachial plexopathies. Neurology 38:546, 1988. 3. Barohn, R. J., and Saperstein, D. S.: Guillain-Barre syndrome and chronic inflammatory demyelinating polyneuropathy. Semin. Neurol. 18:49, 1998.

4. Bennett, R. L., and Knowlton, G. C.: Overwork weakness in partially denervated skeletal muscle. Clin. Orthop. 12:22, 1958. 5. Bensman, A.: Strenuous exercise can impair muscle function in Guillain-Barre patients. JAMA 214:468, 1970. 6. Benzon, H. T.: Epidural steroid injections for low back pain and lumbosacral radiculopathy. Pain 24:277, 1986. 7. Boonstra, A. M., vanWeerden, T. W., Eisma, W. H., et al.: The effect of low-frequency electrical stimulation on denervation atrophy in man. Scand. J. Rehabil. Med. 19:127, 1987. 8. British Association of Physical Medicine: Pain in the neck and arm: a multicentre trial of the effects of physiotherapy. Br. Med. J. 1:253, 1966. 9. Cabric, M., and Appell, H. J.: Effect of electrical stimulation of high and low frequency on maximum isometric force and some morphological characteristics in man. Int. J. Sports Med. 8:256, 1987. 10. Caggiano, E., Emrey, T., Shirley, S., et al.: Effects of electrical stimulation or voluntary contraction for strengthening quadriceps femoris muscles in an aged male population. J. Orthop. Sports Phys. Ther. 20:22, 1994. 11. Clarke, H. M., and Curtis, C. G.: An approach to obstetrical brachial plexus injuries. Hand Clin. 11:563, 1995. 12. Currier, D. P., Lehman, J., and Lightoot, P.: Electrical stimulation in exercise of the quadriceps femoris muscle. Phys. Ther. 59:1508, 1979. 13. Currier, D. P., and Mann, R.: Muscular strength development by electrical stimulation in normal individuals. Phys. Ther. 63:915, 1983. 14. DeLateur, B. J., Lehmann, J., Stonebridge, J., et al.: Isotonic vs. isometric exercises: a double shift, transfer-of-training study. Arch. Phys. Med. Rehabil. 53:212, 1972. 15. Delorme, T. L.: Restoration of muscle power by heavy resistance exercise. J. Bone Joint Surg. 27:645, 1945. 16. Delorme, T. L., and Watkins, A. L.: Techniques of progressive resistance exercise. Arch. Phys. Med. 29:263, 1948. 17. Devroede, G., and Lamarche, J.: Functional importance of extrinsic parasympathetic innervation in the distal colon and rectum in man. Gastroenterology 66:273, 1974. 18. Ebbeling, C. B., and Clarkson, P. M.: Exercise-induced muscle damage and adaptation. Sports Med. 7:207, 1989. 19. Eberstein, A., and Eberstein, S.: Electrical stimulation of denervated muscle: is it worth it? Med. Sci. Sports Exerc. 28:1463, 1996. 20. Ernstoff, B., Wetterqvist, H., Kvist, H., et al.: Endurance training effect on individuals with postpoliomyelitis. Arch. Phys. Med. Rehabil. 77:843, 1996. 21. Feasby, T. E.: Inflammatory-demyelinating polyneuropathies. Neurol. Clin. 10:651, 1992. 22. Feldman, R. M., and Soskolne, C. L.: The use of non-fatiguing strengthening exercises in post-polio syndrome. Birth Defects Orig. Artic. Ser. 23:335, 1987. 23. Fishbain, D. A., Chabal, C., Abbot, A., et al.: Transcutaneous electrical nerve stimulation (TENS) treatment outcome in long-term users. Clin. J. Pain 12:201, 1996. 24. Fisher, M.: An unusual variant of acute idiopathic polyneuritis (syndrome of ophthalmoplegia, ataxia and areflexia). N. Engl. J. Med. 255:57, 1956. 25. Frontera, W. R.: Exercise in physical medicine and rehabilitation. In Grabois, M., Garrison, S. J., Hart, K. A., and

Rehabilitation Management of Neuropathies

26.

27.

28.

29. 30. 31.

32. 33.

34.

35. 36.

37.

38. 39.

40.

41. 42.

43.

44. 45.

46.

Lehmkuhl, L. D. (eds.): Physical Medicine and Rehabilitation: The Complete Approach. Boston, Blackwell Science, p. 487, 2000. Gawne, A. C., and Halstead, L. S.: Post-polio syndrome: pathophysiology and clinic management. Crit. Rev. Phys. Rehabil. Med. 7:147, 1995. Graham, C., and Laskow-McCarthey, P.: Exercise options for persons with diabetic complications. Diabetes Educ. 16:212, 2000. Greenwald, A. G., Schute, P. C., and Shively, J. L.: Brachial plexus birth palsy: a 10-year report on the incidence and prognosis. J. Pediatr. Orthop. 4:689, 1984. Grunfeld, C.: Diabetic foot ulcers: Etiology, treatment, and prevention. Adv. Intern. Med. 37:103, 1991. Guillain-Barre Syndrome Study Group: Plasmapheresis and acute Guillain-Barre syndrome. Neurology 35:1096, 1984. Helen, P. A., Walker, S. C., and Pulium, G. F.: Recurrence of neuropathic ulceration following healing in a total contact cast. Arch. Phys. Med. Rehabil. 72:967, 1991. Hennessey, W. J., and Johnson, E. W.: Lower limb orthoses. Phys. Med. Rehabil. 16:326, 2000. Herbison, G. J, Jaweed, M. M., Ditunno, J. F., et al.: Effect of overwork during reinnervation of rat muscle. Exp. Neurol. 41:1, 1973. Herbison, G. J., Jaweed, M. M., and Ditunno, J. F. Jr.: Exercise therapies in peripheral neuropathies. Arch. Phys. Med. Rehabil. 64:201, 1983. Honet, J. C., and Puri, K.: Cervical radiculitis: treatment and results in 82 patients. Arch. Phys. Med. Rehabil. 57:12, 1976. Hortobagyi, T., Hill, J. P., Houmard, J. A., et al.: Adaptive response to muscle lengthening and shortening in humans. J. Appl. Physiol. 80:765, 1996. Hutton, R. S.: Neuromuscular basis of stretching exercises. In Komi, P. V. (ed.): Strength and Power in Sport. Boston, Blackwell Science, p. 29, 1992. Jensen, S. B.: Sexual relationships in couples with a diabetic partner. J. Sex Marital Ther. 11:259, 1985. Jones, D. A., Rutherford, O. M., Parker, D. F.: Physiological changes in skeletal muscles as a result of strength training. Q. J. Exp. Physiol. 74:233, 1989. Jones, D. R., Speier, J., Canine, K., et al.: Cardiorespiratory responses to aerobic training by patients with postpoliomyelitis sequelae. JAMA 261:3255, 1989. Kantrowitz, A.: Electronic Physiologic Aids. Brooklyn, NY, Maimonides Hospital, p. 4, 1960. Klausen, K., Secher, N. H., Clausen, J. P., et al.: Central and regional circulatory adaptations to one-leg training. J. Appl. Physiol. 52:976, 1982. Kottke, R. J.: The neurophysiology of motor function. In Kottke, F. J., and Lehmann, J. F. (eds.): Krusen’s Handbook of Physical Medicine and Rehabilitation, 4th ed. Philadelphia, W. B. Saunders, p. 234, 1990. Leffert, R. D.: Brachial Plexus Injuries. New York, Churchill Livingstone, 1985. Liberson, W. T., Holmquest, H. J., Scot, D., et al.: Functional electrotherapy: stimulation of the peroneal nerve synchronized with the swing phase of gait of hemiplegic patients. Arch. Phys. Med. Rehabil. 42:101, 1961. Madoff, R. D., Williams, J. G., and Caushaj, P. F.: Fecal incontinence. N. Engl. J. Med. 326:1002, 1992.

2635

47. McArdle, W. D., Katch, F. I., and Katch, V. L.: Essentials of Exercise Physiology. Philadelphia, Lea & Febiger, 1994. 48. McKee, P., and Morgan, L.: Orthotics in Rehabilitation: Splinting the Hand and Body. Philadelphia, F. A. Davis, 1998. 49. Melzack, R., and Wall, P. D.: Pain mechanism: a new theory. Science 150:171, 1965. 50. Michael, J. W., Isbell, M. A., and Harrelson, J. M.: Orthotic management of diabetic neuropathic arthropathy. J. Prosthet. Orthop. 4:45, 1991. 51. Michlow, B. J., Clarke, H. M., Curtis, C. G., et al.: The natural history of obstetrical brachial plexus palsy. Plast. Reconstr. Surg. 93:675, 1994. 52. Mueller, E. A.: Influence of training and of inactivity on muscle strength. Arch. Phys. Med. Rehabil. 51:449, 1970. 53. Mysiw, W. J., and Jackson, R. D.: Electrical stimulation. Phys. Med. Rehabil. 22:459, 2000. 54. NPUA Panel: Pressure Ulcers: Incidence, Economics, Risk Assessment. Consensus Development Conference Statement. West Dundee, IL, S-N Publications, 1989. 55. Patel, A. T., Garber, L. M., and Redford, J. B.: Upper limb orthotic devices. Phys. Med. Rehabil. 15:311, 2000. 56. Pearce, J., and Moll, J. M. H.: Conservative treatment and natural history of acute lumbar disc lesions. J. Neurol. Neurosurg. Psychiatry 30:13, 1967. 57. Perry, J., and Fleming, C.: Polio: long-term problems. Orthopedics 8:877, 1985. 58. Raman, P. T., and Taori, G. M.: Prognostic significance of electrodiagnostic studies in the Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 39:163, 1976. 59. Rendell, M. S., Raijfer, J., Wicker, P. A., et al.: Sildenafil for treatment of erectile dysfunction in men with diabetes: a randomized controlled trial. JAMA 281:421, 1999. 60. Richard, R. R.: Operative treatment for irreparable lesions of the brachial plexus. In Gelberman, R. (ed.): Operative Nerve Repair and Reconstruction. Philadelphia, J. B. Lippincott, 1991. 61. Ropper, A. H.: The Guillain-Barre syndrome. N. Engl. J. Med. 326:1130, 1992. 62. Saal, J. A.: Intervertebral disc herniation: advances in non-operative treatment. Phys. Med. Rehabil. Satte Art Rev. 4:175, 1990. 63. Sady, S. P., Wortman, M., and Blanke, D.: Flexibility training: ballistic, static or proprioceptive neuromuscular facilitation? Arch. Phys. Med. Rehabil. 63:261, 1982. 64. Salcido, R., Hart, D., and Smith, A. M.: The prevention and management of pressure ulcers. In Braddom, R. L. (ed.): Physical Medicine and Rehabilitation. Philadelphia, W. B. Saunders, p. 630, 1996. 65. Scott, O. M., Vrbova, G., Hydes, S. A., et al.: Responses of muscles of patients with Duchenne muscular dystrophy to chronic electrical stimulation. J. Neurol. Neurosurg. Psychiatry 49:1427, 1986. 66. Stanish, W. D., and McVicar, S. F.: Flexibility in injury prevention. In Renstron, P. A. F. H. (ed.): Basic Principles of Prevention and Care. Boston, Blackwell Science, p. 262, 1993. 67. Theriault, R., Boulay, M. R., Theriault, G., et al.: Electrical stimulation-induced changes in performance and fiber type proportion of human knee extensor muscles. Eur. J. Appl. Physiol. Occup. Physiol. 74:311, 1996.

2636

Diseases of the Peripheral Nervous System

68. Tobin, G. W., and Brocklehurst, J. C.: Faecal incontinence in residential homes for the elderly: prevalence, aetiology and management. Age Ageing 15:41, 1986. 69. Virani, A., Mailis, A., Shapiro, L. E., et al.: Drug interactions in human neuropathic pain pharmacotherapy. Pain 73:3, 1997. 70. William, H. B.: The value of continuous electrical muscle stimulation using a completely implantable system in the preservation of muscle function following motor nerve injury and repair: an experimental study. Microsurgery 17:589, 1996.

71. Wolf, D. L., Ariel, G. B., Sar, D., et al.: The effect of muscle stimulation during resistive training on performance parameters. Am. J. Sports Med. 14:18, 1986. 72. Wynn Parry, C. B.: Pain in avulsion injuries of the brachial plexus. Neurosurgery 15:960, 1984. 73. Young, R. J., Ewing, D. J., and Clarke, B. F.: Chronic and remitting painful diabetic polyneuropathy: correlations with clinical features and subsequent changes in neurophysiology. Diabetes Care 11:34, 1988.

118 Mechanisms and Pharmacologic Management of Neuropathic Pain MICHAEL C. ROWBOTHAM

Pharmacotherapy in Perspective Clinical Mechanisms Underlying Neuropathic Pain The Evidence Base and Generalizability of Clinical Trial Results Antidepressants

Anticonvulsants Clinical Use of Sodium Channel–Blocking Anticonvulsants Anticonvulsants That Do Not Block Sodium Channels: Gabapentin, Valproic Acid, Levetiracetam

This chapter focuses on pharmacologic management of chronic neuropathic pain. To supplement the pharmacologic orientation, a brief review of pain mechanisms believed to be relevant for common neuropathic pain disorders is included. Substantial progress has been made in validating animal models of acute pain and chronic neuropathic pain. These animal surrogates, extremely useful for elucidating pain mechanisms and developing new treatment targets, are now widely used for preclinical testing of analgesic drug candidates. The clinical science of unraveling pain mechanisms lags far behind the basic sciences. At present, many clinical syndromes encountered by the neurologist have no analogous animal pain model, and likewise, many clinical syndromes have never been the subject of a controlled treatment trial. The variety and power of assessment methods applicable for evaluating pain mechanisms and potential therapies is expanding. These tools include quantitative sensory testing, imaging of peripheral nerves, skin biopsy assessment of cutaneous innervation, experimental pain models, defined applications of pain-producing and pain-relieving substances, and drug infusion tests. Despite these advances, relieving chronic pain associated with injury or disease affecting the peripheral or central nervous system (CNS) remains a difficult treatment challenge. Fortunately, welldesigned and appropriately controlled clinical trials expand the list of proven treatments with every passing year. Four medication categories often employed to treat neuropathic

Topical Medications Opioids Tramadol The Research Landscape and Changing Clinical Practice Summary and Conclusions

pain are reviewed: antidepressants, anticonvulsants, opioids, and topical agents. All have limitations, and none provides even moderate or better relief in more than about 60% of treated patients. There is ample room for future improvement in each category and a continuing need for innovative treatments.

PHARMACOTHERAPY IN PERSPECTIVE Pharmacologic management is not a cure and should be considered an integral component of a more comprehensive approach to treatment. A discussion of the many widely used nonpharmacologic approaches, including physical therapy, psychological treatments, invasive procedures (e.g., neural blockade, dorsal column stimulation), and various complementary and alternative medicine interventions, is beyond the scope of this chapter. Pharmacotherapy is best used within a treatment context where education, support, and reassurance characterize the relationship between the patient and physician. In many cases, use of nonpharmacologic techniques may lower pain sufficiently for patients to rely on a minimum of medications. Although the package inserts for prescription medications list dosage ranges and contraindications, only a limited number of the medications to be reviewed here have U.S. Food and Drug Administration (FDA) approval 2637

2638

Diseases of the Peripheral Nervous System

for treatment of pain. Medication doses and titration schedules should be individually adjusted as necessary based on frequent and careful evaluation of side effects, treatment adherence, and pain relief. Use of doses higher than the maximum specified in the package labeling is not recommended. For opioids, optimum dosing levels are uncertain and controversial. For medications without a labeled pain indication, drugs with minimal risk that have demonstrated efficacy in one or more (ideally related) neuropathic pain syndromes are preferred. When efficacy has not been established for a particular medication, even more caution is needed to assure acceptable safety and tolerability in light of the patient’s medical condition, age, pain severity, degree of suffering, and previous treatment history.

CLINICAL MECHANISMS UNDERLYING NEUROPATHIC PAIN As listed in Table 118–1, neurologists encounter a broad range of peripheral nervous system and CNS diseases that can be associated with acute or chronic neuropathic pain. For purposes of treatment of the symptom of pain, there may not be that much difference between the syndromes except their peripheral or central location. Several reviews cover the many different mechanisms for the induction and maintenance of neuropathic pain in depth.40,49,126,182,185 Neuropathic pain mechanisms fall into two major categories: peripheral and central. The mechanisms may be linked in cascades, making it unlikely that any individual patient has neuropathic pain as a result of a single mechanism only. However, there may be a dominant and treatable mechanism that can reduce pain to minimal levels. Neuropathic pain often results from increased activity in primary afferent nociceptors. In some patients, reducing activity in ectopic impulse generators located in injured or abnormally functioning primary afferent fibers will dramatically reduce pain. The peripheral mechanism that is best understood is ectopic impulse activity resulting from changes in sodium channels.40 In teased axon recordings from dorsal roots in rats and rabbits with chronic nerve injury, Wall and Gutnick172 and Kirk84 demonstrated an increased level of ongoing discharge in afferent A and C fibers compared to normal intact dorsal root and dorsal root just after acute nerve section. Subsequent studies identified two principal sources of this activity: the nerve injury site (e.g., neuroma or nerve compression zone) and the associated dorsal root ganglia (DRGs).75,171 The development of ectopic hyperexcitability is thought to be due in part to remodeling of the electrical properties of the axon membrane via local increases in sodium channel density and changes in channel distribution. In addition to spontaneous firing, ectopic discharge can be produced by gentle tapping and by a

Table 118–1. Common Types of Acute and Chronic Neuropathic Pain Peripheral Neuropathic Pain Acute and chronic inflammatory demyelinating polyradiculoneuropathy Alcoholic polyneuropathy Chemotherapy-induced polyneuropathy Complex regional pain syndrome type 2* Entrapment neuropathies (e.g., carpal tunnel syndrome) Fabry’s disease Herpes zoster HIV sensory neuropathy* Iatrogenic neuralgias (e.g., postmastectomy pain, post-thoracotomy pain)* Idiopathic sensory neuropathy* Mononeuritis multiplex Nerve compression or infiltration by tumor Nutritional deficiency–related neuropathies Painful diabetic neuropathy—focal and distal symmetrical* Phantom limb pain Postherpetic neuralgia* Postradiation plexopathy Post-traumatic neuralgias Radiculopathy—cervical, thoracic, lumbosacral* Root avulsion (brachial plexus) Toxic exposure–related neuropathies Tic douloureux (trigeminal neuralgia)* Vasculitic neuropathy Central Neuropathic Pain Central poststroke pain* Compressive myelopathy from spinal stenosis, tumor, or space-occupying lesion HIV myelopathy Multiple sclerosis–related pain* Parkinson’s disease–related pain Postischemic myelopathy Postradiation myelopathy Post-traumatic spinal cord injury pain* Syringomyelia *Multiple controlled trials of pharmacotherapy for pain.

range of chemical stimuli. Abnormal afferent barrages that propagate into the CNS will directly elicit paresthesias, dysesthesias, and pain. For example, continuous discharge in C fibers produces continuous sensations of burning pain, and intermittent spontaneous bursts in A␦ or A␤ fibers produce lancinating dysesthesias or paresthesias.33,109 The CNS may contribute to the maintenance of neuropathic pain either as a result of a “normal” response to nociceptive input, or as a result of injury to the CNS. Central somatosensory pathways react to peripheral nociceptor input, either ectopic or evoked, by producing a pain whose

Mechanisms and Pharmacologic Management of Neuropathic Pain

magnitude and duration accurately reflects the nature, amplitude, and timing of the nociceptor input. Because the CNS is passive and peripheral input–dependent, preventing or blocking the nociceptor input immediately eliminates the pain. It is uncertain if any chronic neuropathic pain disorders fall into this category, except perhaps uncomplicated trigeminal neuralgia. Nociceptive input that is either low level and prolonged or brief and intense, regardless of the functional status of the primary afferent source, initiates reversible and activitydependent changes in synaptic efficacy in the dorsal horn of the spinal cord and more rostral structures. This complex cascade is described in simple terms as “central sensitization.”86,165 The response to nociceptor inputs becomes exaggerated and prolonged (hyperalgesia), dynamic low-threshold primary afferent input begins to evoke pain (allodynia), and pain is experienced outside the area of injury or input as the receptive fields of the CNS target neurons enlarge (secondary hyperalgesia).24,182 The pain experienced is thus the peripheral input amplified by the induced CNS responsiveness. The CNS changes are dependent or contingent on continuing peripheral input, especially through nociceptors. If this input is removed, the hyperalgesia, allodynia, and secondary hyperalgesia will gradually subside. If the input returns, the changes are promptly reinstated. This mechanism will contribute whenever the peripheral input exceeds the threshold for activation of central sensitization, whether or not it arises from intense activation of “normal” nociceptors, or from increased activity resulting from inflammation or acute tissue/nerve injury, or as ectopic input from subacute or chronically injured fibers and their neighboring noninjured primary afferents (nociceptor or non-nociceptor). The locus of central sensitization includes all CNS structures (ascending sensory relay pathways, the dorsal horn of the spinal cord, brainstem, thalamus, and cortex, as well as modulating circuits in the central gray matter of the brainstem) and all CNS processes in which activity in nociceptive somatosensory pathways can be initiated, facilitated, or prolonged through increases in excitability or reductions in inhibition. As a broad concept, this mechanism is the most important one for understanding and treating neuropathic pain. Treatment should therefore be directed at the source of the increased peripheral input or the process of CNS sensitization—preferably both. Direct injury to the CNS resulting from, for example, nerve root avulsion, spinal cord injury, multiple sclerosis, or thalamic stroke, can lead to spontaneous pathologic epileptiform-like discharges in projection circuits that are independent of peripheral input. In addition, there may be an exaggerated and sometimes clinically bizarre response (hyperpathia) to sensory inflow. As with central sensitization from peripheral nociceptive input, the abnormal CNS discharges can induce functional changes in other CNS structures. The pain pathology is central, and treatment is

2639

central, but sufficiently little is known about mechanisms that no specific therapies have emerged for true central pain. Removing peripheral input may temporarily reduce the pain but of course cannot return the CNS to its preinjury state. Although controversial, it is possible that intense and prolonged nociceptor input can induce functional and anatomic changes in CNS targets. Once established, these changes no longer require peripheral input for their maintenance. Temporarily removing peripheral input (e.g., by nerve block) does not alter the pain. Even though the original problem lies in the periphery, the refractoriness to treatment is conceptualized as centralization of the pain. This vague term is sometimes applied in a blanket manner to all patients whose pain fails to respond to the prescribed therapy, even when psychological and other factors are critically important for its continuation. For purposes of clarity, it is preferred to reserve the term central pain for those patients in whom the initiating injury is in the CNS. In those patients whose pain can be shown to be clearly stimulus evoked, the disorder requires a peripheral input. When allodynia, secondary hyperalgesia, and temporal stimulus-response mismatches are present, the mechanisms responsible are central, whether input dependent or autonomous. Differentiating autonomous central pain from central sensitization is difficult because they may present in identical ways, as an abnormal amplification of sensory responses. Similarly, identifying peripheral ectopic versus central spontaneous activity as the source of spontaneous pain is a problem because both may present as pain referred to an area of denervation. At the cellular and circuit level, increases in excitability may be produced by increases in excitatory transmitter release, enhanced responsiveness of postsynaptic excitatory transmitter receptors on the membrane, and direct depolarization of the postsynaptic membrane. Trafficking of receptors or ion channels to the synaptic membrane or post-translational changes in receptor/ion channel kinetics are not the only explanations for the maintenance of heightened excitability. Longer lasting changes may reflect transcription-dependent alterations in protein/peptide expression in primary afferents or postsynaptic neurons. Novel expression of transmitters and growth factors (e.g., in large myelinated sensory fibers) after peripheral nerve injury may enable changes in function. The most persistent form of increased excitability would be the result of changes in synaptic connectivity caused by axonal sprouting establishing novel excitatory inputs to projection neurons. Inhibitory neural activity can be phasic, activated in response to stimuli, and tonic. Inhibition occurs at both preand postsynaptic loci at all sites in the CNS, and can be mediated by local inhibitory interneurons as well as inputs from brainstem nuclei. Loss of inhibition, similar to increases in excitability, follows different time courses depending on the

2640

Diseases of the Peripheral Nervous System

responsible mechanism. Induction of apoptosis in inhibitory interneurons will produce an irreversible loss of inhibitory input. Treatment can be specifically targeted to the underlying mechanisms. These include reducing transmitter release by opiates and alpha-2-delta calcium channel–binding drugs, inhibiting postsynaptic excitatory receptors such as the N-methyl-D-aspartate (NMDA) or AMPA/kainate receptors, potentiating inhibitory transmitters by reducing transmitter uptake or by agonist administration, and by use-dependent sodium channel blockers. Many analgesic targets are re-regulated after nerve injury, including sodium and potassium channels, the alpha-2-delta calcium channel accessory protein, cannabinoid, and opiate receptors. Woolf, Bennett, Max, Koltzenburg, and others have proposed a change to a mechanism-based classification from the current classifications of pain based on disease, duration, and location.183,184 By identifying in individual patients which neural mechanisms are responsible for their pain, treatment specifically targeting those mechanisms can be employed. Such a shift in paradigm requires that we have ways of knowing what mechanisms are responsible for an individual patient’s pain.71 Substantial research is needed to develop and validate new pain assessment tools. The greatest progress along this line has come from the study and treatment of postherpetic neuralgia (PHN). The clinical mechanisms underlying PHN pain have been evaluated using thermal/mechanical quantitative sensation testing, electromyography, allodynia assessment, infrared thermography, response to provocative tests of epinephrine or histamine injection and capsaicin application, response to inactivation of cutaneous nociceptors with local anesthetic skin infiltration, and skin biopsy visualization of epidermal nerve fibers.11,13,31,60,61,110,115,117,134,135,142 Based on the results of these studies, it has been proposed that PHN pain spans a mechanistic spectrum between “irritable nociceptors” and deafferentationinduced changes in the CNS.49,138 These mechanisms are not mutually exclusive and likely coexist in the majority of PHN patients. Evidence of irritable nociceptors and the effects of deafferentation have been inferred in individual pain patients with other types of neuropathic pain using clinical examination and provocative tests. Even for investigators with access to sophisticated tools, the ability to infer dominant pain mechanisms in individual patients is relatively primitive. There remain many pain mechanisms proven relevant in preclinical studies for which there is no clinically available treatment. Clinical trials using target-selective drugs in subject cohorts who have had their pain dissected into its component mechanisms are needed. No pharmacotherapy trials reported to date have used detailed testing of this scope in the subject screening phase. For clinical practice to change from disease-based treatment to mechanism-based treatment, it will be necessary to show that the mechanism-based approach produces better results.

THE EVIDENCE BASE AND GENERALIZABILITY OF CLINICAL TRIAL RESULTS One of the very first controlled trials of medications for neuropathic pain was the trial by Watson and colleagues of the tricyclic antidepressant (TCA) amitriptyline for PHN.177 Following that came a series of studies of antidepressants carried out by Max and co-workers at the National Institutes of Health, most following a crossover design and reporting results only on those subjects completing both treatment periods.96–98 Each of these studies reported outcomes in fewer than 60 subjects, and most had fewer than 30 completing subjects. The publication of two randomized placebo-controlled trials of gabapentin for PHN and diabetic peripheral neuropathy (DPN) in 1998 represent the beginning of a marked shift in the evidence base for neuropathic pain therapy.8,130 Together, these two trials enrolled nearly 400 subjects, and the data analysis followed a conservative, modified intent-to-treat (ITT) approach. With such large numbers of subjects and the reporting of results in all drug-exposed subjects, the statistical power of each study was greatly increased. As emphasized by Moore and co-workers, sample size is extremely important.104 Since 1998, a large number of industry-sponsored controlled trials utilizing reasonably similar outcome measures and a similar approach to data analysis have been performed. This trend reflects the fact that many of the trials are part of registration programs by the manufacturers of the different drugs. Before labeling approval can be granted, placebo-controlled trials (with an auditable database) that demonstrate both safety and efficacy in large groups of subjects treated for a year or more must be performed. Access to the full database from each study for rigorous meta-analyses would greatly improve the evidence base for treatment of neuropathic pain. Unfortunately, most of these studies have been reported only in poster form at national and international meetings. Full reports in peerreviewed journals may take years to appear, so the data remain unavailable for rigorous independent analysis. For example, pregabalin, an investigational compound with a mechanism of action thought to be nearly identical to gabapentin, has been evaluated in placebo-controlled clinical trials in the United States and elsewhere enrolling more than 6000 subjects with disorders as diverse as PHN, DPN, chronic low back pain, osteoarthritis pain, and fibromyalgia.146 Only the first of the many individual clinical trials has been published in full form.45 Examples also abound with compounds already available in the marketplace. Topiramate, an anticonvulsant that acts on sodium channels and reduces excitatory amino acid neurotransmitter release, was the subject of multiple, negative, placebo-controlled trials for DPN totaling more than 1000 subjects.127 A few of the smaller studies with experimental compounds that failed

Mechanisms and Pharmacologic Management of Neuropathic Pain

to show analgesia have been published in full format, such as trials of a sodium channel blocker and an NMDA-glycine site antagonist,173,174 but many other “negative” trials of experimental compounds remain unpublished. In addition, as new compounds come into the marketplace, initial published data on their efficacy may come only from prospective open-label studies or anecdotal case reports. Randomized controlled trials may be years away, if they are performed at all. Although crossover trials may be more efficient than parallel-design studies, certain statistical assumptions need to be fulfilled to utilize data from the second treatment phase. This hurdle is not always met (see Moulin et al.105). Enriched-enrollment studies, in which subjects are selected based on factors beyond pain diagnosis and basic demographic variables, ensure a more homogeneous study sample at the expense of the generalizability of the results. In most cases, response to prior treatment is the selection factor.21,55 Compared to small single-center trials involving 20 to 40 subjects, large multicenter trials can establish statistical significance even when differences between placebo and active drug appear relatively small, raising the question of defining “clinically meaningful” as opposed to “statistically significant.” Taking advantage of the extraordinarily large database generated by the pregabalin trials, Farrar and colleagues have made substantial progress in defining what magnitude of pain intensity change is clinically meaningful.48,131 Multiple systematic reviews have been published in recent years.83,101,102,156 Older reports, including nearly all studies published before 1997, are small single-center trials. Although the number needed to treat (NNT) has been calculated for many of these trials, many frequently cited trials have not used a conservative ITT data analysis scheme. Some reported results only for completers, in part because of the crossover design used. This includes most of the TCA trials and the controlled-release oxycodone study in PHN.85,96,98,175 Small sample sizes makes the NNT figures a less reliable guide, and excluding dropouts from the data analysis very likely results in an overestimate of treatment benefit. There is no incentive or requirement for the pharmaceutical industry to evaluate the analgesic efficacy of compounds that are not labeled for pain and are no longer under patent protection (or are soon to lose patent protection). Without independently funded trials, it is unlikely we will ever see another study using, for example, mexiletine, fluoxetine, or valproic acid for neuropathic pain. If rigorous evidence-based medicine becomes the standard, only the newer, and more expensive, compounds will have an adequate evidence base. The true relative efficacy of different treatments for neuropathic pain will remain uncertain until large prospective trials directly compare drugs with different mechanisms of action. The majority of the published prospective controlled pharmacotherapy trials were performed in only two chronic

2641

neuropathic pain syndromes, DPN and PHN. Moreover, the FDA has approved medications for the treatment of only two specific neuropathic pain syndromes, trigeminal neuralgia (carbamazepine) and PHN (gabapentin, lidocaine patch 5%). Extrapolating trial results from one chronic neuropathic syndrome to another is speculative, but opioids, TCAs, and the anticonvulsant gabapentin have been the subject of controlled clinical trials in multiple types of neuropathic pain with generally similar results. How should the evidence base apply to complex regional pain syndrome type I, in which controlled trials of analgesic pharmacotherapy are lacking and the chronic pain is believed to be due to nervous system dysfunction without permanent injury to a nerve trunk?12,52 What about chronic neuropathic back pain (i.e., cervical and lumbar radiculopathic pain), probably the most prevalent chronic pain syndrome with a neuropathic component? There are no accepted diagnostic criteria for separating out the neuropathic component, but a combination of neuropathic, skeletal, and myofascial mechanisms likely accounts for most of the pain. Subgroup analysis of a randomized placebo-controlled trial suggested that those patients who had chronic radicular low back pain responded best to treatment with nortriptyline.5 Some types of neuropathic pain already have distinct treatment guidelines; for tic douloureux (trigeminal neuralgia), the algorithm emphasizes carbamazepine, phenytoin, and baclofen.90

ANTIDEPRESSANTS Although patients may find using drugs labeled only for depression to be stigmatizing, no medication in any category has proven unequivocally superior to TCAs for chronic neuropathic pain. TCAs were the first medication category proven effective in double-blind, placebo-controlled trials.177 Contrary to the assumptions of the initial trials, pain relief and relief of depression are independent effects.96,154 Essentially every blinded clinical trial of TCAs has found them efficacious.83,102 A wide variety of neuropathic pain disorders have been evaluated, including central poststroke pain and spinal cord injury pain.26,88 With TCAs, nearly all patients experience the anticholinergic side effects of dry mouth and constipation, and many also experience urinary hesitancy and blurred vision. Cognitive impairment, sedation, orthostatic hypotension, and sexual dysfunction are also common and poorly tolerated. TCAs can be lethal in an accidental or intentional overdose; the likelihood of successful suicide is 8- to 16-fold higher for TCAs compared to non-TCAs such as trazodone and fluoxetine.77 Non-TCAs are in widespread use for pain: selective serotonin reuptake inhibitors (SSRIs), antidepressants that alter both serotonergic (5-hydroxytryptamine [5-HT]) and noradrenergic (norepinephrine [NE]) neurotransmission,

2642

Diseases of the Peripheral Nervous System

and NE-selective antidepressants. Evidence for analgesic efficacy with the newer antidepressants continues to accrue, but only a small number of controlled studies have been conducted in neuropathic pain disorders.2,92,98,144 For the SSRIs, in a recent review including all types of pain and all types of literature reports, there were 19 publications dealing with fluoxetine, 3 with sertraline, 6 with paroxetine, 3 with fluvoxamine, 3 with citalopram, and 9 with trazodone. Except for paroxetine, controlled studies have not shown SSRIs to be superior to placebo for neuropathic pain. Direct comparisons of TCAs with non-TCAs have rarely been performed; the crossover trial comparing the TCA imipramine and the SSRI paroxetine for diabetic neuropathy pain by Sindrup and colleagues remains the most instructive.155 Paroxetine was equal to imipramine in the majority of subjects, inferior in 30% to 40%, and essentially never superior. Antidepressants that are either relatively NE selective (maprotiline, bupropion) or that affect both 5-HT and NE (venlafaxine and nefazodone) may be more effective analgesics than SSRIs. Bupropion, almost completely selective for NE, was found to be effective for mixed neuropathic pain in a small placebo-controlled crossover trial,147 while maprotiline appeared effective, but less so, when compared to a TCA for PHN.176 Full reports of controlled clinical trials for neuropathic pain have not yet appeared for nefazodone.79 Preclinical studies, human experimental pain data, anecdotal clinical reports, and controlled trial data support analgesic effects for venlafaxine.2,47,87,144,153,160,163 The presumed mechanisms underlying antidepressant analgesia, facilitation of descending serotonergic and noradrenergic pain-modulatory pathways, may not be the only mechanisms. TCAs, in addition to blocking 5-HT and NE reuptake, are relatively potent sodium channel blockers and may act as NMDA receptor blockers, and some have significant sympatholytic effects.36 Sodium channel–blocking effects should not be underestimated as a mechanism of action of TCAs (see next section); amitriptyline is comparable to bupivacaine as a nerve-blocking agent57 and is more potent at blocking use-dependent sodium channel currents than carbamazepine.19 To make matters even more complex, there is some evidence that analgesic effects of antidepressants involve opioid receptors, although the overall importance of this mechanism is uncertain. 120,145 Furthermore, venlafaxine possesses similarities to the analgesic tramadol.94 With the introduction of many non-TCAs in the past decade, it is possible that adequately powered clinical trials could find one or more drugs that are safer than TCAs and equally effective at relieving pain. However, the newer antidepressants should not be assumed to be completely safe, and withdrawal reactions have been described for essentially all antidepressants.18,62,187 In the

meantime, the following general treatment recommendations can be made: 1. If a TCA were tried at doses producing blood levels in the therapeutic range for depression, without pain relief, the available literature would suggest that other TCAs, SSRIs, and monoamine oxidase inhibitors are also likely to fail.154,155,179,181 2. If a TCA relieves the patient’s pain but the side effects prove intolerable, the non-TCAs should be tried in the hope of finding one that is analgesic with fewer side effects. 3. Starting with the newer non-TCAs and switching to a TCA only if pain is not relieved is a valid approach. 4. Patients with safety contraindications to TCAs or intolerable side effects at subtherapeutic doses of a TCA should be given a trial of a non-TCA before abandoning this drug category.

ANTICONVULSANTS Anticonvulsants are a broad category tied together only by their ability to suppress epileptic seizures. Anticonvulsants and related local anesthetic drugs have been used for decades to treat chronic pain. Systemic administration of procaine, and later lidocaine, has been in use since at least the 1930s to treat a variety of acute and chronic pain states. In his 1953 monograph, Bonica noted that there was anecdotal evidence for an analgesic effect of intravenous (IV) procaine in disorders as diverse as postoperative pain, burn pain, acute traumatic pain (such as fractures and sprains), chronic musculoskeletal pain and arthritis, acute herpes zoster, PHN, various neuralgias, and causalgia/reflex dystrophies.16 Central pain has also been evaluated in this way.6 Of the oral anticonvulsants, carbamazepine was demonstrated in 1976 to have dramatic efficacy in the treatment of trigeminal neuralgia.44 There is sufficient evidence to firmly establish anticonvulsants and local anesthetic drugs as specialized analgesics for neuropathic pain.50,101,137,158,167 Outside of trigeminal neuralgia, comparisons between anticonvulsants are rare, even in the preclinical literature.32 With the possible exception of gabapentin, neither anticonvulsants nor local anesthetics have clearly established analgesic efficacy for musculoskeletal or nociceptive pain.119 Anticonvulsants differ in their mechanism of action. Many drugs have in common an ability to block sodium channels in a use-dependent manner, but also exert their effect through non–sodium channel mechanisms. Other anticonvulsant drugs are highly effective in controlling pain, but seem to do so without blocking sodium channels. Proposed mechanisms of analgesic actions for non–sodium channel blocking drugs revolve around effects on sensitized central neurons, such as direct or indirect inhibition of the release of excitatory amino acids, blockade of neuronal

Mechanisms and Pharmacologic Management of Neuropathic Pain

calcium channels, and augmentation of CNS inhibitory pathways by increasing ␥-aminobutyric acid (GABA)-ergic transmission. Although little studied, interactions with opioid receptors do not appear to contribute significantly to analgesic effects of anticonvulsants.14

Clinical Use of Sodium Channel–Blocking Anticonvulsants A variety of anticonvulsant and local anesthetic drugs suppress abnormal discharge originating at nerve injury sites and associated DRGs via sodium channel blockade. These include the anticonvulsants carbamazepine, phenytoin, and lamotrigine; the antiarrhythmic mexiletine; and lidocaine and lidocaine-like local anesthetics.27 Each of these prevents the generation of spontaneous ectopic impulses at concentrations two to three orders of magnitude lower than are required to block normal impulse propagation. Because the process of ectopic impulse generation is so sensitive to sodium channel blockade, these drugs can be given systemically or regionally without fatal toxicity from failure of normal nerve conduction. Carbamazepine is approved by the FDA for treatment of trigeminal neuralgia. Like the local anesthetics lidocaine and mexiletine, carbamazepine reduces spontaneous activity in experimental neuromas.20 Success has been documented in trigeminal neuralgia and DPN, but not in PHN or central pain.23,88,108,129,143 Carbamazepine has been reported to be effective in a variety of other neuropathic pain disorders, but available studies are small and/or uncontrolled. The main drawbacks to carbamazepine are sedation, ataxia, the need for regular monitoring of hematologic and hepatic function, drug-drug interactions, and the rare occurrence of irreversible aplastic anemia. Oxcarbazepine is structurally similar to carbamazepine, is active in animal pain models, and except for syndrome of inappropriate secretion of antidiuretic hormone and dermatologic reactions, appears to have a more benign adverse event profile than carbamazepine.69,81 No controlled clinical trials of oxcarbazepine for neuropathic pain have been published. By blocking voltage-dependent sodium channels and suppressing peripherally generated ectopic impulse activity, lamotrigine has been shown to reduce central release of the excitatory transmitters glutamate and aspartate.30,164 Animal studies have demonstrated analgesic effects of lamotrigine.66,107 Clinical trial evidence in favor of lamotrigine is slowly accumulating.100 In a study of postoperative pain, 200-mg single doses of lamotrigine reduced both postoperative pain and analgesic requirements.17 A placebo-controlled study of 100 patients with a variety of neuropathic pain complaints showed no reduction in pain from a daily dose of 200 mg.99 In contrast, a maintenance dose of 400 mg lamotrigine was superior to placebo as an add-on therapy for trigeminal neuralgia.188 Placebo-controlled trials have been recently reported in

2643

DPN, central pain, and human immunodeficiency virus (HIV)-associated neuropathy.46,151,169 Doses up to 300 to 400 mg/day have been used anecdotally and in open trials with success in central pain and trigeminal neuralgia.25,64,91 Drawbacks specific to lamotrigine are relatively high frequencies of rash (including rare fatalities from StevensJohnson syndrome), and drug-drug interactions. Among the drugs marketed as antiarrhythmics, mexiletine has received the most study. Structurally similar to lidocaine, mexiletine is a local anesthetic and a class IB antiarrhythmic. Pain, dysesthesias, and paresthesias were reduced in patients with chronic painful DPN in a randomized, double-blind, placebo-controlled crossover study by Dejgard and colleagues.37 Relatively high doses of mexiletine were administered based on the subject’s weight. In studies of DPN and various nerve injuries using lower doses of 450 to 750 mg/day, the analgesic effect has been modest.28,111,157,186 Two recent studies of patients with HIV-related neuropathy using a relatively low dose of 600 mg/day failed to demonstrate an analgesic effect of mexiletine.78,80 Specific drawbacks to mexiletine are cardiac contraindications, worsening of arrhythmias, drug-drug interactions, upper gastrointestinal (GI) distress (common and frequently treatment limiting), and tremor. The antiarrhythmic flecainide is occasionally used as therapy for neuropathic pain, but this drug has significant proarrhythmic potential.70 Phenytoin is now used infrequently for pain. Uncontrolled studies reported phenytoin to be effective in the treatment of trigeminal neuralgia.68 Double-blind, placebocontrolled studies have demonstrated analgesia with phenytoin in DPN and Fabry’s disease, a small fiber neuropathy.29,89 Phenytoin, like lidocaine, can be given parenterally and can be used to provide immediate relief for patients who are having severe and frequent attacks of trigeminal neuralgia.1 Specific drawbacks to phenytoin are its complex kinetics, drug-drug interactions, and combined signs of neurotoxicity and cardiac conduction effects at very high blood levels. Topiramate has been used for seizure disorders since 1996, but only recently reported as an analgesic agent.9 Topiramate has been shown to reduce neuronal activity by voltage-dependent blockade of Na⫹ channels, enhancing GABA at GABAA receptors, and by antagonizing the kainate subtype of glutamate receptors. Specific drawbacks to topiramate are a higher incidence of sedation and psychomotor slowing, carbonic anhydrase inhibition, renal stones, and drug-drug interactions. Topiramate can produce narrowangle glaucoma and also has the unusual side effect of producing weight loss in 10% to 20% of patients.72,114 Unfortunately, the initial release of results from large-scale clinical trials for DPN pain did not indicate clinical efficacy.127 A somewhat similar anticonvulsant, zonisamide, has not been the subject of published controlled trials for neuropathic pain.

2644

Diseases of the Peripheral Nervous System

Anticonvulsants That Do Not Block Sodium Channels: Gabapentin, Valproic Acid, Levetiracetam Gabapentin was approved for marketing in the United States in 1995 as adjunctive therapy in the treatment of seizures. Gabapentin was developed as an analogue to the neurotransmitter GABA, but has since been shown not to interact with either GABAA or GABAB receptors.162 Because gabapentin is excreted unmetabolized, active metabolites cannot explain the analgesic effect in humans. GABA-ergic function could be potentiated without direct interaction with GABA sites by increasing the concentration of GABA in neuronal tissue through release of GABA from nerve terminals, enzyme effects, or decreased GABA breakdown.65,162 Another possible mechanism is mobilization of intracellular GABA via gabapentin-sensitive transporters. However, the best evidence to date comes from studies rank-ordering effects of gabapentin and related drugs on a specific gabapentin-binding protein found in brain and spinal cord. The alpha-2-delta binding site is a subunit of a voltage-gated calcium channel found in high density in the cerebral cortex, superficial dorsal horn, cerebellum, and hippocampus.56 Gabapentin binding is primarily postsynaptic and is resistant to neonatal capsaicin treatment and rhizotomy. GABA itself has no activity at this binding site, but gabapentin analogues (including pregabalin) that bind to the alpha-2-delta subunit show analgesic activity. Initial evidence of pain relief from gabapentin came from case reports, chart reviews, and open-label clinical trials in a variety of painful conditions.103 Two large multicenter, double-blind, placebo-controlled clinical trials reported in 1998 demonstrated that gabapentin at a target dose of 3600 mg/day reduced pain from PHN and from DPN. Both were analyzed using a conservative ITT analysis. In the study of PHN, pain was reduced by 33%, compared to a reduction of 8% with placebo.130,138 In the DPN study, pain was reduced by 39% with gabapentin, compared to 22% with placebo.8 At least eight published double-blind, placebo-controlled, randomized clinical trials of gabapentin show efficacy for chronic neuropathic pain. These studies examined patients with PHN, DPN, mixed neuropathic pain syndromes, phantom limb pain, GuillainBarré syndrome, and acute and chronic spinal cord injury pain.15,59,112,128,148,159 Gabapentin at daily dosages up to 3600 mg significantly reduces pain compared to placebo; improvements in sleep, mood, and quality of life were also demonstrated in some of the trials. In two trials of DPN and spinal cord injury pain with small sample sizes or relatively lower dosages, the evidence of efficacy was more limited. On the basis of the results of the two large randomized trials for PHN, the FDA approved gabapentin for the treatment of PHN at a target dose of 1800 mg/day. Based on a calculation of the NNT, a measure of efficacy,

gabapentin is comparable to TCAs for neuropathic pain. Gabapentin has also been found effective in postoperative pain and in suppressing experimentally induced hyperalgesia in healthy volunteers.41,42 In the United States, gabapentin is used more frequently than any other anticonvulsant for chronic pain. This popularity is due to ease of monitoring, a relatively low incidence of serious adverse events, and a perception of efficacy bolstered by evidence from placebo-controlled trials. Gabapentin has a short half-life and is absorbed via a saturable active transport mechanism. As with nearly all anticonvulsants, the most common side effects are dizziness, ataxia, and somnolence. Compared to other anticonvulsants, drug-drug interactions are not a problem. The dose must be adjusted in patients with renal failure, but no adjustment is necessary in patients with hepatic disorders because the drug is excreted unmetabolized. Valproic acid and divalproex sodium are structurally unrelated to any of the other anticonvulsants. Its analgesic mechanism of action is unknown, but valproic acid is known to increase GABA-ergic neurotransmission, increase brain GABA concentrations, and alter brain levels of excitatory amino acids.34 Valproic acid has proved useful in the prophylaxis of migraine in controlled clinical trials,150 but has been little studied for neuropathic pain. Valproate did not relieve central pain compared to placebo.43 In an openlabel study, Peiris et al. reported partial or complete control of trigeminal neuralgia in 9 of 20 patients with valproic acid in doses of up to 1200 mg/day.113 A therapeutic range of plasma levels for pain control has not been established with valproic acid. Side effects and potentially serious toxicity have greatly limited the use of valproic acid for chronic pain. The combination of sedation, GI side effects, hair loss (usually reversible), abnormal elevation of liver transaminases and ammonia, potentially fatal hepatotoxicity, inhibition of platelet aggregation, drug-drug interactions, and numerous other potential hematologic and nonhematologic effects make pretreatment screening and close follow-up mandatory. Levetiracetam has been the subject of one published small, open-label trial and one animal pain model study.3,136,140 The drug is well tolerated and requires minimal monitoring. The mechanism of action is unknown, but thought to be related to effects on neuronal calcium channels.

TOPICAL MEDICATIONS A vast number of compounds have been applied topically for pain control. A thorough review of all the topical agents tested for treatment of neuropathic pain is beyond the scope of this chapter.83,141 PHN has received far more study than any other neuropathic pain disorder. Topical capsaicin has been intensively studied, albeit with conflicting results.178,180

Mechanisms and Pharmacologic Management of Neuropathic Pain

Topical aspirin and nonsteroidal anti-inflammatory drugs have undergone limited study, with encouraging initial results but little in the way of long-term controlled treatment studies.35,82 Numerous drugs are available in topical form through compounding pharmacies, including clonidine, gabapentin, and ketamine. A cream containing the TCA doxepin is now available commercially. Topical lidocaine gel for PHN has received limited study.133 Controlled trial evidence indicates topical lidocaine patch is an effective treatment strategy for PHN through a direct effect of the local anesthetic and a “protective” effect of the patch vehicle on painfully sensitive skin.55,132,142 For both preparations, blood levels are well below the minimum for systemic toxicity.22 In 1999, the FDA approved lidocaine patches for the specific indication of PHN. Open-label studies suggest that lidocaine patches may have utility for neuropathic pain disorders other than PHN.39

OPIOIDS Opioid therapy remains controversial, but the evidence from controlled trials is now quite strong.51,189 Uncontrolled and mostly retrospective reports long ago suggested that opioids were effective as long-term therapy with a low risk of addiction.53,95,121,123,161,168,170,181 Portenoy and colleagues used the response to uncontrolled infusions to propose a continuum of opioid responsiveness in which neuropathic pain might require higher doses, but would still respond.122 Controlled evidence showing that neuropathic pain is responsive to IV opioid administration appeared in 1991. The first study, a three-session, double-blind, crossover trial, compared 1-hour-long IV infusions of 0.3 mg/kg of morphine (average total dose of 19.2 mg), lidocaine 5 mg/kg (average total dose of 316 mg), and saline placebo in 19 patients with established PHN.139 Preinfusion pain intensity visual analogue scale (VAS) ratings declined by 33% during the morphine sessions, compared to a 13% decline during the placebo sessions. The two-session crossover trial reported by Dellemijn and Vanneste included 50 patients with different types of pain, and compared the analgesic effect of IV fentanyl with either saline placebo or the benzodiazepine diazepam.38 Diazepam was included as a control agent in the belief that it would modulate the subject’s emotional experience of pain without producing intrinsic analgesic effects. Each infusion lasted a total of 5 hours: fentanyl at a rate of 5 ␮g/kg/h and diazepam at a rate of 0.2 ␮g/kg/h. Fentanyl reduced both the affective and sensory dimensions of pain, with a maximum 65% pain reduction with fentanyl compared to 15% to 20% for control infusions. Diazepam produced sedation that was nearly as great as with fentanyl, but was no more effective in relieving pain than saline placebo. The Dellemijn and Vanneste study is especially important because they specifically controlled for the sedating effects

2645

of opioids by including a benzodiazepine. Specific types of neuropathic pain have been studied, including pain from CNS injury. Kalman and co-workers studied 14 patients with well-defined multiple sclerosis and, using very strict criteria for morphine responsiveness, found only 4 patients to be morphine responders.76 Attal studied IV morphine for poststroke pain.7 Collectively, these studies provide compelling evidence in favor of a specific analgesic effect of opioids on both the intensity and the unpleasantness of neuropathic pain of peripheral nervous system origin, but pain from CNS injury appears to be less responsive. Five randomized, controlled trials of oral opioid therapy for neuropathic pain have been published. All reported benefit, making oral opioid therapy a firmly established treatment option for neuropathic pain. Watson and Babul described a crossover trial of twice-daily controlled-release oxycodone in 50 patients with PHN.175 Treatment periods were 4 weeks long, and the maximum dose was 30 mg twice a day. A total of 38 subjects completed the trial (dropout rate ⫽ 24%). In completing subjects, the mean total daily dose of oxycodone was 45 mg and pain intensity VAS during the last week of treatment was 35 mm, significantly lower than the 54 mm during treatment with placebo. Significantly more adverse events were reported during oxycodone therapy, particularly constipation, nausea, and sedation. Despite the greater reduction of pain during oxycodone therapy, there were no treatment differences on the Beck Depression Inventory or on any of the six scales of the Profile of Mood States. In another trial using controlledrelease oxycodone, patients with DPN reported significant improvements in pain, interference with activities, and sleep compared with placebo; the average dosage of oxycodone in this trial was 37 mg daily, with a range of 10 to 99 mg.58 Raja and colleagues compared controlled-release morphine with the TCA nortriptyline and a placebo for PHN in a complex three-period crossover design.125 Although pain reduction with morphine was superior to placebo and nearly superior to nortriptyline, patients in that study were much more likely to drop out during the opioid therapy phase. Controlled-release morphine titrated to a maximum dosage of 300 mg daily was superior to placebo in patients with phantom limb pain.67 In a double-blind, randomized study that compared two different dosage levels of levorphanol in patients with a variety of peripheral and central neuropathic pain syndromes, patients receiving the higher dosage reported significantly greater pain reduction, but there were no differences between the groups in mood, sleep, and interference with daily activities.136,140 Although patients with central poststroke pain were the least likely to report improvement and only 30% with this disorder completed the trial, levorphanol was better tolerated and somewhat more effective for other types of pain from CNS injury. Prescribing habits have changed in the past decade, although the true incidence of iatrogenic addiction remains unresolved.73,74 Opioids have numerous side

2646

Diseases of the Peripheral Nervous System

effects, most of which can be managed with dose adjustment, a bowel regimen, and in some cases antinauseants. Sedation and a feeling of cognitive impairment are quite variable, but usually improve over time. Importantly, measures of cognitive functioning were administered in two of the five longer term trials and oral opioid analgesics did not impair performance. For opioids, as well as for TCAs (and probably sedating anticonvulsants), there is an increased risk of hip fracture in the elderly.149 Because tolerance is a major concern of clinicians, more data need to be gathered prospectively on how important an issue this really is in treating patients with neuropathic pain. Although opioids are likely to remain a controversial therapy for chronic neuropathic pain, the evidence accumulated thus far suggests opioids have efficacy comparable to the best available antidepressants and anticonvulsants.10

TRAMADOL Tramadol is FDA-approved for use in the United States for “moderate to moderately severe” pain. This atypical drug weakly inhibits NE and 5-HT reuptake, but the primary metabolite shows low-affinity binding to μ opioid receptors. It remains unsettled how much μ opioid receptor binding contributes to overall pain reduction because tramadol analgesia is only partially reversible by naloxone.124 Two randomized, placebo-controlled trials for neuropathic pain have been reported. Harati and co-workers reported the results of a 6-week-long multicenter, parallel design study in 131 subjects with chronic neuropathic pain caused by distal symmetrical DPN.63 A total of 82 subjects completed the study (dropout rate ⫽ 37%), and took a mean dose of 210 mg/day of tramadol. There were more “adverse event” dropouts in the tramadol group, and more “lack of efficacy” dropouts in the placebo group. A conservative ITT data analysis was performed. Subjects treated with tramadol had significantly less pain and greater pain relief at all time points from day 14 of treatment to day 42 at the end of the treatment period. Pain intensity on a 5-point Likert scale was reduced from 2.5 to 1.4, compared to a change from 2.6 to 2.2 for those subjects assigned to placebo. Tramadol produced moderate relief, compared to slight relief with placebo. Sleep, current health perception, psychological distress, and overall role functioning scales were not improved with tramadol therapy. Sindrup and co-workers reported a crossover study in 45 patients with polyneuropathy of diverse etiologies.152 The total noncompletion rate was 25%, with 7 of 9 adverse event dropouts during tramadol treatment. Data analysis on the 34 subjects completing the trial showed that tramadol reduced pain by approximately 33% (compared to no change in pain during placebo) at a mean dose of 347 mg/day. Reductions in spontaneous and touch-evoked pain were both significant and correlated with each other during tramadol treatment.

THE RESEARCH LANDSCAPE AND CHANGING CLINICAL PRACTICE The best medications in each of the categories discussed above provided satisfactory pain relief, defined as a 50% or greater reduction in pain intensity or moderate pain relief, in approximately 50% to 60% of clinical trial subjects. The probability of complete relief with any single drug is about 20%. In clinical practice, combining medications from the different categories is common, but very few data have been prospectively gathered on the overall success of polypharmacy regimens. Unfortunately, there are no prospectively proven polypharmacy algorithms. With the exception of the recent crossover study by Raja and colleagues, few studies have directly compared the different medication classes.125 Clinical trials recruit patients who are motivated to try new therapies and can meet strict medical, psychological, and diagnostic entry criteria. It is not known if requiring study participants to withdraw from most or all of their other analgesics before beginning study medication affects outcomes. In clinical trials, subjects who voluntarily discontinue participation usually do so because of a lack of relief, unpleasant side effects, or both; that is, they failed to achieve meaningful benefit. Therefore, the more conservative ITT analysis is preferred because it includes all drug-exposed subjects in the outcome analysis. ITT analyses were rare in the analgesic trials of the 1980s and early 1990s, but are now the norm, especially in pharmaceutical industry–sponsored multicenter trials of new analgesics. The great majority of new compounds in development fail to progress beyond Phase II or Phase III. Many of the clinical trials remain unpublished, with a few exceptions.173,174 No category appears to be completely spared, including sodium channel blockers, nerve growth factor, various strategies for blocking the NMDA receptor complex, and NK-1 antagonists. The high failure rate makes it uncertain how many new compounds will enter clinical development programs for neuropathic pain, although sales of gabapentin have demonstrated that the size of the market is large enough to justify aggressive development strategies. Human experimental pain models have a role in the study of pain mechanisms and in the testing of new analgesic drugs. Several human experimental models simulate clinical pain conditions by intentionally inducing temporary cutaneous hyperalgesia using either thermal or chemical stimulation or a combination of both. Such models offer information about temporary inflammatory changes and sensitization of both the peripheral nervous system and the CNS.4,118,126,166 The capsaicin injection, burn injury, and heat/capsaicin sensitization models are examples of skinbased models; models of muscle and visceral hyperalgesia are also used. To a limited extent each has been used to assess the analgesic activity of both experimental drugs and

Mechanisms and Pharmacologic Management of Neuropathic Pain

drugs in clinical use.116 For obvious ethical reasons there are no human models of experimental nerve injury. An IV drug infusion of a highly selective drug may be a useful way to predict the long-term therapeutic response to the analogous oral agent. The response to IV lidocaine partially predicts response to oral local anesthetics for arrhythmia control.106 However, the evidence that the lidocaine infusion test is useful in predicting analgesic response to analogous oral medications is very limited.54,93 Because lidocaine is also a potent anticonvulsant medication when given systemically, central actions independent of modulation of ectopic impulse generation in dysfunctional peripheral nerve fibers need to be considered. The applicability of this general approach remains limited by the number of selective IV drugs that have closely analogous oral compounds.

SUMMARY AND CONCLUSIONS The number of proven treatment options are steadily expanding, but there remain a large number of patients suffering from unrelieved chronic neuropathic pain. None of the currently available options provides moderate or better relief in more than 50% to 60% of clinical trial patients. Our understanding of pain mechanisms is growing in scope and sophistication. Animal models as well as human models of pain are available for testing drugs before embarking on difficult and costly clinical trials. Compared to a decade ago, clinical trials methodology has greatly improved. Large multicenter controlled clinical trials are now the rule, resulting in an evidence base that will soon allow more valid comparisons across drug classes. There remains a very substantial lag time for the results of large multicenter trials to be published. Hopefully, all large trials will be published, including those with negative outcomes. The tools available for research into clinical mechanisms are also improving, although a proven mechanism-based pharmacotherapy algorithm remains many years away.

REFERENCES 1. Albert, H. H.: Infusion therapy of acute trigeminal neuralgia using phenytoin i.v. MMW Muench. Med. Wochenschr. 120:529, 1978. 2. Ansari, A.: The efficacy of newer antidepressants in the treatment of chronic pain: a review of current literature. Harv. Rev. Psychiatry 7:257, 2000. 3. Ardid, D., et al.: Levetiracetam (Keppra), a new antiepileptic drug, is effective in neuropathic but not acute pain models in rats. Neurology 56(Suppl. 3):A350, 2001. 4. Arendt-Nielsen, L.: Induction and assessment of experimental pain from human skin, muscle, and viscera. In T.J. Jensen T. S., and Wiesenfeld-Hallin, Z. (eds.): Proceedings of the 8th World Congress on Pain. Seattle, IASP Press, p. 393, 1997.

2647

5. Atkinson, J. H., Slater, M. A., Williams, R. A., et al.: A placebo-controlled randomized clinical trial of nortriptyline for chronic low back pain. Pain 76:287, 1998. 6. Attal, N., Gaude, V., Brasseur, L., et al.: Intravenous lidocaine in central pain: a double-blind, placebo-controlled, psychophysical study. Neurology 54:564, 2000. 7. Attal, N., Guirimand, G., Brasseur, L., et al.: Effects of IV morphine in central pain: a randomized placebo-controlled study. Neurology 58:554, 2002. 8. Backonja, M., Beydoun, A., Edwards, K. R., et al.: Gabapentin for the symptomatic treatment of painful neuropathy in patients with diabetes mellitus: a randomized controlled trial. JAMA 280:1831, 1998. 9. Bajwa, Z. H., Sami, N., Warfield, C. A., and Wootton, J.: Topiramate relieves refractory intercostal neuralgia. Neurology 52:1917, 1999. 10. Ballantyne, J. C., and Mao, J.: Opioid therapy for chronic pain. N. Engl. J. Med. 349:1943, 2003. 11. Baron, R., Haendler, G., and Schulte, H.: Afferent large fiber polyneuropathy predicts the development of postherpetic neuralgia. Pain 73:231, 1997. 12. Baron, R., Levine, J. D., and Fields, H. L.: Causalgia and reflex sympathetic dystrophy: does the sympathetic nervous system contribute to the generation of pain? Muscle Nerve 22:678, 1999. 13. Baron, R., and Saguer, M.: Postherpetic neuralgia: are C-nociceptors involved in signaling and maintenance of tactile allodynia? Brain 116:1477, 1993. 14. Blackburn-Munro, G., Dickinson, T., and FleetwoodWalker, S. M.: Non-opioid actions of lamotrigine within the rat dorsal horn after inflammation and neuropathic nerve damage. Neurosci. Res. 39:385, 2001. 15. Bone, M., Critchley, P., and Buggy, D. J.: Gabapentin in postamputation phantom limb pain: a randomized, double-blind, placebo-controlled, cross-over study. Reg. Anesth. Pain Med. 27:481, 2002. 16. Bonica, J. J.: The Management of Pain, with Special Emphasis on the Use of Analgesic Block in Diagnosis, Prognosis, and Therapy. Philadelphia, Lea & Febiger, p. 1533, 1953. 17. Bonicalzi, V., Canavero, S., Cerutti, F., et al.: Lamotrigine reduces total postoperative analgesic requirement: a randomized double-blind, placebo-controlled pilot study. Surgery 122:567, 1997. 18. Boyd, I. W.: Venlafaxine withdrawal reactions. Med. J. Aust. 169:91, 1998. 19. Brau, M. E., Dreimann, M., Olschewski, A., et al.: Effect of drugs used for neuropathic pain management on tetrodotoxin-resistant Na(⫹) currents in rat sensory neurons. Anesthesiology 94:137, 2001. 20. Burchiel, K. J.: Carbamazepine inhibits spontaneous activity in experimental neuromas. Exp. Neurol. 102:249, 1988. 21. Byas-Smith, M. G., Max, M. B., Muir, J., and Kingman, A.: Transdermal clonidine compared to placebo in painful diabetic neuropathy using a two-stage ‘enriched enrollment’ design. Pain 60:267, 1995. 22. Campbell B. J., Rowbotham M. C., Davies P., et al.: Systemic absorption of topical lidocaine in normal volunteers, patients with postherpetic neuralgia, and patients with acute herpes zoster. J. Pharm. Sci. 91:1343, 2002.

2648

Diseases of the Peripheral Nervous System

23. Campbell, F. G., Graham, J. G., and Zilkha, K. J.: Clinical trial of carbamazepine (Tegretol) in trigeminal neuralgia. J. Neurol. Neurosurg. Psychiatry 29:265, 1966. 24. Campbell, J. N., Raja, S. N., Meyer, R. A., and Mackinnon, S. E.: Myelinated afferents signal the hyperalgesia associated with nerve injury. Pain 32:89, 1988. 25. Canavero, S., and Bonicalzi, V.: Lamotrigine control of central pain. Pain 68:179, 1996. 26. Cardenas, D. D., Warms, C. A., Turner, J. A., et al.: Efficacy of amitriptyline for relief of pain in spinal cord injury: results of a randomized controlled trial. Pain 96:365, 2002. 27. Catterall, W. A.: Common modes of drug action on Na ⫹ channels: local anaesthetics, antiarrhythmics, and anticonvulsants. Trends Pharmacol. Sci., p. 57, 1987. 28. Chabal, C., Jacobson, L., Mariano, A., et al.: The use of oral mexiletine for the treatment of pain after peripheral nerve injury. Anesthesiology 76:513, 1992. 29. Chadda, V. S., and Mathur, M. S.: Double blind study of the effects of diphenylhydantoin sodium on diabetic neuropathy. J. Assoc. Physicians India 26:403, 1978. 30. Cheung, H., Kamp, D., and Harris, E.: An in vitro investigation of the action of lamotrigine on neuronal voltage-activated sodium channels. Epilepsy Res. 13:107, 1992. 31. Choi, B., and Rowbotham, M. C.: Effect of adrenergic receptor activation on post-herpetic neuralgia pain and sensory disturbances. Pain 69:55, 1997. 32. Christensen, D., Gautron, M., Guilbaud, G., and Kayser, V.: Effect of gabapentin and lamotrigine on mechanical allodynialike behaviour in a rat model of trigeminal neuropathic pain. Pain 93:147, 2001. 33. Cline, M. A., Ochoa, J., and Torebjork, H. E.: Chronic hyperalgesia and skin warming caused by sensitized C nociceptors. Brain 112:621, 1989. 34. Cutrer, F. M., Limmroth, V., and Moskowitz, M. A.: Possible mechanisms of valproate in migraine prophylaxis. Cephalalgia 17:93, 1997. 35. DeBenedittis, G., Besana, F., and Lorenzetti, A.: A new topical treatment for acute herpetic neuralgia and postherpetic neuralgia: the aspirin/diethyl ether mixture. An open-label study plus a double-blind controlled clinical trial. Pain 48:383, 1992. 36. Deffois, A., Fage, D., and Carter, C.: Inhibition of synaptosomal veratridine-induced sodium influx by antidepressants and neuroleptics used in chronic pain. Neurosci. Lett. 220:117, 1996. 37. Dejgard, A., Petersen, P., and Kastrup, J.: Mexiletine for treatment of chronic painful diabetic neuropathy. Lancet 1:9, 1988. 38. Dellemijn, P., and Vanneste, J.: Randomized double-blind active-placebo-controlled crossover trial of intravenous fentanyl in neuropathic pain. Lancet 349:753, 1997. 39. Devers, A., and Galer, B. S.: Topical lidocaine patch relieves a variety of neuropathic pain conditions: an open-label study. Clin. J. Pain 16:205, 2000. 40. Devor, M., and Seltzer, Z.: Pathophysiology of damaged nerves in relation to chronic pain. In Walls, P. D., and Melzack, R. (eds.): Textbook of Pain, 4th ed. Edinburgh, Churchill Livingstone, p. 129, 1999. 41. Dirks, J., Fredensborg, B. B., Christensen D., et al.: A randomized study of the effects of single-dose gabapentin versus

42.

43.

44.

45.

46.

47.

48.

49.

50.

51. 52.

53.

54.

55.

56.

57.

58.

placebo on postoperative pain and morphine consumption after mastectomy. Anesthesiology 97:560, 2002. Dirks, J., Petersen, K. L., Rowbotham, M. C., and Dahl, J. B.: Gabapentin suppresses cutaneous hyperalgesia following heat/capsaicin sensitization. Anesthesiology 97:102, 2002. Drewes, A. M., Andreasen, A., and Poulsen, L. H.: Valproate for treatment of chronic central pain after spinal cord injury: a double-blind cross-over study. Paraplegia 32:565, 1994. Dunsker, S. B., and Mayfield, F. H.: Carbamazepine in the treatment of the flashing pain syndrome. J. Neurosurg. 45:49, 1976. Dworkin, R. H., Corbin, A. E., Young, J. P., et al.: Pregabalin for the treatment of postherpetic neuralgia: a randomized, placebo-controlled trial. Neurology 60:1274, 2003. Eisenberg, E., Lurie, Y., Braker, C., et al.: Lamotrigine reduces painful diabetic neuropathy: a randomized, controlled study. Neurology 57:505, 2001. Enggaard, T. P., Klitgaard, N. A., Gram, L. F., et al.: Specific effect of venlafaxine on single and repetitive experimental painful stimuli in humans. Clin. Pharmacol. Ther. 69:245, 2001. Farrar, J. T., Young, J. P. Jr., LaMoreaux, L., et al.: Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain 94:149, 2001. Fields, H. L., Rowbotham, M., and Baron, R.: Postherpetic neuralgia: irritable nociceptors and deafferentation. Neurobiol. Dis. 5:209, 1998. Fields, H. L., Rowbotham, M. C., and Devor, M.: Excitability blockers: anticonvulsants and low concentration local anesthetics in the treatment of chronic pain. In Dickenson, A., and Besson, J. M. (eds.): Handbook of Experimental Pharmacology. Berlin, Springer-Verlag, p. 93, 1997. Foley, K. M.: Opioids and chronic neuropathic pain. N. Engl. J. Med. 348:1279, 2003. Forouzanfar, T., Köke, A. J., van Kleef, M., and Weber, W. E.: Treatment of complex regional pain syndrome type 1. Eur. J. Pain 6:105, 2002. France, R. D., Urban, B. J., and Keefe, F. J.: Long-term use of narcotic analgesics in chronic pain. Soc. Sci. Med. 19:1379, 1984. Galer, B. S., Harle, J., and Rowbotham, M. C.: Response to intravenous lidocaine infusion predicts subsequent response to oral mexiletine: a prospective study. J. Pain Symptom Manage. 12:161, 1996. Galer, B. S., Rowbotham, M. C., Perander, J., and Friedman, E.: Topical lidocaine patch relieves postherpetic neuralgia more effectively than a vehicle topical patch: results of an enriched enrollment study. Pain 80:533, 1999. Gee, N. S., Brown, J. P., Dissanayake, V. U., et al.: The novel anticonvulsant drug, gabapentin (Neurontin), binds to the alpha2delta subunit of a calcium channel. J. Biol. Chem. 271:5768, 1996. Gerner, P., Mujtaba, M., Sinnott, C. J., and Wang, G. K.: Amitriptyline versus bupivacaine in rat sciatic nerve blockade. Anesthesiology 94:661, 2001. Gimbel, J. S., Richards, P., and Portenoy, R. .K.: Controlledrelease oxycodone for pain in diabetic neuropathy: a randomized controlled trial. Neurology 60:927, 2003.

Mechanisms and Pharmacologic Management of Neuropathic Pain 59. Gorson, K. C., Schott, C., Herman, R., et al.: Gabapentin in the treatment of painful diabetic neuropathy: a placebocontrolled, double-blind, crossover trial. J. Neurol. Neurosurg. Psychiatry 66:251, 1999. 60. Haanpää, M., Dastidar, P., Weinberg, A., et al.: CSF and MRI findings in patients with acute herpes zoster. Neurology 51:1405, 1998. 61. Haanpää, M., Laippala, P., and Nurmikko, T.: Pain and somatosensory dysfunction in acute herpes zoster. Clin. J. Pain 15:78, 1999. 62. Haddad, P.: Newer antidepressants and the discontinuation syndrome. J. Clin. Psychiatry 58(Suppl. 7):17; discussion 22, 1997. 63. Harati, Y., Gooch, C., Swenson, M., et al.: Double-blind randomized trial of tramadol for the treatment of the pain of diabetic neuropathy. Neurology 50:1842, 1998. 64. Harbison, J., Dennehy, F., and Keating, D.: Lamotrigine for pain with hyperalgesia. Ir. Med. J. 90:56, 1997. 65. Honmou, O., Kocsis, J. D., and Richerson, G. B.: Gabapentin potentiates the conductance increase induced by nipecotic acid in CA1 pyramidal neurons in vitro. Epilepsy Res. 20:193, 1995. 66. Hunter, J. C., Gogas, K. R., Hedley, L. R., et al.: The effect of novel anti-epileptic drugs in rat experimental models of acute and chronic pain. Eur. J. Pharmacol. 324:153, 1997. 67. Huse, E., Larbig, W., Flor, H., and Birbaumer, N.: The effect of opioids on phantom limb pain and cortical reorganization. Pain 90:47, 2001. 68. Iannone, A., Baker, A. B., and Morrell, F.: Dilantin in the treatment of trigeminal neuralgia. Neurology 8:126, 1958. 69. Ichikawa, K., Koyama, N., Kiguchi, S., et al.: Inhibitory effect of oxcarbazepine on high-frequency firing in peripheral nerve fibers. Eur. J. Pharmacol. 420:119, 2001. 70. Ichimata, M., Ikebe, H., Yoshitake, S., et al.: Analgesic effects of flecainide on postherpetic neuralgia. Int. J. Clin. Pharmacol. Res. 21:15, 2001. 71. Jensen, T. S., and Baron, R.: Translation of symptoms and signs into mechanisms in neuropathic pain. Pain 102:1, 2003. 72. Jones, M. W.: Topiramate—safety and tolerability. Can. J. Neurol. Sci. 25:S13, 1998. 73. Joranson, D. E., Gilson, A. M., Dahl, J. L., and Haddox, J. D.: Pain management, controlled substances, and state medical board policy: a decade of change. J. Pain Symptom Manage. 23:138, 2002. 74. Joranson, D. E., Ryuan, K. M., Gilson, A. M., and Dahl, J. L.: Trends in medical use and abuse of opioid analgesics. JAMA 283:1710, 2000. 75. Kajander, K. C., Wakisaka, S., and Bennett, G. J.: Spontaneous discharge originates in the dorsal root ganglion at the onset of a painful peripheral neuropathy in the rat. Neurosci. Lett. 138:225, 1992. 76. Kalman, S., Österberg, A., Sörensen, J., et al.: Morphine responsiveness in a group of well-defined multiple sclerosis patients: a study with i.v. morphine. Eur. J. Pain 6:69, 2002. 77. Kapur, S., Mieczkowski, T., and Mann, J. J.: Antidepressant medications and the relative risk of suicide attempt and suicide [see comments]. JAMA 268:3441, 1992. 78. Kemper, C. A., Kent, G., Burton, S., and Deresinski, S. C.: Mexiletine for HIV-infected patients with painful peripheral neuropathy: a double-blind, placebo-controlled, crossover

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93. 94. 95.

96.

2649

treatment trial. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 19:367, 1998. Khouzam, H. R.: The antidepressant nefazodone: a review of its pharmacology, clinical efficacy, adverse effects, dosage, and administration. J. Psychosoc. Nurs. Ment. Health Serv. 38:20, 2000. Kieburtz, K., Simpson, D., Yiannoutsos, C., et al.: A randomized trial of amitriptyline and mexiletine for painful neuropathy in HIV infection. AIDS Clinical Trial Group 242 Protocol Team. Neurology 51:1682, 1998. Kiguchi, S., Ichikawa, K., and Kojima, M.: Suppressive effects of oxcarbazepine on tooth pulp-evoked potentials recorded at the trigeminal spinal tract nucleus in cats. Clin. Exp. Pharmacol. Physiol. 28:169, 2001. King, R. B.: Topical aspirin in chloroform and the relief of pain due to herpes zoster and postherpetic neuralgia. Arch. Neurol. 50:1046, 1993. Kingery, W. S.: A critical review of controlled clinical trials for peripheral neuropathic pain and complex regional pain syndromes [see comments]. Pain 73:123, 1997. Kirk, E. J.: Impulses in dorsal spinal nerve rootlets in cats and rabbits arising from dorsal root ganglia isolated from the periphery. J. Comp. Neurol. 155:165, 1974. Kishore-Kumar, R., Max, M. B., Schafer, S. C., et al.: Desipramine relieves postherpetic neuralgia. Clin. Pharmacol. Ther. 47:305, 1990. LaMotte, R. H., Shain, C. N., Simone, D. A., and Tsai, E. F.: Neurogenic hyperalgesia: psychophysical studies of underlying mechanisms. J. Neurophysiol. 66:190, 1991. Lang, E., Hord, A. H., and Denson, D.: Venlafaxine hydrochloride (Effexor) relieves thermal hyperalgesia in rats with an experimental mononeuropathy. Pain 68:151, 1996. Leijon, G., and Boivie, J.: Central post-stroke pain—a controlled trial of amitriptyline and carbamazepine. Pain 36:27, 1989. Lockman, L. A., Hunninghake, D. B., Krivit, W., and Desnick, R. J.: Relief of pain of Fabry’s disease by diphenylhydantoin. Neurology 23:871, 1973. Loeser, J. D.: Cranial neuralgias. In Loeser, J. D., Butler, S. H., Chapman, C. R., and Turk, D. C. (eds.): Bonica’s Management of Pain, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 855, 2001. Lunardi, G., Leandri, M., Albano, C., et al.: Clinical effectiveness of lamotrigine and plasma levels in essential and symptomatic trigeminal neuralgia. Neurology 48:1714, 1997. Lynch, M. E.: Antidepressants as analgesics: a review of randomized controlled trials. J. Psychiatry Neurosci. 26:30, 2001. Mao, J., and Chen, L. L.: Systemic lidocaine for neuropathic pain relief. Pain 87:7, 2000. Markowitz, J. S., and Patrick, K. S.: Venlafaxine-tramadol similarities. Med. Hypotheses 51:167, 1998. Maruta, T., Swanson, D., and Finlayson, R.: Drug abuse and dependency in patients with chronic pain. Mayo Clin. Proc. 54:241, 1979. Max, M. B., Culnane, M., Schafer, S. C., et al.: Amitriptyline relieves diabetic neuropathy pain in patients with normal or depressed mood. Neurology 37:589, 1987.

2650

Diseases of the Peripheral Nervous System

97. Max, M. B., Kishore-Kumar, R., Schafer, S. C., et al.: Efficacy of desipramine in painful diabetic neuropathy: a placebocontrolled trial. Pain 45:3, 1991. 98. Max, M. B., Lynch, S. A., Muir, J., et al.: Effects of desipramine, amitriptyline, and fluoxetine on pain in diabetic neuropathy. N. Engl. J. Med. 326:1250, 1992. 99. McCleane, G.: 200 mg daily of lamotrigine has no analgesic effect in neuropathic pain: a randomised, double-blind, placebo controlled trial. Pain 83:105, 1999. 100. McCleane, G. J.: Lamotrigine in the management of neuropathic pain: a review of the literature. Clin. J. Pain 16:321, 2000. 101. McQuay, H., Carroll, D., Jadad, A. R., et al.: Anticonvulsant drugs for management of pain: a systematic review. BMJ 311:1047, 1995. 102. McQuay, H. J., Tramer, M., Nye, B. A., et al.: A systematic review of antidepressants in neuropathic pain. Pain 68:217, 1996. 103. Mellegers, M. A., Furlan, A. D., and Mailis, A.: Gabapentin for neuropathic pain: systematic review of controlled and uncontrolled literature. Clin. J. Pain 17:284, 2001. 104. Moore, R. A., Gavaghan, D., Tramer, M. R., et al.: Size is everything: large amounts of information are needed to overcome random effects estimating direction and magnitude of treatment effects. Pain 78:209, 1998. 105. Moulin, D. E., Iezzi, A., Amireh, R., et al.: Randomised trial of oral morphine for chronic non-cancer pain. Lancet 347:143, 1996. 106. Murray, K. T., Barbey, J. T., Kopelman, H. A., et al.: Mexiletine and tocainide: a comparison of antiarrhythmic efficacy, adverse effects, and predictive value of lidocaine testing. Clin. Pharmacol. Ther. 45:553, 1989. 107. Nakamura-Craig, M., and Follenfant, R. L.: Effect of lamotrigine in the acute and chronic hyperalgesia induced by PGE2 and in the chronic hyperalgesia in rats with streptozotocin-induced diabetes. Pain 63:33, 1995. 108. Nicol, C. F.: A four year double-blind study of Tegretol in facial pain. Headache 9:54, 1969. 109. Nordin, M., Nystrom, B., Wallin, U., and Hagbarth, K. E.: Ectopic sensory discharges and paresthesiae in patients with disorders of peripheral nerves, dorsal roots and dorsal columns. Pain 20:231, 1984. 110. Oaklander, A. L., Romans, K., Horasek, S., et al.: Unilateral postherpetic neuralgia is associated with bilateral sensory neuron damage. Ann. Neurol. 44:789, 1998. 111. Oskarsson, P., Ljunggren, J. G., and Lins, P. E.: Efficacy and safety of mexiletine in the treatment of painful diabetic neuropathy. The Mexiletine Study Group. Diabetes Care 20:1594, 1997. 112. Pandey, C. K., Bose, N., Garg, G., et al.: Gabapentin for the treatment of pain in Guillain-Barré syndrome: a doubleblinded, placebo-controlled, crossover study. Anesth. Analg. 95:1719, 2002. 113. Peiris, J. B., Perera, G. L., Devendra, S. V., and Lionel, N. D.: Sodium valproate in trigeminal neuralgia. Med. J. Aust. 2:278, 1980. 114. Perucca, E.: A pharmacological and clinical review on topiramate, a new antiepileptic drug. Pharmacol. Res. 35:241, 1997.

115. Petersen, K. L., Fields, H. L., Brennum, J., et al.: Capsaicin activation of ‘irritable’ nociceptors in post-herpetic neuralgia. Pain 88:125, 2000. 116. Petersen, K. L., Jones, B., Segredo, V., et al.: The effect of remifentanil on pain and secondary hyperalgesia associated with the heat/capsaicin sensitization model in healthy volunteers. Anesthesiology 94:15, 2001. 117. Petersen, K. L., Rice, F., Suess, F., et al.: Relief of postherpetic neuralgia by surgical removal of painful skin. Pain 98:119, 2002. 118. Petersen, K. L., and Rowbotham, M. C.: A new human experimental pain model: the heat/capsaicin sensitization model. Neuroreport 10:1511, 1999. 119. Petersen, K. L., and Rowbotham, M. C.: Will ion channel blockers be useful for management of non-neuropathic pain? J. Pain 1(3, Suppl. 1):26, 2000. 120. Pick, C. G., Paul, D., Eison, M. S., and Pasternak, G. W.: Potentiation of opioid analgesia by the antidepressant nefazodone. Eur. J. Pharmacol. 211:375, 1992. 121. Portenoy, R. K., and Foley, K. M.: Chronic use of opioid analgesics in non-malignant pain: report of 38 cases. Pain 25:171, 1986. 122. Portenoy, R. K., Foley, K. M., and Inturrisi, C. E.: The nature of opioid responsiveness and its implications for neuropathic pain: new hypotheses derived from studies of opioid infusions. Pain 43:273, 1990. 123. Porter, J., and Jick, H.: Addiction rare in patients treated with narcotics [letter]. N. Engl. J. Med. 302:123, 1980. 124. Raffa, R. B., Friderichs, E., Reimann, W., et al: Opioid and nonopioid components independently contribute to the mechanism of action of tramadol, an ‘atypical’ opioid analgesic. J. Pharmacol. Exp. Ther. 260:275, 1992. 125. Raja, S. N., Haythornthwaite, J. A., Pappagallo, M., et al: Opioids versus antidepressants in postherpetic neuralgia: a randomized, placebo-controlled trial. Neurology 59:1015, 2002. 126. Raja, S. N., Meyer, R. A., and Campbell, J. N.: Peripheral mechanisms of somatic pain. Anesthesiology 68:571, 1988. 127. Reuters News Service: Johnson & Johnson’s pain drug trial unsuccessful. September 17, 2001. 128. Rice, A. S., and Maton, S.: Gabapentin in postherpetic neuralgia: a randomised, double blind, placebo controlled study. Pain 94:215, 2001. 129. Rockliff, B. W., and Davis, E. H.: Controlled sequential trials of carbamazepine in trigeminal neuralgia. Arch. Neurol. 15:129, 1966. 130. Rowbotham, M., Harden, N., Stacey, B., et al: Gabapentin for the treatment of postherpetic neuralgia: a randomized controlled trial. JAMA 280:1837, 1998. 131. Rowbotham, M. C.: What is a ‘clinically meaningful’ reduction in pain? Pain 94:131, 2001. 132. Rowbotham, M. C., Davies, P. S., Verkempinck, C., and Galer, B. S.: Lidocaine patch: double-blind controlled study of a new treatment method for post-herpetic neuralgia. Pain 65:39, 1996. 133. Rowbotham, M. C., and Fields, H. L.: Post-herpetic neuralgia: the relation of pain complaint, sensory disturbance, and skin temperature. Pain 39:129, 1989.

Mechanisms and Pharmacologic Management of Neuropathic Pain 134. Rowbotham, M. C., and Fields, H. L.: Topical lidocaine reduces pain in post-herpetic neuralgia. Pain 38:297, 1989. 135. Rowbotham, M. C., and Fields, H. L.: The relationship of pain, allodynia and thermal sensation in post-herpetic neuralgia. Brain 119:347, 1996. 136. Rowbotham, M. C., Manville, N., and Ren, J.: Pilot tolerability and efficacy study of levetiracetam for post-herpetic neuralgia. Neurology 61:866, 2003. 137. Rowbotham, M. C., and Petersen, K. L.: Anticonvulsants and local anesthetic drugs. In Loeser, J. L. (ed.): Bonica’s Management of Pain, 3rd ed. Philadelphia, Lippincott Williams & Wilkins, p. 1727, 2000. 138. Rowbotham, M. C., Petersen, K. L., and Fields, H. L.: Post-herpetic neuralgia: more than one disorder? Pain Forum 7:231, 1998. 139. Rowbotham, M. C., Reisner-Keller, L. A., and Fields, H. L.: Both intravenous lidocaine and morphine reduce the pain of postherpetic neuralgia. Neurology 41:1024, 1991. 140. Rowbotham, M. C., Twilling, L., Davies, P. S., et al: Oral opioid therapy for chronic peripheral and central neuropathic pain. N. Engl. J. Med. 348:1223, 2003. 141. Rowbotham, M. C., and Watson, C. P. N.: Topical agents for post-herpetic neuralgia. In Watson, C. P. N., and Gershon, A. A. (eds.): Herpes Zoster and Postherpetic Neuralgia, 2nd ed. Amsterdam, Elsevier, p. 231, 2001. 142. Rowbotham, M. C., Yosipovitch, G., Connolly, M. K., et al: Cutaneous innervation density in the allodynic form of postherpetic neuralgia. Neurobiol. Dis. 3:205, 1996. 143. Rull, J. A., Quibrera, R., Gonzalez-Millan, H., and Lozano Castaneda, O.: Symptomatic treatment of peripheral diabetic neuropathy with carbamazepine (Tegretol): double blind crossover trial. Diabetologia 5:215, 1969. 144. Sawynok, J., Esser, M. J., and Reid, A. R.: Antidepressants as analgesics: an overview of central and peripheral mechanisms of action. J. Psychiatry Neurosci. 26:21, 2001. 145. Schreiber, S., Backer, M. M., and Pick, C. G.: The antinociceptive effect of venlafaxine in mice is mediated through opioid and adrenergic mechanisms. Neurosci. Lett. 273:85, 1999. 146. Selak, I.: Pregabalin (Pfizer). Curr. Opin. Invest. Drugs 2:828, 2001. 147. Semenchuk, M. R., Sherman, S., and Davis, B.: Doubleblind, randomized trial of bupropion SR for the treatment of neuropathic pain. Neurology 57:1583, 2001. 148. Serpell, M. G., for the Neuropathic Pain Study Group: Gabapentin in neuropathic pain syndromes: a randomized, double-blind, placebo-controlled trial. Pain 99:557, 2002. 149. Shorr, R. I., Griffin, M. R., Daugherty, J. R., and Ray, W. A: Opioid analgesics and the risk of hip fracture in the elderly: codeine and propoxyphene. J. Gerontol. 47:M111, 1992. 150. Silberstein, S. D.: Divalproex sodium in headache: literature review and clinical guidelines. Headache 36:547, 1996. 151. Simpson, D. M., Olney, R., McArthur, J. C., et al.: A placebo-controlled trial of lamotrigine for painful HIVassociated neuropathy. Neurology 54:2115, 2000. 152. Sindrup, S. H., Andersen, G., Madsen, C., et al.: Tramadol relieves pain and allodynia in polyneuropathy: a randomised, double-blind, controlled trial. Pain 83:85, 1999.

2651

153. Sindrup, S. H., Bach, F. W., Madsen, C., et al.: Venlafaxine versus imipramine in painful polyneuropathy: a randomized, controlled trial. Neurology 60:1284, 2003. 154. Sindrup, S. H., Brosen, K., and Gram, L. F.: Antidepressants in pain treatment: antidepressant or analgesic effect? Clin. Neuropharmacol. 15:636A, 1992. 155. Sindrup, S. H., Brosen, K., and Gram, L. F.: The mechanism of action of antidepressants in pain treatment: controlled cross-over studies in diabetic neuropathy. Clin. Neuropharmacol. 15:380A, 1992. 156. Sindrup, S. H., and Jensen, T. S.: Pharmacologic treatment of pain in polyneuropathy. Neurology 55:915, 2000. 157. Stracke, H., Meyer, U. E., Schumacher, H. E., and Federlin, K.: Mexiletine in the treatment of diabetic neuropathy. Diabetes Care 15:1550, 1992. 158. Swerdlow, M., and Cundill, J. G.: Anticonvulsant drugs used in the treatment of lancinating pain: a comparison. Anaesthesia 36:1129, 1981. 159. Tai, Q., Kirshblum, S., Chen, B., et al.: Gabapentin in the treatment of neuropathic pain after spinal cord injury: a prospective, randomized double-blind, crossover trial. J. Spinal Cord Med. 25:100, 2002. 160. Tasmuth, T., Hartel, B., and Kalso, E.: Venlafaxine in neuropathic pain following treatment of breast cancer. Eur. J. Pain 6:17, 2002. 161. Taub, A.: Opioid analgesics in the treatment of chronic intractable pain of non-neoplastic origin. In Kitihata, L. M., and Collins, J. D. (eds.): Narcotic Analgesics in Anesthesiology. Baltimore, Williams & Wilkins, p. 199, 1982. 162. Taylor, C. P., Gee, N. S., Su, T. Z., et al.: A summary of mechanistic hypotheses of gabapentin pharmacology. Epilepsy Res. 29:233, 1998. 163. Taylor, K., and Rowbotham, M. C.: Venlafaxine hydrochloride and chronic pain. West. J. Med. 165:147, 1996. 164. Teoh, H., Fowler, L. J., and Bowery, N. G.: Effect of lamotrigine on the electrically-evoked release of endogenous amino acids from slices of dorsal horn of the rat spinal cord. Neuropharmacology 34:1273, 1995. 165. Torebjork, H. E., Lundberg, L. E., and LaMotte, R. H.: Central changes in processing of mechanoreceptive input in capsaicin-induced secondary hyperalgesia in humans. J. Physiol. (Lond.) 448:765, 1992. 166. Treede, R. D., Meyer, R. A., Raja, S. N., and Campbell, J. N.: Peripheral and central mechanisms of cutaneous hyperalgesia. Prog. Neurobiol. 38:397, 1992. 167. Tremont-Lukats, I. W., Megeff, C., and Backonja, M. M.: Anticonvulsants for neuropathic pain syndromes: mechanisms of action and place in therapy. Drugs 60:1029, 2000. 168. Urban, B. J., France, R. D., Steinberger, E. K., et al.: Long-term use of narcotic/antidepressant medication in the management of phantom limb pain. Pain 24:191, 1986. 169. Vestergaard, K., Andersen, G., Gottrup, H., et al.: Lamotrigine for central poststroke pain: a randomized controlled trial. Neurology 56:184, 2001. 170. Wall, P. D.: Neuropathic pain [editorial]. Pain 43:267, 1990. 171. Wall, P. D., and Devor, M.: Sensory afferent impulses originate from dorsal root ganglia as well as from the periphery in normal and nerve injured rats. Pain 17:321, 1983.

2652

Diseases of the Peripheral Nervous System

172. Wall, P. D., and Gutnick, M.: Properties of afferent nerve impulses originating from a neuroma. Nature 248:740, 1974. 173. Wallace, M. S., Rowbotham, M., Bennett, G. J., et al.: A multicenter, double-blind, randomized, placebo-controlled crossover evaluation of a short course of 4030W92 in patients with chronic neuropathic pain. J. Pain 3:227, 2002. 174. Wallace, M. S., Rowbotham, M., Katz, N., et al.: A doubleblind, placebo-controlled trial of a glycine antagonist in neuropathic pain. Neurology 59:1694, 2002. 175. Watson, C. P., and Babul, N.: Efficacy of oxycodone in neuropathic pain: a randomized trial in postherpetic neuralgia. Neurology 50:1837, 1998. 176. Watson, C. P., Chipman, M., Reed, K., et al.: Amitriptyline versus maprotiline in postherpetic neuralgia: a randomized, double-blind, crossover trial. Pain 48:29, 1992. 177. Watson, C. P., Evans, R. J., Reed, K., et al.: Amitriptyline versus placebo in postherpetic neuralgia. Neurology 32:671, 1982. 178. Watson, C. P., Tyler, K. L., Bickers, D. R., et al.: A randomized vehicle-controlled trial of topical capsaicin in the treatment of postherpetic neuralgia. Clin. Ther. 15:510, 1993. 179. Watson, C. P., Vernich, L., Chipman, M., and Reed, K.: Nortriptyline versus amitriptyline in postherpetic neuralgia: a randomized trial. Neurology 51:1166, 1998. 180. Watson, C. P. N.: Topical capsaicin as an adjuvant analgesic. J. Pain Symptom Manage. 9:425, 1994.

181. Watson, C. P. N., Evans, R. J., Watt, V. R., and Birkett, N.: Postherpetic neuralgia: 208 cases. Pain 35:289, 1988. 182. Woolf, C. J.: Excitability changes in central neurons following peripheral damage. In Willis, W. D. J. (ed.): Hyperalgesia and Allodynia: the Bristol-Myers-Squibb Symposium on Pain Research. New York, Raven Press, p. 221, 1992. 183. Woolf, C. J., Bennett, G. J., Doherty, M., et al.: Towards a mechanism-based classification of pain? [editorial]. Pain 77:227, 1998. 184. Woolf, C. J., and Decosterd, I.: Implications of recent advances in the understanding of pain pathophysiology for the assessment of pain in patients. Pain Suppl. 6:S141, 1999. 185. Woolf, C. J., and Mannion, R. J.: Neuropathic pain: aetiology, symptoms, mechanisms and management. Lancet 35:1959, 1999. 186. Wright, J. M., Oki, J. C., and Graves, L. 3rd: Mexiletine in the symptomatic treatment of diabetic peripheral neuropathy. Ann. Pharmacother. 31:29, 1997. 187. Zajecka, J., Tracy, K. A., and Mitchell, S.: Discontinuation symptoms after treatment with serotonin reuptake inhibitors: a literature review. J. Clin. Psychiatry 58:291, 1997. 188. Zakrzewska, J. M., Chaudhry, Z., Nurmikko, T. J., et al.: Lamotrigine (Lamictal) in refractory trigeminal neuralgia: results from a double-blind placebo controlled crossover trial. Pain 73:223, 1997. 189. Zochodne, D. W., and Max, M. B.: An old acquaintance: opioids and neuropathic pain. Neurology 60:894, 2003.

119 Management of Autonomic Failure PHILLIP A. LOW AND WOLFGANG SINGER

Neurogenic Orthostatic Hypotension Definition Epidemiology Causes Clinical Manifestations Course and Prognosis Pathophysiology and Pathogenesis Treatment Nonpharmacologic Management Pharmacologic Management

Treatment for Periods of Orthostatic Decompensation Hospital Management of Severe Orthostatic Hypotension Treatment of Early Morning Orthostatic Hypotension Treatment of Postprandial Orthostatic Hypotension Treatment of Nocturnal Hypertension

The autonomic nervous system is complex in its organization and function, and diffuse in its distribution. Because the autonomic nervous system innervates all organs, the manifestations of dysautonomia are diverse and myriad. There are also numerous diseases of the autonomic nervous system with lesions at the level of the brain or spinal cord or peripheral nerve. This chapter must therefore be selective and incomplete. We have chosen a detailed focus on the management of orthostatic hypotension with a briefer description of gastrointestinal and genitourinary autonomic failure. Additional information is provided in Chapters 12, 14, and 15. Chapters 10, 11, and 13 deal with closely related subjects.

Neurogenic Orthostatic Hypotension DEFINITION Neurogenic orthostatic hypotension is defined as orthostatic hypotension that results from a lesion of autonomic neural pathways: central, preganglionic, or postganglionic. By this definition orthostatic hypotension that results from pharmacologic agents (such as hypotensive agents) and hypovolemia is excluded.

Gastrointestinal and Genitourinary Dysautonomia Bladder Dysfunction Impotence Gastrointestinal Symptoms Gastroparesis Chronic Intestinal Pseudo-obstruction Constipation Diarrhea Fecal Incontinence

We define orthostatic hypotension as a fall in systolic blood pressure (BP) greater than or equal to 30 mm Hg at 1 minute of standing up or on head-up tilt.80 This definition is more conservative than the definition provided by a Consensus Conference committee convened in 1995, sponsored by the American Autonomic Society and cosponsored by the American Academy of Neurology. This committee defined orthostatic hypotension as “a reduction of systolic blood pressure of at least 20 mm Hg or diastolic blood pressure of at least 10 mm Hg within 3 minutes of standing up.”81 The use of a tilt table in the head-up position at an angle of at least 60 degrees is an alternative. The committee recommended that the confounding variables of food ingestion, time of day, state of hydration, ambient temperature, recent recumbency, postural deconditioning, hypertension, medications, gender, and age be considered. Orthostatic hypotension may be symptomatic or asymptomatic. If the patient has symptoms suggestive of, but does not have, documented orthostatic hypotension, repeated measurements of BP should be performed. The values chosen by the committee are reasonable screening values but are associated with a 5% false-positive rate. A value of 30 mm Hg fall in systolic BP would reduce the frequency of false positives to 1%.56 A standing time of 3 minutes does not add much in diagnostic yield compared to 1 minute. In a recent study of 2653

2654

Diseases of the Peripheral Nervous System

Table 119–1. Symptom-Grade of Orthostatic Intolerance Grade 0: No orthostatic symptoms.* Grade 1: Orthostatic symptoms are infrequent and do not restrict activities of daily living. Grade 2: Frequent orthostatic symptoms developing at least once a week. Some limitation in activities of daily living. Grade 3: Orthostatic symptoms develop on most occasions. Able to stand ⱖ 1 min on most occasions. Limitation in most of activities of daily living. Grade 4: Symptoms consistently develop on orthostasis. Able to stand ⬍ 1 min on most occasions. Syncope/presyncope is common if patient attempts to stand.

hypotension) for some autonomic disorders that are best estimates based on current data. For adult diabetic patients (combined type 1 and type 2 diabetes mellitus), the percentage of patients with orthostatic hypotension evaluated from 1987 to 1997 was 10%. The mean age of the Rochester diabetic patient cohort over this decade was 60.6 ⫾ 11.7 years. In a population-based study, Bower et al.7 reported an average annual incidence rate (new cases per 100,000 person-years) for ages 50 to 99 years of 3.0 for multiple system atrophy (MSA) in Olmsted County, Minnesota. Schrag et al.,82 in a cross-sectional estimate of the prevalence of MSA in London, England, reported a prevalence of 4.4 per 100,000.

*Syncope, dizziness, visual disturbances, or neck pain, relieved on lying flat.

CAUSES 69 patients with orthostatic hypotension, we reported that 66 patients developed orthostatic hypotension by 1 minute, with the remaining three patients developing orthostatic hypotension by 2 minutes.34 More importantly, these patients should have an autonomic laboratory study to confirm the presence, severity, and distribution of adrenergic failure. It is possible to generate a formal grading scale (Table 119–1).53 The grade is based on the components of 1. Frequency and severity of symptoms 2. Standing time before onset of symptoms and of presyncope 3. Influence on activities of daily living 4. BP data

EPIDEMIOLOGY The prevalence of orthostatic hypotension is not known with certainty. If the orthostatic hypotension associated with aging is included, then the prevalence is quite high in subjects older than 70 years of age. Table 119–2 provides prevalences (as prevalence in the population and as percentage of the affected population with orthostatic Table 119–2. Estimated Prevalence of Orthostatic Hypotension in Different Autonomic Disorders Disorder

Prevalence

Reference

Aging Diabetes Other autonomic neuropathies Multiple system atrophy

14%–20% 10%* 10–50/100,00

Low53

Pure autonomic failure

10–30/100,000

4–15/100,000

Estimated Fearnley and Lees31 Schrag et al.82 Estimated

*Prevalence for adult combined type 1 and type 2 diabetes mellitus for the Rochester Diabetic Cohort for 1997 (principal investigator: P. J. Dyck).

There are many causes of neurogenic orthostatic hypotension.58 The most important are MSA, pure autonomic failure (PAF), diabetic autonomic neuropathy, and autoimmune, including paraneoplastic autonomic neuropathies. In some practices, tetraplegia is a common cause of orthostatic hypotension. In a prospective study of 90 consecutive patients referred for suspected orthostatic hypotension, evaluated in the Mayo Autonomic Reflex Laboratory, and confirmed to have orthostatic hypotension, the diagnoses were PAF: 33% MSA: 26% Autonomic neuropathy: 17% Diabetic autonomic neuropathy: 14% Miscellaneous: 8% Olivopontocerebellar atrophy: 2%

CLINICAL MANIFESTATIONS Orthostatic BP varies with many factors. Patients with orthostatic hypotension will be asymptomatic a lot of the time and symptomatic at least some of the time. To adequately characterize the severity of orthostatic hypotension, it is necessary to define orthostatic BP and its relationship to orthostatic stressors. It is also important to recognize the full range of orthostatic symptoms. The symptoms of orthostatic hypotension vary by age. Table 119–3 provides a distribution of symptoms based on a prospective study of 90 consecutive patients with confirmed orthostatic hypotension.59 Lightheadedness is common, as expected. However, difficulty in concentrating and thinking is present in about half the patients. In patients over the age of 70 years, this may be the most common symptom and, while subtle, seriously impairs quality of life. Palpitations, tremulousness, anxiety, and nausea are symptoms of autonomic overactivity and occur in patients who have only partial autonomic failure,

2655

Management of Autonomic Failure

Table 119–3. Symptoms of Orthostatic Intolerance Lightheadedness (dizziness) Weakness or tiredness Cognitive (thinking/concentrating) Blurred vision Tremulousness Vertigo Pallor Anxiety Palpitations Clammy feeling Nausea

88% 72% 47% 47% 38% 37% 31% 29% 26% 19% 18%

typically seen in the autonomic neuropathies and in younger patients. Symptoms are typically worse in the early AM, after meals, with a rise in core temperature, with prolonged standing, and with activity. The early AM severity of orthostatic hypotension relates to the nocturnal diuresis present in many of these patients.3 Postprandial worsening is almost the rule, occurring within 30 minutes of a meal and lasting about an hour. Patients will commonly recognize that they have more orthostatic symptoms following a hot bath, after being in a hot tub, or on a hot day. Indeed, any stress that results in vasodilatation of skin or muscle vessels will worsen symptoms. Patients who get up in the middle of the night out of a warm bed are vasodilated and have worse orthostatic hypotension. Similarly, symptoms may be worse after ingestion of alcohol because of vasodilatation. With physical activity sufficient to cause muscle vasodilatation, orthostatic hypotension is also worse. Most symptoms are due to cerebral hypoperfusion. When orthostatic hypotension is severe and sustained, syncope will occur. Syncope after diagnosis tends to be relatively uncommon, as patients learn to recognize the symptoms of orthostatic hypotension and take corrective steps. On examination, orthostatic hypotension is confirmed. Patients with partial preservation of baroreflexes may have a marked tachycardia, indicating that the vagal component of the baroreflex is still intact. There are prominent oscillations in BP at the baroreflex loop frequency (about 0.1 Hz), with some modest improvement in BP over time. With generalized and more complete autonomic failure, orthostatic hypotension is associated with a fixed heart rate, indicating that both the cardiovagal and peripheral adrenergic limbs of the baroreflex have failed. Patients will also note that certain maneuvers will reduce orthostatic hypotension.23 For instance, patients have less orthostatic hypotension when they pace than when they stand perfectly still. Muscle contraction results in a pumping action, increasing venous return. They also notice that certain physical maneuvers such as bending forward, crossing their legs in a scissors movement, and

toe-rise will alleviate symptoms. The application of these physical countermaneuvers to reduce standing BP is covered below under Nonpharmacologic Management.

COURSE AND PROGNOSIS Prognosis depends on the specific disorders. Patients with classic MSA have a median survival of about 7 years from the time of diagnosis of the disease.55,104 However, Wenning et al.104 reported a median survival of 9.5 years, calculated by Kaplan-Meier analysis. Similar results have been reported,79 and the sporadic olivopontocerebellar variety has been suggested to have a longer survival than the striatonigral variety.79 The differences in survival reported by different investigators likely relate to the criteria used to define MSA. The downhill course is marked by increasing rigidity, urinary incontinence, and sometimes marked stridor, which may require tracheotomy. Death in MSA is frequently due to respiratory obstruction or failure after worsening rigidity, akinesia, and bladder disorder. With the appreciation of a spectrum of severities, an attempt has been made to relate the severity and distribution of autonomic and nonautonomic involvement to outcome. We reviewed the clinical and autonomic features of all patients with extrapyramidal and cerebellar disorders studied in the Mayo Autonomic Reflex Laboratory from 1983 to 1989.80 Orthostatic BP reduction, percentage of anhidrosis on thermoregulatory sweat test, quantitative sudomotor axon reflex test, forearm response, and heart rate response to deep breathing strongly regressed with severity of clinical involvement. The severity and distribution of autonomic failure at the time of first evaluation was predictive of a greater rate of progression 2 years later. Saito et al.79 came to the same conclusion. We concluded that the earlier and the more severe the involvement of the autonomic nervous system, and to a lesser extent the striatonigral system, the poorer the prognosis. Information on the clinical features, progression, and outcome in PAF is quite limited. Some patients with PAF have continued relatively symptom free for many years with standing BPs in the region of 80 mm Hg. The natural history of PAF is that of a slow progression taking place over some 10 to 15 years.55 The percentage of PAF that evolves into MSA is probably about 10%. The development of orthostatic hypotension worsens the prognosis. Ewing et al.28 reported a mortality rate of 50% at 2 1/2 years in patients with symptomatic diabetic autonomic neuropathy. However, these patients had longstanding clinical autonomic neuropathy and died of renal failure. Subsequent studies suggested that autonomic failure worsens prognosis, but less dismally than was originally thought.40 Based on the clinical and laboratory features of 229 patients with primary systemic amyloid seen at Mayo Clinic, Kyle and Greipp reported that median survival

2656

Diseases of the Peripheral Nervous System

from the time of diagnosis for patients with peripheral neuropathy, carpal tunnel syndrome, orthostatic hypotension, and cardiac failure was 60, 45, 9.5, and 6.5 months, respectively.49 The Mayo Clinic experience is that subacute autoimmune autonomic neuropathy runs a chronic debilitating course with significant residual deficits in the majority of patients. Patients seem to improve substantially over the first year, followed by a slower rate of improvement over the subsequent 4 years.88 Overall, approximately one in three patients makes a good functional recovery.

PATHOPHYSIOLOGY AND PATHOGENESIS The maintenance of postural normotension without an excessive heart rate increment depends on an adequate blood volume and the integration of many reflex and humoral systems in several key vascular beds, including the striated muscle, splanchnic-mesenteric, and cerebrovascular beds. An adequate blood volume is essential. Hypovolemia will regularly cause orthostatic hypotension, even if vascular reflexes are intact. Hypovolemia can also be relative. Denervation decreases vascular tone and increases vascular capacity. Patients with adrenergic failure will be relatively hypovolemic, although their plasma volume is normal. These patients can improve orthostatic intolerance if their plasma volume is expanded, hence the importance of volume expansion in the treatment of orthostatic hypotension. Reduced red cell mass, or the normocytic, normochromic anemia of chronic autonomic failure, will also aggravate orthostatic hypotension. Correcting anemia with erythropoietin (Epo) will improve orthostatic intolerance.4,44 Two sets of baroreflexes, the arterial (or high pressure) and venous (or low pressure), are mainly responsible for the reflex control of blood pressure and the circulation. When systemic pulse pressure or mean arterial pressure falls, baroreceptors are unloaded in the carotid sinus and aortic arch.2 These are high-pressure arterial baroreceptors. Afferent traffic from the carotid sinus travels via the glossopharyngeal nerve, and that from the carotid arch via the vagus nerve, to synapse in the nucleus of the tractus solitarius. From this structure, cardiovagal fibers travel via polysynaptic fibers to the nucleus ambiguus and dorsal motor nucleus of the vagus and thence as the vagus nerve to the sinoatrial node. Sympathetic function is regulated via the rostroventrolateral nucleus of the medulla to the intermediolateral column of the thoracic cord to provide sympathetic innervation to the heart and periphery (arterioles and venules).48 In addition to arterial baroreceptors, there are also low-pressure venous baroreceptors. The effective stimulus is a reduction in central venous pressure (i.e., responsive to changes in volume). Cardiopulmonary

receptors in the heart and lungs send mainly nonmyelinated vagal fibers to the nucleus of the tractus solitarius. The central pathways and efferents are the same as for arterial baroreceptors. The splanchnic-mesenteric capacitance bed is a largevolume, low-resistance system of great importance in the maintenance of postural normotension in humans. It constitutes 25% to 30% of total blood volume.78 Unlike muscle veins, the splanchnic veins have an abundance of smooth muscle and a rich sympathetic innervation. The mesenteric capacitance bed is markedly responsive to both arterial and venous baroreflexes. Venoconstriction is ␣adrenergic receptor mediated.93 The nerve supply to the mesenteric bed is largely from the greater splanchnic nerve, with its cell body in the intermediolateral column (mainly T4 to T9) and synapses at the celiac ganglion, from whence postganglionic adrenergic fibers go on to supply effector cells. Much clinical and research evidence supports the importance of the splanchnic outflow in the maintenance of postural normotension in humans. Postural hypotension occurs regularly when bilateral splanchnic neurectomy is performed, whereas neither bilateral lumbar sympathectomy nor cardiac denervation alone will cause postural hypotension.105,107 In patients with complete spinal cord lesions, postural hypotension becomes most pronounced when the splanchnic outflow is affected (above T6). Abnormalities in the splanchnic autonomic outflow have been found in human diabetic neuropathy, indicating that preganglionic fibers can be affected.60 Cerebral vasoregulation is important in ensuring adequate and stable flow to the brain in spite of changing systemic BP. The maintenance of constant blood flow in spite of variations in BP is termed autoregulation.89 Within a mean BP range of approximately 50 to 150 mm Hg, a change in blood pressure results in an insignificant change in cerebral perfusion. Previous studies12,24,26 in patients with orthostatic hypotension have demonstrated an expansion of the autoregulated range at both the upper and lower limits, so that cerebral perfusion remained relatively constant with the patient in the supine position (when supine hypertension might be present) and in response to standing (when orthostatic hypotension occurs).

TREATMENT There are four goals in the treatment of orthostatic hypotension (Table 119–4). The first is to improve orthostatic BP without excessive supine hypertension. There are three important consequences of autonomic and especially baroreflex failure: orthostatic hypotension, supine hypertension, and loss of diurnal variation in BP. Patients with neurogenic orthostatic hypotension have an excessive

Management of Autonomic Failure

2657

Table 119–4. Goals in the Treatment of Orthostatic Hypotension

Table 119–5. Patient Education in Management of Orthostatic Hypotension

1. Improve orthostatic BP without excessive supine hypertension. 2. Improve standing time (minimum of 2 min). 3. Relieve orthostatic symptoms. 4. Improve ability to perform activities of daily living.

1. Maintenance of postural normotension and abnormal mechanisms responsible for neurogenic orthostatic hypotension 2. Appropriate use of a BP-symptom log 3. Recognition of orthostatic stressors and their minimization 4. Training in appropriate physical countermaneuvers 5. Ability to manage orthostatic hypotension in all circumstances, including management of a. Stable orthostatic hypotension b. AM worsening c. Postprandial hypotension d. Periods of orthostatic decompensation e. Supine hypertension

fall in BP on standing and an excessively high BP while recumbent. Also, in contrast to normal subjects (who have significantly lower BP while asleep), patients with neurogenic orthostatic hypotension may have higher nocturnal BP. It is always possible to relieve symptoms of orthostatic hypotension, but it is much more difficult to do so without unacceptable supine hypertension, because patients with generalized autonomic failure have baroreflex failure, with the loss of postural regulation of BP as a regular component of such failure. A reasonable practical goal is a regimen that relieves symptoms most of the day with a supine BP that does not usually exceed 180/110 mm Hg. Patients with neurogenic orthostatic hypotension will have greater fluctuations than normal (loss of baroreflexes or “buffer nerves”) and will often have supine hypertension. The value we cite is quite acceptable because patients are taught to avoid lying flat. The second goal is to improve standing time. If patients are able to remain on their feet for 2 minutes, they can usually be taught to extend that time by resorting to physical countermaneuvers or other “tricks.” The third goal is to relieve orthostatic symptoms. The fourth goal, related to the second goal, is to improve the patient’s ability to perform orthostatic activities of daily living. Patients with asymptomatic orthostatic hypotension do not require treatment. However, that statement needs to be qualified by the recognition that most patients with orthostatic hypotension will have symptoms at some time. The orthostatic stresses may be time of day (early AM), Aa meal, a rise in core temperature, physical activity, or reduced salt or fluid intake. Older patients may become symptomatic after a period of bed rest or on taking a diuretic. It is necessary to recognize the patient with iatrogenic orthostatic hypotension. Common causative drugs include antihypertensive drugs and calcium channel entry blocking drugs. Angiotensinconverting enzyme inhibitors are widely used in hypertensive and diabetic patients. Insulin, levodopa, and tricyclic antidepressants can also cause vasodilatation and orthostatic hypotension in predisposed subjects. The first-line treatment of orthostatic hypotension is patient education (Table 119–5) and standard treatment (Table 119–6). Patient education and nonpharmacologic treatment of orthostatic hypotension should take precedence over pharmacologic treatment.

Nonpharmacologic Management Patient Education Patient education is critically important, and there is still insufficient emphasis on its implementation. The twin elements of switching to patient-centered decision making and employing a dynamic model of management of orthostatic hypotension require detailed patient education (see Table 119–5). Patients should understand in simple terms the maintenance of postural normotension and its practical implications (importance of blood volume, activating the humoral mechanism by postural training). They need to understand and recognize the orthostatic stressors, their pathophysiology in simple terms, and their management. For instance, the patient who knows that blood pools in the abdomen and pelvis after a meal and BP will fall might have a siesta. The patient who uses sildenafil needs to know that it drops the BP and he should remain in bed for a couple of additional hours if possible. Once the implications are explained, patients have much improved ability to recognize and manage the worsening orthostatic hypotension caused by these stressors. Suggestions on some modifications in activities of daily living include how to handle early AM orthostatic hypotension and postprandial orthostatic hypotension. In a patient-centered management plan, the patient will come up with additional useful modifications. Table 119–6. Standard Initial Treatment of Orthostatic Hypotension 1. 2. 3. 4. 5. 6.

High-salt diet (10–20 g/d) High fluid intake (20–40 oz/d); coffee or tea beneficial Elevating head of bed 4 inches Correcting anemia Maintaining postural stimuli Physical countermaneuvers

2658

Diseases of the Peripheral Nervous System

Patients should be shown how to keep a BP and symptom log. This is extremely useful when there is a change in management, when there is worsening in symptoms, and in the fine-tuning of BP management. Oftentimes it is uncertain whether this worsening is related to worsening orthostatic hypotension or some other mechanism. The patient or preferably spouse or other support person should use an automated sphygmomanometer to take BP measurements supine and after standing for 1 minute. By jotting down symptoms and corresponding BPs, the patient will gain insights into the relationship of symptoms and BP and the role of orthostatic stressors. Recordings (supine and after 1 minute of standing up) should be taken • On awakening

• • • •

After a meal During a time of maximal orthostatic tolerance During a time of poor orthostatic tolerance Before and 1 hour after medication

Patients need to know the sodium content of common food items and be educated on the significance of a high sodium intake, because there is a close relationship between salt intake and BP. The salt intake should be between 150 and 250 mEq of sodium (10 to 20 g of salt). Patients need to be on a high-potassium diet because of their high sodium intake, especially if they are also taking fludrocortisone. Some patients are intensely sensitive to salt intake and can fine-tune their plasma volume and BP control with salt intake manipulation alone. Foods with a high salt content include fast foods such as hamburgers, hot dogs, chicken pieces, fries, and fish fries. Canned soups, chili, ham, bacon, sausage, additives such as soy sauce, and commercially processed canned products also have a high sodium content. Fruits, especially bananas, and vegetables have a high potassium content. Patients should have a snack with more fluids at night. The postprandial orthostatic hypotension will reduce some of the nocturnal hypertension. The patient should have at least one glass or cup of fluid with meals and at least two at other times each day to obtain 1.5 to 2.5 L/day. Patient education on the relationship between orthostatic stressors such as duration of standing, exercise, meals, and ambient heat and symptoms and BP is important. Learning how to manage situations of increased orthostatic stress is very important. Recognizing orthostatic stressors and learning strategies to cope with these situations provide the patient with a great sense of empowerment. Physical countermaneuvers are very helpful in prolonging the duration that the patient can be upright. With simple explanation, patients recognize that they can alleviate orthostatic symptoms by utilizing the muscle pump of dependent muscles. For instance, they recognize that they are more symptomatic when they stand still than when they are pacing. They can be taught to modify simple maneuvers to enhance their efficacy, typically by increas-

ing the duration (to about 30 seconds) and/or the vigor of muscle contraction and by recruiting additional muscles. Commonly used countermaneuvers that have been demonstrated to be efficacious include maneuvers such as toe-raise, leg crossing, thigh contraction, and bending at the waist.6,91 The patient soon learns that these maneuvers can be used repetitively. Volume Expansion All patients with neurogenic orthostatic hypotension require the standard treatment of orthostatic hypotension aimed at expanding blood volume. Generous fluid intake is critically important and often neglected in the elderly. They need five to eight 8-ounce glasses of fluid per day. Salt supplementation is essential. Most patients will manage with added salt with their meals. The occasional patient will prefer to use salt tablets (available as 0.5- and 1-g tablets). It is important to recognize that the vasoconstrictors may be ineffective when plasma volume is significantly reduced. Many patients who have inadequate control of orthostatic hypotension have an inadequate salt intake. This can be verified by checking the 24-hour urinary sodium. Patients who have a value below 170 mmol/24 h can be treated with supplemental sodium 0.5 to 1 g tid and their weight, symptoms, and urinary sodium checked 1 or 2 weeks later.27 Postural Adjustment The head of the bed is elevated 4 inches for at least two reasons. First, it reduces nocturia, probably by stimulating renin release.74 Second, it reduces supine hypertension. During the day, it is important to maintain adequate orthostatic stress. Patients with orthostatic hypotension who are tilted up repeatedly develop gradual attenuation of orthostatic hypotension.41,62 This likely results from the release of renin and arginine vasopressin, which require more sustained or repetitive orthostatic stress. Another mechanism is suggested to be the effect of extravasated plasma around veins in providing a vascular cuff, increasing venomotor tone. Compression Garments In some subjects, the use of a tightly fitting body stocking results in amelioration of orthostatic hypotension and associated symptoms.85 These have to be well fitted and put on prior to arising. The stockings work by reducing the venous capacitance bed. Disadvantages are the cumbersome application and discomfort in hot weather. Some available sources are Jobst (1-800-537-1063), Barton-Carey (1-800-421-0444), Sigvaris (1-800-322-7744), and Camp (1-800-492-1088). A measured Jobst stocking (to generate a pressure of 30 mm Hg) is especially useful. Some subjects find that a tightly fitting girdle or abdominal binder is similarly successful. In a study using a modified G-suit, in which individual compartments (calves, thighs, abdomen) can be separately compressed, we found that calf com-

2659

Management of Autonomic Failure

pression was ineffective, combined compression of calves and thighs was mildly effective, and abdominal compression at 30 to 40 mm Hg provided about two thirds of the benefits of a well-fitting Jobst stocking.23 Physical Countermaneuvers Certain physical countermaneuvers, all involving the contraction of certain muscle groups of the lower extremities, reduce venous capacitance and increase venous return.91 Most patients recognize that their symptoms can be partially alleviated if they move around instead of standing perfectly still. Physical countermaneuvers are an extension of normal muscle contraction. The main difference is that the patient is taught to contract more vigorously for a longer duration (typically for 30 seconds) and recruit more muscles. These maneuvers, once learned, can significantly prolong standing time.106 They include crossing of the legs and contracting the leg muscles of one leg against the other, slow stepping or marching on the spot, propping the leg up on a chair, or contraction of the thighs (Table 119–7).6 Possibly the most useful maneuver is leg crossing. The subject is asking to make contact with or hold onto something (for support), cross the legs, and contract one leg against the other for 30 seconds. Typically systolic BP increases by 20 to 30 mm Hg. The maneuver can be repeated as needed. Training can enhance the efficacy of some more complicated maneuvers.6 Resistance Training Patients with chronic orthostatic hypotension typically have poor muscle tone as a result of factors such as old age, the disorder causing orthostatic hypotension, and the debilitating nature of chronic orthostatic hypotension itself. There is suggestive evidence that resistance training11,54,92 and endurance training19 will improve orthostatic hypotension, possibly by expanding plasma volume18 and increasing sympathetic tone.73 Increasing the size and strength of dependent muscle37 should enhance the efficacy of physical countermaneuvers. We recommend the use of resistance training focusing on dependent muscles (abdominal, glutei, quadriceps, gastrocnemius-soleus). Patients are encouraged to undertake weight training three times a week to build up these muscles.54 Water Drinking Recently, a simple method of increasing BP by simply drinking one or two 8-ounce glasses of water has been reported, and its mechanism of action studied.46 The effect Table 119–7. Some Useful Physical Countermaneuvers 1. 2. 3. 4.

Leg cross Marching Leg prop Thigh contraction

of drinking tap water on seated BP was studied in a large number of patients with severe neurogenic orthostatic hypotension and young and old control subjects. Systolic BP increased profoundly with water drinking, reaching a maximum increase of 33 ⫾ 5 mm Hg in MSA patients and 37 ⫾ 7 mm Hg in PAF patients after 30 minutes. The pressor response was proportional to the presence of residual sympathetic function and was almost completely abolished when sympathetic neurotransmission is abolished by intravenous ganglion blockade. Systolic BP increased by 11 ⫾ 2.4 mm Hg in elderly but not in young controls. Plasma norepinephrine increased in both groups while renin, vasopressin, and blood volume did not change in any group. Water drinking, therefore, increases BP by increasing sympathetic activity.

Pharmacologic Management Drug treatment is an important part of the overall therapeutic regimen and, if well used, will greatly enhance BP control.29 The main drugs are fludrocortisone and midodrine (Table 119–8). Midodrine and Fludrocortisone The optimal approach to the treatment of orthostatic hypotension is to modestly expand plasma volume without inducing significant supine hypertension and to add a drug that increases total systemic resistance, hence reducing the orthostatic fall in BP during times of greatest need. Midodrine hydrochloride is a prodrug that is converted to desglymidodrine, an active metabolite, with both drugs being active on ␣1-receptors on arterioles and venules.69,97,112 Treatment with midodrine has several advantages over previously available ␣-agonists. Because it is a prodrug, it is absorbed as an only partially active compound, potentially reducing gastrointestinal adverse effects. Unlike sympathomimetic drugs such as ephedrine, the amphetamines, and methylphenidate, which have significant central nervous system stimulant effects, midodrine does not cross the blood-brain barrier and is devoid of these central stimulant adverse effects.57 In a double-blind, placebo-controlled study, midodrine was demonstrated to significantly improve standing BP and reduce orthostatic lightheadedness (Fig. 119–1).57 The correlation between the effects of desglymidodrine on BP and its blood levels is excellent.111 The duration of action of a single dose is approximately 4 hours, corresponding to blood levels of desglymidodrine in excess of 15 ng/mL for the 10-mg dose (Fig. 119–2).111 The safest approach to volume expansion is by oral salt supplementation and the use of midodrine during the waking hours only. The minimum effective dose of midodrine is 5 mg, and most patients respond best to 10 mg. However, there is sufficient variability that a trial of 2.5 mg is reasonable for some patients. The optimal approach is to give a test dose of the medication and check standing and

2660

Diseases of the Peripheral Nervous System

Table 119–8. Doses and Common Side Effects of Some Medications Used to Treat Orthostatic Hypotension Treatment

Starting and Typical Dose

Common Side Effects/Complications

Fludrocortisone Midodrine Phenylpropanolamine Ephedrine

0.1 mg/day; 0.1 mg bid 2.5 mg tid; 10 mg tid 25 mg/day; 25 mg tid 25 mg tid

Dihydroergotamine

0.5–1 mg IM or SC Oral and nasal available Oral, nasal, suppository 5.4 mg tid

Hypokalemia, edema, cardiac failure Supine hypertension, scalp pruritus Supine hypertension, tachyphylaxis Nervousness, tachycardia, supine hypertension, tachyphylaxis Ergotism, supine hypertension

Ergotamine Yohimbine

25 ␮g SC bid; 100 ␮g SC bid 50 mg/kg 3/wk SC; same dose weekly 30–60 mg AM and noon; Mestinon, 180 mg/day

Octreotide Erythropoietin Pyridostigmine

Improvement in Standing Systolic BP, mm Hg

supine BP 1 hour later. The onset of action is between 0.5 and 1 hour and the duration of action is between 2 and 4 hours, corresponding to the peak blood levels of midodrine and its active metabolite, desglymidodrine, respectively (see Fig. 119–2).111 Some patients have a short duration of action of midodrine, lasting less than 4 hours. Since one of the mechanisms of hypertensive swings is severe hypotension, these patients do best with increasing the frequency of dosing (to every 3 hours) during the period of maximal orthostatic stress. Another approach is

30

P<.001 P<.001 P<.001 P<.001

25 20 15 10 5 0 5

Visit 1 Run In

Visit 2

Visit 3

Visit 4

Visit 5

Double-blind Period

Visit 6 Washout

FIGURE 119–1 Mean increment in standing systolic blood pressure (BP) recorded at different visits for midodrine (light shading) and placebo (dark shading) (mean ⫾ standard error of the mean). Statistical significance for midodrine versus placebo was P ⬍ .001 at visits 2, 3, 4, and 5.

Same as above Anxiety, tremor, palpitations, diarrhea, supine hypertension Nausea, abdominal pain, hypertension Supine hypertension, polycythemia Abdominal cramps, diarrhea, sweating

to take a small dose 2 hours after the initial dose. For these patients a supplemental dose of 2.5 to 5.0 mg could be tried. Patients should generally avoid midodrine after 6:00 PM. Some patients need significant tailoring of their dose schedule. For instance, some patients have severe orthostatic hypotension on awakening and have near-normal standing BP in the late afternoon. One regimen that works well for some patients is a schedule of 20 mg on awakening, 10 mg 4 hours later, and 5 mg in the afternoon. A good way to establish optimal dosing is to take supine and standing BP measurements before and after each dose for a couple of days and repeat that after a change in dosing. For patients who cannot take enough salt or do not respond adequately to midodrine, fludrocortisone (Florinef), 0.1 mg once or twice daily, can be added to provide volume expansion and sensitize vascular smooth muscle.20 Uncommonly, the dose of fludrocortisone may be increased to 0.4 or 0.6 mg daily in the patient with refractory orthostatic hypotension. Because the regulatory reflexes are greatly impaired, it is necessary to slightly overexpand the plasma volume of these patients. Reasonable clues indicating adequate volume expansion include a weight gain of 3 to 5 pounds. A better index is the 24-hour urinary sodium excretion. Patients with adequate sodium intake should excrete 170 mEq/24 h or more. Mild dependent edema is to be expected. The potential risks are congestive heart failure and excessive supine hypertension.17 Patients who are placed on fludrocortisone should have their supine and standing blood pressure checked in 2 weeks. Alternative or Second-Line Pharmacologic Agents Many of the second-line drugs used to treat orthostatic hypotension (see Table 119–8) are either of limited or historical interest only. Ephedrine, in a dose of 25 mg tid, is

Management of Autonomic Failure

2661

FIGURE 119–2 Blood levels of midodrine and its active metabolite desglymidodrine.

inconsistently effective.35 An indirectly acting ␣-agonist with ␤-agonist effects,30 ephedrine suffers from the disadvantages of inconsistent absorption, central stimulant effect, and tachyphylaxis (reduced efficacy with continuing use).30 Methylphenidate (Ritalin; 10 mg tid), also indirectly acting, is somewhat more efficacious. However, it causes central stimulation, resulting in palpitations, tremulousness, and anxiety, and there is a small risk of addiction. The duration of action is quite variable, ranging from 1 to 4 hours. Phenylpropanolamine (Propagest; 25 mg tid) is less well absorbed, has a shorter duration of action, and is generally less reliable than midodrine. As of November 2000, the Food and Drug Administration has recommended the removal of phenylpropanolamine from the market because of a concern for hemorrhagic stroke. The nonsteroidal anti-inflammatory drugs ibuprofen (400 to 800 mg tid) and indomethacin (25 to 50 mg tid) are only modestly efficacious. They probably work by inhibition of prostaglandin synthesis, blocking vasodilator prostaglandins.33,98 The drugs should be taken with meals or with a glass of milk to reduce the gastric irritant effect. Dihydroergotamine (DHE-45) is poorly absorbed, causes supine hypertension, and has the theoretical problem of ergotism.42,97 Related to DHE is the use of intranasal ergotamine (1 to 2 puffs). The drug should be used with caution and has the same contraindications as in its use for migraine. ␤-Blockers, either nonspecific ␤-blockers such as propranolol (Inderal)10 or those with partial agonist effects (e.g., pindolol), have been recommended previously65 but have not stood the test of time22 and cannot be currently recommended for the treatment of orthostatic hypotension. Clonidine (Catapres), a centrally acting ␣2-agonist, is rarely efficacious in the treatment of orthostatic hypotension, and its indication is doubtful. It has a central ␣2-agonist action that causes hypotension and a partial peripheral ␣1-agonist effect that causes hypertension by venoconstriction. In patients with tetraplegia or MSA (central or

preganglionic lesion), clonidine may work by its peripheral ␣-agonist effect on supersensitive postsynaptic ␣-receptors without a central inhibitory action. Metoclopramide and yohimbine are of doubtful clinical efficacy at the time of writing. Novel Applications of Old Drugs Erythropoietin. Serum Epo levels are partially controlled by the sympathetic outflow to the kidney. About 25% of patients with chronic orthostatic hypotension have a normocytic, normochromic anemia. In a recent study of 18 patients with MSA, 32 idiopathic Parkinson’s disease (PD) patients, 23 controls with iron deficiency anemia, and 18 healthy individuals, 4 MSA patients had unexplained anemia, whereas none of the PD patients was anemic. Serum Epo levels were suppressed in relation to anemia in MSA patients compared to elevated Epo levels in iron deficiency anemia patients, indicating Epo deficiency in the anemic MSA patients.109 Serum Epo levels in PD patients were within the normal range. Erythropoietin therapy was effective in treating the orthostatic hypotension in four diabetic anemic patients with orthostatic hypotension. The effect of Epo in raising standing BP might be mediated through a number of different mechanisms. Postulated mechanisms improving vasomotor tone include increased norepinephrine levels, quenching of nitric oxide by increased concentration of hemoglobin, increased vascular sensitivity to angiotensin II, and possibly direct pressor effects on vascular smooth muscle cells.108 Pyridostigmine. The major disadvantage of current pharmacotherapeutic options is the nonselective elevation of BP regardless of the body position, with supine hypertension as a common, troublesome, and typically dose-limiting side effect.17,36,77,83,111 In a recent study, we hypothesized that, because ganglionic transmission is low in the supine position and is greatly increased during orthostasis, enhancing ganglionic transmission by inhibition of acetylcholinesterase might selectively reduce orthostatic hypotension without

2662

Diseases of the Peripheral Nervous System

causing supine hypertension.86 We investigated the effect of pyridostigmine (60 mg) on 15 patients with neurogenic orthostatic hypotension. One hour after dosing, acetylcholinesterase inhibition had significantly increased orthostatic BP and total systemic resistance without a significant increase in supine BP. The improvement of orthostatic BP was associated with significant improvement of orthostatic symptoms. This novel approach may comprise an alternative or supplemental tool in the treatment of orthostatic hypotension, being of special value for patients with high supine BP.86

Treatment for Periods of Orthostatic Decompensation Patients with neurogenic orthostatic hypotension can have periods of orthostatic decompensation. During these times, they have greater orthostatic hypotension or are less responsive to pressor agents. These patients will need to be checked for a cause of decompensation. Causes include hypovolemia, hypokalemia, anemia, deconditioning related to a recent period of recumbency, and another illness, including pump (cardiac) failure; however, often no cause is found. Volume expansion is very important. As a first-aid measure, drinking two glasses of water will transiently elevate BP.46 An approach that results in more sustained improvement is the “bouillon treatment”: The patient drinks about five 8-ounce servings of one of these salty soups in one-half day. An alternative is supplemental sodium chloride (2 g tid) and a minimum of eight 8-ounce servings of fluids in 1 day for 2 days. Salt tablets should be taken with food. If the patient does not improve on this regimen or is not retaining the fluid, desmopressin (1 puff in each nostril qhs) is used for 1 week. The dose of vasoconstrictor can be adjusted upward. This is one time when a tight-fitting body stocking (e.g., Jobst stocking) could be very beneficial. Finally, fludrocortisone (0.2 mg tid) can be used for 1 week. The drug is traditionally considered to be slowly cumulative in its action, but clinical experience suggests a more rapid mode of action. Intravenous saline (1 to 2 L) is used if all these measures are unsuccessful.

Hospital Management of Severe Orthostatic Hypotension Some patients with severe orthostatic hypotension need acute hospital management. In addition to a search for the cause and specific treatment of orthostatic hypotension, hospital management is aimed at improving orthostatic tolerance to the degree that subsequent management can be continued on an outpatient basis. A regimen of treatment over approximately 3 days is suggested. These patients are volume depleted, either absolutely or rela-

tively (because of increased capacity caused by denervation). Intravenous infusion of 1 to 2 L of isotonic saline is needed to expand plasma volume. Volume expansion acutely is critically important because hypovolemia greatly reduces the effectiveness of vasoconstrictors in increasing blood pressure, significantly affecting the sensitivity of cardiopulmonary but not carotid-cardiac baroreflex responses to ␣-agonists.95 In elderly patients, care needs to be exercised to avoid heart failure. Postural training is needed. The head of the bed is progressively elevated over the 3 days. The patient spends an increasing period of time seated and standing. Fludrocortisone (0.2 mg/day) is commenced, as is NaCl (1 g tid) and a high fluid intake. During this hospitalization, the patient receives education on dietary salt content, maintenance of postural normotension, physical countermaneuvers, and the management of periods of increased orthostatic stress and supine hypertension. BP is measured supine and standing 1 minute before and 1 hour after 10 mg midodrine, and supine and standing values are recorded hourly to establish the optimal dose and duration of action.

Treatment of Early Morning Orthostatic Hypotension The most common time of day that orthostatic hypotension is worse is on awakening. This occurs in some subjects because of excessive nocturia. Many patients are improved by sleeping with the bed elevated (reducing nocturia). A common routine is to drink two cups of strong coffee (250 mg caffeine), take vasoconstrictors (typically midodrine), and read the newspaper or engage in some such activity for 30 minutes prior to getting up out of bed.

Treatment of Postprandial Orthostatic Hypotension Patients often have postprandial accentuation of their orthostatic hypotension.52,70,94 The condition can occur in any type of neurogenic orthostatic hypotension but is particularly common in diabetic autonomic neuropathy.43 It often occurs against the background of gastrointestinal autonomic neuropathy, highlighting the great importance of the splanchnic mesenteric bed in orthostatic blood pressure control.66 This is a large-volume (20% to 30% of total blood volume) capacitance bed that, unlike other venous beds, is exquisitely baroreflex responsive.78,84 Among patients with mild postprandial orthostatic hypotension, some find that the worsening can be reduced by frequent small meals. Other patients find that certain foods are the most troublesome and avoid these.67 Some patients report that hot drinks or hot foods need to be avoided. Carbohydrates and alcohol are especially troublesome.68

Management of Autonomic Failure

Ibuprofen (400 to 800 mg) or indomethacin (25 to 50 mg) with the meal is well tolerated and should be tried. The next step is the administration of a vasoconstrictor such as midodrine (10 mg). A problem with vasoconstrictors is the aggravation of gastroparesis. Rarely, symptoms suggestive of gut ischemia can occur. If all these approaches are inadequate, the somatostatin analogue octreotide can be used with the meal.1,76 The dose is 25 ␮g by subcutaneous injection. The dose can be increased, if necessary, to 100 to 200 ␮g. This is by far the most efficacious agent but requires parenteral administration.

Treatment of Nocturnal Hypertension Normal subjects have a diurnal variation in BP, with lower nocturnal BP. Patients with neurogenic orthostatic hypotension have a loss of diurnal variation of BP with nocturnal hypertension.14,45 To minimize the problems of nocturnal hypertension, pressor medications should not be given after 6:00 PM. The head of the bed should be elevated, resulting in a lower intracranial BP. A nighttime snack with a glass of fluids (not coffee or tea) will result in some postprandial hypotension and can be used as a means to increase fluid intake and reduce nocturnal hypertension. Patients who enjoy a glass of wine should have it at this time for its vasodilator effect. Occasionally, it is not possible to control orthostatic hypotension without significant nocturnal hypertension. In these patients, the following drugs have been reported to be efficacious. Hydralazine (Apresoline; 25 mg) can be given at night. This drug, with its sodium-retaining properties, is especially suitable. Nifedipine (Procardia; 10 mg) or another calcium channel blocker, extended-release nifedipine (30 mg), and a nitroglycerin patch are alternatives.30 In a comparison of 30 mg nifedipine (n ⫽ 10) and 0.025 to 0.2 mg/h nitroglycerin patch, both improved supine hypertension with the nitroglycerin patch being superior. It relieved supine hypertension without AM worsening of supine orthostatic hypotension, while nifedipine significantly worsened hypotension and enhanced nocturnal sodium excretion.47

Gastrointestinal and Genitourinary Dysautonomia BLADDER DYSFUNCTION There are multiple modalities of treatment for patients with disturbance of bladder function resulting from neurologic disease (see, e.g., Bradley8). The medical forms of treatment include behavioral therapy, catheterization,

2663

pharmacologic agents, and prevention and treatment of urinary tract infections. The surgical interventions used for patients who are not successfully treated with medical interventions include denervation of the detrusor, “clam” augmentation cystoplasty, urethral sphincterotomy, or artificial urinary sphincter implantation.8,9,71 Electrical stimulation for lower motor neuron nonreflexic bladders by directly stimulating the detrusor muscle or the nerves at the conus medullaris, anterior root, or pelvic nerve level8,38 and spiral intraurethral prosthetics101 are currently under investigation. This discussion focuses on the medical treatments that the neurologist may provide for patients with neurogenic bladders. Behavioral therapy is considered the first line of therapy in patients with neurogenic bladder because of potential benefits without risks. The forms of therapy used include habit training by means of voiding diaries and schedules so that the patient will have a voiding stimulus. Cooperative and highly motivated patients may use bladder retraining for control of urgency and detrusor hyperreflexia. The operant conditioning paradigm is used in biofeedback to help patients primarily with stress of postprostatectomy incontinence to develop the ability to suppress detrusor reflex contraction while simultaneously contracting the pelvic floor muscles.8 Cooperative, nondemented patients with neurologic disease that does not impair upper extremity motor control may employ the technique of clean, intermittent selfcatheterization.50 Patients with lower motor neuron bladder dysfunction and detrusor hyperreflexia treated with anticholinergic agent suppression of reflex activity may use this technique. The potential risks of urinary tract infections, renal stone formation, and possible renal deterioration must be kept in mind.8 Indwelling catheters or condom catheters are used when other methods fail. However, appropriate precautions must be used to prevent complications such as skin breakdown or urinary tract infections. Pharmacologic treatments include cholinergic agonists, anticholinergics, adrenergic agonists, calcium channel antagonists, and muscle relaxants. Patients with peripheral nervous system lesions causing detrusor areflexia have been treated with bethanechol, a cholinergic agonist, at several hundred milligrams per day without impressive results. Detrusor hyperreflexia has been treated with clean, intermittent self-catheterization and anticholinergic agents, including propantheline bromide (15 to 30 mg every 4 to 6 hours in adults), emepromium chloride (100 mg every 8 hours), and oxybutynin chloride (5 mg three to four times per day). Oxybutynin, which causes side effects of dry mouth, constipation, and esophageal reflux, has shown the greatest efficacy and has even been used successfully in some cases with imipramine.64,102 Imipramine, a tricyclic antidepressant with anticholinergic and ␣1-agonist actions, has shown limited efficacy when used alone.15

2664

Diseases of the Peripheral Nervous System

Prazosin, an ␣-adrenergic antagonist, has been used at low doses with gradual increase to 20 mg daily to treat excessive tone of urethral smooth muscles.39 The ␣-adrenergic agonist phenylpropanolamine, in doses of 50 to 75 mg twice per day, has served as treatment to increase urethral smooth muscle tone in women with stress incontinence. Midodrine, in doses similar to those used in the treatment of orthostatic hypotension, is preferable to phenylpropanolamine. Estrogen has also been used with some success in these patients. Flavoxate, an antispasmodic, and terodiline, an antimuscarinic and calcium channel blocker, have not been proven to provide significant benefit in clinical trials.16,90,100,110 Bladder distention should be avoided as a precaution in patients with spinal cord injury rostral to the conus medullaris. This helps to prevent the potentially lifethreatening disorder of autonomic dysreflexia, which may be associated with severe paroxysmal hypertension. Nifedipine and other calcium channel antagonists, in regularly scheduled oral doses or as needed for acute circumstances, may be helpful in treatment of these patients.51,72

IMPOTENCE The treatment of erectile sexual dysfunction has progressed significantly in the last 20 years. Useful interventions range from oral medications to complex surgical procedures. This subject has recently been reviewed.21,25,87 The current discussion focuses on the treatment of neurogenic erectile failure. However, most available forms of therapy are not specific for the underlying etiology causing the problem. In addition, counseling or sex therapy, which are the mainstays of therapy in psychogenic impotence, are useful as adjunctive therapy for patients with organic erectile failure. A number of general measures should be followed by any patient with impotence. These include avoidance of cigarettes, alcohol, and other drugs, treatment of general medical illnesses, reduction of stress, and improvement in sleep habits. Oral drug therapy with yohimbine, an ␣2-adrenergic receptor antagonist, may be beneficial in some patients with organic impotence. However, controlled trials have shown it to be only slightly better than placebo, with the best results noted in patients who have contributing psychological factors. The usual dosage is 5.4 mg three times per day, and onset of benefit is 3 days to 3 weeks after start of therapy.21,25 Sildenafil, an orally active type 5 cyclic GMP–specific phosphodiesterase, has been proven effective for organic erectile failure.5 External vacuum constriction devices may be useful in patients with neurogenic or other types of erectile failure. Negative pressure is applied to the penis with a vacuum pump and this allows blood to fill the corpus cavernosum. A constrictive rubber ring is then placed at the base of the

penis to impede venous flow of blood out of the penis. This safe, yet cumbersome, treatment is discontinued by a minority of the patients who use it because of pain from the constricting band or lack of spontaneity when using the device.99 Penile self-injection therapy using a 30-gauge needle to inject vasoactive agents into the corporeal body is an effective form of therapy for many men with erectile failure. Papaverine, prostaglandin E1, and a compound containing papaverine, prostaglandin, and phentolamine (Tri-Mix) are the agents commonly used. The use of Tri-Mix has advantages such as more predictable erection and lower cost than treatment with the other medications.21,25,87 The doses must be adjusted according to patient response, while keeping in mind that patients with neurogenic etiologies may be more sensitive to this form of treatment and require lower doses of any of the intracavernosal vasoactive agents. Pain, bleeding, infection, and prolonged erections are short-term complications. The restriction of injections to no more than 10 per month helps to lessen chances of the more serious long-term complication, scarring of the cavernosa, which eventually prevents effectiveness of injections and makes prosthetic implantation difficult. Surgical interventions are performed in select patients if the therapies noted above are ineffective. The surgical possibilities include implantation of penile prostheses (semi-rigid/malleable or inflatable types), arterial revascularization, and vein ligation for venous leakage. These types of treatments are used much less frequently than medical treatments, which are now quite successful in most patients with erectile failure. In addition, the techniques and success rates of arterial revascularization and venous ligation are being reassessed. Surgeries that allow perfusion of iliac arteries and spare pelvic autonomic nerves (aortoiliac procedures) can restore erectile function as well as prevent iatrogenic impotence.21

GASTROINTESTINAL SYMPTOMS Gastrointestinal symptoms are common in patients with autonomic failure of different etiologies. For effective management of gastrointestinal symptoms, it is essential to have an accurate diagnosis. A recent review outlines the strategy necessary in the diagnostic evaluation of patients with altered gastrointestinal function caused by autonomic dysfunction.75 The general therapeutic principles of gastrointestinal motility disorders include restoration of hydration and nutrition, stimulation of motility, suppression of bacterial overgrowth with antibiotics, and surgery in selected cases. We now review the current specific treatment modalities in the different gastrointestinal syndromes that are present in patients with autonomic failure.

2665

Management of Autonomic Failure

Gastroparesis Prokinetic agents and antiemetics combined with a digestible diet low in fat and fiber are the main forms of treatment. The most commonly used prokinetic medications are metoclopramide and cisapride. Metoclopramide is a peripheral cholinergic agonist and dopamine antagonist with antiemetic properties. The prokinetic effects are limited to the upper gastrointestinal tract, esophagus, and stomach. Its clinical use is limited by potential side effects, including extrapyramidal syndromes, tardive dyskinesia, and hyperprolactinemia. Therefore, the lowest dose possible is used, such as 10 mg tid before meals and 10 mg at bedtime. Cisapride is a relatively effective cholinergic agonist that stimulates transit and motor function at all levels of the gastrointestinal system. The usual dose is 20 mg PO tid, 30 minutes before meals and at bedtime. Unfortunately, because of its cardiac complications, the drug is only available via a company-sponsored program. Intravenous erythromycin is an effective prokinetic drug in cases of acute gastroparesis. The usual dose is 3 mg/kg every 8 hours. Domperidone is another dopamine antagonist useful for gastric stasis, but it has not yet been approved for use in the United States. Clonidine, a partial ␣-agonist, has been used in patients with diabetic gastroparesis with mixed results.

Chronic Intestinal Pseudo-obstruction These patients suffer from nausea, vomiting, early satiety, weight loss, abdominal discomfort, and obstinate constipation. Cisapride is the only prokinetic drug that has been shown to be efficacious in this syndrome. In addition, different antibiotics are used to suppress bacterial overgrowth. The most commonly used antibiotics are metronidazole (500 mg bid), doxycycline (100 mg tid), and ciprofloxacin (500 mg bid). These are usually given for 10 days each month, but selected patients may require chronic treatment.

Constipation Constipation is a frequent symptom. Common causes of constipation, such as little fiber in the diet, lack of exercise, and ignoring the urge to defecate, may be easily detected and corrected. Constipation may also be secondary to an obstructing lesion, disordered defecation, and altered colonic motility, as in patients with autonomic failure. An appropriate gastrointestinal evaluation should be performed prior to treatment. The management of intractable constipation has been recently reviewed.13 The therapeutic approach in patients with altered colonic motility may be divided in steps. The first step is proper hydration, fiber supplementation (ispaghula or psyllium), and osmotic laxatives such as magnesium salts (2 tablets of milk of magnesia qid) or stool softeners such as docusate (2 tablets bid). If this approach is not effective, the second step includes the

use of stimulant cathartics such as oral bisacodyl or a bisacodyl suppository supplemented with oral cisapride (if available). If this treatment is unsuccessful, the third level includes the combination of digital stimulation, laxatives, and enemas given every 1 or 2 days. Sacral-anterior root stimulation has been recently proposed as an alternative method to promote sigmoid and rectal contractions and sphincter relaxation to help facilitate defecation in patients with altered neuronal control.61 Surgical procedures (i.e., colostomy or subtotal colectomy) are reserved for patients with intractable colonic inertia that has failed previous treatments. New methods such as electrical pacing of the stomach are currently under investigation and may be of use in the future.

Diarrhea Several mechanisms are implicated, including osmotic diarrhea from bacterial overgrowth, rapid transit from uncoordinated bowel motor activity, dysfunction of the anorectal sphincter, and infiltration of the gut smooth muscle (i.e., amyloidosis). In patients with diabetic autonomic neuropathy, a superimposed pancreatic exocrine insufficiency or gluten-sensitive enteropathy may be present. The treatment is focused at identifying the proper mechanism. If osmotic diarrhea is identified, then a course of antibiotics such as the ones previously described is usually helpful. Whenever bacterial overgrowth is demonstrated, antibiotic treatment is required. For the treatment of diarrhea sometimes seen in diabetic autonomic neuropathy, tetracycline (250 to 500 mg) as a single dose at the onset of the diarrheal attack may abort symptoms dramatically in about half the cases.63 Erythromycin is sometimes effective in diabetic diarrhea. This drug also acts as a motilin agonist. Cholestyramine may be tried because it has been used to treat postvagotomy diarrhea, but its effectiveness has been questioned. Clonidine improves fluid absorption in experimental diabetes and has been reported to reduce diabetic diarrhea in humans.32

Fecal Incontinence Reversible causes such as overuse of laxatives or overflow incontinence resulting from fecal impaction should be sought first. Biofeedback may be of use in patients with sufficient residual innervation of the sphincters and pelvic nerves.

REFERENCES 1. Armstrong, E., and Mathias, C. J.: The effects of the somatostatin analogue, octreotide, on postural hypotension, before and after food ingestion, in primary autonomic failure. Clin. Auton. Res. 1:135, 1991.

2666

Diseases of the Peripheral Nervous System

2. Auer, R. N., and Coulter, K. C.: The nature and time course of neuronal vacuolation induced by the N-methyl-D-aspartate antagonist MK-801. Acta Neuropathol. (Berl.) 87:1, 1994. 3. Bannister, R.: Multiple system atrophy and pure autonomic failure. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management. Boston, Little, Brown, p. 517, 1993. 4. Biaggioni, I., Goldstein, D. S., Atkinson, T., and Robertson, D.: Dopamine-beta-hydroxylase deficiency in humans. Neurology 40:370, 1990. 5. Boolell, M., Gepi-Attee, S., Gingell, J. C., and Allen, M. J.: Sildenafil, a novel effective oral therapy for male erectile dysfunction. Br. J. Urol. 78:257, 1996. 6. Bouvette, C. M., McPhee, B. R., Opfer-Gehrking, T. L., and Low, P. A.: Role of physical countermaneuvers in the management of orthostatic hypotension: efficacy and biofeedback augmentation. Mayo Clin. Proc. 71:847, 1996. 7. Bower, J. H., Maraganore, D. M., McDonnell, S. K., and Rocca, W. A.: Incidence of progressive supranuclear palsy and multiple system atrophy in Olmsted County, Minnesota, 1976 to 1990. Neurology 49:1284, 1997. 8. Bradley, W. E.: Autonomic regulation of the urinary bladder. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management. Boston, Little, Brown, p. 105, 1993. 9. Bramble, F. J.: The clam cystoplasty. Br. J. Urol. 66:337, 1990. 10. Brevetti, G., Chiariello, M., Giudice, P., et al.: Effective treatment of orthostatic hypotension by propranolol in the Shy-Drager syndrome. Am. Heart J. 102:938, 1981. 11. Brilla, L. R., Stephens, A. B., Knutzen, K. M., and Caine, D.: Effect of strength training on orthostatic hypotension in other adults. J. Cardiopulm. Rehabil. 18:295, 1998. 12. Brooks, D. J., Redmond, S., Mathias, C. J., et al.: The effect of orthostatic hypotension on cerebral blood flow and middle cerebral artery velocity in autonomic failure, with observations on the action of ephedrine. J. Neurol. Neurosurg. Psychiatry 52:962, 1989. 13. Camilleri, M., Thompson, W. G., Fleshman, J. W., and Pemberton, J. H.: Clinical management of intractable constipation. Ann. Intern. Med. 121:520, 1994. 14. Carvalho, M. J., van Den Meiracker, A. H., Boomsma, F., et al.: Diurnal blood pressure variation in progressive autonomic failure. Hypertension 35:892, 2000. 15. Castleden, C. M., George, C. F., Renwick, A. G., and Asher, M. J.: Imipramine—a possible alternative to current therapy for urinary incontinence in the elderly. J. Urol. 125:318, 1981. 16. Chapple, C. R., Parkhouse, H., Gardener, C., and Milroy, E. J.: Double-blind, placebo-controlled, cross-over study of flavoxate in the treatment of idiopathic detrusor instability. Br. J. Urol. 66:491, 1990. 17. Chobanian, A. V., Volicer, L., Tifft, C. P., et al.: Mineralocorticoid-induced hypertension in patients with orthostatic hypotension. N. Engl. J. Med. 301:68, 1979. 18. Convertino, V. A.: Blood volume: its adaptation to endurance training. Med. Sci. Sports Exerc. 23:1338, 1991. 19. Convertino, V. A.: Endurance exercise training: conditions of enhanced hemodynamic responses and tolerance to LBNP. Med. Sci. Sports Exerc. 25:705, 1993. 20. Davies, I. B., Bannister, R. G., Sever, P. S., and Wilcox, C. S.: Fludrocortisone in the treatment of postural hypotension:

21. 22.

23.

24.

25. 26.

27.

28.

29.

30.

31. 32.

33.

34.

35.

36. 37.

38.

altered sensitivity to pressor agents. Br. J. Clin. Pharmacol. 6:444P, 1978. De Palma, R. G.: New developments in the diagnosis and treatment of impotence. West. J. Med. 164:54, 1996. Dejgaard, A., and Hilsted, J.: No effect of pindolol on postural hypotension in type 1 (insulin-dependent) diabetic patients with autonomic neuropathy: a randomised doubleblind controlled study. Diabetologia 31:281, 1988. Denq, J. C., Opfer-Gehrking, T. L., Giuliani, M., et al.: Efficacy of compression of different capacitance beds in the amelioration of orthostatic hypotension. Clin. Auton. Res. 7:321, 1997. Depresseux, J. C., Rousseau, J. J., and Franck, G.: The autoregulation of cerebral blood flow, the cerebrovascular reactivity and their interaction in the Shy-Drager syndrome. Eur. Neurol. 18:295, 1979. DeWire, D. M.: Evaluation and treatment of erectile dysfunction. Am. Fam. Physician 53:2101, 1996. Eldar, M., Battler, A., Neufeld, H. N., et al.: Transluminal carbon dioxide-laser catheter angioplasty for dissolution of atherosclerotic plaques. J. Am. Coll. Cardiol. 3:135, 1984. El-Sayed, H., and Hainsworth, R.: Salt supplementation increases plasma volume and orthostatic tolerance in patients with unexplained syncope. Heart 75:134, 1996. Ewing, D. J., Campbell, I. W., and Clarke, B. F.: The natural history of diabetic autonomic neuropathy. Q. J. Med. 49:95, 1980. Fealey, R. D., and Robertson, D.: Management of orthostatic hypotension. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management. Boston, Little, Brown, p. 731, 1993. Fealey, R. D., and Robertson, D.: Management of orthostatic hypotension. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management, 2nd ed. Philadelphia, Lippincott–Raven, p. 763, 1997. Fearnley, J. M., and Lees, A. J.: Striatonigral degeneration: a clinicopathological study. Brain 113:1823, 1990. Fedorak, R. N., Field, M., and Chang, E. B.: Treatment of diabetic diarrhea with clonidine. Ann. Intern. Med. 102:197, 1985. Galassi, G.: Indomethacin and propranolol combined treatment for Shy-Drager syndrome. J. Neurol. Neurosurg. Psychiatry 45:472, 1982. Gehrking, J. A., Hines, S. M., Benrud-Larson, L. M., et al.: What is the minimum duration of head-up tilt necessary to detect orthostatic hypotension? Clin. Auton. Res. 12:318, 2002. Ghrist, D. G., and Brown, G. E.: Postural hypotension with syncope: its successful treatment with ephedrine. Am. J. Med. Sci. 175:336, 1928. Gilden, J. L.: Midodrine in neurogenic orthostatic hypotension: a new treatment. Int. Angiol. 12:125, 1993. Hakkinen, K., and Hakkinen, A.: Neuromuscular adaptations during intensive strength training in middle-aged and elderly males and females. Electromyogr. Clin. Neurophysiol. 35:137, 1995. Hasan, S. T., Robson, W. A., Pridie, A. K., and Neal, D. E.: Transcutaneous electrical nerve stimulation and temporary S3 neuromodulation in idiopathic detrusor instability. J. Urol. 155:2005, 1996.

Management of Autonomic Failure 39. Hedlund, H., and Andersson, K. E.: Effects of prazosin in men with symptoms of bladder neck obstruction and a nonhyperplastic prostate. Scand. J. Urol. Nephrol. 23:251, 1989. 40. Hilsted, J., and Low, P. A.: Diabetic autonomic neuropathy. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management, 2nd ed. Philadelphia, Lippincott–Raven, p. 487, 1997. 41. Hoeldtke, R. D., Cavanaugh, S. T., and Hughes, J. D.: Treatment of orthostatic hypotension: interaction of pressor drugs and tilt table conditioning. Arch. Phys. Med. Rehabil. 69:895, 1988. 42. Hoeldtke, R. D., and Davis, K. M.: The orthostatic tachycardia syndrome: evaluation of autonomic function and treatment with octreotide and ergot alkaloids. J. Clin. Endocrinol. Metab. 73:132, 1991. 43. Hoeldtke, R. D., O’Dorisio, T. M., and Boden, G.: Prevention of postprandial hypotension with somatostatin. Ann. Intern. Med. 103:889, 1985. 44. Hoeldtke, R. D., and Streeten, D. H.: Treatment of orthostatic hypotension with erythropoietin. N. Engl. J. Med. 329:611, 1993.] 45. Imai, Y., Abe, K., Munakata, M., et al.: Circadian blood pressure variations under different pathophysiological conditions. J. Hypertens. 8:S125, 1990. 46. Jordan, J., Shannon, J. R., Black, B. K., et al.: The pressor response to water drinking in humans: a sympathetic reflex? Circulation 101:504, 2000. 47. Jordan, J., Shannon, J. R., Pohar, B., et al.: Contrasting effects of vasodilators on blood pressure and sodium balance in the hypertension of autonomic failure. J. Am. Soc. Nephrol. 10:35, 1999. 48. Joyner, M. J., and Shepherd, J. T.: Autonomic control of circulation. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management. Boston, Little, Brown, p. 55, 1993. 49. Kyle, R. A., and Greipp, P. R.: Amyloidosis: clinical and laboratory features in 229 cases. Mayo Clin. Proc. 58:665, 1983. 50. Lapides, J., Diokno, A. C., Silber, S. J., and Lowe, B. S.: Clean, intermittent self-catheterization in the treatment of urinary tract disease. J. Urol. 107:458, 1972. 51. Lindan, R., Leffler, E. J., and Kedia, K. R.: A comparison of the efficacy of an alpha-I-adrenergic blocker in the slow calcium channel blocker in the control of autonomic dysreflexia. Paraplegia 23:34, 1985. 52. Lipsitz, L. A., Nyquist, R. P. Jr., Wei, J. Y., and Rowe, J. W.: Postprandial reduction in blood pressure in the elderly. N. Engl. J. Med. 309:81, 1983. 53. Low, P. A.: Laboratory evaluation of autonomic function. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management, 2nd ed. Philadelphia, Lippincott–Raven, p. 179, 1997. 54. Low, P. A., Allison, T., McPhee, B., et al.: Efficacy of resistance training in improving orthostatic tolerance in POTS. Clin. Auton. Res. 8:279, 1998. 55. Low, P. A., and Bannister, R.: Multiple system atrophy and pure autonomic failure. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management, 2nd ed. Philadelphia, Lippincott–Raven, p. 555, 1997.

2667

56. Low, P. A., Denq, J. C., Opfer-Gehrking, T. L., et al.: Effect of age and gender on sudomotor and cardiovagal function and blood pressure response to tilt in normal subjects. Muscle Nerve 20:1561, 1997. 57. Low, P. A., Gilden, J. L., Freeman, R., et al.: Efficacy of midodrine vs placebo in neurogenic orthostatic hypotension: a randomized, double-blind multicenter study. Midodrine Study Group. JAMA 277:1046, 1997. 58. Low, P. A., and McLeod, J. G.: Autonomic neuropathies. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management, 2nd ed. Philadelphia, Lippincott–Raven, pp. 463–486, 1997. 59. Low, P. A., Opfer-Gehrking, T. L., McPhee, B. R., et al.: Prospective evaluation of clinical characteristics of orthostatic hypotension. Mayo Clin. Proc. 70:617, 1995. 60. Low, P. A., Walsh, J. C., Huang, C. Y., and McLeod, J. G.: The sympathetic nervous system in diabetic neuropathy: a clinical and pathological study. Brain 98:341, 1975. 61. MacDonagh, R. P., Sun, W. M., Smallwood, R., et al.: Control of defecation in patients with spinal injuries by stimulation of sacral anterior nerve roots. BMJ 300:1494, 1990. 62. Maclean, A. R., and Ellen, E. V.: Orthostatic hypotension and orthostatic tachycardia: treatment with the ‘head-up’ bed. JAMA 115:2163, 1940. 63. Malins, J. M., and French, J. M.: Diabetic diarrhoea. Q. J. Med. 26:467, 1957. 64. Malone-Lee, J. G.: The clinical efficacy of oxybutynin. Rev. Contemp. Pharmacother. 5:195, 1994. 65. Man In’T Veld, A.J., and Schalekamp, M. A.: Pindolol acts as beta-adrenoceptor agonist in orthostatic hypotension: therapeutic implications. Br. Med. J. (Clin. Res. Ed). 282:929, 1981. 66. Mathias, C. J., da Costa, D. F., Fosbraey, P., et al.: Cardiovascular, biochemical and hormonal changes during food-induced hypotension in chronic autonomic failure. J. Neurol. Sci. 94:255, 1989. 67. Mathias, C. J., Holly, E., Armstrong, E., et al.: The influence of food on postural hypotension in three groups with chronic autonomic failure—clinical and therapeutic implications. J. Neurol. Neurosurg. Psychiatry 54:726, 1991. 68. Maule, S., Chaudhuri, K. R., Thomaides, T., et al.: Effects of oral alcohol on superior mesenteric artery blood flow in normal man, horizontal and tilted. Clin. Sci. 84:419, 1993. 69. McTavish, D., and Goa, K. L.: Midodrine: a review of its pharmacological properties and therapeutic use in orthostatic hypotension and secondary hypotensive disorders. Drugs 38:757, 1989. 70. Micieli, G., Martignoni, E., Cavallini, A., et al.: Postprandial and orthostatic hypotension in Parkinson’s disease. Neurology 37:386, 1987. 71. Mundy, A. R., and Stephenson, T. P.: “Clam” ileocystoplasty for the treatment of refractory urge incontinence. Br. J. Urol. 57:641, 1985. 72. Murnaghan, G. F.: Neurogenic disorders of the bladder in parkinsonism. Br. J. Urol. 33:403, 1961. 73. Ng, A. V., Callister, R., Johnson, D. G., and Seals, D. R.: Endurance exercise training is associated with elevated basal sympathetic nerve activity in healthy older humans. J. Appl. Physiol. 77:1366, 1994.

2668

Diseases of the Peripheral Nervous System

74. Oparil, S., Vassaux, C., Sanders, C. A., and Haber, E.: Role of renin in acute postural homeostasis. Circulation 41:89, 1970. 75. Prather, C. M., and Camilleri, M.: Gastrointestinal dysfunction: approach to management. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management, 2nd ed. Philadelphia, Lippincott–Raven, p. 597, 1997. 76. Raimbach, S. J., Cortelli, P., Kooner, J. S., et al.: Prevention of glucose-induced hypotension by the somatostatin analogue octreotide (SMS 201-995) in chronic autonomic failure: haemodynamic and hormonal changes. Clin. Sci. 77:623, 1989. 77. Robertson, D., and Davis, T. L.: Recent advances in the treatment of orthostatic hypotension. Neurology 45:S26, 1995. 78. Rowell, L. B., Detry, J. M., Blackmon, J. R., and Wyss, C.: Importance of the splanchnic vascular bed in human blood pressure regulation. J. Appl. Physiol. 32:213, 1972. 79. Saito, Y., Matsuoka, Y., Takahashi, A., and Ohno, Y.: Survival of patients with multiple system atrophy. Intern. Med. 33:321, 1994. 80. Sandroni, P., Ahlskog, J. E., Fealey, R. D., and Low, P. A.: Autonomic involvement in extrapyramidal and cerebellar disorders. Clin. Auton. Res. 1:147, 1991. 81. Schatz, I. J., Bannister, R., Freeman, R. L., et al.: Consensus statement on the definition of orthostatic hypotension, pure autonomic failure, and multiple system atrophy. Neurology 46:1470, 1996. 82. Schrag, A., Ben-Shlomo, Y., and Quinn, N. P.: Prevalence of progressive supranuclear palsy and multiple system atrophy: a cross-sectional study. Lancet 354:1771, 1999. 83. Shannon, J., Jordan, J., Costa, F., et al.: The hypertension of autonomic failure and its treatment. Hypertension 30:1062, 1997. 84. Shepherd, J. T., and Vanhoutte, P. M.: The Human Cardiovascular System: Facts and Concepts. New York, Raven Press, 1979. 85. Sheps, S. G.: Use of an elastic garment in the treatment of orthostatic hypotension. Cardiology 61:271, 1976. 86. Singer, W., Opfer-Gehrking, T. L., McPhee, B. R., et al.: Acetylcholinesterase inhibition: a novel approach in the treatment of neurogenic orthostatic hypotension. J. Neurol. Neurosurg. Psychiatry 74:1294, 2003. 87. Stewart, J. D.: Autonomic regulation of sexual function. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management. Boston, Little, Brown, p. 117, 1993. 88. Suarez, G. A., Fealey, R. D., Camilleri, M., and Low, P. A.: Idiopathic autonomic neuropathy: clinical, neurophysiologic, and follow-up studies on 27 patients. Neurology 44:1675, 1994. 89. Symon, L.: Pathological regulation in cerebral ischemia. In Wood, J. H. (ed.): Cerebral Blood Flow: Physiologic and Clinical Aspects. New York, McGraw-Hill, p. 423, 1987. 90. Tapp, A., Fall, M., Norgaard, J., et al.: Terodiline: a dose titrated, multicenter study of the treatment of idiopathic detrusor instability in women. J. Urol. 142:1027, 1989. 91. ten Harkel, A. D., van Lieshout, J. J., and Wieling, W.: Effects of leg muscle pumping and tensing on orthostatic

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102. 103.

104.

105.

106.

107.

arterial pressure: a study in normal subjects and patients with autonomic failure. Clin. Sci. 87:553, 1994. Tesch, P. A., Hjort, H., and Balldin, U. I.: Effects of strength training on G tolerance. Aviat. Space Environ. Med. 54:691, 1983. Thirlwell, M. P., and Zsoter, T. T.: The effect of propranolol and atropine on venomotor reflexes in man: venous reflexes—effect of propranolol and atropine. J. Med. 3:65, 1972. Thomaides, T., Bleasdale-Barr, K., Chaudhuri, K. R., et al.: Cardiovascular and hormonal responses to liquid food challenge in idiopathic Parkinson’s disease, multiple system atrophy, and pure autonomic failure. Neurology 43:900, 1993. Thompson, C. A., Tatro, D. L., Ludwig, D. A., and Convertino, V. A.: Baroreflex responses to acute changes in blood volume in humans. Am. J. Physiol. 259:R792, 1990. Thompson, J. M., Jennings, G. L., Chin, J. P., and Esler, M. D.: Measurement of human sympathetic nervous responses to stressors by microneurography. J. Auton. Nerv. Syst. 49:277, 1994. Thulesius, O., Gjores, J. E., and Berlin, E.: Vasoconstrictor effect of midodrine, ST 1059, noradrenaline, etilefrine and dihydroergotamine on isolated human veins. Eur. J. Clin. Pharmacol. 16:423, 1979. Tsuda, Y., Kimura, K., Yoneda, S., et al.: Hemodynamics in Shy-Drager syndrome and treatment with indomethacin. Eur. Neurol. 22:421, 1983. Turner, L. A., Althof, S. E., Levine, S. B., et al.: External vacuum devices in the treatment of erectile dysfunction: a one-year study of sexual and psychosocial impact. J. Sex Marital Ther. 17:81, 1991. van der Klauw, M. M., van Rey, F. J., and Stricker, B. H.: Polymorph ventricular tachycardia with torsades de pointes caused by administration of terodiline (Mictrol). Ned. Tijdschr. Geneeskd. 136:91, 1992. Vicente, J., Salvador, J., and Chechile, G.: Spiral urethral prosthesis as an alternative to surgery in high risk patients with benign prostatic hyperplasia: prospective study. J. Urol. 142:1504, 1989. Wall, L. L.: The management of detrusor instability. Clin. Obstet. Gynecol. 33:367, 1990. Wallin, B. G., and Elam, M.: Microneurography and autonomic dysfunction. In Low, P. A. (ed.): Clinical Autonomic Disorders: Evaluation and Management, 2nd ed. Philadelphia, Lippincott–Raven, p. 233, 1997. Wenning, G. K., Ben Shlomo, Y., Magalhaes, M., et al.: Clinical features and natural history of multiple system atrophy: an analysis of 100 cases. Brain 117:835, 1994. White, J. C., and Smithwick, R. H.: The Autonomic Nervous System: Anatomy, Physiology and Surgical Application. London, Macmillan, 1944. Wieling, W., van Lieshout, J. J., and van Leeuwen, A. M.: Physical manoeuvres that reduce postural hypotension in autonomic failure. Clin. Auton. Res. 3:57, 1993. Wilkins, R. W., Culbertson, J. W., and Inglefinger, F. J.: The effect of splanchnic sympathectomy in hypertensive patients upon estimated hepatic blood flow in the upright as

Management of Autonomic Failure contrasted with the horizontal position. J. Clin. Invest. 30:312, 1951. 108. Winkler, A. S., Landau, S., and Watkins, P. J.: Erythropoietin treatment of postural hypotension in anemic type 1 diabetic patients with autonomic neuropathy: a case study of four patients. Diabetes Care 24:1121, 2001. 109. Winkler, A. S., Marsden, J., Parton, M., et al.: Erythropoietin deficiency and anaemia in multiple system atrophy. Mov. Disord. 16:233, 2001.

2669

110. Wiseman, P. A., Malone-Lee, J., and Rai, G. S.: Terodiline with bladder retraining for treating detrusor instability in elderly people. BMJ 302:994, 1991. 111. Wright, R. A., Kaufmann, H. C., Perera, R., et al.: A doubleblind, dose-response study of midodrine in neurogenic orthostatic hypotension. Neurology 51:120, 1998. 112. Zachariah, P. K., Bloedow, D. C., Moyer, T. P., et al.: Pharmacodynamics of midodrine, an antihypotensive agent. Clin. Pharmacol. Ther. 39:586, 1986.

120 Treatment of Cardiovascular Autonomic Failure ROY FREEMAN

Overview of Cardiovascular Autonomic Failure Treatment

Nonpharmacologic Measures Volume Expansion Sympathomimetic Agents

OVERVIEW OF CARDIOVASCULAR AUTONOMIC FAILURE The assumption of the upright posture results in a complex sequence of physiologic reactions in response to the pooling of 500 to 1000 mL of blood in the lower extremities and splanchnic circulation. There is a decrease in venous return to the heart, and the reduced ventricular filling pressure results in diminished cardiac output and blood pressure. These hemodynamic changes provoke a baroreceptor-initiated compensatory reflex that is modulated via the central nervous system and executed by the peripheral efferent autonomic outflow. Specifically, the lower arterial pressure unloads the baroreceptors—the terminals of afferent fibers of the glossopharyngeal and vagus nerves—that are situated in the carotid sinus and aortic arch. This leads to a reduction in the afferent impulses that are relayed from these receptors via the glossopharyngeal and vagus nerves to the nucleus tractus solitarius (NTS) in the dorsomedial medulla. The reduced baroreceptor afferent activity produces a decrease in vagal nerve activity (mediated by the neuroanatomic connections of the NTS to the nucleus ambiguus) and an increase in sympathetic efferent activity (mediated by the NTS projections to the caudal ventrolateral medulla and from there to the rostral ventrolateral medulla). The synapse of the afferent glossopharyngeal and vagus nerves with the NTS and the synapse of the neurons of the NTS with the caudal ventrolateral medulla are glutamatergic. The synapse of neurons from the caudal ventrolateral medulla with the rostral ventrolateral medulla is GABAergic. The rostral ventrolateral

Supplementary Therapy and Experimental Agents Conclusion

medulla neurons project to preganglionic neurons in the thoracic and lumbar intermediolateral column that in turn project to the peripheral sympathetic efferents, the final effector neurons (Fig. 120–1).16,49,55,92 Thus, in response to a decreasing blood pressure, sympathetic outflow to the heart, kidney, and peripheral vasculature increases. This results in constriction of the capacitance and resistance vessels and an increase in heart rate and myocardial contractility. There is also baroreceptormediated release of vasopressin from the posterior pituitary gland. Vasopressin has a pressor effect and also conserves fluid volume to support blood pressure over the longer term.40,46 These compensatory mechanisms increase the peripheral resistance, venous return, and cardiac output, and thus limit the fall in blood pressure. If these mechanisms fail, orthostatic hypotension and the symptoms of cerebral hypoperfusion occur. The normal response to the assumption of the erect posture is a fall in systolic blood pressure (5 to 10 mm Hg), an increase in diastolic blood pressure (5 to 10 mm Hg), and an increase in the pulse rate (10 to 25 beats/min).9,91 Orthostatic hypotension is defined as a decrease in systolic blood pressure of greater than 20 mm Hg or diastolic blood pressure of greater that 10 mm Hg upon standing or head-up tilt accompanied by symptoms of cerebral hypoperfusion.19 Orthostatic hypotension, the cardinal feature of cardiovascular autonomic failure, is the most incapacitating symptom of autonomic dysfunction. Severely afflicted patients are unable to leave the supine position without experiencing symptoms of presyncope or losing consciousness. Orthostatic hypotension is not always the first symptom of autonomic 2671

2672

Diseases of the Peripheral Nervous System

FIGURE 120–1 Pathways within the lower brainstem and spinal cord that subserve the baroreceptor and chemoreceptor reflex control of sympathetic outflow to the heart and blood vessels. 䉭, excitatory synaptic inputs; 䉱, inhibitory synaptic inputs. CVLM ⫽ caudal ventrolateral medulla; IML ⫽ intermediolateral cell column in the spinal cord; KF ⫽ Kölliker-Fuse nucleus in pons; NTS ⫽ nucleus tractus solitarius; RVLM ⫽ rostral ventrolateral medulla. (From Dampney, R. A., Coleman, M. J., Fontes, M. A., et al.: Central mechanisms underlying short- and long-term regulation of the cardiovascular system. Clin. Exp. Pharmacol. Physiol. 29:261, 2002, with permission.)

failure,83,98 although it is the symptom that frequently leads patients to seek medical attention. Patients typically present with lightheadedness and presyncopal complaints that occur in response to sudden postural change, meals, exertion, or prolonged standing. Complaints less easily recognized as hypotensive in origin, such as generalized weakness, fatigue, cognitive slowing, leg buckling, visual blurring, headache, and neck pain, also may be present. The visual complaints most likely represent retinal or occipital lobe ischemia. Neck pain typically occurs in the suboccipital, posterior cervical, and shoulder region (the “coat-hanger headache”) and may be the only symptom of orthostatic hypotension. This symptom is most likely a consequence of neck muscle ischemia.80 Some patients report anginal pains despite normal coronary arteries. Loss of consciousness may be of gradual onset or can occur suddenly, raising the possibility of a seizure or cardiac cause. Some patients may display coarse jerking movements and, rarely, focal neurologic findings. The appearance of focal findings may suggest underlying cerebrovascular disease.20 Patients with autonomic failure accommodate to their orthostatic hypotension and frequently are able to tolerate significant falls in blood pressure (even to the 40- to 50-mm Hg level) without symptoms. This may in part be due to a shift to the left of the lower limit of cerebral autoregulation, the capacity to maintain constant cerebral blood flow despite changes in the perfusion pressure. Orthostatic hypotension occurs as a consequence of primary degeneration of the central or peripheral autonomic nervous system. Multiple system atrophy (MSA),

or the Shy-Drager syndrome, is a central autonomic degenerative disorder in which the autonomic dysfunction is accompanied by degeneration of other neuronal groups in the extrapyramidal, pyramidal, and cerebellar systems.1,28,74,89,99 Orthostatic hypotension may also accompany Parkinson’s disease and other central nervous system disorders.29,31,93 The treatment of Parkinson’s disease with L-dihydroxyphenylalanine (L-DOPA) and dopamine agonists may exacerbate this tendency.86 Orthostatic hypotension may rarely occur early in the course of Parkinson’s disease, prior to the initiation of treatment.8 Pure autonomic failure (PAF) results in progressive autonomic failure characteristically without other associated signs or symptoms of peripheral neuropathy or central nervous system disease.10,23,33 Peripheral autonomic dysfunction may also accompany small fiber peripheral neuropathies such as those seen in diabetes, amyloidosis, immune-mediated neuropathies, hereditary sensory and autonomic neuropathies (types III and IV), and inflammatory neuropathies. Less frequently, orthostatic hypotension is associated with the peripheral neuropathies that accompany porphyria, vitamin B12 deficiency, and human immunodeficiency virus infection.58,65 The hallmark of both central and peripheral causes of neurogenic orthostatic hypotension is the failure to release the catecholamine norepinephrine appropriately upon standing. Norepinephrine, which along with the other endogenous catechols, epinephrine and dopamine, is synthesized from tryosine (Fig. 120–2), is normally released

Treatment of Cardiovascular Autonomic Failure

Nonpharmacologic Measures

COO NH2 HO

Tyrosine

tyrosine

COO

H

NH2 HO

DOPA L-aromatic amino-acid decarboxylase

OH H

H NH2

NH2

dopamine β-hydroxylase

HO

HO

Dopamine

2673

Norepinephrine

FIGURE 120–2 The biosynthesis of norepinephrine from tyrosine. (Modified from Freeman, R., Landsberg, L., and Young, J.: The treatment of neurogenic orthostatic hypotension with 3,4-DL-threodihydroxyphenylserine: a randomized, placebo-controlled, crossover trial. Neurology 53:2151, 1999.)

into the synaptic cleft in response to standing, resulting in a two- to threefold increase in plasma norepinephrine.30,32,52,106 Patients with neurogenic orthostatic hypotension caused by PAF or a small fiber peripheral neuropathy typically have lower resting norepinephrine levels, as a result of degeneration of postganglionic neurons. Patients with MSA (in which the postganglionic neuron remains intact) have normal supine norepinephrine levels. Neither patient group demonstrates an increase in plasma norepinephrine upon standing.52,106 However, considerable overlap exists between these two patient groups, which limits the ability to make a definitive diagnosis based on plasma catecholamine measurements in individual patients.2,15,32

TREATMENT The therapeutic interventions to treat patients with orthostatic hypotension should be implemented in stages and depend, in large part, on the severity of symptoms. Some patients may be substantially improved by education, counseling, removal of hypotensive medications, and other nonpharmacologic interventions, while more severely afflicted patients require pharmacologic interventions.

Patient education is the cornerstone of the management of orthostatic hypotension. Throughout the day, patients are subject to a number of orthostatic demands for which there are simple but effective countermeasures. The time spent with patients emphasizing these practical management principles is of inestimable value (Table 120–1). Patients with orthostatic hypotension should move from a supine to a standing position in gradual stages, particularly in the morning, when orthostatic tolerance is lowest.69 Maneuvers such as straining, coughing, and raising the arms above the head should be avoided. These maneuvers, which increase the intrathoracic pressure, may reduce venous return and induce symptoms of hypotension. Similarly, isometric exercise should be discouraged. Prolonged recumbency can markedly reduce orthostatic tolerance, and physical activity and exercise should be strongly advocated. Isotonic exercise is recommended to avoid the reduction in venous return that is associated with straining. Patients should be advised that the intramuscular vasodilation that accompanies aerobic exercise may exacerbate orthostatic hypotension. Recumbent or seated aerobic exercise is preferable, such as on a rowing machine or reclining bicycle. Swimming is well tolerated since the hydrostatic pressure that accompanies swimming increases orthostatic tolerance during this activity. The hemodynamic consequences of exercise, warm weather, fevers, hot baths, and showers may present challenges to orthostatic tolerance. This is of particular importance in patients who lack appropriate thermoregulatory sweating. Patients should be made aware of a number of physical countermaneuvers that can help maintain blood pressure during daily activities. Such maneuvers include leg-crossing; stooping; squatting; tensing of the muscles of the leg, abdomen, buttock, or whole body; and muscle tensing.94 These maneuvers reduce venous pooling and thereby increase central blood volume and cardiac filling pressures. Cardiac output is increased, with a consequent increase in blood pressure and cerebral perfusion.102 Table 120–1. Nonpharmacologic Measures Gradual staged movements with postural change Avoid straining, coughing, and other maneuvers that increase intrathoracic pressure Avoid prolonged recumbency Encourage isotonic exercise Physical countermaneuvers Discontinue hypotensive medications Raise head of the bed 10 to 20 degrees Custom-fitted elastic stockings Minimize postprandial hypotension

2674

Diseases of the Peripheral Nervous System

The use of custom-fitted elastic stockings permits the application of a graded pressure to the lower extremity and abdomen. These stockings minimize peripheral blood pooling in the lower extremities and splanchnic circulation. It is essential that such stockings extend to the waist because most peripheral pooling occurs in the splanchnic circulation. These stockings are poorly tolerated by many patients, particularly those with painful peripheral neuropathies or motor dysfunction. Antigravity suits have been useful to some patients. Patients with autonomic failure and even normal elderly are susceptible to significant falls in blood pressure associated with meals.54,81 Postprandial hypotension can be minimized by avoiding large meals, ingesting meals low in carbohydrates, and minimizing alcohol intake. Patients should be advised against activities or sudden standing immediately after eating. The recognition and removal of potential reversible causes of orthostatic hypotension is an important management step. Medications such as diuretics, antihypertensive agents, antianginal agents, antidepressants, and ␣ adrenoreceptor antagonists to treat benign prostatic hyperplasia are the most common offending agents. Paradoxically, the use of antihypertensive agents occurs frequently in patients with orthostatic hypotension, because of the coexistence of supine hypertension. Many patients will present with orthostatic hypotension shortly after treatment for supine hypertension is initiated.87 Orthostatic hypotension also may first manifest when a tricyclic antidepressant is introduced as therapy for a painful neuropathy. Nortriptyline and desipramine are tricyclic agents with a relatively low incidence of postural hypotension, and their use may minimize this problem.82 In contrast, patients with autonomic failure are sensitive to the hypertensive effects of over-the-counter medications such as sympathomimetic-containing diet pills and cold remedies.4 Even topical ophthalmic solutions may produce similar unwanted effects.77 Many patients with neurogenic orthostatic hypotension have severe supine hypertension even before treatment of hypotension is initiated, although the hypertension is often exacerbated by treatment with fludrocortisone and pressor agents.87 The prolonged duration of action of fludrocortisone makes it a common offender. In some series, supine hypertension is present in over 50% of patients.87 The cause of supine hypertension that is not due to therapy in patients with autonomic failure is not definitively established. It does not appear to be due to increased plasma volume, cardiac output, plasma norepinephrine, or plasma renin.87 Baroreflex dysfunction in the presence of residual sympathetic outflow may be partially responsible.88 Severe supine hypertension often limits therapeutic intervention; however, surprisingly, most patients tolerate sustained supine blood pressures without untoward effect, and hypertensive end-organ damage, such as cerebrovascular

disease, nephropathy, and cardiomyopathy, is not seen as frequently as might be anticipated. The use of short-acting oral antihypertensive agents or transdermal nitroglycerin (0.025 to 0.1 mg/h) should be considered in patients with severe, sustained supine hypertension. Alternately, ingestion of a snack or small amounts of alcohol before retiring may be of benefit, although these interventions do not lead to a sustained reduction in blood pressure. Orthostatic tolerance, should the patient need to leave the supine position, may be severely impaired by these interventions, particularly since patients with autonomic failure lack the ability to generate the appropriate compensatory reflexes. These interventions increase the likelihood of falls, and this risk should be balanced against the perceived consequences of supine hypertension.

Volume Expansion Administration of Fluid and Sodium Chloride Plasma volume expansion is essential to improve orthostatic tolerance. The high nocturnal blood pressure causes a pressure natriuresis and results in nocturnal volume and sodium chloride depletion. This phenomenon is most likely responsible for the observation that patients with neurogenic orthostatic hypotension are more symptomatic in the morning hours. Raising the head of the bed 10 to 20 degrees (6 to 10 inches)59 reduces supine hypertension, activates the renin-angiotensin-aldosterone system, and decreases the nocturnal diuresis. The excessive natriuresis and reduction in central blood volume can be attenuated or minimized by increasing sodium intake with high-sodium foods or salt tablets and by increasing fluid ingestion.91 Patients should have a daily dietary intake of at least 10 g (185 mmol) of sodium. The increase in dietary sodium should be accompanied by an increase in fluid intake of 2 to 2.5 L (in adults) per day. A 24-hour urinary sodium excretion of greater than 170 mEq/24 h and a urinary volume of greater than 1500 mL is a measure of acceptable fluid and salt intake. An early morning body weight gain of about 1 to 2 kg usually implies adequate extracellular fluid volume expansion.101 Ingestion of approximately 500 mL of tap water elicits a marked pressor response and improvement in symptoms in patients with autonomic failure. The pressor response, a systolic blood pressure increase of over 30 mm Hg in some patients, is evident within 5 minutes after the water ingestion. The peak response occurs 20 to 30 minutes later and the effect lasts for up to 1 hour. There is some interpatient variability and not all patients respond to this intervention. The mechanisms underlying the pressor effect are not established. There is an increase in venous plasma norepinephrine following this intervention, suggesting that sympathetic nervous system activation plays a role in this response.43

Treatment of Cardiovascular Autonomic Failure

9-␣-Fluorohydrocortisone 9-␣-Fluorohydrocortisone (fludrocortisone acetate), a synthetic mineralocorticoid, may be used if patients are unable to increase plasma volume effectively with fluid and salt.11,34,95 This agent has a long duration of action and is well tolerated by most patients. Fludrocortisone increases the blood volume and enhances the sensitivity of blood vessels to circulating catecholamines.18,34 Other potential modes of action include enhancing norepinephrine release from sympathetic neurons and increasing vascular fluid content.85 Sodium retention and plasma volume return to normal with long-term use, although the pressor effect persists as a result of increased peripheral vascular resistance.13 Treatment is initiated with a 0.1-mg tablet and can be increased to 0.3 to 0.5 mg daily.101 Treatment may unfortunately be limited by supine hypertension resulting from an increase in the peripheral vascular resistance.13 Other side effects include ankle edema, hypokalemia, headache, and, rarely, congestive heart failure. Potassium supplementation is usually required, particularly when higher doses are used. Vasopressin Analogues The vasopressin analogues may be used to supplement volume expansion. The postural release of arginine-vasopressin is reduced in some patients with autonomic failure,46 particularly those patients with autonomic failure caused by a central neurodegenerative process, in which there may be loss of vasopressin neurons in the suprachiasmatic nucleus of the hypothalamus.72 In normal individuals, daytime urinary volumes are 33% more than nocturnal volumes. This is mediated by argininevasopressin, which has a circadian rhythm that peaks during the night. Thus the increase in nocturnal urinary excretion in patients with autonomic failure, which is in part due to the increase in supine blood pressures, is also due to loss of vasopressin neurons in the suprachiasmatic nucleus of the hypothalamus and the ensuing loss of the normal argininevasopressin circadian rhythm.73 The potent, synthetic vasopressin analogue desmopressin acetate (DDAVP) acts on the V2 receptors in the collecting ducts of the renal tubules and has minimal V1 receptor vasoconstricting potential. This agent, which can be administered as a nasal spray (5 to 40 ␮g), orally (100 to 800 ␮g), or intramuscularly (2 to 4 ␮g), prevents nocturia and weight loss and reduces the morning postural fall in blood pressure in patients with autonomic failure. Fluid and electrolyte status should be carefully monitored during therapy to avoid water intoxication and hyponatremia.63 Low doses of intranasal desmopressin (5 ␮g) may be sufficient to attenuate the nocturnal diuresis without placing the patient at risk for hyponatremia.84 The vasopressin analogues of vasopressin (V1 receptor agonists), such as lysine-vasopressin nasal spray and intramuscular triglycllysine vasopressin, may also increase blood pressure and

2675

peripheral vascular resistance and improve symptoms of orthostatic hypotension.50

Sympathomimetic Agents Since neurogenic orthostatic hypotension is in large part a consequence of the failure to release the ␣1 adrenoreceptor agonist norepinephrine from sympathetic neurons, the administration of sympathomimetic agents is central to the management of patients with cardiovascular autonomic failure. Thus, if symptoms of orthostatic intolerance persist once intravascular volume is replenished with varying combinations of fluid, sodium chloride, fludrocortisone acetate, and vasopressin analogues (Table 120–2), the direct or indirect adrenergic agonists and antagonists should be administered (Table 120–3). The pressor response of these agents is due to the reduction in venous capacity and the constriction of the resistance vessels. The available ␣1 adrenoreceptor agonists include those with direct and indirect effects (ephedrine, pseudoephedrine, and phenylpropanolamine), those with only direct effects (midodrine and phenylephrine), and those with only indirect effects (methylphenidate and dextroamphetamine sulfate). The peripheral selective direct, ␣1 adrenoreceptor agonist midodrine is the only agent approved by the Food and Drug Administration (FDA) for the treatment of orthostatic hypotension.57 The pressor effect of midodrine is due to both arterial and venous constriction. The efficacy of this agent has been demonstrated in double-blind, placebo-controlled studies.45,57,105 Midodrine, the prodrug, is activated to desglymidodrine, the active ␣ adrenoreceptor agonist. Midodrine is rapidly absorbed from the gastrointestinal tract. There is 93% bioavailability after oral administration. The peak plasma concentration of midodrine occurs in 20 to 40 minutes and the half-life is 30 minutes. Midodrine has an elimination half-life of 0.5 hours and is undetectable in plasma 2 hours after an oral dose. There is minimal protein binding, and the agent undergoes enzymatic hydrolysis (deglycination) in the systemic circulation to form the active agent, desglymidodrine. Desglymidodrine is 15 times more potent than midodrine and is primarily responsible for the therapeutic effect. The elimination half-life of desglymidodrine is 2 to 4 hours. Desglymidodrine is predominantly excreted by the kidneys. Patient sensitivity to this agent varies, and the dose should be titrated from 2.5 to 10 mg three times a day. The peak

Table 120–2. Volume Expansion Measures Fluid and salt administration Acute water ingestion 9-␣-Fluorohydrocortisone Vasopressin analogues

2676

Diseases of the Peripheral Nervous System

Table 120–3. Sympathomimetic Agents Midodrine Ephedrine Pseudoephedrine Phenylpropanolamine Methylphenidate Dextroamphetamine

effect of this agent occurs 1 hour after ingestion.105 Potential side effects of midodrine include pilomotor reactions, pruritus, supine hypertension, gastrointestinal complaints, and urinary retention. Central nervous system side effects occur infrequently. The mixed ␣ adrenoreceptor agonists—which act directly on the ␣ adrenoreceptor and release norepinephrine from the postganglionic sympathetic neuron—include ephedrine, pseudoephedrine, and phenylpropanolamine. There are structural, pharmacologic, and therapeutic differences between these agents. Ephedrine is an agonist of ␣, ␤1, and ␤2 receptors. The ␤2 vasodilatory effects may attenuate the pressor effect of this drug. Pseudoephedrine, a stereoisomer of ephedrine, and phenylpropanolamine (d,l-norephedrine), an ephedrine metabolite, have similar pharmacologic and therapeutic properties. Phenylpropanolamine may have less ␤ adrenoreceptor agonism than ephedrine and fewer central sympathomimetic effects.42 Typical doses of these indirect agonists are ephedrine, 25 to 50 mg three times a day; pseudoephedrine, 30 to 60 mg three times a day; and phenylpropanolamine, 12.5 to 25 mg three times a day. Because the effectiveness of the indirect agonists is at least in part due to the release of norepinephrine from the postganglionic neuron, these medications are in theory most likely to benefit patients with partial or incomplete lesions.4,17,24,27 The indirect agonists, methylphenidate and dextroamphetamine, that release norepinephrine from postganglionic neurons are infrequently used to treat orthostatic hypotension. There are few head-to-head comparisons of the ␣ adrenoreceptor agonists. In a small clinical trial midodrine (mean dose 8.4 mg tid) improved standing blood pressure and orthostatic tolerance more than ephedrine (22.3 mg tid).22 In another trial, phenylpropanolamine (12.5 mg) and yohimbine (5.4 mg) produced equivalent increases in standing systolic blood pressure while methylphenidate failed to increase standing systolic blood pressure significantly.42 The use of the sympathomimetic agents (with the possible exception of midodrine) may be complicated by tachyphylaxis, although efficacy may be regained after a short drug holiday. The central sympathomimetic side effects such as anxiety, tremulousness, and tachycardia that invariably accompany the use of these agents are frequently intolerable to patients. Midodrine, which does not cross

the blood-brain barrier in significant amounts, does not have these central sympathomimetic side effects. Although phenylpropanolamine may result in less central nervous system stimulation than the other agents, recent data suggest that phenylpropanolamine increases the risk of hemorrhagic stroke in women.48 The FDA is examining the toxicity of this and other indirect and mixed ␣ adrenoreceptor antagonists and has suggested that phenylpropanolamine be removed from all drug products. Severe hypertension is an important adverse effect of all sympathomimetic agents. Patients should not take these medications for 4 hours prior to recumbency. Patients taking midodrine report piloerection, pruritus, and tingling. These are most likely peripheral sympathomimetic effects of the drug.

Supplementary Therapy and Experimental Agents There are rare patients who do not respond to first-line interventions and require additional agents to treat their symptoms. The supplementary and experimental agents can be considered in such patients (Table 120–4). Caffeine The methylxanthine caffeine has a well-established pressor effect that is in part due to blockade of vasodilating adenosine receptors. Caffeine improves orthostatic hypotension and attenuates postprandial hypotension in patients with autonomic failure. Typical caffeine doses are 100 to 250 mg three times a day, either as tablets or caffeinated beverages (one cup of coffee contains approximately 85 mg of caffeine and one cup of tea contains 50 mg of caffeine).70 Erythropoietin Erythropoietin increases standing blood pressure and improves orthostatic tolerance in patients with orthostatic hypotension.39,75 This agent corrects the normochromic

Table 120–4. Supplementary Measures and Experimental Measures Caffeine Erythropoietin Cyclooxygenase inhibitors ␤-Blockers Clonidine Yohimbine Somatostatin Dihydroxyphenylserine Dihydroergotamine Dopamine antagonists Monoamine oxidase inhibition with tyramine Acetylcholinesterase inhibition

Treatment of Cardiovascular Autonomic Failure

normocytic anemia that frequently accompanies autonomic failure6,104 and diabetic autonomic neuropathy.103 Recombinant human erythropoietin, or erythropoietin alfa, is administered subcutaneously or intravenously at doses between 25 and 75 U/kg three times a week until a hematocrit that approaches normal is attained. Lower maintenance doses (approximately 25 U/kg three times a week) may then be used. Iron supplementation is usually required, particularly during the period when the hematocrit is increasing. The mechanism of action for the pressor effect of this agent is unresolved. Possibilities include the increase in red cell mass and central blood volume, alterations in blood viscosity, and direct or indirect neurohumoral effects on the vascular wall. There is evidence that the effect of erythropoietin is related to vascular tone regulation mediated by the interaction between hemoglobin and the vasodilator nitric oxide.76 Supine hypertension may accompany the use of this agent.39,75 There is no support for the use of erythropoietin in the treatment of orthostatic tachycardia.36 Cyclooxygenase Inhibitors The cyclooxygenase inhibitors, such as indomethacin, flurbiprofen, and other nonsteroidal anti-inflammatory agents, may also reduce orthostatic hypotension.97 These medications are rarely effective as monotherapy, but may be used to supplement treatment with 9-␣-fluorohydrocortisone or a sympathomimetic agent. The probable mode of action of the prostaglandin synthesis inhibitors is to limit the vasodilating effects of circulating prostaglandins and arachidonic acid derivatives. These agents may also increase the central circulating blood volume and enhance vascular sensitivity to circulating pressor amines.51 ␤-Blockers Nonselective ␤-blockers, particularly those with intrinsic sympathomimetic activity such as pindolol and xamoterol, may have a limited place in the treatment of orthostatic hypotension despite the well-acknowledged negative chronotropy and inotropy associated with these medications.14,66,68,100 The suggested mechanism of action of these medications is the blockade of vasodilating ␤2 receptors, allowing unopposed ␣ adrenoreceptor–mediated vasoconstrictor effects to dominate. Congestive heart failure may be a serious side effect of this medication.12,62 The pressor effect of these agents has not been demonstrated in all clinical trials.53 Clonidine Clonidine is an ␣2 antagonist that usually produces a central, sympatholytic effect and a consequent decrease in blood pressure. In patients with autonomic failure, who have little central sympathetic efferent activity, the effect of this agent on postsynaptic ␣2 adrenoreceptors may predominate. These receptors may be more numerous on veins than arterioles. The use of clonidine (0.1 to 0.6 mg/day) could therefore

2677

result in an increase in venous return without a significant increase in peripheral vascular resistance. The mechanism of action is still not clearly defined.96 The use of this agent, at least theoretically, is limited to patients with severe central autonomic dysfunction in whom there is no ostensible effect of further sympatholysis and the peripheral effect may dominate. The hypertensive effect is inconsistent, and in some patients residual sympathetic activity could be inhibited. The agent may cause profound hypotension in patients with autonomic failure. Other side effects include dry mouth and sedation.78,79 Yohimbine Yohimbine is a centrally and peripherally active selective ␣2 adrenoreceptor antagonist that increases sympathetic nervous system efferent output by antagonizing central or presynaptic ␣2 adrenoreceptors or both. Yohimbine (2.5 to 5.4 mg tid) enhances norepinephrine release from postganglionic sympathetic neurons and produces a modest pressor effect. This agent, theoretically, should be more effective in patients who have some residual sympathetic nervous system output, although this has not always been borne out in clinical studies. Side effects of yohimbine include anxiety, tremor, palpitations, diarrhea, and supine hypertension.42,71 Somatostatin Somatostatin and somatostatin analogues such as octreotide, the long-acting synthetic octapeptide, attenuate the pancreatic and gastrointestinal hormone response to food ingestion and other stimuli. These agents attenuate the postprandial blood pressure fall and reduce orthostatic hypotension in patients with autonomic failure. Mechanisms of action for these medications include a local effect on splanchnic vasculature by inhibiting the release of vasoactive gastrointestinal peptides, enhanced cardiac output, and an increase in forearm and splanchnic vascular resistance. Subcutaneous doses of octreotide range from 25 to 200 ␮g. Side effects of nausea and abdominal cramps limit the use of these agents.37,38 Dihydroxyphenylserine Dihydroxyphenylserine (DOPS) is a synthetic, nonphysiologic amino acid norepinephrine precursor that is decarboxylated by the ubiquitous L-amino acid decarboxylase (L-AADC) to L-norepinephrine in both animals and humans (Fig. 120–3). The important role played by norepinephrine in the maintenance of upright blood pressure provides the rationale for the use of this agent to treat neurogenic orthostatic hypotension. Of the four stereoisomers D- and L-threo-DOPS and D- and L-erythro-DOPS, only L-threo-DOPS is pharmacologically and biologically active.3 This agent effectively treats patients with dopamine ␤-hydroxylase deficiency who are unable to synthesize norepinephrine and epinephrine in the central and peripheral nervous systems. Because the conversion

2678

Diseases of the Peripheral Nervous System

OH H

COO

NH2 HO L-DOPS L-aromatic amino-acid decarboxylase

OH H NH2 HO

Norepinephrine

FIGURE 120–3 The biosynthesis of norepinephrine from L-DOPS. (Modified from Freeman, R., Landsberg, L., and Young, J.: The treatment of neurogenic orthostatic hypotension with 3,4-DL-threodihydroxyphenylserine: a randomized, placebo-controlled, crossover trial. Neurology 53:2151, 1999.)

of DOPS to norepinephrine bypasses the ␤-hydroxylation step, DOPS is the ideal therapeutic agent for this inherited disorder.5,61 In addition, both racemic DL-DOPS and L-DOPS produce a clinically significant pressor effect that is associated with an increase in the plasma concentration of norepinephrine. The effect is present in patients with orthostatic hypotension25,47,64 and postprandial hypotension26 caused by peripheral and central autonomic disorders. Of note, oral administration of carbidopa, which inhibits L-AADC outside the blood-brain barrier, blunts both the increase in plasma norepinephrine and the pressor response to L-DOPS, suggesting that DOPS will not be effective if administered to parkinsonian patients on L-DOPA combination therapy.47 This drug is not commercially available at present. Dihydroergotamine Dihydroergotamine, an ergot alkaloid that interacts with ␣-adrenergic receptors, has a selective venoconstrictor effect.21,35 This medication may increase venous return in patients with orthostatic hypotension without producing a significant increase in peripheral vascular resistance. Although dihydroergotamine is an effective pressor intravenously and intramuscularly, low oral bioavailability results in an inconsistent effect when taken orally.41 The use of inhaled ergotamine may avoid the inconsistent bioavailability.7 Dopamine Antagonists The dopamine antagonists metoclopramide and domperidone may also treat orthostatic hypotension. These agents most likely inhibit the vasodilating and natriuretic effect of dopamine or increase noradrenaline release as a result of

blockade of prejunctional inhibitory dopaminergic receptors. The risk of tardive dyskinesia and other extrapyramidal side effects limits the long-term use of these agents.56 Metoclopramide may have a hypotensive effect in some patients with autonomic failure.60 Monoamine Oxidase Inhibition with Tyramine There were initial optimistic reports on the use of the indirect-acting agent tyramine, which releases norepinephrine from neuronal storage pools, in combination with a monoamine oxidase inhibitor to prevent breakdown of the released norepinephrine.44,67 The use of this combination of pharmacologic agents is unfortunately limited by severe supine hypertension, an unpredictable response, and, in some cases, the failure to abolish orthostatic symptoms. Acetylcholinesterase Inhibition The acetylcholinesterase inhibitor pyridostigmine, 60 mg administered orally, increased head-up tilt blood pressure and reduced orthostatic hypotension in patients with neurogenic orthostatic hypotension in an open-label trial. The blood pressure increase was associated with an increase in peripheral resistance. Supine blood pressure and peripheral resistance increased modestly. Twenty percent of patients reported cholinergic side effects. The rationale for the use of this agent is that inhibition of acetylcholinesterase would enhance sympathetic ganglionic neurotransmission.90

CONCLUSION Left untreated, orthostatic hypotension can be an incapacitating symptom of autonomic failure. Fortunately, with the help of nonpharmacologic and pharmacologic interventions, most patients’ quality of life can be improved substantially. Treatment should be initiated in stages. First, patients should be counseled as to the nature of the underlying disorder. Simultaneously, reversible causes of orthostatic hypotension should be removed. Should symptoms persist, pharmacologic treatment is implemented. Many patients will be substantially improved by education, counseling, removal of hypotensive medications, and other nonpharmacologic interventions, while more severely afflicted patients require pharmacologic interventions.

REFERENCES 1. Bannister, R., and Oppenheimer, D. R.: Degenerative diseases of the nervous system associated with autonomic failure. Brain 95:457, 1972. 2. Bannister, R., Sever, P., and Gross, M.: Cardiovascular reflexes and biochemical responses in progressive autonomic failure. Brain 100:327, 1977.

Treatment of Cardiovascular Autonomic Failure 3. Bartholini, J., Constantinidis, J., Puig, M., et al.: The stereoisomers of 3,4-dihydroxyphenylserine as precursors of norepinephrine. J. Pharmacol. Exp. Ther. 193:523, 1975. 4. Biaggioni, I., Onrot, J., Stewart, C. K., and Robertson, D.: The potent pressor effect of phenylpropanolamine in patients with autonomic impairment. JAMA 258:236, 1987. 5. Biaggioni, I., and Robertson, D.: Endogenous restoration of noradrenaline by precursor therapy in dopamine-betahydroxylase deficiency. Lancet 2:1170, 1987. 6. Biaggioni, I., Robertson, D., Krantz, S., et al.: The anemia of primary autonomic failure and its reversal with recombinant erythropoietin. Ann. Intern. Med. 121:181, 1994. 7. Biaggioni, I., Zygmunt, D., Haile, V., and Robertson, D.: Pressor effect of inhaled ergotamine in orthostatic hypotension. Am. J. Cardiol. 65:89, 1990. 8. Bonuccelli, U., Lucetti, C., Del Dotto, P., et al.: Orthostatic hypotension in de novo Parkinson disease. Arch. Neurol. 60:1400, 2003. 9. Borst, C., Weiling, W., van Brederode, J. F. M., et al.: Mechanisms of initial heart rate response to postural change. Am. J. Physiol. 243:H676, 1982. 10. Bradbury, S., and Eggleston, C.: Postural hypotension: report of 3 cases. Am. Heart J. 1:73, 1925. 11. Campbell, I. W., Ewing, D. J., and Clarke, B. F.: 9-Alphafluorohydrocortisone in the treatment of postural hypotension in diabetic autonomic neuropathy. Diabetes 24:381, 1975. 12. Chobanian, A. V., Volicer, L., Liang, C. S., et al.: Use of propranolol in the treatment of idiopathic orthostatic hypotension. Trans. Assoc. Am. Physicians 90:324, 1977. 13. Chobanian, A. V., Volicer, L., Tifft, C. P., et al.: Mineralocorticoid-induced hypertension in patients with orthostatic hypotension. N. Engl. J. Med. 301:68, 1979. 14. Cleophas, T. J., Grabowsky, I., Niemeyer, M. G., et al.: Paradoxical pressor effects of beta-blockers in standing elderly patients with mild hypertension: a beneficial side effect. Circulation 105:1669, 2002. 15. Cohen, J., Low, P., Fealey, R., et al.: Somatic and autonomic function in progressive autonomic failure and multiple system atrophy. Ann. Neurol. 22:692, 1987. 16. Dampney, R. A., Coleman, M. J., Fontes, M. A., et al.: Central mechanisms underlying short- and long-term regulation of the cardiovascular system. Clin. Exp. Pharmacol. Physiol. 29:261, 2002. 17. Davies, B., Bannister, R., and Sever, P.: Pressor amines and monoamine-oxidase inhibitors for treatment of postural hypotension in autonomic failure: limitations and hazards. Lancet 1:172, 1978. 18. Davies, I. B., Bannister, R. G., Sever, P. S., and Wilcox, C. S.: Fludrocortisone in the treatment of postural hypotension: altered sensitivity to pressor agents [proceedings]. Br. J. Clin. Pharmacol. 6:444P, 1978. 19. The definition of orthostatic hypotension, pure autonomic failure, and multiple system atrophy. J. Auton. Nerv. Syst. 58:123, 1996. 20. Dobkin, B. H.: Orthostatic hypotension as a risk factor for symptomatic occlusive cerebrovascular disease. Neurology 39:30, 1989. 21. Fouad, F. M., Tarazi, R. C., and Bravo, E. L.: Dihydroergotamine in idiopathic orthostatic hypotension: short-term intramuscular and long-term oral therapy. Clin. Pharmacol. Ther. 30:782, 1981.

2679

22. Fouad-Tarazi, F. M., Okabe, M., and Goren, H.: Alpha sympathomimetic treatment of autonomic insufficiency with orthostatic hypotension. Am. J. Med. 99:604, 1995. 23. Freeman, R.: Pure autonomic failure. In Robertson, D., and Biaggioni, I. (eds.): Disorders of the Autonomic Nervous System. Luxembourg, Harwood Academic Publishers, p. 83, 1995. 24. Freeman, R.: Treatment of orthostatic hypotension— midodrine and other pressor drugs. In Robertson, D., Low, P. A., and Polinsky, R. J. (eds.): Primer on the Autonomic Nervous System. New York, Academic Press, p. 326, 1996. 25. Freeman, R., Landsberg, L., and Young, J.: The treatment of neurogenic orthostatic hypotension with 3,4-DL-threodihydroxyphenylserine: a randomized, placebo-controlled, crossover trial. Neurology 53:2151, 1999. 26. Freeman, R., Young, J. B., Landsberg, L., and Lipsitz, L. A.: The treatment of postprandial hypotension in autonomic failure with 3,4-DL-threo-dihydroxyphenylserine. Neurology 47:1414, 1996. 27. Ghrist, D. G., and Brown, G. E.: Postural hypertension with syncope: its successful treatment with ephedrine. Am. J. Med. Sci. 175:336, 1928. 28. Gilman, S., Low, P. A., Quinn, N., et al.: Consensus statement on the diagnosis of multiple system atrophy. J. Auton. Nerv. Syst. 74:189, 1998. 29. Goetz, C. G., Lutge, W., and Tanner, C. M.: Autonomic dysfunction in Parkinson’s disease. Neurology 36:73, 1986. 30. Goldstein, D. S., Holmes, C., Sharabi, Y., et al.: Plasma levels of catechols and metanephrines in neurogenic orthostatic hypotension. Neurology 60:1327, 2003. 31. Goldstein, D. S., Holmes, C. S., Dendi, R., et al.: Orthostatic hypotension from sympathetic denervation in Parkinson’s disease. Neurology 58:1247, 2002. 32. Goldstein, D. S., Polinsky, R. J., Garty, M., et al.: Patterns of plasma levels of catechols in neurogenic orthostatic hypotension. Ann. Neurol. 26:558, 1989. 33. Hague, K., Lento, P., Morgello, S., et al.: The distribution of Lewy bodies in pure autonomic failure: autopsy findings and review of the literature. Acta Neuropathol. (Berl). 94:192, 1997. 34. Hickler, R. B., Thompson, G. R., Fox, L. M., and Hamlin, J. T.: Successful treatment of orthostatic hypotension with 9-alphafluorohydrocortisone. N. Engl. J. Med. 261:788, 1959. 35. Hoeldtke, R. D., Cavanaugh, S. T., Hughes, J. D., and Polansky, M.: Treatment of orthostatic hypotension with dihydroergotamine and caffeine. Ann. Intern. Med. 105:168, 1986. 36. Hoeldtke, R. D., Horvath, G. G., and Bryner, K. D.: Treatment of orthostatic tachycardia with erythropoietin. Am. J. Med. 99:525, 1995. 37. Hoeldtke, R. D., Horvath, G. G., Bryner, K. D., and Hobbs, G. R.: Treatment of orthostatic hypotension with midodrine and octreotide. J. Clin. Endocrinol. Metab. 83:339, 1998. 38. Hoeldtke, R. D., and Israel, B. C.: Treatment of orthostatic hypotension with octreotide. J. Clin. Endocrinol. Metab. 68:1051, 1989. 39. Hoeldtke, R. D., Streeten, D. H. P., and Streeten, D. H.: Treatment of orthostatic hypotension with erythropoietin. N. Engl. J. Med. 329:611, 1993.

2680

Diseases of the Peripheral Nervous System

40. Jacob, G., Ertl, A. C., Shannon, J. R., et al.: Effect of standing on neurohumoral responses and plasma volume in healthy subjects. J. Appl. Physiol. 84:914, 1998. 41. Jennings, G., Esler, M., and Holmes, R.: Treatment of orthostatic hypotension with dihydroergotamine. Br. Med. J. 2:307, 1979. 42. Jordan, J., Shannon, J. R., Biaggioni, I., et al.: Contrasting actions of pressor agents in severe autonomic failure. Am. J. Med. 105:116, 1998. 43. Jordan, J., Shannon, J. R., Black, B. K., et al.: The pressor response to water drinking in humans: a sympathetic reflex? Circulation 101:504, 2000. 44. Karet, F. E., Dickerson, J. E., Brown, J., and Brown, M. J.: Bovril and moclobemide: a novel therapeutic strategy for central autonomic failure. Lancet 344:1263, 1994. 45. Kaufmann, H., Brannan, T., Krakoff, L., et al.: Treatment of orthostatic hypotension due to autonomic failure with a peripheral alpha-adrenergic agonist (midodrine). Neurology 38:951, 1988. 46. Kaufmann, H., Oribe, E., Miller, M., et al.: Hypotensioninduced vasopressin release distinguishes between pure autonomic failure and multiple system atrophy with autonomic failure. Neurology 42:590, 1992. 47. Kaufmann, H., Saadia, D., Voustianiouk, A., et al.: Norepinephrine precursor therapy in neurogenic orthostatic hypotension. Circulation 108:724, 2003. 48. Kernan, W. N., Viscoli, C. M., Brass, L. M., et al.: Phenylpropanolamine and the risk of hemorrhagic stroke. N. Engl. J. Med. 343:1826, 2000. 49. Kircheim, H. R.: Systemic arterial baroreceptor reflexes. Physiol. Rev. 56:100, 1976. 50. Kochar, M. S.: Hemodynamic effects of lysinevasopressin in orthostatic hypotension. Am. J. Kidney Dis. 6:49, 1985. 51. Kochar, M. S., and Itskovitz, H. D.: Treatment of idiopathic orthostatic hypotension (Shy-Drager syndrome) with indomethacin. Lancet 1:1011, 1978. 52. Lake, C. R., Ziegler, M. G., and Kopin, I. J.: Use of plasma norepinephrine for evaluation of sympathetic neuronal function in man. Life Sci. 18:1315, 1976. 53. Leslie, P. J., Thompson, C., Clarke, B. F., and Ewing, D. J.: A double-blind crossover study of oral xamoterol in postural hypotension due to diabetic autonomic neuropathy. Clin. Auton. Res. 1:119, 1991. 54. Lipsitz, L. A., Ryan, S. M., Parker, J. A., et al.: Hemodynamic and autonomic nervous system responses to mixed meal ingestion in healthy young and old subjects and dysautonomic patients with postprandial hypotension. Circulation 87:391, 1993. 55. Loewy, A. D., and Neil, J. J.: The role of descending monoaminergic systems in central control of blood pressure. Fed. Proc. 40:56, 1981. 56. Lopes de Faria, S. R., Zanella, M. T., Andriolo, A., et al.: Peripheral dopaminergic blockade for the treatment of diabetic orthostatic hypotension. Clin. Pharmacol. Ther. 44:670, 1988. 57. Low, P. A., Gilden, J. L., Freeman, R., et al.: Efficacy of midodrine vs placebo in neurogenic orthostatic hypotension: a randomized, double blind multicenter study. Midodrine Study Group. JAMA 277:1046, 1997.

58. Low, P. A., Vernino, S., and Suarez, G.: Autonomic dysfunction in peripheral nerve disease. Muscle Nerve 27:646, 2003. 59. MacLean, A. R., and Allen, B. V.: Orthostatic hypotension and orthostatic tachycardia: treatment with the “head-up” bed. JAMA 115:2162, 1940. 60. Magnifico, F., Pierangeli, G., Barletta, G., et al.: The cardiovascular effects of metoclopramide in multiple system atrophy and pure autonomic failure. Clin. Auton. Res. 11:163, 2001. 61. Man in’t Veld, A. J., Boomsma, F., van den Meiracker, A. H., and Schalekamp, M. A.: Effect of unnatural noradrenaline precursor on sympathetic control and orthostatic hypotension in dopamine-beta-hydroxylase deficiency. Lancet 2:1172, 1987. 62. Man in’t Veld, A. J., and Schalekamp, M. A.: Pindolol acts as beta-adrenoceptor agonist in orthostatic hypotension: therapeutic implications. Br. Med. J. (Clin. Res. Ed.) 282:929, 1981. 63. Mathias, C. J., Fosbraey, P., da Costa, D. F., et al.: The effect of desmopressin on nocturnal polyuria, overnight weight loss, and morning postural hypotension in patients with autonomic failure. Br. Med. J. (Clin. Res. Ed.) 293:353, 1986. 64. Mathias, C. J., Senard, J. M., Braune, S., et al.: L-threodihydroxyphenylserine (L-threo-DOPS; Droxidopa) in the management of neurogenic orthostatic hypotension: a multi-national, multi-center, dose-ranging study in multiple system atrophy and pure autonomic failure. Clin. Auton. Res. 11:235, 2001. 65. McLeod, J. G.: Invited review: autonomic dysfunction in peripheral nerve disease. Muscle Nerve 15:3, 1992. 66. Mehlsen, J., and Trap-Jensen, J.: Xamoterol, a new selective beta-1-adrenoceptor partial agonist, in the treatment of postural hypotension. Acta Med. Scand. 219:173, 1986. 67. Nanda, R. N., Johnson, R. H., and Keogh, H. J.: Treatment of neurogenic orthostatic hypotension with a monoamine oxidase inhibitor and tyramine. Lancet 2:1164, 1976. 68. Obara, A., Yamashita, H., Onodera, S., et al.: Effect of xamoterol in Shy-Drager syndrome. Circulation 85:606, 1992. 69. Omboni, S., Smit, A. A., van Lieshout, J. J., et al.: Mechanisms underlying the impairment in orthostatic tolerance after nocturnal recumbency in patients with autonomic failure. Clin. Sci. (Lond.) 101:609, 2001. 70. Onrot, J., Goldberg, M. R., Biaggioni, I., et al.: Hemodynamic and humoral effects of caffeine in autonomic failure: therapeutic implications for postprandial hypotension. N. Engl. J. Med. 313:549, 1985. 71. Onrot, J., Goldberg, M. R., Biaggioni, I., et al.: Oral yohimbine in human autonomic failure. Neurology 37:215, 1987. 72. Ozawa, T., Oyanagi, K., Tanaka, H., et al.: Suprachiasmatic nucleus in a patient with multiple system atrophy with abnormal circadian rhythm of arginine-vasopressin secretion into plasma. J. Neurol. Sci. 154:116, 1998. 73. Ozawa, T., Tanaka, H., Nakano, R., et al.: Nocturnal decrease in vasopressin secretion into plasma in patients with multiple system atrophy. J. Neurol. Neurosurg. Psychiatry 67:542, 1999. 74. Parikh, S. M., Diedrich, A., Biaggioni, I., and Robertson, D.: The nature of the autonomic dysfunction in multiple system atrophy. J. Neurol. Sci. 200:1, 2002.

Treatment of Cardiovascular Autonomic Failure 75. Perera, R., Isola, L., and Kaufmann, H.: Effect of recombinant erythropoietin on anemia and orthostatic hypotension in primary autonomic failure. Clin. Auton. Res. 5:211, 1995. 76. Rao, S. V., and Stamler, J. S.: Erythropoietin, anemia, and orthostatic hypotension: the evidence mounts. Clin. Auton. Res. 12:141, 2002. 77. Robertson, D.: Contraindication to the use of ocular phenylephrine in idiopathic orthostatic hypotension. Am. J. Ophthalmol. 87:819, 1979. 78. Robertson, D., Goldberg, M. R., Hollister, A. S., et al.: Clonidine raises blood pressure in idiopathic orthostatic hypotension. Am. J. Med. 74:193, 1983. 79. Robertson, D., Goldberg, M. R., Tung, C. S., et al.: Use of alpha 2 adrenoreceptor agonists and antagonists in the functional assessment of the sympathetic nervous system. J. Clin. Invest. 78:576, 1986. 80. Robertson, D., Kincaid, D. W., Haile, V., and Robertson, R. M.: The head and neck discomfort of autonomic failure: an unrecognized aetiology of headache. Clin. Auton. Res. 4:993, 1994. 81. Robertson, D., Wade, D., and Robertson, R. M.: Postprandial alterations in cardiovascular hemodynamics in autonomic dysfunctional states. Am. J. Cardiol. 48:1048, 1981. 82. Roose, S. P., Glassman, A. H., Siris, S. G., et al.: Comparison of imipramine- and nortriptyline-induced orthostatic hypotension: a meaningful difference. J. Clin. Psychopharmacol. 1:316, 1981. 83. Sakakibara, R., Hattori, T., Uchiyama, T., et al.: Urinary dysfunction and orthostatic hypotension in multiple system atrophy: which is the more common and earlier manifestation? J. Neurol. Neurosurg. Psychiatry 68:65, 2000. 84. Sakakibara, R., Matsuda, S., Uchiyama, T., et al.: The effect of intranasal desmopressin on nocturnal waking in urination in multiple system atrophy patients with nocturnal polyuria. Clin. Auton. Res. 13:106, 2003. 85. Schatz, I. J., Miller, M. J., and Frame, B.: Corticosteroids in the management of orthostatic hypotension. Cardiology 61(Suppl. 1):280, 1976. 86. Senard, J. M., Brefel-Courbon, C., Rascol, O., and Montastruc, J. L.: Orthostatic hypotension in patients with Parkinson’s disease: pathophysiology and management. Drugs Aging 18:495, 2001. 87. Shannon, J., Jordan, J., Costa, F., et al.: The hypertension of autonomic failure and its treatment. Hypertension 30:1062, 1997. 88. Shannon, J. R., Jordan, J., Diedrich, A., et al.: Sympathetically mediated hypertension in autonomic failure. Circulation 101:2710, 2000. 89. Shy, G. M., and Drager, G. A.: A neurological syndrome associated with orthostatic hypotension. Arch. Neurol. 2:511, 1960. 90. Singer, W., Opfer-Gehrking, T. L., McPhee, B. R., et al.: Acetylcholinesterase inhibition: a novel approach in the treatment of neurogenic orthostatic hypotension. J. Neurol. Neurosurg. Psychiatry 74:1294, 2003.

2681

91. Smit, A. A., Halliwill, J. R., Low, P. A., and Wieling, W.: Pathophysiological basis of orthostatic hypotension in autonomic failure. J. Physiol. (Lond.) 519(Pt. 1):1, 1999. 92. Spyer, K. M.: Neural organisation and control of the baroreceptor reflex. Rev. Physiol. Biochem. Pharmacol. 88:23, 1981. 93. van Dijk, J. G., Haan, J., Zwinderman, K., et al.: Autonomic nervous system dysfunction in Parkinson’s disease: relationships with age, medication, duration, and severity. J. Neurol. Neurosurg. Psychiatry 56:1090, 1993. 94. van Lieshout, J. J., ten Harkel, A. D., and Wieling, W.: Physical manoeuvres for combating orthostatic dizziness in autonomic failure. Lancet 339:897, 1992. 95. van Lieshout, J. J., ten Harkel, A. D., and Wieling, W.: Fludrocortisone and sleeping in the head-up position limit the postural decrease in cardiac output in autonomic failure. Clin. Auton. Res. 10:35, 2000. 96. Victor, R. G., and Talman, W. T.: Comparative effects of clonidine and dihydroergotamine on venomotor tone and orthostatic tolerance in patients with severe hypoadrenergic orthostatic hypotension. Am. J. Med. 112:361, 2002. 97. Watt, S. J., Tooke, J. E., Perkins, C. M., and Lee, M. R.: The treatment of idiopathic orthostatic hypotension: a combined fludrocortisone and flurbiprofen regime. Q. J. Med. 50:205, 1981. 98. Wenning, G. K., Scherfler, C., Granata, R., et al.: Time course of symptomatic orthostatic hypotension and urinary incontinence in patients with postmortem confirmed parkinsonian syndromes: a clinicopathological study [see comments]. J. Neurol. Neurosurg. Psychiatry 67:620, 1999. 99. Wenning, G. K., Tison, F., Ben Shlomo, Y., et al.: Multiple system atrophy: a review of 203 pathologically proven cases. Mov. Disord. 12:133, 1997. 100. West, J. N., Stallard, T. J., Dimmitt, S. B., et al.: Xamoterol in the treatment of orthostatic hypotension associated with multiple system atrophy (Shy-Drager syndrome). Q. J. Med. 74:209, 1990. 101. Wieling, W., van Lieshout, J. J., and Hainsworth, R.: Extracellular fluid volume expansion in patients with posturally related syncope. Clin. Auton. Res. 12:242, 2002. 102. Wieling, W., van Lieshout, J. J., and van Leeuwen, A. M.: Physical manoeuvres that reduce postural hypotension in autonomic failure. Clin. Auton. Res. 3:57, 1993. 103. Winkler, A. S., Landau, S., Watkins, P., and Chaudhuri, K. R.: Observations on haematological and cardiovascular effects of erythropoietin treatment in multiple system atrophy with sympathetic failure. Clin. Auton. Res. 12:203, 2002. 104. Winkler, A. S., Marsden, J., Parton, M., et al.: Erythropoietin deficiency and anaemia in multiple system atrophy. Mov. Disord. 16:233, 2001. 105. Wright, R. A., Kaufmann, H. C., Perera, R., et al.: A doubleblind, dose-response study of midodrine in neurogenic orthostatic hypotension. Neurology 51:120, 1998. 106. Ziegler, M. G., Lake, C. R., and Kopin, I. J.: The sympatheticnervous-system defect in primary orthostatic hypotension. N. Engl. J. Med. 296:293, 1977.

121 Management of Gut Dysmotility ADIL E. BHARUCHA AND MICHAEL CAMILLERI

Regulation of Gastrointestinal Sensorimotor Function by Extrinsic Nerves Gastrointestinal Manifestations of Autonomic Neuropathy Oropharyngeal Dysphagia Gastroparesis and Chronic Intestinal Pseudo-obstruction Constipation

Megacolon Diarrhea Fecal Incontinence Assessment of Autonomic Function in the Gastrointestinal Tract Pancreatic Polypeptide Response to Hypoglycemia or to Modified Sham Feeding by the “Chew and Spit” Technique

The influence of autonomic nerves on gastrointestinal motility was first demonstrated by severe diarrhea in animals after a ganglionectomy. Studies by Bayliss and Starling10–12 provided the foundation for understanding extrinsic neural regulation of gastrointestinal motility. Autonomic dysfunction can be associated with symptomatic or asymptomatic gastrointestinal sensory or motor dysfunction at one or more levels in the gastrointestinal tract, ranging from the pharynx to the anal sphincter. Diabetic gastroparesis is the preeminent example of a gastrointestinal motility disorder attributed to autonomic neuropathy. Autonomic dysfunction is frequent, occurring in approximately 66% of patients with diabetes and other neurologic syndromes who have symptoms of gastrointestinal dysmotility.15 The spectrum of gastrointestinal motility disturbances attributable to an autonomic neuropathy may be wider because one or more autonomic function tests are abnormal in approximately 25% of patients with a functional gastrointestinal disorder; these patients do not have other manifestations or an underlying disease associated with a neurologic disorder.2,15 However it is unknown if the association between an autonomic neuropathy and a functional gastrointestinal disorder is coincidental or reflects cause and effect. This chapter focuses on autonomic neuropathies associated with gastrointestinal dysmotility, but also reviews other neurologic disorders associated with gastrointestinal dysmotility.

R-R Interval Response to Deep Breathing Mesenteric Flow in Response to Tilt-Table Testing Extrinsic Neurologic Disorders Causing Gut Dysmotility Central Neurologic Disorders Autonomic System Degenerations Peripheral Neuropathy

REGULATION OF GASTROINTESTINAL SENSORIMOTOR FUNCTION BY EXTRINSIC NERVES Gastrointestinal motility is regulated by the intrinsic or enteric nervous system, which contains the same number (~100 million) of neurons as the spinal cord. The enteric nervous system is organized into ganglionated plexuses, of which two predominant plexuses are the myenteric and submucous. The extrinsic innervation of the gastrointestinal tract does not merely represent a link between the center (i.e., central nervous system) and the effector, but a connection between two centers, that is, the central or “big” brain and peripheral or “little” brain74 (Fig. 121–1). The enteric nervous system contains a number of neural circuits that control motor functions, local blood flow, and mucosal transport and secretions and modulate immune and endocrine functions.47 In addition to neurons and nerve fibers, these plexuses also contain interstitial cells of Cajal, which probably serve as pacemakers, conductors of electrical activity, and mediators of neurotransmission from enteric neurons.177 The myenteric plexus is primarily responsible for controlling motility, while the submucous plexus regulates mucosal secretion/absorption. These functions involve an integrated response to sensory input from the lumen (via chemo- or osmoreceptors) or the wall (via mechanoreceptors). Integration occurs in intrinsic or 2683

2684

Diseases of the Peripheral Nervous System

Extrinsic

Intrinsic

Parasympathetic “Little” brain

Vagal

ICCs Sympathetic Smooth muscle cell with receptors for transmitters modulates peristaltic reflex

Thoracolumbar

Inter

Inter

Motor Ach

Motor IPAN: sensory

NO

Sub P

VIP ATP

Ascending contraction

Descending relaxation

Distention by bolus

FIGURE 121–1 Control of gut motility. Sympathetic and parasympathetic pathways link the “big” brain (i.e., the central nervous system) to the “little” brain (i.e., the enteric nervous system). For clarity, only sympathetic and parasympathetic pathways to the upper gut are shown in this figure. The peristaltic reflex (insert) involves contraction proximal to and relaxation distal to a distending bolus, facilitating a broad propagation of gastrointestinal contents; afferent impulses are conveyed by intrinsic (IPAN) and extrinsic primary afferent neurons. Interactions between transmitters (e.g., peptides, amines) and receptors alter muscle membrane potentials by stimulating bidirectional ion fluxes. In turn, membrane characteristics dictate whether or not the muscle cell contracts. Primary excitatory neurotransmitters are acetylcholine (Ach) and tachykinins (e.g., substance P [Sub P]). The main inhibitory neurotransmitters are nitric oxide (NO), vasoactive intestinal peptide (VIP), and ATP. (Adapted from Camilleri, M., and Phillips, S. F.: Disorders of small intestinal motility. Gastroenterol. Clin. North Am. 18:405, 1989.)

extrinsic pathways and modulates function at a local level (e.g., peristaltic reflex), or over longer distances (e.g., the interdigestive migrating motor complex that spreads from the stomach to the ileum) (Fig. 121–2). Extrinsic sympathetic and parasympathetic pathways modulate the enteric nervous system.108,136 In general, extrinsic innervation has the most pronounced effects on the gastrointestinal tract lined by skeletal muscle (i.e., upper esophageal sphincter and external anal sphincter). Within segments lined by smooth muscle, the extrinsic nerves influence the stomach and distal colon to a greater degree than the small bowel. Thus Cannon observed that gastrointestinal smooth muscle functions reasonably well without extrinsic nerves after recovering from the immediate insult34; the extrinsic nerves primarily integrate activity in widely separated regions, often via reflexes that synapse in prevertebral ganglia.92 Parasympathetic and sympathetic pathways have

opposing effects on gastrointestinal motor function. Parasympathetic pathways are excitatory to smooth muscle, but inhibit sphincters. Sympathetic nerves inhibit intrinsic excitatory cholinergic neural circuits, but do not directly innervate muscle. Consequently reduced sympathetic inhibitory input (“the brake”) after extrinsic denervation may lead to excessive or uncoordinated gastrointestinal phasic pressure activity, including high-amplitude propagated contractions,100 which may account for diarrhea in patients who have an autonomic neuropathy42 (Fig. 121–3). Other consequences of extrinsic denervation include adrenergic hypersensitivity that may potentially compensate for loss of sympathetic tonic inhibition.144 Conversely, increased sensitivity at muscarinic M1 receptors may contribute to excessive motor activity after extrinsic denervation.99 Extrinsic pathways convey afferent input from the gut to the central nervous system via a classic three-neuron

2685

Management of Gut Dysmotility

FIGURE 121–2 Tracing showing normal upper gastrointestinal motility in the fasting (left) and fed (right) states. The fasting tracing shows phase III of the interdigestive migrating motor complex, normally propagated from the antroduodenal junction to the jejunum. Note irregular but persistent antral and intestinal phasic pressure activity after feeding. (From Malagelada, J.-R., Camilleri, M., and Stanghellini, V.: Manometric Diagnosis of Gastrointestinal Motility Disorders. New York, Thieme, 1986, by permission of Mayo Foundation.)

pathway through the brainstem reticular formation or the thalamus to higher centers33 (Fig. 121–4). These afferents are predominantly unmyelinated C afferent fibers that respond to low-threshold, polymodal stimulation and course along splanchnic sympathetic fibers to the central nervous system. Pain perception is dependent on higher intensity stimulation of naked nerve endings that act as mechanoreceptors. Descending pathways serve to modulate incoming afferent signals by inhibition or facilitation of dorsal horn neurons.

GASTROINTESTINAL MANIFESTATIONS OF AUTONOMIC NEUROPATHY Disorders of the autonomic nervous system manifest primarily as abnormalities in motor or sensory function; less frequently, absorption, secretion, or other digestive processes may be abnormal. Disorders at several levels of the neural axis (e.g., dysautonomias or autonomic neuropathies affecting peripheral pathways, spinal cord

transection, multiple sclerosis [MS], or strokes) may alter sensory or motor function in the gastrointestinal tract. The main gastrointestinal manifestations of autonomic neuropathies include dysphagia, gastroparesis and chronic intestinal pseudo-obstruction, constipation, diarrhea, and fecal incontinence.

Oropharyngeal Dysphagia Oropharyngeal dysphagia refers to difficulty in swallowing, with a holdup in the cervical area.46 Because patients with distal esophageal obstruction may also experience cervical dysphagia, the location of the symptom does not always reflect the site of holdup. Oropharyngeal dysphagia is generally caused by a disease, often systemic, affecting any level of the swallowing pathway, rather than an oropharyngeal disorder. Dysphagia afflicts up to 45% of patients with a stroke admitted to a hospital; however only 0.4% of survivors have dysphagia 6 months after a stroke.134 Dysphagia is a common49,85,184 and occasionally the presenting symptom of inflammatory myopathy,

2686

Diseases of the Peripheral Nervous System

FIGURE 121–3 Colonic high-amplitude propagated contractions (HAPCs) in a denervated bowel loop in the dog. Tracing shows several colonic HAPCs (arrows) accompanied by reduction in barostat balloon volume. IL ⫽ ileum; C ⫽ colon. Distances refer to location of sensors relative to the ileo-cecal sphincter (ICS). (From Leelakusolvong, S., Bharucha, A. E., Sarr, M. G., et al.: Effect of extrinsic denervation on muscarinic neurotransmission in the canine ileocolonic region. Neurogastroenterol. Motil. 15:173, 2003, with permission.)

occurring in 30% to 60% of cases. Pharyngeal dysphagia can also occur in up to 20% of cases of mixed connective tissue disease overlap syndrome, which may include clinical and serologic features of Sjögren’s syndrome, lupus,

or scleroderma.80,167 Significant dysphagia affects 30% to 60% of myasthenics,131 is noted at diagnosis in approximately 20% of cases,131,172 and may dominate the clinical picture. thalamus

afferent

dorsal horn

reticular formation

DRG spinothalamic & spinoreticular tracts

prevertebral ganglia

efferent DRG =dorsal root ganglion

FIGURE 121–4 Pathways involved in visceral sensation. As with somatic sensation, a three-neuron chain links gut to conscious perception. Sensation leads to reflex modulation of gut function, and descending spinal pathways reduce afferent signals conveyed through dorsal horn neuron. (Adapted from Camilleri, M., Saslow, S. B., and Bharucha, A. E.: Gastrointestinal sensation: mechanisms and relation to functional gastrointestinal disorders. Gastroenterol. Clin. North Am. 25:247, 1996.)

Management of Gut Dysmotility

Clinical Assessment Symptoms of oral and pharyngeal dysfunction should be elicited; oral dysfunction manifests as drooling from the mouth, inability to chew or propel the bolus from the mouth, sialorrhea or xerostomia, difficulty initiating swallowing, piecemeal swallowing, and dysarthria. Symptoms of pharyngeal dysfunction also include nasal regurgitation, the need to swallow repeatedly to clear food or fluid from the pharynx, coughing or choking during meals, a “gurgly” voice, and dysphonia. Physical findings may reveal cranial nerve dysfunction, neuromuscular disease, cerebellar dysfunction, or a movement disorder. Muscle fasciculation, ptosis, wasting, and weakness or fatigability should be sought to detect underlying motor neuron disease, myopathy, or myasthenia. Examination of the neck may disclose masses, lymph nodes, or a goiter; axial motion of the larynx and hyoid bone can be assessed while the patient swallows with the examiner’s index and middle fingers resting lightly on the hyoid and laryngeal cartilages, respectively. Laryngeal ascent is frequently reduced in neurogenic dysphagia. The gag reflex does not predict severity of pharyngeal dysfunction or aspiration risk, and is absent in 20% to 40% of healthy adults.46 Unfortunately even a detailed bedside swallowing assessment is not sufficient for assessing aspiration risk in patients with dysphagia; in one study the clinical examination failed to identify 14% of patients with aspiration identified by the criterion standard, fiberoptic endoscopic evaluation of swallowing (FEES).97 After a stroke, aspiration predicted risk of pneumonia.83 Evaluation Nasoendoscopy of the oropharynx is useful to evaluate for malignant, neurogenic, or myogenic etiology of swallowing dysfunction. A videofluoroscopic swallowing study, also termed a modified barium swallow (MBS), captures the dynamic components of swallowing, assessing for aspiration, providing insights into etiology of dysphagia (e.g., Zenker’s diverticulum), quantifying severity of swallowing dysfunction, and suggesting which compensatory swallowing maneuvers will compensate for dysfunction.105 The pharynx and larynx are visualized immediately before and after the swallow by FEES; silent aspiration of pooled secretions can be visualized. FEES is comparable to MBS for identifying aspiration.6 FEES and MBS primarily analyze the motor component of swallowing, but only indirectly assess the sensory component. In the presence of laryngopharyngeal sensory deficits, protective reflexes may not be properly initiated, leading to dysphagia and aspiration. In addition to standard endoscopic evaluation of swallowing, sensory testing assesses laryngopharyngeal sensory discrimination thresholds by endoscopically delivering air pulse stimuli to the mucosa innervated by the superior laryngeal nerve. Laryngopharyngeal sensation is assessed by recording the

2687

threshold pressure of the air pulse required to elicit the laryngeal adductor reflex (LAR), a sensorimotor reflex.7 When the pharyngeal squeeze was preserved, the LAR did not predict penetration or aspiration of pureed consistencies, and was therefore of questionable utility.8 Conversely, when pharyngeal squeeze was impaired, aspiration of pureed consistencies occurred more frequently (in 78% of subjects) compared to when laryngeal sensation was normal (in 18%). Thus sensory testing may help stratify the risk of aspiration in patients with pharyngeal weakness.

Management The rationale and evidence for various therapies have been detailed in a recent review.46 Videofluoroscopic assessment facilitates selection of dietary modifications and postural adjustments, and the decision to institute nonoral feeding. Other factors influencing the latter decision include natural history and prognosis of the underlying illness, patient’s cognitive ability, and estimated likelihood that therapeutic maneuvers will compensate for dysfunction. Thick liquids are less likely to spill over the tongue base, reducing the risk for pneumonia after a stroke. Swallowing therapy, including postural manipulation, strengthening exercises, and thermal stimulation, reduces the degree of aspiration demonstrated by radiography. However, the effect of swallowing therapy on bolus clearance is unclear, and there is no evidence of long-term benefit or improved oral intake. Cricopharyngeal myotomy is effective for alleviating dysphagia attributable to a Zenker’s diverticulum, but there are few data to support efficacy in neuropathic or myopathic dysphagia. Because swallowing may improve considerably in the first 2 weeks after a stroke, long-term decisions should be delayed for that period. Nonoral feeding is intended to nourish patients and reduce the risk of aspiration pneumonia. However, tube feeding cannot be expected to prevent aspiration of oral secretions, and its effect on the risk of aspirating regurgitated gastric contents is unclear.63 Jejunostomy feeding is not associated with a lower rate of pneumonia than feeding through a gastrostomy.64,96 Cuffed tracheostomy tubes prevent aspiration when the cuff is inflated, but they may cause tracheal stenosis because tubes are rarely needed in most neuromuscular disorders associated with dysphagia. Other approaches used to reduce aspiration resulting from acute vocal fold paralysis include thyroplasty and arytenoid adduction35 and endoscopic Teflon injections140 into the paralyzed vocal fold. These procedures reduced aspiration in over 90% of patients. However, they are not easily reversed, and thyroplasty must be performed in the operating room. An alternative approach is vocal cord augmentation with percutaneously injected Gelfoam paste.3

2688

Diseases of the Peripheral Nervous System

Gastroparesis and Chronic Intestinal Pseudo-obstruction Gastroparesis is a clinical syndrome in which patients have upper gastrointestinal symptoms that are at least partly attributable to markedly delayed gastric emptying. However, the relationship between symptoms and delayed gastric emptying is unclear. In fact, many diabetics with delayed gastric emptying are asymptomatic.88 Conversely, gastric emptying may be only slightly delayed in other patients with severe symptoms. Indeed, mild delays in gastric emptying are relatively common, occurring in up to one third of patients with nonulcer dyspepsia.152 Therefore, the diagnosis of gastroparesis should be reserved for patients who have severe symptoms and markedly delayed gastric emptying, recognizing that the boundaries that separate mildly from markedly delayed gastric emptying are not defined. Intestinal pseudo-obstruction is an uncommon syndrome characterized by acute, recurrent, or chronic symptoms suggestive of small or large intestinal obstruction without radiologic, surgical, or endoscopic evidence of mechanical obstruction. The term pseudo- suggests that the symptoms mimic, but are not caused by, mechanical bowel obstruction. Thus gastroparesis and intestinal pseudo-obstruction result from impaired gut motility, involving predominantly the stomach in gastroparesis and the small intestine in intestinal pseudo-obstruction. Patients with intestinal pseudoobstruction frequently have delayed gastric emptying. Etiology Either disorder may be caused by extrinsic neurologic diseases at any level of the brain-gut connection, neuronal dysfunction in the myenteric plexus,45 or a visceral myopathy. Common autonomic neuropathies associated with intestinal pseudo-obstruction include diabetes mellitus,45 amyloidosis,9 paraneoplastic disorders,98 and idiopathic autonomic neuropathy.27,28

Clinical Assessment Symptoms range from postprandial abdominal discomfort to recurrent postprandial emesis resulting in weight loss and malnutrition. Bowel habits may range from constipation to diarrhea. Diarrhea may suggest bacterial overgrowth, or loss of tonic-inhibitory sympathetic input resulting in increased motility.42 Physical examination may reveal a succussion splash, abdominal distention, and tympany to percussion. Diagnostic Evaluation The evaluation is guided by clinical features and comprises the following steps (Fig. 121–5). Excluding Mechanical Obstruction. It is critical to distinguish mechanical gastric outlet or small intestinal obstruction from a motility disorder. Flat and upright radiographs of the abdomen should always be obtained, if possible when patients have severe symptoms (e.g., abdominal pain or distention). Air-fluid levels and dilated small intestinal loops can occur in mechanical and pseudoobstruction; a distinct cutoff between dilated and normalcaliber small bowel loops suggests mechanical obstruction. Upper gastrointestinal endoscopy is useful for excluding mechanical gastric outlet obstruction or an active peptic ulcer and obtaining a small bowel aspirate to diagnose bacterial overgrowth when indicated. An upper gastrointestinal barium radiograph is necessary when mechanical small intestinal obstruction is suspected. Barium retention is a potential hazard in patients with severe dysmotility; patients should receive an osmotic laxative after the test if obstruction is excluded. Confirming Diagnosis of Motility Disorder. A scintigraphic gastric emptying and/or small bowel transit study demonstrates the presence and severity of delayed emptying29,163,173 (Fig. 121–6). A transit study may not be necessary when clinical symptoms and diagnostic tests (e.g., dilated small intestinal loops and/or air-fluid levels on

FIGURE 121–5 Algorithm for the investigation of patients with suspected gastrointestinal dysmotility. (From Camilleri, M.: Study of human gastroduodenojejunal motility: applied physiology in clinical practice. Dig. Dis. Sci. 38:785, 1993, with permission.)

Management of Gut Dysmotility

2689

FIGURE 121–6 Scintigraphic images at baseline (0 hour) and 1, 2, and 4 hours after ingestion of two scrambled eggs labeled with 99mTc-sulfur colloid with one slice of whole-wheat bread and one glass of whole milk (total 300 kcal). Gastric emptying was delayed at 1 (normal range ⫽ 11% to 39%), 2 (40% to 76%), and 4 (84% to 98%) hours after the meal.

plain radiographs) are consistent with intestinal pseudoobstruction. Not all patients with diabetes mellitus and upper gastrointestinal symptoms have delayed gastric emptying; indeed, emptying may be normal or even rapid.129 Intubated studies of pressure profiles by manometric transducers in the distal stomach and small bowel identify motor dysfunction (Fig. 121–7), differentiate neuropathic from myopathic processes, and exclude mechanical obstruction that may have been missed on previous radiographic studies of the small bowel.65 Ascertaining Etiology. Autonomic function tests are abnormal in patients with extrinsic nerve dysfunction but normal in patients with intrinsic nerve dysfunction.15 Brain imaging is indicated if there is no obvious cause for dysmotility, particularly if autonomic tests suggest a central lesion. In smokers, the presence of circulating paraneoplastic antibodies should lead to a chest computed tomography (CT) scan, which will reveal a small cell lung cancer.107 If manometry reveals a myopathy, a thorough family history, serologic tests for a connective tissue disorder, fat or rectal biopsy to rule out amyloidosis, and appropriate tests for a muscular dystrophy or mitochondrial cytopathy should be undertaken. Caveats to Diagnostic Testing. Because narcotics markedly influence motility and impair transit, they must be discontinued for at least 48 hours before assessing

transit. In addition, several technical factors influence gastric emptying studies. These factors include the phase of the meal evaluated (solid vs. homogenized solid vs. caloric liquid vs. noncaloric liquid), the duration for which scans are taken after meal ingestion, and the confidence limits of the normative data provided. It is particularly important to obtain scans for 4 hours after ingestion of meals. A recent multicenter study estimated the normal range for gastric emptying.164,165 Finally, symptoms may be caused not only by delayed gastric emptying but also by other disturbances (e.g., impaired gastric accommodation and visceral hypersensitivity) in patients with an autonomic neuropathy.17 Management The objectives are to maintain adequate hydration and nutrition, restore normal intestinal propulsion, and prevent or treat complications (e.g., bacterial overgrowth) (Table 121–1).26 The specific measures depend on the severity of the illness and may be categorized as follows. Nonpharmacologic Measures. All patients who tolerate oral feeding should be advised to eat several small meals daily, reducing dietary fat and fiber content because lipids delay gastric emptying and digestion of solids (especially nondigestibles) is impaired in patients with gastric antral hypomotility. Patients with more severe delays in gastric emptying may benefit from blenderized foods or liquid formula supplements. Gastric emptying of liquids is

2690

Diseases of the Peripheral Nervous System

FIGURE 121–7 Left, Myopathic pattern of intestinal pseudo-obstruction caused by systemic sclerosis. Note the low amplitude of phasic pressure activity compared to control subject (center). Right, Neuropathic intestinal pseudo-obstruction in diabetes mellitus. Note the absence of antral contractions and persistence of cyclical fasting-type motility in the postprandial period. (From Camilleri, M.: Medical treatment of chronic intestinal pseudo-obstruction. Pract. Gastroenterol. 15:10, 1991, with permission.)

relatively preserved compared to solids even in severe gastroparesis; liquid nutritional supplements help provide sustenance. Consideration should be given to withdrawing medications that can impair motility, such as opiates, phenothiazines, tricyclic antidepressants with anticholinergic effects, and calcium channel blockers. Antiemetic and Procontractile Agents. In the short term, the central and peripheral dopamine2 (D2) receptor Table 121–1. Principles for Managing Gut Dysmotility • Consider, and if necessary exclude, mechanical obstruction. • Consider potentially reversible metabolic causes or medications causing gut dysmotility. • Enteral feeding is preferred to parenteral feeding. • Liquids are emptied faster than solids, even in patients with gastroparesis. • Consider antibiotic prophylaxis for bacterial overgrowth. • Prokinetic and antiemetic agents should be used judiciously. • Surgery is rarely necessary except for placing enteral feeding tubes.

antagonist metoclopramide (Reglan IM; 10 to 15 mg before meals) improves symptoms in gastroparesis; the effect on gastric emptying is unclear.117 Long-term use of metoclopramide may lead to hyperprolactinemia and extrapyramidal side effects, which are generally irreversible. Domperidone (Motilium; 20 mg qid) is a dopaminergic D2 receptor antagonist available in Canada and Europe but not the United States; it reduces symptoms and accelerates gastric emptying for liquids in gastroparesis.151 Extrapyramidal side effects are rare with domperidone because the compound does not cross the blood-brain barrier; the antiemetic effect is mediated through central D2 receptors that are not protected by the blood-brain barrier. There is strong evidence to support the use of intravenous erythromycin (3 mg/kg IV q8h) for 48 to 72 hours in gastroparesis25,87; the drug stimulates motilin receptors, inducing giant antral contractions at this dose, and improving symptoms and accelerating gastric emptying. However, erythromycin may cause abdominal cramps and vomiting, and rarely ventricular arrhythmias in people who have a prolonged corrected Q-T interval, or when

2691

Management of Gut Dysmotility

administered with terfenadine, astemizole, or cisapride. Though long-term oral erythromycin is often prescribed for gastroparesis, its effects wane rapidly as a result of tachyphylaxis, and are barely evident at 1 month. The somatostatin analogue octreotide induces small intestinal contractile activity fronts.78 While octreotide (50 ␮g qhs) improved gastrointestinal symptoms in scleroderma in an uncontrolled study,150 the drug delayed small bowel transit in healthy subjects.174 The cholinesterase inhibitor neostigmine also enhances small bowel contractility, and case reports suggest it may temporarily relieve small bowel distention in acute-on-chronic intestinal pseudoobstruction.16 Gastric Pacing. Gastric pacing refers to electrical stimulation at a frequency that is at or slightly higher than that of the intrinsic gastric slow wave.166 The objective is to create an artificial slow wave (3/min) that entrains and coordinates gastric electrical activity, inducing propagative motor activity and accelerated emptying. In uncontrolled studies, gastric pacing improved symptoms, gastric emptying, and requirement for tube feeding in gastroparesis.116 However, patients were allowed to continue prokinetic medications during the study. Another approach involves high-frequency stimulation at around 12/min. The third approach employs a computercontrolled stimulation protocol delivered in sequential fashion from the proximal to the distal stomach using successive rings of electrodes affixed to the gastric mucosa.121 In contrast to pacing, this option does not depend on this intrinsic gastric myoelectrical rhythm, and delivers a current pulse 1000 times the frequency of the slow wave. Larger controlled studies are necessary to ascertain the benefits of gastric electrical stimulation on symptoms and emptying. It is also conceivable that the effects of electrical stimulation may not be entirely attributable to motor or myoelectric propagation, but may also result from inhibition of transmission in afferent nerve pathways. Suppression of Bacterial Overgrowth. Intestinal dysmotility predisposes to bacterial overgrowth.168 An empirical approach to suppress bacterial overgrowth utilizes rotating brief courses of antibiotics with the intent of avoiding emergence of resistant strains. Options include norfloxacin (400 mg bid), amoxicillin–clavulanic acid (500 mg tid), doxycycline (100 mg bid), and metronidazole (250 mg tid), each given for 7 days every month in rotation. Relief of Distention. A venting gastrostomy/jejunostomy can be placed surgically or by endoscopy and is used to relieve distention, nausea, or vomiting. Care should be taken to avoid electrolyte imbalances (i.e., alkalosis, hypokalemia) from excessive drainage.

Surgical Therapy. There are few indications for surgical therapy in chronic intestinal pseudo-obstruction.124 Patients with severe persistent intestinal distention may require venting; an enterostomy may also permit feeding. Resection of an apparently localized segment of intestinal pseudoobstruction is generally inadvisable, because the apparently nondilated bowel segment may also be involved by disease. Bypass may be effective for megaduodenum.4,112

Constipation Constipation encompasses a variety of symptoms including infrequent bowel movements and/or hard stools and difficult stool evacuation.181 Etiology One or more mechanisms (i.e., diminished rectal sensation, impaired colonic propulsion, or pelvic floor dysfunction) may cause constipation in patients with an autonomic neuropathy. Patients with reduced rectal sensation, as in diabetes mellitus and MS, may not experience the urge to defecate.36 The pelvic floor may not relax normally during defecation in MS and Parkinson’s disease39,115; anticholinergic agents may also delay transit in Parkinson’s disease. In addition to impaired colonic transit and/or pelvic floor function, paraplegics may lack the ability to increase intraabdominal pressure during expulsion. Clinical Assessment A digital rectal examination may identify impacted stool or pelvic floor dysfunction. Normally the puborectalis muscle relaxes and the perineum descends by 1 to 2 cm as patients try to expel the examiner’s finger. In pelvic floor dysfunction, the puborectalis muscle contracts instead of relaxing; perineal descent may be inadequate or excessive (⬎4 cm). Diagnostic Evaluation An obstructing colonic lesion must be considered, and if necessary excluded by barium radiograph or colonoscopy, particularly if patients are older than 50 years, have anemia, or have blood in the stool. Colonic transit can be assessed by radiopaque markers120 or scintigraphy.173 The radiopaque marker technique is simple, sensitive, and widely available. There are several variations of the technique. In a widely used method, patients ingest 24 radiopaque markers on each of 3 consecutive days; the number of markers in the abdomen is counted on abdominal plain radiographs taken on days 4 and 7. Bony landmarks are used to apportion distribution of markers within a specific colonic segment (i.e., right colon, left colon, and rectosigmoid colon). The total number of markers in each segment on days 4 and 7 provides the estimated mean colonic transit time for that segment. The upper limit of normal total colonic transit time is approximately 75 hours.

2692

Diseases of the Peripheral Nervous System

Scintigraphic assessment of colonic transit utilizes 111 In-labeled charcoal pellets contained in a capsule with a pH-sensitive methacrylate coating.173 The pH-sensitive coating is designed to disintegrate in the alkaline pH of the terminal ileum, releasing the isotope in the cecum. Scans taken over the next 24 to 48 hours monitor the progression of the isotope as it travels around the colon; the average colonic distribution of the isotope is expressed as its geometric center.173 Pelvic floor function is assessed by anorectal manometry, balloon expulsion test, and pelvic floor imaging during defecation by scintigraphic or barium defecography if clinically indicated. These tests are discussed in detail in Chapter 13. Delayed colonic transit does not necessarily indicate colonic motor dysfunction because the delay may be secondary to pelvic floor dysfunction.

Diagnostic Evaluation The diagnosis is established by increased colonic diameter on plain radiographs of the abdomen. In acute megacolon, cecal diameter is greater than 12 cm. In chronic megacolon, the sigmoid colonic diameter at the pelvic brim is greater than 6.5 cm. It is critical to exclude distal colonic obstructive lesion by a contrast (Hypaque) enema or flexible endoscopy before administering neostigmine (see below).

Management Fiber from foods and supplements should be taken twice daily with fluids/meals, and increased gradually over several weeks to the target of 12 to 20 g fiber/day (Table 121–2).104 Though fiber supplementation may cause bloating, this symptom often subsides over time; alternatively another fiber supplement (e.g., methylcellulose [Citrucel]) should be tried. The next step is to add an inexpensive osmotic agent such as milk of magnesia titrated to produce soft, but not liquid stools. If necessary, stimulant laxatives such as bisacodyl (Dulcolax) or polyethylene glycol should be considered. These agents have been reviewed in detail in recent publications.104 Other “colonic prokinetics” are in development; the use of prostaglandins (misoprostol) and colchicine to treat constipation is not recommended. Disordered evacuation responds poorly to laxatives, but often (~75%) responds to biofeedback therapy.57

Management While mechanical obstruction is being excluded, potentially enabling medications should be discontinued and metabolic disturbances corrected. The rectum should be decompressed with an indwelling tube and tap water enemas. Intravenous neostigmine (1 mg) initially is generally effective and safe for patients with colonic distention unresponsive to such conservative therapies; though the recommended dose is 2 mg,135 even neostigmine 0.75 mg IV can increase colonic motor activity in healthy subjects.95 Contraindications to neostigmine include active bronchospasm, heart rate less than 60 beats/min or systolic blood pressure less than 90 mm Hg, and sick sinus syndrome or history of second- or third-degree atrioventricular block without a pacemaker. Because neostigmine may cause bradycardia, patients should be monitored closely, though not necessarily in an intensive care unit setting, after administration. Atropine (0.6 to 1.2 mg IV) is an effective antidote for the undesirable muscarinic side effects of neostigmine. Endoscopic decompression is necessary for patients who do not respond to or relapse after neostigmine, or in whom neostigmine is contraindicated. Signs of peritonitis may imply colonic perforation, and surgery will be needed, often on an emergent basis.

Megacolon

Diarrhea

Colonic dilatation or megacolon may be acute as in Ogilvie’s syndrome, chronic, or toxic.14 Acute megacolon is attributed to a sympathetically mediated reflex response to a number of serious medical or surgical conditions in elderly patients, often after orthopedic fractures or surgeries. Chronic megacolon may be congenital (i.e., Hirschsprung’s disease) or acquired, reflecting endstage colonic motor dysfunction. Toxic megacolon is secondary to colonic inflammation, as in ulcerative or infectious colitis.

Loss of the sympathetic “brake” in autonomic neuropathy, attributable to diabetes168 or of unclear etiology, may be associated with diarrhea.42

Clinical Assessment Patients with acute megacolon are often relatively asymptomatic, despite marked abdominal distention and percussion tympany.

Clinical Assessment It is important to specifically ask patients whether they are incontinent because up to 50% of patients with fecal incontinence may not volunteer the symptom.101 It is important to inquire about ingestion of substances (e.g., fructose or sorbitol) that are used to sweeten beverages in amounts sufficient to cause diarrhea even in subjects who do not suffer from malabsorption. For example, each can of a regular soda contains approximately 40 g of fructose, which is higher than the dose of 25 g used to assess for fructose malabsorption in the laboratory!

2693

Management of Gut Dysmotility

Table 121–2. Management of Constipation Type Generic and Trade Name (Dosage) Fiber Bran (1 cup/day) Psyllium (Metamucil, 1 tsp. up to tid) Methylcellulose (Citrucel, 1 tsp. up to tid)

Mechanism of Action qStool bulk, pcolonic transit time, q GI motility

Polyethylene glycol (Miralax, 17 g/day) Suppository Glycerin (up to daily) Bisacodyl (Dulcolax, 10 mg daily) Oral Stimulant Bisacodyl (Dulcolax, 10 mg PO up to 3 times/wk) Anthraquinones (senna, cascara) Senokot (2 tab qd to 4 tab bid) Perdiem (plain) (1–2 tsp. qd) Peri-Colace (1–2 tab qd) Saline Laxative Magnesia

Bloating, flatulence, iron and calcium malabsorption; increase dose gradually; methylcellulose may cause less bloating

12–72

Ineffective for constipation

Nonabsorbable disaccharides q osmolality and are metabolized by colonic bacteria into short-chain fatty acids, which accelerate colonic transit Osmotically q intraluminal fluids

24–48

Sweet tasting (lactulose), transient abdominal cramps, flatulence

0.5–1

Incontinence caused by potency

Evacuation induced by local rectal stimulation

0.25–1 0.25–1

Rectal irritation Irritation

Similar to senna (see anthraquinones) Electrolyte transport altered by q intraluminal fluids; myenteric plexus stimulated; motility q

Fluid osmotically drawn into small bowel lumen; CCK stimulated; colon transit time p

Milk of magnesia (15–30 mL qd or bid) Lubricant Mineral oil (15–45 mL)

Enema (All per Rectum) Mineral oil retention (100–250 mL qd) Tap water (500 mL) Phosphate (Fleet’s 1 unit) Soapsuds (1500 mL)

Side Effects

Weeks

Stool Softener Docusate sodium (Colace, 100 mg bid) Hyperosmolar Agents Sorbitol (15–30 mL qd or bid) Lactulose (Chronulac, 15–30 mL qd or bid)

Time to Onset of Action (h)

6–8 8–12

1–3

1–3

Stool lubricant

Stool softened and lubricated (mineral oil) Evacuation induced by distended colon; mechanical lavage (others)

6–8

6–8 (mineral oil)

Incontinence, hypokalemia, abdominal cramps Degeneration of Meissner’s and Auerbach’s plexuses (unproven), malabsorption, abdominal cramps, dehydration, melanosis coli Magnesium toxicity, dehydration, abdominal cramps, incontinence Avoid in renal failure

Lipid pneumonia, fat-soluble vitamin malabsorption, dehydration, incontinence Incontinence, mechanical trauma

5–15 min (others) Hyperphosphatemia

q ⫽ increase; p ⫽ delay; CCK ⫽ cholecystokinin. From Bharucha, A. E., and Camilleri, M.: GI dysmotility and sphincter dysfunction. In Noseworthy, J. H. (ed.): Neurological Therapeutics: Principles and Practice. London, Martin Dunitz, 2003, with permission.

2694

Diseases of the Peripheral Nervous System

Diagnostic Tests The differential diagnosis of causes of diarrhea unrelated to neurologic disorders is beyond the scope of this chapter. The presence and severity of malabsorption can be ascertained by quantifying fecal fat and weight in a 48-hour stool collection. Though the upper limit of normal for stool fat content is approximately 7%, values ranging between 7% and 13% are often encountered in patients with rapidtransit diarrhea, and do not necessarily indicate a mucosal or other disease process causing malabsorption.127 A small bowel radiograph and/or mucosal biopsy may be necessary in patients with significant malabsorption; the biopsy may reveal amyloid deposition. In diabetics, the biopsy may demonstrate changes consistent with associated disorders (i.e., bacterial overgrowth or celiac sprue). Specific search for small bowel diverticula or dilated loops may require x-ray films taken upright to demonstrate air-fluid (barium) levels. In the absence of malabsorption, it is useful to ascertain whether diarrhea is attributable to a secretory or osmotic process. Osmotic diarrhea is caused by ingestion of carbohydrates (e.g., fructose, magnesium, or other laxatives), and subsides with fasting. Secretory diarrheas subside with fasting; common causes include infections, stimulant laxatives, inflammatory bowel disease, celiac sprue, and bile acid malabsorption. Management Malabsorption from bacterial overgrowth is treated with rotating antibiotics; with absence of malabsorption, medications that retard small bowel and colonic transit provide symptomatic relief. For example, by restoring the sympathetic brake, clonidine may reduce diarrhea in diabetic autonomic neuropathy.61 However, many patients do not tolerate the medication because of somnolence and hypotension. More often, treatment involves oral loperamide (2 to 16 mg/day) or subcutaneous preprandial octreotide (25 to 100 mg tid).

Fecal Incontinence Fecal incontinence refers to involuntary leakage of stool from the anus. Common neurologic disorders associated with fecal incontinence include MS, Parkinson’s disease, multiple system atrophy, Alzheimer’s disease, strokes, diabetic neuropathy, and spinal cord lesions.13 In the absence of neurologic disease, “idiopathic” fecal incontinence is frequently due to anal sphincter weakness resulting from sphincter defects caused by obstetric trauma, perhaps aggravated by stretch-induced pudendal nerve injury; the latter has been attributed to prolonged straining in constipated patients, or obstetric trauma. Fecal incontinence may have devastating consequences on lifestyle, and even contribute to institutionalization.

Clinical Assessment Important aspects include frequency and type of incontinence episodes, bowel habits, and impact of incontinence on quality of life. External sphincter weakness often manifests as urge incontinence or stress incontinence during coughing, sneezing, or laughing. Internal anal sphincter weakness, as occurs in diabetic autonomic neuropathy or systemic sclerosis, often manifests as nocturnal incontinence or “passive” incontinence (i.e., without prior warning).13,175 Diminished rectal sensation may impair ability to distinguish between flatus and stool in the rectum. Inspection of the anus with and without straining in the lateral decubitus and seated position will detect rectal prolapse; a digital rectal exam can assess resting and squeeze tone in the anal canal, and exclude fecal impaction. Diagnostic Evaluation54 Colonoscopy, or proctoscopy and barium enema, are necessary to exclude mucosal lesions. Tests to assess anorectal function—anal manometry to measure anal canal pressures, rectal sensation of balloon distention, endoanal ultrasound for anal sphincter morphology, and defecating proctography for perineal descent—are discussed in Chapter 13. The identification of pudendal neuropathy by pudendal nerve latencies is suboptimal because the technique lacks sensitivity and specificity for identifying pudendal nerve damage. Needle electromyography (EMG) of the external sphincter provides a sensitive measure of denervation (fibrillation potentials) and can usually identify myopathic (small polyphasic motor unit potentials), neurogenic (large polyphasic motor unit potentials), or mixed injury. Management Perianal skin care and incontinence pads are essential. Regularization of bowel habits is critical to managing incontinence, particularly for patients with diarrhea. Fiber supplements and laxatives, used judiciously, may lessen constipation. Rectal irrigation with an enema after a bowel movement may avoid stool incontinence. Antidiarrheal agents such as loperamide (Imodium) are extremely useful for diarrhea; the dose is 2 to 4 mg, given 30 minutes before meals, up to a maximum of 16 mg/day. The dose of loperamide must be titrated to reduce diarrhea without causing constipation. Pelvic floor retraining by biofeedback therapy is useful for enhancing anal sphincter tone and improving rectal sensation. In patients who have severe weakness or lack rectal sensation, there may be little that can be achieved with medical or biofeedback therapy. In these situations, colostomy may provide a more manageable solution to the chronic incontinence. The role of other surgical approaches (anal sphincter and/or pelvic floor repair, creation of a new anal sphincter without or with electrical stimulation) for fecal incontinence is unclear. There are no randomized trials comparing surgical intervention to nonsurgical approaches for fecal incontinence.

2695

Management of Gut Dysmotility

ASSESSMENT OF AUTONOMIC FUNCTION IN THE GASTROINTESTINAL TRACT While general autonomic reflex tests106 are useful to assess the function of the autonomic control of the viscera, three tests are specific for gut autonomic innervation.

Pancreatic Polypeptide Response to Hypoglycemia or to Modified Sham Feeding by the “Chew and Spit” Technique Plasma pancreatic polypeptide concentrations increase by at least 25 pg/mL during sham feeding, reflecting a vagal response. The modified sham feeding test seems to be a more sensitive means of detecting vagal dysfunction than the postprandial response of plasma pancreatic polypeptide.27,75 The coexistence of antral hypomotility with abnormal pancreatic polypeptide responses to sham feeding further supports the presence of vagal dysfunction or impaired gastric emptying of solids in these situations.

R-R Interval Response to Deep Breathing This is a test of cardiovagal reflexes.58,59 However, Buysschaert et al.21 showed that this assessment is a useful surrogate for testing abdominal vagal function, consistent with the concept that vagal denervation, as with most forms of diabetic neuropathy, commences caudally and progresses in a cranial direction.

Mesenteric Flow in Response to Tilt-Table Testing37,69 Splanchnic blood flow is under baroreflex control, and appropriate regulation is important in the maintenance of postural normotension. Evaluation of mesenteric flow in response to eating and head-up tilt provides important information on intra-abdominal sympathetic adrenergic function, and the ability of the patient to cope with orthostatic stress. Superior mesenteric artery flow in response to perturbations such as tilting and meal ingestion assesses sympathetic adrenergic function in the abdomen. While useful, this technique requires considerable expertise and is not widely available.

EXTRINSIC NEUROLOGIC DISORDERS CAUSING GUT DYSMOTILITY This review concentrates on neurologic diseases of extrinsic neural control and smooth muscle that affect gut motility.

Central Neurologic Disorders Autonomic Epilepsy and Migraine Both of these disorders are infrequent causes of nausea, vomiting, and/or abdominal pain. Postpolio Dysphagia Individuals with postpolio syndrome with bulbar involvement frequently have dysphagia and aspiration. Attention to the position of the patient’s head during swallowing and changing food consistency to semisolid food can decrease the incidence of choking and aspiration. Brainstem Tumors Brainstem lesions can present with isolated gastrointestinal motor dysfunction. In the absence of increased intracranial pressure, symptoms likely result from direct effects of the mass in the brainstem, with distortion of the vomiting center on the floor of the fourth ventricle. Motor dysfunction is typically evident on manometric or radionuclide studies of the stomach and small bowel.183 Although most reports are associated with vomiting, colonic and anorectal dysfunction have also been described.179 The presence of autonomic neuropathy in a patient with a motility disorder should prompt a search for a structural lesion in the central nervous system, particularly if the tests suggest a preganglionic sympathetic lesion. Magnetic resonance imaging is considered preferable to CT for detecting brainstem lesions.183 Parkinson’s Disease Gastrointestinal symptoms occurring more commonly in Parkinson’s disease than in unaffected individuals include abnormal salivation, dysphagia, nausea, constipation, and defecatory dysfunction.56 The prevalence of these symptoms correlates with the duration and severity of Parkinson’s disease, not with diet or treatment. Dysphagia is frequently associated with choking and disordered salivation, and may be attributable to direct involvement of oropharyngeal and upper esophageal sphincter muscles by the primary disease process. Constipation often manifests not by decreased stool frequency, but by disturbed stool consistency and excessive straining, suggestive of defecatory dysfunction resulting from paradoxical puborectalis contraction during defecation.115 Degeneration of dopaminergic neurons in the myenteric plexus is considered important in the pathogenesis of constipation related to Parkinson’s disease.146

Autonomic System Degenerations Pandysautonomia/Selective Dysautonomias Pandysautonomia or selective dysautonomias are characterized by preganglionic or postganglionic lesions affecting both sympathetic and parasympathetic nerves. Vomiting, paralytic ileus, constipation, and a chronic pseudo-obstruction

2696

Diseases of the Peripheral Nervous System

syndrome have been reported in acute, subacute, or familial dysautonomias. Motor dysfunctions have been documented in the esophagus, the stomach, and the small bowel. Selective cholinergic dysautonomia affects parasympathetic nerves and sympathetic nerves to sweat glands, and may also impair upper and lower gastrointestinal motor activity. This picture may follow a viral infection, such as infectious mononucleosis.170 Idiopathic Orthostatic Hypotension Idiopathic orthostatic hypotension is sometimes associated with motor dysfunction of the gut, such as esophageal dysmotility, gastric stasis, alteration in bowel movements, and fecal incontinence.32 Cardiovascular abnormalities usually precede gut involvement. The precise site of the lesion causing the gut dysmotility is unknown. Multiple System Atrophy Multiple system atrophy is associated with esophageal dysmotility, pseudo-obstruction, constipation, and fecal incontinence.145 In contrast to amyotrophic lateral sclerosis, the anterior horn cells supplying the urethral and anal sphincters, termed Onuf’s nucleus, are often affected early in the course of multiple system atrophy, causing urinary and bowel dysfunction. In one study, anal sphincter EMG was abnormal in 82% of patients with multiple system atrophy.132

Peripheral Neuropathy Acute Peripheral Neuropathy Autonomic dysfunction is associated with nausea, vomiting, abdominal cramps, constipation, Guillain-Barré syndrome, herpes zoster, Epstein-Barr virus, botulism from Clostridium botulinum type B, or cytomegalovirus.20 Human immunodeficiency virus (HIV)–induced diarrhea may be another manifestation of autonomic dysfunction (see below). Diabetes Mellitus Epidemiology. At a tertiary referral center, up to 76% of diabetics had at least one gastrointestinal symptom.62 Twenty percent of diabetics registered at a general clinical research center had gastroparesis.43 In community-based studies, the prevalence of gastrointestinal symptoms is much lower. Only 1% of patients had gastroparesis and 0.6% had nocturnal diarrhea in the Rochester Diabetic Neuropathy Study,55 while constipation and laxative use were the only symptoms more prevalent in type 1 and not type 2 diabetics, compared to age- and gender-matched healthy controls, in another detailed questionnaire-based study in Olmsted County, Minnesota.111 In contrast, a questionnaire-based community survey from Australia reported more diarrhea, fecal incontinence, dysphagia, and postprandial fullness among diabetics; 95% of this cohort

had type 2 diabetes.23 The relatively high prevalence of functional gastrointestinal disorders (e.g., irritable bowel syndrome, constipation, and dyspepsia) in the general population confounds estimation of the prevalence of diabetic neuropathy based on symptoms alone. Diabetic enteropathy refers to all gastrointestinal complications of diabetes, including dysphagia, heartburn, nausea and vomiting, abdominal pain, constipation, diarrhea, and fecal incontinence (Fig. 121–8). Asymptomatic gastric retention was termed gastroparesis diabeticorum by Kassander in 1985.88 Since then, the term gastroparesis has been used in reference to upper gastrointestinal symptoms associated with delayed gastric emptying. More recently the term diabetic dyspepsia has been coined to convey that not all upper gastrointestinal symptoms in diabetics are attributable to delayed gastric emptying.119 Symptoms of a diabetic enteropathy typically occur in patients with type 1 or insulin-dependent diabetes mellitus who have a peripheral neuropathy, often in association with other manifestations of autonomic neuropathy (e.g., orthostatic hypotension). Factors that interact to produce gastrointestinal dysfunction include extrinsic nerve dysfunction, intrinsic or enteric nerve dysfunction, and hyperglycemia. In the gut, these factors have been studied most thoroughly for gastric sensorimotor dysfunction. Diabetic Extrinsic Nerve Dysfunction. Impaired antral motor function, attributed at least in part to vagal dysfunction, may impair trituration and retard gastric emptying of solids; pylorospasm may also contribute to delayed gastric emptying.31,118 However, the correlation between symptom severity and delayed gastric emptying is relatively weak,88 implicating a role for other mechanisms including impaired gastric accommodation and visceral hypersensitivity. The stomach normally accommodates or relaxes after meals, increasing volume by at least two- to threefold, thereby providing room and time for antral trituration of solids into smaller particles before pyloric emptying. Impaired gastric accommodation has been associated with early satiety in functional gastrointestinal disorders.91,159 Until recently, the gastric accommodation reflex was considered to be exclusively mediated by the vagus nerve. However, a recent study questioned that concept, observing normal postprandial accommodation in diabetics with a cardiovagal autonomic neuropathy.52 Conceivably, enteric neurons adapt to preserve gastric accommodation after vagal denervation, as observed in rats 30 days after vagotomy.160 Impaired accommodation may also explain why gastric emptying of liquids is accelerated in some diabetics.66 Hypersensitivity refers to exaggerated gastrointestinal perception of luminal distention; there is no evidence for gastrointestinal hypersensitivity in diabetic neuropathy. The various factors contributing to diarrhea and fecal incontinence are depicted in Figure 121–9. Sympathetic

Management of Gut Dysmotility

2697

FIGURE 121–8 Gastrointestinal manifestations of diabetes may be secondary to complications of the disease per se or other conditions associated with diabetes.

denervation is associated with abnormal motor patterns,30,142 and may also reduce net intestinal absorption of fluids and electrolytes. Bacterial overgrowth and osmotic loads of artificial sweeteners frequently aggravate diarrhea; gluten-sensitive enteropathy is uncommonly associated with type 1 diabetes, and pancreatic exocrine insufficiency is rarely associated with type 1 diabetes. Glycemic control may influence gastrointestinal function directly or by altering counter-regulatory hormones that also influence these functions. These hormones include glucagon, glucagon-like peptide-1 (GLP-1), amylin, epinephrine, somatostatin, growth hormone, and cortisol. Hyperglycemia can modulate gastrointestinal sensorimotor function, but this magnitude of effect is unclear. Epidemiologic studies have reported conflicting results on the association between gastrointestinal symptoms and measured or self-reported glycemic control.22,23 Clamped hyperglycemia retarded gastric emptying for liquids and solids in healthy subjects and diabetics.67 From a clinical perspective, the hyperglycemia-induced gastric emptying delay in these studies was relatively small (e.g., a difference in median half-time of ~15 and 5 minutes for solids and liquids, respectively). Moreover, gastric emptying of solids and accommodation were normal in non–insulindependent diabetics without autonomic neuropathy, despite increased blood glucose and hemoglobin A1c levels.66 Perhaps the most convincing evidence for hyperglycemic

effects is the gastric liquid emptying delay in nonobese diabetic (NOD) and streptozotocin diabetic mice with hyperglycemia. Gastric emptying normalized 12 hours after insulin treatment.178 Hyperglycemic clamping did not affect fasting colonic motor function, colonic perception of distention, or contractile responses to a meal.110 Humoral Mediators. GLP-1 is released from the lower small intestine and colon in response to nutrients. Pharmacologic concentrations of GLP-1 markedly delay gastric emptying, increase insulin, and suppress glucagon secretion in response to eating.51 Thus GLP-1 may be useful to prevent postprandial hyperglycemia, and may contribute to the ileal brake, that is, inhibition of upper gastrointestinal motility by unabsorbed nutrients in the small intestine. Amylin is co-secreted with insulin by pancreatic beta cells in response to nutrients; type 1 diabetes is characterized by insulin and amylin deficiency. Pharmacologic infusions of amylin or the more stable analogue pramlintide inhibit gastric emptying143,171 and reduce glucagon secretion in animals and humans. Pramlintide may delay gastric emptying by centrally inhibiting vagal function, as evidenced by pramlintide-induced inhibition of plasma pancreatic polypeptide after a meal. Enteric Nerve Dysfunction. Nitric oxide is a critical inhibitory neurotransmitter throughout the gastrointestinal

2698

Diseases of the Peripheral Nervous System

FIGURE 121–9 Algorithm for managing diarrhea in patients with diabetes mellitus. IAS ⫽ internal anal sphincter; EAS ⫽ external anal sphincter. (From Camilleri, M., and Bharucha, A. E.: Overview of gastrointestinal dysfunction in diseases of central and peripheral nervous system. In Quigley, E. M. M., and Pfeiffer, R. F. [eds.]: Neurogastroenterology: Gastrointestinal Dysfunction in Neurological Diseases. Oxford, UK, Butterworth-Heinemann [in press], with permission.)

tract. Expression of neuronal nitric oxide synthase (nNOS) protein and messenger RNA is reduced in pyloric myenteric neurons of diabetic mice with delayed gastric emptying; insulin restored pyloric nNOS protein and restored gastric emptying.178 Sildenafil, a cyclic GMP phosphodiesterase inhibitor, also normalized gastric emptying in diabetic mice, suggesting NOS downregulation is involved in the pathogenesis of diabetic gastropathy.178 Reduction in the number of interstitial cells of Cajal, as observed in the distal stomach of spontaneously diabetic mice130 and a full-thickness jejunal biopsy from a patient with type 1 diabetes,79 may also contribute to dysmotility in diabetes.

Psychological Factors. Symptoms in diabetics are significantly influenced by psychological factors.43,109

Management. Because management of gastrointestinal symptoms has been detailed above, only additional options specific to diabetes are discussed here. The ␣-glucosidase inhibitor acarbose accelerated ascending and transverse colonic transit in diabetics with delayed colonic transit.139 Recombinant neurotrophic factors did not improve nerve conduction in diabetic polyneuropathy, but patients experienced diarrhea. Subsequent studies showed that these factors accelerate colonic transit.48

Paraneoplastic Neuropathy Gastrointestinal symptoms (from achalasia to pseudoobstruction) may be paraneoplastic manifestations. A circulating immunoglobulin G antibody (called ANNA-1 or anti-Hu) directed against enteric neuronal nuclei102 has been detected in these patients, suggesting that the enteric plexus is the major target of this immune paraneoplastic process. More than 95% of patients with ANNA-1 have small cell carcinoma of the lung, though this paraneoplastic syndrome may also be associated with kidney cancer or pulmonary carcinoid.41 Several patients also have evidence of extrinsic visceral neuropathies,41,148 suggesting a more extensive neuropathologic process. The chest radiograph is frequently negative, and a chest CT scan is indicated when the syndrome is suspected, typically in middle-aged smokers with recent onset of nausea, vomiting, or feeding intolerance. ANNA-1 is associated with paraneoplastic neuropathies, 14% of which are restricted to GI manifestations.107 Amyloid Neuropathy Familial or acquired amyloid neuropathy may lead to constipation, diarrhea, or steatorrhea. Intestinal myoelectric disturbances in amyloidosis32 mimic those of ganglionectomy.113 Chronic Sensory and Autonomic Neuropathy of Unknown Cause This is a rare, nonfamilial form of slowly progressive neuropathy affecting a number of autonomic functions.

Management of Gut Dysmotility

Patients may have only a chronic autonomic disturbance (e.g., abnormal sweating or blood pressure control or gastrointestinal dysfunction) for many years before peripheral sensory symptoms become apparent. A recent review of the Mayo Clinic experience of 27 patients with idiopathic autonomic neuropathy identified gastrointestinal involvement in 19 patients.157 Porphyria Acute intermittent porphyria and hereditary coproporphyria may present with abdominal pain, nausea, vomiting, and constipation; dilatation and impaired motor function may be seen in any part of the intestinal tract. These manifestations may result from extrinsic autonomic dysfunction or intrinsic neuromyopathy. HIV Infection Neurologic disease may manifest at any phase of the infection with HIV-1 virus. Autonomic dysfunction may be associated with chronic diarrhea, possibly as a result of increased extrinsic parasympathetic activity to the gut44 or damage to the enteric plexuses.76 Spinal Cord Injury Ileus is a frequent but transient complication after spinal cord injury. Colonic and anorectal dysfunction, manifested as chronic constipation and/or fecal incontinence, affects up to 80% of patients with spinal cord injury93 and significantly impairs quality of life. These symptoms result from interruption of supraspinal control of the sacral parasympathetic supply to the colon, pelvic floor, and anal sphincters.156,158 Manifestations of upper motor neuron involvement include preserved resting anal sphincter tone, loss of rectoanal sensation, and the loss of the ability to relax or squeeze pelvic floor muscles during defecation or continence; these disturbances cause constipation. The lower motor neuron pathways to the colon and/or anal sphincters are affected in lesions at vertebral level T10 or below, causing external sphincter denervation and reduced anal sphincter pressures, predisposing to fecal incontinence. Colonic transit is frequently delayed, and intraluminal assessments disclose a reduced gastrocolonic response to eating.73 Moreover, patients with spinal cord injury may be unable to contract abdominal muscles during defecation. Disturbances of upper gastrointestinal motility are relatively uncommon after spinal cord injury. Gastroesophageal reflux disease after spinal cord injury may be attributable to ineffective esophageal acid clearance mechanisms89 and may present with high-grade esophagitis,147 perhaps because patients with lesions above T7 have reduced esophageal sensation and lack typical symptoms of gastroesophageal reflux (e.g., heartburn, and regurgitation). The prevalence of gallstones is higher than that in the general

2699

population, ranging from 17% to 31%5,90,122; patients with lesions above T10 also have a higher prevalence of biliary sludge.161 The etiology of gallstones after spinal cord injury is unclear.161 Management. Comprehensive bowel care programs incorporate modifications of diet and activity, oral and rectal medications, and scheduled bowel care intended to effectively evacuate the colon, eliminate incontinence, and prevent secondary complications of neurogenic bowel dysfunction. Consensus guidelines on bowel programs were developed by the Consortium for Spinal Cord Medicine153 and were published as a consumer guide.154 Fiber supplementation and fluid intake sufficient to produce urinary output of 2 to 3 L daily are generally recommended. However, a bran supplement (40 g daily) for 3 weeks did not affect evacuation frequency, time required for bowel care, or stool weight after spinal cord injury; colonic transit was either unchanged or prolonged.155 Thus the effects of fiber supplementation in spinal cord injury needs further study. Bowel care should be scheduled at least 3 times every week. For upper motor neuron spinal cord injury, the patient or an attendant can initiate defecation by stimulation; typical stimuli in order of intensity include abdominal massage, digital perianal stimulation, rectal stimulation, glycerine suppository, and bisacodyl suppository. Repeated digital rectal stimulation after onset of fecal evacuation activates the rectoanal inhibitory reflex, thereby relaxing the internal anal sphincter, and induces colonic contractions. Digital rectal stimulation is continued for 30 seconds to 1 minute on each occasion and repeated every 3 to 10 minutes to sustain defecation. Among other methods used for bowel care, large-volume tap water enemas are used infrequently,153 except for patients who have incontinence or spina bifida.18 Pulsed irrigation enhanced evacuation devices have been used to clear impaction.137 Electrical stimulation of S2-S4 sacral nerve roots by an implanted device is approved by the Food and Drug Administration for stimulating defecation.19,40,68,169 A selective S2-S3 rhizotomy is performed to reduce anal sphincter tone. However, the device is currently not commercially available in the United States. Preliminary results suggest that magnetic stimulation of the rectum or abdominal wall is an effective option103; an abdominal belt with implanted electrodes has also been used to stimulate the abdominal muscles and expel stool by increasing intra-abdominal pressure.60 Patients with spinal cord injury at T12 or below often have reduced anal sphincter tone and lack spinal cord–mediated reflexes that facilitate defecation. Consequently, bowel care is more challenging compared to when the injury is above T12. The rectum must be cleared of stool more frequently, perhaps once daily, to prevent fecal incontinence. Bowel care is often timed to coincide with the colonic response after meals. Rectal manipulation involves digital stimulation alternating with

2700

Diseases of the Peripheral Nervous System

manual evacuation. Between bowel care sessions, the Valsalva maneuver is discouraged to avoid increased intraabdominal pressure and inadvertent expulsion of stool. Lastly, measures to avoid falls while transferring to a commode125 and pressure ulcers are imperative.126,153 The patient’s position on the commode chair should be adjusted every 15 minutes during a typical bowel care session, which can last 1 hour or longer. Colostomy is a reasonable option for patients who are unsatisfied with a comprehensive bowel care program because the program is cumbersome and/or bowel evacuation is incomplete. Quality of life is often improved after a colostomy138,141; however, pressure ulcers are generally multifactorial and may not heal.53 Potential complications include diversion colitis,94 bowel obstruction, and stomal complications (e.g., ischemia, retraction, prolapse). The Malone procedure involves enema delivery via a surgically created continent catheterizable appendicocecostomy. The procedure has been extensively used in children with spina bifida; 95% of children continued to use the stoma 6 years after it was created.70 Less than 80% of adults continued to use the technique after the stoma was fashioned, perhaps because of concurrent colonic motor dysfunction and/or impaired anal sphincter relaxation, precluding evacuation of feces. Multiple Sclerosis Dysphagia in MS is related to disease severity and brainstem involvement and probably caused by disordered neuromotor sequencing of laryngeal events and pharyngeal constrictor weakness.1,24,50 Constipation and/or fecal incontinence are reported in direct questioning by 39% to 73% of patients with MS38,82,128; both symptoms frequently coexist.81 After spasticity and incoordination, bowel and bladder dysfunction were the most important comorbidities that impaired ability to work.84 Bowel disturbances are probably attributable to disordered extrinsic nervous control of gut function; altered toileting habits secondary to emotional factors and behavioral changes, medications (e.g., anticholinergics for bladder dysfunction, antispastic agents, antidepressants), and systemic factors may also contribute. Potential systemic factors include muscle weakness, spasticity, fatigue, and restricted mobility. Constipated MS patients have delayed colonic transit and/or pelvic floor dysfunction; prolonged colonic transit may either reflect colonic motor dysfunction72,77 or be secondary to pelvic floor dysfunction.39,71,162,180 Extrinsic neural pathways regulating gastrointestinal sensorimotor function may be affected at one or more levels; mechanisms of impaired gut transit are incompletely understood. Somatosensory evoked potentials recorded from the brain are often delayed in MS patients with bowel symptoms. However, potentials recorded at the lumbar spine were normal in most patients, suggesting predominantly higher

spinal or cortical involvement. Motor pathways were also affected, as evidenced by conduction times that are delayed from cortex to lumbar spine, or cortex to pelvic floor striated muscle.114 Without appropriate controls, it is unclear whether these disturbances are causally related to bowel dysfunction. Anal resting pressures are generally normal, but squeeze pressures, reflecting voluntary contraction of the external sphincter, are often reduced in incontinent patients. Paradoxical contraction of the external anal sphincter during defecation may cause constipation. Reduced anorectal sensation, reported in some but not all studies, may contribute to constipation or fecal incontinence.36,39,86,149,176 Postural changes and/or modification of meal quantity and consistency restored swallowing function in greater than 95% of patients with mild or moderate dysphagia secondary to MS; these strategies were ineffective in three patients with severe dysphagia.24 Pelvic floor incoordination responded to biofeedback therapy in 5 of 13 MS patients with constipation with or without fecal incontinence.182 A response was more likely in patients with limited disability and nonprogressive disease course. Myasthenia Gravis Dysphagia occurs in 30% to 60% of myasthenics, is noted at diagnosis in approximately 20% of cases, and may not be associated with typical ocular signs. Therefore, edrophonium (Tensilon) testing and EMG should be considered in the appropriate clinical setting. Mitochondrial Cytopathies Organ disturbances in these syndromes reflect the importance of mitochondria to normal function of skeletal and smooth muscle and the nervous system. The mitochondrial disorder affecting the gut is called mitochondrial neurogastrointestinal encephalomyopathy (MNGIE), but is also referred to as mitochondrial encephalomyopathy with sensorimotor polyneuropathy, ophthalmoplegia, and pseudoobstruction; oculogastrointestinal muscular dystrophy; or familial visceral myopathy type II. This is an autosomal recessive condition with gastrointestinal and hepatic manifestations that may present at any age—typically with hepatomegaly or hepatic failure in the neonate, seizures or diarrhea in infancy, and hepatic failure or chronic intestinal pseudo-obstruction in children and adults. MNGIE is characterized clinically by severe gastrointestinal dysmotility, external ophthalmoplegia, ptosis, peripheral neuropathy, and leukoencephalopathy. The small intestine is dilated or has multiple diverticula, and the amplitude of contractions is typical of a myopathic disorder.123 Some patients have a combination of intestinal dysmotility and Kearns-Sayre syndrome, or transfer dysphagia resulting from abnormal coordination and propagation of the swallow through the pharynx and the skeletal muscle portion of the esophagus, possible reflecting

Management of Gut Dysmotility

cranial nerve disorders. This becomes even more devastating when the smooth muscle portion of the esophagus is affected by the associated MNGIE. Apart from the obvious external ophthalmoplegia, these patients manifest skeletal muscle pain and cramps and systemic (lactic) acidosis. Circulating muscle enzyme levels (creatine phosphokinase, alanine aminotransferase, aldolase, etc.) are elevated, and muscle biopsy shows characteristic ragged red fibers on modified Gomori stain. This appearance results from the hypertrophy of mitochondria in the subsarcolemmal position in a few muscle fibers and the lack of mitochondria in other muscle fibers. In the intestine, there is hypertrophy of the circular muscle layer and atrophy of the longitudinal muscle, and megamitochondria in myenteric neurons and muscle cells.133

REFERENCES 1. Abraham, S. S., and Yun, P. T.: Laryngopharyngeal dysmotility in multiple sclerosis. Dysphagia 17:69, 2002. 2. Aggarwal, A., Cutts, T. F., Abell, T. L., et al.: Predominant symptoms in irritable bowel syndrome correlate with specific autonomic nervous system abnormalities [see comment]. Gastroenterology 106:945, 1994. 3. Anderson, T. D., and Mirza, N.: Immediate percutaneous medialization for acute vocal fold immobility with aspiration. Laryngoscope 111:1318, 2001. 4. Anuras, S., Shirazi, S., Faulk, D. L., et al.: Surgical treatment in familial visceral myopathy. Ann. Surg. 189:306, 1979. 5. Apstein, M. D., and Dalecki-Chipperfield, K.: Spinal cord injury is a risk factor for gallstone disease. Gastroenterology 92:966, 1987. 6. Aviv, J. E.: Prospective, randomized outcome study of endoscopy versus modified barium swallow in patients with dysphagia. Laryngoscope 110:563, 2000. 7. Aviv, J. E., Kim, T., Sacco, R. L., et al.: FEESST: a new bedside endoscopic test of the motor and sensory components of swallowing. Ann. Otol. Rhinol. Laryngol. 107(5 Pt. 1):378, 1998. 8. Aviv, J. E., Spitzer, J., Cohen, M., et al.: Laryngeal adductor reflex and pharyngeal squeeze as predictors of laryngeal penetration and aspiration. Laryngoscope 112:338, 2002. 9. Battle, W. M., Rubin, M. R., Cohen, S., and Snape, W. J. Jr.: Gastrointestinal-motility dysfunction in amyloidosis. N. Engl. J. Med. 301:24, 1979. 10. Bayliss, W., and Starling, E.: The movements and innervation of the small intestine. J. Physiol. (Lond.) 24:99, 1899. 11. Bayliss, W., and Starling, E.: The movements and the innervation of the large intestine. J. Physiol. (Lond.) 26:107, 1900. 12. Bayliss, W., and Starling, E.: The movements and innervation of the small intestine. J. Physiol. (Lond.) 26:125, 1901. 13. Bharucha, A.: Fecal incontinence. Gastroenterology 124:1672, 2003. 14. Bharucha, A., and Phillips, S.: Megacolon: acute, toxic, and chronic. Curr. Treat. Op. Gastroenterol. 2:517, 1999.

2701

15. Bharucha, A. E., Camilleri, M., Low, P. A., and Zinsmeister, A. R.: Autonomic dysfunction in gastrointestinal motility disorders. Gut 34:397, 1993. 16. Borgaonkar, M. R., and Lumb, B.: Acute on chronic intestinal pseudoobstruction responds to neostigmine. Dig. Dis. Sci. 45:1644, 2000. 17. Bredenoord, A. J., Chial, H. J., Camilleri, M., et al.: Gastric accommodation and emptying in evaluation of patients with upper gastrointestinal symptoms. Clin. Gastroenterol. Hepatol. 1:264, 2003. 18. Briel, J. W., Schouten, W. R., Vlot, E. A., et al.: Clinical value of colonic irrigation in patients with continence disturbances. Dis. Colon Rectum 40:802, 1997. 19. Brindley, G.: Control of the bladder and urethral sphincters by surgically implanted electrical stimulators. In Chrisholm, D. (ed.): Scientific Foundation of Urology. London, Heinemann, p. 464, 1982. 20. Burns, T. M., Lawn, N. D., Low, P. A., et al.: Adynamic ileus in severe Guillain-Barre syndrome. Muscle Nerve 24:963, 2001. 21. Buysschaert, M., Donckier, J., Dive, A., et al.: Gastric acid and pancreatic polypeptide responses to sham feeding are impaired in diabetic subjects with autonomic neuropathy. Diabetes 34:1181, 1985. 22. Bytzer, P., Hammer, J., Talley, N. J., et al.: Gastrointestinal symptoms in diabetes mellitus are associated with diabetic complications but not with current diabetic control. Gastroenterology 118:A468, 2000. 23. Bytzer, P., Talley, N. J., Leemon, M., et al.: Prevalence of gastrointestinal symptoms associated with diabetes mellitus: a population-based survey of 15,000 adults. Arch. Intern. Med. 161:1989, 2001. 24. Calcagno, P., Ruoppolo, G., Grasso, M. G., et al.: Dysphagia in multiple sclerosis—prevalence and prognostic factors. Acta Neurol. Scand. 105:40, 2002. 25. Camilleri, M.: The current role of erythromycin in the clinical management of gastric emptying disorders [editorial; comment]. Am. J. Gastroenterol. 88:169, 1993. 26. Camilleri, M.: Appraisal of medium- and long-term treatment of gastroparesis and chronic intestinal dysmotility. Am. J. Gastroenterol. 89:1769, 1994. 27. Camilleri, M., Balm, R. K., and Low, P. A.: Autonomic dysfunction in patients with chronic intestinal pseudo-obstruction. Clin. Auton. Res. 3:95, 1993. 28. Camilleri, M., and Fealey, R. D.: Idiopathic autonomic denervation in eight patients presenting with functional gastrointestinal disease: a causal association? Dig. Dis. Sci. 35:609, 1990. 29. Camilleri, M., Hasler, W. L., Parkman, H. P., et al.: Measurement of gastrointestinal motility in the GI laboratory. Gastroenterology 115:747, 1998. 30. Camilleri, M., and Malagelada, J. R.: Abnormal intestinal motility in diabetics with the gastroparesis syndrome. Eur. J. Clin. Invest. 14:420, 1984. 31. Camilleri, M., Malagelada, J. R., Brown, M. L., et al.: Relation between antral motility and gastric emptying of solids and liquids in humans. Am. J. Physiol. 249(5 Pt. 1): G580, 1985. 32. Camilleri, M., Malagelada, J. R., Stanghellini, V., et al.: Gastrointestinal motility disturbances in patients with orthostatic hypotension. Gastroenterology 88:1852, 1985.

2702

Diseases of the Peripheral Nervous System

33. Camilleri, M., Saslow, S. B., and Bharucha, A. E.: Gastrointestinal sensation: mechanisms and relation to functional gastrointestinal disorders. Gastroenterol. Clin. North Am. 25:247, 1996. 34. Cannon, W.: The motor activities of the stomach and small intestine after splanchnic and vagus section. Am. J. Physiol. 17:429, 1906. 35. Carrau, R. L., Pou, A., Eibling, D. E., et al.: Laryngeal framework surgery for the management of aspiration. Head Neck 21:139, 1999. 36. Caruana, B. J., Wald, A., Hinds, J. P., and Eidelman, B. H.: Anorectal sensory and motor function in neurogenic fecal incontinence: comparison between multiple sclerosis and diabetes mellitus. Gastroenterology 100:465, 1991. 37. Chaudhuri, K. R., Thomaides, T., and Mathias, C. J.: Abnormality of superior mesenteric artery blood flow responses in human sympathetic failure. J. Physiol. (Lond.) 457:477, 1992. 38. Chia, Y. W., Fowler, C. J., Kamm, M. A., et al.: Prevalence of bowel dysfunction in patients with multiple sclerosis and bladder dysfunction. J. Neurol. 242:105, 1995. 39. Chia, Y. W., Gill, K. P., Jameson, J. S., et al.: Paradoxical puborectalis contraction is a feature of constipation in patients with multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 60:31, 1996. 40. Chia, Y. W., Lee, T. K., Kour, N. W., et al.: Microchip implants on the anterior sacral roots in patients with spinal trauma: does it improve bowel function? Dis. Colon Rectum 39:690, 1996. 41. Chinn, J. S., and Schuffler, M. D.: Paraneoplastic visceral neuropathy as a cause of severe gastrointestinal motor dysfunction. Gastroenterology 95:1279, 1988. 42. Choi, M. G., Camilleri, M., O’Brien, M. D., et al.: A pilot study of motility and tone of the left colon in patients with diarrhea due to functional disorders and dysautonomia. Am. J. Gastroenterol. 92:297, 1997. 43. Clouse, R. E., and Lustman, P. J.: Gastrointestinal symptoms in diabetic patients: lack of association with neuropathy [see comment]. Am. J. Gastroenterol. 84:868, 1989. 44. Coker, R. J., Horner, P., Bleasdale-Barr, K., et al.: Increased gut parasympathetic activity and chronic diarrhoea in a patient with the acquired immunodeficiency syndrome. Clin. Auton. Res. 2:295, 1992. 45. Colemont, L. J., and Camilleri, M.: Chronic intestinal pseudo-obstruction: diagnosis and treatment. Mayo Clin. Proc. 64:60, 1989. 46. Cook, I. J., and Kahrilas, P. J.: AGA technical review on management of oropharyngeal dysphagia. Gastroenterology 116:455, 1999. 47. Costa, M., Brookes, S. J., and Hennig, G. W.: Anatomy and physiology of the enteric nervous system. Gut 47(Suppl. 4):iv15; discussion iv26, 2000. 48. Coulie, B., Szarka, L. A., Camilleri, M., et al.: Recombinant human neurotrophic factors accelerate colonic transit and relieve constipation in humans [see comment]. Gastroenterology 119:41, 2000. 49. Dalakas, M. C.: Polymyositis, dermatomyositis and inclusionbody myositis [see comment]. N. Engl. J. Med. 325:1487, 1991.

50. De Pauw, A., Dejaeger, E., D’Hooghe, B., and Carton, H.: Dysphagia in multiple sclerosis. Clin. Neurol. Neurosurg. 104:345, 2002. 51. Delgado-Aros, S., Kim, D. Y., Burton, D. D., et al.: Effect of GLP-1 on gastric volume, emptying, maximum volume ingested, and postprandial symptoms in humans. Am. J. Physiol. Gastrointest. Liver Physiol. 282:G424, 2002. 52. Delgado-Aros, S., Vella, A., Camilleri, M. C., et al.: Effects of glucagon-like peptide-1 and feeding on gastric volumes in diabetes mellitus with cardio-vagal dysfunction. Neurogastroenterol. Motil. (in press). 53. Deshmukh, G. R., Barkel, D. C., Sevo, D., and Hergenroeder, P.: Use or misuse of colostomy to heal pressure ulcers. Dis. Colon Rectum 39:737, 1996. 54. Diamant, N. E., Kamm, M. A., Wald, A., and Whitehead, W. E.: AGA technical review on anorectal testing techniques. Gastroenterology 116:735, 1999. 55. Dyck, P. J., Karnes, J. L., O’Brien, P. C., et al.: The Rochester Diabetic Neuropathy Study: reassessment of tests and criteria for diagnosis and staged severity. Neurology 42:1164, 1992. 56. Edwards, L. L., Pfeiffer, R. F., Quigley, E. M., et al.: Gastrointestinal symptoms in Parkinson’s disease. Mov. Disord. 6:151, 1991. 57. Enck, P.: Biofeedback training in disordered defecation: a critical review. Dig. Dis. Sci. 38:1953, 1993. 58. Ewing, D. J., Campbell, I. W., Murray, A., et al.: Immediate heart-rate response to standing: simple test for autonomic neuropathy in diabetes. Br. Med. J. 1:145, 1978. 59. Ewing, D. J., and Clarke, B. F.: Diagnosis and management of diabetic autonomic neuropathy. Br. Med. J. (Clin. Res. Ed.) 285:916, 1982. 60. Fajardo, N., Pasiliao, R., Duncan, R., et al.: The effect of neuromuscular stimulation of the abdominal muscles on defecation after spinal cord injury. Gastroenterology 120(5 Suppl. 1):A-109, 2001. 61. Fedorak, R. N., Field, M., and Chang, E. B.: Treatment of diabetic diarrhea with clonidine. Ann. Intern. Med. 102:197, 1985. 62. Feldman, M., and Schiller, L. R.: Disorders of gastrointestinal motility associated with diabetes mellitus. Ann. Intern. Med. 98:378, 1983. 63. Finucane, T. E., Christmas, C., and Travis, K.: Tube feeding in patients with advanced dementia: a review of the evidence [see comment]. JAMA 282:1365, 1999. 64. Fox, K. A., Mularski, R. A., Sarfati, M. R., et al.: Aspiration pneumonia following surgically placed feeding tubes. Am. J. Surg. 170:564; discussion 566, 1995. 65. Frank, J. W., Sarr, M. G., and Camilleri, M.: Use of gastroduodenal manometry to differentiate mechanical and functional intestinal obstruction: an analysis of clinical outcome. Am. J. Gastroenterol. 89:339, 1994. 66. Frank, J. W., Saslow, S. B., Camilleri, M., et al.: Mechanism of accelerated gastric emptying of liquids and hyperglycemia in patients with type II diabetes mellitus. Gastroenterology 109:755, 1995. 67. Fraser, R. J., Horowitz, M., Maddox, A. F., et al.: Hyperglycaemia slows gastric emptying in type 1 (insulindependent) diabetes mellitus. Diabetologia 33:675, 1990.

Management of Gut Dysmotility 68. Frost, F., Hartwig, D., Jaeger, R., et al.: Electrical stimulation of the sacral dermatomes in spinal cord injury: effect on rectal manometry and bowel emptying. Arch. Phys. Med. Rehabil. 74:696, 1993. 69. Fujimura, J., Camilleri, M., Low, P. A., et al.: Effect of perturbations and a meal on superior mesenteric artery flow in patients with orthostatic hypotension. J. Auton. Nerv. Syst. 67:15, 1997. 70. Gerharz, E. W., Vik, V., Webb, G., et al.: The value of the MACE (Malone antegrade colonic enema) procedure in adult patients. J. Am. Coll. Surg. 185:544, 1997. 71. Gill, K. P., Chia, Y. W., Henry, M. M., and Shorvon, P. J.: Defecography in multiple sclerosis patients with severe constipation. Radiology 191:553, 1994. 72. Glick, M. E., Meshkinpour, H., Haldeman, S., et al.: Colonic dysfunction in multiple sclerosis. Gastroenterology 83:1002, 1982. 73. Glick, M. E., Meshkinpour, H., Haldeman, S., et al.: Colonic dysfunction in patients with thoracic spinal cord injury. Gastroenterology 86:287, 1984. 74. Gonella, J., Bouvier, M., and Blanquet, F.: Extrinsic nervous control of motility of small and large intestines and related sphincters. Physiol. Rev. 67:902, 1987. 75. Greydanus, M. P., Vassallo, M., Camilleri, M., et al.: Neurohormonal factors in functional dyspepsia: insights on pathophysiological mechanisms. Gastroenterology 100(5 Pt. 1):1311, 1991. 76. Griffin, G. E., Miller, A., Batman, P., et al.: Damage to jejunal intrinsic autonomic nerves in HIV infection. AIDS 2:379, 1988. 77. Haldeman, S., Glick, M., Bhatia, N. N., et al.: Colonometry, cystometry, and evoked potentials in multiple sclerosis. Arch. Neurol. 39:698, 1982. 78. Haruma, K., Wiste, J. A., and Camilleri, M.: Effect of octreotide on gastrointestinal pressure profiles in health and in functional and organic gastrointestinal disorders. Gut 35:1064, 1994. 79. He, C. L., Soffer, E. E., Ferris, C. D., et al.: Loss of interstitial cells of Cajal and inhibitory innervation in insulindependent diabetes. Gastroenterology 121:427, 2001. 80. Hietaharju, A., Jaaskelainen, S., Kalimo, H., and Hietarinta, M.: Peripheral neuromuscular manifestations in systemic sclerosis (scleroderma). Muscle Nerve 16:1204, 1993. 81. Hinds, J. P., Eidelman, B. H., and Wald, A.: Prevalence of bowel dysfunction in multiple sclerosis: a population survey. Gastroenterology 98:1538, 1990. 82. Hinds, J. P., and Wald, A.: Colonic and anorectal dysfunction associated with multiple sclerosis. Am. J. Gastroenterol. 84:587, 1989. 83. Holas, M. A., DePippo, K. L., and Reding, M. J.: Aspiration and relative risk of medical complications following stroke. Arch. Neurol. 51:1051, 1994. 84. Hornby, A.: The MS sufferer in the community. Nurs. Times 74(19 Suppl.):130, 1978. 85. Jacob, H., Berkowitz, D., McDonald, E., et al.: The esophageal motility disorder of polymyositis: a prospective study. Arch. Intern. Med. 143:2262, 1983. 86. Jameson, J. S., Rogers, J., Chia, Y. W., et al.: Pelvic floor function in multiple sclerosis. Gut 35:388, 1994.

2703

87. Janssens, J., Peeters, T. L., Vantrappen, G., et al.: Improvement of gastric emptying in diabetic gastroparesis by erythromycin: preliminary studies [see comments]. N. Engl. J. Med. 322:1028, 1990. 88. Kassander, P.: Asymptomatic gastric retention in diabetics (gastroparesis diabeticorum). Ann. Intern. Med. 48:797, 1958. 89. Katzka, D. A.: Motility abnormalities in gastroesophageal reflux disease. Gastroenterol. Clin. North Am. 28:905, 1999. 90. Ketover, S. R., Ansel, H. J., Goldish, G., et al.: Gallstones in chronic spinal cord injury: is impaired gallbladder emptying a risk factor? Arch. Phys. Med. Rehabil. 77:1136, 1996. 91. Kim, D. Y., Myung, S. J., and Camilleri, M.: Novel testing of human gastric motor and sensory functions: rationale, methods, and potential applications in clinical practice. Am. J. Gastroenterol. 95:3365, 2000. 92. Kreulen, D., and Szurszewski, J. H.: Reflex pathways in the abdominal prevertebral ganglia: evidence for the colocolonic inhibitory reflex. J. Physiol. (Lond.) 295:21, 1979. 93. Krogh, K., Nielsen, J., Djurhuus, J. C., et al.: Colorectal function in patients with spinal cord lesions. Dis. Colon Rectum 40:1233, 1997. 94. Lai, J. M., Chuang, T. Y., Francisco, G. E., and Strayer, J. R.: Diversion colitis: a cause of abdominal discomfort in spinal cord injury patients with colostomy. Arch. Phys. Med. Rehabil. 78:670, 1997. 95. Law, N. M., Bharucha, A. E., Undale, A. S., and Zinsmeister, A. R.: Cholinergic stimulation enhances colonic motor activity, transit, and sensation in humans. Am. J. Physiol. Gastrointest. Liver Physiol. 281:G1228, 2001. 96. Lazarus, B. A., Murphy, J. B., and Culpepper, L.: Aspiration associated with long-term gastric versus jejunal feeding: a critical analysis of the literature. Arch. Phys. Med. Rehabil. 71:46, 1990. 97. Leder, S. B., and Espinosa, J. F.: Aspiration risk after acute stroke: comparison of clinical examination and fiberoptic endoscopic evaluation of swallowing [see comment]. Dysphagia 17:214, 2002. 98. Lee, H. R., Lennon, V. A., Camilleri, M., and Prather, C. M.: Paraneoplastic gastrointestinal motor dysfunction: clinical and laboratory characteristics. Am. J. Gastroenterol. 96:373, 2001. 99. Leelakusolvong, S., Bharucha, A. E., Sarr, M. G., et al.: Effect of extrinsic denervation on muscarinic neurotransmission in the canine ileocolonic region. Neurogastroenterol. Motil. 15:173, 2003. 100. Leelakusolvong, S., Sarr, M. G., Miller, S. M., et al.: Role of extrinsic innervation in modulating nitrergic transmission in the canine ileocolonic region. Am. J. Physiol. Gastrointest. Liver Physiol. 283:G230, 2002. 101. Leigh, R. J., and Turnberg, L. A.: Faecal incontinence: the unvoiced symptom. Lancet 1:1349, 1982. 102. Lennon, V. A., Sas, D. F., Busk, M. F., et al.: Enteric neuronal autoantibodies in pseudoobstruction with small-cell lung carcinoma [see comments]. Gastroenterology 100:137, 1991. 103. Lin, V. W., Nino-Murcia, M., Frost, F., et al.: Functional magnetic stimulation of the colon in persons with spinal cord injury. Arch. Phys. Med. Rehabil. 82:167, 2001.

2704

Diseases of the Peripheral Nervous System

104. Locke, G. R., Pemberton, J. H., and Phillips, S.: AGA technical review on constipation. Gastroenterology 119:1766, 2000. 105. Logemann, J. A.: Evaluation and Treatment of Swallowing Disorders. San Diego, CA, College Hill Press, 1983. 106. Low, P. A: Composite autonomic scoring scale for laboratory quantification of generalized autonomic failure. Mayo Clin. Proc. 68:748, 1993. 107. Lucchinetti, C. F., Kimmel, D. W., and Lennon, V. A.: Paraneoplastic and oncologic profiles of patients seropositive for type 1 antineuronal nuclear autoantibodies. Neurology 50:652, 1998. 108. Lundgren, O.: Sympathetic input into the enteric nervous system. Gut 47(Suppl. 4):iv33; discussion iv36, 2000. 109. Lustman, P. J., Anderson, R. J., Freedland, K. E., et al.: Depression and poor glycemic control: a meta-analytic review of the literature. Diabetes Care 23:934, 2000. 110. Maleki, D., Camilleri, M., Zinsmeister, A. R., and Rizza, R. A.: Effect of acute hyperglycemia on colorectal motor and sensory function in humans. Am. J. Physiol. 273(4 Pt. 1):G859, 1997. 111. Maleki, D., Locke, G. R. 3rd, Camilleri, M., et al.: Gastrointestinal tract symptoms among persons with diabetes mellitus in the community. Arch. Intern. Med. 160:2808, 2000. 112. Mansell, P. I., Tattersall, R. B., Balsitis, M., et al.: Megaduodenum due to hollow visceral myopathy successfully managed by duodenoplasty and feeding jejunostomy. Gut 32:334, 1991. 113. Marlett, J. A., and Code, C. F.: Effects of celiac and superior mesenteric ganglionectomy on interdigestive myoelectric complex in dogs. Am. J. Physiol. 237:E432, 1979. 114. Mathers, S. E., Ingram, D. A., and Swash, M.: Electrophysiology of motor pathways for sphincter control in multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 53:955, 1990. 115. Mathers, S. E., Kempster, P. A., Law, P. J., et al.: Anal sphincter dysfunction in Parkinson’s disease. Arch. Neurol. 46:1061, 1989. 116. McCallum, R. W., Chen, J. D., Lin, Z., et al.: Gastric pacing improves emptying and symptoms in patients with gastroparesis [see comment]. Gastroenterology 114:456, 1998. 117. McCallum, R. W., Ricci, D. A., Rakatansky, H., et al.: A multicenter placebo-controlled clinical trial of oral metoclopramide in diabetic gastroparesis. Diabetes Care 6:463, 1983. 118. Mearin, F., Camilleri, M., and Malagelada, J. R.: Pyloric dysfunction in diabetics with recurrent nausea and vomiting. Gastroenterology 90:1919, 1986. 119. Mearin, F., and Malagelada, J. R.: Gastroparesis and dyspepsia in patients with diabetes mellitus. Eur. J. Gastroenterol. Hepatol. 7:717, 1995. 120. Metcalf, A. M., Phillips, S. F., Zinsmeister, A. R., et al.: Simplified assessment of segmental colonic transit. Gastroenterology 92:40, 1987. 121. Mintchev, M. P., Sanmiguel, C. P., Amaris, M., and Bowes, K. L.: Microprocessor-controlled movement of solid gastric content using sequential neural electrical stimulation [see comment]. Gastroenterology 118:258, 2000.

122. Moonka, R., Stiens, S. A., Resnick, W. J., et al.: The prevalence and natural history of gallstones in spinal cord injured patients. J. Am. Coll. Surg. 189:274, 1999. 123. Mueller, L. A., Camilleri, M., and Emslie-Smith, A. M.: Mitochondrial neurogastrointestinal encephalomyopathy: manometric and diagnostic features. Gastroenterology 116:959, 1999. 124. Murr, M. M., Sarr, M. G., and Camilleri, M.: The surgeon’s role in the treatment of chronic intestinal pseudoobstruction. Am. J. Gastroenterol. 90:2147, 1995. 125. Nelson, A., Malassigne, P., Amerson, T., et al.: Descriptive study of bowel care practices and equipment in spinal cord injury. Sci. Nurs. 10:65, 1993. 126. Nelson, A. L., Malassigne, P., and Murray, J.: Comparison of seat pressures on three bowel care/shower chairs in spinal cord injury. Sci. Nurs. 11:105, 1994. 127. Newcomer, A. D.: Can diarrhea cause secondary steatorrhea? [see comment]. Gastroenterology 102:2163, 1992. 128. Nordenbo, A. M., Andersen, J. R., and Andersen, J. T.: Disturbances of ano-rectal function in multiple sclerosis. J. Neurol. 243:445, 1996. 129. Nowak, T. V., Johnson, C. P., Kalbfleisch, J. H., et al.: Highly variable gastric emptying in patients with insulin dependent diabetes mellitus [see comment]. Gut 37:23, 1995. 130. Ordog, T., Takayama, I., Cheung, W. K., et al.: Remodeling of networks of interstitial cells of Cajal in a murine model of diabetic gastroparesis. Diabetes 49:1731, 2000. 131. Osserman, K. E., and Genkins, G.: Studies in myasthenia gravis: review of a twenty-year experience in over 1200 patients. Mt. Sinai J. Med. 38:497, 1971. 132. Palace, J., Chandiramani, V. A., and Fowler, C. J.: Value of sphincter electromyography in the diagnosis of multiple system atrophy. Muscle Nerve 20:1396, 1997. 133. Perez-Atayde, A. R., Fox, V., Teitelbaum, J. E., et al.: Mitochondrial neurogastrointestinal encephalo-myopathy: diagnosis by rectal biopsy. Am. J. Surg. Pathol. 22:1141, 1998. 134. Perry, L., and Love, C. P.: Screening for dysphagia and aspiration in acute stroke: a systematic review. Dysphagia 16:7, 2001. 135. Ponec, R. J., Saunders, M. D., and Kimmey, M. B.: Neostigmine for the treatment of acute colonic pseudoobstruction [see comment]. N. Engl. J. Med. 341:137, 1999. 136. Powley, T. L.: Vagal input to the enteric nervous system. Gut 47(Suppl. 4):iv30; discussion iv36, 2000. 137. Puet, T. A., Jackson, H., and Amy, S.: Use of pulsed irrigation evacuation in the management of the neuropathic bowel. Spinal Cord 35:694, 1997. 138. Randell, N., Lynch, A. C., Anthony, A., et al.: Does a colostomy alter quality of life in patients with spinal cord injury? A controlled study. Spinal Cord 39:279, 2001. 139. Ron, Y., Wainstein, J., Leibovitz, A., et al.: The effect of acarbose on the colonic transit time of elderly long-term care patients with type 2 diabetes mellitus. J. Gerontol. A Biol. Sci. Med. Sci. 57:M111, 2002. 140. Rontal, E., Rontal, M., Morse, G., and Brown, E. M.: Vocal cord injection in the treatment of acute and chronic aspiration. Laryngoscope 86:625, 1976. 141. Saltzstein, R., and Romano, J.: The efficacy of colostomy as a bowel management alternative in selected spinal cord injured patients. J. Am. Paraplegia Soc. 13:9, 1990.

Management of Gut Dysmotility 142. Samsom, M., Jebbink, R., Akkermans, L., et al.: Abnormalities of antroduodenal motility in type 1 diabetes. Diabetes Care 19:21, 1996. 143. Samsom, M., Szarka, L. A., Camilleri, M., et al.: Pramlintide, an amylin analog, selectively delays gastric emptying: potential role of vagal inhibition. Am. J. Physiol. Gastrointest. Liver Physiol. 278:G946, 2000. 144. Shibata, C., Balsiger, B. M., Anding, W. J., and Sarr, M. G.: Adrenergic denervation-hypersensitivity in ileal circular smooth muscle after small bowel transplantation in rats. Dig. Dis. Sci. 42:2213, 1997. 145. Shy, G. M., and Drager, G. A.: A neurological syndrome associated with orthostatic hypotension: a clinical-pathologic study. Arch. Neurol. 2:511, 1960. 146. Singaram, C., Ashraf, W., Gaumnitz, E. A., et al.: Dopaminergic defect of enteric nervous system in Parkinson’s disease patients with chronic constipation. Lancet 346:861, 1995. 147. Singh, G., and Triadafilopoulos, G.: Gastroesophageal reflux disease in patients with spinal cord injury. J. Spinal Cord Med. 23:23, 2000. 148. Sodhi, N., Camilleri, M., Camoriano, J. K., et al.: Autonomic function and motility in intestinal pseudoobstruction caused by paraneoplastic syndrome. Dig. Dis. Sci. 34:1937, 1989. 149. Sorensen, M., Lorentzen, M., Petersen, J., and Christiansen, J.: Anorectal dysfunction in patients with urologic disturbance due to multiple sclerosis. Dis. Colon Rectum 34:136, 1991. 150. Soudah, H. C., Hasler, W. L., and Owyang, C.: Effect of octreotide on intestinal motility and bacterial overgrowth in scleroderma [see comment]. N. Engl. J. Med. 325:1461, 1991. 151. Soykan, I., Sarosiek, I., and McCallum, R. W.: The effect of chronic oral domperidone therapy on gastrointestinal symptoms, gastric emptying, and quality of life in patients with gastroparesis. Am. J. Gastroenterol. 92:976, 1997. 152. Stanghellini, V., Tosetti, C., Paternico, A., et al.: Risk indicators of delayed gastric emptying of solids in patients with functional dyspepsia [see comment]. Gastroenterology 110:1036, 1996. 153. Stiens, SA, Braunschweig, C, Cowel, J, et al., for the Consortium for Spinal Cord Medicine: Neurogenic Bowel Management in Adults with Spinal Cord Injury (Clinical Practice Guidelines). Washington, DC, Paralyzed Veterans of America, p. 39, 1998. 154. Stiens, SA, Braunschweig, C, Cowel, J, et al., for the Consortium for Spinal Cord Medicine: Neurogenic Bowel: What You Should Know. A Guide for People with Spinal Cord Injury (Clinical Practice Consumer Guidelines). Washington, DC, Paralyzed Veterans of America, p. 53, 1999. 155. Stiens, S. A., Fajardo, N., and Korsten, M. A.: The gastrointestinal system after spinal cord injury. In Lin, V. W. (ed.): Spinal Cord Medicine: Principles and Practice. New York, Demos Publishing, p. 321, 2003. 156. Stone, J. M., Nino-Murcia, M., Wolfe, V. A., and Perkash, I.: Chronic gastrointestinal problems in spinal cord injury patients: a prospective analysis. Am. J. Gastroenterol. 85:1114, 1990. 157. Suarez, G. A., Fealey, R. D., Camilleri, M., and Low, P. A.: Idiopathic autonomic neuropathy: clinical, neurophysio-

158.

159.

160.

161.

162.

163.

164.

165.

166.

167.

168.

169. 170.

171.

172. 173.

174.

175.

2705

logic, and follow-up studies on 27 patients. Neurology 44:1675, 1994. Sun, W. M., Read, N. W., and Donnelly, T. C.: Anorectal function in incontinent patients with cerebrospinal disease. Gastroenterology 99:1372, 1990. Tack, J., Piessevaux, H., Coulie, B., et al.: Role of impaired gastric accommodation to a meal in functional dyspepsia [see comment]. Gastroenterology 115:1346, 1998. Takahashi, T., and Owyang, C.: Characterization of vagal pathways mediating gastric accommodation reflex in rats. J. Physiol. (Lond.) 504(Pt. 2):479, 1997. Tandon, R. K., Jain, R. K., and Garg, P. K.: Increased incidence of biliary sludge and normal gall bladder contractility in patients with high spinal cord injury. Gut 41:682, 1997. Taylor, M. C., Bradley, W. E., Bhatia, N., et al.: The conus demyelination syndrome in multiple sclerosis. Acta Neurol. Scand. 69:80, 1984. Thomforde, G. M., Camilleri, M., Phillips, S. F., and Forstrom, L. A.: Evaluation of an inexpensive screening scintigraphic test of gastric emptying. J. Nucl. Med. 36:93, 1995. Tougas, G., Chen, Y., Coates, G., et al.: Standardization of a simplified scintigraphic methodology for the assessment of gastric emptying in a multicenter setting. Am. J. Gastroenterol. 95:78, 2000. Tougas, G., Eaker, E. Y., Abell, T. L., et al.: Assessment of gastric emptying using a low fat meal: establishment of international control values. Am. J. Gastroenterol. 95:1456, 2000. Tougas, G., and Huizinga, J. D.: Gastric pacing as a treatment for intractable gastroparesis: shocking news? [see comment]. Gastroenterology 114:598, 1998. Tymms, K. E., and Webb, J.: Dermatopolymyositis and other connective tissue diseases: a review of 105 cases. J. Rheumatol. 12:1140, 1985. Valdovinos, M. A., Camilleri, M., and Zimmerman, B. R.: Chronic diarrhea in diabetes mellitus: mechanisms and an approach to diagnosis and treatment. Mayo Clin. Proc. 68:691, 1993. Varma, J. S.: Autonomic influences on colorectal motility and pelvic surgery. World J. Surg. 16:811, 1992. Vassallo, M., Camilleri, M., Caron, B. L., and Low, P. A.: Gastrointestinal motor dysfunction in acquired selective cholinergic dysautonomia associated with infectious mononucleosis. Gastroenterology 100:252, 1991. Vella, A., Lee, J. S., Camilleri, M., et al.: Effects of pramlintide, an amylin analogue, on gastric emptying in type 1 and 2 diabetes mellitus. Neurogastroenterol. Motil. 14:123, 2002. Viets, H.: Diagnosis of myasthenia gravis in patients with dysphagia. JAMA 134:987, 1947. von der Ohe, M. R., and Camilleri, M.: Measurement of small bowel and colonic transit: indications and methods. Mayo Clin. Proc. 67:1169, 1992. von der Ohe, M. R., Camilleri, M., Thomforde, G. M., and Klee, G. G.: Differential regional effects of octreotide on human gastrointestinal motor function. Gut 36:743, 1995. Wald, A., and Tunuguntla, A. K.: Anorectal sensorimotor dysfunction in fecal incontinence and diabetes mellitus: modification with biofeedback therapy. N. Engl. J. Med. 310:1282, 1984.

2706

Diseases of the Peripheral Nervous System

176. Waldron, D. J., Horgan, P. G., Patel, F. R., et al.: Multiple sclerosis: assessment of colonic and anorectal function in the presence of faecal incontinence. Int. J. Colorectal Dis. 8:220, 1993. 177. Ward, S. M., and Sanders, K. M.: Interstitial cells of Cajal: primary targets of enteric motor innervation. Anat. Rec. 262:125, 2001. 178. Watkins, C. C., Sawa, A., Jaffrey, S., et al.: Insulin restores neuronal nitric oxide synthase expression and function that is lost in diabetic gastropathy. J. Clin. Invest. 106:373, 2000. [erratum appears in J. Clin. Invest. 106:803, 2000.] 179. Weber, J., Denis, P., Mihout, B., et al.: Effect of brain-stem lesion on colonic and anorectal motility: study of three patients. Dig. Dis. Sci. 30:419, 1985. 180. Weber, J., Grise, P., Roquebert, M., et al.: Radiopaque markers transit and anorectal manometry in 16 patients

181.

182.

183.

184.

with multiple sclerosis and urinary bladder dysfunction. Dis. Colon Rectum 30:95, 1987. Whitehead, W. E., Wald, A., Diamant, N., et al.: Functional disorders of the anus and rectum. Gut 45(Suppl. II):II55, 1999. Wiesel, P. H., Norton, C., Roy, A. J., et al.: Gut focused behavioural treatment (biofeedback) for constipation and faecal incontinence in multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 69:240, 2000. Wood, J. R., Camilleri, M., Low, P. A., and Malagelada, J. R.: Brainstem tumor presenting as an upper gut motility disorder. Gastroenterology 89:1411, 1985. Wortman, R.: Inflammatory disease of muscle and other myopathies. In Kelley, W., Harris, E., Ruddy, S., and Sledge, C. (eds.): Textbook of Rheumatology. Philadelphia, W. B. Saunders, p. 1177, 1997.

122 Peripheral Neuropathies in Infants and Children: Polyneuropathies, Mononeuropathies, Plexopathies, and Radiculopathies TED M. BURNS, MONIQUE M. RYAN, BASIL T. DARRAS, AND H. ROYDEN JONES, JR.

Pediatric Polyneuropathies: An Overview Epidemiology Approach to Evaluation of a Child with a Polyneuropathy Infantile Polyneuropathies Infantile Acquired Acute Polyneuropathies Infantile Chronic Polyneuropathies Infantile Hereditary Metabolic Polyneuropathies

Polyneuropathies of Later Childhood Acquired Acute Childhood Polyneuropathies Acquired Chronic Childhood Polyneuropathies Hereditary Chronic Childhood Polyneuropathies Pediatric Mononeuropathies Evaluation of Pediatric Mononeuropathies

A unique spectrum of neuropathies, mononeuropathies, plexopathies, and radiculopathies is seen within the pediatric population. Etiology, presentation, evaluation, management, course, and other aspects of disease are frequently quite different from analogous conditions of adulthood. This chapter reviews these differences and emphasizes those characteristics unique to the pediatric population. More detailed discussion of a particular disease can usually be found in other chapters of this book.

PEDIATRIC POLYNEUROPATHIES: AN OVERVIEW Epidemiology Two thirds to three quarters of pediatric polyneuropathies are inherited, Charcot-Marie-Tooth disease (CMT) being the most common.143,167,198,355,361,466 The prevalence of all types of CMT has been estimated at 1 in 2500 to 1 in 10,000.131,441

Upper Extremity Mononeuropathies Lower Extremity Mononeuropathies Plexopathies of Infants and Children Brachial Plexopathy Brachial Plexopathies in Older Children Lumbosacral Plexopathies Pediatric Radiculopathies Cervical Radiculopathies Lumbosacral Radiculopathies

Acute and chronic inflammatory polyradiculoneuropathies make up the majority of the acquired polyneuropathies of childhood. Of the acquired varieties of childhood acute inflammatory demyelinating polyradiculoneuropathy (or Guillain-Barré syndrome [GBS]) has an estimated incidence of 1 in 100,000 per year. Childhood chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is less common than childhood GBS.31,59,60,62,208,232,237,443 The prevalence of childhood CIDP is estimated at less than 1 in 100,000 and the incidence is probably one tenth that of childhood GBS.316,361 One very significant difference between the adult and pediatric populations is that polyneuropathies associated with systemic diseases such as diabetes mellitus, chronic renal failure, malignancy, paraproteinemia, collagen-vascular disease, chronic alcohol use, or nutritional deficiency are very rare or nonexistent in childhood. Large case series of pediatric polyneuropathy patients provide an indication of the frequency of various disorders. They are of limited value because of referral bias, and because genetic information was often not sufficient. 2707

2708

Diseases of the Peripheral Nervous System

Sixty-one American children with a polyneuropathy, evaluated at Vanderbilt University between 1971 and 1977, included 22 with acquired disorders: 17 with GBS, and 5 with a systemic disease. Of those 23 with hereditary mechanisms, 11 had CMT, 7 had miscellaneous mechanisms and 5 had a spinocerebellar degeneration. Another 16 children had no demonstrable cause.143 In a 1988 Australian study using sural nerve biopsy in children with either a subacute or chronic polyneuropathy, a genetic disorder was found in at least 71% of cases. CMT type 1 (CMT1) was most common, followed by CMT type 2 (CMT2), Friedreich’s ataxia, and rare polyneuropathies associated with inborn errors of metabolism. There was selection bias in this study because it excluded the performance of nerve biopsy in children with a family history of probable CMT1 or GBS. In contrast to the high incidence of CMT, CIDP was diagnosed in only 11% of these pediatric polyneuropathies. In a later 12-year period, of 249 pediatric polyneuropathy cases from the Children’s Hospital of Philadelphia, hereditary motor and sensory polyneuropathy (i.e., CMT) was diagnosed in just 42% and GBS or CIDP was diagnosed in 41% of patients.442 Here also, polyneuropathies associated with collagen-vascular disease, metabolic disorders, and toxins were uncommon. Referral bias may have contributed to the higher frequency of acquired demyelinating polyneuropathies in this series. CMT1 is a much more common inherited motor and sensory polyneuropathy than is CMT2, especially in the pediatric population.48,120,121,205,355,367 The pooled European series of 287 cases of inherited peripheral neuropathy occurring in infancy and childhood identified CMT1 in 51% of patients, while CMT2 was identified in 21% of these children. CMT type 3 (Déjérine-Sottas disease [DSD]), defined in this series as autosomal recessive congenital demyelinating neuropathy, was diagnosed in another 12% of these cases. Subsequently, many instances of DSD have been found to be due to de novo mutations in the same genes that cause CMT1.48,367,380 The X-linked form of CMT (CMTX), now identified as the second most common form of CMT, was not recognized at the time of these review series. Hereditary sensory neuropathies were diagnosed in only 3% of cases. To date, large studies, such as those reported from the children’s hospitals of Sydney and Philadelphia, have not been evaluated or reported routinely using modern DNA techniques. Similarly, one wonders how carefully children with acute polyneuropathies have been routinely screened for toxic etiologies, or for the presence of an underlying hereditary lesion in those experiencing a severe medication reaction, such as may occur when vincristine is given to a child with unsuspected CMT.211,315

Approach to Evaluation of a Child with a Polyneuropathy The differential diagnosis of a pediatric patient’s polyneuropathy may be facilitated by use of an algorithm similar

to that used in adult cases, based on clinical history and a targeted physical and neurophysiologic examination (Figs. 122–1 and 122–2). Clinical History The first principle is to determine whether a suspected neuropathy is acquired or inherited. The age of onset and temporal evolution provide important clues. Inherited polyneuropathies are increasingly more likely the earlier the onset of polyneuropathy. Of those cases with a congenital or infantile-onset polyneuropathy, the vast majority are inherited. Occasionally the neuropathy may be due to GBS or CIDP.226,298,402 In the older child the majority of polyneuropathies are still most commonly inherited, although GBS or CIDP occur more frequently than in infants. Age of onset is also helpful, because the various inherited polyneuropathies tend to present at certain ages. The differential diagnosis of a newly diagnosed demyelinating polyneuropathy in a baby or late toddler includes not only the demyelinating forms of CMT, GBS, and CIDP, but also globoid cell leukodystrophy (GLD) up to age 6 months and, later, metachromatic leukodystrophy (MLD). In contrast, the latter are much less likely in children older than age 4 years, although we rarely do find MLD at this age. There are many other examples of inherited polyneuropathies with predilections for certain ages of onset. A thorough family history and examination of the parents and siblings is particularly important. It is one of the most useful diagnostic studies. The clinician should ask whether any individuals in the family have had difficulties walking or high arches of the feet. Subsequently, living family members with suspicious symptoms (or even without symptoms) should be interviewed and examined. Given the high likelihood that most pediatric polyneuropathies have a genetic basis, the family history must always receive the highest priority in the child’s evaluation. The onset of polyneuropathy is difficult to determine in young children. This is especially true since children are unlikely to report loss of strength, clumsiness, or gait trouble or sensory loss. Parents may not recognize such abnormalities unless they are definite. The parents’ uncertainty as to the time of onset is a clue that the polyneuropathy may have a genetic basis. The time when a school nurse, pediatrician, or orthopedic surgeon first observes problems such as foot deformity or gait problems to the attention of the parent or child is unlikely to be the onset of the disease. Children who report an acute or subacute onset or fluctuating course probably have an acquired neuropathy. Physicians need to ask specific questions about the child’s present and past activities, milestones, and motor skills. Was the child ever able to walk or run? Is the child falling or tripping more frequently? What specific limitations does the child have now in skills that he or she could once perform? In general, an unequivocal loss of skills or milestones raises

2709

Peripheral Neuropathies in Infants and Children

• CNS disease is the most common cause of infantile hypotania • Of NM etiologies, SMA, myopathies and disorders of NM junction more common causes of infantile weakness than polyneuropathy

History, examination, EMG, CSF examination

Examination and EMG suggests sensorimotor polyneuropathy

Demyelinating

Axonal

CNS demyelination? No

Yes

Check molgenwww.uia.ac.be/CMTMutations/ or other sources for updates on mutations responsible for inherited polyneuropathies of the CMT and DSD phenotypes

Consider ruling out SMA

Diaphragmatic weakness? Genetic testing for PMP22, MPZ and EGR2 point mutations

GLD, MLD or other neurometabolic disease?

Yes

No

No No Autosomal recessive CMT? GDAP1, MT MR2, NDRG1, PRX, EGR2 mutation? No

Sural or other sensory nerve biopsy

Consider SMA, CMT2C, HMN type VII, HMN IV, SMARD (dSMA VI). Genetic testing available only for SMA at time of writing. No

LMNA or PRX mutation? Infantile neuroaxonal dystrophy? Hereditary tyrosinemia? Other? No

Infantile GBS or CIDP much less likely. Consider treatment trial.

FIGURE 122–1 Suggested diagnostic algorithm for infant-onset polyneuropathy, including the “floppy-infant” presentation. CIDP ⫽ chronic inflammatory demyelinating polyradiculoneuropathy; CMT ⫽ Charcot-Marie-Tooth disease; CNS ⫽ central nervous system; CSF ⫽ cerebrospinal fluid; DSD ⫽ Déjérine-Sottas disease; EMG ⫽ electromyography; GBS ⫽ Guillain-Barré syndrome; GLD ⫽ globoid cell leukodystrophy; HMN ⫽ hereditary motor neuropathy; MLD ⫽ metachromatic leukodystrophy; NM ⫽ neuromuscular; SMA ⫽ spinal muscular atrophy; SMARD ⫽ spinal muscular atrophy with respiratory distress.

the question of an acquired etiology such as GBS or CIDP and may point away from CMT, but does not exclude other inherited etiologies, particularly inborn errors of metabolism, which usually present with developmental regression or stagnation. It should be noted, however, that the loss of skills in GBS or CIDP is limited to the gross motor area, whereas with inborn neurometabolic errors, such as MLD, the impact on the child’s development is often global (i.e., it includes cognition as well as motor performance).

Ideally, as with adult polyneuropathies, the child neurologist tries to define which peripheral nerve modalities are affected. However, it is often difficult for the child to accurately convey telltale neuropathic symptoms, Weakness may come to the attention of parents because of frequent falls, altered or clumsy gait, inability to run, or difficulty arising from the floor or a seated position. Descriptions of sensory disturbance are often lacking, inaccurate, or unorthodox. Crying, fussiness, or other expressions of discontent,

2710

Diseases of the Peripheral Nervous System

Childhood polyneuropathy

Acute

Chronic

GBS mimickers: GBS or CIDP Axonal Transverse myelitis, (first episode) Lyme disease, Botulism, Tick paralysis, Poliomyelitis, Diphtheria, CNS disease? Porphyria, HIV-related, Drugs and toxins Yes No

CTX, HSAN IV, mitochondrial, toxic, porphyria, GAN, NAD

HSAN, CMT2, Fabry, XP, GAN, mitochondrial, abetalipoproteinemia, Tangier, toxic

Primarily demyelinating

CNS disease? Yes

GLD, MLD, ALD, Cockayne

No

CMT, CIDP, Refsum, Tangier

FIGURE 122–2 Suggested diagnostic algorithm for childhood-onset polyneuropathy. ALD ⫽ adrenoleukodystrophy; CIDP ⫽ chronic inflammatory demyelinating polyradiculoneuropathy; CMT ⫽ Charcot-Marie-Tooth disease; CNS ⫽ central nervous system; CTX ⫽ cerebrotendinous xanthomatosis; GAN ⫽ giant axonal neuropathy; GBS ⫽ Guillain-Barré syndrome; GLD ⫽ globoid cell leukodystrophy; HIV ⫽ human immunodeficiency virus; HSAN ⫽ hereditary sensory and autonomic neuropathy; MLD ⫽ metachromatic leukodystrophy; NAD ⫽ neuroaxonal dystrophy; XP ⫽ xeroderma pigmentosum.

particularly in situations that may provoke neuropathic pain (e.g., ambulation), may warrant consideration as symptoms of neuropathic pain. Ataxia, sometimes manifested as only a reluctance to walk, must be investigated for a central or peripheral sensory etiology. The child and his or her family must be queried about other symptoms that may reflect concomitant systemic involvement. For example, a recent poor school performance may suggest concomitant central nervous system (CNS) involvement (e.g., MLD or mitochondrial disorders). The presence of any ophthalmologic, gastrointestinal, autonomic, and skin symptoms must also be sought and investigated (Table 122–1). Toxic etiologies, particularly glue or gasoline sniffing and, in agricultural settings, arsenic, parathion, and buckthorn, are necessary considerations in the older age group. The Neurologic Examination Etiologic clues are often observed at the time of the general examination (see Fig. 122–3). Pes cavus and hammer toe deformities are typical complications of long-standing neuropathies that are almost always inherited. Pes planus and valgus angulation of the forefoot are less common but often have similar implications.58,120,121,205,471 Pes cavus is not, however, a sine qua non of a hereditary polyneuropathy. At the Children’s Hospital, Boston (CHB), only 50% of children seen in the electromyography (EMG) laboratory for evaluation of such orthopedic abnormalities had nerve conduction studies (NCSs) compatible with a hereditary

polyneuropathy.245 The presence of enlarged peripheral nerves is a nonspecific finding occasionally observed in long-standing demyelinating and remyelinating (hypertrophic) polyneuropathies. This includes hereditary forms (i.e., CMT1) as well as acquired lesions such as CIDP. A hypertrophic greater auricular nerve may be seen or palpated. Enlargement of other nerves also needs to be sought at the medial aspect of the biceps brachii and humerus, near the olecranon groove at the elbow, and near the fibular head. Occasionally, a careful general physical examination may offer further diagnostic clues, particularly in the form of dermatologic and ophthalmologic abnormalities associated with specific childhood polyneuropathies. The formal neurologic examination is important for providing details as to the nature, distribution, and severity of peripheral nervous system (PNS) impairment. Rarely, signs of concomitant CNS involvement are unexpectedly uncovered. The examiner should assess whether the motor findings have a predominantly distal, generalized, or proximal distribution. Most polyneuropathies in children, as in adults, have a distal predominance. Predominantly proximal weakness is occasionally seen, usually in the lower extremities, and most commonly points toward the diagnosis of acute or chronic inflammatory demyelinating polyradiculoneuropathy. Almost all childhood peripheral neuropathies are bilaterally symmetrical. Very rarely, asymmetrical findings compatible with the diagnosis of mononeuritis multiplex (MNM) are seen in adolescents

Peripheral Neuropathies in Infants and Children

Table 122–1. Unique Clinical Features for Inherited Polyneuropathies Clinical Feature

Polyneuropathy

Prominent ataxia

FA, other SCA, abetalipoproteinemia, CTX, Refsum, mitochondrial (NARP), XP, vitamin E deficiency, CDG Fabry, porphyria, tyrosinemia CMT, DSD, HSAN, FA CMT, DSD, FA GLD, MLD, ALD, mitochondrial, toxins, CTX, Cockayne, GAN, NAD CTX, Tangier, Fabry, mitochondrial Refsum, hypo- or abetalipoproteinemia, mitochondrial, SCA, CDG FA, Cockayne, NAD, GAN, SCA Fabry, CTX, XP, ALD, Cockayne GAN CTX Tangier Diphtheria Diphtheria, Lyme, sarcoidosis Diphtheria, tyrosinemia Cockayne Mitochondrial, Refsum, Cockayne, XP Mitochondrial, Cockayne Hypo- or abetalipoproteinemia, vitamin E deficiency Mitochondrial SLE-associated neuropathy, rarely other Paralysis of shoulder and biceps Paralysis of shoulder, biceps brachii, and forearm extensors Complete paralysis of limb Above with Horner’s syndrome

Prominent pain Foot deformity Scoliosis CNS disease Cataracts Retinitis pigmentosa

Optic atrophy Skin changes Kinky hair Xanthomas Orange tonsils Pharyngitis Cranial neuropathies Fever Dwarfism Hearing loss Short stature Malabsorption Lipomatosis Systemic inflammatory Group I (C5-6) Group II (C5-7) Group III (C5-T1) Group IV (C5-T1)

ALD ⫽ adrenoleukodystrophy; CDG ⫽ carbohydrate-deficient glycoprotein syndrome; CMT ⫽ Charcot-Marie-Tooth disease; CNS ⫽ central nervous system; CTX ⫽ cerebrotendinous xanthomatosis; DSD ⫽ Déjérine-Sottas disease; FA ⫽ Friedreich’s ataxia; GAN ⫽ giant axonal neuropathy; GLD ⫽ globoid cell leukodystrophy; HSAN ⫽ hereditary sensory and autonomic neuropathy; MLD ⫽ metachromatic leukodystrophy; NAD ⫽ neuroaxonal dystrophy; NARP ⫽ neuropathy, ataxia, and xeroderma pigmentosum; SCA ⫽ spinocerebellar ataxia; SLE ⫽ systemic lupus erythematosus; XP ⫽ xeroderma pigmentosum.

with acquired neuropathies.418 Truncal and appendicular muscle tone is best assessed in the young child or infant by vertical and ventral suspension. Formal testing of muscle strength is often difficult in young children. An indirect assessment of muscle power is best obtained by functional testing (i.e., by one’s observations in the younger child while walking, running, rising from a chair or the floor, stepping up stairs or onto the ledge of an examining table, and manipulating objects and toys). The sensory examination is even

2711

more difficult in children, especially those under the age of 5 years. However, testing of touch (by tickle), vibration sense (looking for a behavioral response such as stilling or giggling), and temperature sensation is sometimes valuable. The muscle stretch reflexes should be present at any age. Hypo- or areflexia is in keeping with PNS involvement, including the anterior horn cell at the spinal cord level, whereas hyperreflexia suggests CNS pathology. It is important to note that the plantar responses may be extensor to 12 to 18 months of age. Reflex testing with reinforcement is sometimes necessary. Testing of cerebellar function is also very useful. The presence of appendicular ataxia, dysarthria, nystagmus, or abnormal eye movements may point toward a cerebellar rather than a sensory ataxia. Just as with the history, a very important part of the clinical assessment is the observation and examination of parents and siblings for signs of PNS disease. Even where such symptoms are denied, examination of other family members will often reveal muscle weakness, reflex abnormalities, orthopedic deformities, or myotonia. Nerve Conduction and Needle Electromyography The foremost role of electrophysiologic testing is to identify a neuropathic basis for clinical abnormalities as well as to establish the underlying pathophysiology of such disorders (i.e., the differentiation of a primary demyelinating processes from a predominantly axonal pathology). Information on the distribution, severity, and temporal course of neuropathy can also be obtained from the electrodiagnostic assessment.49,322 Parents accompanying their child to the EMG laboratory often have been misinformed or have received inadequate information in reference to the purpose and clinical value of EMG. Ideally, it is helpful to have the electromyographer speak with parents at the time the appointment is scheduled. If not, when the child and parents arrive in the EMG laboratory, the electromyographer should undertake a detailed discussion of the purpose and nature of the procedure with the parent. An understanding of the test purpose and performance helps to alleviate the anxiety and sometimes guilt the parents may feel about putting their child through a somewhat uncomfortable procedure. Parents are also more willing to assist in comforting the child during the study when they appreciate the empathy of the electromyographer and understand this study’s aims and potential benefits. It is often helpful to find a means to involve the older child in the procedure (e.g., having her or him watch the recording screen as it “builds mountains” each time the motor or sensory action potentials appear on the EMG screen). Similarly, one may ask the child to identify the sound of the motor unit potentials as “rain” or “a motorcycle.” For those children who do not tolerate the procedure, it is preferable to initially plan or to later reattempt the EMG under sedation. The major disadvantage of this

2712

Diseases of the Peripheral Nervous System

approach is the diminished opportunity to carefully examine motor unit potentials under voluntary control, but one can usually gain a reasonable sense of motor unit potential characteristics by reflex stimulation of the extremity (e.g., by tickling the sole of the foot to observe activation of the tibialis anterior or iliopsoas). The appropriate interpretation of pediatric EMG studies depends on the clinical neurophysiologist’s appreciation of the normal neurophysiologic maturational changes of infancy and early childhood. Nerve conduction attributes and motor unit potentials evolve from infancy, with values less than 50% of adult values, and reach adult values between ages 3 and 5 years. These changes occur in parallel with the maturation of peripheral nerve myelin.23,63,68,166,282,324,420,485,493 Techniques of pediatric EMG are discussed in detail in dedicated texts.243,462 In brief, we begin with the NCSs, initially with sensory testing, which is generally easy to perform and in children usually requires very low stimulus intensities.231 Motor and sensory NCSs are performed by standard technique, taking care to use the minimal stimulus required to obtain a just supramaximal compound muscle action potential (CMAP). Sensory nerve action potentials (SNAPs) are normal in infantile spinal muscular atrophy (SMA) and in congenital myopathies, the two most common neuromuscular causes of the “floppy infant.”231 A normal median SNAP in hypotonic infants is usually sufficient evidence to exclude congenital peripheral nerve lesions distal to the dorsal root ganglion. Further assessment is best made using the medial plantar nerve in infants, or the sural nerve in the older child. Normal SNAPs in these nerves provide good evidence for an intact peripheral sensory system, making a primary sensorimotor polyneuropathy unlikely. Abnormal SNAPs in the setting of infantile hypotonia suggests the possibility of either a congenital inherited polyneuropathy or, much less likely, a very-early-onset acquired autoimmune polyneuropathy.226,298,402 When performing needle EMG in a child, one examines as few muscles as are necessary to establish a diagnosis. We use a (very small) 25-gauge needle. In the older, potentially cooperative child, we liken the insertion of the needle to placing a microphone into the muscle. Generalized processes are commonly less problematic than mononeuropathies, plexopathies, or mononeuritis multiplex, wherein a detailed needle examination is required to adequately define the distribution of the nerve lesion. Differences in voluntary motor unit potential size in infants and younger children are also important to recognize. Motor unit potential amplitudes are smaller in younger children and increase in size with age.

Laboratory Testing Currently, DNA analysis has an increasingly important role in the investigation of chronic childhood polyneuropathies. This is particularly valuable for those children whose electro-

diagnostic studies demonstrate a uniform demyelinating process.92 Nerve biopsy is now less frequently necessary in the assessment of childhood polyneuropathies. Molecular genetic testing advances are obviating the need for nerve biopsy in a number of inherited peripheral nerve disorders wherein diagnosis was formerly contingent upon or supported by identification of typical pathologic changes in the nerve per se. When a particular phenotype such as CMT, GBS, or CIDP is not readily defined, screening laboratory investigations include a complete blood count, fasting blood sugar, renal function tests, and an erythrocyte sedimentation rate (ESR) test. Other clinically appropriate investigations are outlined later within this as well as other chapters in this text.

INFANTILE POLYNEUROPATHIES This section discusses a number of neuropathies, some unique to this pediatric age group. Approximately three quarters of “floppy babies” have primary disorders affecting the CNS.231,364,398,461 Twenty percent to 25% of floppy babies will ultimately be found to have neuromuscular disorders, infantile SMA, and congenital myopathies or, less commonly, are infantile polyneuropathy.190,231,363,364,398 Absence of SNAPs on EMG, if not due to technical difficulties, is an important electrodiagnostic clue to the diagnosis of neonatal polyneuropathies. A sural nerve biopsy may then prove to be quite helpful in defining these rare instances. Unfortunately, the most treatable disorders (e.g., the acquired immune-mediated forms such as GBS or CIDP) are extremely rare in infants.

Infantile Acquired Acute Polyneuropathies Guillain-Barré Syndrome GBS very rarely presents during infancy. Characteristically, these babies develop an acute hypotonia, weakness, and sometimes an associated respiratory distress as well as feeding difficulties.233 Even less commonly, GBS develops in utero or in the immediate neonatal period.226,298,402 Diagnosis of an acute infantile inflammatory demyelinating polyneuropathy is relatively straightforward if the mother was previously similarly affected during her pregnancy.298,402 In contrast, GBS is a more difficult diagnosis where there is no maternal history of weakness or sensory loss.13,226 Electrodiagnostic findings are those of acquired demyelination with slowing and dispersed CMAPs, sometimes with conduction block, as detailed in another section of this text. Needle examination sometimes demonstrates acute denervation.13,226 The EMG is particularly useful in the differentiation of infantile GBS from other causes of acquired infantile hypotonia (e.g., infantile botulism, myasthenia gravis, and, rarely, poliomyelitis).240 The cerebrospinal (CSF) protein is usually elevated in GBS; the cell count should be normal. In most instances recovery is

Peripheral Neuropathies in Infants and Children

spontaneous and complete, although one infant subsequently developed CIDP.370,444 There is a single report of a rapid recovery after intravenous immunoglobulin (IVIg) therapy in a 2-week-old infant with GBS.402 Childhood GBS is discussed in more detail later in this chapter. Differential Diagnosis of Infantile GBS Hereditary Tyrosinemia. This is a very rare cause of an acute, sometimes recurrent infantile polyneuropathy.170,325 Affected children experience “crises” typified by a polyneuropathy, weakness, pain, axial and extensor hypertonia (sometimes even opisthotonic posturing), renal hypertension, liver dysfunctions, and fever.325 Tyrosinemia is caused by a defect of fumarylacetate hydrolase.170,325 Treatment includes restriction of dietary tyrosine as well as phenylalanine and therapy with oral hematin and 2-nitro-4trifluoromethylbenzoyl-1,3-cyclohexanedione (NTBC).170,393 Liver transplantation is occasionally required. Spinal Muscular Atrophy. SMA is the most common neuromuscular cause of the floppy infant syndrome.85,96,363,398,474 SMA is usually recognized at 4 to 12 weeks of age in babies presenting with marked generalized hypotonia, decreased spontaneous limb movement, hyporeflexia, tongue fasciculations, and paradoxical respirations.519 On occasion, one evaluates a “previously healthy” baby in the neonatal intensive care unit because of persisting ventilator dependence after an episode of bronchiolitis or other respiratory illness. When such infants cannot be weaned from a respirator secondary to a global weakness, the possibility of GBS or SMA needs careful consideration.240 Infantile Botulism. This relatively rare presynaptic neuromuscular transmission defect may also mimic infantile GBS.85,238 Infantile botulism always deserves consideration in the differential diagnosis of GBS with acute hypotonia and/or respiratory distress. Typically this occurs in infants primarily between the ages of 10 days and 6 months. Feeding difficulties, constipation, weak cry, and loss of head control are typical. Respiratory crisis may suddenly ensue, rarely as early as 5 hours after onset of symptoms. The diagnosis is best made acutely by EMG demonstration of a facilitation of 23% to 313% (mean 73%) with repetitive motor nerve stimulation.80 Subsequent stool culture for the specific Clostridium botulinum organism will confirm the precise diagnosis. Maternally Transmitted Transient Myasthenia Gravis. This should be expected before birth and easily recognized postpartum. This postsynaptic disorder occurs in about 15% of infants delivered to mothers with autoimmune myasthenia gravis. Some of the pre- and postsynaptic congenital myasthenic syndromes also deserve consideration in the differential diagnosis of GBS.

2713

Infantile Poliomyelitis. Fortunately this is now a very uncommon disorder in the industrialized countries. As recently as the last decade, however, this was an important consideration in the rare instance of a previously well infant who developed an acute, often asymmetrical or monomelic weakness, with hypotonia, soon after initial immunization with the Sabin oral (live) type 3 polio vaccine.30,94 An important clue to diagnosis is the significant pleocytosis (100 to 500 white blood cells) seen in association with a modest elevation of the CSF protein.30,94 New guidelines mandate use of the Salk (killed) vaccine for the initial poliomyelitis immunization in the United States. Although one of the authors (H.R.J.) was personally consulted in reference to three instances of vaccine-related poliomyelitis in this country in the mid-1990s, he has had no knowledge of similar cases since the new immunization guidelines were introduced. Unimmunized or immunocompromised children remain at risk for poliomyelitis infection from exposure to unsuspected contacts. For instance, in the less privileged countries the Sabin vaccine remains the primary vehicle for poliomyelitis prophylaxis. Thus such infants are at risk for vaccine-related poliomyelitis. There are rare reports of an adult-onset poliomyelitis clinical picture secondary to the West Nile virus. Thus we need to be vigilant to this occurring in childhood although as of Sept. 2004, we are unaware of such. Myopathies. Occasionally, we are asked to see a newborn baby in the neonatal intensive care unit with profound hypotonia, sometimes with arthrogryposis, who is found to have a primary myopathy.238,462 It is not clear that there are any primary myopathies that have presented with acute hypotonia mimicking GBS in a previously healthy infant.

Infantile Chronic Polyneuropathies Acquired CIDP in infancy is exceedingly rare.370,444 This disorder occurs primarily in the setting of a baby who develops infantile GBS who then later experiences exacerbations leading to a chronic disorder.442,444 The vast majority of chronic demyelinating neuropathies of infancy are inherited. Inherited Hereditary (Nonmetabolic) Motor and Sensory Neuropathies. The hereditary motor and sensory neuropathies of the classic CMT phenotype are discussed in detail in Chapters 71, 73, 75, and 76.308,357 DÉJÉRINE-SOTTAS DISEASE AND CONGENITAL HYPOMYELINATING NEUROPATHY. Historically, the classic inherited demyelinating and hypomyelinating polyneuropathies of infancy have been referred to as DéjérineSottas disease and congenital hypomyelinating neuropathy

2714

Diseases of the Peripheral Nervous System

(CHN).252,257,367,380,473 Affected infants have a severe nonprogressive weakness, hypotonia, areflexia, and respiratory failure.160 EMG is very helpful for recognizing these babies among the group of floppy infants typically evaluated in a child neurology clinic.96 Typical parameters found in these CHN and DSD babies include extremely slow motor nerve conduction velocities (e.g., 3 to 8 m/s) with low-amplitude CMAPs and, most importantly, unelicitable SNAPs. Neurogenic changes are seen on the needle examination. The CSF protein level is significantly elevated in some instances. Sural nerve biopsy demonstrates a severe hypomyelinating neuropathy, with or without onion bulb formation. DSD and CHN were classically believed to be inherited in an autosomal recessive fashion. However, it is now recognized that many cases are due to de novo dominant point mutations of MPZ, EGR2, or PMP22.48,98,99,367,380,495 Occasionally some babies with CHN have associated Waardenburg-Hirschsprung disease. These cases are associated with dominant mutations of SOX10, a transcription factor in neural crest development.223,379 Despite these recent findings, DNA analysis is often unrevealing in a significant proportion of these infants.380 AUTOSOMAL RECESSIVE CHARCOT-MARIE-TOOTH DISEASE (ARCMT OR CMT4). This form of inherited polyneuropathy often presents during infancy. There is a clinical as well as electrodiagnostic overlap between CMT4 and DSD, CHN, CMT1, and CMT2, complicating classification.380 An updated listing of mutations for CMT can be found at molgen-ww.uia.ac.be/CMTMutations. ARCMT is discussed in more detail in Chapter 75. At the time of this writing, mutations for demyelinating ARCMT have been identified in at least five genes (GDAP1, MTMR2, NDRG1, PRX, and EGR2), and additional loci are also identified on chromosomes 5q and 11p15. Mutation of an LMNA gene also causes one form of axonal ARCMT.101 Other congenital axonal polyneuropathies are extremely rare in infancy.160 A number of sporadic and familial cases are reported. Autosomal recessive inheritance is suspected in many cases.326,487,506 Those children with associated diaphragmatic weakness may fit within the clinical spectrum of spinal muscular atrophy with respiratory distress (SMARD, or distal hereditary motor neuronopathy type VI).18,326,506 At least a minority of such cases is caused by mutations in the gene for immunoglobulin μ-binding protein-2 (IGHMBP2).189 These babies must be distinguished from those with infantile SMA.267 CHARCOT-MARIE-TOOTH DISEASE TYPE 2C. CMT2 is discussed in detail in Chapter 73.319 CMT2C may also present in infancy with stridor and intermittent breathing difficulties.122 Weakness of laryngeal muscles is responsible for the respiratory stridor in infants, and a hoarse voice may be a manifestation of vocal cord paralysis in infants, older

children, and adolescents. CMT2C patients have respiratory muscle weakness caused by intercostal muscle and diaphragmatic involvement. Sensory loss is usually mild, but when this is detectable clinically and/or electrodiagnostically, it helps distinguish CMT2C from distal SMA with vocal cord paralysis (distal SMA type VII, or distal hereditary motor neuronopathy type VII) and SMARD. The gene for CMT2C maps to chromosome 12q23-24,260 and distal SMA type VII links to a locus on chromosome 2q14.312

Infantile Hereditary Metabolic Polyneuropathies Infants with a number of inborn errors of metabolism may either present with or develop a significant polyneuropathy during the course of the evolution of these very rare disorders. They are discussed in more detail in Chapters 79, 80, and 84. Lysosomal Storage Disorders: Sphingolipidoses (Leukodystrophies) (see Chapter 79) GLD and MLD are rare autosomal recessive disorders that present in infancy (GLD; first 6 months) or early childhood (MLD; age 2 to 4 years) with features of progressive central and peripheral demyelination.60 GLD and MLD have devastating consequences without treatment, and thus prompt diagnosis is important in order to enable initiation of appropriate therapy. Both GLD and MLD most often present with CNS manifestations that usually serve to alert the clinician to a more systemic disorder rather than one confined to the PNS. However, on occasion these disorders initially may have a primary peripheral manifestation, sometimes mimicking CMT, and are rarely clinically prominent during infancy. These disorders are primarily demyelinating on electrodiagnostic testing.60,115,307 Globoid Cell Leukodystrophy. GLD is caused by a deficiency of the enzyme galactocerebrosidase and most often presents as its “classic” form, Krabbe’s disease, between 3 and 8 months of age with motor retardation, significant irritability, seizures, and hypertonicity.196 The GLD-associated demyelinating polyneuropathy is usually overshadowed by other clinical features, although occasional cases with acute-onset or clinically prominent polyneuropathy are reported in infancy and childhood.196,268,307 Bone marrow transplantation may be beneficial when performed early in the disease course, but sadly may not reverse the neurologic manifestations.271,272 Hematopoietic stem cell transplantation, performed within weeks of birth in presymptomatic patients, appears promising.271 Metachromatic Leukodystrophy. MLD is described in the subsequent section on later childhood metabolic polyneuropathies, and described in detail in Chapter 79. Rarely, the clinical onset is during late infancy, between

Peripheral Neuropathies in Infants and Children

18 and 24 months, when it presents with either ataxia or a primary polyneuropathy. In later childhood it is characterized by loss of motor milestones, with emerging signs of CNS deterioration. MLD is caused by one of three inborn errors of metabolism: (1) deficiency of arylsulfatase A, (2) multiple sulfatase deficiency, or (3) deficiency of the sphingolipid activator protein saposin B. All forms of MLD are transmitted as autosomal recessive traits. Lipid Disorders Of these various inborn errors of fat metabolism, abetalipoproteinemia and hypobetalipoproteinemia are the primary lipid disorders to present during infancy. Abetalipoproteinemia is an autosomal recessive disorder that leads to failure of chylomicron formation, and impaired absorption of dietary lipids and fat-soluble vitamins. Approximately one third of affected children have neurologic problems during the first year of life, beginning with a peripheral neuropathy that usually predates visual loss and sometimes an ataxia. Affected patients are treated with vitamin E, A, and K supplementation.60,335 Familial hypobetalipoproteinemia is genetically separate but clinically indistinguishable from abetalipoproteinemia.515 Mitochondrial Disorders (see Chapter 84) Leigh Disease (Subacute Necrotizing Encephalomyelopathy). Typically, subacute necrotizing encephalomyelopathy (SNE) begins in either early infancy or childhood. During the early months of life, SNE typically presents with either lethargy or coma. Less focused symptoms include hypotonia and feeding and swallowing difficulties. SNE is characterized by multisystem involvement with variable neurologic manifestations, often including a peripheral neuropathy, ataxia and intention tremor, involuntary movements, external ophthalmoplegia, loss of vision, impaired hearing, vascular-type headaches, seizures, and respiratory irregularities. Autonomic system involvement is characterized by cardiac instability and hypertension.376,377 A lactic acidosis is frequently detectable during acute events, although an elevated lactate may only be found in CSF. Nerve conduction velocities may reveal a primarily demyelinating disorder, making SNE another disorder to consider in the child with combined multifocal neurologic findings involving both the CNS and PNS and very slow nerve conduction velocities. Cranial magnetic resonance imaging (MRI) typically demonstrates abnormalities in the basal ganglia, particularly the putamen, thalamus, and brainstem. The underlying biochemical defects include various energy metabolism abnormalities as outlined in Chapter 84 and by Moser and Percy.333 In many instances, a specific enzymatic defect is not identified. The clinical presentation relates in part to the percentage of mutant DNA found in the affected child. Early-onset Leigh disease occurs when this is at high levels (⬎95%).

2715

However, if the percentage of mutant mitochondrial DNA (mtDNA) is less than 75%, the clinical phenotype may be that of neuropathy, ataxia, and retinitis pigmentosa (NARP syndrome) (discussed later). There is still no effective therapy available. Home ventilators, and other forms of supportive care, are important. The EMG findings are well defined.145 Mitochondrial Trifunctional Protein and Long-Chain 3-Hydroxyacyl Coenzyme A Dehydrogenase Deficiencies. These two deficiency syndromes of mitochondrial trifunctional protein (MTP) and long-chain 3-hydroxyacyl coenzyme A dehydrogenase (LCHAD) occur in early infancy. Progressive sensorimotor neuropathy, with a pigmentary retinopathy and cardiomyopathy, are typical of these disorders. Peripheral nerve histology reveals both axonal loss and demyelination with predominant changes in sensory fibers. In addition to the peripheral neuropathy, they are characterized by lethargy and coma, hypotonia, hepatomegaly with lipid storage, hypoglycemia, hypoketonemia, and hyperammonemia with acidosis. It is typical to have 3-hydroxydicarboxylic aciduria. Diagnosis can be confirmed by measuring LCHAD in cultured skin fibroblasts. LCHAD is part of the MTP complex, the third enzyme in very-long-chain fatty acid ␤-oxidation. It is possible that some children with LCHAD deficiency actually have MTP deficiency. Treatment requires frequent feeding, avoidance of fasting, and alteration of fat intake by substitution of medium-chain for long-chain fatty acids. Docosahexanoic acid replacement appears to improve the sensorimotor neuropathy. Nevertheless, death in infancy is common. Amino Acid Disorders (see Chapter 80) Tyrosinemia is the prime example of these rare disorders that in many ways mimic intermittent porphyria, although tyrosinemia typically presents in infancy and intermittent porphyria presents later. Tyrosinemia was discussed earlier in this section vis-à-vis the differential diagnosis of GBS.

POLYNEUROPATHIES OF LATER CHILDHOOD Acquired Acute Childhood Polyneuropathies Guillain-Barré Syndrome (GBS) This is the most common cause of acute flaccid paralysis in childhood.52,139,143,200,237,443 It is discussed in detail in Chapter 98. By definition, GBS has a rapid course, reaching a maximal deficit within 4 weeks.52 Eighty percent of children reach their clinical nadir within 2 weeks of onset.52 The primary features of GBS are a rapidly progressive limb weakness with lost or diminished muscle stretch reflexes. Affected children often initially present with difficulty climbing stairs, squatting, and rising. Pain is often prominent

2716

Diseases of the Peripheral Nervous System

in childhood GBS,51 and may confound the initial diagnosis, particularly in younger children and especially in toddlers. Typically there has been a preceding upper respiratory or gastrointestinal infection. The relation to prior immunizations is not well defined in children.93,477 We have not seen GBS in children present with AIDS that was not previously recognized.394 At CHB, muscle pain was the presenting symptom in 20% of children with GBS. In some instances this was associated with marked irritability, vomiting, and headache. This is in effect a “pseudo-meningoencephalopathic” clinical picture characterized by signs of meningeal irritation, including stiff neck and Kernig’s or Brudzinski’s signs.51 Childhood GBS may also present with an acute sensory ataxia, especially in toddlers.171,233 Cranial nerve palsies, usually of the facial or oculomotor nerves, are relatively common.52 An internal ophthalmoplegia, with nonreactive pupillary responses, is rare in GBS; such a finding should prompt consideration of other potential diagnoses, including diphtheria, tick paralysis, and botulism.26 Dysautonomia, as well as respiratory insufficiency,283 are uncommon presenting signs of pediatric GBS. One 6-year-old, recently treated for a brainstem tumor, developed acute respiratory failure followed by a rapid evolution of complete flaccid paralysis and loss of all brainstem reflexes totally mimicking brain death.26 A normal electroencephalogram led to the EMG that made the diagnosis. CSF analysis in children with GBS usually demonstrates the classic albuminocytologic dissociation. The presence of more than 50 leukocytes per milliliter is very atypical of pediatric GBS and should prompt consideration of poliomyelitis,30,94 CNS lymphoma,192 human immunodeficiency virus (HIV), and Lyme disease.384 EMG findings are similar to those in adult GBS, although criteria associated with a poor outcome in adult studies of GBS, namely lowamplitude CMAPs and fibrillation potentials, are common in childhood and do not predict persistent deficits.52 MRI often demonstrates hypertrophic and gadoliniumenhancing cauda equina and lumbar roots,84,87,367 a nonspecific finding also present with CIDP. More subtle nerve root enhancement is reported in severe CMT1. A child treated with vincristine who has unsuspected CMT may present with an acute neuropathy mimicking GBS. Thus enhancement per se does not exclude an inherited demyelinating polyneuropathy, 368 CIDP, sarcoidosis, lymphoma, and leptomeningeal carcinoma. An MRI of the thoracic as well as the cervical spinal cord may be required to exclude a myelopathic lesion such as a spinal cord tumor in those children whose symptoms include severe back pain, or whose findings demonstrate retained muscle stretch reflexes, possible Babinski’s signs, suggestion of a cord lesion, or persistent sphincter dysfunction for more than 24 hours. At time of presentation, every child with suspected GBS must be admitted to the hospital for monitoring as well as treatment. The importance of zealous cardiovascular and

autonomic monitoring in all cases of GBS must be emphasized.52,233,404 The specific immunomodulatory treatment of pediatric GBS depends on the degree of neurologic compromise. Children with mild GBS, who maintain the ability to ambulate unassisted, usually do not require specific treatment. In contrast, those children who are either unable to walk unassisted or have early respiratory compromise need to be considered for therapy with either IVIg or plasma exchange (PE). Although there have been no controlled randomized trials of these therapies in childhood GBS, a number of retrospective studies have suggested that treatment with PE or IVIg hastens the recovery of a child with GBS to independent ambulation.139,188,228,241,258,281,347,512 IVIg offers the most practical therapy for childhood GBS because of its ease of administration.482 Clinical responses to IVIg may be dramatic.2,12 A total dose of 2 g/kg of IVIg is given over 2 to 5 days and is generally well tolerated at all ages.233 This therapy has been administered to infants as young as 2 weeks.402 Some pediatric studies demonstrated that IVIg was effective sooner than with PE; in one, a compressed dose schedule was 1.0 g/kg on two different days.481 Another study reported the interesting finding that the total dose of 2.0 g/kg could be safely administered to children in just one dose.520 Plasma exchange is also a safe and effective treatment of GBS for children as young as 18 months of age, and is usually performed four to six times on an alternate-day schedule.248,258,347 Comparisons of PE with IVIg have been limited because of differences between treatment groups, including severity of GBS and time to treatment. Because of its ease of administration in younger children, IVIg is now the generally preferred therapy. When the GBS proves to be refractory to IVIg, PE is indicated. Pediatric GBS has a shorter clinical course and more complete recovery than is typical of the adult forms.52 Most children have minimal residual impairment by 1 to 4 months from onset.52,139,228,258 In contrast to the adult patient, EMG markers of severe axonal damage do not invariably predict poor outcome for children with GBS52; however, their initial course may be more turbulent. A 1994–1999 evaluation of 23 children included 10 with predominant demyelinating GBS and 8 with the primary axonal form. Although 38% of the children with the axonal form required intubation, versus 10% of those with the primary demyelinating form, all had an equal recovery at 6 months, and each child was fully recovered at 12 months no matter what the primary physiology of the GBS. A small percentage of infants and children presenting with acute GBS later develop chronic inflammatory polyneuropathy.370,388,437 Currently a fatal outcome is very uncommon in childhood GBS; it is probably no more than 1% to 2%.233 Mimickers of Childhood GBS Acute Myelopathies. Immediate diagnostic considerations in a child with an acute flaccid paralysis and presumed GBS

Peripheral Neuropathies in Infants and Children

include spinal cord tumors and transverse myelitis.262 Both of these disease entities may produce a rapidly progressive paralysis, hyporeflexia, and back pain. Sphincter dysfunction is common with spinal cord lesions, but unusual and typically transient in GBS.233 Transverse myelitis may be associated with marked elevation of the CSF protein and with significant CSF pleocytosis.262 On occasion we have seen a child with clinical history initially suggesting GBS but in whom the presence of Babinski’s signs has suggested a cord lesion, yet no other signs of myelopathy ever develop. One extensive review of 206 children seen at 70 children’s hospitals demonstrated that 198 had either GBS or CIDP, whereas only 8 children, ages 2 to 13, had simultaneous involvement of both nerve roots and spinal cord. Each of these children had classic findings of both GBS and transverse myelitis. As might be expected, the prognosis was significantly worse in this subcategory. Follow-up of seven of the eight children noted that two were unable to walk while the remaining five had a residual paraparesis requiring orthoses for ambulation. Tick Paralysis. This uncommon diagnosis must always be considered in all children developing an acute flaccid paralysis. The skin and particularly the scalp of all children suspected of having GBS must be carefully inspected to be certain that there is no attached tick. This is particularly common in girls, whose longer hairstyles frequently hide a tick. Its removal will lead to early and dramatic improvement in North American children but a delayed response in children with the Australian variant.187 An important differential diagnostic point is that pupillary involvement is common in tick paralysis but only rarely observed in GBS.187 EMG findings include very diminished CMAP amplitudes but preservation of motor conduction velocities, motor distal latencies, and SNAPs.187,460 Toxins. Vincristine, heavy metals, glue and gasoline sniffing, buckthorn wild cherry, and fish toxins, such as those of shellfish and ciguatera, warrant consideration in the differential diagnosis of GBS.61,380a Typically vincristine toxicity is associated with the gradual evolution of a primary sensory polyneuropathy.380a However, children who have overt or covert CMT1 and concomitantly later require vincristine may present with an acute motor and sensory neuropathy not unlike GBS. Children living in agricultural communities are at risk for buckthorn, arsenic, and organophosphate pesticide (OPP) poisoning. With the exception of buckthorn in northern Mexico and the southwestern United States, most of these are quite uncommon in North America but still deserve consideration. In contrast, in Paraguay there was a 30% incidence of OPP exposure identified in children with what appeared to be an otherwise typical clinical presentation of childhood GBS.208

2717

Inborn Errors of Metabolism. A very few of these uncommon disorders, such as porphyria and Leigh disease, enhanced by medications, particularly barbiturates, may also precipitate symptoms resembling childhood GBS, as described in greater detail later in this chapter.77 Hypokalemic paralysis, presenting as a manifestation of proximal renal tubular acidosis, mimicked GBS in a 14-year-old girl who presented with paraplegia and areflexia. Sepsis and Asthma. Children with overwhelming sepsis or status asthmaticus may develop a critical illness neuromuscular syndrome, either a myopathy or a polyneuropathy, mimicking GBS.49,240 Polyneuropathy associated with critical illness is discussed in Chapter 87. Infectious Diseases. As with GBS, Lyme disease may also be associated with a painful neuropathy. Lyme disease has a similar predilection for the cranial nerves but is occasionally complicated in childhood by an acute inflammatory demyelinating polyradiculoneuropathy.35,505 Additionally, at CHB we evaluated an adolescent presenting with rapidly evolving cranial neuropathies and ataxia mimicking the Miller Fisher variant of GBS (M. Rivkin et al., personal communication, 2003). He had a rapidly migratory erythematosus rash with bull’s-eye centers, normal protein but 35 lymphocytes in his CSF, and very positive acute Lyme antibody results. Although rarely a GBS picture is the presenting manifestation of acquired immunodeficiency syndrome (AIDS) in adults, we are unaware of this occurring in children. In contrast, there are a few reports of children with congenital AIDS who later developed GBS.15,384 Acute Sensory Neuropathy or Neuronopathy of Childhood This self-limited acute, severe, hyperpathic symmetrical polyneuropathy is associated with variable autonomic symptoms that occasionally occur during childhood. EMG demonstrates that sensory NCS and blink responses are lost while standard motor as well as late responses are normal.149,342 Both myelinated and unmyelinated fibers are affected when a sural nerve biopsy is performed. Recovery may be prolonged.149 Diphtheria This uncommon polyneuropathy should be considered in any child with a recent pharyngitis, fever, and subsequent acute peripheral neuropathy, with features similar to GBS, particularly when associated with a prominent associated bulbar palsy. A higher level of suspicion exists when the child comes from an environment where it is possible that he or she did not receive appropriate immunization. One of the authors (H.R.J.) has seen eight cases of diphtheria in his career, however, seven occurred at Philadelphia’s General

2718

Diseases of the Peripheral Nervous System

Hospital in 1963. In each instance the child’s parents had inadvertently thought appropriate immunization had been given. However, on review of the specific immunization records, none had received the full complement of diphtheria, tetanus, and pertussis prophylaxis.

Acquired Chronic Childhood Polyneuropathies Chronic Inflammatory Demyelinating Polyradiculoneuropathy Epidemiology. CIDP is also discussed in Chapter 99. Childhood CIDP is an important entity, responsible for up to 10% of cases of childhood neuropathy,355,442 although it is significantly less common than GBS. Approximately 10% of patients with CIDP experience onset in childhood.119,311 In most children CIDP presents subacutely, although the onset of childhood CIDP may be more rapid than that of adult CIDP.385,388,416,437 Antecedent infections or vaccinations occur less frequently with CIDP than in GBS.266,343,416,437 In most cases the parents and child can suggest a date of onset; this contrasts with the inherited neuropathies, wherein it is difficult to ascertain a precise onset. CIDP very rarely presents during early infancy370,444 or during ages 1 to 3 years.343,437 However, some investigators found a higher incidence during the first decade.343,416 Clinical Presentation. Gait difficulties are the most common childhood presentation for CIDP.78,209,266,343,388,416,437 Subsequently, frequent falls, abnormal or clumsy gait, inability to run, or difficulty arising from the floor or a seat lead to discovery of the associated weakness, which sometimes mimics a myopathy. Less commonly, the upper extremities are equally or more affected.78,209,343,416,486 Functional disability is typically quite significant.416,437 Paresthesias and distal sensory loss are often difficult to identify in children.78,266,343,416,437 However, their definition clinically or with EMG is important to the differentiation of this polyneuropathy or polyradiculoneuropathy from various primary motor unit disorders such as SMA type III, a myopathy, or a rare neuromuscular junction disorder. Rarely the cranial nerves are affected.110 Reduced or absent muscle stretch reflexes are typical, although retention of these reflexes is occasionally seen in some pediatric CIDP patients.416 The cranial nerves are occasionally involved in childhood CIDP; facial paresis occurs in as many as 25% of children.343,416 In contrast, diplopia, dysarthria, and dysphagia are uncommon,266,343,416,437 as are tremor and ataxia.416 Significant involvement of the respiratory muscles is rare.438 A relapsing-remitting course is common in childhood CIDP.266,388,416,437 This was seen at CHB in 10 of our 11 children.388 The mean duration from symptom onset to presentation was 3 to 5 months at CHB and in an Australian series.416

Very unusually, a child with CIPD has a concomitant CNS disease affecting both spinal and brain white matter. One 5-year-old child presented with a 3-month history of progressive gait difficulty and frequent falls. Her examination demonstrated increased tone, hyperreflexia, and Babinski’s signs as well as pes cavus. MRI demonstrated two separate expansile intramedullary spinal cord lesions as well as cauda equina enhancement. Biopsy was unremarkable; corticosteroids led to almost complete improvement. She had a recurrence 3 years later. This time she had both diffuse weakness and absent muscle stretch reflexes, as well as Babinski’s signs. Spinal MRI results were similar to those obtained previously; however, brain MRI also demonstrated patch white matter lesions. Prednisone was helpful once again. Laboratory Findings CEREBROSPINAL FLUID. Elevation of the CSF protein (⬎35 mg/dL) without pleocytosis (⬍10 cells/mm3) is typical in childhood CIDP.209,266,343,416 More than 10 white cells per cubic millimeter suggests either Lyme polyradiculoneuropathy or, very rarely, an HIVassociated disorder. However, in contrast to adults, we are unaware of HIV in children presenting as an acquired inflammatory demyelinating polyneuropathy. ELECTROMYOGRAPHY. Some of the specific EMG clinical research criteria recognized for the diagnosis of CIDP in adults apply to children, and are detailed in Chapter 99.217,323,344,502 These include nonuniform nerve conduction slowing, conduction block at sites not prone to entrapment, and abnormal temporal dispersion.289,324 However, conduction block is not as commonly demonstrated in CIDP in the pediatric population, possibly because pediatric electromyographers tend to limit testing in young children.59,343,416 Thus some children with clinical CIDP may not fulfill specific EMG criteria. Progressive worsening of nerve conduction velocities, distal latencies, and F-wave latencies are often observed in children with clinical deterioration. Paradoxically, clinical improvement may occur even with static or worsening neurophysiologic abnormalities.416 MAGNETIC RESONANCE IMAGING. Spinal MRI is an adjunct means to diagnose CIDP.86 Irregular thickening and nerve root gadolinium enhancement seen with CIDP contrast with the relatively homogeneous changes of inherited neuropathies.321,416 However, subtle nerve root enhancement is also seen in some instances of severe inherited demyelinating polyneuropathies.368 Differential Diagnosis. Generally CIDP is a straightforward diagnosis to establish, especially when there has been a preceding instance of GBS or there are significant clinical sensory findings in the presence of hypoactive muscle stretch reflexes. SMA type III (Kugelberg-Welander syndrome) also presents with a gradual onset of symmetrical limb weakness, often having a proximal predominance, and

Peripheral Neuropathies in Infants and Children

diminished muscle stretch reflexes, thus entering the CIDP differential diagnosis. DNA testing for the specific gene defect, the survival motor neuron gene (SMN) at chromosome 12.2-q13.3, is the best means to diagnose SMA type III; the EMG can be very helpful early on in the evaluation. Similarly, the predominance of proximal weakness may also suggest a myopathy such as dermatomyositis. Usually there is an elevation of creatine kinase, although this has been relatively minimal in children in our experience. However, when the creatine phosphokinase is normal in a child with acquired proximal weakness, a myopathy is much less likely. In this setting one needs to consider the very rare possibility of the Lambert-Eaton syndrome; there are two single reports in children and adolescents. Rarely CIDP may mimic lumbar spinal stenosis.178 Treatment INTRAVENOUS IMMUNOGLOBULIN. IVIg is the most common first-line treatment for childhood CIDP. This is because of its superior side effect profile, efficacy, and ease of administration. IVIg is generally effective in pediatric CIDP in children naïve to all forms of therapy,266,343,388,416,437 as well as cases refractory to treatment with prednisone79,416 and other immunosuppressive therapies.486 A small number of children will have a sustained response to only a single course of IVIg.266 However, occasionally children do not respond to IVIg but have a favorable response to corticosteroids.438 A typical pediatric CIDP regimen for IVIg treatment consists of an initial dose of 2.0 g/kg over 2 to 5 days, followed by monthly maintenance doses of 0.4 to 2.0 g/kg. Pretesting for a significant but rare immunoglobulin A deficiency is important because children with such a deficiency are predisposed to a serious allergic reaction to IVIg. Once a sustained response occurs, the IVIg dosage and dose frequency are gradually tapered and ultimately withdrawn. Drawbacks to IVIg therapy are similar to those noted in adults. ORAL CORTICOSTEROIDS, IVIG, AND PLASMA EXCHANGE. These are the mainstays of therapy for childhood CIDP. They are generally indicated only for significant, persistent disability, particularly when this limits the patient’s ability to walk. If these therapeutic interventions will eventually be successful, clinical improvement usually becomes evident within 2 to 3 weeks. Prednisone was the first accepted therapy for childhood CIDP, and it remains an effective, cheap, and convenient treatment for this condition; however, its long-term side effects make it a second choice to an initial trial with IVIg. Although no prospective randomized trials of corticosteroid therapy have been carried out for childhood CIDP, the efficacy of prednisone has been demonstrated in a number of retrospective reviews.78,103,209,266,343,370,416,438 Initial treatment with oral prednisone is usually at a dose of 1 to 1.5 mg/kg daily for 2 to 4 weeks, with subsequent slow tapers over a

2719

period of months once a therapeutic response has been obtained. Long-term maintenance therapy is often required, sometimes in conjunction with chronic IVIg therapy. In childhood CIDP, relapses are common when one attempts tapering corticosteroid therapy; these may ultimately necessitate reinstitution of the “steroid-sparing” IVIg therapy.343,416,438 PLASMA EXCHANGE. PE is used less frequently in the treatment of childhood CIDP, primarily because of technical difficulties with venous access and the exchange of large volumes in small children. When other therapies are unsuccessful or inconsistent, PE is a safe and effective therapy for some children.266,343 This effect may be transient,266,416 and prolonged forms of maintenance treatment may be required.40,343,437,438 OTHER THERAPEUTIC OPTIONS. Other treatments include azathioprine, 266,343,416 methotrexate, 343,416 cyclosporine, and cyclophosphamide.343,416 Experience with these agents is limited. These are second-line therapies for children who fail adequate trials of IVIg, corticosteroids, and/or PE. Prognosis. The long-term outcome is generally quite good in childhood CIDP. Most children followed for long periods eventually achieve sustained clinical remission with no or minimal residual deficits.78,266,388,416,438 Occasionally children do have significant persistent weakness.343 Childhood Chronic Polyneuropathy Associated with Systemic Disease Connective Tissue Disorders (see Chapters 105 and 106). Polyneuropathy associated with connective tissue diseases is very rare in children. Systemic lupus erythematosus occasionally occurs with a distal symmetrical polyneuropathy143,180,206,369,425,451 or childhood CIDP,25,181,311 and, rarely, with MNM (Fig. 122–3).418 One adolescent with chronic juvenile rheumatoid arthritis had a progressive polyneuropathy,143 while another study noted a tendency to asymptomatic focal NCS slowing.387 Rarely, microscopic polyangiitis,373 polyarteritis nodosa,109,159,418 Wegener’s granulomatosis,352,410 the Churg-Strauss syndrome,193 and Henoch-Schönlein purpura36 are associated with a pediatric polyneuropathy. Infectious Polyneuropathies (see Chapters 91, 92, 95, and 107). Leprosy is the most common infectious cause of neuropathy worldwide and, although symptomatic pediatric cases are relatively uncommon,419,446 they are seen in endemic areas. Very rarely, neurosarcoidosis causes facial palsies, hearing loss, and polyneuropathy in childhood despite its peak incidence in the third and fourth decade.159,286,314 Lyme disease has a similar predilection for the cranial nerves but is occasionally complicated in childhood by an acute inflammatory demyelinating

2720

Diseases of the Peripheral Nervous System

Other causes of toxic neuropathy, including alcohol, heavy metals, and exposure to organophosphorous esters, are exceedingly rare in children.359 Mercury,409,459 thallium, and lead poisoning in children usually causes an encephalopathy rather than a polyneuropathy; however, on rare occasions, it is important to consider these heavy metals in the differential diagnosis of a pediatric peripheral neuropathy.

FIGURE 122–3 Sural nerve biopsy in an adolescent with systemic lupus erythematosus and mononeuritis multiplex. See Color Plate

polyneuropathy.35,505 Diphtheria has resurged within eastern and central Europe in the past decade.293 Typically children develop a severe pharyngitis, fever, bulbar palsy, myocarditis, and some weeks later an acute peripheral neuropathy similar to GBS. The bulbar palsy is often the primary symptom bringing the patient to a neurologist’s attention. In the experience of one of us (H.R.J.) with a few pediatric diphtheria cases presenting with bulbar symptoms in the United States, diagnosis became clear when it was found that these children had not received adequate immunizations. Metabolic Polyneuropathies (see Chapters 85 and 87). Rarely, juvenile diabetics develop a symptomatic polyneuropathy even though screening EMG studies identify subclinical slowing of motor nerve conduction in as many as 72% of patients.112,124,163,168,186,215,285,500,514 In the (rare) patient with significant symptoms, neuropathy may be reversed by improved glycemic control.500 There is a single brief report relating diabetic ketoacidosis to MNM in a child21; however, the authors’ clinical criteria for MNM diagnosis are very atypical at best. Most children with chronic renal failure have no or minimal clinical neuropathic symptoms; however, uremia is an uncommon cause of a childhood polyneuropathy.8,19 EMG findings correlate with the severity of renal impairment and may be reversible after renal transplantation.19 Toxic Polyneuropathies (see Chapters 112 and 113). On rare occasions, a few toxins are associated with a chronic polyneuropathy in children. A neuropathy associated with glue- or gasoline-sniffing neuropathy is related to inhalation of n-hexane, an aromatic hydrocarbon.61,162,251,277,360 Adolescents sniff glue and gasoline vapors for their euphoric effect.210,259,269 MNM has also been linked to a systemic necrotizing angiitis caused by amphetamine abuse.450

Medication-Related Polyneuropathies. Chemotherapy is the most common cause of drug-related neuropathy in childhood. Common causative agents including the vinca alkaloids, cytosine arabinoside, paclitaxel, thalidomide, and cisplatin. Vincristine frequently causes a relatively benign axonal sensorimotor and autonomic neuropathy. However, very rarely, an acute severe vincristine neuropathy may develop and cause severe weakness in children with undiagnosed CMT.211,218 Thalidomide was withdrawn from the market in 1961 because of teratogenic effects but has been reintroduced for its antiangiogenic and immunomodulatory properties.73 Thalidomide neuropathy is primarily a sensory axonal polyneuropathy reported thus far in adults only. Severe proprioceptive sensory deficit and motor involvement do occur in severe cases.169 We have seen three children, two at CHB and one at Westmeade Children’s Hospital in Sydney, who developed a polyneuropathy probably induced by thalidomide. Electrodiagnostic studies showed a moderately severe sensory axonopathy. The course of thalidomide neuropathy in children is not clear, because few studies have been carried out. Two of our three cases demonstrated progressive polyneuropathies, and the third demonstrated ongoing EMG changes despite cessation of therapy.156 Nutritional Polyneuropathies. Vitamin deficiencies primarily warrant consideration in infants and children with malabsorptive syndromes or as a complication of maternal dietary restriction in breast-fed infants. Nutritional polyneuropathies are discussed in detail in Chapter 89. Vitamin B12 deficiency is a very uncommon cause of pediatric polyneuropathy.274,302 Breast-fed infants are at risk of B12 deficiency where there is a history of maternal pernicious anemia or strict vegetarianism.274 Autoimmune pernicious anemia is extremely rare in childhood. Vitamin B1 (thiamine) deficiency is very rare in Western civilization, although beriberi neuropathy can occur in breast- fed infants of thiamine-deficient mothers. Typical clinical characteristics include dependent edema, neck stiffness, dyspnea, and acute axonal peripheral neuropathy, possibly associated with aphonia resulting from laryngeal nerve involvement.508 Vitamin E deficiency associated with cystic fibrosis, other malabsorptive syndromes, or isolated defects of vitamin E absorption causes posterior column and dorsal root ganglia degeneration resulting in ataxia and impaired joint position and vibration sensation.130,204,408,499 Other signs include gaze paresis, nystagmus, and retinitis pigmentosa.

Peripheral Neuropathies in Infants and Children

An axonal polyneuropathy, possibly related to nutritional deficiencies, has also been described in pediatric celiac disease.439 Paraneoplastic Neuropathies (see Chapter 110). These are extremely uncommon in the pediatric population. Lymphoma has occasionally been associated with neuropathy or neuronopathy in children.275,413 Paraproteinemic polyneuropathy is unknown in childhood in our experience.

Hereditary Chronic Childhood Polyneuropathies Hereditary Motor and Sensory Neuropathies (see Chapters 70, 71, 73, 75, and 76) Charcot-Marie-Tooth disease (CMT) is the most common cause of a childhood polyneuropathy. These are a genetically heterogeneous group of sensorimotor polyneuropathies sharing a similar clinical phenotype: weakness and wasting of the distal limb muscles (predominantly peroneal), pes cavus, absent or decreased muscle stretch reflexes, and distal sensory loss.58,120,121,205,471 Classification is discussed in Chapter 69.44,58,120,121,205 CMT1 usually first manifests clinically during the second decade, whereas CMT2 commonly presents after age 20 years.58,120,121,205,225,455,471 The most frequent initial signs of CMT1 include foot deformity (pes cavus or cavus varus, clawing of toes, atrophy of intrinsic foot muscles), difficulty with heel walking, lower limb areflexia, and sometimes nerve enlargement.14,37,120 At CHB only about half of the children referred with pes cavus have demyelinating findings on EMG.118,245,414 The X-linked form of CMT affects males much more severely than females. In contrast to CMT1, EMGs of CMTX patients may mimic findings seen in CIDP.417 Males with CMTX are symptomatic from the first or second decade of life, whereas females present later on.45,111,201 Hereditary neuropathy with liability to pressure palsy (HNPP) also frequently presents in the second decade of life.71,299 Characteristic NCS abnormalities are sometimes seen even in clinically asymptomatic family members, especially in CMTX, CMT2, and HNPP.198,319,336 When evaluating children with a negative family history, the clinician needs to consider the likelihood of a de novo mutation, as occurs in 25% of children with CMT. Genetic testing may be valuable even for children with no family history of a polyneuropathy. DNA testing is commercially available for CMT1, DSD, and CMTX as outlined in Chapter 64. Such de novo mutations are especially common in DSD and CHN.29,48,367,380 Inborn Errors of Metabolism and Peripheral Neuropathies Pediatric neuropathies are occasionally associated with inborn errors of metabolism, most having an autosomal recessive or X-linked inheritance. Although symptoms of

2721

CNS dysfunction predominate in these children, a concomitant peripheral neuropathy raises concern of a more generalized metabolic disorder.60 Occasionally some children present primarily with a peripheral neuropathy; sometimes these children are thought to have CMT.64,196,197,268,300,307 These neuropathies are variously classified; Moser and Percy provided an easily understood schema.333 Because these entities are detailed in other portions of this text, the subsequent discussion is limited to a review of their childhood clinical as well as neurophysiologic manifestations. Lysosomal Storage Disorders SPHINGOLIPIDOSES (LEUKODYSTROPHIES) Globoid Cell Leukodystrophy (see Chapter 79). This usually presents during the first 6 months of infancy with an extremely irritable, inconsolable baby. Very occasionally the polyneuropathy of GLD predominates while the CNS manifestations are subtle or even unidentifiable when the polyneuropathy is first recognized.60,195 Late-onset GLD presents between 2 and 6 years with dementia, optic atrophy, cortical blindness, and spastic quadriplegia. Both forms may have a demyelinating sensorimotor polyneuropathy with uniform slowing, and well-defined CMAPs. Metachromatic Leukodystrophy (see Chapter 79). MLD usually presents in toddlers with rapid motor regression and concomitant development of spasticity, dysarthria, dementia, and ataxia.172,300 A peripheral neuropathy is usually invariable, although not always a predominant clinical feature.197,300 However, this is not always the case. One 2-year-old child seen at CHB was initially diagnosed as having CMT at age 18 months.64 However, CNS signs usually appear within another 8 to 12 months.429 Nerve conduction velocities are in the demyelinating range307 and typically show uniform slowing of nerve conduction without temporal dispersion or conduction block.290 However, occasionally these children have dispersed CMAPs and asymmetrical findings compatible with CIDP.64 Bone marrow transplantation is sometimes helpful.303 Fabry’s Disease (see Chapter 81). Fabry’s disease is a rare X-linked recessive disorder whose clinical hallmarks include attacks of severe neuropathic pain, angiokeratomas, and progressive renal disease.53,350,433 Onset is typically in childhood or adolescence.359 Angiokeratomas, which are blood-filled papules occurring in the umbilical area, loins, and genitalia, appear before age 10 years. Other complications include cardiac conduction disturbances and stroke. Fabrazyme has been approved for enzyme replacement therapy for Fabry’s disease.132,426

MUCOPOLYSACCHARIDOSES (SEE CHAPTER 79). These disorders have many clinical features in common, including psychomotor retardation, deposition of mucopolysaccharides within arterial smooth muscle and heart valves, hepatosplenomegaly, joint contractures, and varying skeletal anomalies. As to their peripheral neuropathic presentation, these disorders primarily present as mononeuropathies,

2722

Diseases of the Peripheral Nervous System

particularly the carpal tunnel syndrome as discussed later in the section on mononeuropathies. O LIGOSACCHARIDOSES ( SEE C HAPTER 79). These disorders are similar to the mucopolysaccharidoses. Inheritance is in an autosomal fashion. Fucosidosis. Median nerve entrapment may develop secondary to mucopolysaccharide accumulation in the transverse carpal ligament. Mannosidosis. A peripheral neuropathy is seen in the delayed-onset form of ␤-mannosidosis. These children may have recurrent infections and associated psychomotor decline. Sialidosis Type II. This lysosomal disorder of neuraminidase function is characterized by myoclonus and development of nonpigmentary macular degeneration (a “cherry-red spot”). Rarely, the type II form is associated with a demyelinating polyneuropathy.452 Schindler Disease. This is a very rare glycoproteinosis associated with a central and peripheral axonopathy. Recently an acute and chronic pediatric polyneuropathy has been recognized. The early-onset acute form presents in a previously healthy toddler with a rapidly evolving psychomotor deterioration, an exaggerated startle response with myoclonic epilepsy, and subsequently severe spasticity and cortical blindness. Cutaneous angiomas occur mimicking Fabry’s disease. Nerve biopsy demonstrates axonal spheroids resembling neuroaxonal dystrophy.4,392,494 CHÉDIAK-HIGASHI SYNDROME (SEE CHAPTER 79). This autosomal recessive disorder is associated with defective pigmentation and susceptibility to infection and to lymphoreticular malignancies. Affected children may develop a demyelinating neuropathy with conduction block.292 Peroxisomal Disorders. Three peroxisomal disorders are associated with a clinically significant childhood polyneuropathy. REFSUM’S DISEASE (SEE CHAPTER 79). This demyelinating polyneuropathy and associated pigmentary retinal degeneration, characterized by night blindness, is an autosomal recessive disorder that usually presents during childhood. More than 40% of these children become symptomatic by age 10 years, and the remainder by age 20.227,440 X-LINKED ADRENOLEUKODYSTROPHY (SEE CHAPTER 79). Very atypically, the spinal form of X-linked adrenoleukodystrophy (X-ALD), adrenomyeloneuropathy, is identified during early childhood.331,332,334,407 In this instance NCSs are compatible with a combined axonal loss and multifocal demyelinating polyneuropathy.74 About one quarter of these patients partially fulfilled the neurophysiologic criteria of primary demyelination. A polyneuropathy does not occur in childhood cerebral X-ALD.428

HYPEROXALURIA. This disorder usually has its onset before age 5, presenting with nephrolithiasis and obstructive uropathy. A peripheral neuropathy, myopathy, and arthritis are other complications. Nerve involvement may result in painful paresthesias and distal weakness or may be totally subclinical. NCSs are compatible with an axonal pathology.164,328 Definitive diagnosis requires a liver biopsy. Treatment of advanced primary hyperoxaluria usually necessitates combined liver-kidney transplantation. This may substantially alleviate neuropathic symptoms.42,164,202,328 Lipid Disorders CEREBROTENDINOUS XANTHOMATOSIS. Cerebrotendinous xanthomatosis is a very uncommon autosomal recessive lipid storage disorder.54,106,255,337 Clinical features include a demyelinating peripheral neuropathy, tendon xanthomas, progressive cerebellar ataxia, spasticity, and cataracts.491 Onset usually occurs between ages 8 and 18 years.39,46,421,491 TANGIER DISEASE (SEE CHAPTER 82). The peripheral neuropathy of Tangier disease may present during childhood as a relapsing, multifocal MNM, or syringomyelia-like phenotype neuropathy. It primarily affects the upper extremities,378,381 and is typically associated with yellowish lipid laden tonsils. Tangier disease is characterized by diminished serum high-density lipoproteins and a defect in the cellular efflux of lipids to high-density lipoproteins.396,509 A BETALIPOPROTEINEMIA AND H YPOBETALIPOPRO These diseases are discussed earlier in the section on lipid storage diseases in infants. Clinical onset is in the first or second decade.348,431,501 TEINEMIA.

Mitochondrial Disorders (see Chapter 84). Infant-onset mitochondrial disorders are discussed in an earlier section. An ataxic sensory polyneuropathy is a clinically predominant finding in both NARP and Leigh disease.341 There is a clinical continuum between these two diseases related to the amount of mutant mtDNA at nucleotide 8993. The extent of mutant mtDNA determines the phenotypic variability. When mutant mtDNA is less than 75%, the NARP phenotype is seen. When mutant mtDNA exceeds 95%, Leigh disease phenotype results. NARP may be seen in early childhood, although it usually appears in adolescence. A sensory neuropathy, ataxia, retinitis pigmentosa, and sometimes an element of proximal weakness as well as cognitive decline typify the child with NARP. Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE) is an autosomal recessive disorder caused by mutations in the thymidine phosphorylase (TP) gene.345,467 Gastrointestinal dysmotility is the most prominent manifestation. Peripheral neuropathy, ptosis, ophthalmoparesis, and cerebral leukodystrophy are also characteristic. Symptom onset is in the first through fourth decades, with mean age of onset of 18 years of age.

Peripheral Neuropathies in Infants and Children

Amino Acid Disorders (see Chapter 80). Tyrosinemia is primarily discussed in the section on polyneuropathies of infancy earlier in this chapter. The associated polyneuropathy is predominantly axonal, motor, and often painful. This affects proximal more than distal muscles and upper more than lower extremities.158 Facial and ocular palsies may be present.129 Other Inborn Errors of Metabolism HEPATIC PORPHYRIA (SEE CHAPTER 80). Acute intermittent porphyria (AIP) is the only metabolic inherited polyneuropathy having an autosomal dominant genotype. Episodes of the dominantly inherited hepatic porphyrias are of acute onset and may be provoked by infection or medications.158 Clinical features include polyneuropathy, abdominal pain, constipation, vomiting, mental status change, and autonomic disturbances. The recurrent crises of porphyria resemble those of hereditary tyrosinemia. Porphyria occurs rarely during the first decade of life and just occasionally during the second decade.108,128,158 The typical patient has intermittent, primarily visceral pain. When there is a significant peripheral neuropathic component, limb paresthesias, dysesthesias, and weakness typically proximal and greater in the upper than lower extremities are found. Much less commonly the facial, masticatory, swallowing, and respiratory muscles are involved. An autonomic neuropathy that occurs only during exacerbations features hypertension,158 postural hypotension,108,128,158 tachycardia, and gastrointestinal and bladder dysfunction. Typically, the peripheral neuropathy found with AIP has features of both segmental demyelination and axonal degeneration. Generally these tend to occur when the affected individuals have experienced multiple clinical episodes. Charles Bolton (personal communication, 1996) noted a pure axonal motor neuropathy when he performed NCS on a child with porphyria. COCKAYNE’S SYNDROME. This is a very uncommon autosomal recessive disorder of young children secondary to abnormalities of transcription-coupled DNA repair. Clinically it is characterized by dwarfism and progressive neurologic dysfunction.339,362 Although affected children are normally proportioned at birth, they subsequently develop marked postnatal growth failure. Other common findings include sensorineural hearing loss, photosensitivity, optic atrophy, ataxia, and developmental delay.339,362 The associated demyelinating peripheral neuropathy seen in Cockayne’s syndrome is usually not a presenting feature.329,339 No specific treatment is available. XERODERMA PIGMENTOSUM. This rare autosomal recessive disorder is associated with photosensitivity, telangiectasia, predisposition to cutaneous malignancies, and ataxia. NCSs and sural nerve biopsies demonstrate both demyelination and axon loss.250

2723

Hereditary Sensory and Autonomic Neuropathies (see Chapter 78) Hereditary Sensory and Autonomic Neuropathy Type I. Hereditary sensory and autonomic neuropathy type I (HSAN I) presents in children during their first or second decade with impairment of pain and temperature sensation, distal muscle wasting, and weakness.33,97,278,295 Because affected individuals are insensate to pain, repetitive trauma can lead to (painless) ulceration, cellulites, and osteomyelitis.265,470 SNAPs are characteristically absent on NCSs. Motor NCSs may be normal, and needle EMG may reflect changes of distal chronic reinnervation. Hereditary Sensory and Autonomic Neuropathy Type II. This congenital sensory neuropathy also has a childhood onset.220,221,351 Manifestations of HSAN type II (HSAN II) usually begin in early childhood with numbness in the hands and feet, aggravated by cold, and reduced pain sensation. Touch is most severely affected. A congenital subtype of HSAN II with neonatal hypotonia and sensory impairment of the entire body, not just limbs, also occurs.150 Hereditary Sensory and Autonomic Neuropathy Type III (Familial Dysautonomia or Riley-Day Syndrome).17,445 HSN type III is typically seen in infants with Ashkenazi ancestry.3 This rare autosomal recessive illness usually manifests with poor feeding, recurrent pulmonary infections, and autonomic dysfunction typified by defective lacrimation, excessive perspiration, skin blotching, recurrent fevers, and labile blood pressure. Neurologic examination demonstrates absence of pain sensation, lack of fungiform papillae, loss of muscle stretch reflexes, and absent corneal reflexes.3,55 SNAPs are absent. Motor nerve conduction velocity is slightly reduced or normal.3,55 Hereditary Sensory and Autonomic Neuropathy Types IV and V (Congenital Insensitivity to Pain with Anhidrosis). These phenotypically similar autosomal recessive disorders are characterized by loss of pain sensation, impaired or absent sweating, and recurrent fevers, with preservation of muscle strength and muscle stretch reflexes, beginning during early childhood.123,305 A mutilating acropathy, neurogenic arthropathy, and ulcers may develop.123 Sweating is absent in HSAN type IV and reduced but not absent in HSAN type V. Unmyelinated fibers are virtually absent.179 Hereditary Ataxias Giant Axonal Neuropathy (see Chapters 73 and 75). Giant axonal neuropathy (GAN) is a sensory ataxia usually beginning during the early childhood.219,249 Although classically it is associated with tight curly hair, this finding is not witnessed in all of these children.20,38,165,353,386 GAN is a generalized neuropathologic condition typified by

2724

Diseases of the Peripheral Nervous System

accumulations of intermediate filaments in various cell populations.263 The classic nerve biopsy findings consist of giant axonal swellings, with densely packed neurofilaments.20 The initial clinical features appear during infancy or the toddler stage. Typically there is delayed motor development or the already established gait becomes broad based, clumsy, and unsteady. Subsequently an evolving peripheral weakness with atrophy develops, eventually moving from the lower to the upper extremities. A truncal ataxia gradually appears in association with nystagmus, dysarthria, intention tremor, dysmetria, and mild mental retardation. Other cranial nerves may become affected. Muscle stretch reflexes are diminished or absent. Babinski’s signs are often present. A spinocerebellar degeneration eventually predominates after a few years. Epileptic seizures occasionally occur. Death occurs between adolescence and the third decade. The gene responsible for GAN was localized in five consanguineous families to chromosome 16q24.155 The gene GAN encodes for a cytoskeletal protein, gigaxonin, that has a role in neurofilament architecture.50

X-linked forms occur.144 Marinesco-Sjögren syndrome is a recessive disorder characterized by ataxia, mental retardation, and congenital cataracts. The associated neuropathy is primarily demyelinating.9 The ataxia with isolated vitamin E deficiency due to ␣-tocopherol transfer protein (TTP1) gene mutations usually manifests between 4 and 18 years of age. This is potentially treatable, being either arrested or reversed by dietary vitamin E supplementation.204,511 Thus it is very important to obtain a vitamin E level in each child presenting with an ataxia. All six children with vitamin E deficiency evaluated at CHB had neurologic involvement. The malabsorption of vitamin E was primarily related to chronic cholestasis secondary to bile duct hypoplasia. All children had sensory loss, as well as hyporeflexia and in some instances appendicular/truncal ataxia and muscle weakness.374 EMG studies were abnormal in three (50%). Two children had a primary sensory and one a motor sensory polyneuropathy.

PEDIATRIC MONONEUROPATHIES Friedreich’s Ataxia (see Chapter 73). This is the most common spinocerebellar ataxia of childhood.65,82,83,327,356,422 Onset is in the first or second decade, usually between the ages of 5 and 15 years.203 Sometimes there is a later onset.100 Cardinal features of FA include progressive gait and appendicular ataxia, sensory loss (especially vibration sense), dysarthria, lower limb areflexia, and pyramidal tract signs.203 EMG demonstrates a complete absence of SNAPs but is otherwise normal. Cardiomyopathy with abnormal EKG’s, impaired glucose tolerance, optic nerve atrophy, and skeletal abnormalities, such as pes cavus and scoliosis, are also common.117,203 The degree of clinical involvement is related to the number of trinucleotide repeats.151,327 Ataxia-Telangiectasia. This autosomal recessive disorder is associated with progressive ataxia, oculocutaneous telangiectasia, recurrent sinopulmonary infections, and a predisposition to malignancy.47 Involvement of the peripheral nerves is common but rarely clinically predominant.354 Carbohydrate-Deficient Glycoprotein Syndrome. This spectrum of peptide hormone glycosylation disorders is characterized by congenital hypotonia and dysmorphism with subsequent severe developmental delay and failure to thrive. Ataxia and strokelike episodes are seen in childhood. Peripheral neuropathy is evidenced by mild motor conduction slowing.199,488 Neuroradiologic studies typically demonstrate cerebellar hypoplasia and brainstem atrophy. Other Spinocerebellar Ataxias (see Chapters 73, 75, and 89). A number of other spinocerebellar ataxias occasionally present in the pediatric population with a peripheral neuropathy (especially sensory).346,423 Inheritance is quite variable; autosomal dominant, autosomal recessive, and

Although very rare, mononeuropathies are seen in all pediatric age groups. When found in neonates and toddlers, their distribution and frequency differ from that in adults (Fig. 122–4).140,148 Isolated mononeuropathies are often first detected by an observant parent or pediatrician who notices focal muscle weakness, atrophy, or abnormal positioning of a limb.234 It is particularly common for neonatal lesions to have unusual presentations and etiologies.84 In contrast to adults, in whom median nerve lesions predominate, childhood mononeuropathies are distributed equally between the median, ulnar, radial, peroneal, and sciatic nerves.453

Other

Adult Pediatric

Peroneal

Sciatic

Radial

Ulnar

Median 0

20

40

60

80

%

FIGURE 122–4 Relative distributions of mononeuropathies in 113 children and 712 adults.

Peripheral Neuropathies in Infants and Children

Many pediatric mononeuropathies are related to trauma; however, compression is also a relatively common mechanism,230,234,279,401,497 but entrapment does not occur very often. Very rarely, a mononeuropathy occurs as a manifestation of a few systemic illnesses. In contrast to adults, it is particularly unusual in children to have a mononeuropathy be the presenting sign of a yet to be discerned generalized polyneuropathy. The differential diagnosis of a pediatric mononeuropathy includes intraspinal lesions, plexopathies, and even, very rarely, intraparenchymal muscle lesions such as soft tissue sarcoma.95 Multifocal mononeuropathies are very uncommon in childhood.478,480 When seen acutely, they most commonly reflect an underlying vasculitic or other inflammatory process in the form of MNM.418 Additionally, HNPP should be considered. It is one of the rare multifocal nerve disorders that presents during childhood71 (see Chapter 70). Typically this autosomal dominant cause for recurrent mononeuropathies is not clinically appreciated until the second decade, when it presents with recurrent, often pressure-induced, focal neuropathies. One Ashkenazi child is reported who had the first presentation of his HNPP at birth, having transient bilateral footdrops. Other extremely rare causes of hereditary multiple pediatric mononeuropathies include either Tangier disease (see Chapter 82) or porphyria (see Chapter 80).

Evaluation of Pediatric Mononeuropathies The clinical evaluation of childhood mononeuropathy should include inquiry as to trauma and any activities predisposing to nerve compression. The temporal evolution of symptoms and presence or absence of associated pain may provide clues as to causation. The medical history is directed toward any inherited or systemic illness. EMG studies localize and characterize the nerve lesion, as well as determine its severity. Comparison of side-to-side difference is often valuable in the identification of axon loss lesions in children. Imaging studies, particularly MRI, may occasionally define the precise site of the nerve lesion. Variable fascicular involvement renders precise localization of some mononeuropathies difficult. It is therefore important to include proximal components of the affected peripheral nerve within the imaging field. With persisting or progressive mononeuropathies, surgical exploration may be indicated even in the face of negative imaging. At times a soft tissue entrapment may not be identified even with “appropriate” MRI.141,489 When a mononeuropathy or plexopathy occurs in a neonate, an EMG should be considered within 1 to 2 days of birth. In this setting this examination may help to differentiate lesions of intrauterine onset from those related to obstetric trauma.243,244,304 Similarly, MRI has been useful on occasion to define isolated mononeuropathies in the older

2725

child.498 On occasion, computed tomography (CT) is also a valuable adjunct.288

Upper Extremity Mononeuropathies Median Nerve Infantile Median Neuropathies. Pediatric median neuropathies at the wrist are uncommon, particularly in very young children. When they do occur, the child is too young to accurately describe his or her discomfort. At times, children will simply shake their hands in distress. One 8-month-old baby seen at CHB had dysesthesias presumably so severe as to lead to self-mutilation, with severe macerating ulcers of the hand. He was referred for consideration of congenital insensitivity to pain; however, his ulcers were confined bilaterally to a median nerve distribution (Fig. 122–5). EMG demonstrated severe bilateral median nerve entrapments at the wrist. Additionally, there was a striking family history of carpal tunnel syndrome (CTS).463 We presume that he chewed on his fingers because of the seemingly unbearable distress of the dysesthesias. Decompression led to complete resolution of his ulcers within a few months. Median Neuropathies in the Older Child. A typical CTS is very uncommon even in older children. In this age group, sensory manifestations present with unusual characteristics because, of course, the child is unable to provide the history so typical of this lesion in the older child or adult. A hand tremor in a 6-year-old child is an example of an unusual presenting manifestation of a CTS. This child was thought to have an unidentified primary movement disorder until the CTS was discovered (J. Brown, University of Western Ontario, personal communication). Decompression again was successful. Results of the neurologic examination of small children with CTS vary from normal to mild sensory abnormalities or

FIGURE 122–5 Macerated hand with congenital carpal tunnel syndrome. See Color Plate

2726

Diseases of the Peripheral Nervous System

eventually thenar weakness and atrophy, reduced movements of the fingers, repetitive hand shaking, hand tremor, or ulcerations. Very rarely, an acquired thenar muscle atrophy is observed without sensory involvement. This is particularly characteristic of the various mucopolysaccharidoses.104,234 Distal median neuropathies are very rare in young children.383 When present, these are usually associated with a hereditary congenital predisposition. Thus this diagnosis may only be made in extreme cases at this age.463 Another reported case was a child with a profound “tremor” (J. Brown, personal communication, 2001). In contrast, “classic CTS” occasionally occurs in older children or teenagers. These cases are often idiopathic, although sometimes certain athletic activities, such as skiing, predispose to such mononeuropathies.104,234 Rarely, some hereditary inborn errors of metabolism such as mucopolysaccharidoses, particularly Hurler’s syndrome, predispose to thenar atrophy. However, this is strictly a motor involvement with no “classic CTS” sensory symptomatology.194 Proximal median neuropathies in children are most commonly related to trauma, particularly bony fractures. Less common pediatric mechanisms include blunt nerve trauma, lacerations, blood pressure cuffs, and arterial or venous punctures.104 Rarely, entrapments secondary to a number of uncommon anatomic structures lead to other proximal median neuropathies. These structures include a ligament of Struthers, congenital fibromuscular bands, pronator teres, and bicipital aponeurosis. Children with proximal median nerve entrapment secondary to congenital constriction bands may also have concurrent involvement of the radial and ulnar nerves. Multilevel compression of the proximal and distal segments of the median nerve has resulted from lipofibromas, hamartomas, neurofibromas, and hemangiomas. Other causes of median nerve compression in children include osteoid osteoma,104 juvenile cutaneous mucinosis,104 calcified flexor digitorum superficialis tendon, and abscess. Rarely, an anterior interosseous neuropathy (AIN) occurs in children. Some typically these develop concomitant with an autoimmune brachial neuritis.146,338 On rare occasions, supracondylar fractures of the humerus also lead to a pediatric AIN neuropathy.338 Diagnostic studies important to the evaluation of a child with a median neuropathy include EMGs with appreciation of the maturation-dependent NCS parameters.324,462 Radiographs of the distal humerus are indicated when there is a possible median nerve entrapment at the ligament of Struthers. MRI is helpful for anatomic assessment of soft tissue infiltrative or compressive lesions.104,463 Plain roentgenograms of the hand are important in suspected cases of congenital thenar hypotrophy.67 Testing for a mucopolysaccharidosis or mucolipidosis is indicated with primary median motor neuropathies, especially if the child

has associated dysmorphic features, organomegaly, and other signs of storage disorders. Ulnar Nerve In contrast to the other primary mononeuropathies, we are unaware of any ulnar neuropathies presenting during infancy: The youngest in our experience with 21 childhood ulnar neuropathies was 5 years old.148 These usually occur at the elbow, and only rarely at the wrist. Trauma is the most common cause. Proximal ulnar neuropathies may result from supracondylar or medial epicondylar fractures or elbow lacerations. Distal radial fractures can cause ulnar neuropathies at the wrist. Fractures may cause nerve damage via blunt injury, entrapment, compression, or laceration. Delayed or “tardy” ulnar nerve palsy is described in children following elbow trauma, with posttraumatic bony changes and fibrosis.148 Rarely, ulnar nerve compression occurs during surgery or sleep in children, or is related to wheelchairs or bicycle hand grips.148 Other unique ulnar nerve compressions have occurred with hemorrhage in hemophilia, or extravasation of intravenous fluids with subsequent compartment syndrome.148 Ulnar mononeuropathies also occur in children related to leprosy,446 focal hypertrophic neuropathy, burns, repetitive throwing injuries in baseball pitchers, tumors,148 and hamartomas. Cubital tunnel syndrome with ulnar nerve entrapment in children has been documented at surgery.148 Other causes of pediatric ulnar nerve entrapments include persistent epitrochleoanconeus muscles, congenital constriction bands, or elbow fractures and/or dislocations.148 The management of pediatric ulnar neuropathy is determined by the underlying pathophysiology. At CHB, 12 of our 21 children required nerve decompression or transposition, or surgical repair for acute fractures and lacerations, cubital tunnel lesions, neuromas, and a nerve laceration.148 A 2-month to 6-year follow-up was available for 15 of these 21 children (71%). Children sustaining a nontraumatic ulnar neuropathy had a more favorable outcome, with 83% improving versus 56% with traumatic lesions.148 Radial Nerve Infantile Radial Neuropathies. These result from presumed intrauterine compression by uterine contraction rings,147,330 prolonged labor, subcutaneous fat necrosis,291 and subcutaneous abscess or hematoma (Fig. 122–6).409 At birth, the radial nerve is at risk for injury secondary to a humeral fracture, hematoma, blood pressure monitoring, and prolonged birth-related external compression.230 Radial Neuropathies in Older Children. These occur at the level of the axilla, spiral groove segment, posterior interosseous nerve, or superficial sensory branches. Most are due to fractures or lacerations.142 Pediatric radial nerve

Peripheral Neuropathies in Infants and Children

2727

compression from a Milwaukee brace, and idiopathic lesions.148b,157a,234 Musculocutaneous Nerve. Three childhood examples of musculocutaneous neuropathy include a child with nerve compression by a body cast, another idiopathic case leading to biceps atrophy, and a third with an isolated involvement of the lateral antebrachial cutaneous nerve.148b,234 A 16-year-old boy had painless biceps weakness and atrophy secondary to a musculocutaneous mononeuropathy superimposed upon an unsuspected diffuse HNPP confirmed by DNA analysis.148a FIGURE 122–6 Radial neuropathy in a newborn. (From In Jones, H. R., De Vivo, D. C., and Darras, B. T. [eds.]: Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 313, 2003, with permission.)

Suprascapular Nerve. Entrapment and trauma are the primary mechanisms for these childhood neuropathies.148a One teenager had an entrapment from a transverse ligament at the suprascapular notch.283a Some are related to athletic injuries. Very rarely, compression may occur secondary to a ganglion cyst.148a

trauma is also reported with injection injuries and arthroscopic elbow surgery. Childhood radial nerve compression usually occurs at the level of the spiral groove. Other pediatric radial compressive lesions have occurred during surgery or sleep, from crutches, and from compartment syndrome secondary to infiltration of a chemotherapeutic infusion.142 Very rarely, radial neuropathies are the presenting feature of HNPP.318 Radial nerve entrapments are found within the triceps muscle, and sometimes within the interseptal fascia of children.142,230 Benign tumors and other lesions rarely lead to radial neuropathies in children. These include lipomas, ganglia, fibromas, neuromas, and hemangiomas.142,230 EMG is important for localization and differential diagnosis (see Chapter 60). Early EMG studies are also indicated with lacerations and fractures to assess for nerve continuity with consideration of surgical exploration and nerve repair. MRI is sometimes also important for evaluation of pediatric radial neuropathies, looking for tumors or congenital anatomic variants leading to entrapment. Slowly progressive radial or posterior interosseous nerve dysfunction may require surgical exploration to search for fascial bands or other focal lesions even when MRI is nondiagnostic. A full recovery usually occurs with acute radial nerve compression at the spiral groove. Most children with traumatic radial mononeuropathies make very good recoveries.142 Other Pediatric Upper Extremity Mononeuropathies

Thoracodorsal Nerve. One child developed right latissimus dorsi atrophy following a chest tube insertion.148b We are unaware of other reported cases of thoracodorsal nerve lesions in children.

Axillary Neuropathies. These have occurred in children with sports-related shoulder trauma as well as a humeral exostosis.148b,230,234 Long Thoracic Neuropathies. Various mechanisms for pediatric long thoracic mononeuropathies include sportsrelated injuries (e.g., tennis, weight lifting), shoulder trauma,

Lower Extremity Mononeuropathies Sciatic Nerve Sciatic mononeuropathies comprised 24% of all childhood mononeuropathies evaluated at the CHB EMG laboratory over 14 years.140,141 These are as common as peroneal neuropathies in childhood. Pediatric sciatic nerve lesions occur within the pelvis, at the sciatic notch, or in the thigh, often presenting with weakness and paresthesias, a chronic progressive footdrop, or pes cavus deformity.28 Isolated involvement of the sciatic peroneal division occasionally mimics a peroneal lesion at the fibula head. Because the more laterally placed peroneal division of the sciatic nerve is prone to early outside compression, peroneal trunk fibers have greater susceptibility to stretch and other injuries.141,253 Thus a primary sciatic nerve lesion must be considered in the differential diagnosis of any child with a footdrop. Infantile Sciatic Neuropathies. Rarely, these neuropathies occur in neonates subsequent to traction during breech deliveries.141,242 However, most cases are idiopathic; the majority of children seen at CHB (16 of 21) had excellent recoveries except for those delivered by cesarean section, who had a poor outcome.391 Prenatal compression of the sciatic nerve is possibly related to reduced fetal movement, oligohydramnios, or structural uterine abnormalities leading to external compression. This occurs with Asherman’s syndrome (i.e., decreased amniotic fluid), amniotic fluid bands, abnormal uterine contractions during prolonged labor,

2728

Diseases of the Peripheral Nervous System

or uterine abnormalities including adhesions, synechiae, and bands.432 Occasionally some infants have a necrotic ischemic lesion (eschar) overlying the sciatic nerve at the site of intrauterine compression. One CHB newborn was “pavulonized”; within 3 days this baby lost all motor function below one knee. A pressure sore was noted over the right buttock. Seventeen days later EMG demonstrated no peroneal or tibial CMAPs. Sciaticinnervated muscles had just a few motor unit potentials with many fibrillation potentials. However, at age 8 weeks this baby’s weakness totally resolved.242 Newborn sciatic nerve entrapment defined by EMG was found in a baby with a weak and smaller foot.287 Surgical exploration at the sciatic notch demonstrated a congenital iliac anomaly with nerve entrapment. This case also illustrates the importance of early EMG shortly after birth providing documentation of a prenatal nerve injury.234 Another rare, neonatal severe sciatic lesion occurs secondary to inadvertent umbilical artery injection of an analeptic agent. The nutrient artery of the fetal sciatic nerve is a branch of the umbilical artery and usually closes at birth. However, when it remains patent, injection of this medication into the umbilical artery may cause a severe ischemic sciatic neuropathy.140,157

Sciatic Neuropathies in Older Children. Nontraumatic pediatric sciatic neuropathies are most commonly due to compressive lesions secondary to long leg and body casts or braces, abnormal lower extremity positioning, and hematomas.141,229,427,492,516,517 Compression occurred in a third of sciatic neuropathies at CHB.141 Focal entrapment of the sciatic nerve is very unusual in childhood. When present, it usually occurs in relation to myofascial or fibrovascular bands,489 or congenital iliac anomalies.287,464 Identification of lesions causing focal entrapment is sometimes aided by MRI.498 Occasionally MRI is unrevealing, and surgical exploration is necessary to identify a treatable lesion (Fig. 122–7).489 Surgery in the lithotomy position occasionally leads to devastating sciatic neuropathies in children.141,273,496 Proposed pathophysiologic mechanisms for these sciatic neuropathies include ischemia, stretch injury, and external compression. One postulate suggests the sciatic nerve blood supply in affected children is an anatomic variant with a persistent fetal artery leading to a nerve infarct possibly related to pressure or stretch.402a Rarely, traumatic dislocation and fracture-dislocations of the hip lead to laceration, traction, or compression of the sciatic nerve in children.141 Heterotopic ossification may

Superior gluteal nerve Gluteus medius muscle Gluteus minimus muscle Tensor fasciae latae muscle Sciatic nerve Posterior cutaneous nerve Semitendinosus muscle

Gluteus maximus muscle Site of lesion

Semimembranosus muscle Adductor magnus muscle

Biceps, long head Biceps, short head

B

Tibial nerve

Gastrocnemius muscle

A

Common peroneal nerve

FIGURE 122–7 Sciatic neuropathy in an adolescent. (From Escolar, D., Ryan, M., and Jones, H. R.: Lower extremity mononeuropathies. In Jones, H. R., De Vivo, D. C., and Darras, B. T. [eds.]: Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 324, 2003, with permission.)

2729

Peripheral Neuropathies in Infants and Children

develop subsequently, in rare cases entrapping the sciatic nerve.81,294 Crush injuries to the sciatic nerve in victims of an earthquake are also reported. Intragluteal medication injections previously also caused sciatic neuropathies but are rare today.90,176 Some were successfully repaired with end-to-end anastomosis.176 Rarely, ischemia and vasculitis lead to childhood sciatic neuropathies.107,140 Our CHB experience includes one case with a hypereosinophilic syndrome and another with meningococcemia. Rarely, hemophiliacs develop hematomas leading to sciatic nerve compression.127 Primary sciatic nerve tumors occur in children on rare occasions, primarily presenting with severe pain in the posterior leg extending into the foot. This distribution should raise suspicion of a sciatic nerve rather than a nerve root disorder. Such tumors may have little or no weakness.41,469 Back pain, although more common with intraspinal lesions, is in rare cases associated with pure sciatic neuropathies.469 At CHB, one teenager’s lymphoma presented with a rapidly progressive sciatic neuropathy.242 Occasionally no mechanism is identified for a child’s sciatic neuropathy despite MRI and surgical exploration.137,242,424 Six patients ages 10 to 27 years had idiopathic progressive lower extremity mononeuropathies.137 A painless sciatic neuropathy developed in two adolescents. Numbness in the ball and heel of the foot was the presenting complaint in one child; the other had muscle weakness and atrophy. Once a sciatic neuropathy is defined, neuroimaging is indicated to attempt to disclose its cause.1,152,498 Because of its large size, the sciatic nerve is relatively easily visualized on MRI. Surgical exploration may prove fruitful in the occasional instance wherein no pathophysiologic mechanism is identified.489 The prognosis for pediatric sciatic neuropathies is generally worse than that of other childhood mononeuropathies. Seven of our 16 children at CHB had no clinical improvement, and only 3 had a complete resolution.141 More favorable outcomes are reported, possibly related to etiologic differences.261 Peroneal Nerve Almost all of the neuropathies of the peroneal nerve seen at CHB (16 of 17 children, or 94%) were localized to the fibular head.241 Infantile Peroneal Neuropathies. These occasionally occur in newborns.84,88,244,432,458 An EMG performed 18 hours after birth provided compelling evidence that this neuropathy had an intrauterine onset (Fig. 122–8).244 The etiology of some of these neonatal mononeuropathies is unclear.88 Their frequent rapid (within 4 to 9 weeks) and complete recovery suggests a neurapraxic mechanism. Other etiologies included intravenous solution infiltration in premature infants,270 and footboard compression.153 Intrauterine rings might also contribute.84

0.5mV 5 ms/D

Peroneal STIM (0.2 ms) at knee

Right

Left

FIGURE 122–8 Newborn peroneal neuropathy.

Peroneal Neuropathy in the Older Child. The majority of pediatric peroneal neuropathies at CHB (10 of 17) occurred secondary to nerve compression.241 The majority were associated with therapeutic modalities, including casts, traction, straps, and intravenous footboards. Peroneal neuropathies associated with excessive weight loss, particularly in teenagers with anorexia,241 as well as chronic leg crossing or prolonged compression during water skiing on a knee board,480a have also occurred in children.503 Here NCSs demonstrate conduction block at the fibular head.241,254 A spontaneous nontraumatic anterior tibial compartment syndrome caused a transient adolescent peroneal neuropathy.445a As with any compression neuropathy, HNPP always requires consideration as the underlying mechanism.160 Common or deep peroneal nerve entrapment warrants consideration in children presenting with a variety of symptoms.448 These include chronic progressive footdrop, repeated ankle sprains secondary to peroneal muscle weakness, or cavus foot deformities.435 Causes of pediatric peroneal entrapment include bony exostoses at the fibular head288 and talotibial exostosis derived from osteochondromas.123a Entrapment of this nerve also occurs on rare occasions within the tendon of the peroneus longus.140,435 Rarely, scarring resulting from repetitive ankle sprain entraps the superficial peroneal nerve; an MRI can define the diagnosis, leading to surgical correction.90a,457 The fascia of the lateral compartment at the origin of the superficial peroneal nerve is another site of potential entrapment. Rarely, this leads to bilateral neuropathies.310 A variety of soft tissue and primary neural lesions are also found related to various peroneal neuropathies, including schwannomas, hemangiomas, and perineuromas.140,241,348 Additionally, synovial cysts, intraneural ganglion cysts, and a ganglion in the anterior compartment of the leg have all

2730

Diseases of the Peripheral Nervous System

led to pediatric peroneal neuropathies presumably secondary to compression and not entrapment. However, both mechanisms may be operative in some instances.140 One child presented with a painful footdrop.345a An adolescent had a fluctuating footdrop worsened by sporting activities secondary to an intraneural peroneal cyst.21a,30a One 3-year-old child, with a ganglion cyst overlying the fibula head, presented with “in-toeing.” Knee joint swelling secondary to anaphylactic purpura caused a common peroneal neuropathy in a 2-year-old.399a Occult nerve tumors sometimes present with a long history of slowly progressive pain and weakness, or sometimes just paresthesias.241,458 These are typically schwannomas and hemangiomas.42a Rarely, intraneural perineuromas, also referred to as localized hypertrophic neuropathy, cause a progressive peroneal neuropathy. This benign condition was previously difficult to diagnose, frequently requiring surgical exploration.241 Today, magnetic resonance (MR) neurography and ultrasound techniques have improved the diagnosis of these nerve tumors.308a,436 Peroneal neuropathy in children very rarely can be the presenting sign of a systemic disease, including a vasculitis related to either systemic lupus erythematosus240 or diabetic neuropathy.25a,127a,285 Leprosy involving the peroneal nerve occurred in an immigrant boy from Vietnam who presented with a 1-year history of unilateral footdrop.74a In some children the etiology of a peroneal neuropathy remains unknown despite surgical exploration and pathologic analysis of the nerve.137 Trauma to the peroneal nerve is surprisingly uncommon in comparison to the other pediatric mononeuropathies.241 In addition to EMG study, MRI, MR neurography, and ultrasonography may be useful in identifying the location and etiology of nerve injuries.293a,308a,436 Precise localization of nerve lesions enables more directed approaches with surgical exploration.1,498 Early surgical repair is indicated in those children in whom significant improvement is not seen within 3 to 4 months of nerve injury.43a Primary repair of childhood peripheral nerve lesions often results in significant recovery. The outcome of pediatric peroneal mononeuropathies is variable. In the CHB series, 13 of 17 children had a complete or significant improvement, and 4 patients made a poor recovery.241 EMG findings are prognostically helpful; absent or low-amplitude CMAPs are commonly associated with an unfavorable outcome, and children with conduction block at the fibular head have a better prognosis.241 Femoral Nerve Femoral neuropathies are extremely rare in children.465 Most children present with weakness, pain, or numbness of the anterior thigh, or with atrophy of the quadriceps femoris.65 Only three children with femoral neuropathy were identified in an unpublished 16-year preliminary retrospec-

tive review at CHB. Two developed subsequent to surgical procedures to the pelvis or hip joint (H.R. Jones and J. Srinivasan, unpublished data). Lateral Femoral Cutaneous Nerve of the Thigh. This is an uncommon focal neuropathy in children. The clinical manifestations are primarily sensory, with an elliptical loss of sensation over the lateral thigh. Pediatric lateral femoral cutaneous neuropathies occur as complications of orthopedic harnesses,359 osteoid osteomas,182 hemangiomas, and blunt trauma to the thigh.294,301 Sometimes no mechanism is defined.234,399 The diagnosis of lateral femoral cutaneous neuropathy is usually a clinical one. When symptoms are atypical or persistent, however, further investigation may be indicated. We have evaluated two young adults with soft tissue sarcomas presenting with thigh pain, one of which mimicked a lateral femoral cutaneous neuropathy.96 Thus the potential for a similar mechanism requires consideration in children, particularly adolescents. Tibial Nerve Tibial mononeuropathies are rare, but the tibial nerve can concomitantly be affected with the peroneal nerve at the popliteal fossa. One case report of “tarsal tunnel syndrome” in girls with foot pain was not confirmed by NCS or EMG.7 Sural Nerve Isolated sural mononeuropathies are extremely uncommon in children. There is a single report of reversible sural neuropathy secondary to compression from a tight ankle bracelet in an adolescent girl.395

PLEXOPATHIES OF INFANTS AND CHILDREN Brachial Plexopathy (see Chapter 55) Two main syndromes of brachial plexopathy are seen in childhood: birth-related perinatal lesions and a much broader spectrum of less common disorders occurring variously during infancy, childhood, or adolescence.246 Perinatal Brachial Plexopathies Brachial plexus palsies complicate 0.5 to 2 in 1000 live births.24,56,256,411,456,484 These are classified into four anatomic groups (see Table 122–1).126,340 Clinical Picture. Paralysis of the upper roots, derived primarily from cervical nerve roots C5-C7 (Group I, Erb’s palsy), is by far the most common presentation.5,136,184,484 A typical infant with Erb’s palsy lies with the affected arm adducted at the shoulder and internally rotated, the elbow

Peripheral Neuropathies in Infants and Children

extended, the forearm pronated, and the fingers and wrists flexed. This characteristic posture is commonly referred to as the “waiter’s tip” position (Fig. 122–9). A concomitant Horner’s syndrome indicates added involvement of the T1 nerve root, and is often accompanied by severe intrinsic hand weakness (C8 root). The presence of long thoracic, dorsal scapular, and C8-T1 sympathetic nerve dysfunction implies a very proximal plexopathy, usually at the level of the nerve root entry zone secondary to complete avulsion of the spinal nerves from the cord. Infants whose brachial plexopathies involve these levels have a relatively poor prognosis for recovery.6 Quantification of motor function is a major challenge when assessing these babies. Because the Medical Research Council muscle grading system is not applicable to infants, a number of alternative grading scales have been proposed.75,173,320 Sensory assessment of infants is also difficult. In general it is usually possible to determine only whether the infant responds to painful stimuli.

2731

EMG patterns consistent with a prenatal, intrauterine brachial plexus injury include the early presence of fibrillation potentials, presumably too soon to have had their onset subsequent to birth.264,365 MRI is also a useful means to evaluate for signs of an avulsion (Fig. 122–10).304

Predisposing Risk Factors. These include gestational diabetes, birth weight greater than 4500 grams, shoulder dystocia, and instrumental vaginal delivery.105,175,184,284,313, 411,479 Additionally, similar brachial plexopathies have occurred after cesarean sections. Presumed mechanisms in such infants include abnormal fetal positioning and amniotic bands.382

Prognosis. At least two thirds of infants with obstetric brachial plexopathies have a good to excellent outcome. One quarter have moderate residual damage, and 8% to 15% have significant permanent disability.135,136,256,320,415 Lesions confined to the upper trunk and lateral cord clearly have the best prognosis, although some late (at age ⫾ 5 years) impairment of hand function may be seen even in children with C5-C6 lesions, presumably as a result of impaired shoulder rotation secondary to the initial upper trunk brachial plexopathy.456 Infants with Erb’s palsy who have early signs of recovery, within the first 2 months of life, almost always have a seemingly complete resolution of motor deficit by age 5 years. However, recent studies are causing some reevaluation of this statement.456 A poor prognosis is predicted by (1) lower plexus lesions with associated Horner’s syndrome, (2) total plexus involvement, (3) the absence of appreciable recovery by 6 months of age,340 (4) concomitant phrenic nerve palsy,6,256,340 and (5) nerve root avulsion.

Diagnostic Evaluation: Electromyography/MRI. EMG determine the more precise distribution and severity of neonatal brachial plexopathy.235,284,449,468 Results may occasionally be falsely reassuring in some neonates. This partially relates to difficulties in assessment of motor unit numbers and firing rates in neonates.57,483 There are three specific goals of EMG in evaluation of the infant with a brachial plexus lesion: (1) establishment of the level and extent of plexus injury, (2) accurate identification of preganglionic lesions and/or root avulsion, and (3) definition of the underlying pathophysiology of the lesion (i.e., neurapraxia, axonotmesis, and/or neurotmesis). However, these goals can be very difficult to achieve in children.235,246 Neonatal plexus lesions often have both preganglionic and postganglionic components. This complicates identification of presence or absence of root avulsion.510 There is no well-defined set of standards to use for determination of precise timing of the nerve damage in the infant in comparison to the adult.74 Although NCS loss parameters in the older child with complete axonal disruption (i.e., CMAPs disappear within a week of injury, and SNAPs may persist for 10 to 11 days),74 are similar to those in adults, such a profile is not defined for babies. Similarly, standard needle EMG changes indicative of acute denervation (i.e., fibrillation potentials at 10 to 14 days) do not apply to infants.243,297,462 Acute denervation potentials may occur as early as the fourth day after injury.183,304

Surgical Repair. The relationship between EMG findings, prognosis, and indications for surgery remains controversial for the infant born with a brachial plexopathy. The primary challenge lies in distinguishing those children with potential for significant spontaneous recovery from those, if any, who require nerve surgery to aid their progress. Assessments of elbow flexion together with elbow, wrist, thumb, and finger extension at age 3 months may best predict recovery at 12 months.320 The appropriate timing for surgery is often debated. Children with incomplete lesions should be followed, with serial clinical evaluations, for at least 3 months after birth. Most infants recovering elbow flexion by age 2 months eventually develop full recovery of limb function. A near-normal recovery is also reported in babies with some evidence of elbow flexion by 3 months of age.6 Surgical intervention is sometimes carried out in babies who fail to develop a biceps contraction by age 3 to 4 months, although some of these babies will proceed to develop a complete and spontaneous recovery.135,191,256,320 Other surgical indications include a complete brachial plexopathy with a flail arm and Horner’s syndrome, or a C5-C6 palsy with no recovery in biceps function at age 3 to 6 months and an EMG consistent with an axonal lesion. Spontaneous recovery in these babies is relatively uncommon, whereas some recovery of function after surgery is reported in as many as 80% of children173,174,468 but is not invariable.456

FIGURE 122–9 Newborn brachial plexopathy. (From Jones, H. R., Miller, T. A., and Wilbourn, A.: Brachial and lumbosacral plexus lesions. In Jones, H. R., De Vivo, D. C., and Darras, B. T. [eds.]: Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 248, 2003, with permission.)

Peripheral Neuropathies in Infants and Children

2733

initial therapy and partial reinnervation, these children may develop internal rotation and adduction contractures at the shoulder, flexion contracture of the elbow, and dislocation of the radial head. Therapeutically these are best treated with tendon transfer procedures.43,126 Concomitant posterior subluxation or dislocation of the glenohumeral joint requires surgical release, reduction, and reconstruction of the capsule.43,191 Other children develop hypoplasia of the arm. Physiotherapy is indicated to prevent joint contractures. This focuses upon maintenance of passive range of flexion, extension, and rotation of the forearm and elbow, and external rotation of the shoulder.390 Gentle passive stretching techniques can easily be taught to parents. Other Observations. It is also very important to be cognizant of the psychological and social consequences of plexopathies. They require careful multidisciplinary attention.174 Parenthetically, one must take note that these neonatal plexopathies are inappropriately often referred to as obstetric brachial plexopathies. In our litigious society, the term obstetric is being utilized in an adversarial format as synonymous with one specific etiologic inference—the obstetrician—rather than in reference to the purely temporal relationship of the lesion to the obstetric process per se. As noted earlier, there are other associated risk factors apropos of this plexopathy; some are totally unrelated to the skill of the attending obstetrician.

FIGURE 122–10 Magnetic resonance image of infantile brachial plexopathy. (From Jones, H. R., Miller, T. A., and Wilbourn, A.: Brachial and lumbosacral plexus lesions. In Jones, H. R., De Vivo, D. C., and Darras, B. T. [eds.]: Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 250, 2003, with permission.)

The prognosis for natural shoulder function recovery is much better than that achieved in the hand because of the shorter distances required for reinnervation.173,174 Preganglionic root avulsions generally are irreparable,191,256,468 although some heroic procedures to repair such lesions have been undertaken soon after birth.191,340,468 Rehabilitation. Children who have persistent loss of C5-C6 function have poor external rotation, shoulder elevation, elbow flexion, and forearm supination. Despite

Other Infantile Brachial Plexopathies Acquired brachial plexopathies are very uncommon during early infancy. The septic plexopathy associated with proximal humerus osteomyelitis is an eminently treatable entity that requires early recognition for successful therapy (Fig. 122–11).76,235 These require a differentiation from an inherent unsuspected neonatal brachial plexopathy or a pseudo-palsy secondary to limb pain or deformity. Important diagnostic clues include fever, an elevated ESR, and blood cultures positive for bacterial infection. Although a bone scan may aid diagnosis if it is positive for a humeral lesion, it is not always initially positive.235 Therefore, when confronted with any febrile infant with a possibly acute brachial plexopathy, a search for proximal humeral osteomyelitis is indicated. Rarely, a pyogenic cervical osteomyelitis, with an associated paraspinal abscess and neonatal Erb’s palsy, also occurs during infancy. Staphylococcus aureus is usually the primary organism, although rarely gram-negative organisms are isolated.430 Certain mass lesions, including malignant rhabdoid tumors and plexiform neoplasms, have mimicked neonatal plexopathies.224 Features distinguishing these lesions from an idiopathic neonatal brachial plexopathy include (1) the onset of weakness after the first postpartum day, (2) a progressive course, (3) history of normal delivery

2734

Diseases of the Peripheral Nervous System

B

A

FIGURE 122–11 Osteomyelitis-associated brachial plexopathy. (From Jones, H. R., Miller, T. A., and Wilbourn, A.: Brachial and lumbosacral plexus lesions. In Jones, H. R., De Vivo, D. C., and Darras, B. T. [eds.]: Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 262, 2003, with permission.)

and birth weight, (4) the appearance of a mass in the supraclavicular region without radiographic evidence of clavicle fracture, and (5) scratch marks on a weak arm.10,11 Neonatal hemangiomatosis is another rare mechanism.296 Neonatal brachial plexopathies are also very occasionally associated with an exostosis of the first rib.10 The differential diagnosis of these various other brachial plexus lesions in a previously healthy infant also includes anterior horn cell and peripheral nerve lesions.10,11,30,142,148

Brachial Plexopathies in Older Children Most childhood brachial plexopathies have a well-defined etiologic mechanism, trauma being the most common cause. Other lesions, such as brachial plexitis, hereditary mechanisms, thoracic outlet, and tumors, are relatively uncommon. In about 10% of children, there is no identifiable mechanism. Trauma. This is the most common mechanism leading to a brachial plexus lesion in the older child. Additionally, one sees children with probable autoimmune (Parsonage-Turner brachial plexitis) as well as hereditary plexopathies. Child abuse must always be considered in any child with an “idiopathic” plexopathy. We evaluated one toddler with a recurrent brachial plexopathy. An abusive parent swinging the child by an arm and then throwing him proved to be the cause.246 One needs to be cautious in

arriving at such a conclusion, because childhood brachial plexopathies also occur spontaneously, particularly in cases of HNPP or hereditary or idiopathic brachial plexus neuropathy. Adolescents are particularly prone to athletic injuries, including burners or stingers from football and other contact sports injuries,212,400,503,507 motorcycling accidents, and potentially even gunshot wounds. Auto accidents predispose to airbag deployment with its potential for plexus injuries.317 Occasionally a brachial plexopathy develops as an iatrogenic complication of pediatric surgery. Idiopathic Brachial Plexus Neuritis. This is also known as the Parsonage-Turner syndrome or neuralgic amyotrophy. It is the most common form of a nontraumatic brachial plexopathy in childhood. The youngest child was just 3 months old.476 About one third of patients have bilateral, although asymmetrical, involvement. One case of a painless diffuse brachial plexitis presenting with flaccid weakness of the entire arm and weak grasp 4 days after a febrile illness had a poor recovery. This compares to the excellent recovery of an idiopathic brachial plexus neuropathy confined to the proximal arm. Genetically Determined Plexopathies. Hereditary brachial plexus neuropathy (HBPN) (hereditary neuralgic amyotrophy) is an unusual cause of childhood brachial plexopathy.72,114 Affected children have repeated episodes of shoulder and arm pain, followed by weakness and

Peripheral Neuropathies in Infants and Children

atrophy with gradual recovery. Onset as early as 7 weeks of age is reported,114 although most patients do not become symptomatic until adolescence.72 One of the phenotypic markers of HBPN is a characteristic facies, with hypotelorism and a long nasal bridge.72,114,372 Most children with HBPN (Fig. 122–12) (see Chapter 70) present with recurrent mononeuropathies, primarily of nerves liable to focal compression. In 10% of HNPP patients, there is primary involvement of the brachial plexus. Presentation in such cases is also with a painless weakness, often after minimal or seemingly coincidental trauma.32 Again this contrasts with the marked discomfort associated with idiopathic and hereditary brachial plexitis.490 Pediatric pressure-related brachial plexopathies independent of HBPN are very uncommon, but there have been occasional reports. These compression plexopathies have been reported related to infant seat harnesses375 and

2735

rucksacks.212,405 The latter study reported that 15 of 16 sports-related plexopathies were secondary to mountain climbing with packs as heavy as 40 kg. These lesions were proposed to occur secondary to axillary compression.212 Our extremely limited experience with these adolescent brachial plexopathies suggests that better structure and padding of today’s backpacking equipment, as well as waistbands to support the load, may account for the relative infrequency of these plexopathies at CHB. Other Childhood Brachial Plexopathies. Illicit heroin use is another unusual atraumatic brachial plexopathy.70 This mechanism merits occasional consideration in adolescents. Thoracic outlet syndrome is a very rare consideration when evaluating a child with an apparent medial plexus lesion.403 At CHB we have seen just one well-defined instance of this syndrome in an adolescent in over 25 years.177,409,504 One 4-yearold child with Klippel-Feil anomaly associated with cervical ribs developed thenar atrophy and weakness.358 Neoplasms rarely affect the brachial plexus in children.10,11,224 These particularly need to be considered in those cases with no history of antecedent trauma, fever, or hereditary predisposition. MRI is very useful for detecting the occasional unsuspected neoplasm (Fig. 122–13).10,11 Additionally, a very unusual brachial plexopathy has occurred following therapy for childhood acute lymphoblastic leukemia.207

Lumbosacral Plexopathies These lesions are extremely uncommon in childhood. Although it is important to recognize that the generic term lumbosacral plexus is used as a means of logical organization, on occasion the clinical lesion may predominate in either the lumbar or sacral portions. Usually however, there is a significant overlap in pediatric patients, reflecting their relatively small body size. There are relatively few mechanisms to consider. MRI is the diagnostic study of choice for suspected lumbosacral plexopathy and is useful in excluding the possibility of pseudomeningocele or occult myelomeningocele. Electromyography is not necessary where MRI is clearly diagnostic; however, it is very useful when the MRI is non-specific.

FIGURE 122–12 Hereditary neuralgic amyotrophy. (From Jones, H. R., Miller, T. A., and Wilbourn, A.: Brachial and lumbosacral plexus lesions. In Jones, H. R., De Vivo, D. C., and Darras, B. T. [eds.]: Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 263, 2003, with permission.)

Perinatal Lumbosacral Plexopathies Occasional cases of lumbosacral plexopathy are reported after complicated breech deliveries.134,216 Recovery in such cases is commonly incomplete. Prognosis varies from modest216 to good. Lumbosacral Plexopathies in Older Children A 16-year review of our experience at CHB defined just 6 children with a lumbosacral plexopathy in comparison to 54 with brachial plexopathies.235

2736

Diseases of the Peripheral Nervous System

FIGURE 122–13 Brachial plexus tumor. (Courtesy of Diana Escolar, M.D.)

Stretch Injuries. Most lumbosacral plexopathies are related to discrete traumatic insults, particularly stretch injuries. One was a “Raggedy Ann”–like stretch injury occurring when a loving 3-year-old sister proudly towed her 4-month-old brother around the room by his leg. He had a good improvement 7 months later.246

Neoplasms. Lymphoma and rarely other pelvic tumors occasionally involve the lumbosacral plexus in children. Such lesions always merit consideration in any child with an insidious onset of pain, paresthesias, and leg weakness compatible with a lesion at the level of the lumbosacral plexus.246,371

Other Trauma. A number of stretch injuries are associated with conjoint trauma to the pelvis, sacrum, or hip with dislocation.22,27,513 Such injuries include fractures, knife injuries, gunshot wounds, and motor vehicle accidents.69 After an acute abdominal crush injury, one 13-month-old developed an almost flaccid left leg with areflexia. Over the next few days, a similar loss of right leg function occurred.125

Miscellaneous Mechanisms. Lumbosacral plexopathies associated with hemophilia, vaccinations,306,492 and surgery are also reported.69,472

Hereditary Liability to Pressure Palsies. In some children with lumbosacral plexopathies there are some underlying, complementary mechanisms. These include HNPP185 and Ehlers-Danlos syndrome.160,161 Lumbosacral Plexitis. These lesions, which are analogous to idiopathic brachial plexus neuropathy, also occur in childhood. Affected children develop acute pain, leg weakness, and atrophy spontaneously or after viral illness.22,69,472 The prognosis for recovery is generally good. To our knowledge, diabetic lumbosacral radiculoplexus neuropathy (also known as diabetic amyotrophy) has not been reported in the pediatric population.

PEDIATRIC RADICULOPATHIES Lumbosacral and cervical radiculopathies are very rare in childhood. Most occur in adolescence, although instances as early as age 27 months are reported.371 Most pediatric radiculopathies involve the lumbosacral nerve roots. Cervical radiculopathies are extremely uncommon and may be mimicked by other rare cervical spine lesions such as congenital spinal stenosis. It is possible that a proportion of nerve root injuries go unrecognized in children, because they are rare, the accompanying clinical signs may be relatively subtle, and plain radiographs are commonly nondiagnostic. Differentiation between nerve root lesions and mononeuropathies is aided by both MRI and electromyography. Today, MRI is the diagnostic study of choice. EMG is usually

Peripheral Neuropathies in Infants and Children

2737

straightforward in the adolescent; however, in younger children needle examination of the paraspinal musculature is especially challenging within the cervical paraspinals. In anxious or restless children, it can be very difficult to achieve adequate relaxation to be completely confident of the absence of such insertional activity. Nerve root avulsion and brachial plexus injuries often occur concomitantly in children with traumatic injuries. It is always important to consider the possibility of coexisting more proximal nerve root injury in children whose EMG findings are compatible with a primary plexus lesion.304 Similar concerns arise in the (rare) pediatric cases of infiltrating malignancy, in which differential involvement of the nerve roots and plexus can be difficult to ascertain.

of cervical radiculopathies.309 These adolescents typically present with unrelenting neck and/or throat pain, often followed by torticollis and sometimes subluxation at this joint. Usually there is a preceding history of an influenzalike illness, otitis, tonsillitis, or pharyngitis. Subsequently, these children develop the insidious onset of a progressively severe and localized spinal pain that often becomes unremitting, sometimes radiating into the shoulder. A torticollis may also precipitously develop. Other neurologic manifestations may include greater occipital nerve paresthesias, myelopathy with transient or fixed quadriparesis, and rarely death. Other diagnostic possibilities include brachial plexitis or hereditary brachial plexus neuropathy and HNPP.222,454

Cervical Radiculopathies

Lumbosacral Radiculopathies

Cervical radiculopathies are very uncommon in adolescence and, to our knowledge, have not been reported in young children.239 Of 561 patients with cervical radiculopathy seen at the Mayo Clinic, only 41 cases (7%) were ages 15 to 29 years.389 We suspect that the distribution of affected nerve roots in children mimics that seen in adults, although this impression is difficult to substantiate. Contact athletic activities such as football or wrestling may be significant predisposing and triggering factors. One has to differentiate root lesions from other traumatic lesions at the level of the brachial plexus, particularly the stinger/burner injury. These athletic injuries represent either a C5/6 cervical radiculopathy or a mild brachial plexus upper trunk stretch injury. Because clinical presentation is consistent with either or both of these lesion sites and because they have no apparent long-term sequelae, stingers have not been further defined with MRI or EMG. MRI is the most useful investigation for suspected radiculopathy. Myelography is rarely required.475

Most radiculopathies of childhood are lumbosacral in location.102,154,276 In one series from two large North American centers, a pediatric patient with lumbosacral radiculopathy was seen, on average, once every 3 years.154 A higher incidence of juvenile disc herniation is reported in Japan.276 The youngest reported case was a 27-month-old boy who fell from his crib.397

Differential Diagnosis This includes a number of conditions presenting with progressive upper extremity weakness and atrophy. However, the lack of neck and shoulder pain so characteristic of most cervical radiculopathies makes this a less difficult differential. These other processes include congenital spinal stenosis and, very rarely, juvenile motor neuron disorders (see Chapter 53). Hirayama disease is an idiopathic disorder, more common in males, that presents during adolescence with painless hand weakness (see Chapter 53). The weakness is slowly progressive but usually stabilizes by the middle of the third decade. There may be mild involvement of the contralateral arm.213,258a Juvenile motor neuron disease very rarely occurs.91,235,246 Clinical and electromyographic findings mirror those of amyotrophic lateral sclerosis in adults.91,247 Grisel syndrome, an uncommon infectious process primarily leading to subluxation of the atlantoaxial joint, is another condition to consider in the differential diagnosis

Predisposing Factors These include a recent growth spurt434 or trauma related to athletic endeavors, lifting, or falls.102,154,276 Lumbosacral radiculopathies may develop after scoliosis repair in as many as 29% of children.116 Clinical Features Most cases are adolescents presenting with back pain and/or sciatica.102,154,276 Shillito voiced the concern that, in some adolescents, “lumbar disc disease may be missed because of the absence of sciatica.”434 Presentation with painless scoliosis is occasionally reported.434 Focal weakness is seen in 32% to 60% of adolescents, most commonly affecting the L5 myotome.102,154,276 Sensory loss is variable.154 Most children have a positive straight leg raising test or pain on back motion.434 Sphincter dysfunction is extremely uncommon.102,154,276 Diagnostic Evaluation Imaging Modalities. MRI is the study of choice for a suspected lumbosacral radiculopathy. CT imaging is sometimes simpler, briefer, and “kinder to the child.”434 It provides a clear differentiation between bone, disc, ligaments, and ossified ligaments. Routine spinal radiographs demonstrate anomalies of spinal segmentation in as many as 30% of affected children, although these may not directly contribute to nerve root compression.102,434 Myelography is rarely performed today. EMG Examination. This occasionally demonstrates lumbar radiculopathy even with normal myelography.138 However, a

2738

Diseases of the Peripheral Nervous System

comparison with MRI in children has been reported. The false-negative rate of pediatric EMG in this situation is unknown but may be significant given the difficulty in adequately assessing the paraspinal musculature in children. Differential Diagnosis Infections. Disc space infection or epidural abscess always requires consideration in children with back pain, nerve root symptoms, and an abnormal gait, especially if there is a fever.168a,397 Diskitis may sometimes lead to an associated disc protrusion.78a This is usually associated with fever, severe pain, and refusal to bear weight. An MRI and/or CT will help to exclude spontaneous disc space infections, because these lesions have a presentation similar to that of classic disc disease. Ankylosing Spondylitis. This is a very important consideration in the differential diagnosis of sciatica in adolescents, especially in boys. Because of the potential for ankylosing spondylitis to eventually lead to crippling spinal arthritis, it is incumbent on clinicians to recognize this disorder early on in the illness. Often, these adolescents have a concomitant history of early morning stiffness and “jelling” associated with diminished chest excursion on examination. Neoplasms. Nerve root tumors are uncommon but also present with radicular pain, paresthesias, and progressive weakness. The various lesions identified include schwannoma, hemangiomas, osteoblastomas, and leukemic infiltration.16,280,406,412 Even with leukemic infiltration, the child may present with what appears to be a single nerve root lesion. Sacral bone tumors may present early with a combination of radicular pain, paresthesias, and progressive weakness mimicking a lumbosacral radiculopathy. One review of 16 children reported that 13 had Ewing’s sarcoma and the other 3 cases were osteoblastoma, hemangiopericytoma, and chordoma.269a Occasionally congenital tumors are associated with a tethering of the spinal cord. These may or may not be associated with other vertebral column developmental anomalies. Sometimes these children present with painless footdrop or foot deformities. In the two instances we evaluated at Lahey Clinic and at CHB, there was no associated back pain. This emphasizes the need to always consider the possibility of an intraspinal lesion with any apparent mononeuropathy or plexopathy. Primary Spinal Scoliosis Surgery. This also predisposes patients to pediatric lumbosacral radiculopathies. One report noted a 29% incidence following surgery.116 Direct surgical trauma was responsible in 50% of these children. The remaining injuries were secondary to traction in the lumbosacral area.

Congenital Gynecologic Anomalies. Very rarely, these unique lesions can mimic a lumbar radiculopathy in a female adolescent.97a The pain in one adolescent interfered with sleeping and sitting but, interestingly, was relieved by bending forward or lying on her side. Examination demonstrated typical findings of an L5 radiculopathy. Gynecologic evaluation revealed no vaginal opening. Pelvic ultrasonography demonstrated a midpelvic mass. At surgery, this proved to be a congenital absence of the lower one third of the vagina. Vaginoplasty released 750 mL of menstrual fluids. Subsequently she had immediate relief of her low back pain. The dilated imperforate vagina with the hematometra likely compressed her lumbosacral plexus, mimicking a radiculopathy.97a Noting that only about half of back pain referrals to a pediatric orthopedic clinic had spinal disease, and that no diagnosis was reached in the other children, it was suggested that pelvic ultrasonography be performed with the evaluation of any indeterminate back pain and lumbar radiculopathy. The incidence of this type of congenital anomaly in adolescent girls is 1 in 5000. Therapy Conservative management is commonly successful in childhood lumbosacral radiculopathy.102,154,276 The Mayo Clinic study suggests that fewer than 1 in 25 children with lumbar radiculopathies eventually require a laminectomy. In those who do ultimately require surgery, a good result can usually be anticipated, although reoperation may be required in as many as 10% to 25% of these adolescents.102,138,154,276

REFERENCES 1. Aagaard, B. D., Maravilla, K. R., and Kliot, M.: MR neurography: MR imaging of peripheral nerves. Magn. Reson. Imaging Clin. N. Am. 6:179, 1998. 2. Abd-Allah, S. A., Jansen, P. W., Ashwal, S., and Perkin, R. M.: Intravenous immunoglobulin as therapy for pediatric Guillain-Barré syndrome. J. Child. Neurol. 12:376, 1997. 3. Aguayo, A. J., Nair, C. P. V., and Bray, G. M.: Peripheral nerve abnormalities in the Riley-Day syndrome. Arch. Neurol. 24:106–116, 1971. 4. Aicardi, J., and Castelein, P.: Infantile neuroaxonal dystrophy. Brain 102:727, 1979. 5. Al Qattan, M. M., Clarke, H. M., and Curtis, C. G.: Klumpke’s birth palsy: does it really exist? J. Hand Surg. Br. 20:19, 1995. 6. Al Qattan, M. M., Clarke, H. M., and Curtis, C. G. : The prognostic value of concurrent Horner’s syndrome in total obstetric brachial plexus injury. J. Hand Surg. Br. 25:166, 2000. 7. Albrektsson, B., Rydholm, A., and Rydholm, U.: The tarsal tunnel syndrome in children. J. Bone Joint Surg. Br. 64:215, 1982.

Peripheral Neuropathies in Infants and Children 8. Alderson, K., Seay, A., Brewer, E., and Petajan, J.: Neuropathies in children with chronic renal failure treated by hemodialysis. Neurology 35(Suppl. 1):94, 1985. 9. Alexianu, M., Christodorescu, D., Vasilescu, C., et al.: Sensorimotor neuropathy in a patient with MarinescoSjogren syndrome. Eur. Neurol. 22:222, 1983. 10. Alfonso, I., Alfonso, D. T., and Papazian, O.: Focal upper extremity neuropathy in neonates. Semin. Pediatr. Neurol. 7:4, 2000. 11. Alfonso, I., Papazian, O., Prieto, G., et al.: Neoplasm as a cause of brachial plexus palsy in neonates. Pediatr. Neurol. 22:309, 2000. 12. Al-Qudah, A. A.: Immunoglobulins in the treatment of Guillain-Barré syndrome in early childhood. J. Child Neurol. 9:178, 1994. 13. Al-Qudah, A. A., Shahar, E., Logan, W. J., and Murphy, E. G.: Neonatal Guillain-Barré syndrome. Pediatr. Neurol. 4:255, 1998. 14. Amato, A. A., Gronseth, G. S., Callerme, K. J., et al.: Tomaculous neuropathy: a clinical and electrophysiological study in patients with and without 1.5-Mb deletions in chromosome 17p11.2. Muscle Nerve 19:16, 1996. 15. Amit, R. Parker, M., Newman, G., et al.: Guillain-Barré syndrome in an immunosuppressed child (abstract). Pediatr. Neurol. 11:107, 1994. 16. Anderson, S. C., Baquis, G. D., Jackson, A., et al.: Ventral polyradiculopathy with pediatric acute lymphatic leukemia. Muscle Nerve 25:106, 2002. 17. Anderson, S. L., Coli, R., Daly, I. W., et al.: Familial dysautonomia is caused by mutations of the IKAP gene. Am. J. Hum. Genet. 68:753, 2001. 18. Appleton, R., Riordan, A., Tedman, B., et al.: Congenital peripheral neuropathy presenting as apnoea and respiratory insufficiency. Dev. Med. Child Neurol. 36:545, 1994. 19. Arbus, G. S., Barnor, N. A., Hsu, A. C., et al.: Effect of chronic renal failure, dialysis and transplantation on motor nerve conduction velocity in children. Can. Med. Assoc. J. 113:517, 1975. 20. Asbury, A. K., Gale, M. K., Cox, S. C., et al.: Giant axonal neuropathy: a unique case with segmental neurofilamentous masses. Acta Neuropathol. (Berl.) 20:237, 1972. 21. Atkin, S. L., Coady, A. M., Horton, D., et al.: Multiple cerebral haematomata and peripheral nerve palsies associated with a case of juvenile diabetic ketoacidosis. Diabet. Med. 12:267, 1995. 21a. Aulisa, L., Tamburrelli, F., Padua, R., et al.: Intraneural cyst of the peroneal nerve. Childs Nerv. Syst. 14:222, 1998. 22. Awerbuch, G. I., Nigro, M. A., Dabrowski, E., and Levin, J. R.: Childhood lumbosacral plexus neuropathy. Pediatr. Neurol. 5:314, 1989. 23. Baer, R. D., and Johnson, E. W.: Motor nerve conduction velocities in normal children. Arch. Phys. Med. Rehabil. 46:698, 1965. 24. Bager, B.: Perinatally acquired brachial plexus palsy— a persisting challenge. Acta Paediatr. 86:1214, 1997. 25. Bailey, A. A., Sayre, G. P., and Clark, E. C.: Neuritis associated with systemic lupus erythematosus. Arch. Neurol. Psychiatry 75:251, 1956.

2739

25a. Barkai, L., Kempler, P., Vamosi, I., et al.: Peripheral sensory nerve dysfunction in children and adolescents with type 1 diabetes mellitus. Diabet. Med. 15:228, 1998. 26. Bakshi, N., Maselli, R., Gospe, S. M., et al.: Fulminating demyelinating neuropathy mimicking cerebral death. Muscle Nerve 20:1595, 1997. 27. Barquet, A.: Traumatic anterior dislocation of the hip in childhood. Injury 13:435, 1982. 28. Bassett, G. S., Monforte-Munoz, H., Mitchell, W. G., and Rowland J. M.: Cavus deformity of the foot secondary to a neuromuscular choristoma (hamartoma) of the sciatic nerve: a case report. J. Bone Joint Surg. Am. 79:1398, 1997. 29. Baxter, R. V., Ben Othmane, K., Rochelle, J. M., et al.: Ganglioside-induced differentiation-associated protein-1 is mutant in Charcot-Marie-Tooth disease type 4A/8q21. Nat. Genet. 30:21, 2002. 30. Beausoleil, J. L., Nordgren, R. E., and Modlin, J. F.: Vaccine-associated paralytic poliomyelitis. J. Child Neurol. 9:334, 1994. 30a. Beck, T. D. Jr., Miller, K. E., and Kruse, R. W.: An unusual presentation of intoeing in a child. J. Am. Osteopath. Assoc. 98:48, 1998. 31. Beghi, E., Kurland, L. T., Mulder, D. W., et al.: GuillainBarré syndrome: clinicoepidemiologic features and effect of influenza vaccine. Arch. Neurol. 42:1053, 1985. 32. Behse, F., Buchthal, F., Carlsen, F., and Knappeis, G. G.: Hereditary neuropathy with liability to pressure palsies: electrophysiological and histopathological aspects. Brain 95:777, 1972. 33. Bejaoui, K., Wu, C., Scheffler, M. D., et al.: SPTLC1 is mutated in hereditary sensory neuropathy, type 1. Nat. Genet. 27:261, 2001. 34. Bell, W. E.: Acquired immobility of the right arm. Semin. Pediatr. Neurol. 3:173, 1996. 35. Belman, A. L., Iyer, M., Coyle, P. K., and Dattwyler, R.: Neurologic manifestations in children with North American Lyme disease. Neurology 43:2609, 1993. 36. Belman, A. L., Leicher, C. R., Moshe, S. L., and Mezey, A. P.: Neurologic manifestations of Schoenlein-Henoch purpura: report of three cases and review of the literature. Pediatrics 75:687, 1985. 37. Berciano, J., García, A., and Combarros, O.: Initial semeiology in children with Charcot-Marie-Tooth disease 1A duplication. Muscle Nerve 27:34, 2003. 38. Berg, B. O., Rosenberg, S. H., and Asbury, A. K.: Giant axonal neuropathy. Pediatrics 49:894, 1972. 39. Berginer, V. M., Salen, G., and Shefer, S.: Long term treatment of cerebrotendinous xanthomatosis with chenodeoxycholic acid. N. Engl. J. Med. 311:1649, 1984. 40. Beydoun, S. R., Engel, W. K., Karofsky, P., and Schwartz, M. U.: Long-term plasmapheresis therapy is effective and safe in children with chronic relapsing dysimmune polyneuropathy. Rev. Neurol. 146:123, 1990. 41. Bickels, J., Kahanovitz, N., Rubert, C. K., et al.: Extraspinal bone and soft-tissue tumors as a cause of sciatica. Clinical diagnosis and recommendations: analysis of 32 cases. Spine 24:1611, 1999. 42. Bilbao, J. M., Berry, H., Marotta, J., and Ross, R. C.: Peripheral neuropathy in oxalosis: a case report with electron microscopic observations. Can. J. Neurol. Sci. 3:63, 1976.

2740

Diseases of the Peripheral Nervous System

42a. Bilge, T., Kaya, A., Alatli, M., et al.: Hemangioma of the peroneal nerve: case report and review of the literature. Neurosurgery 25:649, 1989. 43. Birch, R.: Brachial plexus injuries. J. Bone Joint Surg. Br. 78:986, 1996. 43a. Birch, R., and Achan, P.: Peripheral nerve repairs and their results in children. Hand. Clin. 16:579, 2000. 44. Birouk, N., Gouider, R., LeGuern, E., et al.: Charcot-MarieTooth disease type 1A with 17p11.2 duplication: clinical and electrophysiological phenotype study and factors influencing disease severity in 119 cases. Brain 120:813, 1997. 45. Birouk, N., LeGuern, E., Maisonobe, T., et al.: X-linked Charcot-Marie-Tooth disease with connexin 32 mutations: clinical and electrophysiologic study. Neurology 50:1074, 1998. 46. Björkhem, I., and Bober, K. M.: Inborn errors in bile acid biosynthesis and storage of sterols other than cholesterol. In Scriver, C., Beaudet, A. L., Sly, W. S., and Valle, D. (eds.): The Metabolic and Molecular Basis of Inherited Disease, 7th ed. New York, McGraw-Hill, p. 2273, 1995. 47. Boder, E., and Sedgwick, P.: Ataxia telangiectasia: a review of 101 cases. Dev. Med. 8:110, 1963. 48. Boerkoel, C. F., Takshima, H., Garcia, C. A., et al.: CharcotMarie-Tooth disease and related neuropathies: mutation distribution and genotype-phenotype correlation. Ann. Neurol. 51:190, 2002. 49. Bolton, C. F.: Polyneuropathies. In Jones, H. R., Bolton, C. F., and Harper, C. M. (eds.): Pediatric Clinical Electromyography. Philadelphia, Lippincott–Raven, p. 251, 1996. 50. Bomont, P., Cavalier, L., Blondeau, F., et al.: The gene encoding gigaxonin, a new member of the cytoskeletal BTB/kelch repeat family, is mutated in giant axonal neuropathy. Nat. Genet. 26:370, 2000. 51. Bradshaw, D. Y., and Jones, H. R.: “Pseudo-meningoencephalopathic” presentation of pediatric Guillain-Barre syndrome. J. Child Neurol. 16:505, 2001. 52. Bradshaw, D. Y., and Jones, H. R. Jr.: Guillain-Barré syndrome in children: clinical course, electrodiagnosis, and prognosis. Muscle Nerve 15:500, 1992. 53. Brady, R. O., Gal, A. E., Bradley, R. M., et al.: Enzymatic defect in Fabry’s disease: ceramidetrihexosidase deficiency. N. Engl. J. Med. 276:1163, 1967. 54. Brooks-Wilson, A., Marcil, M., Clee, S. M., et al.: Mutations in ABC1 in Tangier disease and familial high density lipoprotein deficiency. Nat. Genet. 22:336, 1999. 55. Brown, J. C., and Johns, R. J.: Nerve conduction in familial dysautonomia (Riley-Day) syndrome. JAMA 201:200, 1967. 56. Brown, K. L.: Review of obstetrical palsies: nonoperative treatment. Clin. Plast. Surg. 11:181, 1984. 57. Brown, T., Cupido, C., Scarfone, H., et al.: Developmental apraxia arising from neonatal brachial plexus palsy. Neurology 55:24, 2000. 58. Buchthal, F., and Behse, F.: Peroneal muscular atrophy (PMA) and related disorders. I. Clinical manifestations as related to biopsy findings, nerve conduction and electromyography. Brain 100:41, 1977. 59. Burns, T. M., Dyck, P. J., Darras, B. T., and Jones, H. R.: Chronic inflammatory demyelinating polyradiculoneuropathy. In Jones, H. R., DeVivo, D. C., and Darras, B. T. (eds.):

60.

61. 62. 63.

64.

65.

66.

67.

68.

69. 70.

71.

72.

73.

74.

74a. 75. 76. 77. 78.

Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, ButterworthHeinemann, p. 445, 2003. Burns, T. M., Ryan, M. R., Darras, B. T., and Jones, H. R.: Current therapeutic strategies for patients with polyneuropathies secondary to inherited metabolic disorders. Mayo Clin. Proc. 78:858, 2003. Burns, T. M., Shneker, B. F., and Juel, V. C.: Gasoline sniffing multifocal neuropathy. Pediatr. Neurol. 25:419, 2001. Byers, R. K., and Taft, L. T.: Chronic multiple periphery neuropathy in childhood. Pediatrics 20:517, 1957. Cai, F., and Zhang, J.: Study of nerve conduction and late responses in normal Chinese infants, children and adults. J. Child Neurol. 12:13, 1997. Cameron, C., Kang, P. B., Burns, T. M., et al: Multifocal slowing of nerve conduction in metachromatic leukodystrophy. Muscle Nerve 29:531, 2004. Campuzano, V., Montermini, L., Molto, M. D., et al.: Friedreich ataxia: autosomal recessive disease caused by an intronic GAA triplet repeat expansion. Science 271:1423, 1996. Carter, G. T., McDonald, C. M., Chan, T. T., and Margherita, A. J.: Isolated femoral mononeuropathy to the vastus lateralis: EMG and MRI findings. Muscle Nerve 18:341, 1995. Cavanagh, N. P., Yates, D. A., and Sutcliffe, J.: Thenar hypoplasia with associated radiologic abnormalities. Muscle Nerve 2:431, 1979. Cerra, D., and Johnson, E. W.: Motor nerve conduction velocity in premature infants. Arch. Phys. Med. Rehabil. 43:160, 1962. Chad, D. A., and Bradley, W. G.: Lumbosacral plexopathy. Semin. Neurol. 7:97, 1987. Challenor, Y. B., Richter, R. W., Bruun, B., and Pearson, J.: Nontraumatic plexitis and heroin addiction. JAMA 225:958, 1973. Chance, P. F., Alderson, M. K., Leppig, K. A., et al.: DNA deletion associated with hereditary neuropathy with liability to pressure palsies. Cell 72:143, 1993. Chance, P. F., Lensch, M. W., Lipe, H., et al.: Hereditary neuralgic amyotrophy and hereditary neuropathy with liability to pressure palsies: two distinct genetic disorders. Neurology 44:2253, 1994. Chaudhry, V., Chaudhry, M., Crawford, T. O., et al.: Toxic neuropathy in patients with pre-existing neuropathy. Neurology 60:337, 2003. Chaudhry, V., and Cornblath, D. R.: Wallerian degeneration in human nerves: serial electrophysiological studies. Muscle Nerve 15:687, 1992. Choe, W.: Leprosy presenting as unilateral foot drop in an immigrant boy. Postgrad. Med. J. 70:111, 1994. Clarke, H. M., and Curtis, C. G.: An approach to obstetrical brachial plexus injuries. Hand Clin. 11:563, 1995. Clay, S. A.: Osteomyelitis as a cause of brachial plexus neuropathy. Am. J. Dis. Child. 136:1054, 1982. Coker, S. B.: Leigh disease presenting as Guillain-Barré syndrome. Pediatr. Neurol. 9:61, 1993. Colan, R. V., Snead, O. C. 3rd, Oh, S. J., and Benton, J. W.: Steroid-responsive polyneuropathy with subacute onset in childhood. J. Pediatr. 97:374, 1980.

Peripheral Neuropathies in Infants and Children 78a. Conforti, R., Scuotto, A., Muras, I., et al.: Les hernies discales de adolescents. [Herniated disk adolescents.] J. Neuroradiol. 20:60, 1993. 79. Cook, J. D., Delgado, M. R., and Soutter-Glass, D.: Treatment of childhood autoimmune polyneuropathy: gamma globulin. Neurology 37:253, 1987. 80. Cornblath, D. R., Sladky, J. T., and Sumner, A. J.: Clinical electrophysiology of infantile botulism. Muscle Nerve 6:448, 1983. 81. Cornwall, R., and Radomisli, T. E.: Nerve injury in traumatic dislocation of the hip. Clin. Orthop. 377:84, 2000. 82. Cossée, M., Dürr, A., Schmitt, M., et al.: Friedreich’s ataxia: point mutations and clinical presentations of compound heterozygotes. Ann. Neurol. 45:200, 1999. 83. Cossée, M., Schmitt, M., Campuzano, V., et al.: Evolution of the Friedreich’s ataxia trinucleotide repeat expansion: founder effect and premutations. Proc. Natl. Acad. Sci. U. S. A. 94:7452, 1997. 84. Craig, W. S., and Clark, J. M. P.: Of peripheral nerve palsies in the newborn. J. Obstet. Gynecol. Br. Empire 65:229, 1958. 85. Crawford, T. O.: From enigmatic to problematic: the new molecular genetics of childhood spinal muscular atrophy. Neurology 46:335, 1996. 86. Crino, P. B., Grossman, R. I., and Rostami, A.: Magnetic resonance imaging of the cauda equina in chronic inflammatory demyelinating polyneuropathy. Ann. Neurol. 33:311, 1993. 87. Crino, P. B., Zimmerman, R., Laskowitz, D., et al.: Magnetic resonance imaging of the cauda equina in Guillain-Barré syndrome. Neurology 44:1334, 1994. 88. Crumrine, P. K., Koenigsberger, M. R., and Chutorian, A. M.: Footdrop in the neonate with neurologic and electrophysiologic data. J. Pediatr. 86:779, 1975. 89. Cuesta, A., Pedrola, L., Sevilla, T., et al.: The gene encoding ganglioside-induced differentiation-associated protein 1 is mutated in axonal Charcot-Marie-Tooth type 4A disease. Nat. Genet. 30:22, 2002. 90. Curtis, P. H. Jr., and Tucker, H. J.: Sciatic palsy in infants and prevention of impairment resulting from intragluteal injections JAMA 174:1586, 1960. 90a. Daghino, W., Pasquali, M., and Faletti, C.: Superficial peroneal nerve entrapment in a young athlete: the diagnostic contribution of magnetic resonance imaging. J. Foot Ankle Surg. 36:170, 1997. 91. Dantes, M., and McComas, A.: The extent and time course of motor neuron involvement in amyotrophic lateral sclerosis. Muscle Nerve 14:416, 1991. 92. Darras, B. T., and Jones, H. R.: Diagnosis of pediatric neuromuscular disorders in the era of DNA analysis. Pediatr. Neurol. 23:289, 2000. 93. DaSilveira, C. M., Salisbury, D. M., and de Quadros, C. A.: Measles vaccination and Guillain-Barre syndrome. Lancet 349:14, 1997. 94. David, W. S., and Doyle, J. J.: Acute infantile weakness: a case of vaccine associated poliomyelitis. Muscle Nerve 20:747, 1997. 95. David, W. S., and Jones, H. R.: Electromyography and biopsy correlation with suggested protocol for evaluation of the floppy infant. Muscle Nerve 17:424, 1994.

2741

96. David, W. S., and Jones, H. R.: Soft tissue sarcoma presenting as atypical extremity pain. Muscle Nerve 17:1071, 1994. 97. Dawkins, J. L., Hulme, D. J., Brahmbhatt, S. B., et al.: Mutations in SPTLC1, encoding serine palmitoyltransferase, long chain base subunit-1, cause hereditary sensory neuropathy type 1. Nat. Genet. 27:309, 2001. 97a. Deathe, A. B.: Hematometra as a cause of lumbar radiculopathy: a case report. Spine 18:1920, 1993. 98. De Jonghe, P., Timmerman, V., Ceuterick, C., et al.: The Thr124Met mutation in the peripheral myelin protein zero (MPZ) gene is associated with a clinically distinct Charcot-Marie-Tooth phenotype. Brain 122:281, 1999. 99. De Jonghe, P., Timmerman, V., Nelis, E., et al.: CharcotMarie-Tooth disease and related peripheral neuropathies. J. Peripher. Nerv. Syst. 2:370, 1997. 100. De Michele, G., Filla, A., Barbieri, F., et al.: Late onset recessive ataxia with Friedreich’s disease phenotype. J. Neurol. Neurosurg. Psychiatry 52:1398, 1989. 101. De Sandre-Giovannoli, A., Chaouch, M., Kozlov, S., et al.: Homozygous defects in LMNA, encoding lamin A/C nuclear-envelope proteins, cause autosomal recessive axonal neuropathy in human (Charcot-Marie-Tooth disorder type 2) and mouse. Am. J. Hum. Genet. 70:726, 2002. 102. DeOrio, J. K., and Bianco, A. J. Jr.: Lumbar disc excision in children and adolescents. J. Bone Joint Surg. Am. 64:991, 1982. 103. DeVivo, D. C., and Engel, W. K.: Remarkable recovery of a steroid-responsive recurrent polyneuropathy. J. Neurol. Neurosurg. Psychiatry 33:62, 1970. 104. Deymeer, F., and Jones, H. R.: Pediatric median mononeuropathies: a clinical and electromyographic study. Muscle Nerve 17:755, 1994. 105. Dodds, S. D., and Wolfe, S. W.: Perinatal brachial plexus palsy. Curr. Opin. Pediatr. 12:40, 2000. 106. Donaghy, M., King, R. H. M., McKeran, R. O., et al.: Cerebrotendinous xanthomatosis: clinical, electrophysiological and nerve biopsy findings, and response to treatment with chenodeoxycholic acid. J. Neurol. 237:216, 1990. 107. Dorfman, L. J., Ransom, B. R., Forno, L. S., and Kelts, A.: Neuropathy in the hypereosinophilic syndrome. Muscle Nerve 6:291, 1983. 108. Doss, M., Schneider, J., Von Tiepermann, R., and Brandt, A.: New type of acute porphyria with porphobilinogen synthase (delta-aminolevulinic acid dehydratase) defect in the homozygous state. Clin. Biochem. 15:52, 1982. 109. Draaisma, J. M., Fiselier, T. J., and Mullaart, R. A.: Mononeuritis multiplex in a child with cutaneous polyarteritis. Neuropediatrics 23:28, 1992. 110. Duarte, J., Cruz Martinez, A., Rodriguez, F., et al.: Hypertrophy of multiple cranial nerves and spinal roots in chronic inflammatory demyelinating neuropathy. J. Neurol. Neurosurg. Psychiatry 67:685, 1999. 111. Dubourg, O., Tardieu, S., Birouk, N., et al.: Clinical, electrophysiological and molecular genetic characteristics of 93 patients with X-linked Charcot-Marie-Tooth disease. Brain 124:1958, 2001. 112. Duck, S. C., Wei, F., Parke, J., and Swick, H. M.: Role of height and glycosylated hemoglobin in abnormal nerve conduction in pediatric patients with type I diabetes mellitus after 4–9 yr of disease. Diabetes Care 14:386, 1991.

2742

Diseases of the Peripheral Nervous System

113. Dunn, H. G.: Nerve conduction studies in children with Friedreich’s ataxia and ataxia-telangiectasia. Dev. Med. Child Neurol. 15:324, 1973. 114. Dunn, H. G., Daube, J. R., and Gomez, M. R.: Heredofamilial branchial plexus neuropathy (hereditary neuralgic amyotrophy with branchial predilection) in childhood. Dev. Med. Child Neurol. 20:28, 1978. 115. Dunn, H. G., Lake, B. D., Dolman, C. L., and Wilson, J.: The neuropathy of Krabbe’s infantile cerebral sclerosis (globoid cell leucodystrophy). Brain 2:329, 1969. 116. Dunne, J. W., Silbert, P. L., and Wren, M.: A prospective study of acute radiculopathy after scoliosis surgery. Clin. Exp. Neurol. 28:180, 1991. 117. Durr, A., Cossee, M., Agid, Y., et al.: Clinical and genetic abnormalities in patients with Friedreich’s ataxia. N. Engl. J. Med. 335:1169, 1996. 118. Dyck, P. J., Karnes, J. L., and Lambert, E. H.: Longitudinal study of neuropathic deficit and nerve conduction abnormalities in hereditary motor and sensory neuropathy type 1. Neurology 39:1302, 1989. 119. Dyck, P. J., Lais, A. C., Ohta, M., et al.: Chronic inflammatory polyradiculoneuropathy. Mayo Clin. Proc. 50:621, 1975. 120. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. I. Neurologic, genetic and electrophysiologic findings in hereditary polyneuropathies. Arch. Neurol. 18:603, 1968. 121. Dyck, P. J., and Lambert, E. H.: Lower motor and primary sensory neuron diseases with peroneal muscular atrophy. II. Neurologic, genetic and electrophysiologic findings in various neuronal degenerations. Arch. Neurol. 18:619, 1968. 122. Dyck, P. J., Litchy, W. J., Minnerath, S., et al.: Hereditary motor and sensory neuropathy with diaphragm and vocal cord paresis. Ann. Neurol. 35:608, 1994. 123. Dyck, P. J., Mellinger, J. F., Reagan, T. J., et al.: Not “indifference to pain” but varieties of hereditary sensory and autonomic neuropathy. Brain 106:373, 1983. 123a. Edlich, H. S., Fariss, B. L., Phillips, V. A., et al.: Talotibial exostoses with entrapment of the deep peroneal nerve. J. Emerg. Med. 5:109, 1987. 124. Eeg-Olofsson, O., and Peterson, I.: Childhood diabetic neuropathy: a clinical and neurophysiological study. Acta Paediatr. Scand. 55:163, 1966. 125. Egel, R. T., Cueva, J. P., and Adair, R. L.: Posttraumatic childhood lumbosacral plexus neuropathy. Pediatr. Neurol. 12:62, 1995. 126. Egloff, D. V., Raffoul, W., Bonnard, C., and Stalder, J.: Palliative surgical procedures to restore shoulder function in obstetric brachial palsy: critical analysis of Narakas’ series. Hand Clin. 11:597, 1995. 127. Ehrmann, L., Lechner, K., Mamoli, B., et al.: Peripheral nerve lesions in haemophilia. J. Neurol. 225:175, 1981. 127a. el Bahri-Ben Mrad, F., Gouider, R., Fredj, M., et al.: Childhood diabetic neuropathy: a clinical and electrophysiological study. Funct. Neurol. 15:35, 2000. 128. Elder, G. H.: Hepatic porphyrias in children. J. Inherit. Metab. Dis. 20:237, 1997. 129. Elder, G. H., and Hift, R. J.: Treatment of acute porphyria. Hosp. Med. 62:422, 2001. 130. Elias, E., Muller, D. P. R., and Scott, J.: Association of spinocerebellar disorders with cystic fibrosis or chronic

131.

132.

133.

134. 135.

136.

137.

138.

139.

140.

141.

142.

143. 144.

145.

146. 147. 148.

childhood cholestasis and very low serum vitamin E. Lancet 2:1319, 1981. Emery, A. H.: Population frequencies of inherited neuromuscular diseases—a world survey. Neuromuscul. Disord. 1:19, 1991. Emmert, S., Schneider, T. D., Khan, S. G., and Kraemer, K. H.: The human XPG gene: gene architecture, alternative splicing and single nucleotide polymorphisms. Nucleic Acids Res. 29:1443, 2001. Eng, C. M., Guffon, N., Wilcox, W. R., et al., for the International Collaborative Fabry Disease Study Group: Safety and efficacy of recombinant human alpha galactosidase A replacement therapy in Fabry’s disease. N. Engl. J. Med. 345:9, 2001. Eng, G. D.: Brachial plexus palsy in newborn infants. Pediatrics 48:18, 1971. Eng, G. D., Binder, H., Getson, P., and O’Donnell, R.: Obstetrical brachial plexus palsy (OBPP) outcome with conservative management. Muscle Nerve 19:884, 1996. Eng, G. D., Koch, B., and Smokvia, M.: Brachial plexus palsy in neonates and children. Arch. Phys. Med. Rehabil. 59:458, 1978. Engstrom, J. W., Layzer, R. B., Olney, R. K., and Edwards, M. B.: Idiopathic, progressive mononeuropathy in young people. Arch. Neurol. 50:20, 1993. Epstein, J. A., Epstein, N. E., Marc, J., et al.: Lumbar intervertebral disk herniation in teenage children: recognition and management of associated anomalies. Spine 9:427, 1984. Epstein, M. A., and Sladky, J. T.: The role of plasmapheresis in childhood Guillain-Barré syndrome. Ann. Neurol. 28:65, 1990. Escolar, D., Ryan, M., and Jones, H. R.: Lower extremity mononeuropathies. In Jones, H. R., De Vivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 321, 2003. Escolar, D. M., and Jones, H. R.: Pediatric sciatic mononeuropathies: a clinical and electromyographic analysis [abstract]. Muscle Nerve 17:108, 1994. Escolar, D. M., and Jones, H. R.: Pediatric radial mononeuropathies: a clinical and electromyographic study of sixteen children with review of the literature. Muscle Nerve 19:876, 1996. Evans, O. B.: Polyneuropathy in childhood. Pediatrics 64:96, 1979. Evidente, V. G. H., Gwinn-Hardy, K. A., Cavinees, J. N., and Gilman, S.: Hereditary ataxias. Mayo Clin. Proc. 75:475, 2000. Eymard, B., Penicaud, A., Leger, J. M., et al.: Peripheral nerve in mitochondrial disease: clinical and electrophysiological data. A study of 28 cases. Rev. Neurol. 147:508, 1991. Fearn, C. B., and Goodfellow, J. W.: Anterior interosseous nerve palsy. J. Bone Joint Surg. 47:91, 1965. Feldman, G. V.: Radial nerve palsies in the newborn. Arch. Dis. Child. 32:469, 1957. Felice, K. J., and Jones, H. R.: Pediatric ulnar mononeuropathy: report of 21 electromyography-documented cases and review of the literature. J. Child Neurol. 11:116, 1996.

Peripheral Neuropathies in Infants and Children 148a. Felice, K. J., Leicher, C. R., and DiMario, F. J. Jr.: Hereditary neuropathy with liability to pressure palsies in children. Pediatr. Neurol. 21:818, 1999. 148b. Felice, K. J., and Jones, H. R.: Upper extremity mononeuropathies in infants and children. In Jones, H. R., De Vivo, D. C., and Darras, D. F. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence. Boston, Butterworth-Heinemann, p. 321, 2003. 149. Fernandez, J. M., Davalos, A., Ferrer, I., et al.: Acute sensory neuronopathy: report of a child with remarkable clinical recovery. Neurology 44:762, 1994. 150. Ferriere, G., Guzzetta, F., Kulakoski, S., and Evrard, P.: Nonprogressive type II hereditary sensory and autonomic neuropathy: a homogeneous clinicopathologic entity. J. Child Neurol. 7:364, 1992. 151. Filla, A., De Michele, G., Cavalcanti, F., et al.: The relationship between trinucleotide (GAA) repeat length and clinical features in Friedreich ataxia. Am. J. Hum. Genet. 59:554, 1996. 152. Filler, A. G., Kliot, M., Howe, F. A., et al.: Application of magnetic resonance neurography in the evaluation of patients with peripheral nerve pathology. J. Neurosurg. 85:299, 1996. 153. Fischer, A. Q., and Strasburger, J.: Footdrop in the neonate secondary to use of footboards. J. Pediatr. 101:1003, 1982. 154. Fisher, R. G., and Saunders, R. L.: Lumbar disc protrusions in children. J. Neurosurg. 54:480, 1981. 155. Flanigan, K. M., Crawford, T. O., Griffin, J. W., et al.: Localization of the giant axonal neuropathy gene to chromosome 16q24. Ann. Neurol. 43:143, 1998. 156. Fleming, F. J., Chaitow, J., Jones, H. R. Jr., et al.: Thalidomide neuropathy in childhood. Submitted for publication. 157. Fok, T. F., Ha, M. H., Leung, K. W., and Wong, W.: Sciatic nerve palsy complicating umbilical arterial catheterization. Eur. J. Pediatr. 145:308, 1986. 157a. Foo, C. L., and Swann, M.: Isolated paralysis of the serratus anterior: a report of 20 cases. J. Bone Joint. Surg. [Br] 65:552, 1983. 158. Ford, F. R.: Diseases of the Nervous System in Infancy, Childhood and Adolescence, 5th ed. Springfield, Charles C Thomas, pp. 765–767, 1966. 159. Ford, F. R.: Diseases of the Nervous System in Infancy, Childhood and Adolescence, 5th ed. Springfield, Charles C Thomas, p. 825, 1966. 160. Gabreels-Festen, A. A., Gabreels, F. J., Joosten, E. M., et al.: Hereditary neuropathy with liability to pressure palsies in childhood. Neuropediatrics 23:138, 1992. 161. Galan, E., and Kousseff, B. G.: Peripheral neuropathy in Ehlers-Danlos syndrome. Pediatr. Neurol. 12:242, 1995. 162. Galassi, R., Passaglia, P., Cirignotti, F., and Lugaresi, E.: Peripheral neuropathy due to gasoline sniffing. A case report. Eur. Neurol. 19:419, 1980. 163. Gallai, V., Firenze, C., Mazzota, G., and Del Gatto, F.: Neuropathy in children and adolescents with diabetes mellitus. Acta Neurol. Scand. 78:136, 1988. 164. Galloway, G., Giuliani, M. J., Burns, D. K., and Lacomis, D.: Neuropathy associated with hyperoxaluria: improvement after combined renal and liver transplantations. Brain Pathol. 8:247, 1998.

2743

165. Gambarelli, D., Hassoun, J., Pellissier, J. F., et al.: Giant axonal neuropathy: involvement of peripheral nerve, myenteric plexus and extraneuronal area. Acta Neuropathol. (Berl.) 39:261, 1977. 166. Gamstorp, I.: Normal conduction velocity of ulnar, median and peroneal nerves in infancy, childhood and adolescence. Acta Paediatr. Scand. 146:68, 1963. 167. Gamstorp, I.: Polyneuropathy in childhood. Acta Paediatr. Scand. 57:230, 1968. 168. Gamstorp, I., Shelburne, S. A. Jr., Engleson, G., et al.: Peripheral neuropathy in juvenile diabetics. Diabetes 15:411, 1966. 168a. Garcia, F. F., Semba, C. P., and Sartoris, D. J.: Diagnostic imaging of childhood spinal infection. Orthop. Rev. 22:321, 1993. 169. Giannini, F., Volpi, N., Rossi, S., et al.: Thalidomide-induced neuropathy: a ganglionopathy? Neurology 60:877, 2003. 170. Gibbs, T. C., Payan, J., Brett, E. M., et al.: Peripheral neuropathy as the presenting feature of tyrosinaemia type I and effectively treated with an inhibitor of 4-hydroxyphenylpyruvate dioxygenase. J. Neurol. Neurosurg. Psychiatry 56:1129, 1993. 171. Gieron-Korthals, M. A., Westberry, K. R., and Emmanuel, P. J.: Acute childhood ataxia: 10 year experience. J. Child Neurol. 9:381, 1994. 172. Gieselmann, A., Polten, J., Kreysing, J., and von Figura, K.: Molecular genetics of metachromatic leukodystrophy. J. Inherit. Metab. Dis. 17:500, 1994. 173. Gilbert, A.: Long-term evaluation of brachial plexus surgery in obstetrical palsy. Hand Clin. 11:583, 1995. 174. Gilbert, A., Brockman, R., and Carlioz, H.: Surgical treatment of brachial plexus birth palsy. Clin. Orthop. 264:39, 1991. 175. Gilbert, W. M., Nesbitt, T. S., and Danielsen, B.: Associated factors in 1611 cases of brachial plexus injury. Obstet. Gynecol. 93:536, 1999. 176. Giles, F. H., and Matson, D. D.: Sciatic nerve injury following misplaced intragluteal injections. J. Pediatr. 76:247, 1970. 177. Gilliatt, R. W., Willison, R. G., Dietz, V., and Williams, I. R.: Peripheral nerve conduction in patients with a cervical rib and band. Ann. Neurol. 4:124, 1978. 178. Ginsberg, L., Platts, A. D., and Thomas, P. K.: Chronic inflammatory demyelinating polyneuropathy mimicking a lumbar spinal stenosis syndrome. J. Neurol. Neurosurg. Psychiatry 59:189, 1995. 179. Goebel, H. H., Veit, S., and Dyck, P. J.: Confirmation of virtual unmyelinated fiber absence in hereditary sensory neuropathy type IV. J. Neuropathol. Exp. Neurol. 39:670, 1980. 180. Gold, A. P., and Yahr, M. D.: Childhood lupus erythematosus. Trans. Am. Neurol. Assoc. 85:96, 1960. 181. Goldberg, M., and Chitanondh, H.: Polyneuritis with albuminocytologic dissociation in the spinal fluid in systemic lupus erythematosus: report of a case, with review of pertinent literature. Am. J. Med. 27:342, 1959. 182. Goldberg, V. M., and Jacobs, B.: Osteoid osteoma of the hip in children. Clin. Orthop. 106:41, 1975. 183. Gonik, B., McCormick, E. M., Verweij, B. H., et al.: The timing of congenital brachial plexus injury: a study

2744

184.

185.

186.

187.

187a.

188.

189.

190.

191.

192.

193.

194.

195. 196. 197.

198.

199.

200.

Diseases of the Peripheral Nervous System of electromyography findings in the newborn piglet. Am. J. Obstet. Gynecol. 178:688, 1998. Gordon, M., Rich, H., Deutschberger, J., and Green, M.: The immediate and long-term outcome of obstetric birth trauma. I. Brachial plexus paralysis. Am. J. Obstet. Gynecol. 117:51, 1973. Gouider, R., LeGuren, E., Gugenheim, M., et al.: Clinical, electrophysiologic, and molecular correlations in 13 families with hereditary neuropathy with liability to pressure palsies and a chromosome 17p11.2 deletion. Neurology 45:2018, 1995. Grant, I. A., O’Brien, P., and Dyck, P.J.: Neuropathy tests and normative results. In Dyck, P. J., and Thomas, P. K. (eds.): Diabetic Neuropathy, 2nd ed. Philadelphia, W. B. Saunders, p. 123, 1999. Grattan-Smith, P. J., Morris, J. G., Johnston, H. M., et al.: Clinical and neurophysiological features of tick paralysis. Brain 120:1975, 1997. Gregg, J. R., Labosky, D., Harty, M., et al.: Serratus anterior paralysis in the young athlete. J. Bone Joint. Surg. [Am] 61:825, 1979. Griesemer, D. A., and Johnson, M. I.: Guillain-Barré syndrome and plasmapheresis in childhood. Ann. Neurol. 29:688, 1991. Grohmann, K., Schuelke, M., Diers, A., et al.: Mutations in the gene encoding immunoglobulin mu-binding protein 2 cause spinal muscular atrophy with respiratory distress type 1. Nat. Genet. 29:75, 2001. Grohmann, K., Wienker, T. F., Saar, K., et al.: Diaphragmatic spinal muscular atrophy with respiratory distress is heterogeneous, and one form is linked to chromosome 11q13-q21. Am. J. Hum. Genet. 65:1459, 1999. Grossman, J. A.: Early operative intervention for birth injuries to the brachial plexus. Semin. Pediatr. Neurol. 7:36, 2000. Gucuyener, K., Keskil, S., Baykaner, M. K., et al.: Co-incidence of Guillain-Barré syndrome and spinal cord compression in non-Hodgkins lymphoma. Neuropediatrics 24:36, 1994. Guillevin, L., Cohen, P., Gayraud, M., et al.: ChurgStrauss syndrome : clinical study and long-term follow-up of 96 patients. Medicine 78:26, 1999. Haddad, F. S., Jones, D. H. A., Vellodi, A., et al.: Carpal tunnel syndrome in the mucopolysaccharidoses and mucolipidoses. J. Bone Joint Surg. Br. 79:576, 1997. Hagberg, B.: The clinical diagnosis of Krabbe’s infantile leucodystrophy. Acta Paediatr. Scand. 52:213, 1963. Hagberg, B.: Krabbe’s disease: clinical presentation of neurological variants. Neuropediatrics 15(Suppl.):S11, 1984. Hagberg, B., Sourander, P., and Thoren, L.: Peripheral nerve changes in the diagnosis of metachromatic leucodystrophy. Acta Paediatr. Suppl. 135:63, 1962. Hagberg, B., and Westerberg, B.: The nosology of genetic peripheral neuropathies in Swedish children. Dev. Med. Child Neurol. 25:3, 1983. Hagberg, B. A., Blennow, G., Kristiansson, B., and Stibler, H.: Carbohydrate-deficient glycoprotein syndromes: peculiar group of new disorders. Pediatr. Neurol. 9:255, 1993. Hahn, A. F.: Guillain-Barre syndrome. Lancet 352:635, 1998.

201. Hahn, A. F., Bolton, C. F., White, C. M., et al.: Genotypephenotype correlations in X-linked dominant CharcotMarie-Tooth disease. Ann. N. Y. Acad. Sci. 883:366, 1999. 202. Hall, B. M., Walsh, J. C., Horvath, J. S., and Lytton, D. G.: Peripheral neuropathy complicating primary hyperoxaluria. J. Neurol. Sci. 29:343, 1976. 203. Harding, A. E.: Friedreich’s ataxia: a clinical and genetic study of 90 families with an analysis of early diagnosis criteria and intrafamilial clustering of clinical features. Brain 104:589, 1981. 204. Harding, A. E., Matthews, S., Jones, S., et al.: Spinocerebellar degeneration associated with a selective defect of vitamin E absorption. N. Engl. J. Med. 313:32, 1985. 205. Harding, A. E., and Thomas, P. K.: The clinical features of hereditary motor and sensory neuropathy types I and II. Brain 103:259, 1980. 206. Harel, L., Mukamel, M., Brik, R., et al.: Peripheral neuropathy in pediatric systemic lupus erythematosus. Pediatr. Neurol. 27:53, 2002. 207. Harila-Saari, A. H., Vainionpaa, L. K., Kovala, T. T., et al.: Nerve lesions after therapy for childhood acute lymphoblastic leukemia. Cancer 82:200, 1998. 208. Hart, D. E., Rojas, L. A., Rosariio, J. A., et al.: Childhood Guillain-Barré syndrome in Paraguay, 1990 to 1991. Ann. Neurol. 36:859, 1994. 209. Hattori, N., Ichimura, M., Aoki, A., et al.: Clinicopathological features of chronic inflammatory demyelinating polyradiculoneuropathy in childhood. J. Neurol. Sci. 154:66, 1998. 210. Herskowitz, A., Ishii, N., and Schaumburg, H.: n-Hexane neuropathy: a syndrome occurring as a result of industrial exposure. N. Engl. J. Med. 285:82, 1971. 211. Hildebrandt, G., Holler, E., Woenkhaus, M., et al.: Acute deterioration of Charcot-Marie-Tooth disease IA (CMT IA) following 2 mg of vincristine chemotherapy. Ann. Oncol. 11:743, 2000. 212. Hirasawa, Y., and Sakakida, K.: Sports and peripheral nerve injury. Am. J. Sports Med. 11:420, 1983. 213. Hirayama, K., Tsubaki, T., Toyokura, Y., and Okinaka, S.: Juvenile muscular atrophy of unilateral upper extremity. Neurology 13:373, 1963. 214. Hoekstra, M. F.: Responses to DNA damage and regulation of cell cycle checkpoints by the ATM protein kinase family. Curr. Opin. Genet. Dev. 7:170, 1997. 215. Hoffman, W. H., Hart, Z. H., and Frank, R. N.: Correlates of delayed motor nerve conduction and retinopathy in juvenile-onset diabetes mellitus. J. Pediatr. 102:351, 1983. 216. Hope, E. E., Bodensteiner, J. B., and Thong, N.: Neonatal lumbar plexus injury. Arch. Neurol. 42:94, 1985. 217. Hughes, R., Bensa, S., Willison, H., et al.: Randomized controlled trial of intravenous immunoglobulin versus oral prednisolone in chronic inflammatory demyelinating polyradiculoneuropathy. Ann. Neurol. 50:195, 2001. 218. Igarashi, M., Thompson, E. I., and Rivera, G. K.: Vincristine neuropathy in type I and type II Charcot-Marie-Tooth disease (hereditary motor sensory neuropathy). Med. Pediatr. Oncol. 25:113, 1995. 219. Igisu, H., Ohta, M., Tabira, T., et al.: Giant axonal neuropathy: a clinical entity affecting the central as well as the peripheral nervous system. Neurology 25:717, 1975. 220. Indo, Y.: Molecular basis of congenital insensitivity to pain with anhidrosis (CIPA): mutations and polymorphisms in

Peripheral Neuropathies in Infants and Children

221.

222.

223.

224.

225.

226. 227.

228.

229. 229a.

230. 231.

232. 233.

234.

235.

236.

237. 238.

239.

TRKA (NTRK1) gene encoding the receptor tyrosine kinase for nerve growth factor. Hum. Mutat. 18:462, 2001. Indo, Y., Tsuruta, M., Hayashida, Y., et al.: Mutations in the TRKA/NGF receptor gene in patients with congenital insensitivity to pain with anhidrosis. Nat. Genet. 13:485, 1996. Infante, J., Garcia, A., Combarros, O., et al.: Diagnostic strategy for familial and sporadic cases of neuropathy associated with 17p11.2 deletion. Muscle Nerve 24:1149, 2001. Inoue, K., Shilo, K., Boerkoel, C. F., et al.: Congenital hypomyelinating neuropathy, central dysmyelination, and Waardenburg-Hirschsprung disease: phenotypes linked by SOX10 mutation. Ann. Neurol. 52:836, 2002. Inoue, M., Kawano, T., Matsumura, H., et al.: Solitary benign schwannoma of the brachial plexus. Surg. Neurol. 20:103, 1983. Ionasescu, V. V.: Charcot-Marie-Tooth neuropathies: from clinical description to molecular genetics. Muscle Nerve 18:267, 1995. Jackson, A. H., Baquis, G. D., and Shaw, B. L.: Congenital Guillain-Barre syndrome. J. Child Neurol. 5:407, 1996. Jansen, G. A., Ofman, R., Ferdinandusse, S., et al.: Refsum disease is caused by mutations in the phytanoyl-CoA hydroxylase gene. Nat. Genet. 17:190, 1997. Jansen, P. W., Perkin, R. M., and Ashwal, S.: Guillain-Barré syndrome in childhood: natural course and efficacy of plasmapheresis. Pediatr. Neurol. 9:16, 1993. Jellis, J. E., and Helal, B.: Childhood sciatic palsies: congenital and traumatic. Proc. R. Soc. Med. 63:655, 1970. Johnson, P. C., and Kline, D. G.: Localized hypertrophic neuropathy: possible focal perineurial barrier defect. Acta. Neuropathol. 77:514, 1989. Jones, H. R.: Compressive neuropathy in childhood: a report of 14 cases. Muscle Nerve 9:720, 1986. Jones, H. R.: EMG evaluation of the floppy infant: differential diagnosis and technical aspects. Muscle Nerve 13:338, 1990. Jones, H. R.: Guillain-Barré syndrome in children. Curr. Opin. Pediatr. 7:663, 1995. Jones, H. R.: Childhood Guillain-Barré syndrome; a review: clinical presentation, diagnosis and therapy. J. Child Neurol. 11:4, 1996. Jones, H. R.: Mononeuropathies. In Jones, H. R., Bolton, C. F., and Harper, C. M. (eds.): Pediatric Clinical Electromyography. Philadelphia, Lippincott–Raven, p. 171, 1996. Jones, H. R.: Plexus and nerve root lesions. In Jones, H. R., Bolton, C. F., and Harper, C. M. (eds.): Pediatric Clinical Electromyography. Philadelphia, Lippincott–Raven, p. 123, 1996. Jones, H. R.: Guillain-Barré syndrome: perspectives with infants and children. Semin. Pediatr. Neurol. 7:91, 2000. Jones, H. R.: Mononeuropathies of infancy and childhood. Clin. Neurophysiol. 53:396, 2000. Jones, H. R.: Infantile botulism and other acquired pediatric neuromuscular junction disorders of infancy and childhood. In Brown, W., Bolton, C. F., and Aminoff, M. (eds.): Neuromuscular Function and Disease. Philadelphia, W. B. Saunders, p. 1697, 2002. Jones, H. R.: Nerve root lesions. In Jones, H. R., De Vivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders

240. 241.

242.

243.

244.

245.

246.

247.

248.

249.

250.

251. 252. 253.

254.

255.

256. 257.

258.

2745

of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 233, 2003. Jones, H. R., and Darras, B. T.: Acute care pediatric electromyography. Muscle Nerve Suppl 9:S53, 2000. Jones, H. R., Felice, K. J., and Gross, P. T.: Pediatric peroneal mononeuropathy: a clinical and electromyographic study. Muscle Nerve 16:1167, 1993. Jones, H. R., Gianturco, L. E., Gross, P. T., and Buchhalter J.: Sciatic neuropathies in childhood: a report of ten cases and review of the literature. J. Child Neurol. 3:193, 1988. Jones, H. R., Harmon, R. L., Bolton, C. F., and Harper C. M.: An approach to pediatric electromyography. In Jones, H. R., Bolton, C. F., and Harper, C. M. (eds.): Pediatric Clinical Electromyography. Philadelphia, Lippincott–Raven, p. 1, 1996. Jones, H. R., Herbison, G. J., Jacob, S. R., et al.: Intrauterine onset of a mononeuropathy: peroneal neuropathy in a newborn. Muscle Nerve 19:88, 1996. Jones, H. R., and Hsu, P.: Pes cavus as the presenting sign of childhood neuropathies: EMG evaluation in 23 otherwise asymptomatic children [abstract]. Muscle Nerve 20:1070, 1997. Jones, H. R., Miller, T. A., and Wilbourn, A.: Brachial and lumbosacral plexus lesions. In Jones, H. R., De Vivo, D. C., and Darras, B. T. (eds): Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 245, 2003. Jones, H. R. Jr.: Pediatric case studies. In American Association of Electrodiagnostic Medicine (AAEM) Plenary Session: New Developments in Pediatric Neuromuscular Disease. Rochester, MN, AAEM, p. 51, 1993. Jones, H. R. Jr., and Bradshaw, D. Y.: Guillain-Barré syndrome and plasmapheresis in childhood (letter). Ann. Neurol. 29:688, 1991. Jones, M. Z., Nigro, M. A., and Barr, P. S.: Familial “giant axonal neuropathy” [abstract]. J. Neuropathol. Exp. Neurol. 38:324, 1979. Kanda, T., Oda, M., Yonezawa, M., et al.: Peripheral neuropathy in xeroderma pigmentosum. Brain 113:1025, 1990. Karani, V.: Peripheral neuritis after addiction to petrol. Br. Med. J. 1:216, 1966. Kasman, M., Bernstein, L., and Schulman, S.: Chronic polyradiculoneuropathy of infancy. Neurology 26:565, 1976. Katirji, B., and Wilbourn, A. J.: High sciatic lesion mimicking peroneal neuropathy at the fibular head. J. Neurol. Sci. 121:172, 1994. Katirji, M. B., and Wilbourn, A. J.: Common peroneal mononeuropathy: a clinical and electrophysiologic study of 116 lesions. Neurology 38:1723, 1988. Katz, D. A., Scheinberg, L., Horoupian, D. S., and Salen, G.: Peripheral neuropathy in cerebrotendinous xanthomatosis. Arch. Neurol. 42:1008, 1984. Kay, S. P.: Obstetrical brachial palsy. Br. J. Plast. Surg. 51:43, 1998. Kennedy, W. R., Sung, J. H., and Berry, J. F.: A case of congenital hypomyelinating neuropathy: clinical, morphological and chemical studies. Arch. Neurol. 34:337, 1977. Khatri, B. O., Flamini, J. R., Baruah, J. K., et al.: Plasmapheresis with acute inflammatory polyneuropathy. Pediatr. Neurol. 6:17, 1990.

2746

Diseases of the Peripheral Nervous System

258a. Kikuchi, S., and Tashiro, K.: Juvenile muscular atrophy of distal upper extremity-Hirayama disease. In Jones, H. R., De Vivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence. Boston, Butterworth-Heinemann, pp. 167–182, 2003. 259. King, P. J., Morris, J. G., and Pollard, J. D.: Glue sniffing neuropathy. Aust. N. Z. J. Med. 15:293, 1985. 260. Klein, C. J., Cunningham, J. M., Atkinson, E. J., et al.: The gene for HMSN2C maps to 12q23-24: a region of neuromuscular disorders. Neurology 60:1151, 2003. 261. Kline, D. G., Kim, D., Midha, R., et al.: Management and results of sciatic nerve injuries: a 24-year experience. J. Neurosurg. 89:13, 1998. 262. Knebusch, M., Strassburg, H. M., and Reiners, K.: Acute transverse myelitis in childhood: nine cases and review of the literature. Dev. Med. Child Neurol. 40:631, 1998. 263. Koch, T., Schulz, P., Williams, R., and Lampert, P.: Giant axonal neuropathy: a childhood disorder of neurofilaments. Ann. Neurol. 1:438, 1977. 264. Koenigsberger, M. R.: Brachial plexus palsy at birth: intrauterine or due to delivery trauma? Ann. Neurol. 8:228, 1980. 265. Kok, C., Kennerson, M. L., Spring, P. J., et al.: A locus for hereditary sensory neuropathy with cough and gastroesophageal reflux on chromosome 3p22-p24. Am. J. Hum. Genet. 73:632, 2003. 266. Korinthenberg, R.: Chronic inflammatory demyelinating polyradiculoneuropathy in children and their response to treatment. Neuropediatrics 30:190, 1999. 267. Korinthenberg, R., Sauer, M., Ketelsen, U. P., et al.: Congenital axonal neuropathy caused by deletions in the spinal muscular atrophy region. Ann. Neurol. 42:364, 1997. 268. Korn-Libutzki, I., Dor-Wollman, T., Soffer, D., et al.: Early peripheral nervous system manifestations in infantile Krabbe diseases. Pediatr. Neurol. 28:115, 2003. 269. Korobkin, R., Asbury, A., Sumner, A., and Nielsen, S.: Glue-sniffing neuropathy. Arch. Neurol. 32:158, 1975. 269a. Kozlowski, K., Barylak, A., Campbell, J., et al.: Primary sacral bone tumours in children (report of 16 cases with a short literature review). Aust. Radiol. 34:142, 1990. 270. Kreusser, K. L., and Volpe, J. J.: Peroneal palsy produced by intravenous fluid infiltration in a newborn. Dev. Med. Child Neurol. 26:522, 1984. 271. Krivit, W., Peters, C., and Shapiro, E. G.: Bone marrow transplantation as effective treatment of central nervous system disease in globoid cell leukodystrophy, metachromatic leukodystrophy, adrenoleukodystrophy, mannosidosis, fucosidosis, aspartylglucosaminuria, Hurler, MaroteauxLamy, and Sly syndromes, and Gaucher disease type III. Curr. Opin. Neurol. 12:167, 1999. 272. Krivit, W., Shapiro, E. G., Peters, C., et al.: Hematopoietic stem-cell transplantation in globoid-cell leukodystrophy. N. Engl. J. Med. 338:1119, 1998. 273. Kubiak, R., Wilcox, D. T., Spitz, L., and Kiely, E. M.: Neurovascular morbidity from the lithotomy position. J. Pediatr. Surg. 33:1808, 1998. 274. Kühne, T., Bubl, R., and Baumgartner, R.: Maternal vegan diet causing a serious infantile neurological disorder due to vitamin B12 deficiency. Eur. J. Pediatr. 150:205, 1991.

275. Kurczynski, T. W., Choudhury, A. A., Horwitz, S. J., et al.: Remote effects of malignancy on the nervous system in children. Dev. Med. Child Neurol. 22:205, 1980. 276. Kurihara, A., and Kataoka, O.: Lumbar disc herniation of children and adolescents: a review of 70 operated cases and their minimum 5-year follow-up studies. Spine 5:443, 1980. 277. Kuwabara, S., Nakajima, M., Tsuboi, Y., and Hirayama, K.: Multifocal conduction block in n-hexane neuropathy. Muscle Nerve 16:1416, 1993. 278. Lafrenière, R. G., MacDonald, M. L. E., Dubé, M.-P., et al.: Identification of a novel gene (HSN2) causing hereditary sensory and autonomic neuropathy type II through the study of Canadian genetic isolates. Am. J. Hum. Genet. 74:1064, 2004. 279. Lagos, J. C.: Compression neuropathy in childhood. Dev. Med. Child. Neurol. 13:531, 1971. 280. Lahat, E., Rothman, A. S., and Aron, A. M.: Schwannoma presenting as lumbar disc disease in an adolescent girl. Spine 9:695, 1984. 281. Lamont, P. J., Johnston, H. M., Berdoukas, V. A.: Plasmapheresis in children with Guillain-Barré syndrome. Neurology 41:1928, 1991. 282. Lang, H. A., Puusa, A., Hynninen, P., et al.: Evolution of nerve conduction velocity in later children and adolescence. Muscle Nerve 8:38, 1985. 283. Larsen, A., and Tobias, J. D.: Landry-Guillain-Barre syndrome presenting with symptoms of upper airway obstruction. Pediatr. Emerg. Care 10:347, 1994. 283a. Laulund, T., Fedders, O., Sogaard, I., and Kornum, M.: Suprascapular nerve compression syndrome. Surg. Neurol. 22:308, 1984. 284. Laurent, J. P., and Lee, R. T.: Birth-related upper brachial plexus injuries in infants: operative and nonoperative approaches. J. Child Neurol. 9:111, 1994. 285. Lawrence, D. G., and Locke, S.: Neuropathy in children with diabetes mellitus. Br. Med. J. 1:784, 1963. 286. Le Luyer, B., Devaux, A. M., Dailly, R., and Ensel, P.: Polyradiculoneuritis as a manifestation of childhood sarcoidosis. Arch. Fr. Pediatr. 40:175, 1983. 287. Lester, P., and McAlister, W.: Congenital iliac anomaly with sciatic palsy. Radiology 96:397, 1970. 288. Levin, K. H., Wilbourn, A. J., and Jones, H. R.: Childhood peroneal neuropathy from bone tumors. Pediatr. Neurol. 7:308, 1991. 289. Lewis, R. A., and Sumner, A. J.: The electrodiagnostic distinctions between chronic familial and acquired demyelinative neuropathies. Neurology 32:592, 1982. 290. Lewis, R. A., Sumner, A. J., and Shy, M. E.: Electrophysiological features of inherited demyelinating neuropathies: a reappraisal in the era of molecular diagnosis. Muscle Nerve 23:1472, 2000. 291. Lightwood, R.: Radial nerve palsy associated with localized subcutaneous fat necrosis in the newborn. Arch. Dis. Child. 26:436, 1951. 292. Lockman, L. A., Kennedy, W. R., and White, J. G.: The Chediak-Higashi syndrome: electrophysiological and electron microscopic observations on the peripheral neuropathy. J. Pediatr. 70:942, 1967.

Peripheral Neuropathies in Infants and Children 293. Logina, I., and Donaghy, M.: Diphtheritic polyneuropathy: a clinical study and comparison with Guillain-Barre syndrome. J. Neurol. Neurosurg. Psychiatry 67:433, 1999. 293a. Loredo, R., Hodler, J. Pedowitz, R., et al.: MRI of the common peroneal nerve: normal anatomy and evaluation of masses associated with nerve entrapment. J. Comput. Assist. Tomogr. 22:925, 1998. 294. Lorei, M. P., and Hershman, E. B.: Peripheral nerve injuries in athletes: treatment and prevention. Sports Med. 16:130, 1993. 295. Low, P. A., Burke, W. J., and McLeod, J. G.: Congenital sensory neuropathy with selective loss of small myelinated fibers. Ann. Neurol. 3:179, 1978. 296. Lucas, J. W., Holden, K. R., Purohit, D. M., and Cure, J. K.: Neonatal hemangiomatosis associated with brachial plexus palsy. J. Child Neurol. 10:411, 1995. 297. Luco, J. V., and Eyzaguirre, C.: Fibrillation and hypersensitivity to ACh in denervated muscle: effect of length and degenerating nerve fibers. J. Neurophysiol. 18:65, 1955. 298. Luijckx, G. J., Vies, J., de Baets, M., et al.: Guillain-Barre syndrome in mother and child. Lancet 349:27, 1997. 299. Lupski, J. R., Chance, P. F., and Garcia, C. A.: Inherited primary peripheral neuropathies: molecular genetics and clinical implications of CMT1A and HNPP. JAMA 270:2326, 1993. 300. MacFaul, R., Cavanagh, N., Lake, B. D., et al.: Metachromatic leukodystrophy: review of 38 cases. Arch. Dis. Child. 57:168, 1982. 301. Macgregor, J., and Moncur, J. A.: Meralgia paraesthetica— a sports lesion in girl gymnasts. Br. J. Sports Med. 11:16, 1977. 302. MacLean, W. C., and Graham, G. G.: Vegetarianism in children. Am. J. Dis. Child. 134:513, 1980. 303. Malm, G., Ringén, O., Winiarski, J., et al.: Clinical outcome in four children with metachromatic leukodystrophy treated by bone marrow transplantation. Bone Marrow Transplant. 17:1003, 1996. 304. Mancias, P., Slopis, J. M., Yeakley, J. W., and Vriesendorp, F. J.: Combined brachial plexus injury and root avulsion after complicated delivery [letter]. Muscle Nerve 17:1237, 1994. 305. Mardy, S., Miura, Y., Endo, F., et al.: Congenital insensitivity to pain with anhidrosis: novel mutations in the TRKA (NTRK1) gene encoding a high-affinity receptor for nerve growth factor. Am. J. Hum. Genet. 64:1570, 1999. 306. Marin, R., Bryant, P. R., and Eng, G. D.: Lumbosacral plexopathy temporally related to vaccination. Clin. Pediatr. 33:175, 1994. 307. Marks, H. G., Scavina, M. T., Kolodny, E. H., et al.: Krabbe’s disease presenting as a peripheral neuropathy. Muscle Nerve 20:1024, 1997. 308. Marrosu, M. G., Vaccargiu, S., Marrosu, G., et al.: CharcotMarie-Tooth disease type 2 associated with mutation of myelin protein zero gene. Neurology 50:1397, 1998. 308a. Martinoli, C., Bianchi, S., and Derchi, L.: Ultrasonography of peripheral nerves. Semin. Ultrasound CT MR 21:205, 2000. 309. Mathern, G. W., and Batzdorf, U.: Grisel’s syndrome: cervical spine clinical, pathologic, and neurologic manifestations. Clin. Orthop. 244:131, 1989.

2747

310. McAuliffe, T., Fiddian, N., and Browett, J.: Entrapment neuropathy of the superficial peroneal nerve: a bilateral case. J. Bone Joint Surg. Am. 67:62, 1985. 311. McCombe, P. A., Pollard, J. D., and McLeod, J. G.: Chronic inflammatory demyelinating polyradiculoneuropathy: a clinical and electrophysiological study of 92 cases. Brain 110:1617, 1987. 312. McEntagart, M., Norton, N., Williams, H., et al.: Localisation of the gene for distal hereditary motor neuronopathy-VII to chromosome 2q14. Am. J. Hum. Genet. 68:1270, 2001. 313. McFarland, L. V., Raskin, M., Daling, J. R., and Benedetti, T. J.: Erb/Duchenne’s palsy: a consequence of fetal macrosomia and method of delivery. Obstet. Gynecol. 68:784, 1986. 314. McGovern, J. P., and Merritt, D. H.: Sarcoidosis in childhood. Adv. Pediatr. 8:97, 1956. 315. McGuire, S. A., Gospe, S. M. Jr., and Dahl, G.: Acute vincristine neurotoxicity in the presence of hereditary motor and sensory neuropathy type I. Med. Pediatr. Oncol. 17:520, 1989. 316. McLeod, J. G., Pollard, J. D., Macaskill, P., et al.: Prevalence of chronic inflammatory demyelinating polyneuropathy in New South Wales, Australia. Ann. Neurol. 46:910, 1999. 317. Mehlman, C. T., Scott, K. A., Koch, B. L., and Garcia, V. F.: Orthopaedic injuries in children secondary to airbag deployment. J. Bone Joint Surg. Am. 82:895, 2000. 318. Meier, C., and Moll, C.: Hereditary neuropathy with liability to pressure palsies: report of two families and review of the literature. J. Neurol. 228:73, 1982. 319. Mersiyanova, I. V., Perepelov, A. V., Polyakov, A. V., et al.: A new variant of Charcot-Marie-Tooth type 2 is probably the result of a mutation in the neurofilament-light gene. Am. J. Hum. Genet. 67:37, 2000. 320. Michelow, B. J., Clarke, H. M., Curtis, C. G., et al.: The natural history of obstetrical brachial plexus palsy. Plast. Reconstr. Surg. 93:675, 1994. 321. Mildroni, G., de Tilly, L. N., Gray, B., and Vajsar, J.: MRI of the cauda equina in CIDP: clinical correlation. J. Neurol. Sci. 170:36, 1999. 322. Miller, R. G.: Hereditary and acquired polyneuropathies: electrophysiologic aspects. Neurol. Clin. 3:543, 1985. 323. Miller, R. G., Gutmann, L., Lewis, R. A., and Sumner, A. J.: Acquired versus familial demyelinative neuropathies in children. Muscle Nerve 8:205, 1985. 324. Miller, R. G., and Kuntz, N. L.: Nerve conduction studies in infants and children. J. Child Neurol. 1:19, 1986. 325. Mitchell, G., Larochelle, J., Lambert, M., et al.: Neurologic crises in hereditary tyrosinemia. N. Engl. J. Med. 322:432, 1990. 326. Mohan, U., Misra, V. P., Britto, J., et al.: Inherited early onset severe axonal polyneuropathy with respiratory failure and autonomic involvement. Neuromuscul. Disord. 11:395, 2001. 327. Montermini, L., Richter, A., Morgan, K., et al.: Phenotypic variability in Friedreich ataxia: role of the associated GAA triplet repeat expansion. Ann. Neurol. 41:675, 1997. 328. Moorhead, P. J., Cooper, D. J., and Timperley, W. R.: Progressive peripheral neuropathy in patients with primary hyperoxaluria. Br. Med. J. 2:312, 1975. 329. Moosa, A., and Dubowitz, V.: Peripheral neuropathy in Cockayne’s syndrome. Arch. Dis. Child. 45:674, 1970.

2748

Diseases of the Peripheral Nervous System

330. Morgan, L.: Radial nerve paralysis in the newborn. Arch. Dis. Child. 23:137, 1948. 331. Moser, H. W., Moser, A. B., Smith, K. D., et al.: Adrenoleukodystrophy: phenotypic variability and implications for therapy. J. Inherit. Metab. Dis. 15:645, 1992. 332. Moser, H. W., Naidu, S., Kumar, A. J., and Rosenbaum, A. E.: The adrenoleukodystrophies. Crit. Rev. Neurobiol. 3:29, 1987. 333. Moser, H. W., and Percy, A. K.: Peripheral neuropathy in inherited metabolic disease. In Jones, H. K., De Vivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence. Boston, Butterworth Heinemann, pp. 469–492, 2003. 334. Mosser, J., Douar, A. M., Sarde, C. O., et al.: Putative X-linked adrenoleukodystrophy gene shares unexpected homology with ABC transporters. Nature 361:726, 1993. 335. Muller, D. P. R., Lloyd, J. K., and Bird, A. C.: Long-term management of abetalipoproteinemia: possible role for vitamin E. Arch. Dis. Child. 52:209, 1977. 336. Nagamatsu, M., Jenkins, R. B., Schaid, D. J., et al.: Hereditary motor and sensory neuropathy type 2C is genetically distinct from types 2B and 2D. Arch. Neurol. 57:669, 2000. 337. Nakamura, T., Matsuzawa, Y., Takemura, K., et al.: Combined treatment with chenodeoxycholic acid and pravastatin improves plasma levels associated with marked regression of tendon xanthomas in CTX. Metabolism 40:741, 1991. 338. Nakano, K. K., Lundergran, C., and Okihiro, M. M.: Anterior interosseous nerve syndromes: diagnostic methods and alternative treatments. Arch. Neurol. 34:477, 1977. 339. Nance, M. A., and Berry, S. A.: Cockayne syndrome: review of 140 cases. Am. J. Med. Genet. 42:68, 1992. 340. Narakas, A. O.: Obstetrical Brachial Plexus Injuries. The Paralysed Hand. Edinburgh, Churchill Livingstone, p. 116, 1987. 341. Nardin, R. A., and Johns, D. R.: Mitochondrial dysfunction and neuromuscular disease. Muscle Nerve 24:170, 2001. 342. Nass, R., and Chutorian, A.: Dysaesthesias and dysautonomia: a self-limited syndrome of painful dysaesthesias and autonomic dysfunction in childhood. J. Neurol. Neurosurg. Psychiatry 45:162, 1982. 343. Nevo, U., Pestronk, A., Kornberg, A. J., et al.: Childhood chronic inflammatory demyelinating neuropathies: clinical course and long-term follow-up. Neurology 47:98, 1996. 344. Nevo, Y., and Topaloglu, H.: 88th ENMC International Workshop: Childhood Chronic Inflammatory Demyelinating Polyneuropathy (Including Revised Diagnostic Criteria), Naarden, The Netherlands, December 8–10, 2000. Neuromuscul. Disord. 12:195, 2002. 345. Nichino, I., Spinazzola, A., Papadimitriou, A., et al.: Mitochondrial neurogastrointestinal encephalomyopathy: an autosomal recessive disorder due to thymidine phosphorylase mutations. Ann. Neurol. 47:792, 2000. 345a. Nicholson, T. R., Cohen, R. C., and Grattan-Smith, P. J.: Intraneural ganglion of the common peroneal nerve in a 4-year-old boy. J. Child Neurol. 10:213, 1995. 346. Nikali, K., Koskinen, T., Suomalainen, A., et al.: Infantile onset spinocerebellar ataxia represents an allelic disease

347.

348.

349.

350. 351.

352.

353. 354. 355.

356.

357.

358.

359.

360.

361.

362. 363.

364. 365.

distinct from other hereditary ataxias. Pediatr. Res. 36:607, 1994. Niparko, N., Goldie, W. D., Mitchell, W., et al.: The use of plasmapheresis in the management of Guillain-Barré syndrome in pediatric patients [abstract]. Ann. Neurol. 26:448, 1989. Nucci, F., Artico, M., Santoro, A., et al.: Intraneural synovial cyst of the peroneal nerve: report of two cases and review of the literature. Neurosurgery 26:339, 1990. Ohashi, K., Ishibashi, S., Osuga, J., et al.: Novel mutations in the microsomal triglyceride transfer protein gene causing abetalipoproteinemia. J. Lipid Res. 41:1199, 2000. Ohnishi, A., and Dyck, P. J.: Loss of small peripheral sensory neurons in Fabry disease. Arch. Neurol. 31:120, 1974. Ohta, M., Ellefson, R. D., Lambert, E. H., and Dyck, P. J.: Hereditary sensory neuropathy, type II: clinical, electrophysiologic, histologic, and biochemical studies of a Quebec kinship. Arch. Neurol. 29:23, 1973. Orlowski, J. P., Clough, J. D., and Dyment, P. G.: Wegener’s granulomatosis in the pediatric age group. Pediatrics 61:83, 1978. Ouvrier, R. A.: Giant axonal neuropathy: a review. Brain Dev. 11:207, 1989. Ouvrier, R. A.: Update on acute and chronic inflammatory polyneuropathy. J. Child Neurol. 14:53, 1999. Ouvrier, R. A., and McLeod, J. G.: Chronic peripheral neuropathy in childhood: an overview. Aust. Paediatr. J. 24(Suppl.):80, 1988. Ouvrier, R. A., McLeod, J. G., and Conchin, T. E.: Friedreich’s ataxia: early detection and progression of peripheral nerve abnormalities. J. Neurol. Sci. 55:137, 1982. Ouvrier, R. A., McLeod, J. G., Morgan, G. J., et al.: Hereditary motor and sensory neuropathy of neuronal type with onset in early childhood. J. Neurol. Sci. 51:181, 1981. Ouvrier, R. A., and Shield, L.: Focal lesions of peripheral nerves. In Ouvrier, R. A., McLeod, J. G., and Pollard, J. D.: Peripheral Neuropathy in Childhood, 2nd ed. Cambridge, UK, Cambridge University Press, p. 244, 1999. Ouvrier, R. A., McLeod, J. G., and Pollard, J. D.: Neuropathies in metabolic and degenerative disorders. In Peripheral Neuropathy in Childhood, 2nd ed. Cambridge, UK, Cambridge University Press, p. 172, 1999. Ouvrier, R. A., McLeod, J. G., and Pollard, J. D.: Toxic neuropathies. In Peripheral Neuropathy in Childhood, 2nd ed. Cambridge, UK, Cambridge University Press, p. 201, 1999. Ouvrier, R. A., and Wilmshurst, J. M.: Overview of the neuropathies. In Jones, H. R., DeVivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 339, 2003. Ozdirim, E., Topcu, M., Ozon, A., and Cila, A.: Cockayne syndrome: review of 25 cases. Pediatr. Neurol. 15:312, 1996. Packer, R. J., Brown, M. J., and Berman, P. H.: The diagnostic value of electromyography in infantile hypotonia. Am. J. Dis. Child. 136:1057, 1982. Paine, R. S.: The future of the “floppy infant”: a follow-up study of 133 patients. Dev. Med. Child. Neurol. 5:115, 1963. Paradiso, G., Granana, N., and Maza, E.: Prenatal brachial plexus paralysis. Neurology 49:261, 1997.

Peripheral Neuropathies in Infants and Children 366. Parano, E., Uncini, A., De Vivo, D. C., and Lovelace, R. E.: Electrophysiologic correlates of peripheral nervous system maturation in infancy and childhood. J. Child Neurol. 8:336, 1993. 367. Pareyson, D.: Charcot-Marie-Tooth disease and related neuropathies: molecular basis for distinction and diagnosis. Muscle Nerve 22:1498, 1999. 368. Pareyson, D., Testa, D., Morbin, M., et al.: Does CMT1 homozygosity cause more severe disease with root hypertrophy and higher CSF proteins? Neurology 60:1721, 2003. 369. Parikh, S., Swaiman, K. F., and Kim, Y.: Neurologic characteristics of childhood lupus erythematosus. Pediatr. Neurol. 13:198, 1995. 370. Pasternak, J. F., Fulling, K., Nelson, J., and Prensky, A. L.: An infant with chronic relapsing polyneuropathy responsive to steroids. Dev. Med. Child. Neurol. 24:505, 1982. 371. Pasternak, J. F., and Volpe, J. J.: Lumbosacral lipoma with acute deterioration during infancy. Pediatrics 66:125, 1980. 372. Pellegrino, J. E., George, R. A., Biegel, J., et al.: Hereditary neuralgic amyotrophy: evidence for genetic homogeneity and mapping to chromosome 17q25. Hum. Genet. 101:277, 1997. 373. Peñas, P. F., Porras, J. I., Fraga, J., et al.: Microscopic polyangiitis: a systemic vasculitis with positive P-ANCA. Br. J. Dermatol. 134:542, 1996. 374. Perlmutter, D. H., Gross, P. T., Jones, H. R., et al.: Intramuscular vitamin E repletion in children with chronic cholestasis. Am. J. Dis. Child. 141:170, 1987. 375. Peterson, C. R., and Peterson, C. M.: Brachial-plexus injury in an infant from a car safety seat. N. Engl. J. Med. 325:1587, 1991. 376. Peyronnard, J. M., Charron, L., Bellavance, A., and Marchand, L.: Neuropathy and mitochondrial myopathy. Ann. Neurol. 7:262, 1980. 377. Pezeshkpour, G., Krarup, C., Buchtal, F., et al.: Peripheral neuropathy in mitochondrial disease. J. Neurol. Sci. 77:285, 1987. 378. Pietrini, V., Rizzuto, N., Vergani, C., et al.: Neuropathy in Tangier disease: a clinicopathologic study and review of the literature. Acta Neurol. Scand. 72:495, 1985. 379. Pingault, V., Bondurand, N., Le Caignec, C., et al.: The SOX10 transcription factor: evaluation as a candidate gene for central and peripheral hereditary myelin disorders. J. Neurol. 248:496, 2001. 380. Plante-Bordeneuve, V., and Said, G.: Dejerine-Sottas disease and hereditary demyelinating polyneuropathy of infancy. Muscle Nerve 26:608, 2002. 380a. Pleasure, D., and Jones, H. R.: Toxic polyneuropathies. In Jones, H. K., De Vivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence. Boston, Butterworth Heinemann, pp. 519–529, 2003. 381. Pollock, M., Nukada, H., Frith, R. W., et al.: Peripheral neuropathy in Tangier disease. Brain 106:911, 1983. 382. Pollack, R. N., Buchman, A. S., Yaffe, H., and Divon, M. Y.: Obstetrical brachial palsy: pathogenesis, risk factors and prevention. Clin. Obstet. Gynecol. 43:236, 2000. 383. Preston, D. C.: Distal medial neuropathies. Neurol. Clin. 17:407, 1999.

2749

384. Price, L., Raphael, S. A., Lischner, H. W., et al.: Inflammatory, demyelinating polyneuropathy in a child with symptomatic human immunodeficiency virus infection. J. Pediatr. 118:242, 1991. 385. Prineas, J. W., and McLeod, J. D.: Chronic relapsing polyneuritis. J. Neurol. Sci. 27:247, 1976. 386. Prineas, J. W., Ouvrier, R. A., Wright, R. G., et al.: Giant axonal neuropathy: a generalized disorder of cytoplasmic filament formation. J. Neuropathol. Exp. Neurol. 35:458, 1976. 387. Puusa, A., Lang, H. A., and Mäkelä, A.-L.: Nerve conduction velocity in juvenile rheumatoid arthritis. Acta Neurol. Scand. 73:145, 1986. 388. Quijano-Roy, S., Burns, T. M., Darras, B. T., and Jones, H. R.: Childhood CIDP: the Children’s Hospital of Boston experience. J. Peripher. Nerv. Syst. 6:171, 2001. 389. Radharkrishnan, K., Litchy, W. J., O’Fallon, W. M., and Kurland, L. T.: Epidemiology of cervical radiculopathy: a population based study from Rochester, Minnesota, 1976 through 1990. Brain 117:325, 1994. 390. Ramos, L. E., and Zell, J. P.: Rehabilitation program for children with brachial plexus and peripheral nerve injury. Semin. Pediatr. Neurol. 7:44, 2000. 391. Ramos-Fernandez, J. M., Oliete-Garcia, F. M., RoldanAparicio, S., et al.: Neonatal sciatic palsy: etiology and outcome of 21 cases. Rev. Neurol. 26:752, 1998. 392. Ramaekers, V. T., Lake, B. D., Harding, B., et al.: Diagnostic difficulties in infantile neuroaxonal dystrophy: a clinicopathological study of eight cases. Neuropediatrics 18:170, 1987. 393. Rank, J. M., Pascual-Leone, A., Payne, W., et al.: Hematin therapy for the neurologic crisis of tyrosinemia. J. Pediatr. 118:136, 1991. 394. Raphael, S. A., Price, M. L., Lischner, H. W., et al.: Inflammatory demyelinating polyneuropathy in a child with symptomatic human immunodeficiency viral infection. J. Pediatr. 118:242, 1991. 395. Reisin, R., Pardal, A., Ruggieri, V., and Gold, L.: Sural neuropathy due to external pressure: report of three cases. Neurology 44:2408, 1994. 396. Remaley, A. T., Rust, S., Rosier, M., et al.: Human ATP-binding cassette transporter 1 (ABC1): genomic organization and identification of the genetic defect in the original Tangier disease kindred. Proc. Natl. Acad. Sci. U. S. A. 96:12685, 1999. 397. Revuelta, R., De Juambelz, P. P., Fernandez, B., and Flores, J. A.: Lumbar disc herniation in a 27-month-old child: case report. J. Neurosurg. 92(Suppl. 1):98, 2000. 398. Richer, L. P., Shevell, M. I., and Miller, S. P.: Diagnostic profile of neonatal hypotonia: an 11-year study. Pediatr. Neurol. 25:32, 2001. 399. Richer, L. P., Shevell, M. I., Stewart, J., and Poulin, C.: Pediatric meralgia paresthetica. Pediatr. Neurol. 26:321, 2002. 399a. Ritter, F. J., Seay, A. R., and Lahey, M. E.: Peripheral mononeuropathy complicating anaphylactoid purpura. J. Pediatr. 103:77, 1983. 400. Robertson, W. C., Eichman, P. L., and Clancy, W. G.: Upper trunk brachial plexopathy in football players. JAMA 241:1480, 1979.

2750

Diseases of the Peripheral Nervous System

401. Robinson, L. R.: Traumatic injury to peripheral nerves. Muscle Nerve 23:863, 2000. 402. Rolfs, A., and Bolik, A.: Guillain-Barré syndrome in pregnancy: reflections on immunopathogenesis. Acta Neurol. Scand. 89:400, 1994. 402a. Romfh, J. H., and Currier, R. D.: Sciatic neuropathy induced by the lithotomy position. Arch. Neurol. 40:127, 1983. 403. Roos, D. B.: Thoracic outlet syndrome is underdiagnosed. Muscle Nerve 22:126, 1999. 404. Ropper, A. H.: The Guillain-Barré syndrome. N. Engl. J. Med. 326:1130, 1992. 405. Rothner, A. D., Wilbourn, A., and Mercer, R. D.: Rucksack palsy. Pediatrics 56:822, 1975. 406. Rothschild, E. J., Savitz, M. H., Chang, T., et al.: Primary vertebral tumor in an adolescent girl. Spine 9:695, 1984. 407. Rosen, N. L., Lechtenberg, R., Wisniewski, K., et al.: Adrenomyeloneuropathy with onset in early childhood. Ann. Neurol. 17:311, 1985. 408. Rosenblum, J. L., Keating, J. P., Prensky, A. L., and Nelson, J. S.: A progressive neurologic syndrome in children with chronic liver disease. N. Engl. J. Med. 304:503, 1981. 409. Ross, D., Jones, H. R. Jr., Fisher, J., and Konkol, R. J.: Isolated radial nerve lesion in the newborn. Neurology 33:1354, 1983. 410. Rottem, M., Fauci, A. S., Hallahan, C. W., et al.: Wegener granulomatosis in children and adolescents: clinical presentation and outcome. J. Pediatr. 122:26, 1993. 411. Rouse, D. J., Owen, J., Goldenberg, R. L., and Cliver, S. P.: The effectiveness and costs of elective cesarean delivery for fetal macrosomia diagnosed by ultrasound. JAMA 276:1480, 1996. 412. Rovira, A., Rovira, A., Capellades, J., et al.: Lumbar extradural hemangiomas: report of three cases. Am. J. Neuroradiol. 20:27, 1999. 413. Rowland, L. P., and Schneck, S. A.: Neuromuscular disorders associated with malignant neoplastic disease. J. Chron. Dis. 16:777, 1963. 414. Roy, E. P., Gutmann, L., and Riggs, J. E.: Longitudinal conduction studies in hereditary motor and sensory neuropathy type 1. Muscle Nerve 12:52, 1989. 415. Rust, R. S.: Congenital brachial plexus palsy: where have we been and where are we now. Semin. Pediatr. Neurol. 7:58, 2000. 416. Ryan, M. M., Grattan-Smith, P. J., Procopis, P. G., et al.: Childhood chronic inflammatory demyelinating polyneuropathy: clinical course and long-term outcome. Neuromuscul. Disord. 10:398, 2000. 417. Ryan, M. M., and Jones, H. R. Jr.: X-linked CMT presenting as chronic inflammatory demyelinating neuropathy with tremor. Submitted for publication. 418. Ryan, M. M., Tilton, A., De Girolami, U., et al.: Paediatric mononeuritis multiplex: a report of three cases and review of the literature. Neuromuscul. Disord. 13:751, 2003. 419. Sabin, T. D., Swift, T. R., and Jacobson, R. R.: Leprosy. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 1354, 1993. 420. Sacco, G., Buchthal, F., and Rosenfalck, P.: Motor unit potentials at different ages. Arch. Neurol. 6:44, 1962.

421. Salen, G.: Chenodeoxycholic treatment of cerebrotendinous xanthomatosis [reply from the authors]. Neurology 56:695, 2001. 422. Salih, M. A., Ahlstern, G., Stålberg, E., et al.: Friedreich’s ataxia in 13 children: presentation and evolution with electrophysiologic, electrocardiographic, and echocardiographic features. J. Child Neurol. 5:321, 1990. 423. Savitsky, K., Bar-Shira, A., Gilad, S., et al.: A single ataxia telangiectasia gene with a product similar to PI-3 kinase. Science 268:1749, 1995. 424. Sawaya, R. A.: Idiopathic sciatic mononeuropathy. Clin. Neurol. Neurosurg. 101:256, 1999. 425. Scheinberg, L.: Polyneuritis in systemic lupus erythematosus. N. Engl. J. Med. 255:416, 1956. 426. Wilcox, W. R., Banikazemi, M., and Guffon, N., et al.: Long term safety and efficacy of enzyme replacement for Fabry disease. Am. J. Human Genetics 75:65, 2004. 427. Schneegans, E., Amar, G., Isch, C., and Schneegans, D.: Sciatic paralysis of newborn and premature infants. Ann. Pediatr. 15:657, 1968. 428. Shapiro, E., Krivit, W., Lockman, L., et al.: Long-term beneficial effect of bone-marrow transplantation of childhoodonset cerebral X-linked adrenoleukodystrophy. Lancet 356: 713, 2000. 429. Shapiro, E. G., and Lipton, M. E.: White matter dysfunction and its neurophysiological correlates: a longitudinal study of a case of metachromatic leukodystrophy by bone marrow transplantation. J. Clin. Exp. Neuropsychol. 14:610, 1992. 430. Sharma, R. R., Sethu, A. U., Mahapatra, A. K., et al.: Neonatal cervical osteomyelitis with paraspinal abscess and Erb’s palsy: a case report and brief review of the literature. Pediatr. Neurosurg. 32:230, 2000. 431. Sharp, D., Blinderman, L., Combs, K. A., et al.: Cloning and gene defects in microsomal triglyceride transfer protein associated with abetalipoproteinemia. Nature 365:65, 1993. 432. Sheth, D., Gutmann, L., Blumenthal, D. T., et al.: Compressive sciatic neuropathy due to uterine abnormality. Muscle Nerve 17:1486, 1994. 433. Sheth, K. J., and Swick, H. M.: Peripheral nerve conduction in Fabry disease. Ann. Neurol. 7:319, 1980. 434. Shillito, J. Jr.: Pediatric lumbar disc surgery: 20 patients under 15 years of age. Surg. Neurol. 46:14, 1996. 435. Sidi, J.: Weak ankles: a study of common peroneal entrapment neuropathy. Br. Med. J. 3:623, 1969. 436. Simmons, Z., Mahadeen, Z. I., Kothari, M. J., et al.: Localized hypertrophic neuropathy: magnetic resonance imaging findings and long-term follow-up. Muscle Nerve 22:28, 1999. 437. Simmons, Z., Wald, J. J., and Albers, J. W.: Chronic inflammatory demyelinating polyradiculoneuropathy in children: I. Presentation, electrodiagnostic studies, and initial clinical course, with comparison to adults. Muscle Nerve 20:1008, 1997. 438. Simmons, Z., Wald, J. J., and Albers, J. W.: Chronic inflammatory demyelinating polyradiculoneuropathy in children: II. Long-term follow-up, with comparison to adults. Muscle Nerve 20:1569, 1997. 439. Simonati, A., Battistella, P. A., Guariso, G., et al.: Coeliac disease associated with peripheral neuropathy in a child: a case report. Neuropediatrics 29:155, 1998.

Peripheral Neuropathies in Infants and Children 440. Skjeldal, O. H., Stokke, O., Refsum, S., et al.: Clinical and biochemical heterogeneity in conditions with phytanic acid accumulations. J. Neurol. Sci. 77:87, 1987. 441. Skre, H.: Genetic and clinical aspects of Charcot-MarieTooth disease. Clin. Genet. 6:98, 1974. 442. Sladky, J. T.: Immune neuropathies in childhood. Baillieres Clin. Neurol. 5:233, 1996. 443. Sladky, J. T.: Guillain-Barré syndrome. In Jones, H. R., DeVivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 407, 2002. 444. Sladky, J. T., Brown, M. J., and Berman, P. H.: Chronic inflammatory demyelinating polyneuropathy of infancy: a corticosteroid-responsive disorder. Ann. Neurol. 20:76, 1986. 445. Slaugenhaupt, S. A., Blumenfeld, A., Gill, S. P., et al.: Tissuespecific expression of a splicing mutation in the IKBKAP gene causes familial dysautonomia. Am. J. Hum. Genet. 68:598, 2001. 445a. Sloane, A. E., Vajsar, J. Laxer, R. M., et al.: Spontaneous nontraumatic anterior compartment syndrome with peroneal neuropathy and favorable outcome. Neuropediatrics 25:268, 1994. 446. Solbrig, M. V., Mozaffar, T., and Kim, R. C.: Hansen’s disease (leprosy). In Jones, H. R., De Vivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, ButterworthHeinemann, p. 505, 2003. 447. Sotaniemi, K. A.: Slimmer’s paralysis—peroneal neuropathy during weight reduction. J. Neurol. Neurosurg. Psychiatry 47:564, 1984. 448. Sourkes, M., and Stewart, J. D.: Common peroneal neuropathy: a study of selective motor and sensory involvement. Neurology 41:1029, 1991. 449. Smith, S. J.: The role of neurophysiological investigation in traumatic brachial plexus lesions in adults and children. J. Hand Surg. Br. 21:145, 1996. 450. Stafford, C. R., Bogdanoff, B. M., Green, L., and Spector, H. B.: Mononeuropathy multiplex as a complication of amphetamine angiitis. Neurology 25:570, 1975. 451. Steinlin, M. I., Blaser, S. I., Gilday, D. L., et al.: Neurologic manifestations of pediatric systemic lupus erythematosus. Pediatr. Neurol. 13:191, 1995. 452. Steinman, L., Tharp, B. R., Dorfman, L. J., et al.: Peripheral neuropathy in the cherry-red spot-myoclonus syndrome (sialidosis type I). Ann. Neurol. 7:450, 1980. 453. Stewart, J.: Focal peripheral neuropathies. Submitted for publication. 454. Stogbauer, F., Young, P., Kerschensteiner, M., et al.: Recurrent brachial plexus palsies as the only clinical expression of hereditary neuropathy with liability to pressure palsies associated with a de novo deletion of the peripheral myelin protein-22 gene. Muscle Nerve 21:1199, 1998. 455. Street, V. A., Bennett, C. L., Goldy, J. D., et al.: Mutation of a putative protein degradation gene LITAF/SIMPLE in Charcot-Marie-Tooth disease 1C. Neurology 60:22, 2003. 456. Strombeck, C., Krumlinde-Sundholm, L., and Forssberg, H.: Functional outcome at 5 years in children with obstetrical

457.

458. 459.

460. 461.

462.

463.

464.

465.

466. 467.

468. 469.

470. 471.

472. 473.

474.

475.

2751

brachial plexus palsy with and without microsurgical reconstruction. Dev. Med. Child Neurol. 42:148, 2000. Styf, J., and Morberg, P.: The superficial peroneal tunnel syndrome: results of treatment by decompression. J. Bone Joint Surg. Br. 79:801, 1997. Sutherland, S.: Nerve and Nerve Injuries, 2nd ed. Edinburgh, Churchill-Livingston, p. 967, 1978. Swaiman, K. F., and Flagler, D. G.: Mercury poisoning with central and peripheral nervous system involvement treated with penicillamine. Pediatrics 48:639, 1971. Swift, T. R., and Ignacio, O. J.: Tick paralysis: electrophysiologic studies. Neurology 25:1130, 1975. Swoboda, K., and Jones, H. R.: The floppy infant: In Brown, W., Bolton, C. F., and Aminoff, M. (eds.): Neuromuscular Function and Disease. Philadelphia, W. B. Saunders, p. 1552, 2002. Swoboda, K. J., Edelbol-Eeg-Olofsson, K., Harmon, R. L., et al.: Pediatric electromyography. In Jones, H. R., De Vivo, D. C., and Darras, B. T. (eds.): Neuromuscular Disorders of Infancy, Childhood, and Adolescence: A Clinician’s Approach. Boston, Butterworth-Heinemann, p. 35, 2003. Swoboda, K. J., Engle, E. C., Scheindlin, B., et al.: Mutilating hand syndrome in an infant with familial carpal tunnel syndrome. Muscle Nerve 21:104, 1998. Tada, K., Yonenobu, K., and Swanson, A. B.: Congenital constriction band syndrome. J. Pediatr. Orthop. 4:726, 1984. Takao, M., Fukuuchi, Y., Koto, A., et al.: Localized hypertrophic mononeuropathy involving the femoral nerve. Neurology 52:389, 1999. Tasker, W., and Chutorian, A. M.: Chronic polyneuritis of childhood. J. Pediatr. 74:699, 1969. Teitelbaum, J. E., Berde, C. B., Nurko, S., et al.: Diagnosis and management of MNGIE syndrome in children: case report and review of the literature. J. Pediatr. Gastroenterol. Nutr. 35:377, 2002. Terzis, J. K., and Papakonstantinou, K. C.: Management of obstetric brachial plexus palsy. Hand Clin. 15:717, 1999. Thomas, J. E., Piepgras, D. G., Scheithauer, B., et al.: Neurogenic tumors of the sciatic nerve: a clinicopathologic study of 35 cases. Mayo Clin. Proc. 58:640, 1983. Thomas, P. K.: Hereditary sensory neuropathies. Brain Pathol. 3:157, 1993. Thomas, P. K., Marques, W. Jr., Davis, M. B., et al.: The phenotypic manifestations of chromosome 17p11.2 duplication. Brain 120:465, 1997. Thomson, A. J.: Idiopathic lumbosacral plexus neuropathy in two children. Dev. Med. Child Neurol. 35:258, 1993. Timmerman, V., De Jonghe, P., Ceuterick, C., et al.: A de novo mutation in the early growth response 2 (EGR2) gene associated with Dejerine-Sottas syndrome (DSS) phenotype. Am. J. Hum. Genet. 63(Suppl.):2259, 1998. To, W. C., and Traquina, D. N.: Neuralgic amyotrophy presenting with bilateral vocal cord paralysis in a child: a case report. Int. J. Pediatr. Otorhinolaryngol. 48:251, 1999. Trojaborg, W.: Clinical, electrophysiological, and myelographic studies of 9 patients with cervical spinal root avulsions: discrepancies between EMG and x-ray findings. Muscle Nerve 17:913, 1994.

2752

Diseases of the Peripheral Nervous System

476. Tsairis, P., Dyck, P. J., and Mulder, D. W.: Natural history of brachial plexus neuropathy: report on 99 patients. Arch. Neurol. 27:109, 1972. 477. Tuttle, J., Chen, R. T., Rantala, H., et al.: The risk of Guillain Barre syndrome after tetanus-toxoid containing vaccines in adults and children in the United States. Am. J. Public Health 87:2045, 1997. 478. Tyson, J., Malcolm, S., Thomas, P. K., and Harding, A. E.: Deletions of chromosome 17p11.2 in multifocal neuropathies. Ann. Neurol. 39:180, 1996. 479. Ubachs, J. M., Slooff, A. C., and Peeters, L. L.: Obstetric antecedents of surgically treated obstetric brachial plexus injuries. Br. J. Obstet. Gynaecol. 102:813, 1995. 480. Uncini, A., Di Guglielmo, G., Di Muzio, A., et al.: Differential electrophysiological features of neuropathies associated with 17p11.2 deletion and duplication. Muscle Nerve 18:628, 1995. 480a. Vaccaro, A. R., Ludwig, S. C., Klein, G. R., et al.: Bilateral peroneal nerve palsy secondary to a knee board: report of two cases. Am. J. Orthop. 27:746, 1998. 481. Vajsar, J., Sloane, A., Wood, E., and Murphy, E. G.: Plasmapheresis vs intravenous immunoglobulin treatment in childhood Guillain-Barré syndrome. Arch. Pediatr. Adolesc. Med. 148:1210, 1994. 482. Vallée, L., Dulac, O., Nuyts, J.-P., and Vamecq, J.: Intravenous immune globulin is also an efficient therapy of acute Guillain-Barré syndrome in affected children. Neuropediatrics 24:235, 1993. 483. van Dijk, J. G., Pondaag, W., and Malessy, M. J.: Obstetric lesions of the brachial plexus. Muscle Nerve 24:1451, 2001. 484. Vassalos, E., Prevedourakis, C., and ParaschopoulouPrevedouraki, P.: Brachial plexus paralysis in the newborn: an analysis of 169 cases. Am. J. Obstet. Gynecol. 101:554, 1968. 485. Vecchierinin-Blineau, M. R., and Guiheneuc, P.: Electrophysiological study of the peripheral nervous system in children. J. Neurol. Neurosurg. Psychiatry 42:753, 1979. 486. Vedanarayanan, V. V., Kandt, R. S., Lewis, D. V., and DeLong, G. R.: Chronic inflammatory demyelinating polyradiculoneuropathy of childhood: treatment with high-dose intravenous immunoglobulin. Neurology 41:828, 1991. 487. Vedanarayanan, V. V., Smith, S., Subramony, S. H., et al.: Lethal neonatal autosomal recessive axonal sensorimotor polyneuropathy. Muscle Nerve 21:1473, 1998. 488. Veneselli, E., Biancheri, R., Di Rocco, M., and Tortorelli, S.: Neurophysiological findings in a case of carbohydratedeficient glycoproteins (CDG) syndrome type I with phosphomannomutase deficiency. Eur. J. Paediatr. Neurol. 2:239, 1998. 489. Venna, N., Bielawski, M., and Spatz, E. M.: Sciatic nerve entrapment in a child: case report. J. Neurosurg. 75:652, 1991. 490. Verhagen, W. I., Gabreels-Festen, A. A., van Wensen, P. J., et al.: Hereditary neuropathy with liability to pressure palsies: a clinical, electroneurophysiological and morphological study. J. Neurol. Sci. 116:176, 1993. 491. Verrips, A., Hoefsloot, L. H., Steenbergen, G. C. H., et al.: Clinical and molecular genetic characteristics of patients with cerebrotendinous xanthomatosis. Brain 123:908, 2000.

492. Villarejo, F. J., and Pascual, A. M.: Injection injury of the sciatic nerve (370 cases). Childs Nerv. Syst. 9:229, 1993. 493. Wagner, A. L., and Buchthal, F.: Motor and sensory conduction in infancy and childhood: reappraisal. Dev. Med. Child Neurol. 14:189, 1972. 494. Wakai, S., Asanuma, H., Tachi, N., et al.: Infantile neuroaxonal dystrophy: axonal changes in biopsied muscle tissue. Pediatr. Neurol. 9:309, 1993. 495. Warner, L. E., Mancias, P., Butler, I. J., et al.: Mutations in the early growth response 2 (EGR2) gene are associated with hereditary myelinopathies. Nat. Genet. 18:382, 1998. 496. Warner, M. A., Warner, D. O., Harper, C. M., et al.: Lower extremity neuropathies associated with lithotomy positions. Anesthesiology 93:938, 2000. 497. Watkins, R. G.: Nerve injuries in football players. Clin. Sports Med. 5:215, 1986. 498. Weig, S. G., Waite, R. J., and McAvoy, K.: MRI in unexplained mononeuropathy. Pediatr. Neurol. 22:314, 2000. 499. Werlin, S. L., Harb, J. M., Swick, H., and Blank, E.: Neuromuscular dysfunction and ultrastructural pathology in children with chronic cholestasis and vitamin E deficiency. Ann. Neurol. 13:291, 1983. 500. White, N. H., Waltman, S. R., Krupin, T., and Santiago, J. V.: Reversal of neuropathic and gastrointestinal complications related to diabetes mellitus in adolescents with improved metabolic control. J. Pediatr. 99:41, 1981. 501. Wichman, A., Buchthal, F., Pezeshkpour, G. H., and Gregg, R. E.: Peripheral neuropathy in abetalipoproteinemia. Neurology 35:1279, 1985. 502. Wilbourn, A. J.: Differentiating acquired from familial segmental demyelinating neuropathies by EMG. Electroencephalogr. Clin. Neurophysiol. 43:616, 1977. 503. Wilbourn, A. J.: Electrodiagnostic testing of neurologic injuries in athletes. Clin. Sports Med. 9:229, 1990. 504. Wilbourn, A. J.: Brachial plexus disorders. In Dyck, P. J., Thomas, P. K., Griffin, J. W., et al. (eds.): Peripheral Neuropathy, 3rd ed. Philadelphia, W. B. Saunders, p. 903, 1993. 505. Williams, C. L., Strobino, B., Lee, A., et al.: Lyme disease in childhood: clinical and epidemiologic features of ninety cases. Pediatr. Infect. Dis. J. 9:10, 1990. 506. Wilmshurst, J. M., Bye, A., Rittey, C., et al.: Severe infantile axonal neuropathy with respiratory failure. Muscle Nerve 24:760, 2001. 507. Wroble, R. R., and Albright, J. P.: Neck and low back injuries in wrestling. Clin. Sports Med. 5:295, 1986. 508. Yabuki, S., Nakaya, K., Sugimura, T., et al.: Juvenile polyneuropathy due to vitamin B1 deficiency: clinical observations and pathogenetic analysis of 24 cases. Folia Psychiatr. Neurol. 30:517, 1976. 509. Yao, J. K., Herbert, P. N., Fredrickson, D. S., et al.: Biochemical studies in a patient with a Tangier syndrome. J. Neuropathol. Exp. Neurol. 37:138, 1978. 510. Yilmaz, K., Caliskan, M., Oge, E., et al.: Clinical assessment, MRI, and EMG in congenital brachial plexus palsy. Pediatr. Neurol. 21:705, 1999. 511. Yokota, T., Shiojiri, T., Gotoda, T., et al.: Friedreich-like ataxia with retinitis pigmentosa caused by the His101Gln

Peripheral Neuropathies in Infants and Children mutation of the alpha-tocopherol transfer protein gene. Ann. Neurol. 41:826, 1997. 512. Yoshioka, M., Kuroki, S., and Mizue, H.: Plasmapheresis in the treatment of Guillain-Barré syndrome in childhood. Pediatr. Neurol. 1:329, 1985. 513. Young, N. L., Davis, R. J., Bell, D. F., and Redmond, D. M.: Electromyographic and nerve conduction changes after tibial lengthening by the Ilizarov method. J. Pediatr. Orthop. 13:473, 1993. 514. Young, R. J., Ewing, D. J., and Clarke, B. F.: Nerve function and metabolic control in teenage diabetics. Diabetes 32:142, 1983.

2753

515. Young, S. G., Hubl, S. T., Smith, R. S., et al.: Familial hypobetalipoproteinemia caused by a mutation in the apolipoprotein B gene that results in a truncated species of apolipoprotein B (B-31). J. Clin. Invest. 85:993, 1990. 516. Yuen, E. C., and So, Y. T.: Sciatic neuropathy. Neurol. Clin. 17:617, 1999. 517. Yuen, E. C., So, Y. T., and Olney, R. K.: The electrophysiologic features of sciatic neuropathy in 100 patients. Muscle Nerve 18:414, 1995. 518. Zerras, K., and Rudnik-Schöneborn, S.: Natural history in proximal spinal muscular atrophy. Arch. Neurol. 52:518, 1995.

Index

Note: Page numbers followed by the letter f refer to figures and those followed by t refer to tables. A wave, 926, 927f, 994 A-60, of axoplasmic cortex, 47 AAN (acute autonomic neuropathy), Epstein-Barr virus and, 2121 A␤ fibers, 168t, 169, 2638 ABC syndrome, 1150 ABCA1, 1909 expression of, 1913 genetics of, 1912–1913 human mutations of, 1914, 1914t in familial hypoalphalipoproteinemia, 1905 in intracellular trafficking, 1913 in Tangier disease, 1905, 1906, 1911–1914 knockout mouse models of, 1914 other functions of, 1913–1914 regulation of, 1913 topologic arrangement of transporter for, 1913, 1915f Abdominal innervation, 25–30 autonomic, 30–31 Abdominal pain, 2610 Abducens nerve, 1192–1201 blood supply of, 1195 clinical features of sixth nerve palsy, 1197 course of, 1193–1194, 1194f histology of, 1195, 1196f in angiostrongyliasis, 2164 in diabetes, 1960 muscles innervated by, 1192 neoplasms affecting, 1199–1200 nucleus of, 1193 traumatic injury of, 1198 vascular diseases affecting, 1201 Abductor tunnel, 1503 Abetalipoproteinemia childhood, 2722 infantile, 2715 ABRs. See Auditory brainstem responses (ABRs). Abscess, pelvic, 1385f Acanthamoeba infection, 2161 keratitis, 2154t, 2162 clinical features of, 2162 pathogenesis of, 2162 transmission of, 2162 life cycle and morphology of etiologic agent in, 2162, 2163f

Acceptability of health outcome measures, 1057–1058 Accessory ganglion, 17 Accessory nerve, 1286 anatomy of, 1286 course of, 1286 cranial, 1277 injury during surgery in neck, 1286, 1519, 1520f lesions of, 1286 clinical features of, 1286 muscles innervated by, 956t nerve conduction studies of, 909 Accommodation, 113, 114 ACE (angiotensin-converting enzyme), in sarcoidosis, 2419, 2422 Acetylcholine (ACh) action on urethral smooth muscle, 302 carbamate-induced accumulation of, 2509 congenital myasthenic syndrome with reduced quantal release and CNS symptoms, 842 effect on anal sphincter, 286t in autonomic ganglia, 233, 240, 241 in bladder, 301, 304 in cardiac neural control, 221, 223–224, 223f–224f in cutaneous vascular responses to body heating, 333 in enteric nervous system, 271 in Horner’s syndrome, 208 in male sexual response cycle, 314, 315 in micturition reflex, 312 in neuromuscular transmission, 832 in sweat gland innervation, 1106 paucity of synaptic vesicles and reduced quantal release of, 841–842 therapy for, 860 storage in synaptic vesicles, 831 transmitter quantum of, 831 vesicular transporter for, 832 Acetylcholine receptor (AChR), 831–832 deficiency in myasthenia gravis, 833, 835f expression by dorsal root ganglion neurons, 183 in bladder, 301, 304 in sweat gland innervation, 1106 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Acetylcholine receptor (AChR) (Continued) kinetic abnormalities of, 845, 847, 847t fast-channel syndromes, 851f, 851–852 slow-channel syndromes, 847f–850f, 847–851 muscle, subunits of, 844–845 mutations causing deficiency of, 852–855 clinical features of, 852–853, 853f end-plate studies in, 853, 854f molecular pathogenesis of, 853–855, 855f Acetylcholinesterase (AChE), 831–832 activity after axotomy, 693 axonal transport of, 389, 394–395 carbamate inhibition of, 2509 end-plate, deficiency of, 842–844 clinical features of, 842–843, 843f–844f molecular pathogenesis of, 843–844, 846f pathogenesis and pathology of, 843, 844f–845f therapy for, 860 in ganglion cells of lower urinary tract, 301 organophosphate inhibition of, 2515–2516 3-Acetylpyridine chromatolysis induced by, 686 intracytoplasmic vacuolation induced by, 688 ACh. See Acetylcholine (ACh). Achalasia, 1281, 1285 sphincteric, 260–261 AChE. See Acetylcholinesterase (AChE). AChR. See Acetylcholine receptor (AChR). Acid cholesteryl ester hydrolase deficiency, 1864 Acid ␣-1,4-glycosidase deficiency, 709 Acid phosphatase, 417 Acid-sensing ion channels (ASICs), 187t, 188 Acoustic meatus, internal, 1219 Acoustic neuroma, blink reflex in, 921t Acoustic reflex, 1256 Acousticofacial cistern, 1219 Acquired immune deficiency syndrome (AIDS). See Human immunodeficiency virus (HIV) infection. Acrocyanosis, carpal tunnel syndrome and, 1440 Acrodermatitis chronica atrophicans, 2109. See also Lyme disease.

i

ii Acrodystrophy autosomal recessive HMSN II with, 1785 neuropathy and, 1151 Acromegaly, 2045–2046 carpal tunnel syndrome in, 1445, 2045 polyneuropathy in, 2045–2046 cause of, 2045 clinical features of, 2045 pathology of, 2045, 2046f treatment of, 2046 Acrylamide, 2506–2508 clinical features of exposure to, 2506 effect on axonal transport, 396–398, 397t metabolism of, 2506 monitoring exposure to, 2506 neuropathy due to, 2506–2508 diagnosis of, 2506 electrophysiologic findings in, 2507–2508 morphology of, 2507 pathogenesis of, 2508 treatment of, 2506 pathology induced by, 717, 2507 intracytoplasmic vacuolation, 688 toxicity of, 2506 use and sources of exposure to, 2506 Actin dynamics of, 448f, 448–453 depolymerization, 448f, 448–449, 452 polymerization and nucleation, 450–452 retrograde F-actin flow, 452 role of Rho GTPases, 450, 450f–451f expression after axotomy, 690 filopodia formed by, 448, 448f lamellipodia formed by, 448, 448f of axoplasmic cortex, 47 perineurial, 39 treadmilling of, 449 Action potentials compound, 1015–1026 fiber composition and, 1015–1016 fiber diameter and, 1015, 1018 fiber groups and sensation, 1016–1017 nomenclature for, 1015–1016 letter designations of, 1015–1016 roman numeral classification, 1016 of sural nerve, 1017–1026 fiber diameter and, 1018 in Déjérine-Sottas neuropathy, 1646f in dominantly inherited amyloidosis, 1021f, 1021–1022 in Friedreich’s ataxia, 1020f, 1020–1021, 1817f in hereditary motor and sensory neuropathy type I, 1025–1026, 1027f in hereditary sensory neuropathy type I, 1022, 1023f, 1817f in hereditary sensory neuropathy type II, 1022–1023, 1024f in normal nerve, 1018–1020, 1019f in relapsing inflammatory polyradiculoneuropathy, 1024–1025, 1026f–1027f in tabes dorsalis, 1022, 1022f, 1817f in uremic neuropathy, 1023–1024, 1025f method of study of, 1017–1018 reconstruction of, 1016 sources of distortion in recording of, 1017

Index Action potentials (Continued) compound muscle, 900 in brachial plexopathy, 1351 in Charcot-Marie-Tooth disease type 1A, 1662 in chronic inflammatory demyelinating polyradiculoneuropathy, 2234 in chronic obstructive pulmonary disease–associated neuropathy, 2022 in common peroneal neuropathy, 1498 in congenital myasthenic syndromes resembling Lambert-Eaton myasthenic syndrome, 842 slow-channel syndromes, 848 sodium channel myasthenia, 858 with reduced quantal release and central nervous system symptoms, 842 in critical illness polyneuropathy, 2027 in diagnosis of conduction block, 2283 in end-plate acetylcholinesterase deficiency, 547 in femoral neuropathy, 1491 in giant axonal neuropathy, 1651 in Guillain-Barré syndrome acute inflammatory demyelinating polyneuropathy, 2200 axonal forms, 2201 pediatric, 2716 in hepatitis C–associated neuropathy, 2019 in hereditary motor and sensory neuropathy, 1637–1638, 1647 type II, 1718, 1722 in Lambert-Eaton myasthenic syndrome, 837 in mice deficient in porphobilinogen deaminase, 551 in monoclonal gammopathy of undetermined significance, 2265 in multifocal motor neuropathy, 2182 in neurapraxia, 902 in neuropathy associated with chronic hepatic disease, 2018 in paraneoplastic subacute sensory neuronopathy, 2475 in pediatric polyneuropathies, 2712 in perinatal brachial plexopathy, 2731 in POEMS syndrome, 2457–2458 in tick paralysis, 2717 in true neurologic thoracic outlet syndrome, 1360 in uremic neuropathy, 2023, 2023f after renal transplantation, 2023f, 2025 in intrafusal fibers, 138 nodes of Ranvier in generation of, 415 of neurons of enteric nervous system, 262–263 potassium channels and, 95, 99, 100f sensory nerve effect of dorsal root ganglion pathology on, 1323, 1324, 1327 factors affecting amplitude of, 904 gender differences in, 904 in acrylamide neuropathy, 2506 in brachial plexopathy, 1351 immune, 2302 perinatal, 2731 in chronic obstructive pulmonary disease–associated neuropathy, 2022 in common peroneal neuropathy, 1498 in critical illness myopathy, 2030 in critical illness polyneuropathy, 2027 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Action potentials (Continued) in cryptogenic sensory polyneuropathy, 2323 in femoral neuropathy, 1490 in giant axonal neuropathy, 1651 in Guillain-Barré syndrome acute inflammatory demyelinating polyneuropathy, 2200 axonal forms, 2201 in hepatitis C–associated neuropathy, 2019 in hereditary motor and sensory neuropathy, 1637–1638 in lumbosacral plexus lesions, 1378 in neuropathy associated with chronic hepatic disease, 2018 in nonmalignant inflammatory sensory polyganglionopathy, 2312–2313 in pediatric polyneuropathies, 2712 in primary biliary cirrhosis–associated sensory polyneuropathy, 2020 in radiculopathy, 1324 in sciatic neuropathy, 1495 in sensory variant of chronic inflammatory demyelinating polyradiculoneuropathy, 2229 in tick paralysis, 2717 in true neurologic thoracic outlet syndrome, 1360 in uremic neuropathy, 2023 after renal transplantation, 2025 sodium channels and, 95 Activating transcription factor 3, 692 Activities of daily living (ADLs), 2621–2622 Acute autonomic neuropathy (AAN), Epstein-Barr virus and, 2121 Acute febrile polyneuritis, 2198 Acute idiopathic neuritis. See Guillain-Barré syndrome (GBS). Acute inflammatory demyelinating polyneuropathy (AIDP), 2199–2200 after bone marrow transplantation, 2498 antiganglioside antibodies and, 592, 596 clinical manifestations of, 2199–2201 in children, 2199 progression and plateau phase, 2200 regression phase, 2200–2201 stage of invasion, 2199–2200 compared with chronic inflammatory demyelinating polyradiculoneuropathy, 2222–2223 electrodiagnosis of, 2200 epidemiology of, 2199 in HIV infection, 2136–2137, 2224 clinical features of, 2136–2137 disease-stage specificity of, 2129 etiology of, 2137 pathology of, 2137 treatment of, 2137 infections associated with, 2200t, 2223 Campylobacter jejuni, 592 lymphoma and, 2494 pathogenesis of, 2211–2212 pathology of, 806f, 806–807, 2200t, 2202–2204, 2203f–2206f axonal degeneration, 2204 demyelination, 2202–2204 inflammation, 2202 recovery from, 2222 sensory-predominant, 2211, 2315t severity of, 2200 temporal evolution of, 2222–2223

Index Acute intermittent porphyria (AIP), 1167, 1883–1891. See also Porphyric neuropathy. clinical manifestations of, 1886 drug-induced attacks of, 1884–1885, 1887t gastrointestinal symptoms of, 2699 hormonal influences in, 1885 in children, 2723 incidence of, 1886 inheritance of, 1886 laboratory studies in, 1889t, 1889–1890 latent form of, 1886–1887 mechanisms of neuronal damage in, 1885–1886 mouse model of, 1886 neurologic disease in, 1888 pathologic changes in, 1886 treatment of, 1885, 1890–1891 Acute lymphoblastic leukemia (ALL), 2496–2497 Acute motor axonal neuropathy (AMAN), 575, 2178t, 2201 clinical features of, 2201 electrophysiology of, 2201 IgG anti-GD1a antibodies in, 2207–2210 IgG anti-GM1 antibodies in, 591, 596, 2178t, 2184–2185, 591, 596, 2178t, 2184–2185, 2207–2210 disease pathogenesis and, 2184–2185 testing for, 2184 treatment based on, 2184 infections associated with, 2200t pathogenesis of, 2206–2210 antecedent Campylobacter infection, 591, 2206–2207 antigens, 2200t, 2207–2209 experimental evidence, 2209–2210 immunopathology, 2207, 2208f pathology of, 2200t, 2204 chromatolysis, 686 respiratory insufficiency in, 2201 Acute motor-sensory axonal neuropathy (AMSAN), 2201 antigens in, 2200t clinical features of, 2201 infections associated with, 2200t pathology of, 2200t, 2204 chromatolysis, 686 Acute myelogenous leukemia (AML), 2497 Acute sensory neuropathy of childhood, 2717 Acyclovir for Bell’s palsy, 1230 for Ramsay Hunt syndrome, 1232 A␦ fibers, 168t, 169, 870, 1005, 2638 Adaptive equipment, 2625 Adaptive immune system, 574, 609–610 ADDQoL (Audit of Diabetes-Dependent Quality of Life), 1058 Adenosine effects on nicotinic transmission in bladder, 304 in enteric nervous system, 271 Adenosine triphosphate (ATP) action on bladder smooth muscle, 301 action on urethral smooth muscle, 302 in cardiac neural control, 224 in male sexual response cycle, 314–315 mitochondrial generation of, 1938 receptors for, 182 role in axonal transport, 387, 388, 393 secretion by sympathetic ganglion neurons, 233, 240 urothelial cell release of, 300

Adhesion molecules, 377, 2341–2342 immunoglobulin superfamily of, 1684, 2342 in vasculitis, 2343 integrins, 2342 leukocyte emigration and, 2342–2343, 2343f P0, 417, 1682, 1684–1685, 1687f selectins, 2341–2342 Adie’s syndrome, 205, 210–212, 211f ADLs (activities of daily living), 2621–2622 Adrenal cortical adenoma, 2596 Adrenaline. See Epinephrine. Adrenergic agents, effect on anal sphincter, 286t ␣-Adrenergic receptors expression by dorsal root ganglion neurons, 183 in bladder, 304 ␤-Adrenergic receptors, in bladder, 304 Adrenoleukodystrophy (ALD), 1866–1868, 1867f clinical features of, 1867 demyelination in, 1167, 1867 genetics of, 1867 hyperpigmentation in, 1168 inheritance of, 1866 neonatal, 1868–1869 pathogenesis of, 1867–1868 pathology of, 710, 1867, 1868f phenotypes of, 1867 treatment of, 1868 vs. amyotrophic lateral sclerosis, 1536 X-linked, in children, 2722 Adrenomyeloneuropathy (AMN), 1867 adrenal insufficiency in, 1867 clinical features of, 1867 demyelination in, 1167, 1867 in children, 2722 pathogenesis of, 1867–1868 pathology of, 710, 1867 ␣-Adrenoreceptor agonists, for orthostatic hypotension, 2675–2676 Adult polyglucosan body diseases, 709–710 Advanced glycation end products (AGEs), 514 age-related formation of, 492 in diabetic neuropathy, 1970, 1975–1976 receptor for, 514, 1975–1975 Adynamic ileus, 2610, 2616 Afferent fibers of dorsal root ganglion neurons, 165–166 adaptation rate of, 166, 167t–168t afferent receptive properties of, 165–166 conduction velocity of, 166 cutaneous, muscle, and visceral, 167t–168t, 170, 186 functional classes of, 166–169, 167t–168t low-threshold mechanoreceptors, 166–169 nociceptors, 169 thermoreceptors, 169 nociceptors, 165–166, 168t of vagus nerve, 1278 disorders of, 1283–1284, 1284f to lower urinary tract, 305, 308, 309f to pelvic floor, 282–283 AFO (ankle-foot orthosis), 2623 Age-related changes, 483–501, 702–703 brachial plexopathies, 1342 chromatolysis, 686 in ankle reflexes, 1176 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

iii Age-related changes (Continued) in autonomic neurons, 492–495, 702 mechanisms of, 498–500, 499f neurodegeneration and loss of function, 494–495 selective vulnerability to neurodegeneration, 492–494, 493f–495f in dorsal root ganglion, 484–486, 487, 702t, 702–703, 703t in eighth cranial nerve, 1253–1254 in muscle strength, 1176 in muscle sympathetic nerve activity, 332f, 332–333 in nerve conduction velocity, 486, 671, 907, 907f, 908t, 1176 in nerve fiber density, 37 epidermal, 879, 879f in nerve trunks, 486–492 functional impairments in axons, 486 mechanisms of, 496–498 myelinated axons, 483–484, 487–491, 489f–492f structural impairments in axons, 486–487, 488f–489f unmyelinated axons, 491–492 vascular changes, 492 in perineurial compliance, 656 in peripheral neurons, 702–703, 703t lipofuscin, 702–703, 703t melanin, 703 neurofibrillary tangles, 703 nodules of Nageotte, 702, 702t in pupil size, 208f, 208–209 in sensation thresholds, 1067, 1176 in sensory neurons, 483–486, 702 functional impairments, 483 mechanisms of, 495–496 signaling molecules, 484–486 structural impairments, 483–484, 484f–485f in Valsalva ratio in males and females, 1115, 1115f–1116f neuron death, 701 on nerve blood flow, 671 selective vulnerability to, 483 vs. polyneuropathy, 1176 AGEs (advanced glycation end products), 514 age-related formation of, 492 in diabetic neuropathy, 1970 receptor for, 514 Aggresomes, in Pmp22 mutant mice, 1572 AIDP. See Acute inflammatory demyelinating polyneuropathy (AIDP). AIDS (acquired immune deficiency syndrome). See Human immunodeficiency virus (HIV) infection. AIMS2 (Arthritis Impact Measurement Scale), 1058–1059, 1060 AIP. See Acute intermittent porphyria (AIP). ALA (␦-aminolevulinic acid), in lead neuropathy, 2530 ALA synthase. See ␦-Aminolevulinic acid (ALA) synthase. Albers-Schönberg disease, facial nerve paralysis and, 1236 Alcoholic neuropathy intracytoplasmic vacuolation and, 688 macro electromyography in, 999 thiamine deficiency and, 2051 ALD. See Adrenoleukodystrophy (ALD). Aldicarb intoxication, 2509, 2510

iv Alimentary tract ganglioneuromatosis, in MEN type 2B, 1806 Alkylating agents, 637–638 adverse effects of, 638 mechanism of action of, 637–638 pharmacokinetics of, 638 use in neuropathy, 638 ALL (acute lymphoblastic leukemia), 2496–2497 Allergic granulomatous angiitis, 2337. See also Churg-Strauss syndrome (CSS). All-median hand, 929 Allodynia, 1146 epidermal denervation and, 870 HIV-associated sensory neuropathy and, 2131 in diabetic neuropathy, 1954 in hereditary brachial plexus neuropathy, 1754 in small fiber neuropathy, 2322 mechanism of, 2639 NMDA antagonists for, 527 Allografts, nerve, 1424 Allyl chloride neuropathy, 2508 Almitrine neuropathy, 2022, 2554 ALS. See Amyotrophic lateral sclerosis (ALS). ALS2 gene mutations, in amyotrophic lateral sclerosis type 2, 1535, 1598–1599 ALS/PDC (amyotrophic lateral sclerosis/ parkinsonism-dementia complex), 1600 Aluminum chloride, 715 Aluminum phosphate, 715 Alzheimer’s disease, 703 AMAN. See Acute motor axonal neuropathy (AMAN). Ambulation aids, 2624, 2630 Ameba, free-living, 2161 Amebic keratitis, 2154t, 2162 clinical features of, 2162 life cycle and morphology of etiologic agent in, 2162, 2163f pathogenesis of, 2162 transmission of, 2162 American trypanosomiasis, 260, 1151, 1281, 2154t, 2156–2160 cardiac lesions in, 2159 clinical phases of, 2156–2157, 2158f congenital, 2156 epidemiology of, 2156 etiologic agent in, 2156 life cycle of, 2156, 2157f gastrointestinal megasyndromes in, 2159–2160 pathogenesis of, 2157 pathophysiology of, 2157–2159 peripheral neuropathy in, 2160 reactivation of, 2157 transmission of, 2156 ␥-Aminobutyric acid (GABA) blockade by ␦-aminolevulinic acid synthase in porphyria, 1885 in cardiac neural control, 218, 219f, 221, 222 in micturition reflex, 312 Aminoguanidine, for diabetic neuropathy, 1976 ␣-Aminohydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, effect of inflammation on, 525 ␦-Aminolevulinic acid (ALA), in lead neuropathy, 2530 ␦-Aminolevulinic acid (ALA) synthase blockade of GABA transmission by, 1885 drug effects on, 1885 glucose levels and induction of, 1885 hormone induction of, 1885 in heme biosynthesis, 1883–1884

Index Amiodarone neuropathy, 717, 2554–2555 Amitriptyline for diabetic neuropathy, 1971, 1971t for HIV-associated sensory polyneuropathies, 2135 for neuropathic pain, 2640 AML (acute myelogenous leukemia), 2497 AMN. See Adrenomyeloneuropathy (AMN). AMPA (␣-aminohydroxy-5-methyl4-isoxazolepropionic acid) receptors, effect of inflammation on, 525 Amphiphysin-1, 586 Amphiphysin-2, 586 AMSAN. See Acute motor-sensory axonal neuropathy (AMSAN). Amyloid neuropathy, 2430–2433 at time of amyloidosis diagnosis, 2430–2431, 2431f autonomic dysfunction in, 1141, 2431 carpal tunnel syndrome and, 1165, 1447, 1923, 1924, 2431f, 2432–2433 clinical features of, 2431 anhidrosis, 1150 gastrointestinal, 2444, 2698 electromyography in, 2431–2432 gender and severity of, 1548 hereditary, 1921–1932, 2440–2444 age at onset of, 2440 apolipoprotein A-I–related, 1921, 1922, 1931–1932 classification of, 1921, 1922, 2428t-2429t, 2440t clinical features of, 2440–2441 compound action potential of sural nerve in, 1021f, 1021–1022 cranial nerve involvement in, 2440–2441 diagnosis of, 2443 electrocardiographic findings in, 2441 familial amyloid polyneuropathy, 2440–2441 gelsolin-related, 1921, 1922, 1932 historical overview of, 1921–1922, 2440 inheritance of, 2440 pathology of, 2443–2444 therapy for, 2444 transthyretin-related, 809, 1921, 1922–1931 central nervous system involvement in, 1924 clinical course of, 1924–1925 clinical features of, 1923–1924, 2441 diagnosis of, 1928–1929, 1929f differential diagnosis of, 1928 epidemiology of, 1924, 1927 genetics of, 1925–1927, 2442–2443 other system involvement in, 1924 pathogenesis of, 1927–1928, 2442–2443 pathology of, 1925, 1926f, 2443–2444 prognosis for, 1931 treatment of, 1929–1931, 2444 lower motor neuron disease and, 1304 pathogenesis of, 2432 pathology of, 712–713, 809–810, 811f–814f, 1925, 1926f, 2432, 2443–2444 nerve thickening, 1152 small fiber sensory loss, 1147 upper limb involvement in, 1138 Amyloidoma of peripheral nerve, 2596 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Amyloidosis, 1921–1922, 2427–2444 axonal degeneration in, 2430 classification of, 2428, 2428t–2429t cranial nerve involvement in, 2433 facial nerve palsy, 1232 dialysis-related, 1921, 2428t–2429t distribution by type of, 2429t familial, 1921–1932, 2440–2444. See also Amyloid neuropathy, hereditary. historical overview of, 1921–1922, 2427 localized, 1921, 2429t microscopic appearance and ultrastructure of amyloid, 2427–2428 nerve biopsy in, 2430 oculoleptomeningeal, 1924 primary, 1921, 2428t–2429t, 2428–2440 cardiovascular involvement in, 2433 clinical features of, 2430–2431 diagnosis of, 2435–2437, 2436f, 2437t frequency of, 2429–2430 laboratory findings in, 2435, 2435f other neurologic features of, 2433 other organ involvement in, 2433–2434, 2434f peripheral neuropathy in, 2430–2433. See also Amyloid neuropathy. prognosis for, 2437, 2437t, 2438f renal disease in, 2433 systemic symptoms and physical findings in, 2430, 2430f treatment of, 2437–2440 chemotherapy for, 2437–2439 definition of response to, 2439 general therapeutic measures for, 2439–2440 stem cell transplantation for, 2439 secondary, 1921, 2428, 2428t–2429t leprosy and, 2087 senile, 2428t–2429t stains for, 2428, 2428t systemic forms of, 1921, 2428t with monoclonal proteins, 2255, 2259–2260, 2260t, 2265, 2265t, 2269t, 2435, 2435t vs. POEMS syndrome, 2462t Amyotonia congenita, 1604t Amyotrophic lateral sclerosis (ALS), 1304–1306, 1533–1540 age at onset of, 1534 anti-neurofilament antibodies in, 586 axonal transport in, 395 transgenic mouse models of, 401–402, 1588–1589 clinical features of, 1536 diagnosis of, 1537 certainty of, 1537 delivering to patient, 1537 differential diagnosis of, 1536 polyneuropathy, 1176 electromyography in, 944f, 952f, 954f macro, 992f–993f, 996 scanning, 996f, 996–997 single-fiber, 990, 993f, 996f, 996–997 epidemiology of, 1534 etiology of, 1534–1535 environmental risk factors, 1535 genetics factors, 1534–1535 hypotheses of, 1538–1539 viruses, 1535 extramotor involvement in, 1536–1537 autonomic dysfunction, 1537 dementia, 1537, 1600 emotional lability, 1536 executive dysfunction, 1537

Index Amyotrophic lateral sclerosis (ALS) (Continued) familial, 1534–1535, 1595–1600 ALS type 2 (ALS2 mutations), 1535, 1598–1599 ALS type 4 (SETX mutations), 1599 ALS/parkinsonism-dementia complex, 1600 apolipoprotein E4 allele in, 1535 autosomal dominant, loci linked to, 1600 ciliary neurotrophic factor gene in, 1535 early descriptions of, 1595 genetic counseling about, 1596 linkage analysis studies in, 1596 neurofilament heavy subunit in, 1534, 1599 pathology of, 1595–1596 patterns of inheritance of, 1596, 1597t recurrence risk for, 1535 screening for, 1596 SOD1 mutations in, 1534–1535, 1539, 1586, 1596–1598 survival motor neuron genes in, 1535 vascular endothelial growth factor polymorphisms in, 1534–1535 with dynactin mutation, 1599 with predominantly lower motor neuron findings, 1299 gender distribution of, 1534 historical overview of, 1533 incidence of, 1534 lead exposure and, 2534 lower motor neuron involvement in, 1299, 1304–1305, 1533, 1538 lymphoproliferative disorders and, 2494–2495 management of, 2633–2634 in intensive care unit, 2607–2617. See also Intensive care unit (ICU). mechanical ventilation, 2633 neurotrophic factors, 380–381, 383 quality of life and, 2633–2634 transgenic models in screening for new therapeutic agents, 1590–1591 monoclonal gammopathy and, 2269–2270 monomelic, 1536 morphometric studies in, 770f nerve excitability studies in, 126 pathogenesis of, 1306 pathology of, 1305f, 1305–1306, 1537–1538, 1595–1596 axonopathy, 1538 Bunina bodies, 1305, 1538 chromatolysis, 686, 1305 Lewy body–like inclusions, 1538 neurofilaments, 1295, 1305, 1306, 1538, 1539, 1588 ubiquitin-immunoreactive inclusions, 1538 prevalence of, 1534 progressive spinal muscular atrophy and, 1304 reduction of glial glutamate transporter EAAT2 in, 1588 respiratory failure in, 2633 sporadic, 1533–1540 synaptophysin in, 1305 transgenic animal models of, 1585–1591 apoptosis and, 1590, 1590f axonal transport, 401–402, 1588–1589 double transgenic models, 1589 in screening for new therapeutic agents, 1590–1591 neurofilament transgenic models, 1588–1589 SOD1 transgenic models, 1585–1588, 1586f

Amyotrophic lateral sclerosis (ALS) (Continued) treatments for, 1539–1540 anecdotal therapies, 1540 clinical trials of, 1540 future of, 1540 riluzole, 1539 supportive therapies, 1539–1540 upper motor neuron involvement in, 1533, 1538 Amyotrophic lateral sclerosis/ parkinsonism-dementia complex (ALS/PDC), 1600 Amyotrophic type of myelopathy hand, 1311, 1311f pathology of, 1313 Amyotrophy, neuralgic. See Brachial plexus neuropathy, immune (IBPN). AN. See Auditory neuropathy (AN). Anaerobic metabolism, 668 Anal canal, 279 afferent fibers to, 282–283 ectoderm and endoderm of, 280 embryology of, 280 Anal manometry, 310t for fecal incontinence, 308, 309f for obstructed defecation, 306 Anal pressure, 2627 in obstructed defecation, 286 Anal sphincter external, 279, 284, 2627 during defecation, 284 during micturition, 284 electromyography of, 288–289, 290t inadequate relaxation of, 286 innervation of, 279, 282, 2627 muscle fibers of, 284 nerve to, 28 squeeze response of, 284, 285 tonic activity of, 284 internal, 279, 283–284, 2627 adrenergic innervation of, 284–285 during micturition, 284 innervation of, 279 neurotransmitter effects on, 286t pharmacology of, 284–285, 286t tone of, 283 Analgesics for neuropathic pain, 2326, 2637–2638, 2640–2647 anticonvulsants, 2642–2644 antidepressants, 2641–2642 evidence base and clinical trials of, 2640–2641 HIV-associated sensory polyneuropathies, 2135–2136 opioids, 2645–2646 overview of, 2637–2638 postherpetic neuralgia, 2120 research and changing clinical practice for, 2646–2647 topical agents, 2644–2645 tramadol, 2646 Analytic studies of peripheral neuropathy, 1182–1184, 1183f Anatomy brachial plexus, 18, 19–22, 1339–1341, 1340f central nervous system–peripheral nervous system transitional zone, 67–77 cranial nerves, 1192–1195, 1194f lower urinary tract, 299–301 lumbosacral plexus, 1375–1377, 1376f medial brachial fascial compartment, 1368, 1369f Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

v Anatomy (Continued) myelinated nerve fibers, 42–62 nerve trunks and spinal roots, 35–42 peripheral nervous system gross, 18–32 microscopic, 35–77 skin, 870–872, 871f unmyelinated nerve fibers, 62–67 variation in sensation thresholds by anatomic site, 1067 ANCA (antineutrophil cytoplasmic antibody)–mediated vasculitis, 2346t, 2346–2347, 2347f, 2351, 2352 genetics of, 2364 treatment of, 2375t Andermann syndrome, 1738, 1770t, 1781–1782 Andersen’s disease, 709 Anesthesia congenital trigeminal, 1211 intrathecal, chromatolysis induced by, 686 paralysis after, 1363–1364 Aneurysms false, nerve injury from, 1518 lumbosacral plexus compression by, 1381, 1382f ocular palsy due to, 1201 Angiitis. See also Vasculitis. allergic granulomatous, 2337. See also Churg-Strauss syndrome (CSS). cutaneous leukocytoclastic, 2361–2362, 2405 classification and definition of, 2338t, 2339 incidence of, 2340t microscopic polyangiitis, 2337, 2339, 2353t, 2355 antineutrophil cytoplasmic antibodies in, 2352, 2355 classification and definition of, 2338t, 2339, 2340t genetics of, 2363 incidence of, 2340t relative prevalence of, 2341t sarcoidosis and, 2417 Angiography, pulmonary, 2616 Angiostrongyliasis, 2154t, 2162–2164 clinical features of, 2164 cranial nerve involvement in, 2164 etiologic agent in, 2162 life cycle of, 2162 ocular, 2164 pathology of, 2162–2164 transmission of, 2162 Angiotensin-converting enzyme (ACE), in sarcoidosis, 2419, 2422 Anhidrosis, 1150 congenital insensitivity to pain with, 704, 886, 1837 facial, in Horner’s syndrome, 207 in diabetic patients, 1959 Anisocoria, 212 in Adie’s syndrome, 211f, 211–212 in Horner’s syndrome, 205–207 simple, 209, 209f Ankle motion and moment during gait, 1099, 1100f Ankle-foot orthosis (AFO), 2623 Ankylosing spondylitis cauda equina syndrome and, 1334 vs. pediatric lumbosacral radiculopathy, 2738 Ankyrin ankyrin G, 415, 549 in CRASH syndrome, 468 of axoplasmic cortex, 47

vi Annexin-1, 567 Anorectal angle, 284, 285f Anorectum. See also Anal canal; Anal sphincter; Rectum. anatomy of, 280–281, 281f dysfunction in neurologic disorders, 286–292 fecal incontinence, 287–293, 290t, 293t obstructed defecation, 286–287, 287f embryology of, 279–280 function of, 279 innervation of, 279 nerve supply to, 281f, 281–283 afferent innervation, 282–283 parasympathetic nerves, 282 regulation by central nervous system pathways, 282 sympathetic nerves, 281–282 Anosmin-1, in Kallman’s syndrome, 469 Ansa cervicalis (hypoglossi), 18 Ansa subclavia, 17f, 23 Antebrachial cutaneous nerve lateral, 15f, 20f, 1470f, 1472, 1472f lesions of, 1471–1474 nerve conduction studies of, 916, 1473 medial, 15f, 20, 1470f lesions of, 1483–1484 nerve conduction studies of, 916 posterior, 22 Anterior horn cells, 1296, 1297 disease of degeneration in, 772 ␤,␤⬘-iminodipropionitrile-induced, 1296 in amyotrophic lateral sclerosis, 1305f, 1305–1306, 1537–1538 in amyotrophic type of myelopathy hand, 1313 in Hopkins syndrome, 2031 monoclonal gammopathy and, 2270 poliomyelitis, 1300–1301 vs. polyradiculopathy, 1327 somatotopic organization of, 1297 Anterior interosseous nerve syndrome, 1453–1454 in children, 2726 Antibodies, 563, 567, 2177–2190 anti-acetylcholine receptor, in myasthenia gravis, 833 anti-amphiphysin, 586, 2472t anti-CV2, 586, 595, 2472t, 2476 anti–endothelial cell antibody–mediated vasculitis, 2345–2346 anti-Hu, 584–585, 594, 715, 1173, 2472, 2472t, 2473–2474, 2474f, 2475, 2476 anti-La, 1172 anti-myelin, in experimental autoimmune neuritis, 619–620 antineutrophil cytoplasmic antibody–mediated vasculitis, 2346t, 2346–2347, 2347f, 2351, 2352, 2375t antinuclear, 1172–1173, 2358, 2472t anti-Ri, 2472t, 2476 anti-Ro, 1172 anti-Yo, 2472t, 2477 B-cell production of, 563 IgM monoclonal protein reactive with, 2266–2268 in chronic inflammatory demyelinating polyradiculoneuropathy, 2179t, 2239, 2245–2246 in connective tissue diseases, 1172–1173

Index Antibodies (Continued) in paraneoplastic neuropathy, 2471–2472, 2472t autonomic neuropathy, 2480–2481 motor neuropathy, 2479 sensorimotor neuropathy, 2476–2477 subacute sensory neuronopathy, 584–586, 2473–2474, 2474f, 2475 to heparan sulfate, 2190 to neurofilaments, 586, 2190 to trisulfated heparin disaccharide, 2190 to ␤-tubulin, 2190 with binding to gangliosides, 590, 591, 2182–2187 diseases related to, 596 GD1a ganglioside, 2186 GD1b gangliosides, 597, 2186–2187 GM1 ganglioside, 2182–2185 IgG anti-GM1 antibodies, 2184–2185 IgM anti-GM1 antibodies, 2182–2184 GM1b ganglioside, 2186 GM2 and GalNac-GD1a gangliosides, 2186 GQ1b ganglioside, 1173, 2185 GT1a ganglioside, 2185 in Guillain-Barré syndrome, 590, 596, 2285 in multifocal motor neuropathy and conduction block, 596, 2284–2286, 2286f, 2288–2290 infection and, 590–591 molecular mimicry, 591–592 with binding to peripheral nerve glycoproteins and glycolipids, 1173, 2187–2190 galactocerebroside, 2190 GALOP antigen, 2189 myelin-associated glycoprotein, 1173, 2187–2188, 2255 PMP22 and P0, 2188–2189 sulfated glucuronyl paragloboside, 2188 sulfatide, 2189 Antibody testing, 1172–1173, 1174t, 2177–2182 clinical validation of, 2181 cost-effectiveness of, 2181 in cryptogenic sensory polyneuropathy, 2326 methodology for, 2177–2178 enzyme-linked immunosorbent assay, 2178–2180 immunocytochemistry, 2180 immunofixation electrophoresis, 2181 thin-layer chromatography, 2180–2181 Western blot, 2180 pathogenic significance of, 2181–2182 varying prevalence and, 2181 Anticholinesterase agents, for myasthenia gravis, 860 Anticoagulant therapy, electromyography in patient receiving, 938 Anticonvulsants, for neuropathic pain, 2642–2644 in diabetic patients, 1971, 1971t mechanisms of action of, 2642 non-sodium channel–blocking, 2644 sodium channel–blocking, 2643 Antidepressants for neuropathic pain, 2640, 2641–2642 for patients with amyotrophic lateral sclerosis, 1539–1540 Antidiarrheal agents, 293t Antigen-presenting cells (APCs), 559, 560f, 574, 609, 610 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Antigens, 573–598 antigen compartments of peripheral nervous system, 575, 576f autoantigens in animal models, 593–594, 611t, 611–612 galactocerebroside-induced EAN, 593 ganglioside-induced neuropathy, 594 MAG-induced neuropathy, 594 P0 EAN, 593 P2 EAN, 593–594 PMP22 EAN, 593 autoantigens in human neuropathies, 594–598 axonopathies, 594–595 channelopathies, 597–598 Charcot-Marie-Tooth disease, 596 galactocerebroside, 596 ganglioside antibody–related diseases, 596 inflammatory demyelinating polyradiculoneuropathies, 595–596 neuronopathies, 594 paraproteinemic neuropathies, 596–597 protein antigens, 597 calcium and potassium channels and, 590 ganglioside, 586–590 in Guillain-Barré syndrome, 2200t infection and, 590–591 molecular mimicry, 591–592 microbial, 590–591 neurofilament, 586 neuronal, 584–586 of T and B cells, 574 Schwann cell and myelin, 577–584. See also Proteins. Antigen-specific therapy, for experimental autoimmune neuritis, 625–626, 626f Antimicrobial therapy for chronic intestinal pseudo-obstruction, 2665 for diarrhea, 2665 in diabetic patients, 1960 for leprosy, 2103–2104 for Lyme disease, 2114 for neurosyphilis, 1316 to suppress bacterial overgrowth in gastroparesis, 2691 Antineutrophil cytoplasmic antibody (ANCA)–mediated vasculitis, 2346t, 2346–2347, 2347f, 2351, 2352 genetics of, 2364 treatment of, 2375t Antioxidants, 510–511, 511t for protection against ischemia-perfusion injury, 524, 526f, 676 for treatment of experimental diabetic neuropathy, 516t, 516–518, 517f reduced levels of in diabetic peripheral nerve, 514–515, 515t, 673 in experimental diabetes, 512, 515, 515t in ischemia, 521 Antiretroviral toxic neuropathy (ATN), 2131–2136 clinical features of, 2131–2132 electrophysiologic findings in, 2132 incidence of, 2131, 2171f pathogenesis of, 2133–2135, 2134f–2135f pathology of, 2132 treatment of, 2135–2136 Antisulfatide neuropathy, 2179t, 2180t

Index Antiviral therapy for Bell’s palsy, 1230–1231, 2118 for cytomegalovirus infection, 2122, 2123 for herpes zoster, 2120 for Ramsay Hunt syndrome, 1232 for vasculitic neuropathy, 2336, 2376–2378 Aortic arch, baroreceptors in, 323 Aortic plexus, 25, 31, 281f Aorticorenal ganglion, 31 APCs (antigen-presenting cells), 559, 560f, 574, 609, 610 Apolipoproteins, 1905, 1907t apo A-I amyloidosis, 1921, 1922, 1931–1932 apo E4 allele in amyotrophic lateral sclerosis, 1535 clinical phenotypes of apo A-I deficiency diseases, 1909, 1910t in familial high-density lipoprotein deficiency, 1912 in Tangier disease, 1906, 1908, 1912 Apomorphine, for erectile dysfunction, 318 Apoptosis, 695–699, 697f–698f biochemical events in, 700–701 caspases in, 700f, 701, 1590, 1590f in amyotrophic lateral sclerosis, transgenic models of, 1590, 1590f regulation of, 701 regulation of Schwann cell survival, 354f, 354–356 triggers for, 700 AR (axoplasmic reticulum), 47 Arachnoid, 1324f Arachnoiditis chronic, 1332–1333 ossificans, 1333 Arcade of Frohse, 1474 Arcade of Struthers, 1478 Argyll Robertson pupil, 205, 212 Arnold’s nerve, 1277 Arp2/3, in actin nucleation, 452 Arrector pilorum muscles, 870 innervation of, 871 quantitation of, 881 Arrhythmias, cardiac, 226–227, 2611 in glossopharyngeal neuralgia, 1276, 1276f in Guillain-Barré syndrome, 2626 in trichinosis, 2164 vagal efferent activity and, 1281 Arsenic neuropathy, 1168, 2064–2065, 2535–2538 clinical features of, 2535–2536 laboratory studies in, 2536–2537 pathology of, 2537–2538 treatment of, 2538 use and sources of exposure to arsenic, 2535 ART18 monoclonal antibody, for experimental autoimmune neuritis, 623 Artemin, 179 Arterial pressure. See Blood pressure (BP). Arteries. See also specific arteries. anatomy of, 801 supplying nerves, 800–801, 803f Arteriovenous fistula, nerve injury from, 1518 Arteritis. See Vasculitis. Arthritis in giardiasis, 2167–2168 in Lyme disease, 2112 in sarcoidosis, 2420 rheumatoid. See Rheumatoid arthritis (RA). Arthritis Impact Measurement Scale (AIMS2), 1058–1059, 1060 Arylsulfatase A deficiency, 710, 1847, 1852–1853, 2715

Ashkenazi Jews, familial dysautonomia in, 886, 1832–1833 ASICs (acid-sensing ion channels), 187t, 188 ASN (Axon–Schwann cell network), 53–54, 417 Aspartylglucosaminuria, 1864 Aspirin for deep venous thrombosis prophylaxis, 2609 for erythromelalgia, 1175, 2498 topical, for postherpetic neuralgia, 2644 Association studies, 1184 Asthma, vs. childhood Guillain-Barré syndrome, 2717 Ataxia calcium channel mutations in, 590 carbohydrate-deficiency, 1738 Cayman Island, 1738 Friedreich’s, 705–706, 1738–1739. See also Friedreich’s ataxia. hereditary, in children, 2723–2724 in chronic inflammatory demyelinating polyradiculoneuropathy, 2224 in Miller Fisher syndrome, 2185 in nonmalignant inflammatory sensory polyganglionopathy, 2311 in tropical ataxic neuropathy, 2068 locomotor, 2310 neuropathy, ataxia, and retinitis pigmentosa (NARP), 1941, 1942, 1946–1947 sensory ataxic neuropathy, dysarthria, and ophthalmoparesis (SANDO), 1946 spastic, of Charlevoix-Saguenay, 1738 spinocerebellar. See Spinocerebellar ataxia (SCA). Ataxia-telangiectasia, 705, 1738, 1741 in children, 2724 Ataxic neuropathy, 1147 Ataxin-1, 706 Ataxin-3, 706 ATN. See Antiretroviral toxic neuropathy (ATN). Atopic dermatitis, skin biopsy in, 888 ATP. See Adenosine triphosphate (ATP). Atrioventricular node, 217 Atropine blockade of sweat response by, 333 for carbamate neurotoxicity, 2509 for carotid sinus hypersensitivity, 1277 for organophosphate neurotoxicity, 2516 Audiometry, 1255–1256 pure tone, 1255–1256 in hereditary sensorimotor neuropathy, 1257f in patient with bilateral eosinophil granulomas of petrous bones, 1266f speech, 1256 Audit of Diabetes-Dependent Quality of Life (ADDQoL), 1058 Auditory brainstem responses (ABRs), 1256 age-related changes in, 1254 in auditory neuropathy, 1260 in hereditary sensorimotor neuropathy, 1258f in patient with bilateral eosinophil granulomas of petrous bones, 1267f in X-linked Charcot-Marie-Tooth disease, 1797 Auditory nerve, 1255f Auditory neuropathy (AN), 1259–1261 acquired, 1259–1261 diagnosis of, 1259–1260 etiology of, 1260–1261, 1261t histopathology and pathophysiology of, 1261, 1262f management of, 1261–1262 inherited, 1259 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

vii Auerbach’s plexus, 18, 252, 252f, 2627 congenital absence of enteric neurons in, 705 in regulation of gut motility, 2683 Auricular nerve, 1223f posterior, 1220 Auriculotemporal syndrome, 1283 Autoantibodies. See Antibodies. Autografts, nerve, 1423–1424 alternatives to, 1423–1424 allografts, 1424 conduits, 1424 for long interstump gap, 1424 rationale for use of, 1423 Autoimmunity, 567–569, 574, 575 B-cell tolerance, 567 breakdown of tolerance, 568–569 in Lambert-Eaton myasthenic syndrome, 834–838 in myasthenia gravis, 833–834 macrophages in, 560 T-cell tolerance, 567, 568f Autonomic dysfunction, 1141, 1150–1151, 1166, 2653–2665 bladder dysfunction, 2626–2627, 2663–2664 cardiovascular, 217, 226–227, 2653–2654, 2671–2678. See also Cardiovascular autonomic failure; Orthostatic hypotension (OH), neurogenic. in diabetic patients, 1957, 1958 overview of, 2671–2673, 2672f, 2673f treatment of, 2626, 2630, 2656–2657, 2673–2681 detection of, 1103–1104 evaluating response to therapy for, 1105 gastrointestinal symptoms of, 2664–2665, 2685–2694 constipation, 2610, 2665, 2691–2692, 2665 diarrhea, 2616, 2616t, 2665, 2692–2694 fecal incontinence, 287–292, 2627–2628, 2665, 2694, 2665 gastroparesis and chronic intestinal pseudo-obstruction, 259–260, 2665, 2688–2691 in idiopathic orthostatic hypotension, 2696 in multiple system atrophy, 2696 in pandysautonomia/selective dysautonomias, 2695–2696 megacolon, 2692 oropharyngeal dysphagia, 2633, 2685–2687 impotence, 2628, 2664 in acute inflammatory demyelinating polyneuropathy, 2200 in amyloidosis, 1923, 2431 familial, 2441 in amyotrophic lateral sclerosis, 1537 in chronic obstructive pulmonary disease–associated neuropathy, 2022 in cryptogenic sensory polyneuropathy, 2322, 2324 in diabetic neuropathy, 1956–1960 pathology of, 1967–1968 in familial dysautonomia, 1831–1833, 2722 in Guillain-Barré syndrome, 2608, 2611, 2630 in children, 2716 in hereditary motor and sensory neuropathy, 1635–1636 in hereditary sensory and autonomic neuropathies, 1809–1839 in Lambert-Eaton myasthenic syndrome, 834 in Leigh disease, 2715 in leprosy, 2093

viii Autonomic dysfunction (Continued) in lumbosacral radiculoplexus neuropathy, 1999 diabetic, 1996–1999 in neuropathy associated with chronic hepatic disease, 2018 in primary biliary cirrhosis–associated sensory polyneuropathy, 2021 in Sjögren’s syndrome, 1309 in uremic neuropathy, 2025 intensive care unit care for, 2611 monitoring course of, 1105 morphometric studies in, 771f spinal ganglion somas, 756, 760f neurogenic orthostatic hypotension, 2653–2663 paraneoplastic, 2479–2481 quantitation of, 1103–1127. See also Autonomic function tests. Autonomic dysreflexia, 2664 Autonomic function tests, 1103–1127 autonomic reflex laboratory for, 1105 baroreflex sensitivity determination, 1123–1124 beat-to-beat blood pressure recordings, 1111–1113 response to head-up tilt, 1112–1113 response to prolonged tilt, 1123 response to Valsalva maneuver, 1111–1112, 1112f biochemical and hormonal evaluation, 1116–1119 plasma catecholamines, 1116–1118, 1117f–1120f total body and regional noradrenaline spillover, 1118–1119 urinary catecholamines and metabolites, 1118 central autonomic pathways, 1125–1127, 1126f–1127f clonidine–growth hormone stimulation test, 1126–1127, 1127f neuroimaging, 1125–1126, 1126f Composite Autonomic Severity Score, 1113t, 1113–1116 heart rate response to deep breathing, 1114–1115, 1115t normative database for, 1114, 1114t orthostatic blood pressure, 1115–1116, 1116t QSART, 1114, 1114t–1115t Valsalva ratio, 1115, 1115f–1116f, 1116t denervation supersensitivity determination, 1124 blood pressure and heart rate response to intravenous tyramine, 1124 discontinuing medications before, 1105–1106 electrophysiologic assessment of sympathetic neural activity, 1121–1122, 1121f–1122f evaluation of splanchnic-mesenteric vasculature, 1124–1125, 1125f goals of, 1105 heart rate recordings to evaluate cardiovagal function, 1109–1111 comparison of tests, 1110–1111 other tests, 1123 patient preparation for, 1109 response to deep breathing, 1109, 1109t, 1110f response to prolonged tilt, 1123 response to standing, 1110 Valsalva ratio, 1109–1110, 1110t

Index Autonomic function tests (Continued) imaging of sympathetic innervation to organs, 1119, 1120f MIBG SPECT study, 1120–1121 in cryptogenic sensory polyneuropathy, 2324 indications for, 1103–1105, 1104t isoproterenol infusion test, 1124 nonroutine tests, 1119–1127 orthostatic stress tests, 1123 overview of, 1103, 1104t patient preparation for, 1105–1106 quantitative sudomotor axon reflex test, 1106–1109 commercial equipment for, 1108 interpretation of, 1108t, 1108–1109 physiologic basis for, 1106–1108, 1107f recording sites for, 1108 reproducibility of, 1108 sweat volumes recorded from same and distant compartments in, 1106, 1107t routine tests, 1106–1119 salivation test, 1116 sequence of, 1106 skin vasomotor reflexes, 1122–1123 vs. nerve biopsy in diabetic neuropathy, 736 Autonomic nervous system, 15–18. See also Enteric nervous system (ENS); Parasympathetic nervous system; Sympathetic nervous system. age-related changes in autonomic neurons, 492–495, 493f–495f, 702 mechanisms of, 498–500, 499f anatomy of, 15–18 assessing function in gastrointestinal system, 2695 cardiac, 217–227 definition of, 15 historical identification of, 222 in abdomen and pelvis, 30–31 in blood pressure regulation, 323, 667, 2654 in head, neck, and upper limb, 22–23 in lower urinary tract, 299–314 in sex organs, 314–318 in thorax, 23–25, 24f, 25 neurobiology of ganglia of, 233–245 organization of, 15 Autonomic neuropathy, 1141 Epstein-Barr virus and, 2121 in HIV infection, 2138 in mitochondrial disease, 1947 management of, 2626 symptoms of, 2626 Autonomic reflex laboratory, 1105 Autonomic symptom scores, 1036 Autosomal dominant inheritance, 1546–1547 Autosomal recessive inheritance, 1547–1548 Autosomal recessive neuropathy, 1775–1776, 1776f Axillary arteriography, medial brachial fascial compartment syndrome after, 1368 Axillary artery, 1341 Axillary hematoma, brachial plexus injury from, 1370 Axillary nerve, 1340f distribution of, 15f, 22, 22f, 1466f, 1469 involvement in vasculitic neuropathy, 2365t muscles innervated by, 956t, 1469 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Axillary neuropathy, 1469–1471 clinical manifestations of, 1470 diagnosis of, 1470–1471 etiologies of, 1469–1470 iatrogenic, 1470 in children, 2727 prognosis for, 1471 traumatic, 1469–1470 treatment of, 1471 Axolemma E face of, 47 function of, 47 intramembranous particles of, 47 juxtaparanodal, 54 microscopic anatomy of, 45–47 nodal, 415–416, 415f–416f P face of, 47 paranodal, 417, 549 Axon(s), 35 age-related changes in, 486–492 functional impairments, 486 mechanisms of, 496–498 myelinated axons, 483–484, 487–491, 489f–492f recovery after injury, 498 selective vulnerability to, 497 structural impairments, 486–487, 488f–489f unmyelinated axons, 491–492 anatomic sites of origin of, 1296 caliber of, 433–443 extrinsic factors influencing, 435–436 myelin-associated glycoprotein, 436 myelination, 435f, 435–436 signaling pathways from myelinating Schwann cell to axonal cytoskeleton, 436 functional correlates of, 433 in PMP22-deficient mice, 543 influence on myelination, 436–438, 438f internodal length and, 414 intrinsic factors influencing, 439–443 alterations in velocity of neurofilament transport, 441f, 441–442 in large-caliber myelinated fibers, 439–440, 440f neurofilament gene expression, 442–443 neurofilament transport, 440 myelin thickness and, 414–415 neurofilament deficiency and, 547 normal range of, 433 paranodal regions and, 417 Schwann cells and, 411 timing of myelination and, 411 cytology of, 433–435, 434f cytoskeleton of, 433 effects of ischemia on, 119–121, 121f, 125–126, 804, 805f effects of neurofilament deficiency on, 547–548 excitability of. See Nerve excitability studies. glycogenosomes in, 487, 488f–489f guidance to targets, 447–469 actin dynamics in, 450f–451f, 450–453 chemoaffinity theory of, 447 growth cone cytoskeletal dynamics in, 448f–449f, 448–450 in development, 453–465 ephrin/Eph signaling, 453–456, 454f netrin/DCC/UNC-5 signaling, 459–462, 460f semaphorin/plexin signaling, 462–465, 463f Slit/Robo signaling, 456–459, 458f

Index Axon(s) (Continued) in disease, 465–469 CRASH syndrome/MASA syndrome/L1 syndrome, 467–468 epilepsy, 465–466 Kallmann’s syndrome, 468–469 schizophrenia, 466–467 ligand-receptor systems for, 447, 453 outlook for, 469 in diphtheria, 2150 in Remak fiber, 36 injury of, 1403 internode of. See Internode(s). ion channels on, 95 potassium channels, 96f, 99–104 sodium channels, 96f, 96–98 length of, 667 modulation of discharge frequency by, 1007 effect of hyperpolarization, 1007, 1007f–1008f unidirectional block, 1007, 1009f myelinated, 42, 45–51 age-related changes in, 483–484, 487–491, 489f–492f at central nervous system–peripheral nervous system interface, 68–70, 73f axolemma of, 45–47 axoplasm of, 47–51 organization of, 45–51, 46f, 48f, 50f periaxonal space of, 45 of dorsal root ganglion cells, 1296 of P0-deficient mice, 537, 538f paranode of. See Paranode(s). pathologic alterations of, 812–819 acute axonal degeneration, 819 acute neuronal degeneration, 816–818 agenesis and maldevelopment, 816, 817f anterograde and retrograde effects, 816 axonal atrophy, myelin remodeling, and axonal degeneration, 819 axonal spheroids, distal axonal atrophy, myelin remodeling, and axonal degeneration, 819 by functional classes of neurons, 813–814 by levels within neurons, 814f, 814–815 by segmental level, 812–813 chronic neuronal degeneration, axonal spheroids, axonal atrophy, myelin remodeling, and axonal degeneration, 818–819 classification of, 816 temporal selectivity of neurons, 815–816 wallerian degeneration, 819 regrowth of, 1404, 1421–1423 across interstump gap, 1421f, 1421–1422 cellular outgrowth into regeneration chamber, 1422 precision of reinnervation, 1422–1423, 1423f rate of, 1411 response to injury, 1463–1464, 1512 degeneration, 1406–1407, 1407f delayed degeneration, 1408 role of Ca2⫹ in, 1407–1408, 1408f sealing of severed ends, 1406, 1407f unmyelinated, 62–67 age-related changes in, 491–492 at central nervous system–peripheral nervous system interface, 70–71 development of, 64–65, 64f–65f diameter of, 64 historical studies of, 62–64 morphometry of, 66–67, 66f–67f

Axon(s) (Continued) organization of, 65 relationship with Schwann cells, 66 spatial relationships between Schwann cells, collagen and, 66 visualization of, 64 velocity of electrical impulses conducted by, 36 Axon bundles, 411, 412f Axon reaction. See Chromatolysis. Axon reflex, 1007 Axonal atrophy, 487–491, 497 axonal spheroids, myelin remodeling, axonal degeneration and, 819 chronic neuronal degeneration and, 818–819 myelin remodeling, axonal degeneration and, 819 secondary segmental demyelination and, 780–788, 784f–786f in Friedreich’s ataxia, 783, 785f in uremic neuropathy, 781 permanent axotomy studies of, 787–788, 787f–790f retrograde effects of nerve transection, 784–786 Axonal constriction, 1512 Axonal degeneration, 1154, 1166, 1405, 1406–1407, 1407f acute, 819 age-related, 487, 488f axonal atrophy, myelin remodeling and, 819 axonal spheroids, distal axonal atrophy, myelin remodeling and, 819 chronic neuronal degeneration, axonal spheroids, axonal atrophy, myelin remodeling and, 818–819 delayed, 1408 demyelinative internal strangulation and, 777 early cooling for slowing of, 1410 in acrylamide neuropathy, 2507, 2508 in acute inflammatory demyelinating polyneuropathy, 2204 in amiodarone neuropathy, 2555 in amyloidosis, 2430 in arsenic poisoning, 2537 in brachial plexus lesions, 1344 in chronic inflammatory demyelinating polyradiculoneuropathy, 2239, 2246 in cobalamin deficiency, 1308 in colchicine neuropathy, 2557 in critical illness polyneuropathy, 2027–2028, 2028f in dapsone neuropathy, 2558 in diabetic neuropathy, 1965 in disulfiram neuropathy, 2558 in eosinophilia-myalgia syndrome, 2571 in gold neuropathy, 2544 in hereditary motor and sensory neuropathy with P0 mutation, 1695, 1698–1699 in HIV-associated sensory polyneuropathies, 2132 in hypothyroid polyneuropathy, 2042 in isoniazid neuropathy, 2560 in lysosomal disorders, 1846 in metronidazole neuropathy, 2560 in misonidazole neuropathy, 2561 in niacin deficiency, 2057 in nitrofurantoin neuropathy, 2562 in nonmalignant inflammatory sensory polyganglionopathy, 2313 in organophosphate-induced delayed peripheral neuropathy, 2516–2517 in podophyllin neuropathy, 2566 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

ix Axonal degeneration (Continued) in POEMS syndrome, 2457, 2458 in primary biliary cirrhosis–associated sensory polyneuropathy, 2021 in pyridoxine neuropathy, 2566 in severe demyelination, 777 in Tangier disease, 1905 in taxane neuropathy, 2569 in uremic neuropathy, 2024 in vinca alkaloid neuropathy, 2572 induction by intermittent cooling and rewarming, 1405, 1406f nerve conduction studies in, 902–904, 903f perineurial cells in, 40 teased fiber analysis of, 745f, 748, 749f wallerian, 4, 772–776, 780f–781f, 819, 1405–1406. See also Wallerian degeneration. wallerian-like, 1406 within intact Schwann cell tube, 1405 Axonal dystrophy, 487, 497, 778–780 axonal spheroids in, 778–780 definition of, 778 diseases associated with, 778 giant, 705, 778, 782f, 1736–1738 historical background of, 778 in vitamin E deficiency, 497, 2056 infantile and juvenile, 1744 neurofilaments in, 705, 778 pathology of, 705 skin biopsy in, 888 Axonal neuropathy acute motor. See Acute motor axonal neuropathy (AMAN). autosomal recessive, 1741–1742 chromatolysis and, 686 giant, 705, 888, 1782–1783, 1783f mitochondrial respiratory chain dysfunction and, 1946–1947 with loss of fast-conducting fibers, 903 Axonal regeneration, 1404, 1421–1423. See also Nerve regeneration. across interstump gap, 1421f, 1421–1422 cellular outgrowth into regeneration chamber, 1422 precision of reinnervation, 1422–1423, 1423f rate of, 1411 regeneration stagger, 1421 teased fiber analysis of, 748, 751f vs. axonal regrowth, 1404 Axonal spheroids. See Axonal dystrophy. Axonal swellings. See Axonal dystrophy. Axonal transport, 36, 387–402, 667 age-related changes in, 496–497 chromatolysis and, 390, 686–687 components of, 387–388 definition of, 387 evidence for abnormalities in human neuropathies, 394–395, 394f–395f fast anterograde, 387, 388–389, 388f–389f cargoes of, 389 fast retrograde, 387, 389–390 historical studies of, 387 in acrylamide neuropathy, 2508 in amyotrophic lateral sclerosis, 395, 1588–1589 in compression neuropathies, 1399 in diabetic neuropathy, 1978

x Axonal transport (Continued) in experimental neuropathies, 396–400, 397t acrylamide, 396–398 p-bromophenylacetylurea, 400 diabetes, 396 hexacarbons, 399 ␤,␤⬘-iminodipropionitrile, 398–399 zinc pyridinethione, 399–400 in genetically based animal neuropathies, 400–402 in lysosomal disorders, 1846 intracytoplasmic vacuolation and, 688 motors for, 388, 390–391, 391f dyneins, 391 kinesins, 391 overview of, 387–388 physiologic role of, 387 potential causes of defects of, 393–394 cytoskeletal disturbances, 393–394 disturbed ionic balance, 394 impaired glycolysis or oxidative phosphorylation, 393 ischemia, 393 local pressure, 393 toxic effects on transport motors, 394 slow, 387, 390 stages of, 392, 402, 402f theoretical implications of abnormalities of, 392–393 anterograde transport defects, 392 loading stage defects, 392 retrograde transport defects, 393 turnaround defects, 393 video-enhanced contrast–differential interference contrast microscopy studies of, 389, 395, 395f Axonin-1, in CRASH syndrome, 468 Axonopathy(ies), 1137 autoantigens in, 594–595 Axonotmesis, 1404, 1512 brachial plexus injury with, 1344, 2631 nerve conduction studies in, 902–903, 903f Axon–Schwann cell interactions, 35–36, 1408 functions of axonally derived signals, 1408 functions of Schwann cell signals, 1408 in developing peripheral nervous system, 4–5 in hereditary motor and sensory neuropathy, 1646–1648, 1651f–1652f related to P0, 1698–1699 potassium channels in, 106 role of axolemma in, 47 techniques for study of, 535–536 Axon–Schwann cell network (ASN), 53–54, 417 Axoplasm, 433–435, 47–51 components of, 47 cortex of, 47 cytoskeleton of, 49–51 microtrabecular matrix, 49–51 microtubules, 49 neurofilaments, 49 granular material of, 51 microscopic anatomy of, 47–51 microtubule domain of, 51 neurofilament domain of, 51 of constricted axon segment, 59–60 organelles of, 47–49 dense lamellar bodies, 47 membranous cisterns, 49 mitochondria, 47 multivesicular bodies, 47 reticulum, 47 vesicotubular profiles, 47 subaxolemmal domain of, 51

Index Axoplasmic reticulum (AR), 47 Axotomy chronic, 1411f, 1411–1412 nerve growth factor expression after, 177, 694 peripheral neuron pathology induced by, 684–702 biochemical alterations, 689–695 chromatolysis, 684–687 intracytoplasmic vacuolation, 687–689 neuron death, 695–702 retrograde effects of, 784–786 sequential changes in myelinated fibers after, 787–788, 787f–790f AZA. See Azathioprine (AZA). Azathioprine (AZA), 639–640 adverse effects of, 640 for chronic inflammatory demyelinating polyradiculoneuropathy, 640, 2235–2236 in children, 2719 in diabetic patients, 1955 for myasthenia gravis, 639–640 for sarcoidosis, 2423 for vasculitis, 2337 interaction with allopurinol, 639–640 mechanism of action of, 639 pharmacokinetics of, 639–640 use in neuropathy, 640 AZT (zidovudine), 2130

B cells, 574, 609–610 activation of, 563 antibody production by, 563 tolerance of, 574, 610 B7-1, 563 B7-2, 563 Back, innervation of, 31–32 Baclofen for hemifacial spasm, 1245 for trigeminal neuralgia, 2641 Bacteremia, 2026 Bacteria–related vasculitis, 2351 BAERs (brainstem auditory evoked responses), in X-linked Charcot-Marie-Tooth disease, 1797 Bag fibers, 137, 138 Baker’s cyst, tibial neuropathy and, 1501–1503 BAL. See British antilewisite (BAL). Balamuthia infection, 2161 Balance, 1253 problems with, 483 tests of, 1256 Bands of Bünger, 65, 66, 776, 781f, 793, 1405, 1413, 1422–1423 Bannwarth’s syndrome, 1331. See also Lyme disease. Baroreceptors control of muscle sympathetic nerve activity by, 325 effect of glossopharyngeal and vagus nerve block on, 325, 326f, 1274, 1275f modified Oxford test of, 325, 325f innervation of, 323 orthostatic hypotension and, 2656 threshold for function of, 324 Baroreflex in blood pressure regulation, 323–324, 324f, 2671, 2672f in muscle sympathetic nerve activity, 325–326, 325f–326f resetting during exercise, 328–329, 329f sensitivity determination, 1123–1124 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Basal lamina, 577 antigens of, 576 axon–Schwann cell interactions in formation of, 5 connection between myelinating Schwann cells and, 424 in Schwann cell development, 362–363 Basilar artery, 1220 Batten disease, 708 adult form of, 708 gene mutations in, 708 pathology of, 708 BBE (Bickerstaff’s brainstem encephalitis), 2185, 2205 B-cell receptor, 563 B-cell tolerance, 567 Bcl-2, 1590 BDNF. See Brain-derived neurotrophic factor (BDNF). Becker muscular dystrophy, 1545 Behavioral therapy, for neurogenic bladder, 2663 Behçet’s disease genetics of, 2363 vasculitic neuropathy and, 2340t, 2355t, 2361 Bell’s palsy, 1142, 1223–1231 age distribution of, 1223–1224 bilateral, 1225–1226, 1226f disorders associated with, 1225 recurrent, 1225–1226 blink reflex in, 921t, 1226–1227 clinical features of, 1224–1226 congenital/pediatric, 1236–1237 diabetes mellitus and, 1225, 1234 electromyography in, 1226, 1227 electroneurography in, 1226, 1227 epidemiology of, 1223 family history of, 1236 gender distribution of, 1223 herpes simplex virus infection and, 1228–1229, 2117, 2118 in diabetic patients, 1961 incidence of, 1223 Lyme disease and, 1226f, 1229–1230 maximal stimulation test in, 1226 natural history of, 1224–1225 nerve conduction studies in, 908 nerve excitability testing in, 1226, 1227 neuroimaging in, 1226f, 1228, 1228f neuropathology of, 1230 prognosis for, 1225, 1231 prognostic testing in, 1227 recurrent, 1225 seasonality of, 1223 sequelae of, 1225 tests for detection of, 1227–1228 treatment of, 1230–1231 medical therapy, 1230–1231 glucocorticoids, 637, 1230–1231 surgical decompression, 1231 Bence Jones proteins, 1174, 2257, 2258 amyloid and, 2428 cryoglobulins, 2257 Benedikt’s syndrome, 1201 Beriberi, 2051–2053, 2072 causes of, 2051, 2072 clinical features of, 2052 diagnosis of, 2052 dry, 2052 epidemiology of, 2072 historical overview of, 2051–2052 management of, 2052–2053 pathogenesis of, 2052 wet, 2052, 2073

Index Bernard-Horner syndrome, 1522, 1525 Betamethasone, for radiculopathy due to disc herniation, 1326 Bethanechol effect on anal sphincter, 286t for neurogenic bladder, 2663 B-FABP (brain-specific fatty acid–binding protein), 343, 344f, 344–345, 347 Bickerstaff’s brainstem encephalitis (BBE), 2185, 2205 Bilharziasis. See Schistosomiasis. Biochemistry changes after peripheral axotomy, 689–695 evaluation in autonomic function tests, 1116–1119 plasma catecholamines, 1116–1118, 1117f–1120f total body and regional noradrenaline spillover, 1118–1119 urinary catecholamines and metabolites, 1118 neuronal death and, 700f, 700–701 of lysosomal storage disorders, 1846 of Riley-Day syndrome, 1833 Biofeedback therapy, for fecal incontinence, 292, 293t Biopsy. See Nerve biopsy; Skin biopsy. Bizarre high frequency discharges, 943 Bladder. See also Urinary tract, lower. anatomy of, 299–300 detrusor region of, 299 innervation of, 301–306, 302f afferent pathways, 305 neural modulatory mechanisms, 304 parasympathetic pathways, 301–302, 303f somatic pathways, 304–305 sympathetic pathways, 302–304 urothelial-afferent interactions, 305–306 neurotransmitter release by urothelial cells of, 300 pressure within, 301, 306 smooth muscle of, 300 sphincter between urethra and, 300 storage and elimination of urine from, 300f, 300–301 reflexes controlling, 303t, 306, 306f–307f organization of, 310–312, 311f trigone of, 299–300 umbrella cells of, 300 Bladder atony, 1150–1151 Bladder dysfunction, 2626–2627, 2663–2664 disorders associated with, 2626–2627 Parkinson’s disease, 312 spinal cord injury, 300f, 313–314 management of, 2627, 2663–2664 behavioral therapy, 2663 in intensive care unit, 2610 pharmacologic, 2663–2664 self-catheterization, 2663 neurogenic, 313–314 failure to eliminate urine, 313 failure to store urine, 313 in spinal cord injury, 313–314 overflow incontinence, 2627 Blink reflex, 918–919, 920t, 921t in Bell’s palsy, 921t, 1226–1227 in hemifacial spasm, 1240, 1241t in single-fiber electromyography, 989 in trigeminal neuralgia, 921t

␤-Blockers, for orthostatic hypotension, 2677 Blood flow, 667–677. See also Ischemia. cutaneous vascular responses to temperature changes, 333, 334f experimental nerve ischemia, 671–676 factors that modulate effects of, 672, 672f in diabetes, 512, 513f, 672–674, 673f, 673t, 1980 models of, 671–672 reperfusion injury and, 674f, 674–676 severity of ischemic fiber degeneration in, 672, 672f temperature effects on, 672 in compression neuropathies, 1396–1399 measuring in skin, 1122 nerve blood flow:blood pressure curve, 667–668 nerve microvasculature, 667–668 pathogenesis of nerve ischemia, 676–677 peripheral nerve adaptation to hypoxic stress, 668 poor autoregulation of peripheral nerve, 667–668 ␣-receptor subtypes and blood flow in active muscles, 331 regulation in nerve, 668–671 extrinsic mechanisms of, 669–671 maturation and aging, 671 perineurial pinch and endoneurial edema, 669–670 vasoconstrictors, 670, 670t vasodilators, 670t, 670–671 intrinsic mechanisms of, 668–669 blood viscosity, 669 capillary length, 669 capillary radius, 669 pressure gradient, 669 Poiseuille equation for prediction of, 668 splanchnic-mesenteric, 1124–1125, 1125f Blood pressure (BP). See also Hypertension; Hypotension. autoregulation of, 323, 667, 2654 baroreflexes in regulation of, 323–324, 324f, 2671, 2672f determination of baroreflex sensitivity, 1123–1124 muscle sympathetic nerve activity, 325–326, 325f–326f beat-to-beat recordings of, 1111–1113 response to head-up tilt, 1112–1113, 1115–1116, 1116t response to prolonged tilt, 1123 response to Valsalva maneuver, 1111–1112, 1112f diurnal variation in, 2663 effect of pressure gradient on blood flow, 669 aging and, 671 effects of glossopharyngeal nerve block on, 325, 326f, 325, 326f, 1274, 1275f effects of mental stress on, 331–332 effects of vagus nerve block on, 325, 326f exercise effects on, 328–331 heart rate, sympathetic outflow and, 323–324 managing fluctuations of, 2611 muscle sympathetic nerve activity and, 324–325, 325f nerve blood flow:blood pressure curve, 667–668 neurogenic orthostatic hypotension, 1150, 2653–2663. See also Orthostatic hypotension (OH), neurogenic. Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xi Blood pressure (BP) (Continued) patient log of, 2658 response to intravenous tyramine, 1124 response to orthostatic challenge, 326–327, 327f vasopressin in regulation of, 2671 Blood supply to nerve, 667–671, 800–801, 803f Blood vessels. See Vasculature. Blood viscosity, 669 Blood volume, orthostatic hypotension and, 2656, 2658 Blood-brain barrier, 575 poliovirus crossing, 1300 Blood-nerve barrier (BNB). See Blood-nerve interface (BNI). Blood-nerve interface (BNI), 39, 559, 575, 576, 610, 651–662 assessment of blood-nerve exchange of solutes at, 654 dynamics of blood-nerve exchange, 652–654 contribution of perineurium and microvessels, 652–653 epineurium, 654 perineurium as specialized connective tissue, 654 transcapillary permeation, 653 transperineurial permeation, 653 transporter molecules, 653–654 endoneurial homeostasis and, 657–660 during development, 659–660 in lead neuropathy, 657–658, 658f in nerve degeneration and regeneration, 658–659, 658f–659f functional complexity of, 652 in experimental autoimmune neuritis, 615–616, 616f oxidative stress and effects of ischemia on, 520, 676–677 permeability of, 652 morphologic studies of, 651–652 regulation of, 660–661 cellular control, 661 molecular basis of, 661 neurovascular unit, 660 protective function of, 651 T cell passage across, 559, 560f, 563 terminology for, 651, 661–662 BMPs (bone morphogenetic proteins), in neural crest development, 346 BNB (blood-nerve barrier). See Blood-nerve interface (BNI). Bogorad syndrome, 1275–1276 Bolton’s neuropathy. See Critical illness polyneuropathy. Bone lesions, 1151 in POEMS syndrome, 2453–2454, 2456t, 2458–2459, 2459f–2460f in sarcoidosis, 2420 lead lines, 2530, 2530f Bone marrow transplantation, 2498, 2498t. See also Hematopoietic stem cell transplantation (HSCT). for adrenoleukodystrophy, 1868 for Chédiak-Higashi syndrome, 1864 for globoid cell leukodystrophy, 2714 for metachromatic leukodystrophy, 1854 Bone morphogenetic proteins (BMPs), in neural crest development, 346 Bonnier syndrome, 1274 Borrelia burgdorferi, 2109, 2110. See also Lyme disease. Bors ice water test, 314

xii Botulinum toxin for hemifacial spasm, 1245 Horner’s syndrome and, 208 Botulism gastrointestinal atony in, 1281 infantile, 2713 vs. Guillain-Barré syndrome, 2202 Bowel dysfunction, 287–292, 2627–2628, 2665, 2694. See also Constipation; Diarrhea. after spinal cord injury, 2699 clinical features of, 288 denervation and, 287–288 diagnostic tests for, 288–292, 290t disorders associated with, 2627 in multiple sclerosis, 2700 management of, 292–293, 293t, 2628 Bowel habit modification, 292 BP. See Blood pressure (BP). BPAU (p-bromophenylacetylurea), effect on axonal transport, 400 Brace syndrome, 1359 Brachial amyotrophic diplegia, 1299 Brachial cutaneous nerve lateral, 15f, 22, 1470f medial, 15f, 20, 1470f posterior, 22, 1470f Brachial plexus anatomic regions of, 1341–1342, 1341f–1342f anatomy of, 18, 19–22, 1339–1341, 1340f blood supply to, 1341 branches of, 19–22 deep, 18 cords of, 1340, 1340f divisions of, 1340, 1340f location of, 19 muscles innervated by, 956t nerve conduction studies of, 914–915, 915t prefixation or postfixation of, 19 roots of, 1339–1340, 1340f terminal nerves of, 1340f, 1340–1341 trunks of, 1340, 1340f variations of, 19, 1341 Brachial plexus lesions, 1341–1371, 1521–1528 age distribution of, 1342 amputation for, 2631–2632 anatomic vulnerability to, 1342 assessment of, 1349–1353 clinical examination of, 1349–1350 clinical features of, 1346 clinical presentations of, 1346–1349 combined suprascapular/infraclavicular, 1349 degree of nerve fiber injury in, 1344–1345, 1363 axon loss lesions, 1344–1345 demyelinating conduction block, 1344 demographics of, 1342–1343 diagnosis of, 1522 electrodiagnostic examination of, 1351–1352, 1352f limitations of, 1352 needle electromyography, 1351–1352 nerve conduction studies, 1351 etiology of, 1521, 2630–2631 extent of involvement in, 1345 gender distribution of, 1343 histamine test for, 1522 imaging of, 1350–1351, 1522–1523

Index Brachial plexus lesions (Continued) in children, 2730–2735 infantile, 2733–2734, 2734f older children, 2734–2735 genetically determined, 2734–2735 idiopathic, 2734 neoplastic, 2735, 2736f traumatic, 2734 perinatal, 1362–1363, 1525–1528, 2632, 2730–2733. See also Brachial plexus lesions, obstetric. incidence of, 1343 infraclavicular, 1348–1349, 1365–1370, 2631 iatrogenic, 1367–1370 after orthopedic procedures, 1369–1370 medial brachial fascial compartment syndrome, 1368–1369, 1369f miscellaneous lesions, 1370 radiation-induced lesion, 1367–1368, 2490 nontraumatic, 1367 traumatic, 1365–1367 humeral neck fractures and dislocations, 1365–1366 miscellaneous lesions, 1367 thoracic outlet syndrome, 1366–1367 intradural, 1512 intraoperative nerve action potentials in, 1353 laterality of, 1343 level of, 1521–1522, 1521f–1522f longitudinal extent of damage in, 1345 mechanism of, 1343, 1521 natural history of, 1345–1346 obstetric, 1362–1363, 1525–1528, 2632, 2730–2733 clinical classification of, 1525, 1525t clinical features of, 2730–2731, 2732f course and prognosis for, 1526 degree of nerve fiber injury in, 1363 diagnostic investigation of, 1363, 1525, 1526–1527, 1527t, 2731, 2733f etiology of, 1362–1363, 1525 gender distribution of, 1343, 1362 Horner’s syndrome and, 2731 incidence of, 1362, 1525–1526, 2730 laterality of, 1343, 1362 operative repair of, 1363, 1527–1528, 2731–2733 prognosis for, 1363, 1526, 2731 rehabilitation for, 2632, 2733 risk factors for, 1362, 1526, 2731 pain of, 1514–1515, 2631 physiologic investigation of, 1522 preganglionic vs. postganglionic, 1342f radiation-induced, 1367–1368, 2490 clinical features of, 1367 electrodiagnostic evaluation of, 1368 incidence of, 1367 natural history of, 1368 pathophysiology of, 1367 risk factors for, 1367 types of disorders, 1367 vs. neoplastic infiltration, 1361, 1362t, 1368 rehabilitation for, 2630–2632 repair of, 1523–1525 for complete lesion, 1523–1524 for reconnection of spinal cord and avulsed nerves, 1524–1525 for rupture or avulsion of C5/C6 with intact (C7) C8/T1, 1523, 1524f patterns of injury and, 1523, 1523t results of, 1524, 1524t timing of, 1523 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Brachial plexus lesions (Continued) retroclavicular, 1348 somatosensory evoked potentials in, 1352–1353 spinal nerve root injury and, 1327 supraclavicular, 1346–1348, 1353–1365, 2631 iatrogenic, 1362 “classic” postoperative paralysis, 1363–1364 miscellaneous lesions, 1365 obstetric paralysis, 1362–1363, 1525–1528, 2632 post–disputed thoracic outlet syndrome surgery, 1365 post–median sternotomy lesion, 1364–1365 lower plexus lesions, 1347 middle plexus lesions, 1347 nontraumatic, 1359–1362 miscellaneous lesions, 1360–1362, 1362t neoplastic infiltration, 1360–1362, 1362t true neurologic thoracic outlet syndrome, 1359–1360, 1360f pan-supraclavicular lesions, 1347–1348 traumatic, 1353–1359 burner syndrome, 1358 closed traction injuries, 1353–1356 electricity injuries, 1359 miscellaneous lesions, 1359 open injuries, 1356–1357 pack palsy, 1357–1358 root avulsion injuries, 1354t, 1354–1356 upper plexus lesions, 1346–1347 thoracic outlet syndromes, 1370t, 1370–1371 traumatic, 2631–2632 infraclavicular, 1365–1367 supraclavicular, 1353–1359 vs. cervical radiculopathy, 1327 Brachial plexus neuropathy hereditary (HBPN), 1753–1766 age at onset of, 1754 chronic undulating form of, 1754, 1756 classic (relapsing-remitting) form of, 1754, 1756 clinical features of, 1754–1756, 1755f–1756f cranial nerve palsies and, 1755 differential diagnosis of, 1753, 1758–1760 hereditary neuropathy with liability to pressure palsies, 1670t, 1670–1671 immune brachial plexus neuropathy, 1754, 1757, 1758, 2304 dysmorphic features and, 1755–1756, 1756f, 1765 genetic counseling about, 1765–1766 historical overview of, 1753–1754 in children, 2734–2735 laboratory findings in, 1760–1761 molecular genetics of, 1753, 1754, 1764–1765, 1765 natural history of, 1756 pain of, 1753, 1754 pathogenesis of, 1763–1764 pathology of biopsied nerves in, 1761–1763, 1762f–1763f pattern of inheritance of, 1754, 1759f, 1764 pattern of weakness in, 1754–1755 pedigrees of families affected by, 1757f, 1759f postpartum attacks of, 1756, 1757f, 1757–1758

Index Brachial plexus neuropathy (Continued) precipitating or associated factors in, 1757–1758 sensory loss and, 1755 treatment of, 1763, 1766 immune (IBPN), 1140, 1754, 1757, 1758, 1993, 2299–2306 antecedent illnesses and conditions associated with, 2301, 2302t clinical features of, 2300 diagnosis of, 2301 differential diagnosis of, 2303t, 2303–2304 hereditary brachial plexus neuropathy, 1754, 1757, 1758, 2304 electrophysiologic findings in, 2302–2303 epidemiology of, 2300 historical overview of, 2299–2300 in children, 2734, 2735f long thoracic neuropathy and, 1464–1466 microvasculitis in, 2406 neurologic examination in, 2301 pathogenesis of, 2304 pathology of, 2011, 2304–2306, 2305f pattern of weakness in, 2300–2301 posterior interosseous neuropathy and, 1476 prognosis for, 2306 suprascapular neuropathy and, 1467–1469 symptoms of, 2300–2301 treatment of, 2306, 2632 lymphoma and, 2495 radiation therapy–induced, 2490 rehabilitation for, 2630–2632 Bradycardia due to cardiac vagal overactivity, 1281–1282, 1284 in acute inflammatory demyelinating polyneuropathy, 2200 in glossopharyngeal neuralgia, 1276, 1276f Bradykinin in cutaneous vascular responses to body heating, 333 receptors for, 182 Brain. See also Central nervous system (CNS). central projections of Golgi tendon organ, 158–159 in amyloidosis, 1924 in angiostrongyliasis, 2162–2164 in coenuriasis, 2169 in cysticercosis, 2168 in familial dysautonomia, 1832 in malaria, 2172 in schistosomiasis, 2160 pathways controlling lower urinary tract, 309–310 pontine micturition center, 311–312 spinobulbospinal micturition reflex pathway, 310–311, 311f suprapontine control of micturition, 312 pathways in sexual response cycle, 316, 317–318 tumors of affecting glossopharyngeal nerve, 1274 facial nerve paralysis due to, 1235, 1236 hypoglossal palsy due to, 1287 ocular palsies due to, 1199–1200, 1200f Brain-derived neurotrophic factor (BDNF), 174, 177–178, 377, 378, 380, 496 binding with tyrosine receptor kinase B, 177 clinical trials of, 380 in Guillain-Barré syndrome, 635 downregulation by neurotrophin-3, 177 expression after axotomy, 177–178, 694

Brain-derived neurotrophic factor (BDNF) (Continued) expression by dorsal root ganglion neurons, 177–178 inflammation and, 177 neurons responsive to, 380 role in myelination, 361, 380 structure of, 378 to increase axonal regrowth after chronic axotomy, 1411 upregulation by nerve growth factor, 176, 177 Brain-specific fatty acid–binding protein (B-FABP), 343, 344f, 344–345, 347 Brainstem auditory evoked responses (BAERs), in X-linked Charcot-Marie-Tooth disease, 1797 Brainstem tumors gastrointestinal manifestations of, 2695 glioma, 1199 Branchial arches, 11 British antilewisite (BAL) for arsenic poisoning, 2538 for gold intoxication, 2544 for lead poisoning, 2535 for mercury intoxication, 2540 Brn2 in Schwann cell development, 342 role in myelination, 359–360 p-Bromophenylacetylurea (BPAU), effect on axonal transport, 400 Brown-Séquard syndrome, partial, 1522 Brudzinski’s sign, in pediatric Guillain-Barré syndrome, 2716 Bruns-Garland syndrome. See Diabetic lumbosacral radiculoplexus neuropathy (DLRPN). Buccinator muscle, innervation of, 1222 Bulbar hereditary motor neuronopathy, 1604t, 1605t, 1617 Bulbar muscular atrophy, 705 Bulbospinal muscular atrophy scapuloperoneal, 1139, 1615, 1616 X-linked, 705, 706, 1299, 1316, 1599–1600, 1605t, 1612 clinical features of, 1612, 1612f differential diagnosis of, 1612 electrophysiologic findings in, 1316, 1612 genetics of, 1316, 1536, 1548, 1600, 1612 pathology of, 1612 pattern of involvement in, 1139 sensory involvement in, 1316 vs. amyotrophic lateral sclerosis, 1536 Bundle of His, 217 Büngner’s bands, in sarcoidosis, 2421 Bunina bodies, 1305, 1538 Bupivacaine, for radiculopathy due to disc herniation, 1326 Bupropion, for neuropathic pain, 2642 Burn neuropathy, 2030–2031 Burner syndrome, 1358 Burning feet in diabetic neuropathy, 1954 in subclinical hereditary sensory and autonomic neuropathy type I, 1825–1826 in tropical myeloneuropathies, 2063 Cuban epidemic neuropathy, 2066 Strachan’s syndrome, 2064 tropical ataxic neuropathy, 2069, 2070t tropical spastic paraparesis, 2070t skin biopsy in, 884 Butafox, 2515 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xiii Buttock innervation of, 28, 32 lumbosacral plexus injury due to injections into, 1380 Butylated hydroxytoluene, for experimental diabetic neuropathy, 516, 516t

C fibers, 1005–1006, 2638 modulation of discharge frequency by axon, 1007 effect of hyperpolarization, 1007, 1007f–1008f unidirectional block, 1007, 1009f nociceptors, 168t, 169, 870, 1005–1006, 2638 effects of hyperpolarization of, 1007, 1007f–1008f itch fibers, 1006 low-threshold, 1007 mechano-insensitive, 1006 polymodal, 1006 sensitization in erythromelalgia, 1009–1010, 1010f warmth detection by, 1005 Cachexia, diabetic neuropathic, 1954, 1995 CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarct and leukoencephalopathy), skin biopsy in, 887 Cadherins, 616 E-cadherin, in myelination, 418f, 419t, 420 N-cadherin in epilepsy, 466 in Schwann cell development, 343 Cadmium, 716 Caffeine, for orthostatic hypotension, 2676 Calbindin, 690 Calcaneal nerve, medial, 1502f, 1503, 1506 neuropathies of, 1506 Calcineurin inhibitors, 638–639 Calcitonin gene–related peptide (CGRP) aging effects on, 486 effect on nerve blood flow, 670, 670t expression after axotomy, 690, 692f expression by dorsal root ganglion neurons, 176t, 184–185 functions of, 184–185 immunostaining for, 173f in bladder, 301, 305 in chronic constriction model, 525 in diabetic neuropathy, 1978 in epidermal nerve fibers, 872 in ganglionic neurons, 236, 240, 242, 242f in sweat gland innervation, 1106 upregulation by nerve growth factor, 176 Calcium in axonal degeneration, 1407–1408, 1408f in axonal transport, 394 in neuron death, 701 intracellular accumulation in ischemia, 677 Calcium channel blockers for nocturnal hypertension in patients with neurogenic orthostatic hypotension, 2663 to prevent autonomic dysreflexia, 2664 Calcium channels, 191 antibodies to, in Lambert-Eaton myasthenic syndrome, 590, 597–598, 834 axotomy-induced alterations of, 191, 694–695 expression by dorsal root ganglion neurons, 191 in cardiac neural control, 223, 223f–224f in mechanotransduction, 188 in neuromuscular transmission, 832

xiv Caloric test, 1256 cAMP (cyclic adenosine monophosphate), in netrin signaling, 461 Campylobacter infection acute inflammatory demyelinating polyradiculoneuropathy and, 592 Bickerstaff’s brainstem encephalitis and, 2205 epidemiology of, 2206 Guillain-Barré syndrome and, 590–591, 2198, 2199, 2285 acute inflammatory demyelinating polyneuropathy, 2211–2212 axonal forms, 2206–2207 Fisher syndrome, 2206, 2210 molecular mimicry and, 592 multifocal motor neuropathy and conduction block and, 2290 CAMs. See Cellular adhesion molecules (CAMs). Canine neuropathies, inherited, 712 Cannabinoid receptors, expression by dorsal root ganglion neurons, 183 CANOMAD (chronic ataxic neuropathy with ophthalmoplegia, M protein, agglutination, and diasialosyl antibodies) syndrome, 2180t, 2187 Capillaries endoneurial, permeability of, 652–653 epineurial, 42 in diabetic neuropathy, 1966, 1968f large-diameter, 668 occlusion of, 804–805 oxygen delivery to nerve, 668–669, 669f capillary length and, 669 capillary radius and, 669 Capsaicin chromatolysis induced by, 686 excitotoxic mechanism of neurotoxicity of, 699 for diabetic neuropathy, 1972 hyperalgesia induced by, 870 intracytoplasmic vacuolation induced by, 688 receptor for, 872 research model for chemical denervation, 889–890, 889f–890f topical, for postherpetic neuralgia, 2644 Carbamates, 2508–2510 clinical features of exposure to, 2509–2510 metabolism of, 2509 toxicity of, 2509 use and sources of exposure to, 2508–2509 Carbamazepine for diabetic neuropathy, 1971t for Fabry’s disease, 1900 for glossopharyngeal neuralgia, 1277 for hemifacial spasm, 1245 for neuromyotonia, 1144 for neuropathic pain, 2643 for superior laryngeal neuralgia, 1284 for trigeminal neuralgia, 1215, 2641, 2642, 2643 for vestibular paroxysmia, 1264 side effects of, 1215 Carbohydrate-deficiency ataxia, 1738 in children, 2724 Carbon disulfide, 2510–2511 metabolism of, 2510 neurotoxicity of, 2510 clinical features of, 2510 electrophysiology of, 2510–2511 morphology of, 2510 pathogenesis of, 2511 use and sources of exposure to, 2510

Index Carbonic anhydrase, 176t, 178 Carboplatin neurotoxicity, 2544–2545 Cardiac arrhythmias, 226–227, 2611 in glossopharyngeal neuralgia, 1276, 1276f in Guillain-Barré syndrome, 2626 in trichinosis, 2164 vagal efferent activity and, 1281 “Cardiac brain,” 220 Cardiac function, neural control of, 217–227 assessing cardiac neural drive, 225–226 neck chamber technique for, 226 vasoactive drugs for, 225–226 cardiac vagal and sympathetic nervous system, 217–220, 218f, 2671–2672, 2672f cardiac parasympathetic tone, 218–220, 220f–221f cardiac plexus and “cardiac brain” hypothesis, 220 cardiac sympathetic tone, 217–218, 219f chemical transmission at cardiac neuroeffector junction, 222–224, 223f historical studies of, 222–223 postjunctional signaling, 223–224, 224f clinical outcome of impairment of, 226–227 communication between sympathetic and parasympathetic systems, 224–225 accentuated antagonism, 225 neurochemical transmission in central nervous system, 220–222 Cardiac nerves, 217 Cardiac neuroeffector junction, 222–224 Cardiac plexus, 25, 220 Cardiac receptors, vagal, 1284 Cardiac vagal efferent activity, 1280–1281 Cardiofacial syndrome, 1237 Cardiotrophin-like cytokine (CLC), 380 Cardiovagal function, heart rate recordings for evaluation of, 1109–1111 comparison of tests, 1110–1111 diving response, 1123 patient preparation for, 1109 response to coughing, 1123 response to deep breathing, 1109, 1109t, 1110f response to prolonged tilt, 1123 response to standing, 1110 Valsalva ratio, 1109–1110, 1110t Cardiovascular autonomic failure, 217, 226–227, 2653–2654, 2671–2678. See also Orthostatic hypotension (OH), neurogenic. in diabetic patients, 1957, 1958 heart rate abnormalities, 1957, 1958 impaired cardiac function, 1958 management of, 1960 orthostatic hypotension, 1150, 1957, 1958, 1958t painless myocardial infarction, 1958 overview of, 2671–2673, 2672f, 2673f treatment of, 2626, 2630, 2656–2657, 2673–2681 ␤-blockers, 2677 caffeine, 2676 clonidine, 2677 compression garments, 2658–2659, 2674 cyclooxygenase inhibitors, 2677 dietary, 2658 dihydroergotamine, 2660t, 2661, 2678 dihydrophenylserine, 2677–2678, 2678f dopamine antagonists, 2661, 2678 erythropoietin, 2676–2677 for early morning symptoms, 2662 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Cardiovascular autonomic failure (Continued) for nocturnal hypertension, 2663 for periods of orthostatic decompensation, 2662 for postprandial symptoms, 2662–2663, 2674 goals of, 2656–2657, 2657t in hospital, 2662 in patients with supine hypertension, 2674 nonpharmacologic, 2657–2659, 2673t, 2673–2674 patient education in, 2657t, 2657–2658, 2673 pharmacologic, 2659–2662, 2660t, 2660f–2661f, 2674–2681 physical countermaneuvers, 2659, 2659t, 2673 postural adjustment, 2658 pyridostigmine, 2678 resistance training, 2659 somatostatin, 2677 sympathomimetic agents, 2675–2676, 2676t tyramine, 2678 volume expansion, 2658, 2659, 2674–2675, 2675t water drinking, 2658, 2659, 2662, 2674 yohimbine, 2677 Cardiovascular system involvement, 1150 in amyloidosis, 2433 in Fabry’s disease, 1894 in POEMS syndrome, 2460 in Tangier disease, 1908 in trichinosis, 2164 Carotid nerve, internal, 23 Carotid plexus, 1224f internal, 23 Carotid sinus baroreceptors in, 323 hypersensitivity of, 1277, 1282 Carotid sinus nerve, 1274 Carotid-cavernous sinus fistula, 1201 Carpal tunnel, 1437, 1437f anatomy of, 1437, 1437f anomalous structures of, 1447 Carpal tunnel syndrome (CTS), 1435–1452 anatomy of, 1436f–1437f, 1436–1438 carpal tunnel pressures in, 1443 clinical criteria for diagnosis of, 1441–1442 conditions associated with, 1444t, 1444–1447 acromegaly, 1445, 2045 amyloidosis, 1165, 1447, 1923, 1924, 2431f, 2432–2433 anomalous carpal tunnel structures, 1447 computer use, 1446–1447 diabetes mellitus, 1444–1445, 1964 dracunculiasis, 2170 hereditary neuropathy with liability to pressure palsies, 1165, 1448, 1668 Hurler’s syndrome, 1865 hyperthyroidism, 1445 hypothyroidism, 1445, 2039–2040 infections, 1447 menopause and hormone replacement therapy, 1445 miscellaneous conditions, 1447 occupational factors, 1446, 1446t pregnancy, 1445 renal failure, 2025 rheumatoid arthritis and other inflammatory conditions, 1445

Index Carpal tunnel syndrome (CTS) (Continued) Scheie’s syndrome, 1865 tendonitis and degenerative arthritis, 1445–1446 trauma, 1444 tumors/mass lesions, 1447 congenital, 2725, 2725f differential diagnosis of, 1448t, 1448–1449 endoneurial fluid flow in, 657 etiology of, 1442–1443 familial, 1447–1448 historical recognition of, 1435–1436 incidence of, 1438 laboratory diagnosis of, 1449–1451 electromyography, 1449 MRI, 1450–1451, 1450f–1451f ultrasound, 1449–1450 medical treatment of, 1451 glucocorticoids, 637, 1451 nerve conduction studies in, 912f, 1436 collision technique for, 926–928, 928f postoperative, 1452 with Martin-Gruber anastomosis, 929, 929f nocturnal paresthesias in, 1141, 1142 pathophysiology of, 1443–1444 pediatric, 1448, 1448t, 2724f, 2725f, 2725–2726 polyneuropathy and, 1165 provocative tests for, 1441 risk factors for, 1442–1443, 1443t sensory and motor examination in, 1440–1441 severity scales for by EMG criteria, 1442 clinical, 1442 staging system for, 1442 surgical treatment of, 1452 median nerve conduction studies after, 1452 preoperative severity and outcome of, 1452 recurrent paresthesias after, 1452 symptoms of, 1435, 1438t, 1438–1440 correlation with physical examination and nerve conduction studies, 1441 factors causing aggravation or relief of, 1440 proximal radiation of, 1439, 1440t sensory symptoms and pain, 1438–1439, 1439f Carvedilol, for experimental diabetic neuropathy, 516, 516t Cas (Crk-associated tyrosine kinase substrate), 455 CASE-IV. See Computer-Assisted Sensory Examination System IV (CASE-IV). Caspases in apoptosis, 700f, 701, 1590, 1590f in ischemic injury, 524 Caspr (contactin-associated protein), 549 mice deficient in, 550, 551f Caspr2 (contactin-associated protein 2), 549–550 CASS. See Composite Autonomic Severity Score (CASS). Cassava neurotoxicity, 2069t, 2073–2075 clinical features of, 2074 epidemiology of, 2073 konzo, 2074–2075 pathogenesis of, 2074 tropical ataxic neuropathy and, 2070, 2073 tropical spastic paraparesis and, 2074 urinary test for, 2075

Castleman’s disease, 2453, 2454, 2455, 2459–2460 Catalase, 511 in experimental diabetes, 512, 673 Catecholamines biosynthesis of, 1117f–1118f in autonomic ganglia, 241 in familial dysautonomia, 1833 measuring plasma levels of, 1116–1118, 1117f–1120f measuring urinary levels of, 1118 Cauda equina, 1323 Cauda equina syndrome after spinal anesthesia, 1333 bladder dysfunction and, 2626–2627 causes of, 1334 due to lumbosacral disc herniation, 1328–1329, 1329f in ankylosing spondylitis, 1334 in chronic inflammatory demyelinating polyradiculoneuropathy, 2225 in sarcoidosis, 1334 vascular disease and, 1334 Causalgia, 1148, 1513–1514 after nerve injury with false aneurysm and arteriovenous fistula, 1518 due to penetrating missile injury, 1513–1514, 1514f Cavernous sinus cranial nerves in, 1193–1195 tumors of, 1200 Cayman Island ataxia, 1738 CCFDN (congenital cataracts, facial dysmorphism, and neuropathy), 1775, 1776f Marinesco-Sjögren syndrome with myoglobinuria and, 1776 CCK (cholecystokinin), 255 downregulation by nerve growth factor, 177 expression after axotomy, 691 in ganglionic neurons, 242 test for sphincter of Oddi innervation, 255 CD4, 37, 2342 CD4⫹ cells, 561–563 CD8, 37, 2342 CD8⫹ cells, 561–563 CD9, 346 CD22, 575 CD33, 575 CD47, 575 CD200 (Ox-2), 575 Cdc42, in actin dynamics, 450, 451f, 452 CDT. See Cooling detection threshold (CDT). Celiac ganglion, 234 Celiac plexus, 30, 31 Cellular adhesion molecules (CAMs), 97–98 categories of, 616 in experimental autoimmune neuritis, 615–617, 616f, 623, 623t, 625 in nodes of Ranvier, 56 in sodium channel segregation, 97–98 mice deficient in L1, 548, 549f Cellular neurofibroma, 2593 Cellular schwannoma, 2588, 2590f Central canal stenosis, 1328 Central nervous system (CNS), 1295. See also Brain; Spinal cord. enteric nervous system properties mimicking those of, 253 evaluation of central autonomic pathways, 1125–1127, 1126f–1127f in angiostrongyliasis, 2162–2164 in cardiac neural control, 220–222 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xv Central nervous system (CNS) (Continued) in chronic inflammatory demyelinating polyradiculoneuropathy, 2233 in coenuriasis, 2169 in cysticercosis, 2168 in Fabry’s disease, 1894 in hereditary sensory and autonomic neuropathies, 1809 in HIV infection, 2129, 2130 in Lyme disease, 2112 in of metachromatic leukodystrophy, 1848, 1849 in sarcoidosis, 2419 in schistosomiasis, 2160 in transthyretin-related amyloid neuropathy, 1924 in X-linked Charcot-Marie-Tooth disease, 1797 pathways controlling lower urinary tract, 306–310 pathways in sexual response cycle, 316, 317–318 pediatric polyneuropathies with involvement of, 2710 reflex pathways in sexual response cycle, 317–318 regulation of anorectal nerve supply by, 282 role in neuropathic pain, 2638–2639 Central nervous system–peripheral nervous system transitional zone, 67–77, 68f blood supply to, 77 central tissue projections of, 68 comparative anatomy of, 77, 78f development of, 74f, 75–77 fiber transition in, 68–71, 71f–72f glial fringe of, 68 glial islands and, 71 historical studies of, 75 mechanical features of, 77 myelinated axons at, 68–70, 73f nerves lacking clearly defined barrier, 71 overlap of tissues at, 68, 69f peripheral tissue insertions of, 68 tissue composition of, 71, 75 unmyelinated axons at, 70–71 variations in morphology of, 68, 69f–70f Central pontine myelinolysis, 2029–2030 Central tissue projection (CTP), 68 Cephalic (cranial) parasympathetic ganglion, 17 Ceramide deficiency, 1859 Ceramide trihexoside accumulation, in Fabry’s disease, 1893, 1898–1900 Cerebellar artery, anterior inferior, 1220 Cerebellar degeneration, lymphoma and, 2495 Cerebellopontine angle tumors, 1199 differential diagnosis of, 1265 Cerebellopontine cistern, 1219 Cerebral autosomal dominant arteriopathy with subcortical infarct and leukoencephalopathy (CADASIL), skin biopsy in, 887 Cerebral schistosomiasis, 2160 Cerebro-oculofacioskeletal syndrome, 1259 Cerebroside sulfatase deficiency, 1847 Cerebrospinal fluid (CSF) analysis, 1174–1175 in acute inflammatory demyelinating polyradiculoneuropathy, 2315t in arsenic poisoning, 2537 in brachial plexus neuropathy hereditary, 1760 immune, 2303

xvi Cerebrospinal fluid (CSF) analysis (Continued) in chronic inflammatory demyelinating polyradiculoneuropathy, 2222, 2226, 2233–2234, 2315t, 2718 in cytomegalovirus infection, 2122, 2123 in Déjérine-Sottas neuropathy, 1625 in diabetic lumbosacral radiculoplexus neuropathy, 1963 in diabetic motor neuropathy, 1955 in diphtheria, 2148 in gold neuropathy, 2544 in Guillain-Barré syndrome childhood, 2716 infantile, 2712 in Hopkins syndrome, 2031 in hypothyroid polyneuropathy, 2041 in lead poisoning, 2531 in leptomeningeal metastasis, 1330 in Lewis-Sumner syndrome, 2232 in lumbosacral plexus lesions, 1378 in Lyme disease, 2111 in monoclonal gammopathy of undetermined significance, 2264 in multifocal motor neuropathy and conduction block, 2286 in neurolymphomatosis, 2492 in nonmalignant inflammatory sensory polyganglionopathy, 2312 in paraneoplastic neuropathy demyelinating neuronopathy, 2478 subacute sensory neuronopathy, 2475 in POEMS syndrome, 2453, 2457 in poliomyelitis, 1300 infantile, 2713 in primary biliary cirrhosis–associated sensory polyneuropathy, 2020 in radiculopathy, 1325 infectious polyradiculopathy, 1331 in sarcoidosis, 2419 in thallium intoxication, 2541 in transverse myelitis, 2717 in uremic neuropathy, 2024 in West Nile virus infection, 1301 Cerebrotendinous xanthomatosis, 2722 Cervical arteries, 1341 Cervical ganglion, 17f, 234 inferior, 23 middle, 23 superior, 23, 493 Cervical nerve(s), 23, 31–32 transverse, 18 Cervical plexus, 18–19, 31 anatomy of, 18–19 muscles innervated by, 956t Cervical radiculopathies, pediatric, 2737 Cervical radiculoplexus neuropathy diabetic, 1993 nondiabetic. See Brachial plexus neuropathy, immune (IBPN). Cervical rib and band syndrome, 1359–1360 cause of, 1359, 1360f clinical features of, 1359 diagnosis of, 1359–1360 treatment of, 1360 Cervical spine, 1324f Cervical spondylotic myelopathy, 1310–1313, 1328 amyotrophic type of myelopathy hand in, 1311, 1311f animal models of, 1313 chronic, with age-related degenerative changes, 1311 clinical features of, 1310

Index Cervical spondylotic myelopathy (Continued) definition of, 1310 differential diagnosis of, 1311 neuroimaging in, 1311–1313, 1312f ossification of posterior longitudinal ligament and, 1310 pathology of, 1313 Cervicofacial nerve, 1220 Cervicothoracic (stellate) ganglion, 17f, 23, 25 cGMP (cyclic guanosine monophosphate) in penile erection, 315 in semaphorin signaling, 465 CGRP. See Calcitonin gene–related peptide (CGRP). CGT (UDP-galactose:ceramide galactosyltransferase), 423 mice deficient in, 550, 551f Chagas’ disease, 260, 1151, 1281, 2154t, 2156–2160 cardiac lesions in, 2159 clinical phases of, 2156–2157, 2158f congenital, 2156 epidemiology of, 2156 etiologic agent in, 2156 life cycle of, 2156, 2157f gastrointestinal megasyndromes in, 2159–2160 pathogenesis of, 2157 pathophysiology of, 2157–2159 peripheral neuropathy in, 2160 reactivation of, 2157 transmission of, 2156 Chain fibers, 137, 138 Chain (paravertebral) ganglia, 234 Channelopathies, 597–598 Charcot joints, 1146 in diabetic patients, 1972 in familial dysautonomia, 1832 with subclinical hereditary sensory and autonomic neuropathy type I, 1823–1825, 1825f Charcot-Marie-Tooth (CMT) disease, 1623–1653. See also Hereditary motor and sensory neuropathy (HMSN). autosomal dominant, 1626, 1626f autosomal recessive, 1626, 1627t, 1769–1785, 2714 classification of, 1625t, 1625–1626, 1627t clinical features of, 1624, 1631–1636 epidemiology of, 1185–1186, 1186t, 1630–1631, 2707–2708 gene identification studies in, 1627t, 1628–1630, 1632f hearing loss in, 1168, 1259 historical background of, 1624, 1681 in children, 2707–2708 macro electromyography in, 993f, 998–999 nerve conduction studies in, 1636–1639, 1637f–1638f, 1639t, 1681, 1682 nomenclature for, 1623 pathology of, 703–704, 1639–1641, 1640f–1644f, 1644t quality-of-life issues in, 1060 spinal form of, 1604t transgenic models of, 1561–1579 treatment of, 1652–1653 trigeminal neuralgia and, 1213 X-linked, 1626, 1791–1800 ChAT. See Choline acetyltransferase (ChAT). Chédiak-Higashi syndrome, 1863–1864 clinical features of, 1863 genetics of, 1863–1864 in children, 2722 pathology of, 710 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Chelation therapy for arsenic poisoning, 2538 for gold intoxication, 2544 for lead poisoning, 2534–2535 for mercury intoxication, 2540 for thallium intoxication, 2543 Chemoaffinity theory of axonal guidance, 447 Chemokines and chemokine receptors, 2342 in experimental autoimmune neuritis, 617 in HIV-associated distal symmetrical polyneuropathy, 2133, 2133f in wallerian degeneration, 1415–1416 Chemoreceptors activation of, 327 muscle sympathetic nerve activity responses to, 327–328, 328f central chemoreflexes and autonomic function, 327–328 expression by dorsal root ganglion neurons, 166, 168t, 188 in enteric nervous system, 254, 255–256 Chemotherapy. See also specific chemotherapeutic agents. for amyloidosis, 2437–2439 for leptomeningeal metastasis, 1330 for monoclonal gammopathy and peripheral neuropathy, 2268–2269, 2269t for POEMS syndrome, 2463 peripheral neuropathy induced by, 2490t, 2490–2491, 2498t Cherry-red spot–myoclonus syndrome, 1865 Chest radiography, 1174 Chickenpox, 2118 Child-Pugh severity classification of hepatic disease, 2018 Children. See Pediatric patients. Chlorambucil, 637–638 adverse effects of, 638 for monoclonal gammopathy and peripheral neuropathy, 2268–2269 mechanism of action of, 637–638 pharmacokinetics of, 638 use in neuropathy, 638 3-Chlorophene neuropathy, 2508 Chloroquine neuropathy, 2555–2556 Chlorphos, 2515 CHN. See Congenital hypomyelinating neuropathy (CHN). Cholecystokinin (CCK), 255 downregulation by nerve growth factor, 177 expression after axotomy, 691 in ganglionic neurons, 242 test for sphincter of Oddi innervation, 255 Cholesteatoma, congenital, 1236 Cholesterol, 1905–1906 in familial high-density lipoprotein deficiency, 1912 in myelin membrane, 423, 577 in Tangier disease, 1906–1909, 1912 regulation of cellular homeostasis of, 1911 reverse transport pathway for, 1906 synthesis of, 1911 Cholesterol ester storage disease, 1864 Cholestyramine, for diarrhea, 2683 Choline, 832 Choline acetyltransferase (ChAT), 832 activity after axotomy, 693 deficiency of, 839–841 clinical features of, 839–840, 840f end-plate studies in, 840 molecular pathogenesis of, 840–841, 841f therapy for, 860 Cholinergic dysautonomia, 1281

Index Cholinergic receptor. See Acetylcholine receptor (AChR). Chondroitin sulfate antibodies to, 597 IgM monoclonal protein reactive with, 2268 in wallerian degeneration, 1418–1419, 1419f Chorda tympani, 1220, 1222, 1223f–1224f Choristoma, neuromuscular, 2602 Chromaffin cells, preganglionic innervation of, 237 Chromaffin system, 17 Chromatolysis, 684–687 axonal transport and, 390, 686–687 axotomy-induced, 684–686, 685f diseases associated with, 686 factors affecting extent and duration of, 686 age, 686 location of injury, 686 regeneration, 686 type of peripheral neuron, 686 histologic and ultrastructural features of, 684, 685f toxin-induced, 686, 715 Chronaxie, 114 Chronic ataxic neuropathy with ophthalmoplegia, M protein, agglutination, and diasialosyl antibodies (CANOMAD) syndrome, 2180t, 2187 Chronic immune sensory polyradiculopathy, 808, 809f–810f Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP), 2179t, 2221–2246 after bone marrow transplantation, 2498, 2498t as separate nosologic entity, 2222–2223 autoantibodies in, 2179t, 2239, 2245–2246 P0, 595 P2, 595–596 PMP22, 595 blink reflex in, 920t cerebrospinal fluid analysis in, 2222, 2226, 2315t, 2718 chronic lymphocytic leukemia and, 2496 clinical features of, 2223–2225, 2225t age at onset, 2224 atypical presentations, 2229 cranial nerve involvement, 1139, 2224, 2225 early symptoms and signs, 2224 motor involvement, 1141, 2224, 2231 neurologic findings, 2225 preceding infections or immunizations, 2223–2224 sensory variant, 2229, 2231–2232, 2315t clinical spectrum of, 2231–2233 CIDP and demyelinating hereditary motor and sensory polyneuropathy, 2233 CIDP with central nervous system involvement, 2233 multifocal motor and sensory demyelinating neuropathy, 2232–2233, 2278, 2279, 2336 purely motor presentation, 2231 purely sensory presentation, 2229, 2231–2232, 2315t conduction block in, 125–126, 2222, 2227t, 2228 course of, 1141, 1167, 2222–2223, 2225–2226, 2229 definition of, 2221

Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) (Continued) diagnostic criteria for, 2225–2228, 2226t electrodiagnostic criteria, 2227t, 2227–2228 proposed criteria for clinical trials, 2226–2228 differential diagnosis of, 2221, 2228t, 2228–2229 acute inflammatory demyelinating polyneuropathy, 2222–2223, 2228 diabetic radiculoplexus neuropathy, 2000, 2228–2229 in children, 2718–2719 monoclonal gammopathy of undetermined significance, 2228, 2255, 2264–2265 multifocal motor neuropathy and conduction block, 2229 POEMS syndrome, 2457–2458, 2462–2463 subacute idiopathic demyelinating polyradiculoneuropathy, 2222–2223 electrophysiology of, 2227t, 2227–2228, 2234, 2457–2458 epidemiology of, 2223 experimental autoimmune neuritis as model for, 621, 2239 historical descriptions of, 2221–2222 human leukocyte antigens in, 2223 idiopathic, 2225 in children, 2224, 2230–2231, 2718–2719 cerebrospinal fluid analysis in, 2718 clinical features of, 2230–2231, 2718 course of, 2230 diagnosis of, 2231, 2707–2710 differential diagnosis of, 2718–2719 electromyography in, 2718 epidemiology of, 2230, 2718 immunotherapy for, 2231, 2719 infantile, 2713 magnetic resonance imaging in, 2718 prognosis for, 2231, 2719 in diabetic patients, 1955–1956, 2229 in HIV infection, 2136–2137, 2224, 2229 clinical features of, 2136–2137 disease-stage specificity of, 2129 etiology of, 2137 treatment of, 2137 laboratory findings in, 2226, 2233–2234 cerebrospinal fluid analysis, 2222, 2226, 2233–2234, 2315t, 2718 nerve biopsy, 2234, 2242f–2244f, 2243 lymphoma and, 2494 melanoma and, 2228, 2245 multifocal, 808, 2229 overview of, 2221 paraneoplastic, 2478 pathogenesis of, 2243–2246 axonal loss, 2246 cellular immune responses, 2243–2245 humoral immune responses, 2245–2246 pathology of, 806–808, 807f–808f, 2226, 2239–2243, 2240f–2244f chromatolysis, 686 demyelination, 2239–2240 diagnostic pathologic features, 2241–2242 edema, 2240, 2241, 2243f endoneurial edema, 2240, 2241, 2243f interstitial alterations, 2240–2241 onion bulb formation, 2239, 2240f, 2242, 2242f, 2243f polyglucosan bodies, 783f Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xvii Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) (Continued) prevalence of, 2223 prognosis for, 2229–2230 proximal symmetrical and upper limb–predominant, 1138 skin biopsy in, 887 treatment of, 635, 2179t, 2230, 2234–2239 approaches to management, 2239 azathioprine, 640, 2235–2236 corticosteroids, 637, 2222, 2223, 2234–2235, 2236–2237 cyclophosphamide, 2236 cyclosporine, 639, 2179t, 2236 history of, 2234–2236 interferons, 639 intravenous immunoglobulin, 643–644, 2230, 2235, 2237–2239 mycophenolate mofetil, 641, 2236 plasma exchange, 642, 2179t, 2235, 2237 review of clinical trials and long-term therapy, 2236–2239 stem cell transplantation, 2236 tacrolimus, 639 upper limb–predominant, 1164 Chronic lymphocytic leukemia (CLL), 2495–2496 Chronic myeloid leukemia (CML), 2496 Chronic obstructive pulmonary disease (COPD), neuropathy and, 2021–2022 Chronic progressive external ophthalmoplegia (CPEO), 1941 Chronic renal failure, nerve excitability studies in, 124–125 Chronic sensory and autonomic neuropathy of unknown cause, gastrointestinal symptoms of, 2698–2699 Churg-Strauss syndrome (CSS), 2337, 2353t, 2355–2356 antineutrophil cytoplasmic antibodies in, 2352, 2355 classification and definition of, 2338t, 2340t diagnostic criteria for, 2338, 2356 incidence of, 2340t pediatric polyneuropathy and, 2719 phases of, 2356 relative prevalence of, 2341t treatment of, 2336, 2375t Chylomicron, 1906, 1923t CIDP. See Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). Ciliary ganglion, 17, 23, 203–204, 204f, 234 Ciliary nerves, 203–204, 205f Ciliary neurotrophic factor (CNTF), 180, 377, 378, 380 expression after axotomy, 180, 694 expression by dorsal root ganglion neurons, 180 null allele in amyotrophic lateral sclerosis, 1535 receptors for, 380 structure of, 378 upregulation by interleukin-6, 180 Circulating immune complex–mediated vasculitis, 2344–2345, 2345f Cirrhosis hepatic Child-Pugh severity classification of, 2018 somatic and autonomic polyneuropathy in, 2017–2018, 2018t primary biliary, sensory neuropathy in, 2020–2021

xviii Cisapride for gastroparesis, 1960, 2665 for intestinal pseudo-obstruction, 2665 Cisplatin neurotoxicity, 2490, 2490t, 2498t, 2544–2545 brachial plexus lesion from, 1370 nerve growth factor protection against, 382–383 Citalopram, for pain, 2642 c-Jun expression after axotomy, 692 CK (creatine kinase) in hereditary motor neuronopathy, 1606 in multifocal motor neuropathy and conduction block, 2286 CLA (cutaneous leukocytoclastic angiitis), 2361–2362 classification and definition of, 2338t, 2339 incidence of, 2340t Clarithromycin, for leprosy, 2103 Claude’s syndrome, 1201 Claudication, in amyloidosis, 2433 Clavicular fracture, brachial plexus lesion due to, 1366–1367 Claw hand, 1151 CLC (cardiotrophin-like cytokine), 380 CLF-1 (cytokine-like factor-1), 380 Clinical course of peripheral neuropathy, 1141–1142, 1167 Clinical Neuropathy Assessment (CNA), 1040–1041, 1042f–1045f Clinical patterns of peripheral neuropathy, 1137–1155 clinical course, 1141–1142 intermittent symptoms, 1141–1142 diagnosis, 1154–1155 confirmation of neuropathy, 1154 establishing etiology, 1154–1155 differential diagnosis, 1152–1154 disorders simulating peripheral neuropathy, 1152–1153 neuropathies simulating other conditions, 1153–1154 distribution and patterns of involvement, 1137–1140 complex distributions, 1139 cranial nerve involvement, 1139–1140 distal symmetrical polyneuropathy, 1138, 1138f–1140f focal and multifocal neuropathies, 1140, 1140t proximal symmetrical and upper limb–predominant polyneuropathy, 1138–1139 functional selectivity, 1141 autonomic neuropathy, 1141 motor neuropathy, 1141 sensory neuropathy, 1141 symptoms and signs, 1142–1152 autonomic involvement, 1150–1151 negative motor symptoms and signs, 1142 nerve thickening, 1151–1152 positive motor symptoms and signs, 1142–1146, 1143t sensory symptoms, 1146 negative sensory symptoms and signs, 1146–1148, 1147f positive sensory symptoms and signs, 1148–1150, 1149f skeletal deformity, 1151 trophic changes, 1151 Clioquinol neurotoxicity, 1153, 1296

Index Clitoris deep dorsal vein of, 280f dorsal nerve of, 28 Clivus tumors, 1199 CLL (chronic lymphocytic leukemia), 2495–2496 Cloaca, 279–280 Clofazimine, for leprosy, 2103–2104 Clonazepam for hemifacial spasm, 1245 for restless legs syndrome, 1145 Clonidine for diarrhea, 2665, 2694 for gastroparesis, 2665 for orthostatic hypotension, 2661, 2677 for pain, 527 topical, 2645 Clonidine–growth hormone stimulation test, 1126–1127, 1127f Clostridium botulinum, 2202. See also Botulism. Clunial nerve inferior, 28 middle, 32 superior, 32 CMAPs. See Compound muscle action potentials (CMAPs). CMSs. See Congenital myasthenic syndromes (CMSs). CMT disease. See Charcot-Marie-Tooth (CMT) disease. CMV infection. See Cytomegalovirus (CMV) infection. CNA (Clinical Neuropathy Assessment), 1040–1041, 1042f–1045f CNL (chronic myeloid leukemia), 2496 CNS. See Central nervous system (CNS). CNTF. See Ciliary Neurotrophic factor (CNTF). Cobalamin deficiency. See Vitamin B12 deficiency. Cocaine test for Horner’s syndrome, 207–208 Coccygeal nerve, 30, 32 Coccygeal plexus, 30 Coccygeus muscle, 280, 280f nerve to, 28 Cochlea, 1253, 1255f Cochlear implant for auditory neuropathy, 1261 for eighth nerve aplasia/hypoplasia, 1259 Cochlear microphonic, 1256 in auditory neuropathy, 1260 Cochlear nerve, 1253. See also Vestibulocochlear nerve. Cockayne’s syndrome, 1741 in children, 2723 neurofibrillary tangles in, 705 Codominant inheritance, 1547 Coenuriasis, 2168–2169 clinical features of, 2169 epidemiology of, 2169 etiologic agent in, 2168–2169 Cofilin, 452 Cognitive deficits, in amyotrophic lateral sclerosis, 1537 Colchicine for amyloidosis, 2438 microtubule disaggregation induced by, 389, 393 neuropathy induced by, 2556–2557 Cold detection fibers, 1005 Cold therapy, 2630 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Cold transduction, 187 Collagen, 35–37, 377 endoneurial, 37 epineurial, 37, 42 fibrous long-spacing, 37–38, 38f nerve fiber sheaths of, 37 of Renaut bodies, 38 perineurial, 37, 38f, 39 Collagen pockets, 66 Collapsin response mediator protein (CRMP), 585–586 Collapsins, 585 Colon afferent fibers to, 283 assessment of colonic transit, 2691–2692 colonic transit in patient with obstructed defecation, 287, 287f ganglionic transmission and peristaltic activity of, 244, 245f parasympathetic nerves to, 282 Colonoscopy, 2691 Colostomy, 2700 Compartment syndrome, 1517–1518 decompression for, 1518 gluteal, 1494–1495 medial brachial fascial, 1368–1369, 1369f Complement, 564, 575 generation by Schwann cells, 567 in circulating immune complex–mediated vasculitis, 2344–2345, 2345f in demyelination in IgM monoclonal gammopathy with neuropathy, 2267 in neuron death, 701 macrophage expression of, 559 Complex regional pain syndrome (CRPS) type 1 adrenergic sweating in, 1106 detecting sympathetic dysfunction in, 1105 Complex repetitive discharges, 941, 943, 946f Composite Autonomic Severity Score (CASS), 1113t, 1113–1116 heart rate response to deep breathing, 1114–1115, 1115t in diabetic lumbosacral radiculoplexus neuropathy, 1999, 2002, 2002t normative database for, 1114, 1114t orthostatic blood pressure, 1115–1116, 1116t QSART, 1114, 1114t–1115t Valsalva ratio, 1115, 1115f–1116f, 1116t Composite scores derived from nerve tests, 971–972, 976–977, 979f as measures of severity of neuropathy over time, 981 functional abnormality when test values fall within normal range, 979–981 in genetic linkage analysis, 981 vs. individual attributes of nerve conduction abnormality, 977–979, 980f–982f Composite scores of neurologic signs, 1038–1041 electronic case report form of Clinical Neuropathy Assessment, 1040–1041, 1042f–1045f history of, 1038 MNSI questionnaire for diabetic neuropathy, 1040, 1041f nerve conduction tests as impairment measures, 1039–1040 Neuropathy Impairment Score, 1038–1039, 1096 Total Neuropathy Score, 1040

Index Compound action potentials, 1015–1026 fiber composition and, 1015–1016 fiber diameter and, 1015, 1018 fiber groups and sensation, 1016–1017 nomenclature for, 1015–1016 letter designations of, 1015–1016 roman numeral classification, 1016 of sural nerve, 1017–1026 fiber diameter and, 1018 in Déjérine-Sottas neuropathy, 1646f in dominantly inherited amyloidosis, 1021f, 1021–1022 in Friedreich’s ataxia, 1020f, 1020–1021, 1817f in hereditary motor and sensory neuropathy type I, 1025–1026, 1027f in hereditary sensory neuropathy type I, 1022, 1023f, 1817f in hereditary sensory neuropathy type II, 1022–1023, 1024f in normal nerve, 1018–1020, 1019f in relapsing inflammatory polyradiculoneuropathy, 1024–1025, 1026f–1027f in tabes dorsalis, 1022, 1022f, 1817f in uremic neuropathy, 1023–1024, 1025f method of study of, 1017–1018 reconstruction of, 1016 sources of distortion in recording of, 1017 Compound muscle action potentials (CMAPs), 900 in brachial plexopathy, 1351 in Charcot-Marie-Tooth disease type 1A, 1662 in chronic inflammatory demyelinating polyradiculoneuropathy, 2234 in chronic obstructive pulmonary disease–associated neuropathy, 2022 in common peroneal neuropathy, 1498 in congenital myasthenic syndromes resembling Lambert-Eaton myasthenic syndrome, 842 slow-channel syndromes, 848 sodium channel myasthenia, 858 with reduced quantal release and central nervous system symptoms, 842 in critical illness polyneuropathy, 2027 in diagnosis of conduction block, 2283 in end-plate acetylcholinesterase deficiency, 547 in femoral neuropathy, 1491 in giant axonal neuropathy, 1651 in Guillain-Barré syndrome acute inflammatory demyelinating polyneuropathy, 2200 axonal forms, 2201 pediatric, 2716 in hepatitis C–associated neuropathy, 2019 in hereditary motor and sensory neuropathy, 1637–1638, 1647 type II, 1718, 1722 in Lambert-Eaton myasthenic syndrome, 837 in mice deficient in porphobilinogen deaminase, 551 in monoclonal gammopathy of undetermined significance, 2265 in multifocal motor neuropathy, 2182 in neurapraxia, 902 in neuropathy associated with chronic hepatic disease, 2018 in paraneoplastic subacute sensory neuronopathy, 2475 in pediatric polyneuropathies, 2712

Compound muscle action potentials (CMAPs) (Continued) in perinatal brachial plexopathy, 2731 in POEMS syndrome, 2457–2458 in tick paralysis, 2717 in true neurologic thoracic outlet syndrome, 1360 in uremic neuropathy, 2023, 2023f after renal transplantation, 2023f, 2025 Compression garments, for orthostatic hypotension, 2658–2659, 2674 Compression neuropathies, 1140, 1140t, 1391–1400. See also Mononeuropathy(ies). clinical features of, 798 compression vs. entrapment, 797–798 conduction block in, 2280–2281 cubital tunnel syndrome, 1479–1480 differential mechanical forces in, 1391–1396 role in acute nerve compression, 1391–1394, 1392f–1394f role in chronic nerve compression, 1394–1396, 1395f due to compression by hematoma, 1518 endoneurial fluid flow in, 656–657 impaired axonal transport in, 1399 impaired nerve trunk gliding capacity in, 1399–1400 in diabetes, 1140, 1964t, 1964–1965 ischemia and increased endoneurial fluid pressure in, 1396–1399, 1397f–1398f ischemia of nerves compressed by swollen muscles, 1517–1518 mechanism of nerve injury in, 800, 801f–802f myokymia in, 1143 nerve crush and, 798–799 of axillary nerve, 1470 of brachial plexus, 1359 of long thoracic nerve, 1464 of median nerve, 1435–1454 in children, 2725–2726 of peroneal nerve, in children, 2729 of radial nerve, 1475 in children, 2726–2727 of sciatic nerve, 1494 in children, 2727–2728 of suprascapular nerve, 1467 of tibial nerve, 1501 of trigeminal nerve, 1214 of ulnar nerve, 1478–1482 in children, 2726 overview of, 1391 wallerian degeneration in, 776 with polyneuropathy, 1140, 1165 Computed tomography (CT) in brachial plexus injury, 1350, 1522 in hydatidosis, 2170 in lumbosacral plexus lesions, 1377 in POEMS syndrome, 2460f of facial canal, 1221 of temporal bone fracture, 1233f to detect malignancy underlying paraneoplastic neuropathy, 2482 Computer use, carpal tunnel syndrome and, 1446–1447 Computer-Assisted Sensory Examination System IV (CASE-IV), 1038, 1066–1067, 1069f–1070f, 1071t. See also Quantitative sensation testing (QST). algorithm of testing and finding threshold for, 1076 baseline conditions of stimulus for, 1072 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xix Computer-Assisted Sensory Examination System IV (CASE-IV) (Continued) calibration of stimulus magnitude for, 1074–1075, 1074f–1075f components of, 1079–1080 concordance of results with neuropathic impairment, 1088 cooling detection threshold using, 1085–1086, 1085f–1086f heat-pain threshold testing using, 1086–1087 ideal stimuli and testing algorithms for, 1072 patterns of abnormality with, 1067 stimulus waveform for, 1073 subjective differences among anatomic sites, 1067 vibration detection threshold using, 1080–1082 calibration of stimuli, 1080 forced-choice algorithm, 1076–1077, 1078f, 1080–1081 4, 2, 1 stepping algorithm with null stimuli, 1070, 1081–1082, 1081f–1082f vibratory transducer, 1080 vs. clinical assessment, 1066 Concomitant blocking, 993f, 994 Conduction. See also Nerve conduction studies (NCSs). after nerve transection (rat), 772–773 distal slowing of, 902 in demyelinating neuropathies, 125–126 ischemia effects on, 119–121, 121f nerve excitability studies of, 113–126 temperature effects on, 125, 906–907 Conduction block, 903, 904f, 1512, 2280–2282 criteria for, 931–934, 2281, 2281t, 2283 definition of, 933, 2280 demyelinating, 1464 in brachial plexopathies, 1344 detection of difficulty in, 2281 inching technique for, 934, 2283 in chronic inflammatory demyelinating polyradiculoneuropathy, 125–126, 2222, 2227t, 2228 in common peroneal neuropathy, 1498 in compressive neuropathies, 2280–2281 in immune brachial plexus neuropathy, 2302 interphase cancellation and misdiagnosis of, 2283 multifocal motor neuropathy with, 122–124, 935f, 1306–1307, 2282f, 2282–2284. See also Multifocal motor neuropathy and conduction block (MMN-CB). progression of, 2283 sites of, 2283–2284 paralysis or weakness due to, 1142 partial, 932 pathologic basis of, 2281 prolonged, 1512 temporal dispersion in recording of, 2281–2282 transient in nondegenerative nerve injuries, 1405 in vasculitic neuropathy, 2367 Conduction velocity, 113. See also Nerve conduction studies (NCSs). age effects on, 486, 671, 907, 907f, 908t, 1176 calculation of, 901–902 fiber diameter and, 1015, 1016 in axonal neuropathy with loss of fast-conducting fibers, 903 in diabetic neuropathy, 1978–1979

xx Conduction velocity (Continued) in lead neuropathy, 2529 in myelopathies, 903 latency and in motor nerves, 900–902, 902f in sensory nerves, 904 of dorsal root ganglion neurons, 165, 166 Conduits, for nerve injury, 1424 Confocal microscopy, of cutaneous nerves, 876, 877f Conformation-sensitive gel electrophoresis (CSGE), 1557, 1558f Congenital cataracts, facial dysmorphism, and neuropathy (CCFDN), 1775, 1776f Marinesco-Sjögren syndrome with myoglobinuria and, 1776 Congenital cranial nerve defects, 1197–1198 Congenital hypomyelinating neuropathy (CHN), 1625, 1769 autosomal recessive, 1777, 1778f clinical features of, 1625, 1632 infantile, 2713–2714 vs. Déjérine-Sottas neuropathy, 1625 Waardenburg-Hirschsprung disease and, 2714 with EGR2 mutation, 1707–1709, 1708f, 1709t clinical features of, 1710t with P0 gene mutation, 579 with SOX10 mutation, 2714 Congenital insensitivity to pain with anhidrosis, 704, 886, 1837 in children, 2723 skin biopsy in, 886 Congenital megacolon, 705 Congenital myasthenic syndromes (CMSs), 838–860 choline acetyltransferase deficiency, 839–841 clinical features of, 839–840, 840f end-plate studies in, 840 molecular pathogenesis of, 840–841, 841f therapy for, 860 classification of, 839t end-plate acetylcholinesterase deficiency, 842–844 clinical features of, 842–843, 843f–844f molecular pathogenesis of, 843–844, 846f pathogenesis and pathology of, 843, 844f–845f therapy for, 860 fast-channel syndromes, 847t, 851–852 caused by selective abnormality of gating, 852 caused by unstable kinetics, 852 clinical features of, 851 end-plate studies in, 851, 851f low-affinity syndromes, 852 molecular pathogenesis of, 851f, 851–852 therapy for, 860 investigation of, 839t mutations causing AChR deficiency with/without minor kinetic abnormality, 852–855 clinical features of, 852–853, 853f end-plate studies in, 853, 854f molecular pathogenesis of, 853–855, 855f paucity of synaptic vesicles and reduced quantal release, 841–842 postsynaptic syndromes, 844–847 resembling Lambert-Eaton myasthenic syndrome, 842

Index Congenital myasthenic syndromes (CMSs) (Continued) slow-channel syndromes, 847t, 847–851 clinical features of, 847, 847f end-plate studies in, 848–849, 848f–850f molecular pathogenesis of, 848f, 849–851 therapy for, 860 sodium channel myasthenia, 857–859 clinical features of, 857, 858f end-plate studies in, 857–858 molecular pathogenesis of, 858–859, 859f relation to other sodium channel disorders, 859 therapy for, 860 syndromes caused by rapsyn deficiency, 855–857 clinical features of, 856 end-plate studies in, 857 molecular pathogenesis of, 856f, 857 therapy for, 860 therapy for, 860 with episodic apnea, 840 with plectin deficiency, 859 therapy for, 860 with reduced quantal release and central nervous system symptoms, 842 Congenital tomaculous neuropathy, 823 Congestive heart failure in amyloidosis, 2433 in POEMS syndrome, 2460 muscle sympathetic nerve activity in, 333 Conjunctival hyperemia, in Horner’s syndrome, 207 Connective tissue disease (CTD) dysphagia in, 2686 facial nerve paralysis and, 1234 laboratory investigation for, 1172–1173 pediatric polyneuropathy and, 2719, 2720f vasculitic neuropathy and, 2340t, 2341t, 2357–2361 Connective tissues of peripheral nerves, 35–42, 36f endoneurium, 36–39 epineurium, 41–42 in diabetic neuropathy, 1967 perineurium, 39–41 Connexin 29 (Cx 29), 420, 1795 Connexin 32 (Cx 32), 413, 418f, 419t, 420 localization of, 1686f mutations of analysis of, 1728, 1773–1774 in X-linked Charcot-Marie-Tooth disease, 543, 704, 1259, 1628–1629, 1727–1728, 1729f, 1791–1793, 1792f with hearing loss, 1775 role in gap junction formation, 1576, 1793–1795 in myelin sheath, 1774 Schwann cell expression of, 61, 1576 structure of, 1794f tissues expression pattern of, 1793 transgenic mouse studies of, 543–545, 1576–1578 Cx32 mutant mice, 543–545, 544f–545f, 1577–1578 alterations at molecular level, 1577 functional gap junctions in Cx32-deficient mice, 1577–1578 xenograft model, 1578 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Connexin 32 (Cx 32) (Continued) Cx32-deficient mice, 1576–1577 electrophysiology of, 1577 histology of, 1576–1577 random Cx32 inactivation in female mice, 1577 Constipation, 2610, 2665, 2691–2692 after spinal cord injury, 2699 clinical assessment of, 2691 diagnostic evaluation of, 2691–2692 etiology of, 2691 in multiple sclerosis, 2700 in Parkinson’s disease, 2695 management of, 2692, 2693t Constitutional symptoms, 1167 Construct validity of health outcome measures, 1057 Contactin-associated protein (Caspr), 549 mice deficient in, 550, 551f Contactin-associated protein 2 (Caspr2), 549–550 Continence. See also Incontinence. fecal, 284–286, 2627 external anal sphincter control of, 284, 2627 mechanisms of, 285–286 puborectalis disruption and, 284 urinary, 300, 300f, 2626–2627 detrusor contraction and, 2626 internal and external sphincter control of, 2626 Continuous muscle fiber activity, 946 Contracture(s), 2622, 2628–2629 edema and, 2629 in hereditary motor and sensory neuropathy, 1636 Volkmann’s, 1517 Cooling detection threshold (CDT), 1038. See also Quantitative sensation testing (QST). algorithm of testing and finding threshold for, 1076, 1077f calibration of stimulus magnitude for, 1075, 1075f CASE III testing of, 1069f CASE IV testing of, patterns of abnormality on, 1067 instructions for 4, 2, 1 step testing algorithm with null stimuli for, 1070 reproducibility of vibration detection threshold results, 1091, 1092f stimulus waveform for, 1073 subjective differences among anatomic sites, 1067 using CASE IV, 1085–1086, 1085f–1086f COPD (chronic obstructive pulmonary disease), neuropathy and, 2021–2022 Copper deficiency of, 2058 in diabetic patients, 512 Coproporphyria, hereditary (HCP), 1883–1891. See also Porphyric neuropathy. clinical manifestations of, 1886, 1887 neurologic disease, 1888 inheritance of, 1887 laboratory studies in, 1889t, 1889–1890 mechanisms of neuronal damage in, 1885–1886 treatment of, 1890–1891 Cornea clouding of, in Tangier disease, 1908 congenital trigeminal anesthesia and ulceration of, 1211

Index Corpora amylacea, 709 Corpus callosum agenesis, in Andermann syndrome, 1738 Corticoanal pathways, assessment of, 289–290 Corticobulbar palsy, 1533 Corticospinal impulse transmission, 989 Corticosteroids. See also Glucocorticoids (GCs). adverse effects of, 637, 2381–2383, 2382t–2383t osteoporosis, 2382–2383, 2383t for amyloidosis, 2437–2438 for Bell’s palsy, 637, 1230–1231 for brachial plexus neuropathy hereditary, 1763, 1766 immune, 2306 for carpal tunnel syndrome, 637, 1451 for chronic inflammatory demyelinating polyradiculoneuropathy, 637, 2179t, 2222, 2234–2235, 2236–2237 in children, 2719, 2231 in diabetic patients, 1956 for diphtheria, 2148 for Guillain-Barré syndrome, 2199, 2213 for herpes zoster, 2120 for leprosy reactions, 637, 2104 for Lewis-Sumner syndrome, 2232–2233 for lumbosacral radiculoplexus neuropathy, 2012 diabetic, 2011–2012, 2013f for nonmalignant inflammatory sensory polyganglionopathy, 2317 for POEMS syndrome, 2463 for radiculopathy, 2634 due to disc herniation, 1326 for Ramsay Hunt syndrome, 1232 for sarcoidosis, 2423 for temporal arteritis, 2362 for vasculitic neuropathy, 637, 2336–2337, 2375t–2376t, 2378–2380 mechanism of action of, 636 pharmacokinetics of, 636–637 use in neuropathy, 637 Corynebacterium diphtheriae, 2147–2150. See also Diphtheria and diphtheritic neuropathy. Costal cervical artery, 1341 Cough reflex, 1283–1284 Coughing, heart rate response to, 1123 COX (cyclooxygenase) enzymes, 182 COX (cyclooxygenase) inhibitors, for orthostatic hypotension, 2677 CPEO (chronic progressive external ophthalmoplegia), 1941 Cranial nerves, 17, 1191–1203 compared with spinal nerves, 11–12 congenital defects of, 1197 EGR2 mutations and involvement of, 1709 eighth (vestibulocochlear), 1253–1268 age-related changes of, 1253–1254 anatomy of, 1253 aplasia/hypoplasia of, 1256, 1259, 1259t compression by osseous lesions of internal auditory canal, 1264 course of, 1253, 1255f diagnostic approach to disorders of, 1254–1256 drug toxicity and, 1268 in acquired auditory neuropathy, 1259–1261, 1262f in autoimmune disorders, 1268 in vestibular neur(on)itis, 1262–1263, 1263f inherited disorders of, 1259, 1260f–1261f

Cranial nerves (Continued) neurovascular cross-compression of, 1263–1264 noise-induced damage of, 1268 secondary degeneration of, 1268 traumatic injury of, 1268 vestibular schwannoma, 1264t–1265t, 1264–1268, 1265f–1267f eleventh (accessory), 1286 anatomy of, 1286 course of, 1286 cranial, 1277 injury during surgery in neck, 1286, 1519, 1520f lesions of, 1286 clinical features of, 1286 muscles innervated by, 956t nerve conduction studies of, 909 fifth (trigeminal), 1207–1215 anatomy and physiology of, 1207–1209 divisions and branches, 1207, 1208f motor root, 1209 sensory root, 1208 trigeminal ganglion, 1207 trigeminal nuclei, 1208–1209 blink reflex to test integrity of, 918–919, 920t, 921t congenital anesthesia of, 1211 course of, 1194f differential diagnosis of lesions of, 1209, 1210t facial numbness due to lesions of, 1209 idiopathic neuropathy of, 1211–1212 in perineurial tumor spread, 1210 in trigeminal neuralgia, 1213–1215, 1214t in trigeminal trophic syndrome, 1212 investigation of lesions of, 1212–1213 microvascular decompression of, 1215 other tumors of, 1210–1211 procedures to anesthetize territory of, 1215 role in taste sensation, 1207 schwannomas of, 1210 territory supplied by, 1207, 1209f toxic lesions of, 1211 traumatic injury of, 1209 fourth (trochlear), 1192–1201 assessing function in third nerve palsy, 1197 blood supply of, 1195 clinical features of fourth nerve palsy, 1195–1197 course of, 1193–1194, 1194f histology of, 1195, 1196f in diabetes, 1960 muscles innervated by, 1192 neoplasms affecting, 1199–1200, 1200f nucleus of, 1193 traumatic injury of, 1198 vascular diseases affecting, 1201 in cavernous sinus, 1193–1195 involvement in amyloidosis, 2433 familial, 2440–2441 involvement in angiostrongyliasis, 2164 involvement in brachial plexus neuropathy hereditary, 1755 immune, 2301 involvement in chronic inflammatory demyelinating polyradiculoneuropathy, 1139, 2224, 2225, 2225t involvement in cytomegalovirus infection, 2122 involvement in diabetes, 1140, 1960–1961 involvement in Guillain-Barré syndrome, in children, 2716 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxi Cranial nerves (Continued) involvement in herpes zoster, 2119–2120 involvement in hypothyroidism, 2040 involvement in Lyme disease, 1140, 2111 involvement in polyneuropathies, 1139–1140 involvement in sarcoidosis, 1140, 1203, 1268, 2415, 2417–2418 involvement in trichinosis, 2164–2165 ninth (glossopharyngeal), 1273–1277 anatomy of, 1273–1274 carotid sinus nerve, 1274 motor fibers, 1273 parasympathetic fibers, 1273 sensory fibers, 1273–1274 carotid sinus baroreceptor innervation by, 323 blood pressure effects of nerve block, 325, 326f, 325, 326f, 1274, 1275f carotid sinus hypersensitivity and, 1277 clinical features of lesions of, 1274 course of, 1273 extramedullary lesions of, 1274–1275 in glossopharyngeal neuralgia, 1214, 1276f, 1276–1277 intramedullary lesions of, 1274 testing function of, 1274 parasitic infections of, 2154t seventh (facial), 1219–1246 acute paralysis of, 1223–1236 idiopathic, 1223–1231. See also Bell’s palsy. in amyloidosis, 1232 in benign intracranial hypertension, 1235 in connective tissue diseases, 1234 in diabetes mellitus, 1234, 1961 in granulomatous disease, 1232 in HIV infection, 1234 in pregnancy, 1234–1235 miscellaneous causes of, 1235–1236 Ramsay Hunt syndrome, 1231–1232 traumatic, 1232–1233, 1233f tumor-related, 1235 blink reflex to test integrity of, 918–919, 920t, 921t blood supply of, 1220–1221 congenital/pediatric paralysis of, 1236–1238 in cardiofacial syndrome, 1237 in leukemia, 1237 in Melkersson-Rosenthal syndrome, 1237–1238 infectious, 1236–1237 traumatic, 1237 course of, 1194f, 1219–1220 functional anatomy of, 1222, 1223f–1224f gross anatomy of, 1219–1220 in geniculate neuralgia, 1246 in Guillain-Barré syndrome, 1203 in hemifacial spasm, 1238–1245 in internal auditory canal, 1254, 1255f in leprosy, 2089f, 2092–2093, 2095f in sarcoidosis, 1203, 2417–2418 inherited/familial palsy of, 1236 motor root of, 1219–1220 nerve conduction studies of, 908–909, 909t neuroimaging of, 1221–1222, 1221f–1222f sensory root of, 1219–1220 terminal branches of, 1220 tumors of, 1235 imaging of, 1221–1222, 1221f–1222f, 1235 vulnerability to pathologic processes, 1219

xxii Cranial nerves (Continued) sixth (abducens), 1192–1201 blood supply of, 1195 clinical features of sixth nerve palsy, 1197 course of, 1193–1194, 1194f histology of, 1195, 1196f in diabetes, 1960 muscles innervated by, 1192 neoplasms affecting, 1199–1200 nucleus of, 1193 traumatic injury of, 1198 vascular diseases affecting, 1201 tenth (vagus), 17, 1277–1285 anatomy of, 1277–1279 somatic sensory fibers, 1278 visceral afferent fibers, 1278 visceral efferent fibers, 1278–1279 aortic baroreceptor innervation by, 323 blood pressure effects of nerve block, 325, 326f auricular branch of, 1277 clinical features of lesions of, 1279 laryngeal paralysis, 1279 palatal and pharyngeal paralysis, 1279 course of, 1277–1278 disorders of autonomic function of, 1280–1284 vagal afferents, 1283–1284, 1284f vagal efferents, 1280–1283, 1281f–1283f extramedullary lesions of, 1285 intramedullary lesions of, 1284–1285 pharyngeal branch of, 1277 tests of autonomic function of, 1279–1280 enteroinsular peptide release, 1280 esophageal, gastric, and intestinal motility, 1280 gastric acid secretion, 1280 third (oculomotor), 1191–1202 aberrant regeneration of, 1198–1199 blood supply of, 1195 clinical features of third nerve palsy, 1195 course of, 1193–1195, 1194f histology of, 1195, 1196f in diabetes, 804, 1201–1202, 1960–1961 muscles innervated by, 1191–1192 neoplasms affecting, 1199–1200 nuclei of, 1192f, 1192–1193 traumatic injury of, 1198 vascular diseases affecting, 1201 twelfth (hypoglossal), 18, 1286–1288 anatomy of, 1286–1287 course of, 1286–1287 dural branches of, 1287 lesions of, 1287 clinical features of, 1287 glossodynia, 1288 muscles innervated by, 1286 Craniodiaphysial dysplasia, 1264 CRASH syndrome, axon guidance in, 467–468 axonin-1 and associated kinases, 468 cytoskeleton association via ankyrin, 468 fibroblast growth factor receptor and Src kinase, 468 L1 and modulation of semaphorin signaling, 468 C-reactive protein, 575 Creatine kinase (CK) in hereditary motor neuronopathy, 1606 in multifocal motor neuropathy and conduction block, 2286 Criterion validity of health outcome measures, 1057

Index Critical illness myopathy, 2030 Critical illness polyneuropathy, 1141, 2017, 2026–2030 after bone marrow transplantation, 2498t axonal degeneration in, 2027–2028, 2028f definitions related to, 2026–2027 differential diagnosis of, 2029–2030 electrophysiology of, 2027 historical descriptions of, 2027 overview of, 2026 possible mechanisms of, 2028–2029 Crk-associated tyrosine kinase substrate (Cas), 455 CRMP (collapsin response mediator protein), 585–586 Crohn’s disease, 1167 Cronbach’s alpha, 1056–1057 Crow-Fukase syndrome. See POEMS syndrome. CRPS (complex regional pain syndrome) type 1 adrenergic sweating in, 1106 detecting sympathetic dysfunction in, 1105 Crush artifact, 769, 771f–772f Crush injury, 798–799 Crutch palsy, 1370 Cryoglobulinemia, 2257–2258 clinical and laboratory features of, 2462t hepatitis C and, 2018–2020, 2019f Raynaud’s phenomenon and, 1168, 2019 types I, II, and III, 2018–2019 Cryoglobulinemic vasculitis, 2354t, 2405 classification and definition of, 2338t, 2340t hepatitis C–associated, 2019f, 2335, 2350–2351 treatment of, 2375t incidence of, 2340t malignancy and, 2351 Cryptococcus neoformans, in HIV infection, 2140 Cryptogenic sensory polyneuropathy (CSP), 2321–2331 age distribution of, 2321 clinical course of, 2322–2323 clinical findings in, 2322 diagnosis of, 2323–2326 autonomic testing, 2324 electrodiagnostic tests, 2323–2324 quantitative sensation testing, 2324 serologic tests, 2326 skin biopsy, 2325–2326 sural nerve biopsy, 2325, 2325f differential diagnosis of, 2327t, 2327–2331 diabetic polyneuropathy, 2327 inherited neuropathy, 2329–2331, 2330t orthopedic disease, 2331 Sjögren’s syndrome, 2328t, 2328–2329, 2329f gender distribution of, 2322 heterogeneity of, 2321 monoclonal gammopathy and, 2326 symptoms of, 2322 terminology for, 2321 treatment of, 2326–2327 utility of serial evaluation in, 2323 CSF. See Cerebrospinal fluid (CSF) analysis. CSGE (conformation-sensitive gel electrophoresis), 1557, 1558f CSP. See Cryptogenic sensory polyneuropathy (CSP). CSS. See Churg-Strauss syndrome (CSS). CTD. See Connective tissue disease (CTD). CTDP1 gene mutations, 1770t, 1775–1776 CTP (central tissue projection), 68 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

CTS. See Carpal tunnel syndrome (CTS). Cuban epidemic neuropathy, 2063, 2066 clinical features of, 2066 dorsolateral myelopathy, 2066 optic neuropathy, 2066 peripheral neuropathy, 2066 sensorineural deafness, 2066 epidemiology of, 2066 etiology of, 2066 Cubital tunnel syndrome, 1479–1483. See also Ulnar neuropathy. in Hurler’s syndrome, 1865 nerve conduction studies in, 910 Cuneocerebellar tract, 158 Current-threshold relationship, 118f, 118–119, 120f Cutaneoanal reflex, 284 Cutaneous leukocytoclastic angiitis (CLA), 2361–2362, 2405 classification and definition of, 2338t, 2339 incidence of, 2340t Cutaneous nerves, 869–891. See also Skin biopsy. contributions of skin biopsy to diagnosis, 883–888 methods for analysis of, 872–876 biopsy methods, 872, 874f choice of biopsy/blister location, 872, 873f epidermal sheet preparations, 876 polymerase chain reaction techniques, 876 skin blister method, 875f, 875–876 staining biopsy specimens, 873–874, 874f–875f microscopy of, 876, 877f overview of, 869–870 quantitation of, 876–883 by tracing, 878, 878f correlation between epidermal nerve and sural nerve, 883 correlation with tests of unmyelinated nerve function, 882–883 dermal nerves, 879 epidermal nerves, 876–879 epidermal nerves in skin blisters and epidermal sheets, 879, 880f manual counting, 878–879 morphologic changes in sensory neuropathy, 882 nerve counting rules, 877f, 877–878 nerves to arrector pilorum muscles, 881 nerves to arteries, 881 nerves to hair follicles, 881–882 normal epidermal nerve fiber density, 879, 879f, 880t subepidermal neural plexus rating, 879–880, 881f sweat gland nerves, 880–881, 882f utility of skin biopsy, 883 research uses of skin biopsy, 888–891 sensory receptors, 1064, 1064f skin anatomy and, 870–872, 871f Cutaneous neurofibroma, 2591 CV2 antigen, 585–586, 595 Cx 29 (connexin 29), 420, 1795 Cx 32. See Connexin 32 (Cx 32). Cyanide neurotoxicity, 2069, 2073–2075 clinical features of, 2074 epidemiology of, 2073 konzo, 2074–2075 pathogenesis of, 2074 tropical ataxic neuropathy and, 2070, 2073 tropical spastic paraparesis and, 2074 urinary test for, 2075

Index CYC. See Cyclophosphamide (CYC). Cycas circinalis, 2063 Cyclic adenosine monophosphate (cAMP), in netrin signaling, 461 Cyclic guanosine monophosphate (cGMP) in penile erection, 315 in semaphorin signaling, 465 Cyclin-dependent kinases, as Schwann cell mitogens, 357 Cyclooxygenase (COX) enzymes, 182 Cyclooxygenase (COX) inhibitors, for orthostatic hypotension, 2677 Cyclophilins, 638 Cyclophosphamide (CYC), 637–638 adverse effects of, 638, 2383, 2384t hemorrhagic cystitis, 638 for anti-MAG neuropathy, 2179t for antisulfatide neuropathy, 2179t for chronic inflammatory demyelinating polyradiculoneuropathy, 638, 2236 in children, 2719 in diabetic patients, 1956 for diabetic lumbosacral radiculoplexus neuropathy, 2011 for GALOP syndrome, 2179t, 2189 for hepatitis C–associated neuropathy, 2020 for IgM paraproteinemia, 638 for monoclonal gammopathy and peripheral neuropathy, 2268–2269 for multifocal motor neuropathy and conduction block, 638, 2179t, 2182–2184, 2290, 2291, 2293f, 2293–2294 for vasculitic neuropathy, 638, 2336–2337, 2375t–2376t, 2378–2380 paraneoplastic, 2374, 2479 mechanism of action of, 637–638 pharmacokinetics of, 638 use in neuropathy, 638 Cyclosporiasis, 2155t, 2173 Guillain-Barré syndrome and, 2173 transmission of, 2173 Cyclosporine, 638–639 adverse effects of, 639 for chronic inflammatory demyelinating polyradiculoneuropathy, 639, 2179t, 2236 in children, 2719 for sarcoidosis, 2423 for vasculitic neuropathy, 2381 mechanism of action of, 638–639 use in neuropathy, 639 Cyst(s) Baker’s, 1501–1503 hydatid, in liver, 2169, 2170 paraglenoid, 1467 Cysticercosis, 2155t, 2168 clinical features of, 2168 etiologic agent in, 2163f, 2168 life cycle of, 2168 peripheral neuropathy in, 2168 transmission of, 2168 Cystometrograms, 300f Cytochrome P-450 system, heme biosynthesis and, 1884–1885 Cytokine-like factor-1 (CLF-1), 380 Cytokines, 565t anti-inflammatory, 2344 calcineurin activation of, 639 effects of intravenous immunoglobulin on, 643 in experimental autoimmune neuritis, 617–618, 618f–619f

Cytokines (Continued) in HIV-associated distal symmetrical polyneuropathy, 2133 in ischemia-reperfusion injury, 522–523, 525f, 675, 677 in POEMS syndrome, 2454–2455 in vasculitis, 2343–2344 in wallerian degeneration, 1413–1415, 1416f–1417f neuropoietic, 180–181 pro-inflammatory, 181–182, 2342, 2344 glucocorticoid effects on, 636 release by macrophages, 560 Cytomegalovirus (CMV) infection, 2121–2123 antiganglioside antibodies and, 591 anti-MAG paraproteinemic polyneuropathy and, 590 congenital, 2121 cranial nerve involvement in, 2122 diagnosis of, 2122–2123 Guillain-Barré syndrome and, 590, 2122 acute inflammatory demyelinating polyneuropathy, 2211 in HIV infection, 2139 disease-stage specificity of, 2130 distal symmetrical polyneuropathy and, 2132 prevalence of, 2130 in HIV-infected patients, 2121, 2123 latent, 2121 mononeuritis multiplex and, 2122 myeloradiculopathy and, 2122 neurologic complications of, 2121 neuropathies associated with, 2118t pathology of, 713 polyradiculopathy due to, 1331 treatment of, 2123 vasculitic neuropathy and, 2335 treatment of, 2336, 2378 virology of, 2121 Cytoplasmic neuronopathy, 2201 Cytosine arabinosine, peripheral neuropathy induced by, 2490, 2490t, 2498t

d4T (stavudine), 2133–2135, 2134f–2135f, 2563 D-39 (Diabetes-39), 1058 DAG (diacylglycerol), in diabetic neuropathy, 1977 Dantrolene, for spasticity, 1540 3,4-DAP for congenital myasthenic syndromes, 860 for Lambert-Eaton myasthenic syndrome, 860 Dapsone for leprosy, 2103 neuropathy induced by, 2557–2558 Davidenkow syndrome, 1139, 1615 vs. scapuloperoneal hereditary motor neuronopathy, 1616 DCA (dichloracetylene), 2520 DCC. See Netrin/DCC/UNC-5. DCCT (Diabetes Control and Complications Trial), 1952, 1969 DCP (Diabetes Care Profile), 1058, 1060 DCRPN (diabetic cervical radiculoplexus neuropathy), 1993, 2303 clinical features of, 2000 diabetic lumbosacral radiculoplexus neuropathy and, 1962, 1995, 2000 historical overview of, 1995 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxiii DDAVP (desmopressin), for orthostatic hypotension, 1958t, 2675 ddc (zalcitabine), 2130, 2131, 2133–2135, 2134f–2135f, 2563 ddI (didanosine), 2133–2135, 2134f–2135f, 2563 DDPAC (dementia-disinhibitionparkinsonism-amyotrophy complex), 1600 Deafness. See Hearing loss. Deconditioning, 2628–2629 Decubitus ulcers, 2629 Deep breathing, heart rate response to, 1109, 1109t, 1110f Deep venous thrombosis (DVT) immobility and, 2629 in intensive care unit patients, 2616 prophylaxis for, 2609, 2629 Defecation, 279, 284–287, 2627–2628 external anal sphincter during, 284 mechanisms of, 285f, 285–286 obstructed, 286–287 balloon expulsion test for, 286–287, 287f colonic transit in, 287, 287f parasympathetic nerves in, 282 Defecation index, 306 Defecography, 290t, 291 Deferoxamine, for experimental diabetic neuropathy, 516, 516t Déjérine-Sottas neuropathy (DSN), 1688, 1769 clinical features of, 1624–1625, 1632 diagnostic criteria for, 1625 gene mutations causing, 703, 1625, 2714 EGR2, 1707, 1708f, 1710t P0, 579, 1688, 1690t–1691t SOX10, 2714 historical background of, 1624–1625, 1681 in children, 2708, 2713–2714 inheritance pattern of, 1624, 1625, 2714 nerve conduction studies in, 1624, 1625 pathology of, 819, 1625, 1641–1643, 1645f–1649f periaxin in, 580 vs. congenital hypomyelinating neuropathy, 1625 Delayed-type hypersensitivity, in vasculitis, 2348 Delirium, 1168 Dementia Alzheimer’s, 703 amyotrophic lateral sclerosis and, 1537, 1600 Dementia-disinhibition-parkinsonismamyotrophy complex (DDPAC), 1600 Demyelination age-related, 487–491 axonal caliber and, 435 axonal degeneration and, 777 conduction block and, 931, 2281 multifocal motor neuropathy with, 2286–2287 immune, 820 impulse conduction and, 125–126 in acromegalic neuropathy, 2045 in acute inflammatory demyelinating polyneuropathy, 2202–2204, 2203f–2206f in adrenoleukodystrophy, 1167, 1867 in adrenomyeloneuropathy, 1167, 1867 in brachial plexopathies, 1344 in cervical spondylotic myelopathy, 1313 in cherry-red spot–myoclonus syndrome, 1865 in chloroquine neuropathy, 2556 in chronic inflammatory demyelinating polyradiculoneuropathy, 2239–2241, 2240f–2241f

xxiv Demyelination (Continued) in compression neuropathies, 1398, 1398f in diabetic motor neuropathy, 1955 in diabetic neuropathy, 1965, 1966f in diphtheria, 2148–2149, 2130f–2131f in disulfiram neuropathy, 2558 in eosinophilia-myalgia syndrome, 2571 in experimental autoimmune neuritis, 615, 615f in experimental diabetic neuropathy, 517f, 517–518 in gold neuropathy, 2544 in hereditary motor and sensory neuropathy, 1638–1639, 1640, 1644f–1645f Déjérine-Sottas neuropathy, 1641–1642, 1645f–1648f transgenic animals overexpressing PMP22, 1563, 1563f with P0 mutation, 1695 in hereditary sensory and autonomic neuropathy type II, 1829f, 1830 in HIV-associated sensory polyneuropathies, 2132 in hypothyroid polyneuropathy, 2041–2042 in IgM monoclonal gammopathy with neuropathy, 2262, 2263 role of complement in, 2267 in Krabbe’s disease, 1855, 1857, 1857f in leprosy, 2100f in lysosomal disorders, 1846 in metachromatic leukodystrophy, 1847, 1849–1852 in multiple sulfatase deficiency, 1855 in P0-deficient mice, 537–538, 539f–540f in paraneoplastic neuropathy, 2477–2478 in perhexilene neuropathy, 2564 in phenytoin neuropathy, 2565 in POEMS syndrome, 2457, 2458 in suramin neuropathy, 2567 in tacrolimus neuropathy, 2568 in taxane neuropathy, 2569 ischemic, 522 mitochondrial diseases with, 1944–1946 Leigh disease, 1944–1945, 1945f mitochondrial neurogastrointestinal encephalopathy, 1945–1946 nerve conduction studies in, 902–904, 907–908 PMP22 point mutations as cause of, 1665–1667, 1666f transgenic animal models of, 1566–1571, 1567t, 1569f–1571f polyneuropathies with, 1166t, 1166–1167 potassium channels in, 10 segmental, 3–4, 788–790, 1154, 1166 definition of, 790 fiber diameter and compound action potential in, 1024–1025, 1027f in lead intoxication, 820–823, 821f–822f internodal, 790 macrophage-mediated, 560 paranodal, 790 primary, 790 remyelination and, 790 secondary axonal atrophy and, 780–788, 784f, 790 in Friedreich’s ataxia, 783, 785f in uremic neuropathy, 781, 2024 vs. primary demyelination, 790–791 teased fiber analysis of, 748, 749f teased fiber analysis of, 748, 749f, 753 uniform vs. nonuniform, 1167 Demyelinative internal strangulation, 777

Index Denaturing high-performance liquid chromatography (dHPLC), 1557–1559 Dendrites, of autonomic ganglion neurons, 235 Dendritic cells, 574 Dendritic glomeruli, 235 Dendritic nests, 235 Denervation acute, of Schwann cells, 1408–1410 chemical, capsaicin as research model for, 889–890, 889f–890f chronic, 1412, 1414f determination of denervation supersensitivity, 1124 muscle atrophy and, 1142 Schwann cell death and, 1413 single-fiber electromyography in, 991–994 Dense lamellar bodies (DLBs) axonal, 47 of constricted axon segment, 59 of juxtaparanodal segment, 53 Deoxyribonucleic acid (DNA), 1545 analysis in evaluation of pediatric polyneuropathies, 2712 mitochondrial, 1549–1550, 1938–1940, 1939f in antiretroviral toxic neuropathy, 2134 maternal inheritance of mutations of, 1938–1940 penetrance of mutations of, 1939–1940 peripheral nerve disease associated with mutations of, 1937–1947. See also Mitochondrial respiratory chain dysfunction. structure of, 1938, 1939f mutations in repair genes for, 1741 TDP1, 1739–1741 neuronal pathology and defects of repair of, 705 polymerase chain reaction for analysis of, 1553–1554 Southern blot for analysis of, 1554 testing for hereditary motor and sensory neuropathies, 2721 Depolarization effect on excitability abnormalities in multifocal motor neuropathy, 123, 124f in neuromuscular transmission, 832 uremic neuropathy and, 124–125 Dermatitis, atopic, skin biopsy in, 888 Dermatomes, 12–14 distribution of, 12–14, 13f, 14f underlying muscles and, 13–14 Dermatomyositis electromyography in, 944f intravenous immunoglobulin for, 643 vs. pediatric chronic inflammatory demyelinating polyradiculoneuropathy, 2719 Dermis, 870 innervation of, 871, 871f quantitation of, 879–880 papillary, 870 Descriptive studies of peripheral neuropathy, 1181–1182, 1182t Desert hedgehog (Dhh), as marker of Schwann cell development, 343, 345 Desglymidodrine, 2675 Desipramine, for diabetic neuropathy, 1971, 1971t Desmin, 434 Desmopressin (DDAVP), for orthostatic hypotension, 1958t, 2675 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Detrusor muscle, 299 detrusor-sphincter dyssynergia, 313 during micturition, 301 in bladder continence, 2626–2627 treatment for hyperreflexia of, 2663–2664 Dexamethasone, for amyloidosis, 2438 Dextroamphetamine, for orthostatic hypotension, 2676 DFP (diisopropylfluorophosphate), 2515 Dhh (desert hedgehog), as marker of Schwann cell development, 343, 345 DHP (Diabetes Health Profile), 1058 Diabetes Care Profile (DCP), 1058, 1060 Diabetes Control and Complications Trial (DCCT), 1952, 1969 Diabetes Health Profile (DHP), 1058 Diabetes Impact Measurement Scale (DIMS), 1058 Diabetes Quality of Life Measure (DQOL), 1058 Diabetes-39 (D-39), 1058 Diabetes-Specific Quality of Life Scale (DSQOLS), 1058 Diabetic amyotrophy, 1140 Diabetic cervical radiculoplexus neuropathy (DCRPN), 1993, 2303 clinical features of, 2000 diabetic lumbosacral radiculoplexus neuropathy and, 1962, 1995, 2000 historical overview of, 1995 Diabetic lumbosacral radiculoplexus neuropathy (DLRPN), 1962–1964, 1993, 2340t clinical features of, 1962, 1995–1999, 1997t–1998t autonomic symptoms, 1996 pain, 1996 proximal and distal symptoms, 1996 sensory loss, 1996 vs. diabetic polyneuropathy, 1995–1996 weakness, 1996 clinical testing in, 1962–1963, 1963t, 1997t, 2001–2002, 1964t nerve conduction, EMG, and quantitative autonomic testing, 1963, 2001, 2002t quantitative sensation testing, 1996, 2002, 2003t diabetic cervical radiculoplexus neuropathy and, 1962, 1995, 2000 diabetic thoracic radiculoneuropathy and, 2000 differential diagnosis of, 1963–1964, 1964t historical overview of, 1994–1995 long-term morbidity in, 1999 mechanisms of, 1962, 1994–1995 monophasic course of, 1999 onset of, 1962 pathology of, 2004f–2005f, 2004–2009, 2006t, 2007f–2008f inflammation, 2005 ischemic damage, 2004–2005 microvasculitis, 2004–2007, 2007f–2008f, 2405–2406 teased fiber analysis, 2005, 2006t recovery from, 1963, 1999 subtypes of, 1994, 2004 terminology for, 1993 treatment of, 1963, 2011–2013 analgesia, 1963 immunotherapy, 1963, 2011–2012, 2013f symptomatic, 2012–2013 Diabetic neuropathic cachexia, 1954, 1995

Index Diabetic neuropathies, 1137, 1951–1980 autonomic function testing in, 1104 autonomic involvement in, 1956–1960 bladder dysfunction, 2626–2627 cardiovascular system, 1957, 1958 heart rate abnormalities, 1957, 1958 impaired cardiac function, 1958 management of, 1958t, 1960 orthostatic hypotension, 1150, 1957, 1958, 1958t painless myocardial infarction, 1958 gastrointestinal system, 1280, 1281, 1957, 1958–1959, 2696–2698, 2697f–2698f diarrhea, 1959, 1960, 2665 enteric nerve dysfunction and, 2697–2698 epidemiology of, 2696 extrinsic nerve dysfunction and, 2696–2697 humoral mediators of, 2697 management of, 2698 psychological factors and, 2698 vagal efferent activity and gastroparesis, 1167, 1280, 1281, 1958–1959 hypoglycemic unawareness, 1960 natural history of, 1957 pathology of, 1967–1968 prevalence of, 1957 pupillary dysfunction, 1959 respiratory control disturbances, 1959–1960 risks of, 1957–1958 sudomotor dysfunction, 1150, 1959 urogenital dysfunction, 1957, 1959, 2628 blink reflex in, 920t chronic inflammatory demyelinating polyradiculoneuropathy and, 1955–1956, 2229 classification of, 1953t compression neuropathies, 1140, 1964–1965 carpal tunnel syndrome, 1444–1445, 1964 differential diagnosis of, 1964t, 1965 cough responses in, 1284, 1284f cranial nerve involvement in, 1140, 1960–1961 abducens palsy, 1960 facial nerve paralysis, 1225, 1234, 1961 glossopharyngeal nerve, 1274 oculomotor palsy, 804, 1201–1202, 1960–1961 trochlear palsy, 1960 determining cause of, 1169–1170 differential diagnosis of, 1956, 1956t chronic inflammatory demyelinating polyradiculoneuropathy, 2228–2229 cryptogenic sensory polyneuropathy, 2327 epidemiology of, 1187–1189, 1189t, 1952 experimental antioxidants for treatment of, 516t, 516–518, 517f axonal transport in, 396, 397t nerve ischemia in, 512, 513f, 672–674, 673f, 673t excessive susceptibility to, 673–674 reduced levels of antioxidants in, 512, 515, 515t femoral, 1491, 1993 historical recognition of, 1951 hypoglycemic neuropathy, 1965 in children, 2720 macro electromyography in, 993f, 999

Diabetic neuropathies (Continued) management of, 1970–1972 blood glucose control, 1970–1971 capsaicin ointment, 1972 citalopram, 1971 for at-risk diabetic foot, 1972, 1972t for painful neuropathy, 1971t, 1971–1972 gabapentin, 1971, 1971t, 2640, 2644 ␣-lipoic acid, 516, 516t, 1971t, 1972 neurotrophic factors, 380, 381, 383 OpSite surgical dressing, 1972 topiramate, 2640, 2643 tramadol, 2646 tricyclic antidepressants, 1971 venlafaxine, 1971 minimum criteria for diagnosis of, 1952 MNSI questionnaire for, 1040, 1041f motor neuropathy, 1993, 1955–1956 acute, 1955 chronic inflammatory demyelinating polyradiculoneuropathy, 1955 investigation of, 1955 subacute, 1955 treatment of, 1955–1956 natural history of, 1952 nerve biopsy vs. other neuropathic evaluations in, 736 nerve conduction studies in, 937f nerve excitability studies in, 126 onset of, 1956 painful, 870, 1953, 1954–1955 insulin-induced, 1954–1955 treatment of, 1971t, 1971–1972 pathogenesis of, 1968–1970 advanced glycation end products, 1970 animal models of, 1972–1980 changes in nerve blood flow, 1980 conduction velocity deficit, 1978–1979 endoskeletal, axonal transport, and neurotrophin abnormalities, 1977–1978 glucose toxicity, 1973, 1974f mitogen-activated protein kinases, 1976–1977 nonenzymatic glycation, 1974–1976, 1975f oxidative stress, 511–519, 1976 polyol pathway, 1973–1974 protein kinase C, 1977 resistance to ischemic conduction failure, 1979–1980 essential fatty acids, 1970 hyperglycemia, 1952–1953, 1969, 2327 nerve ischemia/hypoxia, 1970 neurotrophic factors, 1970 oxidative stress, 511–519, 1970 polyol pathway, 1969–1970 pathology of, 701, 2006t, 1965–1968, 1968t blood vessel abnormalities, 1966, 1968f connective tissue changes, 1967 epidermal nerve fiber loss, 870, 873f, 884–885, 885f in autonomic neuropathy, 1967–1968 intracytoplasmic vacuolation, 688–689 myelinated fibers, 1965–1966, 1966f, 1967f pattern of involvement in, 1956 prevalence of, 1951, 1952 proximal, 1993 clinical features of, 1386 vs. lumbosacral plexus lesions, 1386 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxv Diabetic neuropathies (Continued) quality-of-life issues in, 1060 quality-of-life measures in, 1058 rehabilitation for, 2629 related to impaired glucose tolerance and hyperglycemia, 1952–1953, 2327 risk factors for, 1952–1953, 1953t screening for, 1172 sensory involvement in, 1139f, 1141, 1993 distal symmetrical sensorimotor polyneuropathy, 1953–1954 skin biopsy in, 884–885, 885f staging severity of, 1952, 1952t temporal profile of, 1167 truncal radiculoneuropathy, 1961–1962 clinical features of, 1961, 1962f differential diagnosis of, 1962 distribution of sensory loss in, 1961, 1961f investigation of, 1961–1962 management of, 1962 skin biopsy in, 885 Diabetic radiculoplexus neuropathy (DRPN), 1993–2013 cervical. See Diabetic cervical radiculoplexus neuropathy (DCRPN). historical overview of, 1994–1995 lumbosacral. See Diabetic lumbosacral radiculoplexus neuropathy (DLRPN). subtypes of, 1993 thoracic. See Diabetic thoracic radiculoneuropathy (DTRN). vs. chronic inflammatory demyelinating polyradiculoneuropathy, 2000 Diabetic retinopathy, 1168 damage to ciliary nerves during photocoagulation treatment of, 204 Diabetic thoracic radiculoneuropathy (DTRN), 1993 clinical features of, 2000 diabetic lumbosacral radiculoplexus neuropathy and, 2000 historical overview of, 1995 Diacylglycerol (DAG), in diabetic neuropathy, 1977 Diagnosis, 1154–1155. See also Nerve tests. confirmation of neuropathy, 1154 establishing etiology, 1154–1155 Diaphragm paralysis, in hereditary motor and sensory neuropathy IIC, 1629, 1733 Diarrhea, 2616, 2616t, 2665, 2692–2694 clinical assessment of, 2692 diagnostic evaluation of, 2694 in diabetic patients, 1959, 1960, 2665 management of, 2694 Diazepam for carbamate neurotoxicity, 2509 for hemifacial spasm, 1245 for neuropathic pain, 2645 Dichloracetylene (DCA), 2520 Didanosine (ddI), 2133–2135, 2134f–2135f, 2563 Dietary factors copper deficiency, 2058 epidemiologic studies of, 1185 in adrenoleukodystrophy, 1868 in management of constipation, 2692 in management of gastroparesis, 2689–2690 in management of orthostatic hypotension, 1958t, 2658 in metachromatic leukodystrophy, 1854 in Strachan’s syndrome, 2057 niacin deficiency, 2057

xxvi Dietary factors (Continued) thiamine deficiency, 2051–2053 vitamin B6 deficiency, 2056 vitamin B6 toxicity, 717, 2056–2057, 2566–2567 vitamin B12 deficiency, 1139–1140, 2053–2055 vitamin E deficiency, 2055–2056 Differential diagnosis, 1152–1154, 1163–1176 approaches to, 1164–1167 class of affected fiber, 1165–1166 electrophysiology, 1166–1167 pathophysiology, 1166, 1166t pattern, 1164–1165, 1165t temporal profile, 1167 associated symptoms and signs of disease, 1167–1170 neurologic disease, 1170, 1170t non-neurologic disease, 1167–1170, 1169t diagnostic yield in assessment of polyneuropathy, 1163–1164 disorders simulating peripheral neuropathy, 1152–1153, 1175–1176 aging, 1176 erythromelalgia, 1175 motor neuron disease, 1176 myelopathy, 1176 myopathy, 1175–1176 painful legs and moving toes, 1175 family history, 1170 laboratory investigations, 1164, 1170–1175 neuropathies simulating other conditions, 1153–1154 Diffuse infiltrative lymphomatosis syndrome (DILS), 2138–2139 Digastric muscle, innervation of, 1220, 1222 Digenic inheritance, 1546 Dihydroergotamine, for orthostatic hypotension, 2660t, 2661, 2678 Dihydrophenylserine (DOPS), for orthostatic hypotension, 2677–2678, 2678f Diisopropylfluorophosphate (DFP), 2515 DILS (diffuse infiltrative lymphomatosis syndrome), 2138–2139 2,3-Dimercaptopropanesulfonate sodium, for arsenic poisoning, 2538 Dimercaptosuccinic acid (DMSA) for lead poisoning, 2534, 2535 for mercury intoxication, 2540 DIMS (Diabetes Impact Measurement Scale), 1058 Diphenoxylate, for diarrhea, 293t Diphtheria and diphtheritic neuropathy, 714, 2147–2150 axonal transport in, 397t clinical features of, 2147–2148 epidemiology of, 2147 laboratory features of, 2148 molecular pathogenesis and pathophysiology of, 2149–2150 motor involvement in, 1141 pathology of, 714, 820, 2148–2149, 2148f–2149f pediatric, 2717–2718, 2719–2720 segmental demyelination in, 3–4 treatment of, 2148 Diphtheria toxin, 2149–2150 Diplopia binocular vertical, 1196 in Guillain-Barré syndrome, 1202 in multiple sclerosis, 1203

Index Disability definition of, 1054 quantitation of, 1041–1048 fatigue, 1047–1048 general scores, 1041–1045 Functional Independence Measure, 1041–1042 Guillain-Barré Syndrome Disability Scale, 1042–1043, 1046t modified Rankin scale, 1042, 1046t Overall Disability Score Scale, 1044–1045 Health-Related Quality of Life and Clinical Outcomes Score in Peripheral Neuropathy, 1045–1046, 1048t morphometry of nerve or nerve fiber counts in skin biopsy, 1046 other outcomes, 1046 symptoms and, 1033 DISC-1, in schizophrenia, 467 Diskitis, vs. pediatric lumbosacral radiculopathy, 2738 Distal symmetrical polyneuropathy (DSP), 1138, 1138f–1140f in HIV infection, 2129–2133, 2135–2136 clinical features of, 2131–2132 disease-stage specificity of, 2129–2130 electrophysiologic findings in, 2132 incidence of, 2130, 2131, 2131f pathogenesis of, 2132–2133, 2133f pathology of, 2132 risk factors for, 2131 treatment of, 2135–2136 Distribution and patterns of involvement, 1137–1140, 1138f–1140f complex distributions, 1139 cranial nerve involvement, 1139–1140 differential diagnosis based on, 1164–1165, 1165t distal symmetrical polyneuropathy, 1138, 1138f–1140f focal and multifocal neuropathies, 1140, 1140t proximal symmetrical and upper limb-predominant polyneuropathy, 1138–1139 Disulfiram neuropathy, 2510–2511, 2558–2559 Dithiocarbamates, 2511 Diving response, 1123 DLBs (dense lamellar bodies) axonal, 47 of constricted axon segment, 59 of juxtaparanodal segment, 53 DLRPN. See Diabetic lumbosacral radiculoplexus neuropathy (DLRPN). DMSA (dimercaptosuccinic acid) for lead poisoning, 2534, 2535 for mercury intoxication, 2540 DNA. See Deoxyribonucleic acid (DNA). Docetaxel neuropathy, 2490, 2490t, 2568–2570 Dogiel types I and II neurons, 253–254, 254f, 255f Domperidone for gastroparesis, 1960, 2665, 2690 for orthostatic hypotension, 2678 Dopamine biosynthesis of, 1117f effect of dopamine-␤-hydroxylase deficiency on plasma level of, 1117, 1118 in micturition reflex, 312 measuring plasma levels of, 1116–1118, 1119f Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Dopamine agonists, for restless legs syndrome, 1145 Dopamine antagonists, for orthostatic hypotension, 2661, 2678 Dopamine-␤-hydroxylase (DBH) axonal transport of, 389, 389f, 394f, 394–395 deficiency of effect on plasma catecholamines, 1117, 1118, 1119f effect on urinary catecholamines and metabolites, 1118 DOPS (dihydrophenylserine), for orthostatic hypotension, 2677–2678, 2678f Dorsal axial line, 32 Dorsal nerve roots, 11, 1323, 1339. See also Spinal nerve roots. morphometry of, 766f Dorsal rami, 11, 12, 16, 16f cervical, 31–32, 1324f coccygeal, 32 lumbar, 25, 32 sacral, 28, 32 thoracic, 23, 32 Dorsal root entry zone, 67 Dorsal root ganglion (DRG), 574, 1295, 2309 age-related changes in, 484–486, 487, 702t, 702–703, 703t lipofuscin, 702–703, 703t nodules of Nageotte, 702, 702t anatomy of, 1298, 1323, 1324f peripherally and centrally directed axonal processes, 1296 intraspinal, 1323 pathology of axotomy-induced, 684–702 biochemical alterations, 689–695 chromatolysis, 684–687 intracytoplasmic vacuolation, 687–689 neuronal death, 695–702 effect on sensory nerve action potentials, 1323, 1324, 1327 in HIV-associated sensory polyneuropathies, 2132, 2133 in Sjögren’s syndrome, 1309–1310, 1309f–1310f somatosensory neurons of, 163 Dorsal root ganglion (DRG) neurons, 163–191 afferent receptive properties of, 165–166 chemical phenotype of, 171–183 adrenoreceptors, 183 cannabinoid receptors, 183 cholinergic receptors, 183 GM1 receptors, 183 immunostaining for, 173, 173f in culture vs. in vitro neurons, 173 inflammatory mediators and effects of tissue damage, 181–183 neuron size and, 172f neurotrophic factors, 173–181, 176t brain-derived neurotrophic factor, 177–178 fibroblast growth factor and receptors, 180 nerve growth factor, 175–177 neuropoietic cytokines and receptors, 180–181 neurotrophin-3, 178 neurotrophin-4, 178 p75 binding of, 178 receptors for glial cell-line-derived neurotrophic factors, 179–180 Trk receptors, 174–179

Index Dorsal root ganglion (DRG) neurons (Continued) nociceptive and low-threshold mechanoreceptor neurons, 191 oligosaccharides, 183 sources of endogenous substances affecting, 174f–175f classes of cutaneous afferent receptive neurons, 166–169, 167t–168t low-threshold mechanoreceptors, 166–169 nociceptors, 169 thermoreceptors, 169 cutaneous, muscle, and visceral afferents of cytochemistry of, 186 receptive properties of, 170 distance spanned by, 2309 electrical membrane properties of, 170f, 170–171 glial, development of, 163, 347–348 in paraneoplastic subacute sensory neuronopathy, 2476 ion currents and channels expressed by, 189–191 calcium, 191 hyperpolarization-activated channels, 191 potassium, 189–191 sodium, 189 large light (neurofilament-rich), 163–165, 164f membrane receptors and sensory transduction mechanisms of, 186–188, 187t acid-sensing molecules, 188–189 chemoreceptors, 188 mechanoreceptors, 188 thermoreceptors, 186–187 nociceptors, 163, 165–166 origin of, 163 overview of, 163–165 primary afferent neurons, 163 size distributions of, 163–165, 164f small dark (neurofilament-poor), 163–165, 164f soma size, conduction velocity, and neurofilament in, 165 soma size in relation to sensory properties of, 165 transmitters, neuromodulators, peptides, and their receptors expressed by, 184–186 calcitonin gene–related peptide, 184–185 excitatory amino acid transmitters, 184 galanin, 185 neuropeptide Y, 185 opioid receptors, 185–186 opioid-like receptors, 186 pituitary adenyl cyclase–activating protein, 185 somatostatin, 185 substance P, 184–185 vasoactive intestinal peptide, 185 Dorsal scapular artery, 1341 Dorsal scapular nerve, 19, 1340f muscles innervated by, 956t Dorsal spinocerebellar tract (DSCT), 158–159 Double-crush syndrome, tarsal tunnel syndrome and, 1504 Down syndrome, carpal tunnel syndrome and, 1448 Doxepin, topical, 2645 Doxorubicin intracytoplasmic vacuolation induced by, 688 selective toxicity for dorsal root ganglion cells, 1296

DQOL (Diabetes Quality of Life Measure), 1058 Dracunculiasis, 2155t, 2170–2171 clinical features of, 2170 complications of, 2170 epidemiology of, 2170 etiologic agent in, 2163f, 2170 life cycle of, 2170 peripheral neuropathy in, 2170–2171 transmission of, 2170 DRASIC, 187t, 188 DRG. See Dorsal root ganglion (DRG). DRPN. See Diabetic radiculoplexus neuropathy (DRPN). Drug-induced neuropathy, 2553–2573 almitrine, 2554 amiodarone, 2554–2555 chloroquine, 2555–2556 colchicine, 2556–2557 dapsone, 2557–2558 disulfiram, 2510–2511, 2558–2559 ethambutol, 2559 in children, 2720 isoniazid, 2559–2560 metronidazole, 2560 misonidazole, 2561 nitrofurantoin, 2561–2562 nitrous oxide, 2562–2563 nucleoside analogues, 2131–2136, 2134f–2135f, 2563 perhexilene, 2563–2564 phenytoin, 2564–2565 platinum compounds, 2544–2545 podophyllin, 2565–2566 pyridoxine, 2566–2567 suramin, 2567–2568 tacrolimus, 2568 taxanes, 2490, 2490t, 2498t, 2568–2570 thalidomide, 2490–2491, 2498t, 2570 tryptophan, 2570–2571 vinca alkaloids, 2572–2573 DSCT (dorsal spinocerebellar tract), 158–159 DSN. See Déjérine-Sottas neuropathy (DSN). DSP. See Distal symmetrical polyneuropathy (DSP). DSQOLS (Diabetes-Specific Quality of Life Scale), 1058 DTRN (diabetic thoracic radiculoneuropathy), 1993 clinical features of, 2000 diabetic lumbosacral radiculoplexus neuropathy and, 2000 historical overview of, 1995 Duane’s syndrome, 1197 Duchenne muscular dystrophy dystrophin gene mutation in, 1545 electromyography in, 954f X-linked recessive inheritance of, 1548 Duodenum, in Chagas’ disease, 2159–2160 Dura, 11, 1324f Dural pouch, 11 DVT (deep venous thrombosis) immobility and, 2629 in intensive care unit patients, 2616 prophylaxis for, 2609, 2629 Dyck Score, 1045–1046, 1048t Dynactin, in familial amyotrophic lateral sclerosis, 1599 Dynamometers, 1096 Dyneins in axonal transport, 388, 390–391, 391f toxic effects on, 394 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxvii Dynorphin, in ganglionic neurons, 242 Dysarthria, 2633 Dysautonomia. See Autonomic dysfunction. Dysesthesias, 1146 HIV-associated sensory neuropathy and, 2131 in congenital carpal tunnel syndrome, 2725 in hyperthyroid neuropathy, 2041 in nonmalignant inflammatory sensory polyganglionopathy, 2311 in uremic neuropathy, 2023 Dysphagia, 2633, 2685–2687 causes of, 2685–2686 clinical assessment of, 2687 evaluation of, 2687 in amyotrophic lateral sclerosis, 1540 in multiple sclerosis, 2700 in myasthenia gravis, 2686, 2700 in Parkinson’s disease, 2695 management of, 2687 palatal paralysis and, 1279 pharyngeal paralysis and, 1279 postpolio, 2695 Dystonia musculorum, 711 Dystonin, role in myelination, 364 Dystroglycan as laminin receptor, in myelination, 424–425 effect of inactivation on node of Ranvier, 546, 549 mice deficient in, 546 signaling in Schwann cells, 364–365 Dystrophia muscularis mouse mutant, 424 Dystrophin, 424 gene mutations, 1545

EAE (experimental autoimmune encephalomyelitis), 611 epitope spreading in, 621 EAN. See Experimental autoimmune neuritis (EAN). Early growth response 1 gene (EGR1, Krox24), 1707 role in myelination, 361 Early growth response 2 gene (EGR2, Krox20), 1707–1714 chromosomal localization of, 1707 disorders associated with mutations of, 358–359, 1707 autosomal recessive hereditary motor and sensory neuropathy, 1774, 2714 clinical features of, 1710t cranial nerve involvement, 1709 myelinopathies, 1708–1711, 1709t function of, 1707–1708 functional consequences of mutations of, 1711–1713 NAB proteins, 1711 R1 inhibitory domain, 1711 zinc finger domain, 1711–1713, 1712f identification of, 1707 in Schwann cell development, 61, 342, 1708 position of disease-associated mutations of, 1708f protein structure of, 1707 role in myelination, 357–360, 1708 transgenic animal studies of, 1575–1576 Egr2-null mice, 1575 transcriptional activation of late myelin genes, 1575–1576 E-cadherin, in myelination, 418f, 419t, 420 Echinococcus granulosis infection. See Hydatidosis.

xxviii Edema contractures and, 2629 endoneurial effect on blood flow, 669–670 in chronic inflammatory demyelinating polyradiculoneuropathy, 2240, 2241, 2243f in POEMS syndrome, 2455, 2457 interstitial, 794 intramyelinic, 489–491, 491f–492f periocular, in American trypanosomiasis, 2156, 2158f Edinger-Westphal nucleus, 203, 206f, 1193 Edrophonium, effect on plasma noradrenaline level, 1117–1118 EDTA (ethylenediaminetetraacetic acid), for lead poisoning, 2534–2535 EFF (endoneurial fluid flow), 655f, 655–656 driving force for proximodistal flow, 656 in entrapment neuropathies, 656–657 EGR1. See Early growth response 1 gene (EGR1, Krox24). EGR2. See Early growth response 2 gene (EGR2, Krox20). Ehlers-Danlos syndrome, 2736 EHP. See Endoneurial hydrostatic pressure (EHP). Ejaculation, 316–317 retrograde, 1151 Ekbom syndrome, 1168 Elastic stockings, for orthostatic hypotension, 2658–2659, 2674 Elastin, 37 Elaunin, 37 Electrical injuries, of brachial plexus, 1359 Electrical stimulation, 2623–2624 diagnostic, 113 for quantitative sensation testing, 1065 functional, 2623, 2624 neuromuscular, for muscle strengthening, 2623–2624 submaximal, to determine ease of axon excitability, 113–126. See also Nerve excitability studies. to promote nerve regeneration after injury, 1410, 1410f transcutaneous electrical nerve stimulation, 2624 Electrocochleography, 1256 Electrolyte abnormalities in intensive care unit patients, 2616 Electromyography (EMG), 113, 937–958, 1166 age-related effects on, 1176 anal sphincter, 288–289, 290t discharge patterns of motor units in, 952–955 in lower and upper motor neuron disorders, 955 in myopathy, 955 involuntary movement, 955 measurements of turns and amplitude, 953–954 recruitment, 952–953, 953f–954f effect on serum creatine kinase level, 939 electrodes for, 986f hemostasis for, 939 in amiodarone neuropathy, 2555 in amyloidosis, 2431–2432 in amyotrophic lateral sclerosis, 944f, 952f, 954f in arsenic poisoning, 2537 in axillary neuropathy, 1470 in Bell’s palsy, 1226, 1227

Index Electromyography (EMG) (Continued) in brachial plexopathy, 1351–1352 hereditary, 1761 immune, 2302 obstetric, 1526, 1527t, 2731 in carpal tunnel syndrome, 1442, 1449, 2040 in chloroquine neuropathy, 2556 in chronic inflammatory demyelinating polyradiculoneuropathy, 2234, 2457–2458 pediatric, 2718 in coagulopathies, 939 in colchicine neuropathy, 2557 in common peroneal neuropathy, 1498 in congenital hypomyelinating neuropathy, 2714 in congenital myasthenic syndromes caused by rapsyn deficiency, 856 choline acetyltransferase deficiency, 839–840 end-plate acetylcholinesterase deficiency, 842–843, 844f resembling Lambert-Eaton myasthenic syndrome, 842 slow-channel syndromes, 848, 848f sodium channel myasthenia, 857 in critical illness myopathy, 2030 in critical illness polyneuropathy, 2027 in cryptogenic sensory polyneuropathy, 2323 in dapsone neuropathy, 2558 in Déjérine-Sottas disease, 2714 in dermatomyositis, 944f in diabetic lumbosacral radiculoplexus neuropathy, 1963, 1963t, 2001, 2002t in diabetic neuropathy, vs. nerve biopsy, 736 in disulfiram neuropathy, 2558 in Duchenne muscular dystrophy, 954f in evaluation of pediatric mononeuropathies, 2725 in evaluation of pediatric polyneuropathies, 2711–2712 in femoral neuropathy, 1490–1491 in gluteal neuropathy, 1493 in Guillain-Barré syndrome acute inflammatory demyelinating polyneuropathy, 2200 axonal forms, 2201 childhood, 2716 infantile, 2712–2713 in hemifacial spasm, 1241–1242, 1243f in hepatitis C–associated neuropathy, 2019 in hereditary motor and sensory neuropathy type IA, 1662 type II, 1718, 1722 in hereditary motor neuronopathy, 1605 in hereditary sensory and autonomic neuropathy type II, 1828 in hexacarbon neuropathy, 2514 in Hopkins syndrome, 2031 in infantile botulism, 2713 in Kennedy’s disease, 1612 in Lambert-Eaton myasthenic syndrome, 837 in lesion localization, 955–958 innervation of cranial, shoulder girdle, and upper limb muscles, 956t innervation of hip girdle and lower limb muscles, 957t in limb-girdle dystrophy, 954f in long thoracic neuropathy, 1465 in lower motor neuron disorders, 939f in lumbosacral plexus lesions, 1377–1378 in monoclonal gammopathy of undetermined significance, 2265–2266 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Electromyography (EMG) (Continued) in multifocal motor neuropathy and conduction block, 2284 in multiple sclerosis, 947f in myopathy, 939f, 943, 951–952, 954f, 955, 1176 in myotonia, 939f in myotonic dystrophy, 942f in neuromyotonia, 1144, 1144f in nitrofurantoin neuropathy, 2562 in nitrous oxide neuropathy, 2562 in nonmalignant inflammatory sensory polyganglionopathy, 2312–2313 in obturator neuropathy, 1489 in organophosphate-induced delayed peripheral neuropathy, 2518 in painful legs and moving toes syndrome, 1145, 1175 in paraneoplastic subacute sensory neuronopathy, 2475–2476 in patients receiving anticoagulants, 938 in phenytoin neuropathy, 2565 in podophyllin neuropathy, 2566 in POEMS syndrome, 2457–2458 in polymyositis, 939f in porphyrias, 1890 in primary biliary cirrhosis–associated sensory polyneuropathy, 2020 in radial neuropathy, 953f, 1477 in children, 2727 in radiculopathy, 1324–1325 pediatric, 2737–2738 polyradiculopathy, 1327 in sciatic neuropathy, 1495 in suprascapular neuropathy, 1468 in tarsal tunnel syndrome, 1505 in taxane neuropathy, 2569 in tibial neuropathy, 1502 in tick paralysis, 2717 in ulnar neuropathy, 1482–1483 in upper motor neuron disorders, 939f in uremic neuropathy, 2023–2024 in vasculitic neuropathy, 2367 in vinca alkaloid neuropathy, 2572 macro, 985, 991, 997–999 electrode for, 986f, 991 in alcoholic neuropathy, 999 in amyotrophic lateral sclerosis, 992f–993f, 996 in diabetic neuropathy, 993f, 999 in Guillain-Barré syndrome, 993f, 997–998, 998f in hereditary motor and sensory neuropathy, 993f, 998–999 in limb-girdle muscular dystrophy, 992f in patients with history of polio, 993f, 997 normal values for, 991 recording principle for, 991, 991f motor unit potentials in, 947–952 abnormalities of, 949–951 amplitude and area of, 948 conditions associated with types of, 947–948, 948t duration of, 949, 950t inn lower motor neuron vs. myopathic disorders, 951–952, 953f–954f jiggle of, 949 phases of, 949, 951f–952f physiologic factors affecting, 947 profile of, 948f rise time of, 948–949 of voiding responses in infant, paraplegic patient, and adult, 300f, 301

Index Electromyography (EMG) (Continued) potentials associated with needle insertion for, 939–941 end-plate activities, 940, 941f injury potentials, 939–940, 940t myotonic discharge, 940–941, 942f, 942t scanning, 985, 990f, 990–991 applications of, 985, 991 concentric needle electrode for, 986f, 990 in amyotrophic lateral sclerosis, 996f, 996–997 in primary myopathies, 990, 999 in reinnervation, 991 recording principle for, 990, 990f sequence of abnormalities in, 958 single-fiber, 985–997, 986f–988f applications of, 985 during axonal stimulation in myasthenia gravis and Lambert-Eaton myasthenic syndrome, 995–996 during intramuscular electrical stimulation, 988f, 988–989 during voluntary activation, 985–987 electrode for, 985, 986f in amyotrophic lateral sclerosis, 993f, 996 in denervation and reinnervation, 991–994, 993f in myasthenia gravis, 987, 993f, 994–995, 995f in syringomyelia, 997 multielectrode studies, 989–990 signs of abnormal nerve excitability on, 994 signs of abnormal transmission in nerve branches on, 993f, 994 studies of spinal events with, 989 steps of, 938f, 939 to assess muscle strength, 1097 types of spontaneous activity in, 941–947, 943t, 944f–945f complex repetitive discharges, 941, 943, 946f continuous muscle fiber activity, 946 cramp, 946–947, 1143 fasciculation potentials, 941–942, 945–946 fibrillation potentials, 941, 942 grading of, 942 myokymic discharges, 942, 946, 947f, 1142–1143 positive sharp waves, 941, 942–943 single-fiber discharge and denervation, 943 typical findings on, 939f Electron microscopy early studies, 3 of teased fibers, 753–754, 755f Electroneurography, in Bell’s palsy, 1226, 1227 Electronystagmography, 1256 Electrophoresis immunofixation, 2181 to identify monoclonal proteins, 2256–2258, 2256f–2258f Elephantiasis neuromatosa, 2591 ELISA (enzyme-linked immunosorbent assay), 2178–2180 procedure for, 2178–2179 sensitivity and specificity of, 2179–2180 Elsberg syndrome, 1331 Elzholz bodies, in Schwann cells, 771, 775f, 790 Emepromium chloride, for detrusor hyperreflexia, 2663 Emery-Dreifuss muscular dystrophy, 1741–1742 vs. scapuloperoneal hereditary motor neuronopathy, 1616

EMG. See Electromyography (EMG). Emotional lability, in amyotrophic lateral sclerosis, 1536, 1539 EMS (eosinophilia-myalgia syndrome), 2570–2571 Encephalitis Bickerstaff’s brainstem, 2185, 2205 cytomegalovirus, 2121 Epstein-Barr virus, 2120 herpes simplex virus, 2117 West Nile virus, 1301–1302 Encephalomyelitis Epstein-Barr virus, 2120 experimental autoimmune, 611, 621 paraneoplastic, 1200 Encephalomyelocele, 1333 Encephalomyelopathy mitochondrial. See Mitochondrial respiratory chain dysfunction. subacute necrotizing, 1941, 1944–1945, 1945f, 2715 Encephalopathy, lead, 2528 Endocrinopathies, in POEMS syndrome, 2458 Endoneurial fluid, 36, 39, 651, 655–657 continuity between cerebrospinal fluid and, 655 turnover of, 656 Endoneurial fluid flow (EFF), 655f, 655–656 driving force for proximodistal flow, 656 in entrapment neuropathies, 656–657 Endoneurial hydrostatic pressure (EHP), 655 developmental increase of, 656 effect of endoneurial protein concentration on, 656 in entrapment neuropathies, 657, 1396–1397 in lead neuropathy, 658, 658f in nerve degeneration and regeneration, 658–659, 659f Endoneurium, 36–39, 576 anatomy of, 35, 36–39, 36f antigen presentation of, 574 as regulated physiologic space, 651 capillary permeability of, 652–654 collagen of, 37 debris removal from, 791–792, 792f edema of effect on blood flow, 669–670, 672 in chronic inflammatory demyelinating polyradiculoneuropathy, 2240, 2241, 2243f fibroblasts of, 37 homeostasis of, 657–660 controlled variables of, 657 during development, 659–660 in lead neuropathy, 657–658, 658f in nerve degeneration and regeneration, 658–659, 658f–659f in compression neuropathies, 1396–1399, 1397f–1398f in diabetic neuropathy, 1966, 1967, 1968f macrophages of, 559–560 pericytes of, 576 Renaut bodies in, 38–39, 39f transporters of, 653–654 vascular bed of, 574, 652 Endothelial dysfunction, diabetic, oxidative stress and, 518, 518f Endothelin effect on nerve blood flow, 670, 670t in Schwann cell development, 60, 342, 346, 353–354 receptors for, 183 Endotracheal intubation, 2613–2614 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxix End-plate activities associated with electromyographic needle insertion, 940, 941f End-plate potential (EPP) amplitude of, 832–833 in congenital myasthenic syndromes associated with plectin deficiency, 859 choline acetyltransferase deficiency, 840, 841f end-plate acetylcholinesterase deficiency, 843 slow-channel syndromes, 846 sodium channel myasthenia, 858 in myasthenia gravis, 833–834 miniature, 832 in choline acetyltransferase deficiency, 840 in myasthenia gravis, 833 Enemas, for fecal incontinence, 293t Enkephalins effect on transmission in bladder ganglia, 304 expression by dorsal root ganglion neurons, 185 in ganglionic neurons, 236, 240, 242 actions of, 243 in micturition reflex, 312 Enoxaparin, for deep venous thrombosis prophylaxis, 2609 ENS. See Enteric nervous system (ENS). Enteral feeding, 2610 Enteric ganglia, 252f, 252–253 Enteric nervous system (ENS), 15, 235, 249–272, 2683 age-related changes in, 483, 494, 495f anatomy of, 18 conceptual model of, 251, 251f extrinsic innervation of gastrointestinal tract, 2683–2685, 2684f–2686f glial cells of, 341 histoanatomy of, 252f, 252–253 ganglia, 253 myenteric plexus, 252–253 submucosal plexus, 252–253 historical perspective on understanding of, 250f, 250–251 in disinhibitory motor disorders, 259t, 259–261 chronic intestinal pseudo-obstruction, 259–260 sphincteric achalasia, 260–261 in functional gastrointestinal disorders, 249, 251, 271–272 integrative control of gut systems by, 249–250 interneurons of, 251, 251f, 253, 256–257 mechanoreceptors in, 254 motor neurons of, 251, 251f, 253, 257–259 excitatory musculomotor neurons, 258 inhibitory musculomotor neurons, 258–259 secretomotor neurons, 257f, 257–258 neuronal electrical behavior in, 261–263, 262f action potentials, 262–263 resting membrane potentials, 262 neurons of, 249, 251, 251f, 253 Dogiel types I and II, 253–254, 254f, 255f properties mimicking those of central nervous system, 253 sensory neurons of, 251, 251f, 253, 254–256 chemoreception, 255–256 mechanoreception, 256 multimodal properties of, 254 visceral hypersensitivity in irritable bowel syndrome, 254–255

xxx Enteric nervous system (ENS) (Continued) synaptic transmission in, 263–271 fast excitatory postsynaptic potentials, 264f, 264–265 presynaptic facilitation, 270 presynaptic inhibition, 270–271 slow excitatory postsynaptic potentials, 264f, 265–269 slow inhibitory postsynaptic potentials, 269–270 Enteric plexus, 18 Enteroviruses, amyotrophic lateral sclerosis and, 1535 Entrapment neuropathies. See Compression neuropathies. Environmental modifications, 2625 Enzyme activity after axotomy, 693 Enzyme-linked immunosorbent assay (ELISA), 2178–2180 procedure for, 2178–2179 sensitivity and specificity of, 2179–2180 EO (ethylene oxide) neurotoxicity, 2511 effect on axonal transport, 397t Eosinophilia-myalgia syndrome (EMS), 2570–2571 Eosinophilic meningoencephalitis, 2164 Eosinophilic radiculomyeloencephalitis, 2164 Ephaptic transmission, 994, 1150 Ephedrine for carotid sinus hypersensitivity, 1277 for orthostatic hypotension, 2660t, 2660–2661, 2676 Ephexin, regulation of Rho GTPases by, 454 Ephrin/Eph, 453–456 biologic functions of, 453 forward signaling by, 453–454, 454f Abl and Arg kinase, 454–455 FAK and Cas, 455 inactivation of MAPK pathway, 455 regulation of Rho GTPases by ephexin, 454 modulation of signaling by, 456 receptor complexes of, 453 family of, 453 structure of, 453 reverse signaling by, 454f, 455–456 GPI-linked A-type ephrins, 456 phosphorylated ephrin-B and Grb4, 455–456 phosphorylation by tyrosine kinases, 455 phosphorylation-independent interactions, 456 Epidemiologic studies, 1181–1189 clinical applications of, 1184–1189 methodology of, 1181–1184 analytic studies, 1182–1184, 1183f case-control studies, 1183, 1183f cohort studies, 1183, 1183f cross-sectional studies, 1181 descriptive studies, 1181–1182, 1182t genetic studies, 1184 longitudinal studies, 1181 of Charcot-Marie-Tooth disease, 1185–1186, 1186t of diabetic neuropathy, 1187–1189, 1189t of familial clusters, 1184 of Guillain-Barré syndrome, 1186–1187, 1188t of nutrient-deficient diets, 1185 of postherpetic neuralgia, 1186 of tri-o-cresyl phosphate intoxication, 1185

Index Epidermal nerve fibers, 871–872. See also Cutaneous nerves. density of by age, 879, 879f by location, 880t correlation with sural nerve morphometry, 883 correlation with tests of unmyelinated nerve function, 882–883 in chronic inflammatory demyelinating polyradiculoneuropathy, 887 in diabetic neuropathy, 870, 873f, 884–885, 885f in Fabry’s disease, 887 in giant axonal neuropathy, 888 in HIV-associated sensory neuropathy, 870, 885 in painful sensory neuropathies, 883–884, 884f in postherpetic neuralgia, 870, 887 in postural tachycardia syndrome, 887–888 in sensory ganglionopathies, 887 normal, 873f, 879, 880t loss of, 870 neuropathic pain and, 870 morphologic changes in sensory neuropathy, 882 neurotrophic requirements of, 872 quantitation of, 876–879 by tracing, 878, 878f in skin blisters and epidermal sheets, 879 manual counting, 878–879 rules for, 877f, 877–878 remodeling of, 871–872 Epidermal sheet preparations, 876 quantification of epidermal nerves in, 879 Epidermis, 870 innervation of, 870, 871f Epidermolysis bullosa simplex, 859 Epilepsy autonomic, 2695 axon guidance in, 465–466 glycosaminoglycans, proteoglycans, and N-cadherin, 466 mossy fiber sprouting, 466 semaphorin signaling, 466 Epinephrine biosynthesis of, 1117f Horner’s syndrome and, 208 in sweat gland innervation, 1106 measuring plasma levels of, 1116–1118, 1119f measuring urinary levels of, 1118 Epineurial arteries, 801 Epineurium, 35, 41–42, 654 adipose tissue of, 42 anatomy of, 36f, 41–42 capillaries of, 42 collagen of, 37, 42 epifascicular, 41 function of, 654 interfascicular, 41 Epithelioid malignant peripheral nerve sheath tumor, 2597–2598 EPP. See End-plate potential (EPP). EPSPs. See Excitatory postsynaptic potentials (EPSPs). Epstein-Barr virus (EBV) infection, 2120–2121 latent, 2121 neurologic complications of, 2120 neuropathies associated with, 2118t, 2121 pathology of, 2121 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Epstein-Barr virus (EBV) infection (Continued) polyradiculopathy due to, 1331 treatment of, 2121 virology of, 2121 ErbB3, in glial development from neural crest cells, 347 Erb’s palsy, 2632, 2730–2733, 2732f. See also Obstetric brachial plexus palsy (OBPP). Erb’s point, nerve conduction studies at median nerve, 909, 909f radial nerve, 911, 915f ulnar nerve, 913f Erectile dysfunction, 1151, 2628, 2664 in diabetic patients, 1957, 1959, 2628 Ergotamine, for orthostatic hypotension, 2660t, 2661 Erythema migrans, 2109, 2110–2111, 2111f. See also Lyme disease. Erythema nodosum in sarcoidosis, 2420 leprosum, 2086–2087 Erythromelalgia, 1150, 1175, 2497–2498 sensitization of C fibers in, 1009–1010, 1010f Erythromycin for diarrhea, 2665 for gastroparesis, 1960, 2665, 2690–2691 Erythropoietin for orthostatic hypotension, 2660t, 2661, 2676–2677 in Schwann cell proliferation, 1409 E-selectin, 2342 Esophageal motility testing, 1280 Esophageal plexus, 25 Esophagus, in Chagas’ disease, 2159 Ethambutol neuropathy, 2559 Ethanol intracytoplasmic vacuolation induced by, 688 macro electromyography in alcoholic neuropathy, 999 thiamine deficiency and, 2051 Ethionamide, for leprosy, 2103 Ethylene oxide (EO) neurotoxicity, 2511 effect on axonal transport, 397t Ethylenediaminetetraacetic acid (EDTA), for lead poisoning, 2534–2535 Ethylmercury, 2539 Etiologic diagnosis, 1154–1155 Etoposide (VP-16) for vasculitic neuropathy, 2381 peripheral neuropathy induced by, 2490, 2490t, 2498t EURODIAB IDDM Complications Study, 1957 Evacuation proctography, 290t, 291 Evening primrose oil, for diabetic neuropathy, 1970 Excitatory amino acids, 527 Excitatory postsynaptic potentials (EPSPs) fast ganglionic transmission and, 244 in enteric nervous system, 264f, 264–265 slow effects on ganglionic transmission, 243 in enteric nervous system, 264f, 265–269 ionic mechanisms of, 243 muscarinic vs. peptidergic, 242 Excitotoxic neuron death, 699, 701 Executive dysfunction, in amyotrophic lateral sclerosis, 1537

Index Exercise autonomic responses to, 328–331 baroreflex resetting, 328–329, 329f central command, 328 ␣-receptor subtypes and blood flow in active muscles, 331 sensory feedback from active muscles, 328, 329f sympathetic vasoconstriction and metabolic vasodilation, 330, 330f in motor neuron disease, 2633 therapeutic, 2622–2623 Experimental autoimmune encephalomyelitis (EAE), 611 epitope spreading in, 621 Experimental autoimmune neuritis (EAN), 574, 609–628 actively induced, 611–620 adhesion molecules in, 615–617, 616f, 623, 623t characteristics of, 612–614 chemokines in, 617 cytokines in, 617–618, 618f–619f effector mechanisms of myelin destruction in, 618–620, 620f electrophysiology of, 614f, 614–615, 627f mast cells in, 620 matrix metalloproteinases in, 617, 617f, 623 mode of immunization and antigens in, 593–594, 611t, 611–612, 613f P0-induced, 593 P2-induced, 574, 593–594 pathology of, 615, 615f PMP22-induced, 593 polymorphonuclear leukocytes in, 620 scoring progression of, 612 adoptive transfer, 620–621 animal models of, 611t, 611–612 chronic relapsing, 621 galactocerebroside-induced, 593, 621, 622f Guillain-Barré syndrome and, 2199 pathology of, 2199, 2242f segmental demyelination in, 4, 619 treatments for, 623t, 623–628 adhesion molecules and matrix metalloproteinases, 625 antigen presentation and co-stimulation, 623–625 future of, 628 macrophage- and antibody-directed, 626–627, 627f related to phases of immune response, 623, 624f survival factors in peripheral nervous system, 627–628 T-cell-directed, 625–626, 626f termination of immune response, 627, 628f, 635 External anal sphincter, 279, 284, 2627 during defecation, 284 during micturition, 284 electromyography of, 288–289, 290t inadequate relaxation of, 286 innervation of, 279, 282, 2627 muscle fibers of, 284 nerve to, 28 squeeze response of, 284, 285 tonic activity of, 284 External urethral sphincter, 304 Extracellular matrix age-related changes in, 427–428 in myelination, 424

Extrafusal fibers, 132 vs. intrafusal fibers, 137, 138 Extraocular muscles in acute inflammatory demyelinating polyneuropathy, 2200 in third nerve palsy, 1195 innervation of, 1191–1192 aberrant, 1197 Exuberant callus syndrome, 1366 Eye. See also Pupil(s). corneal ulceration and congenital trigeminal anesthesia, 1211 in amebic keratitis, 2162 in angiostrongyliasis, 2164 in coenuriasis, 2169 in Fabry’s disease, 1896 in lepromatous leprosy, 2086 in Miller Fisher syndrome, 1202–1203 in onchocerciasis, 2171 in Refsum’s disease, 1168, 1869 in sarcoidosis, 2418 in Tangier disease, 1908 in toxoplasmosis, 2173 intensive care unit care for, 2609 ocular palsy, 1192, 1195–1203 pupil of, 203–213

F responses, 989 F waves, 921–926 central latency of, 922–923, 923f clinical value and limitations of, 923 comparison between two nerves in same limb, 925t conduction velocity of, 923 normal values for, 924t, 926 recording procedures for, 921 Fabrazyme, 2721 Fabry’s disease, 1893–1901, 2330t carpal tunnel syndrome and, 1447 clinical features of, 1893–1894 diagnosis of, 1900 prenatal, 1900 genetics of, 1548, 1894 histochemistry and ultrastructural changes in, 1846, 1896–1898, 1899f enzymatic defect, 1893, 1894, 1898–1900 nature of accumulating lipid, 1898 source of accumulating glycolipid, 1900 historical background of, 1893 in children, 2721 molecular defects in, 1900 pathology and organ system involvement in, 709, 772, 777f, 1894–1896 acute neuronal degeneration, 818 cardiovascular and pulmonary systems, 1894 central and peripheral nervous systems, 1894–1896, 1895f, 1896t, 1897f–1898f eye, 1896 gastrointestinal system, 1896, 1900 kidney, 1893–1894, 1900 liver, 1896 reticuloendothelial system, 1896 skeletal muscle, 1896 skin, 1168, 1893, 1894, 1899f pattern of inheritance of, 1894 skin biopsy in, 887 small fiber sensory loss in, 1147, 1896 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxxi Fabry’s disease (Continued) treatment of, 1900–1901 enzyme replacement therapy, 1893, 1900–1901, 1901f–1902f gene therapy, 1901 substrate depletion therapy, 1901 Face validity of health outcome measures, 1057 Facial canal, 1219–1220 labyrinthine segment of, 1220 mastoid segment of, 1220 neuroimaging of, 1221–1222, 1221f–1222f tympanic segment of, 1220 vascular supply of, 1220–1221 Facial expression muscles, innervation of, 1222, 1223 Facial nerve, 1219–1246 acute paralysis of, 1223–1236 idiopathic, 1223–1231. See also Bell’s palsy. in amyloidosis, 1232 in benign intracranial hypertension, 1235 in connective tissue diseases, 1234 in diabetes mellitus, 1234 in granulomatous disease, 1232 in HIV infection, 1234 in pregnancy, 1234–1235 miscellaneous causes of, 1235–1236 Ramsay Hunt syndrome, 1231–1232 traumatic, 1232–1233, 1233f tumor-related, 1235 blink reflex to test integrity of, 918–919, 920t, 921t blood supply of, 1220–1221 congenital/pediatric paralysis of, 1197, 1236–1238 in cardiofacial syndrome, 1237 in leukemia, 1237 in Melkersson-Rosenthal syndrome, 1237–1238 infectious, 1236–1237 traumatic, 1237 course of, 1194f, 1219–1220 functional anatomy of, 1222, 1223f–1224f gross anatomy of, 1219–1220 in angiostrongyliasis, 2164 in geniculate neuralgia, 1246 in Guillain-Barré syndrome, 1203 in children, 2716 in hemifacial spasm, 1238–1245 in internal auditory canal, 1254, 1255f in leprosy, 2089f, 2092–2093, 2095f in sarcoidosis, 1203, 2417–2418 in trichinosis, 2164–2165 inherited/familial palsy of, 1236 motor root of, 1219–1220 nerve conduction studies of, 908–909, 909t neuroimaging of, 1221–1222, 1221f–1222f sensory root of, 1219–1220 terminal branches of, 1220 tumors of, 1235 imaging of, 1221–1222, 1221f–1222f, 1235 vulnerability to pathologic processes, 1219 FAHI (Functional Assessment of Human Immunodeficiency Virus Infection), 1059 Fainting, 326 FAK (focal adhesion kinase), 425 as target of Eph receptors, 455 Falls, 483 False aneurysm, nerve injury from, 1518 Famciclovir, for herpes zoster, 2120

xxxii Familial amyloid polyneuropathy (FAP), 1921–1932, 2330t, 2440–2444 age at onset of, 2440 apolipoprotein A-I–related, 1921, 1922, 1931–1932 classification of, 1921, 1922, 2428t–2429t, 2440t clinical features of, 2440–2441 cranial nerve involvement in, 2440–2441 diagnosis of, 2443 electrocardiographic findings in, 2441 familial amyloid polyneuropathy, 2440–2441 gelsolin-related, 1921, 1922, 1932 historical overview of, 1921–1922, 2440 inheritance of, 2440 pathology of, 2443–2444 therapy for, 2444 transthyretin-related, 1921, 1922–1931 central nervous system involvement in, 1924 clinical course of, 1924–1925 clinical features of, 1923–1924, 2441 diagnosis of, 1928–1929, 1929f differential diagnosis of, 1928 epidemiology of, 1924, 1927 genetics of, 1925–1927, 2442–2443 other system involvement in, 1924 pathogenesis of, 1927–1928, 2442–2443 pathology of, 1925, 1926f, 2443–2444 prognosis for, 1931 treatment of, 1929–1931, 2444 Familial dysautonomia, 1831–1833, 2722 autonomic dysfunction in, 1141, 1831 biochemistry of, 1833 clinical features of, 886, 1831–1832 genetics of, 886, 1627t, 1832 inheritance of, 1832–1833 nerve conduction studies in, 1832 pathology of, 1832 skin biopsy in, 886 sweating in, 1150 Familial hypoalphalipoproteinemia (FHA), 1905, 1908 cholesterol homeostasis in, 1912 Familial infantile myasthenia, 839 Familial Mediterranean fever (FMF), 2428t colchicine for, 2438 inheritance of, 2440 Family history, 1170, 2708 FAP. See Familial amyloid polyneuropathy (FAP). Farber’s disease, 1859 Fas, 565–566 Fasciculation potentials, 941–942, 945–946 in amyotrophic lateral sclerosis, 996 Fasciculations, 1142 in amyotrophic lateral sclerosis, 1536 in motor units, 994 in peripheral nerve hyperexcitability syndromes, 1143–1144 Fascioscapulohumeral hereditary motor neuronopathy, 1604t, 1616–1617 Fast excitatory postsynaptic potentials in enteric nervous system, 264f, 264–265 Fast-channel congenital myasthenic syndromes, 847t, 851–852 caused by selective abnormality of gating, 852 caused by unstable kinetics, 852 clinical features of, 851 end-plate studies in, 851, 851f low-affinity syndromes, 852 molecular pathogenesis of, 851f, 851–852 therapy for, 860

Index Fatigue, scoring severity of, 1047–1048 Fatty acids in diabetic neuropathy, 1970 lipid peroxidation from, 510, 510f very-long-chain age-related increase in, 497–498 in adrenoleukodystrophy, 1867–1868 Fazio-Londe disease, 1604t, 1605t, 1617 Fc receptors antibody binding to, 564 glucocorticoid downregulation of, 636 intravenous immunoglobulin blockade of, 643 on macrophages, 560 Fecal continence, 284–286, 2627 external anal sphincter control of, 284, 2627 mechanisms of, 285–286 puborectalis disruption and, 284 Fecal incontinence, 287–292, 2627–2628, 2665, 2694 after spinal cord injury, 2699 clinical assessment of, 2694 clinical features of, 288 denervation and, 287–288 diagnostic evaluation of, 288–292, 290t, 2694 disorders associated with, 2627, 2694 in multiple sclerosis, 2700 management of, 292–293, 293t, 2628, 2694 Felty’s syndrome, 2357 Femoral artery, 801 Femoral cutaneous nerve lateral, 16f, 26, 1376 lesions of, 1492–1493. See also Meralgia paresthetica. in children, 2730 nerve conduction studies of, 918 posterior, 16f, 28 lesions of, 1493 structures innervated by, 1376 Femoral nerve, 1375–1376 compression of, 1376f, 1376–1377 by hemorrhage, 1378 obstetric causes of, 1381 distribution of, 16f, 26, 1489, 1489f involvement in vasculitic neuropathy, 2365t lesions of, vs. lumbosacral plexus lesion, 1377 muscles innervated by, 957t, 1489 nerve conduction studies of, 917, 919t Femoral neuropathy, 1489–1491 clinical manifestations of, 1490 diabetic, 1491, 1993 diagnosis of, 1490–1491 etiology of, 1489–1490 at inguinal ligament, 1490 in pelvis, 1490 total hip replacement arthroplasty, 1512, 1519–1520 in children, 2730 prognosis for, 1491 treatment of, 1491 Fentanyl, for neuropathic pain, 2645 FES (functional electrical stimulation), 2623, 2624. See also Electrical stimulation. FGE (formyl glycin–generating enzyme), in multiple sulfatase deficiency, 1855 FGF. See Fibroblast growth factor (FGF). FGFR (fibroblast growth factor receptor), 176t in CRASH syndrome, 468 in Kallman’s syndrome, 469 FHA (familial hypoalphalipoproteinemia), 1905, 1908 cholesterol homeostasis in, 1912 Fiber, dietary, 2692, 2693t Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Fibrillation potentials, 941, 942, 944f Fibrillations, 1142 Fibroblast growth factor (FGF), 180 expression after axotomy, 180, 694 expression by dorsal root ganglion neurons, 180 family of, 180 in Schwann cell development, 344f, 354, 356 Fibroblast growth factor receptor (FGFR), 176t in CRASH syndrome, 468 in Kallman’s syndrome, 469 Fibroblasts endoneurial, 37 in Remak fiber maturation, 65 in Tangier disease, 1909, 1912 perineurial, 39 synthesis of nerve growth factor by, 379 Fibronectin, 377 Fibular tunnel, 1496 Field units, 166 Fight or flight response, 331 Filopodium(a), 448, 448f Firearm recoil palsy, 1359 Fish eye disease, 1909, 1910t Fisher syndrome (FS), 575, 2178t, 2630, 2199 blink reflex in, 920t cranial nerve involvement in, 1139 IgG binding to GQ1b ganglioside in, 596, 1173, 2178t, 2185 antibody testing for, 2185 disease pathogenesis and, 2185 infections associated with, 2200t Campylobacter, 2206, 2210 intravenous immunoglobulin for, 643 ocular effects of, 1202–1203 pathogenesis of, 2210–2211 anti-GQ1b antibodies, 2210 experimental studies, 2210–2211 inciting events and immune response, 2210 pathology of, 2200t, 2205 FK506, 638–639 adverse effects of, 639 as protection against antiretroviral toxic neuropathy, 2136, 2563 mechanism of action of, 638–639 neuropathy induced by, 2568 to increase axonal regrowth after chronic axotomy, 1411, 1411f use in neuropathy, 639 FK506 binding proteins (FKBPs), 638, 639 FKBPs (FK506 binding proteins), 638, 639 “Flail arm” syndrome, 1299 Flecainide, for neuropathic pain, 2643 Flexibility, 2622 “Floppy babies,” 2712–2713 Fludarabine, for monoclonal gammopathy and peripheral neuropathy, 2269, 2269t Fludrocortisone for carotid sinus hypersensitivity, 1277 for orthostatic hypotension, 1958t, 1960, 2660, 2660t, 2675 Flufenamic acid, for transthyretin amyloidosis, 2444 Fluid management for orthostatic hypotension, 2658, 2674–2675, 2675t in intensive care unit, 2610 Fluorescence resonance energy transfer (FRET), 1555 Fluorescent in situ hybridization, 1551, 1552f Fluoride-resistant acid phosphatase (FRAP), 693

Index Fluoxetine, for pain, 2642 Flurbiprofen, for orthostatic hypotension, 2677 Fluvoxamine, for pain, 2642 FMF. See Familial Mediterranean fever (FMF). Focal adhesion kinase (FAK), 425 as target of Eph receptors, 455 Fodrin, of axoplasmic cortex, 47 Folic acid deficiency, 2073 Foot deformities, 1151, 1168, 1168f in hereditary motor and sensory neuropathy, 1624, 1634, 1634f, 1636, 1653 in hereditary sensory and autonomic neuropathy, 1815f, 1816, 1838 Foot ulcers, 2629 diabetic, 1972, 1972t in hereditary motor and sensory neuropathy II, 1719 Footdrop diagnosis of, 1324 differential diagnosis of, 1377, 1498, 1499t electromyography in, 944f in hydatidosis, 2170 in lumbosacral radiculoplexus neuropathy, 1999 diabetic, 1996, 1999 in trichinosis, 2164 peroneal neuropathies with, 1497 postpartum, 1380 Footwear, 2629 Formin-1, in actin nucleation, 451 Formyl glycin–generating enzyme (FGE), in multiple sulfatase deficiency, 1855 Foscarnet, for cytomegalovirus infection, 2122, 2123 Fourth nerve palsy, 1195–1197 bilateral, 1196–1197 congenital, 1197 diabetic, 1201 due to infections and toxins, 1202 due to neoplasms, 1199–1200, 1200f due to vascular diseases and aneurysms, 1201 traumatic, 1198 Fracture clavicular, 1366–1367 humeral neck, 1365–1366 Monteggia’s, 1475 scapular, 1367 temporal bone, 1268 facial nerve paralysis and, 1233, 1233f Fragile X syndrome, 705 FRAP (fluoride-resistant acid phosphatase), 693 Frataxin, 706, 1739 Free radicals. See Oxidative stress; Reactive oxygen species (ROS). FRET (fluorescence resonance energy transfer), 1555 Frey’s syndrome, 1283 Friedreich’s ataxia, 705–706, 1738–1739 auditory and vestibular neuropathy in, 1259 axonal atrophy and segmental demyelination in, 783, 785f clinical features of, 1739 compound action potential of sural nerve in, 1020f, 1020–1021 distribution of sensation abnormality in, 1089f FRX in, 1739 in children, 2708, 2724 large fiber sensory loss in, 1147 loss of mechanoreception in, 1064 pathology of, 706, 1739, 1739f–1740f prevalence of, 705 sensory involvement in, 1141 skin biopsy in, 886

Frontotemporal dementia, amyotrophic lateral sclerosis and, 1537, 1600 FRX, in Friedreich’s ataxia, 1739 FS. See Fisher syndrome (FS). Fucosidosis, 1864, 2722 Fumarylacetate hydrolase deficiency, 2713 Functional Assessment of Cancer Treatment–General (FACT-G), 1059 Functional Assessment of Human Immunodeficiency Virus Infection (FAHI), 1059 Functional electrical stimulation (FES), 2623, 2624. See also Electrical stimulation. Functional Independence Measure, 1041–1042 Fungicides, carbamate, 2508–2510 Fyn kinase, 425 in CRASH syndrome, 468

GABA (␥-aminobutyric acid) blockade by ␦-aminolevulinic acid synthase in porphyria, 1885 in cardiac neural control, 218, 219f, 221, 222 in micturition reflex, 312 Gabapentin, 2644 adverse effects of, 2644 for diabetic neuropathy, 1971, 1971t, 2640, 2644 for hemifacial spasm, 1245 for HIV-associated sensory polyneuropathies, 2135 for neuropathic pain, 2644 for postherpetic neuralgia, 2640, 2641, 2644 mechanism of analgesic action of, 2644 topical, 2645 GACT-G (Functional Assessment of Cancer Treatment–General), 1059 Gait abnormalities of, 483 gait disorder, autoantibody, late-age onset, and polyneuropathy (GALOP) syndrome, 2179t, 2189 in hereditary motor and sensory neuropathies, 1624, 1631, 1634 in pediatric chronic inflammatory demyelinating polyradiculoneuropathy, 2718 in transgenic mice deficient in PMP22, 1566 in transgenic mice overexpressing PMP22, 1562, 1562f in transgenic mice with Pmp22 point mutations, 1567–1568 quantifying muscle function during, 1098–1099, 1099f–1100f GAJ (glia-axon-junction) complex, 55 Galactocerebroside (GalC) deficiency of, 2714 in mice, 550, 584 IgG or IgM antibody binding to, 2190 in experimental autoimmune neuritis, 593, 611t, 621, 622f in human neuropathies, 596 in myelin membrane, 423, 580 localization of, 578f structure of, 582t Galactolipids, in myelination, 423–424 Galactosialidosis, 1864 ␣-Galactosidase A deficiency of, 709, 887, 1893, 1894, 1898–1900 replacement therapy for Fabry’s disease, 1900–1901, 1901f–1902f Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxxiii ␤-Galactosidase deficiency, 707, 1860 Galactosulfatide. See Sulfatide. Galactosylceramide lipidosis, 1855–1859 clinical features of, 1856f, 1856–1857 diagnosis of, 1856, 1857 genetics of, 1858–1859 historical description of, 1855 in children, 2708, 2721 infants, 2714 nerve conduction studies in, 1856 pathogenesis of, 1858 pathology of, 710, 1857, 1857f–1859f Schwann cell expression of, 61 stages of, 1856 Galactosylceramide sulfotransferase (GST) deficiency, in mice, 550, 551f, 584 Galactosylsulfatide, in metachromatic leukodystrophy, 1847, 1852 Galanin downregulation by nerve growth factor, 177 expression after axotomy, 185, 690–691 expression by dorsal root ganglion neurons, 185 in bladder, 301, 305 in ganglionic neurons, 242 receptors for, 185 GalC. See Galactocerebroside (GalC). Gallbladder, in Chagas’ disease, 2160 GALOP (gait disorder, autoantibody, late-age onset, and polyneuropathy) syndrome, 2179t, 2189 Galvanic stimulation, 1256 GAN. See Giant axonal neuropathy (GAN). Ganciclovir, for cytomegalovirus infection, 2122, 2123 Ganglion(a) accessory, 17 anatomy of gross, 234–235 microscopic, 235 nerve fibers and synaptic structures, 235–236 presynaptic vs. preganglionic fibers, 236 aorticorenal, 31 celiac, 234 cephalic (cranial) parasympathetic, 17 cervical, 17f, 234 inferior, 23 middle, 23 superior, 23, 493 cervicothoracic (stellate), 17f, 23, 25 ciliary, 203–204, 204f, 234 dorsal root. See Dorsal root ganglion (DRG). enteric, 252f, 252–253 geniculate, 1220 impar, 234 lumbar, 30, 234 mesenteric inferior, 234, 238f, 243 superior, 234 nodose, 1277, 1278, 1279 otic, 234, 17 parasympathetic, 233–235, 235 pelvic, 234, 30 phrenic, 31 previsceral, 234 pterygopalatine, 17 renal, 31 sacral, 30 satellite, 234 Scarpa’s, 1253 sphenopalatine, 234

xxxiv Ganglion(a) (Continued) spinal, 11 morphometry of neuron somas in, 755–756, 758t, 759f–760f spiral, 1253 submandibular, 234, 17 sympathetic, 15–16, 234–235 age-related changes in, 493–494 in abdomen, 30 in head, neck, and upper limb, 15–16 in thorax, 25 morphometry of neuron somas of, 756, 760f paravertebral (chain), 234 prevertebral, 234–235 trigeminal, 1207 vertebral, 23 Ganglion cysts, paraglenoid, 1467 Ganglioneuroma, 2595 alimentary tract ganglioneuromatosis in multiple endocrine neoplasia type 2B, 1806, 1806f, 2595 Ganglionic pathways functional, 234 neurochemistry of, 233, 240–243 actions of neuropeptides, 242–243 neuropeptide phenotype, 240t, 240–241 neurotransmitter phenotype, 240 presynaptic fibers, 241–242 acetylcholine, 241 catecholamines, 241 neuropeptides, 241–242, 242f Ganglionic transmission, 236–240 activity of ganglionic neurons and effectiveness of preganglionic impulses, 239 definition of, 233, 236 divergence and convergence in, 237 in prevertebral sympathetic ganglia, 243–245, 244f–245f rate of nerve firing and, 238–239 specification of autonomic outflow as to target and function in, 237–238, 238f specification of preganglionic input to ganglia and, 237 spontaneous activity of ganglionic neurons, 239–240 Ganglionopathy, sensory, 1148, 2322 nonmalignant inflammatory sensory polyganglionopathy, 2309–2317 clinical patterns of, 2311 course of, 2311–2312 differential diagnosis of, 2314–2317, 2315t electrophysiologic findings in, 2312–2313 historical review of, 2310–2311 laboratory findings in, 2312 pathology of, 2313–2314, 2314f radiologic features of, 2312 symptoms and signs of, 2311 treatment of, 2317 pattern of involvement in, 1139 skin biopsy in, 887 vs. radiculopathy, 1324 Gangliosides, 587–588, 2182 accessibility to antibodies, 588 animal models of neuropathy induced by, 594 antibodies with binding to, 590–591, 2178t–2180t, 2182–2187 diseases related to, 596 GD1a ganglioside, 2186 GD1b ganglioside, 597 GD1b gangliosides, 597, 2186–2187

Index Gangliosides (Continued) GM1 ganglioside, 2182–2185 IgG anti-GM1 antibodies, 2184–2185 IgM anti-GM1 antibodies, 2182–2184 GM1b ganglioside, 2186 GM2 and GalNac-GD1a gangliosides, 2186 GQ1b ganglioside, 2185 GT1a ganglioside, 2185 in Guillain-Barré syndrome, 590, 596, 2200t, 2285 acute inflammatory demyelinating polyneuropathy, 2211 axonal forms, 2207–2210 Fisher syndrome, 2210–2211 in multifocal motor neuropathy and conduction block, 596, 2284–2286, 2286f, 2288–2290 infection and, 590–591 molecular mimicry, 591–592 at neuromuscular junction, 590 definition of, 2182, 2285 distribution of, 587, 588, 589f expression of GM1 receptors by dorsal root ganglion neurons, 183 functions of, 587–588, 2182 in myelin membrane, 577, 580, 584 in myelination, 424 localization to specific nerve sites and fiber types, 588–590 mice deficient in GD1a and GT1b, 548 monoclonal IgM reactive with, 2267–2268 nomenclature for, 586–587 sialic acids in, 2285 structures of, 582t–583t, 586–587, 2183f synthesis of, 587, 587f Gangliosidosis, 1846, 1846f, 1859–1860 classification of, 1859 GM1, 1859, 1860 GM2, 1859–1860 Gap junctions in developing nerve, 39 in myelin sheath, 1796 role of connexins in, 1576, 1793–1795 functional gap junctions in Cx32-deficient mice, 1577–1578 GAP-43. See Growth-associated protein-43 (GAP-43). GAPs (guanosine triphosphatase-activating proteins), 450, 450f–451f GARS gene mutations, in hereditary motor and sensory neuropathy IID, 1629, 1731 Gas exchange measurements, 2612–2613 Gastric acid secretion, vagal function and, 1280 Gastric emptying assessment of, 1280 vagal efferent activity and, 1281, 1281f Gastric pacing, 2691 Gastric surgery, polyneuropathy after, 2057 Gastrin-releasing peptide, in ganglionic neurons, 242 Gastrocolic reflex, 2627–2628 Gastroesophageal reflux disease after spinal cord injury, 2699 hereditary sensory and autonomic neuropathy type I ⫾ cough and, 1823 Gastrointestinal disorders, 1167 arsenic-induced, 2535–2536 assessing autonomic function, 2695 brainstem tumors and, 2695 disinhibitory motor disorders, 259t, 259–261 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Gastrointestinal disorders (Continued) due to autonomic dysfunction, 2664–2665, 2685–2694 constipation, 2610, 2665, 2691–2692 diarrhea, 2616, 2616t, 2665, 2692–2694 fecal incontinence, 287–292, 2627–2628, 2665, 2694 gastroparesis and chronic intestinal pseudo-obstruction, 259–260, 2665, 2688–2691 megacolon, 2692 oropharyngeal dysphagia, 2633, 2685–2687 enteric nervous system malfunctions and, 249, 251, 271–272 idiopathic orthostatic hypotension and, 2696 in amyloidosis, 2434, 2698 familial, 2441 in autonomic epilepsy and migraine, 2695 in botulism, 1281 in Chagas’ disease, 1281, 2158f, 2159–2160 in chronic sensory and autonomic neuropathy of unknown cause, 2698–2699 in cysticercosis, 2168 in diabetes mellitus, 1280, 1281, 1957, 1958–1959, 2696–2698, 2697f–2698f enteric nerve dysfunction and, 2697–2698 epidemiology of, 2696 extrinsic nerve dysfunction and, 2696–2697 humoral mediators of, 2697 management of, 1960, 2698 psychological factors and, 2698 vagal efferent activity and gastroparesis, 1167, 1280, 1281, 1958–1959 in Fabry’s disease, 1896, 1900 in giardiasis, 2167 in HIV infection, 2699 in multiple sclerosis, 2700 in multiple system atrophy, 2696 in myasthenia gravis, 2700 in pandysautonomia/selective dysautonomias, 2695–2696 in paraneoplastic neuropathy, 2698 in Parkinson’s disease, 2695 in peripheral neuropathies, 2696 in porphyria, 2699 in schistosomiasis, 2160 in strongyloidiasis, 2161 in trichinosis, 2164 intensive care unit care for, 2609–2610, 2616 mitochondrial neurogastrointestinal encephalopathy, 1942, 1945–1946, 2700–2701 postpolio dysphagia, 2695 spinal cord injury and, 2699–2700 vagal efferent activity and, 1281, 1281f Gastroparesis, 1280, 2665, 2688–2691, 2688f clinical assessment of, 2688 diagnostic evaluation of, 2688–2689, 2688f–2690f etiology of, 2688 gastric emptying time in, 2688 in diabetic patients, 1167, 1280, 1281, 1958–1959 management of, 1960, 2689–2691, 2690t antiemetic and procontractile agents, 2665, 2690–2691 gastric pacing, 2691 nonpharmacologic, 2689–2690 relief of distention, 2691 suppression of bacterial overgrowth, 2691 surgical, 2691

Index Gastroplasty, 1167 Gastrostomy tube placement, 2610 Gaucher’s disease, 1846, 1909, 1910t, 1860–1861 pathology of, 706–707 GBS. See Guillain-Barré syndrome (GBS). GCRs (glucocorticoid receptors), 637 GCs. See Glucocorticoids (GCs). GDAP1 gene mutations in autosomal recessive hereditary motor and sensory neuropathy, 1770t, 1773, 1779–1780, 2714 in hereditary motor and sensory neuropathy IVA, 424, 1630, 1735 with hoarseness, 1780 with pyramidal involvement, 1780 GDIs (guanine nucleotide dissociation inhibitors), 450, 450f–451f GDNF. See Glial cell line–derived neurotrophic factor (GDNF). GDPA1 gene, 424 GEFs (guanine nucleotide exchange factors), 450, 450f–451f Gelsolin amyloidosis, 1921, 1922, 1932 Gender differences in amyotrophic lateral sclerosis incidence, 1534 in brachial plexopathies, 1343 in carpal tunnel syndrome incidence, 1443 in nerve conduction, 904, 907 in sensation thresholds, 1067 in severity of X-linked dominant disorders, 1548–1549 in Valsalva ratio, 1115, 1115f–1116f Gene mutations. See also specific genes. analysis of genes with limited number of, 1554–1555, 1555f in adrenoleukodystrophy, 1867 in amyotrophic lateral sclerosis, 1534–1535, 1539, 1586, 1596–1599 in auditory neuropathy, 1259, 1261f in autosomal dominant disorders, 1546–1547 in autosomal recessive disorders, 1547–1548 in Batten disease, 708 in Chédiak-Higashi syndrome, 1863–1864 in congenital hypomyelinating neuropathy, 579 in congenital myasthenic syndromes acetylcholinesterase receptor deficiency with or without minor kinetic abnormality, 853–855, 855f caused by rapsyn deficiency, 856f, 857 choline acetyltransferase deficiency, 840–841, 841f end-plate acetylcholinesterase deficiency, 843–844, 846f fast-channel syndromes, 851f, 851–852 slow-channel syndromes, 848f, 849–851 sodium channel myasthenia, 858–859, 859f in Farber’s disease, 1859 in Gaucher’s disease, 1860 in GM1 gangliosidosis, 1860 in GM2 gangliosidoses, 1859–1860 in hereditary motor and sensory neuropathies, 1627t, 1628–1630, 1718, 1718f type IA, 1659 type IB, 579 type II, 1719t X-linked, 543, 704 in hereditary sensory and autonomic neuropathy, 1810, 1811, 1811t HSAN I, 1819–1822, 1822f

Gene mutations (Continued) in Krabbe’s disease, 1858–1859 in metachromatic leukodystrophy, 1852–1854 in mucolipidosis IV, 1865 in multiple sulfatase deficiency, 1855 in muscular dystrophy, 424 in neuronal ceroid-lipofuscinoses, 1863, 1864t in Niemann-Pick disease, 708, 1862 in Pompe’s disease, 1864 in Refsum’s disease, 1869 in Wolman’s disease, 1864 in X-linked dominant disorders, 1548–1549 in X-linked recessive disorders, 1548 in Y-linked disorders, 1549 methodologies for analysis of, 1551–1559 of dystrophin gene, 1545 PMP22 point mutations as cause of demyelinating neuropathies, 1665–1667, 1666f recombinant DNA technology for detection of, 1553–1554 Gene polymorphisms, 1551–1552 of vascular endothelial growth factor, in familial amyotrophic lateral sclerosis, 1534–1535 restriction fragment length, 1551–1552 single-nucleotide, 1553 single-strand conformation, 1557 Gene therapy, for Fabry’s disease, 1901 Genetic counseling for familial amyotrophic lateral sclerosis, 1596 for hereditary brachial plexus neuropathy, 1765–1766 for hereditary motor and sensory neuropathies, 1652 for mendelian disorders, 1549 Genetic disorders diagnostic methodologies for, 1551–1559 analysis of genes with large number of known mutations, 1556, 1556f analysis of genes with large number of unique mutations, 1557–1559, 1558f analysis of genes with limited number of mutations, 1554–1555, 1555f fluorescent in situ hybridization, 1551, 1552f genetic linkage, 1551–1553 polymerase chain reaction, 1553–1554 incomplete penetrance of, 1546 inheritance patterns for, 1545–1551, 1595 imprinting, 1550–1551 Mendelian, 1545–1549, 1595 mitochondrial, 1549–1550 multifactorial, 1551 prenatal diagnosis of, 1549 presymptomatic diagnosis of, 1549 variable expression of, 1546 Genetic epidemiology, 1184 Genetic linkage analysis, 1184, 1551–1553 composite scores in, 981 Genetic testing, 1171 in pediatric polyneuropathies, 2712 Geniculate fossa, 1220 Geniculate ganglion, 1220, 1224f, 1220 sensory systems of, 1222, 1223f Geniculate neuralgia, 1246 Genioglossus muscle, innervation of, 1286 Geniohyoid muscle, innervation of, 1286 Genitofemoral nerve distribution of, 16f, 26, 1488, 1488f lesions of, 1488 structures innervated by, 1376 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxxv Genitourinary dysfunction, 1150–1151 in diabetic patients, 1957, 1959, 2628 Genome, 1545, 1553 Genotype, 1546 Germinal mosaicism, 1546 GFAP (glial fibrillary acidic protein), 343, 434 Giant axonal neuropathy (GAN), 705, 1651, 1736–1738, 1782–1783, 1783f in children, 2723–2724 pathology of, 705, 778, 782f, 1736–1738, 1737f skin biopsy in, 888 teased fiber analysis in, 1736f vs. neuroaxonal dystrophy, 1738 Giant cell arteritis, 2337, 2354t, 2362 carpal tunnel syndrome and, 1445 classification and definition of, 2338t, 2340t cytokines in, 2344 diagnostic criteria for, 2338 genetics of, 2363 incidence of, 2340t polymyalgia rheumatica and, 2362 treatment of, 2336, 2376t Giardiasis, 2154t, 2167–2168 clinical features of, 2167–2168 etiologic agent in, 2163f, 2167 life cycle of, 2167 peripheral neuropathy in, 2168 reactive arthritis in, 2167–2168 transmission of, 2167 Gigaxonin gene mutations, 705, 1651, 1770t, 1782–1783 GJB1 gene mutations. See also Connexin 32 (Cx 32). in X-linked Charcot-Marie-Tooth disease, 543, 704, 1259, 1628–1629, 1727–1728, 1729f, 1791–1793, 1792f GLD. See Globoid cell leukodystrophy (GLD). Glia limitans, 67, 68 Glia-axon-junction (GAJ) complex, 55 Glial cell line–derived neurotrophic factor (GDNG), 179–180, 378, 380 aging effects on, 496 dependence of dorsal root ganglion neurons on, 176 epidermal nerve fibers and, 872 expression after axotomy, 179, 694 expression on dorsal root ganglion neurons, 179 family of, 179 receptors for, 176t, 179–180 role in myelination, 64, 362 sources of, 179 structure of, 378 Glial cells, peripheral, 341 developmental potential of, 349–351 neural crest origins of, 346–348 Glial elements of enteric nervous system, 253 Glial fibrillary acidic protein (GFAP), 343, 434 Glial fringe, 68 Glial islands, 71 Globoid cell leukodystrophy (GLD), 1855–1859 clinical features of, 1856f, 1856–1857 diagnosis of, 1856, 1857 genetics of, 1858–1859 historical description of, 1855 in children, 2708, 2721 infants, 2714 nerve conduction studies in, 1856 pathogenesis of, 1858 pathology of, 710, 1857, 1857f–1859f stages of, 1856

xxxvi Glossodynia, 1288 Glossopharyngeal nerve, 1273–1277 anatomy of, 1273–1274 carotid sinus nerve, 1274 motor fibers, 1273 parasympathetic fibers, 1273 sensory fibers, 1273–1274 carotid sinus baroreceptor innervation by, 323 blood pressure effects of nerve block, 325, 326f, 1274, 1275f carotid sinus hypersensitivity and, 1277 clinical features of lesions of, 1274 course of, 1273 extramedullary lesions of, 1274–1275 intramedullary lesions of, 1274 testing function of, 1274 Glossopharyngeal neuralgia, 1276–1277 age distribution of, 1276 cardiac effects of, 1276, 1276f clinical features of, 1276 etiologies of, 1276–1277 idiopathic, 1276 pain of, 1274, 1276 treatment of, 1277 trigeminal neuralgia and, 1214, 1276 Glove-and-stocking distribution of sensory neuropathy, 1138, 1139f, 1140 Glucocerebroside, in Gaucher’s disease, 706 Glucocorticoid receptors (GCRs), 637 Glucocorticoid response elements (GREs), 637 Glucocorticoids (GCs), 635, 636–637. See also Corticosteroids. adverse effects of, 637 for experimental autoimmune neuritis, 627, 628f half-life of, 637 in inflammation, 636 mechanism of action of, 636 pharmacokinetics of, 636–637 receptor binding of, 637 use in neuropathy, 637 Glucose for porphyria, 1885, 1891 neuropathy related to impaired tolerance of, 1953 Glucosylceramidase deficiency, 706 Glucosylceramide lipidosis, 1846, 1909, 1910t, 1860–1861 pathology of, 706–707 Glutamate axon dystrophy and, 497 effect of nerve injury on, 184 expression by dorsal root ganglion neurons, 184 in amyotrophic lateral sclerosis, 1538 in cardiac neural control, 218, 219f, 220–221 in excitotoxic neuron death, 699 in inflammation, 184, 525 in micturition reflex, 312 NMDA receptors for, 524–527 effect of inflammation on, 525 in amyotrophic lateral sclerosis, 1538 in chronic constriction model, 525–527 localization of, 525 role in pain, 524–527, 2640 ␥-Glutamylcysteine synthetase, in diabetic patients, 511 Glutathione, reduced (GSH), 511, 514 for treatment of experimental diabetic neuropathy, 516, 516t in diabetic patients, 511 peripheral nerve levels, 515, 515t

Index Glutathione, reduced (GSH) (Continued) in experimental diabetic neuropathy, 512, 513f, 516, 517 in ischemia, 521 Gluteal artery inferior, 800–801, 1376 superior, 1376 Gluteal compartment syndrome, 1494–1495 Gluteal nerve inferior, 28 distribution of, 1493 lesions of, 1493 structures innervated by, 957t, 1376 superior, 28 distribution of, 1493 involvement in vasculitic neuropathy, 2365t lesions of, 1493 structures innervated by, 957t, 1376 Glycation, nonenzymatic, in diabetic neuropathy, 1974–1976, 1975f Glycine, in micturition reflex, 312 Glycogen granules, in hypothyroid neuropathy, 2042, 2043f–2044f Glycogenosis type II, 1864 pathology of, 709, 1864 Glycogenosomes in axons, 487, 488f–489f Glycolipids in myelin membrane, 423–424, 577, 580–584 monoclonal IgM reactive with, 2267 structures of, 581t–583t Glycosaminoglycans, in epilepsy, 466 Glycosphingolipids, definition of, 2285 GM1 gangliosidosis, 707, 1859, 1860 GM1 receptor, expression by dorsal root ganglion neurons, 183 GM2 activator protein deficiency, 707 GM2 gangliosidoses, 1859–1860 pathology of, 707–708 Gold neuropathy, 1170, 2543–2544 Goldenhar’s syndrome, 1211 Golgi tendon organ, 151–159 central projections of, 158–159 connection between axoplasmic reticulum and, 47 discharge frequency-tension relation of, 154f, 155–156 distribution of, 153 dynamic sensitivity of, 155–156 in control of movement, 151 in proprioception, 159 innervation of, 151, 153 location of, 151 muscle fiber insertion into, 151–152 recordings during natural movements, 156 recovery from nerve lesions, 157–158 reflex actions of, 158 responses to muscle contraction, 153–155, 154f septal cells of, 153 size of, 151 static sensitivity of, 155 stretch response of, 153, 156–157 structure of, 151–153, 152f transduction of, 156–157 unloading response of, 152, 154 Golgi-Rezzonico spiral, 55 Golli proteins, 419–420 Gonadal mosaicism, 1546 Gout, carpal tunnel syndrome and, 1447 gp120, in HIV-associated distal symmetrical polyneuropathy, 2133 Gradenigo’s syndrome, 1202 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Graft-versus-host disease, 2498 Granular cell tumor, 2595 malignant, 2595, 2600 Granular malignant peripheral nerve sheath tumor, 2599, 2599f Granuloma annulare, carpal tunnel syndrome and, 1445 Granulomas consequences of formation in peripheral nervous system, 2417 in leprosy, 2084, 2100, 2417 in sarcoidosis, 2416–2417, 2421, 2421f–2422f Grb4, phosphorylated ephrin-B and, 455–456 Great auricular nerve, 18 in leprosy, 2089f GREs (glucocorticoid response elements), 637 Grisel syndrome, 2737 Growth cone, neuronal, 447–450 actin dynamics and, 450f–451f, 450–453 attractive and repulsive guidance cues from, 448, 448f cytoskeletal dynamics of, 448f–449f, 448–450 function of, 447–448 receptors for, 447–448 Growth factors, 377–383. See also Neurotrophic factors. Growth-associated protein-43 (GAP-43) expression after axotomy, 690, 691f expression in dorsal root ganglion cells after sciatic nerve crush, 1297 in Schwann cell development, 346 upregulation by nerve growth factor, 176 GSH. See Glutathione, reduced (GSH). Guanine nucleotide dissociation inhibitors (GDIs), 450, 450f–451f Guanine nucleotide exchange factors (GEFs), 450, 450f–451f Guanosine triphosphatase-activating proteins (GAPs), 450, 450f–451f Guillain-Barré syndrome (GBS), 2178t, 2197–2213 after bone marrow transplantation, 2498, 2498t autoantibodies in, 2178t antiganglioside, 590, 596, 2200t, 2285 GM1b ganglioside, 2186 GM2 and GalNac-GD1a gangliosides, 2186 P2, 595–596 PMP22, 595 to galactocerebroside, 596 to heparan sulfate glycosaminoglycans, 576 autonomic involvement in, 2608, 2611, 2630 blink reflex in, 920t chronic lymphocytic leukemia and, 2496 classification of, 2199, 2200t clinical manifestations of, 2199–2201 acute inflammatory demyelinating polyneuropathy, 2199–2200 axonal forms, 2200–2201 Fisher syndrome, 2199 cranial nerve involvement in, 1139 glossopharyngeal nerve, 1274, 1275 differential diagnosis of, 2201–2202 botulism, 2202 myasthenia gravis, 2202 neoplasms, 2202 paralytic rabies, 2202 poliomyelitis, 1300 polymyositis, 1152, 2201 porphyric neuropathy, 1888, 1889 tick paralysis, 2173–2174, 2201–2202

Index Guillain-Barré syndrome (GBS) (Continued) electromyography in macro, 993f, 997–998, 998f single-fiber, 997–998 epidemiology of, 1186–1187, 1188t experimental autoimmune neuritis and, 2199 heterogeneity of, 2197, 2199 history of, 2197–2199 early descriptions, 2197–2198 mechanisms and treatments, 2199 syndromic boundaries, 2198–2199 immunology of, 575 in children, 2715–2717 cerebrospinal fluid analysis in, 2716 clinical features of, 2715–2716 diagnosis of, 2707–2709 during infancy, 2712–2713 electrophysiology of, 2716 immunotherapy for, 2716 magnetic resonance imaging in, 2716 mimickers of, 2716–2717 prognosis for, 2716 in diabetic patients, 1955 in HIV infection, 2136–2137 clinical features of, 2136–2137 disease-stage specificity of, 2129 etiology of, 2137 pathology of, 2137 prevalence of, 2130 treatment of, 2137 infections associated with, 590–591, 1187, 2171–2174 Campylobacter jejuni, 591, 592, 2198, 2199 cyclosporiasis, 2173 cytomegalovirus, 590, 2122 malaria, 2171–2172 toxoplasmosis, 2172–2173 long-term disability from, 635 management of, 2199, 2212–2213, 2630 immunotherapy, 635, 2199, 2213 glucocorticoids, 637 intravenous immunoglobulin, 643 plasma exchange, 641, 642 in intensive care unit, 2212–2213, 2607–2617, 2609t. See also Intensive care unit (ICU). monitoring and education in early stages, 2212 rehabilitation program, 2630 mortality from, 2201 motor involvement in, 1141 myokymia in, 1142 nerve excitability studies in, 126 overview of, 2197 paralysis due to conduction block in, 1142 paraneoplastic, 2478 pathogenesis of, 1187, 2197, 2199, 2205–2212 acute inflammatory demyelinating polyneuropathy, 2211–2212 axonal forms, 2206–2210 Fisher syndrome, 2210–2211 pathology of, 2202–2205 acute inflammatory demyelinating polyneuropathy, 2202–2204, 2203f–2206f acute motor axonal neuropathy, 2204–2205, 2208f acute motor-sensory axonal neuropathy, 2204–2205 Fisher syndrome, 2205 sarcoidosis and, 2419 segmental demyelination in, 4

Guillain-Barré syndrome (GBS) (Continued) temporal profile of, 1167 ulcerative colitis and, 1167 upper limb-predominant, 1164 Guillain-Barré Syndrome Disability Scale, 1042–1043, 1046t Guinea worm disease. See Dracunculiasis. Gustatory piloerection, 1283 Gustatory sweating, 1150, 1282–1283, 1283f, 1959 Gustolacrimal reflex, 1275–1276 Gut dysmotility, 2683–2700 assessment of autonomic function, 2695 caused by extrinsic neurologic disorders, 2695–2701 acute peripheral neuropathy, 2696 amyloid neuropathy, 2698 autonomic epilepsy and migraine, 2695 brainstem tumors, 2695 chronic sensory and autonomic neuropathy of unknown cause, 2698–2699 diabetes mellitus, 2696–2698, 2697f–2698f HIV infection, 2699 idiopathic orthostatic hypotension, 2696 mitochondrial neurogastrointestinal encephalomyopathy, 1942, 1945–1946, 2700–2701 multiple sclerosis, 2700 multiple system atrophy, 2696 myasthenia gravis, 2700 pandysautonomia/selective dysautonomias, 2695–2696 paraneoplastic neuropathy, 2698 Parkinson’s disease, 2695 porphyria, 2699 postpolio dysphagia, 2695 spinal cord injury, 2699–2700 gastrointestinal symptoms of autonomic neuropathy, 2664–2665, 2685–2694 constipation, 2610, 2665, 2691–2692, 2665 diarrhea, 2616, 2616t, 2665, 2692–2694 fecal incontinence, 287–292, 2627–2628, 2665, 2694, 2665 gastroparesis and chronic intestinal pseudo-obstruction, 259–260, 2665, 2688–2691 megacolon, 2692 oropharyngeal dysphagia, 2633, 2685–2687 regulation of gastrointestinal sensorimotor function by extrinsic nerves, 2683–2685, 2684f–2686f Guyon’s canal, 1478 Gynecologic anomalies, vs. pediatric lumbar radiculopathy, 2738

H current, 191 H reflex, 919–921, 922f in premature neonates, 919–921 in single-fiber electromyography, 989 normal values for, 922t Haber-Weiss reaction, 510 Haemophilus influenzae infection, Guillain-Barré syndrome and, 590, 591 Hair cells inner, 1253 carboplatin toxicity to, 1268 in auditory neuropathy, 1261, 1262f noise-induced damage of, 1268 outer, 1253 testing function of, 1256 vestibular, 1253, 1254f Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxxvii Hair follicles, 870 innervation of, 871 quantitation of, 881–882 Haloxon, 2515 Hamartoma, lipofibromatous, 2602, 2602f Hammer toe deformity, 1168, 1168f in children, 2710 in hereditary sensory and autonomic neuropathy type I, 1815f Hand in hereditary motor and sensory neuropathies, 1624, 1631 orthoses for, 2623, 2630 paralysis of, 2630 Handgrip dynamometer, 1096 Handicap definition of, 1033, 1054 quantitation of, 1048–1049 Hansen’s disease. See Leprosy. HAQ (Health Assessment Questionnaire), 1059 Harris-Benedict formula, 2610 HAT/QoL (HIV/AIDS-Targeted Quality of Life), 1059 HBPN. See Brachial plexus neuropathy, hereditary (HBPN). HCP. See Hereditary coproporphyria (HCP). Hcy (homocysteine), vitamin B12 deficiency and, 2053, 2054 HDL. See High-density lipoprotein (HDL). Head, innervation of, 18–23 autonomic, 22–23 Head injury, cranial nerve injury and, 1198 facial nerve paralysis, 1232–1233, 1233f Head-up tilt, blood pressure response to, 1112–1113, 1115–1116, 1116t Health Assessment Questionnaire (HAQ), 1059 Health outcomes definitions of impairment, disability and handicap, 1033, 1054 quality of life and, 1054. See also Quality of life (QOL). quantitation of, 1031–1049 requirements of measures of, 1056–1058 acceptability, 1057–1058 reliability, 1056–1057 responsiveness, 1057 validity, 1057 Health-Related Quality of Life and Clinical Outcomes Score in Peripheral Neuropathy, 1045–1046, 1048t Hearing, 1253 assessment of, 1065 acoustic reflex, 1256 auditometric, 1255–1256 auditory brainstem responses, 1256 electrophysiologic, 1256 otoacoustic emissions, 1256 Hearing loss, 1168 age-related, 496, 1254 due to vestibular paroxysmia, 1264 due to vestibular schwannoma, 1264 in auditory neuropathy, 1259 in cassava (cyanide) intoxication, 2074 in Charcot-Marie-Tooth disease, 1168, 1259 in Cuban epidemic neuropathy, 2066 in hereditary motor and sensory neuropathy, Lom type, 1259 in hypothyroidism, 2040 in sarcoidosis, 2418 with optic neuropathy in Tanzania, 2067

xxxviii Heart. See also Cardiac and Cardiovascular entries. in amyloidosis, 2433 familial, 2441 in diphtheria, 2147 in Lyme disease, 2112 in transthyretin amyloidosis, 1923 neural control of cardiac function, 217–227 Heart rate (HR) abnormalities in diabetic patients, 1957, 1958 blood pressure, sympathetic outflow and, 323–324, 324f cardiac vagal efferent activity and, 1280 effect of vagal efferents on variability of, 225 effects of mental stress on, 331–332 exercise effects on, 328–331 in carotid sinus hypersensitivity, 1277 in glossopharyngeal neuralgia, 1276, 1276f recordings to evaluate cardiovagal function, 1109–1111 comparison of tests, 1110–1111 patient preparation for, 1109 response to deep breathing, 1109, 1109t, 1110f in Composite Autonomic Severity Score, 1113t, 1114–1115, 1115t response to intravenous tyramine, 1124 response to prolonged tilt, 1123 response to standing, 1110 Valsalva ratio, 1109–1110, 1110t response to orthostatic challenge, 326–327, 327f resting, 217 spectral analysis of variability during exercise, 225 sympathovagal balance and, 225 Heat shock proteins (HSPs), 690 Heat therapy, 2630 Heat transduction, 186–187 Heat-pain threshold (HPT), 1086–1087, 1087f–1088f. See also Quantitative sensation testing (QST). CASE IV testing of, 1086–1087 patterns of abnormality on, 1067 instructions for nonrepeating ascending algorithm with null stimuli for, 1070–1071 Heavy metals. See also Metal neuropathy. pathology induced by, 715–716, 716f, 2527–2545 screening for, 1172 Heerfordt’s syndrome, 2418, 2420 Hemangioblastoma, 2596 Hemangioma, 2596 of facial nerve, 1221–1222, 1222f Hemangiomatosis, neonatal, 2734 Hematin for porphyria, 1891 for tyrosinemia, 2713 Hematoma iliopsoas, 1378, 1379f nerve compression by, 1518 Hematopoietic stem cell transplantation (HSCT). See also Bone marrow transplantation. for adrenoleukodystrophy, 1868 for amyloidosis, 2439 Heme biosynthesis, 1883–1884, 1884f drug effects on, 1884–1885 other factors affecting, 1885

Index Hemifacial spasm, 1145–1146, 1238–1245 bilateral, 1238 clinical features of, 1238 differential diagnosis of, 1239 during sleep, 1238 electrodiagnostic testing in, 1240–1242, 1241f, 1243f etiologies of, 1239f, 1239–1240 familial, 1238–1239 geniculate neuralgia and, 1246 hypertension and, 1240 neuroimaging in, 1239f pathology of, 1242 treatment of, 1242–1245 botulinum toxin, 1245 medical therapy, 1245 other surgical procedures, 1245 surgical microvascular decompression, 1242–1245 trigeminal neuralgia and, 1213–1214 Hemimasticatory spasm, 1146 Hemizygote, 1548 Hemodialysis for uremic neuropathy, 124–125, 2022, 2024 prevalence of polyneuropathy in patients undergoing, 2023 Hemorrhage into nerve, 794–797, 798f–800f lumbosacral plexus compression by, 1378, 1379f Hemorrhagic cystitis, cyclophosphamide-induced, 638 Hemosiderin, in macrophages, 797, 798f Henoch-Schönlein purpura (HSP), 2354t, 2362, 2405 classification and definition of, 2338t, 2340t genetics of, 2363 incidence of, 2340t pediatric polyneuropathy and, 2719 Heparan sulfate, antibodies to, 2190 Heparan sulfate proteoglycans, in Kallman’s syndrome, 469 Heparin for deep venous thrombosis prophylaxis, 2609, 2629 in modulating action of growth factors, 377 Hepatic disorders Child-Pugh severity classification of, 2018 in amyloidosis, 2434 in Fabry’s disease, 1896 in POEMS syndrome, 2455 in sarcoidosis, 2420 neuropathies associated with, 2017–2021 autoimmune hepatitis in Sjögren’s syndrome, 2021 cholestatic hepatic disease and vitamin E deficiency, 2021 Guillain-Barré syndrome and viral hepatitis, 2018 in hepatitis B, 2020 in hepatitis C, 2018–2020, 2019f Navajo neuropathy, 2021 sensory neuropathy and primary biliary cirrhosis, 2020–2021 somatic and autonomic polyneuropathy, 2017–2018, 2018t Hepatic heme biosynthesis, 1883–1884, 1884f Hepatitis B virus infection neuropathies associated with, 2020 polyarteritis nodosa and, 2020, 2335, 2336, 2349–2350 treatment of, 2375t, 2376 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Hepatitis C virus infection cryoglobulinemic vasculitis and, 2019f, 2020, 2335, 2350–2351 treatment of, 2375t, 2376 neuropathies associated with, 2018–2020 cryoglobulinemia and, 2018–2020 electrophysiology of, 2019 nerve biopsy in, 2019f, 2019–2020 signs and symptoms of, 2019 treatment of, 2020 Hepatocyte growth factor (HGF), 356 Hepatoma, 1168 Hereditary brachial plexus neuropathy. See Brachial plexus neuropathy, hereditary (HBPN). Hereditary coproporphyria (HCP), 1883–1891. See also Porphyric neuropathy. clinical manifestations of, 1886, 1887 neurologic disease, 1888 inheritance of, 1887 laboratory studies in, 1889t, 1889–1890 mechanisms of neuronal damage in, 1885–1886 treatment of, 1890–1891 Hereditary high-density lipoprotein deficiency. See Tangier disease (TD). Hereditary motor and sensory neuropathy (HMSN), 1623–1653, 2330 age at onset of, 1651 autosomal dominant, 1625t, 1625–1626, 1626f, 1627t autosomal recessive, 1626, 1627t, 1769–1785, 2714 classification of, 1770t genetic defects in, 1770t, 1771 infantile, 2714 phenotypes of, 1770t type I, 1771–1777 chromosome 5q32/KIAA1985, 1771–1773, 1756f chromosome 8q13-q21.1/GDAP1, 1773 chromosome 8q24/NDRG1, 1773f, 1773–1774 chromosome 10q21-q22/EGR2, 1774 chromosome 11p15/SBF2 (MTMR13), 1774 chromosome 11q23/MTMR2, 1774–1775, 1775f chromosome 17p11.2/PMP22, 1775 chromosome 18q23-qter/CTDP1, 1775–1776, 1776f chromosome 19q13/PRX, 1776–1777, 1777f congenital hypomyelination neuropathy, 1777, 1778f with myelin folding, 1777 type II, 1770t, 1778–1785 chromosome 1q21/LMNA, 1778–1779 chromosome 5q/SMN1, 1779 chromosome 8q21.1/GDAP1, 1779–1780 chromosome 10q23.2, 1780, 1781f chromosome 11q13-q21/IGHMBP2, 1780–1781, 1782f chromosome 15q13-q15/SLC12A6, 1781–1782 chromosome 16q24/gigaxonin, 1782–1783 chromosome 19q13.3, 1783 with acrodystrophy, 1785 with deafness, mental retardation, and absence of large myelinated fibers, 1784–1785 with early onset, 1783–1784, 1784f

Index Hereditary motor and sensory neuropathy (HMSN) (Continued) blink reflex in, 920t chronic inflammatory demyelinating polyradiculoneuropathy and, 2233 classification of, 1625t, 1625–1628, 1627t, 1681, 1707–1708 clinical course of, 1141, 1651–1652 clinical features of, 1631–1636 autonomic dysfunction, 1635–1636 bone and joint abnormalities, 1636 classic phenotype, 1631–1632 hearing loss, 1168, 1259 hypertrophic nerves, 1636, 1636f sensory dysfunction, 1635 tremor, 1624, 1631, 1636 weakness, 1631–1635, 1632t description of syndromes, 1624–1625 Charcot-Marie-Tooth disease, 1624 congenital hypomyelinating neuropathy, 1625 Déjérine-Sottas neuropathy, 1624–1625 Roussy-Lévy syndrome, 1624 diagnostic approach for various forms of, 1718t epidemiology of, 1185–1186, 1186t, 1630–1631 gene identification studies in, 703, 1659, 1623, 1626, 1627t, 1628–1630, 1632f, 1718, 1718f, 1719t genetic counseling about, 1652 giant axonal neuropathy, 705, 1651, 1736f–1737f, 1736–1738, 1782–1783, 1783f health insurance for patients with, 1652 hereditary brachial plexus neuropathy, 1753–1766 historical overview of, 1624–1630, 1681–1682 in children, 2707–2708, 2721 infants, 2713–2714 macro electromyography in, 993f, 998–999 multiple endocrine neoplasia type 2B, 1805–1807 myokymia in, 1143 nerve conduction studies in, 1636–1639, 1637f–1638f, 1639t, 1681, 1682 nomenclature for, 1623 outcome of, 1651–1652 pathology of, 703–704, 1152, 1639–1648 patterns of inheritance of, 1769 quality-of-life issues in, 1060 related to early growth response 2 gene, 1707–1714 transgenic models of, 1575–1576 related to P0 mutations, 1681–1701 classic phenotype of, 1692–1694 differential diagnosis of, 1699 early-onset neuropathy, 1688–1692, 1689f, 1690t–1691t, 1694–1695 epidemiology and genetics of, 1688 genetic testing for, 1709 late-onset neuropathy, 1692, 1693t, 1695–1696 management of, 1699–1701 pathogenesis of, 1696–1699 intracellular trafficking, 1696 Schwann cell–axon interactions and axonal degeneration, 1698–1699 tertiary structure of extracellular domain, 1696–1698, 1697t truncation of cytoplasmic domain, 1698

Hereditary motor and sensory neuropathy (HMSN) (Continued) pathology of, 1694–1696, 1695f phenotypic variability of, 1694 transgenic models of, 1572–1575 related to PMP22 altered dosage or mutation, 1659–1673 Charcot-Marie-Tooth disease type 1A, 1659–1665, 1671–1673 hereditary neuropathy with liability to pressure palsies, 1659, 1667–1673 transgenic models of, 1561–1572 transgenic models of, 1561–1579 connexin 32, 543–545, 544f–545f, 1576–1578 early growth response 2 gene, 1575–1576 P0, 536–542, 537f–542f, 1572–1575 periaxin, 545–546, 546f, 1578–1579 peripheral myelin protein 22, 542–543, 1561–1572 trembler mouse, 704, 1566–1572, 1647, 1698 treatment of, 1652–1653 type I, 1625t, 1626, 1769 autosomal recessive, 1770t compared with HMSN II, 1717–1718 compound action potential of sural nerve in, 1025–1026, 1027f electron microscopy of teased fibers in, 755f genetics of, 1627t, 1628, 1628f, 1681 in children, 2708, 2721 pathology of, 1639–1641, 1640f–1644f, 1644t pattern of involvement in, 1138, 1138f Schwann cell–axon interactions in, 1646–1648, 1651f–1652f with EGR2 mutation, 1707, 1708f, 1709, 1709t clinical features of, 1710t type IA nerve conduction studies in, 1682 nerve excitability studies in, 126 PMP22 in, 419, 542, 579, 596, 1628, 1659 increased PMP22 gene dosage, 1663–1665, 1664t with chromosome 17p12 duplication, 1628, 1659–1665, 1671–1673 case studies of, 1663 clinical findings associated with, 1660–1661, 1661t detection of, 1660, 1660f electrophysiologic features of, 1661–1663 generation by unequal crossing over, 1671–1672, 1672f management of, 1663 molecular diagnostics for detection of, 1672–1673, 1673f pathology of, 1661, 1662f structure of genomic region, 1671 type IB genetic testing for, 1700 genetics of, 1628, 1681–1682, 1682f, 1688 management of, 1699–1701 nerve conduction studies in, 1682 P0 mutations in, 579, 1682, 1688, 1690t–1691t type IF, 1732 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xxxix Hereditary motor and sensory neuropathy (HMSN) (Continued) type II, 1625t, 1626, 1717–1734, 1769 axonal dysfunction in, 1719–1722, 1720f–1721f clinical and molecular mimickers of, 1722–1724, 1723f common features of, 1717 compared with HMSN I, 1717–1718 definition of, 1717 genetics of, 1627t, 1629, 1718f, 1719t molecular genetic classification, 1724–1725 identifying axonally affected individuals, 1722 in children, 2708, 2721 multiple system involvement in, 1719 nerve conduction studies in, 1717–1718 pathology of, 1643–1646, 1650f, 1650t pattern of involvement in, 1138 phenotype of, 1688 sensory involvement in, 1722–1723 subtypes of, 1629 terminology for, 1717 with P0 mutations, 1693t, 1725–1727, 1726f–1727f type IIA (Kif1B mutations), 1629, 1728–1730 type IIB (RAB7 mutations), 1629, 1730–1731 type IIC, 1732–1734, 1733f–1734f infantile, 2714 type IID (GARS mutations), 1629, 1731 type IIE (NEFL mutations), 1629, 1732 type IIF, 1734 type III (Déjérine-Sottas neuropathy), 1625t, 1626, 1688, 1769 clinical features of, 1624–1625, 1632 diagnostic criteria for, 1625 gene mutations causing, 703, 1625, 2714 EGR2, 1707, 1708f, 1710t P0, 579, 1688, 1690t–1691t historical background of, 1624–1625, 1681 in children, 2708 infants, 2713–2714 inheritance pattern of, 1624, 1625, 2714 nerve conduction studies in, 1624, 1625 pathology of, 819, 1625, 1641–1643, 1645f–1649f periaxin in, 580 vs. congenital hypomyelinating neuropathy, 1625 type IV (Refsum’s disease), 1625t, 1626, 1866, 1869–1871 clinical features of, 1869 course of, 1141 demyelination in, 1167 genetics of, 1627t, 1629–1630, 1869 hearing loss in, 1168 histologic changes in, 1870–1871, 1871f infantile, 1866, 1869, 1872 nerve thickening in, 1152 pathology of, 710, 1869–1870 phytanic acid elevation in, 1866, 1869 pigmentary retinopathy in, 1168, 1869 type IVA, 1629–1630, 1735–1736, 1770t GDAP1 mutation in, 424, 1630, 1735 type IVB, 1630, 1769, 1770t type IVC, 1630 type IVD, 1630 type IVF, 1630, 1770t periaxin in, 424, 580, 1630 type V (with spastic paraplegia), 1625t, 1626, 1742t, 1742–1743, 1771

xl Hereditary motor and sensory neuropathy (HMSN) (Continued) type VI (with optic atrophy), 1625t, 1626, 1627t, 1649–1650, 1743–1744, 1771 type VII (with retinitis pigmentosa), 1625t, 1626, 1627t, 1650, 1744, 1771 with corpus callosum agenesis (Andermann syndrome), 1738, 1770t, 1781–1782 with proximal involvement, 1734 X-linked Charcot-Marie-Tooth disease, 1626, 1791–1800 animal models of, 1800 brainstem auditory evoked responses in, 1797 central nervous system involvement in, 1797 discovery of, 1791 gender differences in severity of, 1629 genetics of, 1627t, 1628–1629 genotype-phenotype correlations in, 1799–1800 GJB1 (connexin 32) mutations causing, 543, 704, 1259, 1628–1629, 1727–1728, 1729f, 1791–1796, 1792f in children, 2721 medical issues in, 1800 nerve conduction studies in, 1797–1798, 1798f neuromuscular manifestations of, 1796f, 1796–1797 nonuniform demyelination in, 1167 pathology of, 1798–1799, 1799f type 2, 1734 type 3, 1734–1735 Hereditary motor neuronopathy (HMN), 1603–1618. See also Spinal muscular atrophy (SMA). bulbar, 1604t, 1605t, 1617 classification of, 1603–1605, 1604t, 1605t diagnosis of, 1605–1606 distal, 1604t, 1613t, 1613–1615, 1614f electromyography in, 1605 etiology of, 1618 fascioscapulohumeral, 1604t, 1616–1617 genetic linkage studies in, 1603 heterogeneity of, 1603 management of, 1618 mode of inheritance of, 1603, 1604t muscle biopsy in, 1605–1606 nerve conduction studies in, 1605 proximal, 1604t, 1606–1612 nosology of, 1606, 1607t type I (acute infantile), 1606–1607, 1608f type II (chronic childhood), 1607–1610, 1610f type III (autosomal recessive, of adult onset), 1610–1611 types IV and V (autosomal dominant, of juvenile and adult onset), 1611–1612 Ryukyuan, 1617–1618 scapuloperoneal, 1604t, 1613t, 1615f, 1615–1616 segmental, 1604t, 1613t, 1618 site of motor weakness in, 1604t, 1605 vs. hexosaminidase A deficiency, 1606 with oculopharyngeal involvement, 1604t, 1617 X-linked bulbospinal neuronopathy, 1612, 1612f Hereditary neuralgic amyotrophy (HNA), 1670t, 1670–1671

Index Hereditary neuropathy with liability to pressure palsies (HNPP), 1140, 1659 age at onset of, 1667 carpal tunnel syndrome and, 1165, 1448, 1668 electrophysiologic features of, 1668 fluorescent in situ hybridization in, 1552f in children, 2725, 2736 inheritance of, 1667 lumbosacral plexus lesions and, 1385–1386 management of, 1669 nonuniform demyelination in, 1167 pathology of, 823, 824f, 1668, 1668f PMP22 deficiency in, 542, 579, 1627t decreased PMP22 gene dosage, 1564–1565, 1669–1670 transgenic animal studies of, 1565–1566, 1670 behavioral effects of, 1566 electrophysiology of, 1566 histology of, 1566 sites of nerve compression in, 1667 teased fiber analysis in, 745f, 748, 750f vs. hereditary brachial plexus neuropathy, 1753, 1759–1760 vs. hereditary neuralgic amyotrophy, 1670t, 1670–1671 with chromosome 17p12 deletion, 1628, 1667–1673 case studies of, 1668–1669 clinical findings associated with, 1667t, 1667–1668 generation by unequal crossing over, 1671–1672, 1672f molecular diagnostics for detection of, 1672–1673, 1673f structure of genomic region, 1671 Hereditary sensory and autonomic neuropathy (HSAN), 1809–1839 autosomal dominant (type I), 1814–1826, 2330, 2722 ⫾ gastroesophageal reflux ⫾ cough, 1823 burning feet with subclinical HSAN I, 1825–1826 clinical features of, 1814, 1815f differential diagnosis of, 1814 distal lower limb sensory and autonomic neuropathy ⫹ acral mutilation, 1816–1819 clinical features of, 1816–1817 compound action potentials in, 1817, 1817f pathology of, 1817–1819, 1818f–1821f inheritance of, 1627t, 1814–1815 lancinating pain and restless legs with subclinical HASN I, 1826 neuropathic arthropathy with subclinical HSAN I, 1823, 1825f terminology for, 1814 with acral mutilation and peroneal weakness (RAB7 mutation), 1822–1823, 1824f with acral mutilation and peroneal weakness ⫾ lancinating pain ⫾ hearing loss ⫾ hyperhidrosis (SPRLC1 mutation), 1819–1822 autosomal recessive, 1826–1837 type II (“HSN2” gene), 2722, 1826–1830 biochemistry of, 1830 clinical features of, 1827f, 1827–1828 course of, 1828 electrophysiology of, 1828 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Hereditary sensory and autonomic neuropathy (HSAN) (Continued) genetics and mechanism of, 1627t, 1826, 1828–1830 hallmarks of, 1826 pathology of, 1829f, 1830 treatment of, 1830 type III (IKBKAP gene), 2722, 1831–1833 autonomic dysfunction in, 1141, 1831 biochemistry of, 1833 clinical features of, 1831–1832 genetics of, 1627t, 1832 inheritance of, 1832–1833 nerve conduction studies in, 1832 pathology of, 1832 skin biopsy in, 886 type IV (TRKA gene), 886, 1833–1834, 2722 clinical features of, 1833–1834 genetics of, 1627t, 1833 type V (possibly TRKA and NGFB genes), 2722, 1834–1837 pathology of, 1834–1836, 1835f–1836f central nervous system involvement in, 1809 classification of, 1813–1814 clinical features of, 1809–1810 differential diagnosis of, 1809, 1810 algorithm for, 1810f hereditary motor and sensory neuropathy, 1723 nonmalignant inflammatory sensory polyganglionopathy, 2315–2316 electrophysiology of, 1812, 1812f genetic defects in, 1810, 1811, 1811t indifference to pain in, 1811–1813 injury avoidance in, 1811 mode of inheritance of, 1809 nerve biopsy in, 1810–1811 other manifestations of congenital HSAN, 1837–1838 acromutilating, paralyzing neuropathy with corneal ulceration in Navajo children, 1837 congenital insensitivity to pain with anhidrosis, 1837 congenital nonprogressive sensory neuropathy with tonic pupils, 1837 congenital sensory neuropathy, 1838 recessive form of, 1838 hereditary sensory neuropathy with spastic paraplegia, 1837 HSAN and neurotrophic keratitis, 1837 mechanisms and treatment of acral mutilation and arthropathy in, 1838–1839 pansensory familial neuropathy, 1838 overview of, 1809–1811 pathology of, 704 congenital failure of neuronal development, 816 variable expression of, 1809 Hereditary sensory neuropathy in children, 2708 type I compound action potential of sural nerve in, 1022, 1023f SPTLC1 mutations in, 424 type II, compound action potential of sural nerve in, 1022–1023, 1024f Hereditary spastic paraparesis, 1536 Heredopathia atactica polyneuritiformis. See Refsum’s disease.

Index Herpes simplex virus (HSV) infection, 2117–2118 Bell’s palsy and, 1228–1229, 2118 congenital, 2117 diagnosis of, 2118 genomic organization of virus, 2118 latent, 2118 lumbosacral plexus lesions due to, 1384–1385 neurologic complications of, 2117 neuropathies associated with, 2118t polyradiculopathy due to, 1331 radiculopathy and, 2118 Herpes zoster, 2119–2120 clinical features of, 2119–2120 cranial nerve involvement in, 2119–2120 epidemiology of, 2119 in HIV infection, 2139 in lymphoma patients, 2490 lumbosacral plexus lesions due to, 1384 oticus, 2120 with facial nerve palsy, 1231–1232. See also Ramsay Hunt syndrome. pathology of, 713 acute neuronal degeneration, 817–818 intracytoplasmic vacuolation, 689 postherpetic neuralgia and, 2120 risk factors for, 2119 treatment of, 2120 Herpesviruses, 2117–2123, 2118t Heteroplasmy, 1550, 1940 Heterozygote, 1546 compound, 1547 Hexacarbons, 2511–2514 metabolism of, 2512 neurotoxicity of, 2512 clinical features of, 2512–2513 effect on axonal transport, 397t, 399 electrophysiology of, 2514 morphology of, 715, 2512–2513 pathogenesis of, 2514 use and sources of exposure to, 2511–2512 n-Hexane, 715, 2511–2514 2,5-Hexanedione, 397t, 399, 686, 715, 2511–2514 Hex-LM1 (sialosyllactosaminyl paragloboside), 584 Hexosaminidase deficiency, 707–708, 1536, 1859–1860 vs. hereditary motor neuronopathy, 1606 HGF (hepatocyte growth factor), 356 HHV-5 (human herpesvirus 5). See Cytomegalovirus (CMV) infection. High-density lipoprotein (HDL), 1906 composition of, 1906 deficiency of clinical phenotypes of, 1909, 1910t familial, 1905, 1908, 1912 hereditary, 1905–1916. See also Tangier disease (TD). Hip girdle muscles, innervation of, 957t Hirschsprung’s disease, 705, 1151, 2692 absence of rectoanal inhibitory reflex in, 283 genetics of, 1546 Histamine effect on endoneurial capillary permeability, 652 in enteric nervous system, 265–266, 271 receptors for, 182 Histamine test for brachial plexus lesions, 1522 HIV infection. See Human immunodeficiency virus (HIV) infection. HIV Overview of Problems–Evaluation System (HOPES), 1059

HIV/AIDS-Targeted Quality of Life (HAT/QoL), 1059 HL (Hodgkin lymphoma). See Lymphoma. HLA. See Human leukocyte antigens (HLA). HMSN. See Hereditary motor and sensory neuropathy (HMSN). HNA (hereditary neuralgic amyotrophy), 1670t, 1670–1671 HNPP. See Hereditary neuropathy with liability to pressure palsies (HNPP). Hodgkin lymphoma (HL). See Lymphoma. Hodgkin-Huxley model of nerve excitation, 95 Hoffmann’s sign, in amyloidosis, 2433 Home modifications, 2625 Homocysteine (Hcy), vitamin B12 deficiency and, 2053, 2054 Homoplasmy, 1940 Homovanillic acid, effect of dopamine␤-hydroxylase deficiency on urinary level of, 1118 Homozygote, 1546 HOPES (HIV Overview of Problems–Evaluation System), 1059 Hopkins syndrome, 2031 Horizontal gaze palsy, congenital, 1198 Hormone replacement therapy, carpal tunnel syndrome and, 1445 Horner’s syndrome, 204–208, 1150, 1197 clinical signs of, 204–207 conjunctival and lid hyperemia in, 207 dilation lag in, 205–207, 207f etiology of, 204 facial anhidrosis in, 207 in Pancoast’s syndrome, 1362 localization of lesion causing, 204–205 miosis in, 205 perinatal brachial plexopathy and, 2731 ptosis in, 205 pupil pharmacology as related to autonomic denervation in, 207 sweat patterns in, 204–205 sympathetic pharmacology of, 207–208 botulinum toxin, 208 cocaine, 207–208 epinephrine, 208 hydroxyamphetamine, 208 phenylephrine, 208 tyramine, 208 HPT. See Heat-pain threshold (HPT). HR. See Heart rate (HR). HSAN. See Hereditary sensory and autonomic neuropathy (HSAN). HSCT (hematopoietic stem cell transplantation). See also Bone marrow transplantation. for adrenoleukodystrophy, 1868 for amyloidosis, 2439 “HSN2” gene, in hereditary sensory and autonomic neuropathy type II, 1810, 1811t, 1826–1830 HSP. See Henoch-Schönlein purpura (HSP). HSPs (heat shock proteins), 690 HSV. See Herpes simplex virus (HSV) infection. 5-HT (5-hydroxytryptamine). See Serotonin (5-HT). HTLV-1. See Human T-lymphotropic virus type 1 (HTLV-1). Hu antigens, 584–585, 585f, 594 Human genome, 1545 Human Genome Project, 95, 1184, 1551–1553 Human herpesvirus 5 (HHV-5). See Cytomegalovirus (CMV) infection. Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xli Human immunodeficiency virus (HIV) infection, 2129–2140 autonomic neuropathy in, 2138 carpal tunnel syndrome and, 1447 central nervous system complications of, 2129, 2130 childhood Guillain-Barré syndrome and, 2717 facial nerve paralysis and, 1234 gastrointestinal symptoms of, 2699 inflammatory demyelinating polyneuropathy in, 2136–2137, 2224, 2229 clinical features of, 2136–2137 etiology of, 2137 pathology of, 2137 stage of disease and, 2129 treatment of, 2137 lower motor neuron disease in, 1299, 1303 multiple mononeuropathy in, 2137–2138, 2350 clinical features of, 2137–2138 etiology of, 2138 pathology of, 2138 treatment of, 2138, 2375t, 2378 neuropathy in diffuse infiltrative lymphomatosis syndrome of, 2138–2139 opportunistic infections and, 2139–2140 Cryptococcus neoformans, 2140 cytomegalovirus, 2130, 2132, 2139 herpes zoster, 2119–2120, 2139 Mycobacterium avium-intracellulare, 2139 pathology of, 713–714 epidermal nerve fiber loss, 870, 885 neuropathic pain and, 870 peripheral neuropathies in, 2129, 2130t animal models of, 2133 associated with HIV-1 seroconversion, 2136 disease-stage specificity of, 2129 epidemiology of, 2130–2131 incidence and prevalence of, 1313, 2129, 2130–2131, 2131f pathogenesis of, 2129 quality-of-life measures in, 1059 sensory polyneuropathies in, 2131–2136 clinical features of, 2131–2132 electrophysiologic findings in, 2132 pathogenesis of, 2132–2135 antiretroviral toxic neuropathy, 2133–2135, 2134f–2135f distal symmetrical polyneuropathy, 2132–2133, 2133f pathology of, 2132 prevalence of, 2131, 2131f skin biopsy in diagnosis of, 885 treatment of, 2135–2136 vacuolar myelopathy in, 1313–1314 vasculitic neuropathy in, 2137–2138, 2335, 2350 treatment of, 2138, 2375t, 2378 Human leukocyte antigens (HLA) in chronic inflammatory demyelinating polyradiculoneuropathy, 2223 in myasthenia gravis, 833 in vasculitides, 2363 Human T-lymphotropic virus type 1 (HTLV-1), 713, 714, 2075 adult T-cell leukemia and, 2496 amyotrophic lateral sclerosis and, 1535 pseudo–amyotropic lateral sclerosis and, 2075 tropical spastic paraparesis and, 2063, 2066, 2071, 2076

xlii Humeral neck fractures and dislocations, brachial plexus injury from, 1365–1366 Hunter’s disease, 709 Hunter’s syndrome, 1864 Hurler’s syndrome, 1864, 1865, 2726 Hydatidosis, 2155t clinical features of, 2169 epidemiology of, 2169 etiologic agent in, 2166f, 2169 life cycle of, 2169 pathology of, 2169 peripheral neuropathy in, 2170 transmission of, 2169 Hydralazine, for nocturnal hypertension in patients with neurogenic orthostatic hypotension, 2663 Hydration in intensive care unit, 2610 Hydrogen peroxide, 509–510 Hydroxyamphetamine, Horner’s syndrome and, 208 5-Hydroxytryptamine (5-HT). See Serotonin (5-HT). Hyoglossus muscle, innervation of, 1286 Hypalgesia, 12 Hyperalgesia, 1146 capsaicin-induced, 870 in nonmalignant inflammatory sensory polyganglionopathy, 2311 in small fiber neuropathy, 2322 mechanism of, 2639 NMDA antagonists for, 527 reduced substance P and, 527 secondary, 2639 Hyperbaric oxygen, protective effects against ischemia-reperfusion injury, 672, 676 Hypercalcemia, 2629 infantile, facial nerve paralysis and, 1236 Hypercapnia, central chemoreceptor activation and, 327–328 Hyperemia, conjunctival, 207 Hyperesthesia, 1146 in diabetic neuropathy, 1953 management of, 2626 Hyperglycemia auto-oxidative lipid peroxidation and, 512 diabetic neuropathy and, 1952–1953, 1969, 1973, 1974t, 2327 effect on nerve blood flow, 672–673 gastroparesis and, 1959 Hyperoxaluria, 2722 Hyperpathia, 1146 Hyperpigmentation, 1168 in POEMS syndrome, 2455, 2457f, 2458 Hyperpnea, chemoreceptor activation and, 327–328 Hyperpolarization, 125, 191 activity-dependent, 125 effect of, 1007, 1007f–1008f in multifocal motor neuropathy, 123, 123f Hypersensitivity, 1066–1067, 1089 Hypersensitivity vasculitis, 2337, 2354t, 2361–2362 classification of, 2340t diagnostic criteria for, 2338 drug-induced, 2351 in lymphoproliferative disorders, 2351 Hypertension cryptogenic sensory polyneuropathy and, 2324 due to glossopharyngeal nerve block, 325, 326f, 1274, 1275f hemifacial spasm and, 1240 induced by sympathomimetic agents, 2676

Index Hypertension (Continued) intensive care unit management of, 2611 muscle sympathetic nerve activity in, 333 supine, in patients with neurogenic orthostatic hypotension, 2663, 2674 Hyperthyroid neuropathy, 2043–2044 Hyperthyroidism, carpal tunnel syndrome and, 1445 Hypertrichosis, in POEMS syndrome, 2455 Hypertrophic nerves, 1152 biopsy of, 737 in acromegaly, 2045 in children, 2710 in hereditary motor and sensory neuropathies, 1152, 1636, 1636f, 1641 in leprosy, 1152, 2097–2098, 2100 Hypertropia, 1196–1197 Hyperventilation, heart rate response to, 1109 Hypesthesia, 12 Hypoalgesia, 1146 Hypoalphalipoproteinemia, familial, 1905 Hypobetalipoproteinemia childhood, 2722 infantile, 2715 Hypocalcemia, 2616, 2616t Hypoesthesia, 1146 Hypogastric nerve, 31, 234–235, 281, 281f, 301 Hypogastric plexus inferior, 18, 31, 281, 281f superior, 30, 31, 281, 281f, 302 Hypoglossal canal, 1287 Hypoglossal nerve, 18, 1286–1288 anatomy of, 1286–1287 course of, 1286–1287 dural branches of, 1287 lesions of, 1287 clinical features of, 1287 glossodynia, 1288 muscles innervated by, 1286 Hypoglycemic neuropathy, 1965 Hypoglycemic unawareness, in diabetic patients, 1960 Hypokalemia, 2616t Hypomagnesemia, 2616 Hypomyelination, congenital, 1688 P0 mutations in, 1688, 1690t–1691t Hyponatremia, 2616, 2616t in Guillain-Barré syndrome, 2200 Hyposensitivity, 1066–1067, 1089 Hypotension after carotid endarterectomy, 1277 drug-induced, 2674 due to cardiac vagal overactivity, 1284 in carotid sinus hypersensitivity, 1277 in glossopharyngeal neuralgia, 1276, 1276f intensive care unit management of, 2611 neurogenic orthostatic, 1150, 2653–2663, 2671–2681. See also Orthostatic hypotension (OH), neurogenic. Hypothermia, protective effects against ischemia-reperfusion injury, 672, 676 Hypothyroidism, 2039–2042 carpal tunnel syndrome in, 1445, 2039–2040 clinical features of, 2039–2040 diagnosis of, 2040 incidence of, 2039 pathology of, 2040 treatment of, 2040 cranial nerve involvement in, 2040 deafness in, 2040 in POEMS syndrome, 2458 neurologic complications of, 2039 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Hypothyroidism (Continued) polyneuropathy in, 2040–2042 clinical features of, 2040–2041 morphology of, 2041–2042, 2043f–2044f nerve conduction studies in, 2041, 2041f–2042f pathology of, 2042 treatment of, 2042 Hypovolemia orthostatic hypotension and, 2656, 2658 peripheral nerve adaptation to, 668 Hypoxemia, 2615 Hypoxia. See also Ischemia. in chronic obstructive pulmonary disease, 2021–2022 in diabetic neuropathy, 512, 513f, 1970 mechanisms of injury in peripheral nerve disease, 2022 muscle sympathetic nerve activity responses to, 327 peripheral nerve adaptation to, 668 Hysterical symptoms, 1153

IAC (internal auditory canal) eighth nerve in, 1253, 1255f osseous lesions of, 1264 temporal bone metastases in, 1268 Iatropathic nerve injury, 1512, 1516–1521 causes of, 1517 during total hip arthroplasty, 1519–1520, 1520f from compression by hematoma, 1518 from drug injection, 1517 from false aneurysm and arteriovenous fistula, 1518, 1519f from interscalene block, 1517 in field of operation, 1518–1519, 1520f ischemic, 1517–1518, 1518f malpractice claims related to, 1516 to small nerves, 1521 IB4 (isolectin B4), 173f, 176t, 179–180 IBPN. See Brachial plexus neuropathy, immune (IBPN). IBS (irritable bowel syndrome) after acute infectious enteritis, 256 visceral hypersensitivity in, 254–255 Ibuprofen, for orthostatic hypotension, 2661, 2663 ICAM-1. See Intercellular adhesion molecule-1 (ICAM-1). ICAM-2 (intercellular adhesion molecule-2), 2342 I-cell disease, 1864–1865 pathology of, 709 ICU. See Intensive care unit (ICU). Id proteins in oligodendrocyte differentiation, 354 role in myelination, 361 IDPN (␤,␤’-iminodipropionitrile) effect on axonal transport, 392, 394, 397t, 398–399, 441f, 441–442 neuropathology induced by, 1296–1297 IENFs (intraepidermal nerve fibers), 2325–2326 IFN-␥ (interferon-␥) in experimental autoimmune neuritis, 617, 618f–619f in wallerian degeneration, 1415, 1417f IFs (intermediate filaments), 434 classes of, 434 protein composition of, 434 IgA. See Immunoglobulin A (IgA).

Index IgD (immunoglobulin D), 563 monoclonal, 2256 IgE (immunoglobulin E), 563 monoclonal, 2256 IgG. See Immunoglobulin G (IgG). IGHMBP2 gene mutations, 1770t, 1780–1781 IgM. See Immunoglobulin M (IgM). IHCs. See Inner hair cells (IHCs). IKBKAP gene, in familial dysautonomia, 886, 1810, 1811t, 1831–1833 Ileus adynamic, 2610, 2616 after spinal cord injury, 2699 Iliac artery aneusysm, lumbosacral plexus compression by, 1381, 1382f Iliacus muscle, hematoma of, 1378, 1379f Iliococcygeus muscle, 280, 280f Iliohypogastric nerve distribution of, 16f, 26, 1488, 1488f structures innervated by, 1376 Iliohypogastric neuropathy, 1488 Ilioinguinal nerve distribution of, 16f, 26, 1488, 1488f structures innervated by, 1376 Ilioinguinal neuropathy, 1488 Iliolumbar artery, 1376 Imaging studies, 1174. See also specific imaging modalities. 123 I-meta-iodobenzylguanadine (MIBG) single-photon emission computed tomography of sympathetic innervation to organs, 1119–1121, 1120f ␤,␤’-Iminodipropionitrile (IDPN) effect on axonal transport, 392, 394, 397t, 398–399, 441f, 441–442 neuropathology induced by, 1296–1297 Imipramine for detrusor hyperreflexia, 2663 for diabetic neuropathy, 2642, 1971, 1971t Immobility, complications of, 2616, 2628–2629 Immune brachial plexus neuropathy. See Brachial plexus neuropathy, immune (IBPN). Immune response, 559–569, 573–604, 609–610 adaptive immune system, 574, 609–610 cell surface receptors in, 574–575, 609 immunocompetent cellular components, 559–567 B lymphocytes, 563–564, 566f macrophages, 559–560 mast cells, 564–565 Schwann cells, 565–567 T lymphocytes, 560–563, 561f–564f, 565t immunologic privilege, 610 in chronic inflammatory demyelinating polyradiculoneuropathy, 2243–2246 in experimental autoimmune neuritis, 615–620, 623, 624f in leprosy, 2084 in myasthenia gravis, 833 in paraneoplastic neuropathy, 2481–2482 in peripheral nervous system, 559–569, 560f, 573–575 in vasculitis, 2335, 2344–2348 inflammatory mediators and, 181–183 innate immune system, 574–575, 609 peripheral nerve antigens and, 573–598. See also Antigens. termination of, 569

Immune response (Continued) tolerance and autoimmunity, 567–569, 574, 575 B-cell tolerance, 567 breakdown of tolerance, 568–569 T-cell tolerance, 567, 568f Immune surveillance, 559, 610 Immune-mediated polyneuropathies, 2177–2190 antibody testing in, 2177–2182 autoantibodies associated with, 2177–2190, 2178t–2180t. See also Antibodies. classification of, 2177 targets of pathologic processes in, 2177 Immunocompromised persons cytomegalovirus infection in, 2121 herpes zoster in, 2119–2120 human immunodeficiency virus infection in, 2129–2140 parasitic infections in, 2153 Immunocytochemistry, 3, 2180 Immunofixation electrophoresis, 2181 to identify monoclonal proteins, 2257, 2257f–2258f Immunoglobulin A (IgA), 563 cryoglobulins, 2257 in Henoch-Schönlein purpura, 2362 in Waldenström’s macroglobulinemia, 2271 monoclonal, neuropathy and, 2256, 2261–2264, 2268 Immunoglobulin D (IgD), 563 monoclonal, 2256 Immunoglobulin E (IgE), 563 monoclonal, 2256 Immunoglobulin G (IgG), 563 binding to galactocerebroside, 2190 binding to GD1b ganglioside, 2187 binding to GM1b ganglioside, 2186 binding to GM2 and GalNac-GD1a gangliosides, 2186 binding to GQ1b ganglioside, 2185 binding to GT1a ganglioside, 2178t, 2185 binding to heparan sulfate, 2190 cryoglobulins, 2257 IgG anti-GM1 antibodies, 2184–2185 antibody testing for, 2184 treatment of motor neuropathies based on, 2184 disease pathogenesis and, 2184–2185 in acute motor axonal neuropathy, 2184–2185 in Waldenström’s macroglobulinemia, 2271 monoclonal, neuropathy and, 2256, 2261–2264, 2268 Immunoglobulin M (IgM), 563 binding to ␤-tubulin, 2190 binding to galactocerebroside, 2190 binding to GD1b ganglioside, 2186 binding to GM2 and GalNac-GD1a gangliosides, 2186 binding to heparan sulfate, 2190 binding to myelin-associated glycoprotein, 2187–2188 binding to neurofilaments, 2190 binding to sulfated glucuronyl paragloboside, 2188 binding to sulfatide, 2189, 2268 in lipid membrane environment, 2189 cryoglobulins, 2257 IgM anti-GM1 antibodies, 2182–2184 antibody testing for, 2182 treatment of motor neuropathies based on, 2182–2184 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xliii Immunoglobulin M (IgM) (Continued) disease pathogenesis and, 2184 in multifocal motor neuropathy, 2182 in Waldenström’s macroglobulinemia, 2271 monoclonal gammopathy and neuropathy, 1145, 1173, 2255, 2256, 2261–2264, 2266–2268 paraproteinemic neuropathy, 62, 2315t Immunoglobulin superfamily of adhesion molecules, 1684, 2342 Immunoglobulins, 563–564, 616 Immunohistochemistry, in neuron death, 700f, 700–701 Immunologic privilege, 610 Immunomodulatory therapies, 635, 641–644 intravenous immunoglobulin, 642–644 plasma exchange, 641–642 Immunophilin binding agents, 638–639 adverse effects of, 639 mechanism of action of, 638–639 use in neuropathy, 639 Immunosuppressive agents, 635–641 alkylating agents, 637–638 azathioprine, 639–640 for radiculoplexus neuropathies, 2011–2012, 2013f for vasculitic neuropathy, 2375t–2376t, 2378–2381 limiting adverse effects of, 2381–2383, 2382t–2384t glucocorticoids, 636–637 immunophilin binding agents, 638–639 mycophenolate mofetil, 640–641 Immunotherapy, 635–644 agents used for, 635–644 immunomodulatory agents, 635, 641–644 immunosuppressive agents, 635–641 determining therapeutic effect of, 636 for chronic inflammatory demyelinating polyradiculoneuropathy, 635, 2179t, 2230, 2234–2239 in children, 2231, 2719 in diabetic patients, 1955–1956 for cryptogenic sensory polyneuropathy, 2326–2327 for diabetic lumbosacral radiculoplexus neuropathy, 1963, 2011–2012, 2013f for Guillain-Barré syndrome, 635, 2199, 2213 in children, 2716 in diabetic patients, 1955 for leprosy, 637, 2084, 2104 for Lewis-Sumner syndrome, 2232–2233 for paraneoplastic neuropathy, 2482 strategies for, 635–636 IMP (intramuscular pressure) measurement, 1097–1099, 1098f–1099f, 1099t Impairment definition of, 1033, 1054 quantitation of, 1038–1041 Impar ganglion, 234 Impotence, 1151, 2628, 2664 Imprinting, 1550–1551 IMPs (intramembranous particles), of axolemma, 47, 55 In situ hybridization, 3 Inborn errors of metabolism Fabry’s disease, 1893–1901 hereditary amyloid neuropathy, 1921–1932 in children, 2723 lysosomal disorders, 1845–1865 mitochondrial respiratory chain dysfunction, 1937–1947 pediatric polyneuropathy and, 2721

xliv Inborn errors of metabolism (Continued) peroxisomal disorders, 1845, 1865–1873 porphyric neuropathy, 1883–1891 Tangier disease, 1905–1916 vs. childhood Guillain-Barré syndrome, 2717 Inching technique, 934, 2283 Incidence of peripheral neuropathy, 1181–1182, 1182t Incontinence fecal, 287–292, 2627–2628, 2665 clinical features of, 288 denervation and, 287–288 diagnostic tests for, 288–292, 290t disorders associated with, 2627 management of, 292–293, 293t, 2628 urinary, 2626–2627 disorders associated with, 2626–2627 management of, 2627, 2663–2664 overflow, 2627 Indifference to pain, 1811–1813 Indomethacin, for orthostatic hypotension, 2661, 2663, 2677 Industrial agents, 2505–2520 acrylamide, 2506–2508 allyl chloride, 2508 carbamates, 2508–2510 carbon disulfide, 2510–2511 dichloracetylene, 2520 ethylene oxide, 2511 hexacarbons, 2511–2514 methyl bromide, 2514–2515 organophosphates, 2515–2519 polychlorinated biphenyls, 2519 pyriminil, 2519 trichloroethylene, 2520 Infants. See Pediatric patients. Infection(s) acute inflammatory demyelinating polyneuropathy and, 592, 2200t, 2223 antiganglioside antibodies and, 590–591 carpal tunnel syndrome due to, 1447 chronic inflammatory demyelinating polyradiculoneuropathy and, 2223–2224 diagnostic investigation for, 1172 diphtheria, 2147–2150 facial nerve paralysis due to, 1235–1236 Bell’s palsy, 1228–1230 in children, 1236–1237 Guillain-Barré syndrome and, 590–591, 1187, 2200t, 2206–2207 herpesvirus, 2117–2123 HIV, 2129–2140 human T-lymphocyte virus type 1, 713, 714, 1535, 2063, 2066, 2071, 2075, 2496 immune brachial plexus neuropathy and, 2299, 2301, 2302t leprosy, 2081–2105 lumbosacral plexus lesions due to, 1384–1385, 1385f Lyme disease, 2109–2115 microbial antigens and, 590–591 nosocomial, 2615–2616, 2616t ocular palsies due to, 1202–1203 parasitic, 2153–2174 pediatric polyneuropathy and, 2719–2720 polyradiculopathy due to, 1331–1332 tropical myeloneuropathies and, 2069t vasculitis due to, 2335, 2340t, 2348–2351 Infectious mononucleosis, hypoglossal nerve palsy and, 1287

Index Inflammation brain-derived growth factor upregulation in, 177 effect on glutamate receptors, 525 electrophysiologic properties of dorsal root ganglion neurons in chronic pain models with, 171 glucocorticoid effects on, 636 glutamate in, 184 hereditary brachial plexus neuropathy and, 1753, 1764 in acute inflammatory demyelinating polyneuropathy, 2202 in cryptogenic sensory polyneuropathy, 2325, 2325f in diabetic lumbosacral radiculoplexus neuropathy, 2005 ischemia and, 522, 524f leukemia inhibitory factor upregulation in, 181 lumbosacral plexopathy with, 1386 nerve growth factor upregulation in, 176, 177 nerve injury and, 1413–1414 neurokinin-1 receptors and, 185 neurotrophin-4 upregulation in, 178 pro-inflammatory mediators, 181–183 ATP receptors, 182 bradykinin receptors, 182 endothelin receptors, 183 histamine receptors, 182 nitric oxide, 183 peptides, 183 prostaglandins and COX enzymes, 182 serotonin, 183 reactive oxygen species and, 514 tumor necrosis factor-␣ upregulation in, 181 vasculitis, 2335–2385 Inflammatory bowel disease, 1167 Inflammatory demyelinating polyneuropathy acute, 2197–2213. See also Guillain-Barré syndrome (GBS). chronic, 2221–2246. See also Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP). pathology of, 806–808, 806f–808f Inflammatory Neuropathy Overall Disability Sum Score (ODSS), 1044–1045, 1047t Inflammatory Neuropathy Sensory Sum Score, 1037, 1037t, 1049 Inflammatory sensory polyganglionopathy malignant, 2315t nonmalignant, 2309–2317 clinical patterns of, 2311 course of, 2311–2312 differential diagnosis of, 2314–2317, 2315t electrophysiologic findings in, 2312–2313 historical review of, 2310–2311 idiopathic, 2315t laboratory findings in, 2312 pathology of, 2313–2314, 2314f radiologic features of, 2312 Sjögren syndrome and, 2315t symptoms and signs of, 2311 treatment of, 2317 pathology of, 772, 778f Infraspinatus muscle, 1466–1468 Inguinal ligament, femoral neuropathy at, 1490 INH (isoniazid) neuropathy, 1296, 2559–2560 Inheritance patterns, 1545–1551, 1595 imprinting, 1550–1551 Mendelian, 1545–1549, 1595 mitochondrial, 1549–1550, 1938–1940 multifactorial, 1551 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Inhibitory postsynaptic potentials (IPSPs), slow, in enteric nervous system, 269t, 269–270 Injection injury of nerve, 1517 Innate immune system, 574–575, 609 Inner hair cells (IHCs), 1253 carboplatin toxicity to, 1268 in auditory neuropathy, 1261, 1262f noise-induced damage of, 1268 testing function of, 1256 Insecticides carbamates, 2508–2510 organophosphates, 2515–2519 Insulin neuritis induced by, 1954–1955 orthostatic hypotension induced by, 1958 Insulin-like growth factor I (IGF-I), 378, 380–381 clinical trials of, 380–381 epidermal nerve fibers and, 872 in acromegaly, 2045 in neuronal growth and development, 380 receptor for, 380 Insulin-like growth factor II (IGF-II), 378, 380 in Schwann cell survival, 355 Integrins, 616, 2342 as laminin receptors, in myelination, 424–425 expression in promyelinating vs. myelinating Schwann cells, 424 in Schwann cell migration, 364 mice deficient in ␤1 and ␤4 integrins, 546 signaling in Schwann cells, 363–364 Intensive care unit (ICU), 2607–2617 criteria for admission to, 2608, 2608t general aspects of care in, 2608–2611 deep venous thrombosis prophylaxis, 2609 gastrointestinal and urinary tract complications, 2609–2610 nutrition and hydration, 2610–2611 skin, eye, and mouth care, 2609 history of, 2607 management of Guillain-Barré syndrome in, 2212–2213, 2607–2609, 2609t management of respiratory failure in, 2607–2608, 2611–2615 effects of acute neuromuscular disease on respiration, 2611–2612, 2612f intubation and mechanical ventilation, 2613–2614, 2613t–2614t monitoring pulmonary function, 2612f, 2612–2613 principles of respiratory physiology, 2611 tracheostomy, 2614–2615 weaning from ventilation, 2614–2615 withdrawal of respiratory support and palliation, 2617 management of systemic complications in, 2615–2616, 2616t reasons for management of acute neuromuscular diseases in, 2607–2608 Intercellular adhesion molecule-1 (ICAM-1), 2342 in experimental autoimmune neuritis, 616, 625 in ischemic-reperfusion oxidative injury, 522, 524f, 677 in vasculitis, 2343 Intercellular adhesion molecule-2 (ICAM-2), 2342 Intercostal arteries, 1341

Index Intercostal nerve(s) first, 24 second, 24 seventh to eleventh, 24, 24f third, 24 typical, 23–24, 24f Intercostobrachial nerve, 15f Interdigital neuropathies, 1506–1507 Joplin’s neuroma, 1506–1507 Morton’s neuroma, 1507 Interferon-␣ for hepatitis B–associated polyarteritis nodosa, 2336, 2376 for hepatitis C–associated cryoglobulinemic vasculitis, 2020, 2336, 2376–2378 for leprosy, 2104 Interferon-␤, for chronic inflammatory demyelinating polyradiculoneuropathy, 2236 Interferon-␥, 565t, 2342, 2344 actions of, 565t in experimental autoimmune neuritis, 617, 618f–619f in wallerian degeneration, 1415, 1417f Interleukin-1, 560, 565t, 2342 actions of, 565t Schwann cell production of, 566 Interleukin-1␤ expression in dorsal root ganglion neurons, 181 in enteric nervous system, 265–266, 271 in ischemia-reperfusion injury, 522, 525f, 675, 677 in POEMS syndrome, 2454–2455 in wallerian degeneration, 1415, 1416f Interleukin-2, 565t Interleukin-4, 565t, 2344 Interleukin-5, 565t Interleukin-6, 180, 560 expression after axotomy, 180 expression in dorsal root ganglion neurons, 180 family of, 180 gp130 signaling and, 180 in enteric nervous system, 271 in ischemic-reperfusion oxidative injury, 523 in POEMS syndrome, 2454 in wallerian degeneration, 1415, 1416f receptor for, 180 Schwann cell production of, 566 synthesis of, 180 Interleukin-8, 2342, 2344 Interleukin-10, 560, 565t, 2344 actions of, 565t in experimental autoimmune neuritis, 618 in ischemic-reperfusion oxidative injury, 523 in wallerian degeneration, 1415, 1416f Interleukin-12, 565t in experimental autoimmune neuritis, 618 in wallerian degeneration, 1415, 1417f Interleukin-13, 565t, 2344 Interleukin-18, 565t in experimental autoimmune neuritis, 617 in wallerian degeneration, 1415, 1417f Interleukin-21, 565t Interleukin-23, 565t Intermediate filaments (IFs), 434 classes of, 434 protein composition of, 434 Intermesenteric plexus, 30, 31

Internal anal sphincter, 279, 283–284, 2627 adrenergic innervation of, 284–285 during micturition, 284 innervation of, 279 neurotransmitter effects on, 286t pharmacology of, 284–285, 286t tone of, 283 Internal auditory canal (IAC) eighth nerve in, 1253, 1255f osseous lesions of, 1264 temporal bone metastases in, 1268 Internal consistency reliability of health outcome measures, 1056 Internal iliac artery, 1376 Internal jugular vein catheterization, brachial plexus injury from, 1365 Internal urethral sphincter, 300 International Classification of Impairments, Disabilities and Handicaps, 1054 Interneurons of enteric nervous system, 251, 251f, 253, 256–257 spinal, in lower urinary tract control, 308–309, 309f Internode(s), 42, 413f, 788 average diameter of myelinated fibers related to shape of sample sections of, 758–759, 761f axonal caliber and length of, 414 conduction velocity and diameter of fiber at, 1024–1025, 1027f end regions of, 42, 43f establishment of, 411–415 in hereditary motor and sensory neuropathy, 1640, 1644f, 1644t main internodal region of, 42, 43f–44f Schwann cells of, 788 teased fiber analysis of length and diameter of, 752–753, 753t Interosseous nerve of forearm anterior, 20f anterior interosseous nerve syndrome, 1453–1454 muscles innervated by, 956t pseudo anterior interosseous nerve syndrome, 1454 posterior, 1474 lesions of, 1475–1478 clinical manifestations of, 1476 diagnosis of, 1476–1478 etiologies of, 1475–1476 treatment of, 1478 muscles innervated by, 956t Interosseous nerve of leg, 29 Interscalene block, nerve injury during, 1517 Interscalene nerve block, brachial plexus lesion after, 1365 Interstitial cells of Cajal, 2683 Intervertebral disc herniation lumbosacral, presenting as cauda equina syndrome, 1328–1329, 1329f radiculopathy due to, 1325–1326 Intervertebral foramina, 1323 effects of spondylosis on, 1328 Intestinal motility testing, 1280 Intestinal pseudo-obstruction, 259–260, 2688–2691 autonomic neuropathies with, 2688 chronic, 2665 clinical assessment of, 2688 definition of, 2688 diagnostic evaluation of, 2688–2689, 2688f–2690f Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xlv Intestinal pseudo-obstruction (Continued) etiology of, 2688 idiopathic, 260 in Chagas’ disease, 260 in inflammatory neuropathies, 260 in noninflammatory neuropathies, 260 management of, 2689–2691, 2690t myopathic vs. neuropathic, 259–260 paraneoplastic syndrome, 260 Intraepidermal nerve fibers (IENFs), 2325–2326 Intrafusal fibers, 132, 137–138 bag fibers, 137, 138 chain fibers, 137, 138 contractile poles of, 137 contraction of, 138 difference from extrafusal fibers, 137 of cat soleus, 132–135 Intramembranous particles (IMPs), of axolemma, 47, 55 Intramuscular pressure (IMP) measurement, 1097–1099, 1098f–1099f, 1099t Intraocular muscles innervation of, 1192 parasympathetic innervation of, 203–204, 206f sympathetic innervation of, 203, 204f–206f Intravenous immunoglobulin (IVIg), 642–644 administration of, 642 adverse effects of, 644, 644t dosage of, 642 for acute and subacute neuropathies, 2178t for acute motor axonal neuropathy, 2184 for chronic inflammatory demyelinating polyradiculoneuropathy, 643–644, 2179t, 2230, 2235, 2237–2239 in children, 2231, 2719 in diabetic patients, 1955 for dermatomyositis, 643 for GALOP syndrome, 2189 for Guillain-Barré syndrome, 643, 2199, 2213 childhood, 2716 in diabetic patients, 1955 infantile, 2713 for immune brachial plexus neuropathy, 2306 for Kawasaki disease, 643 for Lewis-Sumner syndrome, 2233 for lumbosacral radiculoplexus neuropathy, 2012 diabetic, 2011–2012 for monoclonal gammopathy and peripheral neuropathy, 644, 2269, 2269t for multifocal motor neuropathy and conduction block, 644, 2182, 2290–2294, 2292f–2293f for paraneoplastic neuropathy, 2482 for small fiber sensory neuropathy, 2327, 2479 for thrombocytopenia, 642, 643 for vasculitic neuropathy, 644, 2381 for West Nile virus infection, 1302 indications for, 643–644 informed consent for use of, 642 mechanisms of action of, 642–643, 643t preparation of, 642 Intubation, endotracheal, 2613–2614 Ions and ion channels acid-sensing ion channels, 187t, 188 axotomy-induced alterations of, 694–695 calcium, 191 cloning of, 95 functions of, 95

xlvi Ions and ion channels (Continued) genes for, 95 neuropathy linked to mutations of, 95 in action potential, 95 in axonal transport, 394 in mechanotransduction, 188 in neuromuscular transmission, 832 patch-clamp recording of, 95 potassium, 99–108, 189–191. See also Potassium channels. sodium, 96–99, 189. See also Sodium channels. voltage-gated, 95, 590 IPSPs (inhibitory postsynaptic potentials), slow, in enteric nervous system, 269t, 269–270 Iris, 203 dilator muscle of, 203 effects of damage on light, near, and drug responses of, 209 innervation of, 203–204 Horner’s syndrome and, 204–208, 207f parasympathetic, 203–204, 206f sympathetic, 203, 204f–206f testing integrity of, 213 movement of, 209 sphincter muscle of, 203, 204f differential diagnosis of dysfunction of, 210 Irritable bowel syndrome (IBS) after acute infectious enteritis, 256 visceral hypersensitivity in, 254–255 Ischemia, 800–805 age-related resistance to, 671 blood supply to nerves and, 800–801, 803f demyelination due to, 522 effect on axonal transport, 393 effects on axons, 119–121, 121f, 125–126, 804, 805f experimental nerve ischemia, 671–676 factors that modulate effects of, 672, 672f in diabetes, 512, 513f, 672–674, 673f, 673t excessive susceptibility to, 673–674 models of, 671–672 reperfusion injury and, 521, 674f, 674–676 severity of ischemic fiber degeneration in, 672, 672f temperature effects on, 672 in compression neuropathy, 1396–1399, 1397f–1398f in diabetic lumbosacral radiculoplexus neuropathy, 2004–2005 in diabetic neuropathy, 1970 resistance to ischemic conduction failure, 1979–1980 lumbosacral plexus lesions due to, 1380 nerve injury due to, 1517, 1518f cellular vulnerability to, 804, 805f microvasculitis and, 2413f spatial distribution of, 802–803 neuroprotection against ischemiareperfusion injury, 524, 526f, 676 of nerves compressed by swollen muscles, 1517–1518 oxidative stress and, 519–524 adrenergic contribution to, 520 cytokines in, 522–523, 525f, 675, 677 effects on blood-nerve barrier, 520, 676–677 endoneurial contents and, 521 footprints of, 521 genetic susceptibility to, 523–524 inflammatory response and, 522, 524f molecular pathogenesis of, 524

Index Ischemia (Continued) neuropathology of, 521–522, 523f reperfusion injury, 521, 674–676 paresthesias induced by, 121, 1141–1142 pathogenesis of nerve ischemia, 676–677 peripheral nerve adaptation to hypoxic stress, 668 vascular processes causing, 801–802 capillary and precapillary occlusion, 804–805 peripheral vascular disease, 802 vulnerability of lumbosacral nerve roots to, 1324 Isochronism, 114 Isokinetic dynamometer, 1096 Isolectin B4 (IB4), 173f, 176t, 179–180 cutaneous, muscle, and visceral expression of, 176t Isoniazid (INH) neuropathy, 1296, 2559–2560 Isoproterenol infusion test, 1124 Issacs’ syndrome, 1144 antibodies to potassium channels in, 590, 597 Itch fibers, 1006 IVIg. See Intravenous immunoglobulin (IVIg). Ixodes ticks, in Lyme disease transmission, 2109, 2110

Jackson’s syndrome, 1285 Jake leg paralysis, 2075 Jamaican myeloneuropathy, 2066 Jaw (masseter) reflex, 921, 923t Jitter bimodal, 994 in single-fiber electromyography, 986–987 JJ316 monoclonal antibody, for experimental autoimmune neuritis, 625 Joplin’s neuroma, 1506–1507 Juxtaparanodal segment, 417, 1684 microscopic anatomy of, 44f, 51, 53–55 potassium channels of, 54–55, 549–550, 577 transgenic mouse studies of, 536 transition between paranodal segment and, 51–52, 54f

KAFO (knee-ankle-foot orthosis), 2623 Kainate receptors, effect of inflammation on, 525 Kal-1, in Kallman’s syndrome, 469 Kala-azar, 2165–2167 Kallman’s syndrome, axon guidance in, 468–469 anosmin-1, heparan sulfate proteoglycans, and fibroblast growth factor receptor, 469 Kal-1 and anosmin-1, 469 Kawasaki disease classification and definition of, 2338t facial nerve paralysis and, 1236 intravenous immunoglobulin for, 643 Kearns-Sayre syndrome (KSS), 1941, 2700 Kennedy’s disease, 705, 706, 1299, 1316, 1599–1600, 1605t, 1612 clinical features of, 1612, 1612f differential diagnosis of, 1612 amyotrophic lateral sclerosis, 1536 electrophysiologic findings in, 1316, 1612 genetics of, 1316, 1536, 1548, 1600, 1612 pathology of, 1612 pattern of involvement in, 1139 sensory involvement in, 1316 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Keratinocytes, epidermal, 870–871 Keratins, 434 Keratitis amebic, 2154t, 2162 clinical features of, 2162 life cycle and morphology of etiologic agent in, 2162, 2163f pathogenesis of, 2162 transmission of, 2162 hereditary sensory and autonomic neuropathy and, 1837 in onchocerciasis, 2171 Kernig’s sign, in pediatric Guillain-Barré syndrome, 2716 Ketamine, topical, 2645 KIAA1985 gene mutations, 1770t, 1771–1773 KIF1B gene mutations, in Charcot-MarieTooth disease type 2A, 1629, 1728–1730 Kinesins in axonal transport, 388, 390–391, 391f toxic effects on, 394 Klippel-Feil syndrome, 2735 Knee-ankle-foot orthosis (KAFO), 2623 Konzo, 2074–2075 clinical features of, 2074 epidemiology of, 2073–2074 urinary test for, 2075 Krabbe’s disease, 1855–1859 clinical features of, 1856f, 1856–1857 diagnosis of, 1856, 1857 genetics of, 1858–1859 historical description of, 1855 in children, 2708, 2721 infants, 2714 nerve conduction studies in, 1856 pathogenesis of, 1858 pathology of, 710, 1857, 1857f–1859f stages of, 1856 Krogh-Erlang equation, 668, 669f Krox20. See Early growth response 2 gene (EGR2, Krox20). Krox24. See Early growth response 1 gene (EGR1, Krox24). KSS (Kearns-Sayre syndrome), 1941, 2700 Kufs’ disease, 708 Kugelberg-Welander disease, 704, 1604t, 1605t, 1607t, 1610–1611 vs. pediatric chronic inflammatory demyelinating polyradiculoneuropathy, 2718–2719 Kyphoscoliosis, 1151

L1 in CRASH syndrome, 467–468 in schizophrenia, 467 mice deficient in, 548, 549f Labial nerves, 28 Laboratory investigations, 1164, 1170–1175 cerebrospinal fluid analysis, 1174–1175 imaging studies, 1174 in pediatric polyneuropathies, 2712 monoclonal gammopathy and antiglycolipid antibodies, 1173, 1174t nerve biopsy, 1173–1174 screening tests, 1170–1173 Lacrimal glands, innervation of, 1222 Lacrimation, paroxysmal, 1275–1276 Lactic acidosis, in Leigh disease, 2715 Lama2 gene, 424

Index Lambert-Eaton myasthenic syndrome (LEMS), 590, 834–838 autoantibodies to calcium channels in, 590, 597–598, 834, 1200 clinical features of, 834, 837 congenital myasthenic syndrome resembling, 842 therapy for, 860 differential diagnosis of, 1152 pediatric chronic inflammatory demyelinating polyradiculoneuropathy, 2719 electromyography in, 837 extraocular muscle paralysis in, 1200 non-neoplastic, 837 pathogenesis and pathology of, 837–838, 838f single-fiber electromyography during axonal stimulation in, 995–996 small-cell lung carcinoma and, 597, 834, 1200 therapy for, 860 Lamellipodium(a), 448, 448f Lamina densa, 38 Lamina lucida, 38 Laminin, 4, 377 age-related changes in, 427–428 in Schwann cell development, 362–363 isoforms of, 363 laminin 2, 424, 576 Laminin receptors dystroglycan complexes, 424–425 integrins, 424–425 Lamivudine, for hepatitis B–associated polyarteritis nodosa, 2336, 2376 Lamotrigine adverse effects of, 2643 for diabetic neuropathy, 1971 for HIV-associated sensory polyneuropathies, 2135, 2643 for neuropathic pain, 2643 for trigeminal neuralgia, 2643 interaction with ritonavir, 2135 Langerhans cells, epidermal, 870 Lapique’s equation, 116 Large fiber sensory loss, 1147f, 1147–1148, 1165–1166 Laryngeal nerve recurrent, 1278 lesions of, 1279, 1285 superior, 1278 lesions of, 1279, 1285 Laryngeal paralysis, 1279 Laryngeal receptors, vagal, 1283–1284 Lasègue’s sign, 1377 Late subexcitability, 118 Lathyrism, 2063, 2068, 2071, 2069t, 2075 Laxatives, 2665, 2692, 2693t LCAT (lecithin:cholesterol acyltransferase), 1906 familial deficiency of, 1909, 1910t LCHAD (long-chain 3-hydroxyacyl coenzyme A dehydrogenase) deficiency, 2715 LD (Leigh disease), 1941, 1944–1945, 1945f, 2715, 2722 LDL (low-density lipoprotein), 1906 Lead neuropathy, 2063, 2527–2535 anemia and, 2529–2530 axonal transport in, 397t blood and urine indices for, 2528t clinical features of, 2528–2529 diagnostic criteria for, 2529 electrophysiology of, 2529 endoneurial homeostasis in, 657–658, 658f historical overview of, 2527–2528 laboratory studies in, 2529–2531

Lead neuropathy (Continued) lead lines in bones and, 2530, 2530f motor involvement in, 1141 pathology of, 715, 791, 792f, 2531–2534, 2532f–2534f Schwann cell damage, 820–823, 821f–822f sources of lead exposure, 2527–2528 treatment of, 2534–2535 Leber’s hereditary optic neuropathy (LHON), 1550, 1940, 1941 Lecithin:cholesterol acyltransferase (LCAT), 1906 familial deficiency of, 1909, 1910t Lectin injection, intraneuronal, chromatolysis induced by, 686 Lectins, 575 Leigh disease (LD), 1941, 1944–1945, 1945f, 2715, 2722 Leishmaniasis, 2154t, 2165–2167 clinical features of, 2165 cutaneous, 2165 etiologic agent in, 2165, 2166f life cycle of, 2165 mucocutaneous, 2165 peripheral neuropathy in, 2165–2167 transmission of, 2165 visceral, 2165 LEMS. See Lambert-Eaton myasthenic syndrome (LEMS). Leonine facies, in leprosy, 2084 Leprosy, 574, 2081–2105 absorption in, 2098, 2098f amyloidosis in, 2087 autonomic dysfunction in, 2093 bacterial etiology of, 2083–2084 blood-nerve exchange in, 653 borderline, 2083, 2084, 2085f, 2086 neuropathology of, 2101f, 2102 sensory loss in, 2093–2095, 2096f clinical features of, 2082 clinicopathologic spectrum of, 2083f complications of, 2086–2087 deformity from, 2087 differential diagnosis of, 2096–2098 distal polyneuropathy, 2096–2097 mononeuropathy and mononeuropathy multiplex, 2097 nerve enlargement, 2097–2098 electrophysiologic testing in, 2098–2099 epidemiology of, 2082 evaluation and classification of patients with, 2082–2084 history of, 2081–2082 indeterminate, 2083, 2083f, 2084, 2085f lepromatous, 2083, 2083f, 2084–2086, 2085f neuropathology of, 2101f, 2101–2102 sensory loss in, 2089–2093, 2090f–2095f leprosy reactions in, 2086–2087 downgrading reactions, 2086 in nerves, 2102–2103 neuritis in, 2095 treatment of, 2104 type 1 (reversal) reactions, 2086 type 2 (erythema nodosum leprosum) reactions, 2086–2087 morphologic index in, 2083 multibacillary, 2083 muscle involvement in, 2103 Mycobacterium leprae adhesion to Schwann cells, 424, 574 nerve biopsy in, 736 nerve compression in, 2096 nerve conduction studies in, 2098–2099 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xlvii Leprosy (Continued) nerves commonly involved in, 2089f neurofibromatosis and, 2099 paralysis in, 2091 pathology of, 714, 2099–2103, 2100f–2101f granulomas, 2084, 2100, 2417 nerve enlargement, 1152, 2097–2098, 2100 paucibacillary, 2083 pediatric polyneuropathy and, 2719 sensory loss in, 1141, 1147, 2087–2095 temperature-linked pattern of, 2092–2095, 2093f, 2097 skin biopsy in, 887, 2083, 2086 socioeconomic impact of, 2087 stigma of, 2081–2082 susceptibility to, 2082 transmission of, 2082 treatment of, 2103–2105 antibacterial therapy, 2103–2104 for insensitivity, 2105 for leprosy reactions, 2104 immunotherapy, 637, 2084, 2104 social rehabilitation, 2105 surgical, 2104–2105 tuberculoid, 2083, 2083f, 2084, 2085f neuropathology of, 2100 sensory loss in, 2088–2089, 2088f–2089f Leptomeningeal metastasis, 1316, 1329–1330 vs. Guillain-Barré syndrome, 2202 Leptophos, 2515 Leukemia, 2495–2497, 2600 acute lymphoblastic, 2496–2497 acute myelogenous, 2497 chronic lymphocytic, 2495–2496 chronic myeloid, 2496 facial nerve paralysis and, 1237 natural killer cell, 2496 peripheral neuropathies related to bone marrow transplantation for, 2498, 2498t T-cell, 2496 Leukemia inhibitory factor (LIF), 181, 691 expression after axotomy, 181, 694 in Schwann cell survival, 355 receptor for, 181 transport by dorsal root ganglion neurons, 181 upregulation in inflammation, 181 Leukocytes emigration from vasculature at sites of inflammation, 2342–2343, 2343f production of reactive oxygen species by, 510, 514 Leukodystrophy(ies), 1847–1859, 2330t. See also specific leukodystrophies. adrenoleukodystrophy, 1866–1868, 1867f globoid cell, 1855–1859, 1856f–1859f infantile, 2714 lysosomal, 1847–1859 metachromatic, 1847–1854, 1848f–1853f infantile, 2714–2715 multiple sulfatase deficiency, 1854–1855, 1854f–1855f sulfatide activator protein-B deficiency, 1855 pathology of, 710 Leukotrienes, glucocorticoid effects on, 636 Levator ani anatomy of, 280, 280f nerve to, 28, 281f Levetiracetam for diabetic neuropathy, 1971 for pain, 2644

xlviii Levorphanol, for neuropathic pain, 2645 Lewis-Sumner syndrome, 2278, 2279, 2336, 2232–2233 cerebrospinal fluid analysis in, 2232 clinical features of, 2232 electromyography in, 2232 magnetic resonance imaging in, 2232 nerve conduction studies in, 2232 nosologic position of, 2233 treatment of, 2232–2233 LFA-1 (lymphocyte function–associated antigen-1), 2342 in experimental autoimmune neuritis, 616, 625 Lhermitte’s sign, 1153, 1176, 1310 LHON (Leber’s hereditary optic neuropathy), 1550, 1940, 1941 Lidocaine neurotoxicity in sensory neurons, 699 spinal cauda equina syndrome after, 1333 syndrome of transient neurologic symptoms after, 1333 topical, for postherpetic neuralgia, 2641, 2645 LIF. See Leukemia inhibitory factor (LIF). Ligament of Struthers, median nerve entrapment by, 1453 Ligamentum flavum, 1324f Light-near dissociation, 212 LIM kinase (LIMK), in actin depolymerization, 452 Limb-girdle dystrophy, electromyography in, 954f Limb-girdle muscular dystrophy, macro electromyography in, 992f Limbs, cutaneous nerve distribution in, 14, 15f, 16f Lingual nerve, 1223f ␥-Linolenic acid, for diabetic neuropathy, 1970 Lipid peroxidation, 510, 510f auto-oxidative, 512–513 in diabetic patients, 511–512 in experimental diabetes, 512 role of nerve growth factor in, 514 Lipid rafts, 588 Lipids in myelin membrane, 423, 577 role in myelination, 423–424 Lipocortin-1, 567 Lipocyte, in Schwann cell, 771, 776f Lipofibromatous hamartoma, 2602, 2602f Lipofuscin age-related accumulation of, 702–703, 703t in spastic paraplegia, 703 in Tangier disease, 703 stress-related increase in, 703 vitamin E protection against accumulation of, 703 ␣-Lipoic acid for diabetic neuropathy, 516, 516t, 1971t, 1972 in ischemic-reperfusion oxidative injury, 521 protective effects against ischemia-reperfusion injury, 524, 526f, 676 Lipolysis, in diabetic neuropathy, 514 Lipoma(s), 2596 subcutaneous, 1168

Index Lipoprotein(s), 1905–1906, 1907t classification of, 1905–1906 components of, 1905 high-density (HDL), 1906 clinical phenotypes of deficiency of, 1909, 1910t composition of, 1906 familial deficiency of, 1905, 1908 hereditary deficiency of, 1905–1916. See also Tangier disease (TD). low-density (LDL), 1906 very-low-density (VLDL), 1906 LIS-1, in schizophrenia, 467 Liver heme biosynthesis in, 1883–1884, 1884f hydatid cysts in, 2169, 2170 in amyloidosis, 2434 in Fabry’s disease, 1896 in POEMS syndrome, 2455 in sarcoidosis, 2420 neuropathies associated with disorders of, 2017–2021 autoimmune hepatitis in Sjögren’s syndrome, 2021 cholestatic hepatic disease and vitamin E deficiency, 2021 Guillain-Barré syndrome and viral hepatitis, 2018 in hepatitis B, 2020 in hepatitis C, 2018–2020, 2019f Navajo neuropathy, 2021 sensory neuropathy and primary biliary cirrhosis, 2020–2021 somatic and autonomic polyneuropathy, 2017–2018, 2018t Liver transplantation for transthyretin amyloidosis, 1930–1931, 2444 for tyrosinemia, 2713 LM1 (sialosynneolactotetraosylceramide), 583t, 584 LMNA gene mutations, 1741–1742, 1770t, 1778–1779, 2714 Load cells, 1096 Localized interdigital neuritis, 1507, 2601 Long Q-T syndrome, 226–227 Long thoracic nerve, 19, 1340f distribution of, 1464, 1464f muscles innervated by, 956t Long thoracic neuropathy, 1464–1466 clinical manifestations of, 1465 diagnosis of, 1465 etiologies of, 1464–1465 iatrogenic, 1464 in children, 2727 prognosis for, 1465–1466 traumatic, 1464 treatment of, 1465 Long-chain 3-hydroxyacyl coenzyme A dehydrogenase (LCHAD) deficiency, 2715 Longitudinal ligament, anterior and posterior, 1324f Loperamide effect on anal sphincter, 286t for diarrhea, 293t, 2694 Low-density lipoprotein (LDL), 1906 Lower limb cutaneous nerve distribution in, 14, 16f in amyotrophic lateral sclerosis, 1536 in hereditary motor and sensory neuropathies, 1624, 1631–1635, 1634f Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Lower limb (Continued) innervation of, 14, 16f, 25–31, 957t lower motor neuron syndromes affecting, 1299–1300 mononeuropathies of, 1487–1508. See also Mononeuropathy(ies). in children, 2727–2730 Low-threshold mechanoreceptors (LTMs), 166–169, 167t, 1005 chemical phenotype of, 191 cutaneous A␣␤-fiber, 166–169 cutaneous A␦-fiber, 169 cutaneous C-fiber, 169 electrical membrane properties of, 170–171 microneurography of, 1005 proprioceptive or cutaneous, 186 vs. nociceptors, 170–171 LPA (lysophosphatidic acid), 355, 364 L-selectin, 2341–2342 LTMs. See Low-threshold mechanoreceptors (LTMs). Luft’s disease, 1940 Lumbar ganglion, 30, 234 Lumbar plexus, 25–26 anatomy of, 1375–1376 blood supply of, 1376 Lumbosacral plexus, 26, 1375–1386 anatomy of, 1375–1377, 1376f blood supply of, 1376 branches of, 26–28 lesions of, 1375–1386 causes of, 1378–1386 cerebrospinal fluid analysis in, 1378 CT and MRI in, 1377 diagnosis of, 1377 differential diagnosis of, 1499t due to hemorrhage, 1378, 1379f due to intra-arterial injections, 1380 due to obstetric and gynecologic complications, 1380–1381 due to surgical conditions and operations, 1381, 1382f electromyography in, 1377–1378 idiopathic neuropathy, 1386 in children, 2735–2736 in hereditary liability to pressure palsies, 1385–1386 infectious, 1384–1385, 1385f inflammatory, 1386 intradural injury, 1512 investigation of, 1377–1378 ischemic, 1380 traumatic, 1375, 1376, 1378–1380 tumors, 1381–1382, 1382f–1384f, 2736, 2738 vasculitic, 1386 vs. root lesions, 1377 muscles innervated by, 957t, 1376 nerve conduction studies of, 917 prefixation and postfixation of, 26 radiation plexopathy of, 1382–1384, 2490 clinical features of, 1382–1383 diagnostic studies in, 1383 electromyography in, 1378 vs. tumor plexopathy, 1384t variations of, 26 Lumbosacral polyradiculopathy, 734f, 1328–1329 Lumbosacral radiculopathies, pediatric, 2737–2738

Index Lumbosacral radiculoplexus neuropathy (LRPN), 1994 clinical features of, 1997t–1998t, 1999–2000 clinical testing in, 1997t, 2001–2002 nerve conduction, EMG, and quantitative autonomic testing, 2001, 2002t quantitative sensation testing, 2002, 2003t diabetic. See Diabetic lumbosacral radiculoplexus neuropathy (DLRPN). historical overview of, 1995 immunotherapy for, 2012 long-term morbidity in, 1999 pathology of, 2005f, 2006t, 2008f, 2009–2011, 2010f–2011f microvasculitis, 2405, 2408f teased fiber analysis, 2006t, 2009, 2011f recovery from, 1999 symptomatic treatment of, 2012–2013 Lung cancer, Pancoast’s syndrome, 1361–1362 Lung disease. See also Respiratory failure. in amyloidosis, 2433 in Fabry’s disease, 1894 in POEMS syndrome, 2460 in sarcoidosis, 2420 Lung receptors, vagal, 1284, 1284f Lupus. See Systemic lupus erythematosus (SLE). Luse bodies, 38 Lyme disease, 2109–2114 arthritis in, 2112 cardiac involvement in, 2112 carpal tunnel syndrome and, 1447 central nervous system involvement in, 2112 chronic, 2112 clinical features of, 2110–2112 cranial nerve involvement in, 1140, 2111 Bell’s palsy, 1226f, 1229–1230 facial nerve paralysis in children, 1236–1237 diagnosis of, 2113–2114 differential diagnosis of, 2114 childhood Guillain-Barré syndrome, 2717 etiologic agent for, 2109, 2110 experimental neuroborreliosis, 2113 geographic distribution of, 2109 historical overview of, 2109 incidence of, 2109–2110 neurobiology of, 2110 neurologic manifestations of, 2111–2112, 2351 pathology of, 2112–2113, 2112f–2113f pediatric polyneuropathy and, 2719 polyradiculopathy in, 1331 serologic tests for, 2114 skin lesions of, 1168, 2110–2111, 2111f ticks in transmission of, 2109, 2110 treatment of, 2114 vasculitis and, 2335, 2336, 2351 Lymphadenopathy in American trypanosomiasis, 2156 in leprosy, 2086 in POEMS syndrome, 2455 in sarcoidosis, 2420 Lymphocyte function–associated antigen-1 (LFA-1), 2342 in experimental autoimmune neuritis, 616, 625 Lymphoma, 2489–2495, 2600 acute inflammatory demyelinating polyneuropathy and, 2494 causes of peripheral nerve involvement in, 2489, 2490t

Lymphoma (Continued) chronic inflammatory demyelinating polyradiculoneuropathy and, 2494 classification of, 2489 Guillain-Barré syndrome and, 2202 intravascular, 2493 microvasculitis and, 2411f monoclonal proteins and, 2255, 2259–2260, 2260t motor neuron disease and, 2494–2495 nerve infiltration by, 1383f, 2491–2493, 2492f neuromyotonia and, 2495 paraneoplastic cerebellar degeneration and, 2495 paraneoplastic neuropathies in, 2495 brachial plexus neuropathy, 2495 subacute sensory neuronopathy or polyganglionopathy, 2495 paraneoplastic vasculitis and, 2493–2494 paraproteinemia and, 2494 pathology of, 714 peripheral neuropathies related to bone marrow transplantation for, 2498, 2498t peripheral neuropathy induced by chemotherapy for, 2490t, 2490–2491 primary leptomeningeal, 2491 Lyonization, 1548 Lysophosphatidic acid (LPA), 355, 364 Lysosomal storage disorders, 1845–1865 accumulating biochemical compounds in, 1846 Chédiak-Higashi syndrome, 1863–1864 classification of, 1846–1847 glycogenosis type II (Pompe’s disease), 1864 morphologic and clinical manifestations of, 1845–1846, 1846f–1847f mucolipidosis IV, 1865, 1866f mucopolysaccharidoses, 1864 neuronal ceroid-lipofuscinoses, 1862–1863, 1862f–1863f, 1864t oligosaccharidoses, 1864–1865, 1865f pathology of, 706–710 Schwann cells in, 1846 sequential degradation in, 1847 sphingolipidoses, 1847–1862 ceramide deficiency (Farber’s disease), 1859 gangliosidoses, 1859–1860 glucosylceramide lipidosis (Gaucher’s disease), 1860–1861 lysosomal leukodystrophies, 1847–1859 childhood, 2721 infantile, 2714 Niemann-Pick disease, 1861f, 1861–1862 Wolman’s disease, 1864

M protein, 1173 immunofixation electrophoresis for identification of, 2181, 2257, 2257f–2258f in POEMS syndrome, 2454, 2455 MAC (membrane attack complex), in myasthenia gravis, 834 Mac-1, 2342 Macro electromyography, 985, 991, 997–999 electrode for, 986f, 991 in alcoholic neuropathy, 999 in amyotrophic lateral sclerosis, 992f–993f, 996 in diabetic neuropathy, 993f, 999 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xlix Macro electromyography (Continued) in Guillain-Barré syndrome, 993f, 997–998, 998f in hereditary motor and sensory neuropathy, 993f, 998–999 in limb-girdle muscular dystrophy, 992f in patients with history of polio, 993f, 997 normal values for, 991 recording principle for, 991, 991f Macrodystrophia lipomatosa or hamartoma, 811–812 Macroglobulinemia, 686, 2255, 2259–2260, 2260t, 2271 Macroglossia, in amyloidosis, 2430, 2430f Macrophages, 559–560 activation of Fc receptors on, 560 as antigen-presenting cells, 559, 560f endoneurial, 574, 1415, 37, 559–560 functions of, 560 hemosiderin in, 797, 798f in autoimmunity, 560 in chronic inflammatory demyelinating polyradiculoneuropathy, 2239, 2240, 2241, 2245 in experimental autoimmune neuritis, 615, 618–620, 2239, 2242f inhibition of, 626–627 in Farber’s disease, 1859 in HIV-associated distal symmetrical polyneuropathy, 2133 in lysosomal disorders, 1846 in metachromatic leukodystrophy, 1849 in P0-deficient mice, 537–538, 541f in sarcoidosis, 2421 in Tangier disease, 1909, 1912 in wallerian degeneration, 1415–1417 pro-inflammatory cytokines released by, 560 response to nerve injury, 574 segmental demyelination mediated by, 560 MADSAM (multifocal acquired demyelinating sensory and motor) neuropathy, 2278, 2279 MAG. See Myelin-associated glycoprotein (MAG). Magnetic resonance imaging (MRI) functional, during sexual arousal, 317 in brachial plexopathy, 1350–1351, 1522–1523 hereditary, 1755f, 1760 immune, 2303 perinatal, 2731, 2733f in carpal tunnel syndrome, 1450–1451, 1450f–1451f in cervical spondylotic myelopathy, 1311–1313, 1312f in chronic immune sensory polyradiculopathy, 809f in chronic inflammatory demyelinating polyradiculoneuropathy, pediatric, 2718 in evaluation of pediatric mononeuropathies, 2725 in Guillain-Barré syndrome, pediatric, 2716 in lateral antebrachial cutaneous neuropathy, 1473 in Leigh disease, 2715 in leptomeningeal metastasis, 1330 in Lewis-Sumner syndrome, 2232 in lumbosacral plexus lesions, 1377 in lumbosacral polyradiculoneuropathy, 734f in multifocal motor neuropathy and conduction block, 2287 in nonmalignant inflammatory sensory polyganglionopathy, 2312 in radiculopathy, 1325 pediatric, 2737

l Magnetic resonance imaging (MRI) (Continued) in Ramsay Hunt syndrome, 1231 in sarcoidosis, 734f, 2419 in Sjögren’s syndrome, 1310, 1310f in suprascapular neuropathy, 1468 in tarsal tunnel syndrome, 1504 in West Nile virus infection, 1302 lumbar disc herniation on, 1329f of facial nerve, 1221–1222, 1221f–1222f, 1235 in Bell’s palsy, 1226f, 1228, 1228f in hemifacial spasm, 1239f in Lyme disease, 1226f of schwannoma, 2587 of spine, 1174 pelvic, in evaluation of fecal incontinence, 290t, 291 to evaluate central autonomic pathways, 1125 MAI (Mycobacterium avium-intracellulare), in HIV infection, 2139 Maillard reaction, 1974 Major dense line, 62, 63f, 413 absence in shiverer mice, 420, 536 in P0-deficient mice, 536 myelin basic protein and development of, 420, 536 Major histocompatibility complex (MHC) molecules, 574, 575, 2342 in small-cell lung carcinoma, 585 macrophage expression of, 559 Schwann cell expression of, 62, 565, 574 T cell recognition of, 560–563 MAL (myelin and lymphocyte) protein, 418f, 419t, 421 Malabsorption, tropical, 2068, 2071–2072 Malaria, 2155t, 2171–2172 cerebral, 2172 clinical features of, 2172 epidemiology of, 2171 etiologic agent in, 2166f, 2171 life cycle of, 2172 Guillain-Barré syndrome and, 2172 pathology of, 2172 transmission of, 2171–2172 Malignancy-associated vasculitis, 2351–2352 treatment of, 2374, 2375t, 2377t Malignant inflammatory sensory polyganglionopathy (MISP), 2315t Malignant peripheral nerve sheath tumor (MPNST), 2586t, 2596–2599 epithelioid, 2597–2598 granular, 2599, 2599f gross appearance of, 2596–2597, 2596f–2597f histology of, 2597, 2597f–2598f in neurofibromatosis 1, 2596 in transformation from tumors other than neurofibroma, 2599 metastasis of, 2597 nomenclature for, 2596 prognosis for, 2597 recurrence of, 2597 sites of, 2596 with divergent differentiation, 2598–2599, 2598f–2599f Malignant triton tumor, 2598 Mammalian diaphanous (mDia), in actin nucleation, 451f, 451–452 Mandibular nerve, anatomy of, 1207, 1208f Mannosidosis, 1846, 1864 in children, 2722 pathology of, 709

Index Manual muscle testing, 1095–1096, 1096t relative value for lower extremity muscles, 1095, 1096f scoring system for, 1095, 1096t MAPK (mitogen-activated protein kinase) diabetic neuropathy and, 1976–1977 ephrin-mediated neurite retraction by inactivation of, 455 Maprotiline, for neuropathic pain, 2642 Marcus Gunn phenomenon, 816 Marinesco-Sjögren syndrome, 1738 in children, 2724 with myoglobinuria and CCFDN syndrome, 1776 Martin-Gruber anastomosis, 1437 nerve conduction studies with, 928–929, 929f MASA syndrome, 467–468 Mash1, 343 Massachusetts Male Aging Study, 1957 Masseter (jaw) reflex, 921, 923t Mast cells, 574, 564–565 endoneurial, 37 enteric, 266 in experimental autoimmune neuritis, 620 synthesis of nerve growth factor by, 379 Matrix metalloproteinases (MMPs) in diabetic motor neuropathy, 1955 in experimental autoimmune neuritis, 617, 617f, 623, 625 in POEMS syndrome, 2455 in wallerian degeneration, 1418f, 1418–1419 MMP-3 in Schwann cell proliferation, 1409 MAX-1, in netrin signaling, 461 Maxillary nerve, anatomy of, 1207, 1208f Maximal stimulation test, in Bell’s palsy, 1226 Maytansine, microtubule disaggregation induced by, 389 MBFC. See Medial brachial fascial compartment (MBFC) syndrome. MBP. See Myelin basic protein (MBP). MCB (membranous cytoplasmic bodies), 707–708 mDia (mammalian diaphanous), in actin nucleation, 451f, 451–452 Meatal foramen, 1219 MEC-2, 187t, 188 Mechanical ventilation, 2613–2614 in amyotrophic lateral sclerosis, 2633–2634 laboratory criteria for, 2613, 2613t modes of, 2613–2614, 2614t nosocomial pneumonia associated with, 2615–2616, 2616t weaning from, 2614–2615 withdrawal of respiratory support, 2617 Mechanoreceptors, 1064 expression by dorsal root ganglion neurons, 166–169, 188 fast-adapting and slow-adapting, 166, 1005 functional significance of, 1005 Golgi tendon organ, 151–159 in enteric nervous system, 254, 256 low-threshold (LTMs), 166–169, 167t, 1005 chemical phenotype of, 191 cutaneous A␣␤-fiber, 166–169 cutaneous A␦-fiber, 169 cutaneous C-fiber, 169 electrical membrane properties of, 170–171 microneurography of, 1005 proprioceptive or cutaneous, 186 vs. nociceptors, 170–171 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Medial brachial fascial compartment (MBFC) syndrome, 1368–1369 anatomy of, 1368, 1369f diagnostic evaluation of, 1368 management of, 1369 mechanism of, 1368 predisposing factors for, 1368 Median nerve anatomy of, 1340f, 1436f–1437f, 1436–1438, 1436f–1437f, 1436–1438 Martin-Gruber anastomosis, 1437 Riche-Cannieu anastomosis, 1437 distribution of, 15f, 20f, 20–21, 1470f in leprosy, 2089f, 2093 involvement in vasculitic neuropathy, 2365t muscles innervated by, 956t, 1436–1437, 1436f–1437f nerve conduction studies of, 901f–903f, 902, 905f, 908t, 909–910, 910t, 911f–912f, 913t, 914t collision technique for, 926–928, 928f in multifocal motor neuropathy, 935f reproducibility of, 936, 938f with Martin-Gruber anastomosis, 928–929, 929f Volkmann’s contracture with infarction of, 1517 Median neuropathy, 1435–1454 anterior interosseous nerve syndrome, 1453–1454 carpal tunnel syndrome, 1435–1452 entrapment by ligament of Struthers, 1453 in arm, 1452–1453 in children, 1448, 1448t, 2724f, 2725f, 2725–2726 in renal failure, 2026 proximal forearm entrapment, 1453 Median sternotomy, brachial plexus lesion after, 1364–1365 Medical Outcomes Study–HIV (MOS-HIV), 1059 Medullary thyroid carcinoma (MTC), in multiple endocrine neoplasia type 2B, 1805–1807 Mees’ lines, 1168 Megacolon, 2692 causes of, 2692 congenital, 705, 1151 diagnostic evaluation of, 2692 in Chagas’ disease, 2158f, 2159 management of, 2692 Megaesophagus, in Chagas’ disease, 2159 Meissner’s corpuscles, 484, 870, 871, 1005, 1064, 1064f Meissner’s plexus, 18, 2627 Melanin, aging and, 703 Melanocytes, 870 Melanoma, 2478, 2600 chronic inflammatory demyelinating polyradiculoneuropathy and, 2228, 2245 Melanotic schwannoma, 2588–2591, 2590f MELAS (mitochondrial encephalomyopathy, lactic acidosis, and strokelike episodes), 1941, 1943, 1944, 1946 Melkersson-Rosenthal syndrome, 1237–1238 Melphalan, 637–638 for amyloidosis, 2437–2438 for POEMS syndrome, 2463 Membrane attack complex (MAC), in myasthenia gravis, 834 Membranous cytoplasmic bodies (MCBs), 707–708

Index MEN. See Multiple endocrine neoplasia (MEN) type 2B. Mendelian inheritance, 1545–1549, 1595 autosomal dominant, 1546–1547 autosomal recessive, 1547–1548 genetic counseling about, 1549 X-linked dominant, 1548–1549 X-linked recessive, 1548 Y-linked, 1549 Meningeal artery, middle, 1220 Meningioma, 2596 trigeminal, 1210–1211 Meningitis aseptic, intravenous immunoglobulin-induced, 644 Epstein-Barr virus, 2120 herpes simplex virus, 2117 in Lyme disease, 2111 in sarcoidosis, 2419, 2421 Meningoencephalitis, eosinophilic, 2164 Meningomyelocele, 1333 Meningoradiculitis, in Lyme disease, 2111 Menopause, carpal tunnel syndrome and, 1445 MEPC (miniature end-plate current), 832 MEPP (miniature end-plate potential), 832 in choline acetyltransferase deficiency, 840 in myasthenia gravis, 833 Meralgia paresthetica, 1492–1493 diagnosis of, 1492 etiology of, 1492 intermittent symptoms in, 1141 treatment of, 1493 6-Mercaptopurine (6MP), 639 Mercury intoxication, 2538–2540 clinical features of, 2538–2539 effect on axonal transport, 397t laboratory studies in, 2539 pathology of, 688, 715, 2539–2540 selective toxicity for dorsal root ganglion cells, 1296 treatment of, 2540 Merkel cell–neurite complex, aging of, 484, 496 Merkel cells, 871, 1005 loss of, in P0-deficient mice, 537 Merlin in Schwann cell proliferation, 363–364 schwannomas associated with mutations of, 363–364, 2586 MERRF (myoclonic epilepsy with ragged red fibers), 1941 Mesaxon, 3, 5, 412, 413f Mesenteric blood flow evaluation of, 1124–1125, 1125f in response to tilt-table testing, 2695 Mesenteric ganglion inferior, 234, 238f, 243 superior, 234 Mesenteric plexus, 31 Meso-2,3-dimercaptosuccinic acid, for arsenic poisoning, 2538 Metachromatic leukodystrophy (MLD), 1847–1854, 2714–2715 adult form of, 1848 diagnosis of, 1848, 1849 prenatal, 1853–1854 enzyme defect in, 1847, 2715 genetics of, 1852–1854 historical recognition of, 1847–1848 in children, 2708–2710, 2721 juvenile form, 1848–1849 late infantile form, 1848, 1848f–1853f, 2715 laboratory findings in, 1849

Metachromatic leukodystrophy (MLD) (Continued) nerve conduction velocity in, 1849 pathogenesis of, 1851–1852 pathology of, 710, 814f, 815, 819–820, 1849–1851, 1850f–1853f demyelination, 1847, 1849–1852 prismatic inclusions, 1850 tuffstone bodies, 1850, 1852f–1853f types of inclusion material, 1849–1851 zebra bodies, 1850, 1851f treatment of, 1854 variants of, 1847, 1854–1855 multiple sulfatase deficiency, 1854–1855, 1854f–1855f sulfatide activator protein-B deficiency, 1855 Metal neuropathy, 715–716, 716f, 2527–2545 arsenic, 1168, 2064–2065, 2535–2538 effects on axonal transport, 397t gold, 2543–2544 lead, 2527–2535 mercury, 2538–2540 platinum, 2544–2545 thallium, 2540–2543 Metastatic radiculopathy, 1329–1331 Methamidophos toxicity, 2075 Methemoglobinemia, 2613 Methotrexate for chronic inflammatory demyelinating polyradiculoneuropathy, 2179t in children, 2719 for sarcoidosis, 2423 Methyl bromide, 2514–2515 Methyl ethyl ketone, 2512 Methyl isobutyl ketone, 2512 Methyl n-butyl ketone, 397t, 399, 2511–2514 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)–induced Parkinson’s disease, 312 2-Methylacyl-CoA racemase deficiency, peroxisomal, 1866, 1872–1873 N-Methyl-D-aspartate (NMDA) antagonists, for pain, 527 N-Methyl-D-aspartate (NMDA) receptors, 524–527 effect of inflammation on, 525 in amyotrophic lateral sclerosis, 1538 in chronic constriction model, 525–527 localization of, 525 role in pain, 524–527, 2640 Methylmalonic acid (MMA) measurement of, 1172 vitamin B12 deficiency and, 2053, 2054 Methylmercury intoxication, 397t, 1296, 2538–2540. See also Mercury intoxication. Methylphenidate, for orthostatic hypotension, 2661, 2676 Methylprednisolone for brachial plexus neuropathy hereditary, 1763, 1766 immune, 2306 for chronic inflammatory demyelinating polyradiculoneuropathy, 637 for Guillain-Barré syndrome, 2213 for lumbosacral radiculoplexus neuropathy, 2012 diabetic, 2012, 2013f 3-Methyoxytyramine, effect of dopamine-␤-hydroxylase deficiency on urinary level of, 1118 Metoclopramide for gastroparesis, 1960, 2665, 2690 for orthostatic hypotension, 2661, 2678 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

li Metronidazole neuropathy, 2560 Mexiletine adverse effects of, 2643 for diabetic neuropathy, 1971 for neuropathic pain, 2643 MFs. See Myelinated fibers (MFs). MG. See Myasthenia gravis (MG). MGUS. See Monoclonal gammopathy of undetermined significance (MGUS). MHC. See Major histocompatibility complex (MHC) molecules. MIBG (123I-meta-iodobenzylguanadine) single-photon emission computed tomography of sympathetic innervation to organs, 1119–1121, 1120f Microbial antigens, 590–591 Microfilaments, 434 Microneurography, 324, 324f, 1003–1011 future perspectives in, 1011 in C8 radiculopathy, 1149f in disorders of sensory nerve function, 1007–1010 neuropathic changes in unmyelinated fibers, 1008 paresthesias and spontaneous activity in myelinated fibers, 1007–1008 sensitization of C fibers in erythromelalgia, 1009–1010, 1010f intraneural stimulation in, 1004 of low-threshold mechanoreceptors, 1005 of nociceptors, 1005–1006 of thermoreceptors, 1005 technique of, 1003–1004, 1004f Microsatellite repeat markers, 1552 Microscopic polyangiitis (MPA), 2337, 2339, 2353t, 2355 antineutrophil cytoplasmic antibodies in, 2352, 2355 classification and definition of, 2338t, 2339, 2340t genetics of, 2363 incidence of, 2340t pediatric polyneuropathy and, 2719 relative prevalence of, 2341t Microscopy, of cutaneous nerves, 876, 877f Microtrabecular matrix, axonal, 49–51 Microtubules (MTs), 434 drug-induced disaggregation of, 389, 393 fragility of, 49 in axonal dystrophy, 778, 782f in axonal transport, 388–389, 393 in peripheral domain, actin dynamics and, 448–449 in unmyelinated fibers, 65 juxtaparanodal, 54 length of, 389 microscopic anatomy of, 49 number of, 49 of constricted axon segment, 60 of growth cone, 448, 448f paranodal, 417 structure of, 389 tubulin subunits of, 389 Microvasculature of peripheral nerve, 667–668 as nutritive capacitance system, 668 in compression neuropathies, 1396–1399, 1397f–1398f large capillaries of, 668 nerve blood flow:blood pressure curve, 667–668 regulation of nerve blood flow, 668–671, 669f, 670t

lii Microvasculitis, 2405–2407. See also Vasculitis. disease association with, 2405–2406 diabetic lumbosacral radiculoplexus neuropathy, 2004–2007, 2007f–2008f, 2405–2406 microvessel pathology in, 2406–2407 neovascularization and, 2412f nerve pathology in, 2406f–2413f, 2407 overview of, 2405 Micturition, 282, 2626, 300f, 300–301. See also Urinary continence; Urinary incontinence. anal sphincter activity during, 284 in chronic inflammatory demyelinating polyradiculoneuropathy, 2225 neurogenic dysfunction of, 313–314 failure to eliminate urine, 313 failure to store urine, 313 in spinal cord injury, 313–314 neurotransmitters in micturition reflex pathways, 312 organization of urine storage and voiding reflexes, 300f, 300–301, 310–312 pontine micturition center, 311–312 somatic pathways to urethral sphincter, 310 spinobulbospinal micturition reflex pathway, 310–311, 311f suprapontine control of micturition, 312 sympathetic pathways, 310 positron emission tomography studies during, 312 reflexes controlling, 303t, 306, 306f–307f responses in infant, paraplegic patient, and adult, 300f Midazolam, for ventilator withdrawal, 2617 Midbrain tumors, 1199 Midodrine for orthostatic hypotension, 1960, 2659–2660, 2660t, 2660f–2661f, 2675–2676 for urinary incontinence, 2664 Migraine familial hemiplegic, calcium channel mutations in, 590 gastrointestinal manifestations of, 2695 Migrainous neuralgia, 1214 Migrant sensory neuritis of Wartenberg, 1141 Military personnel, brachial plexus lesions in, 1359 Miniature end-plate current (MEPC), 832 Miniature end-plate potential (MEPP), 832 in choline acetyltransferase deficiency, 840 in myasthenia gravis, 833 Minipolymyoclonus, 1144 Minocycline, for leprosy, 2103 Miosis in Horner’s syndrome, 205 in neonate, 208 Mipafox, 2515 Misonidazole neuropathy, 2561 MISP (malignant inflammatory sensory polyganglionopathy), 2315t Mitochondria axonal, 47, 48f of internodal end region, 53 production of reactive oxygen species by, 510, 514 Mitochondrial DNA (mtDNA), 1549–1550, 1938 in antiretroviral toxic neuropathy, 2134 maternal inheritance of mutations of, 1938–1940 penetrance of mutations of, 1939–1940 structure of, 1938, 1939f

Index Mitochondrial encephalomyopathy, lactic acidosis, and strokelike episodes (MELAS), 1941, 1943, 1944, 1946 Mitochondrial inheritance, 1549–1550, 1938–1940 heteroplasmy and, 1550, 1940 maternal, 1938–1940 segregation and, 1940 threshold effect and, 1940 Mitochondrial neurogastrointestinal encephalopathy (MNGIE), 1942, 1945–1946, 2700–2701 in children, 2722 Mitochondrial respiratory chain, 1938–1940, 1939f Mitochondrial respiratory chain dysfunction, 1937–1947 axonal neuropathy and, 1946–1947 acute and subacute, with A3243G mutation, 1946 autonomic neuropathy, 1947 neuropathy, ataxia, and retinitis pigmentosa, 1941, 1942, 1946–1947 sensory ataxic neuropathy, dysarthria, and ophthalmoparesis, 1946 classification of, 1937–1938, 1938t clinical features of, 1940–1941 common phenotypes of, 1940–1941 demyelinating neuropathy and, 1944–1946 Leigh disease, 1944–1945, 1945f, 2715 mitochondrial neurogastrointestinal encephalopathy, 1942, 1945–1946, 2700–2701 diagnosis of, 1940–1941 ragged red fibers, 1940, 1941f evidence for association with peripheral nerve disease, 1937 frequency of peripheral nerve disease in, 1942 in children, 2722 pathology of peripheral nerve disease in, 1943 Mitochondrial trifunctional protein (MTP) deficiency, 2715 Mitogen-activated protein kinase (MAPK) diabetic neuropathy and, 1976–1977 ephrin-mediated neurite retraction by inactivation of, 455 Miyoshi myopathy, 1175, 1176 MK-801, for pain, 527 MLCK (myosin light chain kinase), 452 MLCP (myosin light chain phosphatase), 452 MLD. See Metachromatic leukodystrophy (MLD). MMA (methylmalonic acid) measurement of, 1172 vitamin B12 deficiency and, 2053, 2054 MMCs (myoenteric migrating complexes), 1280 MMN-CB. See Multifocal motor neuropathy and conduction block (MMN-CB). MMPs. See Matrix metalloproteinases (MMPs). MND. See Motor neuron disease (MND). MNGIE (mitochondrial neurogastrointestinal encephalopathy), 1942, 1945–1946, 2700–2701 in children, 2722 MNSI questionnaire, 1040, 1041f Möbius syndrome, 1197–1198 congenital trigeminal anesthesia and, 1211 facial paralysis in, 1225, 1236 Modified Rankin scale, 1042, 1046t MODS (multiple organ dysfunction syndrome), 2026–2027 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Molecular mimicry, 591–592 in chronic inflammatory demyelinating polyradiculoneuropathy, 2245 in Guillain-Barré syndrome, 2197, 2199, 2205–2206, 2245 Monoclonal gammopathy, 1173, 2255–2271 cryptogenic sensory polyneuropathy and, 2326 motor neuron disease and, 2269–2270 multiple myeloma and, 2270–2271 solitary plasmacytoma and, 2259–2261, 2260t, 2271 Waldenström’s macroglobulinemia and, 2259, 2271 Monoclonal gammopathy of undetermined significance (MGUS), 1173, 2255, 2259–2269 association of sensorimotor peripheral neuropathy with, 2261–2262 characterization of, 2259–2261 clinical features of, 2255, 2264, 2462t differential diagnosis of neuropathy in, 2264–2265, 2265t amyloidosis, 2265 chronic inflammatory demyelinating polyradiculoneuropathy, 2228, 2255, 2264–2265 multiple myeloma, 2265 paraneoplastic neuropathy, 2265 electromyography in, 2265–2266, 2266f etiology of neuropathy in, 2262 genetic factors, 2262 experimental studies of, 2263 IgG and IgA monoclonal proteins and neuropathy in, 2268 IgM monoclonal proteins and neuropathy in, 2266–2267 myelin-associated glycoprotein, 2266–2267, 2267f reaction with other components, 2267–2268 incidence of, 2259 laboratory features of, 2462t lymphoma and, 2494 neuropathy associated with IgM vs. IgG and IgA in, 2255, 2263–2264 anti–myelin-associated glycoprotein antibodies and, 2255 pathology of, 2262–2263 risk of progression of, 2259–2261, 2260t, 2260f–2261f terminology for, 2259 treatment of neuropathy in, 2255, 2268–2269, 2269t intravenous immunoglobulin, 644, 2269 Monoclonal proteins, 2255–2256 amyloidosis with, 2255, 2259–2260, 2260t, 2265, 2265t, 2269t, 2435, 2435t analysis of serum for, 2256–2258 cryoglobulinemia, 2257–2258 electrophoresis, 2256f, 2256–2257 immunofixation, 2181, 2257, 2257f serum viscometry, 2257 analysis of urine for, 2258, 2258f diseases associated with, 810, 2255, 2259–2261, 2260t clinical and laboratory features of, 2462t lymphoma and, 2494 in chronic lymphocytic leukemia, 2496 in nonmalignant inflammatory sensory polyganglionopathy, 2312 in POEMS syndrome, 2458–2459, 2459f incidence of, 2259 nomenclature for, 2256

Index Monomelic amyotrophy of lower limb, 1299 Mononeuritis multiplex, 1140, 2336. See also Multiple mononeuropathy. cytomegalovirus, 2122 diabetic, 1993 Mononeuropathy(ies), 1138, 1140 compressive, 1140, 1140t, 1391–1400 in renal failure, 1140, 2025–2026 of lower limb, 1487–1508 common peroneal, 1496–1500 femoral, 1489–1491 genitofemoral, 1488, 1488f gluteal, 1493 iliohypogastric, 1488, 1488f ilioinguinal, 1488, 1488f in children, 2727–2730 interdigital, 1506–1507 Joplin’s neuroma, 1506–1507 Morton’s neuroma, 1507 lateral femoral cutaneous, 1492–1493 medial calcaneal, 1506 obturator, 1488–1489 pathophysiology of, 1487–1488 piriformis syndrome, 1496 plantar, 1505–1506 posterior femoral cutaneous, 1493 saphenous, 1491–1492 sciatic, 1493–1496 sural, 1507–1508 tarsal tunnel syndrome, 1503–1505 tibial, 1500–1503 of upper limb, 1463–1484 axillary, 1469–1471 in children, 2725–2727 long thoracic, 1464–1466 medial antebrachial cutaneous, 1483–1484 median, 1435–1454 musculocutaneous, 1471–1474 pathophysiology of, 1463–1464 radial, 1474–1478 suprascapular, 1466–1469 ulnar, 1478–1483 pediatric, 2724–2730. See also Pediatric patients. vs. radiculopathy, 1324 Monteggia’s fracture, posterior interosseous nerve injury and, 1475 Morphine for neuropathic pain, 2645 for ventilator withdrawal, 2617 Morphometry in amyotrophic lateral sclerosis, 770f in autonomic dysfunction, 771f of nerve or nerve fiber counts in skin biopsy, 1046 of neuron somas in motor neuron columns, 754–755, 756f–757f, 759f of neuron somas in spinal ganglia, 755–756, 758t, 759f–760f of paranode-node-paranode region in cat, 52t–53t of peripheral nerves, 756–763 computer imaging of myelinated fiber composition, 759–763, 761f assessing variability in myelinated fiber density, 762–763 hardware, 760 myelinated fiber measurements from transverse sections, 761–762, 762t system approaches, 760–761 histologic processing, 757 photography of myelinated fiber composition, 759

Morphometry (Continued) sampled sections of internodes and average diameter, 758–759, 761f section thickness and axon-myelin relationship, 757–758 of peroneal nerve, 768f of sciatic nerve, 767f of tibial nerve, 767f of unmyelinated fibers, 66–67, 66f–67f Morquio syndrome type B, 1860 Morton’s neuroma, 1507, 2601 Mosaicism, germinal, 1546 MOS-HIV (Medical Outcomes Study–HIV), 1059 Mossy fiber sprouting, in epilepsy, 466 Motilin, in enteric nervous system, 266 Motor nerve conduction studies, 900–903, 901f, 905f. See also Nerve conduction studies (NCSs). absent responses on, 903, 905f in diabetic neuropathy, 1978–1979 in hereditary motor neuronopathy, 1605 increased latency with normal amplitude on, 902–903 latency and conduction velocity in, 900–902, 902f reduced amplitude with normal latency on, 902, 903f–904f types of abnormalities of, 902, 902f waveform, amplitude, and duration on, 900 Motor neuron disease (MND) amyotrophic lateral sclerosis, 1304–1306, 1533–1540 course of, 1534 differential diagnosis of, 1298 in multifocal motor neuropathy with conduction block, 1306–1307 lower, 1298–1300 acquired syndromes, 1299 amyotrophic lateral sclerosis and, 1304–1306 differential diagnosis of, 1536 electromyography in, 939f, 951, 953f, 955 hereditary motor neuronopathy, 1299, 1603–1618 in HIV infection, 1303 inherited syndromes, 1299 other causes of, 1304 overflow incontinence and, 2627 paraneoplastic, 1299, 1302–1303 poliomyelitis, 1300–1301 postirradiation, 1299, 1303 primary site of pathology in, 1298–1299 progressive bulbar palsy, 1533 progressive muscular atrophy, 1533 upper and lower extremity involvement in, 1299–1300 West Nile virus infection, 1301–1302 lymphoproliferative disorders and, 2494–2495 management of, 2633–2634 monoclonal gammopathy and, 2269–2270 paraneoplastic, 2479 spinal cord injury and, 2699 sporadic, 1533–1540 upper corticobulbar palsy, 1533 differential diagnosis of, 1536 electromyography in, 939f, 955 primary lateral sclerosis, 1533 vs. polyneuropathy, 1176 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

liii Motor neurons, 575 morphometry of neuron somas in motor neuron columns, 754–755, 756f–757f, 759f of enteric nervous system, 251, 251f, 253, 257f, 257–259 Motor neuropathy, 1141 diabetic, 1955–1956 in chronic inflammatory demyelinating polyradiculoneuropathy, 1141, 2224, 2225t, 2231 in sarcoidosis, 2418 multifocal, with conduction block, 122–124, 1306–1307, 2277–2294. See also Multifocal motor neuropathy and conduction block (MMN-CB). paraneoplastic, 2479 Motor symptoms, 1142–1146, 1165 negative, 1142 positive, 1142–1146 Motor syndromes, 1298–1307 amyotrophic lateral sclerosis, 1304–1306 classification of, 1298 HIV infection, 1303 localizing portions of motor neuron involved in, 1298 lower motor neuron disease, 1298–1300 multifocal motor neuropathy with conduction block, 1306–1307 other causes of, 1304 paraneoplastic lower motor neuron syndrome, 1302–1303 poliomyelitis, 1300–1301 postirradiation lower motor neuron syndrome, 1303 West Nile virus infection, 1301–1302 Motor unit(s) classification by contractile properties of, 155 discharge patterns in electromyography, 952–955 in lower and upper motor neuron disorders, 955 in myopathy, 955 involuntary movement, 955 measurements of turns and amplitude, 953–954 recruitment, 952–953, 953f–954f electrophysiologic studies of, 985–999 Golgi tendon organ response to contraction of, 153–155, 154f measuring fiber density of, 987f, 987–988 Motor unit potentials (MUPs), 947–952 abnormalities of, 949–951 amplitude and area of, 948 conditions associated with types of, 947–948, 948t duration of, 949, 950t in brachial plexopathy, 1351 in true neurologic thoracic outlet syndrome, 1360 in vasculitic neuropathy, 2367 inn lower motor neuron vs. myopathic disorders, 951–952, 953f–954f jiggle of, 949 phases of, 949, 951f–952f physiologic factors affecting, 947 profile of, 948f rise time of, 948–949 Mouse motor neuron degeneration, 711 Mouth, intensive care unit care for, 2609 6MP (6-mercaptopurine), 639 MPA. See Microscopic polyangiitis (MPA).

liv MPNST. See Malignant peripheral nerve sheath tumor (MPNST). MPTP (1-methyl-4-phenyl-1,2,3,6tetrahydropyridine)–induced Parkinson’s disease, 312 MPZ. See P0 (myelin protein zero). MRI. See Magnetic resonance imaging (MRI). MS. See Multiple sclerosis (MS). MSA. See Multiple system atrophy (MSA). MSD. See Multiple sulfatase deficiency (MSD). MSNA. See Muscle sympathetic nerve activity (MSNA). MTC (medullary thyroid carcinoma), in multiple endocrine neoplasia type 2B, 1805–1807 mtDNA. See Mitochondrial DNA (mtDNA). MTMR2 gene mutations in autosomal recessive hereditary motor and sensory neuropathy, 1770t, 1774–1775, 2714 in hereditary motor and sensory neuropathy IVB1, 1630 MTMR13 gene mutations, 1770t, 1774 MTP (mitochondrial trifunctional protein) deficiency, 2715 MTs. See Microtubules (MTs). Mucolipidosis pathology of, 709 type I, 1864 type II, 1864 type III, 1864 type IV, 1865, 1866f Mucopolysaccharidoses, 1846, 1864–1865 classification of, 1864 in children, 2721–2722 pathology of, 708–709, 1864 Mucosulfatidosis, 1854. See also Multiple sulfatase deficiency (MSD). Müller’s muscle, paralysis in Horner’s syndrome, 205 Multifactorial disorders, 1551 Multifocal acquired demyelinating sensory and motor (MADSAM) neuropathy, 2232, 2278, 2279 Multifocal motor and sensory demyelinating neuropathy, 2232–2233 cerebrospinal fluid analysis in, 2232 clinical features of, 2232 electromyography in, 2232 magnetic resonance imaging in, 2232 nerve conduction studies in, 2232 nosologic position of, 2233 treatment of, 2232–2233 Multifocal motor neuropathy and conduction block (MMN-CB), 122–124, 575, 935f, 1306–1307, 2179t, 2277–2294 age at onset of, 2278 clinical features of, 122, 1142, 2278–2279, 2279t, 2280f diagnostic criteria for, 2279, 2280t differential diagnosis of, 2279–2280 amyotrophic lateral sclerosis, 1536 chronic inflammatory demyelinating polyradiculoneuropathy, 2229 electrophysiology of, 1306–1307, 2182, 2280–2284, 2281t conduction block, 2280–2284, 2282f electromyography, 2284 nerve conduction studies outside areas of conduction block, 2284 nerve excitability studies, 122–124, 123f, 124f transcranial magnetic stimulation, 2284 etiology and pathogenesis of, 2287–2290

Index Multifocal motor neuropathy and conduction block (MMN-CB) (Continued) antiganglioside antibodies, 1307, 2288–2290 Campylobacter jejuni/coli infection, 2290 immunology, 575 historical descriptions of, 2277–2278 IgM anti-GM1 antibodies in, 2179t, 2182–2184 antibody testing for, 2182 treatment based on, 2182–2184 disease pathogenesis and, 2184 in children, 2725 in sarcoidosis, 2418 laboratory findings in, 2284–2286 antiganglioside antibodies, 596, 2284–2286, 2286f cerebrospinal fluid protein, 2286 creatine kinase, 2286 magnetic resonance imaging in, 2287 natural history of, 2278, 2280 pathology of, 2286–2287, 2287f–2288f pathophysiology of, 1307 prevalence of, 2278 treatment of, 2182–2184, 2290–2294 cyclophosphamide, 2290, 2291 intravenous immune globulin, 644, 2290–2291, 2292f recommendations for, 2291–2294, 2293f Multiple endocrine neoplasia (MEN) type 2B, 1805–1807 alimentary tract ganglioneuromatosis in, 1806, 1806f, 2595 congenital maldevelopment and pathology of, 816, 817f facies in, 1805, 1806, 1806f historical overview of, 1805 neurologic manifestations of, 1807 pattern of inheritance of, 1805 ret mutation in, 1805 skeletal abnormalities in, 1806, 1806f Multiple mononeuropathy, 2336 in adolescents, 2711 in hepatitis B infection, 2020 in HIV infection, 2137–2138, 2350 in necrotizing vasculitis, 802 in sarcoidosis, 2418 vs. leprosy, 2097 Multiple myeloma, 2255, 2259–2260, 2260t, 2265t, 2269t, 2270–2271 amyloidosis with, 2429–2430 vs. POEMS syndrome, 2453–2454, 2462t Multiple neuritis in West Indies. See Strachan’s syndrome. Multiple organ dysfunction syndrome (MODS), 2026–2027 Multiple sclerosis (MS) blink reflex in, 920t chronic inflammatory demyelinating polyradiculoneuropathy and, 2233 diplopia in, 1203 eighth nerve pathology in, 1268 electromyography in, 947f gastrointestinal symptoms of, 2700 obstructed defecation in, 306 potassium channel blockers for, 104 trigeminal neuralgia and, 1214 vs. sensory neuropathy, 1153 Multiple sulfatase deficiency (MSD), 1854–1855, 2715 clinical features of, 1854, 1854f enzyme defects in, 1854 inheritance of, 1854 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Multiple sulfatase deficiency (MSD) (Continued) laboratory findings in, 1854–1855 pathogenesis of, 1855 pathology of, 1855, 1855f Multiple system atrophy (MSA) airway dysfunction in, 1284–1285 autonomic function testing in, 1103–1104 clonidine–growth hormone stimulation test, 1127, 1127f plasma catecholamine levels, 1117, 1119f SPECT imaging of sympathetic innervation to organs, 1119–1121, 1120f cardiac vagal activity in, 226, 1280 gastrointestinal manifestations of, 2696 Multivesicular bodies (MVBs) axonal, 47 of constricted axon segment, 59 of juxtaparanodal segment, 53 ␮ granule of Reich, in Schwann cells, 771, 775f, 791 MUPs. See Motor unit potentials (MUPs). Muscle. See also specific muscles. age-related changes in, 1176 atrophy in hereditary motor and sensory neuropathies, 1624, 1636 transgenic mice overexpressing PMP22, 1564 denervation atrophy of, 1142 effects of immobility on strength of, 2629 Golgi tendon organ response to contraction of, 153–155, 154f in Fabry’s disease, 1896 innervation of cranial, shoulder girdle, and upper limb muscles, 956t innervation of hip girdle and lower limb muscles, 957t involvement in leprosy, 2103 neurally mediated vasodilation in, 331–332 neuromuscular electrical stimulation for strengthening of, 2623–2624 smooth muscle layers of bladder, 300 stiffness of, in peripheral nerve hyperexcitability syndromes, 1143 Muscle action potentials. See Compound muscle action potentials (CMAPs). Muscle biopsy in amiodarone neuropathy, 2555 in chloroquine neuropathy, 2556 in colchicine neuropathy, 2557 in critical illness polyneuropathy, 2028, 2028f in hereditary motor neuronopathy, 1605–1606 in sarcoidosis, 2421, 2422f–2423f Muscle cramps, 1143 in amyotrophic lateral sclerosis, 1536, 1540 in hereditary motor and sensory neuropathy, 1624, 1631 in peripheral nerve hyperexcitability syndromes, 1143 in uremic neuropathy, 2023 on electromyography, 946–947, 1143 Muscle fibers extrafusal, 132, 137, 138 insertion into Golgi tendon organ, 151–152, 152f intrafusal, 132, 137–138 Muscle spindle(s), 131–147 abundance of, 135–136 action potentials evoked in, 138 capsule of, 132 coefficient of variation for number of, 135 development of, 136–137 equatorial region of, 132, 138

Index Muscle spindle(s) (Continued) inner capsule cells of, 132 innervation of, 132, 138–147 motor, 144–147, 145f sensory, 132, 138–144, 139f–143f intrafusal fibers of, 132, 137–138 junctional potentials evoked in, 138 of cat soleus, 132–135, 133f, 134f organization of, 132 periaxial space of, 132 sensory complement of, 136 structural and functional complexity of, 131 Muscle strength testing, 1095–1099 clinical examination, 1095–1096, 0432f, 1096t direct muscle force measurement for, 1097 electromyographic, 1097 in children, 2711 instrumented methods for, 1096–1097 intramuscular pressure measurement for, 1097–1099, 1098f–1099f, 1099t noninstrumented methods for, 1096 quantitation of, 1037 scoring results of, 1036–1037 Muscle stretch reflexes, 1038 in Guillain-Barré syndrome, 2199–2200 in pediatric chronic inflammatory demyelinating polyradiculoneuropathy, 2718 testing in children, 2711 Muscle sympathetic nerve activity (MSNA) autonomic responses to exercise and, 328–331 baroreflex resetting, 328–329, 329f central command, 328 ␣-receptor subtypes and blood flow in active muscles, 331 sensory feedback from active muscles, 328, 329f sympathetic vasoconstriction and metabolic vasodilation, 330, 330f baroreceptor control of, 325, 325f blood pressure and, 324–325 effect of glossopharyngeal and vagus nerve block on, 325, 326f characteristics of, 325 in aging, 332f, 332–333 in congestive heart failure, 333 in hypertension, 333 microneurography for measurement of, 324, 324f responses to chemoreceptor activation, 327–328, 328f responses to mental stress, 331–332 neurally mediated dilation in skeletal muscle, 331–332 responses to orthostatic challenge, 326–327, 327f site for recording of, 325 Muscle twitch, 1038 Muscle weakness, 1165 differential diagnosis in critically ill patients, 2029–2030 due to conduction block, 1142 in acute inflammatory demyelinating polyneuropathy, 2200 in amyotrophic lateral sclerosis, 1536 in chronic inflammatory demyelinating polyradiculoneuropathy, 2224, 2225, 2225t in cryptogenic sensory polyneuropathy, 2322 in diabetic lumbosacral radiculoplexus neuropathy, 1996–1999 in hereditary motor and sensory neuropathies, 1631–1635, 1632t in hydatidosis, 2170

Muscle weakness (Continued) in hyperthyroid neuropathy, 2041 in uremic neuropathy, 2023 rehabilitation for, 2625 related to immobility, 2629 scoring of, 1036–1037 Muscles of facial expression, innervation of, 1222, 1223 Muscular atrophy bulbar, 705 bulbospinal scapuloperoneal, 1139, 1615, 1616 X-linked, 705, 706, 1299, 1316, 1599–1600, 1605t, 1612 clinical features of, 1612, 1612f differential diagnosis of, 1612 electrophysiologic findings in, 1316, 1612 genetics of, 1316, 1536, 1548, 1600, 1612 pathology of, 1612 pattern of involvement in, 1139 sensory involvement in, 1316 vs. amyotrophic lateral sclerosis, 1536 peroneal, 1138, 1623, 1613, 1624. See also Hereditary motor and sensory neuropathy (HMSN). spinal. See Spinal muscular atrophy (SMA). Muscular dystrophy Becker, 1545 Duchenne dystrophin gene mutation in, 1545 electromyography in, 954f X-linked recessive inheritance of, 1548 Emery-Dreifuss, 1741–1742 vs. scapuloperoneal hereditary motor neuronopathy, 1616 Lama2 gene mutations in, 424 Musculocutaneous nerve, 1340f distribution of, 15f, 20f, 21, 1471–1472, 1472f involvement in vasculitic neuropathy, 2365t muscles innervated by, 956t nerve conduction studies of, 915, 1473 Musculocutaneous neuropathy, 1471–1474 clinical manifestations of, 1473 diagnosis of, 1473–1474 etiologies of, 1472–1473 iatrogenic, 1473 in children, 2727 prognosis for, 1474 traumatic, 1472 treatment of, 1474 Musculomotor neurons of enteric nervous system, 258–259 excitatory, 258 inhibitory, 258–259 Mutations. See Gene mutations. Mutilated foot mouse, 711 MVBs (multivesicular bodies) axonal, 47 of constricted axon segment, 59 of juxtaparanodal segment, 53 Myasthenia gravis (MG), 590, 833–834 anti-neurofilament antibodies in, 586 classification of, 833 clinical features of, 833 dysphagia, 2686, 2700 maternally transmitted transient, 2713 pathogenesis of, 833–834 pathology of, 834, 835f–837f single-fiber electromyography in, 987, 993f, 994–995, 995f during axonal stimulation, 995–996 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lv Myasthenia gravis (MG) (Continued) susceptibility to development of, 833 therapy for, 860 azathioprine, 639–640 vs. Guillain-Barré syndrome, 2202 vs. peripheral neuropathy, 1152 Mycobacterium avium-intracellulare (MAI), in HIV infection, 2139 Mycobacterium leprae, 2081, 2083–2084. See also Leprosy. adhesion to Schwann cells, 424, 574, 2099 host response to, 2084 mechanism of nerve damage due to, 2099 portals of entry for, 2082 susceptibility to infection with, 2082 Mycophenolate mofetil, 640–641 adverse effects of, 640–641 for chronic inflammatory demyelinating polyradiculoneuropathy, 2236 in diabetic patients, 1956 mechanism of action of, 640 pharmacokinetics of, 640 use in neuropathy, 641 Mycoplasma pneumoniae infection antiganglioside antibodies and, 591, 596 Bickerstaff’s brainstem encephalitis and, 2205 galactocerebroside antibodies and, 596 Guillain-Barré syndrome and, 590 Mydriasis, drug-induced and toxic, 212 Myelin abnormalities in hereditary motor and sensory neuropathy with P0 mutation, 1694–1696, 1695f age-related changes in, 487–491, 489f–492f early electron microscopic studies of, 3 factors affecting compaction of, 62 oral, for experimental autoimmune neuritis, 623 phagocytosis of, 1417 remodeling in normal nerves, 788–790 thickness of, axonal caliber and, 414–415 Myelin and lymphocyte (MAL) protein, 418f, 419t, 421 Myelin basic protein (MBP), 6, 419–420, 579 accessibility to antibodies, 577 age-related changes in, 491 in myelination, 418f, 419–420 isoforms of, 579 localization of, 578f mice deficient in, 579 Schwann cell expression of, 61 structure and function of, 419t Myelin bubbles (balloons), 489–491, 491f–492f Myelin folding, autosomal recessive HMSN I with, 1777 Myelin genes, regulation of for myelination, 421–423 signal transduction for, 422 transcription factors in, 421–423 Myelin lamella, 416 Myelin membrane accessibility to antibody, 577 composition of, 423 growth of, 3 lipid content of, 423, 577 proteins of, 6, 417–421, 577–584 Myelin protein zero (MPZ). See P0 (myelin protein zero). Myelin sheath, 42 abnormal, 63, 63f age-related thickness of, 37 contour of, 43 development of. See Myelination.

lvi Myelin sheath (Continued) functional gap junctions in, 1796 in IgM monoclonal gammopathy with neuropathy, 2262, 2263, 2265–2266, 2266f in wallerian degeneration, 4 less-dense line of, 62, 63f major dense line of, 62, 63f, 413 microscopic anatomy of, 62, 63f P0 in, 1685–1688 proteins of, 6, 417–421, 418f, 419t Myelin spiral, 3, 5 Myelin-associated glycoprotein (MAG), 6, 413, 575, 577, 580 accessibility to antibodies, 577 animal models of neuropathy induced by, 594 antibodies to, 1173, 2179t, 2187–2188 antibody testing for, 2188 disease pathogenesis and, 2188 electrophysiologic testing in patients with, 2187 in IgM neuropathy of monoclonal gammopathy of undetermined significance, 2187, 2255, 2266–2267, 2267f nerve biopsy in patients with, 2187–2188 axonal caliber and, 436 in myelination, 418f, 420, 547 isoforms of, 420, 546, 580, 580f localization of, 578f, 580 mice deficient in, 424, 546–548, 547f molecular weight of, 2266 paraproteinemic polyneuropathy and, 596–597 cytomegalovirus infection and, 590 Schwann cell expression of, 61 structure and function of, 419t, 420 Myelinated fibers (MFs), 42–62 age-related changes in proportion of, 37 axonal atrophy and secondary segmental demyelination of, 780–788, 784f–786f axonal caliber and, 435 axonal dystrophy of, 778 compound action potentials of, 1015–1016 conduction velocity and diameter of, 1015, 1016 decreased density in nonmalignant inflammatory sensory polyganglionopathy, 2313, 2314f in chronic inflammatory demyelinating polyradiculoneuropathy, 2239–2240, 2241f in diabetic neuropathy, 1965–1966, 1966f, 1967f in Fabry’s disease, 1894, 1896t, 1897f in hereditary motor and sensory neuropathy Déjérine-Sottas neuropathy, 1642, 1645f, 1647f type I, 1639, 1640f, 1643f–1644f, 1651f type II, 1643, 1650f with P0 mutation, 1695 in hereditary sensory and autonomic neuropathy type I, 1817–1819, 1818f–1821f in hereditary sensory and autonomic neuropathy type II, 1829f, 1830 in hereditary sensory and autonomic neuropathy type V, 1834–1836, 1835f–1836f in sarcoidosis, 2421 inner axonal zone of, 42–43 internodes of, 42, 43f–44f

Index Myelinated fibers (MFs) (Continued) microneurography of paresthesias and spontaneous activity in, 1007–1008 microscopic anatomy of, 42–62, 43f–46f morphometric measurements of, 43, 757–763 computer imaging method, 759–763 assessing variability in density, 762–763 hardware, 760 measurements from transverse sections, 761–762, 762t system approaches, 760–761 histologic processing, 757 photographic method, 759 samples sections of internodes and average diameter, 758–759, 761f section thickness and axon-myelin relationship, 757–758 myelin sheath of, 62, 63f normal numbers of, 763–765, 764f–771f of sural nerve, 37 organization of myelinated axons, 45–51, 46f, 48f, 50f outer Schwann cell zone of, 42–43 paranode-node-paranode region of, 43–45, 45f, 51–60 response to injury, 1463–1464 Schwann cells of, 60–62, 61f–62f study of teased fibers, 742–754. See also Teased fiber analysis. wallerian degeneration of, 773–776, 779f–781f Myelination, 5–6, 411–425 cell-cell signaling leading to, 357 effect on axonal caliber, 435f, 435–436 EGR2 mutations associated with defects of, 1707–1713 extracellular matrix, cellular receptors, and Schwann cell cytoskeleton in, 424–425 influence of axonal caliber on, 436–438, 438f lipids in, 423–424 molecular biology of, 417–425 morphology of, 411–417 myelin internode, 411–415, 412f–414f node of Ranvier, 415f–416f, 415–417 myelin gene regulation for, 421–423 potassium channels and, 104–105 proteins involved in, 417–421, 418f, 419t, 1684, 1685f. See also specific proteins. myelin basic protein, 418f, 419–420 myelin-associated glycoprotein, 418f, 420, 547 P0, 417–419, 418f, 1684, 1685–1688 transgenic animal studies of, 536–542, 575f–580f, 1572–1575 peripheral myelin protein 22, 418f, 419, 419t, 579 transgenic animal studies of, 542–543, 1561–1571 regulation of biosynthesis of, 423 transgenic animal models of abnormalities of, 536–548 signaling pathways and growth factors regulating, 361–362 timing of, 64 transcription factors involved in, 357–361, 358t, 359f Myelitis Epstein-Barr virus, 2120 herpes simplex virus, 2117 transverse, in children, 2717 Myelography, 1325 arachnoiditis induced by, 1332–1333 of brachial plexus lesion, 1350 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Myeloid cells, 574 Myelomeningocele, 1333 Myeloneuritis, Epstein-Barr virus, 2120 Myelopathy. See also Spinal cord disease. AIDS-related vacuolar, 1313–1314 cervical spondylotic, 1310–1313, 1328 diabetic, 1993 in critically ill patients, 2029 in Sjögren’s syndrome, 1308–1310 nerve conduction velocity in, 903 tropical myeloneuropathies, 2063–2077 vitamin B12 deficiency and, 1307–1308, 2054 vs. childhood Guillain-Barré syndrome, 2716–2717 vs. peripheral neuropathy, 1176 Myeloradiculopathy, cytomegalovirus, 2122 Myenteric plexus, 18, 252, 252f, 2627 congenital absence of enteric neurons in, 705 in regulation of gut motility, 2683 Myocardial infarction, painless, in diabetic patients, 1958 Myoclonic epilepsy with ragged red fibers (MERRF), 1941 Myoclonus, 1145 Myoenteric migrating complexes (MMCs), 1280 Myokymia, 942, 946, 947f, 1142–1143 in gold neuropathy, 2544 in gold toxicity, 1170 in peripheral nerve hyperexcitability syndromes, 1143 in radiation-induced nerve lesions, 2490 Myopathy(ies) critical illness, 2030 drug-induced amiodarone, 2554–2555 chloroquine, 2556 colchicine, 2556–2557 electromyography in, 939f, 943, 951–952, 954f, 955, 1176 in amyloidosis, 2434, 2434f in thyrotoxicosis, 2043 infantile, 2713 Miyoshi, 1175, 1176 Nonaka, 1175 sarcoid, 2419 vs. peripheral neuropathy, 1152, 1175–1176 Welander’s, 1152, 1175 Myosin light chain kinase (MLCK), 452 Myosin light chain phosphatase (MLCP), 452 Myotonia, electromyography in, 939f, 942t Myotonic discharge associated with electromyographic needle insertion, 940–941, 942f, 942t Myotonic dystrophy electromyography in, 942f trinucleotide repeats in, 705 vs. peripheral neuropathy, 1152 Myotubes, 136–137 primary, 136, 137 secondary, 136 Myxedema. See Hypothyroidism. Myxoma, nerve sheath, 2594–2595

Naegleria infection, 2161 Narcotic analgesics, for HIV-associated sensory polyneuropathies, 2135–2136 NARP (neuropathy, ataxia, and retinitis pigmentosa), 1941, 1942, 1946–1947 in children, 2722 Nasogastric tube placement, 2610

Index Nasopharyngeal tumors, 1200 Natural killer cell leukemia, 2496 Natural killer (NK) cells, 574 Nausea, 2610 Navajo neuropathy, 1837, 2021 N-cadherin in epilepsy, 466 in Schwann cell development, 343 NCAM (neural cell adhesion molecule), in schizophrenia, 467 NCLs. See Neuronal ceroid-lipofuscinoses (NCLs). NCP1 (Neurexin-Caspr-Paranodin), 417, 549. See also Contactin-associated protein (Caspr). NCSs. See Nerve conduction studies (NCSs). NDRG1 gene mutations in autosomal recessive hereditary motor and sensory neuropathy, 1770t, 1773–1774, 2714 in hereditary motor and sensory neuropathy IVD, 1630 NDS (Neurologic Disability Score), 1038 NE. See Norepinephrine (NE). Neck, innervation of, 18–23 Necrosis, neuronal, 699 Nefazodone, for neuropathic pain, 2642 NEFH gene mutations, in amyotrophic lateral sclerosis, 1534 NEFL gene mutations, in hereditary motor and sensory neuropathy IIE, 1629, 1732 Neostigmine, for gastroparesis, 2691 Nephrotic syndrome hepatitis C–associated neuropathy and, 2020 in amyloidosis, 2431f, 2433 treatment of, 2440 Nerve allografts, 1424 Nerve autografts, 1423 alternatives to, 1423–1424 allografts, 1424 conduits, 1424 for long interstump gap, 1424 rationale for use of, 1423 Nerve biopsy, 734–739, 1173–1174 at sites of nerve enlargement or enhancement, 737 histologic methods for, 740, 741f–742f in amyloidosis, 2430 in auditory neuropathy, 1261, 1262f in chronic inflammatory demyelinating polyradiculoneuropathy, 2234, 2242f–2244f, 2243 in chronic obstructive pulmonary disease–associated neuropathy, 2022 in congenital insensitivity to pain with anhidrosis, 886 in cryptogenic sensory polyneuropathy, 2325, 2325f in diabetic neuropathy, 1965–1966, 1966f in familial dysautonomia, 1832 in hepatitis B–associated neuropathy, 2020 in hepatitis C–associated neuropathy, 2019f, 2019–2020 in hereditary sensory and autonomic neuropathy, 1810–1811 in hereditary sensory and autonomic neuropathy type V, 1834–1837, 1835f–1836f in metachromatic leukodystrophy, 1849 in paraneoplastic subacute sensory neuronopathy, 2476 in pediatric polyneuropathies, 2712 in POEMS syndrome, 2458

Nerve biopsy (Continued) in primary biliary cirrhosis–associated sensory polyneuropathy, 2020–2021 in restless legs syndrome, 886 in sarcoidosis, 735f, 2421, 2421f–2422f in tropical ataxic neuropathy, 2068–2069 in vasculitic neuropathy, 736, 2369, 2369t–2370t indications for, 734–735 of cutaneous nerves without MRI enlargement or enhancement, 737 superficial peroneal nerve and other nerves, 739 sural nerve, 737–739, 738f of sural nerve, 37, 737–739, 738f peripheral nerves that may be biopsied, 737–739 symptoms after, 739 uses of, 733–734 vs. other neuropathic evaluations, 735–737 vs. study of necropsy specimens, 733–734 Nerve block effect on baroreceptor activity, 325, 326f interscalene, nerve injury during, 1517 Nerve conduction studies (NCSs), 899–937, 1154, 1166–1167 age effects on, 486, 907, 907f, 908t, 1176 as impairment measures, 1039–1040 averaging technique in, 900 clinical considerations for, 936–937 clinical values and limitations of, 907–908 collision technique for, 926–928, 928f criteria for conduction block in, 931–934 gender differences in, 904, 907 height effects on, 907 human reflexes and late responses in, 918–926 A wave, 926, 927f blink reflex, 918–919, 920t, 921t F waves, 921–926, 923f, 924t, 925t H reflex, 919–921, 922f, 922t masseter reflex, 921, 923t T reflex, 919 in acrylamide neuropathy, 2507 in amiodarone neuropathy, 2555 in arsenic poisoning, 2537 in axillary neuropathy, 1470 in brachial plexopathy, 1351 hereditary, 1761 immune, 2302–2303 obstetric, 1526, 1527t, 2731 in carbon disulfide intoxication, 2510–2511 in carpal tunnel syndrome, 912f, 926–928, 928f, 2040 in Charcot-Marie-Tooth disease type 1A, 1661–1663 in chloroquine neuropathy, 2556 in chronic inflammatory demyelinating polyradiculoneuropathy, 2234 in colchicine neuropathy, 2557 in common peroneal neuropathy, 1498 in critical illness polyneuropathy, 2027 in cryptogenic sensory polyneuropathy, 2323 in dapsone neuropathy, 2558 in Déjérine-Sottas neuropathy, 1624, 1625 in demyelinating polyneuropathies, 1167 in diabetic lumbosacral radiculoplexus neuropathy, 1963, 1963t, 2001, 2002t in diabetic neuropathy, 1955, 1978–1979 vs. nerve biopsy, 736 in distal spinal muscular atrophy, 1299 in disulfiram neuropathy, 2558 in eosinophilia-myalgia syndrome, 2571 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lvii Nerve conduction studies (NCSs) (Continued) in experimental autoimmune neuritis, 614f, 614–615, 627f in familial dysautonomia, 1832 in femoral neuropathy, 1490–1491 in giant axonal neuropathy, 1651 in gold neuropathy, 2544 in Guillain-Barré syndrome acute inflammatory demyelinating polyneuropathy, 2200 axonal forms, 2201 in hereditary motor and sensory neuropathy, 1636–1639, 1637f–1638f, 1639t, 1681, 1682, 1681, 1682 type II, 1717–1718 with P0 mutations early-onset neuropathy, 1690t–1691t, 1692 late-onset neuropathy, 1692, 1693t in hereditary motor neuronopathy, 1605 in hereditary neuralgic amyotrophy, 1671 in hereditary neuropathy with liability to pressure palsies, 1668 in hereditary sensory and autonomic neuropathy type II, 1828 in hereditary vs. acquired demyelinating disorders, 907–908 in hexacarbon neuropathy, 2514 in hyperthyroid neuropathy, 2043–2044 in hypothyroid polyneuropathy, 2041, 2041f–2042f in Kennedy’s disease, 1612 in Krabbe’s disease, 1856 in lead neuropathy, 2529 in Leigh disease, 2715 in leprosy, 2098–2099 in metachromatic leukodystrophy, 1849 in metronidazole neuropathy, 2560 in misonidazole neuropathy, 2561 in monoclonal gammopathy of undetermined significance, 2265 in multifocal motor neuropathy, 935f, 2284 in musculocutaneous neuropathy, 1473 in neuropathy associated with chronic hepatic disease, 2018 in nitrofurantoin neuropathy, 2561–2562 in nitrous oxide neuropathy, 2562 in nonmalignant inflammatory sensory polyganglionopathy, 2312–2313 in nucleoside analogue neuropathy, 2563 in paraneoplastic neuropathy demyelinating neuronopathy, 2478 subacute sensory neuronopathy, 2475–2476 in pediatric polyneuropathies, 2710, 2712 in perhexilene neuropathy, 2564 in phenytoin neuropathy, 2565 in podophyllin neuropathy, 2566 in POEMS syndrome, 2457–2458 in poliomyelitis, 1300 in porphyrias, 1890 in pyridoxine neuropathy, 2566 in radial neuropathy, 1477 in radiculopathy, 905, 1325 polyradiculopathy, 1327 in sciatic neuropathy, 1495 in suprascapular neuropathy, 1468 in suramin neuropathy, 2567 in tacrolimus neuropathy, 2568 in tarsal tunnel syndrome, 1504–1505 in thalidomide neuropathy, 2570 in tibial neuropathy, 1502

lviii Nerve conduction studies (NCSs) (Continued) in true neurologic thoracic outlet syndrome, 1360 in ulnar neuropathy, 1482 in uremic neuropathy, 2023–2024 in vasculitic neuropathy, 2366–2369 in vinca alkaloid neuropathy, 2572 in X-linked Charcot-Marie-Tooth disease, 1797–1798, 1798f late responses for evaluation of long pathways in, 934–936 needle electrodes in, 900 of anomalous innervations, 928–930 accessory deep peroneal nerve, 930, 930f Martin-Gruber anastomosis, 928–929, 929f of motor nerves, 900–903, 901f absent responses on, 903, 905f increased latency with normal amplitude on, 902–903 latency and conduction velocity in, 900–902, 902f reduced amplitude with normal latency on, 902, 903f–904f types of abnormalities of, 902, 902f waveform, amplitude, and duration on, 900 of sensory nerves, 901f, 903–905 latency and conduction velocity in, 904 types of abnormalities on, 904–905 waveform, amplitude, and duration on, 903–904 of specific nerves, 908–918 accessory nerve, 909 brachial plexus, 914–915, 915t common and deep peroneal nerves, 907, 908t, 916, 917t, 918f, 918t facial nerve, 908–909, 909t femoral nerve, 917, 919t lateral antebrachial cutaneous nerve, 916 lateral femoral cutaneous nerve, 918 lumbosacral plexus, 917 medial antebrachial cutaneous nerve, 916 median nerve, 901f–903f, 902, 905f, 908t, 909–910, 910t, 911f–912f, 913t, 914t musculocutaneous nerve, 915 phrenic nerve, 914 radial nerve, 911–914, 915f saphenous nerve, 917–918 superficial peroneal nerve, 916 sural nerve, 916–917, 919f tibial nerve, 905f, 908t, 916, 916f, 917t ulnar nerve, 907, 907f, 908t, 910, 913f, 913t, 914t phase cancellation in, 931, 933f physiologic variation among nerve segments in, 905–906 principles of, 899–900 reproducibility of, 936, 937f–938f segmental stimulation in short increments for, 934 sources of error in, 926 stimulus intensity and possible risk of, 900 surface electrodes in, 900 temperature effects on, 125, 906–907 temporal dispersion in, 931 pathologic, 931 physiologic, 931, 932f vs. nerve excitability studies, 114–115

Index Nerve degeneration. See also Axonal degeneration. regeneration after, 793, 794f–796f microfascicular, 793, 797f transgenic models of, 535–552 Nerve excitability studies, 113–126 automated recording of multiple nerve excitability properties, 119, 120f biophysical basis of current measures for, 115f, 115–119 recovery cycles, 117f, 117–118 strength-duration behavior, 113–114, 114f, 116f, 116–117 threshold electrotonus and currentthreshold relationship, 118f, 118–119 effect of ischemia on, 119–121, 121f history of, 113–114 impulse conduction in demyelinating neuropathies, 124–125 in amyotrophic lateral sclerosis, 126 in Bell’s palsy, 1226, 1227 in Charcot-Marie-Tooth disease type 1A, 126 in demyelinating neuropathies, 125–126 in diabetic neuropathy, 126 in Guillain-Barré syndrome, 126 in multifocal motor neuropathy, 122–124, 123f, 124f in neuromyotonia, 126 in uremic neuropathy, 124–125 limitations of, 126 membrane potential and, 119–121 modality and regional differences between axons, 121–122 length-dependent difference, 122 sensory vs. motor axons, 121–122 potential clinical role of, 126 vs. nerve conduction studies, 114–115 Nerve fascicles, 35 blood supply to, 42 cellular composition of, 37 collagen of, 37 connective tissues surrounding, 35–42, 36f Nerve fibers afferent. See Afferent fibers. age-related changes in density of, 37 collagenous sheaths of, 37 definition of, 35 density in pediatric patients, 37 epidermal. See Epidermal nerve fibers. fascicles of, 35 ganglionic, 235–236 intraepidermal, 2325–2326 myelinated, 42–62. See also Myelinated fibers (MFs). of sural nerve, 37 peripheral, development of, 4–6 postganglionic. See Postganglionic fibers. preganglionic. See Preganglionic fibers. Remak, 36, 62, 64, 65, 491–492 teased. See Teased fiber analysis. unmyelinated, 62–67. See also Unmyelinated fibers. Nerve growth factor (NGF), 174, 175–177, 378–380 age-related changes in, 495, 499, 499f, 500 binding with tyrosine receptor kinase A, 175 cellular internalization and transport of, 379–380 clinical applications of, 383 dependence of dorsal root ganglion neurons on, 175–176 discovery of, 377, 378–379 epidermal nerve fibers and, 872 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Nerve growth factor (NGF) (Continued) expression after axotomy, 177, 694 fast retrograde axonal transport of, 390 in diabetic neuropathy, 1970, 1978 in familial dysautonomia, 1833 in inflammation, 176, 177 in nociception, 177 in Schwann cell apoptosis, 355 molecules upregulated by, 177 neuroprotective effects of, 382–383 peptides downregulated by, 177 peripheral nervous system targets for, 379 recombinant human, for HIV-associated sensory polyneuropathies, 2136 role in lipid peroxidation, 514 signaling pathways of, 378f, 379, 381–382, 382f structure of, 378, 379 synthesis of, 175, 379 Nerve infiltration, 1383f, 2491–2493, 2492f in chronic lymphocytic leukemia, 2495–2496 in lymphoma, 1383f, 2491–2493, 2492f Nerve injury. See also Traumatic injury. arterial injury and, 1404f causalgia after, 1148 classification of, 1404–1405, 1512, 1512t axonolysis within intact Schwann cell tube, 1405 degenerative vs. nondegenerative, 1405 perineurial damage, 1405 transection and separation of nerve ends, 1405 clinical diagnosis of, 1512–1513 early signs of, 1513 electrophysiologic properties of dorsal root ganglion neurons after, 171 iatropathic, 1512, 1516–1521 causes of, 1517 during total hip arthroplasty, 1519–1520, 1520f from compression by hematoma, 1518 from drug injection, 1517 from false aneurysm and arteriovenous fistula, 1518, 1519f from interscalene block, 1517 in field of operation, 1518–1519, 1520f ischemic, 1517–1518, 1518f malpractice claims related to, 1516 to small nerves, 1521 overview of, 1403–1404 pain from, 1513–1515 causalgia, 1513–1514, 1514f central pain in brachial plexus injury, 1514–1515 neurostenalgia, 1514, 1515f posttraumatic neuralgia, 1514 perineurial, 39 plasticity with CNS neurons after, 1150 surgical indications for, 1511 timing of surgical repair of, 1410–1411 Tinel’s sign in, 1513 traumatic neuroma at site of, 2600–2601, 2601f Nerve injury repair mechanisms, 1403–1425 axonal regrowth, 1421–1423 across interstump gap, 1421f, 1421–1422 cellular outgrowth into regeneration chamber, 1422 precision of reinnervation, 1422–1423, 1423f

Index Nerve injury repair mechanisms (Continued) axonal responses, 1406–1408 degeneration, 1406–1407, 1407f delayed degeneration, 1408 role of Ca2⫹ in, 1407–1408, 1408f sealing of injured axons, 1406, 1407f factors affecting, 1404 from chronic axotomy and chronic denervation, 1410f, 1410–1413 chronic axotomy, 1411f, 1411–1412 chronic denervation, 1412, 1412f Schwann cell death and axon-independent Schwann cell, 1413, 1413f–1415f future of, 1425 in knockout mice, 1419 inflammation and wallerian degeneration, 1413–1417 cytokine network, 1413–1415, 1416f macrophages, 1415–1417 matrix metalloproteinases and chondroitin sulfate proteoglycans, 1418–1419, 1418f–1419f nerve autografts, 1423 alternatives to, 1423–1424 rationale for use of, 1423 neuropathic pain and, 1419–1421 chronic constriction injury model and role of TNF-␣ in, 1420–1421 spinal nerve ligation model of, 1420, 1420f of acutely denervated Schwann cells, 1408–1410 phenotype of, 1409–1410 Schwann cell proliferation, 1409 overview of, 1403–1404 recovery of Golgi tendon organ, 157–158 termino-lateral neurorrhaphy, 1424–1425 Nerve microenvironment, 667–668 Nerve regeneration, 793, 794f–796f. See also Axonal regeneration. aberrant regeneration of oculomotor nerve, 1198–1199 abnormal regeneration of vagus nerve, 1282–1283, 1283f after crush injury, 799 blood-nerve interface and endoneurial homeostasis in, 658–659, 658f–659f chromatolysis and, 686 electrical stimulation for promotion of, 1410, 1410f microfascicular, 793, 797f of proximal portions of nerves, 1296 of unmyelinated fibers, 67 skin biopsy in human models of, 888–890, 889f–890f teased fiber analysis of, 748, 751f Nerve sheath myxoma, 2594–2595 Nerve tests. See also specific tests. composite scores derived from, 971–972, 976–977, 979f as measures of severity of neuropathy over time, 981 functional abnormality when test values fall within normal range, 979–981 in genetic linkage analysis, 981 vs. individual attributes of nerve conduction abnormality, 977–979, 980f–982f compound action potentials of sural nerve in vitro, 1017–1026 cost-effectiveness of prospectively assessed normal controls for, 973 diagnostic yield in assessment of polyneuropathy, 1163–1164

Nerve tests (Continued) electromyography, 937–958, 1166 macro, 985, 991, 991f–992f, 997–999 scanning, 985, 990f, 990–991 single-fiber, 985–997, 986f–988f expressed as normal deviates, 972 expressed as percentiles, 971–972 microneurography, 324, 324f, 1003–1011 need for normal values in, 972 nerve conduction studies, 899–937, 1166–1167 selection of reference cohort for, 973 statistical methods for computing normal values for, 973–976 multiple explanatory variables, 975–976 multiple independent variables and multiple dependent variables, 976 single dependent variable, 973–975, 974t single independent variable, 975 tests for which reference values can be set, 972 variables influencing neuropathic end point in normal subject cohort for, 976, 976t–978t Nerve trunks age-related changes in, 486–492, 488f–492f mechanisms of, 496–498 impaired gliding capacity of, in compression neuropathies, 1399–1400 spiral bands of Fontana on surface of, 42 thickening of, 1151–1152 Nervus descendens cervicalis, 18 Nervus intermedius (nerve of Wrisberg), 1219, 1224f Netrin/DCC/UNC-5, 459–462, 460f modulation of signaling by, 461–462 cross talk with Robo/Slit signaling, 459, 462 intracellular cAMP levels, 461 proteolytic cleavage of DCC receptor, 462 sequestration of netrin, 462 receptor complexes of, 459–460 structure of DCC and UNC-5, 459–460 signaling by, 460–461 at transcriptional level, 461 coupling of Rac and PAK by Nck, 460–461 function of UNC-34/Mena in, 461 MAX-1 in DCC-independent signaling, 461 tyrosine phosphatase Shp-2 binding of UNC-5, 461 Neural cell adhesion molecule (NCAM), in schizophrenia, 467 Neural crest, 4, 346 cells derived from, 346 Notch signaling during development of, 346–347 origins of Schwann cells in, 341, 346–348 Neural tube defects, 1333–1334 Neuralgia geniculate, 1246 glossopharyngeal, 1214 migrainous, 1214 obturator, 1489 postherpetic, 870, 887, 1186, 2120 posttraumatic, 1514 superior laryngeal, 1284 trigeminal, 1213–1215, 1214t Neuralgic amyotrophy. See Brachial plexus neuropathy, immune (IBPN). Neurapraxia, 1404, 1512 brachial plexus injury with, 2631 nerve conduction studies in, 902–903 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lix Neuregulin-1 in Schwann cell development and survival, 60, 342, 346–347, 351–356 role in myelination, 362 Neurexin-Caspr-Paranodin (NCP1), 417, 549. See also Contactin-associated protein (Caspr). Neurilemmoma. See Schwannoma(s). Neurinoma. See Schwannoma(s). Neuritis Epstein-Barr virus, 2120 experimental autoimmune, 609–628. See also Experimental autoimmune neuritis (EAN). in leprosy reactions, 2095 localized interdigital, 1507, 2601 Neuroaxonal dystrophy. See Axonal dystrophy. Neuroborreliosis, 2111–2112. See also Lyme disease. experimental, 2113 Neurocardiology, 226 Neurochemistry of ganglionic pathways, 240–243 Neurocysticercosis, 2168 Neuroeffector junction, 233 Neurofascin, in sodium channel segregation, 97–98 Neurofascin 155, 418f, 419t, 420–421 Neurofascin 186, 415, 421, 549 Neurofibrillary tangles (NFTs) diseases associated with, 703 in ALS/parkinsonism-dementia complex, 1600 in Cockayne’s syndrome, 705 in progressive supranuclear palsy, 703, 704f Neurofibroma(s), 2586t, 2591–2593 atypical, 2593 cellular, 2593 cutaneous, 815f, 2591 histology of, 2591–2593, 2592f in elephantiasis neuromatosa, 2591 intraneural, 2591, 2591f Luse bodies in, 38 malignant peripheral nerve sheath tumor arising in transformation from, 2596, 2598f of brachial plexus, 1360–1361 pelvic, 1383f plexiform, 2591, 2592f Neurofibromatosis (NF), 2585 leprosy and, 2099 nerve thickening in, 1152 type 1 (NF1) diagnostic criteria for, 1265t ganglioneuromas in, 2595 genetics of, 1264 malignant peripheral nerve sheath tumors in, 2596 neurofibromas in, 2591 type 2 (NF2) diagnostic criteria for, 1265t genetics of, 1264 schwannoma in, 2586 vestibular, 1265f Neurofibrosarcoma. See Malignant peripheral nerve sheath tumor (MPNST). Neurofilaments (NFs), 343, 434 antibodies to, 586, 2190 as major determinant of axon size, 49 axonal caliber and, 434–443 effect on neurofilament spacing, 435–436, 437f in large-caliber myelinated fibers, 439–440, 440f

lx Neurofilaments (NFs) (Continued) axonal transport defects and aggregations of, 395 density of, 49, 440, 440f after axotomy, 690, 787–788 in P0-deficient mice, 537 expression of gene for, 442 regulation of, 442–443 in acrylamide neuropathy, 2507 in amyotrophic lateral sclerosis, 1295, 1305, 1306, 1538, 1539 familial ALS, 1534, 1599 transgenic animal models, 1588–1589 double transgenic models, 1589 in axonal dystrophy, 705, 778, 782f in carbon disulfide intoxication, 2510–2511 in dorsal root ganglion neurons, 164f, 165 in giant axonal neuropathy, 1736, 1737f in unmyelinated fibers, 65 in Werdnig-Hoffmann disease, 690 mice deficient in genes for, 548–549 microscopic anatomy of, 49 monomers of, 49 myelination and phosphorylation of, 435f, 435–436 protein subunits of, 434f, 434–435, 586 immunostaining for NF200, 173f transport of, 440 alterations in velocity of, 441f, 441–442 vinca alkaloid–induced accumulation of, 2572 Neurogastroenterology, 272 Neurogenic bladder. See Bladder dysfunction. Neurogenic bowel. See Bowel dysfunction. Neurogenic claudication, 1328 Neurogenic sarcoma. See Malignant peripheral nerve sheath tumor (MPNST). Neurogenins, 343 Neurokinin A, expression by dorsal root ganglion neurons, 184 Neurokinin receptors, 185 Neurokinin-1, in pain, 527 Neurologic Disability Score (NDS), 1038 Neurological Symptom Score (NSS), 1034t, 1034–1035 Neurolymphomatosis, 1383f, 2491–2493, 2492f Neuroma, 1405 acoustic, 921t ganglioneuroma, 2595 Joplin’s, 1506–1507 Morton’s (plantar), 1507, 2601 palisaded encapsulated, 2602 traumatic, 2600–2601, 2601f Neuromuscular choristoma, 2602 Neuromuscular electrical stimulation (NMES), 2623–2624 Neuromuscular junction (NMJ) conduction delay at, 902 gangliosides at, 590 postsynaptic region of, 831 presynaptic region of, 831 safety margin of transmission at, 831–833 synaptic space of, 831 Neuromuscular junction (NMJ) disease, 831–860 autoimmune disorders, 833–838 Lambert-Eaton myasthenic syndrome, 834–838 myasthenia gravis, 833–834 classification of, 832t congenital myasthenic syndromes, 838–860 choline acetyltransferase deficiency, 839–841 classification of, 839t

Index Neuromuscular junction (NMJ) disease (Continued) end-plate acetylcholinesterase deficiency, 842–844 fast-channel syndromes, 851–852 investigation of, 839t mutations causing AChR deficiency with/without minor kinetic abnormality, 852–855 paucity of synaptic vesicles and reduced quantal release, 841–842 postsynaptic syndromes, 844–847 resembling Lambert-Eaton myasthenic syndrome, 842 slow-channel syndromes, 847–851 sodium channel myasthenia, 857–859 syndromes caused by rapsyn deficiency, 855–857 with plectin deficiency, 859 with reduced quantal release and central nervous system symptoms, 842 therapy for, 860 vulnerability to autoimmune attack in Guillain-Barré syndrome, 590 Neuromuscular transmission organophosphate-induced abnormalities of, 2516 single-fiber electromyography studies of, 986–987 Neuromyopathy, 1170 Neuromyotonia acquired, 590, 597 cause of, 1143 definition of, 1143 electromyography in, 955 extra discharges in, 994 lymphoma and, 2495 muscle activity in, 1143–1144, 1144f myokymia in, 1143 nerve excitability studies in, 126 pharmacotherapy for, 1144 vs. Schwartz-Jampel syndrome, 1144 Neuronal antigens, 584–586 Neuronal ceroid-lipofuscinoses (NCLs), 1846, 1862–1863 clinical features of, 1863 genetic forms of, 1863, 1864t inheritance of, 1862 pathology of, 708, 1862–1863, 1862f–1863f Neuronal death, 695–702 apoptotic, 695–699, 697f–698f axotomy-induced, 695, 696f biochemical events in, 700f, 700–701 excitotoxic, 699 factors affecting magnitude of, 701–702 necrotic, 699 other types of, 699 Neuronal degeneration, 772, 776f–778f acute, 816–818 in herpes zoster, 817–818 in lysosomal storage diseases, 818 in poliomyelitis, 816–817 chronic, with axonal spheroids, axonal atrophy, myelin remodeling, and axonal degeneration, 818–819 Neuronal intranuclear inclusions (NIIs), 705–706 in Kennedy’s disease, 706 in neuronal intranuclear hyaline inclusion disease, 706 in spinocerebellar ataxias, 706 in transgenic mice, 705 in trinucleotide repeat disorders, 705 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Neuronopathy(ies), 1137, 2322 distribution and patterns of involvement in, 1139 motor involvement in, 1141 paraneoplastic, autoantigens in, 594 Neuronophagia, 695, 696f Neurons morphometry of somas of, 754–756 in intermediolateral column, 756 in motor neuron columns, 754–755, 756f–757f, 759f in spinal ganglia, 755–756, 758t, 759f–760f in sympathetic trunk ganglion, 756, 760f neurotransmitter phenotype of, 240 of autonomic ganglia, 235 of dorsal root ganglion, 2309 diseases affecting segments of, 2309 of enteric nervous system, 249, 251, 251f, 253 pathology of. See Pathology of peripheral nerves. postganglionic, 233–234 ganglionic transmission and rate of firing of, 238–239 preganglionic, 233–234 divergence and convergence of, 237 effectiveness of impulses from, 239 ganglionic transmission and rate of firing of, 238–239 specification of input to ganglia, 237 response to traumatic injury, 1403–1404 Neuron-specific tubulin, 343 Neuropathy(ies). See also specific types. acute motor axonal, 575, 2178t, 2201 acute motor-sensory axonal, 2200t, 2201 amyloid, 2430–2433 hereditary, 1921–1932, 2440–2444 auditory, 1259–1261 axillary, 1469–1471 brachial plexus hereditary, 1753–1766 immune, 1140, 1754, 1757, 1758, 1993, 2299–2306 clinical course of, 1141–1142 compression, 1140, 1140t, 1391–1400 confirmation of, 1154 definition of, 1137 Déjérine-Sottas, 1688, 1769 diabetic, 1137, 1951–1980 diagnosis of, 1154–1155 differential diagnosis of, 1152–1154, 1163–1176, 1152–1154 diphtheritic, 714, 2147–2150 distribution and patterns of involvement in, 1137–1140, 1138f–1140f, 1137–1140 drug-induced, 2553–2573 due to industrial agents, 2505–2520 epidemiologic studies of, 1181–1189 establishing etiology of, 1154–1155 functional selectivity of, 1141 giant axonal, 705, 1651, 1736–1738, 1782–1783, 1783f hereditary motor and sensory, 1623–1653, 2330 hereditary sensory and autonomic, 1809–1839 hyperthyroid, 2043–2044 hypoglycemic, 1965 interdigital, 1506–1507 lower limb, 1487–1508 common peroneal, 1496–1500 femoral, 1489–1491 genitofemoral, 1488, 1488f gluteal, 1493 iliohypogastric, 1488, 1488f

Index Neuropathy(ies) (Continued) ilioinguinal, 1488, 1488f interdigital, 1506–1507 Joplin’s neuroma, 1506–1507 Morton’s neuroma, 1507 lateral femoral cutaneous, 1492–1493 medial calcaneal, 1506 obturator, 1488–1489 pathophysiology of, 1487–1488 piriformis syndrome, 1496 plantar, 1505–1506 posterior femoral cutaneous, 1493 saphenous, 1491–1492 sciatic, 1493–1496 sural, 1507–1508 tarsal tunnel syndrome, 1503–1505 tibial, 1500–1503 metal-induced, 2527–2545 multifocal motor and sensory demyelinating, 2232–2233 multifocal motor neuropathy and conduction block, 122–124, 575, 935f, 1306–1307, 2179t, 2277–2294 Navajo, 1837, 2021 paraneoplastic, 714–715, 1141, 1168, 2316, 2471–2482, 2495 porphyric, 1883–1891 pupil involvement in, 213t small fiber sensory, 1146–1147, 1165–1166, 2321 symptoms and signs of, 1142–1152 tropical myeloneuropathies, 2063–2077 upper limb, 1463–1484 axillary, 1469–1471 long thoracic, 1464–1466 medial antebrachial cutaneous, 1483–1484 median, 1435–1454 musculocutaneous, 1471–1474 pathophysiology of, 1463–1464 radial, 1474–1478 suprascapular, 1466–1469 ulnar, 1478–1483 uremic, 1137, 2022–2026 vasculitic, 2335–2385 Neuropathy, ataxia, and retinitis pigmentosa (NARP), 1941, 1942, 1946–1947 in children, 2722 Neuropathy Impairment Score (NIS), 1038–1039, 1043f, 1096 in diabetic lumbosacral radiculoplexus neuropathy, 1999 Neuropathy Symptoms and Change (NSC) score, 1035, 1044f–1045f, 2012 Neuropathy Symptoms Profile (NSP), 1035 Neuropathy target esterase (NTE), 2518–2519 Neuropeptide Y (NPY) downregulation by nerve growth factor, 177 expression after axotomy, 185, 691 expression by dorsal root ganglion neurons, 185 in bladder, 301, 304 in cardiac neural control, 224 in chronic constriction model, 525 in enteric nervous system, 271 in ganglionic neurons, 240 Neuropeptides aging and expression of, 484–486 expression after axotomy, 690–692, 691f–692f expression by dorsal root ganglion neurons, 184–185

Neuropeptides (Continued) in enteric nervous system, 271 in ganglionic neurons, 240–243 actions of, 242–243 phenotype of, 240t, 240–241 Neuroprotection against antiretroviral toxic neuropathy, 2136, 2563 against ischemic-reperfusion oxidative injury, 524 effects of nerve growth factor, 382–383 immune-mediated, 635 Neuropsychological changes, 1168 Neurorrhaphy, termino-lateral, 1424–1425 Neurosarcoidosis, 2416. See also Sarcoidosis. pediatric polyneuropathy and, 2719 Neurostenalgia, 1514, 1515f Neurosyphilis, 1314–1316 amyotrophy in, 1315 arteritis in, 1315 meningovascular, 1314–1315 polyradiculopathy in, 1331 serologic tests for, 1316 spinal cord pathology in, 1315 spinal thrombosis in, 1315 stages of, 1314–1315 tabes dorsalis in, 1153, 1315, 2310, 2316 treatment of, 1316 Neurotensin, in ganglionic neurons, 242 actions of, 243 Neurothekeoma, 2594–2595 Neurotmesis, 1404, 1512 brachial plexus injury with, 1344, 2631 Neurotransmitters in cardiac neural control, 218, 219f, 220–225 in micturition reflex, 312 in sexual response cycle, 314–317 of enteric nervous system, 253 role in cell growth and differentiation, 377 secretion by ganglionic neurons, 233, 240 urothelial cell release of, 300 Neurotrophic factors, 377–383. See also specific neurotrophins. age-related changes in expression of, 495–499, 499f characteristics of, 173–174 clinical applications of, 383, 635 experimental autoimmune neuritis, 627–628, 635 Guillain-Barré syndrome, 635 definition of, 173, 377 epidermal nerve fiber requirements for, 872 expression after axotomy, 693–694, 1421 expression by dorsal root ganglion neurons, 173–181 effect on survival, 174 families of, 173–174, 378 functions of, 377, 378f in diabetic neuropathy, 1970, 1978 molecular weights of, 377 naming of, 377–378 production by immunocompetent cells, 635 receptors for, 173, 176t signaling pathways for, 381–383 p75NTR, 178, 379, 381–382, 382f tyrosine receptor kinase A, 175–177, 378f, 379, 381 tyrosine receptor kinase B, 177–178 tyrosine receptor kinase C, 178 sources of, 173, 174f–175f structures of, 378 timing of receptivity for, 377 tissue localization of, 377 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxi Neurotrophin 3 (NT-3), 174, 178, 378, 379, 380 age-related changes in expression of, 496, 499f binding with tyrosine kinase receptor C, 178 cellular expression of, 178, 380 clinical applications of, 383 expression after axotomy, 178, 694 expression in dorsal root ganglion neurons, 178 in diabetic neuropathy, 1978 in Schwann cell survival, 355 intracytoplasmic vacuolation induced by, 688 role in myelination, 361 structure of, 378 Neurotrophin 4 (NT-4), 174, 178, 378, 379, 380 expression in dorsal root ganglion neurons, 178 in inflammation, 178 structure of, 378 Neurotrophin 5 (NT-5), 174, 378 Neurotrophin 6 (NT-6), 378, 379 Neurotrophin 7 (NT-7), 378, 379 Neurotrophins, fast retrograde axonal transport of, 390 Neurovascular bundle, 20 Neurturin, 179 NF. See Neurofibromatosis (NF). NFAT (nuclear factor of activated T cell), 639 in netrin signaling, 461 NF-␬B. See Nuclear factor-␬B (NF-␬B). NFs. See Neurofilaments (NFs). NFTs. See Neurofibrillary tangles (NFTs). NG2, 549 NGF. See Nerve growth factor (NGF). NGFB gene, in hereditary sensory and autonomic neuropathy type V, 1810, 1811t, 1834–1837 NHL (non-Hodgkin lymphoma). See Lymphoma. Niacin deficiency, 2057, 2073 Nicotine, effect on anal sphincter, 286t Niemann-Pick disease, 1847f, 1861–1862, 1909, 1910t clinical features of, 1861 gene mutations in, 708, 1862 pathology of, 708, 1861f, 1861–1862 Nifedipine, for nocturnal hypertension in patients with neurogenic orthostatic hypotension, 2663 NIIs. See Neuronal intranuclear inclusions (NIIs). NIS (Neuropathy Impairment Score), 1038–1039, 1043f, 1096 in diabetic lumbosacral radiculoplexus neuropathy, 1999 NISP. See Nonmalignant inflammatory sensory polyganglionopathy (NISP). Nissl substance, in dorsal root ganglion neurons, 165 Nitric oxide (NO), 183 antinociceptive effects of, 527 effect on nerve blood flow, 670, 670t in cardiac neural control, 221, 225 in critical illness polyneuropathy, 2029 in excitotoxic neuron death, 701 in micturition reflex, 312 in rectoanal inhibitory reflex, 284 in sexual response cycle, 314–317 Schwann cell secretion of, 566–567 urothelial cell release of, 300 Nitric oxide synthase (NOS), 183, 225 activity after axotomy, 183, 693 endothelial, in penile erection, 315

lxii Nitric oxide synthase (NOS) (Continued) in critical illness polyneuropathy, 2029 in intramural ganglia of bladder, 301 in ischemia-reperfusion injury, 675 2-Nitro-4-trifluoromethylbenzoyl-1,3cyclohexanedione (NTBC), for tyrosinemia, 2713 Nitrofurantoin neuropathy, 2561–2562 Nitrogen mustard, 637 Nitroglycerin, for nocturnal hypertension in patients with neurogenic orthostatic hypotension, 2663 Nitrous oxide neuropathy, 2562–2563 NK (natural killer) cells, 574 NMDA receptors. See N-Methyl-D-aspartate (NMDA) receptors. NMES (neuromuscular electrical stimulation), 2623–2624 NMJ. See Neuromuscular junction (NMJ). NO. See Nitric oxide (NO). Nociceptin, expression by dorsal root ganglion neurons, 187 Nociceptors, 1005–1006, 2638–2640 A␤ fibers, 168t, 169, 2638 activation thresholds for, 165 A␦ fibers, 168t, 169, 870, 1005, 2638 C fibers, 168t, 169, 870, 1005–1006, 2638 effects of hyperpolarization of, 1007, 1007f–1008f itch fibers, 1006 low-threshold, 1007 mechano-insensitive, 1006 modulation of discharge frequency by axon, 1007, 1007f–1009f polymodal, 1006 sensitization in erythromelalgia, 1009–1010, 1010f warmth detection by, 1005 cutaneous, 168t, 169, 871–872 definition of, 165 dorsal root ganglia neurons, 163, 165–166, 168t, 169 chemical phenotype of, 191 electrical membrane properties of, 170–171 after peripheral nerve injury, 171 during chronic peripheral inflammation, 171 low-threshold mechanoreceptors vs. nociceptors, 170–171 non-nociceptive neurons, 171 mechanisms of neuropathic pain, 2638–2640 microneurography of, 1005–1006 nerve growth factor and, 176–177 polymodal, 166 “silent,” 168t, 169 Nodal gap, 50f, 56 metallophilia of matrix substance of, 56 NG2 of, 56 Nodes of Ranvier, 3, 36, 788 architecture of, 415f–416f, 415–417 dystroglycan in, 546, 549 axonal caliber at, 435 axonal compartment of, 57 definition of, 42 development of, 415 function of, 415 gangliosides at, 590 microscopic anatomy of, 42–43, 43f, 45f, 51, 55–57, 58f Schwann cell compartment of, 56–57 Schwann cell: axon membrane ratio of, 57, 57f

Index Nodes of Ranvier (Continued) sodium channels at, 55, 56, 96f, 96–98, 415, 577 transgenic mouse studies of, 536, 549–550, 551f Caspr deficiency, 550 contactin deficiency, 550 galactocerebroside and sulfatide deficiency, 550 Nodose ganglion, 1277, 1278, 1279 Nodular prurigo, 888 Nodules of Nageotte, 702, 702t in Friedreich’s ataxia, 706 in paraneoplastic neuropathies, 714 subacute sensory neuronopathy, 2476 in toxic neuropathies, 715 Nonaka myopathy, 1175 Nonenzymatic glycation, in diabetic neuropathy, 1974–1976, 1975f Non-Hodgkin lymphoma (NHL). See Lymphoma. Nonmalignant inflammatory sensory polyganglionopathy (NISP), 2309–2317 clinical patterns of, 2311 course of, 2311–2312 differential diagnosis of, 2314–2317, 2315t electrophysiologic findings in, 2312–2313 historical review of, 2310–2311 idiopathic, 2315t laboratory findings in, 2312 pathology of, 2313–2314, 2314f radiologic features of, 2312 Sjögren syndrome and, 2311, 2312, 2313, 2314f, 2315t symptoms and signs of, 2311 treatment of, 2317 Nonsteroidal anti-inflammatory drugs for orthostatic hypotension, 2677 topical, for postherpetic neuralgia, 2644 Nonsystemic vasculitic neuropathy (NSVN), 2339, 2340t, 2341t, 2355t, 2363, 2363t outcome of, 2383–2385 treatment of, 2375t Noradrenaline. See Norepinephrine (NE). Norepinephrine (NE) aging effects on levels of, 332, 332f biosynthesis of, 1117f–1118f effect on nerve blood flow, 670, 670t fast anterograde axonal transport of, 389 Horner’s syndrome and, 203, 207–208 in cardiac neural control, 223f, 223–224 in chronic constriction model, 525 in enteric nervous system, 271 in iris innervation, 203 in micturition reflex, 312 in neurogenic orthostatic hypotension, 2672–2673 measuring plasma levels of, 1116–1118, 1120f measuring total body and regional spillover of, 1118–1119 measuring urinary levels of, 1118 secretion by sympathetic neurons, 233, 240 synthesis of from dihydrophenylserine, 2678, 2678f from tyrosine, 2672, 2673f Nortriptyline for postherpetic neuralgia, 2645 for radicular low back pain, 2641 NOS (nitric oxide synthase), 225 activity after axotomy, 693 Nosocomial infections, 2615–2616, 2616t Notalgia paresthetica, 888 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Notch, in peripheral gliogenesis, 346–347 NPY. See Neuropeptide Y (NPY). Nr-CAM, 415, 549 NRTIs (nucleoside reverse transcriptase inhibitors), 2131–2136, 2134f–2135f, 2563 NSC (Neuropathy Symptoms and Change) score, 1035, 1044f–1045f, 2012 NSP (Neuropathy Symptoms Profile), 1035 NSS (Neurological Symptom Score), 1034t, 1034–1035 NSVN (nonsystemic vasculitic neuropathy), 2339, 2340t, 2341t, 2355t, 2363, 2363t outcome of, 2383–2385 treatment of, 2375t NT3. See Neurotrophin 3 (NT-3). NTBC (2-nitro-4-trifluoromethylbenzoyl1,3-cyclohexanedione), for tyrosinemia, 2713 NTE (neuropathy target esterase), 2518–2519 NTS (nucleus tractus solitarius), 218, 218f, 2671, 2672f Nuclear factor of activated T cell (NFAT), 639 in netrin signaling, 461 Nuclear factor-␬B (NF-␬B), 381 advanced glycation end product formation and, 514 calcineurin activation of, 639 glucocorticoid effects on, 636 in Schwann cell development, 342 role in myelination, 360–361 Nucleoside reverse transcriptase inhibitors (NRTIs), 2131–2136, 2134f–2135f, 2563 Nucleus ambiguus, 218, 218f Nucleus pulposus, 1324f herniation of, 1324 hypalgesia and, 12 spinal nerve root compression from, 1325–1326 Nucleus tractus solitarius (NTS), 218, 218f, 2671, 2672f NUDEL, in schizophrenia, 467 Numb chin syndrome, 1209, 2600 Numbness, 1146, 1165 facial, due to trigeminal nerve lesions, 1209–1213, 1210t in carpal tunnel syndrome, 1438–1440 in chronic inflammatory demyelinating polyradiculoneuropathy, 2224 in cryptogenic sensory polyneuropathy, 2322 Nutritional deficiency, 2051–2058 after gastric surgery, 2057 epidemiologic studies of, 1185 of copper, 2058 of niacin, 2057 of thiamine, 2051–2053 of vitamin B6, 2056 of vitamin B12, 2053–2055 of vitamin E, 2055–2056 tropical malabsorption and malnutrition in soldiers during World War II, 2067–2068 tropical myeloneuropathies due to, 2071–2072 Cuban epidemic neuropathy, 2063, 2066 Strachan’s syndrome, 2057, 2064 tropical ataxic neuropathy, 2069–2071 tropical malabsorption, 2072 tropical malabsorption-malnutrition, 2071–2072 vitamin deficiencies, 2072 Nutritional management in intensive care unit, 2610–2611 Nystagmus in vestibular neur(on)itis, 1262 tests for, 1256

Index O4, 61 OAEs. See Otoacoustic emissions (OAEs). Obesity, carpal tunnel syndrome and, 1443 OBPP. See Obstetric brachial plexus palsy (OBPP). Obstetric brachial plexus palsy (OBPP), 1362–1363, 1525–1528, 2632, 2730–2733 clinical classification of, 1525, 1525t clinical features of, 2730–2731, 2732f course and prognosis for, 1526 degree of nerve fiber injury in, 1363 diagnostic investigation of, 1363, 1525, 1526–1527, 1527t, 2731, 2733f etiology of, 1362–1363, 1525 gender distribution of, 1343, 1362 Horner’s syndrome and, 2731 incidence of, 1362, 1525–1526 laterality of, 1343, 1362 operative repair of, 1363, 1527–1528, 2731–2733 prognosis for, 1363, 1526, 2731 rehabilitation for, 2632, 2733 risk factors for, 1362, 1526, 2731 Obstetric lumbosacral plexus injury, 1376, 1376f, 1379f, 1380–1381 Obturator internus, nerve to, 28 Obturator nerve accessory, 28 distribution of, 16f, 27–28, 1488f muscles innervated by, 957t, 1376, 1488–1489 Obturator neuralgia, 1489 Obturator neuropathy, 1488–1489 Occipital nerve greater, 31 lesser, 18 third, 32 Occupational factors, in carpal tunnel syndrome, 1446, 1446t Oct6 in Schwann cell development, 342, 345 role in myelination, 357–360 Octreotide for gastroparesis, 2691 for orthostatic hypotension, 1958t, 2660t, 2663 Ocular muscles, in myasthenia gravis, 833 Ocular palsy, 1192, 1195–1203 clinical features of, 1195–1197 due to congenital cranial nerve defects, 1197–1198 due to infections and toxins, 1202–1203 due to neoplasms, 1199–1200, 1200f due to vascular diseases and aneurysms, 1201–1202 in diabetes, 1201–1202 traumatic, 1198 Ocular symptoms, 1168. See also Eye. Oculoauriculovertebral dysplasia, 1211 Oculocardiac reflex, 1282 Oculoleptomeningeal amyloidosis (OLMA), 1924 Oculomotor nerve, 1191–1202 aberrant regeneration of, 1198–1199 blood supply of, 1195 clinical features of third nerve palsy, 1195 course of, 1193–1195, 1194f histology of, 1195, 1196f in diabetes, 804, 1201–1202, 1960–1961 muscles innervated by, 1191–1192 neoplasms affecting, 1199–1200 nuclei of, 1192f, 1192–1193

Oculomotor nerve (Continued) traumatic injury of, 1198 vascular diseases affecting, 1201 ODSS (Inflammatory Neuropathy Overall Disability Sum Score), 1044–1045, 1047t Ofloxacin, for leprosy, 2103 Ogilvie’s syndrome, 2692 OH. See Orthostatic hypotension (OH), neurogenic. OHCs (outer hair cells), 1253 noise-induced damage of, 1268 testing function of, 1256 Olfactory ensheathing cells, 341 Oligodendrocytes, 411 Id proteins in differentiation of, 354 transitional, 69 Oligosaccharide glycoconjugates expression after axotomy, 693 expression by dorsal root ganglion neurons, 183 Oligosaccharidoses, 1864–1865 in children, 2722 Olivopontocerebellar degeneration, 1259 OLMA (oculoleptomeningeal amyloidosis), 1924 Omeprazole, 2609 Onchocerciasis, 2155t, 2171 clinical features of, 2171 epidemiology of, 2171 etiologic agent in, 2163f, 2171 life cycle of, 2171 ocular changes in, 2171 pathology of, 2171 peripheral neuropathy in, 2171 transmission of, 2171 Onion bulb formation CMT1A duplication and, 1661, 1662f in acromegalic neuropathy, 2045, 2046f in Andermann syndrome, 1738 in chronic inflammatory demyelinating polyradiculoneuropathy, 2239, 2240f, 2242, 2242f, 2243f in hereditary motor and sensory neuropathies, 1624, 1636, 1639, 1641f–1643f, 1642, 1646, 1649f, 1651t, 1695 in inflammatory demyelinating disorders, 806, 807f–808f in metachromatic leukodystrophy, 1849 in Refsum’s disease, 1870, 1871f in Roussy-Lévy syndrome, 1624 vs. pseudo-onion bulb formation in intraneural perineurioma, 2593, 2594f Onuf’s nucleus, 282, 286, 308 in amyotrophic lateral sclerosis, 1538 in poliomyelitis, 1538 OPA1 mutations, 1744 Ophthalmic nerve, 1207, 1208f Ophthalmoplegia chronic progressive external, 1941 drug-induced, 1202 in acute inflammatory demyelinating polyneuropathy, 2200 in Fisher syndrome, 1202, 2185, 2199 in mitochondrial diseases, 1941 in multiple sclerosis, 1203 in pediatric Guillain-Barré syndrome, 2716 in Tolosa-Hunt syndrome, 1202 OPIDP. See Organophosphate-induced delayed peripheral neuropathy (OPIDP). Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxiii Opioid receptors, 185–186 Opioids endogenous in bladder, 305 in cardiac neural control, 221–222 in micturition reflex, 312 for insulin neuritis, 1955 for neuropathic pain, 2645–2646 for restless legs syndrome, 1145 Oppenheim disease, 1604t OPs. See Organophosphates (OPs). OpSite surgical dressing, 1972 Optic atrophy hereditary motor and sensory neuropathy with, 1625t, 1626, 1627t, 1649–1650, 1743–1744, 1771 in angiostrongyliasis, 2164 in onchocerciasis, 2171 Optic nerve, 12 Optic neuritis, retrobulbar, in Cuba, 2065 Optic neuropathy in Cuban epidemic neuropathy, 2066 in Guillain-Barré syndrome, 2200 in sarcoidosis, 2418 in Tanzania, 2066–2067 in vitamin B12 deficiency, 2054 Leber’s hereditary, genetics of, 1550, 1940, 1941 paraneoplastic, anti-neurofilament, 586 Oral lesions, 1167 Organ of Corti, 1253, 1254f Organophosphate-induced delayed peripheral neuropathy (OPIDP), 717, 2515–2519 clinical features of, 2516–2517 electrophysiology of, 2518 laboratory findings in, 2517 pathogenesis of, 2518–2519 pathology of, 2517–2518 prognosis for, 2517 Organophosphates (OPs), 2515–2519 childhood Guillain-Barré syndrome vs. poisoning with, 2717 chronic OP-induced neuropsychiatric disorder, 2517 clinical features of cholinergic reactions to, 2516 effects on axonal transport, 397t toxicology of, 2515–2516 tropical myeloneuropathies due to, 2069t, 2075 uses and sources of exposure to, 2515 Ornithine decarboxylase, 690 Oropharyngeal dysphagia, 2633, 2685–2687 causes of, 2685–2686 clinical assessment of, 2687 evaluation of, 2687 in multiple sclerosis, 2700 in myasthenia gravis, 2686, 2700 in Parkinson’s disease, 2695 management of, 2687 postpolio, 2695 Orthopedic disease, vs. cryptogenic sensory polyneuropathy, 2331 Orthopedic procedures, brachial plexus injury from, 1369–1370 Orthoses, 2623, 2629–2630 ankle-foot, 2623 for brachial plexus injury, 2631 hand, 2623, 2630 knee-ankle-foot, 2623 thoracic-lumbar-sacral, 2623 wrist-hand, 2623

lxiv Orthostatic hypotension (OH), neurogenic, 1150, 2653–2663, 2671–2681 autonomic function testing in, 1104 blood pressure response to head-up tilt, 1112–1113, 1115–1116, 1116t orthostatic stress tests, 1123 response to prolonged tilt, 1123 cardiovascular autonomic failure and, 2671–2673 causes of, 2654 clinical features of, 2654–2655, 2655t gut dysmotility, 2696 course and prognosis for, 2655–2656 definition of, 2653–2654 epidemiology of, 2654 grading scale for, 2654, 2654t immobility and, 2629 in acute inflammatory demyelinating polyneuropathy, 2200 in amyloidosis, 2431, 2431f familial, 2444 treatment of, 2439, 2444 in diabetic patients, 1150, 1957, 1958 muscle sympathetic nerve activity responses to, 326–327, 327f pathophysiology and pathogenesis of, 2656 prevalence of, 2654, 2654t treatment of, 1958t, 1960, 2626, 2630, 2656–2663, 2673–2681 ␤-blockers, 2677 caffeine, 2676 clonidine, 2661, 2677 compression garments, 2658–2659, 2674 cyclooxygenase inhibitors, 2677 dietary, 2658 dihydroergotamine, 2661, 2678 dihydrophenylserine, 2677–2678, 2678f dopamine antagonists, 2661, 2678 erythropoietin, 2661, 2676–2677 for early morning symptoms, 2662 for periods of orthostatic decompensation, 2662 for postprandial symptoms, 2662–2663, 2674 goals of, 2656–2657, 2657t in hospital, 2662 in patients with supine hypertension, 2663, 2674 nonpharmacologic, 2657–2659, 2673t, 2673–2674 patient education, 2657t, 2657–2658, 2673 pharmacologic, 2659–2662, 2660t, 2660f–2661f, 2674–2681 physical countermaneuvers, 2659, 2659t, 2673 postural adjustment, 2658 pyridostigmine, 2661–2662, 2678 resistance training, 2659 somatostatin, 2677 sympathomimetic agents, 2659–2660, 2660f–2661f, 2675–2676, 2676t tyramine, 2678 volume expansion, 2658, 2659, 2674–2675, 2675t water drinking, 2658, 2659, 2662, 2674 yohimbine, 2661, 2677 Osteoarthritis, carpal tunnel syndrome and, 1446 Osteoma of internal auditory canal, 1264 Osteomyelitis of skull base, facial nerve paralysis and, 1235–1236 pediatric brachial plexopathy and, 2733, 2734f

Index Osteopetrosis, 1264 facial nerve paralysis and, 1236 Osteoporosis, corticosteroid-induced, 2382–2383, 2383t Osteosclerotic myeloma. See POEMS syndrome. Otic ganglion, 17, 234, 17 Otitis media, facial nerve paralysis and, 1236 Otoacoustic emissions (OAEs), 1256 in auditory neuropathy, 1260 in hereditary sensorimotor neuropathy, 1257f–1258f in patient with bilateral eosinophil granulomas of petrous bones, 1266f–1267f Otoferlin, 1259, 1261f Outer hair cells (OHCs), 1253 noise-induced damage of, 1268 testing function of, 1256 Ovarian plexus, 281 Overall Disability Sum Score (ODSS), 1044–1045, 1047t Overuse injury, carpal tunnel syndrome due to, 1446, 1446t Ox-2 (CD200), 575 Oxaliplatin, trigeminal nerve lesions due to, 1211 Oxcarbazepine, 2643 for diabetic neuropathy, 1971 Oxford Community Diabetes Study, 1957 Oxidative stress, 509–527 diabetic neuropathy and, 511–519, 1970, 1976 antioxidants for, 516t, 516–518, 517f evidence in non-neural tissue in human subjects, 511–512 experimental diabetes, 512 mechanisms of, 512–514 advanced glycation end products and nuclear factor-␬B, 514 auto-oxidative lipid peroxidation, 512–513, 514f excessive lipolysis, 514 growth factor deficit, 514 inflammation, 514 ischemia/hypoxia, 512, 513f polyol pathway overactivity, 514 protein kinase C, 514 pathogenesis of endothelial dysfunction in, 518, 518f pathogenesis of Schwann cell injury in, 519 pathogenesis of sensory neuropathy in, 518–519, 519f–520f peripheral nerve evidence of, 514–516 increased diabetic pro-oxidant status, 515–516 increased free radical generation, 515 reduced free radical defenses, 514–515, 515t free radical biology, 509–511, 510f. See also Reactive oxygen species (ROS). indices of, 513, 514f mechanisms leading to, 511 nerve ischemia and, 519–524 adrenergic contribution to, 520 effects on blood-nerve barrier, 520 endoneurial contents and, 521 footprints of, 521 genetic susceptibility to, 523–524 inflammatory response and, 522, 524f molecular pathogenesis of, 524 neuropathology of, 521–522, 523f neuroprotection against ischemia-reperfusion injury, 524, 526f Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Oxidative stress (Continued) reperfusion injury, 521 role of cytokines in, 522–523, 525f overview of, 509 peripheral nerve disease and, 511 protective mechanisms against, 510–511, 511t Oxybutynin, for detrusor hyperreflexia, 2663 Oxycodone, for postherpetic neuralgia, 2641, 2645 Oxygen delivery to nerve, 668–669, 669f Oxygen requirements of nerve, 668 Oxytalan, 37, 38 Oxytocin, 218

P0 (myelin protein zero), 6, 577–579, 1171, 1682–1688 accessibility to antibodies, 577 age-related changes in, 491 antibodies to, 2188–2189 as adhesion molecule, 417, 1682, 1684–1685, 1687f as marker of Schwann cell development, 343, 345 functions of, 1572 gene for, 1682 hereditary motor and sensory neuropathies related to, 579, 1681–1701 classic phenotype of, 1692–1694 differential diagnosis of, 1699 epidemiology of, 1688 genetics of, 1688 in early-onset neuropathy, 1688–1692, 1689f, 1690t–1691t, 1694–1695 in late-onset neuropathy, 1692, 1693t, 1695–1696 management of, 1699–1701 pathogenesis of, 1696–1699 pathology of, 1694–1696, 1695f phenotypic variability of, 1694 Roussy-Lévy syndrome, 1624 type II, 1683t, 1725–1727, 1726f–1727f in chronic inflammatory demyelinating polyradiculoneuropathy, 595 in developing myelin sheath, 1685–1688 in experimental autoimmune neuritis, 593, 611t in myelination, 417–419, 418f, 1684 induction of autoimmune neuritis in mice by, 574 localization of, 578f, 1682, 1686f mutations of, 1683f, 1688–1694 in trembler mouse, 704 on polyacrylamide gel electrophoresis, 577, 578f phosphorylation of, 1682–1684 posttranslational modifications of, 1682 Schwann cell expression of, 6, 61, 419, 1682 structure and function of, 419t, 577–579 structure of, 1682 transgenic mouse studies of, 536–542, 579, 1572–1575 epitope-tagged, toxic gain-of-function model, 1574f, 1574–1575 heterozygous deficiency, 537–540, 539f–541f homozygous deficiency, 536–537, 537f–538f, 1573 overexpression, 540–542, 542f, 1573, 1573f, 1573t

Index P1. See Myelin basic protein (MBP). P2, 6, 577, 579 in experimental autoimmune neuritis, 574, 593–594, 611t in human inflammatory demyelinating polyradiculoneuropathies, 595–596 P2X1 receptors, in bladder, 301 P2X3, 187t, 188 P2Y1, 187t, 188 p21-activated kinase (PAK) in actin depolymerization, 452 in netrin signaling, 460–461 in semaphorin signaling, 464 p75NTR, in neurotrophin signaling, 178, 379, 381–382, 382f PACAP (pituitary adenyl cyclase–activating protein) expression by dorsal root ganglion neurons, 185 in bladder, 305 in micturition reflex, 312 Pacemaker for carotid sinus hypersensitivity, 1277 for glossopharyngeal neuralgia, 1276 Pacinian corpuscles, 40, 42f, 341, 484, 773f, 1005, 1064, 2309 Pack palsy, 1357–1358 Paclitaxel neuropathy, 2490, 2490t, 2498t, 2568–2570 Pain, 1146, 1165–1166 congenital insensitivity to, with anhidrosis, 704, 886 disinhibition of central mechanisms of, 1150 experimental models of, 527 from nerve injury, 1513–1515 causalgia, 1513–1514, 1514f central pain in brachial plexus injury, 1514–1515, 2631 neurostenalgia, 1514, 1515f posttraumatic neuralgia, 1514 gate theory of, 2623 glossodynia, 1288 in brachial plexus neuropathy hereditary, 1753, 1754 immune, 2300 in cryptogenic sensory polyneuropathy, 2322 in diabetic lumbosacral radiculoplexus neuropathy, 1996–1999 in diabetic neuropathy, 1953, 1954–1955 management of, 1971t, 1971–1972 in erythromelalgia, 2497–2498 in Fabry’s disease, 1893 in geniculate neuralgia, 1246 in glossopharyngeal neuralgia, 1274, 1276 in Guillain-Barré syndrome, in children, 2716 in hereditary motor and sensory neuropathy, 1631 type IB, 1701 in hereditary sensory and autonomic neuropathy, type I, 1816–1817 in herpes zoster, 2119 in leprosy, 2087–2088 in lumbosacral radiculoplexus neuropathy, 1999 diabetic, 1996–1999 in migrainous neuralgia, 1214 in nonmalignant inflammatory sensory polyganglionopathy, 2311 in postherpetic neuralgia, 2120 in postpolio syndrome, 2632 in radiculopathy, 2634 in superior laryngeal neuralgia, 1284

Pain (Continued) in trigeminal neuralgia, 1213 indifference to, 1811–1813 insensitivity of schizophrenics to, 1813 management of, 2626 neuropathic, 1148–1150, 2637–2647 after axotomy, 1419–1421 causalgia, 1148 chronic constriction injury model and role of TNF-␣ in, 1420–1421 disorders associated with, 2638t epidermal nerve fiber loss and, 870 mechanisms of, 1148–1150, 1149f, 2638–2640 central pain, 1150, 2638–2639 peripheral pain, 2638 postherpetic neuralgia, 2640 role of NMDA receptors, 524–527, 2640 methods for evaluation of, 2637 nonpharmacologic management of, 2637 pharmacologic management of, 2326, 2637–2638, 2640–2647 anticonvulsants, 2642–2644 antidepressants, 2641–2642 evidence base and clinical trials of, 2640–2641 in diabetic patients, 1971t, 1971–1972 opioids, 2645–2646 overview of, 2637–2638 research and changing clinical practice for, 2646–2647 topical agents, 2644–2645 tramadol, 2646 spinal nerve ligation model of, 1420, 1420f tumor necrosis factor-␣ upregulation in, 181 Pain scores, 1035 Painful arm and moving fingers syndrome, 1145 Painful legs and moving toes syndrome, 1145, 1175 PAK (p21-activated kinase) in actin depolymerization, 452 in netrin signaling, 460–461 in semaphorin signaling, 464 Palatal paralysis, 1279 Palatine nerve, 1223f Palatopharyngeal syndrome of Avellis, 1284 Palisaded encapsulated neuroma, 2602 Pallbearer’s palsy, 1359 Palmar digital nerves, 21 PAN. See Polyarteritis nodosa (PAN). Pancoast’s syndrome, 1361–1362 Pancreatic polypeptide response to sham feeding, 2695 vagal function and release of, 1280 Pandysautonomia, 1141 Papilledema, in POEMS syndrome, 2457 Paraganglioma, 2596 Paraglenoid cysts, 1467 Paralysis brachial plexus “classic” postoperative paralysis, 1363–1364 obstetric, 1362–1363, 1525–1528, 2632 diaphragmatic, 1629, 1733 due to conduction block, 1142 facial nerve, 1223–1238 acute, 1223–1236 congenital/pediatric, 1236–1238 in leprosy, 2091 in Möbius syndrome, 1225, 1236 in Sjögren’s syndrome, 1234 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxv Paralysis (Continued) jake leg, 2075 laryngeal, 1279 of extraocular muscles, 1200 of hand, 2630 of Müller’s muscle, 205 of serratus anterior muscle, 1465 palatal, 1279 pharyngeal, 1279 postoperative (postnarcosis), 1363–1364 rucksack, 1357–1358 “Saturday night,” 1141, 1142 tourniquet, 1517 tri-nerve, 1475 vocal cord, 1285 in hereditary motor and sensory neuropathy IIC, 1629, 1733 Paralytic ileus, 2610, 2616 Paraneoplastic cerebellar degeneration, 2495 Paraneoplastic encephalomyelitis, 1200 Paraneoplastic intestinal pseudo-obstruction, 260 Paraneoplastic lower motor neuron syndrome, 1302–1303 Paraneoplastic neuropathy, 714–715, 1141, 1168, 2316, 2471–2482, 2495 antibodies in, 2471–2472, 2472t autonomic, 2479–2481 autoantibodies in, 2480–2481 clinical features of, 2480 pathology of, 2480 demyelinating, 2477–2478 detection of underlying malignancy in, 2482 gastrointestinal symptoms of, 2698 historical descriptions of, 2471–2472 immunopathogenesis of, 2481–2482 in children, 2721 incidence of, 2472–2473 malignancies associated with, 2472, 2472t motor, 2479 pathology of, 714–715 sensorimotor, 2476–2477 sensory, 1141, 2178t, 2316 autoantibodies in, 2178t small fiber, 2479 subacute, 2473–2476 age at onset of, 2474 autoantibodies in, 584–586, 2473–2474, 2474f, 2475 clinical features of, 2474 course of, 2474 differential diagnosis of, 2475 electrophysiology of, 2475–2476 lymphoma and, 2495 pathology of, 2476, 2477f treatment of, 2476 underlying malignancy in, 2475 treatment of, 2482 vasculitic, 2374, 2377t, 2478–2479, 2493–2494 clinical features of, 2478 laboratory findings in, 2478 pathology of, 2405, 2478 response to treatment for, 2479 underlying malignancy in, 2478–2479 Paranode(s), 42, 415f–416f, 416–417, 549–550, 1684 axolemma of, 417, 549 axon size and, 417 function of, 417 in Caspr-deficient mice, 550 in hereditary motor and sensory neuropathy, 1640

lxvi Paranode(s) (Continued) subcellular components of, 55, 417, 549 adherens, 417 microtubules, 417 molecular composition of, 417, 549–550 septate-like junctions, 417, 549 terminal cytoplasmic spiral of, 55 transgenic mouse studies of, 536 Paranode-node-paranode (PNP) region, 43f, 43–45, 45f, 51–60 components of, 51 constricted (CON) segment of, 52, 57–60 internodal end region of, 53 juxtaparanodal segment of, 51, 53–55 morphometry of, in cat, 52t–53t node of Ranvier of, 55–57, 57f–58f node-paranodal apparatus, 60 paranodal segment of, 55 transition between juxtaparanodal and paranodal segments of, 51–52, 54f Paranodin. See Contactin-associated protein (Caspr). Paraplegia, carpal tunnel syndrome and, 1446 Paraplegin mutations, 1743 Paraproteinemic polyneuropathy in lymphoproliferative disorders, 2494 myelin-associated glycoprotein and, 596–597 cytomegalovirus infection and, 590 Parasitic infections, 2153–2174 affecting cranial nerves, 2161–2165 angiostrongyliasis, 2162–2164 free-living amebae, 2161–2162 amebic keratitis, 2162 trichinosis, 2164–2165 affecting nerves innervating visceral organs, 2156–2161 American trypanosomiasis, 2156–2160, 2157f–2158f schistosomiasis, 2160–2161 strongyloidiasis, 2161 affecting peripheral nerves, 2165–2171 coenuriasis, 2168–2169 cysticercosis, 2168 dracunculiasis, 2170–2171 giardiasis, 2167–2168 hydatidosis, 2169–2170 leishmaniasis, 2165–2167 onchocerciasis, 2171 geographic distribution of, 2154t–2155t Guillain-Barré syndrome and, 2171 cyclosporiasis, 2173 malaria, 2171–2172 tick paralysis, 2173–2174 toxoplasmosis, 2172–2173 in immunocompromised hosts, 2153 peripheral neuropathic lesions caused by, 2153, 2154t–2155t Parasympathetic ganglia, 233–235 neuropeptides in, 240t, 240–243 Parasympathetic nervous system, 15, 233 anatomy of, 17–18 cardiac, 218–220, 220f–221f communication with sympathetic system, 224–225 in abdomen, pelvis, and lower limb, 30. See also Vagus nerve. in head and neck, 22. See also Cranial nerves. to iris, 203–204, 206f defects of, 210f, 210–212 in lower urinary tract, 301–302 in thorax, 25. See also Vagus nerve.

Index Parasympathetic nervous system (Continued) in urogenital system, 314–315 supplying pelvic floor, 281f, 282 in defecation, 282 Paravertebral (chain) ganglia, 234 Paresthesias, 1148, 1165 during ischemic period, 121 in amyloidosis, 2431 in carpal tunnel syndrome, 1141, 1142, 1438–1439 after surgery, 1452 in chronic inflammatory demyelinating polyradiculoneuropathy, 2224 in cryptogenic sensory polyneuropathy, 2322 in diabetic neuropathy, 1953 in hyperthyroid neuropathy, 2040 in large fiber neuropathies, 1148 in leprosy, 2087 in peripheral nerve hyperexcitability syndromes, 1144 in restless legs syndrome, 886 in small fiber neuropathies, 1148 in uremic neuropathy, 2023 intermittent, 1141 mechanisms of, 1149 microneurography in, 1007–1008 postischemic, 1141 Parkinson dementia complex and amyotrophic lateral sclerosis from Guam, neurofibrillary tangles in, 703 Parkinson’s disease (PD) autonomic function testing in, 1104 clonidine–growth hormone stimulation test, 1127, 1127f SPECT imaging of sympathetic innervation to organs, 1119–1121, 1120f bladder hyperactivity in, 312 cardiac dysautonomia and, 226 experimentally induced, 312 gastrointestinal manifestations of, 2695 obstructed defecation in, 306 Paroxetine, for pain, 2642 Paroxysmal lacrimation, 1275–1276 PARP (poly(ADP-ribose) polymerase), 514, 518, 677 Parsonage-Turner syndrome. See Brachial plexus neuropathy, immune (IBPN). Parvulins, 638 Patellar plexus, 26–27 Pathology of peripheral nerves, 683–717, 733–823 age-related, 702t, 702–703, 703t axotomy-induced, 684–702 biochemical alterations, 689–695 chromatolysis, 684–687, 685f intracytoplasmic vacuolation, 687–689, 688f–689f neuron death, 695–702 cutaneous nerves, 869–891. See also Skin biopsy. contributions of skin biopsy to diagnosis, 883–888 methods for analysis of, 872–876 microscopy of, 876, 877f quantitation of, 876–883 skin anatomy and, 870–872, 871f debris removal from endoneurium, 791–792, 792f differential diagnosis based on, 1166, 1166t fiber alterations, 772–788 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Pathology of peripheral nerves (Continued) axonal atrophy, myelin remodeling, and degeneration, 780–788, 784f–786f axonal degeneration in severe demyelination, 777 axonal dystrophy, 705, 778–780, 782f–783f demyelinative internal strangulation, 777 neuronal degeneration, 772, 776f–778f wallerian degeneration, 772–776, 780f–781f in hereditary neuropathies, 703–706, 706 Charcot-Marie-Tooth disease, 703–704 defects in DNA repair, 705 disorders caused by trinucleotide repeat instability, 705 Friedreich’s ataxia, 705–706, 783 giant axonal neuropathy, 705, 778, 782f, 1736–1738, 1737f hereditary sensory and autonomic neuropathy types I through V, 704 Hirschsprung’s disease, 705 Kennedy’s disease, 706 neuroaxonal dystrophy, 705 neuronal intranuclear hyaline inclusion disease, 706 spinal muscular atrophy, 704–705 spinocerebellar ataxias, 706, 707f in infective and inflammatory neuropathies, 713–715 cytomegalovirus, 713 diphtheria, 714, 820, 2148–2149, 2148f–2149f HIV, 713–714 leprosy, 714, 2081–2105 lymphoma, 714 other viruses, 714 paraneoplastic neuropathy, 714–715 poliovirus, 713 sensory ganglionitis in Sjögren’s syndrome, 714 varicella zoster and herpes zoster, 713 in lysosomal storage diseases, 706–710 acute neuronal degeneration, 818 adult polyglucosan body diseases, 709–710 Andersen’s disease, 709 Batten disease, 708 Chédiak-Higashi syndrome, 710 Fabry’s disease, 709, 772, 777f, 818 Gaucher’s disease, 706–707 GM1 gangliosidosis, 707 GM2 gangliosidosis, 707–708 leukodystrophies, 710 mannosidosis, 709 mucolipidosis, 709 mucopolysaccharidosis, 708–709 Niemann-Pick disease, 708 Pompe’s disease, 709 Refsum’s disease, 710 Tangier disease, 710 in neurologic mutants, 710–712 dystonia musculorum, 711 inherited dog neuropathies, 712 mouse motor neuron degeneration, 711 sprawling mouse and mutilated foot rat, 711 wobbler mouse, 711 in neuromuscular junction disease, 831–860 autoimmune disorders, 833–838 Lambert-Eaton myasthenic syndrome, 834–838 myasthenia gravis, 833–834

Index Pathology of peripheral nerves (Continued) classification of, 832t congenital myasthenic syndromes, 838–860 choline acetyltransferase deficiency, 839–841 classification of, 839t end-plate acetylcholinesterase deficiency, 842–844 fast-channel syndromes, 851–852 investigation of, 839t mutations causing AChR deficiency with/without minor kinetic abnormality, 852–855 paucity of synaptic vesicles and reduced quantal release, 841–842 postsynaptic syndromes, 844–847 resembling Lambert-Eaton myasthenic syndrome, 842 slow-channel syndromes, 847–851 sodium channel myasthenia, 857–859 syndromes caused by rapsyn deficiency, 855–857 with plectin deficiency, 859 with reduced quantal release and central nervous system symptoms, 842 therapy for, 860 vulnerability to autoimmune attack in Guillain-Barré syndrome, 590 in neuropathies of metabolic disorders, 712–713 amyloidosis, 712–713, 809–810, 811f–814f, 1925, 1926f, 2432, 2443–2444 diabetes mellitus, 712 porphyria, 712 interstitial, 793–810 compressive, 797–800, 801f–802f due to intraneural injection of foreign substance, 805, 806f from edema and hemorrhage, 794–797, 798f–800f in amyloidosis, 712–713, 809–810, 811f–814f, 1925, 1926f, 2432, 2443–2444 in chronic immune sensory polyradiculopathy, 808, 809f–810f in inflammatory demyelinating disorders, 806–808, 806f–808f in leprosy, 714, 2081–2105 in multifocal motor neuropathy, 2286–2287, 2287f–2288f in sarcoidosis, 734f–735f, 808–809, 2421, 2421f–2423f ischemic, 800–805, 803f–805f tumor-related, 810, 2585–2600 nerve regeneration and, 793, 794f–796f microfascicular, 793, 797f parenchymatous, 812–823 neuronal (axonal), 812–819 acute axonal degeneration, 819 acute neuronal degeneration, 816–818 agenesis and maldevelopment, 816, 817f anterograde and retrograde effects, 816 axonal atrophy, myelin remodeling, and axonal degeneration, 819 axonal spheroids, distal axonal atrophy, myelin remodeling, and axonal degeneration, 819 by functional classes of neurons, 813–814

Pathology of peripheral nerves (Continued) by levels within neurons, 814f, 814–815 by segmental level, 812–813 chronic neuronal degeneration, axonal spheroids, axonal atrophy, myelin remodeling, and axonal degeneration, 818–819 classification of, 816 temporal selectivity of neurons, 815–816 wallerian degeneration, 819 of Schwann cells, 819–823 developmental abnormality, 819 immune demyelination, 820 in congenital tomaculous neuropathy, 823 in Déjérine-Sottas disease, 819 in inherited tendency to pressure palsy, 823, 824f in lead intoxication, 820–823, 821f–822f in lysosomal storage diseases, 819–820, 820f toxin-induced, 820 perineurial, 810–812 in macrodystrophia lipomatosa or hamartoma, 811–812 in neurofibromatosis, 811, 815f segmental demyelination and remyelination, 788–791 myelin remodeling in normal nerves, 788–790 vs. secondary demyelination, 790–791 study of, 733–770 artifacts of surgical removal and of histologic preparation for, 765–770 crush artifact, 769, 771f–772f ice crystal formation, 770, 772f teasing artifacts, 743–744, 746f, 769, 772f histologic methods for, 740–742, 741f–742f morphometry of neuron somas in, 754–756 morphometry of peripheral nerves in, 756–763 nerve biopsy for, 734–739 postmortem spinal cord and nerve removal and processing for, 739f, 739–740 Renaut corpuscles in, 770, 773f role of, 733–734 Schwann cell inclusions in, 771, 774f–776f teased fiber analysis for, 742–754 toxin-induced, 715–717 acrylamide, 717 amiodarone, 717 hexacarbons, 715 metals, 715–716, 716f organophosphates, 717 pyridoxine, 717 vincristine, 716–717 Patterns of involvement, 1137–1140, 1138f–1140f differential diagnosis based on, 1164–1165, 1165t PAX3, in Schwann cell development, 1708 Paxillin, 425 PBC (primary biliary cirrhosis), sensory neuropathy in, 2020–2021 PCBs (polychlorinated biphenyls), 2519 PCR. See Polymerase chain reaction (PCR). PD. See Parkinson’s disease (PD). PDE (phosphodiesterase), in penile erection, 315 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxvii PDGF (platelet-derived growth factor) in vasculitis, 2344 PFGF-BB in Schwann cell development and survival, 355–357 PECAM-1 (platelet/endothelial cell adhesion molecule-1), 2343 Pectoral nerve, 20 muscles innervated by, 956t Pediatric patients, 2707–2738 cardiac vagal tone in, 1282 cholestatic hepatic disease and vitamin E deficiency in, 2021 facial nerve paralysis in, 1236–1237 hereditary motor neuronopathies in, 1603, 1604t, 1605, 1605t infantile and juvenile neuroaxonal dystrophy in, 1744 mononeuropathies in, 2724–2730 differential diagnosis of, 2725 evaluation of, 2725 frequency of, 2724, 2724f lower extremity, 2727–2730 femoral nerve, 2730 lateral femoral cutaneous nerve, 2730 peroneal nerve, 2724f, 2729f, 2729–2730 sciatic nerve, 2724f, 2727–2729, 2728f sural nerve, 2730 tibial nerve, 2730 mechanisms of, 2725 multifocal, 2725 upper extremity, 2725–2727 axillary nerve, 2727 long thoracic nerve, 2727 median nerve, 1448, 1448t, 2724f, 2725f, 2725–2726 musculocutaneous nerve, 2727 radial nerve, 2724f, 2726–2727, 2727f suprascapular nerve, 2727 thoracodorsal nerve, 2727 ulnar nerve, 2724f, 2726 Navajo neuropathy in, 2021 nerve conduction velocity in, 907, 908t nerve fiber density in, 37 plexopathies in, 2730–2736 brachial plexopathy, 2730–2735 lumbosacral plexopathy, 2735–2736 polyneuropathies in, 2707–2724 age at onset of, 2708 approach to evaluation of, 2708–2712 clinical history, 2708–2710, 2711t diagnostic algorithms, 2709f–2710f electrodiagnostic testing, 2711–2712 laboratory testing, 2712 neurologic examination, 2710–2711 epidemiology of, 2707–2708 infantile, 2712–2715 amino acid disorders, 2715 autosomal recessive Charcot-MarieTooth disease, 2714 Charcot-Marie-Tooth disease type 2C, 2714 chronic inflammatory demyelinating polyradiculoneuropathy, 2713 congenital hypomyelinating neuropathy, 2713–2714 Déjérine-Sottas disease, 2713–2714 differential diagnosis of, 2713 globoid cell leukodystrophy, 2714 Guillain-Barré syndrome, 2712–2713 hereditary tyrosinemia, 2713 infantile botulism, 2713 infantile poliomyelitis, 2713

lxviii Pediatric patients (Continued) Leigh disease, 2715 lipid disorders, 2715 maternally transmitted transient myasthenia gravis, 2713 metachromatic leukodystrophy, 2714–2715 mitochondrial trifunctional protein and long-chain 3-hydroxyacyl coenzyme A dehydrogenase deficiencies, 2715 myopathies, 2713 spinal muscular atrophy, 2713 inherited vs. acquired, 2707–2708 later childhood, 2715–2724 abetalipoproteinemia and hypobetalipoproteinemia, 2722 acute sensory neuropathy or neuronopathy, 2717 ataxia-telangiectasia, 2724 carbohydrate-deficient glycoprotein syndrome, 2724 cerebrotendinous xanthomatosis, 2722 Chédiak-Higashi syndrome, 2722 chronic inflammatory demyelinating polyradiculoneuropathy, 2230–2231, 2718–2719 Cockayne’s syndrome, 2723 diphtheria, 2717–2718 Fabry’s disease, 2721 Friedreich’s ataxia, 2724 fucosidosis, 2722 giant axonal neuropathy, 2723–2724 globoid cell leukodystrophy, 2721 Guillain-Barré syndrome, 2715–2717 hepatic porphyria, 2723 hereditary motor and sensory neuropathies, 2721 hereditary sensory and autonomic neuropathies, 2723 hyperoxaluria, 2722 in connective tissue disease, 2719, 2720f infectious, 2719–2720 mannosidosis, 2722 medication-related, 2720 metabolic, 2720 metachromatic leukodystrophy, 2721 mitochondrial disorders, 2722 mucopolysaccharidoses, 2721–2722 nutritional, 2720–2721 other spinocerebellar ataxias, 2724 paraneoplastic, 2721 Refsum’s disease, 2722 Schindler disease, 2722 sialidosis type II, 2722 Tangier disease, 2722 toxic, 2720 tyrosinemia, 2723 xeroderma pigmentosum, 2723 X-linked adrenoleukodystrophy, 2722 PEG (percutaneous endoscopic gastrostomy), 2610–2611 Pelizaeus-Merzbacher disease, 1841 Pellagra, 2057, 2073 Pelvic diaphragm, 280 Pelvic floor. See also Anorectum. anatomy of, 280, 280f anorectal manifestations of dysfunction of, 286–292 balloon expulsion test for, 286–287, 287f fecal incontinence, 287–293, 290t, 293t obstructed defecation, 286–287

Index Pelvic floor (Continued) assessing function of, 2692 nerve supply to, 281f, 281–283 afferent innervation, 282–283 parasympathetic nerves, 282 regulation by central nervous system pathways, 282 sympathetic nerves, 281–282 Pelvic ganglion, 234, 30 Pelvic nerves, 301 Pelvic plexus, 234, 302, 234 Pelvis femoral neuropathy in, 1490 innervation of, 25–30 autonomic, 30–31 Penetrance, incomplete, 1546 Penicillamine, for lead poisoning, 2534–2535 Penicillin, for neurosyphilis, 1316 Penis dorsal nerve of, 28 erection of, 315–316. See also Sexual response cycle. central mechanisms of, 316 peripheral mechanisms of, 315–316 psychogenic, 316 PEP syndrome (plasma cell dyscrasia, endocrinopathy, and polyneuropathy). See POEMS syndrome. Peptides. See also Neuropeptides. inflammatory, 183 Percentile values for nerve tests, 971–972 Percutaneous endoscopic gastrostomy (PEG), 2610–2611 Perhexilene neuropathy, 2563–2564 Periarteritis nodosa, 2337. See also Polyarteritis nodosa (PAN). facial nerve paralysis and, 1234 Periaxin (PRX), 418f, 419t, 421, 580 antibodies to, 597 localization of, 1686f mutations in gene for, 424, 580, 1578 in autosomal recessive hereditary motor and sensory neuropathy, 1770t, 1776–1777, 1777f, 2714 in hereditary motor and sensory neuropathy IVF, 424, 580, 1630 protein encoded by murine gene for, 1578 Schwann cell expression of, 61, 1578 transgenic mice deficient in, 545–546, 546f, 1578–1579 behavior of, 1578–1579 electrophysiology of, 1579 histology of, 1579 sensory involvement in, 1579 Periaxonal space, 45, 46f, 55 Pericytes as antigen-presenting cells, 574 endoneurial, 37, 576 Perifascicular fluid, 652 Perikaryon, Schwann cell, 43, 44f Perineal body, 280 Perineal nerve, 281f, 282, 28 Perineurial arteries, 801 Perineurial window, 39, 1405 Perineurioma, 2586t, 2593–2594, 2594f intraneural, 2593–2594, 2594f soft tissue, 2594 Perineurium, 35, 39–41, 574, 576, 651 age-related decrease in compliance of, 656 anatomy of, 36f, 39–41, 40f, 41f, 652 as specialized connective tissue, 654 blood vessels traversing, 40 collagen of, 37, 38f, 39 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Perineurium (Continued) damage to, 1405 diffusion barrier properties of, 39, 40, 652–653 effect of perineurial pinch on blood flow, 669–670 embryologic origin of, 654 enzymes of, 39 fibroblasts of, 39 function of, 652 GLUT-1 transporters of, 654 in axonal degeneration, 40 in lysosomal disorders, 1846 injury to, 39 origin of cells of, 39 termination of, 40–41 tight junctions of, 38f, 39, 40 transperineurial permeation, 652, 653 Perinodal space, 56 Peripheral myelin protein 22 (PMP22), 577, 579, 1171 antibodies to, 2188–2189 association of calnexin with, 423 expression in non-neuronal tissues, 1561 functions of, 1561 gene for, 579 in Charcot-Marie-Tooth disease type 1A, 419, 542, 579, 596, 1659–1665, 1671–1673, 1775 increased PMP22 gene dosage, 1561, 1663–1665, 1664t transgenic animal models of, 543, 1561–1572, 1562t in experimental autoimmune neuritis, 593, 611t in hereditary neuropathy with liability to pressure palsies, 542, 579, 1564–1565, 1659, 1667–1673 in human inflammatory demyelinating polyradiculoneuropathies, 595 in myelination, 418f, 419, 419t, 579 induction of autoimmune neuritis in mice by, 574 localization of, 578f, 1686f point mutations of demyelinating neuropathies caused by, 1665–1667, 1666f transgenic models of, 1566–1571, 1567t regulation of biosynthesis of, 423 Schwann cell expression of, 61, 423, 579, 1561 after injury, 1409 as developmental marker, 346 structure and function of, 419t, 579 transgenic animal studies of, 542–543, 1561–1572 aggregate formation as source of toxicity and phenotype variability, 1572 deficiency in mice, 542–543, 1564–1566 behavioral effects of, 1566 electrophysiology of, 1566 histology of, 1566 intracellular trafficking abnormalities, 1571–1572 overexpression in mice and rats, 543, 1561–1564 behavioral effects of, 1561–1562, 1562f histology of, 1562–1564, 1563f inducible PMP22-overexpressing model, 1564, 1565f muscle pathology due to, 1564 rat model of progesterone therapy for, 1564

Index Peripheral myelin protein 22 (PMP22) (Continued) point mutations, 1566–1571, 1567t behavioral effects of, 1566–1568 electrophysiology of, 1571 histology of, 1568–1571, 1569f–1571f muscle pathology due to, 1571 Pmp22 expression and, 1571 protein aggregation, 1572 Peripheral nerve(s). See also Neurons. antigens of, 573–598. See also Antigens. blood flow to, 667–677 blood supply to, 800–801, 803f compression of, 14–15 development of nerve fibers, 4–6 distribution of, 12f, 12–14 cutaneous, in limbs, 14, 15f, 16f dermatomal, 12–14, 13f, 14f types of branches of, 14–15 early electron microscopic studies of, 3 general features of, 11–12 hyperexcitability syndromes of, 1143t, 1143–1144 electromyography in, 1144, 1144f pharmacotherapy for, 1144 immune responses in, 573–575 in abdomen, pelvis, and lower limb, 25–31, 27f in head, neck, and upper limb, 18–23, 20f–22f in thorax, 23–25, 24f microenvironment of, 667 molecular composition of, 575–590 morphometry of, 756–763 computer imaging of myelinated fiber composition, 759–763, 761f assessing variability in myelinated fiber density, 762–763 hardware, 760 myelinated fiber measurements from transverse sections, 761–762, 762t system approaches, 760–761 histologic processing, 757 photography of myelinated fiber composition, 759 sampled sections of internodes and average diameter, 758–759, 761f section thickness and axon-myelin relationship, 757–758 operating on, 1511–1528. See also Surgery. pathology of, 683–717, 733–823. See also Pathology of peripheral nerves. proximal portions of, 1295–1316 combined motor and sensory syndromes affecting, 1310–1316 AIDS-related vacuolar myelopathy, 1313–1314 bulbospinal muscular atrophy, 1306 cervical spondylotic myelopathy, 1310–1313 neurosyphilis, 1314–1316 implications for nerve regeneration and repair, 1296 motor syndromes affecting, 1298–1307 amyotrophic lateral sclerosis, 1304–1306 classification of, 1298 HIV infection, 1303 lower motor neuron disease, 1298–1300 multifocal motor neuropathy with conduction block, 1306–1307 other causes of, 1304 paraneoplastic lower motor neuron syndrome, 1302–1303

Peripheral nerve(s) (Continued) poliomyelitis, 1300–1301 postirradiation lower motor neuron syndrome, 1303 West Nile virus infection, 1301–1302 overview of pathologic processes affecting, 1296–1297 sensory syndromes affecting, 1307–1310 cobalamin deficiency, 1307–1308 Sjögren’s syndrome, 1308–1310 spinal cord anatomy and, 1297–1298 Peripheral nervous system (PNS) aging in, 483–501 apoptosis in, 695–699, 697f–698f autonomic nervous system, 15–18 development of axon–Schwann cell interactions in, 4–5 Schwann cells, 341–365 general features of, 11–18 glial cells of, 341 gross anatomy of, 18–32 immune reactions in, 559–569 transitional zone with central nervous system, 67–77, 68f. See also Central nervous system–peripheral nervous system transitional zone. Peripheral tissue insertions, 68 Peripheral vascular disease, 801 Peripherin, 343, 434 expression after axotomy, 690, 691f Pernicious anemia diagnosis of, 2054–2055 Schilling test for, 2054 vitamin B12 deficiency due to, 2053 Peroneal muscular atrophy, 1138, 1623, 1613, 1624. See also Hereditary motor and sensory neuropathy (HMSN). Peroneal nerve accessory deep, 29 nerve conduction studies of, 930, 930f as site for recording peripheral sympathetic traffic, 325 blood supply to, 801 common, 1376 clinical manifestations of neuropathies of, 1497 diagnosis of neuropathies of, 1497–1498 distribution of, 27f, 29, 1494f, 1496, 1496f, 1500f high neuropathy, 1497 nerve conduction studies of, 907, 908t, 916, 917t, 918f, 918t neuropathy at femoral neck, 1497 prognosis for neuropathy of, 1500 treatment for neuropathy of, 1499–1500 deep, 16f, 30, 1496f muscles innervated by, 957t, 1496–1497 nerve conduction studies of, 916, 917t, 918f, 918t neuropathy at femoral neck, 1497 in leprosy, 2089f, 2093 involvement in vasculitic neuropathy, 2365t morphometry of, 768f pediatric lesions of, 2729f, 2729–2730 superficial, 16f, 29, 1496f biopsy of, 739 muscles innervated by, 957t, 1497 nerve conduction studies of, 916 Peroxisomal disorders, 1845, 1865–1873 adrenoleukodystrophy and adrenomyeloneuropathy, 1866–1868, 1867f classification of, 1866 in children, 2722 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxix Peroxisomal disorders (Continued) infantile forms of, 1866 2-methylacyl-CoA racemase deficiency, 1872–1873 neonatal adrenoleukodystrophy, 1868–1869 Refsum’s disease, 1869–1871, 1871f–1873f infantile, 1872 ultrastructural changes in, 1866 Peroxynitrite, in critical illness polyneuropathy, 2029 Persephin, 178 Pes cavus, 1151, 1168 in children, 2710 in hereditary sensory and autonomic neuropathy type I, 1815f Pes planus, 1151 in children, 2710 Pesticide neurotoxicity carbamates, 2508–2510 organophosphates, 2515–2519 tropical neuropathies due to, 2063, 2069t, 2075 PET (positron emission tomography) to detect malignancy underlying paraneoplastic neuropathy, 2482 to evaluate central autonomic pathways, 1125–1126, 1126f Petrosal artery, 1220 Petrosal nerve, greater, 1224f superficial, 1222, 1223f PGBs (polyglucosan bodies), 709–710, 779–780, 783f Phagocytosis, 575 glucocorticoid inhibition of, 636 myelin, 1417 Phalen’s maneuver, in carpal tunnel syndrome, 1441 Pharyngeal paralysis, 1279 Pharyngeal-cervical-brachial variants, IgG binding to GT1a ganglioside in, 2178t, 2185 Phase cancellation, 931 duration-dependent, 931 model for, 931, 933f Phenotype, 1546 Phenylacetic acid mustard, 637, 638 Phenylephrine effect on anal sphincter, 286t for fecal incontinence, 292 Horner’s syndrome and, 208 Phenylpropanolamine for orthostatic hypotension, 2660t, 2661, 2676 for urinary incontinence, 2664 Phenytoin for Fabry’s disease, 1900 for neuromyotonia, 1144 for neuropathic pain, 2643 for trigeminal neuralgia, 2641, 2643 neuropathy induced by, 2564–2565 Pheochromocytoma ganglioneuroma and, 2595 in multiple endocrine neoplasia type 2B, 1805–1807 plasma catecholamines in, 1118, 1120f PHN. See Postherpetic neuralgia (PHN). Phosphodiesterase (PDE), in penile erection, 315 Phospholipid, 1905–1906 Phosphoramide mustard, 637 Photosensitivity, cutaneous, 1168 Phox2, 343 Phrenic ganglion, 31

lxx Phrenic nerve, 18–19 accessory, 19 muscles innervated by, 956t nerve conduction studies of, 914 paralysis of, immune brachial plexus neuropathy and, 2301 Phytanic acid excess, 1625t, 1626, 1866, 1869. See also Refsum’s disease. Pia mater, 1324f Pigmentary retinopathy, 1168 in Refsum’s disease, 1168, 1869 Pinocytosis, perineurial, 38f, 40 ␲ granules, in Schwann cells, 771, 774f–775f, 791 Piriformis muscle, 280f nerve to, 28 Piriformis syndrome, 1496 Pituitary adenyl cyclase–activating protein (PACAP) expression by dorsal root ganglion neurons, 185 in bladder, 305 in micturition reflex, 312 Pituitary tumors, 1199–1200 PKC (protein kinase C) in diabetic neuropathy, 1977 in oxidative stress, 514 role in P0 phosphorylation, 1683–1684 Plantar digital nerves, 29, 1505 Joplin’s neuroma, 1506–1507 Morton’s neuroma, 1507 Plantar nerve lateral, 16f, 29 distribution of, 1502f, 1503 in tarsal tunnel syndrome, 1504 muscles innervated by, 957t, 1505–1506 medial, 16f, 29 distribution of, 1502f, 1503 in tarsal tunnel syndrome, 1504 muscles innervated by, 957t, 1505 Plantar neuroma, 1507, 2601 Plantar neuropathies, 1506 Plasma cell dyscrasia, endocrinopathy, and polyneuropathy (PEP syndrome). See POEMS syndrome. Plasma exchange (PE), 641–642 adverse effects of, 641 for acute motor axonal neuropathy, 2184 for chronic inflammatory demyelinating polyradiculoneuropathy, 642, 2179t, 2230, 2235, 2237 in children, 2231, 2719 in diabetic patients, 1955 for Guillain-Barré syndrome, 641, 642, 2199, 2213 in children, 2716 in diabetic patients, 1955 for hepatitis C–associated neuropathy, 2020 for IgA and IgG paraproteinemias, 642 for immune brachial plexus neuropathy, 2306 for Lewis-Sumner syndrome, 2233 for monoclonal gammopathy and peripheral neuropathy, 2268–2269, 2269t for POEMS syndrome, 2463 for Refsum’s disease, 642 for vasculitic neuropathy, 2336–2337, 2376, 2380–2381 for West Nile virus infection, 1302 procedure for, 641 schedule of, 641 use in neuropathy, 641, 642

Index Plasmacytoma in POEMS syndrome, 2453–2454 solitary, 2259–2261, 2260t, 2271 Plasmacytosis, in amyloidosis, 2435 Plasmapheresis. See Plasma exchange (PE). Plasmodium infection. See Malaria. Plasmolipin, 421 Platelet-activating factor, in enteric nervous system, 271 Platelet-derived growth factor (PDGF) in vasculitis, 2344 PDGF-BB in Schwann cell development and survival, 355–357 Platelet/endothelial cell adhesion molecule-1 (PECAM-1), 2343 Platinum neuropathy, 2544–2545 clinical features of, 2544–2545 laboratory studies in, 2545 pathology of, 715, 2545 treatment of, 2545 Platysma muscle, innervation of, 1222 Plectin deficiency, congenital myasthenic syndrome with, 859 therapy for, 860 Plexiform neurofibroma, 2591, 2592f Plexiform schwannoma, 2588 Plexin. See Semaphorin/plexin. Plexus aortic, 25, 31, 281f Auerbach’s (myenteric), 18, 252, 252f, 2627, 2683, 18, 252, 252f, 18, 252, 252f, 2627 congenital absence of enteric neurons in, 705 in regulation of gut motility, 2683 brachial. See Brachial plexus. cardiac, 25, 220 carotid, 1224f internal, 23 celiac, 30, 31 cervical, 18–19, 31 coccygeal (sacrococcygeal), 30 enteric, 18 esophageal, 25 hypogastric inferior, 18, 31, 281, 281f superior, 30, 31, 281, 281f intermesenteric, 30, 31 internal carotid, 23 lumbar, 25–26 lumbosacral. See Lumbosacral plexus. Meissner’s, 18, 2627 mesenteric, 31 ovarian, 281 patellar, 26–27 pelvic, 234 prevertebral, 16, 17 abdominal, 30–31 thoracic, 25 prostatic, 31 pulmonary, 25 rectal, 31, 281, 281f sacral, 18, 28–30, 32 solar, 234 subepidermal neural, 871, 872, 879–880 submucosal, 252, 252f, 2683 subsartorial, 26 testicular, 281 ureteric, 281 uterovaginal, 281, 281f, 31 vertebral, 23 vesical, 281, 281f PLP (proteolipid protein), 61, 346 Plumbism, 2063 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

PMC (pontine micturition center), 311–312 PMP22. See Peripheral myelin protein 22 (PMP22). PMR (polymyalgia rheumatica), 2362, 2363 PNET (primitive neuroectodermal tumor), 2600 Pneumonia, nosocomial, 2615–2616, 2616t PNP. See Paranode-node-paranode (PNP) region. PNS. See Peripheral nervous system (PNS). P-NSS (Positive Neuropathic Sensory Symptom) score, 1035, 1035t Podophyllin neuropathy, 2565–2566 POEMS syndrome, 1168, 2255, 2264, 2265t, 2269t, 2453–2464 age at onset of, 2455 bone marrow and bone involvement in, 2458–2459, 2459f–2460f cardiovascular involvement in, 2460 Castleman’s disease and, 2453, 2454, 2455, 2459–2460 causes of death in, 2464 cerebrospinal fluid analysis in, 2453, 2457 clinical features of, 2453, 2455–2461, 2456t course and prognosis for, 2463f, 2463–2464 based on number of features at presentation, 2461f diagnosis of, 2461–2463 diagnostic criteria for, 2453, 2454t differential diagnosis of, 2462t, 2462–2463 chronic inflammatory demyelinating polyradiculoneuropathy, 2457–2458, 2462–2463 monoclonal gammopathy of undetermined significance, 2462 electromyography in, 2457–2458 endocrine abnormalities in, 2458 history of, 2453–2454 laboratory features of, 2461 nerve biopsy in, 2458 overview of, 2453 pathogenesis of, 2454–2455 pathology of, 2458 polyneuropathy in, 2457 pulmonary involvement in, 2460 renal involvement in, 2460–2461 skin changes in, 2457f, 2458 terminology for, 2453 treatment strategies for, 2463 Poiseuille equation, 668, 672 POLG gene mutations, 1946 Poliomyelitis, 1299, 1300–1301, 1332 amyotrophic lateral sclerosis and, 1535 cerebrospinal fluid analysis in, 1300 clinical features of, 1300 differential diagnosis of, 1153 Guillain-Barré syndrome, 1300, 2201 immune brachial plexus neuropathy, 2304 electrophysiologic testing in, 1300 epidemiology of, 1300, 2198 immunization against, 2713, 2199 infantile, 2713 loss of deep tendon reflexes in, 1300 macro electromyography in patients with history of, 993f, 997 mortality from, 1300 pathology of, 696f, 713, 772, 776f, 1300–1301 acute neuronal degeneration, 816–817 chromatolysis, 686, 1301 lower motor neuron atrophy, 1301 postpolio syndrome and, 2632–2633 recovery from, 1301 vaccine-related, 2713

Index Polyarteritis nodosa (PAN), 2337, 2352–2355, 2353t age distribution of, 2352 classic, 2339, 2352–2355 classification and definition of, 2338t, 2339, 2340t clinical features of, 2352 diagnostic criteria for, 2338 genetics of, 2363 hepatitis B–associated, 2020, 2335, 2336, 2349–2350, 2375t incidence of, 2340t, 2352 laboratory findings in, 2352–2355 microscopic form of, 2337, 2339, 2353t, 2355 antineutrophil cytoplasmic antibodies in, 2352, 2355 classification and definition of, 2338t, 2339, 2340t genetics of, 2363 incidence of, 2340t pediatric polyneuropathy and, 2719 relative prevalence of, 2341t pathology of, 803 pediatric polyneuropathy and, 2719 relative prevalence of, 2341t treatment of, 2336, 2375t Polychlorinated biphenyls (PCBs), 2519 Polycythemia vera (PV), 2497 Polyganglionopathy, inflammatory sensory malignant, 2315t nonmalignant, 2309–2317 clinical patterns of, 2311 course of, 2311–2312 differential diagnosis of, 2314–2317, 2315t electrophysiologic findings in, 2312–2313 historical review of, 2310–2311 laboratory findings in, 2312 pathology of, 2313–2314, 2314f radiologic features of, 2312 Sjögren syndrome and, 2311, 2312, 2313, 2314f, 2315t symptoms and signs of, 2311 treatment of, 2317 Polygenic inheritance, 1551 Polyglucosan bodies (PGBs), 709–710, 779–780, 783f Polymerase chain reaction (PCR), 1553–1554 for analysis of genes with large number of known mutations, 1556 for analysis of genes with large number of unique mutations, 1557–1559, 1558f for analysis of genes with limited number of mutations, 1554–1555, 1555f for analysis of short tandem repeats, 1552 multiplex, 1552 techniques in skin, 876 Polymorphisms, 1551–1552 of vascular endothelial growth factor, in familial amyotrophic lateral sclerosis, 1534–1535 restriction fragment length, 1551–1552 single-nucleotide, 1553 single-strand conformation, 1557 Polymorphonuclear leukocytes, 574, 575 in experimental autoimmune neuritis, 620 Polymyalgia rheumatica, carpal tunnel syndrome and, 1445 Polymyalgia rheumatica (PMR), 2362, 2363 Polymyositis electromyography in, 939f vs. Guillain-Barré syndrome, 1152

Polyneuropathy(ies), 1137. See also specific types. chronic inflammatory demyelinating polyradiculoneuropathy, 2221–2246 compressive mononeuropathies and, 1140, 1165 critical illness, 1141, 2017, 2026–2030 cryptogenic sensory, 2321–2331 demyelinating, 1166, 1166t diagnostic yield in assessment of, 1163–1164 differential diagnosis of, 1163–1176 approaches to, 1164–1167 symptoms and signs of neurologic disease, 1170, 1170t symptoms and signs of non-neurologic disease, 1167–1170, 1169t distal, 2096–2097 distal symmetrical, 1138, 1138f–1140f in HIV infection, 2129–2133, 2135–2136 familial amyloid, 1921–1932, 2330t, 2440–2444 Guillain-Barré syndrome, 2197–2213 immune-mediated, 2177–2190 paraproteinemic, 596–597, 590, 2494 patterns of involvement in, 1137–1140 complex distributions, 1139 cranial nerve involvement, 1139–1140 distal symmetrical, 1138, 1138f–1140f proximal symmetrical and upper limb–predominant, 1138–1139, 1140f, 1164–1165 pediatric, 2707–2724. See also Pediatric patients. with axonal degeneration, 1154, 1166 with segmental demyelination, 1154, 1166 Polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes. See POEMS syndrome. Polyol pathway hypothesis of diabetic neuropathy, 1969–1970, 1973–1974 Poly(ADP-ribose) polymerase (PARP), 514, 518, 677 Polyradiculoneuropathy, 1137 chronic inflammatory demyelinating, 2179t, 2221–2246 diabetic, 1993 in amyloidosis, 2430 Polyradiculopathy, 1137, 1327–1334 chronic arachnoiditis, 1332–1333 congenital anomalies causing, 1333–1334 cytomegalovirus, 2122 diabetic, 1332 differential diagnosis of, 1327 etiologies of, 1327–1328 in ankylosing spondylitis, 1334 in sarcoidosis, 1334 infectious causes of, 1331–1332 miscellaneous causes of, 1334 neoplastic, 1329–1331 radiation-induced, 1332 spondylotic, 1328–1329, 1329f toxic meningoradiculopathy, 1333 vasculitic, 1334 vs. anterior horn cell disease, 1327 Pompe’s disease, 1864 pathology of, 709, 1864 Pontine micturition center (PMC), 311–312 Popliteal artery, 801 Porphobilinogen deaminase–deficient mice, 551–552 Porphyria cutanea tarda, 1890 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxi Porphyric neuropathy, 1883–1891 differential diagnosis of, 1888–1889 electrodiagnostic studies in, 1890 gastrointestinal symptoms of, 2699 heme biosynthesis and, 1883–1884, 1884f drug effects on, 1884–1885 other factors affecting, 1885 in children, 2723 laboratory studies in, 1889t, 1889–1890 mechanism of neuronal damage in, 1885–1886 motor involvement in, 1141 mouse model of, 1886 pathologic changes in, 712, 1886 chromatolysis, 712 intracytoplasmic vacuolation, 688, 712 prognosis for, 1888, 1891 proximal distribution of, 1138, 1139 temporal profile of, 1167 treatment of, 1890–1891 upper limb–predominant, 1164 vs. Guillain-Barré syndrome, 1888, 1889 Porphyrin(s) drug effects on metabolism of, 1884–1885, 1887t glucose levels and metabolism of, 1885 hormonal influences on, 1885 in heme biosynthesis, 1883–1884, 1884f laboratory measurement of, 1889t, 1889–1890 structure of, 1884f Port-wine stains, 886 Porus acusticus, 1219 Positive Neuropathic Sensory Symptom (P-NSS) score, 1035, 1035t Positive sharp waves, 941, 942–943, 944f Positron emission tomography (PET) during sexual arousal, 317 during voiding, 312 to detect malignancy underlying paraneoplastic neuropathy, 2482 to evaluate central autonomic pathways, 1125–1126, 1126f Postganglionic fibers cardiac, 217 for abdomen and pelvis, 30 for head, neck, and upper limb, 23 for thorax, 25 of parasympathetic nervous system, 17 of sympathetic nervous system, 15–17, 17f Postherpetic neuralgia (PHN), 2119, 2120 analgesics for, 2120 gabapentin, 2640, 2641, 2644 oxycodone, 2641, 2645 topical agents, 2641, 2644–2645 epidemiology of, 1186 epidermal nerve fiber loss in, 870, 887 pain mechanisms in, 2640 skin biopsy in, 887 Post–median sternotomy brachial plexus lesion, 1364–1365 Postmortem nerve studies, 733–734 histologic methods for, 740–742 spinal cord and nerve removal and processing, 739f, 739–740 Postoperative (postnarcosis) paralysis, 1363–1364 Postpolio syndrome, 1299, 2632–2633 management of, 2632–2633 insulin-like growth factor I for, 381 pain of, 2632 symptoms of, 2632 gastrointestinal, 2695 Postprandial hypotension, 2662–2663, 2674

lxxii Postural tachycardia syndrome (POTS) autonomic function testing in, 1104 skin biopsy in, 887–888 Potassium, for thallium intoxication, 2543 Potassium buffering, 106 Potassium channels, 99–108, 189–191 antibodies to, 590 in Isaacs’ syndrome, 590, 597 peripheral nerve hyperexcitability syndromes and, 1143t, 1144 effects of ischemia on, 119–121 expression after axotomy, 190–191 expression by dorsal root ganglion neurons, 189–191 genes for, 95 in action potential repolarization, 95 in inflammation, 191 of juxtaparanodes, 54–55, 549–550, 577 on axons, 96f, 99–104 function of, 95, 99, 100f, 102–104 in constricted axon segment, 57–59 mechanisms for clustering of, 101f, 101–102, 103f molecular identity of, 99–101 on Schwann cells, 56, 98f, 104–108, 105f expression linked to myelinogenesis, 104–105 functions of, 106–108 molecular identity and localization of, 105–106 pore structure of, 95 voltage-gated, 190–191 background potassium channels, 190 calcium-activated potassium currents, 190 delayed rectifier currents, 190 fast transient currents, 190 inwardly rectifying channels, 190 M currents, 190 POTS (postural tachycardia syndrome) autonomic function testing in, 1104 skin biopsy in, 887–888 Prazosin, for detrusor hyperreflexia, 2664 Prednisone for amyloidosis, 2437–2438 for Bell’s palsy, 1230–1231 for chronic inflammatory demyelinating polyradiculoneuropathy, 2234–2235, 2236–2237 in children, 2231, 2719 for cryptogenic sensory polyneuropathy, 2327 for hepatitis C–associated neuropathy, 2020 for hereditary motor and sensory neuropathy, 1653 for lumbosacral radiculoplexus neuropathy, 2012 diabetic, 2011–2012 for monoclonal gammopathy and peripheral neuropathy, 2268–2269 for POEMS syndrome, 2463 for Ramsay Hunt syndrome, 1232 for sarcoidosis, 2423 for vasculitic neuropathy, 638, 2336–2337, 2378–2380 paraneoplastic, 2374 limiting adverse effects of, 2381–2382, 2382 Preganglionic fibers cardiac, 217 for abdomen and pelvis, 30 for head, neck, and upper limb, 23 for lower urinary tract, 308 for thorax, 25

Index Preganglionic fibers (Continued) of parasympathetic nervous system, 17–18 of sympathetic nervous system, 15–17, 17f age-related changes in, 493 Pregnancy carpal tunnel syndrome in, 1445 facial nerve paralysis in, 1234–1235 hereditary brachial plexus neuropathy and, 1756, 1757f, 1757–1758, 1764 Prenatal diagnosis of genetic disorders, 1549 Presacral nerve, 31, 281 Presbycusis, 1254 Pressure palsies. See Hereditary neuropathy with liability to pressure palsies (HNPP). Pressure sores, 2629 Presynaptic facilitation in enteric nervous system, 270 Presynaptic inhibition in enteric nervous system, 270–271, 271t Prevalence of peripheral neuropathy, 1181–1182, 1182t Prevertebral ganglia, 234–235 ganglionic transmission in, 243–245, 244f–245f Prevertebral plexus, 16, 17 abdominal, 30–31 thoracic, 25 Previsceral ganglion, 234 Primary biliary cirrhosis (PBC), sensory neuropathy in, 2020–2021 Primitive neuroectodermal tumor (PNET), 2600 Prion disease, lower motor neuron disease and, 1304 Prismatic inclusions, in metachromatic leukodystrophy, 1850 Probucol, for experimental diabetic neuropathy, 516, 516t Procaine, for pain, 2642 Procarbazine neuropathy, 2490, 2490t Profilin, 451, 451f Progesterone rat model for Charcot-Marie-Tooth disease type 1A therapy with, 1564 role in myelination, 361–362 stimulation of PMP22 expression by, 1665 Progesterone antagonist therapy, for hereditary motor and sensory neuropathy, 1653 Progressive supranuclear palsy (PSP), 703, 704f Promyelin stage, 411, 412f Pronase injection, intraneuronal, intracytoplasmic vacuolation induced by, 688 Pronator syndrome, 1453 Propantheline for carotid sinus hypersensitivity, 1277 for detrusor hyperreflexia, 2663 Propofol, for ventilator withdrawal, 2617 Proprioception aging and, 483 deficits of, 1165 Prosaptide, for HIV-associated sensory polyneuropathies, 2136 Prostaglandins as inflammatory mediators, 182 effect on nerve blood flow, 670t, 670–671 glucocorticoid effects on, 636 in ischemic-reperfusion oxidative injury, 520, 677 Prostatic plexus, 31 Protein kinase C (PKC) in diabetic neuropathy, 1977 in oxidative stress, 514 in P0 phosphorylation, 1683–1684 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Protein truncation test (PTT), 1557, 1558f Proteins. See also specific proteins. antibodies to, 597 cytoskeletal, changes after axotomy, 690, 691f myelin, 6, 417–421, 577–584 age-related changes in, 491 in myelination, 417–421, 418f, 419t, 1684 polyacrylamide gel electrophoresis of, 577, 578f regulation of biosynthesis of, 423 transgenic animal studies of abnormalities of, 536–548 neurofilament, 586 Proteoglycans in epilepsy, 466 in Luse bodies, 38 Proteolipid protein (PLP), 61, 346 Proteomics, 1545 Prothionamide, for leprosy, 2103 Proton pump inhibitors, 2609 PRX. See Periaxin (PRX). P-selectin, 2342 Pseudo anterior interosseous nerve syndrome, 1454 Pseudoathetosis, 2311 Pseudobulbar palsy, 1533 Pseudoephedrine, for orthostatic hypotension, 2676 Pseudo-Horner’s syndrome, 204 Pseudomyotonia, 943, 1143 Pseudo–onion bulb formation, 2593, 2594f Pseudosyringomyelic pattern, 1146–1147 Pseudotumor cerebri, 1201 facial nerve paralysis and, 1235 Psoas hematoma, 1378, 1379f Psoriasis, 886 PSP (progressive supranuclear palsy), 703, 704f Pterygopalatine ganglion, 17, 1223f–1224f, 17 Pterygopalatine nerve, 1223f Ptosis, in Horner’s syndrome, 205 PTT (protein truncation test), 1557, 1558f Pubococcygeus muscle, 280, 280f Puborectalis dyssynergia, 306 Puborectalis muscle, 280, 280f effect on anorectal angle, 284 paradoxical contraction of, 286 Pudendal nerve, 281f, 282, 28, 301, 304, 315 terminal motor latencies, in fecal incontinence, 288, 290t Pudendoanal reflex, 284 Pulmonary angiography, 2616 Pulmonary embolus, 2616 Pulmonary function testing, 2612f, 2612–2613 Pulmonary plexus, 25 Pulmonary stretch receptors, vagal, 1284, 1284f Pulse oximetry, 2613 Pupil(s), 203–213 age and size of, 208f, 208–209 anatomy of iris innervation, 203–204, 204f–206f parasympathetic, 203–204 defects of, 210f, 210–212 sympathetic, 203 Argyll Robertson, 205, 212 as measure of autonomic dysfunction, 212–213 cycle time of, 209 drug-induced and toxic mydriasis of, 212 effects of iris damage on light, near, and drug responses of, 209 efferent defects of, 210–213 equipment for testing function of, 212t in Adie’s syndrome, 205, 210–212, 211f

Index Pupil(s) (Continued) in Horner’s syndrome, 204–208, 207f in simple anisocoria, 209, 209f light-near dissociation of, 212 peripheral neuropathies with involvement of, 213t responses in diabetic patients, 1959 tonic, 210, 211f congenital nonprogressive sensory neuropathy with, 1837 unrest and hippus of, 209 Pure tone audiometry, 1255–1256 in hereditary sensorimotor neuropathy, 1257f in patient with bilateral eosinophil granulomas of petrous bones, 1266f Purkinje fibers, 217 Purpura, 1168 PV (polycythemia vera), 2497 Pyridostigmine, for orthostatic hypotension, 2660t, 2661–2662, 2678 Pyridoxine deficiency of, 2056 neuropathy induced by, 717, 2056–2057, 2566–2567 Pyriminil (Vacor), 2519

QOL. See Quality of life (QOL). QSART. See Quantitative sudomotor axon reflex test (QSART). QST. See Quantitative sensation testing (QST). QSWEAT unit, 1108 QTc interval prolongation, 226–227, 1150 in chronic obstructive pulmonary disease–associated neuropathy, 2022 in neuropathy associated with chronic hepatic disease, 2018 in primary biliary cirrhosis–associated sensory polyneuropathy, 2021 Quadrilateral space, 1469 Quadrilateral space syndrome, 1470 Quadruatus femoris, nerve to, 28 Quality of life (QOL), 1053–1060 assessing effect of interventions on, 1053 assessment in peripheral neuropathies, 1049, 1058–1060 diabetes-specific measures, 1058 generic measures, 1059–1060 HIV/AIDS–specific measures, 1059 rheumatoid arthritis–specific measures, 1058–1059 conceptualization of, 1054 definition of, 1033, 1054 dimensions of, 1055 key issues in peripheral neuropathy, 1060 measurement of, 1053 generic and specific measures, 1055–1056 mode of administration, 1056 related to health outcomes of impairment, disability and handicap, 1054 requirements of health outcome measures, 1056–1058 acceptability, 1057–1058 reliability, 1056–1057 responsiveness, 1057 validity, 1057 uses of measurements of, 1053–1054 who should provide information about, 1055 asking relatives, 1055

Quantitating neuropathic symptoms, impairments, and outcomes, 1031–1049 autonomic impairment, 1036, 1103–1127. See also Autonomic function tests. composite scores of neurologic signs, 1038–1041 electronic case report form of Clinical Neuropathy Assessment, 1040–1041, 1042f–1045f history of, 1038 MNSI questionnaire for diabetic neuropathy, 1040, 1041f nerve conduction tests as impairment measures, 1039–1040 Neuropathy Impairment Score, 1038–1039, 1096 Total Neuropathy Score, 1040 definitions related to, 1033–1034 disability scores, 1041–1048 fatigue, 1047–1048 general scores, 1041–1045, 1046t–1047t Health-Related Quality of Life and Clinical Outcomes Score in Peripheral Neuropathy, 1045–1046, 1048t morphometry of nerve or nerve fiber counts in skin biopsy, 1046 other outcomes, 1046 end points to be scored, 1034 handicap, 1048–1049 history of scoring neurologic signs, 1032–1033 muscle stretch reflexes, 1038 muscle twitch, 1038 muscle weakness/strength, 1036–1037 ordinal scales for, 1033 quality of life, 1049 sensory loss, 1037–1038 Inflammatory Neuropathy Sensory Sum Score, 1037, 1037t touch-pressure sensation, 1038 vibration sensation, 1038 surrogate measures for, 1033–1034 symptoms scores for, 1034–1036 autonomic scores, 1036 Neurological Symptom Score, 1034t, 1034–1035 Neuropathy Symptoms and Change Score, 1035 Neuropathy Symptoms Profile, 1035 pain scores, 1035 Positive Neuropathic Sensory Symptom Score, 1035, 1036f Total Symptom Score, 1035, 1035t volunteer symptoms, 1034 Quantitation of cutaneous nerves, 876–883 by tracing, 878, 878f correlation between epidermal nerve and sural nerve, 883 correlation with tests of unmyelinated nerve function, 882–883 dermal nerves, 879 epidermal nerves, 876–879 in skin blisters and epidermal sheets, 879, 880f normal epidermal nerve fiber density, 879, 879f, 880t manual counting, 878–879 morphologic changes in sensory neuropathy, 882 nerve counting rules, 877f, 877–878 nerves to arrector pilorum muscles, 881 nerves to arteries, 881 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxiii Quantitation of cutaneous nerves (Continued) nerves to hair follicles, 881–882 subepidermal neural plexus rating, 879–880, 881f sweat gland nerve density, 880–881, 882f utility of skin biopsy, 883 Quantitative sensation testing (QST), 1037–1038, 1063–1091, 1166 as educational tool, 1066 assessing sensation and vision and hearing, 1064–1065 commercially available devices and instruments for, 1068, 1071t criteria and standards for L-QST using natural stimuli, 1068–1077 algorithm of testing and finding threshold, 1075–1077, 1076f–1078f ambient environment, 1068 baseline conditions of stimulus, 1072 calibration of stimulus magnitude, 1074–1075, 1074f–1075f comparison of results among different devices and systems, 1072 4, 2, 1 step testing algorithm with null stimuli: cooling detection threshold, 1070 4, 2, 1 step testing algorithm with null stimuli: vibration detection threshold, 1070, 1081–1082, 1081f–1082f ideal stimuli and testing algorithms, 1072 nonrepeating ascending algorithm with null stimuli: heat-pain threshold, 1070–1071 null stimuli, 1073–1074 patient instructions, 1069–1071 stimulus waveform, 1073, 1073f subject or patient factors, 1068 cutaneous sensory receptors, 1064, 1064f defining threshold for, 1065 electrical stimulation for, 1065 factors affecting sensation, 1065 for hyposensitivity and hypersensitivity, 1066–1067, 1089 in cryptogenic sensory polyneuropathy, 2324 in diabetic neuropathy, vs. nerve biopsy, 736 in hereditary brachial plexus neuropathy, 1761 in lumbosacral radiculoplexus neuropathy, 2002, 2003t diabetic, 1996, 2002, 2003t in medical practice, 1088–1090 assessing impairment, 1090 developing useful sensation-testing approaches, 1089 distribution of sensation abnormality, 1089, 1089f–1090f hyposensitivity and hypersensitivity, 1066–1067, 1089 inferring anatomic-pathologic abnormalities of fibers, 1089–1090 inferring improvement or worsening, 1090 in research, 1090–1091, 1091f–1092f in suramin neuropathy, 2567 in taxane neuropathy, 2569 in uremic neuropathy, 2024 in vitamin B12 deficiency polyneuropathy, 2055 natural stimulation for, 1065 reasons for testing different modalities of skin sensation, 1066 sensitivity, reproducibility, and concordance with neuropathic impairment, 1088, 1091, 1092f

lxxiv Quantitative sensation testing (QST) (Continued) specifying modality being tested, 1065 subjective differences among anatomic sites for, 1067 thermal sensation, 1084–1088 cooling detection threshold using CASE IV, 1085–1086, 531f–532f estimating percentile abnormality, 1087–1088, 1088t warm detection threshold, 1086 warmth or heat-pain sensation, 1086–1087, 1087f–1088f threshold variation with anatomic site, age, sex, and anthropomorphic characteristics, 1067 touch-pressure sensation, 1077–1079, 1078f two-point discrimination, 1079 types of, 1067–1068 laboratory QST (L-QST), 1068, 1069f–1070f quantitative QST, 1068 semiquantitative QST, 1067–1068 uses of, 1063–1064, 1066, 1088 vibration sensation, 1079–1083 CASE IV, 1079–1080 quality assurance of vibration detection threshold, 1082, 1083f use of vibration detection threshold, 1082–1083 vibration detection threshold as measure of mechanoreceptor function, 1080 vibration detection threshold in health, 1082, 1084f vibration detection threshold using CASE IV, 1080–1082, 1081f–1082f vs. clinical assessment, 1065–1066 Quantitative sudomotor axon reflex test (QSART), 1106–1109 commercial equipment for, 1108 in Composite Autonomic Severity Score, 1113t, 1114, 1114t–1115t in distal small fiber neuropathy, 1104 in preganglionic vs. postganglionic lesions, 1108, 1108t in small fiber neuropathy, 2324 interpretation of, 1108–1109 persistent sweat activity on, 1108 physiologic basis for, 1106–1108, 1107f recording sites for, 1108 reproducibility of, 1108 resting sweat activity on, 1108 sweat volumes recorded from same and distant compartments in, 1106, 1107t

R73 monoclonal antibody, for experimental autoimmune neuritis, 623 RA. See Rheumatoid arthritis (RA). RAB7 gene in Charcot-Marie-Tooth disease type 2B, 1629, 1730–1731 in hereditary sensory and autonomic neuropathy type I, 1810, 1811t, 1822–1823, 1824f Rabies, vs. Guillain-Barré syndrome, 2202 Rac in actin dynamics, 450, 451f, 452 in netrin signaling, 460–461 Radial nerve, 1340f distribution of, 15f, 21–22, 22f, 1470f, 1474 in leprosy, 2089f

Index Radial nerve (Continued) involvement in vasculitic neuropathy, 2365t muscles innervated by, 956t, 1474 nerve conduction studies of, 911–914, 915f “Saturday night” paralysis of, 1141, 1142 superficial, 1474 Radial neuropathy, 1474–1478 clinical manifestations of, 1476 diagnosis of, 1476–1478 electromyography in, 953f etiologies of, 1474–1476 iatrogenic, 1475 in children, 2724f, 2726–2727, 2727f in renal failure, 2026 traumatic, 1475 treatment of, 1478 Radial tunnel syndrome, 1476 Radiation exposure chromatolysis due to, 686 intracytoplasmic vacuolation induced by, 688 Radiation therapy brachial plexopathy induced by, 1367–1368, 2490 for leptomeningeal metastasis, 1330 for metastatic radiculopathy, 1331 for POEMS syndrome, 2453, 2463 lumbosacral plexopathy induced by, 1378, 1382–1384, 1384t, 2490 postirradiation lower motor neuron syndrome after, 1303 radiculopathy after, 1332 Radiculitis, in Lyme disease, 2111 Radiculomyeloencephalitis, eosinophilic, 2164 Radiculopathy(ies), 1324–1334 affecting multiple spinal nerve roots, 1327–1334. See also Polyradiculopathy. affecting single spinal nerve roots, 1325–1327 cerebrospinal fluid analysis in, 1325 compressive causes of, 1334 due to intervertebral disc herniation, 1325–1326 diabetic, 1332 diagnosis of, 1324–1325 due to spinal root trauma, 1326–1327 electrophysiologic studies in, 905, 1324–1325 herpes simplex virus infection and, 2118 in children, 2736–2738 cervical, 2737 lumbosacral, 2737–2738 infectious causes of, 1331–1332 intraspinal medications for, 1326 L5, 1499t magnetic resonance imaging in, 1325 management of, 2634 metastatic, 1329–1331 myokymia in, 1143 radiation-induced, 1332 spondylotic, 1311, 1328–1329, 1329f toxic, 1333 vs. mononeuropathy, 1324 Radiculoplexus neuropathies, 1993–2013 clinical features of, 1995–2000, 1997t–1998t clinical testing in, 1997t, 2001–2002, 2002t–2003t diabetic, 1993–2013 cervical. See Diabetic cervical radiculoplexus neuropathy (DCRPN). Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Radiculoplexus neuropathies (Continued) lumbosacral. See Diabetic lumbosacral radiculoplexus neuropathy (DLRPN). thoracic. See Diabetic thoracic radiculoneuropathy (DTRN). historical overview of, 1994–1995 immunotherapy for, 2011–2012, 2013f microvasculitis in, 2405 nondiabetic, 1993–1994 cervical, 1994, 2000. See also Brachial plexus neuropathy, immune (IBPN). lumbosacral, 1994, 2000 thoracic, 1994, 2000 pathology of, 2004–2011 symptomatic treatment of, 2012–2013 Radiography chest, 1174 of brachial plexus lesions, 1350 Ragged red fibers (RRFs), 1940, 1941f RAIR (rectoanal inhibitory reflex), 283–284 Rami communicantes, 15–16, 17f of head, neck, and upper limb sympathetic ganglia, 23 of lumbar sympathetic ganglia, 30 of sacral sympathetic ganglia, 30 of thoracic sympathetic ganglia, 25 Ramsay Hunt syndrome, 1231–1232, 2120 clinical features of, 1231 definition of, 1231 epidemiology of, 1231 herpes simplex virus infection and, 1229 medical therapy for, 1232 neuroimaging in, 1231 neuropathology of, 1232 pathophysiology of, 1231–1232 surgical decompression for, 1232 Ramus(i) dorsal, 11, 12, 16, 16f cervical, 31–32, 1324f coccygeal, 32 lumbar, 25, 32 sacral, 28, 32 thoracic, 23, 32 ventral, 11, 12, 15–16, 16f, 18, 19, 23, 1339 cervical, 1324f lumbar, 25, 1375 sacral, 28, 1376 thoracic, 23, 24 Range of motion (ROM), 2622 effects of immobility on, 2628–2629 in motor neuron disease, 2633 Ranitidine, 2610 Rankin scale, modified, 1042, 1046t Rapamycin, 638–639 adverse effects of, 639 mechanism of action of, 639 use in neuropathy, 639 Rapsyn deficiency, congenital myasthenic syndrome caused by, 855–857 clinical features of, 856 end-plate studies in, 857 molecular pathogenesis of, 856f, 857 therapy for, 860 RASQoL (Rheumatoid Arthritis-Specific Quality of Life scale), 1059 Raynaud’s phenomenon, 1168 carpal tunnel syndrome and, 1440 cryoglobulinemia and, 1168, 2019 Sjögren’s syndrome and, 1168, 1308 skin biopsy in, 888 RDNS (Rochester Diabetic Neuropathy Study), 1996, 1997t, 2001, 2005, 1952, 1956

Index Reactive oxygen species (ROS), 509–511. See also Oxidative stress. age-related increase in, 500, 501 axon dystrophy and, 497 factors that increase activity of, 511 generation of, 509–510, 510f biologic generators, 510 in experimental diabetic neuropathy, 515, 677 in ischemia-reperfusion, 520, 521, 677 polyol pathway overactivity, 514 half-lives of, 510 protective mechanisms against, 510–511, 511t scavengers of, 511 Recombinant DNA technology, 1553–1554 Recovery cycle, 117f, 117–118, 120f in multifocal motor neuropathy, 123, 123f Rectal balloon extension test, 286–287, 287f Rectal nerve, inferior, 281, 281f, 282, 28 Rectal plexus, 31, 281, 281f Rectoanal inhibitory reflex (RAIR), 283–284 Rectorectal reflex, 2627 Rectosigmoid hyposensory perception, 255 Rectum, 280f, 2627. See also Anorectum. anatomy of, 280–281, 281f assessment of rectoanal sensation, 290t, 291, 292f compliance of, 291–292 digital examination of, 2691 distention by stool, 282, 285 embryology of, 279–280 Reference values for nerve tests, 972 selection of reference cohort, 973 statistical methods for computation of, 973–976, 974t Reflex sympathetic dystrophy, carpal tunnel syndrome and, 1441 Refractoriness, 117 Refsum’s disease, 1625t, 1626, 1738, 1866, 1869–1871 clinical features of, 1869 course of, 1141 demyelination in, 1167 genetics of, 1627t, 1629–1630, 1869 hearing loss in, 1168 histologic changes in, 1870–1871, 1871f in children, 2722 infantile, 1866, 1869, 1872 nerve thickening in, 1152 pathology of, 710, 1869–1870 phytanic acid elevation in, 1866, 1869 pigmentary retinopathy in, 1168, 1869 treatment of, 642 Rehabilitation, 2621–2634 for acute and subacute polyradiculoneuropathy, 2630 for brachial plexus injuries and disease, 2630–2632 obstetric brachial plexopathy, 2632, 2733 for generalized sensorimotor peripheral neuropathy, 2629–2630 for motor neuron disease, 2633–2634 for postpolio syndrome, 2632–2633 for radiculopathy, 2634 for specific complications, 2625–2629 autonomic dysfunction and pain, 2626 bladder incontinence, 2626–2627 bowel incontinence, 2627–2628 related to immobility and deconditioning, 2628–2629 sensory loss, 2625 sexual dysfunction, 2628 weakness, 2625

Rehabilitation (Continued) goal of, 2621 interventions and techniques of, 2622–2625 adaptive equipment and environmental modifications, 2625 ambulation aids, 2624 electrical stimulation, 2623–2624 orthoses, 2623 therapeutic exercise, 2622–2623 wheelchairs, 2624–2625 overview and principles of, 2621–2622 physician’s role in, 2621–2622 restorative approaches to, 2621 Reich granules, 53, 62, 62f, 63f Reinnervation axonal regrowth, 1421f, 1421–1422 precision of, 1422–1423, 1423f preferential motor, 1422 scanning electromyography in, 991 single-fiber electromyography in, 992–994, 993f bimodal jitter, 994 concomitant blocking, 994 Relapsing inflammatory polyradiculoneuropathy, compound action potential of sural nerve in, 1024–1025, 1026f–1027f Relapsing polychondritis, vasculitic neuropathy and, 2340t Reliability, of health outcome measures, 1056–1057 Remak fibers, 36, 62, 64, 65. See also Unmyelinated fibers. aging and, 491–492 arrangement of axons in, 65 number of axons in, 65, 65f Remyelination, 3–4, 489, 790 conduction block and, 2281 in diphtheria, 2149 teased fiber analysis of, 748, 749f–750f Renal disease, 1168 in Fabry’s disease, 1893–1894, 1900 in POEMS syndrome, 2460–2461 neuropathies associated with, 2022–2026. See also Uremic neuropathy. Renal ganglion, 31 Renal transplantation, effect on uremic neuropathy, 2024–2025 Renaut bodies, 770, 773f, 1405 endoneurial, 38–39, 39f in hypothyroid polyneuropathy, 2041 Reperfusion injury, 521. See also Ischemia. cytokines in, 522–523, 525f, 675 in experimental diabetes, 673f, 673–674 neuroprotection against, 524, 526f, 676 pathogenesis of, 676–677 pathology of, 675–676 Resiniferatoxin, 699 Respiratory control in diabetic patients, 1959–1960 Respiratory disease and neuropathy, 2021–2022 chronic obstructive pulmonary disease, 2021–2022 mechanisms of hypoxic injury, 2022 Respiratory failure, 2611–2615 effects of acute neuromuscular disease on respiration, 2611–2612, 2612f in amyotrophic lateral sclerosis, 1536, 2633 management of, 1540 in Guillain-Barré syndrome, 2200 acute inflammatory demyelinating polyneuropathy, 2200 acute motor axonal neuropathy, 2201 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxv Respiratory failure (Continued) in children, 2716 malaria-associated, 2172 in tick paralysis, 2174 intensive care unit management of, 2607–2608, 2611–2615 intubation and mechanical ventilation, 2613–2614, 2613t–2614t monitoring pulmonary function, 2612f, 2612–2613 principles of respiratory physiology, 2611 tracheostomy, 2614–2615 weaning from ventilation, 2614–2615 Resting membrane potential, of neurons of enteric nervous system, 262 Restless legs syndrome (RLS), 1145 age at onset of, 886 crurum dolorosum, 886 crurum paresthetica, 886 in uremic neuropathy, 1145, 2023 nerve biopsy in, 886 skin biopsy in, 886 subclinical hereditary sensory and autonomic neuropathy type I with lancinating pain and, 1826 treatment of, 1145 Restriction fragment length polymorphisms (RFLPs), 1551–1552 Reticuloendothelial system, in Fabry’s disease, 1896 Retinitis pigmentosa, 1625t, 1626 hereditary motor and sensory neuropathy with, 1625t, 1626, 1627t, 1650, 1744, 1771 Retinopathy, 1168 diabetic, 1168 damage to ciliary nerves during photocoagulation treatment of, 204 pigmentary, in Refsum’s disease, 1168, 1869 Retrograde ejaculation, 1151 in diabetic patients, 1959 Retroviruses, amyotrophic lateral sclerosis and, 1535 Reverse cholesterol transport pathway, 1906 in Tangier disease, 1906–1908 RFLPs (restriction fragment length polymorphisms), 1551–1552 Rheumatoid arthritis (RA), 2357 carpal tunnel syndrome and, 1445 clinical features of, 2357 eighth nerve damage in, 1268 juvenile, 2719 pathogenesis of, 2357 peripheral neuropathy and, 2357 anti–endothelial cell antibodies in, 576 quality-of-life measures in, 1058–1059 prevalence of peripheral neuropathy and, 1172 vasculitis and, 2337, 2340t, 2353t, 2357–2358 incidence of, 2340t pathology of, 803, 804f relative prevalence of, 2341t Rheumatoid Arthritis-Specific Quality of Life scale (RASQoL), 1059 Rheumatoid factor, 1172 Rho guanosine triphosphatases (Rho GTPases) in netrin signaling, 460–461 in semaphorin signaling, 464 protein regulators of, 450 regulation by Eph-associated ephexin, 454 role in actin dynamics, 450f–451f, 450–452

lxxvi Rho kinase (ROCK), in actin dynamics, 451, 451f, 452 RhoA, in actin dynamics, 450, 451f Ribavirin, for hepatitis C–associated cryoglobulinemic vasculitis, 2336 Ribonucleic acid (RNA), 1545, 1550 Riche-Cannieu anastomosis, 1437 Rifampin, for leprosy, 2103 Riley-Day syndrome, 2722 autonomic dysfunction in, 1141, 1831 biochemistry of, 1833 clinical features of, 886, 1831–1832 genetics of, 886, 1627t, 1832 inheritance of, 1832–1833 nerve conduction studies in, 1832 pathology of, 1832 skin biopsy in, 886 sweating in, 1150 Riluzole, for amyotrophic lateral sclerosis, 1538–1539 Rituximab for anti-MAG neuropathy, 2179t for CANOMAD syndrome, 2180t for distal lower motor neuron syndrome, 2180t for monoclonal gammopathy and peripheral neuropathy, 2269, 2269t for multifocal motor neuropathy, 2179t, 2182–2184 River blindness. See Onchocerciasis. RLS. See Restless legs syndrome (RLS). RNA (ribonucleic acid), 1545, 1550 Robo. See Slit/Robo. Rochester Diabetic Neuropathy Study (RDNS), 1996, 1997t, 2001, 2005, 1952, 1956 ROCK (Rho kinase), in actin dynamics, 451, 451f, 452 ROM (range of motion), 2622 effects of immobility on, 2628–2629 in motor neuron disease, 2633 Romaña’s sign, in American trypanosomiasis, 2156, 2158f ROS. See Reactive oxygen species (ROS). Rostral spinocerebellar tract, 158 Rostral ventrolateral medulla (RVLM), 217–218 Roussy-Lévy syndrome, 1145, 1170, 1624, 1662, 1681 RP-67580, for hyperalgesia, 527 RRFs (ragged red fibers), 1940, 1941f Rucksack paralysis, 1357–1358 Ruffini corpuscles, 484, 1005 RVLM (rostral ventrolateral medulla), 217–218 Ryukyuan hereditary motor neuronopathy, 1617–1618

S100, 61, 343 Sabin polio vaccine, 2713 Saccule, 1253 Sacral bone tumors, 2738 Sacral ganglion, 30 Sacral nerve stimulation for fecal incontinence, 293, 293t for urge urinary incontinence, 293 Sacral plexus, 18, 28–30, 32, 28–30, 32, 18, 28–30, 32 anatomy of, 1376 branches to buttock and lower limb, 28–30 branches to pelvic structures, 28 location of, 28

Index Sacral reflexes, 284 Sacrococcygeal plexus, 30 Salivary glands, in Chagas’ disease, 2159 Salivation test, 1116 Salk polio vaccine, 2713 Salla disease, 1865f Salt supplementation, for orthostatic hypotension, 2658, 2659, 2662 SAN (sinoatrial node), 217 Sandbag palsy, 1359 Sandhoff’s disease, 1860 SANDO (sensory ataxic neuropathy, dysarthria, and ophthalmoplegia), 1942, 1946 Sanfilippo’s syndrome, 1864 Saphenous nerve, 16f, 1376 distribution of, 1489f, 1491 lesions of, 1491–1492 nerve conduction studies of, 917–918 Sarcoidosis, 2415–2423 age distribution of, 2416 angiitis and, 2417 as multisystem disease, 2415, 2419–2420 central nervous system involvement in, 2419 cerebrospinal fluid analysis in, 2419 clinical features of, 2419 course of, 2420 cranial nerve involvement in, 1140, 1203, 1268, 2415, 2417–2418 facial nerve palsy, 1232 diagnosis of, 2415, 2421–2422 electrophysiologic findings in, 2420 epidemiology of, 2415–2416 exocrine gland involvement in, 2420 eye involvement in, 2418 genetics of, 2416 granuloma formation in, 2416–2417 hearing loss in, 2418 historical overview of, 2415 liver in, 2420 lung disease in, 2420 lymphadenopathy in, 2420 myopathy in, 2419 neuroimaging in, 2419 neurosarcoidosis, 2416 osteoarticular lesions in, 2420 pathology of, 734f–735f, 808–809, 2421 muscle biopsy, 2421, 2422f–2423f nerve biopsy, 735f, 2421, 2421f–2422f pediatric polyneuropathy and, 2719 peripheral neuropathy in, 2418–2419 Guillain-Barré–like syndrome, 2419 multifocal neuropathy with conduction block, 2418 multiple mononeuropathy, 2418 prognosis for, 2423 purely motor neuropathy, 2418 polyradiculopathy in, 1334 skin lesions in, 2420 spleen in, 2420 treatment of, 2423 Sarcoma(s), 2600 neurogenic. See Malignant peripheral nerve sheath tumor (MPNST). of brachial plexus, 1360 Satellite cells, 235, 341, 1295 Satellite ganglia, 234 “Saturday night” paralysis, 1141, 1142 SCA. See Spinocerebellar ataxia (SCA). SCa (slow component a), in axonal transport, 387, 390 Scanning electromyography, 990–991 applications of, 985, 991 concentric needle electrode for, 986f, 990 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Scanning electromyography (Continued) in amyotrophic lateral sclerosis, 996f, 996–997 in primary myopathies, 990, 999 in reinnervation, 991 recording principle for, 990, 990f Scapula fracture of, 1367 long thoracic neuropathy and winging of, 1465 muscle stabilizers of, 1464 operative stabilization of, 1465 Scapuloperoneal hereditary motor neuronopathy, 1604t, 1613t, 1615f, 1615–1616 Scarpa’s ganglion, 1253 SCb (slow component b), in axonal transport, 387, 390 SCBM (Schwann cell basement membrane), in nerve regeneration, 793, 794f–796f Schindler disease, 2722 Schistosomiasis, 2154t–2155t, 2160–2161 cerebral, 2160 chronic, 2160 clinical stages of, 2160 epidemiology of, 2160 etiologic agent in, 2160 life cycle of, 2160 gastrointestinal, 2160 pathogenesis of, 2160 pathology of, 2158f peripheral neuropathy in, 2160 spinal, 2160 transmission of, 2160 urinary, 2160–2161 Schizophrenia axon guidance in, 466–467 DISC-1/NUDEL/LIS-1, 467 neural cell adhesion molecule and L1, 467 semaphorin signaling, 467 Wnt signaling, 467 pain insensitivity in, 1813 Schmidt-Lanterman incisures, 3, 5, 6, 42, 43, 61f, 413 function of, 417 identification of, 413–414 protein localization in, 418f, 420, 580 Schmidt’s syndrome, 1285 Schwann cell basement membrane (SCBM), in nerve regeneration, 793, 794f–796f Schwann cell families, 5 Schwann cell precursors, 411, 412f Schwann cells, 3, 35, 42 acutely denervated, 1408–1410 phenotype of, 1409–1410 as antigen-presenting cells, 610 axon-independent, 1413, 1414f–1415f basal lamina of, vs. collagen sheath, 37 cytoskeletal reorganization during myelination, 424 death of, 1413 development of, 60–61, 64, 341–365 developmental potential of early glia, 349–351 dystroglycan signaling in, 364–365 endothelin in, 60, 342, 346, 353–354 integrin signaling and interactions with actin cytoskeleton in, 363–364 laminin in, 362–363 neural activity and, 362 neural crest origins for, 341, 346–348 neuregulin-1 in, 60, 342, 346–347, 351–356

Index Schwann cells (Continued) outline of, 341–343, 342f regulation of transition from precursors, 353–354 transcription factors in, 342, 343, 1708 differentiation of, 341–343 markers for, 343–346, 344f reversal of, 343 growth-arrested, 362 immature, 411, 412f in adrenoleukodystrophy, 1867, 1868f in adrenomyeloneuropathy, 1867 in diabetic neuropathy, 1965–1966, 1967f in diphtheria, 2131, 2132 in Farber’s disease, 1859 in Krabbe’s disease, 1857, 1857f in leprosy, 424, 574, 2099 in lysosomal disorders, 1846 in metachromatic leukodystrophy, 1849, 1850, 1850f in mucolipidosis IV, 1865, 1866f in multiple sulfatase deficiency, 1855f in neuronal ceroid-lipofuscinoses, 1862, 1863f in Niemann-Pick disease, 1861f, 1861–1862 in Refsum’s disease, 1870–1871, 1871f–1873f in sialic acid storage diseases, 1865, 1865f in Tangier disease, 1909, 1911f–1912f in wallerian degeneration, 4 inclusions in, 771, 791–792 lipocyte, 771, 776f locations of, 771 ␮ granule of Reich, 771, 775f, 791 ␲ granules, 771, 774f–775f, 791 interactions with axons. See Axon–Schwann cell interactions. intraperiod line of, 413, 577 major histocompatibility complex molecules expressed by, 62, 565, 574 mesaxon of, 412 migration of, 60, 364 mitogens for, 344f, 356–357 myelinating, 60–62, 61f–62f differentiation of, 341–343, 359f proteins expressed by, 61 signaling pathways and growth factors regulating myelination, 361–362 transcription factors involved in myelination, 357–361, 358t, 359f nonmyelinating differentiation of, 341–343, 359f proteins expressed by, 61 of internodes, 42–43, 788 of node of Ranvier, 56–57 origin of, 60 P0 expression by, 6, 419, 1682 pathogenesis of injury in diabetes, 519 pathologic alterations of, 819–823 developmental abnormality, 819 immune demyelination, 820 in congenital tomaculous neuropathy, 823 in Déjérine-Sottas disease, 819 in inherited tendency to pressure palsy, 823, 824f in lead intoxication, 820–823, 821f–822f in lysosomal storage diseases, 819–820, 820f toxin-induced, 820 potassium channels on, 56, 98f, 104–108, 105f precursors of, 341–343, 344f, 348, 349f regulation of transition from, 353–354

Schwann cells (Continued) proliferation of, 61, 342, 361, 1409 cell culture studies on, 356–357 PMP22 in, 579 relationship of unmyelinated axons with, 66 role in immune response, 565–567 signals that control survival of, 354f, 354–356 sodium channels on, 98f, 98–99 synthesis of nerve growth factor by, 379 transitional, 69 Schwannoma(s), 2586t, 2586–2591 cellular, 2588, 2590f granular cell, 2595 intraosseous, 2587 intraspinal, 2586–2587 Luse bodies in, 38 macroscopic features of, 2587, 2587f–2588f malignant. See Malignant peripheral nerve sheath tumor (MPNST). malignant peripheral nerve sheath tumor arising in transformation from, 2599 melanotic, 2588–2591, 2590f merlin mutations and, 363–364, 2586 microscopic features of, 2587–2588, 2589f of brachial plexus, 1360–1361 of facial nerve, 1221, 1221f on magnetic resonance imaging, 1221, 2587 plexiform, 2588 sites of, 2586–2587 trigeminal, 1210 vestibular, 1264–1268 Schwartz-Jampel syndrome, 1144 Sciatic nerve anatomy of, 1493–1494 blood supply to, 800–801 distribution of, 27f, 28–29, 1494, 1494f, 1500f growth-associated protein-43 expression in dorsal root ganglion cells after crush of, 1297 injection of foreign substance into, 805 ischemia of, 1494 lymphomatous infiltration of, 1383f morphometry of, 767f pediatric tumors of, 2729 structures innervated by, 957t, 1376, 1494 Sciatic neuropathy, 1493–1496 clinical manifestations of, 1494–1495 diagnosis of, 1495 differential diagnosis of, 1499t etiology of, 1493 total hip replacement arthroplasty, 1494–1496, 1519–1520, 1520f in children, 2724f, 2727–2729, 2728f piriformis syndrome, 1496 prognosis for, 1495–1496 treatment of, 1495 Sciatica catamenial, 1381 vs. lumbosacral plexus lesion, 1377 SCIP (suppressed cAMP-inducible POU), 1708 SCLC. See Small-cell lung carcinoma (SCLC). Scleroderma, 2360–2361 sensory polyneuropathies and, 2361 vasculitic neuropathy and, 2361 Scoliosis surgery, radiculopathy after, 2738 Scoring neurologic signs. See Quantitating neuropathic symptoms, impairments, and outcomes. Screening laboratory tests, 1170–1173 Scrotal nerve, posterior, 282 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxvii Scrotal nerves, 28 Sebaceous glands, 870 Secretomotor neurons, in enteric nervous system, 257f, 257–258 Seizures, 465–466 focal vs. generalized, 466 sodium channel gene mutations and, 97 Selectins, 616, 2341–2342 Selective serotonin reuptake inhibitors (SSRIs), for neuropathic pain, 2641–2642 Self-tolerance, 574, 575, 610 Semaphorin/plexin, 462–468 modulation of signaling by, 465 intracellular cGMP levels, 465 local protein synthesis for Sema3A signaling, 465 receptor complexes of, 462–464 signaling by, 464–466 downstream kinases, 464–465 in CRASH syndrome, 468 in epilepsy, 466 in schizophrenia, 467 Rho GTPases, 464 role of redox signaling, 465 Sema7A promotion of axon outgrowth, 465 Semmes-Weinstein monofilaments, 1038 Semon’s law, 1279 Sensation fiber groups and, 1016–1017 quantitative testing of, 1037–1038, 1063–1091. See also Quantitative sensation testing (QST). Sensorimotor polyneuropathy, rehabilitation for, 2629–2630 Sensory ataxic neuropathy, dysarthria, and ophthalmoplegia (SANDO), 1942, 1946 Sensory ganglionopathy, 1148, 2322 nonmalignant inflammatory, 2309–2317 clinical patterns of, 2311 course of, 2311–2312 differential diagnosis of, 2314–2317, 2315t electrophysiologic findings in, 2312–2313 historical review of, 2310–2311 laboratory findings in, 2312 pathology of, 2313–2314, 2314f radiologic features of, 2312 symptoms and signs of, 2311 treatment of, 2317 pattern of involvement in, 1139 skin biopsy in, 887 Sensory nerve action potentials (SNAPs) effect of dorsal root ganglion pathology on, 1323, 1324, 1327 factors affecting amplitude of, 904 gender differences in, 904 in acrylamide neuropathy, 2506 in brachial plexopathy, 1351 immune, 2302 perinatal, 2731 in chronic obstructive pulmonary disease–associated neuropathy, 2022 in common peroneal neuropathy, 1498 in critical illness myopathy, 2030 in critical illness polyneuropathy, 2027 in cryptogenic sensory polyneuropathy, 2323 in femoral neuropathy, 1490 in giant axonal neuropathy, 1651 in Guillain-Barré syndrome acute inflammatory demyelinating polyneuropathy, 2200 axonal forms, 2201

lxxviii Sensory nerve action potentials (SNAPs) (Continued) in hepatitis C–associated neuropathy, 2019 in hereditary motor and sensory neuropathy, 1637–1638 in lumbosacral plexus lesions, 1378 in neuropathy associated with chronic hepatic disease, 2018 in nonmalignant inflammatory sensory polyganglionopathy, 2312–2313 in pediatric polyneuropathies, 2712 in primary biliary cirrhosis–associated sensory polyneuropathy, 2020 in radiculopathy, 1324 in sciatic neuropathy, 1495 in sensory variant of chronic inflammatory demyelinating polyradiculoneuropathy, 2229 in tick paralysis, 2717 in true neurologic thoracic outlet syndrome, 1360 in uremic neuropathy, 2023 after renal transplantation, 2025 Sensory nerve conduction studies, 901f, 903–905. See also Nerve conduction studies (NCSs). in hereditary motor neuronopathy, 1605 in radiculopathy, 1324 polyradiculopathy, 1327 latency and conduction velocity in, 904 types of abnormalities on, 904–905 waveform, amplitude, and duration on, 903–904 Sensory neurons age-related changes in, 483–486, 702 functional impairments, 483 mechanisms of, 495–496 signaling molecules, 484–486 structural impairments, 483–484, 485f–486f chromatolysis of, 686 in mice heterozygously deficient for P0, 538 of autonomic ganglia, 236 of enteric nervous system, 251, 251f, 253, 254–256 response to injury, 1463–1464 Sensory neuropathy, 1137, 1141 acute, autoantibodies in, 2178t clinical course of, 1141 congenital, 1838 recessive form of, 1838 congenital nonprogressive, with tonic pupils, 1837 cryptogenic, 2321–2331 diabetic oxidative stress and, 518–519, 519f–520f symmetrical distal polyneuropathy, 1953–1954 disorders associated with, 2309–2310 disorders mimicking, 1153 genitofemoral, 1488 glove-and-stocking distribution of, 1138, 1139f, 1140 hereditary, with spastic paraplegia, 1837 hereditary sensory and autonomic neuropathies, 1809–1839 iliohypogastric, 1488 ilioinguinal, 1488 in amyloidosis, 1147 in carpal tunnel syndrome, 1438t, 1438–1439, 1439f in chronic inflammatory demyelinating polyradiculoneuropathy, 2229, 2231–2232, 2315t

Index Sensory neuropathy (Continued) in distal polyneuropathy, 2096–2097 in Fabry’s disease, 1147, 1896, 1898f in Friedreich’s ataxia, 1147 in hereditary brachial plexus neuropathy, 1755 in hereditary motor and sensory neuropathies, 1635 type II, 1722–1723 in HIV infection, 2131–2136 in leprosy, 1141, 1147, 2087–2095, 2088f–2096f in lumbosacral radiculoplexus neuropathy, 1999 diabetic, 1996–1999 in POEMS syndrome, 2457 in primary biliary cirrhosis, 2020–2021 in Tangier disease, 1147 in tropical ataxic neuropathy, 2068–2071 large fiber sensory loss, 1147f, 1147–1148, 1165–1166 painful, skin biopsy in, 883–884, 884f pansensory familial neuropathy, 1838 paraneoplastic, 1141, 2316, 1141, 2178t, 2316 small fiber, 2479 subacute, 2473–2476 pattern of involvement in, 1138, 1139f–1140f quantitation of sensory loss, 1037–1038 Inflammatory Neuropathy Sensory Sum Score, 1037, 1037t touch-pressure sensation, 1038 vibration sensation, 1038 rehabilitation for, 2625 small fiber sensory loss, 1146–1147, 1165–1166 Sensory polyganglionopathy malignant inflammatory, 2315t nonmalignant inflammatory, 2309–2317 clinical patterns of, 2311 course of, 2311–2312 differential diagnosis of, 2314–2317, 2315t electrophysiologic findings in, 2312–2313 historical review of, 2310–2311 laboratory findings in, 2312 pathology of, 2313–2314, 2314f radiologic features of, 2312 Sjögren syndrome and, 2311, 2312, 2313, 2314f, 2315t symptoms and signs of, 2311 treatment of, 2317 Sensory polyradiculopathy, disorders associated with, 2309–2310 Sensory receptors of skin, 1004–1007, 1064, 1064f low-threshold C fibers, 1007 low-threshold mechanoreceptors, 1005 nociceptors, 1005–1006 thermoreceptors, 1005 Sensory symptoms, 1146–1150, 1165–1166 negative, 1146–1148 positive, 1148–1150 Sensory syndromes, 1307–1310 cobalamin deficiency, 1307–1308 Sjögren’s syndrome, 1308–1310 Sepsis, 2026–2030. See also Critical illness polyneuropathy. vs. childhood Guillain-Barré syndrome, 2717 Septic shock, 2026 Serine palmitoyltransferase, 424 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Serotonin (5-HT) effect on nerve blood flow, 670 in cardiac neural control, 221 in enteric nervous system, 255–256, 266, 267, 271 in micturition reflex, 312 in pain response, 183 receptors for, 183 Serratus anterior muscle, 1464 electromyography of, 1465 lesions of nerve to, 1464–1466. See also Long thoracic neuropathy. Sertraline, for pain, 2642 Serum amyloid P component, 2428 SETX gene mutations, in amyotrophic lateral sclerosis type 4, 1599 Sexual dysfunction, 1151, 2628, 2664 in diabetic patients, 1957, 1959, 2628 sildenafil for, 1959, 2664 Sexual response cycle, 314–318 central reflex pathways in, 317–318 emission-ejaculation phase of, 316–317 erection, 315–316 central mechanisms of, 316 peripheral mechanisms of, 315–316 glandular secretion during, 316 male vs. female, 314 neural control of, 314t, 314–315 parasympathetic pathways, 314–315 somatic pathways, 315 sympathetic pathways, 315 SF-36 (Short Form-36), 1048, 1049, 1059–1060 SFEMG. See Single-fiber electromyography (SFEMG). SFN. See Small fiber sensory neuropathy (SFN). SGLPG (sulfated glucuronyllactosaminyl paragloboside), 583t, 584 SGPG (sulfated glucuronyl paragloboside), 584, 597 IgM binding to, 2188 structure of, 583t Short Form-36 (SF-36), 1048, 1049, 1059–1060 Short tandem repeats (STRs), 1552 Short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing (SUNCT syndrome), 1214 Shoulder dislocation axillary neuropathy due to, 1469–1470 brachial plexus lesion due to, 1365–1366 from orthopedic procedures, 1369–1370 suprascapular neuropathy due to, 1467 Shoulder girdle brachial plexus injury from orthopedic procedures on, 1369–1370 innervation of muscles of, 956t neuritis of. See Brachial plexus neuropathy, immune (IBPN). Shoulder pain, in hereditary brachial plexus neuropathy, 1754 Shy-Drager syndrome, 282 airway dysfunction in, 1284–1285 cardiac vagal activity in, 226, 1280 Sialic acids, 2285 Sialidosis, 1864–1865, 1865f, 1865 type II, in children, 2722 Sialoadhesin, 575 Sialosyl paragloboside, 584 Sialosyllactosaminyl paragloboside (Hex-LM1), 584 Sialosynneolactotetraosylceramide (LM1), 583t, 584 Sicca syndrome, 1150, 2329

Index SIDP (subacute idiopathic demyelinating polyradiculoneuropathy), 2222–2223 Siglecs, 575 myelin-associated glycoprotein, 420, 547 Signal transduction in enteric nervous system, 266–267 in Schwann cell development, 341–365 neurotrophins and, 378f, 381–383, 382f pathways and growth factors regulating myelination, 361–362 signaling molecules in aging sensory neurons, 484–486 Sildenafil, for erectile dysfunction, 2664 in diabetic patients, 1959 Single-fiber discharge and denervation, 943, 944f Single-fiber electromyography (SFEMG), 985–997 applications of, 985 during axonal stimulation in myasthenia gravis and Lambert-Eaton myasthenic syndrome, 995–996 during intramuscular electrical stimulation, 988f, 988–989 axonal stimulation, 988–989 direct stimulation, 988 during voluntary activation, 985–987 fiber density, 987, 987f jitter, 986–987 propagation velocity, 986 electrode for, 985, 986f in amyotrophic lateral sclerosis, 993f, 996f, 996–997 in Bell’s palsy, 1226 in denervation and reinnervation, 991–994, 993f in myasthenia gravis, 987, 993f, 994–995, 995f in syringomyelia, 997 multielectrode studies, 989–990 signs of abnormal nerve excitability on, 994 A waves, 994 ephaptic transmission, 994 extra discharges, 994 fasciculations, 994 signs of abnormal transmission in nerve branches on, 993f, 994 studies of spinal events with, 989 cortical stimulation, 989 F response, 989 firing pattern, 989 H reflex, 989 motor neuron excitability, 989 other reflexes, 989 Single-gene inheritance. See Mendelian inheritance. Single-nucleotide polymorphisms (SNPs), 1553 Single-photon emission computed tomography (SPECT), imaging of sympathetic innervation to organs, 1119–1121, 1120f Single-strand conformation polymorphism (SSCP), 1557 Sinoatrial node (SAN), 217 Sinuvertebral nerve, 31 Sirolimus, 638–639 adverse effects of, 639 mechanism of action of, 639 use in neuropathy, 639 SIRS (systemic inflammatory response syndrome), 2017, 2026–2030. See also Critical illness polyneuropathy. Sit-to-stand test, 1096

Sixth nerve palsy, 1197 congenital, 1197 diabetic, 1201 due to infections and toxins, 1202 due to neoplasms, 1199–1200 traumatic, 1198 Sjögren’s syndrome, 1308–1310, 2328–2329, 2358–2360 autoantibodies in, 1172, 2359 autoimmune hepatitis and, 2021 autonomic dysfunction in, 1309 clinical features of, 1308, 2328, 2359 cryptogenic sensory polyneuropathy and, 2329, 2329f diagnostic criteria for, 2328, 2328t, 2360t dorsal root ganglion damage in, 1309–1310, 1309f–1310f facial nerve paralysis in, 1234 laboratory investigation for, 1172 microvasculitis in, 2405 nonmalignant inflammatory sensory polyganglionopathy and, 2311, 2312, 2313, 2314f, 2315t prevalence of neuropathy in, 2329 primary vs. secondary, 2358 Raynaud’s phenomenon and, 1168, 1308 sensory involvement in, 714, 1141, 1308–1310, 2328–2329, 2359–2360 vasculitic neuropathy and, 2340t, 2360 incidence of, 2340t pathology of, 803 primary biliary cirrhosis and, 2021 relative prevalence of, 2341t Skeletal deformities, 1151 of foot, 1151, 1168, 1168f Skeletal survey, 1174 Skin anatomy of, 870–872, 871f decubitus ulcers, 2629 diseases of, 1167–1168 in cutaneous leishmaniasis, 2165–2167 in Fabry’s disease, 1168, 1893, 1894, 1899f in POEMS syndrome, 2456t, 2457f, 2458 in sarcoidosis, 2420 intensive care unit care for, 2609 lesions in leprosy, 1168, 2084–2086, 2085f lesions in Lyme disease, 1168, 2110–2111, 2111f measuring vasomotor reflexes in, 1122–1123 mounding phenomenon of, in hypothyroid polyneuropathy, 2041 nerves of, 869–883. See also Cutaneous nerves. methods for analysis of, 872–876 microscopy of, 876, 877f quantitation of, 876–883 pigmentary changes in Strachan’s syndrome, 2064 sensory receptors of, 1004–1007, 1064, 1064f low-threshold C fibers, 1007 low-threshold mechanoreceptors, 1005 nociceptors, 1005–1006 thermoreceptors, 1005 sympathetic skin response, 1121–1122, 1121f–1122f vascular responses to temperature changes, 333, 334f Skin biopsy choice of location for, 872, 873f depth of, 872 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxix Skin biopsy (Continued) fixatives for, 872–873 in diagnosis, 883–888 atopic dermatitis, 888 CADASIL, 887 chronic inflammatory demyelinating polyradiculoneuropathy, 887 congenital insensitivity to pain with anhidrosis, 886 cryptogenic sensory polyneuropathy, 2325–2326 diabetic neuropathy, 884–885, 885f diabetic truncal neuropathy, 885 Fabry’s disease, 887 familial dysautonomia, 886 Friedreich’s ataxia, 886 giant axonal neuropathy, 888 HIV-associated sensory neuropathies, 885 leprosy, 887, 2083, 2086 neuroaxonal dystrophies, 888 nodular prurigo, 888 notalgia paresthetica, 888 other conditions, 888 painful sensory neuropathy, 883–884, 884f pediatric disorders, 888 postherpetic neuralgia, 887 postural tachycardia syndrome, 887–888 post-wine stains, 886 psoriasis, 886 Raynaud’s syndrome, 888 restless legs syndrome, 886 sensory ganglionopathies, 887 systemic sclerosis, 888 Meissner’s corpuscles on, 1064f methods for, 872–873 morphometry of nerve or nerve fiber counts in, 1046 punch instrument for, 872, 874f research uses of, 888–891 animal models, 890–891, 891f human models of nerve regeneration, 888–890, 889f–890f staining specimens from, 873–874, 874f–875f utility of, 883 Skin blisters choice of location for, 872 method for, 875f, 875–876 quantification of epidermal nerves in, 879, 880f research uses of, 889 Skin sympathetic nerve activity (SSNA) characteristics of, 325 response to exercise, 328 site for recording of, 325 SLC12A6 gene mutations, 1738, 1770t, 1781–1782 Slit/Robo, 456–459, 458f modulation of signaling by, 458–459 cross talk with DCC/netrin signaling, 459, 462 intracellular cGMP levels, 459 receptor degradation or proteolytic cleavage, 458–459 receptor complexes of, 456–457 Robo structure and differential expression patterns, 456–457 Slit proteins, 457 Slit structure and proteolytic cleavage, 457 signaling by, 457–458 Abl and Mena, 457 Abl target capulet, 457

lxxx Slit/Robo (Continued) linkage to Rho GTPases and actin dynamics, 458 protein tyrosine phosphatases, 457–458 Slow component a (SCa), in axonal transport, 387, 390 Slow component b (SCb), in axonal transport, 387, 390 Slow excitatory postsynaptic potentials in enteric nervous system, 264f, 265–269 functional significance of, 267–269, 268f ionic mechanisms of, 265 mediators of, 265–266, 266t signal transduction in AH neurons, 266–267 signal transduction in S neurons, 267 Slow inhibitory postsynaptic potentials in enteric nervous system, 269–270 functional significance of, 270 mediators of, 269t, 269–270 Slow-channel congenital myasthenic syndromes, 847t, 847–851 clinical features of, 847, 847f end-plate studies in, 848–849, 848f–850f molecular pathogenesis of, 848f, 849–851 therapy for, 860 SMA. See Spinal muscular atrophy (SMA). Small fiber sensory neuropathy (SFN), 1146–1147, 1165–1166, 2321 autonomic function testing in, 1104, 2324 clinical course of, 2323 electrodiagnostic testing in, 2323–2324 intravenous immunoglobulin for, 2327, 2479 nerve biopsy in, 870, 2325 pain in, 2322 paraneoplastic, 2479 quantitative sensation testing in, 2324 skin biopsy in, 883–884, 884f Small intestine in Chagas’ disease, 2159–2160 in coenuriasis, 2168–2169 in cyclosporiasis, 2173 in cysticercosis, 2168 in giardiasis, 2167 Small-cell lung carcinoma (SCLC) antibodies in, 2471–2472, 2472t anti-amphiphysin, 586 anti-CV2, 586 anti-Hu, 584, 585, 594, 715, 2473–2474, 2474f, 2475 Lambert-Eaton myasthenic syndrome and, 597, 834, 1200 neurofilament expression in, 586 paraneoplastic neuropathy and, 2471–2473 autonomic neuropathy, 2480 subacute sensory neuronopathy, 2473–2476 SMN1 gene mutations, 1535, 1770t, 1779 SMN2 gene mutations, 1535 Snail fever. See Schistosomiasis. SNAPs. See Sensory nerve action potentials (SNAPs). SNE (subacute necrotizing encephalomyelopathy), 1941, 1944–1945, 1945f, 2715 SNP (subepidermal neural plexus), 871, 871f, 872, 873f in pediatric neuroaxonal dystrophy, 888 quantitation of nerves in, 879–880, 881f Social rehabilitation, for leprosy, 2105 SOD1. See Superoxide dismutase 1 (SOD1). Sodium channel myasthenia, 857–859 clinical features of, 857, 858f end-plate studies in, 857–858 molecular pathogenesis of, 858–859, 859f

Index Sodium channel myasthenia (Continued) relation to other sodium channel disorders, 859 therapy for, 860 Sodium channels, 96–99, 189 ␣ subunits, 189 antibodies against, 590 in Lambert-Eaton myasthenic syndrome, 590 axotomy-induced alterations of, 695 ␤ subunits, 189 effect of depolarization on, 126 expression by dorsal root ganglion neurons, 189 genes for, 95 mutations of, 97 of neuromuscular junction, 831 on axons, 96f, 96–98 at nodes of Ranvier, 55, 56, 96–98, 415, 577 function of, 95, 99 in constricted axon segment, 57–59 mechanisms of segregation of, 97–98 molecular identity of, 96–97 on Schwann cells, 98f, 98–99 role in neuropathic pain, 2638 upregulation by nerve growth factor, 176 Sodium-potassium pump effects of ischemia on, 119–121 in multifocal motor neuropathy, 123–124 Soft tissue perineurioma, 2594 Solar plexus, 234 Soleus muscle of cat, 132–135, 133f, 134f Somatosensory evoked potentials in brachial plexopathy, 1352–1353 in sensory variant of chronic inflammatory demyelinating polyradiculoneuropathy, 2229 in tropical spastic paraparesis, 2076 Somatostatin expression after axotomy, 691 expression by dorsal root ganglion neurons, 176t, 185 for orthostatic hypotension, 2677 functions of, 185 in ganglionic neurons, 240 receptors for, 185 Southern blot, 1554 Sox10 in Déjerine-Sottas disease and congenital hypomyelinating neuropathy, 2714 in early glial development from neural crest cells, 347–348 in late stages of peripheral glial development, 360 in Schwann cell development, 60, 342, 1708 SP. See Substance P (SP). Spanish Civil War neuropathy, 2065 Spanish Town disease in Jamaica, 2065 Spartin mutations, 1743 Spasm(s) hemifacial, 1145–1146, 1238–1245 hemimasticatory, 1146 of amputation stumps, 1145 Spastic paraplegia autosomal dominant, 1742t, 1742–1743 autosomal recessive, 1742, 1742t, 1743 lipofuscin in, 703 hereditary motor and sensory neuropathy with, 1625t, 1626, 1742t, 1742–1743, 1771 hereditary sensory neuropathy with, 1837 vs. amyotrophic lateral sclerosis, 1536 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Spastin mutations, 1743 SPECT (single-photon emission computed tomography), imaging of sympathetic innervation to organs, 1119–1121, 1120f Spectrin, 415 Speech audiometry, 1256 Sphenopalatine ganglion, 234 Spheroids, axonal. See Axonal dystrophy. Sphincteric achalasia, 260–261 Sphingolipid activator protein 1 deficiency, 710 Sphingolipid activator protein C deficiency, 706 Sphingolipid-cholesterol rafts, 423 Sphingolipidoses, 1846, 1847–1862 ceramide deficiency (Farber’s disease), 1859 gangliosidoses, 1859–1860 glucosylceramide lipidosis (Gaucher’s disease), 1860–1861 lysosomal leukodystrophies, 1847–1859 childhood, 2721 globoid cell leukodystrophy (Krabbe’s disease), 1855–1859, 1856f–1859f infantile, 2714 metachromatic leukodystrophy, 1847–1854, 1848f–1853f multiple sulfatase deficiency, 1854–1855, 1854f–1855f sulfatide activator protein-B deficiency, 1855 Niemann-Pick disease, 1861f, 1861–1862 Sphingolipids, 1847 Sphingomyelin lipidosis, 708 Spinal cord. See also Central nervous system (CNS). anatomy of, 307–308, 1297–1298 compression of, in chronic inflammatory demyelinating polyradiculoneuropathy, 2225 injury of cardiac vagal overactivity and, 1273, 1282f–1283f gastrointestinal manifestations of, 2699–2700 lower urinary tract dysfunction and, 300f, 313–314 plasma catecholamine levels in, 1117 sexual dysfunction and, 317 morphologic response to peripheral axonal injury, 1298 pathways controlling lower urinary tract, 307–309 afferent projections, 308, 309f efferent pathways, 308 spinal interneurons, 308–309, 309f pathways in sexual response cycle, 317–318 postmortem removal and processing of, 739f, 739–740 tethered, 1333–1334, 2738 Spinal cord disease, 1295–1316 combined motor and sensory syndromes, 1310–1316 AIDS-related vacuolar myelopathy, 1313–1314 bulbospinal muscular atrophy, 1316 cervical spondylotic myelopathy, 1310–1313 neurosyphilis, 1314–1316 in hereditary motor and sensory neuropathy, 1641

Index Spinal cord disease (Continued) motor syndromes, 1298–1307 amyotrophic lateral sclerosis, 1304–1306 classification of, 1298 lower motor neuron disease, 1298–1304 in HIV infection, 1303 other causes of, 1304 paraneoplastic lower motor neuron syndrome, 1302–1303 poliomyelitis, 1300–1301 postirradiation lower motor neuron syndrome, 1303 West Nile virus infection, 1301–1302 multifocal motor neuropathy and conduction block, 1306–1307 overview of peripheral nerve involvement in, 1296–1297 peripheral neuropathy related to, 1295–1316 sensory syndromes, 1307–1316 Sjögren’s syndrome, 1308–1310 vitamin B12 deficiency, 1307–1308, 2054 sexual dysfunction and, 2628 tropical myeloneuropathies, 2063–2077 vs. peripheral neuropathy, 1153 Spinal dysraphism, 1333–1334 Spinal ganglia, 11 morphometry of neuron somas in, 755–756, 758t, 759f–760f Spinal metastasis, 1330–1331 Spinal muscular atrophy (SMA), 1603. See also Hereditary motor neuronopathy (HMN). classification of, 1603–1605, 1604t, 1605t differential diagnosis of motor axonopathies, 1153 peripheral neuropathy, 1152, 1176 distal, 1299, 1604t, 1613t, 1613–1615, 1614f genetics of, 1548 infantile, 1603, 1604t, 1605t, 1606–1607, 2712, 2713 inherited canine, 712 lower motor neuron disease in, 1299 macro electromyography in, 993f pathology of, 705 progressive, 1299 overlap with amyotrophic lateral sclerosis, 1304 proximal, 1138, 1604t, 1606–1612, 1607t segmental, 1299, 1604t, 1613t, 1618 severe congenital (type 0), 1779 spectrum of, 704 terminology for, 1603 type I, 704, 1603, 1604t, 1605t, 1606–1607, 1607t, 1608f type II, 1604t, 1605t, 1607t, 1607–1610, 1610f type III, 704, 1604t, 1605t, 1607t, 1610–1611 types IV and V, 1604t, 1605t, 1607t, 1611–1612 upper and lower extremity involvement in, 993f Spinal nerve roots, 1295, 1323–1334 anatomy of, 11, 1297–1298, 1323–1324 gross, 1323, 1324f microscopic, 1324 attachment zones of rootlets of, 67, 68f avulsion from spinal cord, 1327 blood supply of, 1324 collagen in, 37 congenital anomalies affecting, 1333–1334 diagnosis of radiculopathy affecting, 1324–1325

Spinal nerve roots (Continued) disorders affecting single nerve roots, 1325–1327 acute intervertebral disc herniation, 1325–1326 miscellaneous causes, 1327 trauma, 1326–1327 dorsal, 11, 1323, 1339 morphometry of, 766f electromyography in localization of lesion to, 955–957 in cervical spine, 1323, 1324f in chronic arachnoiditis, 1332–1333 in polyradiculopathy, 1327–1334 diabetic, 1332 miscellaneous causes, 1334 in radiation radiculopathy, 1332 in spondylosis, 1328–1329, 1329f in toxic meningoradiculopathy, 1333 infections affecting, 1331–1332 neoplastic disease of, 1329–1331 pediatric lesions of, 2736–2738 cervical, 2737 lumbosacral, 2737–2738 rootlet sheaths of, 77 sheaths of, 11, 1324 transitional zone with peripheral nervous system, 67–77. See also Central nervous system–peripheral nervous system transitional zone. tumors of, 2738 ventral, 11, 1323, 1339 morphometry of, 763–765, 764f–765f vulnerability to ischemia, 1324 Spinal nerves, 11–12 distribution of, 12, 12f general features of, 11 meningeal branches of, 31 Spinal schistosomiasis, 2160 Spinal stenosis cervical, 1328 lumbosacral polyradiculopathy due to, 1328–1329 Spinobulbar muscular atrophy. See Bulbospinal muscular atrophy. Spinobulbospinal micturition reflex pathway, 310–311, 311f Spinocerebellar ataxia (SCA), 2330t autosomal recessive, 1738–1741 Friedreich’s ataxia, 1738–1739 spinocerebellar atrophy with axonal neuropathy, 1738–1741 in children, 2724 pathology of, 706, 707f type 1, 706 type 2, 706 type 3, 706 type 6, 706 type 7, 706, 707f Spinocerebellar atrophy, calcium channel mutations in, 590 Spinocerebellar atrophy with axonal neuropathy (SCAN1), 1738–1741 Spinocerebellar degeneration, auditory and vestibular neuropathy in, 1259 Spinocerebellar tracts, 158 Spinoglenoid ligament, 1466 hypertrophy of, 1467 Spinoglenoid notch, 1466–1468 Spiral bands of Fontana, 42 Spiral ganglion, 1253 degeneration of, 1268 Spirometry, 2612–2613 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxxi Splanchnic blood flow, 2695 Splanchnic nerve(s), 17 greater, 25 lesser, 25 lowest, 25 lumbar, 30, 281, 281f pelvic, 18, 28, 31, 281, 281f sacral, 30, 31 Splanchnic-mesenteric capacitance bed evaluation of, 1124–1125, 1125f orthostatic hypotension and, 2656, 2662 Spleen in Fabry’s disease, 1896 in POEMS syndrome, 2455 in sarcoidosis, 2420 Spondylosis cervical, with myelopathy, 1310–1313 radiculopathy due to, 1328–1329 Spondylotic radiculopathy, 1311, 1328–1329, 1329f Spontaneous repetitive motor action potentials (SRMAPs), organophosphate-induced, 2516 Spranger syndrome, 1865 Sprawling mouse, 711 SPTLC1 gene, 424 in hereditary sensory and autonomic neuropathy type I, 1810, 1811t, 1819–1823, 1822f SPTLC2 gene, 424 Src kinase, in CRASH syndrome, 468 SRMAPs (spontaneous repetitive motor action potentials), organophosphate-induced, 2516 SSCP (single-strand conformation polymorphism), 1557 SSN. See Subacute sensory neuronopathy (SSN), paraneoplastic. SSNA (skin sympathetic nerve activity) characteristics of, 325 response to exercise, 328 site for recording of, 325 SSR (sympathetic skin response), 1121–1122, 1121f–1122f SSRIs (selective serotonin reuptake inhibitors), for neuropathic pain, 2641–2642 Stair-climbing test, 1096 Stapedius muscle, nerve to, 1220, 1222, 1224f Statistical methods for computing normal test values, 973–976 multiple explanatory variables, 975–976 multiple independent variables and multiple dependent variables, 976 single dependent variable, 973–975, 974t single independent variable, 975 Status asthmaticus, 2031 Status epilepticus, 466 Stavudine (d4T), 2133–2135, 2134f–2135f, 2563 Stellate (cervicothoracic) ganglion, 17f, 23, 25 Stem cell transplantation for adrenoleukodystrophy, 1868 for amyloidosis, 2439 for chronic inflammatory demyelinating polyradiculoneuropathy, 2236 Sternotomy, median, brachial plexus lesion after, 1364–1365 Stevens-Johnson syndrome, lamotrigine-induced, 2643 Stiff-man syndrome, 942 Stocking-and-glove distribution of sensory neuropathy, 1138, 1139f, 1140

lxxxii Stockings, elastic, for orthostatic hypotension, 2658–2659, 2674 Stomach, in Chagas’ disease, 2159 Stool softeners, 2692, 2693t Strachan’s syndrome, 2057, 2064–2065 clinical features of, 2064 etiology of, 2064–2065 arsenic intoxication, 2064–2065 neuropathology of, 2064 skin pigmentation in, 2064 “Strawberry picker’s palsy,” 1142 Strength-duration curves, 113–114, 114f, 116f, 116–117, 120f Streptomycin, for leprosy, 2103 Stress effects on blood pressure and heart rate, 331 lipofuscin accumulation and, 703 vasodilation in skeletal muscle and, 331–332 Stress gastric ulcers, prophylaxis for, 2609–2610 Stroke, dysphagia after, 2685, 2687 Stromelysin-1, in Schwann cell proliferation, 1409 Strongyloidiasis, 2161 STRs (short tandem repeats), 1552 Styloglossus muscle, innervation of, 1286, 1287 Stylohyoid artery, 1221 Stylohyoid muscle, innervation of, 1220, 1222 Stylomastoid artery, 1221 Stylomastoid foramen, 1220–1221, 1224f Stylopharyngeus muscle, innervation of, 1273 Subacute idiopathic demyelinating polyradiculoneuropathy (SIDP), 2222–2223 Subacute necrotizing encephalomyelopathy (SNE), 1941, 1944–1945, 1945f, 2715 Subacute sensory neuronopathy (SSN), paraneoplastic, 2473–2476 age at onset of, 2474 autoantibodies in, 2473–2474, 2474f, 2475 clinical features of, 2474 course of, 2474 differential diagnosis of, 2475 electrophysiology of, 2475–2476 pathology of, 2476, 2477f treatment of, 2476 underlying malignancy in, 2475 Subarachnoid space, spinal nerve roots in, 1323–1324 Subclavian artery, 1341 Subclavian nerve, 19 Subclavian vein catheterization, brachial plexus injury from, 1365 Subcostal nerve, 23, 24–25 Subepidermal neural plexus (SNP), 871, 871f, 872, 873f in pediatric neuroaxonal dystrophy, 888 quantitation of nerves in, 879–880, 881f Sublingual glands, innervation of, 1222 Submandibular ganglion, 17, 234, 17 Submandibular glands, innervation of, 1222 Submucosal plexus, 252, 252f, 2683 Suboccipital nerve, 31 Subsartorial plexus, 26 Subscapular artery, 1341 Subscapular nerve lower, 20 muscles innervated by, 956t upper, 20 Substance P (SP) aging effects on, 486 effect on nerve blood flow, 670, 670t excitatory amino acids and, 527

Index Substance P (SP) (Continued) expression after axotomy, 690–691 expression by dorsal root ganglion neurons, 176t, 184–185 fast anterograde axonal transport of, 389 functions of, 184 in autonomic ganglia, 236, 242 in bladder, 301 in diabetic neuropathy, 1978 in enteric nervous system, 265–266 in epidermal nerve fibers, 872 in hyperalgesia, 527 in micturition reflex, 312 in sweat gland innervation, 1106 receptors for, 185 upregulation by nerve growth factor, 176 Sudomotor axon reflex test. See Quantitative sudomotor axon reflex test (QSART). Sudomotor nerve density, 880–881, 882f Sulfated glucuronyl paragloboside (SGPG), 584, 597 IgM binding to, 2188 structure of, 583t Sulfated glucuronyllactosaminyl paragloboside (SGLPG), 583t, 584 Sulfatide IgM antisulfatide antibodies, 2189, 2268 antibody testing for, 2189 clinical disease associated with, 2189 disease pathogenesis and, 2189 in myelin membrane, 423, 580 localization of, 578f mice deficient in, 550, 584 structure of, 583t Sulfatide activator protein-B deficiency, 1855 SUNCT syndrome (short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing), 1214 Superexcitability, 117 in multifocal motor neuropathy, 123, 123f Superior laryngeal neuralgia, 1284 Superoxide dismutase 1 (SOD1), 511 activity after axotomy, 693 enzymatic pathway of, 1586, 1586f for treatment of experimental diabetic neuropathy, 516, 516t in diabetic patients, 511 peripheral nerve levels, 515, 515t in experimental diabetes, 512, 515t, 673 in neuronal death, 699 mutations of, 1586 in amyotrophic lateral sclerosis, 1534–1535, 1539, 1586, 1596–1598 double transgenic models, 1589 SOD1 transgenic models, 1585–1588, 1586f protective effects against ischemia-reperfusion injury, 676 Suppressed cAMP-inducible POU (SCIP), 1708 Supraclavicular ligament, 1466 Supraclavicular nerve, 15f, 18, 1470f Suprascapular artery, 1341 Suprascapular nerve, 19, 1340f brachial plexus lesion after blockade of, 1365 distribution of, 1466, 1466f muscles innervated by, 956t, 1466f Suprascapular neuropathy, 1466–1469 clinical manifestations of, 1468 diagnosis of, 1468 etiologies of, 1466–1468 iatrogenic, 1467 in children, 2727 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Suprascapular neuropathy (Continued) prognosis for, 1469 traumatic, 1467 treatment of, 1468–1469 Suprascapular notch, 1466–1468 Supraspinatus muscle, 1466–1468 Sural cutaneous nerve lateral, 16f medial, 16f, 29 Sural nerve, 29 biopsy of, 37, 737–739, 738f histologic methods for, 740, 741f–742f myelinated and unmyelinated fiber counts, 37 symptoms after, 739 compound action potentials of, 1017–1026 fiber diameter and, 1018 in dominantly inherited amyloidosis, 1021f, 1021–1022 in Friedreich’s ataxia, 1020f, 1020–1021 in hereditary motor and sensory neuropathy type I, 1025–1026, 1027f in hereditary sensory neuropathy type I, 1022, 1023f in hereditary sensory neuropathy type II, 1022–1023, 1024f in normal nerve, 1018–1020, 1019f in relapsing inflammatory polyradiculoneuropathy, 1024–1025, 1026f–1027f in tabes dorsalis, 1022, 1022f in uremic neuropathy, 1023–1024, 1025f method of study of, 1017–1018 sources of distortion in recording of, 1017 distribution of, 1507, 1508f epidermal nerve fiber density and morphometry of, 883 in leprosy, 2089f lesions of, 1507–1508 in children, 2730 morphometry of, 768f–769f nerve conduction studies of, 904–905, 916–917, 919f Suramin neuropathy, 2567–2568 SURF1 gene mutations, 1945 Surgery for axillary neuropathy, 1471 for Bell’s palsy, 1231 for brachial plexus lesion, 1523–1525 obstetric, 1525–1528, 2731–2733 for carotid sinus hypersensitivity, 1277 for carpal tunnel syndrome, 1452 for chronic intestinal pseudo-obstruction, 2691 for fecal incontinence, 292–293 for hemifacial spasm, 1242–1245 for hereditary neuropathy with liability to pressure palsies, 1669 for leprosy, 2104–2105 for peroneal neuropathy, 1499–1500 for radial neuropathy, 1478 for Ramsay Hunt syndrome, 1232 for serratus anterior paralysis, 1465 for suprascapular neuropathy, 1469 for tarsal tunnel syndrome, 1505 for ulnar neuropathy, 1483 on peripheral nerves, 1511–1528 alternative methods of repair, 1516 electrophysiologic studies during, 1516 iatropathic nerve injury from, 1516–1521 indications for, after nerve injury, 1511 principles of, 1515–1516 reasons for, 1511 Swallow syncope, 1277

Index Swallowing difficulty, 2633, 2685–2687 causes of, 2685–2686 clinical assessment of, 2687 evaluation of, 2687 hypoglossal nerve lesions and, 1287 in multiple sclerosis, 2700 in myasthenia gravis, 2686, 2700 in Parkinson’s disease, 2695 management of, 2687 palatal paralysis and, 1279 pharyngeal paralysis and, 1279 postpolio, 2695 Sweat glands, innervation of, 1106 nerve density, 880–881, 882f Sweating, 333, 1150 atropine blockade of, 333 gustatory, 1150, 1282–1283, 1283f, 1959 in diabetic patients, 1150, 1959 quantitative sudomotor axon reflex test of, 1106–1109 regulation of, 1106 thermoregulatory sweat test, in distal small fiber neuropathy, 1104 Sweet’s syndrome, 2458 Swellings, axonal. See Axonal dystrophy. Sympathetic ganglia, 15–16, 234–235 age-related changes in, 493–494 in abdomen, 30 in head, neck, and upper limb, 15–16 in thorax, 25 morphometry of neuron somas of, 756, 760f neuropeptides in, 240t, 240–243 paravertebral (chain), 234 prevertebral, 234–235 Sympathetic nerves age-related changes in neurons, 492–495, 493f–495f, 702t, 702–703, 703t aging and activity of, 332f, 332–333 baroreflexes and, 323–324, 324f characteristics of peripheral sympathetic traffic, 325 cutaneous vascular responses to body heating, 333, 334f direct measurement of activity via microneurography, 324, 324f imaging of sympathetic innervation to organs, 1119, 1120f MIBG SPECT study, 1120–1121 in congestive heart failure, 333 in hypertension, 333 muscle activity of autonomic responses to exercise and, 328–331 baroreflex resetting, 328–329, 329f central command, 328 ␣-receptor subtypes and blood flow in active muscles, 331 sensory feedback from active muscles, 328, 329f sympathetic vasoconstriction and metabolic vasodilation, 330, 330f baroreceptor control of, 325, 325f blood pressure and, 324–325 effect of glossopharyngeal and vagus nerve block on, 325, 326f characteristics of, 325 in aging, 332f, 332–333 in congestive heart failure, 333 in hypertension, 333 microneurography for measurement of, 324, 324f responses to chemoreceptor activation, 327–328, 328f

Sympathetic nerves (Continued) responses to mental stress, 331–332 neurally mediated dilation in skeletal muscle, 331–332 responses to orthostatic challenge, 326–327, 327f site for recording of, 325 responses to chemoreceptor activation, 327–328, 328f responses to exercise, 328–331 responses to sympathetic challenge, 326–327, 327f site for recording traffic of, 325 skin activity of characteristics of, 325 response to exercise, 328 site for recording of, 325 to iris, 203, 204f–206f pharmacology in Horner’s syndrome, 207–208 to pelvic floor, 281f, 281–282 vasodilators and vasoconstrictors, 323 Sympathetic nervous system, 15–17, 17f, 233 anatomy of, 15–17, 17f cardiac, 217–218, 219f communication with parasympathetic system, 224–225 in abdomen, pelvis, and lower limb, 30–31 in head, neck, and upper limb, 23 in lower urinary tract, 302–304 in thorax, 25 in urogenital system, 315 Sympathetic skin response (SSR), 1121–1122, 1121f–1122f Sympatholysis, functional, 330 Sympathomimetic agents, for orthostatic hypotension, 2675–2676, 2676t Symphysis pubis, 280f Symptoms and signs of associated disease, 1167–1170 neurologic disease, 1170, 1170t non-neurologic disease, 1167–1170, 1169t constitutional symptoms, 1167 of peripheral neuropathy, 1142–1152 autonomic, 1150–1151, 1166 disability status and, 1033 intermittent, 1141–1142 motor, 1142–1146, 1165 negative, 1142 positive, 1142–1146 nerve thickening, 1151–1152 onset and progression of, 1141, 1167 quantitation of, 1031–1049 sensory, 1146–1150, 1165–1166 negative, 1146–1148 positive, 1148–1150 skeletal deformity, 1151 trophic changes, 1151 Synaptic neuropils, of enteric nervous system, 253 Synaptic structures, of autonomic ganglia, 235–236 Synaptic transmission, in enteric nervous system, 263–271 Synaptic vesicle exocytosis, 832 Synaptophysin, in amyotrophic lateral sclerosis, 1305 Syncope, 326 autonomic function testing in, 1104–1105 isoproterenol infusion test, 1124 due to cardiac vagal overactivity, 1281–1282, 1284 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxxiii Syncope (Continued) emotional, 1282 in amyloidosis, 2433 treatment of, 2439 in carotid sinus hypersensitivity, 1277 in glossopharyngeal neuralgia, 1276 swallow, 1277 Syndrome of crocodile tears, 1275–1276 ␣-Synuclein, in neuronal death, 699 Syphilis. See Neurosyphilis. Syringomyelia, 997 Systemic inflammatory response syndrome (SIRS), 2017, 2026–2030. See also Critical illness polyneuropathy. Systemic lupus erythematosus (SLE), 1172, 2358 diagnostic criteria for, 2358, 2359t eighth nerve damage in, 1268 neuropathy and, 2358 pathogenesis of, 2358 pediatric polyneuropathy and, 2719, 2720f vasculitic neuropathy and, 2340t, 2354t, 2358 incidence of, 2340t pathology of, 803 relative prevalence of, 2341t Systemic sclerosis, 1172 skin biopsy in, 888

T cells, 560–563, 574, 609–610 activation of, 562, 562f crossing of blood-nerve barrier by, 559, 560f, 563 cytokine profiles of, 2344 cytotoxic T lymphocyte-mediated vasculitis, 2348, 2349f in chronic inflammatory demyelinating polyradiculoneuropathy, 2243–2245 in experimental autoimmune neuritis, 617–618, 620–621 in P0-deficient mice, 537–538 in sarcoidosis, 2416–2417, 2421 interaction with antigen-presenting cells, 562–563, 563f stages of response, 564f subpopulations of, 561–562 CD4⫹(helper) cells, 561–563 CD8⫹(cytotoxic) cells, 561–563 thymic education of, 561 tolerance of, 574, 610 T reflex, 919 TA. See Temporal arteritis (TA). Tabes dorsalis, 1315, 2310, 2316. See also Neurosyphilis. clinical features of, 1315 compound action potential of sural nerve in, 1022, 1022f electrophysiologic studies in, 1315 glossopharyngeal nerve involvement in, 1274 vs. sensory neuropathy, 1153 Tachycardia, 226 postural tachycardia syndrome autonomic function testing in, 1104 skin biopsy in, 887–888 Tacrolimus, 638–639 adverse effects of, 639 as protection against antiretroviral toxic neuropathy, 2136, 2563 mechanism of action of, 638–639 neuropathy induced by, 2568

lxxxiv Tacrolimus (Continued) to increase axonal regrowth after chronic axotomy, 1411, 1411f use in neuropathy, 639 Tactile sensitivity, aging and, 484 Taenia (Multiceps) spp. infection. See Coenuriasis. Taenia solium infection. See Cysticercosis. TAG-1, 418f, 419t, 421 Takatsuki syndrome. See POEMS syndrome. Takayasu’s arteritis classification and definition of, 2338t diagnostic criteria for, 2338 genetics of, 2363 incidence of, 2340t TAN. See Tropical ataxic neuropathy (TAN). Tangier disease (TD), 1905–1916, 2330t ABCA1 in, 1905, 1906, 1911–1914 expression of, 1913 genetics of, 1912–1913 human mutations of, 1914, 1914t in intracellular trafficking, 1913 knockout mouse models of, 1914 other functions of, 1913–1914 regulation of, 1913 topologic arrangement of transporter for, 1913, 1915f diagnosis of, 1908–1909 differential diagnosis of, 1909, 1910t familial high-density lipoprotein deficiency, 1905, 1908 spinal cord disease, 1153 distribution of involvement in, 1164 proximal, 1139, 1140f sensation abnormality, 1090f epidemiology of, 1908 extraneural, 1908 high-density lipoprotein and plasma cholesterol transport in, 1906–1908 in children, 2722 inheritance of, 1905 lipofuscin in, 703 lipoprotein metabolism and, 1905–1906, 1907t neuropathophysiology of, 1914–1915 neuropathy in, 1908 course of, 1908 subtypes of, 1908 orange tonsils of, 1167 pathogenesis of, 1911–1912 pathology of, 710, 1909, 1911f–1912f small fiber sensory loss in, 1147 treatment of, 1915–1916 Tapia’s syndrome, 1285 Tardy posterior interosseous neuropathies, 1475 Tardy ulnar palsy, 1480 nerve conduction studies in, 910 Tarsal tunnel syndrome (TTS), 1503–1505 anatomy of, 1502f, 1503 anterior vs. medial, 1503 clinical manifestations of, 1504 diagnosis of, 1504–1505 etiology of, 1504 prognosis for, 1505 treatment of, 1505 Taste sensation, 1222 trigeminal nerve lesions and, 1207 Taxane-induced neuropathy, 2490, 2490t, 2498t, 2568–2570 Tay-Sachs disease, 1860. See also GM2 gangliosidoses. TCAs (tricyclic antidepressants) adverse effects of, 2641 for neuropathic pain, 2640, 2641–2642

Index TCE (trichloroethylene), 1211, 2520 T-cell depletion, for vasculitic neuropathy, 2381 T-cell leukemia, 2496 T-cell receptor, 560–561, 561f T-cell tolerance, 567, 568f TCS (terminal cytoplasmic spiral), 55 TD. See Tangier disease (TD). TDP1 gene mutations, 1739–1741 Teased fiber analysis, 742–754 background of, 742 classification of pathologic abnormalities on, 747–749 axonal degeneration, 748, 749f empty nerve strands, 748 myelin wrinkling, 747–748, 748f normal fiber, 747, 747f proximal segmental demyelination and distal axonal degeneration, 748, 751f regeneration after axonal degeneration, 748, 751f remyelination, 748, 750f segmental demyelination, 748, 749f segmental demyelination and remyelination, 748, 749f tomacula, 745f, 748, 750f clinical usefulness of, 750 comparison of frequency of abnormalities on, 751 electron microscopic assessment for, 753–754, 755f fixation and teasing artifacts in, 743–744, 746f, 769, 772f for clustering of demyelination, 753, 754f frequency of graded abnormalities in control nerves, 752, 752t histologic technique for, 743 in acute inflammatory demyelinating polyneuropathy, 2202 in axonal degeneration, 745f in chronic inflammatory demyelinating polyradiculoneuropathy, 2241f in critical illness polyneuropathy, 2028, 2028f in diabetic neuropathy, 1965 in Friedreich’s ataxia, 1740f in giant axonal neuropathy, 1736f in hereditary motor and sensory neuropathy Déjérine-Sottas neuropathy, 1642, 1647f type I, 1640, 1643f–1644f type II, 1644, 1650t with P0 mutation, 1695–1696 in hereditary neuropathy with liability to pressure palsy, 745f in hereditary sensory and autonomic neuropathy type I, 1820f–1821f in leprosy, 2100f in lumbosacral radiculoplexus neuropathy, 2006t, 2009, 2011f diabetic, 2005, 2006t in tropical ataxic neuropathy, 2070 internode length and diameter on, 752–753, 753t of amputated nerve stumps, 787f of wallerian degeneration, 773 sampling for, 744–747 sources of variability in grading abnormalities on, 752 teasing technique for, 743, 744f–745f types of evaluation in, 747 uses of, 743 Teloglia, 341 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Temperature cutaneous vascular responses to changes in, 333, 334f effects on nerve conduction, 125, 907, 907f, 908t protective effects of hypothermia against ischemia-reperfusion injury, 672, 676 Temporal arteritis (TA), 2337, 2354t, 2362 carpal tunnel syndrome and, 1445 classification and definition of, 2338t, 2340t cytokines in, 2344 diagnostic criteria for, 2338 genetics of, 2363 incidence of, 2340t polymyalgia rheumatica and, 2362 treatment of, 2336, 2376t Temporal artery, superficial, 1221 Temporal bone fracture of, 1268 facial nerve paralysis and, 1233, 1233f metastatic carcinoma of, 1268 Temporal dispersion, 931 pathologic, 931 physiologic, 931, 932f Temporal profile of symptoms, 1141, 1167 Temporofacial nerve, 1220 Tenascin C, 549 Tendon areflexia, 1146 in large fiber neuropathies, 1147 in peripheral nerve hyperexcitability syndromes, 1143 in poliomyelitis, 1300 in sensory ganglionopathy, 1148 Teniposide neuropathy, 2490 Tenosynovitis, carpal tunnel syndrome and, 1445–1446 TENS (transcutaneous electrical nerve stimulation), 2623, 2624 Terminal cytoplasmic spiral (TCS), 55 Termino-lateral neurorrhaphy, 1424–1425 Testicular plexus, 281 Test-retest reliability of health outcome measures, 1056 Tetanus, intracytoplasmic vacuolation in, 688 Tethered cord, 1333–1334, 2738 Tetracycline, for neurosyphilis, 1316 TGF-␣. See Transforming growth factor-␣ (TGF-␣). TGF-␤. See Transforming growth factor-␤ (TGF-␤). Thalidomide for leprosy reactions, 2104 for sarcoidosis, 2423 neuropathy induced by, 2490t, 2490–2491, 2498t, 2570 in children, 2720 Thallium neuropathy, 2540–2543 clinical features of, 2540–2541 laboratory studies in, 2541 tissue distribution and pathology of, 2541–2543, 2542f–2543f treatment of, 2543 use and sources of thallium exposure, 2540 Therapeutic exercise, 2622–2623 in motor neuron disease, 2633 Therapeutic heat and cold, 2630 Thermal sensation testing, 1084–1088 algorithm of testing and finding threshold for, 1076, 1077f calibration of stimulus magnitude for, 1075, 1075f

Index Thermal sensation testing (Continued) cooling detection threshold using CASE IV, 1085f, 1085–1086 estimating percentile abnormality, 1087–1088, 1088t warm detection threshold, 1086 warmth or heat-pain sensation, 1086–1087, 1087f–1088f Thermoreceptors, 1005 expression by dorsal root ganglion neurons, 168t, 169, 186–187 cold transduction, 187 heat transduction, 186–187 in enteric nervous system, 254 microneurography of, 1005 Thermoregulatory sweat test (TST) in distal small fiber neuropathy, 1104 in preganglionic vs. postganglionic lesions, 1108, 1108t Thiamine deficiency, 2051–2053, 2072–2073 causes of, 2051 clinical features of, 2052 diagnosis of, 2052 in children, 2720 management of, 2052–2053 pathogenesis of, 2052 Thin-layer chromatography, 2180–2181 Thiocyanate (cassava) toxicity, 2071, 2073–2075 clinical features of, 2073–2074 epidemiology of, 2073 konzo, 2074–2075 pathogenesis of, 2074 tropical ataxic neuropathy and, 2069, 2074 tropical spastic paraparesis and, 2074 urinary test for, 2075 6-Thiouric acid, 639 Third nerve palsy, 1195 diabetic, 804, 1201–1202 due to infections and toxins, 1202 due to neoplasms, 1199–1200 due to vascular diseases and aneurysms, 1201 traumatic, 1198 Thoracic inlet tumor, 1361–1362 Thoracic innervation, 23–25, 24f Thoracic nerves, 23, 32 Thoracic outlet syndrome(s) (TOS), 1370–1371 disputed neurologic, 1370–1371 brachial plexus injury after surgery for, 1365 in children, 2734 traumatic, 1366–1367 clavicular fracture and, 1366 clinical features of, 1366 disorders described by, 1366 electrodiagnostic evaluation of, 1366 exuberant callus syndrome, 1366 treatment of, 1366 true neurologic, 1359–1360 cause of, 1359, 1360f clinical features of, 1359 diagnosis of, 1359–1360 treatment of, 1360 types of, 1370t Thoracic radiculoplexus neuropathy diabetic, 1993 nondiabetic, 1994 Thoracic-lumbar-sacral orthosis (TLSO), 2623 Thoracoabdominal nerves, 24, 24f Thoracoacromial artery, 1341 Thoracodorsal nerve, 20 pediatric lesions of, 2727

THR (total hip replacement) arthroplasty nerve compression by hematoma after, 1518 sciatic or femoral neuropathy after, 1494–1496, 1512, 1519–1520, 1520f Threshold electrotonus, 118f, 118–119, 120f in multifocal motor neuropathy, 122–123, 123f Threshold tracking, 115 Thrombocytopenia, intravenous immunoglobulin for, 642, 643 Thrombocytosis, in amyloidosis, 2435 Thrombosis immobility and, 2629 in intensive care unit patients, 2616 in POEMS syndrome, 2460 prophylaxis for, 2609, 2629 Thromboxane B2, glucocorticoid effects on, 636 Thymosin ␤4, in actin polymerization, 451, 451f Thymus, in myasthenia gravis, 833 Thyrohyoid muscle, innervation of, 1286 Thyroid disease, 1172, 2039–2044 carpal tunnel syndrome and, 1445 hyperthyroid neuropathy, 2043–2044 hypothyroid neuropathy, 2039–2042, 2041f–2044f medullary thyroid carcinoma in multiple endocrine neoplasia type 2B, 1805–1807 Tibial nerve blood supply to, 801 distribution of, 16f, 27f, 29, 1494f, 1500, 1500f in leprosy, 2089f in tarsal tunnel, 1503 in tarsal tunnel syndrome, 1503–1505 involvement in vasculitic neuropathy, 2365t morphometry of, 767f muscles innervated by, 957t, 1500 nerve conduction studies of, 905f, 908t, 916, 916f, 917t Tibial neuropathy, 1500–1503 clinical manifestations of, 1501–1502, 1502f diagnosis of, 1502–1503 etiology of, 1500–1501 in children, 2730 prognosis for, 1503 treatment of, 1503 Tick paralysis, 2155t, 2173–2174 clinical features of, 2174 epidemiology of, 2173–2174 etiologic agents in, 2173 vs. childhood Guillain-Barré syndrome, 2201–2202, 2717 Tight junctions, perineurial, 38f, 39, 40 Timofeew’s corpuscles, 40 TIMPs (tissue inhibitors of matrix metalloproteinase), in POEMS syndrome, 2455 Tinel’s sign in brachial plexus injury, 1522 in carpal tunnel syndrome, 1441 in pronator syndrome, 1453 in tarsal tunnel syndrome, 1504 on day of nerve injury, 1513 Tinnitus due to vestibular schwannoma, 1264 in vestibular paroxysmia, 1264 Tissue inhibitors of matrix metalloproteinase (TIMPs), in POEMS syndrome, 2455 TLSO (thoracic-lumbar-sacral orthosis), 2623 TMNs. See Tropical myeloneuropathies (TMNs). TMS (transcranial magnetic stimulation), in multifocal motor neuropathy and conduction block, 2284 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxxv TNF-␣. See Tumor necrosis factor-␣ (TNF-␣). Tocainide, for neuromyotonia, 1144 ␣-Tocopherol. See Vitamin E. TOCP (triorthocresylphosphate), 1185, 2515, 2075 Tolerance and autoimmunity, 567–569, 574, 575 B-cell tolerance, 567 breakdown of tolerance, 568–569 T-cell tolerance, 567, 568f Toll receptors, 574 Tolosa-Hunt syndrome, 1202 Tomacula. See also Hereditary neuropathy with liability to pressure palsies (HNPP). in hereditary motor and sensory neuropathy with P0 mutation, 1695, 1695f in hereditary neuropathy with liability to pressure palsies, 1565 in Pmp22-deficient transgenic mice, 1566 teased fiber analysis of, 745f, 748, 751f Tongue, 1288 congenital atrophy of, 1287 effect of hypoglossal nerve lesion on, 1287 in glossodynia, 1288 innervation of, 1286 Tonic pupil, 210, 211f Topiramate adverse effects of, 2643 for diabetic neuropathy, 1971, 2640, 2643 Topoisomerase I, 1741 TOS. See Thoracic outlet syndrome(s) (TOS). Total hip replacement (THR) arthroplasty nerve compression by hematoma after, 1518 sciatic or femoral neuropathy after, 1494–1496, 1512, 1519–1520, 1520f Total Symptom Score (TSS), 1035, 1035t Touch-pressure sensation testing, 1038, 1077–1079, 1078f with CASE III system, 1069f Touch-Pressure System III, 1078, 1078f Tourniquet paralysis, 1517 Toxic neuropathies drug-induced, 2553–2573. See also Drug-induced neuropathy. effects on axonal transport motors, 394 from industrial agents, 2505–2520 acrylamide, 2506–2508 allyl chloride, 2508 carbamates, 2508–2510 carbon disulfide, 2510–2511 dichloracetylene, 2520 ethylene oxide, 2511 hexacarbons, 2511–2514 methyl bromide, 2514–2515 organophosphates, 2515–2519 polychlorinated biphenyls, 2519 pyriminil, 2519 trichloroethylene, 2520 from metals, 2527–2545 arsenic, 1168, 2064–2065, 2535–2538 gold, 2543–2544 lead, 2527–2535 mercury, 2538–2540 platinum, 2544–2545 thallium, 2540–2543 in children, 2720 intensive care unit management of, 2607 meningoradiculopathy, 1333 mydriasis and, 212 pathology of, 715–717, 716f axonal dystrophy, 778 chromatolysis, 686 intracytoplasmic vacuolation, 688 Schwann cell alterations, 820

lxxxvi Toxic neuropathies (Continued) sensory polyganglionopathy, 2310, 2315t, 2316 trigeminal nerve lesions from, 1211 tropical myeloneuropathies, 2069t, 2073–2076 cassava (cyanide), 2073–2075 lathyrism, 2075 pesticides, 2075–2076 vs. childhood Guillain-Barré syndrome, 2717 Toxoplasmosis, 2155t, 2172–2173 clinical features of, 2173 congenital, 2173 etiologic agent in, 2166f, 2172 life cycle of, 2172–2173 Guillain-Barré syndrome and, 2173 transmission of, 2172 TP gene mutations, 2722 Tracheostomy, 2614–2615 Tramadol for HIV-associated sensory polyneuropathies, 2135 for neuropathic pain, 2646 Transcapillary permeation, 653 Transcranial magnetic stimulation (TMS), in multifocal motor neuropathy and conduction block, 2284 Transcription, 1545 Transcription factors calcineurin activation of, 639 expression after axotomy, 692–693 glucocorticoid effects on, 636 in myelin gene regulation, 421–423 in myelination, 357–361, 358t, 359f in Schwann cell development, 342, 343, 1708 in Schwann cell survival, 354–356 Transcutaneous electrical nerve stimulation (TENS), 2623, 2624 Transforming growth factor-␣ (TGF-␣), 181, 560, 565t actions of, 565t expression after axotomy, 181, 694 expression by dorsal root ganglion neurons, 181 receptor for, 181 Transforming growth factor-␤ (TGF-␤), 181 effect on myelination, 362 expression by dorsal root ganglion neurons, 181 in experimental autoimmune neuritis, 618 in Schwann cell development and apoptosis, 355–356 receptors for, 181 Transgenic animal models, 535–552 mice deficient in porphobilinogen deaminase, 551–552 of amyotrophic lateral sclerosis, 401–402, 1585–1591 of inherited neuropathy, 1561–1579 related to axon, 548–549 axonal transport, 401–402, 1588–1589 repair after injury, 1419 related to connexin 32, 1576–1578 Cx32-deficient mice, 1576–1577 electrophysiology of, 1577 histology of, 1576–1577 random inactivation in female mice, 1577 Cx32-mutant mice, 543–545, 544f–545f, 1577–1578 alterations at molecular level, 1577 functional gap junctions in Cx32-deficient mice, 1577–1578 xenograft model, 1578

Index Transgenic animal models (Continued) related to early growth response gene 2, 1575–1576 Egr2-null mice, 1575 transcriptional activation of late myelin genes, 1575–1576 related to myelin sheath organization and maintenance, 535–548 mice and rats overexpressing PMP22, 543 mice deficient in ␤1 and ␤4 integrins, 546 mice deficient in connexin 32, 543–545, 544f–545f mice deficient in myelin-associated glycoprotein, 546–548, 547f mice deficient in periaxin and dystroglycan, 545–546, 546f mice deficient in PMP22, 542–543 mice heterozygously deficient for P0, 537–540, 539f–541f mice homozygously deficient for P0, 536–537, 537f–538f mice overexpressing P0, 540–542, 542f related to organization of node of Ranvier, 549–550, 551f mice deficient in Caspr, 550 mice deficient in contactin, 550 mice deficient in galactocerebroside and sulfatide, 550 related to P0, 1572–1575 epitope-tagged, toxic gain-of-function model, 1574f, 1574–1575 heterozygous deficiency, 537–540, 539f–541f homozygous deficiency, 536–537, 537f–538f, 1573 overexpression, 540–542, 542f, 1573, 1573f, 1573t related to periaxin deficiency in mice, 545–546, 546f, 1578–1579 behavior of, 1578–1579 electrophysiology of, 1579 histology of, 1579 sensory involvement in, 1579 related to peripheral myelin protein 22, 542–543, 1561–1572 aggregate formation as source of toxicity and phenotype variability, 1572 deficiency in mice, 542–543, 1564–1566 behavioral effects of, 1566 electrophysiology of, 1566 histology of, 1566 intracellular trafficking abnormalities, 1571–1572 overexpression in mice and rats, 543, 1561–1564 behavioral effects of, 1561–1562, 1562f histology of, 1562–1564, 1563f inducible PMP22-overexpressing model, 1564, 1565f muscle pathology due to, 1564 rat model of progesterone therapy for, 1564 point mutations, 1566–1571, 1567t behavioral effects of, 1566–1568 electrophysiology of, 1571 histology of, 1568–1570, 1569f–1571f muscle pathology due to, 1571 Pmp22 expression and, 1571 protein aggregation, 1572 Transitional node, 68–69 Transitional zone (TZ). See Central nervous system-peripheral nervous system transitional zone. Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Translation, 1545 Transperineurial permeation, 652, 653 Transthyretin (TTR) amyloidosis, 809, 1921, 1922–1931 clinical course of, 1924–1925 clinical features of, 1923–1924 central nervous system involvement, 1924 other system involvement, 1924 other TTR point mutations, 1924 TTR amyloidosis Met 30, 1923–1924 diagnosis of, 1928–1929, 1929f differential diagnosis of, 1928 epidemiology of, 1927 genetics of, 1925–1927 pathogenesis of, 1927–1928 pathology of, 1925, 1926f prognosis for, 1931 treatment of, 1929–1931 Transverse facial artery, 1221 Traumatic injury. See also Nerve injury. amebic keratitis due to, 2162 burn neuropathy, 2030–2031 carpal tunnel syndrome due to, 1444 causalgia due to, 1513–1514, 1514f iridoplegia due to, 209 neuralgia after, 1514 of axillary nerve, 1469–1470 of brachial plexus, 2631–2632 in children, 2734 infraclavicular, 1365–1367 supraclavicular, 1353–1359 of cranial nerves eighth, 1268 fifth, 1209 fourth, 1198 seventh, 1232–1233, 1233f, 1237 sixth, 1198 third, 1198 of lateral antebrachial cutaneous nerve, 1473 of long thoracic nerve, 1464 in children, 2727 of lumbosacral plexus, 1375, 1376, 1378–1380 of median nerve, in children, 2726 of musculocutaneous nerve, 1472 of radial nerve, 1475 in children, 2726–2727 of sciatic nerve, in children, 2728–2729 of spinal nerve roots, 1326–1327 of suprascapular nerve, 1467 in children, 2727 of ulnar nerve, 1479 in children, 2726 tarsal tunnel syndrome due to, 1504 Traumatic neuroma, 2600–2601, 2601f Trazodone, for pain, 2642 Treadmilling of actin filaments, 449 TREK1, 187, 188 Trembler mouse, 704, 1566–1572, 1647, 1698 behavioral features of, 1566–1568 Trembler, 1566–1567 Trembler-J, 1567 Trembler-m1H, -m2H, and -m3H, 1567–1568 Trembler-Ncnp, 1567 histology of, 1568–1570 Trembler, 1568 Trembler-J, 1568–1569, 1569f–1570f Trembler-m1H, -m2H, -m3H, 1570 Trembler-Ncnp, 1569–1570, 1571f intracellular trafficking abnormalities in, 1571–1572

Index Trembler mouse (Continued) Pmp22 point mutations in, 1567t protein aggregation in, 1572 as source of toxicity and phenotype variability, 1572 Tremor amiodarone-induced, 2554 in chronic inflammatory demyelinating polyradiculoneuropathy, 2224 in hereditary motor and sensory neuropathies, 1624, 1631, 1636 in trembler mouse, 1567 neuropathic, 1144–1145 Treponema pallidum infection. See Neurosyphilis. Trichinosis, 2154t–2155t, 2164–2165 clinical features of, 2164 etiologic agent in, 2164 life cycle of, 2164 pathology of, 2163f, 2164 peripheral neuropathy in, 2164–2165 transmission of, 2164 Trichlorfon, 2515 Trichloroethylene (TCE), 1211, 2520 Tricyclic antidepressants (TCAs) adverse effects of, 2641 for neuropathic pain, 2640, 2641–2642 in diabetic patients, 1971, 1971t Trigeminal ganglion, 1207 tumors of, 1210–1211, 2600, 2600f Trigeminal nerve, 1207–1215 anatomy and physiology of, 1207–1209 divisions and branches, 1207, 1208f motor root, 1209 sensory root, 1208 trigeminal ganglion, 1207 trigeminal nuclei, 1208–1209 blink reflex to test integrity of, 918–919, 920t, 921t congenital anesthesia of, 1211 course of, 1194f differential diagnosis of lesions of, 1209, 1210t facial numbness due to lesions of, 1209 idiopathic neuropathy of, 1211–1212 in perineurial tumor spread, 1210 in trigeminal neuralgia, 1213–1215, 1214t in trigeminal trophic syndrome, 1212 investigation of lesions of, 1212–1213 microvascular decompression of, 1215 other tumors of, 1210–1211 procedures to anesthetize territory of, 1215 role in taste sensation, 1207 schwannomas of, 1210 territory supplied by, 1207, 1209f toxic lesions of, 1211 traumatic injury of, 1209 Trigeminal neuralgia, 1213–1215 age-associated incidence of, 1213 blink reflex in, 921t causes of, 1214, 1214t clinical features of, 1213 differential diagnosis of, 1214t glossopharyngeal neuralgia and, 1214 hemifacial spasm and, 1213–1214 idiopathic, 1213 in multiple sclerosis, 1214 medical treatment of, 1215 choice of, 1215 drug therapy, 1215 carbamazepine, 1215, 2641, 2642, 2643 lamotrigine, 2643 phenytoin, 2643 valproate, 2644

Trigeminal neuralgia (Continued) microvascular decompression, 1215 procedures to anesthetize nerve territory, 1215 pathology of, 1214 patient support group for, 1215 triggers for, 1213 vascular compression and, 1214 Triglyceride, 1905–1906 Trimethyl tin, pathology induced by, 688, 715, 716f Tri-nerve paralysis, 1475 Trinucleotide repeat expansions, 1547, 1548, 1554 in Friedreich’s ataxia, 1739 neuronal pathology and, 705–706 Triorthocresylphosphate (TOCP), 1185, 2515, 2075 Trisomic rescue, 1548 Trisulfated heparin disaccharide, antibodies to, 2190 TRKA gene, in hereditary sensory and autonomic neuropathy, 1810, 1811t, 1833–1834 Trochlear nerve, 1192–1201 assessing function in third nerve palsy, 1197 blood supply of, 1195 clinical features of fourth nerve palsy, 1195–1197 course of, 1193–1194, 1194f histology of, 1195, 1196f muscles innervated by, 1192 neoplasms affecting, 1199–1200, 1200f nucleus of, 1193 traumatic injury of, 1198 vascular diseases affecting, 1201 Trophic changes, 1151 Tropical ataxic neuropathy (TAN), 2068–2070 cassava and, 2070, 2073 clinical features of, 2069, 2070t epidemiology of, 2070, 2073 etiology of, 2070–2071 nerve biopsy in, 2068–2069 neuropathology of, 2070 pathogenesis of, 2070 Tropical myeloneuropathies (TMNs), 2063–2077 classification of, 2068–2069 tropical ataxic neuropathy, 2068–2070, 2070t tropical spastic paraparesis, 2068, 2071 epidemic outbreaks of, 2063–2067 Cruickshank’s Jamaican myeloneuropathy, 2066 Cuban epidemic neuropathy, 2066 Domingo Mádan’s retrobulbar optic neuritis in Cuba, 2065 optic neuropathy in Tanzania, 2066–2067 Peraita’s Spanish Civil War neuropathy, 2065 Scott’s Spanish Town disease in Jamaica, 2065 Strachan’s multiple neuritis of West Indies, 2064–2065 epidemiology of, 2063, 2076 etiologies of, 2069t, 2071–2076 HTLV-1 infection, 2076 neurotoxins, 2073–2076 cassava (cyanide), 2073–2075 lathyrism, 2075 pesticides, 2075–2076 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxxvii Tropical myeloneuropathies (TMNs) (Continued) nutritional deficiencies, 2071–2073 tropical malabsorption, 2071 tropical malabsorption-malnutrition, 2071–2072 vitamin deficiencies, 2072 history of, 2063–2068 in soldiers and POWs during World War II, 2063, 2067–2068 clinical features of, 2068 etiology of, 2068 neuropathology of, 2068 tropical malabsorption and malnutrition, 2068 Tropical spastic paraparesis (TSP), 714, 2063, 2066, 2068, 2070–2071 cassava and, 2073 clinical features of, 2070t, 2071 epidemiology of, 2071 etiology of, 2071 neuropathy in, 2076 pseudo–amyotropic lateral sclerosis form of, 2076 vs. amyotrophic lateral sclerosis, 1536 Troyer’s syndrome, 1743 TRPA1, 187, 187t TRPM8, 187, 187t TRPV1, 186–187, 187t TRPV2, 187, 187t TRPV3, 187, 187t TRPV4, 187, 187t, 188 Truncal radiculoneuropathy, diabetic, 1961–1962 clinical features of, 1961, 1962f differential diagnosis of, 1962 distribution of sensory loss in, 1961, 1961f investigation of, 1961–1962 management of, 1962 skin biopsy in, 885 Trunk(s) lumbosacral, 26 of brachial plexus, 19 sympathetic, 15–16, 17f abdominal, 30 cervical, 23 lumbar, 26, 30 thoracic, 23, 25 Trypanosoma cruzi infection. See American trypanosomiasis. Tryptophan neuropathy, 2570–2571 TSP. See Tropical spastic paraparesis (TSP). TSS (Total Symptom Score), 1035, 1035t TST (thermoregulatory sweat test) in distal small fiber neuropathy, 1104 in preganglionic vs. postganglionic lesions, 1108, 1108t TTP1 gene mutations, 2724 TTR amyloidosis. See Transthyretin (TTR) amyloidosis. TTS. See Tarsal tunnel syndrome (TTS). Tubulin, 36 antibodies to ␤-tubulin, 2190 expression after axotomy, 690 Tuffstone bodies, in metachromatic leukodystrophy, 1850, 1851, 1852f–1853f Tumor necrosis factor-␣ (TNF-␣), 181, 560, 565t, 2342, 2344 actions of, 565t effects of intravenous immunoglobulin on, 643 expression in dorsal root ganglion neurons, 181

lxxxviii Tumor necrosis factor-␣ (TNF-␣) (Continued) glucocorticoid effects on, 636 in experimental autoimmune neuritis, 617–618, 625 in HIV-associated distal symmetrical polyneuropathy, 2133 in inflammatory and neuropathic pain, 181 in ischemia-reperfusion injury, 522–523, 525f, 675, 677 in neuropathic pain, 1420–1421 in POEMS syndrome, 2454–2455 in wallerian degeneration, 1415, 1417f receptors for, 181 upregulation after nerve injury, 181 Tumor-like conditions of peripheral nerve, 2585, 2600–2602 classification of, 2586t lipofibromatous hamartoma, 2602, 2602f localized interdigital neuritis, 1507, 2601 neuromuscular choristoma, 2602 palisaded encapsulated neuroma, 2602 traumatic neuroma, 2600–2601, 2601f Tumors of peripheral nerve, 810, 2585–2600. See also specific tumors. affecting brachial plexus, 1360–1362, 1362t, 1367 in children, 2735, 2736f Pancoast’s syndrome, 1361–1362 primary neoplasms, 1360–1361 secondary neoplasms, 1361 vs. radiation-induced lesions, 1361, 1362t affecting lumbosacral plexus, 1381–1382, 1382–1384, 2736, 2738 affecting ocular nerves, 1199–1200, 1200f affecting spinal nerve roots, 1329–1331 affecting trigeminal ganglion, 1210–1211, 2600, 2600f benign, 2586–2596 ganglioneuroma, 2595 granular cell tumor, 2595 nerve sheath myxoma and neurothekeoma, 2594–2595 neurofibroma, 2591f–2592f, 2591–2593 non-neurogenic tumors, 2596 perineurioma, 2593–2594, 2594f schwannoma and variants, 2586–2591, 2587f–2590f carpal tunnel syndrome due to, 1447 classification of, 2585, 2586t differentiation of, 2585 leptomeningeal metastasis from, 1329–1330 leukemia, 2495–2497, 2600 lymphoma, 2489–2495, 2600 malignant, 2596–2600 malignant peripheral nerve sheath tumor and variants, 2596–2599, 2596f–2599f metastatic tumors, 2600, 2600f primitive neuroectodermal tumor, 2600 polyneuropathy and, 1168–1169 suprascapular neuropathy from, 1467 syndromic, 2585 vasculitic neuropathy associated with, 2351–2352 treatment of, 2374, 2375t, 2377t Tuning fork test of vibration sensation, 1038 Two-point discrimination, 1079 Tympanic membrane, 1224f Tympanometry, 1256 Type C retrovirus, 714 Tyramine for orthostatic hypotension, 2678 Horner’s syndrome and, 208

Index Tyrosine hydroxylase in bladder, 301, 305 in familial dysautonomia, 1833 in sweat gland innervation, 1106 Tyrosine kinase(s) activity after axotomy, 693 phosphorylation of ephrins by, 455 Tyrosine receptor kinase(s) (Trk), 173–178, 176t cutaneous, muscle, and visceral expression of, 176t expression on dorsal root ganglion neurons, 174 family of, 174 ligand binding of, 174 TrkA, 174–175 expression on dorsal root ganglion neurons, 174–175 immunostaining for, 173f in neurotrophin signaling, 378f, 379, 381–382 nerve growth factor binding with, 175–177 TrkB, 177 brain-derived neurotrophic factor binding with, 177–178 neurotropin-4 binding with, 178 TrkC, 178 neurotrophin-3 binding with, 178 Tyrosinemia, hereditary, 2713, 2715, 2723 TZ (transitional zone). See Central nervous system–peripheral nervous system transitional zone.

UDP-galactose:ceramide galactosyltransferase (CGT), 423 mice deficient in, 550, 551f U.K. Prospective Diabetes Study (UKPDS), 1953, 1969 Ulcerative colitis, 1167 Ulnar collateral nerve, 22 Ulnar nerve anatomy of, 1340f, 1478–1479 Martin-Gruber anastomosis, 1437 distribution of, 15f, 21, 21f, 1470f, 1478 in leprosy, 2089f, 2091–2092 involvement in vasculitic neuropathy, 2365t muscles innervated by, 956t, 1478–1479 nerve conduction studies of, 907, 907f, 908t, 910, 913f, 913t, 914t collision technique for, 926–928, 928f in all-median hand, 929 with Martin-Gruber anastomosis, 928–929, 929f prognosis for, 1483 treatment of, 1483 Ulnar neuropathy, 1478–1483 clinical manifestations of, 1481 diagnosis of, 1481–1483 etiologies of, 1479–1481 in arm, 1479 in elbow segment, 1479–1480 in forearm, 1480 in wrist and hand, 1480–1481 in children, 2724f, 2726 in diabetes, 1964 in renal failure, 2026 Ulnar tunnel, 1478 Ultrasound endoanal, 290t, 290–291, 291f in carpal tunnel syndrome diagnostic, 1449–1450 therapeutic, 1451 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Umbrella cells of bladder, 300 UNC-5. See Netrin/DCC/UNC-5. Uniparental disomy, 1548 Unmyelinated fibers collagen pockets of, 66 compound action potentials of, 1016 density of, 66–67 age-related changes in, 65 dermal, 871 development of, 64–65, 64f–65f diameter of, 64, 65, 67 historical studies of, 62–64 in Fabry’s disease, 1894, 1896t, 1897f in hereditary motor and sensory neuropathy, 1639–1640, 1640f Déjérine-Sottas neuropathy, 1642 in hereditary sensory and autonomic neuropathy type I, 1818, 1818f–1819f in hereditary sensory and autonomic neuropathy type II, 1829f, 1830 in hereditary sensory and autonomic neuropathy type V, 1835f–1836f, 1836 in sarcoidosis, 2421 microneurography of neuropathic changes in, 1008 microscopic anatomy of, 62–67 morphometry of, 66–67, 66f–67f of sural nerve, 37 organization of, 65 regeneration of, 67 relationship with Schwann cells, 66 response to injury, 1463–1464 visualization of, 64 wallerian degeneration of, 776 Upper limb cutaneous nerve distribution in, 14, 15f in amyotrophic lateral sclerosis, 1536 in multifocal motor neuropathy and conduction block, 2278–2279, 2280f innervation of, 18–23, 956t lower motor neuron syndromes affecting, 1299 mononeuropathies of, 1463–1484. See also Mononeuropathy(ies). in children, 2725–2727 predominance of polyneuropathy in, 1138–1139, 1140f, 1164–1165 Uremic neuropathy, 1137, 2022–2026 compound action potential of sural nerve in, 1023–1024, 1025f hemodialysis for, 124–125, 2022, 2024 historical descriptions of, 2022 in children, 2720 macro electromyography in, 993f mononeuropathies, 1140, 2025–2026 carpal tunnel syndrome, 2025 ischemic forearm neuropathy, 2026 nerve excitability studies in, 124–125 polyneuropathy, 2023–2025 as index of severity of renal failure, 2023 autonomic abnormalities in, 2025 effect of renal transplantation on, 2024–2025 electrophysiology of, 2023f, 2023–2024 etiology of, 2025 nerve biopsy in, 2024 prevalence in hemodialysis patients, 2023 rapidly progressive, 2024 symptoms of, 2023 restless legs and, 1145, 2023 segmental demyelination in, 781, 2024 sensory involvement in, 1141, 2023

Index Ureteric plexus, 281 Urethra, 280f, 300 during voiding response, 301 innervation of, 301–304 parasympathetic pathways, 301–302 sympathetic pathways, 302–304 pressure within, 301 Urethral sphincter external, 304 internal, 300 somatic pathways to, 310 Urinalysis, for monoclonal proteins, 2258, 2258f Urinary bladder. See Bladder. Urinary catecholamines, 1118 Urinary continence, 300, 300f, 2626–2627 detrusor contraction and, 2626 internal and external sphincter control of, 2626 Urinary incontinence, 2626–2627 disorders associated with, 2626–2627 management of, 2627 overflow, 2627 Urinary schistosomiasis, 2160–2161 Urinary tract, lower, 299–314 anatomy of, 299–301 central nervous pathways controlling, 306–310 in brain, 309–310 in spinal cord, 307–309, 309f innervation of, 301–306, 302f afferent pathways, 305 neural modulatory mechanisms, 304 parasympathetic pathways, 301–302, 303f somatic pathways, 304–305 sympathetic pathways, 302–304 urothelial-afferent interactions, 305–306 neurogenic dysfunction of, 313–314 failure to eliminate urine, 313 failure to store urine, 313 spinal cord injury, 313–314 neurotransmitters in micturition reflex pathways, 312 organization of urine storage and voiding reflexes, 300f, 300–301, 310–312 pontine micturition center, 311–312 somatic pathways to urethral sphincter, 310 spinobulbospinal micturition reflex pathway, 310–311, 311f suprapontine control of micturition, 312 sympathetic pathways, 310 receptors for putative transmitters in, 308t reflex mechanisms controlling, 303t, 306, 306f–307f Urinary tract infection, nosocomial, 2616, 2616t Urine-plasma barrier, 300 Uroplakins, 300 Urothelial-afferent interactions, 305–306 Urothelium, 300 Usher’s syndrome, 1259 Uterovaginal plexus, 31, 281, 281f, 31 Utricle, 1253 Utrophin, 424 Uveoparotid fever, 2418, 2420

Vaccines chronic inflammatory demyelinating polyradiculoneuropathy preceded by, 2223–2224 for poliomyelitis, 2713, 2199

VAChT (vesicular acetylcholine transporter), 301, 832 Vacor (pyriminil), 2519 Vacuolar myelopathy, AIDS-related, 1313–1314 Vacuolation, intracytoplasmic, 687–689 axotomy-induced, histologic and ultrastructural features of, 687–688, 688f–689f diseases associated with, 688–689 radiation-induced, 688 toxin-induced, 688 Vagina, 280f Vagus nerve, 17, 1277–1285 anatomy of, 1277–1279 somatic sensory fibers, 1278 visceral afferent fibers, 1278 visceral efferent fibers, 1278–1279 aortic baroreceptor innervation by, 323 blood pressure effects of nerve block, 325, 326f auricular branch of, 1277 cardiac innervation by, 217–220, 220f–221f clinical features of lesions of, 1279 laryngeal paralysis, 1279 palatal and pharyngeal paralysis, 1279 course of, 1277–1278 disorders of autonomic function of, 1280–1284 vagal afferents, 1283–1284 cardiac receptors, 1284 laryngeal receptors, 1283–1284 pulmonary receptors, 1284, 1284f vagal efferents, 1280–1283 abnormal regeneration, 1282–1283, 1283f overactivity, 1281–1282, 1282f–1283f underactivity, 1280–1281, 1281f extramedullary lesions of, 1285 intramedullary lesions of, 1284–1285 pharyngeal branch of, 1277 tests of autonomic function of, 1279–1280 enteroinsular peptide release, 1280 esophageal, gastric, and intestinal motility, 1280 gastric acid secretion, 1280 Valacyclovir, for herpes zoster, 2120 Validity, of health outcome measures, 1057 Valleix phenomenon, 1504 Valproic acid adverse effects of, 2644 for neuropathic pain, 2644 Valsalva maneuver, blood pressure response to, 1111–1112, 1112f Valsalva ratio (VR), 1109–1110, 1110t age and gender effects on, 1115, 1115f–1116f in Composite Autonomic Severity Score, 1115 Vanilloid receptor type 1 (VR1), 872 Vanilloid receptor–like receptor type 1, 872 Varicella zoster virus (VZV) infection, 2118–2120 genomic organization of virus, 2119 herpes zoster, 2119–2120 latent, 2119 neurologic complications of reactivation of, 2119 neuropathies associated with, 2118t neuropathy without rash in, 2120 pathology of, 713 postherpetic neuralgia, 2120 reactivation in lymphoma patients, 2490 zoster sine herpete, 2120 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

lxxxix Variegate porphyria (VP), 1883–1891. See also Porphyric neuropathy. clinical manifestations of, 1886, 1888 neurologic disease, 1888 incidence of, 1887 inheritance of, 1888 laboratory studies in, 1889t, 1889–1890 latent form of, 1890 mechanisms of neuronal damage in, 1885–1886 treatment of, 1890–1891 Vascular cell adhesion molecule-1 (VCAM-1), 2342 in experimental autoimmune neuritis, 616, 625 in vasculitis, 2343 Vascular endothelial growth factor (VEGF) in POEMS syndrome, 2455 in vasculitis, 2344 polymorphisms of, in familial amyotrophic lateral sclerosis, 1534–1535 Vasculature age-related changes in vessels supplying nerve trunks, 492 blood supply of third, fourth, and sixth cranial nerves, 1195 blood supply to nerves, 800–801, 803f causes of ischemic nerve damage, 801–802 peripheral vascular disease, 801 endothelial proliferation in monoclonal gammopathy with neuropathy, 2262–2263 in diabetic neuropathy, 1966, 1968f intravascular lymphomatosis, 2493 microvasculature of peripheral nerve, 667–668 as nutritive capacitance system, 668 large capillaries of, 668 nerve blood flow:blood pressure curve, 667–668 regulation of nerve blood flow, 668–671 splanchnic-mesenteric, evaluation of, 1124–1125, 1125f trigeminal nerve compression by, 1214 Vasculitic lumbosacral plexus neuropathy, 1386 Vasculitic polyradiculopathy, 1334 Vasculitis, 2335–2385. See also specific vasculitides. age distribution of, 2339 angiitis and sarcoidosis, 2417 classification and nomenclature for, 2338t, 2338–2339, 2339f, 2405 definition of, 2335, 2337 diagnostic criteria for, 2338 epidemiology of, 2339 etiology of, 2348–2352 drugs, 2335, 2336, 2340t, 2351 infections, 2335, 2340t, 2348–2351 malignancy, 2351–2352 genetics of, 2363–2364 hypersensitivity, 2354t, 2361–2362 Henoch-Schönlein purpura, 2362 idiopathic, 2335 in connective tissue diseases, 2357–2361 Behçet’s disease, 2361 rheumatoid arthritis, 2357–2358 scleroderma, 2360–2361 Sjögren’s syndrome, 2358–2360, 2360t systemic lupus erythematosus, 2358, 2359t in diabetic lumbosacral radiculoplexus neuropathy, 2004–2007, 2007f–2008f

xc Vasculitis (Continued) in HIV infection, 2137–2138 clinical features of, 2137–2138 disease-stage specificity of, 2129 etiology of, 2138 pathology of, 2138 prevalence of, 2130 treatment of, 2138 incidence of, 2339, 2340t ischemic nerve damage and, 802–803 microvasculitis, 2405–2407 disease association with, 2405–2406 diabetic lumbosacral radiculoplexus neuropathy, 2004–2007, 2007f–2008f, 2405–2406 microvessel pathology in, 2406–2407 neovascularization and, 2412f nerve pathology in, 2406f–2413f, 2407 overview of, 2405 mortality from, 2336, 2337, 2366 nerve biopsy in, 736, 2369, 2369t–2370t nonsystemic vasculitic neuropathy, 2339, 2340t, 2341t, 2355t, 2363, 2363t paraneoplastic, 2374, 2377t, 2493–2494 pathogenesis of, 2335, 2339–2348 adhesion molecules, 2341–2343 chemokines, 2342 cytokines, 2343–2344 humoral and cellular immunity, 2344–2348 anti–endothelial cell antibody– mediated vasculitis, 2345–2346 antineutrophil cytoplasmic antibody–mediated vasculitis, 2346t, 2346–2347, 2347f, 2351, 2352 cellular immunity, 2348 circulating immune complex–mediated vasculitis, 2344–2345, 2345f cytotoxic T lymphocyte–mediated vasculitis, 2348, 2349f delayed-type hypersensitivity, 2348 leukocyte emigration, 2342–2343, 2343f primary vs. secondary, 2338 relapse rates for, 2337 sex distribution of, 2339 systemic, with neuropathy, 2337–2338 classification of, 2339, 2340t clinical features of, 2337, 2353t–2354t, 2364–2365, 2365f diagnosis of, 2336, 2366–2370, 2369t differential diagnosis of, 2366, 2367t electrodiagnostic studies in, 2366–2369 epineurial lesions in, 2370, 2374f evolution of, 2336 historical descriptions of, 2337 laboratory findings in, 2353t–2354t, 2366, 2368t muscle biopsy in, 2336 natural history of, 2366 nerves commonly involved in, 2365, 2365t outcome of, 2383–2385 paraneoplastic, 2374, 2377t, 2478–2479 pathology of, 2336, 2337, 2353t–2354t, 2369–2370, 2370t, 2371f–2373f prevalence of types of, 2341t treatment of, 2370–2383, 2375t–2376t antigen removal, 2373–2376 antiviral therapy, 2376–2378 for paraneoplastic vasculitic neuropathy, 2374, 2377t immunosuppressive therapy, 637, 2378–2381 limiting adverse effects of, 2381–2383, 2382t–2384t

Index Vasculitis (Continued) systemic necrotizing, 2352–2357 Churg-Strauss syndrome, 2355–2356 classic polyarteritis nodosa, 2352–2355 microscopic polyangiitis, 2355 Wegener’s granulomatosis, 2356–2357 temporal arteritis, 2362 treatment of, 2336–2337 Vasoactive intestinal peptide (VIP), 4 downregulation by nerve growth factor, 177 expression after axotomy, 691 expression by dorsal root ganglion neurons, 185 in bladder, 301, 305 in cardiac neural control, 224 in ganglionic neurons, 240, 242 in male sexual response cycle, 314–316 in micturition reflex, 312 in sweat gland innervation, 1106 Vasoconstriction cutaneous responses to body cooling, 333 ␣-receptor subtypes and blood flow in active muscles, 331 regulation of nerve blood flow, 670, 670t sympathetic, 330, 330f Vasoconstrictor nerves, 323, 325 Vasodilation cutaneous responses to body heating, 333 metabolic, 330, 330f neurally mediated, in skeletal muscle, 331–332 ␣-receptor subtypes and blood flow in active muscles, 331 regulation of nerve blood flow, 670t, 670–671 Vasodilator nerves, 323 Vasoglossopharyngeal neuralgia, 1276. See also Glossopharyngeal neuralgia. Vasomotor reflexes in skin, 1122–1123 Vasopressin effect on nerve blood flow, 670, 670t in blood pressure regulation, 2671 Vasopressin analogues, for orthostatic hypotension, 2675 VCAM-1 (vascular cell adhesion molecule-1), 2342 in experimental autoimmune neuritis, 616, 625 in vasculitis, 2343 VCR (vincristine) neuropathy, 2490, 2490t, 2572–2573 effect on axonal transport, 393, 397t microtubule disaggregation in, 393 vs. childhood Guillain-Barré syndrome, 2717 VDT. See Vibration detection threshold (VDT). VEC-DIC (video-enhanced contrast–differential interference contrast microscopy), studies of axonal transport, 389, 395, 395f VEGF (vascular endothelial growth factor), in vasculitis, 2344 VEMPs (vestibular evoked myogenic potentials), 1256 Venlafaxine, for neuropathic pain, 2642 Ventral nerve roots, 11, 1323, 1339. See also Spinal nerve roots. morphometry of, 763–765, 764f–765f Ventral rami, 11, 12, 15–16, 16f, 18, 19, 23, 1339 cervical, 1324f lumbar, 25, 1375 sacral, 28, 1376 thoracic, 23, 24 Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

Ventral spinocerebellar tract, 158 Ventricular fibrillation, 226 Vernet’s syndrome, 1285 Vertebral artery, 1341 Vertebral ganglion, 23 Vertebral plexus, 23 Vertebrobasilar disease, ocular palsy and, 1201 Vertigo in vestibular neur(on)itis, 1262 in vestibular paroxysmia, 1264 Very late antigen-4 (VLA-4), 2342 in experimental autoimmune neuritis, 616, 625 Very-long-chain fatty acids (VLCAs) age-related increase in, 497–498 in adrenoleukodystrophy, 1867–1868 Very-low-density lipoprotein (VLDL), 1906 Vesical plexus, 281, 281f Vesicular acetylcholine transporter (VAChT), 301, 832 Vesiculotubular profiles, axonal, 47, 48f Vestibular compensation, 1263 Vestibular evoked myogenic potentials (VEMPs), 1256 Vestibular hair cells, 1253, 1254f Vestibular nerve, 1253, 1255f. See also Vestibulocochlear nerve. Vestibular neur(on)itis (VN), 1262–1263 diagnosis of, 1262 etiology of, 1262 histopathology and pathophysiology of, 1262–1263, 1263f signs and symptoms of, 1262 Vestibular paroxysmia, 1264 Vestibular schwannoma (VS), 1264–1268 clinical evolution of, 1264t diagnosis of, 1264 differential diagnosis of, 1265–1268, 1266f–1267f genetics of, 1264, 1265t histopathology of, 1264 in neurofibromatosis 2, 1264, 1265f management of, 1264–1265 Vestibular tests, 1256 Vestibulocochlear nerve, 1253–1268 age-related changes of, 1253–1254 anatomy of, 1253 aplasia/hypoplasia of, 1256, 1259, 1259t compression by osseous lesions of internal auditory canal, 1264 course of, 1253, 1255f diagnostic approach to disorders of, 1254–1256 drug toxicity and, 1268 in acquired auditory neuropathy, 1259–1261, 1262f in autoimmune disorders, 1268 in vestibular neur(on)itis, 1262–1263, 1263f inherited disorders of, 1259, 1260f–1261f neurovascular cross-compression of, 1263–1264 noise-induced damage of, 1268 secondary degeneration of, 1268 traumatic injury of, 1268 vestibular schwannoma, 1264t–1265t, 1264–1268, 1265f–1267f Vialetto–van Laere syndrome, 1604t, 1605t, 1617 Vibration detection threshold (VDT), 1034, 1082–1083 as measure of mechanoreceptor function, 1080 in health, 1082, 1084f

Index Vibration detection threshold (VDT) (Continued) quality assurance of, 1082, 1083f reproducibility of, 1091, 1092f Vibration sensation, 1079–1083. See also Quantitative sensation testing (QST). aging and, 486 algorithm of testing and finding threshold for, 1076, 1076f calibration of stimulus magnitude for, 1074f, 1074–1075 CASE III testing of, 1069f CASE IV testing of, 1079–1080, 1080–1082, 1081f–1082f patterns of abnormality on, 1067 vs. clinical assessment, 1066 instructions for 4, 2, 1 step testing algorithm with null stimuli for, 1070 stimulus waveform for, 1073, 1073f subjective differences among anatomic sites, 1067 testing in children, 2711 tuning fork method for testing of, 1038 Video-enhanced contrast–differential interference contrast microscopy (VEC-DIC), studies of axonal transport, 389, 395, 395f Video-oculography (VOG), 1256 Vidian nerve, 1223f Vimentin, 434 Vinblastine neuropathy, 2572 chromatolysis in, 686–687 in patients with CMT1A duplication, 1661 microtubule disaggregation in, 389, 393 Vincristine (VCR) neuropathy, 716–717, 2490, 2490t, 2572–2573 effect on axonal transport, 393, 397t in children, 2720 microtubule disaggregation in, 393 vs. childhood Guillain-Barré syndrome, 2717 VIP. See Vasoactive intestinal peptide (VIP). Viral infections amyotrophic lateral sclerosis and, 1535 cytomegalovirus, 2121–2123 Epstein-Barr virus, 2120–2121 herpes simplex virus, 2117–2118 HIV, 2129–2140 microvasculitis and, 2405 varicella zoster virus, 2118–2120 West Nile virus, 1301–1302 Visceral leishmaniasis, 2165–2167 Vision assessment, 1065 Vitamin A, for abetalipoproteinemia, 2715 Vitamin B1 deficiency, 2051–2053, 2072 causes of, 2051 clinical features of, 2052 diagnosis of, 2052 in children, 2720 management of, 2052–2053 pathogenesis of, 2052 Vitamin B6 deficiency of, 2056 neuropathy induced by, 717, 2056–2057, 2566–2567 Vitamin B12 deficiency, 1307–1308, 2053–2055 animal models of, 1308 causes of, 1307, 2053 diagnosis of, 2054–2055 management of, 2055 myelopathy due to, 2054 neuropathy due to, 1139–1140, 1307–1308, 2053–2055 clinical features of, 1307, 2053–2054 electrodiagnostic testing in, 1308, 2055

Vitamin B12 deficiency (Continued) in children, 2720 quantitative sensation testing in, 2055 sensory involvement in, 1141 upper limb involvement in, 1138, 1164 vs. nonmalignant inflammatory sensory polyganglionopathy, 2316 pathogenesis of, 2053 spinal cord pathology in, 1308 subclinical, 1308 tropical myeloneuropathy due to, 2073 Vitamin B12 measurement, 1172 Vitamin E deficiency of, 1738, 2055–2056 axon dystrophy and, 497, 2056 cholestatic hepatic disease and, 2021 clinical features of, 2056 diagnosis of, 2056 effect on axonal transport, 397t in children, 2720 ataxia due to TTP1 mutations, 2724 management of, 2056 pathogenesis of, 2055–2056 tropical myeloneuropathy due to, 2072 for abetalipoproteinemia, 2715 for experimental diabetic neuropathy, 512, 516, 516t in diabetic patients, 511 lipofuscin accumulation and, 703 to prevent ischemic-reperfusion oxidative injury, 524 Vitamin K, for abetalipoproteinemia, 2715 VLA-4 (very late antigen-4), 2342 in experimental autoimmune neuritis, 616, 625 VLCAs (very-long-chain fatty acids) age-related increase in, 497–498 in adrenoleukodystrophy, 1867–1868 VLDL (very-low-density lipoprotein), 1906 VN. See Vestibular neur(on)itis (VN). Vocal cord paralysis, 1285 in hereditary motor and sensory neuropathy IIC, 1629, 1733 VOG (video-oculography), 1256 Voiding. See Micturition. Volkmann’s contracture, 1517 Volume expansion, for orthostatic hypotension, 2658, 2659, 2674–2675, 2675t von Frey’s hairs, 1038 von Recklinghausen’s disease. See Neurofibromatosis (NF), type 1 (NF1). VP. See Variegate porphyria (VP). VP-16 (etoposide) for vasculitic neuropathy, 2381 peripheral neuropathy induced by, 2490, 2490t, 2498t VR (Valsalva ratio), 1109–1110, 1110t age and gender effects on, 1115, 1115f–1116f in Composite Autonomic Severity Score, 1115 VR1 (vanilloid receptor type 1), 872 VS. See Vestibular schwannoma (VS). VZV. See Varicella zoster virus (VZV) infection.

“Waiter’s tip” position, 2731, 2732f Waldenström’s macroglobulinemia, 2259, 2271, 2433 Wallenberg’s syndrome, 921t Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

xci Wallerian degeneration, 4, 387, 772–776, 780f–781f, 819, 1405–1406 blood-nerve exchange in, 653 causes of, 776, 819 cytokine expression in, 1413–1415, 1416f–1417f endoneurial homeostasis in, 658–659, 658f–659f in axonal forms of Guillain-Barré syndrome, 2204 in IgM monoclonal gammopathy with neuropathy, 2262, 2263 in P0-deficient mice, 537 in paraneoplastic subacute sensory neuronopathy, 2476 in sarcoidosis, 2421 in thalidomide neuropathy, 2570 macrophages in, 1415–1417 matrix metalloproteinases and chondroitin sulfate proteoglycans in, 1418–1419, 1418f–1419f nerve conduction studies in, 902, 903f of myelinated fibers, 773–776, 779f–781f of unmyelinated fibers, 776 Schwann cell basement membrane in, 795f Warfarin, electromyography in patient receiving, 938 Warm detection threshold, 1086 reproducibility of test results, 1091, 1092f Warmth detection fibers, 1005 Water drinking, for orthostatic hypotension, 2658, 2659, 2662, 2674 Weber’s syndrome, 1201 Wegener’s granulomatosis (WG), 2337, 2353t, 2356–2357 antineutrophil cytoplasmic antibodies in, 2352, 2356 classification and definition of, 2338t, 2340t cytokines in, 2344 diagnosis of, 2357 diagnostic criteria for, 2338 facial nerve palsy in, 1232 genetics of, 2363 incidence of, 2340t pathology of, 803 pediatric polyneuropathy and, 2719 phases of, 2356 relative prevalence of, 2341t Weight loss or gain, 1167 Weight loss surgery, polyneuropathy after, 2057 Welander’s myopathy, 1152, 1175 Werdnig-Hoffmann disease (WHD), 704, 1603, 1604t, 1606–1607, 1607t aberrant neurofilament phosphorylation and glycosylation in, 690, 705 arrested form of, 1607t, 1607–1610, 1610f chromatolysis in, 686, 705 clinical course of, 1607 clinical features of, 1606–1607, 1608f differential diagnosis of, 1607 electromyography in, 1605 incidence of, 1606 life expectancy in, 1607 prenatal diagnosis of, 1607 recurrence risk to siblings in, 1607 West Nile virus infection, 1301–1302 animal models of, 1302 cerebrospinal fluid analysis in, 1301 clinical features of, 1301 diagnosis of, 1302 electrophysiologic testing in, 1301

xcii West Nile virus infection (Continued) epidemiology of, 1301 lower motor neuron disease in, 1299, 1301–1302 mortality from, 1302 neuroimaging in, 1302 pathogenesis of weakness in, 1302 pathology of, 1302 transmission of, 1301 treatment of, 1302 vs. Guillain-Barré syndrome, 2201 vs. immune brachial plexus neuropathy, 2304 Western blot, 2180 WG. See Wegener’s granulomatosis (WG). WHD. See Werdnig-Hoffmann disease (WHD). Wheelchairs, 2624–2625 WHO (wrist-hand orthosis), 2623 Wnt signaling, in schizophrenia, 467 Wobbler mouse, 711, 1585 Wolman’s disease, 1846, 1864, 1909, 1910t Wrist-hand orthosis (WHO), 2623

Xanthomas, in primary biliary cirrhosis– associated sensory polyneuropathy, 2021 Xanthomatosis, cerebrotendinous, 2722 Xeroderma pigmentosum, 705, 1741 in children, 2722

Index X-linked adrenoleukodystrophy, 2722 X-linked bulbospinal muscular atrophy, 705, 706, 1299, 1316, 1599–1600, 1605t, 1612 clinical features of, 1612, 1612f differential diagnosis of, 1612 amyotrophic lateral sclerosis, 1536 electrophysiologic findings in, 1316, 1612 genetics of, 1316, 1536, 1548, 1600, 1612 pathology of, 1612 pattern of involvement in, 1139 sensory involvement in, 1316 X-linked Charcot-Marie-Tooth disease, 1626, 1791–1800 animal models of, 1800 brainstem auditory evoked responses in, 1797 central nervous system involvement in, 1797 discovery of, 1791 genotype-phenotype correlations in, 1799–1800 GJB1 mutations causing, 543, 704, 1259, 1628–1629, 1791–1796, 1792f medical issues in, 1800 nerve conduction studies in, 1797–1798, 1798f neuromuscular manifestations of, 1796f, 1796–1797 pathology of, 1798–1799, 1799f X-linked dominant inheritance, 1548–1549

Volume 1 pp 1–1190 • Volume 2 pp 1191–2754

X-linked hereditary motor and sensory neuropathy gender differences in severity of, 1629 genetics of, 1627t, 1628–1629 nonuniform demyelination in, 1167 X-linked recessive inheritance, 1548

Y-linked inheritance, 1549 Yohimbine for erectile dysfunction, 2664 for orthostatic hypotension, 2660t, 2661, 2677

Zalcitabine (ddC), 2130, 2131, 2133–2135, 2134f–2135f, 2563 Zebra bodies, 709 in metachromatic leukodystrophy, 1850, 1851, 1851f in Sanfilippo syndrome, 1864 in Schwann cells, 771, 774f–775f Zidovudine (AZT), 2130 Zinc pyridinethione (ZPT), effect on axonal transport, 397t, 399–400 Zonisamide, 2643 Zonula occludens, perineurial, 40f Zoster sine herpete, 2120 ZPT (zinc pyridinethione), effect on axonal transport, 397t, 399–400 Zygomatic nerve, 1220